On 29 March 2017 at 10:11, Eric Dumazet <eric.duma...@gmail.com> wrote:
> Should we also fix IPv6 or is this bug only about IPv4 ?

In IPv6 the second connect() returns ENETUNREACH (rather than failing
yet returning 0 as it does in IPv4).
This should probably incorrect behaviour and should the operation
should succeed.

Reply via email to