[RS690M] newest mesa update makes kwin compositing constantly crash to kdm

Bug #446578 reported by marcinq
44
This bug affects 6 people
Affects Status Importance Assigned to Milestone
Mesa
Fix Released
High
mesa (Debian)
Fix Released
Unknown
mesa (Ubuntu)
Fix Released
Undecided
Unassigned

Bug Description

After the recent mesa update kwin composite in kubuntu is broken. Loging in from kdm works, the desktop starts, but when opening any aplication window (for example system settings) Xorg crashes and I'm back at kdm login. From this point also changing to terminal ctrl+alt+f1 doesn't work as by doing that I'm left with kdm bitmap and some graphical garbage is produced at the top of the screen.

When compositing is disabled everything is ok, also strangely when booting up the kernel with radeon.modset=1 and using KMS compositing in kde works as it should and the bug is gone.

My xorg.conf is unmodified default one so I'm not attaching it. The card is radeon x1200 (rs690) on a toshiba a215-7416 laptop.

Related branches

Revision history for this message
In , Jure Repinc (jlp) wrote :

Created an attachment (id=29824)
xorg.conf

Revision history for this message
In , Jure Repinc (jlp) wrote :

Created an attachment (id=29825)
Xorg.0.log.old

Revision history for this message
In , Jure Repinc (jlp) wrote :

Created an attachment (id=29826)
kdm.log

Revision history for this message
In , Fatih Aşıcı (fatih-pardus) wrote :

Same problem with HP Compaq nc6400 containing:

01:00.0 VGA compatible controller [0300]: ATI Technologies Inc M52 [Mobility Radeon X1300] [1002:714a]

I have bisected on 7.6 branch. This is the result:

284a7af274bc148f112bd0ebb40583923ee26b49 is first bad commit
commit 284a7af274bc148f112bd0ebb40583923ee26b49
Author: Pauli Nieminen <email address hidden>
Date: Sun Sep 20 22:24:35 2009 +0300

    radeon: Fix legacy bo not to reuse dma buffers before refcount is 1.
····
    This should help detecting possible memory leaks with dma buffers and prevent
    possible visual corruption if data would be overwriten too early.

:040000 040000 2fada933dbda54c97f8c47678f2bb03ff05d09bf 1b9c0f0425df9851994f7fa1e095c9a51e79d2c4 M src

Revision history for this message
In , Michel Dänzer (michel-daenzer) wrote :

Any ideas Pauli?

Revision history for this message
In , Steffen-schloenvoigt (steffen-schloenvoigt) wrote :

Same problem here on Mobility x1400 with latest packages from the openSuse Xorg:X11 repository. Happens for me only when KDE 4 desktop-effects are enabled, though. If I disable them, I can work as usual. Enabling them leads to X crashing within a couple of seconds. Couldn't find the error message in Xorg.0.log but kdm.log catched it as well.

Revision history for this message
In , Bugzi09-fdo-tormod (bugzi09-fdo-tormod) wrote :

Does anyone know a way of reproducing this other than installing KDE?

Revision history for this message
In , Val-gaav (val-gaav) wrote :

On my radeon x1200 (rs690) the desktop loads but when opening any application window (for example system settings) Xorg crashes and I'm back at kdm login. From this point on changing to terminal ctrl+alt+f1 doesn't work as by doing that I'm left with kdm bitmap and some graphical garbage is produced at the top of the screen.

When compositing is disabled everything is ok, also strangely when booting up the kernel with radeon.modeset=1 and using KMS compositing in kde works as it should and the bug is gone.

Revision history for this message
marcinq (marcinq) wrote : [karmic] newest mesa update makes kwin compositing constantly crash to kdm on radeon x1200 (rs690)

After the recent mesa update kwin composite in kubuntu is broken. Loging in from kdm works, the desktop starts, but when opening any aplication window (for example system settings) Xorg crashes and I'm back at kdm login. From this point also changing to terminal ctrl+alt+f1 doesn't work as by doing that I'm left with kdm bitmap and some graphical garbage is produced at the top of the screen.

When compositing is disabled everything is ok, also strangely when booting up the kernel with radeon.modset=1 and using KMS compositing in kde works as it should and the bug is gone.

My xorg.conf is unmodified default one so I'm not attaching it. The card is radeon x1200 (rs690) on a toshiba a215-7416 laptop.

Revision history for this message
marcinq (marcinq) wrote :
Bryce Harrington (bryce)
summary: - [karmic] newest mesa update makes kwin compositing constantly crash to
- kdm on radeon x1200 (rs690)
+ [RS690M] newest mesa update makes kwin compositing constantly crash to
+ kdm
tags: added: kubuntu
Revision history for this message
Bryce Harrington (bryce) wrote :

Hi Marcinq,

This sounds like it might be a release critical bug for us. You could help us a LOT if you could collect a full backtrace on this crash.

For information on how to collect backtraces, directions are at http://wiki.ubuntu.com/X/Backtracing - let me know if you have any questions.

Can you also attach the output of `dmesg` and your /var/log/dpkg.log for reference?

tags: added: regression-potential
Revision history for this message
Tormod Volden (tormodvolden) wrote :

Thanks for your bug report. Can you please also attach your gdm log, from /var/log/gdm/ there is :0.log and :0.log.1 etc, and one of them has hopefully a stacktrace or other error messages.

Changed in mesa (Ubuntu):
assignee: nobody → Tormod Volden (tormodvolden)
status: New → Incomplete
Revision history for this message
Tormod Volden (tormodvolden) wrote :

Would you have the chance to try a 32-bit live CD? I wonder if it is a 64-bit issue.

Revision history for this message
marcinq (marcinq) wrote :
Revision history for this message
marcinq (marcinq) wrote :
Revision history for this message
marcinq (marcinq) wrote :

Apport unfortunately doesn't catch this issue (I've checked /var/crash too after getting no pop up)..

I can try the gdb/remote method but only tomorrow when I'm likely to have access to a second machine.

I'll download the 32 bit CD and check up on this too.

Revision history for this message
Tormod Volden (tormodvolden) wrote :

From the kdm log:
X: radeon_bo_legacy.c:207: legacy_is_pending: Assertion `bo_legacy->is_pending <= bo->cref' failed.

Revision history for this message
Tormod Volden (tormodvolden) wrote :

marcinq, thanks for posting the logs so quickly. Upstream is aware of the issue, and it is confirmed by other users. It would be nice if you also subscribe to the upstream bug report http://bugs.freedesktop.org/show_bug.cgi?id=24131 in case they need people to test things.

Good that there is KMS as a workaround :)

Changed in mesa (Ubuntu):
status: Incomplete → Confirmed
assignee: Tormod Volden (tormodvolden) → nobody
Revision history for this message
Bryce Harrington (bryce) wrote :

marcinq, in the upstream bug report there is a specific change flagged when the crash first appeared. To better understand the cause of this crash it would be worthwhile to test if dropping that change makes the crash go away.

For your convenience in testing this I've created a ubuntu package with this patch reverted. Could you please install from this ppa and restart X, and see if you are able to start without kwin crashing. This will give some good insight into how we can avoid this problem for kubuntu users.

https://edge.launchpad.net/~bryceharrington/+archive/beige

Revision history for this message
In , Bugzi09-fdo-tormod (bugzi09-fdo-tormod) wrote :

I confirmed the crash on RV515 on 32-bit by installing kwin and that reverting the commit fixes it.

Revision history for this message
marcinq (marcinq) wrote :

Thanks, that's really convenient :) Yes it does solve the issue. No crashes in that version.

However that version produces some white lines garbage on screen when moving windows, or when the dim effect for kdesudo is active. Those lines stay there and are refreshed/gone after next window move, but that produces new ones. I'm attaching a picture with that problem. Once again this does not happen when using kms, but kms is a lot slower on this card then normal/old mode.

Revision history for this message
Tormod Volden (tormodvolden) wrote :

marcinq, I think your white lines might be bug 446425.

Revision history for this message
Tormod Volden (tormodvolden) wrote :

I am able to reproduce both issues on RV515, by installing kwin and running "kwin --replace", then right click on the window frame to "Configure Window Behaviour" and "Enable Desktop Effects". Just moving windows around and switching between them with Alt-Tab does trigger the issues.

Revision history for this message
Tormod Volden (tormodvolden) wrote :

I have verified that 7.6.0-1ubuntu2~bug446578~1 fixes the issue in this report.

Changed in mesa:
status: Unknown → Confirmed
Revision history for this message
NoN (tsm339) wrote :

Could this bug be related to what I'm experiencing on my old-ish R200 ? Whenever I run KDE 4.3 with desktop effects enabled, opening any menu or window results in a freeze- the mouse can still be moved but the menus don't respond to clicks and the keyboard is completely unresponsive, requiring a hard reboot. I'm also getting some trash on the screen for a second, every time I log into KDE. Don't know if that's related.

Everything works fine without Kwin effects and compositing works fine in GNOME.

Revision history for this message
Tom Pringle (thomas-b-pringle) wrote :

I filed a bug to report this problem which was marked as a duplicate of this bug.

I see this problem on 2 64bit installations of karmic. Can I use the package sited above?

>Tormod Volden wrote on 2009-10-09: #14
>I have verified that 7.6.0-1ubuntu2~bug446578~1 fixes the issue in this report.

 I added the ppa to my sources.list and updated but upgrade didn't pick up any new mesa packages. What else am I supposed to do?

Thanks,
Tom

Revision history for this message
Tom Pringle (thomas-b-pringle) wrote :

I also tried setting radeon.modeset=1 on the kernel command line but it did not seem to help. I am not sure I put this parameter in the correct place though.

I hit the shift key during boot to bring up the grub menu
selected 'e' to edit the boot menu entry
then appended the radeon.modeset=1 to the command line starting with linux
then ctrl-x to boot

Revision history for this message
Tormod Volden (tormodvolden) wrote :

Tom, since there is now a new 7.6.0-1ubuntu2 in karmic, the 7.6.0-1ubuntu2~bug446578~1 will not be installed. You can try this to force the wanted version:
 sudo apt-get install libgl1-mesa-dri=7.6.0-1ubuntu2~bug446578~1 libgl1-mesa-glx=7.6.0-1ubuntu2~bug446578~1 libglu1-mesa=7.6.0-1ubuntu2~bug446578~1 mesa-utils=7.6.0-1ubuntu2~bug446578~1

It sounds like you set the modeset option correctly though.

Revision history for this message
tzekwangteo (tkteo) wrote :

I can report that forcing the install of 7.6.0-1ubuntu2~bug446578~1 on my notebook PC with ATI Radeon Xpress 200M graphics chip has solved the problem thus far. I can (re)enable compositing and desktop effects on Kubuntu Karmic Beta with KDE 4.3.2 and not experience the crash back to kdm login screen.

I am running the latest kernel 2.6.31-13. So I tried both booting without Radeon KMS enabled and with "radeon.modeset=1". Enabling "radeon.modeset=1" slows down my system, but on the other hand, leaving Radeon KMS off by current default creates the situation whereby I can see certain visual artifacts on the screen.

Revision history for this message
Nathan (nathanbailey) wrote :

I can also report that I am getting kwin crashing, however it is not as "fatal." I am still able to open applications and I don't end up at the login screen when it happens. But it is happening regularly and apport is not able to collect for it. I am running KDE amd64 with kernel 2.6.31-13 on a HP HDX 16 with nVidia GeForce 9600M GT and nVidia driver version 185.18.36 (with of course compositing / desktop effects). Since it is not really bothering me I don't think I'll change anything unless someone wants me to for testing sake. If you need any files let me know and I'll be glad to post these.

Revision history for this message
Tormod Volden (tormodvolden) wrote :

Nathan, if you using the proprietary nvidia drivers, you are not using the mesa libraries.

Revision history for this message
In , Modestas Vainius (geromanas) wrote :
Download full text (3.2 KiB)

1:00.0 VGA compatible controller: ATI Technologies Inc Radeon Mobility X700 (PCIE)

ii xserver-xorg-core 2:1.6.4-2 ii xserver-xorg-video-radeon 1:6.12.3-1
ii libgl1-mesa-dri 7.6-1

Xorg crashes with such a backtrace:

Program received signal SIGABRT, Aborted.
0x00007fdcfacafd25 in *__GI_raise (sig=<value optimized out>) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
64 ../nptl/sysdeps/unix/sysv/linux/raise.c: Toks failas ar aplankas neegzistuoja.
        in ../nptl/sysdeps/unix/sysv/linux/raise.c
(gdb) bt
#0 0x00007fdcfacafd25 in *__GI_raise (sig=<value optimized out>) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
#1 0x00007fdcfacb2de1 in *__GI_abort () at abort.c:88
#2 0x00007fdcfaca8f99 in *__GI___assert_fail (assertion=0x7fdcee613a28 "bo_legacy->is_pending <= bo->cref",
    file=0x7fdcee6139a0 "radeon_bo_legacy.c", line=207, function=0x7fdcee613ae0 "legacy_is_pending") at assert.c:78
#3 0x00007fdcee4619df in legacy_is_pending (bo=<value optimized out>) at radeon_bo_legacy.c:207
#4 0x00007fdcee461ba4 in legacy_track_pending (bom=0x87eee0, debug=0) at radeon_bo_legacy.c:251
#5 0x00007fdcee461f3a in bo_open (bom=0x87eee0, handle=<value optimized out>, size=<value optimized out>, alignment=<value optimized out>,
    domains=<value optimized out>, flags=<value optimized out>) at radeon_bo_legacy.c:448
#6 0x00007fdcee467f1e in _radeon_bo_open (rmesa=0x15f8370, size=32) at ../radeon/radeon_bo_drm.h:110
#7 radeonRefillCurrentDmaRegion (rmesa=0x15f8370, size=32) at radeon_dma.c:196
#8 0x00007fdcee468243 in radeonAllocDmaRegion (rmesa=0x15f8370, pbo=0x15fb050, poffset=0x15fb058, bytes=32,
    alignment=<value optimized out>) at radeon_dma.c:250
#9 0x00007fdcee44b0fb in r300AllocDmaRegions (ctx=0x1661360, arrays=0x1d973c8, prim=0x1d95a44, nr_prims=1, ib=0x0,
    index_bounds_valid=<value optimized out>, min_index=0, max_index=3) at r300_draw.c:529
#10 r300TryDrawPrims (ctx=0x1661360, arrays=0x1d973c8, prim=0x1d95a44, nr_prims=1, ib=0x0, index_bounds_valid=<value optimized out>,
    min_index=0, max_index=3) at r300_draw.c:653
#11 r300DrawPrims (ctx=0x1661360, arrays=0x1d973c8, prim=0x1d95a44, nr_prims=1, ib=0x0, index_bounds_valid=<value optimized out>,
    min_index=0, max_index=3) at r300_draw.c:706
#12 0x00007fdcee508610 in vbo_exec_vtx_flush (exec=0x1d95800, unmap=1 '\1') at vbo/vbo_exec_draw.c:378
#13 0x00007fdcee5046bd in vbo_exec_FlushVertices_internal (ctx=<value optimized out>, unmap=159 '\237') at vbo/vbo_exec_api.c:759
#14 0x00007fdcee504772 in vbo_exec_FlushVertices (ctx=0xb9f, flags=1) at vbo/vbo_exec_api.c:781
#15 0x00007fdcee4d755e in _mesa_set_scissor (ctx=0x1661360, x=310, y=403, width=2, height=1) at main/scissor.c:75
#16 0x00007fdcf9f83446 in __glXDisp_Render (cl=<value optimized out>, pc=0x21c1a8c "\24") at ../../glx/glxcmds.c:1823
#17 0x00007fdcf9f87572 in __glXDispatch (client=0x1460550) at ../../glx/glxext.c:568
#18 0x000000000044d454 in Dispatch () at ../../dix/dispatch.c:456
#19 0x000000000043337a in main (argc=8, arg...

Read more...

Revision history for this message
js (solard3ity-deactivatedaccount-deactivatedaccount) wrote :

Same issue here to.. turn effects on open Systemsettings or right-klick... wooooms back to KDM
Everything went fine and as stable like never before, before updating MESA..
I also tried the X-Udates PPA ( X-SWAT ) to see if it would get better, but nope!.. Why did we update Mesa, I dont understand???!!!

I am running a 32bit have this issue on a AMD 740G ( Radeon 2100 ) and another system with a ATI card ..

Revision history for this message
Tormod Volden (tormodvolden) wrote :

Jason, please read Bryce's comment above.

Revision history for this message
js (solard3ity-deactivatedaccount-deactivatedaccount) wrote :

Ohh.. yeah.. okay, ill try that out :-) thanks... ( I should have read everything first )

Revision history for this message
Bryce Harrington (bryce) wrote :

marcinc, thanks for testing. The white lines seem less severe than a crash, so I guess this patch is worth adding. Please file a separate bug report on the white line issue.

Revision history for this message
Tormod Volden (tormodvolden) wrote :

Bryce, note that the white lines are not caused by the patch. But without the patch the server crashes before he can see any. I am pretty sure the white lines issue is the one you fixed in the last upload.

Revision history for this message
Launchpad Janitor (janitor) wrote :

This bug was fixed in the package mesa - 7.6.0-1ubuntu3

---------------
mesa (7.6.0-1ubuntu3) karmic; urgency=low

  * Add 109_revert-dma-reuse.patch: Fix assertion failure leading to crash
    on kwin when compositing is enabled. This reverts commit 284a7af27.
    (LP: #446578)

 -- Bryce Harrington <email address hidden> Tue, 13 Oct 2009 00:52:32 -0700

Changed in mesa (Ubuntu):
status: Confirmed → Fix Released
Revision history for this message
In , agd5f (agd5f) wrote :

*** Bug 24508 has been marked as a duplicate of this bug. ***

Revision history for this message
marcinq (marcinq) wrote :

yes, I couldn't test that white line patch from bug 446425 because it still was effected by this bug...

Just to cofirm with latest mesa upgrade both bugs are gone for me and compositing works perfectly :)

Revision history for this message
js (solard3ity-deactivatedaccount-deactivatedaccount) wrote :

I can also confirm that the white line bug is gone, composite speed is good and smooth + till now, no crash! :-)
Awesome.... I Love it....
Thanks!..

Revision history for this message
In , Modestas Vainius (geromanas) wrote :

FYI, this bug is screwing many KDE 4 users using their desktop environment with OpenGL effects (enabled by default) on radeon cards. The bug triggers semi-random X session crashes and people lose data and otherwise get annoyed. Moreover, this is a regression from meta 7.5. The origin of the crashes is not very obvious without deep investigation which a typical user is unable to do. Please consider this problem of higher priority.

Revision history for this message
In , Michel Dänzer (michel-daenzer) wrote :

As Pauli seems to be MIA, any better ideas than just reverting the bisected commit?

Revision history for this message
In , Julien Cristau (jcristau) wrote :

*** Bug 24822 has been marked as a duplicate of this bug. ***

Revision history for this message
In , agd5f (agd5f) wrote :

Reverted: 13b5a624b1899c457279907d58046dfb3c95addc

Revision history for this message
In , Bugzi09-fdo-tormod (bugzi09-fdo-tormod) wrote :

I wonder if reverting that commit brought all the "radeon_dma.c:210: radeonRefillCurrentDmaRegion: Assertion `dma_bo->bo->cref == 1' failed" reported in https://launchpad.net/bugs/459961 though.

Changed in mesa (Debian):
status: Unknown → Confirmed
Changed in mesa:
status: Confirmed → Fix Released
Changed in mesa (Debian):
status: Confirmed → Fix Released
Changed in mesa:
importance: Unknown → High
Revision history for this message
OmniUni (omniuni) wrote :

Does anyone know where we are at with this bug? It seems to be the one that is affecting me, I am fully up-to-date using KUbuntu 10.10, and I can't enable KWin. I have good graphics acceleration otherwise. I have tried xorg-edgers and x-updates, and they don't help.

Changed in mesa:
importance: High → Unknown
Changed in mesa:
importance: Unknown → High
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.