Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Skip to content

Commit 0983315

Browse files
tglsfdctglsfdc
committed
Release notes for 9.3.3, 9.2.7, 9.1.12, 9.0.16, 8.4.20.
1 parent fa1f0d7 commit 0983315

File tree

4 files changed

+1564
-0
lines changed

4 files changed

+1564
-0
lines changed

‎doc/src/sgml/release-8.4.sgml

+320
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,326 @@
11
<!-- doc/src/sgml/release-8.4.sgml -->
22
<!-- See header comment in release.sgml about typical markup -->
33

4+
<sect1 id="release-8-4-20">
5+
<title>Release 8.4.20</title>
6+
7+
<note>
8+
<title>Release Date</title>
9+
<simpara>2014-02-20</simpara>
10+
</note>
11+
12+
<para>
13+
This release contains a variety of fixes from 8.4.19.
14+
For information about new features in the 8.4 major release, see
15+
<xref linkend="release-8-4">.
16+
</para>
17+
18+
<para>
19+
The <productname>PostgreSQL</> community will stop releasing updates
20+
for the 8.4.X release series in July 2014.
21+
Users are encouraged to update to a newer release branch soon.
22+
</para>
23+
24+
<sect2>
25+
<title>Migration to Version 8.4.20</title>
26+
27+
<para>
28+
A dump/restore is not required for those running 8.4.X.
29+
</para>
30+
31+
<para>
32+
However, if you are upgrading from a version earlier than 8.4.19,
33+
see <xref linkend="release-8-4-19">.
34+
</para>
35+
36+
</sect2>
37+
38+
<sect2>
39+
<title>Changes</title>
40+
41+
<itemizedlist>
42+
43+
<listitem>
44+
<para>
45+
Fix possible mis-replay of WAL records when some segments of a
46+
relation aren't full size (Greg Stark, Tom Lane)
47+
</para>
48+
49+
<para>
50+
The WAL update could be applied to the wrong page, potentially many
51+
pages past where it should have been. Aside from corrupting data,
52+
this error has been observed to result in significant <quote>bloat</>
53+
of standby servers compared to their masters, due to updates being
54+
applied far beyond where the end-of-file should have been. This
55+
failure mode does not appear to be a significant risk during crash
56+
recovery, only when initially synchronizing a standby created from a
57+
base backup taken from a quickly-changing master.
58+
</para>
59+
</listitem>
60+
61+
<listitem>
62+
<para>
63+
Ensure that insertions into non-leaf GIN index pages write a full-page
64+
WAL record when appropriate (Heikki Linnakangas)
65+
</para>
66+
67+
<para>
68+
The previous coding risked index corruption in the event of a
69+
partial-page write during a system crash.
70+
</para>
71+
</listitem>
72+
73+
<listitem>
74+
<para>
75+
Fix race conditions during server process exit (Robert Haas)
76+
</para>
77+
78+
<para>
79+
Ensure that signal handlers don't attempt to use the
80+
process's <varname>MyProc</> pointer after it's no longer valid.
81+
</para>
82+
</listitem>
83+
84+
<listitem>
85+
<para>
86+
Fix unsafe references to <varname>errno</> within error reporting
87+
logic (Christian Kruse)
88+
</para>
89+
90+
<para>
91+
This would typically lead to odd behaviors such as missing or
92+
inappropriate <literal>HINT</> fields.
93+
</para>
94+
</listitem>
95+
96+
<listitem>
97+
<para>
98+
Fix possible crashes from using <function>ereport()</> too early
99+
during server startup (Tom Lane)
100+
</para>
101+
102+
<para>
103+
The principal case we've seen in the field is a crash if the server
104+
is started in a directory it doesn't have permission to read.
105+
</para>
106+
</listitem>
107+
108+
<listitem>
109+
<para>
110+
Clear retry flags properly in OpenSSL socket write
111+
function (Alexander Kukushkin)
112+
</para>
113+
114+
<para>
115+
This omission could result in a server lockup after unexpected loss
116+
of an SSL-encrypted connection.
117+
</para>
118+
</listitem>
119+
120+
<listitem>
121+
<para>
122+
Fix length checking for Unicode identifiers (<literal>U&amp;"..."</>
123+
syntax) containing escapes (Tom Lane)
124+
</para>
125+
126+
<para>
127+
A spurious truncation warning would be printed for such identifiers
128+
if the escaped form of the identifier was too long, but the
129+
identifier actually didn't need truncation after de-escaping.
130+
</para>
131+
</listitem>
132+
133+
<listitem>
134+
<para>
135+
Fix possible crash due to invalid plan for nested sub-selects, such
136+
as <literal>WHERE (... x IN (SELECT ...) ...) IN (SELECT ...)</>
137+
(Tom Lane)
138+
</para>
139+
</listitem>
140+
141+
<listitem>
142+
<para>
143+
Ensure that <command>ANALYZE</> creates statistics for a table column
144+
even when all the values in it are <quote>too wide</> (Tom Lane)
145+
</para>
146+
147+
<para>
148+
<command>ANALYZE</> intentionally omits very wide values from its
149+
histogram and most-common-values calculations, but it neglected to do
150+
something sane in the case that all the sampled entries are too wide.
151+
</para>
152+
</listitem>
153+
154+
<listitem>
155+
<para>
156+
In <literal>ALTER TABLE ... SET TABLESPACE</>, allow the database's
157+
default tablespace to be used without a permissions check
158+
(Stephen Frost)
159+
</para>
160+
161+
<para>
162+
<literal>CREATE TABLE</> has always allowed such usage,
163+
but <literal>ALTER TABLE</> didn't get the memo.
164+
</para>
165+
</listitem>
166+
167+
<listitem>
168+
<para>
169+
Fix <quote>cannot accept a set</> error when some arms of
170+
a <literal>CASE</> return a set and others don't (Tom Lane)
171+
</para>
172+
</listitem>
173+
174+
<listitem>
175+
<para>
176+
Fix checks for all-zero client addresses in pgstat functions (Kevin
177+
Grittner)
178+
</para>
179+
</listitem>
180+
181+
<listitem>
182+
<para>
183+
Fix possible misclassification of multibyte characters by the text
184+
search parser (Tom Lane)
185+
</para>
186+
187+
<para>
188+
Non-ASCII characters could be misclassified when using C locale with
189+
a multibyte encoding. On Cygwin, non-C locales could fail as well.
190+
</para>
191+
</listitem>
192+
193+
<listitem>
194+
<para>
195+
Fix possible misbehavior in <function>plainto_tsquery()</>
196+
(Heikki Linnakangas)
197+
</para>
198+
199+
<para>
200+
Use <function>memmove()</> not <function>memcpy()</> for copying
201+
overlapping memory regions. There have been no field reports of
202+
this actually causing trouble, but it's certainly risky.
203+
</para>
204+
</listitem>
205+
206+
<listitem>
207+
<para>
208+
Accept <literal>SHIFT_JIS</> as an encoding name for locale checking
209+
purposes (Tatsuo Ishii)
210+
</para>
211+
</listitem>
212+
213+
<listitem>
214+
<para>
215+
Fix misbehavior of <function>PQhost()</> on Windows (Fujii Masao)
216+
</para>
217+
218+
<para>
219+
It should return <literal>localhost</> if no host has been specified.
220+
</para>
221+
</listitem>
222+
223+
<listitem>
224+
<para>
225+
Improve error handling in <application>libpq</> and <application>psql</>
226+
for failures during <literal>COPY TO STDOUT/FROM STDIN</> (Tom Lane)
227+
</para>
228+
229+
<para>
230+
In particular this fixes an infinite loop that could occur in 9.2 and
231+
up if the server connection was lost during <literal>COPY FROM
232+
STDIN</>. Variants of that scenario might be possible in older
233+
versions, or with other client applications.
234+
</para>
235+
</listitem>
236+
237+
<listitem>
238+
<para>
239+
Fix misaligned descriptors in <application>ecpg</> (MauMau)
240+
</para>
241+
</listitem>
242+
243+
<listitem>
244+
<para>
245+
In <application>ecpg</>, handle lack of a hostname in the connection
246+
parameters properly (Michael Meskes)
247+
</para>
248+
</listitem>
249+
250+
<listitem>
251+
<para>
252+
Fix performance regression in <filename>contrib/dblink</> connection
253+
startup (Joe Conway)
254+
</para>
255+
256+
<para>
257+
Avoid an unnecessary round trip when client and server encodings match.
258+
</para>
259+
</listitem>
260+
261+
<listitem>
262+
<para>
263+
In <filename>contrib/isn</>, fix incorrect calculation of the check
264+
digit for ISMN values (Fabien Coelho)
265+
</para>
266+
</listitem>
267+
268+
<listitem>
269+
<para>
270+
Ensure client-code-only installation procedure works as documented
271+
(Peter Eisentraut)
272+
</para>
273+
</listitem>
274+
275+
<listitem>
276+
<para>
277+
In Mingw and Cygwin builds, install the <application>libpq</> DLL
278+
in the <filename>bin</> directory (Andrew Dunstan)
279+
</para>
280+
281+
<para>
282+
This duplicates what the MSVC build has long done. It should fix
283+
problems with programs like <application>psql</> failing to start
284+
because they can't find the DLL.
285+
</para>
286+
</listitem>
287+
288+
<listitem>
289+
<para>
290+
Don't generate plain-text <filename>HISTORY</>
291+
and <filename>src/test/regress/README</> files anymore (Tom Lane)
292+
</para>
293+
294+
<para>
295+
These text files duplicated the main HTML and PDF documentation
296+
formats. The trouble involved in maintaining them greatly outweighs
297+
the likely audience for plain-text format. Distribution tarballs
298+
will still contain files by these names, but they'll just be stubs
299+
directing the reader to consult the main documentation.
300+
The plain-text <filename>INSTALL</> file will still be maintained, as
301+
there is arguably a use-case for that.
302+
</para>
303+
</listitem>
304+
305+
<listitem>
306+
<para>
307+
Update time zone data files to <application>tzdata</> release 2013i
308+
for DST law changes in Jordan and historical changes in Cuba.
309+
</para>
310+
311+
<para>
312+
In addition, the zones <literal>Asia/Riyadh87</>,
313+
<literal>Asia/Riyadh88</>, and <literal>Asia/Riyadh89</> have been
314+
removed, as they are no longer maintained by IANA, and never
315+
represented actual civil timekeeping practice.
316+
</para>
317+
</listitem>
318+
319+
</itemizedlist>
320+
321+
</sect2>
322+
</sect1>
323+
4324
<sect1 id="release-8-4-19">
5325
<title>Release 8.4.19</title>
6326

0 commit comments

Comments
 (0)