xref: /linux/Documentation/networking/timestamping.rst (revision 45d8b572fac3aa8b49d53c946b3685eaf78a2824)
1.. SPDX-License-Identifier: GPL-2.0
2
3============
4Timestamping
5============
6
7
81. Control Interfaces
9=====================
10
11The interfaces for receiving network packages timestamps are:
12
13SO_TIMESTAMP
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.
22
23SO_TIMESTAMPNS
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
30  respectively.
31
32IP_MULTICAST_LOOP + SO_TIMESTAMP[NS]
33  Only for multicast:approximate transmit timestamp obtained by
34  reading the looped packet receive timestamp.
35
36SO_TIMESTAMPING
37  Generates timestamps on reception, transmission or both. Supports
38  multiple timestamp sources, including hardware. Supports generating
39  timestamps for stream sockets.
40
41
421.1 SO_TIMESTAMP (also SO_TIMESTAMP_OLD and SO_TIMESTAMP_NEW)
43-------------------------------------------------------------
44
45This socket option enables timestamping of datagrams on the reception
46path. Because the destination socket, if any, is not known early in
47the network stack, the feature has to be enabled for all packets. The
48same is true for all early receive timestamp options.
49
50For interface details, see `man 7 socket`.
51
52Always use SO_TIMESTAMP_NEW timestamp to always get timestamp in
53struct __kernel_sock_timeval format.
54
55SO_TIMESTAMP_OLD returns incorrect timestamps after the year 2038
56on 32 bit machines.
57
581.2 SO_TIMESTAMPNS (also SO_TIMESTAMPNS_OLD and SO_TIMESTAMPNS_NEW)
59-------------------------------------------------------------------
60
61This option is identical to SO_TIMESTAMP except for the returned data type.
62Its struct timespec allows for higher resolution (ns) timestamps than the
63timeval of SO_TIMESTAMP (ms).
64
65Always use SO_TIMESTAMPNS_NEW timestamp to always get timestamp in
66struct __kernel_timespec format.
67
68SO_TIMESTAMPNS_OLD returns incorrect timestamps after the year 2038
69on 32 bit machines.
70
711.3 SO_TIMESTAMPING (also SO_TIMESTAMPING_OLD and SO_TIMESTAMPING_NEW)
72----------------------------------------------------------------------
73
74Supports multiple types of timestamp requests. As a result, this
75socket option takes a bitmap of flags, not a boolean. In::
76
77  err = setsockopt(fd, SOL_SOCKET, SO_TIMESTAMPING, &val, sizeof(val));
78
79val is an integer with any of the following bits set. Setting other
80bit returns EINVAL and does not change the current state.
81
82The socket option configures timestamp generation for individual
83sk_buffs (1.3.1), timestamp reporting to the socket's error
84queue (1.3.2) and options (1.3.3). Timestamp generation can also
85be enabled for individual sendmsg calls using cmsg (1.3.4).
86
87
881.3.1 Timestamp Generation
89^^^^^^^^^^^^^^^^^^^^^^^^^^
90
91Some bits are requests to the stack to try to generate timestamps. Any
92combination of them is valid. Changes to these bits apply to newly
93created packets, not to packets already in the stack. As a result, it
94is possible to selectively request timestamps for a subset of packets
95(e.g., for sampling) by embedding an send() call within two setsockopt
96calls, one to enable timestamp generation and one to disable it.
97Timestamps may also be generated for reasons other than being
98requested by a particular socket, such as when receive timestamping is
99enabled system wide, as explained earlier.
100
101SOF_TIMESTAMPING_RX_HARDWARE:
102  Request rx timestamps generated by the network adapter.
103
104SOF_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.
108
109SOF_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.
112
113SOF_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.
119
120SOF_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.
133
134SOF_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.
142
143
1441.3.2 Timestamp Reporting
145^^^^^^^^^^^^^^^^^^^^^^^^^
146
147The other three bits control which timestamps will be reported in a
148generated control message. Changes to the bits take immediate
149effect at the timestamp reporting locations in the stack. Timestamps
150are only reported for packets that also have the relevant timestamp
151generation request set.
152
153SOF_TIMESTAMPING_SOFTWARE:
154  Report any software timestamps when available.
155
156SOF_TIMESTAMPING_SYS_HARDWARE:
157  This option is deprecated and ignored.
158
159SOF_TIMESTAMPING_RAW_HARDWARE:
160  Report hardware timestamps as generated by
161  SOF_TIMESTAMPING_TX_HARDWARE when available.
162
163
1641.3.3 Timestamp Options
165^^^^^^^^^^^^^^^^^^^^^^^
166
167The interface supports the options
168
169SOF_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.
177
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. For stream sockets, also set
183  SOF_TIMESTAMPING_OPT_ID_TCP, see the section below.
184
185  The counter starts at zero. It is initialized the first time that
186  the socket option is enabled. It is reset each time the option is
187  enabled after having been disabled. Resetting the counter does not
188  change the identifiers of existing packets in the system.
189
190  This option is implemented only for transmit timestamps. There, the
191  timestamp is always looped along with a struct sock_extended_err.
192  The option modifies field ee_data to pass an id that is unique
193  among all possibly concurrently outstanding timestamp requests for
194  that socket.
195
196SOF_TIMESTAMPING_OPT_ID_TCP:
197  Pass this modifier along with SOF_TIMESTAMPING_OPT_ID for new TCP
198  timestamping applications. SOF_TIMESTAMPING_OPT_ID defines how the
199  counter increments for stream sockets, but its starting point is
200  not entirely trivial. This option fixes that.
201
202  For stream sockets, if SOF_TIMESTAMPING_OPT_ID is set, this should
203  always be set too. On datagram sockets the option has no effect.
204
205  A reasonable expectation is that the counter is reset to zero with
206  the system call, so that a subsequent write() of N bytes generates
207  a timestamp with counter N-1. SOF_TIMESTAMPING_OPT_ID_TCP
208  implements this behavior under all conditions.
209
210  SOF_TIMESTAMPING_OPT_ID without modifier often reports the same,
211  especially when the socket option is set when no data is in
212  transmission. If data is being transmitted, it may be off by the
213  length of the output queue (SIOCOUTQ).
214
215  The difference is due to being based on snd_una versus write_seq.
216  snd_una is the offset in the stream acknowledged by the peer. This
217  depends on factors outside of process control, such as network RTT.
218  write_seq is the last byte written by the process. This offset is
219  not affected by external inputs.
220
221  The difference is subtle and unlikely to be noticed when configured
222  at initial socket creation, when no data is queued or sent. But
223  SOF_TIMESTAMPING_OPT_ID_TCP behavior is more robust regardless of
224  when the socket option is set.
225
226SOF_TIMESTAMPING_OPT_CMSG:
227  Support recv() cmsg for all timestamped packets. Control messages
228  are already supported unconditionally on all packets with receive
229  timestamps and on IPv6 packets with transmit timestamp. This option
230  extends them to IPv4 packets with transmit timestamp. One use case
231  is to correlate packets with their egress device, by enabling socket
232  option IP_PKTINFO simultaneously.
233
234
235SOF_TIMESTAMPING_OPT_TSONLY:
236  Applies to transmit timestamps only. Makes the kernel return the
237  timestamp as a cmsg alongside an empty packet, as opposed to
238  alongside the original packet. This reduces the amount of memory
239  charged to the socket's receive budget (SO_RCVBUF) and delivers
240  the timestamp even if sysctl net.core.tstamp_allow_data is 0.
241  This option disables SOF_TIMESTAMPING_OPT_CMSG.
242
243SOF_TIMESTAMPING_OPT_STATS:
244  Optional stats that are obtained along with the transmit timestamps.
245  It must be used together with SOF_TIMESTAMPING_OPT_TSONLY. When the
246  transmit timestamp is available, the stats are available in a
247  separate control message of type SCM_TIMESTAMPING_OPT_STATS, as a
248  list of TLVs (struct nlattr) of types. These stats allow the
249  application to associate various transport layer stats with
250  the transmit timestamps, such as how long a certain block of
251  data was limited by peer's receiver window.
252
253SOF_TIMESTAMPING_OPT_PKTINFO:
254  Enable the SCM_TIMESTAMPING_PKTINFO control message for incoming
255  packets with hardware timestamps. The message contains struct
256  scm_ts_pktinfo, which supplies the index of the real interface which
257  received the packet and its length at layer 2. A valid (non-zero)
258  interface index will be returned only if CONFIG_NET_RX_BUSY_POLL is
259  enabled and the driver is using NAPI. The struct contains also two
260  other fields, but they are reserved and undefined.
261
262SOF_TIMESTAMPING_OPT_TX_SWHW:
263  Request both hardware and software timestamps for outgoing packets
264  when SOF_TIMESTAMPING_TX_HARDWARE and SOF_TIMESTAMPING_TX_SOFTWARE
265  are enabled at the same time. If both timestamps are generated,
266  two separate messages will be looped to the socket's error queue,
267  each containing just one timestamp.
268
269New applications are encouraged to pass SOF_TIMESTAMPING_OPT_ID to
270disambiguate timestamps and SOF_TIMESTAMPING_OPT_TSONLY to operate
271regardless of the setting of sysctl net.core.tstamp_allow_data.
272
273An exception is when a process needs additional cmsg data, for
274instance SOL_IP/IP_PKTINFO to detect the egress network interface.
275Then pass option SOF_TIMESTAMPING_OPT_CMSG. This option depends on
276having access to the contents of the original packet, so cannot be
277combined with SOF_TIMESTAMPING_OPT_TSONLY.
278
279
2801.3.4. Enabling timestamps via control messages
281^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
282
283In addition to socket options, timestamp generation can be requested
284per write via cmsg, only for SOF_TIMESTAMPING_TX_* (see Section 1.3.1).
285Using this feature, applications can sample timestamps per sendmsg()
286without paying the overhead of enabling and disabling timestamps via
287setsockopt::
288
289  struct msghdr *msg;
290  ...
291  cmsg			       = CMSG_FIRSTHDR(msg);
292  cmsg->cmsg_level	       = SOL_SOCKET;
293  cmsg->cmsg_type	       = SO_TIMESTAMPING;
294  cmsg->cmsg_len	       = CMSG_LEN(sizeof(__u32));
295  *((__u32 *) CMSG_DATA(cmsg)) = SOF_TIMESTAMPING_TX_SCHED |
296				 SOF_TIMESTAMPING_TX_SOFTWARE |
297				 SOF_TIMESTAMPING_TX_ACK;
298  err = sendmsg(fd, msg, 0);
299
300The SOF_TIMESTAMPING_TX_* flags set via cmsg will override
301the SOF_TIMESTAMPING_TX_* flags set via setsockopt.
302
303Moreover, applications must still enable timestamp reporting via
304setsockopt to receive timestamps::
305
306  __u32 val = SOF_TIMESTAMPING_SOFTWARE |
307	      SOF_TIMESTAMPING_OPT_ID /* or any other flag */;
308  err = setsockopt(fd, SOL_SOCKET, SO_TIMESTAMPING, &val, sizeof(val));
309
310
3111.4 Bytestream Timestamps
312-------------------------
313
314The SO_TIMESTAMPING interface supports timestamping of bytes in a
315bytestream. Each request is interpreted as a request for when the
316entire contents of the buffer has passed a timestamping point. That
317is, for streams option SOF_TIMESTAMPING_TX_SOFTWARE will record
318when all bytes have reached the device driver, regardless of how
319many packets the data has been converted into.
320
321In general, bytestreams have no natural delimiters and therefore
322correlating a timestamp with data is non-trivial. A range of bytes
323may be split across segments, any segments may be merged (possibly
324coalescing sections of previously segmented buffers associated with
325independent send() calls). Segments can be reordered and the same
326byte range can coexist in multiple segments for protocols that
327implement retransmissions.
328
329It is essential that all timestamps implement the same semantics,
330regardless of these possible transformations, as otherwise they are
331incomparable. Handling "rare" corner cases differently from the
332simple case (a 1:1 mapping from buffer to skb) is insufficient
333because performance debugging often needs to focus on such outliers.
334
335In practice, timestamps can be correlated with segments of a
336bytestream consistently, if both semantics of the timestamp and the
337timing of measurement are chosen correctly. This challenge is no
338different from deciding on a strategy for IP fragmentation. There, the
339definition is that only the first fragment is timestamped. For
340bytestreams, we chose that a timestamp is generated only when all
341bytes have passed a point. SOF_TIMESTAMPING_TX_ACK as defined is easy to
342implement and reason about. An implementation that has to take into
343account SACK would be more complex due to possible transmission holes
344and out of order arrival.
345
346On the host, TCP can also break the simple 1:1 mapping from buffer to
347skbuff as a result of Nagle, cork, autocork, segmentation and GSO. The
348implementation ensures correctness in all cases by tracking the
349individual last byte passed to send(), even if it is no longer the
350last byte after an skbuff extend or merge operation. It stores the
351relevant sequence number in skb_shinfo(skb)->tskey. Because an skbuff
352has only one such field, only one timestamp can be generated.
353
354In rare cases, a timestamp request can be missed if two requests are
355collapsed onto the same skb. A process can detect this situation by
356enabling SOF_TIMESTAMPING_OPT_ID and comparing the byte offset at
357send time with the value returned for each timestamp. It can prevent
358the situation by always flushing the TCP stack in between requests,
359for instance by enabling TCP_NODELAY and disabling TCP_CORK and
360autocork. After linux-4.7, a better way to prevent coalescing is
361to use MSG_EOR flag at sendmsg() time.
362
363These precautions ensure that the timestamp is generated only when all
364bytes have passed a timestamp point, assuming that the network stack
365itself does not reorder the segments. The stack indeed tries to avoid
366reordering. The one exception is under administrator control: it is
367possible to construct a packet scheduler configuration that delays
368segments from the same stream differently. Such a setup would be
369unusual.
370
371
3722 Data Interfaces
373==================
374
375Timestamps are read using the ancillary data feature of recvmsg().
376See `man 3 cmsg` for details of this interface. The socket manual
377page (`man 7 socket`) describes how timestamps generated with
378SO_TIMESTAMP and SO_TIMESTAMPNS records can be retrieved.
379
380
3812.1 SCM_TIMESTAMPING records
382----------------------------
383
384These timestamps are returned in a control message with cmsg_level
385SOL_SOCKET, cmsg_type SCM_TIMESTAMPING, and payload of type
386
387For SO_TIMESTAMPING_OLD::
388
389	struct scm_timestamping {
390		struct timespec ts[3];
391	};
392
393For SO_TIMESTAMPING_NEW::
394
395	struct scm_timestamping64 {
396		struct __kernel_timespec ts[3];
397
398Always use SO_TIMESTAMPING_NEW timestamp to always get timestamp in
399struct scm_timestamping64 format.
400
401SO_TIMESTAMPING_OLD returns incorrect timestamps after the year 2038
402on 32 bit machines.
403
404The structure can return up to three timestamps. This is a legacy
405feature. At least one field is non-zero at any time. Most timestamps
406are passed in ts[0]. Hardware timestamps are passed in ts[2].
407
408ts[1] used to hold hardware timestamps converted to system time.
409Instead, expose the hardware clock device on the NIC directly as
410a HW PTP clock source, to allow time conversion in userspace and
411optionally synchronize system time with a userspace PTP stack such
412as linuxptp. For the PTP clock API, see Documentation/driver-api/ptp.rst.
413
414Note that if the SO_TIMESTAMP or SO_TIMESTAMPNS option is enabled
415together with SO_TIMESTAMPING using SOF_TIMESTAMPING_SOFTWARE, a false
416software timestamp will be generated in the recvmsg() call and passed
417in ts[0] when a real software timestamp is missing. This happens also
418on hardware transmit timestamps.
419
4202.1.1 Transmit timestamps with MSG_ERRQUEUE
421^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
422
423For transmit timestamps the outgoing packet is looped back to the
424socket's error queue with the send timestamp(s) attached. A process
425receives the timestamps by calling recvmsg() with flag MSG_ERRQUEUE
426set and with a msg_control buffer sufficiently large to receive the
427relevant metadata structures. The recvmsg call returns the original
428outgoing data packet with two ancillary messages attached.
429
430A message of cm_level SOL_IP(V6) and cm_type IP(V6)_RECVERR
431embeds a struct sock_extended_err. This defines the error type. For
432timestamps, the ee_errno field is ENOMSG. The other ancillary message
433will have cm_level SOL_SOCKET and cm_type SCM_TIMESTAMPING. This
434embeds the struct scm_timestamping.
435
436
4372.1.1.2 Timestamp types
438~~~~~~~~~~~~~~~~~~~~~~~
439
440The semantics of the three struct timespec are defined by field
441ee_info in the extended error structure. It contains a value of
442type SCM_TSTAMP_* to define the actual timestamp passed in
443scm_timestamping.
444
445The SCM_TSTAMP_* types are 1:1 matches to the SOF_TIMESTAMPING_*
446control fields discussed previously, with one exception. For legacy
447reasons, SCM_TSTAMP_SND is equal to zero and can be set for both
448SOF_TIMESTAMPING_TX_HARDWARE and SOF_TIMESTAMPING_TX_SOFTWARE. It
449is the first if ts[2] is non-zero, the second otherwise, in which
450case the timestamp is stored in ts[0].
451
452
4532.1.1.3 Fragmentation
454~~~~~~~~~~~~~~~~~~~~~
455
456Fragmentation of outgoing datagrams is rare, but is possible, e.g., by
457explicitly disabling PMTU discovery. If an outgoing packet is fragmented,
458then only the first fragment is timestamped and returned to the sending
459socket.
460
461
4622.1.1.4 Packet Payload
463~~~~~~~~~~~~~~~~~~~~~~
464
465The calling application is often not interested in receiving the whole
466packet payload that it passed to the stack originally: the socket
467error queue mechanism is just a method to piggyback the timestamp on.
468In this case, the application can choose to read datagrams with a
469smaller buffer, possibly even of length 0. The payload is truncated
470accordingly. Until the process calls recvmsg() on the error queue,
471however, the full packet is queued, taking up budget from SO_RCVBUF.
472
473
4742.1.1.5 Blocking Read
475~~~~~~~~~~~~~~~~~~~~~
476
477Reading from the error queue is always a non-blocking operation. To
478block waiting on a timestamp, use poll or select. poll() will return
479POLLERR in pollfd.revents if any data is ready on the error queue.
480There is no need to pass this flag in pollfd.events. This flag is
481ignored on request. See also `man 2 poll`.
482
483
4842.1.2 Receive timestamps
485^^^^^^^^^^^^^^^^^^^^^^^^
486
487On reception, there is no reason to read from the socket error queue.
488The SCM_TIMESTAMPING ancillary data is sent along with the packet data
489on a normal recvmsg(). Since this is not a socket error, it is not
490accompanied by a message SOL_IP(V6)/IP(V6)_RECVERROR. In this case,
491the meaning of the three fields in struct scm_timestamping is
492implicitly defined. ts[0] holds a software timestamp if set, ts[1]
493is again deprecated and ts[2] holds a hardware timestamp if set.
494
495
4963. Hardware Timestamping configuration: SIOCSHWTSTAMP and SIOCGHWTSTAMP
497=======================================================================
498
499Hardware time stamping must also be initialized for each device driver
500that is expected to do hardware time stamping. The parameter is defined in
501include/uapi/linux/net_tstamp.h as::
502
503	struct hwtstamp_config {
504		int flags;	/* no flags defined right now, must be zero */
505		int tx_type;	/* HWTSTAMP_TX_* */
506		int rx_filter;	/* HWTSTAMP_FILTER_* */
507	};
508
509Desired behavior is passed into the kernel and to a specific device by
510calling ioctl(SIOCSHWTSTAMP) with a pointer to a struct ifreq whose
511ifr_data points to a struct hwtstamp_config. The tx_type and
512rx_filter are hints to the driver what it is expected to do. If
513the requested fine-grained filtering for incoming packets is not
514supported, the driver may time stamp more than just the requested types
515of packets.
516
517Drivers are free to use a more permissive configuration than the requested
518configuration. It is expected that drivers should only implement directly the
519most generic mode that can be supported. For example if the hardware can
520support HWTSTAMP_FILTER_PTP_V2_EVENT, then it should generally always upscale
521HWTSTAMP_FILTER_PTP_V2_L2_SYNC, and so forth, as HWTSTAMP_FILTER_PTP_V2_EVENT
522is more generic (and more useful to applications).
523
524A driver which supports hardware time stamping shall update the struct
525with the actual, possibly more permissive configuration. If the
526requested packets cannot be time stamped, then nothing should be
527changed and ERANGE shall be returned (in contrast to EINVAL, which
528indicates that SIOCSHWTSTAMP is not supported at all).
529
530Only a processes with admin rights may change the configuration. User
531space is responsible to ensure that multiple processes don't interfere
532with each other and that the settings are reset.
533
534Any process can read the actual configuration by passing this
535structure to ioctl(SIOCGHWTSTAMP) in the same way.  However, this has
536not been implemented in all drivers.
537
538::
539
540    /* possible values for hwtstamp_config->tx_type */
541    enum {
542	    /*
543	    * no outgoing packet will need hardware time stamping;
544	    * should a packet arrive which asks for it, no hardware
545	    * time stamping will be done
546	    */
547	    HWTSTAMP_TX_OFF,
548
549	    /*
550	    * enables hardware time stamping for outgoing packets;
551	    * the sender of the packet decides which are to be
552	    * time stamped by setting SOF_TIMESTAMPING_TX_SOFTWARE
553	    * before sending the packet
554	    */
555	    HWTSTAMP_TX_ON,
556    };
557
558    /* possible values for hwtstamp_config->rx_filter */
559    enum {
560	    /* time stamp no incoming packet at all */
561	    HWTSTAMP_FILTER_NONE,
562
563	    /* time stamp any incoming packet */
564	    HWTSTAMP_FILTER_ALL,
565
566	    /* return value: time stamp all packets requested plus some others */
567	    HWTSTAMP_FILTER_SOME,
568
569	    /* PTP v1, UDP, any kind of event packet */
570	    HWTSTAMP_FILTER_PTP_V1_L4_EVENT,
571
572	    /* for the complete list of values, please check
573	    * the include file include/uapi/linux/net_tstamp.h
574	    */
575    };
576
5773.1 Hardware Timestamping Implementation: Device Drivers
578--------------------------------------------------------
579
580A driver which supports hardware time stamping must support the
581SIOCSHWTSTAMP ioctl and update the supplied struct hwtstamp_config with
582the actual values as described in the section on SIOCSHWTSTAMP.  It
583should also support SIOCGHWTSTAMP.
584
585Time stamps for received packets must be stored in the skb. To get a pointer
586to the shared time stamp structure of the skb call skb_hwtstamps(). Then
587set the time stamps in the structure::
588
589    struct skb_shared_hwtstamps {
590	    /* hardware time stamp transformed into duration
591	    * since arbitrary point in time
592	    */
593	    ktime_t	hwtstamp;
594    };
595
596Time stamps for outgoing packets are to be generated as follows:
597
598- In hard_start_xmit(), check if (skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)
599  is set no-zero. If yes, then the driver is expected to do hardware time
600  stamping.
601- If this is possible for the skb and requested, then declare
602  that the driver is doing the time stamping by setting the flag
603  SKBTX_IN_PROGRESS in skb_shinfo(skb)->tx_flags , e.g. with::
604
605      skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
606
607  You might want to keep a pointer to the associated skb for the next step
608  and not free the skb. A driver not supporting hardware time stamping doesn't
609  do that. A driver must never touch sk_buff::tstamp! It is used to store
610  software generated time stamps by the network subsystem.
611- Driver should call skb_tx_timestamp() as close to passing sk_buff to hardware
612  as possible. skb_tx_timestamp() provides a software time stamp if requested
613  and hardware timestamping is not possible (SKBTX_IN_PROGRESS not set).
614- As soon as the driver has sent the packet and/or obtained a
615  hardware time stamp for it, it passes the time stamp back by
616  calling skb_tstamp_tx() with the original skb, the raw
617  hardware time stamp. skb_tstamp_tx() clones the original skb and
618  adds the timestamps, therefore the original skb has to be freed now.
619  If obtaining the hardware time stamp somehow fails, then the driver
620  should not fall back to software time stamping. The rationale is that
621  this would occur at a later time in the processing pipeline than other
622  software time stamping and therefore could lead to unexpected deltas
623  between time stamps.
624
6253.2 Special considerations for stacked PTP Hardware Clocks
626----------------------------------------------------------
627
628There are situations when there may be more than one PHC (PTP Hardware Clock)
629in the data path of a packet. The kernel has no explicit mechanism to allow the
630user to select which PHC to use for timestamping Ethernet frames. Instead, the
631assumption is that the outermost PHC is always the most preferable, and that
632kernel drivers collaborate towards achieving that goal. Currently there are 3
633cases of stacked PHCs, detailed below:
634
6353.2.1 DSA (Distributed Switch Architecture) switches
636^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
637
638These are Ethernet switches which have one of their ports connected to an
639(otherwise completely unaware) host Ethernet interface, and perform the role of
640a port multiplier with optional forwarding acceleration features.  Each DSA
641switch port is visible to the user as a standalone (virtual) network interface,
642and its network I/O is performed, under the hood, indirectly through the host
643interface (redirecting to the host port on TX, and intercepting frames on RX).
644
645When a DSA switch is attached to a host port, PTP synchronization has to
646suffer, since the switch's variable queuing delay introduces a path delay
647jitter between the host port and its PTP partner. For this reason, some DSA
648switches include a timestamping clock of their own, and have the ability to
649perform network timestamping on their own MAC, such that path delays only
650measure wire and PHY propagation latencies. Timestamping DSA switches are
651supported in Linux and expose the same ABI as any other network interface (save
652for the fact that the DSA interfaces are in fact virtual in terms of network
653I/O, they do have their own PHC).  It is typical, but not mandatory, for all
654interfaces of a DSA switch to share the same PHC.
655
656By design, PTP timestamping with a DSA switch does not need any special
657handling in the driver for the host port it is attached to.  However, when the
658host port also supports PTP timestamping, DSA will take care of intercepting
659the ``.ndo_eth_ioctl`` calls towards the host port, and block attempts to enable
660hardware timestamping on it. This is because the SO_TIMESTAMPING API does not
661allow the delivery of multiple hardware timestamps for the same packet, so
662anybody else except for the DSA switch port must be prevented from doing so.
663
664In the generic layer, DSA provides the following infrastructure for PTP
665timestamping:
666
667- ``.port_txtstamp()``: a hook called prior to the transmission of
668  packets with a hardware TX timestamping request from user space.
669  This is required for two-step timestamping, since the hardware
670  timestamp becomes available after the actual MAC transmission, so the
671  driver must be prepared to correlate the timestamp with the original
672  packet so that it can re-enqueue the packet back into the socket's
673  error queue. To save the packet for when the timestamp becomes
674  available, the driver can call ``skb_clone_sk`` , save the clone pointer
675  in skb->cb and enqueue a tx skb queue. Typically, a switch will have a
676  PTP TX timestamp register (or sometimes a FIFO) where the timestamp
677  becomes available. In case of a FIFO, the hardware might store
678  key-value pairs of PTP sequence ID/message type/domain number and the
679  actual timestamp. To perform the correlation correctly between the
680  packets in a queue waiting for timestamping and the actual timestamps,
681  drivers can use a BPF classifier (``ptp_classify_raw``) to identify
682  the PTP transport type, and ``ptp_parse_header`` to interpret the PTP
683  header fields. There may be an IRQ that is raised upon this
684  timestamp's availability, or the driver might have to poll after
685  invoking ``dev_queue_xmit()`` towards the host interface.
686  One-step TX timestamping do not require packet cloning, since there is
687  no follow-up message required by the PTP protocol (because the
688  TX timestamp is embedded into the packet by the MAC), and therefore
689  user space does not expect the packet annotated with the TX timestamp
690  to be re-enqueued into its socket's error queue.
691
692- ``.port_rxtstamp()``: On RX, the BPF classifier is run by DSA to
693  identify PTP event messages (any other packets, including PTP general
694  messages, are not timestamped). The original (and only) timestampable
695  skb is provided to the driver, for it to annotate it with a timestamp,
696  if that is immediately available, or defer to later. On reception,
697  timestamps might either be available in-band (through metadata in the
698  DSA header, or attached in other ways to the packet), or out-of-band
699  (through another RX timestamping FIFO). Deferral on RX is typically
700  necessary when retrieving the timestamp needs a sleepable context. In
701  that case, it is the responsibility of the DSA driver to call
702  ``netif_rx()`` on the freshly timestamped skb.
703
7043.2.2 Ethernet PHYs
705^^^^^^^^^^^^^^^^^^^
706
707These are devices that typically fulfill a Layer 1 role in the network stack,
708hence they do not have a representation in terms of a network interface as DSA
709switches do. However, PHYs may be able to detect and timestamp PTP packets, for
710performance reasons: timestamps taken as close as possible to the wire have the
711potential to yield a more stable and precise synchronization.
712
713A PHY driver that supports PTP timestamping must create a ``struct
714mii_timestamper`` and add a pointer to it in ``phydev->mii_ts``. The presence
715of this pointer will be checked by the networking stack.
716
717Since PHYs do not have network interface representations, the timestamping and
718ethtool ioctl operations for them need to be mediated by their respective MAC
719driver.  Therefore, as opposed to DSA switches, modifications need to be done
720to each individual MAC driver for PHY timestamping support. This entails:
721
722- Checking, in ``.ndo_eth_ioctl``, whether ``phy_has_hwtstamp(netdev->phydev)``
723  is true or not. If it is, then the MAC driver should not process this request
724  but instead pass it on to the PHY using ``phy_mii_ioctl()``.
725
726- On RX, special intervention may or may not be needed, depending on the
727  function used to deliver skb's up the network stack. In the case of plain
728  ``netif_rx()`` and similar, MAC drivers must check whether
729  ``skb_defer_rx_timestamp(skb)`` is necessary or not - and if it is, don't
730  call ``netif_rx()`` at all.  If ``CONFIG_NETWORK_PHY_TIMESTAMPING`` is
731  enabled, and ``skb->dev->phydev->mii_ts`` exists, its ``.rxtstamp()`` hook
732  will be called now, to determine, using logic very similar to DSA, whether
733  deferral for RX timestamping is necessary.  Again like DSA, it becomes the
734  responsibility of the PHY driver to send the packet up the stack when the
735  timestamp is available.
736
737  For other skb receive functions, such as ``napi_gro_receive`` and
738  ``netif_receive_skb``, the stack automatically checks whether
739  ``skb_defer_rx_timestamp()`` is necessary, so this check is not needed inside
740  the driver.
741
742- On TX, again, special intervention might or might not be needed.  The
743  function that calls the ``mii_ts->txtstamp()`` hook is named
744  ``skb_clone_tx_timestamp()``. This function can either be called directly
745  (case in which explicit MAC driver support is indeed needed), but the
746  function also piggybacks from the ``skb_tx_timestamp()`` call, which many MAC
747  drivers already perform for software timestamping purposes. Therefore, if a
748  MAC supports software timestamping, it does not need to do anything further
749  at this stage.
750
7513.2.3 MII bus snooping devices
752^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
753
754These perform the same role as timestamping Ethernet PHYs, save for the fact
755that they are discrete devices and can therefore be used in conjunction with
756any PHY even if it doesn't support timestamping. In Linux, they are
757discoverable and attachable to a ``struct phy_device`` through Device Tree, and
758for the rest, they use the same mii_ts infrastructure as those. See
759Documentation/devicetree/bindings/ptp/timestamper.txt for more details.
760
7613.2.4 Other caveats for MAC drivers
762^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
763
764Stacked PHCs, especially DSA (but not only) - since that doesn't require any
765modification to MAC drivers, so it is more difficult to ensure correctness of
766all possible code paths - is that they uncover bugs which were impossible to
767trigger before the existence of stacked PTP clocks.  One example has to do with
768this line of code, already presented earlier::
769
770      skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
771
772Any TX timestamping logic, be it a plain MAC driver, a DSA switch driver, a PHY
773driver or a MII bus snooping device driver, should set this flag.
774But a MAC driver that is unaware of PHC stacking might get tripped up by
775somebody other than itself setting this flag, and deliver a duplicate
776timestamp.
777For example, a typical driver design for TX timestamping might be to split the
778transmission part into 2 portions:
779
7801. "TX": checks whether PTP timestamping has been previously enabled through
781   the ``.ndo_eth_ioctl`` ("``priv->hwtstamp_tx_enabled == true``") and the
782   current skb requires a TX timestamp ("``skb_shinfo(skb)->tx_flags &
783   SKBTX_HW_TSTAMP``"). If this is true, it sets the
784   "``skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS``" flag. Note: as
785   described above, in the case of a stacked PHC system, this condition should
786   never trigger, as this MAC is certainly not the outermost PHC. But this is
787   not where the typical issue is.  Transmission proceeds with this packet.
788
7892. "TX confirmation": Transmission has finished. The driver checks whether it
790   is necessary to collect any TX timestamp for it. Here is where the typical
791   issues are: the MAC driver takes a shortcut and only checks whether
792   "``skb_shinfo(skb)->tx_flags & SKBTX_IN_PROGRESS``" was set. With a stacked
793   PHC system, this is incorrect because this MAC driver is not the only entity
794   in the TX data path who could have enabled SKBTX_IN_PROGRESS in the first
795   place.
796
797The correct solution for this problem is for MAC drivers to have a compound
798check in their "TX confirmation" portion, not only for
799"``skb_shinfo(skb)->tx_flags & SKBTX_IN_PROGRESS``", but also for
800"``priv->hwtstamp_tx_enabled == true``". Because the rest of the system ensures
801that PTP timestamping is not enabled for anything other than the outermost PHC,
802this enhanced check will avoid delivering a duplicated TX timestamp to user
803space.
804