Re: [patch] cifs: change && to ||
Dan Carpenter <[email protected]> Mon, 26 Apr 2010 12:46:43 +0200
| Newsgroups | gmane.network.samba.internals,gmane.linux.kernel.janitors,gmane.linux.file-systems.cifs |
|---|---|
| Message-ID | <20100426104643.GR29093@bicker> |
On Mon, Apr 26, 2010 at 06:44:01AM -0400, Jeff Layton wrote: > On Mon, 26 Apr 2010 12:10:06 +0200 > Dan Carpenter <[email protected]> wrote: > > > This is a typo, if pvolume_info were NULL it would oops. > > > > This function is used in clean up and error handling. The current code > > never passes a NULL pvolume_info, but it could pass a NULL *pvolume_info > > if the kmalloc() failed. > > > > Signed-off-by: Dan Carpenter <[email protected]> > > > > diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c > > index d9566bf..85a35a4 100644 > > --- a/fs/cifs/connect.c > > +++ b/fs/cifs/connect.c > > @@ -2262,7 +2262,7 @@ cleanup_volume_info(struct smb_vol **pvolume_info) > > { > > struct smb_vol *volume_info; > > > > - if (!pvolume_info && !*pvolume_info) > > + if (!pvolume_info || !*pvolume_info) > > return; > > > > volume_info = *pvolume_info; > > Good catch. I wonder though whether we can just eliminate that check > altogether? In any case, this is clearly correct in the short term. Nah, we can't. If the kmalloc() that allocates *pvolume_info fails it would lead to an OOPs. regards, dan carpenter > > Acked-by: Jeff Layton <[email protected]>