public inbox for [email protected]  

help / color / mirror / Atom feed

From: Jeff Davis <[email protected]>
To: Alvaro Herrera <[email protected]>
To: Bharath Rupireddy <[email protected]>
Cc: Tom Lane <[email protected]>
Cc: Andres Freund <[email protected]>
Cc: [email protected]
Cc: Jaime Casanova <[email protected]>
Subject: Re: LogwrtResult contended spinlock
Date: Fri, 05 Apr 2024 11:22:40 -0700
Message-ID: <[email protected]> (raw)
In-Reply-To: <[email protected]>
References: <[email protected]>

On Fri, 2024-04-05 at 13:54 +0200, Alvaro Herrera wrote:
> Couldn't push: I tested with --disable-atomics --disable-spinlocks
> and
> the tests fail because the semaphore for the atomic variables is not
> always initialized.  This is weird -- it's like a client process is
> running at a time when StartupXLOG has not initialized the variable
> ...
> so the initialization in the other place was not completely wrong.

It looks like you solved the problem and pushed the first patch. Looks
good to me.

Patch 0002 also looks good to me, after a mostly-trivial rebase (just
remember to initialize logCopyResult).

Minor comments:

* You could consider using a read barrier before reading the Copy ptr,
or using the pg_atomic_read_membarrier_u64() variant. I don't see a
correctness problem, but it might be slightly more clear and I don't
see a lot of downside.

* Also, I assume that the Max() call in
pg_atomic_monotonic_advance_u64() is just for clarity? Surely the
currval cannot be less than _target when it returns. I'd probably just
do Assert(currval >= _target) and then return currval.

Regards,
	Jeff Davis

view thread (3+ messages)

reply

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Reply to all the recipients using the --to and --cc options:
  reply via email

  To: [email protected]
  Cc: [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected]
  Subject: Re: LogwrtResult contended spinlock
  In-Reply-To: <[email protected]>

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox