Date: Thu, 30 Mar 2000 17:49:32 -0800 (PST) From: Matthew Dillon <dillon@apollo.backplane.com> To: Bob Bishop <rb@gid.co.uk> Cc: freebsd-current@FreeBSD.ORG Subject: Re: SMP buildworld times / performance tests Message-ID: <200003310149.RAA71632@apollo.backplane.com> References: <Pine.BSF.4.10.10003301158360.14426-100000@backup.af.speednet.com.au> <200003300604.WAA68031@apollo.backplane.com> <v0422080ab508dbb83f59@[195.238.22.136]> <l03130306b509abed48dc@[194.32.164.2]>
next in thread | previous in thread | raw e-mail | index | archive | help
:>:At 10:04 PM -0800 2000/3/29, Matthew Dillon wrote:
:>[...]
:> 6 minutes 20 seconds (about 7%).
:
:I'm seeing the same order of improvement still.
:
:--
:Bob Bishop (0118) 977 4017 international code +44 118
:rb@gid.co.uk fax (0118) 989 4254 between 0800 and 1800 UK
Ok, excellent. I figured out why my I/O rate was still high -- I
also hadn't turned on softupdates for /var/tmp, but it didn't make
much of a difference since I use -pipe in my compiler options so
I am still hanging on around 5-7% too.
That's still significant, but not as fun a number as the first one :-).
I am coordinating one more patch set with Mike that makes sigprocmask
and the core copyout function MP safe (so both copyin and copyout are
MP safe). Copyout is basically MP safe already, sigprocmask
needed only minor adjustments. Then I'm going to turn the code loose
in 5.0 and, in a week or two, backport it to 4.0. Then it will be up to
the rest of the community to push the MP lock further, I don't have
as much time on my hands as I used to :-).
-Matt
Matthew Dillon
<dillon@backplane.com>
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?200003310149.RAA71632>