Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   Frederic Weisbecker <frederic@kernel.org>
To:     Thomas Gleixner <tglx@linutronix.de>
Cc:     LKML <linux-kernel@vger.kernel.org>,
        Frederic Weisbecker <frederic@kernel.org>,
        Anna-Maria Gleixner <anna-maria@linutronix.de>,
        Peter Zijlstra <peterz@infradead.org>,
        Juri Lelli <juri.lelli@redhat.com>
Subject: [PATCH 6/9] timer: Expand clk forward logic beyond nohz
Date:   Tue,  7 Jul 2020 03:32:50 +0200
Message-Id: <20200707013253.26770-7-frederic@kernel.org>
In-Reply-To: <20200707013253.26770-1-frederic@kernel.org>
References: <20200707013253.26770-1-frederic@kernel.org>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

As for next_expiry, we want to expand base->clk catch up logic beyond
nohz in order to avoid triggering useless softirqs.

If we want to fire softirqs only to expire pending timers, periodic
base->clk increments must be skippable for random amounts of time.
Therefore we must prepare to catch-up with missing updates whenever we
need an up-to-date base clock.

Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Anna-Maria Gleixner <anna-maria@linutronix.de>
Cc: Juri Lelli <juri.lelli@redhat.com>
---
 kernel/time/timer.c | 26 ++++----------------------
 1 file changed, 4 insertions(+), 22 deletions(-)

diff --git a/kernel/time/timer.c b/kernel/time/timer.c
index 501d36ef0b75..ffa2c956d968 100644
--- a/kernel/time/timer.c
+++ b/kernel/time/timer.c
@@ -885,19 +885,12 @@ get_target_base(struct timer_base *base, unsigned tflags)
 
 static inline void forward_timer_base(struct timer_base *base)
 {
-#ifdef CONFIG_NO_HZ_COMMON
 	unsigned long jnow;
 
-	/*
-	 * We only forward the base when we are idle or have just come out of
-	 * idle (must_forward_clk logic), and have a delta between base clock
-	 * and jiffies. In the common case, run_timers will take care of it.
-	 */
-	if (likely(!base->must_forward_clk))
+	if (!base->must_forward_clk)
 		return;
 
 	jnow = READ_ONCE(jiffies);
-	base->must_forward_clk = base->is_idle;
 	if ((long)(jnow - base->clk) < 2)
 		return;
 
@@ -912,7 +905,6 @@ static inline void forward_timer_base(struct timer_base *base)
 			return;
 		base->clk = base->next_expiry;
 	}
-#endif
 }
 
 
@@ -1666,10 +1658,8 @@ u64 get_next_timer_interrupt(unsigned long basej, u64 basem)
 		 * logic is only maintained for the BASE_STD base, deferrable
 		 * timers may still see large granularity skew (by design).
 		 */
-		if ((expires - basem) > TICK_NSEC) {
-			base->must_forward_clk = true;
+		if ((expires - basem) > TICK_NSEC)
 			base->is_idle = true;
-		}
 	}
 	raw_spin_unlock(&base->lock);
 
@@ -1768,16 +1758,7 @@ static inline void __run_timers(struct timer_base *base)
 	/*
 	 * timer_base::must_forward_clk must be cleared before running
 	 * timers so that any timer functions that call mod_timer() will
-	 * not try to forward the base. Idle tracking / clock forwarding
-	 * logic is only used with BASE_STD timers.
-	 *
-	 * The must_forward_clk flag is cleared unconditionally also for
-	 * the deferrable base. The deferrable base is not affected by idle
-	 * tracking and never forwarded, so clearing the flag is a NOOP.
-	 *
-	 * The fact that the deferrable base is never forwarded can cause
-	 * large variations in granularity for deferrable timers, but they
-	 * can be deferred for long periods due to idle anyway.
+	 * not try to forward the base.
 	 */
 	base->must_forward_clk = false;
 
@@ -1790,6 +1771,7 @@ static inline void __run_timers(struct timer_base *base)
 		while (levels--)
 			expire_timers(base, heads + levels);
 	}
+	base->must_forward_clk = true;
 	raw_spin_unlock_irq(&base->lock);
 	timer_base_unlock_expiry(base);
 }
-- 
2.26.2