Fix calls to malloc, realloc in prefix.c
Kaveh R. Ghazi
ghazi@caip.rutgers.edu
Thu Feb 17 14:23:00 GMT 2000
> From: Zack Weinberg <zack@wolery.cumb.org>
>
> You have inspired me... try this patch on for size. Does a bunch of
> stuff:
Great thanks! Comments below.
> - compile fix-header with $(CC) and add some commentary about why
Err, even though it should never run in a cross compile, I don't
entirely agree with this.
But I guess it should be okay... maybe... :-)
Anyway to be consistent, you should also change the includes of
hconfig.h to config.h in any file affected.
> GCC_NEED_DECLARATIONS(bcopy bzero bcmp \
> - index rindex getenv atol sbrk abort atof strerror getcwd getwd \
> + index rindex getenv atol sbrk abort atof getcwd getwd \
> strsignal putc_unlocked fputs_unlocked strstr environ)
Don't take strerror decl check out here. Its still called in fixincl.c
so we should still provide a decl if the system doesn't. (At least until
all calls are changed to xstrerror.)
> Index: system.h
> --- system.h 2000/01/17 15:47:29 1.59
> +++ system.h 2000/02/17 18:41:13
> @@ -394,17 +394,6 @@ extern PTR calloc PARAMS ((size_t, size_
> extern PTR realloc PARAMS ((PTR, size_t));
> #endif
>
> -#ifdef HAVE_STRERROR
> -# ifdef NEED_DECLARATION_STRERROR
> -# ifndef strerror
> -extern char *strerror PARAMS ((int));
> -# endif
> -# endif
> -#else /* ! HAVE_STRERROR */
> -extern int sys_nerr;
> -extern char *sys_errlist[];
> -#endif /* HAVE_STRERROR */
> -
Likewise, you can nuke the sys_nerr/sys_errlist, but keep the strerror
prototype. E.g.:
#ifdef NEED_DECLARATION_STRERROR
# ifndef strerror
extern char *strerror PARAMS ((int));
# endif
#endif
> - cpp_error (pfile, "%s: %s", name, my_strerror (errno));
> + cpp_error (pfile, "%s: %s", name, strerror (errno));
> }
>
> void
> @@ -473,5 +439,5 @@ cpp_notice_from_errno (pfile, name)
> cpp_reader *pfile;
> const char *name;
> {
> - cpp_notice (pfile, "%s: %s", name, my_strerror (errno));
> + cpp_notice (pfile, "%s: %s", name, strerror (errno));
Why not use xstrerror here? Keep us close to being able to remove the
prototype of strerror. :-)
--Kaveh
--
Kaveh R. Ghazi Engagement Manager / Project Services
ghazi@caip.rutgers.edu Qwest Internet Solutions
More information about the Gcc-patches
mailing list