[ADMGPU] SDWA peephole optimization pass.

Summary:
First iteration of SDWA peephole.

This pass tries to combine several instruction into one SDWA instruction. E.g. it converts:
'''
    V_LSHRREV_B32_e32 %vreg0, 16, %vreg1
    V_ADD_I32_e32 %vreg2, %vreg0, %vreg3
    V_LSHLREV_B32_e32 %vreg4, 16, %vreg2
'''
Into:
'''
   V_ADD_I32_sdwa %vreg4, %vreg1, %vreg3 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:DWORD
'''

Pass structure:
    1. Iterate over machine instruction in basic block and try to apply "SDWA patterns" to each of them. SDWA patterns match machine instruction into either source or destination SDWA operand. E.g. ''' V_LSHRREV_B32_e32 %vreg0, 16, %vreg1''' is matched to source SDWA operand '''%vreg1 src_sel:WORD_1'''.
    2. Iterate over found SDWA operands and find instruction that could be potentially coverted into SDWA. E.g. for source SDWA operand potential instruction are all instruction in this basic block that uses '''%vreg0'''
    3. Iterate over all potential instructions and check if they can be converted into SDWA.
    4. Convert instructions to SDWA.

This review contains basic implementation of SDWA peephole pass. This pass requires additional testing fot both correctness and performance (no performance testing done).
There are several ways this pass can be improved:
    1. Make this pass work on whole function not only basic block. As I can see this can be done right now without changes to pass.
    2. Introduce more SDWA patterns
    3. Introduce mnemonics to limit when SDWA patterns should apply

Reviewers: vpykhtin, alex-t, arsenm, rampitec

Subscribers: wdng, nhaehnle, mgorny

Differential Revision: https://reviews.llvm.org/D30038

llvm-svn: 298365

llvm/lib/Target/AMDGPU/AMDGPU.h

@@ -37,6 +37,7 @@ FunctionPass *createAMDGPUCFGStructurizerPass();
 FunctionPass *createSITypeRewriter();
 FunctionPass *createSIAnnotateControlFlowPass();
 FunctionPass *createSIFoldOperandsPass();
+FunctionPass *createSIPeepholeSDWAPass();
 FunctionPass *createSILowerI1CopiesPass();
 FunctionPass *createSIShrinkInstructionsPass();
 FunctionPass *createSILoadStoreOptimizerPass(TargetMachine &tm);
@@ -58,6 +59,9 @@ extern char &AMDGPULowerIntrinsicsID;
 void initializeSIFoldOperandsPass(PassRegistry &);
 extern char &SIFoldOperandsID;
 
+void initializeSIPeepholeSDWAPass(PassRegistry &);
+extern char &SIPeepholeSDWAID;
+
 void initializeSIShrinkInstructionsPass(PassRegistry&);
 extern char &SIShrinkInstructionsID;
 

llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp

@@ -94,6 +94,11 @@ static cl::opt<bool> InternalizeSymbols(
   cl::init(false),
   cl::Hidden);
 
+static cl::opt<bool> EnableSDWAPeephole(
+  "amdgpu-sdwa-peephole",
+  cl::desc("Enable SDWA peepholer"),
+  cl::init(false));
+
 // Enable address space based alias analysis
 static cl::opt<bool> EnableAMDGPUAliasAnalysis("enable-amdgpu-aa", cl::Hidden,
   cl::desc("Enable AMDGPU Alias Analysis"),
@@ -109,6 +114,7 @@ extern "C" void LLVMInitializeAMDGPUTarget() {
   initializeSIFixSGPRCopiesPass(*PR);
   initializeSIFixVGPRCopiesPass(*PR);
   initializeSIFoldOperandsPass(*PR);
+  initializeSIPeepholeSDWAPass(*PR);
   initializeSIShrinkInstructionsPass(*PR);
   initializeSIFixControlFlowLiveIntervalsPass(*PR);
   initializeSILoadStoreOptimizerPass(*PR);
@@ -683,6 +689,10 @@ bool GCNPassConfig::addGlobalInstructionSelect() {
 
 void GCNPassConfig::addPreRegAlloc() {
   addPass(createSIShrinkInstructionsPass());
+  if (EnableSDWAPeephole) {
+    addPass(&SIPeepholeSDWAID);
+    addPass(&DeadMachineInstructionElimID);
+  }
   addPass(createSIWholeQuadModePass());
 }
 

llvm/lib/Target/AMDGPU/CMakeLists.txt

@@ -89,6 +89,7 @@ add_llvm_target(AMDGPUCodeGen
   SIMachineFunctionInfo.cpp
   SIMachineScheduler.cpp
   SIOptimizeExecMasking.cpp
+  SIPeepholeSDWA.cpp
   SIRegisterInfo.cpp
   SIShrinkInstructions.cpp
   SITypeRewriter.cpp

llvm/lib/Target/AMDGPU/SIInstrInfo.h

@@ -768,6 +768,9 @@ namespace AMDGPU {
   LLVM_READONLY
   int getVOPe32(uint16_t Opcode);
 
+  LLVM_READONLY
+  int getSDWAOp(uint16_t Opcode);
+
   LLVM_READONLY
   int getCommuteRev(uint16_t Opcode);
 

llvm/lib/Target/AMDGPU/SIInstrInfo.td

@@ -1441,6 +1441,15 @@ def getVOPe32 : InstrMapping {
   let ValueCols = [["4", "0"]];
 }
 
+// Maps ordinary instructions to their SDWA counterparts
+def getSDWAOp : InstrMapping {
+  let FilterClass = "VOP";
+  let RowFields = ["OpName"];
+  let ColFields = ["AsmVariantName"];
+  let KeyCol = ["Default"];
+  let ValueCols = [["SDWA"]];
+}
+
 def getMaskedMIMGOp : InstrMapping {
   let FilterClass = "MIMG_Mask";
   let RowFields = ["Op"];