Add support for Xen time abstractions. To avoid expensive traps into
the hypervisor, the passage of time is extrapolated from the local TSC
and a set of timestamps and scaling factors exported to the guest via
shared memory. Xen also provides a periodic interrupt facility which
is used to drive updates of xtime and jiffies, and perform the usual
process accounting and profiling.
Signed-off-by: Ian Pratt <[email protected]>
Signed-off-by: Christian Limpach <[email protected]>
Signed-off-by: Chris Wright <[email protected]>
---
arch/i386/kernel/Makefile | 6
arch/i386/mach-xen/time.c | 1045 ++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 1050 insertions(+), 1 deletion(-)
--- xen-subarch-2.6.orig/arch/i386/kernel/Makefile
+++ xen-subarch-2.6/arch/i386/kernel/Makefile
@@ -9,8 +9,11 @@ obj-y := process.o semaphore.o signal.o
pci-dma.o i386_ksyms.o i387.o dmi_scan.o bootflag.o \
quirks.o i8237.o topology.o
+timers-y := timers/
+timers-$(CONFIG_XEN) :=
+
obj-y += cpu/
-obj-y += timers/
+obj-y += $(timers-y)
obj-y += acpi/
obj-$(CONFIG_X86_BIOS_REBOOT) += reboot.o
obj-$(CONFIG_MCA) += mca.o
@@ -45,6 +48,7 @@ obj-$(CONFIG_SCx200) += scx200.o
hw_irq-y := i8259.o
hw_irq-$(CONFIG_XEN) := ../mach-xen/evtchn.o
+time-$(CONFIG_XEN) := ../mach-xen/time.o
# vsyscall.o contains the vsyscall DSO images as __initdata.
# We must build both images before we can assemble it.
--- /dev/null
+++ xen-subarch-2.6/arch/i386/mach-xen/time.c
@@ -0,0 +1,1045 @@
+/*
+ * linux/arch/i386/kernel/time.c
+ *
+ * Copyright (C) 1991, 1992, 1995 Linus Torvalds
+ *
+ * This file contains the PC-specific time handling details:
+ * reading the RTC at bootup, etc..
+ * 1994-07-02 Alan Modra
+ * fixed set_rtc_mmss, fixed time.year for >= 2000, new mktime
+ * 1995-03-26 Markus Kuhn
+ * fixed 500 ms bug at call to set_rtc_mmss, fixed DS12887
+ * precision CMOS clock update
+ * 1996-05-03 Ingo Molnar
+ * fixed time warps in do_[slow|fast]_gettimeoffset()
+ * 1997-09-10 Updated NTP code according to technical memorandum Jan '96
+ * "A Kernel Model for Precision Timekeeping" by Dave Mills
+ * 1998-09-05 (Various)
+ * More robust do_fast_gettimeoffset() algorithm implemented
+ * (works with APM, Cyrix 6x86MX and Centaur C6),
+ * monotonic gettimeofday() with fast_get_timeoffset(),
+ * drift-proof precision TSC calibration on boot
+ * (C. Scott Ananian <[email protected]>, Andrew D.
+ * Balsa <[email protected]>, Philip Gladstone <[email protected]>;
+ * ported from 2.0.35 Jumbo-9 by Michael Krause <[email protected]>).
+ * 1998-12-16 Andrea Arcangeli
+ * Fixed Jumbo-9 code in 2.1.131: do_gettimeofday was missing 1 jiffy
+ * because was not accounting lost_ticks.
+ * 1998-12-24 Copyright (C) 1998 Andrea Arcangeli
+ * Fixed a xtime SMP race (we need the xtime_lock rw spinlock to
+ * serialize accesses to xtime/lost_ticks).
+ */
+
+#include <linux/errno.h>
+#include <linux/sched.h>
+#include <linux/kernel.h>
+#include <linux/param.h>
+#include <linux/string.h>
+#include <linux/mm.h>
+#include <linux/interrupt.h>
+#include <linux/time.h>
+#include <linux/delay.h>
+#include <linux/init.h>
+#include <linux/smp.h>
+#include <linux/module.h>
+#include <linux/sysdev.h>
+#include <linux/bcd.h>
+#include <linux/efi.h>
+#include <linux/mca.h>
+#include <linux/sysctl.h>
+#include <linux/percpu.h>
+#include <linux/kernel_stat.h>
+#include <linux/posix-timers.h>
+
+#include <asm/io.h>
+#include <asm/smp.h>
+#include <asm/irq.h>
+#include <asm/msr.h>
+#include <asm/delay.h>
+#include <asm/mpspec.h>
+#include <asm/uaccess.h>
+#include <asm/processor.h>
+#include <asm/timer.h>
+#include <asm/sections.h>
+
+#include "mach_time.h"
+
+#include <linux/timex.h>
+#include <linux/config.h>
+
+#include <asm/hpet.h>
+
+#include <asm/arch_hooks.h>
+
+#include <xen/evtchn.h>
+#include <xen/interface/vcpu.h>
+
+#if defined (__i386__)
+#include <asm/i8259.h>
+#endif
+
+int pit_latch_buggy; /* extern */
+
+#if defined(__x86_64__)
+unsigned long vxtime_hz = PIT_TICK_RATE;
+struct vxtime_data __vxtime __section_vxtime; /* for vsyscalls */
+volatile unsigned long __jiffies __section_jiffies = INITIAL_JIFFIES;
+unsigned long __wall_jiffies __section_wall_jiffies = INITIAL_JIFFIES;
+struct timespec __xtime __section_xtime;
+struct timezone __sys_tz __section_sys_tz;
+#endif
+
+unsigned int cpu_khz; /* Detected as we calibrate the TSC */
+EXPORT_SYMBOL(cpu_khz);
+
+extern unsigned long wall_jiffies;
+
+DEFINE_SPINLOCK(rtc_lock);
+EXPORT_SYMBOL(rtc_lock);
+
+#if defined (__i386__)
+#include <asm/i8253.h>
+#endif
+
+DEFINE_SPINLOCK(i8253_lock);
+EXPORT_SYMBOL(i8253_lock);
+
+extern struct init_timer_opts timer_tsc_init;
+extern struct timer_opts timer_tsc;
+#define timer_none timer_tsc
+struct timer_opts *cur_timer __read_mostly = &timer_tsc;
+
+/* These are peridically updated in shared_info, and then copied here. */
+struct shadow_time_info {
+ u64 tsc_timestamp; /* TSC at last update of time vals. */
+ u64 system_timestamp; /* Time, in nanosecs, since boot. */
+ u32 tsc_to_nsec_mul;
+ u32 tsc_to_usec_mul;
+ int tsc_shift;
+ u32 version;
+};
+static DEFINE_PER_CPU(struct shadow_time_info, shadow_time);
+static struct timespec shadow_tv;
+static u32 shadow_tv_version;
+
+/* Keep track of last time we did processing/updating of jiffies and xtime. */
+static u64 processed_system_time; /* System time (ns) at last processing. */
+static DEFINE_PER_CPU(u64, processed_system_time);
+
+/* How much CPU time was spent blocked and how much was 'stolen'? */
+static DEFINE_PER_CPU(u64, processed_stolen_time);
+static DEFINE_PER_CPU(u64, processed_blocked_time);
+
+/* Current runstate of each CPU (updated automatically by the hypervisor). */
+static DEFINE_PER_CPU(struct vcpu_runstate_info, runstate);
+
+/* Must be signed, as it's compared with s64 quantities which can be -ve. */
+#define NS_PER_TICK (1000000000LL/HZ)
+
+static inline void __normalize_time(time_t *sec, s64 *nsec)
+{
+ while (*nsec >= NSEC_PER_SEC) {
+ (*nsec) -= NSEC_PER_SEC;
+ (*sec)++;
+ }
+ while (*nsec < 0) {
+ (*nsec) += NSEC_PER_SEC;
+ (*sec)--;
+ }
+}
+
+/* Does this guest OS track Xen time, or set its wall clock independently? */
+static int independent_wallclock = 0;
+static int __init __independent_wallclock(char *str)
+{
+ independent_wallclock = 1;
+ return 1;
+}
+__setup("independent_wallclock", __independent_wallclock);
+
+/* Permitted clock jitter, in usecs, beyond which a warning will be printed. */
+static unsigned long permitted_clock_jitter = 10000UL;
+static int __init __permitted_clock_jitter(char *str)
+{
+ permitted_clock_jitter = simple_strtoul(str, NULL, 0);
+ return 1;
+}
+__setup("permitted_clock_jitter=", __permitted_clock_jitter);
+
+int tsc_disable __devinitdata = 0;
+
+static void delay_tsc(unsigned long loops)
+{
+ unsigned long bclock, now;
+
+ rdtscl(bclock);
+ do {
+ rep_nop();
+ rdtscl(now);
+ } while ((now - bclock) < loops);
+}
+
+struct timer_opts timer_tsc = {
+ .name = "tsc",
+ .delay = delay_tsc,
+};
+
+/*
+ * Scale a 64-bit delta by scaling and multiplying by a 32-bit fraction,
+ * yielding a 64-bit result.
+ */
+static inline u64 scale_delta(u64 delta, u32 mul_frac, int shift)
+{
+ u64 product;
+#ifdef __i386__
+ u32 tmp1, tmp2;
+#endif
+
+ if (shift < 0)
+ delta >>= -shift;
+ else
+ delta <<= shift;
+
+#ifdef __i386__
+ __asm__ (
+ "mul %5 ; "
+ "mov %4,%%eax ; "
+ "mov %%edx,%4 ; "
+ "mul %5 ; "
+ "xor %5,%5 ; "
+ "add %4,%%eax ; "
+ "adc %5,%%edx ; "
+ : "=A" (product), "=r" (tmp1), "=r" (tmp2)
+ : "a" ((u32)delta), "1" ((u32)(delta >> 32)), "2" (mul_frac) );
+#else
+ __asm__ (
+ "mul %%rdx ; shrd $32,%%rdx,%%rax"
+ : "=a" (product) : "0" (delta), "d" ((u64)mul_frac) );
+#endif
+
+ return product;
+}
+
+#if defined (__i386__)
+int read_current_timer(unsigned long *timer_val)
+{
+ rdtscl(*timer_val);
+ return 0;
+}
+#endif
+
+void init_cpu_khz(void)
+{
+ u64 __cpu_khz = 1000000ULL << 32;
+ struct vcpu_time_info *info;
+ info = &HYPERVISOR_shared_info->vcpu_info[0].time;
+ do_div(__cpu_khz, info->tsc_to_system_mul);
+ if (info->tsc_shift < 0)
+ cpu_khz = __cpu_khz << -info->tsc_shift;
+ else
+ cpu_khz = __cpu_khz >> info->tsc_shift;
+}
+
+static u64 get_nsec_offset(struct shadow_time_info *shadow)
+{
+ u64 now, delta;
+ rdtscll(now);
+ delta = now - shadow->tsc_timestamp;
+ return scale_delta(delta, shadow->tsc_to_nsec_mul, shadow->tsc_shift);
+}
+
+static unsigned long get_usec_offset(struct shadow_time_info *shadow)
+{
+ u64 now, delta;
+ rdtscll(now);
+ delta = now - shadow->tsc_timestamp;
+ return scale_delta(delta, shadow->tsc_to_usec_mul, shadow->tsc_shift);
+}
+
+static void __update_wallclock(time_t sec, long nsec)
+{
+ long wtm_nsec, xtime_nsec;
+ time_t wtm_sec, xtime_sec;
+ u64 tmp, wc_nsec;
+
+ /* Adjust wall-clock time base based on wall_jiffies ticks. */
+ wc_nsec = processed_system_time;
+ wc_nsec += sec * (u64)NSEC_PER_SEC;
+ wc_nsec += nsec;
+ wc_nsec -= (jiffies - wall_jiffies) * (u64)NS_PER_TICK;
+
+ /* Split wallclock base into seconds and nanoseconds. */
+ tmp = wc_nsec;
+ xtime_nsec = do_div(tmp, 1000000000);
+ xtime_sec = (time_t)tmp;
+
+ wtm_sec = wall_to_monotonic.tv_sec + (xtime.tv_sec - xtime_sec);
+ wtm_nsec = wall_to_monotonic.tv_nsec + (xtime.tv_nsec - xtime_nsec);
+
+ set_normalized_timespec(&xtime, xtime_sec, xtime_nsec);
+ set_normalized_timespec(&wall_to_monotonic, wtm_sec, wtm_nsec);
+
+ ntp_clear();
+}
+
+static void update_wallclock(void)
+{
+ struct shared_info *s = HYPERVISOR_shared_info;
+
+ do {
+ shadow_tv_version = s->wc_version;
+ rmb();
+ shadow_tv.tv_sec = s->wc_sec;
+ shadow_tv.tv_nsec = s->wc_nsec;
+ rmb();
+ } while ((s->wc_version & 1) | (shadow_tv_version ^ s->wc_version));
+
+ if (!independent_wallclock)
+ __update_wallclock(shadow_tv.tv_sec, shadow_tv.tv_nsec);
+}
+
+/*
+ * Reads a consistent set of time-base values from Xen, into a shadow data
+ * area.
+ */
+static void get_time_values_from_xen(void)
+{
+ struct shared_info *s = HYPERVISOR_shared_info;
+ struct vcpu_time_info *src;
+ struct shadow_time_info *dst;
+
+ src = &s->vcpu_info[smp_processor_id()].time;
+ dst = &per_cpu(shadow_time, smp_processor_id());
+
+ do {
+ dst->version = src->version;
+ rmb();
+ dst->tsc_timestamp = src->tsc_timestamp;
+ dst->system_timestamp = src->system_time;
+ dst->tsc_to_nsec_mul = src->tsc_to_system_mul;
+ dst->tsc_shift = src->tsc_shift;
+ rmb();
+ } while ((src->version & 1) | (dst->version ^ src->version));
+
+ dst->tsc_to_usec_mul = dst->tsc_to_nsec_mul / 1000;
+}
+
+static inline int time_values_up_to_date(int cpu)
+{
+ struct vcpu_time_info *src;
+ struct shadow_time_info *dst;
+
+ src = &HYPERVISOR_shared_info->vcpu_info[cpu].time;
+ dst = &per_cpu(shadow_time, cpu);
+
+ rmb();
+ return (dst->version == src->version);
+}
+
+/*
+ * This is a special lock that is owned by the CPU and holds the index
+ * register we are working with. It is required for NMI access to the
+ * CMOS/RTC registers. See include/asm-i386/mc146818rtc.h for details.
+ */
+volatile unsigned long cmos_lock = 0;
+EXPORT_SYMBOL(cmos_lock);
+
+/* Routines for accessing the CMOS RAM/RTC. */
+unsigned char rtc_cmos_read(unsigned char addr)
+{
+ unsigned char val;
+ lock_cmos_prefix(addr);
+ outb_p(addr, RTC_PORT(0));
+ val = inb_p(RTC_PORT(1));
+ lock_cmos_suffix(addr);
+ return val;
+}
+EXPORT_SYMBOL(rtc_cmos_read);
+
+void rtc_cmos_write(unsigned char val, unsigned char addr)
+{
+ lock_cmos_prefix(addr);
+ outb_p(addr, RTC_PORT(0));
+ outb_p(val, RTC_PORT(1));
+ lock_cmos_suffix(addr);
+}
+EXPORT_SYMBOL(rtc_cmos_write);
+
+/*
+ * This version of gettimeofday has microsecond resolution
+ * and better than microsecond precision on fast x86 machines with TSC.
+ */
+void do_gettimeofday(struct timeval *tv)
+{
+ unsigned long seq;
+ unsigned long usec, sec;
+ unsigned long max_ntp_tick;
+ s64 nsec;
+ unsigned int cpu;
+ struct shadow_time_info *shadow;
+ u32 local_time_version;
+
+ cpu = get_cpu();
+ shadow = &per_cpu(shadow_time, cpu);
+
+ do {
+ unsigned long lost;
+
+ local_time_version = shadow->version;
+ seq = read_seqbegin(&xtime_lock);
+
+ usec = get_usec_offset(shadow);
+ lost = jiffies - wall_jiffies;
+
+ /*
+ * If time_adjust is negative then NTP is slowing the clock
+ * so make sure not to go into next possible interval.
+ * Better to lose some accuracy than have time go backwards..
+ */
+ if (unlikely(time_adjust < 0)) {
+ max_ntp_tick = (USEC_PER_SEC / HZ) - tickadj;
+ usec = min(usec, max_ntp_tick);
+
+ if (lost)
+ usec += lost * max_ntp_tick;
+ }
+ else if (unlikely(lost))
+ usec += lost * (USEC_PER_SEC / HZ);
+
+ sec = xtime.tv_sec;
+ usec += (xtime.tv_nsec / NSEC_PER_USEC);
+
+ nsec = shadow->system_timestamp - processed_system_time;
+ __normalize_time(&sec, &nsec);
+ usec += (long)nsec / NSEC_PER_USEC;
+
+ if (unlikely(!time_values_up_to_date(cpu))) {
+ /*
+ * We may have blocked for a long time,
+ * rendering our calculations invalid
+ * (e.g. the time delta may have
+ * overflowed). Detect that and recalculate
+ * with fresh values.
+ */
+ get_time_values_from_xen();
+ continue;
+ }
+ } while (read_seqretry(&xtime_lock, seq) ||
+ (local_time_version != shadow->version));
+
+ put_cpu();
+
+ while (usec >= USEC_PER_SEC) {
+ usec -= USEC_PER_SEC;
+ sec++;
+ }
+
+ tv->tv_sec = sec;
+ tv->tv_usec = usec;
+}
+
+EXPORT_SYMBOL(do_gettimeofday);
+
+int do_settimeofday(struct timespec *tv)
+{
+ time_t sec;
+ s64 nsec;
+ unsigned int cpu;
+ struct shadow_time_info *shadow;
+
+ if ((unsigned long)tv->tv_nsec >= NSEC_PER_SEC)
+ return -EINVAL;
+
+ cpu = get_cpu();
+ shadow = &per_cpu(shadow_time, cpu);
+
+ write_seqlock_irq(&xtime_lock);
+
+ /*
+ * Ensure we don't get blocked for a long time so that our time delta
+ * overflows. If that were to happen then our shadow time values would
+ * be stale, so we can retry with fresh ones.
+ */
+ for (;;) {
+ nsec = tv->tv_nsec - get_nsec_offset(shadow);
+ if (time_values_up_to_date(cpu))
+ break;
+ get_time_values_from_xen();
+ }
+ sec = tv->tv_sec;
+ __normalize_time(&sec, &nsec);
+
+ if (independent_wallclock) {
+ nsec -= shadow->system_timestamp;
+ __normalize_time(&sec, &nsec);
+ __update_wallclock(sec, nsec);
+ }
+
+ write_sequnlock_irq(&xtime_lock);
+
+ put_cpu();
+
+ clock_was_set();
+ return 0;
+}
+
+EXPORT_SYMBOL(do_settimeofday);
+
+static int set_rtc_mmss(unsigned long nowtime)
+{
+ int retval;
+
+ WARN_ON(irqs_disabled());
+
+ if (independent_wallclock || !(xen_start_info->flags & SIF_INITDOMAIN))
+ return 0;
+
+ /* gets recalled with irq locally disabled */
+ spin_lock_irq(&rtc_lock);
+ if (efi_enabled)
+ retval = efi_set_rtc_mmss(nowtime);
+ else
+ retval = mach_set_rtc_mmss(nowtime);
+ spin_unlock_irq(&rtc_lock);
+
+ return retval;
+}
+
+/* monotonic_clock(): returns # of nanoseconds passed since time_init()
+ * Note: This function is required to return accurate
+ * time even in the absence of multiple timer ticks.
+ */
+unsigned long long monotonic_clock(void)
+{
+ int cpu = get_cpu();
+ struct shadow_time_info *shadow = &per_cpu(shadow_time, cpu);
+ u64 time;
+ u32 local_time_version;
+
+ do {
+ local_time_version = shadow->version;
+ barrier();
+ time = shadow->system_timestamp + get_nsec_offset(shadow);
+ if (!time_values_up_to_date(cpu))
+ get_time_values_from_xen();
+ barrier();
+ } while (local_time_version != shadow->version);
+
+ put_cpu();
+
+ return time;
+}
+EXPORT_SYMBOL(monotonic_clock);
+
+unsigned long long sched_clock(void)
+{
+ return monotonic_clock();
+}
+
+#if defined(CONFIG_SMP) && defined(CONFIG_FRAME_POINTER)
+unsigned long profile_pc(struct pt_regs *regs)
+{
+ unsigned long pc = instruction_pointer(regs);
+
+#ifdef __x86_64__
+ /* Assume the lock function has either no stack frame or only a single word.
+ This checks if the address on the stack looks like a kernel text address.
+ There is a small window for false hits, but in that case the tick
+ is just accounted to the spinlock function.
+ Better would be to write these functions in assembler again
+ and check exactly. */
+ if (in_lock_functions(pc)) {
+ char *v = *(char **)regs->rsp;
+ if ((v >= _stext && v <= _etext) ||
+ (v >= _sinittext && v <= _einittext) ||
+ (v >= (char *)MODULES_VADDR && v <= (char *)MODULES_END))
+ return (unsigned long)v;
+ return ((unsigned long *)regs->rsp)[1];
+ }
+#else
+ if (in_lock_functions(pc))
+ return *(unsigned long *)(regs->ebp + 4);
+#endif
+
+ return pc;
+}
+EXPORT_SYMBOL(profile_pc);
+#endif
+
+irqreturn_t timer_interrupt(int irq, void *dev_id, struct pt_regs *regs)
+{
+ s64 delta, delta_cpu, stolen, blocked;
+ u64 sched_time;
+ int i, cpu = smp_processor_id();
+ struct shadow_time_info *shadow = &per_cpu(shadow_time, cpu);
+ struct vcpu_runstate_info *runstate = &per_cpu(runstate, cpu);
+
+ write_seqlock(&xtime_lock);
+
+ do {
+ get_time_values_from_xen();
+
+ /* Obtain a consistent snapshot of elapsed wallclock cycles. */
+ delta = delta_cpu =
+ shadow->system_timestamp + get_nsec_offset(shadow);
+ delta -= processed_system_time;
+ delta_cpu -= per_cpu(processed_system_time, cpu);
+
+ /*
+ * Obtain a consistent snapshot of stolen/blocked cycles. We
+ * can use state_entry_time to detect if we get preempted here.
+ */
+ do {
+ sched_time = runstate->state_entry_time;
+ barrier();
+ stolen = runstate->time[RUNSTATE_runnable] +
+ runstate->time[RUNSTATE_offline] -
+ per_cpu(processed_stolen_time, cpu);
+ blocked = runstate->time[RUNSTATE_blocked] -
+ per_cpu(processed_blocked_time, cpu);
+ barrier();
+ } while (sched_time != runstate->state_entry_time);
+ } while (!time_values_up_to_date(cpu));
+
+ if ((unlikely(delta < -(s64)permitted_clock_jitter) ||
+ unlikely(delta_cpu < -(s64)permitted_clock_jitter))
+ && printk_ratelimit()) {
+ printk("Timer ISR/%d: Time went backwards: "
+ "delta=%lld delta_cpu=%lld shadow=%lld "
+ "off=%lld processed=%lld cpu_processed=%lld\n",
+ cpu, delta, delta_cpu, shadow->system_timestamp,
+ (s64)get_nsec_offset(shadow),
+ processed_system_time,
+ per_cpu(processed_system_time, cpu));
+ for (i = 0; i < num_online_cpus(); i++)
+ printk(" %d: %lld\n", i,
+ per_cpu(processed_system_time, i));
+ }
+
+ /* System-wide jiffy work. */
+ while (delta >= NS_PER_TICK) {
+ delta -= NS_PER_TICK;
+ processed_system_time += NS_PER_TICK;
+ do_timer(regs);
+ }
+
+ if (shadow_tv_version != HYPERVISOR_shared_info->wc_version) {
+ update_wallclock();
+ clock_was_set();
+ }
+
+ write_sequnlock(&xtime_lock);
+
+ /*
+ * Account stolen ticks.
+ * HACK: Passing NULL to account_steal_time()
+ * ensures that the ticks are accounted as stolen.
+ */
+ if ((stolen > 0) && (delta_cpu > 0)) {
+ delta_cpu -= stolen;
+ if (unlikely(delta_cpu < 0))
+ stolen += delta_cpu; /* clamp local-time progress */
+ do_div(stolen, NS_PER_TICK);
+ per_cpu(processed_stolen_time, cpu) += stolen * NS_PER_TICK;
+ per_cpu(processed_system_time, cpu) += stolen * NS_PER_TICK;
+ account_steal_time(NULL, (cputime_t)stolen);
+ }
+
+ /*
+ * Account blocked ticks.
+ * HACK: Passing idle_task to account_steal_time()
+ * ensures that the ticks are accounted as idle/wait.
+ */
+ if ((blocked > 0) && (delta_cpu > 0)) {
+ delta_cpu -= blocked;
+ if (unlikely(delta_cpu < 0))
+ blocked += delta_cpu; /* clamp local-time progress */
+ do_div(blocked, NS_PER_TICK);
+ per_cpu(processed_blocked_time, cpu) += blocked * NS_PER_TICK;
+ per_cpu(processed_system_time, cpu) += blocked * NS_PER_TICK;
+ account_steal_time(idle_task(cpu), (cputime_t)blocked);
+ }
+
+ /* Account user/system ticks. */
+ if (delta_cpu > 0) {
+ do_div(delta_cpu, NS_PER_TICK);
+ per_cpu(processed_system_time, cpu) += delta_cpu * NS_PER_TICK;
+ if (user_mode(regs))
+ account_user_time(current, (cputime_t)delta_cpu);
+ else
+ account_system_time(current, HARDIRQ_OFFSET,
+ (cputime_t)delta_cpu);
+ }
+
+ /* Local timer processing (see update_process_times()). */
+ run_local_timers();
+ if (rcu_pending(cpu))
+ rcu_check_callbacks(cpu, user_mode(regs));
+ scheduler_tick();
+ run_posix_cpu_timers(current);
+
+ return IRQ_HANDLED;
+}
+
+static void init_missing_ticks_accounting(int cpu)
+{
+ struct vcpu_register_runstate_memory_area area;
+ struct vcpu_runstate_info *runstate = &per_cpu(runstate, cpu);
+
+ memset(runstate, 0, sizeof(*runstate));
+
+ area.addr.v = runstate;
+ HYPERVISOR_vcpu_op(VCPUOP_register_runstate_memory_area, cpu, &area);
+
+ per_cpu(processed_blocked_time, cpu) =
+ runstate->time[RUNSTATE_blocked];
+ per_cpu(processed_stolen_time, cpu) =
+ runstate->time[RUNSTATE_runnable] +
+ runstate->time[RUNSTATE_offline];
+}
+
+/* not static: needed by APM */
+unsigned long get_cmos_time(void)
+{
+ unsigned long retval;
+
+ spin_lock(&rtc_lock);
+
+ if (efi_enabled)
+ retval = efi_get_time();
+ else
+ retval = mach_get_cmos_time();
+
+ spin_unlock(&rtc_lock);
+
+ return retval;
+}
+EXPORT_SYMBOL(get_cmos_time);
+
+static void sync_cmos_clock(unsigned long dummy);
+
+static DEFINE_TIMER(sync_cmos_timer, sync_cmos_clock, 0, 0);
+
+static void sync_cmos_clock(unsigned long dummy)
+{
+ struct timeval now, next;
+ int fail = 1;
+
+ /*
+ * If we have an externally synchronized Linux clock, then update
+ * CMOS clock accordingly every ~11 minutes. Set_rtc_mmss() has to be
+ * called as close as possible to 500 ms before the new second starts.
+ * This code is run on a timer. If the clock is set, that timer
+ * may not expire at the correct time. Thus, we adjust...
+ */
+ if (!ntp_synced())
+ /*
+ * Not synced, exit, do not restart a timer (if one is
+ * running, let it run out).
+ */
+ return;
+
+ do_gettimeofday(&now);
+ if (now.tv_usec >= USEC_AFTER - ((unsigned) TICK_SIZE) / 2 &&
+ now.tv_usec <= USEC_BEFORE + ((unsigned) TICK_SIZE) / 2)
+ fail = set_rtc_mmss(now.tv_sec);
+
+ next.tv_usec = USEC_AFTER - now.tv_usec;
+ if (next.tv_usec <= 0)
+ next.tv_usec += USEC_PER_SEC;
+
+ if (!fail)
+ next.tv_sec = 659;
+ else
+ next.tv_sec = 0;
+
+ if (next.tv_usec >= USEC_PER_SEC) {
+ next.tv_sec++;
+ next.tv_usec -= USEC_PER_SEC;
+ }
+ mod_timer(&sync_cmos_timer, jiffies + timeval_to_jiffies(&next));
+}
+
+void notify_arch_cmos_timer(void)
+{
+ mod_timer(&sync_cmos_timer, jiffies + 1);
+}
+
+static long clock_cmos_diff, sleep_start;
+
+static struct timer_opts *last_timer;
+static int timer_suspend(struct sys_device *dev, pm_message_t state)
+{
+ /*
+ * Estimate time zone so that set_time can update the clock
+ */
+ clock_cmos_diff = -get_cmos_time();
+ clock_cmos_diff += get_seconds();
+ sleep_start = get_cmos_time();
+ last_timer = cur_timer;
+ cur_timer = &timer_none;
+ if (last_timer->suspend)
+ last_timer->suspend(state);
+ return 0;
+}
+
+static int timer_resume(struct sys_device *dev)
+{
+ unsigned long flags;
+ unsigned long sec;
+ unsigned long sleep_length;
+
+#ifdef CONFIG_HPET_TIMER
+ if (is_hpet_enabled())
+ hpet_reenable();
+#endif
+ sec = get_cmos_time() + clock_cmos_diff;
+ sleep_length = (get_cmos_time() - sleep_start) * HZ;
+ write_seqlock_irqsave(&xtime_lock, flags);
+ xtime.tv_sec = sec;
+ xtime.tv_nsec = 0;
+ write_sequnlock_irqrestore(&xtime_lock, flags);
+ jiffies += sleep_length;
+ wall_jiffies += sleep_length;
+ if (last_timer->resume)
+ last_timer->resume();
+ cur_timer = last_timer;
+ last_timer = NULL;
+ touch_softlockup_watchdog();
+ return 0;
+}
+
+static struct sysdev_class timer_sysclass = {
+ .resume = timer_resume,
+ .suspend = timer_suspend,
+ set_kset_name("timer"),
+};
+
+
+/* XXX this driverfs stuff should probably go elsewhere later -john */
+static struct sys_device device_timer = {
+ .id = 0,
+ .cls = &timer_sysclass,
+};
+
+static int time_init_device(void)
+{
+ int error = sysdev_class_register(&timer_sysclass);
+ if (!error)
+ error = sysdev_register(&device_timer);
+ return error;
+}
+
+device_initcall(time_init_device);
+
+#ifdef CONFIG_HPET_TIMER
+extern void (*late_time_init)(void);
+/* Duplicate of time_init() below, with hpet_enable part added */
+static void __init hpet_time_init(void)
+{
+ xtime.tv_sec = get_cmos_time();
+ xtime.tv_nsec = (INITIAL_JIFFIES % HZ) * (NSEC_PER_SEC / HZ);
+ set_normalized_timespec(&wall_to_monotonic,
+ -xtime.tv_sec, -xtime.tv_nsec);
+
+ if ((hpet_enable() >= 0) && hpet_use_timer) {
+ printk("Using HPET for base-timer\n");
+ }
+
+ cur_timer = select_timer();
+ printk(KERN_INFO "Using %s for high-res timesource\n",cur_timer->name);
+
+ time_init_hook();
+}
+#endif
+
+/* Dynamically-mapped IRQ. */
+DEFINE_PER_CPU(int, timer_irq);
+
+extern void (*late_time_init)(void);
+static void setup_cpu0_timer_irq(void)
+{
+ per_cpu(timer_irq, 0) =
+ bind_virq_to_irqhandler(
+ VIRQ_TIMER,
+ 0,
+ timer_interrupt,
+ SA_INTERRUPT,
+ "timer0",
+ NULL);
+ BUG_ON(per_cpu(timer_irq, 0) < 0);
+}
+
+void __init time_init(void)
+{
+#ifdef CONFIG_HPET_TIMER
+ if (is_hpet_capable()) {
+ /*
+ * HPET initialization needs to do memory-mapped io. So, let
+ * us do a late initialization after mem_init().
+ */
+ late_time_init = hpet_time_init;
+ return;
+ }
+#endif
+ get_time_values_from_xen();
+
+ processed_system_time = per_cpu(shadow_time, 0).system_timestamp;
+ per_cpu(processed_system_time, 0) = processed_system_time;
+ init_missing_ticks_accounting(0);
+
+ update_wallclock();
+
+ init_cpu_khz();
+ printk(KERN_INFO "Xen reported: %u.%03u MHz processor.\n",
+ cpu_khz / 1000, cpu_khz % 1000);
+
+#if defined(__x86_64__)
+ vxtime.mode = VXTIME_TSC;
+ vxtime.quot = (1000000L << 32) / vxtime_hz;
+ vxtime.tsc_quot = (1000L << 32) / cpu_khz;
+ sync_core();
+ rdtscll(vxtime.last_tsc);
+#endif
+
+ /* Cannot request_irq() until kmem is initialised. */
+ late_time_init = setup_cpu0_timer_irq;
+}
+
+/* Convert jiffies to system time. */
+u64 jiffies_to_st(unsigned long j)
+{
+ unsigned long seq;
+ long delta;
+ u64 st;
+
+ do {
+ seq = read_seqbegin(&xtime_lock);
+ delta = j - jiffies;
+ /* NB. The next check can trigger in some wrap-around cases,
+ * but that's ok: we'll just end up with a shorter timeout. */
+ if (delta < 1)
+ delta = 1;
+ st = processed_system_time + (delta * (u64)NS_PER_TICK);
+ } while (read_seqretry(&xtime_lock, seq));
+
+ return st;
+}
+EXPORT_SYMBOL(jiffies_to_st);
+
+/*
+ * stop_hz_timer / start_hz_timer - enter/exit 'tickless mode' on an idle cpu
+ * These functions are based on implementations from arch/s390/kernel/time.c
+ */
+void stop_hz_timer(void)
+{
+ unsigned int cpu = smp_processor_id();
+ unsigned long j;
+
+ /* We must do this /before/ checking rcu_pending(). */
+ cpu_set(cpu, nohz_cpu_mask);
+ smp_mb();
+
+ /* Leave ourselves in 'tick mode' if rcu or softirq pending. */
+ if (rcu_pending(cpu) || local_softirq_pending()) {
+ cpu_clear(cpu, nohz_cpu_mask);
+ j = jiffies + 1;
+ } else {
+ j = next_timer_interrupt();
+ }
+
+ BUG_ON(HYPERVISOR_set_timer_op(jiffies_to_st(j)) != 0);
+}
+
+void start_hz_timer(void)
+{
+ cpu_clear(smp_processor_id(), nohz_cpu_mask);
+}
+
+/* No locking required. We are only CPU running, and interrupts are off. */
+void time_resume(void)
+{
+ init_cpu_khz();
+
+ get_time_values_from_xen();
+
+ processed_system_time = per_cpu(shadow_time, 0).system_timestamp;
+ per_cpu(processed_system_time, 0) = processed_system_time;
+ init_missing_ticks_accounting(0);
+
+ update_wallclock();
+}
+
+#ifdef CONFIG_SMP
+static char timer_name[NR_CPUS][15];
+
+void local_setup_timer(unsigned int cpu)
+{
+ int seq;
+
+ BUG_ON(cpu == 0);
+
+ do {
+ seq = read_seqbegin(&xtime_lock);
+ /* Use cpu0 timestamp: cpu's shadow is not initialised yet. */
+ per_cpu(processed_system_time, cpu) =
+ per_cpu(shadow_time, 0).system_timestamp;
+ init_missing_ticks_accounting(cpu);
+ } while (read_seqretry(&xtime_lock, seq));
+
+ sprintf(timer_name[cpu], "timer%d", cpu);
+ per_cpu(timer_irq, cpu) =
+ bind_virq_to_irqhandler(
+ VIRQ_TIMER,
+ cpu,
+ timer_interrupt,
+ SA_INTERRUPT,
+ timer_name[cpu],
+ NULL);
+ BUG_ON(per_cpu(timer_irq, cpu) < 0);
+}
+
+void local_teardown_timer(unsigned int cpu)
+{
+ BUG_ON(cpu == 0);
+ unbind_from_irqhandler(per_cpu(timer_irq, cpu), NULL);
+}
+#endif
+
+/*
+ * /proc/sys/xen: This really belongs in another file. It can stay here for
+ * now however.
+ */
+static ctl_table xen_subtable[] = {
+ {
+ .ctl_name = 1,
+ .procname = "independent_wallclock",
+ .data = &independent_wallclock,
+ .maxlen = sizeof(independent_wallclock),
+ .mode = 0644,
+ .proc_handler = proc_dointvec
+ },
+ {
+ .ctl_name = 2,
+ .procname = "permitted_clock_jitter",
+ .data = &permitted_clock_jitter,
+ .maxlen = sizeof(permitted_clock_jitter),
+ .mode = 0644,
+ .proc_handler = proc_doulongvec_minmax
+ },
+ { 0 }
+};
+static ctl_table xen_table[] = {
+ {
+ .ctl_name = 123,
+ .procname = "xen",
+ .mode = 0555,
+ .child = xen_subtable},
+ { 0 }
+};
+static int __init xen_sysctl_init(void)
+{
+ (void)register_sysctl_table(xen_table, 0);
+ return 0;
+}
+__initcall(xen_sysctl_init);
--
> # We must build both images before we can assemble it.
> --- /dev/null
> +++ xen-subarch-2.6/arch/i386/mach-xen/time.c
> @@ -0,0 +1,1045 @@
> +/*
> + * linux/arch/i386/kernel/time.c
no it's not
is there any reason why this is a full copy rather than just a
subarching of the modified functions, or some simple hooks into the
generic function?
Duplication like this is evil because it means too many places need
duplicated bugfixes and rework (and such rework is underway)
* Arjan van de Ven ([email protected]) wrote:
> Duplication like this is evil because it means too many places need
> duplicated bugfixes and rework (and such rework is underway)
Yes, that one is already on the todo list. Will work on consolidation
there.
thanks,
-chris
Chris Wright wrote:
> * Arjan van de Ven ([email protected]) wrote:
>> Duplication like this is evil because it means too many places need
>> duplicated bugfixes and rework (and such rework is underway)
>
> Yes, that one is already on the todo list. Will work on consolidation
> there.
>
> thanks,
> -chris
>
These patches from the VMI patch series sent out last week could be
easily used to abstract time when running on Xen to accomplish the
consolidation and transparency with native goals:
http://lkml.org/lkml/2006/3/13/172
http://lkml.org/lkml/2006/3/13/173
Unlike this Xen specific patch, the VMI time patches already provide
transparency with native -- the timer_vmi.c device is used when booting
on a hypervisor that provides this functionality, otherwise one of the
traditional time sources is fallen back to.
Also, they remove much of the code duplication by taking advantage of
the the i386 time interface mechanisms (and can be easily updated for
the time rework that is underway).
Finally, the vmi nicely abstracts the details of the formula that is
used to compute system_time from the HYPERVISOR_shared_info time
parameters by hiding the computation underneath the kernel <->
hypervisor interface. Note that efficiency is not lost since the code
to compute system_time, etc would execute in the "ROM" (a.k.a. hypercall
page) without needing to call all the way down into the hypervisor.
Hiding these details from the Linux code would be nice since it leads to
easier to understand linux code (it doesn't have to deal with the
details of how time is interpolated from hardware rdtsc), and allows the
calculation for "system_time", etc to change without breaking the
kernel<->hypervisor interface. It also allows for more freedom as to
how the time interface is implemented; hypervisors aren't forced to
implement system_time using a particular algorithm. The Xen interface
needlessly imposes one particular algorithm to compute system_time from
rdtsc.
See this post for a description of how the Xen time interface could be
mapped to the timer_vmi.c patches I linked above:
http://lists.xensource.com/archives/html/xen-devel/2006-02/msg00876.html
Dan
> --- xen-subarch-2.6.orig/arch/i386/kernel/Makefile
> +++ xen-subarch-2.6/arch/i386/kernel/Makefile
> @@ -9,8 +9,11 @@ obj-y := process.o semaphore.o signal.o
> pci-dma.o i386_ksyms.o i387.o dmi_scan.o bootflag.o \
> quirks.o i8237.o topology.o
>
> +timers-y := timers/
> +timers-$(CONFIG_XEN) :=
> +
You need to disable CONFIG_HPET_TIMER for CONFIG_XEN, otherwise
select_timer is undefined since you don't include timers here.
Eli
* Eli Collins ([email protected]) wrote:
> >--- xen-subarch-2.6.orig/arch/i386/kernel/Makefile
> >+++ xen-subarch-2.6/arch/i386/kernel/Makefile
> >@@ -9,8 +9,11 @@ obj-y := process.o semaphore.o signal.o
> > pci-dma.o i386_ksyms.o i387.o dmi_scan.o bootflag.o \
> > quirks.o i8237.o topology.o
> >
> >+timers-y := timers/
> >+timers-$(CONFIG_XEN) :=
> >+
>
> You need to disable CONFIG_HPET_TIMER for CONFIG_XEN, otherwise
> select_timer is undefined since you don't include timers here.
Thanks, you're right. It's done the main Xen tree, didn't make it into
this patchset, fixed.
thanks,
-chris