On 24 Feb 2015, at 16:25, Eric Dumazet [off-list ref] wrote:
On Tue, 2015-02-24 at 16:07 +0200, Eyal Birger wrote:
quoted
quoted
rxpc is buggy right now anyway, as it reads skb->mark _and_ uses
sock_recv_ts_and_drops(), so skb->mark value is pretty much void.
Yes, but it does not seem to use sock_queue_rcv_skb() which sets skb->dropcount.
Well, SO_RXQ_OVFL is broken for these sockets, as we'll report a
dropcount if skb->mark was set.
Right. That feature seems broken.
Fixing it would mean either
avoiding the use of skb->mark in
rxrpc or aliasing dropcount with
something else.
I am also not sure what should be the
semantics of SO_RCVMARK on these
sockets.