On Tue, Apr 18, 2023 at 10:53:23AM +0200, Samuel Thibault wrote:
> Guillaume Nault, le mar. 18 avril 2023 10:34:03 +0200, a ecrit:
> > On Mon, Apr 17, 2023 at 12:07:04AM +0200, Samuel Thibault wrote:
> > > sax.sa_family = AF_PPPOX;
> > > sax.sa_protocol = PX_PROTO_OL2TP;
> > > sax.pppol2tp.fd = tunnel_fd;
> > > @@ -406,12 +407,64 @@ Sample userspace code:
> > > /* session_fd is the fd of the session's PPPoL2TP socket.
> > > * tunnel_fd is the fd of the tunnel UDP / L2TPIP socket.
> > > */
> > > - fd = connect(session_fd, (struct sockaddr *)&sax, sizeof(sax));
> > > - if (fd < 0 ) {
> > > + ret = connect(session_fd, (struct sockaddr *)&sax, sizeof(sax));
> > > + if (ret < 0 ) {
> >
> > Now you also need to close session_fd.
>
> ? No, we need it for PPPIOCGCHAN, and also PPPIOCGL2TPSTATS.
connect() failed. You can't do anything with this socket.
> I'll put return session_fd instead.
What's the point of returning session_fd if connect() failed?
How will the caller know if session_fd is connected or not?
Why would it even be interested in a half-created session fd?
> > > +The ppp<ifunit> interface can then be configured as usual with SIOCSIFMTU,
> > > +SIOCSIFADDR, SIOCSIFDSTADDR, SIOCSIFNETMASK, and activated by setting IFF_UP
> > > +with SIOCSIFFLAGS
> > > +
> > > + - Tunnel switching is supported by bridging channels::
> >
> > This is a PPP feature not an L2TP one.
> >
> > PPPIOCBRIDGECHAN's description
> > belongs to Documentation/networking/ppp_generic.rst, where it's already
> > documented.
>
> Yes but that's hard to find out when you're looking from the L2TP end.
That's why I proposed linking to ppp_generic.rst.
> > If necessary, you can link to ppp_generic.rst here.
> >
> > Also, calling this feature 'tunnel switching' is misleading.
>
> That's how I have seen it is called in L2TP jargon.
That still doesn't describe the kernel feature. We can add a 'so called
"tunnel switching" in L2TP jargon' into parenthesis to give a hint to
the people using this terminology.
> Samuel
>
