diff options
| author | Felix Kuehling <Felix.Kuehling@amd.com> | 2012-02-23 19:16:12 -0500 |
|---|---|---|
| committer | Dave Airlie <airlied@redhat.com> | 2012-02-29 10:23:50 +0000 |
| commit | 81ffbbedc37c6043e5f5b123da926aa7dd8ad60a (patch) | |
| tree | f092c02ae2baca58e695f59a3d25f9881dfe43b2 /drivers/gpu/drm/radeon/radeon_display.c | |
| parent | 884a53ef43eb69dfd48408659b9606e581aee7ba (diff) | |
drm/radeon: fix deferred page-flip detection logic on Avivo-based ASICs
This fixes page-flip-related flickering observed on Iconia Tab W500.
The update_pending status returned by radeon_page_flip is very accurate on
Avivo-based ASICs when vpos is negative.
Experiments were conducted on several ASIC generations ranging from RS690
to Cayman where the page flip was artificially timed to occur at a specific
vpos. With negative vpos, overriding update_pending always lead to
flickering.
The same experiment on RV380 and RV410 showed that update_pending is not
accurate with negative vpos. In most cases update_pending == 1 is returned
although the flip would complete before the start of the next frame.
Therefore I left the behaviour unchanged for pre-AVIVO ASICs for
performance reasons, although this may result in flickering in rare cases.
This change also makes the logic a little easier to understand.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Diffstat (limited to 'drivers/gpu/drm/radeon/radeon_display.c')
| -rw-r--r-- | drivers/gpu/drm/radeon/radeon_display.c | 13 |
1 files changed, 11 insertions, 2 deletions
diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index 7cb062daa71..1f98e5fbc73 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -303,8 +303,17 @@ void radeon_crtc_handle_flip(struct radeon_device *rdev, int crtc_id) if (update_pending && (DRM_SCANOUTPOS_VALID & radeon_get_crtc_scanoutpos(rdev->ddev, crtc_id, &vpos, &hpos)) && - (vpos >=0) && - (vpos < (99 * rdev->mode_info.crtcs[crtc_id]->base.hwmode.crtc_vdisplay)/100)) { + ((vpos >= (99 * rdev->mode_info.crtcs[crtc_id]->base.hwmode.crtc_vdisplay)/100) || + (vpos < 0 && !ASIC_IS_AVIVO(rdev)))) { + /* crtc didn't flip in this target vblank interval, + * but flip is pending in crtc. Based on the current + * scanout position we know that the current frame is + * (nearly) complete and the flip will (likely) + * complete before the start of the next frame. + */ + update_pending = 0; + } + if (update_pending) { /* crtc didn't flip in this target vblank interval, * but flip is pending in crtc. It will complete it * in next vblank interval, so complete the flip at |