[Cake] [PATCH net-next v3] Add Common Applications Kept Enhanced (cake) qdisc
kbuild test robot
lkp at intel.com
Fri Apr 27 06:54:59 EDT 2018
Thank you for the patch! Yet something to improve:
[auto build test ERROR on net-next/master]
config: i386-allmodconfig (attached as .config)
compiler: gcc-7 (Debian 7.3.0-16) 7.3.0
# save the attached .config to linux build tree
All errors (new ones prefixed by >>):
>> net/sched/sch_cake.c:68:10: fatal error: pkt_sched.h: No such file or directory
vim +68 net/sched/sch_cake.c
3 /* COMMON Applications Kept Enhanced (CAKE) discipline
5 * Copyright (C) 2014-2018 Jonathan Morton <chromatix99 at gmail.com>
6 * Copyright (C) 2015-2018 Toke Høiland-Jørgensen <toke at toke.dk>
7 * Copyright (C) 2014-2018 Dave Täht <dave.taht at gmail.com>
8 * Copyright (C) 2015-2018 Sebastian Moeller <moeller0 at gmx.de>
9 * (C) 2015-2018 Kevin Darbyshire-Bryant <kevin at darbyshire-bryant.me.uk>
10 * Copyright (C) 2017 Ryan Mounce <ryan at mounce.com.au>
12 * The CAKE Principles:
13 * (or, how to have your cake and eat it too)
15 * This is a combination of several shaping, AQM and FQ techniques into one
16 * easy-to-use package:
18 * - An overall bandwidth shaper, to move the bottleneck away from dumb CPE
19 * equipment and bloated MACs. This operates in deficit mode (as in sch_fq),
20 * eliminating the need for any sort of burst parameter (eg. token bucket
21 * depth). Burst support is limited to that necessary to overcome scheduling
22 * latency.
24 * - A Diffserv-aware priority queue, giving more priority to certain classes,
25 * up to a specified fraction of bandwidth. Above that bandwidth threshold,
26 * the priority is reduced to avoid starving other tins.
28 * - Each priority tin has a separate Flow Queue system, to isolate traffic
29 * flows from each other. This prevents a burst on one flow from increasing
30 * the delay to another. Flows are distributed to queues using a
31 * set-associative hash function.
33 * - Each queue is actively managed by Cobalt, which is a combination of the
34 * Codel and Blue AQM algorithms. This serves flows fairly, and signals
35 * congestion early via ECN (if available) and/or packet drops, to keep
36 * latency low. The codel parameters are auto-tuned based on the bandwidth
37 * setting, as is necessary at low bandwidths.
39 * The configuration parameters are kept deliberately simple for ease of use.
40 * Everything has sane defaults. Complete generality of configuration is *not*
41 * a goal.
43 * The priority queue operates according to a weighted DRR scheme, combined with
44 * a bandwidth tracker which reuses the shaper logic to detect which side of the
45 * bandwidth sharing threshold the tin is operating. This determines whether a
46 * priority-based weight (high) or a bandwidth-based weight (low) is used for
47 * that tin in the current pass.
49 * This qdisc was inspired by Eric Dumazet's fq_codel code, which he kindly
50 * granted us permission to leverage.
53 #include <linux/module.h>
54 #include <linux/types.h>
55 #include <linux/kernel.h>
56 #include <linux/jiffies.h>
57 #include <linux/string.h>
58 #include <linux/in.h>
59 #include <linux/errno.h>
60 #include <linux/init.h>
61 #include <linux/skbuff.h>
62 #include <linux/jhash.h>
63 #include <linux/slab.h>
64 #include <linux/vmalloc.h>
65 #include <linux/reciprocal_div.h>
66 #include <net/netlink.h>
67 #include <linux/version.h>
> 68 #include "pkt_sched.h"
69 #include <linux/if_vlan.h>
70 #include <net/pkt_sched.h>
71 #include <net/tcp.h>
72 #include <net/flow_dissector.h>
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 62952 bytes
Desc: not available
More information about the Cake