Hi,
On Tue, 3 Feb 2009, Junio C Hamano wrote:
> Johannes Schindelin <johannes.schindelin@xxxxxx> writes:
>
> > The variable 'err' was not necessarily initialized before it was used.
> >
> > Signed-off-by: Johannes Schindelin <johannes.schindelin@xxxxxx>
> > ---
> > refs.c | 2 +-
> > 1 files changed, 1 insertions(+), 1 deletions(-)
> >
> > diff --git a/refs.c b/refs.c
> > index b13e01b..ded7ec4 100644
> > --- a/refs.c
> > +++ b/refs.c
> > @@ -929,7 +929,7 @@ static int log_ref_write(const char *ref_name, const unsigned char *old_sha1,
> > int delete_ref(const char *refname, const unsigned char *sha1, int delopt)
> > {
> > struct ref_lock *lock;
> > - int err, i = 0, ret = 0, flag = 0;
> > + int err = 0, i = 0, ret = 0, flag = 0;
> > struct stat loginfo;
> > int log = !lstat(git_path("logs/%s", refname), &loginfo);
>
> Sorry, I do not see it.
>
> There are two uses of "err" in this function, both of which looks like:
>
> if (err && errno != ENOENT)
>
> but both of these places have
>
> err = unlink(...)
>
> immediately before it.
I should know better by now than to base my work on the tip of my Git
tree. Sure enough I have an attempt at resolving the "deleting a branch
deletes its reflog, too" issue, which has that bug.
Sorry for the noise,
Dscho
--
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
