From: tip-bot for Ashwin Chaugule <ashwinc@quicinc.com>
To: linux-tip-commits@vger.kernel.org
Cc: linux-kernel@vger.kernel.org, hpa@zytor.com, mingo@redhat.com,
tglx@linutronix.de, ashwinc@quicinc.com
Subject: [tip:timers/core] hrtimer: Eliminate needless reprogramming of clock events device
Date: Tue, 15 Sep 2009 09:09:39 GMT [thread overview]
Message-ID: <tip-6198f0b6b0473bb13992fd7fe9ef8145df2d0a30@git.kernel.org> (raw)
In-Reply-To: <4AA00165.90609@codeaurora.org>
Commit-ID: 6198f0b6b0473bb13992fd7fe9ef8145df2d0a30
Gitweb: http://git.kernel.org/tip/6198f0b6b0473bb13992fd7fe9ef8145df2d0a30
Author: Ashwin Chaugule <ashwinc@quicinc.com>
AuthorDate: Tue, 1 Sep 2009 23:03:33 -0400
Committer: Thomas Gleixner <tglx@linutronix.de>
CommitDate: Tue, 15 Sep 2009 11:02:31 +0200
hrtimer: Eliminate needless reprogramming of clock events device
On NOHZ systems the following timers,
- tick_nohz_restart_sched_tick (tick_sched_timer)
- hrtimer_start (tick_sched_timer)
are reprogramming the clock events device far more often than needed
because there is no check to see if the currently removed or restarted
hrtimer is:
1) the one which previously armed the clock events device.
2) going to be replaced by another timer which has the same expiry time.
Avoid the reprogramming in hrtimer_force_reprogram when the new expiry
value which is evaluated from the clock bases is equal to
cpu_base->expires_next.
This results in faster application startup time by ~4%.
[ tglx: simplified initial solution ]
Signed-off-by: Ashwin Chaugule <ashwinc@quicinc.com>
LKML-Reference: <4AA00165.90609@codeaurora.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
kernel/hrtimer.c | 52 ++++++++++++++++++++++++++++++++++------------------
1 files changed, 34 insertions(+), 18 deletions(-)
diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c
index e2f91ec..b53e7ef 100644
--- a/kernel/hrtimer.c
+++ b/kernel/hrtimer.c
@@ -486,13 +486,14 @@ static inline int hrtimer_hres_active(void)
* next event
* Called with interrupts disabled and base->lock held
*/
-static void hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base)
+static void
+hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base, int skip_equal)
{
int i;
struct hrtimer_clock_base *base = cpu_base->clock_base;
- ktime_t expires;
+ ktime_t expires, expires_next;
- cpu_base->expires_next.tv64 = KTIME_MAX;
+ expires_next.tv64 = KTIME_MAX;
for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++, base++) {
struct hrtimer *timer;
@@ -508,10 +509,15 @@ static void hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base)
*/
if (expires.tv64 < 0)
expires.tv64 = 0;
- if (expires.tv64 < cpu_base->expires_next.tv64)
- cpu_base->expires_next = expires;
+ if (expires.tv64 < expires_next.tv64)
+ expires_next = expires;
}
+ if (skip_equal && expires_next.tv64 == cpu_base->expires_next.tv64)
+ return;
+
+ cpu_base->expires_next.tv64 = expires_next.tv64;
+
if (cpu_base->expires_next.tv64 != KTIME_MAX)
tick_program_event(cpu_base->expires_next, 1);
}
@@ -594,7 +600,7 @@ static void retrigger_next_event(void *arg)
base->clock_base[CLOCK_REALTIME].offset =
timespec_to_ktime(realtime_offset);
- hrtimer_force_reprogram(base);
+ hrtimer_force_reprogram(base, 0);
spin_unlock(&base->lock);
}
@@ -707,7 +713,8 @@ static int hrtimer_switch_to_hres(void)
static inline int hrtimer_hres_active(void) { return 0; }
static inline int hrtimer_is_hres_enabled(void) { return 0; }
static inline int hrtimer_switch_to_hres(void) { return 0; }
-static inline void hrtimer_force_reprogram(struct hrtimer_cpu_base *base) { }
+static inline void
+hrtimer_force_reprogram(struct hrtimer_cpu_base *base, int skip_equal) { }
static inline int hrtimer_enqueue_reprogram(struct hrtimer *timer,
struct hrtimer_clock_base *base,
int wakeup)
@@ -850,19 +857,28 @@ static void __remove_hrtimer(struct hrtimer *timer,
struct hrtimer_clock_base *base,
unsigned long newstate, int reprogram)
{
- if (timer->state & HRTIMER_STATE_ENQUEUED) {
- /*
- * Remove the timer from the rbtree and replace the
- * first entry pointer if necessary.
- */
- if (base->first == &timer->node) {
- base->first = rb_next(&timer->node);
- /* Reprogram the clock event device. if enabled */
- if (reprogram && hrtimer_hres_active())
- hrtimer_force_reprogram(base->cpu_base);
+ ktime_t expires;
+
+ if (!(timer->state & HRTIMER_STATE_ENQUEUED))
+ goto out;
+
+ /*
+ * Remove the timer from the rbtree and replace the first
+ * entry pointer if necessary.
+ */
+ if (base->first == &timer->node) {
+ base->first = rb_next(&timer->node);
+ /* Reprogram the clock event device. if enabled */
+ if (reprogram && hrtimer_hres_active()) {
+ expires = ktime_sub(hrtimer_get_expires(timer),
+ base->offset);
+ if (base->cpu_base->expires_next.tv64 == expires.tv64)
+ hrtimer_force_reprogram(base->cpu_base, 1);
}
- rb_erase(&timer->node, &base->active);
}
+
+ rb_erase(&timer->node, &base->active);
+out:
timer->state = newstate;
}
next prev parent reply other threads:[~2009-09-15 9:10 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-08-27 21:51 [RFC] [PATCH 1/1] hrtimers: Cache next hrtimer Ashwin Chaugule
2009-08-27 21:56 ` Ashwin Chaugule
2009-08-27 22:51 ` Thomas Gleixner
2009-08-27 23:09 ` Ashwin Chaugule
2009-08-27 23:16 ` Thomas Gleixner
2009-08-28 5:56 ` Ashwin Chaugule
2009-08-28 11:17 ` Thomas Gleixner
2009-08-28 16:34 ` Ashwin Chaugule
2009-08-28 18:19 ` Thomas Gleixner
2009-08-28 20:27 ` Ashwin Chaugule
2009-08-30 6:06 ` Ashwin Chaugule
2009-08-30 8:36 ` Thomas Gleixner
2009-08-31 4:17 ` Ashwin Chaugule
2009-08-31 7:08 ` Thomas Gleixner
2009-09-01 3:13 ` Ashwin Chaugule
2009-09-03 17:48 ` Ashwin Chaugule
2009-09-15 9:09 ` tip-bot for Ashwin Chaugule [this message]
2009-09-15 15:19 ` [tip:timers/core] hrtimer: Eliminate needless reprogramming of clock events device tip-bot for Ashwin Chaugule
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=tip-6198f0b6b0473bb13992fd7fe9ef8145df2d0a30@git.kernel.org \
--to=ashwinc@quicinc.com \
--cc=hpa@zytor.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-tip-commits@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=tglx@linutronix.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Powered by JetHome