Merge tag 'timers-core-2021-08-30' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull timer updates from Thomas Gleixner:
 "Updates for timekeeping, timers and related drivers:

  Core code:

   - Cure a couple of correctness issues in the posix CPU timer code to
     prevent that the tick dependency for NOHZ full is kept alive for no
     reason.

   - Avoid expensive double reprogramming of the clockevent device in
     hrtimer_start_range_ns().

   - Avoid pointless SMP function calls when the clock was set to avoid
     disturbing CPUs which do not have any affected timers queued.

   - Make the clocksource watchdog test work correctly when CONFIG_HZ is
     less than 100.

  Drivers:

   - Prefer the ARM architected timer over the Exynos timer which is way
     more expensive to access.

   - Add device tree bindings for new Ingenic SoCs

   - The usual improvements and cleanups all over the place"

* tag 'timers-core-2021-08-30' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (29 commits)
  clocksource: Make clocksource watchdog test safe for slow-HZ systems
  dt-bindings: timer: Add ABIs for new Ingenic SoCs
  clocksource/drivers/fttmr010: Pass around less pointers
  clocksource/drivers/mediatek: Optimize systimer irq clear flow on shutdown
  clocksource/drivers/ingenic: Use bitfield macro helpers
  clocksource/drivers/sh_cmt: Fix wrong setting if don't request IRQ for clock source channel
  dt-bindings: timer: convert rockchip,rk-timer.txt to YAML
  clocksource/drivers/exynos_mct: Mark MCT device as CLOCK_EVT_FEAT_PERCPU
  clocksource/drivers/exynos_mct: Prioritise Arm arch timer on arm64
  hrtimer: Unbreak hrtimer_force_reprogram()
  hrtimer: Use raw_cpu_ptr() in clock_was_set()
  hrtimer: Avoid more SMP function calls in clock_was_set()
  hrtimer: Avoid unnecessary SMP function calls in clock_was_set()
  hrtimer: Add bases argument to clock_was_set()
  time/timekeeping: Avoid invoking clock_was_set() twice
  timekeeping: Distangle resume and clock-was-set events
  timerfd: Provide timerfd_resume()
  hrtimer: Force clock_was_set() handling for the HIGHRES=n, NOHZ=y case
  hrtimer: Ensure timerfd notification for HIGHRES=n
  hrtimer: Consolidate reprogramming code
  ...

Documentation/devicetree/bindings/timer/rockchip,rk-timer.txt

-Rockchip rk timer
-
-Required properties:
-- compatible: should be:
-  "rockchip,rv1108-timer", "rockchip,rk3288-timer": for Rockchip RV1108
-  "rockchip,rk3036-timer", "rockchip,rk3288-timer": for Rockchip RK3036
-  "rockchip,rk3066-timer", "rockchip,rk3288-timer": for Rockchip RK3066
-  "rockchip,rk3188-timer", "rockchip,rk3288-timer": for Rockchip RK3188
-  "rockchip,rk3228-timer", "rockchip,rk3288-timer": for Rockchip RK3228
-  "rockchip,rk3229-timer", "rockchip,rk3288-timer": for Rockchip RK3229
-  "rockchip,rk3288-timer": for Rockchip RK3288
-  "rockchip,rk3368-timer", "rockchip,rk3288-timer": for Rockchip RK3368
-  "rockchip,rk3399-timer": for Rockchip RK3399
-- reg: base address of the timer register starting with TIMERS CONTROL register
-- interrupts: should contain the interrupts for Timer0
-- clocks : must contain an entry for each entry in clock-names
-- clock-names : must include the following entries:
-  "timer", "pclk"
-
-Example:
-	timer: timer@ff810000 {
-		compatible = "rockchip,rk3288-timer";
-		reg = <0xff810000 0x20>;
-		interrupts = <GIC_SPI 72 IRQ_TYPE_LEVEL_HIGH>;
-		clocks = <&xin24m>, <&cru PCLK_TIMER>;
-		clock-names = "timer", "pclk";
-	};

Documentation/devicetree/bindings/timer/rockchip,rk-timer.yaml

+# SPDX-License-Identifier: GPL-2.0
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/timer/rockchip,rk-timer.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Rockchip Timer Device Tree Bindings
+
+maintainers:
+  - Daniel Lezcano <daniel.lezcano@linaro.org>
+
+properties:
+  compatible:
+    oneOf:
+      - const: rockchip,rk3288-timer
+      - const: rockchip,rk3399-timer
+      - items:
+          - enum:
+              - rockchip,rv1108-timer
+              - rockchip,rk3036-timer
+              - rockchip,rk3066-timer
+              - rockchip,rk3188-timer
+              - rockchip,rk3228-timer
+              - rockchip,rk3229-timer
+              - rockchip,rk3288-timer
+              - rockchip,rk3368-timer
+              - rockchip,px30-timer
+          - const: rockchip,rk3288-timer
+  reg:
+    maxItems: 1
+
+  interrupts:
+    maxItems: 1
+
+  clocks:
+    minItems: 2
+    maxItems: 2
+
+  clock-names:
+    items:
+      - const: pclk
+      - const: timer
+
+required:
+  - compatible
+  - reg
+  - interrupts
+  - clocks
+  - clock-names
+
+additionalProperties: false
+
+examples:
+  - |
+    #include <dt-bindings/interrupt-controller/arm-gic.h>
+    #include <dt-bindings/clock/rk3288-cru.h>
+
+    timer: timer@ff810000 {
+        compatible = "rockchip,rk3288-timer";
+        reg = <0xff810000 0x20>;
+        interrupts = <GIC_SPI 72 IRQ_TYPE_LEVEL_HIGH>;
+        clocks = <&cru PCLK_TIMER>, <&xin24m>;
+        clock-names = "pclk", "timer";
+    };

drivers/clocksource/exynos_mct.c

@@ -51,6 +51,15 @@
 
 #define TICK_BASE_CNT	1
 
+#ifdef CONFIG_ARM
+/* Use values higher than ARM arch timer. See 6282edb72bed. */
+#define MCT_CLKSOURCE_RATING		450
+#define MCT_CLKEVENTS_RATING		500
+#else
+#define MCT_CLKSOURCE_RATING		350
+#define MCT_CLKEVENTS_RATING		350
+#endif
+
 enum {
 	MCT_INT_SPI,
 	MCT_INT_PPI
@@ -206,7 +215,7 @@ static void exynos4_frc_resume(struct clocksource *cs)
 
 static struct clocksource mct_frc = {
 	.name		= "mct-frc",
-	.rating		= 450,	/* use value higher than ARM arch timer */
+	.rating		= MCT_CLKSOURCE_RATING,
 	.read		= exynos4_frc_read,
 	.mask		= CLOCKSOURCE_MASK(32),
 	.flags		= CLOCK_SOURCE_IS_CONTINUOUS,
@@ -456,8 +465,9 @@ static int exynos4_mct_starting_cpu(unsigned int cpu)
 	evt->set_state_oneshot = set_state_shutdown;
 	evt->set_state_oneshot_stopped = set_state_shutdown;
 	evt->tick_resume = set_state_shutdown;
-	evt->features = CLOCK_EVT_FEAT_PERIODIC | CLOCK_EVT_FEAT_ONESHOT;
-	evt->rating = 500;	/* use value higher than ARM arch timer */
+	evt->features = CLOCK_EVT_FEAT_PERIODIC | CLOCK_EVT_FEAT_ONESHOT |
+			CLOCK_EVT_FEAT_PERCPU;
+	evt->rating = MCT_CLKEVENTS_RATING,
 
 	exynos4_mct_write(TICK_BASE_CNT, mevt->base + MCT_L_TCNTB_OFFSET);
 

drivers/clocksource/ingenic-sysost.c

@@ -4,6 +4,7 @@
  * Copyright (c) 2020 周琰杰 (Zhou Yanjie) <zhouyanjie@wanyeetech.com>
  */
 
+#include <linux/bitfield.h>
 #include <linux/bitops.h>
 #include <linux/clk.h>
 #include <linux/clk-provider.h>
@@ -34,8 +35,6 @@
 /* bits within the OSTCCR register */
 #define OSTCCR_PRESCALE1_MASK	0x3
 #define OSTCCR_PRESCALE2_MASK	0xc
-#define OSTCCR_PRESCALE1_LSB	0
-#define OSTCCR_PRESCALE2_LSB	2
 
 /* bits within the OSTCR register */
 #define OSTCR_OST1CLR			BIT(0)
@@ -98,7 +97,7 @@ static unsigned long ingenic_ost_percpu_timer_recalc_rate(struct clk_hw *hw,
 
 	prescale = readl(ost_clk->ost->base + info->ostccr_reg);
 
-	prescale = (prescale & OSTCCR_PRESCALE1_MASK) >> OSTCCR_PRESCALE1_LSB;
+	prescale = FIELD_GET(OSTCCR_PRESCALE1_MASK, prescale);
 
 	return parent_rate >> (prescale * 2);
 }
@@ -112,7 +111,7 @@ static unsigned long ingenic_ost_global_timer_recalc_rate(struct clk_hw *hw,
 
 	prescale = readl(ost_clk->ost->base + info->ostccr_reg);
 
-	prescale = (prescale & OSTCCR_PRESCALE2_MASK) >> OSTCCR_PRESCALE2_LSB;
+	prescale = FIELD_GET(OSTCCR_PRESCALE2_MASK, prescale);
 
 	return parent_rate >> (prescale * 2);
 }
@@ -151,7 +150,8 @@ static int ingenic_ost_percpu_timer_set_rate(struct clk_hw *hw, unsigned long re
 	int val;
 
 	val = readl(ost_clk->ost->base + info->ostccr_reg);
-	val = (val & ~OSTCCR_PRESCALE1_MASK) | (prescale << OSTCCR_PRESCALE1_LSB);
+	val &= ~OSTCCR_PRESCALE1_MASK;
+	val |= FIELD_PREP(OSTCCR_PRESCALE1_MASK, prescale);
 	writel(val, ost_clk->ost->base + info->ostccr_reg);
 
 	return 0;
@@ -166,7 +166,8 @@ static int ingenic_ost_global_timer_set_rate(struct clk_hw *hw, unsigned long re
 	int val;
 
 	val = readl(ost_clk->ost->base + info->ostccr_reg);
-	val = (val & ~OSTCCR_PRESCALE2_MASK) | (prescale << OSTCCR_PRESCALE2_LSB);
+	val &= ~OSTCCR_PRESCALE2_MASK;
+	val |= FIELD_PREP(OSTCCR_PRESCALE2_MASK, prescale);
 	writel(val, ost_clk->ost->base + info->ostccr_reg);
 
 	return 0;

drivers/clocksource/sh_cmt.c

@@ -579,7 +579,8 @@ static int sh_cmt_start(struct sh_cmt_channel *ch, unsigned long flag)
 	ch->flags |= flag;
 
 	/* setup timeout if no clockevent */
-	if ((flag == FLAG_CLOCKSOURCE) && (!(ch->flags & FLAG_CLOCKEVENT)))
+	if (ch->cmt->num_channels == 1 &&
+	    flag == FLAG_CLOCKSOURCE && (!(ch->flags & FLAG_CLOCKEVENT)))
 		__sh_cmt_set_next(ch, ch->max_match_value);
  out:
 	raw_spin_unlock_irqrestore(&ch->lock, flags);
@@ -621,8 +622,10 @@ static struct sh_cmt_channel *cs_to_sh_cmt(struct clocksource *cs)
 static u64 sh_cmt_clocksource_read(struct clocksource *cs)
 {
 	struct sh_cmt_channel *ch = cs_to_sh_cmt(cs);
-	unsigned long flags;
 	u32 has_wrapped;
+
+	if (ch->cmt->num_channels == 1) {
+		unsigned long flags;
 		u64 value;
 		u32 raw;
 
@@ -635,6 +638,9 @@ static u64 sh_cmt_clocksource_read(struct clocksource *cs)
 		raw_spin_unlock_irqrestore(&ch->lock, flags);
 
 		return value + raw;
+	}
+
+	return sh_cmt_get_counter(ch, &has_wrapped);
 }
 
 static int sh_cmt_clocksource_enable(struct clocksource *cs)
@@ -697,7 +703,7 @@ static int sh_cmt_register_clocksource(struct sh_cmt_channel *ch,
 	cs->disable = sh_cmt_clocksource_disable;
 	cs->suspend = sh_cmt_clocksource_suspend;
 	cs->resume = sh_cmt_clocksource_resume;
-	cs->mask = CLOCKSOURCE_MASK(sizeof(u64) * 8);
+	cs->mask = CLOCKSOURCE_MASK(ch->cmt->info->width);
 	cs->flags = CLOCK_SOURCE_IS_CONTINUOUS;
 
 	dev_info(&ch->cmt->pdev->dev, "ch%u: used as clock source\n",

drivers/clocksource/timer-fttmr010.c

@@ -271,9 +271,7 @@ static irqreturn_t ast2600_timer_interrupt(int irq, void *dev_id)
 }
 
 static int __init fttmr010_common_init(struct device_node *np,
-		bool is_aspeed,
-		int (*timer_shutdown)(struct clock_event_device *),
-		irq_handler_t irq_handler)
+				       bool is_aspeed, bool is_ast2600)
 {
 	struct fttmr010 *fttmr010;
 	int irq;
@@ -374,8 +372,6 @@ static int __init fttmr010_common_init(struct device_node *np,
 				     fttmr010->tick_rate);
 	}
 
-	fttmr010->timer_shutdown = timer_shutdown;
-
 	/*
 	 * Setup clockevent timer (interrupt-driven) on timer 1.
 	 */
@@ -383,8 +379,18 @@ static int __init fttmr010_common_init(struct device_node *np,
 	writel(0, fttmr010->base + TIMER1_LOAD);
 	writel(0, fttmr010->base + TIMER1_MATCH1);
 	writel(0, fttmr010->base + TIMER1_MATCH2);
-	ret = request_irq(irq, irq_handler, IRQF_TIMER,
-			  "FTTMR010-TIMER1", &fttmr010->clkevt);
+
+	if (is_ast2600) {
+		fttmr010->timer_shutdown = ast2600_timer_shutdown;
+		ret = request_irq(irq, ast2600_timer_interrupt,
+				  IRQF_TIMER, "FTTMR010-TIMER1",
+				  &fttmr010->clkevt);
+	} else {
+		fttmr010->timer_shutdown = fttmr010_timer_shutdown;
+		ret = request_irq(irq, fttmr010_timer_interrupt,
+				  IRQF_TIMER, "FTTMR010-TIMER1",
+				  &fttmr010->clkevt);
+	}
 	if (ret) {
 		pr_err("FTTMR010-TIMER1 no IRQ\n");
 		goto out_unmap;
@@ -432,23 +438,17 @@ static int __init fttmr010_common_init(struct device_node *np,
 
 static __init int ast2600_timer_init(struct device_node *np)
 {
-	return fttmr010_common_init(np, true,
-			ast2600_timer_shutdown,
-			ast2600_timer_interrupt);
+	return fttmr010_common_init(np, true, true);
 }
 
 static __init int aspeed_timer_init(struct device_node *np)
 {
-	return fttmr010_common_init(np, true,
-			fttmr010_timer_shutdown,
-			fttmr010_timer_interrupt);
+	return fttmr010_common_init(np, true, false);
 }
 
 static __init int fttmr010_timer_init(struct device_node *np)
 {
-	return fttmr010_common_init(np, false,
-			fttmr010_timer_shutdown,
-			fttmr010_timer_interrupt);
+	return fttmr010_common_init(np, false, false);
 }
 
 TIMER_OF_DECLARE(fttmr010, "faraday,fttmr010", fttmr010_timer_init);

drivers/clocksource/timer-mediatek.c

@@ -60,9 +60,9 @@
  * SYST_CON_EN: Clock enable. Shall be set to
  *   - Start timer countdown.
  *   - Allow timeout ticks being updated.
- *   - Allow changing interrupt functions.
+ *   - Allow changing interrupt status,like clear irq pending.
  *
- * SYST_CON_IRQ_EN: Set to allow interrupt.
+ * SYST_CON_IRQ_EN: Set to enable interrupt.
  *
  * SYST_CON_IRQ_CLR: Set to clear interrupt.
  */
@@ -75,6 +75,7 @@ static void __iomem *gpt_sched_reg __read_mostly;
 static void mtk_syst_ack_irq(struct timer_of *to)
 {
 	/* Clear and disable interrupt */
+	writel(SYST_CON_EN, SYST_CON_REG(to));
 	writel(SYST_CON_IRQ_CLR | SYST_CON_EN, SYST_CON_REG(to));
 }
 
@@ -111,6 +112,9 @@ static int mtk_syst_clkevt_next_event(unsigned long ticks,
 
 static int mtk_syst_clkevt_shutdown(struct clock_event_device *clkevt)
 {
+	/* Clear any irq */
+	mtk_syst_ack_irq(to_timer_of(clkevt));
+
 	/* Disable timer */
 	writel(0, SYST_CON_REG(to_timer_of(clkevt)));
 

fs/timerfd.c

@@ -115,6 +115,22 @@ void timerfd_clock_was_set(void)
 	rcu_read_unlock();
 }
 
+static void timerfd_resume_work(struct work_struct *work)
+{
+	timerfd_clock_was_set();
+}
+
+static DECLARE_WORK(timerfd_work, timerfd_resume_work);
+
+/*
+ * Invoked from timekeeping_resume(). Defer the actual update to work so
+ * timerfd_clock_was_set() runs in task context.
+ */
+void timerfd_resume(void)
+{
+	schedule_work(&timerfd_work);
+}
+
 static void __timerfd_remove_cancel(struct timerfd_ctx *ctx)
 {
 	if (ctx->might_cancel) {

include/dt-bindings/clock/ingenic,sysost.h

@@ -13,4 +13,23 @@
 #define OST_CLK_PERCPU_TIMER2	3
 #define OST_CLK_PERCPU_TIMER3	4
 
+#define OST_CLK_EVENT_TIMER		1
+
+#define OST_CLK_EVENT_TIMER0	0
+#define OST_CLK_EVENT_TIMER1	1
+#define OST_CLK_EVENT_TIMER2	2
+#define OST_CLK_EVENT_TIMER3	3
+#define OST_CLK_EVENT_TIMER4	4
+#define OST_CLK_EVENT_TIMER5	5
+#define OST_CLK_EVENT_TIMER6	6
+#define OST_CLK_EVENT_TIMER7	7
+#define OST_CLK_EVENT_TIMER8	8
+#define OST_CLK_EVENT_TIMER9	9
+#define OST_CLK_EVENT_TIMER10	10
+#define OST_CLK_EVENT_TIMER11	11
+#define OST_CLK_EVENT_TIMER12	12
+#define OST_CLK_EVENT_TIMER13	13
+#define OST_CLK_EVENT_TIMER14	14
+#define OST_CLK_EVENT_TIMER15	15
+
 #endif /* __DT_BINDINGS_CLOCK_INGENIC_OST_H__ */

include/linux/hrtimer.h

@@ -318,16 +318,12 @@ struct clock_event_device;
 
 extern void hrtimer_interrupt(struct clock_event_device *dev);
 
-extern void clock_was_set_delayed(void);
-
 extern unsigned int hrtimer_resolution;
 
 #else
 
 #define hrtimer_resolution	(unsigned int)LOW_RES_NSEC
 
-static inline void clock_was_set_delayed(void) { }
-
 #endif
 
 static inline ktime_t
@@ -351,13 +347,13 @@ hrtimer_expires_remaining_adjusted(const struct hrtimer *timer)
 						    timer->base->get_time());
 }
 
-extern void clock_was_set(void);
 #ifdef CONFIG_TIMERFD
 extern void timerfd_clock_was_set(void);
+extern void timerfd_resume(void);
 #else
 static inline void timerfd_clock_was_set(void) { }
+static inline void timerfd_resume(void) { }
 #endif
-extern void hrtimers_resume(void);
 
 DECLARE_PER_CPU(struct tick_device, tick_cpu_device);
 

include/linux/posix-timers.h

@@ -82,12 +82,19 @@ static inline bool cpu_timer_enqueue(struct timerqueue_head *head,
 	return timerqueue_add(head, &ctmr->node);
 }
 
-static inline void cpu_timer_dequeue(struct cpu_timer *ctmr)
+static inline bool cpu_timer_queued(struct cpu_timer *ctmr)
 {
-	if (ctmr->head) {
+	return !!ctmr->head;
+}
+
+static inline bool cpu_timer_dequeue(struct cpu_timer *ctmr)
+{
+	if (cpu_timer_queued(ctmr)) {
 		timerqueue_del(ctmr->head, &ctmr->node);
 		ctmr->head = NULL;
+		return true;
 	}
+	return false;
 }
 
 static inline u64 cpu_timer_getexpires(struct cpu_timer *ctmr)

include/linux/sched/signal.h

@@ -714,6 +714,12 @@ static inline void unlock_task_sighand(struct task_struct *task,
 	spin_unlock_irqrestore(&task->sighand->siglock, *flags);
 }
 
+#ifdef CONFIG_LOCKDEP
+extern void lockdep_assert_task_sighand_held(struct task_struct *task);
+#else
+static inline void lockdep_assert_task_sighand_held(struct task_struct *task) { }
+#endif
+
 static inline unsigned long task_rlimit(const struct task_struct *task,
 		unsigned int limit)
 {

kernel/signal.c

@@ -1413,6 +1413,21 @@ struct sighand_struct *__lock_task_sighand(struct task_struct *tsk,
 	return sighand;
 }
 
+#ifdef CONFIG_LOCKDEP
+void lockdep_assert_task_sighand_held(struct task_struct *task)
+{
+	struct sighand_struct *sighand;
+
+	rcu_read_lock();
+	sighand = rcu_dereference(task->sighand);
+	if (sighand)
+		lockdep_assert_held(&sighand->siglock);
+	else
+		WARN_ON_ONCE(1);
+	rcu_read_unlock();
+}
+#endif
+
 /*
  * send signal info to all the members of a group
  */

kernel/time/clocksource-wdtest.c

@@ -19,6 +19,8 @@
 #include <linux/prandom.h>
 #include <linux/cpu.h>
 
+#include "tick-internal.h"
+
 MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Paul E. McKenney <paulmck@kernel.org>");
 
@@ -34,9 +36,6 @@ static u64 wdtest_jiffies_read(struct clocksource *cs)
 	return (u64)jiffies;
 }
 
-/* Assume HZ > 100. */
-#define JIFFIES_SHIFT	8
-
 static struct clocksource clocksource_wdtest_jiffies = {
 	.name			= "wdtest-jiffies",
 	.rating			= 1, /* lowest valid rating*/

kernel/time/clocksource.c

@@ -306,12 +306,12 @@ void clocksource_verify_percpu(struct clocksource *cs)
 		return;
 	cpumask_clear(&cpus_ahead);
 	cpumask_clear(&cpus_behind);
-	get_online_cpus();
+	cpus_read_lock();
 	preempt_disable();
 	clocksource_verify_choose_cpus();
 	if (cpumask_weight(&cpus_chosen) == 0) {
 		preempt_enable();
-		put_online_cpus();
+		cpus_read_unlock();
 		pr_warn("Not enough CPUs to check clocksource '%s'.\n", cs->name);
 		return;
 	}
@@ -337,7 +337,7 @@ void clocksource_verify_percpu(struct clocksource *cs)
 			cs_nsec_min = cs_nsec;
 	}
 	preempt_enable();
-	put_online_cpus();
+	cpus_read_unlock();
 	if (!cpumask_empty(&cpus_ahead))
 		pr_warn("        CPUs %*pbl ahead of CPU %d for clocksource %s.\n",
 			cpumask_pr_args(&cpus_ahead), testcpu, cs->name);

kernel/time/jiffies.c

@@ -10,28 +10,9 @@
 #include <linux/init.h>
 
 #include "timekeeping.h"
+#include "tick-internal.h"
 
 
-/* Since jiffies uses a simple TICK_NSEC multiplier
- * conversion, the .shift value could be zero. However
- * this would make NTP adjustments impossible as they are
- * in units of 1/2^.shift. Thus we use JIFFIES_SHIFT to
- * shift both the nominator and denominator the same
- * amount, and give ntp adjustments in units of 1/2^8
- *
- * The value 8 is somewhat carefully chosen, as anything
- * larger can result in overflows. TICK_NSEC grows as HZ
- * shrinks, so values greater than 8 overflow 32bits when
- * HZ=100.
- */
-#if HZ < 34
-#define JIFFIES_SHIFT	6
-#elif HZ < 67
-#define JIFFIES_SHIFT	7
-#else
-#define JIFFIES_SHIFT	8
-#endif
-
 static u64 jiffies_read(struct clocksource *cs)
 {
 	return (u64) jiffies;

kernel/time/posix-cpu-timers.c

@@ -291,6 +291,8 @@ static void thread_group_start_cputime(struct task_struct *tsk, u64 *samples)
 	struct thread_group_cputimer *cputimer = &tsk->signal->cputimer;
 	struct posix_cputimers *pct = &tsk->signal->posix_cputimers;
 
+	lockdep_assert_task_sighand_held(tsk);
+
 	/* Check if cputimer isn't running. This is accessed without locking. */
 	if (!READ_ONCE(pct->timers_active)) {
 		struct task_cputime sum;
@@ -405,6 +407,55 @@ static int posix_cpu_timer_create(struct k_itimer *new_timer)
 	return 0;
 }
 
+static struct posix_cputimer_base *timer_base(struct k_itimer *timer,
+					      struct task_struct *tsk)
+{
+	int clkidx = CPUCLOCK_WHICH(timer->it_clock);
+
+	if (CPUCLOCK_PERTHREAD(timer->it_clock))
+		return tsk->posix_cputimers.bases + clkidx;
+	else
+		return tsk->signal->posix_cputimers.bases + clkidx;
+}
+
+/*
+ * Force recalculating the base earliest expiration on the next tick.
+ * This will also re-evaluate the need to keep around the process wide
+ * cputime counter and tick dependency and eventually shut these down
+ * if necessary.
+ */
+static void trigger_base_recalc_expires(struct k_itimer *timer,
+					struct task_struct *tsk)
+{
+	struct posix_cputimer_base *base = timer_base(timer, tsk);
+
+	base->nextevt = 0;
+}
+
+/*
+ * Dequeue the timer and reset the base if it was its earliest expiration.
+ * It makes sure the next tick recalculates the base next expiration so we
+ * don't keep the costly process wide cputime counter around for a random
+ * amount of time, along with the tick dependency.
+ *
+ * If another timer gets queued between this and the next tick, its
+ * expiration will update the base next event if necessary on the next
+ * tick.
+ */
+static void disarm_timer(struct k_itimer *timer, struct task_struct *p)
+{
+	struct cpu_timer *ctmr = &timer->it.cpu;
+	struct posix_cputimer_base *base;
+
+	if (!cpu_timer_dequeue(ctmr))
+		return;
+
+	base = timer_base(timer, p);
+	if (cpu_timer_getexpires(ctmr) == base->nextevt)
+		trigger_base_recalc_expires(timer, p);
+}
+
+
 /*
  * Clean up a CPU-clock timer that is about to be destroyed.
  * This is called from timer deletion with the timer already locked.
@@ -439,7 +490,7 @@ static int posix_cpu_timer_del(struct k_itimer *timer)
 		if (timer->it.cpu.firing)
 			ret = TIMER_RETRY;
 		else
-			cpu_timer_dequeue(ctmr);
+			disarm_timer(timer, p);
 
 		unlock_task_sighand(p, &flags);
 	}
@@ -498,15 +549,9 @@ void posix_cpu_timers_exit_group(struct task_struct *tsk)
  */
 static void arm_timer(struct k_itimer *timer, struct task_struct *p)
 {
-	int clkidx = CPUCLOCK_WHICH(timer->it_clock);
+	struct posix_cputimer_base *base = timer_base(timer, p);
 	struct cpu_timer *ctmr = &timer->it.cpu;
 	u64 newexp = cpu_timer_getexpires(ctmr);
-	struct posix_cputimer_base *base;
-
-	if (CPUCLOCK_PERTHREAD(timer->it_clock))
-		base = p->posix_cputimers.bases + clkidx;
-	else
-		base = p->signal->posix_cputimers.bases + clkidx;
 
 	if (!cpu_timer_enqueue(&base->tqhead, ctmr))
 		return;
@@ -703,7 +748,8 @@ static int posix_cpu_timer_set(struct k_itimer *timer, int timer_flags,
 	timer->it_overrun_last = 0;
 	timer->it_overrun = -1;
 
-	if (new_expires != 0 && !(val < new_expires)) {
+	if (val >= new_expires) {
+		if (new_expires != 0) {
 			/*
 			 * The designated time already passed, so we notify
 			 * immediately, even if the thread never runs to
@@ -712,7 +758,19 @@ static int posix_cpu_timer_set(struct k_itimer *timer, int timer_flags,
 			cpu_timer_fire(timer);
 		}
 
-	ret = 0;
+		/*
+		 * Make sure we don't keep around the process wide cputime
+		 * counter or the tick dependency if they are not necessary.
+		 */
+		sighand = lock_task_sighand(p, &flags);
+		if (!sighand)
+			goto out;
+
+		if (!cpu_timer_queued(ctmr))
+			trigger_base_recalc_expires(timer, p);
+
+		unlock_task_sighand(p, &flags);
+	}
  out:
 	rcu_read_unlock();
 	if (old)
@@ -1346,8 +1404,6 @@ void set_process_cpu_timer(struct task_struct *tsk, unsigned int clkid,
 			}
 		}
 
-		if (!*newval)
-			return;
 		*newval += now;
 	}
 

kernel/time/posix-timers.c

@@ -336,7 +336,7 @@ void posixtimer_rearm(struct kernel_siginfo *info)
 int posix_timer_event(struct k_itimer *timr, int si_private)
 {
 	enum pid_type type;
-	int ret = -1;
+	int ret;
 	/*
 	 * FIXME: if ->sigq is queued we can race with
 	 * dequeue_signal()->posixtimer_rearm().

kernel/time/tick-common.c

@@ -470,6 +470,13 @@ void tick_resume_local(void)
 		else
 			tick_resume_oneshot();
 	}
+
+	/*
+	 * Ensure that hrtimers are up to date and the clockevents device
+	 * is reprogrammed correctly when high resolution timers are
+	 * enabled.
+	 */
+	hrtimers_resume_local();
 }
 
 /**

kernel/time/tick-internal.h

@@ -165,3 +165,35 @@ DECLARE_PER_CPU(struct hrtimer_cpu_base, hrtimer_bases);
 
 extern u64 get_next_timer_interrupt(unsigned long basej, u64 basem);
 void timer_clear_idle(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);
+
+/* Since jiffies uses a simple TICK_NSEC multiplier
+ * conversion, the .shift value could be zero. However
+ * this would make NTP adjustments impossible as they are
+ * in units of 1/2^.shift. Thus we use JIFFIES_SHIFT to
+ * shift both the nominator and denominator the same
+ * amount, and give ntp adjustments in units of 1/2^8
+ *
+ * The value 8 is somewhat carefully chosen, as anything
+ * larger can result in overflows. TICK_NSEC grows as HZ
+ * shrinks, so values greater than 8 overflow 32bits when
+ * HZ=100.
+ */
+#if HZ < 34
+#define JIFFIES_SHIFT	6
+#elif HZ < 67
+#define JIFFIES_SHIFT	7
+#else
+#define JIFFIES_SHIFT	8
+#endif

kernel/time/timekeeping.c

@@ -1323,8 +1323,8 @@ int do_settimeofday64(const struct timespec64 *ts)
 	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 @@ static int timekeeping_inject_offset(const struct timespec64 *ts)
 	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(const struct timespec64 *delta)
 	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
 
@@ -1810,8 +1810,10 @@ void timekeeping_resume(void)
 
 	touch_softlockup_watchdog();
 
+	/* Resume the clockevent device(s) and hrtimers */
 	tick_resume();
-	hrtimers_resume();
+	/* Notify timerfd as resume is equivalent to clock_was_set() */
+	timerfd_resume();
 }
 
 int timekeeping_suspend(void)
@@ -2125,7 +2127,7 @@ static u64 logarithmic_accumulation(struct timekeeper *tk, u64 offset,
  * timekeeping_advance - Updates the timekeeper to the current time and
  * current NTP tick length
  */
-static void timekeeping_advance(enum timekeeping_adv_mode mode)
+static bool timekeeping_advance(enum timekeeping_adv_mode mode)
 {
 	struct timekeeper *real_tk = &tk_core.timekeeper;
 	struct timekeeper *tk = &shadow_timekeeper;
@@ -2196,9 +2198,8 @@ static void timekeeping_advance(enum timekeeping_adv_mode mode)
 	write_seqcount_end(&tk_core.seq);
 out:
 	raw_spin_unlock_irqrestore(&timekeeper_lock, flags);
-	if (clock_set)
-		/* Have to call _delayed version, since in irq context*/
-		clock_was_set_delayed();
+
+	return !!clock_set;
 }
 
 /**
@@ -2207,7 +2208,8 @@ static void timekeeping_advance(enum timekeeping_adv_mode mode)
  */
 void update_wall_time(void)
 {
-	timekeeping_advance(TK_ADV_TICK);
+	if (timekeeping_advance(TK_ADV_TICK))
+		clock_was_set_delayed();
 }
 
 /**
@@ -2387,8 +2389,9 @@ int do_adjtimex(struct __kernel_timex *txc)
 {
 	struct timekeeper *tk = &tk_core.timekeeper;
 	struct audit_ntp_data ad;
-	unsigned long flags;
+	bool clock_set = false;
 	struct timespec64 ts;
+	unsigned long flags;
 	s32 orig_tai, tai;
 	int ret;
 
@@ -2423,6 +2426,7 @@ int do_adjtimex(struct __kernel_timex *txc)
 	if (tai != orig_tai) {
 		__timekeeping_set_tai_offset(tk, tai);
 		timekeeping_update(tk, TK_MIRROR | TK_CLOCK_WAS_SET);
+		clock_set = true;
 	}
 	tk_update_leap_state(tk);
 
@@ -2433,10 +2437,10 @@ int do_adjtimex(struct __kernel_timex *txc)
 
 	/* Update the multiplier immediately if frequency was set directly */
 	if (txc->modes & (ADJ_FREQUENCY | ADJ_TICK))
-		timekeeping_advance(TK_ADV_FREQ);
+		clock_set |= timekeeping_advance(TK_ADV_FREQ);
 
-	if (tai != orig_tai)
-		clock_was_set();
+	if (clock_set)
+		clock_was_set(CLOCK_REALTIME);
 
 	ntp_notify_cmos_timer();