[Intel-gfx] [PATCH 9/9] drm/atomic: Add encoder_mask to crtc_state.

[Maarten Lankhorst]

Op 03-12-15 om 10:53 schreef Daniel Vetter:
> On Tue, Nov 24, 2015 at 10:34:36AM +0100, Maarten Lankhorst wrote:
>> This allows iteration over encoders without requiring connection_mutex.
>>
>> Signed-off-by: Maarten Lankhorst <maarten.lankhorst at linux.intel.com>
>> ---
>>  drivers/gpu/drm/drm_atomic_helper.c  | 4 ++++
>>  drivers/gpu/drm/i915/intel_display.c | 3 +++
>>  include/drm/drm_crtc.h               | 2 ++
>>  3 files changed, 9 insertions(+)
>>
>> diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c
>> index fb79c54b2aed..f3fd8f131f92 100644
>> --- a/drivers/gpu/drm/drm_atomic_helper.c
>> +++ b/drivers/gpu/drm/drm_atomic_helper.c
>> @@ -269,6 +269,8 @@ mode_fixup(struct drm_atomic_state *state)
>>  			continue;
>>  
>>  		drm_mode_copy(&crtc_state->adjusted_mode, &crtc_state->mode);
>> +
>> +		crtc_state->encoder_mask = 0;
> Hm, I think a small function to set the best_encoder (like we do to set
> the crtc for connector or planes) would be good. Otherwise we'll frob
> around the code and forget this, and much confusion will ensue.
>
> That helper should probably be in core drm_atomic.c, like the other
> set_foo_for_bar helpers.
As always only i915 assigns best_encoder outside drm_atomic_helper_check_modeset, and the i915 calls can't be fixed because of hw readout. :(
At the time of mode_fixup all encoders should have been updated, so I'm not sure adding a helper for best_encoder would help much..

~Maarten