i965: Add a debug flag for counting cycles spent in each compiled shader.

This can be used for two purposes: Using hand-coded shaders to determine
per-instruction timings, or figuring out which shader to optimize in a
whole application.

Note that this doesn't cover the instructions that set up the message to
the URB/FB write -- we'd need to convert the MRF usage in these
instructions to GRFs so that our offsets/times don't overwrite our
shader outputs.

Reviewed-by: Kenneth Graunke <[email protected]> (v1)

v2: Check the timestamp reset flag in the VS, which is apparently
    getting set fairly regularly in the range we watch, resulting in
    negative numbers getting added to our 32-bit counter, and thus large
    values added to our uint64_t.
v3: Rebase on reladdr changes, removing a new safety check that proved
    impossible to satisfy.  Add a comment to the AOP defs from Ken's
    review, and put them in a slightly more sensible spot.
v4: Check timestamp reset in the FS as well.

1 files changed, 24 insertions, 4 deletions

diff --git a/src/mesa/drivers/dri/i965/brw_context.h b/src/mesa/drivers/dri/i965/brw_context.h
index 1abaee3db1d..dc25cabcf53 100644
--- a/src/mesa/drivers/dri/i965/brw_context.h
+++ b/src/mesa/drivers/dri/i965/brw_context.h
@@ -559,14 +559,15 @@ struct brw_vs_prog_data {
 #define SURF_INDEX_FRAG_CONST_BUFFER (BRW_MAX_DRAW_BUFFERS + 1)
 #define SURF_INDEX_TEXTURE(t)        (BRW_MAX_DRAW_BUFFERS + 2 + (t))
 #define SURF_INDEX_WM_UBO(u)         (SURF_INDEX_TEXTURE(BRW_MAX_TEX_UNIT) + u)
-
+#define SURF_INDEX_WM_SHADER_TIME    (SURF_INDEX_WM_UBO(12))
 /** Maximum size of the binding table. */
-#define BRW_MAX_WM_SURFACES          SURF_INDEX_WM_UBO(BRW_MAX_WM_UBOS)
+#define BRW_MAX_WM_SURFACES          (SURF_INDEX_WM_SHADER_TIME + 1)
 
 #define SURF_INDEX_VERT_CONST_BUFFER (0)
 #define SURF_INDEX_VS_TEXTURE(t)     (SURF_INDEX_VERT_CONST_BUFFER + 1 + (t))
 #define SURF_INDEX_VS_UBO(u)         (SURF_INDEX_VS_TEXTURE(BRW_MAX_TEX_UNIT) + u)
-#define BRW_MAX_VS_SURFACES          SURF_INDEX_VS_UBO(BRW_MAX_VS_UBOS)
+#define SURF_INDEX_VS_SHADER_TIME    (SURF_INDEX_VS_UBO(12))
+#define BRW_MAX_VS_SURFACES          (SURF_INDEX_VS_SHADER_TIME + 1)
 
 #define SURF_INDEX_SOL_BINDING(t)    ((t))
 #define BRW_MAX_GS_SURFACES          SURF_INDEX_SOL_BINDING(BRW_MAX_SOL_BINDINGS)
@@ -651,6 +652,13 @@ struct brw_tracked_state {
    void (*emit)( struct brw_context *brw );
 };
 
+enum shader_time_shader_type {
+   ST_NONE,
+   ST_VS,
+   ST_FS8,
+   ST_FS16,
+};
+
 /* Flags for brw->state.cache.
  */
 #define CACHE_NEW_BLEND_STATE            (1<<BRW_BLEND_STATE)
@@ -1089,6 +1097,16 @@ struct brw_context
 
    uint32_t num_instances;
    int basevertex;
+
+   struct {
+      drm_intel_bo *bo;
+      struct gl_shader_program **programs;
+      enum shader_time_shader_type *types;
+      uint64_t *cumulative;
+      int num_entries;
+      int max_entries;
+      double report_time;
+   } shader_time;
 };
 
 /*======================================================================
@@ -1144,7 +1162,9 @@ void brwInitFragProgFuncs( struct dd_function_table *functions );
 int brw_get_scratch_size(int size);
 void brw_get_scratch_bo(struct intel_context *intel,
 			drm_intel_bo **scratch_bo, int size);
-
+void brw_init_shader_time(struct brw_context *brw);
+void brw_collect_and_report_shader_time(struct brw_context *brw);
+void brw_destroy_shader_time(struct brw_context *brw);
 
 /* brw_urb.c
  */