linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Thomas Gleixner <tglx@linutronix.de>
To: LKML <linux-kernel@vger.kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>,
	Anna-Maria Behnsen <anna-maria@linutronix.de>,
	Marcelo Tosatti <mtosatti@redhat.com>,
	Frederic Weisbecker <frederic@kernel.org>
Subject: [patch V2 08/10] hrtimer: Add bases argument to clock_was_set()
Date: Tue, 13 Jul 2021 15:39:53 +0200	[thread overview]
Message-ID: <20210713135158.691083465@linutronix.de> (raw)
In-Reply-To: 20210713133945.063650594@linutronix.de

From: Thomas Gleixner <tglx@linutronix.de>

clock_was_set() unconditionaly invokes retrigger_next_event() on all online
CPUs. This was necessary because that mechanism was also used for resume
from suspend to idle which is not longer the case.

The bases arguments allows the callers of clock_was_set() to hand in a mask
which tells clock_was_set() which of the hrtimer clock bases are affected
by the clock setting. This mask will be used in the next step to check
whether a CPU base has timers queued on a clock base affected by the event
and avoid the SMP function call if there are none.

Add a @bases argument, provide defines for the active bases masking and
fixup all callsites.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 kernel/time/hrtimer.c       |    4 ++--
 kernel/time/tick-internal.h |    9 ++++++++-
 kernel/time/timekeeping.c   |   14 +++++++-------
 3 files changed, 17 insertions(+), 10 deletions(-)

--- a/kernel/time/hrtimer.c
+++ b/kernel/time/hrtimer.c
@@ -880,7 +880,7 @@ static void hrtimer_reprogram(struct hrt
  * in the tick, which obviously might be stopped, so this has to bring out
  * the remote CPU which might sleep in idle to get this sorted.
  */
-void clock_was_set(void)
+void clock_was_set(unsigned int bases)
 {
 	if (!hrtimer_hres_active() && !tick_nohz_active)
 		goto out_timerfd;
@@ -894,7 +894,7 @@ void clock_was_set(void)
 
 static void clock_was_set_work(struct work_struct *work)
 {
-	clock_was_set();
+	clock_was_set(CLOCK_SET_WALL);
 }
 
 static DECLARE_WORK(hrtimer_work, clock_was_set_work);
--- a/kernel/time/tick-internal.h
+++ b/kernel/time/tick-internal.h
@@ -166,7 +166,14 @@ DECLARE_PER_CPU(struct hrtimer_cpu_base,
 extern u64 get_next_timer_interrupt(unsigned long basej, u64 basem);
 void timer_clear_idle(void);
 
-void clock_was_set(void);
+#define CLOCK_SET_WALL							\
+	(BIT(HRTIMER_BASE_REALTIME) | BIT(HRTIMER_BASE_REALTIME_SOFT) |	\
+	 BIT(HRTIMER_BASE_TAI) | BIT(HRTIMER_BASE_TAI_SOFT))
+
+#define CLOCK_SET_BOOT							\
+	(BIT(HRTIMER_BASE_BOOTTIME) | BIT(HRTIMER_BASE_BOOTTIME_SOFT))
+
+void clock_was_set(unsigned int bases);
 void clock_was_set_delayed(void);
 
 void hrtimers_resume_local(void);
--- a/kernel/time/timekeeping.c
+++ b/kernel/time/timekeeping.c
@@ -1323,8 +1323,8 @@ int do_settimeofday64(const struct times
 	write_seqcount_end(&tk_core.seq);
 	raw_spin_unlock_irqrestore(&timekeeper_lock, flags);
 
-	/* signal hrtimers about time change */
-	clock_was_set();
+	/* Signal hrtimers about time change */
+	clock_was_set(CLOCK_SET_WALL);
 
 	if (!ret)
 		audit_tk_injoffset(ts_delta);
@@ -1371,8 +1371,8 @@ error: /* even if we error out, we forwa
 	write_seqcount_end(&tk_core.seq);
 	raw_spin_unlock_irqrestore(&timekeeper_lock, flags);
 
-	/* signal hrtimers about time change */
-	clock_was_set();
+	/* Signal hrtimers about time change */
+	clock_was_set(CLOCK_SET_WALL);
 
 	return ret;
 }
@@ -1746,8 +1746,8 @@ void timekeeping_inject_sleeptime64(cons
 	write_seqcount_end(&tk_core.seq);
 	raw_spin_unlock_irqrestore(&timekeeper_lock, flags);
 
-	/* signal hrtimers about time change */
-	clock_was_set();
+	/* Signal hrtimers about time change */
+	clock_was_set(CLOCK_SET_WALL | CLOCK_SET_BOOT);
 }
 #endif
 
@@ -2440,7 +2440,7 @@ int do_adjtimex(struct __kernel_timex *t
 		clock_set |= timekeeping_advance(TK_ADV_FREQ);
 
 	if (clock_set)
-		clock_was_set();
+		clock_was_set(CLOCK_REALTIME);
 
 	ntp_notify_cmos_timer();
 


  parent reply	other threads:[~2021-07-13 13:54 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-07-13 13:39 [patch V2 00/10] hrtimer: Reprogramming and clock_was_set() overhaul Thomas Gleixner
2021-07-13 13:39 ` [patch V2 01/10] hrtimer: Avoid double reprogramming in __hrtimer_start_range_ns() Thomas Gleixner
2021-08-10 16:02   ` [tip: timers/core] " tip-bot2 for Thomas Gleixner
2021-07-13 13:39 ` [patch V2 02/10] hrtimer: Consolidate reprogramming code Thomas Gleixner
2021-08-10 16:02   ` [tip: timers/core] " tip-bot2 for Peter Zijlstra
2021-08-12  7:19     ` Mike Galbraith
2021-08-12 14:11       ` Thomas Gleixner
2021-08-12 14:32         ` Thomas Gleixner
2021-08-12 15:04           ` Mike Galbraith
2021-08-12 15:22             ` Thomas Gleixner
2021-08-12 15:31               ` Mike Galbraith
2021-08-12 16:58                 ` Thomas Gleixner
2021-08-12 15:27             ` Mike Galbraith
2021-08-12 20:32           ` [PATCH] hrtimer: Unbreak hrtimer_force_reprogram() Thomas Gleixner
2021-08-12 20:40             ` [tip: timers/core] " tip-bot2 for Thomas Gleixner
2021-08-13  7:58             ` [PATCH] " Thomas Gleixner
2021-08-12 20:31         ` [PATCH] hrtimer: Use raw_cpu_ptr() in clock_was_set() Thomas Gleixner
2021-08-12 20:40           ` [tip: timers/core] " tip-bot2 for Thomas Gleixner
2021-08-13  6:42           ` [PATCH] " Marek Szyprowski
     [not found]   ` <CGME20210812130945eucas1p117fc1e90f31c8d9fd177932cd1a18512@eucas1p1.samsung.com>
2021-08-12 13:09     ` [patch V2 02/10] hrtimer: Consolidate reprogramming code Marek Szyprowski
2021-07-13 13:39 ` [patch V2 03/10] hrtimer: Ensure timerfd notification for HIGHRES=n Thomas Gleixner
2021-08-10 16:02   ` [tip: timers/core] " tip-bot2 for Thomas Gleixner
2021-07-13 13:39 ` [patch V2 04/10] hrtimer: Force clock_was_set() handling for the HIGHRES=n, NOHZ=y case Thomas Gleixner
2021-08-10 16:02   ` [tip: timers/core] " tip-bot2 for Thomas Gleixner
2021-07-13 13:39 ` [patch V2 05/10] timerfd: Provide timerfd_resume() Thomas Gleixner
2021-08-10 16:02   ` [tip: timers/core] " tip-bot2 for Thomas Gleixner
2021-07-13 13:39 ` [patch V2 06/10] timekeeping: Distangle resume and clock-was-set events Thomas Gleixner
2021-08-10 16:02   ` [tip: timers/core] " tip-bot2 for Thomas Gleixner
2021-07-13 13:39 ` [patch V2 07/10] time/timekeeping: Avoid invoking clock_was_set() twice Thomas Gleixner
2021-08-10 16:02   ` [tip: timers/core] " tip-bot2 for Thomas Gleixner
2021-07-13 13:39 ` Thomas Gleixner [this message]
2021-08-10 16:02   ` [tip: timers/core] hrtimer: Add bases argument to clock_was_set() tip-bot2 for Thomas Gleixner
2021-07-13 13:39 ` [patch V2 09/10] hrtimer: Avoid unnecessary SMP function calls in clock_was_set() Thomas Gleixner
2021-08-10 16:02   ` [tip: timers/core] " tip-bot2 for Marcelo Tosatti
2021-07-13 13:39 ` [patch V2 10/10] hrtimer: Avoid more " Thomas Gleixner
2021-08-10 16:02   ` [tip: timers/core] " tip-bot2 for Thomas Gleixner

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=20210713135158.691083465@linutronix.de \
    --to=tglx@linutronix.de \
    --cc=anna-maria@linutronix.de \
    --cc=frederic@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mtosatti@redhat.com \
    --cc=peterz@infradead.org \
    /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;
as well as URLs for NNTP newsgroup(s).