[PATCH ipsec-next v2 0/17] Add IP-TFS mode to xfrm

[Christian Hopps <chopps@chopps.org>]

From: Christian Hopps <chopps@labn.net>

Summary of Changes
------------------

This patchset adds a new xfrm mode implementing on-demand IP-TFS. IP-TFS
(AggFrag encapsulation) has been standardized in RFC9347.

  Link: https://www.rfc-editor.org/rfc/rfc9347.txt

This feature supports demand driven (i.e., non-constant send rate)
IP-TFS to take advantage of the AGGFRAG ESP payload encapsulation. This
payload type supports aggregation and fragmentation of the inner IP
packet stream which in turn yields higher small-packet bandwidth as well
as reducing MTU/PMTU issues. Congestion control is unimplementated as
the send rate is demand driven rather than constant.

In order to allow loading this fucntionality as a module a set of
callbacks xfrm_mode_cbs has been added to xfrm as well.

Patchset Changes:
-----------------

  23 files changed, 3252 insertions(+), 19 deletions(-)
  Documentation/networking/xfrm_sysctl.rst |   30 +
  include/net/netns/xfrm.h                 |    6 +
  include/net/xfrm.h                       |   40 +
  include/uapi/linux/in.h                  |    2 +
  include/uapi/linux/ip.h                  |   16 +
  include/uapi/linux/ipsec.h               |    3 +-
  include/uapi/linux/snmp.h                |    3 +
  include/uapi/linux/xfrm.h                |    9 +-
  net/ipv4/esp4.c                          |    3 +-
  net/ipv6/esp6.c                          |    3 +-
  net/netfilter/nft_xfrm.c                 |    3 +-
  net/xfrm/Makefile                        |    1 +
  net/xfrm/trace_iptfs.h                   |  218 +++
  net/xfrm/xfrm_compat.c                   |   10 +-
  net/xfrm/xfrm_device.c                   |    4 +-
  net/xfrm/xfrm_input.c                    |   14 +-
  net/xfrm/xfrm_iptfs.c                    | 2741 ++++++++++++++++++++++++++++++
  net/xfrm/xfrm_output.c                   |    6 +
  net/xfrm/xfrm_policy.c                   |   26 +-
  net/xfrm/xfrm_proc.c                     |    3 +
  net/xfrm/xfrm_state.c                    |   60 +
  net/xfrm/xfrm_sysctl.c                   |   38 +
  net/xfrm/xfrm_user.c                     |   32 +

Patchset Structure:
-------------------

The first 8 commits are changes to the xfrm infrastructure to support
the callbacks as well as more generic IP-TFS additions that may be used
outside the actual IP-TFS implementation.

  - iptfs: config: add CONFIG_XFRM_IPTFS
  - iptfs: uapi: ip: add ip_tfs_*_hdr packet formats
  - iptfs: uapi: IPPROTO_AGGFRAG AGGFRAG in ESP
  - iptfs: sysctl: allow configuration of global default values
  - iptfs: netlink: add config (netlink) options
  - iptfs: xfrm: Add mode_cbs module functionality
  - iptfs: xfrm: add generic iptfs defines and functionality

The last 9+1 commits constitute the IP-TFS implementation constructed in
layers to make review easier. The first 9 commits all apply to a single
file `net/xfrm/xfrm_iptfs.c`, the last commit adds a new tracepoint
header file along with the use of these new tracepoint calls.

  - iptfs: impl: add new iptfs xfrm mode impl
  - iptfs: impl: add user packet (tunnel ingress) handling
  - iptfs: impl: share page fragments of inner packets
  - iptfs: impl: add fragmenting of larger than MTU user packets
  - iptfs: impl: add basic receive packet (tunnel egress) handling
  - iptfs: impl: handle received fragmented inner packets
  - iptfs: impl: add reusing received skb for the tunnel egress packet
  - iptfs: impl: add skb-fragment sharing code
  - iptfs: impl: handle reordering of received packets
  - iptfs: impl: add tracepoint functionality

Patchset History:
-----------------

RFCv1 (11/10/2023)

RFCv1 -> RFCv2 (11/12/2023)

  Updates based on feedback from Simon Horman, Antony,
  Michael Richardson, and kernel test robot.

RFCv2 -> v1 (2/19/2024)

  Updates based on feedback from Sabrina Dubroca, kernel test robot

v1 -> v2 (5/19/2024)

  Updates based on feedback from Sabrina Dubroca, Simon Horman, Antony.

  o Add handling of new netlink SA direction attribute (Antony).
  o Split single patch/commit of xfrm_iptfs.c (the actual IP-TFS impl)
    into 9+1 distinct layered functionality commits for aiding review.
  - xfrm: fix return check on clone() callback
  - xfrm: add sa_len() callback in xfrm_mode_cbs for copy to user
  - iptfs: remove unneeded skb free count variable
  - iptfs: remove unused variable and "breadcrumb" for future code.
  - iptfs: use do_div() to avoid "__udivd13 missing" link failure.
  - iptfs: remove some BUG_ON() assertions questioned in review.