[llvm] [X86] X86FixupInstTuning - prefer VPBLENDD to VPBLENDW shuffles on AVX2+ targets (PR #144269)

[Phoebe Wang via llvm-commits]

================
@@ -270,6 +290,13 @@ bool X86FixupInstTuningPass::processInstruction(
     return ProcessBLENDToMOV(X86::VMOVSSrr, 0xF, 0x1) ||
            ProcessBLENDToMOV(X86::VMOVSDrr, 0xF, 0x3);
 
+  case X86::VPBLENDWrri:
+    // TODO: Add X86::VPBLENDWrmi handling
+    return ProcessBLENDWToBLENDD(X86::VPBLENDDrri, 4);
+  case X86::VPBLENDWYrri:
+    // TODO: Add X86::VPBLENDWYrmi handling
+    return ProcessBLENDWToBLENDD(X86::VPBLENDDYrmi, 8);
----------------
phoebewang wrote:

I don't see a test for it either.

https://github.com/llvm/llvm-project/pull/144269