Date: Tue, 13 Jul 1999 14:10:45 -0400 (EDT) From: "Brian F. Feldman" <green@FreeBSD.org> To: Matthew Dillon <dillon@apollo.backplane.com> Cc: Peter Jeremy <jeremyp@gsmx07.alcatel.com.au>, freebsd-current@FreeBSD.org Subject: Re: LOCK overheads (was Re: "objtrm" problem probably found) Message-ID: <Pine.BSF.4.10.9907131409350.86113-100000@janus.syracuse.net> In-Reply-To: <199907131555.IAA78738@apollo.backplane.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, 13 Jul 1999, Matthew Dillon wrote:
>
> :Here's another data point:
> :{"/home/green"}$ cc -O2 locktests.c -o locktests
> :{"/home/green"}$ ./locktests
> :mode 0 10.92 ns/loop nproc=1 lcks=EMPTY
> :mode 1 17.99 ns/loop nproc=1 lcks=no
> :mode 2 17.97 ns/loop nproc=2 lcks=no
> :mode 3 166.33 ns/loop nproc=1 lcks=yes
> :mode 4 167.91 ns/loop nproc=2 lcks=yes
> :mode 5 42.16 ns/loop nproc=1 lcks=no
> :mode 6 41.60 ns/loop nproc=2 lcks=no
> :mode 7 188.86 ns/loop nproc=1 lcks=yes
> :mode 8 191.56 ns/loop nproc=2 lcks=yes
> :
> :This is a K6-2 350. Locks are pretty expensive on them.
> :
> : Brian Fundakowski Feldman _ __ ___ ____ ___ ___ ___
>
> Wow, now that *is* expensive! The K6 must be implementing it in
> microcode for it to be that bad.
Sometimes you wonder what they're doing with that RISC86 translation.
I have a feeling it's overhead for the OpenPIC implementation, really.
>
> -Matt
> Matthew Dillon
> <dillon@backplane.com>
>
Brian Fundakowski Feldman _ __ ___ ____ ___ ___ ___
green@FreeBSD.org _ __ ___ | _ ) __| \
FreeBSD: The Power to Serve! _ __ | _ \._ \ |) |
http://www.FreeBSD.org/ _ |___/___/___/
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.10.9907131409350.86113-100000>