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):
60 This option is identical to SO_TIMESTAMP except for the returned data type.
61 Its struct timespec allows for higher resolution (ns) timestamps than the
62 timeval of SO_TIMESTAMP (ms).
64 Always use SO_TIMESTAMPNS_NEW timestamp to always get timestamp in
65 struct __kernel_timespec format.
67 SO_TIMESTAMPNS_OLD returns incorrect timestamps after the year 2038
70 1.3 SO_TIMESTAMPING (also SO_TIMESTAMPING_OLD and SO_TIMESTAMPING_NEW)
71 ----------------------------------------------------------------------
73 Supports multiple types of timestamp requests. As a result, this
74 socket option takes a bitmap of flags, not a boolean. In::
76 err = setsockopt(fd, SOL_SOCKET, SO_TIMESTAMPING, &val, sizeof(val));
78 val is an integer with any of the following bits set. Setting other
79 bit returns EINVAL and does not change the current state.
81 The socket option configures timestamp generation for individual
82 sk_buffs (1.3.1), timestamp reporting to the socket's error
83 queue (1.3.2) and options (1.3.3). Timestamp generation can also
84 be enabled for individual sendmsg calls using cmsg (1.3.4).
87 1.3.1 Timestamp Generation
88 ^^^^^^^^^^^^^^^^^^^^^^^^^^
90 Some bits are requests to the stack to try to generate timestamps. Any
91 combination of them is valid. Changes to these bits apply to newly
92 created packets, not to packets already in the stack. As a result, it
93 is possible to selectively request timestamps for a subset of packets
94 (e.g., for sampling) by embedding an send() call within two setsockopt
95 calls, one to enable timestamp generation and one to disable it.
96 Timestamps may also be generated for reasons other than being
97 requested by a particular socket, such as when receive timestamping is
98 enabled system wide, as explained earlier.
100 SOF_TIMESTAMPING_RX_HARDWARE:
101 Request rx timestamps generated by the network adapter.
103 SOF_TIMESTAMPING_RX_SOFTWARE:
104 Request rx timestamps when data enters the kernel. These timestamps
105 are generated just after a device driver hands a packet to the
106 kernel receive stack.
108 SOF_TIMESTAMPING_TX_HARDWARE:
109 Request tx timestamps generated by the network adapter. This flag
110 can be enabled via both socket options and control messages.
112 SOF_TIMESTAMPING_TX_SOFTWARE:
113 Request tx timestamps when data leaves the kernel. These timestamps
114 are generated in the device driver as close as possible, but always
115 prior to, passing the packet to the network interface. Hence, they
116 require driver support and may not be available for all devices.
117 This flag can be enabled via both socket options and control messages.
119 SOF_TIMESTAMPING_TX_SCHED:
120 Request tx timestamps prior to entering the packet scheduler. Kernel
121 transmit latency is, if long, often dominated by queuing delay. The
122 difference between this timestamp and one taken at
123 SOF_TIMESTAMPING_TX_SOFTWARE will expose this latency independent
124 of protocol processing. The latency incurred in protocol
125 processing, if any, can be computed by subtracting a userspace
126 timestamp taken immediately before send() from this timestamp. On
127 machines with virtual devices where a transmitted packet travels
128 through multiple devices and, hence, multiple packet schedulers,
129 a timestamp is generated at each layer. This allows for fine
130 grained measurement of queuing delay. This flag can be enabled
131 via both socket options and control messages.
133 SOF_TIMESTAMPING_TX_ACK:
134 Request tx timestamps when all data in the send buffer has been
135 acknowledged. This only makes sense for reliable protocols. It is
136 currently only implemented for TCP. For that protocol, it may
137 over-report measurement, because the timestamp is generated when all
138 data up to and including the buffer at send() was acknowledged: the
139 cumulative acknowledgment. The mechanism ignores SACK and FACK.
140 This flag can be enabled via both socket options and control messages.
143 1.3.2 Timestamp Reporting
144 ^^^^^^^^^^^^^^^^^^^^^^^^^
146 The other three bits control which timestamps will be reported in a
147 generated control message. Changes to the bits take immediate
148 effect at the timestamp reporting locations in the stack. Timestamps
149 are only reported for packets that also have the relevant timestamp
150 generation request set.
152 SOF_TIMESTAMPING_SOFTWARE:
153 Report any software timestamps when available.
155 SOF_TIMESTAMPING_SYS_HARDWARE:
156 This option is deprecated and ignored.
158 SOF_TIMESTAMPING_RAW_HARDWARE:
159 Report hardware timestamps as generated by
160 SOF_TIMESTAMPING_TX_HARDWARE when available.
163 1.3.3 Timestamp Options
164 ^^^^^^^^^^^^^^^^^^^^^^^
166 The interface supports the options
168 SOF_TIMESTAMPING_OPT_ID:
169 Generate a unique identifier along with each packet. A process can
170 have multiple concurrent timestamping requests outstanding. Packets
171 can be reordered in the transmit path, for instance in the packet
172 scheduler. In that case timestamps will be queued onto the error
173 queue out of order from the original send() calls. It is not always
174 possible to uniquely match timestamps to the original send() calls
175 based on timestamp order or payload inspection alone, then.
177 This option associates each packet at send() with a unique
178 identifier and returns that along with the timestamp. The identifier
179 is derived from a per-socket u32 counter (that wraps). For datagram
180 sockets, the counter increments with each sent packet. For stream
181 sockets, it increments with every byte.
183 The counter starts at zero. It is initialized the first time that
184 the socket option is enabled. It is reset each time the option is
185 enabled after having been disabled. Resetting the counter does not
186 change the identifiers of existing packets in the system.
188 This option is implemented only for transmit timestamps. There, the
189 timestamp is always looped along with a struct sock_extended_err.
190 The option modifies field ee_data to pass an id that is unique
191 among all possibly concurrently outstanding timestamp requests for
195 SOF_TIMESTAMPING_OPT_CMSG:
196 Support recv() cmsg for all timestamped packets. Control messages
197 are already supported unconditionally on all packets with receive
198 timestamps and on IPv6 packets with transmit timestamp. This option
199 extends them to IPv4 packets with transmit timestamp. One use case
200 is to correlate packets with their egress device, by enabling socket
201 option IP_PKTINFO simultaneously.
204 SOF_TIMESTAMPING_OPT_TSONLY:
205 Applies to transmit timestamps only. Makes the kernel return the
206 timestamp as a cmsg alongside an empty packet, as opposed to
207 alongside the original packet. This reduces the amount of memory
208 charged to the socket's receive budget (SO_RCVBUF) and delivers
209 the timestamp even if sysctl net.core.tstamp_allow_data is 0.
210 This option disables SOF_TIMESTAMPING_OPT_CMSG.
212 SOF_TIMESTAMPING_OPT_STATS:
213 Optional stats that are obtained along with the transmit timestamps.
214 It must be used together with SOF_TIMESTAMPING_OPT_TSONLY. When the
215 transmit timestamp is available, the stats are available in a
216 separate control message of type SCM_TIMESTAMPING_OPT_STATS, as a
217 list of TLVs (struct nlattr) of types. These stats allow the
218 application to associate various transport layer stats with
219 the transmit timestamps, such as how long a certain block of
220 data was limited by peer's receiver window.
222 SOF_TIMESTAMPING_OPT_PKTINFO:
223 Enable the SCM_TIMESTAMPING_PKTINFO control message for incoming
224 packets with hardware timestamps. The message contains struct
225 scm_ts_pktinfo, which supplies the index of the real interface which
226 received the packet and its length at layer 2. A valid (non-zero)
227 interface index will be returned only if CONFIG_NET_RX_BUSY_POLL is
228 enabled and the driver is using NAPI. The struct contains also two
229 other fields, but they are reserved and undefined.
231 SOF_TIMESTAMPING_OPT_TX_SWHW:
232 Request both hardware and software timestamps for outgoing packets
233 when SOF_TIMESTAMPING_TX_HARDWARE and SOF_TIMESTAMPING_TX_SOFTWARE
234 are enabled at the same time. If both timestamps are generated,
235 two separate messages will be looped to the socket's error queue,
236 each containing just one timestamp.
238 New applications are encouraged to pass SOF_TIMESTAMPING_OPT_ID to
239 disambiguate timestamps and SOF_TIMESTAMPING_OPT_TSONLY to operate
240 regardless of the setting of sysctl net.core.tstamp_allow_data.
242 An exception is when a process needs additional cmsg data, for
243 instance SOL_IP/IP_PKTINFO to detect the egress network interface.
244 Then pass option SOF_TIMESTAMPING_OPT_CMSG. This option depends on
245 having access to the contents of the original packet, so cannot be
246 combined with SOF_TIMESTAMPING_OPT_TSONLY.
249 1.3.4. Enabling timestamps via control messages
250 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
252 In addition to socket options, timestamp generation can be requested
253 per write via cmsg, only for SOF_TIMESTAMPING_TX_* (see Section 1.3.1).
254 Using this feature, applications can sample timestamps per sendmsg()
255 without paying the overhead of enabling and disabling timestamps via
260 cmsg = CMSG_FIRSTHDR(msg);
261 cmsg->cmsg_level = SOL_SOCKET;
262 cmsg->cmsg_type = SO_TIMESTAMPING;
263 cmsg->cmsg_len = CMSG_LEN(sizeof(__u32));
264 *((__u32 *) CMSG_DATA(cmsg)) = SOF_TIMESTAMPING_TX_SCHED |
265 SOF_TIMESTAMPING_TX_SOFTWARE |
266 SOF_TIMESTAMPING_TX_ACK;
267 err = sendmsg(fd, msg, 0);
269 The SOF_TIMESTAMPING_TX_* flags set via cmsg will override
270 the SOF_TIMESTAMPING_TX_* flags set via setsockopt.
272 Moreover, applications must still enable timestamp reporting via
273 setsockopt to receive timestamps::
275 __u32 val = SOF_TIMESTAMPING_SOFTWARE |
276 SOF_TIMESTAMPING_OPT_ID /* or any other flag */;
277 err = setsockopt(fd, SOL_SOCKET, SO_TIMESTAMPING, &val, sizeof(val));
280 1.4 Bytestream Timestamps
281 -------------------------
283 The SO_TIMESTAMPING interface supports timestamping of bytes in a
284 bytestream. Each request is interpreted as a request for when the
285 entire contents of the buffer has passed a timestamping point. That
286 is, for streams option SOF_TIMESTAMPING_TX_SOFTWARE will record
287 when all bytes have reached the device driver, regardless of how
288 many packets the data has been converted into.
290 In general, bytestreams have no natural delimiters and therefore
291 correlating a timestamp with data is non-trivial. A range of bytes
292 may be split across segments, any segments may be merged (possibly
293 coalescing sections of previously segmented buffers associated with
294 independent send() calls). Segments can be reordered and the same
295 byte range can coexist in multiple segments for protocols that
296 implement retransmissions.
298 It is essential that all timestamps implement the same semantics,
299 regardless of these possible transformations, as otherwise they are
300 incomparable. Handling "rare" corner cases differently from the
301 simple case (a 1:1 mapping from buffer to skb) is insufficient
302 because performance debugging often needs to focus on such outliers.
304 In practice, timestamps can be correlated with segments of a
305 bytestream consistently, if both semantics of the timestamp and the
306 timing of measurement are chosen correctly. This challenge is no
307 different from deciding on a strategy for IP fragmentation. There, the
308 definition is that only the first fragment is timestamped. For
309 bytestreams, we chose that a timestamp is generated only when all
310 bytes have passed a point. SOF_TIMESTAMPING_TX_ACK as defined is easy to
311 implement and reason about. An implementation that has to take into
312 account SACK would be more complex due to possible transmission holes
313 and out of order arrival.
315 On the host, TCP can also break the simple 1:1 mapping from buffer to
316 skbuff as a result of Nagle, cork, autocork, segmentation and GSO. The
317 implementation ensures correctness in all cases by tracking the
318 individual last byte passed to send(), even if it is no longer the
319 last byte after an skbuff extend or merge operation. It stores the
320 relevant sequence number in skb_shinfo(skb)->tskey. Because an skbuff
321 has only one such field, only one timestamp can be generated.
323 In rare cases, a timestamp request can be missed if two requests are
324 collapsed onto the same skb. A process can detect this situation by
325 enabling SOF_TIMESTAMPING_OPT_ID and comparing the byte offset at
326 send time with the value returned for each timestamp. It can prevent
327 the situation by always flushing the TCP stack in between requests,
328 for instance by enabling TCP_NODELAY and disabling TCP_CORK and
331 These precautions ensure that the timestamp is generated only when all
332 bytes have passed a timestamp point, assuming that the network stack
333 itself does not reorder the segments. The stack indeed tries to avoid
334 reordering. The one exception is under administrator control: it is
335 possible to construct a packet scheduler configuration that delays
336 segments from the same stream differently. Such a setup would be
343 Timestamps are read using the ancillary data feature of recvmsg().
344 See `man 3 cmsg` for details of this interface. The socket manual
345 page (`man 7 socket`) describes how timestamps generated with
346 SO_TIMESTAMP and SO_TIMESTAMPNS records can be retrieved.
349 2.1 SCM_TIMESTAMPING records
350 ----------------------------
352 These timestamps are returned in a control message with cmsg_level
353 SOL_SOCKET, cmsg_type SCM_TIMESTAMPING, and payload of type
355 For SO_TIMESTAMPING_OLD::
357 struct scm_timestamping {
358 struct timespec ts[3];
361 For SO_TIMESTAMPING_NEW::
363 struct scm_timestamping64 {
364 struct __kernel_timespec ts[3];
366 Always use SO_TIMESTAMPING_NEW timestamp to always get timestamp in
367 struct scm_timestamping64 format.
369 SO_TIMESTAMPING_OLD returns incorrect timestamps after the year 2038
372 The structure can return up to three timestamps. This is a legacy
373 feature. At least one field is non-zero at any time. Most timestamps
374 are passed in ts[0]. Hardware timestamps are passed in ts[2].
376 ts[1] used to hold hardware timestamps converted to system time.
377 Instead, expose the hardware clock device on the NIC directly as
378 a HW PTP clock source, to allow time conversion in userspace and
379 optionally synchronize system time with a userspace PTP stack such
380 as linuxptp. For the PTP clock API, see Documentation/driver-api/ptp.rst.
382 Note that if the SO_TIMESTAMP or SO_TIMESTAMPNS option is enabled
383 together with SO_TIMESTAMPING using SOF_TIMESTAMPING_SOFTWARE, a false
384 software timestamp will be generated in the recvmsg() call and passed
385 in ts[0] when a real software timestamp is missing. This happens also
386 on hardware transmit timestamps.
388 2.1.1 Transmit timestamps with MSG_ERRQUEUE
389 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
391 For transmit timestamps the outgoing packet is looped back to the
392 socket's error queue with the send timestamp(s) attached. A process
393 receives the timestamps by calling recvmsg() with flag MSG_ERRQUEUE
394 set and with a msg_control buffer sufficiently large to receive the
395 relevant metadata structures. The recvmsg call returns the original
396 outgoing data packet with two ancillary messages attached.
398 A message of cm_level SOL_IP(V6) and cm_type IP(V6)_RECVERR
399 embeds a struct sock_extended_err. This defines the error type. For
400 timestamps, the ee_errno field is ENOMSG. The other ancillary message
401 will have cm_level SOL_SOCKET and cm_type SCM_TIMESTAMPING. This
402 embeds the struct scm_timestamping.
405 2.1.1.2 Timestamp types
406 ~~~~~~~~~~~~~~~~~~~~~~~
408 The semantics of the three struct timespec are defined by field
409 ee_info in the extended error structure. It contains a value of
410 type SCM_TSTAMP_* to define the actual timestamp passed in
413 The SCM_TSTAMP_* types are 1:1 matches to the SOF_TIMESTAMPING_*
414 control fields discussed previously, with one exception. For legacy
415 reasons, SCM_TSTAMP_SND is equal to zero and can be set for both
416 SOF_TIMESTAMPING_TX_HARDWARE and SOF_TIMESTAMPING_TX_SOFTWARE. It
417 is the first if ts[2] is non-zero, the second otherwise, in which
418 case the timestamp is stored in ts[0].
421 2.1.1.3 Fragmentation
422 ~~~~~~~~~~~~~~~~~~~~~
424 Fragmentation of outgoing datagrams is rare, but is possible, e.g., by
425 explicitly disabling PMTU discovery. If an outgoing packet is fragmented,
426 then only the first fragment is timestamped and returned to the sending
430 2.1.1.4 Packet Payload
431 ~~~~~~~~~~~~~~~~~~~~~~
433 The calling application is often not interested in receiving the whole
434 packet payload that it passed to the stack originally: the socket
435 error queue mechanism is just a method to piggyback the timestamp on.
436 In this case, the application can choose to read datagrams with a
437 smaller buffer, possibly even of length 0. The payload is truncated
438 accordingly. Until the process calls recvmsg() on the error queue,
439 however, the full packet is queued, taking up budget from SO_RCVBUF.
442 2.1.1.5 Blocking Read
443 ~~~~~~~~~~~~~~~~~~~~~
445 Reading from the error queue is always a non-blocking operation. To
446 block waiting on a timestamp, use poll or select. poll() will return
447 POLLERR in pollfd.revents if any data is ready on the error queue.
448 There is no need to pass this flag in pollfd.events. This flag is
449 ignored on request. See also `man 2 poll`.
452 2.1.2 Receive timestamps
453 ^^^^^^^^^^^^^^^^^^^^^^^^
455 On reception, there is no reason to read from the socket error queue.
456 The SCM_TIMESTAMPING ancillary data is sent along with the packet data
457 on a normal recvmsg(). Since this is not a socket error, it is not
458 accompanied by a message SOL_IP(V6)/IP(V6)_RECVERROR. In this case,
459 the meaning of the three fields in struct scm_timestamping is
460 implicitly defined. ts[0] holds a software timestamp if set, ts[1]
461 is again deprecated and ts[2] holds a hardware timestamp if set.
464 3. Hardware Timestamping configuration: SIOCSHWTSTAMP and SIOCGHWTSTAMP
465 =======================================================================
467 Hardware time stamping must also be initialized for each device driver
468 that is expected to do hardware time stamping. The parameter is defined in
469 include/uapi/linux/net_tstamp.h as::
471 struct hwtstamp_config {
472 int flags; /* no flags defined right now, must be zero */
473 int tx_type; /* HWTSTAMP_TX_* */
474 int rx_filter; /* HWTSTAMP_FILTER_* */
477 Desired behavior is passed into the kernel and to a specific device by
478 calling ioctl(SIOCSHWTSTAMP) with a pointer to a struct ifreq whose
479 ifr_data points to a struct hwtstamp_config. The tx_type and
480 rx_filter are hints to the driver what it is expected to do. If
481 the requested fine-grained filtering for incoming packets is not
482 supported, the driver may time stamp more than just the requested types
485 Drivers are free to use a more permissive configuration than the requested
486 configuration. It is expected that drivers should only implement directly the
487 most generic mode that can be supported. For example if the hardware can
488 support HWTSTAMP_FILTER_V2_EVENT, then it should generally always upscale
489 HWTSTAMP_FILTER_V2_L2_SYNC_MESSAGE, and so forth, as HWTSTAMP_FILTER_V2_EVENT
490 is more generic (and more useful to applications).
492 A driver which supports hardware time stamping shall update the struct
493 with the actual, possibly more permissive configuration. If the
494 requested packets cannot be time stamped, then nothing should be
495 changed and ERANGE shall be returned (in contrast to EINVAL, which
496 indicates that SIOCSHWTSTAMP is not supported at all).
498 Only a processes with admin rights may change the configuration. User
499 space is responsible to ensure that multiple processes don't interfere
500 with each other and that the settings are reset.
502 Any process can read the actual configuration by passing this
503 structure to ioctl(SIOCGHWTSTAMP) in the same way. However, this has
504 not been implemented in all drivers.
508 /* possible values for hwtstamp_config->tx_type */
511 * no outgoing packet will need hardware time stamping;
512 * should a packet arrive which asks for it, no hardware
513 * time stamping will be done
518 * enables hardware time stamping for outgoing packets;
519 * the sender of the packet decides which are to be
520 * time stamped by setting SOF_TIMESTAMPING_TX_SOFTWARE
521 * before sending the packet
526 /* possible values for hwtstamp_config->rx_filter */
528 /* time stamp no incoming packet at all */
529 HWTSTAMP_FILTER_NONE,
531 /* time stamp any incoming packet */
534 /* return value: time stamp all packets requested plus some others */
535 HWTSTAMP_FILTER_SOME,
537 /* PTP v1, UDP, any kind of event packet */
538 HWTSTAMP_FILTER_PTP_V1_L4_EVENT,
540 /* for the complete list of values, please check
541 * the include file include/uapi/linux/net_tstamp.h
545 3.1 Hardware Timestamping Implementation: Device Drivers
546 --------------------------------------------------------
548 A driver which supports hardware time stamping must support the
549 SIOCSHWTSTAMP ioctl and update the supplied struct hwtstamp_config with
550 the actual values as described in the section on SIOCSHWTSTAMP. It
551 should also support SIOCGHWTSTAMP.
553 Time stamps for received packets must be stored in the skb. To get a pointer
554 to the shared time stamp structure of the skb call skb_hwtstamps(). Then
555 set the time stamps in the structure::
557 struct skb_shared_hwtstamps {
558 /* hardware time stamp transformed into duration
559 * since arbitrary point in time
564 Time stamps for outgoing packets are to be generated as follows:
566 - In hard_start_xmit(), check if (skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)
567 is set no-zero. If yes, then the driver is expected to do hardware time
569 - If this is possible for the skb and requested, then declare
570 that the driver is doing the time stamping by setting the flag
571 SKBTX_IN_PROGRESS in skb_shinfo(skb)->tx_flags , e.g. with::
573 skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
575 You might want to keep a pointer to the associated skb for the next step
576 and not free the skb. A driver not supporting hardware time stamping doesn't
577 do that. A driver must never touch sk_buff::tstamp! It is used to store
578 software generated time stamps by the network subsystem.
579 - Driver should call skb_tx_timestamp() as close to passing sk_buff to hardware
580 as possible. skb_tx_timestamp() provides a software time stamp if requested
581 and hardware timestamping is not possible (SKBTX_IN_PROGRESS not set).
582 - As soon as the driver has sent the packet and/or obtained a
583 hardware time stamp for it, it passes the time stamp back by
584 calling skb_hwtstamp_tx() with the original skb, the raw
585 hardware time stamp. skb_hwtstamp_tx() clones the original skb and
586 adds the timestamps, therefore the original skb has to be freed now.
587 If obtaining the hardware time stamp somehow fails, then the driver
588 should not fall back to software time stamping. The rationale is that
589 this would occur at a later time in the processing pipeline than other
590 software time stamping and therefore could lead to unexpected deltas
593 3.2 Special considerations for stacked PTP Hardware Clocks
594 ----------------------------------------------------------
596 There are situations when there may be more than one PHC (PTP Hardware Clock)
597 in the data path of a packet. The kernel has no explicit mechanism to allow the
598 user to select which PHC to use for timestamping Ethernet frames. Instead, the
599 assumption is that the outermost PHC is always the most preferable, and that
600 kernel drivers collaborate towards achieving that goal. Currently there are 3
601 cases of stacked PHCs, detailed below:
603 3.2.1 DSA (Distributed Switch Architecture) switches
604 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
606 These are Ethernet switches which have one of their ports connected to an
607 (otherwise completely unaware) host Ethernet interface, and perform the role of
608 a port multiplier with optional forwarding acceleration features. Each DSA
609 switch port is visible to the user as a standalone (virtual) network interface,
610 and its network I/O is performed, under the hood, indirectly through the host
611 interface (redirecting to the host port on TX, and intercepting frames on RX).
613 When a DSA switch is attached to a host port, PTP synchronization has to
614 suffer, since the switch's variable queuing delay introduces a path delay
615 jitter between the host port and its PTP partner. For this reason, some DSA
616 switches include a timestamping clock of their own, and have the ability to
617 perform network timestamping on their own MAC, such that path delays only
618 measure wire and PHY propagation latencies. Timestamping DSA switches are
619 supported in Linux and expose the same ABI as any other network interface (save
620 for the fact that the DSA interfaces are in fact virtual in terms of network
621 I/O, they do have their own PHC). It is typical, but not mandatory, for all
622 interfaces of a DSA switch to share the same PHC.
624 By design, PTP timestamping with a DSA switch does not need any special
625 handling in the driver for the host port it is attached to. However, when the
626 host port also supports PTP timestamping, DSA will take care of intercepting
627 the ``.ndo_do_ioctl`` calls towards the host port, and block attempts to enable
628 hardware timestamping on it. This is because the SO_TIMESTAMPING API does not
629 allow the delivery of multiple hardware timestamps for the same packet, so
630 anybody else except for the DSA switch port must be prevented from doing so.
632 In code, DSA provides for most of the infrastructure for timestamping already,
633 in generic code: a BPF classifier (``ptp_classify_raw``) is used to identify
634 PTP event messages (any other packets, including PTP general messages, are not
635 timestamped), and provides two hooks to drivers:
637 - ``.port_txtstamp()``: The driver is passed a clone of the timestampable skb
638 to be transmitted, before actually transmitting it. Typically, a switch will
639 have a PTP TX timestamp register (or sometimes a FIFO) where the timestamp
640 becomes available. There may be an IRQ that is raised upon this timestamp's
641 availability, or the driver might have to poll after invoking
642 ``dev_queue_xmit()`` towards the host interface. Either way, in the
643 ``.port_txtstamp()`` method, the driver only needs to save the clone for
644 later use (when the timestamp becomes available). Each skb is annotated with
645 a pointer to its clone, in ``DSA_SKB_CB(skb)->clone``, to ease the driver's
646 job of keeping track of which clone belongs to which skb.
648 - ``.port_rxtstamp()``: The original (and only) timestampable skb is provided
649 to the driver, for it to annotate it with a timestamp, if that is immediately
650 available, or defer to later. On reception, timestamps might either be
651 available in-band (through metadata in the DSA header, or attached in other
652 ways to the packet), or out-of-band (through another RX timestamping FIFO).
653 Deferral on RX is typically necessary when retrieving the timestamp needs a
654 sleepable context. In that case, it is the responsibility of the DSA driver
655 to call ``netif_rx_ni()`` on the freshly timestamped skb.
660 These are devices that typically fulfill a Layer 1 role in the network stack,
661 hence they do not have a representation in terms of a network interface as DSA
662 switches do. However, PHYs may be able to detect and timestamp PTP packets, for
663 performance reasons: timestamps taken as close as possible to the wire have the
664 potential to yield a more stable and precise synchronization.
666 A PHY driver that supports PTP timestamping must create a ``struct
667 mii_timestamper`` and add a pointer to it in ``phydev->mii_ts``. The presence
668 of this pointer will be checked by the networking stack.
670 Since PHYs do not have network interface representations, the timestamping and
671 ethtool ioctl operations for them need to be mediated by their respective MAC
672 driver. Therefore, as opposed to DSA switches, modifications need to be done
673 to each individual MAC driver for PHY timestamping support. This entails:
675 - Checking, in ``.ndo_do_ioctl``, whether ``phy_has_hwtstamp(netdev->phydev)``
676 is true or not. If it is, then the MAC driver should not process this request
677 but instead pass it on to the PHY using ``phy_mii_ioctl()``.
679 - On RX, special intervention may or may not be needed, depending on the
680 function used to deliver skb's up the network stack. In the case of plain
681 ``netif_rx()`` and similar, MAC drivers must check whether
682 ``skb_defer_rx_timestamp(skb)`` is necessary or not - and if it is, don't
683 call ``netif_rx()`` at all. If ``CONFIG_NETWORK_PHY_TIMESTAMPING`` is
684 enabled, and ``skb->dev->phydev->mii_ts`` exists, its ``.rxtstamp()`` hook
685 will be called now, to determine, using logic very similar to DSA, whether
686 deferral for RX timestamping is necessary. Again like DSA, it becomes the
687 responsibility of the PHY driver to send the packet up the stack when the
688 timestamp is available.
690 For other skb receive functions, such as ``napi_gro_receive`` and
691 ``netif_receive_skb``, the stack automatically checks whether
692 ``skb_defer_rx_timestamp()`` is necessary, so this check is not needed inside
695 - On TX, again, special intervention might or might not be needed. The
696 function that calls the ``mii_ts->txtstamp()`` hook is named
697 ``skb_clone_tx_timestamp()``. This function can either be called directly
698 (case in which explicit MAC driver support is indeed needed), but the
699 function also piggybacks from the ``skb_tx_timestamp()`` call, which many MAC
700 drivers already perform for software timestamping purposes. Therefore, if a
701 MAC supports software timestamping, it does not need to do anything further
704 3.2.3 MII bus snooping devices
705 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
707 These perform the same role as timestamping Ethernet PHYs, save for the fact
708 that they are discrete devices and can therefore be used in conjunction with
709 any PHY even if it doesn't support timestamping. In Linux, they are
710 discoverable and attachable to a ``struct phy_device`` through Device Tree, and
711 for the rest, they use the same mii_ts infrastructure as those. See
712 Documentation/devicetree/bindings/ptp/timestamper.txt for more details.
714 3.2.4 Other caveats for MAC drivers
715 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
717 Stacked PHCs, especially DSA (but not only) - since that doesn't require any
718 modification to MAC drivers, so it is more difficult to ensure correctness of
719 all possible code paths - is that they uncover bugs which were impossible to
720 trigger before the existence of stacked PTP clocks. One example has to do with
721 this line of code, already presented earlier::
723 skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
725 Any TX timestamping logic, be it a plain MAC driver, a DSA switch driver, a PHY
726 driver or a MII bus snooping device driver, should set this flag.
727 But a MAC driver that is unaware of PHC stacking might get tripped up by
728 somebody other than itself setting this flag, and deliver a duplicate
730 For example, a typical driver design for TX timestamping might be to split the
731 transmission part into 2 portions:
733 1. "TX": checks whether PTP timestamping has been previously enabled through
734 the ``.ndo_do_ioctl`` ("``priv->hwtstamp_tx_enabled == true``") and the
735 current skb requires a TX timestamp ("``skb_shinfo(skb)->tx_flags &
736 SKBTX_HW_TSTAMP``"). If this is true, it sets the
737 "``skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS``" flag. Note: as
738 described above, in the case of a stacked PHC system, this condition should
739 never trigger, as this MAC is certainly not the outermost PHC. But this is
740 not where the typical issue is. Transmission proceeds with this packet.
742 2. "TX confirmation": Transmission has finished. The driver checks whether it
743 is necessary to collect any TX timestamp for it. Here is where the typical
744 issues are: the MAC driver takes a shortcut and only checks whether
745 "``skb_shinfo(skb)->tx_flags & SKBTX_IN_PROGRESS``" was set. With a stacked
746 PHC system, this is incorrect because this MAC driver is not the only entity
747 in the TX data path who could have enabled SKBTX_IN_PROGRESS in the first
750 The correct solution for this problem is for MAC drivers to have a compound
751 check in their "TX confirmation" portion, not only for
752 "``skb_shinfo(skb)->tx_flags & SKBTX_IN_PROGRESS``", but also for
753 "``priv->hwtstamp_tx_enabled == true``". Because the rest of the system ensures
754 that PTP timestamping is not enabled for anything other than the outermost PHC,
755 this enhanced check will avoid delivering a duplicated TX timestamp to user
projects / linux-2.6-microblaze.git / blob