summaryrefslogtreecommitdiffstats
path: root/samples/bpf/tcp_bpf.readme
blob: fee746621aecec98301ca5cbc125ebb61e4f8f29 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
This file describes how to run the tcp_*_kern.o tcp_bpf (or socket_ops)
programs. These programs attach to a cgroupv2. The following commands create
a cgroupv2 and attach a bash shell to the group.

  mkdir -p /tmp/cgroupv2
  mount -t cgroup2 none /tmp/cgroupv2
  mkdir -p /tmp/cgroupv2/foo
  bash
  echo $$ >> /tmp/cgroupv2/foo/cgroup.procs

Anything that runs under this shell belongs to the foo cgroupv2. To load
(attach) one of the tcp_*_kern.o programs:

  bpftool prog load tcp_basertt_kern.o /sys/fs/bpf/tcp_prog
  bpftool cgroup attach /tmp/cgroupv2/foo sock_ops pinned /sys/fs/bpf/tcp_prog
  bpftool prog tracelog

"bpftool prog tracelog" will continue to run printing the BPF log buffer.
The tcp_*_kern.o programs use special print functions to print logging
information (if enabled by the ifdef).

If using netperf/netserver to create traffic, you need to run them under the
cgroupv2 to which the BPF programs are attached (i.e. under bash shell
attached to the cgroupv2).

To remove (unattach) a socket_ops BPF program from a cgroupv2:

  bpftool cgroup attach /tmp/cgroupv2/foo sock_ops pinned /sys/fs/bpf/tcp_prog