| History log of /src/sys/kern/subr_time.c |
| Revision | | Date | Author | Comments |
| 1.41 |
| 22-Dec-2024 |
riastradh | kern: Move some purely arithmetic routines to subr_time_arith.c.
Preparation for testing and fixing:
PR kern/58922: itimer(9): arithmetic overflow
PR kern/58925: itimer(9) responds erratically to clock wound back
PR kern/58926: itimer(9) integer overflow in overrun counting
PR kern/58927: itimer(9): overrun accounting is broken
|
| 1.40 |
| 22-Dec-2024 |
riastradh | subr_time.c: Sort includes.
No functional change intended.
Preparation for factoring out arithmetic to test and fix:
PR kern/58922: itimer(9): arithmetic overflow
PR kern/58925: itimer(9) responds erratically to clock wound back
PR kern/58926: itimer(9) integer overflow in overrun counting
PR kern/58927: itimer(9): overrun accounting is broken
|
| 1.39 |
| 10-Oct-2024 |
kre | PR kern/58733 - avoid ts2timo() clobbering its arg
See the PR for the gory details - in the TIMER_ABSTIME case
ts2timo() should not (really *must* not) alter the timespec
it is passed (in that case it should be const - but for now
anyway, cannot be for the TIMER_RELTIME case, and there is
just one of them!)
XXX pullup -10
XXX pullup -9 (will need a patch).
|
| 1.38 |
| 08-Jul-2023 |
riastradh | clock_gettime(2): Fix CLOCK_PROCESS/THREAD_CPUTIME_ID.
Use same calculation as getrusage, not some ad-hoc arithmetic of
internal scheduler parameters that are periodically rewound.
PR kern/57512
XXX pullup-8
XXX pullup-9
XXX pullup-10
|
| 1.37 |
| 29-Apr-2023 |
isaki | White space fix.
|
| 1.36 |
| 09-Apr-2023 |
riastradh | kern: KASSERT(A && B) -> KASSERT(A); KASSERT(B)
|
| 1.35 |
| 28-Jun-2022 |
riastradh | branches: 1.35.4;
kern: Avoid arithmetic overflow in gettimeleft.
Sprinkle assertions in to verify we're monotonically counting the
time left down to zero.
Reported-by: syzbot+5f6a6329d139810dfe3c@syzkaller.appspotmail.com
|
| 1.34 |
| 26-Jun-2022 |
riastradh | kern: Use timespecsubok in ts2timo.
Should fix arithmetic overflow.
Reported-by: syzbot+4393a753a7c787abe6e2@syzkaller.appspotmail.com
|
| 1.33 |
| 26-Jun-2022 |
riastradh | kern: New functions timespecaddok, timespecsubok.
Return false if timespecadd or timespecsub with the same arguments
would overflow (possibly in an intermediate calculation), true if OK.
Typical usage:
sys_wotsit(...)
{
...
if (!timespecsubok(x, y))
return EINVAL;
timespecub(x, y, xydelta);
...
}
|
| 1.32 |
| 13-Mar-2022 |
riastradh | kern: Fix fencepost error in ts2timo overflow checks.
Triggered by
clock_settime({.tv_sec=0, .tv_nsec=0})
clock_nanosleep({.tv_sec=LLONG_MIN, .tv_nsec=0})
so that, by the time we enter ts2timo (after a few nanoseconds have
passed), we end up with
tsd = {.tv_sec=0, .tv_nsec=nonzero}
ts = {.tv_sec=LLONG_MIN, .tv_nsec=0}
and the subtraction ts - tsd leads to a borrow from tv_sec.
Reported-by: syzbot+14818113e9d0b45bca64@syzkaller.appspotmail.com
|
| 1.31 |
| 21-Sep-2021 |
christos | don't opencode kauth_cred_get()
|
| 1.30 |
| 18-Mar-2021 |
nia | restore flags-as-bitmask, just in case another function is passing its
flags here.
|
| 1.29 |
| 18-Mar-2021 |
nia | ts2timo(9): refactor TIMER_ABSTIME handling
- only use *start for output of the original time.
for clarity purposes, use the temporary variable for everything else.
- add a check for integer underflow
Reported-by: syzbot+17b5072d5ed262a966d3@syzkaller.appspotmail.com
|
| 1.28 |
| 18-Mar-2021 |
nia | revert previous
|
| 1.27 |
| 18-Mar-2021 |
nia | ts2timo(9): further deobfuscation.
we want to check against NULL, so use it, rather than some random value on
the stack
|
| 1.26 |
| 18-Mar-2021 |
nia | ts2timo(9): refactor for clarity
- 'flags' is not a boolean...
- actually, it is, but it should simply be named "absolute".
- convert tests for if (flags) to if (flags != TIMER_RELTIME)
- hoist function calls out of if expressions (requested by uwe)
still needs fixing:
- need to check for overflow before timespecsub.
|
| 1.25 |
| 23-May-2020 |
ad | branches: 1.25.2;
Move proc_lock into the data segment. It was dynamically allocated because
at the time we had mutex_obj_alloc() but not __cacheline_aligned.
|
| 1.24 |
| 11-May-2020 |
riastradh | Remove timedwaitclock.
This did not fix the bug I hoped it would fix in futex, and needs
more design thought. Might redo it somewhat differently later.
|
| 1.23 |
| 04-May-2020 |
riastradh | New timedwaitclock_setup.
C99 initializers would have been nice, but part of the struct is
explicit parameters and part of the struct is implicit state, and
-Wmissing-field-initializers can't discriminate between them
(although for some reason it doesn't always fire!).
Instead, just do:
struct timedwaitclock T;
timedwaitclock_setup(&T, timeout, clockid, flags, epsilon);
while (...) {
error = timedwaitclock_begin(&T, &timo);
if (error)
...
error = waitwhatever(timo);
timedwaitclock_end(&T);
...
}
|
| 1.22 |
| 03-May-2020 |
thorpej | Move timedwaitclock_begin() and timedwaitclock_end() to subr_time.c
so they can be used by other things.
|
| 1.21 |
| 04-Oct-2019 |
kamil | Avoid signed integer overflow in ts2timo() for ts->tv_nsec
The condition would be rechecked later again after subtracting start time
and most invalid inputs rejected. In corner cases the current code can
accept certain invalid inputs that will pass checks later and behave like
valid ones (due to signed integer overflow).
Reported-by: syzbot+3a4a07b62558bbbd3baa@syzkaller.appspotmail.com
|
| 1.20 |
| 08-Dec-2017 |
christos | branches: 1.20.4; 1.20.8;
make _lwp_park return the remaining time to sleep in the "ts" argument
if it is a relative timestamp, as discussed in tech-kern.
XXX: pullup-8
|
| 1.19 |
| 05-Jan-2017 |
pgoyette | Use the new magic BINTIME_SCALE_* macros instead of magic numbers.
No functional change.
|
| 1.18 |
| 23-Apr-2016 |
christos | branches: 1.18.2;
Add clock_getcpuclockid2(2) as well as CLOCK_{PROCESS,THREAD}_CPUTIME_ID.
|
| 1.17 |
| 22-May-2013 |
christos | branches: 1.17.10;
Make ts2timo(9) always return the absolute start time if the start argument
is present, and handle the TIMER_ABSTIME case in nanosleep1(9).
|
| 1.16 |
| 21-May-2013 |
bouyer | ts2timo(): if TIMER_ABSTIME is set and start is not NULL, initialize
it to 0. Some callers (e.g. nanosleep1()) expect *start to always be
initialized and would use random values from stack otherwise.
While there, remove an always-true conditionnal.
|
| 1.15 |
| 01-Apr-2013 |
christos | undo previous and move the test to the timeout function since 0,0 means
disable timer/interval.
|
| 1.14 |
| 01-Apr-2013 |
christos | do the timeout test centrally.
|
| 1.13 |
| 01-Apr-2013 |
martin | ts2timo: return ETIMEDOUT instead of failing an assertion when the
calculated difference to the target time is zero.
|
| 1.12 |
| 31-Mar-2013 |
christos | instead of doing the tests twice fix the *fix() routines to return ETIMEDOUT
if seconds are negative. Accorting to TOG, this is not an error as linux
claims. Also make an assert stricter.
|
| 1.11 |
| 29-Mar-2013 |
martin | Move clock_gettime1() to subr_time.c (which is included in rump kernels)
|
| 1.10 |
| 29-Mar-2013 |
christos | Centralize the computation of struct timespec to the int timo.
Make lwp_park take the regular arguments for specifying what kind
of timeout we supply like clock_nanosleep(), namely clockid_t and flags.
|
| 1.9 |
| 18-Dec-2011 |
christos | branches: 1.9.6;
Fix monotonic interval timers.
|
| 1.8 |
| 26-Jan-2011 |
drochner | branches: 1.8.4; 1.8.8;
fix and cleanup for tvtohz():
-assume (KASSERT) that the timeval given is normalized, and remove
some partial fixup which I don't see what it is good for
(I'm ready to back that out if someone tells a reason)
-catch overflows due to conversion of time_t (from tv_sec) to
integer -- this function doesn't do 64-bit arithmetics (which makes
sense because relative times which don't fit into 32 bits can be
considered nonsense here), and before a huge tv_sec could lead to
a zero hz result, violating the caller's assumptions (in particular
trigger a diagnostic panic in abstimeout2timo())
|
| 1.7 |
| 26-Apr-2010 |
rmind | branches: 1.7.2; 1.7.4;
Revert rev1.6, it shall return ETIMEDOUT for pthread calls.
|
| 1.6 |
| 23-Apr-2010 |
rmind | mq_timed{send,receive}: as required by POSIX, return EINVAL on invalid
timeout and thread would have blocked, instead of ETIMEDOUT. Change is
to abstimeout2timo(), thus also affects _lwp_park(2).
Reported by Stathis Kamperis some months ago.
|
| 1.5 |
| 01-Nov-2009 |
rmind | branches: 1.5.4;
- Move inittimeleft() and gettimeleft() to subr_time.c, where they belong.
- Move abstimeout2timo() there too and export. Use it in lwp_park().
|
| 1.4 |
| 15-Jul-2008 |
christos | Use more timespecs internally. From Alexander Shishkin and me.
Welcome to 4.99.70, 30 more to go for 100.
|
| 1.3 |
| 22-Dec-2007 |
yamt | branches: 1.3.6; 1.3.10; 1.3.12; 1.3.14; 1.3.16;
reduce #ifdef __HAVE_TIMECOUNTER.
|
| 1.2 |
| 29-Nov-2007 |
ad | branches: 1.2.2; 1.2.6;
Pull in sys/intr.h
|
| 1.1 |
| 09-Aug-2007 |
pooka | branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; 1.1.10; 1.1.16;
Shuffle routines which just roll values around from kern_clock.c
and kern_time.c to subr_time.c.
|
| 1.1.16.2 |
| 27-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.1.16.1 |
| 08-Dec-2007 |
mjf | Sync with HEAD.
|
| 1.1.10.4 |
| 21-Jan-2008 |
yamt | sync with head
|
| 1.1.10.3 |
| 07-Dec-2007 |
yamt | sync with head
|
| 1.1.10.2 |
| 03-Sep-2007 |
yamt | sync with head.
|
| 1.1.10.1 |
| 09-Aug-2007 |
yamt | file subr_time.c was added on branch yamt-lazymbuf on 2007-09-03 14:41:06 +0000
|
| 1.1.8.1 |
| 09-Jan-2008 |
matt | sync with HEAD
|
| 1.1.6.2 |
| 20-Aug-2007 |
ad | Sync with HEAD.
|
| 1.1.6.1 |
| 09-Aug-2007 |
ad | file subr_time.c was added on branch vmlocking on 2007-08-20 22:07:07 +0000
|
| 1.1.4.3 |
| 03-Dec-2007 |
joerg | Sync with HEAD.
|
| 1.1.4.2 |
| 16-Aug-2007 |
jmcneill | Sync with HEAD.
|
| 1.1.4.1 |
| 09-Aug-2007 |
jmcneill | file subr_time.c was added on branch jmcneill-pm on 2007-08-16 11:03:37 +0000
|
| 1.1.2.2 |
| 15-Aug-2007 |
skrll | Sync with HEAD.
|
| 1.1.2.1 |
| 09-Aug-2007 |
skrll | file subr_time.c was added on branch nick-csl-alignment on 2007-08-15 13:49:14 +0000
|
| 1.2.6.1 |
| 02-Jan-2008 |
bouyer | Sync with HEAD
|
| 1.2.2.1 |
| 26-Dec-2007 |
ad | Sync with head.
|
| 1.3.16.1 |
| 19-Oct-2008 |
haad | Sync with HEAD.
|
| 1.3.14.1 |
| 18-Jul-2008 |
simonb | Sync with head.
|
| 1.3.12.1 |
| 18-Sep-2008 |
wrstuden | Sync with wrstuden-revivesa-base-2.
|
| 1.3.10.2 |
| 11-Mar-2010 |
yamt | sync with head
|
| 1.3.10.1 |
| 04-May-2009 |
yamt | sync with head.
|
| 1.3.6.1 |
| 28-Sep-2008 |
mjf | Sync with HEAD.
|
| 1.5.4.1 |
| 05-Mar-2011 |
rmind | sync with head
|
| 1.7.4.1 |
| 08-Feb-2011 |
bouyer | Sync with HEAD
|
| 1.7.2.1 |
| 06-Jun-2011 |
jruoho | Sync with HEAD.
|
| 1.8.8.1 |
| 18-Feb-2012 |
mrg | merge to -current.
|
| 1.8.4.2 |
| 22-May-2014 |
yamt | sync with head.
for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.
this commit was splitted into small chunks to avoid
a limitation of cvs. ("Protocol error: too many arguments")
|
| 1.8.4.1 |
| 17-Apr-2012 |
yamt | sync with head
|
| 1.9.6.2 |
| 03-Dec-2017 |
jdolecek | update from HEAD
|
| 1.9.6.1 |
| 23-Jun-2013 |
tls | resync from head
|
| 1.17.10.2 |
| 05-Feb-2017 |
skrll | Sync with HEAD
|
| 1.17.10.1 |
| 29-May-2016 |
skrll | Sync with HEAD
|
| 1.18.2.1 |
| 07-Jan-2017 |
pgoyette | Sync with HEAD. (Note that most of these changes are simply $NetBSD$
tag issues.)
|
| 1.20.8.2 |
| 13-Oct-2024 |
martin | Pull up following revision(s) (requested by kre in ticket #1909):
share/man/man9/ts2timo.9: revision 1.4
share/man/man9/ts2timo.9: revision 1.5
sys/kern/subr_time.c: revision 1.39 (patch)
ts2timo() uses struct timespec, those don't have a tv_usec
field, they have tv_nsec instead. EINVAL will happen if the
tv_nsec field is invalid, not the non-existant tv_usec field.
PR kern/58733 - avoid ts2timo() clobbering its arg
See the PR for the gory details - in the TIMER_ABSTIME case
ts2timo() should not (really *must* not) alter the timespec
it is passed (in that case it should be const - but for now
anyway, cannot be for the TIMER_RELTIME case, and there is
just one of them!).
|
| 1.20.8.1 |
| 11-Oct-2024 |
martin | Pull up following revision(s) (requested by riastradh in ticket #1896):
sys/sys/resourcevar.h: revision 1.58
sys/kern/subr_time.c: revision 1.36
sys/kern/subr_time.c: revision 1.37
sys/kern/subr_time.c: revision 1.38
sys/kern/kern_resource.c: revision 1.190
sys/kern/kern_resource.c: revision 1.191
tests/lib/libc/sys/t_clock_gettime.c: revision 1.4
tests/lib/libc/sys/t_clock_gettime.c: revision 1.5
tests/lib/libc/sys/t_clock_gettime.c: revision 1.6
(all via patch)
kern: KASSERT(A && B) -> KASSERT(A); KASSERT(B)
White space fix.
kern_resource.c: Fix brace placement.
No functional change intended.
t_clock_gettime: Add test for PR kern/57512.
clock_gettime(2): Fix CLOCK_PROCESS/THREAD_CPUTIME_ID.
Use same calculation as getrusage, not some ad-hoc arithmetic of
internal scheduler parameters that are periodically rewound.
PR kern/57512
t_clock_gettime: Avoid signed/unsigned comparison.
|
| 1.20.4.1 |
| 13-Apr-2020 |
martin | Mostly merge changes from HEAD upto 20200411
|
| 1.25.2.1 |
| 03-Apr-2021 |
thorpej | Sync with HEAD.
|
| 1.35.4.2 |
| 11-Oct-2024 |
martin | Pull up following revision(s) (requested by riastradh in ticket #943):
sys/sys/resourcevar.h: revision 1.58
sys/kern/subr_time.c: revision 1.36
sys/kern/subr_time.c: revision 1.37
sys/kern/subr_time.c: revision 1.38
sys/kern/kern_resource.c: revision 1.190
sys/kern/kern_resource.c: revision 1.191
tests/lib/libc/sys/t_clock_gettime.c: revision 1.4
tests/lib/libc/sys/t_clock_gettime.c: revision 1.5
tests/lib/libc/sys/t_clock_gettime.c: revision 1.6
kern: KASSERT(A && B) -> KASSERT(A); KASSERT(B)
White space fix.
kern_resource.c: Fix brace placement.
No functional change intended.
t_clock_gettime: Add test for PR kern/57512.
clock_gettime(2): Fix CLOCK_PROCESS/THREAD_CPUTIME_ID.
Use same calculation as getrusage, not some ad-hoc arithmetic of
internal scheduler parameters that are periodically rewound.
PR kern/57512
XXX pullup-8
XXX pullup-9
XXX pullup-10
t_clock_gettime: Avoid signed/unsigned comparison.
|
| 1.35.4.1 |
| 11-Oct-2024 |
martin | Pull up following revision(s) (requested by kre in ticket #942):
share/man/man9/ts2timo.9: revision 1.4
share/man/man9/ts2timo.9: revision 1.5
sys/kern/subr_time.c: revision 1.39
ts2timo() uses struct timespec, those don't have a tv_usec
field, they have tv_nsec instead. EINVAL will happen if the
tv_nsec field is invalid, not the non-existant tv_usec field.
PR kern/58733 - avoid ts2timo() clobbering its arg
See the PR for the gory details - in the TIMER_ABSTIME case
ts2timo() should not (really *must* not) alter the timespec
it is passed (in that case it should be const - but for now
anyway, cannot be for the TIMER_RELTIME case, and there is
just one of them!).
|
