[i945] X crash in ExaCheckPutImage() playing Klondike Solitaire

These attachments are from /var/log/gdm, from 17:14:45 to 17:15:16, immediately following the crash, before the reset.

Bad news: x server crashed again. This is the first time since 15-11-2008. Again, I was playing Solitaire Aisleriot. Firefox and Thunderbird were idle in the background.
Good news: this time "low graphics mode" did function ("your screen, graphics card, and input devices settings could not be detected"), and a troubleshooting program too. I choose to save the configuration and errors messages. Tar file attached.

Once more !!! I had only had the time to post and try restart Solitaire. OK, I'll refrain from playing Solitaire for a while.
Tar file attached.
Half an hour ago, I did Ctrl+Alt+F1, login, sudo shutdown - r now. This time, after saving the tar file, I accepted to run Ubuntu in low graphics and the display restarted in the normal (1280 per 1024) configuration.

[This is an automated message]

Hi dominique-meeus-d,

Please attach the output of `lspci -vvnn` too.

[This is an automatic notice.]

We'd like to forward your bug upstream, however upstream requires
that you first test it against their newer driver code.

To save you the effort of building the driver from source, we've built
packages for the driver and its new dependencies.

So you have a couple options:

 1. Download and test .debs for intrepid, from:
     https://edge.launchpad.net/~intel-gfx-testing/+archive

 -or-

 2. Download and test the Jaunty alpha-2 (or newer) Live CD,
     (which includes a beta of the new xserver 1.6 as well).
     See http://cdimage.ubuntu.com/releases/9.04/ for ISOs

Thanks ahead of time! You can simply reply to this email to report your
findings.

P.S., if you wish to forward your bug upstream yourself, please follow
these directions to do so:
  http://intellinuxgraphics.org/how_to_report_bug.html

Bryce Harrington a écrit ce qui suit, le 20/12/08 02:37 :
> ...Please attach the output of `lspci -vvnn` too...
*~$ lspci -vvnn
00:00.0 Host bridge [0600]: Intel Corporation 82945G/GZ/P/PL Memory
Controller Hub [8086:2770] (rev 02)
    Subsystem: ASUSTeK Computer Inc. Device [1043:817a]
    Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B- DisINTx-
    Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort-
<TAbort- <MAbort+ >SERR- <PERR- INTx-
    Latency: 0
    Capabilities: <access denied>
    Kernel driver in use: agpgart-intel
    Kernel modules: intel-agp

00:02.0 VGA compatible controller [0300]: Intel Corporation 82945G/GZ
Integrated Graphics Controller [8086:2772] (rev 02)
    Subsystem: ASUSTeK Computer Inc. Device [1043:817a]
    Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B- DisINTx-
    Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort-
<TAbort- <MAbort- >SERR- <PERR- INTx-
    Latency: 0
    Interrupt: pin A routed to IRQ 16
    Region 0: Memory at dfe00000 (32-bit, non-prefetchable) [size=512K]
    Region 1: I/O ports at 8800 [size=8]
    Region 2: Memory at e0000000 (32-bit, prefetchable) [size=256M]
    Region 3: Memory at dfe80000 (32-bit, non-prefetchable) [size=256K]
    Capabilities: <access denied>
    Kernel modules: intelfb

00:1b.0 Audio device [0403]: Intel Corporation 82801G (ICH7 Family) High
Definition Audio Controller [8086:27d8] (rev 01)
    Subsystem: ASUSTeK Computer Inc. Device [1043:8290]
    Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR- FastB2B- DisINTx-
    Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort-
<TAbort- <MAbort- >SERR- <PERR- INTx-
    Latency: 0, Cache Line Size: 16 bytes
    Interrupt: pin A routed to IRQ 19
    Region 0: Memory at dfef8000 (64-bit, non-prefetchable) [size=16K]
    Capabilities: <access denied>
    Kernel driver in use: HDA Intel
    Kernel modules: snd-hda-intel

00:1c.0 PCI bridge [0604]: Intel Corporation 82801G (ICH7 Family) PCI
Express Port 1 [8086:27d0] (rev 01)
    Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR+ FastB2B- DisINTx+
    Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort-
<TAbort- <MAbort- >SERR- <PERR- INTx-
    Latency: 0, Cache Line Size: 16 bytes
    Bus: primary=00, secondary=03, subordinate=03, sec-latency=0
    I/O behind bridge: 0000e000-0000efff
    Secondary status: 66MHz- FastB2B- ParErr- DEVSEL=fast >TAbort-
<TAbort- <MAbort- <SERR- <PERR-
    BridgeCtl: Parity- SERR+ NoISA+ VGA- MAbort- >Reset- FastB2B-
        PriDiscTmr- SecDiscTmr- DiscTmrStat- DiscTmrSERREn-
    Capabilities: <access denied>
    Kernel driver in use: pcieport-driver
    Kernel modules: shpchp

00:1c.1 PCI bridge [0604]: Intel Corporation 82801G (ICH7 Family) PCI
Express Port 2 [8086:27d2] (rev 01)
    Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr-
Stepping- SERR+ FastB2B- DisINTx+
    Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort-
<TAbort- <MAbort- >SERR- <PERR- INTx-
    Latency: 0, Cache Line Size: 16 bytes
    Bus: ...

Bryce Harrington a écrit ce qui suit, le 20/12/08 03:37 :
> ...
>
> 1. Download and test .debs for intrepid, from:
> https://edge.launchpad.net/~intel-gfx-testing/+archive
>
I did, using Synaptic, and this are the "installed" lines resulting in
/var/log/dpkg.log
2008-12-20 18:00:48 status installed man-db 2.5.2-2
2008-12-20 18:00:49 status installed libdrm2 2.4.1-0ubuntu7~intrepid
2008-12-20 18:00:49 status installed libdrm-intel1 2.4.1-0ubuntu7~intrepid
2008-12-20 18:00:49 status installed xserver-xorg-video-intel
2:2.5.1-1ubuntu5~intrepid
2008-12-20 18:00:49 status installed libc6 2.8~20080505-0ubuntu7
> Thanks ahead of time! You can simply reply to this email to report your
> findings.
>
I do not crash the X server on command. I did about three times in a
week a month ago. I didn't for a month. Then twice in half an hour. So I
can not tell you if these packages are doing any good... until the next
crash. I can't think of other "findings" I could report on. (But you or
upstream could suggest positive tests.) If you don't hear from me on
this bug for a long while, then it will definitely be a positive answer.
> P.S., if you wish to forward your bug upstream yourself, please follow
> these directions to do so: http://intellinuxgraphics.org/how_to_report_bug.html
>
I see that you intended to do so. I prefer to leave it to you. I am a
conscious user. I am not sure I would call myself a hacker. I suppose
that you know better what and how to report.
--

Amitiés, Dominique,
<email address hidden>, +32 (473) 61 31 75, http://www.meeus-d.be/

Okay, thanks. Let's give it some time for testing to be certain. If we don't hear from you in 2 months we'll assume it to be fixed.

Didn't last two month. Again playing Klondike. No other user program running. Everything froze while dragging a card. The screen didn't disappear or go into low graphics, just froze. Only the mouse pointer moved (a hand, because frozen while dragging). Nothing else would happen, no way to stop the game, no way to start another program, no opening a menu, no logging out, nothing worked, no keyboard shortcut not even Ctrl+Alt+F1. I had to do a reset to look at /var/log and send this post.

Again. After the restart, Nautilus, Thunderbird, Firefox, Gedit were left open. I did a try at Klondike. Again a general freeze without loosing the image of the screen and with pointer (hand) still moving, but nothing else working. Had to reset.

Thanks for testing and confirming it still is an issue. Looks like it's something that your Solitaire program is triggering; I'll update the title. This should go upstream; setting to Triaged for now.

Btw, if you can boil it down to an exact set of steps to reproduce the freeze, that could help a lot.

Hi Dominique, thanks for collecting all this information. I've forwarded the bug upstream to https://bugs.freedesktop.org/show_bug.cgi?id=19433 . Please subscribe to that bug report in case upstream needs further information or wishes you to test something.

Also, if you have additional tips for reproducing it, please let us know. I played around a bit with klondike solitaire but it didn't cause a crash for me.

Hi,

I have exactly the problem described here, also only with Solitaire - seems to freeze routinely after dragging around 5 cards, and the mouse cursor moves but nothing else happens.

Tony

Upstream says they believe this issue is fixed with a newer version of the driver. Can someone please test and verify this?

Just in case... the game software is AisleRiot, now 2.24.1.1. I play only the Klondike game and I do very often to release the stress after some big work, certainly once a day at least. But it doesn't make X freeze or crash very often: see dates above, sometimes twice the same day, sometimes not for a month. This is the most difficult thing to test.
I did install a new driver on 2008-12-20 and it DID crash twice on 2008-12-25. It didn't since.
The driver is still xserver-xorg-video-intel 2:2.5.1-1ubuntu5~intrepid and I still have http://ppa.launchpad.net/intel-gfx-testing/ubuntu intrepid main in my sources list. Is there a newer driver? Is some other package implied?

Playing Klondike in Solitaire Aisleriot. Froze while dragging a card. That card had two images (was shown in two slightly different positions of the movement).

Xorg didn't crash but everything froze but the mouse pointer. No menu, no clicking, no Ctrl+Alt+F1. Reset.

Nothing in /var/crash. I attach Xorg.0.log.old as the only file containing something written at the time the Gnome session froze.

Same information at the same time in /va/log/gdm/:0.log.1

Dominique, please provide your feedback on the upstream bug at http://bugs.freedesktop.org/show_bug.cgi?id=19433 else they will close the bug report and won't investigate it.

Dominique, upstream would like you to test against the current (2.6.x) driver. The best way to test this is to download an ISO image of jaunty, which you can do at http://cdimages.ubuntu.com/releases/jaunty/, burn to CD, boot, and reproduce the issue.

Bryce Harrington a écrit ce qui suit, le 05/02/09 06:19 :
> Dominique, upstream would like you to test against the current (2.6.x)
> driver. The best way to test this is to download an ISO image of
> jaunty, which you can do at http://cdimages.ubuntu.com/releases/jaunty/,
> burn to CD, boot, and reproduce the issue.
Bryce, you seem to overlook the fact that this bug is absolutely
unpredictable and I can not figure out how to "reproduce the issue". I
play Klondike about everyday day and it freezes only about once a month
(since a few month, this seems to be one more 8.10 regression, like
Bluetooth). I have no idea of something special I would be doing when it
freezes. The only thing I can say is that as far as I remember it
happens only while dragging a card, but I would not call dragging a card
"special" in a card game. To give you an order of magnitude: dragging a
card causes freezing once in nearly ten thousand dragged cards. I
strongly believe that things do not happen without a cause, but here the
cause is hard to figure. It must be a queer combination of particular
circumstances, dragging a card at a particular state of the whole system.

This being so, boot a Jaunty CD and "reproduce the issue" would mean
working all the time on a live Jaunty CD for weeks or months. (Mind the
fact that with an unpredictable bug of a monthly mean frequency, you
have to wait many months to conclude that the bug is corrected.) It
would be more realistic to install Jaunty for good.

Well, after all, this is not impossible. This is a home computer, not a
professional computer. I do a lot of things on this computer that I do
not want to loose but I rsync -u everything new every day on a USB stick
(to have it at work in my pocket just in case I need something there
too). My mail is in IMAP thus never lost. OK, when I think of it I could
consider installing Jaunty if you answer this question: how buggy is
Jaunty in the present state? is it really a pain working with it? is it
completely unusable at times? how would you compare it with Debian
testing (that some people do use)? is it just a matter of a minor bug
moderately annoying once in a while? I do not ask you to decide for me.
You give me your evaluation of the present buggyness or stability of
Jaunty. I understand that it is my responsibility to appreciate your
answer and to decide, or not, to test Jaunty (glad if this can help) and
I promise not to complain if I do.

In the post above, I meant Debian unstable (some people do use unstable).

Hi Dominique,

As you see on the upstream bug, they believe this issue is fixed in the latest releases of several X components. These versions are all present in Jaunty.

Personally, I would like to see better proof from upstream that the issue is indeed resolved, but it would not surprise me to know that the issue is resolved with the newer components.

I can understand your hesitance at running an unreleased version of Ubuntu, but unfortunately without knowing from upstream specifically what caused and fixed the crash, we have little other way to prove whether the bug is gone or not.

I myself have been running jaunty on both my laptop and desktop since alpha-1, however I'm a developer and can deal with various problems if I run into them. So far in Jaunty I've run into a number of regressions, but they were mostly on the annoyance scale; most I reported as bugs and got fixes within a matter of weeks, although there is still a flash-breakage-on-amd64 I'm waiting to see fixed, and some other misc. issues. On my desktop (using -ati), I have an issue where the mouse cursor is sometimes slightly corrupted, but it goes away when I move it. I would classify all of these merely as minor annoyances. I've not run into anything that prevented me from using the system: network has been solid throughout jaunty, X has been okay (many complain about performance regressions on -intel though, but it's been acceptable for my modest needs on my laptop), I've had no kernel issues (although my laptop's speaker re-activated and produces an annoying beep when I boot), and all the Gnome software has been okay with a few hiccups (the worst was that opening url's from Gnome programs to firefox broke, but that's been fixed now). By now I think Jaunty has gotten pretty stable and the worst issues seem ironed out. Also I think the number of major changes from here until release will be pretty limited, and developers will be focusing more on just bug fixing here on out.

Aside from the problems, there is one improvement which may have bearing on this issue. With Jaunty I've improved the automatic apport-based X crash handler, such that it is more often able to capture and post to Launchpad when a crash occurs. So, if you do run Jaunty and happen to reproduce the error, it now may be able to automatically capture the full backtrace for you, no need to run gdb manually. (At least, I hope -- it's been working out okay for many).

Running Jaunty since this morning (in Belgium, winter, UTC+1). Let's see for Klondike.

I have the same problem on Ubuntu 8.10. But I do have it everytime I'm on battery. here's a copy of a message I posted somewhere else:

Critical bug with AisleRiot

It always happens on battery, never had the problem otherwise.
I start a game of Aislriot, fullscreen or not, I start playing, and at some point when I move a card the screen freezes and goes black. A cross appears and disappears in the middle of the screen, indicating that X is trying to restart or something. Rebooting Ubuntu normaly doesn't solve it, I just end on terminal tty1. Ctrl Alt F7 leads me to a blinking white cursor, terminal style. I have to reboot the computer and start Ubuntu in safe mode to repair X with the xfix function. It sometimes is not enough and some data on my disk are severly damaged. It happened three times in a row now and I'm pretty sure it will happen again if I start a new game. The xorg.conf.datetime saved does not seem to differ from the restored one.

It only happens with AisleRiot, I can watch films and play Wine games without problem.

I'm running Ubuntu 8.10 with the Netbook Remix interface on a Asus Eee901, which as only SSD drives. The kernel is Array.org lean.

Seems solved for me with the patch I had two days ago. Before it I had crashes in every games, and for now i have been able to complete two games without problem. It needs further testing, but for now it seems cleared! Thank you.

Thanks for letting us know the issue is resolved.