diff options
author | Lorenzo Bianconi <lorenzo@kernel.org> | 2020-07-14 15:56:39 +0200 |
---|---|---|
committer | Daniel Borkmann <daniel@iogearbox.net> | 2020-07-16 17:00:32 +0200 |
commit | 28b1520ebf81ced970141640d90279ac7b9f1f9a (patch) | |
tree | de19ec6e4d9ead265f11cceb9de9bf96d94475ec /kernel/bpf | |
parent | 9216477449f33cdbc9c9a99d49f500b7fbb81702 (diff) | |
download | linux-stable-28b1520ebf81ced970141640d90279ac7b9f1f9a.tar.gz linux-stable-28b1520ebf81ced970141640d90279ac7b9f1f9a.tar.bz2 linux-stable-28b1520ebf81ced970141640d90279ac7b9f1f9a.zip |
bpf: cpumap: Implement XDP_REDIRECT for eBPF programs attached to map entries
Introduce XDP_REDIRECT support for eBPF programs attached to cpumap
entries.
This patch has been tested on Marvell ESPRESSObin using a modified
version of xdp_redirect_cpu sample in order to attach a XDP program
to CPUMAP entries to perform a redirect on the mvneta interface.
In particular the following scenario has been tested:
rq (cpu0) --> mvneta - XDP_REDIRECT (cpu0) --> CPUMAP - XDP_REDIRECT (cpu1) --> mvneta
$./xdp_redirect_cpu -p xdp_cpu_map0 -d eth0 -c 1 -e xdp_redirect \
-f xdp_redirect_kern.o -m tx_port -r eth0
tx: 285.2 Kpps rx: 285.2 Kpps
Attaching a simple XDP program on eth0 to perform XDP_TX gives
comparable results:
tx: 288.4 Kpps rx: 288.4 Kpps
Co-developed-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Jesper Dangaard Brouer <brouer@redhat.com>
Link: https://lore.kernel.org/bpf/2cf8373a731867af302b00c4ff16c122630c4980.1594734381.git.lorenzo@kernel.org
Diffstat (limited to 'kernel/bpf')
-rw-r--r-- | kernel/bpf/cpumap.c | 17 |
1 files changed, 15 insertions, 2 deletions
diff --git a/kernel/bpf/cpumap.c b/kernel/bpf/cpumap.c index b3a8aea81ee5..4c95d0615ca2 100644 --- a/kernel/bpf/cpumap.c +++ b/kernel/bpf/cpumap.c @@ -239,7 +239,7 @@ static int cpu_map_bpf_prog_run_xdp(struct bpf_cpu_map_entry *rcpu, if (!rcpu->prog) return n; - rcu_read_lock(); + rcu_read_lock_bh(); xdp_set_return_frame_no_direct(); xdp.rxq = &rxq; @@ -267,6 +267,16 @@ static int cpu_map_bpf_prog_run_xdp(struct bpf_cpu_map_entry *rcpu, stats->pass++; } break; + case XDP_REDIRECT: + err = xdp_do_redirect(xdpf->dev_rx, &xdp, + rcpu->prog); + if (unlikely(err)) { + xdp_return_frame(xdpf); + stats->drop++; + } else { + stats->redirect++; + } + break; default: bpf_warn_invalid_xdp_action(act); /* fallthrough */ @@ -277,9 +287,12 @@ static int cpu_map_bpf_prog_run_xdp(struct bpf_cpu_map_entry *rcpu, } } + if (stats->redirect) + xdp_do_flush_map(); + xdp_clear_return_frame_no_direct(); - rcu_read_unlock(); + rcu_read_unlock_bh(); /* resched point, may call do_softirq() */ return nframes; } |