Bug 887919 - xserver crashes at startup 1.13.0-15.fc18, downgrade to -11 fixes it
Summary: xserver crashes at startup 1.13.0-15.fc18, downgrade to -11 fixes it
Keywords:
Status: CLOSED WORKSFORME
Alias: None
Product: Fedora
Classification: Fedora
Component: xorg-x11-server
Version: 18
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: X/OpenGL Maintenance List
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2012-12-17 15:47 UTC by Cole Robinson
Modified: 2013-02-02 18:42 UTC (History)
2 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2013-02-02 18:42:27 UTC
Type: Bug
Embargoed:


Attachments (Terms of Use)
Xorg.0.log (43.77 KB, text/x-log)
2012-12-17 15:47 UTC, Cole Robinson
no flags Details
dmesg output (98.01 KB, text/plain)
2013-01-06 19:01 UTC, Cole Robinson
no flags Details

Description Cole Robinson 2012-12-17 15:47:46 UTC
Created attachment 664936 [details]
Xorg.0.log

After upgrading to xorg-x11-server-1.13.0-15.fc18 and rebooting, X fails to start after plymouth+systemd do their thing. Failure scenario sucks hard since either my keyboard isn't working or none of the VTs are available, so I had to reboot into rescue mode and downgrade. server-1.13.0-11.fc18 works fine.

Backtrace from the log:

[    17.125] (EE) 
[    17.125] (EE) Backtrace:
[    17.125] (EE) 0: /usr/bin/Xorg (xorg_backtrace+0x36) [0x46c346]
[    17.125] (EE) 1: /usr/bin/Xorg (0x400000+0x71c79) [0x471c79]
[    17.125] (EE) 2: /lib64/libpthread.so.0 (0x3ec1c00000+0xf000) [0x3ec1c0f000]
[    17.125] (EE) 3: /lib64/libdrm_radeon.so.1 (radeon_bo_set_tiling+0x0) [0x7f7ce7fb3750]
[    17.125] (EE) 4: /usr/lib64/xorg/modules/drivers/radeon_drv.so (0x7f7ce81c0000+0x43361) [0x7f7ce8203361]
[    17.125] (EE) 5: /usr/lib64/xorg/modules/drivers/radeon_drv.so (0x7f7ce81c0000+0x434e1) [0x7f7ce82034e1]
[    17.125] (EE) 6: /usr/bin/Xorg (AddGPUScreen+0x91) [0x4399c1]
[    17.126] (EE) 7: /usr/bin/Xorg (InitOutput+0x27c) [0x48d15c]
[    17.126] (EE) 8: /usr/bin/Xorg (0x400000+0x27ea6) [0x427ea6]
[    17.126] (EE) 9: /lib64/libc.so.6 (__libc_start_main+0xf5) [0x3ec1021a05]
[    17.126] (EE) 10: /usr/bin/Xorg (0x400000+0x283cd) [0x4283cd]
[    17.126] (EE) 
[    17.126] (EE) Segmentation fault at address 0x20
[    17.126] 
Fatal server error:
[    17.126] Caught signal 11 (Segmentation fault). Server aborting
[    17.126] 
[    17.126] (EE)

Comment 1 Cole Robinson 2012-12-19 01:02:08 UTC
Just noticed this in /var/log/messages:

Dec 18 19:52:11 colepc kernel: [  110.652078]  [<ffffffffa019c671>] radeon_gem_o
bject_create+0x61/0x100 [radeon]
Dec 18 19:52:11 colepc kernel: [  110.652092]  [<ffffffffa019c9fb>] radeon_gem_c
reate_ioctl+0x6b/0x120 [radeon]
Dec 18 19:52:11 colepc kernel: [  110.652105]  [<ffffffffa0035393>] drm_ioctl+0x
4d3/0x580 [drm]
Dec 18 19:52:11 colepc kernel: [  110.652123]  [<ffffffffa019c990>] ? radeon_gem
_pwrite_ioctl+0x30/0x30 [radeon]


So I maybe that's causing the server crash. And probably explains why machine appears to completely lock.

Comment 2 Dave Airlie 2012-12-19 04:21:38 UTC
can you boot to runlevel 3 (edit kernel command line in grub, add 3 to the end)

and attach the complete dmesg from that?

it sounds like something is screwing up on the secondary GPU you have.

Comment 3 Cole Robinson 2012-12-20 17:17:27 UTC
Will be away from the HW for a week, I'll retry after the holidays.

Comment 4 Cole Robinson 2013-01-06 19:01:33 UTC
Created attachment 673439 [details]
dmesg output

Didn't see any backtrace in dmesg this time. And with all latest packages I could at least ctrl-alt-f2.

Indeed this was related to my second video card: I originally plugged it in for possible virt passthrough testing, only to find that F18 couldn't give any useful output out of it. It's the only OS I've tried though. Once I popped out the card, things are working fine. Though I'm happy to plug it back in for testing.

For reference the card is this:

http://www.newegg.com/Product/Product.aspx?Item=N82E16814161376

Comment 5 Cole Robinson 2013-02-02 18:42:27 UTC
Just closing since this is no longer affecting me.


Note You need to log in before you can comment on or make changes to this bug.