[Sparc] Add .uahalf and .uaword directives

[dcederman]

Summary

Adding these makes it easier to assemble the output from GCC which generates a lot of .uahalf and .uaword directives.

GAS treats .uahalf and .half the same unless the --enforce-aligned-data flag is used. I could not find a similar flag for LLVM so it seems that .half does not have any alignment requirement and is treated the same as .uahalf should be. If that would change later on then the tests in sparc-directives.s would fail due to bad alignment.

[jyknight]

Oddly, the oracle reference manual says that only .half has an alignment requirement (in that it _fails_ if the current position is odd). The docs for .word explicitly say it doesn't have an alignment requirement. *boggle*. Seems useless to actually implement such a requirement, just adding aliases seems fine.

llvm/trunk/lib/Target/Sparc/AsmParser/SparcAsmParser.cpp

bool expandSET(MCInst &Inst, SMLoc IDLoc,
                  SmallVectorImpl<MCInst> &Instructions);
 
 public:
   SparcAsmParser(const MCSubtargetInfo &sti, MCAsmParser &parser,
                 const MCInstrInfo &MII,
                 const MCTargetOptions &Options)
       : MCTargetAsmParser(Options, sti, MII), Parser(parser) {
     Parser.addAliasForDirective(".half", ".2byte");
+    Parser.addAliasForDirective(".uahalf", ".2byte");
     Parser.addAliasForDirective(".word", ".4byte");
+    Parser.addAliasForDirective(".uaword", ".4byte");
     Parser.addAliasForDirective(".nword", is64Bit() ? ".8byte" : ".4byte");
     if (is64Bit())
       Parser.addAliasForDirective(".xword", ".8byte");
 
     // Initialize the set of available features.
     setAvailableFeatures(ComputeAvailableFeatures(getSTI().getFeatureBits()));
   }
 };

llvm/trunk/test/MC/Sparc/sparc-directives.s

 ! RUN: llvm-mc %s -arch=sparc   -show-encoding | FileCheck %s --check-prefix=SPARC32
 ! RUN: llvm-mc %s -arch=sparcv9 -show-encoding | FileCheck %s --check-prefix=SPARC64
 
         ! '.proc' is documented to do nothing in the binutils assembler.
         ! so it should do nothing for clang either, i.e. not be an error.
         .proc 1
 
         ! SPARC32: .byte 24
         ! SPARC64: .byte 24
         .byte 24
 
         ! SPARC32: .half 1024
         ! SPARC64: .half 1024
+        ! SPARC32: .half 1024
+        ! SPARC64: .half 1024
         .half 1024
+        .uahalf 1024
 
         ! SPARC32: .word 65536
         ! SPARC64: .word 65536
+        ! SPARC32: .word 65536
+        ! SPARC64: .word 65536
         .word 65536
+        .uaword 65536
 
         ! SPARC32: .word 65536
         ! SPARC64: .xword 65536
         .nword 65536