[NVPTX] Add an NVPTX-specific TargetTransformInfo

[jingyue]

Summary

It currently only implements hasBranchDivergence, and will be extended
in later diffs.

Split from D6188.

[jingyue]

+ jholewinski@nvidia

[jholewinski]

LGTM. Thanks!

[meheff]

Small nit.

lib/Target/NVPTX/NVPTXTargetTransformInfo.cpp

76

this is missing the corresponding "/// @{" doxygen member grouping.

[jingyue]

Fixed by removing the closing "/// @}". Thanks!

lib/Target/NVPTX/CMakeLists.txt

 set(LLVM_TARGET_DEFINITIONS NVPTX.td)
 
 
 tablegen(LLVM NVPTXGenRegisterInfo.inc -gen-register-info)
 tablegen(LLVM NVPTXGenInstrInfo.inc -gen-instr-info)
 tablegen(LLVM NVPTXGenAsmWriter.inc -gen-asm-writer)
 tablegen(LLVM NVPTXGenDAGISel.inc -gen-dag-isel)
 tablegen(LLVM NVPTXGenSubtargetInfo.inc -gen-subtarget)
 add_public_tablegen_target(NVPTXCommonTableGen)
 
 set(NVPTXCodeGen_sources
+  NVPTXAllocaHoisting.cpp
+  NVPTXAsmPrinter.cpp
+  NVPTXAssignValidGlobalNames.cpp
   NVPTXFavorNonGenericAddrSpaces.cpp
   NVPTXFrameLowering.cpp
-  NVPTXInstrInfo.cpp
+  NVPTXGenericToNVVM.cpp
   NVPTXISelDAGToDAG.cpp
   NVPTXISelLowering.cpp
+  NVPTXImageOptimizer.cpp
+  NVPTXInstrInfo.cpp
+  NVPTXLowerAggrCopies.cpp
+  NVPTXLowerStructArgs.cpp
+  NVPTXMCExpr.cpp
+  NVPTXPrologEpilogPass.cpp
   NVPTXRegisterInfo.cpp
+  NVPTXReplaceImageHandles.cpp
   NVPTXSubtarget.cpp
   NVPTXTargetMachine.cpp
-  NVPTXLowerAggrCopies.cpp
-  NVPTXutil.cpp
-  NVPTXAllocaHoisting.cpp
-  NVPTXAsmPrinter.cpp
+  NVPTXTargetTransformInfo.cpp
   NVPTXUtilities.cpp
+  NVPTXutil.cpp
   NVVMReflect.cpp
-  NVPTXGenericToNVVM.cpp
-  NVPTXAssignValidGlobalNames.cpp
-  NVPTXPrologEpilogPass.cpp
-  NVPTXMCExpr.cpp
-  NVPTXReplaceImageHandles.cpp
-  NVPTXImageOptimizer.cpp
-  NVPTXLowerStructArgs.cpp
   )
 
 add_llvm_target(NVPTXCodeGen ${NVPTXCodeGen_sources})
 
 add_subdirectory(TargetInfo)
 add_subdirectory(InstPrinter)
 add_subdirectory(MCTargetDesc)

lib/Target/NVPTX/NVPTX.h

inline static const char *NVPTXCondCodeToString(NVPTXCC::CondCodes CC) {
   case NVPTXCC::GT:
     return "gt";
   case NVPTXCC::GE:
     return "ge";
   }
   llvm_unreachable("Unknown condition code");
 }
 
+ImmutablePass *createNVPTXTargetTransformInfoPass(const NVPTXTargetMachine *TM);
 FunctionPass *
 createNVPTXISelDag(NVPTXTargetMachine &TM, llvm::CodeGenOpt::Level OptLevel);
 ModulePass *createNVPTXAssignValidGlobalNamesPass();
 ModulePass *createGenericToNVVMPass();
 FunctionPass *createNVPTXFavorNonGenericAddrSpacesPass();
 ModulePass *createNVVMReflectPass();
 ModulePass *createNVVMReflectPass(const StringMap<int>& Mapping);
 MachineFunctionPass *createNVPTXPrologEpilogPass();

lib/Target/NVPTX/NVPTXTargetMachine.h

public:
   TargetPassConfig *createPassConfig(PassManagerBase &PM) override;
 
   // Emission of machine code through MCJIT is not supported.
   bool addPassesToEmitMC(PassManagerBase &, MCContext *&, raw_ostream &,
                          bool = true) override {
     return true;
   }
 
+  /// \brief Register NVPTX analysis passes with a pass manager.
+  void addAnalysisPasses(PassManagerBase &PM) override;
+
 }; // NVPTXTargetMachine.
 
 class NVPTXTargetMachine32 : public NVPTXTargetMachine {
   virtual void anchor();
 public:
   NVPTXTargetMachine32(const Target &T, StringRef TT, StringRef CPU,
                        StringRef FS, const TargetOptions &Options,
                        Reloc::Model RM, CodeModel::Model CM,

lib/Target/NVPTX/NVPTXTargetMachine.cpp

 };
 } // end anonymous namespace
 
 TargetPassConfig *NVPTXTargetMachine::createPassConfig(PassManagerBase &PM) {
   NVPTXPassConfig *PassConfig = new NVPTXPassConfig(this, PM);
   return PassConfig;
 }
 
+void NVPTXTargetMachine::addAnalysisPasses(PassManagerBase &PM) {
+  // Add first the target-independent BasicTTI pass, then our NVPTX pass. This
+  // allows the NVPTX pass to delegate to the target independent layer when
+  // appropriate.
+  PM.add(createBasicTargetTransformInfoPass(this));
+  PM.add(createNVPTXTargetTransformInfoPass(this));
+}
+
 void NVPTXPassConfig::addIRPasses() {
   // The following passes are known to not play well with virtual regs hanging
   // around after register allocation (which in our case, is *all* registers).
   // We explicitly disable them here.  We do, however, need some functionality
   // of the PrologEpilogCodeInserter pass, so we emulate that behavior in the
   // NVPTXPrologEpilog pass (see NVPTXPrologEpilogPass.cpp).
   disablePass(&PrologEpilogCodeInserterID);
   disablePass(&MachineCopyPropagationID);

lib/Target/NVPTX/NVPTXTargetTransformInfo.cpp

+//===-- NVPTXTargetTransformInfo.cpp - NVPTX specific TTI pass ---------===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+// \file
+// This file implements a TargetTransformInfo analysis pass specific to the
+// NVPTX target machine. It uses the target's detailed information to provide
+// more precise answers to certain TTI queries, while letting the target
+// independent and default TTI implementations handle the rest.
+//
+//===----------------------------------------------------------------------===//
+
+#include "NVPTXTargetMachine.h"
+#include "llvm/Analysis/LoopInfo.h"
+#include "llvm/Analysis/TargetTransformInfo.h"
+#include "llvm/Analysis/ValueTracking.h"
+#include "llvm/Support/Debug.h"
+#include "llvm/Target/CostTable.h"
+#include "llvm/Target/TargetLowering.h"
+using namespace llvm;
+
+#define DEBUG_TYPE "NVPTXtti"
+
+// Declare the pass initialization routine locally as target-specific passes
+// don't have a target-wide initialization entry point, and so we rely on the
+// pass constructor initialization.
+namespace llvm {
+void initializeNVPTXTTIPass(PassRegistry &);
+}
+
+namespace {
+
+class NVPTXTTI final : public ImmutablePass, public TargetTransformInfo {
+  const NVPTXTargetMachine *TM;
+  const NVPTXSubtarget *ST;
+  const NVPTXTargetLowering *TLI;
+
+  /// Estimate the overhead of scalarizing an instruction. Insert and Extract
+  /// are set if the result needs to be inserted and/or extracted from vectors.
+  unsigned getScalarizationOverhead(Type *Ty, bool Insert, bool Extract) const;
+
+public:
+  NVPTXTTI() : ImmutablePass(ID), TM(nullptr), ST(nullptr), TLI(nullptr) {
+    llvm_unreachable("This pass cannot be directly constructed");
+  }
+
+  NVPTXTTI(const NVPTXTargetMachine *TM)
+      : ImmutablePass(ID), TM(TM), ST(TM->getSubtargetImpl()),
+        TLI(TM->getSubtargetImpl()->getTargetLowering()) {
+    initializeNVPTXTTIPass(*PassRegistry::getPassRegistry());
+  }
+
+  void initializePass() override { pushTTIStack(this); }
+
+  void getAnalysisUsage(AnalysisUsage &AU) const override {
+    TargetTransformInfo::getAnalysisUsage(AU);
+  }
+
+  /// Pass identification.
+  static char ID;
+
+  /// Provide necessary pointer adjustments for the two base classes.
+  void *getAdjustedAnalysisPointer(const void *ID) override {
+    if (ID == &TargetTransformInfo::ID)
+      return (TargetTransformInfo *)this;
+    return this;
+  }
+
+  bool hasBranchDivergence() const override;
+
+  /// @}
+};
+
+} // end anonymous namespace
+
+INITIALIZE_AG_PASS(NVPTXTTI, TargetTransformInfo, "NVPTXtti",
+                   "NVPTX Target Transform Info", true, true, false)
+char NVPTXTTI::ID = 0;
+
+ImmutablePass *
+llvm::createNVPTXTargetTransformInfoPass(const NVPTXTargetMachine *TM) {
+  return new NVPTXTTI(TM);
+}
+
+bool NVPTXTTI::hasBranchDivergence() const { return true; }