This is the mail archive of the libc-alpha@sourceware.org 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: PowerPC ugly symbol versioning


On Tue, Aug 20, 2013 at 02:19:15PM -0500, Ryan S. Arnold wrote:
> On Thu, Aug 8, 2013 at 11:52 PM, Alan Modra <amodra@gmail.com> wrote:
> > -# if __WORDSIZE == 64
> > -symbol_version (__novmx_longjmp,_longjmp,GLIBC_2.3);
> > -symbol_version (__novmxlongjmp,longjmp,GLIBC_2.3);
> > -symbol_version (__novmxsiglongjmp,siglongjmp,GLIBC_2.3);
> > -# else
> > -symbol_version (__novmx_longjmp,_longjmp,GLIBC_2.0);
> > -symbol_version (__novmxlongjmp,longjmp,GLIBC_2.0);
> > -symbol_version (__novmxsiglongjmp,siglongjmp,GLIBC_2.0);
> > -# endif
> > +compat_symbol (libc, __novmx_longjmp, _longjmp, GLIBC_2_0);
> > +compat_symbol (libc, __novmxlongjmp, longjmp, GLIBC_2_0);
> > +compat_symbol (libc, __novmxsiglongjmp, siglongjmp, GLIBC_2_0);
> 
> Doesn't eliminating the __WORDSIZE == 64 check erroneously export
> glibc 2.0 symbol availability for this symbol?

No.  compat_symbol uses VERSION_libc_GLIBC_2_0 for the symbol version,
and VERSION_libc_GLIBC_2_0 is defined to GLIBC_2.3 when shlib-versions
says the earliest symbol set is GLIBC_2.3.

-- 
Alan Modra
Australia Development Lab, IBM


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