Horton

Re: Refactoring backend fork+exec code

Thomas Munro <thomas.munro@gmail.com>

View thread
From: Thomas Munro <thomas.munro@gmail.com>
To: Andres Freund <andres@anarazel.de>
Cc: Tristan Partin <tristan@neon.tech>, Heikki Linnakangas <hlinnaka@iki.fi>, pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2023-11-30T20:49:05Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Avoid 0-length memcpy to NULL with EXEC_BACKEND

  2. Tighten check for --forkchild argument when spawning child process

  3. Add missing PGDLLIMPORT markings

  4. Remove assertions that some compiler say are tautological

  5. Move code for backend startup to separate file

  6. Refactor postmaster child process launching

  7. Move some functions from postmaster.c to a new source file

  8. Split registration of Win32 deadchild callback to separate function

  9. Improve log messages referring to background worker processes

  10. Move initialization of the Port struct to the child process

  11. Pass CAC as an argument to the backend process

  12. Set socket options in child process after forking

  13. Fix references to renamed function in comments

  14. Use MyBackendType in more places to check what process this is

  15. Remove MyAuxProcType, use MyBackendType instead

  16. Replace BackendIds with 0-based ProcNumbers

  17. Redefine backend ID to be an index into the proc array

  18. Remove superfluous 'pgprocno' field from PGPROC

  19. Fix incorrect comment on how BackendStatusArray is indexed

  20. Refactor how InitProcess is called

  21. Pass BackgroundWorker entry in the parameter file in EXEC_BACKEND mode

  22. Refactor CreateSharedMemoryAndSemaphores

  23. Silence Valgrind complaint with EXEC_BACKEND

  24. Adjust obsolete comment explaining set_stack_base().

  25. Print lwlock stats also for aux processes, when built with LWLOCK_STATS

  26. Move canAcceptConnections check from ProcessStartupPacket to caller.

  27. Fix crash on syslogger startup

  28. Refactor ListenSocket array.

  29. Use FD_CLOEXEC on ListenSockets

  30. Don't leak descriptors into subprograms.

  31. Use WaitEventSet API for postmaster's event loop.

  32. Rearrange postmaster's startup sequence for better syslogger results.

  33. Fix and enhance the assertion of no palloc's in a critical section.

On Fri, Dec 1, 2023 at 9:31 AM Andres Freund <andres@anarazel.de> wrote:
> On 2023-11-30 12:44:33 -0600, Tristan Partin wrote:
> > >  +        /*
> > >  +         * Set reference point for stack-depth checking.  This might seem
> > >  +         * redundant in !EXEC_BACKEND builds; but it's not because the postmaster
> > >  +         * launches its children from signal handlers, so we might be running on
> > >  +         * an alternative stack. XXX still true?
> > >  +         */
> > >  +        (void) set_stack_base();
> >
> > Looks like there is still this XXX left. Can't say I completely understand
> > the second sentence either.
>
> We used to start some child processes of postmaster in signal handlers. That
> was fixed in
>
> commit 7389aad6366
> Author: Thomas Munro <tmunro@postgresql.org>
> Date:   2023-01-12 12:34:23 +1300
>
>     Use WaitEventSet API for postmaster's event loop.
>
>
> In some cases signal handlers run on a separate stack, which meant that the
> set_stack_base() we did in postmaster would yield a completely bogus stack
> depth estimation.  So this comment should likely have been removed. Thomas?

Right, I should delete that comment in master and 16.  While wondering
what to write instead, my first thought is that it is better to leave
the actual call there though, because otherwise there is a small
difference in stack reference point between EXEC_BACKEND and
!EXEC_BACKEND builds, consumed by a few postmaster stack frames.  So
the new comment would just say that.

(I did idly wonder if there is a longjmp trick to zap those frames
post-fork, not looked into and probably doesn't really improve
anything we care about...)