Tell us about your PDF experience.

C/C++ preprocessor reference

Article • 08/03/2021 • 2 minutes to read

The C/C++ preprocessor reference explains the preprocessor as it is implemented in

Microsoft C/C++. The preprocessor performs preliminary operations on C and C++ files
before they are passed to the compiler. You can use the preprocessor to conditionally
compile code, insert files, specify compile-time error messages, and apply machine-
specific rules to sections of code.

In Visual Studio 2019 the /Zc:preprocessor compiler option provides a fully conformant
C11 and C17 preprocessor. This is the default when you use the compiler flag /std:c11
or /std:c17 .

In this section
Preprocessor

Provides an overview of the traditional and new conforming preprocessors.

Preprocessor directives

Describes directives, typically used to make source programs easy to change and easy to
compile in different execution environments.

Preprocessor operators

Discusses the four preprocessor-specific operators used in the context of the #define
directive.

Predefined macros

Discusses predefined macros as specified by the C and C++ standards and by Microsoft
C++.

Pragmas

Discusses pragmas, which offer a way for each compiler to offer machine- and operating
system-specific features while retaining overall compatibility with the C and C++
languages.

Related sections
C++ language reference

Provides reference material for the Microsoft implementation of the C++ language.
C language reference

Provides reference material for the Microsoft implementation of the C language.

C/C++ build reference

Provides links to topics discussing compiler and linker options.

Visual Studio projects - C++

Describes the user interface in Visual Studio that enables you to specify the directories
that the project system will search to locate files for your C++ project.
Preprocessor
Article • 08/03/2021 • 2 minutes to read

The preprocessor is a text processor that manipulates the text of a source file as part of
the first phase of translation. The preprocessor doesn't parse the source text, but it does
break it up into tokens to locate macro calls. Although the compiler ordinarily invokes
the preprocessor in its first pass, the preprocessor can also be invoked separately to
process text without compiling.

The reference material on the preprocessor includes the following sections:

Preprocessor directives

Preprocessor operators

Predefined macros

Pragmas

Microsoft Specific

You can obtain a listing of your source code after preprocessing by using the /E or /EP
compiler option. Both options invoke the preprocessor and send the resulting text to
the standard output device, which, in most cases, is the console. The difference between
the two options is that /E includes #line directives, and /EP strips out these directives.

END Microsoft Specific

Special terminology
In the preprocessor documentation, the term "argument" refers to the entity that is
passed to a function. In some cases, it's modified by "actual" or "formal," which
describes the argument expression specified in the function call, and the argument
declaration specified in the function definition, respectively.

The term "variable" refers to a simple C-type data object. The term "object" refers to
both C++ objects and variables; it's an inclusive term.

See also
C/C++ preprocessor reference

Phases of translation
MSVC new preprocessor overview
Article • 02/01/2022 • 7 minutes to read

We're updating the Microsoft C++ preprocessor to improve standards conformance, fix
longstanding bugs, and change some behaviors that are officially undefined. We've also
added new diagnostics to warn on errors in macro definitions.

Starting in Visual Studio 2019 version 16.5, preprocessor support for the C++20
standard is feature-complete. These changes are available by using the /Zc:preprocessor
compiler switch. An experimental version of the new preprocessor is available in earlier
versions starting in Visual Studio 2017 version 15.8. You can enable it by using the
/experimental:preprocessor compiler switch. The default preprocessor behavior remains
the same as in previous versions.

New predefined macro

You can detect which preprocessor is in use at compile time. Check the value of the
predefined macro _MSVC_TRADITIONAL to tell if the traditional preprocessor is in use.
This macro is set unconditionally by versions of the compiler that support it,
independent of which preprocessor is invoked. Its value is 1 for the traditional
preprocessor. It's 0 for the conforming preprocessor.

C++

#if !defined(_MSVC_TRADITIONAL) || _MSVC_TRADITIONAL

// Logic using the traditional preprocessor

#else

// Logic using cross-platform compatible preprocessor

#endif

Behavior changes in the new preprocessor

The initial work on the new preprocessor has been focused on making all macro
expansions conform to the standard. It lets you use the MSVC compiler with libraries
that are currently blocked by the traditional behaviors. We tested the updated
preprocessor on real world projects. Here are some of the more common breaking
changes we found:

Macro comments
The traditional preprocessor is based on character buffers rather than preprocessor
tokens. It allows unusual behavior such as the following preprocessor comment trick,
which doesn't work under the conforming preprocessor:

C++

#if DISAPPEAR

#define DISAPPEARING_TYPE /##/

#else

#define DISAPPEARING_TYPE int

#endif

// myVal disappears when DISAPPEARING_TYPE is turned into a comment

DISAPPEARING_TYPE myVal;

The standards-conforming fix is to declare int myVal inside the appropriate

#ifdef/#endif directives:

C++

#define MYVAL 1

#ifdef MYVAL

int myVal;

#endif

L#val
The traditional preprocessor incorrectly combines a string prefix to the result of the
stringizing operator (#) operator:

C++

#define DEBUG_INFO(val) L"debug prefix:" L#val

// ^

// this prefix

const wchar_t *info = DEBUG_INFO(hello world);

In this case, the L prefix is unnecessary because the adjacent string literals are
combined after macro expansion anyway. The backward-compatible fix is to change the
definition:

C++
 #define DEBUG_INFO(val) L"debug prefix:" #val

// ^

// no prefix

The same issue is also found in convenience macros that "stringize" the argument to a
wide string literal:

C++

// The traditional preprocessor creates a single wide string literal token

#define STRING(str) L#str

You can fix the issue in various ways:

Use string concatenation of L"" and #str to add prefix. Adjacent string literals are
combined after macro expansion:

C++

#define STRING1(str) L""#str

Add the prefix after #str is stringized with additional macro expansion

C++

#define WIDE(str) L##str

#define STRING2(str) WIDE(#str)

Use the concatenation operator ## to combine the tokens. The order of

operations for ## and # is unspecified, although all compilers seem to evaluate
the # operator before ## in this case.

C++

#define STRING3(str) L## #str

Warning on invalid ##
When the token-pasting operator (##) doesn't result in a single valid preprocessing
token, the behavior is undefined. The traditional preprocessor silently fails to combine
the tokens. The new preprocessor matches the behavior of most other compilers and
emits a diagnostic.
 C++

// The ## is unnecessary and does not result in a single preprocessing

token.

#define ADD_STD(x) std::##x

// Declare a std::string

ADD_STD(string) s;

Comma elision in variadic macros

The traditional MSVC preprocessor always removes commas before empty __VA_ARGS__
replacements. The new preprocessor more closely follows the behavior of other popular
cross-platform compilers. For the comma to be removed, the variadic argument must be
missing (not just empty) and it must be marked with a ## operator. Consider the
following example:

C++

void func(int, int = 2, int = 3);

// This macro replacement list has a comma followed by __VA_ARGS__

#define FUNC(a, ...) func(a, __VA_ARGS__)

int main()

// In the traditional preprocessor, the

// following macro is replaced with:

// func(10,20,30)

FUNC(10, 20, 30);

// A conforming preprocessor replaces the

// following macro with: func(1, ), which

// results in a syntax error.

FUNC(1, );

In the following example, in the call to FUNC2(1) the variadic argument is missing in the
macro being invoked. In the call to FUNC2(1, ) the variadic argument is empty, but not
missing (notice the comma in the argument list).

C++

#define FUNC2(a, ...) func(a , ## __VA_ARGS__)

int main()

// Expands to func(1)

FUNC2(1);

// Expands to func(1, )

 FUNC2(1, );

In the upcoming C++20 standard, this issue has been addressed by adding __VA_OPT__ .
New preprocessor support for __VA_OPT__ is available starting in Visual Studio 2019
version 16.5.

C++20 variadic macro extension

The new preprocessor supports C++20 variadic macro argument elision:

C++

#define FUNC(a, ...) __VA_ARGS__ + a

int main()

int ret = FUNC(0);

return ret;

This code isn't conforming before the C++20 standard. In MSVC, the new preprocessor
extends this C++20 behavior to lower language standard modes ( /std:c++14 ,
/std:c++17 ). This extension matches the behavior of other major cross-platform C++

compilers.

Macro arguments are "unpacked"

In the traditional preprocessor, if a macro forwards one of its arguments to another
dependent macro then the argument doesn't get "unpacked" when it's inserted. Usually
this optimization goes unnoticed, but it can lead to unusual behavior:

C++

// Create a string out of the first argument, and the rest of the arguments.

#define TWO_STRINGS( first, ... ) #first, #__VA_ARGS__

#define A( ... ) TWO_STRINGS(__VA_ARGS__)

const char* c[2] = { A(1, 2) };

// Conforming preprocessor results:

// const char c[2] = { "1", "2" };

// Traditional preprocessor results, all arguments are in the first string:

// const char c[2] = { "1, 2", };

When expanding A() , the traditional preprocessor forwards all of the arguments
packaged in __VA_ARGS__ to the first argument of TWO_STRINGS, which leaves the
variadic argument of TWO_STRINGS empty. That causes the result of #first to be "1, 2"
rather than just "1". If you're following along closely, then you may be wondering what
happened to the result of #__VA_ARGS__ in the traditional preprocessor expansion: if the
variadic parameter is empty it should result in an empty string literal "" . A separate
issue kept the empty string literal token from being generated.

Rescanning replacement list for macros

After a macro is replaced, the resulting tokens are rescanned for additional macro
identifiers to replace. The algorithm used by the traditional preprocessor for doing the
rescan isn't conforming, as shown in this example based on actual code:

C++

#define CAT(a,b) a ## b

#define ECHO(...) __VA_ARGS__

// IMPL1 and IMPL2 are implementation details

#define IMPL1(prefix,value) do_thing_one( prefix, value)

#define IMPL2(prefix,value) do_thing_two( prefix, value)

// MACRO chooses the expansion behavior based on the value passed to

macro_switch

#define DO_THING(macro_switch, b) CAT(IMPL, macro_switch) ECHO(( "Hello",

b))

DO_THING(1, "World");

// Traditional preprocessor:

// do_thing_one( "Hello", "World");

// Conforming preprocessor:

// IMPL1 ( "Hello","World");

Although this example may seem a bit contrived, we've seen it in real-world code.

To see what's going on, we can break down the expansion starting with DO_THING :

1. DO_THING(1, "World") expands to CAT(IMPL, 1) ECHO(("Hello", "World"))

2. CAT(IMPL, 1) expands to IMPL ## 1 , which expands to IMPL1
3. Now the tokens are in this state: IMPL1 ECHO(("Hello", "World"))
4. The preprocessor finds the function-like macro identifier IMPL1 . Since it's not
followed by a ( , it isn't considered a function-like macro invocation.
5. The preprocessor moves on to the following tokens. It finds the function-like
macro ECHO gets invoked: ECHO(("Hello", "World")) , which expands to ("Hello",
 "World")

6. IMPL1 is never considered again for expansion, so the full result of the expansions
is: IMPL1("Hello", "World");

To modify the macro to behave the same way under both the new preprocessor and the
traditional preprocessor, add another layer of indirection:

C++

#define CAT(a,b) a##b

#define ECHO(...) __VA_ARGS__

// IMPL1 and IMPL2 are macros implementation details

#define IMPL1(prefix,value) do_thing_one( prefix, value)

#define IMPL2(prefix,value) do_thing_two( prefix, value)

#define CALL(macroName, args) macroName args

#define DO_THING_FIXED(a,b) CALL( CAT(IMPL, a), ECHO(( "Hello",b)))

DO_THING_FIXED(1, "World");

// macro expands to:

// do_thing_one( "Hello", "World");

Incomplete features before 16.5

Starting in Visual Studio 2019 version 16.5, the new preprocessor is feature-complete for
C++20. In previous versions of Visual Studio, the new preprocessor is mostly complete,
although some preprocessor directive logic still falls back to the traditional behavior.
Here's a partial list of incomplete features in Visual Studio versions before 16.5:

Support for _Pragma

C++20 features
Boost blocking bug: Logical operators in preprocessor constant expressions aren't
fully implemented in the new preprocessor before version 16.5. On some #if
directives, the new preprocessor can fall back to the traditional preprocessor. The
effect is only noticeable when macros incompatible with the traditional
preprocessor get expanded. It can happen when building Boost preprocessor slots.
Phases of translation
Article • 08/03/2021 • 2 minutes to read

C and C++ programs consist of one or more source files, each of which contains some
of the text of the program. A source file, together with its include files, files that are
included using the #include preprocessor directive, but not including sections of code
removed by conditional-compilation directives such as #if , is called a translation unit.

Source files can be translated at different times. In fact, it's common to translate only
out-of-date files. The translated translation units can be processed into separate object
files or object-code libraries. These separate, translated translation units are then linked
to form an executable program or a dynamic-link library (DLL). For more information
about files that can be used as input to the linker, see LINK input files.

Translation units can communicate using:

Calls to functions that have external linkage.

Calls to class member functions that have external linkage.

Direct modification of objects that have external linkage.

Direct modification of files.

Interprocess communication (for Microsoft Windows-based applications only).

The following list describes the phases in which the compiler translates files:

Character mapping

Characters in the source file are mapped to the internal source representation. Trigraph
sequences are converted to single-character internal representation in this phase.

Line splicing

All lines ending in a backslash (\) immediately followed by a newline character are joined
with the next line in the source file, forming logical lines from the physical lines. Unless
it's empty, a source file must end in a newline character that's not preceded by a
backslash.

Tokenization

The source file is broken into preprocessing tokens and white-space characters.
Comments in the source file are replaced with one space character each. Newline
characters are retained.
Preprocessing

Preprocessing directives are executed and macros are expanded into the source file. The
#include statement invokes translation starting with the preceding three translation

steps on any included text.

Character-set mapping

All source character set members and escape sequences are converted to their
equivalents in the execution character set. For Microsoft C and C++, both the source
and the execution character sets are ASCII.

String concatenation

All adjacent string and wide-string literals are concatenated. For example, "String "
"concatenation" becomes "String concatenation" .

Translation

All tokens are analyzed syntactically and semantically; these tokens are converted into
object code.

Linkage

All external references are resolved to create an executable program or a dynamic-link

library.

The compiler issues warnings or errors during phases of translation in which it

encounters syntax errors.

The linker resolves all external references and creates an executable program or DLL by
combining one or more separately processed translation units along with standard
libraries.

See also
Preprocessor
Preprocessor directives
Article • 08/03/2021 • 2 minutes to read

Preprocessor directives, such as #define and #ifdef , are typically used to make source
programs easy to change and easy to compile in different execution environments.
Directives in the source file tell the preprocessor to take specific actions. For example,
the preprocessor can replace tokens in the text, insert the contents of other files into the
source file, or suppress compilation of part of the file by removing sections of text.
Preprocessor lines are recognized and carried out before macro expansion. Therefore, if
a macro expands into something that looks like a preprocessor command, it isn't
recognized by the preprocessor.

Preprocessor statements use the same character set as source file statements, with the
exception that escape sequences aren't supported. The character set used in
preprocessor statements is the same as the execution character set. The preprocessor
also recognizes negative character values.

The preprocessor recognizes the following directives:

#define

#elif

#else

#endif

#error

#if

#ifdef

#ifndef

#import

#include

#line

#pragma

#undef

#using

The number sign ( # ) must be the first nonwhite-space character on the line containing
the directive. White-space characters can appear between the number sign and the first
letter of the directive. Some directives include arguments or values. Any text that follows
a directive (except an argument or value that is part of the directive) must be preceded
by the single-line comment delimiter ( // ) or enclosed in comment delimiters ( /* */ ).
Lines containing preprocessor directives can be continued by immediately preceding the
end-of-line marker with a backslash ( \ ).

Preprocessor directives can appear anywhere in a source file, but they apply only to the
rest of the source file, after they appear.

See also
Preprocessor operators

Predefined macros

c/c++ preprocessor reference

#define directive (C/C++)
Article • 08/03/2021 • 4 minutes to read

The #define creates a macro, which is the association of an identifier or parameterized

identifier with a token string. After the macro is defined, the compiler can substitute the
token string for each occurrence of the identifier in the source file.

Syntax
#define identifier token-stringopt

#define identifier ( identifieropt, ... , identifieropt ) token-stringopt

Remarks
The #define directive causes the compiler to substitute token-string for each occurrence
of identifier in the source file. The identifier is replaced only when it forms a token. That
is, identifier is not replaced if it appears in a comment, in a string, or as part of a longer
identifier. For more information, see Tokens.

The token-string argument consists of a series of tokens, such as keywords, constants, or

complete statements. One or more white-space characters must separate token-string
from identifier. This white space is not considered part of the substituted text, nor is any
white space that follows the last token of the text.

A #define without a token-string removes occurrences of identifier from the source file.
The identifier remains defined and can be tested by using the #if defined and #ifdef
directives.

The second syntax form defines a function-like macro with parameters. This form
accepts an optional list of parameters that must appear in parentheses. After the macro
is defined, each subsequent occurrence of identifier( identifieropt, ..., identifieropt ) is
replaced with a version of the token-string argument that has actual arguments
substituted for formal parameters.

Formal parameter names appear in token-string to mark the locations where actual
values are substituted. Each parameter name can appear multiple times in token-string,
and the names can appear in any order. The number of arguments in the call must
match the number of parameters in the macro definition. Liberal use of parentheses
guarantees that complex actual arguments are interpreted correctly.
The formal parameters in the list are separated by commas. Each name in the list must
be unique, and the list must be enclosed in parentheses. No spaces can separate
identifier and the opening parenthesis. Use line concatenation — place a backslash ( \ )
immediately before the newline character — for long directives on multiple source lines.
The scope of a formal parameter name extends to the new line that ends token-string.

When a macro has been defined in the second syntax form, subsequent textual
instances followed by an argument list indicate a macro call. The actual arguments that
follows an instance of identifier in the source file are matched to the corresponding
formal parameters in the macro definition. Each formal parameter in token-string that is
not preceded by a stringizing ( # ), charizing ( #@ ), or token-pasting ( ## ) operator, or not
followed by a ## operator, is replaced by the corresponding actual argument. Any
macros in the actual argument are expanded before the directive replaces the formal
parameter. (The operators are described in Preprocessor operators.)

The following examples of macros with arguments illustrate the second form of the
#define syntax:

// Macro to define cursor lines

#define CURSOR(top, bottom) (((top) << 8) | (bottom))

// Macro to get a random integer with a specified range

#define getrandom(min, max) \

((rand()%(int)(((max) + 1)-(min)))+ (min))

Arguments with side effects sometimes cause macros to produce unexpected results. A
given formal parameter may appear more than one time in token-string. If that formal
parameter is replaced by an expression with side effects, the expression, with its side
effects, may be evaluated more than one time. (See the examples under Token-Pasting
Operator (##).)

The #undef directive causes an identifier's preprocessor definition to be forgotten. See

The #undef Directive for more information.

If the name of the macro being defined occurs in token-string (even as a result of
another macro expansion), it is not expanded.

A second #define for a macro with the same name generates a warning unless the
second token sequence is identical to the first.

Microsoft Specific
Microsoft C/C++ lets you redefine a macro if the new definition is syntactically identical
to the original definition. In other words, the two definitions can have different
parameter names. This behavior differs from ANSI C, which requires that the two
definitions be lexically identical.

For example, the following two macros are identical except for the parameter names.
ANSI C does not allow such a redefinition, but Microsoft C/C++ compiles it without
error.

#define multiply( f1, f2 ) ( f1 * f2 )

#define multiply( a1, a2 ) ( a1 * a2 )

On the other hand, the following two macros are not identical and will generate a
warning in Microsoft C/C++.

#define multiply( f1, f2 ) ( f1 * f2 )

#define multiply( a1, a2 ) ( b1 * b2 )

END Microsoft Specific

This example illustrates the #define directive:

#define WIDTH 80

#define LENGTH ( WIDTH + 10 )

The first statement defines the identifier WIDTH as the integer constant 80 and defines
LENGTH in terms of WIDTH and the integer constant 10. Each occurrence of LENGTH is

replaced by ( WIDTH + 10 ). In turn, each occurrence of WIDTH + 10 is replaced by the

expression ( 80 + 10 ). The parentheses around WIDTH + 10 are important because they
control the interpretation in statements such as the following:

var = LENGTH * 20;

After the preprocessing stage the statement becomes:

C
 var = ( 80 + 10 ) * 20;

which evaluates to 1800. Without parentheses, the result is:

var = 80 + 10 * 20;

which evaluates to 280.

Microsoft Specific

Defining macros and constants with the /D compiler option has the same effect as using
a #define preprocessing directive at the start of your file. Up to 30 macros can be
defined by using the /D option.

END Microsoft Specific

See also
Preprocessor directives
#error directive (C/C++)
Article • 08/03/2021 • 2 minutes to read

The #error directive emits a user-specified error message at compile time, and then
terminates the compilation.

Syntax
#error token-string

Remarks
The error message that this directive emits includes the token-string parameter. The
token-string parameter is not subject to macro expansion. This directive is most useful
during preprocessing, to notify the developer of a program inconsistency, or the
violation of a constraint. The following example demonstrates error processing during
preprocessing:

C++

#if !defined(__cplusplus)

#error C++ compiler required.

#endif

See also
Preprocessor directives
#if, #elif, #else, and #endif directives
(C/C++)
Article • 01/25/2023 • 5 minutes to read

The #if directive, with the #elif, #else, and #endif directives, controls compilation of
portions of a source file. If the expression you write (after the #if) has a nonzero value,
the line group immediately following the #if directive is kept in the translation unit.

Grammar
conditional :

 if-part elif-partsopt else-partopt endif-line

if-part :

 if-line text

if-line :

 #if constant-expression

 #ifdef identifier

 #ifndef identifier

elif-parts :

 elif-line text

 elif-parts elif-line text

elif-line :

 #elif constant-expression

else-part :

 else-line text

else-line :

 #else

endif-line :

 #endif

Remarks
Each #if directive in a source file must be matched by a closing #endif directive. Any
number of #elif directives can appear between the #if and #endif directives, but at most
one #else directive is allowed. The #else directive, if present, must be the last directive
before #endif.

The #if, #elif, #else, and #endif directives can nest in the text portions of other #if
directives. Each nested #else, #elif, or #endif directive belongs to the closest preceding
#if directive.

All conditional-compilation directives, such as #if and #ifdef, must match a closing
#endif directive before the end of file. Otherwise, an error message is generated. When
conditional-compilation directives are contained in include files, they must satisfy the
same conditions: There must be no unmatched conditional-compilation directives at the
end of the include file.

Macro replacement is done within the part of the line that follows an #elif command, so
a macro call can be used in the constant-expression.

The preprocessor selects one of the given occurrences of text for further processing. A
block specified in text can be any sequence of text. It can occupy more than one line.
Usually text is program text that has meaning to the compiler or the preprocessor.

The preprocessor processes the selected text and passes it to the compiler. If text
contains preprocessor directives, the preprocessor carries out those directives. Only text
blocks selected by the preprocessor are compiled.

The preprocessor selects a single text item by evaluating the constant expression
following each #if or #elif directive until it finds a true (nonzero) constant expression. It
selects all text (including other preprocessor directives beginning with #) up to its
associated #elif, #else, or #endif.

If all occurrences of constant-expression are false, or if no #elif directives appear, the

preprocessor selects the text block after the #else clause. When there's no #else clause,
and all instances of constant-expression in the #if block are false, no text block is
selected.

The constant-expression is an integer constant expression with these additional

restrictions:

Expressions must have integral type and can include only integer constants,
character constants, and the defined operator.

The expression can't use sizeof or a type-cast operator.

The target environment may be unable to represent all ranges of integers.

 The translation represents type int the same way as type long , and unsigned int
the same way as unsigned long .

The translator can translate character constants to a set of code values different
from the set for the target environment. To determine the properties of the target
environment, use an app built for that environment to check the values of the
LIMITS.H macros.

The expression must not query the environment, and must remain insulated from
implementation details on the target computer.

Preprocessor operators

defined
The preprocessor operator defined can be used in special constant expressions, as
shown by the following syntax:

defined( identifier )

defined identifier

This constant expression is considered true (nonzero) if the identifier is currently defined.
Otherwise, the condition is false (0). An identifier defined as empty text is considered
defined. The defined operator can be used in an #if and an #elif directive, but nowhere
else.

In the following example, the #if and #endif directives control compilation of one of
three function calls:

#if defined(CREDIT)

credit();

#elif defined(DEBIT)

debit();

#else

printerror();

#endif

The function call to credit is compiled if the identifier CREDIT is defined. If the identifier
DEBIT is defined, the function call to debit is compiled. If neither identifier is defined,
the call to printerror is compiled. Both CREDIT and credit are distinct identifiers in C
and C++ because their cases are different.

The conditional compilation statements in the following example assume a previously

defined symbolic constant named DLEVEL .

#if DLEVEL > 5

#define SIGNAL 1

#if STACKUSE == 1

#define STACK 200

#else

#define STACK 100

#endif

#else

#define SIGNAL 0

#if STACKUSE == 1

#define STACK 100

#else

#define STACK 50

#endif

#endif

#if DLEVEL == 0

#define STACK 0

#elif DLEVEL == 1

#define STACK 100

#elif DLEVEL > 5

display( debugptr );

#else

#define STACK 200

#endif

The first #if block shows two sets of nested #if, #else, and #endif directives. The first set
of directives is processed only if DLEVEL > 5 is true. Otherwise, the statements after
#else are processed.

The #elif and #else directives in the second example are used to make one of four
choices, based on the value of DLEVEL . The constant STACK is set to 0, 100, or 200,
depending on the definition of DLEVEL . If DLEVEL is greater than 5, then the statement

#elif DLEVEL > 5

display(debugptr);

is compiled, and STACK isn't defined.

A common use for conditional compilation is to prevent multiple inclusions of the same
header file. In C++, where classes are often defined in header files, constructs like this
one can be used to prevent multiple definitions:

C++

/* EXAMPLE.H - Example header file */

#if !defined( EXAMPLE_H )

#define EXAMPLE_H

class Example

//...

};

#endif // !defined( EXAMPLE_H )

The preceding code checks to see if the symbolic constant EXAMPLE_H is defined. If so,
the file has already been included and doesn't need reprocessing. If not, the constant
EXAMPLE_H is defined to mark EXAMPLE.H as already processed.

__has_include
Visual Studio 2017 version 15.3 and later: Determines whether a library header is
available for inclusion:

C++

#ifdef __has_include

# if __has_include(<filesystem>)

# include <filesystem>

# define have_filesystem 1

# elif __has_include(<experimental/filesystem>)

# include <experimental/filesystem>

# define have_filesystem 1

# define experimental_filesystem

# else

# define have_filesystem 0

# endif

#endif

See also
Preprocessor directives
#ifdef and #ifndef directives (C/C++)
Article • 08/03/2021 • 2 minutes to read

The #ifdef and #ifndef preprocessor directives have the same effect as the #if
directive when it's used with the defined operator.

Syntax
#ifdef identifier

#ifndef identifier

These directives are equivalent to:

#if defined identifier

#if !defined identifier

Remarks
You can use the #ifdef and #ifndef directives anywhere #if can be used. The #ifdef
identifier statement is equivalent to #if 1 when identifier has been defined. It's

equivalent to #if 0 when identifier hasn't been defined, or has been undefined by
the #undef directive. These directives check only for the presence or absence of
identifiers defined with #define , not for identifiers declared in the C or C++ source
code.

These directives are provided only for compatibility with previous versions of the
language. The defined( identifier ) constant expression used with the #if directive
is preferred.

The #ifndef directive checks for the opposite of the condition checked by #ifdef . If the
identifier hasn't been defined, or if its definition has been removed with #undef , the
condition is true (nonzero). Otherwise, the condition is false (0).

Microsoft Specific

The identifier can be passed from the command line using the /D option. Up to 30
macros can be specified with /D .
The #ifdef directive is useful for checking whether a definition exists, because a
definition can be passed from the command line. For example:

C++

// ifdef_ifndef.CPP

// compile with: /Dtest /c

#ifndef test

#define final

#endif

END Microsoft Specific

See also
Preprocessor directives
#import directive (C++)
Article • 08/03/2021 • 6 minutes to read

C++ Specific

Used to incorporate information from a type library. The content of the type library is
converted into C++ classes, mostly describing the COM interfaces.

Syntax
#import "filename" [attributes]

#import <filename> [attributes]

Parameters
filename

Specifies the type library to import. The filename can be one of the following kinds:

The name of a file that contains a type library, such as an .olb, .tlb, or .dll file. The
keyword, file: , can precede each filename.

The progid of a control in the type library. The keyword, progid: , can precede
each progid. For example:

C++

#import "progid:my.prog.id.1.5"

For more on progids, see Specifying the Localization ID and Version Number.

When you use a 32-bit cross compiler on a 64-bit operating system, the compiler
can only read the 32-bit registry hive. You might want to use the native 64-bit
compiler to build and register a 64-bit type library.

The library ID of the type library. The keyword, libid: , can precede each library ID.
For example:

C++

#import "libid:12341234-1234-1234-1234-123412341234" version("4.0")

lcid("9")

 If you don't specify version or lcid , the rules applied to progid: are also applied
to libid: .

An executable (.exe) file.

A library (.dll) file containing a type library resource (such as an .ocx).

A compound document holding a type library.

Any other file format that can be understood by the LoadTypeLib API.

attributes

One or more #import attributes. Separate attributes with either a space or comma. For
example:

C++

#import "..\drawctl\drawctl.tlb" no_namespace, raw_interfaces_only

-or-

C++

#import "..\drawctl\drawctl.tlb" no_namespace raw_interfaces_only

Remarks

Search order for filename

filename is optionally preceded by a directory specification. The file name must name an
existing file. The difference between the two syntax forms is the order in which the
preprocessor searches for the type library files when the path is incompletely specified.

Syntax Action
form

Quoted Instructs the preprocessor to look for type library files first in the directory of the file
form that contains the #import statement, and then in the directories of whatever files
include ( #include ) that file. The preprocessor then searches along the paths shown
below.
 Syntax Action
form

Angle- Instructs the preprocessor to search for type library files along the following paths:

bracket
form 1. The PATH environment variable path list

2. The LIB environment variable path list

3. The path specified by the /I compiler option, except it the compiler is searching for a
type library that was referenced from another type library with the no_registry attribute.

Specify the localization ID and version number

When you specify a progid, you can also specify the localization ID and version number
of the progid. For example:

C++

#import "progid:my.prog.id" lcid("0") version("4.0)

If you don't specify a localization ID, a progid is chosen according to the following rules:

If there's only one localization ID, that one is used.

If there's more than one localization ID, the first one with version number 0, 9, or
409 is used.

If there's more than one localization ID and none of them are 0, 9, or 409, the last
one is used.

If you don't specify a version number, the most recent version is used.

Header files created by import

#import creates two header files that reconstruct the type library contents in C++
source code. The primary header file is similar to the one produced by the Microsoft
Interface Definition Language (MIDL) compiler, but with additional compiler-generated
code and data. The primary header file has the same base name as the type library, plus
a .TLH extension. The secondary header file has the same base name as the type library,
with a .TLI extension. It contains the implementations for compiler-generated member
functions, and is included ( #include ) in the primary header file.

If importing a dispinterface property that uses byref parameters, #import doesn't

generate a __declspec(property) statement for the function.
Both header files are placed in the output directory specified by the /Fo (name object
file) option. They're then read and compiled by the compiler as if the primary header file
was named by a #include directive.

The following compiler optimizations come with the #import directive:

The header file, when created, is given the same timestamp as the type library.

When #import is processed, the compiler first checks if the header exists and is
up-to-date. If yes, then it doesn't need to be re-created.

The #import directive also participates in minimal rebuild and can be placed in a
precompiled header file. For more information, see Creating precompiled header files.

Primary type library header file

The primary type library header file consists of seven sections:

Heading boilerplate: Consists of comments, #include statement for COMDEF.H

(which defines some standard macros used in the header), and other miscellaneous
setup information.

Forward references and typedefs: Consists of structure declarations such as struct

IMyInterface and typedefs.

Smart pointer declarations: The template class _com_ptr_t is a smart pointer. It

encapsulates interface pointers, and eliminates the need to call AddRef , Release ,
and QueryInterface functions. It also hides the CoCreateInstance call when
creating a new COM object. This section uses the macro statement
_COM_SMARTPTR_TYPEDEF to establish typedefs of COM interfaces as template

specializations of the _com_ptr_t template class. For example, for interface

IMyInterface , the .TLH file will contain:

TLH

_COM_SMARTPTR_TYPEDEF(IMyInterface, __uuidof(IMyInterface));

which the compiler will expand to:

C++

typedef _com_ptr_t<_com_IIID<IMyInterface, __uuidof(IMyInterface)> >

IMyInterfacePtr;

 Type IMyInterfacePtr can then be used in place of the raw interface pointer
IMyInterface* . Consequently, there's no need to call the various IUnknown
member functions

Typeinfo declarations: Primarily consists of class definitions and other items

exposing the individual typeinfo items returned by ITypeLib:GetTypeInfo . In this
section, each typeinfo from the type library is reflected in the header in a form
dependent on the TYPEKIND information.

Optional old-style GUID definition: Contains initializations of the named GUID

constants. These names have the form CLSID_CoClass and IID_Interface , similar
to the ones generated by the MIDL compiler.

#include statement for the secondary type library header.

Footer boilerplate: Currently includes #pragma pack(pop) .

All sections, except the heading boilerplate and footer boilerplate section, are enclosed
in a namespace with its name specified by the library statement in the original IDL file.
You can use the names from the type library header by an explicit qualification using the
namespace name. Or, you can include the following statement:

C++

using namespace MyLib;

immediately after the #import statement in the source code.

The namespace can be suppressed by using the no_namespace) attribute of the

#import directive. However, suppressing the namespace may lead to name collisions.
The namespace can also be renamed by the rename_namespace attribute.

The compiler provides the full path to any type library dependency required by the type
library it's currently processing. The path is written, in the form of comments, into the
type library header (.TLH) that the compiler generates for each processed type library.

If a type library includes references to types defined in other type libraries, then the .TLH
file will include comments of the following sort:

TLH

//

// Cross-referenced type libraries:

//

 // #import "c:\path\typelib0.tlb"

//

The actual filename in the #import comment is the full path of the cross-referenced
type library, as stored in the registry. If you encounter errors that are caused by missing
type definitions, check the comments at the head of the .TLH to see which dependent
type libraries may need to be imported first. Likely errors are syntax errors (for example,
C2143, C2146, C2321), C2501 (missing decl-specifiers), or C2433 ('inline' not permitted
on data declaration) while compiling the .TLI file.

To resolve dependency errors, determine which of the dependency comments aren't

otherwise provided for by system headers, and then provide an #import directive at
some point before the #import directive of the dependent type library.

#import attributes
#import can optionally include one or more attributes. These attributes tell the compiler
to modify the contents of the type-library headers. A backslash (\) symbol can be used
to include additional lines in a single #import statement. For example:

C++

#import "test.lib" no_namespace \

rename("OldName", "NewName")

For more information, see #import attributes.

END C++ Specific

See also
Preprocessor directives

Compiler COM support

#import attributes (C++)
Article • 08/03/2021 • 2 minutes to read

Provides links to attributes used with the #import directive.

Microsoft Specific

The following attributes are available to the #import directive.

Attribute Description

auto_rename Renames C++ reserved words by appending two underscores (__) to

the variable name to resolve potential name conflicts.

auto_search Specifies that, when a type library is referenced with #import and
itself references another type library, the compiler can do an implicit
#import for the other type library.

embedded_idl Specifies that the type library is written to the .tlh file with the
attribute-generated code preserved.

exclude Excludes items from the type library header files being generated.

high_method_prefix Specifies a prefix to be used in naming high-level properties and

methods.

high_property_prefixes Specifies alternate prefixes for three property methods.

implementation_only Suppresses the generation of the .tlh header file (the primary header
file).

include() Disables automatic exclusion.

inject_statement Inserts its argument as source text into the type-library header.

named_guids Tells the compiler to define and initialize GUID variables in old style,
of the form LIBID_MyLib , CLSID_MyCoClass , IID_MyInterface , and
DIID_MyDispInterface .

no_auto_exclude Disables automatic exclusion.

no_dual_interfaces Changes the way the compiler generates wrapper functions for dual
interface methods.

no_implementation Suppresses the generation of the .tli header, which contains the
implementations of the wrapper member functions.

no_namespace Specifies that the namespace name is not generated by the

compiler.
 Attribute Description

no_registry Tells the compiler not to search the registry for type libraries.

no_search_namespace Has the same functionality as the no_namespace attribute but is

used on type libraries that you use the #import directive with the
auto_search attribute.

no_smart_pointers Suppresses the creation of smart pointers for all interfaces in the
type library.

raw_dispinterfaces Tells the compiler to generate low-level wrapper functions for

dispinterface methods and properties that call IDispatch::Invoke
and return the HRESULT error code.

raw_interfaces_only Suppresses the generation of error-handling wrapper functions and

property declarations that use those wrapper functions.

raw_method_prefix Specifies a different prefix to avoid name collisions.

raw_native_types Disables the use of COM support classes in the high-level wrapper
functions and forces the use of low-level data types instead.

raw_property_prefixes Specifies alternate prefixes for three property methods.

rename Works around name collision problems.

rename_namespace Renames the namespace that contains the contents of the type
library.

rename_search_namespace Has the same functionality as the rename_namespace attribute but

is used on type libraries that you use the #import directive with the
auto_search attribute.

tlbid Allows for loading libraries other than the primary type library.

END Microsoft Specific

See also
#import directive
auto_rename import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Renames C++ reserved words by appending two underscores (__) to the variable name
to resolve potential name conflicts.

Syntax
#import type-library auto_rename

Remarks
This attribute is used when importing a type library that uses one or more C++ reserved
words (keywords or macros) as variable names.

END C++ Specific

See also
#import attributes

#import directive
auto_search import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Specifies that, when a type library is referenced with #import and itself references
another type library, the compiler can do an implicit #import for the other type library.

Syntax
#import type-library auto_search

Remarks
END C++ Specific

See also
#import attributes

#import directive
embedded_idl import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Specifies whether the type library is written to the .tlh file with the attribute-generated
code preserved.

Syntax
#import type-library embedded_idl [ ( { "emitidl" | "no_emitidl" } ) ]

Parameters
"emitidl"

Type information imported from type-library is present in the IDL generated for the
attributed project. This behavior is the default, and is in effect if you don't specify a
parameter to embedded_idl .

"no_emitidl"

Type information imported from type-library isn't present in the IDL generated for the
attributed project.

Example
C++

// import_embedded_idl.cpp

// compile with: /LD

#include <windows.h>

[module(name="MyLib2")];

#import "\school\bin\importlib.tlb" embedded_idl("no_emitidl")

END C++ Specific

See also
#import attributes

#import directive
exclude import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Excludes items from the type library header files being generated.

Syntax
#import type-library exclude( "Name1" [ , "Name2" ... ] )

Parameters
Name1

First item to be excluded.

Name2

(Optional) Second and later items to be excluded, if necessary.

Remarks
Type libraries may include definitions of items defined in system headers or other type
libraries. This attribute can take any number of arguments, where each is a top-level
type library item to be excluded.

END C++ Specific

See also
#import attributes

#import directive
high_method_prefix import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Specifies a prefix to be used in naming high-level properties and methods.

Syntax
#import type-library high_method_prefix( "Prefix" )

Parameters
Prefix

Prefix to be used.

Remarks
By default, high-level error-handling properties and methods are exposed by member
functions named without a prefix. The names are from the type library.

END C++ Specific

See also
#import attributes

#import directive
high_property_prefixes import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Specifies alternate prefixes for three property methods.

Syntax
#import type-library high_property_prefixes( "GetPrefix" , "PutPrefix" ,
"PutRefPrefix" )

Parameters
GetPrefix

Prefix to be used for the propget methods.

PutPrefix

Prefix to be used for the propput methods.

PutRefPrefix

Prefix to be used for the propputref methods.

Remarks
By default, high-level error-handling propget , propput , and propputref methods are
exposed by member functions named with prefixes Get , Put , and PutRef , respectively.

END C++ Specific

See also
#import attributes

#import directive
implementation_only import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Suppresses the generation of the .tlh primary type-library header file.

Syntax
#import type-library implementation_only

Remarks
This file contains all the declarations used to expose the type-library contents. The .tli
header file, with the implementations of the wrapper member functions, will be
generated and included in the compilation.

When this attribute is specified, the content of the .tli header is in the same
namespace as the one normally used in the .tlh header. In addition, the member
functions are not declared as inline.

The implementation_only attribute is intended for use in conjunction with the

no_implementation attribute as a way of keeping the implementations out of the
precompiled header (PCH) file. An #import statement with the no_implementation
attribute is placed in the source region used to create the PCH. The resulting PCH is
used by a number of source files. An #import statement with the implementation_only
attribute is then used outside the PCH region. You're required to use this statement only
once in one of the source files. It generates all the required wrapper member functions
without additional recompilation for each source file.

７ Note

The implementation_only attribute in one #import statement must be used in

conjunction with another #import statement, of the same type library, with the
no_implementation attribute. Otherwise, compiler errors are generated. This is
because wrapper class definitions generated by the #import statement with the
no_implementation attribute are required to compile the implementations

generated by the implementation_only attribute.

END C++ Specific

See also
#import attributes

#import directive
include() import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Disables automatic exclusion.

Syntax
#import type-library include( "Name1" [,"Name2" ... ] )

Parameters
Name1

First item to be forcibly included.

Name2

Second item to be forcibly included (if necessary).

Remarks
Type libraries may include definitions of items defined in system headers or other type
libraries. #import attempts to avoid multiple definition errors by automatically excluding
such items. If some items shouldn't be excluded automatically, you may see Compiler
Warning (level 3) C4192. You can use this attribute to disable the automatic exclusion.
This attribute can take any number of arguments, one for each name of a type-library
item to be included.

END C++ Specific

See also
#import attributes

#import directive
inject_statement import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Inserts its argument as source text into the type-library header.

Syntax
#import type-library inject_statement( "source-text" )

Parameters
source-text

Source text to be inserted into the type library header file.

Remarks
The text is placed at the beginning of the namespace declaration that wraps the type-
library contents in the header file.

END C++ Specific

See also
#import attributes

#import directive
named_guids import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Tells the compiler to define and initialize GUID variables in the old style, of the form
LIBID_MyLib , CLSID_MyCoClass , IID_MyInterface , and DIID_MyDispInterface .

Syntax
#import type-library named_guids

END C++ Specific

See also
#import attributes

#import directive
no_auto_exclude import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Disables automatic exclusion.

Syntax
#import type-library no_auto_exclude

Remarks
Type libraries may include definitions of items defined in system headers or other type
libraries. #import attempts to avoid multiple definition errors by automatically excluding
such items. It causes Compiler Warning (level 3) C4192 to be issued for each item to be
excluded. You can disable the automatic exclusion by using this attribute.

END C++ Specific

See also
#import attributes

#import directive
no_dual_interfaces import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Changes the way the compiler generates wrapper functions for dual interface methods.

Syntax
#import type-library no_dual_interfaces

Remarks
Normally, the wrapper calls the method through the virtual function table for the
interface. With no_dual_interfaces, the wrapper instead calls IDispatch::Invoke to
invoke the method.

END C++ Specific

See also
#import attributes

#import directive
no_implementation import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Suppresses the generation of the .tli header, which contains the implementations of
the wrapper member functions.

Syntax
#import type-library no_implementation

Remarks
If this attribute is specified, the .tlh header, with the declarations to expose type-library
items, will be generated without an #include statement to include the .tli header file.

This attribute is used in conjunction with implementation_only.

END C++ Specific

See also
#import attributes

#import directive
no_namespace import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Specifies that the compiler doesn't generate a namespace name.

Syntax
#import type-library no_namespace

Remarks
The type-library contents in the #import header file are normally defined in a
namespace. The namespace name is specified in the library statement of the original
IDL file. If the no_namespace attribute is specified, then this namespace isn't generated
by the compiler.

If you want to use a different namespace name, then use the rename_namespace
attribute instead.

END C++ Specific

See also
#import attributes

#import directive
no_registry import attribute
Article • 08/03/2021 • 2 minutes to read

no_registry tells the compiler not to search the registry for type libraries imported with
#import .

Syntax
#import type-library no_registry

Parameters
type-library

A type library.

Remarks
If a referenced type library isn't found in the include directories, the compilation fails
even if the type library is in the registry. no_registry propagates to other type libraries
implicitly imported with auto_search .

The compiler never searches the registry for type libraries that are specified by file name
and passed directly to #import .

When auto_search is specified, the additional #import directives are generated by using
the no_registry setting of the initial #import . If the initial #import directive was
no_registry, an auto_search -generated #import is also no_registry.

no_registry is useful if you want to import cross-referenced type libraries. It keeps the
compiler from finding an older version of the file in the registry. no_registry is also
useful if the type library isn't registered.

See also
#import attributes

#import directive
no_search_namespace import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Has the same functionality as the no_namespace attribute, but is used on type libraries
where you use the #import directive with the auto_search attribute.

Syntax
#import type-library no_search_namespace

END C++ Specific

See also
#import attributes

#import directive
no_smart_pointers import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Suppresses the creation of smart pointers for all interfaces in the type library.

Syntax
#import type-library no_smart_pointers

Remarks
By default, when you use #import , you get a smart pointer declaration for all interfaces
in the type library. These smart pointers are of type _com_ptr_t.

END C++ Specific

See also
#import attributes

#import directive
raw_dispinterfaces import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Tells the compiler to generate low-level wrapper functions for dispinterface methods,
and for properties that call IDispatch::Invoke and return the HRESULT error code.

Syntax
#import type-library raw_dispinterfaces

Remarks
If this attribute isn't specified, only high-level wrappers are generated, which throw C++
exceptions on failure.

END C++ Specific

See also
#import attributes

#import directive
raw_interfaces_only import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Suppresses the generation of error-handling wrapper functions, and property

declarations that use those wrapper functions.

Syntax
#import type-library raw_interfaces_only

Remarks
The raw_interfaces_only attribute also causes the default prefix used in naming the non-
property functions to be removed. Normally, the prefix is raw_ . If this attribute is
specified, the function names are taken directly from the type library.

This attribute allows you to expose only the low-level contents of the type library.

END C++ Specific

See also
#import attributes

#import directive
raw_method_prefix
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Specifies a different prefix to avoid name collisions.

Syntax
#import type-library raw_method_prefix( "Prefix" )

Parameters
Prefix

The prefix to be used.

Remarks
Low-level properties and methods are exposed by member functions named using a
default prefix of raw_, to avoid name collisions with the high-level error-handling
member functions.

７ Note

The effects of the raw_method_prefix attribute are unchanged by the presence of

the raw_interfaces_only attribute. The raw_method_prefix always takes precedence
over raw_interfaces_only in specifying a prefix. If both attributes are used in the
same #import statement, then the prefix specified by the raw_method_prefix
attribute is used.

END C++ Specific

See also
#import attributes

#import directive
raw_native_types import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Disables the use of COM support classes in the high-level wrapper functions, and forces
the use of low-level data types instead.

Syntax
#import type-library raw_native_types

Remarks
By default, the high-level error-handling methods use the COM support classes _bstr_t
and _variant_t in place of the BSTR and VARIANT data types and raw COM interface
pointers. These classes encapsulate the details of allocating and deallocating memory
storage for these data types, and greatly simplify type casting and conversion
operations.

END C++ Specific

See also
#import attributes

#import directive
raw_property_prefixes import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Specifies alternate prefixes for three property methods.

Syntax
#import type-library raw_property_prefixes( "GetPrefix" , "PutPrefix" , "PutRefPrefix"
)

Parameters
GetPrefix

Prefix to use for the propget methods.

PutPrefix

Prefix to use for the propput methods.

PutRefPrefix

Prefix to use for the propputref methods.

Remarks
By default, low-level propget , propput , and propputref methods are exposed by
member functions named using prefixes of get_ , put_ , and putref_ , respectively. These
prefixes are compatible with the names used in the header files generated by MIDL.

END C++ Specific

See also
#import attributes

#import directive
rename import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Works around name collision problems.

Syntax
#import type-library rename( "OldName" , "NewName" )

Parameters
OldName

Old name in the type library.

NewName

Name to be used instead of the old name.

Remarks
When the rename attribute is specified, the compiler replaces all occurrences of
OldName in type-library with the user-supplied NewName in the resulting header files.

The rename attribute can be used when a name in the type library coincides with a
macro definition in the system header files. If this situation isn't resolved, the compiler
may issue various syntax errors, such as Compiler Error C2059 and Compiler Error C2061.

７ Note

The replacement is for a name used in the type library, not for a name used in the
resulting header file.

For example, suppose a property named MyParent exists in a type library, and a macro
GetMyParent is defined in a header file and used before #import . Since GetMyParent is

the default name of a wrapper function for the error-handling get property, a name
collision will occur. To work around the problem, use the following attribute in the
#import statement:
 C++

#import MyTypeLib.tlb rename("MyParent","MyParentX")

which renames the name MyParent in the type library. An attempt to rename the
GetMyParent wrapper name will fail:

C++

#import MyTypeLib.tlb rename("GetMyParent","GetMyParentX")

It's because the name GetMyParent only occurs in the resulting type library header file.

END C++ Specific

See also
#import attributes

#import directive
rename_namespace import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Renames the namespace that contains the contents of the type library.

Syntax
#import type-library rename_namespace( "NewName" )

Parameters
NewName

The new name of the namespace.

Remarks
The rename_namespace attribute takes a single argument, NewName, which specifies
the new name for the namespace.

To remove the namespace, use the no_namespace attribute instead.

END C++ Specific

See also
#import attributes

#import directive
rename_search_namespace import
attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Has the same functionality as the rename_namespace attribute, but is used on type
libraries where you use the #import directive along with the auto_search attribute.

Syntax
#import type-library rename_search_namespace( "NewName" )

Parameters
NewName

The new name of the namespace.

Remarks
END C++ Specific

See also
#import attributes

#import directive
tlbid import attribute
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Allows for loading libraries other than the primary type library.

Syntax
#import type-library-dll tlbid( number )

Parameters
number

The number of the type library in type-library-dll.

Remarks
If multiple type libraries are built into a single DLL, it's possible to load libraries other
than the primary type library by using tlbid.

For example:

C++

#import <MyResource.dll> tlbid(2)

is equivalent to:

C++

LoadTypeLib("MyResource.dll\\2");

END C++ Specific

See also
#import attributes

#import directive
#include directive (C/C++)
Article • 02/18/2022 • 4 minutes to read

Tells the preprocessor to include the contents of a specified file at the point where the
directive appears.

Syntax
#include " path-spec "

#include < path-spec >

Remarks
You can organize constant and macro definitions into include files (also known as header
files) and then use #include directives to add them to any source file. Include files are
also useful for incorporating declarations of external variables and complex data types.
The types may be defined and named only once in an include file created for that
purpose.

The path-spec is a file name that may optionally be preceded by a directory

specification. The file name must name an existing file. The syntax of the path-spec
depends on the operating system on which the program is compiled.

For information about how to reference assemblies in a C++ application that's compiled
by using /clr, see #using directive.

Both syntax forms cause the #include directive to be replaced by the entire contents of
the specified file. The difference between the two forms is the order of the paths that
the preprocessor searches when the path is incompletely specified. The following table
shows the difference between the two syntax forms.

Syntax Action
Form
 Syntax Action
Form

Quoted The preprocessor searches for include files in this order:

form
1) In the same directory as the file that contains the #include statement.

2) In the directories of the currently opened include files, in the reverse order in which
they were opened. The search begins in the directory of the parent include file and
continues upward through the directories of any grandparent include files.

3) Along the path that's specified by each /I compiler option.

4) Along the paths that are specified by the INCLUDE environment variable.

Angle- The preprocessor searches for include files in this order:

bracket
form 1) Along the path that's specified by each /I compiler option.

2) When compiling occurs on the command line, along the paths that are specified by
the INCLUDE environment variable.

The preprocessor stops searching as soon as it finds a file that has the given name. If
you enclose a complete, unambiguous path specification for the include file between
double quotation marks ( " " ), the preprocessor searches only that path specification
and ignores the standard directories.

If the file name that's enclosed in double quotation marks is an incomplete path
specification, the preprocessor first searches the parent file's directory. A parent file is
the file that contains the #include directive. For example, if you include a file named
file2 in a file named file1, file1 is the parent file.

Include files can be nested: An #include directive can appear in a file that's named by
another #include directive. For example, file2 could include file3. In this case, file1 would
still be the parent of file2, but it would be the grandparent of file3.

When include files are nested and when compiling occurs on the command line,
directory searching begins in the directory of the parent file. Then it proceeds through
the directories of any grandparent files. That is, searching begins relative to the directory
that contains the source that's currently being processed. If the file isn't found, the
search moves to directories that are specified by the /I (Additional include directories)
compiler option. Finally, the directories that are specified by the INCLUDE environment
variable are searched.
Within the Visual Studio development environment, the INCLUDE environment variable is
ignored. The values specified in the project properties for include directories are used
instead. For more information about how to set the include directories in Visual Studio,
see Include Directories and Additional Include Directories.

This example shows file inclusion by using angle brackets:

#include <stdio.h>

The example adds the contents of the file named stdio.h to the source program. The
angle brackets cause the preprocessor to search the directories that are specified by the
INCLUDE environment variable for stdio.h , after it searches directories that are specified

by the /I compiler option.

The next example shows file inclusion by using the quoted form:

#include "defs.h"

The example adds the contents of the file that's specified by defs.h to the source
program. The quotation marks mean that the preprocessor first searches the directory
that contains the parent source file.

Nesting of include files can continue up to 10 levels. When processing of the nested
#include is finished, the preprocessor continues to insert the enclosing parent include

file into the original source file.

Microsoft-specific

To locate the source files to include, the preprocessor first searches the directories
specified by the /I compiler option. If the /I option isn't present, or if it fails, the
preprocessor uses the INCLUDE environment variable to find any include files within
angle brackets. The INCLUDE environment variable and /I compiler option can contain
multiple paths, separated by semicolons ( ; ). If more than one directory appears as part
of the /I option or within the INCLUDE environment variable, the preprocessor searches
them in the order in which they appear.

For example, the command

cmd
 CL /ID:\msvc\include myprog.c

causes the preprocessor to search the directory D:\msvc\include\ for include files such
as stdio.h . The commands

cmd

SET INCLUDE=D:\msvc\include

CL myprog.c

have the same effect. If both sets of searches fail, a fatal compiler error is generated.

If the file name is fully specified for an include file that has a path that includes a colon
(for example, F:\MSVC\SPECIAL\INCL\TEST.H ), the preprocessor follows the path.

For include files that are specified as #include "path-spec" , directory search begins in
the directory of the parent file and then proceeds through the directories of any
grandparent files. That is, the search begins relative to the directory that contains the
source file that's being processed. If there's no grandparent file and the file still isn't
found, the search continues as if the file name were enclosed in angle brackets.

END Microsoft-specific

See also
Preprocessor directives

/I (Additional include directories)

#line directive (C/C++)
Article • 03/14/2022 • 2 minutes to read

The #line directive tells the preprocessor to set the compiler's reported values for the
line number and filename to a given line number and filename.

Syntax
#line digit-sequence ["filename"]

Remarks
The compiler uses the line number and optional filename to refer to errors that it finds
during compilation. The line number usually refers to the current input line, and the
filename refers to the current input file. The line number is incremented after each line is
processed.

The digit-sequence value can be any integer constant within the range from 0 to
2147483647, inclusive. Macro replacement can be used on the preprocessing tokens,
but the result must evaluate to the correct syntax. The filename can be any combination
of characters and must be enclosed in double quotation marks ( " " ). If filename is
omitted, the previous filename remains unchanged.

You can alter the source line number and filename by writing a #line directive. The
#line directive sets the value for the line that immediately follows the directive in the
source file. The translator uses the line number and filename to determine the values of
the predefined macros __FILE__ and __LINE__ . You can use these macros to insert self-
descriptive error messages into the program text. For more information on these
predefined macros, see Predefined macros.

The __FILE__ macro expands to a string whose contents are the filename, surrounded
by double quotation marks ( " " ).

If you change the line number and filename, the compiler ignores the previous values
and continues processing with the new values. The #line directive is typically used by
program generators. It's used to cause error messages to refer to the original source file,
instead of to the generated program.

Example
The following examples illustrate #line and the __LINE__ and __FILE__ macros.

In the first example, the line number is set to 10, then to 20, and the filename is changed
to hello.cpp.

C++

// line_directive.cpp

// Compile by using: cl /W4 /EHsc line_directive.cpp

#include <stdio.h>

int main()

printf( "This code is on line %d, in file %s\n", __LINE__, __FILE__ );

#line 10

printf( "This code is on line %d, in file %s\n", __LINE__, __FILE__ );

#line 20 "hello.cpp"

printf( "This code is on line %d, in file %s\n", __LINE__, __FILE__ );

printf( "This code is on line %d, in file %s\n", __LINE__, __FILE__ );

Output

This code is on line 7, in file line_directive.cpp

This code is on line 10, in file line_directive.cpp

This code is on line 20, in file hello.cpp

This code is on line 21, in file hello.cpp

In this example, the macro ASSERT uses the predefined macros __LINE__ and __FILE__
to print an error message about the source file if a given assertion isn't true.

#define ASSERT(cond) if( !(cond) )\

{printf( "assertion error line %d, file(%s)\n", \

__LINE__, __FILE__ );}

See also
Preprocessor directives
Null directive
Article • 08/03/2021 • 2 minutes to read

The null preprocessor directive is a single number sign (#) alone on a line. It has no
effect.

Syntax
#

See also
Preprocessor directives
#undef directive (C/C++)
Article • 08/03/2021 • 2 minutes to read

Removes (undefines) a name previously created with #define .

Syntax
#undef identifier

Remarks
The #undef directive removes the current definition of identifier. Consequently,
subsequent occurrences of identifier are ignored by the preprocessor. To remove a
macro definition using #undef, give only the macro identifier, not a parameter list.

You can also apply the #undef directive to an identifier that has no previous definition.
This ensures that the identifier is undefined. Macro replacement isn't performed within
#undef statements.

The #undef directive is typically paired with a #define directive to create a region in a
source program in which an identifier has a special meaning. For example, a specific
function of the source program can use manifest constants to define environment-
specific values that do not affect the rest of the program. The #undef directive also
works with the #if directive to control conditional compilation of the source program.
For more information, see The #if, #elif, #else, and #endif directives.

In the following example, the #undef directive removes definitions of a symbolic

constant and a macro. Note that only the identifier of the macro is given.

#define WIDTH 80

#define ADD( X, Y ) ((X) + (Y))

#undef WIDTH

#undef ADD

Microsoft Specific
Macros can be undefined from the command line using the /U option, followed by the
macro names to be undefined. The effect of issuing this command is equivalent to a
sequence of #undef macro-name statements at the beginning of the file.

END Microsoft Specific

See also
Preprocessor directives
#using directive (C++/CLI)
Article • 06/30/2022 • 2 minutes to read

Imports metadata into a program compiled with /clr.

Syntax
#using file [ as_friend ]

Parameters
file

A Microsoft intermediate language (MSIL) .dll , .exe , .netmodule , or .obj file. For
example,

#using <MyComponent.dll>

as_friend

Specifies that all types in file are accessible. For more information, see Friend Assemblies
(C++).

Remarks
file can be a Microsoft intermediate language (MSIL) file that you import for its
managed data and managed constructs. If a DLL contains an assembly manifest, then all
the DLLs referenced in the manifest are imported. The assembly you're building will list
file in the metadata as an assembly reference.

Perhaps file doesn't contain an assembly (file is a module), and you don't intend to use
type information from the module in the current (assembly) application. You may
indicate the module is part of the assembly by using /ASSEMBLYMODULE. The types in
the module would then be available to any application that referenced the assembly.

An alternative to use #using is the /FU compiler option.

.exe assemblies passed to #using should be compiled by using one of the .NET Visual
Studio compilers (Visual Basic or Visual C#, for example). Attempting to import
metadata from an .exe assembly compiled with /clr will result in a file load exception.

７ Note
 A component that is referenced with #using can be run with a different version of
the file imported at compile time, causing a client application to give unexpected
results.

In order for the compiler to recognize a type in an assembly (not a module), it needs to
be forced to resolve the type. You can force it, for example, by defining an instance of
the type. There are other ways to resolve type names in an assembly for the compiler.
For example, if you inherit from a type in an assembly, the type name becomes known
to the compiler.

When importing metadata built from source code that used __declspec(thread), the
thread semantics aren't persisted in metadata. For example, a variable declared with
__declspec(thread) , compiled in a program that is built for the .NET Framework

common language runtime, and then imported via #using , won't have
__declspec(thread) semantics on the variable.

All imported types (both managed and native) in a file referenced by #using are
available, but the compiler treats native types as declarations, not definitions.

mscorlib.dll is automatically referenced when compiling with /clr .

The LIBPATH environment variable specifies the directories to search when the compiler
resolves file names passed to #using .

The compiler searches for references along the following path:

A path specified in the #using statement.

The current directory.

The .NET Framework system directory.

Directories added with the /AI compiler option.

Directories on LIBPATH environment variable.

Examples
You can build an assembly that references a second assembly that itself references a
third assembly. You only have to explicitly reference the third assembly from the first
one if you explicitly use one of its types.

Source file using_assembly_A.cpp :

 C++

// using_assembly_A.cpp

// compile with: /clr /LD

public ref class A {};

Source file using_assembly_B.cpp :

C++

// using_assembly_B.cpp

// compile with: /clr /LD

#using "using_assembly_A.dll"

public ref class B {

public:

void Test(A a) {}

void Test() {}

};

In the following sample, there's the compiler doesn't report an error about referencing
using_assembly_A.dll, because the program doesn't use any of the types defined in
using_assembly_A.cpp.

C++

// using_assembly_C.cpp

// compile with: /clr

#using "using_assembly_B.dll"

int main() {

B b;

b.Test();

See also
Preprocessor directives
Preprocessor operators
Article • 08/03/2021 • 2 minutes to read

Four preprocessor-specific operators are used in the context of the #define directive.
See the following table for a summary of each. The stringizing, charizing, and token-
pasting operators are discussed in the next three sections. For information on the
defined operator, see The #if, #elif, #else, and #endif directives.

Operator Action

Stringizing Causes the corresponding actual argument to be enclosed in double quotation

operator (#) marks

Charizing Causes the corresponding argument to be enclosed in single quotation marks

operator (#@) and to be treated as a character (Microsoft-specific)

Token-pasting Allows tokens used as actual arguments to be concatenated to form other

operator (##) tokens

defined Simplifies the writing of compound expressions in certain macro directives

operator

See also
Preprocessor directives

Predefined macros

c/c++ preprocessor reference

Stringizing operator (#)
Article • 08/03/2021 • 2 minutes to read

The number-sign or "stringizing" operator (#) converts macro parameters to string

literals without expanding the parameter definition. It's used only with macros that take
arguments. If it precedes a formal parameter in the macro definition, the actual
argument passed by the macro invocation is enclosed in quotation marks and treated as
a string literal. The string literal then replaces each occurrence of a combination of the
stringizing operator and formal parameter within the macro definition.

７ Note

The Microsoft C (versions 6.0 and earlier) extension to the ANSI C standard that
previously expanded macro formal arguments appearing inside string literals and
character constants is no longer supported. Code that relied on this extension
should be rewritten using the stringizing (#) operator.

White space that precedes the first token and follows the last token of the actual
argument is ignored. Any white space between the tokens in the actual argument is
reduced to a single white space in the resulting string literal. Thus, if a comment occurs
between two tokens in the actual argument, it's reduced to a single white space. The
resulting string literal is automatically concatenated with any adjacent string literals that
are separated only by white space.

Further, if a character contained in the argument usually requires an escape sequence

when used in a string literal, for example, the quotation mark ( " ) or backslash ( \ )
character, the necessary escape backslash is automatically inserted before the character.

The Microsoft C++ stringizing operator doesn't behave correctly when it's used with
strings that include escape sequences. In this situation, the compiler generates Compiler
Error C2017.

Examples
The following example shows a macro definition that includes the stringizing operator,
and a main function that invokes the macro:

C++

// stringizer.cpp

#include <stdio.h>

 #define stringer( x ) printf_s( #x "\n" )

int main() {

stringer( In quotes in the printf function call );

stringer( "In quotes when printed to the screen" );

stringer( "This: \" prints an escaped double quote" );

The stringer macros are expanded during preprocessing, producing the following
code:

C++

int main() {

printf_s( "In quotes in the printf function call" "\n" );

printf_s( "\"In quotes when printed to the screen\"" "\n" );

printf_s( "\"This: \\\" prints an escaped double quote\"" "\n" );

Output

In quotes in the printf function call

"In quotes when printed to the screen"

"This: \" prints an escaped double quote"

The following sample shows how you can expand a macro parameter:

C++

// stringizer_2.cpp

// compile with: /E

#define F abc

#define B def

#define FB(arg) #arg

#define FB1(arg) FB(arg)

FB(F B)

FB1(F B)

See also
Preprocessor operators
Charizing operator (#@)
Article • 08/03/2021 • 2 minutes to read

Microsoft Specific

The charizing operator can be used only with arguments of macros. If #@ precedes a
formal parameter in the definition of the macro, the actual argument is enclosed in
single quotation marks and treated as a character when the macro is expanded. For
example:

C++

#define makechar(x) #@x

causes the statement

C++

a = makechar(b);

to be expanded to

C++

a = 'b';

The single-quotation character ( ' ) can't be used with the charizing operator.

END Microsoft Specific

See also
Preprocessor operators
Token-pasting operator (##)
Article • 08/03/2021 • 2 minutes to read

The double-number-sign or token-pasting operator (##), which is sometimes called the

merging or combining operator, is used in both object-like and function-like macros. It
permits separate tokens to be joined into a single token, and therefore, can't be the first
or last token in the macro definition.

If a formal parameter in a macro definition is preceded or followed by the token-pasting

operator, the formal parameter is immediately replaced by the unexpanded actual
argument. Macro expansion is not performed on the argument prior to replacement.

Then, each occurrence of the token-pasting operator in token-string is removed, and the
tokens preceding and following it are concatenated. The resulting token must be a valid
token. If it is, the token is scanned for possible replacement if it represents a macro
name. The identifier represents the name by which the concatenated tokens will be
known in the program before replacement. Each token represents a token defined
elsewhere, either within the program or on the compiler command line. White space
preceding or following the operator is optional.

This example illustrates use of both the stringizing and token-pasting operators in
specifying program output:

C++

#define paster( n ) printf_s( "token" #n " = %d", token##n )

int token9 = 9;

If a macro is called with a numeric argument like

C++

paster( 9 );

the macro yields

C++

printf_s( "token" "9" " = %d", token9 );

which becomes

C++
 printf_s( "token9 = %d", token9 );

Example
C++

// preprocessor_token_pasting.cpp
#include <stdio.h>

#define paster( n ) printf_s( "token" #n " = %d", token##n )

int token9 = 9;

int main()

paster(9);

Output

token9 = 9

See also
Preprocessor operators
Macros (C/C++)
Article • 08/03/2021 • 2 minutes to read

The preprocessor expands macros in all lines except preprocessor directives, lines that
have a # as the first non-white-space character. It expands macros in parts of some
directives that aren't skipped as part of a conditional compilation. Conditional
compilation directives allow you to suppress compilation of parts of a source file. They
test a constant expression or identifier to determine which text blocks to pass on to the
compiler, and which ones to remove from the source file during preprocessing.

The #define directive is typically used to associate meaningful identifiers with constants,
keywords, and commonly used statements or expressions. Identifiers that represent
constants are sometimes called symbolic constants or manifest constants. Identifiers that
represent statements or expressions are called macros. In this preprocessor
documentation, only the term "macro" is used.

When the name of a macro is recognized in the program source text, or in the
arguments of certain other preprocessor commands, it's treated as a call to that macro.
The macro name is replaced by a copy of the macro body. If the macro accepts
arguments, the actual arguments following the macro name are substituted for formal
parameters in the macro body. The process of replacing a macro call with the processed
copy of the body is called expansion of the macro call.

In practical terms, there are two types of macros. Object-like macros take no arguments.
Function-like macros can be defined to accept arguments, so that they look and act like
function calls. Because macros don't generate actual function calls, you can sometimes
make programs run faster by replacing function calls with macros. (In C++, inline
functions are often a preferred method.) However, macros can create problems if you
don't define and use them with care. You may have to use parentheses in macro
definitions with arguments to preserve the proper precedence in an expression. Also,
macros may not correctly handle expressions with side effects. For more information,
see the getrandom example in The #define directive.

Once you've defined a macro, you can't redefine it to a different value without first
removing the original definition. However, you can redefine the macro with exactly the
same definition. Thus, the same definition may appear more than once in a program.

The #undef directive removes the definition of a macro. Once you've removed the
definition, you can redefine the macro to a different value. The #define directive and The
#undef directive discuss the #define and #undef directives, respectively.
For more information, see,

Macros and C++

Variadic Macros

Predefined macros

See also
C/C++ preprocessor reference
Macros and C++
Article • 08/03/2021 • 2 minutes to read

C++ offers new capabilities, some of which supplant the ones offered by the ANSI C
preprocessor. These new capabilities enhance the type safety and predictability of the
language:

In C++, objects declared as const can be used in constant expressions. It allows

programs to declare constants that have type and value information. They can
declare enumerations that can be viewed symbolically with the debugger. When
you use the preprocessor #define directive to define constants, it's not as precise,
and not type-safe. No storage is allocated for a const object, unless the program
contains an expression that takes its address.

The C++ inline function capability supplants function-type macros. The advantages
of using inline functions over macros are:

Type safety. Inline functions are subject to the same type checking as normal
functions. Macros aren't type-safe.

Correct handling of arguments that have side effects. Inline functions evaluate
the expressions supplied as arguments before the function body is entered.
Therefore, there's no chance that an expression with side effects will be unsafe.

For more information on inline functions, see inline, __inline, __forceinline.

For backward compatibility, all preprocessor facilities that existed in ANSI C and in
earlier C++ specifications are preserved for Microsoft C++.

See also
Predefined macros

Macros (C/C++)
Variadic macros
Article • 08/03/2021 • 2 minutes to read

Variadic macros are function-like macros that contain a variable number of arguments.

Remarks
To use variadic macros, the ellipsis may be specified as the final formal argument in a
macro definition, and the replacement identifier __VA_ARGS__ may be used in the
definition to insert the extra arguments. __VA_ARGS__ is replaced by all of the arguments
that match the ellipsis, including commas between them.

The C Standard specifies that at least one argument must be passed to the ellipsis to
ensure the macro doesn't resolve to an expression with a trailing comma. The traditional
Microsoft C++ implementation suppresses a trailing comma if no arguments are passed
to the ellipsis. When the /Zc:preprocessor compiler option is set, the trailing comma isn't
suppressed.

Example
C++

// variadic_macros.cpp

#include <stdio.h>

#define EMPTY

#define CHECK1(x, ...) if (!(x)) { printf(__VA_ARGS__); }

#define CHECK2(x, ...) if ((x)) { printf(__VA_ARGS__); }

#define CHECK3(...) { printf(__VA_ARGS__); }

#define MACRO(s, ...) printf(s, __VA_ARGS__)

int main() {

CHECK1(0, "here %s %s %s", "are", "some", "varargs1(1)\n");

CHECK1(1, "here %s %s %s", "are", "some", "varargs1(2)\n"); // won't

print

CHECK2(0, "here %s %s %s", "are", "some", "varargs2(3)\n"); // won't

print

CHECK2(1, "here %s %s %s", "are", "some", "varargs2(4)\n");

// always invokes printf in the macro

CHECK3("here %s %s %s", "are", "some", "varargs3(5)\n");

MACRO("hello, world\n");

 MACRO("error\n", EMPTY); // would cause error C2059, except VC++

// suppresses the trailing comma

Output

here are some varargs1(1)

here are some varargs2(4)

here are some varargs3(5)

hello, world

error

See also
Macros (C/C++)
Predefined macros
Article • 08/08/2022 • 16 minutes to read

The Microsoft C/C++ compiler (MSVC) predefines certain preprocessor macros,

depending on the language (C or C++), the compilation target, and the chosen compiler
options.

MSVC supports the predefined preprocessor macros required by the ANSI/ISO C99, C11,
and C17 standards, and the ISO C++14, C++17, and C++20 standards. The
implementation also supports several more Microsoft-specific preprocessor macros.
Some macros are defined only for specific build environments or compiler options.
Except where noted, the macros are defined throughout a translation unit as if they
were specified as /D compiler option arguments. When defined, the macros are
expanded to the specified values by the preprocessor before compilation. The
predefined macros take no arguments and can't be redefined.

Standard predefined identifier

The compiler supports this predefined identifier specified by ISO C99 and ISO C++11.

__func__ The unqualified and unadorned name of the enclosing function as a

function-local static const array of char .

C++

void example(){

printf("%s\n", __func__);

} // prints "example"

Standard predefined macros

The compiler supports these predefined macros specified by the ISO C99, C11, C17, and
ISO C++17 standards.

__cplusplus Defined as an integer literal value when the translation unit is

compiled as C++. Otherwise, undefined.

__DATE__ The compilation date of the current source file. The date is a constant

length string literal of the form Mmm dd yyyy. The month name Mmm is the same
as the abbreviated month name generated by the C Runtime Library (CRT) asctime
function. The first character of date dd is a space if the value is less than 10. This
macro is always defined.

__FILE__ The name of the current source file. __FILE__ expands to a character

string literal. To ensure that the full path to the file is displayed, use /FC (Full Path
of Source Code File in Diagnostics). This macro is always defined.

__LINE__ Defined as the integer line number in the current source file. The value of

the __LINE__ macro can be changed by using a #line directive. The integral type
of the value of __LINE__ can vary depending on context. This macro is always
defined.

__STDC__ Defined as 1 when compiled as C and if the /Za compiler option is

specified. Starting in Visual Studio 2022 version 17.2, it's also defined as 1 when
compiled as C and if the /std:c11 or /std:c17 compiler option is specified.
Otherwise, undefined.

__STDC_HOSTED__ Defined as 1 if the implementation is a hosted implementation,

one that supports the entire required standard library. Otherwise, defined as 0.

__STDC_NO_ATOMICS__ Defined as 1 if the implementation doesn't support optional

standard atomics. The MSVC implementation defines it as 1 when compiled as C

and one of the /std C11 or C17 options is specified.

__STDC_NO_COMPLEX__ Defined as 1 if the implementation doesn't support optional

standard complex numbers. The MSVC implementation defines it as 1 when
compiled as C and one of the /std C11 or C17 options is specified.

__STDC_NO_THREADS__ Defined as 1 if the implementation doesn't support optional

standard threads. The MSVC implementation defines it as 1 when compiled as C

and one of the /std C11 or C17 options is specified.

__STDC_NO_VLA__ Defined as 1 if the implementation doesn't support standard

variable length arrays. The MSVC implementation defines it as 1 when compiled as

C and one of the /std C11 or C17 options is specified.

__STDC_VERSION__ Defined when compiled as C and one of the /std C11 or C17

options is specified. It expands to 201112L for /std:c11, and 201710L for /std:c17.

__STDCPP_DEFAULT_NEW_ALIGNMENT__ When /std:c17 or later is specified, this macro

expands to a size_t literal that has the value of the alignment guaranteed by a
call to alignment-unaware operator new . Larger alignments are passed to an
 alignment-aware overload, such as operator new(std::size_t, std::align_val_t) .
For more information, see /Zc:alignedNew (C++17 over-aligned allocation).

__STDCPP_THREADS__ Defined as 1 if and only if a program can have more than one

thread of execution, and compiled as C++. Otherwise, undefined.

__TIME__ The time of translation of the preprocessed translation unit. The time is a

character string literal of the form hh:mm:ss, the same as the time returned by the
CRT asctime function. This macro is always defined.

Microsoft-specific predefined macros

MSVC supports these additional predefined macros.

__ATOM__ Defined as 1 when the /favor:ATOM compiler option is set and the

compiler target is x86 or x64. Otherwise, undefined.

__AVX__ Defined as 1 when the /arch:AVX, /arch:AVX2, or /arch:AVX512 compiler

options are set and the compiler target is x86 or x64. Otherwise, undefined.

__AVX2__ Defined as 1 when the /arch:AVX2 or /arch:AVX512 compiler option is set

and the compiler target is x86 or x64. Otherwise, undefined.

__AVX512BW__ Defined as 1 when the /arch:AVX512 compiler option is set and the
compiler target is x86 or x64. Otherwise, undefined.

__AVX512CD__ Defined as 1 when the /arch:AVX512 compiler option is set and the

compiler target is x86 or x64. Otherwise, undefined.

__AVX512DQ__ Defined as 1 when the /arch:AVX512 compiler option is set and the

compiler target is x86 or x64. Otherwise, undefined.

__AVX512F__ Defined as 1 when the /arch:AVX512 compiler option is set and the

compiler target is x86 or x64. Otherwise, undefined.

__AVX512VL__ Defined as 1 when the /arch:AVX512 compiler option is set and the
compiler target is x86 or x64. Otherwise, undefined.

_CHAR_UNSIGNED Defined as 1 if the default char type is unsigned. This value is

defined when the /J (Default char type is unsigned) compiler option is set.
Otherwise, undefined.

__CLR_VER Defined as an integer literal that represents the version of the Common
Language Runtime (CLR) used to compile the app. The value is encoded in the
form Mmmbbbbb , where M is the major version of the runtime, mm is the minor
version of the runtime, and bbbbb is the build number. __CLR_VER is defined if the
/clr compiler option is set. Otherwise, undefined.

C++

// clr_ver.cpp

// compile with: /clr

using namespace System;

int main() {

Console::WriteLine(__CLR_VER);
}

_CONTROL_FLOW_GUARD Defined as 1 when the /guard:cf (Enable Control Flow

Guard) compiler option is set. Otherwise, undefined.

__COUNTER__ Expands to an integer literal that starts at 0. The value is incremented

by 1 every time it's used in a source file, or in included headers of the source file.
__COUNTER__ remembers its state when you use precompiled headers. This macro is

always defined.

This example uses __COUNTER__ to assign unique identifiers to three different

objects of the same type. The exampleClass constructor takes an integer as a
parameter. In main , the application declares three objects of type exampleClass ,
using __COUNTER__ as the unique identifier parameter:

C++

// macro__COUNTER__.cpp

// Demonstration of __COUNTER__, assigns unique identifiers to

// different objects of the same type.

// Compile by using: cl /EHsc /W4 macro__COUNTER__.cpp

#include <stdio.h>

class exampleClass {

int m_nID;

public:

// initialize object with a read-only unique ID

exampleClass(int nID) : m_nID(nID) {}

int GetID(void) { return m_nID; }

};

int main()

// __COUNTER__ is initially defined as 0

exampleClass e1(__COUNTER__);

// On the second reference, __COUNTER__ is now defined as 1

 exampleClass e2(__COUNTER__);

// __COUNTER__ is now defined as 2

exampleClass e3(__COUNTER__);

printf("e1 ID: %i\n", e1.GetID());

printf("e2 ID: %i\n", e2.GetID());

printf("e3 ID: %i\n", e3.GetID());

// Output

// ------------------------------

// e1 ID: 0

// e2 ID: 1

// e3 ID: 2

return 0;

__cplusplus_cli Defined as the integer literal value 200406 when compiled as

C++ and a /clr compiler option is set. Otherwise, undefined. When defined,
__cplusplus_cli is in effect throughout the translation unit.

C++

// cplusplus_cli.cpp

// compile by using /clr

#include "stdio.h"

int main() {

#ifdef __cplusplus_cli

printf("%d\n", __cplusplus_cli);

#else

printf("not defined\n");

#endif

__cplusplus_winrt Defined as the integer literal value 201009 when compiled as

C++ and the /ZW (Windows Runtime Compilation) compiler option is set.
Otherwise, undefined.

_CPPRTTI Defined as 1 if the /GR (Enable Run-Time Type Information) compiler

option is set. Otherwise, undefined.

_CPPUNWIND Defined as 1 if one or more of the /GX (Enable Exception Handling),

/clr (Common Language Runtime Compilation), or /EH (Exception Handling

Model) compiler options are set. Otherwise, undefined.

_DEBUG Defined as 1 when the /LDd, /MDd, or /MTd compiler option is set.

Otherwise, undefined.
_DLL Defined as 1 when the /MD or /MDd (Multithreaded DLL) compiler option is

set. Otherwise, undefined.

__FUNCDNAME__ Defined as a string literal that contains the decorated name of the

enclosing function. The macro is defined only within a function. The __FUNCDNAME__
macro isn't expanded if you use the /EP or /P compiler option.

This example uses the __FUNCDNAME__ , __FUNCSIG__ , and __FUNCTION__ macros to

display function information.

C++

// Demonstrates functionality of __FUNCTION__, __FUNCDNAME__, and

__FUNCSIG__ macros

void exampleFunction()

printf("Function name: %s\n", __FUNCTION__);

printf("Decorated function name: %s\n", __FUNCDNAME__);

printf("Function signature: %s\n", __FUNCSIG__);

// Sample Output

// -------------------------------------------------

// Function name: exampleFunction

// Decorated function name: ?exampleFunction@@YAXXZ

// Function signature: void __cdecl exampleFunction(void)

__FUNCSIG__ Defined as a string literal that contains the signature of the enclosing

function. The macro is defined only within a function. The __FUNCSIG__ macro isn't
expanded if you use the /EP or /P compiler option. When compiled for a 64-bit
target, the calling convention is __cdecl by default. For an example of usage, see
the __FUNCDNAME__ macro.

__FUNCTION__ Defined as a string literal that contains the undecorated name of the

enclosing function. The macro is defined only within a function. The __FUNCTION__
macro isn't expanded if you use the /EP or /P compiler option. For an example of
usage, see the __FUNCDNAME__ macro.

_INTEGRAL_MAX_BITS Defined as the integer literal value 64, the maximum size (in

bits) for a non-vector integral type. This macro is always defined.

C++

// integral_max_bits.cpp

#include <stdio.h>

int main() {

 printf("%d\n", _INTEGRAL_MAX_BITS);

__INTELLISENSE__ Defined as 1 during an IntelliSense compiler pass in the Visual

Studio IDE. Otherwise, undefined. You can use this macro to guard code the
IntelliSense compiler doesn't understand, or use it to toggle between the build and
IntelliSense compiler. For more information, see Troubleshooting Tips for
IntelliSense Slowness .

_ISO_VOLATILE Defined as 1 if the /volatile:iso compiler option is set. Otherwise,

undefined.

_KERNEL_MODE Defined as 1 if the /kernel (Create Kernel Mode Binary) compiler

option is set. Otherwise, undefined.

_M_AMD64 Defined as the integer literal value 100 for compilations that target x64

processors or ARM64EC. Otherwise, undefined.

_M_ARM Defined as the integer literal value 7 for compilations that target ARM
processors. Undefined for ARM64, ARM64EC, and other targets.

_M_ARM_ARMV7VE Defined as 1 when the /arch:ARMv7VE compiler option is set for

compilations that target ARM processors. Otherwise, undefined.

_M_ARM_FP Defined as an integer literal value that indicates which /arch compiler
option was set for ARM processor targets. Otherwise, undefined.

A value in the range 30-39 if no /arch ARM option was specified, indicating the
default architecture for ARM was set ( VFPv3 ).

A value in the range 40-49 if /arch:VFPv4 was set.

For more information, see /arch (ARM).

_M_ARM64 Defined as 1 for compilations that target ARM64. Otherwise, undefined.

_M_ARM64EC Defined as 1 for compilations that target ARM64EC. Otherwise,

undefined.

_M_CEE Defined as 001 if any /clr (Common Language Runtime Compilation)

compiler option is set. Otherwise, undefined.

_M_CEE_PURE Deprecated beginning in Visual Studio 2015. Defined as 001 if the

/clr:pure compiler option is set. Otherwise, undefined.

_M_CEE_SAFE Deprecated beginning in Visual Studio 2015. Defined as 001 if the

/clr:safe compiler option is set. Otherwise, undefined.

_M_FP_CONTRACT Available beginning in Visual Studio 2022. Defined as 1 if the

/fp:contract or /fp:fast compiler option is set. Otherwise, undefined.

_M_FP_EXCEPT Defined as 1 if the /fp:except or /fp:strict compiler option is set.

Otherwise, undefined.

_M_FP_FAST Defined as 1 if the /fp:fast compiler option is set. Otherwise, undefined.

_M_FP_PRECISE Defined as 1 if the /fp:precise compiler option is set. Otherwise,

undefined.

_M_FP_STRICT Defined as 1 if the /fp:strict compiler option is set. Otherwise,

undefined.

_M_IX86 Defined as the integer literal value 600 for compilations that target x86
processors. This macro isn't defined for x64 or ARM compilation targets.

_M_IX86_FP Defined as an integer literal value that indicates the /arch compiler
option that was set, or the default. This macro is always defined when the
compilation target is an x86 processor. Otherwise, undefined. When defined, the
value is:

0 if the /arch:IA32 compiler option was set.

1 if the /arch:SSE compiler option was set.

2 if the /arch:SSE2 , /arch:AVX , /arch:AVX2 , or /arch:AVX512 compiler option

was set. This value is the default if an /arch compiler option wasn't specified.
When /arch:AVX is specified, the macro __AVX__ is also defined. When
/arch:AVX2 is specified, both __AVX__ and __AVX2__ are also defined. When

/arch:AVX512 is specified, __AVX__ , __AVX2__ , __AVX512BW__ , __AVX512CD__ ,

__AVX512DQ__ , __AVX512F__ , and __AVX512VL__ are also defined.

For more information, see /arch (x86).

_M_X64 Defined as the integer literal value 100 for compilations that target x64

processors or ARM64EC. Otherwise, undefined.

_MANAGED Defined as 1 when the /clr compiler option is set. Otherwise, undefined.
_MSC_BUILD Defined as an integer literal that contains the revision number element

of the compiler's version number. The revision number is the fourth element of the
period-delimited version number. For example, if the version number of the
Microsoft C/C++ compiler is 15.00.20706.01, the _MSC_BUILD macro evaluates to 1.
This macro is always defined.

_MSC_EXTENSIONS Defined as 1 if the on-by-default /Ze (Enable Language

Extensions) compiler option is set. Otherwise, undefined.

_MSC_FULL_VER Defined as an integer literal that encodes the major, minor, and

build number elements of the compiler's version number. The major number is the
first element of the period-delimited version number, the minor number is the
second element, and the build number is the third element. For example, if the
version number of the Microsoft C/C++ compiler is 15.00.20706.01, the
_MSC_FULL_VER macro evaluates to 150020706. Enter cl /? at the command line to

view the compiler's version number. This macro is always defined.

_MSC_VER Defined as an integer literal that encodes the major and minor number

elements of the compiler's version number. The major number is the first element
of the period-delimited version number and the minor number is the second
element. For example, if the version number of the Microsoft C/C++ compiler is
17.00.51106.1, the _MSC_VER macro evaluates to 1700. Enter cl /? at the command
line to view the compiler's version number. This macro is always defined.

Visual Studio version _MSC_VER

Visual Studio 6.0 1200

Visual Studio .NET 2002 (7.0) 1300

Visual Studio .NET 2003 (7.1) 1310

Visual Studio 2005 (8.0) 1400

Visual Studio 2008 (9.0) 1500

Visual Studio 2010 (10.0) 1600

Visual Studio 2012 (11.0) 1700

Visual Studio 2013 (12.0) 1800

Visual Studio 2015 (14.0) 1900

Visual Studio 2017 RTW (15.0) 1910

Visual Studio 2017 version 15.3 1911

 Visual Studio version _MSC_VER

Visual Studio 2017 version 15.5 1912

Visual Studio 2017 version 15.6 1913

Visual Studio 2017 version 15.7 1914

Visual Studio 2017 version 15.8 1915

Visual Studio 2017 version 15.9 1916

Visual Studio 2019 RTW (16.0) 1920

Visual Studio 2019 version 16.1 1921

Visual Studio 2019 version 16.2 1922

Visual Studio 2019 version 16.3 1923

Visual Studio 2019 version 16.4 1924

Visual Studio 2019 version 16.5 1925

Visual Studio 2019 version 16.6 1926

Visual Studio 2019 version 16.7 1927

Visual Studio 2019 version 16.8, 16.9 1928

Visual Studio 2019 version 16.10, 16.11 1929

Visual Studio 2022 RTW (17.0) 1930

Visual Studio 2022 version 17.1 1931

Visual Studio 2022 version 17.2 1932

To test for compiler releases or updates in a given version of Visual Studio or after,
use the >= operator. You can use it in a conditional directive to compare _MSC_VER
against that known version. If you have several mutually exclusive versions to
compare, order your comparisons in descending order of version number. For
example, this code checks for compilers released in Visual Studio 2017 and later.
Next, it checks for compilers released in or after Visual Studio 2015. Then it checks
for all compilers released before Visual Studio 2015:

C++

#if _MSC_VER >= 1910

// . . .

#elif _MSC_VER >= 1900

 // . . .

#else

// . . .

#endif

To test for compiler versions that share major and minor numbers, use the major,
minor, and build numbers in _MSC_FULL_VER for comparisons. The compilers in
Visual Studio 2019 version 16.9 have an _MSC_FULL_VER value of 192829500 or
greater. The compilers in Visual Studio 2019 version 16.11 have an _MSC_FULL_VER
value of 192930100 or greater.

For more information, see Visual C++ Compiler Version in the Microsoft C++
Team Blog.

_MSVC_LANG Defined as an integer literal that specifies the C++ language standard

targeted by the compiler. It's set only in code compiled as C++. The macro is the
integer literal value 201402L by default, or when the /std:c++14 compiler option is
specified. The macro is set to 201703L if the /std:c++17 compiler option is
specified. The macro is set to 202002L if the /std:c++20 compiler option is
specified. It's set to a higher, unspecified value when the /std:c++latest option is
specified. Otherwise, the macro is undefined. The _MSVC_LANG macro and /std
(Specify language standard version) compiler options are available beginning in
Visual Studio 2015 Update 3.

__MSVC_RUNTIME_CHECKS Defined as 1 when one of the /RTC compiler options is set.

Otherwise, undefined.

_MSVC_TRADITIONAL :
Available beginning with Visual Studio 2017 version 15.8: Defined as 0 when the
preprocessor conformance mode /experimental:preprocessor compiler option is
set. Defined as 1 by default, or when the /experimental:preprocessor- compiler
option is set, to indicate the traditional preprocessor is in use.
Available beginning with Visual Studio 2019 version 16.5: Defined as 0 when the
preprocessor conformance mode /Zc:preprocessor compiler option is set.
Defined as 1 by default, or when the /Zc:preprocessor- compiler option is set,
to indicate the traditional preprocessor is in use (essentially, /Zc:preprocessor
replaces the deprecated /experimental:preprocessor ).

C++

#if !defined(_MSVC_TRADITIONAL) || _MSVC_TRADITIONAL

// Logic using the traditional preprocessor

#else

 // Logic using cross-platform compatible preprocessor

#endif

_MT Defined as 1 when /MD or /MDd (Multithreaded DLL) or /MT or /MTd

(Multithreaded) is specified. Otherwise, undefined.

_NATIVE_WCHAR_T_DEFINED Defined as 1 when the /Zc:wchar_t compiler option is set.

Otherwise, undefined.

_OPENMP Defined as integer literal 200203, if the /openmp (Enable OpenMP 2.0

Support) compiler option is set. This value represents the date of the OpenMP
specification implemented by MSVC. Otherwise, undefined.

C++

// _OPENMP_dir.cpp

// compile with: /openmp

#include <stdio.h>

int main() {

printf("%d\n", _OPENMP);

_PREFAST_ Defined as 1 when the /analyze compiler option is set. Otherwise,

undefined.

__SANITIZE_ADDRESS__ Available beginning with Visual Studio 2019 version 16.9.

Defined as 1 when the /fsanitize=address compiler option is set. Otherwise,

undefined.

__TIMESTAMP__ Defined as a string literal that contains the date and time of the last

modification of the current source file, in the abbreviated, constant length form
returned by the CRT asctime function, for example, Fri 19 Aug 13:32:58 2016 . This
macro is always defined.

_VC_NODEFAULTLIB Defined as 1 when the /Zl (Omit Default Library Name)

compiler option is set. Otherwise, undefined.

_WCHAR_T_DEFINED Defined as 1 when the default /Zc:wchar_t compiler option is set.

The _WCHAR_T_DEFINED macro is defined but has no value if the /Zc:wchar_t-
compiler option is set, and wchar_t is defined in a system header file included in
your project. Otherwise, undefined.

_WIN32 Defined as 1 when the compilation target is 32-bit ARM, 64-bit ARM, x86,

or x64. Otherwise, undefined.

 _WIN64 Defined as 1 when the compilation target is 64-bit ARM or x64. Otherwise,

undefined.

_WINRT_DLL Defined as 1 when compiled as C++ and both /ZW (Windows Runtime

Compilation) and /LD or /LDd compiler options are set. Otherwise, undefined.

No preprocessor macros that identify the ATL or MFC library version are predefined by
the compiler. ATL and MFC library headers define these version macros internally.
They're undefined in preprocessor directives made before the required header is
included.

_ATL_VER Defined in <atldef.h> as an integer literal that encodes the ATL version
number.

_MFC_VER Defined in <afxver_.h> as an integer literal that encodes the MFC version

number.

See also
Macros (C/C++)

Preprocessor operators

Preprocessor directives
Preprocessor grammar summary
(C/C++)
Article • 08/03/2021 • 2 minutes to read

This article describes the formal grammar of the C and C++ preprocessor. It covers the
syntax of preprocessing directives and operators. For more information, see
Preprocessor and Pragma directives and the __pragma and _Pragma keywords.

Definitions for the grammar summary

Terminals are endpoints in a syntax definition. No other resolution is possible. Terminals
include the set of reserved words and user-defined identifiers.

Nonterminals are placeholders in the syntax. Most are defined elsewhere in this syntax
summary. Definitions can be recursive. The following nonterminals are defined in the
Lexical conventions section of the C++ Language Reference:

constant , constant-expression , identifier , keyword , operator , punctuator

An optional component is indicated by the subscripted opt. For example, the following
syntax indicates an optional expression enclosed in curly braces:

{ expression opt }

Document conventions
The conventions use different font attributes for different components of the syntax. The
symbols and fonts are as follows:

Attribute Description

nonterminal Italic type indicates nonterminals.

#include Terminals in bold type are literal reserved words and symbols that must be entered
as shown. Characters in this context are always case sensitive.

opt Nonterminals followed by opt are always optional.

default Characters in the set described or listed in this typeface can be used as terminals in
typeface statements.
A colon ( : ) following a nonterminal introduces its definition. Alternative definitions are
listed on separate lines.

In code syntax blocks, these symbols in the default typeface have a special meaning:

Symbol Description

[] Square brackets surround an optional element.

{|} Curly braces surround alternative elements, separated by vertical bars.

... Indicates the previous element pattern can be repeated.

In code syntax blocks, commas ( , ), periods ( . ), semi-colons ( ; ), colons ( : ),

parentheses ( ( ) ), double-quotes ( " ), and single-quotes ( ' ) are literals.

Preprocessor grammar
control-line :

  #define identifier token-string opt

  #define identifier ( identifier opt , ... , identifier opt ) token-string opt

  #include " path-spec "

  #include < path-spec >

  #line digit-sequence " filename " opt

  #undef identifier
  #error token-string

  #pragma token-string

constant-expression :

  defined( identifier )

  defined identifier

 any other constant expression

conditional :

  if-part elif-parts opt else-part opt endif-line

if-part :

  if-line text

if-line :

  #if constant-expression

  #ifdef identifier
  #ifndef identifier

elif-parts :

  elif-line text

  elif-parts elif-line text

elif-line :

  #elif constant-expression

else-part :

  else-line text

else-line :

  #else

endif-line :

  #endif

digit-sequence :

  digit

  digit-sequence digit

digit : one of

 0 1 2 3 4 5 6 7 8 9

token-string :

 String of token

token :

  keyword

  identifier

  constant

  operator

  punctuator

filename :

 Legal operating system filename

path-spec :

 Legal file path

text :

 Any sequence of text

７ Note

The following nonterminals are expanded in the Lexical conventions section of the
C++ Language Reference: constant , constant-expression , identifier , keyword ,
operator , and punctuator .

See also
C/C++ preprocessor reference
Pragma directives and the __pragma and
_Pragma keywords
Article • 03/31/2022 • 4 minutes to read

Pragma directives specify machine-specific or operating system-specific compiler

features. A line that starts with #pragma specifies a pragma directive. The Microsoft-
specific __pragma keyword enables you to code pragma directives within macro
definitions. The standard _Pragma preprocessor operator, introduced in C99 and
adopted by C++11, is similar.

Syntax
#pragma token-string

__pragma( token-string ) // two leading underscores - Microsoft-specific extension

_Pragma( string-literal ) // C99

Remarks
Each implementation of C and C++ supports some features unique to its host machine
or operating system. Some programs, for example, must exercise precise control over
the location of data in memory, or control the way certain functions receive parameters.
The #pragma directives offer a way for each compiler to offer machine- and operating
system-specific features, while maintaining overall compatibility with the C and C++
languages.

Pragma directives are machine-specific or operating system-specific by definition, and

are typically different for every compiler. A pragma can be used in a conditional
directive, to provide new preprocessor functionality. Or, use one to provide
implementation-defined information to the compiler.

The token-string is a series of characters representing a specific compiler instruction and

arguments, if any. The number sign ( # ) must be the first non-white-space character on
the line that contains the pragma. White-space characters can separate the number sign
and the word "pragma". Following #pragma , write any text that the translator can parse
as preprocessing tokens. The argument to #pragma is subject to macro expansion.

The string-literal is the input to _Pragma . Outer quotes and leading/trailing whitespace
are removed. \" is replaced with " and \\ is replaced with \ .
The compiler issues a warning when it finds a pragma that it doesn't recognize, and
continues compilation.

The Microsoft C and C++ compilers recognize the following pragma directives:

alloc_text

auto_inline

bss_seg

check_stack

code_seg

comment

component

conform 1

const_seg

data_seg

deprecated

detect_mismatch

endregion

fenv_access

float_control

fp_contract

function

hdrstop

include_alias

init_seg 1

inline_depth

inline_recursion

intrinsic

loop 1

make_public

managed

message

omp

once

optimize

pack

pointers_to_members 1

pop_macro

push_macro

region

runtime_checks

section

setlocale

strict_gs_check

system_header

unmanaged

vtordisp 1

warning

1 Supported only by the C++ compiler.

Pragma directives and compiler options

Some pragma directives provide the same functionality as compiler options. When a
pragma is reached in source code, it overrides the behavior specified by the compiler
option. For example, if you specified /Zp8, you can override this compiler setting for
specific sections of the code with pack:

cmd

cl /Zp8 some_file.cpp

C++

// some_file.cpp - packing is 8

// ...

#pragma pack(push, 1) - packing is now 1

// ...

#pragma pack(pop) - packing is 8 again

// ...

The __pragma keyword

The compiler also supports the Microsoft-specific __pragma keyword, which has the
same functionality as the #pragma directive. The difference is, the __pragma keyword is
usable inline in a macro definition. The #pragma directive isn't usable in a macro
definition, because the compiler interprets the number sign character ('#') in the
directive as the stringizing operator (#).

The following code example demonstrates how the __pragma keyword can be used in a
macro. This code is excerpted from the mfcdual.h header in the ACDUAL sample in
"Compiler COM Support Samples":

C++

#define CATCH_ALL_DUAL \

CATCH(COleException, e) \

{ \

_hr = e->m_sc; \

} \

AND_CATCH_ALL(e) \

{ \

__pragma(warning(push)) \

__pragma(warning(disable:6246)) /*disable _ctlState prefast warning*/ \

AFX_MANAGE_STATE(pThis->m_pModuleState); \

__pragma(warning(pop)) \

_hr = DualHandleException(_riidSource, e); \

} \

END_CATCH_ALL \

return _hr; \

The _Pragma preprocessing operator

_Pragma is similar to the Microsoft-specific __pragma keyword. It was introduced into

the C standard in C99, and the C++ standard in C++11. It's available in C only when you
specify the /std:c11 or /std:c17 option. For C++, it's available in all /std modes,
including the default.

Unlike #pragma , _Pragma allows you to put pragma directives into a macro definition.
The string literal should be what you would otherwise put following a #pragma
statement. For example:

#pragma message("the #pragma way")

_Pragma ("message( \"the _Pragma way\")")

Quotation marks and back-slashes should be escaped, as shown above. A pragma string
that isn't recognized is ignored.

The following code example demonstrates how the _Pragma keyword could be used in
an assert-like macro. It creates a pragma directive that suppresses a warning when the
condition expression happens to be constant.

The macro definition uses the do ... while(0) idiom for multi-statement macros so
that it can be used as though it were one statement. For more information, see C multi-
line macro on Stack Overflow. The _Pragma statement in the example only applies to
the line of code that follows it.

// Compile with /W4

#include <stdio.h>

#include <stdlib.h>

#define MY_ASSERT(BOOL_EXPRESSION) \

do { \

_Pragma("warning(suppress: 4127)") /* C4127 conditional expression

is constant */ \

if (!(BOOL_EXPRESSION)) { \

printf("MY_ASSERT FAILED: \"" #BOOL_EXPRESSION "\" on %s(%d)",

__FILE__, __LINE__); \

exit(-1); \

} \

} while (0)

int main()

MY_ASSERT(0 && "Note that there is no warning: C4127 conditional

expression is constant");

return 0;

See also
C/C++ preprocessor reference

C pragma directives

Keywords
alloc_text pragma
Article • 08/03/2021 • 2 minutes to read

Names the code section where the specified function definitions are placed. The pragma
must occur between a function declarator and the function definition for the named
functions.

Syntax
#pragma alloc_text( "text-section" , function_1 [ , function_2 ... ] )

Remarks
The alloc_text pragma doesn't handle C++ member functions or overloaded functions.
It's applicable only to functions declared with C linkage, that is, functions declared with
the extern "C" linkage specification. If you attempt to use this pragma on a function
with C++ linkage, a compiler error is generated.

Since function addressing using __based isn't supported, specifying section locations
requires the use of the alloc_text pragma. The name specified by text-section should
be enclosed in double quotation marks.

The alloc_text pragma must appear after the declarations of any of the specified
functions and before the definitions of these functions.

Functions referenced in an alloc_text pragma should be defined in the same module

as the pragma. Otherwise, if an undefined function is later compiled into a different text
section, the error may or may not be caught. Although the program will usually run
correctly, the function won't be allocated in the intended sections.

Other limitations on alloc_text are as follows:

It can't be used inside a function.

It must be used after the function has been declared, but before the function has
been defined.

See also
Pragma directives and the __pragma and _Pragma keywords
auto_inline pragma
Article • 08/03/2021 • 2 minutes to read

Excludes any functions defined within the range where off is specified from being
considered as candidates for automatic inline expansion.

Syntax
#pragma auto_inline( [ { on | off } ] )

Remarks
To use the auto_inline pragma, place it before and immediately after, not inside, a
function definition. The pragma takes effect as soon as the first function definition after
the pragma is seen.

See also
Pragma directives and the __pragma and _Pragma keywords
bss_seg pragma
Article • 08/03/2021 • 2 minutes to read

Specifies the section (segment) where uninitialized variables are stored in the object
(.obj) file.

Syntax
#pragma bss_seg( [ "section-name" [ , "section-class" ] ] )

#pragma bss_seg( { push | pop } [ , identifier ] [ , "section-name" [ , "section-class"

]] )

Parameters
push

(Optional) Puts a record on the internal compiler stack. A push can have an identifier
and section-name.

pop

(Optional) Removes a record from the top of the internal compiler stack. A pop can have
an identifier and section-name. You can pop multiple records using just one pop
command by using the identifier. The section-name becomes the active BSS section
name after the pop.

identifier

(Optional) When used with push , assigns a name to the record on the internal compiler
stack. When used with pop , the directive pops records off the internal stack until
identifier is removed. If identifier isn't found on the internal stack, nothing is popped.

"section-name"

(Optional) The name of a section. When used with pop , the stack is popped and section-
name becomes the active BSS section name.

"section-class"

(Optional) Ignored, but included for compatibility with versions of Microsoft C++ earlier
than version 2.0.

Remarks
A section in an object file is a named block of data that's loaded into memory as a unit.
A BSS section is a section that contains uninitialized data. In this article, the terms
segment and section have the same meaning.

The bss_seg pragma directive tells the compiler to put all uninitialized data items from
the translation unit into a BSS section named section-name. In some cases, use of
bss_seg can speed load times by grouping uninitialized data into one section. By

default, the BSS section used for uninitialized data in an object file is named .bss . A
bss_seg pragma directive without a section-name parameter resets the BSS section

name for the subsequent uninitialized data items to .bss .

Data allocated using the bss_seg pragma does not retain any information about its
location.

For a list of names that shouldn't be used to create a section, see /SECTION.

You can also specify sections for initialized data (data_seg), functions (code_seg), and
const variables (const_seg).

You can use the DUMPBIN.EXE application to view object files. Versions of DUMPBIN for
each supported target architecture are included with Visual Studio.

Example
C++

// pragma_directive_bss_seg.cpp

int i; // stored in .bss

#pragma bss_seg(".my_data1")

int j; // stored in .my_data1

#pragma bss_seg(push, stack1, ".my_data2")

int l; // stored in .my_data2

#pragma bss_seg(pop, stack1) // pop stack1 from stack

int m; // stored in .my_data1

int main() {

See also
Pragma directives and the __pragma and _Pragma keywords
check_stack pragma
Article • 08/03/2021 • 2 minutes to read

Instructs the compiler to turn off stack probes if off (or - ) is specified, or to turn on
stack probes if on (or + ) is specified.

Syntax
#pragma check_stack( [{ on | off }] )

#pragma check_stack { + | - }

Remarks
This pragma takes effect at the first function defined after the pragma is seen. Stack
probes are neither a part of macros nor of functions that are generated inline.

If you don't give an argument for the check_stack pragma, stack checking reverts to the
behavior specified on the command line. For more information, see Compiler options.
The interaction of the #pragma check_stack and the /Gs option is summarized in the
following table.

Using the check_stack Pragma

Syntax Compiled with

Action

/Gs option?

#pragma check_stack( ) or
Yes Turns off stack checking for functions that follow

#pragma check_stack

#pragma check_stack( ) or
No Turns on stack checking for functions that follow

#pragma check_stack

#pragma check_stack(on)
Yes or No Turns on stack checking for functions that follow

or #pragma check_stack +
 Syntax Compiled with
Action

/Gs option?

#pragma check_stack(off)
Yes or No Turns off stack checking for functions that follow

or #pragma check_stack -

See also
Pragma directives and the __pragma and _Pragma keywords
code_seg pragma
Article • 08/03/2021 • 2 minutes to read

Specifies the text section (segment) where functions are stored in the object (.obj) file.

Syntax
#pragma code_seg( [ "section-name" [ , "section-class" ] ] )

#pragma code_seg( { push | pop } [ , identifier ] [ , "section-name" [ , "section-

class" ] ] )

Parameters
push

(Optional) Puts a record on the internal compiler stack. A push can have an identifier
and section-name.

pop

(Optional) Removes a record from the top of the internal compiler stack. A pop can have
an identifier and section-name. You can pop multiple records using just one pop
command by using the identifier. The section-name becomes the active text section
name after the pop.

identifier

(Optional) When used with push , assigns a name to the record on the internal compiler
stack. When used with pop , the directive pops records off the internal stack until
identifier is removed. If identifier isn't found on the internal stack, nothing is popped.

"section-name"

(Optional) The name of a section. When used with pop , the stack is popped and section-
name becomes the active text section name.

"section-class"

(Optional) Ignored, but included for compatibility with versions of Microsoft C++ earlier
than version 2.0.

Remarks
A section in an object file is a named block of data that's loaded into memory as a unit.
A text section is a section that contains executable code. In this article, the terms
segment and section have the same meaning.

The code_seg pragma directive tells the compiler to put all subsequent object code from
the translation unit into a text section named section-name. By default, the text section
used for functions in an object file is named .text . A code_seg pragma directive
without a section-name parameter resets the text section name for the subsequent
object code to .text .

The code_seg pragma directive doesn't control placement of object code generated for
instantiated templates. Nor does it control code generated implicitly by the compiler,
such as special member functions. To control that code, we recommend you use the
__declspec(code_seg(...)) attribute instead. It gives you control over placement of all
object code, including compiler-generated code.

For a list of names that shouldn't be used to create a section, see /SECTION.

You can also specify sections for initialized data (data_seg), uninitialized data (bss_seg),
and const variables (const_seg).

You can use the DUMPBIN.EXE application to view object files. Versions of DUMPBIN for
each supported target architecture are included with Visual Studio.

Example
This example shows how to use the code_seg pragma directive to control where object
code is put:

C++

// pragma_directive_code_seg.cpp

void func1() { // stored in .text

#pragma code_seg(".my_data1")

void func2() { // stored in my_data1

#pragma code_seg(push, r1, ".my_data2")

void func3() { // stored in my_data2

#pragma code_seg(pop, r1) // stored in my_data1

void func4() {

 int main() {

See also
code_seg (__declspec)

Pragma directives and the __pragma and _Pragma keywords

comment pragma
Article • 08/03/2021 • 2 minutes to read

Places a comment record into an object file or executable file.

Syntax
#pragma comment( comment-type [ , "comment-string" ] )

Remarks
The comment-type is one of the predefined identifiers, described below, that specifies
the type of comment record. The optional comment-string is a string literal that provides
additional information for some comment types. Because comment-string is a string
literal, it obeys all the rules for string literals on use of escape characters, embedded
quotation marks ( " ), and concatenation.

compiler
Places the name and version number of the compiler in the object file. This comment
record is ignored by the linker. If you supply a comment-string parameter for this record
type, the compiler generates a warning.

lib
Places a library-search record in the object file. This comment type must be
accompanied by a comment-string parameter that has the name (and possibly the path)
of the library that you want the linker to search. The library name follows the default
library-search records in the object file. The linker searches for this library the same way
as if you specified it on the command line, as long as the library isn't specified by using
/nodefaultlib. You can place multiple library-search records in the same source file. Each
record appears in the object file in the same order it's found in the source file.

If the order of the default library and an added library is important, compiling with the
/Zl switch will prevent the default library name from being placed in the object module.
A second comment pragma then can be used to insert the name of the default library
after the added library. The libraries listed with these pragma directives will appear in
the object module in the same order they're found in the source code.
linker
Places a linker option in the object file. You can use this comment-type to specify a
linker option instead of passing it to the command line or specifying it in the
development environment. For example, you can specify the /include option to force the
inclusion of a symbol:

#pragma comment(linker, "/include:__mySymbol")

Only the following (comment-type) linker options are available to be passed to the linker
identifier:

/DEFAULTLIB

/EXPORT

/INCLUDE

/MANIFESTDEPENDENCY

/MERGE

/SECTION

user
Places a general comment in the object file. The comment-string parameter contains the
text of the comment. This comment record is ignored by the linker.

Examples
The following pragma causes the linker to search for the EMAPI.LIB library while linking.
The linker searches first in the current working directory, and then in the path specified
in the LIB environment variable.

#pragma comment( lib, "emapi" )

The following pragma causes the compiler to place the name and version number of the
compiler in the object file:
 C

#pragma comment( compiler )

For comments that take a comment-string parameter, you can use a macro in any place
where you would use a string literal, as long as the macro expands to a string literal. You
can also concatenate any combination of string literals and macros that expand to string
literals. For example, the following statement is acceptable:

#pragma comment( user, "Compiled on " __DATE__ " at " __TIME__ )

See also
Pragma directives and the __pragma and _Pragma keywords
component pragma
Article • 08/03/2021 • 2 minutes to read

Controls the collection of browse information or dependency information from within

source files.

Syntax
#pragma component( browser, { on | off } [ , references [ , name ] ] )

#pragma component( minrebuild, { on | off } )

#pragma component( mintypeinfo, { on | off } )

Remarks

Browser
You can turn collecting on or off, and you can specify particular names to be ignored as
information is collected.

Using on or off controls the collection of browse information from the pragma forward.
For example:

C++

#pragma component(browser, off)

stops the compiler from collecting browse information.

７ Note

To turn on the collecting of browse information with this pragma, browse

information must first be enabled.

The references option can be used with or without the name argument. Using
references without name turns on or off the collecting of references (other browse

information continues to be collected, however). For example:

C++
 #pragma component(browser, off, references)

stops the compiler from collecting reference information.

Using references with name and off prevents references to name from appearing in
the browse information window. Use this syntax to ignore names and types you are not
interested in and to reduce the size of browse information files. For example:

C++

#pragma component(browser, off, references, DWORD)

ignores references to DWORD from that point forward. You can turn collecting of
references to DWORD back on by using on :

C++

#pragma component(browser, on, references, DWORD)

This is the only way to resume collecting references to name; you must explicitly turn on
any name that you have turned off.

To prevent the preprocessor from expanding name (such as expanding NULL to 0), put
quotes around it:

C++

#pragma component(browser, off, references, "NULL")

Minimal rebuild
The deprecated /Gm (Enable Minimal Rebuild) feature requires the compiler to create
and store C++ class dependency information, which takes disk space. To save disk
space, you can use #pragma component( minrebuild, off ) whenever you don't need to
collect dependency information, for instance, in unchanging header files. Insert #pragma
component( minrebuild, on ) after unchanging classes to turn dependency collection

back on.

Reduce type information

The mintypeinfo option reduces the debugging information for the region specified.
The volume of this information is considerable, impacting .pdb and .obj files. You cannot
debug classes and structures in the mintypeinfo region. Use of the mintypeinfo option
can be helpful to avoid the following warning:

cmd

LINK : warning LNK4018: too many type indexes in PDB "filename", discarding
subsequent type information

For more information, see the /Gm (Enable Minimal Rebuild) compiler option.

See also
Pragma directives and the __pragma and _Pragma keywords
conform pragma
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Specifies the run-time behavior of the /Zc:forScope compiler option.

Syntax
#pragma conform( name [ , show ] [ , { on | off } ] [ [ , { push | pop } ] [ ,

identifier [ , { on | off } ] ] ] )

Parameters
name

Specifies the name of the compiler option to be modified. The only valid name is
forScope .

show

(Optional) Causes the current setting of name (true or false) to be displayed by means of
a warning message during compilation. For example, #pragma conform(forScope, show) .

on , off

(Optional) Setting name to on enables the /Zc:forScope compiler option. The default is
off .

push

(Optional) Pushes the current value of name onto the internal compiler stack. If you
specify identifier, you can specify the on or off value for name to be pushed onto the
stack. For example, #pragma conform(forScope, push, myname, on) .

pop

(Optional) Sets the value of name to the value at the top of the internal compiler stack
and then pops the stack. If identifier is specified with pop , the stack will be popped back
until it finds the record with identifier, which will also be popped; the current value for
name in the next record on the stack becomes the new value for name. If you specify
pop with an identifier that is not in a record on the stack, the pop is ignored.

identifier

(Optional) Can be included with a push or pop command. If identifier is used, then an
on or off specifier can also be used.

Example
C++

// pragma_directive_conform.cpp

// compile with: /W1

// C4811 expected

#pragma conform(forScope, show)

#pragma conform(forScope, push, x, on)

#pragma conform(forScope, push, x1, off)

#pragma conform(forScope, push, x2, off)

#pragma conform(forScope, push, x3, off)

#pragma conform(forScope, show)

#pragma conform(forScope, pop, x1)

#pragma conform(forScope, show)

int main() {}

See also
Pragma directives and the __pragma and _Pragma keywords
const_seg pragma
Article • 08/03/2021 • 2 minutes to read

Specifies the section (segment) where const variables are stored in the object (.obj) file.

Syntax
#pragma const_seg( [ "section-name" [ , "section-class" ] ] )

#pragma const_seg( { push | pop } [ , identifier ] [ , "section-name" [ , "section-

class" ] ] )

Parameters
push

(Optional) Puts a record on the internal compiler stack. A push can have an identifier
and section-name.

pop

(Optional) Removes a record from the top of the internal compiler stack. A pop can have
an identifier and section-name. You can pop multiple records using just one pop
command by using the identifier. The section-name becomes the active const section
name after the pop.

identifier

(Optional) When used with push , assigns a name to the record on the internal compiler
stack. When used with pop , the directive pops records off the internal stack until
identifier is removed. If identifier isn't found on the internal stack, nothing is popped.

"section-name"

(Optional) The name of a section. When used with pop , the stack is popped and section-
name becomes the active const section name.

"section-class"

(Optional) Ignored, but included for compatibility with versions of Microsoft C++ earlier
than version 2.0.

Remarks
A section in an object file is a named block of data that's loaded into memory as a unit.
A const section is a section that contains constant data. In this article, the terms segment
and section have the same meaning.

The const_seg pragma directive tells the compiler to put all constant data items from
the translation unit into a const section named section-name. The default section in the
object file for const variables is .rdata . Some const variables, such as scalars, are
automatically inlined into the code stream. Inlined code doesn't appear in .rdata . A
const_seg pragma directive without a section-name parameter resets the section name

for the subsequent const data items to .rdata .

If you define an object that requires dynamic initialization in a const_seg , the result is
undefined behavior.

For a list of names that shouldn't be used to create a section, see /SECTION.

You can also specify sections for initialized data (data_seg), uninitialized data (bss_seg),
and functions (code_seg).

You can use the DUMPBIN.EXE application to view object files. Versions of DUMPBIN for
each supported target architecture are included with Visual Studio.

Example
C++

// pragma_directive_const_seg.cpp
// compile with: /EHsc

#include <iostream>

const int i = 7; // inlined, not stored in .rdata

const char sz1[]= "test1"; // stored in .rdata

#pragma const_seg(".my_data1")

const char sz2[]= "test2"; // stored in .my_data1

#pragma const_seg(push, stack1, ".my_data2")

const char sz3[]= "test3"; // stored in .my_data2

#pragma const_seg(pop, stack1) // pop stack1 from stack

const char sz4[]= "test4"; // stored in .my_data1

int main() {

using namespace std;

// const data must be referenced to be put in .obj

cout << sz1 << endl;

cout << sz2 << endl;

 cout << sz3 << endl;

cout << sz4 << endl;

Output

test1

test2

test3

test4

See also
Pragma directives and the __pragma and _Pragma keywords
data_seg pragma
Article • 08/03/2021 • 2 minutes to read

Specifies the data section (segment) where initialized variables are stored in the object
(.obj) file.

Syntax
#pragma data_seg( [ "section-name" [ , "section-class" ] ] )

#pragma data_seg( { push | pop } [ , identifier ] [ , "section-name" [ , "section-

class" ] ] )

Parameters
push

(Optional) Puts a record on the internal compiler stack. A push can have an identifier
and section-name.

pop

(Optional) Removes a record from the top of the internal compiler stack. A pop can have
an identifier and section-name. You can pop multiple records using just one pop
command by using the identifier. The section-name becomes the active data section
name after the pop.

identifier

(Optional) When used with push , assigns a name to the record on the internal compiler
stack. When used with pop , pops records off the internal stack until identifier is
removed. If identifier isn't found on the internal stack, nothing is popped.

identifier enables multiple records to be popped with a single pop command.

"section-name"

(Optional) The name of a section. When used with pop , the stack is popped and section-
name becomes the active data section name.

"section-class"

(Optional) Ignored, but included for compatibility with versions of Microsoft C++ earlier
than version 2.0.
Remarks
A section in an object file is a named block of data that's loaded into memory as a unit.
A data section is a section that contains initialized data. In this article, the terms segment
and section have the same meaning.

The default section in the .obj file for initialized variables is .data . Variables that are
uninitialized are considered to be initialized to zero and are stored in .bss .

The data_seg pragma directive tells the compiler to put all initialized data items from
the translation unit into a data section named section-name. By default, the data section
used for initialized data in an object file is named .data . Variables that are uninitialized
are considered to be initialized to zero, and are stored in .bss . A data_seg pragma
directive without a section-name parameter resets the data section name for the
subsequent initialized data items to .data .

Data allocated using data_seg doesn't retain any information about its location.

For a list of names that shouldn't be used to create a section, see /SECTION.

You can also specify sections for const variables (const_seg), uninitialized data (bss_seg),
and functions (code_seg).

You can use the DUMPBIN.EXE application to view object files. Versions of DUMPBIN for
each supported target architecture are included with Visual Studio.

Example
C++

// pragma_directive_data_seg.cpp

int h = 1; // stored in .data

int i = 0; // stored in .bss

#pragma data_seg(".my_data1")

int j = 1; // stored in .my_data1

#pragma data_seg(push, stack1, ".my_data2")

int l = 2; // stored in .my_data2

#pragma data_seg(pop, stack1) // pop stack1 off the stack

int m = 3; // stored in .my_data1

int main() {

See also
Pragma directives and the __pragma and _Pragma keywords
deprecated pragma
Article • 08/03/2021 • 2 minutes to read

The deprecated pragma lets you indicate that a function, type, or any other identifier
may no longer be supported in a future release or should no longer be used.

７ Note

For information about the C++14 [[deprecated]] attribute, and guidance on when
to use that attribute instead of the Microsoft __declspec(deprecated) modifier or
the deprecated pragma, see Attributes in C++.

Syntax
#pragma deprecated( identifier1 [ , identifier2 ... ] )

Remarks
When the compiler encounters an identifier specified by a deprecated pragma, it issues
compiler warning C4995.

You can deprecate macro names. Place the macro name in quotes or else macro
expansion will occur.

Because the deprecated pragma works on all matching identifiers, and does not take
signatures into account, it is not the best option for deprecating specific versions of
overloaded functions. Any matching function name that is brought into scope triggers
the warning.

We recommend you use the C++14 [[deprecated]] attribute, when possible, instead of
the deprecated pragma. The Microsoft-specific __declspec(deprecated) declaration
modifier is also a better choice in many cases than the deprecated pragma. The
[[deprecated]] attribute and __declspec(deprecated) modifier allow you to specify

deprecated status for particular forms of overloaded functions. The diagnostic warning
only appears on references to the specific overloaded function the attribute or modifier
applies to.

Example
 C++

// pragma_directive_deprecated.cpp

// compile with: /W3

#include <stdio.h>

void func1(void) {

void func2(void) {

int main() {

func1();

func2();

#pragma deprecated(func1, func2)

func1(); // C4995

func2(); // C4995

The following sample shows how to deprecate a class:

C++

// pragma_directive_deprecated2.cpp

// compile with: /W3

#pragma deprecated(X)

class X { // C4995

public:

void f(){}

};

int main() {

X x; // C4995

See also
Pragma directives and the __pragma and _Pragma keywords
detect_mismatch pragma
Article • 08/03/2021 • 2 minutes to read

Places a record in an object. The linker checks these records for potential mismatches.

Syntax
#pragma detect_mismatch( "name" , "value" )

Remarks
When you link the project, the linker throws a LNK2038 error if the project contains two
objects that have the same name but each has a different value. Use this pragma to
prevent inconsistent object files from linking.

Both name and value are string literals and obey the rules for string literals with respect
to escape characters and concatenation. They are case-sensitive and cannot contain a
comma, equal sign, quotation marks, or the null character.

Example
This example creates two files that have different version numbers for the same version
label.

C++

// pragma_directive_detect_mismatch_a.cpp

#pragma detect_mismatch("myLib_version", "9")

int main ()

return 0;

// pragma_directive_detect_mismatch_b.cpp

#pragma detect_mismatch("myLib_version", "1")

If you compile both of these files by using the command line cl

pragma_directive_detect_mismatch_a.cpp pragma_directive_detect_mismatch_b.cpp , you

will receive the error LNK2038.

See also
Pragma directives and the __pragma and _Pragma keywords
execution_character_set pragma
Article • 02/01/2022 • 2 minutes to read

Specifies the execution character set used for string and character literals. This directive
isn't needed for literals marked with the u8 prefix.

Syntax
#pragma execution_character_set( " target " )

Parameters
target

Specifies the target execution character set. Currently the only target execution set
supported is "utf-8" .

Remarks
This compiler directive is obsolete in Visual Studio 2015 Update 2 and later versions. We
recommend that you use the /execution-charset:utf-8 or /utf-8 compiler options
together with the u8 prefix on narrow character and string literals that contain extended
characters. For more information about the u8 prefix, see String and character literals.
For more information about the compiler options, see /execution-charset (Set execution
character set) and /utf-8 (Set source and execution character sets to UTF-8).

The #pragma execution_character_set("utf-8") directive tells the compiler to encode

narrow character and narrow string literals in your source code as UTF-8 in the
executable. This output encoding is independent of how the source file is encoded.

By default, when the compiler encodes narrow characters and narrow strings, it uses the
current code page as the execution character set. Unicode or DBCS characters outside
the range of the current code page get converted to the default replacement character
in the output. Unicode and DBCS characters are truncated to their low-order byte, which
is almost never what you intend. To specify the UTF-8 encoding for literals in the source
file, use a u8 prefix. The compiler passes these UTF-8 encoded strings to the output
unchanged. Narrow character literals prefixed by u8 must fit in a byte, or they're
truncated on output.
By default, Visual Studio uses the current code page as the source character set used to
interpret your source code for output. When a file is read in, Visual Studio interprets it
according to the current code page unless the file has a code page set. Or, unless a
byte-order mark (BOM) or UTF-16 characters are detected at the beginning of the file.
You can't set UTF-8 as the current code page in some versions of Windows. When the
automatic detection finds UTF-8 encoded source files without a BOM in those versions,
Visual Studio assumes they're encoded in the current code page. Characters in the
source file that are outside the range of the specified or automatically detected code
page can cause compiler warnings and errors.

See also
Pragma directives and the __pragma and _Pragma keywords

/execution-charset (Set execution character set)

/utf-8 (Set source and execution character sets to UTF-8)

fenv_access pragma
Article • 10/25/2021 • 2 minutes to read

Disables ( on ) or enables ( off ) optimizations that could change floating-point

environment flag tests and mode changes.

Syntax
#pragma fenv_access ( { on | off } )

Remarks
By default, fenv_access is off . The compiler assumes your code doesn't access or
manipulate the floating-point environment. If environment access isn't required, the
compiler can do more to optimize your floating-point code.

Enable fenv_access if your code tests floating-point status flags, exceptions, or sets
control mode flags. The compiler disables floating-point optimizations, so your code can
access the floating-point environment consistently.

The /fp:strict command-line option automatically enables fenv_access . For more

information on this and other floating-point behavior, see /fp (Specify Floating-Point
Behavior).

There are restrictions on the ways you can use the fenv_access pragma in combination
with other floating-point settings:

You can't enable fenv_access unless precise semantics are enabled. Precise
semantics can be enabled either by the float_control pragma, or by using the
/fp:precise or /fp:strict compiler options. The compiler defaults to /fp:precise if
no other floating-point command-line option is specified.

You can't use float_control to disable precise semantics when fenv_access(on) is

set.

The fenv_access(on) directive disables generation of floating-point contractions,

machine instructions that combine floating-point operations. fenv_access(off) restores
the previous behavior for contractions. This behavior is new in Visual Studio 2022.
Previous compiler versions could generate contractions by default under
fenv_access(on) . For more information about floating-point contractions, see

/fp:contract.

The kinds of optimizations that are subject to fenv_access are:

Global common subexpression elimination

Code motion

Constant folding

Other floating-point pragma directives include:

float_control

fp_contract

Examples
This example sets fenv_access to on to set the floating-point control register for 24-bit
precision:

C++

// pragma_directive_fenv_access_x86.cpp

// compile with: /O2 /arch:IA32

// processor: x86

#include <stdio.h>

#include <float.h>

#include <errno.h>

#pragma fenv_access (on)

int main() {

double z, b = 0.1, t = 0.1;

unsigned int currentControl;

errno_t err;

err = _controlfp_s(&currentControl, _PC_24, _MCW_PC);

if (err != 0) {

printf_s("The function _controlfp_s failed!\n");

return -1;

z = b * t;

printf_s ("out=%.15e\n",z);

Output

out=9.999999776482582e-03

If you comment out #pragma fenv_access (on) from the previous sample, the output is
different. It's because the compiler does compile-time evaluation, which doesn't use the
control mode.

C++

// pragma_directive_fenv_access_2.cpp

// compile with: /O2 /arch:IA32

#include <stdio.h>

#include <float.h>

int main() {

double z, b = 0.1, t = 0.1;

unsigned int currentControl;

errno_t err;

err = _controlfp_s(&currentControl, _PC_24, _MCW_PC);

if (err != 0) {

printf_s("The function _controlfp_s failed!\n");

return -1;

z = b * t;

printf_s ("out=%.15e\n",z);

Output

out=1.000000000000000e-02

See also
Pragma directives and the __pragma and _Pragma keywords
float_control pragma
Article • 10/25/2021 • 3 minutes to read

Specifies floating-point behavior for a function.

Syntax
#pragma float_control

#pragma float_control( precise, { on | off } [ , push ] )

#pragma float_control( except, { on | off } [ , push ] )

#pragma float_control( { push | pop } )

Options
precise , on | off , push

Specifies whether to enable ( on ) or disable ( off ) precise floating-point semantics. For

information on differences with the /fp:precise compiler option, see the Remarks
section. The optional push token pushes the current setting for float_control on the
internal compiler stack.

except , on | off , push

Specifies whether to enable ( on ) or disable ( off ) floating-point exception semantics.

The optional push token pushes the current setting for float_control on the internal
compiler stack.

except can only be set to on when precise is also set to on .

push

Pushes the current float_control setting on to the internal compiler stack.

pop

Removes the float_control setting from the top of the internal compiler stack, and
makes that the new float_control setting.

Remarks
The float_control pragma doesn't have the same behavior as the /fp compiler option.
The float_control pragma only governs part of the floating-point behavior. It must be
combined with fp_contract and fenv_access pragma directives to recreate the /fp
compiler options. The following table shows the equivalent pragma settings for each
compiler option:

Option float_control(precise, float_control(except, fp_contract(*) fenv_access(*)

*) *)

/fp:strict on on off on

/fp:precise on off off * off

/fp:fast off off on off

* In versions of Visual Studio before Visual Studio 2022, the /fp:precise behavior
defaulted to fp_contract(on) .

In other words, you may need to use several pragma directives in combination to
emulate the /fp:fast , /fp:precise , and /fp:strict command-line options.

There are restrictions on the ways you can use the float_control and fenv_access
floating-point pragma directives in combination:

You can only use float_control to set except to on if precise semantics are
enabled. Precise semantics can be enabled either by the float_control pragma, or
by using the /fp:precise or /fp:strict compiler options.

You can't use float_control to turn off precise when exception semantics are
enabled, whether by a float_control pragma or a /fp:except compiler option.

You can't enable fenv_access unless precise semantics are enabled, whether by a
float_control pragma or a compiler option.

You can't use float_control to turn off precise when fenv_access is enabled.

These restrictions mean the order of some floating-point pragma directives is

significant. To go from a fast model to a strict model using pragma directives, use the
following code:

C++

#pragma float_control(precise, on) // enable precise semantics

#pragma fenv_access(on) // enable environment sensitivity

#pragma float_control(except, on) // enable exception semantics

To go from a strict model to a fast model by using the float_control pragma, use the
following code:

C++

#pragma float_control(except, off) // disable exception semantics

#pragma fenv_access(off) // disable environment sensitivity

#pragma float_control(precise, off) // disable precise semantics

#pragma fp_contract(on) // enable contractions

If no options are specified, float_control has no effect.

The float_control directive disables contractions when it turns on precise or except .

Use of float_control to turn off precise or except restores the previous setting for
contractions. You can use the fp_contract pragma directive to change the compiler
behavior on contractions. float_control(push) and float_control(pop) push and pop
the setting for contractions as part of the float_control setting on to the internal
compiler stack. This behavior is new in Visual Studio 2022. The float_control directive
in previous compiler versions did not affect contraction settings.

Example
The following sample shows how to catch an overflow floating-point exception by using
pragma float_control .

C++

// pragma_directive_float_control.cpp

// compile with: /EHa

#include <stdio.h>

#include <float.h>

double func( ) {

return 1.1e75;

#pragma float_control (except, on)

int main( ) {
float u[1];

unsigned int currentControl;

errno_t err;

err = _controlfp_s(&currentControl, ~_EM_OVERFLOW, _MCW_EM);

if (err != 0)

printf_s("_controlfp_s failed!\n");

 try {

u[0] = func();

printf_s ("Fail");

return(1);

catch (...) {

printf_s ("Pass");

return(0);

Output

Pass

See also
Pragma directives and the __pragma and _Pragma keywords

fenv_access pragma

fp_contract pragma
fp_contract pragma
Article • 03/31/2022 • 2 minutes to read

Determines whether floating-point contraction takes place. A floating-point contraction

is an instruction such as Fused-Multiply-Add (FMA) that combines two separate floating
point operations into a single instruction. Use of these instructions can affect floating-
point precision, because instead of rounding after each operation, the processor may
round only once after both operations.

Syntax
#pragma fp_contract ( { on | off } )

Remarks
When you use the default compiler options, fp_contract is off , which tells the compiler
to preserve individual floating-point instructions. Set fp_contract to on to use floating-
point contraction instructions where possible. This behavior is new in Visual Studio 2022
version 17.0. In previous compiler versions, fp_contract defaulted to on .

For more information on floating-point behavior, see /fp (Specify floating-point

behavior).

Other floating-point pragma directives include:

fenv_access

float_control

Example
The /fp:fast compiler option enables contractions by default, but the #pragma
fp_contract (off) directive in this example turns them off. The code generated from

this sample won't use a fused-multiply-add instruction even when it's available on the
target processor. If you comment out #pragma fp_contract (off) , the generated code
may use a fused-multiply-add instruction if it's available.

C++
 // pragma_directive_fp_contract.cpp

// On x86 and x64 compile with: /O2 /fp:fast /arch:AVX2

#include <stdio.h>

// remove the following line to enable FP contractions

#pragma fp_contract (off)

int main() {

double z, b, t;

for (int i = 0; i < 10; i++) {

b = i * 5.5;

t = i * 56.025;

z = t * i + b;

printf("out = %.15e\n", z);

Output

out = 0.000000000000000e+00

out = 6.152500000000000e+01

out = 2.351000000000000e+02

out = 5.207249999999999e+02

out = 9.184000000000000e+02

out = 1.428125000000000e+03

out = 2.049900000000000e+03

out = 2.783725000000000e+03

out = 3.629600000000000e+03

out = 4.587525000000000e+03

See also
Pragma directives and the __pragma and _Pragma keywords
function pragma
Article • 08/03/2021 • 2 minutes to read

Tells the compiler to generate calls to functions specified in the pragma's argument list,
instead of inlining them.

Syntax
#pragma function( function1 [ , function2 ... ] )

Remarks
Intrinsic functions are normally generated as inline code, not as function calls. If you use
the intrinsic pragma or the /Oi compiler option to tell the compiler to generate intrinsic
functions, you can use the function pragma to explicitly force a function call. Once a
function pragma is seen, it takes effect at the first function definition that contains a
specified intrinsic function. The effect continues to the end of the source file, or to the
appearance of an intrinsic pragma specifying the same intrinsic function. You can only
use the function pragma outside of a function, at the global level.

For lists of the functions that have intrinsic forms, see intrinsic pragma.

Example
C++

// pragma_directive_function.cpp

#include <ctype.h>

#include <stdio.h>

#include <stdlib.h>

#include <string.h>

// use intrinsic forms of memset and strlen

#pragma intrinsic(memset, strlen)

// Find first word break in string, and set remaining

// chars in string to specified char value.

char *set_str_after_word(char *string, char ch) {

int i;

int len = strlen(string); /* NOTE: uses intrinsic for strlen */

for(i = 0; i < len; i++) {

if (isspace(*(string + i)))

 break;

for(; i < len; i++)

*(string + i) = ch;

return string;

// do not use strlen intrinsic

#pragma function(strlen)

// Set all chars in string to specified char value.

char *set_str(char *string, char ch) {

// Uses intrinsic for memset, but calls strlen library function

return (char *) memset(string, ch, strlen(string));

int main() {

char *str = (char *) malloc(20 * sizeof(char));

strcpy_s(str, sizeof("Now is the time"), "Now is the time");

printf("str is '%s'\n", set_str_after_word(str, '*'));

printf("str is '%s'\n", set_str(str, '!'));

}

Output

str is 'Now************'

str is '!!!!!!!!!!!!!!!'

See also
Pragma directives and the __pragma and _Pragma keywords
hdrstop pragma
Article • 08/03/2021 • 2 minutes to read

Gives you more control over precompilation file names, and over the location at which
the compilation state is saved.

Syntax
#pragma hdrstop [ ( "filename" ) ]

Remarks
The filename is the name of the precompiled header file to use or create (depending on
whether /Yu or /Yc is specified). If filename doesn't contain a path specification, the
precompiled header file is assumed to be in the same directory as the source file.

If a C or C++ file contains a hdrstop pragma when compiled with /Yc , the compiler
saves the state of the compilation up to the location of the pragma. The compiled state
of any code that follows the pragma isn't saved.

Use filename to name the precompiled header file in which the compiled state is saved.
A space between hdrstop and filename is optional. The file name specified in the
hdrstop pragma is a string, and is subject to the constraints of any C or C++ string. In
particular, you must enclose it in quotation marks and use the escape character
(backslash, \ ) to specify directory names. For example:

#pragma hdrstop( "c:\\projects\\include\\myinc.pch" )

The name of the precompiled header file is determined according to the following rules,
in order of precedence:

1. The argument to the /Fp compiler option

2. The filename argument to #pragma hdrstop

3. The base name of the source file with a PCH extension

If none of the /Yc and /Yu options or the hdrstop pragma specifies a file name, the
base name of the source file is used as the base name of the precompiled header file.
You can also use preprocessing commands to perform macro replacement as follows:

#define INCLUDE_PATH "c:\\progra~1\\devstsu~1\\vc\\include\\"

#define PCH_FNAME "PROG.PCH"

#pragma hdrstop( INCLUDE_PATH PCH_FNAME )

The following rules govern where the hdrstop pragma can be placed:

It must appear outside any data or function declaration or definition.

It must be specified in the source file, not within a header file.

Example
C

#include <windows.h> // Include several files

#include "myhdr.h"

__inline Disp( char *szToDisplay ) // Define an inline function

// ... // Some code to display string

#pragma hdrstop

In this example, the hdrstop pragma appears after two files have been included and an
inline function has been defined. This location might, at first, seem to be an odd
placement for the pragma. Consider, however, that using the manual precompilation
options, /Yc and /Yu , with the hdrstop pragma makes it possible for you to precompile
entire source files, or even inline code. The Microsoft compiler doesn't limit you to
precompiling only data declarations.

See also
Pragma directives and the __pragma and _Pragma keywords
include_alias pragma
Article • 08/03/2021 • 2 minutes to read

Specifies that when alias_filename is found in a #include directive, the compiler

substitutes actual_filename in its place.

Syntax
#pragma include_alias( "alias_filename" , "actual_filename" )

#pragma include_alias( <alias_filename> , <actual_filename> )

Remarks
The include_alias pragma directive allows you to substitute files that have different
names or paths for the file names included by source files. For example, some file
systems allow longer header filenames than the 8.3 FAT file system limit. The compiler
can't just truncate the longer names to 8.3, because the first eight characters of the
longer header filenames may not be unique. Whenever the compiler sees the
alias_filename string in a #include directive, it substitutes the name actual_filename
instead. Then it loads the actual_filename header file. This pragma must appear before
the corresponding #include directives. For example:

C++

// First eight characters of these two files not unique.

#pragma include_alias( "AppleSystemHeaderQuickdraw.h", "quickdra.h" )

#pragma include_alias( "AppleSystemHeaderFruit.h", "fruit.h" )

#pragma include_alias( "GraphicsMenu.h", "gramenu.h" )

#include "AppleSystemHeaderQuickdraw.h"

#include "AppleSystemHeaderFruit.h"

#include "GraphicsMenu.h"

The alias to search for must match the specification exactly. The case, spelling, and the
use of double quotation marks or angle brackets must all match. The include_alias
pragma does simple string matching on the filenames. No other filename validation is
performed. For example, given the following directives,

C++
 #pragma include_alias("mymath.h", "math.h")

#include "./mymath.h"

#include "sys/mymath.h"

no alias substitution is done, since the header file strings don't match exactly. Also,
header filenames used as arguments to the /Yu and /Yc compiler options, or the
hdrstop pragma, aren't substituted. For example, if your source file contains the

following directive,

C++

#include <AppleSystemHeaderStop.h>

the corresponding compiler option should be

/YcAppleSystemHeaderStop.h

You can use the include_alias pragma to map any header filename to another. For
example:

C++

#pragma include_alias( "api.h", "c:\version1.0\api.h" )

#pragma include_alias( <stdio.h>, <newstdio.h> )

#include "api.h"

#include <stdio.h>

Don't mix filenames enclosed in double quotation marks with filenames enclosed in
angle brackets. For example, given the above two #pragma include_alias directives, the
compiler does no substitution on the following #include directives:

C++

#include <api.h>

#include "stdio.h"

Furthermore, the following directive generates an error:

C++

#pragma include_alias(<header.h>, "header.h") // Error

The filename reported in error messages, or as the value of the predefined __FILE__
macro, is the name of the file after the substitution is done. For example, see the output
after the following directives:

C++

#pragma include_alias( "VERYLONGFILENAME.H", "myfile.h" )

#include "VERYLONGFILENAME.H"

An error in VERYLONGFILENAME.H produces the following error message:

Output

myfile.h(15) : error C2059 : syntax error

Also note that transitivity isn't supported. Given the following directives,

C++

#pragma include_alias( "one.h", "two.h" )

#pragma include_alias( "two.h", "three.h" )

#include "one.h"

the compiler searches for the file two.h rather than three.h .

See also
Pragma directives and the __pragma and _Pragma keywords
init_seg pragma
Article • 08/03/2021 • 3 minutes to read

C++ Specific

Specifies a keyword or code section that affects the order in which startup code is
executed.

Syntax
#pragma init_seg( { compiler | lib | user | "section-name" [ , func-name ] } )

Remarks
The terms segment and section have the same meaning in this article.

Because code is sometimes required to initialize global static objects, you must specify
when to construct the objects. In particular, it's important to use the init_seg pragma
in dynamic-link libraries (DLLs), or in libraries that require initialization.

The options to the init_seg pragma are:

compiler

Reserved for Microsoft C run-time library initialization. Objects in this group are
constructed first.

lib

Available for third-party class-library vendors' initializations. Objects in this group are
constructed after the ones marked as compiler , but before any others.

user

Available to any user. Objects in this group are constructed last.

section-name

Allows explicit specification of the initialization section. Objects in a user-specified

section-name aren't implicitly constructed. However, their addresses are placed in the
section named by section-name.

The section-name you give will contain pointers to helper functions that will construct
the global objects declared after the pragma in that module.
For a list of names you shouldn't use when creating a section, see /SECTION.

func-name

Specifies a function to be called in place of atexit when the program exits. This helper
function also calls atexit with a pointer to the destructor for the global object. If you
specify a function identifier in the pragma of the form,

C++

int __cdecl myexit (void (__cdecl *pf)(void))

then your function will be called instead of the C run-time library's atexit . It allows you
to build a list of the destructors to call when you're ready to destroy the objects.

If you need to defer initialization (for example, in a DLL) you may choose to specify the
section name explicitly. Your code must then call the constructors for each static object.

There are no quotes around the identifier for the atexit replacement.

Your objects will still be placed in the sections defined by the other XXX_seg pragma
directives.

The objects that are declared in the module aren't automatically initialized by the C run-
time. Your code has to do the initialization.

By default, init_seg sections are read only. If the section name is .CRT , the compiler
silently changes the attribute to read only, even if it's marked as read, write.

You can't specify init_seg more than once in a translation unit.

Even if your object doesn't have a user-defined constructor, one explicitly defined in
code, the compiler may generate one for you. For example, it may create one to bind v-
table pointers. When needed, your code calls the compiler-generated constructor.

Example
C++

// pragma_directive_init_seg.cpp

#include <stdio.h>

#pragma warning(disable : 4075)

typedef void (__cdecl *PF)(void);

int cxpf = 0; // number of destructors we need to call

PF pfx[200]; // pointers to destructors.

int myexit (PF pf) {

pfx[cxpf++] = pf;

return 0;

struct A {

A() { puts("A()"); }
~A() { puts("~A()"); }

};

// ctor & dtor called by CRT startup code

// because this is before the pragma init_seg

A aaaa;

// The order here is important.

// Section names must be 8 characters or less.

// The sections with the same name before the $

// are merged into one section. The order that

// they are merged is determined by sorting

// the characters after the $.

// InitSegStart and InitSegEnd are used to set

// boundaries so we can find the real functions

// that we need to call for initialization.

#pragma section(".mine$a", read)

__declspec(allocate(".mine$a")) const PF InitSegStart = (PF)1;

#pragma section(".mine$z",read)

__declspec(allocate(".mine$z")) const PF InitSegEnd = (PF)1;

// The comparison for 0 is important.

// For now, each section is 256 bytes. When they

// are merged, they are padded with zeros. You

// can't depend on the section being 256 bytes, but

// you can depend on it being padded with zeros.

void InitializeObjects () {

const PF *x = &InitSegStart;

for (++x ; x < &InitSegEnd ; ++x)

if (*x) (*x)();

void DestroyObjects () {

while (cxpf>0) {

--cxpf;

(pfx[cxpf])();

// by default, goes into a read only section

#pragma init_seg(".mine$m", myexit)

A bbbb;

A cccc;

 int main () {
InitializeObjects();

DestroyObjects();

Output

A()

A()

A()

~A()

~A()

~A()

See also
Pragma directives and the __pragma and _Pragma keywords
inline_depth pragma
Article • 08/03/2021 • 2 minutes to read

Specifies the inline heuristic search depth. Functions at a depth in the call graph greater
than the specified value aren't inlined.

Syntax
#pragma inline_depth( [ n ] )

Remarks
This pragma controls the inlining of functions marked inline and __inline, or inlined
automatically under the /Ob compiler option. For more information, see /Ob (Inline
function expansion).

n can be a value between 0 and 255, where 255 means unlimited depth in the call graph.
A value of 0 inhibits inline expansion. When n isn't specified, the default value 254 is
used.

The inline_depth pragma controls the number of times a series of function calls can be
expanded. For example, assume the inline depth is 4. If A calls B, and B then calls C, all
three calls are expanded inline. However, if the closest inline depth expansion is 2, only
A and B are expanded, and C remains as a function call.

To use this pragma, you must set the /Ob compiler option to 1 or higher. The depth set
using this pragma takes effect at the first function call after the pragma.

The inline depth can be decreased during expansion, but not increased. If the inline
depth is 6, and during expansion the preprocessor encounters an inline_depth pragma
with a value of 8, the depth remains 6.

The inline_depth pragma has no effect on functions marked with __forceinline .

７ Note

Recursive functions can be substituted inline to a maximum depth of 16 calls.

See also
Pragma directives and the __pragma and _Pragma keywords

inline_recursion
inline_recursion pragma
Article • 08/03/2021 • 2 minutes to read

Controls the inline expansion of direct or mutually recursive function calls.

Syntax
#pragma inline_recursion( [ { on | off } ] )

Remarks
Use this pragma to control functions marked as inline and __inline or functions that the
compiler automatically expands under the /Ob2 option. Use of this pragma requires an
/Ob compiler option setting of either 1 or 2. The default state for inline_recursion is
off. This pragma takes effect at the first function call after the pragma is seen and
doesn't affect the definition of the function.

The inline_recursion pragma controls how recursive functions are expanded. If

inline_recursion is off, and if an inline function calls itself, either directly or indirectly,

the function is expanded only one time. If inline_recursion is on, the function is
expanded multiple times until it reaches the value set with the inline_depth pragma, the
default value for recursive functions that is defined by the inline_depth pragma, or a
capacity limit.

See also
Pragma directives and the __pragma and _Pragma keywords

inline_depth

/Ob (Inline function expansion)

intrinsic pragma
Article • 08/24/2022 • 2 minutes to read

Specifies that calls to functions specified in the pragma's argument list are intrinsic.

Syntax
#pragma intrinsic( function_1 [ , function_2 ... ] )

Remarks
The intrinsic pragma tells the compiler that a function has known behavior. The
compiler may call the function and not replace the function call with inline instructions,
if it will result in better performance.

The library functions with intrinsic forms are listed below. Once an intrinsic pragma is
seen, it takes effect at the first function definition containing a specified intrinsic
function. The effect continues to the end of the source file or to the appearance of a
function pragma specifying the same intrinsic function. The intrinsic pragma can be

used only outside of a function definition, at the global level.

The following functions have intrinsic forms, and the intrinsic forms are used when you
specify /Oi:

abs

_disable

_enable

fabs

_inp

_inpw

labs

_lrotl

_lrotr

memcmp

memcpy

memset

_outp

_outpw

_rotl

_rotr

strcat

strcmp

strcpy

strlen

_strset

Programs that use intrinsic functions are faster because they don't have the overhead of
function calls. However, they may be larger because of the additional code generated.

x86-specific example
The _disable and _enable intrinsics generate kernel-mode instructions to disable or
enable interrupts, and could be useful in kernel-mode drivers.

Compile the following code from the command line with cl -c -FAs sample.c and look
at sample.asm to see that they turn into x86 instructions CLI and STI:

C++

// pragma_directive_intrinsic.cpp
// processor: x86

#include <dos.h> // definitions for _disable, _enable

#pragma intrinsic(_disable)

#pragma intrinsic(_enable)

void f1(void) {

_disable();

// do some work here that should not be interrupted

_enable();

int main() {

Intrinsic floating-point functions

These floating-point functions don't have true intrinsic forms. Instead, they have
versions that pass arguments directly to the floating-point chip, rather than pushing
them on the stack:

acos

asin
cosh

fmod

pow

sinh

tanh

These floating-point functions have true intrinsic forms when you specify /Oi and
/fp:fast (or any option that includes /Oi : /Ox, /O1, and /O2):

atan

atan2

cos

exp

log

log10

sin

sqrt

tan

You can use /fp:strict or /Za to override generation of true intrinsic floating-point
options. In this case, the functions are generated as library routines that pass arguments
directly to the floating-point chip instead of pushing them onto the program stack.

See #pragma function for information and an example on how to enable and disable
intrinsics for a block of source text.

See also
Pragma directives and the __pragma and _Pragma keywords

Compiler intrinsics
loop pragma
Article • 08/03/2021 • 2 minutes to read

Controls how loop code is to be considered by the auto-parallelizer, or excludes a loop

from consideration by the auto-vectorizer.

Syntax
#pragma loop( hint_parallel( n ) )

#pragma loop( no_vector )

#pragma loop( ivdep )

Parameters
hint_parallel( n )

A hint to the compiler that this loop should be parallelized across n threads, where n is a
positive integer literal or zero. If n is zero, the maximum number of threads is used at
run time. It's a hint to the compiler, not a command. There's no guarantee that the loop
will be parallelized. If the loop has data dependencies, or structural issues, then it won't
be parallelized. For example, it isn't parallelized if it stores to a scalar that's used beyond
the loop body.

The compiler ignores this option unless the /Qpar compiler switch is specified.

no_vector

By default, the auto-vectorizer attempts to vectorize all loops that it evaluates may
benefit from it. Specify this pragma to disable the auto-vectorizer for the loop that
follows.

ivdep

A hint to the compiler to ignore vector dependencies for this loop.

Remarks
To use the loop pragma, place it immediately before, not in, a loop definition. The
pragma takes effect for the scope of the loop that follows it. You can apply multiple
pragma directives to a loop, in any order, but you must state each one in a separate
pragma statement.
See also
Auto-parallelization and auto-vectorization

Pragma directives and the __pragma and _Pragma keywords

make_public pragma
Article • 08/03/2021 • 2 minutes to read

Indicates that a native type should have public assembly accessibility.

Syntax
#pragma make_public( type )

Parameters
type

The name of the type you want to have public assembly accessibility.

Remarks
make_public is useful for when the native type you want to reference is from a header

file that you can't change. If you want to use the native type in the signature of a public
function in a type with public assembly visibility, the native type must also have public
assembly accessibility, or the compiler will issue a warning.

make_public must be specified at global scope. It's only in effect from the point at which
it's declared through to the end of the source code file.

The native type may be implicitly or explicitly private. For more information, see Type
visibility.

Examples
The following sample is the contents of a header file that contains the definitions for
two native structs.

C++

// make_public_pragma.h

struct Native_Struct_1 { int i; };

struct Native_Struct_2 { int i; };

The following code sample consumes the header file. It shows that, unless you explicitly
mark the native structs as public by using make_public , the compiler will generate a
warning when you attempt to use the native structs in the signature of public function in
a public managed type.

C++

// make_public_pragma.cpp

// compile with: /c /clr /W1

#pragma warning (default : 4692)

#include "make_public_pragma.h"

#pragma make_public(Native_Struct_1)

public ref struct A {

void Test(Native_Struct_1 u) {u.i = 0;} // OK

void Test(Native_Struct_2 u) {u.i = 0;} // C4692

};

See also
Pragma directives and the __pragma and _Pragma keywords
managed and unmanaged pragma
Article • 08/03/2021 • 2 minutes to read

Enable function-level control to compile functions as managed or unmanaged.

Syntax
#pragma managed

#pragma unmanaged

#pragma managed( [ push, ] { on | off } )

#pragma managed(pop)

Remarks
The /clr compiler option provides module-level control for compiling functions either as
managed or unmanaged.

An unmanaged function is compiled for the native platform. Execution of that portion of
the program will be passed to the native platform by the common language runtime.

Functions are compiled as managed by default when /clr is used.

When applying a managed or unmanaged pragma:

Add the pragma preceding a function, but not within a function body.

Add the pragma after #include statements. Don't use it before any #include
statements.

The compiler ignores the managed and unmanaged pragma if /clr isn't used in the
compilation.

When a template function is instantiated, the pragma state when the template is defined
determines if it's managed or unmanaged.

For more information, see Initialization of mixed assemblies.

Example
C++
 // pragma_directives_managed_unmanaged.cpp

// compile with: /clr

#include <stdio.h>

// func1 is managed

void func1() {

System::Console::WriteLine("In managed function.");

// #pragma unmanaged

// push managed state on to stack and set unmanaged state

#pragma managed(push, off)

// func2 is unmanaged

void func2() {

printf("In unmanaged function.\n");

// #pragma managed

#pragma managed(pop)

// main is managed

int main() {

func1();

func2();

Output

In managed function.

In unmanaged function.

See also
Pragma directives and the __pragma and _Pragma keywords
message pragma
Article • 08/03/2021 • 2 minutes to read

Sends a string literal to the standard output without terminating the compilation.

Syntax
#pragma message( message-string )

Remarks
A typical use of the message pragma is to display informational messages at compile
time.

The message-string parameter can be a macro that expands to a string literal, and you
can concatenate such macros with string literals in any combination.

If you use a predefined macro in the message pragma, the macro should return a string.
Otherwise, you'll have to convert the output of the macro to a string.

The following code fragment uses the message pragma to display messages during
compilation:

C++

// pragma_directives_message1.cpp
// compile with: /LD

#if _M_IX86 >= 500

#pragma message("_M_IX86 >= 500")

#endif

#pragma message("")

#pragma message( "Compiling " __FILE__ )

#pragma message( "Last modified on " __TIMESTAMP__ )

#pragma message("")

// with line number

#define STRING2(x) #x

#define STRING(x) STRING2(x)

#pragma message (__FILE__ "[" STRING(__LINE__) "]: test")

#pragma message("")

See also
Pragma directives and the __pragma and _Pragma keywords
omp pragma
Article • 08/03/2021 • 2 minutes to read

Takes one or more OpenMP directives, along with any optional directive clauses.

Syntax
#pragma omp directive

Remarks
For more information, see OpenMP directives.

See also
Pragma directives and the __pragma and _Pragma keywords
once pragma
Article • 08/03/2021 • 2 minutes to read

Specifies that the compiler includes the header file only once, when compiling a source
code file.

Syntax
#pragma once

Remarks
The use of #pragma once can reduce build times, as the compiler won't open and read
the file again after the first #include of the file in the translation unit. It's called the
multiple-include optimization. It has an effect similar to the include guard idiom, which
uses preprocessor macro definitions to prevent multiple inclusions of the contents of
the file. It also helps to prevent violations of the one definition rule: the requirement that
all templates, types, functions, and objects have no more than one definition in your
code.

For example:

C++

// header.h

#pragma once

// Code placed here is included only once per translation unit

We recommend the #pragma once directive for new code because it doesn't pollute the
global namespace with a preprocessor symbol. It requires less typing, it's less
distracting, and it can't cause symbol collisions. Symbol collisions are errors caused when
different header files use the same preprocessor symbol as the guard value. It isn't part
of the C++ Standard, but it's implemented portably by several common compilers.

There's no advantage to use of both the include guard idiom and #pragma once in the
same file. The compiler recognizes the include guard idiom, and implements the
multiple-include optimization the same way as the #pragma once directive if no non-
comment code or preprocessor directive comes before or after the standard form of the
idiom:
 C++

// header.h

// Demonstration of the #include guard idiom.

// Note that the defined symbol can be arbitrary.

#ifndef HEADER_H_ // equivalently, #if !defined HEADER_H_

#define HEADER_H_

// Code placed here is included only once per translation unit

#endif // HEADER_H_

We recommend the include guard idiom when code must be portable to compilers that
don't implement the #pragma once directive, to maintain consistency with existing code,
or when the multiple-include optimization is impossible. It can occur in complex projects
when file system aliasing or aliased include paths prevent the compiler from identifying
identical include files by canonical path.

Be careful not to use #pragma once or the include guard idiom in header files designed
to be included multiple times, that use preprocessor symbols to control their effects. For
an example of this design, see the <assert.h> header file. Also be careful to manage
your include paths to avoid creating multiple paths to included files, which can defeat
the multiple-include optimization for both include guards and #pragma once .

See also
Pragma directives and the __pragma and _Pragma keywords
optimize pragma
Article • 03/01/2022 • 2 minutes to read

Specifies optimizations on a function-by-function basis.

Syntax
#pragma optimize( " [ optimization-list ] ", { on | off } )

Remarks
The optimize pragma must appear outside a function. It takes effect at the first function
defined after the pragma is seen. The on and off arguments turn options specified in
the optimization-list on or off.

The optimization-list can be zero or more of the parameters shown in the following
table.

Parameters of the optimize Pragma

Parameter(s) Type of optimization

g Enable global optimizations. Deprecated. For more information, see /Og (Global
optimizations).

s or t Specify short or fast sequences of machine code.

y Generate frame pointers on the program stack.

These parameters are the same letters used with the /O compiler options. For example,
the following pragma is equivalent to the /Os compiler option:

C++

#pragma optimize( "s", on )

Using the optimize pragma with the empty string ( "" ) is a special form of the directive:

When you use the off parameter, it turns all the optimizations, g , s , t , and y , off.
When you use the on parameter, it resets the optimizations to the ones that you
specified using the /O compiler option.

C++

#pragma optimize( "", off )

/* unoptimized code section */

#pragma optimize( "", on )

See also
Pragma directives and the __pragma and _Pragma keywords
pack pragma
Article • 07/08/2022 • 3 minutes to read

Specifies the packing alignment for structure, union, and class members.

Syntax
#pragma pack( show )

#pragma pack( push [ , identifier ] [ , n ] )

#pragma pack( pop [ , { identifier | n } ] )

#pragma pack( [ n ] )

Parameters
show

(Optional) Displays the current byte value for packing alignment. The value is displayed
by a warning message.

push

(Optional) Pushes the current packing alignment value on the internal compiler stack,
and sets the current packing alignment value to n. If n isn't specified, the current
packing alignment value is pushed.

pop

(Optional) Removes the record from the top of the internal compiler stack. If n isn't
specified with pop , then the packing value associated with the resulting record on the
top of the stack is the new packing alignment value. If n is specified, for example,
#pragma pack(pop, 16) , n becomes the new packing alignment value. If you pop using

an identifier , for example, #pragma pack(pop, r1) , then all records on the stack are
popped until the record that has identifier is found. That record gets popped, and the
packing value associated with the record found on the top of the stack becomes the
new packing alignment value. If you pop using an identifier that isn't found in any
record on the stack, then the pop is ignored.

The statement #pragma pack (pop, r1, 2) is equivalent to #pragma pack (pop, r1)
followed by #pragma pack(2) .

identifier

(Optional) When used with push , assigns a name to the record on the internal compiler
stack. When used with pop , pops records off the internal stack until identifier is
removed. If identifier isn't found on the internal stack, nothing is popped.

(Optional) Specifies the value, in bytes, to be used for packing. If the compiler option
/Zp isn't set for the module, the default value for n is 8. Valid values are 1, 2, 4, 8, and
16. The alignment of a member is on a boundary that's either a multiple of n , or a
multiple of the size of the member, whichever is smaller.

Remarks
To pack a class is to place its members directly after each other in memory. It can mean
that some or all members can be aligned on a boundary smaller than the default
alignment of the target architecture. pack gives control at the data-declaration level. It
differs from compiler option /Zp, which only provides module-level control. pack takes
effect at the first struct , union , or class declaration after the pragma is seen. pack has
no effect on definitions. Calling pack with no arguments sets n to the value set in the
compiler option /Zp . If the compiler option isn't set, the default value is 8 for x86, ARM,
and ARM64. The default is 16 for x64 native and ARM64EC.

If you change the alignment of a structure, it may not use as much space in memory.
However, you may see a loss of performance or even get a hardware-generated
exception for unaligned access. You can modify this exception behavior by using
SetErrorMode.

For more information about how to modify alignment, see these articles:

alignof

align

__unaligned

x64 structure alignment examples

２ Warning

In Visual Studio 2015 and later you can use the standard alignas and alignof
operators, which unlike __alignof and __declspec( align ) are portable
across compilers. The C++ standard doesn't address packing, so you must still
use pack (or the corresponding extension on other compilers) to specify
alignments smaller than the target architecture's word size.
Examples
The following sample shows how to use the pack pragma to change the alignment of a
structure.

C++

// pragma_directives_pack.cpp

#include <stddef.h>

#include <stdio.h>

struct S {

int i; // size 4

short j; // size 2
double k; // size 8

};

#pragma pack(2)

struct T {

int i;

short j;

double k;

};

int main() {

printf("%zu ", offsetof(S, i));

printf("%zu ", offsetof(S, j));

printf("%zu\n", offsetof(S, k));

printf("%zu ", offsetof(T, i));

printf("%zu ", offsetof(T, j));

printf("%zu\n", offsetof(T, k));

Output

0 4 8

0 4 6

The following sample shows how to use the push, pop, and show syntax.

C++

// pragma_directives_pack_2.cpp

// compile with: /W1 /c

#pragma pack() // n defaults to 8; equivalent to /Zp8

#pragma pack(show) // C4810

#pragma pack(4) // n = 4

 #pragma pack(show) // C4810

#pragma pack(push, r1, 16) // n = 16, pushed to stack

#pragma pack(show) // C4810

// pop to the identifier and then set

// the value of the current packing alignment:

#pragma pack(pop, r1, 2) // n = 2 , stack popped

#pragma pack(show) // C4810

See also
Pragma directives and the __pragma and _Pragma keywords
pointers_to_members pragma
Article • 08/03/2021 • 2 minutes to read

C++ Specific

Specifies whether a pointer to a class member can be declared before its associated
class definition. Used to control the pointer size, and the code required to interpret the
pointer.

Syntax
#pragma pointers_to_members( best_case )

#pragma pointers_to_members( full_generality [ , most-general-representation ]

)

Remarks
You can place a pointers_to_members pragma in your source file as an alternative to
using the /vmb or /vmg and /vmm, /vms, /vmv compiler options or the Microsoft-
specific inheritance keywords.

The pointer-declaration argument specifies whether you've declared a pointer to a

member before or after the associated function definition. The pointer-declaration
argument is one of these two symbols:

full_generality

Generates safe, sometimes nonoptimal code. Use full_generality if any pointer to

a member is declared before the associated class definition. This argument always
uses the pointer representation specified by the most-general-representation
argument. Equivalent to /vmg .

best_case

Generates optimal code using best-case representation for all pointers to

members. Requires you to define the class before you declare a pointer to a
member. The default is best_case .

The most-general-representation argument specifies the smallest pointer

representation that the compiler should use to reference safely any pointer to a member
of a class in a translation unit. The argument can be one of these values:
 single_inheritance

The most general representation is single-inheritance pointer to member function.

Equivalent to /vmg /vms . Causes an error if the inheritance model of a class
definition is either multiple or virtual.

multiple_inheritance

The most general representation is multiple-inheritance pointer to member

function. Equivalent to /vmg /vmm . Causes an error if the inheritance model of a
class definition is virtual.

virtual_inheritance

The most general representation is virtual-inheritance pointer to member function.

Equivalent to /vmg /vmv . Never causes an error. virtual_inheritance is the default
argument when #pragma pointers_to_members(full_generality) is used.

Ｕ Caution

We advise you to put the pointers_to_members pragma only in the source code file
that you want to affect, and only after any #include directives. This practice
reduces the risk that the pragma will affect other files, and that you'll accidentally
specify multiple definitions for the same variable, function, or class name.

Example
C++

// Specify single-inheritance only

#pragma pointers_to_members( full_generality, single_inheritance )

END C++ Specific

See also
Pragma directives and the __pragma and _Pragma keywords
pop_macro pragma
Article • 08/03/2021 • 2 minutes to read

Sets the value of the macro-name macro to the value on the top of the stack for this
macro.

Syntax
#pragma pop_macro( "macro-name" )

Remarks
A push_macro for macro-name must be issued before you can do a pop_macro .

Example
C++

// pragma_directives_pop_macro.cpp

// compile with: /W1

#include <stdio.h>

#define X 1

#define Y 2

int main() {

printf("%d",X);

printf("\n%d",Y);

#define Y 3 // C4005

#pragma push_macro("Y")

#pragma push_macro("X")

printf("\n%d",X);

#define X 2 // C4005

printf("\n%d",X);

#pragma pop_macro("X")

printf("\n%d",X);

#pragma pop_macro("Y")

printf("\n%d",Y);

Output

See also
Pragma directives and the __pragma and _Pragma keywords
push_macro pragma
Article • 08/03/2021 • 2 minutes to read

Saves the value of the macro-name macro on the top of the stack for this macro.

Syntax
#pragma push_macro(" macro-name ")

Remarks
You can retrieve the value for macro-name with pop_macro .

See pop_macro pragma for a sample.

See also
Pragma directives and the __pragma and _Pragma keywords
region and endregion pragma
Article • 08/03/2021 • 2 minutes to read

#pragma region lets you specify a block of code that you can expand or collapse when

using the outlining feature of the Visual Studio editor.

Syntax
#pragma region name

#pragma endregion comment

Parameters
comment

(Optional) A comment to display in the code editor.

name

(Optional) The name of the region. This name displays in the code editor.

Remarks
#pragma endregion marks the end of a #pragma region block.

A #pragma region block must be terminated by a #pragma endregion directive.

Example
C++

// pragma_directives_region.cpp

#pragma region Region_1

void Test() {}

void Test2() {}

void Test3() {}

#pragma endregion Region_1

int main() {}

See also
Pragma directives and the __pragma and _Pragma keywords
runtime_checks pragma
Article • 08/03/2021 • 2 minutes to read

Disables or restores the /RTC compiler option settings.

Syntax
#pragma runtime_checks( " [ runtime-check-options ] ", { restore | off } )

Remarks
You can't enable a run-time check that wasn't enabled by a compiler option. For
example, if you don't specify /RTCs on the command line, specifying #pragma
runtime_checks( "s", restore) won't enable stack frame verification.

The runtime_checks pragma must appear outside a function, and takes effect at the first
function defined after the pragma is seen. The restore and off arguments turn options
specified in the runtime_checks pragma on or off.

The runtime-check-options can be zero or more of the parameters shown in the

following table.

Parameters of the runtime_checks Pragma

Parameter(s) Type of run-time check

s Enables stack (frame) verification.

c Reports when a value is assigned to a smaller data type that results in a data loss.

u Reports when a variable is used before it's defined.

These parameters are the same ones used with the /RTC compiler option. For example:

C++

#pragma runtime_checks( "sc", restore )

Using the runtime_checks pragma with the empty string ( "" ) is a special form of the
directive:
 When you use the off parameter, it turns the run-time error checks listed in the
table above, off.

When you use the restore parameter, it resets the run-time error checks to the
ones that you specified using the /RTC compiler option.

C++

#pragma runtime_checks( "", off )

/* runtime checks are off in this region */

#pragma runtime_checks( "", restore )

See also
Pragma directives and the __pragma and _Pragma keywords
section pragma
Article • 08/03/2021 • 2 minutes to read

Creates a section in an OBJ file.

Syntax
#pragma section( " section-name " [ , attributes ] )

Remarks
The terms segment and section have the same meaning in this article.

Once a section is defined, it remains valid for the rest of the compilation. However, you
must use __declspec(allocate), or nothing is placed in the section.

section-name is a required parameter that becomes the name of the section. The name
must not conflict with any standard section names. See /SECTION for a list of names you
shouldn't use when creating a section.

attributes is an optional parameter consisting of one or more comma-separated

attributes to assign to the section. Possible attributes are:

Attribute Description

read Allows read operations on data.

write Allows write operations on data.

execute Allows code to be executed.

shared Shares the section among all processes that load the image.

nopage Marks the section as not pageable. Useful for Win32 device drivers.

nocache Marks the section as not cacheable. Useful for Win32 device drivers.

discard Marks the section as discardable. Useful for Win32 device drivers.

remove Marks the section as not memory-resident. For virtual device drivers (VxD) only.

If you don't specify any attributes, the section has read and write attributes.
Example
In this example, the first section pragma identifies the section and its attributes. The
integer j isn't put into mysec because it wasn't declared using __declspec(allocate) .
Instead, j goes into the data section. The integer i does go into mysec because of its
__declspec(allocate) storage-class attribute.

C++

// pragma_section.cpp

#pragma section("mysec",read,write)

int j = 0;

__declspec(allocate("mysec"))

int i = 0;

int main(){}

See also
Pragma directives and the __pragma and _Pragma keywords
setlocale pragma
Article • 11/08/2022 • 2 minutes to read

Defines the locale, the country/region and language to use when translating wide-
character constants and string literals.

Syntax
#pragma setlocale( " [ locale-string ] " )

Remarks
Because the algorithm for converting multibyte characters to wide characters may vary
by locale, or the compilation may take place in a different locale from where an
executable file will be run, this pragma provides a way to specify the target locale at
compile time. It guarantees wide-character strings are stored in the correct format.

The default locale-string is the empty string, specified by #pragma setlocale( "" ) .

The "C" locale maps each character in the string to its value as a wchar_t . Other valid
values for setlocale are the entries found in the Language strings list. For example, you
could specify:

C++

#pragma setlocale("dutch")

The ability to specify a language string depends on the code page and language ID
support on your computer.

See also
Pragma directives and the __pragma and _Pragma keywords
strict_gs_check pragma
Article • 08/03/2021 • 2 minutes to read

This pragma provides enhanced security checking.

Syntax
#pragma strict_gs_check( [ push, ] { on | off } )

#pragma strict_gs_check( pop )

Remarks
Instructs the compiler to insert a random cookie in the function stack to help detect
some categories of stack-based buffer overrun. By default, the /GS compiler option
doesn't insert a cookie for all functions. For more information, see /GS (Buffer Security
Check).

Compile by using /GS to enable strict_gs_check .

Use this pragma in code modules that are exposed to potentially harmful data.
strict_gs_check is an aggressive pragma, and is applied to functions that might not
need this defense, but is optimized to minimize its effect on the performance of the
resulting application.

Even if you use this pragma, you should strive to write secure code. That is, make sure
that your code has no buffer overruns. strict_gs_check might protect your application
from buffer overruns that do remain in your code.

Example
In this sample, a buffer overrun occurs when we copy an array to a local array. When you
compile this code with /GS , no cookie is inserted in the stack, because the array data
type is a pointer. Adding the strict_gs_check pragma forces the stack cookie into the
function stack.

C++

// pragma_strict_gs_check.cpp

// compile with: /c

 #pragma strict_gs_check(on)

void ** ReverseArray(void **pData,

size_t cData)

// *** This buffer is subject to being overrun!! ***

void *pReversed[20];

// Reverse the array into a temporary buffer

for (size_t j = 0, i = cData; i ; --i, ++j)

// *** Possible buffer overrun!! ***

pReversed[j] = pData[i];

// Copy temporary buffer back into input/output buffer

for (size_t i = 0; i < cData ; ++i)

pData[i] = pReversed[i];

return pData;

See also
Pragma directives and the __pragma and _Pragma keywords

/GS (Buffer security check)

system_header pragma
Article • 08/03/2021 • 2 minutes to read

Treat the rest of the file as external for diagnostics reports.

Syntax
#pragma system_header

Remarks
The system_header pragma tells the compiler to show diagnostics at the level specified
by the /external:Wn option for the rest of the current source file. For more information
on how to specify external files and the external warning level to the compiler, see
/external.

The system_header pragma doesn't apply past the end of the current source file. In
other words, it doesn't apply to files that include this file. The system_header pragma
applies even if no other files are specified as external to the compiler. However, if no
/external:Wn option level is specified, the compiler may issue a diagnostic and uses the
same warning level it applies to non-external files. Other pragma directives that affect
warning behavior still apply after a system_header pragma. The effect of #pragma
system_header is similar to the warning pragma:

C++

// If n represents the warning level specified by /external:Wn,

// #pragma system_header is roughly equivalent to:

#pragma warning( push, n )

// . . .

// At the end of the file:

#pragma warning( pop )

The system_header pragma is available starting in Visual Studio 2019 version 16.10.

Example
This sample header shows how to mark the contents of a file as external:
 C++

// library.h

// Use /external:Wn to set the compiler diagnostics level for this file's
contents

#pragma once

#ifndef _LIBRARY_H // include guard for 3rd party interop

#define _LIBRARY_H

#pragma system_header

// The compiler applies the /external:Wn diagnostic level from here to the
end of this file.

// . . .

// You can still override the external diagnostic level for warnings
locally:

#pragma warning( push )

#pragma warning( error : 4164 )

// . . .

#pragma warning(pop)

// . . .

#endif

See also
/external

warning pragma

/Wn (Compiler warning level)

Pragma directives and the __pragma and _Pragma keywords
vtordisp pragma
Article • 08/03/2021 • 2 minutes to read

Controls the addition of the hidden vtordisp construction/destruction displacement

member. The vtordisp pragma is C++-specific.

Syntax
#pragma vtordisp( [ push, ] n )

#pragma vtordisp(pop)

#pragma vtordisp()

#pragma vtordisp( [ push, ] { on | off } )

Parameters
push

Pushes the current vtordisp setting on the internal compiler stack, and sets the new
vtordisp setting to n. If n isn't specified, the current vtordisp setting is unchanged.

pop

Removes the top record from the internal compiler stack, and restores the vtordisp
setting to the removed value.

Specifies the new value for the vtordisp setting. Possible values are 0 , 1 , or 2 ,
corresponding to the /vd0 , /vd1 , and /vd2 compiler options. For more information, see
/vd (Disable Construction Displacements).

on

Equivalent to #pragma vtordisp(1) .

off

Equivalent to #pragma vtordisp(0) .

Remarks
The vtordisp pragma is applicable only to code that uses virtual bases. If a derived class
overrides a virtual function that it inherits from a virtual base class, and if a constructor
or destructor for the derived class calls that function using a pointer to the virtual base
class, the compiler might introduce extra hidden vtordisp fields into classes with virtual
bases.

The vtordisp pragma affects the layout of classes that follow it. The /vd0 , /vd1 , and
/vd2 compiler options specify the same behavior for complete modules. Specifying 0 or
off suppresses the hidden vtordisp members. Turn off vtordisp only if there's no

possibility that the class's constructors and destructors call virtual functions on the
object pointed to by the this pointer.

Specifying 1 or on , the default, enables the hidden vtordisp members where they're
necessary.

Specifying 2 enables the hidden vtordisp members for all virtual bases with virtual
functions. #pragma vtordisp(2) might be necessary to ensure correct performance of
dynamic_cast on a partially constructed object. For more information, see Compiler
Warning (level 1) C4436.

#pragma vtordisp() , with no arguments, restores the vtordisp setting to its initial
setting.

C++

#pragma vtordisp(push, 2)

class GetReal : virtual public VBase { ... };

#pragma vtordisp(pop)

See also
Pragma directives and the __pragma and _Pragma keywords
warning pragma
Article • 01/25/2023 • 3 minutes to read

Enables selective modification of the behavior of compiler warning messages.

Syntax
#pragma warning(

  warning-specifier : warning-number-list

 [ ; warning-specifier : warning-number-list ... ] )

#pragma warning( push [ , n ] )

#pragma warning( pop )

Remarks
The following warning-specifier parameters are available.

warning- Meaning
specifier

1, 2, 3, Apply the given level to the specified warnings. Also turns on a specified warning that
4 is off by default.

default Reset warning behavior to its default value. Also turns on a specified warning that is
off by default. The warning will be generated at its default, documented, level.

For more information, see Compiler warnings that are off by default.

disable Don't issue the specified warning messages.

error Report the specified warnings as errors.

once Display the specified message(s) only one time.

suppress Pushes the current state of the pragma on the stack, disables the specified warning
for the next line, and then pops the warning stack so that the pragma state is reset.

The following code statement illustrates that a warning-number-list parameter can

contain multiple warning numbers, and that multiple warning-specifier parameters can
be specified in the same pragma directive.

C++
 #pragma warning( disable : 4507 34; once : 4385; error : 164 )

This directive is functionally equivalent to the following code:

C++

// Disable warning messages 4507 and 4034.

#pragma warning( disable : 4507 34 )

// Issue warning C4385 only once.

#pragma warning( once : 4385 )

// Report warning C4164 as an error.

#pragma warning( error : 164 )

The compiler adds 4000 to any warning number that is between 0 and 999.

Warning numbers in the range 4700-4999 are associated with code generation. For
these warnings, the state of the warning in effect when the compiler reaches the
function definition remains in effect for the rest of the function. Use of the warning
pragma in the function to change the state of a warning number larger than 4699 only
takes effect after the end of the function. The following example shows the correct
placement of a warning pragma to disable a code-generation warning message, and
then to restore it.

C++

// pragma_warning.cpp

// compile with: /W1

#pragma warning(disable:4700)

void Test() {

int x;

int y = x; // no C4700 here

#pragma warning(default:4700) // C4700 enabled after Test ends

int main() {

int x;

int y = x; // C4700

Notice that throughout a function body, the last setting of the warning pragma will be
in effect for the whole function.

Push and pop

The warning pragma also supports the following syntax, where the optional n parameter
represents a warning level (1 through 4).

#pragma warning( push [ , n ] )

#pragma warning( pop )

The pragma warning( push ) stores the current warning state for every warning. The
pragma warning( push, n ) stores the current state for every warning and sets the
global warning level to n.

The pragma warning( pop ) pops the last warning state pushed onto the stack. Any
changes that you made to the warning state between push and pop are undone.
Consider this example:

C++

#pragma warning( push )

#pragma warning( disable : 4705 )

#pragma warning( disable : 4706 )

#pragma warning( disable : 4707 )

// Some code

#pragma warning( pop )

At the end of this code, pop restores the state of every warning (includes 4705, 4706,
and 4707) to what it was at the start of the code.

When you write header files, you can use push and pop to guarantee that warning-state
changes made by a user don't prevent the headers from compiling correctly. Use push
at the start of the header and pop at the end. For example, you may have a header that
doesn't compile cleanly at warning level 4. The following code changes the warning level
to 3, and then restores the original warning level at the end of the header.

C++

#pragma warning( push, 3 )

// Declarations/definitions

#pragma warning( pop )

For more information about compiler options that help you suppress warnings, see /FI
and /w.

See also
Pragma directives and the __pragma and _Pragma keywords