]> git.ipfire.org Git - thirdparty/postgresql.git/commit
Consistently test for in-use shared memory.
authorNoah Misch <noah@leadboat.com>
Sat, 13 Apr 2019 05:36:38 +0000 (22:36 -0700)
committerNoah Misch <noah@leadboat.com>
Sat, 13 Apr 2019 05:39:52 +0000 (22:39 -0700)
commit808e1e75fb376ff7b39d0c3712fa85ff6e0f0720
tree763af2dcfa764e74621354f52b2cab21d9b20b17
parent77300684867f93aebd29de106f8adcf0797bcb21
Consistently test for in-use shared memory.

postmaster startup scrutinizes any shared memory segment recorded in
postmaster.pid, exiting if that segment matches the current data
directory and has an attached process.  When the postmaster.pid file was
missing, a starting postmaster used weaker checks.  Change to use the
same checks in both scenarios.  This increases the chance of a startup
failure, in lieu of data corruption, if the DBA does "kill -9 `head -n1
postmaster.pid` && rm postmaster.pid && pg_ctl -w start".  A postmaster
will no longer stop if shmat() of an old segment fails with EACCES.  A
postmaster will no longer recycle segments pertaining to other data
directories.  That's good for production, but it's bad for integration
tests that crash a postmaster and immediately delete its data directory.
Such a test now leaks a segment indefinitely.  No "make check-world"
test does that.  win32_shmem.c already avoided all these problems.  In
9.6 and later, enhance PostgresNode to facilitate testing.  Back-patch
to 9.4 (all supported versions).

Reviewed (in earlier versions) by Daniel Gustafsson and Kyotaro HORIGUCHI.

Discussion: https://postgr.es/m/20190408064141.GA2016666@rfd.leadboat.com
src/Makefile.global.in
src/backend/port/sysv_shmem.c
src/backend/port/win32_shmem.c
src/backend/postmaster/postmaster.c
src/backend/storage/ipc/ipci.c
src/backend/utils/init/postinit.c
src/include/storage/ipc.h
src/include/storage/pg_shmem.h
src/test/perl/PostgresNode.pm
src/test/recovery/t/017_shm.pl [new file with mode: 0644]
src/tools/msvc/vcregress.pl