Re: [PATCH v2 net-next] tcp: allow drivers to tweak TSQ logic

2017-11-28 Thread Eric Dumazet
On Tue, 2017-11-28 at 14:10 +0100, Felix Fietkau wrote: > On 2017-11-12 00:54, Eric Dumazet wrote: > > From: Eric Dumazet > > > > I had many reports that TSQ logic breaks wifi aggregation. > > > > Current logic is to allow up to 1 ms of bytes to be queued into > > qdisc > >

Re: [PATCH v2 net-next] tcp: allow drivers to tweak TSQ logic

2017-11-28 Thread Felix Fietkau
On 2017-11-12 00:54, Eric Dumazet wrote: > From: Eric Dumazet > > I had many reports that TSQ logic breaks wifi aggregation. > > Current logic is to allow up to 1 ms of bytes to be queued into qdisc > and drivers queues. > > But Wifi aggregation needs a bigger budget to

Re: [PATCH v2 net-next] tcp: allow drivers to tweak TSQ logic

2017-11-13 Thread David Miller
From: Eric Dumazet Date: Sat, 11 Nov 2017 15:54:12 -0800 > From: Eric Dumazet > > I had many reports that TSQ logic breaks wifi aggregation. > > Current logic is to allow up to 1 ms of bytes to be queued into qdisc > and drivers queues. > > But

Re: [net-next] tcp: allow drivers to tweak TSQ logic

2017-11-13 Thread Johannes Berg
On Sat, 2017-11-11 at 15:38 -0800, Eric Dumazet wrote: > > Hm. I wish we wouldn't have to do this on every skb, but perhaps it > > doesn't matter that much. > > Yes, it does not matter, even at 40Gbit ;) :) > Same cache line already ;) Hah. It seemed so far away, but actually looking at the

Re: [net-next] tcp: allow drivers to tweak TSQ logic

2017-11-12 Thread Toke Høiland-Jørgensen
Johannes Berg writes: > Thanks Eric! > >> We expect wifi drivers to set this field to smaller values (tests have >> been done with values from 6 to 9) > > I suppose we should test each driver or so. I think that for TXQ-enabled drivers we could probably set this in

Re: [PATCH v2 net-next] tcp: allow drivers to tweak TSQ logic

2017-11-12 Thread Neal Cardwell
On Sat, Nov 11, 2017 at 6:54 PM, Eric Dumazet wrote: > > From: Eric Dumazet > > I had many reports that TSQ logic breaks wifi aggregation. > > Current logic is to allow up to 1 ms of bytes to be queued into qdisc > and drivers queues. > > But Wifi

[PATCH v2 net-next] tcp: allow drivers to tweak TSQ logic

2017-11-11 Thread Eric Dumazet
From: Eric Dumazet I had many reports that TSQ logic breaks wifi aggregation. Current logic is to allow up to 1 ms of bytes to be queued into qdisc and drivers queues. But Wifi aggregation needs a bigger budget to allow bigger rates to be discovered by various TCP

Re: [net-next] tcp: allow drivers to tweak TSQ logic

2017-11-11 Thread Eric Dumazet
On Sat, 2017-11-11 at 15:27 +0100, Johannes Berg wrote: > Thanks Eric! > > > We expect wifi drivers to set this field to smaller values (tests have > > been done with values from 6 to 9) > > I suppose we should test each driver or so. > > > They would have to use following template : > > > >

Re: [net-next] tcp: allow drivers to tweak TSQ logic

2017-11-11 Thread Johannes Berg
Thanks Eric! > We expect wifi drivers to set this field to smaller values (tests have > been done with values from 6 to 9) I suppose we should test each driver or so. > They would have to use following template : > > if (skb->sk && skb->sk->sk_pacing_shift != MY_PACING_SHIFT) >

[PATCH net-next] tcp: allow drivers to tweak TSQ logic

2017-11-09 Thread Eric Dumazet
From: Eric Dumazet I had many reports that TSQ logic breaks wifi aggregation. Current logic is to allow up to 1 ms of bytes to be queued into qdisc and drivers queues. But Wifi aggregation needs a bigger budget to allow bigger rates to be discovered by various TCP