Date: Sat, 16 Jun 2007 08:53:35 +0200 From: Gary Jennejohn <garyj@jennejohn.org> To: Craig Rodrigues <rodrigc@crodrigues.org> Cc: current@freebsd.org Subject: Re: recent nvidia-driver trouble Message-ID: <20070616085335.9d5d4032.garyj@jennejohn.org> In-Reply-To: <20070615190107.GA2111@crodrigues.org> References: <1181929982.1086.6.camel@worf> <20070615203136.402a2df5.garyj@jennejohn.org> <20070615190107.GA2111@crodrigues.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, 15 Jun 2007 15:01:07 -0400 Craig Rodrigues <rodrigc@crodrigues.org> wrote: > On Fri, Jun 15, 2007 at 08:31:36PM +0200, Gary Jennejohn wrote: > > I've seen this also. It's a panic due to a NULL-pointer dereference > > someplace in the nvidia-driver. I suspect that it's related to the > > recent locking/mutex/whatever changes but can't prove it. All the > > crash dumps which I have are corrupted so I can't do a backtrace :-( > > > > The bug seems to be in the binary blob part of the driver, which > > means we have to wait for Nvidia to get around to fixing it. > > Have you reported this problem to Nvidia? They don't regularly > monitor FreeBSD mailing lists, but I've found that they > are fairly responsive to e-mails sent to: > freebsd-gfx-bugs@nvidia.com > I thought about it but haven't done anything yet. I don't feel that I really have enough information for a meaningful bug report because I can't analyze any crash dumps. Maybe that wouldn't matter to Nvidia. My suspicion about the binary blob is based on the panic output, which points at a macro which examines td_refcnt. I find it hard to believe that the pointer to the thread is NULL when this macro is used. That's why I suspect that the real culprit is someplace in the locking used by Nvidia, but without a usable crash dump it's hard to verify. --- Gary Jennejohn / garyjATjennejohnDOTorg gjATfreebsdDOTorg garyjATdenxDOTde
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20070616085335.9d5d4032.garyj>