This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC 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: 3.1 on Darwin?


lucier@math.purdue.edu wrote:
> 
> In the thread beginning
> 
> http://gcc.gnu.org/ml/gcc/2001-10/msg00578.html
> 
> it was generally acknowledged that there's a problem with Darwin's
> headers wrt isalpha and relatives. It seems that it should be relatively
> easy to fix with fixheaders (coming from someone who nothing about
> fixheaders ;-).  Dale Johansen  outlined a fix in
> 
> http://gcc.gnu.org/ml/gcc/2001-10/msg00605.html
> 
> Is there any interest in getting this in for 3.1?  I don't know if
> a PR has been filed for it.

Yes, in fact I spent a some time this weekend tangling with it.
Adding to fixincludes requires the use of autogen, which itself
requires the building and installation of guile, which worked, but
then autogen was unable to finish building because of syntax
errors in autogen's libtool script, presumably in the Darwin
specific parts, at which point I swore at the machine for a bit
and spent the remainder of the weekend doing more enjoyable
things.

Just to bump up my motivation :-), /usr/include/ctype.h is
now fixed in latest Darwin sources, so any fixincludes addition
will only be necessary for a few months (and for systems that
don't update their OS of course).  I suppose my Plan B will be
to find a platform that autogen *does* work on, and simulate a
fix-needing ctype.h somehow.

Stan


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