Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> writes:
> Actually, thinking about it, the +20 was not just fairly arbitrary and
> would have been better off with a comment, it was also not a very good
> number to begin with - it just was the old tempfile pattern rounded up to
> the next "even" number.
>
> The _correct_ number to use is +39, with a better tempfile pattern.
>
> That needs a few comments, though. So I added those too.
>
> This patch is totally unimportant, but it does mean that a really
> traditional filesystem can now do the final rename in-place, because the
> temporary file is not just in the same directory as the final one, but ti
> also has the same length, so old-style filesystems can literally just edit
> the name in place and mark the buffer dirty.
That's interesting and somewhat amusing ;-)
> So take this or leave it as you want - it's really not very important, I
> just wrote this because I was thinking about what really goes on at a very
> low level when we do that final atomic rename to create the actual
> directory entry. This makes it theoretically just a tiny bit more atomic
> on old-fashioned filesystems.
However, this may be more important fix. We want to make sure that
adjust_shared_perm() is called on the success codepath, especially not
when mkdir() does _not_ fail.
diff --git a/sha1_file.c b/sha1_file.c
index 500584b..e300562 100644
--- a/sha1_file.c
+++ b/sha1_file.c
@@ -2125,7 +2125,7 @@ static int create_tmpfile(char *buffer, size_t bufsiz, const char *filename)
if (fd < 0 && dirlen) {
/* Make sure the directory exists */
buffer[dirlen-1] = 0;
- if (mkdir(buffer, 0777) && adjust_shared_perm(buffer))
+ if (mkdir(buffer, 0777) || adjust_shared_perm(buffer))
return -1;
/* Try again */
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
