Bug 69768

Created attachment 86469 [details]
dmesg kernel 3.11.1

Created attachment 86470 [details]
message 3.11.1

Created attachment 86471 [details]
suspend to ram fail

Created attachment 86472 [details]
message crash

Created attachment 86473 [details]
message 3.11.1

i use mesa 9.2 and suse 12.3

As mentioned in IRC, according to your logs your 3.10.9 suspend issues are not caused by nouveau

Whereas for the 3.11 one it seems that we've introduced a regression

I would recommend a git bisect between 3.10 and 3.11 as there are quite a few commits related to your card. I would check dmesg on each step to establish which commit introduced the following messages (and most likely the whole s2r issue)

nouveau  [     DRM] MM: using COPY0 for buffer copies
nouveau E[    PBUS][0000:01:00.0] MMIO write of 0x00000401 FAULT at 0x002010 [ IBUS TIMEOUT ]
nouveau E[    PBUS][0000:01:00.0] MMIO write of 0x01000000 FAULT at 0x03a54c [ IBUS TIMEOUT ]

There are others having trouble with NVC1 on 3.11+, see bug 69203.

ok i used bisect...

git bisect log
# bad: [6e4664525b1db28f8c4e1130957f70a94c19213e] Linux 3.11
# good: [8bb495e3f02401ee6f76d1b1d77f3ac9f079e376] Linux 3.10
git bisect start 'v3.11' 'v3.10' '--' 'drivers/gpu/drm/nouveau'
# bad: [b054aadfb0030b9717bb22f4283bfe5aec13440b] drm/nvf0/gr: magic sequence that makes PGRAPH come out of hiding
git bisect bad b054aadfb0030b9717bb22f4283bfe5aec13440b
# good: [a8004a9edd8c6ee86c3e263c9082ddf64797a667] drm/nvc0-/gr: bump maximum gpc/tpc limits
git bisect good a8004a9edd8c6ee86c3e263c9082ddf64797a667
# good: [0d4a1450c95801c21ba4db109303fbad62378b91] drm/nouveau/vdec: fork vp3 implementations from vp2
git bisect good 0d4a1450c95801c21ba4db109303fbad62378b91
# bad: [c03ff9e8fa5fc0186158b99a89f613325ff352cf] drm/nvc0-/gr: pull out a group of separately context-switched gpc regs
git bisect bad c03ff9e8fa5fc0186158b99a89f613325ff352cf
# good: [05f9a5bc58381f58095d8789e1c2d4e18758c2bc] drm/nouveau/bsp/nv84: initial vp2 engine implementation
git bisect good 05f9a5bc58381f58095d8789e1c2d4e18758c2bc
# good: [791dc143ed2c441f5202d8721609d94dce9fcf88] drm/nvd0-/disp: handle case where display engine is missing/disabled
git bisect good 791dc143ed2c441f5202d8721609d94dce9fcf88


don't know if that help?

(In reply to comment #9)
> ok i used bisect...
> 
> git bisect log
> # bad: [6e4664525b1db28f8c4e1130957f70a94c19213e] Linux 3.11
> # good: [8bb495e3f02401ee6f76d1b1d77f3ac9f079e376] Linux 3.10
> git bisect start 'v3.11' 'v3.10' '--' 'drivers/gpu/drm/nouveau'
> # bad: [b054aadfb0030b9717bb22f4283bfe5aec13440b] drm/nvf0/gr: magic
> sequence that makes PGRAPH come out of hiding
> git bisect bad b054aadfb0030b9717bb22f4283bfe5aec13440b
> # good: [a8004a9edd8c6ee86c3e263c9082ddf64797a667] drm/nvc0-/gr: bump
> maximum gpc/tpc limits
> git bisect good a8004a9edd8c6ee86c3e263c9082ddf64797a667
> # good: [0d4a1450c95801c21ba4db109303fbad62378b91] drm/nouveau/vdec: fork
> vp3 implementations from vp2
> git bisect good 0d4a1450c95801c21ba4db109303fbad62378b91
> # bad: [c03ff9e8fa5fc0186158b99a89f613325ff352cf] drm/nvc0-/gr: pull out a
> group of separately context-switched gpc regs
> git bisect bad c03ff9e8fa5fc0186158b99a89f613325ff352cf
> # good: [05f9a5bc58381f58095d8789e1c2d4e18758c2bc] drm/nouveau/bsp/nv84:
> initial vp2 engine implementation
> git bisect good 05f9a5bc58381f58095d8789e1c2d4e18758c2bc
> # good: [791dc143ed2c441f5202d8721609d94dce9fcf88] drm/nvd0-/disp: handle
> case where display engine is missing/disabled
> git bisect good 791dc143ed2c441f5202d8721609d94dce9fcf88
> 
> 
> don't know if that help?
Seems like there was one more step to complete the bisection.

Either way your issue/dmesg looks exactly the same as bug 69203, despite that you've notice it at a later stage (when you s2ram).

*** This bug has been marked as a duplicate of bug 69203 ***

what command i will need to type to do the step missing?