Date: Thu, 17 Apr 2003 13:20:42 +1000 (EST) From: Bruce Evans <bde@zeta.org.au> To: John Baldwin <jhb@FreeBSD.org> Cc: src-committers@FreeBSD.org Subject: Re: cvs commit: src/sys/dev/wl if_wl.c Message-ID: <20030417131012.D2575@gamplex.bde.org> In-Reply-To: <XFMail.20030416161124.jhb@FreeBSD.org> References: <XFMail.20030416161124.jhb@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 16 Apr 2003, John Baldwin wrote: > On 16-Apr-2003 John Hay wrote: > > On Wed, Apr 16, 2003 at 02:06:25PM -0400, John Baldwin wrote: > >> Note that some of the spl calls covered structures other than > >> just the softc. I'd prefer that we leave spl calls in until > >> all the structures they cover are locked. > > > > Ok, should I put them all back? Or just the ones being used as > > interrupt disablers? And what about the places where there > > should have been splx()es but weren't? > > I would put back any spl's covering things not covered by the > locks you added. :) Adding in missing spl's is not necessarily > required, so doing or not doing that is your choice. I think it should be rare to not need locks in much the same places than there were spl's. Anything else would tend to give even coarser locking than that provided by spls. I found it instructive that in a couple of places where I had fixed spl's being held too long (across M_WAITOK malloc()s), mutex locking had to add locks in much the same places that I had added spl's. Bruce
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030417131012.D2575>