MIR Serialization: Serialize machine instruction names.

This commit implements initial machine instruction serialization. It
serializes machine instruction names. The instructions are represented
using a YAML sequence of string literals and are a part of machine
basic block YAML mapping.

This commit introduces a class called 'MIParser' which will be used to
parse the machine instructions and operands.

Reviewers: Duncan P. N. Exon Smith

Differential Revision: http://reviews.llvm.org/D10481


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@240295 91177308-0d34-0410-b5e6-96231b3b80d8

include/llvm/CodeGen/MIRYamlMapping.h

@@ -22,6 +22,8 @@
 #include "llvm/Support/YAMLTraits.h"
 #include <vector>
 
+LLVM_YAML_IS_SEQUENCE_VECTOR(std::string)
+
 namespace llvm {
 namespace yaml {
 
@@ -31,7 +33,8 @@ struct MachineBasicBlock {
   bool IsLandingPad = false;
   bool AddressTaken = false;
   // TODO: Serialize the successors and liveins.
-  // TODO: Serialize machine instructions.
+
+  std::vector<std::string> Instructions;
 };
 
 template <> struct MappingTraits<MachineBasicBlock> {
@@ -41,6 +44,7 @@ template <> struct MappingTraits<MachineBasicBlock> {
     YamlIO.mapOptional("alignment", MBB.Alignment);
     YamlIO.mapOptional("isLandingPad", MBB.IsLandingPad);
     YamlIO.mapOptional("addressTaken", MBB.AddressTaken);
+    YamlIO.mapOptional("instructions", MBB.Instructions);
   }
 };
 

lib/CodeGen/MIRParser/CMakeLists.txt

@@ -1,4 +1,5 @@
 add_llvm_library(LLVMMIRParser
+  MIParser.cpp
   MIRParser.cpp
   )
 

lib/CodeGen/MIRParser/MIParser.cpp

@@ -0,0 +1,106 @@
+//===- MIParser.cpp - Machine instructions parser implementation ----------===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+// This file implements the parsing of machine instructions.
+//
+//===----------------------------------------------------------------------===//
+
+#include "MIParser.h"
+#include "llvm/ADT/StringMap.h"
+#include "llvm/CodeGen/MachineBasicBlock.h"
+#include "llvm/CodeGen/MachineFunction.h"
+#include "llvm/CodeGen/MachineInstr.h"
+#include "llvm/Support/raw_ostream.h"
+#include "llvm/Support/SourceMgr.h"
+#include "llvm/Target/TargetSubtargetInfo.h"
+#include "llvm/Target/TargetInstrInfo.h"
+
+using namespace llvm;
+
+namespace {
+
+class MIParser {
+  SourceMgr &SM;
+  MachineFunction &MF;
+  SMDiagnostic &Error;
+  StringRef Source;
+  /// Maps from instruction names to op codes.
+  StringMap<unsigned> Names2InstrOpCodes;
+
+public:
+  MIParser(SourceMgr &SM, MachineFunction &MF, SMDiagnostic &Error,
+           StringRef Source);
+
+  /// Report an error at the current location with the given message.
+  ///
+  /// This function always return true.
+  bool error(const Twine &Msg);
+
+  MachineInstr *parse();
+
+private:
+  void initNames2InstrOpCodes();
+
+  /// Try to convert an instruction name to an opcode. Return true if the
+  /// instruction name is invalid.
+  bool parseInstrName(StringRef InstrName, unsigned &OpCode);
+};
+
+} // end anonymous namespace
+
+MIParser::MIParser(SourceMgr &SM, MachineFunction &MF, SMDiagnostic &Error,
+                   StringRef Source)
+    : SM(SM), MF(MF), Error(Error), Source(Source) {}
+
+bool MIParser::error(const Twine &Msg) {
+  // TODO: Get the proper location in the MIR file, not just a location inside
+  // the string.
+  Error =
+      SMDiagnostic(SM, SMLoc(), SM.getMemoryBuffer(SM.getMainFileID())
+                                    ->getBufferIdentifier(),
+                   1, 0, SourceMgr::DK_Error, Msg.str(), Source, None, None);
+  return true;
+}
+
+MachineInstr *MIParser::parse() {
+  StringRef InstrName = Source;
+  unsigned OpCode;
+  if (parseInstrName(InstrName, OpCode)) {
+    error(Twine("unknown machine instruction name '") + InstrName + "'");
+    return nullptr;
+  }
+
+  // TODO: Parse the rest of instruction - machine operands, etc.
+  const auto &MCID = MF.getSubtarget().getInstrInfo()->get(OpCode);
+  auto *MI = MF.CreateMachineInstr(MCID, DebugLoc());
+  return MI;
+}
+
+void MIParser::initNames2InstrOpCodes() {
+  if (!Names2InstrOpCodes.empty())
+    return;
+  const auto *TII = MF.getSubtarget().getInstrInfo();
+  assert(TII && "Expected target instruction info");
+  for (unsigned I = 0, E = TII->getNumOpcodes(); I < E; ++I)
+    Names2InstrOpCodes.insert(std::make_pair(StringRef(TII->getName(I)), I));
+}
+
+bool MIParser::parseInstrName(StringRef InstrName, unsigned &OpCode) {
+  initNames2InstrOpCodes();
+  auto InstrInfo = Names2InstrOpCodes.find(InstrName);
+  if (InstrInfo == Names2InstrOpCodes.end())
+    return true;
+  OpCode = InstrInfo->getValue();
+  return false;
+}
+
+MachineInstr *llvm::parseMachineInstr(SourceMgr &SM, MachineFunction &MF,
+                                      StringRef Src, SMDiagnostic &Error) {
+  return MIParser(SM, MF, Error, Src).parse();
+}

lib/CodeGen/MIRParser/MIParser.h

@@ -0,0 +1,31 @@
+//===- MIParser.h - Machine Instructions Parser ---------------------------===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+// This file declares the function that parses the machine instructions.
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIB_CODEGEN_MIRPARSER_MIPARSER_H
+#define LLVM_LIB_CODEGEN_MIRPARSER_MIPARSER_H
+
+#include "llvm/ADT/StringRef.h"
+
+namespace llvm {
+
+class MachineInstr;
+class MachineFunction;
+class SMDiagnostic;
+class SourceMgr;
+
+MachineInstr *parseMachineInstr(SourceMgr &SM, MachineFunction &MF,
+                                StringRef Src, SMDiagnostic &Error);
+
+} // end namespace llvm
+
+#endif

lib/CodeGen/MIRParser/MIRParser.cpp

@@ -13,6 +13,7 @@
 //===----------------------------------------------------------------------===//
 
 #include "llvm/CodeGen/MIRParser/MIRParser.h"
+#include "MIParser.h"
 #include "llvm/ADT/StringRef.h"
 #include "llvm/ADT/StringMap.h"
 #include "llvm/ADT/STLExtras.h"
@@ -79,7 +80,7 @@ public:
   /// Initialize the machine basic block using it's YAML representation.
   ///
   /// Return true if an error occurred.
-  bool initializeMachineBasicBlock(MachineBasicBlock &MBB,
+  bool initializeMachineBasicBlock(MachineFunction &MF, MachineBasicBlock &MBB,
                                    const yaml::MachineBasicBlock &YamlMBB);
 
 private:
@@ -218,18 +219,29 @@ bool MIRParserImpl::initializeMachineFunction(MachineFunction &MF) {
     }
     auto *MBB = MF.CreateMachineBasicBlock(BB);
     MF.insert(MF.end(), MBB);
-    if (initializeMachineBasicBlock(*MBB, YamlMBB))
+    if (initializeMachineBasicBlock(MF, *MBB, YamlMBB))
       return true;
   }
   return false;
 }
 
 bool MIRParserImpl::initializeMachineBasicBlock(
-    MachineBasicBlock &MBB, const yaml::MachineBasicBlock &YamlMBB) {
+    MachineFunction &MF, MachineBasicBlock &MBB,
+    const yaml::MachineBasicBlock &YamlMBB) {
   MBB.setAlignment(YamlMBB.Alignment);
   if (YamlMBB.AddressTaken)
     MBB.setHasAddressTaken();
   MBB.setIsLandingPad(YamlMBB.IsLandingPad);
+  // Parse the instructions.
+  for (const auto &MISource : YamlMBB.Instructions) {
+    SMDiagnostic Error;
+    if (auto *MI = parseMachineInstr(SM, MF, MISource, Error)) {
+      MBB.insert(MBB.end(), MI);
+      continue;
+    }
+    reportDiagnostic(Error);
+    return true;
+  }
   return false;
 }
 

lib/CodeGen/MIRPrinter.cpp

@@ -21,6 +21,8 @@
 #include "llvm/Support/MemoryBuffer.h"
 #include "llvm/Support/raw_ostream.h"
 #include "llvm/Support/YAMLTraits.h"
+#include "llvm/Target/TargetInstrInfo.h"
+#include "llvm/Target/TargetSubtargetInfo.h"
 
 using namespace llvm;
 
@@ -39,6 +41,17 @@ public:
   void convert(yaml::MachineBasicBlock &YamlMBB, const MachineBasicBlock &MBB);
 };
 
+/// This class prints out the machine instructions using the MIR serialization
+/// format.
+class MIPrinter {
+  raw_ostream &OS;
+
+public:
+  MIPrinter(raw_ostream &OS) : OS(OS) {}
+
+  void print(const MachineInstr &MI);
+};
+
 } // end anonymous namespace
 
 namespace llvm {
@@ -83,6 +96,25 @@ void MIRPrinter::convert(yaml::MachineBasicBlock &YamlMBB,
   YamlMBB.Alignment = MBB.getAlignment();
   YamlMBB.AddressTaken = MBB.hasAddressTaken();
   YamlMBB.IsLandingPad = MBB.isLandingPad();
+
+  // Print the machine instructions.
+  YamlMBB.Instructions.reserve(MBB.size());
+  std::string Str;
+  for (const auto &MI : MBB) {
+    raw_string_ostream StrOS(Str);
+    MIPrinter(StrOS).print(MI);
+    YamlMBB.Instructions.push_back(StrOS.str());
+    Str.clear();
+  }
+}
+
+void MIPrinter::print(const MachineInstr &MI) {
+  const auto &SubTarget = MI.getParent()->getParent()->getSubtarget();
+  const auto *TII = SubTarget.getInstrInfo();
+  assert(TII && "Expected target instruction info");
+
+  OS << TII->getName(MI.getOpcode());
+  // TODO: Print the instruction flags, machine operands, machine mem operands.
 }
 
 void llvm::printMIR(raw_ostream &OS, const Module &M) {

test/CodeGen/MIR/X86/lit.local.cfg

@@ -0,0 +1,2 @@
+if not 'X86' in config.root.targets:
+    config.unsupported = True

test/CodeGen/MIR/X86/machine-instructions.mir

@@ -0,0 +1,24 @@
+# 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 X86 machine instructions
+# correctly.
+
+--- |
+
+  define i32 @inc(i32 %a) {
+  entry:
+    %b = mul i32 %a, 11
+    ret i32 %b
+  }
+
+...
+---
+# CHECK: name: inc
+name:            inc
+body:
+ - name:         entry
+   instructions:
+     # CHECK:      - IMUL32rri8
+     # CHECK-NEXT: - RETQ
+     - IMUL32rri8
+     - RETQ
+...

test/CodeGen/MIR/X86/unknown-instruction.mir

@@ -0,0 +1,20 @@
+# 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 unknown instruction is
+# encountered.
+
+--- |
+
+  define i32 @foo() {
+  entry:
+    ret i32 0
+  }
+
+...
+---
+name:            foo
+body:
+ - name:         entry
+   instructions:
+     # CHECK: 1:1: unknown machine instruction name 'retJust0'
+     - retJust0
+...