<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd"> <HTML ><HEAD ><TITLE >Release 9.6.11</TITLE ><META NAME="GENERATOR" CONTENT="Modular DocBook HTML Stylesheet Version 1.79"><LINK REV="MADE" HREF="mailto:pgsql-docs@postgresql.org"><LINK REL="HOME" TITLE="PostgreSQL 9.6.15 Documentation" HREF="index.html"><LINK REL="UP" TITLE="Release Notes" HREF="release.html"><LINK REL="PREVIOUS" TITLE="Release 9.6.12" HREF="release-9-6-12.html"><LINK REL="NEXT" TITLE="Release 9.6.10" HREF="release-9-6-10.html"><LINK REL="STYLESHEET" TYPE="text/css" HREF="stylesheet.css"><META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=ISO-8859-1"><META NAME="creation" CONTENT="2019-08-11T16:06:51"></HEAD ><BODY CLASS="SECT1" ><DIV CLASS="NAVHEADER" ><TABLE SUMMARY="Header navigation table" WIDTH="100%" BORDER="0" CELLPADDING="0" CELLSPACING="0" ><TR ><TH COLSPAN="4" ALIGN="center" VALIGN="bottom" ><A HREF="index.html" >PostgreSQL 9.6.15 Documentation</A ></TH ></TR ><TR ><TD WIDTH="10%" ALIGN="left" VALIGN="top" ><A TITLE="Release 9.6.12" HREF="release-9-6-12.html" ACCESSKEY="P" >Prev</A ></TD ><TD WIDTH="10%" ALIGN="left" VALIGN="top" ><A HREF="release.html" ACCESSKEY="U" >Up</A ></TD ><TD WIDTH="60%" ALIGN="center" VALIGN="bottom" >Appendix E. Release Notes</TD ><TD WIDTH="20%" ALIGN="right" VALIGN="top" ><A TITLE="Release 9.6.10" HREF="release-9-6-10.html" ACCESSKEY="N" >Next</A ></TD ></TR ></TABLE ><HR ALIGN="LEFT" WIDTH="100%"></DIV ><DIV CLASS="SECT1" ><H1 CLASS="SECT1" ><A NAME="RELEASE-9-6-11" >E.5. Release 9.6.11</A ></H1 ><DIV CLASS="FORMALPARA" ><P ><B >Release date: </B >2018-11-08</P ></DIV ><P > This release contains a variety of fixes from 9.6.10. For information about new features in the 9.6 major release, see <A HREF="release-9-6.html" >Section E.16</A >. </P ><DIV CLASS="SECT2" ><H2 CLASS="SECT2" ><A NAME="AEN131713" >E.5.1. Migration to Version 9.6.11</A ></H2 ><P > A dump/restore is not required for those running 9.6.X. </P ><P > However, if you are upgrading from a version earlier than 9.6.9, see <A HREF="release-9-6-9.html" >Section E.7</A >. </P ></DIV ><DIV CLASS="SECT2" ><H2 CLASS="SECT2" ><A NAME="AEN131718" >E.5.2. Changes</A ></H2 ><P ></P ><UL ><LI ><P > Fix corner-case failures in <CODE CLASS="FUNCTION" >has_<TT CLASS="REPLACEABLE" ><I >foo</I ></TT >_privilege()</CODE > family of functions (Tom Lane) </P ><P > Return NULL rather than throwing an error when an invalid object OID is provided. Some of these functions got that right already, but not all. <CODE CLASS="FUNCTION" >has_column_privilege()</CODE > was additionally capable of crashing on some platforms. </P ></LI ><LI ><P > Avoid O(N^2) slowdown in regular expression match/split functions on long strings (Andrew Gierth) </P ></LI ><LI ><P > Fix parsing of standard multi-character operators that are immediately followed by a comment or <TT CLASS="LITERAL" >+</TT > or <TT CLASS="LITERAL" >-</TT > (Andrew Gierth) </P ><P > This oversight could lead to parse errors, or to incorrect assignment of precedence. </P ></LI ><LI ><P > Avoid O(N^3) slowdown in lexer for long strings of <TT CLASS="LITERAL" >+</TT > or <TT CLASS="LITERAL" >-</TT > characters (Andrew Gierth) </P ></LI ><LI ><P > Fix mis-execution of SubPlans when the outer query is being scanned backwards (Andrew Gierth) </P ></LI ><LI ><P > Fix failure of <TT CLASS="COMMAND" >UPDATE/DELETE ... WHERE CURRENT OF ...</TT > after rewinding the referenced cursor (Tom Lane) </P ><P > A cursor that scans multiple relations (particularly an inheritance tree) could produce wrong behavior if rewound to an earlier relation. </P ></LI ><LI ><P > Fix <CODE CLASS="FUNCTION" >EvalPlanQual</CODE > to handle conditionally-executed InitPlans properly (Andrew Gierth, Tom Lane) </P ><P > This resulted in hard-to-reproduce crashes or wrong answers in concurrent updates, if they contained code such as an uncorrelated sub-<TT CLASS="LITERAL" >SELECT</TT > inside a <TT CLASS="LITERAL" >CASE</TT > construct. </P ></LI ><LI ><P > Fix character-class checks to not fail on Windows for Unicode characters above U+FFFF (Tom Lane, Kenji Uno) </P ><P > This bug affected full-text-search operations, as well as <TT CLASS="FILENAME" >contrib/ltree</TT > and <TT CLASS="FILENAME" >contrib/pg_trgm</TT >. </P ></LI ><LI ><P > Disallow pushing sub-<TT CLASS="LITERAL" >SELECT</TT >s containing window functions, <TT CLASS="LITERAL" >LIMIT</TT >, or <TT CLASS="LITERAL" >OFFSET</TT > to parallel workers (Amit Kapila) </P ><P > Such cases could result in inconsistent behavior due to different workers getting different answers, as a result of indeterminacy due to row-ordering variations. </P ></LI ><LI ><P > Ensure that sequences owned by a foreign table are processed by <TT CLASS="LITERAL" >ALTER OWNER</TT > on the table (Peter Eisentraut) </P ><P > The ownership change should propagate to such sequences as well, but this was missed for foreign tables. </P ></LI ><LI ><P > Ensure that the server will process already-received <TT CLASS="LITERAL" >NOTIFY</TT > and <TT CLASS="LITERAL" >SIGTERM</TT > interrupts before waiting for client input (Jeff Janes, Tom Lane) </P ></LI ><LI ><P > Fix over-allocation of space for <CODE CLASS="FUNCTION" >array_out()</CODE >'s result string (Keiichi Hirobe) </P ></LI ><LI ><P > Fix memory leak in repeated SP-GiST index scans (Tom Lane) </P ><P > This is only known to amount to anything significant in cases where an exclusion constraint using SP-GiST receives many new index entries in a single command. </P ></LI ><LI ><P > Ensure that <CODE CLASS="FUNCTION" >ApplyLogicalMappingFile()</CODE > closes the mapping file when done with it (Tomas Vondra) </P ><P > Previously, the file descriptor was leaked, eventually resulting in failures during logical decoding. </P ></LI ><LI ><P > Fix logical decoding to handle cases where a mapped catalog table is repeatedly rewritten, e.g. by <TT CLASS="LITERAL" >VACUUM FULL</TT > (Andres Freund) </P ></LI ><LI ><P > Prevent starting the server with <TT CLASS="VARNAME" >wal_level</TT > set to too low a value to support an existing replication slot (Andres Freund) </P ></LI ><LI ><P > Avoid crash if a utility command causes infinite recursion (Tom Lane) </P ></LI ><LI ><P > When initializing a hot standby, cope with duplicate XIDs caused by two-phase transactions on the master (Michael Paquier, Konstantin Knizhnik) </P ></LI ><LI ><P > Fix event triggers to handle nested <TT CLASS="COMMAND" >ALTER TABLE</TT > commands (Michael Paquier, Álvaro Herrera) </P ></LI ><LI ><P > Propagate parent process's transaction and statement start timestamps to parallel workers (Konstantin Knizhnik) </P ><P > This prevents misbehavior of functions such as <CODE CLASS="FUNCTION" >transaction_timestamp()</CODE > when executed in a worker. </P ></LI ><LI ><P > Fix transfer of expanded datums to parallel workers so that alignment is preserved, preventing crashes on alignment-picky platforms (Tom Lane, Amit Kapila) </P ></LI ><LI ><P > Fix WAL file recycling logic to work correctly on standby servers (Michael Paquier) </P ><P > Depending on the setting of <TT CLASS="VARNAME" >archive_mode</TT >, a standby might fail to remove some WAL files that could be removed. </P ></LI ><LI ><P > Fix handling of commit-timestamp tracking during recovery (Masahiko Sawada, Michael Paquier) </P ><P > If commit timestamp tracking has been turned on or off, recovery might fail due to trying to fetch the commit timestamp for a transaction that did not record it. </P ></LI ><LI ><P > Randomize the <CODE CLASS="FUNCTION" >random()</CODE > seed in bootstrap and standalone backends, and in <SPAN CLASS="APPLICATION" >initdb</SPAN > (Noah Misch) </P ><P > The main practical effect of this change is that it avoids a scenario where <SPAN CLASS="APPLICATION" >initdb</SPAN > might mistakenly conclude that POSIX shared memory is not available, due to name collisions caused by always using the same random seed. </P ></LI ><LI ><P > Allow DSM allocation to be interrupted (Chris Travers) </P ></LI ><LI ><P > Avoid failure in a parallel worker when loading an extension that tries to access system caches within its init function (Thomas Munro) </P ><P > We don't consider that to be good extension coding practice, but it mostly worked before parallel query, so continue to support it for now. </P ></LI ><LI ><P > Properly handle turning <TT CLASS="VARNAME" >full_page_writes</TT > on dynamically (Kyotaro Horiguchi) </P ></LI ><LI ><P > Fix possible crash due to double <CODE CLASS="FUNCTION" >free()</CODE > during SP-GiST rescan (Andrew Gierth) </P ></LI ><LI ><P > Avoid possible buffer overrun when replaying GIN page recompression from WAL (Alexander Korotkov, Sivasubramanian Ramasubramanian) </P ></LI ><LI ><P > Fix missed fsync of a replication slot's directory (Konstantin Knizhnik, Michael Paquier) </P ></LI ><LI ><P > Fix unexpected timeouts when using <TT CLASS="VARNAME" >wal_sender_timeout</TT > on a slow server (Noah Misch) </P ></LI ><LI ><P > Ensure that hot standby processes use the correct WAL consistency point (Alexander Kukushkin, Michael Paquier) </P ><P > This prevents possible misbehavior just after a standby server has reached a consistent database state during WAL replay. </P ></LI ><LI ><P > Ensure background workers are stopped properly when the postmaster receives a fast-shutdown request before completing database startup (Alexander Kukushkin) </P ></LI ><LI ><P > Update the free space map during WAL replay of page all-visible/frozen flag changes (Álvaro Herrera) </P ><P > Previously we were not careful about this, reasoning that the FSM is not critical data anyway. However, if it's sufficiently out of date, that can result in significant performance degradation after a standby has been promoted to primary. The FSM will eventually be healed by updates, but we'd like it to be good sooner, so work harder at maintaining it during WAL replay. </P ></LI ><LI ><P > Avoid premature release of parallel-query resources when query end or tuple count limit is reached (Amit Kapila) </P ><P > It's only okay to shut down the executor at this point if the caller cannot demand backwards scan afterwards. </P ></LI ><LI ><P > Don't run atexit callbacks when servicing <TT CLASS="LITERAL" >SIGQUIT</TT > (Heikki Linnakangas) </P ></LI ><LI ><P > Don't record foreign-server user mappings as members of extensions (Tom Lane) </P ><P > If <TT CLASS="COMMAND" >CREATE USER MAPPING</TT > is executed in an extension script, an extension dependency was created for the user mapping, which is unexpected. Roles can't be extension members, so user mappings shouldn't be either. </P ></LI ><LI ><P > Make syslogger more robust against failures in opening CSV log files (Tom Lane) </P ></LI ><LI ><P > Fix <SPAN CLASS="APPLICATION" >psql</SPAN >, as well as documentation examples, to call <CODE CLASS="FUNCTION" >PQconsumeInput()</CODE > before each <CODE CLASS="FUNCTION" >PQnotifies()</CODE > call (Tom Lane) </P ><P > This fixes cases in which <SPAN CLASS="APPLICATION" >psql</SPAN > would not report receipt of a <TT CLASS="LITERAL" >NOTIFY</TT > message until after the next command. </P ></LI ><LI ><P > Fix possible inconsistency in <SPAN CLASS="APPLICATION" >pg_dump</SPAN >'s sorting of dissimilar object names (Jacob Champion) </P ></LI ><LI ><P > Ensure that <SPAN CLASS="APPLICATION" >pg_restore</SPAN > will schema-qualify the table name when emitting <TT CLASS="LITERAL" >DISABLE</TT >/<TT CLASS="LITERAL" >ENABLE TRIGGER</TT > commands (Tom Lane) </P ><P > This avoids failures due to the new policy of running restores with restrictive search path. </P ></LI ><LI ><P > Fix <SPAN CLASS="APPLICATION" >pg_upgrade</SPAN > to handle event triggers in extensions correctly (Haribabu Kommi) </P ><P > <SPAN CLASS="APPLICATION" >pg_upgrade</SPAN > failed to preserve an event trigger's extension-membership status. </P ></LI ><LI ><P > Fix <SPAN CLASS="APPLICATION" >pg_upgrade</SPAN >'s cluster state check to work correctly on a standby server (Bruce Momjian) </P ></LI ><LI ><P > Enforce type <TT CLASS="TYPE" >cube</TT >'s dimension limit in all <TT CLASS="FILENAME" >contrib/cube</TT > functions (Andrey Borodin) </P ><P > Previously, some cube-related functions could construct values that would be rejected by <CODE CLASS="FUNCTION" >cube_in()</CODE >, leading to dump/reload failures. </P ></LI ><LI ><P > In <TT CLASS="FILENAME" >contrib/postgres_fdw</TT >, don't try to ship a variable-free <TT CLASS="LITERAL" >ORDER BY</TT > clause to the remote server (Andrew Gierth) </P ></LI ><LI ><P > Fix <TT CLASS="FILENAME" >contrib/unaccent</TT >'s <CODE CLASS="FUNCTION" >unaccent()</CODE > function to use the <TT CLASS="LITERAL" >unaccent</TT > text search dictionary that is in the same schema as the function (Tom Lane) </P ><P > Previously it tried to look up the dictionary using the search path, which could fail if the search path has a restrictive value. </P ></LI ><LI ><P > Fix build problems on macOS 10.14 (Mojave) (Tom Lane) </P ><P > Adjust <SPAN CLASS="APPLICATION" >configure</SPAN > to add an <TT CLASS="OPTION" >-isysroot</TT > switch to <TT CLASS="VARNAME" >CPPFLAGS</TT >; without this, PL/Perl and PL/Tcl fail to configure or build on macOS 10.14. The specific sysroot used can be overridden at configure time or build time by setting the <TT CLASS="VARNAME" >PG_SYSROOT</TT > variable in the arguments of <SPAN CLASS="APPLICATION" >configure</SPAN > or <SPAN CLASS="APPLICATION" >make</SPAN >. </P ><P > It is now recommended that Perl-related extensions write <TT CLASS="LITERAL" >$(perl_includespec)</TT > rather than <TT CLASS="LITERAL" >-I$(perl_archlibexp)/CORE</TT > in their compiler flags. The latter continues to work on most platforms, but not recent macOS. </P ><P > Also, it should no longer be necessary to specify <TT CLASS="OPTION" >--with-tclconfig</TT > manually to get PL/Tcl to build on recent macOS releases. </P ></LI ><LI ><P > Fix MSVC build and regression-test scripts to work on recent Perl versions (Andrew Dunstan) </P ><P > Perl no longer includes the current directory in its search path by default; work around that. </P ></LI ><LI ><P > On Windows, allow the regression tests to be run by an Administrator account (Andrew Dunstan) </P ><P > To do this safely, <SPAN CLASS="APPLICATION" >pg_regress</SPAN > now gives up any such privileges at startup. </P ></LI ><LI ><P > Allow btree comparison functions to return <TT CLASS="LITERAL" >INT_MIN</TT > (Tom Lane) </P ><P > Up to now, we've forbidden datatype-specific comparison functions from returning <TT CLASS="LITERAL" >INT_MIN</TT >, which allows callers to invert the sort order just by negating the comparison result. However, this was never safe for comparison functions that directly return the result of <CODE CLASS="FUNCTION" >memcmp()</CODE >, <CODE CLASS="FUNCTION" >strcmp()</CODE >, etc, as POSIX doesn't place any such restriction on those functions. At least some recent versions of <CODE CLASS="FUNCTION" >memcmp()</CODE > can return <TT CLASS="LITERAL" >INT_MIN</TT >, causing incorrect sort ordering. Hence, we've removed this restriction. Callers must now use the <TT CLASS="LITERAL" >INVERT_COMPARE_RESULT()</TT > macro if they wish to invert the sort order. </P ></LI ><LI ><P > Fix recursion hazard in shared-invalidation message processing (Tom Lane) </P ><P > This error could, for example, result in failure to access a system catalog or index that had just been processed by <TT CLASS="COMMAND" >VACUUM FULL</TT >. </P ><P > This change adds a new result code for <CODE CLASS="FUNCTION" >LockAcquire</CODE >, which might possibly affect external callers of that function, though only very unusual usage patterns would have an issue with it. The API of <CODE CLASS="FUNCTION" >LockAcquireExtended</CODE > is also changed. </P ></LI ><LI ><P > Save and restore SPI's global variables during <CODE CLASS="FUNCTION" >SPI_connect()</CODE > and <CODE CLASS="FUNCTION" >SPI_finish()</CODE > (Chapman Flack, Tom Lane) </P ><P > This prevents possible interference when one SPI-using function calls another. </P ></LI ><LI ><P > Avoid using potentially-under-aligned page buffers (Tom Lane) </P ><P > Invent new union types <TT CLASS="TYPE" >PGAlignedBlock</TT > and <TT CLASS="TYPE" >PGAlignedXLogBlock</TT >, and use these in place of plain char arrays, ensuring that the compiler can't place the buffer at a misaligned start address. This fixes potential core dumps on alignment-picky platforms, and may improve performance even on platforms that allow misalignment. </P ></LI ><LI ><P > Make <TT CLASS="FILENAME" >src/port/snprintf.c</TT > follow the C99 standard's definition of <CODE CLASS="FUNCTION" >snprintf()</CODE >'s result value (Tom Lane) </P ><P > On platforms where this code is used (mostly Windows), its pre-C99 behavior could lead to failure to detect buffer overrun, if the calling code assumed C99 semantics. </P ></LI ><LI ><P > When building on i386 with the <SPAN CLASS="APPLICATION" >clang</SPAN > compiler, require <TT CLASS="OPTION" >-msse2</TT > to be used (Andres Freund) </P ><P > This avoids problems with missed floating point overflow checks. </P ></LI ><LI ><P > Fix <SPAN CLASS="APPLICATION" >configure</SPAN >'s detection of the result type of <CODE CLASS="FUNCTION" >strerror_r()</CODE > (Tom Lane) </P ><P > The previous coding got the wrong answer when building with <SPAN CLASS="APPLICATION" >icc</SPAN > on Linux (and perhaps in other cases), leading to <SPAN CLASS="APPLICATION" >libpq</SPAN > not returning useful error messages for system-reported errors. </P ></LI ><LI ><P > Update time zone data files to <SPAN CLASS="APPLICATION" >tzdata</SPAN > release 2018g for DST law changes in Chile, Fiji, Morocco, and Russia (Volgograd), plus historical corrections for China, Hawaii, Japan, Macau, and North Korea. </P ></LI ></UL ></DIV ></DIV ><DIV CLASS="NAVFOOTER" ><HR ALIGN="LEFT" WIDTH="100%"><TABLE SUMMARY="Footer navigation table" WIDTH="100%" BORDER="0" CELLPADDING="0" CELLSPACING="0" ><TR ><TD WIDTH="33%" ALIGN="left" VALIGN="top" ><A HREF="release-9-6-12.html" ACCESSKEY="P" >Prev</A ></TD ><TD WIDTH="34%" ALIGN="center" VALIGN="top" ><A HREF="index.html" ACCESSKEY="H" >Home</A ></TD ><TD WIDTH="33%" ALIGN="right" VALIGN="top" ><A HREF="release-9-6-10.html" ACCESSKEY="N" >Next</A ></TD ></TR ><TR ><TD WIDTH="33%" ALIGN="left" VALIGN="top" >Release 9.6.12</TD ><TD WIDTH="34%" ALIGN="center" VALIGN="top" ><A HREF="release.html" ACCESSKEY="U" >Up</A ></TD ><TD WIDTH="33%" ALIGN="right" VALIGN="top" >Release 9.6.10</TD ></TR ></TABLE ></DIV ></BODY ></HTML >