Cyclone V Device Handbook Volume 3 Hard Processor System Technical Reference Manual PDF

Cyclone V Device Handbook Volume 3: Hard Processor
System Technical Reference Manual
Cyclone V Device Handbook

Volume 3: Hard Processor System Technical Reference
Manual
101 Innovation Drive

San Jose, CA 95134
www.altera.com
cv_5v4-1.3
© 2012 Altera Corporation. All rights reserved. ALTERA, ARRIA, CYCLONE, HARDCOPY, MAX, MEGACORE, NIOS, QUARTUS and STRATIX words and logos
are trademarks of Altera Corporation and registered in the U.S. Patent and Trademark Office and in other countries. All other words and logos identified as
trademarks or service marks are the property of their respective holders as described at www.altera.com/common/legal.html. Altera warrants performance of its ISO
semiconductor products to current specifications in accordance with Altera's standard warranty, but reserves the right to make changes to any products and 9001:2008
services at any time without notice. Altera assumes no responsibility or liability arising out of the application or use of any information, product, or service Registered
described herein except as expressly agreed to in writing by Altera. Altera customers are advised to obtain the latest version of device specifications before relying
on any published information and before placing orders for products or services.
Portions © 2011 ARM Limited. Used with permission. All rights reserved. ARM, the ARM Powered logo, AMBA, Jazelle, StrongARM, Thumb, and TrustZone are
registered trademarks of ARM Limited. The ARM logo, Angel, ARMulator, AHB, APB, ASB, ATB, AXI, CoreSight, Cortex, EmbeddedICE, ModelGen, MPCore,
Multi-ICE, NEON, PrimeCell, ARM7TDMI, ARM7TDMI-S, ARM9TDMI, ARM9E-S, ARM966E-S, ETM7, ETM9, TDMI and STRONG are trademarks of ARM
Limited. All other products or services mentioned herein may be trademarks of their respective owners. Neither the whole nor any part of the information
contained in, or the product described in, this document may be adapted or reproduced in any material form except with the prior written permission of the
copyright holder. The product described in this document is subject to continuous developments and improvements. All particulars of the product and its use
contained in this document are given by ARM in good faith. However, all warranties implied or expressed, including but not limited to implied warranties of
merchantability, or fitness for purpose, are excluded. This document is intended only to assist the reader in the use of the product. ARM Limited shall not be liable
for any loss or damage arising from the use of any information in this document, or any error or omission in such information, or any incorrect use of the product.
Where the term ARM is used it means “ARM or any of its subsidiaries as appropriate”. This document is Non-Confidential. The right to use, copy and disclose this
document may be subject to license restrictions in accordance with the terms of the agreement entered into by ARM and the party that ARM delivered this
document to. The information in this document is final, that is for a developed product.
Portions © 2011 Synopsys, Inc. Used with permission. All rights reserved. Synopsys & DesignWare are registered trademarks of Synopsys, Inc. All documentation
is provided "as is" and without any warranty. Synopsys expressly disclaims any and all warranties, express, implied, or otherwise, including the implied warranties
of merchantability, fitness for a particular purpose, and non-infringement, and any warranties arising out of a course of dealing or usage of trade.
Portions © 2011 Cadence Design Systems, Inc. Used with permission. All rights reserved worldwide. Cadence and the Cadence logo are registered trademarks of
Cadence Design Systems, Inc. All others are the property of their respective holders.
Portions © 2011 Robert Bosch GmbH. Used with permission. All rights reserved.
November 2012 Altera Corporation Cyclone V Device Handbook

Volume 3: Hard Processor System Technical Reference Manual
Contents
Chapter Revision Dates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiii
Section I. Overview
Chapter 1. Introduction to the Hard Processor System

Features of the HPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–2
HPS Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–4
MPU Subsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–5
Interconnect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–5
Memory Controllers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–6
SDRAM Controller Subsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–6
NAND Flash Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–7
Quad SPI Flash Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–7
SD/MMC Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–8
Support Peripherals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–8
Clock Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–8
Reset Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–8
System Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–8
Scan Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–9
Timers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–9
Watchdog Timers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–9
DMA Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–9
FPGA Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–10
Interface Peripherals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–10
EMACs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–10
USB Controllers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–10
I2C Controllers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–11
UARTs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–11
CAN Controllers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–11
SPI Master Controllers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–12
SPI Slave Controllers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–12
GPIO Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–12
On-Chip Memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–13
On-Chip RAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–13
Boot ROM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–13
Endian Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–13
HPS-FPGA Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–13
Address Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–14
Address Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–15
SDRAM Address Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–16
MPU Address Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–16
L3 Address Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–17
Peripheral Region Address Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–18
Document Revision History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1–19
Chapter 2. Clock Manager

Features of the Clock Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–1
Clock Manager Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–3

iv Contents
Functional Description of the Clock Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–3

Clock Manager Building Blocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–3
PLLs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–3
Dividers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–5
Clock Gating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–5
Control and Status Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–5
Hardware-Managed and Software-Managed Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–5
Clock Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–6
OSC1 Clock Group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–6
Main Clock Group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–6
Peripheral Clock Group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–9
SDRAM Clock Group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–11
Flash Controller Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–12
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–13
Cold Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–13
Warm Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–13
Safe Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–13
Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–14
Clock Usage By Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–14
Clock Manager Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–16
Chapter 3. Reset Manager

Reset Manager Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–2
HPS External Reset Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–3
Reset Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–3
Module Reset Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–5
Slave Interface and Status Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–7
Functional Description of the Reset Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–8
Reset Sequencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–8
Cold Reset Assertion Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–10
Warm Reset Assertion Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–10
Cold and Warm Reset Deassertion Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–11
Reset Pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–11
Reset Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–12
Altering Warm Reset System Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–12
Reset Handshaking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–13
Reset Manager Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–13
Section II. System Interconnect
Chapter 4. Interconnect
Features of the Interconnect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–1
Interconnect Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–2
L3 Masters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–3
L3 Slaves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–4
L4 Slaves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–4
Functional Description of the Interconnect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–5
Master-to-Slave Connectivity Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–6
Address Remapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–6
Master Caching and Buffering Overrides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–8
Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–9
Cyclone V Device Handbook November 2012 Altera Corporation

Contents v
Slave Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–9

Master Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–9
Arbitration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–9
Cyclic Dependency Avoidance Schemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–9
Single Slave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–10
Single Slave Per ID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–10
Single Active Slave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–10
Interconnect Master Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–10
Interconnect Slave Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–11
Upsizing Data Width Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–13
Incrementing Bursts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–13
Wrapping Bursts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–13
Fixed Bursts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–13
Bypass Merge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–14
Downsizing Data Width Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–14
Incrementing Bursts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–14
Wrapping Bursts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–14
Fixed Bursts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–14
Bypass Merge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–14
Lock Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–14
FIFO Buffers and Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–15
Data Release Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–15
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–15
Interconnect Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–15
Chapter 5. HPS-FPGA AXI Bridges

Features of the AXI Bridges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–1
AXI Bridges Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–2
Functional Description of the AXI Bridges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–3
The Global Programmers View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–3
FPGA-to-HPS Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–3
FPGA-to-HPS Access to ACP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–4
FPGA-to-HPS Bridge Slave Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–4
HPS-to-FPGA Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–6
HPS-to-FPGA Bridge Master Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–7
Lightweight HPS-to-FPGA Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–9
Lightweight HPS-to-FPGA Bridge Master Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–10
Clocks and Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–12
FPGA-to-HPS Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–12
HPS-to-FPGA Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–13
Lightweight HPS-to-FPGA Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–13
GPV Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–13
Data Width Sizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–14
HPS-FPGA AXI Bridges Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5–14
Section III. Cortex-A9 Microprocessor
Chapter 6. Cortex-A9 Microprocessor Unit Subsystem

Features of the Cortex-A9 MPU Subsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–1
Cortex-A9 MPU Subsystem Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–2
Cortex-A9 MPU Subsystem Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–3

vi Contents
Cortex-A9 MPCore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–4

Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–4
Implementation Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–5
Cortex-A9 Processor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–5
Interactive Debugging Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–6
L1 Caches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–6
Preload Engine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–7
Floating Point Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–7
NEON Multimedia Processing Engine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–8
Memory Management Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–9
Performance Monitoring Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–11
MPCore Timers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–11
Generic Interrupt Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–12
Global Timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–17
Snoop Control Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–18
Accelerator Coherency Port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–19
ACP ID Mapper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–20
Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–20
Implementation Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–21
L2 Cache . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–24
Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–24
ECC Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–25
Implementation Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–26
Debugging Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–28
Program Trace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–28
Event Trace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–29
Cross-Triggering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–29
Cortex-A9 MPU Subsystem Register Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–29
Section IV. Debug and Trace
Chapter 7. CoreSight Debug and Trace

Features of CoreSight Debug and Trace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–1
ARM CoreSight Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–2
CoreSight Debug and Trace Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–3
Functional Description of CoreSight Debug and Trace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–4
Debug Access Port (DAP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–4
System Trace Macrocell (STM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–4
Trace Funnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–5
Embedded Trace FIFO (ETF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–5
AMBA Trace Bus Replicator (Replicator) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–5
Embedded Trace Router (ETR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–5
Trace Port Interface Unit (TPIU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–6
Embedded Cross Trigger (ECT) System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–6
Cross Trigger Interface (CTI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–7
Cross Trigger Matrix (CTM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–8
Program Trace Macrocell (PTM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–10
HPS Debug APB Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–11
CoreSight Debug and Trace Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–11
ROM Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–11
STM Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–12
CTI Trigger Connections to Outside the Debug System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–13

Contents vii
csCTI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–13
FPGA-CTI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–14
Configuring Embedded Cross-Trigger Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–14
Debug Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–15
Debug Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–16
CoreSight Debug and Trace Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . 7–17
Section V. Memory and Memory Controllers
Chapter 8. SDRAM Controller Subsystem

Features of the SDRAM Controller Subsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–1
SDRAM Controller Subsystem Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . 8–2
SDRAM Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–2
DDR PHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–2
SDRAM Controller Subsystem Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–3
MPU Subsystem Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–3
L3 Interconnect Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–3
CSR Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–3
FPGA-to-HPS SDRAM Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–3
Memory Controller Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–4
MPFE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–5
Command Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–5
Write Data Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–6
Read Data Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–6
Single-Port Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–6
Command Generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–6
Timer Bank Pool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–6
Arbiter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–7
Rank Timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–7
Write Data Buffer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–7
ECC Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–7
AFI Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–7
CSR Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–7
Functional Description of the SDRAM Controller Subsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–7
MPFE Operational Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–7
Operation Ordering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–7
Multiport Scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–8
SDRAM Burst Scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–9
Clocking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–10
Single-Port Controller Operational Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–10
SDRAM Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–10
ECC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–11
Interleaving Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–12
AXI-Exclusive Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–14
Memory Protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–14
SDRAM Power Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–16
DDR PHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–17
Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–17
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–18
Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–18
Protocol Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–18
SDRAM Controller Subsystem Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–21

viii Contents
Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–21
Timing Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–21
SDRAM Controller Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8–21
Chapter 9. On-Chip Memory

On-Chip RAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9–1
Features of the On-Chip RAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9–1
On-Chip RAM Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9–2
Functional Description of the On-Chip RAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9–2
Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9–2
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9–2
Boot ROM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9–3
Features of the Boot ROM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9–3
Boot ROM Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9–3
Functional Description of the Boot ROM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9–3
Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9–4
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9–4
On-Chip Memory Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9–4
Document Revision History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9–4
Chapter 10. NAND Flash Controller

NAND Flash Controller Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–1
NAND Flash Controller Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–2
Functional Description of the NAND Flash Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–2
Discovery and Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–2
Bootstrap Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–3
Configuration by Host . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–4
Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–5
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–5
Indexed Addressing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–6
Command Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–6
MAP00 Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–7
MAP01 Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–8
MAP10 Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–9
MAP11 Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–10
Data DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–11
Multitransaction DMA Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–13
Burst DMA Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–15
ECC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–15
Main Area Transfer Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–16
Spare Area Transfer Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–16
Main+Spare Area Transfer Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–17
Preserving Bad Block Markers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–17
Error Correction Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–18
Interface Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–19
NAND Flash Controller Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–19
Basic Flash Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–19
NAND Flash Controller Optimization Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–19
Device Initialization Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–20
Device Operation Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–21
ECC Enabling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–21
NAND Flash Controller Performance Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–22

Contents ix
Interrupt and DMA Enabling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–22

Timing Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–23
Registers to Ignore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–24
Flash-Related Special Function Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–24
Erase Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–24
Lock Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–25
Transfer Mode Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–25
Read-Modify-Write Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–27
Copy-back Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–28
Pipeline Read-Ahead and Write-Ahead Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–29
NAND Flash Controller Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–32
Document Revision History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–32
Chapter 11. SD/MMC Controller

Features of the SD/MMC Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–1
SD/MMC Controller Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–3
Functional Description of the SD/MMC Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–4
SD/MMC/CE-ATA Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–4
BIU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–6
Slave Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–6
Register Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–6
Interrupt Controller Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–7
FIFO Buffer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–8
Internal DMA Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–8
Host Bus Burst Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–13
Host Data Buffer Alignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–13
Buffer Size Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–13
Internal DMA Controller Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–13
Internal DMA Controller FSM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–14
Abort During Internal DMA Transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–15
FIFO Buffer Overflow and Underflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–15
Fatal Bus Error Scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–16
CIU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–16
Command Path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–17
Data Path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–22
Clock Control Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–31
Error Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–32
Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–33
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–34
SD/MMC Controller Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–35
Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–35
Enumerated Card Stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–37
Clock Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–40
Controller/DMA/FIFO Buffer Reset Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–41
Enabling FIFO Buffer ECC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–42
Non-Data Transfer Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–42
Data Transfer Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–44
Single-Block or Multiple-Block Read . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–46
Single-Block or Multiple-Block Write . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–48
Stream Read and Write . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–50
Packed Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–50
Transfer Stop and Abort Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–51
STOP_TRANSMISSION (CMD12) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–51

x Contents
ABORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–51
Internal DMA Controller Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–52
Internal DMA Controller Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–52
Internal DMA Controller Transmission Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–53
Internal DMA Controller Reception Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–54
Commands for SDIO Card Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–54
Suspend and Resume Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–54
Read-Wait Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–57
CE-ATA Data Transfer Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–57
Reset and Card Device Discovery Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–57
ATA Task File Transfer Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–57
ATA Task File Transfer Using the RW_MULTIPLE_REGISTER (RW_REG) Command . . . . . 11–58
ATA Payload Transfer Using the RW_MULTIPLE_BLOCK (RW_BLK) Command . . . . . . . . 11–59
CE-ATA CCS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–60
Reduced ATA Command Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–62
Card Read Threshold . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–64
Recommended Usage Guidelines for Card Read Threshold . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–64
Card Read Threshold Programming Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–64
Card Read Threshold Programming Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–65
Interrupt and Error Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–66
Booting Operation for eMMC and MMC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–67
Boot Operation by Holding Down the CMD Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–67
Boot Operation for eMMC Card Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–68
Boot Operation for Removable MMC4.3, MMC4.4 and MMC4.41 Cards . . . . . . . . . . . . . . . . . . 11–72
Alternative Boot Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–73
Alternative Boot Operation for eMMC Card Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–74
Alternative Boot Operation for MMC4.3 Cards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–78
SD/MMC Controller Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–79
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11–79
Chapter 12. Quad SPI Flash Controller

Features of the Quad SPI Flash Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–1
Quad SPI Flash Controller Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–2
Functional Description of the Quad SPI Flash Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–3
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–3
Data Slave Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–3
Register Slave Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–3
Direct Access Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–4
Indirect Access Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–4
Indirect Read Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–5
Indirect Write Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–6
Consecutive Reads and Writes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–7
Local Memory Buffer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–7
DMA Peripheral Request Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–8
STIG Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–9
SPI Legacy Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–10
Configuring the Flash Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–10
XIP Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–11
Write Protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–12
Data Slave Sequential Access Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–12
Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–12
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–13
Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–13

Contents xi

Quad SPI Flash Controller Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–14
Setting Up the Quad SPI Flash Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–14
Indirect Read Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–15
Indirect Write Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–16
XIP Mode Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–17
Entering XIP Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–17
Exiting XIP Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–18
XIP Mode at Power on Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–19
Quad SPI Flash Controller Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–19
Section VI. Peripherals
Chapter 13. FPGA Manager

Features of the FPGA Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13–1
FPGA Manager Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13–2
Functional Description of the FPGA Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13–3
FPGA Manager Building Blocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13–3
Fabric I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13–3
Monitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13–3
FPGA Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13–4
Power Up Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13–5
Reset Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13–5
Configuration Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13–6
Initialization Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13–6
User Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13–7
Clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13–7
Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13–7
FPGA Manager Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13–7
Chapter 14. System Manager

Features of the System Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–1
System Manager Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–2
Functional Description of the System Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–3
Boot Configuration and System Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–4
Additional Module Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–4
Scan Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–4
DMA Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–5
NAND Flash Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–5
EMAC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–6
USB 2.0 OTG Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–6
SD/MMC Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–7
Watchdog Timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–7
Boot ROM Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–7
Freeze Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–9
FPGA Interface Enables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–10
ECC and Parity Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–10
Pin Multiplexing Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–11
Preloader Handoff Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–11
Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–11
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–12

xii Contents
System Manager Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–12

Chapter 15. Scan Manager

Features of the Scan Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15–1
Scan Manager Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15–2
Functional Description of the Scan Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15–4
Configuring HPS I/O Scan Chains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15–4
Communicating with the JTAG TAP Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15–5
JTAG-AP FIFO Buffer Access and Byte Command Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15–5
Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15–6
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15–6
Scan Manager Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15–6
Chapter 16. DMA Controller

Features of the DMA Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–1
DMA Controller Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–3
Functional Description of the DMA Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–3
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–4
Operating States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–4
Stopped . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–6
Executing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–6
Cache Miss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–7
Updating PC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–7
Waiting For Event . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–7
At Barrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–7
Waiting For Peripheral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–7
Faulting Completing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–7
Faulting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–7
Killing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–7
Completing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–7
Initializing the DMAC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–8
How to Set the Security State of the DMA Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–8
How to Set the Security State for the Interrupt Outputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–8
How to Set the Security State for a Peripheral Request Interface . . . . . . . . . . . . . . . . . . . . . . . . . . 16–8
Using the Slave Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–9
Issuing Instructions to the DMAC using a Slave Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–9
Peripheral Request Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–10
Request Acceptance Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–12
Peripheral Length Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–12
DMAC Length Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–13
Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–14
Burst Only Request . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–14
No Flush Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–14
No Acknowledge Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–14
Using Events and Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–15
Using an Event to Restart DMA Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–15
Interrupting the MPU Subsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–16
Aborts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–16
Abort Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–16
Abort Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–16
Watchdog Abort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–17

Contents xiii
Abort Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–18

Security Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–20
DMA Manager Thread in Secure State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–20
DMA Manager Thread in Non-Secure State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–21
DMA Channel Thread in Secure State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–21
DMA Channel Thread in Non-Secure State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–22
Constraints and Limitations of Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–24
DMA Channel Arbitration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–24
DMA Channel Prioritization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–24
Instruction Cache Latency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–24
AXI Data Transfer Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–24
AXI Bursts Crossing 4 KB Boundaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–24
AXI Burst Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–24
AXI Write Addresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–24
AXI Write Data Interleaving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–24
Programming Restrictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–25
Fixed Unaligned Bursts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–25
Endian Swap Size Restrictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–25
Updating DMA Channel Control Registers During a DMA Cycle . . . . . . . . . . . . . . . . . . . . . . . 16–25
Resource Sharing Between DMA Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–26
DMA Controller Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–27
Instruction Syntax Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–27
Instruction Set Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–28
Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–29
DMAADDH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–29
DMAADNH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–29
DMAEND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–30
DMAFLUSHP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–31
DMAGO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–31
DMAKILL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–32
DMALD[S | B] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–33
DMALDP<S | B> . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–34
DMALP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–35
DMALPEND[S | B] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–36
DMALPFE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–38
DMAMOV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–38
DMANOP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–39
DMARMB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–39
DMASEV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–40
DMAST[S | B] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–40
DMASTP<S | B> . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–41
DMASTZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–42
DMAWFE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–43
DMAWFP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–43
DMAWMB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–44
Assembler Directives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–45
DCD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–45
DCB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–45
DMALP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–45
DMALPFE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–46
DMAMOV CCR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–46
MFIFO Buffer Usage Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–47
About MFIFO Buffer Usage Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–47
Aligned Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–48

xiv Contents
Unaligned Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–50

Fixed Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–54
DMA Controller Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–56
Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–57
Chapter 17. Ethernet Media Access Controller

Features of the Ethernet MAC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–1
MAC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–1
PHY Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–2
DMA Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–2
Management Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–2
Acceleration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–2
Other Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–2
EMAC Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–3
EMAC to RGMII Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–3
PHY Management Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–4
MDIO Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–4
I2C External PHY Management Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–5
IEEE 1588 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–5
Functional Description of the EMAC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–6
Host Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–7
Slave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–7
Master . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–7
External PHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–8
Transmit and Receive Data FIFO Buffers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–8
IEEE 1588-2002 Time Stamps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–8
Reference Timing Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–10
System Time Register Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–10
Transmit Path Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–13
Receive Path Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–13
Timestamp Error Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–13
Frequency Range of Reference Timing Clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–13
IEEE 1588-2008 Advanced Timestamps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–14
Peer-to-Peer PTP Transparent Clock (P2P TC) Message Support . . . . . . . . . . . . . . . . . . . . . . . . 17–14
Clock Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–15
Reference Timing Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–15
Transmit Path Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–15
Receive Path Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–15
Auxiliary Snapshot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–15
IEEE 802.3az Energy Efficient Ethernet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–16
LPI Timers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–16
Checksum Offload . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–17
Frame Filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–17
Source Address or Destination Address Filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–17
VLAN Filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–18
Layer 3 and Layer 4 Filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–19
Clocks and Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–20
Clock Gating for EEE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–20
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–20
Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–20
Ethernet MAC Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–21
DMA Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–21
Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–22

Contents xv
Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–25
Reception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–31
Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–34
Error Response to DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–35
Descriptor Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–35
Descriptor Endianness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–36
Normal Descriptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–36
Transmit Descriptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–36
Receive Descriptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–40
Descriptor Format With IEEE 1588-2005 Timestamps Enabled . . . . . . . . . . . . . . . . . . . . . . . . . . 17–44
Alternate or Enhanced Descriptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–47
Transmit Descriptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–48
Receive Descriptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–54
Initializing DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–60
Initializing MAC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–61
Performing Normal Receive and Transmit Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–62
Stopping and Starting Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–63
Programming Guidelines for Energy Efficient Ethernet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–63
Entering and Exiting the TX LPI Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–63
Gating Off the CSR Clock in the LPI Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–64
Programming Guidelines for Flexible Pulse-Per-Second (PPS) Output . . . . . . . . . . . . . . . . . . . . . . 17–65
Generating Single Pulse on PPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–65
Generating a Pulse Train on PPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–65
Generating an Interrupt without Affecting the PPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–66
Ethernet MAC Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17–67
Chapter 18. USB 2.0 OTG Controller

Features of the USB OTG Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–2
Supported PHYs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–3
USB OTG Controller Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–4
Functional Description of the USB OTG Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–5
USB OTG Controller Block Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–5
Master Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–5
Slave Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–6
Application Interface Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–6
Packet FIFO Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–6
SPRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–6
MAC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–7
Wakeup and Power Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–8
PHY Interface Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–8
ULPI PHY Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–9
Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–9
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–9
Reset Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–9
Hardware Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–10
Software Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–10
Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–11
USB OTG Controller Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–12
Enabling SPRAM ECCs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–12
Host Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–12
Host Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–12
Host Transaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–13
Device Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–14

xvi Contents
Device Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–14

Device Transaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–14
USB OTG Controller Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–15
Chapter 19. SPI Controller

Features of the SPI Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–1
SPI Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–1
SPI Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–2
Functional Description of the SPI Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–3
Protocol Details and Standards Compliance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–3
SPI Controller Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–3
Serial Bit-Rate Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–4
Transmit and Receive FIFO Buffers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–5
SPI Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–6
Transfer Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–7
Transmit and Receive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–7
Transmit Only . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–7
Receive Only . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–7
EEPROM Read . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–7
SPI Master . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–8
RXD Sample Delay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–8
Data Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–9
Master SPI and SSP Serial Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–9
Master Microwire Serial Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–10
SPI Slave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–11
Slave SPI and SSP Serial Transfers † . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–12
Serial Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–13
Partner Connection Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–13
Motorola SPI Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–13
Texas Instruments Synchronous Serial Protocol (SSP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–14
National Semiconductor Microwire Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–15
DMA Controller Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–17
Slave Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–18
Control and Status Register Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–18
Data Register Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–18
Clocks and Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–18
SPI Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–18
Master SPI and SSP Serial Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–19
Master Microwire Serial Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–21
Slave SPI and SSP Serial Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–23
Slave Microwire Serial Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–24
Software Control for Slave Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–24
DMA Controller Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–25
DMA Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–25
Transmit FIFO Buffer Underflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–25
Transmit Watermark Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–25
Transmit FIFO Buffer Overflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–27
Receive FIFO Buffer Overflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–28
Choosing Receive Watermark Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–28
Receive FIFO Buffer Underflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–28
SPI Controller Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–29

Contents xvii
Chapter 20. I2C Controller

Features of the I2C Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–1
I2C Controller Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–2
Functional Description of the I2C Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–3
Feature Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–3
Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–4
START and STOP Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–5
Combined Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–5
Protocol Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–5
START and STOP Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–5
Addressing Slave Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–6
Transmitting and Receiving Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–7
START BYTE Transfer Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–9
Multiple Master Arbitration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–10
Clock Synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–11
Clock Frequency Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–12
Minimum High and Low Counts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–12
SDA Hold Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–13
DMA Controller Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–14
Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–14
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–14
Interface Pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–14
I2C Controller Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–15
Slave Mode Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–15
Initial Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–15
Slave-Transmitter Operation for a Single Byte . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–15
Slave-Receiver Operation for a Single Byte . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–17
Slave-Transfer Operation for Bulk Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–17
Master Mode Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–18
Initial Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–18
Dynamic IC_TAR or IC_10BITADDR_MASTER Update . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–19
Master Transmit and Master Receive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–19
Disabling the I2C Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–19
DMA Controller Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–20
Transmit FIFO Underflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–20
Transmit Watermark Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–20
Transmit FIFO Overflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–22
Receive FIFO Overflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–22
Receive Watermark Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–23
Receive FIFO Underflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–23
I2C Controller Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20–23
Chapter 21. UART Controller

UART Controller Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–1
UART Controller Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–2
Functional Description of the UART Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–3
FIFO Buffer Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–3
Automatic Flow Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–4
Automatic RTS mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–4
Automatic CTS mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–4
Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–5
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–5

xviii Contents
Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–5
Programmable THRE Interrupt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–6
UART Controller Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–8
DMA Controller Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–8
Transmit FIFO Underflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–8
Transmit Watermark Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–8
Transmit FIFO Overflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–10
Receive FIFO Overflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–11
Receive Watermark Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–11
Receive FIFO Underflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–11
UART Controller Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21–12
Chapter 22. General-Purpose I/O Interface

Features of the GPIO Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22–1
GPIO Interface Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22–1
Functional Description of the GPIO Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22–2
Debounce Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22–2
Pin Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22–2
GPIO Interface Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22–2
GPIO Interface Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22–3
Chapter 23. Timer

Features of the Timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23–1
Timer Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23–2
Functional Description of the Timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23–2
Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23–3
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23–3
Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23–3
Timer Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23–4
Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23–4
Enabling or Disabling the Timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23–4
Loading the Timer Countdown Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23–4
Servicing Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23–5
Clearing Interrupt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23–5
Checking Interrupt Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23–5
Masking Interrupt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23–5
Timer Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23–5
Chapter 24. Watchdog Timer

Features of the Watchdog Timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–1
Watchdog Timer Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–2
Functional Description of the Watchdog Timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–2
Counter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–2
Pause Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–3
Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–3
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–4
Watchdog Timer Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–4
Setting the Timeout Period Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–4
Selecting the Output Response Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–4
Enabling and Initially Starting a Watchdog Timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–4

Contents xix
Reloading a Watchdog Counter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–4

Pausing a Watchdog Timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–5
Disabling and Stopping a Watchdog Timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–5
Watchdog Timer State Machine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–5
Watchdog Timer Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24–6
Chapter 25. CAN Controller

Features of the CAN Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–1
CAN Controller Block Diagram and System Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–2
Functional Description of the CAN Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–3
Message Object . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–3
Message Object Control Flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–3
Message Object Mask Bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–5
CAN Message Bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–6
Message Interface Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–7
DMA Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–7
Automatic Retransmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–8
Test Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–8
Silent Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–8
Loopback Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–8
Combined Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–9
L4 Slave Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–9
Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–9
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–10
Software Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–10
Hardware Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–10
Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–10
Error Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–10
Status Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–11
Message Object Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–11
CAN Controller Programming Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–11
Software Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–11
CAN Message Transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–12
Message Object Reconfiguration for Frame Reception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–13
Message Object Reconfiguration for Frame Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–13
CAN Controller Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25–14
Section VII. Hard Processor System User Guide
Chapter 26. Introduction to the HPS Component

Chapter 27. Instantiating the HPS Component

Configuring FPGA Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–1
General Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–2
Boot and Clock Selection Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–3
AXI Bridges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–3
FPGA-to-HPS SDRAM Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–3
Reset Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–5
DMA Peripheral Request . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–5
Configuring Peripheral Pin Multiplexing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–5

xx Contents
Configuring Peripherals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–5

Connecting Unassigned Pins to GPIO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–6
Resolving Pin Multiplexing Conflicts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–6
Configuring HPS Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–7
User Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–7
PLL Reference Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–8
Configuring the External Memory Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–8
Selecting PLL Output Frequency and Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–9
Using the Address Span Extender Component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–9
Generating and Compiling the HPS Component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27–10
Chapter 28. HPS Component Interfaces

Memory-Mapped Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–1
FPGA-to-HPS Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–1
ACP Sideband Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–2
HPS-to-FPGA and Lightweight HPS-to-FPGA Bridges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–2
Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–4
Alternative Clock Inputs to HPS PLLs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–4
User Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–4
AXI Bridge FPGA Interface Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–4
SDRAM Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–4
Resets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–5
HPS-to-FPGA Reset Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–5
HPS External Reset Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–5
Debug and Trace Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–5
Trace Port Interface Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–5
FPGA System Trace Macrocell Events Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–6
FPGA Cross Trigger Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–6
Debug APB Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–6
Peripheral Signal Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–6
DMA Controller Peripheral Request Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–6
Other Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–7
MPU Standby and Event Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–7
FPGA-to-HPS Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–7
General-Purpose Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28–8
Chapter 29. Simulating the HPS Component

HPS Simulation Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–1
Clock and Reset Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–2
Clock Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–2
Reset Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–3
FPGA-to-HPS AXI Slave Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–4
HPS-to-FPGA AXI Master Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–4
Lightweight HPS-to-FPGA AXI Master Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–5
HPS-to-FPGA MPU General-Purpose I/O Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–6
HPS-to-FPGA MPU Event Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–6
FPGA-to-HPS Interrupts Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–6
HPS-to-FPGA Debug APB Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–7
FPGA-to-HPS System Trace Macrocell (STM) Hardware Event Interface . . . . . . . . . . . . . . . . . . . . . 29–7

Contents xxi
HPS-to-FPGA Cross-Trigger Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–8

HPS-to-FPGA Trace Port Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–8
FPGA-to-HPS DMA Handshake Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–9
Simulation Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–10
Specifying HPS Simulation Model in Qsys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–10
Generating HPS Simulation Model in Qsys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–13
Running HPS RTL Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–13
Running HPS Post-Fit Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29–14
Section VIII. Appendices
Appendix A. Booting and Configuration

HPS Boot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–3
Boot Process Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–4
Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–4
Boot ROM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–4
Preloader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–4
Boot Loader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–5
Boot ROM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–5
Boot ROM Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–5
Loading the Preloader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–7
Shared Memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–8
L4 Watchdog 0 Timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–10
HPS State on Entry to the Preloader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–10
Preloader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–11
Typical Preloader Boot Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–11
Flash Memory Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–14
NAND Flash Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–14
SD/MMC Flash Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–15
SPI and Quad SPI Flash Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–17
FPGA Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–19
Full Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–20
Partial Reconfiguration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–21
Document Revision History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A–22
Additional Information
How to Contact Altera . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Info–1
Typographic Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Info–1

xxii Contents

Chapter Revision Dates
The chapters in this document, Cyclone V Device Handbook, were revised on the
following dates. Where chapters or groups of chapters are available separately, part
numbers are listed.
Chapter 1. Introduction to the Hard Processor System

Revised: November 2012
Part Number: cv_54001-1.3
Chapter 2. Clock Manager

Chapter 3. Reset Manager

Chapter 4. Interconnect
Chapter 5. HPS-FPGA AXI Bridges

Chapter 6. Cortex-A9 Microprocessor Unit Subsystem

Chapter 7. CoreSight Debug and Trace

Chapter 8. SDRAM Controller Subsystem

Chapter 9. On-Chip Memory

Chapter 10. NAND Flash Controller

Chapter 11. SD/MMC Controller

Chapter 12. Quad SPI Flash Controller

xxiv Chapter Revision Dates

Chapter 13. FPGA Manager

Chapter 14. System Manager

Chapter 15. Scan Manager

Chapter 16. DMA Controller

Chapter 17. Ethernet Media Access Controller

Chapter 18. USB 2.0 OTG Controller

Chapter 19. SPI Controller

Chapter 20. I2C Controller

Chapter 21. UART Controller

Chapter 22. General-Purpose I/O Interface

Chapter 23. Timer

Chapter 24. Watchdog Timer

Chapter 25. CAN Controller


Chapter Revision Dates xxv
Chapter 26. Introduction to the HPS Component

Revised: June 2012
Chapter 27. Instantiating the HPS Component

Chapter 28. HPS Component Interfaces

Chapter 29. Simulating the HPS Component

Appendix A. Booting and Configuration

Part Number: cv_5400A-1.3

xxvi Chapter Revision Dates

Section I. Overview
This section includes the following chapters:

■ Chapter 1, Introduction to the Hard Processor System
■ Chapter 2, Clock Manager
■ Chapter 3, Reset Manager
f For information about the revision history for chapters in this section, refer to
“Document Revision History” in each individual chapter.

I–2 Section I: Overview

1. Introduction to the Hard Processor
System
November 2012
cv_54001-1.3
cv_54001-1.3
The Cyclone® V SoC FPGA device is a single-die system on a chip (SoC) that consists
of two distinct parts—a hard processor system (HPS) portion and an FPGA portion.
Figure 1–1 shows a high-level block diagram of the Altera SoC FPGA device. Blocks
connected to device pins have symbols (square with X) adjacent to them in the figure.
Figure 1–1. Altera SoC FPGA Device Block Diagram
Altera SoC FPGA Device

HPS Portion FPGA Portion
Flash SDRAM Controller

Controllers Subsystem Control User HSSI
Block I/O Transceivers
Cortex-A9 MPU Subsystem

HPS-FPGA
Interfaces FPGA Fabric
On-Chip Support (LUTs, RAMs, Multipliers & Routing)
Memories Peripherals
Interface Hard Hard Memory

PLLs Debug PLLs
Peripherals PCIe Controllers
The HPS contains a microprocessor unit (MPU) subsystem with single or dual ARM®
Cortex™-A9 MPCore processors, flash memory controllers, an SDRAM controller
subsystem, on-chip memories, support peripherals, interface peripherals, debug
capabilities, and PLLs. The dual-processor HPS supports symmetric (SMP) and
asymmetric (AMP) multiprocessing.
The FPGA portion of the device contains the FPGA fabric, a control block (CB), phase-
locked loops (PLLs), and depending on the device variant, high-speed serial interface
(HSSI) transceivers, hard PCI Express® (PCIe®) controllers, and hard memory
controllers.
f For information about the FPGA portion of the device, refer to Cyclone V Device
Overview.

November 2012
Subscribe
1–2 Chapter 1: Introduction to the Hard Processor System
Features of the HPS
The HPS and FPGA portions of the device are distinctly different. The HPS boots
(from any of multiple boot sources, including the FPGA fabric and external flash
devices) and the FPGA gets configured (through the HPS or any external source
supported by the device).
f For more information, refer to the Booting and Configuration appendix in volume 3 of
the Cyclone V Device Handbook.
The HPS and FPGA portions of the device each have their own pins. Pins are not
freely shared between the HPS and the FPGA fabric. The HPS I/O pins are configured
by software executing in the HPS. Software executing on the HPS accesses control
registers in the system manager to assign HPS I/O pins to the available HPS modules.
The FPGA I/O pins are configured by an FPGA configuration image through the HPS
or any external source supported by the device.
The MPU subsystem can boot from flash devices connected to the HPS pins. Or, when
the FPGA portion is configured by an external source, the MPU subsystem can boot
from memory available to the FPGA portion of the device.
The HPS and FPGA portions of the device have separate external power supplies and
independently power on. You can power on the HPS without powering on the FPGA
portion of the device. But to power on the FPGA portion, the HPS must already be on
or powered on at the same time as the FPGA portion. You can also turn off the FPGA
portion of the device while leaving the HPS power on.
Features of the HPS

The following list contains the main modules of the HPS:
■ An MPU subsystem featuring dual ARM Cortex-A9 MPCore processors
■ An SDRAM controller subsystem
■ One general-purpose direct memory access (DMA) controller
■ Two Ethernet media access controllers (EMACs)
■ Two USB 2.0 On-The-Go (OTG) controllers
■ One NAND flash controller
■ One quad SPI flash controller
■ One Secure Digital (SD) / MultiMediaCard (MMC) controller
■ Two serial peripheral interface (SPI) master controllers
■ Two SPI slave controllers
■ Four inter-integrated circuit (I2C) controllers
■ 64 KB on-chip RAM
■ 64 KB on-chip boot ROM
■ Two UARTs
■ Four timers
■ Two watchdog timers

Chapter 1: Introduction to the Hard Processor System 1–3
Features of the HPS
■ Three general-purpose I/O (GPIO) interfaces

■ Two controller area network (CAN) controllers (certain device variants only)
■ ARM CoreSight™ debug components
■ Debug Access Port (DAP)
■ Trace Port Interface Unit (TPIU)
■ System Trace Macrocell (STM)
■ Program Trace Macrocell (PTM)
■ Embedded Trace Router (ETR)
■ Embedded Cross Trigger (ECT)
■ A system manager
■ A clock manager
■ A reset manager
■ A scan manager
■ An FPGA manager
■ One FPGA-to-HPS bridge
■ Two HPS-to-FPGA bridges

HPS Block Diagram and System Integration

Figure 1–2 shows a block diagram of most modules in the HPS. Debug modules are
not shown.
Figure 1–2. HPS Block Diagram
FPGA Portion FPGA to HPS HPS to FPGA Lightweight HPS to FPGA
Control 1-6
Masters Slaves Slaves
Block Masters
32-, 64- & 128-Bit AXI 32-, 64- & 128-Bit AXI 32-Bit AXI
FPGA FPGA-to-HPS HPS-to-FPGA Lightweight

Manager Bridge Bridge HPS-to-FPGA Bridge
32-Bit 64-Bit AXI 64-Bit AXI 32-Bit AXI

L4, 32-Bit Bus
L3 Interconnect MPU Subsystem

(NIC-301)
ARM Cortex-A9
MPCore
32-Bit CPU0 CPU1
DAP
64-Bit ACP ID SCU
ACP
Mapper
32-Bit
ETR
L3 Main
32-Bit Switch L2
SD/MMC 64-Bit Cache
L3 Master
Peripheral 32-Bit
STM
EMAC 32-Bit Switch 32-Bit
(2)
32-Bit
Boot ROM
USB 32-Bit
OTG 64-Bit
On-Chip RAM
(2)
32-Bit SDRAM
Controller
32-Bit 64-Bit Subsystem
NAND 32-Bit
Flash DMA
32-Bit
32-Bit L3 Slave Peripheral Switch Quad
32-Bit
32-Bit SPI
Flash
L4, 32-Bit Bus
CAN Timer I 2C Watchdog UART GPIO SPI Clock Reset Scan System
(2) (4) (4) Timer (2) (3) (4) Manager Manager Manager Manager
(2)
The following sections list features of the HPS modules and reference related chapters
that provide more detailed information. The HPS incorporates third-party intellectual
property (IP) from several vendors. Each chapter in this handbook identifies
additional third-party IP documentation, if available from the third-party vendors.
1 No clock information is listed in the following summary sections. For comprehensive

clock information for all modules, refer to the Clock Manager chapter in volume 3 of

MPU Subsystem
The MPU subsystem provides the following functionality:
■ ARM Cortex-A9 MPCore
■ One or two ARM Cortex-A9 processors in a cluster
■ NEON™ SIMD coprocessor and VFPv3 per processor
■ Snoop Control Unit (SCU) to ensure coherency within the cluster
■ Accelerator coherency port (ACP) that accepts coherency memory access
requests.
■ Interrupt controller
■ One general-purpose timer and one watchdog timer per processor
■ Debug and trace features
■ 32-KB instruction and 32-KB data level 1 (L1) caches per processor
■ Memory management unit (MMU) per processor
■ ARM L2-310 level 2 (L2) cache
■ Shared 512-KB L2 cache
■ ACP ID mapper
■ Maps the 12-bit ID from the level 3 (L3) interconnect to the 3-bit ID supported
by the ACP
As shown in Figure 1–2, the L2 cache has one 64-bit master port connected to the L3
interconnect and one 64-bit master port connected directly to the SDRAM controller
subsystem. A programmable address filter in the L2 cache controls which portions of
the 32-bit physical address space use which master.
f For more information, refer to the Cortex-A9 MPU System chapter in volume 3 of the
Cyclone V Device Handbook.
Interconnect
The interconnect consists of the L3 interconnect and level 4 (L4) buses. The L3
interconnect is one ARM NIC-301 module composed of the following switches:
■ L3 main switch
■ Connects the master, slaves, and other subswitches
■ Provides 64-bit switching capabilities
■ L3 master peripheral switch
■ Connects master ports of peripherals with integrated DMA controllers to the
L3 main switch
■ L3 slave peripheral switch
■ Connects slave ports of peripherals to the L3 main switch

The L4 buses are each connected to a master in the L3 slave peripheral switch. Each L4
bus is 32 bits wide and is connected to multiple slaves. Each L4 bus operates on a
separate clock source.
f For more information, refer to the Interconnect chapter in volume 3 of the Cyclone V
Device Handbook.
Memory Controllers
The HPS provides the memory controllers described in this section.
SDRAM Controller Subsystem

The SDRAM controller subsystem is mastered by HPS masters and FPGA fabric
masters. The FPGA-to-HPS SDRAM interface is compatible with the hard memory
controllers in the FPGA portion of the device and with hard SDRAM controllers in
non-HPS FPGA devices, such as Stratix IV FPGAs.
The SDRAM controller subsystem implements the following high-level features:
■ Support for double data rate 2 (DDR2), DDR3, and low-power double data rate 2
(LPDDR2) devices
■ Software-configurable priority scheduling on individual SDRAM bursts
■ Error correction code (ECC) support, including calculation, single-bit error
correction and write-back, and error counters
■ Fully-programmable timing parameter support for all JEDEC-specified timing
parameters
■ All ports support memory protection and mutual accesses
■ Support for ARM Advanced Microcontroller Bus Architecture (AMBA®)
Advanced eXtensible Interface (AXI™) quality of service (QoS) for the fabric
interfaces
The SDRAM controller subsystem is composed of the SDRAM controller and the DDR
PHY.
SDRAM Controller
The SDRAM controller contains a multiport front end (MPFE) that accepts requests
from HPS masters and from soft logic in the FPGA fabric via the FPGA-to-HPS
SDRAM interface.
The SDRAM controller offers the following features:
■ Up to 4 GB address range
■ 8-, 16-, and 32-bit data widths
■ Optional ECC support
■ Low-voltage 1.35V DDR3L and 1.2V DDR3U support
■ Full memory device power management support
■ Two chip selects

The SDRAM controller provides the following features to maximize memory

performance:
■ Command reordering (look-ahead bank management)
■ Data reordering (out of order transactions)
■ Deficit round-robin arbitration with aging for bandwidth management
■ High-priority bypass for latency sensitive traffic
DDR PHY
The DDR PHY interfaces the single-port memory controller to the HPS memory I/O.
f For more information, refer to the SDRAM Controller Subsystem chapter in volume 3 of
NAND Flash Controller

The NAND flash controller is based on the Cadence® Design IP® NAND Flash
Memory Controller and offers the following features:
■ Supports single-level cell (SLC) and multilevel cell (MLC) NAND flash devices
■ Integrated descriptor-based DMA controller
■ 8-bit ONFI 1.0 NAND flash devices
■ Programmable page sizes of 512 bytes, 2 KB, 4 KB, and 8 KB
■ Supports 32, 64, 128, 256, 384, and 512 pages per block
■ Programmable hardware ECC for SLC and MLC devices
■ 512 bytes ECC sector size with 4-, 8-, or 16-bit correction
■ 1 KB ECC sector size with 24-bit correction
f For more information, refer to the NAND Flash Controller chapter in volume 3 of the
Quad SPI Flash Controller

The quad SPI flash controller is based on Cadence Quad SPI Flash Controller
(QSPI_FLASH_CTRL) and offers the following features:
■ Supports SPIx1, SPIx2, or SPIx4 (quad SPI) serial NOR flash devices
■ Supports direct access and indirect access modes.
■ Supports single I/O, dual I/O, quad I/O instructions
■ Programmable data frame size of 8, 16, or 32 bits
■ Support up to four chip selects
f For more information, refer to the Quad SPI Controller chapter in volume 3 of the

SD/MMC Controller
The SD/MMC controller is based on Synopsys® DesignWare® Mobile Storage Host
(DWC_mobile_storage) controller and offers the following features:
■ Integrated descriptor-based DMA
■ Supports CE-ATA digital protocol commands
■ Supports single card
■ Single data rate (SDR) mode only
■ Programmable card width: x1, x4, or x8
■ Programmable card types: SD, SDIO, or MMC version 4.3 and 4.4 devices
■ Up to 64 KB programmable block size
f For more information, refer to the SD/MMC Controller chapter in volume 3 of the
Support Peripherals
HPS provides the support peripherals described in this section.
Clock Manager
The clock manager offers the following features:
■ Manages clocks for HPS
■ Supports dynamic clock tuning
f For more information, refer to the Clock Manager chapter in volume 3 of the Cyclone V
Device Handbook.
Reset Manager
The reset manager offers the following features:
■ Manages reset for HPS
f For more information, refer to the Reset Manager chapter in volume 3 of the Cyclone V
Device Handbook.
System Manager
The system manager offers the following features:
■ ECC monitoring and control
■ Pin multiplexing
■ Low-level control of peripheral features not accessible through the control and
status registers (CSRs)
■ Freeze controller that places I/O elements into a safe state for configuration
f For more information, refer to the System Manager chapter in volume 3 of the

Scan Manager
The scan manager offers the following features:
■ Drives serial scan-chains to FPGA JTAG and HPS I/O bank configuration
f For more information, refer to the Scan Manager chapter in volume 3 of the Cyclone V
Device Handbook.
Timers
The four timers are based on the Synopsys DesignWare APB Timers (DW_apb_timers)
peripheral and offer the following features:
■ 32-bit timer resolution
■ Support free-running timer mode
■ Programmable time-out period up to approximately 86 seconds (assuming a
50 MHz clock)
■ Interrupt generation
f For more information, refer to the Timer chapter in volume 3 of the Cyclone V Device
Handbook.
Watchdog Timers
The two watchdog timers are based on the Synopsys DesignWare APB Watchdog
Timer (DW_apb_wdt) peripheral and offer the following features:
■ 32-bit timer resolution
■ Interrupt request
■ Reset request
■ Programmable time-out period up to approximately 86 seconds (assuming a
50 MHz clock)
f For more information, refer to the Watchdog Timer chapter in volume 3 of the Cyclone V
Device Handbook.
DMA Controller
The DMA controller provides high-bandwidth data transfers for modules without
integrated DMA controllers. The DMA controller is based on the ARM Corelink™
DMA Controller (DMA-330) and offers the following features:
■ Microcoded to support flexible transfer types
■ Supports up to 8 channels
■ Supports flow control with 31 peripherals handshake interfaces
f For more information, refer to the DMA Controller chapter in volume 3 of the

FPGA Manager
The FPGA manager offers the following features:
■ Manages configuration of the FPGA portion of the device
■ Mimics passive parallel 32-bit configuration
■ Partial reconfiguration
■ Compressed FPGA configuration images
■ Advanced Encryption Standard (AES) encrypted FPGA configuration images
■ Monitors configuration-related signals in FPGA
■ Provides 32 general-purpose inputs and 32 general-purpose outputs to the FPGA
fabric
f For more information, refer to the FPGA Manager chapter in volume 3 of the Cyclone V
Device Handbook.
Interface Peripherals
HPS provides the interface peripherals described in this section.
EMACs
The two EMACs are based on the Synopsys DesignWare 3504-0 Universal
10/100/1000 Ethernet MAC (DWC_gmac) and offer the following features:
■ Supports 10-, 100-, and 1000-Mbps standards
■ Supports RGMII external PHY interface
■ Integrated DMA controllers
f For more information, refer to the Ethernet Media Access Controller chapter in volume 3
of the Cyclone V Device Handbook.
USB Controllers
The two USB 2.0 On-The-Go (OTG) controllers are based on the Synopsys DesignWare
Cores USB 2.0 Hi-Speed On-The-Go (DWC_otg) controller and offer the following
features:
■ Supports USB 2.0 host and device operation
■ Dual-role device (device and host functions)
■ High-speed (480 Mbps)
■ Full-speed (12 Mbps)
■ Low-speed (1.5 Mbps)
■ Supports USB 1.1 (full-speed & low-speed)
■ Integrated descriptor-based scatter-gather DMA (SGDMA)
■ Support for external ULPI PHY
■ Up to 16 bidirectional endpoints, including control endpoint

■ Up to 16 host channels
■ Supports generic root hub
■ Automatic ping capability
■ Configurable to OTG 1.3 and OTG 2.0 modes
f For more information, refer to the USB 2.0 OTG Controller chapter in volume 3 of the
I2C Controllers
The four I2C controllers are based on Synopsys DesignWare APB I2C (DW_apb_i2c)
controller and offer the following features:
■ Two controllers support I2C management interfaces
■ Support both 100 KBps and 400 KBps modes
■ Support both 7-bit and 10-bit addressing modes
■ No support for mixed-address mode
■ Support master and slave operating mode
■ Direct access for host processor
■ DMA controller may be used for large transfers
f For more information, refer to the I2C Controller chapter in volume 3 of the Cyclone V
Device Handbook.
UARTs
The two UART modules are based on Synopsys DesignWare APB Universal
Asynchronous Receiver/Transmitter (DW_apb_uart) peripheral and offer the
following features:
■ 16550 compatible UART
■ Supports the auto flow control as specified in 16750 specification
■ Supports IrDA 1.0 SIR mode
■ Programmable baud rate up to 115.2 Kbps
f For more information, refer to the UART Controller chapter in volume 3 of the
CAN Controllers
The two CAN controllers are based on the Bosch® D_CAN controller and offer the
following features:
■ Compliant with CAN protocol specification 2.0 part A & B
■ Programmable communication rate up to 1 Mbps

■ Holds up to 128 messages

■ Supports 11-bit standard and 29-bit extended identifiers
■ Programmable interrupt scheme
■ Available on certain device variants only
f For more information, refer to the Controller Area Network Controller chapter in
volume 4 of the Cyclone V Device Handbook.
SPI Master Controllers

The two SPI master controllers are based on Synopsys DesignWare Synchronous
Serial Interface (SSI) controller (DW_apb_ssi) and offer the following features:
■ Programmable data frame size from 4 to 16 bits
■ Supports full and half duplex
■ Support up to two chip selects
f For more information, refer to the SPI Controller chapter in volume 3 of the Cyclone V
Device Handbook.
SPI Slave Controllers

The two SPI slave controllers are based on Synopsys DesignWare Synchronous Serial
Interface (SSI) controller (DW_apb_ssi) and offer the following features:
■ Programmable data frame size from 4 to 16 bits
■ Supports full and half duplex
f For more information, refer to the SPI Controller chapter in volume 3 of the Cyclone V
Device Handbook.
GPIO Interfaces
The three GPIO interfaces are based on Synopsys DesignWare APB General Purpose
Programming I/O (DW_apb_gpio) peripheral and offer the following features:
■ Supports digital de-bounce
■ Configurable interrupt mode
■ Supports up to 71 I/O pins and 14 input-only pins, based on device variant
f For more information, refer to the General-Purpose I/O Interface chapter in volume 3 of

Endian Support
On-Chip Memory
On-chip memory consists of the two modules described in this section.
On-Chip RAM
The on-chip RAM offers the following features:
■ 64 KB size
■ 64-bit slave interface
■ High performance for all burst lengths
f For more information, refer to the On-Chip Memory chapter in volume 3 of the
Boot ROM
The boot ROM offers the following features:
■ 64 KB size
■ Contains the code required to support HPS boot from cold or warm reset
■ Used exclusively for booting the HPS
f For more information, refer to the On-Chip Memory chapter in volume 3 of the
Endian Support
The HPS is natively a little-endian system. All HPS slaves are little-endian.
The processors masters are software configurable to interpret data as little-endian or
big-endian, byte-invariant (BE8). All other masters, including the USB interface, are
little-endian.
The FPGA-to-HPS, HPS-to-FPGA, and lightweight HPS-to-FPGA interfaces are
little-endian.
If a processor is set to BE8 mode, software must convert endianness for accesses to
peripherals and DMA linked lists in memory.
The ARM Cortex-A9 MPU supports a single instruction to change the endianness of
the processor and provides the REV and REV16 instructions to swap the endianness
of bytes or half-words respectively. The MMU page tables are software configurable to
be organized as little-endian or BE8.
The ARM DMA controller is software configurable to perform byte lane swapping
during a transfer.
HPS-FPGA Interfaces
The HPS-FPGA interfaces provide a variety of communication channels between the
HPS and the FPGA fabric. The HPS is highly integrated with the FPGA fabric,
resulting in thousands of connecting signals. The HPS-FPGA interfaces include:

Address Map
■ FPGA-to-HPS bridge—a high-performance AXI bus with a configurable data

width of 32-, 64-, and 128-bits, allowing the FPGA fabric to master transactions to
the slaves in the HPS. This interface allows the FPGA fabric to have full visibility
into the HPS address space. This interface also provides access to the coherent
memory interface.
f For information about the coherent memory interface, refer to the Cortex-A9
MPU System chapter in volume 3 of the Cyclone V Device Handbook.
■ HPS-to-FPGA bridge—a high-performance AXI bus with a configurable data

width of 32-, 64-, and 128-bits, allowing the HPS to master transactions to slaves in
the FPGA fabric.
■ Lightweight HPS-to-FPGA bridge—an AXI bus with a 32-bit fixed data width,
allowing the HPS to master transactions to slaves in the FPGA fabric.
■ FPGA-to-HPS SDRAM interface—a configurable interface to the MPFE of the
SDRAM controller. You can configure the following parameters:
■ AXI-3 or Avalon® Memory-Mapped (Avalon-MM) protocol
■ Up to six ports
■ 32-, 64-, 128-, or 256-bit data width of each port
■ FPGA clocks and resets—provide flexible clocks to and from the HPS.
■ HPS-to-FPGA JTAG—allows the HPS to master the FPGA JTAG chain.
■ TPIU trace—sends trace data created in the HPS to the FPGA fabric.
■ FPGA System Trace Macrocell (STM) events—an interface that allows the FPGA
fabric to send hardware events stored in the HPS trace using STM.
■ FPGA cross-trigger—an interface that allows triggers to and from the CoreSight
trigger system.
■ DMA peripheral interface—multiple peripheral-request channels.
■ FPGA manager interface—signals that communicate with FPGA fabric for boot
and configuration.
■ Interrupts—allow soft IP to supply interrupts directly to the MPU interrupt
controller.
■ MPU standby and events—signals that notify the FPGA fabric that the MPU is in
standby mode and signals that wake up Cortex-A9 processors from a wait for
event (WFE) state.
Address Map
The address map specifies the addresses of slaves, such as memory and peripherals,
as viewed by the MPU and other masters. The HPS has multiple address spaces,
defined in the following section.

Address Map
Address Spaces
Table 1–1 shows the HPS address spaces and their sizes.
Table 1–1. HPS Address Spaces

Name Description Size
MPU MPU subsystem 4 GB
L3 L3 interconnect 4 GB
SDRAM SDRAM controller subsystem 4 GB
Address spaces are divided into one or more nonoverlapping contiguous regions. For
example, the MPU address space has the peripheral, FPGA slaves, SDRAM window,
and boot regions.
Figure 1–3 shows the relationships between the HPS address spaces. The figure is not
to scale.
Figure 1–3. HPS Address Space Relationships

4 GB
Peripheral Region Lightweight Peripheral Region
FPGA
FPGA Slaves FPGA
Slaves Slaves
Region Region
3 GB
ACP
Window
SDRAM 2 GB
Region
SDRAM
Window
SDRAM
Window 1 GB
Boot Region
RAM / SDRAM (ROM/RAM/SDRAM)
0 GB
L3 MPU SDRAM
The window regions provide access to other address spaces. The thin black arrows
indicate which address space is accessed by a window region (arrows point to
accessed address space). For example, accesses to the ACP window in the L3 address
space map to a 1 GB region of the MPU address space.
The SDRAM window in the MPU address space can grow and shrink at the top and
bottom (short, blue vertical arrows) at the expense of the FPGA slaves and boot
regions. For specific details, refer to “MPU Address Space”.
The ACP window can be mapped to any 1 GB region in the MPU address space (blue
vertical bidirectional arrow), on gigabyte-aligned boundaries.

Address Map
Table 1–2 shows the base address and size of each region that is common to the L3 and
MPU address spaces.
Table 1–2. Common Address Space Regions

Identifier Region Name Base Address Size
FPGASLAVES FPGA slaves 0xC0000000 960 MB
LWFPGASLAVES Lightweight FPGA slaves 0xFF200000 2 MB
PERIPH Peripheral 0xFC000000 64 MB
SDRAM Address Space

The SDRAM address space is up to 4 GB. The entire address space can be accessed
through the FPGA-to-HPS SDRAM interface from the FPGA fabric. The total amount
of SDRAM addressable from the other address spaces varies. For specific details, refer
to “MPU Address Space” and “L3 Address Space”.
MPU Address Space

The MPU address space is 4 GB and applies to addresses generated inside the MPU.
The MPU address space contains the following regions:
■ The SDRAM window region provides access to a large, configurable portion of the
4 GB SDRAM address space. The MPU L2 cache controller contains a master
connected to the L3 interconnect and a master connected to the SDRAM. The
address filtering start and end registers in the L2 cache controller define the
SDRAM window boundaries. The boundaries are megabyte-aligned. Addresses
within the boundaries route to the SDRAM master. Addresses outside the
boundaries route to the L3 interconnect master.
Figure 1–3 shows the reset values of the SDRAM window boundaries. By default,
processor accesses to locations between 0x100000 (1 MB) to 0xC0000000 (3 GB) are
made to the SDRAM controller, accesses to all other locations are made to the L3
interconnect. Addresses in the SDRAM window match addresses in the SDRAM
address space. Thus, the lowest 1 MB of the SDRAM is not visible to the MPU
unless the L2 address filter start register is set to 0.
f For more information about L2 address filtering, refer to the Cortex-A9

■ The boot region is 1 MB starting at address 0x0 and is visible to the MPU only
when the L2 address filter start register is set to 0x100000. The L3 interconnect
Global Programmers View (GPV) remap control register determines if the boot
region is mapped to the on-chip RAM or the boot ROM.
f For information about the L3 GPV remap control register bits, refer to the
Interconnect chapter in volume 3 of the Cyclone V Device Handbook.
The boot region is mapped to the boot ROM on reset. Only the lowest 64 KB of the
boot region are legal addresses because the on-chip RAM and boot ROM are only
64 KB.

Address Map
1 When the L2 address filter start register is set to 0, SDRAM obscures access
to the boot region. This technique can be used to gain access to the lowest
SDRAM addresses after booting completes.
■ The FPGA slaves region provides access to 960 MB of slaves in the FPGA fabric
through the HPS-to-FPGA bridge. If the top of the SDRAM window increases in
the MPU address space (by writing to the L2 address filter end register), the lower
portion of the FPGA slaves region is obscured from the MPU subsystem.
■ The peripheral region contains 64 MB at the top of the address space. The
peripheral region includes all slaves connected to the L3 interconnect, L4 buses,
and internally-decoded MPU registers (SCU and L2). The boot ROM and on-chip
RAM are always mapped into the peripheral region (independent of the boot
region contents). The lightweight FPGA slaves are also mapped in the peripheral
region and provide access to 2 MB of slaves in the FPGA fabric through the
lightweight HPS-to-FPGA bridge.
Table 1–3 shows the base address and size of each MPU address space region that is
not included in Table 1–2.
Table 1–3. MPU Default Address Space Regions

MPUBOOT Boot region 0x00000000 1 MB
MPUSDRAM SDRAM window 0x00100000 3071 MB
L3 Address Space
The L3 address space is 4 GB and applies to all L3 masters except the MPU subsystem.
The L3 address space configurations contain the following regions:
■ The peripheral region is the same as the peripheral region in the MPU address
space except that the boot ROM and internal MPU registers (SCU and L2) are not
accessible.
■ The FPGA slaves region provides access to 960 MB of slaves in the FPGA fabric
through the HPS-to-FPGA bridge.
■ The SDRAM window region is 2 GB and provides access to the bottom 2 GB of the
SDRAM address space. The L3 interconnect GPV remap register determines if the
64 KB starting at address 0x0 is mapped to the on-chip RAM or the SDRAM. The
SDRAM is mapped to address 0x0 on reset.
f For information about the L3 GPV remap control register bits, refer to the
■ The ACP window region is 1 GB and provides access to a configurable

gigabyte-aligned region of the MPU address space. Registers in the ACP ID
mapper control which gigabyte-aligned region of the MPU address space is
accessed by the ACP window region. The ACP window region is used by L3
masters to perform coherent accesses into the MPU address space.
f For more information about the ACP ID mapper, refer to the Cortex-A9

Address Map
Table 1–4 shows the base address and size of each L3 address space region that is not
included in Table 1–2.
Table 1–4. L3 Address Space Regions

L3SDRAM SDRAM window 0x00000000 2 GB
L3LOWOCRAM On-chip RAM when present 0x00000000 64 KB
L3ACP ACP window 0x80000000 1 GB
Peripheral Region Address Map

Table 1–5 lists the slave identifier, slave title, base address, and size of each slave in the
peripheral region. The Slave Identifier column lists the names used in the HPS register
map. The Slave Title column contains the module name for modules with only one
slave and module names plus a suffix for modules with more than one slave.
Table 1–5. Peripheral Region Address Map (Part 1 of 2)

Slave Identifier Slave Title Base Address Size
STM STM 0xFC000000 48 MB
DAP DAP 0xFF000000 2 MB
FPGA slaves accessed with
LWFPGASLAVES 0xFF200000 2 MB
lightweight FPGA-to-HPS AXI bridge
Lightweight FPGA-to-HPS AXI bridge
LWHPS2FPGAREGS 0xFF400000 1 MB
GPV
HPS2FPGAREGS HPS-to-FPGA AXI bridge GPV 0xFF500000 1 MB
FPGA2HPSREGS FPGA-to-HPS AXI bridge GPV 0xFF600000 1 MB
EMAC0 EMAC0 0xFF700000 8 KB
EMAC1 EMAC1 0xFF702000 8 KB
SDMMC SD/MMC 0xFF704000 4 KB
QSPIREGS Quad SPI flash controller registers 0xFF705000 4 KB
FPGAMGRREGS FPGA manager registers 0xFF706000 4 KB
ACPIDMAP ACP ID mapper registers 0xFF207000 4 KB
GPIO0 GPIO0 0xFF208000 4 KB
GPIO1 GPIO1 0xFF209000 4 KB
GPIO2 GPIO2 0xFF20A000 4 KB
L3REGS L3 interconnect GPV 0xFF800000 1 MB
NANDDATA NAND controller data 0xFF900000 1 MB
QSPIDATA Quad SPI flash data 0xFFA00000 1 MB
USB0 USB0 OTG controller registers 0xFFB00000 256 KB
USB1 USB1 OTG controller registers 0xFFB40000 256 KB
NANDREGS NAND controller registers 0xFFB80000 64 KB
FPGAMGRDATA FPGA manager configuration data 0xFFB90000 4 KB
CAN0 CAN0 controller registers 0xFFC00000 4 KB
CAN1 CAN1 controller registers 0xFFC01000 4 KB

Document Revision History
Table 1–5. Peripheral Region Address Map (Part 2 of 2)

Slave Identifier Slave Title Base Address Size
UART0 UART0 0xFFC02000 4 KB
UART1 UART1 0xFFC03000 4 KB
I2C0 I2C0 0xFFC04000 4 KB
SPTIMER0 SP Timer0 0xFFC08000 4 KB
SPTIMER1 SP Timer1 0xFFC09000 4 KB
SDRAM controller subsystem
SDRREGS 0xFFC20000 128 KB
registers
OSC1TIMER0 OSC1 Timer0 0xFFD00000 4 KB
OSC1TIMER1 OSC1 Timer1 0xFFD01000 4 KB
L4WD0 Watchdog0 0xFFD02000 4 KB
L4WD1 Watchdog1 0xFFD03000 4 KB
CLKMGR Clock manager 0xFFD04000 4 KB
RSTMGR Reset manager 0xFFD05000 4 KB
SYSMGR System manager 0xFFD08000 16 KB
DMANONSECURE DMA nonsecure registers 0xFFE00000 4 KB
DMASECURE DMA secure registers 0xFFE01000 4 KB
SPIS0 SPI slave0 0xFFE02000 4 KB
SPIS1 SPI slave1 0xFFE03000 4 KB
SPIM0 SPI master0 0xFFF00000 4 KB
SPIM1 SPI master1 0xFFF01000 4 KB
SCANMGR Scan manager registers 0xFFF02000 4 KB
ROM Boot ROM 0xFFFD0000 64 KB
MPUSCU MPU SCU registers 0xFFFEC000 8 KB
MPUL2 MPU L2 cache controller registers 0xFFFEF000 4 KB
OCRAM On-chip RAM 0xFFFF0000 64 KB

Table 1–6 shows the revision history for this document.
Table 1–6. Document Revision History

Date Version Changes
November 2012 1.3 Minor updates.
June 2012 1.2 Updated address spaces section.
May 2012 1.1 Added peripheral region address map.
January 2012 1.0 Initial release.


2. Clock Manager
November 2012
cv_54002-1.2
cv_54002-1.2
The hard processor system (HPS) clock generation is centralized in the clock manager.
The clock manager is responsible for providing software-programmable clock control
to configure all clocks generated in the HPS. Clocks are organized in clock groups. A
clock group is a set of clock signals that originate from the same clock source. A
phase-locked loop (PLL) clock group is a clock group where the clock source is a
common PLL voltage-controlled oscillator (VCO).
Features of the Clock Manager

The clock manager offers the following features:
■ Generates and manages clocks in the HPS
■ Contains the following PLL clock groups:
■ Main—contains clocks for the Cortex™-A9 microprocessor unit (MPU)
subsystem, level 3 (L3) interconnect, level 4 (L4) peripheral bus, and debug
■ Peripheral—contains clocks for PLL-driven peripherals
■ SDRAM—contains clocks for the SDRAM subsystem
■ Allows scaling of the MPU subsystem clocks without disabling peripheral and
SDRAM clock groups
■ Generates clock gate controls for enabling and disabling most clocks
■ Initializes and sequences clocks for the following events:
■ Cold reset
■ Safe mode request from reset manager on warm reset

November 2012
Subscribe
2–2 Chapter 2: Clock Manager
Features of the Clock Manager
■ Allows software to program clock characteristics, such as the following items

discussed later in this chapter:
■ Input clock source for SDRAM and peripheral PLLs
■ Multiplier range, divider range, and six post-scale counters for each PLL
■ Output phases for SDRAM PLL outputs
■ VCO enable for each PLL
■ Bypass modes for each PLL
■ Gate off individual clocks in all PLL clock groups
■ Clear loss of lock status for each PLL
■ Safe mode for hardware-managed clocks
■ General-purpose I/O (GPIO) debounce clock divide
■ Allows software to observe the status of all writable registers
■ Supports interrupting the MPU subsystem on PLL-lock and loss-of-lock
■ Supports clock gating at the signal level
1 The clock manager is not responsible for the following functional behaviors:
■ Selection or management of the clocks for the FPGA-to-HPS, HPS-to-FPGA,

and FPGA-to-HPS SDRAM interfaces. The FPGA logic designer is responsible
for selecting and managing these clocks.
■ Software must not program the clock manager with illegal values. If it does, the
behavior of the clock manager is undefined and could stop the operation of the
HPS. The only guaranteed means for recovery from an illegal clock setting is a
cold reset.
■ When re-programming clock settings, there are no automatic glitch-free clock
transitions. Software must follow a specific sequence to ensure glitch-free clock
transitions. Refer to “Hardware-Managed and Software-Managed Clocks” on
page 2–5.

Chapter 2: Clock Manager 2–3
Clock Manager Block Diagram and System Integration
Clock Manager Block Diagram and System Integration

Figure 2–1 shows the major components of the clock manager and its integration in
the HPS.
Figure 2–1. Clock Manager Block Diagram
FPGA Portion
f2h_sdram_ref_clk f2h_periph_ref_clk
Clock Manager
Flash Controller Clocks
Flash
Divider
Controllers
Main Clock Group

Main MPU, L3, L4
Dividers & Debug
PLL
Peripheral Clock Group

Peripheral PLL-Driven
Dividers Peripherals
PLL
SDRAM Clock Group

EOSC2 SDRAM
SDRAM
Controller
PLL
Subsystem
OSC1 Clock Group

EOSC1 osc1_clk OSC1-Driven
Peripherals
Reset reset_manager_safe_mode_req Control Control & Status

Manager Logic Registers
L4 Bus (osc1_clk)
The next section describes the functional blocks internal to the clock manager. Refer to
Figure 2–3 to Figure 2–6 for more detailed versions of the grey boxes in Figure 2–1.
Functional Description of the Clock Manager

This section describes the functional operation of the clock manager.
Clock Manager Building Blocks

The clock manager has the following major building blocks.
PLLs
The clock manager contains three PLLs: main, peripherals, and SDRAM. These PLLs
generate the majority of clocks in the HPS. There is no phase control between the
clocks generated by the three PLLs.

Each PLL has the following features:

■ Phase detector and output lock signal generation
■ Registers to set VCO frequency
■ Multiplier range is 1 to 4096
■ Divider range is 1 to 64
■ Six post-scale counters (C0-C5) with a range of 1 to 512
■ PLL can be enabled to bypass all outputs to the osc1_clk clock for glitch-free
transitions
The SDRAM PLL has the following additional feature:
■ Phase shift of 1/8 per step
■ Phase shift range is 0 to 7
Equation 2–1 shows equations for FREF, FVCO, and FOUT. The values for M, N, and C are
stored in registers accessible to software.
Equation 2–1. FREF, FVCO, and FOUT Equations

FREF = FIN / N
FVCO = FREF × M = FIN × M/N
FOUT = FVCO / (Ci × K) = FREF × M/ (Ci× K) = (FIN × M)/ (N × Ci × K)
where:
(1) FVCO = VCO frequency.
(2) FIN = input frequency.
(3) FREF = reference frequency.
(4) M = numerator, part of the clock feedback path.
(5) N = denominator, part of the input clock reference path.
(6) Ci = post-scale counter, where i is 0-5 for each of the six counters.
(7) K is an internal post-scale counter in the main PLL, where K = 2 for C0, and K = 4 for C1 and C2. K = 1 for
C3, C4, and C5 in the main PLL and for all Ci counters in the peripheral and SDRAM PLLs.
f Minimum and maximum VCO frequencies for the main, peripheral, and SDRAM
PLLs vary by device speed grade. For specific details, refer the Cyclone V Device
Datasheet.

Figure 2–2 shows the block diagram of each PLL. Values listed for M, N, and C are
actually one greater than the values stored in the CSRs.
Figure 2–2. PLL Block Diagram
FREF (1) (2)

FVCO FOUT
FIN
N Phase Shift C0 Divide 0
(1 - 64) PFD VCO CLKOUT0
(1/8 Per Step) (1 - 512) × K 1
FOUT
Phase Shift C1 Divide 0
CLKOUT1
FFB (1/8 Per Step) (1 - 512) × K 1
M
(1 - 4096) FOUT
CLKOUT2
(1/8 Per Step) (1 - 512) × K 1
FOUT
CLKOUT3
(1/8 Per Step) (1 - 512) 1
FOUT
CLKOUT4
(1/8 Per Step) (1 - 512) 1
FOUT
CLKOUT5
(1/8 Per Step) (1 - 512) 1
PLL Bypass Path
Bypass
Multiplexer
Notes to Figure 2–2:

(1) Phase shift is only available for SDRAM PLL outputs.
(2) In the main PLL, K=2 for C0, and K=4 for C1 and C2. K=1 in the peripheral and SDRAM PLLs.
Dividers
Dividers subdivide the C0-C5 clocks produced by the PLL to lower frequencies. The
main PLL C0-C2 clocks have an additional internal post-scale counter.
Clock Gating
Clock gating enables and disables clock signals.
Control and Status Registers

The clock manager contains registers used to configure and observe the clock
manager.
Hardware-Managed and Software-Managed Clocks

When changing values on clocks, the terms hardware-managed and software-managed
define who is responsible for successful transitions. Software-managed clocks require
that software manually gate off any clock affected by the change, wait for any PLL
lock if required, then gate the clocks back on. Hardware-managed clocks use
hardware to ensure that a glitch-free transition to a new clock value occurs. There are
three hardware-managed sets of clocks in the HPS, namely, clocks generated from the
main PLL outputs C0, C1, and C2. All other clocks in the HPS are software-managed
clocks.

Clock Groups
The clock manager contains one clock group for each PLL and one clock group for the
EOSC1 pin.
OSC1 Clock Group

The clock in the OSC1 clock group is derived directly from the EOSC1 pin. This clock is
never gated or divided. It is used as a PLL input and also by HPS logic that does not
operate on a clock output from a PLL.
Table 2–1 lists the clock in the OSC1 clock group.
Table 2–1. OSC1 Clock Group Clock
Name Frequency Clock Source Destination
OSC1-driven peripherals listed in Table 2–9

osc1_clk 10 to 50 MHz EOSC1 pin
on page 2–14
Main Clock Group

The main clock group consists of a PLL, dividers, and clock gating. The clocks in the
main clock group are derived from the main PLL. The main PLL is always sourced
from the EOSC1 pin of the device.
Table 2–2 lists the main PLL output assignments.
Table 2–2. Main PLL Output Assignments

PLL Output Counter Clock Name Frequency Phase Shift Control
C0 mpu_base_clk osc1_clk to varies (1) No
C1 main_base_clk osc1_clk to varies (1) No
C2 dbg_base_clk osc1_clk/4 to mpu_base_clk/2 No
C3 main_qspi_base_clk Up to 432 MHz No

Main
Up to 250 MHz for the NAND flash
C4 main_nand_sdmmc_base_clk controller and up to 200 MHz for the No
SD/MMC controller
osc1_clk to 125 MHz for driving
C5 cfg_h2f_user0_base_clk configuration and 100 MHz for the No
user clock
Note to Table 2–2:
(1) The maximum frequency depends on the speed grade of the device.
The counter outputs from the main PLL can have their frequency further divided by
programmable dividers external to the PLL. Transitions to a different divide value
occur on the fastest output clock, one clock cycle prior to the slowest clock’s rising
edge. For example, cycle 15 of the divide-by-16 divider for the main C2 output and
cycle 3 of the divide-by-4 divider for the main C0 output.

Figure 2–3 shows how each counter output from the main PLL can have its frequency
further divided by programmable post-PLL dividers. Green-colored clock gating logic
is directly controlled by software writing to a register. Orange-colored clock gating
logic is controlled by hardware. Orange-colored clock gating logic allows hardware to
seamlessly transition a synchronous set of clocks, for example, all the MPU subsystem
clocks.
Figure 2–3. Main Clock Group Divide and Gating
Divide
by 4 mpu_periph_clk
Divide mpu_l2_ram_clk
by 2
Main mpu_base_clk
C0 Clock Gate mpu_clk
PLL
main_base_clk
C1 Clock Gate Clock Gate l4_main_clk
dbg_base_clk l3_main_clk
C2 Clock Gate
Divide l3_mp_clk
main_qspi_base_clk by 1 or 2 Clock Gate
C3 To Flash Divide
Controller by 1 or 2 l3_sp_clk
main_nand_sdmmc_base_clk
C4 Clocks
Divide by l4_mp_clk
cfg_h2f_user0_base_clk 1, 2, 4, 8, or 16 Clock Gate
C5
periph_base_clk (from Peripheral PLL C4)

Divide by l4_sp_clk
1, 2, 4, 8, or 16 Clock Gate
Divide by Clock Gate dbg_at_clk

1, 2, or 4
Divide dbg_clk
by 2 or 4 Clock Gate
Divide by Clock Gate dbg_trace_clk

1, 2, 4, 8, or 16
Clock Gate dbg_timer_clk
Clock Gate cfg_clk
Clock Gate h2f_user0_clock
The clocks derived from main PLL C0-C2 outputs are hardware-managed, meaning
hardware ensures that a clean transition occurs, and can have the following control
values changed dynamically by software write accesses to the control registers:
■ PLL bypass
■ PLL numerator, denominator, and counters
■ External dividers
For these registers, hardware detects that the write has occurred and performs the
correct sequence to ensure that a glitch-free transition to the new clock value occurs.
These clocks can pause during the transition.

Table 2–3 lists the clocks in the main clock group.
Table 2–3. Main Clock Group Clocks

System Clock Name Frequency Constraints and Notes
mpu_clk Main PLL C0 Clock for MPU subsystem, including CPU0 and CPU1
mpu_l2_ram_clk mpu_clk/2 Clock for MPU level 2 (L2) RAM
Clock for MPU snoop control unit (SCU) peripherals, such as the
mpu_periph_clk mpu_clk/4
general interrupt controller (GIC)
l3_main_clk Main PLL C1 Clock for L3 main switch
l3_main_clk or
l3_mp_clk Clock for L3 master peripherals (MP) switch
l3_main_clk/2
l3_mp_clk or
l3_sp_clk Clock for L3 slave peripherals (SP) switch
l3_mp_clk/2
l4_main_clk Main PLL C1 Clock for L4 main bus
osc1_clk/16 to 100 MHz
l4_mp_clk divided from main PLL C1 Clock for L4 MP bus
or peripheral PLL C4
osc1_clk/16 to 100 MHz
l4_sp_clk divided from main PLL C1 Clock for L4 SP bus
or peripheral PLL C4
osc1_clk/4 to main PLL
dbg_at_clk Clock for CoreSight™ debug trace bus
C2/2
osc1_clk/16 to main PLL
dbg_trace_clk Clock for CoreSight™ debug Trace Port Interface Unit (TPIU)
C2
dbg_timer_clk osc1_clk to main PLL C2 Clock for the trace timestamp generator
dbg_at_clk/2 or
dbg_clk Clock for Debug Access Port (DAP) and debug peripheral bus
dbg_at_clk/4
main_qspi_clk Main PLL C3 Quad SPI flash internal logic clock
main_nand_sdmmc_clk Main PLL C4 Input clock to flash controller clocks block
osc1_clk to 125_MHz
cfg_clk FPGA manager configuration clock
divided from main PLL C5
osc1_clk to 100_MHz
h2f_user0_clock Auxiliary user clock to the FPGA fabric
divided from main PLL C5
Changing Values That Affect Main Clock Group PLL Lock

To change any value that affects VCO lock of the main clock group PLL, including the
hardware-managed clocks, software must put the main PLL in bypass mode, which
causes all the main PLL output clocks to be driven by the osc1_clk clock. Software
must detect PLL lock by reading the lock status register prior to taking the main PLL
out of bypass mode.
Once a PLL is locked, changes to any PLL VCO frequency that are 20 percent or less
do not cause the PLL to lose lock. Iteratively changing the VCO frequency in
increments of 20 percent or less allow a slow ramp of the VCO base frequency without
loss of lock.For example, to change a VCO frequency by 40% without losing lock,
change the frequency by 20%, then change it again by 16.7%.

Peripheral Clock Group

The peripheral clock group consists of a PLL, dividers, and clock gating. The clocks in
the peripheral clock group are derived from the peripheral PLL. The peripheral PLL
can be programmed to be sourced from the EOSC1 pin, the EOSC2 pin, or the
f2h_periph_ref_clk clock provided by the FPGA fabric.
The counter outputs from the main PLL can have their frequency further divided by
external dividers. Transitions to a different divide value occur on the fastest output
clock, one clock cycle prior to the slowest clock’s rising edge. For example, cycle 15 of
the divide-by-16 divider for the main C2 output and cycle 3 of the divide-by-4 divider
for the C1 output.
Table 2–4 lists the Peripheral PLL output assignments.
Table 2–4. Peripheral PLL Output Assignments

C0 emac0_base_clk Up to 250 MHz No
C1 emac1_base_clk Up to 250 MHz No
C2 periph_qspi_base_clk Up to 432 MHz No

Up to 250 MHz for the
NAND flash controller and
C3 periph_nand_sdmmc_base_clk No
Peripheral up to 200 MHz for the
SD/MMC controller
Up to 240 MHz for the SPI
masters and up to
C4 periph_base_base_clk No
200 MHz for the scan
manager
C5 h2f_user1_base_clk osc1_clk to 100 MHz No

Figure 2–4 shows programmable post-PLL dividers and clock gating for the
peripheral clock group. Clock gate blocks in the diagram indicate clocks which may
be gated off under software control. Software is expected to gate these clocks off prior
to changing any PLL or divider settings that might create incorrect behavior on these
clocks.
Figure 2–4. Peripheral Clock Group Divide and Gating
Peripheral C0 emac0_base_clk
Clock Gate emac0_clk
PLL
emac1_base_clk
C1 Clock Gate emac1_clk
periph_qspi_base_clk
C2 To Flash Controller Clocks
periph_nand_sdmmc_base_clk
C3 To Flash Controller Clocks
periph_base_clk To main PLL group
C4 l4_mp_clk & l4_sp_clk
h2f_user1_base_clk multiplexer
C5 Divide by
1, 2, 4, 8, or 16 Clock Gate usb_mp_clk
Divide by
1, 2, 4, 8, or 16 Clock Gate spi_m_clk
Divide by
1, 2, 4, 8, or 16 Clock Gate can0_clk
Divide by
1, 2, 4, 8, or 16 Clock Gate can1_clk
24-Bit
Divider Clock Gate gpio_db_clk
Clock Gate h2f_user1_clock
Table 2–5 lists the clocks in the peripheral clock group.
Table 2–5. Peripheral Clock Group Clocks (Part 1 of 2)

System Clock Name Frequency Divided From Constraints and Notes
usb_mp_clk Up to 200 MHz Peripheral PLL C4 Clock for USB
Up to 240 MHz for the
SPI masters and up to
spi_m_clk Peripheral PLL C4 Clock for L4 SPI master bus and scan manager
200 MHz for the scan
manager
EMAC0 clock. The 250 MHz clock is divided
internally by the EMAC into the typical
emac0_clk Up to 250 MHz Peripheral PLL C0
125/25/2.5 MHz speeds for 1000/100/10 Mbps
operation.
EMAC1 clock. The 250 MHz clock is divided
internally by the EMAC into the typical
emac1_clk Up to 250 MHz Peripheral PLL C1
125/25/2.5 MHz speeds for 1000/100/10 Mbps
operation.
Main PLL C1 or
l4_mp_clk Up to 100 MHz Clock for L4 master peripheral bus
peripheral PLL C4
Main PLL C1 or
l4_sp_clk Up to 100 MHz Clock for L4 slave peripheral bus
peripheral PLL C4
can0_clk Up to 100 MHz Peripheral PLL C4 Controller area network (CAN) controller 0 clock

Table 2–5. Peripheral Clock Group Clocks (Part 2 of 2)

can1_clk Up to 100 MHz Peripheral PLL C4 CAN controller 1 clock
gpio_db_clk Up to 32 KHz Peripheral PLL C4 Used to debounce GPIO0, GPIO1, and GPIO2
h2f_user1_clock Peripheral PLL C5 Peripheral PLL C5 Auxiliary user clock to the FPGA fabric
SDRAM Clock Group

The SDRAM clock group consists of a PLL and clock gating. The clocks in the SDRAM
clock group are derived from the SDRAM PLL. The SDRAM PLL can be programmed
to be sourced from the EOSC1 pin, the EOSC2 pin, or the f2h_sdram_ref_clk clock
provided by the FPGA fabric.
The counter outputs from the SDRAM PLL can be gated off directly under software
control. The divider values for each clock are set by registers in the clock manager.
Table 2–6 lists the SDRAM PLL output assignments.
Table 2–6. SDRAM PLL Output Assignments

C0 ddr_dqs_base_clk Up to varies (1) Yes
C1 ddr_2x_dqs_base_clk Up to ddr_dqs_base_clk x 2 Yes
SDRAM
C2 ddr_dq_base_clk Up to ddr_dqs_base_clk Yes
C5 h2f_user2_base_clk osc1_clk to varies (1) Yes

(1) The maximum frequency depends on the speed grade of the device.
Figure 2–5 shows clock gating for SDRAM PLL clock group. Clock gate blocks in the
diagram indicate clocks which may be gated off under software control. Software is
expected to gate these clocks off prior to changing any PLL or divider settings that
might create incorrect behavior on these clocks.
Figure 2–5. SDRAM Clock Group Divide and Gating
SDRAM ddr_dqs_base_clk
C0 Clock Gate ddr_dqs_clk
PLL
ddr_2x_dqs_base_clk
C1 Clock Gate ddr_2x_dqs_clk
ddr_dq_base_clk
C2 Clock Gate ddr_dq_clk
C3 Unused
C4 Unused
h2f_user2_base_clk
C5 Clock Gate h2f_user2_clock
The SDRAM PLL output clocks can be phase shifted in real time in increments of 1/8
the VCO frequency. Maximum number of phase shift increments is 4096.

Table 2–7 lists the clocks in the SDRAM clock group.
Table 2–7. SDRAM Clock Group Clocks

Name Frequency Constraints and Notes
ddr_dqs_clk SDRAM PLL C0 Clock for MPFE, single-port controller, CSR access, and PHY
ddr_2x_dqs_clk SDRAM PLL C1 Clock for PHY
ddr_dq_clk SDRAM PLL C2 Clock for PHY
h2f_user2_clock SDRAM PLL C5 Auxiliary user clock to the FPGA fabric
Flash Controller Clocks

Flash memory peripherals can be driven by the main PLL, the peripheral PLL, or from
clocks provided by the FPGA fabric, as shown in Figure 2–6.
Figure 2–6. Flash Peripheral Clock Divide and Gating
f2h_periph_ref_clk
main_nand_sdmmc_base_clk Clock Gate sdmmc_clk
nand_x_clk
f2h_periph_ref_clk
main_nand_sdmmc_base_clk Clock Gate Divide by 4 Clock Gate nand_clk
f2h_periph_ref_clk
main_qspi_base_clk Clock Gate qspi_clk
periph_qspi_base_clk
Table 2–8 lists the flash controller clocks.
Table 2–8. Flash Controller Clocks

Peripheral PLL C2, main PLL C3,
qspi_clk Up to 432 MHz Clock for quad SPI, typically 108 and 80 MHz
or f2h_periph_ref_clk
Peripheral PLL C3, main PLL C4,
nand_x_clk Up to 250 MHz NAND flash controller master and slave clock
or f2h_periph_ref_clk
Peripheral PLL C3, main PLL C4, Main clock for NAND flash controller, sets
nand_clk nand_x_clk/4
or f2h_periph_ref_clk base frequency for NAND transactions
■ Less than or equal to memory maximum
operating frequency
Peripheral PLL C3, main PLL C4, ■ 45% to 55% duty cycle
sdmmc_clk Up to 200 MHz
or f2h_periph_ref_clk ■ Typical frequencies are 26 and 52 MHz
■ SD/MMC has a subclock tree divided down
from this clock

Resets
Cold Reset
Cold reset places the hardware-managed clocks into safe mode, the software-
managed clocks into their default state, and asynchronously resets all registers in the
clock manager.
For more information, refer to “Safe Mode”.
Warm Reset
Registers in the clock manager control how the clock manager responds to warm
reset. Typically, software places the clock manager into a safe state in order to generate
a known set of clocks for the ROM code to boot the system. The behavior of the
system on warm reset as a whole, including how the FPGA fabric, boot code, and
debug systems are configured to behave, must be carefully considered when choosing
how the clock manager responds to warm reset.
The reset manager can request that the clock manager go into safe mode as part of the
reset manager’s warm reset sequence. Before asserting safe mode to the clock
manager, the reset manager ensures that the reset signal is asserted on all modules
that receive warm reset.
f For more information, refer to “Reset Sequencing” in the Reset Manager chapter in
Safe Mode
Safe mode is enabled in the HPS by the assertion of a safe mode request from the reset
manager or by a cold reset. Assertion of the safe mode request from the reset manager
sets the safe mode bit in the clock manager control register. No other control register
bits are affected by the safe mode request from the reset manager.
When safe mode is enabled, the main PLL hardware-managed clocks (C0-C2) are
bypassed to the osc1_clk clock and are directly generated from the osc1_clk clock.
While in safe mode, clock manager register settings, which control clock behavior, are
not changed. However, the hardware bypasses these settings and uses safe, default
settings.
The hardware-managed clocks are forced to their safe mode values such that the
following conditions occur:
■ The hardware-managed clocks are bypassed to the osc1_clk clock, including
counters in the main PLL.
■ Programmable dividers select the reset default values.
■ The flash controller clocks multiplexer selects the output from the peripheral PLL.
■ All clocks are enabled.
A write by software is the only way to clear the safe mode bit (safemode) of the ctrl
register.

1 Before coming out of safe mode, all registers and clocks need to be configured
correctly. It is possible to program the clock manager in such a way that only a cold
reset can return the clocks to a functioning state. Altera strongly recommends using
Altera-provided libraries to configure and control HPS clocks.
Interrupts
The clock manger provides one interrupt output, enabled using the interrupt enable
register (intren). The source of the interrupt is six inputs, namely, an achieving lock
and a losing lock bit in the interrupt status register (inter) for each PLL.
Clock Usage By Module

Table 2–9 lists every clock input generated by the clock manager to all modules in the
HPS. System clock names are global for the entire HPS and system clocks with the
same name are phase-aligned at all endpoints.
Table 2–9. Clock Usage By Module (Part 1 of 3)

Module Name System Clock Name Use
mpu_clk Main clock for the MPU subsystem
mpu_periph_clk Clock for peripherals inside the MPU subsystem
dbg_at_clk Trace bus clock
MPU subsystem dbg_clk Debug clock
Clock for the L2 cache and Accelerator Coherency Port
mpu_l2_ram_clk
(ACP) ID mapper
l4_mp_clk Clock for the ACP ID mapper control slave
l3_main_clk Clock for the L3 main switch
Clock for the System Trace Macrocell (STM) slave and
dbg_at_clk
Embedded Trace Router (ETR) master connections
dbg_clk Clock for the DAP master connection
l3_mp_clk Clock for the L3 master peripheral switch
Clock for the L4 MP bus, Secure Digital (SD) /
l4_mp_clk
MultiMediaCard (MMC) master, and EMAC masters
usb_mp_clk Clock for the USB masters and slaves
nand_x_clk Clock for the NAND master
Interconnect
cfg_clk Clock for the FPGA manager configuration data slave
l3_sp_clk Clock for the L3 slave peripheral switch
l3_main_clk Clock for the L4 SPIS bus master
Clock for the ACP ID mapper slave and L2 master
mpu_l2_ram_clk
connections
osc1_clk Clock for the L4 OSC1 bus master
spi_m_clk Clock for the L4 SPIM bus master
l4_sp_clk Clock for the L4 SP bus master
l4_mp_clk Clock for the quad SPI bus slave
Boot ROM l3_main_clk Clock for the boot ROM
On-chip RAM l3_main_clk Clock for the on-chip RAM


l4_main_clk Clock for the DMA
DMA controller dbg_at_clk Clock synchronous to the STM module
l4_mp_clk Clock synchronous to the quad SPI flash
Clock for the control block (CB) data interface and
cfg_clk
FPGA manager configuration data slave
l4_mp_clk Clock for the control slave
l3_main_clk Clock for the data slave
HPS-to-FPGA bridge
l4_mp_clk Clock for the global programmer's view (GPV) slave
l3_main_clk Clock for the data master
FPGA-to-HPS bridge
l4_mp_clk Clock for the GPV slave
Lightweight HPS-to-FPGA
l4_mp_clk Clock for the GPV masters, and the data and GPV slave
bridge
l4_mp_clk Clock for the control slave
Quad SPI flash controller
qspi_clk Reference clock for serialization
l4_mp_clk Clock for the master and slave
SD/MMC controller
sdmmc_clk Clock for the SD/MMC internal logic
l4_mp_clk Clock for the master
EMAC 0 emac0_clk EMAC 0 internal logic clock
osc1_clk IEEE 1588 timestamp clock
l4_mp_clk Clock for the master
EMAC 1 emac1_clk EMAC 1 internal logic clock
osc1_clk IEEE 1588 timestamp clock
USB 0 usb_mp_clk Clock for the master and slave
USB 1 usb_mp_clk Clock for the master and slave
nand_x_clk NAND high-speed master and slave clock
NAND flash controller
nand_clk NAND flash clock
OSC1 timer 0 osc1_clk Clock for the OSC1 timer 0
OSC1 timer 1 osc1_clk Clock for the OSC1 timer 1
SP timer 0 l4_sp_clk Clock for the SP timer 0
SP timer 1 l4_sp_clk Clock for the SP timer 1
I2C controller 0 l4_sp_clk Clock for the I2C 0
UART controller 0 l4_sp_clk Clock for the UART 0
UART controller 1 l4_sp_clk Clock for the UART 1
l4_sp_clk Clock for the slave
CAN controller 0
can0_clk CAN 0 controller clock
CAN controller 1
can1_clk CAN 1controller clock

Clock Manager Address Map and Register Definitions

l4_mp_clk Clock for the slave
GPIO interface 0
gpio_db_clk Debounce clock
GPIO interface 1
GPIO interface 2
System manager osc1_clk Clock for the system manager
l4_sp_clk Clock for the control slave
ddr_dq_clk Off-chip data clock
ddr_dqs_clk Clock for the MPFE, single-port controller, CSRs, and PHY
SDRAM subsystem
ddr_2x_dqs_clk Off-chip strobe data clock
mpu_l2_ram_clk Clock for the slave connected to MPU subsystem L2 cache
l3_main_clk Clock for the slave connected to L3 interconnect
L4 watchdog timer 0 osc1_clk Clock for the L4 watchdog timer 0
L4 watchdog timer 1 osc1_clk Clock for the L4 watchdog timer 1
SPI master controller 0 spi_m_clk Clock for the SPI master 0
SPI master controller 1 spi_m_clk Clock for the SPI master 1
SPI slave controller 0 l4_main_clk Clock for the SPI slave 0
SPI slave controller 1 l4_main_clk Clock for the SPI slave 1
l4_mp_clk System bus clock
dbg_clk Debug clock
Debug subsystem
dbg_at_clk Trace bus clock
dbg_trace_clk Trace port clock
osc1_clk Clock for the reset manager
Reset manager
Scan manager spi_m_clk Clock for the scan manager
Timestamp generator dbg_timer_clk Clock for the timestamp generator
Clock Manager Address Map and Register Definitions

f The address map and register definitions reside in the hps.html file that accompanies
this handbook volume. Click the link to open the file.
To view the module description and base address, scroll to and click the link for the
following module instance:
■ clkmgr
To then view the register and field descriptions, scroll to and click the register names.
The register addresses are offsets relative to the base address of each module instance.
f The base addresses of all modules are also listed in the Introduction to the Hard
Processor System chapter in volume 3 of the Cyclone V Device Handbook.



■ Reorganized and expanded functional description section.
May 2012 1.1
■ Added address map and register definitions section.


3. Reset Manager
November 2012
cv_54003-1.2
cv_54003-1.2
This chapter provides an overview of HPS reset manager. The reset manager
generates module reset signals based on reset requests from the various sources in the
HPS and FPGA fabric, and software writing to the module-reset control registers. The
reset manager ensures that a reset request from the FPGA fabric can occur only after
the FPGA portion of the system-on-a-chip (SoC) device is configured.
The HPS contains three reset domains. Each reset domain can be reset independently.
Each register in the HPS that can be reset belongs to one particular reset domain.
Table 3–1 shows the HPS reset domains.
Table 3–1. HPS Reset Domains

Domain Name Domain Logic
JTAG test access port (TAP) controller, which is used by the debug access
TAP
port (DAP).
All debug logic including most of the DAP, CoreSight™ components
Debug connected to the debug peripheral bus, trace, the microprocessor unit (MPU)
subsystem, and the FPGA fabric.
All HPS logic except what is in the TAP and debug reset domains. Includes
System
nondebug logic in the FPGA fabric connected to the HPS reset signals.
The HPS supports the following reset types:

■ Cold reset (power-on reset)
■ Used to ensure the HPS is placed in a default state sufficient for software to
boot
■ Triggered by a power-on reset and other sources
■ Resets all HPS logic that can be reset
■ Affects all reset domains
■ Warm reset
■ Occurs after HPS has already been through a cold reset
■ Used to recover system from a non-responsive condition
■ Resets a subset of the HPS state reset by a cold reset
■ Only affects the system reset domain, which allows debugging (including
trace) to operate through the warm reset

November 2012
Subscribe
3–2 Chapter 3: Reset Manager
Reset Manager Block Diagram and System Integration
■ Debug reset
■ Occurs after HPS has already been through a cold reset
■ Used to recover debug logic from a non-responsive condition
■ Only affects the debug reset domain

Figure 3–1 shows a block diagram of the reset manager in the SoC device.
Figure 3–1. Reset Manager Block Diagram
FPGA Portion
FPGA Fabric
f2h_dbg_rst_req_n
Control
Block f2h_cold_rst_req_n h2f_rst_n
f2h_warm_rst_req_n h2f_cold_rst_n
load_csr usermode
HPS
Reset Manager
fpga_config_complete HPS
Scan Manager Reset Request Modules
Scan Manager
nPOR
nRST
Watchdog Reset Request[1:0] Signal Module
MPU Reset
Assertion / Reset
Controller
De-Assertion Signals
Debug Reset Request
DAP
POR Voltage Reset Request

POR Voltage
Detector (mpumodrst,
permodrst,
per2modrst,
System Watchdog Reset Request[1:0]
brgmodrst,
Watchdog (2)
(swcoldrstreq and and
swwarmrstreq bits of ctrl) miscmodrst)
CSRs
Slave Interface
L4 Peripheral Bus (osc1_clk)
Note to Figure 3–1:

(1) Reset-related handshaking signals to other HPS modules and to the clock manager module are omitted from this figure for clarity.

Chapter 3: Reset Manager 3–3
HPS External Reset Sources

Table 3–2 describes the reset sources external to the HPS. All signals are synchronous
to the osc1_clk clock.
Table 3–2. HPS External Reset Sources

Source Description
f2h_cold_rst_req_n Cold reset request from FPGA fabric (active low)
f2h_warm_rst_req_n Warm reset request from FPGA fabric (active low)
f2h_dbg_rst_req_n Debug reset request from FPGA fabric (active low)
h2f_cold_rst_n Cold-only reset to FPGA fabric (active low)
h2f_rst_n Cold or warm reset to FPGA fabric (active low)
Cold-only reset from FPGA control block (CB) and scan
load_csr
manager
nPOR Power-on reset pin (active low)
nRST Warm reset pin (active low)
1 The reset signals from the HPS to the FPGA fabric must be synchronized to your user
logic clock domain.
Reset Controller
The reset controller performs the following functions:
■ Accepts reset requests from the FPGA CB, FPGA fabric, modules in the HPS, and
reset pins
■ Generates an individual reset signal for each module instance for all modules in
the HPS
■ Provides reset handshaking signals to support system reset behavior
The reset controller generates module reset signals from external reset requests and
internal reset requests. External reset requests originate from sources external to the
reset manager. Internal reset requests originate from control registers in the reset
manager.

Figure 3–2 shows the reset controller signals.
Figure 3–2. Reset Controller Signals
Reset Controller
POR Voltage Monitor Many Signals (refer to Table 3-3)
nPOR Pin Cold
Reset Module Resets
FPGA Fabric (f2h_cold_rst_req_n) Requests
FPGA CB & Scan Manager
Debug Domain
dbg_rst_n
Reset
nRST Pin
Warm TAP Domain
FPGA Fabric (f2h_warm_rst_req_n) JTAG TAP (DAP)
Reset Reset
MPU Watchdog Reset [1:0]
Requests MPU Clock
System Watchdog Reset [1:0] Gating CPUCLKOFF[1:0]
CDBGRSTREQ (DAP) Debug

Reset
FPGA Fabric (f2h_dbg_rst_req_n)
Requests
CDBGRSTACK (DAP)
ETR ETR
Reset
SDRAM Self-Refresh Reset SDRAM Self-Refresh
Handshaking
FPGA Manager Handshaking Outputs FPGA Manager
SCAN Manager Inputs SCAN Manager
FPGA Fabric FPGA
Reset Manager osc1_clk

APB Slave Interface
The reset controller supports the following cold reset requests:

■ Power-on reset (POR) voltage monitor
■ Cold reset request pin (nPOR)
■ FPGA fabric
■ FPGA CB and scan manager
■ Software cold reset request bit (swcoldrstreq) of the control register (ctrl)
The reset controller supports the following warm reset requests:
■ Warm reset request pin (nRST)
■ FPGA fabric
■ Software warm reset request bit (swwarmrstreq) of the ctrl register
■ MPU watchdog reset requests for CPU0 and CPU1
■ System watchdog timer 0 and 1 reset requests
The reset controller supports the following debug reset requests:
■ CDBGRSTREQ from DAP
■ FPGA fabric

Module Reset Signals

Table 3–3 lists the module reset signals. The module reset signals are organized in
groups for the MPU, peripherals, bridges, the level 3 (L3) interconnect, and
miscellaneous.
Checkmarks in the Cold Reset, Warm Reset, and Debug Reset columns denote reset
signals asserted by each type of reset. For example, writing a 1 to the swwarmrstreq bit
in the ctrl register resets all the modules that have a checkmark in the Warm Reset
column.
Checkmarks in the Software Deassert column denote reset signals that are left
asserted by the reset manager. To activate the related modules, software can deassert
these reset signals as needed by writing to the following reset manager registers:
■ MPU module reset register (mpumodrst)
■ Peripheral module reset register (permodrst)
■ Peripheral 2 module reset register (per2modrst)
■ Bridge module reset register (brgmodrst)
Table 3–3. Generated Module Resets (Part 1 of 3)

Reset Cold Warm Debug Software
Group Module Reset Signal Description
Domain Reset Reset Reset Deassert
mpu_cpu_rst_n[0] Resets each processor in the MPU System v v
mpu_cpu_rst_n[1] Resets each processor in the MPU System v v v
Resets both per-processor
mpu_wd_rst_n System v v
MPU watchdogs in the MPU
Resets Snoop Control Unit (SCU)
mpu_scu_periph_rst_n System v v
and peripherals
mpu_l2_rst_n Level 2 (L2) cache reset System v v


emac_rst_n[1:0] Resets each EMAC System v v v
usb_rst_n[1:0] Resets each USB System v v v
nand_flash_rst_n Resets NAND flash controller System v v v
qspi_flash_rst_n Resets quad SPI flash controller System v v v
Resets each system watchdog
watchdog_rst_n[1:0] System v v v
timer
osc1_timer_rst_n[1:0] Resets each OSC1 timer System v v v
sp_timer_rst_n[1:0] Resets each SP timer System v v v
i2c_rst_n[3:0] Resets each I2C controller System v v v
PER uart_rst_n[1:0] Resets each UART System v v v
spim_rst_n[1:0] Resets SPI master controller System v v v
spis_rst_n[1:0] Resets SPI slave controller System v v v
sdmmc_rst_n Resets SD/MMC controller System v v v
can_rst_n[1:0] Resets each CAN controller System v v v
gpio_rst_n[2:0] Resets each GPIO interface System v v v
dma_rst_n Resets DMA controller System v v v
Resets SDRAM subsystem (resets
sdram_rst_n logic associated with cold or warm System v v v
reset)
DMA controller request interface
PER2 dma_periph_if_rst_n[7:0] from FPGA fabric to DMA System v v v
controller
Resets HPS-to-FPGA AMBA®
hps2fpga_bridge_rst_n Advanced eXtensible Interface System v v v
(AXI™) bridge
Bridge
fpga2hps_bridge_rst_n Resets FPGA-to-HPS AXI bridge System v v v
Resets lightweight HPS-to-FPGA
lwhps2fpga_bridge_rst_n System v v v
AXI bridge


boot_rom_rst_n Resets boot ROM System v v
onchip_ram_rst_n Resets on-chip RAM System v v
Resets system manager (resets
sys_manager_rst_n logic associated with cold or warm System v v
reset)
Resets system manager (resets
sys_manager_cold_rst_n logic associated with cold reset System v
only)
fpga_manager_rst_n Resets FPGA manager System v v
acp_id_mapper_rst_n Resets ACP ID mapper System v v
Resets user logic in FPGA fabric
h2f_rst_n (resets logic associated with cold System v v
or warm reset)
Resets user logic in FPGA fabric
h2f_cold_rst_n (resets logic associated with cold System v
reset only)
rst_pin_rst_n Pulls nRST pin low System v v
timestamp_cold_rst_n Resets debug timestamp to 0x0 System v
MISC
Resets clock manager (resets logic
clk_manager_cold_rst_n System v
associated with cold reset only)
scan_manager_rst_n Resets scan manager System v v
Resets freeze controller (resets
frz_ctrl_cold_rst_n logic associated with cold reset System v
only)
Resets debug masters and slaves
sys_dbg_rst_n connected to L3 interconnect and System v v
level 4 (L4) buses
Resets debug components
including DAP, trace, MPU debug
dbg_rst_n Debug v v
logic, and any user debug logic in
the FPGA fabric
Resets portion of TAP controller in
tap_cold_rst_n the DAP that must be reset on a TAP v
cold reset
Resets SDRAM subsystem (resets
sdram_cold_rst_n logic associated with cold reset System v
only)
Resets L3 interconnect and L4
L3 l3_rst_n System v v
buses
Slave Interface and Status Register

The reset manager slave interface is used to control and monitor the reset states.

Functional Description of the Reset Manager
The status register (stat) in the reset manager contains the status of the reset
requester. The register contains a bit for each reset request. The stat register captures
all reset requests that have occurred. Software is responsible for clearing the bits.

The reset manager generates reset signals to modules in the HPS and to the FPGA
fabric. The following actions generate reset signals:
■ Software writing a 1 to the swcoldrstreq or swwarmrstreq bits in the ctrl register.
Writing either bit causes the reset controller to perform a reset sequence.
■ Software writing to the mpumodrst, permodrst, per2modrst, brgmodrst, or
miscmodrst module reset control registers.
■ Asserting reset request signals triggers the reset controller. All external reset
requests cause the reset controller to perform a reset sequence.
f For information about the required duration of reset request signal

assertion, refer to the Cyclone V Device Datasheet.
Multiple reset requests can be driven to the reset manager at the same time. Cold reset
requests take priority over warm and debug reset requests. Higher priority reset
requests preempt lower priority reset requests. There is no priority difference among
reset requests within the same domain.
If a cold reset request is issued while another cold reset is already underway, the reset
manager extends the reset period for all the module reset outputs until all cold reset
requests are removed. If a cold reset request is issued while the reset manager is
removing other modules out of the reset state, the reset manager returns those
modules back to the reset state.
If a warm reset request is issued while another warm reset is already underway, the
first warm reset completes before the second warm reset begins. If the second warm
reset request is removed before the first warm reset completes, the warm first reset is
extended to meet the timing requirements of the second warm reset request.
The nPOR pin can be used to extend the cold reset beyond what the POR voltage
monitor automatically provides. The use of the nPOR pin is optional and can be tied
high when it is not required.
Reset Sequencing
The reset controller sequences resets without software assistance. Module reset
signals are asserted asynchronously at the same time. The reset manager deasserts the
module reset signals synchronous to the osc1_clk clock. Module reset signals are
deasserted in groups in a fixed sequence. All module reset signals in a group are
deasserted at the same time.
The reset manager sends a safe mode request to the clock manager to put the clock
manager in safe mode, which creates a fixed and known relationship between the
osc1_clk clock and all other clocks generated by the clock manager.
f For information about safe mode options, refer to the Clock Manager chapter in

After the reset manager releases the MPU subsystem from reset, CPU1 is left in reset
and CPU0 begins executing code from the reset vector address. Software is
responsible for deasserting CPU1 and other resets, as shown in Table 3–3. Software
deasserts resets by writing the mpumodrst, permodrst, per2modrst, brgmodrst, and
miscmodrst module-reset control registers.
Software can also bypass the reset controller and generate reset signals directly
through the module-reset control registers. In this case, software is responsible for
asserting module reset signals, driving them for the appropriate duration, and
deasserting them in the correct order. The clock manager is not typically in safe mode
during this time, so software is responsible for knowing the relationship between the
clocks generated by the clock manager. Software must not assert a module reset signal
that would prevent software from deasserting the module reset signal. For example,
software should not assert the module reset to the processor executing the software.
Figure 3–3 shows the timing diagram for cold reset.
Figure 3–3. Cold Reset Timing Diagram
nPOR pin (1) (2)
clk_mgr_cold_rst_n
l3_rst_n
miscmod_rst_n
dbg_rst_n
mpu_clkoff
mpu_rst_n[0]
mpu_wd_rst_n
mpu_scu_rst_n
mpu_periph_rst_n
mpu_l2_rst_n
Software
peripheral resets brings out
of reset
32 96 100 200 32 32

(1) Cold reset can be initiated from several other sources. For the complete list, refer to “Reset Controller” on page 3–3.
(2) This dependency applies to all the reset signals.

Figure 3–4 shows the timing diagram for warm reset.
Figure 3–4. Warm Reset Timing Diagram
nRST pin (1)
h2f_pending_rst_req_n (2)
(and other wait
request handshakes)
f2h_pending_rst_ack_n
safe_mode_req
cm_rm_safe_mode_ack
l3_rst_n
miscmod_rst_n
mpu_clkoff
mpu_rst_n[0]
mpu_wd_rst_n
mpu_scu_rst_n
mpu_periph_rst_n
mpu_l2_rst_n
Software
peripheral resets brings out
of reset
8 nRST Pin Count (3) 256 (3) 100 200 32 32

(1) Warm reset can be initiated from several other sources. For the complete list, refer to “Reset Controller” on page 3–3.
(2) For information about the wait request reset handshaking, refer to “Reset Handshaking” on page 3–13.
(3) When the nRST pin count is zero, the 256 cycle stretch count is skipped and the start of the deassertion sequence is determined by the safe mode
acknowledge signal or the user releasing the warm reset button, whichever occurs later.
The cold and warm reset sequences consist of different reset assertion sequences and
the same deassertion sequence. The following sections describe the sequences.
Cold Reset Assertion Sequence

The following list describes the assertion steps for cold reset shown in Figure 3–3:
1. Assert module resets.
2. Wait for 32 cycles. Deassert clock manager cold reset.
3. Wait for 96 cycles (so clocks can stabilize).
4. Proceed to the “Cold and Warm Reset Deassertion Sequence”.
Warm Reset Assertion Sequence

The following list describes the assertion steps for warm reset shown in Figure 3–4:

1. Optionally, handshake with the embedded trace router (ETR) and wait for
acknowledge.
2. Optionally, handshake with the FPGA fabric and wait for acknowledge.
3. Optionally, handshake with the SDRAM controller, scan manager, and FPGA
manager, and wait for acknowledges.
4. Assert module resets (except the MPU watchdog timer resets when the MPU
watchdog timers are the only request sources).
5. Wait for 8 cycles and send a safe mode request to the clock manager.
6. Wait for the greater of the nRST pin count + 256 stretch count, or the warm reset
counter, or the clock manager safe mode acknowledge, then deassert all
handshakes except warm reset ETR handshake (which is deasserted by software).
7. Proceed to the “Cold and Warm Reset Deassertion Sequence”.
Cold and Warm Reset Deassertion Sequence

The following list describes the deassertion steps for both cold and warm reset shown
in Figure 3–3 and Figure 3–4:
1. Deassert L3 reset.
2. Wait for 100 cycles. Deassert resets for miscellaneous-type and debug (cold only)
modules.
3. Wait for 200 cycles. Assert mpu_clkoff for CPU0 and CPU1.
4. Wait for 32 cycles. Deassert resets for MPU modules.
5. Wait for 32 cycles. Deassert mpu_clkoff for CPU0 and CPU1.
6. Peripherals remain held in reset until software brings them out of reset.
Reset Pins
Figure 3–5 shows all HPS pins related to reset.
Figure 3–5. Reset Pins
SoC Device
HPS
nTRST
TMS ARM DAP
TCK
nPOR
nRST Reset Manager

rst_pin_rst_n

The test reset (nTRST), test mode select (TMS), and test clock (TCK) pins are associated
with the TAP reset domain and are used to reset the TAP controller in the DAP. These
pins are not connected to the reset manager.
The nPOR and nRST pins are used to request cold and warm resets respectively. The
nRST pin is an open drain output as well. Any warm reset request causes the reset
manager to drive the rst_pin_rst_n signal output low, which drives the nRST pin low.
The amount of time the reset manager pulls nRST low is controlled by the nRST pin
count field (nrstcnt) of the reset cycles count register (counts). This technique can be
used to reset external devices (such as external memories) connected to the HPS.
Reset Effects
The following list describes how reset affects HPS logic:
■ The TAP reset domain ignores warm reset.
■ The debug reset domain ignores warm reset.
■ System reset domain cold resets ignore warm reset.
■ Each module defines reset behavior individually.
1 For more information, refer to the individual chapters in volume 3 of the

Altering Warm Reset System Response

Registers in the clock manager, system manager, and reset manager control how
warm reset affects the HPS. You can control the impact of a warm reset on the clocks
and I/O elements.
1 Altera strongly recommends using Altera-provided libraries to configure and control

this functionality.
The default warm reset behavior takes all clocks and I/O elements through a cold
reset response. As your software becomes more stable or for debug purposes, you can
alter the system response to a warm reset. The following suggestions provide ways to
alter the system response to a warm reset. None of the register bits that control these
items are affected by warm reset.
■ Boot from on-chip RAM—enables warm boot from on-chip RAM instead of the
boot ROM. When enabled, the boot ROM code validates the RAM code and jumps
to it, making no changes to clocks or any other system settings prior to executing
user code from on-chip RAM.
■ Disable safe mode on warm reset—allows software to transition through a warm
reset without affecting the clocks. Because the boot ROM code indirectly
configures the clock settings after warm reset, Altera recommends to only disable
safe mode when the HPS is not booting from a flash device.

Reset Manager Address Map and Register Definitions
■ Disable safe mode on warm reset for the debug clocks—keeps the debug clocks
from being affected by the assertion of safe mode request on a warm reset. This
technique allows fast debug clocks, such as trace, to continue running through a
warm reset. When enabled, the clock manager puts the debug clocks to their safe
frequencies to respond to a safe mode request from the reset manager on a warm
reset. Disable safe mode on warm reset for the debug clocks only when you are
running the debug clocks off the main PLL VCO and you are certain the main PLL
cannot be impacted by the event which caused the warm reset.
■ Use the osc1_clk clock for debug control—keeps the debug base clock (main PLL
C2 output) always bypassed to the osc1_clk external clock, independent of other
clock manager settings. When implemented, disabling safe mode on warm reset
for the debug clocks has no effect.
f For more information about safe mode, refer to the Clock Manager chapter in volume 3
Reset Handshaking
The reset manager participates in several reset handshaking protocols to ensure other
modules are safely reset.
Before issuing a warm reset, the reset manager performs a handshake with several
modules to allow them to prepare for a warm reset. The handshake logic ensures the
following conditions:
■ ETR master has no pending master transactions to the L3 interconnect
■ Optionally preserve SDRAM contents during warm reset by issuing self-refresh
mode request
■ FPGA manager stops generating configuration clock
■ Scan manager stops generating JTAG and I/O configuration clocks
■ Warns the FPGA fabric of the forthcoming warm reset
Similarly, the handshake logic associated with ETR also occurs during the debug reset
to ensure that the ETR master has no pending master transactions to the L3
interconnect before the debug reset is issued. This action ensures that when ETR
undergoes a debug reset, the reset has no adverse effects on the system domain
portion of the ETR.
Reset Manager Address Map and Register Definitions

To view the module description and base address, scroll to and click the link for any of
the following module instances:
■ rstmgr



■ Added cold and warm reset timing diagrams.
November 2012 1.2
■ Minor updates.
Added reset controller, functional description, and address map and register definitions
May 2012 1.1
sections.

Section II. System Interconnect

■ Chapter 4, Interconnect
■ Chapter 5, HPS-FPGA AXI Bridges

II–2 Section II: System Interconnect

4. Interconnect
November 2012
cv_54004-1.2
cv_54004-1.2
The hard processor system (HPS) level 3 (L3) interconnect and level 4 (L4) peripheral
buses are implemented with the ARM® CoreLink™ Network Interconnect (NIC-301).
The NIC-301 provides a foundation for a high-performance HPS interconnect based
on the ARM Advanced Microcontroller Bus Architecture (AMBA®) Advanced
eXtensible Interface (AXI™), Advanced High-Performance Bus (AHB™), and
Advanced Peripheral Bus (APB™) protocols. The L3 interconnect implements a
multilayer, nonblocking architecture that supports multiple simultaneous
transactions between masters and slaves, including the Cortex™-A9 microprocessor
unit (MPU) subsystem. The interconnect provides five independent L4 buses to access
control and status registers (CSRs) of peripherals, managers, and memory controllers
f Additional information is available in the AMBA Network Interconnect (NIC-301)

Technical Reference Manual, which you can download from the ARM website
(infocenter.arm.com).
Features of the Interconnect

The L3 interconnect has the following characteristics:
■ Main internal data width of 64 bits
■ Programmable master priority with single-cycle arbitration
■ Full pipelining to prevent master stalls
■ Programmable control for FIFO buffer transaction release
■ Security of the following types:
■ Secure
■ Nonsecure
■ Per transaction security
■ Five independent L4 buses
trademarks or service marks are the property of their respective holders as described at www.altera.com/common/legal.html. Altera warrants performance of its
semiconductor products to current specifications in accordance with Altera's standard warranty, but reserves the right to make changes to any products and
services at any time without notice. Altera assumes no responsibility or liability arising out of the application or use of any information, product, or service
Portions © 2011 ARM Limited. Used with permission. All rights reserved. ARM, the ARM Powered logo, AMBA, Jazelle, StrongARM, Thumb, and TrustZone are ISO
registered trademarks of ARM Limited. The ARM logo, Angel, ARMulator, AHB, APB, ASB, ATB, AXI, CoreSight, Cortex, EmbeddedICE, ModelGen, MPCore, 9001:2008
Multi-ICE, NEON, PrimeCell, ARM7TDMI, ARM7TDMI-S, ARM9TDMI, ARM9E-S, ARM966E-S, ETM7, ETM9, TDMI and STRONG are trademarks of ARM Registered
Limited. All other products or services mentioned herein may be trademarks of their respective owners. Neither the whole nor any part of the information contained
in, or the product described in, this document may be adapted or reproduced in any material form except with the prior written permission of the copyright holder.
The product described in this document is subject to continuous developments and improvements. All particulars of the product and its use contained in this
document are given by ARM in good faith. However, all warranties implied or expressed, including but not limited to implied warranties of merchantability, or
fitness for purpose, are excluded. This document is intended only to assist the reader in the use of the product. ARM Limited shall not be liable for any loss or
damage arising from the use of any information in this document, or any error or omission in such information, or any incorrect use of the product. Where the term
ARM is used it means “ARM or any of its subsidiaries as appropriate”. This document is Non-Confidential. The right to use, copy and disclose this document may
be subject to license restrictions in accordance with the terms of the agreement entered into by ARM and the party that ARM delivered this document to. The
information in this document is final, that is for a developed product.

November 2012
Subscribe
4–2 Chapter 4: Interconnect
Interconnect Block Diagram and System Integration

Figure 4–1 shows a block diagram of the L3 interconnect and L4 buses.
Figure 4–1. Interconnect Block Diagram
S
Lightweight Legend
FPGA FPGA-to-HPS HPS-to-FPGA HPS-to-FPGA Bridge
M: Master
32-Bit APB Bus Manager Bridge Bridge S S: Slave
(L4_MP, l4_mp_clk) S M S (GPV) Switch Connection
32-Bit AXI 64-Bit AXI 64-Bit AXI
(cfg_clk) (l3_main_clk) (l3_main_clk) 32-Bit AXI
GPIO (l4_mp_clk)
(3) L3 Interconnect MPU Subsystem
S M S M
(NIC-301) (GPV) (mpu_clk)
ARM Cortex-A9
DAP MPCore
32-Bit AHB (dbg_clk) CPU0 CPU1
ACP ID
S M S (GPV) 64-Bit AXI Mapper 64-Bit AXI SCU
L3 Master (mpu_l2_ram_clk) (mpu_l2_ram_clk)
ETR M S M S ACP
32-Bit AXI Peripheral M M
(dbg_at_clk) S
M S Switch 64-Bit AXI 64-Bit AXI
l3_mp_clk L3 Main Switch (1) (mpu_clk) (mpu_clk)
SD/MMC l3_main_clk S S
32-Bit AHB 64-Bit AXI
(mpu_l2_ram_clk) L2
(l4_mp_clk) (GPV) S M Cache
S M S M
32-Bit AXI
EMAC (dbg_at_clk)
32-Bit AXI 32-Bit AXI M S STM
(2) (l4_mp_clk) (l3_mp_clk) 32-Bit AXI 64-Bit AXI
S M S M S (l3_main_clk) (mpu_l2_ram_clk)
Boot
M S ROM
USB 32-Bit AHB 64-Bit AXI
OTG (l3_main_clk) On-Chip S
(usb_mp_clk)
S (2) M S M S RAM
32-Bit AXI (l3_main_clk) SDRAM
32-Bit AXI M S M S Controller
NAND DMA
(nand_x_clk) 32-Bit AXI Subsystem
S M S M
(l3_sp_clk)
S 64-Bit AXI
L3 Slave Peripheral Switch S (l4_main_clk) S S
32-Bit AXI
l3_sp_clk
(nand_x_clk) M S
M
32-Bit AXI (nand_x_clk) 32-Bit APB (l4_main_clk)
M M
32-Bit AHB (usb_mp_clk) 32-Bit APB Bus
M
32-Bit APB (l4_mp_clk) SPI Slave (L4_MAIN,
M (2) l4_main_clk)
S S M M M M S
Quad SPI 32-Bit AHB
Flash (l4_mp_clk) 32-Bit APB Bus
32-Bit APB Bus 32-Bit APB Bus
(L4_OSC1, osc1_clk) (L4_SPI_M, spi_m_clk) (L4_SP, I4_sp_clk)
S S S S S S S S S S S
System OSC1 Watchdog Clock Reset Scan SPI SP I 2C UART CAN
Manager Timer (2) (2) Manager Manager Manager Master (2) Timer (2) (4) (2) (2)

(1) For L3 main switch connection details, refer to Table 4–1.
The L3 interconnect is a partially-connected switch fabric; not all masters can access
all slaves. For more information, refer to “Master-to-Slave Connectivity Matrix” on
page 4–6.
Internally, the L3 interconnect is partitioned into the following subswitches:
■ L3 main switch
■ Main switch used to transfer high-throughput 64-bit data
■ Operates at up to half the MPU main clock frequency
■ Provides masters with low-latency connectivity to AXI bridges, on-chip
memories, SDRAM, and FPGA manager

Chapter 4: Interconnect 4–3
■ L3 master peripheral switch

■ Used to connect memory-mastering peripherals to the main switch
■ 32-bit data width
■ Operates at up to half the main switch clock frequency
■ L3 slave peripheral switch
■ Used to provide access to level 3 and 4 slave interfaces for masters of the
master peripheral and main switches
■ 32-bit data width
■ Five independent L4 buses
The L3 master and slave peripheral switches are fully-connected crossbars. The L3
main switch is a partially-connected crossbar. Table 4–1 shows the connectivity matrix
of all the master and slave interfaces of the L3 main switch. Checkmarks denote
connections.
Table 4–1. L3 Main Switch Connectivity Matrix

Slaves
ACP ID Mapper Data
Subsystem L3 Data
Peripheral Switch
SDRAM Controller
FPGA Manager
On-Chip RAM
HPS-to-FPGA
Boot ROM
L3 Slave
Bridge
STM
Masters
L3 Master Peripheral Switch v v v v
L2 Cache Master 0 v v v v v v
FPGA-to-HPS Bridge v v v v v
DMA v v v v v v v
DAP v v v v v v
L3 Masters
The following list contains all of the master interfaces connected to the L3
interconnect:
■ MPU subsystem—L2 cache master 0 connected to the L3 main switch
■ FPGA-to-HPS bridge—Connected to the L3 main switch
■ DMA—Connected to the L3 main switch
■ EMAC0—Connected to the L3 master peripheral switch
■ EMAC1—Connected to the L3 master peripheral switch
■ USB0—Connected to the L3 master peripheral switch
■ USB1—Connected to the L3 master peripheral switch
■ NAND—Connected to the L3 master peripheral switch
■ SD/MMC—Connected to the L3 master peripheral switch

■ ETR—Connected to the L3 master peripheral switch

■ DAP—Connected to the L3 main switch
L3 Slaves
The following list contains all of the slave interfaces connected to the L3 interconnect:
■ USB0—CSR slave interface connected to the L3 slave peripheral switch
■ USB1—CSR slave interface connected to the L3 slave peripheral switch
■ NAND registers—CSR slave interface connected to the L3 slave peripheral switch
■ NAND data—Command and data slave interface connected to the L3 slave
peripheral switch
■ Quad SPI flash—Data slave interface connected to the L3 slave peripheral switch
■ FPGA manager—Data slave interface connected to the L3 main switch
■ HPS-to-FPGA bridge—Data slave interface connected to the L3 main switch
■ Lightweight HPS-to-FPGA bridge—Data slave interface connected to the L3 slave
peripheral switch
■ ACP ID mapper—Data slave interface connected to the L3 main switch
■ STM—Connected to the L3 main switch
■ Boot ROM—Connected to the L3 main switch
■ On-chip RAM—Connected to the L3 main switch
■ SDRAM controller subsystem—SDRAM multi-port front end slave interface
connected to the L3 main switch
L4 Slaves
Each of the L4 slaves is an APB slave connected to one of the five following L4 buses:
■ L4 slave peripheral (SP) bus—APB for peripherals that do not require fast access
■ SDRAM controller subsystem—CSR access
■ SP timer 0—CSR access
■ SP timer 1—CSR access
■ I2C0—CSR access
■ I2C1—CSR access
■ I2C2 (associated with EMAC0)—CSR access
■ I2C3 (associated with EMAC1)—CSR access
■ UART0—CSR access
■ UART1—CSR access
■ CAN0—CSR access
■ CAN1—CSR access

Functional Description of the Interconnect
■ L4 master peripheral (MP) bus—APB that provides access to primarily the L3

master peripherals.
■ ACP ID mapper—CSR access
■ FPGA manager—CSR access
■ DAP—CSR access
■ Quad SPI flash—CSR access
■ SD/MMC—CSR access
■ EMAC0—CSR access
■ EMAC1—CSR access
■ GPIO0—CSR access
■ L4 oscillator 1 (OSC1) bus—APB dedicated to peripherals that operate on the
external oscillator 1 domain.
■ OSC1 timer 0—CSR access
■ OSC1 timer 1—CSR access
■ Watchdog 0—CSR access
■ Watchdog 1—CSR access
■ Clock manager—CSR access
■ Reset manager—CSR access
■ System manager—CSR access
■ L4 main bus—APB dedicated to the DMA and SPI slaves
■ DMA_s—Access to the DMA controllers secure registers
■ DMA_ns—Nonsecure access to the DMA controller nonsecure registers
■ SPI slave 0—CSR access
■ SPI slave 1—CSR access
■ L4 SPI master (SPIM) bus—APB dedicated to the SPI masters and scan manager.
■ SPI master 0—CSR access
■ SPI master 1—CSR access
■ Scan manager—CSR access

This section provides a functional description of the interconnect.

Master-to-Slave Connectivity Matrix

The interconnect is a partially-connected crossbar. Table 4–2 shows the connectivity
matrix of all the master and slave interfaces of the interconnect. Checkmarks denote
connections.
Table 4–2. Interconnect Connectivity Matrix

Slaves
Lightweight HPS-to-FPGA
NAND Command and Data
Quad SPI Flash Data
ACP ID Mapper Data

HPS-to-FPGA Bridge
L4 MAIN Bus Slaves
L4 SPIM Bus Slaves

L4 OSC1 Bus Slaves
Subsystem L3 Data
SDRAM Controller
L4 MP Bus Slaves
L4 SP Bus Slaves
USB OTG 0/1 CSR
FPGA Manager
On-Chip RAM
NAND CSR
Boot ROM
Bridge
STM
Masters
L2 Cache Master 0 v v v v v v v v v v v v v v v
FPGA-to-HPS Bridge v v v v v v v v v v v v v v
DMA v v v v v v v v v v v v v v v v
EMAC 0/1 v v v v
USB OTG 0/1 v v v v
NAND v v v v
SD/MMC v v v v
ETR v v v
DAP v v v v v v v v v v v v v v v
Address Remapping
The interconnect supports address remapping through the remap register. Remapping
allows software to control which memory device (SDRAM, on-chip RAM, or boot
ROM) is accessible at address 0x0 and the accessibility of the HPS-to-FPGA and
lightweight HPS-to-FPGA bridges. The remap register is one of the NIC-301 Global
Programmers View (GPV) registers and maps into the address space of the following
L3 masters:
■ MPU
■ FPGA-to-HPS bridge
■ DAP
The remapping bits in the remap register are not mutually exclusive. The lowest order
remap bit has higher priority when multiple slaves are remapped to the same address.
Each bit allows different combinations of address maps to be formed. There is only
one remapping register available in the GPV, so modifying the remap register affects
all memory maps of all the masters of the interconnect.
The effects of the remap bits can be categorized in the following groups:
■ MPU master interface
■ L2 cache master 0 interface

■ Non-MPU master interfaces

■ DMA master interface
■ Master peripheral interfaces
■ Debug Access Port (DAP) master interface
■ FPGA-to-HPS bridge master interface
Figure 4–2 shows the interconnect address map for all MPU and non-MPU masters.
The figure is not to scale.
Figure 4–2. Address Map per Master
0xFFFF_FFFF
0xFFFF_0000 On-Chip RAM On-Chip RAM
0xFFFE_C000 SCU & L2 Registers (1)
Boot ROM
0xFFFD_0000
Peripherals & Peripherals &
L3 GPV L3 GPV (5)
0xFF40_0000
Lightweight FPGA Slaves (2) Lightweight FPGA Slaves (2)
0xFF20_0000
DAP DAP (5)
0xFF00_0000
STM STM (5), (6)
0xFC00_0000
FPGA Slaves (2) FPGA Slaves (2)
0xC000_0000
ACP
0x8000_0000
SDRAM (3)
SDRAM
0x0010_0000
0x0001_0000
0x0000_0000 Boot ROM or On-Chip RAM (4) On-Chip RAM or SDRAM (4)
MPU non-MPU
(1) The SCU and L2 cache registers are located in the MPU subsystem and are not accessible from the L3 interconnect.
(2) This address range is not always accessible. For more information, refer to Table 4–3.
(3) The MPU subsystem has one master that connects to the interconnect and another master that connects directly to
the SDRAM controller subsystem. The address filter registers in the MPU L2 control which MPU addresses are sent
to each master. This figure assumes the filter registers contain their reset values.
(4) This address range is configurable. For more information, refer to Table 4–3.
(5) This address range is not accessible from the master peripheral interfaces. For more information, refer to
“Master-to-Slave Connectivity Matrix” on page 4–6.
(6) This address range is not accessible from the DAP interface. For more information, refer to “Master-to-Slave
Connectivity Matrix” on page 4–6.
For the MPU L3 master, either the boot ROM or on-chip RAM maps to address 0x0
and obscures the lowest 64 K of SDRAM. The address space from 0x0001_0000 to
0x0010_0000 is not accessible because the MPU L2 filter registers only have a
granularity of 1 MB. After booting completes, the MPU can change address filtering to
use the lowest 1 MB of SDRAM.
For non-MPU masters, either the on-chip RAM or the SDRAM maps to address 0x0.
When mapped to address 0x0, the on-chip RAM obscures the lowest 64 K of SDRAM
for non-MPU masters.

Table 4–3 lists how the remap bits affect the memory maps.
Table 4–3. Memory Map Remap Bits

Bit Name Bit Offset Description
When set to 0, the boot ROM maps to address 0x0 for the MPU L3
master. When set to 1, the on-chip RAM maps to address 0x0 for the
MPU L3 master. This bit has no effect on non-MPU masters.
mpuzero 0
Note that regardless of this setting, the boot ROM also always maps to
address 0xfffd_0000 and the on-chip RAM also always maps to
address 0xffff_0000 for the MPU L3 master.
When set to 0, the SDRAM maps to address 0x0 for the non-MPU L3
masters. When set to 1, the on-chip RAM maps to address 0x0 for the
nonmpuzero 1 non-MPU masters. This bit has no effect on the MPU L3 master.
Note that regardless of this setting, the on-chip RAM also always
maps to address 0xffff_0000 for the non-MPU L3 masters.
Reserved 2 Must always be set to 0.
When set to 1, the HPS-to-FPGA bridge slave port is visible to the L3
hps2fpga 3 masters. When set to 0, accesses to the associated address range
return an AXI decode error to the master.
When set to 1, the lightweight HPS-to-FPGA bridge slave port is
lwhp2fpga 4 visible to the L3 masters. When set to 0, accesses to the associated
address range return an AXI decode error to the master.
Reserved 31:5 Must always be set to 0.
1 L2 filter registers in the MPU subsystem, not the interconnect, allow the SDRAM to be
remapped to address 0 for the MPU. For more information about the MPU subsystem,
refer to the Cortex-A9 MPU System chapter in volume 3 of the Cyclone V Device
Handbook.
Master Caching and Buffering Overrides

Some of the masters of the interconnect do not have the ability to drive the caching
and buffering signals of their AXI and AHB interfaces. In order to ensure that these
masters can perform transfers efficiently, the registers are available from the system
manager so that you can enable cacheable and bufferable transactions. The following
masters have their caching and buffering signals driven by the system manager:
■ EMAC0 and EMAC1
■ USB OTG 0 and USB OTG 1
■ NAND flash
■ SD/MMC
At reset time, the system manager drives the cache and buffering signals for these
masters low. In other words, the masters listed do not support cacheable or bufferable
accesses until you enable them after reset. There is no synchronization between the
system manager and the interconnect, so avoid changing these settings when any of
the masters are active. For more information about enabling or disabling this feature,
refer to the System Manager chapter in volume 3 of the Cyclone V Device Handbook.

Security
Slave Security
The interconnect enforces security through the slave settings. The slave settings are
controlled by the address region control registers accessible through the GPV
registers. Each L3 and L4 slave has its own security check and programmable security
settings. After reset, every slave of the interconnect is set to a secure state (referred to
as boot secure). The only accesses allowed to secure slaves are by secure masters.
The GPV can only be accessed by secure masters. The security state of the interconnect
is not accessible through the GPV as the security registers are write-only. Any
nonsecure accesses to the GPV receive a DECERR response, and no register access is
provided. Updates to the security settings through the GPV do not take effect until all
transactions to the affected slave have completed.
Master Security
Masters of the interconnect are either secure, nonsecure, or the security is set on a per
transaction basis. The DAP is capable of performing only secure accesses. The L2
cache master 0, FPGA-to-HPS-bridge, and DMA perform secure and nonsecure
accesses on a per transaction basis. All other interconnect masters perform nonsecure
accesses. For more information, refer to “Interconnect Master Properties” on
page 4–10.
Accesses to secure slaves by unsecure masters result in a response of DECERR and the
transaction does not reach the slave.
Arbitration
At the entry point to the interconnect, all transactions are allocated a local quality of
service (QoS) value that you can programmatically configure. The arbitration of the
transaction throughout the infrastructure uses this QoS value. The QoS controls for
each master connected to the interconnect are separated into read and write QoS
priority values.
At any arbitration node, a fixed priority exists for transactions with different QoS
values. The highest QoS value has the highest priority. If there are coincident
transactions at an arbitration node with the same QoS value that require arbitration,
then the interconnect uses a least recently used (LRU) algorithm.
Cyclic Dependency Avoidance Schemes

The AXI protocol permits re-ordering of transactions. As a result, when routing
concurrent multiple transactions from a single point of divergence to multiple slaves,
the interconnect might need to enforce rules to prevent deadlock.
Each master of the interconnect is configured with one of three possible cyclic
dependency avoidance schemes (CDAS). The same CDAS scheme is configured for
both read and write transactions, but they operate independently. The CDAS
implementation for the masters is described in“Interconnect Master Properties” on
page 4–10.
The following schemes are available:

■ Single Slave
■ Single Slave Per ID
■ Single Active Slave
Single Slave
Single slave (SS) ensures the following conditions at a slave interface of a switch:
■ All outstanding read transactions are to a single end destination.
■ All outstanding write transactions are to a single end destination.
If a master issues another transaction to a different destination than the current
destination for that transaction type (read or write), the network stalls the transactions
until all the outstanding transactions of that type have completed.
Single Slave Per ID

Single slave per ID (SSPID) ensures the following conditions at a slave interface of a
switch:
■ All outstanding read transactions with the same ID go the same destination.
■ All outstanding write transactions with the same ID go the same destination.
When a master issues a transaction, the following situations can occur:
■ If the transaction has an ID that does not match any outstanding transactions, it
passes the CDAS.
■ If the transaction has an ID that matches the ID of an outstanding transaction, and
the destinations also match, it passes the CDAS.
■ If the transaction has an ID that matches the ID of an outstanding transaction, and
the destinations do not match, the transaction fails the CDAS check and stalls.
Single Active Slave

Single active slave (SAS) is the same as the SSPID scheme, with an added check for
write transactions. SAS ensures that a master cannot issue a new write address until
all of the data from the previous write transaction has been sent.
Interconnect Master Properties

The interconnect connects to various slave interfaces through the L3 main switch and
L3 slave peripheral switch.
Table 4–4 shows all the master interfaces connected to the interconnect.
Table 4–4. Interconnect Master Interfaces (Part 1 of 2)

Interface GPV Buffer
Master Clock Switch Security CDAS Issuance (1) Type
Width Access Depth (2)
L2 cache L3 main Per 2, 2, 2,
64 mpu_l2_ram_clk Yes SSPID 7, 12, 19 AXI
M0 switch Transaction 2, 2
FPGA-to- L3 main Per 2, 2, 6,
64 l3_main_clk Yes SAS 16, 16, 32 AXI
HPS bridge switch Transaction 6, 2

Table 4–4. Interconnect Master Interfaces (Part 2 of 2)

Interface GPV Buffer
Master Clock Switch Security CDAS Issuance (1) Type
Width Access Depth (2)
L3 main Per 2, 2, 2,
DMA 64 l4_main_clk No SSPID 8, 8, 8 AXI
switch Transaction 2, 2
L3 master
2, 2, 2,
EMAC 0/1 32 l4_main_clk peripheral Nonsecure No SSPID 16, 16, 32 AXI
2, 2
switch
L3 master
USB OTG
32 usb_mp_clk peripheral Nonsecure No SSPID 2, 2, 4 2, 2, 2 AHB
0/1
switch
L3 master
2, 2, 2,
NAND 32 nand_x_clk peripheral Nonsecure No SSPID 1, 8, 9 AXI
2, 2
switch
L3 master
SD/MMC 32 l4_mp_clk peripheral Nonsecure No SSPID 2, 2, 4 2, 2, 2 AHB
switch
L3 master
2, 2, 2,
ETR 32 dbg_at_clk peripheral Nonsecure No SSPID 32, 1, 32 AXI
2, 2
switch
L3 main
DAP 32 dbg_clk Secure Yes SS 1, 1, 1 2, 2, 2 AHB
switch
Notes to Table 4–4:
(1) Issuance is based on the number of read, write, and total transactions.
(2) The FIFO buffer depth for AXI is based on the AW, AR, R, W, and B channels. For AHB and APB, the depth is based on W, A, and D channels.
Interconnect Slave Properties

The interconnect connects to various slave interfaces through the L3 main switch, L3
slave peripheral switch, and the five L4 peripheral buses. After reset, all slave
interfaces are set to the secure state.
Table 4–5 shows all the slave interfaces connected to the interconnect.
Table 4–5. Interconnect Slave Interfaces (Part 1 of 3)

Interface Buffer Interface
Slave Clock Mastered By Acceptance (1)
Width Depth (2) Type
SDRAM
32 l4_sp_clk L4 SP bus master 1, 1, 1 2, 2, 2 APB
subsystem CSR
SP timer 0/1 32 l4_sp_clk L4 SP bus master 1, 1, 1 2, 2, 2 APB
I2C 0/1/2/3 32 l4_sp_clk L4 SP bus master 1, 1, 1 2, 2, 2 APB
UART 0/1 32 l4_sp_clk L4 SP bus master 1, 1, 1 2, 2, 2 APB
CAN 0/1 32 l4_sp_clk L4 SP bus master 1, 1, 1 2, 2, 2 APB
GPIO 0/1/2 32 l4_mp_clk L4 SP bus master 1, 1, 1 2, 2, 2 APB
ACP ID mapper
32 l4_mp_clk L4 SP bus master 1, 1, 1 2, 2, 2 APB
CSR
FPGA manager
CSR
DAP CSR 32 l4_mp_clk L4 SP bus master 1, 1, 1 2, 2, 2 APB


Quad SPI flash
CSR
SD/MMC CSR 32 l4_mp_clk L4 SP bus master 1, 1, 1 2, 2, 2 APB
EMAC 0/1 CSR 32 l4_mp_clk L4 SP bus master 1, 1, 1 2, 2, 2 APB
L4 OSC1 bus
System manager 32 osc1_clk 1, 1, 1 2, 2, 2 APB
master
L4 OSC1 bus
OSC1 timer 0/1 32 osc1_clk 1, 1, 1 2, 2, 2 APB
master
L4 OSC1 bus
Watchdog 0/1 32 osc1_clk 1, 1, 1 2, 2, 2 APB
master
L4 OSC1 bus
Clock manager 32 osc1_clk 1, 1, 1 2, 2, 2 APB
master
L4 OSC1 bus
Reset manager 32 osc1_clk 1, 1, 1 2, 2, 2 APB
master
DMA secure CSR 32 l4_main_clk L4 main bus master 1, 1, 1 2, 2, 2 APB
DMA nonsecure
32 l4_main_clk L4 main bus master 1, 1, 1 2, 2, 2 APB
CSR
SPI slave 0/1 32 l4_main_clk L4 main bus master 1, 1, 1 2, 2, 2 APB
Scan manager 32 spi_m_clk L4 main bus master 1, 1, 1 2, 2, 2 APB
SPI master 0/1 32 spi_m_clk L4 main bus master 1, 1, 1 2, 2, 2 APB
Lightweight
L3 slave peripheral
HPS-to-FPGA 32 l4_main_clk 16, 16, 32 2, 2, 2, 2, 2 AXI
switch
bridge
L3 slave peripheral
USB OTG 0/1 32 usb_mp_clk 1, 1, 1 2, 2, 2 AHB
switch
L3 slave peripheral
NAND CSR 32 nand_x_clk 1, 1, 1 2, 2, 2 AXI
switch
NAND command L3 slave peripheral
32 nand_x_clk 1, 1, 1 2, 2, 2 AXI
and data switch
Quad SPI flash L3 slave peripheral
32 l4_mp_clk 1, 1, 1 2, 2, 2 AHB
data switch
FPGA manager
32 cfg_clk L3 main switch 1, 2, 3 2, 2, 2, 32, 2 AXI
data
HPS-to-FPGA
64 l3_main_clk L3 main switch 16, 16, 32 2, 2, 6, 6, 2 AXI
bridge
ACP ID mapper
64 mpu_l2_ram_clk L3 main switch 13, 5, 18 2, 2, 2, 2, 2 AXI
data
STM 32 dbg_at_clk L3 main switch 1, 2, 2 2, 2, 2, 2, 2 AXI
On-chip boot ROM 32 l3_main_clk L3 main switch 1, 1, 2 0, 0, 0, 0, 0 AXI
On-chip RAM 64 l3_main_clk L3 main switch 2, 2, 2 0, 0, 0, 8, 0 AXI


SDRAM
32 l3_main_clk L3 main switch 16, 16, 16 2, 2, 2, 2, 2 AXI
subsystem L3 data
(1) Acceptance is based on the number of read, write, and total transactions.
(2) The FIFO buffer depth for AXI is based on the AW, AR, R, W, and B channels. For AHB and APB, the depth is based on the W, A, and D channels.
Upsizing Data Width Function

The upsizing function combines narrow transactions into wider transactions to
increase the overall system bandwidth. Upsizing only packs data for read or write
transactions that are cacheable. If the interconnect splits input-exclusive transactions
into more than one output bus transaction, it removes the exclusive information from
the multiple transactions it creates.
The upsizing function can expand the data width by the following ratios:
■ 1:2
■ 1:4
If multiple responses from created transactions are combined into one response, then
the following order of priority applies:
■ DECERR is the highest priority
■ SLVERR is the next highest priority
■ OKAY is the lowest priority.
f For more information about AXI terms such as DECERR, WRAP, and INCR, refer to the
AMBA AXI Protocol Specification v1.0, which you can download from the ARM website
Incrementing Bursts
The interconnect converts all input INCR bursts that complete within a single output
data width to an INCR1 burst of the minimum SIZE possible, and packs all INCR bursts
into INCR bursts of the optimal size possible for maximum data throughout.
Wrapping Bursts
All WRAP bursts are either passed through unconverted as WRAP bursts, or converted to
one or two INCR bursts of the output bus. The interconnect converts input WRAP bursts
that have a total payload less than the output data width to a single INCR burst.
Fixed Bursts
All FIXED bursts pass through unconverted.

Bypass Merge
If the programmable bit bypass_merge is enabled, the interconnect does not alter any
transactions that could pass through legally without alteration. Bypass merge is
accessible through the GPV registers and is only accessible to secure masters.
Downsizing Data Width Function

The downsizing function reduces the data width of a transaction to match the optimal
data width at the destination. Downsizing does not merge multiple-transaction data
narrower than the destination bus if the transactions are marked as noncacheable.
The downsizing function reduces the data width by the following ratios:
■ 2:1
■ 4:1
Incrementing Bursts
The interconnect converts INCR bursts that fall within the maximum payload size of
the output data bus to a single INCR burst. It converts INCR bursts that are greater than
the maximum payload size of the output data bus to multiple INCR bursts.
INCR bursts with a size that matches the output data width pass through unconverted.
The interconnect packs INCR bursts with a SIZE smaller than the output data width to
match the output width whenever possible, using the upsizing function. For more
information, refer to“Upsizing Data Width Function” on page 4–13.
Wrapping Bursts
The interconnect always converts WRAP bursts to WRAP bursts of twice the length, up to
the output data width maximum size of WRAP16, and treats the WRAP burst as two INCR
bursts that can each be converted into one or more INCR bursts.
Fixed Bursts
The interconnect converts FIXED bursts to one or more INCR1 or INCRn bursts
depending on the downsize ratio.
Bypass Merge
If the programmable bit bypass_merge is enabled, the interconnect does not perform
any packing of beats to match the optimal SIZE for maximum throughput, up to the
output data width SIZE. Bypass merge is accessible through the GPV registers and is
only accessible to secure masters.
If an exclusive transaction is split into multiple transactions at the output of the
downsizing function, the exclusive flag is removed and the master never receives an
EXOKAY response. Response priority is the same as for the upsizing function, described
in “Upsizing Data Width Function” on page 4–13.
Lock Support
Lock is not supported by the interconnect. For atomic accesses, masters can perform
exclusive accesses when sharing data located in the HPS SDRAM.

Interconnect Address Map and Register Definitions
f For more information about exclusive access support, refer to the SDRAM Controller
Subsystem chapter in volume 3 of the Cyclone V Device Handbook.
FIFO Buffers and Clocks

The interconnect contains FIFO buffers in the majority of the interfaces exposed to the
HPS master and slaves, as well as between the subswitches. These FIFO buffers also
provide clock domain crossing for masters and slaves that operate at a different clock
frequency than the switch they connect to.
Data Release Mechanism

For network ports containing write data FIFO buffers with a depth of four or greater,
you can set a write tidemark function, wr_tidemark. This tidemark level stalls the
release of the transaction until one of the following situations occurs:
■ The interconnect receives the WLAST beat of a burst.
■ The write data FIFO buffer becomes full.
■ The number of occupied slots in the write data FIFO buffer exceeds the write
tidemark.
For more information about which interfaces contain write data FIFO buffers with a
depth of four or greater, refer to “Interconnect Master Properties” on page 4–10.
Resets
The interconnect has one reset signal. The reset manager drives this signal to the
SD/MMC controller on a cold or warm reset. On reset, the boot ROM is mapped to
address 0x0. The DAP virtually maps to ID 2.
f For more information about resets, refer to the Reset Manager chapter in volume 3 of
the Cyclone V Device Handbook. For information about virtual ID mapping by the ACP
ID mapper, refer to the Cortex-A9 MPU System chapter in volume 3 of the Cyclone V
Device Handbook.
Interconnect Address Map and Register Definitions

■ l3regs



■ Added main switch connectivity matrix table.
■ Rearranged functional description sections.
June 2012 1.1
■ Simplified address remapping section.

5. HPS-FPGA AXI Bridges
November 2012
cv_54005-1.1
cv_54005-1.1
This chapter describes the bridges in the hard processor system (HPS) used to
communicate data between the FPGA fabric and HPS logic. The bridges use the
Advanced Microcontroller Bus Architecture (AMBA®) Advanced eXtensible Interface
(AXI™) protocol, and are based on the AMBA Network Interconnect (NIC-301).
f Additional information is available in the AMBA AXI Protocol Specification v1.0 and
the AMBA Network Interconnect (NIC-301) Technical Reference Manual, which you can
download from the ARM website (infocenter.arm.com).
The HPS contains the following HPS-FPGA AXI bridges:

■ FPGA-to-HPS Bridge
■ HPS-to-FPGA Bridge
■ Lightweight HPS-to-FPGA Bridge
Features of the AXI Bridges

The HPS-FPGA AXI bridges allow masters in the FPGA fabric to communicate with
slaves in the HPS logic and vice versa. For example, you can instantiate additional
memories or peripherals in the FPGA fabric, and master interfaces belonging to
components in the HPS logic can access them. You can also instantiate components
such as a Nios® II processor in the FPGA fabric and their master interfaces can access
memories or peripherals in the HPS logic.
The AXI bridges provide the features listed in Table 5–1.
Table 5–1. AXI Bridge Features

Lightweight
FPGA-to-HPS HPS-to-FPGA
Feature HPS-to-FPGA
Bridge Bridge
Bridge
Supports the AMBA AXI3 interface protocol v v v
Implements clock crossing and manages the
transfer of data across the clock domains in the v v v
HPS logic and the FPGA fabric
Performs data width conversion between the HPS
v v v
logic and the FPGA fabric
Allows configuration of FPGA interface widths at
v v
instantiation time

November 2012
Subscribe
5–2 Chapter 5: HPS-FPGA AXI Bridges
AXI Bridges Block Diagram and System Integration
Each bridge consists of an AXI master-slave pair with one interface exposed to the
FPGA fabric and the other exposed to the HPS logic. The HPS-to-FPGA and
lightweight HPS-to-FPGA bridges expose an AXI master interface that you can
connect to AXI or Avalon-MM slave interfaces in the FPGA fabric. The FPGA-to-HPS
bridge exposes an AXI slave interface that you can connect to AXI master or
Avalon-MM interfaces in the FPGA fabric.
f For information about configuring the AXI bridges, refer to the Instantiating the HPS
Component chapter in volume 3 of the Cyclone V Device Handbook.
AXI Bridges Block Diagram and System Integration

Figure 5–1 shows a block diagram of the AXI bridges in the context of the FPGA fabric
and the L3 interconnect to the HPS. Each master (M) and slave (S) interface is shown
with its data width(s). The clock domain for each interconnect is shown in
parentheses. The clock domains are described in “Clocks and Resets” on page 5–12.
Figure 5–1. AXI Bridge Connectivity
FPGA Fabric
32, 64, or 128 Bits 32 Bits 32, 64, or 128 Bits

(h2f_axi_clk) (h2f_lw_axi_clk) (f2h_axi_clk)
32 Bits 32 Bits
M S
AXI S (l4_mp_clk) (l4_mp_clk) S AXI
AHB AHB
(GPV) M M M (GPV)
HPS-to-FPGA AHB AXI AHB FPGA-to-HPS
Bridge Bridge
Lightweight
S HPS-to-FPGA Bridge M
AXI AXI
S
AXI (GPV)
64 Bits 32 Bits 64 Bits

(l3_main_clk) (l4_mp_clk) (l3_main_clk)
M M S
AXI AXI AXI
(L3 Main Switch) (L3 Slave Peripheral Switch) (L3 Main Switch)
L3 Interconnect
The HPS-to-FPGA bridge is mastered by the level 3 (L3) main switch and the
lightweight HPS-to-FPGA bridge is mastered by the L3 slave peripheral switch.
The FPGA-to-HPS bridge masters the L3 main switch, allowing any master
implemented in the FPGA fabric to access most slaves in the HPS. For example, the
FPGA-to-HPS bridge can access the accelerator coherency port (ACP) of the
Cortex-A9 MPU subsystem to perform cache-coherent accesses to the SDRAM
subsystem.

Chapter 5: HPS-FPGA AXI Bridges 5–3
Functional Description of the AXI Bridges
All three bridges contain global programmers view (GPV) registers. The GPV
registers control the behavior of the bridge. Access to the GPV registers of all three
bridges is provided through the lightweight HPS-to-FPGA bridge.
f For more information about connectivity, such as which masters have access to each
bridge, refer to the Interconnect chapter in volume 3 of the Cyclone V Device Handbook.
The Global Programmers View

The HPS-to-FPGA bridge includes a set of registers called the GPV. The GPV provides
settings to control the bridge properties and behavior. Access to the GPV registers of
all three bridges is provided through the lightweight HPS-to-FPGA bridge.
The GPV registers can only be accessed by secure masters in the HPS or the FPGA
fabric.
FPGA-to-HPS Bridge
The FPGA-to-HPS bridge provides access to the peripherals and memory in the HPS.
This access is available to any master implemented in the FPGA fabric. You can
configure the bridge slave, which is exposed to the FPGA fabric, to support 32-, 64-, or
128-bit data. The master interface of the bridge, connected to the L3 interconnect, has
a data width of 64 bits.
Table 5–2 lists the properties of the FPGA-to-HPS bridge, including the configurable
slave interface exposed to the FPGA fabric.
Table 5–2. FPGA-to-HPS Bridge Properties

Bridge Property FPGA Slave Interface L3 Master Interface
Data width (1) 32, 64, or 128 bits 64 bits
Clock domain f2h_axi_clk l3_main_clk
Byte address width 32 bits 32 bits
ID width 8 bits 8 bits
Read acceptance 16 transactions 16 transactions
Write acceptance 16 transactions 16 transactions
Total acceptance 32 transactions 32 transactions
(1) The bridge slave data width is user-configurable at the time you instantiate the HPS component in your system.
The FPGA-to-HPS bridge contains a GPV, described in “The Global Programmers

View”. The GPV registers provide settings that adjust the bridge slave properties
when the FPGA slave interface is configured to be 32 or 128 bits wide. The slave
issuing capability can be adjusted, through the fn_mod register, to allow one or
multiple transactions to be outstanding in the HPS. The slave bypass merge feature
can also be enabled, through the bypass_merge bit in the fn_mod2 register. This feature
ensures that the upsizing and downsizing logic does not alter any transactions when
the FPGA slave interface is configured to be 32 or 128 bits wide.

1 It is critical to provide the correct l4_mp_clk clock to support access to the GPV, as
described in “GPV Clocks” on page 5–13.
FPGA-to-HPS Access to ACP

When the error correction code (ECC) option is enabled in the level 2 (L2) cache
controller, all accesses from the FPGA-to-HPS bridge to the ACP must be 64 bits wide
and aligned on 8-byte boundaries after up- or downsizing takes place.
Table 5–3 lists some possible master and FPGA-to-HPS bridge slave configurations
that support accesses to the L2 cache with ECC enabled.
Table 5–3. FPGA Master and FPGA-to-HPS Bridge Configurations

Soft Logic Master Soft Logic Master Soft Logic Master FPGA-to-HPS Bridge
Soft Logic Master Burst Length
Width Alignment Burst Size (Width) Slave Width
32 bits 8 bytes 4 bytes 2, 4, 6, 8, 10, 12, 14, or 16 beats 32 bits
64 bits 8 bytes 8 bytes 1 to 16 beats 32 bits
128 bits 8 or 16 bytes 8 or 16 bytes 1 to 16 beats 32 bits
f For more information about the ECC option of the L2 cache, refer to the Cortex-A9
Microprocessor Unit Subsystem chapter in volume 3 of the Cyclone V Device Handbook.
FPGA-to-HPS Bridge Slave Signals

The FPGA-to-HPS bridge slave address channels support user-sideband signals,
routed to the ACP in the MPU subsystem. All the signals have a fixed width except
the data and write strobes for the read and write data channels. The variable width
signals depend on the data width setting of the bridge. Table 5–4 through Table 5–8
list all the signals exposed by the FPGA-to-HPS slave interface to the FPGA fabric.
Table 5–4 lists the slave write address channel signals.
Table 5–4. FPGA-to-HPS Bridge Slave Write Address Channel Signals

Signal Width Direction Description
AWID 8 bits Input Write address ID
AWADDR 32 bits Input Write address
AWLEN 4 bits Input Burst length
AWSIZE 3 bits Input Burst size
AWBURST 2 bits Input Burst type
Lock type—Valid values are 00 (normal
AWLOCK 2 bits Input
access) and 01 (exclusive access)
AWCACHE 4 bits Input Cache policy type
AWPROT 3 bits Input Protection type

Table 5–4. FPGA-to-HPS Bridge Slave Write Address Channel Signals

AWVALID 1 bit Input Write address channel valid
AWREADY 1 bit Output Write address channel ready
AWUSER 5 bits Input User sideband signals
Table 5–5 lists the slave write data channel signals.
Table 5–5. FPGA-to-HPS Bridge Slave Write Data Channel Signals

WID 8 bits Input Write ID
WDATA 32, 64, or 128 bits Input Write data
WSTRB 4, 8, or 16 bits Input Write data strobes
WLAST 1 bit Input Write last data identifier
WVALID 1 bit Input Write data channel valid
WREADY 1 bit Output Write data channel ready
Table 5–6 lists the slave write response channel signals.
Table 5–6. FPGA-to-HPS Bridge Slave Write Response Channel Signals

BID 8 bits Output Write response ID
BRESP 2 bits Output Write response
BVALID 1 bit Output Write response channel valid
BREADY 1 bit Input Write response channel ready
Table 5–7 lists the slave read address channel signals.
Table 5–7. FPGA-to-HPS Bridge Slave Read Address Channel Signals

ARID 8 bits Input Read address ID
ARADDR 32 bits Input Read address
ARLEN 4 bits Input Burst length
ARSIZE 3 bits Input Burst size
ARBURST 2 bits Input Burst type
ARLOCK 2 bits Input
ARCACHE 4 bits Input Cache policy type
ARPROT 3 bits Input Protection type
ARVALID 1 bit Input Read address channel valid
ARREADY 1 bit Output Read address channel ready
ARUSER 5 bits Input Read user sideband signals

Table 5–8 lists the slave read data channel signals.
Table 5–8. FPGA-to-HPS Bridge Slave Read Data Channel Signals

RID 8 bits Output Read ID
RDATA 32, 64, or 128 bits Output Read data
RRESP 2 bits Output Read response
RLAST 1 bit Output Read last data identifier
RVALID 1 bit Output Read data channel valid
RREADY 1 bit Input Read data channel ready
HPS-to-FPGA Bridge
The HPS-to-FPGA bridge provides a configurable-width, high-performance master
interface to the FPGA fabric. The bridge provides most masters in the HPS with access
to logic, peripherals, and memory implemented in the FPGA. The effective size of the
address space is 0x3FFF0000, or 1 gigabyte (GB) minus 64 megabytes (MB). The
address space size is less than 1 GB because 64 MB is occupied by peripherals,
lightweight HPS-to-FPGA bridge, on-chip RAM, and boot ROM in the HPS. You can
configure the bridge master exposed to the FPGA fabric for 32-, 64-, or 128-bit data.
The amount of address space exposed to the MPU subsystem can also be reduced
through the L2 cache address filtering mechanism.
f For detailed information about which masters have access to each bridge, refer to the
Interconnect chapter in volume 3 of the Cyclone V Device Handbook. For details about L2
cache address filtering, refer to the Cortex-A9 Microprocessor Unit Subsystem chapter in
The slave interface of the bridge in the HPS logic has a data width of 64 bits. The
bridge provides width adaptation and clock crossing logic that allows the logic in the
FPGA to operate in any clock domain, asynchronous from the HPS.
1 The HPS-to-FPGA bridge is accessed if the MPU boots from the FPGA. Before the
MPU boots from the FPGA, the FPGA portion of the SoC device must be configured,
and the HPS-to-FPGA bridge must be remapped into addressable space.
f For more information about enabling the HPS-to-FPGA bridge, refer to the

Table 5–9 lists the properties of the HPS-to-FPGA bridge, including the configurable
master interface exposed to the FPGA fabric.
Table 5–9. HPS-to-FPGA Bridge Properties

Bridge Property L3 Slave Interface FPGA Master Interface
Data width (1) 64 bits 32, 64, or 128 bits
Clock domain l3_main_clk h2f_axi_clk
(1) The bridge master data width is user-configurable at the time you instantiate the HPS component in your system.
The HPS-to-FPGA bridge’s GPV, described in “The Global Programmers View” on

page 5–3, provides settings to adjust the bridge master properties. The master issuing
capability can be adjusted, through the fn_mod register, to allow one or multiple
transactions to be outstanding in the FPGA fabric. The master bypass merge feature
can also be enabled, through the bypass_merge bit in the fn_mod2 register. This feature
ensures that the upsizing and downsizing logic does not alter any transactions when
the FPGA master interface is configured to be 32 or 128 bits wide.
1 It is critical to provide the correct l4_mp_clk clock to support access to the GPV, as
described in “GPV Clocks” on page 5–13.
HPS-to-FPGA Bridge Master Signals

All the HPS-to-FPGA bridge master signals have a fixed width except the data and
write strobes for the read and write data channels. The variable-width signals depend
on the data width setting of the bridge interface exposed to the FPGA logic. Table 5–10
through Table 5–14 list all the signals exposed by the HPS-to-FPGA master interface to
the FPGA fabric.

Table 5–10 lists the master write address channel signals.
Table 5–10. HPS-to-FPGA Bridge Master Write Address Channel Signals

AWID 12 bits Output Write address ID
AWADDR 30 bits Output Write address
AWLEN 4 bits Output Burst length
AWSIZE 3 bits Output Burst size
AWBURST 2 bits Output Burst type
AWLOCK 2 bits Output
AWCACHE 4 bits Output Cache policy type
AWPROT 3 bits Output Protection type
AWVALID 1 bit Output Write address channel valid
AWREADY 1 bit Input Write address channel ready
Table 5–11 lists the master write data channel signals.
Table 5–11. HPS-to-FPGA Bridge Master Write Data Channel Signals

WID 12 bits Output Write ID
WDATA 32, 64, or 128 bits Output Write data
WSTRB 4, 8, or 16 bits Output Write data strobes
WLAST 1 bit Output Write last data identifier
WVALID 1 bit Output Write data channel valid
WREADY 1 bit Input Write data channel ready
Table 5–12 lists the master write response channel signals.
Table 5–12. HPS-to-FPGA Bridge Master Write Response Channel Signals

BID 12 bits Input Write response ID
BRESP 2 bits Input Write response
BVALID 1 bit Input Write response channel valid
BREADY 1 bit Output Write response channel ready

Table 5–13 lists the master read address channel signals.
Table 5–13. HPS-to-FPGA Bridge Master Read Address Channel Signals

ARID 12 bits Output Read address ID
ARADDR 30 bits Output Read address
ARLEN 4 bits Output Burst length
ARSIZE 3 bits Output Burst size
ARBURST 2 bits Output Burst type
ARLOCK 2 bits Output
ARCACHE 4 bits Output Cache policy type
ARPROT 3 bits Output Protection type
ARVALID 1 bit Output Read address channel valid
ARREADY 1 bit Input Read address channel ready
Table 5–14 lists the master read data channel signals.
Table 5–14. HPS-to-FPGA Bridge Master Read Data Channel Signals

RID 12 bits Input Read ID
RDATA 32, 64, or 128 bits Input Read data
RRESP 2 bits Input Read response
RLAST 1 bit Input Read last data identifier
RVALID 1 bit Input Read data channel valid
RREADY 1 bit Output Read data channel ready
Lightweight HPS-to-FPGA Bridge

The lightweight HPS-to-FPGA bridge provides a lower-performance interface to the
FPGA fabric. This interface is useful for accessing the control and status registers of
soft peripherals. The bridge provides a 2 MB address space and access to logic,
peripherals, and memory implemented in the FPGA fabric. The MPU subsystem,
direct memory access (DMA) controller, and debug access port (DAP) can use the
lightweight HPS-to-FPGA bridge to access the FPGA fabric or GPV. Master interfaces
in the FPGA fabric can also use the lightweight HPS-to-FPGA bridge to access the
GPV registers in all three bridges.
f For detailed information about which masters have access to each bridge, refer to the
The bridge master exposed to the FPGA fabric has a fixed data width of 32 bits. The
slave interface of the bridge in the HPS logic has a fixed data width of 32 bits.

Use the lightweight HPS-to-FPGA bridge as a secondary, lower-performance master

interface to the FPGA fabric. With a fixed width and a smaller address space, the
lightweight bridge is useful for low-bandwidth traffic, such as memory-mapped
register accesses to FPGA peripherals. This approach diverts traffic from the
high-performance HPS-to-FPGA bridge, and can improve both CSR access latency
and overall system performance.
Table 5–15 lists the properties of the lightweight HPS-to-FPGA bridge, including the
master interface exposed to the FPGA fabric.
Table 5–15. Lightweight HPS-to-FPGA Bridge Properties

Bridge Property L3 Slave Interface FPGA Master Interface
Data width 32 bits 32 bits
Clock domain l4_mp_clk h2f_lw_axi_clk
The lightweight HPS-to-FPGA bridge has three master interfaces, as shown in

Figure 5–1 on page 5–2. The master interface connected to the FPGA fabric provides a
lightweight interface from the HPS to custom logic in the FPGA fabric. The two other
master interfaces, connected to the HPS-to-FPGA and FPGA-to-HPS bridges, allow
you to access the GPV registers for each bridge.
The lightweight HPS-to-FPGA bridge also has a GPV to control the behavior of its
four interfaces (one slave and three masters). The GPV is described in “The Global
Programmers View” on page 5–3.
The GPV allows you to set the bridge’s issuing capabilities to support single or
multiple transactions. The GPV also lets you set a write tidemark through the
wr_tidemark register, to control how much data is buffered in the bridge before data is
written to slaves in the FPGA fabric.
1 It is critical to provide correct clock settings for the lightweight HPS-to-FPGA bridge,
even if your design does not use this bridge. The l4_mp_clk clock is required for GPV
access on the HPS-to-FPGA and FPGA-to-HPS bridges.
Lightweight HPS-to-FPGA Bridge Master Signals

All the lightweight HPS-to-FPGA bridge master signals have a fixed width.
Table 5–16 through Table 5–20 list all the signals exposed by the lightweight
HPS-to-FPGA master interface to the FPGA fabric.

Table 5–16 lists the master write address channel signals.
Table 5–16. Lightweight HPS-to-FPGA Bridge Master Write Address Channel Signals
AWID 12 bits Output Write address ID
AWADDR 21 bits Output Write address
AWLEN 4 bits Output Burst length
AWSIZE 3 bits Output Burst size
AWBURST 2 bits Output Burst type
AWLOCK 2 bits Output
AWCACHE 4 bits Output Cache policy type
AWPROT 3 bits Output Protection type
AWVALID 1 bit Output Write address channel valid
AWREADY 1 bit Input Write address channel ready
Table 5–17 lists the master write data channel signals.
Table 5–17. Lightweight HPS-to-FPGA Bridge Master Write Data Channel Signals
WID 12 bits Output Write ID
WDATA 32 bits Output Write data
WSTRB 4 bits Output Write data strobes
WLAST 1 bit Output Write last data identifier
WVALID 1 bit Output Write data channel valid
WREADY 1 bit Input Write data channel ready
Table 5–18 lists the master write response channel signals.
Table 5–18. Lightweight HPS-to-FPGA Bridge Master Write Response Channel Signals
BID 12 bits Input Write response ID
BRESP 2 bits Input Write response
BVALID 1 bit Input Write response channel valid
BREADY 1 bit Output Write response channel ready

Table 5–19 lists the master read address channel signals.
Table 5–19. Lightweight HPS-to-FPGA Bridge Master Read Address Channel Signals
ARID 12 bits Output Read address ID
ARADDR 21 bits Output Read address
ARLEN 4 bits Output Burst length
ARSIZE 3 bits Output Burst size
ARBURST 2 bits Output Burst type
ARLOCK 2 bits Output
ARCACHE 4 bits Output Cache policy type
ARPROT 3 bits Output Protection type
ARVALID 1 bit Output Read address channel valid
ARREADY 1 bit Input Read address channel ready
Table 5–20 lists the master read data channel signals.
Table 5–20. Lightweight HPS-to-FPGA Bridge Master Read Data Channel Signals
RID 12 bits Input Read ID
RDATA 32 bits Input Read data
RRESP 2 bits Input Read response
RLAST 1 bit Input Read last data identifier
RVALID 1 bit Input Read data channel valid
RREADY 1 bit Output Read data channel ready
Clocks and Resets

FPGA-to-HPS Bridge
The master interface of the bridge in the HPS logic operates in the l3_main_clk clock
domain. The slave interface exposed to the FPGA fabric operates in the f2h_axi_clk
clock domain provided by the user logic. The bridge provides clock crossing logic that
allows the logic in the FPGA to operate in any clock domain, asynchronous from the
HPS.
f For information about the f2h_axi_clk clock, refer to the HPS Component Interfaces
chapter in volume 3 of the Cyclone V Device Handbook. For information about the
l3_main_clk and l4_mp_clk clocks, refer to the Clock Manager chapter in volume 3 of
The FPGA-to-HPS bridge has one reset signal, fpga2hps_bridge_rst_n. The reset
manager drives this signal to FPGA-to-HPS bridge on a cold or warm reset.
f For more information about the reset manager, refer to the Reset Manager chapter in

HPS-to-FPGA Bridge
The master interface into the FPGA fabric operates in the h2f_axi_clk clock domain.
The h2f_axi_clk clock is provided by user logic. The slave interface of the bridge in
the HPS logic operates in the l3_main_clk clock domain. The bridge provides clock
crossing logic that allows the logic in the FPGA to operate in any clock domain,
asynchronous from the HPS.
f For information about the l3_main_clk clock, refer to the Clock Manager chapter in
volume 3 of the Cyclone V Device Handbook. For information about the h2f_axi_clk
clock, refer to the HPS Component Interfaces chapter in volume 3 of the Cyclone V
Device Handbook.
The HPS-to-FPGA bridge has one reset signal, hps2fpga_bridge_rst_n. The reset
manager drives this signal to HPS-to-FPGA bridge on a cold or warm reset.
For more information about the reset manager, refer to the Reset Manager chapter in
Lightweight HPS-to-FPGA Bridge

The master interface into the FPGA fabric operates in the h2f_lw_axi_clk clock
domain provided by custom logic in the FPGA fabric. The slave interface of the bridge
in the HPS logic operates in the l4_mp_clk clock domain. The bridge provides clock
crossing logic that allows the logic in the FPGA to operate in any clock domain,
asynchronous from the HPS.
f For information about the l4_mp_clk clock, refer to the Clock Manager chapter in
volume 3 of the Cyclone V Device Handbook. For information about the h2f_lw_axi_clk
clock, refer to the HPS Component Interfaces chapter in volume 3 of the Cyclone V
Device Handbook.
The lightweight HPS-to-FPGA bridge has one reset signal, lwhps2fpga_bridge_rst_n.

The reset manager drives this signal to the lightweight HPS-to-FPGA bridge on a cold
or warm reset.
GPV Clocks
The FPGA-to-HPS and HPS-to-FPGA bridges have GPV slave interfaces, mastered by
the lightweight HPS-to-FPGA bridge. These interfaces operate in the l4_mp_clk clock
domain. Even if you do not use the lightweight HPS-to-FPGA bridge in your FPGA
design, you must ensure that a valid l4_mp_clk clock is being generated, so that the
GPV registers in the HPS-to-FPGA and FPGA-to-HPS bridges can be programmed.
The GPV logic in all three bridges is in the l4_mp_clk domain. For information about
the GPV, refer to “The Global Programmers View” on page 5–3.

HPS-FPGA AXI Bridges Address Map and Register Definitions
Data Width Sizing

The HPS-to-FPGA and FPGA-to-HPS bridges allow 32-, 64-, and 128-bit interfaces to
be exposed to the FPGA fabric. For 32-bit and 128-bit interfaces, the bridge performs
data width conversion to the fixed 64-bit interface within the HPS. This conversion is
called upsizing in the case of data being converted from a 64-bit interface to a 128-bit
interface. It is called downsizing in the case of data being converted from a 64-bit
interface to a 32-bit interface. If an exclusive access is split into multiple transactions,
the transactions lose their exclusive access information.
During the upsizing or downsizing process, transactions can also be resized using a
data merging technique. For example, in the case of a 32-bit to 64-bit upsizing, if the
size of each beat entering the bridge’s 32-bit interface is only two bytes, the bridge can
merge up to four beats to form a single 64-bit beat. Similarly, in the case of a 128-bit to
64-bit downsizing, if the size of each beat entering the bridge’s 128-bit interface is only
four bytes, the bridge can merge two beats to form a single 64-bit beat.
The bridges do not perform transaction merging for accesses marked as noncacheable.
1 You can set the bypass_merge bit in the GPV to prevent the bridge from merging data
and responses. If the bridge merges multiple responses into a single response, that
response is the one with the highest priority. The response types have the following
priorities:
1. DECERR
2. SLVERR
3. OKAY
HPS-FPGA AXI Bridges Address Map and Register Definitions

To view the module description and base address, scroll to and click the following
links for the module instance:
■ fpga2hpsregs
■ hps2fpgaregs
Processor chapter in volume 3 of the Cyclone V Device Handbook.


Table 5–21 lists the revision history for this document.

November 2012 1.1 Described GPV.


Section III. Cortex-A9 Microprocessor

■ Chapter 6, Cortex-A9 Microprocessor Unit Subsystem

III–2 Section III: Cortex-A9 Microprocessor

6. Cortex-A9 Microprocessor Unit
Subsystem
November 2012
cv_54006-1.2
cv_54006-1.2
The hard processor system (HPS) in the Altera® SoC FPGA device includes a
stand-alone, full-featured ARM® Cortex™-A9 MPCore™, single- or dual-core 32-bit
application processor. The Cortex-A9 MPU subsystem is composed of a Cortex-A9
MPCore, a level 2 (L2) cache, an Accelerator Coherency Port (ACP) ID mapper, and
debugging modules.
Features of the Cortex-A9 MPU Subsystem

The Altera Cortex-A9 MPU subsystem provides the following features:
■ One or two Cortex-A9 processors
■ Private interval and watchdog timer for each processor
■ Global timer
■ TrustZone® system security extensions
■ Symmetric multiprocessing (SMP) and asymmetric multiprocessing (AMP) modes
■ Debugging modules

November 2012
Subscribe
6–2 Chapter 6: Cortex-A9 Microprocessor Unit Subsystem
Cortex-A9 MPU Subsystem Block Diagram and System Integration
Cortex-A9 MPU Subsystem Block Diagram and System Integration

Figure 6–1 shows a dual-core MPU subsystem in the context of the HPS, with the L2
cache. The L2 cache can access either the level 3 (L3) interconnect fabric or the
SDRAM.
Figure 6–1. Cortex-A9 MPU Subsystem with L3 Interconnect
L3 Interconnect MPU Subsystem

(NIC-301) ARM Cortex-A9 MPCore
Interrupts
CPU0 CPU1
ACP ID Debug Infrastructure

ACP SCU
Mapper
L2 Cache
M0
M1
SDRAM
Controller
Subsystem

Chapter 6: Cortex-A9 Microprocessor Unit Subsystem 6–3
Cortex-A9 MPU Subsystem Components
Figure 6–2 shows a block diagram of the Altera Cortex-A9 MPU subsystem.
Figure 6–2. Cortex-A9 MPU Subsystem Internals
Cortex-A9 MPU Subsystem
ARM Cortex-A9 MPCore
CPU0 CPU1 (Dual-Core HPS Only)
ARM Cortex-A9 Processor ARM Cortex-A9 Processor
NEON Media SIMD NEON Media SIMD

Processing Engine with FPU Processing Engine with FPU
MMU MMU
32 KB 32 KB 32 KB 32 KB
Instruction Data Instruction Data
Cache Cache Cache Cache
CPU0 Private Interval Timer CPU0 Private Watchdog Timer CPU1 Private Interval Timer CPU1 Private Watchdog Timer
GIC (Generic Interrupt Controller)
Global Timer
Accelerator Coherency Port Snoop Control Unit
ACP ID Mapper 512 KB L2 Cache
Debugging Modules
CoreSight Multicore Debug and Trace
Cross Triggering
Event Trace
CPU0 Performance Monitor CPU1 Performance Monitor
CPU0 Program Trace CPU1 Program Trace

The Altera Cortex-A9 MPU subsystem consists of the following hardware blocks:
■ ARM Cortex-A9 MPCore
■ ARM L2C-310 L2 cache controller

■ ACP ID mapper
■ Debugging and trace features
This section describes the components of the Cortex-A9 MPU subsystem.
Cortex-A9 MPCore
The MPU subsystem includes a stand-alone, full-featured ARM Cortex-A9 MPCore
single- or dual-core 32-bit application processor. The processor, like other HPS
masters, can access IP in the FPGA fabric using through the HPS-to-FPGA bridges.
Functional Description
The ARM Cortex-A9 MPCore contains the following blocks:
■ One or two Cortex-A9 Revision r3p0 processors operating in SMP or AMP mode
■ Snoop control unit (SCU)
■ Private interval timer for each processor core
■ Private watchdog timer for each processor core
■ Global timer
Each transaction originating from the Altera Cortex-A9 MPU subsystem can be
flagged as secure or nonsecure.

Implementation Details
Table 6–1 shows the parameter settings for the Altera Cortex-A9 MPCore.
Table 6–1. Cortex-A9 MPCore Processor Configuration

Feature Options
Cortex-A9 processors 1 or 2
Instruction cache size per Cortex-A9 processor 32 KB
Data cache size per Cortex-A9 processor 32 KB
TLB size per Cortex-A9 processor 128 entries
Media Processing Engine with NEON™ technology per Cortex-A9 processor (1) Included
Preload Engine per Cortex-A9 processor Included
Number of entries in the Preload Engine FIFO per Cortex-A9 processor 16
Jazelle DBX extension per Cortex-A9 processor Full
PTM interface per Cortex-A9 processor Included
Support for parity error detection (2) Included
ARM_BIST Included
Master ports Two
Accelerator Coherency Port Included
(1) Includes support for floating-point operations.
(2) For a description of the parity error scheme and parity error signals, Refer to the Cortex-A9 Technical Reference
Manual, Revision r3p0, available on the ARM website (infocenter.arm.com).
f For further information about Cortex-A9 MPCore configurable options, refer to the
Introduction chapter of the Cortex-A9 MPCore Technical Reference Manual, Revision r3p0,
available on the ARM website (infocenter.arm.com).
Cortex-A9 Processor
Each Cortex-A9 processor includes the following hardware blocks:
■ ARM NEON™ single instruction, multiple data (SIMD) coprocessor with vector
floating-point (VFP) v3 double-precision floating point unit for media and signal
processing acceleration
■ Single- and double-precision IEEE-754 floating point math support
■ Integer and polynomial math support
■ Level 1 (L1) cache with parity checking
■ 32 KB four-way set-associative instruction cache
■ 32 KB four-way set-associative data cache
■ CoreSight™ Program Trace Macrocell (PTM) supporting instruction trace
Each Cortex-A9 processor supports the following features:
■ Dual-issue superscalar pipeline with advanced branch prediction
■ Out-of-order (OoO) dispatch and speculative instruction execution

■ 2.5 million instructions per second (MIPS) per MHz, based on the Dhrystone 2.1
benchmark
■ 128-entry translation lookaside buffer (TLB)
■ TrustZone security extensions
■ Configurable data endianness
■ Jazelle® DBX Extensions for byte-code dynamic compiler support
■ The Cortex-A9 processor architecture supports the following instruction sets:
■ The ARMv7-A performance-optimized instruction set
■ The memory-optimized Thumb®-2 mixed instruction set
■ Improves energy efficiency
■ 31% smaller memory footprint
■ 38% faster than the original Thumb instruction set
■ The Thumb instruction set—supported for legacy applications
■ Each processor core in the Altera HPS includes a memory management unit
(MMU) to support the memory management requirements of common modern
operating systems.
The Cortex-A9 processors are designated CPU0 and CPU1.
f Detailed documentation of ARM Cortex-A9 series processors, Revision r3p0, is

Interactive Debugging Features

Each Cortex-A9 processor has built-in debugging capabilities, including the following
features:
■ Six hardware breakpoints, including two with Context ID comparison capability
■ Four watchpoints
The interactive debugging features can be controlled by external JTAG tools or by
processor-based monitor code.
f For more information about the interactive debugging system, refer to the Debug
chapter of the Cortex-A9 Technical Reference Manual, Revision r3p0, available on the
ARM website (infocenter.arm.com).
L1 Caches
Cache memory that is closely coupled with an associated processor is called level 1, or
L1 cache. Each Cortex-A9 processor has two independent 32 KB L1 caches—one for
instructions and one for data—allowing simultaneous instruction fetches and data
access.
Each L1 cache is four-way set associative, with 32 bytes per line, and supports parity
checking.

Preload Engine
The preload engine (PLE) is a hardware block that enables the L2 cache to preload
selected regions of memory. The PLE signals the L2 cache when a cache line will be
needed in the L2 cache, by making the processor data master port start fetching the
data. The processor data master does not complete the fetch or return the data to the
processor. However, the L2 cache can then proceed to load the cache line. The data is
only loaded to the L2 cache, not to the L1 cache or processor registers.
The preload functionality is under software control. The following PLE control
parameters must be programmed:
■ Programmed parameters, including the following:
■ Base address
■ Length of stride
■ Number of blocks
■ A valid bit
■ TrustZone memory protection for the cache memory, with an NS (non-secure) state
bit
■ A translation table base (TTB) address
■ An Address Space Identifier (ASID) value
f For more information about the PLE, refer to the Preload Engine chapter of the
Cortex-A9 Technical Reference Manual, Revision r3p0, available on the ARM website
Floating Point Unit

Each ARM Cortex-A9 processor includes full support for IEEE-754 floating point
operations. The floating-point unit (FPU) fully supports half-, single-, and
double-precision variants of the following operations:
■ Add
■ Subtract
■ Multiply
■ Divide
■ Multiply and accumulate (MAC)
■ Square root
The FPU also converts between floating-point data formats and integers, including
special operations to round towards zero required by high-level languages.

NEON Multimedia Processing Engine

The NEON multimedia processing engine (MPE) provides hardware acceleration for
media and signal processing applications. Each ARM Cortex-A9 processor includes
an ARM NEON MPE that supports SIMD processing, as shown in Figure 6–3. The
NEON processing engine accelerates multimedia and signal processing algorithms
such as video encoding and decoding, 2-D and 3-D graphics, audio and speech
processing, image processing, telephony, and sound synthesis.
Figure 6–3. Single Instruction, Multiple Data (SIMD) Processing

Single Instruction Multiple Data
Source Register
Source Register
Op Op Op Op
Destination Register
The Cortex-A9 NEON MPE performs the following types of operations:

■ SIMD and scalar single-precision floating-point computations
■ Scalar double-precision floating-point computation
■ SIMD and scalar half-precision floating-point conversion
■ 8-bit, 16-bit, 32-bit, and 64-bit signed and unsigned integer SIMD computation
■ 8-bit or 16-bit polynomial computation for single-bit coefficients
The following operations are available:
■ Addition and subtraction
■ Multiplication with optional accumulation (MAC)
■ Maximum or minimum value driven lane selection operations
■ Inverse square root approximation
■ Comprehensive data-structure load instructions, including register-bank-resident
table lookup
f For more information about the Cortex-A9 NEON MPE, refer to the Cortex-A9
NEON™ Media Processing Engine Technical Reference Manual, Revision r3p0, which you
can download from the ARM website (infocenter.arm.com).

Memory Management Unit

The MMU is used in conjunction with the L1 and L2 caches to translate virtual
addresses used by software to physical addresses used by hardware. Each processor
has a private MMU.
The MMU supports the TLBs shown in Table 6–2.
Table 6–2. Supported TLBs

TLB Type Memory Type Number of Entries Associativity
Micro TLB Instruction 32 Fully associative
Micro TLB Data 32 Fully associative
Main TLB Instruction and Data 128 Two-way associative
The main TLB has the following features:

■ Lockable entries using the lock-by-entry model
■ Supports hardware page table walks to perform look-ups in the L1 data cache
f For more information about the MMU, refer to the Memory Management Unit chapter
of the Cortex-A9 Technical Reference Manual, Revision r3p0, available on the ARM
website (infocenter.arm.com).
The MPU address map is divided into the following regions:

■ The boot region
■ The SDRAM region
■ The FPGA slaves region
■ The HPS peripherals region
This section describes the location and contents of each region.
The Boot Region

The boot region is 1 MB in size, based at address 0. After power-on, or after reset of
the L3 interconnect, the boot region is occupied by the boot ROM, allowing the
Cortex-A9 MPCore to boot. Although the boot region size is 1 MB, accesses beyond
64 KB are illegal because the boot ROM is only 64 KB.
As shown in Figure 6–4, this 1 MB region can be subsequently remapped to the
bottom 1 MB of SDRAM. For more information, refer to “The SDRAM Region”.
1 Alternatively, the boot region can be mapped to the 64 KB on-chip RAM. For more
information, refer to the Interconnect chapter in volume 3 of the Cyclone V Device
Handbook.

Figure 6–4. The MPCore Address Map

Addresses Are Not To Scale
0xFFFF FFFF
HPS Peripherals (64 MB)
0xFC00 0000
HPS-to-FPGA
(FPGA-Based Peripherals)
0xC000 0000 (3 GB) L2 Cache Filtering
0x8000 0000 (2 GB)
SDRAM
Mapping Options
On-Chip RAM (64 KB)
0x4000 0000 (1 GB)

Boot ROM (64 KB)
SDRAM (1 MB)
0x1000 0000 (1 MB) (Mapping Provided by
Boot Region L2 Cache Filtering)
0
The SDRAM Region

The SDRAM region starts at address 0x100000 (1 MB). The top of the region is
determined by the L2 cache filter.
The L2 cache contains a filtering mechanism that routes accesses to the SDRAM and
L3 interconnect. The filter defines a filter range with start and end addresses. Any
access within this filter range is routed to the SDRAM subsystem. Accesses outside of
this filter range are routed to the L3 interconnect.
The start and end addresses are specified in the following register fields:
■ reg12_addr_filtering_start.address_filtering_start
■ reg12_address_filtering_end.address_filtering_end
To remap the lower 1MB of SDRAM into the boot region, set the filter start address to
0x0 to ensure accesses between 0x0 and 0xFFFFF are routed to the SDRAM.
Independently, you can set the filter end address in 1 MB increments above
0xC0000000 to extend the upper bounds of the SDRAM region. However, you achieve
this extended range at the expense of the FPGA peripheral address span. Depending
on the address filter settings in the L2 cache, the top of the SDRAM region can range
from 0xBFFFFFFF to 0xFBFFFFFF.
For more information about the L2 cache, refer to “L2 Cache” on page 6–24.

The FPGA Slaves Region

The Cortex-A9 MPU subsystem supports the variable-sized FPGA slaves region to
communicate with FPGA-based peripherals. This region can start as low as
0xC0000000, depending on the L2 cache filter settings. The top of the FPGA slaves
region is located at 0xFBFFFFFF. As a result, the size of the FPGA slaves region can
range from 0 to 0x3F000000 bytes.
The HPS Peripherals Region

The HPS peripherals region is the top 64 MB in the address space, starting at
0xFC000000 and extending to 0xFFFFFFFF. The HPS peripherals region is always
allocated to the HPS dedicated peripherals for the Altera Cortex-A9 MPU subsystem.
Performance Monitoring Unit

Each Cortex-A9 processor has a Performance Monitoring Unit (PMU). The PMU
supports 58 events to gather statistics on the operation of the processor and memory
system. Six counters in the PMU accumulate the events in real time. The PMU
counters are accessible either from the processor itself, using the Coprocessor 14
(CP14) interface, or from an external debugger. The events are also supplied to the
PTM and can be used for trigger or trace.
f For more information about the PMU, refer to the Performance Monitoring Unit chapter
of the Cortex-A9 Technical Reference Manual, Revision r3p0, available on the ARM
MPCore Timers
There is one interval timer and one watchdog timer for each processor.
Each timer is private, meaning that only its associated processor can access it. If the
watchdog timer is not needed, it can be configured as a second interval timer.
Each private interval and watchdog timer has the following features:
■ A 32-bit counter that optionally generates an interrupt when it reaches zero
■ Configurable starting values for the counter
■ An eight-bit prescaler value to qualify the clock period
The timers are configurable to either single-shot or auto-reload mode. The timer
blocks are clocked by mpu_periph_clk, running at ¼ the rate of mpu_clk.
f For more information about private timers, refer to “About the private timer and
watchdog blocks” in the Global timer, Private timers, and Watchdog registers chapter of
the Cortex-A9 MPCore Technical Reference Manual, Revision r3p0, available on the ARM

Generic Interrupt Controller

The Generic Interrupt Controller (GIC) supports up to 180 interrupt sources,
including dedicated peripherals and IP implemented in the FPGA fabric. In a
dual-core system, the GIC is shared by both Cortex-A9 processors. Each processor
also has 16 banked software-generated interrupts and 16 banked private peripheral
interrupts.
The configuration and control for the GIC is memory-mapped and accessed through
the SCU. The GIC are clocked by mpu_periph_clk, running at ¼ the rate of mpu_clk.
f For more information about the GIC, refer to the Interrupt Controller chapter of the
Cortex-A9 MPCore Technical Reference Manual, Revision r3p0, available on the ARM
Table 6–3 shows the interrupt map.
Table 6–3. GIC Interrupt Map (Part 1 of 6)

GIC
Combined
Interrupt Source Block Interrupt Name Triggering
Interrupts
Number (1)
32 CortexA9_0 cpu0_parityfail (2) Edge
33 CortexA9_0 cpu0_parityfail_BTAC Edge
34 CortexA9_0 cpu0_parityfail_GHB Edge
35 CortexA9_0 cpu0_parityfail_I_Tag Edge
36 CortexA9_0 cpu0_parityfail_I_Data Edge
37 CortexA9_0 cpu0_parityfail_TLB Edge
38 CortexA9_0 cpu0_parityfail_D_Outer Edge
39 CortexA9_0 cpu0_parityfail_D_Tag Edge
40 CortexA9_0 cpu0_parityfail_D_Data Edge
41 CortexA9_0 cpu0_deflags0 Level
48 CortexA9_1 cpu1_parityfail (3) Edge
49 CortexA9_1 cpu1_parityfail_BTAC Edge
50 CortexA9_1 cpu1_parityfail_GHB Edge
51 CortexA9_1 cpu1_parityfail_I_Tag Edge
52 CortexA9_1 cpu1_parityfail_I_Data Edge
53 CortexA9_1 cpu1_parityfail_TLB Edge
54 CortexA9_1 cpu1_parityfail_D_Outer Edge
55 CortexA9_1 cpu1_parityfail_D_Tag Edge
56 CortexA9_1 cpu1_parityfail_D_Data Edge


GIC
Combined
Interrupts
Number (1)
64 SCU scu_parityfail0 — Edge
65 SCU scu_parityfail1 — Edge
66 SCU scu_ev_abort Edge
67 L2-Cache l2_ecc_byte_wr_IRQ — Edge
68 L2-Cache l2_ecc_corrected_IRQ — Edge
69 L2-Cache l2_ecc_uncorrected_IRQ — Edge
70 L2-Cache l2_combined_IRQ (4) Level
71 DDR ddr_ecc_error_IRQ — Level
72 FPGA FPGA_IRQ0 — Level or Edge


GIC
Combined
Interrupts
Number (1)


GIC
Combined
Interrupts
Number (1)
136 DMA dma_IRQ0 — Level
144 DMA dma_irq_abort — Level
145 DMA dma_ecc_corrected_IRQ Level
146 DMA dma_ecc_uncorrected_IRQ Level
147 EMAC0 emac0_IRQ (5) Level
148 EMAC0 emac0_tx_ecc_corrected_IRQ Level
149 EMAC0 emac0_tx_ecc_uncorrected_IRQ Level
150 EMAC0 emac0_rx_ecc_corrected_IRQ Level
151 EMAC0 emac0_rx_ecc_uncorrected_IRQ Level
152 EMAC1 emac1_IRQ (5) Level
153 EMAC1 emac1_tx_ecc_corrected_IRQ Level
154 EMAC1 emac1_tx_ecc_uncorrected_IRQ Level
155 EMAC1 emac1_rx_ecc_corrected_IRQ Level
156 EMAC1 emac1_rx_ecc_uncorrected_IRQ Level
157 USB0 usb0_IRQ Level
158 USB0 usb0_ecc_corrected_IRQ Level
159 USB0 usb0_ecc_uncorrected_IRQ Level
160 USB1 usb1_IRQ Level
161 USB1 usb1_ecc_corrected_IRQ Level
162 USB1 usb1_ecc_uncorrected_IRQ Level
163 CAN0 can0_sts_IRQ Level
164 CAN0 can0_mo_IRQ Level
165 CAN0 can0_ecc_corrected_IRQ Level
166 CAN0 can0_ecc_uncorrected_IRQ Level
167 CAN1 can1_sts_IRQ Level
168 CAN1 can1_mo_IRQ Level
169 CAN1 can1_ecc_corrected_IRQ Level
170 CAN1 can1_ecc_uncorrected_IRQ Level
171 SDMMC sdmmc_IRQ Level
172 SDMMC sdmmc_porta_ecc_corrected_IRQ Level
173 SDMMC sdmmc_porta_ecc_uncorrected_IRQ Level


GIC
Combined
Interrupts
Number (1)
174 SDMMC sdmmc_portb_ecc_corrected_IRQ Level
175 SDMMC sdmmc_portb_ecc_uncorrected_IRQ Level
176 NAND nand_IRQ Level
177 NAND nandr_ecc_corrected_IRQ Level
178 NAND nandr_ecc_uncorrected_IRQ Level
179 NAND nandw_ecc_corrected_IRQ Level
180 NAND nandw_ecc_uncorrected_IRQ Level
181 NAND nande_ecc_corrected_IRQ Level
182 NAND nande_ecc_uncorrected_IRQ Level
183 QSPI qspi_IRQ Level
184 QSPI qspi_ecc_corrected_IRQ Level
185 QSPI qspi_ecc_uncorrected_IRQ Level
186 SPI0 spi0_IRQ (6) Level
190 I2C0 i2c0_IRQ (7) Level
194 UART0 uart0_IRQ Level
195 UART1 uart1_IRQ Level
196 GPIO0 gpio0_IRQ — Level
199 Timer0 timer_l4sp_0_IRQ (8) Level
200 Timer1 timer_l4sp_1_IRQ (8) Level
201 Timer2 timer_osc1_0_IRQ (8) Level
202 Timer3 timer_osc1_1_IRQ (8) Level
203 Watchdog0 wdog0_IRQ — Level
204 Watchdog1 wdog1_IRQ — Level
205 Clock manager clkmgr_IRQ Level
206 Clock manager mpuwakeup_IRQ Level
207 FPGA manager fpga_man_IRQ (9) Level
208 CoreSight nCTIIRQ[0] Level
209 CoreSight nCTIIRQ[1] Level


GIC
Combined
Interrupts
Number (1)
210 On-chip RAM ram_ecc_corrected_IRQ Level
211 On-chip RAM ram_ecc_uncorrected_IRQ Level
(1) To ensure that you are using the correct GIC interrupt number, your code should refer to the symbolic interrupt
name, as shown in the Interrupt Name column. Symbolic interrupt names are defined in a header file distributed
with the source installation for your operating system.
(2) This interrupt combines the interrupts named cpu0_parityfail_*.
(3) This interrupt combines the interrupts named cpu1_parityfail_*.
(4) This interrupt combines the following interrupts: DECERRINTR, ECNTRINTR, ERRRDINTR, ERRRTINTR, ERRWDINTR,
ERRWTINTR, PARRDINTR, PARRTINTR, and SLVERRINTR.
(5) This interrupt combines sbd_intr_o, lpi_intr_o, and pmt_intr_o.
(6) This interrupt combines the following interrupts: ssi_txe_intr, ssi_txo_intr, ssi_rxf_intr,
ssi_rxo_intr, ssi_rxu_intr, and ssi_mst_intr.
(7) This interrupt combines the following interrupts: ic_rx_under_intr, ic_rx_full_intr, ic_tx_over_intr,
ic_tx_empty_intr, ic_rd_req_intr, ic_tx_abrt_intr, ic_rx_done_intr, ic_activity_intr,
ic_stop_det_intr, ic_start_det_intr, and ic_gen_call_intr.
(8) This interrupt combines TIMINT1 and TIMINT2.
(9) This interrupt combines the following interrupts: fpga_man_irq[7..0].
Global Timer
The MPU features a global 64-bit, auto-incrementing timer, which is primarily used by
the operating system.
The global timer is accessible by the processors using memory-mapped access
through the SCU. The global timer has the following features:
■ 64-bit incrementing counter with an auto-incrementing feature. It continues
incrementing after sending interrupts.
■ Memory-mapped in the private memory region.
■ Accessed at reset in Secure State only. It can only be set once, but secure code can
read it at any time.
■ Accessible to both Cortex-A9 processors in the MPCore.
Each Cortex-A9 processor has a private 64-bit comparator that generates a private
interrupt when the counter reaches the specified value. Each Cortex-A9 processor
uses the banked ID, ID27, for this interrupt. ID27 is sent to the GIC as a Private
Peripheral Interrupt (PPI).
The global timer are clocked by mpu_periph_clk, running at ¼ the rate of mpu_clk.
f For more information about the global timer, refer to “About the Global Timer” in the
Global timer, Private timers, and Watchdog registers chapter of the Cortex-A9 MPCore
Technical Reference Manual, Revision r3p0, available on the ARM website

Snoop Control Unit

The SCU manages data traffic for the Cortex-A9 processors and the memory system,
including the L2 cache. In a multi-master system, the processors and other masters
can operate on shared data. The SCU ensures that each processor operates on the most
up-to-date copy of data, maintaining cache coherency.
The SCU is used to connect the Cortex-A9 processors and the ACP to the L2 cache
controller. The SCU performs the following functions:
■ When the processors are set to SMP mode, the SCU maintains data cache
coherency between the processors.
1 The SCU does not maintain coherency of the instruction caches.
■ Initiates L2 cache memory accesses

■ Arbitrates between processors requesting L2 access
■ Manages ACP access with cache coherency capabilities.
Figure 6–5 shows the SCU in a dual-processor system, illustrating the flow of data
among the L1 data caches and the SCU.
Figure 6–5. Coherent Memory, Snoop Control Unit, and Accelerator Coherency Port
ARM Cortex-A9 ARM Cortex-A9

32-Bit Dual-Issue 32-Bit Dual-Issue
Superscalar Superscalar
RISC Processor RISC Processor
L1 Data 32 KB L1 Data 32 KB
Cache Instruction Cache Instruction
FPGA Fabric
Cache Cache
ID Mapper
Snoop Control Unit (SCU)

HPS Level 3 (L3) Accelerator
Mastering Interconnect Coherency
Level 2 (L2) Unified Cache
Peripherals Port (ACP)
ACP
Unidirectional Bidirectional Coherent

Coherency Coherency Memory
f For more information about the SCU, refer to the Snoop Control Unit chapter of the

When the processor writes to any coherent memory location, the SCU ensures that the
relevant data is coherent (updated, tagged, or invalidated). Similarly, the SCU
monitors read operations from a coherent memory location. If the required data is
already stored within the other processor’s L1 cache, the data is returned directly to
the requesting processor. If the data is not in L1 cache, the SCU issues a read to the L2
cache. If the data is not in the L2 cache memory, the read is finally forwarded to main
memory. The primary goal is to minimize power consumption and maximize overall
memory performance.
The SCU maintains bidirectional coherency between the L1 data caches belonging to
the processors. When one processor writes to a location in its L1 cache, if the same
location is cached in the other L1 cache, the SCU updates it.
Non-coherent data passes through as a standard read or write operation.
The SCU also arbitrates between the Cortex-A9 processors if both attempt
simultaneous access to the L2 cache, and manages accesses from the ACP.
Accelerator Coherency Port

The ACP allows peripherals—including FPGA-based peripherals—to maintain data
coherency with the Cortex-A9 MPCore processors and the SCU. As shown in
Figure 6–5 on page 6–18, dedicated peripherals in the HPS, and those built in FPGA
logic, access the coherent memory through the ACP ID mapper and the ACP. For
information about the ACP ID mapper, refer to “ACP ID Mapper” on page 6–20.
The high-bandwidth peripherals, including the FPGA data ports, connect to the L3
interconnect.
Burst Sizes and Byte Strobes

The ACP improves system performance for hardware accelerators in the FPGA fabric.
However, in order to achieve high levels of performance, you must use the burst types
listed in Table 6–4. The other burst types have significantly lower performance.
Table 6–4. Recommended Burst Types

Burst Type Beats Width (Bits) Address Type Byte Strobes
Wrapping 4 64 64-bit aligned Asserted
Incrementing 4 64 32-bit aligned Asserted
1 If the slave port of the FPGA-to-HPS bridge is not 64 bits wide, you must supply
bursts to the FPGA-to-HPS bridge that are upsized or downsized to the burst types
above. For example, if the slave data width of the FPGA-to-HPS bridge is 32 bits, then
bursts of eight beats by 32 bits are required to access the ACP efficiently.
c If the address and burst size of the transaction to the ACP matches either of the
conditions above, the logic in the MPU assumes the transaction has all its byte strobes
set. If the byte strobes are not all set, then the write does not actually overwrite all the
bytes in the word. Instead, the cache assumes the whole cache line is valid. If this line
is dirty (and therefore gets written out to SDRAM), data corruption might occur.

Exclusive and Locked Accesses

The ACP does not support exclusive accesses to coherent memory. The ACP supports
exclusive accesses to non-coherent memory; however, it is important that the
exclusive access transaction is not affected by the upsizing and downsizing logic of
the FPGA-to-HPS bridge or the L3 interconnect. If the exclusive access is broken into
multiple transactions due to the sizing logic, the exclusive access bit is cleared by the
bridge or interconnect and the exclusive access fails.
1 Altera recommends that exclusive accesses bypass the ACP altogether, either through
the 32-bit slave port of the SDRAM controller connected directly to the L3
interconnect or through the FPGA-to-SDRAM interface.
f For more information about the exclusive access support of the SDRAM controller
subsystem, refer to the SDRAM Controller Subsystem chapter in volume 3 of the
Cyclone® V Device Handbook.
The ACP ID mapper does not support locked accesses. To ensure mutually exclusive
access to shared data, use the exclusive access support built into the SDRAM
controller.
ACP ID Mapper
The ACP ID mapper is situated between the level 3 (L3) interconnect and the MPU
subsystem ACP slave. It is responsible for mapping 12-bit Advanced Microcontroller
Bus Architecture (AMBA®) Advanced eXtensible Interface (AXI™) IDs (input IDs)
from the L3 interconnect to 3-bit AXI IDs (output IDs) supported by the ACP slave
port.
The ACP ID mapper also implements a 1 GB coherent window into 4 GB address
space.
The ACP slave supports up to six masters. However, custom peripherals
implemented in the FPGA fabric can have a larger number of masters that need to
access the ACP slave. The ACP ID mapper allows these masters to access the ACP.
The ACP ID mapper resides between the interconnect and the ACP slave of the MPU
subsystem. It has the following characteristics:
■ Support for up to six concurrent ID mappings
■ 1 GB coherent window into 4 GB MPCore address space
■ Remaps the 5-bit user sideband signals used by the Snoop Control Unit (SCU) and
L2 cache.
f For more information about AXI user sideband signals, refer to the CoreLink Level 2
Cache Controller L2C-310 Technical Reference Manual, which you can download from the

The ACP is accessed by masters that require access to coherent memory. The ACP
slave port can be accessed by the master peripherals of the L3 interconnect, as well as
by masters implemented in the FPGA fabric (via the FPGA-to-HPS bridge). Figure 6–1
on page 6–2 shows the ACP ID mapper.
The ACP ID mapper supports the following ID mapping modes:
■ Dynamic mapping
■ Fixed mapping
Software can select the ID mapping on a per-ID basis. For input IDs that are
configured for fixed mapping, there is a one-to-one mapping from input IDs to output
IDs. When an input ID is configured for dynamic mapping, it is automatically
mapped to an available output ID. The dynamic mode is more flexible because the
hardware handles the mapping. The hardware mapping allows you to use one output
ID for more than one input ID. Output IDs are assigned to input IDs on a first-come,
first-served basis.
Out of the total of eight output IDs, only six are available to masters of the L3
interconnect. The first two output IDs (0 and 1) are dedicated to the Cortex-A9
processor cores in the MPU subsystem, leaving the last six output IDs (2-7) available
to the ACP ID mapper. Output IDs 2-6 support fixed and dynamic modes of operation
while output ID 7 supports dynamic only.
The operating modes are programmable through accesses to the control and status
registers in the ACP ID mapper, available via the level 4 peripheral bus connection. At
reset time, the ACP ID mapper defaults to dynamic ID mapping for all output IDs
except ID 2, which resets to a fixed mapping for the Debug Access Port (DAP) input
ID.
Table 6–1 summarizes the expected usage of the 3-bit output IDs, and their settings at
reset.
Table 6–5. ID Intended Usage

Output
Reset State Intended Use
ID
7 Dynamic Dynamic mapping only
6
5
Dynamic Fixed or dynamic, programmed by software.
4
3
Fixed at Assigned to the input ID of the DAP at reset. After reset, can be either
2
0x001 (DAP) fixed or dynamic, programmed by software.
1
— Not used by the ACP ID Mapper
0

For masters that cannot drive the AXI user sideband signal of incoming transactions,
the ACP ID mapper can control overriding this signal. The ACP ID mapper can also
control which 1 GB coherent window into memory is accessed by masters of the L3
interconnect. Each fixed mapping can be assigned a different user sideband signal and
memory window to allow specific settings for different masters. All dynamic
mappings share a common user sideband signal and memory window setting.
Transaction Capabilities
At any one time, the ACP ID mapper can accept and issue up to 15 transactions per ID
mapping. Read and write ID mappings are managed in separate lists, allowing more
unique input IDs to be remapped at any given time. If a master issues a series of reads
and writes with the same input ID, there are no ordering restrictions.
Because there are only six output IDs available, there can be no more than six read and
six write transactions with unique IDs in progress at any one time. The write
acceptance of the ACP slave is five transactions, and the read acceptance is 13
transactions. Only four coherent read transactions per ID mapping can be outstanding
at one time.
Dynamic Mapping Mode

In dynamic mode, every unique input ID that is received from the L3 master port is
assigned to an unused output ID. The new output ID is applied to the transaction as it
is issued to the ACP slave of the SCU. Any transaction that arrives to the ACP ID
mapper with an input ID that matches an already-in-progress transaction is mapped
to the same output ID. Once all transactions on an ID mapping have completed, that
output ID is released and can be used again for other input IDs.
Fixed Mapping Mode

In fixed mode, output IDs 2 through 6 can be assigned by software to a specific 12-bit
input ID. This ability makes it possible to use the lock-by-master feature of the L2
cache controller, because the input transaction ID from the master is always assigned
to a specific output ID. Unlike dynamic mode, ID 7 is not available for fixed mapping
because it is reserved for dynamic mode only to avoid system deadlocks.
The ACP ID mapper has two banks of registers to control the behavior of the
mappings, namely, a request bank and a read-only status bank. Both banks contain
the same number of registers. To change the settings for a particular mapping (either a
specific fixed ID, or all dynamic mappings), software should write to the appropriate
register in the request bank. The hardware examines the request, and only applies the
change when safe to do so, which is when there are no outstanding transactions with
the output ID. When the change is applied, the status register is updated. Software
should check that the change has actually taken place by polling the corresponding
status register.

Table 6–6 shows the input IDs issued from the interconnect for each HPS peripheral
master that can access the ACP ID mapper.
Table 6–6. HPS Peripheral Master Input IDs

Interconnect Master ID (1)
DMA 00000xxxx011
EMAC0 10000xxxx001
EMAC1 10000xxxx010
USB0 100000000011
USB1 100000000110
NAND 1xxxxxxxx100
ETR 100000000000
DAP 000000000001
SD/MMC 100000000101
FPGA-to-HPS bridge 0xxxxxxxx100
(1) Values are in binary. The letter x denotes variable ID bits each master passes with each transaction.
Control of the AXI User Sideband Signals

The ACP ID mapper module allows control of the AXI user sideband signal values.
Not all masters drive these signals, so the ACP ID mapper makes it possible to drive
the 5-bit user sideband signal with either a default value (in dynamic mode) or
specific values (in fixed mode).
There are registers available to configure the default values of the user sideband
signals for all transactions, and fixed values of these signals for particular transactions
in fixed mapping mode. In dynamic mode, the user sideband signals of incoming
transactions are mapped with the default values stored in the register. In fixed
mapping mode, the input ID of the transaction is mapped to the 3-bit output ID and
the user sideband signals of the transaction are mapped with the values stored in the
register that corresponds to the output ID. One important exception, however, is that
the ACP ID mapper always allows user sideband signals from the FPGA-to-HPS
bridge to pass through to the ACP regardless of the user sideband value associated
with the ID.

Memory Region Remap

The ACP ID mapper has 1 GB of address space, which is by default a view into the
bottom 1 GB of SDRAM. The mapper also allows transactions to be routed to different
1 GB-sized memory regions, called pages, in both dynamic and fixed modes. The two
most significant bits of incoming 32-bit AXI address signals are replaced with the 2-bit
user-configured address page decode information. The page decoder uses the values
shown in Table 6–7.
Table 6–7. Page Decoder Values

Page Address Range
0 0x00000000—0x3FFFFFFF
1 0x40000000—0x7FFFFFFF
2 0x80000000—0xBFFFFFFF
3 0xC0000000—0xFFFFFFFF
With this page decode information, a master can read or write to any 1 GB region of
the 4 GB memory space while maintaining cache coherency with the MPU subsystem.
Using this feature, a debugger can have a coherent view into main memory, without
having to stop the processor. For example, at reset the DAP input ID (0x001) is
mapped to output ID 2, so the debugger can vary the 1 GB window that the DAP
accesses without affecting any other traffic flow to the ACP.
L2 Cache
The MPU subsystem includes a secondary 512 KB L2 shared, unified cache memory.
The L2 cache is much larger than the L1 cache. The L2 cache has significantly lower
latency than external memory. The L2 cache is up to eight-way associative,
configurable down to one-way (direct mapped). Like the L1 cache, the L2 cache can be
locked by cache line, locked by way, or locked by bus master.
The L2 cache implements error correction codes (ECCs) and ECC error reporting. The
cache can report a number of events to the processor and operating system.
The L2 cache consists of the ARM L2C-310 L2 cache controller configured as follows:
■ 512 KB total memory
■ Eight-way associativity
■ Physically addressed, physically tagged
■ Line length of 32 bytes

■ Critical first word linefills

■ Support for all AXI cache modes, as shown in Table 6–8.
Table 6–8. AXI Cache Mode Support

Cache Mode
Write-through (1)
Write-back (1)
Read allocate
Write allocate
Read and write allocate
(1) Restrictions exist when using ECCs. For more information about SEU protection, refer to the System Manager
chapter in volume 3of the Cyclone V Device Handbook.
■ Single event upset (SEU) protection

■ Parity on Tag RAM
■ ECC on L2 Data RAM
f For more information about SEU protection, refer to the System Manager
chapter in volume 3 of the Cyclone V Device Handbook.
■ Two slave ports mastered by the SCU

■ Two master ports connected to the following slave ports:
■ SDRAM controller, 64 bit slave port width
■ L3 interconnect, 64 bit slave port width
■ Cache lockdown capabilities as follows:
■ Line lockdown
■ Lockdown by way
■ Lockdown by master (both processors and ACP masters)
■ TrustZone support
■ Cache event monitoring. For more information, refer to “L2 Cache Event
Monitoring” on page 6–27.
Figure 6–1 on page 6–2 shows the L2 cache. The L2 cache can access either the L3
interconnect fabric or the SDRAM. The L2 cache address filtering determines how
much address space is allocated to the HPS-to-FPGA bridge and how much is
allocated to SDRAM, as described in “Memory Management Unit” on page 6–9.
ECC Support
The L2 cache has the option of using ECCs to protect against SEU errors in the cache
RAM.

Enabling ECCs does not affect the performance of the L2 cache. The ECC bits are
calculated only for writes to the data RAM that are 64 bits wide (8 bytes, or
one-quarter of the cache line length). The ECC logic does not perform a
read-modify-write when calculating the ECC bits. The ECC protection bits are not
valid in the following cases:
■ Data is written that is not 64-bit aligned in memory
■ Data is written that is less than 64 bits in width
In these cases the Byte Write Error interrupt is asserted. Cache data is still written
when such an error occurs. However, the ECC error detection and correction
continues to function. Therefore, the cache data is likely to be incorrect on subsequent
reads.
To use ECCs, the software and system must meet the following requirements:
■ L1 and L2 cache must be configured as write-back allocate for any cacheable
memory region
■ FPGA soft IP using the ACP must only perform the following types of data writes:
■ 64-bit aligned in memory
■ 64 bit wide accesses
f For more information about SEU errors, refer to the System Manager chapter in
Table 6–9 shows the parameter settings for the cache controller.
Table 6–9. Cache Controller Configuration

Feature Meaning
Cache way size 64 KB
Number of cache ways 8 ways
RAM latencies 2 cycles of latency
Parity logic Parity logic enabled
Lockdown by master Lockdown by master enabled
Lockdown by line Lockdown by line enabled
AXI ID width on slave ports 6 AXI ID bits on slave ports
Address filtering Address filtering logic enabled
Speculative read Logic for supporting speculative read enabled
Presence of ARUSERMx and AWUSERMx
Sideband signals enabled
sideband signals
f For further information about cache controller configurable options, refer to the
CoreLink Level 2 Cache Controller L2C-310 Technical Reference Manual, Revision r3p2,

L2 Cache Lockdown Capabilities

The L2 cache has three methods to lock data in the cache RAMs:
■ Lockdown by line—Used to lock lines in the cache. This is commonly used for
loading critical sections of software into the cache temporarily.
■ Lockdown by way—Allows any or all of the eight cache ways to be locked. This is
commonly used for loading critical data or code into the cache.
■ Lockdown by master—Allows cache ways to be dedicated to a single master port.
This allows a large cache to look like smaller caches to multiple master ports. The
L2 cache can be mastered by CPU0, CPU1, or the six ACP masters, for a total of
eight possible master ports.
f For more information about L2 cache lockdown capabilities, refer to “Cache

operation” in the Functional Overview chapter of the CoreLink Level 2 Cache Controller
L2C-310 Technical Reference Manual, Revision r3p2, available on the ARM website
L2 Cache Event Monitoring

The L2 cache supports the built-in cache event monitoring signals shown in
Table 6–10. The L2 cache can count two of the events at any one time.
Table 6–10. L2 Cache Events

Event Description
CO Eviction (cast out) of a line from the L2 cache.
DRHIT Data read hit in the L2 cache.
DRREQ Data read lookup to the L2 cache. Subsequently results in a hit or miss.
DWHIT Data write hit in the L2 cache.
DWREQ Data write lookup to the L2 cache. Subsequently results in a hit or miss.
DWTREQ Data write lookup to the L2 cache with write-through attribute. Subsequently results in a hit or miss.
EPFALLOC Prefetch hint allocated into the L2 cache.
EPFHIT Prefetch hint hits in the L2 cache.
EPFRCVDS0 Prefetch hint received by slave port S0.
EPFRCVDS1 Prefetch hint received by slave port S1.
IPFALLOC Allocation of a prefetch generated by L2 cache controller into the L2 cache.
IRHIT Instruction read hit in the L2 cache.
IRREQ Instruction read lookup to the L2 cache. Subsequently results in a hit or miss.
SPNIDEN Secure privileged non-invasive debug enable.
SRCONFS0 Speculative read confirmed in slave port S0.
SRCONFS1 Speculative read confirmed in slave port S1.
SRRCVDS0 Speculative read received by slave port S0.
SRRCVDS1 Speculative read received by slave port S1.
WA Allocation into the L2 cache caused by a write, with write-allocate attribute, miss.

f For more information about the built-in L2 event monitoring capability, refer to
“Implementation details” in the Functional Overview chapter of the CoreLink Level 2
Cache Controller L2C-310 Technical Reference Manual, Revision r3p2, available on the
In addition, the L2 cache events can be captured and timestamped using dedicated
debugging circuitry.
f For more information about L2 event capture, refer to the Debug chapter of the
Debugging Modules
The MPU subsystem includes debugging resources through ARM CoreSight on-chip
debugging and trace. The following functionality is included:
■ Individual program trace for each processor
■ Event trace for the Cortex-A9 MPCore
■ Cross triggering between processors and other HPS debugging features
Program Trace
Each processor has an independent PTM that provides real-time instruction flow
trace. The PTM is compatible with a number of third-party debugging tools.
The PTM provides trace data in a highly compressed format. The trace data includes
tags for specific points in the program execution flow, called waypoints. Waypoints
are specific events or changes in the program flow.
The PTM recognizes and tags the waypoints listed in Table 6–11.
Table 6–11. Waypoints Supported by the PTM

Type Additional Waypoint Information
Indirect branches Target address and condition code
Direct branches Condition code
Instruction barrier instructions —
Exceptions Location where the exception occurred
Changes in processor instruction set state —
Changes in processor security state —
Context ID changes —
Entry to and return from debug state when
—
Halting debug mode is enabled

Cortex-A9 MPU Subsystem Register Implementation
The PTM optionally provides additional information for waypoints, including the
following.
■ Processor cycle count between waypoints
■ Global timestamp values
f For information about global timestamps, refer to the CoreSight Debug and
Trace chapter in volume 3 of the Cyclone V Device Handbook.
■ Target addresses for direct branches
f For more information about the PTM, refer to the CoreSight PTM-A9 Technical Reference
Manual, Revision r1p0, available on the ARM website (infocenter.arm.com).
Event Trace
Events from each processor can be used as inputs to the PTM. The PTM can use these
events as trace and trigger conditions.
For more information about the event trace, refer to “Performance Monitoring Unit”
on page 6–11.
f For more information about the trigger and trace capabilities, refer to the CoreSight
PTM-A9 Technical Reference Manual, Revision r1p0, available on the ARM website
Cross-Triggering
The PTM can export trigger events and perform actions on trigger inputs. The
cross-trigger signals interface with other HPS debugging components including the
FPGA fabric. Also, a breakpoint in one processor can trigger a break in the other.
f For detailed information about cross-triggering, refer to the CoreSight Debug and Trace
f For more information about debugging hardware in the MPU, refer to the CoreSight
Debug and Trace chapter in volume 3 of the Cyclone V Device Handbook.
Cortex-A9 MPU Subsystem Register Implementation

The following configurations are available through registers in the Cortex-A9
subsystem:
■ All processor-related controls, including the MMU and L1 caches, are controlled
using the Coprocessor 15 (CP15) registers of each individual processor.
■ All SCU registers, including control for the timers and GIC, are memory map
accessible
■ All L2 cache registers are memory-mapped.

f For an address map of peripheral slave ports, including the SCU and L2 cache, refer to
the Introduction to the Hard Processor System chapter in volume 3 of the Cyclone V
Device Handbook. For detailed definitions of the registers for the Altera Cortex-A9
MPU subsystem, refer to the Cortex-A9 MPCore Technical Reference Manual, Revision
r3p0, and the CoreLink Level 2 Cache Controller L2C-310 Technical Reference Manual,
Revision r3p2, available on the ARM website (infocenter.arm.com).


■ Add description of the ACP ID mapper
May 2012 1.1
■ Consolidate redundant information

Section IV. Debug and Trace
cv_54001-1.0

■ Chapter 7, CoreSight Debug and Trace

January 2012
Subscribe
6–2 Chapter :
Cyclone V Device Handbook January 2012 Altera Corporation

7. CoreSight Debug and Trace
November 2012
cv_54007-1.2
cv_54007-1.2
The hard processor system (HPS) debug infrastructure provides visibility and control
of the HPS modules, the ARM® Cortex™-A9 microprocessor unit (MPU) subsystem,
and user logic implemented in the FPGA fabric. The debug system design
incorporates ARM® CoreSight™ components.
The HPS contains the following ARM CoreSight debug components:
■ “Debug Access Port (DAP)” on page 7–4
■ “System Trace Macrocell (STM)” on page 7–4
■ “Trace Funnel” on page 7–5
■ “Embedded Trace FIFO (ETF)” on page 7–5
■ “AMBA Trace Bus Replicator (Replicator)” on page 7–5
■ “Embedded Trace Router (ETR)” on page 7–5
■ “Trace Port Interface Unit (TPIU)” on page 7–6
■ “Embedded Cross Trigger (ECT) System” on page 7–6
■ “Program Trace Macrocell (PTM)” on page 7–10
Features of CoreSight Debug and Trace

The CoreSight debug and trace system offers the following features:
■ Real-time program flow instruction trace through a separate PTM for each
processor
■ Host debugger JTAG interface
■ Connections for cross-trigger and STM-to-FPGA interfaces, which enable soft IP
generation of triggers and system trace messages
■ Instruction trace interface through TPIU for trace analysis tools
■ Custom message injection through STM into trace stream for delivery to host
debugger
■ STM and PTM trace sources multiplexed into a single stream through the Trace
Funnel

November 2012
Subscribe
7–2 Chapter 7: CoreSight Debug and Trace
ARM CoreSight Documentation
■ Capability to route trace data to any slave accessible to the ETR AXI master
connected to the level 3 (L3) interconnect
■ Capability for the following SoC modules to trigger each other through the
embedded cross-trigger system:
■ FPGA fabric
■ A9-0 processor
■ A9-1 processor
■ PTM-0
■ PTM-1
■ STM
■ ETF
■ ETR
■ TPIU
■ csCTI
■ CTI-0
■ CTI-1
■ FPGA-CTI
■ csCTM
■ CTM
ARM CoreSight Documentation

The following ARM CoreSight specifications and documentation provide a more
thorough description of the ARM CoreSight components in the HPS debug system:
■ CoreSight Technology, System Design Guide, ARM DGI 0012D
■ CoreSight Architecture Specification, ARM IHI 0029B
■ ARM Debug Interface v5, Architecture Specification, ARM IHI 0031A
■ Embedded Cross Trigger Technical Reference Manual, ARM DDI 0291A
■ CoreSight Components Technical Reference Manual, ARM DDI 0314H
■ CoreSight Program Flow Trace, Architecture Specification, ARM IHI 0035A
■ CoreSight PTM-A9 Technical Reference Manual, ARM DDI 0401B
■ CoreSight System Trace Macrocell Technical Reference Manual, ARM DDI 0444A
■ System Trace Macrocell, Programmers' Model Architecture Specification, ARM IHI 0054
■ CoreSight Trace Memory Controller Technical Reference Manual, ARM DDI 0461B
f You can download the documents from the ARM website (infocenter.arm.com).

Chapter 7: CoreSight Debug and Trace 7–3
CoreSight Debug and Trace Block Diagram and System Integration
CoreSight Debug and Trace Block Diagram and System Integration

Figure 7–1 shows an overview block diagram of the HPS CoreSight debug and trace
system.
Figure 7–1. Debug System Block Diagram
Events
from FPGA
Funnel HPS Debug System
PTM-0 ATB On-Chip
0 Hardware Events
PTM-1 ATB Trace RAM
1 CTI Triggers
To DMA
ATB 2 ATB
3 ETF
.
STM .
Hardware Events .
[31:4] [3:0] To Trace Pins [7:0] To Pin
L3 Interconnect Main Switch 7 Multiplexer &
Replicator
Trace Pins
ATB ATB
L3 Interconnect
Output Trace [31:0]
Master Peripheral Switch ETR TPIU To FPGA
HPS Debug
Configuration ROM
System AHB
Debug APB
To FPGA
System APB
DAP
HPS JTAG Pins
I[3:2]
O[1:0] O[3:2]
csCTI
Debug I[7:4] I[1:0]
APB O[5:4] O[7:6]
0
FPGA-
csCTM 2 Triggers
CTI
1 to/from
FPGA
MPU Debug Subsystem

Debug APB
Timestamp 0 CTM 1 MPU Debug

Generator Configuration
ROM
CTI-0 CTI-1
PTM-0 ATB PTM-1 ATB

PTM-0 A9-0 A9-1 PTM-1

Functional Description of CoreSight Debug and Trace

CoreSight systems provide all the infrastructure you require to debug, monitor, and
optimize the performance of a complete HPS design. CoreSight technology addresses
the requirement for a multicore debug and trace solution with high bandwidth for
whole systems beyond the processor core.
CoreSight technology provides the following features:
■ Cross-trigger support between SoC subsystems
■ High data compression
■ Multisource trace in a single stream
■ Standard programming models for standard tool support
f For more information about the CoreSight technology, refer to the CoreSight
Components Technical Reference Manual and the CoreSight Technology System Design
Guide, which you can download from the ARM website (infocenter.arm.com).
The following sections provide brief descriptions of ARM CoreSight components

provided in the HPS debug system.
Debug Access Port (DAP)

The DAP provides the necessary ports for a host debugger to connect to and
communicate with the HPS through a JTAG interface connected to dedicated HPS
pins that is independent of the JTAG for the FPGA. The JTAG interface provided with
the DAP allows a host debugger to access various modules inside the HPS.
Additionally, a debug monitor executing on either processor can access different HPS
components by interfacing with the system Advanced Microcontroller Bus
Architecture (AMBA®) Advanced Peripheral Bus (APB™) slave port of the DAP. The
system APB slave port occupies 2 MB of address space in the HPS. Both the JTAG port
and system APB port have access to the debug APB master port of the DAP. As shown
in Figure 7–1, all CoreSight components are connected to the debug APB.
A host debugger can access any HPS memory-mapped resource in the system via the
DAP system master port. Requests made over the DAP system master port are
impacted by reads and writes to peripheral registers.
f For more information, refer to the CoreSight Components Technical Reference Manual,
which you can download from the ARM website (infocenter.arm.com).
System Trace Macrocell (STM)

The STM allows messages to be injected into the trace stream for delivery to the host
debugger receiving the trace data. These messages can be sent via stimulus ports or
the hardware event interface. The STM allows the messages to be time stamped.

The STM provides an AMBA Advanced eXtensible Interface (AXI™) slave interface
used to create trace events. The interface can be accessed by the MPU subsystem,
direct memory access (DMA) controller, and masters implemented as soft logic in the
FPGA fabric via the FPGA-to-HPS bridge. The AXI slave interface supports three
address segments, where each address segment is 16 MB and each segment supports
up to 65536 channels. Each channel occupies 256 bytes of address space.
The STM also provides 32 hardware event pins. The higher-order 28 pins (31:4) are
connected to the FPGA fabric, allowing logic inside FPGA to insert messages into the
trace stream. When the STM detects a rising edge on an event pin, a message
identifying the event is inserted into the stream. The lower four event pins (3:0) are
connected to csCTI.
f For more information, refer to the CoreSight System Trace Macrocell Technical Reference
Manual, which you can download from the ARM website (infocenter.arm.com).
Trace Funnel
The Trace Funnel multiplexes three trace sources into a single trace stream. Port 0 of
the Trace Funnel is connected to the PTM for CPU 0. Port 1 of the Trace Funnel is
connected to the PTM for CPU 1. Port 3 of the Trace Funnel is connected to the STM.
Port 2 and Port 4 through Port 7 are not used.
Embedded Trace FIFO (ETF)

The output of the Trace Funnel is sent to the ETF. The ETF is used as an elastic buffer
between trace generators (STM, PTM) and trace destinations. The ETF stores up to
32 KB of trace data in the on-chip trace RAM.
AMBA Trace Bus Replicator (Replicator)

The Replicator broadcasts trace data from the ETF to the embedded trace router (ETR)
and trace port interface unit (TPIU).
Embedded Trace Router (ETR)

The ETR can route trace data to the HPS on-chip RAM, the HPS SDRAM, and any
memory in the FPGA fabric connected to the HPS-to-FPGA bridge. The ETR receives
trace data from the Replicator. By default, the buffer to receive the trace data resides in
SDRAM at offset 0x00100000 and is 32 KB. You can override this default configuration
by programming registers in the ETR.

Trace Port Interface Unit (TPIU)

The TPIU is a bridge between on-chip trace sources and an off-chip trace port. The
TPIU receives trace data from the Replicator and drives the trace data to a trace port
analyzer.
The trace output from the TPIU is software programmable and can be set to either 8 or
32 bits wide. The trace output is routed to an 8-bit HPS I/O interface and a 32-bit
interface to the FPGA fabric. The trace data sent to the FPGA fabric can be transported
off-chip using available serializer/deserializer (SERDES) resources in the FPGA.
Embedded Cross Trigger (ECT) System

The ECT system provides a mechanism for HPS modules to trigger each other. The
ECT consists of the following modules:
■ Cross Trigger Interface (CTI)
■ Cross Trigger Matrix (CTM)

Figure 7–2 shows how CTIs and CTMs are used in a generic ECT setup. The red line
depicts an trigger input to one CTI generating a trigger output in another CTI.
Though the signal travels throughout channel 2, it only enters and exits through
trigger inputs and outputs you configure.
Figure 7–2. Generic ECT System
CTI
Trigger Outputs
Trigger Inputs
Channel
Interface
Channel Channel
Interface CTM Channel 0 Interface
Channel 1
Trigger CTI Channel 2 CTI Trigger
Interface Channel 3 Interface
Channel
Interface
CTI
Trigger
Interface
Cross Trigger Interface (CTI)

CTIs allow trigger sources and sinks to interface with the ECT. Each CTI supports up
to eight trigger inputs and eight trigger outputs, and is connected to a CTM.
Figure 7–2 shows the relationship of trigger inputs, trigger outputs, and CTM
channels of a CTI.

Figure 7–3 shows trigger input and trigger output connections in detail.
Figure 7–3. CTI Connections
CTI
Channel
Connection
Trigger Inputs Trigger Outputs
The HPS debug system contains the following four CTIs:

■ csCTI—performs cross triggering between the STM, ETF, ETR, and TPIU.
■ FPGA-CTI—exposes the cross-triggering system to the FPGA fabric.
■ CTI-0 and CTI-1—reside in the MPU debug subsystem. Each CTI is associated
with a processor and the processor’s associated PTM.
Cross Trigger Matrix (CTM)

A CTM is a transport mechanism for triggers traveling from one CTI to one or more
CTIs or CTMs. The HPS contains two CTMs. One CTM connects csCTI and FPGA-
CTI; the other connects CTI-0 and CTI-1. The two CTMs are connected together,
allowing triggers to be transmitted between the MPU debug subsystem, the debug
system, and the FPGA fabric.
Each CTM has four ports and each port has four channels. Each CTM port can be
connected to a CTI or another CTM.
Figure 7–4 shows the structure of a CTM channel. Paths inside the CTM are purely
combinatorial.

Each CTI trigger input can be connected through a CTM to one or more trigger
Figure 7–4. CTM Channel Structure
In Out
Channel 0
outputs under control by a debugger. Figure 7–5 shows a pictorial representation of

CTI trigger connections. The red lines depict the impact one trigger input can have on
the entire system.

Figure 7–5. CTI Trigger Connections
CTI
Trigger Outputs
Trigger Inputs
Channel
Interface
Channel Channel
Interface CTM Channel 0 Interface
Channel 1
Trigger CTI Channel 2 CTI Trigger
Interface Channel 3 Interface
Channel
Interface
CTI
Trigger
Interface
Program Trace Macrocell (PTM)

The PTM performs real-time program flow instruction tracing and provides a variety
of filters and triggers that can be used to trace specific portions of code.
The HPS contains two PTMs. Each PTM is paired with a processor and CTI. Trace data
generated from the PTM can be transmitted off-chip using HPS pins, or to the FPGA
fabric, where it can be pre-processed and transmitted off-chip using high-speed FPGA
pins.
f For more information, refer to the CoreSight PTM-A9 Technical Reference Manual, which
you can download from the ARM website (infocenter.arm.com).

CoreSight Debug and Trace Programming Model
HPS Debug APB Interface

The HPS can extend the CoreSight debug control bus into the FPGA fabric. The debug
interface is an APB-compatible interface with built-in clock crossing.
f For more information, refer to the HPS Component Interfaces chapter in volume 3 of the

This section describes programming model details specific to Altera’s implementation
of the ARM CoreSight technology.
f For programming interface details of each CoreSight component, refer to the CoreSight
Components Technical Reference Manual, which you can download from the ARM
The debug components can be configured to cause triggers when certain events occur.
For example, soft logic in the FPGA fabric can signal an event which triggers an STM
message injection into the trace stream. CoreSight components are configured
through memory-mapped registers, located at offsets relative to the CoreSight
component base address. CoreSight component base addresses are accessible through
a ROM table.
ROM Table
Table 7–1 contains entries found in the ROM table portion of the DAP.
Table 7–1. DAP ROM Table

ROM Entry Offset[30:12] Description
0x0 0x00001 ETF
0x1 0x00002 CTI
0x2 0x00003 TPIU
0x3 0x00004 Trace Funnel
0x4 0x00005 STM
0x5 0x00006 ETR
0x6 0x00007 FPGA-CTI
0x7 0x00100 A9ROM
0x8 0x00080 FPGAROM
0x9 0x00000 End of ROM
A host debugger can access this table at 0x8000_0000 through the DAP. HPS masters
can access this ROM at 0xFF00_0000. Registers for a particular CoreSight component
are accessed by adding the register offset to the CoreSight component base address,
and adding that total to the base address of the ROM table.

The base address of the ROM table is different when accessed from the debugger (at
0x8000_0000) than when accessed from any HPS master (at 0xFF00_0000). For
example, the CTI output enable register, CTIOUTEN[2] at offset 0xA8, can be accessed
by the host debugger at 0x8000_20A8. To derive that value, add the host debugger
access address to the ROM table of 0x8000_0000, to the CTI component base address
of 0x0000_2000, to the CTIOUTEN[2] register offset of 0xA8.
STM Channels
The STM AXI slave is connected to the MPU, DMA, and FPGA-to-HPS bridge
masters. Each master has up to 65536 channels where each channel occupies 256 bytes
of address space, for a total of 16 MB per master. The HPS address map allocates
48 MB of consecutive address space to the STM AXI slave port, divided in three 16 MB
segments.
Table 7–2 lists the address allocation for the STM address segments.
Table 7–2. STM AXI Slave Port Address Allocation

Segment Start Address End Address
0 0xFC00_0000 0xFCFF_FFFF
1 0xFD00_0000 0xFDFF_FFFF
2 0xFE00_0000 0xFEFF_FFFF
Each of the three masters can access any one of the three address segments. Your
software design determines which master uses which segment, based on the value of
bits 24 and 25 in the write address, AWADDRS[25:24]. Software must restrict each
master to use only one of the three segments.
Table 7–3 lists the fields of the STM address.
Table 7–3. STM AXI Address Fields

AXI Signal Fields Description
These bits index the 256 bytes of the stimulus port. For more information, refer to the System
AWADDRS[7:0] Trace Macrocell, Programmers' Model Architecture Specification, which you can download from
the ARM website (infocenter.arm.com).
AWADDRS[23:8] These bits identify the 65536 stimulus ports associated with a master.
AWADDRS[25:24] These bits identify the three masters. Only 0, 1, and 2 are valid values.
AWADDRS[31:26] Always 0x3F. Bits 24 to 31 combine to access 0xFC00_0000 through 0xFEFF_FFFF.
Each STM message contains a master ID that tells the host debugger which master is
associated with the message. The STM master ID is determined by combining a
portion of the AWADDRS signal and the AWPROT protection bit. Table 7–4 shows how the
STM master ID is calculated.
Table 7–4. STM Master ID Calculation

Master ID Bits AXI Signal Bits Notes
The lowest two bits are sufficient to determine which
Master ID[5:0] AWADDRS[29:24]
master, but CoreSight uses a seven-bit master ID.
Master ID[6] AWPROT[1] 0 indicates secure; 1 indicates nonsecure.

In addition to access through STM channels, the higher-order 28 (31:4) of the 32 event
signals are attached to the FPGA through the FPGA-CTI. These event signals allow
the FPGA fabric to send additional messages using the STM.
CTI Trigger Connections to Outside the Debug System

The following CTIs in the HPS debug system connect to outside the debug system:
■ csCTI
■ FPGA-CTI
csCTI
This section lists the trigger input, output, and output acknowledge pin connections
implemented for csCTI in the debug system. The trigger input acknowledge signals
are not connected to pins.
Table 7–5 lists the trigger input pin connections implemented for csCTI.
Table 7–5. Trigger Input Signals

Number Signal Source
7 ASYNCOUT STM
6 TRIGOUTHETE STM
5 TRIGOUTSW STM
4 TRIGOUTSPTE STM
3 ACQCOMP ETR
2 FULL ETR
1 ACQCOMP ETF
0 FULL ETF
Table 7–6 lists the trigger output pin connections implemented for csCTI.
Table 7–6. Trigger Output Signals

Number Signal Destination
7 TRIGIN ETF
6 FLUSHIN ETF
5 HWEVENTS[3:2] STM
4 HWEVENTS[1:0] STM
3 TRIGIN TPIU
2 FLUSHIN TPIU
1 TRIGIN ETR
0 FLUSHIN ETR

Table 7–7 lists the trigger output acknowledge pin connections implemented for
csCTI.
Table 7–7. Trigger Output Acknowledge Signals

Number Signal Source
7 0 —
6 0 —
5 0 —
4 0 —
3 TRIGINACK TPIU
2 FLUSHINACK TPIU
1 0 —
0 0 —
FPGA-CTI
FPGA-CTI connects the debug system to the FPGA fabric. FPGA-CTI has all of its
triggers available to the FPGA fabric.
Configuring Embedded Cross-Trigger Connections

CTI interfaces are programmable through a memory-mapped register interface.
f The specific registers are described in the CoreSight Components Technical Reference
Manual, which you can download from the ARM website (infocenter.arm.com).
To access registers in any CoreSight component through the debugger, the register
offsets must be added to the CoreSight component’s base address. That combined
value must then be added to the address at which the ROM table is visible to the
debugger (0x80000000).
Each CTI has two interfaces, the trigger interface and the channel interface. The
trigger interface is the interface between the CTI and other components. It has eight
trigger signals, which are hardwired to other components. The channel interface is the
interface between a CTI and its CTM, with four bidirectional channels. The mapping
of trigger interface to channel interface (and vice versa) in a CTI is dynamically
configured. You can enable or disable each CTI trigger output and CTI trigger input
connection individually.
For example, you can configure trigger input 0 in the FPGA-CTI to route to channel 3,
and configure trigger output 3 in the FPGA-CTI and trigger output 7 in CTI-0 in the
MPU debug subsystem to route from channel 3. This configuration causes a trigger at
trigger input 0 in FPGA-CTI to propagate to trigger output 3 in the FPGA-CTI and
trigger output 7 in CTI-0. Propagation can be single-to-single, single-to-multiple,
multiple-to-single, and multiple-to-multiple.
A particular soft logic signal in the FPGA connected to a trigger input in the
FPGA-CTI can be configured to trigger a flush of trace data to the TPIU. For example,
you can configure channel 0 to trigger output 2 in csCTI. Then configure trigger input
T3 to channel 0 in FPGA-CTI. Trace data is flushed to the TPIU when a trigger is
received at trigger output 2 in csCTI.

Another soft logic signal in the FPGA connected to trigger input T2 in FPGA-CTI can
be configured to trigger an STM message. csCTI output triggers 4 and 5 are wired to
the STM CoreSight component in the HPS. For example, configure channel 1 to trigger
output 4 in csCTI. Then configure trigger input T2 to channel 1 in FPGA-CTI. Refer to
Figure 7–1.
Another soft logic signal in the FPGA fabric connected to trigger input T1 in
FPGA-CTI can be configured to trigger a breakpoint on CPU 1. Trigger output 1 in
CTI-1 is wired to the debug request (EDBGRQ) signal of CPU-1. For example,
configure channel 2 to trigger output 1 in CTI-1. Then configure trigger input T1 to
channel 2 in FPGA-CTI.
Debug Clocks
The CoreSight system uses several different clocks. Table 7–8 provides a list of these
clocks. Port Name is the name of the clock signal inputs described for individual
CoreSight debug components in the ARM documentation. Signal Name is the name of
the clock signal used with other HPS components.
Table 7–8. CoreSight Clocks

Port Name Clock Source Signal Name Description
ATCLK Clock manager dbg_at_clk Trace bus clock.
CTICLK (for Cross trigger interface clock for csCTI. It can be synchronous
Clock manager dbg_at_clk
csCTI) or asynchronous to CTMCLK.
CTICLK (for
FPGA fabric fpga_cti_clk Cross trigger interface clock for FPGA-CTI.
FPGA-CTI)
CTICLK (for CTI-0 Cross trigger interface clock for CTI-0 and CTI-1. It can be
Clock manager mpu_clk
and CTI-1) synchronous or asynchronous to CTMCLK.
CTMCLK (for Cross trigger matrix clock for csCTM. It can be synchronous
Clock manager dbg_clk
csCTM) or asynchronous to CTICLK.
Cross trigger matrix clock for CTM. It can be synchronous or
CTMCLK (for CTM) Clock manager mpu_clk
asynchronous to CTICLK.
DAPCLK Clock manager dbg_clk DAP internal clock. It must be equivalent to PCLKDBG.
PCLKDBG Clock manager dbg_clk Debug APB (DAPB) clock.
Used by the AHB-Lite master inside the DAP. It is
HCLK Clock manager dbg_clk asynchronous to DAPCLK. In the HPS, the AHB-Lite port uses
same clock as DAPCLK.
Used by the APB slave port inside the DAP. It is asynchronous
PCLKSYS Clock manager l4_mp_clk
to DAPCLK.
JTAG interface dap_tck The SWJ-DP clock driven by the external debugger through
either the JTAG interface or the FPGA fabric. It is
SWCLKTCK
FPGA fabric tpiu_traceclkin asynchronous to DAPCLK. When through the JTAG interface,
this clock is the same as TCK of the JTAG interface.
TPIU trace clock input. It is asynchronous to ATCLK. In the
TRACECLKIN Clock manager dbg_trace_clk HPS, this clock can come from the clock manager or the
FPGA fabric.
f For more information about the CoreSight port names, refer to table 6-2 in the
CoreSight Technology System Design Guide, which you can download from the ARM

Debug Resets
The CoreSight system uses several resets. Table 7–9 provides a list of these resets. Port
Name is the name of the clock signal inputs described for individual CoreSight debug
components in the ARM documentation. Signal Name is the name of the clock signal
used with other HPS components.
Table 7–9. CoreSight Resets

Port Name Clock Source Signal Name Description
ATRESETn Reset manager dbg_rst_n Trace bus reset. It resets all registers in ATCLK domain.
CTI reset signal. It resets all registers in CTICLK domain. In the
nCTIRESET Reset manager dbg_rst_n HPS, there are four instances of CTI. All four use the same
reset signal.
DAPRESETn Reset manager dbg_rst_n DAP internal reset. It is connected to PRESETDBGn.
PRESETDBGn Reset manager dbg_rst_n Debug APB reset. Resets all registers clocked by PCLKDBG.
SoC-provided reset signal that resets all of the AMBA on-chip
HRESETn Reset manager sys_dbg_rst_n interconnect. Use this signal to reset the DAP AHB-Lite master
port.
PRESETSYSn Reset manager sys_dbg_rst_n Resets system APB slave port of DAP.
nCTMRESET Reset manager dbg_rst_n CTM reset signal. It resets all signals clocked by CTMCLK.
True power on reset signal to the DAP SWJ-DP. It must only
nPOTRST Reset manager tap_cold_rst_n
reset at power-on.
Resets the DAP TAP controller inside the SWJ-DP. This signal
nTRST JTAG interface nTRST pin
is driven by the host using the JTAG connector.
Reset signal for TPIU. Resets all registers in the TRACECLKIN
TRESETn Reset manager dbg_rst_n
domain.
f For more information about the CoreSight port names, refer to table 6-3 in the
CoreSight Technology System Design Guide, which you can download from the ARM
The ETR stall enable field (etrstallen) of the ctrl register in the reset manager
controls whether the ETR is requested to stall its AXI master interface to the L3
interconnect before a warm or debug reset.
f For more information about reset handshaking, refer to the Reset Manager chapter in
The level 4 (L4) watchdog timers can be paused during debugging to prevent reset
while the processor is stopped at a breakpoint.
f For more information, refer to the Watchdog Timer chapter in volume 3 of the Cyclone V
Device Handbook.

CoreSight Debug and Trace Address Map and Register Definitions
CoreSight Debug and Trace Address Map and Register Definitions

f The address map resides in the hps.html file that accompanies this handbook volume.
The register definitions reside in separate ARM documentation. Click the link to open
the file.
To view the debug-related module descriptions and base addresses, scroll to and click
the following links:
■ stm
■ dap
■ dmanonsecure
■ dmasecure
■ mpuscu
■ mpul2
To then view the register and field descriptions, click the link in the module
description to access the appropriate ARM documentation. The register addresses are
offsets relative to the base address of each module instance.


Added functional description, programming model, and address map and register definitions
June 2012 1.1
sections.


Section V. Memory and Memory
Controllers

■ Chapter 8, SDRAM Controller Subsystem
■ Chapter 9, On-Chip Memory
■ Chapter 10, NAND Flash Controller
■ Chapter 11, SD/MMC Controller
■ Chapter 12, Quad SPI Flash Controller

V–2 Section V: Memory and Memory Controllers

8. SDRAM Controller Subsystem
November 2012
cv_54008-1.1
cv_54008-1.1
The hard processor system (HPS) SDRAM controller subsystem provides efficient
access to external SDRAM for the ARM® Cortex™-A9 microprocessor unit (MPU)
subsystem, the level 3 (L3) interconnect, and the FPGA fabric. The SDRAM controller
provides an interface between the FPGA fabric and HPS. The interface accepts
Advanced Microcontroller Bus Architecture (AMBA®) Advanced eXtensible Interface
(AXI™) and Avalon® Memory-Mapped (Avalon-MM) transactions, converts those
commands to the correct commands for the SDRAM, and manages the details of the
SDRAM access.
Features of the SDRAM Controller Subsystem

The SDRAM controller subsystem offers the following features:
■ Support for double data rate 2 (DDR2), DDR3, and low-power DDR2
(LPDDR2) SDRAM
■ User-configurable timing parameters
■ Up to 4 Gb density parts
■ Two chip selects
■ Integrated error correction code (ECC), 24- and 40-bit widths
■ User-configurable memory width of 8, 16, 16+ECC, 32, 32+ECC
■ Command reordering (look-ahead bank management)
■ Data reordering (out of order transactions)
■ User-controllable bank policy on a per port basis for either closed page or
conditional open page accesses
■ User-configurable priority support with both absolute and relative priority
scheduling
■ Flexible FPGA fabric interface configuration with up to 6 ports and data widths
up to 256 bits wide using Avalon-MM and AXI interfaces.
■ Power management supporting self refresh, partial array self-refresh (PASR),
power down, and LPDDR2 deep power down

November 2012
Subscribe
8–2 Chapter 8: SDRAM Controller Subsystem
SDRAM Controller Subsystem Block Diagram and System Integration

The SDRAM controller subsystem connects to the MPU subsystem, the main switch of
the L3 interconnect, and the FPGA fabric. The memory interface consists of the
SDRAM controller, the physical layer (PHY), control and status registers (CSRs), and
their associated interfaces.
Figure 8–1 shows a high-level block diagram of the SDRAM controller subsystem.
Figure 8–1. SDRAM Controller Subsystem High-Level Block DIagram
SDRAM Controller Subsystem
SDRAM Controller
MPU 64-Bit AXI
Subsystem
Altera
PHY
32-Bit AXI Interface HPS
L3 DDR External
Multi-Port Single-Port I/O
Interconnect PHY Memory
Front End Controller Pins
FPGA-to-HPS
SDRAM Interface
32- to 256-Bit
AXI or
FPGA Avalon-MM
Fabric
Control & Status Registers
Register Slave Interface
SDRAM Controller
The SDRAM controller provides high performance data access and run-time
programmability. The controller reorders data to reduce row conflicts and bus turn-
around time by grouping read and write transactions together, allowing for efficient
traffic patterns and reduced latency.
The SDRAM controller consists of a multiport front end (MPFE) and a single-port
controller. The MPFE provides multiple independent interfaces to the single-port
controller. The single-port controller communicates with and manages each external
memory device. For more information, refer to “Memory Controller Architecture” on
page 8–4.
DDR PHY
The DDR PHY provides a physical layer interface between the memory controller and
memory devices, which performs read and write memory operations. The DDR PHY
has dataflow components, control components, and calibration logic that handle the
calibration for the SDRAM interface timing.

Chapter 8: SDRAM Controller Subsystem 8–3
SDRAM Controller Subsystem Interfaces

The following sections describe the SDRAM controller subsystem interfaces.
MPU Subsystem Interface

The SDRAM controller is connected to the MPU subsystem with a dedicated 64-bit
AXI interface, operating on the mpu_l2_ram_clk clock domain.
L3 Interconnect Interface
The SDRAM controller is connected to the L3 interconnect with a dedicated 32-bit AXI
interface, operating on the l3_main_clk clock domain.
CSR Interface
The CSR interface is connected the level 4 (L4) bus and operates on the l4_sp_clk
clock domain. The MPU subsystem uses the CSR interface to configure the controller
and PHY, for example, setting the memory timing parameter values or placing the
memory to a low power state. The CSR interface also provides access to the status
registers in the controller and PHY.
FPGA-to-HPS SDRAM Interface

The FPGA-to-HPS SDRAM interface provides masters implemented in the FPGA
fabric access to the SDRAM controller subsystem in the HPS. The interface has three
ports types that are used to construct the following AXI or Avalon-MM interfaces:
■ Command ports—issue read and write commands, and for receive write
acknowledge responses
■ 64-bit read data ports—receive data returned from a memory read
■ 64-bit write data ports—transmit write data
The FPGA-to-HPS SDRAM interface supports six command ports, allowing up to six
Avalon-MM interfaces or three AXI interfaces. Each command port can be used to
implement either a read or write command port for AXI, or be used as part of an
Avalon-MM interface. The AXI and Avalon-MM interfaces can be configured to
support 32-, 64-, 128-, and 256-bit data.
Table 8–1 lists the FPGA-to-HPS SDRAM controller interface ports connected to the
FPGA.
Table 8–1. FPGA-to-HPS SDRAM Controller Port Types

Port Type Number
Command 6
64-bit read data 4
64-bit write data 4
The FPGA-to-HPS SDRAM controller interface can be configured with the following
characteristics:
■ Avalon-MM interfaces and AXI interfaces can be mixed and matched as required
by the fabric logic, within the bounds of the number of ports provided to the
fabric.

Memory Controller Architecture
■ Each Avalon-MM or AXI interface of the FPGA-to-HPS SDRAM interface operates

on an independent clock domain.
■ The FPGA-to-HPS SDRAM interfaces are configured during FPGA configuration.
Table 8–2 shows the number of ports needed to configure different bus protocols,
based on type and data width.
Table 8–2. FPGA-to-HPS SDRAM Port Utilization

Bus Protocol Command Read Data Write Data
32- or 64-bit AXI 2 (1) 1 1
128-bit AXI 2 (1) 2 (2) 2 (2)
256-bit AXI 2 (1) 4 (2) 4 (2)
32- or 64-bit Avalon-MM 1 1 1
128-bit Avalon-MM 1 2 2
256-bit Avalon-MM 1 4 4
32- or 64-bit Avalon-MM write-only 1 0 1
128-bit Avalon-MM write-only 1 0 2
256-bit Avalon-MM write-only 1 0 4
32- or 64-bit Avalon-MM read-only 1 1 0
128-bit Avalon-MM read-only 1 2 0
256-bit Avalon-MM read-only 1 4 0
(1) Because the AXI protocol allows simultaneous read and write commands to be issued, two SDRAM
control ports are required to form an AXI interface.
(2) Because the native size of the data ports is 64 bits, extra read and write ports are required to form an
AXI interface.

The SDRAM controller consists of an MPFE, a single-port controller, and an interface
to the CSRs.

Figure 8–2 shows a block diagram of the SDRAM controller portion of the SDRAM
controller subsystem.
Figure 8–2. SDRAM Controller Block DIagram
SDRAM Controller
Multi-Port Front End Single-Port Controller

Read Data
6
Data Reorder
FIFO Buffer
Buffers
ECC
Generation
Write Data &
Checking
6 FIFO Write Data
Data POP
FIFO Buffer
FPGA Logic
Buffers Altera
Fabric PHY
Interface
Command
6 Write
WR Acknowledge Rank Timer
Acknowledge Queues
10
Timer
Command Command
Scheduler Bank Arbiter
FIFO Generator
Buffers Pool
Control & Status Register Interface
MPFE
The MPFE is responsible for scheduling pending transactions from the configured
interfaces and sending the scheduled memory transactions to the single-port
controller. The MPFE handles all functions related to individual ports.
The MPFE consists of the following three primary sub-blocks.
Command Block
The command block accepts read and write transactions from the FPGA fabric and the
HPS. When the command FIFO buffer is full, the command block applies
backpressure by deasserting the ready signal. For each pending transaction, the
command block calculates the next SDRAM burst needed to progress on that
transaction. The command block schedules pending SDRAM burst commands based
on the user-supplied configuration, available write data, and unallocated read data
space.

Write Data Block

The write data block transmits data to the single-port controller. The write data block
maintains write data FIFO buffers and clock boundary crossing for the write data. The
write data block informs the command block of the amount of pending write data for
each transaction so that the command block can calculate eligibility for the next
SDRAM write burst.
Read Data Block

The read data block receives data from the single-port controller. Depending on the
port state, the read data block either buffers the data in its internal buffer or passes the
data straight to the clock boundary crossing FIFO buffer. The read data block reorders
out-of-order data for Avalon-MM ports.
In order to prevent the read FIFO buffer from overflowing, the read data block
informs the command block of the available buffer area so the command block can
pace read transaction dispatch.
Single-Port Controller
The single-port logic is responsible for following actions:
■ Queuing the pending SDRAM bursts
■ Choosing the most efficient burst to send next
■ Keeping the SDRAM pipeline full
■ Ensuring all SDRAM timing parameters are met
Transactions passed to the single-port logic for a single page in SDRAM are
guaranteed to be executed in order, but transactions can be reordered between pages.
Each SDRAM burst read or write is converted to the appropriate Altera PHY interface
(AFI) command to open a bank on the correct row for the transaction (if required),
execute the read or write command, and precharge the bank (if required).
The single-port logic implements command reordering (looking ahead at the
command sequence to see which banks can be put into the correct state to allow a read
or write command to be executed) and data reordering (allowing data transactions to
be dispatched even if the data transactions are executed in an order different than
they were received from the multiport logic).
Command Generator
The command generator accepts commands from the MPFE and from the internal
ECC logic, and provides those commands to the timer bank pool.
Timer Bank Pool

The timer bank pool is a parallel queue that operates with the arbiter to enable data
reordering. The timer bank pool tracks incoming requests, ensures that all timing
requirements are met, and, on receiving write-data-ready notifications from the write
data buffer, passes the requests to the arbiter.

Functional Description of the SDRAM Controller Subsystem
Arbiter
The arbiter determines the order in which requests are passed to the memory device.
When the arbiter receives a single request, that request is passed immediately. When
multiple requests are received, the arbiter uses arbitration rules to determine the order
to pass requests to the memory device.
Rank Timer
The rank timer performs the following functions:
■ Maintains rank-specific timing information
■ Ensures that only four activates occur within a specified timing window
■ Manages the read-to-write and write-to-read bus turnaround time
■ Manages the time-to-activate delay between different banks
Write Data Buffer

The write data buffer receives write data from the MPFE and passes the data to the
PHY, on approval of the write request.
ECC Block
The ECC block consists of an encoder and a decoder-corrector, which can detect and
correct single-bit errors, and detect double-bit errors. The ECC block can correct
single- bit errors and detect double-bit errors resulting from noise or other
impairments during data transmission.
AFI Interface
The AFI interface provides communication between the controller and the PHY.
CSR Interface
The CSR interface is accessible from the L4 bus. The interface allows code executing in
the HPS MPU and FPGA fabric to configure and monitor the SDRAM controller.

This section provides a functional description of the SDRAM controller subsystem.
MPFE Operational Behavior

This section describes the operational behavior of the MPFE.
Operation Ordering
Requests to the same SDRAM page arriving at a given port are executed in the order
in which they are received. Requests arriving at different ports have no guaranteed
order of service, except when a first transaction has completed before the second
arrives.

Operation ordering is defined and enforced within a port, but not between ports. All
transactions received on a single port for overlapping addresses execute in order.
Transactions received on different ports have no guaranteed order unless the second
transaction is presented after the first has completed.
Avalon-MM does not support write acknowledgement. When a port is configured to
support Avalon-MM, you should read from the location that was previously written
to ensure that the write operation has completed. When a port is configured to
support AXI, the master accessing the port can safely issue a read operation to the
same address as a write operation as soon as the write has been acknowledged. To
keep write latency low, writes are acknowledged as soon as the transaction order is
guaranteed—meaning that any operations received on any port to the same address
as the write operation are executed after the write operation.
To ensure that the overall latency of traffic is as low as possible, the single port logic
can return read data out of order to the multi-port logic which will reorder it when
transactions return out of order. A large percentage of traffic reordering will be
between ports and transactions only are ordered within a port. For traffic which is
reordered between ports but not within a port, no reordering needs to be done.
Eliminating unnecessary reordering reduces average latency.
Multiport Scheduling
Multiport scheduling is governed by two factors, the absolute priority of a request
and the weighting of a port.
The evaluation of absolute priority ensures that ports carrying higher-priority traffic
are served ahead of ports carrying lower-priority traffic. The scheduler recognizes
eight priority levels (0-7), with higher values representing higher priorities. For
example, any transaction with priority seven is scheduled before transactions of
priority six or lower.
When ports carry traffic of the same absolute priority, relative priority is determined
based on port weighting. Port weighting is a five-bit value (0-31), and is determined
by a deficit-weighted round robin (DWRR) algorithm, which corrects for past over-
servicing or under-servicing of a port. Each port has an associated weight which is
updated every cycle, with a user-configured weight added to it and the amount of
traffic served subtracted from it. The port with the highest weighting is considered the
most eligible.
To ensure that high-priority traffic is served quickly and that long and short bursts are
effectively interleaved between ports, incoming transactions longer than a single
SDRAM burst are scheduled as a series of SDRAM bursts, with each burst arbitrated
separately.
To ensure that lower priority ports do not build up large running weights while
higher priority ports monopolize bandwidth, the controller's DWRR weights are
updated only when a port matches the scheduled priority. Therefore, if three ports are
being accessed, two being priority seven and one being priority four, the weights for
both ports at priority seven are updated but the port with priority four remains
unchanged.

Multiport scheduling is performed between all of the ports connected to the FPGA
fabric and internally in the HPS to determine which transaction is serviced next.
Arbitration is performed on a SDRAM burst basis to ensure that a long transaction
does not lock other transactions or cause latency to significantly increase for
high-priority ports.
Arbitration supports both absolute and relative priority. Absolute priority is intended
for applications where one master should always get priority above or below others.
Relative priority is supported through a programmable weight field which controls
scheduling between ports at the same priority.
The scheduler is work-conserving. Write operations can only be scheduled when
enough data for the SDRAM burst has been received. Read operations can only be
scheduled when sufficient internal memory is free and the port is not occupying too
much of the read buffer.
The multiport scheduling configuration can be updated while traffic is flowing. Both
priority and weight for a port can be updated without interrupting traffic on a port.
Updates are used in scheduling decisions within 10 memory clock cycles of being
updated, so priority can be updated frequently if needed.
Read Data Handling

The MPFE contains a read buffer shared by all ports. If a port is capable of receiving
returned data then the read buffer is bypassed. If the size of a read transaction is
smaller than twice the memory interface width, the buffer RAM cannot be bypassed.
SDRAM Burst Scheduling

SDRAM burst scheduling recognizes addresses that access the same row/bank
combination, known as open page accesses. Operations to a page are served in the
order in which they are received by the single-port controller.
Selection of SDRAM operations is a two-stage process. First, each pending transaction
must wait for its timers to be eligible for execution. Next, the transaction arbitrates
against other transactions that are also eligible for execution.
The following rules govern transaction arbitration:
■ High-priority operations take precedence over lower-priority operations
■ If multiple operations are in arbitration, read operations have precedence over
write operations
■ If multiple operations still exist, the oldest is served first
A high-priority transaction in the SDRAM burst scheduler wins arbitration for that
bank immediately if the bank is idle and the high-priority transaction's chip select,
row, or column fields of the address do not match an address already in the single-
port controller. If the bank is not idle, other operations to that bank yield until the
high-priority operation is finished. If the chip select, row, and column fields match an
earlier transaction, the high-priority transaction yields until the earlier transaction is
completed.

Clocking
The FPGA fabric ports of the MPFE can be clocked at different frequencies.
Synchronization is maintained by clock-domain crossing logic in the MPFE.
Command ports can operate on different clock domains, but the data ports associated
with a given command port must be attached to the same clock as that command port.
For example, a command port paired with a read and write port to form an
Avalon-MM interface must operate at the same clock frequency as the data ports
associated with it.
Single-Port Controller Operational Behavior

This section describes the operational behavior of the single-port controller.
SDRAM Interface
The SDRAM interface is up to 40 bits wide and can accommodate 8-bit, 16-bit, 16-bit
plus ECC, 32-bit, or 32-bit plus ECC configurations. The SDRAM interface supports
LPDDR2, DDR2, and DDR3 memory protocols.
Command and Data Reordering

The heart of the SDRAM controller is a command and data reordering engine.
Command reordering allows banks for future transactions to be opened before the
current transaction finishes. Data reordering allows transactions to be serviced in a
different order than they were received when that new order allows for improved
utilization of the SDRAM bandwidth. Operations to the same bank and row are
performed in order to ensure that operations which impact the same address preserve
the data integrity.
Figure 8–3 shows the relative timing for a write/read/write/read command sequence
performed in order and then the same command sequence performed with data
reordering. Data reordering allows the write and read operations to occur in bursts,
without bus turnaround timing delay or bank reassignment.
Figure 8–3. Data Reordering Effect

Data Reordering Off
Command WR RD WR RD
Address B0R0 B1R0 B0R0 B1R0
Data Reordering On
Command WR WR RD RD
Address B0R0 B0R0 B1R0 B1R0
The SDRAM controller schedules among all pending row and column commands
every clock cycle.
Bank Policy
The bank policy of the SDRAM controller allows users to request that a transaction's
bank remain open after an operation has finished so that future accesses do not delay
in activating the same bank and row combination. The controller supports only eight
simultaneously-opened banks, so an open bank might get closed if the bank resource
is needed for other operations.

Open bank resources are allocated dynamically as SDRAM burst transactions are
scheduled. Bank allocation is requested automatically by the controller when an
incoming transaction spans multiple SDRAM bursts or by the extended command
interface. When a bank must be reallocated, the least-recently-used open bank is used
as the replacement.
If the controller determines that the next pending command will cause the bank
request to not be honored, the bank might be held open or closed depending on the
pending operation. A request to close a bank with a pending operation in the timer
bank pool to the same row address causes the bank to remain open. A request to leave
a bank open with a pending command to the same bank but a different row address
causes a precharge operation to occur.
Write Combining
The SDRAM controller combines write operations from successive bursts on a port
where the starting address of the second burst is one greater than the ending address
of the first burst and the resulting burst length does not overflow the 11-bit
burst-length counters. Write combining does not occur if the previous bus command
has finished execution before the new command has been received.
Burst Length Support

The controller supports burst lengths of 2, 4, 8, and 16, and data widths of 8, 16, and 32
bits for non-ECC operation, and widths of 24 and 40 operations with ECC enabled.
Table 8–3 shows the type of SDRAM for each burst length.
Table 8–3. SDRAM Burst Lengths

Burst Length SDRAM
4 LPDDR2, DDR2
8 DDR2, DDR3, LPDDR2
16 LPDDR2
Width Matching
The SDRAM controller automatically performs data width conversion.
ECC
The single-port controller supports memory ECC calculated by the controller. The
controller ECC employs standard Hamming logic to detect and correct single-bit
errors and detect double-bit errors. The controller ECC is available for 16-bit and 32-
bit widths, each requiring an additional 8 bits of memory, resulting in an actual
memory width of 24-bits and 40-bits, respectively.
Controller ECC provides the following features:
■ Byte writes—The memory controller performs a read-modify-write operation to
ensure that the ECC data remains valid when a subset of the bits of a word is being
written. If an entire word is being written (but less than a full burst) and the DM
pins are connected, no read is necessary and only that word is updated. If
controller ECC is disabled, byte-writes have no performance impact.

■ ECC write backs—When a read operation detects a correctable error, the memory
location is scheduled for a read-modify-write operation to correct the single-bit
error. ECC write backs are enabled and disabled through the
cfg_enable_ecc_code_overwrites field in the ctrlcfg register.
■ Notification of ECC errors—The memory controller provides interrupts for single-
bit and double-bit errors. The status of interrupts and errors are recorded in status
registers, as follows:
■ The dramsts register records interrupt status.
■ The dramintr register records interrupt masks.
■ The sbecount register records the single-bit error count.
■ The dbecount register records the double-bit error count.
■ The erraddr register records the address of the most recent error.
Byte Writes
Byte writes with ECC enabled are executed as a read-modify-write. Typical operations
only use a single entry in the timer bank pool. Controller ECC enabled sub-word
writes use two entries. The first operation is a read and the second operation is a
write. These two operations are transferred to the timer bank pool with an address
dependency so that the write cannot be performed until the read data has returned.
This approach ensures that any subsequent operations to the same address (from the
same port) are executed after the write operation, because they are ordered on the row
list after the write operation.
If an entire word is being written (but less than a full burst), then no read is necessary
and only that word is updated.
ECC Write Backs

If the controller ECC is enabled and a read operation results in a correctable ECC
error, the controller corrects the location in memory, if write backs are enabled. The
correction results in scheduling a new read-modify-write. A new read is performed at
the location to ensure that a write operation modifying the location is not overwritten.
The actual ECC correction operation is performed as a read-modify-write operation.
User Notification of ECC Errors

The following methods notify you of an ECC error:
For the MPU subsystem, an interrupt signal provides notification and the ECC error
information is stored in the status registers.
f For more information, refer to the Cortex-A9 Microprocessor Unit SubSystem chapter in
Interleaving Options
The controller supports the following address-interleaving options:
■ Noninterleaved
■ Bank interleave without chip select interleave
■ Bank interleave with chip select interleave

All of the interleaving examples use 512 megabits (Mb) x 16 DDR3 chips and are
documented as byte addresses. For RAMs with smaller address fields, the order of the
fields stays the same but the widths may change.
Noninterleaved
RAM mapping is noninterleaved.
Figure 8–4 shows noninterleaved address decoding.
Figure 8–4. Noninterleaved Address Decoding
Address Decoding
(512 Mb x 16 DDR3 DRAM)
DDR 3 DDR 3
512 x 16 512 x16
DDR 3 DDR 3
512 x 16 512 x16
Controller
28 24 20 16 12 8 4 0
S B (2 :0 ) R( 15 :0 ) C( 9 :0 )
Address Nomenclature
C =Column R= Row B =Bank S = Chip Select
Bank Interleave Without Chip Select Interleave

Bank interleave without chip select interleave swaps row and bank from the
noninterleaved address mapping. This interleaving allows smaller data structures to
spread across all banks in a chip.
Figure 8–5 shows bank interleave without chip select interleave address decoding.
Figure 8–5. Bank Interleave Without Chip Select Interleave Address Decoding
28 24 20 16 12 8 4 0
S R(15:0) B(2:0) C(9:0)
Bank Interleave with Chip Select Interleave

Bank interleave with chip select interleave moves the row address to the top, followed
by chip select, then bank, and finally column address. This interleaving allows smaller
data structures to spread across multiple banks and chips (giving access to 16 total
banks for multithreaded access to blocks of memory). Memory timing is degraded
when switching between chips.

Figure 8–6 shows bank interleave with chip select interleave address decoding.
Figure 8–6. Bank Interleave With Chip Select Interleave Address Decoding
28 24 20 16 12 8 4 0
R(15:0) S B(2:0) C(9:0)
AXI-Exclusive Support
The single-port controller supports AXI-exclusive operations. The controller
implements a table shared across all masters, which can store up to 16 pending writes.
Table entries are allocated on an exclusive read and table entries are deallocated on a
successful write to the same address by any master.
Any exclusive write operation that is not present in the table returns an exclusive fail
as acknowledgement to the operation. If the table is full when the exclusive read is
performed, the table replaces a random entry.
1 When using AXI-exclusive operations, accessing the same location from Avalon-MM
interfaces can result in unpredictable results.
Memory Protection
The single-port controller has address protection to allow the software to configure
basic protection of memory from all masters in the system. If the system has been
designed exclusively with AXI masters, TrustZone® is supported. Ports that use
Avalon-MM can be configured for port level protection.
f For information about TrustZone®, refer to the ARM website (www.arm.com).
Memory protection is based on physical addresses in memory. You can set rules to
allow or disallow accesses to a range of memory, or to enable only secure accesses to a
range of memory (or a combination of the two).
Secure and non-secure regions are specified by rules containing a starting address and
ending address with 1 MB boundaries for both addresses. You can override the port
defaults and allow or disallow all transactions.
The memory protection table, which is an internal table addressed through the CSR
interface, contains rules to permit or deny memory access. You can configure up to a
maximum of twenty rules to control memory access. Table 8–4 lists the fields that you
can specify for each rule.
Table 8–4. Fields for Rules in Memory Protection Table (Part 1 of 2)

Field Width Description
Valid 1 Set to 1 to activate the rule. Set to 0 to deactivate the rule.
Specifies the set of ports to which the rule applies, with one bit
representing each port, as follows: bits 0 to 5 correspond to FPGA
Port Mask (1) 10
fabric ports 0 to 5, bit 6 corresponds to AXI L3 switch read, bit 7 is
the CPU read, bit 8 is L3 switch write, and bit 9 is the CPU write.

Table 8–4. Fields for Rules in Memory Protection Table (Part 2 of 2)

Field Width Description
Low transfer ID of the rules to which this rule applies. Incoming
transactions match if they are greater than or equal to this value.
TID_low (1) 12
Ports with smaller TIDs have the TID shifted to the lower bits and
zero padded at the top.
High transfer ID of the rules to which this rule applies. Incoming
TID_high (1) 12
transactions match if they are less than or equal to this value.
Points to a 1MB block and is the lower address. Incoming addresses
Address_low 12
match if they are greater than or equal to this value.
Upper limit of address. Incoming addresses match if they are less
Address_high 12
than or equal to this value.
A value of 00 indicates that the protection bit is not set; a value of 01
Protection 2 sets the protection bit. Systems that do not set AXI protection to a
known value should program this for either protection value.
Fail/allow 1 Set this value to 1 to force the operation to fail or succeed.
(1) Although TID and Port Mask could be redundant, including both in the table allows possible compression of rules.
If masters connected to a port do not have contiguous TIDs, a port-based rule might be more efficient than a TID-
based rule, in terms of the number of rules needed.
A port has a default access status of either allow or fail, and rules with the opposite
allow/fail value can override the default. The system evaluates each transaction
against every rule in the memory protection table. A transaction received on a port
which by default allows access, would fail only if a rule with the fail bit matches the
transaction. Conversely, a port which by default prevents access, would allow access
only if a rule allows that transaction to pass.
Exclusive transactions are security checked on the read operation only. A write
operation can occur only if a valid read is marked in the internal exclusive table.
Consequently, a master performing an exclusive read followed by a write, can write to
memory only if the exclusive read was successful.
Example of Configuration for TrustZone

For a TrustZone configuration, memory is divided into a range of memory accessible
by secure masters and a range of memory accessible by nonsecure masters. The two
memory address ranges may have a range of memory that overlaps.
This example implements the following memory configuration:
■ 2 GB total RAM size
■ 0—512 MB dedicated secure area
■ 513—576 MB shared area
■ 577—2048 MB dedicated nonsecure area

SDRAM Power Management
In this example, each port is configured by default to disallow all accesses. Table 8–5
shows the two rules programmed into the memory protection table.
Table 8–5. Rules in Memory Protection Table for Example Configuration

Rule # Port Mask TID Low TID High Address Low Address High Prot Fail/Allow
1 0’b1111111111 0 4095 0 576 b01 allow
2 0’b1111111111 0 4095 512 2047 b00 allow
The port mask value, TID Low, and TID High, apply to all ports and all transfers
within those ports. Each access request is evaluated against the memory protection
table, and will fail unless a rule matches allowing a transaction to complete
successfully.
Table 8–6 shows the result for a sample set of transactions.
Table 8–6. Result for a Sample Set of Transactions

Operation Source Address Prot Result Comments
Read CPU 4096 1 Allow Matches rule 1.
536, 870, 912
Write CPU 1 Allow Matches rule 1.
(512 MB)
Does not match rule 1 (out of range of the
L3 attached 605, 028, 350
Write 1 Fail address field), does not match rule 2
masters (577 MB)
(protection bit incorrect).
L3 attached Does not match rule 1 (prot value wrong),
Read 4096 0 Fail
masters does not match rule 2 (not in address range).
536, 870, 912
Write CPU 0 Allow Matches rule 2.
(512 MB)
L3 attached 605, 028, 350
Write 0 Allow Matches rule 2.
masters (577 MB)
If you did not want any overlap between the memory blocks, you could specify the
address ranges in the two rules of Table 8–5 to be mutually exclusive. Depending on
your desired TrustZone configuration, you can add rules to the memory protection
table to create multiple blocks of protected or unprotected space.
SDRAM Power Management

The SDRAM controller subsystem supports the following power saving features in
the SDRAM:
■ Partial array self-refresh (PASR)
■ Power down
■ Deep power down for LPDDR2
Power-saving mode initiates either due to a user command or from inactivity.

DDR PHY
Power-down mode is initiated by writing to the appropriate control register. It forces

the SDRAM burst-scheduling bank-management logic to close all banks and issue the
power down command. You can program the controller to enable power-down when
the SDRAM burst-scheduling queue is empty for a specified number of clock cycles.
The SDRAM automatically reactivates when an active SDRAM command is received.
Other power-down modes are performed only under user control.
DDR PHY
The DDR PHY connects the memory controller and external memory devices in the
speed critical command path.
The DDR PHY implements the following functions:
■ Calibration—the DDR PHY supports the JEDEC-specified steps to synchronize the
memory timing between the controller and the SDRAM chips. The calibration
algorithm is implemented in software.
■ Memory device initialization—the DDR PHY performs the mode register write
operations to initialize the devices. The DDR PHY handles re-initialization after a
deep power down.
■ Single-data-rate to double-data-rate conversion.
Clocks
All clocks are assumed to be asynchronous with respect to the ddr_dqs_clk memory
clock. All transactions are synchronized to memory clock domain.
Table 8–7 shows the SDRAM controller subsystem clock domains.
Table 8–7. SDRAM Controller Subsystem Clock Domains

Clock Name Description
ddr_dq_clk Clock for PHY
ddr_dqs_clk Clock for MPFE, single-port controller, CSR access, and PHY
ddr_2x_dqs_clk Clock for PHY
l4_sp_clk Clock for CSR interface
mpu_l2_ram_clk Clock for MPU interface
l3_main_clk Clock for L3 interface
Six separate clocks used for the FPGA-to-HPS SDRAM ports to the
f2h_sdram_clk[5:0]
FPGA fabric
In terms of clock relationships, the FPGA fabric connects the appropriate clocks to
write data, read data, and command ports for the constructed ports.
Device Handbook.

Resets
Resets
The SDRAM controller subsystem supports a full reset (cold reset) and a warm reset,
which may or may not preserve the contents of memory. In order to preserve the
memory contents, the reset manager can request that the single-port controller place
the SDRAM in self-refresh mode prior to issuing the warm reset. If memory contents
are preserved, the PHY and the memory timing logic is not reset, but the rest of the
controller is reset.
Device Handbook.
Initialization
The SDRAM controller subsystem has CSRs which control the operation of the
controller including DRAM type, DRAM timing parameters and relative port
priorities. It also has a small set of bits which depend on the FPGA fabric to configure
ports between the memory controller and the FPGA fabric; these bits are set for you
when you configure your implementation using the HPS GUI in Qsys.
The CSRs are configured using a dedicated slave interface, which provides accesses to
registers. This region controls all SDRAM operation, MPFE scheduler configuration,
and PHY calibration.
The FPGA fabric interface configuration is programmed into the FPGA fabric and the
values of these register bits can be read by software. The ports can be configured
without software developers needing to know how the FPGA-to-HPS SDRAM
interface has been configured.
Protocol Details
Avalon-MM Bidirectional Port
The Avalon-MM bidirectional ports are standard Avalon-MM ports used to dispatch
read and write operations. Each configured Avalon-MM bidirectional port consists of
the signals listed in Table 8–8.
Table 8–8. Avalon-MM Bidirectional Port Signals (Part 1 of 2)

Name Bits Direction Function
clk 1 In Clock for the Avalon-MM interface
read 1 In Indicates read transaction
write 1 In Indicates write transaction
address 32 In Address of the transaction
readdata 32, 64, 128, or 256 Out Read data return
readdatavalid 1 Out Valid cycle flag for read data return
writedata 32, 64, 128, or 256 In Write data for a transaction
4 (32-bit data),
8(64-bit data),
byteenable In Byte enables for each write byte
16(128-bit data),
32(256-bit data)

Initialization
Table 8–8. Avalon-MM Bidirectional Port Signals (Part 2 of 2)

Indicates need for additional cycles to
waitrequest 1 Out
complete a transaction
burstcount 11 In Transaction burst length
The read and write interfaces are configured to the same size. The byte-enable size
scales with the data bus size.
f For information about the Avalon-MM protocol, refer to the Avalon Interface
Specifications.
Avalon-MM Write Port

The Avalon-MM write ports are standard Avalon-MM ports used only to dispatch
write operations. Each configured Avalon-MM write port consists of the signals listed
in Table 8–9.
Table 8–9. Avalon-MM Write Port Signals

reset 1 In Reset
clk 1 In Clock
write 1 In Indicates write transaction
writedata 32, 64, 128, or 256 In Write data for a transaction
4 (32-bit data),
8(64-bit data),
byteenable In Byte enables for each write byte
16(128-bit data),
32(256-bit data)
Indicates need for additional cycles to
waitrequest 1 Out
complete a transaction
Specifications.
Avalon-MM Read Port

The Avalon-MM read ports are standard Avalon-MM ports used only to dispatch read
operations. Each configured Avalon-MM read port consists of the signals listed in
Table 8–10.
Table 8–10. Avalon-MM Read Port Signals (Part 1 of 2)

reset 1 In Reset
clk 1 In Clock
read 1 In Indicates read transaction

Initialization
Table 8–10. Avalon-MM Read Port Signals (Part 2 of 2)

readdata 32, 64, 128, or 256 Out Read data return
readdatavalid 1 Out Flags valid cycles for read data return
Indicates the need for additional cycles to
complete a transaction. Needed for read
waitrequest 1 Out
operations when delay is needed to accept
the read command.
Specifications.
AXI Port
The AXI port uses an AXI-3 interface.
f For information about the AXI-3 interface, refer to the AMBA Open Specifications on
the ARM website (www.arm.com).
f For information about the AXI interface ports in the high-performance II controller
(HPC II), refer to the Functional Description—HPC II Controller chapter, in the External
Memory Interface Handbook.
Each configured AXI port consists of the signals listed in Table 8–11. Each AXI
interface signal is independent of the other interfaces for all signals, including clock
and reset.
Table 8–11. AXI Port Signals (Part 1 of 2)

ARESETn 1 In Reset
ACLK 1 In Clock
Write Address Channel Signals
AWID 4 In Write identification tag
AWADDR 32 In Write address
AWLEN 4 In Write burst length
AWSIZE 3 In Width of the transfer size
AWBURST 2 In Burst type
AWREADY 1 Out Indicates ready for a write command
AWVALID 1 In Indicates valid write command.
Write Data Channel Signals
WID 4 In Write data transfer ID
WDATA 32, 64, 128 or 256 In Write data
Byte-based write data strobe. Each bit
WSTRB 4, 8, 16, 32 In width corresponds to 8 bit wide transfer for
32-bit wide to 256-bit wide transfer.

SDRAM Controller Subsystem Programming Model
Table 8–11. AXI Port Signals (Part 2 of 2)

WLAST 1 In Last transfer in a burst
WVALID 1 In Indicates write data+strobes are valid
WREADY 1 Out Indicates ready for write data and strobes
Write Response Channel Signals
BID 4 Out Write response transfer ID
BRESP 2 Out Write response status
BVALID 1 Out Write response valid signal
BREADY 1 In Write response ready signal
Read Address Channel Signals
ARID 4 In Read identification tag
ARADDR 32 In Read address
ARLEN 4 In Read burst length
ARSIZE 3 In Width of the transfer size
ARBURST 2 In Burst type
ARREADY 1 Out Indicates ready for a read command
ARVALID 1 In Indicates valid read command
Read Data Channel Signals
RID 4 Out Read data transfer ID
RDATA 32, 64, 128 or 256 Out Read data
RRESP 2 Out Read response status
RLAST 1 Out Last transfer in a burs
RVALID 1 Out Indicates read data is valid
RREADY 1 In Read data channel ready signal
SDRAM Controller Subsystem Programming Model
Initialization
SDRAM controller configuration occurs through software programming of the
configuration registers using the CSR interface. Initialization of the SDRAM controller
has two separate regions with different controls.
Timing Parameters
The SDRAM controller supports a complete set of timing parameters, configurable at
run time.
SDRAM Controller Address Map and Register Definitions

Click the link to open the file.

■ sdr


November 2012 1.1 Added address map and register definitions section.

9. On-Chip Memory
November 2012
cv_54009-1.1
cv_54009-1.1
The hard processor system (HPS) contains the following on-chip memory:
■ On-Chip RAM
■ Boot ROM
The on-chip RAM provides 64 KB of general-purpose RAM. The boot ROM contains
the code required to boot the HPS from cold or warm reset. Both memories connect to
the level 3 (L3) interconnect.
On-Chip RAM
This section describes the HPS on-chip RAM.
Features of the On-Chip RAM

The on-chip RAM offers the following features:
■ 64-bit interface
■ 64 KB size
■ Single-ported RAM
■ Read acceptance of two, write acceptance of two, and a total acceptance of two.
■ Error correction code (ECC) support
■ Sustained ideal throughput (operating frequency times the data width) during
read after read, write after write, write after read, and read after write.
f For information about the operating frequency, refer to the Clock Manager

November 2012
Subscribe
9–2 Chapter 9: On-Chip Memory
On-Chip RAM
On-Chip RAM Block Diagram and System Integration

Figure 9–1 shows a block diagram of the on-chip RAM.
Figure 9–1. On-Chip RAM Block Diagram
NIC-301 On-Chip RAM

L3 Interconnect ECC from System Manager
64-Bit
(l3_main_clk)
M S
ECC Interrupts
to Interrupt Controller
Transfers between memory and the NIC-301 L3 interconnect happen through a 64-bit
interface, gated by the l3_main_clk interconnect clock. ECC logic detects single-bit,
corrected and double-bit, uncorrected errors. The memory has a read acceptance of
two, a write acceptance of two, and a total acceptance of two with round-robin
arbitration.
The entire RAM is either secure or nonsecure. Security is enforced by the NIC-301 L3
interconnect.
f For more information about security, refer to the Interconnect chapter in volume 3 of
Functional Description of the On-Chip RAM

The on-chip RAM serves as a general-purpose memory accessible from the FPGA.
The on-chip RAM uses an 64-bit slave interface. The slave interface supports transfers
between memory and the NIC-301 L3 interconnect. All reads and writes are serviced
in order.
Clocks
The on-chip RAM is driven by the l3_main_clk interconnect clock.
f For information about the operating frequency and maximum throughput, refer to the
Clock Manager chapter in volume 3 of the Cyclone V Device Handbook.
Resets
The contents of the RAM remain unchanged on a cold or warm reset. Reset only clears
the state associated with the slave interface.
The on-chip RAM reset is driven by the onchip_ram_rst_n interconnect reset signal.
Device Handbook.

Chapter 9: On-Chip Memory 9–3
Boot ROM
Boot ROM
This section describes hardware aspects of the HPS boot ROM.
f For information about the boot ROM software, refer to the Booting and Configuration
appendix in volume 3 of the Cyclone V Device Handbook.
Features of the Boot ROM

The boot ROM offers the following features:
■ 64 KB size
■ Single-ported ROM
■ Read acceptance of two
■ Sustained ideal throughput (operating frequency times the data width) during
read after read.
f For information about the operating frequency, refer to the Clock Manager
Boot ROM Block Diagram and System Integration

Figure 9–2 shows a block diagram of the boot ROM.
Figure 9–2. Boot ROM Block Diagram
NIC-301 Boot ROM

L3 Interconnect
32-Bit
(l3_main_clk)
M S
Transfers between memory and the NIC-301 L3 interconnect happen through a 32-bit
interface, gated by the l3_main_clk interconnect clock.
The entire RAM is either secure or nonsecure. Security is enforced by the NIC-301 L3
interconnect.
f For more information about security, refer to the Interconnect chapter in volume 3 of
Functional Description of the Boot ROM

The boot ROM is used only for booting the system. On a cold or warm reset of the
microprocessor unit (MPU) subsystem, MPU0 executes the pre-bootloader code
stored in the boot ROM.

9–4 Chapter 9: On-Chip Memory
On-Chip Memory Address Map and Register Definitions
f For information about the boot ROM software, refer to the Booting and Configuration
appendix in volume 3 of the Cyclone V Device Handbook.
The boot ROM uses an 32-bit slave interface. The slave interface supports transfers
between memory and the NIC-301 L3 interconnect. All writes return an error
response.
Clocks
The boot ROM is driven by the l3_main_clk interconnect clock.
f For information about the operating frequency and maximum throughput, refer to the
Clock Manager chapter in volume 3 of the Cyclone V Device Handbook.
Resets
The contents of the ROM remain unchanged on a cold or warm reset. Reset only clears
the state associated with the slave interface.
The boot ROM reset is driven by the boot_rom_rst_n interconnect clock.
Device Handbook.
On-Chip Memory Address Map and Register Definitions

There are no registers for on-chip memory.
Click the link to open the file.
To view the module descriptions and base addresses, scroll to and click the links for
■ rom
■ ocram


November 2012 1.1 Added address map section.

10. NAND Flash Controller
November 2012
cv_54010-1.2
cv_54010-1.2
The hard processor system (HPS) provides a NAND flash controller to interface with
external NAND flash memory in Altera® system-on-a-chip (SoC) FPGA systems. You
can use external flash memory to store a processor boot image, software, or as extra
storage capacity for large applications or user data. The HPS NAND flash controller is
based on the Cadence® Design IP® NAND Flash Memory Controller.
NAND Flash Controller Features

The NAND flash controller provides the following functionality and features:
■ Supports one x8 NAND flash device
■ Supports Open NAND Flash Interface (ONFI) 1.0
■ Supports NAND flash memories from Hynix, Samsung, Toshiba, Micron, and ST
Micro
■ Supports programmable 512 byte (4-, 8-, or 16-bit correction) or 1024 byte (24-bit
correction) error correction code (ECC) sector size
■ Supports pipeline read-ahead and write commands for enhanced read/write
throughput
■ Supports devices with 32, 64, 128, 256, 384, or 512 pages per block
■ Supports multiplane devices
■ Supports page sizes of 512 bytes, 2 kilobytes (KB), 4 KB, or 8 KB
■ Supports single-level cell (SLC) and multi-level cell (MLC) devices with
programmable correction capabilities
■ Provides internal direct memory access (DMA)
■ Provides programmable access timing
semiconductor products to current specifications in accordance with Altera's standard warranty, but reserves the right to make changes to any products and ISO
services at any time without notice. Altera assumes no responsibility or liability arising out of the application or use of any information, product, or service 9001:2008
described herein except as expressly agreed to in writing by Altera. Altera customers are advised to obtain the latest version of device specifications before relying Registered

November 2012
Subscribe
10–2 Chapter 10: NAND Flash Controller
NAND Flash Controller Block Diagram and System Integration
NAND Flash Controller Block Diagram and System Integration

Figure 10–1 shows integration of the NAND flash controller in the HPS. The flash
controller receives commands and data from the host through the command and data
slave interface. The host accesses the flash controller’s control and status registers
(CSRs) through the register slave interface. The flash controller handles all command
sequencing and flash device interactions. The bootstrap interface supports
configuration of the NAND flash controller when booting the HPS from NAND flash
memory. The flash controller generates interrupts to the HPS Cortex™-A9 MPCore™
processor generic interrupt controller. The DMA master interface provides accesses to
and from the flash controller through the controller's built-in DMA.
Figure 10–1. NAND Flash Controller Block Diagram

Controller Core
Map 00
DMA Master
Command Request
Interface
DMA Data
Map 01
Sequencer/Status NAND Device
Map 10 Module
Asynchronous FIFO Buffers
Interface to
Pin Multiplexer
Command & Data
Slave Interface
Map 11
L3
Interconnect Map Block
ECC Logic
and Buffer MPU
Interrupt Generic
Register Slave
Registers Module Interrupt

Interface
Controller
Bootstrap Write Read

Interface FIFO FIFO
ECC Status and

Control Signals
System
Manager
Functional Description of the NAND Flash Controller

This section describes the functionality of the NAND flash controller.
Discovery and Initialization

The NAND flash controller requires a specific initialization sequence after the HPS
receives power and the flash device is stable. During initialization, the flash controller
queries the flash device and configures itself according to one of the following flash
device types:
■ ONFI 1.0 compliant devices
■ Legacy (non-ONFI) NAND devices

Chapter 10: NAND Flash Controller 10–3
The NAND flash controller identifies ONFI-compliant connected devices using ONFI
discovery protocol, by sending the Read Electronic Signature command. For
devices that do not recognize this command (especially for 512-byte page size
devices), software must write to the system manager to assert the
bootstrap_512B_device signal to identify the device type before reset is de-asserted.
To support booting and initialization, the rdy_busy_in pin must be connected. The
NAND flash controller sends the reset command to the connected device.
The NAND flash controller performs the following initialization steps:
1. If the system manager is asserting bootstrap_inhibit_init, the flash controller
goes directly to step 7.
2. When the device is ready, the flash controller sends the ONFI Read ID command to
read the ONFI signature from the device, to determine whether an ONFI or a
legacy device is connected.
3. If the data returned by the device has an ONFI signature, the flash controller then
reads the device parameter page. The flash controller stores the relevant device
feature information in internal memory control registers, enabling it to correctly
program other registers in the flash device, and goes to step 5.
4. If the data does not have a valid ONFI signature, the flash controller assumes that
it is a legacy (non-ONFI) device. The flash controller then performs the following
steps:
a. Sends the reset command to the device
b. Reads the device signature information
c. Stores the relevant values into internal memory controller registers
5. The flash controller resets the device. At the same time, it verifies the width of the
memory interface. The HPS supports one 8-bit NAND flash device. As a result, the
flash controller always detects an 8-bit memory interface.
6. The flash controller sends the Page Load command to block 0, page 0 of the device,
configuring direct read access, so the processor can boot from that page. The
processor can start reading from the first page of the device, which is the expected
location of the preloader software.
1 The system manager can bypass this step by asserting

bootstrap_inhibit_b0p0_load before reset is deasserted.
7. The flash controller sends the reset command to the device.

8. The flash controller sets the value of the rst_comp bit in the intr_status0 register
in the status group.
Bootstrap Interface
The NAND flash controller provides a bootstrap interface that allows software to
override the default behavior of the flash controller. The bootstrap interface contains
four bits, which when set appropriately allow the flash controller to skip the
initialization phase and begin loading from flash memory immediately after reset.
These bits are driven by software through the system manager. They are sampled by
the NAND flash controller when the controller is released from reset.

f For more information about the bootstrap interface control bits, refer to the System
Manager chapter in volume 3 of the Cyclone V Device Handbook.
Table 10–1 lists the relevant bootstrap setting bits, found in the system manager’s
bootstrap register, in the nandgrp group. This table also lists recommended bootstrap
settings for a 512 byte page device.
Table 10–1. Recommended Bootstrap Settings for 512 Byte Page Device
Register Value
noinit 1 (1)
page512 1
noloadb0p0 1
■ 1—flash device supports two-cycle addressing
tworowaddr
■ 0—flash device support three-cycle addressing
(1) When this register is set, the NAND flash controller expects the host to program the related device parameter
registers. For more information, refer to “Configuration by Host”.
Configuration by Host
If the system manager sets bootstrap_inhibit_init to 1, the NAND flash controller
does not perform the discovery and initialization process. In this case, the host
processor must configure the flash controller.
When performance is not a concern in the design, the timing registers can be left
unprogrammed.
Table 10–2 shows the recommended configuration by host settings to enable the basic
read, write, and erase operations for a single-plane, 512 bytes/page device.
Table 10–2. Recommended Bootstrap Settings for 512 Byte Page Device
Register (1) Value
devices_connected 1
device_width 0 indicating an 8-bit NAND flash device
number_of_planes 1 indicating a single-plane device
The value of this register must reflect the flash device’s page
device_main_area_size
main area size.
The value of this register must reflect the flash device’s page
device_spare_area_size
spare area size.
The value of this register must reflect number of pages per
pages_per_block
block in the flash device.
(1) All registers are in the config group.
Each NAND page has a main area and a spare area. The main area is intended for data
storage. The spare area is intended for ECC and maintenance data, such as wear
leveling information. Each block consists of a group of pages.

The sizes of the main and spare areas, and the number of blocks in a page, depend on
the specific NAND device connected to the NAND flash controller. Therefore, the
device-dependent registers, device_main_area_size, device_spare_area_size, and
pages_per_block, must be programmed to match the characteristics of the device.
If your software does not perform the discovery and initialization sequence, the
software must include an alternative method to determine the correct value of the
device-dependent registers. The HPS boot ROM code enables discovery and
initialization by default (that is, bootstrap_inhibit_init = 0).
Clocks
Table 10–3 lists the NAND flash controller clock inputs.
Table 10–3. Clock Inputs to NAND Flash Controller

Clock Signal Description
nand_x_clk Clock for master and slave interfaces and the ECC sector buffer
nand_clk Clock for the NAND flash controller
The frequency of nand_x_clk is four times the frequency of nand_clk.
f For more information about the clock inputs, refer to the Clock Manager chapter in
Resets
The NAND flash controller has one reset signal. The reset manager drives this signal
to the NAND flash controller on a cold or warm reset.
Before the NAND flash controller comes out of the reset state, the pin multiplexers for
the flash external interface must be configured.

Indexed Addressing
The NAND flash controller uses indexed addressing to reduce the address span
consumed by the flash controller. Indirect addressing is controlled by two registers,
accessed through the command and data slave interface in the nanddata map, as
described in Table 10–4.
Table 10–4. Register Map for Indexed Addressing

Register Name Offset Address Usage
Software writes the 32-bit control information consisting
of MAP command type, block, and page address. The
Control 0x0
upper four bits must be set to 0. For specific usage of the
Control register, refer to Table 10–5 through Table 10–8.
The Data register is a page-size window into the NAND
flash. By reading from or writing to locations starting at
Data 0x10 this offset, the software reads directly from or writes
directly to the page and block of NAND flash memory
specified by the Control register.
The host uses indexed addressing as follows:

1. Program the 32-bit index-address field into the Control register at offset 0x0 of the
data/command slave port. This action provides the flash address parameters to
the NAND flash controller.
2. Perform 32-bit read or write in the Data register at offset 0x10 of the
data/command slave port.
3. Perform additional 32-bit reads and writes if they are in the same page and block
in flash memory.
It is unnecessary to write to the control register for every data transfer if a group of
data transfers targets the same page and block address. For example, you can write
the control register at the beginning of a page with the block and page address, and
then read or write the entire page by directing consecutive transactions to the Data
register.
Command Mapping
The NAND flash controller supports several flash controller-specific MAP
commands, providing an abstraction level for programming a NAND flash device. By
using the MAP commands, you can avoid directly programming device-specific
commands. Using this abstraction layer provides enhanced performance. Commands
take multiple cycles to send off-chip. The MAP commands let you initiate commands
and let the flash controller sequence them off-chip to the NAND device.
The NAND flash controller supports the following flash controller-specific MAP
commands:
■ MAP00 Commands—boot-read or buffer read/write during read-modify-write
operations
■ MAP01 Commands—memory arrays read/write

■ MAP10 Commands—NAND flash controller commands

■ MAP11 Commands—low-level direct access
MAP00 Commands
MAP00 commands access a page buffer in the NAND flash device. Addressing
always begins at 0x0 and extends to the page size specified by the
device_main_area_size and device_spare_area_size registers in the config group.
You can use this command to perform a boot read. Use MAP00 commands in
read-modify-write (RMW) operations to read or write any word in the buffer. MAP00
commands allow a direct data path to the page buffer in the device.
The host can access the page buffer directly using the MAP00 commands only if there
are no other MAP01 or MAP10 commands active on the NAND flash controller.
Table 10–5 lists the address bits as interpreted by the NAND flash controller for a
MAP00 command.
Table 10–5. MAP00 Address Mapping

Address Bits Name Description
31:28 (reserved) Set to 0
27:26 CMD_MAP Set to 0
Data width-aligned buffer address on the memory device.
12:2 BUFF_ADDR
Maximum page access is 8 KB.
The usage of this command under normal operations is limited to the following
situations:
■ It can be used to perform an Execute-in-Place (XIP) boot from the device; reading
directly from the page buffer while booting directly from the device.
■ MAP00 commands can be used to perform RMW operations where MAP00 writes
are used to modify a read page in the device page buffer. Because the NAND flash
controller does not perform ECC correction during such an operation, this method
is not recommended in an MLC device.
■ In association with MAP11 commands, MAP00 commands provide a way for the
host to directly access the device bypassing the hardware abstractions provided by
NAND flash controller with MAP01 and MAP10 commands. This method is also
used for debugging, or for issuing an operation that the flash controller might not
support with MAP01 or MAP10 commands.
Restrictions:
■ MAP00 commands cannot be used with MAP01 commands to read part of a page.
Accesses using MAP01 commands must perform a complete page transfer.
■ No ECC is performed during a MAP00 data access.
■ DMA must be disabled (the flag bit of the dma_enable register in the dma group
must be set to 0) while performing MAP00 operations.

MAP01 Commands
MAP01 commands transfer complete pages between the host memory and a specific
page of the NAND flash device. Because the NAND flash controller supports only
page addresses, the entire page must be read or written at once. The actual number of
commands required depends on the size of the data transfer. You must use the same
address until the entire page is transferred, even if multiple commands are required.
When the NAND flash controller receives a read command, it issues a load operation
on the device, waits for the load to complete, and then returns read data. Read data
must be read from the start of the page to the end of the page. Write data must be
written from the start of the page to the end of the page.
When the NAND flash controller receives confirmation of the transfer, it issues
commands to program the data into the device. The flash controller ignores the byte
enables for read and write commands and transfers the entire data width.
MAP01 command.

23:<M> (1) BLK_ADDR Block address in the device
(<M>-1):0 (1) PAGE_ADDR Page address in the device
(1) <M> depends on the number of pages per block in the device. <M> = ceil ( log2 (<device pages per block> )).
Therefore, use the following values:
32 pages per block: <M>=5
The NAND flash controller incorporates ECC on-the-fly correction that corrects data
read from the device internally before transferring the data out from the flash
controller. The ECC sector buffers store data, while the ECC engine computes the
error location.

Use the MAP01 command as follows:

■ A complete page must be read or written using a MAP01 command. During such
transfers, every transaction from the host must have the same block and page
address. The NAND flash controller internally keeps track of how much of data it
reads or writes.
■ MAP00 commands cannot be used in between using MAP01 commands for
reading or writing a page.
■ DMA must be disabled (the flag bit of the dma_enable register in the dma group
must be set to 0) while the host is performing MAP01 operations directly. If the
host issues MAP01 commands to the NAND flash controller while DMA is
enabled, the flash controller discards the request and generates an unsup_cmd
interrupt.
MAP10 Commands
MAP10 commands provide an interface to the control plane of the NAND flash
controller. MAP10 commands control special functions of the flash device, such as
erase, lock, unlock, copy back, and page spare area access. Data passed in this
command pathway targets the NAND flash controller rather than the flash device.
Unlike other command types, the data (input or output) related to these transactions
does not affect the contents of the flash device. Rather, this data specifies and
performs the exact commands of the flash controller. Only the lower 16 bits of the
Data register contain the relevant information.
MAP10 command.

23:<M> (1) BLK_ADDR Block address in the device
(<M>-1):0 (1) PAGE_ADDR Page address in the device
(1) <M> depends on the number of pages per block in the device. <M> = ceil ( log2 (<device pages per block> )).
Therefore, use the following values:

Table 10–8 lists the special functions defined by MAP10 command.

Table 10–8. MAP10 Operations
Command Function
0x01 Sets block address for erase and initiates operation
0x10 Sets unlock start address
0x11 Sets unlock end address and initiates unlock
0x21 Initiates a lock of all blocks
0x31 Initiates a lock-tight of all blocks
0x41 Sets up for spare area access
0x42 Sets up for default area access
0x43 Sets up for main+spare area access
0x60 Loads page to the buffer for a RMW operation
0x61 Sets the destination address for the page buffer in RMW operation
0x62 Writes the page buffer for a RMW operation
0x1000 Sets copy source address
0x11<PP> Sets copy destination address and initiates a copy of <PP> pages
0x20<PP> Sets up a pipeline read-ahead of <PP> pages
0x21<PP> Sets up a pipeline write of <PP> pages

■ MAP10 commands should be used to issue commands to the device, such as erase,
copy-back, lock, or unlock.
■ MAP10 pipeline commands should also be used to read or write consecutive
multiple pages from the flash device within a device block boundary. The host
must first issue a MAP10 pipeline read or write command and then issue MAP01
commands to do the actual data transfers. The MAP10 pipeline read or write
command instructs the NAND flash controller to use high-performance
commands such as cache or multiplane because the flash controller has knowledge
of multiple consecutive pages to be read. The pages must not cross a block
boundary. If a block boundary is crossed, the flash controller generates an
unsupported command (unsup_cmd) interrupt and drops the command.
■ Up to four pipeline read or write commands can be issued to the NAND flash
controller.
■ While the NAND flash controller is performing MAP10 pipeline read or write
commands, DMA must be disabled (the flag bit of the dma_enable register in the
dma group must be set to 0). DMA must be disabled because the host is directly
transferring data from and to the flash device through the flash controller.
MAP11 Commands
MAP11 commands provide direct access to the NAND flash controller’s address and
control cycles, allowing software to issue the commands directly to the flash device
using the Command and Data registers. The MAP11 command is useful if the flash
device supports a device-specific command not supported by standard flash
commands. It can also be useful for low-level debugging.

MAP11 commands provide a direct control path to the flash device. These commands
execute command, address, and data read/write cycles directly on the NAND device
interface. Command, address, and write data values are placed in the Data register.
On a read, the returned data also appears in the Data register. The indirect address
register encodes the control operation type. Command and address cycles to the
device must be a write transaction on the host bus. For data cycles, the type of
transaction on the host bus (read/write) determines the data cycle type on the device
interface. The host can issue only single-beat accesses to the data slave port while
using MAP11 commands.
MAP11 command.
Table 10–9. MAP11 Addressing Mapping

Sets the control type as follows:
■ 0 = Command cycle
1:0 TYPE
■ 1 = Address cycle
■ 2 = Data Read/Write Cycle

■ Use MAP11 commands only in special cases, for debugging or sending
device-specific commands that are not supported by the NAND flash controller.
■ DMA must be disabled before you use MAP11 operations.
■ The host can use only single beat access transfers when using MAP11 commands.
1 MAP11 commands provide direct, unstructured access to the NAND flash device.
Incorrect use can lead to unpredictable behavior.
Data DMA
The DMA transfers data with minimal host involvement. Software initiates data DMA
with the MAP10 command.
The flag bit of the dma_enable register in the dma group enables data DMA
functionality. Only enable or disable this functionality when there are no active
transactions pending in the NAND flash controller. When the DMA is enabled, the
flash controller initiates one DMA transfer per MAP10 command over the DMA
master interface. When the DMA is disabled, all operations with the flash controller
occur through the data/command slave interface.

The NAND flash controller supports up to four outstanding DMA commands, and
ignores additional DMA commands. If software issues more than four outstanding
DMA commands, the flash controller issues the unsup_cmd interrupt. On receipt of a
DMA command, the flash controller performs command sequencing to transfer the
number of pages requested in the DMA command. The DMA master reads or writes
page data from the system memory in programmed burst-length chunks. After the
DMA command completes, the flash controller issues an interrupt, and starts working
on the next queued DMA command.
Pipelining allows the NAND flash controller to optimize its performance while
executing back-to-back commands of the same type.
With certain restrictions, non-DMA MAP10 commands can be issued to the NAND
flash controller while the flash controller is servicing DMA transactions. MAP00,
MAP01, and MAP11 commands cannot be issued while DMA mode is enabled
because the flash controller is operating in an extremely tightly-coupled,
high-performance data transfer mode. On receipt of erroneous commands (MAP00,
MAP01 or MAP11), the flash controller issues an unsup_cmd interrupt to inform the
host about the violating command.
When the host issues a data DMA command the NAND flash controller transfers data
between the flash device and host memory if data DMA is enabled (the flag bit of the
dma_enable register in the dma group is set to 1). On the completion of the transfer the
flash controller informs the host by asserting an interrupt.
■ A data DMA command is a type of MAP10 command. This command is
interpreted by the data DMA engine and not by the flash controller core.
■ No MAP01, MAP00, or MAP11 commands are allowed when DMA is enabled.
■ Before the flash controller can accept data DMA commands, DMA must be
enabled by setting the flag bit of the dma_enable register in the dma group.
■ When DMA is enabled and the DMA engine initiates data transfers, ECC can be
enabled for as-needed data correction concurrent with the data transfer.
■ MAP10 commands are used along with data movements similar to MAP01
commands.
■ With the exception of data DMA commands and MAP10 pipeline read and write
commands, all other MAP10 commands such as erase, lock, unlock, and copy-back
are forwarded to the flash controller.
■ At any time, up to four outstanding data DMA commands can be handled by flash
controller. During multi-page operations, the DMA transfer must not cross a flash
block boundary. If it does, the flash controller generates an unsupported command
(unsup_cmd) interrupt and drops the command.
■ Data DMA commands are typically multi-page read and write commands with an
associated pointer in host memory. The multi-page data is transferred to or from
the host memory starting from the host memory pointer.
■ Data DMA uses the flash_burst_length register in the dma group to determine
the burst length value to drive on the interconnect. The data DMA hardware does
not account for the interconnect’s boundary crossing restrictions. The host must
initialize the starting host address so that the DMA master burst does not cross a
4 KB boundary.

There are two methods for initiating a DMA transaction: the multitransaction DMA
command, and the burst DMA command.
Multitransaction DMA Command

To initiate DMA with a multitransaction DMA command, you send four
command-data pairs to the NAND flash controller’s data and control slave port, as
shown in Table 10–10 through Table 10–13.
The NAND flash controller processes multitransaction DMA commands only if it
receives all four command-data pairs in order. The flash controller responds to
out-of-order commands with an unsup_cmd interrupt. The flash controller also
responds with an unsup_cmd interrupt if sequenced commands are interleaved with
other flash controller MAP commands.
Table 10–10 through Table 10–13 show the format of each command-data pair.
Table 10–10. Command-Data Pair 1

31:28 27:26 25:24 23:<M> (1) (<M> – 1):0 (1)
Command 0x0 0x2 0x0 Block address Page address
31:16 15:12 11:8 7:0
0x0 = Read
Data 0x0 0x2 <PP>= Number of pages
0x1 = Write
(1) <M> depends on the number of pages per block in the device. <M> = ceil ( log2 (<device pages per block> )). Therefore, use the following
values:

31:28 27:26 25:24 23:8 7:0
Command 0x0 0x2 0x0 memory address high (1) 0x0
31:16 15:12 11:8 7:0
Data 0x0 0x2 0x2 0x0

(1) The buffer address in host memory, which must be aligned to 32 bits


31:28 27:26 25:24 23:8 7:0
Command 0x0 0x2 0x0 memory address low (1) 0x0
31:16 15:12 11:8 7:0


31:28 27:26 25:24 23:17 16 15:8 7:0
INT (1)
Command 0x0 0x2 0x0 0x0 Burst length 0x0
31:16 15:12 11:8 7:0

(1) INT specifies the host interrupt to be generated at the end of the complete DMA transfer. INT controls the value of the dma_cmd_comp bit of the
intr_status0 register in the status group at the end of the DMA transfer. INT can take on one of the following values:
0—Do not interrupt host. The dma_cmd_comp bit is set to 0.
1—Interrupt host. The dma_cmd_comp bit is set to 1.
For more information, refer to “Indexed Addressing” on page 10–6.

If you want the NAND flash controller DMA to perform cacheable accesses then you
must configure the cache bits by writing the l3master register in the nandgrp group in
the system manager. The NAND flash controller DMA must be idle before you use the
system manager to change its cache capabilities.
f For more information about the system manager, refer to the System Manager chapter
in volume 3 of the Cyclone V Device Handbook.
You can issue non-DMA MAP10 commands while the NAND flash controller is in
DMA mode. For example, you might trigger a host-initiated page move between
DMA commands, to achieve wear leveling. However, do not interleave non-DMA
MAP10 commands between the command-data pairs in a set of multitransaction
DMA commands. You must issue all four command-data pairs shown in Table 10–10
through Table 10–13 before sending a different command.
Do not issue MAP00, MAP01 or MAP11 commands while DMA is enabled.
MAP10 commands in multitransaction format are written to the Data register at offset
0x10 in nanddata, the same as MAP10 commands in increment four (INCR4) format
(described in “Burst DMA Command”).

Burst DMA Command

You can initiate a DMA transfer by sending a command to the NAND flash controller
as a burst transaction of four 16-bit accesses. This form of DMA command might be
useful for initiating DMA transfers from custom IP in the FPGA fabric. Most
processor cores cannot use this form of DMA command, because they cannot control
the width of the burst.
When DMA is enabled, the NAND flash controller recognizes the MAP10 pipeline
DMA command, in the format shown in Table 10–14, as an INCR4 command. The
address decoding for MAP10 pipeline DMA command remains the same, as shown in
Table 10–7 on page 10–9. Table 10–14 lists the MAP10 burst DMA command structure.
The burst DMA command carries the same information as the multitransaction DMA
command-data pairs, but in a very different format.
MAP10 commands in INCR4 format are written to the Data register at offset 0x10 in
nanddata, the same as MAP10 commands in multitransaction format (described in
“Multitransaction DMA Command”).
Table 10–14. MAP10 Burst DMA (INCR4) Command Structure

Data Beat 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Beat 0 0x2 0x0: read, 0x1: write <PP>= number of pages
Beat 1 (1) memory address high
Beat 2 (1) memory address low
INT Burst length
Beat 3 0x0
(2)

(2) INT specifies the host interrupt to be generated at the end of the complete DMA transfer. INT controls the value of the dma_cmd_comp bit of the
intr_status0 register in the status group at the end of the DMA transfer. INT can take on one of the following values:
0—Do not interrupt host. The dma_cmd_comp bit is set to 0.
1—Interrupt host. The dma_cmd_comp bit is set to 1.
You can optionally send the 16-bit fields in Table 10–14 to the NAND flash controller
as four separate bursts of length 1 in sequential order. Altera recommends this
method.
If you want the NAND flash controller DMA to perform cacheable accesses then you
must configure the cache bits by writing the l3master register in the nandgrp group in
the system manager. The NAND flash controller DMA must be idle before you use the
system manager to modify its cache capabilities.
f For more information about the system manager, refer to the System Manager chapter
ECC
The NAND flash controller incorporates ECC logic to calculate and correct bit errors.
The flash controller uses a Bose-Chaudhuri-Hocquenghem (BCH) algorithm for
detection of multiple errors in a page.

The NAND flash controller supports 512- and 1024-byte ECC sectors. The flash
controller inserts ECC check bits for every 512 or 1024 bytes of data, depending on the
selected sector size. After 512 or 1024 bytes, the flash controller writes the ECC check
bit information to the device page.
ECC information is striped in between 512 or 1024 bytes of data across the page. The
NAND flash controller reads ECC information in the same pattern and the presence of
errors is calculated according to 512 or 1024 bytes of data read.
Table 10–15 lists the relationship between different correction capabilities, sector sizes,
and the required check-bit sizes written into the spare area.
Table 10–15. Correction Capability, Sector Size, and Check-bit Size
Correction Sector Size in Bytes Check-bit Size in Bytes
4 512 8
8 512 14
16 512 26
24 1024 46
The NAND flash controller provides the following ECC programming modes that
software uses to format a page:
■ Main Area Transfer Mode
■ Spare Area Transfer Mode
■ Main+Spare Area Transfer Mode
Main Area Transfer Mode

In main area transfer mode, when ECC is enabled, the NAND flash controller inserts
ECC check bits in the data stream on writes and strips ECC check bits on reads.
Software does not need to manage the ECC sectors when writing a page. ECC
checking is performed by the flash controller, so software simply transfers the data.
If ECC is turned off, the NAND flash controller does not read or write ECC check bits.
Figure 10–2 shows the main area transfer mode programming model.
Figure 10–2. Main Area Transfer Mode Programming Model for ECC
Sector 0 Sector 1 Sector 2 Sector 3
Spare Area Transfer Mode

The NAND flash controller does not introduce or interpret ECC check bits in spare
area transfer mode, and acts as pass through for data transfer. Figure 10–3 shows the
spare area transfer mode programming model.
Figure 10–3. Spare Area Transfer Mode Programming Model for ECC
Sector 3 ECC3 Flags

Main+Spare Area Transfer Mode

In main+spare area transfer mode, the NAND flash controller expects software to
format a page as shown in Figure 10–4. When ECC is enabled during a write
operation, the flash controller-generated ECC check bits replace the ECC check bit
data provided by software. During read operations, the flash controller forwards the
ECC check bits from the device to the host. If ECC is disabled, page data received
from the software is written to the device, and read data received from the device is
forwarded to the host.
Figure 10–4. Main+Spare Area Transfer Mode Programming Model for ECC
Sector 0 ECC0 Sector 1 ECC1 Sector 2 ECC2 Sector 3 ECC3 Flags
Preserving Bad Block Markers

When flash device manufacturers test their devices at the time of manufacture, they
mark any bad device blocks that are found. Each bad block is marked at specific,
known offsets, typically at the base of the spare area. A bad block marker is any byte
value other than 0xFF (the normal state of erased flash).
Bad block markers can be overwritten by the last sector data in a page when ECC is
enabled. This happens because the NAND flash controller also uses the main area of a
page to store ECC information, which causes the last sector to spill over into the spare
area. It is necessary for the system to preserve the bad block information prior to
writing data, to ensure the correct identification of bad blocks in the flash device.
You can configure the NAND flash controller to skip over a specified number of bytes
when it writes the last sector in a page to the spare area. This option allows the flash
controller to preserve bad block markers. To use this option, write the desired offset to
the spare_area_skip_bytes register in the config group. For example, if the device
page size is 2 KB, and the device manufacturer stores the bad block markers in the
first two bytes in the spare area, set the spare_area_skip_bytes register to 2. When
the flash controller writes the last sector of the page that overlaps with the spare area,
it starts at offset 2 in the spare area, skipping the bad block marker at offset 0. A value
of 0 (default) specifies that no bytes are skipped. The value of spare_area_skip_bytes
must be an even number. For example, if the bad block marker is a single byte, set
spare_area_skip_bytes to 2.
In main area transfer mode, the NAND flash controller does not skip the bad block
marker. Instead, it overrides the bad block marker with the value programmed in the
spare_area_marker register in the config group. This 8-bit register is used in
conjunction with the spare_area_skip_bytes register in the config group to
determine which bytes in the spare area of a page should be written with a the new
marker value. For example, to mark a block as good set the spare_area_marker
register to 0xFF and set the spare_area_skip_bytes register to the number of bytes
that the marker should be written to, starting from the base of the spare area.
In the spare area transfer mode, the NAND flash controller ignores the
spare_area_skip_bytes and spare_area_marker registers. The flash controller
transfers the data exactly as received from the host or device.

In the main+spare area transfer mode, the NAND flash controller starts writing the
last sector in a page into the spare area, starting at the offset specified in the
spare_area_skip_bytes register. However, the area containing the bad block
identifier information is overwritten by the data the host writes into the page. The
host writes both the data sectors and the bad block markers. The flash controller
depends on the host software to set up the bad block markers properly before writing
the data.
For more information about the formatting of this data, refer to Figure 10–4 on
page 10–17.
Figure 10–5 shows an example of how the NAND flash controller can skip over a bad
block marker. In this example, the flash device has a 2-KB page with a 64-byte spare
area. A 14-byte sector ECC is shown, with 8 byte per sector correction.
Figure 10–5. Bad Block Marker
Bad Block Marker

2-KByte Main Area 64-Byte Spare Area
Sector 0 ECC 0 Sector 1 ECC 1 Sector 2 ECC 2 Sector 3 Sector 3 ECC 3 Other Flags
512 Bytes 14 Bytes 512 Bytes 14 Bytes 512 Bytes 14 Bytes 470 Bytes 2 Bytes 42 Bytes 14 Bytes 6 Bytes
(Skip)
f For detailed information about configuring the NAND flash controller for default,
spare, or main+spare area transfer mode, refer to “Transfer Mode Operations” on
page 10–25.
Error Correction Status

The ECC error correction information (ECCCorInfo_b01) register, in the ecc group,
contains error correction information for each read or write that the NAND flash
controller performs. The ECCCorInfo_b01 register contains ECC error correction
information in the max_errors_b0 and uncor_err_b0 fields.
At the end of data correction for the transaction in progress, ECCCorInfo_b01 holds the
maximum number of corrections applied to any ECC sector in the transaction. In
addition, this register indicates whether the transaction as a whole has correctable
errors, uncorrectable errors, or no errors at all. A transaction has no errors when none
of the ECC sectors in the transaction has any errors. The transaction is marked as
uncorrectable if any one of the sectors is uncorrectable. The transaction is marked as
correctable if any one sector has correctable errors and none is uncorrectable.
At the end of each transaction, the host must read this register. The value of this
register provides data to the host about the block. The host can take corrective action
after the number of correctable errors encountered reaches a particular threshold
value.

NAND Flash Controller Programming Model
Interface Signals
Table 10–16 lists I/O pin use for the NAND flash interface signals.
Table 10–16. NAND Flash Interface Signals

Signal Width I/O Description
ad 8 in/out Command, address and data for the flash device
ale 1 out Address latch enable
ce_n 1 out Output Active-low chip enable
cle 1 out Command latch enable
re_n 1 out Active-low read enable signal
rb 1 in Ready/busy signal
we_n 1 out Active-low write enable signal
wp_n 1 out Active-low write protect signal
The HPS I/O pins support a single x8 device.

This section describes how the NAND flash controller is to be programmed by
software running on the microprocessor unit (MPU).
1 If you write a configuration register and follow it up with a data operation that is
dependent on the value of this configuration register, Altera recommends that you
read the value of the register before performing the data operation. This read
operation ensures that the posted write of the register is completed and takes effect
before the data operation is issued to the NAND flash controller.
Basic Flash Programming

This section describes the steps that must be taken by software to access and control
NAND flash controller.
NAND Flash Controller Optimization Sequence

The software must configure the flash device for interrupt or polling mode, using the
bank0 bit of the rb_pin_enabled register in the config group. If the device is in polling
mode, the software must also program the additional registers, to select the times and
frequencies of the polling. Program the following registers in the config group:
■ Set the rb_pin_enabled register to the desired mode of operation for each flash
device.
■ For polling mode, set the load_wait_cnt register to the appropriate value
depending on the speed of operation of the NAND flash controller, and the
desired wait value.
■ For polling mode, set the program_wait_cnt register to the appropriate value by
software depending on the speed of operation of the NAND flash controller, and
the desired wait value.

■ For polling mode, set the erase_wait_cnt register to the appropriate value by
■ For polling mode, set the int_mon_cyccnt register to the appropriate value by
At any time, the software can change any flash device from interrupt mode to polling
mode or vice-versa, using the bank0 bit of the rb_pin_enabled register.
The software needs to ensure that the particular flash device does not have any
outstanding transactions before changing the mode of operation for that particular
flash device.
Device Initialization Sequence

At initialization, host software must program the following registers in the config
group:
■ Set the devices_connected register to 1.
■ Set the device_width register to 8
■ Set the device_main_area_size register to the appropriate value.
■ Set the device_spare_area_size register to the appropriate value.
■ Set the pages_per_block register according to the parameters of the flash device.
■ Set the number_of_planes register according to the parameters of the flash device.
■ If the device allows two ROW address cycles, the flag bit of the
two_row_addr_cycles register must be set to 1. The host program can ensure this
condition either of the following ways:
■ Set the flag bit of the bootstrap_two_row_addr_cycles register to 1 prior to
the NAND flash controller’s reset initialization sequence, causing the flash
controller to initialize the bit automatically.
■ Set the flag bit of the two_row_addr_cycles register directly to 1.
■ Clear the chip_enable_dont_care register in the config group to 0.
The NAND flash controller can identify the flash device features, allowing you to
initialize the flash controller registers to interface correctly with the device, as
described in “Discovery and Initialization” on page 10–2.
However, a few NAND devices do not follow any universally accepted identification
protocol. If connected to such a device, the NAND flash controller cannot identify it
correctly. If you are using such a device, your software must use other means to
ensure that the initialization registers are set up correctly.

Device Operation Control

This section provides a list of registers that you need to program while choosing to
use multi-plane or cache operations on the device. If the device does not support
multi-plane operations or cache operations, then these registers can be left at their
power-on reset values with no impact on the functionality of the NAND flash
controller. Even if the device supports these sequences, the software may choose not
to use these sequences and can leave these registers at their power-on reset values.
Program the following registers in the config group to achieve the best performance
from a given device:
■ Set flag bit in the multiplane_operation register in the config group to 1 if the
device supports multi-plane operations to access the data on the flash device
connected to the NAND flash controller. If the flash controller is set up for
multi-plane operations, the number of pages to be accessed is always a multiple of
the number of planes in the device.
■ If the NAND flash controller is configured for multi-plane operation, and if the
device has support for multi-plane read command sequence, then set the
multiplane_read_enable register in the config group.
■ If the device implements multiplane address restrictions, set the flag bit in the
multiplane_addr_restrict register to 1.
■ Initialize the die_mask and first_block_of_next_plane registers as per device
requirements.
■ If the device supports cache command sequences, enable the cache_write_enable
and cache_read_enable registers in the config group.
■ Clear the flag bit of the copyback_disable register in the config group to 0 if the
device does not support the copyback command sequences. The register defaults
to enabled state.
■ The read_mode, write_mode and copyback_mode registers, in the config group,
currently need not be written by software, because the NAND flash controller is
capable of using the correct sequences based on a combination of some
multi-plane or cache-related settings of the NAND flash controller and the
manufacturer ID. If at some future time these settings change, program the
registers to accommodate the change.
ECC Enabling
Before you start any data operation on the flash device, you need to decide whether
you want to have the ECC enabled or disabled. If the ECC needs to be enabled, then
set up the appropriate correction level depending on the page size and the spare area
available on the device.
Set the flag bit in the ecc_enable register in the config group to 1 to enable ECC. If
enabled, the following registers in the config group need to be programmed
accordingly, else they can be ignored.
■ Initialize the ecc_correction register to the appropriate correction level.
■ Program the spare_area_skip_bytes and spare_area_marker registers in the
config group if the software needs to preserve the bad block marker.
For detailed information about ECCs, refer to “ECC” on page 10–15.

NAND Flash Controller Performance Registers

These registers specify the size of the bursts on the device interface, which maximizes
the overall performance on the NAND flash controller.
Initialize the flash_burst_length register in the dma group to a value which
maximizes the performance of the device interface by minimizing the number of
bursts required to transfer a page.
Interrupt and DMA Enabling

Prior to initiating any data operation on the NAND flash controller, the software must
set appropriate interrupt status register bits. If the software chooses to use the DMA
logic in the flash controller, then the appropriate DMA enable and interrupts bits in
the register space must be set.
■ Set the flag bit in the global_int_enable register in the config group to 1, to
enable global interrupt.
■ Set the relevant bits of the intr_en0 register in the status group to 1 before
sending any operations if the flash controller is in interrupt mode.
■ Enable DMA if your application needs DMA mode. Enable DMA by setting the
flag bit of the dma_enable register in the dma group. Altera recommends that the
software reads back this register to ensure that the mode change is accepted before
sending a DMA command to the flash controller.
■ If the DMA is enabled, then set up the appropriate bits of the dma_intr_en register
in the dma group.
Order of Interrupt Status Bits Assertion

The following interrupt status bits, in the intr_status0 register in the status group,
are listed in the order of interrupt bit setting:
1. time_out—All other interrupt bits are set to 0 when the watchdog time_out bit is
asserted.
2. dma_cmd_comp—This interrupt status bit is the last to be asserted during a DMA
operation to transfer data. This bit signifies the completion of data transfer
sequence.
3. pipe_cpybck_cmd_comp—This bit is asserted when a copyback command or the
last page of a pipeline command completes.
4. locked_blk—This bit is asserted when a program (or erase) is performed on a
locked block.
5. INT_act—No relationship with other interrupt status bits. Indicates a transition
from 0 to 1 on the ready_busy pin value for that flash device.
6. rst_comp—No relationship with other interrupt status bits. Occurs after a reset
command has completed.
7. For an erase command:
a. erase_fail (if failure)
b. erase_comp

8. For a program command:

a. locked_blk (if performed on a locked block)
b. pipe_cmd_err (if the pipeline sequence is broken by a MAP01 command)
c. page_xfer_inc (at the end of each page data transfer)
d. program_fail (if failure)
e. pipe_cpybck_cmd_comp
f. program_comp
g. dma_cmd_comp (If DMA enabled)
9. For a read command:
a. pipe_cmd_err (if the pipeline sequence is broken by a MAP01 command)
b. page_xfer_inc (at the end of each page data transfer)
c. pipe_cpybck_cmd_comp
d. load_comp
e. ecc_uncor_error (if failure)
f. dma_cmd_comp (If DMA enabled)
Timing Registers
You must optimize the following registers for your flash device’s speed grade and
clock frequency. The NAND flash controller operates correctly with the power-on
reset values. However, functioning with power-on reset values is a non-optimal mode
that provides loose timing (large margins to the signals).
Set the following registers in the config group to optimize the NAND flash controller
for the speed grade of the connected device and frequency of operation of the flash
controller:
■ twhr2_and_we_2_re
■ tcwaw_and_addr_2_data
■ re_2_we
■ acc_clks
■ rdwr_en_lo_cnt
■ rdwr_en_hi_cnt
■ max_rd_delay
■ cs_setup_cnt
■ re_2_re

Registers to Ignore
You do not need to initialize the following registers in the config group:
■ The transfer_spare_reg register. The data transfer mode can be initialized using
MAP10 commands.
■ The write_protect register need not be initialized unless you are testing the write
protection feature.
Flash-Related Special Function Operations

This section describes all the special functions that can be performed on the flash
memory. The functions are defined by MAP10 commands as described in “Command
Mapping” on page 10–6.
Erase Operations
Before data can be written to flash, an erase cycle must occur. The NAND flash
memory controller supports single block and multi-plane erases. The controller
decodes the block address from the indirect addressing shown in Table 10–7 on
page 10–9.
Single Block Erase

A single command is needed to complete a single-block erase, as follows:
1. Write to the command register, setting the CMD_MAP field to 2 and the BLK_ADDR field
to the desired erase block.
2. Write 0x01 to the Data register.
For a single block erase, the register multiplane_operation in the config group must
be reset.
After device completes erase operation, the controller generates an erase_comp
interrupt. If the erase operation fails, the erase_fail interrupt is issued. The failing
block's address is updated in the err_block_addr0 register in the status group.
Multi-Plane Erase
For multi-plane erases, the number_of_planes register in the config group holds the
number of planes in the flash device, and the block address specified must be aligned
to the number of planes in the device. The NAND flash controller consecutively
erases each block of the memory, up to the number of planes available. Issue this
command as follows:
to the desired erase block.
For multi-plane erase, the register multiplane_operation in the config group must
be set.
After the device completes erase operation on all planes, the NAND flash controller
generates an erase_comp interrupt. If the erase operation fails on any of the blocks in a
multi-plane erase command, an erase_fail interrupt is issued. The failing block's
address is updated in the err_block_addr0 register in the status group.

Lock Operations
The NAND flash controller supports the following features:
■ Flash locking—The NAND flash controller supports all flash locking operations.
The flash device itself might have limited support for these functions. If the device
does not support locking functions, the flash controller ignores these commands.
■ Lock-tight—With the lock-tight feature, the NAND flash controller can prevent
lock status from being changed. After the memory is locked tight, the flash
controller must be reset before any flash area can be locked or unlocked.
Unlocking a Span of Memory Blocks

To unlock several blocks of memory, perform the following steps:
to the starting address of the area to unlock.
to the ending address of the area to unlock.
When unlocking a range of blocks, the start block address must be less than the end
block address. Otherwise, the NAND flash controller exhibits undetermined behavior.
Locking All Memory Blocks

To lock the entire memory:
to any memory address.
Setting Lock-Tight on All Memory Blocks

After the lock-tight is applied, unlocked areas cannot be locked, and locked areas
cannot be unlocked.
To lock-tight the entire memory:
to any memory address.
To disable the lock-tight, reset the memory controller.
Transfer Mode Operations

You can configure the NAND flash controller in one of the following modes of data
transfer:
■ Default area transfer mode
■ Spare area transfer mode
■ Main+spare area transfer mode

The NAND flash controller determines the default transfer mode from the setting of
transfer_spare_reg register in the config group. Use MAP10 commands to
dynamically change the transfer mode from the existing mode to the new mode. All
subsequent commands are in the new mode of transfer. You must consider that
transfer modes can be changed at logical data transfer boundaries. For example:
■ At the beginning or end of a page in case of single page read or write.
■ At the beginning or end of a complete multi-page pipeline read or write
command.
Refer to“MAP10 Commands” on page 10–9 for detailed information about the
MAP10 commands. Table 10–17 lists the functionality of the MAP10 transfer mode
commands, and their mappings to the transfer_spare_reg register in the config
group.
Table 10–17. transfer_spare_reg and MAP10 Transfer Mode Commands

MAP10 Transfer Mode Resulting NAND Flash
transfer_spare_reg
Commands Controller Mode
0 0x42 Main (1)
0 0x41 Spare
0 0x43 Main+spare
1 0x42 Main+spare (1)
1 0x41 Spare
1 0x43 Main+spare
(1) Default access mode (0x42) maps to either main (only) or main+spare mode, depending on the value of
transfer_spare_reg.
Configure for Default Area Access

You only need to configure for default area access if the transfer mode was previously
changed to spare area or main+spare area. To configure default area access:
to any block.
The NAND flash controller determines the default area transfer mode from the setting
of the transfer_spare_reg register in the config group. If it is set to 1, then the
transfer mode becomes main+spare area, otherwise it is main area.
Configure for Spare Area Access

To access only the spare area of the flash device, use the MAP10 command to set up
the NAND flash controller to read or write only the spare area on the device. After the
flash controller is set up, use MAP01 read and write commands to access the spare
area of the appropriate block and page addresses. To configure the NAND flash
controller to access the spare area only, perform the following steps:
to the target block.

Configure for Main+Spare Area Access

To configure the NAND flash controller to access the main+spare area:
to the target block.
Read-Modify-Write Operations
To read a specific page or modify a few words, bytes, or bits in a page, use the RMW
operations. A read command copies the desired data from flash memory to a page
buffer. You can then modify the information in the buffer using MAP00 buffer read
and write commands and issue another command to write that information back to
the memory.
The read-modify-write command operates on an entire page. This command is also
useful for a copy type operation, where most of a page is saved to a new location. In
this type of operation, the NAND flash controller reads the data, modifies a specified
number of words in the page, and then writes the modified page to a new location.
1 Because the data is modified within the page buffer of the flash device, the NAND
flash controller ECC hardware is not used in RMW operations. Software must update
the ECC during RMW operations.
1 For a read-modify-write command to work with hardware ECC, the entire page must
be read into system memory, modified, then written back to flash without relying on
the RMW feature.
Read-Modify-Write Operation Flow

1. The flow starts by reading a page from the memory:
■ Write to the command register, setting the CMD_MAP field to 2 and the BLK_ADDR
field to the starting address of the desired block.
■ Write 0x60 to the Data register.
This step makes the page available to you in the page buffer in the flash device.
2. Next provide destination page address:
field to the destination address of the desired block.
This step initiates the page program and provides the destination address to the
device.
3. Use MAP00 page buffer read and write commands to modify the data in the page
buffer.

4. Write the page buffer data back to memory:

field to the same destination address.
This step performs the write.
After the device completes the load operation, the NAND flash controller issues a
load_comp interrupt. A program_comp interrupt is issued when the host issues the
write command and the device completes the program operation.
If the page program operation (as a part of an RMW operation) results in a program
failure in the device, program_fail interrupt is issued. The failing page's block and
page address is updated in the err_block_addr0 and err_page_addr0 registers in the
status group.
Copy-back Operations
The NAND flash controller supports copy back operations. However, the flash device
might have limited support for this function. If you attempt to perform a copy-back
operation on a device that does not support copy-back, the NAND flash controller
triggers an interrupt. An interrupt is also triggered if the source block is not specified
before the destination block is specified, or if the destination block is not specified in
the next command following a source block specification.
The NAND flash controller cannot do ECC validation in case of copy-back
commands. The flash controller copies the ECC data, but does not check it during the
copy operation. Altera recommends that you use copy-back only if the ECC
implemented in the flash controller is strong enough so that the next access can
correct accumulated errors.
The 8-bit value <PP> specifies the number of pages for copy-back. With this feature,
the NAND flash controller can copy multiple consecutive pages with a single
command. When you issue a copy-back command, the flash controller performs the
operation in the background. The flash controller puts other commands on hold until
the current copy-back completes.
For a multi-plane device, if the flag bit in the multiplane_operation register in the
config group is set to 1, multi-plane copy-back is available as an option. In this case,
the block address specified must be plane-aligned and the value <PP> must specify
the total number of pages to copy as a multiple of the number of planes. The block
address continues incrementing, keeping the page address fixed, for the total number
of planes in the device before incrementing the page address.
A pipe_cpyback_cmd_comp interrupt is generated when the flash controller has
completed copy-back operation of all <PP> pages. If any page program operation (as
a part of copy back operation) results in a program failure in the device, the
program_fail interrupt is issued. The failing page's block and page address is
updated in the err_block_addr0 and err_page_addr0 registers in the status group.

Copying a Memory Area (Single Plane)

To copy <PP> pages from one memory location to another:
to the starting address of the area to be copied.
to the starting address of the new area to be written.
4. Write 0x11<PP> to the Data register, where <PP> is the number of pages to copy.
Copying a Memory Area (Multi Plane)

To copy <PP> pages from one memory location to another:
1. Set the flag bit of the multiplane_operation register in the config group to 1.
to the starting address of the area to be copied. The address must be plane-aligned.
to the starting address of the new area to be written. This address must also be
plane-aligned.
5. Write 0x11<PP> to the Data register, where <PP> is the number of pages to copy.
The parameter <PP> must be a multiple of the number of planes in the device.
Pipeline Read-Ahead and Write-Ahead Operations

The NAND flash controller supports pipeline read-ahead and write-ahead operations.
However, the flash device might have limited support for this function. If the device
does not support pipeline read-ahead or write-ahead, the flash controller processes
these commands as standard reads or writes.
The NAND flash controller can handle at the most four outstanding pipeline
commands, queued up in the order in which the flash controller received the
commands. The flash controller operates on the pipeline command at the head of the
queue until all the pages corresponding to the pipeline command are executed. The
flash controller then pops the pipeline command at the head of the queue and
proceeds to work on the next pipeline command in the queue.
The pipeline read-ahead function allows for a continuous reading of the flash
memory. On receiving a pipeline read command, the flash controller immediately
issues a load command to the device. While data is read out with MAP01 commands
in a consecutive or multi-plane address pattern, the flash controller maintains
additional cache or multi-plane read command sequencing for continuous streaming
of data from the flash device.
The pipeline write-ahead function allows for a continuous writing of the flash
memory. While data is written with MAP01 commands in a consecutive or
multi-plane address pattern, the NAND flash controller maintains cache or
multi-plane command sequences for continuous streaming of data into the flash
device.

MAP01 commands must read or write pages in the same sequence that the pipelined
commands were issued to the NAND flash controller. If the host issues multiple
pipeline commands, pages must be read or written in the order the pipeline
commands were issued. It is not possible to read or write pages for a second pipeline
command before completing the first pipeline command. If the pipeline sequence is
broken by a MAP01 command, the pipe_cmd_err interrupt is issued, and the flash
controller clears the pipeline command queue. The flash controller services the
violating incoming MAP01 read or write request with a normal page read or write
sequence.
For a multi-plane device that supports multi-plane programming, you must set the
flag bit of the multiplane_operation register in the config group to 1. In this case,
the data is interleaved into page-size chunks to consecutive blocks.
Pipeline read-ahead commands can read data from the queue in this interleaved
fashion. The parameter <PP> denotes the total number of pages in multiples of the
number of planes available, and the block address must be plane-aligned, which
keeps the page address constant while incrementing the block address for each
page-size chunk of data. After reading from every plane, the NAND flash controller
increments the page address and resets the block address to the initial address. You
can also use pipeline write-ahead commands in multi-plane mode. The write
operation works similarly to the read operation, holding the page address constant
while incrementing the block address until all planes are written.
1 The same four-entry queue is used to queue the address and page count for pipeline
read-ahead and write-ahead commands. This commonality requires that you use
MAP01 commands to read out all pages for a pipeline read-ahead command before
the next pipeline command can be processed. Similarly, you must write to all pages
pertaining to pipeline write-ahead command before the next pipeline command can
be processed.
Since the value of the flag bit of the multiplane_operation register in the config
group determines pipeline read-ahead or write-ahead behavior, it can only be
changed when the pipeline registers are empty.
When the host issues a pipeline read-ahead command, and the flash controller is idle,
the load operation happens immediately.
1 The read-ahead command does not return the data to the host, and the write-ahead
command does not write data to the flash address. The NAND flash controller loads
the read data. The read data is returned to the host only when the host issues MAP01
commands to read the data. Similarly, the flash controller loads the write data, and
writes it to the flash only when the host issues MAP01 commands to write the data.
A pipe_cpyback_cmd_comp interrupt is generated when the NAND flash controller

has finished processing a pipeline command and has discarded that command from
its queue. At this point of time, the host can send another pipeline command. A
pipeline command is popped from the queue, and an interrupt is issued when the
flash controller has started processing the last page of pipeline command. Hence, the
pipe_cpyback_cmd_comp interrupt is issued prior to the last page load in case of
pipeline read command and start of data transfer of the last page to be programmed
in case of pipeline writes command.

An additional program_comp interrupt is generated when the last page program

operation completes in case of pipeline write command.
If the device command set requires the NAND flash controller to issue a load
command for the last page in the pipeline read command, a load_comp interrupt is
generated after the last page load operation completes.
For pipeline write commands, if any page program results in a failure in the device, a
program_fail interrupt is issued. The failing page's block and page address is
updated in the err_block_addr0 and err_page_addr0 registers in the status group.
Pipeline commands sequence advanced commands in the device like cache and
multi-plane. When the NAND flash controller receives a multi-page read or write
pipeline command, it sequences commands sent to the device depending on settings
in the following registers, in the config group:
■ cache_read_enable
■ cache_write_enable
■ multiplane_operation
For a device that supports cache read sequences, the flag bit of the
cache_read_enable register must be set to 1. The NAND flash controller sequences
each multi-page pipeline read command as a cache read sequence. For a device that
supports cache program command sequences, cache_write_enable must be set. The
flash controller sequences each multi-page write pipeline command as a cache write
sequence.
For a device that has multi-planes and supports multi-plane program commands, the
NAND flash controller register multiplane_operation, in the config group, must be
set. On receiving the multi-page pipeline write command, the flash controller
sequences the device with multi-plane program commands and expects that the host
transfers data to the flash controller in an even-odd block increment addressing mode.
Set Up a Single Area for Pipeline Read-Ahead

To set up an area for pipeline read-ahead, perform the following steps:
to the starting address of the block to pre-read.
2. Write 0x20<PP> to the Data register, where the 0 sets this command as a
read-ahead and <PP> is the number of pages to pre-read. The pages must not cross
a block boundary. If a block boundary is crossed, the NAND flash controller
generates an unsupported command (unsup_cmd) interrupt and drops the
command.
The read-ahead command is a hint to the flash device to start loading the next page in
the page buffer as soon as the previous page buffer operation has completed. After
you set up the read-ahead, use a MAP01 command to actually read the data. In the
MAP01 command, specify the same starting address as in the read-ahead.
If the read command received following a pipeline read-ahead request is not to a
pre-read page, then an interrupt bit is set to 1 and the pipeline read-ahead or
write-ahead registers are cleared. You must issue a new pipeline read-ahead request
must be issued to re-load the same data. You must use MAP01 commands to read all
of the data that is pre-read before the NAND flash controller returns to the idle state.

NAND Flash Controller Address Map and Register Definitions
Set Up a Single Area for Pipeline Write-Ahead

To set up an area for pipeline write-ahead:
to the starting address of the block to pre-write.
2. Write 0x21<PP> to the Data register, where the value 1 sets this command as a
write-ahead and <PP> is the number of pages to pre-write. The pages must not
cross a block boundary. If a block boundary is crossed, the NAND flash controller
generates an unsupported command (unsup_cmd) interrupt and drops the
command.
After you set up the write-ahead, use a MAP01 command to actually write the data. In
the MAP01 command, specify the same starting address as in the write-ahead.
If the write command received following a pipeline write-ahead request is not to a
pre-written page, then an interrupt bit is set to 1 and the pipeline read-ahead or
write-ahead registers are cleared. You must issue a new pipeline write-ahead request
to configure the write logic.
You must use MAP01 commands to write all of the data that is pre-written before the
NAND flash controller returns to the idle state.
NAND Flash Controller Address Map and Register Definitions

■ nanddata
■ nandregs


■ Supports one 8-bit device
November 2012 1.2 ■ Show additional supported block sizes
■ Bad block marker handling
May 2012 1.1 Added programming model section.



11. SD/MMC Controller
November 2012
cv_54011-1.1
cv_54011-1.1
The hard processor system (HPS) provides a Secure Digital/MultiMediaCard

(SD/MMC) controller for interfacing to external SD and MMC flash cards, secure
digital I/O (SDIO) devices, and Consumer Electronics Advanced Transport
Architecture (CE-ATA) hard drives. The SD/MMC controller enables you to store
boot images and boot the processor system from the removable flash card. You can
also use the flash card to expand the on-board storage capacity for larger applications
or user data. Other applications include interfacing to embedded SD (eSD) and
embedded MMC (eMMC) nonremovable flash devices.
The SD/MMC controller is based on the Synopsys® DesignWare® Mobile Storage
Host (DWC_mobile_storage) controller.
f This document refers to SD/SDIO commands, which are documented in detail in the
Physical Layer Simplified Specification, Version 3.01 and the SDIO Simplified Specification
Version 2.00 as described in “References” on page 11–79.
Features of the SD/MMC Controller

The HPS SD/MMC controller offers the following features:
■ Supports HPS boot from mobile storage
■ Supports the following standards or card types:
■ SD, including eSD—version 3.0
■ SDIO, including embedded SDIO (eSDIO)—version 3.0
■ CE-ATA—version 1.1
■ MMC, including eMMC—version 4.41, 1-bit, 4-bit, and 8-bit (in some packages,
as described in Table 11–2 on page 11–3)
■ Integrated descriptor-based direct memory access (DMA)
■ Internal 4 KB receive and transmit FIFO buffer
described herein except as expressly agreed to in writing by Altera. Altera customers are advised to obtain the latest version of device specifications before relying ISO
on any published information and before placing orders for products or services. 9001:2008
Registered
†Paragraphs marked with the dagger (†) symbol are Synopsys Proprietary. Used with permission.

November 2012
Subscribe
11–2 Chapter 11: SD/MMC Controller
Features of the SD/MMC Controller
Table 11–1 shows various SD card device types and the supported voltages, bus
modes, and speeds.
The SD/MMC controller does not directly support voltage switching, card interrupts,
or back-end power control of eSDIO card devices. However, you can connect these
signals to general-purpose I/Os (GPIOs).
Table 11–1. SD Card Use Cases

Bus Speed Modes Supported
Voltages
Bus Modes Supported Default
Card Device Supported High Speed SDR12 SDR25 (1)
Speed
Type
12.5 MBps 25 MBps 12.5 MBps 25 MBps
3.3 V 1.8 V SPI 1 bit 4 bit 8 bit
25 MHz 50 MHz 25 MHz 50 MHz
SDSC (SD) v — v v — — v v — —
SDHC v v (2) v v v — v v v v
SDXC v v (2) v v v — v v v v
eSD v v (2) v v v — v v v v
SDIO v v (2) v v v — v v v v
eSDIO v v (2) v v v v (3) v v v v

(1) SDR25 speed mode requires 1.8-V signaling. Note that even if a card supports UHS-I modes (for example SDR50, SDR104, DDR50) it can still
communicate at the lower speeds (for example SDR12, SDR25).
(2) Controls the voltage switch output to support 1.8-V signalling for SD.
(3) Optional 8-bit bus mode for eSDIO is not supported in all FPGA packages.
1 Card form factors (such as mini and micro) are not enumerated in Table 11–1 because
they do not impact the card interface functionality.

Chapter 11: SD/MMC Controller 11–3
SD/MMC Controller Block Diagram and System Integration
Table 11–2 shows various MMC card device types and the supported voltage, bus
modes, and bus speeds.
The SD/MMC controller does not contain a reset output as part of the external card
interface. To reset the flash card device, consider using a general purpose output pin.
Table 11–2. MMC Use Cases

Voltages Bus Speed Modes
Max Bus Modes Supported
Max Data Supported Supported
Clock
Card Device Type Rate
Speed Default High
(MBps) 3.3 V 1.8 V SPI (1) 1 bit 4 bit 8 bit
(MHz) Speed Speed
MMC 20 2.5 v — v v — — v —
RSMMC 20 10 v — v v v — v v
MMCPlus 50 (3) 25 v — — v v v (2) v v
MMCMobile 50 6.5 v v — v — — v v
eMMC 50 25 v v — v v — v v
(1) SPI mode is obsolete in the MMC 4.41 specification.
(2) The optional 8-bit bus mode is not supported in all FPGA packages.
(3) Supports a maximum clock rate of 50 MHz instead of 52 MHz (specified in MMC specification).
SD/MMC Controller Block Diagram and System Integration

The SD/MMC controller includes a bus interface unit (BIU) and a card interface unit
(CIU). The BIU provides a slave interface for a host to access the control and status
registers (CSRs). Additionally, this unit also provides independent FIFO buffer access
through a DMA interface. The DMA controller is responsible for exchanging data
between the system memory and FIFO buffer. The DMA registers are accessible by the
host to control the DMA operation. The CIU supports the SD, MMC, and CE-ATA
protocols on the controller, and provides clock management through the clock control
block. The interrupt control block for generating an interrupt connects to the generic
interrupt controller in the ARM® Cortex™-A9 microprocessor unit (MPU) subsystem.

Functional Description of the SD/MMC Controller
Figure 11–1 shows a block diagram of the SD/MMC controller and how it integrates
in the HPS.
Figure 11–1. SD/MMC Controller Connectivity
SD/MMC Controller
Bus Interface Unit Card Interface Unit

L3 Interconnect
Master DMA
Interface Controller Data Path
Control
FIFO FIFO Command I/O Pins

Buffer Synchronizer Buffer Path Control
Control Control
Clock
L4 Bus
Slave Register Control

Interface Block
Storage
FIFO Buffer
MPU Interrupt
Subsystem Control

This section describes the SD/MMC controller components and how the controller
operates.
SD/MMC/CE-ATA Protocol
The SD/MMC/CE-ATA protocol is based on command and data bit streams that are
initiated by a start bit and terminated by a stop bit. Additionally, the SD/MMC
controller provides a reference clock and is the only master interface that can initiate a
transaction.
■ Command—a token transmitted serially on the CMD pin that starts an operation.
■ Response—a token from the card transmitted serially on the CMD pin in response to
certain commands.
■ Data—transferred serially using the data pins for data movement commands.

Figure 11–2 illustrates an example of a multiple-block read operation. The clock is a

representative only and does not show the exact number of clock cycles.
Figure 11–2. Multiple-Block Read Operation
From Host From Card Data from Stop Command

to Card to Host Card to Host Stops Data Transfer
sdmmc_cclk_out
sdmmc_cmd Command Response Command Response
sdmmc_data Data Block CRC Data Block CRC Data Block CRC
Block Read Operation Data Stop Operation
Multiple Block Read Operation
Figure 11–3 illustrates an example of a command token sent by the host in a

multiple-block write operation.
Figure 11–3. Multiple-Block Write Operation
From Host From Card Data from OK Response & Stop Command
to Card to Host Host to Card Busy from Card Stops Data Transfer
sdmmc_cclk_out
sdmmc_cmd Command Response Command Response
sdmmc_data Data Block CRC Busy Data Block CRC Busy
Block Write Operation Data Stop Operation
Multiple Block Read Operation

BIU
The BIU interfaces with the CIU, and is connected to the level 3 (L3) interconnect and
level 4 (L4) peripheral buses. The BIU consists of the following primary functional
blocks:
■ Slave interface
■ Register block
■ FIFO buffer
■ Internal DMA controller
Slave Interface
The host processor accesses the SD/MMC controller registers and data FIFO buffers
through the slave interface.
Register Block
The register block is part of the BIU and provides read and write access to the CSRs.
All registers reside in the BIU clock domain. When a command is sent to a card by
setting the start command bit (start_cmd) of the command register (cmd) to 1, all
relevant registers needed for the CIU operation are transferred to the CIU block.
During this time, software must not write to the registers that are transferred from the
BIU to the CIU. The software must wait for the hardware to reset the start_cmd bit to
0 before writing to these registers again. The register unit has a hardware locking
feature to prevent illegal writes to registers.
After a command start is issued by setting the start_cmd bit of the cmd register, the
following registers cannot be rewritten until the command is accepted by the CIU:
■ Command (cmd)
■ Command argument (cmdarg)
■ Byte count (bytcnt)
■ Block size (blksiz)
■ Clock divider (clkdiv)
■ Clock enable (clkena)
■ Clock source (clksrc)
■ Timeout (tmout)
■ Card type (ctype)
The hardware resets the start_cmd bit once the CIU accepts the command. If a host
write to any of these registers is attempted during this locked time, the write is
ignored and the hardware lock write error bit (hle) is set to 1 in the raw interrupt
status register (rintsts). Additionally, if the interrupt is enabled and not masked for a
hardware lock error, an interrupt is sent to the host.

After a command is accepted, you can send another command to the CIU—which has
a one-deep command queue—under the following conditions:
■ If the previous command is not a data transfer command, the new command is
sent to the SD/MMC/CE-ATA card once the previous command completes.
■ If the previous command is a data transfer command and if the wait previous data
complete bit (wait_prvdata_complete) of the cmd register is set to 1 for the new
command, the new command is sent to the SD/MMC/CE-ATA card only when
the data transfer completes.
■ If the wait_prvdata_complete bit is 0, the new command is sent to the
SD/MMC/CE-ATA card as soon as the previous command is sent. Typically, use
this feature to stop or abort a previous data transfer or query the card status in the
middle of a data transfer.
Interrupt Controller Unit

The interrupt controller unit generates an interrupt that depends on the rintsts
register, the interrupt mask register (intmask), and the interrupt enable bit
(int_enable) of the control register (ctrl). Once an interrupt condition is detected, the
controller sets the corresponding interrupt bit in the rintsts register. The bit in the
rintsts register remains set to 1 until the software resets the bit to 0 by writing a 1 to
the interrupt bit; writing a 0 leaves the bit untouched.
The interrupt port is an active-high, level-sensitive interrupt. The interrupt port is
active only when at least one bit in the rintsts register is set to 1, the corresponding
intmask register bit is 1, and the int_enable bit of the ctrl register is 1.
The following bits are available as top-level ports for debug purposes:
■ All bits of the intmask register
■ All bits of the rintsts register
■ The int_enable bit of the ctrl register
The int_enable bit of the ctrl register is set to 0 on power-on, and the intmask
register bits are set to 0x0000000, which masks all the interrupts.
The following conditions can cause the interrupt to occur:
■ End-bit error on read
■ No cyclic redundancy code (CRC) on write
■ Auto command done
■ Start-bit error
■ Hardware locked write error
■ FIFO buffer underflow or overflow error
■ Data starvation by host timeout
■ Data read timeout or boot data start
■ Response timeout or boot ACK received
■ Data CRC error
■ Response CRC error

■ Receive FIFO buffer data request

■ Transmit FIFO buffer data request
■ Data transfer over
■ Command done
■ Response error
The Receive FIFO Data Request and Transmit FIFO Data Request interrupts are set by
level-sensitive interrupt sources. Therefore, the interrupt source must be first cleared
before you can reset the interrupt’s corresponding bit in the rintsts register to 0.
For example, on receiving the Receive FIFO Data Request interrupt, the FIFO buffer
must be emptied so that the FIFO buffer count is not greater than the RX watermark,
which causes the interrupt to be triggered.
The rest of the interrupts are triggered by single clock-pulse-width sources.
FIFO Buffer
The SD/MMC controller has a 4-KB data FIFO buffer for storing transmit and receive
data. The FIFO buffer memory supports error correction codes (ECCs). Both interfaces
to the FIFO buffer support single and double bit error injection. The enable and error
injection pins are inputs driven by the system manager and the status pins are outputs
driven to the MPU subsystem.
The SD/MMC controller provides outputs to notify the system manager when
single-bit correctable errors are detected (and corrected), and when double-bit
(uncorrectable) errors are detected. The system manager generates an interrupt to the
GIC when an ECC error is detected.
Internal DMA Controller

The internal DMA controller has a CSR and a single transmit or receive engine, which
transfers data from system memory to the card and vice versa. The controller uses a
descriptor mechanism to efficiently move data from source to destination with
minimal host processor intervention. You can set up the controller to interrupt the
host processor in situations such as transmit and receive data transfer completion
from the card, as well as other normal or error conditions. The DMA controller and
the host driver communicate through a single data structure.
The internal DMA controller transfers the data received from the card to the data
buffer in the system memory, and transfers transmit data from the data buffer in the
memory to the controller’s FIFO buffer. Descriptors that reside in the system memory
act as pointers to these buffers.
A data buffer resides in the physical memory space of the system memory and
consists of complete or partial data. The buffer status is maintained in the descriptor.
Data chaining refers to data that spans multiple data buffers. However, a single
descriptor cannot span multiple data buffers.

A single descriptor is used for both reception and transmission. The base address of
the list is written into the descriptor list base address register (dbaddr). A descriptor
list is forward linked. The last descriptor can point back to the first entry to create a
ring structure. The descriptor list resides in the physical memory address space of the
host. Each descriptor can point to a maximum of two data buffers.
Internal DMA Controller Descriptors

The internal DMA controller uses these types of descriptor structures:
■ Dual-buffer structure—The distance between two descriptors is determined by the
skip length value written to the descriptor skip length field (dsl) of the bus mode
register (bmod).
■ Chain structure—Each descriptor points to a unique buffer, and to the next
descriptor in a linked list.
Figure 11–4 and Figure 11–5 illustrate the internal DMA controller dual-buffer
descriptor structure and chain descriptor structure respectively.
Figure 11–4. Dual-Buffer Descriptor Structure
Data Buffer 1
Descriptor A
Data Buffer 2
The Distance Between 2

Data Buffer 1
Descriptors Is Determined
by the DSL Value Programmed Descriptor B
in the BMOD Register Data Buffer 2
Data Buffer 1
Descriptor C
Data Buffer 2
Figure 11–5. Chain Descriptor Structure
Data Buffer
Descriptor A
Data Buffer
Descriptor B
Data Buffer
Descriptor C

Table 11–3 shows the internal format of a descriptor. The descriptor address must be
aligned to the 32-bit bus. Each descriptor contains 16 bytes of control and status
information. For information about each of the bits of the descriptor, refer to
Table 11–4 on page 11–11 through Table 11–7 on page 11–12.
Table 11–3. Descriptor Format

Name Offset 31 30 29 ... 26 25 ... 13 12 ... 6 5 4 3 2 1 0
OWN
CES
DIC
DES0 0 — ER CH FS LD —
DES1 4 — BS2 BS1
DES2 8 BAP1
DES3 12 BAP2 or Next Descriptor Address

The DES0 field in the internal DMA controller descriptor contains control and status
information. Table 11–4 lists the bits in this descriptor.
Table 11–4. Internal DMA Controller DES0 Descriptor Field

Bits Name Description
When set to 1, this bit indicates that the descriptor is owned by the internal DMA
controller.
31 OWN
When this bit is set to 0, it indicates that the descriptor is owned by the host. The
internal DMA controller resets this bit to 0 when it completes the data transfer.
The CES bit indicates whether a transaction error occurred. The CES bit is the
logical OR of the following error bits in the rintsts register.
■ End-bit error (ebe)
■ Response timeout (rto)
30 Card Error Summary (CES) ■ Response CRC (rcrc)
■ Start-bit error (sbe)
■ Data read timeout (drto)
■ Data CRC for receive (dcrc)
■ Response error (re)
29:6 Reserved —
When set to 1, this bit indicates that the descriptor list reached its final
descriptor. The internal DMA controller returns to the base address of the list,
5 End of Ring (ER)
creating a descriptor ring. ER is meaningful for only a dual-buffer descriptor
structure.
Second Address Chained When set to 1, this bit indicates that the second address in the descriptor is the
4 next descriptor address rather than the second buffer address. When this bit is
(CH) set to 1, BS2 (DES1[25:13]) must be all zeros.
When set to 1, this bit indicates that this descriptor contains the first buffer of the
3 First Descriptor (FS) data. If the size of the first buffer is 0, next descriptor contains the beginning of
the data.
When set to 1, this bit indicates that the buffers pointed to by this descriptor are
2 Last Descriptor (LD)
the last buffers of the data.
Disable Interrupt on When set to 1, this bit prevents the setting of the TI/RI bit of the internal DMA
1 controller status register (idsts) for the data that ends in the buffer pointed to
Completion (DIC) by this descriptor.
0 Reserved —

The DES1 descriptor field contains the buffer size. Table 11–5 lists the bits in this
descriptor.

31:26 Reserved —
These bits indicate the second data buffer byte size. The
buffer size must be a multiple of four. When the buffer
25:13 Buffer 2 Size (BS2)
size is not a multiple of four, the resulting behavior is
undefined. This field is not valid if DES0[4] is set to 1.
Indicates the data buffer byte size, which must be a
multiple of four bytes. When the buffer size is not a
multiple of four, the resulting behavior is undefined. If
this field is 0, the DMA ignores the buffer and proceeds
12:0 Buffer 1 Size (BS1) to the next descriptor for a chain structure, or to the next
buffer for a dual-buffer structure.
If there is only one descriptor and only one buffer to be
programmed, you need to use only buffer 1 and not
buffer 2.
The DES2 descriptor field contains the address pointer to the data buffer. Table 11–6
lists the bits in this descriptor.

These bits indicate the physical address of the first data
Buffer Address Pointer 1
31:0 buffer. The internal DMA controller ignores DES2 [1:0],
(BAP1)
because it only performs 32-bit-aligned accesses.
The DES3 descriptor field contains the address pointer to the next descriptor if the
present descriptor is not the last descriptor in a chained descriptor structure or the
second buffer address for a dual-buffer structure. Table 11–7 lists the bits in this
descriptor.

These bits indicate the physical address of the second
buffer when the dual-buffer structure is used. If the
Second Address Chained (DES0[4]) bit is set to 1, this
Buffer Address Pointer 2 address contains the pointer to the physical memory
31:0 (BAP2) or Next Descriptor where the next descriptor is present.
Address
If this is not the last descriptor, the next descriptor
address pointer must be aligned to 32 bits. Bits 1 and 0
are ignored.

Host Bus Burst Access

The internal DMA controller attempts to execute fixed-length burst transfers on the
master interface if configured using the fixed burst bit (fb) of the bmod register. The
maximum burst length is indicated and limited by the programmable burst length
(pbl) field of the bmod register. When descriptors are being fetched, the master
interface always presents a burst size of four to the interconnect.
The internal DMA controller initiates a data transfer only when sufficient space to
accommodate the configured burst is available in the FIFO buffer or the number of
bytes to the end of transfer is less than the configured burst-length. When the DMA
master interface is configured for fixed-length bursts, it transfers data using the most
efficient combination of INCR4/8/16 and SINGLE transactions. If the DMA master
interface is not configured for fixed length bursts, it transfers data using INCR
(undefined length) and SINGLE transactions.
Host Data Buffer Alignment

The transmit and receive data buffers in system memory must be aligned to a 32-bit
boundary.
Buffer Size Calculations

The driver knows the amount of data to transmit or receive. For transmitting to the
card, the internal DMA controller transfers the exact number of bytes from the FIFO
buffer, indicated by the buffer size field of the DES1 descriptor field.
If a descriptor is not marked as last (with the LD bit of the DES0 field set to 0) then the
corresponding buffer(s) of the descriptor are considered full, and the amount of valid
data in a buffer is accurately indicated by its buffer size field. If a descriptor is marked
as last, the buffer might or might not be full, as indicated by the buffer size in the
DES1 field. The driver is aware of the number of locations that are valid. The driver is
expected to ignore the remaining, invalid bytes.
Internal DMA Controller Interrupts

Interrupts can be generated as a result of various events. The idsts register contains
all the bits that might cause an interrupt. The internal DMA controller interrupt
enable register (idinten) contains an enable bit for each of the events that can cause
an interrupt to occur.
There are two summary interrupts—the normal interrupt summary bit (nis) and the
abnormal interrupt summary bit (ais)—in the idsts register. The nis bit results from
a logical OR of the transmit interrupt (ti) and receive interrupt (ri) bits in the idsts
register. The ais bit is a logical OR result of the fatal bus error interrupt (fbe),
descriptor unavailable interrupt (du), and card error summary interrupt (ces) bits in
the idsts register.
Interrupts are cleared by writing a 1 to the corresponding bit position. If a 0 is written
to an interrupt’s bit position, the write is ignored, and does not clear the interrupt.
When all the enabled interrupts within a group are cleared, the corresponding
summary bit is set to 0. When both the summary bits are set to 0, the interrupt signal
is de-asserted.

Interrupts are not queued. If another interrupt event occurs before the driver has
responded to the previous interrupt, no additional interrupts are generated. For
example, the ri bit of the idsts register indicates that one or more data has been
transferred to the host buffer.
An interrupt is generated only once for simultaneous, multiple events. The driver
must scan the idsts register for the interrupt cause. The final interrupt signal from the
controller is a logical OR of the interrupts from the BIU and internal DMA controller.
Internal DMA Controller FSM

The following steps show the internal DMA controller functional state machine (FSM)
operations:
1. The internal DMA controller performs four accesses to fetch a descriptor.
2. The DMA controller stores the descriptor information internally. If it is the first
descriptor, the controller issues a FIFO buffer reset and waits until the reset is
complete.
3. The internal DMA controller checks each bit of the descriptor for the correctness. If
bit mismatches are found, the appropriate error bit is set to 1 and the descriptor is
closed by setting the OWN bit in the DES0 field to 1.
The rintsts register indicates one of the following conditions:
■ Response timeout
■ Response CRC error
■ Data receive timeout
■ Response error
4. The DMA waits for the RX watermark to be reached before writing data to system
memory, or the TX watermark to be reached before reading data from system
memory. The RX watermark represents the number of bytes to be locally stored in
the FIFO buffer before the DMA writes to memory. The TX watermark represents
the number of free bytes in the local FIFO buffer before the DMA reads data from
memory.
5. If the value of the programmable burst length (PBL) field is larger than the
remaining amount of data in the buffer, single transfers are initiated. If dual
buffers are being used, and the second buffer contains no data (buffer size = 0), the
buffer is skipped and the descriptor is closed.
6. The OWN bit in descriptor is set to 0 by the internal DMA controller after the data
transfer for one descriptor is completed. If the transfer spans more than one
descriptor, the DMA controller fetches the next descriptor. If the transfer ends with
the current descriptor, the internal DMA controller goes to idle state after setting
the ri bit or the ti bit of the idsts register. Depending on the descriptor structure
(dual buffer or chained), the appropriate starting address of descriptor is loaded. If
it is the second data buffer of dual buffer descriptor, the descriptor is not fetched
again.

Abort During Internal DMA Transfer

If the host issues an SD/SDIO STOP_TRANSMISSION command (CMD12) to the
card while data transfer is in progress, the internal DMA controller closes the present
descriptor after completing the data transfer until a Data Transfer Over (DTO)
interrupt is asserted. Once a STOP_TRANSMISSION command is issued, the DMA
controller performs single burst transfers.
1. For a card write operation, the internal DMA controller keeps writing data to the
FIFO buffer after fetching it from the system memory until a DTO interrupt is
asserted. This is done to keep the card clock running so that the
STOP_TRANSMISSION command is reliably sent to the card.
2. For a card read operation, the internal DMA controller keeps reading data from
the FIFO buffer and writes to the system memory until a DTO interrupt is
generated. This is required because DTO interrupt is not generated until and
unless all the FIFO buffer data is emptied.
1 For a card write abort, only the current descriptor during which a
STOP_TRANSMISSION command is issued is closed by the internal DMA controller.
The remaining unread descriptors are not closed by the internal DMA controller.
1 For a card read abort, the internal DMA controller reads the data out of the FIFO
buffer and writes them to the corresponding descriptor data buffers. The remaining
unread descriptors are not closed.
FIFO Buffer Overflow and Underflow

During normal data transfer conditions, FIFO buffer overflow and underflow does
not occur. However, if there is a programming error, a FIFO buffer overflow or
underflow can result. For example, consider the following scenarios.
For transmit:
■ PBL=4
■ TX watermark = 1
For these programming values, if the FIFO buffer has only one location empty, the
DMA attempts to read four words from memory even though there is only one word
of storage available. This results in a FIFO Buffer Overflow interrupt.
For receive:
■ PBL=4
■ RX watermark = 1
For these programming values, if the FIFO buffer has only one location filled, the
DMA attempts to write four words, even though only one word is available. This
results in a FIFO Buffer Underflow interrupt.
The driver must ensure that the number of bytes to be transferred, as indicated in the
descriptor, is a multiple of four bytes. For example, if the bytcnt register = 13, the
number of bytes indicated in the descriptor must be rounded up to 16 because the
length field must always be a multiple of four bytes.
Table 11–8 lists the legal PBL and FIFO buffer watermark values for internal DMA
controller data transfers operations.

Table 11–8. PBL and Watermark Levels

PBL (Number of transfers) TX/RX Watermark Value
1 greater than or equal to 1
Fatal Bus Error Scenarios

A fatal bus error occurs due to an error response through the master interface. This
error is a system error, so the software driver must not perform any further setup on
the controller. The only recovery mechanism from such scenarios is to perform one of
the following tasks:
■ Issue a reset to the controller through the reset manager.
■ Issue a program controller reset by writing to the controller reset bit
(controller_reset) of the ctrl register.
CIU
The CIU interfaces with the BIU and SD/MMC cards or devices. The host processor
writes command parameters to the SD/MMC controller’s BIU control registers and
these parameters are then passed to the CIU. Depending on control register values,
the CIU generates SD/MMC command and data traffic on the card bus according to
the SD/MMC protocol. The control register values also decide whether the command
and data traffic is directed to the CE-ATA card, and the SD/MMC controller controls
the command and data path accordingly.
The following list describes the CIU operation restrictions:
■ After a command is issued, the CIU accepts another command only to check read
status or to stop the transfer.
■ Only one data transfer command can be issued at a time.
■ During an open-ended card write operation, if the card clock is stopped because
the FIFO buffer is empty, the software must first fill the data into the FIFO buffer
and start the card clock. It can then issue only an SD/SDIO
STOP_TRANSMISSION (CMD12) command to the card.
■ During an SDIO/COMBO card transfer, if the card function is suspended and the
software wants to resume the suspended transfer, it must first reset the FIFO
buffer and start the resume command as if it were a new data transfer command.

■ When issuing SD/SDIO card reset commands (GO_IDLE_STATE,

GO_INACTIVE_STATE or CMD52_reset) while a card data transfer is in progress,
the software must set the stop abort command bit (stop_abort_cmd) in the cmd
register to 1 so that the controller can stop the data transfer after issuing the card
reset command.
■ If the card clock is stopped because the FIFO buffer is full during a card read, the
software must read at least two FIFO buffer locations to start the card clock.
■ If CE-ATA card device interrupts are enabled (the nIEN bit is set to 0 in the ATA
control register), a new RW_BLK command must not be sent to the same card
device if there is a pending RW_BLK command in progress (the RW_BLK
command used in this document is the RW_MULTIPLE_BLOCK MMC command
defined by the CE-ATA specification). Only the Command Completion Signal
Disable (CCSD) command can be sent while waiting for the Command
Completion Signal (CCS).
■ For the same card device, a new command is allowed for reading status
information, if interrupts are disabled in the CE-ATA card (the nIEN bit is set to 1 in
the ATA control register).
■ Open-ended transfers are not supported for the CE-ATA card devices.
■ The send_auto_stop signal is not supported (software must not set the
send_auto_stop bit in the cmd register) for CE-ATA transfers.
The CIU consists of the following primary functional blocks:
■ Command path
■ Data path
■ Clock control
Command Path
The command path performs the following functions:
■ Load card command parameters
■ Send commands to card bus
■ Receive responses from card bus
■ Send responses to BIU
■ Load clock parameters
■ Drives the P-bit on command pin
A new command is issued to the controller by writing to the BIU registers and setting
the start_cmd bit in the cmd register. The command path loads the new command
(command, command argument, timeout) and sends an acknowledgement to the BIU.

After the new command is loaded, the command path state machine sends a
command to the card bus—including the internally generated seven-term CRC
(CRC-7)—and receives a response, if any. The state machine then sends the received
response and signals to the BIU that the command is done, and then waits for eight
clock cycles before loading a new command. In CE-ATA data payload transfer
(RW_MULTIPLE_BLOCK) commands, if the card device interrupts are enabled (the
nIEN bit is set to 0 in the ATA control register), the state machine performs the
following actions after receiving the response:
■ Does not drive the P-bit; it waits for CCS, decodes and goes back to idle state, and
then drives the P-bit.
■ If the host wants to send the CCSD command and if eight clock cycles are expired
after the response, it sends the CCSD pattern on the command pin.
Load Command Parameters

Commands or responses are loaded in the command path in the following situations:
■ New command from BIU—When the BIU sends a new command to the CIU, the
start_cmd bit is set to 1 in the cmd register.
■ Internally-generated send_auto_stop—When the data path ends, the SD/SDIO
STOP command request is loaded.
■ Interrupt request (IRQ) response with relative card address (RCA) 0x000—When
the command path is waiting for an IRQ response from the MMC and a “send irq
response” request is signaled by the BIU, the send IRQ request bit
(send_irq_response) is set to 1 in the ctrl register.
Loading a new command from the BIU in the command path depends on the
following cmd register bit settings:
■ update_clock_registers_only—If this bit is set to 1 in the cmd register, the
command path updates only the clkena, clkdiv, and clksrc registers. If this bit is
set to 0, the command path loads the cmd, cmdarg, and tmout registers. It then
processes the new command, which is sent to the card.
■ wait_prvdata_complete—If this bit is set to 1, the command path loads the new
command under one of the following conditions:
■ Immediately, if the data path is free (that is, there is no data transfer in
progress), or if an open-ended data transfer is in progress (bytcnt = 0).
■ After completion of the current data transfer, if a predefined data transfer is in
progress.

Send Command and Receive Response

After a new command is loaded in the command path (the
update_clock_registers_only bit in the cmd register is set to 0), the command path
state machine sends out a command on the card bus. Figure 11–6 illustrates the
command path state machine.
Figure 11–6. Command Path State Machine
load_new_cmd Command tNCC Done

Idle
response_expected = 0
Transmit wait_tncc
Command
Send IRQ
Response
Request
Response Done/
Receive Response Timeout
response_expected = 1 Response
The command path state machine performs the following functions, according to cmd
register bit values:
1. send_initialization—Initialization sequence of 80 clock cycles is sent before
sending the command.
2. response_expected—A response is expected for the command. After the
command is sent out, the command path state machine receives a 48-bit or 136-bit
response and sends it to the BIU. If the start bit of the card response is not received
within the number of clock cycles (as set up in the tmout register), the rto bit and
command done (CD) bit are set to 1 in the rintsts register, to signal to the BIU. If
the response-expected bit is set to 0, the command path sends out a command and
signals a response done to the BIU, which causes the cmd bit to be set to 1 in the
rintsts register.
3. response_length—If this bit is set to 1, a 136-bit long response is received; if it is
set to 0, a 48-bit short response is received.
4. check_response_crc—If this bit is set to 1, the command path compares CRC-7
received in the response with the internally-generated CRC-7. If the two do not
match, the response CRC error is signaled to the BIU, that is, the rcrc bit is set to 1
in the rintsts register.
Send Response to BIU

If the response_expected bit is set to 1 in the cmd register, the received response is sent
to the BIU. Response register 0 (resp0) is updated for a short response, and the
response register 3 (resp3), response register 2 (resp2), response register 1 (resp1),
and resp0 registers are updated on a long response, after which the cmd bit is set to 1
in the rintsts register. If the response is for an AUTO_STOP command sent by the
CIU, the response is written to the resp1 register, after which the auto command done
bit (acd) is set to 1 in the rintsts register.

A correct card response contains the fields listed in Table 11–9. The command path
verifies the contents of the card response.
Table 11–9. Card Response Fields

Field Contents
Response transmission bit 0
Command index Command index of the sent command
End bit 1
The command index is not checked for a 136-bit response or if the

check_response_crc bit in the cmd register is set to 0. For a 136-bit response and
reserved CRC 48-bit responses, the command index is reserved, that is, 0b111111.
f For more information about response values, refer to Physical Layer Simplified
Specification, Version 3.01 as described in “References” on page 11–79.
Driving P-bit on CMD Line

The command path drives a one-cycle pull-up bit (P-bit) to 1 on the CMD line
between two commands if a response is not expected. If a response is expected, the
P-bit is driven after the response is received and before the start of the next command.
While accessing a CE-ATA card device, for commands that expect a CCS, the P-bit is
driven after the response only if the interrupts are disabled in the CE-ATA card (the
nIEN bit is set to 1 in the ATA control register), that is, the CCS expected bit
(ccs_expected) in the cmd register is set to 0. If the command expects the CCS, the
P-bit is driven only after receiving the CCS.
Polling the CCS

CE-ATA card devices generate the CCS to notify the host controller of the normal ATA
command completion or ATA command termination. After receiving the response
from the card, the command path state machine performs the functions illustrated in
Figure 11–7 according to cmd register bit values.
Figure 11–7. CE-ATA Command Path State Machine
ccs_expected = 1 Response Transmit

End Bit CMD12
ccs_expected = 0
cmd_in = 0 Command send_auto_stop_ccsd

wait_CCS wait_tncc
Idle
okay_to_send_ccsd
counter_zero
send_CCSD

The following describe some of the details in Figure 11–7:

1. Response end bit state—The state machine receives the end bit of the response
from the card device. If the ccs_expected bit of the cmd register is set to 1, the state
machine enters the wait CCS state.
2. Wait CCS—The state machine waits for the CCS from the CE-ATA card device.
While waiting for the CCS, the following events can happen:
a. Software sets the send CCSD bit (send_ccsd) in the ctrl register, indicating not
to wait for CCS and to send the CCSD pattern on the command line.
b. Receive the CCS on the CMD line.
3. Send CCSD command—Sends the CCSD pattern (0b00001) on the CMD line.
CCS Detection and Interrupt to Host Processor

If the ccs_expected bit in the cmd register is set to 1, the CCS from the CE-ATA card
device is indicated by setting the data transfer over bit (dto) in the rintsts register.
The controller generates a DTO interrupt if this interrupt is not masked.
For the RW_MULTIPLE_BLOCK commands, if the CE-ATA card device interrupts are
disabled (the nIEN bit is set to 1 in the ATA control register)— that is, the
ccs_expected bit is set to 0 in the cmd register—there are no CCSs from the card. When
the data transfer is over—that is, when the requested number of bytes are
transferred—the dto bit in the rintsts register is set to 1.
CCS Timeout
If the command expects a CCS from the card device (the ccs_expected bit is set to 1 in
the cmd register), the command state machine waits for the CCS and remains in the
wait CCS state. If the CE-ATA card fails to send out the CCS, the host software must
implement a timeout mechanism to free the command and data path. The controller
does not implement a hardware timer; it is the responsibility of the host software to
maintain a software timer.
In the event of a CCS timeout, the host must issue a CCSD command by setting the
send_ccsd bit in the ctrl register. The controller command path state machine sends
the CCSD command to the CE-ATA card device and exits to an idle state. After
sending the CCSD command, the host must also send an SD/SDIO
STOP_TRANSMISSION command to the CE-ATA card to abort the outstanding ATA
command.
Send CCSD Command

If the send_ccsd bit in the ctrl register is set to 1, the controller sends a CCSD pattern
on the CMD line. The host can send the CCSD command while waiting for the CCS or
after a CCS timeout happens.
After sending the CCSD pattern, the controller sets the cmd bit in the rintsts register
and also generates an interrupt to the host if the Command Done interrupt is not
masked.
1 Within the CIU block, if the send_ccsd bit in the ctrl register is set to 1 on the same
clock cycle as CCS is sampled, the CIU block does not send a CCSD pattern on the
CMD line. In this case, the dto and cmd bits in the rintsts register are set to 1.

1 Due to asynchronous boundaries, the CCS might have already happened and the
send_ccsd bit is set to 1. In this case, the CCSD command does not go to the CE-ATA
card device and the send_ccsd bit is not set to 0. The host must reset the send_ccsd bit
to 0 before the next command is issued.
If the send auto stop CCSD (send_auto_stop_ccsd) bit in the ctrl register is set to 1,
the controller sends an internally generated STOP_TRANSMISSION command
(CMD12) after sending the CCSD pattern. The controller sets the acd bit in the
rintsts register.
I/O transmission delay (NACIO Timeout)

The host software maintains the timeout mechanism for handling the I/O
transmission delay (NACIO cycles) time-outs while reading from the CE-ATA card
device. The controller neither maintains any timeout mechanism nor indicates that
NACIO cycles are elapsed while waiting for the start bit of a data token. The I/O
transmission delay is applicable for read transfers using the RW_REG and RW_BLK
commands; the RW_REG and RW_BLK commands used in this document refer to the
RW_MULTIPLE_REGISTER and RW_MULTIPLE_BLOCK MMC commands defined
by the CE-ATA specification.
1 After the NACIO timeout, the application must abort the command by sending the
CCSD and STOP commands, or the STOP command. The Data Read Timeout (DRTO)
interrupt might be set to 1 while a STOP_TRANSMISSION command is transmitted
out of the controller, in which case the data read timeout boot data start bit (bds) and
the dto bit in the rintsts register are set to 1.
Data Path
The data path block reads the data FIFO buffer and transmits data on the card bus
during a write data transfer, or receives data and writes it to the FIFO buffer during a
read data transfer. The data path loads new data parameters—data expected,
read/write data transfer, stream/block transfer, block size, byte count, card type,
timeout registers—whenever a data transfer command is not in progress. If the data
transfer expected bit (data_expected) in the cmd register is set to 1, the new command
is a data transfer command and the data path starts one of the following actions:
■ Transmits data if the read/write bit = 1
■ Receives data if read/write bit = 0

Data Transmit
The data transmit state machine, illustrated in Figure 11–8, starts data transmission
two clock cycles after a response for the data write command is received. This occurs
even if the command path detects a response error or response CRC error. If a
response is not received from the card because of a response timeout, data is not
transmitted. Depending upon the value of the transfer mode bit (transfer_mode) in
the cmd register, the data transmit state machine puts data on the card data bus in a
stream or in blocks.
Figure 11–8. Data Transmit State Machine
load_new_cmd,
data_expected, Write
Data & Block Transfer Data Tx Stop Data Command
Idle
load_new_command,
data_expected, Write
Data & Stream Transfer
Stop Data Command
Tx Tx
Data Block Data Stream
Byte Count Byte Count
Remaining != 0 Remaining = 0
Data Not Busy or Suspend/Stop
Data Command
Rx
Block Done CRC Status
Stream Data Transmit

If the transfer_mode bit in the cmd register is set to 1, the transfer is a stream-write
data transfer. The data path reads data from the FIFO buffer from the BIU and
transmits in a stream to the card data bus. If the FIFO buffer becomes empty, the
card clock is stopped and restarted once data is available in the FIFO buffer.
If the bytcnt register is reset to 0, the transfer is an open-ended stream-write data
transfer. During this data transfer, the data path continuously transmits data in a
stream until the host software issues an SD/SDIO STOP command. A stream data
transfer is terminated when the end bit of the STOP command and end bit of the
data match over two clock cycles.
If the bytcnt register is written with a nonzero value and the send_auto_stop bit
in the cmd register is set to 1, the STOP command is internally generated and
loaded in the command path when the end bit of the STOP command occurs after
the last byte of the stream write transfer matches. This data transfer can also
terminate if the host issues a STOP command before all the data bytes are
transferred to the card bus.
Single Block Data

If the transfer_mode bit in the cmd register is set to 0 and the bytcnt register value
is equal to the value of the block_size register, a single-block write-data transfer
occurs. The data transmit state machine sends data in a single block, where the
number of bytes equals the block size, including the internally-generated 16-term
CRC (CRC-16).

If the ctype register is set for a 1-bit, 4-bit, or 8-bit data transfer, the data is
transmitted on 1, 4, or 8 data lines, respectively, and CRC-16 is separately
generated and transmitted for 1, 4, or 8 data lines, respectively.
After a single data block is transmitted, the data transmit state machine receives
the CRC status from the card and signals a data transfer to the BIU. This happens
when the dto bit in the rintsts register is set to 1.
If a negative CRC status is received from the card, the data path signals a data
CRC error to the BIU by setting the dcrc bit in the rintsts register.
Additionally, if the start bit of the CRC status is not received by two clock cycles
after the end of the data block, a CRC status start-bit error (SBE) is signaled to the
BIU by setting the sbe bit in the rintsts register.
Multiple Block Data

A multiple-block write-data transfer occurs if the transfer_mode bit in the cmd
register is set to 0 and the value in the bytcnt register is not equal to the value of
the block_size register. The data transmit state machine sends data in blocks,
where the number of bytes in a block equals the block size, including the
internally-generated CRC-16.
If the ctype register is set to 1-bit, 4-bit, or 8-bit data transfer, the data is
transmitted on 1, 4, or 8 data lines, respectively, and CRC-16 is separately
generated and transmitted on 1, 4, or 8 data lines, respectively.
After one data block is transmitted, the data transmit state machine receives the
CRC status from the card. If the remaining byte count becomes 0, the data path
signals to the BIU that the data transfer is done. This happens when the dto bit in
the rintsts register is set to 1.
If the remaining data bytes are greater than zero, the data path state machine
starts to transmit another data block.
If a negative CRC status is received from the card, the data path signals a data
CRC error to the BIU by setting the dcrc bit in the rintsts register, and continues
further data transmission until all the bytes are transmitted.
If the CRC status start bit is not received by two clock cycles after the end of a data
block, a CRC status SBE is signaled to the BIU by setting the ebe bit in the rintsts
register and further data transfer is terminated.
If the send_auto_stop bit is set to 1 in the cmd register, the SD/SDIO STOP
command is internally generated during the transfer of the last data block, where
no extra bytes are transferred to the card. The end bit of the STOP command
might not exactly match the end bit of the CRC status in the last data block.
If the block size is less than 4, 16, or 32 for card data widths of 1 bit, 4 bits, or 8 bits,
respectively, the data transmit state machine terminates the data transfer when all
the data is transferred, at which time the internally-generated STOP command is
loaded in the command path.
If the bytcnt is zero (the block size must be greater than zero) the transfer is an
open-ended block transfer. The data transmit state machine for this type of data
transfer continues the block-write data transfer until the host software issues an
SD/SDIO STOP or STOP_TRANSMISSION (CMD12) command.

Data Receive
The data-receive state machine, illustrated in Figure 11–9, receives data two clock
cycles after the end bit of a data read command, even if the command path detects a
response error or response CRC error. If a response is not received from the card
because a response timeout occurs, the BIU does not receive a signal that the data
transfer is complete. This happens if the command sent by the controller is an illegal
operation for the card, which keeps the card from starting a read data transfer.
If data is not received before the data timeout, the data path signals a data timeout to
the BIU and an end to the data transfer done. Based on the value of the transfer_mode
bit in the cmd register, the data-receive state machine gets data from the card data bus
in a stream or block(s).
Figure 11–9. Data Receive State Machine
load_new_cmd,
data_expected, Read
Data & Block Transfer Data Rx Stop Data Command
Idle
load_new_command,
data_expected, Read
Data & Stream Transfer
Stop Data Command
Rx Rx
Data Block Data Stream
Byte Count Byte Count
Remaining != 0 Remaining = 0
or Stop Data Command
Read
Block Done Wait
Stream Data Read

A stream-read data transfer occurs if the transfer_mode bit in the cmd register is
set to 1, at which time the data path receives data from the card and writes it to
the FIFO buffer. If the FIFO buffer becomes full, the card clock stops and restarts
once the FIFO buffer is no longer full.
An open-ended stream-read data transfer occurs if the bytcnt register is set to 0.
During this type of data transfer, the data path continuously receives data in a
stream until the host software issues an SD/SDIO STOP command. A stream data
transfer terminates two clock cycles after the end bit of the STOP command.
If the bytcnt register contains a nonzero value and the send_auto_stop bit in the
cmd register is set to 1, a STOP command is internally generated and loaded into
the command path, where the end bit of the STOP command occurs after the last
byte of the stream data transfer is received. This data transfer can terminate if the
host issues an SD/SDIO STOP or STOP_TRANSMISSION (CMD12) command
before all the data bytes are received from the card.

Single-block Data Read

If the ctype register is set to a 1-bit, 4-bit, or 8-bit data transfer, data is received
from 1, 4, or 8 data lines, respectively, and CRC-16 is separately generated and
checked for 1, 4, or 8 data lines, respectively. If there is a CRC-16 mismatch, the
data path signals a data CRC error to the BIU. If the received end bit is not 1, the
BIU receives an End-bit Error (EBE).
Multiple-block Data Read

If the transfer_mode bit in the cmd register is set to 0 and the value of the bytcnt
register is not equal to the value of the block_size register, the transfer is a
multiple-block read-data transfer. The data-receive state machine receives data in
blocks, where the number of bytes in a block is equal to the block size, including
the internally-generated CRC-16.
If the ctype register is set to a 1-bit, 4-bit, or 8-bit data transfer, data is received
from 1, 4, or 8 data lines, respectively, and CRC-16 is separately generated and
checked for 1, 4, or 8 data lines, respectively. After a data block is received, if the
remaining byte count becomes zero, the data path signals a data transfer to the
BIU.
If the remaining data bytes are greater than zero, the data path state machine
causes another data block to be received. If CRC-16 of a received data block does
not match the internally-generated CRC-16, a data CRC error to the BIU and data
reception continue further data transmission until all bytes are transmitted.
Additionally, if the end of a received data block is not 1, data on the data path
signals terminate the bit error to the CIU and the data-receive state machine
terminates data reception, waits for data timeout, and signals to the BIU that the
data transfer is complete.
If the send_auto_stop bit in the cmd register is set to 1, the SD/SDIO STOP
command is internally generated when the last data block is transferred, where
no extra bytes are transferred from the card. The end bit of the STOP command
might not exactly match the end bit of the last data block.
If the requested block size for data transfers to cards is less than 4, 16, or 32 bytes
for 1-bit, 4-bit, or 8-bit data transfer modes, respectively, the data-transmit state
machine terminates the data transfer when all data is transferred, at which point
the internally-generated STOP command is loaded in the command path. Data
received from the card after that are then ignored by the data path.
If the bytcnt register is 0 (the block size must be greater than zero), the transfer is
an open-ended block transfer. For this type of data transfer, the data-receive state
machine continues the block-read data transfer until the host software issues an
SD/SDIO STOP or STOP_TRANSMISSION (CMD12) command.

Auto Stop
The controller internally generates an SD/SDIO STOP command and is loaded in the
command path when the send_auto_stop bit in the cmd register is set to 1. The
AUTO_STOP command helps to send an exact number of data bytes using a stream
read or write for the MMC, and a multiple-block read or write for SD memory transfer
for SD cards. The software must set the send_auto_stop bit according to details listed
in Table 11–10.

Table 11–10. Auto-Stop Generation

The following list describes conditions for the AUTO_STOP command:

■ Stream-read for MMC with byte count greater than zero—The controller generates
an internal STOP command and loads it into the command path so that the end bit
of the STOP command is sent when the last byte of data is read from the card and
no extra data byte is received. If the byte count is less than six (48 bits), a few extra
data bytes are received from the card before the end bit of the STOP command is
sent.
■ Stream-write for MMC with byte count greater than zero—The controller
generates an internal STOP command and loads it into the command path so that
the end bit of the STOP command is sent when the last byte of data is transmitted
on the card bus and no extra data byte is transmitted. If the byte count is less than
six (48 bits), the data path transmits the data last to meet these condition.
■ Multiple-block read memory for SD card with byte count greater than zero—If the
block size is less than four (single-bit data bus), 16 (4-bit data bus), or 32 (8-bit data
bus), the AUTO_STOP command is loaded in the command path after all the bytes
are read. Otherwise, the STOP command is loaded in the command path so that
the end bit of the STOP command is sent after the last data block is received.
■ Multiple-block write memory for SD card with byte count greater than zero—If
the block size is less than three (single-bit data bus), 12 (4-bit data bus), or 24 (8-bit
data bus), the AUTO_STOP command is loaded in the command path after all
data blocks are transmitted. Otherwise, the STOP command is loaded in the
command path so that the end bit of the STOP command is sent after the end bit of
the CRC status is received.
■ Precaution for host software during auto-stop—When an AUTO_STOP command
is issued, the host software must not issue a new command to the controller until
the AUTO_STOP command is sent by the controller and the data transfer is
complete. If the host issues a new command during a data transfer with the
AUTO_STOP command in progress, an AUTO_STOP command might be sent
after the new command is sent and its response is received. This can delay sending
the STOP command, which transfers extra data bytes. For a stream write, extra
data bytes are erroneous data that can corrupt the card data. If the host wants to
terminate the data transfer before the data transfer is complete, it can issue an
SD/SDIO STOP or STOP_TRANSMISSION (CMD12) command, in which case the
controller does not generate an AUTO_STOP command.

Non-Data Transfer Commands that Use Data Path

Some SD/SDIO non-data transfer commands (commands other than read and write
commands) also use the data path. Table 11–11 lists the commands and their register
setup requirements.
Table 11–11. Non-Data Transfer Commands and Requirements
SEND_NUM_WR_BLOCKS
SEND_WRITE_PROT
PROGRAM_CSD
LOCK_UNLOCK
SD_STATUS
SEND_SCR
(ACMD13)
(ACMD22)
(ACMD51)
(CMD27)
(CMD30)
(CMD42)
cmd Register Setup
Cmd_index 0x1B=27 0x1E=30 0x2A=42 0x0D=13 0x16=22 0x33=51
Response_expect 1 1 1 1 1 1
Response_length 0 0 0 0 0 0
Check_response_crc 1 1 1 1 1 1
Data_expected 1 1 1 1 1 1
Read/write 1 0 1 0 0 0
Transfer_mode 0 0 0 0 0 0
Send_auto_stop 0 0 0 0 0 0
Wait_prevdata_complete 0 0 0 0 0 0
Stop_abort_cmd 0 0 0 0 0 0
cmdarg Register Setup

32-bit
write
Stuff
Stuff bits protect Stuff bits Stuff bits Stuff bits
bits
data
address
blksiz Register Setup
16 4 Num_bytes (1) 64 4 8
bytcnt Register Setup
16 4 Num_bytes (1) 64 4 8
(1) Num_bytes = Number of bytes specified as per the lock card data structure. Refer to the SD specification and the
MMC specification.

Table 11–12.
Clock Control Block

The clock control block provides different clock frequencies required for
SD/MMC/CE-ATA cards. The clock control block has one clock divider, which is
used to generate different card clock frequencies.
The clock frequency of a card depends on the following clock ctrl register settings:
■ clkdiv register—Internal clock dividers are used to generate different clock
frequencies required for the cards. The division factor for the clock divider can be
set by writing to the clkdiv register. The clock divider is an 8-bit value that
provides a clock division factor from 1 to 510; a value of 0 represents a
clock-divider bypass, a value of 1 represents a divide by 2, a value of 2 represents a
divide by 4, and so on.
■ clksrc register—Set this register to 0 as clock is divided by clock divider 0.

■ clkena register—The cclk_out card output clock can be enabled or disabled under
the following conditions:
■ cclk_out is enabled when the cclk_enable bit in the clkena register is set to 1
and disabled when set to 0.
■ Low-power mode can be enabled by setting the cclk_low_power bit of the
clkena register to 1. If low-power mode is enabled to save card power, the
cclk_out signal is disabled when the card is idle for at least eight card clock
cycles. Low-power mode is enabled when a new command is loaded and the
command path goes to a non-idle state.
Under the following conditions, the card clock is stopped or disabled:
■ Clock can be disabled by writing to the clkena register.
■ When low-power mode is selected and the card is idle for at least eight clock
cycles.
■ FIFO buffer is full, data path cannot accept more data from the card, and data
transfer is incomplete—to avoid FIFO buffer overflow.
■ FIFO buffer is empty, data path cannot transmit more data to the card, and data
transfer is incomplete—to avoid FIFO buffer underflow.
1 The card clock must be disabled through the clkena register before the host software
changes the values of the clkdiv and clksrc registers.
Error Detection
Errors can occur during card operations within the CIU in the following situations.
Response
■ Response timeout—did not receive the response expected with response start bit
within the specified number of clock cycles in the timeout register.
■ Response CRC error—response is expected and check response CRC requested;
response CRC-7 does not match with the internally-generated CRC-7.
■ Response error—response transmission bit is not 0, command index does not
match with the command index of the send command, or response end bit is not 1.
Data Transmit
■ No CRC status—during a write data transfer, if the CRC status start bit is not
received for two clock cycles after the end bit of the data block is sent out, the data
path performs the following actions:
■ Signals no CRC status error to the BIU
■ Terminates further data transfer
■ Signals data transfer done to the BIU

■ Negative CRC—if the CRC status received after the write data block is negative
(that is, not 0b010), the data path signals a data CRC error to the BIU and continues
with the data transfer.
■ Data starvation due to empty FIFO buffer—if the FIFO buffer becomes empty
during a write data transmission, or if the card clock stopped and the FIFO buffer
remains empty for a data-timeout number of clock cycles, the data path signals a
data-starvation error to the BIU and the data path continues to wait for data in the
FIFO buffer.
Data Receive
■ Data timeout—during a read-data transfer, if the data start bit is not received
before the number of clock cycles specified in the timeout register, the data path
does the following action:
■ Signals a data-timeout error to the BIU
■ Terminates further data transfer
■ Signals data transfer done to BIU
■ Data SBE—during a 4-bit or 8-bit read-data transfer, if the all-bit data line does not
have a start bit, the data path signals a data SBE to the BIU and waits for a data
timeout, after which it signals that the data transfer is done.
■ Data CRC error—during a read-data-block transfer, if the CRC-16 received does
not match with the internally generated CRC-16, the data path signals a data CRC
error to the BIU and continues with the data transfer.
■ Data EBE—during a read-data transfer, if the end bit of the received data is not 1,
the data path signals an EBE to the BIU, terminates further data transfer, and
signals to the BIU that the data transfer is done.
■ Data starvation due to FIFO buffer full—during a read data transmission and
when the FIFO buffer becomes full, the card clock stops. If the FIFO buffer remains
full for a data-timeout number of clock cycles, the data path signals a data
starvation error to the BIU, by setting the data starvation host timeout bit (hto) in
rintsts register to 1, and the data path continues to wait for the FIFO buffer to
empty.
Clocks
The SD/MMC controller clocks are listed in Table 11–13.
Table 11–13. SD/MMC Controller Clocks (Part 1 of 2)

Clock Name Direction Description
sdmmc_clk In Clock for SD/MMC controller CIU
l4_mp_clk In Clock for SD/MMC controller BIU
sdmmc_cclk_out Out Generated output clock for card
Phase-shifted clock of sdmmc_clk used to sample the command and data from
sdmmc_sample_clk Internal
the card

Table 11–13. SD/MMC Controller Clocks (Part 2 of 2)

Clock Name Direction Description
Phase-shifted clock of sdmmc_clk for controller to drive command and data to
sdmmc_drv_clk Internal
the card to meet hold time requirements
sdmmc_clk_divided Internal Divide-by-four clock of sdmmc_clk
Figure 11–10 shows the connection of various clocks to the controller.
Figure 11–10. SD/MMC Controller Clock Connections
l4_mp_clk sdmmc_cclk_out
sdmmc_clk Divide sdmmc_clk_divided

by 4
SD/MMC
Controller
Phase sdmmc_drv_clk Core
Shifter
sdmmc_sample_clk
Phase
Shifter
sdmmc_fb_clk_in
The sdmmc_clk clock from the clock manager is divided by four (sdmmc_clk_divided
clock) before passed to the phase shifters and controller. The phase shifters are used to
generate the sdmmc_drv_clk and sdmmc_sample_clk clocks. These phase shifters
provide up to eight phases shift which include 0, 45, 90, 135, 180, 225, 270, and 315
degrees. The sdmmc_sample_clk clock can be driven by the output from the phase
shifter. The selections of phase shift degree and sdmmc_sample_clk source are done in
the system manager. For information about setting the phase shift and selecting the
source of the sdmmc_sample_clk clock, refer to “Clock Setup” on page 11–40.
The controller generates the sdmmc_cclk_out clock, which is driven to the card. For
more information about the generation of the sdmmc_cclk_out clock, refer to “Clock
Control Block” on page 11–31.
Resets
The SD/MMC controller has one reset signal. The reset manager drives this signal to
the SD/MMC controller on a cold or warm reset.
Device Handbook.

SD/MMC Controller Programming Model
Interface Signals
Table 11–14 lists I/O pin use of the SD/MMC controller interface signals.
Table 11–14. SD/MMC Controller Interface I/O Pins

sdmmc_cclk_out 1 Out Clock from controller to the card
sdmmc_cmd 1 In/Out Card command
sdmmc_pwren 1 Out External device power enable
sdmmc_data 8 In/Out Card data
Initialization
This section describes how to initialize the SD/MMC controller.
Figure 11–11 shows the initialization flow of the SD/MMC controller. After the power
and clock to the controller are stable, the controller active-low reset is asserted. The
reset sequence initializes the registers, FIFO buffer pointers, DMA interface controls,
and state machines in the controller.
Figure 11–11. SD/MMC Controller Initialization Sequence
Assert Active-Low Reset
Enable Power to Card
Set Interrupt Masks
Enumerate Card Stack
Set the Clock Source Assignments
Set Other Controller Registers

Software must perform the following steps after the power-on-reset:

1. Before enabling power to the card, confirm that the voltage setting to the voltage
regulator is correct.
2. Enable power to the card by setting the power enable bit (power_enable) in the
power enable register (pwren) to 1. Wait for the power ramp-up time before
proceeding to the next step.
3. Set the interrupt masks by resetting the appropriate bits to 0 in the intmask
register.
4. Set the int_enable bit of the ctrl register to 1.
1 Altera recommends that you write 0xFFFFFFFF to the rintsts register to

clear any pending interrupts before setting the int_enable bit to 1.
5. Discover the card stack according to the card type. For discovery, you must restrict
the clock frequency to 400 kHz in accordance with SD/MMC/CE-ATA standards.
For more information, refer to “Enumerated Card Stack” on page 11–37.
6. Set the clock source assignments. Set the card frequency using the clkdiv and
clksrc registers of the controller. For more information, refer to “Clock Setup” on
page 11–40.
7. The following common registers and fields can be set during initialization process:
■ The response timeout field (response_timeout) of the tmout register. A typical
value is 0x64.
■ The data timeout field (data_timeout) of the tmout register, highest of the
following:
■ 10 * NAC
= 10 * ((TAAC * FOP) + (100 * NSAC))
where:
NAC = card device total access time
TAAC = Time-dependent factor of the data access time

FOP = The card clock frequency used for the card operation
NSAC = Worst-case clock rate-dependent factor of the data access time
■ Host FIFO buffer latency
On read: Time elapsed before host starts reading from a full FIFO buffer
On write: Time elapsed before host starts writing to an empty FIFO buffer
■ Debounce counter register (debnce). A typical debounce value is 25 ms.
■ TX watermark field (tx_wmark) of the FIFO threshold watermark register
(fifoth). Typically, the threshold value is set to 512, which is half the FIFO
buffer depth.
■ RX watermark field (rx_wmark) of the fifoth register. Typically, the threshold
value is set to 511.
These registers do not need to be changed with every SD/MMC/CE-ATA
command. Set them to a typical value according to the SD/MMC/CE-ATA
specifications.
Enumerated Card Stack

The card stack performs the following tasks:
■ Discovers the connected card
■ Sets the relative Card Address Register (RCA) in the connected card
■ Reads the card specific information
■ Stores the card specific information locally
The card connected to the controller can be an MMC, CE-ATA, SD or SDIO (including
IO ONLY, MEM ONLY and COMBO) card. To identify the connected card type, the
following discovery sequence is needed:
1. Reset the card width 1 or 4 bit (card_width2) and card width 8 bit (card_width1)
fields in the ctype register to 0.
2. Identify the card type as SD, MMC, SDIO or SDIO-COMBO:
a. Send an SD/SDIO IO_SEND_OP_COND (CMD5) command with argument 0
to the card.
b. Read resp0 on the controller. The response to the IO_SEND_OP_COND
command gives the voltage that the card supports.
c. Send the IO_SEND_OP_COND command, with the desired voltage window in
the arguments. This command sets the voltage window and makes the card
exit the initialization state.
d. Check bit 27 in resp0.
■ If bit 27 is 0, the SDIO card is IO ONLY. In this case, proceed to step 5.
■ If bit 27 is 1, the card type is SDIO COMBO. Continue with the following
steps.

3. Only continue with this step if the SDIO card type is COMBO or there is no
response received from the previous IO_SEND_OP_COND command. Otherwise,
skip to step 5.
a. Send the SD/SDIO SEND_IF_COND (CMD8) command with the following
arguments:
■ Bit[31:12] = 0x0 (reserved bits)
■ Bit[11:8] = 0x1 (supply voltage value)
■ Bit[7:0] = 0xAA (preferred check pattern by SD memory cards compliant
with SDIO Simplified Specification Version 2.00 and later.)
f Refer to SDIO Simplified Specification Version 2.00 as described in

“References” on page 11–79:
b. If a response is received to the previous SEND_IF_COND command, the card

supports SD High-Capacity, compliant with SD Specifications, Part 1, Physical
Layer Simplified Specification Version 2.00.
If no response is received, proceed to step e.
c. Send the SD_SEND_OP_COND (ACMD41) command with the following
arguments:
■ Bit[31] = 0x0 (reserved bits)
■ Bit[30] = 0x1 (high capacity status)
■ Bit[24] = 0x1 (S18R --supports voltage switching for 1.8V)
■ Bit[23:0] = supported voltage range
d. If a response is received to the previous SD_SEND_OP_COND command, the
card type is SDHC. Otherwise, the card is MMC or CE-ATA. In either case, skip
the following steps and proceed to step 5.
e. If a response is not received to the initial SEND_IF_COND command, the card
does not support High Capacity SD2.0. Next, issue the GO_IDLE_STATE
command followed by the SD_SEND_OP_COND command with the
following arguments:
■ Bit[31] = 0x0 (reserved bits)
■ Bit[30] = 0x0 (high capacity status)
■ Bit[23:0] = supported voltage range
f. If a response is received to the previous SD_SEND_OP_COND command, the
card is SD type. Otherwise, the card is MMC or CE-ATA.

1 You must issue the SEND_IF_COND command prior to the first

SD_SEND_OP_COND command, to initialize the High Capacity SD
memory card. The card returns busy as a response to the
SD_SEND_OP_COND command when any of the following conditions are
true:
■ The card executes its internal initialization process.

■ A SEND_IF_COND command is not issued before the
SD_SEND_OP_COND command.
■ The ACMD41 command is issued. In the command argument, the Host
Capacity Support (HCS) bit is set to 0, for a high capacity SD card.
4. Use the following sequence to determine whether the card is a CE-ATA 1.1,
CE-ATA 1.0, or MMC device:
a. Determine whether the card is a CE-ATA v1.1 card device by attempting to
select ATA mode. Send the SD/SDIO SEND_IF_COND command, querying
byte 504 (S_CMD_SET) of the EXT_CSD register block in the external card.
■ If bit 4 is set to 1, the card device supports ATA mode. Send the
SWITCH_FUNC (CMD6) command, setting the ATA bit (bit 4) of the
EXT_CSD register slice 191 (CMD_SET) to 1. This command selects ATA
mode and activates the ATA command set.
You can verify the currently selected mode by reading it back from byte 191
of the EXT_CSD register.
Skip to step 5.
■ If the card device does not support ATA mode, it might be an MMC card or
a CE-ATA v1.0 card. Continue to Step b.
b. Determine whether the card is a CE-ATA 1.0 card device or an MMC card
device by sending the RW_REG command. If a response is received and the
response data contains the CE-ATA signature, the card is a CE-ATA 1.0 card
device. Otherwise, the card is an MMC card device.
5. At this point, the software has determined the card type as SD/SDHC, SDIO or
SDIO-COMBO. Now it must enumerate the card stack according to the type that
has been discovered.

6. Set the card clock source frequency to the frequency of identification clock rate,
400 KHz. Use one of the following discovery command sequences:
■ For an SD card or an SDIO memory section, send the following SD/SDIO
command sequence:
■ GO_IDLE_STATE
■ SEND_IF_COND
■ SD_SEND_OP_COND (ACMD41)
■ ALL_SEND_CID (CMD2)
■ SEND_RELATIVE_ADDR (CMD3)
■ For an SDIO card, send the following command sequence:
■ IO_SEND_OP_COND.
■ If the function count is valid, send the SEND_RELATIVE_ADDR command.
■ For an MMC, send the following command sequence:
■ GO_IDLE_STATE
■ SEND_OP_COND (CMD1)
■ ALL_SEND_CID
■ SEND_RELATIVE_ADDR
7. You can change the card clock frequency after discovery by writing a value to the
clkdiv register that divides down the sdmmc_clk clock.
The following list shows typical clock frequencies for various types of cards:
■ SD memory card, 25 MHz
■ MMC card device, 12.5 MHz
■ Full speed SDIO, 25 MHz
■ Low speed SDIO, 400 kHz
Clock Setup
The following registers of the SD/MMC controller allow software to select the desired
clock frequency for the card:
■ clksrc
■ clkdiv
■ clkena
The controller loads these registers when it receives an update clocks command, as
described in this section. To change the card clock frequency, perform the following
steps:
1. Before disabling the clocks, ensure that the card is not busy with any previous data
command. To do so, verify that the data_busy bit of the status register (status) is
0.
2. Reset the cclk_enable bit of the clkena register to 0, to disable the card clock
generation.

3. Reset the clksrc register to 0.

4. Set the following bits in the cmd register to 1:
■ update_clk_regs_only—Specifies the update clocks command
■ wait_prvdata_complete—Ensures that clock parameters do not change until
any ongoing data transfer is complete
■ start_cmd—Initiates the command
5. Wait until the start_cmd and update_clk_regs_only bits change to 0. There is no
interrupt when the clock modification completes. The controller does not set the
command_done bit in the rintsts register upon command completion. The
controller might signal a hardware lock error if it already has another command in
the queue. In this case, return to Step 4.
For information about hardware lock errors, refer to “Interrupt and Error
Handling” on page 11–66.
6. Reset the sdmmc_clk_enable bit to 0 in the enable register of the clock manager
peripheral PLL group (perpllgrp).
7. In the control register (ctrl) of the SDMMC controller group (sdmmcgrp) in the
system manager, set the drive clock phase shift select (drvsel) and sample clock
phase shift select (smplsel) bits to specify the required phase shift value.
8. Set the sdmmc_clk_enable bit in the Enable register of the clock manager
perpllgrp group to 1.
9. Set the clkdiv register of the controller to the correct divider value for the required
clock frequency.
10. Set the cclk_enable bit of the clkena register to 1, to enable the card clock
generation.
You can also use the clkena register to enable low-power mode, which
automatically stops the sdmmc_cclk_out clock when the card is idle for more than
eight clock cycles.
Controller/DMA/FIFO Buffer Reset Usage

The following list shows the effect of reset on various parts in the SD/MMC
controller:
■ Controller reset—resets the controller by setting the controller_reset bit in the
ctrl register to 1. Controller reset resets the CIU and state machines, and also
resets the BIU-to-CIU interface. Because this reset bit is self-clearing, after issuing
the reset, wait until this bit changes to 0.
■ FIFO buffer reset—resets the FIFO buffer by setting the FIFO reset bit (fifo_reset)
in the ctrl register to 1. FIFO buffer reset resets the FIFO buffer pointers and
counters in the FIFO buffer. Because this reset bit is self-clearing, after issuing the
reset, wait until this bit changes to 0.

■ DMA reset—resets the internal DMA controller logic by setting the DMA reset bit
(dma_reset) in the ctrl register to 1, which immediately terminates any DMA
transfer in progress. Because this reset bit is self-clearing, after issuing the reset,
wait until this bit changes to 0.
1 Ensure that the DMA is idle before performing a DMA reset. Otherwise, the
L3 interconnect might be left in an indeterminate state.
Altera recommends setting the controller_reset, fifo_reset, and dma_reset bits in

the ctrl register to 1 first, and then resetting the rintsts register to 0 using another
write, to clear any resultant interrupt.
Enabling FIFO Buffer ECC

To protect the FIFO buffer data with ECC, you must enable the ECC feature before
performing any operations with the SD/MMC controller. Perform the following steps
to enable the FIFO buffer ECC feature:
1. Verify there are no commands committed to the controller.
2. Ensure that the FIFO buffer is initialized. Initialize the FIFO buffer by writing 0 to
all 1024 FIFO buffer locations. A FIFO buffer write to any address from 0x200 to
the maximum FIFO buffer size is valid.
3. Set the SDMMC RAM ECC single and double, correctable error interrupt status
bits (serrporta, derrporta, serrportb, and derrportb) to 1 in the sdmmc register in
the eccgrp group of the system manager, to clear any previously-detected ECC
errors.
4. Reset the FIFO buffer by setting the fifo_reset bit to 1 in the ctrl register. This
action resets pointers and counters in the FIFO buffer. This reset bit is self-clearing,
so after issuing the reset, wait until the bit changes to 0.
5. Set the en bit in sdmmc register in eccgrp group of the system manager to 1, to
enable ECC for the FIFO buffer in SD/MMC controller.
Non-Data Transfer Commands

To send any non-data transfer command, the software needs to write the cmd register
and the cmdarg register with appropriate parameters. Using these two registers, the
controller forms the command and sends it to the CMD pin. The controller reports
errors in the command response through the error bits of the rintsts register.
When a response is received—either erroneous or valid—the controller sets the
command_done bit in the rintsts register to 1. A short response is copied to resp0,
while a long response is copied to all four response registers (resp0, resp1, resp2, and
resp3). For long responses, bit 31 of resp3 represents the MSB and bit 0 of resp0
represents the LSB.
For basic and non-data transfer commands, perform the following steps:
1. Write the cmdarg register with the appropriate command argument parameter.
2. Write the cmd register with the settings in Table 11–15 on page 11–44.

3. Wait for the controller to accept the command. The start_cmd bit changes to 0
when the command is accepted.
The following actions occur when the command is loaded into the controller:
■ If no previous command is being processed, the controller accepts the
command for execution and resets the start_cmd bit in the cmd register to 0. If a
previous command is being processed, the controller loads the new command
in the command buffer.
■ If the controller is unable to load the new command—that is, a command is
already in progress, a second command is in the buffer, and a third command is
attempted—the controller generates a hardware lock error.
4. Check if there is a hardware lock error.
5. Wait for command execution to complete. After receiving either a response from a
card or response timeout, the controller sets the command_done bit in the rintsts
register to 1. Software can either poll for this bit or respond to a generated
interrupt (if enabled).
6. Check if the response timeout boot acknowledge received (bar), rcrc, or re bit is
set to 1. Software can either respond to an interrupt raised by these errors or poll
the re, rcrc, and bar bits of the rintsts register. If no response error is received,
the response is valid. If required, software can copy the response from the
response registers.
1 Software cannot modify clock parameters while a command is being executed.

Table 11–15. cmd Register Settings for Non-Data Transfer Command

Parameter Value Comment
Default
start_cmd 1 This bit resets itself to 0 after the command is committed.
use_hold_reg 1 or 0 Choose the value based on the speed mode being used.
update_clk_regs_only 0 Indicates that the command is not a clock update command
data_expected 0 Indicates that the command is not a data command
card_number 1 For one card
Command Set this parameter to the command number. For example, set to 8 for the
cmd_index
Index SD/SDIO SEND_IF_COND (CMD8) command.
1 for card reset commands such as the SD/SDIO GO_IDLE_STATE
send_initialization 0 or 1 command
0 otherwise
1 for a command to stop data transfer, such as the SD/SDIO
stop_abort_cmd 0 or 1 STOP_TRANSMISSION command
0 otherwise
1 for R2 (long) response
response_length 0 or 1
0 for short response
0 for commands with no response, such as SD/SDIO GO_IDLE_STATE,
response_expect 0 or 1 SET_DSR (CMD4), or GO_INACTIVE_STATE (CMD15).
1 otherwise
User-Selectable
Before sending a command on the command line, the host must wait for
completion of any data command already in process. Altera recommends
wait_prvdata_complete 1
that you set this bit to 1, unless the current command is to query status or
stop data transfer when transfer is in progress.
1 if the response includes a valid CRC, and the software is required to
check_response_crc 1 or 0 crosscheck the response CRC bits.
0 otherwise
Data Transfer Commands

Data transfer commands transfer data between the memory card and the controller.
To issue a data command, the controller requires a command argument, total data
size, and block size. Data transferred to or from the memory card is buffered by the
controller FIFO buffer.
Before issuing a data transfer command, software must confirm that the card is not
busy and is in a transfer state, by performing the following steps:
1. Issue an SD/SDIO SEND_STATUS (CMD13) command. The controller sends the
status of the card as the response to the command.
2. Check the card’s busy status.

3. Wait until the card is not busy.

4. Check the card’s transfer status. If the card is in the stand-by state, issue an
SD/SDIO SELECT/DESELECT_CARD (CMD7) command to place it in the
transfer state.
1 During CE-ATA RW_BLK write transfers, the MMC busy signal might be asserted
after the last block. If the CE-ATA card device interrupt is disabled (the nIEN bit in the
card device’s ATA control register is set to 1), the dto bit in the rintsts register is set
to 1 even though the card sends MMC BUSY. The host cannot issue the CMD60
command to check the ATA busy status after a CMD61 command. Instead, the host
must perform one of the following actions:
■ Issue the SEND_STATUS command and check the MMC busy status before
issuing a new CMD60 command
■ Issue the CMD39 command and check the ATA busy status before issuing a new
CMD60 command
For the data transfer commands, software must set the ctype register to the bus width
that is programmed in the card.
The controller generates an interrupt for different conditions during data transfer,
which are reflected in the following rintsts register bits:
1. dto—Data transfer is over or terminated. If there is a response timeout error, the
controller does not attempt any data transfer and the Data Transfer Over bit is
never set.
2. Transmit FIFO data request bit (txdr)—The FIFO buffer threshold for transmitting
data is reached; software is expected to write data, if available, into the FIFO
buffer.
3. Receive FIFO data request bit (rxdr)—The FIFO buffer threshold for receiving data
is reached; software is expected to read data from the FIFO buffer.
4. hto—The FIFO buffer is empty during transmission or is full during reception.
Unless software corrects this condition by writing data for empty condition, or
reading data for full condition, the controller cannot continue with data transfer.
The clock to the card is stopped.
5. bds—The card has not sent data within the timeout period.
6. dcrc—A CRC error occurred during data reception.
7. sbe—The start bit is not received during data reception.
8. ebe—The end bit is not received during data reception, or for a write operation. A
CRC error is indicated by the card.
Conditions 6, 7, and 8 indicate that the received data might have errors. If there is a
response timeout, no data transfer occurs.

Single-Block or Multiple-Block Read

To implement a single-block or multiple-block read, the software performs the
following steps:
1. Write the data size in bytes to the bytcnt register. For a multi-block read, bytcnt
must be a multiple of the block size.
2. Write the block size in bytes to the blksiz register. The controller expects data to
return from the card in blocks of size blksiz.
3. If the read round trip delay, including the card delay, is greater than half of
sdmmc_clk_divided, write to the card threshold control register (cardthrctl) to
ensure that the card clock does not stop in the middle of a block of data being
transferred from the card to the host. For more information, refer to “Card Read
Threshold” on page 11–64.
1 If the card read threshold enable bit (cardrdthren) is 0, the host system
must ensure that the RX FIFO buffer does not become full during a read
data transfer by ensuring that the RX FIFO buffer is read at a rate faster than
that at which data is written into the FIFO buffer. Otherwise, an overflow
might occur.
4. Write the cmdarg register with the beginning data address for the data read.
5. Write the cmd register with the parameters listed in Table 11–16. For SD and MMC
cards, use the SD/SDIO READ_SINGLE_BLOCK (CMD17) command for a
single-block read and the READ_MULTIPLE_BLOCK (CMD18) command for a
multiple-block read. For SDIO cards, use the IO_RW_EXTENDED (CMD53)
command for both single-block and multiple-block transfers. After writing to the
cmd register, the controller starts executing the command. When the command is
sent to the bus, the Command Done interrupt is generated.
6. Software must check for data error interrupts, reported in the dcrc, bds, sbe, and
ebe bits of the rintsts register. If required, software can terminate the data
transfer by sending an SD/SDIO STOP command.

7. Software must check for host timeout conditions in the rintsts register:
■ Receive FIFO buffer data request
■ Data starvation from host—the host is not reading from the FIFO buffer fast
enough to keep up with data from the card. To correct this condition, software
must perform the following steps:
■ Read the fifo_count field of the status register
■ Read the corresponding amount of data out of the FIFO buffer
In both cases, the software must read data from the FIFO buffer and make space in
the FIFO buffer for receiving more data.
8. When a DTO interrupt is received, the software must read the remaining data from
the FIFO buffer.
Table 11–16. cmd Register Settings for Single-Block and Multiple-Block Reads
Default
start_cmd 1 This bit resets itself to 0 after the command is committed.
use_hold_reg 1 or 0 Choose the value based on speed mode being used.
update_clk_regs_only 0 Does not need to update clock parameters
data_expected 1 Data command
transfer_mode 0 Block transfer
1 for a card reset command such as the SD/SDIO GO_IDLE_STATE
send_initialization 0 command
0 otherwise
1 for a command to stop data transfer such as the SD/SDIO
stop_abort_cmd 0 STOP_TRANSMISSION command
0 otherwise
send_auto_stop 0 or 1 Set according to Table 11–10 on page 11–28
read_write 0 Read from card
1 for R2 (long) response
response_length 0
0 for short response
0 for commands with no response, such as SD/SDIO GO_IDLE_STATE,
response_expect 1 or 0 SET_DSR, and GO_INACTIVE_STATE.
1 otherwise
User-Selectable
0—sends command to CIU immediately
wait_prvdata_complete 1 or 0
1—sends command after previous data transfer ends
0—Controller must not check response CRC
check_response_crc 1 or 0
1—Controller must check response CRC
Set this parameter to the command number. For example, set to 17 or 18
Command
cmd_index for SD/SDIO READ_SINGLE_BLOCK (CMS17) or
Index
READ_MULTIPLE_BLOCK (CMD18).

Single-Block or Multiple-Block Write

The following steps comprise a single-block or multiple-block write:
1. Write the data size in bytes to the bytcnt register. For a multi-block write, bytcnt
must be a multiple of the block size.
2. Write the block size in bytes to the blksiz register. The controller sends data in
blocks of size blksiz each.
3. Write the cmdarg register with the data address to which data must be written.
4. Write data into the FIFO buffer. For best performance, the host software should
write data continuously until the FIFO buffer is full.
5. Write the cmd register with the parameters listed in Table 11–17 on page 11–49. For
SD and MMC cards, use the SD/SDIO WRITE_BLOCK (CMD24) command for a
single-block write and the WRITE_MULTIPLE_BLOCK (CMD25) command for a
multiple-block writes. For SDIO cards, use the IO_RW_EXTENDED command for
both single-block and multiple-block transfers.
After writing to the cmd register, the controller starts executing a command if there
is no other command already being processed. When the command is sent to the
bus, a Command Done interrupt is generated.
6. Software must check for data error interrupts; that is, for dcrc, bds, and ebe bits of
the rintsts register. If required, software can terminate the data transfer early by
sending the SD/SDIO STOP command.

7. Software must check for host timeout conditions in the rintsts register:
■ Transmit FIFO buffer data request
■ Data starvation by the host—the controller wrote data to the card faster than
the host could supply the data.
In both cases, the software must write data into the FIFO buffer.
There are two types of transfers: open-ended and fixed length.
■ Open-ended transfers—For an open-ended block transfer, the byte count is 0.
At the end of the data transfer, software must send the STOP_TRANSMISSION
command (CMD12).
■ Fixed-length transfers—The byte count is nonzero. You must already have
written the number of bytes to the bytcnt register. The controller issues the
STOP command for you if you set the send_auto_stop bit of the cmd register to
1. After completion of a transfer of a given number of bytes, the controller
sends the STOP command. Completion of the AUTO_STOP command is
reflected by the Auto Command Done interrupt. A response to the
AUTO_STOP command is written to the resp1 register.
If software does not set the send_auto_stop bit in the cmd register to 1, software
must issue the STOP command just like in the open-ended case.
When the dto bit of the rintsts register is set, the data command is complete.
Table 11–17. cmd Register Settings for Single-Block and Multiple-Block Write (Part 1 of 2)
Default
This bit resets itself to 0 after the command is committed (accepted by the
start_cmd 1
BIU).
use_hold_reg 1 or 0 Choose the value based on speed mode being used.
update_clk_regs_only 0 Does not need to update clock parameters
data_expected 1 Data command
transfer_mode 0 Block transfer
Can be 1, but only for card reset commands such as SD/SDIO
send_initialization 0
GO_IDLE_STATE
Can be 1 for commands to stop data transfer such as SD/SDIO
stop_abort_cmd 0
STOP_TRANSMISSION
send_auto_stop 0 or 1 Set according to Table 11–10 on page 11–28
read_write 1 Write to card
response_length 0 Can be 1 for R2 (long) response
Can be 0 for commands with no response. For example, SD/SDIO
response_expect 1
GO_IDLE_STATE, SET_DSR, GO_INACTIVE_STATE etc.

Table 11–17. cmd Register Settings for Single-Block and Multiple-Block Write (Part 2 of 2)
Default
User-Selectable
0—Sends command to the CIU immediately
1—Sends command after previous data transfer ends
0—Controller must not check response CRC
check_response_crc 1
1—Controller must check response CRC
Set this parameter to the command number. For example, set to 24 for
Command
cmd_index SD/SDIO WRITE_BLOCK (CMD24) or 25 for WRITE_MULTIPLE_BLOCK
Index
(CMD25).
Stream Read and Write

In a stream transfer, if the byte count is equal to 0, the software must also send the
SD/SDIO STOP command. If the byte count is not 0, when a given number of bytes
completes a transfer, the controller sends the STOP command automatically.
Completion of this AUTO_STOP command is reflected by the Auto_command_done
interrupt. A response to an AUTO_STOP command is written to the resp1 register. A
stream transfer is allowed only for card interfaces with a 1-bit data bus.
A stream read requires the same steps as the block read described in“Single-Block or
Multiple-Block Read” on page 11–46, except for the following bits in the cmd register:
■ transfer_mode = 0x1 (for stream transfer)
■ cmd_index = 20 (SD/SDIO CMD20)
A stream write requires the same steps as the block write mentioned in “Single-Block
or Multiple-Block Write” on page 11–48, except for the following bits in the cmd
register:
■ transfer_mode = 0x1 (for stream transfer)
■ cmd_index = 11 (SD/SDIO CMD11)
Packed Commands
To reduce overhead, read and write commands can be packed in groups of
commands—either all read or all write—that transfer the data for all commands in the
group in one transfer on the bus. Use the SD/SDIO SET_BLOCK_COUNT (CMD23)
command to state ahead of time how many blocks will be transferred. Then issue a
single READ_MULTIPLE_BLOCK or WRITE_MULTIPLE_BLOCK command to read
or write multiple blocks.
■ SET_BLOCK_COUNT—set block count (number of blocks transferred using the
READ_MULTIPLE_BLOCK or WRITE_MULTIPLE_BLOCK command)
■ READ_MULTIPLE_BLOCK—multiple-block read command
■ WRITE_MULTIPLE_BLOCK—multiple-block write command
Packed commands are organized in packets by the application software and are
transparent to the controller.

f For more information about packed commands, refer to JEDEC Standard No. 84-A441,
as referenced in “References” on page 11–79.
Transfer Stop and Abort Commands

This section describes stop and abort commands. The SD/SDIO
STOP_TRANSMISSION command can terminate a data transfer between a memory
card and the controller. The ABORT command can terminate an I/O data transfer for
only an SDIO card.
STOP_TRANSMISSION (CMD12)
The host can send the STOP_TRANSMISSION (CMD12) command on the CMD pin at
any time while a data transfer is in progress. Perform the following steps to send the
STOP_TRANSMISSION command to the SD/SDIO card device:
1. Set the wait_prvdata_complete bit of the cmd register to 0.
2. Set the stop_abort_cmd in the cmd register to 1, which ensures that the CIU stops.
The STOP_TRANSMISSION command is a non-data transfer command. For more
information, refer to “Non-Data Transfer Commands” on page 11–42.
ABORT
The ABORT command can only be used with SDIO cards. To abort the function that is
transferring data, program the ABORT function number in the ASx[2:0] bits at
address 0x06 of the card common control register (CCCR) in the card device, using the
IO_RW_DIRECT (CMD52) command. The CCCR is located at the base of the card
space 0x0 – 0xFF.
The ABORT command is a non-data transfer command. For more information, refer to
“Non-Data Transfer Commands” on page 11–42.
Perform the following steps to send the ABORT command to the SDIO card device:
1. Set the cmdarg register to include the appropriate command argument parameters
listed in Table 11–18.
2. Send the IO_RW_DIRECT command by setting the following fields of the cmd
register:
■ Set the command index to 0x52 (IO_RW_DIRECT).
■ Set the stop_abort_cmd bit of the cmd register to 1 to inform the controller that
the host aborted the data transfer.
■ Set the wait_prvdata_complete bit of the cmd register to 0.
3. Wait for the cmd bit in the rintsts register to change to 1.
4. Read the response to the IO_RW_DIRECT command (R5) in the response registers
for any errors.
f For more information about response values, refer to the Physical Layer Simplified
Specification, Version 3.01 as described in “References” on page 11–79.

Table 11–18. cmdarg Register Settings for SD/SDIO ABORT Command

Bits Contents Value
31 R/W flag 1
30:28 Function number 0, for access to the CCCR in the card device
27 RAW flag 1, if needed to read after write
26 Don't care -
25:9 Register address 0x06
8 Don't care -
7:0 Write data Function number to abort
Internal DMA Controller Operations

For better performance, you can use the internal DMA controller to transfer data
between the host and the controller. This section describes the internal DMA
controller’s initialization process, and transmission sequence, and reception sequence.
Internal DMA Controller Initialization

To initialize the internal DMA controller, perform the following steps:
1. Set the required bmod register bits:
■ If the internal DMA controller enable bit (de) of the bmod register is set to 0
during the middle of a DMA transfer, the change has no effect. Disabling only
takes effect for a new data transfer command.
■ Issuing a software reset immediately terminates the transfer. Prior to issuing a
software reset, Altera recommends the host reset the DMA interface by setting
the dma_reset bit of the ctrl register to 1.
■ The pbl field of the bmod register is read-only and a direct reflection of the
contents of the DMA multiple transaction size field
(dw_dma_multiple_transaction_size) in the fifoth register.
■ The fb bit of the bmod register has to be set appropriately for system
performance.
2. Write to the idinten register to mask unnecessary interrupt causes according to
the following guidelines:
■ When a Descriptor Unavailable interrupt is asserted, the software needs to
form the descriptor, appropriately set its own bit, and then write to the poll
demand register (pldmnd) for the internal DMA controller to re-fetch the
descriptor.
■ It is always appropriate for the software to enable abnormal interrupts because
any errors related to the transfer are reported to the software.

3. Populate either a transmit or receive descriptor list in memory. Then write the base
address of the first descriptor in the list to the internal DMA controller’s descriptor
list base address register (dbaddr). The DMA controller then proceeds to load the
descriptor list from memory.
The next two sections, “Internal DMA Controller Transmission Sequences” and
“Internal DMA Controller Reception Sequences” on page 11–54, describe this step
in detail.
Internal DMA Controller Transmission Sequences

To use the internal DMA controller to transmit data, perform the following steps:
1. The host sets up the Descriptor fields (DES0—DES3) for transmission and sets the
OWN bit (DES0[31]) to 1. The host also loads the data buffer in system memory
with the data to be written to the SD card.
2. The host writes the appropriate write data command (SD/SDIO WRITE_BLOCK
or WRITE_MULTIPLE_BLOCK) to the cmd register. The internal DMA controller
determines that a write data transfer needs to be performed.
3. The host sets the required transmit threshold level in the tx_wmark field in the
fifoth register.
4. The internal DMA controller engine fetches the descriptor and checks the OWN
bit. If the OWN bit is set to 0, the host owns the descriptor. In this case, the internal
DMA controller enters the suspend state and asserts the Descriptor Unable
interrupt. The host then needs to set the descriptor OWN bit to 1 and release the
DMA controller by writing any value to the pldmnd register.
5. The host must write the descriptor base address to the dbaddr register.
6. The internal DMA controller waits for the Command Done (CD) bit in the rintsts
register to be set to 1, with no errors from the BIU. This condition indicates that a
transfer can be done.
7. The internal DMA controller engine waits for a DMA interface request from BIU.
The BIU divides each transfer into smaller chunks. Each chunk is an internal
request to the DMA. This request is generated based on the transmit threshold
value.
8. The internal DMA controller fetches the transmit data from the data buffer in the
system memory and transfers the data to the FIFO buffer in preparation for
transmission to the card.
9. When data spans across multiple descriptors, the internal DMA controller fetches
the next descriptor and continues with its operation with the next descriptor. The
Last Descriptor bit in the descriptor DES0 field indicates whether the data spans
multiple descriptors or not.
10. When data transmission is complete, status information is updated in the idsts
register by setting the ti bit to 1, if enabled. Also, the OWN bit is set to 0 by the
DMA controller by updating the DES0 field of the descriptor.

Internal DMA Controller Reception Sequences

To use the internal DMA controller to receive data, perform the following steps:
1. The host sets up the descriptor fields (DES0—DES3) for reception, sets the OWN
(DES0 [31]) to 1.
2. The host writes the read data command to the cmd register in BIU. The internal
DMA controller determines that a read data transfer needs to be performed.
3. The host sets the required receive threshold level in the rx_wmark field in the
fifoth register.
4. The internal DMA controller engine fetches the descriptor and checks the OWN
bit. If the OWN bit is set to 0, the host owns the descriptor. In this case, the internal
DMA controller enters suspend state and asserts the Descriptor Unable interrupt.
The host then needs to set the descriptor OWN bit to 1 and release the DMA
controller by writing any value to the pldmnd register.
5. The host must write the descriptor base address to the dbaddr register.
6. The internal DMA controller waits for the CD bit in the rintsts register to be set to
1, with no errors from the BIU. This condition indicates that a transfer can be done.
7. The internal DMA controller engine waits for a DMA interface request from the
BIU. The BIU divides each transfer into smaller chunks. Each chunk is an internal
request to the DMA. This request is generated based on the receive threshold
value.
8. The internal DMA controller fetches the data from the FIFO buffer and transfers
the data to system memory.
9. When data spans across multiple descriptors, the internal DMA controller fetches
the next descriptor and continues with its operation with the next descriptor. The
Last Descriptor bit in the descriptor indicates whether the data spans multiple
descriptors or not.
10. When data reception is complete, status information is updated in the idsts
register by setting the ri bit to 1, if enabled. Also, the OWN bit is set to 0 by the
DMA controller by updating the DES0 field of the descriptor.
Commands for SDIO Card Devices

This section describes the commands to temporarily halt the transfers between the
controller and SDIO card device.
Suspend and Resume Sequence

For SDIO cards, a data transfer between an I/O function and the controller can be
temporarily halted using the SUSPEND command. This capability might be required
to perform a high-priority data transfer with another function. When desired, the
suspended data transfer can be resumed using the RESUME command.

The SUSPEND and RESUME operations are implemented by writing to the

appropriate bits in the CCCR (Function 0) of the SDIO card. To read from or write to
the CCCR, use the controller’s IO_RW_DIRECT command.
Suspend
To suspend data transfer, perform the following steps:
1. Check if the SDIO card supports the SUSPEND/RESUME protocol. This can be
done through the SBS bit in the CCCR at offset 0x08 of the card.
2. Check if the data transfer for the required function number is in process. The
function number that is currently active is reflected in the function select bits (FSx)
of the CCCR, bits 3:0 at offset 0x0D of the card.
1 If the bus status bit (BS), bit 0 at address 0xC, is 1, only the function number
given by the FSx bits is valid.
3. To suspend the transfer, set the bus release bit (BR), bit 2 at address 0xC, to 1.
4. Poll the BR and BS bits of the CCCR at offset 0x0C of the card until they are set to 0.
The BS bit is 1 when the currently-selected function is using the data bus. The BR
bit remains 1 until the bus release is complete. When the BR and BS bits are 0, the
data transfer from the selected function is suspended.
5. During a read-data transfer, the controller can be waiting for the data from the
card. If the data transfer is a read from a card, the controller must be informed after
the successful completion of the SUSPEND command. The controller then resets
the data state machine and comes out of the wait state. To accomplish this, set the
abort read data bit (abort_read_data) in the ctrl register to 1.
6. Wait for data completion, by polling until the dto bit is set to 1 in the rintsts
register.
To determine the number of pending bytes to transfer, read the transferred CIU
card byte count (tcbcnt) register of the controller. Subtract this value from the total
transfer size. You use this number to resume the transfer properly.
Resume
To resume the data transfer, perform the following steps:
1. Check that the card is not in a transfer state, which confirms that the bus is free for
data transfer.
2. If the card is in a disconnect state, select it using the SD/SDIO
SELECT/DESELECT_CARD command. The card status can be retrieved in
response to an IO_RW_DIRECT or IO_RW_EXTENDED command.
3. Check that a function to be resumed is ready for data transfer. Determine this state
by reading the corresponding RF<n> flag in CCCR at offset 0x0F of the card. If
RF<n> = 1, the function is ready for data transfer.
1 For detailed information about the RF<n> flags, refer to SDIO Simplified
Specification Version 2.00 as described in “References” on page 11–79:

4. To resume transfer, use the IO_RW_DIRECT command to write the function

number at the FSx bits in the CCCR, bits 3:0 at offset 0x0D of the card. Form the
command argument for the IO_RW_DIRECT command and write it to the cmdarg
register. Bit values are listed in Table 11–19.
Table 11–19. cmdarg Bit Values for RESUME Command

Bits Content Value
31 R/W flag 1
30:28 Function number 0, for CCCR access
27 RAW flag 1, read after write
26 Don't care -
25:9 Register address 0x0D
8 Don't care -
7:0 Write data Function number that is to be resumed
5. Write the block size value to the blksiz register. Data is transferred in units of this
block size.
6. Write the byte count value to the bytcnt register. Specify the total size of the data
that is the remaining bytes to be transferred. It is the responsibility of the software
to handle the data.
To determine the number of pending bytes to transfer, read the transferred CIU
card byte count register (tcbcnt). Subtract this value from the total transfer size
to calculate the number of remaining bytes to transfer.
7. Write to the cmd register similar to a block transfer operation. When the cmd
register is written, the command is sent and the function resumes data transfer. For
more information, refer to “Single-Block or Multiple-Block Read” on page 11–46
and “Single-Block or Multiple-Block Write” on page 11–48.
8. Read the resume data flag (DF) of the SDIO card device. Interpret the DF flag as
follows:
■ DF=1—The function has data for the transfer and begins a data transfer as soon
as the function or memory is resumed.
■ DF=0—The function has no data for the transfer. If the data transfer is a read,
the controller waits for data. After the data timeout period, it issues a data
timeout error.

Read-Wait Sequence
Read_wait is used with SDIO cards only. It temporarily stalls the data transfer, either
from functions or memory, and allows the host to send commands to any function
within the SDIO card device. The host can stall this transfer for as long as required.
The controller provides the facility to signal this stall transfer to the card. To signal the
stall, perform the following steps:
1. Check if the card supports the read_wait facility by reading the SDIO card’s SRW
bit, bit 2 at offset 0x8 in the CCCR.
2. If this bit is 1, all functions in the card support the read_wait facility. Use the
SD/SDIO IO_RW_DIRECT command to read this bit.
3. If the card supports the read_wait signal, assert it by setting the read wait bit
(read_wait) in the ctrl register to 1.
4. Reset the read_wait bit to 0 in the ctrl register.
CE-ATA Data Transfer Commands

This section describes CE-ATA data transfer commands. For information about the
basic settings and interrupts generated for different conditions, refer to “Data Transfer
Commands” on page 11–44.
Reset and Card Device Discovery Overview

Before starting any CE-ATA operations, the host must perform a MMC reset and
initialization procedure. The host and card device must negotiate the MMC transfer
(MMC TRAN) state before the card enters the MMC TRAN state.
f For information about the MMC TRAN state, MMC reset and initialization, refer to
JEDEC Standard No. 84-A441, as referenced in “References” on page 11–79.
The host must follow the existing MMC discovery procedure to negotiate the MMC
TRAN state. After completing normal MMC reset and initialization procedures, the
host must query the initial ATA task file values using the RW_REG or CMD39
command.
By default, the MMC block size is 512 bytes—indicated by bits 1:0 of the srcControl
register inside the CE-ATA card device. The host can negotiate the use of a 1 KB or
4 KB MMC block sizes. The card indicates MMC block sizes that it can support
through the srcCapabilities register in the MMC; the host reads this register to
negotiate the MMC block size. Negotiation is complete when the host controller
writes the MMC block size into the srcControl register bits 1:0 of the card.
ATA Task File Transfer Overview

ATA task file registers are mapped to addresses 0x00h through 0x10h in the MMC
register space. The RW_REG command is used to issue the ATA command, and the
ATA task file is transmitted in a single RW_REG MMC command sequence.
The host software stack must write the task file image to the FIFO buffer before setting
the cmdarg and cmd registers in the controller. The host processor then writes the
address and byte count to the cmdarg register before setting the cmd register bits.
For the RW_REG command, there is no CCS from the CE-ATA card device.

ATA Task File Transfer Using the RW_MULTIPLE_REGISTER (RW_REG)

Command
This command involves data transfer between the CE-ATA card device and the
controller. To send a data command, the controller needs a command argument, total
data size, and block size. Software receives or sends data through the FIFO buffer.
To implement an ATA task file transfer (read or write), perform the following steps:
1. Write the data size in bytes to the bytcnt register. bytcnt must equal the block size,
because the controller expects a single block transfer.
2. Write the block size in bytes to the blksiz register.
3. Write the cmdarg register with the beginning register address.
You must set the cmdarg, cmd, blksiz, and bytcnt registers according to Table 11–20
through Table 11–23.
Table 11–20. cmdarg Register Settings for ATA Task File Transfer
Bit Value Comment
31 1 or 0 Set to 0 for read operation or set to 1 for write operation
30:24 0 Reserved (bits set to 0 by host processor)
23:18 0 Starting register address for read or write (DWORD aligned)
17:16 0 Register address (DWORD aligned)
7:2 16 Number of bytes to read or write (integral number of DWORD)
1:0 0 Byte count in integral number of DWORD
Table 11–21. cmd Register Settings for ATA Task File Transfer (Part 1 of 2)
Bit Value Comment
start_cmd 1 -
ccs_expected 0 CCS is not expected
read_ceata_device 0 or 1 Set to 1 if RW_BLK or RW_REG read
update_clk_regs_only 0 No clock parameters update command
card_num 0 -
send_initialization 0 No initialization sequence
stop_abort_cmd 0 -
send_auto_stop 0 -
Block transfer mode. Block size and byte count must match
transfer_mode 0
number of bytes to read or write
read_write 1 or 0 1 for write and 0 for read
data_expected 1 Data is expected
response_length 0 -
response_expect 1 -
Com Set this parameter to the command number. For example, set
cmd_index mand to 24 for SD/SDIO WRITE_BLOCK (CMD24) or 25 for
index WRITE_MULTIPLE_BLOCK (CMD25).

Table 11–21. cmd Register Settings for ATA Task File Transfer (Part 2 of 2)
Bit Value Comment
■ 0 for send command immediately
■ 1 for send command after previous DTO interrupt
■ 0 for not checking response CRC
■ 1 for checking response CRC
Table 11–22. blksiz Register Settings for ATA Task File Transfer
Bits Value Comment
31:16 0 Reserved bits set to 0
For accessing entire task file (16, 8-bit registers). Block size of 16
15:0 (block_size) 16
bytes
Table 11–23. bytcnt Register Settings for ATA Task File Transfer
Bits Value Comment
For accessing entire task file (16, 8-bit registers). Byte count
31:0 16
value of 16 is used with the block size set to 16.
ATA Payload Transfer Using the RW_MULTIPLE_BLOCK (RW_BLK) Command

This command involves data transfer between the CE-ATA card device and the
controller. To send a data command, the controller needs a command argument, total
data size, and block size. Software receives or sends data through the FIFO buffer.
To implement an ATA payload transfer (read or write), perform the following steps:
1. Write the data size in bytes to the bytcnt register.
2. Write the block size in bytes to the blksiz register. The controller expects a
single/multiple block transfer.
3. Write to the cmdarg register to indicate the data unit count.
You must set the cmdarg, cmd, blksiz, and bytcnt registers according to Table 11–24
through Table 11–27.
Table 11–24. cmdarg Register Settings for ATA Payload Transfer

Bits Value Comment
31 1 or 0 Set to 0 for read operation or set to 1 for write operation
15:8 Data count Data Count Unit [15:8]
7:0 Data count Data Count Unit [7:0]

Table 11–25. cmd Register Settings for ATA Payload Transfer

Bits Value Comment
start_cmd 1 -
CCS is expected. Set to 1 for the RW_BLK command if
ccs_expected 1 interrupts are enabled in CE-ATA card device (the
nIEN bit is set to 0 in the ATA control register)
read_ceata_device 0 or 1 Set to 1 for a RW_BLK or RW_REG read command
update_clk_regs_only 0 No clock parameters update command
card_num 0 -
send_initialization 0 No initialization sequence
stop_abort_cmd 0 -
send_auto_stop 0 -
Block transfer mode. Byte count must be integer
transfer_mode 0
multiple of 4kB. Block size can be 512, 1k or 4k bytes
read_write 1 or 0 1 for write and 0 for read
data_expected 1 Data is expected
response_length 0 -
response_expect 1 -
Set this parameter to the command number. For
Command example, set to 24 for SD/SDIO WRITE_BLOCK
cmd_index
index (CMD24) or 25 for WRITE_MULTIPLE_BLOCK
(CMD25).
■ 0 for send command immediately
■ 1 for send command after previous DTO interrupt
■ 0 for not checking response CRC
■ 1 for checking response CRC
Table 11–26. blksiz Register Settings for ATA Payload Transfer

Bits Value Comment
31:16 0 Reserved bits set to 0
512, 1024 or MMC block size can be 512, 1024 or 4096 bytes as
15:0 (block_size)
4096 negotiated by host
Table 11–27. bytcnt Register Settings for ATA Payload Transfer

Bits Value Comment
Byte count must be an integer multiple of the block
size. For ATA media access commands, byte count
31:0 <n>*block_size must be a multiple of 4 KB.
(<n>*block_size = <x>*4 KB, where <n> and <x> are
integers)
CE-ATA CCS
This section describes disabling the CCS, recovery after CCS timeout, and recovery
after I/O read transmission delay (NACIO) timeout.

Disabling the CCS

While waiting for the CCS for an outstanding RW_BLK command, the host can
disable the CCS by sending a CCSD command:
■ Send a CCSD command—the controller sends the CCSD command to the CE-ATA
card device if the send_ccsd bit is set to 1 in the ctrl register of the controller. This
bit can be set only after a response is received for the RW_BLK command.
■ Send an internal stop command—send an internally-generated SD/SDIO
STOP_TRANSMISSION (CMD12) command after sending the CCSD pattern. If
the send_auto_stop_ccsd bit of the ctrl register is also set to 1 when the controller
is set to send the CCSD pattern, the controller sends the internally-generated STOP
command to the CMD pin. After sending the STOP command, the controller sets the
acd bit in the rintsts register to 1.
Recovery after CCS Timeout

If a timeout occurs while waiting for the CCS, the host needs to send the CCSD
command followed by a STOP command to abort the pending ATA command. The
host can set up the controller to send an internally-generated STOP command after
sending the CCSD pattern:
■ Send CCSD command—set the send_ccsd bit in the ctrl register to 1.
■ Send external STOP command—terminate the data transfer between the CE-ATA
card device and the controller. For more information about sending the STOP
command, refer to “Transfer Stop and Abort Commands” on page 11–51.
■ Send internal STOP command—set the send_auto_stop_ccsd bit in the ctrl
register to 1, which tells the controller to send the internally-generated STOP
command. After sending the STOP command, the controller sets the acd bit in the
rintsts register to 1. The send_auto_stop_ccsd bit must be set to 1 along with
setting the send_ccsd bit.
Recovery after I/O Read Transmission Delay (NACIO) Timeout

If the I/O read transmission delay (NACIO) timeout occurs for the CE-ATA card device,
perform one of the following steps to recover from the timeout:
■ If the CCS is expected from the CE-ATA card device (that is, the ccs_expected bit
is set to 1 in the cmd register), follow the steps in “Recovery after CCS Timeout” on
page 11–61.
■ If the CCS is not expected from the CE-ATA card device, perform the following
steps:
a. Send an external STOP command.
b. Terminate the data transfer between the controller and CE-ATA card device.

Reduced ATA Command Set

It is necessary for the CE-ATA card device to support the reduced ATA command
subset. This section describes the reduced command set.
The IDENTIFY DEVICE Command

The IDENTIFY DEVICE command returns a 512-byte data structure to the host that
describes device-specific information and capabilities. The host issues the IDENTIFY
DEVICE command only if the MMC block size is set to 512 bytes. Any other MMC
block size has indeterminate results.
The host issues a RW_REG command for the ATA command, and the data is retrieved
with the RW_BLK command.
The host controller uses the following settings while sending a RW_REG command
for the IDENTIFY DEVICE ATA command. The following list shows the primary bit
settings:
■ cmd register setting: data_expected bit set to 0
■ cmdarg register settings:
■ Bit [31] set to 0
■ Bits [7:2] set to 128
■ All other bits set to 0
■ Task file settings:
■ Command field of the ATA task file set to 0xEC
■ Reserved fields of the task file set to 0
■ bytcnt register and block_size field of the blksiz register: set to 16
The host controller uses the following settings for data retrieval (RW_BLK command):
■ cmd register settings:
■ ccs_expected set to 1
■ data_expected set to 1
■ Bit [31] set to 0 (read operation)
■ Bits [15:0] set to 1 (data unit count = 1)
The READ DMA EXT Command

The READ DMA EXT command reads a number of logical blocks of data from the
card device using the Data-In data transfer protocol. The host uses a RW_REG
command to issue the ATA command and the RW_BLK command for the data
transfer.

The WRITE DMA EXT Command

The WRITE DMA EXT command writes a number of logical blocks of data to the card
device using the Data-Out data transfer protocol. The host uses a RW_REG command
to issue the ATA command and the RW_BLK command for the data transfer.
The STANDBY IMMEDIATE Command

This ATA command causes the card device to immediately enter the most aggressive
power management mode that still retains internal device context. No data transfer
(RW_BLK) is expected for this command.
For card devices that do not provide a power savings mode, the STANDBY
IMMEDIATE command returns a successful status indication. The host issues a
RW_REG command for the ATA command, and the status is retrieved with the
SD/SDIO CMD39 or RW_REG command. Only the status field of the ATA task file
contains the success status; there is no error status.
The host controller uses the following settings while sending the RW_REG command
for the STANDBY IMMEDIATE ATA command:
■ Bits [7:2] set to 4
■ Command field of the ATA task file set to 0xE0
The FLUSH CACHE EXT Command

For card devices that buffer/cache written data, the FLUSH CACHE EXT command
ensures that buffered data is written to the card media. For cards that do not buffer
written data, the FLUSH CACHE EXT command returns a success status. No data
transfer (RW_BLK) is expected for this ATA command.
The host issues a RW_REG command for the ATA command, and the status is
retrieved with the SD/SDIO CMD39 or RW_REG command. There can be error status
for this ATA command, in which case fields other than the status field of the ATA task
file are valid.
The host controller uses the following settings while sending the RW_REG command
for the STANDBY IMMEDIATE ATA command:
■ Bits [7:2] set to 4


■ Command field of the ATA task file set to 0xEA
Card Read Threshold

When an application needs to perform a single or multiple block read command, the
application must set the cardthrctl register with the appropriate card read threshold
size in the card read threshold field (cardrdthreshold) and set the cardrdthren bit to
1. This additional information specified in the controller ensures that the controller
sends a read command only if there is space equal to the card read threshold available
in the RX FIFO buffer. This in turn ensures that the card clock is not stopped in the
middle a block of data being transmitted from the card. Set the card read threshold to
the block size of the transfer to guarantee there is a minimum of one block size of
space in the RX FIFO buffer before the controller enables the card clock.
The card read threshold is required when the round trip delay is greater than half of
sdmmc_clk_divided.
Recommended Usage Guidelines for Card Read Threshold

1. The cardthrctl register must be set before setting the cmd register for a data read
command.
2. The cardthrctl register must not be set while a data transfer command is in
progress.
3. The cardrdthreshold field of the cardthrctl register must be set to at the least the
block size of a single or multiblock transfer. A cardrdthreshold field setting
greater than or equal to the block size of the read transfer ensures that the card
clock does not stop in the middle of a block of data.
4. If the round trip delay is greater than half of the card clock period, card read
threshold must be enabled and the card threshold must be set as per guideline 3 to
guarantee that the card clock does not stop in the middle of a block of data.
5. If the cardrdthreshold field is set to less than the block size of the transfer, the
host must ensure that the receive FIFO buffer never overflows during the read
transfer. Overflow can cause the card clock from the controller to stop. The
controller is not able to guarantee that the card clock does not stop during a read
transfer.
1 If the cardrdthreshold field of the cardthrctl register, and the rx_wmark and
dw_dma_multiple_transaction_size fields of the fifoth register are set incorrectly,
the card clock might stop indefinitely, with no interrupts generated by the controller.
Card Read Threshold Programming Sequence

Most cards, such as SDHC or SDXC, support block sizes that are either specified in the
card or are fixed to 512 bytes. For SDIO, MMC, and standard capacity SD cards that
support partial block read (READ_BL_PARTIAL set to 1 in the CSD register of the
card device), the block size is variable and can be chosen by the application.

To use the card read threshold feature effectively and to guarantee that the card clock
does not stop because of a FIFO Full condition in the middle of a block of data being
read from the card, follow these steps:
1. Choose a block size that is a multiple of four bytes.
2. Enable card read threshold feature. The card read threshold can be enabled only if
the block size for the given transfer is less than or equal to the total depth of the
FIFO buffer:
(block size / 4) ≤ 1024
3. Choose the card read threshold value:
■ If (block size / 4) ≥ 512, choose cardrdthreshold such that
cardrdthreshold ≤ (block size / 4) in bytes
■ If (block size / 4) < 512, choose cardrdthreshold such that
cardrdthreshold = (block size / 4) in bytes
4. Set the dw_dma_multiple_transaction_size field in the fifoth register to the
number of transfers that make up a DMA transaction. For example, size = 1 means
4 bytes are moved. The possible values for the size are 1, 4, 8, 16, 32, 64, 128, and
256 transfers. Select the size so that the value (block size / 4) is evenly divided by
the size.
5. Set the rx_wmark field in the fifoth register to the size – 1.
For example, if your block size is 512 bytes, legal values of
dw_dma_multiple_transaction_size and rx_wmark are listed in Table 11–28.
Table 11–28. Legal Values of dw_dma_multiple_transaction_size and rx_wmark for Block Size = 512
Block Size dw_dma_multiple_transaction_size rx_wmark
512 1 0
512 4 3
512 8 7
512 16 15
512 32 31
512 64 63
512 128 127
Card Read Threshold Programming Examples

This section shows examples of how to program the card read threshold.
■ Choose a block size that is a multiple of 4 (the number of bytes per FIFO location),
and less than 4096 (1024 FIFO locations). For example, a block size of 3072 bytes is
legal, because 3072 / 4 = 768 FIFO locations.
■ For DMA mode, choose the size so that block size is a multiple of the size. For
example size = 128, where block size%size = 0.

■ Set the rx_wmark field = size – 1. For example, the rx_wmark field = 128 – 1 = 127.
■ Because block size > ½ FifoDepth, set the cardrdthreshold field to the block size.
For example, the cardrdthreshold field = 3072 bytes.
Figure 11–12. FIFO Buffer content when Card Read Threshold is set to 768
SIZE = 128
FIFO Locations
256 FIFO Locations

Read Data Data Read
Unfilled During One
to Host from Card
Block Read
Block Size = cardrdthreshold = 768
One Block of Data Filled 768 FIFO Locations
FIFO Depth = 1,024

rx_wmark = 127
FIFO Locations
Interrupt and Error Handling

This section describes how to use interrupts to handle errors. On power-on or reset,
interrupts are disabled (the int_enable bit in the ctrl register is set to 0), and all the
interrupts are masked (the intmask register default is 0). The controller error handling
includes the following types of errors:
■ Response and data timeout errors—For response time-outs, the host software can
retry the command. For data time-outs, the controller has not received the data
start bit from the card, so software can either retry the whole data transfer again or
retry from a specified block onwards. By reading the contents of the tcbcnt
register later, the software can decide how many bytes remain to be copied (read).
■ Response errors—Set to 1 when an error is received during response reception. If
the response received is invalid, the software can retry the command.
■ Data errors—Set to 1 when a data receive error occurs. Examples of data receive
errors:
■ Data CRC
■ Start bit not found
■ End bit not found
These errors can be occur on any block. On receipt of an error, the software can
issue an SD/SDIO STOP or SEND_IF_COND command, and retry the command
for either the whole data or partial data.
■ Hardware locked error—Set to 1 when the controller cannot load a command
issued by software. When software sets the start_cmd bit in the cmd register to 1,
the controller tries to load the command. If the command buffer already contains a
command, this error is raised, and the new command is discarded, requiring the
software to reload the command.

■ FIFO buffer underrun/overrun error—If the FIFO buffer is full and software tries
to write data to the FIFO buffer, an overrun error is set. Conversely, if the FIFO
buffer is empty and the software tries to read data from the FIFO buffer, an
underrun error is set. Before reading or writing data in the FIFO buffer, the
software must read the FIFO buffer empty bit (fifo_empty) or FIFO buffer full bit
(fifo_full) in the status register.
■ Data starvation by host timeout—This condition occurs when software does not
service the FIFO buffer fast enough to keep up with the controller. Under this
condition and when a read transfer is in process, the software must read data from
the FIFO buffer, which creates space for further data reception. When a transmit
operation is in process, the software must write data to fill the FIFO buffer so that
the controller can write the data to the card.
■ CE-ATA errors
■ CRC error on command—If a CRC error is detected for a command, the
CE-ATA card device does not send a response, and a response timeout is
expected from the controller. The ATA layer is notified that an MMC transport
layer error occurred.
■ Write operation—Any MMC transport layer error known to the card device
causes an outstanding ATA command to be terminated. The ERR bits are set in
the ATA status registers and the appropriate error code is sent to the Error
Register (Error) on the ATA card device.
If the device interrupt bit of the CE-ATA card (the nIEN bit in the ATA control
register) is set to 0, the CCS is sent to the host.
If the device interrupt bit is set to 1, the card device completes the entire data
unit count if the host controller does not abort the ongoing transfer.
1 During a multiple-block data transfer, if a negative CRC status is received

from the card device, the data path signals a data CRC error to the BIU by
setting the dcrc bit in the rintsts register to 1. It then continues further
data transmission until all the bytes are transmitted.
■ Read operation—If MMC transport layer errors are detected by the host
controller, the host completes the ATA command with an error status. The host
controller can issue a CCSD command followed by a STOP_TRANSMISSION
(CMD12) command to abort the read transfer. The host can also transfer the
entire data unit count bytes without aborting the data transfer.
Booting Operation for eMMC and MMC

This section describes how to set up the controller for eMMC and MMC boot
operation.
Boot Operation by Holding Down the CMD Line

The controller can boot from MMC4.3, MMC4.4, and MMC4.41 cards by holding
down the CMD line.
f For information about this boot method, refer to the following specifications, as
referenced in “References” on page 11–79:

■ JEDEC Standard No. 84-A441

■ JEDEC Standard No. JESD84-A43
Boot Operation for eMMC Card Device

Figure 11–13 illustrates the steps to perform the boot process for eMMC card devices.
The detailed steps are described following the flow chart.
Figure 11–13. Flow for eMMC Boot Operation
Start
Step 1
Step 2 to Step 10
expect_boot_ack = 1 expect_boot_ack = 0
Step 11 Step 12
No Start Start Pattern & No Start Pattern & No Boot Data Boot Data
Pattern Boot Data Received Boot Data Received Received Received
Step 11.a Step 11.b Step 11.c Step 12.a Step 12.b
Successfully NAC Timeout

Received Boot Successfully
Data Received Boot
Step 11.f Data
Stop
1. The software driver performs the following checks:

■ If the eMMC card device supports boot operation (the
BOOT_PARTITION_ENABLE bit is set to 1 in the EXT_CSD register of the
eMMC card).
■ The BOOT_SIZE_MULT and BOOT_BUS_WIDTH values in the EXT_CSD
register, to be used during the boot process.

2. The software sets the following bits:

■ Sets masks for interrupts, by setting the appropriate bits to 0 in the intmask
register.
■ Sets the global int_enable bit of the ctrl register to 1. Other bits in the ctrl
register must be set to 0.
1 Altera recommends that you write 0xFFFFFFFF to the rintsts and idsts
registers to clear any pending interrupts before setting the int_enable bit.
For internal DMA controller mode, the software driver needs to unmask all
the relevant fields in the idinten register.
3. If the software driver needs to use the internal DMA controller to transfer the boot
data received, it must perform the following steps:
■ Set up the descriptors as described in “Internal DMA Controller Transmission
Sequences” on page 11–53 and “Internal DMA Controller Reception
Sequences” on page 11–54.
■ Set the use_internal_dmac bit of the ctrl register to 1.
4. Set the card device frequency to 400 kHz using the clkdiv registers. For more
information, refer to “Clock Setup” on page 11–40.
5. Set the data_timeout field of the tmout register equal to the card device total
access time, NAC.
6. Set the blksiz register to 0x200 (512 bytes).
7. Set the bytcnt register to a multiple of 128 KB, as indicated by the
BOOT_SIZE_MULT value in the card device.
8. Set the rx_wmark field in the fifoth register. Typically, the threshold value can be
set to 512, which is half the FIFO buffer depth.
9. Set the following fields in the cmd register:
■ Initiate the command by setting start_cmd = 1
■ Enable boot (enable_boot) = 1
■ Expect boot acknowledge (expect_boot_ack):
■ If a start-acknowledge pattern is expected from the card device, set
expect_boot_ack to 1.
■ If a start-acknowledge pattern is not expected from the card device, set
■ Card number (card_number) = 0
■ data_expected = 1
■ Reset the remainder of cmd register bits to 0
10. If no start-acknowledge pattern is expected from the card device
(expect_boot_ack set to 0) proceed to step 12.

11. This step handles the case where a start-acknowledge pattern is expected
(expect_boot_ack was set to 1 in step 9).
a. If the Boot ACK Received interrupt is not received from the controller within
50 ms of initiating the command (step 9), the software driver must set the
following cmd register fields:
■ start_cmd = 1
■ Disable boot (disable_boot)= 1
■ card_number = 0
■ All other fields = 0
The controller generates a Command Done interrupt after deasserting the CMD
pin of the card interface.
If internal DMA controller mode is used for the boot process, the controller
performs the following steps after the Boot ACK Received timeout:
■ The DMA descriptor is closed.
■ The ces bit in the idsts register is set, indicating the Boot ACK Received
timeout.
■ The ri bit of the idsts register is not set.
b. If the Boot ACK Received interrupt is received, the software driver must clear
this interrupt by writing 1 to the ces bit in the idsts register.
Within 0.95 seconds of the Boot ACK Received interrupt, the Boot Data Start
interrupt must be received from the controller. If this does not occur, the
software driver must write the following cmd register fields:
■ start_cmd = 1
■ disable_boot = 1
■ card_number = 0
The controller generates a Command Done interrupt after deasserting the CMD
pin of the card interface.
■ The DMA descriptor is closed
■ The ces bit in the idsts register is set, indicating Boot Data Start timeout
■ The ri bit of the idsts register is not set
c. If the Boot Data Start interrupt is received, it indicates that the boot data is
being received from the card device. When the DMA engine is not in internal
DMA controller mode, the software driver can then initiate a data read from
the controller based on the rxdr interrupt bit in the rintsts register.
In internal DMA controller mode, the DMA engine starts transferring the data
from the FIFO buffer to the system memory as soon as the level set in the
rx_wmark field of the fifoth register is reached.

At the end of a successful boot data transfer from the card, the following
interrupts are generated:
■ The cmd bit and dto bit in the rintsts register
■ The ri bit in the idsts register, in internal DMA controller mode only
d. If an error occurs in the boot ACK pattern (0b010) or an EBE occurs:
■ The controller automatically aborts the boot process by pulling the CMD
line high
■ The controller generates a Command Done interrupt
■ The controller does not generate a Boot ACK Received interrupt
■ The application aborts the boot transfer
e. In internal DMA controller mode:
■ If the software driver creates more descriptors than required by the
received boot data, the extra descriptors are not closed by the controller.
Software cannot reuse the descriptors until they are closed.
■ If the software driver creates fewer descriptors than required by the
received boot data, the controller generates a Descriptor Unavailable
interrupt and does not transfer any further data to system memory.
f. If NAC is violated between data block transfers, the DRTO interrupt is asserted.
In addition, if there is an error associated with the start or end bit, the SBE or
EBE interrupt is also generated.
The boot operation for eMMC card devices is complete. Do not execute the
remaining step (12).

12. This step handles the case where no start-acknowledge pattern is expected
a. If the Boot Data Start interrupt is not received from the controller within 1
second of initiating the command (step 9), the software driver must write the
cmd register with the following fields:
■ start_cmd = 1
■ disable_boot = 1
■ card_number = 0
The controller generates a Command Done interrupt after deasserting the
CMD line of the card. In internal DMA controller mode, the descriptor is closed
and the ces bit in the idsts register is set to 1, indicating a Boot Data Start
timeout.
b. If a Boot Data Start interrupt is received, it indicates that the boot data is being
received from the card device. When the DMA engine is not in internal DMA
controller mode, the software driver can then initiate a data read from the
controller based on the rxdr interrupt bit in the rintsts register.
from the FIFO buffer to the system memory as soon as the level specified in the
At the end of a successful boot data transfer from the card, the following
c. In internal DMA controller mode:
The boot operation for eMMC card devices is complete.
Boot Operation for Removable MMC4.3, MMC4.4 and MMC4.41 Cards

Removable MMC4.3, MMC4.4, and MMC4.41 cards differ with respect to eMMC in
that the controller is not aware whether these cards support the boot mode of
operation when plugged in. Thus, the controller must:
1. Discover these cards as it would discover MMC4.0/4.1/4.2 cards for the first time
2. Know the card characteristics
3. Decide whether to perform a boot operation or not

For these cards, the software driver must perform the following steps:
1. Discover the card as described in “Enumerated Card Stack” on page 11–37.
2. Read the EXT_CSD register of the card and examine the following fields:
■ BOOT_PARTITION_ENABLE
■ BOOT_SIZE_MULT
■ BOOT_INFO
3. If necessary, the software can manipulate the boot information in the card.
f For more information, refer to “Access to Boot Partition” in the following

specifications, as referenced in “References” on page 11–79:

4. If the host processor needs to perform a boot operation at the next power-up cycle,
it can manipulate the EXT_CSD register contents by using a SWITCH_FUNC
command.
5. After this step, the software driver must power down the card by writing to the
pwren register.
6. From here on, use the same steps as in “Alternative Boot Operation for eMMC
Card Devices” on page 11–74.
Alternative Boot Operation

The alternative boot operation differs from the previous boot operation in that
software uses the SD/SDIO GO_IDLE_STATE command to boot the card, rather than
holding down the CMD line of the card. The alternative boot operation can be
performed only if bit 0 in the BOOT_INFO register is set to 1. BOOT_INFO is located
at offset 228 in the EXT_CSD registers.
f For detailed information about alternative boot operation, refer to the following
specifications, as referenced in “References” on page 11–79:


Alternative Boot Operation for eMMC Card Devices

Figure 11–14 illustrates the sequence of steps required to perform the alternative boot
operation for eMMC card devices. The detailed steps are described following the flow
chart.
Figure 11–14. Flow for eMMC Alternative Boot Operation
Start
Step 1
Step 2 to Step 12
expect_boot_ack = 1 expect_boot_ack = 0
Step 13 Step 15
No Start Start Pattern & No Start Pattern & No Boot Data Boot Data
Pattern Boot Data Received Boot Data Received Received Received
Step 14.a Step 14.b Step 14.e Step 16.a Step 16.b
Successfully NAC Timeout

Received Boot Successfully
Data (14.d) Received Boot
Step 14.h Data (16.d)
Stop
1. The software driver checks:

■ If the eMMC card device supports alternative boot operation (the BOOT_INFO
bit is set to 1 in the eMMC card).
■ The BOOT_SIZE_MULT and BOOT_BUS_WIDTH values in the card device to
use during the boot process.
2. The software sets the following bits:
■ Sets masks for interrupts by resetting the appropriate bits to 0 in the intmask
register.
■ Sets the int_enable bit of the ctrl register to 1. Other bits in the ctrl register
must be set to 0.
1 Altera recommends writing 0xFFFFFFFF to the rintsts register and idsts

register to clear any pending interrupts before setting the int_enable bit.
For internal DMA controller mode, the software driver needs to unmask all
the relevant fields in the idinten register.

3. If the software driver needs to use the internal DMA controller to transfer the boot
data received, it must perform the following actions:
■ Set up the descriptors as described in “Internal DMA Controller Transmission
Sequences” on page 11–53 and “Internal DMA Controller Reception
Sequences” on page 11–54.
■ Set the use internal DMAC bit (use_internal_dmac) of the ctrl register to 1.
4. Set the card device frequency to 400 kHz using the clkdiv registers. For more
information, refer to “Clock Setup” on page 11–40. Ensure that the card clock is
running.
5. Wait for a time that ensures that at least 74 card clock cycles have occurred on the
card interface.
6. Set the data_timeout field of the tmout register equal to the card device total
access time, NAC.
7. Set the blksiz register to 0x200 (512 bytes).
8. Set the bytcnt register to multiples of 128K bytes, as indicated by the
BOOT_SIZE_MULT value in the card device.
9. Set the rx_wmark field in the fifoth register. Typically, the threshold value can be
set to 512, which is half the FIFO buffer depth.
10. Set the cmdarg register to 0xFFFFFFFA.
11. Initiate the command, by setting the cmd register with the following fields:
■ start_cmd = 1
■ enable_boot = 1
■ expect_boot_ack:
■ If a start-acknowledge pattern is expected from the card device, set
■ If a start-acknowledge pattern is not expected from the card device, set
■ card_number = 0
■ data_expected = 1
■ cmd_index = 0
■ Set the remainder of cmd register bits to 0.
12. If no start-acknowledge pattern is expected from the card device
(expect_boot_ack set to 0) jump to step 15.
13. Wait for the Command Done interrupt.

14. This step handles the case where a start-acknowledge pattern is expected
a. If the Boot ACK Received interrupt is not received from the controller within
50 ms of initiating the command (step 11), the start pattern was not received.
The software driver must discontinue the boot process and start with normal
discovery.
■ The ces bit in the idsts register is set to 1, indicating the Boot ACK
Received timeout.
b. If the Boot ACK Received interrupt is received, the software driver must clear
this interrupt by writing 1 to it.
Within 0.95 seconds of the Boot ACK Received interrupt, the Boot Data Start
interrupt must be received from the controller. If this does not occur, the
software driver must discontinue the boot process and start with normal
discovery.
■ The ces bit in the idsts register is set to 1, indicating Boot Data Start
timeout.
c. If the Boot Data Start interrupt is received, it indicates that the boot data is
being received from the card device. When the DMA engine is not in internal
DMA controller mode, the software driver can then initiate a data read from
the controller based on the rxdr interrupt bit in the rintsts register.
d. The software driver must terminate the boot process by instructing the
controller to send the SD/SDIO GO_IDLE_STATE command:
■ Reset the cmdarg register to 0.
■ Set the start_cmd bit of the cmd register to 1, and all other bits to 0.
e. At the end of a successful boot data transfer from the card, the following

f. If an error occurs in the boot ACK pattern (0b010) or an EBE occurs:

■ The controller does not generate a Boot ACK Received interrupt.
■ The controller detects Boot Data Start and generates a Boot Data Start
interrupt.
■ The controller continues to receive boot data.
■ The application must abort the boot process after receiving a Boot Data
Start interrupt.
g. In internal DMA controller mode:
h. If NAC is violated between data block transfers, a DRTO interrupt is asserted.
Apart from this, if there is an error associated with the start or end bit, the SBE
or EBE interrupt is also generated.
The alternative boot operation for eMMC card devices is complete. Do not execute
the remaining steps (15 and 16).
15. Wait for the Command Done interrupt.
16. This step handles the case where a start-acknowledge pattern is not expected
a. If the Boot Data Start interrupt is not received from the controller within 1
second of initiating the command (step 11), the software driver must
discontinue the boot process and start with normal discovery.
In internal DMA controller mode:
■ The ces bit in the idsts register is set to 1, indicating Boot Data Start
timeout.
b. If a Boot Data Start interrupt is received, the boot data is being received from
the card device. When the DMA engine is not in internal DMA controller
mode, the software driver can then initiate a data read from the controller
based on the rxdr interrupt bit in the rintsts register.
c. The software driver must terminate the boot process by instructing the
controller to send the SD/SDIO GO_IDLE_STATE (CMD0) command:
■ Reset the cmdarg register to 0.
■ Set the start_cmd bit in the cmd register to 1, and all other bits to 0.

d. At the end of a successful boot data transfer from the card, the following
e. In internal DMA controller mode:
The alternative boot operation for eMMC card devices is complete.
Alternative Boot Operation for MMC4.3 Cards

Removable MMC4.3 cards differ with respect to eMMC in that the controller is not
aware whether these cards support the boot mode of operation. Thus, the controller
must:
1. Discover these cards as it would discover MMC4.0/4.1/4.2 cards for the first time
2. Know the card characteristics
3. Decide whether to perform a boot operation or not
For these cards, the software driver must perform the following steps:
1. Discover the card as described in “Enumerated Card Stack” on page 11–37.
2. Read the MMC card device’s EXT_CSD registers and examine the following fields:
■ BOOT_PARTITION_ENABLE
■ BOOT_SIZE_MULT
■ BOOT_INFO
f For more information, refer to “Access to Boot Partition” in JEDEC Standard

No. JESD84-A43, as referenced in “References” on page 11–79.
3. If the host processor needs to perform a boot operation at the next power-up cycle,
it can manipulate the contents of the EXT_CSD registers in the MMC card device,
by using a SWITCH_FUNC command.
4. After this step, the software driver must power down the card by writing to the
pwren register.
5. From here on, use the same steps as in “Alternative Boot Operation for eMMC
Card Devices” on page 11–74.

SD/MMC Controller Address Map and Register Definitions
1 Ignore the EBE if it is generated during an abort scenario.
If a boot acknowledge error occurs, the boot acknowledge received

interrupt times out.
In internal DMA controller mode, the application needs to depend on the

descriptor close interrupt instead of the data done interrupt.
SD/MMC Controller Address Map and Register Definitions

link:
■ sdmmc
References
f The following industry specifications provide detailed information about the
standards and protocols implemented by the SD/MMC controller.
The following specifications are available on the JEDEC website (www.jedec.org):

■ JEDEC Standard No. 84-A441—Embedded MultiMediaCard (e•MMC) e•MMC/Card
Product Standard, High Capacity, including Reliable Write, Boot, Sleep Modes, Dual
Data Rate, Multiple Partitions Supports, Security Enhancement, Background Operation
and High Priority Interrupt (MMCA, 4.41)
■ JEDEC Standard No. 84-A44—Embedded MultiMediaCard (e•MMC) e•MMC/Card
Product Standard, High Capacity, including Reliable Write, Boot, Sleep Modes, Dual
Data Rate, Multiple Partitions Supports and Security Enhancement (MMCA, 4.4)
■ JEDEC Standard No. JESD84-A43—Embedded MultiMediaCard (eMMC) eMMC/Card
Product Standard, High Capacity, including Reliable Write, Boot, and Sleep Modes
(MMCA, 4.3)
The following specifications are available on the SD Association website
(www.sdcard.org):
■ Physical Layer Simplified Specification, Version 3.01—SD Specifications Part 1 Physical
Layer Simplified Specification Version 3.01
■ SDIO Simplified Specification Version 2.00—SD Specifications Part E1 SDIO Simplified
Specification Version 2.00



■ Added programming model section.
■ Reorganized programming information.
November 2012 1.1 ■ Added information about ECCs.
■ Added pin listing.
■ Updated clocks section.
January 2012 1.0 Initial release

12. Quad SPI Flash Controller
November 2012
cv_54012-1.2
cv_54012-1.2
The hardware processor system (HPS) provides a quad serial peripheral interface
(SPI) flash controller for access to serial NOR flash devices. The quad SPI flash
controller supports standard SPI flash devices as well as high-performance dual and
quad SPI flash devices. The quad SPI flash controller is based on Cadence® Quad SPI
Flash Controller (QSPI_FLASH_CTRL).
Features of the Quad SPI Flash Controller

The quad SPI flash controller supports the following features:
■ Single, dual, and quad I/O commands
■ Device frequencies up to 108 MHz
■ Direct access and indirect access modes
■ External direct memory access (DMA) controller support for indirect transfers
■ Configurable polarity, phase, and delay
■ Programmable write-protected regions
■ Local buffering with error correction code (ECC) logic for indirect transfers
■ Up to four devices
■ eXecute-In-Place (XIP) flash devices
semiconductor products to current specifications in accordance with Altera's standard warranty, but reserves the right to make changes to any products and ISO
services at any time without notice. Altera assumes no responsibility or liability arising out of the application or use of any information, product, or service 9001:2008
described herein except as expressly agreed to in writing by Altera. Altera customers are advised to obtain the latest version of device specifications before relying Registered

November 2012
Subscribe
12–2 Chapter 12: Quad SPI Flash Controller
Quad SPI Flash Controller Block Diagram and System Integration
Quad SPI Flash Controller Block Diagram and System Integration

Figure 12–1 shows the quad SPI flash controller block diagram.
Figure 12–1. Quad SPI Flash Controller Block Diagram
Quad SPI Flash Controller

DMA
DMA Peripheral DMA Peripheral
Controller Request Request Controller
Interface
SPI PHY
ECC
Signals Indirect
System TX
SRAM Access
Manager FIFO
Controller
Direct Flash
RX
Access Command
FIFO
Controller Generator SPI Flash
Data Slave Data Slave Device
Interface Controller Interface
L3 STIG
Interconnect
CSRs SPI
Control
Logic
L4 Peripheral Bus
The quad SPI controller consists of the following blocks and interfaces:
■ Register slave interface—Slave interface that provides access to the control and
status registers (CSRs)
■ Data slave controller—Slave interface and controller that provides the following
functionality:
■ Performs data transfers to and from the level 3 (L3) interconnect
■ Validates incoming accesses
■ Performs byte or half-word reordering
■ Performs write protection
■ Forwards transfer requests to direct and indirect controller
■ Direct access controller—provides memory-mapped slaves direct access to the
flash memory

Chapter 12: Quad SPI Flash Controller 12–3
Functional Description of the Quad SPI Flash Controller
■ Indirect access controller—provides higher-performance access to the flash

memory through local buffering and software transfer requests
■ Software triggered instruction generator (STIG)—generates flash commands
through the flash command register (flashcmd) and provides low-level access to
flash memory
■ Flash command generator—generates flash command and address instructions
based on instructions from the direct and indirect access controllers or the STIG
■ DMA peripheral request controller—issues requests to the DMA peripheral
request interface to communicate with the external DMA controller
■ SPI PHY—serially transfers data and commands to the external SPI flash devices

This section describes the functions of the quad SPI flash controller.
Overview
1 Terms used in this section are defined in detail in the sections that follow.
The quad SPI flash controller uses the register slave interface to select the operation
modes and configure the data slave interface for data transfers. The quad SPI flash
controller uses the data slave interface for direct and indirect accesses, and the register
slave interface for STIG operation and SPI legacy mode accesses.
Accesses to the data slave are forwarded to the direct or indirect access controller. If
the access address is within the configured indirect address range, the access is sent to
the indirect access controller.
Data Slave Interface

The quad SPI flash controller uses the data slave interface for direct, indirect, and SPI
legacy mode accesses. For information about these modes, refer to the following
sections.
The data slave is 32 bits wide. Byte, half-word, and word accesses are permitted.
For write accesses, only incrementing bursts are supported, and only of sizes 1, 4, 8,
and 16 transfers. For read accesses, all burst types and sizes are supported.

The quad SPI flash controller uses the register slave interface to configure the quad
SPI controller through the quad SPI configuration registers, and to access flash
memory under software control, through the flashcmd register in the STIG.

Direct Access Mode

In direct access mode, an access to the data slave triggers a read or write command to
the flash memory. To use the direct access mode, enable the direct access controller
with the enable direct access controller bit (endiracc) of the quad SPI configuration
register (cfg).
An external master, for example a processor, triggers the direct access controller with
a read or write operation to the data slave interface. The data slave exposes a 1 MB
window into the flash device. You can remap this window to any 1 MB location
within the flash device.
Figure 12–2 shows an example remapping.
Figure 12–2. Data Slave Remapping Example
Offsets
0x01000000
0x00300000
Map to Offset
0x00200000 0x00200000
0x00100000
Data
1 MB
Slave
Address Range Map to Offset 0
Access 0x00000000
Data Slave 16-MB Flash Memory
To remap the data slave to access other 1 MB regions of the flash device, enable
address remapping in the enable AHB address remapping field (enahbremap) of the
cfg register. All incoming data slave accesses remap to the offset specified in the
remap address register (remapaddr).
The 20 LSBs of incoming addresses are used for accessing the 1 MB region and the
higher bits are ignored.
1 The quad SPI controller does not issue any error status for accesses that lie outside the
connected flash memory space.
Indirect Access Mode

In indirect access mode, flash data is temporarily buffered in the quad SPI controller’s
SRAM. Software controls and triggers indirect accesses through the register slave
interface. The controller transfers data through the data slave interface.

Indirect Read Operation

An indirect read operation reads data from the flash memory, places the data into the
SRAM, and transfers the data to an external master through the data slave interface.
The indirect read operations are controlled by the following registers:
■ Indirect read transfer register (indrd)
■ Indirect read transfer watermark register (indrdwater)
■ Indirect read transfer start address register (indrdstaddr)
■ Indirect read transfer number bytes register (indrdcnt)
■ Indirect address trigger register (indaddrtrig)
These registers need to be configured prior to issuing indirect read operations. The
start address needs to be defined in the indrdstaddr register and the total number of
bytes to be fetched is specified in the indircnt register. Writing 1 to the start indirect
read bit (start) of the indrd register triggers the indirect read operation from the flash
memory to populate the SRAM with the returned data.
To read data from the flash device into the SRAM, an external master issues 32-bit
read transactions to the data slave interface. The address of the read access must be in
the indirect address range. You can configure the indirect address through the
indaddrtrig register. The external master can issue 32-bit reads until the last word of
an indirect transfer. On the final read, the external master may issue a 32-bit, 16-bit or
8-bit read to complete the transfer. If there are less than four bytes of data to read on
the last transfer, the external master can still issue a 32-bit read and the quad SPI
controller will pad the upper bits of the response data with zeros.
Assuming the requested data is present in the SRAM at the time the data slave read is
received by the quad SPI controller, the data is fetched from SRAM and the response
to the read burst is achieved with minimum latency. If the requested data is not
immediately present in the SRAM, the data slave interface enters a wait state until the
data has been read from flash memory into SRAM. Once the data has been read from
SRAM by the external master, the quad SPI controller frees up the associated resource
in the SRAM. If the SRAM is full, reads on the SPI interface are backpressured until
space is available in the SRAM. The quad SPI controller completes any current read
burst, waits for SRAM to free up, and issues a new read burst at the address where the
previous burst was terminated.
The processor can also use the SRAM fill level in the SRAM fill register (sramfill) to
control when data should be fetched from the SRAM.
Another alternative is to use the fill level watermark of the SRAM, which you
configure in the indrdwater register. When the SRAM fill level passes the watermark
level, the indirect transfer watermark interrupt is generated. You can disable the
watermark feature by writing zero to the indrdwater register.
For the final bytes of data read by the quad SPI controller and placed in the SRAM, if
the watermark level is greater than zero, the indirect transfer watermark interrupt is
generated even when the actual SRAM fill level has not risen above the watermark.
If the address of the read access is outside the range of the indirect trigger address,
one of the following actions occurs:
■ When direct access mode is enabled, the read uses direct access mode.

■ When direct access mode is disabled, the slave returns an error back to the
requesting master.
You can cancel an indirect operation by setting the cancel indirect read bit (cancel) of
the indrd register to 1. For more information, refer to “Indirect Read Operation” on
page 12–15.
Indirect Write Operation

An indirect write operation programs data from the SRAM to the flash memory. The
indirect write operations are controlled by the following registers:
■ Indirect write transfer register (indwr)
■ Indirect write transfer watermark register (indwrwater)
■ Indirect write transfer start address register (indwrstaddr)
■ Indirect write transfer number bytes register (indwrcnt)
■ indaddrtrig register
These registers need to be configured prior to issuing indirect write operations. The
start address needs to be defined in the indwrstaddr register and the total number of
bytes to be written is specified in the indwrcnt register. The start indirect write bit
(start) of the indwr register triggers the indirect write operation from the SRAM to
the flash memory.
To write data from the SRAM to the flash device, an external master issues 32-bit write
transactions to the data slave. The address of the write access must be in the indirect
address range. You can configure the indirect address through the indaddrtrig
register. The external master can issue 32-bit writes until the last word of an indirect
transfer. On the final write, the external master may issue a 32-bit, 16-bit or 8-bit write
to complete the transfer. If there are less than four bytes of data to write on the last
transfer, the external master can still issue a 32-bit write and the quad SPI controller
discards the extra bytes.
The SRAM size can limit the amount of data that the quad SPI controller can accept
from the external master. If the SRAM is not full at the point of the write access, the
data is pushed to the SRAM with minimum latency. If the external master attempts to
push more data to the SRAM than the SRAM can accept, the quad SPI controller
backpressures the external master with wait states. When the SRAM resource is freed
up by pushing the data from SRAM to the flash memory, the SRAM is ready to receive
more data from the external master. When the SRAM holds an equal or greater
number of bytes than the size of a flash page, or when the SRAM holds all the
remaining bytes of the current indirect transfer, the quad SPI controller initiates a
write operation to the flash memory.
The processor can also use the SRAM fill level, in the sramfill register, to control
when to write more data into the SRAM.

Alternatively, you can configure the fill level watermark of the SRAM in the
indwrwater register. When the SRAM fill level falls below the watermark level, an
indirect transfer watermark interrupt is generated to tell software to write the next
page of data to SRAM. Because the quad SPI controller initiates non-end-of-data
writes to the flash memory only when the SRAM contains a full flash page of data,
you must set the watermark level to a value greater than one flash page to avoid the
system stalling. You can disable the watermark feature by writing zero to the
indwrwater register.
If the address of the write access is outside the range of the indirect trigger address,
one of the following actions occurs:
■ When direct access mode is enabled, the write uses direct access mode.
■ When direct access mode is disabled, the slave returns an error back to the
requesting master.
You can cancel an indirect operation by setting the cancel indirect write bit (cancel) of
the indwr register to 1.
For more information, refer to “Indirect Write Operation” on page 12–16.
Consecutive Reads and Writes

It is possible to trigger two indirect operations at a time by triggering the start bit of
the indrd or indwr register twice in short succession. The second operation can be
triggered while the first operation is in progress. For example, software may trigger
an indirect read or write operation while an indirect write operation is in progress.
The corresponding start and count registers must be configured properly before
software triggers each transfer operation.
This approach allows for a short turnaround time between the completion of one
indirect operation and the start of a second operation. Any attempt to queue more
than two operations causes the indirect read reject interrupt to be generated.
Local Memory Buffer

The SRAM local memory buffer is a 128 by 32-bit (512 total bytes) memory and
includes support for ECC. The ECC logic provides outputs to notify the system
manager when single-bit correctable errors are detected (and corrected) and when
double-bit uncorrectable errors are detected. The ECC logic also allows the injection of
single- and double-bit errors for test purposes.
The SRAM has two partitions, with the lower partition reserved for indirect read
operations and the upper partition for indirect write operations. The size of the
partitions is specified in the SRAM partition register (srampart), based on 32-bit word
sizes. For example, to specify four bytes of storage, write the value 1. The value
written to the indirect read partition size field (addr) defines the number of entries
reserved for indirect read operations. For example, write the value 32 (0x20) to
partition the 128-entry SRAM to 32 entries (25%) for read usage and 96 entries (75%)
for write usage.

DMA Peripheral Request Controller

The DMA peripheral request controller is only used for the indirect mode of operation
where data is temporarily stored in the SRAM. The quad SPI flash controller uses the
DMA peripheral request interface to trigger the external DMA into performing data
transfers between memory and the quad SPI controller.
There are two DMA peripheral request interfaces, one for indirect reads and one for
indirect writes. The DMA peripheral request controller can issue two types of DMA
requests, single or burst, to the external DMA. The number of bytes for each single or
burst request is specified in the number of single bytes (numsglreqbytes) and number
of burst bytes (numburstreqbytes) fields of the DMA peripheral register (dmaper). The
DMA peripheral request controller splits the total amount of data to be transferred
into a number of DMA burst and single requests by dividing the total number of bytes
by the number of bytes specified in the burst request, and then dividing the remainder
by the number of bytes in a single request.
1 When programming the DMA controller, the burst request size must match the burst
request size set in the quad SPI controller to avoid quickly reaching an overflow or
underflow condition.
For indirect reads, the DMA peripheral request controller only issues DMA requests
after the data has been retrieved from flash memory and written to SRAM. The rate at
which DMA requests are issued depends on the watermark level. The indrdwater
register defines the minimum fill level in bytes at which the DMA peripheral request
controller can issue the DMA request. The higher this number is, the more data that
must be buffered in SRAM before the external DMA moves the data. When the SRAM
fill level passes the watermark level, the transfer watermark reached interrupt is
generated.
For example, consider the following conditions:
■ The total amount of data to be read using indirect mode is 256 bytes
■ The SRAM watermark level is set at 128 bytes
■ Software configures the burst type transfer size to 64 bytes
Under these conditions, the DMA peripheral request controller issues the first DMA
burst request when the SRAM fill level passes 128 bytes (the watermark level). The
DMA peripheral request controller triggers consecutive DMA burst requests as long
as there is sufficient data in the SRAM to perform burst type requests. In this example,
DMA peripheral request controller can issue at least two consecutive DMA burst
requests to transfer a total of 128 bytes. If there is sufficient data in the SRAM, the
DMA peripheral request controller requests the third DMA burst immediately.
Otherwise the DMA peripheral request controller waits for the SRAM fill level to pass
the watermark level again to trigger the next burst request. When the watermark level
is triggered, there is sufficient data in the SRAM to perform the third and fourth burst
requests to complete the entire transaction.

For the indirect writes, the DMA peripheral request controller issues DMA requests
immediately after the transfer is triggered and continues to do so until the entire
indirect write transfer has been transferred. The rate at which DMA requests are
issued depends on the watermark level. The indwrwater register defines the
maximum fill level in bytes at which the controller can issue the first DMA burst or
single request. When the SRAM fill level falls below the watermark level, the transfer
watermark reached interrupt is generated. When there is one flash page of data in the
SRAM, the quad SPI controller initiates the write operation from SRAM to the flash
memory.
Software can disable the DMA peripheral request interface with the endma field of the
cfg register. If a master other than the DMA performs the data transfer for indirect
operations, the DMA peripheral request interface must be disabled. By default, the
indirect watermark registers are set to zero, which means the DMA peripheral request
controller can issue DMA request as soon as possible.
f For more information about the HPS DMA controller, refer to the DMA Controller
STIG Operation
The STIG provides software a method to access the flash device registers directly. The
flashcmd register uses the following parameters to define the command to be issued
to the flash device:
■ Instruction opcode
■ Number of address bytes
■ Number of dummy bytes
■ Number of write data bytes
■ Write data
■ Number of read data bytes
The address is specified through the flash command address register (flashcmdaddr).
Once these settings have been specified, software can trigger the command with the
execute command field (execcmd) of the flashcmd register and wait for its completion
by polling the command execution status bit (cmdexecstat) of the flashcmd register. A
maximum of eight data bytes may be read from the flash command read data lower
(flashcmdrddatalo) and flash command read data upper (flashcmdrddataup)
registers or written to the flash command write data lower (flashcmdwrdatalo) and
flash command write data upper (flashcmdwrdataup) registers per command.
Commands issued through the STIG have a higher priority than all other read
accesses and therefore interrupt any read commands being requested by the direct or
indirect controllers. However, the STIG does not interrupt a write sequence that may
have been issued through the direct or indirect access controller. In these cases, it
might take a long time for the cmdexecstat bit of the flashcmd register indicates the
operation is complete.
1 Altera recommends using the STIG instead of the SPI legacy mode to access the flash
device registers and perform erase operations.

SPI Legacy Mode

SPI legacy mode allows software to access the internal TX FIFO and RX FIFO buffers
directly, thus bypassing the direct, indirect and STIG controllers. Software accesses the
TX FIFO and RX FIFO buffers by writing any value to any address through the data
slave while legacy mode is enabled. You can enable legacy mode with the legacy IP
mode enable bit (enlegacyip) of the cfg register.
Legacy mode allows the user to issue any flash instruction to the flash device, but
imposes a heavy software overhead in order to manage the fill levels of the FIFO
buffers effectively. The legacy SPI mode is bidirectional in nature, with data
continuously being transferred both directions while the chip select is enabled. If the
driver only needs to read data from the flash device, dummy data must be written to
ensure the chip select stays active, and vice versa for write transactions.
For example, to perform a basic read of four bytes to a flash device that has three
address bytes, software must write a total of eight bytes to the TX FIFO buffer. The
first byte would be the instruction opcode, the next three bytes are the address, and
the final four bytes would be dummy data to ensure the chip select stays active while
the read data is returned. Similarly, because eight bytes were written to the TX FIFO
buffer, software should expect eight bytes to be returned in the RX FIFO buffer. The
first four bytes of this would be discarded, leaving the final four bytes holding the
data read from the device.
Because the TX FIFO and RX FIFO buffers are four bytes deep each, software must
maintain the FIFO buffer levels to ensure the TX FIFO buffer does not underflow and
the RX FIFO buffer does not overflow. Interrupts are provided to indicate when the fill
levels pass the watermark levels, which are configurable through the TX threshold
register (txtresh) and RX threshold register (rxtresh).
Configuring the Flash Device

For read and write accesses, software must initialize the device read instruction
register (devrd) and the device write instruction register (devwr). These registers
include fields to initialize the instruction opcodes that should be used as well as the
instruction type, and whether the instruction uses single, dual or quad pins for
address and data transfer. To ensure the quad SPI controller can operate from a reset
state, the opcode registers reset to opcodes compatible with single I/O flash devices.
The quad SPI flash controller uses the instruction transfer width field (instwidth) of
the devrd register to set the instruction transfer width for both reads and writes. There
is no instwidth field in the devwr register. If instruction type is set to dual or quad
mode, the address transfer width (addrwidth) and data transfer width (datawidth)
fields of both registers are redundant because the address and data type is based on
the instruction type. Thus, software can support the less common flash instructions
where the opcode, address, and data are sent on two or four lanes. For most

instructions, the opcodes are sent serially to the flash device, even for dual and quad
instructions. One of the flash devices that supports instructions that can send the
opcode over two or four lanes is the Micron N25Q128. For reference, Table 12–1 shows
how software should configure the quad SPI controller for each specific read and
write instruction supported by the Micron N25Q128 device.
Table 12–1. Quad SPI Configuration for Micron N25Q128 Device

Lanes Used Lanes Used
Lanes Used instwidth addrwidth datawidth
Instruction to Send to Send
By Opcode Value Value Value
Address Data
Read Instructions
Read 1 1 1 0 0 0
Fast read 1 1 1 0 0 0
Dual output fast read (DOFR) 1 1 2 0 0 1
Dual I/O fast read (DIOFR) 1 2 2 0 1 1
Quad output fast read (QOFR) 1 1 4 0 0 2
Quad I/O fast read (QIOFR) 1 4 4 0 2 2
Dual command fast read (DCFR) 2 2 2 1 Don’t care Don’t care
Quad command fast read (QCFR) 4 4 4 2 Don’t care Don’t care
Write Instructions
Page program 1 1 1 0 0 0
Dual input fast program (DIFP) 1 1 2 0 0 1
Dual input extended fast program
1 2 2 0 1 1
(DIEFP)
Quad input fast program (QIFP) 1 1 4 0 0 2
Quad input extended fast program
1 4 4 0 2 2
(QIEFP)
Dual command fast program
2 2 2 1 Don’t care Don’t care
(DCFP)
Quad command fast program
4 4 4 2 Don’t care Don’t care
(QCFP)
XIP Mode
The quad SPI controller supports XIP mode, if the flash devices support XIP mode.
Depending on the flash device, XIP mode puts the flash device in read-only mode,
reducing command overhead.
The quad SPI controller must instruct the flash device to enter XIP mode by sending
the mode bits. When the enter XIP mode on next read bit (enterxipnextrd) of the cfg
register is set to 1, the quad SPI controller and the flash device are ready to enter XIP
mode on the next read instruction. When the enter XIP mode immediately bit
(enterxipimm) of the cfg register is set to 1, the quad SPI controller and flash device
enter XIP mode immediately.
When the enterxipnextrd or enterxipimm bit of the cfg register is set to 0, the quad
SPI controller and flash device exit XIP mode on the next read instruction.
For more information, refer to “XIP Mode Operations” on page 12–17.

Write Protection
You can program the controller to write protect a specific region of the flash device.
The protected region is defined as a set of blocks, specified by a starting and ending
block. Writing to an area of protected flash region memory generates an error and
triggers an interrupt.
You define the block size by specifying the number of bytes per block through the
number of bytes per block field (bytespersubsector) of the device size register
(devsz). The lower write protection register (lowwrprot) specifies the first flash block
in the protected region. The upper write protection register (uppwrprot) specifies the
last flash block in the protected region.
The write protection enable bit (en) of the write protection register (wrprot) enables
and disables write protection. The write protection inversion bit (inv) of the wrprot
register flips the definition of protection so that the region specified by lowwrprt and
uppwrprt is unprotected and all flash memory outside that region is protected.
Data Slave Sequential Access Detection

The quad SPI flash controller detects sequential accesses to the data slave interface by
comparing the current access with the previous access. An access is sequential when it
meets the following conditions:
■ The address of the current access sequentially follows the address of the previous
access.
■ The direction of the current access (read or write) is the same as previous access.
■ The size of the current access (byte, half-word, or word) is the same as previous
access.
When the access is detected as nonsequential, the sequential access to the flash device
is terminated and a new sequential access begins. Altera recommends accessing the
data slave sequentially. Sequential access has less command overhead, and therefore,
increases data throughput.
Clocks
There are two clock inputs to the quad SPI controller (l4_mp_clk and qspi_clk) and
one clock output (sclk_out). The quad SPI flash controller uses the l4_mp_clk clock to
clock the data slave transfers and register slave accesses. The qspi_clk clock is the
reference clock for the quad SPI controller and is used to serialize the data and drive
the external SPI interface. The sclk_out clock is the clock source for the connected
flash devices.
The qspi_clk clock must be greater than two times the l4_mp_clk. The sclk_out clock
is derived by dividing down the qspi_clk clock by the baud rate divisor field
(bauddiv) of the cfg register.
Device Handbook.

Resets
A single reset signal (qspi_flash_rst_n) is provided as an input to the quad SPI
controller. The reset manager drives the signal on a cold or warm reset.
Device Handbook.
Interrupts
All interrupt sources are combined to create a single level-sensitive, active-high
interrupt (qspi_intr_in). Software can determine the source of the interrupt by
reading the interrupt status register (irqstat). By default, the interrupt source is
cleared when software writes the interrupt status register. The interrupts are
individually maskable through the interrupt mask register (irqmask). Table 12–2 lists
the interrupt sources in the irqstat register.
Table 12–2. Interrupt Sources in the irqstat Register

Interrupt Source Description
When 0, no underflow has been detected. When 1, the data
slave write data is being supplied too slowly. This situation
can occur when data slave write data is being supplied too
Underflow detected
slowly to keep up with the requested write operation This
bit is reset only by a system reset and cleared to zero only
when the register is written to.
Indirect operation complete The controller has completed a triggered indirect operation.
An indirect operation was requested but could not be
Indirect read reject accepted because two indirect operations are already in the
queue.
Protected area write attempt A write to a protected area was attempted and rejected.
An illegal data slave access has been detected. Data slave
wrapping bursts and the use of split and retry accesses can
Illegal data slave access detected cause this interrupt. It is usually an indicator that soft
masters in the FPGA fabric are attempting to access the
HPS in an unsupported way.
Transfer watermark reached The indirect transfer watermark level has been reached.
This condition occurs only in legacy SPI mode. When 0, no
overflow has been detected. When 1, an over flow to the RX
FIFO buffer has occurred. This bit is reset only by a system
Receive overflow
reset and cleared to zero only when this register is written
to. If a new write to the RX FIFO buffer occurs at the same
time as a register is read, this flag remains set to 1.
This condition occurs only in legacy SPI mode. When 0, the
TX FIFO not full
TX FIFO buffer is full. When 1, the TX FIFO buffer is not full.
TX FIFO full
TX FIFO buffer is not full. When 1, the TX FIFO buffer is full.
RX FIFO not empty RX FIFO buffer is empty. When 1, the RX FIFO buffer is not
empty.

Quad SPI Flash Controller Programming Model
Table 12–2. Interrupt Sources in the irqstat Register

Interrupt Source Description
RX FIFO full
RX FIFO buffer is not full. When 1, the RX FIFO buffer is full.
Indirect Read Partition of SRAM is full and unable to
Indirect read partition overflow
immediately complete indirect operation
Interface Signals
The quad SPI controller provides four chip select outputs to allow control of up to
four external quad SPI flash devices. The outputs serve different purposes depending
on whether the device is used in single, dual, or quad operation mode. Table 12–3 lists
the I/O pin use of the quad SPI controller interface signals for each operation mode.
Table 12–3. Interface Signals

Signal Mode Direction Function
Single Output Data output 0
data[0]
Dual or quad Bidirectional Data I/O 0
Single Input Data input 0
data[1]
Dual or quad Bidirectional Data I/O 1
Single or dual Output Active low write protect
data[2]
Quad Bidirectional Data I/O 2
Single, dual,
data[3] Bidirectional Data I/O 3
or quad
ss_n[0] Active low slave select 0
Single, dual,
ss_n[2] Output Active low slave select 2
or quad
sclk Serial clock

This section describes the programming model for the quad SPI controller.
Setting Up the Quad SPI Flash Controller

The following steps describe how to set up the quad SPI controller:
1. Wait until any pending operation has completed.
2. Disable the quad SPI controller with the quad SPI enable field (en) of the cfg
register.
3. Update the instwidth field of the devrd register with the instruction type you
wish to use for indirect and direct writes and reads.
4. If mode bit enable bit (enmodebits) of the devrd register is enabled, update the
mode bit register (modebit).

5. Update the devsz register as needed. Parts or all of this register might have been
updated after initialization. The number of address bytes is a key configuration
setting required for performing reads and writes. The number of bytes per page is
required for performing any write. The number of bytes per device block is only
required if the write protect feature is used.
6. Update the device delay register (delay). This register allows the user to adjust
how the chip select is driven after each flash access. Each device may have
different timing requirements. If the serial clock frequency is increased, these
timing requirements become more critical. The numbers specified in this register
are based on the period of the qspi_clk clock.
For example, an some devices need 50 ns minimum time before the slave select can
be reasserted after it has been deasserted. When the device is operating at
100 MHz, the clock period is 10 ns, so 40 ns extra is required. If the qspi_clk clock
is running at 400 MHz (2.5 ns period), specify a value of at least 16 to the clock
delay for chip select deassert field (nss) of the delay register.
7. Update the remapaddr register as needed. This register only affects direct access
mode.
8. Set up and enable the write protection registers (wrprot, lowwrprot, and
uppwrprot), when write protection is required.
9. Enable required interrupts though the irqmask register.
10. Set up the bauddiv field of the cfg register to define the required clock frequency
of the target device.
11. Update the read data capture register (rddatacap) as needed. This register delays
when the read data is captured and can help when the read data path from the
device to the quad SPI controller is long and the device clock frequency is high.
12. Enable the quad SPI controller with the en field of the cfg register.
Indirect Read Operation

The following steps describe the general software flow to set up the quad SPI
controller for indirect read operation with the DMA disabled:
1. Perform the steps described in “Setting Up the Quad SPI Flash Controller” on
page 12–14.
2. Set the flash memory start address in the indrdstaddr register.
3. Set the number of bytes to be transferred in the indrdcnt register.
4. Set the indirect transfer trigger address in the indaddrtrig register.
5. Set up the required interrupts through the irqmask register.
6. If the watermark level is used, set the SRAM watermark level through the
indrdwater register.
7. Start the indirect read operation by setting the start field of the indrd register to 1.
8. Either use the watermark level interrupt or poll the SRAM fill level in the sramfill
register to determine when there is sufficient data in the SRAM.

9. Issue a read transaction to the indirect address to access the SRAM. Repeat step 8 if
more read transactions are needed to complete the indirect read transfer.
10. Either use the indirect complete interrupt to determine when the indirect read
operation has completed or poll the completion status of the indirect read
operation through the indirect completion status bit (ind_ops_done_status) of the
indrd register.
controller for indirect read operation with the DMA enabled:
page 12–14.
2. Set the flash memory start address in the indrdstaddr register.
3. Set the number of bytes to be transferred in the indrdcnt register.
5. Set the number of bytes for single and burst type DMA transfers in the dmaper
register.
6. Optionally set the SRAM watermark level in the indrdwater register to control the
rate DMA requests are issued.
7. Start an indirect read access by setting the start field of the indrd register to 1.
8. Either use the indirect complete interrupt to determine when the indirect read
operation has completed or poll the completion status of the indirect read
operation through the ind_ops_done_status field of the indrd register.
Indirect Write Operation

controller for indirect write operation with the DMA disabled:
page 12–14.
2. Set the flash memory start address in the indwrstaddr register.
3. Set up the number of bytes to be transferred in the indwrcnt register.
5. Set up the required interrupts through the interrupt mask register (irqmask).
6. Optionally set the SRAM watermark level in the indwrwater register to control the
rate DMA requests are issued. The value set must be greater than one flash page.
7. Start the indirect write operation by setting the start field of the indwr register to
1.
8. Either use the watermark level interrupt or poll the SRAM fill level in the sramfill
register to determine when there is sufficient space in the SRAM.
9. Issue a write transaction to the indirect address to write one flash page of data to
the SRAM. Repeat step 8 if more write transactions are needed to complete the
indirect write transfer. The final write may be less than one page of data.

controller for indirect write operation with the DMA enabled:
page 12–14.
2. Set the flash memory start address in the indwrstaddr register.
3. Set the number of bytes to be transferred in the indcnt field of the indwr register.
5. Set the number of bytes for single and burst type DMA transfers in the dmaper
register.
6. Optionally set the SRAM watermark level in the indwrwater register to control the
rate DMA requests are issued. The value set must be greater than one flash page.
7. Start the indirect write access by setting the start field of the indirwr register to 1.
8. Either use the indirect complete interrupt to determine when the indirect write
operation has completed or poll the completion status of the indirect write
operation through the ind_ops_done_status field of the indwr register.
XIP Mode Operations

This section describes entering and exiting XIP mode. XIP mode is supported in most
SPI flash devices. However, flash device manufacturers do not use a consistent
standard approach. Most use signature bits that are sent to the device immediately
following the address bytes. Some devices use signature bits and also require a flash
device configuration register write to enable XIP mode.
Entering XIP Mode

The following sections describe the software steps to enter XIP mode for various types
of flash devices.
Micron Quad SPI Flash Devices with Support for Basic-XIP

To enter XIP mode in a Micron quad SPI flash device with support for Basic-XIP,
perform the following steps:
1. Save the values in the mode bits, if you intend to restore them upon exit.
2. Disable the direct access controller and indirect access controller to ensure no new
read or write accesses are sent to the flash device.
3. Set the XIP mode bits in the modebit register to 0x80.
4. Enable the quad SPI controller's XIP mode by setting the enterxipnextrd bit of the
cfg register to 1.
5. Re-enable the direct access controller and, if required, the indirect access controller.
Micron Quad SPI Flash Devices without Support for Basic-XIP

To enter XIP mode in a Micron quad SPI flash device without support for Basic-XIP,
perform the following steps:

read or write accesses will be sent to the flash device.
3. Ensure XIP mode is enabled in the flash device by setting the volatile
configuration register (VCR) bit 3 to 1. Use the flashcmd register to issue the VCR
write command.
5. Enable the quad SPI controller’s XIP mode by setting the enterxipnextrd bit of the
cfg register to 1.
Winbond Quad SPI Flash Devices

To enter XIP mode in a Winbond quad SPI flash device, perform the following steps:
cfg register to 1.
Spansion Quad SPI Flash Devices

To enter XIP mode a Spansion quad SPI flash device, perform the following steps:
3. Set the XIP mode bits in the modebit register to 0xA0.
cfg register to 1.
Exiting XIP Mode

To exit XIP mode, perform the following steps:
2. Restore the mode bits to the values before entering XIP mode, depending on the
flash device and manufacturer.
3. Set the enterxipnextrd bit of the cfg register to 0.
The flash device must receive a read instruction before it can disable its internal XIP
mode state. Thus, XIP mode internally stays active until the next read instruction is
serviced. Ensure that XIP mode is disabled before the end of any read sequence.

Quad SPI Flash Controller Address Map and Register Definitions
XIP Mode at Power on Reset

Some flash devices can be XIP-enabled as a nonvolatile configuration setting,
allowing the flash device to enter XIP mode at power-on reset (POR) without software
intervention. Software cannot discover the XIP state at POR through flash status
register reads because an XIP-enabled flash device can only be accessed through the
XIP read operation. If you known the device will enter XIP mode at POR, have your
initial boot software configure the modebit register and set the enterxipimm bit of the
cfg register to 1.
If you do not known in advance whether or not the device will enter XIP mode at
POR, have your initial boot software issue an XIP mode exit command through the
flashcmd register, then follow the steps in “Entering XIP Mode” on page 12–17.
Software must be aware of the mode bit requirements of the device, because XIP
mode entry and exit varies by device.
Quad SPI Flash Controller Address Map and Register Definitions

■ qspiregs
■ qspidata


Added block diagram and system integration, functional description, programming model,
May 2012 1.1
and address map and register definitions sections.


Section VI. Peripherals

■ Chapter 13, FPGA Manager
■ Chapter 14, System Manager
■ Chapter 15, Scan Manager
■ Chapter 16, DMA Controller
■ Chapter 17, Ethernet Media Access Controller
■ Chapter 18, USB 2.0 OTG Controller
■ Chapter 19, SPI Controller
■ Chapter 20, I2C Controller
■ Chapter 21, UART Controller
■ Chapter 22, General-Purpose I/O Interface
■ Chapter 23, Timer
■ Chapter 24, Watchdog Timer
■ Chapter 25, CAN Controller

VI–2 Section VI: Peripherals

13. FPGA Manager
November 2012
cv_54013-1.3
cv_54013-1.3
The FPGA manager in the hard processor system (HPS) manages and monitors the
FPGA portion of the system on a chip (SoC) FPGA device. The FPGA manager can
configure the FPGA fabric from the HPS, monitor the state of the FPGA, and drive or
sample signals to or from the FPGA fabric.
Features of the FPGA Manager

The FPGA manager provides the following functionality and features:
■ Full configuration and partial reconfiguration of the FPGA portion of the SoC
device
■ Drives 32 general-purpose output signals to the FPGA fabric
■ Receives 32 general-purpose input signals from the FPGA fabric
■ Receives two boot handshaking input signals from the FPGA fabric (used when
the HPS boots from the FPGA)
■ Monitors the FPGA configuration and power status
■ Generates interrupts based on the FPGA status changes
■ Can reset the FPGA

November 2012
Subscribe
13–2 Chapter 13: FPGA Manager
FPGA Manager Block Diagram and System Integration
FPGA Manager Block Diagram and System Integration

Figure 13–1 shows a block diagram of the FPGA manager.
Figure 13–1. FPGA Manager Block Diagram
FPGA Manager FPGA Portion
Configuration Control Block

L3 Data DCLK
Slave
Interconnect Block DATA[31:0]
Interface
MPU nCONFIG
IRQ nSTATUS
CONF_DONE
INIT_DONE
CONFIG_IO Mode
MSEL
Control nCE
Block nCONFIG
nSTATUS
CONF_DONE
PR_REQUEST
L4 Master Peripheral Bus
nCONFIG
nSTATUS
Register CONF_DONE
Slave INIT_DONE
Monitor CRC_ERROR
Interface CVP_CONF_DONE
Block PR_READY
PR_ERROR
PR_DONE
FPGA Fabric
h2f_gp[31:0]
Fabric
f2h_gp[31:0]
I/O
Block f2h_boot_from_fpga_on_failure
f2h_boot_from_fpga_ready
The register slave interface connects to the level 4 (L4) master peripheral bus for
control and status register (CSR) access. The configuration slave interface connects to
the level 3 (L3) interconnect for the microprocessor unit (MPU) subsystem or other
masters to write the FPGA configuration image to the FPGA control block (CB) when
configuring the FPGA portion of the SoC device.
The general-purpose I/O and boot handshaking input interfaces connect to the FPGA
fabric. The FPGA manager also connects to the FPGA CB signals to monitor and
control the FPGA portion of the device.
The FPGA manager consists of the following blocks:
■ Configuration slave interface—accepts and transfers the configuration image to
the data interface.
■ Register slave interface—accesses the CSRs in the FPGA manager.
■ Data—accepts the FPGA configuration image from the configuration slave
interface and sends it to the FPGA CB.
■ Control—controls the FPGA CB.

Chapter 13: FPGA Manager 13–3
Functional Description of the FPGA Manager
■ Monitor—monitors the configuration signals in the FPGA CB and sends interrupts

to the MPU subsystem.
■ Fabric I/O—reads and writes signals from or to the FPGA fabric.
FPGA Manager Building Blocks

The FPGA manager has the following blocks to monitor the signals coming from the
FPGA portion of the device.
Fabric I/O
The fabric I/O block contains the following registers to allow simple low-latency
communication between the HPS and the FPGA fabric:
■ General-purpose input register (gpi)
■ General-purpose output register (gpo)
■ Boot handshaking input register (misci)
These registers are only valid when the FPGA is in user mode. Reading from these
registers while the FPGA is not in user mode provides undefined data.
The 32 general-purpose input signals from the FPGA fabric are read by reading the
gpi register using the register slave interface. The 32 general-purpose output signals
to the FPGA fabric are generated from writes to the gpo register. For more information
about FPGA manager registers, refer to “FPGA Manager Address Map and Register
Definitions” on page 13–7.
The boot handshaking input signals from the FPGA fabric are read by reading the
misci register. The f2h_boot_from_fpga_ready signal indicates to the boot ROM
when logic in the FPGA fabric is ready to accept configuration interface requests from
the HPS-to-FPGA bridge when the boot ROM is booting from the FPGA. The
f2h_boot_from_fpga_on_failure signal serves as a fallback in the event that the boot
ROM code fails to boot from the primary boot flash device. In this case, the boot ROM
code checks these two handshaking signals to determine if it should use the boot code
hosted in the FPGA memory as the next stage in the boot process.
There is no interrupt support for this block.
Monitor
The monitor block is an instance of the Synopsys® DesignWare® GPIO IP
(DW_apb_gpio), which is a separate instance of the IP that comprises the three HPS
GPIO interfaces. The monitor block connects to the configuration signals in the FPGA.
This block monitors key signals related to FPGA configuration such as INIT_DONE,
CRC_ERROR, and PR_DONE. Software configures the monitor block through the register
slave interface, and can either poll FPGA signals or be interrupted. The mon address
map within the FPGA manager register address map contains the monitor registers.
For more information about FPGA manager registers, refer to “FPGA Manager
Address Map and Register Definitions” on page 13–7.

You can program the FPGA manager to treat any of the monitor signals as interrupt
sources. Independent of the interrupt source type, the monitor block always drives an
active-high level interrupt to the MPU. Each interrupt source can be of the following
types:
■ Active-high level
■ Active-low level
■ Rising edge
■ Falling edge
FPGA Configuration
You can configure the FPGA using an external device or through the HPS. This section
only covers configuring the FPGA through the HPS.
f For information about configuring the FPGA using an external device, refer to the
Configuration, Design Security, and Remote System Upgrades in Cyclone V Devices chapter
The FPGA CB uses the FPGA mode select (MSEL) pins to determine which
configuration scheme to use. The MSEL pins must be tied to the appropriate values for
the configuration scheme. Table 13–1 lists supported MSEL values when the FPGA is
configured by the HPS.
HPS software sets the clock-to-data ratio field (cdratio) and configuration data width
bit (cfgwdth) in the control register (ctrl) to match the MSEL pins. The cdratio field
and cfgwdth bit must be set before the start of configuration.
The FPGA manager connects to the configuration logic in the FPGA portion of the
device using a mode similar to how external logic (for example, MAX II or an
intelligent host) configures the FPGA in fast passive parallel (FPP) mode. FPGA
configuration through the HPS supports all the capabilities of FPP mode, including
the following items:
■ FPGA configuration
■ Partial FPGA reconfiguration
■ FPGA I/O configuration, followed by PCI Express® (PCIe®) configuration of the
remainder of FPGA
■ External single event upset (SEU) scrubbing
■ Decompression
■ Advanced Encryption Standard (AES) encryption
■ FPGA DCLK clock used for initialization phase clock
1 The FPGA manager supports a data width of 32 or 16 bits. When configuring the
FPGA fabric from the HPS, Altera recommends that you always set the data width to
32 bits. For partial reconfiguration, the 16-bit data width is the only option.

Table 13–1 lists the supported configuration schemes and their respective MSEL and
control register settings when the HPS configures the FPGA.
Table 13–1. Configuration Schemes for FPGA Configuration by the HPS

Design
Configuration Compression Supports Partial
Security POR Delay (2) MSEL[4..0] (3) cfgwdth cdratio
Scheme Feature Reconfiguration
Feature
Fast 00000 0 1 Yes
Disabled AES Disabled
Standard 00100 0 1 No
Fast 00001 0 2 Yes
FPP ×16 Disabled AES Enabled
Fast 00010 0 4 Yes
Enabled Optional (1)
Fast 01000 1 1 No
Disabled AES Disabled
Fast 01001 1 4 No
FPP ×32 Disabled AES Enabled
Fast 01010 1 8 No
Enabled Optional (1)
(1) You can select to enable or disable this feature.
(2) For information about POR delay, refer to the Configuration, Design Security, and Remote System Upgrades in Cyclone V Devices chapter in
(3) Other MSEL values are allowed when the FPGA is configured from a non-HPS source. For information, refer to the Configuration, Design
Security, and Remote System Upgrades in Cyclone V Devices chapter in volume 1 of the Cyclone V Device Handbook.
Configuring the FPGA portion of the SoC device comprises the following phases:
1. Power up phase
2. Reset phase
3. Configuration phase
4. Initialization phase
5. User mode
Power Up Phase
In this phase, the VCC is ramping up and has yet to reach normal levels. This phase
completes when the on-chip voltage detector determines that the VCC has reached
normal levels.
Reset Phase
The FPGA manager resets the FPGA portion of the SoC device when the FPGA
configuration signal (nCONFIG) is driven low. The HPS configures the FPGA by writing
a 1 to the nconfigpull bit of the ctrl register. This action causes the FPGA portion of
the device to reset and perform the following actions:
1. Clear the FPGA configuration RAM bits

2. Tri-state all FPGA user I/O pins

3. Pull the nSTATUS and CONF_DONE pins low
4. Use the FPGA CB to read the values of the MSEL pins to determine the
configuration scheme
The nconfigpull bit of the ctrl register needs to be set to 0 when the FPGA has
successfully entered the reset phase. Setting the bit releases the FPGA from the reset
phase and transitions to the configuration phase.
1 You must set the cdratio and cfgwdth bits of the ctrl register appropriately before
the FPGA enters the reset phase.
Configuration Phase
To configure the FPGA using the HPS, software sets the axicfgen bit of the ctrl
register to 1. Software then sends configuration data to the FPGA by writing data to
the write data register (data) in the FPGA manager module configuration data
address map. Software polls the CONF_DONE pin by reading the gpio_instatus register
to determine if the FPGA configuration is successful. When configuration is
successful, software sets the axicfgen bit of the ctrl register to 0. The FPGA user I/O
pins are still tri-stated in this phase.
f For more information about configuring the FPGA through the HPS, refer to the
Booting and Configuration appendix in volume 3 of the Cyclone V Device Handbook.
After successfully completing the configuration phase, the FPGA transitions to the
initialization phase. To delay configuring the FPGA, set the confdonepull bit of the
ctrl register to 1.
Initialization Phase
In this phase, the FPGA prepares to enter user mode. The internal oscillator in the
FPGA portion of the device is the default clock source for the initialization phase.
Alternatively, the configuration image can specify the CLKUSR or the DCLK pins as the
clock source. The alternate clock source controls when the FPGA enters user mode.
If DCLK is selected as the clock source, software uses the DCLK count (dclkcnt) register
to drive DCLK pulses to the FPGA. Writing to the cnt field of the dclkcnt register
triggers the FPGA manager to generate the specified number of DCLK pulses. When all
of the DCLK pulses have been sent, the dcntdone bit of the DCLK status (dclkstat)
register is set to 1. Software polls the dcntdone bit to know when all of the DCLK pulses
have been sent.
1 Before another write to the dclkcnt register, software needs to write a value of 1 to the
dcntdone bit to clear the done state.
The FPGA user I/O pins are still tri-stated in this phase. When the initialization phase
completes, the FPGA releases the optional INIT_DONE pin and an external resistor
pulls the pin high.

FPGA Manager Address Map and Register Definitions
User Mode
The FPGA enters the user mode after exiting the initialization phase. The FPGA user
I/O pins are no longer tri-stated in this phase and the configured soft logic in the
FPGA becomes active.
The FPGA remains in user mode until the nCONFIG pin is driven low. If the nCONFIG
pin is driven low, the FPGA reenters the reset phase. The internal oscillator is disabled
in user mode, but is enabled as soon as the nCONFIG pin is driven low.
f For more information about configuring the FPGA through the HPS, refer to the
Booting and Configuration appendix in volume 3 of the Cyclone V Device Handbook.
f For more information about configuring the FPGA in general, refer to

theConfiguration, Design Security, and Remote System Upgrades in Cyclone V Devices
Clock
The FPGA manager has two clock input signals which are asynchronous to each other.
The clock manager generates these two clocks:
■ cfg_clk—the configuration slave interface clock input and also the DCLK output
reference for FPGA configuration. Enable this clock in the clock manager only
when configuration is active or when the configuration slave interface needs to
respond to master requests.
■ l4_mp_clk—the register slave interface clock.
f For more information about the clock manager, refer to the Clock Manager chapter in
Reset
The FPGA manager has one reset signal. The reset manager drives this signal to FPGA
manager on a cold or warm reset. All distributed reset signals in the FPGA manager
are asserted asynchronously at the same time and de-asserted synchronously to their
associated clocks.
FPGA Manager Address Map and Register Definitions

■ fpgamgrregs
■ fpgamgrdata



June 2012 1.2 Updated the FPGA configuration section.
■ Updated the configuration schemes table.
May 2012 1.1 ■ Updated the FPGA configuration section.

14. System Manager
November 2012
cv_54014-1.2
cv_54014-1.2
The system manager in the hard processor system (HPS) contains memory-mapped
control and status registers (CSRs) and logic to control system level functions as well
as other modules in the HPS.
The system manager connects to the following modules in the HPS:
■ Watchdog timers
■ Ethernet media access controllers (EMAC0 and EMAC1)
■ Direct memory access (DMA) controller
■ USB 2.0 On-The-Go (OTG) controllers (USB0 and USB1)
■ Controller area network (CAN) controllers
■ NAND flash controller
■ Secure Digital/MultiMediaCard (SD/MMC) controller
■ Quad serial peripheral interface (SPI) flash controller
■ Microprocessor unit (MPU) subsystem
Features of the System Manager

Software accesses the CSRs in the system manager to control and monitor various
functions in other HPS modules that require external control signals. The system
manager connects to these modules to perform the following functions:
■ Sends pause signals to pause the watchdog timers when the processors in the
MPU subsystem are in debug mode.
■ Freezes the I/O pins after the HPS comes out of cold reset and during serial
configuration.
■ Selects the EMAC level 3 (L3) master signal options.
■ Selects the SD/MMC controller clock options and L3 master signal options.
■ Selects the NAND flash controller bootstrap options and L3 master signal options.
■ Selects USB controller L3 master signal options.
■ Selects whether the CAN controller or the FPGA fabric can issue requests to four
of the DMA controller peripheral request interfaces.
■ Provides control over the DMA security settings when the HPS exits from reset.
■ Provides boot source and clock source information that can be read during the
boot process.

November 2012
Subscribe
14–2 Chapter 14: System Manager
System Manager Block Diagram and System Integration
■ Sends error correction code (ECC) enable signals to all HPS modules with ECC-
protected RAM.
■ Provides the capability to inject errors during testing in modules with ECC-
protected RAM.
■ Controls I/O pin multiplexing in the HPS.
■ Provides information about device-specific HPS options such as the number of
processor cores and the presence of CAN controllers.
■ Triggers parity failures during testing in parity-protected RAMs in the MPU
subsystem.
■ Allows the HPS to access the FPGA JTAG controller through the scan manager for
testing purposes.
System Manager Block Diagram and System Integration

The system manager has a slave interface which connects to the level 4 (L4) bus. The
CSRs interface connects to signals in the FPGA and also to other HPS modules.
Figure 14–1 shows a block diagram of the system manager.
Figure 14–1. System Manager Block Diagram
System Manager
Pause
L4 Watchdog Timers
Watchdog
Debug Pause Debug Status

MPU
Parity Injection Logic with

Parity RAM
ECC & Parity
Generic Interrupt
Interrupts
Controller
Register
Slave Memory-Mapped
Interface CSRs Control Signals
Other Modules (1)
ECC Status and
Control Signals
Modules with
HPS I/O Freeze ECC RAM
Control Controller
HPS CSEL pins FPGA JTAG Control
HPS BSEL pins

FPGA
FPGA Control
Fabric Block

(1) Refer to the list of modules on the following page.

Chapter 14: System Manager 14–3
Functional Description of the System Manager
The system manager consists of the following blocks:

■ CSRs—provide the following functionality:
■ Provide memory-mapped access to control signals for the following HPS
modules:
■ EMACs
■ Debug core
■ Reset manager
■ SD/MMC controller
■ NAND controller
■ SPI masters
■ Quad SPI controller
■ USB controllers
■ DMA controller
■ On-chip RAM
■ CAN controllers
■ Route ECC and parity interrupts to the MPU
■ Store status information received from other HPS modules
■ Controls multiplexing between the FPGA JTAG pins and the scan manager
■ Register slave interface—provides connected masters access to the CSRs in the
system manager.
■ Watchdog debug pause—accepts the debug mode status from the MPU subsystem
and pauses the L4 watchdog timers.
■ Freeze controller—responsible for placing the HPS I/O pins into a safe state so
that they can be configured by software.

This section describes the functional operation of the system manager. The system
manager serves the following purposes:
■ Provides software access to boot configuration and system information
■ Provides software access to control and status signals in other HPS modules
■ Enables and controls ECC and parity in HPS modules
■ Provides freeze signals to the HPS-configurable I/O pins during configuration
■ Enables and disables HPS interfaces
■ Controls I/O pin multiplexing
■ Provides eight registers that software can use to pass information between boot
stages

Boot Configuration and System Information

The system manager provides boot configuration information through the bootinfo
register. The following information is available to software:
■ Sampled value of the HPS boot select (BSEL) pins
■ Sampled value of the HPS clock select (CSEL) pins
The boot configuration information comes from the HPS BSEL and CSEL pins. The BSEL
signal informs software which flash device CPU0 booted from. The CSEL signal
informs software which clock frequency is used for the flash controllers (NAND,
SD/MMC, quad SPI).
f For boot and clock source values, refer to the Booting and Configuration appendix in
The system manager also provides information about the type of HPS variant
available in the system on a chip (SoC) FPGA through the HPS information register
(hpsinfo) so that software can determine how many processor cores are present and
whether the CAN controllers are present in the HPS.
Additional Module Control

Each module in the HPS has its own CSRs, providing access to the internal state of the
module. Other modules, as well as logic in the FPGA fabric, can use these CSRs to
control and monitor functions in the HPS modules. The system manager CSRs
provide access to additional module state information, enabling additional control
and monitoring. Therefore, to fully control each module, you must manipulate both
the module's own CSRs, and CSRs in the system manager. This section describes
system manager CSR usage for each module.
Scan Manager
Either the FPGA JTAG pins or the scan manager can drive JTAG signals to the FPGA.
Registers in the system manager control a multiplexer that determines which source
drives the JTAG signals to the FPGA. Set the FPGA JTAG enable bit (fpgajtagen) of
the control register (ctrl) in the scan manager group (scanmgrgrp) to control this
selection.
1 Software running on the HPS must ensure that the FPGA JTAG pins and scan
manager's connection to the FPGA control block are inactive before changing the
source that drives JTAG signals to the FPGA.
f For more information about the scan manager, refer to the Scan Manager chapter in

DMA Controller
There are eight DMA peripheral request interfaces. The FPGA fabric always has
access to interfaces 0 to 3. Registers in the system manager control a group of
multiplexers that determine whether the FPGA fabric or a CAN controller has access
to each of interfaces 4 to 7. Table 14–1 lists how the channel select bits (chansel[3:0])
of the control register (ctrl) in the DMA controller group (dmagrp) control these
selections.
Table 14–1. DMA Peripheral Request Interfaces 4 to 7 Usage

Channel Select Bit Register Value Usage
0 DMA channel 4 connects to the FPGA fabric
0
1 DMA channel 4 connects to the CAN0 controller
1
2
3
The security state of the DMA controller is controlled by the manager thread security
(mgrnonsecure) and interrupt security (irqnonsecure) bits of the ctrl register, all in
the dmagrp group.
The security state of each DMA peripheral request interface is controlled by the
peripheral nonsecure bits (nonsecure[31:0]) of the peripheral security register
(persecurity) in the dmagrp group. The DMA controller samples these register bits
when it is brought out of reset.
1 Register bits should be accessed only when the DMA master interface is guaranteed
to be in an inactive state.
f For more information about the DMA controller, refer to the DMA Controller chapter

Software uses the bootstrap control register (bootstrap) in the NAND flash controller
group (nandgrp) to modify the default behavior of the NAND flash controller after
reset. The NAND flash controller samples the register bits when it is brought out of
reset.
The following bootstrap register bits control configuration of the NAND flash
controller:
■ Bootstrap inhibit initialization bit (noinit)—inhibits the NAND flash controller
from performing initialization when coming out of reset, and allows software to
program all registers pertaining to device parameters such as page size and width.
■ Bootstrap 512 byte device bit (page512)—informs the NAND flash controller that a
NAND flash device of 512 byte page size is connected to the system.

■ Bootstrap inhibit load block 0 page 0 bit (noloadb0p0)—inhibits the NAND flash
controller from loading page 0 of block 0 of the NAND flash device as part of the
initialization procedure.
■ Bootstrap two row address cycles bit (tworowaddr)—informs the NAND flash
controller that only two ROW address cycles are required instead of the default
three row address cycles.
Registers in the system manager control the L3 master ARCACHE and AWCACHE signals.
Set the NAND arcache (arcache[0]) and NAND awcache (awcache[0]) bits of the
NAND L3 master AxCACHE register (l3master) in the nandgrp group to control these
selections. These bits define the cache attributes for the master transactions of the
DMA engine in the NAND controller.
1 Register bits should be accessed only when the master interface is guaranteed to be in
an inactive state.
f For more information about the NAND flash controller, refer to the NAND Flash
Controller chapter in volume 3 of the Cyclone V Device Handbook.
EMAC
The system manager allows software to select either osc1_clk or fpga_ptp_ref_clk as
the source of the IEEE 1588 reference clock for each EMAC. Set the PTP clock select
(ptpclksel[0]) and (ptpclksel[1) bits of the ctrl register in the EMAC group
(emacgrp) to control this selection.
f For more information about the reference clock, refer to the Clock Manager chapter in
Registers in the system manager control the L3 master ARCACHE and AWCACHE signals.
Set the arcache[0], arcache[1], awcache[0], and awcache[1]bits of the EMAC L3
master AxCACHE register (l3master) in the emacgrp group to control these selections.
These bits define the cache attributes for the master transactions of the DMA engine in
the EMAC controllers.
an inactive state.
f For more information about the EMAC, refer to the Ethernet Media Access Controller
USB 2.0 OTG Controller

Registers in the system manager control whether the USB 2.0 OTG controller L3
master is used for data or opcode access. Set the USB HPROT data/opcode bits
(hprotdata[0] and hprotdata[1]) of the USB L3 master HPROT register (l3master) in
the USB controller group (usbgrp) to specify the access for each USB controller.
You can also use the system manager to specify whether the USB controller L3 master
access is privileged, bufferable, or cacheable for each USB controller by setting the
USB HPROT privileged (hprotpriv[0] and hprotdata[1]), USB HPROT bufferable
(hprotbuff[0] and hprotbuff[1]), and USB HPROT cacheable (hprotcache[0] and
hprotcache[1]) bits of the l3master register in the usbgrp group.

an inactive state.
f For more information about the USB 2.0 OTG controller, refer to the USB 2.0 OTG
SD/MMC Controller
Registers in the system manager control whether the SD/MMC controller L3 master is
used for data or opcode access. Set the SD/MMC hprotdata[0] bit of the SD/MMC
l3master register in the SD/MMC controller group (sdmmcgrp) to specify the access
mode.
You can use the system manager to specify whether the SD/MMC controller L3
master access is privileged, bufferable, or cacheable by setting the SD/MMC
hprotpriv[0], hprotbuff[0], and hprotcache[0] bits of the l3master register in the
sdmmcgrp group.
an inactive state.
The system manager allows software to select the clock’s phase shift for cclk_in_drv
and cclk_in_sample by setting the drive clock phase shift select (drvsel) and sample
clock phase shift select (smplsel) bits of the ctrl register in the sdmmcgrp group. You
can also select which feedback clock (fb_clk) to use as the cclk_in_sample clock. Set
the feedback clock select (fbclksel) bit of the ctrl register to control this selection.
f For more information about the SD/MMC controller, refer to the SD/MMC Controller
Watchdog Timer
The system manager controls the watchdog timer behavior when the CPUs are in
debug mode. The system manager sends a pause signal to the watchdog timers
depending on the setting of the debug mode (mode[0] and mode[1]) bits of the L4
watchdog debug register (wddbg). Each watchdog timer built into the MPU subsystem
is automatically paused when its associated CPU enters debug mode.
f For more information about the watchdog timer, refer to the Watchdog Timer chapter in
Boot ROM Code

Registers in the system manager control whether the boot ROM code configures the
pin multiplexing for boot pins after a warm reset. Set the warm reset configure pin
multiplex for boot pins bit (warmrstcfgpinmux) of the ctrl register in the boot ROM
code register group (romcodegrp) to enable or disable this control.
1 The boot ROM code always configures the pin multiplexing for boot pins after a cold
reset.

Registers in the system manager also control whether the boot ROM code configures
the I/O pins used during the boot process after a warm reset. Set the warm reset
configure IOs for boot pins bit (warmrstcfgio) of the ctrl register in the romcodegrp
group to enable or disable this control. By default, the boot ROM code always
configures the I/O pins used by boot after a cold reset.
When CPU1 is released from reset and the boot ROM code is located at the CPU1 reset
exception address (for a typical case), the boot ROM reset handler code reads the
address stored in the CPU1 start address register (cpu1startaddr) and jumps to that
address to pass control to software.
The preloader state register (initswstate) stores the magic number 0x49535756
written by the preloader to indicate there is a valid preloader software image in the
on-chip RAM.
There can be up to four preloader images stored in flash memory. The preloader last
image loaded register (initswlastld) contains the index of the preloader image
loaded in the on-chip RAM.
The boot ROM software state register (bootromswstate) is a 32-bit general-purpose
register reserved for the boot ROM.
The registers in the warm boot from on-chip RAM group (warmramgrp) in the
romcodegrp group are used by the boot ROM code to support booting from the on-
chip RAM on a warm reset. Table 14–2 lists the registers and their purposes.
Table 14–2. The warmramgrp Registers

Register Name Purpose
Controls whether the boot ROM attempts to boot from the
enable Enable
contents of the on-chip RAM on a warm reset.
Contains the byte offset of the warm boot CRC validation
datastart Data start region in the on-chip RAM. The offset must be word-aligned
to an integer multiple of four.
Contains the length in bytes of the region in the on-chip
length Length
RAM available for warm boot CRC validation.
Contains the byte offset into the on-chip RAM that the boot
execution Execution offset
code jumps to if the CRC validation succeeds.
Contains the expected CRC of the region in the on-chip
crc Expected CRC
RAM.
All the registers in Table 14–2 must be written by software prior to the warm reset
occurring.
The number of wait states applied to the boot ROM's read operation is determined by
the wait state bit (waitstate) of the ctrl register in the romhwgrp group. After the boot
process, software might require reading the code in the boot ROM. If software has
changed the clock frequency of the l3_main_clk after reset, an additional wait state is
necessary to access the boot ROM. Set the waitstate bit of the ctrl register in the
romhwgrp group to add an additional wait state to the read access of the boot ROM.
The enable safe mode warm reset update bit (ensfmdwru) of the ctrl register controls
whether the waitstate bit is updated during a warm reset.

Freeze Controller
The freeze controller provides freeze signals to the HPS configurable I/O elements.
These freeze signals ensure that the I/O pins are in a safe state upon FPGA power on
and until they are configured.
The freeze controller divides the HPS configurable I/O pins into four groups. Each
group has an independent set of freeze signals, referred to as a freeze channel, which
is driven by the freeze controller. Separate freeze channels allow each group of I/O
pins to be independently frozen, configured, and thawed.
The HPS I/O pins are divided into six banks. Each I/O bank is either a vertical (VIO)
or horizontal (HIO) I/O, based on its location on the die.Table 14–3 lists the I/O banks
and bank types that each freeze channel controls.
Table 14–3. Freeze Channels and HPS I/O Banks

Freeze Channel Bank Type HPS I/O Bank
0 VIO I/O bank 7D and bank 7E
1 VIO I/O bank 7B and bank 7C
2 VIO I/O bank 7A
3 HIO I/O bank 6
When the I/O elements are frozen, the configuration registers in the I/O elements are
not placed in a reset state. Instead, the outputs of the configuration registers are gated
off by the freeze signals until software running on the HPS thaws the I/O elements.
The scan manager must configure the I/O elements in each group before software
thaws them.
While frozen, the HPS I/O pins are in the following state:
■ I/O buffers are in tri-state mode
■ Weak pull-up is enabled
■ Outputs of configuration registers are gated
■ Internal registers are reset to their initial states
■ Bus hold is in tri-state mode
■ On-chip termination (OCT) internal timer is reset
Software can generate HPS I/O freeze signals by setting the VIO control (vioctrl)
and HIO control (hioctrl) registers in the freeze control group (frzctrl). The system
manager allows the following options to be set in the HPS VIO and HIO banks:
■ Enable or disable I/O pin configuration
■ Allow I/O pin configuration to control the bus hold circuit
■ Allow I/O pin configuration to control I/O tri-state
■ Allow I/O pin configuration to control weak pull-up resistor
■ Allow I/O pin configuration to control the slew-rate
■ Reset registers in the delay-locked loop (DLL)
■ Reset registers in the OCT

■ Reset registers in the I/O elements and data strobe (DQS)

■ Disable I/O pin configuration in the OCT calibration block, or start OCT
calibration state machine and enable I/O pin configuration in the OCT calibration
block
Software controls each freeze channel through a set of memory-mapped control
register fields in the system manager. The freeze signals for VIO channel 1 is
controlled either by a software state machine, or by a hardware state machine in the
freeze controller. Software chooses between the two freeze signal sources by setting
the source register (src) in the frzctrl group.
The system manager initiates a freeze or thaw operation on VIO channel 1 through the
hardware state machine. Freeze or thaw is initiated by requesting the hardware state
machine to generate a freeze signal sequence. This sequence transitions between
frozen and thawed states by setting the VIO channel 1 freeze/thaw request bit
(vio1req) of the hardware control register (hwctrl) in the frzctrl group.
The system manager allows software to determine the current state of the VIO
channel 1 (frozen or thawed) or to determine when a freeze or thaw request is made
by reading the VIO channel 1 state bit (vio1state) of the hwctrl register in the
frzctrl group.
FPGA Interface Enables

The system manager can enable or disable interfaces between the FPGA and HPS. The
interfaces must be disabled when not in use to avoid undefined behavior.
The global interface bit (intf) of the global disable register (gbl) in the FPGA interface
group (fpgaintfgrp) disables all interfaces between the FPGA and HPS.
1 Ensure that all interfaces between the FPGA and HPS are inactive before disabling
them.
You can set the individual disable register (indiv) in the fpgaintfgrp group to disable
the following interfaces between the FPGA and HPS:
■ Reset request interface
■ JTAG enable interface
■ I/O configuration interface
■ Boundary scan interface
■ Trace interface
■ System Trace Macrocell (STM) interface
■ Cross-trigger interface (CTI)
ECC and Parity Control

The system manager can enable or disable ECC for each of the following HPS
modules with ECC-protected RAM:
■ MPU L2 cache data RAM
■ On-chip RAM

■ USB 2.0 OTG controller (USB0 and USB1) RAM

■ EMAC (EMAC0 and EMAC1) RAM
■ DMA controller RAM
■ CAN controller RAM
■ NAND flash controller RAM
■ Quad SPI flash controller RAM
■ SD/MMC controller RAM
With the exception of L2 cache data RAM, each of the ECC memories generates a
single- and double-bit interrupt to the global interrupt controller (GIC) through the
system manager. Set the registers in the ECC management register group (eccgrp) to
enable or disable the ECC. The L2 cache generates the L2 cache data RAM interrupt
directly.
The system manager can inject single-bit or double-bit errors into each of the ECC
memories for testing purposes. Set the bits in the appropriate memory enable register
to inject errors. For example, to inject a single bit EEC error into the USB0 module, set
the injs bit of the USB0 RAM ECC enable register (usb0) in the ECC management
register group (eccgrp).
The system manager can also inject parity failures into the parity-protected RAM in
the MPU to test the parity failure interrupt handler. Set the bits of the parity fail
injection register (parityinj) to inject parity failures.
Pin Multiplexing Control

In the HPS, many of the pins are available for use by as many as four peripherals. The
system manager allows software to control pin multiplexing selections to select which
peripheral in the HPS has access to each shared pin. Some pins also provide boot
source and clock source information, which are sampled upon deassertion of a cold
reset event. Set the registers in the pin multiplexing control group (pinmuxgrp) to
control this selection.
1 Do not modify the pin multiplexing selection registers after I/O configuration.
Preloader Handoff Information

The system manager provides eight 32-bit registers to store handoff information
between the preloader and the operating system. The preloader can store any
information in these registers. These register contents have no impact on the state of
the HPS hardware. When the operating system kernel boots, it retrieves the
information by reading the preloader to OS handoff information register (handoff)
array in the preloader register group (iswgrp). These registers are reset only by a cold
reset.
Clocks
The system manager is driven by the osc1_clk clock generated by the clock manager.

System Manager Address Map and Register Definitions
Resets
The system manager receives two reset signals from the reset manager. The
sys_manager_rst_n signal is driven on a cold or warm reset and the
sys_manager_cold_rst_n signal is driven only on a cold reset. This functionality
allows the system manager to reset some CSR fields on either a cold or warm reset
and others only on a cold reset.
Table 14–4 lists the registers and groups of registers that are reset by a cold or warm
reset.
Table 14–4. Registers and Corresponding Reset SIgnal

Register or Group Cold Reset Warm Reset
fpgaintfgrp v —
scanmgrgrp v —
frzctrl v —
emacgrp v v
dmagrp v v
sdmmcgrp v v
nandgrp v v
usbgrp v v
eccgrp v —
pinmuxgrp v —
wddbg v —
parityinj v v
romcodegrp v —
romhwgrp v —
System Manager Address Map and Register Definitions

■ sysmgr



May 2012 1.1 Added functional description, address map and register definitions sections.


15. Scan Manager
November 2012
cv_54015-1.2
cv_54015-1.2
The scan manager is used to configure and manage the hard processor system (HPS)
I/O pins, and communicate with the FPGA JTAG test access port (TAP) controller.
The scan manager drives the HPS I/O scan chains to configure the I/O bank
properties before the pins are used by the peripherals in HPS. The scan manager can
also optionally communicate with the FPGA JTAG TAP controller to send commands
for purposes such as managing cyclic redundancy check (CRC) errors detected by the
FPGA control block. When the scan manager communicates with the FPGA JTAG
TAP controller, input on the FPGA JTAG pins is ignored.
The scan manager contains an ARM® JTAG Access Port (JTAG-AP). The JTAG-AP
implements a multiple scan-chain JTAG master interface. One scan chain connects to
the FPGA JTAG and uses the standard JTAG signals. Four other scan chains connect to
the HPS I/O banks, using the JTAG clock and data outputs as a parallel-to-serial
converter.
f For more information about the ARM JTAG-AP, refer to the ARM Debug Interface v5
Architecture Specification, which you can download from the ARM website
Features of the Scan Manager

The scan manager has the following features:
■ Drives all the I/O scan chains for HPS I/O banks
■ Allows the HPS to access the FPGA JTAG TAP controller

November 2012
Subscribe
15–2 Chapter 15: Scan Manager
Scan Manager Block Diagram and System Integration

Figure 15–1 shows a block diagram of the scan manager, showing how it is integrated
in the SoC device.
Figure 15–1. Scan Manager Block Diagram
System fpgajtagen FPGA Portion

Manager (select)
JTAG TAP Control
JTAG Block
Switch Controller
FPGA JTAG Pins
Scan Manager (CONFIG_IO Mode) HPS I/O Pins (1)

Scan Chain 7 IOCSR
Multiplexer
I/O Bank 6
Scan Chain 3 IOCSR 3
Scan Chain 2 I/O Bank 7A

IOCSR 2
JTAG-AP
I/O Bank 7C
I/O Bank 7B
I/O Bank 7E
I/O Bank 7D

(1) Not all devices contain all the banks depicted.
The processor accesses the scan manager through the register slave interface
connected to the level 4 (L4) peripheral bus.
Table 15–1 describes how the ARM JTAG-AP signals are connected in the scan
manager. These signals are internal to the scan manager, are provided here for
reference only, and are not shown in Figure 15–1. The signal, register, and field names
listed in the table match the names used in the ARM Debug Interface v5 Architecture
Specification. The remainder of this chapter uses register and field names as listed in
the Altera SoC register documentation. Refer to Table 15–3 to cross-reference the two
sets of register names.
Table 15–1. ARM JTAG-AP Signal Use in the Scan Manager (Part 1 of 2)
Signal Direction Implementation
SRSTCONNECTED[7:0] Input Tied to 0. The read-only SRSTCONNECTED field in the CSW register always reads as 0.
Tied to 0x8F, which connects only ports 0-3 and 7. The read-only PORTCONNECTED
PORTCONNECTED[7:0] Input field in the CSW register reads as 1 when the PORTSEL register is written with a value
that enables one of the connected ports, and reads as 0 otherwise.

Chapter 15: Scan Manager 15–3
Table 15–1. ARM JTAG-AP Signal Use in the Scan Manager (Part 2 of 2)
Signal Direction Implementation
Tied to 0x8F, so all connected ports are always considered powered on. The PSTA
register does not contain a useful value, so there is no reason for software to access
PORTENABLED[7:0] Input it. Software does not need to monitor the status of ports 0-3 because are always on.
For port 7, software can read the mode field of the stat register in the FPGA manager
to determine the FPGA power status.
nSRSTOUT[7:0] Output Not connected. Writing to the SRST_OUT field of the CSW register has no effect.
nTRST*[7] is connected to the FPGA JTAG TAP controller and nTRST*[6:0] are not
connected. Writing to the TRST_OUT field of the CSW register (the trst bit of the stat
nTRST*[7:0] Output
register in the scan manager) has an effect only when port 7 is enabled by software.
For details, refer to “Communicating with the JTAG TAP Controller” on page 15–5.
The ARM JTAG-AP supports up to eight scan chains. The scan manager uses only
scan chains 0, 1, 2, 3, and 7.
Scan chain 7 of the JTAG-AP connects to FPGA JTAG TAP controller. When the system
manager undergoes a cold reset, this connection is disabled and the FPGA JTAG pins
are connected to the FPGA JTAG TAP controller. You can configure the system
manager to enable the connection, which allows software running on the HPS to
communicate with the FPGA JTAG TAP controller. In this case, software can send
JTAG commands (such as the SHIFT_EDERROR_REG JTAG instruction) to the FPGA
JTAG and get responses to determine details about CRC errors detected by the control
block when the FPGA fabric is in user mode. Through the FPGA manager, software
can determine that a CRC error was detected. For more information about the TAP
controller, refer to “Communicating with the JTAG TAP Controller” on page 15–5.
Scan chains 0 to 3 of the JTAG-AP connect to the configuration information in the HPS
I/O scan chain banks through the I/O configuration shift register (IOCSR)
multiplexer. For more information, refer to “Configuring HPS I/O Scan Chains” on
page 15–4.
1 The I/O scan chains do not use the JTAG protocol. The scan manager uses the
JTAG-AP as a parallel-to-serial converter for the I/O scan chains. The I/O scan chains
are connected only to the serial output data (TDI JTAG signal) and serial clock (TCK
JTAG signal).
The HPS I/O pins are divided into six banks. Each I/O bank is either a vertical (VIO)
or horizontal (HIO) I/O, based on its location on the die. Table 15–2 shows the
mapping of the IOCSR scan chains to the I/O banks.
Table 15–2. Bank Usage of IOCSR Scan Chains

IOCSR Scan Chain Bank Type HPS I/O Bank Usage
0 VIO I/O bank 7D and I/O bank 7E EMAC
SD/MMC, NAND, and quad
1 VIO I/O bank 7B and I/O bank 7C
SPI
Trace, SPI, UART, I2C, and
2 VIO I/O bank 7A
CAN
3 HIO I/O bank 6 SDRAM DDR

Functional Description of the Scan Manager
When the FPGA JTAG TAP controller is in CONFIG_IO mode, the controller can
override the scan manager JTAG-AP and configure the HPS I/O pins. For more
information, refer to “Configuring HPS I/O Scan Chains” on page 15–4.
1 CONFIG_IO mode is commonly used to configure the I/O pin properties prior to
performing boundary scan testing.

This section describes the functional operation of the scan manager. The scan manager
serves the following two purposes:
■ Configuring HPS I/O Scan Chains
■ Communicating with the JTAG TAP Controller
Configuring HPS I/O Scan Chains

The HPS I/O pins are configured through a series of scan chains.
1 The HPS I/O pins need to be frozen before configuring them. For more information,
I/O pin configuration involves such steps as setting the I/O standard and drive
strength for each I/O bank. After a cold reset, all the I/O scan chains in the HPS must
be configured prior to being used to communicate with external devices.
Software uses the scan manager to write configuration data to the scan chains.
Separate I/O configuration data files for FPGA and HPS are generated by the
Quartus® II software when the configuration image for the FPGA portion of the
system-on-a-chip (SoC) device is assembled. The HPS configuration data is written to
the scan manager by software.
Before configuring a specific IO bank, the corresponding scan chain must be enabled
by writing to the bits in en register. The scan manager must not be active during this
process. Software reads the active bit of stat register to determine the scan manager
state.
Alternatively, when the FPGA JTAG TAP controller receives the CONFIG_IO JTAG
instruction, the control block enters CONFIG_IO mode. When the control block is in
CONFIG_IO mode, the controller can override the scan manager JTAG-AP and
configure the HPS I/O pins. The CONFIG_IO instruction configures all configurable
I/O pins in the SoC device including the FPGA I/O pins and the HPS I/O pins. The
FPGA and HPS portions of the device must be both powered on to execute the
CONFIG_IO instruction. External logic connected to the FPGA JTAG pins sends the
CONFIG_IO instruction, which provides I/O configuration data for all FPGA and HPS
I/O pins. While CONFIG_IO mode is active, the HPS is held in cold reset to prevent
software from potentially interfering with the I/O configuration.

Communicating with the JTAG TAP Controller

After the system manager undergoes a cold reset, access to the JTAG TAP controller in
the FPGA control block is through the dedicated FPGA JTAG I/O pins. If necessary,
you can configure your system to use the scan manager to provide the HPS processor
access to the JTAG TAP controller instead. This feature allows the processor to send
JTAG instructions to the FPGA portion of the device.
To connect scan chain 7 between the scan manager and the FPGA JTAG TAP
controller, the following features must be enabled:
■ The scan chain for the FPGA JTAG TAP controller—To enable scan chain 7, set the
fpgajtag field of the en register in the scan manager. For more information, refer
to “Scan Manager Address Map and Register Definitions” on page 15–6.
■ The FPGA JTAG logic source select—This source select determines whether the
scan manager or the dedicated FPGA JTAG pins are connected to the FPGA JTAG
TAP controller in the FPGA portion of the device. On system manager cold reset,
the dedicated FPGA JTAG pins are selected. The source select is configured
through the fpgajtagen bit of the ctrl register in the scanmgrgrp group of the
system manager. The FPGA JTAG pins and scan manager connection to the TAP
controller must both be inactive when switching between them. The mechanism to
ensure both are inactive is user-defined.
f For more information, refer to the System Manager chapter in volume 3 of

1 Before connecting or disconnecting the scan chain between the scan manager and the
FPGA JTAG TAP controller, ensure that both the FPGA JTAG TCK and scan manager
TCK signals are de-asserted. Altera recommends resetting the FPGA JTAG TAP
controller using the scan manager's nTRST signal after the scan manager is connected
to the controller.
JTAG-AP FIFO Buffer Access and Byte Command Protocol

The JTAG-AP contains FIFO buffers for byte commands and responses. The buffers
are accessed through the fifosinglebyte, fifodoublebyte, fifotriplebyte, and
fifoquadbyte registers. The JTAG-AP stalls processor access to the registers when the
buffer does not contain enough data for read access, or when the buffer does not
contain enough free space to accept data for write access.
1 Software should read the rfifocnt and wfifocnt fields of the stat register to
determine the buffer status before performing the access to avoid being stalled by the
JTAG-AP.
JTAG-AP scan chains 0, 1, 2 and 3 are write-only ports connected to the HPS IOCSRs
and JTAG-AP scan chain 7 is a read-write port connected to the FPGA JTAG TAP
controller. The processor can send data to scan chains 0-3, and send and receive data
from scan chain 7 by accessing the command and response FIFO buffers in the
JTAG-AP.
1 Attempting to access data at invalid or non-aligned offsets can produce unpredictable

results that require a reset to recover.

Scan Manager Address Map and Register Definitions
The JTAG commands and TDI data must be sent to the JTAG-AP using an encoded
byte protocol. Similarly, the TDO data received from JTAG-AP is encoded. All
commands are 8 bits wide in the byte command protocol.
f For details about the byte command protocol, refer to the JTAG-AP chapter in the
ARM Debug Interface v5 Architecture Specification, which you can download from the
Clocks
The scan manager is connected to the spi_m_clk clock generated by the clock
manager.
f For more information, including minimum and maximum clock frequencies, refer to
the Clock Manager chapter in volume 3 of the Cyclone V Device Handbook.
The scan manager generates two clocks. One clock routes to the control block of the
FPGA portion of the SoC device with a frequency of spi_m_clk / 6 and runs at a
maximum of 33 MHz. The other clock routes to the HPS I/O scan chains with a
frequency of spi_m_clk / 2 and runs at a maximum frequency of 100 MHz.
1 The spi_m_clk can potentially run faster than the scan manager supports so that SPI
masters can support 60 Mbps rates. When the SPI master is running faster than what
is supported by the scan manager, the scan manager cannot be used and must be held
in reset.
Resets
The reset manager provides the scan_manager_rst_n reset signal to the scan manager
for both cold and warm resets.
Because glitches can happen on the output clocks during a warm reset, the scan
manager temporarily stops generation of the JTAG-AP and I/O configuration clocks.
This action ensures that a warm reset does not cause output clock glitches.
Before asserting warm reset, the reset manager sends a request to the scan manager.
The scan manager stops the output clock generation and acknowledges the reset
manager. The reset manager then issues the warm reset. To enable this warm reset
handshake, configure the scanmgrhsen bit of the reset manager ctrl register.
f For more information about reset handshaking, refer to the Reset Manager chapter in
Scan Manager Address Map and Register Definitions

■ scanmgr

To improve clarity regarding how Altera uses the JTAG-AP, the ARM register names
are changed in the SoC device. Table 15–3 cross references the ARM and Altera names.
Table 15–3. JTAG-AP Register Names

Altera Name ARM Name
stat CSW (control/status word)
en PSEL
fifosinglebyte BWFIFO1 for writes, BRFIFO1 for reads
fifodoublebyte BWFIFO2 for writes, BRFIFO2 for reads
fifotriplebyte BWFIFO3 for writes, BRFIFO3 for reads
fifoquadbyte BWFIFO4 for writes, BRFIFO4 for reads
f For more information about the ARM JTAG-AP, refer to the ARM Debug Interface v5
Architecture Specification, which you can download from the ARM website


November 2012 1.2 Added JTAG-AP descriptions.
Added block diagram and system integration, functional description, and address map and
May 2012 1.1
register definitions sections.


16. DMA Controller
November 2012
cv_54016-1.1
cv_54016-1.1
This chapter describes the direct memory access controller (DMAC) contained in the
hard processor system (HPS). The DMA controller is used to transfer data between
memory and peripherals and other memory locations in the system. The DMA
controller is an instance of the ARM® Corelink™ DMA Controller (DMA-330).
f For more information about ARM’s DMA-330 controller, refer to the CoreLink DMA
Controller DMA-330 Revision: r1p1 Technical Reference Manual, available from the ARM
Features of the DMA Controller

The HPS provides one DMAC to handle the data transfer between memory-mapped
peripherals and memories, off-loading this work from the microprocessor unit (MPU)
subsystem. The DMAC supports memory-to-memory, memory-to-peripheral, and
peripheral-to-memory transfers. The DMAC supports up to eight logical channels for
different levels of service requirements. It provides up to 31 peripheral handshake
interfaces for peripheral hardware flow control.
The DMA controller contains an instruction processing block that enables it to process
program code that controls a DMA transfer. It also contains an ARM Advanced
Microcontroller Bus Architecture (AMBA®) Advanced eXtensible Interface (AXI™)
master interface unit to fetch the program code from system memory into its
instruction cache. The AXI master interface is used to perform DMA data transfer as
well. The DMA instruction execution engine executes the program code from its
instruction cache and schedules read or write AXI instructions through the respective
instruction queues. It also contains a multi-FIFO (MFIFO) data buffer that it uses to
store data that it reads, or writes, during a DMA transfer.
The DMAC provides 11 interrupt outputs to enable efficient communication of events
to the MPU subsystem. The peripheral request interfaces support the connection of
DMA-capable peripherals to enable memory-to-peripheral and
peripheral-to-memory transfers to occur, without intervention from the processor.
Since the HPS supports some peripherals that do not comply with ARM DMA
peripheral interface protocol, adapters are added to allow these peripherals to work
with the DMAC. The following peripheral interface protocols are supported:
■ Synopsys protocol
■ Serial peripheral interface (SPI)
■ Universal asynchronous receiver/transmitter (UART)
■ Inter-integrated circuit (I2C)
■ FPGA

November 2012
Subscribe
16–2 Chapter 16: DMA Controller
Features of the DMA Controller
■ ARM protocol
■ Quad SPI flash controller
■ System trace macrocell (STM™)
■ Bosch protocol
■ Controller area network (CAN)
Dual slave interfaces enable the operation of the DMA controller to be partitioned into
the Secure state and Non-secure state. The network interconnect must be configured
to ensure that only secure transactions can access the secure interface. The slave
interfaces can access status registers and also directly execute instructions in the DMA
controller.
The DMAC has the following features:
■ A small instruction set that provides a flexible method of specifying the DMA
operations. This architecture provides greater flexibility than the fixed capabilities
of a Linked-List Item (LLI) based DMA controller
■ Supports multiple transfer types
■ Memory-to-memory
■ Memory-to-peripheral
■ Peripheral-to-memory
■ Scatter-gather
■ Supports up to eight DMA channels
■ Supports up to eight outstanding AXI read and eight outstanding AXI write
transactions.
■ Enables software to schedule up to 16 outstanding read and 16 outstanding write
instructions
■ Supports 11 interrupt lines into the MPU subsystem
■ 1 for DMA thread abort
■ 8 for events
■ 2 for MFIFO buffer ECC
■ Supports 31 peripheral request interfaces
■ 4 for FPGA
■ 4 shared for FPGA or CAN
■ 8 for I2C
■ 8 for SPI
■ 2 for quad SPI
■ 1 for System Trace Macrocell
■ 4 for UART

Chapter 16: DMA Controller 16–3
DMA Controller Block Diagram and System Integration
DMA Controller Block Diagram and System Integration

Figure 16–1 shows a block diagram of the DMAC and how it is integrated into the rest
of the HPS system.
Figure 16–1. DMA Controller Connectivity
DMA Controller
MFIFO
512x64 MPU Subsystem
Generic Interrupt
Interrupt Controller
AXI-64 Master Interface
Control
Write Instruction Queue

Synopsis
Read Instruction Queue UART / SPI / I2C
Adapter and
Peripheral
Clock
Level 3 Interconnect
Interfaces
Crossing
Bosch
Peripheral Request Adapter and CAN Peripheral
Instruction Cache Instruction Execution Engine
Interface [30:0] Clock Interfaces
Crossing
STM / Quad SPI

Clock
Peripheral
Crossing
Interfaces
CSRs
Reset
System
Initialization
Manager
Interface
Secure Non-Secure
Slave Interface Slave Interface
Level 4 Main Bus
The DMA controller is driven by the l4_main_clk clock, which is used for the
controller logic as well as all the interfaces. The DMA controller accesses the level 3
(L3) main switch with its 64-bit AXI master interface.
The DMA controller provides the following slave interfaces:
■ Non-secure slave interface
■ Secure slave interface
You can use these slave interfaces to access the registers that control the functionality
of the DMA controller. The DMA controller implements TrustZone® secure
technology with one interface operating in the Secure state and the other operating in
the Non-secure state.
Functional Description of the DMA Controller

This section describes the major interfaces and components of the DMAC, and how it
operates.

Overview
The DMAC contains an instruction processing block that enables it to process
program code that controls a DMA transfer. The program code is stored in a region of
system memory that the DMAC accesses using its AXI master interface. The DMAC
stores instructions temporarily in an internal cache.
The DMAC is configured with eight DMA channels, with each channel capable of
supporting a single concurrent thread of DMA operation. In addition, a single DMA
manager thread exists, and you can use it to initialize the DMA channel threads. The
DMAC executes up to one instruction per clock cycle. To ensure that it regularly
executes each active thread, it alternates by processing the DMA manager thread and
then a DMA channel thread. It uses a round-robin process when selecting the next
active DMA channel thread to execute.
The DMAC uses variable-length instructions that consist of one to six bytes. It
provides a separate program counter (PC) register for each DMA channel. When a
thread requests an instruction from an address, the cache performs a lookup. If a
cache hit occurs, then the cache immediately provides the instruction. Otherwise, the
thread is stalled while the DMAC uses the AXI master interface to perform a cache
line fill. If an instruction spans the end of a cache line, the DMAC performs multiple
cache accesses to fetch the instruction.
1 When a cache line fill is in progress, the DMAC enables other threads to access the
cache, but if another cache miss occurs, this stalls the pipeline until the first line fill is
complete.
When a DMA channel thread executes a load or store instruction, the DMAC adds the
instruction to the relevant read or write queue. The DMAC uses these queues as an
instruction storage buffer prior to it issuing the instructions on the AXI. The DMAC
also contains an MFIFO data buffer that it uses to store data that it reads, or writes,
during a DMA transfer.
The DMAC provides multiple interrupt outputs to enable efficient communication of
events to external microprocessors. The peripheral request interfaces support the
connection of DMA-capable peripherals to enable memory-to-peripheral and
peripheral-to-memory DMA transfers to occur, without intervention from a
microprocessor.
Dual slave interfaces enable the operation of the DMAC to be partitioned into the
Secure state and Non-secure states. You can use the slave interfaces to access status
registers and also directly execute instructions in the DMAC.
Operating States
Figure 16–2 shows the transitions among operating states for the DMA manager
thread and DMA channel threads.

The DMAC provides a separate state machine for each thread.
Figure 16–2. Thread Operating States
C
Faulting Stopped
Killing C
C C
C M
Faulting
completing Completing Cache
miss
C
C
Executing
Updating
PC
C
C
Waiting for C C Waiting

peripheral for event
At barrier
1 In Figure 16–2, the DMAC permits that:
■ Only DMA channel threads can use states shown in gray

■ Arcs with no letter designator indicate state transitions for the DMA manager
and DMA channel threads, otherwise use is restricted as follows:
■ C—DMA channel threads only.
■ M—DMA manager thread only.
■ States within the dotted line can transition to the Faulting completing,
Faulting, or Killing
After the DMAC exits from reset, it sets all DMA channel threads to the Stopped state,
and DMA manager thread moves to the Stopped state.
The following sections describe the states:
■ “Stopped” on page 16–6
■ “Executing” on page 16–6
■ “Cache Miss” on page 16–7
■ “Updating PC” on page 16–7
■ “Waiting For Event” on page 16–7
■ “At Barrier” on page 16–7
■ “Waiting For Peripheral” on page 16–7

■ “Faulting Completing” on page 16–7

■ “Faulting” on page 16–7
■ “Killing” on page 16–7
■ “Completing” on page 16–7
Stopped
The thread has an invalid PC and it is not fetching instructions. Depending on the
thread type, you can cause the thread to move to the Executing state by:
■ DMA manager thread—Issuing the DMAGO instruction through the slave
interface.
■ DMA channel thread—Programming the DMA manager thread to execute DMAGO
for a DMA channel thread in the Stopped state.
Executing
The thread has a valid PC and therefore the DMAC includes the thread when it
arbitrates. The thread can then change to one of the following states under the
following conditions:
■ Stopped—When the DMA manager thread executes DMAEND.
■ Cache miss—When the instruction cache does not contain the next instruction for
either the DMA manager thread or the DMA channel thread.
■ Updating PC—When the DMAC calculates the address of the next access in the
cache.
■ Waiting for event—When a thread executes DMAWFE.
■ At barrier—When a DMA channel thread either:
■ Executes DMARMB, DMAWMB, or DMAFLUSHP
■ Updates control registers that affect alignment, refer to “Updating DMA
Channel Control Registers During a DMA Cycle” on page 16–25.
■ Waiting for peripheral—When a DMA channel thread executes DMAWFP.
■ Killing—When a DMA channel thread executes DMAKILL.
■ Faulting completing—For a DMA channel thread when either:
■ The thread executes an undefined or invalid instruction
■ An AXI error occurs during an instruction fetch or data transfer.
■ Faulting—For the DMA manager thread when either:
■ The thread executes an undefined or invalid instruction
■ An AXI error occurs during an instruction fetch.
For a DMA channel thread when a watchdog timeout abort occurs.
■ Completing—When a DMA channel thread executes DMAEND.

Cache Miss
The thread is stalled and the DMAC is performing a cache line fill. After it completes
the cache fill, the thread returns to the Executing state.
Updating PC
The DMAC is calculating the address of the next access in the cache. After it calculates
the PC, the thread returns to the Executing state.
Waiting For Event

The thread is stalled and is waiting for the DMAC to execute DMASEV using the
corresponding event number. After the corresponding event occurs, the thread
returns to the Executing state.
At Barrier
A DMA channel thread is stalled and the DMAC is waiting for transactions on the
AXI to complete. After the AXI transactions complete, the thread returns to the
Executing state.
Waiting For Peripheral

A DMA channel thread is stalled and the DMAC is waiting for the peripheral to
provide the requested data. After the peripheral provides the data, the thread returns
to the Executing state.
Faulting Completing
A DMA channel thread is waiting for the AXI master interface to signal that the
outstanding load or store transactions are complete. After the transactions complete,
the thread moves to the Faulting state.
Faulting
The thread is stalled indefinitely. The thread moves to the Stopped state when you use
the DBGCMD register to instruct the DMAC to execute DMAKILL for that thread.
Killing
the thread moves to the Stopped state.
Completing
the thread moves to the Stopped state.

Initializing the DMAC

The DMAC provides several memory-mapped control signals that initialize its
operating state when it exits from reset. The DMAC is configured so that it does not
automatically begin executing code when it exits from reset. The three
memory-mapped control signals are controlled by the system manager.
f For further information, refer to the System Manager chapter in volume 3 of the
How to Set the Security State of the DMA Manager

The boot_manager_ns signal is the only method to set the security state of the DMA
manager.
When the DMAC exits from reset, it reads the status of the boot_manager_ns signal
and sets the security of the DMA manager.
1 When set, the security state remains constant until a state transition on the dma_rst_n
signal resets the DMAC.
Refer to “DMA Manager Thread in Secure State” on page 16–20 and “DMA Manager
Thread in Non-Secure State” on page 16–21 for a description of how the security state
of the DMA manager affects how the DMAC operates.
How to Set the Security State for the Interrupt Outputs

The DMAC provides the boot_irq_ns[7:0] signals to enable you to assign each
irq[x] signal to a security state.
The boot_irq_ns[7:0] signals are connected to the system manager. Before taking the
DMA out of reset, you should program boot_irq_ns[7:0] through the system
manager to control which interrupt bits are secure.
1 The DMAC samples these bits immediately after it comes out of reset, and then
ignores them until the next reset.
When set, the security state of each irq[x] remains constant until a state transition on
the dma_rst_n signal resets the DMAC.
Refer to “Security Usage” on page 16–20 for a description of how the security state of
the irq[x] signals affects how the DMAC executes the DMAWFE and DMASEV
instructions.
How to Set the Security State for a Peripheral Request Interface

The DMAC provides the boot_periph_ns[31:0] signals to enable you to assign each
peripheral request interface to a security state.
The boot_periph_ns[31:0] signals are connected to the system manager. Before
taking the DMA out of reset, you should program the boot_periph_ns[31:0] signals
through the system manager to control which peripheral interfaces are secure.

1 The DMAC samples these bits immediately after it comes out of reset, and then
ignores them until the next reset. When set, the security state of each peripheral
request interface remains constant until a state transition on the dma_rst_n signal
resets the DMAC.
Refer to “Security Usage” on page 16–20 for how the security state of the peripheral
request interfaces affects how a DMA channel thread executes the DMAWFP, DMALDP,
DMASTP, or DMAFLUSHP instructions.
Using the Slave Interfaces

The slave interfaces connect the DMAC to the level 4 (L4) main bus and enable a
microprocessor to access the registers. Using these registers, a microprocessor can:
■ Access the status of the DMA manager thread
■ Access the status of the DMA channel threads
■ Enable or clear interrupts
■ Enable events
■ Issue an instruction for the DMAC to execute by programming the following
debug registers:
■ DBGCMD register
■ DBGINST0 register
■ DBGINST1 register
Issuing Instructions to the DMAC using a Slave Interface

When the DMAC is operating, you can only issue the following limited subset of
instructions:
■ DMAGO—Starts a DMA transaction using a DMA channel that you specify.
■ DMASEV—Signals the occurrence of an event, or interrupt, using an event number
that you specify.
■ DMAKILL—Terminates a thread.
You must ensure that you use the appropriate slave interface, depending on the
security state in which the boot_manager_ns signal initializes the DMAC to operate.
For example, if the DMAC is in the Secure state, you must issue the instruction using
the secure slave interface, otherwise the DMAC ignores the instruction. You can use
the secure or non-secure slave interface to start or restart a DMA channel when the
DMAC is in the Non-secure state.
1 Before you can issue instructions using the debug instruction registers or the DBGCMD
register, you must read the DBGSTATUS register to ensure that debug is idle, otherwise
the DMAC ignores the instructions.
When the DMAC receives an instruction from a slave interface, it can take several
clock cycles before it can process the instruction, for example, if the pipeline is busy
processing another instruction.

1 Prior to issuing DMAGO, you must ensure that the system memory contains a suitable
program for the DMAC to execute, starting at the address that the DMAGO specifies.
Using DMAGO with the Debug Instruction Registers

This example shows the necessary steps to start a DMA channel thread using the
debug instruction registers.
1. Create a program for the DMA channel.
2. Store the program in a region of system memory.
3. Use one of the slave interfaces on the DMAC to program a DMAGO instruction as
follows:
a. Poll the DBGSTATUS register to ensure that debug is idle, that is, the dbgstatus
bit is 0.
b. Write to the DBGINST0 register and enter the:
■ Instruction byte 0 encoding for DMAGO.
■ Instruction byte 1 encoding for DMAGO.
■ Debug thread bit to 0. This selects the DMA manager thread.
c. Write to the DBGINST1 register with the DMAGO instruction byte [5:2] data. You
must set these four bytes to the address of the first instruction in the program,
that is written to system memory in step b.
4. Instruct the DMAC to execute the instruction that the debug instruction registers
contain by writing zero to the DBGCMD register. The DMAC starts the DMA channel
thread and sets the dbgstatus bit to 1. After the DMAC completes execution of the
instruction, it clears the dbgstatus bit to 0.
Peripheral Request Interface

Figure 16–3 shows that the peripheral request interface consists of a peripheral
request bus and a DMAC acknowledge bus that use the prefixes:
■ dr—The peripheral request bus.
■ da—The DMAC acknowledge bus.
Figure 16–3. Request and Acknowledge Buses on the Peripheral Request Interface
drvalid
drtype[1:0]
drlast
drready Peripheral
Peripheral request DMAC
interface
davalid
datype[1:0]
daready
The peripheral signals the following on the request bus:

■ Request a single transfer

■ Request a burst transfer
■ Acknowledge a flush request.
The peripheral signals the DMAC when it issues the last request of the DMA transfer
sequence.
The DMAC can signal the following on the acknowledge bus:
■ Signal when it completes the requested single transfer
■ Signal when it completes the requested burst transfer
■ Issue a flush request.
The DMAC allows you to assign a peripheral request interface to any of the DMA
channels. When a DMA channel thread executes DMAWFP, the value programmed in the
peripheral [4:0] field specifies the peripheral associated with that DMA channel. For
more information, refer to “DMAWFP” on page 16–43.
The DMAC supports 31 peripheral request handshakes. Each request handshake can
receive up to four outstanding requests, and is assigned a specific peripheral device
ID. Table 16–1 lists the peripheral device ID assignments.
Table 16–1. Peripheral Request Interface Mapping

Request Request
Peripheral Interface Protocol Peripheral Interface Protocol
ID ID
FPGA 0 0 Synopsys SPI Master 0 TX 16 Synopsys
FPGA 1 1 Synopsys SPI Master 0 RX 17 Synopsys
FPGA 2 2 Synopsys SPI Slave 0 TX 18 Synopsys
FPGA 3 3 Synopsys SPI Slave 0 RX 19 Synopsys
FPGA 4 / CAN 0 interface 1 4 Synopsys / Bosch SPI Master 1 TX 20 Synopsys
FPGA 5 / CAN 0 interface 2 5 Synopsys / Bosch SPI Master 1 RX 21 Synopsys
FPGA 6 / CAN 1 interface 1 6 Synopsys / Bosch SPI Slave 1 TX 22 Synopsys
FPGA 7 / CAN 1 interface 2 7 Synopsys / Bosch SPI Slave 1 RX 23 Synopsys
I2C 0 TX 8 Synopsys Quad SPI Flash TX 24 ARM
I2C 0 RX 9 Synopsys Quad SPI Flash RX 25 ARM
I2C 1 TX 10 Synopsys STM 26 ARM
I2C 1 RX 11 Synopsys Reserved 27 —
I2C 2 TX (EMAC) 12 Synopsys UART 0 TX 28 Synopsys
I2C 2 RX (EMAC) 13 Synopsys UART 0 RX 29 Synopsys
I2C 3 TX (EMAC) 14 Synopsys UART 1 TX 30 Synopsys
I2C 3 RX (EMAC) 15 Synopsys UART 1 RX 31 Synopsys
Request interface numbers 4 through 7 are multiplexed between the CAN controllers
and soft logic implemented in the FPGA fabric. The switching between the CAN
controller and FPGA interfaces is controlled by the system manager.

f For more information about controlling request interface numbers 4 through 7 refer to
the System Manager chapter in volume 3 of the Cyclone V Device Handbook.
Request Acceptance Capability

The DMAC is able to accept one active request for each peripheral request interface.
An active request is where the DMAC has not started the requested AXI data
transfers.
Peripheral Length Management

The peripheral request interface enables a peripheral to control the quantity of data
that a DMA cycle contains, without the DMAC being aware of how many data
transfers it contains. The peripheral controls the DMA cycle in one of the following
ways:
■ Selects a single transfer
■ Selects a burst transfer
■ Notifies the DMAC when it commences the final request in the current series
When the DMAC executes a DMAWFP periph instruction, it halts execution of the thread
and waits for the peripheral to send a request. When the peripheral sends the request,
the DMAC sets the state of the request flags depending on the state of the following
signals:
■ drtype_<x>[1:0]—The DMAC sets the state of the request_type flag:
■ drtype_<x>[1:0]=b00—request_type<x> = Single.
■ drtype_<x>[1:0]=b01—request_type<x> = Burst.
■ drlast_<x>—The DMAC sets the state of the request_last flag:
■ drlast_<x>=0—request_last<x> = 0.
■ drlast_<x>=1—request_last<x> = 1.
1 If the DMAC executes a DMAWFP single or DMAWFP burst instruction then the DMAC
sets:
■ The request_type<x> flag to Single or Burst, respectively

■ The request_last<x> flag to 0.
DMALPFE is an assembler directive which forces the associated DMALPEND instruction to
have its nf bit set to 0. This creates a program loop that does not use a loop counter to
terminate the loop.
The DMAC exits the loop when the request_last flag is set to 1.
The DMAC conditionally executes the following instructions, depending on the state
of the request_type and request_last flags:
■ DMALD, DMAST, DMALPEND
When these instructions use the optional B|S suffix then the DMAC executes a
DMANOP if the request_type flag does not match.

■ DMALDP<B|S>, DMASTP<B|S>
The DMAC executes a DMANOP if the request_type<x> flag does not match the B|S
■ DMALPEND
When the nf bit is 0, the DMAC executes a DMANOP if the request_last flag is set.
Use the DMALDB, DMALDPB, DMASTB and DMASTPB instructions if you require the DMAC to
issue a burst transfer when the DMAC receives a burst request. The values in the CCRn
register control the amount of data that the DMAC transfers.
Use the DMALDS, DMALDPS, DMASTS and DMASTPS instructions if you require the DMAC to
issue a single transfer when the DMAC receives a single request. The DMAC ignores
the value of the src_burst_len and dst_burst_len fields in the CCRn register and sets
the arlen[3:0] or awlen[3:0] buses to 0x0.
DMAC Length Management

DMAC length management is when the DMAC controls the total amount of data to
transfer. The peripheral uses the peripheral request interface to notify the DMAC
when it requires the DMAC to transfer data to or from the peripheral. The DMA
channel thread controls how the DMAC responds to the peripheral requests.
The following constraints apply to DMAC length management:
■ The total quantity of data for all the single requests from a peripheral must be less
than the quantity of data for a burst request for that peripheral.
1 The CCRn register controls how much data is transferred for a burst request
and a single request. Altera recommends that you do not update a CCRn
register when a transfer is in progress for channel n.
■ When the peripheral sends a burst request then the peripheral must not send a
single request until the DMAC acknowledges that the burst request is complete.
Use the DMAWFP single instruction when you require the program thread to halt
execution until the peripheral request interface receives any request type. If the head
entry in the request FIFO buffer is of request type:
■ Single—The DMAC pops the entry from the FIFO buffer and continues program
execution.
■ Burst—The DMAC leaves the entry in the FIFO buffer and continues program
execution.
1 The burst request entry remains in the request FIFO buffer until the DMAC
executes a DMAWFP burst instruction or a DMAFLUSHP instruction.
Use the DMAWFP burst instruction when you require the program thread to halt
execution until the peripheral request interface receives a burst request. If the head
entry in the request FIFO buffer is of request type:
■ Single—The DMAC removes the entry from the FIFO buffer and program
execution remains
■ Burst—The DMAC pops the entry from the FIFO buffer and continues program
execution.

Use the DMALDP instruction when you require the DMAC to send an acknowledgement
to the peripheral when it completes the AXI read transfers. Similarly, use the DMASTP
instruction when you require the DMAC to send an acknowledgement to the
peripheral when it completes the AXI write transfers. The DMAC uses the
acknowledge bus to signal a transfer acknowledgement to peripheral <x>.
1 The DMAC sends an acknowledgement for a read transaction when the rvalid and
rlast signals are high and for a write transaction when bvalid signal is high. The
DMAC might send an acknowledgement to the peripheral while the transfer of write
data to the end destination is still in progress.
Use the DMAFLUSHP instruction to reset the request FIFO buffer for the peripheral
request interface. After the DMAC executes DMAFLUSHP, it ignores peripheral requests
until the peripheral acknowledges the flush request. This enables the DMAC and
peripheral to synchronize with each other.
Limitations
The peripherals connected to the DMA peripheral request interface use one of the
following protocols:
■ ARM
■ Synopsys
■ Bosch
For a peripheral using the ARM protocol, the only logic between the DMA and the
peripheral is clock crossing logic. For other protocols, clock crossing and protocol
adaptation logic is located between the DMA and peripheral. The following sections
discuss the limitations of the peripheral handshake interface for peripherals using the
Synopsys or Bosch protocols.
Burst Only Request

The Bosch protocol used by the CAN controllers only supports burst requests. As a
result any time either CAN controller issues the peripheral burst request the DMA
controller receives a burst and single request.
No Flush Support
For peripherals that use the Bosch or Synopsys peripheral request protocols the flush
command is not supported. Issuing a flush to any peripheral interface supporting the
Bosch or Synopsys protocols is dropped before reaching the peripheral.
No Acknowledge Type
For the Bosch peripheral request interface protocol, there is no acknowledge available.
Issuing an acknowledge to either CAN controller as a result is dropped before
reaching the peripheral.
For the Synopsys peripheral request interface protocol, there is no acknowledge type
available. As a result peripherals using the Synopsys protocol cannot distinguish
between acknowledges to burst and single transfers.

Using Events and Interrupts

The DMAC can support eight events and interrupts. The INTEN register is used to
control if each event-interrupt resource is either an event or an interrupt.
When the DMAC executes a DMASEV instruction it modifies the event-interrupt
resource that you specify. The INTEN register sets the event-interrupt resource to
function as an:
■ Event—The DMAC generates an event for the specified event-interrupt resource.
When the DMAC executes a DMAWFE instruction for the same event-interrupt
resource then it clears the event.
■ Interrupt—The DMAC sets the irq<event_num> signal high, where <event_num>
is the number of the specified event resource. To clear the interrupt you must write
to the INTCLR register.
Using an Event to Restart DMA Channels

When you program the INTEN register to generate an event, you can use the DMASEV
and DMAWFE instructions to restart one or more DMA channels.
DMAC executes DMAWFE before DMASEV

To restart a single DMA channel:
1. The first DMA channel executes DMAWFE and then stalls while it waits for the event
to occur.
2. The other DMA channel executes DMASEV using the same event number. This
generates an event, and the first DMA channel restarts. The DMAC clears the
event, one clock cycle after it executes DMASEV.
You can program multiple channels to wait for the same event. For example, if four
DMA channels have all executed DMAWFE for event 2, then when another DMA channel
executes DMASEV for event 2, the four DMA channels all restart at the same time. The
DMAC clears the event, one clock cycle after it executes DMASEV.
DMAC executes DMASEV before DMAWFE

If the DMAC executes DMASEV before another channel executes DMAWFE then the event
remains pending until the DMAC executes DMAWFE. When the DMAC executes DMAWFE
it halts execution for one aclk clock cycle, clears the event and then continues
execution of the channel thread.
For example, if the DMAC executes DMASEV 6 and none of the other threads have
executed DMAWFE 6 then the event remains pending. If the DMAC executes DMAWFE 6
instruction for channel 4 and then executes DMAWFE 6 instruction for channel 3, then:
1. The DMAC halts execution of the channel 4 thread for one aclk clock.
2. The DMAC clears event 6.
3. The DMAC resumes execution of the channel 4 thread.
4. The DMAC halts execution of the channel 3 thread and the thread stalls while it
waits for the next occurrence of event 6.

Interrupting the MPU Subsystem

The DMAC provides the irq[x] signals for use as active-high level-sensitive
interrupts to the MPU subsystem. When you program the INTEN register to generate
an interrupt, after the DMAC executes DMASEV, it sets the corresponding irq[x] signal
high.
The MPU subsystem can clear the interrupt by writing to the INTCLR register.
1 Executing DMAWFE does not clear an interrupt.
If you use the DMASEV instruction to notify a microprocessor when the DMAC
completes a DMALD or DMAST instruction then Altera recommends that you insert a
memory barrier instruction before the DMASEV. Otherwise the DMAC might signal an
interrupt before the AXI transfers complete. Refer to Example 16–1.
Example 16–1. Memory Barrier Instruction

DMALD
DMAST
# Issue a write memory barrier
# Wait for the AXI write transfer to complete before the DMAC
# can send an interrupt
DMAWMB
# The DMAC sends the interrupt
DMASEV
Aborts
This section describes:
■ Abort Types
■ Abort Sources
■ Watchdog Abort
■ Abort Handling
Abort Types
An abort can be classified as either precise or imprecise, depending on whether the
DMAC provides an abort handler with the precise state of the DMAC when the abort
occurs. If an abort is:
■ Precise The DMAC updates the PC register with the address of the instruction that
created the abort.
■ Imprecise The PC register might contain the address of an instruction which did
not cause the abort to occur.
Abort Sources
The DMAC signals a precise abort under the following conditions:
■ A DMA channel thread in the Non-secure state attempts to program its CCRn
register and generate a secure AXI transaction.

■ A DMA channel thread in the Non-secure state executes DMAWFE or DMASEV for an
event that is set as secure. The boot_irq_ns memory-mapped control signals
initialize the security state for an event.
1 For each event, the INTEN register controls if the DMAC generates an event
or signals an interrupt.
■ A DMA channel thread attempts to execute DMAST but the DMAC calculates that
when it eventually performs the store, the MFIFO buffer contains insufficient data
to enable it to complete the store.
■ A DMA channel thread in the Non-secure state executes DMAWFP, DMALDP, DMASTP, or
DMAFLUSHP for a peripheral request interface that is set as secure. The
boot_periph_ns memory-mapped control signals initialize the security state for a
peripheral request interface.
■ A DMA manager thread in the Non-secure state executes DMAGO to attempt to start
a secure DMA channel thread.
■ The DMAC receives an ERROR response on the AXI master interface when it
performs an instruction fetch.
■ A thread executes an undefined instruction.
■ A thread executes an instruction with an operand that is invalid for the
configuration of the DMAC.
1 When the DMAC signals a precise abort, the instruction that triggers the abort is not
executed. Instead, the DMAC executes a DMANOP.
The DMAC signals an imprecise abort under the following conditions:

performs a data load
performs a data store
■ A DMA channel thread executes DMALD or DMAST, and the MFIFO buffer is too small
to hold the required amount of data
■ A DMA channel thread executes DMAST but the thread has not executed sufficient
DMALD instructions
■ A DMA channel thread locks up because of resource starvation, and this causes
the internal watchdog timer to time out.
Watchdog Abort
The DMAC can lock up if one or more DMA channel programs are running and the
MFIFO buffer is too small to satisfy the storage requirements of the DMA programs.
The DMAC contains logic to prevent it from remaining in a state where it is unable to
complete a DMA transfer.
The DMAC detects a lock up when all of the following conditions occur:
■ Load queue is empty

■ Store queue is empty

■ All of the running channels are prevented from executing a DMALD instruction
either because the MFIFO buffer does not have sufficient free space or another
channel owns the load-lock.
When the DMAC detects a lockup it signals an interrupt and can also abort the
contributing channels. The DMAC behavior depends on the state of the wd_irq_only
bit in the WD register.
If:
■ wd_irq_only=0—The DMAC aborts all of the contributing DMA channels and sets
the irq_abort signal high.
■ wd_irq_only=1—The DMAC sets the irq_abort signal high.
For more information, refer to “Resource Sharing Between DMA Channels” on
page 16–26.
Abort Handling
The architecture of the DMAC is not designed to recover from an abort and you must
therefore use an external agent, such as a microprocessor, to terminate a thread when
an abort occurs.

Figure 16–4 shows the operating states for the DMA channel and DMA manager
threads after an abort occurs.
Figure 16–4. Abort Process
DMA channel thread DMA manager thread
Executing Executing
program thread program thread
No Has an abort No Has an abort

occurred? occurred?
Yes Yes
Thread moves to the

Faulting completing state
Active
No AXI transactions
complete?
Yes
Thread moves to the

Faulting state
DMAKILL
No
executed?
Yes
Thread moves to the

Stopped state
After an abort occurs, the action the DMAC takes depends on the thread type:
■ DMA channel thread—The thread immediately moves to the Faulting completing
state. In this state, the DMAC:
■ Sets the irq_abort signal high
■ Stops executing instructions for the DMA channel
■ Invalidates all cache entries for the DMA channel updates the CPCn register to
contain the address of the aborted instruction provided that the abort is precise
■ Does not generate AXI accesses for any instructions remaining in the read
queue and write queue
■ Permits currently active AXI transactions to complete.
1 After the transactions for the DMA channel complete, the thread moves to
the Faulting state.

■ DMA manager thread—The thread immediately moves to the Faulting state and
the DMAC sets the irq_abort signal high.
The external agent can respond to the assertion of the irq_abort signal by:
■ Reading the status of the FSRD register to determine if the DMA manager is
Faulting. In the Faulting state, the FSRD register provides the cause of the abort.
■ Reading the status of the FSRC register to determine if a DMA channel is
Faulting. In the Faulting state, the FSRC register provides the cause of the abort.
To enable a thread in the Faulting state to move to the Stopped state, the external
agent must:
■ Program the DBGINST0 register with the encoding for the DMAKILL instruction.
■ Write to the DBGCMD register.
1 If the aborted thread is secure, you must use the secure slave interface to
update these
After a thread in the Faulting state executes DMAKILL, it moves to the Stopped state.
Security Usage
When the DMAC exits from reset, the status of the configuration signals configures
the security for:
■ DMA manager thread—The DNS bit in the DSR register returns the security state of
the DMA manager thread.
■ Events and interrupts—The INS bit in the CR3 register returns the security state of
the event-interrupt resources.
■ Peripheral request interfaces—The PNS bit in the CR4 register returns the security
state of these interfaces.
Additionally, each DMA channel thread contains a dynamic non-secure bit, CNS, that
is valid when the channel is not in the Stopped state.
DMA Manager Thread in Secure State

If the DNS bit is 0, the DMA manager thread operates in the Secure state and it only
performs secure instruction fetches. When a DMA manager thread in the Secure state
processes:
■ DMAGO—The DMAC uses the status of the ns bit, to set the security state of the
DMA channel thread by writing to the CNS bit for that channel.
■ DMAWFE—The DMAC halts execution of the thread until the event occurs. When the
event occurs, the DMAC continues execution of the thread, irrespective of the
security state of the corresponding INS bit.
■ DMASEV—The DMAC sets the corresponding bit in the INT_EVENT_RIS register,
irrespective of the security state of the corresponding INS bit.

DMA Manager Thread in Non-Secure State

If the DNS bit is 1, the DMA manager thread operates in the Non-secure state, and it
only performs non-secure instruction fetches. When a DMA manager thread in the
Non-secure state processes:
■ DMAGO—The DMAC uses the status of the ns bit, to control if it starts a DMA
channel thread. If:
■ ns = 0—The DMAC does not start a DMA channel thread and instead it:
■ Executes a NOP.
■ Sets the FSRD register.
■ Sets the dmago_err bit in the FTRD register.
■ Moves the DMA manager to the Faulting state.
■ ns = 1—The DMAC starts a DMA channel thread in the Non-secure state and
programs the CNS bit to be non-secure.
■ DMAWFE—The DMAC uses the status of the corresponding INS bit, in the CR3
register, to control if it waits for the event. If:
■ INS = 0—The event is in the Secure state. The DMAC:
■ Executes a NOP.
■ Sets the mgr_evnt_err bit in the FTRD register.
■ INS = 1—The event is in the Non-secure state. The DMAC halts execution of the
thread and waits for the event to occur.
■ DMASEV—The DMAC uses the status of the corresponding INS bit, in the CR3
register, to control if it creates the event interrupt. If:
■ INS = 0—The event-interrupt resource is in the Secure state. The DMAC:
■ Executes a NOP.
■ Sets the mgr_evnt_err bit in the FTRD register.
■ INS = 1—The event-interrupt resource is in the Non-secure state. The DMAC
creates the event interrupt.
DMA Channel Thread in Secure State

When the CNS bit is 0, the DMA channel thread is programmed to operate in the
Secure state and it only performs secure instruction fetches.
When a DMA channel thread in the Secure state processes the following instructions:
■ DMAWFE—The DMAC halts execution of the thread until the event occurs. When the
event occurs, the DMAC continues execution of the thread, irrespective of the
security state of the corresponding INS bit, in the CR3 register.

■ DMASEV—The DMAC creates the event interrupt, irrespective of the security state
of the corresponding INS bit, in the CR3 register.
■ DMAWFP—The DMAC halts execution of the thread until the peripheral signals a
DMA request. When this occurs, the DMAC continues execution of the thread,
irrespective of the security state of the corresponding PNS bit, in the CR4 register.
■ DMALDP and DMASTP—The DMAC sends a message to the peripheral to
communicate that data transfer is complete, irrespective of the security state of the
corresponding PNS bit, in the CR4 register.
■ DMAFLUSHP—The DMAC clears the state of the peripheral and sends a message to
the peripheral to resend its level status, irrespective of the security state of the
corresponding PNS bit, in the CR4 register.
When a DMA channel thread is in the Secure state, it enables the DMAC to perform
secure and non-secure AXI accesses.
DMA Channel Thread in Non-Secure State

When the CNS bit is 1, the DMA channel thread is programmed to operate in the
Non-secure state and it only performs non-secure instruction fetches.
When a DMA channel thread in the Non-secure state processes the following
instructions:
■ DMAWFE—The DMAC uses the status of the corresponding INS bit, in the CR3
register, to control if it waits for the event. If:
■ INS = 0—The event is in the Secure state. The DMAC:
■ Executes a NOP.
■ Sets the appropriate bit in the FSRC register that corresponds to the DMA
channel number.
■ Sets the ch_evnt_err bit in the FTRn register.
■ Moves the DMA channel to the Faulting completing state.
■ INS = 1—The event is in the Non-secure state. The DMAC halts execution of the
thread and waits for the event to occur.
■ DMASEV—The DMAC uses the status of the corresponding INS bit, in the CR3
register, to control if it creates the event. If:
■ INS = 0—The event-interrupt resource is in the Secure state. The DMAC:
■ Executes a NOP.
channel number.
■ Sets the ch_evnt_err bit in the FTRn register.
■ INS = 1—The event-interrupt resource is in the Non-secure state. The DMAC
creates the event interrupt.

■ DMAWFP—The DMAC uses the status of the corresponding PNS bit, in the CR4
register, to control if it waits for the peripheral to signal a request. If:
■ PNS = 0—The peripheral is in the Secure state. The DMAC:
■ Executes a NOP.
channel number.
■ Sets the ch_periph_err bit in the FTRn register.
■ PNS = 1—The peripheral is in the Non-secure state. The DMAC halts execution
of the thread and waits for the peripheral to signal a request.
■ DMALDP and DMASTP—The DMAC uses the status of the corresponding PNS bit, in
the CR4 register, to control if it sends an acknowledgement to the peripheral. If:
■ Executes a NOP.
channel number.
■ PNS = 1—The peripheral is in the Non-secure state. The DMAC sends a
message to the peripheral to communicate when the data transfer is complete.
■ DMAFLUSHP—The DMAC uses the status of the corresponding PNS bit, in the CR4
register, to control if it sends a flush request to the peripheral. If:
■ Executes a NOP.
channel number.
■ PNS = 1—The peripheral is in the Non-secure state. The DMAC clears the state
of the peripheral and sends a message to the peripheral to resend its level
status.
When a DMA channel thread is in the Non-secure state, and a DMAMOV CCR instruction
attempts to program the channel to perform a secure AXI transaction, the DMAC:
1. Executes a DMANOP.
2. Sets the appropriate bit in the FSRC register that corresponds to the DMA channel
number.
3. Sets the ch_rdwr_err bit in the FTRn register.
4. Moves the DMA channel thread to the Faulting completing state.

Constraints and Limitations of Use

This section describes constraints and use limitations.
DMA Channel Arbitration

The DMAC uses a round-robin scheme to service the active DMA channels. To ensure
that the DMAC continues to service the DMA manager, it always services the DMA
manager prior to servicing the next DMA channel.
It is not possible to alter the arbitration process of the DMAC.
DMA Channel Prioritization

The DMAC responds to all active DMA channels with equal priority. It is not possible
to increase the priority of a DMA channel over any other DMA channels.
Instruction Cache Latency

When a cache miss occurs, the latency to service the request is mainly dependent on
the read latency of the memory containing the DMA code. The latency that the DMAC
adds is minimal.
AXI Data Transfer Size

The DMAC can only perform data accesses up to 64 bits in width. If you program the
src_burst_size or dst_burst_size fields to be larger, the DMAC signals a precise
abort. Refer to “Abort Sources” on page 16–16 for more information.
AXI Bursts Crossing 4 KB Boundaries

The AXI specification does not permit AXI bursts to cross 4 KB address boundaries. If
you program the DMAC with a combination of burst start address, size, and length
that would cause a single burst to cross a 4 KB address boundary, then the DMAC
instead generates a pair of bursts with a combined length equal to that specified. This
operation is transparent to the DMAC channel thread program so that, for example,
the DMAC responds to a single DMALD instruction by generating the appropriate pair
of AXI read bursts.
AXI Burst Types

You can program the DMAC to generate only fixed-address or incrementing-address
burst types for data accesses. It does not generate wrapping-address bursts for data
accesses or for instruction fetches.
AXI Write Addresses

The DMAC can issue up to eight outstanding write addresses. The DMAC does not
issue a write address until it has read in all of the data bytes required to fulfill that
write transaction.
AXI Write Data Interleaving

The DMAC does not generate interleaved write data. All write data beats for one
write transaction are output before any write data beat for the next write transaction.

Programming Restrictions
The following sections describe restrictions that apply when programming the
DMAC.
Fixed Unaligned Bursts

The DMAC does not support fixed unaligned bursts. If you program the following
conditions, the DMAC treats this as a programming error:
■ Unaligned read
■ src_inc field is 0 in the CCRn register.
■ The SARn register contains an address that is not aligned to the size of data that
the src_burst_size field contains.
■ Unaligned write
■ dst_inc field is 0 in the CCRn register.
■ The DARn register contains an address that is not aligned to the size of data that
the dst_burst_size field contains.
Endian Swap Size Restrictions

If you program the endian_swap_size field in the CCRn register, to enable a DMA
channel to perform an endian swap, then you must set the corresponding SARn
register and the corresponding DARn register to contain an address that is aligned to
the size that the endian_swap_size field specifies before executing any DMALD or DMAST
instructions.
1 If you update any of endian_swap_size, SARn, or DARn, for example, using a DMAADDH
SAR instruction, then you must ensure that the SARn and DARn registers contain an
address aligned to the size that the endian_swap_size field specifies before executing
any additional DMALD or DMAST instructions.
If you program the src_inc field in the CCRn register to use a fixed address, you must
program the src_burst_size field to select a burst size that is greater than or equal to
the value that the endian_swap_size field specifies. Similarly, if you program the
dst_inc field to select a fixed destination address, you must program the
dst_burst_size field to select a burst size that is greater than or equal to the value
that the endian_swap_size field specifies.
If you program the dst_inc field in the CCRn register to use an incrementing address,
you must program the CCRn register so that dst_burst_len times dst_burst_size is a
multiple of endian_swap_size. For example, if endian_swap_size = 2, 32-bit, and
dst_burst_size = 1, 2 bytes per beat, then you can program dst_burst_len = 1, 3, 5, ...,
15, that is 2, 4, 6, ..., 16 transfers.
Updating DMA Channel Control Registers During a DMA Cycle

Prior to the DMAC executing a sequence of DMALD and DMAST instructions, the values
you program in to the CCRn register, SARn register, and DARn register control the data
byte lane manipulation that the DMAC performs when it transfers the data from the
source address to the destination address.

You can update these registers during a DMA cycle but if you change certain register
fields then it can cause the DMAC to discard data. The following sections describe the
register fields that might have a detrimental impact on a data transfer.
Updates that affect the destination address

If you use a DMAMOV instruction to update the DARn register or CCRn register part
way through a DMA cycle then this might cause a discontinuity in the
destination data stream.
A discontinuity occurs if you change any of the following:
■ endian_swap_size field.
■ dst_inc bit.
■ dst_burst_size field when dst_inc = 0, that is, fixed-address burst.
■ DARn register so that it modifies the destination byte lane alignment. Because the
bus width is 64 bits, you change bits [2:0] in the DARn register.
When a discontinuity in the destination data stream occurs, the DMAC:
1. Halts execution of the DMA channel thread.
2. Completes all outstanding read and write operations for the channel. That is, as if
the DMAC were executing DMARMB and DMAWMB.
3. Discards any residual MFIFO buffer data for the channel.
4. Resumes execution of the DMA channel thread.
Updates that affect the source address

If you use a DMAMOV instruction to update the SARn register or CCRn register part way
through a DMA cycle then this might cause a discontinuity in the source data stream.
A discontinuity occurs if you change any of the following:
■ src_inc bit.
■ src_burst_size field.
■ SARn register so that it modifies the source byte lane alignment. Because the bus
width is 64 bits, you change bits [2:0] in the SARn register.
When a discontinuity in the source data stream occurs, the DMAC:
1. Halts execution of the DMA channel thread.
2. Completes all outstanding read operations for the channel. That is, as if the DMAC
were executing DMARMB
3. Resumes execution of the DMA channel thread. No data is discarded from the
MFIFO buffer.
Resource Sharing Between DMA Channels

DMA channel programs share the MFIFO buffer data storage resource. You must not
start a set of concurrently running DMA channel programs with a resource
requirement that exceeds 512, the size of the MFIFO buffer. If you exceed this limit
then the DMAC might lock up and generate a Watchdog abort, refer to “Watchdog
Abort” on page 16–17.

DMA Controller Programming Model
The DMAC includes a mechanism called the load-lock to ensure that the shared MFIFO
buffer resource is used correctly. The load-lock is either owned by one channel, or it is
free. The channel that owns the load-lock can execute DMALD instructions successfully.
A channel that does not own the load-lock pauses at a DMALD instruction until it takes
ownership of the load-lock.
A channel claims ownership of the load lock when:
■ It executes a DMALD or DMALDP instruction
■ No other channel currently owns the load-lock.
A channel releases ownership of the load-lock when any of the following occur:
■ It executes a DMAST, DMASTP, or DMASTZ
■ It reaches a barrier, that is, it executes DMARMB or DMAWMB
■ It waits, that is, it executes DMAWFP or DMAWFE
■ It terminates normally, that is, it executes DMAEND
■ It aborts for any reason, including DMAKILL.
The MFIFO buffer resource usage of a DMA channel program is measured in MFIFO
buffer entries, and rises and falls as the program proceeds. The MFIFO buffer resource
requirement of a DMA channel program is described using a static requirement and a
dynamic requirement which are affected by the load-lock mechanism.
The static requirement is defined to be the maximum number of MFIFO buffer entries
that a channel is currently using before that channel does one of the following:
■ Executes a WFP or WFE instruction
■ Claims ownership of the load-lock.
The dynamic requirement is defined to be the difference between the static
requirement and the maximum number of MFIFO buffer entries that a channel
program uses at any time during its
To calculate the total MFIFO buffer requirement, add the largest dynamic requirement
to the sum of all the static requirements.
To avoid DMAC lockup, the total MFIFO buffer requirement of the set of channel
programs must be equal to or less than 512, the MFIFO buffer depth.
For more information, refer to “MFIFO Buffer Usage Overview” on page 16–47.

This section describes the instruction set of the DMAC.
Instruction Syntax Conventions

The following conventions are used in assembler syntax prototype lines and their
subfields:
■ < >
Any item bracketed by < and > is mandatory. A description of the item and of how
it is encoded in the instruction is supplied by subsequent text.

■ [ ]
Any item bracketed by [ and ] is optional. A description of the item and of how its
presence or absence is encoded in the instruction is supplied by subsequent
■ Spaces
Single spaces are used for clarity, to separate items. When a space is obligatory in
the assembler syntax, two or more consecutive spaces are used.
Instruction Set Summary

The DMAC instructions:
■ Use a DMA prefix, to provide a unique name-space
■ Have 8-bit opcodes that might use a variable data payload of 0, 8, 16, or 32 bits
■ Use suffixes that are consistent.
Table 16–2 shows a summary of the instruction syntax.
Table 16–2. Instruction Syntax Summary

DMA Manager DMA Channel
Mnemonic Instruction Description
Usage Usage
DMAADDH Add Halfword No Yes Refer to “DMAADDH” on page 16–29.
DMAADNH Add Negative Halfword No Yes Refer to “DMAADNH” on page 16–29.
DMAEND End Yes Yes Refer to “DMAEND” on page 16–30.
DMAFLUSHP Flush and Notify Peripheral No Yes Refer to “DMAFLUSHP” on page 16–31.
DMAGO Go Yes No Refer to “DMAGO” on page 16–31.
DMAKILL Kill Yes Yes Refer to “DMAKILL” on page 16–32.
DMALD Load No Yes Refer to “DMALD[S | B]” on page 16–33.
DMALDP Load and Notify Peripheral No Yes Refer to “DMALDP<S | B>” on page 16–34.
DMALP Loop No Yes Refer to “DMALP” on page 16–35.
Refer to “DMALPEND[S | B]” on
DMALPEND Loop End No Yes
page 16–36.
DMALPFE Loop Forever No Yes Refer to “DMALPFE” on page 16–38.
DMAMOV Move No Yes Refer to “DMAMOV” on page 16–38.
DMANOP No Operation Yes Yes Refer to “DMANOP” on page 16–39.
DMARMB Read Memory Barrier No Yes Refer to “DMARMB” on page 16–39.
DMASEV Send Event Yes Yes Refer to “DMASEV” on page 16–40.
DMAST Store No Yes Refer to “DMAST[S | B]” on page 16–40.
DMASTP Store and Notify Peripheral No Yes Refer to “DMASTP<S | B>” on page 16–41.
DMASTZ Store Zero No Yes Refer to “DMASTZ” on page 16–42.
DMAWFE Wait For Event Yes Yes Refer to “DMAWFE” on page 16–43.
DMAWFP Wait For Peripheral No Yes Refer to “DMAWFP” on page 16–43.
DMAWMB Write Memory Barrier No Yes Refer to “DMAWMB” on page 16–44.

Instructions
The following sections describe the instructions that a DMAC can execute.
DMAADDH
Add Halfword adds an immediate 16-bit value to the SARn register or DARn register, for
the DMA channel thread. This enables the DMAC to support 2D DMA operations.
1 The immediate unsigned 16-bit value is zero-extended before the DMAC adds it to
the address, using 32-bit addition. The DMAC discards the carry bit so that addresses
wrap from 0xFFFFFFFF to 0x00000000.
Figure 16–5 shows the instruction encoding.
Figure 16–5. DMAADDH Encoding
23 16 15 8 7 6 5 4 3 2 1 0
imm[15:8] imm[7:0] 0 1 0 1 0 1 ra 0
Assembler syntax
DMAADDH <address_register>, <16-bit immediate>
where:
<address_register> Selects the address register to use. It must be either:
SAR SARn register and sets ra to 0
DAR DARn register and sets ra to 1
<16-bit immediate> The immediate value to be added to the <address_register>.
Operation
You can only use this instruction in a DMA channel thread.
DMAADNH
Add Negative Halfword adds an immediate negative 16-bit value to the SARn register
or DARn register, for the DMA channel thread. This enables the DMAC to support 2D
DMA operations, or reading or writing an area of memory in a different order to
naturally incrementing addresses.
1 The immediate unsigned 16-bit value is one-extended to 32 bits, to create a value that
is the two's complement representation of a negative number between -65536 and -1,
before the DMAC adds it to the address using 32-bit addition. The DMAC discards
the carry bit so that addresses wrap from 0xFFFFFFFF to 0x00000000. The net effect is
to subtract between 65536 and 1 from the current value in the source or destination
address register.

Figure 16–6. DMAADNH Encoding
23 16 15 8 7 6 5 4 3 2 1 0
imm[15:8] imm[7:0] 0 1 0 1 1 1 ra 0
Assembler syntax
DMAADNH <address_register>, <16-bit immediate>
where:
<address_register> Selects the address register to use. It must be either:
SAR SARn register and sets ra to 0
DAR DARn register and sets ra to 1
<16-bit immediate> The immediate value to be added to the <address_register>.
1 You should specify the 16-bit immediate as the number that is to be represented in the
instruction encoding. For example, DMAADNH DAR, 0xFFF0 causes the value 0xFFFFFFF0
to be added to the current value of the Destination Address register, effectively
subtracting 16 from the DAR.
Operation
DMAEND
End signals to the DMAC that the DMA sequence is complete. After all DMA
transfers are complete for the DMA channel, the DMAC moves the channel to the
Stopped state. It also flushes data from the MFIFO buffer and invalidates all cache
entries for the thread.
Figure 16–7. DMAEND Encoding
7 6 5 4 3 2 1 0
0 0 0 0 0 0 0 0
Assembler syntax
DMAEND
Operation
You can use the instruction with the DMA manager thread and the DMA channel
thread.

DMAFLUSHP
Flush Peripheral clears the state in the DMAC that describes the contents of the
peripheral and sends a message to the peripheral to resend its level status.
Figure 16–8. DMAFLUSHP Encoding
15 11 10 9 8 77 6 5 4 3 2 1 0
periph[4:0] 0 0 0 0 0 1 1 0 1 0 1
Assembler syntax
DMAFLUSHP <peripheral>
where:
<peripheral> 5-bit immediate, value 0-31
Operation
DMAGO
When the DMA manager executes Go for a DMA channel that is in the Stopped state,
it performs the following steps on the DMA channel:
1. Moves a 32-bit immediate into the program counter
2. Sets its security state
3. Updates it to the Executing state.
1 If a DMA channel is not in the Stopped state when the DMA manager executes DMAGO
then the DMAC does not execute DMAGO but instead it executes DMANOP.
Figure 16–9. DMAGO Encoding
15 14 13 12 11 10 8 77 6 5 4 3 2 1 0
0 0 0 0 0 cn[2:0] 1 0 1 0 0 0 ns 0
47 16
imm[31:0]

Assembler syntax
DMAGO <channel_number>, <32-bit_immediate> [, ns]
where:
<channel_number> Selects a DMA channel. It must be one of:
C0 DMA channel 0
C1 DMA channel 1
C2 DMA channel 2
C3 DMA channel 3
C4 DMA channel 4
C5 DMA channel 5
C6 DMA channel 6
C7 DMA channel 7
1 If you provide a channel number that is not available for your configuration
of the DMAC, the DMA manager thread aborts.
<32-bit_immediate> The immediate value that is written to the CPCn register for the
selected <channel_number>.
[ns]
■ If ns is present, the DMA channel operates in the Non-secure state.
■ Otherwise, the execution of the instruction depends on the security state of the
DMA manager:
DMA manager is in the Secure state—DMA channel operates in the Secure
state.
DMA manager is in the Non-secure state—The DMAC aborts.
Operation
You can only use this instruction with the DMA manager thread.
DMAKILL
Kill instructs the DMAC to immediately terminate execution of a thread. Depending
on the thread type, the DMAC performs the following steps:
DMA Manager Thread
1. Invalidates all cache entries for the DMA manager.
2. Moves the DMA manager to the Stopped state.
DMA Channel Thread
1. Moves the DMA channel to the Killing state.
2. Waits for AXI transactions, with an ID equal to the DMA channel number, to
complete.
3. Invalidates all cache entries for the DMA channel.

4. Remove all entries in the MFIFO buffer for the DMA channel.
5. Remove all entries in the read buffer queue and write buffer queue for the DMA
channel.
6. Moves the DMA channel to the Stopped state.
Figure 16–10. DMAKILL Encoding
7 6 5 4 3 2 1 0
0 0 0 0 0 0 0 1
Assembler syntax
DMAKILL
Operation
thread.
1 You must not use the DMAKILL instruction in DMA channel programs. To issue a
DMAKILL instruction, use the DBGINST0 register.
DMALD[S | B]
Load instructs the DMAC to perform a DMA load, using AXI transactions that the
source address registers and channel control registers specify. It places the read data
into the MFIFO buffer and tags it with the corresponding channel number. DMALD is an
unconditional instruction but DMALDS and DMALDB are conditional on the state of the
request_type flag. If the src_inc bit in the channel control registers is set to
incrementing, the DMAC updates the source address registers after it executes
DMALD[S|B].
1 The DMAC sets the value of request_type when it executes a DMAWFP instruction.
Refer to “DMAWFP” on page 16–43.
Figure 16–11. DMALD[S|B] Encoding
7 6 5 4 3 2 1 0
0 0 0 0 0 1 bs x
Assembler syntax
DMALD[S|B]
where:
[S] If S is present, the assembler sets bs to 0 and x to 1. The instruction is conditional
on the state of the request_type flag:

■ request_type = Single
The DMAC performs a DMALD instruction and it sets arlen[3:0]=0x0 so that the
AXI read transaction length is one. The DMAC ignores the value of the
src_burst_len field in the channel control registers.
■ request_type = Burst
The DMAC performs a DMANOP instruction. The DMAC increments the channel PC
to the next instruction. No state change occurs.
[B] If B is present, the assembler sets bs to 1 and x to 1. The instruction is conditional
The DMAC performs a DMANOP instruction. The DMAC increments the channel PC
to the next instruction. No state change occurs.
The DMAC performs a DMALD.
If you do not specify the S or B operand, the assembler sets bs to 0 and x to 0, and the
DMAC always executes a DMA load.
Operation
You can only use this instruction in a DMA channel thread. If you specify the S or B
operand, execution of the instruction is conditional on the state of request_type
matching that of the instruction.
DMALDP<S | B>
Load and notify Peripheral instructs the DMAC to perform a DMA load, using AXI
transactions that source address registers and channel control registers specify. It
places the read data into a FIFO buffer that is tagged with the corresponding channel
number and after it receives the last data item, it sends an acknowledgement to the
peripheral that the data transfer is complete. If the src_inc bit in the channel control
registers is set to incrementing, the DMAC updates source address registers after it
executes DMALDP<S|B>.
Figure 16–12. DMALDP<S|B> Encoding
15 11 10 9 8 77 6 5 4 3 2 1 0
periph[4:0] 0 0 0 0 0 1 0 0 1 bs 1
Assembler syntax
DMALDP<S|B> <peripheral>
where:
<S> When S is present, the assembler sets bs to 0. The instruction is conditional on the
state of the request_type flag:

The DMAC performs a DMALDP instruction and it sets arlen[3:0]=0x0 so that the
AXI read transaction length is one. The DMAC ignores the value of the
src_burst_len field in the channel control registers.
The DMAC performs a DMANOP.
<B> When B is present, the assembler sets bs to 1. The instruction is conditional on the
state of the request_type flag:
The DMAC performs a DMANOP.
The DMAC performs a load using a burst DMA transfer.
<peripheral> 5-bit immediate, value 0-31.
1 The DMAC sets the value of the request_type flag when it executes a DMAWFP
instruction. Refer to “DMAWFP” on page 16–43.
Operation
You can only use this instruction in a DMA channel thread. Execution of the
instruction is conditional on the state of the request_type flag matching that of the
instruction.
DMALP
Loop instructs the DMAC to load an 8-bit value into the loop counter register you
specify.
This instruction indicates the start of a section of instructions, and you set the end of
the section using the DMALPEND instruction. Refer to “DMALPEND[S | B]” on
page 16–36. The DMAC repeats the set of instructions that you insert between DMALP
and DMALPEND until the value in the loop counter register reaches zero.
1 The DMAC saves the value of the PC for the instruction that follows DMALP. After the
DMAC executes DMALPEND, and the loop counter register is not zero, this enables it to
execute the first instruction in the loop.
Figure 16–13. DMALP Encoding
15 8 7 6 5 4 3 2 1 0
iter[7:0] 0 0 1 0 0 0 lc 0

Assembler syntax
DMALP <loop_iterations>
where:
<loop_iterations>
Specifies the number of loops to perform, range 1-256.
■ The assembler determines the loop counter register to use and either:
■ Sets lc to 0, and the DMAC writes the value loop_iterations minus 1 to the
loop counter 0 registers
■ Sets lc to 1, and the DMAC writes the value loop_iterations minus 1 to the
loop counter 1 registers.
Operation
DMALPEND[S | B]
Loop End indicates the last instruction in the program loop but the behavior of the
DMAC depends on whether DMALP or DMALPFE starts the loop. If a loop starts with:
■ DMALP The loop has a defined loop count and DMALPEND[S|B] instructs the DMAC
to read the value of the loop counter register. If a loop counter register returns:
■ Zero—The DMAC executes a DMANOP and therefore exits the loop.
■ Nonzero—The DMAC decrements the value in the loop counter register and
updates the thread PC to contain the address of the first instruction in the
program loop, that is, the instruction that follows the DMALP.
■ DMALPFE The loop has an undefined loop count and the DMAC uses the state of the
request_last flag to control when it exits the loop. If the request_last flag is:
■ 0—The DMAC updates the thread PC to contain the address of the first
instruction in the program loop, that is, the instruction that follows the DMALP.
■ 1—The DMAC executes a DMANOP and therefore exits the loop.
Figure 16–14. DMALPEND[S|B] Encoding
15 8 7 6 5 4 3 2 1 0
backwards_jump[7:0] 0 0 1 nf 1 lc bs x
Assembler syntax
DMALPEND[S|B]
where:
[S] If S is present and the loop starts with DMALP, then the assembler sets bs to 0 and x
to 1. The instruction is conditional on the state of the request_type flag:

■ The DMAC executes the DMALPEND.
■ The DMAC performs a DMANOP and therefore exits the loop.
[B] If B is present and the loop starts with DMALP, then the assembler sets bs to 1 and x
to 1. The instruction is conditional on the state of the request_type flag:
■ The DMAC performs a DMANOP and therefore exits the loop.
■ The DMAC executes the DMALPEND
If you do not specify the S or B operand, the assembler sets bs to 0 and x to 0, and the
DMAC always executes the DMALPEND.
1 You must not specify the S or B operand when a loop starts with DMALPFE. If you do,
the assembler issues a warning message and sets bs to 0, x to 0, and nf to 1. In the
same way as for DMALPFE, the DMAC uses the state of the request_last flag to control
when it exits the loop.
1 The DMAC sets the value of the:
■ request_type flag when it executes a DMAWFP instruction. Refer to “DMAWFP” on

page 16–43.
■ request_last flag to 1 when the corresponding peripheral issues the last request
command through the peripheral request interface. For more information, refer to
“Peripheral Length Management” on page 16–12.
To correctly assign the additional bits in the DMALPEND instruction, that Figure 16–14
shows, the assembler determines the values for:
backwards_jump[7:0] Sets the relative location of the first instruction in the program
loop. The assembler calculates the value for backwards_jump[7:0] by subtracting the
address of the first instruction in the loop from the address of the DMALPEND.
■ nf sets it to:
■ 0 if DMALPFE started the program loop
■ 1 if DMALP started the program loop.
■ lc sets it to:
■ 0 if the loop counter 0 registers contains the loop counter value
■ 1 if the loop counter 1 registers contains the loop counter value
■ 1 if DMALPFE started the program loop.
Operation
operand, execution of the instruction is conditional on the state of the request_type
flag matching that of the instruction.

DMALPFE
The assembler uses Loop Forever to configure certain bits in DMALPEND. Refer to
“DMALPEND[S | B]” on page 16–36.
1 When the assembler encounters DMALPFE, it does not create an instruction for the
DMAC, but instead, it modifies the behavior of DMALPEND. The insertion of DMALPFE in
program code identifies the start of the loop.
Assembler syntax
DMALPFE
DMAMOV
Move instructs the DMAC to move a 32-bit immediate into the following registers:
■ source address registers
■ destination address registers
■ channel control registers
Figure 16–15. DMAMOV Encoding
15 14 13 12 11 10 8 77 6 5 4 3 2 1 0
0 0 0 0 0 rd[2:0] 1 0 1 1 1 1 0 0
47 16
imm[31:0]
Assembler syntax
DMAMOV <destination_register>, <32-bit_immediate>
where:
<destination_register>
The valid registers are:
■ SAR—selects the source address registers and sets rd to b000
■ CCR—selects the channel control registers and sets rd to b001
■ DAR—selects the destination address registers and sets rd to b010
<32-bit_immediate>
A 32-bit value that is written to the specified destination register.
1 For information about using the assembler to program the various fields that the
channel control registers, refer to “DMAMOV CCR” on page 16–46.

Operation
DMANOP
No Operation does nothing. You can use this instruction for code alignment purposes.
Figure 16–16. DMANOP Encoding
7 6 5 4 3 2 1 0
0 0 0 1 1 0 0 0
Assembler syntax
DMANOP
Operation
thread.
DMARMB
Read Memory Barrier forces the DMA channel to wait until all of the executed DMALD
instructions for that channel have been issued on the AXI master interface and have
completed.
This enables write-after-read sequences to the same address location with no hazards.
Figure 16–17. DMARMB Encoding
7 6 5 4 3 2 1 0
0 0 0 1 0 0 1 0
Assembler syntax
DMARMB
Operation

DMASEV
Send Event instructs the DMAC to modify an event-interrupt resource. Depending on
how you program the interrupt enable register, this either:
■ Generates event <event_num>
1 Typically, you use DMAWFE to stall a thread and then another thread executes
DMASEV, using the appropriate event number, to unstall the waiting thread.
Refer to “Using an Event to Restart DMA Channels” on page 16–15.
■ Signals an interrupt using irq<event_num>.

Figure 16–18. DMASEV Encoding
15 11 10 9 8 7 6 5 4 3 2 1 0
event_num[4:0] 0 0 0 0 0 1 1 0 1 0 0
Assembler syntax
DMASEV <event_num>
where:
<event_num> 5-bit immediate, value 0-31
1 The DMAC aborts the thread if you select an event number that is not available.
Operation
thread. For more information, refer to “Using Events and Interrupts” on page 16–15.
DMAST[S | B]
Store instructs the DMAC to transfer data from the FIFO buffer to the location that the
destination address registers specifies, using AXI transactions that the DA register
and channel control registers specify. If the dst_inc bit in the channel control registers
is set to incrementing, the DMAC updates the destination address registers after it
executes DMAST[S|B].
Figure 16–19. DMAST[S|B] Encoding
7 6 5 4 3 2 1 0
0 0 0 0 1 0 bs x

Assembler syntax
DMAST[S|B]
where:
[S] If S is present, the assembler sets bs to 0 and x to 1. The instruction is conditional
■ The DMAC performs a DMAST instruction and it sets awlen[3:0]=0x0 so that the
AXI write transaction length is one. The DMAC ignores the v value of the
dst_burst_len field in the channel control registers.
■ The DMAC performs a DMANOP instruction. The DMAC increments the channel
PC to the next instruction. No state change occurs.
[B] If B is present, the assembler sets bs to 1 and x to 1. The instruction is conditional
■ The DMAC performs a DMANOP instruction. The DMAC increments the channel
PC to the next instruction. No state change occurs.
■ The DMAC performs a DMAST.
■ If you do not specify the S or B operand, the assembler sets bs to 0 and x to 0,
and the DMAC always executes a DMA store.
Operation
operand, execution of the instruction is conditional on the state of the request_type
flag matching that of the instruction.
The DMAC only commences the burst when the MFIFO buffer contains all of the data
necessary to complete the burst transfer.
DMASTP<S | B>
Store and notify Peripheral instructs the DMAC to transfer data from the FIFO buffer
to the location that the destination address registers specifies, using AXI transactions
that the DA register and channel control registers specify. It uses the DMA channel
number to access the appropriate location in the FIFO buffer. After the DMA store is
complete, and the DMAC has received a buffered write response, it issues an
acknowledgement to the peripheral that the data transfer is complete. If the dst_inc
bit in the channel control registers is set to incrementing, the DMAC updates the
destination address registers after it executes DMASTP<S|B>.

Figure 16–20. DMASTP<S|B> Encoding
15 11 10 9 8 77 6 5 4 3 2 1 0
periph[4:0] 0 0 0 0 0 1 0 1 0 bs 1
Assembler syntax
DMASTP<S|B> <peripheral>
where:
<S> Sets bs to 0. This instructs the DMAC to perform:
■ A single DMA store operation if request_type is programmed to Single
1 The DMAC ignores the state of the dst_burst_len field in the channel
control registers and always performs an AXI transfer with a burst length of
one.
■ A DMANOP if request_type is programmed to Burst.

<B> Sets bs to 1. This instructs the DMAC to perform:
■ The DMA store if request_type is programmed to Burst
■ A DMANOP if request_type is programmed to Single.
<peripheral> 5-bit immediate, value 0-31.
Operation
The DMAC only commences the burst when the MFIFO buffer contains all of the data
necessary to complete the burst transfer.
DMASTZ
Store Zero instructs the DMAC to store zeros, using AXI transactions that the
destination address registers and channel control registers specify. If the dst_inc bit in
the channel control registers is set to incrementing, the DMAC updates the
destination address registers after it executes DMASTZ.
Figure 16–21. DMASTZ Encoding
7 6 5 4 3 2 1 0
0 0 0 0 1 1 0 0

Assembler syntax
DMASTZ
Operation
DMAWFE
Wait For Event instructs the DMAC to halt execution of the thread until the event, that
<event_num> specifies, occurs. When the event occurs, the thread moves to the
Executing state and the DMAC clears the event. Refer to “Using Events and
Interrupts” on page 16–15.
Figure 16–22. DMAWFE Encoding
15 11 10 9 8 7 6 5 4 3 2 1 0
event_num[4:0] 0 i 0 0 0 1 1 0 1 1 0
Assembler syntax
DMAWFE <event_num>[, invalid]
where:
<event_num> 5-bit immediate, value 0-31
[invalid] Sets i to 1. If invalid is present, the DMAC invalidates the instruction
cache for the current DMA thread. If invalid is not present, then the assembler sets i
to 0 and the DMAC does not invalidate the instruction cache for the current DMA.
1 The DMAC aborts the thread if you select an event number that is not
available for your configuration of the DMAC.
To ensure cache coherency, you must use invalid when a processor writes
the instruction stream for a DMA channel.
Operation
thread.
DMAWFP
Wait For Peripheral instructs the DMAC to halt execution of the thread until the
specified peripheral signals a DMA request for that DMA channel.

Figure 16–23. DMAWFP Encoding
15 11 10 9 8 7 6 5 4 3 2 1 0
peripheral[4:0] 0 0 0 0 0 1 1 0 0 bs p
Assembler syntax
DMAWFP <peripheral>, <single|burst|periph>
where:
<peripheral> 5-bit immediate, value 0-31
1 The DMAC aborts the thread if you select a peripheral number that is not
available.
<single> Sets bs to 0 and p to 0. This instructs the DMAC to continue executing the
DMA channel thread after it receives a single or burst DMA request. The DMAC sets
the request_type to Single, for that DMA channel.
<burst> Sets bs to 1 and p to 0. This instructs the DMAC to continue executing the
DMA channel thread after it receives a burst DMA request. The DMAC sets the
request_type to Burst.
1 The DMAC ignores single burst DMA requests.
<periph> Sets bs to 0 and p to 1. This instructs the DMAC to continue executing the
DMA channel thread after it receives a single or burst DMA request. The DMAC sets
the request_type to:
Single When it receives a single DMA request.
Burst When it receives a burst DMA request.
Operation
DMAWMB
Write Memory Barrier forces the DMA channel to wait until all of the executed DMAST
instructions for that channel have been issued on the AXI master interface and have
completed.
This permits read-after-write sequences to the same address location with no hazards.
Figure 16–24. DMAWMB Encoding
7 6 5 4 3 2 1 0
0 0 0 1 0 0 1 1

Assembler syntax
DMAWMB
Operation
Assembler Directives
The assembler provides the following additional commands:
■ DCD
■ DCB
■ DMALP
■ DMALPFE
■ DMAMOV CCR
DCD
Assembler directive to place a 32-bit immediate in the instruction stream.
Syntax
DCD imm32
DCB
Assembler directive to place an 8-bit immediate in the instruction stream.
Syntax
DCB imm8
DMALP
Assembler directive to insert an iterative loop.
Syntax
DMALP [<LC0>|<LC1>] <loop_iterations>
where:
<loop_iterations>
An 8-bit value that specifies the number of loops to perform.
1 For clarity in writing assembler instructions, the 8-bit value is the actual
number of iterations of the loop to be executed. The assembler decrements
this by one to create the actual value, 0-255, that the DMAC uses.
[LC0] If LC0 is present, the DMAC stores <loop_iterations> in the loop counter 0
registers.
[LC1] If LC1 is present, the DMAC stores <loop_iterations> in the loop counter 1
registers.

1 If LC0 or LC1 is not present, the assembler determines the loop counter register to use.
DMALPFE
Assembler directive to insert a repetitive loop.
Syntax
DMALPFE
Enables the assembler to clear the nf bit that is present in DMALPEND. Refer to
“DMALPEND[S | B]” on page 16–36.
DMAMOV CCR
Assembler directive that enables you to program the channel control registers using
the specified format.
Syntax
DMAMOV CCR,
[SB<1-16>] [SS<8|16|32|64|128>] [SA<I|F>]
[SP<imm3>] [SC<imm4>]
[DB<1-16>] [DS<8|16|32|64|128>] [DA<I|F>]
[DP<imm3>] [DC<imm4>]
[ES<8|16|32|64|128>]
Table 16–3 shows the argument descriptions and the default values.
Table 16–3. DMAMOV CCR argument description and the default values
Syntax Description Options Default
I = Increment
SA Source address increment. Sets the value of arburst[0] I
F = Fixed
SS Source burst size in bits. Sets the value of arsize[2:0] 8, 16, 32, or 64 8
SB Source burst length. Sets the value of arlen[3:0] 1 to 16 1
SP Source protection 0 to 7 (1) 0
SC Source cache 0 to 15 (1) (2) 0
I = Increment
DA Destination address increment. Sets the value of awburst[0] I
F = Fixed
DS Destination burst size in bits. Sets the value of awsize[2:0] 8, 16, 32, or 64 8
DB Destination burst length. Sets the value of awlen[3:0] 1 to 16 1
DP Destination protection 0 to 7 (1) 0
DC Destination cache 0 to 15 (1) (3) 0
ES Endian swap size, in bits 8, 16, 32, or 64 8
(1) You must use decimal values when programming this immediate value
(2) Because the DMAC ties ARCACHE[3] LOW, the assembler always sets bit 3 to 0 and uses bits [2:0] of your chosen value for SC.
(3) Because the DMAC ties AWCACHE[2] LOW, the assembler always sets bit 2 to 0 and uses bit [3] and bits [1:0] of your chosen value for DC.

MFIFO Buffer Usage Overview

This section shows MFIFO buffer usage for some example DMA channel programs.
About MFIFO Buffer Usage Overview

The MFIFO buffer is a shared resource that is utilized on a first-come, first-served
basis by all currently active channels. To a program, it appears as a set of
variable-depth parallel FIFO buffers, one per channel, with the restriction that the
total depth of all the Fifes cannot exceed the buffer depth, 512. The width of the AXI
master interface is the same as the MFIFO buffer width.
The DMAC is capable of realigning data from the source to the destination. For
example, the DMAC shifts the data by two byte lanes when it reads a word from
address 0x103 and writes to address 0x205. All byte manipulations occur when data
enters the MFIFO buffer, as a result of an AXI read due to a DMALD instruction, so that
the DMAC does not need to manipulate the data when it removes it from the MFIFO
buffer, as a result of an AXI write due to a DMAST instruction. Therefore the storage and
packing of the data in the MFIFO buffer is determined by the destination address and
transfer characteristics.
When a program specifies that incrementing transactions are to be performed to the
destination, the DMAC packs data into the MFIFO buffer to minimize the usage of the
MFIFO buffer entries. For example, the DMAC packs two 32-bit words into a single
entry in the MFIFO buffer when the DMAC has a 64-bit AXI data bus and the
program uses a source address of 0x100, and destination address of 0x200.
In certain situations, the number of entries required to store the data loaded from a
source is not a simple calculation of amount of source data divided by MFIFO buffer
width. The calculation of the number of entries required is not simple when any of the
following occur:
■ The source address is not aligned to the AXI bus width.
■ The destination address is not aligned to the AXI bus width.
■ The transactions are to a fixed destination, that is, a non-incrementing address.
The DMALD and DMAST instructions each specify that an AXI transaction is to be
performed. The amount of data transferred by an AXI transaction depends on the
values programmed in to the CCRn register and the address of the transaction.
f For information about unaligned transfers, refer to the AMBA AXI Protocol
Specification v1.0, available from the ARM website (infocenter.arm.com).
The following sections provide several example DMAC programs together with
illustrations of the MFIFO buffer usage.
1 These sections show MFIFO buffer usage in the following ways:
■ a graph of the number of MFIFO buffer entries versus time

■ a diagram of the byte-lane manipulation that the DMAC performs when data
enters the MFIFO buffer.

1 The numbers 0 and 7 in the MFIFO buffer diagrams indicate the byte lanes in the
MFIFO buffer.
Aligned Transfers
The following sections show examples of aligned transfers.
Simple Aligned Program

In this program, the source address and destination address are aligned with the AXI
data bus width.
Example 16–2. Simple Aligned Program

DMAMOV CCR, SB4 SS64 DB4 DS64
DMAMOV SAR, 0x1000
DMAMOV DAR, 0x4000
DMALP 16
DMALD ; shown as a in Figure 16–25
DMAST ; shown as b in Figure 16–25
DMALPEND
DMAEND
Figure 16–25 shows the MFIFO buffer usage for this program.
Figure 16–25. Simple Aligned Program
Data from
DMALD DMALD
a a a a 7 0
4
a a a a a a a a
a a a a a a a a
a a a a a a a a
a a a a a a a a
0
b b b b Data for
DMAST
DMAST
In Figure 16–25, each DMALD requires four entries and each DMAST removes four entries.
This example has a static requirement of zero MFIFO buffer entries and a dynamic
requirement of four MFIFO buffer entries.

Aligned Asymmetric Program with Multiple Loads

The following program performs four loads for each store and the source address and
destination address are aligned with the AXI data bus width.
Example 16–3. Aligned Asymmetric Program with Multiple Loads

DMAMOV SAR, 0x1000
DMAMOV DAR, 0x4000
DMALP 16
DMALD ; shown as b in Figure 16–26
DMALD ; shown as c in Figure 16–26
DMALD ; shown as d in Figure 16–26
DMAST ; shown as e in Figure 16–26
DMALPEND
DMAEND
Figure 16–26. Aligned Asymmetric Program with Multiple Loads
Data from
4 × DMALD DMALD
d d d 7 0
4
c c c a a a a a a a a
b b b b b b b b b b b
a a a c c c c c c c c
d d d d d d d d
0
e e e Data for
DMAST DMAST
In Figure 16–26, each DMALD requires one entry and each DMAST removes four entries.
Aligned Asymmetric Program with Multiple Stores

The following program performs four stores for each load and the source address and
destination address are aligned with the AXI data bus width.
Example 16–4. Aligned Asymmetric Program with Multiple Stores

DMAMOV SAR, 0x1000
DMAMOV DAR, 0x4000
DMALP 16
DMAST ; shown as c in Figure 16–27
DMAST ; shown as d in Figure 16–27
DMALPEND
DMAEND

Figure 16–27. Aligned Asymmetric Program with Multiple Stores
Data from
DMALD DMALD
a a a 7 0
4
a a a a a a a a
b b b a a a a a a a a
c c c a a a a a a a a
d d d a a a a a a a a
0
e e e Data for
4 × DMAST DMAST
In Figure 16–27, each DMALD requires four entries and each DMAST removes one entry.
Unaligned Transfers
The following sections show examples of unaligned transfers.
Aligned Source Address to Unaligned Destination Address

In this program, the source address is aligned with the AXI data bus width but the
destination address is unaligned. The destination address is not aligned to the
destination burst size so the first DMAST instruction removes less data than the first
DMALD instruction reads. Therefore, a final DMAST of a single word is required to clear
the data from the MFIFO buffer.
Example 16–5. Aligned Source Address to Unaligned Destination Address

DMAMOV SAR, 0x1000
DMAMOV DAR, 0x4004
DMALP 16
DMALD ; shown as a1, ... a, an in Figure 16–28
DMALPEND
DMAEND

Figure 16–28. Aligned to Unaligned Program
Data from
DMALD DMALD
a1 a a an
5 7 0
a1 a1 a1 a1
a1 a1 a1 a1 a1 a1 a1 a1
Data for a1 a1 a1 a1 a1 a1 a1 a1
1 first DMAST a1 a1 a1 a1 a1 a1 a1 a1
b b b b
0 a a a a a1 a1 a1 a1
c
a a a a a a a a
a a a a a a a a
Data for a a a a a a a a
15 × DMAST .
.
.
an an an an
Data for
last DMAST
DMAST
The first DMALD instruction loads four doublewords but because the destination
address is unaligned, the DMAC shifts them by four bytes and therefore it uses five
entries in the MFIFO buffer.
Each DMAST requires only four entries of data and therefore the extra entry remains in
use for the duration of the program until it is emptied by the last DMAST.
This example has a static requirement of one MFIFO buffer entry and a dynamic
Unaligned Source Address to Aligned Destination Address

In this program, the source address is unaligned with the AXI data bus width but the
destination address is aligned. The source address is not aligned to the source burst
size so the first DMALD instruction reads in less data than the DMAST. Therefore, an extra
DMALD is required to satisfy the first DMAST.
Example 16–6. Unaligned Source Address to Aligned Destination Address

DMAMOV SAR, 0x1004
DMAMOV DAR, 0x4000
DMALP 15
DMALD ; shown as b1, ... b, bn in Figure 16–29
DMALPEND
DMALD ; shown as d in Figure 16–29
DMAEND

Figure 16–29. Unaligned to Aligned Program
Data from
DMALD DMALD
7 0
a a a a a a a a
a a a a a a a a
b1 b bn Data for
8 a a a a a a a a
first DMAST b1 b1 b1 b1 a a a a
b1 b1 b1 b1 b1 b1 b1 b1
a d b1 b1 b1 b1 b1 b1 b1 b1
4 b1 b1 b1 b1 b1 b1 b1 b1
c c c
Data for b b b b b1 b1 b1 b1
14 × DMAST .
0
.
e
.
bn bn bn bn bn bn bn bn
bn bn bn bn bn bn bn bn
Data for bn bn bn bn bn bn bn bn
last DMAST d d d d bn bn bn bn
DMAST
1 The DMALD shown as d does not increase the MFIFO buffer usage because it loads four
bytes into an MFIFO buffer entry that the DMAC has already allocated to this
channel.
The first DMALD instruction does not load sufficient data to enable the DMAC to
execute a DMAST and therefore the program includes an additional DMALD, prior to the
start of the loop. After the first DMALD, the subsequent DMALDs align with the source
burst size. This optimizes the p performance but it requires a larger number of MFIFO
buffer entries.
This example has a static requirement of four MFIFO buffer entries and a dynamic

Unaligned Source Address to Aligned Destination Address with Excess Initial Load
This program is an alternative to that described in “Unaligned Source Address to
Aligned Destination Address” on page 16–51. The program uses a different sequence
of source bursts which might be less efficient but requires fewer MFIFO buffer entries.
Example 16–7. Unaligned Source Address to Aligned Destination Address with Excess Initial
Load
DMAMOV SAR, 0x1004
DMAMOV DAR, 0x4000
DMALP 14
DMALD ; shown as c and cn in Figure 16–30
DMAST ; shown as d in Figure 16–30
DMALPEND
DMALD ; shown as e in Figure 16–30
DMALD ; shown as f in Figure 16–30
DMAST ; shown as g in Figure 16–30
DMAEND
Figure 16–30. Unaligned to Aligned with Excess Initial Load
Data from
DMALD DMALD
7 0
a a a a a a a a
a a a a a a a a
Data for a a a a a a a a
first DMAST a a a a a a a a
a c c cn
5 c c c c a a a a
e f
4 c c c c c c c c
c c c c c c c c
Data for c c c c c c c c
1 14 × DMAST .
b d d d
0
g .
.
e e e e cn cn cn cn
e e e e e e e e
Data for e e e e e e e e
last DMAST f f f f e e e e
DMAST
1 The DMALD shown as f does not increase the MFIFO buffer usage because it loads four
bytes into an MFIFO buffer entry that the DMAC has already allocated to this
channel.

The first DMALD instruction loads five beats of data to enable the DMAC to execute the
first DMAST.
After the first DMALD, the subsequent DMALDs are not aligned to the source burst size, for
example the second DMALD reads from address 0x1028. After the loop, the final two
DMALDs read the data required to satisfy the final DMAST.
This example has a static requirement of one MFIFO buffer entry and a dynamic
Aligned Burst Size Unaligned MFIFO Buffer

In this program, the destination address, which is narrower than the MFIFO buffer
width, aligns with the burst size but does not align with the MFIFO buffer width.
Example 16–8. Aligned Burst Size Unaligned MFIFO Buffer

DMAMOV SAR, 0x1000
DMAMOV DAR, 0x4004
DMALP 16
DMALPEND
DMAEND
Figure 16–31. Aligned Burst with Unaligned MFIFO Buffer Width
Data from
DMALD DMALD
a a a a
3 7 0
a a a a
a a a a a a a a
Data for
a a a a
0 b b b b DMAST
DMAST
In this example, the destination address is not 64-bit aligned, it requires three rather
than the expected two MFIFO buffer entries.
requirement of three MFIFO buffer entries.
Fixed Transfers
The following section shows an example of a fixed destination with aligned address.

Fixed Destination with Aligned Address

In this program, the source address and destination address are aligned with the AXI
data bus width, and the destination address is fixed.
Example 16–9. Fixed Destination with Aligned Address

DMAMOV CCR, SB2 SS64 DB4 DS32 DAF
DMAMOV SAR, 0x1000
DMAMOV DAR, 0x4000
DMALP 16
DMALPEND
DMAEND
Figure 16–32. Fixed Destination with Aligned Address
DMALD Data from

a a a a 7 0 DMALD
4
a a a a
a a a a
a a a a
a a a a Data for
0
b b b b DMAST
DMAST
Each DMALD in the program loads two 64-bit data transfers into the MFIFO buffer.
Because the destination address is a 32-bit fixed address then the DMAC splits each
64-bit data item across two entries in the MFIFO buffer.

DMA Controller Registers
DMA Controller Registers

The register map of the DMAC spans a 4 KB region as shown in Figure 16–33.
Figure 16–33. DMAC Summary Register Map
0xFFF
Component ID
0xFE0
0xE14
Configuration
0xE00
0xD0C
Debug
0xD00
0x4FC
AXI and loop counter status
0x400
0x13C
DMA channel thread status
0x100
0x05C
Control
0x000
In Figure 16–33, the register map consists of the following sections.

■ Control registers—allow you to control the DMAC.
■ DMA channel thread status registers—provide the status of the DMA channel
threads.
■ AXI and loop counter status registers—provide the AXI transfer status and the
loop counter status, for each DMA channel thread.

■ Debug registers—enable the following functionality:

■ Allows you to send instructions to a thread when debugging the program
code.
■ Allows system firmware to send instructions to the DMA manager thread, as
“Issuing Instructions to the DMAC using a Slave Interface” on page 16–9
describes.
■ Configuration registers—enable system firmware to discover the configuration of
the DMAC and control the behavior of the watchdog.
■ Component ID registers— enable system firmware to identify peripherals. Do not
attempt to access reserved or unused address locations. Attempting to access these
locations can result in unpredictable behavior.
Address Map and Register Definitions

■ dmanonsecure
■ dmasecure




17. Ethernet Media Access Controller
November 2012
cv_54017-1.2
cv_54017-1.2
The hard processor system (HPS) provides two Ethernet media access controller
(EMAC) peripherals. Each EMAC can be used to transmit and receive data at
10/100/1000 Mbps over Ethernet connections in compliance with the IEEE 802.3
specification. The EMACs are instances of the Synopsys® DesignWare® 3504-0
Universal 10/100/1000 Ethernet MAC (DWC_gmac).
The EMAC has an extensive memory-mapped control and status register (CSR) set,
which can be accessed by the ARM Cortex™-A9® MPCore™.
For an understanding of this chapter, you should be familiar with the basics of IEEE
802.3 media access control (MAC).
f For complete information about IEEE 802.3 MAC, refer to IEEE Std 802.3-2008 Part 3:
Carrier sense multiple access with Collision Detection (CSMA/CD) Access Method and
Physical Layer Specifications, available on the IEEE website
(standards.ieee.org/findstds/).
Features of the Ethernet MAC

Here is a summary of the features supported by the EMAC peripheral.
MAC
■ IEEE 802.3-2008 compliant
■ Data rates of 10/100/1000 Mbps
■ Full duplex and half duplex modes
■ IEEE 802.3x flow control automatic transmission of zero-quanta pause frame on
flow control input deassertion
■ Optional forwarding of received pause control frames to the user
■ Packet bursting and frame extension in 1000 Mbps half-duplex
■ IEEE 802.3x flow control in full-duplex
■ Back-pressure support for half-duplex
■ IEEE 1588-2002 and IEEE 1588-2008 precision networked clock synchronization
■ IEEE 802.3-az, version D2.0 for Energy Efficient Ethernet (EEE)
■ IEEE 802.1Q virtual local area network (VLAN) tag detection for reception frames
Registered

November 2012
Subscribe
17–2 Chapter 17: Ethernet Media Access Controller
Features of the Ethernet MAC
■ Preamble and start-of-frame data (SFD) insertion in transmit and deletion in

receive paths
■ Automatic cyclic redundancy check (CRC) and pad generation controllable on a
per-frame basis
■ Options for automatic pad/CRC stripping on receive frames
■ Programmable frame length supporting standard and jumbo Ethernet frames
(with size up to 16 KB)
■ Programmable inter-frame gap (IFG), from 40 to 96 bit times in steps of 8
PHY Interface
■ Reduced Gigabit Media Independent Interface (RGMII) for 10/100/1000
■ Management Data Input/Output (MDIO) (IEEE 802.3) or I2C PHY management
interface
DMA Interface
■ Programmable burst size for optimal bus utilization
■ Single-channel mode transmit and receive engines
■ Byte-aligned addressing mode for data buffer support
■ Dual-buffer (ring) or linked-list (chained) descriptor chaining
■ Descriptors can each transfer up to 8 KB of data
Management Interface
■ 32-bit host interface to CSR set
■ Comprehensive status reporting for normal operation and transfers with errors
■ Configurable interrupt options for different operational conditions
■ Per-frame transmit/receive complete interrupt control
■ Separate status returned for transmission and reception packets
Acceleration
■ Transmit and receive checksum offload for transmission control protocol (TCP),
user datagram protocol (UDP), or Internet control message protocol (ICMP) over
Internet protocol (IP)
Other Features
■ Supports a variety of flexible address filtering modes
■ Up to 31 additional 48-bit perfect destination address (DA) filters with masks for
each byte
■ Up to 31 48-bit source address (SA) comparison check with masks for each byte

Chapter 17: Ethernet Media Access Controller 17–3
EMAC Block Diagram and System Integration
■ 256-bit hash filter (optional) for multicast and uni-cast DAs

■ Option to pass all multicast addressed frames
■ Promiscuous mode support to pass all frames without any filtering for network
monitoring
■ Passes all incoming packets (as per filter) with a status report

The EMACs are integrated into the HPS portion of the system on a chip (SoC) FPGA
device. They communicate with the I/O pins. Figure 17–1 on page 17–3 shows the
EMAC integration from a high level point of view.
Figure 17–1. EMAC System Integration
HPS
RMII/RGMII
PHY DMA
EMAC
PHY
MDIO/I2C
MDIO CSR
Multiplexer L3 Interconnect
Logic NIC301
I2C Controller
Clock Manager
Reset Manager
EMAC to RGMII Interface

The PHY datapath I/O is described in Table 17–1.
Table 17–1. External PHY Data Interface (Part 1 of 2)

EMAC Port I/O Width Description
Transmit Clock. This is the transmit clock (125/25/2.5 MHz in 1G/100M/10Mbps)
clk_tx_i In 1 provided for the RGMII. All PHY transmit signals generated by the EMAC are
synchronous to this clock.
PHY Transmit Data.This is a group of eight transmit data signals driven by the MAC.
Unused bits in the RGMII interface configuration are tied to low.
phy_txd_o Out 8 RGMII: Bits [3:0] provide the RGMII transmit data. The data bus changes with both
rising and falling edges of the transmit clock (clk_tx_i). The validity of the data is
qualified with phy_txen_o.
Synchronous to: clk_tx_i, clk_tx_180_i
PHY Transmit Data Enable. This signal is driven by the EMAC component.
phy_txen_o Out 1 RGMII: This signal is the control signal (rgmii_tctl) for the transmit data, and is
driven on both edges of the clock.
Synchronous to: clk_tx_i, clk_tx_180_i

Table 17–1. External PHY Data Interface (Part 2 of 2)

EMAC Port I/O Width Description
rst_clk_tx_n_o Out 1 Transmit clock reset output.
Receive clock. Clock frequency is 125/25/2.5 MHz in 1G/100M/10Mbps modes. It is
clk_rx_i In 1 provided by the external PHY. All PHY signals received by the EMAC are
synchronous to this clock.
PHY Receive Data.This is a bundle of eight data signals received from the PHY.
RGMII: Bits [3:0] provide the RGMII receive data. The data bus is sampled with both
phy_rxd_i In 8 rising and falling edges of the receive clock (clk_rx_i). The validity of the data is
qualified with phy_rxdv_i.
Synchronous to: clk_rx_i, clk_rx_180_i
PHY Receive Data Valid.This signal is driven by PHY.
phy_rxdv_i In 1 RGMII: This is the receive control signal used to qualify the data received on
phy_rxd. This signal is sampled on both edges of the clock.
Synchronous to: clk_rx_i, clk_rx_180_i
rst_clk_rx_n_o Out 1 Receive clock reset output.
PHY Interface Select: These pins select one of the PHY interfaces of the EMAC. This
phy_intf_sel_i[1 is sampled only during reset assertion and ignored after that.
In 2
:0] ■ 01: RGMII
■ 00, 10, and 11: Invalid
This is the reference clock to the EMAC. The clock is emac0_clk or emac1_clk
supplied by the clock manager.
clk_ref_i In 1 The system manager drives the phy_intf_sel signal to control which clock is
used.
The clock rate is 250 MHz.
PHY Management Interface

The HPS can provide support for either MDIO or I2C PHY management interfaces.
MDIO Interface
The MDIO interface signals are synchronous to l4_mp_clk in all supported modes.
Table 17–2. PHY MDIO Management Interface

Signal I/O Width Description
Management Data In.The PHY generates this signal to transfer
gmii_mdi_i In 1 register data during a read operation. This signal is driven
synchronously with the gmii_mdc_o clock.
Management Data Out.The EMAC uses this signal to transfer
gmii_mdo_o Out 1
control and data information to the PHY.

Table 17–2. PHY MDIO Management Interface

Signal I/O Width Description
Management Data Output Enable.This enable signal drives the
gmii_mdo_o signal from an external three-state I/O buffer. This
gmii_mdo_o_e Out 1
signal is asserted whenever valid data is driven on the
gmii_mdo_o signal. The active state of this signal is high.
Management Data Clock. The EMAC provides timing reference
for the gmii_mdi_i and gmii_mdo_o signals on MII through
gmii_mdc_o Out 1 this aperiodic clock. The maximum frequency of this clock is 2.5
MHz. This clock is generated from the application clock through
a clock divider.
I2C External PHY Management Interface

Some PHY devices use the I2C instead of MDIO for their control interface. Small form
factor pluggable (SFP) optical or pluggable modules are often among those with this
interface.
The HPS can use two of the four general purpose I2C peripherals for controlling the
PHY devices.
IEEE 1588
The EMAC supports IEEE 1588 operation in all modes with a resolution of one μs. It
can be used by the ARM® Cortex™-A9 microprocessor unit (MPU) subsystem to
maintain synchronization between the time counters internal to the two MACs.

Functional Description of the EMAC

Figure 17–2 illustrates a high level block diagram of the EMAC with its interfaces.
Figure 17–2. EMAC Block Diagram
EMAC
TX FIFO Buffer RX FIFO Buffer

(DPRAM) (DPRAM)
Master DMA TX FIFO Buffer RX FIFO Buffer PHY

Interface Controller Controller Controller Interface
MAC
Slave DMA Operation Mode

Interface CSRs Register
There are two host interfaces to the MAC. The management host interface, a 32-bit
slave interface, provides access to the CSR set. The data interface is a 32-bit interface.
It controls data transfer between the direct memory access (DMA) controller channels
and the rest of the HPS system through the NIC-301 L3 interconnect.
There is a built-in DMA controller which is optimized for data transfer between the
MAC controller and system memory. The DMA controller has independent transmit
and receive engines, and a CSR set. The transmit engine transfers data from system
memory to the device port, while the receive engine transfers data from the device
port to the system memory. The controller uses descriptors to efficiently move data
from source to destination with minimal host intervention.
The EMAC also contains FIFO buffer memory to buffer and regulate the Ethernet
frames between the application system memory and the EMAC controller. On
transmit, the Ethernet frames read into the transmit FIFO buffer (1024 x 42 bits), and
eventually trigger the MAC to perform the transfer. Received Ethernet frames are
stored in the receive FIFO buffer, also indicating the FIFO buffer fill level to the DMA
controller. The DMA controller then initiates the configured burst transfers. Both
receive and transmit transfer status are taken from the MAC and transferred to the
DMA.

Host Interfaces
There are two host interfaces in the EMAC: a slave and a master. The master is
connected to the L3 master peripheral switch interface in the L3 interconnect block.
Slave
The EMAC CSR set access is provided by a slave interface. The slave is connected to
the level 4 (L4) bus.
Master
The DMA interface is provided by a master interface. Two types of data are
transferred on the interface: data descriptors and actual data packets. The interface is
very efficient in transferring full duplex Ethernet packet traffic. Read and write data
transfers from different DMA channels can be performed simultaneously on this port.
The only exceptions to this are transmit descriptor reads and write-backs which
cannot happen simultaneously.
DMA transfers are split into a software configurable number of burst transactions on
the interface. The AXI_Bus_Mode register in the dmagrp group is used to configure
bursting behavior.
The interface assigns a unique ID for each DMA channel and also for each read DMA
or write DMA request in a channel. Data transfers with distinct IDs can be reordered
and interleaved.
Write data transfers are generally performed as posted writes with OK responses
returned as soon as the interconnect has accepted the last beat of a data burst.
Descriptors (status or timestamp) however are always transferred as non-posted
writes in order to prevent race conditions with the transfer complete interrupt logic.
The slave may issue an error response. When that happens, the EMAC disables the
DMA channel which generated the original request and asserts an interrupt signal.
The host needs to reset the EMAC with a hard or soft reset to restart the DMA to
recover from this condition.
The EMAC supports up to 16 outstanding transactions on the interface. Buffering
outstanding transactions smooths out back pressure behavior. This is important when
resource contention bottlenecks arise under high system load conditions.
Cache Control Interface

The system manager provides the values for the master cache outputs through this
interface. These inputs are used as the outputs to the L3 interconnect extending the
capabilities of this block with respect to the cacheable characteristics of master
transfers.
To configure EMAC DMA controller to perform cacheable accesses, configure the
cache bits in the system manager. Cache bits should only be accessed at boot time,
before the EMAC controller is brought out of reset.

External PHY
The following PHY interfaces are supported for the HPS:
■ RGMII for 10/100/1000
The EMAC also has a control interface used for configuration and status monitoring
of the PHY. In this case, the PHY is the slave device. There are two choices of control
interface:
■ MDIO
■ I2C interface
The MDIO interface is built into the EMAC while the I2C interface uses separate I2C
peripherals residing on the HPS. The interfaces are multiplexed externally to the
EMAC.
Transmit and Receive Data FIFO Buffers

Each EMAC component has associated transmit and receive data FIFO buffer
instances. Both FIFO buffer instances are 1024 x 42 bits. The FIFO buffer word consists
of:
■ Data: 32 bits
■ Sideband:
■ End of frame (EOF): one bit
■ Byte enables (BE): two bits
■ Error correction code (ECC): seven bits
The data and sideband are protected by the seven-bit single error correct, double error
detect (SEC-DED) code word. These FIFO buffer RAMs also contain ECC enable, error
injection and status pins. The enable and error injection pins are inputs driven by the
system manager and the status pins are outputs driven to the MPU subsystem.
IEEE 1588-2002 Time Stamps

The IEEE 1588-2002 standard defines the Precision Time Protocol (PTP) which enables
precise synchronization of clocks in a distributed network of devices. The PTP applies
to systems communicating by local area networks supporting multicast messaging.
This protocol enables heterogeneous systems that include clocks of varying inherent
precision, resolution, and stability to synchronize. It is frequently used in automation
systems where a collection of communicating machines such as robots must be
synchronized and hence operate over a common time base. †

The PTP is transported over UDP/IP. The system or network is classified into Master
and Slave nodes for distributing the timing and clock information. Figure 17–3 shows
the process that PTP uses for synchronizing a slave node to a master node by
exchanging PTP messages. †
Figure 17–3. Networked Time Synchronization
Master Clock TIme Slave Clock Time
Sync Message
t1
Data at
Slave Clock
t2m t2 t2
Follow_Up Message
Containing t1 Value
t1, t2
t3m t3 t1, t2, t3

Delay_Req
Message
t4
Delay_Resp Message
Containing t4 Value
t1, t2, t3, t4

Time
As shown in Figure 17–3, the PTP uses the following process: †

1. The master broadcasts the PTP Sync messages to all its nodes. The Sync message
contains the master’s reference time information. The time at which this message
leaves the master’s system is t1. This time must be captured, for Ethernet ports, at
the PHY interface. †
2. The slave receives the Sync message and also captures the exact time, t2, using its
timing reference. †
3. The master sends a Follow_up message to the slave, which contains t1 information
for later use. †
4. The slave sends a Delay_Req message to the master, noting the exact time, t3, at
which this frame leaves the PHY interface. †
5. The master receives the message, capturing the exact time, t4, at which it enters its
system. †
6. The master sends the t4 information to the slave in the Delay_Resp message. †

7. The slave uses the four values of t1, t2, t3, and t4 to synchronize its local timing
reference to the master’s timing reference. †
Most of the PTP implementation is done in the software above the UDP layer.
However, the hardware support is required to capture the exact time when specific
PTP packets enter or leave the Ethernet port at the PHY interface. This timing
information must be captured and returned to the software for the proper
implementation of PTP with high accuracy. †
The EMAC is intended to support IEEE 1588 operation in all modes with a resolution
of one μs. When the two EMACs are operating in an IEEE 1588 environment, the MPU
subsystem is responsible for maintaining synchronization between the time counters
internal to the two MACs. †
The IEEE 1588 interface to the FPGA allows the FPGA to provide an alternate source
for the emac_ptp_ref_clk input as well to allow it to monitor the pulse per second
output from each EMAC controller. †
The EMAC component provides a hardware assisted implementation of the IEEE 1588
protocol. Hardware support is for timestamp maintenance. Timestamps are updated
when receiving any frame on the PHY interface and the receive descriptor is updated
with this value. Timestamps are also updated when the SFD of a frame is transmitted
and updates the transmit descriptor accordingly. †
f For details about the IEEE 1588-2002 standard, refer to IEEE Standard 1588-2002 - IEEE
Standard for a Precision Clock Synchronization Protocol for Networked Measurement and
Control Systems, available on the IEEE Standards Association website
(standards.ieee.org). †
Reference Timing Source

To get a snapshot of the time, the EMAC takes the reference clock input and uses it to
generate the reference time (64-bit) internally and capture timestamps. †
System Time Register Module

The 64-bit time is maintained in this module and updated using the input reference
clock, osc1_clk. The osc1_clk clock comes from the clock manager and the
emac_ptp_ref_clk clock comes from the FPGA fabric. This time is the source for
taking snapshots (timestamps) of Ethernet frames being transmitted or received at the
PHY interface.
The system time counter can be initialized or corrected using the coarse correction
method. In this method, the initial value or the offset value is written to the
Timestamp Update register. For initialization, each EMAC’s system time counter is
written with the value in the Timestamp Update registers, while for system time
correction, the offset value is added to or subtracted from the system time.
In the fine correction method, a slave clock’s frequency drift with respect to the master
clock is corrected over a period of time instead of in one clock, as in coarse correction.
This helps maintain linear time and does not introduce drastic changes (or a large
jitter) in the reference time between PTP sync message intervals. †

In this method, an accumulator sums up the contents of the Timestamp_Addend

register, as shown in Figure 17–4. The arithmetic carry that the accumulator generates
is used as a pulse to increment the system time counter. The accumulator and the
addend are 32-bit registers. Here, the accumulator acts as a high-precision frequency
multiplier or divider.
1 You must connect a PTP clock with a frequency higher than the frequency required for
the specified accuracy. †
This algorithm is depicted in Figure 17–4: †
Figure 17–4. System Time Update Using Fine Method
addend_val[31:0] addend_updt
Addend Register
Accumulator Register
Constant Value
incr_sub_sec_reg
Sub-Second Register
incr_sec_reg
Second Register
The System Time Update logic requires a 50-MHz clock frequency to achieve 20-ns
accuracy. The frequency division ratio (FreqDivisionRatio) is the ratio of the reference
clock frequency to the required clock frequency. Hence, if the reference clock
(clk_ptp_ref_i) is, for example, 66 MHz, this ratio is calculated as 66 MHz / 50 MHz
= 1.32. Hence, the default addend value to be set in the register is 232 / 1.32,
0xC1F07C1F.
If the reference clock drifts lower, to 65 MHz for example, the ratio is 65 / 50, or 1.3
and the value to set in the addend register is 232 / 1.30, or 0xC4EC4EC4. If the clock
drifts higher, to 67 MHz for example, the addend register must be set to 0xBF0B7672.
When the clock drift is nil, the default addend value of 0xC1F07C1F (232 / 1.32) must
be programmed. †

In Figure 17–4, the constant value used to accumulate the sub-second register is
decimal 43, which achieves an accuracy of 20 ns in the system time (in other words, it
is incremented in 20-ns steps).
The software must calculate the drift in frequency based on the Sync messages and
update the Addend register accordingly. †
Initially, the slave clock is set with FreqCompensationValue0 in the Addend register.
This value is as follows: †
FreqCompensationValue0 = 232 / FreqDivisionRatio †
If MasterToSlaveDelay is initially assumed to be the same for consecutive Sync
messages, the algorithm described below must be applied. After a few Sync cycles,
frequency lock occurs. The slave clock can then determine a precise
MasterToSlaveDelay value and re-synchronize with the master using the new value. †
The algorithm is as follows: †
■ At time MasterSyncTimen the master sends the slave clock a Sync message. The
slave receives this message when its local clock is SlaveClockTimen and computes
MasterClockTimen as: †
MasterClockTimen = MasterSyncTimen + MasterToSlaveDelayn †
■ The master clock count for current Sync cycle, MasterClockCountn is given by: †
MasterClockCountn = MasterClockTimen – MasterClockTimen – 1
(assuming that MasterToSlaveDelay is the same for Sync cycles n and n – 1) †
■ The slave clock count for current Sync cycle, SlaveClockCountn is given by: †
SlaveClockCountn = SlaveClockTimen – SlaveClockTimen – 1 †
■ The difference between master and slave clock counts for current Sync cycle,
ClockDiffCountn is given by: †
ClockDiffCountn = MasterClockCountn – SlaveClockCountn †
■ The frequency-scaling factor for slave clock, FreqScaleFactorn is given by: †
FreqScaleFactorn = (MasterClockCountn + ClockDiffCountn) / SlaveClockCountn †
■ The frequency compensation value for Addend register, FreqCompensationValuen
is given by: †
FreqCompensationValuen = FreqScaleFactorn × FreqCompensationValuen – 1 †
In theory, this algorithm achieves lock in one Sync cycle; however, it may take several
cycles, because of changing network propagation delays and operating conditions. †
This algorithm is self-correcting: if for any reason the slave clock is initially set to a
value from the master that is incorrect, the algorithm corrects it at the cost of more
Sync cycles. †

Transmit Path Functions

The MAC captures a timestamp when the SFD of a frame is sent on the PHY interface.
The frames for which you want to capture timestamps are controllable on a per-frame
basis. In other words, each transmit frame can be marked to indicate whether a
timestamp should be captured for that frame. The MAC does not process the
transmitted frames to identify the PTP frames. You need to specify the frames for
which you want to capture timestamps. The MAC returns the timestamp, along with
the Transmit status of the frame, to hardware implemented in the FPGA. You can use
the control bits in the transmit descriptor. The MAC returns the timestamp to the
software inside the corresponding transmit descriptor, thus connecting the timestamp
automatically to the specific PTP frame. †
Receive Path Functions

The MAC captures the timestamp of all frames received on the PHY interface. The
DMA returns the timestamp to the software in the corresponding receive descriptor.
The timestamp is written only to the last receive descriptor. †
Timestamp Error Margin

According to the IEEE 1588 specifications, a timestamp must be captured at the SFD of
the transmitted and received frames at the PHY interface. Because the PHY interface
receive and transmit clocks are not synchronous to the reference timestamp clock
(osc1_clk) a small amount of drift is introduced when a timestamp is moved between
asynchronous clock domains. In the transmit path, the captured and reported
timestamp has a maximum error margin of two PTP clocks. It means that the captured
timestamp has the reference timing source value that is given within two clocks after
the SFD is transmitted on the PHY interface. †
Similarly, in the receive path, the error margin is three PHY interface clocks, plus up to
two PTP clocks. You can ignore the error margin because of the PHY interface clocks
by assuming that this constant delay is present in the system (or link) before the SFD
data reaches the PHY interface of the MAC. †
Frequency Range of Reference Timing Clock

The timestamp information is transferred across asynchronous clock domains, that is,
from MAC clock domain to the FPGA clock domain. Therefore, a minimum delay is
required between two consecutive timestamp captures. This delay is four clock cycles
of the PHY interface and three clock cycles of PTP clocks. If the delay between two
timestamp captures is less than this delay, the MAC does not take a timestamp
snapshot for the second frame. †
The maximum PTP clock frequency is limited by the maximum resolution of the
reference time (20 ns resulting in 50 MHz) and the timing constraints achievable for
logic operating on the PTP clock. In addition, the resolution, or granularity, of the
reference time source determines the accuracy of the synchronization. Therefore, a
higher PTP clock frequency gives better system performance. †

The minimum PTP clock frequency depends on the time required between two
consecutive SFD bytes. Because the PHY interface clock frequency is fixed by the IEEE
1588 specification, the minimum PTP clock frequency required for proper operation
depends upon the operating mode and operating speed of the MAC as shown in
Table 17–3. †
Table 17–3. Minimum PTP Clock Frequency Example

Mode Minimum Gap Between Two SFDs Minimum PTP Frequency
100-Mbps 168 MII clocks
(3 * PTP) + (4 * MII) <= 168 * MII, that is, ~0.5
full-duplex (128 clocks for a 64-byte frame + 24 clocks of MHz ((168 – 4) * 40 ns ÷ 3 = 2180 ns period)
operation min IFG + 16 clocks of preamble)
1000-Mbps 24 GMII clocks
3 * PTP + 4 * GMII <= 24 * GMII, that is, 18.75
half-duplex (4 for a jam pattern sent just after SFD because MHz
operation of collision + 12 IFG + 8 preamble) (1)
(1) For details about jam patterns, refer to IEEE Std 802.3-2008 Part 3: Carrier sense multiple access with Collision Detection (CSMA/CD) Access
Method and Physical Layer Specifications, available on the IEEE website (standards.ieee.org/findstds/).
IEEE 1588-2008 Advanced Timestamps

In addition to the basic timestamp features mentioned in IEEE 1588-2002 Timestamps,
the EMAC supports the following advanced timestamp features defined in the IEEE
1588-2008 standard. †
■ Supports the IEEE 1588-2008 (version 2) timestamp format. †
■ Provides an option to take a timestamp of all frames or only PTP type frames. †
■ Provides an option to take a timestamp of only event messages. †
■ Provides an option to take the timestamp based on the clock type: ordinary,
boundary, end-to-end, or peer-to-peer. †
■ Provides an option to configure the EMAC to be a master or slave for ordinary and
boundary clock. †
■ Identifies the PTP message type, version, and PTP payload in frames sent directly
over Ethernet and sends the status. †
■ Provides an option to measure sub-second time in digital or binary format. †
f For details about the IEEE 1588-2008 standard, refer to IEEE Standard 1588-2008 - IEEE
Standard for a Precision Clock Synchronization Protocol for Networked Measurement and
Control Systems, available on the IEEE Standards Association website
(standards.ieee.org). †
Peer-to-Peer PTP Transparent Clock (P2P TC) Message Support

The IEEE 1588-2008 version supports Peer-to-Peer PTP (Pdelay) messages in addition
to SYNC, Delay Request, Follow-up, and Delay Response messages. †

Clock Types
The EMAC supports the following clock types defined in the IEEE 1588-2008
standard: †
■ Ordinary clock †
■ Boundary clock †
■ End-to-End transparent clock †
■ Peer-to-Peer transparent clock †
Reference Timing Source

The EMAC supports the following reference timing source features defined in the
IEEE 1588-2008 standard: †
■ 48-Bit seconds Field †
■ Fixed pulse-per-second output †
■ Flexible pulse-per-second output †
■ Auxiliary snapshots (timestamps) with external events
Transmit Path Functions

The advanced timestamp feature is supported only through the alternate (enhanced)
descriptors format. †
Receive Path Functions

The MAC processes the received frames to identify valid PTP frames. You can control
the snapshot of the time, to be sent to the application, by using the following
options: †
■ Enable timestamp for all frames. †
■ Enable timestamp for IEEE 1588 version 2 or version 1 timestamp. †
■ Enable timestamp for PTP frames transmitted directly over Ethernet or UDP/IP
Ethernet. †
■ Enable timestamp snapshot for the received frame for IPv4 or IPv6. †
■ Enable timestamp snapshot for EVENT messages (SYNC, DELAY_REQ,
PDELAY_REQ, or PDELAY_RESP) only. †
■ Enable the node to be a master or slave and select the timestamp type. This
controls the type of messages for which timestamps are taken. †
The DMA returns the timestamp to the software inside the corresponding transmit or
receive descriptor. The advanced timestamp feature is supported only with the
32-byte alternate (enhanced) descriptor. †
Auxiliary Snapshot
The auxiliary snapshot feature allows you to store a snapshot (timestamp) of the
system time based on an external event. The event is considered to be the rising edge
of the sideband signal ptp_aux_ts_trig_i. One Auxiliary snapshot input is available.
The depth of the Auxiliary snapshot FIFO buffer is 16. †

The timestamps taken for any input are stored in a common FIFO buffer. The host can
read a register to know which input’s timestamp is available for reading at the top of
this FIFO buffer. The MAC stores these timestamps in a FIFO buffer. Only 64 bits of
the timestamp are stored in the FIFO buffer. When a timestamp is stored, the MAC
indicates this to the host with an interrupt. The value of the timestamp is read through
a FIFO buffer register access. †
IEEE 802.3az Energy Efficient Ethernet

Energy Efficient Ethernet (EEE) standardized by IEEE 802.3-az, version D2.0 is
supported by the EMAC. It is supported by the MAC operating in 10/100/1000 Mbps
rates. EEE is only supported when the EMAC is configured to operate with the RGMII
PHY interface operating in full-duplex mode. It does not support half-duplex mode. †
f For details about the IEEE 802.3az Energy Efficient Ethernet standard, refer to the
IEEE 802.3 Ethernet Working Group website (www.ieee802.org/3/). †
EEE enables the MAC to operate in Low-Power Idle (LPI) mode. Either end point of
an Ethernet link can disable functionality to save power during periods of low link
utilization. The MAC controls whether the system should enter or exit LPI mode and
communicates this to the PHY. †
LPI Timers
Two timers internal to the EMAC are associated with LPI mode: †
■ LPI Link Status (LS) Timer †
■ LPI TW Timer †
The LPI LS timer counts, in ms, the time expired since the link status is up. This timer
is cleared every time the link goes down and is incremented when the link is up again
and the terminal count as programmed by the software is reached. The PHY interface
does not assert the LPI pattern unless the terminal count is reached. This ensures a
minimum time for which no LPI pattern is asserted after a link is established with the
remote station. This period is defined as one second in the IEEE standard 802.3-az,
version D2.0. The LPI LS timer is 10 bits wide. Therefore, the software can program up
to 1023 ms. †
The LPI TW timer counts, in μs, the time expired since the deassertion of LPI. The
terminal count of the timer is the value of resolved transmit TW that is the
auto-negotiated time after which the MAC can resume the normal transmit operation.
The MAC supports the LPI TW timer in units of μs. The LPI TW timer is 16 bits wide.
Therefore, the software can program up to 65535 μs. †
The EMAC generates the LPI interrupt when the transmit or receive channel enters or
exits the LPI state. †

Checksum Offload
Communication protocols such as TCP and UDP implement checksum fields, which
help determine the integrity of data transmitted over a network. Because the most
widespread use of Ethernet is to encapsulate TCP and UDP over IP datagrams, the
EMAC has a Checksum Offload Engine (COE) to support checksum calculation and
insertion in the transmit path, and error detection in the receive path. Supported
offloading types are shown below: †
■ Transmit IP header checksum †
■ Transmit TCP/UDP/ICMP checksum †
■ Receive IP header checksum †
■ Receive full checksum †
Frame Filtering
The EMAC implements the following types of filtering for receive frames. †
Source Address or Destination Address Filtering

The Address Filtering Module checks the destination and source address field of each
incoming packet. †
■ “Unicast Destination Address Filter”
■ “Multicast Destination Address Filter”
■ “Hash or Perfect Address Filter”
■ “Broadcast Address Filter”
■ “Unicast Source Address Filter”
■ “Inverse Filtering Operation (Invert the filter-match result at final output)”
Unicast Destination Address Filter

Up to 128 MAC addresses for unicast perfect filtering are supported. The filter
compares all 48 bits of the received unicast address with the programmed MAC
address for any match. Default MacAddr0 is always enabled, other addresses
MacAddr1–MacAddr127 are selected with an individual enable bit. For
MacAddr1–MacAddr31 addresses, you can mask each byte during comparison with
the corresponding received DA byte. This enables group address filtering for the DA.
The MacAddr32-MacAddr127 addresses do not have mask control and all six bytes of
the MAC address are compared with the received six bytes of DA. †
In hash filtering mode, the filter performs imperfect filtering for unicast addresses
using a 64-bit hash table. It uses the upper six bits of the CRC of the received
destination address to index the content of the hash table. A value of 0 selects Bit 0 of
the selected register, and a value of 111111 binary selects Bit 63 of the Hash Table
register. If the corresponding bit is set to one, the unicast frame is said to have passed
the hash filter; otherwise, the frame has failed the hash filter. †

Multicast Destination Address Filter

The MAC can be programmed to pass all multicast frames. In Perfect Filtering mode,
the multicast address is compared with the programmed MAC Destination Address
registers (1–31). Group address filtering is also supported. In hash filtering mode, the
filter performs imperfect filtering using a 64-bit hash table. For hash filtering, it uses
the upper six bits of the CRC of the received multicast address to index the contents of
the hash table. A value of 0 selects Bit 0 of the selected register and a value of 111111
binary selects Bit 63 of the Hash Table register. If the corresponding bit is set to one,
then the multicast frame is said to have passed the hash filter; otherwise, the frame
has failed the hash filter. †
Hash or Perfect Address Filter

The filter can be configured to pass a frame when its DA matches either the hash filter
or the Perfect filter. This configuration applies to both unicast and multicast frames. †
Broadcast Address Filter

The filter does not filter any broadcast frames in the default mode. However, if the
MAC is programmed to reject all broadcast frames, the filter drops any broadcast
frame. †
Unicast Source Address Filter

The MAC can also perform a perfect filtering based on the source address field of the
received frames. Group filtering with SA is also supported. You can filter a group of
addresses by masking one or more bytes of the address. †
Inverse Filtering Operation (Invert the filter-match result at final output)

For both Destination and Source address filtering, there is an option to invert the
filter-match result at the final output. The result of the unicast or multicast destination
address filter is inverted in this mode. †
VLAN Filtering
The EMAC supports the two kinds of VLAN filtering: †
■ VLAN tag-based filtering †
■ VLAN hash filtering †
VLAN tag-based filtering

In the VLAN tag-based frame filtering, the MAC compares the VLAN tag of the
received frame and provides the VLAN frame status to the application. Based on the
programmed mode, the MAC compares the lower 12 bits or all 16 bits of the received
VLAN tag to determine the perfect match. If VLAN tag filtering is enabled, the MAC
forwards the VLAN-tagged frames along with VLAN tag match status and drops the
VLAN frames that do not match. You can also enable the inverse matching for VLAN
frames. In addition, you can enable matching of SVLAN tagged frames along with the
default Customer Virtual Local Area Network (C-VLAN) tagged frames. †

VLAN hash filtering with a 16-bit hash table

The MAC provides VLAN hash filtering with a 16-bit hash table. The MAC also
supports the inverse matching of the VLAN frames. In inverse matching mode, when
the VLAN tag of a frame matches the perfect or hash filter, the packet should be
dropped. If the VLAN perfect and VLAN hash match are enabled, a frame is
considered as matched if either the VLAN hash or the VLAN perfect filter matches.
When inverse match is set, a packet is forwarded only when both perfect and hash
filters indicate mismatch. †
Layer 3 and Layer 4 Filters

Layer 3 filtering refers to source address and destination address filtering. Layer 4
filtering refers to source port and destination port filtering. The frames are filtered in
the following ways: †
■ Matched frames †
■ Unmatched frames †
■ Non-TCP or UDP IP frames †
Matched Frames
The MAC forwards the frames, which match all enabled fields, to the application
along with the status. The MAC gives the matched field status only if one of the
following conditions is true: †
■ All enabled Layer 3 and Layer 4 fields match. †
■ At least one of the enabled field matches and other fields are bypassed or
disabled. †
Using the CSR set, you can define up to four filters, identified as filter 0 through filter
3. When multiple Layer 3 and Layer 4 filters are enabled, any filter match is
considered as a match. If more than one filter matches, the MAC provides status of the
lowest filter with filter 0 being the lowest and filter 3 being the highest. For example, if
filter 0 and filter 1 match, the MAC gives the status corresponding to filter 0. †
Unmatched Frames
The MAC drops the frames that do not match any of the enabled fields. You can use
the inverse match feature to block or drop a frame with specific TCP or UDP over IP
fields and forward all other frames. You can configure the EMAC so that when a
frame is dropped, it receives a partial frame with appropriate abort status or drops it
completely. †
Non-TCP or UDP IP Frames

By default, all non-TCP or UDP IP frames are bypassed from the Layer 3 and Layer 4
filters. You can optionally program the MAC to drop all non-TCP or UDP over IP
frames. †

Clocks and Resets

The Ethernet MAC controller uses the clocks shown in Table 17–4.
Table 17–4. Clocks

Name Nominal Frequency Functional Usage Notes
If supplied from clock interface, clock
clk_ref_i 250 Mhz Reference Clock to the EMAC is emac0_clk or emac1_clk
Autonegotiates speed down to

clk_tx_i 125/25/2.5 Mhz
10/100Mbps
All PHY signals received by the MAC
clk_rx_i PHY provides this reference to MAC
are synchronous to this clock
Clock Gating for EEE

For the RGMII PHY interface, you can gate the transmit clock for Energy Efficient
Ethernet (EEE) applications. For more information, refer to “Programming Guidelines
for Energy Efficient Ethernet” on page 17–63.
Resets
The Ethernet MAC controller uses the reset signals shown in Table 17–5.
Table 17–5. Resets

Name Nominal Frequency Functional Usage Notes
Used to reset external PHY transmit
rst_clk_tx_n_o Transmit clock reset output
clock domain logic
Used to reset external PHY receive
rst_clk_rx_n_o Receive clock reset output
clock domain logic
Interrupts
Interrupts are generated as a result of specific events in the EMAC and external PHY
device. The interrupt status register indicates all conditions which may trigger an
interrupt and the interrupt enable register determines which interrupts can
propagate.

Ethernet MAC Programming Model
DMA Controller
The DMA has independent transmit and receive engines, and a CSR space. The
transmit engine transfers data from system memory to the device port or MAC
transaction layer (MTL), while the receive engine transfers data from the device port
to the system memory. The controller use descriptors to efficiently move data from
source to destination with minimal Host CPU intervention. The DMA is designed for
packet-oriented data transfers such as frames in Ethernet. The controller can be
programmed to interrupt the Host CPU for situations such as frame transmit and
receive transfer completion, and other normal/error conditions.
The DMA and the Host driver communicate through two data structures: †
■ Control and Status registers (CSR) †
■ Descriptor lists and data buffers †
For information about Control and Status registers, refer to “Ethernet MAC Address
Map and Register Definitions” on page 17–67. Descriptors are described in “Normal
Descriptor” on page 17–36 and “Alternate or Enhanced Descriptors” on page 17–47.
1 You can select an alternative descriptor structure during RTL configuration. The
control bits in this descriptor structure are reassigned so that the application can use a
larger buffer size (8 KB). For a detailed bit map of this alternative descriptor structure,
refer to “Alternate or Enhanced Descriptors” on page 17–47. All descriptions in
“DMA Controller” on page 17–21 refer to the default descriptor structure, not this
new alternative. If you are using the alternate descriptor structure, ignore the
descriptor-specific mapping in “DMA Controller” on page 17–21 and refer to the
alternate descriptor-specific bit maps.
The DMA transfers data frames received by the MAC to the receive Buffer in the Host
memory, and transmit data frames from the transmit Buffer in the Host memory.
Descriptors that reside in the Host memory act as pointers to these buffers. †
There are two descriptor lists; one for reception, and one for transmission. The base
address of each list is written into Register 3 (Receive Descriptor List Address
Register) and Register 4 (Transmit Descriptor List Address Register), respectively. A
descriptor list is forward linked (either implicitly or explicitly). The last descriptor
may point back to the first entry to create a ring structure. Explicit chaining of
descriptors is accomplished by setting the second address chained in both receive and
transmit descriptors (RDES1[24] and TDES1[24]). The descriptor lists resides in the
Host physical memory address space. Each descriptor can point to a maximum of two
buffers. This enables two buffers to be used, physically addressed, rather than
contiguous buffers in memory. †
A data buffer resides in the Host physical memory space, and consists of an entire
frame or part of a frame, but cannot exceed a single frame. Buffers contain only data,
buffer status is maintained in the descriptor. Data chaining refers to frames that span
multiple data buffers. However, a single descriptor cannot span multiple frames. The
DMA skips to the next frame buffer when end-of-frame is detected. Data chaining can
be enabled or disabled. †

The descriptor ring and chain structures are shown in Figure 17–5 and Figure 17–6. †
Figure 17–5. Descriptor Ring Structure
Buffer 1
Descriptor 0
Buffer 21
Buffer 1
Descriptor 1
Buffer 2
Buffer 1
Descriptor 2
Buffer 2
Buffer 1
Descriptor n
Buffer 2
Figure 17–6. Descriptor Chain Structure
Buffer 1
Descriptor 0
Buffer 1
Descriptor 1
Buffer 1
Descriptor 2
Next Descriptor
Initialization
Initialization for the EMAC is as follows.
1. Write to Register 0 (Bus Mode Register) to set Host bus access parameters. †
2. Write to Register 7 (Interrupt Enable Register) to mask unnecessary interrupt
causes. †
3. Create the transmit and receive descriptor lists, and then write to DMA Register 3
(Receive Descriptor List Address Register) and Register 4 (Transmit Descriptor
List Address Register), providing the DMA with the starting address of each list. †

4. Write to Register 1 (MAC Frame Filter), Register 2 (Hash Table High Register), and
Register 3 (Hash Table Low Register) for desired filtering options. †
5. Write to Register 1 (MAC Frame Filter) to configure the operating mode and
enable the transmit operation (Bit 3: Transmitter Enable). The PS and DM bits are
set based on the auto-negotiation result (read from the PHY). †
6. Write to Register 6 (Operation Mode Register) to set Bits 13 and 1 to start
transmission and reception. †
7. Write to Register 0 (MAC Configuration Register) to enable the receive operation
(Bit 2: Receiver Enable). †
The transmit and receive engines enter the Running state and attempt to acquire
descriptors from the respective descriptor lists. The receive and transmit engines
then begin processing receive and transmit operations. The transmit and receive
processes are independent of each other and can be started or stopped
separately. †
Host Bus Burst Access

The DMA attempts to execute fixed-length Burst transfers on the master interface if
configured to do so through FB bit of Register 0 (Bus Mode Register). The maximum
Burst length is indicated and limited by the PBL field (Bits [13:8]) Register 0 (Bus
Mode Register). The receive and transmit descriptors are always accessed in the
maximum possible (limited by PBL or 16 * 8/bus width) burst-size for the 16- bytes to
be read.
The transmit DMA initiates a data transfer only when sufficient space to
accommodate the configured burst is available in MTL transmit FIFO buffer or the
number of bytes till the end of frame (when it is less than the configured burst-length).
The DMA indicates the start address and the number of transfers required to the
master interface. When the interface is configured for fixed-length burst, then it
transfers data using the best combination of INCR4, 8, or 16 and SINGLE transactions.
Otherwise (no fixed-length burst), it transfers data using INCR (undefined length)
and SINGLE transactions.
The receive DMA initiates a data transfer only when sufficient data to accommodate
the configured burst is available in MTL receive FIFO buffer or when the end of frame
(when it is less than the configured burst-length) is detected in the receive FIFO
buffer. The DMA indicates the start address and the number of transfers required to
the master interface. When the interface is configured for fixed-length burst, then it
transfers data using the best combination of INCR4, 8, or 16 and SINGLE transactions.
If the end-of frame is reached before the fixed-burst ends on the interface, then
dummy transfers are performed in order to complete the fixed-burst. Otherwise (FB
bit of Register 0 (Bus Mode Register) is reset), it transfers data using INCR (undefined
length) and SINGLE transactions.
When the interface is configured for address-aligned beats, both DMA engines ensure
that the first burst transfer initiated is less than or equal to the size of the configured
PBL. Thus, all subsequent beats start at an address that is aligned to the configured
PBL. The DMA can only align the address for beats up to size 16 (for PBL > 16),
because the interface does not support more than INCR16.

Host Data Buffer Alignment

The transmit and receive data buffers do not have any restrictions on start address
alignment. For example, in systems with 32-bit memory, the start address for the
buffers can be aligned to any of the four bytes. However, the DMA always initiates
transfers with address aligned to the bus width with dummy data for the byte lanes
not required. This typically happens during the transfer of the beginning or end of an
Ethernet frame. The software driver should discard the dummy bytes based on the
start address of the buffer and size of the frame. †
Example: Buffer Read

If the transmit buffer address is 0x00000FF2 (for 32-bit data bus), and 15 bytes
need to be transferred, then the DMA reads five full words from address
0x00000FF0, but when transferring data to the MTL transmit FIFO buffer, the
extra bytes (the first two bytes) are dropped or ignored. Similarly, the last 3 bytes
of the last transfer are also ignored. The DMA always ensures it transfers a full
32-bit data to the MTL transmit FIFO buffer, unless it is the end-of-frame.
Example: Buffer Write

If the receive buffer address is 0x0000FF2 (for 64-bit data bus) and 16 bytes of a
received frame need to be transferred, then the DMA writes 3 full words from
address 0x00000FF0. But the first 2 bytes of first transfer and the last 6 bytes of the
third transfer have dummy data.
Buffer Size Calculations

The DMA does not update the size fields in the transmit and receive descriptors. The
DMA updates only the status fields (RDES and TDES) of the descriptors. The driver
has to perform the size calculations. †
The transmit DMA transfers the exact number of bytes (indicated by buffer size field
of TDES1) towards the MAC. If a descriptor is marked as first (FS bit of TDES1 is set),
then the DMA marks the first transfer from the buffer as the start of frame. If a
descriptor is marked as last (LS bit of TDES1), then the DMA marks the last transfer
from that data buffer as the end-of frame to the MTL. †
The receive DMA transfers data to a buffer until the buffer is full or the end-of frame
is received from the MTL. If a descriptor is not marked as last (LS bit of RDES0), then
the descriptor’s corresponding buffer(s) are full and the amount of valid data in a
buffer is accurately indicated by its buffer size field minus the data buffer pointer
offset when the FS bit of that descriptor is set. The offset is zero when the data buffer
pointer is aligned to the data bus width. If a descriptor is marked as last, then the
buffer may not be full (as indicated by the buffer size in RDES1). To compute the
amount of valid data in this final buffer, the driver must read the frame length (FL bits
of RDES0[29:16]) and subtract the sum of the buffer sizes of the preceding buffers in
this frame. The receive DMA always transfers the start of next frame with a new
descriptor. †

1 Even when the start address of a receive buffer is not aligned to the data width of
system bus, the system should allocate a receive buffer of a size aligned to the system
bus width. For example, if the system allocates a 1,024-byte (1 KB) receive buffer
starting from address 0x1000, the software can program the buffer start address in the
receive descriptor to have a 0x1002 offset. The receive DMA writes the frame to this
buffer with dummy data in the first two locations (0x1000 and 0x1001). The actual
frame is written from location 0x1002. Thus, the actual useful space in this buffer is
1,022 bytes, even though the buffer size is programmed as 1,024 bytes, because of the
start address offset. †
Transmission
Transmission functions use transmit descriptors, described in detail in “Transmit
Descriptor” on page 17–36.
TX DMA Operation: Default (Non-OSF) Mode

The transmit DMA engine in default mode proceeds as follows: †
1. The Host sets up the transmit descriptor (TDES0-TDES3) and sets the Own bit
(TDES0[31]) after setting up the corresponding data buffer(s) with Ethernet frame
data. †
2. When Bit 13 (ST) of Register 6 (Operation Mode Register) is set, the DMA enters
the Run state. †
3. While in the Run state, the DMA polls the transmit descriptor list for frames
requiring transmission. After polling starts, it continues in either sequential
descriptor ring order or chained order. If the DMA detects a descriptor flagged as
owned by the Host (TDES0[31] = 0), or if an error condition occurs, transmission is
suspended and both the Bit 2 (Transmit Buffer Unavailable) and Bit 16 (Normal
Interrupt Summary) of the Register 5 (Status Register) are set. The transmit Engine
proceeds to Step 9.
4. If the acquired descriptor is flagged as owned by DMA (TDES0[31] = 1), the DMA
decodes the transmit Data Buffer address from the acquired descriptor.
5. The DMA fetches the transmit data from the Host memory and transfers the data
to the MTL for transmission. †
6. If an Ethernet frame is stored over data buffers in multiple descriptors, the DMA
closes the intermediate descriptor and fetches the next descriptor. Steps 3, 4, and 5
are repeated until the end-of-Ethernet-frame data is transferred to the MTL. †
7. When frame transmission is complete, if IEEE 1588 timestamping was enabled for
the frame (as indicated in the transmit status) the timestamp value obtained from
MTL is written to the transmit descriptor (TDES2 and TDES3) that contains the
end-of-frame buffer. The status information is then written to this transmit
descriptor (TDES0). Because the Own bit is cleared during this step, the Host now
owns this descriptor. If timestamping was not enabled for this frame, the DMA
does not alter the contents of TDES2 and TDES3. †

8. Bit 0 (Transmit Interrupt) of Register 5 (Status Register) is set after completing

transmission of a frame that has Interrupt on Completion (TDES1[31]) set in its
Last descriptor. The DMA engine then returns to Step 3. †
9. In the Suspend state, the DMA tries to re-acquire the descriptor (and thereby
return to Step 3) when it receives a Transmit Poll demand and the Underflow
Interrupt Status bit is cleared. †
The TX DMA transmission flow in default mode is shown in Figure 17–7. †

Figure 17–7. TX DMA Operation in Default Mode
Start Start
Tx DMA Stop
Tx DMA
(Re-)Fetch Next
Descriptor
yes
Error?
no
Poll Demand
Tx DMA no
Own Bit Set?
Suspended
yes
Transfer Data
from Buffer(s)
yes
Error?
no
Frame
no Transfer
Complete?
yes
Wait for
Tx Status
Close Intermediate
Descriptor Timestamp yes Write Timestamp
Present? to RDES2 & RDES3
no
Write Status no yes

Error?
Word to TDES0
no yes
Error?

TX DMA Operation: OSF Mode

While in the Run state, the transmit process can simultaneously acquire two frames
without closing the Status descriptor of the first [if Bit 2 (OSF) in Register 6 (Operation
Mode Register) is set]. As the transmit process finishes transferring the first frame, it
immediately polls the transmit descriptor list for the second frame. If the second
frame is valid, the transmit process transfers this frame before writing the first frame’s
status information. †
In OSF mode, the Run state transmit DMA operates in the following sequence: †
1. The DMA operates as described in steps 1–6 of “TX DMA Operation: Default
(Non-OSF) Mode” on page 17–25. †
2. Without closing the previous frame’s last descriptor, the DMA fetches the next
descriptor. †
3. If the DMA owns the acquired descriptor, the DMA decodes the transmit buffer
address in this descriptor. If the DMA does not own the descriptor, the DMA goes
into Suspend mode and skips to Step 7. †
4. The DMA fetches the transmit frame from the Host memory and transfers the
frame to the MTL until the End-of-frame data is transferred, closing the
intermediate descriptors if this frame is split across multiple descriptors. †
5. The DMA waits for the previous frame’s frame transmission status and
timestamp. Once the status is available, the DMA writes the timestamp to TDES2
and TDES3, if such timestamp was captured (as indicated by a status bit). The
DMA then writes the status, with a cleared Own bit, to the corresponding TDES0,
thus closing the descriptor. If timestamping was not enabled for the previous
frame, the DMA does not alter the contents of TDES2 and TDES3. †
6. If enabled, the transmit interrupt is set, the DMA fetches the next descriptor, then
proceeds to Step 3 (when Status is normal). If the previous transmission status
shows an underflow error, the DMA goes into Suspend mode (Step 7). †
7. In Suspend mode, if a pending status and timestamp are received from the MTL,
the DMA writes the timestamp (if enabled for the current frame) to TDES2 and
TDES3, then writes the status to the corresponding TDES0. It then sets relevant
interrupts and returns to Suspend mode. †
8. The DMA can exit Suspend mode and enter the Run state (go to Step 1 or Step 2
depending on pending status) only after receiving a Transmit Poll demand
(Register 1 (Transmit Poll Demand Register)). †
1 As the DMA fetches the next descriptor in advance before closing the current
descriptor, the descriptor chain should have more than two different descriptors for
correct and proper operation. †
The basic flow is described in Figure 17–8. †

Figure 17–8. TX DMA Operation in OSF Mode
Start Start
Tx DMA
Stop
Rx DMA
(Re-)Fetch Next
Descriptor
yes
Error?
Poll Demand
no
Tx DMA no
Own Bit Set?
Suspended
Previous Frame
Status Available yes
Transfer Data
no from Buffer(s)
Timestamp
Present?
yes
yes
Error?
Write Timestamp to RDES2 &
TDES3 for Previous Frame
no no
Frame
yes no Transfer yes Second
Error?
Complete? Frame?
no yes
Write Status Word to Close Intermediate Wait for Previous

Previous Frame’s TDES0 Descriptor Frame’s TX Status
yes Write Timestamp

Timestamp
to TDES2 & TDES3
no Present?
Error? for Previous Frame
no
yes
Write Status Word to no yes
Error?
Previous Frame’s TDES0
no
Error?
yes

Transmit Frame Processing

The transmit DMA expects that the data buffers contain complete Ethernet frames,
excluding preamble, pad bytes, and FCS fields. The DA, SA, and Type/Len fields
contain valid data. If the transmit descriptor indicates that the MAC must disable
CRC or PAD insertion, the buffer must have complete Ethernet frames (excluding
preamble), including the CRC bytes. †
Frames can be data-chained and can span several buffers. Frames must be delimited
by the First Descriptor (TDES1[29]) and the Last Descriptor (TDES1[30]),
respectively. †
As transmission starts, the First Descriptor must have (TDES1[29]) set. When this
occurs, frame data transfers from the Host buffer to the MTL transmit FIFO buffer.
Concurrently, if the current frame has the Last Descriptor (TDES1[30]) clear, the
transmit Process attempts to acquire the Next descriptor. The transmit Process expects
this descriptor to have TDES1[29] clear. If TDES1[30] is clear, it indicates an
intermediary buffer. If TDES1[30] is set, it indicates the last buffer of the frame. †
After the last buffer of the frame has been transmitted, the DMA writes back the final
status information to the Transmit Descriptor 0 (TDES0) word of the descriptor that
has the last segment set in Transmit Descriptor 1 (TDES1[30]). At this time, if Interrupt
on Completion (TDES1[31]) is set, the Bit 0 (Transmit Interrupt) of Register 5 (Status
Register) is set, the Next descriptor is fetched, and the process repeats. †
The actual frame transmission begins after the MTL transmit FIFO buffer has reached
either a programmable transmit threshold (Bits [16:14] of Register 6 (Operation Mode
Register)), or a full frame is contained in the FIFO buffer. There is also an option for
Store and Forward Mode (Bit 21 of Register 6 (Operation Mode Register)). Descriptors
are released (Own bit TDES0[31] clears) when the DMA finishes transferring the
frame. †
1 To ensure proper transmission of a frame and the next frame, you must specify a
non-zero buffer size for the transmit descriptor that has the Last Descriptor
(TDES1[30]) set. †
Transmit Polling Suspended

Transmit polling can be suspended by either of the following conditions: †
■ The DMA detects a descriptor owned by the Host (TDES0[31]=0). To resume, the
driver must give descriptor ownership to the DMA and then issue a Poll Demand
command. †
■ A frame transmission is aborted when a transmit error because of underflow is
detected. The appropriate Transmit Descriptor 0 (TDES0) bit is set. †
If the DMA goes into SUSPEND state because of the first condition, then both Bit 16
(Normal Interrupt Summary) and Bit 2 (Transmit Buffer Unavailable) of Register 5
(Status Register) are set. If the second condition occur, both Bit 15 (Abnormal
Interrupt Summary) and Bit 5 (Transmit Underflow) of Register 5 (Status Register) are
set, and the information is written to Transmit Descriptor 0, causing the suspension. †
In both cases, the position in the transmit List is retained. The retained position is that
of the descriptor following the Last descriptor closed by the DMA. †
The driver must explicitly issue a Transmit Poll Demand command after rectifying the
suspension cause. †

Reception
Receive functions use receive descriptors, described in detail in “Receive Descriptor”
on page 17–40.
The receive DMA engine’s reception sequence is depicted in Figure 17–9 on
page 17–32 and proceeds as follows: †
1. The host sets up receive descriptors (RDES0-RDES3) and sets the Own bit
(RDES0[31]). †
2. When Bit 1 (SR) of Register 6 (Operation Mode Register) is set, the DMA enters the
Run state. While in the Run state, the DMA polls the receive descriptor list,
attempting to acquire free descriptors. If the fetched descriptor is not free (is
owned by the host), the DMA enters the Suspend state and jumps to Step 9. †
3. The DMA decodes the receive data buffer address from the acquired descriptors. †
4. Incoming frames are processed and placed in the acquired descriptor’s data
buffers. †
5. When the buffer is full or the frame transfer is complete, the receive engine fetches
the next descriptor. †
6. If the current frame transfer is complete, the DMA proceeds to Step 7. If the DMA
does not own the next fetched descriptor and the frame transfer is not complete
(EOF is not yet transferred), the DMA sets the Descriptor Error bit in the RDES0
(unless flushing is disabled in Bit 24 of Register 6 (Operation Mode Register)). The
DMA closes the current descriptor (clears the Own bit) and marks it as
intermediate by clearing the Last Segment (LS) bit in the RDES0 value (marks it as
Last Descriptor if flushing is not disabled), then proceeds to Step 8. If the DMA
does own the next descriptor but the current frame transfer is not complete, the
DMA closes the current descriptor as intermediate and reverts to Step 4. †
7. If IEEE 1588 timestamping is enabled, the DMA writes the timestamp (if available)
to the current descriptor’s RDES2 and RDES3. It then takes the receive frame’s
status from the MTL and writes the status word to the current descriptor’s RDES0,
with the Own bit cleared and the Last Segment bit set. †
8. The receive engine checks the latest descriptor’s Own bit. If the host owns the
descriptor (Own bit is 0), the Bit 7 (Receive Buffer Unavailable) of Register 5
(Status Register) is set and the DMA receive engine enters the Suspended state
(Step 9). If the DMA owns the descriptor, the engine returns to Step 4 and awaits
the next frame.
9. Before the receive engine enters the Suspend state, partial frames are flushed from
the receive FIFO buffer. You can control flushing using Bit 24 of Register 6
(Operation Mode Register). †

10. The receive DMA exits the Suspend state when a Receive Poll demand is given or
the start of next frame is available from the MTL’s receive FIFO buffer. The engine
proceeds to Step 2 and refetches the next descriptor. †
Figure 17–9. Receive DMA Operation
Start Start
Rx DMA Stop
Rx DMA
Poll Demand/
New Frame Available (Re-)Fetch Next
Descriptor
Rx DMA yes
Error?
Suspended
yes no
Frame
Transfer no
Own Bit Set?
Complete?
no yes
yes Flush Frame Data no

Disabled? Available?
no yes
Flush the Write Data Wait for Frame

Remaining Frame to Buffer(s) Data
Fetch Next no yes

Error?
Descriptor
Error?
no
Own Bit Frame

no Flush no no yes Timestamp yes
Set For Next Transfer Write Timestamp
Disabled? Descriptor? Complete? Present? to RDES2 & RDES3
yes yes no
Set Descriptor Close RDES0 As Close RDES0 As no yes

Error?
Error Intermediate Descriptor Last Descriptor
no
Error?
yes

If software has enabled timestamping through CSR, when a valid timestamp value is
not available for the frame (for example, because the receive FIFO buffer was full
before the timestamp could be written to it), the DMA writes all-ones to RDES2 and
RDES3. Otherwise (that is, if timestamping is not enabled), the RDES2 and RDES3
remain unchanged. †
Receive Descriptor Acquisition

The receive Engine always attempts to acquire an extra descriptor in anticipation of
an incoming frame. Descriptor acquisition is attempted if any of the following
conditions is satisfied: †
■ Bit 1 (Start or Stop Receive) of Register 6 (Operation Mode Register) has been set
immediately after being placed in the Run state. †
■ The data buffer of current descriptor is full before the frame ends for the current
transfer. †
■ The controller has completed frame reception, but the current receive descriptor is
not yet closed. †
■ The receive process has been suspended because of a host-owned buffer
(RDES0[31] = 0) and a new frame is received. †
■ A Receive poll demand has been issued. †
Receive Frame Processing

The MAC transfers the received frames to the Host memory only when the frame
passes the address filter and frame size is greater than or equal to configurable
threshold bytes set for the receive FIFO buffer of MTL, or when the complete frame is
written to the FIFO buffer in Store-and-Forward mode. †
If the frame fails the address filtering, it is dropped in the MAC block itself (unless Bit
31 (Receive All) of Register 1 (MAC Frame Filter) is set). Frames that are shorter than
64 bytes, because of collision or premature termination, can be removed from the MTL
receive FIFO buffer. †
After 64 (configurable threshold) bytes have been received, the MTL block requests
the DMA block to begin transferring the frame data to the receive Buffer pointed by
the current descriptor. The DMA sets First Descriptor (RDES0[9]) after the DMA Host
interface becomes ready to receive a data transfer (if DMA is not fetching transmit
data from the host), to delimit the frame. The descriptors are released when the Own
(RDES[31]) bit is reset to 0, either as the Data buffer fills up or as the last segment of
the frame is transferred to the receive buffer. If the frame is contained in a single
descriptor, both Last Descriptor (RDES[8]) and First Descriptor (RDES[9]) are set.
The DMA fetches the next descriptor, sets the Last Descriptor (RDES[8]) bit, and
releases the RDES0 status bits in the previous frame descriptor. Then the DMA sets
the Bit 6 (Receive Interrupt) of Register 5 (Status Register). The same process repeats
unless the DMA encounters a descriptor flagged as being owned by the host. If this
occurs, the receive Process sets the Bit 7 (Receive Buffer Unavailable) of Register 5
(Status Register) and then enters the Suspend state. The position in the receive list is
retained. †

Receive Process Suspended

If a new receive frame arrives while the receive Process is in Suspend state, the DMA
refetches the current descriptor in the Host memory. If the descriptor is now owned
by the DMA, the receive Process re-enters the Run state and starts frame reception. If
the descriptor is still owned by the host, by default, the DMA discards the current
frame at the top of the MTL RX FIFO buffer and increments the missed frame counter.
If more than one frame is stored in the MTL EX FIFO buffer, the process repeats. †
The discarding or flushing of the frame at the top of the MTL EX FIFO buffer can be
avoided by disabling Flushing (Bit 24 of Register 6 (Operation Mode Register)). In
such conditions, the receive process sets the Receive Buffer Unavailable status and
returns to the Suspend state. †
Interrupts
Interrupts can be generated as a result of various events. The DMA Register 5 (Status
Register) contains all the bits that might cause an interrupt. Register 7 (Interrupt
Enable Register) contains an enable bit for each of the events that can cause an
interrupt. †
There are two groups of interrupts, Normal and Abnormal, as described in Register 5
(Status Register). Interrupts are cleared by writing a 1 to the corresponding bit
position. When all the enabled interrupts within a group are cleared, the
corresponding summary bit is cleared. When both the summary bits are cleared, the
sbd_intr_o interrupt signal is de-asserted. If the MAC is the cause for assertion of the
interrupt, then any of the GLI, GMI, GPI, TTI, or GLPII bits of Register 5 (Status
Register) are set high, as shown in Figure 17–10.
Figure 17–10. sbd_intr_o Generation (1)
TTI
GPI
GMI
GLI
TI
NIS GLPII/GTMSI
TIE
NIE sbd_intr_o
ERI
ERE
TPS
AIS
TSE
AIE
FBI
FBE

(1) Signals NIS and AIS are registered.
1 The Register 5 (Status Register) is the interrupt status register. The interrupt pin
(sbd_intr_o) is asserted because of any event in this status register only if the
corresponding interrupt enable bit is set in Register 7 (Interrupt Enable Register). †

Interrupts are not queued and if the interrupt event occurs before the driver has
responded to it, no additional interrupts are generated. For example, Bit 6 (Receive
Interrupt) of Register 5 (Status Register) indicates that one or more frames were
transferred to the Host buffer. The driver must scan all descriptors, from the last
recorded position to the first one owned by the DMA. †
An interrupt is generated only once for simultaneous, multiple events. The driver
must scan the Register 5 (Status Register) for the cause of the interrupt. The interrupt
is not generated again unless a new interrupting event occurs, after the driver has
cleared the appropriate bit in Register 5 (Status Register). For example, the controller
generates a Bit 6 (Receive Interrupt) of Register 5 (Status Register) and the driver
begins reading Register 5 (Status Register). Next, Bit 7 (Receive Buffer Unavailable) of
Register 5 (Status Register) occurs. The driver clears the receive interrupt. Even then,
the sbd_intr_o signal is not de-asserted, because of the active or pending Receive
Buffer Unavailable interrupt. †
Bits 7:0 (Interrupt Timer) of Register 9 (Receive Interrupt Watchdog Timer Register) is
given for flexible control of receive Interrupt. When this Interrupt timer is
programmed with a non-zero value, it gets activated as soon as the RX DMA
completes a transfer of a received frame to system memory without asserting the
receive Interrupt because it is not enabled in the corresponding Receive Descriptor
(RDES1[31] in Table 17–12 on page 17–43). When this timer runs out as per the
programmed value, RI bit is set and the interrupt is asserted if the corresponding RI is
enabled in Register 7 (Interrupt Enable Register). This timer gets disabled before it
runs out, when a frame is transferred to memory and the RI is set because it is enabled
for that descriptor. †
Error Response to DMA

For any data transfer initiated by a DMA channel, if the slave replies with an error
response, that DMA stops all operations and updates the error bits and the Fatal Bus
Error bit in the Register 5 (Status Register). The DMA controller can resume operation
only after soft resetting or hard resetting the EMAC and reinitializing the DMA.
Descriptor Overview
This section describes the HPS EMAC DMA descriptors.
The DMA in the Ethernet subsystem transfers data based on a linked list of
descriptors, as explained in “DMA Controller” on page 17–21. The descriptors are
created in the system memory. The EMAC DMA supports the following two types of
descriptors:
■ Normal descriptor—This is the default descriptor format which can have 4
DWORDS. For information about this descriptor type, refer to “Normal
Descriptor” on page 17–36. †
■ Enhanced descriptor—This is the alternate descriptor format which can have 8
DWORDS (32 bytes). For information about this descriptor type, refer to
“Alternate or Enhanced Descriptors” on page 17–47.
Each descriptor contains two buffers, two byte-count buffers, and two address
pointers, which enable the adapter port to be compatible with various types of
memory management schemes. Once configured, you cannot change the descriptor
structure. †

Descriptor Endianness
The descriptor addresses must be aligned to the used bus width (Word, DWord, or
LWord for 32-bit, 64-bit, or 128-bit bus respectively). The data bus is configured as
little-endian.
1 The figures in this section show the normal transmit and receive descriptors. If you
are using the enhanced descriptors, you must ensure that the descriptors are created
accordingly in system memory. For more information about enhanced descriptors,
refer to “Alternate or Enhanced Descriptors” on page 17–47. †
Normal Descriptor
This is the default descriptor format which can have 4 DWORDS. You can use this
descriptor format when advanced features such as IEEE 1588 advanced timestamping
are not enabled.
Transmit Descriptor
The DMA in the EMAC requires at least one descriptor for a transmit frame. In
addition to two buffers, two byte-count buffers, and two address pointers, the
transmit descriptor has control fields which can be used to control the MAC operation
on per-transmit frame basis.
Transmit Descriptor 0 (TDES0)

TDES0 contains the transmitted frame status and the descriptor ownership
information.
Table 17–6. Transmit Descriptor 0 (Part 1 of 3)

Bit Description
OWN: Own Bit
When set, this bit indicates that the descriptor is owned by the DMA. When this bit is reset, it indicates that the
31 descriptor is owned by the Host. The DMA clears this bit either when it completes the frame transmission or when
the buffers allocated in the descriptor are empty. The ownership bit of the First descriptor of the frame should be set
after all subsequent descriptors belonging to the same frame have been set. This avoids a possible race condition
between fetching a descriptor and the driver setting an ownership bit. †
30:18 Reserved †
TTSS: TX Timestamp Status
17 This status bit indicates that a timestamp has been captured for the corresponding transmit frame. When this bit is
set, TDES2 and TDES3 have timestamp values that were captured for the transmit frame. This field is valid only
when the Last Segment control bit (TDES1[30]) in a descriptor is set.
IHE: IP Header Error
16 When set, this bit indicates that the Checksum Offload engine detected an IP header error and consequently did not
modify the transmitted frame for any checksum insertion.


Bit Description
ES: Error Summary
Indicates the logical OR of the following bits:
■ TDES0[16]: IP Header Error
■ TDES0[14]: Jabber Timeout
■ TDES0[13]: Frame Flush
■ TDES0[12]: Payload Checksum Error
15
■ TDES0[11]: Loss of Carrier
■ TDES0[10]: No Carrier
■ TDES0[9]: Late Collision
■ TDES0[8]: Excessive Collision
■ TDES0[2]: Excessive Deferral
■ TDES0[1]: Underflow Error †
JT: Jabber Timeout
14 When set, this bit indicates that the MAC transmitter has experienced a jabber time-out. This bit is only set when the
JD bit of Register 0 (MAC Configuration Register) is not set. †
FF: Frame Flushed
13 When set, this bit indicates that the DMA or MAC transaction layer (MTL) flushed the frame because of a software
flush command given by the CPU. †
12 PCE: Payload Checksum Error
This bit, when set, indicates that the Checksum Offload engine had a failure and did not insert any checksum into
the encapsulated TCP, UDP, or ICMP payload. This failure can be either because of insufficient bytes, as indicated by
12 the IP Header’s Payload Length field, or the MTL starting to forward the frame to the MAC transmitter in
Store-and-Forward mode without the checksum having been calculated yet. This second error condition only occurs
when the transmit FIFO buffer depth is less than the length of the Ethernet frame being transmitted to avoid
deadlock, the MTL starts forwarding the frame when the FIFO buffer is full, even in the store-and-forward mode.
LC: Loss of Carrier
11 When set, this bit indicates that Loss of Carrier occurred during frame transmission (that is, the gmii_crs_i signal
was inactive for one or more transmit clock periods during frame transmission). This is valid only for the frames
transmitted without collision and when the MAC operates in the half-duplex mode. †
NC: No Carrier
10
When set, this bit indicates that the carrier sense signal form the PHY was not asserted during transmission. †
9 Reserved
EC: Excessive Collision
8 When set, this bit indicates that the transmission was aborted after 16 successive collisions while attempting to
transmit the current frame. If Bit 9 (Disable Retry) in Register 0 (MAC Configuration Register) is set, this bit is set
after the first collision and the transmission of the frame is aborted. †
VF: VLAN Frame
7
When set, this bit indicates that the transmitted frame was a VLAN-type frame. †
CC: Collision Count
6:3 This 4-bit counter value indicates the number of collisions occurring before the frame was transmitted. The count is
not valid when the Excessive Collisions bit (TDES0[8]) is set. †


Bit Description
ED: Excessive Deferral
2 When set, this bit indicates that the transmission has ended because of excessive deferral of over 24,288 bit times
(155,680 bits times in 1000-Mbps mode, or in Jumbo frame enabled mode) if Bit 4 (Deferral Check) is set high in
Register 0 (MAC Configuration Register). †
UF: Underflow Error
When set, this bit indicates that the MAC aborted the frame because data arrived late from the Host memory.
1 Underflow Error indicates that the DMA encountered an empty transmit Buffer while transmitting the frame. The
transmission process enters the suspended state and sets both Transmit Underflow (Register 5[5]) and Transmit
Interrupt (Register 5[0]). †
DB: Deferred Bit
0 When set, this bit indicates that the MAC defers before transmission because of the presence of carrier. This bit is
valid only in the half-duplex mode. †

TDES1 contains the buffer sizes and other bits which control the descriptor chain or
ring and the frame being transferred. †
1 Refer to “Buffer Size Calculations” on page 17–24 for further detail on calculating
buffer sizes. †

Bit Description
IC: Interrupt on Completion
31
When set, this bit sets Transmit Interrupt (Register 5[0]) after the present frame has been transmitted. †
LS: Last Segment
30 When set, this bit indicates that the buffer contains the last segment of the frame. When this bit is set, the TBS1 or
TBS2 field should have a non-zero value. †
FS: First Segment
29
When set, this bit indicates that the buffer contains the first segment of a frame. †
CIC: Checksum Insertion Control
These bits control the insertion of checksums in Ethernet frames that encapsulate TCP, UDP, or ICMP over IPv4 or
IPv6 as described below.
■ 0: Do nothing. Checksum Engine is bypassed
■ 1: Insert IPv4 header checksum. Use this value to insert IPv4 header checksum when the frame encapsulates an
IPv4 datagram.
28:27 ■ 2: Insert TCP/UDP/ICMP checksum. The checksum is calculated over the TCP, UDP, or ICMP segment only and
the TCP, UDP, or ICMP pseudo-header checksum is assumed to be present in the corresponding input frame’s
Checksum field. An IPv4 header checksum is also inserted if the encapsulated datagram conforms to IPv4.
■ 3: Insert a TCP/UDP/ICMP checksum that is fully calculated in this engine. In other words, the TCP, UDP, or ICMP
pseudo-header is included in the checksum calculation, and the input frame’s corresponding Checksum field has
an all-zero value. An IPv4 Header checksum is also inserted if the encapsulated datagram conforms to IPv4.
The Checksum engine detects whether the TCP, UDP, or ICMP segment is encapsulated in IPv4 or IPv6 and
processes its data accordingly.


Bit Description
DC: Disable CRC
26 When set, the MAC does not append the CRC to the end of the transmitted frame. This is valid only when the first
segment (TDES1[29]) is set. †
TER: Transmit End of Ring
25 When set, this bit indicates that the descriptor list reached its final descriptor. The returns to the base address of the
list, creating a descriptor ring. †
TCH: Second Address Chained
24 When set, this bit indicates that the second address in the descriptor is the Next descriptor address rather than the
second buffer address. When TDES1[24] is set, TBS2 (TDES1[21–11]) are “don’t care” values. TDES1[25] takes
precedence over TDES1[24]. †
DP: Disable Padding
23 When set, the MAC does not automatically add padding to a frame shorter than 64 bytes. When this bit is reset, the
DMA automatically adds padding and CRC to a frame shorter than 64 bytes and the CRC field is added despite the
state of the DC (TDES1[26]) bit. This is valid only when the first segment (TDES1[29]) is set. †
TTSE: Transmit Timestamp Enable
22 When set, this bit enables IEEE1588 hardware timestamping for the transmit frame referenced by the descriptor.
This field is valid only when the First Segment control bit (TDES1[29]) is set. †
TBS2: Transmit Buffer 2 Size
21:11
These bits indicate the second Data Buffer in bytes. This field is not valid if TDES1[24] is set. †
10:0 These bits indicate the First Data Buffer byte size. If this field is 0, the DMA ignores this buffer and uses Buffer 2 or
next descriptor depending on the value of TCH (Bit 24). †

TDES2 contains the address pointer to the first buffer of the descriptor. †
Table 17–8. Transmit Descriptor 2

Bit Description
Buffer 1 Address Pointer
31:0 These bits indicate the physical address of Buffer 1. There is no limitation on the buffer address alignment. Refer to
“Host Data Buffer Alignment” on page 17–24 for further detail on buffer address alignment. †

TDES3 contains the address pointer either to the second buffer of the descriptor or the
next descriptor. †
Table 17–9. Transmit Descriptor 3

Bit Description
Buffer 2 Address Pointer (Next Descriptor Address)
Indicates the physical address of Buffer 2 when a descriptor ring structure is used. If the Second Address Chained
31:0 (TDES1[24]) bit is set, this address contains the pointer to the physical memory where the Next descriptor is
present. The buffer address pointer must be aligned to the bus width only when TDES1[24] is set. (LSBs are
ignored internally.) †

Receive Descriptor
The DMA in the EMAC requires at least two descriptors when receiving a frame. The
receive state machine of the DMA always attempts to acquire an extra descriptor in
anticipation of an incoming frame. (The size of the incoming frame is unknown).
Before the RX DMA closes a descriptor, it attempts to acquire the next descriptor even
if no frames are received.
In a single descriptor (receive) system, the subsystem generates a descriptor error if
receive buffer is unable to accommodate the incoming frame and the next descriptor
is not owned by the DMA. Thus, the Host is forced to increase either its descriptor
pool or the buffer size. Otherwise, the subsystem starts dropping all incoming
frames. †
Receive Descriptor 0 (RDES0)

RDES0 contains the received frame status, the frame length, and the descriptor
ownership information.
Table 17–10. Receive Descriptor 0 (Part 1 of 3)

Bit Description
OWN: Own Bit
31 When set, this bit indicates that the descriptor is owned by the DMA of the EMAC. When this bit is reset, this bit
indicates that the descriptor is owned by the Host. The DMA clears this bit either when it completes the frame
reception or when the buffers that are associated with this descriptor are full.
AFM: Destination Address Filter Fail
30
When set, this bit indicates a frame that failed in the DA Filter in the MAC. †
FL: Frame Length
These bits indicate the byte length of the received frame that was transferred to host memory (including CRC). This
field is valid when Last Descriptor (RDES0[8]) is set and either the Descriptor Error (RDES0[14]) or Overflow Error
29:16 bit is reset. The frame length also includes the two bytes appended to the Ethernet frame when IP checksum
calculation (Type 1) is enabled and the received frame is not a MAC control frame.
This field is valid when Last Descriptor (RDES0[8]) is set. When the Last Descriptor and Error Summary bits are not
set, this field indicates the accumulated number of bytes that have been transferred for the current frame. †
ES: Error Summary
■ RDES0[0]: Payload Checksum Error
■ RDES0[1]: CRC Error
■ RDES0[3]: Receive Error
15 ■ RDES0[4]: Watchdog Timeout
■ RDES0[6]: Late Collision
■ RDES0[7]: IPC Checksum (Type 2)
■ RDES0[11]: Overflow Error
■ RDES0[14]: Descriptor Error
This field is valid only when the Last Descriptor (RDES0[8]) is set. †
DE: Descriptor Error
14 When set, this bit indicates a frame truncation caused by a frame that does not fit within the current descriptor
buffers, and that the DMA does not own the Next descriptor. The frame is truncated. This field is valid only when the
Last Descriptor (RDES0[8]) is set. †


Bit Description
SAF: Source Address Filter Fail
13
When set, this bit indicates that the SA field of frame failed the SA Filter in the MAC. †
LE: Length Error
12 When set, this bit indicates that the actual length of the frame received and that the Length/ Type field does not
match. This bit is valid only when the Frame Type (RDES0[5]) bit is reset. Length error status is not valid when CRC
error is present. †
OE: Overflow Error
When set, this bit indicates that the received frame is damaged because of buffer overflow in RX FIFO buffer.
11 Note: This bit is set only when the DMA transfers a partial frame to the application. This happens only when the RX
FIFO buffer is operating in the threshold mode. In the store-and-forward mode, all partial frames are dropped
completely in RX FIFO buffer. †
VLAN: VLAN Tag
10 When set, this bit indicates that the frame pointed to by this descriptor is a VLAN frame tagged by the MAC. The
VLAN tagging depends on checking VLAN fields of the received frame based on the Register 7 (VLAN Tag Register)
settings. †
FS: First Descriptor
9 When set, this bit indicates that this descriptor contains the first buffer of the frame. If the size of the first buffer is
0, the second buffer contains the beginning of the frame. If the size of the second buffer is also 0, the next
descriptor contains the beginning of the frame. †
LS: Last Descriptor
8
When set, this bit indicates that the buffers pointed to by this descriptor are the last buffers of the frame †
IPC Checksum Error
When IP Checksum Engine (Type 1) is enabled, this bit, when set, indicates that the 16-bit IPv4 Header checksum
calculated by the MAC did not match the received checksum bytes. Bit 15 (ES) is NOT set when this bit is set in this
7 mode. If this bit is set when Full Checksum Offload Engine (Type 2) is enabled, it indicates an error in the IPv4 or
IPv6 header. This error can be due to inconsistent Ethernet Type field and IP header Version field values, a header
checksum mismatch in IPv4, or an Ethernet frame lacking the expected number of IP header bytes. For more
information, refer to Table 17–11 on page 17–42.
LC: Late Collision
6
When set, this bit indicates that a late collision has occurred while receiving the frame in half-duplex mode. †
FT: Frame Type
5 When set, this bit indicates that the receive frame is an Ethernet-type frame (the LT field is greater than or equal to
0x0600). When this bit is reset, it indicates that the received frame is an IEEE802.3 frame. This bit is not valid for
Runt frames less than 14 bytes. For more information, refer to Table 17–11 on page 17–42. †
RWT: Receive Watchdog Timeout
4 When set, this bit indicates that the receive Watchdog Timer has expired while receiving the current frame and the
current frame is truncated after the Watchdog Timeout. †
RE: Receive Error
3 When set, this bit indicates that the gmii_rxer_i signal is asserted while gmii_rxdv_i is asserted during frame
reception. Error can be of less or no extension, or error (rxd != 0xf) during extension. †
DE: Dribble Bit Error
2 When set, this bit indicates that the received frame has a non-integer multiple of bytes (odd nibbles). This bit is valid
only in MII Mode. †


Bit Description
CE: CRC Error
1 When set, this bit indicates that a CRC error occurred on the received frame. This field is valid only when the Last
Descriptor (RDES0[8]) is set. †
RX MAC Address or Payload Checksum Error
When set, this bit indicates that the RX MAC Address registers value (1 to 15) matched the frame’s DA field. When
reset, this bit indicates that the RX MAC Address Register 0 value matched the DA field. If Full Checksum Offload
0 Engine is enabled, this bit, when set, indicates the TCP, UDP, or ICMP checksum the EMAC calculated does not
match the received encapsulated TCP, UDP, or ICMP segment’s Checksum field. This bit is also set when the
received number of payload bytes does not match the value indicated in the Length field of the encapsulated IPv4 or
IPv6 datagram in the received Ethernet frame. For more information, refer to Table 17–11 on page 17–42. †
When the Full Checksum Offload Engine (Type 2) is enabled, the permutations of Bits
5, 7, and 0 reflect the conditions discussed in Table 17–11. †
Table 17–11. Receive Descriptor 0 when COE (Type 2) is Enabled

Bit 0:
Bit 7: IPC
Bit 5: Frame Payload
Checksum Frame Status
Type Checksum
Error
Error
IEEE 802.3 Type frame (Length field value is less
0 0 0
than 0x0600.) †
IPv4/IPv6 Type frame, no checksum error
1 0 0
detected †
IPv4/IPv6 Type frame with a payload checksum
1 0 1
error (as described for PCE) detected †
IPv4/IPv6 Type frame with an IP header checksum
1 1 0
error (as described for IPC CE) detected †
IPv4/IPv6 Type frame with both IP header and
1 1 1
payload checksum errors detected †
IPv4/IPv6 Type frame with no IP header checksum
0 0 1 error and the payload check bypassed, due to an
unsupported payload †
A Type frame that is neither IPv4 or IPv6 (the
0 1 1 Checksum Offload engine bypasses checksum
completely.) †
0 1 0 Reserved †

RDES1 contains the buffer sizes and other bits that control the descriptor chain or
ring. †

1 For more information about calculating buffer sizes, refer to “Buffer Size
Calculations” on page 17–24. †
Table 17–12. Receive Descriptor 1

Bit Description
Disable Interrupt on Completion
31 When set, this bit prevents the setting of the bit 6 (RI) of the Register 5 (Status Register) for the received frame that
ends in the buffer pointed to by this descriptor. This, in turn, disables the assertion of the interrupt to Host due to RI
for that frame. †
30:26 Reserved †
RER: Receive End of Ring
25 When set, this bit indicates that the descriptor list reached its final descriptor. The DMA returns to the base address
of the list, creating a descriptor Ring. †
RCH: Second Address Chained
second buffer address. When RDES1[24] is set, RBS2 (RDES1[21-11]) is a “don’t care” value. RDES1[25] takes
precedence over RDES1[24]. †
23:22 Reserved †
RBS2: Receive Buffer 2 Size
21:11 These bits indicate the second data buffer size in bytes. The buffer size must be a multiple of 4, even if the value of
RDES3 (buffer2 address pointer) is not aligned to bus width. In the case where the buffer size is not a multiple of 4,
the resulting behavior is undefined. This field is not valid if RDES1[24] is set.
Indicates the first data buffer size in bytes. The buffer size must be a multiple of 4, even if the value of RDES2
10:0 (buffer1 address pointer) is not aligned. In the case where the buffer size is not a multiple of 4, the resulting
behavior is undefined. If this field is 0, the DMA ignores this buffer and uses Buffer 2 or next descriptor depending
on the value of RCH (Bit 24).

RDES2 contains the address pointer to the first data buffer in the descriptor. †
1 For more information about buffer address alignment, refer to“Host Data Buffer
Alignment” on page 17–24. †
Table 17–13. Receive Descriptor 2 (Default Operation)

Bit Description
These bits indicate the physical address of Buffer 1. There are no limitations on the buffer address alignment except
for the following condition: The DMA uses the value programmed in RDES2[1:0] for its address generation when
31:0 the RDES2 value is used to store the start of frame. The DMA performs a write operation with the RDES2[1:0] bits
as 0 during the transfer of the start of frame but the frame data is shifted as per the actual Buffer address pointer.
The DMA ignores RDES2[1:0], if the address pointer is to a buffer where the middle or last part of the frame is
stored.


RDES3 contains the address pointer either to the second data buffer in the descriptor
or to the next descriptor. †
Table 17–14. Receive Descriptor 3

Bit Description
These bits indicate the physical address of Buffer 2 when a descriptor ring structure is used. If the Second Address
Chained (RDES1[24]) bit is set, this address contains the pointer to the physical memory where the Next descriptor
is present.
31:0 If RDES1[24] is set, the buffer (Next descriptor) address pointer must be bus width-aligned (RDES3[1:0] = 0. LSBs
are ignored internally.) However, when RDES1[24] is reset, there are no limitations on the RDES3 value, except for
the following condition: The DMA uses the value programmed in RDES3[1:0] for its buffer address generation when
the RDES3 value is used to store the start of frame. The DMA ignores RDES3[1:0], if the address pointer is to a
buffer where the middle or last part of the frame is stored.
Descriptor Format With IEEE 1588-2005 Timestamps Enabled

The default descriptor format (as described in “Transmit Descriptor” on page 17–36
and “Receive Descriptor” on page 17–40), and field descriptions remain unchanged
when created by software (Own bit is set in DES0). †
However, if the software has enabled IEEE 1588-2005 functionality, the DES2 and
DES3 descriptor fields (refer to Table 17–15) take on a different meaning when the
DMA closes the descriptor (Own bit in DES0 is cleared). †
The DMA updates the DES2 and DES3 with the timestamp value before clearing the
Own bit in DES0. When the EMAC is operating in 32-bit mode, DES2 is updated with
the lower 32 timestamp bits (the Sub-second field, called TSL in subsequent sections)
and DES3 is updated with the upper 32 timestamp bits (the seconds field, called TSH
in subsequent sections).
Table 17–15. Descriptor Fields when DMA Clears the Own Bit
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
DES0
DES1
DES2 Timestamp Low [31:0]
DES3 Timestamp High [31:0]
The following sections describe the details specific to receive and transmit descriptors
in this mode. †

Transmit Descriptor
In addition to the changes described in “Descriptor Format With IEEE 1588-2005
Timestamps Enabled” on page 17–44, the transmit descriptor has additional control
and status bits (TTSE and TTSS, respectively) for timestamping, as shown in
Table 17–16. Software sets the TTSE bit (when the Own bit is set), instructing the
EMAC to generate a timestamp for the corresponding Ethernet frame being
transmitted. The DMA sets the TTSS bit if the timestamp has been updated in the
TDES2 and TDES3 fields when the descriptor is closed (Own bit is cleared).
Table 17–16. Transmit Descriptor Fields - Normal Format

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
T
O
T
TDES0 W RES Status [16:0]
S
N
S
T
T
TDES1 Other Control Fields Buffer 2Byte Count [21:11] Buffer 1 Byte Count [10:0]
S
E
TDES2 TTSL
TDES3 TTSH
Transmit Timestamp Control and Status Fields

The position of these fields is different for normal transmit descriptor and
enhanced format transmit descriptor. The value of this field in both the cases shall
be preserved by the DMA at the time of closing the descriptor. †
Updates to Table 17–14 on page 17–44 and Table 17–6 on page 17–36 for the
default (normal) descriptor format are described below. †
Table 17–17. Transmit Timestamp Status – Normal Descriptor Format Case (TDES0)
Bit Description
TTSS: Transmit Timestamp Status
This field is a status bit indicating that a timestamp was captured for the corresponding transmit frame. When this
17
bit is set, both TDES2 and TDES3 have a timestamp value that was captured for the transmit frame.
This field is valid only when the Last Segment control bit (TDES1[30] in the descriptor) is set. †
Table 17–18. Transmit Timestamp Control – Normal Descriptor Format Case (TDES1)
Bit Description
22 When set, this field enables IEEE1588 hardware timestamping for the transmit frame described by the descriptor.
This field is valid only when the First Segment control bit (TDES1[29] in the descriptor) is set. †

Transmit Timestamp Field

The transmit descriptor format and field descriptions remain unchanged when
they are created by software (when the Own bit is set). However, when the DMA
closes the last descriptor (marked, in the alternative descriptor format, by the LS
bit in TDES1 or TDES0) and IEEE 1588 functionality is enabled (the Own bit is
cleared), the TDES2 and TDES3 descriptor fields are updated with the timestamp,
if taken, for that frame. †
Table 17–19 and Table 17–20 describe the fields that have different meaning when
the descriptor is closed. †
Table 17–19. Transmit Descriptor Fields (TDES2)

Bit Description
TTSL: Transmit Frame Timestamp Low
31:0 This field is updated by DMA with the least significant 32 bits of the timestamp captured for the corresponding
transmit frame. This field has the timestamp only if the Last Segment control bit (LS) in the descriptor is set. †
Table 17–20. Transmit Descriptor Fields (TDES3)

Bit Description
TTSH: Transmit Frame Timestamp High
31:0 This field is updated by DMA with the most significant 32 bits of the timestamp captured for the corresponding
transmit frame. This field has the timestamp only if the Last Segment control bit (LS) in the descriptor is set. †
Receive descriptor
Receive Timestamp
Table 17–21 shows the format of receive descriptor when timestamping is
enabled.
Table 17–21. Receive Timestamp Fields

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
RDES0
RDES1
RDES2 Receive Frame Timestamp Low [31:0]
RDES3 Receive Frame Timestamp High [31:0]
Table 17–22 on page 17–47 and Table 17–23 on page 17–47 describe the fields that
have different meaning for RDES2 and RDES3 when the receive descriptor is
closed and timestamping is enabled.

1 When software disables the Timestamp feature (the TSENA bit in Register
448 is low), the DMA does not update the RDES2 or RDES3 fields of
descriptor before closing the RDES0. †
Table 17–22. Receive Descriptor Fields (RDES2)

Bit Description
RTSL: Receive Frame Timestamp Low
The DMA updates this field with the least significant 32 bits of the timestamp captured for the corresponding
31:0 receive frame. The DMA updates this field only for the last descriptor of the receive frame indicated by Last
Descriptor status bit (RDES0[8]). When this field and the RTSH field in RDES3 show an all-ones value, the
timestamp must be treated as corrupt. †
Table 17–23. Receive descriptor Fields (RDES3)

Bit Description
RTSH: Receive Frame Timestamp High
The DMA updates this field with the most significant 32 bits of the timestamp captured for the corresponding
31:0 receive frame. The DMA updates this field only for the last descriptor of the receive frame indicated by Last
Descriptor status bit (RDES0[8]).
When this field and RDES2’s RTSL field show all-ones values, the timestamp must be treated as corrupt. †
Alternate or Enhanced Descriptors

The alternate (or enhanced) descriptor structure can have 8 DWORDS (32-bytes)
instead of the 4 DWORDS as in the case of normal descriptor format. The features of
the alternate descriptor structure are:
■ The normal descriptor structure allows data buffers of up to 2,048 bytes. The
alternate descriptor structure is implemented to support buffers of up to 8 KB
(useful for Jumbo frames).
■ There is a re-assignment of control and status bits in TDES0, TDES1, RDES0
(advanced timestamp or IPC full offload configuration), and RDES1. †
■ The transmit descriptor stores the timestamp in TDES6 and TDES7 when you
select the advanced timestamp. †
■ This receive descriptor structure is also used for storing the extended status
(RDES4) and timestamp (RDES6 and RDES7) when advanced timestamp, IPC Full
Checksum Offload Engine, or Layer 3 and Layer 4 filter feature is selected. †

■ You can select one of the following options for descriptor structure:
■ If timestamping is enabled in Register 448 (Timestamp Control Register) or
Checksum Offload is enabled in Register 0 (MAC Configuration Register), the
software needs to allocate 32-bytes (8 DWORDS) of memory for every
descriptor. For this, the software should set Bit 7 (Alternate Descriptor Size) of
Register 0 (Bus Mode Register). †
■ If timestamping or Checksum Offload is not enabled, the extended descriptors
(DES4 to DES7) are not required. Therefore, the software can use alternate
descriptors with the default size of 16 bytes (4 DWORDS). For this, the
software should reset Bit 7 (Alternate Descriptor Size) of Register 0 (Bus Mode
Register) to 0. †
■ When alternate descriptor is selected without Timestamp or Receive IPC Full
Checksum Offload Engine (Type 2) feature, the descriptor size is always 4
DWORDs (DES0-DES3). Therefore, the software can use alternate descriptors with
the default size of 16 bytes. †
Transmit Descriptor
The transmit descriptor structure is shown in Table 17–24 on page 17–49. The
application software must program the control bits TDES0[31:18] during descriptor
initialization. When the DMA updates the descriptor, it write backs all the control bits
except the OWN bit (which it clears) and updates the status bits[7:0]. The contents of
the transmitter descriptor word 0 (TDES0) through word 3 (TDES3) are given in
Table 17–26 on page 17–50, Table 17–27 on page 17–53, Table 17–28 on page 17–53, and
Table 17–29 on page 17–53, respectively. †
With the advance timestamp support, the snapshot of the timestamp to be taken can
be enabled for a given frame by setting Bit 25 (TTSE) of TDES0. When the descriptor is
closed (that is, when the OWN bit is cleared), the timestamp is written into TDES6
and TDES7. This is indicated by the status Bit 17 (TTSS) of TDES0 shown in
Table 17–24 on page 17–49. The contents of TDES6 and TDES7 are mentioned in
Table 17–30 on page 17–53 and Table 17–31 on page 17–53. †

1 When advanced timestamp feature is enabled, the software should set Bit 7 of
Register 0 (Bus Mode Register), so that the DMA operates with extended descriptor
size. When this control bit is reset, the TDES4-TDES7 descriptor space is not valid. †
Table 17–24. Transmit Descriptor Fields - Alternate (Enhanced) Format

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
T T
O
T T Ctrl/Status Status
TDES0 W Ctrl [30:26] Ctrl [24:18] Status [16:7]
S S [6:3] [2:0]
N
E S
Ctrl
TDES1 Buffer 2Byte Count [28:16] RES Buffer 1 Byte Count [12:0]
[31:29]
TDES2 Buffer 1 Address [31:0]
TDES3 Buffer 2 Address [31:0] or Next Descriptor Address [31:0]
TDES4 Reserved
TDES5 Reserved
TDES6 Transmit Timestamp Low [31:0]
TDES7 Transmit Timestamp High [31:0]

The DMA always reads or fetches four DWORDS of the descriptor from system
memory to obtain the buffer and control information as shown in Table 17–25. †
Table 17–25. Transmit Descriptor Fetch (Read) for Alternate (Enhanced) Format
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
T Reserve
O SLOT
T d for
TDES0 W Ctrl [30:26] Ctrl [24:18] Reserved for Status [17:7] Number
S Status
N [6:3]
E [2:0]
Ctrl
TDES1 Buffer 2 Byte Count [28:16] RES Buffer 1 Byte Count [12:0]
[31:29]
Table 17–26. Transmit Descriptor Word 0 (TDES0) (Part 1 of 3)

Bit Description
OWN: Own Bit
When set, this bit indicates that the descriptor is owned by the DMA. When this bit is reset, it indicates that the
31 descriptor is owned by the Host. The DMA clears this bit either when it completes the frame transmission or when
the buffers allocated in the descriptor are read completely. The ownership bit of the frame’s first descriptor must be
set after all subsequent descriptors belonging to the same frame have been set. This avoids a possible race
condition between fetching a descriptor and the driver setting an ownership bit. †
IC: Interrupt on Completion
30
When set, this bit sets the Transmit Interrupt (Register 5[0]) after the present frame has been transmitted. †
LS: Last Segment
29 When set, this bit indicates that the buffer contains the last segment of the frame. When this bit is set, the TBS1 or
TBS2 field in TDES1 should have a non-zero value. †
FS: First Segment
28
When set, this bit indicates that the buffer contains the first segment of a frame. †
DC: Disable CRC
27 When this bit is set, the MAC does not append a CRC to the end of the transmitted frame. This is valid only when the
first segment (TDES0[28]) is set. †
DP: Disable Pad
26 When set, the MAC does not automatically add padding to a frame shorter than 64 bytes. When this bit is reset, the
DMA automatically adds padding and CRC to a frame shorter than 64 bytes, and the CRC field is added despite the
state of the DC (TDES0[27]) bit. This is valid only when the first segment (TDES0[28]) is set. †
25 When set, this bit enables IEEE1588 hardware timestamping for the transmit frame referenced by the descriptor.
This field is valid only when the First Segment control bit (TDES0[28]) is set.
24 Reserved


Bit Description
CIC: Checksum Insertion Control
These bits control the checksum calculation and insertion. The following list describes the bit encoding:
■ 0: Checksum Insertion Disabled.
■ 1: Only IP header checksum calculation and insertion are enabled.
23:22 ■ 2: IP header checksum and payload checksum calculation and insertion are enabled, but pseudoheader
checksum is not calculated in hardware.
■ 3: IP Header checksum and payload checksum calculation and insertion are enabled, and pseudoheader
checksum is calculated in hardware.
This field is valid when the First Segment control bit (TDES0[28]) is set.
TER: Transmit End of Ring
of the list, creating a descriptor ring. †
TCH: Second Address Chained
When set, this bit indicates that the second address in the descriptor is the Next descriptor address rather than the
20
second buffer address. When TDES0[20] is set, TBS2 (TDES1[28:16]) is a “don’t care” value.
TDES0[21] takes precedence over TDES0[20]. †
19:18 Reserved
TTSS: Transmit Timestamp Status
17 This field is used as a status bit to indicate that a timestamp was captured for the described transmit frame. When
this bit is set, TDES2 and TDES3 have a timestamp value captured for the transmit frame. This field is only valid
when the descriptor’s Last Segment control bit (TDES0[29]) is set. †
IHE: IP Header Error
When set, this bit indicates that the MAC transmitter detected an error in the IP datagram header. The transmitter
checks the header length in the IPv4 packet against the number of header bytes received from the application and
16 indicates an error status if there is a mismatch. For IPv6 frames, a header error is reported if the main header length
is not 40 bytes. Furthermore, the Ethernet Length/Type field value for an IPv4 or IPv6 frame must match the IP
header version received with the packet. For IPv4 frames, an error status is also indicated if the Header Length field
has a value less than 0x5. †
ES: Error Summary
■ TDES0[14]: Jabber Timeout
■ TDES0[13]: Frame Flush
■ TDES0[11]: Loss of Carrier
■ TDES0[10]: No Carrier
15
■ TDES0[9]: Late Collision
■ TDES0[8]: Excessive Collision
■ TDES0[2]: Excessive Deferral
■ TDES0[1]: Underflow Error
■ TDES0[16]: IP Header Error
■ TDES0[12]: IP Payload Error †


Bit Description
JT: Jabber Timeout
14 When set, this bit indicates the MAC transmitter has experienced a jabber time-out. This bit is only set when Bit 22
(Jabber Disable) of Register 0 (MAC Configuration Register) is not set. †
FF: Frame Flushed
13 When set, this bit indicates that the DMA or MTL flushed the frame because of a software Flush command given by
the CPU. †
IPE: IP Payload Error
When set, this bit indicates that MAC transmitter detected an error in the TCP, UDP, or ICMP IP datagram payload.
12
The transmitter checks the payload length received in the IPv4 or IPv6 header against the actual number of TCP,
UDP, or ICMP packet bytes received from the application and issues an error status in case of a mismatch. †
LC: Loss of Carrier
11 When set, this bit indicates that a loss of carrier occurred during frame transmission (that is, the gmii_crs_i
signal was inactive for one or more transmit clock periods during frame transmission). This is valid only for the
frames transmitted without collision when the MAC operates in the half-duplex mode. †
NC: No Carrier
10
When set, this bit indicates that the Carrier Sense signal form the PHY was not asserted during transmission. †
9 Reserved
EC: Excessive Collision
8 When set, this bit indicates that the transmission was aborted after 16 successive collisions while attempting to
transmit the current frame. If Bit 9 (Disable Retry) bit in the Register 0 (MAC Configuration Register) is set, this bit
is set after the first collision, and the transmission of the frame is aborted. †
VF: VLAN Frame
7
When set, this bit indicates that the transmitted frame is a VLAN-type frame. †
CC: Collision Count (Status field)
6:3 These status bits indicate the number of collisions that occurred before the frame was transmitted. This count is not
valid when the Excessive Collisions bit (TDES0[8]) is set. The EMAC updates this status field only in the half-duplex
mode.
ED: Excessive Deferral
2 When set, this bit indicates that the transmission has ended because of excessive deferral of over 24,288 bit times
(155,680 bits times in 1,000-Mbps mode or if Jumbo frame is enabled) if Bit 4 (Deferral Check) bit in Register 0
(MAC Configuration Register) is set high. †
UF: Underflow Error
When set, this bit indicates that the MAC aborted the frame because the data arrived late from the Host memory.
1 Underflow Error indicates that the DMA encountered an empty transmit buffer while transmitting the frame. The
transmission process enters the Suspended state and sets both Transmit Underflow (Register 5[5]) and Transmit
Interrupt (Register 5[0]). †
DB: Deferred Bit
0 When set, this bit indicates that the MAC defers before transmission because of the presence of carrier. This bit is
valid only in the half-duplex mode. †

Table 17–27. Transmit Descriptor Word 1 (TDES1)

Bit Description
31:29 Reserved
28:16 This field indicates the second data buffer size in bytes. This field is not valid if TDES0[20] is set. For more
information about calculating buffer sizes, refer to “Buffer Size Calculations” on page 17–24. †
15:13 Reserved †
12:0 These bits indicate the first data buffer byte size, in bytes. If this field is 0, the DMA ignores this buffer and uses
Buffer 2 or the next descriptor, depending on the value of TCH (TDES0[20]). †
Table 17–28. Transmit Descriptor 2 (TDES2)

Bit Description
31:0 These bits indicate the physical address of Buffer 1. There is no limitation on the buffer address alignment. For more
information about buffer address alignment, refer to “Host Data Buffer Alignment” on page 17–24. †

Bit Description
Indicates the physical address of Buffer 2 when a descriptor ring structure is used. If the Second Address Chained
31:0 (TDES1[24]) bit is set, this address contains the pointer to the physical memory where the Next descriptor is
present. The buffer address pointer must be aligned to the bus width only when TDES1[24] is set. (LSBs are
ignored internally.) †

Bit Description
TTSL: Transmit Frame Timestamp Low
transmit frame. This field has the timestamp only if the Last Segment bit (LS) in the descriptor is set and Timestamp
status (TTSS) bit is set. †

Bit Description
TTSH: Transmit Frame Timestamp High
receive frame. This field has the timestamp only if the Last Segment bit (LS) in the descriptor is set and Timestamp
status (TTSS) bit is set. †

Receive Descriptor
The structure of the received descriptor is shown in Table 17–32. This can have 32
bytes of descriptor data (8 DWORDs) when advanced timestamp or IPC Full Offload
feature is selected. When either of these features is enabled, the Software should set
Bit 7 of Register 0 (Bus Mode Register) so that the DMA operates with extended
descriptor size. When this control bit is reset, the RDES0[0] is always cleared and the
RDES4-RDES7 descriptor space is not valid. †
Table 17–32. Receive Descriptor Fields - Alternate (Enhanced) Format

31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
O
TDES0 W Status [30:0]
N
C
R
T RES Ctrl
TDES1 Buffer 2Byte Count [28:16] E Buffer 1 Byte Count [12:0]
R [30:29] [15:14]
S
L
TDES4 Extended status [31:0]
TDES5 Reserved
TDES6 Transmit Timestamp Low [31:0]
TDES7 Transmit Timestamp High [31:0]
The contents of RDES0 are identified in Table 17–33. The contents of RDES1 through
RDES3 are identified in Table 17–34 on page 17–56, Table 17–35 on page 17–57, and

Bit Description
OWN: Own Bit
31 When set, this bit indicates that the descriptor is owned by the DMA of the EMAC. When this bit is reset, this bit
indicates that the descriptor is owned by the Host. The DMA clears this bit either when it completes the frame
reception or when the buffers that are associated with this descriptor are full.
AFM: Destination Address Filter Fail
30
When set, this bit indicates a frame that failed in the DA Filter in the MAC. †


Bit Description
FL: Frame Length
These bits indicate the byte length of the received frame that was transferred to host memory (including CRC). This
field is valid when Last Descriptor (RDES0[8]) is set and either the Descriptor Error (RDES0[14]) or Overflow Error
29:16 bits are reset. The frame length also includes the two bytes appended to the Ethernet frame when IP checksum
calculation (Type 1) is enabled and the received frame is not a MAC control frame.
This field is valid when Last Descriptor (RDES0[8]) is set. When the Last Descriptor and Error Summary bits are not
set, this field indicates the accumulated number of bytes that have been transferred for the current frame. †
ES: Error Summary
■ RDES0[1]: CRC Error
■ RDES0[3]: Receive Error
■ RDES0[4]: Watchdog Timeout
15 ■ RDES0[6]: Late Collision
■ RDES0[7]: Giant Frame
■ RDES4[4:3]: IP Header or Payload Error
■ RDES0[11]: Overflow Error
■ RDES0[14]: Descriptor Error
This field is valid only when the Last Descriptor (RDES0[8]) is set. †
DE: Descriptor Error
14 When set, this bit indicates a frame truncation caused by a frame that does not fit within the current descriptor
buffers, and that the DMA does not own the Next descriptor. The frame is truncated. This field is valid only when the
Last Descriptor (RDES0[8]) is set. †
SAF: Source Address Filter Fail
13
When set, this bit indicates that the SA field of frame failed the SA Filter in the MAC. †
LE: Length Error
12 When set, this bit indicates that the actual length of the frame received and that the Length/ Type field does not
match. This bit is valid only when the Frame Type (RDES0[5]) bit is reset. †
OE: Overflow Error
When set, this bit indicates that the received frame was damaged because of buffer overflow in MTL.
11 Note: This bit is set only when the DMA transfers a partial frame to the application. This happens only when the RX
FIFO buffer is operating in the threshold mode. In the store-and-forward mode, all partial frames are dropped
completely in RX FIFO buffer. †
VLAN: VLAN Tag
10 When set, this bit indicates that the frame to which this descriptor is pointing is a VLAN frame tagged by the MAC.
The VLAN tagging depends on checking the VLAN fields of received frame based on the Register 7 (VLAN Tag
Register) setting. †
FS: First Descriptor
9 When set, this bit indicates that this descriptor contains the first buffer of the frame. If the size of the first buffer is
0, the second buffer contains the beginning of the frame. If the size of the second buffer is also 0, the next
descriptor contains the beginning of the frame. †
LS: Last Descriptor
8
When set, this bit indicates that the buffers pointed to by this descriptor are the last buffers of the frame †


Bit Description
Timestamp Available, IP Checksum Error (Type1), or Giant Frame
When advanced timestamp feature is present, when set, this bit indicates that a snapshot of the Timestamp is
written in descriptor words 6 (RDES6) and 7 (RDES7). This is valid only when the Last Descriptor bit (RDES0[8]) is
set.
7 When IP Checksum Engine (Type 1) is selected, this bit, when set, indicates that the 16-bit IPv4 Header checksum
calculated by the EMAC did not match the received checksum bytes.
Otherwise, this bit, when set, indicates the Giant frame Status. Giant frames are larger than 1,518-byte (or
1,522-byte for VLAN or 2,000-byte when Bit 27 (2KPE) of MAC Configuration register is set) normal frames and
larger than 9,018-byte (9,022-byte for VLAN) frame when Jumbo frame processing is enabled.
LC: Late Collision
6
When set, this bit indicates that a late collision has occurred while receiving the frame in the half-duplex mode. †
FT: Frame Type
5 When set, this bit indicates that the receive frame is an Ethernet-type frame (the LT field is greater than or equal to
0x0600). When this bit is reset, it indicates that the received frame is an IEEE802.3 frame. This bit is not valid for
Runt frames less than 14 bytes.
RWT: Receive Watchdog Timeout
4 When set, this bit indicates that the receive Watchdog Timer has expired while receiving the current frame and the
current frame is truncated after the Watchdog Timeout. †
RE: Receive Error
3 When set, this bit indicates that the gmii_rxer_i signal is asserted while gmii_rxdv_i is asserted during frame
reception. Error can be of less or no extension, or error (rxd !=0xf) during extension. †
DE: Dribble Bit Error
2 When set, this bit indicates that the received frame has a non-integer multiple of bytes (odd nibbles). This bit is valid
only in the MII Mode. †
CE: CRC Error
1 When set, this bit indicates that a CRC error occurred on the received frame. This field is valid only when the Last
Descriptor (RDES0[8]) is set. †
Extended Status Available/RX MAC Address
When either advanced timestamp or IP Checksum Offload (Type 2) is present, this bit, when set, indicates that the
extended status is available in descriptor word 4 (RDES4). This is valid only when the Last Descriptor bit
0 (RDES0[8]) is set.
When Advance Timestamp Feature or IPC Full Offload is not selected, this bit indicates RX MAC Address status.
When set, this bit indicates that the RX MAC Address registers value (1 to 15) matched the frame’s DA field. When
reset, this bit indicates that the RX MAC Address Register 0 value matched the DA field. †
Table 17–34. Receive Descriptor Fields 1 (RDES1)

Bit Description
DIC: Disable Interrupt on Completion
31 When set, this bit prevents setting the Status Register’s RI bit (CSR5[6]) for the received frame ending in the buffer
indicated by this descriptor. This, in turn, disables the assertion of the interrupt to Host because of RI for that
frame. †
30:29 Reserved †


Bit Description
These bits indicate the second data buffer size, in bytes. The buffer size must be a multiple of 4, even if the value of
RDES3 (buffer2 address pointer) is not aligned to bus width. If the buffer size is not an appropriate multiple of 4, the
28:16
resulting behavior is undefined. This field is not valid if RDES1[14] is set. For more information about calculating
buffer sizes, refer to “Buffer Size Calculations” on page 17–24.
RER: Receive End of Ring

of the list, creating a descriptor ring. †
RCH: Second Address Chained
second buffer address. When this bit is set, RBS2 (RDES1[28:16]) is a “don’t care” value. RDES1[15] takes
precedence over RDES1[14]. †
13 Reserved †
Indicates the first data buffer size in bytes. The buffer size must be a multiple of 4, even if the value of RDES2
12:0 (buffer1 address pointer) is not aligned. When the buffer size is not a multiple of 4, the resulting behavior is
undefined. If this field is 0, the DMA ignores this buffer and uses Buffer 2 or next descriptor depending on the value
of RCH (Bit 14). For more information about calculating buffer sizes, refer to “Buffer Size Calculations” on
page 17–24.

Bit Description
These bits indicate the physical address of Buffer 1. There are no limitations on the buffer address alignment except
for the following condition: The DMA uses the value programmed in RDES2[1:0] for its address generation when
31:0 the RDES2 value is used to store the start of the frame. The DMA performs a write operation with the RDES2[1:0]
bits as 0 during the transfer of the start of the frame but the frame is shifted as per the actual buffer address pointer.
The DMA ignores RDES2[1:0] if the address pointer is to a buffer where the middle or last part of the frame is
stored. For more information about buffer address alignment, refer to “Host Data Buffer Alignment” on page 17–24.
Table 17–36. Receive descriptor Fields 3 (RDES3)

Bit Description
These bits indicate the physical address of Buffer 2 when a descriptor ring structure is used. If the Second Address
Chained (RDES1[24]) bit is set, this address contains the pointer to the physical memory where the Next descriptor
is present.
31:0 If RDES1[24] is set, the buffer (Next descriptor) address pointer must be bus width-aligned (RDES3[1:0] = 0. LSBs
are ignored internally.) However, when RDES1[24] is reset, there are no limitations on the RDES3 value, except for
the following condition: The DMA uses the value programmed in RDES3 [1:0] for its buffer address generation
when the RDES3 value is used to store the start of frame. The DMA ignores RDES3 [1:0] if the address pointer is to
a buffer where the middle or last part of the frame is stored.

The extended status written is as shown in Table 17–37. The extended status is written
only when there is status related to IPC or timestamp available. The availability of
extended status is indicated by Bit 0 of RDES0. This status is available only when the
Advance Timestamp or IPC Full Offload feature is selected. †
Table 17–37. Receive Descriptor Fields 4 (RDES4) (Part 1 of 2)

Bit Description
31:28 Reserved †
Layer 3 and Layer 4 Filter Number Matched
These bits indicate the number of the Layer 3 and Layer 4 Filter that matched the received frame.
■ 00: Filter 0
■ 01: Filter 1
27:26
■ 10: Filter 2
■ 11: Filter 3
This field is valid only when Bit 24 or Bit 25 is set high. When more than one filter matches, these bits give only the
lowest filter number. †
Layer 4 Filter Match
When set, this bit indicates that the received frame matches one of the enabled Layer 4 Port Number fields. This
status is given only when one of the following conditions is true:
25
■ Layer 3 fields are not enabled and all enabled Layer 4 fields match.
■ All enabled Layer 3 and Layer 4 filter fields match.
When more than one filter matches, this bit gives the layer 4 filter status of filter indicated by Bits [27:26]. †
Layer 3 Filter Match
When set, this bit indicates that the received frame matches one of the enabled Layer 3 IP Address fields.
This status is given only when one of the following conditions is true:
24
■ All enabled Layer 3 fields match and all enabled Layer 4 fields are bypassed.
■ All enabled filter fields match.
When more than one filter matches, this bit gives the layer 3 filter status of filter indicated by Bits [27:26]. †
23:15 Reserved
Timestamp Dropped
14 When set, this bit indicates that the timestamp was captured for this frame but got dropped in the MTL RX FIFO
buffer because of overflow.
PTP Version
13 When set, this bit indicates that the received PTP message is having the IEEE 1588 version 2 format. When reset, it
has the version 1 format.
PTP Frame Type
12 When set, this bit indicates that the PTP message is sent directly over Ethernet. When this bit is not set and the
message type is non-zero, it indicates that the PTP message is sent over UDP-IPv4 or UDP-IPv6. The information
about IPv4 or IPv6 can be obtained from Bits 6 and 7.

Table 17–37. Receive Descriptor Fields 4 (RDES4) (Part 2 of 2)

Bit Description
Message Type
These bits are encoded to give the type of the message received.
■ 0000: No PTP message received
■ 0001: SYNC (all clock types)
■ 0010: Follow_Up (all clock types)
■ 0011: Delay_Req (all clock types)
■ 0100: Delay_Resp (all clock types)
11:8 ■ 0101: Pdelay_Req (in peer-to-peer transparent clock)
■ 0110: Pdelay_Resp (in peer-to-peer transparent clock)
■ 0111: Pdelay_Resp_Follow_Up (in peer-to-peer transparent clock)
■ 1000: Announce
■ 1001: Management
■ 1010: Signaling
■ 1011-1110: Reserved
■ 1111: PTP packet with Reserved message type
IPv6 Packet Received
7 When set, this bit indicates that the received packet is an IPv6 packet. This bit is updated only when Bit 10 (IPC) of
Register 0 (MAC Configuration Register) is set.
IPv4 Packet Received
6 When set, this bit indicates that the received packet is an IPv4 packet. This bit is updated only when Bit 10 (IPC) of
Register 0 (MAC Configuration Register) is set.
IP Checksum Bypassed
5
When set, this bit indicates that the checksum offload engine is bypassed.
IP Payload Error
When set, this bit indicates that the 16-bit IP payload checksum (that is, the TCP, UDP, or ICMP checksum) that the
4 EMAC calculated does not match the corresponding checksum field in the received segment. It is also set when the
TCP, UDP, or ICMP segment length does not match the payload length value in the IP Header field. This bit is valid
when either Bit 7 or Bit 6 is set.
IP Header Error
3 When set, this bit indicates that either the 16-bit IPv4 header checksum calculated by the EMAC does not match the
received checksum bytes, or the IP datagram version is not consistent with the Ethernet Type value. This bit is valid
when either Bit 7 or Bit 6 is set.
IP Payload Type
These bits indicate the type of payload encapsulated in the IP datagram processed by the receive Checksum Offload
Engine (COE). The COE also sets these bits to 0 if it does not process the IP datagram’s payload due to an IP header
error or fragmented IP.
■ 0: Unknown or did not process IP payload
2:0 ■ 1: UDP
■ 2: TCP
■ 3: ICMP
■ 4–7: Reserved
This bit is valid when either Bit 7 or Bit 6 is set.

RDES6 and RDES7 contain the snapshot of the timestamp. The availability of the
snapshot of the timestamp in RDES6 and RDES7 is indicated by Bit 7 in the RDES0
descriptor. The contents of RDES6 and RDES7 are identified in Table 17–38 and

Bit Description
RTSL: Receive Frame Timestamp Low
receive frame. This field is updated by DMA only for the last descriptor of the receive frame which is indicated by
Last Descriptor status bit (RDES0[8]). †

Bit Description
RTSH: Receive Frame Timestamp High
receive frame. This field is updated by DMA only for the last descriptor of the receive frame which is indicated by
Last Descriptor status bit (RDES0[8]). †
Initializing DMA
This section provides the instructions for initializing the DMA/MAC registers in the
proper sequence. Perform the following steps to initialize the DMA:
1. Provide a software reset. This resets all of the EMAC internal registers and logic.
(DMA Register 0 (Bus Mode Register) – bit 0). †
2. Wait for the completion of the reset process (poll bit 0 of the DMA Register 0 (Bus
Mode Register), which is only cleared after the reset operation is completed). †
3. Poll the bits of Register 11 (AHB or AXI Status) to confirm that all previously
initiated (before software-reset) or ongoing transactions are complete.
1 If the application cannot poll the register after soft reset (because of
performance reasons), then it is recommended that you continue with the
next steps and check this register again (as mentioned in step 12) before
triggering the DMA operations. †
4. Program the following fields to initialize the Bus Mode Register by setting values
in DMA Register 0 (Bus Mode Register): †
a. Mixed Burst and AAL
b. Fixed burst or undefined burst †
c. Burst length values and burst mode values. †
d. Descriptor Length (only valid if Ring Mode is used) †
e. TX and RX DMA Arbitration scheme †
5. Program the interface options in Register 10 (AXI Bus Mode Register). If fixed
burst-length is enabled, then select the maximum burst-length possible on the bus
(bits[7:1]). †

6. Create a proper descriptor chain for transmit and receive. In addition, ensure that
the receive descriptors are owned by DMA (bit 31 of descriptor should be set).
When OSF mode is used, at least two descriptors are required. For more
information about descriptors, refer to“Normal Descriptor” on page 17–36 and
“Alternate or Enhanced Descriptors” on page 17–47. †
7. Make sure that your software creates three or more different transmit or receive
descriptors in the chain before reusing any of the descriptors. †
8. Initialize receive and transmit descriptor list address with the base address of the
transmit and receive descriptor (Register 3 (Receive Descriptor List Address
Register) and Register 4 (Transmit Descriptor List Address Register)
respectively). †
9. Program the following fields to initialize the mode of operation in Register 6
(Operation Mode Register)
a. Receive and Transmit Store And Forward †
b. Receive and Transmit Threshold Control (RTC and TTC) †
c. Hardware Flow Control enable †
d. Flow Control Activation and De-activation thresholds for MTL Receive and
Transmit FIFO buffers (RFA and RFD) †
e. Error frame and undersized good frame forwarding enable †
f. OSF Mode †
10. Clear the interrupt requests, by writing to those bits of the status register
(interrupt bits only) that are set. For example, by writing 1 into bit 16, the normal
interrupt summary clears this bit (DMA Register 5 (Status Register)). †
11. Enable the interrupts by programming the Register 7 (Interrupt Enable Register). †
1 Perform step 12 only if you did not perform step 3. †
12. Read Register 11 (AHB or AXI Status) to confirm that all previous transactions are
complete. †
1 If any previous transaction is still in progress when you read the Register 11
(AHB or AXI Status), then it is strongly recommended to check the slave
components addressed by the master interface.
13. Start the receive and transmit DMA by setting SR (bit 1) and ST (bit 13) of the
control register (DMA Register 6 (Operation Mode Register)). †
Initializing MAC
The following MAC Initialization operations can be performed after DMA
initialization. If the MAC initialization is done before the DMA is set-up, then enable
the MAC receiver (last step below) only after the DMA is active. Otherwise, received
frames fills the RX FIFO buffer and overflow. †

1. Program the EMAC Register 4 (GMII Address Register) for controlling the
management cycles for external PHY. For example, Physical Layer Address PA
(bits 15-11). In addition, set bit 0 (GMII Busy) for writing into PHY and reading
from PHY. †
2. Read the 16-bit data of Register 5 (GMII Data Register) from the PHY for link up,
speed of operation, and mode of operation, by specifying the appropriate address
value in bits 15-11 of Register 4 (GMII Address Register). †
3. Provide the MAC address registers (Register 16 (MAC Address0 High Register)
and Register 17 (MAC Address0 Low Register)). Because 128 MAC addresses are
supported, you need to program the MAC addresses accordingly.
4. Program Register 2 (Hash Table High Register) and Register 3 (Hash Table Low
Register).
5. Program the following fields to set the appropriate filters for the incoming frames
in Register 1 (MAC Frame Filter): †
a. Receive All †
b. Promiscuous mode †
c. Hash or Perfect Filter †
d. Unicast, multicast, broadcast, and control frames filter settings †
6. Program the following fields for proper flow control in Register 6 (Flow Control
Register): †
a. Pause time and other pause frame control bits †
b. Receive and Transmit Flow control bits †
c. Flow Control Busy/Backpressure Activate †
7. Program the Interrupt Mask register bits, as required, and if applicable, for your
configuration. †
8. Program the appropriate fields in Register 0 (MAC Configuration Register). For
example, Interframe gap while transmission and jabber disable. Based on the
Auto-negotiation you can set the Duplex mode (bit 11) or port select (bit 15). †
9. Set Bit 3 (TE) and Bit 2 (RE) in Register 0 (MAC Configuration Register). †
1 Do not change the configuration (such as duplex mode, speed, port, or loopback)
when the EMAC DMA is actively transmitting or receiving. The Software should
change these parameters only when the EMAC DMA transmitter and receiver are not
active.
Performing Normal Receive and Transmit Operation

For normal operation, perform the following steps: †
1. For normal transmit and receive interrupts, read the interrupt status. Then, poll
the descriptors, reading the status of the descriptor owned by the Host (either
transmit or receive). †

2. Set appropriate values for the descriptors, ensuring that transmit and receive
descriptors are owned by the DMA to resume the transmission and reception of
data. †
3. If the descriptors are not owned by the DMA (or no descriptor is available), the
DMA goes into SUSPEND state. The transmission or reception can be resumed by
freeing the descriptors and issuing a poll demand by writing 0 into the TX/RX poll
demand register (Register 1 (Transmit Poll Demand Register) and Register 2
(Receive Poll Demand Register)). †
4. The values of the current host transmitter or receiver descriptor address pointer
can be read for the debug process (Register 18 (Current Host Transmit Descriptor
Register) and Register 19 (Current Host Receive Descriptor Register)). †
5. The values of the current host transmit buffer address pointer and receive buffer
address pointer can be read for the debug process (Register 20 (Current Host
Transmit Buffer Address Register) and Register 21 (Current Host Receive Buffer
Address Register)). †
Stopping and Starting Transmission

Perform the following steps to pause the transmission for some time: †
1. Disable the transmit DMA (if applicable), by clearing bit 13 (Start or Stop
Transmission Command) of Register 6 (Operation Mode Register). †
2. Wait for any previous frame transmissions to complete. You can check this by
reading the appropriate bits of Register 9 (Debug Register). †
3. Disable the MAC transmitter and MAC receiver by clearing Bit 3 (TE) and Bit 2
(RE) in Register 0 (MAC Configuration Register). †
4. Disable the receive DMA (if applicable), after making sure that the data in the RX
FIFO buffer is transferred to the system memory (by reading Register 9 (Debug
Register)). †
5. Make sure that both TX FIFO buffer and RX FIFO buffer are empty. †
6. To re-start the operation, first start the DMAs, and then enable the MAC
transmitter and receiver. †
Programming Guidelines for Energy Efficient Ethernet

Entering and Exiting the TX LPI Mode
The Energy Efficient Ethernet (EEE) feature is available in the EMAC. To use it,
perform the following steps during EMAC initialization:
1. Read the PHY register through the MDIO interface, check if the remote end has the
EEE capability, and then negotiate the timer values. †
2. Program the PHY registers through the MDIO interface (including the
RX_CLK_stoppable bit that indicates to the PHY whether to stop RX clock in LPI
mode.) †
3. Program Bits[16:5], LST, and Bits[15:0], TWT, in Register 13 (LPI Timers Control
Register). †

4. Read the link status of the PHY chip by using the MDIO interface and update Bit
17 (PLS) of Register 12 (LPI Control and Status Register) accordingly. This update
should be done whenever the link status in the PHY chip changes. †
5. Set Bit 16 (LPIEN) of Register 12 (LPI Control and Status Register) to make the
MAC enter the LPI state. The MAC enters the LPI mode after completing the
transmission in progress and sets Bit 0 (TLPIEN). †
1 To make the MAC enter the LPI state only after it completes the
transmission of all queued frames in the TX FIFO buffer, you should set Bit
19 (LPITXA) in Register 12 (LPI Control and Status Register). †
1 To switch off the transmit clock during the LPI state, use the
sbd_tx_clk_gating_ctrl_o signal for gating the clock input. †
1 To switch off the CSR clock or power to the rest of the system during the
LPI state, you should wait for the TLPIEN interrupt of Register 12 (LPI
Control and Status Register) to be generated. Restore the clocks before
performing the step 6 when you want to come out of the LPI state. †
6. Reset Bit 16 (LPIEN) of Register 12 (LPI Control and Status Register) to bring the
MAC out of the LPI state. †
The MAC waits for the time programmed in Bits [15:0], TWT, before setting the
TLPIEX interrupt status bit and resuming the transmission. †
Gating Off the CSR Clock in the LPI Mode

You can gate off the CSR clock to save the power when the MAC is in the Low-Power
Idle (LPI) mode. †
Gating Off the CSR Clock in the RX LPI Mode

The following operations are performed when the MAC receives the LPI pattern from
the PHY. †
1. The MAC RX enters the LPI mode and the RX LPI entry interrupt status [RLPIEN
interrupt of Register 12 (LPI_Control_Status)] is set. †
2. The interrupt pin (sbd_intr_o) is asserted. The sbd_intr_o interrupt is cleared
when the host reads the Register 12 (LPI_Control_Status). †
After the sbd_intr_o interrupt is asserted and the MAC TX is also in the LPI mode,
you can gate-off the CSR clock. If the MAC TX is not in the LPI mode when you gate
off the CSR clock, the events on the MAC transmitter do not get reported or updated
in the CSR. †
For restoring the CSR clock, wait for the LPI exit indication from the PHY after which
the MAC asserts the LPI exit interrupt on lpi_intr_o (synchronous to clk_rx_i). The
lpi_intr_o interrupt is cleared when Register 12 is read. †
Gating Off the CSR Clock in the TX LPI Mode

The following operations are performed when Bit 16 (LPIEN) of Register 12 (LPI
Control and Status Register) is set: †
1. The Transmit LPI Entry interrupt (TLPIEN bit of Register 12) is set. †

2. The interrupt pin (sbd_intr_o) is asserted. The sbd_intr_o interrupt is cleared

when the host reads the Register 12. †
After the sbd_intr_o interrupt is asserted and the MAC RX is also in the LPI mode,
you can gate off the CSR clock. If the MAC RX is not in the LPI mode when you gate
off the CSR clock, the events on the MAC receiver do not get reported or updated in
the CSR. †
For restoring the CSR clock, switch on the CSR clock when the MAC has to come out
of the TX LPI mode. †
After the CSR clock is resumed, reset Bit 16 (LPIEN) of Register 12 (LPI Control and
Status Register) to bring the MAC out of the LPI mode. †
Programming Guidelines for Flexible Pulse-Per-Second (PPS) Output

Generating Single Pulse on PPS
To generate single Pulse on PPS: †
1. Program 11 or 10 (for interrupt) in Bits [6:5], TRGTMODSEL, of Register 459 (PPS
Control Register). This instructs the MAC to use the Target Time registers (register
455 and 456) for start time of PPS signal output. †
2. Program the start time value in the Target Time registers (register 455 and 456). †
3. Program the width of the PPS signal output in Register 473 (PPS0 Width
Register). †
4. Program Bits [3:0], PPSCMD, of Register 459 (PPS Control Register) to 0001. This
instructs the MAC to generate single pulse on the PPS signal output at the time
programmed in the Target Time registers (register 455 and 456). †
Once the PPSCMD is executed (PPSCMD bits = 0), you can cancel the pulse
generation by giving the Cancel Start Command (PPSCMD=0011) before the
programmed start time elapses. You can also program the behavior of the next pulse
in advance. To program the next pulse: †
1. Program the start time for the next pulse in the Target Time registers (register 455
and 456). This time should be more than the time at which the falling edge occurs
for the previous pulse. †
2. Program the width of the next PPS signal output in Register 473 (PPS0 Width
Register). †
3. Program Bits [3:0], PPSCMD, of Register 459 (PPS Control Register) to generate a
single pulse after the time at which the previous pulse is de-asserted. This instructs
the MAC to generate single pulse on the PPS signal output, at the time
programmed in Target Time registers. If you give this command before the
previous pulse becomes low, then the new command overwrites the previous
command and the EMAC may generate only 1 extended pulse.
Generating a Pulse Train on PPS

To generate a pulse train on PPS: †

1. Program 11 or 10 (for interrupt) in Bits [6:5], TRGTMODSEL, of Register 459 (PPS

455 and 456) for start time of the PPS signal output. †
2. Program the start time value in the Target Time registers (register 455 and 456). †
3. Program the interval value between the train of pulses on the PPS signal output in
Register 473 (PPS0 Width Register). †
4. Program the width of the PPS signal output in Register 473 (PPS0 Width
Register). †
5. Program Bits[3:0], PPSCMD, of Register 459 (PPS Control Register) to 0010. This
instructs the MAC to generate train of pulses on the PPS signal output with start
time programmed in Target Time registers (register 455 and 456). By default, the
PPS pulse train is free-running unless stopped by ‘STOP Pulse train at time’ or
‘STOP Pulse Train immediately’ commands. †
6. Program the stop value in the Target Time registers (register 455 and 456). Ensure
that Bit 31 (TSTRBUSY) of Register 456 (Target Time Nanoseconds Register) is
reset before programming the Target Time registers (register 455 and 456) again. †
7. Program the PPSCMD field (bit 3:0) of Register 459 (PPS Control Register) to 0100.
This stops the train of pulses on PPS signal output after the programmed stop time
specified in Step 6 elapses. †
You can stop the pulse train at any time by programming 0101 in the PPSCMD field.
Similarly, you can cancel the Stop Pulse train command (given in step 7) by
programming 0110 in the PPSCMD field before the time (programmed in step 6)
elapses. You can cancel the pulse train generation by programming 0011 in the
PPSCMD field before the programmed start time (in step 2) elapses. †
Generating an Interrupt without Affecting the PPS

The Bits [6:5], TRGTMODSEL, of the Register 459 (PPS Control Register) enable you to
program the Target Time registers (register 455 and 456) to do any one of the
following: †
■ Generate only interrupts. †
■ Generate interrupts and the PPS start and stop time. †
■ Generate only PPS start and stop time. †
To program the Target Time registers (register 455 and 456) to generate only interrupt
event: †
1. Program 00 (for interrupt) in Bits [6:5], TRGTMODSEL, of Register 459 (PPS
455 and 456) for target time interrupt. †
2. Program a target time value in the Target Time registers (register 455 and 456). This
instructs the MAC to generate an interrupt when the target time elapses. If Bits
[6:5], TRGTMODSEL, are changed (for example, to control the PPS), then the
interrupt generation is over-written with the new mode and new programmed
Target Time register value.

Ethernet MAC Address Map and Register Definitions
Ethernet MAC Address Map and Register Definitions

To view the module description and base address, scroll to and click the link for either
of the following module instances:
■ emac0
■ emac1




18. USB 2.0 OTG Controller
November 2012
cv_54018-1.2
cv_54018-1.2
The hard processor system (HPS) provides two instances of a USB On-The-Go (OTG)
controller that supports both device and host functions. The controller supports all
high-speed, full-speed, and low-speed transfers in both device and host modes. The
controller is fully compliant with the On-The-Go and Embedded Host Supplement to the
USB Revision 2.0 Specification. The controller can be programmed for both device and
host functions to support data movement over the USB protocol.
The controllers are operationally independent of each other. Each USB OTG controller
supports a single USB port connected through a USB 2.0 Transceiver Macrocell
Interface Plus (UTMI+) Low Pin Interface (ULPI) compliant PHY. The USB OTG
controllers are instances of the Synopsys® DesignWare® Cores USB 2.0 Hi-Speed
On-The-Go (DWC_otg) controller.
The USB OTG controller is optimized for the following applications and systems: †
■ Portable electronic devices †
■ Point-to-point applications (no hub, direct connection to HS, FS, or LS device) †
■ Multi-point applications (as an embedded USB host) to devices (hub and split
support) †
Each of the two USB OTG ports supports both host and device modes, as described in
the On-The-Go and Embedded Host Supplement to the USB Revision 2.0 Specification. The
USB OTG ports support connections for all types of USB peripherals, including the
following peripherals:
■ Mouse
■ Keyboard
■ Digital cameras
■ Network adapters
■ Hard drives
■ Generic hubs
f Additional information is available in the On-The-Go and Embedded Host Supplement to

the USB Revision 2.0 Specification, which you can download from the USB
Implementers Forum website (www.usb.org).
Registered

November 2012
Subscribe
18–2 Chapter 18: USB 2.0 OTG Controller
Features of the USB OTG Controller

The USB OTG controller has the following USB-specific features:
■ Complies with both Revision 1.3 and Revision 2.0 of the On-The-Go and Embedded
Host Supplement to the USB Revision 2.0 Specification
■ Supports software-configurable modes of operation between OTG 1.3 and
OTG 2.0
■ Supports all USB 2.0 speeds:
■ High speed (HS, 480-Mbps)
■ Full speed (FS, 12-Mbps)
■ Low speed (LS, 1.5-Mbps)
1 In host mode, all speeds are supported. However, in device mode, only
high speed and full speed are supported.
■ Supports all USB transaction types:

■ Control transfers
■ Bulk transfers
■ Isochronous transfers
■ Interrupts
■ Supports automatic ping capability
■ Supports Session Request Protocol (SRP) and Host Negotiation Protocol (HNP)
■ Supports suspend, resume, and remote wake
■ Supports up to 16 host channels
1 In host mode, when the number of device endpoints is greater than the
number of host channels, software can reprogram the channels to support
up to 127 devices, each having 32 endpoints (IN + OUT), for a maximum of
4,064 endpoints.
■ Supports up to 16 bidirectional endpoints, including control endpoint 0
1 Only seven periodic device IN endpoints are supported.
■ Supports a generic root hub

■ Performs transaction scheduling in hardware
On the USB PHY layer, the USB OTG controller supports the following features:
■ A single USB port connected to each OTG instance
■ A ULPI connection to an off-chip USB transceiver

Chapter 18: USB 2.0 OTG Controller 18–3
■ Software-controlled access, supporting vendor-specific or optional PHY registers

access to ease debug
■ The OTG 2.0 support for Attach Detection Protocol (ADP) only through an
external (off-chip) ADP controller
On the integration side, the USB OTG controller supports the following features:
■ Different clocks for system and PHY interfaces
■ Dedicated TX FIFO buffer for each device IN endpoint in direct memory access
(DMA) mode
■ Packet-based, dynamic FIFO memory allocation for endpoints for small FIFO
buffers and flexible, efficient use of RAM that can be dynamically sized by
software
■ Ability to change an endpoint's FIFO memory size during transfers
■ Clock gating support during USB suspend and session-off modes
■ PHY clock gating support
■ System clock gating support
■ Data FIFO RAM clock gating support
1 The USB OTG controller does not support the following protocols:
■ Enhanced Host Controller Interface (EHCI)

■ Open Host Controller Interface (OHCI)
■ Universal Host Controller Interface (UHCI)
Supported PHYs
Table 18–1 lists some PHYs that are compatible with the USB OTG.
Table 18–1. Supported PHYs

Manufacturer Part Number
TI TUSB1210
NXP ISP1504
Cypress CY7C68003
SMSC USB3300

USB OTG Controller Block Diagram and System Integration
USB OTG Controller Block Diagram and System Integration

Figure 18–1 is a block diagram showing one USB OTG controller subsystem in the
HPS. Two subsystems are included in the HPS.
Figure 18–1. USB OTG Controller System Integration
L3 Interconnect
Master Slave
Interface Interface
Bus Control
System
Manager ECC Control
SPRAM USB OTG

Controller
IRQ
Clock
Reset
ULPI PHY
Interface
External USB Transceiver
The USB OTG controller connects to the level 3 (L3) interconnect through a slave
interface, allowing other masters to access the control and status registers (CSRs) in
the controller. The controller also connects to the L3 interconnect through a master
interface, allowing the DMA engine in the controller to move data between external
memory and the controller.
A single-port RAM (SPRAM) connected to the USB OTG controller is used to store
USB data packets for both host and device modes. It is configured as FIFO buffers for
receive and transmit data packets on the USB link.
Through the system manager, the USB OTG controller has control to use and test error
correction codes (ECCs) in the SPRAM. Through the system manager, the USB OTG
controller can also control the behavior of the master interface to the L3 interconnect.
The USB OTG controller connects to the external USB transceiver through a ULPI
PHY interface. This interface also connects through pin multiplexers within the HPS.
The pin multiplexers are controlled by the system manager.

Functional Description of the USB OTG Controller
Additional connections on the USB OTG controller include:

■ Clock input from the clock manager to the USB OTG controller
■ Reset input from the reset manager to the USB OTG controller
■ Interrupt line from the USB OTG controller to the microprocessor unit (MPU)
global interrupt controller (GIC).
USB OTG Controller Block Description

Figure 18–2 is a block diagram of the USB OTG controller. The following sections
provide detail about each of the units that comprise the USB OTG controller.
Figure 18–2. USB OTG Controller Block Diagram
L3 Interconnect
USB OTG
Controller Master Interface Slave Interface
Application Interface Unit
SPRAM Packet FIFO Controller
Media Access Controller
Wakeup and PHY Controller
PHY Interface
ULPI PHY Interface

External USB Transceiver
Master Interface
The master interface includes a built-in DMA controller. The DMA controller moves
data between external memory and the media access controller (MAC).
Properties of the master interface are controlled through the USB L3 Master HPROT
Register (l3master) in the system manager. These bits provide access information to
the L3 interconnect, including whether or not transactions are cacheable, bufferable,
or privileged.
1 Bits in the l3master register can be updated only when the master interface is
guaranteed to be in an inactive state.

Slave Interface
The slave interface allows other masters in the system to access the USB OTG
controller’s CSRs. For testing purposes, other masters can also access the SPRAM.
Slave Interface CSR Unit

The slave interface can read from and write to all the CSRs in the USB OTG
controllers. All register accesses are 32 bits.
The CSR is divided into the following groups of registers:
■ Global
■ Host
■ Device
■ Power and clock gating
Some registers are shared between host and device modes, because the controller can
only be in one mode at a time. The controller generates a mode mismatch interrupt if
a master attempts to access device registers when the controller is in host mode, or
attempts to access host registers when the controller is in device mode. Writing to
unimplemented registers is ignored. Reading from unimplemented registers returns
indeterminate values.
Application Interface Unit

The application interface unit (AIU) generates DMA requests based on programmable
FIFO buffer thresholds. The AIU generates interrupts to the GIC for both host and
device modes. A DMA scheduler is included in the AIU to arbitrate and control the
data transfer between packets in system memory and their respective USB endpoints.
Packet FIFO Controller

The Packet FIFO Controller (PFC) connects the AIU with the MAC through data FIFO
buffers located in the SPRAM. In device mode, one FIFO buffer is implemented for
each IN endpoint. In host mode, a single FIFO buffer stores data for all periodic
(isochronous and interrupt) OUT endpoints, and a single FIFO buffer is used for
nonperiodic (control and bulk) OUT endpoints. Host and device mode share a single
receive data FIFO buffer.
SPRAM
An SPRAM implements the data FIFO buffers for host and device modes. The size of
the FIFO buffers can be programmed dynamically.
The SPRAM supports ECCs. ECCs can be enabled through the system manager, by
setting the RAM ECC Enable (en) bit in the USB0 or USB1 RAM ECC Enable Register
(usb0 or usb1), in the ECC Management Register Group (eccgrp). Single-bit and
double-bit errors in each USB instance can be injected using this register.
The SPRAM provides outputs to notify the system manager when single-bit
correctable errors are detected (and corrected), and when double-bit (uncorrectable)
errors are detected. The system manager generates an interrupt to the GIC when an
ECC error is detected.

MAC
The MAC module implements the following functionality:
■ USB transaction support
■ Host protocol support
■ Device protocol support
■ OTG protocol support
■ Link power management (LPM) functions
USB Transactions
In device mode, the MAC decodes and checks the integrity of all token packets. For
valid OUT or SETUP tokens, the following DATA packet is also checked. If the data
packet is valid, the MAC performs the following steps:
1. Writes the data to the receive FIFO buffer
2. Sends the appropriate handshake when required to the USB host.
If a receive FIFO buffer is not available, the MAC sends a NAK response to the host.
The MAC also supports ping protocol.
For IN tokens, if data is available in the transmit FIFO buffer, the MAC performs the
following steps:
1. Reads the data from the FIFO buffer
2. Forms the data packet
3. Transmits the packet to the host
4. Receives the response from the host
5. Sends the updated status to the PFC
In host mode, the MAC receives a token request from the AIU. The MAC performs the
following steps:
1. Builds the token packet
2. Sends the packet to the device
For OUT or SETUP transactions, the MAC also performs the following steps:
1. Reads the data from the transmit FIFO buffer
2. Assembles the data packet
3. Sends the packet to the device
4. Waits for a response
The response from the device causes the MAC to send a status update to the AIU.
For IN or PING transactions, the MAC waits for the data or handshake response from
the device. For data responses, the MAC performs the following steps:
1. Validates the data
2. Writes the data to the receive FIFO buffer

3. Sends a status update to the AIU

4. Sends a handshake to the device, if appropriate
Host Protocol
In host mode, the MAC performs the following functions:
■ Detects connect, disconnect, and remote wakeup events on the USB link
■ Initiates reset
■ Initiates speed enumeration processes
■ Generates Start of Frame (SOF) packets.
Device Protocol
In device mode, the MAC performs the following functions:
■ Handles USB reset sequence
■ Handles speed enumeration
■ Detects USB suspend and resume activity on the USB link
■ Initiates remote wakeup
■ Decodes SOF packets
OTG Protocol
The MAC handles HNP and SRP for OTG operation. HNP provides a mechanism for
swapping host and device roles. SRP provides mechanisms for the host to turn off
VBUS to save power, and for a device to request a new USB session.
LPM Functions
The USB OTG controller supports LPM in both host and device modes. With this
feature, the USB OTG controller can enter a sleep state when a successful LPM
transaction occurs on the USB link.
Wakeup and Power Control

To reduce power, the USB OTG controller supports a power-down mode. In
power-down mode, the controller and the PHY can shut down their clocks. The
controller supports wakeup on the detection of the following events:
■ Resume
■ Remote wakeup
■ Session request protocol
■ New session start
PHY Interface Unit

The USB OTG controller supports synchronous SDR data transmission to a ULPI
PHY. The SDR mode implements an eight-bit data bus.

ULPI PHY Interface

The ULPI PHY interface is synchronous to the ulpi_clk signal coming from the PHY.
Table 18–2 lists the ULPI PHY interface names and related information.
Table 18–2. ULPI PHY Interfaces

Bit
Port Name Direction Description
Width
ULPI Clock
ulpi_clk 1 Input Receives the 60-MHz clock supplied by the high-speed ULPI PHY. All signals
are synchronous to the positive edge of the clock.
ULPI Data Bus Control
ulpi_dir 1 Input 1—The PHY has data to transfer to the USB OTG controller.
0—The PHY does not have data to transfer.
ULPI Next Data Control
ulpi_nxt 1 Input Indicates that the PHY has accepted the current byte from the USB OTG
controller. When the PHY is transmitting, this signal indicates that a new byte
is available for the controller.
ULPI Stop Data Control
ulpi_stp 1 Output The controller drives this signal high to indicate the end of its data stream. The
controller can also drive this signal high to request data from the PHY.
ulpi_data[7:0] 8 Bidirectional Bidirectional data bus. Driven low by the controller during idle.
Clocks
All clocks must be operational when reset is released. No special handling is required
on the clocks.
Table 18–3 lists the USB OTG controller clock inputs.
Table 18–3. USB OTG Controller Clock Inputs

Clock Signal Frequency Functional Usage
Drives the master and slave interfaces, DMA controller, and internal FIFO
usb_mp_clk 60 – 200 MHz
buffers
usb0_ulpi_clk 60 MHz ULPI reference clock for usb0 from external ULPI PHY I/O pin
usb1_ulpi_clk 60 MHz ULPI reference clock for usb1 from external ULPI PHY I/O pin
Resets
The USB OTG controller can be reset either through the hardware reset input or
through software.
Reset Requirements
There must be a minimum of 12 cycles on the ulpi_clk clock before the controller is
taken out of reset. During reset, the USB OTG controller asserts the ulpi_stp signal.
The PHY outputs a clock when it sees the ulpi_stp signal asserted. However, if the
pin multiplexers are not programmed, the PHY does not see the ulpi_stp signal. As a
result, the ulpi_clk clock signal does not arrive at the USB OTG controller.

Software must ensure that the reset is active for a minimum of two usb_mp_clk cycles.
There is no maximum assertion time.
Hardware Reset
Each of the USB OTG controllers has one reset input from the reset manager. The reset
signal is asserted during a cold or warm reset event. The reset manager holds the
controllers in reset until software releases the resets. Software releases resets by
clearing the appropriate USB bits in the Peripheral Module Reset Register (permodrst)
in the HPS reset manager.
The reset input resets the following blocks:
■ The master and slave interface logic
■ The integrated DMA controller
■ The internal FIFO buffers
■ The CSR
The reset input is synchronized to the usb_mp_clk domain. The reset input is also
synchronized to the ULPI clock within the USB OTG controller and is used to reset the
ULPI PHY domain logic.
Software Reset
Software can reset the controller by setting the Core Soft Reset (csftrst) bit in the
Reset Register (grstctl) in the Global Registers (globgrp) group of the USB OTG
controller.
Software resets are useful in the following situations:
■ A PHY selection bit is changed by software. Resetting the USB OTG controller is
part of clean-up to ensure that the PHY can operate with the new configuration or
clock.
■ During software development and debugging.

Interrupts
Each USB OTG controller has a single interrupt output. Interrupts are asserted on the
conditions shown in Table 18–4.
Table 18–4. USB OTG Interrupt Conditions

Condition Mode
Device-initiated remote wakeup is detected. Host mode
Session request is detected from the device. Host mode
Device disconnect is detected. Host mode
Host LPM entry retry has expired or LPM transaction(s) are complete. Host mode
Host periodic TX FIFO buffer is empty (can be further programmed to indicate
Host mode
half-empty).
Host channels interrupt received. Host mode
Incomplete periodic transfer is pending at the end of the microframe. Host mode
Host port status interrupt received. Host mode
External host initiated resume is detected. Device mode
LPM handshake is sent. Device mode
Reset is detected when in suspend or normal mode. Device mode
USB suspend mode is detected. Device mode
Data fetch is suspended due to TX FIFO buffer full or request queue full. Device mode
At least one isochronous OUT endpoint is pending at the end of the microframe. Device mode
At least one isochronous IN endpoint is pending at the end of the microframe. Device mode
At least one IN or OUT endpoint interrupt is pending at the end of the
Device mode
microframe.
The end of the periodic frame is reached. Device mode
Failure to write an isochronous OUT packet to the RX FIFO buffer. The RX FIFO
buffer does not have enough space to accommodate the maximum packet size Device mode
for the isochronous OUT endpoint.
Enumeration has completed. Device mode
Connector ID change. Common modes
Mode mismatch. Software accesses registers belonging to an incorrect mode. Common modes
Nonperiodic TX FIFO buffer is empty. Common modes
RX FIFO buffer is not empty. Common modes
Start of microframe. Common modes
Device connection debounce is complete in host mode. OTG interrupts
A-Device timeout while waiting for B-Device connection. OTG interrupts
Host negotiation is complete. OTG interrupts
Session request is complete. OTG interrupts
Session end is detected in device mode. OTG interrupts

USB OTG Controller Programming Model

For detailed information about using the USB OTG controller, consult your operating
system (OS) driver documentation. The OS vendor provides application
programming interfaces (APIs) to control USB host, device and OTG operation. This
section provides a brief overview of the following software operations:
■ Enabling SPRAM ECCs
■ Host operation
■ Device operation
Enabling SPRAM ECCs

To avoid false ECC errors, you must initialize the ECC bits in the SPRAM before using
ECCs. To initialize the ECC bits, software writes data to all locations in the SPRAM.
The L3 interconnect has access to the SPRAM is accessible through the USB OTG L3
slave interface. Software accesses the SPRAM through the directfifo memory space,
in the USB OTG controller address space.
The SPRAM contains 8192 (32 KB) locations. The L3 slave provides 32-bit access to the
SPRAM. Physically the SPRAM is implemented as a 35-bit memory, with the highest
three bits reserved for the USB OTG controller’s internal use. When a write is
performed to the SPRAM through the L3 slave interface, bits 32 through 34 of the
internal data bus are tied to 1, to enable the ECC bits to be initialized.
The directfifo memory space is described in the controller address map. Refer to
“USB OTG Controller Address Map and Register Definitions” on page 18–15.
1 Software cannot access the SPRAM beyond the 32-KB range. Out-of-range read
transactions return indeterminate data. Out-of-range write transactions are ignored.
Host Operation
Host Initialization
After power up, the USB port is in its default mode. No VBUS is applied to the USB
cable. The following process sets up the USB OTG controller as a USB host.
1. To enable power to the USB port, the software driver sets the Port Power (prtpwr)
bit to 1 in the Host Port Control and Status Register (hprt) of the Host Mode
Registers (hostgrp) group. This action drives the VBUS signal on the USB link.
The controller waits for a connection to be detected on the USB link.
2. When a USB device connects, an interrupt is generated. The Port Connect Detected
(PrtConnDet) bit in hprt is set to 1.
3. Upon detecting a port connection, the software driver initiates a port reset by
setting the Port Reset (prtrst) bit to 1 in hprt.
4. The software driver must wait a minimum of 10 ms so that speed enumeration can
complete on the USB link.
5. After the 10 ms, the software driver sets prtrst back to 0 to release the port reset.

6. The USB OTG controller generates an interrupt. The Port Enable Disable Change
(prtenchng) and Port Speed (prtspd) bits, in hprt, are set to reflect the enumerated
speed of the device that attached.
At this point the port is enabled for communication. Keep alive or SOF packets are
sent on the port. If a USB 2.0-capable device fails to initialize correctly, it is
reported as a USB 1.1 device.
The Host Frame Interval Register (hfir) is updated with the corresponding PHY
clock settings. The hfir, used for sending SOF packets, is in the Host Mode
Registers (hostgrp) group.
7. The software driver must program the following registers in the Global Registers
(globgrp) group, in the order listed:
a. Receive FIFO Size Register (grxfsiz)—selects the size of the receive FIFO
buffer
b. Non-periodic Transmit FIFO Size Register (gnptxfsiz)—selects the size and
the start address of the non-periodic transmit FIFO buffer for nonperiodic
transactions
c. Host Periodic Transmit FIFO Size Register (hptxfsiz)—selects the size and
start address of the periodic transmit FIFO buffer for periodic transactions
8. System software initializes and enables at least one channel to communicate with
the USB device.
Host Transaction
When configured as a host, the USB OTG controller pipes the USB transactions
through one of two request queues (one for periodic transactions and one for
nonperiodic transactions). Each entry in the request queue holds the SETUP, IN, or
OUT channel number along with other information required to perform a transaction
on the USB link. The sequence in which the requests are written to the queue
determines the sequence of transactions on the USB link.
The host processes the requests in the following order at the beginning of each frame
or microframe:
1. Periodic request queue, including isochronous and interrupt transactions
2. Nonperiodic request queue (bulk or control transfers)
The host schedules transactions for each enabled channel in round-robin fashion.
When the host controller completes the transfer for a channel, the controller updates
the DMA descriptor status in the system memory.
For OUT transactions, the host controller uses two transmit FIFO buffers to hold the
packet payload to be transmitted. One transmit FIFO buffer is used for all nonperiodic
OUT transactions and the other is used for all periodic OUT transactions.
For IN transactions, the USB host controller uses one receive FIFO buffer for all
periodic and nonperiodic transactions. The controller holds the packet payload from
the USB device in the receive FIFO buffer until the packet is transferred to the system
memory. The receive FIFO buffer also holds the status of each packet received. The
status entry holds the IN channel number along with other information, including
received byte count and validity status.

For generic hub operations, the USB OTG controller uses SPLIT transfers to
communicate with slower-speed devices downstream of the hub. For these transfers,
the transaction accumulation or buffering is performed in the generic hub, and is
scheduled accordingly. The USB OTG controller ensures that enough transmit and
receive buffers are allocated when the downstream transactions are completed or
when accumulated data is ready to be sent upstream.
Device Operation
Device Initialization
The following process sets up the USB OTG controller as a USB device:
1. After power up, the USB OTG controller must be set to the desired device speed
by writing to the Device Speed (devspd) bits in the Device Configuration Register
(dcfg) in the Device Mode Registers (devgrp) group. After the device speed is set,
the controller waits for a USB host to detect the USB port as a device port.
2. When an external host detects the USB port, the host performs a port reset, which
generates an interrupt to the USB device software. The USB Reset (usbrst) bit in
the Interrupt (port reset) register in the Global Registers (globgrp) group is set.
The device software then sets up the data FIFO buffer to receive a SETUP packet
from the external host. Endpoint 0 is not enabled yet.
3. After completion of the port reset, the operation speed required by the external
host is known. Software reads the device speed status and sets up all the
remaining required transaction fields to enable control endpoint 0.
After completion of this process, the device is receiving SOF packets, and is ready for
the USB host to set up the device’s control endpoint.
Device Transaction
When configured as a device, the USB OTG controller uses a single FIFO buffer to
receive the data for all the OUT endpoints. The receive FIFO buffer holds the status of
the received data packet, including the byte count, the data packet ID (PID), and the
validity of the received data. The DMA controller reads the data out of the FIFO buffer
as the data are received. If a FIFO buffer overflow condition occurs, the controller
responds to the OUT packet with a NAK, and internally rewinds the pointers.
For IN endpoints, the controller uses dedicated transmit buffers for each endpoint.
The application does not need to predict the order in which the USB host will access
the nonperiodic endpoints. If a FIFO buffer underrun condition occurs during
transmit, the controller inverts the cyclic redundancy code (CRC) to mark the packet
as corrupt on the USB link.
The application handles one data packet at a time per endpoint in transaction-level
operations. The software receives an interrupt on completion of every packet. Based
on the handshake response received on the USB link, the application determines
whether to retry the transaction or proceed with the next transaction, until all packets
in the transfer are completed.
IN Transactions
For an IN transaction, the application performs the following steps:

USB OTG Controller Address Map and Register Definitions
1. Enables the endpoint

2. Triggers the DMA engine to write the associated data packet to the corresponding
transmit FIFO buffer
3. Waits for the packet completion interrupt from the controller
When an IN token is received on an endpoint when the associated transmit FIFO
buffer does not contain sufficient data, the controller performs the following steps:
1. Generates an interrupt
2. Returns a NAK handshake to the USB host
If sufficient data is available, the controller transmits the data to the USB host.
OUT Transactions
For an OUT transaction, the application performs the following steps:
1. Enables the endpoint
2. Waits for the packet received interrupt from the USB OTG controller
3. Retrieves the packet from the receive FIFO buffer
When an OUT token or PING token is received on an endpoint where the receive
FIFO buffer does not have sufficient space, the controller performs the following
steps:
1. Generates an interrupt
2. Returns a NAK handshake to USB host
If sufficient space is available, the controller stores the data in the receive FIFO buffer
and returns an ACK handshake to the USB link.
Control Transfers
For control transfers, the application performs the following steps:
1. Waits for the packet received interrupt from the controller
2. Retrieves the packet from the receive buffer
Because the control transfer is governed by USB protocol, the controller always
responds with an ACK handshake.
USB OTG Controller Address Map and Register Definitions

■ usb0
■ usb1



■ Described interrupt generation.
■ Described software initialization in host and device modes.
November 2012 1.2 ■ Described software operation in host and device modes.
■ Simplified features list.
■ Simplified hardware description.
June 2012 1.1 Added information about ECCs.

19. SPI Controller
November 2012
cv_54019-1.2
cv_54019-1.2
The hard processor system (HPS) provides two serial peripheral interface (SPI)
masters and two SPI slaves. The SPI masters and slaves are instances of the Synopsys®
DesignWare® Synchronous Serial Interface (SSI) controller (DW_apb_ssi).
Features of the SPI Controller

The SPI controller has the following features: †
■ Serial master and serial slave controllers – Enable serial communication with
serial-master or serial-slave peripheral devices. †
■ Serial interface operation – Programmable choice of the following protocols:
■ Motorola SPI protocol
■ Texas Instruments Synchronous Serial Protocol
■ National Semiconductor Microwire
■ DMA controller interface integrated with HPS DMA controller
■ SPI master supports rxd sample delay
■ Transmit and receive FIFO buffers are 256 words deep
■ SPI master supports up to four slave selects
■ Programmable master serial bit rate
■ Programmable data item size of 4 to 16 bits
SPI Block Diagram and System Integration

The SPI supports data bus widths of 32 bits. †
Registered

November 2012
Subscribe
19–2 Chapter 19: SPI Controller
SPI Block Diagram and System Integration
SPI Block Diagram

Figure 19–1 shows the functional groupings of the main interfaces to the SPI block. †
Figure 19–1. SPI Block Diagram
SPI Master (2)

MPU
Shift Control FSM Control Interrupt Controller IRQ IRQ

SPI Master Interface
DMA Peripheral
Clock Request Interface
Pre-Scale Tx & Rx FIFO DMA Interface
DMA
Controller
Register Block
Slave Interface
L4 Peripheral Bus
SPI Slave (2)
Slave Interface
Register Block
Clock DMA Interface

Tx & Rx FIFO DMA Interface
SPI Slave Interface Pre-Scale
Shift Control FSM Control Interrupt Controller
The functional groupings of the main interfaces to the SPI block are as follows: †
■ System bus interface
■ DMA peripheral request interface
■ Interrupt interface
■ SPI interface

Chapter 19: SPI Controller 19–3
Functional Description of the SPI Controller
Protocol Details and Standards Compliance

This chapter describes the functional operation of the SPI controller. †
The host processor accesses data, control, and status information about the SPI
controller through the system bus interface. The SPI also interfaces with the DMA
Controller. †
The HPS includes two general-purpose SPI master controllers and two
general-purpose SPI slave controllers.
The SPI controller can connect to any other SPI device using any of the following
protocols:
■ Motorola SPI Protocol †
■ Texas Instruments Serial Protocol (SSP) †
■ National Semiconductor Microwire Protocol †
SPI Controller Overview

In order for the SPI controller to connect to a serial-master or serial-slave peripheral
device, the peripheral must have a least one of the following interfaces: †
■ Motorola SPI protocol – A four-wire, full-duplex serial protocol from Motorola.
The slave select line is held high when the SPI controller is idle or disabled. For
more information, refer to “Motorola SPI Protocol” on page 19–13. †
■ Texas Instruments Serial Protocol (SSP) – A four-wire, full-duplex serial protocol.
The slave select line used for SPI and Microwire protocols doubles as the frame
indicator for the SSP protocol. For more information, refer to “Texas Instruments
Synchronous Serial Protocol (SSP)” on page 19–14. †
■ National Semiconductor Microwire – A half-duplex serial protocol, which uses a
control word transmitted from the serial master to the target serial slave. For more
information, refer to “National Semiconductor Microwire Protocol” on
page 19–15. You can program the FRF (frame format) bit field in the Control
Register 0 (CTRLR0) to select which protocol is used. †
The serial protocols supported by the SPI controller allow for serial slaves to be
selected or addressed using hardware. Serial slaves are selected under the control of
dedicated hardware select lines. The number of select lines generated from the serial
master is equal to the number of serial slaves present on the bus. The serial-master
device asserts the select line of the target serial slave before data transfer begins. This
architecture is illustrated in part A of Figure 19–2. †
When implemented in software, the input select line for all serial slave devices should
originate from a single slave select output on the serial master. In this mode it is
assumed that the serial master has only a single slave select output. If there are
multiple serial masters in the system, the slave select output from all masters can be
logically ANDed to generate a single slave select input for all serial slave devices. †

The main program in the software domain controls selection of the target slave device;
this architecture is illustrated in part B of Figure 19–2. Software would control which
slave is to respond to the serial transfer request from the master device. †
Figure 19–2. Hardware/Software Slave Selection †
A B
Data Bus Data Bus
Master Slave Master Slave
ss_0 ss ss ss
ss_1
Slave Slave
ss ss
ss = Slave Select Line
Serial Bit-Rate Clocks

SPI Master Bit-Rate Clock
The maximum frequency of the SPI master bit-rate clock (sclk_out) is one-half the
frequency of SPI master clock (spi_m_clk). This allows the shift control logic to
capture data on one clock edge of sclk_out and propagate data on the opposite edge,
as shown in Figure 19–3. The sclk_out line toggles only when an active transfer is in
progress. At all other times it is held in an inactive state, as defined by the serial
protocol under which it operates. †
Figure 19–3. Maximum sclk_out/spi_m_clk Ratio

Capture Drive 1 Capture 1 Drive 2 Capture 2 Drive 3 Capture 3
spi_m_clk
sclk_out
txd/rxd MSB
The frequency of sclk_out can be derived from Equation 19–1, where <SPI clock> is
spi_m_clk for the master SPI modules and l4_main_clk for the slave SPI modules. †
Equation 19–1.
Fsclk_out = F<SPI clock> / SCKDV
SCKDV is a bit field in the register BAUDR, holding any even value in the range 2 to
65,534. If SCKDV is 0, then sclk_out is disabled. †

Equation 19–2 describes the frequency ratio restrictions between the bit-rate clock
sclk_out and the SPI master peripheral clock. The SPI master peripheral clock must
be at least double the offchip master clock. †
Equation 19–2. SPI Master Peripheral Clock

Fspi_m_clk >= 2 x (maximum Fsclk_out) †
SPI Slave Bit-Rate Clock

The minimum frequency of l4_main_clk depends on the operation of the slave
peripheral. If the slave device is receive only, the minimum frequency of l4_main_clk is
six times the maximum expected frequency of the bit-rate clock from the master
device (sclk_in). The sclk_in signal is double synchronized to the l4_main_clk
domain, and then it is edge detected; this synchronization requires three l4_main_clk
periods. †
If the slave device is transmit and receive, the minimum frequency of l4_main_clk is
eight times the maximum expected frequency of the bit-rate clock from the master
device (sclk_in). This ensures that data on the master rxd line is stable before the
master shift control logic captures the data. †
The frequency ratio restrictions between the bit-rate clock sclk_in and the SPI slave
peripheral clock are as follows: †
■ Slave (receive only): Fl4_main_clk >= 6 x (maximum Fsclk_in) †
■ Slave: Fl4_main_clk >= 8 x (maximum Fsclk_in) †
Transmit and Receive FIFO Buffers

There are two 16-bit FIFO buffers, a transmit FIFO buffer and a receive FIFO buffer,
with a depth of 256. Data frames that are less than 16 bits in size must be
right-justified when written into the transmit FIFO buffer. The shift control logic
automatically right-justifies receive data in the receive FIFO buffer. †
Each data entry in the FIFO buffers contains a single data frame. It is impossible to
store multiple data frames in a single FIFO buffer location; for example, you may not
store two 8-bit data frames in a single FIFO buffer location. If an 8-bit data frame is
required, the upper 8-bits of the FIFO buffer entry are ignored or unused when the
serial shifter transmits the data. †
1 The transmit and receive FIFO buffers are cleared when the SPI controller is disabled
(SSIENR=0) or reset. For detailed information about reset signals, refer to the Reset
Manager chapter in volume 3 of the Cyclone® V Device Handbook.
The transmit FIFO buffer is loaded by write commands to the SPI data register (DR).
Data are popped (removed) from the transmit FIFO buffer by the shift control logic
into the transmit shift register. The transmit FIFO buffer generates a transmit FIFO
empty interrupt request when the number of entries in the FIFO buffer is less than or
equal to the FIFO buffer threshold value. The threshold value, set through the register
TXFTLR, determines the level of FIFO buffer entries at which an interrupt is generated.
The threshold value allows you to provide early indication to the processor that the
transmit FIFO buffer is nearly empty. A Transmit FIFO Overflow Interrupt is
generated if you attempt to write data into an already full transmit FIFO buffer. †

Data are popped from the receive FIFO buffer by read commands to the SPI data
register (DR). The receive FIFO buffer is loaded from the receive shift register by the
shift control logic. The receive FIFO buffer generates a receive FIFO full interrupt
request when the number of entries in the FIFO buffer is greater than or equal to the
FIFO buffer threshold value plus one. The threshold value, set through register
RXFTLR, determines the level of FIFO buffer entries at which an interrupt is
generated. †
The threshold value allows you to provide early indication to the processor that the
receive FIFO buffer is nearly full. A Receive FIFO Overrun Interrupt is generated
when the receive shift logic attempts to load data into a completely full receive FIFO
buffer. However, the newly received data are lost. A Receive FIFO Underflow
Interrupt is generated if you attempt to read from an empty receive FIFO buffer. This
alerts the processor that the read data are invalid. †
SPI Interrupts
The SPI controller supports combined interrupt requests, which can be masked. The
combined interrupt request is the ORed result of all other SPI interrupts after
masking. All SPI interrupts have active-high polarity level. The SPI interrupts are
described as follows: †
■ Transmit FIFO Empty Interrupt – Set when the transmit FIFO buffer is equal to or
below its threshold value and requires service to prevent an underrun. The
threshold value, set through a software-programmable register, determines the
level of transmit FIFO buffer entries at which an interrupt is generated. This
interrupt is cleared by hardware when data are written into the transmit FIFO
buffer, bringing it over the threshold level. †
■ Transmit FIFO Overflow Interrupt – Set when a master attempts to write data into
the transmit FIFO buffer after it has been completely filled. When set, new data
writes are discarded. This interrupt remains set until you read the transmit FIFO
overflow interrupt clear register (TXOICR). †
■ Receive FIFO Full Interrupt – Set when the receive FIFO buffer is equal to or above
its threshold value plus 1 and requires service to prevent an overflow. The
threshold value, set through a software-programmable register, determines the
level of receive FIFO buffer entries at which an interrupt is generated. This
interrupt is cleared by hardware when data are read from the receive FIFO buffer,
bringing it below the threshold level. †
■ Receive FIFO Overflow Interrupt – Set when the receive logic attempts to place
data into the receive FIFO buffer after it has been completely filled. When set,
newly received data are discarded. This interrupt remains set until you read the
receive FIFO overflow interrupt clear register (RXOICR). †
■ Receive FIFO Underflow Interrupt – Set when a system bus access attempts to
read from the receive FIFO buffer when it is empty. When set, zeros are read back
from the receive FIFO buffer. This interrupt remains set until you read the receive
FIFO underflow interrupt clear register (RXUICR). †
■ Combined Interrupt Request – ORed result of all the above interrupt requests after
masking. To mask this interrupt signal, you must mask all other SPI interrupt
requests. †

Transmit FIFO Overflow, Transmit FIFO Empty, Receive FIFO Full, Receive FIFO
Underflow, and Receive FIFO Overflow interrupts can all be masked independently,
using the Interrupt Mask Register (IMR). †
Transfer Modes
When transferring data on the serial bus, the SPI controller operates in the modes
discussed in this section.
The transfer mode (TMOD) is set by writing to control register 0 (CTRLR0). †
1 The transfer mode setting does not affect the duplex of the serial transfer. TMOD is
ignored for Microwire transfers, which are controlled by the MWCR register. †
Transmit and Receive

When TMOD = 0, both transmit and receive logic are valid. The data transfer occurs
as normal according to the selected frame format (serial protocol). Transmit data are
popped from the transmit FIFO buffer and sent through the txd line to the target
device, which replies with data on the rxd line. The receive data from the target device
is moved from the receive shift register into the receive FIFO buffer at the end of each
data frame. †
Transmit Only
When TMOD = 1, any receive data are ignored. The data transfer occurs as normal,
according to the selected frame format (serial protocol). Transmit data are popped
from the transmit FIFO buffer and sent through the txd line to the target device,
which replies with data on the rxd line. At the end of the data frame, the receive shift
register does not load its newly received data into the receive FIFO buffer. The data in
the receive shift register is overwritten by the next transfer. You should mask
interrupts originating from the receive logic when this mode is entered. †
Receive Only
When TMOD = 2, the transmit data are invalid. In the case of the SPI slave, the
transmit FIFO buffer is never popped in Receive Only mode. The txd output remains
at a constant logic level during the transmission. The data transfer occurs as normal
according to the selected frame format (serial protocol). The receive data from the
target device is moved from the receive shift register into the receive FIFO buffer at
the end of each data frame. You should mask interrupts originating from the transmit
logic when this mode is entered. †
EEPROM Read
1 This transfer mode is only valid for serial masters. †

When TMOD = 3, the transmit data is used to transmit an opcode and/or an address
to the EEPROM device. Typically this takes three data frames (8-bit opcode followed
by 8-bit upper address and 8-bit lower address). During the transmission of the
opcode and address, no data is captured by the receive logic (as long as the SPI master
is transmitting data on its txd line, data on the rxd line is ignored). The SPI master
continues to transmit data until the transmit FIFO buffer is empty. Therefore, you
should ONLY have enough data frames in the transmit FIFO buffer to supply the
opcode and address to the EEPROM. If more data frames are in the transmit FIFO
buffer than are needed, then read data is lost. †
When the transmit FIFO buffer becomes empty (all control information has been sent),
data on the receive line (rxd) is valid and is stored in the receive FIFO buffer; the txd
output is held at a constant logic level. The serial transfer continues until the number
of data frames received by the SPI master matches the value of the NDF field in the
CTRLR1 register plus one. †
1 EEPROM read mode is not supported when the SPI controller is configured to be in
the SSP mode. †
SPI Master
The SPI master initiates and controls all serial transfers with serial-slave peripheral
devices. Figure 19–1 on page 19–2 shows a SPI master. †
The serial bit-rate clock, generated and controlled by the SPI controller, is driven out
on the sclk_out line. When the SPI controller is disabled, no serial transfers can occur
and sclk_out is held in “inactive” state, as defined by the serial protocol under which
it operates. †
RXD Sample Delay

SPI master device is capable of delaying the default sample time of the rxd signal in
order to increase the maximum achievable frequency on the serial bus.
Round trip routing delays on the sclk_out signal from the master and the rxd signal
from the slave can mean that the timing of the rxd signal, as seen by the master, has
moved away from the normal sampling time.
Without the RXD sample delay, you must increase the baud rate for the transfer in
order to ensure that the setup times on the rxd signal are within range. This reduces
the frequency of the serial interface.
Additional logic is included in the SPI master to delay the default sample time of the
rxd signal. This additional logic can help to increase the maximum achievable
frequency on the serial bus. †
By writing to the rsd field of the RX Sample Delay Register (rx_sample_dly), you
specify an additional amount of delay applied to the rxd sample, in number of
spi_m_clk clock cycles, up to 64 cycles. If the rsd field is programmed with a value
exceeding 64, zero delay is applied to the rxd sample.

Round trip routing delays on the sclk_out signal from the master and the rxd signal
from the slave can mean that the timing of the rxd signal, as seen by the master, has
moved away from the normal sampling time. Figure 19–4 illustrates this situation.
Red arrows indicate routing delay between master and slave devices. Blue arrow
indicates sampling delay within slave from receiving sclk_in to driving txd out. †
Figure 19–4. Effects of Round-Trip Routing Delays on sclk_out Signal

dly=0
dly=4
baud-rate = 4
spi_m_clk
sclk_out
txd_mst MSB LSB
rxd_mst MSB LSB
sclk_in
rxd_slv MSB LSB
txd_slv MSB LSB
Red arrows indicate routing delay between master and slave devices
Blue arrow indicates sampling delay within slave from receiving slk_in to driving txd out
Data Transfers
The SPI master starts data transfers when all the following conditions are met:
■ The SPI master is enabled
■ There is at least one valid entry in the transmit FIFO buffer
■ A slave device is selected
When actively transferring data, the busy flag (BUSY) in the status register (SR) is set.
You must wait until the busy flag is cleared before attempting a new serial transfer. †
1 The BUSY status is not set when the data are written into the transmit FIFO buffer. This
bit gets set only when the target slave has been selected and the transfer is underway.
After writing data into the transmit FIFO buffer, the shift logic does not begin the
serial transfer until a positive edge of the sclk_out signal is present. The delay in
waiting for this positive edge depends on the baud rate of the serial transfer. Before
polling the BUSY status, you should first poll the Transit FIFO Empty (TFE) status
(waiting for 1) or wait for (BAUDR * SPI clock) clock cycles. †
Master SPI and SSP Serial Transfers

“Motorola SPI Protocol” on page 19–13 and “Texas Instruments Synchronous Serial
Protocol (SSP)” on page 19–14 describe the SPI and SSP serial protocols, respectively. †

When the transfer mode is “transmit and receive” or “transmit only” (TMOD = 0 or
TMOD = 1, respectively), transfers are terminated by the shift control logic when the
transmit FIFO buffer is empty. For continuous data transfers, you must ensure that the
transmit FIFO buffer does not become empty before all the data have been
transmitted. The transmit FIFO threshold level (TXFTLR) can be used to early interrupt
(Transmit FIFO Empty Interrupt) the processor indicating that the transmit FIFO
buffer is nearly empty. †
When the DMA is used in conjunction with the SPI master, the transmit data level
(DMATDLR) can be used to early request the DMA Controller, indicating that the
transmit FIFO buffer is nearly empty. The FIFO buffer can then be refilled with data to
continue the serial transfer. The user may also write a block of data (at least two FIFO
buffer entries) into the transmit FIFO buffer before enabling a serial slave. This
ensures that serial transmission does not begin until the number of data frames that
make up the continuous transfer are present in the transmit FIFO buffer. †
When the transfer mode is “receive only” (TMOD = 2), a serial transfer is started by
writing one “dummy” data word into the transmit FIFO buffer when a serial slave is
selected. The txd output from the SPI controller is held at a constant logic level for the
duration of the serial transfer. The transmit FIFO buffer is popped only once at the
beginning and may remain empty for the duration of the serial transfer. The end of the
serial transfer is controlled by the “number of data frames” (NDF) field in control
register 1 (CTRLR1). †
If, for example, you want to receive 24 data frames from a serial-slave peripheral, you
should program the NDF field with the value 23; the receive logic terminates the serial
transfer when the number of frames received is equal to the NDF value plus one. This
transfer mode increases the bandwidth of the system bus as the transmit FIFO buffer
never needs to be serviced during the transfer. The receive FIFO buffer should be read
each time the receive FIFO buffer generates a FIFO full interrupt request to prevent an
overflow. †
When the transfer mode is “eeprom_read” (TMOD = 3), a serial transfer is started by
writing the opcode and/or address into the transmit FIFO buffer when a serial slave
(EEPROM) is selected. The opcode and address are transmitted to the EEPROM
device, after which read data is received from the EEPROM device and stored in the
receive FIFO buffer. The end of the serial transfer is controlled by the NDF field in the
control register 1 (CTRLR1). †
1 EEPROM read mode is not supported when the SPI controller is configured to be in
the SSP mode. †
The receive FIFO threshold level (RXFTLR) can be used to give early indication that the
receive FIFO buffer is nearly full. When a DMA is used, the receive data level
(DMARDLR) can be used to early request the DMA Controller, indicating that the
receive FIFO buffer is nearly full. †
Master Microwire Serial Transfers

“National Semiconductor Microwire Protocol” on page 19–15 describes the Microwire
serial protocol in detail. †

Microwire serial transfers from the SPI serial master are controlled by the Microwire
Control Register (MWCR). The MHS bit field enables and disables the Microwire
handshaking interface. The MDD bit field controls the direction of the data frame (the
control frame is always transmitted by the master and received by the slave). The
MWMOD bit field defines whether the transfer is sequential or nonsequential. †
All Microwire transfers are started by the SPI serial master when there is at least one
control word in the transmit FIFO buffer and a slave is enabled. When the SPI master
transmits the data frame (MDD =1), the transfer is terminated by the shift logic when
the transmit FIFO buffer is empty. When the SPI master receives the data frame (MDD
= 1), the termination of the transfer depends on the setting of the MWMOD bit field. If
the transfer is nonsequential (MWMOD = 0), it is terminated when the transmit FIFO
buffer is empty after shifting in the data frame from the slave. When the transfer is
sequential (MWMOD = 1), it is terminated by the shift logic when the number of data
frames received is equal to the value in the CTRLR1 register plus one. †
When the handshaking interface on the SPI master is enabled (MHS =1), the status of
the target slave is polled after transmission. Only when the slave reports a ready status
does the SPI master complete the transfer and clear its BUSY status. If the transfer is
continuous, the next control/data frame is not sent until the slave device returns a
ready status. †
SPI Slave
The SPI slave handles serial communication with transfer initiated and controlled by
serial master peripheral devices.
Figure 19–5 shows an example of a SPI slave in a single-master bus system, including
the following signals: †
■ sclk_in—serial clock to the SPI slave †
■ ss_in_n—slave select input to the SPI slave †
■ ss_oe_n—output enable for the SPI master or slave †
■ txd—transmit data line for the SPI master or slave †
■ rxd—receive data line for the SPI master or slave †
When the SPI serial slave is selected, it enables its txd data onto the serial bus. All data
transfers to and from the serial slave are regulated on the serial clock line (sclk_in),
driven from the SPI master device. Data are propagated from the serial slave on one
edge of the serial clock line and sampled on the opposite edge. †
When the SPI serial slave is not selected, it must not interfere with data transfers
between the serial-master and other serial-slave devices. When the serial slave is not
selected, its txd output is buffered, resulting in a high impedance drive onto the SPI
master rxd line. The buffers shown in Figure 19–5 are external to SPI controller.
spi_oe_n is the SPI slave output enable signal. †

The serial clock that regulates the data transfer is generated by the serial-master
device and input to the SPI slave on sclk_in. The slave remains in an idle state until
selected by the bus master. When not actively transmitting data, the slave must hold
its txd line in a high impedance state to avoid interference with serial transfers to
other slave devices. The SPI slave output enable (ss_oe_n) signal is available for use to
control the txd output buffer. The slave continues to transfer data to and from the
master device as long as it is selected. If the master transmits to all serial slaves, a
control bit (SLV_OE) in the SPI control register 0 (CTRLR0) can be programmed to
inform the slave if it should respond with data from its txd line. †
Figure 19–5. SPI Slave †
HPS
Master Device SPI Slave
DO rxd
DI txd
ss_oe_n
SCLK sclk_in
SS_O ss_in_n
SS_X
DI
DO
SCLK
SS
Slave
Peripheral n
Slave SPI and SSP Serial Transfers †

“Motorola SPI Protocol” on page 19–13 and “Texas Instruments Synchronous Serial
Protocol (SSP)” on page 19–14 contain a description of the SPI and SSP serial
protocols, respectively. †
If the SPI slave is receive only (TMOD=2), the transmit FIFO buffer need not contain
valid data because the data currently in the transmit shift register is resent each time
the slave device is selected. The TXE error flag in the status register (SR) is not set
when TMOD=2. You should mask the Transmit FIFO Empty Interrupt when this
mode is used. †
If the SPI slave transmits data to the master, you must ensure that data exists in the
transmit FIFO buffer before a transfer is initiated by the serial-master device. If the
master initiates a transfer to the SPI slave when no data exists in the transmit FIFO
buffer, an error flag (TXE) is set in the SPI status register, and the previously
transmitted data frame is resent on txd. For continuous data transfers, you must
ensure that the transmit FIFO buffer does not become empty before all the data have

been transmitted. The transmit FIFO threshold level register (TXFTLR) can be used to
early interrupt (Transmit FIFO Empty Interrupt) the processor, indicating that the
transmit FIFO buffer is nearly empty. When a DMA Controller is used, the DMA
transmit data level register (DMATDLR) can be used to early request the DMA
Controller, indicating that the transmit FIFO buffer is nearly empty. The FIFO buffer
can then be refilled with data to continue the serial transfer. †
The receive FIFO buffer should be read each time the receive FIFO buffer generates a
FIFO full interrupt request to prevent an overflow. The receive FIFO threshold level
register (RXFTLR) can be used to give early indication that the receive FIFO buffer is
nearly full. When a DMA Controller is used, the DMA receive data level register
(DMARDLR) can be used to early request the DMA controller, indicating that the receive
FIFO buffer is nearly full. †
Serial Transfers
“National Semiconductor Microwire Protocol” on page 19–15 describes the Microwire
serial protocol in detail, including timing diagrams and information about how data
are structured in the transmit and receive FIFO buffers before and after a serial
transfer. The Microwire protocol operates in much the same way as the SPI protocol.
There is no decode of the control frame by the SPI slave device. †
Partner Connection Interfaces

The SPI can connect to any serial-master or serial-slave peripheral device using one of
the interfaces discussed in the following sections. †
Motorola SPI Protocol

The inactive state of the serial clock is low. The data frame can be 4 to 16 bits in
length. †
Data transmission begins on the falling edge of the slave select signal. The first data
bit is captured by the master and slave peripherals on the first edge of the serial clock;
therefore, valid data must be present on the txd and rxd lines prior to the first serial
clock edge. †
1 The slave select signal takes effect only when used as slave SPI. For master SPI, the
data transmission begins as soon as the output enable signal is deasserted.
Figure 19–6 shows a timing diagram for a single SPI data transfer. †
The following signals are illustrated in the timing diagrams in this section: †
■ sclk_out—serial clock from SPI master †
■ sclk_in—serial clock from SPI slave †
■ ss_0_n—slave select signal from SPI master †
■ ss_oe_n—output enable for the SPI master or slave †

■ txd—transmit data line for the SPI master or slave †

■ rxd—receive data line for the SPI master or slave †
Figure 19–6. SPI Serial Format
sclk_out/in 0
txd MSB LSB

4 - 16 bits
rxd MSB LSB
ss_0_n/ss_in_n
ss_oe_n
There are four possible transfer modes on the SPI controller for performing SPI serial
transactions; refer to “Transfer Modes” on page 19–7. For transmit and receive transfers
(transfer mode field (9:8) of the Control Register 0 = 0), data transmitted from the SPI
controller to the external serial device is written into the transmit FIFO buffer. Data
received from the external serial device into the SPI controller is pushed into the
receive FIFO buffer. †
For transmit only transfers (transfer mode field (9:8) of the Control Register 0 = 1), data
transmitted from the SPI controller to the external serial device is written into the
transmit FIFO buffer. As the data received from the external serial device is deemed
invalid, it is not stored in the SPI receive FIFO buffer. †
For receive only transfers (transfer mode field (9:8) of the Control Register 0 = 2), data
transmitted from the SPI controller to the external serial device is invalid, so a single
dummy word is written into the transmit FIFO buffer to begin the serial transfer. The
txd output from the SPI controller is held at a constant logic level for the duration of
the serial transfer. Data received from the external serial device into the SPI controller
is pushed into the receive FIFO buffer. †
For eeprom_read transfers (transfer mode field [9:8] of the Control Register 0 = 3),
opcode and/or EEPROM address are written into the transmit FIFO buffer. During
transmission of these control frames, received data is not captured by the SPI master.
After the control frames have been transmitted, receive data from the EEPROM is
stored in the receive FIFO buffer.
Texas Instruments Synchronous Serial Protocol (SSP)

Data transfers begin by asserting the frame indicator line (ss_0_n) for one serial clock
period. Data to be transmitted are driven onto the txd line one serial clock cycle later;
similarly data from the slave are driven onto the rxd line. Data are propagated on the
rising edge of the serial clock (sclk_out/sclk_in) and captured on the falling edge.
The length of the data frame ranges from 4 to 16 bits.
1 The slave select signal (ss_0_n) takes effect only when used as slave SPI. For master
SPI, the data transmission begins as soon as the output enable signal is deasserted.

Figure 19–7 shows the timing diagram for a single SSP serial transfer. †
Figure 19–7. SSP Serial Format
sclk_out/in
txd/rxd MSB LSB
ss_0_n/ss_in_n
ss_oe_n
Continuous data frames are transferred in the same way as single data frames. The
frame indicator is asserted for one clock period during the same cycle as the LSB from
the current transfer, indicating that another data frame follows. Figure 19–8 shows the
timing for a continuous SSP transfer. †
Figure 19–8. SSP Serial Format Continuous Transfer
sclk_out/in
txd/rxd MSB LSB MSB
ss_0_n/ss_in_n
ss_oe_n
National Semiconductor Microwire Protocol

For the master SPI, data transmission begins as soon as the output enable signal is
deasserted. One-half serial clock (sclk_out) period later, the first bit of the control is
sent out on the txd line. The length of the control word can be in the range 1 to 16 bits
and is set by writing bit field CFS (bits 15:12) in CTRLR0. The remainder of the control
word is transmitted (propagated on the falling edge of sclk_out) by the SPI serial
master. During this transmission, no data are present (high impedance) on the serial
master's rxd line. †
The direction of the data word is controlled by the MDD bit field (bit 1) in the
Microwire Control Register (MWCR). When MDD=0, this indicates that the SPI serial
master receives data from the external serial slave. One clock cycle after the LSB of the
control word is transmitted, the slave peripheral responds with a dummy 0 bit,
followed by the data frame, which can be 4 to 16 bits in length. Data are propagated
on the falling edge of the serial clock and captured on the rising edge. †
Continuous transfers from the Microwire protocol can be sequential or nonsequential,
and are controlled by the MWMOD bit field (bit 0) in the MWCR. †
Nonsequential continuous transfers occur, with the control word for the next transfer
following immediately after the LSB of the current data word. †
The only modification needed to perform a continuous nonsequential transfer is to
write more control words into the transmit FIFO buffer. †

During sequential continuous transfers, only one control word is transmitted from the
SPI master. The transfer is started in the same manner as with nonsequential read
operations, but the cycle is continued to read further data. The slave device
automatically increments its address pointer to the next location and continues to
provide data from that location. Any number of locations can be read in this manner;
the SPI master terminates the transfer when the number of words received is equal to
the value in the CTRLR1 register plus one. †
When MDD = 1, this indicates that the SPI serial master transmits data to the external
serial slave. Immediately after the LSB of the control word is transmitted, the SPI
master begins transmitting the data frame to the slave peripheral. †
1 The SPI controller does not support continuous sequential Microwire writes, where
MDD = 1 and MWMOD = 1. †
Continuous transfers occur with the control word for the next transfer following
immediately after the LSB of the current data word.
The Microwire handshaking interface can also be enabled for SPI master write
operations to external serial-slave devices. To enable the handshaking interface, you
must write 1 into the MHS bit field (bit 2) on the MWCR register. When MHS is set to
1, the SPI serial master checks for a ready status from the slave device before
completing the transfer, or transmitting the next control word for continuous
transfers. †
After the first data word has been transmitted to the serial-slave device, the SPI
master polls the rxd input waiting for a ready status from the slave device. Upon
reception of the ready status, the SPI master begins transmission of the next control
word. After transmission of the last data frame has completed, the SPI master
transmits a start bit to clear the ready status of the slave device before completing the
transfer. †
In the SPI slave, data transmission begins with the falling edge of the slave select
signal (ss_in_0). One-half serial clock (sclk_in) period later, the first bit of the control
is present on the rxd line. The length of the control word can be in the range of 1 to 16
bits and is set by writing bit field CFS in the CTRLR0 register. The CFS bit field must be
set to the size of the expected control word from the serial master. The remainder of
the control word is received (captured on the rising edge of sclk_in) by the SPI serial
slave. During this reception, no data are driven (high impedance) on the serial slave's
txd line. †
The direction of the data word is controlled by the MDD bit field (bit 1) MWCR
register. When MDD=0, this indicates that the SPI serial slave is to receive data from
the external serial master. Immediately after the control word is transmitted, the serial
master begins to drive the data frame onto the SPI slave rxd line. Data are propagated
on the falling edge of the serial clock and captured on the rising edge. The slave-select
signal is held active-low during the transfer and is deasserted one-half clock cycle
later after the data are transferred. The SPI slave output enable signal is held inactive
for the duration of the transfer. †
When MDD=1, this indicates that the SPI serial slave transmits data to the external
serial master. Immediately after the LSB of the control word is transmitted, the SPI
slave transmits a dummy 0 bit, followed by the 4- to 16-bit data frame on the txd
line. †

Continuous transfers for a SPI slave occur in the same way as those specified for the
SPI master. The SPI slave does not support the handshaking interface, as there is
never a busy period. †
Figure 19–9 shows the timing diagram for a single SPI serial master read from an
external serial slave.
Figure 19–9. Single SPI Serial Master Microwire Serial Transfer (MDD=0)
sclk_out
Control Word
txd MSB LSB

4 - 16 Bits
rxd MSB LSB
ssi_oe_n
Figure 19–10 shows the timing diagram for a single SPI serial slave write to an
external serial master.
Figure 19–10. Single SPI Slave Microwire Serial Transfer (MDD=1)
sclk_out
Data Word
txd 0 MSB LSB

Control Word
rxd MSB LSB
ss_in_0
ssi_oe_n
DMA Controller Interface

The SPI controller supports DMA signaling to indicate when the receive FIFO buffer
has data ready to be read or when the transmit FIFO buffer needs data. It requires two
DMA channels, one for transmit data and one for receive data. The SPI controller can
issue single or burst DMA transfers and accepts burst acknowledges from the DMA.
System software can trigger the DMA burst mode by programming an appropriate
value into the threshold registers. The typical setting of the threshold register value is
half full.
To enable the DMA Controller interface on the SPI controller, you must write the
DMA Control Register (DMACR). Writing a 1 into the TDMAE bit field of DMACR
register enables the SPI transmit handshaking interface. Writing a 1 into the RDMAE
bit field of the DMACR register enables the SPI receive handshaking interface. †

SPI Programming Model
Slave Interface
The host processor accesses data, control, and status information about the SPI
controller through the slave interface. The SPI supports a data bus width of 32 bits.
Accesses to the SPI peripheral are described in the following subsections. †
Control and Status Register Access

Control and status registers within the SPI controller are byte-addressable. The
maximum width of the control or status register in the SPI controller is 16 bits.
Therefore, all read and write operations to the SPI control and status registers require
only one access. †
Data Register Access

The data register (DR) within the SPI controller is 16 bits wide in order to remain
consistent with the maximum serial transfer size (data frame). A write operation to DR
moves data from the slave write data bus into the transmit FIFO buffer. An read
operation from DR moves data from the receive FIFO buffer onto the slave readback
data bus. †
1 The DR register in the SPI controller occupies sixty-four 32-bit locations of the memory
map to facilitate burst transfers. There are no burst transactions on the system bus
itself, but SPI supports bursts on the system interconnect. Writing to any of these
address locations has the same effect as pushing the data from the slave write data bus
into the transmit FIFO buffer. Reading from any of these locations has the same effect
as popping data from the receive FIFO buffer onto the slave readback data bus. The
FIFO buffers on the SPI controller are not addressable.
Clocks and Resets

The SPI controller uses the clock and reset signals shown in Table 19–1.
Table 19–1. SPI Controller Clocks and Resets

Master Slave
SPI clock spi_m_clk l4_main_clk
SPI bit-rate clock sclk_out sclk_in
Reset spim_rst_n spis_rst_n

This section describes the programming model for the SPI controller based on the
following master and slave transfers:
■ Master SPI and SSP Serial Transfers
■ Master Microwire Serial Transfers
■ Slave SPI and SSP Serial Transfers
■ Slave Microwire Serial Transfers
■ Software Control for Slave Selection

Master SPI and SSP Serial Transfers

Figure 19–11 shows the software flow for a master SPI or SSP serial transfer.
Figure 19–11. Master SPI or SSP Serial Transfer Software Flow
Idle
Disable SPI
Configure Master by Writing

CTRLR0, CTRLR1, BAUDR,
TXFTLR, RXFTLR, IMR & SER
Enable SPI
The FIFO can be filled at this point.

Write Data to Transfer begins when the first data
Tx FIFO word is present in the transmit FIFO
and a slave is enabled.
Transfer
in Progress
yes Interrupt Service

Interrupt?
Routine
If the transmit FIFO makes the request

no and all data has not been sent, write
data to the transmit FIFO.
If the receive FIFO makes the request,

read data from the receive FIFO.
Busy?
yes
TMOD = 01
no
Read Rx
FIFO
To complete an SPI or SSP serial transfer from the SPI master, follow these steps:
1. If the SPI master is enabled, disable it by writing 0 to the SSI Enable register
(SSIENR).

2. Set up the SPI master control registers for the transfer; you can set these transfers
in any order.
■ Write Control Register 0 (CTRLLR0). For SPI transfers, you must set the serial
clock polarity and serial clock phase parameters identical to the target slave
device.
■ If the transfer mode is receive only, write Control Register 1 (CTRLR1) with the
number of frames in the transfer minus 1. For example, if you want to receive
four data frames, write this register with 3.
■ Write the Baud Rate Select Register (BAUDR) to set the baud rate for the transfer.
■ Write the Transmit and Receive FIFO Threshold Level registers (TXFTLR and
RXFTLR) to set FIFO buffer threshold levels.
■ Write the IMR register to set up interrupt masks.
■ Write the Slave Enable Register (SER) register here to enable the target slave for
selection. If a slave is enabled here, the transfer begins as soon as one valid data
entry is present in the transmit FIFO buffer. If no slaves are enabled prior to
writing to the Data Register (DR), the transfer does not begin until a slave is
enabled.
3. Enable the SPI master by writing 1 to the SSIENR register.
4. Write data for transmission to the target slave into the transmit FIFO buffer (write
DR). If no slaves were enabled in the SER register at this point, enable it now to
begin the transfer.
5. Poll the BUSY status to wait for the transfer to complete. If a transmit FIFO empty
interrupt request is made, write the transmit FIFO buffer (write DR). If a receive
FIFO full interrupt request is made, read the receive FIFO buffer (read DR).
6. The shift control logic stops the transfer when the transmit FIFO buffer is empty. If
the transfer mode is receive only (TMOD = 2’b10), the shift control logic stops the
transfer when the specified number of frames have been received. When the
transfer is done, the BUSY status is reset to 0.
7. If the transfer mode is not transmit only (TMOD != 01), read the receive FIFO
buffer until it is empty.
8. Disable the SPI master by writing 0 to SSIENR.

Master Microwire Serial Transfers

Figure 19–12 shows the software flow for a Microwire serial transfer.
Figure 19–12. Microwire Serial Transfer Software Flow
Idle
Disable SPI
Configure Master by Writing

CTRLR0, CTRLR1, BAUDR,
TXFTLR, RXFTLR, MWCR,
IMR & SER
Enable SPI
If the master receives data, the user only

Write Control & needs to write control frames into the TX
Data to Tx FIFO FIFO. Transfer begins when the first control
word is present in the transmit FIFO and a
slave is enabled.
Transfer
in Progress

Interrupt?
Routine


Busy?
yes
MWCR[1] = 1
no
Read Rx
FIFO
To complete a Microwire serial transfer from the SPI master, follow these steps:
1. If the SPI master is enabled, disable it by writing 0 to SSIENR.

2. Set up the SPI control registers for the transfer. You can set these registers in any
order.
■ Write CTRLR0 to set transfer parameters. If the transfer is sequential and the SPI
master receives data, write CTRLR1 with the number of frames in the transfer
minus 1. For example, if you want to receive four data frames, write this
register with 3.
■ Write BAUDR to set the baud rate for the transfer.
■ Write TXFTLR and RXFTLR to set FIFO buffer threshold levels.
You can write the SER register to enable the target slave for selection. If a slave is
enabled here, the transfer begins as soon as one valid data entry is present in the
transmit FIFO buffer. If no slaves are enabled prior to writing to the DR register, the
transfer does not begin until a slave is enabled.
3. Enable the SPI master by writing 1 to the SSIENR register.
4. If the SPI master transmits data, write the control and data words into the transmit
FIFO buffer (write DR). If the SPI master receives data, write the control word or
words into the transmit FIFO buffer. If no slaves were enabled in the SER register
at this point, enable now to begin the transfer.
5. Poll the BUSY status to wait for the transfer to complete. If a transmit FIFO empty
interrupt request is made, write the transmit FIFO buffer (write DR). If a receive
FIFO full interrupt request is made, read the receive FIFO buffer (read DR).
6. The shift control logic stops the transfer when the transmit FIFO buffer is empty. If
the transfer mode is sequential and the SPI master receives data, the shift control
logic stops the transfer when the specified number of data frames is received.
When the transfer is done, the BUSY status is reset to 0.
7. If the SPI master receives data, read the receive FIFO buffer until it is empty.
8. Disable the SPI master by writing 0 to SSIENR.

Slave SPI and SSP Serial Transfers

Figure 19–13 shows the software flow for a slave SPI or SSP serial transfer.
Figure 19–13. Slave SPI or SSP Serial Transfer Software Flow
Idle
Disable SPI
Configure Slave by Writing

CTRLR0, CTRLR1, TXFTLR,
RXFTLR, MWCR, & IMR
Enable SPI
Write Data
to Tx FIFO
TMOD = 10
Wait for Master

to Select Slave
Transfer
in Progress

Interrupt?
Routine


Busy?
yes
TMOD = 01
no
Read Rx
FIFO
To complete a continuous serial transfer from a serial master to the SPI slave, follow
these steps:
1. If the SPI slave is enabled, disable it by writing 0 to SSIENR.

2. Set up the SPI control registers for the transfer. You can set these registers in any
order.
■ Write CTRLR0 (for SPI transfers, set SCPH and SCPOL identical to the master
device.
■ Write TXFTLR and RXFTLR to set FIFO buffer threshold levels.
3. Enable the SPI slave by writing 1 to the SSIENR register.
4. If the transfer mode is transmit and receive (TMOD=2'b00) or transmit only
(TMOD=2'b01), write data for transmission to the master into the transmit FIFO
buffer (write DR). If the transfer mode is receive only (TMOD=2'b10), you need not
write data into the transmit FIFO buffer. The current value in the transmit shift
register is retransmitted.
5. The SPI slave is now ready for the serial transfer. The transfer begins when a
serial-master device selects the SPI slave.
6. When the transfer is underway, the BUSY status can be polled to return the transfer
status. If a transmit FIFO empty interrupt request is made, write the transmit FIFO
buffer (write DR). If a receive FIFO full interrupt request is made, read the receive
FIFO buffer (read DR).
7. The transfer ends when the serial master removes the select input to the SPI slave.
When the transfer is completed, the BUSY status is reset to 0.
8. If the transfer mode is not transmit only (TMOD != 01), read the receive FIFO
buffer until empty.
9. Disable the SPI slave by writing 0 to SSIENR.
Slave Microwire Serial Transfers

For SPI slave, the Microwire protocol operates in much the same way as the SPI
protocol. There is no decode of the control frame by the SPI slave.
Software Control for Slave Selection

When using software to select slave devices, the input select lines from serial slave
devices is connected to a single slave select output on the SPI master. The following
examples show the software flow for slave selection:
■ For SPI master:
1. If the SPI master is enabled, disable it by writing 0 to SSIENR.
2. Write CTRLR0 to match the required transfer.
3. If the transfer is receive only, write the number of frames into CTRLR1.
4. Write BAUDR to set the transfer baud rate.
5. Write TXFTLR and RXFTLR to set FIFO buffer threshold levels.
6. Write IMR register to set interrupt masks.
7. Write SER register bit[0] to logic '1' to select slave 1 in this example.

8. Write SSIENR register bit[0] to logic '1' to enable SPI master.

■ For SPI slave:
9. If the SPI slave is enabled, disable it by writing 0 to SSIENR.
10. Write CTRLR0 to match the required transfer.
11. Write TXFTLR and RXFTLR to set FIFO buffer threshold levels.
12. Write IMR register to set interrupt masks.
13. Write SSIENR register bit[0] to logic '1' to enable SPI slave.
14. If the SPI slave transmits data, write data into TX FIFO buffer. Note all other SPI
slaves are disabled (SSIENR = 0) and therefore will not respond to an active level on
their ss_in_n port.
The FIFO buffer depth (FIFO_DEPTH) for both the RX and TX buffers in the SPI
controller is 256 entries.
DMA Controller Operation

To enable the DMA controller interface on the SPI controller, you must write the DMA
Control Register (DMACR). Writing a 1 to the TDMAE bit field of DMACR register enables
the SPI controller transmit handshaking interface. Writing a 1 to the RDMAE bit field
of the DMACR register enables the SPI controller receive handshaking.†
f For details about the DMA controller, refer to the DMA Controller chapter in volume 3
DMA Operation
For more information about the DMA operation, refer to the ARM DMA chapter.
Transmit FIFO Buffer Underflow

During SPI serial transfers, transmit FIFO buffer requests are made to the DMA
Controller whenever the number of entries in the transmit FIFO buffer is less or equal
to the value in DMA Transmit Data Level Register (DMATDLR); also known as the
watermark level. The DMA Controller responds by writing a burst of data to the
transmit FIFO buffer, of length specified as DMA burst length.†
f For details about the DMA burst length microcode setup, refer to the DMA Controller
Data should be fetched from the DMA often enough for the transmit FIFO buffer to
perform serial transfers continuously, that is, when the FIFO buffer begins to empty,
another DMA request should be triggered. Otherwise, the FIFO buffer will run out of
data (underflow). To prevent this condition, you must set the watermark level
correctly.†
Transmit Watermark Level

Consider the example where the assumption is made: †
DMA burst length = FIFO_DEPTH - DMATDLR

Here the number of data items to be transferred in a DMA burst is equal to the empty
space in the transmit FIFO buffer. Consider the following two different watermark
level settings. †
■ Case 1: DMATDLR = 64: †
■ Transmit FIFO watermark level = DMATDLR = 64: †
■ DMA burst length = FIFO_DEPTH - DMATDLR = 192: †
■ SPI transmit FIFO_DEPTH = 256: †
■ Block transaction size = 960: †
Figure 19–14 shows the transmit FIFO buffer when the watermark level equals 64.
Figure 19–14. Transmit FIFO Watermark Level = 64
Transmit
FIFO Buffer
FIFO_DEPTH = 256 FIFO_DEPTH - DMATDLR = 192
Empty Data In DMA

Transmit FIFO
Controller
DMATDLR = 64
Watermark Level
Data Out Full
The number of burst transactions needed equals the block size divided by the
number of data items per burst:
Block transaction size/DMA burst length = 960/192 = 5
The number of burst transactions in the DMA block transfer is 5. But the
watermark level, DMATDLR, is quite low. Therefore, the probability of transmit
underflow is high where the SPI serial transmit line needs to transmit data, but
there is no data left in the transmit FIFO buffer. This occurs because the DMA has
not had time to service the DMA request before the FIFO buffer becomes empty.
■ Case 2: DMATDLR = 192†
■ Transmit FIFO watermark level = DMATDLR = 192 †
■ DMA burst length = FIFO_DEPTH - DMATDLR = 64 †
■ SPI transmit FIFO_DEPTH = 256 †
■ Block transaction size = 960 †

Figure 19–15 shows the transmit FIFO buffer when the watermark level equals 192.
Transmit
FIFO Buffer
FIFO_DEPTH = 256 Empty

Transmit FIFO
FIFO_DEPTH - DMATDLR = 64
Watermark Level
Data In DMA
Full Controller
DMATDLR = 192
Data Out
Number of burst transactions in block: †

Block transaction size/DMA burst length = 960/64 = 15 †
In this block transfer, there are 15 destination burst transactions in a DMA block
transfer. But the watermark level, DMATDLR, is high. Therefore, the probability of
SPI transmit underflow is low because the DMA controller has plenty of time to
service the destination burst transaction request before the SPI transmit FIFO
buffer becomes empty. †
Thus, the second case has a lower probability of underflow at the expense of more
burst transactions per block. This provides a potentially greater amount of bursts
per block and worse bus utilization than the former case. †
Therefore, the goal in choosing a watermark level is to minimize the number of
transactions per block, while at the same time keeping the probability of an underflow
condition to an acceptable level. In practice, this is a function of the ratio of the rate at
which the SPI transmits data to the rate at which the DMA can respond to destination
burst requests. †
Transmit FIFO Buffer Overflow

Setting the DMA transaction burst length to a value greater than the watermark level
that triggers the DMA request may cause overflow when there is not enough space in
the transmit FIFO buffer to service the destination burst request. Therefore, the
following equation must be adhered to in order to avoid overflow: †
DMA burst length <= FIFO_DEPTH - DMATDLR
In case 2: DMATDLR = 192, the amount of space in the transmit FIFO buffer at the time of
the burst request is made is equal to the DMA burst length. Thus, the transmit FIFO
buffer may be full, but not overflowed, at the completion of the burst transaction. †
Therefore, for optimal operation, DMA burst length should be set at the FIFO buffer
level that triggers a transmit DMA request; that is: †
DMA burst length = FIFO_DEPTH - DMATDLR
Adhering to this equation reduces the number of DMA bursts needed for block
transfer, and this in turn improves bus utilization. †

1 The transmit FIFO buffer will not be full at the end of a DMA burst transfer if the SPI
controller has successfully transmitted one data item or more on the serial transmit
line during the transfer. †
Receive FIFO Buffer Overflow

During SPI serial transfers, receive FIFO buffer requests are made to the DMA
whenever the number of entries in the receive FIFO buffer is at or above the DMA
Receive Data Level Register, that is DMATDLR + 1.This is known as the watermark level.
The DMA responds by fetching a burst of data from the receive FIFO buffer. †
Data should be fetched by the DMA often enough for the receive FIFO buffer to accept
serial transfers continuously, that is, when the FIFO buffer begins to fill, another DMA
transfer is requested. Otherwise the FIFO buffer will fill with data (overflow). To
prevent this condition, the user must set the watermark level correctly. †
Choosing Receive Watermark Level

Similar to choosing the transmit watermark level, the receive watermark level,
DMATDLR + 1, should be set to minimize the probability of overflow, as shown in Figure
8. It is a trade off between the number of DMA burst transactions required per block
versus the probability of an overflow occurring. †
Receive FIFO Buffer Underflow

Setting the source transaction burst length greater than the watermark level can cause
underflow where there is not enough data to service the source burst request.
Therefore, the following equation must be adhered to avoid underflow: †
DMA burst length = DMATDLR + 1
If the number of data items in the receive FIFO buffer is equal to the source burst
length at the time of the burst request is made, the receive FIFO buffer may be
emptied, but not underflowed, at the completion of the burst transaction. For optimal
operation, DMA burst length should be set at the watermark level, DMATDLR + 1. †
Adhering to this equation reduces the number of DMA bursts in a block transfer,
which in turn can improve bus utilization. †
1 The receive FIFO buffer will not be empty at the end of the source burst transaction if
the SPI controller has successfully received one data item or more on the serial receive
line during the burst. †

SPI Controller Address Map and Register Definitions
Figure 19–16 shows the receive FIFO buffer.
Figure 19–16. Receive FIFO Buffer
Receive
FIFO Buffer
Empty
Data Out
Transmit FIFO DMA
Watermark Level Controller
DMARDLR + 1
Full
Data In
SPI Controller Address Map and Register Definitions

f The address map and register definitions reside in hps.html file that accompanies this
handbook volume. Click the link to open the file.
■ spis0
■ spis1
■ spim0
■ spim1


Added programming model, address map and register definitions, clocks, and reset
May 2012 1.1
sections.


20. I2C Controller
November 2012
cv_54020-1.2
cv_54020-1.2
The I2C controller provides support for a communication link between integrated
circuits on a board. It is a simple two-wire bus which consists of a serial data line
(SDA) and a serial clock (SCL) for use in applications such as temperature sensors and
voltage level translators to EEPROMs, A/D and D/A converters, CODECs, and many
types of microprocessors. †
The hard processor system (HPS) provides four I2C controllers to enable system
software to communicate serially with I2C buses. Each I2C controller can operate in
master or slave mode, and support standard mode of up to 100 kilobits per
second (Kbps) or fast mode of up to 400 Kbps. These I2C controllers are instances of
the Synopsys® DesignWare® APB I2C (DW_apb_i2c) controller.
1 Each I2C controller must be programmed to operate in either master or slave mode
only. Operating as a master and slave simultaneously is not supported. †
Features of the I2C Controller

The I2C controller has the following features:
■ Maximum clock speed of up to 400 Kbps
■ One of the following I2C operations:
■ A master in an I2C system and programmed only as a master †
■ A slave in an I2C system and programmed only as a slave †
■ 7- or 10-bit addressing †
■ Mixed read and write combined-format transactions in both 7-bit and 10-bit
addressing mode †
■ Bulk transmit mode †
■ Transmit and receive buffers †
■ Handles bit and byte waiting at all bus speeds †
■ DMA handshaking interface †
Registered

November 2012
Subscribe
20–2 Chapter 20: I2C Controller
I2C Controller Block Diagram and System Integration
I2C Controller Block Diagram and System Integration

The I2C controller consists of a slave interface, an I2C interface, and FIFO logic to
buffer data between the two interfaces. †
The host processor accesses data, control, and status information about the I2C
controller through a 32-bit slave interface. Figure 20–1 shows the I2C controller block
diagram.
Figure 20–1. I2C Controller Block Diagram
I2C Controller MPU
Tx & Rx FIFO Interrupt Controller IRQ

I2 CInterface
(to I/O Pins)
Tx & Rx Shift Control
Rx Filter DMA Interface DMA

Controller
Clock
Manager
Register Block
Reset
Manager
Slave Interface
L4 Peripheral Bus
The I2C controller consists of the following modules and interfaces:

■ Slave interface for control and status register (CSR) accesses and DMA transfers,
allowing a master to access the CSRs and the DMA to read or write data directly.
■ Two FIFO buffers for transmit and receive data, which hold the Rx FIFO and Tx
FIFO buffer register banks and controllers, along with their status levels. †
■ Shift logic for parallel-to-serial and serial-to-parallel conversion
■ Rx shift – Receives data into the design and extracts it in byte format. †
■ Tx shift – Presents data supplied by CPU for transfer on the I2C bus. †
■ Control logic responsible for implementing the I2C protocol.
■ DMA interface that generates handshaking signals to the DMA controller in order
to automate the data transfer without CPU intervention. †

Chapter 20: I2C Controller 20–3
Functional Description of the I2C Controller
■ Interrupt controller that generates raw interrupt and interrupt flags, allowing
them to be set and cleared. †
■ Receive filter for detecting events, such as start and stop conditions, in the bus; for
example, start, stop and arbitration lost. †

This section describes the functional operation of the I2C controller.
Feature Usage
The I2C controller can operate in standard mode (with data rates 0 to 100 Kbps) or fast
mode (with data rates less than or equal to 400 Kbps). Additionally, fast mode devices
are downward compatible. For instance, fast mode devices can communicate with
standard mode devices in 0 to 100 Kbps I2C bus system. However, standard mode
devices are not upward compatible and should not be incorporated in a fast-mode I2C
bus system as they cannot follow the higher transfer rate and therefore unpredictable
states would occur. †
You can attach any I2C controller to an I2C-bus and every device can talk with any
master, passing information back and forth. There needs to be at least one master
(such as a microcontroller or DSP) on the bus and there can be multiple masters,
which require them to arbitrate for ownership. Multiple masters and arbitration are
explained later in this chapter. †

Behavior
You can control the I2C controller via software to be either mode:
■ An I2C master only, communicating with other I2C slaves; OR
■ An I2C slave only, communicating with one or more I2C masters.
The master is responsible for generating the clock and controlling the transfer of data.
The slave is responsible for either transmitting or receiving data to/from the master.
The acknowledgement of data is sent by the device that is receiving data, which can
be either a master or a slave. As mentioned previously, the I2C protocol also allows
multiple masters to reside on the I2C bus and uses an arbitration procedure to
determine bus ownership. †
Each slave has a unique address that is determined by the system designer. When a
master wants to communicate with a slave, the master transmits a START/RESTART
condition that is then followed by the slave's address and a control bit (R/W) to
determine if the master wants to transmit data or receive data from the slave. The
slave then sends an acknowledge (ACK) pulse after the address. †
If the master (master-transmitter) is writing to the slave (slave-receiver), the receiver
receives one byte of data. This transaction continues until the master terminates the
transmission with a STOP condition. If the master is reading from a slave
(master-receiver), the slave transmits (slave-transmitter) a byte of data to the master,
and the master then acknowledges the transaction with an ACK pulse. This
transaction continues until the master terminates the transmission by not
acknowledging (NACK) the transaction after the last byte is received, and then the
master issues a STOP condition or addresses another slave after issuing a RESTART
condition. †
Figure 20–2 illustrates the data transfer behavior on the I2C bus.
Figure 20–2. Data transfer on the I2C Bus †

ACK
from Receiver P or R
ACK
SDA MSB LSB from Slave
SCL S or R 1 2 7 8 9 1 2 3-8 9 R or P
Start or Restart Byte Complete SCL Held Low while Stop & Restart
Condition Interrupt within Servicing Interrupts Condition
Slave
The I2C controller is a synchronous serial interface. The SDA line is a bidirectional
signal and changes only while the SCL line is low, except for STOP, START, and
RESTART conditions. The output drivers are open-drain or open-collector to perform
wire-AND functions on the bus. The maximum number of devices on the bus is
limited by only the maximum capacitance specification of 400 pF. Data is transmitted
in byte packages. †

START and STOP Generation

When operating as a master, putting data into the transmit FIFO causes the I2C
controller to generate a START condition on the I2C bus. Allowing the transmit FIFO
to empty causes the I2C controller to generate a STOP condition on the I2C bus. †
When operating as a slave, the I2C controller does not generate START and STOP
conditions, as per the protocol. However, if a read request is made to the I2C
controller, it holds the SCL line low until read data has been supplied to it. This stalls
the I2C bus until read data is provided to the slave I2C controller, or the I2C controller
slave is disabled by writing a 0 to IC_ENABLE register. †
Combined Formats
The I2C controller supports mixed read and write combined format transactions in
both 7-bit and 10-bit addressing modes. †
The I2C controller does not support mixed address and mixed address format—that
is, a 7-bit address transaction followed by a 10-bit address transaction or vice versa—
combined format transactions. †
To initiate combined format transfers, the IC_RESTART_EN bit in the IC_CON register
should be set to 1. With this value set and operating as a master, when the I2C
controller completes an I2C transfer, it checks the transmit FIFO and executes the next
transfer. If the direction of this transfer differs from the previous transfer, the
combined format is used to issue the transfer. If the transmit FIFO is empty when the
current I2C transfer completes, a STOP is issued and the next transfer is issued
following a START condition. †
Protocol Details
This section describes the I2C controller protocols.
START and STOP Conditions

When the bus is idle, both the SCL and SDA signals are pulled high through pull-up
resistors on the bus. When the master wants to start a transmission on the bus, the
master issues a START condition. This is defined to be a high-to-low transition of the
SDA signal while SCL is 1. When the master wants to terminate the transmission, the
master issues a STOP condition. This is defined to be a low-to-high transition of the
SDA line while SCL is 1. †
Figure 20–3 shows the timing of the START and STOP conditions. When data is being
transmitted on the bus, the SDA line must be stable when SCL is 1. †
Figure 20–3. Timing Diagram for START and STOP Condition †
SDA
SCL S P
Start Data Change Data Line Stable Data Change Stop

Condition Allowed Data Valid Allowed Condition

1 The signal transitions for the START or STOP condition, as shown in Figure 20–3,
reflect those observed at the output signals of the master driving the I2C bus. Care
should be taken when observing the SDA or SCL signals at the input signals of the
slave(s), because unequal line delays may result in an incorrect SDA or SCL timing
relationship. †
Addressing Slave Protocol

There are two address formats: the 7-bit address format and the 10-bit address format.
7-Bit Address Format

During the 7-bit address format, the first seven bits (bits 7:1) of the first byte set the
slave address and the LSB bit (bit 0) is the R/W bit as shown in Figure 20–4. When
bit 0 (R/W) is set to 0, the master writes to the slave. When bit 0 (R/W) is set to 1, the
master reads from the slave. †
Figure 20–4. 7-bit Address Format †
MSB LSB
S A6 A5 A4 A3 A2 A1 A0 R/W ACK
Slave Address
S: Start Condition
R/W: Read/Write Pulse
ACK: Acknowledge (sent by slave)
10-Bit Address Format

During 10-bit addressing, two bytes are transferred to set the 10-bit address. The
transfer of the first byte contains the following bit definition. The first five bits (bits
7:3) notify the slaves that this is a 10-bit transfer followed by the next two bits (bits
2:1), which set the slaves address bits 9:8, and the LSB bit (bit 0) is the R/W bit. The
second byte transferred sets bits 7:0 of the slave address. Figure 20–5 shows the 10-bit
address format. †
Figure 20–5. 10-Bit Address Format †
S 1 1 1 1 0 A9 A8 R/W ACK A7 A6 A5 A4 A3 A2 A1 A0 ACK
Reserved for 10-Bit Address
S: Start Condition
ACK: Acknowledge (sent by slave)

Table 20–1 defines the special purpose and reserved first byte addresses. †
Table 20–1. I2C Definition of Bits in First Byte †

Slave Address R/W Bit Description
General call address. The I2Ccontroller places the data in the receive
0000 000 0
buffer and issues a general call interrupt.
START byte. For more details, refer to “START BYTE Transfer
0000 000 1
Protocol” on page 20–9
0000 001 X CBUS address. The I2C controller ignores these accesses.
0000 010 X Reserved
0000 011 X Reserved
0000 1XX X Unused
1111 1XX X Reserved
1111 0XX X 10-bit slave addressing.
(1) ‘X’ indicates do not care.
Transmitting and Receiving Protocol

The master can initiate data transmission and reception to or from the bus, acting as
either a master-transmitter or master-receiver. A slave responds to requests from the
master to either transmit data or receive data to or from the bus, acting as either a
slave-transmitter or slave-receiver, respectively. †
Master-Transmitter and Slave-Receiver

All data is transmitted in byte format, with no limit on the number of bytes
transferred per data transfer. After the master sends the address and R/W bit or the
master transmits a byte of data to the slave, the slave-receiver must respond with the
acknowledge signal (ACK). When a slave-receiver does not respond with an ACK
pulse, the master aborts the transfer by issuing a STOP condition. The slave must
leave the SDA line high so that the master can abort the transfer. †

If the master-transmitter is transmitting data as shown in Figure 20–6, then the

slave-receiver responds to the master-transmitter with an ACK pulse after every byte
of data is received. †
Figure 20–6. Master-Transmitter Protocol †
7-Bit Address
S Slave Address R/W A Data A Data A/A P
0 (Write)
10-Bit Address
Slave Address Slave Address
S First 7 Bits R/W A Second Byte A Data A/A P
11110xxx 0 (Write)
S: Start Condition From Master to Slave

P: Stop Condition From Slave to Master
A: Acknowledge (SDA Low)
A: No Acknowledge (SDA High)
Master-Receiver and Slave-Transmitter

If the master is receiving data as shown in Figure 20–7, then the master responds to
the slave-transmitter with an ACK pulse after a byte of data has been received, except
for the last byte. This is the way the master-receiver notifies the slave-transmitter that
this is the last byte. The slave-transmitter relinquishes the SDA line after detecting the
No Acknowledge (NACK) bit so that the master can issue a STOP condition. †
When a master does not want to relinquish the bus with a STOP condition, the master
can issue a RESTART condition. This is identical to a START condition except it occurs
after the ACK pulse. Operating in master mode, the I2C controller can then
communicate with the same slave using a transfer of a different direction. For a
description of the combined format transactions that the I2C controller supports, refer
to “Combined Formats” on page 20–5. †

1 The I2C controller must be inactive on the serial port (I2C_DYNAMIC_TAR_UPDATE = 1)

before the target slave address register, IC_TAR can be reprogrammed. †
Figure 20–7. Master-Receiver Protocol †
7-Bit Address
S Slave Address R/W A Data A Data A P
1 (Read)
10-Bit Address
Slave Address Slave Address Slave Address
S First 7 bits R/W A Second Byte A R First 7 bits R/W A Data A P
11110xxx 0 (Write) 11110xxx 1 (Read)
S: Start Condition From Master to Slave

R: Restart Condition From Slave to Master
P: Stop Condition
A: Acknowledge (SDA Low)
A: No Acknowledge (SDA High)
START BYTE Transfer Protocol

The START BYTE transfer protocol is set up for systems that do not have an on-board
dedicated I2C hardware module. When the I2C controller is set as a slave, it always
samples the I2C bus at the highest speed supported so that it never requires a START
BYTE transfer. However, when I2C controller is set as a master, it supports the
generation of START BYTE transfers at the beginning of every transfer in case a slave
device requires it. This protocol consists of seven zeros being transmitted followed by
a 1, as illustrated in Figure 20–8. This allows the processor that is polling the bus to
under-sample the address phase until the microcontroller detects a 0. Once the
microcontroller detects a 0, it switches from the under sampling rate to the correct rate
of the master. †
Figure 20–8. START BYTE Transfer †
SDA Dummy Acknowledge
(High)
SCL S 1 2 7 8 9 Sr
ACK
Start Byte 00000001
The START BYTE has the following procedure: †

1. Master generates a START condition. †
2. Master transmits the START byte (0000 0001). †
3. Master transmits the ACK clock pulse. (Present only to conform with the byte
handling format used on the bus) †
4. No slave sets the ACK signal to 0. †

5. Master generates a RESTART (R) condition. †

A hardware receiver does not respond to the START BYTE because it is a reserved
address and resets after the RESTART condition is generated. †
Multiple Master Arbitration

The I2C controller bus protocol allows multiple masters to reside on the same bus. If
there are two masters on the same I2C-bus, there is an arbitration procedure if both try
to take control of the bus at the same time by simultaneously generating a START
condition. Once a master (for example, a microcontroller) has control of the bus, no
other master can take control until the first master sends a STOP condition and places
the bus in an idle state. †
Arbitration takes place on the SDA line, while the SCL line is 1. The master, which
transmits a 1 while the other master transmits 0, loses arbitration and turns off its data
output stage. The master that lost arbitration can continue to generate clocks until the
end of the byte transfer. If both masters are addressing the same slave device, the
arbitration could go into the data phase. †
Upon detecting that it has lost arbitration to another master, the I2C controller stops
generating SCL. †
Figure 20–9 illustrates the timing of when two masters are arbitrating on the bus.
Figure 20–9. Multiple Master Arbitration †
DATA1 MSB 1 DATA1 Loses Arbitration
Matching Data
DATA2 MSB 0
SDA Mirrors DATA2
SDA MSB
SCL
SDA Lines Up with

DATA1 Start Condition
The bus control is determined by address or master code and data sent by competing
masters, so there is no central master nor any order of priority on the bus. †
Arbitration is not allowed between the following conditions: †
■ A RESTART condition and a data bit †
■ A STOP condition and a data bit †
■ A RESTART condition and a STOP condition †
Slaves are not involved in the arbitration process. †

Clock Synchronization
When two or more masters try to transfer information on the bus at the same time,
they must arbitrate and synchronize the SCL clock. All masters generate their own
clock to transfer messages. Data is valid only during the high period of SCL clock.
Clock synchronization is performed using the wired-AND connection to the SCL
signal. When the master transitions the SCL clock to 0, the master starts counting the
low time of the SCL clock and transitions the SCL clock signal to 1 at the beginning of
the next clock period. However, if another master is holding the SCL line to 0, then the
master goes into a HIGH wait state until the SCL clock line transitions to 1. †
All masters then count off their high time, and the master with the shortest high time
transitions the SCL line to 0. The masters then counts out their low time and the one
with the longest low time forces the other master into a HIGH wait state. Therefore, a
synchronized SCL clock is generated, which is illustrated in Figure 20–10. Optionally,
slaves may hold the SCL line low to slow down the timing on the I2C bus. †
Figure 20–10. Multiple Master Clock Synchronization †
Wait State Start Counting High Period
CLKA
CLKB
SCL
SCL Low Transition Resets SCL Transitions High

All Clocks; Start Counting When All Clocks Are in a High State
Their Low Periods

Clock Frequency Configuration

When you configure the I2C controller as a master, the SCL count registers must be set
before any I2C bus transaction can take place in order to ensure proper I/O timing. †
There are four SCL count registers:
■ Standard speed I2C clock SCL high count, IC_SS_SCL_HCNT †
■ Standard speed I2C clock SCL low count, IC_SS_SCL_LCNT †
■ Fast speed I2C clock SCL high count, IC_FS_SCL_HCNT †
■ Fast speed I2C clock SCL low count, IC_FS_SCL_LCNT †
1 It is not necessary to program any of the SCL count registers if the I2C controller is
enabled to operate only as an I2C slave, since these registers are used only to
determine the SCL timing requirements for operation as an I2C master. †
Minimum High and Low Counts

When the I2C controller operates as an I2C master in both transmit and receive
transfers, the minimum value that can be programmed in the SCL low count registers
is 8 while the minimum value allowed for the SCL high count registers is 6. †
The minimum value of 8 for the low count registers is due to the time required for the
I2C controller to drive SDA after a negative edge of SCL. The minimum value of 6 for
the high count register is due to the time required for the I2C controller to sample SDA
during the high period of SCL. †
The I2C controller adds one cycle to the low count register values in order to generate
the low period of the SCL clock.
The I2C controller adds seven cycles to the high count register values in order to
generate the high period of the SCL clock. This is due to the following factors: †
■ The digital filtering applied to the SCL line incurs a delay of four l4_sp_clk
cycles. This filtering includes metastability removal and a 2-out-of-3 majority vote
processing on SDA and SCL edges. †
■ Whenever SCL is driven 1 to 0 by the I2C controller—that is, completing the SCL
high time—an internal logic latency of three l4_sp_clk cycles incurs. †
Consequently, the minimum SCL low time of which the I2C controller is capable is
nine (9) l4_sp_clk periods (8+1), while the minimum SCL high time is thirteen (13)
l4_sp_clk periods (6+1+3+3). †
Calculating High and Low Counts

The calculations below show an example of how to calculate SCL high and low counts
for each speed mode in the I2C controller.

The equation to calculate the proper number of l4_sp_clk clock pulses required for
setting the proper SCL clocks high and low times is as follows: †
Equation 20–1.
IC_HCNT = ceil(MIN_SCL_HIGHtime*OSCFREQ)
IC_LCNT = ceil(MIN_SCL_LOWtime*OSCFREQ)
MIN_SCL_HIGHtime = minimum high period

MIN_SCL_HIGHtime =
4000 ns for 100 kbps
600 ns for 400 kbps
60 ns for 3.4 Mbs, bus loading = 100pF
160 ns for 3.4 Mbs, bus loading = 400pF
MIN_SCL_LOWtime = minimum low period

MIN_SCL_LOWtime =
120 ns for 3.4Mbs, bus loading = 100pF
320 ns for 3.4Mbs, bus loading = 400pF
OSCFREQ = l4_sp_clk clock frequency (Hz)
For example:
OSCFREQ = 100 MHz
I2Cmode = fast, 400 kbps
MIN_SCL_HIGHtime = 600 ns
MIN_SCL_LOWtime = 1300 ns
IC_HCNT = ceil(600 ns * 100 MHz) IC_HCNTSCL PERIOD = 60

IC_LCNT = ceil(1300 ns * 100 MHz) IC_LCNTSCL PERIOD = 130
Actual MIN_SCL_HIGHtime = 60*(1/100 MHz) = 600 ns
Actual MIN_SCL_LOWtime = 130*(1/100 MHz) = 1300 ns †
SDA Hold Time

The I2C protocol specification requires 300 ns of hold time on the SDA signal in
standard and fast speed modes. Board delays on the SCL and SDA signals can mean
that the hold time requirement is met at the I2C master, but not at the I2C slave (or
vice-versa). As each application encounters differing board delays, the I2C controller
contains a software programmable register, IC_SDA_HOLD, to enable dynamic
adjustment of the SDA hold time. †

DMA Controller Interface

The I2C controller supports DMA signaling to indicate when data is ready to be read
or when the transmit FIFO needs data. This support requires 2 DMA channels, one for
transmit data and one for receive data. The I2C controller supports both single and
burst DMA transfers. System software can choose the DMA burst mode by
programming an appropriate value into the threshold registers. The recommended
setting of the FIFO threshold register value is half full.
To enable the DMA controller interface on the I2C controller, you must write to the
DMA control register (DMACR) bits. Writing a 1 into the TDMAE bit field of DMACR
register enables the I2C controller transmit handshaking interface. Writing a 1 into the
RDMAE bit field of the DMACR register enables the I2C controller receive
handshaking interface. †
Clocks
Each I2C controller is connected to the l4_sp_clk clock, which clocks transfers in
standard and fast mode. The clock input is driven by the clock manager.
Device Handbook.
Resets
Each I2C controller has a separate reset signal. The reset manager drives the signals on
a cold or warm reset.
Device Handbook.
Interface Pins
All instances of the I2C controller connect to external pins through pin multiplexers.
Pin multiplexing allows all instances to function simultaneously and independently.
The pins must be connected to a pull-up resistors and the I2C bus capacitance cannot
exceed 400 pF.
Table 20–2 shows I/O pin use of the I2C controller interface.
Table 20–2. I2C Controller Interface Pins

Pin Name Signal Width Direction Description
SCL 1 bit Bidirectional Serial clock
SDA 1 bit Bidirectional Serial data

I2C Controller Programming Model

This section describes the programming model for the I2C controllers based on the
two master and slave operation modes. †
1 Each I2C controller should be set to operate only as an I2C master or as an I2C slave,
never both simultaneously. Ensure that bit 6 (IC_SLAVE_DISABLE) and 0
(IC_MASTER_MODE) of the IC_CON register are never set to 0 and 1, respectively. †
Slave Mode Operation

This section discusses slave mode procedures. †
Initial Configuration
To use the I2C controller as a slave, perform the following steps: †
1. Disable the I2C controller by writing a 0 to bit 0 of the IC_ENABLE register. †
2. Write to the IC_SAR register (bits 9:0) to set the slave address. This is the address to
which the I2C controller responds. †
1 The reset value for the I2C controller slave address is 0x55. If you are using
0x55 as the slave address, you can safely skip this step.
3. Write to the IC_CON register to specify which type of addressing is supported (7- or
10-bit by setting bit 3). Enable the I2C controller in slave-only mode by writing a 0
into bit 6 (IC_SLAVE_DISABLE) and a 0 to bit 0 (MASTER_MODE). †
1 Slaves and masters do not have to be programmed with the same type of
addressing 7- or 10-bit address. For instance, a slave can be programmed
with 7-bit addressing and a master with 10-bit addressing, and vice versa. †
4. Enable the I2C controller by writing a 1 in bit 0 of the IC_ENABLE register. †
Slave-Transmitter Operation for a Single Byte

When another I2C master device on the bus addresses the I2C controller and requests
data, the I2C controller acts as a slave-transmitter and the following steps occur: †
1. The other I2C master device initiates an I2C transfer with an address that matches
the slave address in the IC_SAR register of the I2C controller †
2. The I2C controller acknowledges the sent address and recognizes the direction of
the transfer to indicate that it is acting as a slave-transmitter. †

3. The I2C controller asserts the RD_REQ interrupt (bit 5 of the IC_RAW_INTR_STAT
register) and waits for software to respond. †
If the RD_REQ interrupt has been masked, due to bit 5 of the IC_INTR_MASK register
(M_RD_REQ bit field) being set to 0, then it is recommended that you instruct the
CPU to perform periodic reads of the IC_RAW_INTR_STAT register. †
a. Reads that indicate bit 5 of the IC_RAW_INTR_STAT register (R_RD_REQ bit field)
being set to 1 must be treated as the equivalent of the RD_REQ interrupt being
asserted. †
b. Software must then act to satisfy the I2C transfer. †
c. The timing interval used should be in the order of 10 times the fastest SCL
clock period the I2C controller can handle. For example, for 400 Kbps, the
timing interval is 25 us. †
1 The value of 10 is recommended here because this is approximately the

amount of time required for a single byte of data transferred on the I2C
bus.†
4. If there is any data remaining in the TX FIFO before receiving the read request, the
I2C controller asserts a TX_ABRT interrupt (bit 6 of the IC_RAW_INTR_STAT register)
to flush the old data from the TX FIFO. †
1 Because the I2C controller's TX FIFO is forced into a flushed/reset state

whenever a TX_ABRT event occurs, it is necessary for software to release the
I2C controller from this state by reading the IC_CLR_TX_ABRT register before
attempting to write into the TX FIFO. For more information, refer to the
C_RAW_INTR_STAT register description in the register map.†
If the TX_ABRT interrupt has been masked, due to of IC_INTR_MASK[6] register

(M_TX_ABRT bit field) being set to 0, then it is recommended that the CPU performs
periodic reads of the IC_RAW_INTR_STAT register. †
a. Reads that indicate bit 6 (R_TX_ABRT) being set to 1 must be treated as the
equivalent of the TX_ABRT interrupt being asserted. †
b. There is no further action required from software. †
c. The timing interval used should be similar to that described in the previous
step for the IC_RAW_INTR_STAT[5] register. †
5. Software writes to the DAT bits of the IC_DATA_CMD register with the data to be
written and writes a 0 in bit 8.†
6. Software must clear the RD_REQ and TX_ABRT interrupts (bits 5 and 6, respectively)
of the IC_RAW_INTR_STAT register before proceeding. †
If the RD_REQ and/or TX_ABRT interrupts have been masked, then clearing of the
IC_RAW_INTR_STAT register will have already been performed when either the
R_RD_REQ or R_TX_ABRT bit has been read as 1. †
7. The I2C controller transmits the byte. †
8. The master may hold the I2C bus by issuing a RESTART condition or release the
bus by issuing a STOP condition. †

Slave-Receiver Operation for a Single Byte

When another I2C master device on the bus addresses the I2C controller and is
sending data, the I2C controller acts as a slave-receiver and the following steps occur:†
1. The other I2C master device initiates an I2C transfer with an address that matches
the I2C controller's slave address in the IC_SAR register. †
2. The I2C controller acknowledges the sent address and recognizes the direction of
the transfer to indicate that the I2C controller is acting as a slave-receiver. †
3. I2C controller receives the transmitted byte and places it in the receive buffer. †
1 If the RX FIFO is completely filled with data when a byte is pushed, then an
overflow occurs and the I2C controller continues with subsequent I2C
transfers. Because a NACK is not generated, software must recognize the
overflow when indicated by the I2C controller (by the R_RX_OVER bit in the
IC_INTR_STAT register) and take appropriate actions to recover from lost
data. Hence, there is a real time constraint on software to service the RX
FIFO before the latter overflow as there is no way to reapply pressure to the
remote transmitting master. †
4. I2C controller asserts the RX_FULL interrupt (IC_RAW_INTR_STAT[2] register). †

If the RX_FULL interrupt has been masked, due to setting IC_INTR_MASK[2] register
to 0 or setting IC_TX_TL to a value larger than 0, then it is recommended that the
CPU does periodic reads of the IC_STATUS register. Reads of the IC_STATUS register,
with bit 3 (RFNE) set at 1, must then be treated by software as the equivalent of the
RX_FULL interrupt being asserted. †
5. Software may read the byte from the IC_DATA_CMD register (bits 7:0). †
6. The other master device may hold the I2C bus by issuing a RESTART condition or
release the bus by issuing a STOP condition. †
Slave-Transfer Operation for Bulk Transfers

In the standard I2C protocol, all transactions are single byte transactions and the
programmer responds to a remote master read request by writing one byte into the
slave's TX FIFO. When a slave (slave-transmitter) is issued with a read request
(RD_REQ) from the remote master (master-receiver), at a minimum there should be at
least one entry placed into the slave-transmitter's TX FIFO. The I2C controller is
designed to handle more data in the TX FIFO so that subsequent read requests can
receive that data without raising an interrupt to request more data. Ultimately, this
eliminates the possibility of significant latencies being incurred between raising the
interrupt for data each time had there been a restriction of having only one entry
placed in the TX FIFO. †
This mode only occurs when I2C controller is acting as a slave-transmitter. If the
remote master acknowledges the data sent by the slave-transmitter and there is no
data in the slave's TX FIFO, the I2C controller raises the read request interrupt
(RD_REQ) and waits for data to be written into the TX FIFO before it can be sent to the
remote master. †

If the RD_REQ interrupt is masked, due to bit 5 (M_RD_REQ) of the IC_INTR_STAT register
being set to 0, then it is recommended that the CPU does periodic reads of the
IC_RAW_INTR_STAT register. Reads of IC_RAW_INTR_STAT that return bit 5 (R_RD_REQ) set
to 1 must be treated as the equivalent of the RD_REQ interrupt referred to in this
section.†
The RD_REQ interrupt is raised upon a read request, and like interrupts, must be
cleared when exiting the interrupt service handling routine (ISR). The ISR allows you
to either write 1 byte or more than 1 byte into the TX FIFO. During the transmission of
these bytes to the master, if the master acknowledges the last byte then the slave must
raise the RD_REQ again because the master is requesting for more data. †
If the programmer knows in advance that the remote master is requesting a packet of
n bytes, then when another master addresses the I2C controller and requests data, the
TX FIFO could be written with n number bytes and the remote master receives it as a
continuous stream of data. For example, the I2C controller slave continues to send
data to the remote master as long as the remote master is acknowledging the data sent
and there is data available in the TX FIFO. There is no need to issue RD_REQ again. †
If the remote master is to receive n bytes from the I2C controller but the programmer
wrote a number of bytes larger than n to the TX FIFO, then when the slave finishes
sending the requested n bytes, it clears the TX FIFO and ignores any excess bytes. †
The I2C controller generates a transmit abort (TX_ABRT) event to indicate the clearing
of the TX FIFO in this example. At the time an ACK/NACK is expected, if a NACK is
received, then the remote master has all the data it wants. At this time, a flag is raised
within the slave's state machine to clear the leftover data in the TX FIFO. This flag is
transferred to the processor bus clock domain where the FIFO exists and the contents
of the TX FIFO are cleared at that time. †
Master Mode Operation

This section discusses master mode procedures. †
Initial Configuration
For master mode operation, the target address and address format can be changed
dynamically without having to disable the I2C controller. This feature is only
applicable when the I2C controller is acting as a master because the slave requires the
component to be disabled before any changes can be made to the address. To use the
I2C controller as a master, perform the following steps: †
1. Disable the I2C controller by writing 0 to the IC_ENABLE register. †
2. Write to the IC_CON register to set the maximum speed mode supported for slave
operation (bits 2:1) and to specify whether the I2C controller starts its transfers in
7/10 bit addressing mode when the device is a slave (bit 3). †
3. Write to the IC_TAR register the address of the I2C device to be addressed. It also
indicates whether a General Call or a START BYTE command is going to be
performed by I2C. The desired speed of the I2C controller master-initiated
transfers, either 7-bit or 10-bit addressing, is controlled by the
IC_10BITADDR_MASTER bit field (bit 12). †
4. Enable the I2C controller by writing a 1 in the IC_ENABLE register. †

5. Now write the transfer direction and data to be sent to the IC_DATA_CMD register. If
the IC_DATA_CMD register is written before the I2C controller is enabled, the data
and commands are lost as the buffers are kept cleared when the I2C controller is
not enabled. †
1 For multiple I2C transfers, perform additional writes to the TX FIFO such that the TX
FIFO does not become empty during the I2C transaction. If the TX FIFO is completely
emptied at any stage, then further writes to the TX FIFO results in an independent I2C
transaction. †
Dynamic IC_TAR or IC_10BITADDR_MASTER Update

The I2C controller supports dynamic updating of the IC_TAR (bits 9:0) and
IC_10BITADDR_MASTER (bit 12) bit fields of the IC_TAR register. You can dynamically
write to the IC_TAR register provided the following conditions are met: †
■ The I2C controller is not enabled (IC_ENABLE=0); †
■ The I2C controller is enabled (IC_ENABLE=1); AND I2C controller is NOT engaged
in any Master (TX, RX) operation (IC_STATUS[5]=0); AND I2C controller is enabled
to operate in Master mode (IC_CON[0]=1); AND there are no entries in the TX FIFO
(IC_STATUS[2]=1) †
Master Transmit and Master Receive

The I2C controller supports switching back and forth between reading and writing
dynamically. To transmit data, write the data to be written to the lower byte of the I2C
Rx/Tx Data Buffer and Command Register (IC_DATA_CMD). The CMD bit [8] should be
written to 0 for I2C write operations. Subsequently, a read command may be issued by
writing "don't cares" to the lower byte of the IC_DATA_CMD register, and a 1 should be
written to the CMD bit. The I2C controller in master mode continues to initiate transfers
as long as there are commands present in the transmit FIFO. If the transmit FIFO
becomes empty, the I2C controller inserts a STOP condition after completing the
current transfers. †
Disabling the I2C Controller

The register IC_ENABLE_STATUS is added to allow software to unambiguously
determine when the hardware has completely shutdown in response to the IC_ENABLE
register being set from 1 to 0. †
1. Define a timer interval (ti2c_poll) equal to the 10 times the signaling period for
the highest I2C transfer speed used in the system and supported by the I2C
controller. For example, if the highest I2C transfer mode is 400 Kbps, then
ti2c_poll is 25 us. †
2. Define a maximum time-out parameter, MAX_T_POLL_COUNT, such that if any
repeated polling operation exceeds this maximum value, an error is reported. †
3. Execute a blocking thread/process/function that prevents any further I2C master
transactions to be started by software, but allows any pending transfers to be
completed.
1 This step can be ignored if the I2C controller is programmed to operate as

an I2C slave only. †

4. The variable POLL_COUNT is initialized to zero. †

5. Set IC_ENABLE to 0. †
6. Read the IC_ENABLE_STATUS register and test the IC_EN bit (bit 0). Increment
POLL_COUNT by one. If POLL_COUNT >= MAX_T_POLL_COUNT, exit with the relevant
error code. †
7. If IC_ENABLE_STATUS[0] is 1, then sleep for ti2c_poll and proceed to the previous
step. Otherwise, exit with a relevant success code. †

To enable the DMA controller interface on the I2C controller, you must write the DMA
Control Register (IC_DMA_CR). Writing a 1 to the TDMAE bit field of IC_DMA_CR register
enables the I2C controller transmit handshaking interface. Writing a 1 to the RDMAE bit
field of the IC_DMA_CR register enables the I2C controller receive handshaking
interface.†
The FIFO buffer depth (FIFO_DEPTH) for both the RX and TX buffers in the I2C
Transmit FIFO Underflow

During I2C serial transfers, transmit FIFO requests are made to the DMA controller
whenever the number of entries in the transmit FIFO is less than or equal to the value
in DMA Transmit Data Level Register (IC_DMA_TDLR), also known as the watermark
level. The DMA controller responds by writing a burst of data to the transmit FIFO
buffer, of length specified as DMA burst length. †
Data should be fetched from the DMA often enough for the transmit FIFO to perform
serial transfers continuously, that is, when the FIFO begins to empty, another DMA
request should be triggered. Otherwise, the FIFO will run out of data (underflow)
causing a STOP to be inserted on the I2C bus. To prevent this condition, you must set
the watermark level correctly. †

Consider the example where the assumption is made: †
DMA burst length = FIFO_DEPTH - IC_DMA_TDLR †
space in the transmit FIFO. Consider the following two different watermark level
settings: †

■ Case 1: IC_DMA_TDLR = 16: †

■ Transmit FIFO watermark level = IC_DMA_TDLR = 16: †
■ DMA burst length = FIFO_DEPTH - IC_DMA_TDLR = 48: †
■ I2C transmit FIFO_DEPTH = 64: †
■ Block transaction size = 240: †
Figure 20–11 shows the transmit FIFO when the watermark level equals 16.
Transmit
FIFO Buffer
FIFO_DEPTH = 64 FIFO_DEPTH - IC_DMA_TDLR = 48
Empty Data In DMA

Transmit FIFO
Controller
IC_DMA_TDLR = 16
Watermark Level
Data Out Full
watermark level, IC_DMA_TDLR, is quite low. Therefore, the probability of transmit
underflow is high where the I2C serial transmit line needs to transmit data, but
there is no data left in the transmit FIFO. This occurs because the DMA has not had
time to service the DMA request before the FIFO becomes empty.
■ Case 2: IC_DMA_TDLR = 48 †
■ Transmit FIFO watermark level = IC_DMA_TDLR = 48 †
■ DMA burst length = FIFO_DEPTH - IC_DMA_TDLR = 16 †
■ I2C transmit FIFO_DEPTH = 64 †
Transmit
FIFO Buffer
Transmit FIFO
FIFO_DEPTH - IC_DMA_TDLR = 16
Watermark Level
Data In DMA
Full Controller
IC_DMA_TDLR = 48
Data Out


transfer. But the watermark level, IC_DMA_TDLR, is high. Therefore, the probability
of I2C transmit underflow is low because the DMA controller has plenty of time to
service the destination burst transaction request before the I2C transmit FIFO
becomes empty. †
which the I2C transmits data to the rate at which the DMA can respond to destination
burst requests. †
Transmit FIFO Overflow

Setting the DMA burst length to a value greater than the watermark level that triggers
the DMA request might cause overflow when there is not enough space in the
transmit FIFO to service the destination burst request. Therefore, the following
equation must be adhered to in order to avoid overflow: †
DMA burst length <= FIFO_DEPTH - IC_DMA_TDLR
In case 2: IC_DMA_TDLR = 48, the amount of space in the transmit FIFO at the time of
the burst request is made is equal to the DMA burst length. Thus, the transmit FIFO
may be full, but not overflowed, at the completion of the burst transaction. †
Therefore, for optimal operation, DMA burst length should be set at the FIFO level
that triggers a transmit DMA request; that is: †
DMA burst length = FIFO_DEPTH - IC_DMA_TDLR
1 The transmit FIFO will not be full at the end of a DMA burst transfer if the I2C
controller has successfully transmitted one data item or more on the I2C serial
transmit line during the transfer. †
Receive FIFO Overflow

During I2C serial transfers, receive FIFO requests are made to the DMA whenever the
number of entries in the receive FIFO is at or above the DMA Receive Data Level
Register, that is IC_DMA_RDLR + 1. This is known as the watermark level. The DMA
responds by fetching a burst of data from the receive FIFO. †
Data should be fetched by the DMA often enough for the receive FIFO to accept serial
transfers continuously, that is, when the FIFO begins to fill, another DMA transfer is
requested. Otherwise the FIFO will fill with data (overflow). To prevent this
condition, the user must set the watermark level correctly. †

I2C Controller Address Map and Register Definitions
Receive Watermark Level

Similar to choosing the transmit watermark level described earlier, the receive
watermark level, IC_DMA_RDLR + 1, should be set to minimize the probability of
overflow, as shown in Figure 20–13. It is a trade off between the number of DMA
burst transactions required per block versus the probability of an overflow occurring.
†
Receive FIFO Underflow

DMA burst length = IC_DMA_RDLR + 1
If the number of data items in the receive FIFO is equal to the source burst length at
the time of the burst request is made, the receive FIFO may be emptied, but not
underflowed, at the completion of the burst transaction. For optimal operation, DMA
burst length should be set at the watermark level, IC_DMA_RDLR + 1. †
which in turn can avoid underflow and improve bus utilization. †
1 The receive FIFO will not be empty at the end of the source burst transaction if the I2C
controller has successfully received one data item or more on the I2C serial receive line
during the burst. †
Receive
FIFO Buffer
Empty
Data Out
Transmit FIFO DMA
Watermark Level Controller
IC_DMA_RDLR + 1
Full
Data In
I2C Controller Address Map and Register Definitions

■ i2c0
■ i2c1

■ i2c2
■ i2c3


Added programming model, address map and register definitions, clocks, reset, and
May 2012 1.1
interface pins sections.

21. UART Controller
November 2012
cv_54021-1.2
cv_54021-1.2
The hard processor system (HPS) provides two UART controllers for asynchronous
serial communication. The UART controllers are based on an industry standard 16550
UART controller. The UART controllers are instances of the Synopsys® DesignWare®
APB Universal Asynchronous Receiver/Transmitter (DW_apb_uart) peripheral.
UART Controller Features

The UART controller provides the following functionality and features:
■ Programmable character properties, such as number of data bits per character,
optional parity bits, and number of stop bits †
■ Line break generation and detection †
■ Direct memory access (DMA) controller interface
■ Prioritized interrupt identification †
■ Programmable baud rate
■ False start bit detection †
■ Automatic flow control mode per 16750 standard †
■ Internal loopback mode support
■ 128-bit transmit and receive FIFO buffer depth
■ FIFO buffer status registers †
■ FIFO buffer access mode (for FIFO buffer testing) enables write of receive FIFO
buffer by master and read of transmit FIFO buffer by master †
■ Shadow registers reduce software overhead and provide programmable reset †
■ Transmitter holding register empty (THRE) interrupt mode †
Registered

November 2012
Subscribe
21–2 Chapter 21: UART Controller
UART Controller Block Diagram and System Integration
UART Controller Block Diagram and System Integration

Figure 21–1 shows the integration of the UART controller in the HPS system.
Figure 21–1. UART Block Diagram
UART Controller MPU
To I/O Pins Serial Transmitter/ Interrupt and System IRQ

Receiver Reset Control
RX
TX
Baud Clock
RTS FIFO Buffer
Generator
CTS
DMA DMA
Interface Controller
Clock
Manager
Register Block
Reset
Manager
Slave Interface
L4 Peripheral Bus
Table 21–1 provides a summary description of the major blocks in the UART
controller.
Table 21–1. UART Controller Block Descriptions

Block Description
Slave interface Slave interface between the component and L4 peripheral bus.
Register block Provides main UART control, status, and interrupt generation functions.†
FIFO buffer Provides FIFO buffer control and storage. †
Generates the transmitter and receiver baud clock. With a reference clock
of 100 MHz, the UART controller supports transfer rates of 95 baud to
6.25 Mbaud. This supports communication with all known 16550
Baud clock generator
devices. The baud rate is controlled by programming the interrupt enable
or divisor latch high (IER_DLH) and receive buffer, transmit holding, or
divisor latch low (RBR_THR_DLL) registers.
Converts parallel data written to the UART into serial data and adds all
additional bits, as specified by the control register, for transmission. This
Serial transmitter
makeup of serial data, referred to as a character, exits the block in serial
UART. †

Chapter 21: UART Controller 21–3
Functional Description of the UART Controller
Table 21–1. UART Controller Block Descriptions

Block Description
Converts the serial data character (as specified by the control register)
received in the UART format to parallel form. Parity error detection,
Serial receiver
framing error detection and line break detection is carried out in this
block. †
The UART controller includes a DMA controller interface to indicate when
received data is available or when the transmit FIFO buffer requires data.
The DMA requires two channels, one for transmit and one for receive.
DMA interface The UART controller supports single and burst transfers. You can use
DMA in FIFO buffer and non-FIFO buffer mode.
For more specific information, refer to the DMA Controller chapter in
volume 3 of the Cyclone® V Device Handbook.

The HPS UART is based on an industry-standard 16550 UART. The UART supports
serial communication with a peripheral, modem (data carrier equipment), or data set.
The master (CPU) writes data over the slave bus to the UART. The UART converts the
data to serial format and transmits to the destination device. The UART also receives
serial data and stores it for the master (CPU). †
The UART’s registers control the character length, baud rate, parity generation and
checking, and interrupt generation. The UART’s single interrupt output signal is
supported by several prioritized interrupt types that trigger assertion. You can
separately enable or disable each of the interrupt types with the control registers. †
FIFO Buffer Support

The UART controller includes 128-byte FIFO buffers to buffer transmit and receive
data. FIFO buffer access mode allows the master to write the receive FIFO buffer and
to read the transmit FIFO buffer for test purposes. FIFO buffer access mode is enabled
with the FIFO access register (FAR). Once enabled, the control portions of the transmit
and receive FIFO buffers are reset and the FIFO buffers are treated as empty. †
When FIFO buffer access mode is enabled, you can write data to the transmit FIFO
buffer as normal; however, no serial transmission occurs in this mode and no data
leaves the FIFO buffer. You can read back the data that is written to the transmit FIFO
buffer with the transmit FIFO read (TFR) register. The TFR register provides the current
data at the top of the transmit FIFO buffer. †
Similarly, you can also read data from the receive FIFO buffer in FIFO buffer access
mode. Since the normal operation of the UART is halted in this mode, you must write
data to the receive FIFO buffer to read it back. The receive FIFO write (RFW) register
writes data to the receive FIFO buffer. The upper two bits of the 10-bit register write
framing errors and parity error detection information to the receive FIFO buffer. Bit 9
of RFW indicates a framing error and bit 8 of RFW indicates a parity error. Although you
cannot read these bits back from the receive buffer register, you can check the bits by
reading the line status register (LSR), and by checking the corresponding bits when the
data in question is at the top of the receive FIFO buffer. †

Automatic Flow Control

The UART includes 16750-compatible request-to-send (RTS) and clear-to-send (CTS)
serial data automatic flow control mode. You enable automatic flow control with the
modem control register (MCR.AFCE). †
Automatic RTS mode

Automatic RTS mode becomes active when the following conditions occur: †
■ RTS (MCR.RTS bit and MCR.AFCE bit are both set)
■ FIFO buffers are enabled (IIR_FCR.FIFOE bit is set)
When automatic RTS is enabled, the rts_n output pin is forced inactive (high) when
the receive FIFO buffer level reaches the threshold set by RCVR trigger (IIR_FCR.RT).
When rts_n is connected to the cts_n input pin of another UART device, the other
UART stops sending serial data until the receive FIFO buffer has available space (until
it is completely empty). †
The selectable receive FIFO buffer threshold values are 1, ¼, ½, and 2 less than full.
Because one additional character may be transmitted to the UART after rts_n is
inactive (due to data already having entered the transmitter block in the other UART),
setting the threshold to 2 less than full allows maximum use of the FIFO buffer with a
margin of one character. †
Once the receive FIFO buffer is completely emptied by reading the receiver buffer
register (RBR_THR_DLL), rts_n again becomes active (low), signalling the other UART
to continue sending data.†
Even when you set the correct MCR bits, if the FIFO buffers are disabled through
FCR.FIFOE, automatic flow control is also disabled. When auto RTS is not
implemented or disabled, rts_n is controlled solely by MCR.RTS. Figure 21–2
illustrates automatic RTS operation. In Figure 21–2, the character T is received because
rts_n is not detected prior to the next character entering the sending UART
transmitter. †
Figure 21–2. Automatic RTS Timing
sin
start character T stop start character T+1 stop
rts_n
1 2 3 T T+1
rx_fifo_read
Automatic CTS mode

Automatic CTS mode becomes active when the following conditions occur: †
■ AFCE (MCR.AFCE bit is set)
■ FIFO buffers are enabled (through FIFO buffer control register IIR_FCR.FIFOE) bit
When automatic CTS is enabled (active), the UART transmitter is disabled whenever
the cts_n input becomes inactive (high). This prevents overflowing the FIFO buffer of
the receiving UART. †

If the cts_n input is not deactivated before the middle of the last stop bit, another
character is transmitted before the transmitter is disabled. While the transmitter is
disabled, you can continue to write and even overflow to the transmit FIFO buffer. †
Automatic CTS mode requires the following sequence:
1. The UART status register are read to verify that the transmit FIFO buffer is full
(UART status register USR.TFNF set to zero). †
2. The current FIFO buffer level is read via the transmit FIFO level (TFL) register. †
3. Programmable THRE interrupt mode must be enabled to access the FIFO buffer
full status from the LSR. †
When using the FIFO buffer full status, software can poll this before each write to the
transmit FIFO buffer. When the cts_n input becomes active (low) again, transmission
resumes. If the FIFO buffers are disabled with the IIR_FCR.FIFOE bit, automatic flow
control is also disabled regardless of any other settings. When auto CTS is not
implemented or disabled, the transmitter is unaffected by cts_n. Figure 21–3 timing
diagram shows auto CTS operation.†
Figure 21–3. Automatic CTS Timing
sout
start data bits stop start data bits stop start data bits stop
Disabled
cts_n
Clocks
The UART controller is connected to the l4_sp_clk clock. The clock input is driven by
the clock manager.
Resets
The UART controller is connected to the uart_rst_n reset signal. The reset manager
drives the signal on a cold or warm reset.
For more information, refer to the Reset Manager chapter in volume 3 of the Cyclone V
Device Handbook.
Interrupts
The assertion of the UART interrupt output signal occurs when one of the following
interrupt types are enabled and active: †
Table 21–2. Interrupt Types and Priority †

Interrupt Type Priority
Receiver line status Highest
Received data available Second

Table 21–2. Interrupt Types and Priority †

Interrupt Type Priority
Character timeout indication Second
Transmit holding register empty Third
You can enable the interrupt types with the interrupt enable register (IER_DLH).
Programmable THRE Interrupt

The UART has a programmable THRE interrupt mode to increase system
performance. You enable the programmable THRE interrupt mode with the interrupt
enable register (IER_DLH.PTIME). When the THRE mode is enabled, THRE interrupts
and the dma_tx_req signal are active at and below a programmed transmit FIFO
buffer empty threshold level, as shown in the flowchart in Figure 21–4. †
Figure 21–4. Programmable THRE Interrupt Mode
Clear INTR
FIFO Level > TX yes

Empty Trigger?
no
THRE Interrupt no
Enabled?
yes
Set INTR
FIFO Level > TX no

Empty Trigger?
yes
The threshold level is programmed into IIR_FCR.TET. The available empty thresholds
are empty, 2, ¼, and ½. The optimum threshold value depends on the system's ability
to begin a new transmission sequence in a timely manner. However, one of these
thresholds should prove optimum in increasing system performance by preventing
the transmit FIFO buffer from running empty.

In addition to the interrupt change, line status register (LSR.THRE) also switches from
indicating that the transmit FIFO buffer is empty, to indicating that the FIFO buffer is
full. This change allows software to fill the FIFO buffer for each transmit sequence by
polling LSR.THRE before writing another character. This directs the UART to fill the
transmit FIFO buffer whenever an interrupt occurs and there is data to transmit,
instead of waiting until the FIFO buffer is completely empty. Waiting until the FIFO
buffer is empty reduces performance whenever the system is too busy to respond
immediately. You can increase system efficiency when this mode is enabled in
combination with automatic flow control.
When not selected or disabled, THRE interrupts and LSR.THRE function normally,
reflecting an empty THR or FIFO buffer. Figure 21–5 illustrates THRE interrupt
generation when not in programmable THRE interrupt mode.
Figure 21–5. Interrupt Generation without Programmable THRE Interrupt Mode
Clear INTR
TX FIFO no
Empty?
yes
THRE
no
Interrupt
Enabled?
yes
Set INTR
(INTR Is Asserted If
There Are No Interrupts)
TX FIFO Not no
Empty?
yes

UART Controller Programming Model

This section describes the programming model for the UART controller.

The UART controller includes a DMA controller interface to indicate when the receive
FIFO buffer data is available or when the transmit FIFO buffer requires data. The
DMA requires two channels, one for transmit and one for receive. The UART
controller supports both single and burst transfers.
The FIFO buffer depth (FIFO_DEPTH) for both the RX and TX buffers in the UART
Transmit FIFO Underflow

During UART serial transfers, transmit FIFO requests are made to the DMA controller
whenever the number of entries in the transmit FIFO is less than or equal to the
decoded level of the Transmit Empty Trigger (TET) field in the FIFO Control Register
(IIR_FCR), also known as the watermark level. The DMA controller responds by
writing a burst of data to the transmit FIFO buffer, of length specified as DMA burst
length. †
Data should be fetched from the DMA often enough for the transmit FIFO to perform
serial transfers continuously, that is, when the FIFO begins to empty, another DMA
request should be triggered. Otherwise, the FIFO will run out of data (underflow)
causing a STOP to be inserted on the UART bus. To prevent this condition, you must
set the watermark level correctly. †

Consider the example where the following assumption is made: †
DMA burst length = FIFO_DEPTH - decoded watermark level of IIR_FCR.TET †
space in the transmit FIFO. Consider the following two different watermark level
settings: †
■ Case 1: IIR_FCR.TET = 1, which decodes to a watermark level of 16: †
■ Transmit FIFO watermark level = decoded watermark level of IIR_FCR.TET =
16 †
■ DMA burst length = FIFO_DEPTH - decoded watermark level of IIR_FCR.TET =
112 †
■ UART transmit FIFO_DEPTH = 128 †
■ Block transaction size = 448†

Transmit
FIFO Buffer
FIFO_DEPTH = 128
FIFO_DEPTH - IIR_FCR.TET = 112
Empty
DMA
Transmit FIFO Data In Controller
Watermark Level Decoded watermark
Data Out Full level of IIR_FCR.TET = 16
watermark level, decoded level of IIR_FCR.TET, is quite low. Therefore, the
probability of transmit underflow is high where the UART serial transmit line
needs to transmit data, but there is no data left in the transmit FIFO. This occurs
because the DMA has not had time to service the DMA request before the FIFO
becomes empty.
■ Case 2: IIR_FCR.TET = 3, which decodes to a watermark level of 64: †
■ Transmit FIFO watermark level = decoded watermark level of IIR_FCR.TET =
64 †
■ DMA burst length = FIFO_DEPTH - decoded watermark level of IIR_FCR.TET =
64†
■ UART transmit FIFO_DEPTH = 128 †

Transmit
FIFO Buffer

FIFO_DEPTH - IIR_FCR.TET = 64
Transmit FIFO DMA

Decoded watermark
Watermark Level Full Data In Controller
level of IIR_FCR.TET = 64
Data Out

transfer. But the watermark level, decoded level of IIR_FCR.TET, is high. Therefore,
the probability of UART transmit underflow is low because the DMA controller
has plenty of time to service the destination burst transaction request before the
UART transmit FIFO becomes empty. †
which the UART transmits data to the rate at which the DMA can respond to
destination burst requests. †
Transmit FIFO Overflow

Setting the DMA burst length to a value greater than the watermark level that triggers
the DMA request might cause overflow when there is not enough space in the
transmit FIFO to service the destination burst request. Therefore, the following
equation must be adhered to in order to avoid overflow: †
DMA burst length <= FIFO_DEPTH - decoded watermark level of IIR_FCR.TET
In case 2: decoded watermark level of IIR_FCR.TET = 64, the amount of space in the
transmit FIFO at the time of the burst request is made is equal to the DMA burst
length. Thus, the transmit FIFO may be full, but not overflowed, at the completion of
the burst transaction. †
Therefore, for optimal operation, DMA burst length should be set at the FIFO level
that triggers a transmit DMA request; that is: †
DMA burst length = FIFO_DEPTH - decoded watermark level of IIR_FCR.TET

1 The transmit FIFO will not be full at the end of a DMA burst transfer if the UART
controller has successfully transmitted one data item or more on the UART serial
transmit line during the transfer. †
Receive FIFO Overflow

During UART serial transfers, receive FIFO requests are made to the DMA whenever
the number of entries in the receive FIFO is at or above the decoded level of Receive
Trigger (RT) field in the FIFO Control Register (IIR_FCR). This is known as the
watermark level. The DMA responds by fetching a burst of data from the receive
FIFO. †
Data should be fetched by the DMA often enough for the receive FIFO to accept serial
transfers continuously, that is, when the FIFO begins to fill, another DMA transfer is
requested. Otherwise the FIFO will fill with data (overflow). To prevent this
condition, the user must set the watermark level correctly. †
Receive Watermark Level

Similar to choosing the transmit watermark level described earlier, the receive
watermark level, decoded watermark level of IIR_FCR.RT, should be set to minimize
the probability of overflow, as shown in Figure 21–8. It is a tradeoff between the
number of DMA burst transactions required per block versus the probability of an
overflow occurring. †
Receive FIFO Underflow

DMA burst length = decoded watermark level of IIR_FCR.RT + 1
If the number of data items in the receive FIFO is equal to the source burst length at
the time of the burst request is made, the receive FIFO may be emptied, but not
underflowed, at the completion of the burst transaction. For optimal operation, DMA
burst length should be set at the watermark level, decoded watermark level of
IIR_FCR.RT. †
which in turn can avoid underflow and improve bus utilization. †
1 The receive FIFO will not be empty at the end of the source burst transaction if the
UART controller has successfully received one data item or more on the UART serial
receive line during the burst. †

UART Controller Address Map and Register Definitions
Receive
FIFO Buffer
Empty
Receive FIFO DMA

Watermark Level Data Out Controller
Decoded watermark
Full
Data In level of IIR_FCR.RT
UART Controller Address Map and Register Definitions

■ uart0
■ uart1


May 2012 1.1 Added programming model, address map and register definitions, and reset sections.

22. General-Purpose I/O Interface
November 2012
cv_54022-1.2
cv_54022-1.2
The hard processor system (HPS) provides three general-purpose I/O (GPIO)
interface modules. The GPIO modules are instances of the Synopsys® DesignWare®
APB General Purpose Programming I/O (DW_apb_gpio) peripheral.
Features of the GPIO Interface

The GPIO interface offers the following features:
■ Supports digital de-bounce
■ Configurable interrupt mode
■ Supports up to 71 I/O pins and 14 input-only pins
GPIO Interface Block Diagram and System Integration

Figure 22–1 shows a block diagram of the GPIO interface.
Figure 22–1. GPIO Interface Block Diagram
HLGPI[13..0] GPIO Interface
Reset gpio_rst_n[n] gpio0_intr_in Cortex A9 Subsystem

Interrupt & Core Generic Interrupt
Manager
Control Controller
Clock l4_mp_clk
Manager
Register
GPIO[28..0] Block
GPIO[57..29]
I/O
GPIO[70..58] Slave
Interface
L4 Peripheral Bus
Registered

November 2012
Subscribe
22–2 Chapter 22: General-Purpose I/O Interface
Functional Description of the GPIO Interface
Functional Description of the GPIO Interface

This section provides functional details of the GPIO interface.
Debounce Operation
The GPIO modules provided in the HPS include optional debounce capabilities. The
external signal can be debounced to remove any spurious glitches that are less than
one period of the external debouncing clock, gpio_db_clk.
When input interrupt signals are debounced using the gpio_db_clk debounce clock,
the signals must be active for a minimum of two cycles of the debounce clock to
guarantee that they are registered. Any input pulse widths less than a debounce clock
period are filtered out. If the input signal pulse width is between one and two
debounce clock widths it may or may not be filtered out, depending on its phase
relationship to the debounce clock. If the input pulse spans two rising edges of the
debounce clock, it is registered. If it spans only one rising edge, it is not registered.
Figure 22–2 shows a timing diagram of the debounce circuitry for both cases: a
bounced input signal, and later, a propagated input signal..
Figure 22–2. Debounce Timing With Asynchronous Reset Flip-Flops
gpio_db_clk
gpio_ext_porta
gpio_intr_in
Because the signal is
registered, it generates
the interrupt signal.
Interrupt
The signal is not registered This signal is registered because Cleared
because it does not meet the it meets the debounce clock’s 2-cycle
debounce clock’s 2-cycle requirements
requirement.
1 Enabling the debounce circuitry increases interrupt latency by two clock cycles of the
debounce clock.
Pin Directions
The pins GPIO0 through GPIO70 can be configured to be either input or output signals.
The pins HLGPI0 through HLGPI13 share pins with the HPS DDR controller and are
input-only signals.
GPIO Interface Programming Model

Debounce capability for each of the input signals on port A can be enabled or disabled
under software control by setting the corresponding bits in the gpio_debounce register
accordingly. The debounce clock must be stable and operational before the debounce
capability is enabled.

Chapter 22: General-Purpose I/O Interface 22–3
GPIO Interface Address Map and Register Definitions
Under software control, the direction of the external I/O pad is controlled by a write
to the gpio_swportx_ddr register. When configured as input mode, reading
gpio_ext_porta would read the values on the signal of the external I/O pad. When
configured as output mode, the data written to the gpio_swporta_dr register drives
the output buffer of the I/O pad. The same pins are shared for both input and output
modes, so they cannot be configured as input and output modes at the same time.
GPIO Interface Address Map and Register Definitions

■ gpio0
■ gpio1
■ gpio2



22–4 Chapter 22: General-Purpose I/O Interface

23. Timer
November 2012
cv_54023-1.2
cv_54023-1.2
The hard processor system (HPS) provides four, 32-bit, general-purpose timers
connected to the level 4 (L4) peripheral bus. The timers optionally generate an
interrupt when the 32-bit binary count-down timer reaches zero. The timers are
instances of the Synopsys® DesignWare® APB Timers (DW_apb_timers) peripheral.
1 The microprocessor unit (MPU) subsystem provides additional timers. For

information about the timers in the MPU, refer to the Cortex-A9 MPU System chapter
Features of the Timer

■ Supports interrupt generation
■ Supports free-running mode
■ Supports user-defined count mode
Registered

November 2012
Subscribe
23–2 Chapter 23: Timer
Timer Block Diagram and System Integration
Timer Block Diagram and System Integration

Figure 23–1 shows a block diagram of the timer. Each timer includes a slave interface
for control and status register (CSR) access, a register block, and a programmable
32-bit down counter that generates interrupts on reaching zero. The timer operates on
a single clock domain driven by the clock manager.
Figure 23–1. Timer Block Diagram
Timer
Reset Interrupt and Interrupt

MPU
Manager System Reset Control
Clock
Manager Register Block
Slave Interface
Functional Description of the Timer

The 32-bit timer counts down from a programmed value and generates an interrupt
when the count reaches zero. The timer has an independent clock input connected to
the system clock signal or to an external clock source. †
The timer supports the following modes of operation:
■ Free-running mode—decrementing from the maximum value (0xFFFFFFFF).
Reloads maximum value upon reaching zero.
■ User-defined count mode—generates a periodic interrupt. Decrements from the
user-defined count value loaded from the timer1 load count register
(timer1loadcount). Reloads the user-defined count upon reaching zero.
The initial value for the timer (that is, the value from which it counts down) is loaded
into the timer by the timer1loadcount register. The following events can cause a timer
to load the initial count from the timer1loadcount register: †
■ Timer is enabled after being reset or disabled
■ Timer counts down to 0

Chapter 23: Timer 23–3
Functional Description of the Timer
Clocks
Table 23–1 shows the clock signals and connections associated with the timers.
Table 23–1. Timer Clock Characteristics
Timer System Clock Notes
OSC1 timer 0
osc1_clk —
OSC1 timer 1
SP timer 0
l4_sp_clk Timer must be disabled if clock frequency changes
SP timer 1
SP timer 0 and SP timer 1 must be disabled before l4_sp_clk is changed to another

frequency. You can then re-enable the timer once the clock frequency change takes
effect. You cannot change the frequency of OSC1 timer 0 and OSC1 timer 1.
f For more information about clock performance, refer to the Clock Manager chapter in
Resets
The timers are reset by a cold or warm reset. Resetting the timers produces the
following results:
1. The timer is disabled
2. The interrupt is enabled
3. The timer enters free-running mode
4. The timer count load register value is set to zero
Interrupts
The timer1 interrupt status (timer1intstat) and timer1 end of interrupt (timer1eoi)
registers handle the interrupts. The timer1intstat register allows you to read the
status of the interrupt. Reading from the timer1eoi register returns the value 0 and
clears the interrupt. †
The timer1 control register (timer1controlreg) contains the timer1 interrupt mask bit
(timer1_interrupt_mask)to mask the interrupt. In both the free-running and user-
defined count modes of operation, the timer generates an interrupt signal when the
timer count reaches zero and the interrupt mask bit of the control register is high.
If the timer interrupt is set, then it is cleared when the timer is disabled.

Timer Programming Model
Timer Programming Model

This section describes the programming model for the timer.
Initialization
To initialize the timer, perform the following steps: †
1. Initialize the timer through the timer1controlreg register: †
a. Disable the timer by writing a 0 to the timer1 enable bit (timer1_enable) of the
timer1controlreg register. †
1 Before writing to a timer1 load count register (timer1loadcount), you must

disable the timer by writing a 0 to the timer1_enable bit of the
timer1controlreg register to avoid potential synchronization problems. †
b. Program the timer mode—user-defined count or free-running—by writing a 0

or 1, respectively, to the timer1 mode bit (timer1_mode) of the
c. Set the interrupt mask as either masked or not masked by writing a 1 or 0,
respectively, to the timer1_interrupt_mask bit of the timer1controlreg
register. †
2. Load the timer counter value into the timer1loadcount register. †
3. Enable the timer by writing a 1 to the timer1_enable bit of the timer1controlreg
register. †
Enabling or Disabling the Timer

To enable the timer, write a 1 to the timer1_enable bit of the timer1controlreg
register. To disable the timer, write a 0 to the timer1_enable bit. †
When a timer transitions to the enabled state, the current value of timer1loadcount
register is loaded into the timer counter. †
When the timer enable bit is cleared to 0, the timer counter and any associated
registers in the timer clock domain, are asynchronously reset. †
Loading the Timer Countdown Value

When a timer counter is enabled after being reset or disabled, the count value is
loaded from the timer1loadcount register; this occurs in both free-running and
user-defined count modes. †
When a timer counts down to 0, it loads one of two values, depending on the timer
operating mode: †
■ User-defined count mode—timer loads the current value of the timer1loadcount
register. Use this mode if you want a fixed, timed interrupt. Designate this mode
by writing a 1 to the timer1_mode bit of the timer1controlreg register. †

Chapter 23: Timer 23–5
Timer Address Map and Register Definitions
■ Free-running mode—timer loads the maximum value (0xFFFFFFFF). The timer

max count value allows for a maximum amount of time to reprogram or disable
the timer before another interrupt occurs. Use this mode if you want a single timed
interrupt. Enable this mode by writing a 0 to the timer1_mode bit of the
Servicing Interrupts
This section discusses various interrupt cases and how to service them.
Clearing Interrupt
You can clear an active timer interrupt by reading the timer1eoi register or by
disabling the timer. When the timer is enabled, its interrupt remains asserted until it is
cleared by reading the timer1eoi register. †
If you clear the interrupt at the same time as the timer reaches 0, the interrupt remains
asserted. This is because setting the timer interrupt takes precedence over clearing the
interrupt. †
Checking Interrupt Status

You can query the interrupt status of the timer without clearing its interrupt. To check
the interrupt status, read the timer1intstat register. †
Masking Interrupt
The timer interrupt can be masked using the timer1controlreg register. To mask an
interrupt, write a 1 to the timer1_interrupt_mask bit of the timer1controlreg
register. †
Timer Address Map and Register Definitions

■ osc1timer0
■ osc1timer1
■ sptimer0
■ sptimer1



May 2012 1.1 Added programming model and address map and register definitions sections.

24. Watchdog Timer
November 2012
cv_54024-1.2
cv_54024-1.2
The primary function of a watchdog timer is to provide a way for the system to
recover from an unresponsive state. The hard processor system (HPS) provides two
programmable watchdog timers, which are connected to the level 4 (L4) peripheral
bus. The watchdog timers are instances of the Synopsys® DesignWare® APB
Watchdog Timer (DW_apb_wdt) peripheral.
1 The microprocessor unit (MPU) subsystem provides two additional watchdog timers.
For information about the watchdog timers in the MPU, refer to the Cortex-A9 MPU
System chapter in volume 3 of the Cyclone V Device Handbook.
Features of the Watchdog Timer

The following list describes the features of the watchdog timer:
■ Programmable 32-bit timeout range
■ Timer counts down from a preset value to zero, then performs one of the following
user-configurable operations:
■ Generates a system reset †
■ Generates an interrupt, restarts the timer, and if the timer is not cleared before a
second timeout occurs, generates a system reset
■ Dual programmable timeout period, used when the time to wait after the first start
is different than that required for subsequent restarts †
■ Prevention of accidental restart of the watchdog counter †
■ Prevention of accidental disabling of the watchdog counter †
■ Pause mode for debugging
Registered

November 2012
Subscribe
24–2 Chapter 24: Watchdog Timer
Watchdog Timer Block Diagram and System Integration
Watchdog Timer Block Diagram and System Integration

Figure 24–1 shows a block diagram of the watchdog timer.
Figure 24–1. Watchdog Timer Block Diagram
Watchdog Timer
Reset
Request Reset
Interrupt & Manager
System Pause System Reset
Manager
Control Interrupt
MPU
Register Block
Slave Interface
Each watchdog timer consists of a slave interface for control and status register (CSR)
access, a register block, and a 32-bit down counter that operates on the slave interface
clock (osc1_clk). A pause input, driven by the system manager, optionally pauses the
counter when a CPU is being debugged.
The watchdog timer drives an interrupt request to the MPU and a reset request to the
reset manager.
f For more information, refer to the Cortex-A9 MPU System and Reset Manager chapters
Functional Description of the Watchdog Timer

The watchdog timers are peripherals that can be used to recover from system lockup
typically caused by software or system related issues.
Counter
Each watchdog timer is a programmable down counter that decrements by one on
each clock cycle. The watchdog timer supports 16 fixed timeout period values and
software chooses which timeout periods are desired. A timeout period is 2n osc1_clk
clock periods, where n is an integer from 16 to 31 inclusive. For information about
programming the timeout period values, refer to “Setting the Timeout Period Values”
on page 24–4. For more information about the osc1_clk clock, refer to “Clocks” on
page 24–3.

Chapter 24: Watchdog Timer 24–3
Functional Description of the Watchdog Timer
If the counter reaches zero, the watchdog timer has timed out, indicating an
unrecoverable error has occurred and a system reset is needed. Software must
continually restart the timer (which reloads the counter with the restart timeout
period value) to indicate that the system is functioning normally. Software can reload
the counter at any time by writing to the restart register. For information, refer to
“Reloading a Watchdog Counter” on page 24–4.
Software sets the watchdog timer output response mode to either generate a reset
request on a timeout, or assert an interrupt request and start counting down a second
time. In the former case, the counter wraps and keeps decrementing, even while a
reset request is asserted, until the watchdog timer is reset by the reset manager. In the
latter case, the generated interrupt is passed to the generic interrupt controller (GIC)
in the MPU subsystem. If the interrupt is not serviced by software before a second
timeout occurs, the timer generates a reset request. For information about
programming the output response mode, refer to “Selecting the Output Response
Mode” on page 24–4.
1 If a restart occurs at the same time the watchdog counter reaches zero, an interrupt is
not generated.
Pause Mode
The watchdog timers can be paused during debugging. Pausing the watchdog timers
is controlled by the system manager. The following options are available:
■ Pause the timer while either CPU0 or CPU1 is in debug mode
■ Pause the timer while only CPU1 is in debug mode
■ Pause the timer while only CPU0 is in debug mode
■ Do not pause the timer
When pause mode is enabled, the system manager pauses the watchdog timer while
debugging. When pause mode is disabled, the watchdog timer runs even while
debugging.
When the system manager exits reset, the watchdog pausing feature is enabled for
both CPUs by default.
For information about programming the pause mode, refer to “Pausing a Watchdog
Timer” on page 24–5.
Clocks
Each watchdog timer is connected to the osc1_clk clock so that timer operation is not
dependent on the phase-locked loops (PLLs) in the clock manager. This independence
allows recovery from software that inadvertently programs the PLLs in the clock
manager incorrectly.
Device Handbook.

Watchdog Timer Programming Model
Resets
Watchdog timers are reset by a cold or warm reset from the reset manager, and are
disabled when exiting reset. †
Device Handbook.

The watchdog timer is a little-endian module. This section describes the
programming options for the timers.
Setting the Timeout Period Values

The watchdog timers have a dual timeout period. The counter uses the initial start
timeout period value the first the timer is started. All subsequent restarts use the
restart timeout period. The valid values are 2(16+i) – 1 clock cycles, where i is an integer
from 0 to 15. To set the programmable timeout periods, perform the following steps:
■ To set the initial start timeout period, write i to the timeout period for initialization
field (top_init) of the watchdog timeout range register (wdt_torr).
■ To set the restart timeout period, write i to the timeout period field (top) of the
wdt_torr register.
1 Set the timeout values before enabling the timer.
Selecting the Output Response Mode

The watchdog timers have two output response modes (as described in “Counter” on
page 24–2). To select the desired mode, perform one of the following steps:
■ To generate a system reset request when a timeout occurs, write 0 to the output
response mode bit (rmod) of the watchdog timer control register (wdt_cr).
■ To generate an interrupt and restart the timer when a timeout occurs, write 1 to the
rmod field of the wdt_cr regsiter.
If a restart occurs at the same time the watchdog counter reaches zero, a system reset
is not generated. †
Enabling and Initially Starting a Watchdog Timer

To enable and start a watchdog timer, write the value 1 to the watchdog timer enable
bit (wdt_en) of the wdt_cr register.
Reloading a Watchdog Counter

To reload a watchdog counter, write the value 0x76 to the counter restart register
(wdt_crr). This unique 8-bit value is used as a safety feature to prevent accidental
restarts.

Pausing a Watchdog Timer

Pausing the watchdog timers is controlled by the L4 watchdog debug register (wddbg)
in the system manager.
f For more information, refer to register definitions section of the System Manager
Disabling and Stopping a Watchdog Timer

The watchdog timers are disabled and stopped only by resetting them from the reset
manager.
Device Handbook.
Watchdog Timer State Machine

Figure 24–2 shows a state machine for the watchdog timer functionality.
Figure 24–2. Watchdog Timer State Machine
System Reset
(Timer Disabled)
Software Sets Initial and Restart Timeout Periods (WDT_TORR),

Sets Output Response Mode (WDT_CR.RMOD),
and Enables the Timer (WDT_CR.WDT_EN)
Load Counter with Initial Timeout

Value & Start Timer
Software Reads WDT_EOI

Load Counter or Writes 0x76 to WDT_CRR
Decrement
with Restart Counter > 0
Counter
Timeout Value
Counter == 0 and Counter == 0 and

WDT_CR.RMOD == 1 WDT_CR.RMOD == 0
Assert Interrupt &

Assert System
Load Counter with
Reset Request
Restart Timeout Value
Software Reads WDT_EOI

or Writes 0x76 to WDT_CRR
Decrement
Counter > 0 Counter Counter == 0

Watchdog Timer Address Map and Register Definitions
The state machine illustrates the behavior of the watchdog timer, including the
behavior of both output response modes. Once initialized, the counter decrements at
every clock cycle. The state machine remains in the Decrement Counter state until the
counter reaches zero, or the watchdog timer is restarted. If software reads the
interrupt clear register (wdt_eoi), or writes 0x76 to the wdt_crr register, the state
changes from Decrement Counter to Load Counter with Restart Timeout Value. In this
state, the watchdog counter gets reloaded with the restart timeout value, and then the
state changes back to Decrement Counter.
If the counter reaches zero, the state changes based on the value of the output
response mode setting defined in the rmod bit of the wdt_cr register. If the rmod bit of
the wdt_cr register is 0, the output response mode is to generate a system reset
request. In this case, the state changes to Assert System Reset Request. In response, the
reset manager resets and disables the watchdog timer, and gives software the
opportunity to reinitialize the timer.
If the rmod bit of the wdt_cr register is 1, the output response mode is to generate an
interrupt. In this case, the state changes to Assert Interrupt and Load Counter with
Restart Timeout Value. An interrupt to the processor is generated, and the watchdog
counter is reloaded with the restart timeout value. The state then changes to the
second Decrement Counter state, and the counter resumes decrementing. If software
reads the wdt_eoi register, or writes 0x76 to the wdt_crr register, the state changes
from Decrement Counter to Load Counter with Restart Timeout Value. In this state,
the watchdog counter gets reloaded with the restart timeout value, and then the state
changes back to the first Decrement Counter state. If the counter again reaches zero,
the state changes to Assert System Reset Request. In response, the reset manager
resets the watchdog timer, and gives software the opportunity to reinitialize the timer.
Watchdog Timer Address Map and Register Definitions

■ l4wd0
■ l4wd1



May 2012 1.1 Added programming model and address map and register definitions sections.


25. CAN Controller
November 2012
cv_54025-1.2
cv_54025-1.2
The hardware processor system (HPS) provides two controller area network (CAN)
controllers for serial communication with the Cortex™-A9 microprocessor unit
(MPU) subsystem host processor and the direct memory access (DMA) controller
using the CAN protocol. The CAN controllers are instances of the Bosch® D_CAN
controller and compliant with ISO 11898-1.
Features of the CAN Controller

The CAN controllers offer the following features:
■ Compliant with CAN Protocol Specification 2.0 parts A and B, available from the
Bosch website (www.semiconductors.bosch.de)
■ Programmable communication rate up to 1 Mbps
■ Holds up to 128 messages
■ Error correction code (ECC)
■ 11-bit standard and 29-bit extended identifiers
■ Programmable loopback mode
■ External direct memory access (DMA) controller for large data transfers
■ Automatic retransmission

November 2012
Subscribe
25–2 Chapter 25: CAN Controller
CAN Controller Block Diagram and System Integration
CAN Controller Block Diagram and System Integration

Figure 25–1 shows the CAN controller block diagram.
Figure 25–1. CAN Controller Block Diagram
CAN Controller
CAN Bus Interface CAN Core
CAN_TXD
CAN_RXD
Message
ECC Control Handler
Signals Message
System Message
RAM RAM
Manager
Interface
Interrupt Request Host Processor
(MPU Subsystem)
Register
Block
DMA
DMA Peripheral Controller
Request Interface
Slave Interface
L4 Peripheral Bus
The CAN controller consists of the following modules and interfaces:

■ CAN core
■ Connects to the CAN bus interface
■ Handles all ISO 11898-1 protocol functions
■ Message handler
■ State machine that controls the data transfer between the message RAM and
CAN core.
■ Handles acceptance filtering and the interrupt generation
■ Message RAM
■ Storage for up to 128 messages objects
■ Single bit error correction and double bit error detection
■ Message RAM interface
■ Two separate interfaces, IF1 and IF2

Chapter 25: CAN Controller 25–3
Functional Description of the CAN Controller
■ Register block
■ Control and status registers (CSR) for module setup and indirect message
object access.
■ All host processor accesses to the message RAM are relayed through the
message RAM interface.
■ Level 4 (L4) slave interface for CSR accesses

The CAN controllers perform communication according to the CAN protocol version
2.0 parts A and B. All communication on the CAN bus is through message objects.
The CAN controller stores message objects in its internal message RAM. The host
processor cannot access the message RAM directly, instead the IF1 and IF2 message
interface register sets provide the host processor access to the messages. Messages are
passed between the message RAM and the CAN core by the message handler. The
message handler is also responsible for message level responsibilities such as
acceptance filtering, interrupt generation, and transmission request generation.
Message Object
The message RAM can store up to 128 message objects. To avoid conflicts between
host processor accesses to the message RAM and CAN message reception and
transmission, the host processor does not access the message objects directly. Accesses
are handled through the IF1 and IF2 message interface registers.
Table 25–1 shows the structure of a message object. The first row contains the message
object control flags, the second row contains the message object masks, and the third
row is the CAN message.
Table 25–1. Message Object Structure

MsgVal NewDat MsgLst IntPnd TxIE RxIE RmtEn TxRqst EoB
Msk
UMask MXtd MDir
[28:0]
ID DLC Data 0 Data 1 Data 2 Data 3 Data 4 Data 5 Data 6 Data 7
Xtd Dir
[28:0] [3:0] [7:0] [7:0] [7:0] [7:0] [7:0] [7:0] [7:0] [7:0]
Message Object Control Flags

This section describes the message object control flags.
Message Valid (MsgVal)

■ 0= The message object is ignored by the message handler
■ 1= The message object is configured and should be considered by the message
handler
The host processor must set the MsgVal bit of all unused message objects to 0 before it
resets the initialization bit (Init) of the CAN control register (CCTRL) to initialize the
CAN controller. MsgVal must also be set to 0 when the message object is no longer in
use.

The MsgVal field is directly readable from the message valid registers (MOVALA, MOVALB,
MOVALC, MOVALD, and MOVALX). However, to write to the MsgVal field for a specific
message object, the host processor must write to the message interface registers.
New Data (NewDat)

■ 0= No new data has been written into the data portion of this message object by
the message handler since last time this flag was cleared by the host processor.
■ 1= The message handler or the host processor has written new data to the data
portion of this message object.
The NewDat field is directly readable from the new data registers (MONDA, MONDB, MONDC,
MONDD, and MONDX). However, to write to the NewDat field for a specific message object,
the host processor must write to the message interface registers.
Message Lost (MsgLst)

■ 0= No message lost since last time this bit is was last reset by the host processor
■ 1= The message handler stored a new message into this object when the NewDat bit
was still set, indicating the host processor has lost a message.
MsgLst is valid only in message objects with the message direction bit (Dir) set to
receive.
Interrupt Pending (IntPnd)

■ 0= This message object is not the source of an interrupt.
■ 1= This message object is the source of an interrupt. The interrupt identifier field in
the CAN interrupt register (CIR) points to this message object if there is no other
interrupt source with higher priority.
The IntPnd field is directly readable from the interrupt pending registers (MOIPA,
MOIPB, MOIPC, MOIPD, and MOIPX). However, to write to the IntPnd field for a specific
Transmit Interrupt Enable (TxIE)

■ 0= TxIE is disabled. IntPnd is left unchanged after the successful transmission of a
frame.
■ 1= TxIE is enabled. IntPnd is set after a successful transmission of a frame.
Receive Interrupt Enable (RxIE)

■ 0= RxIE is disabled. IntPnd is left unchanged after a successful reception of a
frame.
■ 1= RxIE is enabled. IntPnd is set after a successful reception of a frame.
Remote Enable (RmtEn)

■ 0= RmtEn is disabled. At the reception of a remote frame, TxRqst is left unchanged.
■ 1= RmtEn is enabled. At the reception of a remote frame, TxRqst is set.
Transmit Request (TxRqst)

■ 0= This message object is not waiting for transmission.

■ 1= The transmission of this message object is requested and is not complete.

The TxRqst field is directly readable from the transmission request registers (MOTRA,
MOTRB, MOTRC, MOTRD, and MOTRX). However, to write to the TxRqst field for a specific
End of Block (EoB)

■ 0= Message object belongs to a FIFO buffer block and is not the last message object
of that FIFO buffer block.
■ 1= Single message object or last message object of a FIFO buffer block.
This bit is used to concatenate two or more message objects (up to 128) to build a FIFO
buffer. For single message objects (not belonging to a FIFO buffer), this bit must
always be set to 1.
Message Object Mask Bits

The message object mask bits, together with the arbitration bits, are used for
acceptance filtering of incoming messages.
Use Acceptance Mask (UMask)

■ 0= Ignore mask. Msk[28:0], MXtd, and Mdir have no effect on acceptance filtering.
For an incoming message to be accepted, all the following conditions must be met:
■ The received message is a data frame with message direction set to 0 (receive)
or is a remote frame with message direction set to 1 (transmit).
■ The received message identifier matches the message identifier (ID[28:0]) of
the message object.
■ The received identifier extension bit matches the identifier extension bit (Xtd)
of the message object.
■ 1= Use mask (Msk[28:0], MXtd, and MDir) for acceptance filtering if the respective
mask bits are set up for acceptance filtering. For an incoming message to be
accepted, all the following conditions must be met:
■ The received message is a data frame with message direction set to 0 (receive)
or is a remote frame with message direction set to 1 (transmit) with the MDir
mask bit enabled.
■ The received message identifier matches the message identifier (ID[28:0]) of
the message object with the Msk[28:0] mask bits enabled
■ The received identifier extension bit matches the identifier extension bit (Xtd)
of the message object, with the MXtd mask bit enabled.
1 If the UMask bit is set to 1, the message object's mask bits have to be programmed
during the initialization of the message object before MsgVal is set to 1.
Identifier Mask (Msk[28:0])

The identifier mask filters the corresponding bits in ID[28:0].
■ 0= The corresponding identifier bit has no effect on the acceptance filter.
■ 1= The corresponding identifier bit is used for acceptance filtering.

Extended Identifier Mask (MXtd)

■ 0= The extended frame identifier bit (Xtd) has no effect on the acceptance
filtering
■ 1= The extended frame identifier bit (Xtd) is used for acceptance filtering
When 11-bit (standard) identifiers are used for a message object, the identifiers of
received data frames are written to bits ID28 to ID18. For acceptance filtering, only
these bits, together with mask bits Msk28 to Msk18, are used.
Mask Message Direction (MDir)

■ 0= The message direction bit (Dir) has no effect on the acceptance filtering.
■ 1= The message direction bit (Dir) is used for acceptance filtering.
(1) Altera recommends always setting MDir to 1. Ignoring the message direction bit is an advanced technique that must be handled with great care.
CAN Message Bits

The arbitration fields ID28-0, Xtd, and Dir are used to define the identifier and type of
outgoing messages and are used (together with the mask fields Msk28-0, MXtd, and
MDir) for acceptance filtering of incoming messages. A received message is stored to
the valid message object with matching identifier when the direction is set to receive a
data frame or transmit a remote frame. Extended frames can be stored only in
message objects with Xtd is set to 1, standard frames in message objects with Xtd is set
to 0. If a received message (data frame or remote frame) matches more than one valid
message object, it is stored to the object with the lowest message number.
Message Identifier (ID[28:0])

■ ID28-ID0: 29-bit identifier (extended frame)
■ ID28-ID18: 11-bit identifier (standard frame)
Extended Frame Identifier (Xtd)

■ 0= The 11-bit (standard) identifier is used for this message object.
■ 1= The 29-bit (extended) identifier is used for this message object.
Message Direction (Dir)

■ 0= receive direction. When TxRqst is set to 1, a remote frame with the identifier of
this message object is transmitted. On reception of a data frame with matching
identifier, that message is stored in this message object.
■ 1= transmit direction. When TxRqst is set to 1, the respective message object is
transmitted as a data frame. On reception of a remote frame with matching
identifier, the TxRqst bit of this message object is set to 1 (if RmtEn = 1).
Data Length Code (DLC[3:0])

DLC specifies the number of data bytes in the data frame. The maximum number is
eight.
The data length code (DLC) of a message object must be defined the same as in all the
corresponding objects with the same identifier in all CAN devices. When the message
handler stores a data frame, it sets the DLC field to the value provided in the received
message.

Data Bytes 0-7 (Data 0[7:0] - Data 7[7:0])

The data bytes in a CAN data frame.
Message Interface Registers

There are two sets of message interface registers, IF1 and IF2, that provide a means
for a host processor or DMA controller to access any message object indirectly.
Message objects are transferred between the message RAM and the message buffer
registers as a single, atomic operation maintaining data consistency across the
message.
Table 25–2 lists the structure of each message interface register set, where x represents
either set 1 or set 2.
Table 25–2. Message Interface Register Set

Register
Register Name Description
Type
IFx command Specifies the transfer direction and selects the
Command IFxCMR
register portions of the message object to transfer
IFx mask Provides access to the Msk, MDir, and MXtd
IFxMSK
register mask fields of the message object
IFx arbitration Provides access to the ID, Dir, Xtd, and MsgVal
IFxARB
register arbitration fields of the message object
Provides access to the DLC, EoB, TxRqst, RmtEn,
Message IFx message
IFxMCTR RxIE, TxIE, UMask, IntPnd, MsgLst, and
buffer control register
NewDat fields of the message object
IFx data A Provides access to data bytes 0-3 of the message
IFxDA
register object
IFx data B Provides access to data bytes 4-7 of the message
IFxDB
register object
DMA Mode
The CAN controller, shown in Figure 25–1, can issue DMA controller requests to
transfer data between one or both of the message interface registers and system
memory. The CAN controller has two DMA request interfaces, called can_if1dma and
can_if2dma. The CAN peripheral request interfaces are shared with the FPGA DMA
peripheral request interfaces. To use the DMA peripheral request interface, the host
processor must access the CAN control register (CCTRL) in the protocol group
(protogrp). The peripheral request interface is selected through the system manager.
f For more information about the selecting the CAN DMA peripheral request interface,
To activate the DMA support feature and initiate a transfer, write a 1 to the DMAactive
bit in the appropriate IF command register (IFxCMR) in the message interface group
(msgifgrp). After the message object transfer is completed, the CAN controller issues
a DMA peripheral request to perform the next message object transfer. The request
remains active until the first read or write to the message interface register.

Automatic Retransmission
The CAN controller provides a means for automatic retransmission of frames that
have lost arbitration or have errors during transmission. Retransmission happens
automatically without user intervention or notification. Normal confirmation is given
when the transmission is successfully complete.
Test Mode
To enable test mode, set the test mode enable bit (Test) in the CCTRL register to 1. This
action activates write access to the CAN test register (CTR). The following sections
describe the available test modes.
Silent Mode
The CAN controller is set in the silent mode by programming the silent mode
(Silent) bit in the test register (CTR) to 1. In silent mode, the CAN controller is able to
receive valid data frames and valid remote frames, but it holds the CAN_TXD pin high,
sending no data to the CAN bus. The silent mode can be used to analyze the traffic on
a CAN bus without affecting it by the transmission of dominant bits (acknowledge
bits, error frames). In ISO 11898-1, the silent mode is called the bus monitoring mode.
Figure 25–2 shows the CAN core in silent mode.
Figure 25–2. CAN Core in Silent Mode
CAN Controller
CAN_RXD RX
CAN_TXD Vcc TX
CAN Core
Loopback Mode
The CAN controller is set in loopback mode by programming the loopback mode
(LBack) bit in the test register (CTR) to 1. In loopback mode, the CAN controller treats
its own transmitted messages as received messages and stores them (if they pass
acceptance filtering) to a receive buffer.
To be independent from external stimulation, the CAN controller ignores
acknowledge errors in loopback mode. In this mode, the CAN controller provides
internal feedback from its transmit (TX) output to its receive (RX) input. The actual
value of the input pin is disregarded by the CAN controller.

Figure 25–3 shows the CAN core in loopback mode.
Figure 25–3. CAN Core in Loopback Mode
CAN Controller
CAN_RXD RX
CAN_TXD TX
CAN Core
Combined Mode
The CAN controller is set in combined loopback and silent mode by programming the
silent mode (Silent) and loopback mode (LBack) bits in the test register (CTR) to 1.
Combined mode can be used for testing the CAN hardware without affecting other
devices connected to the CAN bus. In this mode, the CAN_RXD pin is disconnected from
the CAN core and the CAN_TXD pin is held high.
Figure 25–4 shows the CAN core in combined mode.
Figure 25–4. CAN Core in Combined Mode
CAN Controller
CAN_RXD RX
CAN_TXD Vcc TX
CAN Core
L4 Slave Interface
The host processor accesses data, control, and status information of the CAN
controller through the L4 slave interface. The slave interface supports 32-bit accesses
only.
1 This interface does not support error responses.
Clocks
The CAN controller operates on the l4_sp_clk and can_clk clock inputs. The
l4_sp_clk clock is used by the L4 slave interface and the can_clk is used to operate
the CAN core.
The can_clk clock must be programmed to be at least eight times the CAN bus
interface speed. For example, for a CAN bus interface operating at a 1 Mbps baud
rate, the can_clk clock must be set to at least 8 MHz. The l4_sp_clk clock can operate
at a clock frequency that is equal to or greater than the can_clk frequency.

f For more information about the l4_sp_clk and can_clk clocks, refer to the Clock
Manager chapter in volume 3 of the Cyclone V Device Handbook.
Resets
Each CAN controller can be reset by software or by hardware.
Software Reset
Software initialization is done by setting the Init bit in the CAN control register
(CCTRL) in the protocol group (protogrp) in the CAN controller register map. This bit
is set through the CAN protocol when a bus off condition occurs on the CAN link. The
bit is also set through the hardware reset input described in “Hardware Reset”.
1 Due to the synchronization mechanism between the two clock domains, there might
be a delay until the value written to the Init bit can be read back. To assure that the
previous value written has been accepted, read the Init bit before setting it to a new
value.
1 The bus off recovery sequence cannot be shortened by setting or resetting the Init bit.
For more information about bus off, refer to the CAN Protocol Specification 2.0 parts A
and B, available from the Bosch website (www.semiconductors.bosch.de).
Hardware Reset
Each CAN controller has a separate reset signal. The reset manager drives the signals
on a cold or warm reset. The reset signal is synchronized to both clock domains and
applied to the appropriate logic within the CAN controllers.
Device Handbook.
Interrupts
Each CAN controller generates two interrupt signals. One signal indicates error and
status interrupts and the other signal indicates message object interrupts. Both
interrupt signals connect to the global interrupt controller (GIC). Interrupts are
enabled in the CAN control register (CCTRL) in the protocol group (protogrp). The
CAN interrupt register (CIR) in the protocol group (protogrp) indicates the highest
priority interrupt that is pending.
Error Interrupts
The following error conditions generate interrupts:
■ Bus off—when the transmit error count is equal to or greater than 256, the bus off
(BOff) bit in the CAN status register (CSTS) in the protocol group (protogrp) is set
to 1.
■ Error warning—when either the transmit error counter or the receive error
counters reaches 96, the error warning status (EWarn) bit in the CAN status register
(CSTS) in the protocol group (protogrp) is set to 1.

CAN Controller Programming Model
Status Interrupts
The following status conditions generate interrupts:
■ Receive OK—when the CAN controller receives a message successfully, the RxOK
bit in the CAN status register (CSTS) in the protocol group (protogrp) is set to 1.
■ Transmit OK—when the CAN controller transmits a message successfully, the
TxOK bit in the CAN status register (CSTS) in the protocol group (protogrp) is set to
1.
■ Last error code—when a message is received or transmitted with an error, the LEC
bits in the CAN status register (CSTS) in the protocol group (protogrp) are set
according to the error type.
Message Object Interrupts

The IntPnd bits from the message objects can generate interrupts when the
corresponding message object TxIE bit or RxIE bit is set to 1. Table 25–2 lists the
location of message object interrupt information in the interrupt pending registers.
The interrupt pending registers are located in the message handler group
(msghandgrp).
Table 25–3. Message Object Interrupt Registers

Register Title Message Objects
MOIPA Interrupt pending A register 1 to 32
MOIPB Interrupt pending B register 33 to 64
MOIPC Interrupt pending C register 65 to 96
MOIPD Interrupt pending D register 97 to 128
The MOIPX register allows software to quickly detect which message object group has a
pending interrupt.

This section describes how to operate the CAN controllers.
Software Initialization
The software initialization is started by setting the Init bit in the CAN control register
(CCTRL) to 1. While the Init bit is 1, messages are not transferred to or from the CAN
bus, and the CAN_TXD CAN bus output is held in the high state. Setting the Init bit
does not change any configuration registers.
To initialize the CAN controller, the host processor must program the CAN bit timing
(CBT) register and message objects that will be used for CAN communication. If a
message object is not needed, it is sufficient to set the MsgVal bit of the message object
to not valid (0), which is the default after RAM initialization. You must set up the
entire message object before setting MsgVal bit to valid (1). The message objects are set
up through either message interface register set.
Access to the CAN bit timing (CBT) register is only enabled when the configuration
change enable (CCE) and Init bits in the CAN control register (CCTRL) are both set to 1.

Setting the Init bit to 0 finishes the software initialization. The CAN core
synchronizes itself to the data transfer on the CAN bus by waiting for the bus to reach
an idle state before it can take part in bus activities and message transfers.
The initialization of the message objects is independent of the CAN controller
initialization and can be done anytime, but the message objects should all be
configured to particular identifiers or set to not valid before message transferring
begins.
On power up, the message RAM has to be initialized. To initialize RAM, set the Init
bit to 1, then set the RAMInit bit in the CAN function register (CFR) in the protocol
group (protogrp) to 1. The RAMInit bit returns to 0 when RAM initialization
completes. During RAM initialization, all message objects are cleared to zero and the
RAM ECC bits are initialized. Access to RAM is not allowed prior to or during RAM
initialization.
CAN Message Transfer

Once the CAN controller is initialized, the CAN controller synchronizes itself to the
CAN bus and starts transferring messages.
Received messages are stored to their appropriate message objects, if they pass the
message handler's acceptance filtering. The entire message, including all arbitration
bits, Xtd, Dir, DLC, eight data bytes, the mask and control bits UMask, MXtd, MDir, EoB,
MsgLst, RxIE, TxIE, and RmtEn, is stored in the message object. Masked arbitration bits
might change in the message object when a received message is stored.
The host processor may read or update each message at any time using the message
interface registers. The message handler guarantees data consistency when the host
processor accesses the message object at the same time the message is being
transferred to or from RAM.
Messages to be transmitted are updated by the host processor. If a permanent message
object (arbitration and control bits set up during configuration are unchanged for
multiple CAN transfers) exists for the message, only the data bytes need to be
updated. If several transmit messages are assigned to the same message object (when
the number of message objects is not sufficient), the whole message object has to be
configured before the transmission of this message is requested.
The transmission of any number of message objects may be requested at the same
time they are transmitted, according to their internal priority. The message object
numbers are 1 to 128, with 1 being the lowest internal priority and 128 the highest
priority. Messages may be updated or set to invalid (MsgVal=0) at anytime, even when
their requested transmission is still pending. The old data is discarded when a
message is updated before its pending transmission has started.
Depending on the configuration on the message object, the transmission of a message
may be requested automatically by the reception of a remote frame with a matching
identifier. Remote frames are frames used to request a particular message on the CAN
network.
1 For ease in programming, Altera recommends using one IF message interface for all
receive direction activity and the other IF message interface for all transmit direction
activity.

Message Object Reconfiguration for Frame Reception

To configure a message object to receive data frames, set the Dir field to 0.
To configure a message object to receive remote frames, set the Dir field to 1, set UMask
to 1, and set RmtEn to 0.
To avoid modifying an object while it is being transmitted, you must set MsgVal to 0
before changing any of the following configuration and control bits:
■ ID[28:0]
■ Xtd
■ DLC[3:0]
■ RxIE
■ TxIE
■ RmtEn
■ EoB
■ UMask
■ Msk[28:0]
■ MXtd
■ MDir
The following fields of a message object can be changed without clearing MsgVal:
■ Data0[7:0] to Data7[7:0]
■ TxRqst
■ NewDat
■ MsgLst
■ IntPnd
Message Object Reconfiguration for Frame Transmission

To configure a message object to transmit data frames, set the Dir field to 1, and either
set UMask to 0 or set RmtEn to 1.
Before changing any of the following configuration and control bits, you must set
MsgVal to 0:
■ Dir
■ RxIE
■ TxIE
■ RmtEn
■ EoB
■ UMask
■ Msk[28:0]

CAN Controller Address Map and Register Definitions
■ MXtd
■ MDir
The following fields of a message object can be changed without clearing MsgVal:
■ ID[28:0]
■ Xtd
■ DLC[3:0]
■ Data0[7:0] to Data7[7:0]
■ TxRqst
■ NewDat
■ MsgLst
■ IntPnd
CAN Controller Address Map and Register Definitions

■ can0
■ can1


■ Minor updates.
November 2012 1.2 ■ Expanded reset section.
■ Expanded interrupts seciton.
Added block diagram and system integration, functional description, programming model,
May 2012 1.1
and address map and register definitions sections.

Section VII. Hard Processor System User
Guide

■ Chapter 26, Introduction to the HPS Component
■ Chapter 27, Instantiating the HPS Component
■ Chapter 28, HPS Component Interfaces
■ Chapter 29, Simulating the HPS Component

VII–2 Section VII: Hard Processor System User Guide

26. Introduction to the HPS Component
June 2012
cv_54026-1.0
cv_54026-1.0
The hard processor system (HPS) component is a soft component that you can
instantiate in the FPGA fabric of the Cyclone® V SoC FPGA. It enables other soft
components to interface with the HPS hard logic. The HPS component itself has a
small footprint in the FPGA fabric, because its only purpose is to enable soft logic to
connect to the extensive hard logic in the HPS.
f For a description of the HPS and its integration into the system on a chip (SoC), refer
to the Cyclone V Device Datasheet. For a description of HPS system architecture and
features, refer to the Introduction to the Hard Processor chapter in volume 3 of the
Cyclone V Device Handbook and the CoreSight Debug and Trace chapter in volume 3 of
f For descriptions of individual peripheral architectures and features, refer to the

following chapters and sections:
■ “HPS Block Diagram and System Integration” in the Introduction to the Hard
Processor chapter in volume 3 of the Cyclone V Device Handbook.
■ “Clock Manager Block Diagram and System Integration” in the Clock Manager
■ “Reset Manager Block Diagram and System Integration” in the Reset Manager
■ “Interconnect Block Diagram and System Integration” in the Interconnect chapter
■ “AXI Bridges Block Diagrams and System Integration” in the HPS-FPGA AXI
Bridges chapter in volume 3 of the Cyclone V Device Handbook.
■ “Cortex-A9 MPU Subsystem Block Diagram and System Integration” in the
Cortex-A9 Microprocessor Unit Subsystem chapter in volume 3 of the Cyclone V
Device Handbook.
■ “CoreSight Debug and Trace Block Diagram and System Integration” in the
CoreSight Debug and Trace chapter in volume 3 of the Cyclone V Device Handbook.
■ “SDRAM Controller Subsystem Block Diagram and System Integration” in the
SDRAM Controller Subsystem chapter in volume 3 of the Cyclone V Device Handbook.
■ “On-Chip RAM Block Diagram and System Integration” in the On-Chip Memory
■ “Boot ROM Block Diagram and System Integration” in the On-Chip Memory

June 2012
Subscribe
26–2 Chapter 26: Introduction to the HPS Component
■ “NAND Flash Controller Block Diagram and System Integration” in the NAND
Flash Controller chapter in volume 3 of the Cyclone V Device Handbook.
■ “SD/MMC Controller Block Diagram and System Integration” in the SD/MMC
■ “Quad SPI Flash Controller Block Diagram and System Integration” in the Quad
SPI Flash Controller chapter in volume 3 of the Cyclone V Device Handbook.
■ “FPGA Manager Block Diagram and System Integration” in the FPGA Manager
■ “System Manager Block Diagram and System Integration” in the System Manager
■ “Scan Manager Block Diagram and System Integration” in the Scan Manager
■ “DMA Controller Block Diagram and System Integration” in the DMA Controller
■ “EMAC Block Diagram and System Integration” in the Ethernet Media Access
■ “USB OTG Controller Block Diagram and System Integration” in the USB 2.0 OTG
■ “SPI Block Diagram and System Integration” in the SPI Controller chapter in
■ “I2C Controller Block Diagram and System Integration” in the I2C Controller
■ “UART Controller Block Diagram and System Integration” in the UART Controller
■ “General-Purpose I/O Interface Block Diagram and System Integration” in the
General-Purpose I/O Interface chapter in volume 3 of the Cyclone V Device Handbook.
■ “Timer Block Diagram and System Integration” in the Timer chapter in volume 3 of
■ “Watchdog Timer Block Diagram and System Integration” in the Watchdog Timer
■ “CAN Controller Block Diagram and System Integration” in the CAN Controller
To view the description and base address for a specific peripheral, scroll to and click
the link for the peripheral’s module name.
Cyclone V Device Handbook June 2012 Altera Corporation

Chapter 26: Introduction to the HPS Component 26–3


June 2012 1.0 Initial release.
May 2012 0.1 Preliminary draft.
June 2012 Altera Corporation Cyclone V Device Handbook

26–4 Chapter 26: Introduction to the HPS Component
Cyclone V Device Handbook June 2012 Altera Corporation

27. Instantiating the HPS Component
November 2012
cv_54027-1.1
cv_54027-1.1
You instantiate the hard processor system (HPS) component in Qsys. The HPS is
available in the component library under Embedded Processors. This chapter
describes the parameters available in the HPS component parameter editor, which
opens when you add or edit an HPS component.
f The HPS requires specific device targets. For a detailed list of supported devices, refer
to the Cyclone® V Device Datasheet.
f For general information about using Qsys, refer to the Creating a System with Qsys
chapter in volume 1 of the Quartus® II Handbook.
Configuring FPGA Interfaces

This section describes parameters on the FPGA Interfaces tab.
f For general information about interfaces, refer to the HPS Component Interfaces chapter

November 2012
Subscribe
27–2 Chapter 27: Instantiating the HPS Component
General Interfaces
This section describes parameters in the General group on the FPGA Interfaces tab.
When enabled, the interfaces described in Table 27–1 become visible in the HPS
component.
Table 27–1. General Parameters

Parameter Name Parameter Description Interface Name
Enables interfaces that perform the
following functions:
■ Notify the FPGA fabric that the
Enable MPU standby and event signals microprocessor unit (MPU) is in h2f_mpu_events
standby mode.
■ Wake up an MPCore processor from
a wait for event (WFE) state.
Enables a pair of 32-bit unidirectional
general-purpose interfaces between the
Enable MPU general purpose signals h2f_mpu_gp
FPGA fabric and the FPGA manager in
the HPS portion of the SoC device.
Enables interface for FPGA interrupt f2h_irq0
Enable FPGA-to-HPS Interrupts
signals to the MPU (in the HPS). f2h_irq1
Enables debug interface to the FPGA, h2f_debug_apb

Enable Debug APB interface allowing access to debug components in h2f_debug_apb_sideband
the HPS. (1) h2f_debug_apb_clock
Enables system trace macrocell (STM)
hardware events, allowing logic inside
Enable System Trace Macrocell hardware events f2h_stm_hw_events
the FPGA to insert messages into the
trace stream. (1)
Enables the cross trigger interface (CTI),
which allows trigger sources and sinks h2f_cti
Enable FPGA Cross Trigger interface
to interface with the embedded cross h2f_cti_clock
trigger (ECT). (1)
Enables an interface between the trace
port interface unit (TPIU) and logic in the h2f_tpiu
Enable FPGA Trace Port Interface Unit FPGA. The TPIU is a bridge between
on-chip trace sources and a trace h2f_tpiu_clock_in
port. (1)
(1) For information about this functionality, refer to the CoreSight Debug and Trace chapter in volume 3 of the Cyclone V Device Handbook.

Chapter 27: Instantiating the HPS Component 27–3
Boot and Clock Selection Interfaces

This section describes parameters in the Boot and Clock Selection group in the FPGA
Interfaces tab.
Table 27–2 lists the available parameters.
Table 27–2. Boot and Clock Selection Parameters

Parameter Name Parameter Description
Enables an input to the HPS indicating whether a preloader is
Enable boot from FPGA ready available in on-chip RAM. If the input is asserted, a preloader
image is ready at memory location 0.
Enables an input to the HPS indicating whether a fallback
preloader is available in on-chip RAM. If the input is asserted, a
fallback preloader image is ready at memory location 0. The
Enable boot from FPGA on failure
fallback preloader is to be used only if the HPS boot ROM does
not find a valid preloader image in the selected flash memory
device.
f For detailed information about the HPS boot sequence, refer to the Booting and
Configuration appendix in volume 3 of the Cyclone V Device Handbook.
AXI Bridges
This section describes parameters in the AXI Bridges group on the FPGA Interfaces
tab.
Table 27–3. Bridge Parameters

Enable or disable the FPGA-to-HPS interface; if
FPGA-to-HPS interface width enabled, set the data width to 32, 64, or 128 f2h_axi_slave
bits.
Enable or disable the HPS-to-FPGA interface; if
HPS-to-FPGA interface width enabled, set the data width to 32, 64, or 128 h2f_axi_master
bits.
Enable or disable the lightweight HPS-to-FPGA
Lightweight HPS-to-FPGA interface width interface. When enabled, the data width is 32 h2f_lw_axi_master
bits.
To facilitate accessing these slaves from a memory-mapped master with a smaller

address width, you can use the Altera® Address Span Extender. The address span
extender is discussed in “Using the Address Span Extender Component” on
page 27–9.
f For more information, refer to the Interconnect chapter in volume 3 of the Cyclone V
Device Handbook.

This section describes parameters in the FPGA-to-HPS SDRAM Interface group on
the FPGA Interfaces tab.

You can add one or more SDRAM ports that make the HPS SDRAM subsystem
accessible to the FPGA fabric.
You can configure the slave interface to a data width of 32, 64, 128, or 256 bits. To
facilitate accessing this slave from a memory-mapped master with a smaller address
width, you can use the Altera Address Span Extender. The address span extender is
discussed in “Using the Address Span Extender Component” on page 27–9.
On the FPGA Interfaces tab, in the FPGA to HPS SDRAM Interface table, use + or –
to add or remove FPGA-to-HPS SDRAM interfaces. The Name column denotes the
interface name. Table 27–4 shows the parameters available for each SDRAM interface.
Table 27–4. FPGA-to-HPS SDRAM Interface Parameters

Name Port name (auto assigned as shown in Table 27–5)
Interface type:
■ AXI-3
Type ■ Avalon-MM Bidirectional
■ Avalon-MM Write-only
■ Avalon-MM Read-only
Width 32, 64, 128, or 256
Table 27–5. FPGA-to-HPS SDRAM Port and Interface Names

Port Name Interface Name
f2h_sdram0 f2h_sdram0_data
f For more information, refer to the SDRAM Controller Subsystem chapter in volume 3 of

Configuring Peripheral Pin Multiplexing
Reset Interfaces
This section describes parameters in the Resets group on the FPGA Interfaces tab.
You can enable most resets on an individual basis. Table 27–6 lists the available reset
parameters.
Table 27–6. Reset Parameters

Enable interface for HPS-to-FPGA cold
Enable HPS-to-FPGA cold reset output h2f_cold_reset
reset output
Enable an additional pair of reset
handshake signals allowing soft logic to
Enable HPS warm reset handshake signals h2f_warm_reset_handshake
notify the HPS when it is safe to initiate a
warm reset in the FPGA fabric.
Enable interface for FPGA-to-HPS debug
Enable FPGA-to-HPS debug reset request f2h_debug_reset_req
reset request
Enable interface for FPGA-to-HPS warm
Enable FPGA-to-HPS warm reset request f2h_warm_reset_req
reset request
Enable interface for FPGA-to-HPS cold
Enable FPGA-to-HPS cold reset request f2h_cold_reset_req
reset request
f For more information about the reset interfaces, refer to “Functional Description of
the Reset Manager” in the Reset Manager chapter in volume 3 of the Cyclone V Device
Handbook.
DMA Peripheral Request

This section describes parameters in the DMA Peripheral Request group on the
FPGA Interfaces tab.
You can enable each direct memory access (DMA) controller peripheral request ID
individually. Each request ID enables an interface for FPGA soft logic to request one
of eight logical DMA channels to the FPGA.
1 Peripheral request IDs 4–7 are shared with the controller area network (CAN)
controllers.

This section describes parameters on the Peripheral Pin Multiplexing tab.
Configuring Peripherals
The Peripheral Pin Multiplexing tab contains a group of parameters for each
available type of peripheral. You can enable one or more instances of each peripheral
type by selecting an HPS I/O pin set for each instance. When enabled, some
peripherals also have a mode settings specific to their functions.

Each list in the Peripheral Pin Multiplexing tab has a hint describing in detail the
options available in the list. The hint for each pin multiplexing list shows the I/O pins
used by each available pin set. The hint for each mode list shows the signals used by
each available mode.
View each hint by hovering over the corresponding list.
The tooltips for the combo boxes show useful information. The pin multiplexing
parameters have tables showing the pin destinations and the mode parameters show
which signals are used in which mode.
You can enable the following types of peripherals. For details of peripheral-specific
settings, refer to the chapter for each peripheral:
■ Ethernet Media Access Controller—Ethernet Media Access Controller chapter in
volume 3 of the Cyclone V Device Handbook
■ NAND Flash Controller—NAND Flash Controller chapter in volume 3 of the
■ Quad serial peripheral interface (SPI) Flash Controller—Quad SPI Flash Controller
chapter in volume 3 of the Cyclone V Device Handbook
■ Secure Digital / MultiMediaCard (SD/MMC) Controller—SD/MMC Controller
chapter in volume 3 of the Cyclone V Device Handbook
■ USB 2.0 On-The-Go (OTG) Controllers—USB 2.0 OTG Controller chapter in
volume 3 of the Cyclone V Device Handbook
■ SPI Controllers—SPI Controller chapter in volume 3 of the Cyclone V Device
Handbook
■ UART Controllers— UART Controller chapter in volume 3 of the Cyclone V Device
Handbook
■ Inter-integrated circuit (I2C) Controllers— I2C Controller chapter in volume 3 of the
■ CAN Controllers—CAN Controller chapter in volume 3 of the Cyclone V Device
Handbook.
■ Trace port interface unit (TPIU)—CoreSight Debug and Trace chapter in volume 3 of
the Cyclone V Device Handbook. Enabling the TPIU exposes trace signals to the
device pins.
Connecting Unassigned Pins to GPIO

On the Peripheral Pin Multiplexing tab, the Conflicts table shows pins that are not
assigned to any peripheral. By default, general-purpose I/O (GPIO) is disabled on
these pins. You can enable a pin as GPIO by changing the GPIO Enabled field in the
table. The table also shows the GPIO pin number assigned to that pin.
Resolving Pin Multiplexing Conflicts

Use the Conflicts table to view pins with invalid multiple assignments. The table
shows one of two or more peripheral interfaces assigned to the same pin(s). You can
use the peripherals’ pin configuration to determine which other peripheral(s) are in
conflict, and to resolve the conflict.

Configuring HPS Clocks
f For detailed information about available HPS pin configurations, refer to the
Cyclone V Device Family Pin Connection Guidelines.
Configuring HPS Clocks

This section describes parameters on the HPS Clocks tab.
f For general information about clock signals, refer to the Clock Manager chapter in
User Clocks
This section describes parameters in the User Clocks group on the HPS Clocks tab.
When you enable a user clock, you must manually enter its maximum frequency for
timing analysis. The TimeQuest Timing Analyzer has no other information about how
software running on the HPS configures the phase-locked loop (PLL) outputs. Each
possible clock, including clocks that are available from peripherals, has its own
parameter for describing the clock frequency.
Table 27–7 lists the user clock parameters. The frequencies that you provide are the
maximum expected frequencies. The actual clock frequencies can be modified
through the register interface, for example by software running on the microprocessor
unit (MPU). For further details, refer to “Selecting PLL Output Frequency and Phase”
on page 27–9.
Table 27–7. User Clock Parameters

Clock Interface
Name
Enable HPS-to-FPGA user 0 clock Enable main PLL from HPS to FPGA
h2f_user0_clock
User 0 clock frequency Specify the maximum expected frequency for the main PLL
Enable HPS-to-FPGA user 1 clock Enable peripheral PLL from HPS to FPGA
h2f_user1_clock
User 1 clock frequency Specify the maximum expected frequency for the peripheral PLL
Enable HPS-to-FPGA user 2 clock Enable SDRAM PLL from HPS to FPGA
h2f_user2_clock
User 2 clock frequency Specify the maximum expected frequency for the SDRAM PLL
The clock frequencies you provide are reported in a Synopsys Design Constraints File
(.sdc) generated by Qsys.
f For details about driving these clocks, refer to the Clock Manager chapter in volume 3

Configuring the External Memory Interface
PLL Reference Clocks

This section describes parameters in the PLL Reference Clocks group on the HPS
Clocks tab.
Table 27–8. PLL Reference Clock Parameters

Parameter Name Parameter Description Clock Interface Name
Enable FPGA-to-HPS peripheral PLL reference Enable the interface for FPGA fabric to supply
f2h_periph_ref_clock
clock reference clock to HPS peripheral PLL
Enable FPGA-to-HPS SDRAM PLL reference Enable the interface for FPGA fabric to supply
f2h_sdram_ref_clock
clock reference clock to HPS SDRAM PLL
Device Handbook.
Configuring the External Memory Interface

This section describes parameters on the SDRAM tab.
The HPS supports one memory interface implementing double data rate 2 (DDR2),
double data rate 3 (DDR3), and low-power double data rate 2 (LPDDR2) protocols.
The interface can be up to 40 bits wide with optional error correction code (ECC).
Configuring the HPS SDRAM controller is similar to configuring any other Altera
SDRAM controller. There are several important differences:
■ The HPS parameter editor supports all SDRAM protocols with one tab. When you
parameterize the SDRAM controller, you must specify the memory protocol:
DDR2, DDR3, or LPDDR2.
To select the memory protocol, select DDR2, DDR3, or LPDDR2 from the SDRAM
Protocol list in the PHY Settings tab in the SDRAM tab. After you select the
protocol, settings not applicable to that protocol are disabled.
1 Many HPS SDRAM controller settings are the same as for Altera’s
dedicated DDR2, DDR3, and LPDDR2 controllers. This section only
describes SDRAM parameters that are specific to the HPS component.
■ Because the HPS memory controller is not configurable through the Quartus II
software, the Controller and Diagnostic tabs are not present in the HPS parameter
editor.
■ Some settings, such as the controller settings, are not included because they can
only be configured through the register interface, for example by software running
on the MPU.
■ Unlike the memory interface clocks in the FPGA, the memory interface clocks for
the HPS are initialized by the boot-up code using values provided by the
configuration process. You can accept the values provided by UniPHY, or you can
use your own PLL settings, as described in “Selecting PLL Output Frequency and
Phase”.
1 The HPS does not support external memory interface (EMIF) synthesis generation,
compilation, or timing analysis.

Using the Address Span Extender Component
The HPS memory controller cannot be bonded with a memory controller on the FPGA
portion of the device.
f For detailed information about SDRAM controller parameters, refer to the following
chapters:
■ The Implementing and Parameterizing Memory IP chapter in volume 2 of the External

■ The Functional Description—Hard Memory Interface chapter in volume 3 of the
External Memory Interface Handbook. “EMI-Related HPS Features in SoC Devices”
describes features specific to the HPS SDRAM controller.
Selecting PLL Output Frequency and Phase

You select PLL output frequency and phase with controls in the PHY Settings tab in
the SDRAM tab. In the HPS, PLL frequencies and phases are set by software at
system startup. A PLL might not be able to produce the exact frequency that you
specify in Memory clock frequency. Normally, the Quartus II software sets Achieved
memory clock frequency to the closest achievable frequency, using an algorithm that
tries to balance frequency accuracy against clock jitter. This clock frequency is used for
timing analysis by the TimeQuest analyzer.
It is possible to use a different software algorithm for configuring the PLLs. You can
force the Achieved memory clock frequency box to take on the same value as
Memory clock frequency, by turning on Use specified frequency instead of
calculated frequency in the PHY Settings tab, under Clocks.
1 If you turn on Use specified frequency instead of calculated frequency, the

Quartus II software assumes that the value in the Achieved memory clock frequency
box is correct. If it is not, timing analysis results are incorrect.
Using the Address Span Extender Component

The FPGA-to-HPS bridge and FPGA-to-HPS SDRAM memory-mapped interfaces
expose their entire 4 GB address spaces to the FPGA fabric. The Address Span
Extender component provides a memory-mapped window into the address space that
it masters. Using the address span extender, you can expose portions of the HPS
memory space without needing to expose the entire 4-GB address space.
You can use the address span extender between a soft logic master and an
FPGA-to-HPS bridge or FPGA-to-HPS SDRAM interface. This component reduces the
number of address bits required for a master to address a memory-mapped slave
interface located in the HPS.

Generating and Compiling the HPS Component
Figure 27–1 shows how two address span extender components might be used in a
system with the HPS.
Figure 27–1. Address Span Extender
Qsys System
DMA HPS
4 GB
M S
512 MB FPGA-to-SDRAM
Address Span 4 GB
S Extender M S Interface
512 MB 512 MB
Nios II Window
Processor
M FPGA-to-HPS
1 GB Address Span 4 GB Bridge
M S Extender M S
512 MB
Window
You can also use the address span extender in the HPS-to-FPGA direction, for slave
interfaces in the FPGA. In this case, the HPS-to-FPGA bridge exposes a limited,
variable address space in the FPGA, which can be paged in using the address span
extender.
For example, suppose that the HPS-to-FPGA bridge has a 1 GB span, and the HPS
needs to access three independent 1 GB memories in the FPGA portion of the device.
To achieve this, the HPS programs the address span extender to access one SDRAM
(1 GB) in the FPGA at a time. This technique is commonly called paging or
windowing.
f For more information about the address span extender, refer to “Bridges” in the Qsys
Interconnect and System Design Components chapter in volume 1 of the Quartus II
Handbook.
Generating and Compiling the HPS Component

The process of generating and compiling an HPS design is very similar to the process
for any other Qsys project. Perform the following steps:
1. Generate the design with Qsys. The generated files include an .sdc file containing
clock timing constraints. If simulation is enabled, simulation files are also
generated.
f For information about generating a Qsys project, refer to the Creating a

System with Qsys chapter in volume 1 of the Quartus II Handbook. For a
description of the simulation files generated, refer to “Simulation Flow” in
the Simulating the HPS Component chapter in volume 3 of the Cyclone V
Device Handbook.
2. Add system.qip to the Quartus II project. system.qip is the Quartus II IP File for
the HPS component, generated by Qsys.

3. Perform Analysis and Elaboration with the Quartus II software.

4. Assign constraints to the SDRAM component. When Qsys generates the HPS
component (step 1), it generates the pin assignment Tcl Script File (.tcl) to perform
memory assignments. The script file name is
<qsys_system_name>_pin_assignments.tcl, where <qsys_system_name> is the name
of your Qsys system. Run this script to assign constraints to the SDRAM
component.
f For information about running the pin assignment script, refer to

“MegaWizard Plug-In Manager Flow” in the Implementing and
Parameterizing Memory IP chapter in volume 2 of the External Memory
Interface Handbook.
You do not need to specify pin assignments other than memory assignments.
When you configure pin multiplexing as described in “Configuring Peripheral Pin
Multiplexing” on page 27–5, you implicitly make pin assignments for all HPS
peripherals. Each peripheral is routed exclusively to the pins you specify. HPS I/O
signals are exported to the top level of the Qsys design, with information enabling
the Quartus II software to make pin assignments automatically.
You can view and modify the assignments in the Peripheral Pin Multiplexing tab.
You can also view the assignments in the Quartus fitter report.
5. Compile the design with the Quartus II software.
6. Optionally back-annotate the SDRAM pin assignments, to eliminate pin
assignment warnings the next time you compile the design.
h For information about back-annotating pin assignments, refer to About

Back-Annotating Assignments in Quartus II Help.


■ Added debug interfaces
■ Added boot options
■ Corrected slave address width
November 2012 1.1 ■ Corrected SDRAM interface widths
■ Added TPIU peripheral
■ Added .sdc file generation
■ Added .tcl script for memory assignments


28. HPS Component Interfaces
November 2012
cv_54028-1.1
cv_54028-1.1
This chapter describes the interfaces, including clocks and resets, implemented by the
hard processor system (HPS) component.
The majority of the resets can be enabled on an individual basis. The exception is the
h2f_reset interface, which is always enabled.
You must declare the clock frequency of each HPS-to-FPGA clock for timing purposes.
Each possible clock, including ones that are available from peripherals, has its own
parameter for describing the clock frequency. Declaring the clock frequency for
HPS-to-FPGA clocks specifies how you plan to configure the PLLs and peripherals, to
enable TimeQuest to accurately estimate system timing. It has no effect on PLL
settings.
f For information about instantiating the HPS component, refer to the Instantiating the
HPS Component chapter in volume 3 of the Cyclone® V Device Handbook. For Avalon™
protocol timing, refer to Avalon Interface Specifications. For Advanced Microcontroller
Bus Architecture (AMBA®) Advanced eXtensible Interface (AXI™) protocol timing,
refer to the AMBA AXI Protocol Specification v1.0, which you can download from the
Memory-Mapped Interfaces
FPGA-to-HPS Bridge
Table 28–1. FPGA-to-HPS Bridges and Clocks
Interface Name Description Associated Clock Interface (1)
f2h_axi_slave FPGA-to-HPS AXI slave interface f2h_axi_clock
(1) Refer to “Clocks” on page 28–4 for information about clock interfaces.
The FPGA-to-HPS interface is a configurable data width AXI slave allowing FPGA
masters to issue transactions to the HPS. This interface allows the FPGA fabric to
access the majority of the HPS slaves. This interface also provides a coherent memory
interface.
The FPGA-to-HPS interface is an AXI-3 compliant interface with the following
features:
■ Configurable data width: 32, 64, or 128 bits

November 2012
Subscribe
28–2 Chapter 28: HPS Component Interfaces
■ Accelerator Coherency Port (ACP) sideband signals

■ HPS-side AXI bridge to manage clock crossing, buffering, and data width
conversion.
Other interface standards in the FPGA fabric, such as connecting to Avalon®
Memory-Mapped (Avalon-MM) interfaces, can be supported through the use of soft
logic adaptors. The Qsys system integration tool automatically generates adaptor
logic to connect AXI to Avalon-MM interfaces.
This interface has an address width of 32 bits. To access existing Avalon-MM/AXI
masters, you can use the Altera® Address Span Extender.
f For more information about the FPGA-to-HPS bridge, refer to the HPS-FPGA AXI
Bridges chapter in volume 3 of the Cyclone V Device Handbook. For information about
the address span extender, refer to “Using the Address Span Extender Component” in
the Instantiating the HPS Component chapter in volume 3 of the Cyclone V Device
Handbook.
ACP Sideband Signals

For communication with the ACP on the microprocessor unit (MPU) subsystem, AXI
sideband signals are used to describe the inner cacheable attributes for the
transaction.
f For more information about the ACP sideband signals, refer to the Cortex-A9
Microprocessor Unit Subsystem chapter in volume 3 of the Cyclone V Device Handbook.
HPS-to-FPGA and Lightweight HPS-to-FPGA Bridges

Table 28–2. HPS-to-FPGA and Lightweight HPS-to-FPGA Bridges and Clocks
h2f_axi_master HPS-to-FPGA AXI master interface h2f_axi_clock
HPS-to-FPGA lightweight AXI master
h2f_lw_axi_master h2f_lw_axi_clock
interface
The HPS-to-FPGA interface is a configurable data width AXI master (32, 64, or
128-bit) that allows HPS masters to issue transactions to the FPGA fabric.
The lightweight HPS-to-FPGA interface is a 32-bit AXI master that allows HPS
masters to issue transactions to the FPGA fabric.
Both HPS-to-FPGA interfaces are AXI-3 compliant. The HPS-side AXI bridges manage
clock crossing, buffering, and data width conversion where necessary.
Other interface standards in the FPGA fabric, such as connecting to Avalon-MM
interfaces, can be supported through the use of soft logic adaptors. The Qsys system
integration tool automatically generates adaptor logic to connect AXI to Avalon-MM
interfaces.

Chapter 28: HPS Component Interfaces 28–3
Each AXI bridge accepts a clock input from the FPGA fabric and performs clock
domain crossing internally. The exposed AXI interface operates on the same clock
domain as the clock supplied by the FPGA fabric.
f For more information, refer to the HPS-FPGA AXI Bridges chapter in volume 3 of the

The FPGA-to-HPS SDRAM interface is a direct connection between the FPGA fabric
and the HPS SDRAM controller. This interface is highly configurable, allowing a mix
between number of ports and port width. The interface supports both AXI-3 and
Avalon-MM protocols.
Table 28–3. HPGA-to-HPS SDRAM Interfaces and Clocks

f2h_sdram0_data SDRAM AXI or Avalon-MM port 0 f2h_sdram0_clock
The FPGA-to-HPS SDRAM interface is a configurable interface to the multi-port

SDRAM controller.
The total data width of all interfaces is limited to a maximum of 256 bits in the read
direction and 256 bits in the write direction. The interface is implemented as four
64-bit read ports and four 64-bit write ports. As a result, the minimum data width
used by the interface is 64 bits, regardless of the number or type of interfaces.
You can configure this interface the following ways:
■ AXI-3 or Avalon-MM protocol
■ Number of interfaces
■ Data width of interfaces
The FPGA-to-HPS SDRAM interface supports six command ports, allowing up to six
Avalon-MM interfaces or three bidirectional AXI interfaces.
Each command port is available either to implement a read or write command port
for AXI, or to form part of an Avalon-MM interface.
You can use a mix of Avalon-MM and AXI interfaces, limited by the number of
command/data ports available. Some AXI features are not present in Avalon-MM
interfaces.
This interface has an address width of 32 bits. To access existing Avalon-MM/AXI
masters, you can use the Altera Address Span Extender.

Clocks
f For more information about available combinations of interfaces and ports, refer to
the SDRAM Controller Subsystem chapter in volume 3 of the Cyclone V Device
Handbook. For information about the address span extender, refer to “Using the
Address Span Extender Component” in the Instantiating the HPS Component chapter in
Clocks
The HPS-to-FPGA clock interface supplies physical clocks and resets to the FPGA.
These clocks and resets are generated in the HPS.
Alternative Clock Inputs to HPS PLLs

This section lists alternative clock inputs to HPS PLLs.
■ f2h_periph_ref_clock—FPGA-to-HPS peripheral PLL reference clock. You can
connect this clock input to a clock in your design that is driven by the clock
network on the FPGA side.
■ f2h_sdram_ref_clock—FPGA-to-HPS SDRAM PLL reference clock. You can
connect this clock to a clock in your design that is driven by the clock network on
the FPGA side.
User Clocks
A user clock is a PLL output that is connected to the FPGA fabric rather than the HPS.
You can connect a user clock to logic that you instantiate in the FPGA fabric.
■ h2f_user0_clock—HPS-to-FPGA user clock, driven from main PLL
■ h2f_user1_clock—HPS-to-FPGA user clock, driven from peripheral PLL
■ h2f_user2_clock—HPS-to-FPGA user clock, driven from SDRAM PLL
AXI Bridge FPGA Interface Clocks

The AXI interface has an asynchronous clock crossing in the FPGA-to-HPS bridge.
The FPGA-to-HPS and HPS-to-FPGA interfaces are synchronized to clocks generated
in the FPGA fabric. These interfaces can be asynchronous to one another. The SDRAM
controller’s multiport front end (MPFE) transfers the data between the FPGA and
HPS clock domains.
■ f2h_axi_clock—AXI slave clock for FPGA-to-HPS bridge, generated in FPGA
fabric
■ h2f_axi_clock—AXI master clock for HPS-to-FPGA bridge, generated in FPGA
fabric
■ h2f_lw_axi_clock—AXI master clock for lightweight HPS-to-FPGA bridge,
generated in FPGA fabric
SDRAM Clocks
You can configure the HPS component with up to six FPGA-to-HPS SDRAM clocks.

Resets
Each command channel to the SDRAM controller has an individual clock source from
the FPGA fabric. The interface clock is always supplied by the FPGA fabric, with clock
crossing occurring on the HPS side of the boundary.
The FPGA-to-HPS SDRAM clocks are driven by soft logic in the FPGA fabric.
■ f2h_sdram0_clock—SDRAM clock for port 0
Resets
This section describes the reset interfaces to the HPS component.
f For details about the HPS reset sequences, refer to “Functional Description of the
Reset Manager” in the Reset Manager chapter in volume 3 of the Cyclone V Device
Handbook.
HPS-to-FPGA Reset Interfaces

The following interfaces allow the HPS to reset soft logic in the FPGA fabric:
■ h2f_reset—HPS-to-FPGA cold and warm reset
■ h2f_cold_reset—HPS-to-FPGA cold reset
■ h2f_warm_reset_handshake—Warm reset request and acknowledge interface
between HPS and FPGA
HPS External Reset Sources

The following interfaces allow soft logic in the FPGA fabric to reset the HPS:
■ f2h_cold_reset_req—FPGA-to-HPS cold reset request
■ f2h_warm_reset_req—FPGA-to-HPS warm reset request
■ f2h_dbg_reset_req—FPGA-to-HPS debug reset request
Debug and Trace Interfaces
Trace Port Interface Unit

The TPIU is a bridge between on-chip trace sources and a trace port.
■ h2f_tpiu
■ h2f_tpiu_clock_in

Peripheral Signal Interfaces
FPGA System Trace Macrocell Events Interface

The system trace macrocell (STM) hardware events allow logic in the FPGA to insert
messages into the trace stream.
■ f2h_stm_hw_events
FPGA Cross Trigger Interface

The cross trigger interface (CTI) allows trigger sources and sinks to interface with the
embedded cross trigger (ECT).
■ h2f_cti
■ h2f_cti_clock
Debug APB Interface

The debug Advanced Peripheral Bus (APB™) interface allows debug components in
the FPGA fabric to debug components on the CoreSight™ debug APB.
■ h2f_debug_apb
■ h2f_debug_apb_sideband
■ h2f_debug_apb_reset
■ h2f_debug_apb_clock
Peripheral Signal Interfaces
DMA Controller Peripheral Request Interfaces

The DMA controller interface allows soft IP in the FPGA fabric to communicate with
the DMA controller in the HPS. You can configure up to eight separate interface
channels.
■ f2h_dma_req0—FPGA DMA controller peripheral request interface 0

Other Interfaces
Other Interfaces
MPU Standby and Event Interfaces

MPU standby and event signals are notification signals to the FPGA fabric that the
MPU is in standby. Event signals are used to wake up the Cortex-A9 processors from a
wait for event (WFE) state. The standby and event signals are included in the
following interfaces:
■ h2f_mpu_events—MPU standby and event interface, including the following
signals:
■ h2f_mpu_eventi—Sends an event from logic in the FPGA fabric to the MPU.
This FPGA-to-HPS signal is used to wake up a processor that is in a Wait For
Event state. Asserting this signal has the same effect as executing the SEV
instruction in the Cortex-A9. This signal must be de-asserted until the FPGA
fabric is powered-up and configured.
■ h2f_mpu_evento—Sends an event from the MPU to logic in the FPGA fabric.
This HPS-to-FPGA signal is asserted when an SEV instruction is executed by
one of the Cortex-A9 processors.
■ h2f_mpu_standbywfe[1:0]—Indicates whether each Cortex-A9 processor is in
the WFE state
■ h2f_mpu_standbywfi[1:0]—Indicates whether each Cortex-A9 processor is in
the wait for interrupt (WFI) state
■ h2f_mpu_gp—General purpose interface
The MPU provides signals to indicate when it is in a standby state. These signals are
available to custom hardware designs in the FPGA fabric.
f For more information, refer to the Cortex-A9 Microprocessor Unit Subsystem chapter in
FPGA-to-HPS Interrupts
You can configure the HPS component to provide 64 general-purpose FPGA-to-HPS
interrupts, allowing soft IP in the FPGA fabric to trigger interrupts to the MPU’s
generic interrupt controller (GIC). The interrupts are implemented through the
following 32-bit interfaces:
■ f2h_irq0—FPGA-to-HPS interrupts 0 through 31
■ f2h_irq1—FPGA-to-HPS interrupts 32 through 63
The FPGA-to-HPS interrupts are asynchronous on the FPGA interface. Inside the
HPS, the interrupts are synchronized to the MPU’s internal peripheral clock
(periphclk).

General-Purpose Interfaces
You can use the FPGA manager to supply the h2f_mpu_gp interface, which includes
the following general-purpose signals:
■ 32 FPGA-to-HPS signals
■ 32 HPS-to-FPGA signals
f For more information, refer to the FPGA Manager chapter in volume 3 of the Cyclone V
Device Handbook.


■ Added debug interfaces.
■ Updated HPS-to-FPGA reset interface names.
November 2012 1.1 ■ Updated HPS external reset source interface names.
■ Removed DMA peripheral interface clocks.
■ Referred to Altera Address Span Extender.

29. Simulating the HPS Component
November 2012
cv_54030-1.1
cv_54030-1.1
HPS Simulation Support

This section describes the simulation support for the hard processor system (HPS)
component. The HPS simulation models support interfaces between the HPS and
FPGA fabric, including:
■ Bus functional models (BFMs) for most interfaces between HPS and FPGA fabric
■ A simulation model for the HPS SDRAM memory
The HPS simulation support does not include modules implemented in the HPS, such
as the ARM® Cortex™-A9 MPCore processor.
You specify simulation support files when you instantiate the HPS component in the
Qsys system integration tool. When you enable a particular HPS-FPGA interface,
Qsys provides the corresponding model during the generation process. Refer to
“Simulation Flows” on page 29–10 for a description of the simulation flows.
f For general information about instantiating the component, refer to the Instantiating
the HPS Component chapter in volume 3 of the Cyclone® V Device Handbook.
The HPS simulation support enables you to develop and verify your own FPGA soft
logic or intellectual property (IP) that interfaces to the HPS component.
The simulation model supports the following interfaces:
■ Clock and reset interfaces
■ FPGA-to-HPS Advanced Microcontroller Bus Architecture (AMBA®) Advanced
eXtensible Interface (AXI™) slave interface
■ HPS-to-FPGA AXI master interface
■ Lightweight HPS-to-FPGA AXI master interface
■ FPGA-to-HPS SDRAM interface
■ Microprocessor unit (MPU) general-purpose I/O interface
■ MPU standby and event interface
■ Interrupts interface
■ Direct memory access (DMA) controller peripheral request interface
■ Debug Advanced Peripheral Bus (APB™) interface
■ System Trace Macrocell (STM) hardware event

November 2012
Subscribe
29–2 Chapter 29: Simulating the HPS Component
■ FPGA cross trigger interface

■ FPGA trace port interface
Figure 29–1 on page 29–2 shows the HPS with its BFMs.
Figure 29–1. HPS BFM Block Diagram
Hard Processor System

AXI Protocol Master
IP Components
AXI HPS Master BFM
AXI Protocol Slave

IP Components
AXI HPS Slave BFM
Qsys
Interconnect
SDRAM Memory Model
Avalon-MM
Interrupt Sink BFM IP Components
Conduit BFM
The HPS BFMs use standard function calls from the Altera® BFM application
programming interface (API), as detailed in the remainder of this section.
f For information about the BFM API, refer to the Avalon Verification IP Suite User Guide
and the Mentor Verification IP Altera Edition User Guide.
HPS simulation supports only Verilog HDL or SystemVerilog simulation

environments.
Clock and Reset Interfaces

f For general information about clock and reset interfaces, refer to “Memory-Mapped
Interfaces” in the HPS Component Interfaces chapter in volume 3 of the Cyclone V Device
Handbook.
Clock Interface
Qsys generates the clock source BFM for each clock output interface from the HPS
component. For HPS-to-FPGA user clocks, specify the BFM clock rate in the User
clock frequency field in the HPS Clocks page when instantiating the HPS component
in Qsys.
The HPS-to-FPGA trace port interface unit generates a clock output to the FPGA,
named h2f_tpiu_clock. In simulation, the clock source BFM also represents this clock
output’s behavior.

Chapter 29: Simulating the HPS Component 29–3
Table 29–1 lists all HPS clock output interfaces with the BFM instance name.
Table 29–1. HPS Clock Output Interface Simulation Model

Interface Name BFM Instance Name
h2f_user0_clock h2f_user0_clock
h2f_tpiu_clock h2f_tpiu_clock
The Altera clock source BFM application programming interface (API) applies to all
the BFMs listed in Table 29–1. Your Verilog interfaces use the same API, passing in
different instance names.
Qsys does not generate BFMs for FPGA-to-HPS clock input interfaces.
Reset Interface
The HPS reset request and handshake interfaces are connected to Altera conduit
BFMs for simulation. Table 29–2 lists the name of each interface. You can monitor the
reset request interface state changes or set the interface by using the API listed in
Table 29–2.
Table 29–2. HPS Reset Input Interface Simulation Model

Interface Name BFM Instance Name API Function Names
f2h_cold_reset_req f2h_cold_reset_req get_f2h_cold_rst_req_n()
f2h_debug_reset_req f2h_debug_reset_req get_f2h_dbg_rst_req_n()
f2h_warm_reset_req f2h_warm_reset_req get_f2h_warm_rst_req_n()
set_h2f_pending_rst_req_n()
h2f_warm_reset_handshake h2f_warm_reset_handshake
get_f2h_pending_rst_ack_n()
Table 29–3 lists all HPS reset output interfaces with the BFM instance name. The
Altera reset source BFM application programming interface applies to all the BFMs
listed in Table 29–3.
Table 29–3. HPS Reset Output Interface Simulation Model

h2f_reset h2f_reset
h2f_cold_reset h2f_cold_reset
h2f_debug_apb_reset h2f_debug_apb_reset

The HPS reset output interface is connected to a reset source BFM. Qsys configures the
BFM as shown in Table 29–4.
Table 29–4. Configuration of Reset Source BFM for HPS Reset Output Interface
Parameter BFM Value (1) Meaning
Assert reset high Off This parameter is off, specifying an active-low reset signal from the BFM.
This parameter is 0, specifying that the BFM does not assert the reset signal
Cycles of initial reset 0
automatically.
(1) The parameter value of the instantiated BFM as configured for HPS simulation
FPGA-to-HPS AXI Slave Interface

The FPGA-to-HPS AXI slave interface, f2h_axi_slave, is connected to a Mentor
Graphics AXI slave BFM for simulation. Qsys configures the BFM as shown in
Table 29–5. The BFM clock input is connected to f2h_axi_clock clock.
Table 29–5. Configuration of FPGA-to-HPS AXI Slave BFM

Parameter Value
AXI Address Width 32
AXI Read Data Width 32, 64, 128
AXI Write Data Width 32, 64, 128
AXI ID Width 8
You control and monitor the AXI slave BFM by using the BFM API.
f For more information, refer to the Mentor Verification IP Altera Edition User Guide. For
general information about FPGA-to-HPS AXI slave interfaces, refer to
“Memory-Mapped Interfaces” in the HPS Component Interfaces chapter in volume 3 of
HPS-to-FPGA AXI Master Interface

The HPS-to-FPGA AXI master interface, h2f_axi_master, is connected to a Mentor
Graphics AXI master BFM for simulation. Qsys configures the BFM as shown in
Table 29–6. The BFM clock input is connected to h2f_axi_clock clock.
Table 29–6. Configuration of HPS-to-FPGA AXI Master BFM

Parameter Value
AXI Read and Write Data Width 32, 64, 128
AXI ID Width 12
You control and monitor the AXI master BFM by using the BFM API.

general information about HPS-to-FPGA AXI master interfaces, refer to
Lightweight HPS-to-FPGA AXI Master Interface

The lightweight HPS-to-FPGA AXI master interface, h2f_lw_axi_master, is connected
to a Mentor Graphics AXI master BFM for simulation. Qsys configures the BFM as
shown in Table 29–7. The BFM clock input is connected to h2f_lw_axi_clock clock.
Table 29–7. Configuration of Lightweight HPS-to-FPGA AXI Master BFM

Parameter Value
AXI Read and Write Data Width 32
AXI ID Width 12
You control and monitor the AXI master BFM by using the BFM API.
general information about lightweight HPS-to-FPGA AXI master interfaces, refer to

The HPS component contains a memory interface simulation model to which all of the
FPGA-to-HPS SDRAM interfaces are connected. The model is based on the HPS
implementation and provides cycle-level accuracy, reflecting the true bandwidth and
latency of the interface. However, the model does not have the detailed configuration
provided by the HPS software, and hence does not reflect any inter-port scheduling
that might occur under contention on the real hardware when different priorities or
weights are used.
f For more information, refer to “EMI-Related HPS Features in SoC Devices” in the
Functional Description—Hard Memory Interface chapter in volume 3 of the External

HPS-to-FPGA MPU General-Purpose I/O Interface

The HPS-to-FPGA MPU general-purpose I/O interface is connected to an Altera
conduit BFM for simulation. Table 29–8 lists the name of each interface, along with
API function names for each type of simulation. You can monitor the interface state
changes or set the interface by using the API listed in Table 29–8.
Table 29–8. HPS-to-FPGA MPU General-Purpose I/O Interface Simulation Model

BFM Instance RTL Simulation API Function Post-Fit Simulation
Interface Name
Name Names API Function Names
set_h2f_mpu_gp_out() set_gp_out()
h2f_mpu_gp h2f_mpu_gp
get_h2f_mpu_gp_in() get_gp_in()
HPS-to-FPGA MPU Event Interface

The HPS-to-FPGA MPU event interface is connected to an Altera conduit BFM for
simulation. Table 29–9 lists the name of each interface, along with API function names
for each type of simulation. You can monitor the interface state changes or set the
interface by using the API listed in Table 29–9.
Table 29–9. HPS-to-FPGA MPU Event Interface Simulation Model

BFM Instance RTL Simulation API Function Post-Fit Simulation
Interface Name
Name Names (1) API Function Names
get_h2f_mpu_eventi() get_eventi()
set_h2f_mpu_evento() set_evento()
h2f_mpu_events h2f_mpu_events
set_h2f_mpu_standbywfe() set_standbywfe()
set_h2f_mpu_standbywfi() set_standbywfi()
(1) The usage of conduit get_*() and set_*() API functions is the same as with the general Avalon conduit BFM.
FPGA-to-HPS Interrupts Interface

The FPGA-to-HPS interrupts interface is connected to an Altera Avalon interrupt sink
BFM for simulation. Table 29–10 lists the name of each interface.
Table 29–10. FPGA-to-HPS Interrupts Interface Simulation Model

f2h_irq0 f2h_irq0
f2h_irq1 f2h_irq1
The Altera Avalon interrupt sink BFM API applies to all the BFMs listed in Table 29–3.

HPS-to-FPGA Debug APB Interface

The HPS-to-FPGA debug APB interface is connected to an Altera conduit BFM for
simulation. Table 29–11 lists the name of each interface, along with API function
names for each type of simulation. You can monitor the interface state changes or set
the interface by using the API functions listed in Table 29–11.
Table 29–11. HPS-to-FPGA Debug APB Interface Simulation Model

Post-Fit Simulation API
Interface Name BFM Name RTL Simulation API Function Names
Function Names
set_h2f_dbg_apb_PADDR() set_PADDR()
set_h2f_dbg_apb_PADDR_31() set_PADDR_31()
set_h2f_dbg_apb_PENABLE() set_PENABLE()
get_h2f_dbg_apb_PRDATA() get_PRDATA()
h2f_debug_apb h2f_debug_apb get_h2f_dbg_apb_PREADY() get_PREADY()
set_h2f_dbg_apb_PSEL() set_PSEL()
get_h2f_dbg_apb_PSLVERR() get_PSLVERR()
set_h2f_dbg_apb_PWDATA() set_PWDATA()
set_h2f_dbg_apb_PWRITE() set_PWRITE()
h2f_debug_apb_side h2f_debug_apb_side get_h2f_dbg_apb_PCLKEN() get_PCLKEN()
band band get_h2f_dbg_apb_DBG_APB_DISABLE() get_DBG_APB_DISABLE()
FPGA-to-HPS System Trace Macrocell (STM) Hardware Event Interface

The FPGA-to-HPS STM hardware event interface is connected to an Altera conduit
BFM for simulation. Table 29–12 lists the name of each interface, along with API
function name for each type of simulation. You can monitor the interface state
changes or set the interface by using the API functions listed in Table 29–12.
Table 29–12. FPGA-to-HPS STM Hardware Event Interface Simulation Model

Post-Fit Simulation API
Interface Name BFM Name RTL Simulation API Function Name
Function Name
f2h_stm_hw_events f2h_stm_hw_events get_f2h_stm_hwevents() get_stm_events()

HPS-to-FPGA Cross-Trigger Interface

The HPS-to-FPGA cross-trigger interface is connected to an Altera conduit BFM for
Table 29–13. HPS-to-FPGA Cross-Trigger Interface Simulation Model

Interface BFM Post-Fit Simulation API
RTL Simulation API Function Names
Name Name Function Names
get_h2f_cti_trig_in() get_trig_in()
set_h2f_cti_trig_in_ack() set_trig_inack()
set_h2f_cti_trig_out() set_trig_out()
h2f_cti h2f_cti
get_h2f_cti_trig_out_ack() get_trig_outack()
set_h2f_cti_asicctl() set_asicctl()
get_h2f_cti_fpga_clk_en() get_clk_en()
HPS-to-FPGA Trace Port Interface

The HPS-to-FPGA trace port interface is connected to an Altera conduit BFM for
Table 29–14. HPS-to-FPGA Trace Port Interface Simulation Model

Interface Post-Fit Simulation API
BFM Name RTL Simulation API Function Names
Name Function Names
get_h2f_tpiu_clk_ctl() get_traceclk_ctl()
h2f_tpiu h2f_tpiu
set_h2f_tpiu_data() set_trace_data()

FPGA-to-HPS DMA Handshake Interface

The FPGA-to-HPS DMA handshake interface is connected to an Altera conduit BFM
for simulation. Table 29–15 lists the name for each interface, along with API function
the interface by using the API listed in Table 29–15.
Table 29–15. FPGA-to-HPS DMA Handshake Interface Simulation Model

Interface BFM Instance RTL Simulation API Function Post-Fit Simulation API
Name Name Names (1) Function Names
get_f2h_dma_req0_req() get_channel0_req()
f2h_dma_req0 f2h_dma_req0 get_f2h_dma_req0_single() get_channel0_single()
set_f2h_dma_req0_ack() set_channel0_xx_ack()
(1) The usage of conduit get_*() and set_*() API functions is the same as with the general Avalon conduit BFM.

Simulation Flows
Simulation Flows
This section describes the simulation flows for an HPS-based design.
Altera provides both functional register transfer level (RTL) simulation and post-fitter
gate-level simulation flows. The simulation flows involve the following major steps:
1. Instantiate the HPS component—Refer to “Specifying HPS Simulation Model in
Qsys” on page 29–10.
2. Generate the system in Qsys, including the simulation model—Refer to
“Generating HPS Simulation Model in Qsys” on page 29–13.
3. Run the simulation—Refer to one of the following sections:
■ “Running HPS RTL Simulation” on page 29–13
■ “Running HPS Post-Fit Simulation” on page 29–14
f For general information about simulation, refer to the Simulating Altera Designs
chapter in volume 3 of the Quartus II Handbook.
Specifying HPS Simulation Model in Qsys

The following steps outline how to set up the HPS component for simulation.
1. Add the HPS component from the Qsys Component Library.
2. Configure the component based on your application needs by selecting or
deselecting the HPS-FPGA interfaces.
3. Connect the appropriate HPS interfaces to other components in the system. For
example, connect the FPGA-to-HPS AXI slave interface to an AXI master interface
in another component in the system.
When you create your component, make sure the conduit interfaces have the
correct role names, directions, and widths. Table 29–16 lists the role names,
directions, and widths for all the HPS conduit interfaces.
f For general information about adding the HPS component to your design, refer to the
Instantiating the HPS Component chapter in volume 3 of the Cyclone V Device Handbook.
Table 29–16. HPS Conduit Interfaces (Part 1 of 3)

Role Name Direction Width
h2f_warm_reset_handshake
h2f_pending_rst_req_n Output 1
f2h_pending_rst_ack_n Input 1
h2f_mpu_gp
gp_in Input 32
gp_out Output 32

Simulation Flows

h2f_mpu_events
eventi Input 1
evento Output 1
standbywfe Output 2
standbywfi Output 2
f2h_dma_req0
req0_req Input 1
req0_single Input 1
req0_ack Output 1
f2h_dma_req1
req1_req Input 1
req1_single Input 1
req1_ack Output 1
f2h_dma_req2
req2_req Input 1
req2_single Input 1
req2_ack Output 1
f2h_dma_req3
req3_req Input 1
req3_single Input 1
req3_ack Output 1
f2h_dma_req4
req4_req Input 1
req4_single Input 1
req4_ack Output 1
f2h_dma_req5
req5_req Input 1
req5_single Input 1
req5_ack Output 1
f2h_dma_req6
req6_req Input 1
req6_single Input 1
req6_ack Output 1
f2h_dma_req7
req7_req Input 1
req7_single Input 1
req7_ack Output 1

Simulation Flows

h2f_debug_apb
paddr Input 18
paddr_31 Input 1
penable Input 1
prdata Output 32
pready Output 1
psel Input 1
pslverr Output 1
pwdata Input 32
pwrite Input 1
h2f_debug_apb_sideband
pclken Output 1
dbg_apb_disable Output 1
f2h_stm_hw_events
stm_hwevents Output 28
h2f_cti
trig_in Output 8
trig_in_ack Input 8
trig_out Input 8
trig_out_ack Output 8
asicctl Input 8
fpga_clk_en Output 1
h2f_tpiu
clk_ctl Output 1
data Input 32

Simulation Flows
Generating HPS Simulation Model in Qsys

The following steps outline how to generate the simulation model.
1. Go to the Generation page in Qsys.
2. For RTL simulation, perform the following steps:
a. Set Create simulation model to Verilog.
b. Click Generate.
1 HPS simulation does not support the VHDL simulation environment.
For post-fit simulation, perform the following steps:

a. Turn on the Create HDL design files for synthesis option.
b. Turn on the Create block symbol file (.bsf) option.
c. Click Generate.
f For general information about generating the HPS component, refer to the
Instantiating the HPS Component chapter in volume 3 of the Cyclone V Device Handbook.
For more information about Qsys simulation, refer to “Simulating a Qsys System” in
the Creating a System with Qsys chapter in volume 1 of the Quartus II Handbook.
Running HPS RTL Simulation

Qsys generates scripts for various simulators that you use to complete the simulation
process. Table 29–17 lists simulation tools, script names and the script directories for
various vendors.
Table 29–17. Qsys-Generated Scripts for Various Simulators

Simulator Script Name Directory
Mentor Graphics Modelsim®
msim_setup.tcl <project directory>/<Qsys design name>/simulation/mentor
Altera Edition
Cadence NC-Sim ncsim_setup.sh <project directory>/<Qsys design name>/simulation/cadence
Synopsys VCS vcs_setup.sh <project directory>/<Qsys design name>/simulation/synopsys/vcs
Synopsys VCS-MX vcsmx_setup.sh <project directory>/<Qsys design name>/simulation/synopsys/vcsmx
f For detailed simulation steps, refer to the Mentor Verification IP Altera Edition User
Guide, and to the Qsys Tutorial chapter of the Avalon Verification IP Suite User Guide.

Simulation Flows
Running HPS Post-Fit Simulation

The section describes how you run HPS post-fit simulation. After successful Qsys
generation, perform the following steps:
1. Add the Qsys-generated synthesis file set to your Quartus II project by performing
the following steps:
a. In the Quartus II software, click Settings in the Assignments menu.
b. In the Settings - <your Qsys system name> dialog box, on the Files tab, browse
to <your project directory>/<your Qsys system name>/synthesis/ and select
<your Qsys system name>.qip.
c. Click OK.
2. You can instantiate the Qsys system that contains HPS component as your
Quartus II project top-level entity, if necessary.
3. Compile the design by clicking Start Compilation in the Processing menu.
4. Change the EDA Netlist Writer settings, if necessary, by performing the following
steps:
a. Click Settings in the Assignment menu.
b. On the Simulation tab, under the EDA Tool Settings tab, you can specify the
following EDA Netlist Writer settings:
■ Tool name—The name of the simulation tool
■ Format for output netlist
■ Output directory
c. Click OK.
h For more information about EDA Netlist Writer settings, refer to Simulation
Page (Settings Dialog Box) in Quartus II Help.
5. To create the post-fitter simulation model with Quartus II EDA Netlist Writer, in
the Start menu, point to Processing and click Start EDA Netlist Writer.
6. Compile the necessary simulation files in your simulation tool. Table 29–18 lists
the required libraries and files.
7. Start simulation.

Simulation Flows
Table 29–18. Post-Fit Simulation Files

Library Directory (1) File
verbosity_pkg.sv
Altera Verification IP
<Avalon Verification IP>/lib/ avalon_mm_pkg.sv
Library
avalon_utilities_pkg.sv
Avalon Clock Source BFM <Avalon Verification IP>/altera_avalon_clock_source/ altera_avalon_clock_source.sv
Avalon Reset Source BFM <Avalon Verification IP>/altera_avalon_reset_source/ altera_avalon_reset_source.sv
Avalon MM Slave BFM <Avalon Verification IP>/altera_avalon_mm_slave_bfm/ altera_avalon_mm_slave_bfm.sv
Avalon Interrupt Sink BFM <Avalon Verification IP>/altera_avalon_interrupt_sink/ altera_avalon_interrupt_sink.sv
Mentor AXI Verification IP
<AXI Verification IP>/common/ questa_mvc_svapi.svh
Library
mgc_common_axi.sv
Mentor AXI3 BFM <AXI Verification IP>/axi3/bfm/ mgc_axi_master.sv
mgc_axi_slave.sv
HPS Post-Fit Simulation
<HPS Post-fit Sim>/ All the files in the directory
Library
altera_primitives.v
220model.v
sgate.v
altera_mf.v
Device Simulation
<Device Sim Lib>/ altera_lnsim.sv
Library (2)
cyclonev_atoms.v
arriav_atoms.v
mentor/cyclonev_atoms_ncrypt.v
mentor/arriav_atoms_ncrypt.v
EDA Netlist Writer *.vo
Generated Post-Fit <User project directory>/
*.vho (3)
Simulation Model
*.v
User testbench files <User project directory>/ *.sv
*.vhd (3)
(1) <ACDS install> = Altera Complete Design Suite installation path
<Avalon Verification IP> = <ACDS install>/ip/altera/sopc_builder_ip/verification
<AXI Verification IP> = <ACDS install>/ip/altera/mentor_vip_ae
<HPS Post-fit Sim> = <ACDS install>/ip/altera/hps/postfitter_simulation
<Device Sim Lib> = <ACDS install>/quartus/eda/sim_lib
(2) The device simulation library is not needed with Modelsim-Altera.
(3) Mixed-language simulator is needed for Verilog HDL and VHDL mixed design

For post-fit simulation, you must call the BFM API in your test program with a
specific hierarchy. The hierarchy format is:
<DUT>.\<HPS>|fpga_interfaces|<interface><space>.<BFM>.<API function>
Where:
■ <DUT> is the instance name of the design under test that you instantiated in your
test bench that consists of the HPS component.
■ <HPS> is the HPS component instance name that you use in your Qsys system.
■ <interface> is the instance name for a specific FPGA-to-HPS or HPS-to-FPGA
interface. This name can be found in the fpga_interfaces.sv file located in <project
directory>/<Qsys design name>/synthesis/submodules.
■ <space>—You must insert one space character after the interface instance name.
■ <BFM> is the BFM instance name. In
<ACDS install>/ip/altera/hps/postfitter_simulation, identify the SystemVerilog
file corresponding to the interface type that you are using. The SystemVerilog file
contains the BFM instance name.
For example, a path for the Lightweight HPS-to-FPGA master interface hierarchy
could be formed as follows:
top.dut.\my_hps_component|fpga_interface|hps2fpga_light_weight .h2f_lw_axi_master
Notice the space after “hps2fpga_light_weight”. Omitting this space would cause
simulation failure because the instance name “hps2fpga_light_weight ”, including the
space, is the name used in the post-fit simulation model generated by the Quartus® II
software.


■ Added debug APB, STM hardware event, FPGA cross trigger, FPGA trace port interfaces.
■ Added support for post-fit simulation.
November 2012 1.1
■ Updated some API function names.
■ Removed DMA peripheral clock.

Section VIII. Appendices
This section includes the following appendices:

■ Appendix A, Booting and Configuration

VIII–2 Section VIII: Appendices

A. Booting and Configuration
November 2012
cv_5400A-1.3
cv_5400A-1.3
This appendix describes the booting of the hard processor system (HPS) and the
configuration of the FPGA portion of the Altera system-on-a-chip (SoC) FPGA device.
The HPS boot starts when a processor is released from reset (for example, on power
up) and executes code in the internal boot ROM at the reset exception address. The
boot process ends when the code in the boot ROM jumps to the next stage of the boot
software. This next stage of boot software is referred to as the preloader. The
preloader can be customized and is typically stored external to the HPS in a
nonvolatile flash-based memory.
The processor can boot from the following sources:
■ NAND flash memory through the NAND flash controller
■ Secure Digital/MultiMediaCard (SD/MMC) flash memory through the SD/MMC
flash controller
■ Serial peripheral interface (SPI) and quad SPI flash memory through the quad SPI
flash controller
■ FPGA fabric
The HPS boot supports indirect or direct execution of the preloader depending on the
boot device. With indirect execution, the boot ROM code copies the preloader from
the boot device into the on-chip RAM and jumps to it. Indirect execution is used for
flash memory boot sources. With direct execution, the boot ROM code executes the
preloader directly from the boot device or from the FPGA fabric boot source.
Configuration of the FPGA portion of the device starts when the FPGA portion is
released from the reset state (for example, on power-on). The control block (CB) in the
FPGA portion of the device is responsible for obtaining an FPGA configuration image
and configuring the FPGA. The FPGA configuration ends when the configuration
image has been fully loaded and the FPGA enters user mode. The FPGA
configuration image is provided by users and is typically stored in non-volatile flash-
based memory. The FPGA CB can obtain a configuration image from the HPS through
the FPGA manager or from any of the sources supported by the Cyclone® V FPGAs
family.
f For more information about the memory and peripheral modules used during the
booting and configuration process, refer to their respective chapters:
■ On-Chip Memory chapter in volume 3 of the Cyclone V Device Handbook.

■ FPGA Manager chapter in volume 3 of the Cyclone V Device Handbook.
The following three figures illustrate the possible HPS boot and FPGA configuration
schemes. The arrows in the figures denote the data flow direction.

November 2012
Subscribe
A–2 Appendix A: Booting and Configuration
■ Independent
Figure A–1 shows that the FPGA configuration and HPS boot occur
independently. The FPGA configuration obtains its configuration image from a
non-HPS source, while the HPS boot obtains its preloader from a non-FPGA fabric
source.
Figure A–1. Independent FPGA Configuration and HPS Booting
PCIe

FPGA Portion HPS Portion
Quad SPI
Flash Controller
Configuration Quad SPI / MPU
Sources SPI Flash SD/MMC Boot
Controller Flash Controller Sources
FPGA
Passive Fabric
NAND
Serial Flash Controller
Passive Boot On-Chip

Parallel ROM RAM
■ FPGA first
Figure A–2 shows that the FPGA is first configured through one of its non-HPS
configuration sources and then the HPS boots from the FPGA fabric. The HPS boot
waits for the FPGA fabric to be powered on and in user mode before executing.
The HPS boot ROM code executes the preloader from the FPGA fabric over the
HPS-to-FPGA bridge. The preloader can be obtained from the FPGA RAM or by
accessing an external interface, depending on your design and implementation.
Figure A–2. FPGA Configures First
PCIe

Boot & MPU

Configuration Quad SPI /
Sources SPI Flash
Controller
FPGA
Passive Fabric HPS-to-FPGA
Serial Bridge
Passive Boot
Parallel ROM

Appendix A: Booting and Configuration A–3
HPS Boot
■ HPS first
Figure A–3 shows that the HPS boots first through one of its non-FPGA fabric boot
sources and then software running on the HPS configures the FPGA fabric
through the HPS FPGA manager. The software on the HPS obtains the FPGA
configuration image from any of its flash memory devices or communication
interfaces, for example, the Ethernet media access controller (EMAC). The
software is provided by users and the boot ROM is not involved in configuring the
FPGA fabric.
Figure A–3. HPS Boots First

Quad SPI
Flash Controller
MPU
SD/MMC Boot
Flash Controller Sources
FPGA
FPGA
Fabric Manager
NAND
Flash Controller
EMAC Configuration Source
HPS Boot
Booting software on the HPS is a multi-stage process. Each stage is responsible for
loading the next stage. The first software stage is the boot ROM. The boot ROM code
locates and executes the second stage software, called the preloader. The preloader
locates, and if present, executes the next stage software. The preloader and
subsequent boot stages (if present) are collectively referred to as user software.
Only the boot ROM code is located in the HPS. The user software is located external to
the HPS and is provided by users. The boot ROM code is only aware of the preloader
and not aware of any potential subsequent boot stages.
Figure A–4 illustrates the typical boot flow. There may be more or less boot stages in
the user software than shown and the roles of the boot stages may vary.
Figure A–4. Typical Boot Flow
User Software
Operating
Reset Boot ROM Preloader Boot Loader Application
System

HPS Boot
Boot Process Overview

Reset
The boot process begins when a CPU in the MPU exits from the reset state. When a
CPU exits from reset, it starts running code at the reset exception address. In normal
operation, the boot ROM is mapped at the reset exception address so code starts
running in the boot ROM.
It is possible to map the on-chip RAM or SDRAM at the reset exception address and
run code other than the boot ROM code. However, this chapter assumes that the boot
ROM maps to the reset exception address.
Boot ROM
The boot ROM contains software code that executes after a reset exits. The boot ROM
code determines if it is running on CPU0 or CPU1. If running on CPU1, the boot ROM
code jumps to the value in the CPU1 start address register (cpu1startaddr) in the boot
ROM code register group (romcodegrp) in the system manager. If running on CPU0,
the boot ROM code reads the boot select (BSEL) and clock select (CSEL) values from
the bsel and csel fields of the boot information register (bootinfo) in the system
manager to determine the boot source and to set up the clock manager. The bsel and
csel field values come from the BSEL and CSEL pins. The system manager samples the
values on these pins when coming out of a reset.
Table A–1 lists the bsel field values for each flash memory device selection.
Table A–1. bsel Field Values and Flash Device Selection

bsel Field Value Flash Device
0x0 Reserved
0x1 FPGA (HPS-to-FPGA bridge)
0x2 1.8 V NAND flash memory
0x3 3.0 V NAND flash memory
0x4 1.8 V SD/MMC flash memory with external transceiver
0x5 3.0 V SD/MMC flash memory with internal transceiver
0x6 1.8 V SPI or quad SPI flash memory
0x7 3.0 V SPI or quad SPI flash memory
For indirect execution from flash memory boot sources, the boot ROM code loads the
preloader image from the flash device to the on-chip RAM and passes software
control to the preloader in the on-chip RAM. For direct execution from the FPGA
fabric boot source, the boot ROM code waits until the FPGA portion of the device is in
user mode and then passes software control to the preloader in the FPGA RAM.
Preloader
The function of the preloader is user-defined. However, typical functions include
initializing the SDRAM interface and configuring the HPS I/O pins. Initializing the
SDRAM allows the preloader to load the next stage of the boot software (that might
not fit in the 60 kilobytes (KB) available in the on-chip RAM). A typical next software
stage is the open source boot loader, U-boot.

HPS Boot
The preloader is allowed to load the next stage boot software from any device
available to the HPS. Typical sources include the same flash device that contains the
preloader, a different flash device, or a communication interface such as an EMAC.
Boot Loader
The boot loader loads the operating system and passes software control to the
operating system.
Boot ROM
The function of the boot ROM code is to determine the boot source, initialize the HPS
after a reset, and jump to the preloader. In the case of indirect execution, the boot
ROM code loads the preloader image from the flash memory to on-chip RAM. The
boot ROM performs the following actions to initialize the HPS:
Enable instruction cache, branch predictor, floating point unit, NEON vector unit
■ Sets up the level 4 (l4) watchdog 0 timer
■ Configures the main PLL and peripheral PLL based on the CSEL value
■ Configures I/O elements and pin multiplexing based on the BSEL value
■ Initializes the flash controller to default settings
While executing, the boot ROM code uses the top 4 KB of the on-chip RAM as data
workspace. This area is reserved for the boot ROM code after a reset until the boot
ROM code passes software control to preloader. This limits the maximum size of the
preloader for indirect execution to 60 KB.
Boot ROM Flow

This section describes the software flow from reset until the boot ROM code passes
software control to the preloader.

HPS Boot
Figure A–5 illustrates the boot ROM flow. The boot ROM code can perform a warm
boot from on-chip RAM, a cold boot from the FPGA portion of the device, or a cold
boot from flash memory.
Figure A–5. Boot ROM Flow
Reset
no
CPU 0?
yes
Warm Boot no Cold Boot no Cold Boot from

from On-Chip from
Flash Memory
RAM? FPGA?
yes yes
Valid
Preloader no Fallback no
Image Image Perform Post-
Found? Found? Mortem Dump
yes yes
Execute Code Execute Second Execute Fallback Jump to CPU1

Wait for Reset
in FPGA Stage Boot Loader Image Boot Code
Valid
Preloader no Load Preloader
Image? Image from Flash Memory
yes
Valid
Preloader no Fallback no
Image Perform Post-
Image
Found? Mortem Dump
Found?
yes yes
Execute Code Execute Second Execute Fallback

Wait for Reset
in On-Chip RAM Stage Boot Loader Image
During a cold boot from the FPGA portion of the device, the boot ROM code waits
until the FPGA is ready and then attempts direct execution at address 0x0 across the
HPS-to-FPGA bridge. For example, the boot software could be provided by initialized
on-chip RAM in the FPGA portion of the device at address 0x0. During a cold boot
from flash memory, the boot ROM code attempts to load the first preloader image
from flash memory to on-chip RAM and pass control to the preloader. If the image is
invalid, the boot ROM code attempts to load up to three subsequent images from flash
memory. If there is still no valid image found after the subsequent loads, the boot
ROM code checks the FPGA portion of the device for a fallback image.

HPS Boot
During a warm boot from on-chip RAM, the boot ROM code reads the preloader state
register (initswstate) in the romcodegrp group in the system manager to determine if
there is a valid preloader image in the on-chip RAM. If a valid preloader image is
found in the on-chip RAM, the boot ROM code skips loading a preloader image from
flash memory and instead passes control to the preloader residing in the on-chip
RAM.
If a valid preloader image cannot be found in the on-chip RAM, boot ROM code
attempts to load the last valid preloader image loaded from the flash memory,
identified by the index field of the initial software last image loaded register
(initswlastld) in the romcodegrp group in the system manager. If the image is
invalid, boot ROM code attempts to load up to three subsequent images from flash
memory. If a valid preloader image cannot be found in the on-chip RAM or flash
memory, the boot ROM code checks the FPGA portion of the device for a fallback
image.
Loading the Preloader

For indirect execution, the boot ROM code loads the preloader image from flash
memory into the on-chip RAM and passes control to the preloader. The boot ROM
code checks for a valid image by verifying the header and cyclic redundancy check
(CRC) in the preloader image. Figure A–6 shows the preloader header.
The boot ROM code checks the header for the following information:
■ Validation word—validates the preloader image. The validation word has a fixed
value of 0x31305341.
■ Version—indicates the preloader version.
■ Flags—unused
■ Program length—the total length of the image (in bytes) from offset 0x0 to the end
of code area, including exception vectors and CRC.
■ Checksum—a checksum of all the bytes in the header, from offset 0x40 to 0x49.
The preloader image has a maximum size of 60 KB. This size is limited by the on-chip
RAM size of 64 KB, where 4 KB is reserved as a workspace for the boot ROM data and
stack. The preloader can use this 4 KB region (for its stack and data, for example) after
the boot ROM code passes control to the preloader. This 4 KB region is overwritten by
the boot ROM code on a subsequent reset.

HPS Boot
Figure A–6 shows the preloader image layout in the on-chip RAM after being loaded
from the boot ROM.
Figure A–6. Preloader Image Layout (1)
0xFFFFFFFF
0x4C
Reserved at Reset Checksum
0x49
0xFFFFF000
Program Length
CRC
0x46
Flags
User-Defined Code 0x45
Version
0xFFFF0050 0x44
Entry Point Validation Word
0xFFFF004C 0x40
Header
0xFFFF0040
Exception Vectors
0xFFFF0000
Note to Figure A–6:

(1) Addresses are not represented to scale.
Exception vectors—Exception vectors are located at the start of the on-chip RAM.
Typically, the preloader remaps the lowest region of the memory map to the on-chip
RAM (from the boot ROM) to create easier access to the exception vectors.
Header—contains information such as validation word, version, flags, program
length, and checksum for the boot ROM code to validate the preloader image before
passing control to the preloader.
Entry point—contains the preloader image address. After the boot ROM code
validates the header, the boot ROM code jumps to this address.
User-defined code—typically contains the program code of the preloader.
CRC—contains a CRC of data from address 0xFFFF0000 to
0xFFFF0000+(Program Length*4)–0x0004. The polynomial used to validate the
preloader image is x32+ x26+ x23+ x22+ x16+ x12+ x11+ x10+ x8+ x7+ x5+ x4+ x2+ x + 1.
There is no reflection of the bits. The initial value of the remainder is 0xFFFFFFFF and
the final value is XORed with 0xFFFFFFFF.
Reserved at reset—the top 4 KB is reserved for the boot ROM code after a reset. The
boot ROM code uses this area for internal structures, workspace, and post-mortem
dump. This area includes the shared memory where the boot ROM code passes
information to the preloader.
Shared Memory
The shared memory contains information that the boot ROM code passes to the
preloader. The boot ROM code passes the location of shared memory to the preloader
in register r0, as described in “HPS State on Entry to the Preloader” on page A–10.

HPS Boot
The shared memory contains the following information:

■ Common—contains non-flash-specific settings used by the boot ROM code.
■ Saved hardware register—contains hardware register values that the boot ROM
code saves before the registers are modified by the boot ROM code.
■ Flash device-specific—contains flash device-specific settings used by the boot
ROM code that the preloader may use to continue using the flash device without
reinitialization.
Table A–2 lists the contents of the shared memory block.
Table A–2. Shared Memory Block (Part 1 of 2)

Information
Content Description
Type
Indicates which preloader image (0-3) the boot ROM code loaded from the flash
Flash image device. A nonzero value indicates that there was an error loading image 0 and the
boot ROM code loaded another image.
Indicates the CSEL value used by the boot ROM code. Typically, this value is the value
read from the csel field of the bootinfo register in the romcodegrp group in the
CSEL value used
system manager. However, if the PLL fails to lock, the boot ROM code ignores the
csel field and uses zero to indicate PLL bypass mode.
Indicates the BSEL value used by the boot ROM code. Typically, this value is the value
BSEL value used read from the bsel field of the bootinfo register in the romcodegrp group in the
system manager.
Last page Indicates the address of the last page read from the flash device.
Indicates the page size used by the flash device.
Common
■ For NAND flash memory, the boot ROM code reads this value from the NAND flash
controller.
Page size ■ For SPI and quad SPI flash memory, the boot ROM code configures the page size
through the quad SPI flash controller.
■ For SD/MMC flash memory, the boot ROM code configures the page size through
the SD/MMC flash controller.
Flash device type Indicates the flash device used by the boot ROM code.
Track the completion state of up to 64 individual major steps during the boot
Step complete process. The 64 bit value has one bit set for each major step completed in the boot
process.
CPU0 and CPU1 Indicates the CPU0 and CPU1 crash data. These values are pointers to where the
crash data boot ROM code saves the crash dump in an event of a crash.
Status register (stat)
Contains reset source and event timeout information.
of the reset manager
Control (ctrl)
register in the
Saved Contains information used to control the boot ROM code.
romcodegrp group in
hardware
the system manager
registers
initswstate
register in the
Contains the magic value 0x49535756 when the preloader has reached a valid state.
romcodegrp group in
the system manager

HPS Boot
Table A–2. Shared Memory Block (Part 2 of 2)

Information
Content Description
Type
is_sd_card Indicates the card type. 1 indicates SD card; 0 indicates MMC.
Indicates the addressing mode of card. 1 indicates sector addressing mode; 0
Flash is_sector_mode
indicates byte addressing mode.
device-
specific Contains the relative card address, which is used for host-card communication
rca
during card identification
(SD/MMC)
partition_start_sector Indicates the partition start offset in unit sectors. 0 indicates raw mode.
partition_size Size of the partition in unit sectors. 0 indicates raw mode.
L4 Watchdog 0 Timer
The L4 watchdog 0 timer is reserved for boot ROM use. If a watchdog reset happens
before software control passes to the preloader, boot ROM code attempts to load the
last valid preloader image, identified by the initswlastld register in the romcodegrp
group in the system manager.
HPS State on Entry to the Preloader

When the boot ROM code is ready to pass control to the preloader, the processor
(CPU0) is in the following state:
■ Instruction cache is enabled
■ Branch predictor is enabled
■ Data cache is disabled
■ MMU is disabled
■ Floating point unit is enabled
■ NEON vector unit enabled
■ Processor is in ARM secure supervisor mode
The boot ROM code sets the ARM® Cortex™-A9 MPCore™ registers to the following
values:
■ r0—contains the pointer to the shared memory block, which is used to pass
information from the boot ROM code to the preloader. The shared memory block
is located in the top 4 KB of on-chip RAM.
■ r1—contains the length of the shared memory.
■ r3—unused and set to 0x0.
■ r4—unused and set to 0x0.
All other MPCore registers are undefined.
1 When booting CPU0 using the FPGA boot, or when booting CPU1 using any boot
source, all MPCore registers, caches, the MMU, the floating point unit, and the NEON
vector unit are undefined. HPS subsystems and the PLLs are undefined.

HPS Boot
When the boot ROM code passes control to the preloader, the following conditions
also exist:
■ The boot ROM is still mapped to address 0x0.
■ The L4 watchdog 0 timer is active.
Preloader
The preloader typically performs the following actions:
■ Initialize the SDRAM interface.
■ Configure the remap register of the L3 (NIC-301) GPV registers (l3regs) to map the
on-chip RAM to address 0x0 so that exceptions are handled by the preloader.
1 The on-chip RAM is also accessible at address 0xFFFF0000. The address 0x0
is an alias.
■ Configure the HPS I/O through the scan manager.

■ Configure pin multiplexing through the system manager.
■ Configure HPS clocks through the clock manager.
■ Initialize the flash controller (NAND, SD/MMC, or quad SPI) that contains the
next stage boot software.
■ Load the next stage boot software into the SDRAM and pass control to it.
Typical Preloader Boot Flow

This section describes a typical software flow from the preloader entry point until the
software passes control to the next stage boot software.

HPS Boot
Figure A–7 shows a typical preloader boot flow.
Figure A–7. Typical Preloader Boot Flow
Preloader Entry
Timer & UART Initialization
Low-Level Initialization
SDRAM Interface Initialization (Include
Calibration & PLL Configuration)
Ensure All HPS I/O Banks Are Frozen

by the Freeze Controller
no
Success?
Assert Reset to Affected Peripherals/
Components during PLL Reconfiguration
yes
Clock Reconfiguration
Next Stage Boot Device Initialization
Initiate Scan Manager to Configure

Checking Boot Image’s Checksum
the HPS I/O CSR
(Optional)
Configure Pin Multiplexing through

the System Manager
Checksum no
Passed?
Reset Deassertion through the
Reset Manager
yes
Thaw (Unfreeze) All HPS I/O Banks Copy the Next Stage Boot Image from the
through the Freeze Controller Next Stage Boot Device to the SDRAM
Write Magic Value to the Initial

L3/L4 Interconnect Configuration Software State Register
Pass Control to Next Boot Error Handler

Stage Software in SDRAM
Low-level initialization steps include reconfiguring or disabling the L4 watchdog 0

timer, invalidating the instruction cache and branch predictor, remapping the on-chip
RAM to the lowest memory region, and setting up the data area.
Upon entering the preloader, the L4 watchdog 0 timer is active. The preloader can
either disable, reconfigure, or leave the watchdog timer unchanged. Once enabled
after reset, the watchdog timer cannot be disabled, only paused.
The instruction cache and branch predictor, which were previously enabled by the
boot ROM code, need to be invalidated.

HPS Boot
The preloader needs to remap the exception vector table because the exception vectors
are still pointing to the exception handler in the boot ROM when the preloader starts
executing. By setting the L3 interconnect remap bit 0 to high, the on-chip RAM
mirrors to the lowest region of the memory map. After this remap, the exception
vectors will use the exception handlers in the preloader image.
Figure A–8 shows the memory map before and after remap.
Figure A–8. Remapping the On-Chip RAM (1)
Boot ROM Preloader

0xFFFF_FFFF 0xFFFF_FFFF
On-Chip RAM On-Chip RAM
0xFFFF_0000 0xFFFF_0000
... ...
0xFFFD_FFFF 0xFFFD_FFFF
On-Chip ROM On-Chip ROM
0xFFFD_0000 0xFFFD_0000
... ...
0xBFFF_FFFF 0xBFFF_FFFF
SDRAM SDRAM
0x0010_0000 0x0010_0000
Unused Unused
0x0001_0000 0x0001_0000
Boot ROM On-Chip RAM
0x0 0x0
Before After
Note to Figure A–8:

(1) Addresses are not represented to scale.
The preloader can reconfigure all HPS clocks. During clock reconfiguration, the
preloader asserts reset to the peripherals in the HPS affected by the clock changes.
The preloader configures HPS I/O pins through the scan manager and pin
multiplexing through the system manager. The preloader initiates the freeze
controller in the scan manager to freeze all the I/O pins and put them in a safe state
during I/O configuration and pin multiplexing.
The SDRAM goes through full initialization for cold boot or a partial initialization for
warm boot. For full initialization, the preloader configures the SDRAM PLL before
releasing the SDRAM interface from reset. SDRAM calibration adjusts I/O delays and
FIFO settings to compensate for any board skew or impairment in the board, FPGA
portion of the device, or memory device. For partial initialization, SDRAM PLL
configuration and SDRAM calibration is not necessary.
The preloader looks for a valid next stage boot image in the next stage boot device by
checking the boot image validation data and checksum in the mirror image. Once
validated, the preloader copies the next stage boot image from the next stage boot
device to the SDRAM.
Before software passes control to the next stage boot software, the preloader can write
a valid value (such as 0x49535756) to the preloader initswstate register under the
romcodegrp group in the system manager. This value indicates that there is a valid
boot image in the on-chip RAM. When a warm reset occurs, the boot ROM code can
check the initswstate register for the magic value to determine if it needs to reload
the preloader image into the on-chip RAM.

HPS Boot
Flash Memory Devices

The flash memory devices available for HPS boot are the NAND, SD/MMC, SPI, and
quad SPI. The flash device can store the following kinds of files for booting purposes:
■ Preloader binary file (up to four copies)
■ Boot loader binary file
■ Operating system binary file
■ Application file
■ FPGA programming files
1 The preloader file must be stored in a partition with no file system.
NAND Flash Devices

Figure A–9 shows the NAND flash image layout. The preloader image is located at
offsets which are multiples of the block size. If the image size is less than 64 KB, only
one block size is used. Since a block is the smallest area used for erase operation, any
update to a particular image does not affect other images.
Figure A–9. NAND Flash Image Layout
Preloader Image 3
Preloader Image 2
Multiple of Block Size
Preloader Image 1
Preloader Image 0
0x0
Table A–3 lists the NAND flash driver features supported in the boot ROM code.
Table A–3. NAND Flash Support Features

Feature Driver Support
Open NAND Flash Interface (ONFI) 1.0 raw NAND or electronic signature
Device devices, single layer cell (SLC) and multiple layer cell (MLC) devices with
integrated error correction code (ECC).
CS0 only. Only CS0 is available to the HPS, the other three chip selects are
Chip select
routed out to the FPGA portion of the device.
Bus width x8 only
Page size 512 bytes, 2 KB, 4 KB, or 8 KB.
Page per block 8, 16, 32, or 128
ECC 512-bytes with 8-bit correction

HPS Boot
Table A–4 lists the CSEL pin settings for the quad SPI controller.
Table A–4. NAND Controller CSEL Pin Settings

CSEL Pin
Setting
0 1 2 3
osc1_clk (EOSC1 pin) range 10–50 MHz 10–12.5 MHz 12.5–25 MHz 25–50 MHz
osc1_clk/25, osc1_clk*20/25, osc1_clk*10/25, osc1_clk*5/25,
nand_x_clk /25 device frequency
2 MHz max 9.6 MHz max 9.6 MHz max 9.6 MHz max
osc1_clk, osc1_clk*20, osc1_clk*10, osc1_clk*5,
nand_x_clk controller clock
50 MHz max 240 MHz max 240 MHz max 240 MHz max
mpu_clk
PLL modes Bypassed Locked Locked Locked
f For more information about the NAND flash memory, refer to the NAND Flash
SD/MMC Flash Devices

Figure A–10 shows the SD/MMC flash image layout. The master boot record (MBR) is
located at the first 512 bytes of the memory. The MBR contains information of
partitions (address and size of partition). The preloader image is always stored in
partition A2. Partition A2 is a custom raw partition with no file system.
The start address of each image is based on the following formula:
Start address = start address + (<n> * 64 K), where <n> is the image number
Figure A–10. SD/MMC Flash Image Layout
Preloader Image 3
Preloader Image 2 Partition Type: A2

Partition Size: 64 KB x 4
Preloader Image 1
Preloader Image 0
...
Master Boot Record (MBR) MBR Partition Size: 512 Bytes

0x0
The SD/MMC controller supports two booting modes:

■ MBR (partition) mode
■ The boot image is read from a custom partition (0xA2)
■ The first image is located at the beginning of the partition, at offset 0x0
■ Start address = partition start address

HPS Boot
■ Raw mode
■ If the MBR signature is not found, SD/MMC driver assumes it is in raw mode.
The boot image data is read directly from sectors in the user area and is located
at the first sector of the SD/MMC.
■ The first image is located at the start of the memory card, at offset 0
■ Start address = 0
MBR
The MBR contains the partition table, which is always located on the first sector
(LBA0) with a memory size of 512 bytes. The MBR consists of executable code, four
partition entries, and the MBR signature. A MBR can be created by specific tools like
the FDISK program.
Table A–5 lists the MBR structure.
Table A–5. MBR Structure

Offset Size (Byte) Description
0x000 446 Code area
0x1BE 16 Partition entry for partition 1
0x1CE 16 Partition entry for partition 2
0x1DE 16 Partition entry for partition 3
0x1EE 16 Partition entry for partition 4
0x1FE 2 MBR signature: 0xAA55
The standard MBR structure contains a partition with four 16-bytes entries. Thus,
memory cards using this standard table cannot have more than four primary
partitions or up to three primary partitions and one extended partition.
Each partition type is defined by the partition entry. The boot images are stored in a
primary partition with custom partition type (0xA2). The SD/MMC flash driver does
not support file system, so the boot images are located in partition A2 at fixed
locations. Table A–6 lists the partition entry.
Table A–6. Partition Entry

Offset Size (Byte) Description
0x0 1 Boot indicator. 0x80 indicates that is it bootable.
0x1 3 Starting CHS value
0x4 1 Partition type
0x5 3 Ending CHS value
0x8 4 LBA of first section in partition
0xB 4 Number of sectors in partition

HPS Boot
The boot ROM code configures the SD/MMC controller to default settings for the
supported SD/MMC flash memory. Table A–7 lists the default settings of the
SD/MMC controller.
Table A–7. SD/MMC Controller Default Settings

Parameter Default Register Value
The card type register (ctype) in the
Card type 1 bit
SD/MMC controller registers (sdmmc) = 0x0
Timeout Maximum The timeout register (tmout) = 0xFFFFFFFF
The RX watermark level field (rx_wmark) of
FIFO threshold RX watermark level 1 the FIFO threshold watermark register
(fifoth) = 0x1
Clock source 0 The clock source register (clksrc) = 0x0
Block size 512 The block size register (blksiz) = 0x200
Identification The clock divider register (clkdiv)= 0x10
32
mode (2*16=32)
Clock divider
Data transfer
Bypass The clock divider register (clkdiv)= 0x00
mode
Table A–8 lists the CSEL pin settings for the SD/MMC controller.
Table A–8. SD/MMC Controller CSEL Pin Settings

CSEL Pin
Setting
0 1 2 3
osc1_clk (EOSC1 pin) range 10–50 MHz 10–12.5 MHz 12.5–25 MHz 25–50 MHz
osc1_clk/128, osc1_clk/32, osc1_clk/64, osc1_clk/128,
sdmmc_cclk_out device clock
ID mode 391 KHz max 391 KHz max 391 KHz max 391 KHz max
Controller baud rate divisor 32 32 32 32
osc1_clk/4, osc1_clk*1, osc1_clk/2, osc1_clk/4,
sdmmc_cclk_out device clock
Data transfer 12.5 MHz max 12.5 MHz max 12.5 MHz max 12.5 MHz max
mode Controller baud rate divisor
1 (bypass) 1 (bypass) 1 (bypass) 1 (bypass)
(even numbers only)
osc1_clk, osc1_clk, osc1_clk, osc1_clk*2,
sdmmc_clk controller clock:
mpu_clk
f For more information about the SD/MMC, refer to the SD/MMC Controller chapter in
SPI and Quad SPI Flash Devices

Figure A–11 shows the SPI and quad SPI flash image layout. The preloader image is
always located at offsets which are multiples of a subsector size. If the image is less
than 64 KB, only one subsector is used. If subsector erase is not supported, sector
erase is supported. Since a subsector is the smallest area used for erase operation, any
update to a particular image does not affect other images.

HPS Boot
The first image is located at offset 0 and followed by subsequent images. The start
address of each image is based on the following formula:
Start address = (N * 64K), where N is the image number
Figure A–11. SPI and Quad SPI Flash Image Layout
Preloader Image 3
Preloader Image 2
Multiple of Subsector
Preloader Image 1
Preloader Image 0
0x0
The boot ROM code configures the quad SPI controller to default settings for the
supported SPI or quad SPI flash memory. Table A–9 lists the default settings of the
quad SPI controller.
Table A–9. Quad SPI Controller Default Settings

Parameter Default Setting Register Value
The master mode baud rate divisor field (bauddiv) of
SPI baud rate. Divide by 4 the quad SPI configuration register (cfg) in the quad
SPI controller registers (qspiregs) = 1.
The read opcode in non-XIP mode field (rdopcode) in
Read opcode. Normal read
the device read instruction register (devrd) = 0x3.
The address transfer width field (addrwidth) and data
Single I/O (1 bit
Instruction type. transfer width field (datawidth) of the devrd register
wide)
= 0.
Delay in master reference clocks for the The clock delay for chip select deassert field (nss) in
length that the master mode chip select the quad SPI device delay register (delay).
200 ns
outputs are deasserted between words when
the clock phase is zero. Refer to delay[31:24] in Table A–11 on page A–19.
Delay in master reference clocks between

one chip select being deactivated and the
activation of another. This delay ensures a The clock delay for chip select deactivation field (btwn)
0 ns
quiet period between the selection of two in the delay register = 0x0.
different slaves and requires the transmit
FIFO to be empty.
Delay in master reference clocks between the
last bit of the current transaction and the first The clock delay for last transaction bit field (after) in
bit of the next transaction. If the clock phase the delay register.
20 ns
is zero, the first bit of the next transaction
refers to the cycle in which the chip select is Refer to delay[15:8] in Table A–11 on page A–19.
deselected.
Added delay in master reference clocks The clock delay with qspi_n_ss_out field (init) in
between setting qspi_n_ss_out low and 20 ns the delay register.
first bit transfer. Refer to delay[7:0] in Table A–11 on page A–19.
The number of address bytes field (numaddrbytes) of
Number of address bytes. 3 bytes
the device size register (devsz) = 2.

FPGA Configuration
Table A–10 lists the CSEL pin settings for the quad SPI controller.
Table A–10. Quad SPI Controller CSEL Pin Settings

CSEL Pin
Setting
0 1 2 3
osc1_clk (EOSC1 pin) range 10–50 MHz 20–50 MHz 25–50 MHz 10–25 MHz
osc1_clk/4, osc1_clk/2, osc1_clk*1, osc1_clk*2,
sclk_out device clock
12.5 MHz max 25 MHz max 50 MHz max 50 MHz max
qspi_clk controller clock
Controller baud rate divisor (even numbers only) 4 4 4 4
Flash read instruction (1 dummy byte for
READ READ READ_FAST READ_FAST
READ_FAST)
mpu_clk
The delay register in the quad SPI controller registers (qspiregs) configures relative
delay into the generation of the master output signals.
The SPI or quad SPI flash memory needs to meet the following timing requirements:
■ TSLCH (delay[7:0]): 20 ns
■ TCHSH (delay[15:8]): 20 ns
■ TSHSL (delay[31:24]): 200 ns
Table A–11 lists the delay register configuration with CSEL pin settings.
Table A–11. SPI and Quad SPI Flash Delay Configuration
Tref_clk Tsclk_out Device Delay Register

CSEL Pin
(ns) (ns) delay[7:0] delay[15:8] delay[31:24]
0 20 80 0 0 6
1 10 40 0 0 16
2–3 5 20 0 0 36
The formula to calculate the delay is (Tdelay-Tsclk_out) / Tref_clk.
f For more information about the quad SPI flash memory, refer to the Quad SPI Flash
FPGA Configuration
You can configure the FPGA portion of the SoC device with non-HPS sources or by
utilizing the HPS. Software executing on the HPS configures the FPGA by writing the
configuration image to the FPGA manager in the HPS. Software can control the
configuration process and monitor the FPGA status by accessing the control and
status register (CSR) interface in the FPGA manager.

FPGA Configuration
f For more information about configuring the FPGA through the HPS FPGA manager,
refer to the FPGA Manager chapter in volume 3 of the Cyclone V Device Handbook.
f For more information about configuring the FPGA in general, refer to the
Configuration, Design Security, and Remote System Upgrades in Cyclone V Devices chapter
Full Configuration
The HPS uses the FPGA manager to configure the FPGA portion of the device. The
following sequence suggests one way for software to perform a full configuration:
1. Set the cdratio and cfgwdth bits of the ctrl register in the FPGA manager
registers (fpgamgrregs) to match the characteristics of the configuration image.
These settings are dependant on the MSEL pins input.
2. Set the nce bit of the ctrl register to 0 to enable HPS configuration.
3. Set the en bit of the ctrl register to 1 to give the FPGA manager control of the
configuration input signals.
4. Set the nconfigpull bit of the ctrl register to 1 to pull down the nCONFIG pin and
put the FPGA portion of the device into the reset phase.
5. Poll the mode bit of the stat register and wait until the FPGA enters the reset
phase.
6. Set the nconfigpull bit of the ctrl register to 0 to release the FPGA from reset.
7. Read the mode bit of the stat register and wait until the FPGA enters the
configuration phase.
8. Clear the state of the nSTATUS bits (ns) in the configuration monitor registers
(fpgamgrregs.mon).
9. Set the axicfgen bit of the ctrl register to 1 to enable sending configuration data
to the FPGA.
10. Write the configuration image to the configuration data register (data) in the
FPGA manager module configuration data registers (fpgamgrdata). You can also
choose to use a DMA controller to transfer the configuration image from a
peripheral device to the FPGA manager.
11. Use the fpgamgrregs.mon registers to monitor the CONF_DONE (cd) and nSTATUS (ns)
bits.
a. CONF_DONE = 1 and nSTATUS = 1 indicates successful configuration.
b. CONF_DONE = 0 or nSTATUS = 0 indicates unsuccessful configuration. Complete
steps 12 and 13, then go back and repeat steps 3 to 10 to reload the
configuration image.
12. Set the axicfgen bit of the ctrl register to 0 to disable configuration data on AXI
slave.
13. Send the DCLKs required by the FPGA to enter the initialization phase.
a. If DCLK is unused, write a value of 4 to the DCLK count register (dclkcnt).
b. If DCLK is used, write a value of 20,480 (0x5000) to the dclkcnt register.

FPGA Configuration
14. Poll the dcntdone bit of the DCLK status register (dclkstat) until it changes to 1,
which indicates that all the DCLKs have been sent.
15. Write a 1 to the dcntdone bit of the DCLK status register to clear the completed
status flag.
16. Read the mode bit of the stat register to wait for the FPGA to enter user mode.
17. Set the en bit of the ctrl register to 0 to allow the external pins to drive the
If the HPS resets in the middle of a normal configuration data transfer before entering
user mode, software can assume that the configuration is unsuccessful. After the HPS
resets, software must repeat the steps for full configuration.
Partial Reconfiguration
Partial reconfiguration allows you to reconfigure part of the device while other
sections remain running. The HPS performs partial reconfiguration while the FPGA
portion of the device is in user mode. The following sequence suggests one way for
software to perform a partial configuration:
1. Read the mode bit of the stat register in fpgamgrregs to ensure that the FPGA is in
user mode.
2. Set the cdratio bit of the ctrl register to match the characteristics of the partial
reconfiguration image and set the cfgwdth bit of the ctrl register to 0 for 16-bit
configuration data width.
3. Set the en bit of the ctrl register to 1 to give the FPGA manager control of the
4. Set the prreq bit of the ctrl register to 1 to assert PR_REQUEST.
5. Write a value of 1 to the dclkcnt register to generate DCLK pulses for one clock
cycle.
6. Poll the fpgamgrregs.mon registers to observe the PR_READY (prr) bit.
a. If PR_READY=1, continue to step 7.
b. If PR_READY is 0, then go back and repeat step 5. Note that a minimum of 16
DCLK pulses are required.
7. Write a value of 3 to the dclkcnt register to generate DCLK pulses for three clock
cycles.
8. Set the axicfgen bit of the ctrl register to 1 to enable sending configuration data
to the FPGA.
9. Write the partial reconfiguration image to the data register in the FPGA manager
fpgamgrdata registers. You can also choose to use a DMA to transfer the
configuration image from a peripheral device to the FPGA manager.

10. Poll the fpgamgrregs.mon registers to observe the PR_DONE (prd), PR_READY (prr),
PR_ERROR (pre), and CRC_ERROR (crc) bits until the bits match one of the completion
statuses shown in Table A–12.
Table A–12. Partial Reconfiguration Status

Bits
Partial Reconfiguration Status
PR_DONE PR_READY PR_ERROR CRC_ERROR
1 0 0 0 Completed
0 0 1 0 Completed with error
0 0 0 1 Completed with a SEU event CRC error
11. Set the axicfgen bit of the ctrl register to 0 to disable sending configuration data
to the FPGA.
12. Set the prreq bit of the ctrl register to 0 to deassert PR_REQUEST.
13. Write a value of 128 to the dclkcnt register to generate DCLK pulses for 128 clock
cycles.
14. Poll the dcntdone bit of the dclkstat register until it changes to 1, which indicates
that all the DCLKs have been sent.
15. Write a 1 to the dcntdone bit of the dclkstat register to clear the completed status
flag.
16. Poll the fpgamgrregs.mon registers to observe the PR_DONE (prd), PR_READY (prr),
PR_ERROR (pre), and CRC_ERROR (crc) bits. When all bits are set to 0, the FPGA is
ready for the next transaction.
17. Set the en bit of the ctrl register to 0 to allow the external pins to drive the
If the HPS resets in the middle of a partial reconfiguration, software can assume that
the configuration is unsuccessful. After an HPS warm reset, software must repeat the
steps for partial configuration. After an HPS cold reset, software must repeat the steps
for “Full Configuration” on page A–20.
f For more information about the configuration modes and pin settings, refer to the
FPGA Manager chapter in volume 3 of the Cyclone V Device Handbook.

Table A–13 shows the revision history for this document.
Table A–13. Document Revision History (Part 1 of 2)

■ Expanded shared memory block table.
November 2012 1.3 ■ Added CSEL tables.
■ Additional minor updates.
June 2012 1.2 Updated the HPS boot and FPGA configuration sections.

Table A–13. Document Revision History (Part 2 of 2)

■ Updated the HPS boot section.
May 2012 1.1 ■ Added information about the flash devices used for HPS boot.
■ Added information about the FPGA configuration mode.


Additional Information
This chapter provides additional information about the document and Altera.
How to Contact Altera

To locate the most up-to-date information about Altera products, refer to the
following table.
Contact (1) Contact Method Address

Technical support Website www.altera.com/support
Website www.altera.com/training
Technical training
Email custrain@altera.com
Product literature Website www.altera.com/literature
Nontechnical support (general) Email nacomp@altera.com
(software licensing) Email authorization@altera.com
Note to Table:
(1) You can also contact your local Altera sales office or sales representative.
Typographic Conventions
The following table shows the typographic conventions this document uses.
Visual Cue Meaning

Indicate command names, dialog box titles, dialog box options, and other GUI
Bold Type with Initial Capital
labels. For example, Save As dialog box. For GUI elements, capitalization matches
Letters
the GUI.
Indicates directory names, project names, disk drive names, file names, file name
bold type extensions, software utility names, and GUI labels. For example, \qdesigns
directory, D: drive, and chiptrip.gdf file.
Italic Type with Initial Capital Letters Indicate document titles. For example, Stratix IV Design Guidelines.
Indicates variables. For example, n + 1.
italic type Variable names are enclosed in angle brackets (< >). For example, <file name> and
<project name>.pof file.
Indicate keyboard keys and menu names. For example, the Delete key and the
Initial Capital Letters
Options menu.
Quotation marks indicate references to sections in a document and titles of
“Subheading Title”
Quartus II Help topics. For example, “Typographic Conventions.”

Info–2 Additional Information
Typographic Conventions
Visual Cue Meaning

Indicates signal, port, register, bit, block, and primitive names. For example, data1,
tdi, and input. The suffix n denotes an active-low signal. For example, resetn.
Indicates command line commands and anything that must be typed exactly as it
Courier type appears. For example, c:\qdesigns\tutorial\chiptrip.gdf.
Also indicates sections of an actual file, such as a Report File, references to parts of
files (for example, the AHDL keyword SUBDESIGN), and logic function names (for
example, TRI).
r An angled arrow instructs you to press the Enter key.
1., 2., 3., and Numbered steps indicate a list of items when the sequence of the items is important,
a., b., c., and so on such as the steps listed in a procedure.
■ ■ ■ Bullets indicate a list of items when the sequence of the items is not important.
1 The hand points to information that requires special attention.
h The question mark directs you to a software help system with related information.
f The feet direct you to another document or website with related information.
m The multimedia icon directs you to a related multimedia presentation.
A caution calls attention to a condition or possible situation that can damage or
c
destroy the product or your work.
A warning calls attention to a condition or possible situation that can cause you
w
injury.
The envelope links to the Email Subscription Management Center page of the Altera
website, where you can sign up to receive update notifications for Altera documents.


Cyclone V Device Handbook Volume 3 Hard Processor System Technical Reference Manual PDF

Uploaded by

Copyright:

Available Formats

Cyclone V Device Handbook Volume 3 Hard Processor System Technical Reference Manual PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cyclone V Device Handbook Volume 3 Hard Processor System Technical Reference Manual PDF

Uploaded by

Copyright:

Available Formats

Cyclone V Device Handbook Volume 3: Hard Processor

System Technical Reference Manual

Cyclone V Device Handbook

101 Innovation Drive

November 2012 Altera Corporation Cyclone V Device Handbook

Chapter Revision Dates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiii

Chapter 1. Introduction to the Hard Processor System

Chapter 2. Clock Manager

November 2012 Altera Corporation Cyclone V Device Handbook

Functional Description of the Clock Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2–3

Chapter 3. Reset Manager

Section II. System Interconnect

Cyclone V Device Handbook November 2012 Altera Corporation

Slave Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4–9

Chapter 5. HPS-FPGA AXI Bridges

Section III. Cortex-A9 Microprocessor

Chapter 6. Cortex-A9 Microprocessor Unit Subsystem

November 2012 Altera Corporation Cyclone V Device Handbook

Cortex-A9 MPCore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6–4

Section IV. Debug and Trace

Chapter 7. CoreSight Debug and Trace

Cyclone V Device Handbook November 2012 Altera Corporation

Section V. Memory and Memory Controllers

Chapter 8. SDRAM Controller Subsystem

November 2012 Altera Corporation Cyclone V Device Handbook

Chapter 9. On-Chip Memory

Chapter 10. NAND Flash Controller

Cyclone V Device Handbook November 2012 Altera Corporation

Interrupt and DMA Enabling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10–22

Chapter 11. SD/MMC Controller

November 2012 Altera Corporation Cyclone V Device Handbook

Chapter 12. Quad SPI Flash Controller

Cyclone V Device Handbook November 2012 Altera Corporation

Interface Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12–14

Section VI. Peripherals

Chapter 13. FPGA Manager

Chapter 14. System Manager

November 2012 Altera Corporation Cyclone V Device Handbook

System Manager Address Map and Register Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14–12

Chapter 15. Scan Manager

Chapter 16. DMA Controller

Cyclone V Device Handbook November 2012 Altera Corporation

Abort Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–18

November 2012 Altera Corporation Cyclone V Device Handbook

Unaligned Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16–50

Chapter 17. Ethernet Media Access Controller

Cyclone V Device Handbook November 2012 Altera Corporation

Chapter 18. USB 2.0 OTG Controller

November 2012 Altera Corporation Cyclone V Device Handbook

Device Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18–14

Chapter 19. SPI Controller

Cyclone V Device Handbook November 2012 Altera Corporation

Chapter 20. I2C Controller

Chapter 21. UART Controller

November 2012 Altera Corporation Cyclone V Device Handbook

Chapter 22. General-Purpose I/O Interface

Chapter 23. Timer

Chapter 24. Watchdog Timer

Cyclone V Device Handbook November 2012 Altera Corporation