|
1 | 1 | <!-- doc/src/sgml/release-10.sgml -->
|
2 | 2 | <!-- See header comment in release.sgml about typical markup -->
|
3 | 3 |
|
| 4 | + <sect1 id="release-10-3"> |
| 5 | + <title>Release 10.3</title> |
| 6 | + |
| 7 | + <formalpara> |
| 8 | + <title>Release date:</title> |
| 9 | + <para>2018-03-01</para> |
| 10 | + </formalpara> |
| 11 | + |
| 12 | + <para> |
| 13 | + This release contains a variety of fixes from 10.2. |
| 14 | + For information about new features in major release 10, see |
| 15 | + <xref linkend="release-10"/>. |
| 16 | + </para> |
| 17 | + |
| 18 | + <sect2> |
| 19 | + <title>Migration to Version 10.3</title> |
| 20 | + |
| 21 | + <para> |
| 22 | + A dump/restore is not required for those running 10.X. |
| 23 | + </para> |
| 24 | + |
| 25 | + <para> |
| 26 | + However, if you are upgrading from a version earlier than 10.2, |
| 27 | + see <xref linkend="release-10-2"/>. |
| 28 | + </para> |
| 29 | + </sect2> |
| 30 | + |
| 31 | + <sect2> |
| 32 | + <title>Changes</title> |
| 33 | + |
| 34 | + <itemizedlist> |
| 35 | + |
| 36 | + <listitem> |
| 37 | +<!-- |
| 38 | +Author: Tom Lane < [email protected]> |
| 39 | +Branch: master [159efe4af] 2018-02-19 16:00:31 -0500 |
| 40 | +Branch: REL_10_STABLE [517e0fe86] 2018-02-19 16:00:18 -0500 |
| 41 | +Branch: REL9_6_STABLE [795f2112e] 2018-02-19 16:00:18 -0500 |
| 42 | +Branch: REL9_5_STABLE [340d63bfb] 2018-02-19 16:00:18 -0500 |
| 43 | +Branch: REL9_4_STABLE [e11b6488e] 2018-02-19 16:00:18 -0500 |
| 44 | +Branch: REL9_3_STABLE [ea6d67cf8] 2018-02-19 16:00:18 -0500 |
| 45 | +--> |
| 46 | + <para> |
| 47 | + Fix misbehavior of concurrent-update rechecks with CTE references |
| 48 | + appearing in subplans (Tom Lane) |
| 49 | + </para> |
| 50 | + |
| 51 | + <para> |
| 52 | + If a CTE (<literal>WITH</literal> clause reference) is used in an |
| 53 | + InitPlan or SubPlan, and the query requires a recheck due to trying |
| 54 | + to update or lock a concurrently-updated row, incorrect results could |
| 55 | + be obtained. |
| 56 | + </para> |
| 57 | + </listitem> |
| 58 | + |
| 59 | + <listitem> |
| 60 | +<!-- |
| 61 | +Author: Tom Lane < [email protected]> |
| 62 | +Branch: master [9afd513df] 2018-02-23 13:47:33 -0500 |
| 63 | +Branch: REL_10_STABLE [c458970ad] 2018-02-23 13:47:33 -0500 |
| 64 | +Branch: REL9_6_STABLE [e7c02a54a] 2018-02-23 13:47:33 -0500 |
| 65 | +Branch: REL9_5_STABLE [7dc5f316f] 2018-02-23 13:47:33 -0500 |
| 66 | +Branch: REL9_4_STABLE [f6dd08489] 2018-02-23 13:47:33 -0500 |
| 67 | +Branch: REL9_3_STABLE [71a0d0c5a] 2018-02-23 13:47:33 -0500 |
| 68 | +--> |
| 69 | + <para> |
| 70 | + Fix planner failures with overlapping mergejoin clauses in an outer |
| 71 | + join (Tom Lane) |
| 72 | + </para> |
| 73 | + |
| 74 | + <para> |
| 75 | + These mistakes led to <quote>left and right pathkeys do not match in |
| 76 | + mergejoin</quote> or <quote>outer pathkeys do not match |
| 77 | + mergeclauses</quote> planner errors in corner cases. |
| 78 | + </para> |
| 79 | + </listitem> |
| 80 | + |
| 81 | + <listitem> |
| 82 | +<!-- |
| 83 | +Author: Tom Lane < [email protected]> |
| 84 | +Branch: master [38b41f182] 2018-02-21 18:40:24 -0500 |
| 85 | +Branch: REL_10_STABLE [5ab5d21aa] 2018-02-21 18:40:24 -0500 |
| 86 | +Branch: REL9_6_STABLE [66e203c85] 2018-02-21 18:40:24 -0500 |
| 87 | +Branch: REL9_5_STABLE [9936f1499] 2018-02-21 18:40:24 -0500 |
| 88 | +Branch: REL9_4_STABLE [2d12c5593] 2018-02-21 18:40:24 -0500 |
| 89 | +Branch: REL9_3_STABLE [d3b0a23a2] 2018-02-21 18:40:24 -0500 |
| 90 | +--> |
| 91 | + <para> |
| 92 | + Repair <application>pg_upgrade</application>'s failure to |
| 93 | + preserve <structfield>relfrozenxid</structfield> for materialized |
| 94 | + views (Tom Lane, Andres Freund) |
| 95 | + </para> |
| 96 | + |
| 97 | + <para> |
| 98 | + This oversight could lead to data corruption in materialized views |
| 99 | + after an upgrade, manifesting as <quote>could not access status of |
| 100 | + transaction</quote> or <quote>found xmin from before |
| 101 | + relfrozenxid</quote> errors. The problem would be more likely to |
| 102 | + occur in seldom-refreshed materialized views, or ones that were |
| 103 | + maintained only with <command>REFRESH MATERIALIZED VIEW |
| 104 | + CONCURRENTLY</command>. |
| 105 | + </para> |
| 106 | + |
| 107 | + <para> |
| 108 | + If such corruption is observed, it can be repaired by refreshing the |
| 109 | + materialized view (without <literal>CONCURRENTLY</literal>). |
| 110 | + </para> |
| 111 | + </listitem> |
| 112 | + |
| 113 | + <listitem> |
| 114 | +<!-- |
| 115 | +Author: Tom Lane < [email protected]> |
| 116 | +Branch: master [3486bcf9e] 2018-02-20 11:23:42 -0500 |
| 117 | +Branch: REL_10_STABLE [6753f6c41] 2018-02-20 11:23:34 -0500 |
| 118 | +--> |
| 119 | + <para> |
| 120 | + Fix incorrect <application>pg_dump</application> output for some |
| 121 | + non-default sequence limit values (Alexey Bashtanov) |
| 122 | + </para> |
| 123 | + </listitem> |
| 124 | + |
| 125 | + <listitem> |
| 126 | +<!-- |
| 127 | +Author: Tom Lane < [email protected]> |
| 128 | +Branch: master [5c9f2564f] 2018-02-11 13:24:15 -0500 |
| 129 | +Branch: REL_10_STABLE [1298fccef] 2018-02-11 13:24:15 -0500 |
| 130 | +--> |
| 131 | + <para> |
| 132 | + Fix <application>pg_dump</application>'s mishandling |
| 133 | + of <literal>STATISTICS</literal> objects (Tom Lane) |
| 134 | + </para> |
| 135 | + |
| 136 | + <para> |
| 137 | + An extended statistics object's schema was mislabeled in the dump's |
| 138 | + table of contents, possibly leading to the wrong results in a |
| 139 | + schema-selective restore. Its ownership was not correctly restored, |
| 140 | + either. Also, change the logic so that statistics objects are |
| 141 | + dumped/restored, or not, as independent objects rather than tying |
| 142 | + them to the dump/restore decision for the table they are on. The |
| 143 | + original definition could not scale to the planned future extension to |
| 144 | + cross-table statistics. |
| 145 | + </para> |
| 146 | + </listitem> |
| 147 | + |
| 148 | + <listitem> |
| 149 | +<!-- |
| 150 | +Author: Tom Lane < [email protected]> |
| 151 | +Branch: master [e748e902d] 2018-02-14 14:47:18 -0500 |
| 152 | +Branch: REL_10_STABLE [4081e1136] 2018-02-14 14:47:18 -0500 |
| 153 | +Branch: REL9_6_STABLE [0a2381d13] 2018-02-14 14:47:18 -0500 |
| 154 | +Branch: REL9_5_STABLE [db1c597f3] 2018-02-14 14:47:18 -0500 |
| 155 | +Branch: REL9_4_STABLE [bd8718637] 2018-02-14 14:47:18 -0500 |
| 156 | +Branch: REL9_3_STABLE [457e9e88e] 2018-02-14 14:47:18 -0500 |
| 157 | +--> |
| 158 | + <para> |
| 159 | + Fix incorrect reporting of PL/Python function names in |
| 160 | + error <literal>CONTEXT</literal> stacks (Tom Lane) |
| 161 | + </para> |
| 162 | + |
| 163 | + <para> |
| 164 | + An error occurring within a nested PL/Python function call (that is, |
| 165 | + one reached via a SPI query from another PL/Python function) would |
| 166 | + result in a stack trace showing the inner function's name twice, |
| 167 | + rather than the expected results. Also, an error in a nested |
| 168 | + PL/Python <literal>DO</literal> block could result in a null pointer |
| 169 | + dereference crash on some platforms. |
| 170 | + </para> |
| 171 | + </listitem> |
| 172 | + |
| 173 | + <listitem> |
| 174 | +<!-- |
| 175 | +Author: Tom Lane < [email protected]> |
| 176 | +Branch: master [8af87f411] 2018-02-23 14:39:17 -0500 |
| 177 | +Branch: REL_10_STABLE [0222e897d] 2018-02-23 14:39:18 -0500 |
| 178 | +Branch: REL9_6_STABLE [be87cd2a0] 2018-02-23 14:39:20 -0500 |
| 179 | +Branch: REL9_5_STABLE [753875ed4] 2018-02-23 14:39:21 -0500 |
| 180 | +Branch: REL9_4_STABLE [3ba503bde] 2018-02-23 14:39:22 -0500 |
| 181 | +Branch: REL9_3_STABLE [a7a6051cf] 2018-02-23 14:39:23 -0500 |
| 182 | +--> |
| 183 | + <para> |
| 184 | + Allow <filename>contrib/auto_explain</filename>'s |
| 185 | + <varname>log_min_duration</varname> setting to range up |
| 186 | + to <literal>INT_MAX</literal>, or about 24 days instead of 35 minutes |
| 187 | + (Tom Lane) |
| 188 | + </para> |
| 189 | + </listitem> |
| 190 | + |
| 191 | + <listitem> |
| 192 | +<!-- |
| 193 | +Author: Robert Haas < [email protected]> |
| 194 | +Branch: master [935dee9ad] 2018-02-09 15:54:45 -0500 |
| 195 | +Author: Andres Freund < [email protected]> |
| 196 | +Branch: REL_10_STABLE [2ff2baa22] 2018-02-22 12:54:45 -0800 |
| 197 | +Branch: REL9_6_STABLE [83fce670e] 2018-02-22 12:58:43 -0800 |
| 198 | +--> |
| 199 | + <para> |
| 200 | + Mark assorted GUC variables as <literal>PGDLLIMPORT</literal>, to |
| 201 | + ease porting extension modules to Windows (Metin Doslu) |
| 202 | + </para> |
| 203 | + </listitem> |
| 204 | + |
| 205 | + </itemizedlist> |
| 206 | + |
| 207 | + </sect2> |
| 208 | + </sect1> |
| 209 | + |
4 | 210 | <sect1 id="release-10-2">
|
5 | 211 | <title>Release 10.2</title>
|
6 | 212 |
|
|
0 commit comments