Finally figured out why my desktop keeps rebooting.
Checked the gdm logs and found evidence of a segfault.
Created attachment 346260 [details]
backtrace in gdm log:
(EE) 0: /usr/bin/X (xorg_backtrace+0x4a) [0x81de95a]
(EE) 1: /usr/bin/X (0x8048000+0x19a6c6) [0x81e26c6]
(EE) 2: /lib/libpthread.so.0 (0xb76ea000+0xfc10) [0xb76f9c10]
(EE) 3: /usr/bin/X (mieqProcessDeviceEvent+0x10f) [0x81bebff]
(EE) 4: /usr/bin/X (mieqProcessInputEvents+0x106) [0x81bee56]
(EE) 5: /usr/bin/X (ProcessInputEvents+0x15) [0x80c01a5]
(EE) 6: /usr/bin/X (xf86Wakeup+0x4ad) [0x80c0c7d]
(EE) 7: /usr/bin/X (WakeupHandler+0x6b) [0x80836bb]
(EE) 8: /usr/bin/X (WaitForSomething+0x193) [0x81dbdb3]
(EE) 9: /usr/bin/X (0x8048000+0x36cff) [0x807ecff]
(EE) 10: /usr/bin/X (0x8048000+0x24e85) [0x806ce85]
(EE) 11: /lib/libc.so.6 (__libc_start_main+0xf4) [0xb744a854]
(EE) 12: /usr/bin/X (0x8048000+0x25269) [0x806d269]
(EE) Segmentation fault at address 0xa8
Why would an X crash cause a reboot (unless something more serious is going on)?
Please reopen this bug report when you have analysed the problem further and have attached more information.
Just the desktop, not the entire computer. As in gdm catches the termination and restarts the display with a fresh login.
I said "reboot" as a figure of speech, sorry for the confusion.
1) Please attach the relevant /var/log/Xorg.?.log to this bug report.
afraid it already rotated off of my system.
I'll reopen when it happens again.
Created attachment 351986 [details]
I'm having a similar problem, but it only seems to happen when using the touch screen. Here's my backtrace from Xorg.0.log.old:
[622368.754] (EE) Backtrace:
[622368.754] (EE) 0: /usr/bin/X (xorg_backtrace+0x34) [0x5960e4]
[622368.754] (EE) 1: /usr/bin/X (0x400000+0x199e99) [0x599e99]
[622368.754] (EE) 2: /lib64/libpthread.so.0 (0x7f68c33a5000+0x10bf0) [0x7f68c33b5bf0]
[622368.754] (EE) 3: /usr/bin/X (0x400000+0x5609b) [0x45609b]
[622368.755] (EE) 4: /usr/bin/X (0x400000+0x561e5) [0x4561e5]
[622368.755] (EE) 5: /usr/bin/X (0x400000+0x12156d) [0x52156d]
[622368.755] (EE) 6: /usr/bin/X (0x400000+0x121d3e) [0x521d3e]
[622368.755] (EE) 7: /usr/bin/X (0x400000+0x1222a0) [0x5222a0]
[622368.755] (EE) 8: /usr/bin/X (0x400000+0x1248f8) [0x5248f8]
[622368.755] (EE) 9: /usr/bin/X (0x400000+0x14828a) [0x54828a]
[622368.755] (EE) 10: /usr/bin/X (mieqProcessDeviceEvent+0x1b9) [0x577099]
[622368.755] (EE) 11: /usr/bin/X (0x400000+0x653ee) [0x4653ee]
[622368.755] (EE) 12: /usr/bin/X (0x400000+0x4b0cd) [0x44b0cd]
[622368.755] (EE) 13: /usr/bin/X (0x400000+0x4b48d) [0x44b48d]
[622368.755] (EE) 14: /usr/bin/X (ProcAllowEvents+0xab) [0x44b57b]
[622368.755] (EE) 15: /usr/bin/X (0x400000+0x3af1e) [0x43af1e]
[622368.755] (EE) 16: /usr/bin/X (0x400000+0x29964) [0x429964]
[622368.755] (EE) 17: /lib64/libc.so.6 (__libc_start_main+0xed) [0x7f68c205560d]
[622368.755] (EE) 18: /usr/bin/X (0x400000+0x29cad) [0x429cad]
[622368.755] (EE) Segmentation fault at address 0x88
Fatal server error:
[622368.755] (EE) Caught signal 11 (Segmentation fault). Server aborting
I recompiled xorg-server with FEATURES="nostrip" and -ggdb in CFLAGS and CXXFLAGS, but I'm still not getting more symbols in the backtrace, I'd appreciate some guidance.
Also, I was having these crashes with xorg-server-1.13.4, then I upgraded to 1.14.2 and nothing seemed to change. This backtrace is from 1.14.2.
If you can, ssh into the computer and attach gdb to the X server. The stack trace that is dumped to Xorg.0.log is not always very informative.
(In reply to Chí-Thanh Christopher Nguyễn from comment #10)
> If you can, ssh into the computer and attach gdb to the X server.
Thanks, I actually tried that, and this is what I got:
Program received signal SIGSEGV, Segmentation fault.
0x000000000045609b in xi2mask_isset_for_device (mask=0xa0,
dev=0x8173c0 <xi_all_devices>, event_type=21)
1027 /var/tmp/portage/x11-base/xorg-server-1.14.2/work/xorg-server-1.14.2/dix/inpututils.c: No such file or directory.
Program received signal SIGABRT, Aborted.
0x00007f7b9b21ac15 in raise () from /lib64/libc.so.6
Program terminated with signal SIGABRT, Aborted.
The program no longer exists.
It's possible that kdm restarted it, or maybe I need to add some options
At the gdb prompt following SIGSEGV, enter
so that it outputs the stack trace.
(In reply to Chí-Thanh Christopher Nguyễn from comment #12)
> At the gdb prompt following SIGSEGV, enter
> (gdb) bt
> so that it outputs the stack trace.
That would probably work if the program still existed. But, continuing from that output:
The program no longer exists.
At the first (gdb) prompt the program still exists, at the second it does not exist any more.
Does enter repeat the last command? That would be horrible, but would explain some things...
Created attachment 353636 [details]
Oh my, enter does repeat the last command, and that was the problem. Here's the gdb output including full backtrace.
So this appears to be in touch code. Possibly one of the patches mentioned here
might help fix this problem. In particular
> Xi: if we delivered a TouchEnd to a passive grab, end it
looks related. The patches can all be found at:
The touch fixes are now in xorg-server-22.214.171.1241
Please test and reopen of the issue still exists with that version.
(In reply to Chí-Thanh Christopher Nguyễn from comment #18)
> The touch fixes are now in xorg-server-126.96.36.1991
> Please test and reopen of the issue still exists with that version.
Thanks, I actually contacted the author and he confirmed those patches should fix the problem and advised me to get the latest code from master. I installed xorg-server-9999 and have been using it for about a week, no crashes so far. I'll also switch to 188.8.131.521 and test.
On the other hand, I think the OP is having a slightly different problem, but I guess that could have been fixed too.
Tested with xorg-server-184.108.40.2061 for a few days, it's rock solid so far.
Marking as fixed per comment 20.
Reporter: if you still see this problem with >=xorg-server-220.127.116.111, please reopen this bug.