On Mon, Mar 20, 2017 at 10:33:57PM +0800, Xin Long wrote:
> On Mon, Mar 20, 2017 at 9:29 PM, Neil Horman <nhor...@tuxdriver.com> wrote:
> > On Mon, Mar 20, 2017 at 05:46:27PM +0800, Xin Long wrote:
> >> sctp_stream_free uses struct sctp_stream as a param, but struct sctp_stream
> >> is defined after it's declaration.
> >>
> >> This patch is to declare struct sctp_stream before sctp_stream_free.
> >>
> >> Fixes: a83863174a61 ("sctp: prepare asoc stream for stream reconf")
> >> Signed-off-by: Xin Long <lucien....@gmail.com>
> >> ---
> >>  include/net/sctp/structs.h | 1 +
> >>  1 file changed, 1 insertion(+)
> >>
> >> diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h
> >> index 4f64519..592dece 100644
> >> --- a/include/net/sctp/structs.h
> >> +++ b/include/net/sctp/structs.h
> >> @@ -83,6 +83,7 @@ struct sctp_bind_addr;
> >>  struct sctp_ulpq;
> >>  struct sctp_ep_common;
> >>  struct crypto_shash;
> >> +struct sctp_stream;
> >>
> >>
> >>  #include <net/sctp/tsnmap.h>
> >> --
> >> 2.1.0
> >>
> >>
> >
> > Not sure I follow.  did we run into a compilation failure here?
> no compilation err was found, even now.
> 
> because of
> struct sctp_stream *sctp_stream_new(__u16 incnt, __u16 outcnt, gfp_t gfp);
> void sctp_stream_free(struct sctp_stream *stream);
> 
> I'm not sure why gcc did not complain, but when I moved sctp_stream_new
> below sctp_stream_free:
> void sctp_stream_free(struct sctp_stream *stream);
> struct sctp_stream *sctp_stream_new(__u16 incnt, __u16 outcnt, gfp_t gfp);
> 
> the err showed up.
> 
odd, its almost like gcc considered the return value of sctp_stream_new to be a
forward declaration of the struct.

Anywho, can't hurt to make this chagne
Acked-by: Neil Horman <nhor...@tuxdriver.com>

> So the better thing we should do is to declare struct sctp_stream to
> avoid this potential issue.
> 
> Besides, I'm planing to change the return value type of sctp_stream_new.
> after which, the issue will be exposed.
> 
> >
> > Neil
> >
> 

Reply via email to