Description of problem: I'm running Fedora 12 Beta (via a USB stick) on an IBM Thinkpad X41t (has an Intel Corporation Mobile 915GM/GMS/910GML Express Graphics Controller (rev 03)). This behaviour did not occur in Fedora 12 Alpha. Applications/System Tools/System Monitor, System/About this Computer, and resuming from Suspend all cause X to crash with a similar backtrace: 0: /usr/bin/Xorg (xorg_backtrace+0x3c) [0x80e57bc] 1: /usr/bin/Xorg (0x8048000+0x5eba6) [0x80a6ba6] 2: (vdso) (__kernel_rt_sigreturn+0x0) [0x13640c] 3: /usr/lib/libpixman-1.so.0 (0x399000+0x38b0) [0x39c8b0] 4: /usr/lib/libpixman-1.so.0 (0x399000+0x3f705) [0x3d8705] 5: /usr/lib/libpixman-1.so.0 (0x399000+0x1477a) [0x3ad77a] 6: /usr/lib/libpixman-1.so.0 (0x399000+0x39d50) [0x3d2d50] 7: /usr/lib/libpixman-1.so.0 (0x399000+0x41a7a) [0x3daa7a] 8: /usr/lib/libpixman-1.so.0 (0x399000+0x42c9b) [0x3dbc9b] 9: /usr/lib/libpixman-1.so.0 (0x399000+0x399f4) [0x3d29f4] 10: /usr/lib/libpixman-1.so.0 (0x399000+0x152f4) [0x3ae2f4] 11: /usr/lib/libpixman-1.so.0 (0x399000+0x3b9af) [0x3d49af] 12: /usr/lib/libpixman-1.so.0 (0x399000+0x152f4) [0x3ae2f4] 13: /usr/lib/libpixman-1.so.0 (0x399000+0x46df4) [0x3dfdf4] 14: /usr/lib/libpixman-1.so.0 (0x399000+0x152f4) [0x3ae2f4] 15: /usr/lib/libpixman-1.so.0 (0x399000+0x4d5ca) [0x3e65ca] 16: /usr/lib/libpixman-1.so.0 (0x399000+0x152f4) [0x3ae2f4] 17: /usr/lib/libpixman-1.so.0 (pixman_image_composite+0x177) [0x3d3847] 18: /usr/lib/xorg/modules/libfb.so (fbComposite+0x170) [0x357d70] 19: /usr/lib/xorg/modules/drivers/intel_drv.so (0x754000+0x5b27c) [0x7af27c] 20: /usr/lib/xorg/modules/drivers/intel_drv.so (0x754000+0x59855) [0x7ad855] 21: /usr/bin/Xorg (0x8048000+0xd3977) [0x811b977] 22: /usr/bin/Xorg (CompositePicture+0x290) [0x810ee70] 23: /usr/bin/Xorg (0x8048000+0x1686ed) [0x81b06ed] 24: /usr/bin/Xorg (0x8048000+0x168a09) [0x81b0a09] 25: /usr/bin/Xorg (0x8048000+0xa16d8) [0x80e96d8] 26: /usr/bin/Xorg (0x8048000+0xa17d1) [0x80e97d1] 27: /usr/bin/Xorg (0x8048000+0x4b6e3) [0x80936e3] 28: /usr/bin/Xorg (MapWindow+0x15e) [0x809796e] 29: /usr/bin/Xorg (0x8048000+0x25829) [0x806d829] 30: /usr/bin/Xorg (0x8048000+0x26187) [0x806e187] 31: /usr/bin/Xorg (0x8048000+0x1a875) [0x8062875] 32: /lib/libc.so.6 (__libc_start_main+0xe6) [0x5edb36] 33: /usr/bin/Xorg (0x8048000+0x1a461) [0x8062461] Segmentation fault at address 0x9602ca0 Fatal server error: Caught signal 11 (Segmentation fault). Server aborting Version-Release number of selected component (if applicable): Name : xorg-x11-drv-intel Relocations: (not relocatable) Version : 2.9.0 Vendor: Fedora Project Release : 2.fc12 Build Date: Fri 02 Oct 2009 04:04:21 PM NZDT Install Date: Thu 15 Oct 2009 04:40:43 PM NZDT Build Host: x86-3.fedora.phx.redhat.com How reproducible: Always Steps to Reproduce: 1. Click on System Tools 2. 3. Actual results: X dies, I see a black terminal briefly, and then GDM comes back up. Expected results: System Tools should come up. Additional info: I'm running Fedora 12 Beta from a Live USB stick. I'm not sure if it's relevant, but I also end up seeing this in dmesg Oct 25 09:13:17 localhost kernel: render error detected, EIR: 0x00000010 Oct 25 09:13:17 localhost kernel: page table error Oct 25 09:13:17 localhost kernel: PGTBL_ER: 0x00000010 Oct 25 09:13:17 localhost kernel: [drm:i915_handle_error] *ERROR* EIR stuck: 0x00000010, masking Oct 25 09:13:17 localhost kernel: render error detected, EIR: 0x00000010 Oct 25 09:13:17 localhost kernel: page table error Oct 25 09:13:17 localhost kernel: PGTBL_ER: 0x00000010
I'll note that on another laptop, that opening System Tools (which I now realise is what About this Computer opens) does not crash X, but this other computer uses an AMD/ATI chipset.
I'll note that it doesn't do it with Xephyr inside the same session.
Created attachment 366076 [details] Magic stacktrace Here is a stacktrace that I obtained by doing this via SSH: X :2 gdb --pid=<the PID of the above X session> DISPLAY=:2 gnome-system-monitor I couldn't do it locally because running gdb on X locally locks up my system (for some reason I don't understand). If I could produce a more useful one with fewer values optimised out, I'll gladly do it.
I take it that the two uxa_ functions are the ones earlier reported in intel_drv.so in the less-useful stacktrace? I'll try with exa or xaa later, I guess.
Using the "vesa" driver, I do not run into the same problem. System Monitor appears as expected, no crash. I tried to set Option "AccelMethod" "exa" and Option "AccelMethod" "xaa" and Option "NoAccel" "True" but from what I could tell from the Xorg.0.log, UXA was used anyway.
Created attachment 366094 [details] Xorg.0.log including crash with intel driver This is the Xorg.0.log.old that's left behind after X crashes. Intel and presumably UXA acceleration.
I'll note that trying xorg-x11-drv-intel-2.9.0-3.fc12 xorg-x11-drv-intel-2.8.0-16.20090909.fc12 from koji suffered the same issue. (The beta comes with xorg-x11-drv-intel-2.9.0-2.fc12)
*** This bug has been marked as a duplicate of bug 530803 ***