Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 7 Jun 1998 12:56:27 -0500 (EST)
From:      "John S. Dyson" <toor@dyson.iquest.net>
To:        sos@FreeBSD.ORG
Cc:        toor@dyson.iquest.net, roberto@keltia.freenix.fr, current@FreeBSD.ORG
Subject:   Re: I am getting good reports on the latest -current stuff
Message-ID:  <199806071756.MAA01060@dyson.iquest.net>
In-Reply-To: <199806071737.TAA23667@sos.freebsd.dk> from =?ISO-8859-1?Q?S=F8ren_Schmidt?= at "Jun 7, 98 07:37:04 pm"

next in thread | previous in thread | raw e-mail | index | archive | help
> In reply to John S. Dyson who wrote:
> > > According to John S. Dyson:
> > > > Included in these changes are LL support of per-CPU kernel stacks
> > > > for threaded processes (higher level kernel support isn't there yet),
> > > > a little support for CPU affinity, and much more intelligent usage of
> > > > TLB IPI's.  The number of bugs fixed is overwhelming.
> > > 
> > > Are these patches interesting to UP systems too ? If you have fixed that
> > > many bugs, I guess yes...
> > > 
> > Yes, they should be helpful.  I have not beat on my system under UP like
> > I have using the SMP version of the kernel, but I do know that there were
> > windows in UP mode that this code will fix.  Also, the code has been
> > carefully scheduled to work (nearly) optimally on P5/P6 machines, so
> > there should be a nearly unmeasurable improvement for UP even just with
> > the swtch.s changes.
> 
> Hmm, I just tried the sys*7b patches, and it hangs my dual P6/200 solid
> very fast in a make world, it just locks up, and is totally cactus.
> 
What size memory do you have?  Do you have any hints?

>
> I saw that it turned on WC on the videocards mem (S3 ViRGE/DK 4M), and
> that might be the problem, on the other hand the system wasn't running
> X at the time, so that mem shouldn't be in use....
>
That *might* be a problem, but I highly doubt it.

> 
> Oh, I run with SLICE/DEVFS if that matters, else plain of the hour
> -current sources...
> 
There *might* be a latent bug in the DEVFS code???


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?199806071756.MAA01060>