postgresql/src/include
Tom Lane ac8f2e1ef3 In walreceiver, don't try to do ereport() in a signal handler.
This is quite unsafe, even for the case of ereport(FATAL) where we won't
return control to the interrupted code, and despite this code's use of
a flag to restrict the areas where we'd try to do it.  It's possible
for example that we interrupt malloc or free while that's holding a lock
that's meant to protect against cross-thread interference.  Then, any
attempt to do malloc or free within ereport() will result in a deadlock,
preventing the walreceiver process from exiting in response to SIGTERM.
We hypothesize that this explains some hard-to-reproduce failures seen
in the buildfarm.

Hence, get rid of the immediate-exit code in WalRcvShutdownHandler,
as well as the logic associated with WalRcvImmediateInterruptOK.
Instead, we need to take care that potentially-blocking operations
in the walreceiver's data transmission logic (libpqwalreceiver.c)
will respond reasonably promptly to the process's latch becoming
set and then call ProcessWalRcvInterrupts.  Much of the needed code
for that was already present in libpqwalreceiver.c.  I refactored
things a bit so that all the uses of PQgetResult use latch-aware
waiting, but didn't need to do much more.

These changes should be enough to ensure that libpqwalreceiver.c
will respond promptly to SIGTERM whenever it's waiting to receive
data.  In principle, it could block for a long time while waiting
to send data too, and this patch does nothing to guard against that.
I think that that hazard is mostly theoretical though: such blocking
should occur only if we fill the kernel's data transmission buffers,
and we don't generally send enough data to make that happen without
waiting for input.  If we find out that the hazard isn't just
theoretical, we could fix it by using PQsetnonblocking, but that
would require more ticklish changes than I care to make now.

Back-patch of commit a1a789eb5.  This problem goes all the way back
to the origins of walreceiver; but given the substantial reworking
the module received during the v10 cycle, it seems unsafe to assume
that our testing on HEAD validates this patch for pre-v10 branches.
And we'd need to back-patch some prerequisite patches (at least
597a87ccc and its followups, maybe other things), increasing the risk
of problems.  Given the dearth of field reports matching this problem,
it's not worth much risk.  Hence back-patch to v10 and v11 only.

Patch by me; thanks to Thomas Munro for review.

Discussion: https://postgr.es/m/20190416070119.GK2673@paquier.xyz
2019-06-12 17:29:48 -04:00
..
access Mark a few parallelism-related variables with PGDLLIMPORT. 2019-06-03 21:25:43 -04:00
bootstrap Phase 2 of pgindent updates. 2017-06-21 15:19:25 -04:00
catalog Fix missing role dependencies for some schema and type ACLs. 2018-11-09 20:42:03 -05:00
commands Mark assorted variables PGDLLIMPORT. 2017-12-05 09:24:05 -05:00
common Fix detection of passwords hashed with MD5 or SCRAM-SHA-256 2019-04-23 15:43:38 +09:00
datatype Phase 2 of pgindent updates. 2017-06-21 15:19:25 -04:00
executor Fix failure with initplans used conditionally during EvalPlanQual rechecks. 2018-09-15 13:42:34 -04:00
fe_utils Empty search_path in Autovacuum and non-psql/pgbench clients. 2018-02-26 07:39:47 -08:00
foreign Separate reinitialization of shared parallel-scan state from ExecReScan. 2017-08-30 13:18:16 -04:00
lib Rename rbtree.c functions to use "rbt" prefix not "rb" prefix. 2018-11-06 13:25:24 -05:00
libpq Fix detection of passwords hashed with MD5 or SCRAM-SHA-256 2019-04-23 15:43:38 +09:00
mb Phase 3 of pgindent updates. 2017-06-21 15:35:54 -04:00
nodes Fix handling of targetlist SRFs when scan/join relation is known empty. 2019-03-07 14:21:52 -05:00
optimizer Avoid sharing PARAM_EXEC slots between different levels of NestLoop. 2019-01-11 15:54:09 -05:00
parser Fix INSERT ON CONFLICT UPDATE through a view that isn't just SELECT *. 2018-08-04 19:38:58 -04:00
port Remove overzeleous assertions in pg_atomic_flag code. 2018-04-07 18:30:15 -07:00
portability Phase 2 of pgindent updates. 2017-06-21 15:19:25 -04:00
postmaster Log when a BRIN autosummarization request fails 2018-03-14 12:00:53 -03:00
regex Phase 2 of pgindent updates. 2017-06-21 15:19:25 -04:00
replication In walreceiver, don't try to do ereport() in a signal handler. 2019-06-12 17:29:48 -04:00
rewrite Fix creation of resjunk tlist entries for inherited mixed UPDATE/DELETE. 2017-11-27 17:54:09 -05:00
snowball Phase 2 of pgindent updates. 2017-06-21 15:19:25 -04:00
statistics Phase 2 of pgindent updates. 2017-06-21 15:19:25 -04:00
storage Consistently test for in-use shared memory. 2019-04-12 22:36:42 -07:00
tcop Fix event triggers for partitioned tables 2018-10-06 19:17:46 -03:00
tsearch Fix serious performance problems in json(b) to_tsvector(). 2017-07-18 12:45:51 -04:00
utils Prevent memory leaks associated with relcache rd_partcheck structures. 2019-04-13 13:22:26 -04:00
.gitignore Autoconfiscate selection of 64-bit int type for 64-bit large object API. 2012-10-07 21:52:43 -04:00
c.h Allow btree comparison functions to return INT_MIN. 2018-10-05 16:01:29 -04:00
fmgr.h Phase 2 of pgindent updates. 2017-06-21 15:19:25 -04:00
funcapi.h Add a utility function to extract variadic function arguments 2017-10-25 07:14:21 -04:00
getaddrinfo.h Phase 2 of pgindent updates. 2017-06-21 15:19:25 -04:00
getopt_long.h Phase 2 of pgindent updates. 2017-06-21 15:19:25 -04:00
Makefile Add statistics subdirectory to Makefile. 2017-06-08 11:29:50 -04:00
miscadmin.h Backport: Mark assorted GUC variables as PGDLLIMPORT. 2018-02-22 12:54:45 -08:00
pg_config.h.in Avoid thread-safety problem in ecpglib. 2019-01-21 23:18:58 -05:00
pg_config.h.win32 Stamp 10.8. 2019-05-06 16:48:45 -04:00
pg_config_ext.h.in Autoconfiscate selection of 64-bit int type for 64-bit large object API. 2012-10-07 21:52:43 -04:00
pg_config_ext.h.win32 Autoconfiscate selection of 64-bit int type for 64-bit large object API. 2012-10-07 21:52:43 -04:00
pg_config_manual.h Fix default minimum value for descending sequences 2017-01-23 14:00:58 -05:00
pg_getopt.h Use our own getopt() on OpenBSD. 2019-01-18 15:06:26 -05:00
pg_trace.h Phase 2 of pgindent updates. 2017-06-21 15:19:25 -04:00
pgstat.h Rearrange pgstat_bestart() to avoid failures within its critical section. 2019-05-11 21:27:13 -04:00
pgtar.h Phase 3 of pgindent updates. 2017-06-21 15:35:54 -04:00
pgtime.h Mark assorted variables PGDLLIMPORT. 2017-12-05 09:24:05 -05:00
port.h Fix misc typos in comments. 2019-01-23 13:50:50 +02:00
postgres.h Phase 3 of pgindent updates. 2017-06-21 15:35:54 -04:00
postgres_ext.h Phase 2 of pgindent updates. 2017-06-21 15:19:25 -04:00
postgres_fe.h Phase 2 of pgindent updates. 2017-06-21 15:19:25 -04:00
rusagestub.h Phase 2 of pgindent updates. 2017-06-21 15:19:25 -04:00
windowapi.h Phase 2 of pgindent updates. 2017-06-21 15:19:25 -04:00