|
1 | 1 | <!-- doc/src/sgml/release-8.4.sgml -->
|
2 | 2 | <!-- See header comment in release.sgml about typical markup -->
|
3 | 3 |
|
| 4 | + <sect1 id="release-8-4-17"> |
| 5 | + <title>Release 8.4.17</title> |
| 6 | + |
| 7 | + <note> |
| 8 | + <title>Release Date</title> |
| 9 | + <simpara>2013-04-04</simpara> |
| 10 | + </note> |
| 11 | + |
| 12 | + <para> |
| 13 | + This release contains a variety of fixes from 8.4.16. |
| 14 | + For information about new features in the 8.4 major release, see |
| 15 | + <xref linkend="release-8-4">. |
| 16 | + </para> |
| 17 | + |
| 18 | + <sect2> |
| 19 | + <title>Migration to Version 8.4.17</title> |
| 20 | + |
| 21 | + <para> |
| 22 | + A dump/restore is not required for those running 8.4.X. |
| 23 | + </para> |
| 24 | + |
| 25 | + <para> |
| 26 | + However, this release corrects several errors in management of GiST |
| 27 | + indexes. After installing this update, it is advisable to |
| 28 | + <command>REINDEX</> any GiST indexes that meet one or more of the |
| 29 | + conditions described below. |
| 30 | + </para> |
| 31 | + |
| 32 | + <para> |
| 33 | + Also, if you are upgrading from a version earlier than 8.4.10, |
| 34 | + see the release notes for 8.4.10. |
| 35 | + </para> |
| 36 | + |
| 37 | + </sect2> |
| 38 | + |
| 39 | + <sect2> |
| 40 | + <title>Changes</title> |
| 41 | + |
| 42 | + <itemizedlist> |
| 43 | + |
| 44 | + <listitem> |
| 45 | + <para> |
| 46 | + Reset OpenSSL randomness state in each postmaster child process |
| 47 | + (Marko Kreen) |
| 48 | + </para> |
| 49 | + |
| 50 | + <para> |
| 51 | + This avoids a scenario wherein random numbers generated by |
| 52 | + <filename>contrib/pgcrypto</> functions might be relatively easy for |
| 53 | + another database user to guess. The risk is only significant when |
| 54 | + the postmaster is configured with <varname>ssl</> = <literal>on</> |
| 55 | + but most connections don't use SSL encryption. (CVE-2013-1900) |
| 56 | + </para> |
| 57 | + </listitem> |
| 58 | + |
| 59 | + <listitem> |
| 60 | + <para> |
| 61 | + Fix GiST indexes to not use <quote>fuzzy</> geometric comparisons when |
| 62 | + it's not appropriate to do so (Alexander Korotkov) |
| 63 | + </para> |
| 64 | + |
| 65 | + <para> |
| 66 | + The core geometric types perform comparisons using <quote>fuzzy</> |
| 67 | + equality, but <function>gist_box_same</> must do exact comparisons, |
| 68 | + else GiST indexes using it might become inconsistent. After installing |
| 69 | + this update, users should <command>REINDEX</> any GiST indexes on |
| 70 | + <type>box</>, <type>polygon</>, <type>circle</>, or <type>point</> |
| 71 | + columns, since all of these use <function>gist_box_same</>. |
| 72 | + </para> |
| 73 | + </listitem> |
| 74 | + |
| 75 | + <listitem> |
| 76 | + <para> |
| 77 | + Fix erroneous range-union and penalty logic in GiST indexes that use |
| 78 | + <filename>contrib/btree_gist</> for variable-width data types, that is |
| 79 | + <type>text</>, <type>bytea</>, <type>bit</>, and <type>numeric</> |
| 80 | + columns (Tom Lane) |
| 81 | + </para> |
| 82 | + |
| 83 | + <para> |
| 84 | + These errors could result in inconsistent indexes in which some keys |
| 85 | + that are present would not be found by searches, and also in useless |
| 86 | + index bloat. Users are advised to <command>REINDEX</> such indexes |
| 87 | + after installing this update. |
| 88 | + </para> |
| 89 | + </listitem> |
| 90 | + |
| 91 | + <listitem> |
| 92 | + <para> |
| 93 | + Fix bugs in GiST page splitting code for multi-column indexes |
| 94 | + (Tom Lane) |
| 95 | + </para> |
| 96 | + |
| 97 | + <para> |
| 98 | + These errors could result in inconsistent indexes in which some keys |
| 99 | + that are present would not be found by searches, and also in indexes |
| 100 | + that are unnecessarily inefficient to search. Users are advised to |
| 101 | + <command>REINDEX</> multi-column GiST indexes after installing this |
| 102 | + update. |
| 103 | + </para> |
| 104 | + </listitem> |
| 105 | + |
| 106 | + <listitem> |
| 107 | + <para> |
| 108 | + Fix infinite-loop risk in regular expression compilation (Tom Lane, |
| 109 | + Don Porter) |
| 110 | + </para> |
| 111 | + </listitem> |
| 112 | + |
| 113 | + <listitem> |
| 114 | + <para> |
| 115 | + Fix potential null-pointer dereference in regular expression compilation |
| 116 | + (Tom Lane) |
| 117 | + </para> |
| 118 | + </listitem> |
| 119 | + |
| 120 | + <listitem> |
| 121 | + <para> |
| 122 | + Fix <function>to_char()</> to use ASCII-only case-folding rules where |
| 123 | + appropriate (Tom Lane) |
| 124 | + </para> |
| 125 | + |
| 126 | + <para> |
| 127 | + This fixes misbehavior of some template patterns that should be |
| 128 | + locale-independent, but mishandled <quote><literal>I</></quote> and |
| 129 | + <quote><literal>i</></quote> in Turkish locales. |
| 130 | + </para> |
| 131 | + </listitem> |
| 132 | + |
| 133 | + <listitem> |
| 134 | + <para> |
| 135 | + Fix unwanted rejection of timestamp <literal>1999-12-31 24:00:00</> |
| 136 | + (Tom Lane) |
| 137 | + </para> |
| 138 | + </listitem> |
| 139 | + |
| 140 | + <listitem> |
| 141 | + <para> |
| 142 | + Remove useless <quote>picksplit doesn't support secondary split</> log |
| 143 | + messages (Josh Hansen, Tom Lane) |
| 144 | + </para> |
| 145 | + |
| 146 | + <para> |
| 147 | + This message seems to have been added in expectation of code that was |
| 148 | + never written, and probably never will be, since GiST's default |
| 149 | + handling of secondary splits is actually pretty good. So stop nagging |
| 150 | + end users about it. |
| 151 | + </para> |
| 152 | + </listitem> |
| 153 | + |
| 154 | + <listitem> |
| 155 | + <para> |
| 156 | + Fix possible failure to send a session's last few transaction |
| 157 | + commit/abort counts to the statistics collector (Tom Lane) |
| 158 | + </para> |
| 159 | + </listitem> |
| 160 | + |
| 161 | + <listitem> |
| 162 | + <para> |
| 163 | + Eliminate memory leaks in PL/Perl's <function>spi_prepare()</> function |
| 164 | + (Alex Hunsaker, Tom Lane) |
| 165 | + </para> |
| 166 | + </listitem> |
| 167 | + |
| 168 | + <listitem> |
| 169 | + <para> |
| 170 | + Fix <application>pg_dumpall</> to handle database names containing |
| 171 | + <quote><literal>=</></quote> correctly (Heikki Linnakangas) |
| 172 | + </para> |
| 173 | + </listitem> |
| 174 | + |
| 175 | + <listitem> |
| 176 | + <para> |
| 177 | + Avoid crash in <application>pg_dump</> when an incorrect connection |
| 178 | + string is given (Heikki Linnakangas) |
| 179 | + </para> |
| 180 | + </listitem> |
| 181 | + |
| 182 | + <listitem> |
| 183 | + <para> |
| 184 | + Ignore invalid indexes in <application>pg_dump</> (Michael Paquier) |
| 185 | + </para> |
| 186 | + |
| 187 | + <para> |
| 188 | + Dumping invalid indexes can cause problems at restore time, for example |
| 189 | + if the reason the index creation failed was because it tried to enforce |
| 190 | + a uniqueness condition not satisfied by the table's data. Also, if the |
| 191 | + index creation is in fact still in progress, it seems reasonable to |
| 192 | + consider it to be an uncommitted DDL change, which |
| 193 | + <application>pg_dump</> wouldn't be expected to dump anyway. |
| 194 | + </para> |
| 195 | + </listitem> |
| 196 | + |
| 197 | + <listitem> |
| 198 | + <para> |
| 199 | + Fix <filename>contrib/pg_trgm</>'s <function>similarity()</> function |
| 200 | + to return zero for trigram-less strings (Tom Lane) |
| 201 | + </para> |
| 202 | + |
| 203 | + <para> |
| 204 | + Previously it returned <literal>NaN</> due to internal division by zero. |
| 205 | + </para> |
| 206 | + </listitem> |
| 207 | + |
| 208 | + <listitem> |
| 209 | + <para> |
| 210 | + Update time zone data files to <application>tzdata</> release 2013b |
| 211 | + for DST law changes in Chile, Haiti, Morocco, Paraguay, and some |
| 212 | + Russian areas. Also, historical zone data corrections for numerous |
| 213 | + places. |
| 214 | + </para> |
| 215 | + |
| 216 | + <para> |
| 217 | + Also, update the time zone abbreviation files for recent changes in |
| 218 | + Russia and elsewhere: <literal>CHOT</>, <literal>GET</>, |
| 219 | + <literal>IRKT</>, <literal>KGT</>, <literal>KRAT</>, <literal>MAGT</>, |
| 220 | + <literal>MAWT</>, <literal>MSK</>, <literal>NOVT</>, <literal>OMST</>, |
| 221 | + <literal>TKT</>, <literal>VLAT</>, <literal>WST</>, <literal>YAKT</>, |
| 222 | + <literal>YEKT</> now follow their current meanings, and |
| 223 | + <literal>VOLT</> (Europe/Volgograd) and <literal>MIST</> |
| 224 | + (Antarctica/Macquarie) are added to the default abbreviations list. |
| 225 | + </para> |
| 226 | + </listitem> |
| 227 | + |
| 228 | + </itemizedlist> |
| 229 | + |
| 230 | + </sect2> |
| 231 | + </sect1> |
| 232 | + |
4 | 233 | <sect1 id="release-8-4-16">
|
5 | 234 | <title>Release 8.4.16</title>
|
6 | 235 |
|
|
0 commit comments