1ADJTIMEX(8) System Manager's Manual ADJTIMEX(8)
2
6 adjtimex - display or set the kernel time variables
7
9 adjtimex option]...
10
12 This program gives you raw access to the kernel time variables. Anyone
13 may print out the time variables, but only the superuser may change
14 them.
15 Your computer has two clocks - the "hardware clock" that runs all the
17 time, and the system clock that runs only while the computer is on.
18 Normally, "hwclock --hctosys" should be run at startup to initialize
19 the system clock. The system clock has much better precision (approxi‐
20 mately 1 usec), but the hardware clock probably has better long-term
21 stability. There are three basic strategies for managing these clocks.
22 For a machine connected to the Internet, or equipped with a precision
24 oscillator or radio clock, the best way is to regulate the system clock
25 with ntpd(8). The kernel will automatically update the hardware clock
26 every eleven minutes.
27 In addition, hwclock(8) can be used to approximately correct for a con‐
29 stant drift in the hardware clock. In this case, "hwclock --adjust" is
30 run occasionally. hwclock notes how long it has been since the last
31 adjustment, and nudges the hardware clock forward or back by the appro‐
32 priate amount. The user needs to set the time with "hwclock --set"
33 several times over the course of a few days so hwclock can estimate the
34 drift rate. During that time, ntpd should not be running, or else
35 hwclock will conclude the hardware clock does not drift at all. After
36 you have run "hwclock --set" for the last time, it's okay to start
37 ntpd. Then, "hwclock --systohc" should be run when the machine is shut
38 down. (To see why, suppose the machine runs for a week with ntpd, is
39 shut down for a day, is restarted, and "hwclock --adjust" is run by a
40 startup script. It should only correct for one day's worth of drift.
41 However, it has no way of knowing that ntpd has been adjusting the
42 hardware clock, so it bases its adjustment on the last time hwclock was
43 run.)
44 For a standalone or intermittently connected machine, where it's not
46 possible to run ntpd, you may use adjtimex instead to correct the sys‐
47 tem clock for systematic drift.
48 There are several ways to estimate the drift rate. If your computer
50 can be connected to the net, you might run ntpd for at least several
51 hours and run "adjtimex --print" to learn what values of tick and freq
52 it settled on. Alternately, you could estimate values using as a ref‐
53 erence the CMOS clock (see the --compare and --adjust switches),
54 another host (see --host and --review), or some other source of time
55 (see --watch and --review). You could then add a line to rc.local
56 invoking adjtimex, or configure /etc/init.d/adjtimex or
57 /etc/default/adjtimex, to set those parameters each time you reboot.
58
60 Options may be introduced by either - or --, and unique abbreviations
61 may be used. Here is a summary of the options, grouped by type.
62 Explanations follow.
63 Get/Set Kernel Time Parameters
65 -p --print -t --tick val -f newfreq --frequency newfreq -o val
66 --offset val -s adjustment --singleshot adjustment -S status
67 --status status -m val -R --reset --maxerror val -e val
68 --esterror val -T val --timeconstant val -a[count]
69 --adjust[=count]
70 Estimate Systematic Drifts
72 -c[count] --compare[=count] -i tim --interval tim -l file
73 --logfile file -h timeserver --host timeserver -w --watch
74 -r[file] --review[=file] -u --utc
75 Informative Output
77 --help -v --version -V --verbose
78 -p, --print
80 Print the current values of the kernel time variables. NOTE:
81 The time is "raw", and may be off by up to one timer tick (10
82 msec). "status" gives the value of the time_status variable in
83 the kernel. For Linux 1.0 and 1.2 kernels, the value is as
84 follows:
85 0 clock is synchronized (so the kernel should
86 periodically set the CMOS clock to match the
87 system clock)
88 1 inserting a leap second at midnight
89 2 deleting a leap second at midnight
90 3 leap second in progress
91 4 leap second has occurred
92 5 clock not externally synchronized (so the
93 kernel should leave the CMOS clock alone)
94 For Linux 2.0 kernels, the value is a sum of these:
95 1 PLL updates enabled
96 2 PPS freq discipline enabled
97 4 PPS time discipline enabled
98 8 frequency-lock mode enabled
99 16 inserting leap second
100 32 deleting leap second
101 64 clock unsynchronized
102 128 holding frequency
103 256 PPS signal present
104 512 PPS signal jitter exceeded
105 1024 PPS signal wander exceeded
106 2048 PPS signal calibration error
107 4096 clock hardware fault
108 -t val, --tick val
110 Set the number of microseconds that should be added to the
111 system time for each kernel tick interrupt. For a kernel with
112 USER_HZ=100, there are supposed to be 100 ticks per second, so
113 val should be close to 10000. Increasing val by 1 speeds up the
114 system clock by about 100 ppm, or 8.64 sec/day. tick must be in
115 the range 900000/USER_HZ...1100000/USER_HZ. If val is rejected
116 by the kernel, adjtimex will determine the acceptable range
117 through trial and error and print it. (After completing the
118 search, it will restore the original value.)
119 -f newfreq, --frequency newfreq
121 Set the system clock frequency offset to newfreq. newfreq can
122 be negative or positive, and gives a much finer adjustment than
123 the --tick switch. When USER_HZ=100, the value is scaled such
124 that newfreq = 65536 speeds up the system clock by about 1 ppm,
125 or .0864 sec/day. Thus, all of these are about the same:
126 --tick 9995 --frequency 32768000
127 --tick 10000 --frequency 6553600
128 --tick 10001 --frequency 0
129 --tick 10002 --frequency -6553600
130 --tick 10005 --frequency -32768000
131 To see the acceptable range for newfreq, use --print and look at
132 "tolerance", or try an illegal value (e.g. --tick 0).
133 -s adj, --singleshot adj
135 Slew the system clock by adj usec. (Its rate is changed
136 temporarily by about 1 part in 2000.)
137 -o adj, --offset adj
139 Add a time offset of adj usec. The kernel code adjusts the time
140 gradually by adj, notes how long it has been since the last time
141 offset, and then adjusts the frequency offset to correct for the
142 apparent drift. adj must be in the range -512000...512000.
143 -S status, --status status
145 Set kernel system clock status register to value status. Look
146 here above at the --print switch section for the meaning of
147 status, depending on your kernel.
148 -R, --reset
150 Reset clock status after setting a clock parameter. For early
151 Linux kernels, using the adjtimex(2) system call to set any time
152 parameter the kernel think the clock is synchronized with an
153 external time source, so it sets the kernel variable time_status
154 to TIME_OK. Thereafter, at 11 minute intervals, it will adjust
155 the CMOS clock to match. We prevent this "eleven minute mode"
156 by setting the clock, because that has the side effect of
157 resetting time_status to TIME_BAD. We try not to actually
158 change the clock setting. Kernel versions 2.0.40 and later
159 apparently don't need this. If your kernel does require it, use
160 this option with: -t -T -t -e -m -f -s -o -c -r.
161 -m val, --maxerror val
163 Set maximum error (usec).
164 -e val, --esterror val
166 Set estimated error (usec). The maximum and estimated error are
167 not used by the kernel. They are merely made available to user
168 processes via the adjtimex(2) system call.
169 -T val, --timeconstant val
171 Set phase locked loop (PLL) time constant. val determines the
172 bandwidth or "stiffness" of the PLL. The effective PLL time
173 constant will be a multiple of (2^val). For room-temperature
174 quartz oscillators, David Mills recommends the value 2, which
175 corresponds to a PLL time constant of about 900 sec and a
176 maximum update interval of about 64 sec. The maximum update
177 interval scales directly with the time constant, so that at the
178 maximum time constant of 6, the update interval can be as large
179 as 1024 sec.
180 Values of val between zero and 2 give quick convergence; values
182 between 2 and 6 can be used to reduce network load, but at a
183 modest cost in accuracy.
184 -ccount], --compare=count
186 Periodically compare the system clock with the CMOS clock.
187 After the first two calls, print values for tick and frequency
188 offset that would bring the system clock into approximate
189 agreement with the CMOS clock. CMOS clock readings are adjusted
190 for systematic drift using using the correction in /etc/adjtime
191 — see hwclock(8). The interval between comparisons is 10
192 seconds, unless changed by the --interval switch. The optional
193 argument is the number of comparisons. (If the argument is
194 supplied, the "=" is required.) If the CMOS clock and the
195 system clock differ by more than six minutes, adjtimex will try
196 shifting the time from the CMOS clock by some multiple of one
197 hour, up to plus or minus 13 hours in all. This should allow
198 correct operation, including logging, if the --utc switch was
199 used when the CMOS clock is set to local time (or vice-versa),
200 or if summer time has started or stopped since the CMOS clock
201 was last set.
202 -acount], --adjust[=count]
204 By itself, same as --compare, except the recommended values are
205 actually installed after every third comparison. With --review,
206 the tick and frequency are set to the least-squares estimates.
207 (In the latter case, any count value is ignored.)
208 -i tim, --interval tim
210 Set the interval in seconds between clock comparisons for the
211 --compare and --adjust options.
212 -u, --utc
214 The CMOS clock is set to UTC (universal time) rather than local
215 time.
216 -lfile], --log=file
218 Save the current values of the system and CMOS clocks, and
219 optionally a reference time, to file (default
220 /var/log/clocks.log). The reference time is taken from a
221 network timeserver (see the --host switch) or supplied by the
222 user (see the --watch switch).
223 -h timeserver, --host timeserver
225 Use ntpdate to query the given timeserver for the current time.
226 This will fail if timeserver is not running a Network Time
227 Protocol (NTP) server, or if that server is not synchronized.
228 Implies --log.
229 -w, --watch
231 Ask for a keypress when the user knows the time, then ask what
232 that time was, and its approximate accuracy. Implies --log.
233 -rfile], --review=file
235 Review the clock log file (default /var/log/clocks.log) and
236 estimate, if possible, the rates of the CMOS and system clocks.
237 Calculate least-squares rates using all suitable log entries.
238 Suggest corrections to adjust for systematic drift. With
239 --adjust, the frequency and tick are set to the suggested
240 values. (The CMOS clock correction is not changed.)
241 -V, --verbose
243 Increase verbosity.
244 --help Print the program options.
246 -v, --version
248 Print the program version.
249
251 If your system clock gained 8 seconds in 24 hours, you could set the
252 tick to 9999, and then it would lose 0.64 seconds a day (that is, 1
253 tick unit = 8.64 seconds per day). To correct the rest of the error,
254 you could set the frequency offset to (2^16)*0.64/.0864 = 485452.
255 Thus, putting the following in rc.local would approximately correct the
256 system clock:
257 adjtimex --tick 9999 --freq 485452
259
261 adjtimex adjusts only the system clock — the one that runs while the
262 computer is powered up. To set or regulate the CMOS clock, see
263 hwclock(8).
264
266 Steven S. Dick <ssd at nevets.oau.org>, Jim Van Zandt <jrv at
267 comcast.net>.
268
270 date(1L), gettimeofday(2), settimeofday(2), hwclock(8), ntpdate(8),
271 ntpd(8), /usr/src/linux/include/linux/timex.h,
272 /usr/src/linux/include/linux/sched.h, /usr/src/linux/kernel/time.c,
273 /usr/src/linux/kernel/sched.c
274 May 23, 2006 ADJTIMEX(8)