6c56727456
return through doreti to handle ast's. This is necessary for the clock interrupts to work properly. - Change the clock interrupts on the x86 to be fast instead of threaded. This is needed because both hardclock() and statclock() need to run in the context of the current process, not in a separate thread context. - Kill the prevproc hack as it is no longer needed. - We really need Giant when we call psignal(), but we don't want to block during the clock interrupt. Instead, use two p_flag's in the proc struct to mark the current process as having a pending SIGVTALRM or a SIGPROF and let them be delivered during ast() when hardclock() has finished running. - Remove CLKF_BASEPRI, which was #ifdef'd out on the x86 anyways. It was broken on the x86 if it was turned on since cpl is gone. It's only use was to bogusly run softclock() directly during hardclock() rather than scheduling an SWI. - Remove the COM_LOCK simplelock and replace it with a clock_lock spin mutex. Since the spin mutex already handles disabling/restoring interrupts appropriately, this also lets us axe all the *_intr() fu. - Back out the hacks in the APIC_IO x86 cpu_initclocks() code to use temporary fast interrupts for the APIC trial. - Add two new process flags P_ALRMPEND and P_PROFPEND to mark the pending signals in hardclock() that are to be delivered in ast(). Submitted by: jakeb (making statclock safe in a fast interrupt) Submitted by: cp (concept of delaying signals until ast())
433 lines
11 KiB
C
433 lines
11 KiB
C
/*-
|
|
* Copyright (c) 1982, 1986, 1991, 1993
|
|
* The Regents of the University of California. All rights reserved.
|
|
* (c) UNIX System Laboratories, Inc.
|
|
* All or some portions of this file are derived from material licensed
|
|
* to the University of California by American Telephone and Telegraph
|
|
* Co. or Unix System Laboratories, Inc. and are reproduced herein with
|
|
* the permission of UNIX System Laboratories, Inc.
|
|
*
|
|
* Redistribution and use in source and binary forms, with or without
|
|
* modification, are permitted provided that the following conditions
|
|
* are met:
|
|
* 1. Redistributions of source code must retain the above copyright
|
|
* notice, this list of conditions and the following disclaimer.
|
|
* 2. Redistributions in binary form must reproduce the above copyright
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
* documentation and/or other materials provided with the distribution.
|
|
* 3. All advertising materials mentioning features or use of this software
|
|
* must display the following acknowledgement:
|
|
* This product includes software developed by the University of
|
|
* California, Berkeley and its contributors.
|
|
* 4. Neither the name of the University nor the names of its contributors
|
|
* may be used to endorse or promote products derived from this software
|
|
* without specific prior written permission.
|
|
*
|
|
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
|
|
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
|
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
|
|
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
|
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
|
|
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
|
|
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
|
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
|
|
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
|
|
* SUCH DAMAGE.
|
|
*
|
|
* @(#)kern_clock.c 8.5 (Berkeley) 1/21/94
|
|
* $FreeBSD$
|
|
*/
|
|
|
|
#include "opt_ntp.h"
|
|
|
|
#include <sys/param.h>
|
|
#include <sys/systm.h>
|
|
#include <sys/dkstat.h>
|
|
#include <sys/callout.h>
|
|
#include <sys/kernel.h>
|
|
#include <sys/proc.h>
|
|
#include <sys/resourcevar.h>
|
|
#include <sys/signalvar.h>
|
|
#include <sys/timetc.h>
|
|
#include <sys/timepps.h>
|
|
#include <vm/vm.h>
|
|
#include <sys/lock.h>
|
|
#include <vm/pmap.h>
|
|
#include <vm/vm_map.h>
|
|
#include <sys/sysctl.h>
|
|
|
|
#include <machine/cpu.h>
|
|
#include <machine/ipl.h>
|
|
#include <machine/limits.h>
|
|
#include <machine/mutex.h>
|
|
#include <machine/smp.h>
|
|
|
|
#ifdef GPROF
|
|
#include <sys/gmon.h>
|
|
#endif
|
|
|
|
|
|
static void initclocks __P((void *dummy));
|
|
SYSINIT(clocks, SI_SUB_CLOCKS, SI_ORDER_FIRST, initclocks, NULL)
|
|
|
|
/* Some of these don't belong here, but it's easiest to concentrate them. */
|
|
long cp_time[CPUSTATES];
|
|
|
|
long tk_cancc;
|
|
long tk_nin;
|
|
long tk_nout;
|
|
long tk_rawcc;
|
|
|
|
/*
|
|
* Clock handling routines.
|
|
*
|
|
* This code is written to operate with two timers that run independently of
|
|
* each other.
|
|
*
|
|
* The main timer, running hz times per second, is used to trigger interval
|
|
* timers, timeouts and rescheduling as needed.
|
|
*
|
|
* The second timer handles kernel and user profiling,
|
|
* and does resource use estimation. If the second timer is programmable,
|
|
* it is randomized to avoid aliasing between the two clocks. For example,
|
|
* the randomization prevents an adversary from always giving up the cpu
|
|
* just before its quantum expires. Otherwise, it would never accumulate
|
|
* cpu ticks. The mean frequency of the second timer is stathz.
|
|
*
|
|
* If no second timer exists, stathz will be zero; in this case we drive
|
|
* profiling and statistics off the main clock. This WILL NOT be accurate;
|
|
* do not do it unless absolutely necessary.
|
|
*
|
|
* The statistics clock may (or may not) be run at a higher rate while
|
|
* profiling. This profile clock runs at profhz. We require that profhz
|
|
* be an integral multiple of stathz.
|
|
*
|
|
* If the statistics clock is running fast, it must be divided by the ratio
|
|
* profhz/stathz for statistics. (For profiling, every tick counts.)
|
|
*
|
|
* Time-of-day is maintained using a "timecounter", which may or may
|
|
* not be related to the hardware generating the above mentioned
|
|
* interrupts.
|
|
*/
|
|
|
|
int stathz;
|
|
int profhz;
|
|
static int profprocs;
|
|
int ticks;
|
|
static int psdiv, pscnt; /* prof => stat divider */
|
|
int psratio; /* ratio: prof / stat */
|
|
|
|
/*
|
|
* Initialize clock frequencies and start both clocks running.
|
|
*/
|
|
/* ARGSUSED*/
|
|
static void
|
|
initclocks(dummy)
|
|
void *dummy;
|
|
{
|
|
register int i;
|
|
|
|
/*
|
|
* Set divisors to 1 (normal case) and let the machine-specific
|
|
* code do its bit.
|
|
*/
|
|
psdiv = pscnt = 1;
|
|
cpu_initclocks();
|
|
|
|
/*
|
|
* Compute profhz/stathz, and fix profhz if needed.
|
|
*/
|
|
i = stathz ? stathz : hz;
|
|
if (profhz == 0)
|
|
profhz = i;
|
|
psratio = profhz / i;
|
|
}
|
|
|
|
/*
|
|
* The real-time timer, interrupting hz times per second.
|
|
*/
|
|
void
|
|
hardclock(frame)
|
|
register struct clockframe *frame;
|
|
{
|
|
register struct proc *p;
|
|
|
|
p = curproc;
|
|
if (p != idleproc) {
|
|
register struct pstats *pstats;
|
|
|
|
/*
|
|
* Run current process's virtual and profile time, as needed.
|
|
*/
|
|
pstats = p->p_stats;
|
|
if (CLKF_USERMODE(frame) &&
|
|
timevalisset(&pstats->p_timer[ITIMER_VIRTUAL].it_value) &&
|
|
itimerdecr(&pstats->p_timer[ITIMER_VIRTUAL], tick) == 0) {
|
|
p->p_flag |= P_ALRMPEND;
|
|
aston();
|
|
}
|
|
if (timevalisset(&pstats->p_timer[ITIMER_PROF].it_value) &&
|
|
itimerdecr(&pstats->p_timer[ITIMER_PROF], tick) == 0) {
|
|
p->p_flag |= P_PROFPEND;
|
|
aston();
|
|
}
|
|
}
|
|
|
|
#if defined(SMP) && defined(BETTER_CLOCK)
|
|
forward_hardclock(pscnt);
|
|
#endif
|
|
|
|
/*
|
|
* If no separate statistics clock is available, run it from here.
|
|
*/
|
|
if (stathz == 0)
|
|
statclock(frame);
|
|
|
|
tc_windup();
|
|
ticks++;
|
|
|
|
/*
|
|
* Process callouts at a very low cpu priority, so we don't keep the
|
|
* relatively high clock interrupt priority any longer than necessary.
|
|
*/
|
|
if (TAILQ_FIRST(&callwheel[ticks & callwheelmask]) != NULL) {
|
|
setsoftclock();
|
|
} else if (softticks + 1 == ticks)
|
|
++softticks;
|
|
}
|
|
|
|
/*
|
|
* Compute number of ticks in the specified amount of time.
|
|
*/
|
|
int
|
|
tvtohz(tv)
|
|
struct timeval *tv;
|
|
{
|
|
register unsigned long ticks;
|
|
register long sec, usec;
|
|
|
|
/*
|
|
* If the number of usecs in the whole seconds part of the time
|
|
* difference fits in a long, then the total number of usecs will
|
|
* fit in an unsigned long. Compute the total and convert it to
|
|
* ticks, rounding up and adding 1 to allow for the current tick
|
|
* to expire. Rounding also depends on unsigned long arithmetic
|
|
* to avoid overflow.
|
|
*
|
|
* Otherwise, if the number of ticks in the whole seconds part of
|
|
* the time difference fits in a long, then convert the parts to
|
|
* ticks separately and add, using similar rounding methods and
|
|
* overflow avoidance. This method would work in the previous
|
|
* case but it is slightly slower and assumes that hz is integral.
|
|
*
|
|
* Otherwise, round the time difference down to the maximum
|
|
* representable value.
|
|
*
|
|
* If ints have 32 bits, then the maximum value for any timeout in
|
|
* 10ms ticks is 248 days.
|
|
*/
|
|
sec = tv->tv_sec;
|
|
usec = tv->tv_usec;
|
|
if (usec < 0) {
|
|
sec--;
|
|
usec += 1000000;
|
|
}
|
|
if (sec < 0) {
|
|
#ifdef DIAGNOSTIC
|
|
if (usec > 0) {
|
|
sec++;
|
|
usec -= 1000000;
|
|
}
|
|
printf("tvotohz: negative time difference %ld sec %ld usec\n",
|
|
sec, usec);
|
|
#endif
|
|
ticks = 1;
|
|
} else if (sec <= LONG_MAX / 1000000)
|
|
ticks = (sec * 1000000 + (unsigned long)usec + (tick - 1))
|
|
/ tick + 1;
|
|
else if (sec <= LONG_MAX / hz)
|
|
ticks = sec * hz
|
|
+ ((unsigned long)usec + (tick - 1)) / tick + 1;
|
|
else
|
|
ticks = LONG_MAX;
|
|
if (ticks > INT_MAX)
|
|
ticks = INT_MAX;
|
|
return ((int)ticks);
|
|
}
|
|
|
|
/*
|
|
* Start profiling on a process.
|
|
*
|
|
* Kernel profiling passes proc0 which never exits and hence
|
|
* keeps the profile clock running constantly.
|
|
*/
|
|
void
|
|
startprofclock(p)
|
|
register struct proc *p;
|
|
{
|
|
int s;
|
|
|
|
if ((p->p_flag & P_PROFIL) == 0) {
|
|
p->p_flag |= P_PROFIL;
|
|
if (++profprocs == 1 && stathz != 0) {
|
|
s = splstatclock();
|
|
psdiv = pscnt = psratio;
|
|
setstatclockrate(profhz);
|
|
splx(s);
|
|
}
|
|
}
|
|
}
|
|
|
|
/*
|
|
* Stop profiling on a process.
|
|
*/
|
|
void
|
|
stopprofclock(p)
|
|
register struct proc *p;
|
|
{
|
|
int s;
|
|
|
|
if (p->p_flag & P_PROFIL) {
|
|
p->p_flag &= ~P_PROFIL;
|
|
if (--profprocs == 0 && stathz != 0) {
|
|
s = splstatclock();
|
|
psdiv = pscnt = 1;
|
|
setstatclockrate(stathz);
|
|
splx(s);
|
|
}
|
|
}
|
|
}
|
|
|
|
/*
|
|
* Statistics clock. Grab profile sample, and if divider reaches 0,
|
|
* do process and kernel statistics. Most of the statistics are only
|
|
* used by user-level statistics programs. The main exceptions are
|
|
* p->p_uticks, p->p_sticks, p->p_iticks, and p->p_estcpu.
|
|
*/
|
|
void
|
|
statclock(frame)
|
|
register struct clockframe *frame;
|
|
{
|
|
#ifdef GPROF
|
|
register struct gmonparam *g;
|
|
int i;
|
|
#endif
|
|
register struct proc *p;
|
|
struct pstats *pstats;
|
|
long rss;
|
|
struct rusage *ru;
|
|
struct vmspace *vm;
|
|
|
|
mtx_enter(&sched_lock, MTX_SPIN);
|
|
|
|
if (CLKF_USERMODE(frame)) {
|
|
/*
|
|
* Came from user mode; CPU was in user state.
|
|
* If this process is being profiled, record the tick.
|
|
*/
|
|
p = curproc;
|
|
if (p->p_flag & P_PROFIL)
|
|
addupc_intr(p, CLKF_PC(frame), 1);
|
|
#if defined(SMP) && defined(BETTER_CLOCK)
|
|
if (stathz != 0)
|
|
forward_statclock(pscnt);
|
|
#endif
|
|
if (--pscnt > 0) {
|
|
mtx_exit(&sched_lock, MTX_SPIN);
|
|
return;
|
|
}
|
|
/*
|
|
* Charge the time as appropriate.
|
|
*/
|
|
p->p_uticks++;
|
|
if (p->p_nice > NZERO)
|
|
cp_time[CP_NICE]++;
|
|
else
|
|
cp_time[CP_USER]++;
|
|
} else {
|
|
#ifdef GPROF
|
|
/*
|
|
* Kernel statistics are just like addupc_intr, only easier.
|
|
*/
|
|
g = &_gmonparam;
|
|
if (g->state == GMON_PROF_ON) {
|
|
i = CLKF_PC(frame) - g->lowpc;
|
|
if (i < g->textsize) {
|
|
i /= HISTFRACTION * sizeof(*g->kcount);
|
|
g->kcount[i]++;
|
|
}
|
|
}
|
|
#endif
|
|
#if defined(SMP) && defined(BETTER_CLOCK)
|
|
if (stathz != 0)
|
|
forward_statclock(pscnt);
|
|
#endif
|
|
if (--pscnt > 0) {
|
|
mtx_exit(&sched_lock, MTX_SPIN);
|
|
return;
|
|
}
|
|
/*
|
|
* Came from kernel mode, so we were:
|
|
* - handling an interrupt,
|
|
* - doing syscall or trap work on behalf of the current
|
|
* user process, or
|
|
* - spinning in the idle loop.
|
|
* Whichever it is, charge the time as appropriate.
|
|
* Note that we charge interrupts to the current process,
|
|
* regardless of whether they are ``for'' that process,
|
|
* so that we know how much of its real time was spent
|
|
* in ``non-process'' (i.e., interrupt) work.
|
|
*/
|
|
p = curproc;
|
|
if ((p->p_ithd != NULL) || CLKF_INTR(frame)) {
|
|
p->p_iticks++;
|
|
cp_time[CP_INTR]++;
|
|
} else {
|
|
p->p_sticks++;
|
|
if (p != idleproc)
|
|
cp_time[CP_SYS]++;
|
|
else
|
|
cp_time[CP_IDLE]++;
|
|
}
|
|
}
|
|
pscnt = psdiv;
|
|
|
|
schedclock(p);
|
|
|
|
/* Update resource usage integrals and maximums. */
|
|
if ((pstats = p->p_stats) != NULL &&
|
|
(ru = &pstats->p_ru) != NULL &&
|
|
(vm = p->p_vmspace) != NULL) {
|
|
ru->ru_ixrss += pgtok(vm->vm_tsize);
|
|
ru->ru_idrss += pgtok(vm->vm_dsize);
|
|
ru->ru_isrss += pgtok(vm->vm_ssize);
|
|
rss = pgtok(vmspace_resident_count(vm));
|
|
if (ru->ru_maxrss < rss)
|
|
ru->ru_maxrss = rss;
|
|
}
|
|
|
|
mtx_exit(&sched_lock, MTX_SPIN);
|
|
}
|
|
|
|
/*
|
|
* Return information about system clocks.
|
|
*/
|
|
static int
|
|
sysctl_kern_clockrate(SYSCTL_HANDLER_ARGS)
|
|
{
|
|
struct clockinfo clkinfo;
|
|
/*
|
|
* Construct clockinfo structure.
|
|
*/
|
|
clkinfo.hz = hz;
|
|
clkinfo.tick = tick;
|
|
clkinfo.tickadj = tickadj;
|
|
clkinfo.profhz = profhz;
|
|
clkinfo.stathz = stathz ? stathz : hz;
|
|
return (sysctl_handle_opaque(oidp, &clkinfo, sizeof clkinfo, req));
|
|
}
|
|
|
|
SYSCTL_PROC(_kern, KERN_CLOCKRATE, clockrate, CTLTYPE_STRUCT|CTLFLAG_RD,
|
|
0, 0, sysctl_kern_clockrate, "S,clockinfo","");
|