[Intel-gfx] [PATCH 1/3] drm/i915/dp: Make sure all tiled connectors get added to the state with full modeset
Matt Roper
matthew.d.roper at intel.com
Fri Dec 13 03:11:15 UTC 2019
On Thu, Dec 12, 2019 at 05:18:02PM -0800, Manasi Navare wrote:
> On Thu, Dec 12, 2019 at 04:32:32PM -0800, Matt Roper wrote:
> > On Wed, Dec 11, 2019 at 01:14:23PM -0800, Manasi Navare wrote:
> > > In case of tiled displays, all the tiles are linke dto each other
> >
> > Minor typo on "linked to" here.
>
> I will fix it
>
> >
> > > for transcoder port sync. So in intel_atomic_check() we need to make
> > > sure that we add all the tiles to the modeset and if one of the
> > > tiles needs a full modeset then mark all other tiles for a full modeset.
> > >
> > > Suggested-by: Ville Syrjälä <ville.syrjala at linux.intel.com>
> > > Cc: Ville Syrjälä <ville.syrjala at linux.intel.com>
> > > Cc: José Roberto de Souza <jose.souza at intel.com>
> > > Bugzilla: https://gitlab.freedesktop.org/drm/intel/issues/5
> >
> > I think we're moving to "Closes:" as the annotation here now that it's
> > not actually a bugzilla bug database anymore.
>
> Ok cool, will change that to Closes
>
> >
> > > Signed-off-by: Manasi Navare <manasi.d.navare at intel.com>
> > > ---
> > > drivers/gpu/drm/i915/display/intel_display.c | 78 ++++++++++++++++++++
> > > 1 file changed, 78 insertions(+)
> > >
> > > diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c
> > > index 803993a01ca7..7263eaa66cda 100644
> > > --- a/drivers/gpu/drm/i915/display/intel_display.c
> > > +++ b/drivers/gpu/drm/i915/display/intel_display.c
> > > @@ -14066,6 +14066,80 @@ static int intel_atomic_check_crtcs(struct intel_atomic_state *state)
> > > return 0;
> > > }
> > >
> > > +static int
> > > +intel_dp_modeset_all_tiles(struct drm_i915_private *dev_priv,
> > > + struct intel_atomic_state *state, int tile_grp_id)
> > > +{
> > > + struct drm_connector *conn_iter;
> > > + struct drm_connector_list_iter conn_list_iter;
> > > + struct drm_crtc_state *crtc_state;
> > > +
> > > + drm_connector_list_iter_begin(&dev_priv->drm, &conn_list_iter);
> > > + drm_for_each_connector_iter(conn_iter, &conn_list_iter) {
> > > + struct drm_connector_state *conn_iter_state;
> > > +
> > > + if (!conn_iter->has_tile)
> > > + continue;
> > > + conn_iter_state = drm_atomic_get_connector_state(&state->base,
> > > + conn_iter);
> > > + if (IS_ERR(conn_iter_state)) {
> > > + drm_connector_list_iter_end(&conn_list_iter);
> > > + return PTR_ERR(conn_iter_state);
> > > + }
> > > +
> > > + if (!conn_iter_state->crtc)
> > > + continue;
> > > +
> > > + if (conn_iter->tile_group->id != tile_grp_id)
> > > + continue;
> > > +
> > > + crtc_state = drm_atomic_get_crtc_state(&state->base, conn_iter_state->crtc);
> > > + if (IS_ERR(crtc_state)) {
> > > + drm_connector_list_iter_end(&conn_list_iter);
> > > + return PTR_ERR(conn_iter_state);
> > > + }
> > > + crtc_state->mode_changed = true;
> > > + }
> > > + drm_connector_list_iter_end(&conn_list_iter);
> > > +
> > > + return 0;
> > > +}
> > > +
> > > +static int
> > > +intel_dp_atomic_trans_port_sync_check(struct drm_i915_private *dev_priv,
> > > + struct intel_atomic_state *state)
> > > +{
> > > + struct drm_connector *connector;
> > > + struct drm_crtc_state *crtc_state;
> > > + struct drm_connector_state *connector_state;
> > > + int i, ret, tile_grp_id = 0;
> > > +
> > > + if (INTEL_GEN(dev_priv) < 11)
> > > + return 0;
> > > +
> > > + /* Is tiled, mark all other tiled CRTCs as needing a modeset */
> > > + for_each_new_connector_in_state(&state->base, connector, connector_state, i) {
> > > + if (!connector->has_tile)
> > > + continue;
> > > + if (connector_state->crtc &&
> > > + tile_grp_id != connector->tile_group->id) {
> > > + crtc_state = drm_atomic_get_new_crtc_state(&state->base,
> > > + connector_state->crtc);
> > > + if (!drm_atomic_crtc_needs_modeset(crtc_state))
> > > + continue;
> > > +
> > > + tile_grp_id = connector->tile_group->id;
> > > + } else
> >
> > Minor kernel coding style violation; if we use {} on one branch of an
> > if, we need to use them on all.
> >
>
> Yes i got a checkpatch check warning, will fix it
>
> > > + continue;
> > > +
> > > + ret = intel_dp_modeset_all_tiles(dev_priv, state, tile_grp_id);
> > > + if (ret)
> > > + return ret;
> > > + }
> > > +
> > > + return 0;
> > > +}
> > > +
> > > /**
> > > * intel_atomic_check - validate state object
> > > * @dev: drm device
> > > @@ -14093,6 +14167,10 @@ static int intel_atomic_check(struct drm_device *dev,
> > > if (ret)
> > > goto fail;
> > >
> > > + ret = intel_dp_atomic_trans_port_sync_check(dev_priv, state);
> > > + if (ret)
> > > + goto fail;
> >
> > Should this happen before the drm_atomic_helper_check_modeset() just
> > above (or should we re-call that function if we flag the other tile as
> > needing a modeset)? The kerneldoc on that function says:
> >
> > """
> > Drivers which set &drm_crtc_state.mode_changed [...] _must_ call this
> > function afterwards after that change. It is permitted to call this
> > function multiple times for the same update ...
> > """
> >
>
> IMO, here infact it makes sense to call my function after the drm_atomic_helper_check_modeset()
> because it directly sets the new_crtc_state->mode_changed to true for all tiles if 1 of them needs
> a full modeset.
> And whether that one tile needs a full modeset or not will be decided based on mode changed for that set
> in drm_atomic_helper_check_modeset.
Okay, makes sense. But based on the comment, I believe we still need to
call the helper a second time during/after your function in the event
that we've switched any additional crtcs to mode_changed=true.
I'm also wondering whether we should eventually move most of your logic
here out of i915 and into a general drm helper that other drivers can
utilize too. But I'm okay with doing that as a followup after we've
landed it and verified it as an i915-specific change first.
Matt
>
> Manasi
>
> >
> > Matt
> >
> > > +
> > > for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state,
> > > new_crtc_state, i) {
> > > if (!needs_modeset(new_crtc_state)) {
> > > --
> > > 2.19.1
> > >
> > > _______________________________________________
> > > Intel-gfx mailing list
> > > Intel-gfx at lists.freedesktop.org
> > > https://lists.freedesktop.org/mailman/listinfo/intel-gfx
> >
> > --
> > Matt Roper
> > Graphics Software Engineer
> > VTT-OSGC Platform Enablement
> > Intel Corporation
> > (916) 356-2795
--
Matt Roper
Graphics Software Engineer
VTT-OSGC Platform Enablement
Intel Corporation
(916) 356-2795
More information about the Intel-gfx
mailing list