Date: Wed, 27 Dec 2006 16:50:05 -0500 From: John Baldwin <jhb@freebsd.org> To: Bruce Evans <bde@freebsd.org> Cc: cvs-src@freebsd.org, src-committers@freebsd.org, cvs-all@freebsd.org Subject: Re: cvs commit: src/sys/amd64/include atomic.h src/sys/i386/include atomic.h Message-ID: <200612271650.06432.jhb@freebsd.org> In-Reply-To: <200612272026.kBRKQ1gP000873@repoman.freebsd.org> References: <200612272026.kBRKQ1gP000873@repoman.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wednesday 27 December 2006 15:26, Bruce Evans wrote:
> bde 2006-12-27 20:26:00 UTC
>
> FreeBSD src repository
>
> Modified files:
> sys/amd64/include atomic.h
> sys/i386/include atomic.h
> Log:
> Avoid an instruction in atomic_cmpset_{int_long)() in most cases.
> These functions are used a lot for mutexes, so this reduces the text
> size of an average kernel by about 0.75%. This wasn't intended to
> be a significant optimization, but it somehow increased the maximum
> number of packets per second that can be transmitted by my bge hardware
> from 320000 to 460000 (this benchmark is CPU-bound and remarkably
> sensitive to changes in the text section).
Very cool, thanks! It'd be nice to get rid of the 'sete' as well (if there
were someway to let gcc know that a bool return value was mapped to a flag
in %eflags), but this fix alone is most welcome.
--
John Baldwin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200612271650.06432.jhb>