From: Jesper Dangaard Brouer <jbrouer@redhat•com>
To: Prashant Bhole <prashantbhole.linux@gmail•com>
Cc: "David S . Miller" <davem@davemloft•net>,
"Michael S . Tsirkin" <mst@redhat•com>,
Alexei Starovoitov <ast@kernel•org>,
Daniel Borkmann <daniel@iogearbox•net>,
Jesper Dangaard Brouer <hawk@kernel•org>,
Jason Wang <jasowang@redhat•com>, David Ahern <dsahern@gmail•com>,
Jakub Kicinski <jakub.kicinski@netronome•com>,
John Fastabend <john.fastabend@gmail•com>,
Toshiaki Makita <toshiaki.makita1@gmail•com>,
Martin KaFai Lau <kafai@fb•com>, Song Liu <songliubraving@fb•com>,
Yonghong Song <yhs@fb•com>, Andrii Nakryiko <andriin@fb•com>,
netdev@vger•kernel.org,
Ilias Apalodimas <ilias.apalodimas@linaro•org>
Subject: Re: [RFC net-next 11/14] tun: run XDP program in tx path
Date: Wed, 18 Dec 2019 11:07:32 +0100 [thread overview]
Message-ID: <20191218110732.33494957@carbon> (raw)
In-Reply-To: <20191218081050.10170-12-prashantbhole.linux@gmail.com>
On Wed, 18 Dec 2019 17:10:47 +0900
Prashant Bhole <prashantbhole.linux@gmail•com> wrote:
> +static u32 tun_do_xdp_tx(struct tun_struct *tun, struct tun_file *tfile,
> + struct xdp_frame *frame)
> +{
> + struct bpf_prog *xdp_prog;
> + struct tun_page tpage;
> + struct xdp_buff xdp;
> + u32 act = XDP_PASS;
> + int flush = 0;
> +
> + xdp_prog = rcu_dereference(tun->xdp_tx_prog);
> + if (xdp_prog) {
> + xdp.data_hard_start = frame->data - frame->headroom;
> + xdp.data = frame->data;
> + xdp.data_end = xdp.data + frame->len;
> + xdp.data_meta = xdp.data - frame->metasize;
You have not configured xdp.rxq, thus a BPF-prog accessing this will crash.
For an XDP TX hook, I want us to provide/give BPF-prog access to some
more information about e.g. the current tx-queue length, or TC-q number.
Question to Daniel or Alexei, can we do this and still keep BPF_PROG_TYPE_XDP?
Or is it better to introduce a new BPF prog type (enum bpf_prog_type)
for XDP TX-hook ?
To Prashant, look at net/core/filter.c in xdp_convert_ctx_access() on
how the BPF instruction rewrites are done, when accessing xdp_rxq_info.
> + act = bpf_prog_run_xdp(xdp_prog, &xdp);
> + switch (act) {
> + case XDP_PASS:
> + break;
> + case XDP_TX:
> + /* fall through */
> + case XDP_REDIRECT:
> + /* fall through */
> + default:
> + bpf_warn_invalid_xdp_action(act);
> + /* fall through */
> + case XDP_ABORTED:
> + trace_xdp_exception(tun->dev, xdp_prog, act);
> + /* fall through */
> + case XDP_DROP:
> + xdp_return_frame_rx_napi(frame);
I'm not sure that it is safe to use "napi" variant here, as you have to
be under same RX-NAPI processing loop for this to be safe.
Notice the "rx" part of the name "xdp_return_frame_rx_napi".
> + break;
> + }
> + }
> +
> + return act;
> +}
--
Best regards,
Jesper Dangaard Brouer
MSc.CS, Principal Kernel Engineer at Red Hat
LinkedIn: http://www.linkedin.com/in/brouer
next prev parent reply other threads:[~2019-12-18 10:07 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-12-18 8:10 [RFC net-next 00/14] XDP in tx path Prashant Bhole
2019-12-18 8:10 ` [RFC net-next 01/14] net: add tx path XDP support Prashant Bhole
2019-12-18 8:10 ` [RFC net-next 02/14] tools: sync kernel uapi/linux/if_link.h header Prashant Bhole
2019-12-18 8:10 ` [RFC net-next 03/14] libbpf: API for tx path XDP support Prashant Bhole
2019-12-18 18:20 ` Alexei Starovoitov
2019-12-18 8:10 ` [RFC net-next 04/14] samples/bpf: xdp1, add XDP tx support Prashant Bhole
2019-12-18 8:10 ` [RFC net-next 05/14] net: core: rename netif_receive_generic_xdp() to do_generic_xdp_core() Prashant Bhole
2019-12-18 8:10 ` [RFC net-next 06/14] net: core: export do_xdp_generic_core() Prashant Bhole
2019-12-18 8:10 ` [RFC net-next 07/14] tuntap: check tun_msg_ctl type at necessary places Prashant Bhole
2019-12-18 8:10 ` [RFC net-next 08/14] vhost_net: user tap recvmsg api to access ptr ring Prashant Bhole
2019-12-18 8:10 ` [RFC net-next 09/14] tuntap: remove usage of ptr ring in vhost_net Prashant Bhole
2019-12-18 8:10 ` [RFC net-next 10/14] tun: set tx path XDP program Prashant Bhole
2019-12-18 8:10 ` [RFC net-next 11/14] tun: run XDP program in tx path Prashant Bhole
2019-12-18 10:07 ` Jesper Dangaard Brouer [this message]
2019-12-18 11:48 ` Toke Høiland-Jørgensen
2019-12-18 16:33 ` David Ahern
2019-12-19 2:44 ` Jason Wang
2019-12-18 18:19 ` Alexei Starovoitov
2019-12-19 2:34 ` Prashant Bhole
2019-12-19 10:15 ` Toke Høiland-Jørgensen
2019-12-20 0:07 ` Prashant Bhole
2019-12-20 3:24 ` Jason Wang
2019-12-20 4:46 ` Prashant Bhole
2019-12-20 7:36 ` Jason Wang
2019-12-20 10:11 ` Toke Høiland-Jørgensen
2019-12-20 16:11 ` David Ahern
2019-12-20 22:17 ` Prashant Bhole
2019-12-23 6:05 ` Jason Wang
2019-12-23 8:09 ` Prashant Bhole
2019-12-23 8:34 ` Jason Wang
2019-12-23 11:06 ` Prashant Bhole
2019-12-18 16:29 ` David Ahern
2019-12-19 1:47 ` Prashant Bhole
2019-12-18 8:10 ` [RFC net-next 12/14] tun: add a way to inject tx path packet into Rx path Prashant Bhole
2019-12-18 8:10 ` [RFC net-next 13/14] tun: handle XDP_TX action of tx path XDP program Prashant Bhole
2019-12-18 8:10 ` [RFC net-next 14/14] tun: run xdp prog when tun is read from file interface Prashant Bhole
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=20191218110732.33494957@carbon \
--to=jbrouer@redhat$(echo .)com \
--cc=andriin@fb$(echo .)com \
--cc=ast@kernel$(echo .)org \
--cc=daniel@iogearbox$(echo .)net \
--cc=davem@davemloft$(echo .)net \
--cc=dsahern@gmail$(echo .)com \
--cc=hawk@kernel$(echo .)org \
--cc=ilias.apalodimas@linaro$(echo .)org \
--cc=jakub.kicinski@netronome$(echo .)com \
--cc=jasowang@redhat$(echo .)com \
--cc=john.fastabend@gmail$(echo .)com \
--cc=kafai@fb$(echo .)com \
--cc=mst@redhat$(echo .)com \
--cc=netdev@vger$(echo .)kernel.org \
--cc=prashantbhole.linux@gmail$(echo .)com \
--cc=songliubraving@fb$(echo .)com \
--cc=toshiaki.makita1@gmail$(echo .)com \
--cc=yhs@fb$(echo .)com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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