Re: Speed up transaction completion faster after many relations are accessed in a transaction

On Thu, 9 Nov 2023 at 21:48, Heikki Linnakangas <hlinnaka(at)iki(dot)fi> wrote:
>
> On 18/09/2023 07:08, David Rowley wrote:
> > On Fri, 15 Sept 2023 at 22:37, Heikki Linnakangas <hlinnaka(at)iki(dot)fi> wrote:
> >>> I've added a call to LockAssertNoneHeld(false) in there.
> >>
> >> I don't see it in the patch?
> >
> > hmm. I must've git format-patch before committing that part.
> >
> > I'll try that again... see attached.
>
> This needed a rebase after my ResourceOwner refactoring. Attached.
>
> A few quick comments:
>
> - It would be nice to add a test for the issue that you fixed in patch
> v7, i.e. if you prepare a transaction while holding session-level locks.
>
> - GrantLockLocal() now calls MemoryContextAlloc(), which can fail if you
> are out of memory. Is that handled gracefully or is the lock leaked?

CFBot shows one of the test has aborted at [1] with:
[20:54:28.535] Core was generated by `postgres: subscriber: logical
replication apply worker for subscription 16397 '.
[20:54:28.535] Program terminated with signal SIGABRT, Aborted.
[20:54:28.535] #0 __GI_raise (sig=sig(at)entry=6) at
../sysdeps/unix/sysv/linux/raise.c:50
[20:54:28.535] Download failed: Invalid argument. Continuing without
source file ./signal/../sysdeps/unix/sysv/linux/raise.c.
[20:54:28.627]
[20:54:28.627] Thread 1 (Thread 0x7f0ea02d1a40 (LWP 50984)):
[20:54:28.627] #0 __GI_raise (sig=sig(at)entry=6) at
../sysdeps/unix/sysv/linux/raise.c:50
...
...
[20:54:28.627] #2 0x00005618e989d62f in ExceptionalCondition
(conditionName=conditionName(at)entry=0x5618e9b40f70
"dlist_is_empty(&(MyProc->myProcLocks[i]))",
fileName=fileName(at)entry=0x5618e9b40ec0
"../src/backend/storage/lmgr/proc.c", lineNumber=lineNumber(at)entry=856)
at ../src/backend/utils/error/assert.c:66
[20:54:28.627] No locals.
[20:54:28.627] #3 0x00005618e95e6847 in ProcKill (code=<optimized
out>, arg=<optimized out>) at ../src/backend/storage/lmgr/proc.c:856
[20:54:28.627] i = <optimized out>
[20:54:28.627] proc = <optimized out>
[20:54:28.627] procgloballist = <optimized out>
[20:54:28.627] __func__ = "ProcKill"
[20:54:28.627] #4 0x00005618e959ebcc in shmem_exit
(code=code(at)entry=1) at ../src/backend/storage/ipc/ipc.c:276
[20:54:28.627] __func__ = "shmem_exit"
[20:54:28.627] #5 0x00005618e959ecd0 in proc_exit_prepare
(code=code(at)entry=1) at ../src/backend/storage/ipc/ipc.c:198
[20:54:28.627] __func__ = "proc_exit_prepare"
[20:54:28.627] #6 0x00005618e959ee8e in proc_exit (code=code(at)entry=1)
at ../src/backend/storage/ipc/ipc.c:111
[20:54:28.627] __func__ = "proc_exit"
[20:54:28.627] #7 0x00005618e94aa54d in BackgroundWorkerMain () at
../src/backend/postmaster/bgworker.c:805
[20:54:28.627] local_sigjmp_buf = {{__jmpbuf =
{94665009627112, -3865857745677845768, 0, 0, 140732736634980, 1,
3865354362587970296, 7379258256398875384}, __mask_was_saved = 1,
__saved_mask = {__val = {18446744066192964099, 94665025527920,
94665025527920, 94665025527920, 0, 94665025528120, 8192, 1,
94664997686410, 94665009627040, 94664997622076, 94665025527920, 1, 0,
0, 140732736634980}}}}
[20:54:28.627] worker = 0x5618eb37c570
[20:54:28.627] entrypt = <optimized out>
[20:54:28.627] __func__ = "BackgroundWorkerMain"
[20:54:28.627] #8 0x00005618e94b495c in do_start_bgworker
(rw=rw(at)entry=0x5618eb3b73c8) at
../src/backend/postmaster/postmaster.c:5697
[20:54:28.627] worker_pid = <optimized out>
[20:54:28.627] __func__ = "do_start_bgworker"
[20:54:28.627] #9 0x00005618e94b4c32 in maybe_start_bgworkers () at
../src/backend/postmaster/postmaster.c:5921
[20:54:28.627] rw = 0x5618eb3b73c8
[20:54:28.627] num_launched = 0
[20:54:28.627] now = 0
[20:54:28.627] iter = {cur = 0x5618eb3b79a8, next =
0x5618eb382a20, prev = 0x5618ea44a980 <BackgroundWorkerList>}
[20:54:28.627] #10 0x00005618e94b574a in process_pm_pmsignal () at
../src/backend/postmaster/postmaster.c:5073
[20:54:28.627] __func__ = "process_pm_pmsignal"
[20:54:28.627] #11 0x00005618e94b5f4a in ServerLoop () at
../src/backend/postmaster/postmaster.c:1760

[1] - https://cirrus-ci.com/task/5118173163290624?logs=cores#L51

Regards,
Vignesh