From: Jesper Dangaard Brouer <brouer@redhat.com>
To: netdev@vger.kernel.org
Cc: John Fastabend <john.fastabend@gmail.com>,
Jesper Dangaard Brouer <brouer@redhat.com>
Subject: [PATCH net-next 4/7] xdp: separate xdp_redirect tracepoint in error case
Date: Tue, 29 Aug 2017 16:37:56 +0200 [thread overview]
Message-ID: <150401747611.16384.8021135230122395742.stgit@firesoul> (raw)
In-Reply-To: <150401743083.16384.15778781741742858567.stgit@firesoul>
There is a need to separate the xdp_redirect tracepoint into two
tracepoints, for separating the error case from the normal forward
case.
Due to the extreme speeds XDP is operating at, loading a tracepoint
have a measurable impact. Single core XDP REDIRECT (ethtool tuned
rx-usecs 25) can do 13.7 Mpps forwarding, but loading a simple
bpf_prog at the tracepoint (with a return 0) reduce perf to 10.2 Mpps
(CPU E5-1650 v4 @ 3.60GHz, driver: ixgbe)
The overhead of loading a bpf-based tracepoint can be calculated to
cost 25 nanosec ((1/13782002-1/10267937)*10^9 = -24.83 ns).
Using perf record on the tracepoint event, with a non-matching --filter
expression, the overhead is much larger. Performance drops to 8.3 Mpps,
cost 48 nanosec ((1/13782002-1/8312497)*10^9 = -47.74))
Having a separate tracepoint for err cases, which should be less
frequent, allow running a continuous monitor for errors while not
affecting the redirect forward performance (this have also been
verified by measurements).
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
---
include/trace/events/xdp.h | 22 ++++++++++++++++++----
net/core/filter.c | 37 ++++++++++++++++++++++++-------------
2 files changed, 42 insertions(+), 17 deletions(-)
diff --git a/include/trace/events/xdp.h b/include/trace/events/xdp.h
index 89eba564e199..1eebad55ebd4 100644
--- a/include/trace/events/xdp.h
+++ b/include/trace/events/xdp.h
@@ -94,11 +94,25 @@ DEFINE_EVENT(xdp_redirect_template, xdp_redirect,
TP_ARGS(dev, xdp, to_ifindex, err, map, map_index)
);
-#define _trace_xdp_redirect(dev, xdp, to, err) \
- trace_xdp_redirect(dev, xdp, to, err, NULL, 0);
+DEFINE_EVENT(xdp_redirect_template, xdp_redirect_err,
+ TP_PROTO(const struct net_device *dev,
+ const struct bpf_prog *xdp,
+ int to_ifindex, int err,
+ const struct bpf_map *map, u32 map_index),
+ TP_ARGS(dev, xdp, to_ifindex, err, map, map_index)
+);
+
+#define _trace_xdp_redirect(dev, xdp, to) \
+ trace_xdp_redirect(dev, xdp, to, 0, NULL, 0);
+
+#define _trace_xdp_redirect_err(dev, xdp, to, err) \
+ trace_xdp_redirect_err(dev, xdp, to, err, NULL, 0);
+
+#define trace_xdp_redirect_map(dev, xdp, fwd, map, idx) \
+ trace_xdp_redirect(dev, xdp, fwd ? fwd->ifindex : 0, 0, map, idx);
-#define trace_xdp_redirect_map(dev, xdp, fwd, err, map, idx) \
- trace_xdp_redirect(dev, xdp, fwd ? fwd->ifindex : 0, err, map, idx);
+#define trace_xdp_redirect_map_err(dev, xdp, fwd, map, idx, err) \
+ trace_xdp_redirect_err(dev, xdp, fwd ? fwd->ifindex : 0, err, map, idx);
#endif /* _TRACE_XDP_H */
diff --git a/net/core/filter.c b/net/core/filter.c
index 31eab77cc842..096e78de0b97 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -2515,16 +2515,20 @@ static int xdp_do_redirect_map(struct net_device *dev, struct xdp_buff *xdp,
fwd = __dev_map_lookup_elem(map, index);
if (!fwd) {
err = -EINVAL;
- goto out;
+ goto err;
}
if (ri->map_to_flush && ri->map_to_flush != map)
xdp_do_flush_map();
err = __bpf_tx_xdp(fwd, map, xdp, index);
- if (likely(!err))
- ri->map_to_flush = map;
-out:
- trace_xdp_redirect_map(dev, xdp_prog, fwd, err, map, index);
+ if (unlikely(err))
+ goto err;
+
+ ri->map_to_flush = map;
+ trace_xdp_redirect_map(dev, xdp_prog, fwd, map, index);
+ return 0;
+err:
+ trace_xdp_redirect_map_err(dev, xdp_prog, fwd, map, index, err);
return err;
}
@@ -2543,12 +2547,17 @@ int xdp_do_redirect(struct net_device *dev, struct xdp_buff *xdp,
ri->ifindex = 0;
if (unlikely(!fwd)) {
err = -EINVAL;
- goto out;
+ goto err;
}
err = __bpf_tx_xdp(fwd, NULL, xdp, 0);
-out:
- _trace_xdp_redirect(dev, xdp_prog, index, err);
+ if (unlikely(err))
+ goto err;
+
+ _trace_xdp_redirect(dev, xdp_prog, index);
+ return 0;
+err:
+ _trace_xdp_redirect_err(dev, xdp_prog, index, err);
return err;
}
EXPORT_SYMBOL_GPL(xdp_do_redirect);
@@ -2566,23 +2575,25 @@ int xdp_do_generic_redirect(struct net_device *dev, struct sk_buff *skb,
ri->ifindex = 0;
if (unlikely(!fwd)) {
err = -EINVAL;
- goto out;
+ goto err;
}
if (unlikely(!(fwd->flags & IFF_UP))) {
err = -ENETDOWN;
- goto out;
+ goto err;
}
len = fwd->mtu + fwd->hard_header_len + VLAN_HLEN;
if (skb->len > len) {
err = -EMSGSIZE;
- goto out;
+ goto err;
}
skb->dev = fwd;
-out:
- _trace_xdp_redirect(dev, xdp_prog, index, err);
+ _trace_xdp_redirect(dev, xdp_prog, index);
+ return 0;
+err:
+ _trace_xdp_redirect_err(dev, xdp_prog, index, err);
return err;
}
EXPORT_SYMBOL_GPL(xdp_do_generic_redirect);
next prev parent reply other threads:[~2017-08-29 14:37 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-08-29 14:37 [PATCH net-next 0/7] XDP redirect tracepoints Jesper Dangaard Brouer
2017-08-29 14:37 ` [PATCH net-next 1/7] xdp: remove redundant argument to trace_xdp_redirect Jesper Dangaard Brouer
2017-08-29 14:37 ` [PATCH net-next 2/7] xdp: tracepoint xdp_redirect also need a map argument Jesper Dangaard Brouer
2017-08-29 14:37 ` [PATCH net-next 3/7] xdp: make xdp tracepoints report bpf prog id instead of prog_tag Jesper Dangaard Brouer
2017-08-29 14:37 ` Jesper Dangaard Brouer [this message]
2017-08-29 17:02 ` [PATCH net-next 4/7] xdp: separate xdp_redirect tracepoint in error case Alexei Starovoitov
2017-08-29 14:38 ` [PATCH net-next 5/7] xdp: separate xdp_redirect tracepoint in map case Jesper Dangaard Brouer
2017-08-29 14:38 ` [PATCH net-next 6/7] samples/bpf: xdp_redirect load XDP dummy prog on TX device Jesper Dangaard Brouer
2017-08-31 10:08 ` Tariq Toukan
2017-08-31 10:15 ` Jesper Dangaard Brouer
2017-08-29 14:38 ` [PATCH net-next 7/7] samples/bpf: xdp_monitor tool based on tracepoints Jesper Dangaard Brouer
2017-08-29 17:05 ` Alexei Starovoitov
2017-08-29 17:24 ` Daniel Borkmann
2017-08-29 17:51 ` [PATCH net-next 0/7] XDP redirect tracepoints David Miller
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=150401747611.16384.8021135230122395742.stgit@firesoul \
--to=brouer@redhat.com \
--cc=john.fastabend@gmail.com \
--cc=netdev@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox