[llvm] r234295 - Try a third time to fix MSVC build after r234290

[Duncan P. N. Exon Smith]

> On 2015-Apr-06, at 22:03, Duncan P. N. Exon Smith <dexonsmith at apple.com> wrote:
> 
> Author: dexonsmith
> Date: Tue Apr  7 00:03:47 2015
> New Revision: 234295
> 
> URL: http://llvm.org/viewvc/llvm-project?rev=234295&view=rev
> Log:
> Try a third time to fix MSVC build after r234290
> 
> I have no idea what MSVC means with its error text here :(.
> 
> http://lab.llvm.org:8011/builders/sanitizer-windows/builds/2310
> 
> Modified:
>    llvm/trunk/lib/CodeGen/AsmPrinter/DwarfUnit.cpp
> 
> Modified: llvm/trunk/lib/CodeGen/AsmPrinter/DwarfUnit.cpp
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/CodeGen/AsmPrinter/DwarfUnit.cpp?rev=234295&r1=234294&r2=234295&view=diff
> ==============================================================================
> --- llvm/trunk/lib/CodeGen/AsmPrinter/DwarfUnit.cpp (original)
> +++ llvm/trunk/lib/CodeGen/AsmPrinter/DwarfUnit.cpp Tue Apr  7 00:03:47 2015
> @@ -1015,11 +1015,11 @@ void DwarfUnit::constructTypeDIE(DIE &Bu
>     // Add return type. A void return won't have a type.
>     auto Elements = cast<MDSubroutineType>(CTy)->getTypeArray();
>     if (Elements.size())
> -      if (MDType *RTy = resolve(DITypeRef(Elements[0])))
> +      if (MDType *RTy = resolve(DITypeRef(Elements[0u])))
>         addType(Buffer, RTy);
> 
>     bool isPrototyped = true;
> -    if (Elements.size() == 2 && !DITypeRef(Elements[1]))
> +    if (Elements.size() == 2 && !DITypeRef(Elements[1u]))
>       isPrototyped = false;
> 
>     constructSubprogramArguments(Buffer, Elements);
> @@ -1318,7 +1318,7 @@ void DwarfUnit::applySubprogramAttribute
>   // Add a return type. If this is a type like a C/C++ void type we don't add a
>   // return type.
>   if (Args.size())
> -    if (MDType *Ty = resolve(DITypeRef(Args[0])))
> +    if (MDType *Ty = resolve(DITypeRef(Args[0u])))
>       addType(SPDie, Ty);
> 
>   unsigned VK = SP.getVirtuality();
> 

This one worked.  There was another failure here with a message I
actually understood:

http://lab.llvm.org:8011/builders/clang-cmake-mips/builds/4442

(Not sure why I couldn't figure out what MSVC was saying.)

This end result is terrible, so I'll have to rethink the API for these
arrays -- either remove `operator[]()` or remove `operator MDTuple*()`.