Documentation/networking/timestamping.rst

1 .. SPDX-License-Identifier: GPL-2.0
43 -------------------------------------------------------------
59 -------------------------------------------------------------------
72 ----------------------------------------------------------------------
138   over-report measurement, because the timestamp is generated when all
189   is derived from a per-socket u32 counter (that wraps). For datagram
206   passing a specific ID with control message SCM_TS_OPT_ID (not
212     cmsg->cmsg_level		 = SOL_SOCKET;
213     cmsg->cmsg_type		 = SCM_TS_OPT_ID;
214     cmsg->cmsg_len		 = CMSG_LEN(sizeof(__u32));
230   a timestamp with counter N-1. SOF_TIMESTAMPING_OPT_ID_TCP
280   received the packet and its length at layer 2. A valid (non-zero)
305   ignore the unexpected non-zero value. But it makes behavior subtly
332   cmsg->cmsg_level	       = SOL_SOCKET;
333   cmsg->cmsg_type	       = SO_TIMESTAMPING;
334   cmsg->cmsg_len	       = CMSG_LEN(sizeof(__u32));
352 -------------------------
362 correlating a timestamp with data is non-trivial. A range of bytes
391 relevant sequence number in skb_shinfo(skb)->tskey. Because an skbuff
400 autocork. After linux-4.7, a better way to prevent coalescing is
422 ----------------------------
445 feature. At least one field is non-zero at any time. Most timestamps
452 as linuxptp. For the PTP clock API, see Documentation/driver-api/ptp.rst.
489 is the first if ts[2] is non-zero, the second otherwise, in which
517 Reading from the error queue is always a non-blocking operation. To
549 Desired behavior is passed into the kernel and to a specific device by
582 If the requested fine-grained filtering for incoming packets is not
589     /* possible values for hwtstamp_config->tx_type */
607     /* possible values for hwtstamp_config->rx_filter */
627 --------------------------------------------------------
648 - In hard_start_xmit(), check if (skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)
649   is set no-zero. If yes, then the driver is expected to do hardware time
651 - If this is possible for the skb and requested, then declare
653   SKBTX_IN_PROGRESS in skb_shinfo(skb)->tx_flags , e.g. with::
655       skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
661 - Driver should call skb_tx_timestamp() as close to passing sk_buff to hardware
664 - As soon as the driver has sent the packet and/or obtained a
676 ----------------------------------------------------------
685 3.2.1 DSA (Distributed Switch Architecture) switches
690 a port multiplier with optional forwarding acceleration features.  Each DSA
695 When a DSA switch is attached to a host port, PTP synchronization has to
697 jitter between the host port and its PTP partner. For this reason, some DSA
700 measure wire and PHY propagation latencies. Timestamping DSA switches are
702 for the fact that the DSA interfaces are in fact virtual in terms of network
704 interfaces of a DSA switch to share the same PHC.
706 By design, PTP timestamping with a DSA switch does not need any special
708 host port also supports PTP timestamping, DSA will take care of intercepting
712 anybody else except for the DSA switch port must be prevented from doing so.
714 In the generic layer, DSA provides the following infrastructure for PTP
717 - ``.port_txtstamp()``: a hook called prior to the transmission of
719   This is required for two-step timestamping, since the hardware
722   packet so that it can re-enqueue the packet back into the socket's
725   in skb->cb and enqueue a tx skb queue. Typically, a switch will have a
728   key-value pairs of PTP sequence ID/message type/domain number and the
736   One-step TX timestamping do not require packet cloning, since there is
737   no follow-up message required by the PTP protocol (because the
740   to be re-enqueued into its socket's error queue.
742 - ``.port_rxtstamp()``: On RX, the BPF classifier is run by DSA to
747   timestamps might either be available in-band (through metadata in the
748   DSA header, or attached in other ways to the packet), or out-of-band
751   that case, it is the responsibility of the DSA driver to call
758 hence they do not have a representation in terms of a network interface as DSA
764 mii_timestamper`` and add a pointer to it in ``phydev->mii_ts``. The presence
769 driver.  Therefore, as opposed to DSA switches, modifications need to be done
772 - Checking, in ``.ndo_eth_ioctl``, whether ``phy_has_hwtstamp(netdev->phydev)``
776 - On RX, special intervention may or may not be needed, depending on the
779   ``skb_defer_rx_timestamp(skb)`` is necessary or not - and if it is, don't
781   enabled, and ``skb->dev->phydev->mii_ts`` exists, its ``.rxtstamp()`` hook
782   will be called now, to determine, using logic very similar to DSA, whether
783   deferral for RX timestamping is necessary.  Again like DSA, it becomes the
792 - On TX, again, special intervention might or might not be needed.  The
793   function that calls the ``mii_ts->txtstamp()`` hook is named
818       skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
820 Any TX timestamping logic, be it a plain MAC driver, a DSA switch driver, a PHY
829    the ``.ndo_eth_ioctl`` ("``priv->hwtstamp_tx_enabled == true``") and the
830    current skb requires a TX timestamp ("``skb_shinfo(skb)->tx_flags &
832    "``skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS``" flag. Note: as
840    "``skb_shinfo(skb)->tx_flags & SKBTX_IN_PROGRESS``" was set. With a stacked
847 "``skb_shinfo(skb)->tx_flags & SKBTX_IN_PROGRESS``", but also for
848 "``priv->hwtstamp_tx_enabled == true``". Because the rest of the system ensures