5 (C) Copyright 2007 Rodolfo Giometti <giometti@enneenne.com>
7 This program is free software; you can redistribute it and/or modify
8 it under the terms of the GNU General Public License as published by
9 the Free Software Foundation; either version 2 of the License, or
10 (at your option) any later version.
12 This program is distributed in the hope that it will be useful,
13 but WITHOUT ANY WARRANTY; without even the implied warranty of
14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 GNU General Public License for more details.
22 LinuxPPS provides a programming interface (API) to define in the
23 system several PPS sources.
25 PPS means "pulse per second" and a PPS source is just a device which
26 provides a high precision signal each second so that an application
27 can use it to adjust system clock time.
29 A PPS source can be connected to a serial port (usually to the Data
30 Carrier Detect pin) or to a parallel port (ACK-pin) or to a special
31 CPU's GPIOs (this is the common case in embedded systems) but in each
32 case when a new pulse arrives the system must apply to it a timestamp
33 and record it for userland.
35 Common use is the combination of the NTPD as userland program, with a
36 GPS receiver as PPS source, to obtain a wallclock-time with
37 sub-millisecond synchronisation to UTC.
43 While implementing a PPS API as RFC 2783 defines and using an embedded
44 CPU GPIO-Pin as physical link to the signal, I encountered a deeper
47 At startup it needs a file descriptor as argument for the function
50 This implies that the source has a /dev/... entry. This assumption is
51 OK for the serial and parallel port, where you can do something
52 useful besides(!) the gathering of timestamps as it is the central
53 task for a PPS API. But this assumption does not work for a single
54 purpose GPIO line. In this case even basic file-related functionality
55 (like read() and write()) makes no sense at all and should not be a
56 precondition for the use of a PPS API.
58 The problem can be simply solved if you consider that a PPS source is
59 not always connected with a GPS data source.
61 So your programs should check if the GPS data source (the serial port
62 for instance) is a PPS source too, and if not they should provide the
63 possibility to open another device as PPS source.
65 In LinuxPPS the PPS sources are simply char devices usually mapped
66 into files /dev/pps0, /dev/pps1, etc.
69 PPS with USB to serial devices
70 ------------------------------
72 It is possible to grab the PPS from an USB to serial device. However,
73 you should take into account the latencies and jitter introduced by
74 the USB stack. Users have reported clock instability around +-1ms when
75 synchronized with PPS through USB. With USB 2.0, jitter may decrease
76 down to the order of 125 microseconds.
78 This may be suitable for time server synchronization with NTP because
79 of its undersampling and algorithms.
81 If your device doesn't report PPS, you can check that the feature is
82 supported by its driver. Most of the time, you only need to add a call
83 to usb_serial_handle_dcd_change after checking the DCD status (see
84 ch341 and pl2303 examples).
90 To register a PPS source into the kernel you should define a struct
91 pps_source_info as follows:
93 static struct pps_source_info pps_ktimer_info = {
96 .mode = PPS_CAPTUREASSERT | PPS_OFFSETASSERT |
98 PPS_CANWAIT | PPS_TSFMT_TSPEC,
99 .echo = pps_ktimer_echo,
100 .owner = THIS_MODULE,
103 and then calling the function pps_register_source() in your
104 initialization routine as follows:
106 source = pps_register_source(&pps_ktimer_info,
107 PPS_CAPTUREASSERT | PPS_OFFSETASSERT);
109 The pps_register_source() prototype is:
111 int pps_register_source(struct pps_source_info *info, int default_params)
113 where "info" is a pointer to a structure that describes a particular
114 PPS source, "default_params" tells the system what the initial default
115 parameters for the device should be (it is obvious that these parameters
116 must be a subset of ones defined in the struct
117 pps_source_info which describe the capabilities of the driver).
119 Once you have registered a new PPS source into the system you can
120 signal an assert event (for example in the interrupt handler routine)
123 pps_event(source, &ts, PPS_CAPTUREASSERT, ptr)
125 where "ts" is the event's timestamp.
127 The same function may also run the defined echo function
128 (pps_ktimer_echo(), passing to it the "ptr" pointer) if the user
129 asked for that... etc..
131 Please see the file drivers/pps/clients/pps-ktimer.c for example code.
137 If the SYSFS filesystem is enabled in the kernel it provides a new class:
142 Every directory is the ID of a PPS sources defined in the system and
143 inside you find several files:
145 $ ls -F /sys/class/pps/pps0/
146 assert dev mode path subsystem@
147 clear echo name power/ uevent
150 Inside each "assert" and "clear" file you can find the timestamp and a
153 $ cat /sys/class/pps/pps0/assert
154 1170026870.983207967#8
156 Where before the "#" is the timestamp in seconds; after it is the
157 sequence number. Other files are:
159 * echo: reports if the PPS source has an echo function or not;
161 * mode: reports available PPS functioning modes;
163 * name: reports the PPS source's name;
165 * path: reports the PPS source's device path, that is the device the
166 PPS source is connected to (if it exists).
169 Testing the PPS support
170 -----------------------
172 In order to test the PPS support even without specific hardware you can use
173 the pps-ktimer driver (see the client subsection in the PPS configuration menu)
174 and the userland tools available in your distribution's pps-tools package,
175 http://linuxpps.org , or https://github.com/redlab-i/pps-tools.
177 Once you have enabled the compilation of pps-ktimer just modprobe it (if
178 not statically compiled):
180 # modprobe pps-ktimer
182 and the run ppstest as follow:
184 $ ./ppstest /dev/pps1
185 trying PPS source "/dev/pps1"
186 found PPS source "/dev/pps1"
187 ok, found 1 source(s), now start fetching data...
188 source 0 - assert 1186592699.388832443, sequence: 364 - clear 0.000000000, sequence: 0
189 source 0 - assert 1186592700.388931295, sequence: 365 - clear 0.000000000, sequence: 0
190 source 0 - assert 1186592701.389032765, sequence: 366 - clear 0.000000000, sequence: 0
192 Please note that to compile userland programs, you need the file timepps.h.
193 This is available in the pps-tools repository mentioned above.
199 Sometimes one needs to be able not only to catch PPS signals but to produce
200 them also. For example, running a distributed simulation, which requires
201 computers' clock to be synchronized very tightly. One way to do this is to
202 invent some complicated hardware solutions but it may be neither necessary
203 nor affordable. The cheap way is to load a PPS generator on one of the
204 computers (master) and PPS clients on others (slaves), and use very simple
205 cables to deliver signals using parallel ports, for example.
207 Parallel port cable pinout:
208 pin name master slave
226 18-25 GND *-----------*
228 Please note that parallel port interrupt occurs only on high->low transition,
229 so it is used for PPS assert edge. PPS clear edge can be determined only
230 using polling in the interrupt handler which actually can be done way more
231 precisely because interrupt handling delays can be quite big and random. So
232 current parport PPS generator implementation (pps_gen_parport module) is
233 geared towards using the clear edge for time synchronization.
235 Clear edge polling is done with disabled interrupts so it's better to select
236 delay between assert and clear edge as small as possible to reduce system
237 latencies. But if it is too small slave won't be able to capture clear edge
238 transition. The default of 30us should be good enough in most situations.
239 The delay can be selected using 'delay' pps_gen_parport module parameter.