2d4e511ca2
Update ntp-4.2.8p13 --> 4.2.8p14. The advisory can be found at: http://support.ntp.org/bin/view/Main/SecurityNotice#\ March_2020_ntp_4_2_8p14_NTP_Rele No CVEs have been documented yet. MFC after: now Security: http://support.ntp.org/bin/view/Main/NtpBug3610 http://support.ntp.org/bin/view/Main/NtpBug3596 http://support.ntp.org/bin/view/Main/NtpBug3592
223 lines
4.8 KiB
C
223 lines
4.8 KiB
C
/*
|
|
* timespecops.h -- calculations on 'struct timespec' values
|
|
*
|
|
* Written by Juergen Perlinger (perlinger@ntp.org) for the NTP project.
|
|
* The contents of 'html/copyright.html' apply.
|
|
*
|
|
* Rationale
|
|
* ---------
|
|
*
|
|
* Doing basic arithmetic on a 'struct timespec' is not exceedingly
|
|
* hard, but it requires tedious and repetitive code to keep the result
|
|
* normalised. We consider a timespec normalised when the nanosecond
|
|
* fraction is in the interval [0 .. 10^9[ ; there are multiple value
|
|
* pairs of seconds and nanoseconds that denote the same time interval,
|
|
* but the normalised representation is unique. No two different
|
|
* intervals can have the same normalised representation.
|
|
*
|
|
* Another topic is the representation of negative time intervals.
|
|
* There's more than one way to this, since both the seconds and the
|
|
* nanoseconds of a timespec are signed values. IMHO, the easiest way is
|
|
* to use a complement representation where the nanoseconds are still
|
|
* normalised, no matter what the sign of the seconds value. This makes
|
|
* normalisation easier, since the sign of the integer part is
|
|
* irrelevant, and it removes several sign decision cases during the
|
|
* calculations.
|
|
*
|
|
* As long as no signed integer overflow can occur with the nanosecond
|
|
* part of the operands, all operations work as expected and produce a
|
|
* normalised result.
|
|
*
|
|
* The exception to this are functions fix a '_fast' suffix, which do no
|
|
* normalisation on input data and therefore expect the input data to be
|
|
* normalised.
|
|
*
|
|
* Input and output operands may overlap; all input is consumed before
|
|
* the output is written to.
|
|
*/
|
|
#ifndef TIMESPECOPS_H
|
|
#define TIMESPECOPS_H
|
|
|
|
#include <sys/types.h>
|
|
#include <stdio.h>
|
|
#include <math.h>
|
|
|
|
#include "ntp.h"
|
|
#include "timetoa.h"
|
|
|
|
|
|
/* nanoseconds per second */
|
|
#define NANOSECONDS 1000000000
|
|
|
|
/* predicate: returns TRUE if the nanoseconds are in nominal range */
|
|
#define timespec_isnormal(x) \
|
|
((x)->tv_nsec >= 0 && (x)->tv_nsec < NANOSECONDS)
|
|
|
|
/* predicate: returns TRUE if the nanoseconds are out-of-bounds */
|
|
#define timespec_isdenormal(x) (!timespec_isnormal(x))
|
|
|
|
|
|
|
|
|
|
/* make sure nanoseconds are in nominal range */
|
|
extern struct timespec normalize_tspec(struct timespec x);
|
|
|
|
/* x = a + b */
|
|
static inline struct timespec
|
|
add_tspec(
|
|
struct timespec a,
|
|
struct timespec b
|
|
)
|
|
{
|
|
struct timespec x;
|
|
|
|
x = a;
|
|
x.tv_sec += b.tv_sec;
|
|
x.tv_nsec += b.tv_nsec;
|
|
|
|
return normalize_tspec(x);
|
|
}
|
|
|
|
/* x = a + b, b is fraction only */
|
|
static inline struct timespec
|
|
add_tspec_ns(
|
|
struct timespec a,
|
|
long b
|
|
)
|
|
{
|
|
struct timespec x;
|
|
|
|
x = a;
|
|
x.tv_nsec += b;
|
|
|
|
return normalize_tspec(x);
|
|
}
|
|
|
|
/* x = a - b */
|
|
static inline struct timespec
|
|
sub_tspec(
|
|
struct timespec a,
|
|
struct timespec b
|
|
)
|
|
{
|
|
struct timespec x;
|
|
|
|
x = a;
|
|
x.tv_sec -= b.tv_sec;
|
|
x.tv_nsec -= b.tv_nsec;
|
|
|
|
return normalize_tspec(x);
|
|
}
|
|
|
|
/* x = a - b, b is fraction only */
|
|
static inline struct timespec
|
|
sub_tspec_ns(
|
|
struct timespec a,
|
|
long b
|
|
)
|
|
{
|
|
struct timespec x;
|
|
|
|
x = a;
|
|
x.tv_nsec -= b;
|
|
|
|
return normalize_tspec(x);
|
|
}
|
|
|
|
/* x = -a */
|
|
static inline struct timespec
|
|
neg_tspec(
|
|
struct timespec a
|
|
)
|
|
{
|
|
struct timespec x;
|
|
|
|
x.tv_sec = -a.tv_sec;
|
|
x.tv_nsec = -a.tv_nsec;
|
|
|
|
return normalize_tspec(x);
|
|
}
|
|
|
|
/* x = abs(a) */
|
|
struct timespec abs_tspec(struct timespec a);
|
|
|
|
/*
|
|
* compare previously-normalised a and b
|
|
* return 1 / 0 / -1 if a < / == / > b
|
|
*/
|
|
extern int cmp_tspec(struct timespec a, struct timespec b);
|
|
|
|
/*
|
|
* compare possibly-denormal a and b
|
|
* return 1 / 0 / -1 if a < / == / > b
|
|
*/
|
|
static inline int
|
|
cmp_tspec_denorm(
|
|
struct timespec a,
|
|
struct timespec b
|
|
)
|
|
{
|
|
return cmp_tspec(normalize_tspec(a), normalize_tspec(b));
|
|
}
|
|
|
|
/*
|
|
* test previously-normalised a
|
|
* return 1 / 0 / -1 if a < / == / > 0
|
|
*/
|
|
extern int test_tspec(struct timespec a);
|
|
|
|
/*
|
|
* test possibly-denormal a
|
|
* return 1 / 0 / -1 if a < / == / > 0
|
|
*/
|
|
static inline int
|
|
test_tspec_denorm(
|
|
struct timespec a
|
|
)
|
|
{
|
|
return test_tspec(normalize_tspec(a));
|
|
}
|
|
|
|
/* return LIB buffer ptr to string rep */
|
|
static inline const char *
|
|
tspectoa(
|
|
struct timespec x
|
|
)
|
|
{
|
|
return format_time_fraction(x.tv_sec, x.tv_nsec, 9);
|
|
}
|
|
|
|
/*
|
|
* convert to l_fp type, relative and absolute
|
|
*/
|
|
|
|
/* convert from timespec duration to l_fp duration */
|
|
extern l_fp tspec_intv_to_lfp(struct timespec x);
|
|
|
|
/* x must be UN*X epoch, output will be in NTP epoch */
|
|
static inline l_fp
|
|
tspec_stamp_to_lfp(
|
|
struct timespec x
|
|
)
|
|
{
|
|
l_fp y;
|
|
|
|
y = tspec_intv_to_lfp(x);
|
|
y.l_ui += JAN_1970;
|
|
|
|
return y;
|
|
}
|
|
|
|
/* convert from l_fp type, relative signed/unsigned and absolute */
|
|
extern struct timespec lfp_intv_to_tspec(l_fp x);
|
|
extern struct timespec lfp_uintv_to_tspec(l_fp x);
|
|
|
|
/*
|
|
* absolute (timestamp) conversion. Input is time in NTP epoch, output
|
|
* is in UN*X epoch. The NTP time stamp will be expanded around the
|
|
* pivot time *p or the current time, if p is NULL.
|
|
*/
|
|
extern struct timespec lfp_stamp_to_tspec(l_fp x, const time_t *pivot);
|
|
|
|
#endif /* TIMESPECOPS_H */
|