lkml.org 
[lkml]   [2007]   [May]   [29]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Subject[PATCH] NOHZ: prevent multiplication overflow - stop timer for huge timeouts
From
Date
get_next_timer_interrupt() returns a delta of (LONG_MAX > 1) in case
there is no timer pending. On 64 bit machines this results in a
multiplication overflow in tick_nohz_stop_sched_tick().

Reported by: Dave Miller <davem@davemloft.net>

Make the return value a constant and limit the return value to a 32 bit
value.

When the max timeout value is returned, we can safely stop the tick
timer device. The max jiffies delta results in a 12 days timeout for
HZ=1000.

In the long term the get_next_timer_interrupt() code needs to be
reworked to return ktime instead of jiffies, but we have to wait until
the last users of the original NO_IDLE_HZ code are converted.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

---
include/linux/timer.h | 6 ++++++
kernel/time/tick-sched.c | 16 +++++++++++++++-
kernel/timer.c | 10 +++++++++-
3 files changed, 30 insertions(+), 2 deletions(-)

Index: linux-2.6.21/include/linux/timer.h
===================================================================
--- linux-2.6.21.orig/include/linux/timer.h 2007-05-29 17:58:04.000000000 +0200
+++ linux-2.6.21/include/linux/timer.h 2007-05-29 17:58:06.000000000 +0200
@@ -68,6 +68,12 @@
extern int mod_timer(struct timer_list *timer, unsigned long expires);

/*
+ * The jiffies value which is added to now, when there is no timer
+ * in the timer wheel:
+ */
+#define NEXT_TIMER_MAX_DELTA ((1UL << 30) - 1)
+
+/*
* Return when the next timer-wheel timeout occurs (in absolute jiffies),
* locks the timer base:
*/
Index: linux-2.6.21/kernel/time/tick-sched.c
===================================================================
--- linux-2.6.21.orig/kernel/time/tick-sched.c 2007-05-29 17:58:04.000000000 +0200
+++ linux-2.6.21/kernel/time/tick-sched.c 2007-05-29 17:58:06.000000000 +0200
@@ -233,6 +233,21 @@
if (cpu == tick_do_timer_cpu)
tick_do_timer_cpu = -1;

+ ts->idle_sleeps++;
+
+ /*
+ * delta_jiffies >= NEXT_TIMER_MAX_DELTA signals that
+ * there is no timer pending or at least extremly far
+ * into the future (12 days for HZ=1000). In this case
+ * we simply stop the tick timer:
+ */
+ if (unlikely(delta_jiffies >= NEXT_TIMER_MAX_DELTA)) {
+ ts->idle_expires.tv64 = KTIME_MAX;
+ if (ts->nohz_mode == NOHZ_MODE_HIGHRES)
+ hrtimer_cancel(&ts->sched_timer);
+ goto out;
+ }
+
/*
* calculate the expiry time for the next timer wheel
* timer
@@ -240,7 +255,6 @@
expires = ktime_add_ns(last_update, tick_period.tv64 *
delta_jiffies);
ts->idle_expires = expires;
- ts->idle_sleeps++;

if (ts->nohz_mode == NOHZ_MODE_HIGHRES) {
hrtimer_start(&ts->sched_timer, expires,
Index: linux-2.6.21/kernel/timer.c
===================================================================
--- linux-2.6.21.orig/kernel/timer.c 2007-05-29 17:58:05.000000000 +0200
+++ linux-2.6.21/kernel/timer.c 2007-05-29 17:58:06.000000000 +0200
@@ -625,7 +625,7 @@
static unsigned long __next_timer_interrupt(tvec_base_t *base)
{
unsigned long timer_jiffies = base->timer_jiffies;
- unsigned long expires = timer_jiffies + (LONG_MAX >> 1);
+ unsigned long expires = timer_jiffies + NEXT_TIMER_MAX_DELTA;
int index, slot, array, found = 0;
struct timer_list *nte;
tvec_t *varray[4];
@@ -708,6 +708,14 @@

tsdelta = ktime_to_timespec(hr_delta);
delta = timespec_to_jiffies(&tsdelta);
+
+ /*
+ * Limit the delta to the max value, which is checked in
+ * tick_nohz_stop_sched_tick():
+ */
+ if (delta > NEXT_TIMER_MAX_DELTA)
+ delta = NEXT_TIMER_MAX_DELTA;
+
/*
* Take rounding errors in to account and make sure, that it
* expires in the next tick. Otherwise we go into an endless

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2007-05-29 23:51    [W:0.117 / U:0.920 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site