| Message ID | 1476278901-15750-9-git-send-email-maarten.lankhorst@linux.intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Wed, Oct 12, 2016 at 03:28:21PM +0200, Maarten Lankhorst wrote: > This is the last bit required for making nonblocking modesets work > correctly. The state in intel_crtc->hw_ddb is not updated until > somewhere in atomic commit, while the previous crtc state should be > accurate if the ddb hasn't changed. > > Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Can we get rid of hw_ddb completely and always pull from old state? It looks like the only other place it's still used is skl_update_crtcs() -> skl_ddb_allocation_overlaps(). Matt > --- > drivers/gpu/drm/i915/intel_display.c | 2 +- > drivers/gpu/drm/i915/intel_pm.c | 6 +++++- > 2 files changed, 6 insertions(+), 2 deletions(-) > > diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c > index d3d7d9dc14a8..93e16da0aa51 100644 > --- a/drivers/gpu/drm/i915/intel_display.c > +++ b/drivers/gpu/drm/i915/intel_display.c > @@ -14332,7 +14332,7 @@ static void skl_update_crtcs(struct drm_atomic_state *state, > * new ddb allocation to take effect. > */ > if (!skl_ddb_entry_equal(&cstate->wm.skl.ddb, > - &intel_crtc->hw_ddb) && > + &to_intel_crtc_state(old_crtc_state)->wm.skl.ddb) && > !crtc->state->active_changed && > intel_state->wm_results.dirty_pipes != updated) > vbl_wait = true; > diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c > index 18c62d1eea19..182e6b30b60a 100644 > --- a/drivers/gpu/drm/i915/intel_pm.c > +++ b/drivers/gpu/drm/i915/intel_pm.c > @@ -3091,7 +3091,11 @@ skl_ddb_get_pipe_allocation_limits(struct drm_device *dev, > * we currently hold. > */ > if (!intel_state->active_pipe_changes) { > - *alloc = to_intel_crtc(for_crtc)->hw_ddb; > + /* > + * alloc may be cleared by clear_intel_crtc_state, > + * copy from old state to be sure > + */ > + *alloc = to_intel_crtc_state(for_crtc->state)->wm.skl.ddb; > return; > } > > -- > 2.7.4 > > _______________________________________________ > Intel-gfx mailing list > Intel-gfx@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Op 21-10-16 om 00:09 schreef Matt Roper: > On Wed, Oct 12, 2016 at 03:28:21PM +0200, Maarten Lankhorst wrote: >> This is the last bit required for making nonblocking modesets work >> correctly. The state in intel_crtc->hw_ddb is not updated until >> somewhere in atomic commit, while the previous crtc state should be >> accurate if the ddb hasn't changed. >> >> Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> > Can we get rid of hw_ddb completely and always pull from old state? It > looks like the only other place it's still used is > skl_update_crtcs() -> skl_ddb_allocation_overlaps(). Probably as a followup patch. I kept it only for that reason indeed. I suppose it could be changed to if (updated) state = new_state->ddb else state = old_state->ddb.
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index d3d7d9dc14a8..93e16da0aa51 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -14332,7 +14332,7 @@ static void skl_update_crtcs(struct drm_atomic_state *state, * new ddb allocation to take effect. */ if (!skl_ddb_entry_equal(&cstate->wm.skl.ddb, - &intel_crtc->hw_ddb) && + &to_intel_crtc_state(old_crtc_state)->wm.skl.ddb) && !crtc->state->active_changed && intel_state->wm_results.dirty_pipes != updated) vbl_wait = true; diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 18c62d1eea19..182e6b30b60a 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -3091,7 +3091,11 @@ skl_ddb_get_pipe_allocation_limits(struct drm_device *dev, * we currently hold. */ if (!intel_state->active_pipe_changes) { - *alloc = to_intel_crtc(for_crtc)->hw_ddb; + /* + * alloc may be cleared by clear_intel_crtc_state, + * copy from old state to be sure + */ + *alloc = to_intel_crtc_state(for_crtc->state)->wm.skl.ddb; return; }
This is the last bit required for making nonblocking modesets work correctly. The state in intel_crtc->hw_ddb is not updated until somewhere in atomic commit, while the previous crtc state should be accurate if the ddb hasn't changed. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> --- drivers/gpu/drm/i915/intel_display.c | 2 +- drivers/gpu/drm/i915/intel_pm.c | 6 +++++- 2 files changed, 6 insertions(+), 2 deletions(-)