[REGRESSION] net: SRv6 End.DT6 function is broken in VRF mode

[REGRESSION] net: SRv6 End.DT6 function is broken in VRF mode

[Anton Makarov]

#regzbot introduced: b9132c32e01976686efa26252cc246944a0d2cab

Hi All!

Seems there is a regression of SRv6 End.DT6 function in VRF mode. In the 
following scenario packet is decapsulated successfully on vrf10 
interface but not forwarded to vrf10's slave interface:

ip netns exec r4 ip -6 nexthop add id 1004 encap seg6local action 
End.DT6 vrftable 10 dev vrf10

ip netns exec r4 ip -6 route add fcff:0:4:200:: nhid 1004


In End.DT6 legacy mode everything works good:

ip netns exec r4 ip -6 nexthop add id 1004 encap seg6local action 
End.DT6 table 10 dev vrf10

ip netns exec r4 ip -6 route add fcff:0:4:200:: nhid 1004


The issue impacts even stable v5.18.1. Please help to fix it.


Thanks!

Anton

Re: [REGRESSION] net: SRv6 End.DT6 function is broken in VRF mode

[Andrea Mayer]

On Fri, 3 Jun 2022 15:23:26 +0300
Anton Makarov <am@3a-alliance.com> wrote:

> #regzbot introduced: b9132c32e01976686efa26252cc246944a0d2cab
> 
> Hi All!
> 
> Seems there is a regression of SRv6 End.DT6 function in VRF mode. In the 
> following scenario packet is decapsulated successfully on vrf10 
> interface but not forwarded to vrf10's slave interface:
> 
> ip netns exec r4 ip -6 nexthop add id 1004 encap seg6local action 
> End.DT6 vrftable 10 dev vrf10
> 
> ip netns exec r4 ip -6 route add fcff:0:4:200:: nhid 1004
> 
> 
> In End.DT6 legacy mode everything works good:
> 
> ip netns exec r4 ip -6 nexthop add id 1004 encap seg6local action 
> End.DT6 table 10 dev vrf10
> 
> ip netns exec r4 ip -6 route add fcff:0:4:200:: nhid 1004
> 
> 
> The issue impacts even stable v5.18.1. Please help to fix it.
> 
> 
> Thanks!
> 
> Anton
> 
> 
> 

Hi Anton,

thank you for reporting this issue. I am already working on a fix patch which I
will send shortly.

Andrea

Re: [REGRESSION] net: SRv6 End.DT6 function is broken in VRF mode

[Thorsten Leemhuis]

On 06.06.22 14:33, Andrea Mayer wrote:
> On Fri, 3 Jun 2022 15:23:26 +0300
> Anton Makarov <am@3a-alliance.com> wrote:
> 
>> #regzbot introduced: b9132c32e01976686efa26252cc246944a0d2cab
>>
>> Seems there is a regression of SRv6 End.DT6 function in VRF mode. In the 
>> following scenario packet is decapsulated successfully on vrf10 
>> interface but not forwarded to vrf10's slave interface:
>>
>> ip netns exec r4 ip -6 nexthop add id 1004 encap seg6local action 
>> End.DT6 vrftable 10 dev vrf10
>>
>> ip netns exec r4 ip -6 route add fcff:0:4:200:: nhid 1004

#regzbot fixed-by: a3bd2102e4642

> thank you for reporting this issue. I am already working on a fix patch which I
> will send shortly.

@Andrea: when you fix a reported issue, next time please include a
"Link: <url>" that links to the report, as explained in the
documentation (see submitting-patches.rst). Linus wants these tags(¹)
and my regression tracking efforts rely on them (that'S why I had to
write this mail to tell regression tracking bot with above command that
the issue has been fixed). tia!

Ciao, Thorsten

(¹) see for example:
https://lore.kernel.org/all/CAHk-=wjMmSZzMJ3Xnskdg4+GGz=5p5p+GSYyFBTh0f-DgvdBWg@mail.gmail.com/
https://lore.kernel.org/all/CAHk-=wgs38ZrfPvy=nOwVkVzjpM3VFU1zobP37Fwd_h9iAD5JQ@mail.gmail.com/