[Intel-gfx] 3.7-rc7 IVB forcewake resume timeouts...

[Daniel J Blueman]

On 29 November 2012 02:57, Ben Widawsky <ben at bwidawsk.net> wrote:
> On Wed, 28 Nov 2012 11:13:14 +0800
> Daniel J Blueman <daniel at quora.org> wrote:
>
>> Hi Daniel/Ben,
>>
>> On stock 3.7-rc7, when resuming my Ivy Bridge laptop from suspend, the
>> kernel logs:
>>
>> [drm:__gen6_gt_force_wake_mt_get] *ERROR* Timed out waiting for
>> forcewake to ack request.
>> [drm:__gen6_gt_wait_for_thread_c0] *ERROR* GT thread status wait timed out
>> [drm:__gen6_gt_force_wake_mt_get] *ERROR* Timed out waiting for
>> forcewake to ack request.
>> [drm:__gen6_gt_force_wake_mt_get] *ERROR* Timed out waiting for
>> forcewake old ack to clear.
>> [drm:__gen6_gt_force_wake_mt_get] *ERROR* Timed out waiting for
>> forcewake to ack request.
>> [drm:__gen6_gt_force_wake_mt_get] *ERROR* Timed out waiting for
>> forcewake to ack request.
>> [drm:__gen6_gt_force_wake_mt_get] *ERROR* Timed out waiting for
>> forcewake to ack request.
>> [drm:__gen6_gt_force_wake_mt_get] *ERROR* Timed out waiting for
>> forcewake old ack to clear.
>> [drm:__gen6_gt_force_wake_mt_get] *ERROR* Timed out waiting for
>> forcewake to ack request.
>> [drm:__gen6_gt_force_wake_mt_get] *ERROR* Timed out waiting for
>> forcewake old ack to clear.
>>
>> I'm running xserver-xord-video-intel 2.20.9 and xorg-server 1.13.0.
>>
>> Interested in any debug output?

> Is this a regression? The first one is sort of expected, but the rest
> are no good.
>
> Timing printk and lspci info would be a good start.

This doesn't occur in 3.6, but does occur in 3.7-rc4+, so has
regressed. I've been having a hard time bisecting due to another
kernel-hang issue fixed in 3.7-rc4, so can backport the fixes I need
to bisect or scope the bisection.

I've put a full 3.7-rc7 boot log and suspend-resume cycle at:
http://quora.org/2012/i915-suspend-forcewake.txt
Last boot and suspend-resume cycle, I saw only one forcewake message;
here I see a number of ones and related.

Thanks,
  Daniel
--
Daniel J Blueman