[AMDGPU] Disable VALU sinking and hoisting with WWM

Machine LICM can hoist a VALU instruction from a WWM region.
In this case WQM pass will have to create yet another WWM region
around the hoisted instruction, which is not desired.

Unfortunatelly we cannot tell if an instruction is in the WWM
region, so this patch disables hoisting if WWM is used in the
function.

This works around the bug SWDEV-502411.

Author: rampitec

llvm/lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp

@@ -2773,6 +2773,9 @@ void AMDGPUDAGToDAGISel::SelectINTRINSIC_WO_CHAIN(SDNode *N) {
   case Intrinsic::amdgcn_wwm:
   case Intrinsic::amdgcn_strict_wwm:
     Opcode = AMDGPU::STRICT_WWM;
+    CurDAG->getMachineFunction()
+        .getInfo<SIMachineFunctionInfo>()
+        ->setInitWholeWave();
     break;
   case Intrinsic::amdgcn_strict_wqm:
     Opcode = AMDGPU::STRICT_WQM;

llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp

@@ -1055,8 +1055,12 @@ bool AMDGPUInstructionSelector::selectG_INTRINSIC(MachineInstr &I) const {
   case Intrinsic::amdgcn_softwqm:
     return constrainCopyLikeIntrin(I, AMDGPU::SOFT_WQM);
   case Intrinsic::amdgcn_strict_wwm:
-  case Intrinsic::amdgcn_wwm:
+  case Intrinsic::amdgcn_wwm: {
+    MachineFunction *MF = I.getParent()->getParent();
+    SIMachineFunctionInfo *MFInfo = MF->getInfo<SIMachineFunctionInfo>();
+    MFInfo->setInitWholeWave();
     return constrainCopyLikeIntrin(I, AMDGPU::STRICT_WWM);
+  }
   case Intrinsic::amdgcn_strict_wqm:
     return constrainCopyLikeIntrin(I, AMDGPU::STRICT_WQM);
   case Intrinsic::amdgcn_writelane:

llvm/lib/Target/AMDGPU/SIInstrInfo.cpp

@@ -184,7 +184,11 @@ static bool resultDependsOnExec(const MachineInstr &MI) {
 bool SIInstrInfo::isIgnorableUse(const MachineOperand &MO) const {
   // Any implicit use of exec by VALU is not a real register read.
   return MO.getReg() == AMDGPU::EXEC && MO.isImplicit() &&
-         isVALU(*MO.getParent()) && !resultDependsOnExec(*MO.getParent());
+         isVALU(*MO.getParent()) && !resultDependsOnExec(*MO.getParent()) &&
+         !MO.getParent()
+              ->getMF()
+              ->getInfo<SIMachineFunctionInfo>()
+              ->hasInitWholeWave();
 }
 
 bool SIInstrInfo::isSafeToSink(MachineInstr &MI,

llvm/test/CodeGen/AMDGPU/atomic_optimizations_global_pointer.ll

@@ -8,31 +8,33 @@ define i32 @test(i32 %val, i32 %cond) {
 ; GCN-NEXT:    s_xor_saveexec_b32 s4, -1
 ; GCN-NEXT:    buffer_store_dword v2, off, s[0:3], s32 ; 4-byte Folded Spill
 ; GCN-NEXT:    buffer_store_dword v3, off, s[0:3], s32 offset:4 ; 4-byte Folded Spill
+; GCN-NEXT:    buffer_store_dword v4, off, s[0:3], s32 offset:8 ; 4-byte Folded Spill
 ; GCN-NEXT:    s_waitcnt_depctr 0xffe3
 ; GCN-NEXT:    s_mov_b32 exec_lo, s4
 ; GCN-NEXT:    s_or_saveexec_b32 s4, -1
 ; GCN-NEXT:    v_mov_b32_e32 v2, 0
 ; GCN-NEXT:    v_cndmask_b32_e64 v3, 0, v0, s4
-; GCN-NEXT:    v_mov_b32_dpp v2, v3 row_xmask:1 row_mask:0xf bank_mask:0xf
-; GCN-NEXT:    s_mov_b32 exec_lo, s4
-; GCN-NEXT:    v_mov_b32_e32 v5, 0
 ; GCN-NEXT:    v_mov_b32_e32 v4, v2
+; GCN-NEXT:    v_mov_b32_dpp v4, v3 row_xmask:1 row_mask:0xf bank_mask:0xf
+; GCN-NEXT:    s_mov_b32 exec_lo, s4
+; GCN-NEXT:    v_mov_b32_e32 v5, v4
 ; GCN-NEXT:    v_cmp_eq_u32_e32 vcc_lo, 0, v1
 ; GCN-NEXT:    s_and_saveexec_b32 s4, vcc_lo
 ; GCN-NEXT:  ; %bb.1: ; %if
 ; GCN-NEXT:    s_or_saveexec_b32 s5, -1
-; GCN-NEXT:    v_mov_b32_e32 v2, 0
-; GCN-NEXT:    v_cndmask_b32_e64 v3, 0, v0, s5
-; GCN-NEXT:    v_mov_b32_dpp v2, v3 row_xmask:1 row_mask:0xf bank_mask:0xf
+; GCN-NEXT:    v_mov_b32_e32 v3, 0
+; GCN-NEXT:    v_cndmask_b32_e64 v2, 0, v0, s5
+; GCN-NEXT:    v_mov_b32_dpp v3, v2 row_xmask:1 row_mask:0xf bank_mask:0xf
 ; GCN-NEXT:    s_mov_b32 exec_lo, s5
-; GCN-NEXT:    v_mov_b32_e32 v5, v2
+; GCN-NEXT:    v_mov_b32_e32 v2, v3
 ; GCN-NEXT:  ; %bb.2: ; %end
 ; GCN-NEXT:    s_or_b32 exec_lo, exec_lo, s4
-; GCN-NEXT:    v_add_nc_u32_e32 v0, v4, v5
+; GCN-NEXT:    v_add_nc_u32_e32 v0, v5, v2
 ; GCN-NEXT:    s_xor_saveexec_b32 s4, -1
-; GCN-NEXT:    s_clause 0x1
+; GCN-NEXT:    s_clause 0x2
 ; GCN-NEXT:    buffer_load_dword v2, off, s[0:3], s32
 ; GCN-NEXT:    buffer_load_dword v3, off, s[0:3], s32 offset:4
+; GCN-NEXT:    buffer_load_dword v4, off, s[0:3], s32 offset:8
 ; GCN-NEXT:    s_waitcnt_depctr 0xffe3
 ; GCN-NEXT:    s_mov_b32 exec_lo, s4
 ; GCN-NEXT:    s_waitcnt vmcnt(0)