From: | Robert Haas <robertmhaas(at)gmail(dot)com> |
---|---|
To: | Andreas Seltenreich <seltenreich(at)gmx(dot)de> |
Cc: | "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: [PATCH] Improve spinlock inline assembly for x86. |
Date: | 2016-01-18 18:27:59 |
Message-ID: | CA+Tgmoay6NVAag4kGPpwMSFGmE0ZNvpOgY2enAZhqNNvOyHrDA@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Sun, Jan 17, 2016 at 6:38 AM, Andreas Seltenreich <seltenreich(at)gmx(dot)de> wrote:
> I'm currently experimenting with just-in-time compilation using libfirm.
> While discussing issues with its developers, it was pointed out to me
> that our spinlock inline assembly is less than optimal. Attached is a
> patch that addresses this.
>
> ,----
> | Remove the LOCK prefix from the XCHG instruction. Locking is implicit
> | with XCHG and the prefix wastes a byte. Also remove the "cc" register
> | from the clobber list as the XCHG instruction does not modify any flags.
> |
> | Reported by Christoph Mallon.
> `----
I did a Google search and everybody seems to agree that the LOCK
prefix is redundant. I found a source agreeing with the idea that it
doesn't clobber registers, too:
http://www.oopweb.com/Assembly/Documents/ArtOfAssembly/Volume/Chapter_6/CH06-1.html#HEADING1-85
So I guess it would be safe to change this. Scares me slightly, though.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
From | Date | Subject | |
---|---|---|---|
Next Message | Robert Haas | 2016-01-18 18:43:54 | Re: Proposal: speeding up GIN build with parallel workers |
Previous Message | Robert Haas | 2016-01-18 18:17:22 | Re: Truncating/vacuuming relations on full tablespaces |