Re: Re: Re: possible deadlock in cma_netdev_callback
From: Jason Gunthorpe <jgg@ziepe.ca>
Date: 2020-02-28 16:49:56
Also in:
linux-rdma, lkml
From: Jason Gunthorpe <jgg@ziepe.ca>
Date: 2020-02-28 16:49:56
Also in:
linux-rdma, lkml
On Fri, Feb 28, 2020 at 04:42:02PM +0000, Bernard Metzler wrote:
Well, right, marking a socket via setsockopt SO_BINDTODEVICE does not work - I get -EPERM. Maybe works only from user land since the ifname gets copied in from there. What I tested as working is nailing the scope of wildcard listen via: s->sk->sk_bound_dev_if = netdev->ifindex;
That sounds potentially right
I am not sure what is the right way of limiting the scope of a socket to one interface in kernel mode. Is above line the way to go, or do I miss an interface to do such things? Anybody could help?
I didn't find an alternative, but not a lot of places touching this outside the implementators of a socket type. Jason