[Llvm-bgq-discuss] Patches for r176829-20130309 (the current vesta version)

[Hal Finkel]

----- Original Message -----
> From: "Michael Kruse" <MichaelKruse at meinersbur.de>
> To: "Hal Finkel" <hfinkel at anl.gov>
> Cc: "Jeff Hammond" <jhammond at alcf.anl.gov>, llvm-bgq-discuss at lists.alcf.anl.gov, "Michael Kruse"
> <reply at meinersbur.de>
> Sent: Thursday, April 18, 2013 8:20:59 AM
> Subject: Re: [Llvm-bgq-discuss] Patches for r176829-20130309 (the current vesta version)
> 
> 2013/4/17 Hal Finkel <hfinkel at anl.gov>:
> >> Is the inlining problem related to
> >> http://clang.llvm.org/compatibility.html#inline?
> >
> > Yes, that's correct. Maybe we should turn on the gnu-style inline
> > interpretation in the header file? FWIW, never versions of the
> > glibc headers have fixed this issue, so our problem really is just
> > that we have older system headers. Nevertheless, we should
> > probably make this as easy as possible for the users.
> 
> Why glibc? The problem is the line

Indeed, thanks! You're correct, I was misremembering: On the BG/P the glibc headers are old enough that they still have this problem, on the Q, the glibc headers are fine.

 -Hal 

> 
> #ifndef __INLINE__
> #define __INLINE__ extern inline __attribute__((always_inline))
> #endif
> 
> in /bgsys/drivers/ppcfloor/hwi/include/common/compiler_support.h This
> is gnu89-style inline. gcc should also use C99 semantics with the
> switch -std=c99 or -std=gnu99 . Maybe it's the
> __attribute__((always_inline)) that makes the difference? In my own
> code I use
> 
> #if defined(__GNUC__) && !defined(__GNUC_STDC_INLINE__)
> #define EXTERN_INLINE_DECLARATION extern inline
> #define EXTERN_INLINE_DEFINITION inline
> #else
> #define EXTERN_INLINE_DECLARATION inline
> #define EXTERN_INLINE_DEFINITION extern inline
> #endif
> 
> to distinguish the cases.
> 
> Regards,
> Michael
> 
> --
> Tardyzentrismus verboten!
>