Re: [PATCH] audo-detect and use -moutline-atomics compilation flag for aarch64

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: "Zidenberg, Tsahi" <tsahee(at)amazon(dot)com>
Cc: "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: [PATCH] audo-detect and use -moutline-atomics compilation flag for aarch64
Date: 2020-09-07 22:00:54
Message-ID: 1713568.1599516054@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

"Zidenberg, Tsahi" <tsahee(at)amazon(dot)com> writes:
> Outline-atomics is a gcc compilation flag that adds runtime detection of weather or not the cpu supports atomic instructions. CPUs that don't support atomic instructions will use the old load-exclusive/store-exclusive instructions. If a different compilation flag defined an architecture that unconditionally supports atomic instructions (e.g. -march=armv8.2), the outline-atomic flag will have no effect.

I wonder what version of gcc you intend this for. AFAICS, older
gcc versions lack this flag at all, while newer ones have it on
by default. Docs I can find on the net suggest that it would only
help to supply the flag when using gcc 10.0.x. Is there a sufficient
population of production systems using such gcc releases to make it
worth expending configure cycles on? (That's sort of a trick question,
because the GCC docs make it look like 10.0.x was never considered
to be production ready.)

regards, tom lane

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Thomas Munro 2020-09-07 22:03:28 Re: A micro-optimisation for walkdir()
Previous Message Peter Geoghegan 2020-09-07 21:31:07 Re: v13: show extended stats target in \d