intel: Emit 3DSTATE_VF_STATISTICS dynamically

Pipeline statistics queries should not count BLORP's rectangles.

    (23) How do operations like Clear, TexSubImage, etc. affect the
         results of the newly introduced queries?

      DISCUSSION: Implementations might require "helper" rendering
      commands be issued to implement certain operations like Clear,
      TexSubImage, etc.

      RESOLVED: They don't. Only application submitted rendering
      commands should have an effect on the results of the queries.

Piglit's arb_pipeline_statistics_query-vert_adj exposes this bug when
the driver is hacked to always perform glBufferData via a GPU staging
copy (for debugging purposes).

Reviewed-by: Jason Ekstrand <[email protected]>
Reviewed-by: Lionel Landwerlin <[email protected]>

2 files changed, 24 insertions, 6 deletions

diff --git a/src/mesa/drivers/dri/i965/brw_misc_state.c b/src/mesa/drivers/dri/i965/brw_misc_state.c
index 9bff2c8ac92..b752294250e 100644
--- a/src/mesa/drivers/dri/i965/brw_misc_state.c
+++ b/src/mesa/drivers/dri/i965/brw_misc_state.c
@@ -614,12 +614,6 @@ brw_upload_invariant_state(struct brw_context *brw)
       OUT_BATCH(0);
       ADVANCE_BATCH();
    }
-
-   const uint32_t _3DSTATE_VF_STATISTICS =
-      is_965 ? GEN4_3DSTATE_VF_STATISTICS : GM45_3DSTATE_VF_STATISTICS;
-   BEGIN_BATCH(1);
-   OUT_BATCH(_3DSTATE_VF_STATISTICS << 16 | 1);
-   ADVANCE_BATCH();
 }
 
 /**
diff --git a/src/mesa/drivers/dri/i965/genX_state_upload.c b/src/mesa/drivers/dri/i965/genX_state_upload.c
index e904c4c0cf9..ecffa2e8e86 100644
--- a/src/mesa/drivers/dri/i965/genX_state_upload.c
+++ b/src/mesa/drivers/dri/i965/genX_state_upload.c
@@ -928,6 +928,22 @@ const struct brw_tracked_state genX(cut_index) = {
 };
 #endif
 
+static void
+genX(upload_vf_statistics)(struct brw_context *brw)
+{
+   brw_batch_emit(brw, GENX(3DSTATE_VF_STATISTICS), vf) {
+      vf.StatisticsEnable = true;
+   }
+}
+
+const struct brw_tracked_state genX(vf_statistics) = {
+   .dirty = {
+      .mesa  = 0,
+      .brw   = BRW_NEW_BLORP | BRW_NEW_CONTEXT,
+   },
+   .emit = genX(upload_vf_statistics),
+};
+
 #if GEN_GEN >= 6
 /**
  * Determine the appropriate attribute override value to store into the
@@ -5640,6 +5656,8 @@ genX(init_atoms)(struct brw_context *brw)
 #if GEN_GEN < 6
    static const struct brw_tracked_state *render_atoms[] =
    {
+      &genX(vf_statistics),
+
       /* Once all the programs are done, we know how large urb entry
        * sizes need to be and can decide if we need to change the urb
        * layout.
@@ -5696,6 +5714,8 @@ genX(init_atoms)(struct brw_context *brw)
 #elif GEN_GEN == 6
    static const struct brw_tracked_state *render_atoms[] =
    {
+      &genX(vf_statistics),
+
       &genX(sf_clip_viewport),
 
       /* Command packets: */
@@ -5760,6 +5780,8 @@ genX(init_atoms)(struct brw_context *brw)
 #elif GEN_GEN == 7
    static const struct brw_tracked_state *render_atoms[] =
    {
+      &genX(vf_statistics),
+
       /* Command packets: */
 
       &genX(cc_vp),
@@ -5850,6 +5872,8 @@ genX(init_atoms)(struct brw_context *brw)
 #elif GEN_GEN >= 8
    static const struct brw_tracked_state *render_atoms[] =
    {
+      &genX(vf_statistics),
+
       &genX(cc_vp),
       &genX(sf_clip_viewport),