Re: drm2, i915kms cause instant lock-up

From: Steve Kargl <sgk_at_troutmask.apl.washington.edu>
Date: Tue, 21 Feb 2017 21:37:41 -0800
On Wed, Feb 22, 2017 at 07:29:08AM +0200, Konstantin Belousov wrote:
> On Tue, Feb 21, 2017 at 10:55:11AM -0800, Steve Kargl wrote:
> > On Mon, Feb 20, 2017 at 09:26:58PM -0800, Steve Kargl wrote:
> > > 
> > > Well, the good news seems to be that r313254 and older are 'ok'.
> > > So, something between r313943 and r313254 is triggering a the
> > > problem.  I'm still bisecting, but it might take a day or two.
> > > 
> > 
> > I've been able to narrow the range down to r313854 to r313943.
> > If I had to guess, the issue may be related to 
> > 
> > Author: kib
> > Date: Fri Feb 17 21:08:32 2017
> > New Revision: 313898
> > URL: https://svnweb.freebsd.org/changeset/base/313898
> > 
> > Log:
> >   Merge i386 and amd64 mtrr drivers.
> > 
> > I won't be able to investigate until later tonight (~ 10 hours from now).
> 
> >From what I see in other messages, you are using i386 kernel on Core2
> class machine, am I right ?  Did r313897 worked fine ?
> 
> r313898 has a bug for i386 architecture, which was fixed in r313934.
> Could you compile kernel from r313898 sources with r313934 applied on
> top of it ?  I mean, take r313898 and apply the changes from r313934
> either manually or with patch, but not take any further changes from
> svn after r313898.
> 

I just completed the bisection.  r313934 is the problem.
'svn update -r 313933' boots and I can load drm2.ko.
'svn update -r 313934' boots and hangs on loading the
module.

As I noted in another reply I just sent, around -r313902
I can load the i915kms.ko but console output to vt is sloooow.
Perhaps, a locking issue that 313934 exacerbated.

-- 
Steve
20161221 https://www.youtube.com/watch?v=IbCHE-hONow
Received on Wed Feb 22 2017 - 04:37:42 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:41:10 UTC