Re: [PATCH 6/9] streaming_write_entry: propagate streaming errors

From: Jeff King <peff@peff.net>
Date: Mar 25, 2013, 21:37 UTC
Message-ID: <20130325213737.GC19303@sigill.intra.peff.net>
In-Reply-To: <CAPig+cRjK6mrRm+K4Qzf2CsjT3SYGotZ2PrVLniYzdBRC1Mv2A@mail.gmail.com>

On Mon, Mar 25, 2013 at 05:35:51PM -0400, Eric Sunshine wrote:

Show 18 quoted lines

> On Mon, Mar 25, 2013 at 4:22 PM, Jeff King <peff@peff.net> wrote:
> > diff --git a/entry.c b/entry.c
> > index 17a6bcc..002b2f2 100644
> > --- a/entry.c
> > +++ b/entry.c
> > @@ -126,8 +126,10 @@ static int streaming_write_entry(struct cache_entry *ce, char *path,
> >         fd = open_output_fd(path, ce, to_tempfile);
> >         if (0 <= fd) {
> >                 result = stream_blob_to_fd(fd, ce->sha1, filter, 1);
> > -               *fstat_done = fstat_output(fd, state, statbuf);
> > -               result = close(fd);
> > +               if (!result) {
> > +                       *fstat_done = fstat_output(fd, state, statbuf);
> > +                       result = close(fd);
> > +               }
> 
> Is this intentionally leaking the opened 'fd' when stream_blob_to_fd()
> returns an error?

Good catch. I was so focused on making sure we still called unlink that I forgot about the cleanup side-effect of close.

I'll re-roll it.

-Peff