summaryrefslogtreecommitdiff
path: root/sys/kern/kern_tc.c
AgeCommit message (Expand)Author
2021-06-02timecounter(9): Zero the next pointer after tc_detach.riastradh
2021-04-08Sort includes.simonb
2021-04-08tc_init(): KASSERT that the timecounter hasn't previously beensimonb
2020-05-27- remove #ifdef __FreeBSD__rin
2020-05-27Oops, put back opt_ntp.h for options PPS_{SYNC,DEBUG}.rin
2020-05-27- kern_tc.c does not depend on "options NTP"rin
2020-05-27Tiny optimization; timecounters should not go bad ~ every ticktock.rin
2020-05-23Move proc_lock into the data segment. It was dynamically allocated becausead
2020-01-02- Eliminate the global "boottime" variable, which was being accessedthorpej
2019-12-27s/inital/initial/msaitoh
2019-10-06xc_barrier - convenience function to xc_broadcast() a nop.uwe
2018-07-01Revert "Sprinkle cold conditionals to make tc_ticktock before inittimecounter."riastradh
2018-06-30Sprinkle cold conditionals to make tc_ticktock before inittimecounter.riastradh
2018-02-13Put time_second and time_uptime in different cache lines, probably savesmaxv
2017-11-02C99ify initialization of dummy_timecounter.riastradh
2017-06-09define a copy of getnanotime() named dtrace_getnanotime() so thatchs
2013-09-14ifdef a variable like its usemartin
2013-05-26Extend kernel PPS api with pps_ref_event().kardel
2012-11-13kill some -Wunused-but-set-variable warningspooka
2012-02-21Make time_second and time_uptime volatile, so the compiler knows theymartin
2010-04-13_KERNEL_OPTpooka
2010-02-06__inline -> inlineuebayasi
2009-06-14Make PPS work with fast time counters (> 2GHz)kardel
2009-05-23- Add lwp_pctr(), get an LWP's preemption/ctxsw counter.ad
2009-01-11merge christos-time_tchristos
2008-07-19Initialize timecounter_lock at IPL_HIGH. Prevents a "locking against myself"kardel
2008-05-11Fix locking botch. Pointed out by kardel@.ad
2008-05-08- Add tc_gonebad(): allows timecounter to be flagged as bad and removed atad
2008-04-28Remove clause 3 and 4 from TNF licensesmartin
2008-04-21Make ntp, pmc, reboot, sysarch, time syscalls MPSAFE.ad
2008-02-10Disable the event counters by default, since they create terrible cachead
2008-01-20Now that __HAVE_TIMECOUNTER and __HAVE_GENERIC_TODR are invariants,joerg
2008-01-05Serialize calls to tc_windup().ad
2008-01-03Add tc_detach() for detaching a timecounter. It is good enough todyoung
2007-12-15fix a comment.yamt
2007-11-30Use membar_*().ad
2007-11-23Kill another instance of KAUTH_GENERIC_ISSUSER.elad
2007-11-16Work around broken MD headers.ad
2007-11-15Kill the memory barriers: force users of the outgoing timehand to movead
2007-11-15Oops; fix thinko.ad
2007-11-15Add a bit of locking around timecounter attachment / selection.ad
2007-10-21Adjust a comment. In NetBSD the time-related kernel function prototypessimonb
2007-08-17Timecounters are lockless. Add conservative memory barriers to ensure thatad
2007-03-04Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.christos
2007-02-09Merge newlock2 to head.ad
2007-01-04Consistent usage of KAUTH_GENERIC_ISSUSER.elad
2006-11-01remove some __unused from function parameters.yamt
2006-10-21Avoid printf() the selected timecounter as it is called in the path of hardcl...kardel
2006-10-12- sprinkle __unused on function decls.christos
2006-09-24Reduce hardclock() code path length wrt/ multiply/divide bykardel