On Tue, 2019-04-16 at 21:13 +0800, Herbert Xu wrote: > On Tue, Apr 16, 2019 at 11:39:56AM +0200, Johannes Berg wrote: > > > > > not to mention > > > breaking certain applications that rely on congestion feedback? > > > > This I don't understand. The congestion feedback happens through socket > > buffer space etc. which is still there (as long as nobody sneaks in an > > skb_orphan() call) > > The congestion control happens at two levels. You are right that > the socket buffer acts as one limit. However, other applications > may also rely on the TX queue being full as the throttle (by setting > a sufficiently large socket buffer size). I'm not sure how they'd even realize this, other than packets getting dropped? Which of course we do here as well, we didn't invent something that let's us expand memory arbitrarily ;-) johannes
