[Intel-gfx] [RESEND PATCH 4/6] drm/i915: Update atomic modeset state synchronously

[Maarten Lankhorst]

Op 21-10-16 om 16:21 schreef Ville Syrjälä:
> On Wed, Oct 19, 2016 at 03:55:37PM +0200, Maarten Lankhorst wrote:
>> All of this state should be updated as soon as possible. It shouldn't be
>> done later because then future updates may not depend on it.
>>
>> Signed-off-by: Maarten Lankhorst <maarten.lankhorst at linux.intel.com>
>> ---
>>  drivers/gpu/drm/i915/intel_display.c | 15 ++++++++-------
>>  1 file changed, 8 insertions(+), 7 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
>> index 69b9e91f071e..ba7f7be3aa4f 100644
>> --- a/drivers/gpu/drm/i915/intel_display.c
>> +++ b/drivers/gpu/drm/i915/intel_display.c
>> @@ -14341,14 +14341,8 @@ static void intel_atomic_commit_tail(struct drm_atomic_state *state)
>>  
>>  	drm_atomic_helper_wait_for_dependencies(state);
>>  
>> -	if (intel_state->modeset) {
>> -		memcpy(dev_priv->min_pixclk, intel_state->min_pixclk,
>> -		       sizeof(intel_state->min_pixclk));
>> -		dev_priv->active_crtcs = intel_state->active_crtcs;
>> -		dev_priv->atomic_cdclk_freq = intel_state->cdclk;
>> -
>> +	if (intel_state->modeset)
>>  		intel_display_power_get(dev_priv, POWER_DOMAIN_MODESET);
>> -	}
>>  
>>  	for_each_crtc_in_state(state, crtc, old_crtc_state, i) {
>>  		struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
>> @@ -14558,6 +14552,13 @@ static int intel_atomic_commit(struct drm_device *dev,
>>  	intel_atomic_track_fbs(state);
>>  
>>  	drm_atomic_state_get(state);
>> +	if (intel_state->modeset) {
>> +		memcpy(dev_priv->min_pixclk, intel_state->min_pixclk,
>> +		       sizeof(intel_state->min_pixclk));
>> +		dev_priv->active_crtcs = intel_state->active_crtcs;
>> +		dev_priv->atomic_cdclk_freq = intel_state->cdclk;
>> +	}
> The placement after the _get() looks a bit weird. The other parts of the
> state were flipped over before the _get(), so putting this next to the
> other would make it look less magicy.
>
That's fine, this patch was rebased after thje drm_atomic_state_get was added, which explains the placement.

~Maarten