On Fri, Nov 03, 2017 at 04:24 PM GMT, Mark Michelson wrote:
> Thanks for the reviews Jakub. I've added an in-line comment below.
> Otherwise consider that there is an implicit "Will do!" on all of your
> other suggestions.
>
> On Fri, Nov 3, 2017 at 11:04 AM Jakub Sitnicki <j...@redhat.com> wrote:
[...]
>> > diff --git a/lib/packets.h b/lib/packets.h
>> > index 057935cbf..9d69b93d0 100644
>> > --- a/lib/packets.h
>> > +++ b/lib/packets.h
>> > @@ -976,6 +976,7 @@ BUILD_ASSERT_DECL(ND_PREFIX_OPT_LEN == sizeof(struct
>> ovs_nd_prefix_opt));
>> >
>> > /* Neighbor Discovery option: MTU. */
>> > #define ND_MTU_OPT_LEN 8
>> > +#define ND_MTU_DEFAULT 0
>> > struct ovs_nd_mtu_opt {
>> > uint8_t type; /* ND_OPT_MTU */
>> > uint8_t len; /* Always 1. */
>> > @@ -1015,6 +1016,9 @@ BUILD_ASSERT_DECL(RA_MSG_LEN == sizeof(struct
>> ovs_ra_msg));
>> > #define ND_RA_MANAGED_ADDRESS 0x80
>> > #define ND_RA_OTHER_CONFIG 0x40
>> >
>> > +#define ND_RA_MAX_INTERVAL_DEFAULT 600
>> > +#define ND_RA_MIN_INTERVAL_DEFAULT(max) (max) >= 9 ? (max) / 3 : (max)
>> * 3 / 4
>> > +
>>
>> I don't understand the formula. It generates a sequence that is not
>> always increasing but takes a dip when max == 9. What is the reasoning
>> behind it?
>>
>
> It's based on RFC 4861 section 6.2.1.
>
> MinRtrAdvInterval
> The minimum time allowed between sending
> unsolicited multicast Router Advertisements from
> the interface, in seconds. MUST be no less than 3
> seconds and no greater than .75 *
> MaxRtrAdvInterval.
>
> Default: 0.33 * MaxRtrAdvInterval If
> MaxRtrAdvInterval >= 9 seconds; otherwise, the
> Default is 0.75 * MaxRtrAdvInterval.
>
> Note that this is not the exact quote from the RFC. I have altered the
> default text to include the suggested change from Errata 3154. Does that
> explain the formula better?
>
> I can add the citation to the code so that it makes more sense for someone
> skimming the code.
OK, I see now where it came from. Thanks for explaining it.
Maybe a short reference will do?
/* RFC 4861 MinRtrAdvInterval, MaxRtrAdvInterval */
... so the reader has a hint where to look it up.
-Jakub
_______________________________________________
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
