On Thu, Jun 9, 2016 at 9:38 AM, Justin Pettit <jpet...@ovn.org> wrote:
>
> > On Jun 9, 2016, at 9:25 AM, Darrell Ball <dlu...@gmail.com> wrote:
> >
> >
> >
> > On Thu, Jun 9, 2016 at 12:12 AM, Justin Pettit <jpet...@ovn.org> wrote:
> > A zero prefix length is used to match any IP address, which is useful
> > for defining default routes.
> >
> > Signed-off-by: Justin Pettit <jpet...@ovn.org>
> > ---
> > lib/packets.c | 10 +++++-----
> > 1 file changed, 5 insertions(+), 5 deletions(-)
> >
> > diff --git a/lib/packets.c b/lib/packets.c
> > index 6a55d6f..ecb5339 100644
> > --- a/lib/packets.c
> > +++ b/lib/packets.c
> > @@ -442,9 +442,9 @@ ip_parse_masked_len(const char *s, int *n, ovs_be32
> *ip,
> > /* OK. */
> > } else if (ovs_scan_len(s, n, IP_SCAN_FMT"/%d",
> > IP_SCAN_ARGS(ip), &prefix)) {
> > - if (prefix <= 0 || prefix > 32) {
> > - return xasprintf("%s: network prefix bits not between 0 and
> "
> > - "32", s);
> > + if (prefix < 0 || prefix > 32) {
> > + return xasprintf("%s: IPv4 network prefix bits not between
> 0 and "
> > + "31, inclusive", s);
> >
> >
> > Are we are talking number of bits, not bit indices ?
>
> I'm trying specify the width of a CIDR block. Do you have a suggestion
> for a better way to phrase the error message?
>
I edited the error msgs, per below, in the previous response with an upper
range change; that was my suggestion.
+ return xasprintf("%s: IPv4 network prefix bits not between 0 and
"
+ "32, inclusive", s);
+ "and 128, inclusive", s);
for v4, 0 to 32 is valid; i.e. /0 to /32; => not (between 0 and 32,
inclusive)
for v6, 0 to 128 is valid: i.e. /0 to /128; => not (between 0 and 128,
inclusive)
>
> --Justin
>
>
>
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
