| Age | Commit message (Collapse) | Author |
|---|
|
Low-probability clean-up should be to use rmdir to get rid of
the just-created directory, not unlink.
|
|
to define main with arguments, for C++. Reported by Eric Blake.
* c-stack.m4 (AC_SYS_XSI_STACK_OVERFLOW_HEURISTIC):
Prefer 'int main ()' to 'int main (void)', for C++.
* getcwd-path-max.m4 (gl_FUNC_GETCWD_PATH_MAX): Likewise.
* fsusage.m4 (gl_FILE_SYSTEM_USAGE): Specify a return type
for 'main', for C99 and C++.
|
|
and don't check for fcntl.h).
|
|
|
|
where no absolute path name can be longer than PATH_MAX.
|
|
|
|
Define HAVE_PARTLY_WORKING_GETCWD if getcwd is partly working.
|
|
GL_FUNC_GETCWD_PATH_MAX for consistency. All uses changed.
Use the _ONCE methods, for efficiency.
Check for fcntl.h. In test program, include <errno.h>
and <fcntl.h> if available. Remove old K&R cruft from
test program. Check for common errors in GNU/Linux,
OpenBSD, and Solaris. Just set gl_cv_func_getcwd_path_max;
don't do AC_LIBOBJ, as that's getcwd.m4's job.
|
|
|
|
was being replaced when it shouldn't have been, and vice versa.
|
|
declaration of getcwd *before* checking for the getcwd kernel bug.
Otherwise, configure-time `checking ...' messages would be intermixed.
|
|
is not defined, don't run the test, and don't use the wrapper.
Otherwise, on the Hurd, it would take a long time to create
and remove a hierarchy about 4000 levels deep.
Based on a patch from Robert Millan.
|
|
|
|
|
|
|
|
|
|
|
|
i386-unknown-netbsdelf1.6
|
|
|
|
|
