bpf.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: John Fastabend <john.fastabend@gmail.com>
To: Lorenzo Bianconi <lorenzo@kernel.org>,
	bpf@vger.kernel.org, netdev@vger.kernel.org
Cc: lorenzo.bianconi@redhat.com, davem@davemloft.net,
	kuba@kernel.org, ast@kernel.org, daniel@iogearbox.net,
	shayagr@amazon.com, john.fastabend@gmail.com, dsahern@kernel.org,
	brouer@redhat.com, echaudro@redhat.com, jasowang@redhat.com,
	alexander.duyck@gmail.com, saeed@kernel.org,
	maciej.fijalkowski@intel.com, magnus.karlsson@intel.com,
	tirthendu.sarkar@intel.com, toke@redhat.com
Subject: RE: [PATCH v12 bpf-next 00/18] mvneta: introduce XDP multi-buffer support
Date: Tue, 31 Aug 2021 17:45:10 -0700	[thread overview]
Message-ID: <612ecd169d9c7_6b87208d7@john-XPS-13-9370.notmuch> (raw)
In-Reply-To: <cover.1629473233.git.lorenzo@kernel.org>

Lorenzo Bianconi wrote:
> This series introduce XDP multi-buffer support. The mvneta driver is
> the first to support these new "non-linear" xdp_{buff,frame}. Reviewers
> please focus on how these new types of xdp_{buff,frame} packets
> traverse the different layers and the layout design. It is on purpose
> that BPF-helpers are kept simple, as we don't want to expose the
> internal layout to allow later changes.
> 
> The main idea for the new multi-buffer layout is to reuse the same
> structure used for non-linear SKB. This rely on the "skb_shared_info"
> struct at the end of the first buffer to link together subsequent
> buffers. Keeping the layout compatible with SKBs is also done to ease
> and speedup creating a SKB from an xdp_{buff,frame}.
> Converting xdp_frame to SKB and deliver it to the network stack is shown
> in patch 05/18 (e.g. cpumaps).
> 
> A multi-buffer bit (mb) has been introduced in the flags field of xdp_{buff,frame}
> structure to notify the bpf/network layer if this is a xdp multi-buffer frame
> (mb = 1) or not (mb = 0).
> The mb bit will be set by a xdp multi-buffer capable driver only for
> non-linear frames maintaining the capability to receive linear frames
> without any extra cost since the skb_shared_info structure at the end
> of the first buffer will be initialized only if mb is set.
> Moreover the flags field in xdp_{buff,frame} will be reused even for
> xdp rx csum offloading in future series.
> 

The series is looking really close to me. Couple small comments/questions
inline. Also I think we should call out the potential issues in the cover
letter with regards to backwards compatibility. Something like, 

"
A multi-buffer enabled NIC may receive XDP frames with multiple frags.
If a BPF program does not understand mb layouts its possible to contrive
a BPF program that incorrectly views data_end as the end of data when
there is more data in the payload. Note helpers will generally due the
correct thing, for example perf_output will consume entire payload. But,
it is still possible some programs could do the wrong thing even if in
an edge case. Although we expect most BPF programs not to be impacted
we can't rule out, you've been warned.
"

I can't think of an elegant way around this and it does require at least
some type of opt-in by increasing the MTU limit so I'm OK with it given
I think it should impact few (no?) real programs.

> Typical use cases for this series are:
> - Jumbo-frames
> - Packet header split (please see Google���s use-case @ NetDevConf 0x14, [0])
> - TSO/GRO
> 
> The two following ebpf helpers (and related selftests) has been introduced:
> - bpf_xdp_adjust_data:
>   Move xdp_md->data and xdp_md->data_end pointers in subsequent fragments
>   according to the offset provided by the ebpf program. This helper can be
>   used to read/write values in frame payload.
> - bpf_xdp_get_buff_len:
>   Return the total frame size (linear + paged parts)
> 
> bpf_xdp_adjust_tail and bpf_xdp_copy helpers have been modified to take into
> account xdp multi-buff frames.
> 
> More info about the main idea behind this approach can be found here [1][2].
> 
> Changes since v11:
> - add missing static to bpf_xdp_get_buff_len_proto structure
> - fix bpf_xdp_adjust_data helper when offset is smaller than linear area length.
> 
> Changes since v10:
> - move xdp->data to the requested payload offset instead of to the beginning of
>   the fragment in bpf_xdp_adjust_data()
> 
> Changes since v9:
> - introduce bpf_xdp_adjust_data helper and related selftest
> - add xdp_frags_size and xdp_frags_tsize fields in skb_shared_info
> - introduce xdp_update_skb_shared_info utility routine in ordere to not reset
>   frags array in skb_shared_info converting from a xdp_buff/xdp_frame to a skb 
> - simplify bpf_xdp_copy routine
> 
> Changes since v8:
> - add proper dma unmapping if XDP_TX fails on mvneta for a xdp multi-buff
> - switch back to skb_shared_info implementation from previous xdp_shared_info
>   one
> - avoid using a bietfield in xdp_buff/xdp_frame since it introduces performance
>   regressions. Tested now on 10G NIC (ixgbe) to verify there are no performance
>   penalties for regular codebase
> - add bpf_xdp_get_buff_len helper and remove frame_length field in xdp ctx
> - add data_len field in skb_shared_info struct
> - introduce XDP_FLAGS_FRAGS_PF_MEMALLOC flag
> 
> Changes since v7:
> - rebase on top of bpf-next
> - fix sparse warnings
> - improve comments for frame_length in include/net/xdp.h
> 
> Changes since v6:
> - the main difference respect to previous versions is the new approach proposed
>   by Eelco to pass full length of the packet to eBPF layer in XDP context
> - reintroduce multi-buff support to eBPF kself-tests
> - reintroduce multi-buff support to bpf_xdp_adjust_tail helper
> - introduce multi-buffer support to bpf_xdp_copy helper
> - rebase on top of bpf-next
> 
> Changes since v5:
> - rebase on top of bpf-next
> - initialize mb bit in xdp_init_buff() and drop per-driver initialization
> - drop xdp->mb initialization in xdp_convert_zc_to_xdp_frame()
> - postpone introduction of frame_length field in XDP ctx to another series
> - minor changes
> 
> Changes since v4:
> - rebase ontop of bpf-next
> - introduce xdp_shared_info to build xdp multi-buff instead of using the
>   skb_shared_info struct
> - introduce frame_length in xdp ctx
> - drop previous bpf helpers
> - fix bpf_xdp_adjust_tail for xdp multi-buff
> - introduce xdp multi-buff self-tests for bpf_xdp_adjust_tail
> - fix xdp_return_frame_bulk for xdp multi-buff
> 
> Changes since v3:
> - rebase ontop of bpf-next
> - add patch 10/13 to copy back paged data from a xdp multi-buff frame to
>   userspace buffer for xdp multi-buff selftests
> 
> Changes since v2:
> - add throughput measurements
> - drop bpf_xdp_adjust_mb_header bpf helper
> - introduce selftest for xdp multibuffer
> - addressed comments on bpf_xdp_get_frags_count
> - introduce xdp multi-buff support to cpumaps
> 
> Changes since v1:
> - Fix use-after-free in xdp_return_{buff/frame}
> - Introduce bpf helpers
> - Introduce xdp_mb sample program
> - access skb_shared_info->nr_frags only on the last fragment
> 
> Changes since RFC:
> - squash multi-buffer bit initialization in a single patch
> - add mvneta non-linear XDP buff support for tx side
> 
> [0] https://netdevconf.info/0x14/session.html?talk-the-path-to-tcp-4k-mtu-and-rx-zerocopy
> [1] https://github.com/xdp-project/xdp-project/blob/master/areas/core/xdp-multi-buffer01-design.org
> [2] https://netdevconf.info/0x14/session.html?tutorial-add-XDP-support-to-a-NIC-driver (XDPmulti-buffers section)
> 
> Eelco Chaudron (3):
>   bpf: add multi-buff support to the bpf_xdp_adjust_tail() API
>   bpf: add multi-buffer support to xdp copy helpers
>   bpf: update xdp_adjust_tail selftest to include multi-buffer
> 
> Lorenzo Bianconi (15):
>   net: skbuff: add size metadata to skb_shared_info for xdp
>   xdp: introduce flags field in xdp_buff/xdp_frame
>   net: mvneta: update mb bit before passing the xdp buffer to eBPF layer
>   net: mvneta: simplify mvneta_swbm_add_rx_fragment management
>   net: xdp: add xdp_update_skb_shared_info utility routine
>   net: marvell: rely on xdp_update_skb_shared_info utility routine
>   xdp: add multi-buff support to xdp_return_{buff/frame}
>   net: mvneta: add multi buffer support to XDP_TX
>   net: mvneta: enable jumbo frames for XDP
>   bpf: introduce bpf_xdp_get_buff_len helper
>   bpf: move user_size out of bpf_test_init
>   bpf: introduce multibuff support to bpf_prog_test_run_xdp()
>   bpf: test_run: add xdp_shared_info pointer in bpf_test_finish
>     signature
>   net: xdp: introduce bpf_xdp_adjust_data helper
>   bpf: add bpf_xdp_adjust_data selftest
> 
>  drivers/net/ethernet/marvell/mvneta.c         | 204 ++++++++++-------
>  include/linux/skbuff.h                        |   6 +-
>  include/net/xdp.h                             |  95 +++++++-
>  include/uapi/linux/bpf.h                      |  39 ++++
>  kernel/trace/bpf_trace.c                      |   3 +
>  net/bpf/test_run.c                            | 117 ++++++++--
>  net/core/filter.c                             | 213 +++++++++++++++++-
>  net/core/xdp.c                                |  76 ++++++-
>  tools/include/uapi/linux/bpf.h                |  39 ++++
>  .../bpf/prog_tests/xdp_adjust_data.c          |  55 +++++
>  .../bpf/prog_tests/xdp_adjust_tail.c          | 118 ++++++++++
>  .../selftests/bpf/prog_tests/xdp_bpf2bpf.c    | 151 +++++++++----
>  .../bpf/progs/test_xdp_adjust_tail_grow.c     |  10 +-
>  .../bpf/progs/test_xdp_adjust_tail_shrink.c   |  32 ++-
>  .../selftests/bpf/progs/test_xdp_bpf2bpf.c    |   2 +-
>  .../bpf/progs/test_xdp_update_frags.c         |  41 ++++
>  16 files changed, 1036 insertions(+), 165 deletions(-)
>  create mode 100644 tools/testing/selftests/bpf/prog_tests/xdp_adjust_data.c
>  create mode 100644 tools/testing/selftests/bpf/progs/test_xdp_update_frags.c
> 
> -- 
> 2.31.1
> 



  parent reply	other threads:[~2021-09-01  0:45 UTC|newest]

Thread overview: 38+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-08-20 15:40 [PATCH v12 bpf-next 00/18] mvneta: introduce XDP multi-buffer support Lorenzo Bianconi
2021-08-20 15:40 ` [PATCH v12 bpf-next 01/18] net: skbuff: add size metadata to skb_shared_info for xdp Lorenzo Bianconi
2021-08-31 23:13   ` John Fastabend
2021-09-03 17:13     ` Lorenzo Bianconi
2021-08-20 15:40 ` [PATCH v12 bpf-next 02/18] xdp: introduce flags field in xdp_buff/xdp_frame Lorenzo Bianconi
2021-08-31 23:15   ` John Fastabend
2021-08-20 15:40 ` [PATCH v12 bpf-next 03/18] net: mvneta: update mb bit before passing the xdp buffer to eBPF layer Lorenzo Bianconi
2021-08-31 23:23   ` John Fastabend
2021-09-03 17:23     ` Lorenzo Bianconi
2021-08-20 15:40 ` [PATCH v12 bpf-next 04/18] net: mvneta: simplify mvneta_swbm_add_rx_fragment management Lorenzo Bianconi
2021-08-31 23:34   ` John Fastabend
2021-08-20 15:40 ` [PATCH v12 bpf-next 05/18] net: xdp: add xdp_update_skb_shared_info utility routine Lorenzo Bianconi
2021-08-31 23:38   ` John Fastabend
2021-08-20 15:40 ` [PATCH v12 bpf-next 06/18] net: marvell: rely on " Lorenzo Bianconi
2021-08-31 23:41   ` John Fastabend
2021-09-03 17:27     ` Lorenzo Bianconi
2021-08-20 15:40 ` [PATCH v12 bpf-next 07/18] xdp: add multi-buff support to xdp_return_{buff/frame} Lorenzo Bianconi
2021-08-31 23:43   ` John Fastabend
2021-08-20 15:40 ` [PATCH v12 bpf-next 08/18] net: mvneta: add multi buffer support to XDP_TX Lorenzo Bianconi
2021-08-31 23:44   ` John Fastabend
2021-08-20 15:40 ` [PATCH v12 bpf-next 09/18] net: mvneta: enable jumbo frames for XDP Lorenzo Bianconi
2021-08-31 23:45   ` John Fastabend
2021-08-20 15:40 ` [PATCH v12 bpf-next 10/18] bpf: add multi-buff support to the bpf_xdp_adjust_tail() API Lorenzo Bianconi
2021-09-01  0:10   ` John Fastabend
2021-08-20 15:40 ` [PATCH v12 bpf-next 11/18] bpf: introduce bpf_xdp_get_buff_len helper Lorenzo Bianconi
2021-09-01  0:12   ` John Fastabend
2021-08-20 15:40 ` [PATCH v12 bpf-next 12/18] bpf: add multi-buffer support to xdp copy helpers Lorenzo Bianconi
2021-09-01  0:19   ` John Fastabend
2021-08-20 15:40 ` [PATCH v12 bpf-next 13/18] bpf: move user_size out of bpf_test_init Lorenzo Bianconi
2021-08-20 15:40 ` [PATCH v12 bpf-next 14/18] bpf: introduce multibuff support to bpf_prog_test_run_xdp() Lorenzo Bianconi
2021-08-20 15:40 ` [PATCH v12 bpf-next 15/18] bpf: test_run: add xdp_shared_info pointer in bpf_test_finish signature Lorenzo Bianconi
2021-08-20 15:40 ` [PATCH v12 bpf-next 16/18] bpf: update xdp_adjust_tail selftest to include multi-buffer Lorenzo Bianconi
2021-08-20 15:40 ` [PATCH v12 bpf-next 17/18] net: xdp: introduce bpf_xdp_adjust_data helper Lorenzo Bianconi
2021-09-01  0:36   ` John Fastabend
2021-09-03 17:57     ` Lorenzo Bianconi
2021-08-20 15:40 ` [PATCH v12 bpf-next 18/18] bpf: add bpf_xdp_adjust_data selftest Lorenzo Bianconi
2021-09-01  0:45 ` John Fastabend [this message]
2021-09-07  8:35   ` [PATCH v12 bpf-next 00/18] mvneta: introduce XDP multi-buffer support Lorenzo Bianconi

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=612ecd169d9c7_6b87208d7@john-XPS-13-9370.notmuch \
    --to=john.fastabend@gmail.com \
    --cc=alexander.duyck@gmail.com \
    --cc=ast@kernel.org \
    --cc=bpf@vger.kernel.org \
    --cc=brouer@redhat.com \
    --cc=daniel@iogearbox.net \
    --cc=davem@davemloft.net \
    --cc=dsahern@kernel.org \
    --cc=echaudro@redhat.com \
    --cc=jasowang@redhat.com \
    --cc=kuba@kernel.org \
    --cc=lorenzo.bianconi@redhat.com \
    --cc=lorenzo@kernel.org \
    --cc=maciej.fijalkowski@intel.com \
    --cc=magnus.karlsson@intel.com \
    --cc=netdev@vger.kernel.org \
    --cc=saeed@kernel.org \
    --cc=shayagr@amazon.com \
    --cc=tirthendu.sarkar@intel.com \
    --cc=toke@redhat.com \
    /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;
as well as URLs for NNTP newsgroup(s).