Horton

Re: Fix performance of generic atomics

Tom Lane <tgl@sss.pgh.pa.us>

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Andres Freund <andres@anarazel.de>
Cc: Sokolov Yura <funny.falcon@postgrespro.ru>, Robert Haas <robertmhaas@gmail.com>, Simon Riggs <simon@2ndquadrant.com>, Jeff Janes <jeff.janes@gmail.com>, Jesper Pedersen <jesper.pedersen@redhat.com>, pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2017-09-06T19:54:20Z
Lists: pgsql-hackers 
I wrote:
> Ah.  I was not thinking of touching pg_atomic_read_u32/u64_impl,
> although now that you mention it, it's not clear to me why we
> couldn't simplify

> -	return *(&ptr->value);
> +	return ptr->value;

Just to check, I applied that change to pg_atomic_read_u32_impl and
pg_atomic_read_u64_impl, and recompiled.  I get bit-for-bit the
same backend executable.  Maybe it would have an effect on some
other compiler, but I doubt it, except perhaps at -O0.

			regards, tom lane

Commits

  1. Further marginal hacking on generic atomic ops.

  2. Use more of gcc's __sync_fetch_and_xxx builtin functions for atomic ops.

  3. Remove duplicate reads from the inner loops in generic atomic ops.