Date: Fri, 26 Jul 2002 10:47:16 -0600 From: Scott Long <scott_long@btc.adaptec.com> To: "Brian F. Feldman" <green@freebsd.org> Cc: Josef Karthauser <joe@tao.org.uk>, cvs-committers@freebsd.org, cvs-all@freebsd.org Subject: Re: cvs commit: src/sys/dev/sound/pcm mixer.c sndstat.c sound.c Message-ID: <20020726164716.GA74649@hollin.btc.adaptec.com> In-Reply-To: <200207261643.g6QGhGJ66380@green.bikeshed.org> References: <20020726161514.GC40513@genius.tao.org.uk> <200207261643.g6QGhGJ66380@green.bikeshed.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, Jul 26, 2002 at 12:43:16PM -0400, Brian F. Feldman wrote: > Josef Karthauser <joe@tao.org.uk> wrote: > > On Wed, Jul 24, 2002 at 09:49:46PM -0700, Brian Feldman wrote: > > > green 2002/07/24 21:49:45 PDT > > > > > > Modified files: > > > sys/dev/sound/pcm mixer.c sndstat.c sound.c > > > Log: > > > Fix some of the places where sound(4) can sleep with a lock held. (Help > > > courtesy of fenner). > > > > > > > And the crowd goes wild - or at least can boot unattended again :). > > You know,I don't think this code in general really WANTS to use mutexes! > The locks it's using are over entire huge code segments, and it would be > quite a bit of work to converting it to only locking structures. I don't > know that using sx locks alone would solve the possibility of deadlock that > exists now in the initialization routine paths... > Yes, it seems that the mutexes are used to protect code paths rather than data structures. It all needs a big re-write. Andrew Reiter (sp?) recently commented on this too and implied that he'd like to fix it. Andrew? To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020726164716.GA74649>