[Mesa-dev] [PATCH 5/8] i965/vs: Prepare split_virtual_grfs() for the presence of SENDs from GRFs.

[Eric Anholt]

---
 src/mesa/drivers/dri/i965/brw_vec4.cpp |   49 +++++++++++++++++++++++++-------
 src/mesa/drivers/dri/i965/brw_vec4.h   |    1 +
 2 files changed, 39 insertions(+), 11 deletions(-)

diff --git a/src/mesa/drivers/dri/i965/brw_vec4.cpp b/src/mesa/drivers/dri/i965/brw_vec4.cpp
index 3e628c8..36b5408 100644
--- a/src/mesa/drivers/dri/i965/brw_vec4.cpp
+++ b/src/mesa/drivers/dri/i965/brw_vec4.cpp
@@ -218,6 +218,13 @@ vec4_instruction::is_math()
 	   opcode == SHADER_OPCODE_INT_REMAINDER ||
 	   opcode == SHADER_OPCODE_POW);
 }
+
+bool
+vec4_instruction::is_send_from_grf()
+{
+   return false;
+}
+
 /**
  * Returns how many MRFs an opcode will write over.
  *
@@ -878,27 +885,49 @@ vec4_visitor::opt_register_coalesce()
  *
  * We initially create large virtual GRFs for temporary structures, arrays,
  * and matrices, so that the dereference visitor functions can add reg_offsets
- * to work their way down to the actual member being accessed.
+ * to work their way down to the actual member being accessed.  But when it
+ * comes to optimization, we'd like to treat each register as individual
+ * storage if possible.
  *
- * Unlike in the FS visitor, though, we have no SEND messages that return more
- * than 1 register.  We also don't do any array access in register space,
- * which would have required contiguous physical registers.  Thus, all those
- * large virtual GRFs can be split up into independent single-register virtual
- * GRFs, making allocation and optimization easier.
+ * So far, the only thing that might prevent splitting is a send message from
+ * a GRF on IVB.
  */
 void
 vec4_visitor::split_virtual_grfs()
 {
    int num_vars = this->virtual_grf_count;
    int new_virtual_grf[num_vars];
+   bool split_grf[num_vars];
 
    memset(new_virtual_grf, 0, sizeof(new_virtual_grf));
 
+   /* Try to split anything > 0 sized. */
+   for (int i = 0; i < num_vars; i++) {
+      if (this->virtual_grf_sizes[i] != 1)
+	 split_grf[i] = true;
+      else
+	 split_grf[i] = false;
+   }
+
+   /* Check that the instructions are compatible with the registers we're trying
+    * to split.
+    */
+   foreach_list(node, &this->instructions) {
+      vec4_instruction *inst = (vec4_instruction *)node;
+
+      /* If there's a SEND message loading from a GRF on gen7+, it needs to be
+       * contiguous.  Assume that the GRF for the SEND is always in src[0].
+       */
+      if (inst->is_send_from_grf()) {
+	 split_grf[inst->src[0].reg] = false;
+      }
+   }
+
    /* Allocate new space for split regs.  Note that the virtual
     * numbers will be contiguous.
     */
    for (int i = 0; i < num_vars; i++) {
-      if (this->virtual_grf_sizes[i] == 1)
+      if (!split_grf[i])
          continue;
 
       new_virtual_grf[i] = virtual_grf_alloc(1);
@@ -913,16 +942,14 @@ vec4_visitor::split_virtual_grfs()
    foreach_list(node, &this->instructions) {
       vec4_instruction *inst = (vec4_instruction *)node;
 
-      if (inst->dst.file == GRF &&
-	  new_virtual_grf[inst->dst.reg] &&
+      if (inst->dst.file == GRF && split_grf[inst->dst.reg] &&
 	  inst->dst.reg_offset != 0) {
 	 inst->dst.reg = (new_virtual_grf[inst->dst.reg] +
 			  inst->dst.reg_offset - 1);
 	 inst->dst.reg_offset = 0;
       }
       for (int i = 0; i < 3; i++) {
-	 if (inst->src[i].file == GRF &&
-	     new_virtual_grf[inst->src[i].reg] &&
+	 if (inst->src[i].file == GRF && split_grf[inst->src[i].reg] &&
 	     inst->src[i].reg_offset != 0) {
 	    inst->src[i].reg = (new_virtual_grf[inst->src[i].reg] +
 				inst->src[i].reg_offset - 1);
diff --git a/src/mesa/drivers/dri/i965/brw_vec4.h b/src/mesa/drivers/dri/i965/brw_vec4.h
index 9e700a9..e792ec0 100644
--- a/src/mesa/drivers/dri/i965/brw_vec4.h
+++ b/src/mesa/drivers/dri/i965/brw_vec4.h
@@ -194,6 +194,7 @@ public:
 
    bool is_tex();
    bool is_math();
+   bool is_send_from_grf();
    bool can_reswizzle_dst(int dst_writemask, int swizzle, int swizzle_mask);
    void reswizzle_dst(int dst_writemask, int swizzle);
 };
-- 
1.7.10.4