Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 12 Jan 1999 08:49:40 -0700
From:      Steve Passe <smp@csn.net>
To:        Alfred Perlstein <bright@hotjobs.com>
Cc:        freebsd-smp@FreeBSD.ORG
Subject:   Re: Fatal Trap 
Message-ID:  <199901121549.IAA22317@Ilsa.StevesCafe.com>
In-Reply-To: Your message of "Mon, 11 Jan 1999 19:29:35 EST." <Pine.BSF.4.05.9901111928140.84303-100000@bright.fx.genx.net> 

next in thread | previous in thread | raw e-mail | index | archive | help
Hi,

> On Mon, 11 Jan 1999, Greg Rowe wrote:
 ...
> > interrupt mask = net tty bio cam <-SMP:XXX
> 
> why is this XXX?  is it just an indication that using spl() is a bad thing
> for SMP and that movement away from the spl() model of interupt
> servicing is a good thing?

I'm the one responsible for that line, but I can't honestly remember its
exact significance.  I'm not even sure if it stands for "Symmetric Multi
Processor", or if its my initials (probably the former).  I put it there
at the time I was attempting to decouple the intr aspect of spl() from
the process aspect (ie. cpl vs. ipl).  Anyways, your statement is accurate
in that spl() must go away in favor of mutex() style programming
b4 the "giant lock" problem can be fixed.  I beat my head on that wall
for quite awhile b4 giving up...

--
Steve Passe	| powered by 
smp@csn.net	|            Symmetric MultiProcessor FreeBSD



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-smp" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199901121549.IAA22317>