[PATCH] D25998: AMDGPU/SI: Don't use non-0 waitcnt values when waiting on Flat instructions

[Tom Stellard via llvm-commits]

tstellarAMD added inline comments.


================
Comment at: test/CodeGen/MIR/AMDGPU/waitcnt.mir:50
+    %vgpr0 = FLAT_LOAD_DWORD %vgpr1_vgpr2, 0, 0, 0, implicit %exec, implicit %flat_scr
+    %vgpr3_vgpr4_vgpr5_vgpr6 = FLAT_LOAD_DWORDX4 %vgpr7_vgpr8, 0, 0, 0, implicit %exec, implicit %flat_scr :: (load 16 from %ir.global16)
+    %vgpr0 = V_MOV_B32_e32 %vgpr1, implicit %exec
----------------
arsenm wrote:
> This looks like it has a mem operand although the comment on the check line says it doesn't
The first load is the one without the mem operand, I can clarify this in the comment.


https://reviews.llvm.org/D25998