[PATCH] drm/amdgpu: Fix error in amdgpu_bo_create_restricted()

[Nils Holland]

On Tue, Jan 24, 2017 at 09:53:24AM +0100, Christian König wrote:
> Am 24.01.2017 um 08:43 schrieb Nils Holland:
> > On Tue, Jan 24, 2017 at 10:35:16AM +0900, Michel Dänzer wrote:
> >> On 24/01/17 09:55 AM, Nils Holland wrote:
> >>> So I decided to fix this with the following follow-up patch. I hope
> >>> this is the right approach (vs. reverting the commit and instead using
> >>> a fixed v2 of the original patch).
> >>>
> >>>  From 41775d2c8a14873f522667a57b66cfbe119e28a4 Mon Sep 17 00:00:00 2001
> >>> From: Nils Holland <nholland at tisys.org>
> >>> Date: Tue, 24 Jan 2017 01:36:45 +0100
> >>> Subject: [PATCH] drm/amdgpu: Fix error in amdgpu_bo_create_restricted()
> >> Note that the way your e-mail was formatted, the above commentary and
> >> mail headers would end up in the Git commit log, unless the person
> >> applying the patch pays attention and removes them. Commentary that
> >> isn't to be part of the commit log should be added between the ---
> >> marker below and the actual code diff.
> > Ah, I see! Making the job unnecessarily difficult for a maintainer is
> > no good idea, so I will certainly do this better / correctly the next
> > time I submit something!
> >
> >> Reviewed-by: Michel Dänzer <michel.daenzer at amd.com>
> 
> Reviewed-by: Christian König <christian.koenig at amd.com> as well.

Just a note: I've just had a look at the official amdgpu Git tree for
the first time, and saw that when Alex Deucher commited my patch into
drm-next-4.11-wip yesterday, he obviously already spotted the mistake
and fixed it. That would mean that my follow-up patch is void and can
be discarded. I should probably have verified the state of the patch in
the tree before reacting to the kernel test robot's output (or,
rather: Should have tested well enough to prevent this mistake right
from the start...)

Thanks Alex!

Greetings
Nils