Re: [PATCH] net/ipv6: repeat route lookup with saddr set for ECMP

[Maximilian Moehl]

On Mon Mar 30, 2026 at 9:56 AM CEST, Paolo Abeni wrote:
> On 3/29/26 11:12 AM, Maximilian Moehl wrote:
>> diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
>> index 8e2a6b28cea7..465fce51d017 100644
>> --- a/net/ipv6/ip6_output.c
>> +++ b/net/ipv6/ip6_output.c
>> @@ -1148,6 +1148,18 @@ static int ip6_dst_lookup_tail(struct net *net, const struct sock *sk,
>>  			*dst = NULL;
>>  		}
>>  
>> +		/* If ECMP was involved the initial hash was calculted
>> +		 * with saddr=:: which can result in instability
>> +		 * when it is later re-calculated with the selected
>> +		 * saddr. Lookup the route again with the chosen
>> +		 * saddr to get a stable result.
>> +		 */
>> +		if (fl6->mp_hash) {
>> +			fl6->mp_hash = 0;
>> +			dst_release(*dst);
>> +			*dst = NULL;
>> +		}
>> +
>>  		if (fl6->flowi6_oif)
>>  			flags |= RT6_LOOKUP_F_IFACE;
>>  	}
>
> This apparently breaks ipv6 fib tests (fib_tests.sh):
>
> # IPv6 multipath load balance test
> #      TEST: IPv6 multipath loadbalance                          [FAIL]
>
> see
> https://github.com/linux-netdev/nipa/wiki/How-to-run-netdev-selftests-CI-style
> on how to reproduce the tests.
>
> Also this would deserve additional testcases.

Thank you for the pointer, I will look into the tests.

> Without diving much inside the code I have the feeling this change is
> plugged into the wrong place: multipath selection logic should be
> encapsulated by fib6_select_path().

To be able to move this logic into fib6_select_path(), the call to
ip6_route_get_saddr() would probably have to move as well as we need
to set the source address after the first lookup. This would then also
set the source address as a side-effect.

I can give this a try, just want to confirm that it's the correct way
before doing so or if I'm missing something.

--
Max