[X86] Parse data32 call in .code16 as CALLpcrel32

[MaskRay]

Summary

Used by kexec-tools (PR46942)
In GNU as, tc-i386.c:output_jump uses 4-byte immediate if a data32 prefix is present.

[craig.topper]

LGTM

llvm/lib/Target/X86/AsmParser/X86AsmParser.cpp

bool X86AsmParser::ParseInstruction(ParseInstructionInfo &Info, StringRef Name,
   }
   if (PatchedName == "data32") {
     if (is32BitMode())
       return Error(NameLoc, "redundant data32 prefix");
     if (is64BitMode())
       return Error(NameLoc, "'data32' is not supported in 64-bit mode");
     // Hack to 'data16' for the table lookup.
     PatchedName = "data16";
+
+    if (getLexer().isNot(AsmToken::EndOfStatement)) {
+      StringRef Next = Parser.getTok().getString();
+      // Parse data32 call as calll.
+      if (Next == "call" || Next == "callw") {
+        getLexer().Lex();
+        Name = "calll";
+        PatchedName = Name;
+        isPrefix = false;
+      }
+    }
   }
 
   Operands.push_back(X86Operand::CreateToken(PatchedName, NameLoc));
 
   // Push the immediate if we extracted one from the mnemonic.
   if (ComparisonPredicate != ~0U && !isParsingIntelSyntax()) {
     const MCExpr *ImmOp = MCConstantExpr::create(ComparisonPredicate,
                                                  getParser().getContext());

llvm/test/MC/X86/data-prefix-fail.s

 // 64: lgdtq 0
 // 64: encoding: [0x0f,0x01,0x14,0x25,0x00,0x00,0x00,0x00]
 // 32: data16
 // 32: encoding: [0x66]
 // 32: lgdtl 0
 // 32: encoding: [0x0f,0x01,0x15,0x00,0x00,0x00,0x00]
 // ERR16: error: redundant data16 prefix
 data16 lgdt 0
+
+// ERR16: {{.*}}.s:[[#@LINE+1]]:1: error: redundant data16 prefix
+data16 call 0

llvm/test/MC/X86/x86-16.s

 // CHECK: ljmpw $31438, $31438
 
 call  $0x7ace,$0x7ace
 lcall $0x7ace,$0x7ace
 jmp $0x7ace,$0x7ace
 ljmp  $0x7ace,$0x7ace
 
 // CHECK: calll a
+// CHECK: calll a
+// CHECK: calll a
  calll a
+data32 call a
+data32 callw a
 
 // CHECK: incb  %al # encoding: [0xfe,0xc0]
   incb %al
 
 // CHECK: incw  %ax # encoding: [0x40]
   incw %ax
 
 // CHECK: incl  %eax # encoding: [0x66,0x40]