[bisected regression] Re: scheduling while atomic on radeon rv620, kernel 3.30-rc3

[Dave Airlie]

On Sun, Feb 12, 2012 at 5:20 PM, Mikko Vinni <mmvinni at yahoo.com> wrote:
> Hi again,
>
> I bisected this problem to:
>
> 47492a23a128e953bd5087b1cac909cd8124ca5e is the first bad commit
> commit 47492a23a128e953bd5087b1cac909cd8124ca5e
> Author: Christian König <deathsimple at vodafone.de>
> Date:   Thu Oct 20 12:38:09 2011 +0200
>
>     drm/radeon: add radeon_fence_count_emited function
>
>     Split counting of emited fences out of power
>     management into a seperate function.
>
>     Signed-off-by: Christian König <deathsimple at vodafone.de>
>     Reviewed-by: Jerome Glisse <jglisse at redhat.com>
>     Signed-off-by: Dave Airlie <airlied at redhat.com>
>
> :040000 040000 adbb5199d8ee03a75f65b9ddf6b1d398acfca4d9 67b8d0bfdd5507d82b0c8d8f500f32d140520f0d M    drivers
>
>
>
> That patch (part of it below) moves code into a new function and perhaps changes
> locking a bit. Is the return in the middle ok?

I've just posted a patch to fix that to dri-devel, can you check it
fixes your issue?

Dave.