rpms/postgresql
7
0
Fork 0
Code Issues Pull Requests Releases Activity

- Fix WAL replay of locking an updated tuple

Browse Source 
kudos to Alvaro Herrera

  See http://www.postgresql.org/message-id/CA+Tgmob8vfzYrLToqYr7uJ2moW3Gnv8rZpPtznxVXRPfTHQpCA@mail.gmail.com
This commit is contained in:
Jozef Mlich 2014-03-13 13:50:06 +01:00
parent a2f501b795
commit a21b7d5314
2 changed files with 53 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

46
postgresql-wal-replay-of-locking-an-updated-tuple.patch Normal file
View File

@ -0,0 +1,46 @@
commit 9a57858f1103b89a5674f0d50c5fe1f756411df6
Author: Alvaro Herrera <alvherre@alvh.no-ip.org>
Date: Thu Feb 27 11:13:39 2014 -0300
Fix WAL replay of locking an updated tuple
We were resetting the tuple's HEAP_HOT_UPDATED flag as well as t_ctid on
WAL replay of a tuple-lock operation, which is incorrect when the tuple
is already updated.
Back-patch to 9.3. The clearing of both header elements was there
previously, but since no update could be present on a tuple that was
being locked, it was harmless.
Bug reported by Peter Geoghegan and Greg Stark in
CAM3SWZTMQiCi5PV5OWHb+bYkUcnCk=O67w0cSswPvV7XfUcU5g@mail.gmail.com and
CAM-w4HPTOeMT4KP0OJK+mGgzgcTOtLRTvFZyvD0O4aH-7dxo3Q@mail.gmail.com
respectively; diagnosis by Andres Freund.
diff --git a/src/backend/access/heap/heapam.c b/src/backend/access/heap/heapam.c
index 33fc5a7..d94980b 100644
--- a/src/backend/access/heap/heapam.c
+++ b/src/backend/access/heap/heapam.c
@@ -7721,11 +7721,19 @@ heap_xlog_lock(XLogRecPtr lsn, XLogRecord *record)
fix_infomask_from_infobits(xlrec->infobits_set, &htup->t_infomask,
&htup->t_infomask2);
- HeapTupleHeaderClearHotUpdated(htup);
+
+ /*
+ * Clear relevant update flags, but only if the modified infomask says
+ * there's no update.
+ */
+ if (HEAP_XMAX_IS_LOCKED_ONLY(htup->t_infomask))
+ {
+ HeapTupleHeaderClearHotUpdated(htup);
+ /* Make sure there is no forward chain link in t_ctid */
+ htup->t_ctid = xlrec->target.tid;
+ }
HeapTupleHeaderSetXmax(htup, xlrec->locking_xid);
HeapTupleHeaderSetCmax(htup, FirstCommandId, false);
- /* Make sure there is no forward chain link in t_ctid */
- htup->t_ctid = xlrec->target.tid;
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
UnlockReleaseBuffer(buffer);

8
postgresql.spec
View File

@ -64,7 +64,7 @@ Summary: PostgreSQL client programs
Name: postgresql
%global majorversion 9.3
Version: 9.3.3
Release: 1%{?dist}
Release: 2%{?dist}
# The PostgreSQL license is very similar to other MIT licenses, but the OSI
# recognizes it as an independent license, so we do as well.
@ -108,6 +108,7 @@ Patch3: postgresql-perl-rpath.patch
Patch4: postgresql-config-comment.patch
Patch5: postgresql-var-run-socket.patch
Patch6: postgresql-man.patch
Patch7: postgresql-wal-replay-of-locking-an-updated-tuple.patch
BuildRequires: perl(ExtUtils::MakeMaker) glibc-devel bison flex gawk help2man
BuildRequires: perl(ExtUtils::Embed), perl-devel
@ -335,6 +336,7 @@ benchmarks.
%patch4 -p1
%patch5 -p1
%patch6 -p1
%patch7 -p1
# We used to run autoconf here, but there's no longer any real need to,
# since Postgres ships with a reasonably modern configure script.
@ -1131,6 +1133,10 @@ fi
%endif
%changelog
* Thu Mar 13 2014 Jozef Mlich <jmlich@redhat.com> - 9.3.3-2
- Fix WAL replay of locking an updated tuple
kudos to Alvaro Herrera
* Thu Feb 20 2014 Jozef Mlich <jmlich@redhat.com> - 9.3.3-1
- update to 9.3.3 minor version per release notes:
http://www.postgresql.org/docs/9.3/static/release-9-3-3.html