]> granicus.if.org Git - postgresql/commit
Minor cleanup for access/transam/parallel.c.
authorTom Lane <tgl@sss.pgh.pa.us>
Mon, 1 Aug 2016 20:12:01 +0000 (16:12 -0400)
committerTom Lane <tgl@sss.pgh.pa.us>
Mon, 1 Aug 2016 20:12:01 +0000 (16:12 -0400)
commita5fe473ad79d8d2c85a625621c165e8c601274e4
tree7f15bee616a12c34969b90237fb9f06c7e145244
parent887feefe87b9099eeeec2967ec31ce20df4dfa9b
Minor cleanup for access/transam/parallel.c.

ParallelMessagePending *must* be marked volatile, because it's set
by a signal handler.  On the other hand, it's pointless for
HandleParallelMessageInterrupt to save/restore errno; that must be,
and is, done at the outer level of the SIGUSR1 signal handler.

Calling CHECK_FOR_INTERRUPTS() inside HandleParallelMessages, which itself
is called from CHECK_FOR_INTERRUPTS(), seems both useless and hazardous.
The comment claiming that this is needed to handle the error queue going
away is certainly misguided, in any case.

Improve a couple of error message texts, and use
ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE to report loss of parallel worker
connection, since that's what's used in e.g. tqueue.c.  (Maybe it would be
worth inventing a dedicated ERRCODE for this type of failure?  But I do not
think ERRCODE_INTERNAL_ERROR is appropriate.)

Minor stylistic cleanups.
src/backend/access/transam/parallel.c
src/include/access/parallel.h