radeonsi: fix a GS hang on VI

Broken by one of the cleanups: 0d46c3bc9d09b376d74f7399e1a2d1b0a923640b
Not applicable to stable.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>

diff --git a/src/gallium/drivers/radeonsi/si_pipe.h b/src/gallium/drivers/radeonsi/si_pipe.h
index 3ff4b46..2abd5b5 100644 (file)
--- a/src/gallium/drivers/radeonsi/si_pipe.h
+++ b/src/gallium/drivers/radeonsi/si_pipe.h
@@ -192,6 +192,7 @@ struct si_context {
 
        /* Precomputed states. */
        struct si_pm4_state             *init_config;
+       bool                            init_config_has_vgt_flush;
        struct si_pm4_state             *vgt_shader_config[4];
        /* With rasterizer discard, there doesn't have to be a pixel shader.
         * In that case, we bind this one: */

diff --git a/src/gallium/drivers/radeonsi/si_state_shaders.c b/src/gallium/drivers/radeonsi/si_state_shaders.c
index 77c585f..f673388 100644 (file)
--- a/src/gallium/drivers/radeonsi/si_state_shaders.c
+++ b/src/gallium/drivers/radeonsi/si_state_shaders.c
@@ -1136,6 +1136,20 @@ static void si_emit_spi_ps_input(struct si_context *sctx, struct r600_atom *atom
                                  sctx->force_persample_interp);
 }
 
+/**
+ * Writing CONFIG or UCONFIG VGT registers requires VGT_FLUSH before that.
+ */
+static void si_init_config_add_vgt_flush(struct si_context *sctx)
+{
+       if (sctx->init_config_has_vgt_flush)
+               return;
+
+       si_pm4_cmd_begin(sctx->init_config, PKT3_EVENT_WRITE);
+       si_pm4_cmd_add(sctx->init_config, EVENT_TYPE(V_028A90_VGT_FLUSH) | EVENT_INDEX(0));
+       si_pm4_cmd_end(sctx->init_config, false);
+       sctx->init_config_has_vgt_flush = true;
+}
+
 /* Initialize state related to ESGS / GSVS ring buffers */
 static void si_init_gs_rings(struct si_context *sctx)
 {
@@ -1156,6 +1170,8 @@ static void si_init_gs_rings(struct si_context *sctx)
                return;
        }
 
+       si_init_config_add_vgt_flush(sctx);
+
        /* Append these registers to the init config state. */
        if (sctx->b.chip_class >= CIK) {
                if (sctx->b.chip_class >= VI) {
@@ -1402,6 +1418,8 @@ static void si_init_tess_factor_ring(struct si_context *sctx)
 
        assert(((sctx->tf_ring->width0 / 4) & C_030938_SIZE) == 0);
 
+       si_init_config_add_vgt_flush(sctx);
+
        /* Append these registers to the init config state. */
        if (sctx->b.chip_class >= CIK) {
                si_pm4_set_reg(sctx->init_config, R_030938_VGT_TF_RING_SIZE,