João Valverde
changed
bug 10673
| What |
Removed |
Added |
| Status |
UNCONFIRMED
|
RESOLVED
|
| Resolution |
---
|
NOTABUG
|
Comment # 15
on bug 10673
from João Valverde
(In reply to João Valverde from comment #14)
> (In reply to boaz.brickner from comment #11)
> > Thank you for the detailed explanation!
> >
> > I've tried some more packets and encountered the attached "Home Address and
> > RPL" packet.
> > It contains an IPv6 packet with a Mobile IP extension header as the first
> > extension header and an RPL extension header as the 4th extension header.
> > It seems that in this packet, the CmprI used are taken from the IPv6
> > destination address field and not from the Mobile IP Home Address field
> > (using Wireshark 1.12.8).
> >
> > Why is the behavior different in this packet?
> >
> > Thanks!
> > Boaz.
>
> Hi,
>
> That's because the segments left field in the MIP routing header is zero.
> Wireshark displays an expert info protocol error in that case (RFC 6275
> section 6.4.1).
>
> For the purpose of selecting the destination address the segleft == 0 header
> is treated as a no-op. I'm not opposed to changing that.
This is a different issue, namely whether an invalid header that will generate
an ICMPv6 error (the MIP routing header in this case) should affect the
dissection state. I could see the argument going both ways but I'm inclined to
leave it as is. Seems sensible that segleft == 0 is a no-op (this will affect,
among other things, the IPv6 pseudo-header for checksum calculation).
You are receiving this mail because:
- You are watching all bug changes.