1 .. SPDX-License-Identifier: GPL-2.0
11 The interfaces for receiving network packages timestamps are:
14 Generates a timestamp for each incoming packet in (not necessarily
15 monotonic) system time. Reports the timestamp via recvmsg() in a
16 control message in usec resolution.
17 SO_TIMESTAMP is defined as SO_TIMESTAMP_NEW or SO_TIMESTAMP_OLD
18 based on the architecture type and time_t representation of libc.
19 Control message format is in struct __kernel_old_timeval for
20 SO_TIMESTAMP_OLD and in struct __kernel_sock_timeval for
21 SO_TIMESTAMP_NEW options respectively.
24 Same timestamping mechanism as SO_TIMESTAMP, but reports the
25 timestamp as struct timespec in nsec resolution.
26 SO_TIMESTAMPNS is defined as SO_TIMESTAMPNS_NEW or SO_TIMESTAMPNS_OLD
27 based on the architecture type and time_t representation of libc.
28 Control message format is in struct timespec for SO_TIMESTAMPNS_OLD
29 and in struct __kernel_timespec for SO_TIMESTAMPNS_NEW options
32 IP_MULTICAST_LOOP + SO_TIMESTAMP[NS]
33 Only for multicast:approximate transmit timestamp obtained by
34 reading the looped packet receive timestamp.
37 Generates timestamps on reception, transmission or both. Supports
38 multiple timestamp sources, including hardware. Supports generating
39 timestamps for stream sockets.
42 1.1 SO_TIMESTAMP (also SO_TIMESTAMP_OLD and SO_TIMESTAMP_NEW)
43 -------------------------------------------------------------
45 This socket option enables timestamping of datagrams on the reception
46 path. Because the destination socket, if any, is not known early in
47 the network stack, the feature has to be enabled for all packets. The
48 same is true for all early receive timestamp options.
50 For interface details, see `man 7 socket`.
52 Always use SO_TIMESTAMP_NEW timestamp to always get timestamp in
53 struct __kernel_sock_timeval format.
55 SO_TIMESTAMP_OLD returns incorrect timestamps after the year 2038
58 1.2 SO_TIMESTAMPNS (also SO_TIMESTAMPNS_OLD and SO_TIMESTAMPNS_NEW)
59 -------------------------------------------------------------------
61 This option is identical to SO_TIMESTAMP except for the returned data type.
62 Its struct timespec allows for higher resolution (ns) timestamps than the
63 timeval of SO_TIMESTAMP (ms).
65 Always use SO_TIMESTAMPNS_NEW timestamp to always get timestamp in
66 struct __kernel_timespec format.
68 SO_TIMESTAMPNS_OLD returns incorrect timestamps after the year 2038
71 1.3 SO_TIMESTAMPING (also SO_TIMESTAMPING_OLD and SO_TIMESTAMPING_NEW)
72 ----------------------------------------------------------------------
74 Supports multiple types of timestamp requests. As a result, this
75 socket option takes a bitmap of flags, not a boolean. In::
77 err = setsockopt(fd, SOL_SOCKET, SO_TIMESTAMPING, &val, sizeof(val));
79 val is an integer with any of the following bits set. Setting other
80 bit returns EINVAL and does not change the current state.
82 The socket option configures timestamp generation for individual
83 sk_buffs (1.3.1), timestamp reporting to the socket's error
84 queue (1.3.2) and options (1.3.3). Timestamp generation can also
85 be enabled for individual sendmsg calls using cmsg (1.3.4).
88 1.3.1 Timestamp Generation
89 ^^^^^^^^^^^^^^^^^^^^^^^^^^
91 Some bits are requests to the stack to try to generate timestamps. Any
92 combination of them is valid. Changes to these bits apply to newly
93 created packets, not to packets already in the stack. As a result, it
94 is possible to selectively request timestamps for a subset of packets
95 (e.g., for sampling) by embedding an send() call within two setsockopt
96 calls, one to enable timestamp generation and one to disable it.
97 Timestamps may also be generated for reasons other than being
98 requested by a particular socket, such as when receive timestamping is
99 enabled system wide, as explained earlier.
101 SOF_TIMESTAMPING_RX_HARDWARE:
102 Request rx timestamps generated by the network adapter.
104 SOF_TIMESTAMPING_RX_SOFTWARE:
105 Request rx timestamps when data enters the kernel. These timestamps
106 are generated just after a device driver hands a packet to the
107 kernel receive stack.
109 SOF_TIMESTAMPING_TX_HARDWARE:
110 Request tx timestamps generated by the network adapter. This flag
111 can be enabled via both socket options and control messages.
113 SOF_TIMESTAMPING_TX_SOFTWARE:
114 Request tx timestamps when data leaves the kernel. These timestamps
115 are generated in the device driver as close as possible, but always
116 prior to, passing the packet to the network interface. Hence, they
117 require driver support and may not be available for all devices.
118 This flag can be enabled via both socket options and control messages.
120 SOF_TIMESTAMPING_TX_SCHED:
121 Request tx timestamps prior to entering the packet scheduler. Kernel
122 transmit latency is, if long, often dominated by queuing delay. The
123 difference between this timestamp and one taken at
124 SOF_TIMESTAMPING_TX_SOFTWARE will expose this latency independent
125 of protocol processing. The latency incurred in protocol
126 processing, if any, can be computed by subtracting a userspace
127 timestamp taken immediately before send() from this timestamp. On
128 machines with virtual devices where a transmitted packet travels
129 through multiple devices and, hence, multiple packet schedulers,
130 a timestamp is generated at each layer. This allows for fine
131 grained measurement of queuing delay. This flag can be enabled
132 via both socket options and control messages.
134 SOF_TIMESTAMPING_TX_ACK:
135 Request tx timestamps when all data in the send buffer has been
136 acknowledged. This only makes sense for reliable protocols. It is
137 currently only implemented for TCP. For that protocol, it may
138 over-report measurement, because the timestamp is generated when all
139 data up to and including the buffer at send() was acknowledged: the
140 cumulative acknowledgment. The mechanism ignores SACK and FACK.
141 This flag can be enabled via both socket options and control messages.
144 1.3.2 Timestamp Reporting
145 ^^^^^^^^^^^^^^^^^^^^^^^^^
147 The other three bits control which timestamps will be reported in a
148 generated control message. Changes to the bits take immediate
149 effect at the timestamp reporting locations in the stack. Timestamps
150 are only reported for packets that also have the relevant timestamp
151 generation request set.
153 SOF_TIMESTAMPING_SOFTWARE:
154 Report any software timestamps when available.
156 SOF_TIMESTAMPING_SYS_HARDWARE:
157 This option is deprecated and ignored.
159 SOF_TIMESTAMPING_RAW_HARDWARE:
160 Report hardware timestamps as generated by
161 SOF_TIMESTAMPING_TX_HARDWARE when available.
164 1.3.3 Timestamp Options
165 ^^^^^^^^^^^^^^^^^^^^^^^
167 The interface supports the options
169 SOF_TIMESTAMPING_OPT_ID:
170 Generate a unique identifier along with each packet. A process can
171 have multiple concurrent timestamping requests outstanding. Packets
172 can be reordered in the transmit path, for instance in the packet
173 scheduler. In that case timestamps will be queued onto the error
174 queue out of order from the original send() calls. It is not always
175 possible to uniquely match timestamps to the original send() calls
176 based on timestamp order or payload inspection alone, then.
178 This option associates each packet at send() with a unique
179 identifier and returns that along with the timestamp. The identifier
180 is derived from a per-socket u32 counter (that wraps). For datagram
181 sockets, the counter increments with each sent packet. For stream
182 sockets, it increments with every byte.
184 The counter starts at zero. It is initialized the first time that
185 the socket option is enabled. It is reset each time the option is
186 enabled after having been disabled. Resetting the counter does not
187 change the identifiers of existing packets in the system.
189 This option is implemented only for transmit timestamps. There, the
190 timestamp is always looped along with a struct sock_extended_err.
191 The option modifies field ee_data to pass an id that is unique
192 among all possibly concurrently outstanding timestamp requests for
196 SOF_TIMESTAMPING_OPT_CMSG:
197 Support recv() cmsg for all timestamped packets. Control messages
198 are already supported unconditionally on all packets with receive
199 timestamps and on IPv6 packets with transmit timestamp. This option
200 extends them to IPv4 packets with transmit timestamp. One use case
201 is to correlate packets with their egress device, by enabling socket
202 option IP_PKTINFO simultaneously.
205 SOF_TIMESTAMPING_OPT_TSONLY:
206 Applies to transmit timestamps only. Makes the kernel return the
207 timestamp as a cmsg alongside an empty packet, as opposed to
208 alongside the original packet. This reduces the amount of memory
209 charged to the socket's receive budget (SO_RCVBUF) and delivers
210 the timestamp even if sysctl net.core.tstamp_allow_data is 0.
211 This option disables SOF_TIMESTAMPING_OPT_CMSG.
213 SOF_TIMESTAMPING_OPT_STATS:
214 Optional stats that are obtained along with the transmit timestamps.
215 It must be used together with SOF_TIMESTAMPING_OPT_TSONLY. When the
216 transmit timestamp is available, the stats are available in a
217 separate control message of type SCM_TIMESTAMPING_OPT_STATS, as a
218 list of TLVs (struct nlattr) of types. These stats allow the
219 application to associate various transport layer stats with
220 the transmit timestamps, such as how long a certain block of
221 data was limited by peer's receiver window.
223 SOF_TIMESTAMPING_OPT_PKTINFO:
224 Enable the SCM_TIMESTAMPING_PKTINFO control message for incoming
225 packets with hardware timestamps. The message contains struct
226 scm_ts_pktinfo, which supplies the index of the real interface which
227 received the packet and its length at layer 2. A valid (non-zero)
228 interface index will be returned only if CONFIG_NET_RX_BUSY_POLL is
229 enabled and the driver is using NAPI. The struct contains also two
230 other fields, but they are reserved and undefined.
232 SOF_TIMESTAMPING_OPT_TX_SWHW:
233 Request both hardware and software timestamps for outgoing packets
234 when SOF_TIMESTAMPING_TX_HARDWARE and SOF_TIMESTAMPING_TX_SOFTWARE
235 are enabled at the same time. If both timestamps are generated,
236 two separate messages will be looped to the socket's error queue,
237 each containing just one timestamp.
239 New applications are encouraged to pass SOF_TIMESTAMPING_OPT_ID to
240 disambiguate timestamps and SOF_TIMESTAMPING_OPT_TSONLY to operate
241 regardless of the setting of sysctl net.core.tstamp_allow_data.
243 An exception is when a process needs additional cmsg data, for
244 instance SOL_IP/IP_PKTINFO to detect the egress network interface.
245 Then pass option SOF_TIMESTAMPING_OPT_CMSG. This option depends on
246 having access to the contents of the original packet, so cannot be
247 combined with SOF_TIMESTAMPING_OPT_TSONLY.
250 1.3.4. Enabling timestamps via control messages
251 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
253 In addition to socket options, timestamp generation can be requested
254 per write via cmsg, only for SOF_TIMESTAMPING_TX_* (see Section 1.3.1).
255 Using this feature, applications can sample timestamps per sendmsg()
256 without paying the overhead of enabling and disabling timestamps via
261 cmsg = CMSG_FIRSTHDR(msg);
262 cmsg->cmsg_level = SOL_SOCKET;
263 cmsg->cmsg_type = SO_TIMESTAMPING;
264 cmsg->cmsg_len = CMSG_LEN(sizeof(__u32));
265 *((__u32 *) CMSG_DATA(cmsg)) = SOF_TIMESTAMPING_TX_SCHED |
266 SOF_TIMESTAMPING_TX_SOFTWARE |
267 SOF_TIMESTAMPING_TX_ACK;
268 err = sendmsg(fd, msg, 0);
270 The SOF_TIMESTAMPING_TX_* flags set via cmsg will override
271 the SOF_TIMESTAMPING_TX_* flags set via setsockopt.
273 Moreover, applications must still enable timestamp reporting via
274 setsockopt to receive timestamps::
276 __u32 val = SOF_TIMESTAMPING_SOFTWARE |
277 SOF_TIMESTAMPING_OPT_ID /* or any other flag */;
278 err = setsockopt(fd, SOL_SOCKET, SO_TIMESTAMPING, &val, sizeof(val));
281 1.4 Bytestream Timestamps
282 -------------------------
284 The SO_TIMESTAMPING interface supports timestamping of bytes in a
285 bytestream. Each request is interpreted as a request for when the
286 entire contents of the buffer has passed a timestamping point. That
287 is, for streams option SOF_TIMESTAMPING_TX_SOFTWARE will record
288 when all bytes have reached the device driver, regardless of how
289 many packets the data has been converted into.
291 In general, bytestreams have no natural delimiters and therefore
292 correlating a timestamp with data is non-trivial. A range of bytes
293 may be split across segments, any segments may be merged (possibly
294 coalescing sections of previously segmented buffers associated with
295 independent send() calls). Segments can be reordered and the same
296 byte range can coexist in multiple segments for protocols that
297 implement retransmissions.
299 It is essential that all timestamps implement the same semantics,
300 regardless of these possible transformations, as otherwise they are
301 incomparable. Handling "rare" corner cases differently from the
302 simple case (a 1:1 mapping from buffer to skb) is insufficient
303 because performance debugging often needs to focus on such outliers.
305 In practice, timestamps can be correlated with segments of a
306 bytestream consistently, if both semantics of the timestamp and the
307 timing of measurement are chosen correctly. This challenge is no
308 different from deciding on a strategy for IP fragmentation. There, the
309 definition is that only the first fragment is timestamped. For
310 bytestreams, we chose that a timestamp is generated only when all
311 bytes have passed a point. SOF_TIMESTAMPING_TX_ACK as defined is easy to
312 implement and reason about. An implementation that has to take into
313 account SACK would be more complex due to possible transmission holes
314 and out of order arrival.
316 On the host, TCP can also break the simple 1:1 mapping from buffer to
317 skbuff as a result of Nagle, cork, autocork, segmentation and GSO. The
318 implementation ensures correctness in all cases by tracking the
319 individual last byte passed to send(), even if it is no longer the
320 last byte after an skbuff extend or merge operation. It stores the
321 relevant sequence number in skb_shinfo(skb)->tskey. Because an skbuff
322 has only one such field, only one timestamp can be generated.
324 In rare cases, a timestamp request can be missed if two requests are
325 collapsed onto the same skb. A process can detect this situation by
326 enabling SOF_TIMESTAMPING_OPT_ID and comparing the byte offset at
327 send time with the value returned for each timestamp. It can prevent
328 the situation by always flushing the TCP stack in between requests,
329 for instance by enabling TCP_NODELAY and disabling TCP_CORK and
332 These precautions ensure that the timestamp is generated only when all
333 bytes have passed a timestamp point, assuming that the network stack
334 itself does not reorder the segments. The stack indeed tries to avoid
335 reordering. The one exception is under administrator control: it is
336 possible to construct a packet scheduler configuration that delays
337 segments from the same stream differently. Such a setup would be
344 Timestamps are read using the ancillary data feature of recvmsg().
345 See `man 3 cmsg` for details of this interface. The socket manual
346 page (`man 7 socket`) describes how timestamps generated with
347 SO_TIMESTAMP and SO_TIMESTAMPNS records can be retrieved.
350 2.1 SCM_TIMESTAMPING records
351 ----------------------------
353 These timestamps are returned in a control message with cmsg_level
354 SOL_SOCKET, cmsg_type SCM_TIMESTAMPING, and payload of type
356 For SO_TIMESTAMPING_OLD::
358 struct scm_timestamping {
359 struct timespec ts[3];
362 For SO_TIMESTAMPING_NEW::
364 struct scm_timestamping64 {
365 struct __kernel_timespec ts[3];
367 Always use SO_TIMESTAMPING_NEW timestamp to always get timestamp in
368 struct scm_timestamping64 format.
370 SO_TIMESTAMPING_OLD returns incorrect timestamps after the year 2038
373 The structure can return up to three timestamps. This is a legacy
374 feature. At least one field is non-zero at any time. Most timestamps
375 are passed in ts[0]. Hardware timestamps are passed in ts[2].
377 ts[1] used to hold hardware timestamps converted to system time.
378 Instead, expose the hardware clock device on the NIC directly as
379 a HW PTP clock source, to allow time conversion in userspace and
380 optionally synchronize system time with a userspace PTP stack such
381 as linuxptp. For the PTP clock API, see Documentation/driver-api/ptp.rst.
383 Note that if the SO_TIMESTAMP or SO_TIMESTAMPNS option is enabled
384 together with SO_TIMESTAMPING using SOF_TIMESTAMPING_SOFTWARE, a false
385 software timestamp will be generated in the recvmsg() call and passed
386 in ts[0] when a real software timestamp is missing. This happens also
387 on hardware transmit timestamps.
389 2.1.1 Transmit timestamps with MSG_ERRQUEUE
390 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
392 For transmit timestamps the outgoing packet is looped back to the
393 socket's error queue with the send timestamp(s) attached. A process
394 receives the timestamps by calling recvmsg() with flag MSG_ERRQUEUE
395 set and with a msg_control buffer sufficiently large to receive the
396 relevant metadata structures. The recvmsg call returns the original
397 outgoing data packet with two ancillary messages attached.
399 A message of cm_level SOL_IP(V6) and cm_type IP(V6)_RECVERR
400 embeds a struct sock_extended_err. This defines the error type. For
401 timestamps, the ee_errno field is ENOMSG. The other ancillary message
402 will have cm_level SOL_SOCKET and cm_type SCM_TIMESTAMPING. This
403 embeds the struct scm_timestamping.
406 2.1.1.2 Timestamp types
407 ~~~~~~~~~~~~~~~~~~~~~~~
409 The semantics of the three struct timespec are defined by field
410 ee_info in the extended error structure. It contains a value of
411 type SCM_TSTAMP_* to define the actual timestamp passed in
414 The SCM_TSTAMP_* types are 1:1 matches to the SOF_TIMESTAMPING_*
415 control fields discussed previously, with one exception. For legacy
416 reasons, SCM_TSTAMP_SND is equal to zero and can be set for both
417 SOF_TIMESTAMPING_TX_HARDWARE and SOF_TIMESTAMPING_TX_SOFTWARE. It
418 is the first if ts[2] is non-zero, the second otherwise, in which
419 case the timestamp is stored in ts[0].
422 2.1.1.3 Fragmentation
423 ~~~~~~~~~~~~~~~~~~~~~
425 Fragmentation of outgoing datagrams is rare, but is possible, e.g., by
426 explicitly disabling PMTU discovery. If an outgoing packet is fragmented,
427 then only the first fragment is timestamped and returned to the sending
431 2.1.1.4 Packet Payload
432 ~~~~~~~~~~~~~~~~~~~~~~
434 The calling application is often not interested in receiving the whole
435 packet payload that it passed to the stack originally: the socket
436 error queue mechanism is just a method to piggyback the timestamp on.
437 In this case, the application can choose to read datagrams with a
438 smaller buffer, possibly even of length 0. The payload is truncated
439 accordingly. Until the process calls recvmsg() on the error queue,
440 however, the full packet is queued, taking up budget from SO_RCVBUF.
443 2.1.1.5 Blocking Read
444 ~~~~~~~~~~~~~~~~~~~~~
446 Reading from the error queue is always a non-blocking operation. To
447 block waiting on a timestamp, use poll or select. poll() will return
448 POLLERR in pollfd.revents if any data is ready on the error queue.
449 There is no need to pass this flag in pollfd.events. This flag is
450 ignored on request. See also `man 2 poll`.
453 2.1.2 Receive timestamps
454 ^^^^^^^^^^^^^^^^^^^^^^^^
456 On reception, there is no reason to read from the socket error queue.
457 The SCM_TIMESTAMPING ancillary data is sent along with the packet data
458 on a normal recvmsg(). Since this is not a socket error, it is not
459 accompanied by a message SOL_IP(V6)/IP(V6)_RECVERROR. In this case,
460 the meaning of the three fields in struct scm_timestamping is
461 implicitly defined. ts[0] holds a software timestamp if set, ts[1]
462 is again deprecated and ts[2] holds a hardware timestamp if set.
465 3. Hardware Timestamping configuration: SIOCSHWTSTAMP and SIOCGHWTSTAMP
466 =======================================================================
468 Hardware time stamping must also be initialized for each device driver
469 that is expected to do hardware time stamping. The parameter is defined in
470 include/uapi/linux/net_tstamp.h as::
472 struct hwtstamp_config {
473 int flags; /* no flags defined right now, must be zero */
474 int tx_type; /* HWTSTAMP_TX_* */
475 int rx_filter; /* HWTSTAMP_FILTER_* */
478 Desired behavior is passed into the kernel and to a specific device by
479 calling ioctl(SIOCSHWTSTAMP) with a pointer to a struct ifreq whose
480 ifr_data points to a struct hwtstamp_config. The tx_type and
481 rx_filter are hints to the driver what it is expected to do. If
482 the requested fine-grained filtering for incoming packets is not
483 supported, the driver may time stamp more than just the requested types
486 Drivers are free to use a more permissive configuration than the requested
487 configuration. It is expected that drivers should only implement directly the
488 most generic mode that can be supported. For example if the hardware can
489 support HWTSTAMP_FILTER_PTP_V2_EVENT, then it should generally always upscale
490 HWTSTAMP_FILTER_PTP_V2_L2_SYNC, and so forth, as HWTSTAMP_FILTER_PTP_V2_EVENT
491 is more generic (and more useful to applications).
493 A driver which supports hardware time stamping shall update the struct
494 with the actual, possibly more permissive configuration. If the
495 requested packets cannot be time stamped, then nothing should be
496 changed and ERANGE shall be returned (in contrast to EINVAL, which
497 indicates that SIOCSHWTSTAMP is not supported at all).
499 Only a processes with admin rights may change the configuration. User
500 space is responsible to ensure that multiple processes don't interfere
501 with each other and that the settings are reset.
503 Any process can read the actual configuration by passing this
504 structure to ioctl(SIOCGHWTSTAMP) in the same way. However, this has
505 not been implemented in all drivers.
509 /* possible values for hwtstamp_config->tx_type */
512 * no outgoing packet will need hardware time stamping;
513 * should a packet arrive which asks for it, no hardware
514 * time stamping will be done
519 * enables hardware time stamping for outgoing packets;
520 * the sender of the packet decides which are to be
521 * time stamped by setting SOF_TIMESTAMPING_TX_SOFTWARE
522 * before sending the packet
527 /* possible values for hwtstamp_config->rx_filter */
529 /* time stamp no incoming packet at all */
530 HWTSTAMP_FILTER_NONE,
532 /* time stamp any incoming packet */
535 /* return value: time stamp all packets requested plus some others */
536 HWTSTAMP_FILTER_SOME,
538 /* PTP v1, UDP, any kind of event packet */
539 HWTSTAMP_FILTER_PTP_V1_L4_EVENT,
541 /* for the complete list of values, please check
542 * the include file include/uapi/linux/net_tstamp.h
546 3.1 Hardware Timestamping Implementation: Device Drivers
547 --------------------------------------------------------
549 A driver which supports hardware time stamping must support the
550 SIOCSHWTSTAMP ioctl and update the supplied struct hwtstamp_config with
551 the actual values as described in the section on SIOCSHWTSTAMP. It
552 should also support SIOCGHWTSTAMP.
554 Time stamps for received packets must be stored in the skb. To get a pointer
555 to the shared time stamp structure of the skb call skb_hwtstamps(). Then
556 set the time stamps in the structure::
558 struct skb_shared_hwtstamps {
559 /* hardware time stamp transformed into duration
560 * since arbitrary point in time
565 Time stamps for outgoing packets are to be generated as follows:
567 - In hard_start_xmit(), check if (skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)
568 is set no-zero. If yes, then the driver is expected to do hardware time
570 - If this is possible for the skb and requested, then declare
571 that the driver is doing the time stamping by setting the flag
572 SKBTX_IN_PROGRESS in skb_shinfo(skb)->tx_flags , e.g. with::
574 skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
576 You might want to keep a pointer to the associated skb for the next step
577 and not free the skb. A driver not supporting hardware time stamping doesn't
578 do that. A driver must never touch sk_buff::tstamp! It is used to store
579 software generated time stamps by the network subsystem.
580 - Driver should call skb_tx_timestamp() as close to passing sk_buff to hardware
581 as possible. skb_tx_timestamp() provides a software time stamp if requested
582 and hardware timestamping is not possible (SKBTX_IN_PROGRESS not set).
583 - As soon as the driver has sent the packet and/or obtained a
584 hardware time stamp for it, it passes the time stamp back by
585 calling skb_tstamp_tx() with the original skb, the raw
586 hardware time stamp. skb_tstamp_tx() clones the original skb and
587 adds the timestamps, therefore the original skb has to be freed now.
588 If obtaining the hardware time stamp somehow fails, then the driver
589 should not fall back to software time stamping. The rationale is that
590 this would occur at a later time in the processing pipeline than other
591 software time stamping and therefore could lead to unexpected deltas
594 3.2 Special considerations for stacked PTP Hardware Clocks
595 ----------------------------------------------------------
597 There are situations when there may be more than one PHC (PTP Hardware Clock)
598 in the data path of a packet. The kernel has no explicit mechanism to allow the
599 user to select which PHC to use for timestamping Ethernet frames. Instead, the
600 assumption is that the outermost PHC is always the most preferable, and that
601 kernel drivers collaborate towards achieving that goal. Currently there are 3
602 cases of stacked PHCs, detailed below:
604 3.2.1 DSA (Distributed Switch Architecture) switches
605 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
607 These are Ethernet switches which have one of their ports connected to an
608 (otherwise completely unaware) host Ethernet interface, and perform the role of
609 a port multiplier with optional forwarding acceleration features. Each DSA
610 switch port is visible to the user as a standalone (virtual) network interface,
611 and its network I/O is performed, under the hood, indirectly through the host
612 interface (redirecting to the host port on TX, and intercepting frames on RX).
614 When a DSA switch is attached to a host port, PTP synchronization has to
615 suffer, since the switch's variable queuing delay introduces a path delay
616 jitter between the host port and its PTP partner. For this reason, some DSA
617 switches include a timestamping clock of their own, and have the ability to
618 perform network timestamping on their own MAC, such that path delays only
619 measure wire and PHY propagation latencies. Timestamping DSA switches are
620 supported in Linux and expose the same ABI as any other network interface (save
621 for the fact that the DSA interfaces are in fact virtual in terms of network
622 I/O, they do have their own PHC). It is typical, but not mandatory, for all
623 interfaces of a DSA switch to share the same PHC.
625 By design, PTP timestamping with a DSA switch does not need any special
626 handling in the driver for the host port it is attached to. However, when the
627 host port also supports PTP timestamping, DSA will take care of intercepting
628 the ``.ndo_eth_ioctl`` calls towards the host port, and block attempts to enable
629 hardware timestamping on it. This is because the SO_TIMESTAMPING API does not
630 allow the delivery of multiple hardware timestamps for the same packet, so
631 anybody else except for the DSA switch port must be prevented from doing so.
633 In the generic layer, DSA provides the following infrastructure for PTP
636 - ``.port_txtstamp()``: a hook called prior to the transmission of
637 packets with a hardware TX timestamping request from user space.
638 This is required for two-step timestamping, since the hardware
639 timestamp becomes available after the actual MAC transmission, so the
640 driver must be prepared to correlate the timestamp with the original
641 packet so that it can re-enqueue the packet back into the socket's
642 error queue. To save the packet for when the timestamp becomes
643 available, the driver can call ``skb_clone_sk`` , save the clone pointer
644 in skb->cb and enqueue a tx skb queue. Typically, a switch will have a
645 PTP TX timestamp register (or sometimes a FIFO) where the timestamp
646 becomes available. In case of a FIFO, the hardware might store
647 key-value pairs of PTP sequence ID/message type/domain number and the
648 actual timestamp. To perform the correlation correctly between the
649 packets in a queue waiting for timestamping and the actual timestamps,
650 drivers can use a BPF classifier (``ptp_classify_raw``) to identify
651 the PTP transport type, and ``ptp_parse_header`` to interpret the PTP
652 header fields. There may be an IRQ that is raised upon this
653 timestamp's availability, or the driver might have to poll after
654 invoking ``dev_queue_xmit()`` towards the host interface.
655 One-step TX timestamping do not require packet cloning, since there is
656 no follow-up message required by the PTP protocol (because the
657 TX timestamp is embedded into the packet by the MAC), and therefore
658 user space does not expect the packet annotated with the TX timestamp
659 to be re-enqueued into its socket's error queue.
661 - ``.port_rxtstamp()``: On RX, the BPF classifier is run by DSA to
662 identify PTP event messages (any other packets, including PTP general
663 messages, are not timestamped). The original (and only) timestampable
664 skb is provided to the driver, for it to annotate it with a timestamp,
665 if that is immediately available, or defer to later. On reception,
666 timestamps might either be available in-band (through metadata in the
667 DSA header, or attached in other ways to the packet), or out-of-band
668 (through another RX timestamping FIFO). Deferral on RX is typically
669 necessary when retrieving the timestamp needs a sleepable context. In
670 that case, it is the responsibility of the DSA driver to call
671 ``netif_rx()`` on the freshly timestamped skb.
676 These are devices that typically fulfill a Layer 1 role in the network stack,
677 hence they do not have a representation in terms of a network interface as DSA
678 switches do. However, PHYs may be able to detect and timestamp PTP packets, for
679 performance reasons: timestamps taken as close as possible to the wire have the
680 potential to yield a more stable and precise synchronization.
682 A PHY driver that supports PTP timestamping must create a ``struct
683 mii_timestamper`` and add a pointer to it in ``phydev->mii_ts``. The presence
684 of this pointer will be checked by the networking stack.
686 Since PHYs do not have network interface representations, the timestamping and
687 ethtool ioctl operations for them need to be mediated by their respective MAC
688 driver. Therefore, as opposed to DSA switches, modifications need to be done
689 to each individual MAC driver for PHY timestamping support. This entails:
691 - Checking, in ``.ndo_eth_ioctl``, whether ``phy_has_hwtstamp(netdev->phydev)``
692 is true or not. If it is, then the MAC driver should not process this request
693 but instead pass it on to the PHY using ``phy_mii_ioctl()``.
695 - On RX, special intervention may or may not be needed, depending on the
696 function used to deliver skb's up the network stack. In the case of plain
697 ``netif_rx()`` and similar, MAC drivers must check whether
698 ``skb_defer_rx_timestamp(skb)`` is necessary or not - and if it is, don't
699 call ``netif_rx()`` at all. If ``CONFIG_NETWORK_PHY_TIMESTAMPING`` is
700 enabled, and ``skb->dev->phydev->mii_ts`` exists, its ``.rxtstamp()`` hook
701 will be called now, to determine, using logic very similar to DSA, whether
702 deferral for RX timestamping is necessary. Again like DSA, it becomes the
703 responsibility of the PHY driver to send the packet up the stack when the
704 timestamp is available.
706 For other skb receive functions, such as ``napi_gro_receive`` and
707 ``netif_receive_skb``, the stack automatically checks whether
708 ``skb_defer_rx_timestamp()`` is necessary, so this check is not needed inside
711 - On TX, again, special intervention might or might not be needed. The
712 function that calls the ``mii_ts->txtstamp()`` hook is named
713 ``skb_clone_tx_timestamp()``. This function can either be called directly
714 (case in which explicit MAC driver support is indeed needed), but the
715 function also piggybacks from the ``skb_tx_timestamp()`` call, which many MAC
716 drivers already perform for software timestamping purposes. Therefore, if a
717 MAC supports software timestamping, it does not need to do anything further
720 3.2.3 MII bus snooping devices
721 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
723 These perform the same role as timestamping Ethernet PHYs, save for the fact
724 that they are discrete devices and can therefore be used in conjunction with
725 any PHY even if it doesn't support timestamping. In Linux, they are
726 discoverable and attachable to a ``struct phy_device`` through Device Tree, and
727 for the rest, they use the same mii_ts infrastructure as those. See
728 Documentation/devicetree/bindings/ptp/timestamper.txt for more details.
730 3.2.4 Other caveats for MAC drivers
731 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
733 Stacked PHCs, especially DSA (but not only) - since that doesn't require any
734 modification to MAC drivers, so it is more difficult to ensure correctness of
735 all possible code paths - is that they uncover bugs which were impossible to
736 trigger before the existence of stacked PTP clocks. One example has to do with
737 this line of code, already presented earlier::
739 skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
741 Any TX timestamping logic, be it a plain MAC driver, a DSA switch driver, a PHY
742 driver or a MII bus snooping device driver, should set this flag.
743 But a MAC driver that is unaware of PHC stacking might get tripped up by
744 somebody other than itself setting this flag, and deliver a duplicate
746 For example, a typical driver design for TX timestamping might be to split the
747 transmission part into 2 portions:
749 1. "TX": checks whether PTP timestamping has been previously enabled through
750 the ``.ndo_eth_ioctl`` ("``priv->hwtstamp_tx_enabled == true``") and the
751 current skb requires a TX timestamp ("``skb_shinfo(skb)->tx_flags &
752 SKBTX_HW_TSTAMP``"). If this is true, it sets the
753 "``skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS``" flag. Note: as
754 described above, in the case of a stacked PHC system, this condition should
755 never trigger, as this MAC is certainly not the outermost PHC. But this is
756 not where the typical issue is. Transmission proceeds with this packet.
758 2. "TX confirmation": Transmission has finished. The driver checks whether it
759 is necessary to collect any TX timestamp for it. Here is where the typical
760 issues are: the MAC driver takes a shortcut and only checks whether
761 "``skb_shinfo(skb)->tx_flags & SKBTX_IN_PROGRESS``" was set. With a stacked
762 PHC system, this is incorrect because this MAC driver is not the only entity
763 in the TX data path who could have enabled SKBTX_IN_PROGRESS in the first
766 The correct solution for this problem is for MAC drivers to have a compound
767 check in their "TX confirmation" portion, not only for
768 "``skb_shinfo(skb)->tx_flags & SKBTX_IN_PROGRESS``", but also for
769 "``priv->hwtstamp_tx_enabled == true``". Because the rest of the system ensures
770 that PTP timestamping is not enabled for anything other than the outermost PHC,
771 this enhanced check will avoid delivering a duplicated TX timestamp to user
projects / platform / kernel / linux-starfive.git / blob