> -----Original Message----- > From: Jeff Layton [mailto:jlayton@xxxxxxxxxx] > Sent: Monday, February 04, 2013 9:42 AM > To: Myklebust, Trond > Cc: bfields@xxxxxxxxxxxx; linux-nfs@xxxxxxxxxxxxxxx; chuck.lever@xxxxxxxxxx > Subject: Re: [PATCH] sunrpc: copy scope ID in __rpc_copy_addr6 > > On Mon, 4 Feb 2013 14:22:47 +0000 > "Myklebust, Trond" <Trond.Myklebust@xxxxxxxxxx> wrote: > > > > -----Original Message----- > > > From: linux-nfs-owner@xxxxxxxxxxxxxxx [mailto:linux-nfs- > > > owner@xxxxxxxxxxxxxxx] On Behalf Of Jeff Layton > > > Sent: Monday, February 04, 2013 8:10 AM > > > To: bfields@xxxxxxxxxxxx > > > Cc: linux-nfs@xxxxxxxxxxxxxxx; chuck.lever@xxxxxxxxxx > > > Subject: [PATCH] sunrpc: copy scope ID in __rpc_copy_addr6 > > > > > > When copying an address, we should also copy the scopeid in the > > > event that this is a link-local address and the scope matters. > > > > > > Signed-off-by: Jeff Layton <jlayton@xxxxxxxxxx> > > > --- > > > include/linux/sunrpc/clnt.h | 1 + > > > 1 file changed, 1 insertion(+) > > > > > > diff --git a/include/linux/sunrpc/clnt.h > > > b/include/linux/sunrpc/clnt.h index > > > 47354a2..6a7c261 100644 > > > --- a/include/linux/sunrpc/clnt.h > > > +++ b/include/linux/sunrpc/clnt.h > > > @@ -242,6 +242,7 @@ static inline bool __rpc_copy_addr6(struct > > > sockaddr *dst, > > > > > > dsin6->sin6_family = ssin6->sin6_family; > > > dsin6->sin6_addr = ssin6->sin6_addr; > > > + dsin6->sin6_scope_id = ssin6->sin6_scope_id; > > > return true; > > > } > > > #else /* !(IS_ENABLED(CONFIG_IPV6) */ > > > -- > > > > > > Can we please move these helpers into a separate header, say > include/Linux/sunrpc/inet.h? > > > > Yes, that's long overdue. I'll spin up a patch on top of this one to do that. I'd > suggest that we call it "addr.h" to keep it in parity with the "addr.c" file that's > in the sunrpc directory, since we'll need to move the prototypes for those > functions as well. > > > Oh... Also note that you probably want to apply the above scopeid to the > NFSv4 callback address that is supplied by the client in the setclientid call. > > > > I'm not clear on what you mean here. This patch should make it so that the > server keeps the correct scopeid along with the cb address that it tracks for a > client (something that I don't think it does properly today). > I mean that the NFSv4 client supplies an address in "universal address" format (i.e. a string) as part of the setclientid call. If that address is a link-local address, then it won't contain a scopeid ('cos the scopeid is an identifier that is local to the _server_). In that case the server should apply the scopeid from the nfs4_clientid in the gen_callback() helper. Cheers Trond -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
