Re: PostmasterIsAlive() in recovery (non-USE_POST_MASTER_DEATH_SIGNAL builds)

On Thu, Sep 17, 2020 at 10:47 PM Heikki Linnakangas <hlinnaka(at)iki(dot)fi> wrote:
> Hmm, so for speedy response to postmaster death, you're relying on the
> loops to have other postmaster-death checks besides
> HandleStartupProcInterrupts(), in the form of WL_EXIT_ON_PM_DEATH. That
> seems a bit fragile, at the very least it needs a comment in
> HandleStartupProcInterrupts() to call it out.

Surely that's the direction we want to go in, though, no? Commit
cfdf4dc4 was intended to standardise the way we react to postmaster
death where waiting is involved. I updated the comment in
HandleStartupProcInterrupts() to highlight that the
PostmasterIsAlive() check in there is only for the benefit of
CPU-bound loops.

> Note that there's one more loop in ShutdownWalRcv() that uses pg_usleep().

Updating that one required me to invent a new wait_event for
pg_stat_activity, which seems like progress.

Unfortunately, while I was doing that I realised that WaitLatch()
without WL_SET_LATCH was broken by commit 3347c982bab (in master
only), in a way not previously reached by the tests. So 0001 is a
patch to fix that.