Re: [PATCH net-next 2/4] net: ipv4: Add a sysctl to set multipath hash seed

["Toke Høiland-Jørgensen" <toke@redhat.com>]

Eric Dumazet <edumazet@google.com> writes:

> On Mon, Jun 3, 2024 at 9:30 AM Toke Høiland-Jørgensen <toke@redhat.com> wrote:
>>
>> Eric Dumazet <edumazet@google.com> writes:
>>
>> > On Wed, May 29, 2024 at 1:21 PM Petr Machata <petrm@nvidia.com> wrote:
>> >>
>> >> When calculating hashes for the purpose of multipath forwarding, both IPv4
>> >> and IPv6 code currently fall back on flow_hash_from_keys(). That uses a
>> >> randomly-generated seed. That's a fine choice by default, but unfortunately
>> >> some deployments may need a tighter control over the seed used.
>> >>
>> >> In this patch, make the seed configurable by adding a new sysctl key,
>> >> net.ipv4.fib_multipath_hash_seed to control the seed. This seed is used
>> >> specifically for multipath forwarding and not for the other concerns that
>> >> flow_hash_from_keys() is used for, such as queue selection. Expose the knob
>> >> as sysctl because other such settings, such as headers to hash, are also
>> >> handled that way. Like those, the multipath hash seed is a per-netns
>> >> variable.
>> >>
>> >> Despite being placed in the net.ipv4 namespace, the multipath seed sysctl
>> >> is used for both IPv4 and IPv6, similarly to e.g. a number of TCP
>> >> variables.
>> >>
>> > ...
>> >
>> >> +       rtnl_lock();
>> >> +       old = rcu_replace_pointer_rtnl(net->ipv4.sysctl_fib_multipath_hash_seed,
>> >> +                                      mphs);
>> >> +       rtnl_unlock();
>> >> +
>> >
>> > In case you keep RCU for the next version, please do not use rtnl_lock() here.
>> >
>> > A simple xchg() will work just fine.
>> >
>> > old = xchg((__force struct struct sysctl_fib_multipath_hash_seed
>> > **)&net->ipv4.sysctl_fib_multipath_hash_seed,
>> >                  mphs);
>>
>> We added a macro to do this kind of thing without triggering any of the
>> RCU type linter warnings, in:
>>
>> 76c8eaafe4f0 ("rcu: Create an unrcu_pointer() to remove __rcu from a pointer")
>>
>> So as an alternative to open-coding the cast, something like this could
>> work - I guess it's mostly a matter of taste:
>>
>> old = unrcu_pointer(xchg(&net->ipv4.sysctl_fib_multipath_hash_seed, RCU_INITIALIZER(mphs)));
>
> Good to know, thanks.
>
> Not sure why __kernel qualifier has been put there.

Not sure either. Paul, care to enlighten us? :)

-Toke