Re: Locale Build Failure #2

[Corinna Vinschen <vinschen at redhat dot com>]

On Aug 16 13:40, Joel Sherrill wrote:
> On 8/16/2016 1:35 PM, Craig Howland wrote:
> > On 08/16/2016 02:17 PM, Corinna Vinschen wrote:
> > > -char *
> > > +const char *
> > >   _DEFUN_VOID (__locale_ctype_ptr)
> > >   {
> > >     /* Only check if the current thread/reent has a locale.  ctype_ptr is unused
> > > @@ -992,7 +992,7 @@ _DEFUN_VOID (__locale_ctype_ptr)
> > >   				 : __ctype_ptr__;
> > >   }
> > With the recent general change to drop original K&R-style, as long is this is
> > being edited, how about also
> > -_DEFUN_VOID (__locale_ctype_ptr)
> > +__locale_ctype_ptr(void)
> > (more readable, compares more easily against the "char *__locale_ctype_ptr
> > (void);" prototype)?
> > 
> 
> I am in the middle of just making sure all RTEMS targets build
> with the proposed patch. I am happy to do it again if another
> patch materializes.

Nah, it's the same thing, just without the weird _DEFUN stuff.

> I wondered if const char * to point to a locale was safer also.

-v?


Corinna

-- 
Corinna Vinschen
Cygwin Maintainer
Red Hat

Attachment: signature.asc
Description: PGP signature