This is the mail archive of the libc-alpha@sources.redhat.com mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [patch] MIPS/Linux: Clean up cpp ABI dependencies


Eric Christopher wrote:
> 
> > I strongly believe that <sgidefs.h> has no business defining _ABIO32 in
> > the first place.  It's a GCC invention and a GCC macro.  It certainly
> > isn't (as the name of the header implies) an SGI definition.
> > 
> > The rule IMO should be:
> > 
> >   (1) If you're going to include <sgidefs.h>, use the macros that
> >       <sgidefs.h> defines, i.e. _MIPS_SIM_*.
> > 
> >   (2) If you want to use GCC-defined macros (such as for glibc internals)
> >       and want to retain compatibility with earlier GCC versions that didn't
> >       define them, handle it in the configure script.  Add -D_ABIO32=1 where
> >       necessary, etc.
> > 
> > I really don't see why adding _ABIO32 to <sgidefs.h> is going to help.
> 
> I agree with this. Though, fwiw, I agree with Maciej that glibc should
> be using ABI* internally. It'll get us away from including sgidefs in
> glibc code. It's not like we need to worry about other compilers there.

It will IIRC break on gcc 3.2 and probably earlier, where the builtin
for O32 is not _ABIO32 but _MIPS_SIM_ABI32. OTOH, it's probably
acceptable to break on bugs in old compilers.


Thiemo


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]