[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <f3cbb565-6275-f1c6-e215-f4723ac965bd@gmail.com>
Date: Tue, 1 May 2018 12:25:24 -0600
From: David Ahern <dsahern@...il.com>
To: Vincent Bernat <bernat@...fy.cx>
Cc: netdev@...r.kernel.org, borkmann@...earbox.net, ast@...nel.org,
davem@...emloft.net, shm@...ulusnetworks.com,
roopa@...ulusnetworks.com, brouer@...hat.com, toke@...e.dk,
john.fastabend@...il.com
Subject: Re: [RFC v2 bpf-next 5/9] net/ipv6: Add fib6_lookup
On 5/1/18 12:15 PM, Vincent Bernat wrote:
> ❦ 29 avril 2018 11:07 -0700, David Ahern <dsahern@...il.com> :
>
>> +struct fib6_info *fib6_lookup(struct net *net, int oif, struct flowi6 *fl6,
>> + int flags)
>
> Maybe an EXPORT_SYMBOL_GPL? There is one for __fib_lookup (fib_lookup is
> an inline function).
>
There needs to be an in-tree user relying on it. My intention is to
convert rt6_lookup to fib6_lookup; as I recall all of those just want
the egress device and do not need a dst based response. That might
provide the in-tree user - depending on how the conversion is done.
And I presume you are asking for your benchmark modules. I can send you
my diffs and results.
Powered by blists - more mailing lists