[MC][X86] Fix Intel Operand assembly parsing for .set ids

[niravd]

Summary

Fix intel syntax special case identifier operands that refer to a constant
(e.g. .set <ID> n) to be interpreted as immediate not memory in parsing.

[rnk]

What about this case:

.set FOO 2
cmp eax, FOO [eax]

Does gas assemble that to cmpl %eax, 2(%eax), or does it reject?

[niravd]

Add FOO[...] case folding in subsequent cleanup patch

[rnk]

lib/Target/X86/AsmParser/X86AsmParser.cpp

1441

Can you add a few negative tests exercising these errors? Look at how intel-syntax-ambiguous.s tests erroneous input.

[niravd]

Add variants of relevant clang testsuite to llvm testsuite.

[rnk]

lgtm

lib/Target/X86/AsmParser/X86AsmParser.cpp

if ((Parser.getTok().getKind() == AsmToken::Identifier ||
 
     End = Tok.getEndLoc();
     Parser.Lex();  // Eat the field.
     Disp = NewDisp;
   }
 
   int BaseReg = SM.getBaseReg();
   int IndexReg = SM.getIndexReg();
   int Scale = SM.getScale();
   if (!isParsingInlineAsm()) {
     // handle [-42]
     if (!BaseReg && !IndexReg) {
       if (!SegReg)
         return X86Operand::CreateMem(getPointerWidth(), Disp, Start, End, Size);
       return X86Operand::CreateMem(getPointerWidth(), SegReg, Disp, 0, 0, 1,
                                    Start, End, Size);
     }
std::unique_ptr<X86Operand> X86AsmParser::ParseIntelMemOperand(int64_t ImmDisp,
   if (getLexer().is(AsmToken::LBrac))
     return ParseIntelBracExpression(/*SegReg=*/0, Start, ImmDisp, Size);
   assert(ImmDisp == 0);
 
   const MCExpr *Val;
   if (!isParsingInlineAsm()) {
     if (getParser().parsePrimaryExpr(Val, End))
       return ErrorOperand(Tok.getLoc(), "unknown token in expression");
-
+    if (Val->getKind() == MCExpr::Constant)
+      return X86Operand::CreateImm(Val, Start, End);
     return X86Operand::CreateMem(getPointerWidth(), Val, Start, End, Size);
   }
 
   InlineAsmIdentifierInfo Info;
   StringRef Identifier = Tok.getString();
   if (ParseIntelIdentifier(Val, Identifier, Info,
                            /*Unevaluated=*/false, End))
     return nullptr;
 
-  if (!getLexer().is(AsmToken::LBrac))
+  if (getLexer().isNot(AsmToken::LBrac)){
     return CreateMemForInlineAsm(/*SegReg=*/0, Val, /*BaseReg=*/0, /*IndexReg=*/0,
                                  /*Scale=*/1, Start, End, Size, Identifier, Info);
+  }
 
   Parser.Lex(); // Eat '['
 
   // Parse Identifier [ ImmDisp ]
   IntelExprStateMachine SM(/*ImmDisp=*/0, /*StopOnLBrac=*/true,
                            /*AddImmPrefix=*/false);
   if (ParseIntelExpression(SM, End))
     return nullptr;

test/MC/X86/intel-syntax-encoding.s

// CHECK: encoding: [0xcb]
     retf
 
 // CHECK: encoding: [0xc2,0x08,0x00]
     ret 8
 
 // CHECK: encoding: [0xca,0x08,0x00]
     retf 8
 
+    .set FOO, 2
+    cmp eax, FOO
+// CHECK: encoding: [0x83,0xf8,0x02]