search for: mod_nsec

...a MAXPATHLEN buffer! (The * buffer is used for recursion, but returned unchanged.) @@ -786,13 +786,13 @@ @@ -353,7 +354,7 @@ void set_blocking(int fd); int fd_pair(int fd[2]); void print_child_argv(const char *prefix, char **cmd); - int set_modtime(const char *fname, time_t modtime, uint32 mod_nsec, mode_t mode); + int set_modtime(const char *fname, time_t modtime, uint32 mod_nsec, mode_t mode, uint32 fileflags); -int make_path(char *fname, int flags); +int make_path(char *fname, mode_t mode, int flags); int full_write(int desc, const char *ptr, size_t len); int copy_file(const char *so...

While going through an old todo list I found that these patches had fallen by the way-side. About a year ago I initiated a discussion[1] with the Linux kernel folks regarding the lack of any useable fadvise support on the kernel side. As a result, I was observing extremely poor performance on my server after backup as executable pages were being swapped out in favor of data waiting to be flushed

...g but a fixed length size, when at least for GCC a long is basically a machine word, that is, its size depends on the CPU architecture. ( Windows long type definition is better in that particular case, making it AFAIR an int32. ) Rsync, by the way, takes a much more reasonable approach by defining mod_nsec in set_modtime() as an uint32 ; however, this sanity is incompatible with insanity in the standard ( see e.g. http://pubs.opengroup.org/onlinepubs/7908799/xsh/time.h.html ), what will of course break the bogus nanosecond mtimes anyway. For a discussion, please refer to a following rsync lists mess...

...ne could stuff the whole Unix timestamp in that field, so we would probably want to ban that as well for the sake of general sanity. Now back to rsync: looking at set_modtime() definition in util.c, one can clearly see that rsync developers take a very sane and reasonable approach by declaring mod_nsec as an uint32 -- that is, both unsigned, and having a byte-fixed reasonable size. ( This is imho how tv_nsec field should have been declared! ) Furthermore, rsync 3.1.1 -- see set_file_attrs() in rsync.c -- compares only mtime _seconds_, and takes no attempt to adjust mtimes further if the &quo...