Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 07 Sep 2004 10:36:59 -0600
From:      Scott Long <scottl@samsco.org>
To:        rnejdl@ringofsaturn.com
Cc:        current@freebsd.org
Subject:   Re: Beta 3 Crash with machdep.hlt_logical_cpus
Message-ID:  <413DE3AB.9080707@samsco.org>
In-Reply-To: <33307.12.148.147.242.1094574244.squirrel@mail.ringofsaturn.com>
References:  <33307.12.148.147.242.1094574244.squirrel@mail.ringofsaturn.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Rusty Nejdl wrote:
> I am running on a Dual 1.8 Ghz Xeon box and I have hyperthreading turned
> on in the BIOS such that it appears that I have 4 processors.  I am not
> physically at the machine right now but wanted to report a problem I had
> today.  I changed machdep.hlt_logical_cpus from 0 to 1 while running and
> the system completely froze up.  I don't know what error occurred as I
> wasn't at the machine and have rebooted since then.  I do know that using
> 5.2 I was able to do this and the sytstem would move all the processes to
> the correct CPU's.  Not sure if anyone else has seen this.
> 
> Rusty Nejdl
> 

This is a known problem with the SCHED_ULE scheduler.  Fixing it doesn't
appear to be trivial, so we are considering other alternatives for
RELENG_5 and 5.3, including switching back to the 4BSD scheduler.  The
4BSD scheduler also appears to be much more stable than ULE, especially
with kernel preemption enabled.

Scott



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?413DE3AB.9080707>