[PATCH 0/5] [Maverick] SRU: Restore video from S3 resume on Sandybridge

[Brad Figg]

On 10/19/2010 01:49 PM, Leann Ogasawara wrote:
> On Tue, 2010-10-19 at 20:03 +0100, Colin King wrote:
>> From: Colin Ian King<colin.king at canonical.com>
>>
>> BugLink: http://launchpad.net/bugs/663442
>>
>> SRU Justification:
>>
>> Impact: Resuming from S3 on Sandybridge laptops results in a blank display
>> and no video.
>>
>> Fix: Attached 5 patches restore fence registers correctly, align the display
>> surface so to avoid tiling corruption and correctly check for sandybrige
>> chipset in resume path.
>>
>> Testcase: Without the patches, resume from S3 results in a blank display with
>> no video being displayed. With the patches, S3 can be run multiple times and
>> video is restored correctly on each resume.
>>
>> Chris Wilson (5):
>>    drm/i915: Silence sparse over duplicate members in static initializer
>
> * upstream commit 2d3fa0de689e1a53fdfc3db50732b62e58d0ca6b
>
>>    drm/i915: Use 128k alignment for untiled display surface on i965 (v2)
>
> * upstream commit 534843dabf79da40561148764916e1b2e6bbcebe
>
>>    drm/i915: Include a generation number in the device info
>
> * upstream commit c96c3a8cb7fadcb33d9a5ebe35fcee8b7d0a7946
>
>>    drm/i915: Fix Sandybridge fence registers
>
> * upstream commit e259befd9013e212648c3bd4f6f1fbf92d0dd51d, CC'd to
> stable
>
>>    drm/i915/suspend: s/IS_IRONLAKE/HAS_PCH_SPLIT/
>
> * upstream commit 90eb77baaea35c591bd324b31e9eac032bd603c9
>
>>   drivers/gpu/drm/i915/i915_drv.c      |   61 +++++++++-----------
>>   drivers/gpu/drm/i915/i915_drv.h      |   31 +++-------
>>   drivers/gpu/drm/i915/i915_gem.c      |   37 ++++++++----
>>   drivers/gpu/drm/i915/i915_suspend.c  |  106 ++++++++++++++++++++--------------
>>   drivers/gpu/drm/i915/intel_display.c |    7 ++-
>>   5 files changed, 130 insertions(+), 112 deletions(-)
>
> The 5 patches are officially upstream, one of which is already CC'd to
> stable.  The patches have been fully tested and confirmed to resolve the
> resume issue.  They look reasonable for the most part, but I'd give this
> a medium risk of regression as they do touch a commonly used i915
> driver.  It's also my understanding that the resume from suspend issue
> fixed by these patches is a regression which was introduced when we
> applied a patch [1] for bug 602281.  It'd be good to let these bake in
> -proposed for a bit.
>
> Acked-by: Leann Ogasawara<leann.ogasawara at canonical.com>
>
> Thanks,
> Leann
>
> [1] commit 9a850554ee3f48334574a17d08d584c0d23ebcad
> Author: Jesse Barnes<jbarnes at virtuousgeek.org>
> Date:   Fri Aug 13 15:11:26 2010 -0700
>
>      drm/i915: fix VGA plane disable for Ironlake+
>
>      BugLink: http://launchpad.net/bugs/602281
>
>

I agree completely with the concern these patches touch a very common
driver and the codepath is used by all models of the related hw.

Acked-by: Brad Figg <brad.figg at canonical.com>


-- 
Brad Figg brad.figg at canonical.com http://www.canonical.com