I submitted this proposed change for openssh a while back, designed to
reduce embedded knowledge of old Class A/B/C networks and use the system's
IN_LOOPBACK macro where it exists. I received just one comment:
Ed Maste <emaste@xxxxxxxxxxx> wrote:
> On Mon, 15 Nov 2021 at 11:51, Mike Karels <karels@xxxxxxxxxxx> wrote:
> >
> > +#ifndef IN_LOOPBACK
> > +#define IN_LOOPBACK(i) (((i) & 0xff000000) == 0x7f000000)
> > +#endif
> ...
> > static int
> > sockaddr_is_local(struct sockaddr *hostaddr)
> > {
> > switch (hostaddr->sa_family) {
> > case AF_INET:
> > - return (ntohl(((struct sockaddr_in *)hostaddr)->
> > - sin_addr.s_addr) >> 24) == IN_LOOPBACKNET;
> > + return (IN_LOOPBACK(ntohl(((struct sockaddr_in *)hostaddr)->
> > + sin_addr.s_addr)));
> Looks fine to me.
> IMO #ifndef is reasonable as it's unlikely that IN_LOOPBACK() would be
> implemented as other than a macro anywhere, but could be a full
> autoconf check if that's what OpenSSH folks want.
Any other comments or suggestions? Any general thoughts on doing this?
I'd like to get it committed if possible.
Thanks,
Mike
_______________________________________________
openssh-unix-dev mailing list
openssh-unix-dev@xxxxxxxxxxx
https://lists.mindrot.org/mailman/listinfo/openssh-unix-dev
