sys-utils/hwclock.8.in

.\" hwclock.8.in -- man page for util-linux' hwclock
.\"
.\" 2015-01-07 J William Piggott
.\" Authored new section: DATE-TIME CONFIGURATION.
.\" Subsections: Keeping Time..., LOCAL vs UTC, POSIX vs 'RIGHT'.
.\" Errata and drop outdated language.
.\" Updates for v2.26
.\"
.\"
.\" -------------------------------
.\" Included for troff portability.
.\" Continuation line for .TP header.
.de TQ
.  br
.  ns
.  TP \\$1\" no doublequotes around argument!
..
.
.\" -------------------------------
.
.TH HWCLOCK 8 "January 2015" "util-linux" "System Administration"
.SH NAME
hwclock \- read or set the hardware clock (RTC)
.SH SYNOPSIS
.B hwclock
.RI [ function ]
.RI [ option ...]
.
.SH DESCRIPTION
.B hwclock
is a tool for accessing the Hardware Clock.  It can: display the
Hardware Clock time; set the Hardware Clock to a specified time; set the
Hardware Clock from the System Clock; set the System Clock from the
Hardware Clock; compensate for Hardware Clock drift; correct the System
Clock timescale; set the kernel's timezone, NTP timescale, and epoch
(Alpha only); compare the System and Hardware Clocks; and predict future
Hardware Clock values based on its drift rate.
.PP
Since v2.26 important changes were made to the
.B \-\-hctosys
function, the
.B \-\-directisa
option, and a new option
.B \-\-update\-drift
was added. See their respective sections below.
.
.SH FUNCTIONS
The following functions are mutually exclusive, only one can be given at
a time. If none are given the default is
.BR \-\-show .
.TP
.B \-\-adjust
Add or subtract time from the Hardware Clock to account for systematic
drift since the last time the clock was set or adjusted.  See the
discussion below, under
.BR "The Adjust Function" .
.
.TP
.BR \-c , \ \-\-compare
Periodically compare the Hardware Clock to the System Time and output
the difference every 10 seconds.  This will also print the frequency
offset and tick.
.
.TP
.B \-\-getepoch
.TQ
.B \-\-setepoch
These functions are for Alpha machines only.
.sp
Read and set the kernel's Hardware Clock epoch value.
Epoch is the number of years into AD to which a zero year value in the
Hardware Clock refers.  For example, if you are using the convention
that the year counter in your Hardware Clock contains the number of
full years since 1952, then the kernel's Hardware Clock epoch value
must be 1952.
.sp
The set function requires using the
.B \%\-\-epoch
option.
.sp
This epoch value is used whenever
.B \%hwclock
reads or sets the Hardware Clock.
.
.TP
.B \-\-predict
Predict what the Hardware Clock will read in the future based upon the
time given by the
.B \-\-date
option and the information in
.IR @ADJTIME_PATH@ .
This is useful, for example, to account for drift when setting a
Hardware Clock wakeup (aka alarm). See
.BR \%rtcwake (8).
.sp
Do not use this function if the Hardware Clock is being modified by
anything other than the current operating system's
.B \%hwclock
command, such as \%'11\ minute\ mode' or from dual-booting another OS.
.
.TP
.BR \-r , \ \-\-show
.TQ
.B \-\-get
.br
Read the Hardware Clock and print the time on standard output.
The time shown is always in local time, even if you keep your Hardware Clock
in UTC.  See the
.B \%\-\-localtime
option.
.sp
Showing the Hardware Clock time is the default when no function is specified.
.sp
The
.B \-\-get
function also applies drift correction to the time read, based upon the
information in
.IR @ADJTIME_PATH@ .
Do not use this function if the Hardware Clock is being modified by
anything other than the current operating system's
.B \%hwclock
command, such as \%'11\ minute\ mode' or from dual-booting another OS.
.
.TP
.BR \-s , \ \-\-hctosys
Set the System Clock from the Hardware Clock.  The time read from the Hardware
Clock is compensated to account for systematic drift before using it to set the
System Clock.  See the discussion below, under
.BR "The Adjust Function" .
.sp
The System Clock must be kept in the UTC timescale for date-time
applications to work correctly in conjunction with the timezone configured
for the system.  If the Hardware Clock is kept in local time then the time read
from it must be shifted to the UTC timescale before using it to set the System
Clock.  The
.B \%\-\-hctosys
function does this based upon the information in the
.I @ADJTIME_PATH@
file or the command line arguments
.BR \%\-\-localtime " and " \-\-utc .
Note: no daylight saving adjustment is made. See the discussion below under
.BR "LOCAL vs UTC" .
.sp
The kernel also keeps a timezone value, the
.B \%\-\-hctosys
function sets it to the timezone configured for the system.  The system
timezone is configured by the TZ environment variable or the
.I \%/etc/localtime
file, as
.BR \%tzset (3)
would interpret them.
The obsolete tz_dsttime field of the kernel's timezone value is set
to zero.  (For details on what this field used to mean, see
.BR \%settimeofday (2).)
.sp
When used in a startup script, making the
.B \%\-\-hctosys
function the first caller of
.BR \%settimeofday (2)
from boot, it will set the NTP \%'11\ minute\ mode' timescale via the
.I \%persistent_clock_is_local
kernel variable.  If the Hardware Clock's timescale configuration is
changed then a reboot is required to inform the kernel.  See the
discussion below, under
.BR "Automatic Hardware Clock Synchronization by the Kernel" .
.sp
This is a good function to use in one of the system startup scripts before the
file systems are mounted read/write.
.sp
This function should never be used on a running system. Jumping system time
will cause problems, such as, corrupted filesystem timestamps.  Also, if
something has changed the Hardware Clock, like NTP's \%'11\ minute\ mode', then
.B \%\-\-hctosys
will set the time incorrectly by including drift compensation.
.sp
Drift compensation can be inhibited by setting the drift factor in
.I @ADJTIME_PATH@
to zero.  This setting will be persistent as long as the
.BR \%\-\-update\-drift " option is not used with " \%\-\-systohc
at shutdown (or anywhere else).  Another way to inhibit this is by using the
.BR \%\-\-noadjfile " option when calling the " \%\-\-hctosys
function.  A third method is to delete the
.IR @ADJTIME_PATH@ " file."
.B Hwclock
will then default to using the UTC timescale for the Hardware Clock.  If
the Hardware Clock is ticking local time it will need to be defined in
the file.  This can be done by calling
.BR hwclock\ \-\-localtime\ \-\-adjust ;
when the file is not present this command will not actually
adjust the Clock, but it will create the file with local time
configured, and a drift factor of zero.
.sp
A condition under which inhibiting
.BR hwclock 's
drift correction may be desired is when dual-booting multiple operating
systems.  If while this instance of Linux is stopped, another OS changes
the Hardware Clock's value, then when this instance is started again the
drift correction applied will be incorrect.
.sp
.RB "For " hwclock 's
drift correction to work properly it is imperative that nothing changes
the Hardware Clock while its Linux instance is not running.
.
.TP
.B \-\-set
Set the Hardware Clock to the time given by the
.BR \-\-date
option, and update the timestamps in
.IR @ADJTIME_PATH@ .
With the
.B --update-drift
option (re)calculate the drift factor.
.
.TP
.B \-\-systz
This is an alternate to the
.B \%\-\-hctosys
function that does not read the Hardware Clock nor set the System Clock;
consequently there is not any drift correction.  It is intended to be
used in a startup script on systems with kernels above version 2.6 where
you know the System Clock has been set from the Hardware Clock by the
kernel during boot.
.sp
It does the following things that are detailed above in the
.BR \%\-\-hctosys " function:"
.RS
.IP \(bu 2
Corrects the System Clock timescale to UTC as needed.  Only instead of
accomplishing this by setting the System Clock,
.B hwclock
simply informs the kernel and it handles the change.
.IP \(bu 2
Sets the kernel's NTP \%'11\ minute\ mode' timescale.
.IP \(bu 2
Sets the kernel's timezone.
.PP
The first two are only available on the first call of
.BR \%settimeofday (2)
after boot.  Consequently this option only makes sense when used in a
startup script.  If the Hardware Clocks timescale configuration is
changed then a reboot would be required to inform the kernel.
.RE
.
.TP
.BR \-w , \ \-\-systohc
Set the Hardware Clock from the System Clock, and update the timestamps in
.IR @ADJTIME_PATH@ .
With the
.B --update-drift
option (re)calculate the drift factor.
.
.TP
.BR \-V , \ \-\-version
Display version information and exit.
.
.TP
.BR \-h , \ \-\-help
Display help text and exit.
.
.SH OPTIONS
.
.TP
.BI \-\-adjfile= filename
.RI "Override the default " @ADJTIME_PATH@ " file path."
.
.TP
.B \-\-badyear
Indicate that the Hardware Clock is incapable of storing years outside
the range 1994-1999.  There is a problem in some BIOSes (almost all
Award BIOSes made between 4/26/94 and 5/31/95) wherein they are unable
to deal with years after 1999.  If one attempts to set the year-of-century
value to something less than 94 (or 95 in some cases), the value that
actually gets set is 94 (or 95).  Thus, if you have one of these machines,
.B \%hwclock
cannot set the year after 1999 and cannot use the value of the clock as
the true time in the normal way.
.sp
To compensate for this (without your getting a BIOS update, which would
definitely be preferable), always use
.B \%\-\-badyear
if you have one of these machines.  When
.B \%hwclock
knows it's working with a brain-damaged clock, it ignores the year part of
the Hardware Clock value and instead tries to guess the year based on the
last calibrated date in the adjtime file, by assuming that date is
within the past year.  For this to work, you had better do a
.B \%hwclock\ \-\-set
or
.B \%hwclock\ \-\-systohc
at least once a year!
.sp
Though
.B \%hwclock
ignores the year value when it reads the Hardware Clock, it sets the
year value when it sets the clock.  It sets it to 1995, 1996, 1997, or
1998, whichever one has the same position in the leap year cycle as
the true year.  That way, the Hardware Clock inserts leap days where
they belong.  Again, if you let the Hardware Clock run for more than a
year without setting it, this scheme could be defeated and you could
end up losing a day.
.sp
.B \%hwclock
warns you that you probably need
.B \%\-\-badyear
whenever it finds your Hardware Clock set to 1994 or 1995.
.
.TP
.BI \%\-\-date= date_string
You need this option if you specify the
.B \-\-set
or
.B \%\-\-predict
functions, otherwise it is ignored.
It specifies the time to which to set the Hardware Clock, or the
time for which to predict the Hardware Clock reading.
The value of this option is used as an argument to the
.BR date "(1) program's " \-\-date
option.  For example:
.RS
.IP "" 4
.BI "\%hwclock\ \-\-set\ \-\-date='" 2011-08-14\ 16:45:05 '
.PP
The argument must be in local time, even if you keep your Hardware Clock in
UTC.  See the
.B \%\-\-localtime
option.  The argument must not be a relative time like "+5 minutes", because
.BR \%hwclock 's
precision depends upon correlation between the argument's value and when
the enter key is pressed.
.RE
.
.TP
.B \-\-debug
Display a lot of information about what
.B \%hwclock
is doing internally.  Some of its functions are complex and this output
can help you understand how the program works.
.
.TP
.B \-\-directisa
This option is meaningful for: ISA compatible machines including x86, and
x86_64; and Alpha (which has a similar Hardware Clock interface).  For other
machines, it has no effect.  This option tells
.B \%hwclock
to use explicit I/O instructions to access the Hardware Clock.
Without this option,
.B \%hwclock
will use the rtc device, which it assumes to be driven by the RTC device
driver.  As of v2.26 it will no longer automatically use directisa when
the rtc driver is unavailable; this was causing an unsafe condition that
could allow two processes to access the Hardware Clock at the same time.
Direct hardware access from userspace should only be used for testing,
troubleshooting, and as a last resort when all other methods fail.  See
the
.BR \-\-rtc " option."
.
.TP
.BR \-f , \ \-\-rtc=\fIfilename\fR
.RB "Override " \%hwclock 's
default rtc device file name.  Otherwise it will
use the first one found in this order:
.in +4
.br
.I /dev/rtc
.br
.I /dev/rtc0
.br
.I /dev/misc/rtc
.br
.in
.RB "For " IA-64:
.in +4
.br
.I /dev/efirtc
.br
.I /dev/misc/efirtc
.in
.
.TP
.B \-\-localtime
.TQ
.BR \-u ", " \-\-utc
Indicate which timescale the Hardware Clock is set to.
.sp
The Hardware Clock may be configured to use either the UTC or the local
timescale, but nothing in the clock itself says which alternative is
being used.  The
.BR \%\-\-localtime " or " \-\-utc
options give this information to the
.B \%hwclock
command.  If you specify the wrong one (or specify neither and take a
wrong default), both setting and reading the Hardware Clock will be
incorrect.
.sp
If you specify neither
.BR \-\-utc " nor " \%\-\-localtime
then the one last given with a set function
.RB ( \-\-set ", " \%\-\-systohc ", or " \%\-\-adjust ),
as recorded in
.IR @ADJTIME_PATH@ ,
will be used.  If the adjtime file doesn't exist, the default is UTC.
.sp
Note: daylight saving time changes may be inconsistent when the
Hardware Clock is kept in local time. See the discussion below under
.BR "LOCAL vs UTC" .
.
.TP
.B \-\-noadjfile
Disable the facilities provided by
.IR @ADJTIME_PATH@ .
.B \%hwclock
will not read nor write to that file with this option.  Either
.BR \-\-utc " or " \%\-\-localtime
must be specified when using this option.
.
.TP
.B \-\-test
Do not actually change anything on the system, i.e., the Clocks or
adjtime file.  This is useful, especially in conjunction with
.BR \%\-\-debug ,
in learning about the internal operations of hwclock.
.
.TP
.B \-\-update\-drift
Update the Hardware Clock's drift factor in
.IR @ADJTIME_PATH@ .
It is used with
.BR \-\-set " or " \%\-\-systohc ,
otherwise it is ignored.
.sp
A minimum four hour period between settings is required.  This is to
avoid invalid calculations.  The longer the period, the more precise the
resulting drift factor will be.
.sp
This option was added in v2.26, because
it is typical for systems to call
.B \%hwclock\ \-\-systohc
at shutdown; with the old behaviour this would automatically
(re)calculate the drift factor which caused several problems:
.RS
.IP \(bu 2
When using ntpd with an \%'11\ minute\ mode' kernel the drift factor
would be clobbered to near zero.
.IP \(bu 2
It would not allow the use of 'cold' drift correction.  With most
configurations using 'cold' drift will yield favorable results.  Cold,
means when the machine is turned off which can have a significant impact
on the drift factor.
.IP \(bu 2
(Re)calculating drift factor on every shutdown delivers suboptimal
results.  For example, if ephemeral conditions cause the machine to be
abnormally hot the drift factor calculation would be out of range.
.PP
.RB "Having " \%hwclock
calculate the drift factor is a good starting point, but for optimal
results it will likely need to be adjusted by directly editing the
.I @ADJTIME_PATH@
file.  For most configurations once a machine's optimal drift factor is
crafted it should not need to be changed.  Therefore, the old behavior to
automatically (re)calculate drift was changed and now requires this
option to be used.  See the discussion below, under
.BR "The Adjust Function" .
.RE
.
.SH OPTIONS FOR ALPHA MACHINES ONLY
.
.TP
.B \-\-arc
This option is equivalent to
.BI \%\-\-epoch= 1980
and is used to specify the most common epoch on Alphas
with ARC console (but Ruffians have an epoch of 1900).
.
.TP
.BI \-\-epoch= year
Specifies the year which is the beginning of the Hardware Clock's epoch,
that is the number of years into AD to which a zero value in the
Hardware Clock's year counter refers.  It is used together with the
.B \%\-\-setepoch
option to set the kernel's idea of the epoch of the Hardware Clock.
.sp
For example, on a Digital Unix machine:
.RS
.IP "" 4
.BI hwclock\ \-\-setepoch\ \-\-epoch= 1952
.RE
.
.TP
.B \-\-funky\-toy
.TQ
.B \-\-jensen
These two options specify what kind of Alpha machine you have.  They
are invalid if you do not have an Alpha and are usually unnecessary
if you do;
.B \%hwclock
should be able to determine what it is running on when
.I \%/proc
is mounted.
.sp
.RB "The " \%\-\-jensen
option is used for Jensen models;
.B \%\-\-funky\-toy
means that the machine requires the UF bit instead of the UIP bit in
the Hardware Clock to detect a time transition.  "Toy" in the option
name refers to the Time Of Year facility of the machine.
.
.TP
.B \-\-srm
This option is equivalent to
.BI \%\-\-epoch= 1900
and is used to specify the most common epoch on Alphas
with SRM console.
.
.SH NOTES
.
.SS Clocks in a Linux System
.PP
There are two types of date-time clocks:
.PP
.B The Hardware Clock:
This clock is an independent hardware device, with its own power domain
(battery, capacitor, etc), that operates when the machine is powered off,
or even unplugged.
.PP
On an ISA compatible system, this clock is specified as part of the ISA
standard.  A control program can read or set this clock only to a whole
second, but it can also detect the edges of the 1 second clock ticks, so
the clock actually has virtually infinite precision.
.PP
This clock is commonly called the hardware clock, the real time clock,
the RTC, the BIOS clock, and the CMOS clock.  Hardware Clock, in its
capitalized form, was coined for use by
.BR \%hwclock .
The Linux kernel also refers to it as the persistent clock.
.PP
Some non-ISA systems have a few real time clocks with
only one of them having its own power domain.
A very low power external I2C or SPI clock chip might be used with a
backup battery as the hardware clock to initialize a more functional
integrated real-time clock which is used for most other purposes.
.PP
.B The System Clock:
This clock is part of the Linux kernel and is driven by
a timer interrupt.  (On an ISA machine, the timer interrupt is part of
the ISA standard.)  It has meaning only while Linux is running on the
machine.  The System Time is the number of seconds since 00:00:00
January 1, 1970 UTC (or more succinctly, the number of seconds since
1969 UTC).  The System Time is not an integer, though.  It has virtually
infinite precision.
.PP
The System Time is the time that matters.  The Hardware Clock's basic
purpose is to keep time when Linux is not running so that the System
Clock can be initialized from it at boot.  Note that in DOS, for which
ISA was designed, the Hardware Clock is the only real time clock.
.PP
It is important that the System Time not have any discontinuities such as
would happen if you used the
.BR \%date (1)
program to set it while the system is running.  You can, however, do whatever
you want to the Hardware Clock while the system is running, and the next
time Linux starts up, it will do so with the adjusted time from the Hardware
Clock.  Note: currently this is not possible on most systems because
.B \%hwclock\ \-\-systohc
is called at shutdown.
.PP
The Linux kernel's timezone is set by
.BR hwclock .
But don't be misled -- almost nobody cares what timezone the kernel
thinks it is in.  Instead, programs that care about the timezone
(perhaps because they want to display a local time for you) almost
always use a more traditional method of determining the timezone: They
use the TZ environment variable or the
.I \%/etc/localtime
file, as explained in the man page for
.BR \%tzset (3).
However, some programs and fringe parts of the Linux kernel such as filesystems
use the kernel's timezone value.  An example is the vfat filesystem.  If the
kernel timezone value is wrong, the vfat filesystem will report and set the
wrong timestamps on files. Another example is the kernel's NTP \%'11\ minute\ mode.'
If the kernel's timezone value and/or the
.I \%persistent_clock_is_local
variable are wrong, then the Hardware Clock will be set incorrectly
by \%'11\ minute\ mode.'  See the discussion below, under
.BR "Automatic Hardware Clock Synchronization by the Kernel" .
.PP
.B \%hwclock
sets the kernel's timezone to the value indicated by TZ or
.IR \%/etc/localtime " with the"
.BR \%\-\-hctosys " or " \%\-\-systz " functions."
.PP
The kernel's timezone value actually consists of two parts: 1) a field
tz_minuteswest indicating how many minutes local time (not adjusted
for DST) lags behind UTC, and 2) a field tz_dsttime indicating
the type of Daylight Savings Time (DST) convention that is in effect
in the locality at the present time.
This second field is not used under Linux and is always zero.
See also
.BR \%settimeofday (2).
.
.SS User access and setuid
.PP
Sometimes, you need to install
.B \%hwclock
setuid root.  If you want users other than the superuser to be able to
display the clock value using the direct ISA I/O method, install it setuid
root.  If you have the rtc device interface on your system, or are on a non-ISA
compatible system, there is probably no need for users to have the direct
ISA I/O method, so do not bother.  See the
.BR \-\-rtc " option."
.PP
In any case, hwclock will not allow you to set anything unless you have the
superuser real uid.  (This restriction is not necessary if you haven't
installed setuid root, but it's there for now.)
.
.SS Hardware Clock Access Methods
.PP
.B \%hwclock
uses many different ways to get and set Hardware Clock values.  The most
normal way is to do I/O to the rtc device special file, which is
presumed to be driven by the rtc device driver.  Also, Linux systems
using the rtc framework with udev, are capable of supporting multiple
Hardware Clocks.  This may bring about the need to override the default
rtc device by specifying one with the
.BR \-\-rtc " option."
.PP
However, this method is not always available as older systems do not
have an rtc driver.  On these systems, the method of accessing the
Hardware Clock depends on the system hardware.
.PP
On an ISA compatible system,
.B \%hwclock
can directly access the "CMOS memory" registers that
constitute the clock, by doing I/O to Ports 0x70 and 0x71.  It does
this with actual I/O instructions and consequently can only do it if
running with superuser effective userid.  This method may be used by
specifying the
.BR \%\-\-directisa " option."
.PP
This is a really poor method of accessing the clock, for all the
reasons that userspace programs are generally not supposed to do
direct I/O and disable interrupts.
.B \%hwclock
provides it for testing, troubleshooting, and  because it may be the
only method available on ISA compatible and Alpha systems which do not
have a working rtc device driver.
.PP
In the case of a Jensen Alpha, there is no way for
.B \%hwclock
to execute those I/O instructions, and so it uses instead the
.I \%/dev/port
device special file, which provides almost as low-level an interface to
the I/O subsystem.
.PP
On an m68k system,
.B \%hwclock
can access the clock with the console driver, via the device special file
.IR \%/dev/tty1 .
.SS The Adjust Function
.PP
The Hardware Clock is usually not very accurate.  However, much of its
inaccuracy is completely predictable - it gains or loses the same amount
of time every day.  This is called systematic drift.
.BR \%hwclock "'s " \%\-\-adjust
function lets you apply systematic drift corrections to the
Hardware Clock.
.PP
It works like this:
.BR \%hwclock " keeps a file,"
.IR @ADJTIME_PATH@ ,
that keeps some historical information.  This is called the adjtime file.
.PP
Suppose you start with no adjtime file.  You issue a
.B \%hwclock\ \-\-set
command to set the Hardware Clock to the true current time.
.B \%hwclock
creates the adjtime file and records in it the current time as the
last time the clock was calibrated.
Five days later, the clock has gained 10 seconds, so you issue a
.B \%hwclock\ \-\-set\ \-\-update\-drift
command to set it back 10 seconds.
.B \%hwclock
updates the adjtime file to show the current time as the last time the
clock was calibrated, and records 2 seconds per day as the systematic
drift rate.  24 hours go by, and then you issue a
.B \%hwclock\ \-\-adjust
command.
.B \%hwclock
consults the adjtime file and sees that the clock gains 2 seconds per
day when left alone and that it has been left alone for exactly one
day.  So it subtracts 2 seconds from the Hardware Clock.  It then
records the current time as the last time the clock was adjusted.
Another 24 hours go by and you issue another
.BR \%hwclock\ \-\-adjust .
.B \%hwclock
does the same thing: subtracts 2 seconds and updates the adjtime file
with the current time as the last time the clock was adjusted.
.PP
When you use the
.BR \%\-\-update\-drift " option with " \-\-set " or " \%\-\-systohc ,
the systematic drift rate is (re)calculated by comparing the fully drift
corrected current Hardware Clock time with the new set time, from that
it derives the 24 hour drift rate based on the last calibrated timestamp
from the adjtime file.  This updated drift factor is then saved in
.IR @ADJTIME_PATH@ .
.PP
A small amount of error creeps in when
the Hardware Clock is set, so
.B \%\-\-adjust
refrains from making any adjustment that is less
than 1 second.  Later on, when you request an adjustment again, the accumulated
drift will be more than 1 second and
.B \%\-\-adjust
will make the adjustment including any fractional amount.
.PP
.B \%hwclock\ \-\-hctosys
also uses the adjtime file data to compensate the value read from the Hardware
Clock before using it to set the System Clock.  It does not share the 1 second
limitation of
.BR \%\-\-adjust ,
and will correct sub-second drift values immediately.  It does not
change the Hardware Clock time nor the adjtime file.  This may eliminate
the need to use
.BR \%\-\-adjust ,
unless something else on the system needs the Hardware Clock to be
compensated.
.
.SS The Adjtime File
While named for its historical purpose of controlling adjustments only,
it actually contains other information used by
.B hwclock
from one invocation to the next.
.PP
The format of the adjtime file is, in ASCII:
.PP
Line 1: Three numbers, separated by blanks: 1) the systematic drift rate
in seconds per day, floating point decimal; 2) the resulting number of
seconds since 1969 UTC of most recent adjustment or calibration,
decimal integer; 3) zero (for compatibility with
.BR \%clock (8))
as a decimal integer.
.PP
Line 2: One number: the resulting number of seconds since 1969 UTC of most
recent calibration.  Zero if there has been no calibration yet or it
is known that any previous calibration is moot (for example, because
the Hardware Clock has been found, since that calibration, not to
contain a valid time).  This is a decimal integer.
.PP
Line 3: "UTC" or "LOCAL".  Tells whether the Hardware Clock is set to
Coordinated Universal Time or local time.  You can always override this
value with options on the
.B \%hwclock
command line.
.PP
You can use an adjtime file that was previously used with the
.BR \%clock "(8) program with " \%hwclock .
.
.SS Automatic Hardware Clock Synchronization by the Kernel
.PP
You should be aware of another way that the Hardware Clock is kept
synchronized in some systems.  The Linux kernel has a mode wherein it
copies the System Time to the Hardware Clock every 11 minutes.
This is a good mode to use when you are using something sophisticated
like NTP to keep your System Clock synchronized. (NTP is a way to keep
your System Time synchronized either to a time server somewhere on the
network or to a radio clock hooked up to your system.  See RFC 1305.)
.PP
This mode (we'll call it \%'11\ minute\ mode') is off until something
turns it on.  The NTP daemon ntpd is one thing that turns it on.  You
can turn it off by running anything, including
.BR \%hwclock\ \-\-hctosys ,
that sets the System Clock the old fashioned way.  However, if the NTP daemon is
still running, it will turn 11 minute mode back on again the next time it
synchronizes the System Clock.
.PP
When \%'11\ minute\ mode' is active the 64 bit of the kernel's
.I time_status
variable is unset. The status variable can be checked with the
.BR \%adjtimex\ --print " or " \%ntptime " commands."
.PP
If your system runs with \%'11\ minute\ mode' on, it may need to use either
.BR \%\-\-hctosys " or " \%\-\-systz
in a startup script, especially if the Hardware Clock is configured to use
the local timescale. Unless the kernel is informed of what timescale the
Hardware Clock is using, it may clobber it with the wrong one. The kernel
uses UTC by default.
.PP
The first userspace command to set the System Clock informs the
kernel what timescale the Hardware Clock is using.  This happens via the
.I \%persistent_clock_is_local
kernel variable.  If
.BR \%\-\-hctosys " or " \%\-\-systz
is the first, it will set this variable according to the adjtime file or the
appropriate command-line argument.  Note that when using this capability and the
Hardware Clock timescale configuration is changed, then a reboot is required to
notify the kernel.
.PP
.B \%hwclock\ \-\-adjust
should not be used with NTP \%'11\ minute\ mode.'
.
.SS ISA Hardware Clock Century value
.PP
There is some sort of standard that defines CMOS memory Byte 50 on an ISA
machine as an indicator of what century it is.
.B \%hwclock
does not use or set that byte because there are some machines that
don't define the byte that way, and it really isn't necessary anyway,
since the year-of-century does a good job of implying which century it
is.
.PP
If you have a bona fide use for a CMOS century byte, contact the
.B \%hwclock
maintainer; an option may be appropriate.
.PP
Note that this section is only relevant when you are using the "direct
ISA" method of accessing the Hardware Clock.
ACPI provides a standard way to access century values, when they
are supported by the hardware.
.
.SH DATE-TIME CONFIGURATION
.in +4
.SS Keeping Time without External Synchronization
.in
.PP
This discussion is based on the following conditions:
.IP \(bu 2
Nothing is running that alters the date-time clocks, e.g.,
.BR \%ntpd "(1), cron jobs, et al."
.IP \(bu 2
The system timezone is configured for the correct local time.  See below
.BR "POSIX vs 'RIGHT'" .
.IP \(bu 2
Early in startup the following are called in this order:
.br
.BI \%adjtimex\ \-\-tick \ <value>\  \-\-frequency \ <value>
.br
.B \%hwclock\ \-\-hctosys
.IP \(bu 2
During shutdown the following is called:
.br
.B \%hwclock\ \-\-systohc
.PP
.in +4
.BR * " Systems without " adjtimex " may use " ntptime .
.in
.PP
Whether maintaining precision time with
.BR \%ntpd (1)
or not, it makes sense to configure the system to keep reasonably good
date-time on its own.
.PP
The first step in making that happen is having a clear understanding of
the big picture.  There are two completely separate hardware devices
running at their own speed and drifting away from the 'correct' time at
their own rates.  The methods and software for drift correction are
different for each of them.  However, most systems are configured to
exchange values between these two clocks at startup and shutdown.  Now
the individual device's time keeping errors are transferred back and
forth between each other.  Attempt to configure drift correction for only
one of them, and the other's drift will be overlaid upon it.  If the big
picture is not kept in mind, confusion will soon ensue.
.PP
This problem can be avoided when configuring drift correction for the
System Clock by simply not shutting down the machine.  This, plus the
fact that all of
.BR \%hwclock 's
precision (including calculating drift factors) depends upon the System
Clock's rate being correct, means that configuration of the System Clock
should be done first.
.PP
The System Clock drift is corrected with the
.BR \%adjtimex "(8) command's " \-\-tick " and " \%\-\-frequency
options.  These two work together, tick is the course adjustment and
frequency is the fine adjustment. (For system that do not have an
.BR \%adjtimex " package,"
.BI \%ntptime\ \-f\  <ppm>
may be use instead.)
.PP
Some Linux distributions attempt to automatically calculate the System
Clock drift with
.BR \%adjtimex 's
compare operation.  Trying to correct one
drifting clock by using another drifting clock as a reference is akin to
a dog trying to catch its own tail.  Success may happen eventually, but
great effort and frustration will likely precede it.  This automation may
yield an improvement over no configuration, but expecting optimum
results would be in error.  A better choice for manual configuration
would be
.BR \%adjtimex 's " \-\-log " options.
.PP
It may be more effective to simply track the System Clock drift with
.BR \%ntpdate\ \-q " , or " \%date\ \-Ins
and a precision timepiece, and then calculate the correction manually.
.PP
After setting the tick and frequency values, continue to test and refine the
adjustments until the System Clock keeps good time.  See
.BR \%adjtimex (8)
for more information and the example demonstrating manual drift
calculations.
.PP
Once the System Clock is ticking smoothly, move on to the Hardware Clock.
.PP
As a rule, cold drift will work best for most use cases.  This should be
true even for 24/7 machines whose normal downtime consists of a reboot.
In that case the drift factor value makes little difference, but on the
rare occasion that the machine is shutdown for an extended period then
cold drift should yield better results.
.PP
.B Steps to calculate cold drift:
.IP 1 2
.RB "Confirm that " ntpd "(1) will not be launched at startup."
.IP 2 2
.RI The " System Clock " "time must be correct at shutdown!"
.IP 3 2
Shutdown the system.
.IP 4 2
Let an extended period pass without changing the Hardware Clock.
.IP 5 2
Start the system.
.IP 6 2
.RB "Immediately use " hwclock " to set the correct time with the"
.BR \%\-\-update\-drift " option."
.PP
Note: if step six uses
.RB \%\-\-systohc ,
then the System Clock must be set correctly (step 6a) just before doing so.
.PP
.RB "Having " hwclock
calculate the drift factor is a good starting point, but for optimal
results it will likely need to be adjusted by directly editing the
.I @ADJTIME_PATH@
file.  Continue to test and refine the drift factor until the Hardware
Clock is corrected properly at startup.  To check this, first make sure
that the System Time is correct before shutdown and then use
.BR \%ntpdate\ \-q ", or " \%date\ \-Ins
and a precision timepiece, immediately after startup.
.PP
Both clocks typically use a quartz crystal oscillator.  Crystals are
used for reference oscillators in electronics because by most measures
they produce a very clean and stable sine wave.  Their greatest
shortcoming is that they have a Positive Temperature Coefficient;
meaning that their frequency increases as the temperature increases and
vise versa. Therefore, both the Hardware and System Clock's drift rate
changes with intrinsic and extrinsic machine temperatures.  These
characteristics will vary by machine depending upon its design.
.PP
Drift correction strategies are many, but as a general guide the goal
would be to find a longterm average.  A year long average to take into
account seasonal ambient temperature shifts may be a good target period.
So perhaps the date-time advances a bit in the summer and declines a bit
in the winter, but at the end of a year it balances to zero.
.PP
If this is beginning to sound futile, it is not.  Left on its own a
machine can lose 3 seconds per day or more.  Accumulated drift over a
year may easily exceed half an hour.  Using carefully crafted drift
corrections can make a significant improvement in a machine's ability to
keep reasonably good date-time.
.
.SS LOCAL vs UTC
Keeping the Hardware Clock in a local timescale causes inconsistent
daylight saving time results:
.IP \(bu 2
If Linux is running during a daylight saving time change, the time
written to the Hardware Clock will be adjusted for the change.
.IP \(bu 2
If Linux is NOT running during a daylight saving time change, the time
read from the Hardware Clock will NOT be adjusted for the change.
.PP
The Hardware Clock on an ISA compatible system keeps only a date and time,
it has no concept of timezone nor daylight saving. Therefore, when
.B hwclock
is told that it is in local time, it assumes it is in the 'correct'
local time and makes no adjustments to the time read from it.
.PP
Linux handles daylight saving time changes transparently only when the
Hardware Clock is kept in the UTC timescale. Doing so is made easy for
system administrators as
.B \%hwclock
uses local time for its output and as the argument to the
.BR \%\-\-date " option."
.PP
POSIX systems, like Linux, are designed to have the System Clock operate
in the UTC timescale. The Hardware Clock's purpose is to initialize the
System Clock, so also keeping it in UTC makes sense.
.PP
Linux does, however, attempt to accommodate the Hardware Clock being in
the local timescale. This is primarily for dual-booting with older
versions of MS Windows. From Windows 7 on, the RealTimeIsUniversal
registry key is supposed to be working properly so that its Hardware
Clock can be kept in UTC.
.
.SS POSIX vs 'RIGHT'
A discussion on date-time configuration would be incomplete without
addressing timezones, this is mostly well covered by
.BR tzset (3).
One area that seems to have no documentation is the 'right'
directory of the Time Zone Database, sometimes called tz or zoneinfo.
.PP
There are two separate databases in the zoneinfo system, posix
and 'right'. 'Right' (now named zoneinfo\-leaps) includes leap seconds and posix
does not. To use the 'right' database the System Clock must be set to
\%(UTC\ +\ leap seconds), which is equivalent to \%(TAI\ \-\ 10). This
allows calculating the
exact number of seconds between two dates that cross a leap second
epoch. The System Clock is then converted to the correct civil time,
including UTC, by using the 'right' timezone files which subtract the
leap seconds. Note: this configuration is considered experimental and is
known to have issues.
.PP
To configure a system to use a particular database all of the files
located in its directory must be copied to the root of
.IR \%/usr/share/zoneinfo .
Files are never used directly from the posix or 'right' subdirectories, e.g.,
.RI \%TZ=' right/Europe/Dublin '.
This habit was becoming so common that the upstream zoneinfo project
restructured the system's file tree by moving the posix and 'right'
subdirectories out of the zoneinfo directory and into sibling directories:
.PP
.in +2
.I /usr/share/zoneinfo
.br
.I /usr/share/zoneinfo\-posix
.br
.I /usr/share/zoneinfo\-leaps
.PP
Unfortunately, some Linux distributions are changing it back to the old
tree structure in their packages. So the problem of system
administrators reaching into the 'right' subdirectory persists. This
causes the system timezone to be configured to include leap seconds
while the zoneinfo database is still configured to exclude them. Then
when an application such as a World Clock needs the South_Pole timezone
file; or an email MTA, or
.B hwclock
needs the UTC timezone file; they fetch it from the root of
.I \%/usr/share/zoneinfo
, because that is what they are supposed to do. Those files exclude leap
seconds, but the System Clock now includes them, causing an incorrect
time conversion.
.PP
Attempting to mix and match files from these separate databases will not
work, because they each require the System Clock to use a different
timescale. The zoneinfo database must be configured to use either posix
or 'right', as described above.
.
.SH "ENVIRONMENT VARIABLES"
.I TZ
.
.SH FILES
.br
.I @ADJTIME_PATH@
.br
.I /etc/localtime
.br
.I /dev/rtc
.br
.I /dev/rtc0
.br
.I /dev/misc/rtc
.br
.I /dev/efirtc
.br
.I /dev/misc/efirtc
.br
.I /dev/port
.br
.I /dev/tty1
.br
.I /proc/cpuinfo
.
.SH "SEE ALSO"
.BR date (1),
.BR adjtimex (8),
.BR gettimeofday (2),
.BR settimeofday (2),
.BR crontab (1),
.BR tzset (3)
.
.SH AUTHORS
Written by Bryan Henderson, September 1996 (bryanh@giraffe-data.com),
based on work done on the
.BR \%clock (8)
program by Charles Hedrick, Rob Hooft, and Harald Koenig.
See the source code for complete history and credits.
.
.SH AVAILABILITY
The hwclock command is part of the util-linux package and is available from
ftp://ftp.kernel.org/pub/linux/utils/util-linux/.