On Fri, Dec 20, 2019 at 08:41:12PM +0200, Ville Syrjälä wrote: > On Thu, Dec 19, 2019 at 01:51:15PM -0800, Manasi Navare wrote: > > In case of tiled displays, all the tiles are linke dto each other > > 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. > > > > v2: > > * Change crtc_state scope, remove tile_grp_id (Ville) > > * Use intel_connector_needs_modeset() (Ville) > > * Add modeset_synced_crtcs (Ville) > > * Make sure synced crtcs are forced full modeset > > after fastset check (Ville) > > > > Suggested-by: Ville Syrjälä <ville.syrjala@xxxxxxxxxxxxxxx> > > Cc: Ville Syrjälä <ville.syrjala@xxxxxxxxxxxxxxx> > > Cc: José Roberto de Souza <jose.souza@xxxxxxxxx> > > Cc: Matt Roper <matthew.d.roper@xxxxxxxxx> > > Bugzilla: https://gitlab.freedesktop.org/drm/intel/issues/5 > > Signed-off-by: Manasi Navare <manasi.d.navare@xxxxxxxxx> > > --- > > drivers/gpu/drm/i915/display/intel_display.c | 143 +++++++++++++++++-- > > 1 file changed, 131 insertions(+), 12 deletions(-) > > > > diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c > > index a3f9430493ae..00608d8cef50 100644 > > --- a/drivers/gpu/drm/i915/display/intel_display.c > > +++ b/drivers/gpu/drm/i915/display/intel_display.c > > @@ -13910,18 +13910,6 @@ static void intel_crtc_check_fastset(const struct intel_crtc_state *old_crtc_sta > > new_crtc_state->uapi.mode_changed = false; > > new_crtc_state->update_pipe = true; > > > > - /* > > - * If we're not doing the full modeset we want to > > - * keep the current M/N values as they may be > > - * sufficiently different to the computed values > > - * to cause problems. > > - * > > - * FIXME: should really copy more fuzzy state here > > - */ > > - new_crtc_state->fdi_m_n = old_crtc_state->fdi_m_n; > > - new_crtc_state->dp_m_n = old_crtc_state->dp_m_n; > > - new_crtc_state->dp_m2_n2 = old_crtc_state->dp_m2_n2; > > - new_crtc_state->has_drrs = old_crtc_state->has_drrs; > > } > > > > static int intel_crtc_add_planes_to_state(struct intel_atomic_state *state, > > @@ -14032,6 +14020,105 @@ static int intel_atomic_check_crtcs(struct intel_atomic_state *state) > > return 0; > > } > > > > +static void > > +intel_dp_modeset_synced_crtcs(struct intel_atomic_state *state) > > +{ > > + struct intel_crtc_state *new_crtc_state; > > + struct intel_crtc *crtc; > > + int i; > > + > > + for_each_new_intel_crtc_in_state(state, crtc, > > + new_crtc_state, i) { > > + if (is_trans_port_sync_mode(new_crtc_state)) { > > + new_crtc_state->uapi.mode_changed = true; > > + new_crtc_state->update_pipe = false; > > + } > > Upon further thinking this is not quite right even for the post fastset > check part. You're not checking at all whether these crtcs here are > actually synced to the crtc that still had its needs_modeset() flagged. > So this can end up forcing a modeset on an entirely unrelated crtc. We > don't want that. > > You could fix that by passing in a transcoder bitmask, and then only > flagging a modeset on crtcs whose cpu_transcoder matches the bitmask. > (see intel_modeset_transcoders() in my branch). > > > + } > > +} > > + > > +static void > > +intel_dp_atomic_check_synced_crtcs(struct intel_atomic_state *state) > > +{ > > + struct intel_crtc_state *new_crtc_state; > > + struct intel_crtc *crtc; > > + int i; > > + > > + for_each_new_intel_crtc_in_state(state, crtc, > > + new_crtc_state, i) { > > + if (!is_trans_port_sync_mode(new_crtc_state) || > > + !needs_modeset(new_crtc_state)) > > + continue; > > + > > + intel_dp_modeset_synced_crtcs(state); I guess in this call I can add master trans and sync_mode_slaves_bitmask And then if the crtcs in state are in sync mode and their cpu_trans matches the master trans or transcoders in the slave mask, add them to the modeset? > > + } > > +} > > + > > +static int > > +intel_dp_modeset_all_tiles(struct intel_atomic_state *state, int tile_grp_id) > > +{ > > + struct drm_i915_private *dev_priv = to_i915(state->base.dev); > > + struct drm_connector *connector; > > + struct drm_connector_list_iter conn_list_iter; > > + > > + drm_connector_list_iter_begin(&dev_priv->drm, &conn_list_iter); > > + drm_for_each_connector_iter(connector, &conn_list_iter) { > > + struct drm_connector_state *conn_iter_state; > > + struct drm_crtc_state *crtc_state; > > + > > + if (!(connector->has_tile && > > + connector->tile_group->id == tile_grp_id)) > > + continue; > > + conn_iter_state = drm_atomic_get_connector_state(&state->base, > > + connector); > > + 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; > > + > > + 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_check_tiled_conns(struct intel_atomic_state *state) > > +{ > > + struct drm_i915_private *dev_priv = to_i915(state->base.dev); > > + struct drm_connector *connector; > > + struct drm_connector_state *old_conn_state, *new_conn_state; > > + int i, ret; > > + > > + if (INTEL_GEN(dev_priv) < 11) > > + return 0; > > + > > + /* Is tiled, mark all other tiled CRTCs as needing a modeset */ > > + for_each_oldnew_connector_in_state(&state->base, connector, > > + old_conn_state, new_conn_state, i) { > > + if (!connector->has_tile) > > + continue; > > + if (!intel_connector_needs_modeset(state, old_conn_state, > > + new_conn_state)) > > + continue; > > + > > + ret = intel_dp_modeset_all_tiles(state, connector->tile_group->id); > > + if (ret) > > + return ret; > > + } > > + > > + return 0; > > +} > > + > > /** > > * intel_atomic_check - validate state object > > * @dev: drm device > > @@ -14059,6 +14146,12 @@ static int intel_atomic_check(struct drm_device *dev, > > if (ret) > > goto fail; > > > > + ret = intel_dp_atomic_check_tiled_conns(state); > > + if (ret) > > + goto fail; > > Since this is not sufficient for the pre-compute_config() check I think > we should drop this part for now. It's a somewhat unlikely scenario > after all so we can do it as a followup. > Well we do need this for handling the hotplug cases correctly so that the connector disconnected are added to the modeset which is working well with this function. Manasi > > + > > + intel_dp_atomic_check_synced_crtcs(state); > > + > > for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state, > > new_crtc_state, i) { > > if (!needs_modeset(new_crtc_state)) { > > @@ -14089,6 +14182,32 @@ static int intel_atomic_check(struct drm_device *dev, > > any_ms = true; > > } > > > > + /* > > + * In case of port synced crtcs, if one of the synced crtcs > > + * needs a full modeset, all other synced crtcs should be > > + * forced a full modeset. > > + */ > > + intel_dp_atomic_check_synced_crtcs(state); > > + > > + for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state, > > + new_crtc_state, i) { > > + if (needs_modeset(new_crtc_state)) > > + continue; > > + > > + /* > > + * If we're not doing the full modeset we want to > > + * keep the current M/N values as they may be > > + * sufficiently different to the computed values > > + * to cause problems. > > + * > > + * FIXME: should really copy more fuzzy state here > > + */ > > + new_crtc_state->fdi_m_n = old_crtc_state->fdi_m_n; > > + new_crtc_state->dp_m_n = old_crtc_state->dp_m_n; > > + new_crtc_state->dp_m2_n2 = old_crtc_state->dp_m2_n2; > > + new_crtc_state->has_drrs = old_crtc_state->has_drrs; > > + } > > + > > if (any_ms && !check_digital_port_conflicts(state)) { > > DRM_DEBUG_KMS("rejecting conflicting digital port configuration\n"); > > ret = EINVAL; > > -- > > 2.19.1 > > -- > Ville Syrjälä > Intel _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
