Movatterモバイル変換


[0]ホーム

URL:


Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

Sign up
Appearance settings

Commitdd29549

Browse files
committed
Draft release notes for 9.1.3, 9.0.7, 8.4.11, 8.3.18.
1 parenta445cb9 commitdd29549

File tree

4 files changed

+1616
-0
lines changed

4 files changed

+1616
-0
lines changed

‎doc/src/sgml/release-8.3.sgml

Lines changed: 258 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,264 @@
11
<!-- doc/src/sgml/release-8.3.sgml -->
22
<!-- See header comment in release.sgml about typical markup -->
33

4+
<sect1 id="release-8-3-18">
5+
<title>Release 8.3.18</title>
6+
7+
<note>
8+
<title>Release Date</title>
9+
<simpara>2012-02-27</simpara>
10+
</note>
11+
12+
<para>
13+
This release contains a variety of fixes from 8.3.17.
14+
For information about new features in the 8.3 major release, see
15+
<xref linkend="release-8-3">.
16+
</para>
17+
18+
<sect2>
19+
<title>Migration to Version 8.3.18</title>
20+
21+
<para>
22+
A dump/restore is not required for those running 8.3.X.
23+
</para>
24+
25+
<para>
26+
However, if you are upgrading from a version earlier than 8.3.17,
27+
see the release notes for 8.3.17.
28+
</para>
29+
30+
</sect2>
31+
32+
<sect2>
33+
<title>Changes</title>
34+
35+
<itemizedlist>
36+
37+
<listitem>
38+
<para>
39+
Fix btree index corruption from insertions concurrent with vacuuming
40+
(Tom Lane)
41+
</para>
42+
43+
<para>
44+
An index page split caused by an insertion could sometimes cause a
45+
concurrently-running <command>VACUUM</> to miss removing index entries
46+
that it should remove. After the corresponding table rows are removed,
47+
the dangling index entries would cause errors (such as <quote>could not
48+
read block N in file ...</>) or worse, silently wrong query results
49+
after unrelated rows are re-inserted at the now-free table locations.
50+
This bug has been present since release 8.2, but occurs so infrequently
51+
that it was not diagnosed until now. If you have reason to suspect
52+
that it has happened in your database, reindexing the affected index
53+
will fix things.
54+
</para>
55+
</listitem>
56+
57+
<listitem>
58+
<para>
59+
Allow non-existent values for some settings in <command>ALTER
60+
USER/DATABASE SET</> (Heikki Linnakangas)
61+
</para>
62+
63+
<para>
64+
Allow <varname>default_text_search_config</>,
65+
<varname>default_tablespace</>, and <varname>temp_tablespaces</> to be
66+
set to names that are not known. This is because they might be known
67+
in another database where the setting is intended to be used, or for the
68+
tablespace cases because the tablespace might not be created yet. The
69+
same issue was previously recognized for <varname>search_path</>, and
70+
these settings now act like that one.
71+
</para>
72+
</listitem>
73+
74+
<listitem>
75+
<para>
76+
Track the OID counter correctly during WAL replay, even when it wraps
77+
around (Tom Lane)
78+
</para>
79+
80+
<para>
81+
Previously the OID counter would remain stuck at a high value until the
82+
system exited replay mode. The practical consequences of that are
83+
usually nil, but there are scenarios wherein a standby server that's
84+
been promoted to master might take a long time to advance the OID
85+
counter to a reasonable value once values are needed.
86+
</para>
87+
</listitem>
88+
89+
<listitem>
90+
<para>
91+
Fix regular expression back-references with <literal>*</> attached
92+
(Tom Lane)
93+
</para>
94+
95+
<para>
96+
Rather than enforcing an exact string match, the code would effectively
97+
accept any string that satisfies the pattern sub-expression referenced
98+
by the back-reference symbol.
99+
</para>
100+
101+
<para>
102+
A similar problem still afflicts back-references that are embedded in a
103+
larger quantified expression, rather than being the immediate subject
104+
of the quantifier. This will be addressed in a future
105+
<productname>PostgreSQL</> release.
106+
</para>
107+
</listitem>
108+
109+
<listitem>
110+
<para>
111+
Fix recently-introduced memory leak in processing of
112+
<type>inet</>/<type>cidr</> values (Heikki Linnakangas)
113+
</para>
114+
115+
<para>
116+
A patch in the December 2011 releases of <productname>PostgreSQL</>
117+
caused memory leakage in these operations, which could be significant
118+
in scenarios such as building a btree index on such a column.
119+
</para>
120+
</listitem>
121+
122+
<listitem>
123+
<para>
124+
Avoid double close of file handle in syslogger on Windows (MauMau)
125+
</para>
126+
127+
<para>
128+
Ordinarily this error was invisible, but it would cause an exception
129+
when running on a debug version of Windows.
130+
</para>
131+
</listitem>
132+
133+
<listitem>
134+
<para>
135+
Fix I/O-conversion-related memory leaks in plpgsql
136+
(Andres Freund, Jan Urbanski, Tom Lane)
137+
</para>
138+
139+
<para>
140+
Certain operations would leak memory until the end of the current
141+
function.
142+
</para>
143+
</listitem>
144+
145+
<listitem>
146+
<para>
147+
Improve <application>pg_dump</>'s handling of inherited table columns
148+
(Tom Lane)
149+
</para>
150+
151+
<para>
152+
<application>pg_dump</> mishandled situations where a child column has
153+
a different default expression than its parent column. If the default
154+
is textually identical to the parent's default, but not actually the
155+
same (for instance, because of schema search path differences) it would
156+
not be recognized as different, so that after dump and restore the
157+
child would be allowed to inherit the parent's default. Child columns
158+
that are <literal>NOT NULL</> where their parent is not could also be
159+
restored subtly incorrectly.
160+
</para>
161+
</listitem>
162+
163+
<listitem>
164+
<para>
165+
Fix <application>pg_restore</>'s direct-to-database mode for
166+
INSERT-style table data (Tom Lane)
167+
</para>
168+
169+
<para>
170+
Direct-to-database restores from archive files made with
171+
<option>--inserts</> or <option>--column-inserts</> options fail when
172+
using <application>pg_restore</> from a release dated September or
173+
December 2011, as a result of an oversight in a fix for another
174+
problem. The archive file itself is not at fault, and text-mode
175+
output is okay.
176+
</para>
177+
</listitem>
178+
179+
<listitem>
180+
<para>
181+
Fix error in <filename>contrib/intarray</>'s <literal>int[] &amp;
182+
int[]</> operator (Guillaume Lelarge)
183+
</para>
184+
185+
<para>
186+
If the smallest integer the two input arrays have in common is 1,
187+
and there are smaller values in either array, then 1 would be
188+
incorrectly omitted from the result.
189+
</para>
190+
</listitem>
191+
192+
<listitem>
193+
<para>
194+
Fix error detection in <filename>contrib/pgcrypto</>'s
195+
<function>encrypt_iv()</> and <function>decrypt_iv()</>
196+
(Marko Kreen)
197+
</para>
198+
199+
<para>
200+
These functions failed to report certain types of invalid-input errors,
201+
and would instead return random garbage values for incorrect input.
202+
</para>
203+
</listitem>
204+
205+
<listitem>
206+
<para>
207+
Fix one-byte buffer overrun in <filename>contrib/test_parser</>
208+
(Paul Guyot)
209+
</para>
210+
211+
<para>
212+
The code would try to read one more byte than it should, which would
213+
crash in corner cases.
214+
Since <filename>contrib/test_parser</> is only example code, this is
215+
not a security issue in itself, but bad example code is still bad.
216+
</para>
217+
</listitem>
218+
219+
<listitem>
220+
<para>
221+
Use <function>__sync_lock_test_and_set()</> for spinlocks on ARM, if
222+
available (Martin Pitt)
223+
</para>
224+
225+
<para>
226+
This function replaces our previous use of the <literal>SWPB</>
227+
instruction, which is deprecated and not available on ARMv6 and later.
228+
Reports suggest that the old code doesn't fail in an obvious way on
229+
recent ARM boards, but simply doesn't interlock concurrent accesses,
230+
leading to bizarre failures in multiprocess operation.
231+
</para>
232+
</listitem>
233+
234+
<listitem>
235+
<para>
236+
Use <option>-fexcess-precision=standard</> option when building with
237+
gcc versions that accept it (Andrew Dunstan)
238+
</para>
239+
240+
<para>
241+
This prevents assorted scenarios wherein recent versions of gcc will
242+
produce creative results.
243+
</para>
244+
</listitem>
245+
246+
<listitem>
247+
<para>
248+
Allow use of threaded Python on FreeBSD (Chris Rees)
249+
</para>
250+
251+
<para>
252+
Our configure script previously believed that this combination wouldn't
253+
work; but FreeBSD fixed the problem, so remove that error check.
254+
</para>
255+
</listitem>
256+
257+
</itemizedlist>
258+
259+
</sect2>
260+
</sect1>
261+
4262
<sect1 id="release-8-3-17">
5263
<title>Release 8.3.17</title>
6264

0 commit comments

Comments
 (0)

[8]ページ先頭

©2009-2025 Movatter.jp