[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: GMP-4.0.1 on BeOS
From: |
heikkine |
Subject: |
Re: GMP-4.0.1 on BeOS |
Date: |
Mon, 11 Mar 2002 12:36:40 +0200 (EET) |
User-agent: |
IMP/PHP IMAP webmail program lut20011113 |
Quoting Kevin Ryde <address@hidden>:
> address@hidden writes:
> >
> > There is a file /boot/develop/headers/posix/inttypes.h, in which I
> found the
> > intmax_t typedef, but the problem can't be solved by replacing the
> '#include
> > <stdint.h>' in the code with '#include <inttypes.h>', I tried this and
> it still
> > gives the same error.
>
> You didn't just change it within the "#if HAVE_STDINT_H"? That of
> course won't work since that define is presumably false (ie. left
> undefined by config.h).
I meant that I replaced the #include <stdint.h> in the printf/doprnt.c and
scanf/doscan.c files, not in the configure script.
>
> In any case I think the fix would be to use the following,
>
> #if HAVE_INTTYPES_H
> # include <inttypes.h>
> #else
> # if HAVE_STDINT_H
> # include <stdint.h>
> # endif
> #endif
>
> This is what the configure tests use. It'd be for printf/doprnt.c and
> scanf/doscan.c, and also for printf/repl-vsnprintf.c (though I guess
> that doesn't get used).
>
I agree, this would make GMP compile out-of-the-box on BeOS (at least R5,
hopefully R4.x and below, too). I tried this by extracting the archive,
modifying the three files mentioned above by replacing the normal /* for
intmax_t */ section of the code with the code above, and then running
'./configure', 'make' and 'make check'. All tests passed. I hope 4.0.2 includes
this fix.