[Nouveau] [Bug 93629] [NVE6] complete system freeze, PGRAPH engine fault on channel 2, SCHED_ERROR [ CTXSW_TIMEOUT ]

bugzilla-daemon at freedesktop.org bugzilla-daemon at freedesktop.org
Tue May 17 16:13:59 UTC 2016 

https://bugs.freedesktop.org/show_bug.cgi?id=93629

--- Comment #30 from Dāvis <davispuh at gmail.com> ---
I think I'm getting this same bug with Nvidia GTX 650 Ti on Arch Linux using
unpatched released kernel 4.6, xf86-video-nouveau 1.0.12 and mesa 11.2.2

Sometimes random hangs happen which locks up GPU.
Note that nouveau have never really worked for me on this PC with this GPU for
like more than a year since I installed Linux there. It always have been
getting random hangs like other bug #89912 So I've been using proprietary
Nvidia driver which works fine.



Anyway for this bug, I'm using GNOME Shell and it happened while playing
Minecraft (with java-8-openjdk)

Kernel log:

kernel: fb: switching to nouveaufb from EFI VGA
kernel: Console: switching to colour dummy device 80x25
kernel: nouveau 0000:01:00.0: NVIDIA GK106 (0e6060a1)
kernel: nouveau 0000:01:00.0: bios: version 80.06.21.00.37
kernel: nouveau 0000:01:00.0: fb: 1024 MiB GDDR5
kernel: [TTM] Zone  kernel: Available graphics memory: 12204692 kiB
kernel: [TTM] Zone   dma32: Available graphics memory: 2097152 kiB
kernel: [TTM] Initializing pool allocator
kernel: [TTM] Initializing DMA pool allocator
kernel: nouveau 0000:01:00.0: DRM: VRAM: 1024 MiB
kernel: nouveau 0000:01:00.0: DRM: GART: 1048576 MiB
kernel: nouveau 0000:01:00.0: DRM: TMDS table version 2.0
kernel: nouveau 0000:01:00.0: DRM: DCB version 4.0
kernel: nouveau 0000:01:00.0: DRM: DCB outp 00: 01000f02 00020030
kernel: nouveau 0000:01:00.0: DRM: DCB outp 01: 02000f00 00020030
kernel: nouveau 0000:01:00.0: DRM: DCB outp 02: 08011f82 0f420030
kernel: nouveau 0000:01:00.0: DRM: DCB outp 03: 02022f62 0f420010
kernel: nouveau 0000:01:00.0: DRM: DCB conn 00: 00001030
kernel: nouveau 0000:01:00.0: DRM: DCB conn 01: 00002131
kernel: nouveau 0000:01:00.0: DRM: DCB conn 02: 00010263
kernel: [drm] Supports vblank timestamp caching Rev 2 (21.10.2013).
kernel: [drm] Driver supports precise vblank timestamp query.
kernel: nouveau 0000:01:00.0: DRM: MM: using COPY for buffer copies
kernel: nouveau 0000:01:00.0: DRM: allocated 2560x1440 fb: 0x60000, bo
ffff880601a1c400
kernel: fbcon: nouveaufb (fb0) is primary device
kernel: Console: switching to colour frame buffer device 128x48
kernel: nouveau 0000:01:00.0: fb0: nouveaufb frame buffer device
kernel: [drm] Initialized nouveau 1.3.1 20120801 for 0000:01:00.0 on minor 0
...
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 00200000 [METHOD] ch 6 [003f7d0000
java[1817]] subc 0 mthd 000c data 00000000
kernel: nouveau 0000:01:00.0: gr: TRAP ch 6 [003f7d0000 java[1817]]
kernel: nouveau 0000:01:00.0: gr: DISPATCH 80000002 [CLASS_SUBCH_MISMATCH]
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 02000000 [SEMAPHORE] ch 6
[003f7d0000 java[1817]] subc 0 mthd 001c data 3f800000
kernel: nouveau 0000:01:00.0: gr: ILLEGAL_CLASS ch 6 [003f7d0000 java[1817]]
subc 0 class 0000 mthd 0100 data 00000000
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 00200000 [METHOD] ch 6 [003f7d0000
java[1817]] subc 0 mthd 0030 data 20030700
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 00200000 [METHOD] ch 6 [003f7d0000
java[1817]] subc 0 mthd 0034 data 00001014
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 00200000 [METHOD] ch 6 [003f7d0000
java[1817]] subc 0 mthd 0038 data 00000000
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 00200000 [METHOD] ch 6 [003f7d0000
java[1817]] subc 0 mthd 003c data 015c4d10
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 00200000 [METHOD] ch 6 [003f7d0000
java[1817]] subc 0 mthd 0040 data 200207c0
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 00200000 [METHOD] ch 6 [003f7d0000
java[1817]] subc 0 mthd 0044 data 00000000
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 00200000 [METHOD] ch 6 [003f7d0000
java[1817]] subc 0 mthd 0048 data 015d1fff
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 00400000 [METHODCRC] ch 6
[003f7d0000 java[1817]] subc 0 mthd 007c data a01108e3
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 00200000 [METHOD] ch 6 [003f7d0000
java[1817]] subc 0 mthd 004c data 80000704
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 00200000 [METHOD] ch 6 [003f7d0000
java[1817]] subc 0 mthd 00ec data 00000000
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 00200000 [METHOD] ch 6 [003f7d0000
java[1817]] subc 0 mthd 00f0 data 00000000
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 00200000 [METHOD] ch 6 [003f7d0000
java[1817]] subc 0 mthd 00f4 data 00000000
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 00200000 [METHOD] ch 6 [003f7d0000
java[1817]] subc 0 mthd 00f8 data 3f800000
kernel: nouveau 0000:01:00.0: fifo: PBDMA0: 00200000 [METHOD] ch 6 [003f7d0000
java[1817]] subc 0 mthd 00fc data 3f800000
... loads of more like these ...
kernel: nouveau 0000:01:00.0: gr: ILLEGAL_CLASS ch 6 [003f7d0000 java[1817]]
subc 0 class 0000 mthd 0100 data 00000000
kernel: nouveau 0000:01:00.0: gr: ILLEGAL_CLASS ch 6 [003f7d0000 java[1817]]
subc 0 class 0000 mthd 0104 data 00000000
kernel: nouveau 0000:01:00.0: gr: ILLEGAL_CLASS ch 6 [003f7d0000 java[1817]]
subc 0 class 0000 mthd 0108 data 00000000
kernel: nouveau 0000:01:00.0: gr: ILLEGAL_CLASS ch 6 [003f7d0000 java[1817]]
subc 0 class 0000 mthd 010c data 00000000
kernel: nouveau 0000:01:00.0: gr: ILLEGAL_CLASS ch 6 [003f7d0000 java[1817]]
subc 0 class 0000 mthd 0110 data 00000000
kernel: nouveau 0000:01:00.0: gr: ILLEGAL_CLASS ch 6 [003f7d0000 java[1817]]
subc 0 class 0000 mthd 0114 data 3f800000
kernel: nouveau 0000:01:00.0: gr: ILLEGAL_CLASS ch 6 [003f7d0000 java[1817]]
subc 0 class 0000 mthd 0118 data 3f800000
... again all log spammed with these like ILLEGAL_CLASS ...
[  345.614760] nouveau 0000:01:00.0: gr: ILLEGAL_CLASS ch 6 [003f7d0000
java[1817]] subc 0 class 0186 mthd 1b08 data 000014b6
[  345.614773] nouveau 0000:01:00.0: gr: ILLEGAL_CLASS ch 6 [003f7d0000
java[1817]] subc 0 class 0186 mthd 1b0c data 1000f010
[  428.465923] nouveau 0000:01:00.0: gr: ILLEGAL_CLASS ch 6 [003f7d0000
java[1817]] subc 0 class 0186 mthd 0100 data 00000000
[  428.465961] nouveau 0000:01:00.0: gr: ILLEGAL_CLASS ch 6 [003f7d0000
java[1817]] subc 0 class 0186 mthd 1b00 data 00000000
[  428.465989] nouveau 0000:01:00.0: gr: ILLEGAL_CLASS ch 6 [003f7d0000
java[1817]] subc 0 class 0186 mthd 1b04 data 00238000
[  428.466010] nouveau 0000:01:00.0: gr: ILLEGAL_CLASS ch 6 [003f7d0000
java[1817]] subc 0 class 0186 mthd 1b08 data 000014b7
[  428.466037] nouveau 0000:01:00.0: gr: ILLEGAL_CLASS ch 6 [003f7d0000
java[1817]] subc 0 class 0186 mthd 1b0c data 1000f010
[ 6833.360096] nouveau 0000:01:00.0: fifo: SCHED_ERROR 0a [CTXSW_TIMEOUT]
[ 6833.360109] nouveau 0000:01:00.0: fifo: gr engine fault on channel 6,
recovering...
[ 7195.188708] nouveau 0000:01:00.0: gnome-shell[1177]: failed to idle channel
5 [gnome-shell[1177]]
[ 7210.189442] nouveau 0000:01:00.0: gnome-shell[1177]: failed to idle channel
5 [gnome-shell[1177]]
[ 7210.189607] nouveau 0000:01:00.0: fifo: read fault at 0000013000 engine 07
[PBDMA0] client 07 [HOST_CPU] reason 02 [PTE] on channel 5 [003f8aa000
gnome-shell[1177]]
[ 7210.189772] nouveau 0000:01:00.0: fifo: fifo engine fault on channel 5,
recovering...
[ 7225.206843] nouveau 0000:01:00.0: java[2577]: failed to idle channel 6
[java[2577]]
[ 7240.207579] nouveau 0000:01:00.0: java[2577]: failed to idle channel 6
[java[2577]]

and here GPU lockup, but kernel itself didn't hung and works fine,
using SSH I was able to kill Xorg, unbind VT and unload nouveau and then reload
it again

# echo 0 > /sys/class/vtconsole/vtcon1/bind
# rmmod nouveau
# modprobe nouveau
# echo 1 > /sys/class/vtconsole/vtcon1/bind

then I started Xorg and it works, and in kernel log can see

[ 7354.650288] Console: switching to colour dummy device 80x25
[ 7398.714966] [TTM] Finalizing pool allocator
[ 7398.714972] [TTM] Finalizing DMA pool allocator
[ 7398.715081] [TTM] Zone  kernel: Used memory at exit: 0 kiB
[ 7398.715084] [TTM] Zone   dma32: Used memory at exit: 0 kiB
[ 7398.715664] [drm] Module unloaded
[ 7449.947953] MXM: GUID detected in BIOS
[ 7449.948021] nouveau 0000:01:00.0: NVIDIA GK106 (0e6060a1)
[ 7450.009294] nouveau 0000:01:00.0: bios: version 80.06.21.00.37
[ 7450.010087] nouveau 0000:01:00.0: fb: 1024 MiB GDDR5
[ 7450.064908] [TTM] Zone  kernel: Available graphics memory: 12204692 kiB
[ 7450.064912] [TTM] Zone   dma32: Available graphics memory: 2097152 kiB
[ 7450.064914] [TTM] Initializing pool allocator
[ 7450.064921] [TTM] Initializing DMA pool allocator
[ 7450.064939] nouveau 0000:01:00.0: DRM: VRAM: 1024 MiB
[ 7450.064942] nouveau 0000:01:00.0: DRM: GART: 1048576 MiB
[ 7450.064947] nouveau 0000:01:00.0: DRM: TMDS table version 2.0
[ 7450.064949] nouveau 0000:01:00.0: DRM: DCB version 4.0
[ 7450.064952] nouveau 0000:01:00.0: DRM: DCB outp 00: 01000f02 00020030
[ 7450.064955] nouveau 0000:01:00.0: DRM: DCB outp 01: 02000f00 00020030
[ 7450.064958] nouveau 0000:01:00.0: DRM: DCB outp 02: 08011f82 0f420030
[ 7450.064960] nouveau 0000:01:00.0: DRM: DCB outp 03: 02022f62 0f420010
[ 7450.064963] nouveau 0000:01:00.0: DRM: DCB conn 00: 00001030
[ 7450.064965] nouveau 0000:01:00.0: DRM: DCB conn 01: 00002131
[ 7450.064967] nouveau 0000:01:00.0: DRM: DCB conn 02: 00010263
[ 7450.066173] [drm] Supports vblank timestamp caching Rev 2 (21.10.2013).
[ 7450.066175] [drm] Driver supports precise vblank timestamp query.
[ 7450.122768] nouveau 0000:01:00.0: DRM: MM: using COPY for buffer copies
[ 7450.213563] nouveau 0000:01:00.0: DRM: allocated 2560x1440 fb: 0x60000, bo
ffff8804405c4800
[ 7450.213720] fbcon: nouveaufb (fb0) is primary device
[ 7450.569182] Console: switching to colour frame buffer device 128x48
[ 7450.569972] nouveau 0000:01:00.0: fb0: nouveaufb frame buffer device
[ 7450.587996] [drm] Initialized nouveau 1.3.1 20120801 for 0000:01:00.0 on
minor 0
[ 7479.483820] Console: switching to colour dummy device 80x25
[ 7483.096411] Console: switching to colour frame buffer device 128x48


but then after some while it hung again 

[10425.388235] nouveau 0000:01:00.0: fifo: SCHED_ERROR 0a [CTXSW_TIMEOUT]
[10425.388240] nouveau 0000:01:00.0: fifo: gr engine fault on channel 6,
recovering...

[11378.573422] nouveau 0000:01:00.0: java[4377]: failed to idle channel 6
[java[4377]]
[11393.574071] nouveau 0000:01:00.0: java[4377]: failed to idle channel 6
[java[4377]]

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/nouveau/attachments/20160517/65a3cb6a/attachment-0001.html>

More information about the Nouveau mailing list