Bug 1044304 - Segmentation fault in xorg-x11-drv-nouveau with kernel 3.13.0-0.rc3.git5.1.fc21.x86_64
Summary: Segmentation fault in xorg-x11-drv-nouveau with kernel 3.13.0-0.rc3.git5.1.fc...
Keywords:
Status: CLOSED RAWHIDE
Alias: None
Product: Fedora
Classification: Fedora
Component: xorg-x11-drv-nouveau
Version: rawhide
Hardware: x86_64
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: Ben Skeggs
QA Contact: Fedora Extras Quality Assurance
URL: https://retrace.fedoraproject.org/faf...
Whiteboard: abrt_hash:ca3498110412854c5b5188ff773...
Depends On:
Blocks: F21AlphaBlocker
TreeView+ depends on / blocked
 
Reported: 2013-12-18 04:29 UTC by Dan Mossor [danofsatx]
Modified: 2015-02-02 18:41 UTC (History)
13 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2014-07-10 04:50:17 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
File: dmesg (90.81 KB, text/plain)
2013-12-18 04:29 UTC, Dan Mossor [danofsatx]
no flags Details
backtrace from clean install/update to rawhide (2.94 KB, text/plain)
2013-12-19 10:19 UTC, Dan Mossor [danofsatx]
no flags Details
dmesg from clean install/update to rawhide (94.57 KB, text/plain)
2013-12-19 10:19 UTC, Dan Mossor [danofsatx]
no flags Details
proc_modules from clean install/update to rawhide (3.15 KB, text/plain)
2013-12-19 10:21 UTC, Dan Mossor [danofsatx]
no flags Details
vbios rom dump (58.00 KB, application/octet-stream)
2013-12-19 16:35 UTC, Dan Mossor [danofsatx]
no flags Details

Description Dan Mossor [danofsatx] 2013-12-18 04:29:44 UTC
Additional info:
reporter:       libreport-2.1.10
general protection fault: 0000 [#1] SMP 
Modules linked in: nouveau(+) crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel mxm_wmi i2c_algo_bit drm_kms_helper ttm drm i2c_core wmi video
CPU: 5 PID: 246 Comm: systemd-udevd Not tainted 3.13.0-0.rc3.git5.1.fc21.x86_64 #1
Hardware name: ASUSTeK COMPUTER INC. G55VW/G55VW, BIOS G55VW.217 11/05/2012
task: ffff8805f49eb3a0 ti: ffff8805f4122000 task.ti: ffff8805f4122000
RIP: 0010:[<ffffffffa00417a8>]  [<ffffffffa00417a8>] drm_debugfs_cleanup+0x18/0x60 [drm]
RSP: 0018:ffff8805f4123a98  EFLAGS: 00010202
RAX: 6b6b6b6b6b6b6b6b RBX: ffff8805f4a13138 RCX: ffffffffa00549e1
RDX: ffffffffa0049dce RSI: ffffffffa00549fd RDI: ffff8805f4a13138
RBP: ffff8805f4123aa0 R08: 000000006b6b6b6b R09: 0000000000000000
R10: 0000000000000001 R11: 0000000000000000 R12: ffff8805fb4ae7b0
R13: ffff8805fb4ae848 R14: ffffffffa01d7260 R15: 00000000ffffffea
FS:  00007f6896d84880(0000) GS:ffff880606400000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 000000300047ce70 CR3: 00000005f40fd000 CR4: 00000000001407e0
Stack:
 ffff8805f4a13138 ffff8805f4123ab8 ffffffffa002feee ffff8805f4a13138
 ffff8805f4123ad0 ffffffffa002ff4e ffff8805f3031148 ffff8805f4123ae8
 ffffffffa002ff79 ffff8805f3031148 ffff8805f4123b38 ffffffffa0032529
Call Trace:
 [<ffffffffa002feee>] drm_unplug_minor+0x1e/0x40 [drm]
 [<ffffffffa002ff4e>] drm_put_minor+0x3e/0x50 [drm]
 [<ffffffffa002ff79>] drm_dev_free+0x19/0x80 [drm]
 [<ffffffffa0032529>] drm_get_pci_dev+0xa9/0x140 [drm]
 [<ffffffff813baa2d>] ? __pci_set_master+0x2d/0xd0
 [<ffffffffa016ac3a>] nouveau_drm_probe+0x20a/0x290 [nouveau]
 [<ffffffff813bf455>] local_pci_probe+0x45/0xa0
 [<ffffffff813c06c5>] ? pci_match_device+0xc5/0xd0
 [<ffffffff813c0821>] pci_device_probe+0x111/0x120
 [<ffffffff814a17f5>] driver_probe_device+0x125/0x3a0
 [<ffffffff814a1b43>] __driver_attach+0x93/0xa0
 [<ffffffff814a1ab0>] ? __device_attach+0x40/0x40
 [<ffffffff8149f69b>] bus_for_each_dev+0x6b/0xb0
 [<ffffffff814a117e>] driver_attach+0x1e/0x20
 [<ffffffff814a0d58>] bus_add_driver+0x188/0x260
 [<ffffffffa01fc000>] ? 0xffffffffa01fbfff
 [<ffffffff814a21b4>] driver_register+0x64/0xf0
 [<ffffffffa01fc000>] ? 0xffffffffa01fbfff
 [<ffffffff813bee00>] __pci_register_driver+0x60/0x70
 [<ffffffffa00326da>] drm_pci_init+0x11a/0x130 [drm]
 [<ffffffffa01fc000>] ? 0xffffffffa01fbfff
 [<ffffffffa01fc043>] nouveau_drm_init+0x43/0x1000 [nouveau]
 [<ffffffff8100214a>] do_one_initcall+0xfa/0x1b0
 [<ffffffff8105ced3>] ? set_memory_nx+0x43/0x50
 [<ffffffff8110682b>] load_module+0x1eab/0x26d0
 [<ffffffff81101de0>] ? store_uevent+0x40/0x40
 [<ffffffff811071e6>] SyS_finit_module+0x86/0xb0
 [<ffffffff81768269>] system_call_fastpath+0x16/0x1b
Code: ff ff ff e8 4b e5 fe ff e9 76 ff ff ff e8 e1 2b 03 e1 90 0f 1f 44 00 00 55 48 89 e5 53 48 83 7f 20 00 48 89 fb 48 8b 47 18 74 3a <48> 8b 80 90 04 00 00 48 8b 80 d8 00 00 00 48 85 c0 74 02 ff d0 
RIP  [<ffffffffa00417a8>] drm_debugfs_cleanup+0x18/0x60 [drm]
 RSP <ffff8805f4123a98>

Comment 1 Dan Mossor [danofsatx] 2013-12-18 04:29:49 UTC
Created attachment 838106 [details]
File: dmesg

Comment 2 Dan Mossor [danofsatx] 2013-12-18 04:37:13 UTC
Kernel 3.13.0 won't boot with nouveau driver. Last message entered on display is an error about fb error, nouveau conflicts with EFI VGA, removing generic driver. All display updating ceases at that point. I was able to enter my password to unlock the LUKS partition, and according to the disk activity light it was taken and boot continued. I was not in a situation where I could try to connect to the system remotely, but I am fairly certain it booted with no video.

This happens consistently with 3.13 - with 3.12.05-301, I get the same "font" on boot until that same fb conflict error, when it then switches to nouveau and displays the correct sized font for a 1920x1080 display - grub is not loading the video driver, which is another issue in and of itself.

Comment 3 Dan Mossor [danofsatx] 2013-12-19 10:16:41 UTC
Confirmed, reproducible. I performed a fresh clean install of F20 with KDE desktop after this bug was first reported. I attempted another update to rawhide with never installing the Nvidia drivers, so the system has had native nouveau driver on it since it was installed. Current rawhide version is xorg-x11-drv-nouveau-1.0.9-6.fc21.x86_64.

Booting the 3.12.05 kernel allows the system to start up normally.

The hardware involved is an ASUS G55VW laptop with a discrete Nvidia GTX 660m+, not Optimus enabled.

Comment 4 Dan Mossor [danofsatx] 2013-12-19 10:19:01 UTC
Created attachment 838915 [details]
backtrace from clean install/update to rawhide

Comment 5 Dan Mossor [danofsatx] 2013-12-19 10:19:52 UTC
Created attachment 838916 [details]
dmesg from clean install/update to rawhide

Comment 6 Dan Mossor [danofsatx] 2013-12-19 10:21:07 UTC
Created attachment 838917 [details]
proc_modules from clean install/update to rawhide

Comment 7 Dan Mossor [danofsatx] 2013-12-19 16:35:27 UTC
Created attachment 839065 [details]
vbios rom dump

On the advice of imirkin in the #nouveau channel, I am providing a dump of my vbios to help solve the unkown opcode error.

Comment 8 Dan Mossor [danofsatx] 2013-12-25 06:46:07 UTC
Double checked with many Kernels - the 3.13 branch will not boot with a graphical display with this bug.

Kernels attempted:
kernel-3.13.0-0.rc4.git1.1.fc21.x86_64
kernel-3.13.0-0.rc4.git5.1.fc21.x86_64 
kernel-3.13.0-0.rc5.git0.1.fc21.x86_64

I am still booting with my last remaining Heisenbug kernel, 3.12.5-301.fc20.x86_64.

Proposing as an F21 blocker, as F21 is unusable at the moment with it's native kernel and nouveau on GeForce GTX hardware.

Comment 9 Adam Williamson 2014-03-26 18:33:50 UTC
Is this still a problem with 3.14?

Comment 10 Dan Mossor [danofsatx] 2014-04-28 15:07:50 UTC
Nightly build from 24 APR was able to boot into a GUI. The only live image I tried was the KDE one which had other major issues, but it did boot into the desktop.

Comment 11 Tim Flink 2014-07-09 18:43:36 UTC
Discussed at the 2014-07-09 Fedora 21 alpha blocker review meeting. This bug seems to have gone stale and it's not clear whether or not it's still valid for recent kernels.

Please retest to see if this is still an issue.

Comment 12 Adam Williamson 2014-07-10 04:50:17 UTC
per the feedback on 2014-04-28 the bug was resolved at that point, so I'm gonna go ahead and close this for now (of course, re-open if the segfault has recurred in current kernels, Dan.)

Comment 13 Dan Mossor [danofsatx] 2015-02-02 18:41:01 UTC
Bug was resolved long, long ago. Clearing the needinfo flag.


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