Merge tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-nextgrafted

Pull networking updates from Jakub Kicinski:
 "Core:

   - Add dedicated kmem_cache for typical/small skb->head, avoid having
     to access struct page at kfree time, and improve memory use.

   - Introduce sysctl to set default RPS configuration for new netdevs.

   - Define Netlink protocol specification format which can be used to
     describe messages used by each family and auto-generate parsers.
     Add tools for generating kernel data structures and uAPI headers.

   - Expose all net/core sysctls inside netns.

   - Remove 4s sleep in netpoll if carrier is instantly detected on
     boot.

   - Add configurable limit of MDB entries per port, and port-vlan.

   - Continue populating drop reasons throughout the stack.

   - Retire a handful of legacy Qdiscs and classifiers.

  Protocols:

   - Support IPv4 big TCP (TSO frames larger than 64kB).

   - Add IP_LOCAL_PORT_RANGE socket option, to control local port range
     on socket by socket basis.

   - Track and report in procfs number of MPTCP sockets used.

   - Support mixing IPv4 and IPv6 flows in the in-kernel MPTCP path
     manager.

   - IPv6: don't check net.ipv6.route.max_size and rely on garbage
     collection to free memory (similarly to IPv4).

   - Support Penultimate Segment Pop (PSP) flavor in SRv6 (RFC8986).

   - ICMP: add per-rate limit counters.

   - Add support for user scanning requests in ieee802154.

   - Remove static WEP support.

   - Support minimal Wi-Fi 7 Extremely High Throughput (EHT) rate
     reporting.

   - WiFi 7 EHT channel puncturing support (client & AP).

  BPF:

   - Add a rbtree data structure following the "next-gen data structure"
     precedent set by recently added linked list, that is, by using
     kfunc + kptr instead of adding a new BPF map type.

   - Expose XDP hints via kfuncs with initial support for RX hash and
     timestamp metadata.

   - Add BPF_F_NO_TUNNEL_KEY extension to bpf_skb_set_tunnel_key to
     better support decap on GRE tunnel devices not operating in collect
     metadata.

   - Improve x86 JIT's codegen for PROBE_MEM runtime error checks.

   - Remove the need for trace_printk_lock for bpf_trace_printk and
     bpf_trace_vprintk helpers.

   - Extend libbpf's bpf_tracing.h support for tracing arguments of
     kprobes/uprobes and syscall as a special case.

   - Significantly reduce the search time for module symbols by
     livepatch and BPF.

   - Enable cpumasks to be used as kptrs, which is useful for tracing
     programs tracking which tasks end up running on which CPUs in
     different time intervals.

   - Add support for BPF trampoline on s390x and riscv64.

   - Add capability to export the XDP features supported by the NIC.

   - Add __bpf_kfunc tag for marking kernel functions as kfuncs.

   - Add cgroup.memory=nobpf kernel parameter option to disable BPF
     memory accounting for container environments.

  Netfilter:

   - Remove the CLUSTERIP target. It has been marked as obsolete for
     years, and we still have WARN splats wrt races of the out-of-band
     /proc interface installed by this target.

   - Add 'destroy' commands to nf_tables. They are identical to the
     existing 'delete' commands, but do not return an error if the
     referenced object (set, chain, rule...) did not exist.

  Driver API:

   - Improve cpumask_local_spread() locality to help NICs set the right
     IRQ affinity on AMD platforms.

   - Separate C22 and C45 MDIO bus transactions more clearly.

   - Introduce new DCB table to control DSCP rewrite on egress.

   - Support configuration of Physical Layer Collision Avoidance (PLCA)
     Reconciliation Sublayer (RS) (802.3cg-2019). Modern version of
     shared medium Ethernet.

   - Support for MAC Merge layer (IEEE 802.3-2018 clause 99). Allowing
     preemption of low priority frames by high priority frames.

   - Add support for controlling MACSec offload using netlink SET.

   - Rework devlink instance refcounts to allow registration and
     de-registration under the instance lock. Split the code into
     multiple files, drop some of the unnecessarily granular locks and
     factor out common parts of netlink operation handling.

   - Add TX frame aggregation parameters (for USB drivers).

   - Add a new attr TCA_EXT_WARN_MSG to report TC (offload) warning
     messages with notifications for debug.

   - Allow offloading of UDP NEW connections via act_ct.

   - Add support for per action HW stats in TC.

   - Support hardware miss to TC action (continue processing in SW from
     a specific point in the action chain).

   - Warn if old Wireless Extension user space interface is used with
     modern cfg80211/mac80211 drivers. Do not support Wireless
     Extensions for Wi-Fi 7 devices at all. Everyone should switch to
     using nl80211 interface instead.

   - Improve the CAN bit timing configuration. Use extack to return
     error messages directly to user space, update the SJW handling,
     including the definition of a new default value that will benefit
     CAN-FD controllers, by increasing their oscillator tolerance.

  New hardware / drivers:

   - Ethernet:
      - nVidia BlueField-3 support (control traffic driver)
      - Ethernet support for imx93 SoCs
      - Motorcomm yt8531 gigabit Ethernet PHY
      - onsemi NCN26000 10BASE-T1S PHY (with support for PLCA)
      - Microchip LAN8841 PHY (incl. cable diagnostics and PTP)
      - Amlogic gxl MDIO mux

   - WiFi:
      - RealTek RTL8188EU (rtl8xxxu)
      - Qualcomm Wi-Fi 7 devices (ath12k)

   - CAN:
      - Renesas R-Car V4H

  Drivers:

   - Bluetooth:
      - Set Per Platform Antenna Gain (PPAG) for Intel controllers.

   - Ethernet NICs:
      - Intel (1G, igc):
         - support TSN / Qbv / packet scheduling features of i226 model
      - Intel (100G, ice):
         - use GNSS subsystem instead of TTY
         - multi-buffer XDP support
         - extend support for GPIO pins to E823 devices
      - nVidia/Mellanox:
         - update the shared buffer configuration on PFC commands
         - implement PTP adjphase function for HW offset control
         - TC support for Geneve and GRE with VF tunnel offload
         - more efficient crypto key management method
         - multi-port eswitch support
      - Netronome/Corigine:
         - add DCB IEEE support
         - support IPsec offloading for NFP3800
      - Freescale/NXP (enetc):
         - support XDP_REDIRECT for XDP non-linear buffers
         - improve reconfig, avoid link flap and waiting for idle
         - support MAC Merge layer
      - Other NICs:
         - sfc/ef100: add basic devlink support for ef100
         - ionic: rx_push mode operation (writing descriptors via MMIO)
         - bnxt: use the auxiliary bus abstraction for RDMA
         - r8169: disable ASPM and reset bus in case of tx timeout
         - cpsw: support QSGMII mode for J721e CPSW9G
         - cpts: support pulse-per-second output
         - ngbe: add an mdio bus driver
         - usbnet: optimize usbnet_bh() by avoiding unnecessary queuing
         - r8152: handle devices with FW with NCM support
         - amd-xgbe: support 10Mbps, 2.5GbE speeds and rx-adaptation
         - virtio-net: support multi buffer XDP
         - virtio/vsock: replace virtio_vsock_pkt with sk_buff
         - tsnep: XDP support

   - Ethernet high-speed switches:
      - nVidia/Mellanox (mlxsw):
         - add support for latency TLV (in FW control messages)
      - Microchip (sparx5):
         - separate explicit and implicit traffic forwarding rules, make
           the implicit rules always active
         - add support for egress DSCP rewrite
         - IS0 VCAP support (Ingress Classification)
         - IS2 VCAP filters (protos, L3 addrs, L4 ports, flags, ToS
           etc.)
         - ES2 VCAP support (Egress Access Control)
         - support for Per-Stream Filtering and Policing (802.1Q,
           8.6.5.1)

   - Ethernet embedded switches:
      - Marvell (mv88e6xxx):
         - add MAB (port auth) offload support
         - enable PTP receive for mv88e6390
      - NXP (ocelot):
         - support MAC Merge layer
         - support for the the vsc7512 internal copper phys
      - Microchip:
         - lan9303: convert to PHYLINK
         - lan966x: support TC flower filter statistics
         - lan937x: PTP support for KSZ9563/KSZ8563 and LAN937x
         - lan937x: support Credit Based Shaper configuration
         - ksz9477: support Energy Efficient Ethernet
      - other:
         - qca8k: convert to regmap read/write API, use bulk operations
         - rswitch: Improve TX timestamp accuracy

   - Intel WiFi (iwlwifi):
      - EHT (Wi-Fi 7) rate reporting
      - STEP equalizer support: transfer some STEP (connection to radio
        on platforms with integrated wifi) related parameters from the
        BIOS to the firmware.

   - Qualcomm 802.11ax WiFi (ath11k):
      - IPQ5018 support
      - Fine Timing Measurement (FTM) responder role support
      - channel 177 support

   - MediaTek WiFi (mt76):
      - per-PHY LED support
      - mt7996: EHT (Wi-Fi 7) support
      - Wireless Ethernet Dispatch (WED) reset support
      - switch to using page pool allocator

   - RealTek WiFi (rtw89):
      - support new version of Bluetooth co-existance

   - Mobile:
      - rmnet: support TX aggregation"

* tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1872 commits)
  page_pool: add a comment explaining the fragment counter usage
  net: ethtool: fix __ethtool_dev_mm_supported() implementation
  ethtool: pse-pd: Fix double word in comments
  xsk: add linux/vmalloc.h to xsk.c
  sefltests: netdevsim: wait for devlink instance after netns removal
  selftest: fib_tests: Always cleanup before exit
  net/mlx5e: Align IPsec ASO result memory to be as required by hardware
  net/mlx5e: TC, Set CT miss to the specific ct action instance
  net/mlx5e: Rename CHAIN_TO_REG to MAPPED_OBJ_TO_REG
  net/mlx5: Refactor tc miss handling to a single function
  net/mlx5: Kconfig: Make tc offload depend on tc skb extension
  net/sched: flower: Support hardware miss to tc action
  net/sched: flower: Move filter handle initialization earlier
  net/sched: cls_api: Support hardware miss to tc action
  net/sched: Rename user cookie and act cookie
  sfc: fix builds without CONFIG_RTC_LIB
  sfc: clean up some inconsistent indentings
  net/mlx4_en: Introduce flexible array to silence overflow warning
  net: lan966x: Fix possible deadlock inside PTP
  net/ulp: Remove redundant ->clone() test in inet_clone_ulp().
  ...

1 files changed, 329 insertions, 0 deletions

diff --git a/Documentation/userspace-api/media/cec/cec-pin-error-inj.rst b/Documentation/userspace-api/media/cec/cec-pin-error-inj.rst
new file mode 100644
index 000000000..411d42a74
--- /dev/null
+++ b/Documentation/userspace-api/media/cec/cec-pin-error-inj.rst
@@ -0,0 +1,329 @@
+.. SPDX-License-Identifier: GFDL-1.1-no-invariants-or-later
+
+.. _cec_pin_error_inj:
+
+CEC Pin Framework Error Injection
+=================================
+
+The CEC Pin Framework is a core CEC framework for CEC hardware that only
+has low-level support for the CEC bus. Most hardware today will have
+high-level CEC support where the hardware deals with driving the CEC bus,
+but some older devices aren't that fancy. However, this framework also
+allows you to connect the CEC pin to a GPIO on e.g. a Raspberry Pi and
+you have now made a CEC adapter.
+
+What makes doing this so interesting is that since we have full control
+over the bus it is easy to support error injection. This is ideal to
+test how well CEC adapters can handle error conditions.
+
+Currently only the cec-gpio driver (when the CEC line is directly
+connected to a pull-up GPIO line) and the AllWinner A10/A20 drm driver
+support this framework.
+
+If ``CONFIG_CEC_PIN_ERROR_INJ`` is enabled, then error injection is available
+through debugfs. Specifically, in ``/sys/kernel/debug/cec/cecX/`` there is
+now an ``error-inj`` file.
+
+.. note::
+
+    The error injection commands are not a stable ABI and may change in the
+    future.
+
+With ``cat error-inj`` you can see both the possible commands and the current
+error injection status::
+
+	$ cat /sys/kernel/debug/cec/cec0/error-inj
+	# Clear error injections:
+	#   clear          clear all rx and tx error injections
+	#   rx-clear       clear all rx error injections
+	#   tx-clear       clear all tx error injections
+	#   <op> clear     clear all rx and tx error injections for <op>
+	#   <op> rx-clear  clear all rx error injections for <op>
+	#   <op> tx-clear  clear all tx error injections for <op>
+	#
+	# RX error injection:
+	#   <op>[,<mode>] rx-nack              NACK the message instead of sending an ACK
+	#   <op>[,<mode>] rx-low-drive <bit>   force a low-drive condition at this bit position
+	#   <op>[,<mode>] rx-add-byte          add a spurious byte to the received CEC message
+	#   <op>[,<mode>] rx-remove-byte       remove the last byte from the received CEC message
+	#    any[,<mode>] rx-arb-lost [<poll>] generate a POLL message to trigger an arbitration lost
+	#
+	# TX error injection settings:
+	#   tx-ignore-nack-until-eom           ignore early NACKs until EOM
+	#   tx-custom-low-usecs <usecs>        define the 'low' time for the custom pulse
+	#   tx-custom-high-usecs <usecs>       define the 'high' time for the custom pulse
+	#   tx-custom-pulse                    transmit the custom pulse once the bus is idle
+	#
+	# TX error injection:
+	#   <op>[,<mode>] tx-no-eom            don't set the EOM bit
+	#   <op>[,<mode>] tx-early-eom         set the EOM bit one byte too soon
+	#   <op>[,<mode>] tx-add-bytes <num>   append <num> (1-255) spurious bytes to the message
+	#   <op>[,<mode>] tx-remove-byte       drop the last byte from the message
+	#   <op>[,<mode>] tx-short-bit <bit>   make this bit shorter than allowed
+	#   <op>[,<mode>] tx-long-bit <bit>    make this bit longer than allowed
+	#   <op>[,<mode>] tx-custom-bit <bit>  send the custom pulse instead of this bit
+	#   <op>[,<mode>] tx-short-start       send a start pulse that's too short
+	#   <op>[,<mode>] tx-long-start        send a start pulse that's too long
+	#   <op>[,<mode>] tx-custom-start      send the custom pulse instead of the start pulse
+	#   <op>[,<mode>] tx-last-bit <bit>    stop sending after this bit
+	#   <op>[,<mode>] tx-low-drive <bit>   force a low-drive condition at this bit position
+	#
+	# <op>       CEC message opcode (0-255) or 'any'
+	# <mode>     'once' (default), 'always', 'toggle' or 'off'
+	# <bit>      CEC message bit (0-159)
+	#            10 bits per 'byte': bits 0-7: data, bit 8: EOM, bit 9: ACK
+	# <poll>     CEC poll message used to test arbitration lost (0x00-0xff, default 0x0f)
+	# <usecs>    microseconds (0-10000000, default 1000)
+
+	clear
+
+You can write error injection commands to ``error-inj`` using
+``echo 'cmd' >error-inj`` or ``cat cmd.txt >error-inj``. The ``cat error-inj``
+output contains the current error commands. You can save the output to a file
+and use it as an input to ``error-inj`` later.
+
+Basic Syntax
+------------
+
+Leading spaces/tabs are ignored. If the next character is a ``#`` or the end
+of the line was reached, then the whole line is ignored. Otherwise a command
+is expected.
+
+The error injection commands fall in two main groups: those relating to
+receiving CEC messages and those relating to transmitting CEC messages. In
+addition, there are commands to clear existing error injection commands and
+to create custom pulses on the CEC bus.
+
+Most error injection commands can be executed for specific CEC opcodes or for
+all opcodes (``any``). Each command also has a 'mode' which can be ``off``
+(can be used to turn off an existing error injection command), ``once``
+(the default) which will trigger the error injection only once for the next
+received or transmitted message, ``always`` to always trigger the error
+injection and ``toggle`` to toggle the error injection on or off for every
+transmit or receive.
+
+So '``any rx-nack``' will NACK the next received CEC message,
+'``any,always rx-nack``' will NACK all received CEC messages and
+'``0x82,toggle rx-nack``' will only NACK if an Active Source message was
+received and do that only for every other received message.
+
+After an error was injected with mode ``once`` the error injection command
+is cleared automatically, so ``once`` is a one-time deal.
+
+All combinations of ``<op>`` and error injection commands can co-exist. So
+this is fine::
+
+	0x9e tx-add-bytes 1
+	0x9e tx-early-eom
+	0x9f tx-add-bytes 2
+	any rx-nack
+
+All four error injection commands will be active simultaneously.
+
+However, if the same ``<op>`` and command combination is specified,
+but with different arguments::
+
+	0x9e tx-add-bytes 1
+	0x9e tx-add-bytes 2
+
+Then the second will overwrite the first.
+
+Clear Error Injections
+----------------------
+
+``clear``
+    Clear all error injections.
+
+``rx-clear``
+    Clear all receive error injections
+
+``tx-clear``
+    Clear all transmit error injections
+
+``<op> clear``
+    Clear all error injections for the given opcode.
+
+``<op> rx-clear``
+    Clear all receive error injections for the given opcode.
+
+``<op> tx-clear``
+    Clear all transmit error injections for the given opcode.
+
+Receive Messages
+----------------
+
+``<op>[,<mode>] rx-nack``
+    NACK broadcast messages and messages directed to this CEC adapter.
+    Every byte of the message will be NACKed in case the transmitter
+    keeps transmitting after the first byte was NACKed.
+
+``<op>[,<mode>] rx-low-drive <bit>``
+    Force a Low Drive condition at this bit position. If <op> specifies
+    a specific CEC opcode then the bit position must be at least 18,
+    otherwise the opcode hasn't been received yet. This tests if the
+    transmitter can handle the Low Drive condition correctly and reports
+    the error correctly. Note that a Low Drive in the first 4 bits can also
+    be interpreted as an Arbitration Lost condition by the transmitter.
+    This is implementation dependent.
+
+``<op>[,<mode>] rx-add-byte``
+    Add a spurious 0x55 byte to the received CEC message, provided
+    the message was 15 bytes long or less. This is useful to test
+    the high-level protocol since spurious bytes should be ignored.
+
+``<op>[,<mode>] rx-remove-byte``
+    Remove the last byte from the received CEC message, provided it
+    was at least 2 bytes long. This is useful to test the high-level
+    protocol since messages that are too short should be ignored.
+
+``<op>[,<mode>] rx-arb-lost <poll>``
+    Generate a POLL message to trigger an Arbitration Lost condition.
+    This command is only allowed for ``<op>`` values of ``next`` or ``all``.
+    As soon as a start bit has been received the CEC adapter will switch
+    to transmit mode and it will transmit a POLL message. By default this is
+    0x0f, but it can also be specified explicitly via the ``<poll>`` argument.
+
+    This command can be used to test the Arbitration Lost condition in
+    the remote CEC transmitter. Arbitration happens when two CEC adapters
+    start sending a message at the same time. In that case the initiator
+    with the most leading zeroes wins and the other transmitter has to
+    stop transmitting ('Arbitration Lost'). This is very hard to test,
+    except by using this error injection command.
+
+    This does not work if the remote CEC transmitter has logical address
+    0 ('TV') since that will always win.
+
+Transmit Messages
+-----------------
+
+``tx-ignore-nack-until-eom``
+    This setting changes the behavior of transmitting CEC messages. Normally
+    as soon as the receiver NACKs a byte the transmit will stop, but the
+    specification also allows that the full message is transmitted and only
+    at the end will the transmitter look at the ACK bit. This is not
+    recommended behavior since there is no point in keeping the CEC bus busy
+    for longer than is strictly needed. Especially given how slow the bus is.
+
+    This setting can be used to test how well a receiver deals with
+    transmitters that ignore NACKs until the very end of the message.
+
+``<op>[,<mode>] tx-no-eom``
+    Don't set the EOM bit. Normally the last byte of the message has the EOM
+    (End-Of-Message) bit set. With this command the transmit will just stop
+    without ever sending an EOM. This can be used to test how a receiver
+    handles this case. Normally receivers have a time-out after which
+    they will go back to the Idle state.
+
+``<op>[,<mode>] tx-early-eom``
+    Set the EOM bit one byte too soon. This obviously only works for messages
+    of two bytes or more. The EOM bit will be set for the second-to-last byte
+    and not for the final byte. The receiver should ignore the last byte in
+    this case. Since the resulting message is likely to be too short for this
+    same reason the whole message is typically ignored. The receiver should be
+    in Idle state after the last byte was transmitted.
+
+``<op>[,<mode>] tx-add-bytes <num>``
+    Append ``<num>`` (1-255) spurious bytes to the message. The extra bytes
+    have the value of the byte position in the message. So if you transmit a
+    two byte message (e.g. a Get CEC Version message) and add 2 bytes, then
+    the full message received by the remote CEC adapter is
+    ``0x40 0x9f 0x02 0x03``.
+
+    This command can be used to test buffer overflows in the receiver. E.g.
+    what does it do when it receives more than the maximum message size of 16
+    bytes.
+
+``<op>[,<mode>] tx-remove-byte``
+    Drop the last byte from the message, provided the message is at least
+    two bytes long. The receiver should ignore messages that are too short.
+
+``<op>[,<mode>] tx-short-bit <bit>``
+    Make this bit period shorter than allowed. The bit position cannot be
+    an Ack bit.  If <op> specifies a specific CEC opcode then the bit position
+    must be at least 18, otherwise the opcode hasn't been received yet.
+    Normally the period of a data bit is between 2.05 and 2.75 milliseconds.
+    With this command the period of this bit is 1.8 milliseconds, this is
+    done by reducing the time the CEC bus is high. This bit period is less
+    than is allowed and the receiver should respond with a Low Drive
+    condition.
+
+    This command is ignored for 0 bits in bit positions 0 to 3. This is
+    because the receiver also looks for an Arbitration Lost condition in
+    those first four bits and it is undefined what will happen if it
+    sees a too-short 0 bit.
+
+``<op>[,<mode>] tx-long-bit <bit>``
+    Make this bit period longer than is valid. The bit position cannot be
+    an Ack bit.  If <op> specifies a specific CEC opcode then the bit position
+    must be at least 18, otherwise the opcode hasn't been received yet.
+    Normally the period of a data bit is between 2.05 and 2.75 milliseconds.
+    With this command the period of this bit is 2.9 milliseconds, this is
+    done by increasing the time the CEC bus is high.
+
+    Even though this bit period is longer than is valid it is undefined what
+    a receiver will do. It might just accept it, or it might time out and
+    return to Idle state. Unfortunately the CEC specification is silent about
+    this.
+
+    This command is ignored for 0 bits in bit positions 0 to 3. This is
+    because the receiver also looks for an Arbitration Lost condition in
+    those first four bits and it is undefined what will happen if it
+    sees a too-long 0 bit.
+
+``<op>[,<mode>] tx-short-start``
+    Make this start bit period shorter than allowed. Normally the period of
+    a start bit is between 4.3 and 4.7 milliseconds. With this command the
+    period of the start bit is 4.1 milliseconds, this is done by reducing
+    the time the CEC bus is high. This start bit period is less than is
+    allowed and the receiver should return to Idle state when this is detected.
+
+``<op>[,<mode>] tx-long-start``
+    Make this start bit period longer than is valid. Normally the period of
+    a start bit is between 4.3 and 4.7 milliseconds. With this command the
+    period of the start bit is 5 milliseconds, this is done by increasing
+    the time the CEC bus is high. This start bit period is more than is
+    valid and the receiver should return to Idle state when this is detected.
+
+    Even though this start bit period is longer than is valid it is undefined
+    what a receiver will do. It might just accept it, or it might time out and
+    return to Idle state. Unfortunately the CEC specification is silent about
+    this.
+
+``<op>[,<mode>] tx-last-bit <bit>``
+    Just stop transmitting after this bit.  If <op> specifies a specific CEC
+    opcode then the bit position must be at least 18, otherwise the opcode
+    hasn't been received yet. This command can be used to test how the receiver
+    reacts when a message just suddenly stops. It should time out and go back
+    to Idle state.
+
+``<op>[,<mode>] tx-low-drive <bit>``
+    Force a Low Drive condition at this bit position. If <op> specifies a
+    specific CEC opcode then the bit position must be at least 18, otherwise
+    the opcode hasn't been received yet. This can be used to test how the
+    receiver handles Low Drive conditions. Note that if this happens at bit
+    positions 0-3 the receiver can interpret this as an Arbitration Lost
+    condition. This is implementation dependent.
+
+Custom Pulses
+-------------
+
+``tx-custom-low-usecs <usecs>``
+    This defines the duration in microseconds that the custom pulse pulls
+    the CEC line low. The default is 1000 microseconds.
+
+``tx-custom-high-usecs <usecs>``
+    This defines the duration in microseconds that the custom pulse keeps the
+    CEC line high (unless another CEC adapter pulls it low in that time).
+    The default is 1000 microseconds. The total period of the custom pulse is
+    ``tx-custom-low-usecs + tx-custom-high-usecs``.
+
+``<op>[,<mode>] tx-custom-bit <bit>``
+    Send the custom bit instead of a regular data bit. The bit position cannot
+    be an Ack bit.  If <op> specifies a specific CEC opcode then the bit
+    position must be at least 18, otherwise the opcode hasn't been received yet.
+
+``<op>[,<mode>] tx-custom-start``
+    Send the custom bit instead of a regular start bit.
+
+``tx-custom-pulse``
+    Transmit a single custom pulse as soon as the CEC bus is idle.