[PATCH 03/26] drm/amd/display: Flip All Planes Under OTG Master When Flip Immediate
Rodrigo Siqueira
Rodrigo.Siqueira at amd.com
Thu Oct 3 23:33:21 UTC 2024
From: Austin Zheng <Austin.Zheng at amd.com>
[Why]
The MPO plane will receive a flip but
desktop plane may not receive a flip when GSL is enabled.
As a result, system will be stuck waiting for a flip that was never sent.
[How]
Set update address update flag of all flip_immediate planes
if there are multiple planes.
Reviewed-by: Dillon Varone <dillon.varone at amd.com>
Signed-off-by: Austin Zheng <Austin.Zheng at amd.com>
Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira at amd.com>
---
drivers/gpu/drm/amd/display/dc/core/dc.c | 33 ++++++++++++++++++++++++
1 file changed, 33 insertions(+)
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index f4a5accf1db9..48057ac22cbd 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -2661,6 +2661,29 @@ static enum surface_update_type det_surface_update(const struct dc *dc,
return overall_type;
}
+/* May need to flip the desktop plane in cases where MPO plane receives a flip but desktop plane doesn't
+ * while both planes are flip_immediate
+ */
+static void force_immediate_gsl_plane_flip(struct dc *dc, struct dc_surface_update *updates, int surface_count)
+{
+ bool has_flip_immediate_plane = false;
+ int i;
+
+ for (i = 0; i < surface_count; i++) {
+ if (updates[i].surface->flip_immediate) {
+ has_flip_immediate_plane = true;
+ break;
+ }
+ }
+
+ if (has_flip_immediate_plane && surface_count > 1) {
+ for (i = 0; i < surface_count; i++) {
+ if (updates[i].surface->flip_immediate)
+ updates[i].surface->update_flags.bits.addr_update = 1;
+ }
+ }
+}
+
static enum surface_update_type check_update_surfaces_for_stream(
struct dc *dc,
struct dc_surface_update *updates,
@@ -3177,6 +3200,11 @@ static bool update_planes_and_stream_state(struct dc *dc,
context = dc->current_state;
update_type = dc_check_update_surfaces_for_stream(
dc, srf_updates, surface_count, stream_update, stream_status);
+ /* It is possible to receive a flip for one plane while there are multiple flip_immediate planes in the same stream.
+ * E.g. Desktop and MPO plane are flip_immediate but only the MPO plane received a flip
+ * Force the other flip_immediate planes to flip so GSL doesn't wait for a flip that won't come.
+ */
+ force_immediate_gsl_plane_flip(dc, srf_updates, surface_count);
if (update_type == UPDATE_TYPE_FULL)
backup_planes_and_stream_state(&dc->scratch.current_state, stream);
@@ -4807,6 +4835,11 @@ static bool update_planes_and_stream_v1(struct dc *dc,
update_type = dc_check_update_surfaces_for_stream(
dc, srf_updates, surface_count, stream_update, stream_status);
+ /* It is possible to receive a flip for one plane while there are multiple flip_immediate planes in the same stream.
+ * E.g. Desktop and MPO plane are flip_immediate but only the MPO plane received a flip
+ * Force the other flip_immediate planes to flip so GSL doesn't wait for a flip that won't come.
+ */
+ force_immediate_gsl_plane_flip(dc, srf_updates, surface_count);
if (update_type >= UPDATE_TYPE_FULL) {
--
2.45.2
More information about the amd-gfx
mailing list