[PATCH] drm/ttm: Don't evict BOs outside of the requested placement range
Michel Dänzer
michel at daenzer.net
Fri Oct 10 01:59:19 PDT 2014
On 10.10.2014 17:51, Alan Swanson wrote:
> On Fri, 2014-10-10 at 12:20 +0900, Michel Dänzer wrote:
>> On 09.10.2014 19:22, Alan Swanson wrote:
>>> On 2014-10-09 07:02, Michel Dänzer wrote:
>>>> From: Michel Dänzer <michel.daenzer at amd.com>
>>>>
>>>> The radeon driver uses placement range restrictions for several reasons,
>>>> in particular to make sure BOs in VRAM can be accessed by the CPU, e.g.
>>>> during a page fault.
>>>>
>>>> Without this change, TTM could evict other BOs while trying to satisfy
>>>> the requested placement, even if the evicted BOs were outside of the
>>>> requested placement range. Doing so didn't free up any space in the
>>>> requested placement range, so the (potentially high) eviction cost was
>>>> incurred for no benefit.
>>>>
>>>> Nominating for stable because radeon driver changes in 3.17 made this
>>>> much more noticeable than before.
>>>>
>>>> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=84662
>>>> Cc: stable at vger.kernel.org
>>>> Signed-off-by: Michel Dänzer <michel.daenzer at amd.com>
>>>> ---
>>>> drivers/gpu/drm/ttm/ttm_bo.c | 20 +++++++++++++++++---
>>>> 1 file changed, 17 insertions(+), 3 deletions(-)
>>>>
>> [...]
>>
>>> I believe you need to "s/place/placement/" over this patch.
>>
>> The fpfn and lpfn members were moved from struct ttm_placement to a new
>> struct ttm_place in f1217ed09f827e42a49ffa6a5aab672aa6f57a65.
>>
>> If you mean something else, please elaborate.
>
> This patch failed to build on 3.17.0 so wouldn't be a candidate for
> stable unless the currently drm-next only ttm_place patch also goes to
> stable (else replace ttm_place with ttm_placements in the patch for
> stable)?
Right, I guess I should drop the Cc: stable then and submit a manual
backport of it to the stable list once it has landed in Linus' tree.
--
Earthling Michel Dänzer | http://www.amd.com
Libre software enthusiast | Mesa and X developer
More information about the dri-devel
mailing list