[llvm-commits] [llvm-gcc-4.2] r48471 - in /llvm-gcc-4.2/trunk/gcc/config/rs6000: llvm-rs6000.cpp rs6000.h

[Dale Johannesen]

On Mar 17, 2008, at 3:29 PM, Bill Wendling wrote:

> Hi Dale,
>
> =
> =
> =
> =
> =
> =
> =
> =
> ======================================================================
>> --- llvm-gcc-4.2/trunk/gcc/config/rs6000/rs6000.h (original)
>> +++ llvm-gcc-4.2/trunk/gcc/config/rs6000/rs6000.h Mon Mar 17  
>> 17:22:06 2008
>> @@ -3486,6 +3486,16 @@
>>  #define LLVM_SHOULD_PASS_AGGREGATE_IN_MIXED_REGS(T, TY, E) \
>>    llvm_rs6000_should_pass_aggregate_in_mixed_regs((T), (TY), (E))
>>
>> +// Non-altivec vectors bigger than 4 bytes are returned by sret.
>
> Watch out for C++-style comments in GCC header files. They're
> #included into non-C++ files which cause warnings which turn into
> errors when we build the Apple Way(tm).

Sorry, I did know that.  But it occurs to me to wonder why this is a  
problem.  Both gcc and Apple supported C++ style comments in C as far  
back as the early 90s, and there is really no reason not to accept  
them.  Do we build with -pedantic or something?