[mips] Process numeric register name in the .set assignment directive

[atanasyan]

Summary

Now LLVM assembler cannot process the following code and generates an error. GNU tools support .set assignment directive with numeric register name.

.set r4, 4

test.s:1:11: error: invalid token in expression
  .set r4, $4
           ^

This patch teach assembler to handle such directives correctly. Unfortunately a numeric register name cannot be represented as an expression. That's why we have to maintain a separate StringMap in the MipsAsmParser to keep mapping between aliases names and register numbers.

[smaksimovic]

LGTM

[atanasyan]

Thanks for review.

llvm/trunk/lib/Target/Mips/AsmParser/MipsAsmParser.cpp

class MipsAsmParser : public MCTargetAsmParser {
   bool IsLittleEndian;
   bool IsPicEnabled;
   bool IsCpRestoreSet;
   int CpRestoreOffset;
   unsigned CpSaveLocation;
   /// If true, then CpSaveLocation is a register, otherwise it's an offset.
   bool     CpSaveLocationIsRegister;
 
+  // Map of register aliases created via the .set directive.
+  StringMap<AsmToken> RegisterSets;
+
   // Print a warning along with its fix-it message at the given range.
   void printWarningWithFixIt(const Twine &Msg, const Twine &FixMsg,
                              SMRange Range, bool ShowColors = true);
 
   void ConvertXWPOperands(MCInst &Inst, const OperandVector &Operands);
 
 #define GET_ASSEMBLER_HEADER
 #include "MipsGenAsmMatcher.inc"
#include "MipsGenAsmMatcher.inc"
 
   bool ParseDirective(AsmToken DirectiveID) override;
 
   OperandMatchResultTy parseMemOperand(OperandVector &Operands);
   OperandMatchResultTy
   matchAnyRegisterNameWithoutDollar(OperandVector &Operands,
                                     StringRef Identifier, SMLoc S);
   OperandMatchResultTy matchAnyRegisterWithoutDollar(OperandVector &Operands,
+                                                     const AsmToken &Token,
+                                                     SMLoc S);
+  OperandMatchResultTy matchAnyRegisterWithoutDollar(OperandVector &Operands,
                                                      SMLoc S);
   OperandMatchResultTy parseAnyRegister(OperandVector &Operands);
   OperandMatchResultTy parseImm(OperandVector &Operands);
   OperandMatchResultTy parseJumpTarget(OperandVector &Operands);
   OperandMatchResultTy parseInvNum(OperandVector &Operands);
   OperandMatchResultTy parseRegisterPair(OperandVector &Operands);
   OperandMatchResultTy parseMovePRegPair(OperandVector &Operands);
   OperandMatchResultTy parseRegisterList(OperandVector &Operands);
if (Expr->getKind() == MCExpr::SymbolRef) {
         if (ResTy == MatchOperand_Success) {
           Parser.Lex();
           return true;
         }
         if (ResTy == MatchOperand_ParseFail)
           llvm_unreachable("Should never ParseFail");
       }
     }
+  } else if (Sym->isUnset()) {
+    // If symbol is unset, it might be created in the `parseSetAssignment`
+    // routine as an alias for a numeric register name.
+    // Lookup in the aliases list.
+    auto Entry = RegisterSets.find(Sym->getName());
+    if (Entry != RegisterSets.end()) {
+      OperandMatchResultTy ResTy =
+          matchAnyRegisterWithoutDollar(Operands, Entry->getValue(), S);
+      if (ResTy == MatchOperand_Success) {
+        Parser.Lex();
+        return true;
+      }
+    }
   }
 
   return false;
 }
 
 OperandMatchResultTy
 MipsAsmParser::matchAnyRegisterNameWithoutDollar(OperandVector &Operands,
                                                  StringRef Identifier,
Operands.push_back(MipsOperand::createMSACtrlReg(
         getLexer().getLoc(), *this));
     return MatchOperand_Success;
   }
 
   return MatchOperand_NoMatch;
 }
 
 OperandMatchResultTy
-MipsAsmParser::matchAnyRegisterWithoutDollar(OperandVector &Operands, SMLoc S) {
-  MCAsmParser &Parser = getParser();
-  auto Token = Parser.getLexer().peekTok(false);
-
+MipsAsmParser::matchAnyRegisterWithoutDollar(OperandVector &Operands,
+                                             const AsmToken &Token, SMLoc S) {
   if (Token.is(AsmToken::Identifier)) {
     LLVM_DEBUG(dbgs() << ".. identifier\n");
     StringRef Identifier = Token.getIdentifier();
     OperandMatchResultTy ResTy =
         matchAnyRegisterNameWithoutDollar(Operands, Identifier, S);
     return ResTy;
   } else if (Token.is(AsmToken::Integer)) {
     LLVM_DEBUG(dbgs() << ".. integer\n");
MipsAsmParser::matchAnyRegisterWithoutDollar(OperandVector &Operands,
   }
 
   LLVM_DEBUG(dbgs() << Token.getKind() << "\n");
 
   return MatchOperand_NoMatch;
 }
 
 OperandMatchResultTy
+MipsAsmParser::matchAnyRegisterWithoutDollar(OperandVector &Operands, SMLoc S) {
+  auto Token = getLexer().peekTok(false);
+  return matchAnyRegisterWithoutDollar(Operands, Token, S);
+}
+
+OperandMatchResultTy
 MipsAsmParser::parseAnyRegister(OperandVector &Operands) {
   MCAsmParser &Parser = getParser();
   LLVM_DEBUG(dbgs() << "parseAnyRegister\n");
 
   auto Token = Parser.getTok();
 
   SMLoc S = Token.getLoc();
 
bool MipsAsmParser::parseSetAssignment() {
 
   if (Parser.parseIdentifier(Name))
     return reportParseError("expected identifier after .set");
 
   if (getLexer().isNot(AsmToken::Comma))
     return reportParseError("unexpected token, expected comma");
   Lex(); // Eat comma
 
-  if (Parser.parseExpression(Value))
-    return reportParseError("expected valid expression after comma");
-
+  if (getLexer().is(AsmToken::Dollar) &&
+      getLexer().peekTok().is(AsmToken::Integer)) {
+    // Parse assignment of a numeric register:
+    //   .set r1,$1
+    Parser.Lex(); // Eat $.
+    RegisterSets[Name] = Parser.getTok();
+    Parser.Lex(); // Eat identifier.
+    getContext().getOrCreateSymbol(Name);
+  } else if (!Parser.parseExpression(Value)) {
+    // Parse assignment of an expression including
+    // symbolic registers:
+    //   .set  $tmp, $BB0-$BB1
+    //   .set  r2, $f2
     MCSymbol *Sym = getContext().getOrCreateSymbol(Name);
     Sym->setVariableValue(Value);
+  } else {
+    return reportParseError("expected valid expression after comma");
+  }
 
   return false;
 }
 
 bool MipsAsmParser::parseSetMips0Directive() {
   MCAsmParser &Parser = getParser();
   Parser.Lex();
   if (getLexer().isNot(AsmToken::EndOfStatement))

llvm/trunk/test/MC/Mips/mips_directives.s

$BB0_4:
     b 1332
     j 1328
     jal 1328
     .set  at=$a0
     .set STORE_MASK,$t7
     .set FPU_MASK,$f7
     .set  $tmp7, $BB0_4-$BB0_2
     .set f6,$f6
+    .set r1,$1
 # CHECK:    abs.s   $f6, $f7           # encoding: [0x46,0x00,0x39,0x85]
 # CHECK:    lui     $1, %hi($tmp7)     # encoding: [0x3c,0x01,A,A]
 # CHECK:                               #   fixup A - offset: 0, value: %hi($tmp7), kind: fixup_Mips_HI16
     abs.s  f6,FPU_MASK
-    lui $1, %hi($tmp7)
+    lui r1, %hi($tmp7)
 
 # CHECK:    .set mips32r2
 # CHECK:    ldxc1   $f0, $zero($5)     # encoding: [0x4c,0xa0,0x00,0x01]
 # CHECK:    luxc1   $f0, $6($5)        # encoding: [0x4c,0xa6,0x00,0x05]
 # CHECK:    lwxc1   $f6, $2($5)        # encoding: [0x4c,0xa2,0x01,0x80]
      .set mips32r2
     ldxc1   $f0, $zero($5)
     luxc1   $f0, $6($5)
# CHECK:   lhx     $5, $6($7)           # encoding: [0x7c,0xe6,0x29,0x0a]
    lbux    $7, $10($11)
    lhx     $5, $6($7)
 
 # CHECK:   .set dspr2
 # CHECK:   append   $7, $10, 2          # encoding: [0x7d,0x47,0x10,0x31]
 # CHECK:   balign   $5, $6, 3           # encoding: [0x7c,0xc5,0x1c,0x31]
    .set dspr2
    append   $7, $10, 2
-   balign   $5, $6, 3
+   balign   $5, $6, 3