Xorg 7.4 crashes on exit

[Robert Noland]

On Mon, 2009-03-30 at 16:19 +0100, N.J. Mann wrote:
> In message <1238424768.8491.331.camel at balrog.2hip.net>,
> 	Robert Noland (rnoland at FreeBSD.org) wrote:
> > On Mon, 2009-03-30 at 11:53 +0100, N.J. Mann wrote:
> > > Hi Robert,
> > > 
> > > 
> > > [Moved from svn-src-all@ to x11@ to merge threads.]
> > > 
> > > In message <1237999344.1828.6.camel at balrog.2hip.net>,
> > > 	Robert Noland (rnoland at FreeBSD.org) wrote:
> > > > On Wed, 2009-03-25 at 09:29 +0000, N.J. Mann wrote:
> > > > > In message <200903250150.n2P1ou4g062376 at svn.freebsd.org>,
> > > > > 	Robert Noland (rnoland at FreeBSD.org) wrote:
> > > > > > Author: rnoland
> > > > > > Date: Wed Mar 25 01:50:56 2009
> > > > > > New Revision: 190401
> > > > > > URL: http://svn.freebsd.org/changeset/base/190401
> > > > > > 
> > > > > > Log:
> > > > > >   Rework the management of vblank interrupts a bit.
> > > > > >   
> > > > > >   When a vt switch occurs the irq handler is uninstalled.  Interrupts
> > > > > >   and the state tracking of what was enabled/disabled wasn't working
> > > > > >   properly.  This should resolve the reports of "slow windows" after a
> > > > > >   vt switch, among other things.  The radeon 2d driver seems to work a
> > > > > >   bit more correctly than the Intel driver.  With the Intel driver,
> > > > > >   vblank interrupts will be enabled at system startup and will only
> > > > > >   be disabled after an additional modeset (vt switch, dpms, randr event).
> > > > > >   
> > > > > >   With this patch, I am able to run glxgears synced to vblank and
> > > > > >   vt switch while it is running without ill effects.
> > > > > >   
> > > > > >   MFC after:	3 days
> > > > > > 
> > > > > > Modified:
> > > > > >   head/sys/dev/drm/drm_irq.c
> > > > > 
> > > > > Is this a possible cure for the problem I reported here:
> > > > > http://lists.freebsd.org/pipermail/freebsd-x11/2009-February/007670.html
> > > > 
> > > > It could be... interrupts were not working correcty, especially on
> > > > Intel.  There is likely still an issue with restarting X that I need to
> > > > look into, but I've not seen a problem vt switching or shutting down.
> > > 
> > > Okay I tried out your recent changes this morning.
> > > 
> > > First I did a svn update to r190540 to bring my sources up to date and
> > > then I built a new 7-Stable world and kernel and installed them.  Sadly,
> > > I still got the video mode corruption on exiting X or switching from X
> > > to a VT.  I then merged all the changes from head/sys/dev/drm to
> > > stable/7/sys/dev/drm and tried again.  Unfortunately, I got the same
> > > results.  (Xorg.log and dmesg.boot available if required.)  The changes
> > > I merged were: 190020, 190021, 190022, 190023, 190123, 190124, 190125,
> > > 190163, 190164, 190166, 190170, 190282, 190399, 190400, 19401 and
> > > 190433.  Some of these were probably not relevant, but I did everything
> > > to make it easier for me.  (I am very much still testing the water with
> > > svn and this was the first time I tried doing a merge!)
> > > 
> > > Any ideas where to go from here?
> > 
> > Which chip do you have?  I have a patch now that fixes vblank issues, at
> > least I can't break it.  This corruption issue on shutdown still exists,
> > but I've never had a problem vt switching.  I'm told that something is
> > still wrong with interrupts on g45 though.
> 
> Is this the information you require?
> 
> % grep '^\(vga\|agp\|drm\)' r19054M.dmesg.boot
> vgapci0: <VGA-compatible display> port 0xf140-0xf147 mem 0xff900000-0xff97ffff,0xd0000000-0xdfffffff,0xff700000-0xff7fffff irq 16 at device 2.0 on pci0
> agp0: <Intel G33 SVGA controller> on vgapci0
> vgapci0: Reserved 0x10000000 bytes for rid 0x18 type 3 at 0xd0000000
> vgapci0: Reserved 0x80000 bytes for rid 0x10 type 3 at 0xff900000
> vgapci0: Reserved 0x100000 bytes for rid 0x1c type 3 at 0xff700000
> agp0: detected 7164k stolen memory
> agp0: aperture size is 256M
> drm0: <Intel G33> on vgapci0
> vgapci0: Reserved 0x80000 bytes for rid 0x10 type 3 at 0xff900000
> vgapci0: child drm0 requested pci_enable_busmaster
> vga: vga0 already exists; skipping it
> vga0: <Generic ISA VGA> at port 0x3c0-0x3df iomem 0xa0000-0xbffff on isa0
> 
> The MB is a Intel DG31GL, which incorporates an Intel GMA 3100 graphics
> adapter.

Ok, you can try
http://people.freebsd.org/~rnoland/drm-intel-fix-033009.patch

robert.

> 
> Cheers,
>        Nick.
> -- 
> "Nothing stands for content-free corporate bullshit quite like PowerPoint."
> 
-- 
Robert Noland <rnoland at FreeBSD.org>
FreeBSD
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: This is a digitally signed message part
Url : http://lists.freebsd.org/pipermail/freebsd-x11/attachments/20090330/ffaee2e5/attachment.pgp