MIR Serialization: Serialize global address machine operands.

This commit serializes the global address machine operands.
This commit doesn't serialize the operand's offset and target
flags, it serializes only the global value reference.

Reviewers: Duncan P. N. Exon Smith

Differential Revision: http://reviews.llvm.org/D10671


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@240851 91177308-0d34-0410-b5e6-96231b3b80d8

lib/CodeGen/MIRParser/MILexer.cpp

@@ -106,6 +106,25 @@ static Cursor lexPercent(Cursor C, MIToken &Token) {
   return C;
 }
 
+static Cursor lexGlobalValue(Cursor C, MIToken &Token) {
+  auto Range = C;
+  C.advance(); // Skip the '@'
+  // TODO: add support for quoted names.
+  if (!isdigit(C.peek())) {
+    while (isIdentifierChar(C.peek()))
+      C.advance();
+    Token = MIToken(MIToken::NamedGlobalValue, Range.upto(C),
+                    /*StringOffset=*/1); // Drop the '@'
+    return C;
+  }
+  auto NumberRange = C;
+  while (isdigit(C.peek()))
+    C.advance();
+  Token =
+      MIToken(MIToken::GlobalValue, Range.upto(C), APSInt(NumberRange.upto(C)));
+  return C;
+}
+
 static Cursor lexIntegerLiteral(Cursor C, MIToken &Token) {
   auto Range = C;
   C.advance();
@@ -151,6 +170,8 @@ StringRef llvm::lexMIToken(
       return lexMachineBasicBlock(C, Token, ErrorCallback).remaining();
     return lexPercent(C, Token).remaining();
   }
+  if (Char == '@')
+    return lexGlobalValue(C, Token).remaining();
   if (isdigit(Char) || (Char == '-' && isdigit(C.peek(1))))
     return lexIntegerLiteral(C, Token).remaining();
   MIToken::TokenKind Kind = symbolToken(Char);

lib/CodeGen/MIRParser/MILexer.h

@@ -40,6 +40,8 @@ struct MIToken {
     Identifier,
     NamedRegister,
     MachineBasicBlock,
+    NamedGlobalValue,
+    GlobalValue,
 
     // Other tokens
     IntegerLiteral
@@ -78,7 +80,8 @@ public:
   const APSInt &integerValue() const { return IntVal; }
 
   bool hasIntegerValue() const {
-    return Kind == IntegerLiteral || Kind == MachineBasicBlock;
+    return Kind == IntegerLiteral || Kind == MachineBasicBlock ||
+           Kind == GlobalValue;
   }
 };
 

lib/CodeGen/MIRParser/MIParser.cpp

@@ -14,9 +14,11 @@
 #include "MIParser.h"
 #include "MILexer.h"
 #include "llvm/ADT/StringMap.h"
+#include "llvm/AsmParser/SlotMapping.h"
 #include "llvm/CodeGen/MachineBasicBlock.h"
 #include "llvm/CodeGen/MachineFunction.h"
 #include "llvm/CodeGen/MachineInstr.h"
+#include "llvm/IR/Module.h"
 #include "llvm/Support/raw_ostream.h"
 #include "llvm/Support/SourceMgr.h"
 #include "llvm/Target/TargetSubtargetInfo.h"
@@ -34,6 +36,8 @@ class MIParser {
   MIToken Token;
   /// Maps from basic block numbers to MBBs.
   const DenseMap<unsigned, MachineBasicBlock *> &MBBSlots;
+  /// Maps from indices to unnamed global values and metadata nodes.
+  const SlotMapping &IRSlots;
   /// Maps from instruction names to op codes.
   StringMap<unsigned> Names2InstrOpCodes;
   /// Maps from register names to registers.
@@ -42,7 +46,8 @@ class MIParser {
 public:
   MIParser(SourceMgr &SM, MachineFunction &MF, SMDiagnostic &Error,
            StringRef Source,
-           const DenseMap<unsigned, MachineBasicBlock *> &MBBSlots);
+           const DenseMap<unsigned, MachineBasicBlock *> &MBBSlots,
+           const SlotMapping &IRSlots);
 
   void lex();
 
@@ -62,6 +67,7 @@ public:
   bool parseRegisterOperand(MachineOperand &Dest, bool IsDef = false);
   bool parseImmediateOperand(MachineOperand &Dest);
   bool parseMBBOperand(MachineOperand &Dest);
+  bool parseGlobalAddressOperand(MachineOperand &Dest);
   bool parseMachineOperand(MachineOperand &Dest);
 
 private:
@@ -89,9 +95,11 @@ private:
 
 MIParser::MIParser(SourceMgr &SM, MachineFunction &MF, SMDiagnostic &Error,
                    StringRef Source,
-                   const DenseMap<unsigned, MachineBasicBlock *> &MBBSlots)
+                   const DenseMap<unsigned, MachineBasicBlock *> &MBBSlots,
+                   const SlotMapping &IRSlots)
     : SM(SM), MF(MF), Error(Error), Source(Source), CurrentSource(Source),
-      Token(MIToken::Error, StringRef()), MBBSlots(MBBSlots) {}
+      Token(MIToken::Error, StringRef()), MBBSlots(MBBSlots), IRSlots(IRSlots) {
+}
 
 void MIParser::lex() {
   CurrentSource = lexMIToken(
@@ -250,6 +258,36 @@ bool MIParser::parseMBBOperand(MachineOperand &Dest) {
   return false;
 }
 
+bool MIParser::parseGlobalAddressOperand(MachineOperand &Dest) {
+  switch (Token.kind()) {
+  case MIToken::NamedGlobalValue: {
+    auto Name = Token.stringValue();
+    const Module *M = MF.getFunction()->getParent();
+    if (const auto *GV = M->getNamedValue(Name)) {
+      Dest = MachineOperand::CreateGA(GV, /*Offset=*/0);
+      break;
+    }
+    return error(Twine("use of undefined global value '@") + Name + "'");
+  }
+  case MIToken::GlobalValue: {
+    unsigned GVIdx;
+    if (getUnsigned(GVIdx))
+      return true;
+    if (GVIdx >= IRSlots.GlobalValues.size())
+      return error(Twine("use of undefined global value '@") + Twine(GVIdx) +
+                   "'");
+    Dest = MachineOperand::CreateGA(IRSlots.GlobalValues[GVIdx],
+                                    /*Offset=*/0);
+    break;
+  }
+  default:
+    llvm_unreachable("The current token should be a global value");
+  }
+  // TODO: Parse offset and target flags.
+  lex();
+  return false;
+}
+
 bool MIParser::parseMachineOperand(MachineOperand &Dest) {
   switch (Token.kind()) {
   case MIToken::underscore:
@@ -259,6 +297,9 @@ bool MIParser::parseMachineOperand(MachineOperand &Dest) {
     return parseImmediateOperand(Dest);
   case MIToken::MachineBasicBlock:
     return parseMBBOperand(Dest);
+  case MIToken::GlobalValue:
+  case MIToken::NamedGlobalValue:
+    return parseGlobalAddressOperand(Dest);
   case MIToken::Error:
     return true;
   default:
@@ -314,6 +355,6 @@ bool MIParser::getRegisterByName(StringRef RegName, unsigned &Reg) {
 MachineInstr *
 llvm::parseMachineInstr(SourceMgr &SM, MachineFunction &MF, StringRef Src,
                         const DenseMap<unsigned, MachineBasicBlock *> &MBBSlots,
-                        SMDiagnostic &Error) {
+                        const SlotMapping &IRSlots, SMDiagnostic &Error) {
-  return MIParser(SM, MF, Error, Src, MBBSlots).parse();
+  return MIParser(SM, MF, Error, Src, MBBSlots, IRSlots).parse();
 }

lib/CodeGen/MIRParser/MIParser.h

@@ -22,13 +22,14 @@ namespace llvm {
 class MachineBasicBlock;
 class MachineInstr;
 class MachineFunction;
+struct SlotMapping;
 class SMDiagnostic;
 class SourceMgr;
 
 MachineInstr *
 parseMachineInstr(SourceMgr &SM, MachineFunction &MF, StringRef Src,
                   const DenseMap<unsigned, MachineBasicBlock *> &MBBSlots,
-                  SMDiagnostic &Error);
+                  const SlotMapping &IRSlots, SMDiagnostic &Error);
 
 } // end namespace llvm
 

lib/CodeGen/MIRParser/MIRParser.cpp

@@ -19,6 +19,7 @@
 #include "llvm/ADT/StringMap.h"
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/AsmParser/Parser.h"
+#include "llvm/AsmParser/SlotMapping.h"
 #include "llvm/CodeGen/MachineFunction.h"
 #include "llvm/CodeGen/MachineRegisterInfo.h"
 #include "llvm/CodeGen/MIRYamlMapping.h"
@@ -46,6 +47,7 @@ class MIRParserImpl {
   StringRef Filename;
   LLVMContext &Context;
   StringMap<std::unique_ptr<yaml::MachineFunction>> Functions;
+  SlotMapping IRSlots;
 
 public:
   MIRParserImpl(std::unique_ptr<MemoryBuffer> Contents, StringRef Filename,
@@ -157,7 +159,7 @@ std::unique_ptr<Module> MIRParserImpl::parse() {
           dyn_cast_or_null<yaml::BlockScalarNode>(In.getCurrentNode())) {
     SMDiagnostic Error;
     M = parseAssembly(MemoryBufferRef(BSN->getValue(), Filename), Error,
-                      Context);
+                      Context, &IRSlots);
     if (!M) {
       reportDiagnostic(diagFromLLVMAssemblyDiag(Error, BSN->getSourceRange()));
       return M;
@@ -263,7 +265,8 @@ bool MIRParserImpl::initializeMachineBasicBlock(
   // Parse the instructions.
   for (const auto &MISource : YamlMBB.Instructions) {
     SMDiagnostic Error;
-    if (auto *MI = parseMachineInstr(SM, MF, MISource.Value, MBBSlots, Error)) {
+    if (auto *MI = parseMachineInstr(SM, MF, MISource.Value, MBBSlots, IRSlots,
+                                     Error)) {
       MBB.insert(MBB.end(), MI);
       continue;
     }

lib/CodeGen/MIRPrinter.cpp

@@ -40,16 +40,18 @@ public:
   void print(const MachineFunction &MF);
 
   void convert(yaml::MachineFunction &MF, const MachineRegisterInfo &RegInfo);
-  void convert(yaml::MachineBasicBlock &YamlMBB, const MachineBasicBlock &MBB);
+  void convert(const Module &M, yaml::MachineBasicBlock &YamlMBB,
+               const MachineBasicBlock &MBB);
 };
 
 /// This class prints out the machine instructions using the MIR serialization
 /// format.
 class MIPrinter {
+  const Module &M;
   raw_ostream &OS;
 
 public:
-  MIPrinter(raw_ostream &OS) : OS(OS) {}
+  MIPrinter(const Module &M, raw_ostream &OS) : M(M), OS(OS) {}
 
   void print(const MachineInstr &MI);
   void print(const MachineOperand &Op, const TargetRegisterInfo *TRI);
@@ -83,6 +85,7 @@ void MIRPrinter::print(const MachineFunction &MF) {
   convert(YamlMF, MF.getRegInfo());
 
   int I = 0;
+  const auto &M = *MF.getFunction()->getParent();
   for (const auto &MBB : MF) {
     // TODO: Allow printing of non sequentially numbered MBBs.
     // This is currently needed as the basic block references get their index
@@ -92,7 +95,7 @@ void MIRPrinter::print(const MachineFunction &MF) {
            "Can't print MBBs that aren't sequentially numbered");
     (void)I;
     yaml::MachineBasicBlock YamlMBB;
-    convert(YamlMBB, MBB);
+    convert(M, YamlMBB, MBB);
     YamlMF.BasicBlocks.push_back(YamlMBB);
   }
   yaml::Output Out(OS);
@@ -106,7 +109,7 @@ void MIRPrinter::convert(yaml::MachineFunction &MF,
   MF.TracksSubRegLiveness = RegInfo.subRegLivenessEnabled();
 }
 
-void MIRPrinter::convert(yaml::MachineBasicBlock &YamlMBB,
+void MIRPrinter::convert(const Module &M, yaml::MachineBasicBlock &YamlMBB,
                          const MachineBasicBlock &MBB) {
   assert(MBB.getNumber() >= 0 && "Invalid MBB number");
   YamlMBB.ID = (unsigned)MBB.getNumber();
@@ -124,7 +127,7 @@ void MIRPrinter::convert(yaml::MachineBasicBlock &YamlMBB,
   std::string Str;
   for (const auto &MI : MBB) {
     raw_string_ostream StrOS(Str);
-    MIPrinter(StrOS).print(MI);
+    MIPrinter(M, StrOS).print(MI);
     YamlMBB.Instructions.push_back(StrOS.str());
     Str.clear();
   }
@@ -191,6 +194,13 @@ void MIPrinter::print(const MachineOperand &Op, const TargetRegisterInfo *TRI) {
         OS << '.' << BB->getName();
     }
     break;
+  case MachineOperand::MO_GlobalAddress:
+    // FIXME: Make this faster - print as operand will create a slot tracker to
+    // print unnamed values for the whole module every time it's called, which
+    // is inefficient.
+    Op.getGlobal()->printAsOperand(OS, /*PrintType=*/false, &M);
+    // TODO: Print offset and target flags.
+    break;
   default:
     // TODO: Print the other machine operands.
     llvm_unreachable("Can't print this machine operand at the moment");

test/CodeGen/MIR/X86/global-value-operands.mir

@@ -0,0 +1,49 @@
+# RUN: llc -march=x86-64 -start-after branch-folder -stop-after branch-folder -o /dev/null %s | FileCheck %s
+# This test ensures that the MIR parser parses global value operands correctly.
+
+--- |
+
+  @G = external global i32
+  @0 = external global i32
+
+  define i32 @inc() {
+  entry:
+    %a = load i32, i32* @G
+    %b = add i32 %a, 1
+    ret i32 %b
+  }
+
+  define i32 @inc2() {
+  entry:
+    %a = load i32, i32* @0
+    %b = add i32 %a, 1
+    ret i32 %b
+  }
+
+...
+---
+# CHECK: name: inc
+name: inc
+body:
+  - id: 0
+    name: entry
+    instructions:
+      # CHECK: - '%rax = MOV64rm %rip, 1, _, @G, _'
+      - '%rax = MOV64rm %rip, 1, _, @G, _'
+      - '%eax = MOV32rm %rax, 1, _, 0, _'
+      - '%eax = INC32r %eax'
+      - 'RETQ %eax'
+...
+---
+# CHECK: name: inc2
+name: inc2
+body:
+  - id: 0
+    name: entry
+    instructions:
+      # CHECK: - '%rax = MOV64rm %rip, 1, _, @0, _'
+      - '%rax = MOV64rm %rip, 1, _, @0, _'
+      - '%eax = MOV32rm %rax, 1, _, 0, _'
+      - '%eax = INC32r %eax'
+      - 'RETQ %eax'
+...

test/CodeGen/MIR/X86/undefined-global-value.mir

@@ -0,0 +1,28 @@
+# RUN: not llc -march=x86-64 -start-after branch-folder -stop-after branch-folder -o /dev/null %s 2>&1 | FileCheck %s
+# This test ensures that an error is reported when an invalid global value index
+# is used.
+
+--- |
+
+  @0 = external global i32
+
+  define i32 @inc() {
+  entry:
+    %a = load i32, i32* @0
+    %b = add i32 %a, 1
+    ret i32 %b
+  }
+
+...
+---
+name: inc
+body:
+  - id: 0
+    name: entry
+    instructions:
+      # CHECK: [[@LINE+1]]:37: use of undefined global value '@2'
+      - '%rax = MOV64rm %rip, 1, _, @2, _'
+      - '%eax = MOV32rm %rax, 1, _, 0, _'
+      - '%eax = INC32r %eax'
+      - 'RETQ %eax'
+...

test/CodeGen/MIR/X86/undefined-named-global-value.mir

@@ -0,0 +1,28 @@
+# RUN: not llc -march=x86-64 -start-after branch-folder -stop-after branch-folder -o /dev/null %s 2>&1 | FileCheck %s
+# This test ensures that an error is reported when an undefined global value is
+# used.
+
+--- |
+
+  @G = external global i32
+
+  define i32 @inc() {
+  entry:
+    %a = load i32, i32* @G
+    %b = add i32 %a, 1
+    ret i32 %b
+  }
+
+...
+---
+name: inc
+body:
+  - id: 0
+    name: entry
+    instructions:
+      # CHECK: [[@LINE+1]]:37: use of undefined global value '@GG'
+      - '%rax = MOV64rm %rip, 1, _, @GG, _'
+      - '%eax = MOV32rm %rax, 1, _, 0, _'
+      - '%eax = INC32r %eax'
+      - 'RETQ %eax'
+...