Re: pgsql: Handle lack of DSM slots in parallel btree build.
On Thu, Jan 30, 2020 at 2:34 PM Thomas Munro wrote: > Handle lack of DSM slots in parallel btree build. > > If no DSM slots are available, a ParallelContext can still be > created, but its seg pointer is NULL. Teach parallel btree build > to cope with that by falling back to a regular non-parallel build, > to avoid crashing with a segmentation fault. Uh, this seems to have completely disabled parallel index builds on the master branch. -- Peter Geoghegan
Re: pgsql: Handle lack of DSM slots in parallel btree build.
Hi Peter, On Tue, Feb 4, 2020 at 9:13 AM Peter Geoghegan wrote: > On Thu, Jan 30, 2020 at 2:34 PM Thomas Munro wrote: > > Handle lack of DSM slots in parallel btree build. > > > > If no DSM slots are available, a ParallelContext can still be > > created, but its seg pointer is NULL. Teach parallel btree build > > to cope with that by falling back to a regular non-parallel build, > > to avoid crashing with a segmentation fault. > > Uh, this seems to have completely disabled parallel index builds on > the master branch. Oops. The check needs to move down below InitializeParallelDSM(), and release any extra resources that might have been acquired (snapshot?). I will do some testing in a few hours and post a fix.
pgsql: Add missing break out seqscan loop in logical replication
Add missing break out seqscan loop in logical replication When replica identity is FULL (an admittedly unusual case), the loop that searches for tuples in execReplication.c didn't stop scanning the table when once a matching tuple was found. Add the missing 'break'. Note slight behavior change: we now return the first matching tuple rather than the last one. They are supposed to be indistinguishable anyway, so this shouldn't matter. Author: Konstantin Knizhnik Discussion: https://postgr.es/m/379743f6-ae91-b866-f7a2-5624e6d2b...@postgrespro.ru Branch -- master Details --- https://git.postgresql.org/pg/commitdiff/1c7a0b387d18c517d45e2f6ec7d8b7d1b2d5fe13 Modified Files -- src/backend/executor/execReplication.c | 3 +++ 1 file changed, 3 insertions(+)
pgsql: Add missing break out seqscan loop in logical replication
Add missing break out seqscan loop in logical replication When replica identity is FULL (an admittedly unusual case), the loop that searches for tuples in execReplication.c didn't stop scanning the table when once a matching tuple was found. Add the missing 'break'. Note slight behavior change: we now return the first matching tuple rather than the last one. They are supposed to be indistinguishable anyway, so this shouldn't matter. Author: Konstantin Knizhnik Discussion: https://postgr.es/m/379743f6-ae91-b866-f7a2-5624e6d2b...@postgrespro.ru Branch -- REL_12_STABLE Details --- https://git.postgresql.org/pg/commitdiff/42e3187a8204486823efe70a3ad6db6404047a02 Modified Files -- src/backend/executor/execReplication.c | 3 +++ 1 file changed, 3 insertions(+)
pgsql: Add missing break out seqscan loop in logical replication
Add missing break out seqscan loop in logical replication When replica identity is FULL (an admittedly unusual case), the loop that searches for tuples in execReplication.c didn't stop scanning the table when once a matching tuple was found. Add the missing 'break'. Note slight behavior change: we now return the first matching tuple rather than the last one. They are supposed to be indistinguishable anyway, so this shouldn't matter. Author: Konstantin Knizhnik Discussion: https://postgr.es/m/379743f6-ae91-b866-f7a2-5624e6d2b...@postgrespro.ru Branch -- REL_10_STABLE Details --- https://git.postgresql.org/pg/commitdiff/380bc8829594b27d0924e617ae705422ec53eca1 Modified Files -- src/backend/executor/execReplication.c | 3 +++ 1 file changed, 3 insertions(+)
pgsql: Add missing break out seqscan loop in logical replication
Add missing break out seqscan loop in logical replication When replica identity is FULL (an admittedly unusual case), the loop that searches for tuples in execReplication.c didn't stop scanning the table when once a matching tuple was found. Add the missing 'break'. Note slight behavior change: we now return the first matching tuple rather than the last one. They are supposed to be indistinguishable anyway, so this shouldn't matter. Author: Konstantin Knizhnik Discussion: https://postgr.es/m/379743f6-ae91-b866-f7a2-5624e6d2b...@postgrespro.ru Branch -- REL_11_STABLE Details --- https://git.postgresql.org/pg/commitdiff/beefab6540deca589c2d9a1d0f91fd98f11d303a Modified Files -- src/backend/executor/execReplication.c | 3 +++ 1 file changed, 3 insertions(+)
pgsql: Fix fuzzy error handling in pg_basebackup when opening gzFile
Fix fuzzy error handling in pg_basebackup when opening gzFile First, this code did not bother checking for a failure when calling dup(). Then, per zlib, gzerror() returns NULL for a NULL input, which can happen if passing down to gzdopen() an invalid file descriptor or if there was an allocation failure. No back-patch is done as this would unlikely be a problem in the field. Per Coverity. Reported-by: Tom Lane Branch -- master Details --- https://git.postgresql.org/pg/commitdiff/177be9edf4bb966400db7769d61e479aa0fe0201 Modified Files -- src/bin/pg_basebackup/pg_basebackup.c | 15 ++- 1 file changed, 14 insertions(+), 1 deletion(-)
