[cfe-commits] r61430 - in /cfe/trunk: include/clang/Basic/TokenKinds.def lib/Lex/Preprocessor.cpp lib/Parse/ParseDecl.cpp test/Parser/MicrosoftExtensions.c

[steve naroff]

On Dec 25, 2008, at 10:34 AM, Anton Korobeynikov wrote:

> Hi, Steve
>
>> If you can educate me a bit more, I'm happy to help fix the mingw  
>> build...
> I'm going to commit some amount of amount of code, which will complete
> the semantic checks of stdcall / fastcall and dll* stuff.
>
> Then my plan was to try to figure out, what's missing to have more or
> less decent mingw support. So, I think we'll return to this problem
> really soon. Maybe we need to distinguish somehow "ms extensions
> handled by gcc" and "ms extensions not handled by gcc". Or something
> like this...
>

Hi Anton,

I don't understand what gcc has to to with this (since we are talking  
about clang, not llvm-gcc).

In any event, it sounds like the interesting distinction is "implement/ 
analyze" (for -emit-llvm) vs. "ignore/parse" (for -rewrite-objc).

-fms-extensions-ignore vs. -fms-extensions-implement, or
-fms-extensions-parse vs. -fms-extensions-analyze

Would having 2 explicit switches solve the problem?

snaroff

> -- 
> With best regards, Anton Korobeynikov
> Faculty of Mathematics and Mechanics, Saint Petersburg State  
> University