Re: [PATCH v1 44/59] 9p-local.c: remove unneeded label in local_unlinkat_common()

[Greg Kurz]

On Tue, 07 Jan 2020 13:04:20 +0100
Christian Schoenebeck <address@hidden> wrote:

> On Montag, 6. Januar 2020 19:24:10 CET Daniel Henrique Barboza wrote:
> > 'err_out' can be replaced by 'return ret' in the error conditions
> > the jump was being made.
> > 
> > CC: Greg Kurz <address@hidden>
> > Signed-off-by: Daniel Henrique Barboza <address@hidden>
> > ---
> >  hw/9pfs/9p-local.c | 12 +++++-------
> >  1 file changed, 5 insertions(+), 7 deletions(-)
> > 
> > diff --git a/hw/9pfs/9p-local.c b/hw/9pfs/9p-local.c
> > index ca641390fb..f9bdd2ad7c 100644
> > --- a/hw/9pfs/9p-local.c
> > +++ b/hw/9pfs/9p-local.c
> > @@ -1094,12 +1094,12 @@ static int local_unlinkat_common(FsContext *ctx, int
> > dirfd, const char *name,
> > 
> >              fd = openat_dir(dirfd, name);
> >              if (fd == -1) {
> > -                goto err_out;
> > +                return ret;
> >              }
> >              ret = unlinkat(fd, VIRTFS_META_DIR, AT_REMOVEDIR);
> >              close_preserve_errno(fd);
> >              if (ret < 0 && errno != ENOENT) {
> > -                goto err_out;
> > +                return ret;
> >              }
> >          }
> >          map_dirfd = openat_dir(dirfd, VIRTFS_META_DIR);
> > @@ -1107,16 +1107,14 @@ static int local_unlinkat_common(FsContext *ctx, int
> > dirfd, const char *name, ret = unlinkat(map_dirfd, name, 0);
> >              close_preserve_errno(map_dirfd);
> >              if (ret < 0 && errno != ENOENT) {
> > -                goto err_out;
> > +                return ret;
> >              }
> >          } else if (errno != ENOENT) {
> > -            goto err_out;
> > +            return ret;
> >          }
> >      }
> > 
> > -    ret = unlinkat(dirfd, name, flags);
> > -err_out:
> > -    return ret;
> > +    return unlinkat(dirfd, name, flags);
> >  }
> > 
> >  static int local_remove(FsContext *ctx, const char *path)
> 
> Well, personally I don't see any improvement by these changes. It probably 
> makes the code slightly more elegant, but IMO not more readable. And return 
> constructed functions vs. jump to label constructed functions are more likely 
> to gather missing-cleanup bugs.
> 
> At least this patch does not cause any behaviour change, so I leave that up 
> to 
> you Greg to decide. ;-)
> 

I don't care that much but in the present case, the function has a bug that
can be fixed with a variant of this patch if 'goto err_out' is turned into
'return -1'. I've hence asked Daniel to move this patch out of the series and
to turn it into a real fix that I'll merge directly.

> Best regards,
> Christian Schoenebeck
> 
>