Moshe wrote:
I am attaching a fixed
sysdeps/unix/sysv/linux/openat.c file
and a diff from the original.
You should explain why your version is better. Currently, it looks like
a hack.
--- openat.orig 2006-03-01 07:32:42.000000000 +0200
+++ openat.c.fixed 2006-04-30 19:15:52.000000000 +0300
@@ -27,6 +27,9 @@
#include <sysdep-cancel.h>
#include <not-cancel.h>
+#if !defined OPENAT
+# define OPENAT openat
+#endif
#if !defined OPENAT && !defined __ASSUME_ATFCTS
# define OPENAT openat
So looks like you got __ASSUME_ATFCTS defined. Why?
@@ -70,11 +73,7 @@ int __have_atfcts;
int
-OPENAT_NOT_CANCEL (fd, file, oflag, mode)
- int fd;
- const char *file;
- int oflag;
- mode_t mode;
+OPENAT_NOT_CANCEL (int fd, const char *file, int oflag, mode_t mode)
This is simply a change from K&R to ANSI C. Doesn't gcc still succeed to
compile the old-fashioned code? If it does, why the change?
--
Alexander E. Patrakov
--
http://linuxfromscratch.org/mailman/listinfo/lfs-dev
FAQ: http://www.linuxfromscratch.org/faq/
Unsubscribe: See the above information page
