From mboxrd@z Thu Jan  1 00:00:00 1970
From: Jakub Kicinski <jakub.kicinski@netronome.com>
Subject: Re: [PATCH bpf-next v4 2/2] samples/bpf: Add xdp_sample_pkts
 example
Date: Tue, 5 Jun 2018 15:24:34 -0700
Message-ID: <20180605152434.0149b8a7@cakuba.netronome.com>
References: <152821020087.23694.8231039605257373797.stgit@alrua-kau>
        <152821020092.23694.4231922206556589059.stgit@alrua-kau>
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Cc: netdev@vger.kernel.org
To: Toke =?UTF-8?B?SMO4aWxhbmQtSsO4cmdlbnNlbg==?= <toke@toke.dk>
Return-path: <netdev-owner@vger.kernel.org>
Received: from mail-pl0-f65.google.com ([209.85.160.65]:45019 "EHLO
        mail-pl0-f65.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
        with ESMTP id S1752166AbeFEWYh (ORCPT
        <rfc822;netdev@vger.kernel.org>); Tue, 5 Jun 2018 18:24:37 -0400
Received: by mail-pl0-f65.google.com with SMTP id z9-v6so2429641plk.11
        for <netdev@vger.kernel.org>; Tue, 05 Jun 2018 15:24:37 -0700 (PDT)
In-Reply-To: <152821020092.23694.4231922206556589059.stgit@alrua-kau>
Sender: netdev-owner@vger.kernel.org
List-ID: <netdev.vger.kernel.org>

On Tue, 05 Jun 2018 16:50:00 +0200, Toke H=C3=B8iland-J=C3=B8rgensen wrote:
> Add an example program showing how to sample packets from XDP using the
> perf event buffer. The example userspace program just prints the ethernet
> header for every packet sampled.
>=20
> Signed-off-by: Toke H=C3=B8iland-J=C3=B8rgensen <toke@toke.dk>

> diff --git a/samples/bpf/xdp_sample_pkts_kern.c b/samples/bpf/xdp_sample_=
pkts_kern.c
> new file mode 100644
> index 000000000000..4560522ca015
> --- /dev/null
> +++ b/samples/bpf/xdp_sample_pkts_kern.c
> @@ -0,0 +1,62 @@
> +#include <linux/ptrace.h>
> +#include <linux/version.h>
> +#include <uapi/linux/bpf.h>
> +#include "bpf_helpers.h"
> +
> +#define SAMPLE_SIZE 64ul
> +#define MAX_CPUS 24

That may be a lil' too few for modern HW with hyper-threading on ;) =20
My development machine says:

$ ncpus
28

128, maybe?

> +#define bpf_printk(fmt, ...)					\
> +({								\
> +	       char ____fmt[] =3D fmt;				\
> +	       bpf_trace_printk(____fmt, sizeof(____fmt),	\
> +				##__VA_ARGS__);			\
> +})
> +
> +struct bpf_map_def SEC("maps") my_map =3D {
> +	.type =3D BPF_MAP_TYPE_PERF_EVENT_ARRAY,
> +	.key_size =3D sizeof(int),
> +	.value_size =3D sizeof(u32),
> +	.max_entries =3D MAX_CPUS,
> +};
> +
> +SEC("xdp_sample")
> +int xdp_sample_prog(struct xdp_md *ctx)
> +{
> +	void *data_end =3D (void *)(long)ctx->data_end;
> +	void *data =3D (void *)(long)ctx->data;
> +
> +        /* Metadata will be in the perf event before the packet data. */
> +	struct S {
> +		u16 cookie;
> +		u16 pkt_len;
> +	} __attribute__((packed)) metadata;
> +
> +	if (data + SAMPLE_SIZE < data_end) {
> +		/* The XDP perf_event_output handler will use the upper 32 bits
> +		 * of the flags argument as a number of bytes to include of the
> +		 * packet payload in the event data. If the size is too big, the
> +		 * call to bpf_perf_event_output will fail and return -EFAULT.
> +		 *
> +		 * See bpf_xdp_event_output in net/core/filter.c.
> +		 *
> +		 * The BPF_F_CURRENT_CPU flag means that the event output fd
> +		 * will be indexed by the CPU number in the event map.
> +		 */
> +		u64 flags =3D (SAMPLE_SIZE << 32) | BPF_F_CURRENT_CPU;
> +		int ret;
> +
> +		metadata.cookie =3D 0xdead;
> +		metadata.pkt_len =3D (u16)(data_end - data);
> +
> +		ret =3D bpf_perf_event_output(ctx, &my_map, flags,
> +				      &metadata, sizeof(metadata));
> +		if(ret)

Please run checkpatch --strict on the samples.

> +			bpf_printk("perf_event_output failed: %d\n", ret);
> +	}
> +
> +	return XDP_PASS;
> +}
> +
> +char _license[] SEC("license") =3D "GPL";
> +u32 _version SEC("version") =3D LINUX_VERSION_CODE;