v3dv: no specific separate_segments flag for V3D 7.1

On V3D 7.1 there is not a flag on the Shader State Record to specify
if we are using shared or separate segments. This is done by setting
the vpm input size to 0 (so we need to ensure that the output would be
the max needed for input/output).

We were already doing the latter on the prog_data_vs, so we just need
to use those values, instead of assigning default values.

As we are here, we also add some comments on the compiler part.

Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>

diff --git a/src/broadcom/compiler/qpu_schedule.c b/src/broadcom/compiler/qpu_schedule.c
index 3890bf4..8649470 100644 (file)
--- a/src/broadcom/compiler/qpu_schedule.c
+++ b/src/broadcom/compiler/qpu_schedule.c
@@ -297,6 +297,10 @@ calculate_deps(struct schedule_state *state, struct schedule_node *n)
         /* If the input and output segments are shared, then all VPM reads to
          * a location need to happen before all writes.  We handle this by
          * serializing all VPM operations for now.
+         *
+         * FIXME: we are assuming that the segments are shared. That is
+         * correct right now as we are only using shared, but technically you
+         * can choose.
          */
         bool separate_vpm_segment = false;
 

diff --git a/src/broadcom/compiler/vir.c b/src/broadcom/compiler/vir.c
index 0adcfba..8c536b8 100644 (file)
--- a/src/broadcom/compiler/vir.c
+++ b/src/broadcom/compiler/vir.c
@@ -753,6 +753,10 @@ v3d_vs_set_prog_data(struct v3d_compile *c,
 
         /* Set us up for shared input/output segments.  This is apparently
          * necessary for our VCM setup to avoid varying corruption.
+         *
+         * FIXME: initial testing on V3D 7.1 seems to work fine when using
+         * separate segments. So we could try to reevaluate in the future, if
+         * there is any advantage of using separate segments.
          */
         prog_data->separate_segments = false;
         prog_data->vpm_output_size = MAX2(prog_data->vpm_output_size,

diff --git a/src/broadcom/vulkan/v3dvx_pipeline.c b/src/broadcom/vulkan/v3dvx_pipeline.c
index a640c1d..a72ca3c 100644 (file)
--- a/src/broadcom/vulkan/v3dvx_pipeline.c
+++ b/src/broadcom/vulkan/v3dvx_pipeline.c
@@ -452,14 +452,25 @@ pack_shader_state_record(struct v3dv_pipeline *pipeline)
          prog_data_vs_bin->separate_segments;
       shader.vertex_shader_has_separate_input_and_output_vpm_blocks =
          prog_data_vs->separate_segments;
-#endif
-
       shader.coordinate_shader_input_vpm_segment_size =
          prog_data_vs_bin->separate_segments ?
          prog_data_vs_bin->vpm_input_size : 1;
       shader.vertex_shader_input_vpm_segment_size =
          prog_data_vs->separate_segments ?
          prog_data_vs->vpm_input_size : 1;
+#endif
+
+      /* On V3D 7.1 there isn't a specific flag to set if we are using
+       * shared/separate segments or not. We just set the value of
+       * vpm_input_size to 0, and set output to the max needed. That should be
+       * already properly set on prog_data_vs_bin
+       */
+#if V3D_VERSION == 71
+      shader.coordinate_shader_input_vpm_segment_size =
+         prog_data_vs_bin->vpm_input_size;
+      shader.vertex_shader_input_vpm_segment_size =
+         prog_data_vs->vpm_input_size;
+#endif
 
       shader.coordinate_shader_output_vpm_segment_size =
          prog_data_vs_bin->vpm_output_size;