Date: Wed, 2 Jul 2003 17:42:09 -0400 From: John D <jwd@bsdwins.com> To: Jeff Roberson <jroberson@chesapeake.net> Cc: current@freebsd.org Subject: Re: ULE problems on HTT SMP Message-ID: <20030702214209.GA56428@BSDWins.Com> In-Reply-To: <20030701142802.G17881-100000@mail.chesapeake.net> References: <16128.16880.618174.525346@grasshopper.cs.duke.edu> <20030701142802.G17881-100000@mail.chesapeake.net>
next in thread | previous in thread | raw e-mail | index | archive | help
----- Jeff Roberson's Original Message -----
> On Mon, 30 Jun 2003, Andrew Gallatin wrote:
> > Does this mean that if, as a temporary measure, I disable
> > machdep.cpu_idle_hlt, ULE should work for me?
> >
> Yes, but it needs to be disabled before booting so you'll have to adjust
> it in the code. See i386/i386/mp_machdep.c
Success :-)
Making the above change in mp_machdep.c allows a SMP HTT ULE
kernel to boot and run. I ran a few 'make buildworld's with
the BSD & ULE schedulers:
bankshot# tail bw.ule*.log | grep real
4324.49 real 1829.01 user 1460.72 sys
4324.20 real 1831.00 user 1459.55 sys
4324.50 real 1829.10 user 1462.96 sys
bankshot# tail bw.bsd*.log | grep real
3317.29 real 1859.33 user 1360.96 sys
3237.46 real 1846.39 user 1367.89 sys
3243.40 real 1848.00 user 1369.43 sys
Any tunable ideas to help bring sys and real times
back inline? Dmesg for the machine used above is
http://rtp.freebsd.org/~jwd/dmesg/dmesg.bankshot
Keep up the good work!
-John
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030702214209.GA56428>