Op 10-08-16 om 16:14 schreef Daniel Vetter:
> On Wed, Aug 10, 2016 at 12:46:23PM +0200, Maarten Lankhorst wrote:
>> When doing a atomic commit affecting multiple crtc's, multiple events
>> are generated. The user_data member does not allow you to distinguish,
>> because they all have the same pointer.
>>
>> I've chosen to use crtc_id, because using pipe would create ambiguity
>> when pipe = 0. A test for != 0 is easier to implement, and crtc_id
>> will never be 0.
>>
>> Signed-off-by: Maarten Lankhorst <maarten.lankhorst@xxxxxxxxxxxxxxx>
>> Cc: David Airlie <airlied@xxxxxxxx>
>> Cc: Daniel Stone <daniels@xxxxxxxxxxxxx>
>> ---
>> drivers/gpu/drm/drm_irq.c | 2 ++
>> include/uapi/drm/drm.h | 2 +-
>> 2 files changed, 3 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c
>> index 97c70642dbe1..f01bb0eea31c 100644
>> --- a/drivers/gpu/drm/drm_irq.c
>> +++ b/drivers/gpu/drm/drm_irq.c
>> @@ -1021,6 +1021,7 @@ void drm_crtc_arm_vblank_event(struct drm_crtc *crtc,
>>
>> e->pipe = pipe;
>> e->event.sequence = drm_vblank_count(dev, pipe);
>> + e->event.crtc_id = crtc->base.id;
>> list_add_tail(&e->base.link, &dev->vblank_event_list);
>> }
>> EXPORT_SYMBOL(drm_crtc_arm_vblank_event);
>> @@ -1048,6 +1049,7 @@ void drm_crtc_send_vblank_event(struct drm_crtc *crtc,
>> now = get_drm_timestamp();
>> }
>> e->pipe = pipe;
>> + e->event.crtc_id = crtc->base.id;
>> send_vblank_event(dev, e, seq, &now);
>> }
>> EXPORT_SYMBOL(drm_crtc_send_vblank_event);
>> diff --git a/include/uapi/drm/drm.h b/include/uapi/drm/drm.h
>> index 452675fb55d9..dced4d4517d6 100644
>> --- a/include/uapi/drm/drm.h
>> +++ b/include/uapi/drm/drm.h
>> @@ -850,7 +850,7 @@ struct drm_event_vblank {
>> __u32 tv_sec;
>> __u32 tv_usec;
>> __u32 sequence;
>> - __u32 reserved;
>> + __u32 crtc_id; /* 0 on older kernels that do not support this */
> Same comment as in reply to Michel's patch: Is there really no existing
> userspace which would fail to compile now because we renamed this? Simply
> creating a new drm_event_vblank2 struct would fix that.
Hey,
This comment must have fallen through, I haven't found any case in which this happened. All of them use a memset or allocate it zero'd.
~Maarten
_______________________________________________
dri-devel mailing list
dri-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/dri-devel
