Re: [lwip-users] IPV6 Stack in LWIP 2.0: Is it IPv6 Ready ?

[delamer]

Hi Arpit,

What you are describing is very well possible. I can't check it right 
now. Please list it as a possible bug in Savannah so it doesn't get 
lost/forgotten.

Cheers
Ivan


> Date: Fri, 7 Oct 2016 17:38:38 +0000
> From: "Arpit  Agarwal" <address@hidden>
> To: "address@hidden" <address@hidden>
> Subject: Re: [lwip-users] IPV6 Stack in LWIP 2.0: Is it IPv6 Ready ?
> Message-ID:
>         <address@hidden>
>
> Content-Type: text/plain; charset="us-ascii"
>
> Hi Ivan,
>
>
> I am testing Core Protocols Only.
>
> For e.g. Core Protocols Section 1: (spec.p2) : In my test many of the
> test cases are failing due to improper neighbor cache cleanup. If I am
> not wrong, the reason might be non-handling of one of the transition
> of nd6 state machine (refer link:
> https://njetwork.files.wordpress.com/2014/01/msi_ipv6-nd-state-machine1.png).
> I am unable to find the State change from Reachable to Stale on
> receiving NA with LLADDR change anywhere in the nd6 code.
>
>
> Similary as per RFC 4861 , DAD-NA might not contain lladdr_opt, but
> the stack in lwip seems to drop NA if there is not lladdr_opt.
>
> Refer Below code from nd6.c:
>
>
> /* Unsolicited NA?*/
>     if (ip6_addr_ismulticast(ip6_current_dest_addr())) {
>       /* This is an unsolicited NA.
>        * link-layer changed?
>        * part of DAD mechanism? */
>
>       /* Check that link-layer address option also fits in packet. */
>       if (p->len < (sizeof(struct na_header) + 2)) {
>         /* @todo debug message */
>         pbuf_free(p);
>         ND6_STATS_INC(nd6.lenerr);
>         ND6_STATS_INC(nd6.drop);
>         return;
>       }
>
>       lladdr_opt = (struct lladdr_option *)((u8_t*)p->payload +
> sizeof(struct na_header));
>
>       if (p->len < (sizeof(struct na_header) + (lladdr_opt->length << 
> 3))) {
>         /* @todo debug message */
>         pbuf_free(p);
>         ND6_STATS_INC(nd6.lenerr);
>         ND6_STATS_INC(nd6.drop);
>         return;
>       }
>
>
>
> Please correct me if I am wrong or I am missing something.
>
>
>
> --
> Thanks & Regards,
> ARPIT
>
> ________________________________
> From: lwip-users
> <address@hidden> on behalf
> of address@hidden <address@hidden>
> Sent: Friday, October 7, 2016 10:26:17 PM
> To: address@hidden
> Subject: Re: [lwip-users] IPV6 Stack in LWIP 2.0: Is it IPv6 Ready ?
>
> Hi,
>
> When I did the testing, several years back, I was going for the "IPv6
> Silver" which included the core protocols. I believe all but one of the
> tests passed, I don't remember what was missing, but it didn't seem 
> like
> a big issue for a "lightweight" stack.
>
> The Silver level is no longer supported, the Gold/Phase 2 tests include
> more features, and many are not supported. E.g. DHCPv6, IPSec, etc.
>
> I also think you can get individual test suites for different protocols
> and list which ones you support.
>
> I believe this would make a great M.Sc. Thesis work if anyone out there
> wants to give it a go. I'd be happy to supervise/cooperate, I've done a
> few of those jobs already.
>
> Cheers
> Ivan
>
>
>
> > Date: Fri, 7 Oct 2016 13:08:33 +0000
> > From: "Arpit  Agarwal" <address@hidden>
> > To: "address@hidden" <address@hidden>
> > Subject: Re: [lwip-users] IPV6 Stack in LWIP 2.0: Is it IPv6 Ready ?
> > Message-ID:
> >        
> > <address@hidden>
> >
> > Content-Type: text/plain; charset="iso-8859-1"
> >
> > Hi Simon,
> >
> >
> > Just for information, I am using Self Test Tool version 5.0.0 (IPv6
> > Phase 2 test). Also when I analyzed reason for failed test cases I
> > found functionality missing too in the code. Also first I tested using
> > lwip-contrib unixsim and my port, but still test cases were failing.
> >
> >
> > --
> > Thanks & Regards,
> > ARPIT
>
> _______________________________________________
> lwip-users mailing list
> address@hidden
> https://lists.nongnu.org/mailman/listinfo/lwip-users