Migrate the NVPTX backend asm printer to a per function subtarget.

This involved moving two non-subtarget dependent features (64-bitness
and the driver interface) to the NVPTX target machine and updating
the uses (or migrating around the subtarget use for ease of review).
Otherwise use the cached subtarget or create a default subtarget
based on the TargetMachine cpu and feature string for the module
level assembler emission.

llvm-svn: 229785

lib/Target/NVPTX/NVPTXAsmPrinter.cpp

@@ -164,7 +164,7 @@ void NVPTXAsmPrinter::emitLineNumberAsDotLoc(const MachineInstr &MI) {
 void NVPTXAsmPrinter::EmitInstruction(const MachineInstr *MI) {
   SmallString<128> Str;
   raw_svector_ostream OS(Str);
-  if (nvptxSubtarget.getDrvInterface() == NVPTX::CUDA)
+  if (nvptxSubtarget->getDrvInterface() == NVPTX::CUDA)
     emitLineNumberAsDotLoc(*MI);
 
   MCInst Inst;
@@ -237,8 +237,6 @@ void NVPTXAsmPrinter::lowerImageHandleSymbol(unsigned Index, MCOperand &MCOp) {
 
 void NVPTXAsmPrinter::lowerToMCInst(const MachineInstr *MI, MCInst &OutMI) {
   OutMI.setOpcode(MI->getOpcode());
-  const NVPTXSubtarget &ST = TM.getSubtarget<NVPTXSubtarget>();
-
   // Special: Do not mangle symbol operand of CALL_PROTOTYPE
   if (MI->getOpcode() == NVPTX::CALL_PROTOTYPE) {
     const MachineOperand &MO = MI->getOperand(0);
@@ -251,7 +249,7 @@ void NVPTXAsmPrinter::lowerToMCInst(const MachineInstr *MI, MCInst &OutMI) {
     const MachineOperand &MO = MI->getOperand(i);
 
     MCOperand MCOp;
-    if (!ST.hasImageHandles()) {
+    if (!nvptxSubtarget->hasImageHandles()) {
       if (lowerImageHandleOperand(MI, i, MCOp)) {
         OutMI.addOperand(MCOp);
         continue;
@@ -349,11 +347,11 @@ MCOperand NVPTXAsmPrinter::GetSymbolRef(const MCSymbol *Symbol) {
 
 void NVPTXAsmPrinter::printReturnValStr(const Function *F, raw_ostream &O) {
   const DataLayout *TD = TM.getDataLayout();
-  const TargetLowering *TLI = TM.getSubtargetImpl()->getTargetLowering();
+  const TargetLowering *TLI = nvptxSubtarget->getTargetLowering();
 
   Type *Ty = F->getReturnType();
 
-  bool isABI = (nvptxSubtarget.getSmVersion() >= 20);
+  bool isABI = (nvptxSubtarget->getSmVersion() >= 20);
 
   if (Ty->getTypeID() == Type::VoidTyID)
     return;
@@ -506,14 +504,13 @@ void NVPTXAsmPrinter::EmitFunctionBodyEnd() {
 
 void NVPTXAsmPrinter::emitImplicitDef(const MachineInstr *MI) const {
   unsigned RegNo = MI->getOperand(0).getReg();
-  const TargetRegisterInfo *TRI = TM.getSubtargetImpl()->getRegisterInfo();
+  const TargetRegisterInfo *TRI = nvptxSubtarget->getRegisterInfo();
   if (TRI->isVirtualRegister(RegNo)) {
     OutStreamer.AddComment(Twine("implicit-def: ") +
                            getVirtualRegisterName(RegNo));
   } else {
-    OutStreamer.AddComment(
+    OutStreamer.AddComment(Twine("implicit-def: ") +
-        Twine("implicit-def: ") +
+                           nvptxSubtarget->getRegisterInfo()->getName(RegNo));
-        TM.getSubtargetImpl()->getRegisterInfo()->getName(RegNo));
   }
   OutStreamer.AddBlankLine();
 }
@@ -815,6 +812,14 @@ void NVPTXAsmPrinter::recordAndEmitFilenames(Module &M) {
 }
 
 bool NVPTXAsmPrinter::doInitialization(Module &M) {
+  // Construct a default subtarget off of the TargetMachine defaults. The
+  // rest of NVPTX isn't friendly to change subtargets per function and
+  // so the default TargetMachine will have all of the options.
+  StringRef TT = TM.getTargetTriple();
+  StringRef CPU = TM.getTargetCPU();
+  StringRef FS = TM.getTargetFeatureString();
+  const NVPTXTargetMachine &NTM = static_cast<const NVPTXTargetMachine &>(TM);
+  const NVPTXSubtarget STI(TT, CPU, FS, NTM, NTM.is64Bit());
 
   SmallString<128> Str1;
   raw_svector_ostream OS1(Str1);
@@ -832,7 +837,7 @@ bool NVPTXAsmPrinter::doInitialization(Module &M) {
   Mang = new Mangler(TM.getDataLayout());
 
   // Emit header before any dwarf directives are emitted below.
-  emitHeader(M, OS1);
+  emitHeader(M, OS1, STI);
   OutStreamer.EmitRawText(OS1.str());
 
   // Already commented out
@@ -848,7 +853,8 @@ bool NVPTXAsmPrinter::doInitialization(Module &M) {
     OutStreamer.AddBlankLine();
   }
 
-  if (nvptxSubtarget.getDrvInterface() == NVPTX::CUDA)
+  // If we're not NVCL we're CUDA, go ahead and emit filenames.
+  if (Triple(TM.getTargetTriple()).getOS() != Triple::NVCL)
     recordAndEmitFilenames(M);
 
   GlobalsEmitted = false;
@@ -889,22 +895,23 @@ void NVPTXAsmPrinter::emitGlobals(const Module &M) {
   OutStreamer.EmitRawText(OS2.str());
 }
 
-void NVPTXAsmPrinter::emitHeader(Module &M, raw_ostream &O) {
+void NVPTXAsmPrinter::emitHeader(Module &M, raw_ostream &O,
+                                 const NVPTXSubtarget &STI) {
   O << "//\n";
   O << "// Generated by LLVM NVPTX Back-End\n";
   O << "//\n";
   O << "\n";
 
-  unsigned PTXVersion = nvptxSubtarget.getPTXVersion();
+  unsigned PTXVersion = STI.getPTXVersion();
   O << ".version " << (PTXVersion / 10) << "." << (PTXVersion % 10) << "\n";
 
   O << ".target ";
-  O << nvptxSubtarget.getTargetName();
+  O << STI.getTargetName();
 
-  if (nvptxSubtarget.getDrvInterface() == NVPTX::NVCL)
+  if (STI.getDrvInterface() == NVPTX::NVCL)
     O << ", texmode_independent";
-  if (nvptxSubtarget.getDrvInterface() == NVPTX::CUDA) {
+  if (STI.getDrvInterface() == NVPTX::CUDA) {
-    if (!nvptxSubtarget.hasDouble())
+    if (!STI.hasDouble())
       O << ", map_f64_to_f32";
   }
 
@@ -914,7 +921,7 @@ void NVPTXAsmPrinter::emitHeader(Module &M, raw_ostream &O) {
   O << "\n";
 
   O << ".address_size ";
-  if (nvptxSubtarget.is64Bit())
+  if (static_cast<const NVPTXTargetMachine &>(TM).is64Bit())
     O << "64";
   else
     O << "32";
@@ -924,7 +931,6 @@ void NVPTXAsmPrinter::emitHeader(Module &M, raw_ostream &O) {
 }
 
 bool NVPTXAsmPrinter::doFinalization(Module &M) {
-
   // If we did not emit any functions, then the global declarations have not
   // yet been emitted.
   if (!GlobalsEmitted) {
@@ -986,7 +992,7 @@ bool NVPTXAsmPrinter::doFinalization(Module &M) {
 
 void NVPTXAsmPrinter::emitLinkageDirective(const GlobalValue *V,
                                            raw_ostream &O) {
-  if (nvptxSubtarget.getDrvInterface() == NVPTX::CUDA) {
+  if (nvptxSubtarget->getDrvInterface() == NVPTX::CUDA) {
     if (V->hasExternalLinkage()) {
       if (isa<GlobalVariable>(V)) {
         const GlobalVariable *GVar = cast<GlobalVariable>(V);
@@ -1218,7 +1224,7 @@ void NVPTXAsmPrinter::printModuleLevelGV(const GlobalVariable *GVar,
           AggBuffer aggBuffer(ElementSize, O, *this);
           bufferAggregateConstant(Initializer, &aggBuffer);
           if (aggBuffer.numSymbols) {
-            if (nvptxSubtarget.is64Bit()) {
+            if (static_cast<const NVPTXTargetMachine &>(TM).is64Bit()) {
               O << " .u64 " << *getSymbol(GVar) << "[";
               O << ElementSize / 8;
             } else {
@@ -1316,7 +1322,7 @@ NVPTXAsmPrinter::getPTXFundamentalTypeStr(const Type *Ty, bool useB4PTR) const {
   case Type::DoubleTyID:
     return "f64";
   case Type::PointerTyID:
-    if (nvptxSubtarget.is64Bit())
+    if (static_cast<const NVPTXTargetMachine &>(TM).is64Bit())
       if (useB4PTR)
         return "b64";
       else
@@ -1407,8 +1413,8 @@ static unsigned int getOpenCLAlignment(const DataLayout *TD, Type *Ty) {
 
 void NVPTXAsmPrinter::printParamName(Function::const_arg_iterator I,
                                      int paramIndex, raw_ostream &O) {
-  if ((nvptxSubtarget.getDrvInterface() == NVPTX::NVCL) ||
+  if ((nvptxSubtarget->getDrvInterface() == NVPTX::NVCL) ||
-      (nvptxSubtarget.getDrvInterface() == NVPTX::CUDA))
+      (nvptxSubtarget->getDrvInterface() == NVPTX::CUDA))
     O << *getSymbol(I->getParent()) << "_param_" << paramIndex;
   else {
     std::string argName = I->getName();
@@ -1427,8 +1433,8 @@ void NVPTXAsmPrinter::printParamName(int paramIndex, raw_ostream &O) {
   Function::const_arg_iterator I, E;
   int i = 0;
 
-  if ((nvptxSubtarget.getDrvInterface() == NVPTX::NVCL) ||
+  if ((nvptxSubtarget->getDrvInterface() == NVPTX::NVCL) ||
-      (nvptxSubtarget.getDrvInterface() == NVPTX::CUDA)) {
+      (nvptxSubtarget->getDrvInterface() == NVPTX::CUDA)) {
     O << *CurrentFnSym << "_param_" << paramIndex;
     return;
   }
@@ -1445,12 +1451,12 @@ void NVPTXAsmPrinter::printParamName(int paramIndex, raw_ostream &O) {
 void NVPTXAsmPrinter::emitFunctionParamList(const Function *F, raw_ostream &O) {
   const DataLayout *TD = TM.getDataLayout();
   const AttributeSet &PAL = F->getAttributes();
-  const TargetLowering *TLI = TM.getSubtargetImpl()->getTargetLowering();
+  const TargetLowering *TLI = nvptxSubtarget->getTargetLowering();
   Function::const_arg_iterator I, E;
   unsigned paramIndex = 0;
   bool first = true;
   bool isKernelFunc = llvm::isKernelFunction(*F);
-  bool isABI = (nvptxSubtarget.getSmVersion() >= 20);
+  bool isABI = (nvptxSubtarget->getSmVersion() >= 20);
   MVT thePointerTy = TLI->getPointerTy();
 
   O << "(\n";
@@ -1469,21 +1475,21 @@ void NVPTXAsmPrinter::emitFunctionParamList(const Function *F, raw_ostream &O) {
         if (isImage(*I)) {
           std::string sname = I->getName();
           if (isImageWriteOnly(*I) || isImageReadWrite(*I)) {
-            if (nvptxSubtarget.hasImageHandles())
+            if (nvptxSubtarget->hasImageHandles())
               O << "\t.param .u64 .ptr .surfref ";
             else
               O << "\t.param .surfref ";
             O << *CurrentFnSym << "_param_" << paramIndex;
           }
           else { // Default image is read_only
-            if (nvptxSubtarget.hasImageHandles())
+            if (nvptxSubtarget->hasImageHandles())
               O << "\t.param .u64 .ptr .texref ";
             else
               O << "\t.param .texref ";
             O << *CurrentFnSym << "_param_" << paramIndex;
           }
         } else {
-          if (nvptxSubtarget.hasImageHandles())
+          if (nvptxSubtarget->hasImageHandles())
             O << "\t.param .u64 .ptr .samplerref ";
           else
             O << "\t.param .samplerref ";
@@ -1516,7 +1522,7 @@ void NVPTXAsmPrinter::emitFunctionParamList(const Function *F, raw_ostream &O) {
           // Special handling for pointer arguments to kernel
           O << "\t.param .u" << thePointerTy.getSizeInBits() << " ";
 
-          if (nvptxSubtarget.getDrvInterface() != NVPTX::CUDA) {
+          if (nvptxSubtarget->getDrvInterface() != NVPTX::CUDA) {
             Type *ETy = PTy->getElementType();
             int addrSpace = PTy->getAddressSpace();
             switch (addrSpace) {
@@ -1645,7 +1651,7 @@ void NVPTXAsmPrinter::setAndEmitFunctionVirtualRegisters(
   if (NumBytes) {
     O << "\t.local .align " << MFI->getMaxAlignment() << " .b8 \t" << DEPOTNAME
       << getFunctionNumber() << "[" << NumBytes << "];\n";
-    if (nvptxSubtarget.is64Bit()) {
+    if (nvptxSubtarget->is64Bit()) {
       O << "\t.reg .b64 \t%SP;\n";
       O << "\t.reg .b64 \t%SPL;\n";
     } else {

lib/Target/NVPTX/NVPTXAsmPrinter.h

@@ -138,7 +138,7 @@ class LLVM_LIBRARY_VISIBILITY NVPTXAsmPrinter : public AsmPrinter {
         unsigned int nSym = 0;
         unsigned int nextSymbolPos = symbolPosInBuffer[nSym];
         unsigned int nBytes = 4;
-        if (AP.nvptxSubtarget.is64Bit())
+        if (static_cast<const NVPTXTargetMachine &>(AP.TM).is64Bit())
           nBytes = 8;
         for (pos = 0; pos < size; pos += nBytes) {
           if (pos)
@@ -212,7 +212,7 @@ private:
   void printParamName(Function::const_arg_iterator I, int paramIndex,
                       raw_ostream &O);
   void emitGlobals(const Module &M);
-  void emitHeader(Module &M, raw_ostream &O);
+  void emitHeader(Module &M, raw_ostream &O, const NVPTXSubtarget &STI);
   void emitKernelFunctionDirectives(const Function &F, raw_ostream &O) const;
   void emitVirtualRegister(unsigned int vr, raw_ostream &);
   void emitFunctionExternParamList(const MachineFunction &MF);
@@ -248,8 +248,10 @@ private:
   typedef DenseMap<unsigned, unsigned> VRegMap;
   typedef DenseMap<const TargetRegisterClass *, VRegMap> VRegRCMap;
   VRegRCMap VRegMapping;
-  // cache the subtarget here.
+
-  const NVPTXSubtarget &nvptxSubtarget;
+  // Cache the subtarget here.
+  const NVPTXSubtarget *nvptxSubtarget;
+
   // Build the map between type name and ID based on module's type
   // symbol table.
   std::map<const Type *, std::string> TypeNameMap;
@@ -303,10 +305,10 @@ private:
 public:
   NVPTXAsmPrinter(TargetMachine &TM, std::unique_ptr<MCStreamer> Streamer)
       : AsmPrinter(TM, std::move(Streamer)),
-        nvptxSubtarget(TM.getSubtarget<NVPTXSubtarget>()) {
+        EmitGeneric(static_cast<NVPTXTargetMachine &>(TM).getDrvInterface() ==
+                    NVPTX::CUDA) {
     CurrentBankselLabelInBasicBlock = "";
     reader = nullptr;
-    EmitGeneric = (nvptxSubtarget.getDrvInterface() == NVPTX::CUDA);
   }
 
   ~NVPTXAsmPrinter() {
@@ -314,6 +316,10 @@ public:
       delete reader;
   }
 
+  bool runOnMachineFunction(MachineFunction &F) override {
+    nvptxSubtarget = &F.getSubtarget<NVPTXSubtarget>();
+    return AsmPrinter::runOnMachineFunction(F);
+  }
   void getAnalysisUsage(AnalysisUsage &AU) const override {
     AU.addRequired<MachineLoopInfo>();
     AsmPrinter::getAnalysisUsage(AU);

lib/Target/NVPTX/NVPTXSubtarget.cpp

@@ -12,6 +12,7 @@
 //===----------------------------------------------------------------------===//
 
 #include "NVPTXSubtarget.h"
+#include "NVPTXTargetMachine.h"
 
 using namespace llvm;
 
@@ -43,17 +44,13 @@ NVPTXSubtarget &NVPTXSubtarget::initializeSubtargetDependencies(StringRef CPU,
 }
 
 NVPTXSubtarget::NVPTXSubtarget(const std::string &TT, const std::string &CPU,
-                               const std::string &FS, const TargetMachine &TM,
+                               const std::string &FS,
-                               bool is64Bit)
+                               const NVPTXTargetMachine &TM, bool is64Bit)
     : NVPTXGenSubtargetInfo(TT, CPU, FS), Is64Bit(is64Bit), PTXVersion(0),
-      SmVersion(20), InstrInfo(initializeSubtargetDependencies(CPU, FS)),
+      SmVersion(20), TM(TM),
-      TLInfo((const NVPTXTargetMachine &)TM, *this), TSInfo(TM.getDataLayout()),
+      InstrInfo(initializeSubtargetDependencies(CPU, FS)), TLInfo(TM, *this),
-      FrameLowering(*this) {
+      TSInfo(TM.getDataLayout()), FrameLowering(*this) {}
 
-  Triple T(TT);
+NVPTX::DrvInterface NVPTXSubtarget::getDrvInterface() const {
-
+  return TM.getDrvInterface();
-  if (T.getOS() == Triple::NVCL)
-    drvInterface = NVPTX::NVCL;
-  else
-    drvInterface = NVPTX::CUDA;
 }

lib/Target/NVPTX/NVPTXSubtarget.h

@@ -32,7 +32,6 @@ namespace llvm {
 class NVPTXSubtarget : public NVPTXGenSubtargetInfo {
   virtual void anchor();
   std::string TargetName;
-  NVPTX::DrvInterface drvInterface;
   bool Is64Bit;
 
   // PTX version x.y is represented as 10*x+y, e.g. 3.1 == 31
@@ -41,6 +40,7 @@ class NVPTXSubtarget : public NVPTXGenSubtargetInfo {
   // SM version x.y is represented as 10*x+y, e.g. 3.1 == 31
   unsigned int SmVersion;
 
+  const NVPTXTargetMachine &TM;
   NVPTXInstrInfo InstrInfo;
   NVPTXTargetLowering TLInfo;
   TargetSelectionDAGInfo TSInfo;
@@ -54,7 +54,8 @@ public:
   /// of the specified module.
   ///
   NVPTXSubtarget(const std::string &TT, const std::string &CPU,
-                 const std::string &FS, const TargetMachine &TM, bool is64Bit);
+                 const std::string &FS, const NVPTXTargetMachine &TM,
+                 bool is64Bit);
 
   const TargetFrameLowering *getFrameLowering() const override {
     return &FrameLowering;
@@ -106,7 +107,7 @@ public:
   bool is64Bit() const { return Is64Bit; }
 
   unsigned int getSmVersion() const { return SmVersion; }
-  NVPTX::DrvInterface getDrvInterface() const { return drvInterface; }
+  NVPTX::DrvInterface getDrvInterface() const;
   std::string getTargetName() const { return TargetName; }
 
   unsigned getPTXVersion() const { return PTXVersion; }

lib/Target/NVPTX/NVPTXTargetMachine.cpp

@@ -86,10 +86,13 @@ NVPTXTargetMachine::NVPTXTargetMachine(const Target &T, StringRef TT,
                                        const TargetOptions &Options,
                                        Reloc::Model RM, CodeModel::Model CM,
                                        CodeGenOpt::Level OL, bool is64bit)
-    : LLVMTargetMachine(T, TT, CPU, FS, Options, RM, CM, OL),
+    : LLVMTargetMachine(T, TT, CPU, FS, Options, RM, CM, OL), is64bit(is64bit),
       TLOF(make_unique<NVPTXTargetObjectFile>()),
-      DL(computeDataLayout(is64bit)),
+      DL(computeDataLayout(is64bit)), Subtarget(TT, CPU, FS, *this, is64bit) {
-      Subtarget(TT, CPU, FS, *this, is64bit) {
+  if (Triple(TT).getOS() == Triple::NVCL)
+    drvInterface = NVPTX::NVCL;
+  else
+    drvInterface = NVPTX::CUDA;
   initAsmInfo();
 }
 

lib/Target/NVPTX/NVPTXTargetMachine.h

@@ -25,8 +25,10 @@ namespace llvm {
 /// NVPTXTargetMachine
 ///
 class NVPTXTargetMachine : public LLVMTargetMachine {
+  bool is64bit;
   std::unique_ptr<TargetLoweringObjectFile> TLOF;
   const DataLayout DL; // Calculates type size & alignment
+  NVPTX::DrvInterface drvInterface;
   NVPTXSubtarget Subtarget;
 
   // Hold Strings that can be free'd all together with NVPTXTargetMachine
@@ -40,7 +42,8 @@ public:
   ~NVPTXTargetMachine() override;
   const DataLayout *getDataLayout() const override { return &DL; }
   const NVPTXSubtarget *getSubtargetImpl() const override { return &Subtarget; }
-
+  bool is64Bit() const { return is64bit; }
+  NVPTX::DrvInterface getDrvInterface() const { return drvInterface; }
   ManagedStringPool *getManagedStrPool() const {
     return const_cast<ManagedStringPool *>(&ManagedStrPool);
   }