Subject: Re: mfi(4) and bio_register()
To: Juan RP <juan@xtrarom.org>
From: Juan RP <juan@xtrarom.org>
List: tech-kern
Date: 11/16/2007 11:09:25
On Fri, 16 Nov 2007 10:50:19 +0100
Juan RP <juan@xtrarom.org> wrote:
> On Fri, 16 Nov 2007 10:42:59 +0100
> Martin Husemann <martin@duskware.de> wrote:
>
> > On Fri, Nov 16, 2007 at 10:34:25AM +0100, Juan RP wrote:
> > > Hi,
> > >
> > > There's currently a problem with mfi(4) on mfi_attach() calling
> > > bio_register(). At the point on where mfi_attach() is called
> > > the pseudo-device bio(4) hasn't been initialized so bio_register()
> > > fails when LOCKDEBUG is used with an uninitialized lock.
> >
> > I would blatantly abuse config_interrupts() for this.
>
> Here's a patch that uses config_interrupts():
Hmm, looks like config_interrupts() seems to have the same problem,
the uninitialized lock is still there.
I've updated the kthread patch which is known to work (missed the
kthread_exit() calls in previous):
http://www.netbsd.org/~xtraeme/mfi_biofix3.diff
--
Juan Romero Pardines - The NetBSD Project
http://plog.xtrarom.org - NetBSD/pkgsrc news in Spanish