| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-Id: <20120723010655.768372060@decadent.org.uk>
Date: Mon, 23 Jul 2012 02:07:19 +0100
From: Ben Hutchings <ben@...adent.org.uk>
To: linux-kernel@...r.kernel.org, stable@...r.kernel.org
Cc: torvalds@...ux-foundation.org, akpm@...ux-foundation.org,
alan@...rguk.ukuu.org.uk, Peter Zijlstra <a.p.zijlstra@...llo.nl>,
Ingo Molnar <mingo@...nel.org>
Subject: [ 028/108] sched/nohz: Rewrite and fix load-avg computation -- again
3.2-stable review patch. If anyone has any objections, please let me know.
------------------
From: Peter Zijlstra <a.p.zijlstra@...llo.nl>
commit 5167e8d5417bf5c322a703d2927daec727ea40dd upstream.
Thanks to Charles Wang for spotting the defects in the current code:
- If we go idle during the sample window -- after sampling, we get a
negative bias because we can negate our own sample.
- If we wake up during the sample window we get a positive bias
because we push the sample to a known active period.
So rewrite the entire nohz load-avg muck once again, now adding
copious documentation to the code.
Reported-and-tested-by: Doug Smythies <dsmythies@...us.net>
Reported-and-tested-by: Charles Wang <muming.wq@...il.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@...llo.nl>
Cc: Linus Torvalds <torvalds@...ux-foundation.org>
Cc: Andrew Morton <akpm@...ux-foundation.org>
Link: http://lkml.kernel.org/r/1340373782.18025.74.camel@twins
[ minor edits ]
Signed-off-by: Ingo Molnar <mingo@...nel.org>
[bwh: Backported to 3.2: adjust filenames, context]
Signed-off-by: Ben Hutchings <ben@...adent.org.uk>
---
include/linux/sched.h | 8 ++
kernel/sched/core.c | 275 ++++++++++++++++++++++++++++++++++------------
kernel/sched/idle_task.c | 1 -
kernel/sched/sched.h | 2 -
kernel/time/tick-sched.c | 2 +
5 files changed, 213 insertions(+), 75 deletions(-)
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1892,6 +1892,14 @@ static inline int set_cpus_allowed_ptr(s
}
#endif
+#ifdef CONFIG_NO_HZ
+void calc_load_enter_idle(void);
+void calc_load_exit_idle(void);
+#else
+static inline void calc_load_enter_idle(void) { }
+static inline void calc_load_exit_idle(void) { }
+#endif /* CONFIG_NO_HZ */
+
#ifndef CONFIG_CPUMASK_OFFSTACK
static inline int set_cpus_allowed(struct task_struct *p, cpumask_t new_mask)
{
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -1885,7 +1885,6 @@ static void double_rq_unlock(struct rq *
#endif
-static void calc_load_account_idle(struct rq *this_rq);
static void update_sysctl(void);
static int get_update_sysctl_factor(void);
static void update_cpu_load(struct rq *this_rq);
@@ -3401,11 +3400,73 @@ unsigned long this_cpu_load(void)
}
+/*
+ * Global load-average calculations
+ *
+ * We take a distributed and async approach to calculating the global load-avg
+ * in order to minimize overhead.
+ *
+ * The global load average is an exponentially decaying average of nr_running +
+ * nr_uninterruptible.
+ *
+ * Once every LOAD_FREQ:
+ *
+ * nr_active = 0;
+ * for_each_possible_cpu(cpu)
+ * nr_active += cpu_of(cpu)->nr_running + cpu_of(cpu)->nr_uninterruptible;
+ *
+ * avenrun[n] = avenrun[0] * exp_n + nr_active * (1 - exp_n)
+ *
+ * Due to a number of reasons the above turns in the mess below:
+ *
+ * - for_each_possible_cpu() is prohibitively expensive on machines with
+ * serious number of cpus, therefore we need to take a distributed approach
+ * to calculating nr_active.
+ *
+ * \Sum_i x_i(t) = \Sum_i x_i(t) - x_i(t_0) | x_i(t_0) := 0
+ * = \Sum_i { \Sum_j=1 x_i(t_j) - x_i(t_j-1) }
+ *
+ * So assuming nr_active := 0 when we start out -- true per definition, we
+ * can simply take per-cpu deltas and fold those into a global accumulate
+ * to obtain the same result. See calc_load_fold_active().
+ *
+ * Furthermore, in order to avoid synchronizing all per-cpu delta folding
+ * across the machine, we assume 10 ticks is sufficient time for every
+ * cpu to have completed this task.
+ *
+ * This places an upper-bound on the IRQ-off latency of the machine. Then
+ * again, being late doesn't loose the delta, just wrecks the sample.
+ *
+ * - cpu_rq()->nr_uninterruptible isn't accurately tracked per-cpu because
+ * this would add another cross-cpu cacheline miss and atomic operation
+ * to the wakeup path. Instead we increment on whatever cpu the task ran
+ * when it went into uninterruptible state and decrement on whatever cpu
+ * did the wakeup. This means that only the sum of nr_uninterruptible over
+ * all cpus yields the correct result.
+ *
+ * This covers the NO_HZ=n code, for extra head-aches, see the comment below.
+ */
+
/* Variables and functions for calc_load */
static atomic_long_t calc_load_tasks;
static unsigned long calc_load_update;
unsigned long avenrun[3];
-EXPORT_SYMBOL(avenrun);
+EXPORT_SYMBOL(avenrun); /* should be removed */
+
+/**
+ * get_avenrun - get the load average array
+ * @loads: pointer to dest load array
+ * @offset: offset to add
+ * @shift: shift count to shift the result left
+ *
+ * These values are estimates at best, so no need for locking.
+ */
+void get_avenrun(unsigned long *loads, unsigned long offset, int shift)
+{
+ loads[0] = (avenrun[0] + offset) << shift;
+ loads[1] = (avenrun[1] + offset) << shift;
+ loads[2] = (avenrun[2] + offset) << shift;
+}
static long calc_load_fold_active(struct rq *this_rq)
{
@@ -3422,6 +3483,9 @@ static long calc_load_fold_active(struct
return delta;
}
+/*
+ * a1 = a0 * e + a * (1 - e)
+ */
static unsigned long
calc_load(unsigned long load, unsigned long exp, unsigned long active)
{
@@ -3433,30 +3497,118 @@ calc_load(unsigned long load, unsigned l
#ifdef CONFIG_NO_HZ
/*
- * For NO_HZ we delay the active fold to the next LOAD_FREQ update.
+ * Handle NO_HZ for the global load-average.
+ *
+ * Since the above described distributed algorithm to compute the global
+ * load-average relies on per-cpu sampling from the tick, it is affected by
+ * NO_HZ.
+ *
+ * The basic idea is to fold the nr_active delta into a global idle-delta upon
+ * entering NO_HZ state such that we can include this as an 'extra' cpu delta
+ * when we read the global state.
+ *
+ * Obviously reality has to ruin such a delightfully simple scheme:
+ *
+ * - When we go NO_HZ idle during the window, we can negate our sample
+ * contribution, causing under-accounting.
+ *
+ * We avoid this by keeping two idle-delta counters and flipping them
+ * when the window starts, thus separating old and new NO_HZ load.
+ *
+ * The only trick is the slight shift in index flip for read vs write.
+ *
+ * 0s 5s 10s 15s
+ * +10 +10 +10 +10
+ * |-|-----------|-|-----------|-|-----------|-|
+ * r:0 0 1 1 0 0 1 1 0
+ * w:0 1 1 0 0 1 1 0 0
+ *
+ * This ensures we'll fold the old idle contribution in this window while
+ * accumlating the new one.
+ *
+ * - When we wake up from NO_HZ idle during the window, we push up our
+ * contribution, since we effectively move our sample point to a known
+ * busy state.
+ *
+ * This is solved by pushing the window forward, and thus skipping the
+ * sample, for this cpu (effectively using the idle-delta for this cpu which
+ * was in effect at the time the window opened). This also solves the issue
+ * of having to deal with a cpu having been in NOHZ idle for multiple
+ * LOAD_FREQ intervals.
*
* When making the ILB scale, we should try to pull this in as well.
*/
-static atomic_long_t calc_load_tasks_idle;
+static atomic_long_t calc_load_idle[2];
+static int calc_load_idx;
+
+static inline int calc_load_write_idx(void)
+{
+ int idx = calc_load_idx;
+
+ /*
+ * See calc_global_nohz(), if we observe the new index, we also
+ * need to observe the new update time.
+ */
+ smp_rmb();
+
+ /*
+ * If the folding window started, make sure we start writing in the
+ * next idle-delta.
+ */
+ if (!time_before(jiffies, calc_load_update))
+ idx++;
+
+ return idx & 1;
+}
+
+static inline int calc_load_read_idx(void)
+{
+ return calc_load_idx & 1;
+}
-static void calc_load_account_idle(struct rq *this_rq)
+void calc_load_enter_idle(void)
{
+ struct rq *this_rq = this_rq();
long delta;
+ /*
+ * We're going into NOHZ mode, if there's any pending delta, fold it
+ * into the pending idle delta.
+ */
delta = calc_load_fold_active(this_rq);
- if (delta)
- atomic_long_add(delta, &calc_load_tasks_idle);
+ if (delta) {
+ int idx = calc_load_write_idx();
+ atomic_long_add(delta, &calc_load_idle[idx]);
+ }
}
-static long calc_load_fold_idle(void)
+void calc_load_exit_idle(void)
{
- long delta = 0;
+ struct rq *this_rq = this_rq();
/*
- * Its got a race, we don't care...
+ * If we're still before the sample window, we're done.
*/
- if (atomic_long_read(&calc_load_tasks_idle))
- delta = atomic_long_xchg(&calc_load_tasks_idle, 0);
+ if (time_before(jiffies, this_rq->calc_load_update))
+ return;
+
+ /*
+ * We woke inside or after the sample window, this means we're already
+ * accounted through the nohz accounting, so skip the entire deal and
+ * sync up for the next window.
+ */
+ this_rq->calc_load_update = calc_load_update;
+ if (time_before(jiffies, this_rq->calc_load_update + 10))
+ this_rq->calc_load_update += LOAD_FREQ;
+}
+
+static long calc_load_fold_idle(void)
+{
+ int idx = calc_load_read_idx();
+ long delta = 0;
+
+ if (atomic_long_read(&calc_load_idle[idx]))
+ delta = atomic_long_xchg(&calc_load_idle[idx], 0);
return delta;
}
@@ -3542,66 +3694,39 @@ static void calc_global_nohz(void)
{
long delta, active, n;
- /*
- * If we crossed a calc_load_update boundary, make sure to fold
- * any pending idle changes, the respective CPUs might have
- * missed the tick driven calc_load_account_active() update
- * due to NO_HZ.
- */
- delta = calc_load_fold_idle();
- if (delta)
- atomic_long_add(delta, &calc_load_tasks);
-
- /*
- * It could be the one fold was all it took, we done!
- */
- if (time_before(jiffies, calc_load_update + 10))
- return;
-
- /*
- * Catch-up, fold however many we are behind still
- */
- delta = jiffies - calc_load_update - 10;
- n = 1 + (delta / LOAD_FREQ);
+ if (!time_before(jiffies, calc_load_update + 10)) {
+ /*
+ * Catch-up, fold however many we are behind still
+ */
+ delta = jiffies - calc_load_update - 10;
+ n = 1 + (delta / LOAD_FREQ);
- active = atomic_long_read(&calc_load_tasks);
- active = active > 0 ? active * FIXED_1 : 0;
+ active = atomic_long_read(&calc_load_tasks);
+ active = active > 0 ? active * FIXED_1 : 0;
- avenrun[0] = calc_load_n(avenrun[0], EXP_1, active, n);
- avenrun[1] = calc_load_n(avenrun[1], EXP_5, active, n);
- avenrun[2] = calc_load_n(avenrun[2], EXP_15, active, n);
+ avenrun[0] = calc_load_n(avenrun[0], EXP_1, active, n);
+ avenrun[1] = calc_load_n(avenrun[1], EXP_5, active, n);
+ avenrun[2] = calc_load_n(avenrun[2], EXP_15, active, n);
- calc_load_update += n * LOAD_FREQ;
-}
-#else
-static void calc_load_account_idle(struct rq *this_rq)
-{
-}
+ calc_load_update += n * LOAD_FREQ;
+ }
-static inline long calc_load_fold_idle(void)
-{
- return 0;
+ /*
+ * Flip the idle index...
+ *
+ * Make sure we first write the new time then flip the index, so that
+ * calc_load_write_idx() will see the new time when it reads the new
+ * index, this avoids a double flip messing things up.
+ */
+ smp_wmb();
+ calc_load_idx++;
}
+#else /* !CONFIG_NO_HZ */
-static void calc_global_nohz(void)
-{
-}
-#endif
+static inline long calc_load_fold_idle(void) { return 0; }
+static inline void calc_global_nohz(void) { }
-/**
- * get_avenrun - get the load average array
- * @loads: pointer to dest load array
- * @offset: offset to add
- * @shift: shift count to shift the result left
- *
- * These values are estimates at best, so no need for locking.
- */
-void get_avenrun(unsigned long *loads, unsigned long offset, int shift)
-{
- loads[0] = (avenrun[0] + offset) << shift;
- loads[1] = (avenrun[1] + offset) << shift;
- loads[2] = (avenrun[2] + offset) << shift;
-}
+#endif /* CONFIG_NO_HZ */
/*
* calc_load - update the avenrun load estimates 10 ticks after the
@@ -3609,11 +3734,18 @@ void get_avenrun(unsigned long *loads, u
*/
void calc_global_load(unsigned long ticks)
{
- long active;
+ long active, delta;
if (time_before(jiffies, calc_load_update + 10))
return;
+ /*
+ * Fold the 'old' idle-delta to include all NO_HZ cpus.
+ */
+ delta = calc_load_fold_idle();
+ if (delta)
+ atomic_long_add(delta, &calc_load_tasks);
+
active = atomic_long_read(&calc_load_tasks);
active = active > 0 ? active * FIXED_1 : 0;
@@ -3624,12 +3756,7 @@ void calc_global_load(unsigned long tick
calc_load_update += LOAD_FREQ;
/*
- * Account one period with whatever state we found before
- * folding in the nohz state and ageing the entire idle period.
- *
- * This avoids loosing a sample when we go idle between
- * calc_load_account_active() (10 ticks ago) and now and thus
- * under-accounting.
+ * In case we idled for multiple LOAD_FREQ intervals, catch up in bulk.
*/
calc_global_nohz();
}
@@ -3646,7 +3773,6 @@ static void calc_load_account_active(str
return;
delta = calc_load_fold_active(this_rq);
- delta += calc_load_fold_idle();
if (delta)
atomic_long_add(delta, &calc_load_tasks);
@@ -3654,6 +3780,10 @@ static void calc_load_account_active(str
}
/*
+ * End of global load-average stuff
+ */
+
+/*
* The exact cpuload at various idx values, calculated at every tick would be
* load = (2^idx - 1) / 2^idx * load + 1 / 2^idx * cur_load
*
--- a/kernel/sched_idletask.c
+++ b/kernel/sched_idletask.c
@@ -23,7 +23,6 @@ static void check_preempt_curr_idle(stru
static struct task_struct *pick_next_task_idle(struct rq *rq)
{
schedstat_inc(rq, sched_goidle);
- calc_load_account_idle(rq);
return rq->idle;
}
--- a/kernel/time/tick-sched.c
+++ b/kernel/time/tick-sched.c
@@ -430,6 +430,7 @@ void tick_nohz_stop_sched_tick(int inidl
*/
if (!ts->tick_stopped) {
select_nohz_load_balancer(1);
+ calc_load_enter_idle();
ts->idle_tick = hrtimer_get_expires(&ts->sched_timer);
ts->tick_stopped = 1;
@@ -563,6 +564,7 @@ void tick_nohz_restart_sched_tick(void)
account_idle_ticks(ticks);
#endif
+ calc_load_exit_idle();
touch_softlockup_watchdog();
/*
* Cancel the scheduled timer and restore the tick
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists