latest drm patches

[Paul B. Mahol]

On 10/2/08, Robert Noland <rnoland at freebsd.org> wrote:
> On Thu, 2008-10-02 at 23:47 +0200, Paul B. Mahol wrote:
>> On 10/2/08, Paul B. Mahol <onemda at gmail.com> wrote:
>> > On 10/2/08, Robert Noland <rnoland at freebsd.org> wrote:
>> >> On Thu, 2008-10-02 at 23:16 +0200, Paul B. Mahol wrote:
>> >>> On 10/2/08, Robert Noland <rnoland at freebsd.org> wrote:
>> >>> > On Thu, 2008-10-02 at 22:38 +0200, Paul B. Mahol wrote:
>> >>> >> On 10/2/08, Robert Noland <rnoland at freebsd.org> wrote:
>> >>> >> > I have made new patch sets for both -CURRENT and -STABLE.  They
>> >>> >> > are
>> >>> >> > located at:
>> >>> >> >
>> >>> >> > http://people.freebsd.org/~rnoland/drm-update-7-100108.patch.bz2
>> >>> >> > http://people.freebsd.org/~rnoland/drm-update-8-100108.patch.bz2
>> >>> >> >
>> >>> >> > Note that if your are using RELENG_7, you will need to be very
>> >>> >> > current.
>> >>> >> > i.e. on or after 2008-09-29 16:20:13 -0400.  CURRENT should be at
>> >>> >> > least
>> >>> >> > 2008-09-20 15:56:02 -0400.
>> >>> >> >
>> >>> >> > This is a re-sync to git master, which seems to address many of
>> >>> >> > the
>> >>> >> > issues with the intel chipsets < 965.  I'm still not certain
>> >>> >> > exactly
>> >>> >> > which change fixes them unfortunately.  This update contains a
>> >>> >> > lot
>> >>> >> > of
>> >>> >> > code cleanup and is post gem merge (no, we don't have gem
>> >>> >> > support).
>> >>> >> > It
>> >>> >> > should prove much easier to read the code now.  A lot of thanks
>> >>> >> > goes
>> >>> >> > to
>> >>> >> > vehemens for that work.  I have adapted the code to use cdevpriv
>> >>> >> > for
>> >>> >> > tracking per open file data, which is the reason that you need
>> >>> >> > really
>> >>> >> > current bits to use this patch.  That alleviates the old ugly
>> >>> >> > hack
>> >>> >> > that
>> >>> >> > we used to try and accomplish the task and helped to clean up the
>> >>> >> > open /
>> >>> >> > close behavior a good bit.  This also replaces the hack that was
>> >>> >> > put
>> >>> >> > in
>> >>> >> > place a year or so ago to prevent radeons from locking up with
>> >>> >> > AIGLX
>> >>> >> > enabled.  I have had a couple of radeon testers report that it
>> >>> >> > still
>> >>> >> > works as expected, though I no longer have radeon hardware to
>> >>> >> > test
>> >>> >> > with
>> >>> >> > myself.  Other various fixes from the linux crew and Intel, many
>> >>> >> > of
>> >>> >> > which are muddled in with the gem merge.
>> >>> >> >
>> >>> >> > I am planning to push this into CURRENT pretty soon, possibly
>> >>> >> > even
>> >>> >> > sometime tomorrow once I have a chance to discuss with a few
>> >>> >> > others.
>> >>> >>
>> >>> >> Wow, something is broken again (glxgears doenst work again). I
>> >>> >> checked
>> >>> >> twice (with two versions of agp)
>> >>> >
>> >>> > Are you sure that your -CURRENT meets the requirements that I
>> >>> > stated?
>> >>>
>> >>> FreeBSD dhcppc1 8.0-CURRENT FreeBSD 8.0-CURRENT #6: Sun Sep 28
>> >>> 11:46:55 CEST 2008
>> >>
>> >> Hrm, ok... You say that git from yesterday is working.  I pushed one of
>> >> the 2 differences last night, so let's try this.
>> >>
>> >> cd into your git directory
>> >> git pull
>> >> cd bsd-core && make clean && make all install
>> >>
>> >> Hopefully that will tell us which one has the issue.
>> >
>> > OK:
>> >
>> >> git pull
>> > remote: Counting objects: 7, done.
>> > remote: Compressing objects: 100% (4/4), done.
>> > remote: Total 4 (delta 3), reused 0 (delta 0)
>> > Unpacking objects: 100% (4/4), done.
>> > From git://anongit.freedesktop.org/git/mesa/drm
>> >    073cb5e..81952c7  master     -> origin/master
>> > Updating 073cb5e..81952c7
>> > Fast forward
>> >  bsd-core/drm_vm.c |   21 ++++++++++++---------
>> >  1 files changed, 12 insertions(+), 9 deletions(-)
>> >
>> > Let we see would this change break something :) ...
>> 
>> It did not, DRI works again.
>
> Sigh... ok, please try applying the following patch in REVERSE to your
> kernel tree.
>
> http://people.freebsd.org/~rnoland/0001--FreeBSD-This-eliminates-a-potential-panic-but-can.patch

Bingo, reversing that patch fixed problem.
I reversed it on top of patch from this thread.

I'm interested what comit really fixed drm module unloading.