Eric Dumazet <eric.duma...@gmail.com> writes: > On 04/25/2018 11:34 AM, Toke Høiland-Jørgensen wrote: >> Eric Dumazet <eric.duma...@gmail.com> writes: >> >>> On 04/25/2018 09:52 AM, Jonathan Morton wrote: >>>>> We can see here the high cost of forcing software GSO :/ >>>>> >>>>> Really, this should be done only : >>>>> 1) If requested by the admin ( tc .... gso ....) >>>>> >>>>> 2) If packet size is above a threshold. >>>>> The threshold could be set by the admin, and/or based on a fraction of >>>>> the bandwidth parameter. >>>>> >>>>> I totally understand why you prefer to segment yourself for < 100 Mbit >>>>> links. >>>>> >>>>> But this makes no sense on 10Gbit+ >>>> >>>> It is absolutely necessary, so far as I can see, to segment GSO >>>> superpackets when overhead compensation is selected - as it very >>>> often should be, even on pure Ethernet links. Without that, the >>>> calculation of link occupancy time will be wrong. (The actual >>>> transmission time of an Ethernet frame is rather more than just 14 >>>> bytes longer than the underlying IP packet.) >>> >>> Just fix the overhead compensation computation in the code. >>> >>> skb in a qdisc have everything you need. >>> >>> qdisc_pkt_len_init() has initialized qdisc_skb_cb(skb)->pkt_len with >>> the exact bytes on the wire, and you have gso_segs to perform any >>> adjustement you need to do. >> >> The problem is that may not be the right values. For example, in many >> CPEs there's a built-in switch that strips VLAN tags before the packet >> actually hits the wire. So we do need to be able to get the actual >> packet size. Is it possible to get the sizes of the individual segments >> of a GSO packet? That way we could do the calculation for the whole >> super-packet... > > All segments of GSO packets have the same size, by definition. > > Only the last segment might be smaller, and again this can be inferred > from gso_size and gso_segs
Gotcha. Until we are confident that we've implemented this in a way that works, will this do? @@ -88,6 +88,7 @@ #define CAKE_SET_WAYS (8) #define CAKE_MAX_TINS (8) #define CAKE_QUEUES (1024) +#define CAKE_SPLIT_GSO_THRESHOLD (125000000) /* 1Gbps */ @@ -1437,7 +1439,7 @@ static s32 cake_enqueue(struct sk_buff *skb, struct Qdisc *sch, * or if we need to know individual packet sizes for framing overhead. */ - if (skb_is_gso(skb)) { + if (skb_is_gso(skb) && q->rate_flags & CAKE_FLAG_SPLIT_GSO) { struct sk_buff *segs, *nskb; netdev_features_t features = netif_skb_features(skb); /* signed slen to handle corner case @@ -2337,6 +2339,12 @@ static int cake_change(struct Qdisc *sch, struct nlattr *opt, if (tb[TCA_CAKE_MEMORY]) q->buffer_config_limit = nla_get_u32(tb[TCA_CAKE_MEMORY]); + if (q->rate_bps && (q->rate_bps <= CAKE_SPLIT_GSO_THRESHOLD || + q->rate_flags & CAKE_FLAG_OVERHEAD)) + q->rate_flags |= CAKE_FLAG_SPLIT_GSO; + else + q->rate_flags &= ~CAKE_FLAG_SPLIT_GSO; + if (q->tins) { sch_tree_lock(sch); cake_reconfigure(sch); _______________________________________________ Cake mailing list Cake@lists.bufferbloat.net https://lists.bufferbloat.net/listinfo/cake