[Mesa-dev] [PATCH 28/31] i965/fs: Remove manual splitting of DDY ops in the generator.

[Francisco Jerez]

---
 src/mesa/drivers/dri/i965/brw_fs_generator.cpp | 50 +++-----------------------
 1 file changed, 4 insertions(+), 46 deletions(-)

diff --git a/src/mesa/drivers/dri/i965/brw_fs_generator.cpp b/src/mesa/drivers/dri/i965/brw_fs_generator.cpp
index 82543d4..c3050f9 100644
--- a/src/mesa/drivers/dri/i965/brw_fs_generator.cpp
+++ b/src/mesa/drivers/dri/i965/brw_fs_generator.cpp
@@ -1027,34 +1027,6 @@ fs_generator::generate_ddy(enum opcode opcode,
                            bool negate_value)
 {
    if (opcode == FS_OPCODE_DDY_FINE) {
-      /* From the Ivy Bridge PRM, volume 4 part 3, section 3.3.9 (Register
-       * Region Restrictions):
-       *
-       *     In Align16 access mode, SIMD16 is not allowed for DW operations
-       *     and SIMD8 is not allowed for DF operations.
-       *
-       * In this context, "DW operations" means "operations acting on 32-bit
-       * values", so it includes operations on floats.
-       *
-       * Gen4 has a similar restriction.  From the i965 PRM, section 11.5.3
-       * (Instruction Compression -> Rules and Restrictions):
-       *
-       *     A compressed instruction must be in Align1 access mode. Align16
-       *     mode instructions cannot be compressed.
-       *
-       * Similar text exists in the g45 PRM.
-       *
-       * Empirically, compressed align16 instructions using odd register
-       * numbers don't appear to work on Sandybridge either.
-       *
-       * On these platforms, if we're building a SIMD16 shader, we need to
-       * manually unroll to a pair of SIMD8 instructions.
-       */
-      bool unroll_to_simd8 =
-         (dispatch_width == 16 &&
-          (devinfo->gen == 4 || devinfo->gen == 6 ||
-           (devinfo->gen == 7 && !devinfo->is_haswell)));
-
       /* produce accurate derivatives */
       struct brw_reg src0 = brw_reg(src.file, src.nr, 0,
                                     src.negate, src.abs,
@@ -1072,24 +1044,10 @@ fs_generator::generate_ddy(enum opcode opcode,
                                     BRW_SWIZZLE_ZWZW, WRITEMASK_XYZW);
       brw_push_insn_state(p);
       brw_set_default_access_mode(p, BRW_ALIGN_16);
-      if (unroll_to_simd8) {
-         brw_set_default_exec_size(p, BRW_EXECUTE_8);
-         brw_set_default_compression_control(p, BRW_COMPRESSION_NONE);
-         if (negate_value) {
-            brw_ADD(p, firsthalf(dst), firsthalf(src1), negate(firsthalf(src0)));
-            brw_set_default_compression_control(p, BRW_COMPRESSION_2NDHALF);
-            brw_ADD(p, sechalf(dst), sechalf(src1), negate(sechalf(src0)));
-         } else {
-            brw_ADD(p, firsthalf(dst), firsthalf(src0), negate(firsthalf(src1)));
-            brw_set_default_compression_control(p, BRW_COMPRESSION_2NDHALF);
-            brw_ADD(p, sechalf(dst), sechalf(src0), negate(sechalf(src1)));
-         }
-      } else {
-         if (negate_value)
-            brw_ADD(p, dst, src1, negate(src0));
-         else
-            brw_ADD(p, dst, src0, negate(src1));
-      }
+      if (negate_value)
+         brw_ADD(p, dst, src1, negate(src0));
+      else
+         brw_ADD(p, dst, src0, negate(src1));
       brw_pop_insn_state(p);
    } else {
       /* replicate the derivative at the top-left pixel to other pixels */
-- 
2.7.3