From 2933954b71f10d392764f95eec0f0aa2d103054b Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Mon, 15 Jul 2019 13:03:48 +0200 Subject: PM: sleep: Fix possible overflow in pm_system_cancel_wakeup() It is not actually guaranteed that pm_abort_suspend will be nonzero when pm_system_cancel_wakeup() is called which may lead to subtle issues, so make it use atomic_dec_if_positive() instead of atomic_dec() for the safety sake. Fixes: 33e4f80ee69b ("ACPI / PM: Ignore spurious SCI wakeups from suspend-to-idle") Signed-off-by: Rafael J. Wysocki Acked-by: Thomas Gleixner --- drivers/base/power/wakeup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/base') diff --git a/drivers/base/power/wakeup.c b/drivers/base/power/wakeup.c index ee31d4f8d856..b30c45aad10f 100644 --- a/drivers/base/power/wakeup.c +++ b/drivers/base/power/wakeup.c @@ -859,7 +859,7 @@ EXPORT_SYMBOL_GPL(pm_system_wakeup); void pm_system_cancel_wakeup(void) { - atomic_dec(&pm_abort_suspend); + atomic_dec_if_positive(&pm_abort_suspend); } void pm_wakeup_clear(bool reset) -- cgit v1.2.3 From 56b991849009f5def0443bfb2f48c8321d888e15 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Mon, 15 Jul 2019 23:52:03 +0200 Subject: PM: sleep: Simplify suspend-to-idle control flow After commit 33e4f80ee69b ("ACPI / PM: Ignore spurious SCI wakeups from suspend-to-idle") the "noirq" phases of device suspend and resume may run for multiple times during suspend-to-idle, if there are spurious system wakeup events while suspended. However, this is complicated and fragile and actually unnecessary. The main reason for doing this is that on some systems the EC may signal system wakeup events (power button events, for example) as well as events that should not cause the system to resume (spurious system wakeup events). Thus, in order to determine whether or not a given event signaled by the EC while suspended is a proper system wakeup one, the EC GPE needs to be dispatched and to start with that was achieved by allowing the ACPI SCI action handler to run, which was only possible after calling resume_device_irqs(). However, dispatching the EC GPE this way turned out to take too much time in some cases and some EC events might be missed due to that, so commit 68e22011856f ("ACPI: EC: Dispatch the EC GPE directly on s2idle wake") started to dispatch the EC GPE right after a wakeup event has been detected, so in fact the full ACPI SCI action handler doesn't need to run any more to deal with the wakeups coming from the EC. Use this observation to simplify the suspend-to-idle control flow so that the "noirq" phases of device suspend and resume are each run only once in every suspend-to-idle cycle, which is reported to significantly reduce power drawn by some systems when suspended to idle (by allowing them to reach a deep platform-wide low-power state through the suspend-to-idle flow). [What appears to happen is that the "noirq" resume of devices after a spurious EC wakeup brings some devices into a state in which they prevent the platform from reaching the deep low-power state going forward, even after a subsequent "noirq" suspend phase, and on some systems the EC triggers such wakeups already when the "noirq" suspend of devices is running for the first time in the given suspend/resume cycle, so the platform cannot reach the deep low-power state at all.] First, make acpi_s2idle_wake() use the acpi_ec_dispatch_gpe() return value to determine whether or not the wakeup may have been triggered by the EC (in which case the system wakeup is canceled and ACPI events are processed in order to determine whether or not the event is a proper system wakeup one) and use rearm_wake_irq() (introduced by a previous change) in it to rearm the ACPI SCI for system wakeup detection in case the system will remain suspended. Second, drop acpi_s2idle_sync(), which is not needed any more, and the corresponding global platform suspend-to-idle callback. Next, drop the pm_wakeup_pending() check (which is an optimization only) from __device_suspend_noirq() to prevent it from returning errors on system wakeups occurring before the "noirq" phase of device suspend is complete (as in the case of suspend-to-idle it is not known whether or not these wakeups are suprious at that point), in order to avoid having to carry out a "noirq" resume of devices on a spurious system wakeup. Finally, change the code flow in s2idle_loop() to (1) run the "noirq" suspend of devices once before starting the loop, (2) check for spurious EC wakeups (via the platform ->wake callback) for the first time before calling s2idle_enter(), and (3) run the "noirq" resume of devices once after leaving the loop. Signed-off-by: Rafael J. Wysocki Acked-by: Thomas Gleixner --- drivers/acpi/sleep.c | 47 +++++++++++++++++++++-------------------- drivers/base/power/main.c | 5 ----- include/linux/suspend.h | 1 - kernel/power/suspend.c | 53 ++++++++++++++++++++--------------------------- 4 files changed, 48 insertions(+), 58 deletions(-) (limited to 'drivers/base') diff --git a/drivers/acpi/sleep.c b/drivers/acpi/sleep.c index 3debe1a42655..970ae7c7a3f7 100644 --- a/drivers/acpi/sleep.c +++ b/drivers/acpi/sleep.c @@ -986,33 +986,37 @@ static void acpi_s2idle_wake(void) lpi_check_constraints(); /* - * If IRQD_WAKEUP_ARMED is not set for the SCI at this point, it means - * that the SCI has triggered while suspended, so cancel the wakeup in - * case it has not been a wakeup event (the GPEs will be checked later). + * If IRQD_WAKEUP_ARMED is set for the SCI at this point, the SCI has + * not triggered while suspended, so bail out. */ - if (acpi_sci_irq_valid() && - !irqd_is_wakeup_armed(irq_get_irq_data(acpi_sci_irq))) { + if (!acpi_sci_irq_valid() || + irqd_is_wakeup_armed(irq_get_irq_data(acpi_sci_irq))) + return; + + /* + * If there are EC events to process, the wakeup may be a spurious one + * coming from the EC. + */ + if (acpi_ec_dispatch_gpe()) { + /* + * Cancel the wakeup and process all pending events in case + * there are any wakeup ones in there. + * + * Note that if any non-EC GPEs are active at this point, the + * SCI will retrigger after the rearming below, so no events + * should be missed by canceling the wakeup here. + */ pm_system_cancel_wakeup(); /* - * On some platforms with the LPS0 _DSM device noirq resume - * takes too much time for EC wakeup events to survive, so look - * for them now. + * The EC driver uses the system workqueue and an additional + * special one, so those need to be flushed too. */ - acpi_ec_dispatch_gpe(); + acpi_os_wait_events_complete(); /* synchronize EC GPE processing */ + acpi_ec_flush_work(); + acpi_os_wait_events_complete(); /* synchronize Notify handling */ } -} -static void acpi_s2idle_sync(void) -{ - /* - * Process all pending events in case there are any wakeup ones. - * - * The EC driver uses the system workqueue and an additional special - * one, so those need to be flushed too. - */ - acpi_os_wait_events_complete(); /* synchronize SCI IRQ handling */ - acpi_ec_flush_work(); - acpi_os_wait_events_complete(); /* synchronize Notify handling */ + rearm_wake_irq(acpi_sci_irq); } static void acpi_s2idle_restore(void) @@ -1044,7 +1048,6 @@ static const struct platform_s2idle_ops acpi_s2idle_ops = { .begin = acpi_s2idle_begin, .prepare = acpi_s2idle_prepare, .wake = acpi_s2idle_wake, - .sync = acpi_s2idle_sync, .restore = acpi_s2idle_restore, .end = acpi_s2idle_end, }; diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c index 7fb2c39bc725..f08332fab531 100644 --- a/drivers/base/power/main.c +++ b/drivers/base/power/main.c @@ -1291,11 +1291,6 @@ static int __device_suspend_noirq(struct device *dev, pm_message_t state, bool a if (async_error) goto Complete; - if (pm_wakeup_pending()) { - async_error = -EBUSY; - goto Complete; - } - if (dev->power.syscore || dev->power.direct_complete) goto Complete; diff --git a/include/linux/suspend.h b/include/linux/suspend.h index 9c0ad1a3a727..66ce3871ed61 100644 --- a/include/linux/suspend.h +++ b/include/linux/suspend.h @@ -191,7 +191,6 @@ struct platform_s2idle_ops { int (*begin)(void); int (*prepare)(void); void (*wake)(void); - void (*sync)(void); void (*restore)(void); void (*end)(void); }; diff --git a/kernel/power/suspend.c b/kernel/power/suspend.c index c874a7026e24..907b2be0372f 100644 --- a/kernel/power/suspend.c +++ b/kernel/power/suspend.c @@ -119,48 +119,41 @@ static void s2idle_enter(void) static void s2idle_loop(void) { + int error; + + dpm_noirq_begin(); + error = dpm_noirq_suspend_devices(PMSG_SUSPEND); + if (error) + goto resume; + pm_pr_dbg("suspend-to-idle\n"); + /* + * Suspend-to-idle equals: + * frozen processes + suspended devices + idle processors. + * Thus s2idle_enter() should be called right after all devices have + * been suspended. + * + * Wakeups during the noirq suspend of devices may be spurious, so try + * to avoid them upfront. + */ for (;;) { - int error; - - dpm_noirq_begin(); - - /* - * Suspend-to-idle equals - * frozen processes + suspended devices + idle processors. - * Thus s2idle_enter() should be called right after - * all devices have been suspended. - * - * Wakeups during the noirq suspend of devices may be spurious, - * so prevent them from terminating the loop right away. - */ - error = dpm_noirq_suspend_devices(PMSG_SUSPEND); - if (!error) - s2idle_enter(); - else if (error == -EBUSY && pm_wakeup_pending()) - error = 0; - - if (!error && s2idle_ops && s2idle_ops->wake) + if (s2idle_ops && s2idle_ops->wake) s2idle_ops->wake(); - dpm_noirq_resume_devices(PMSG_RESUME); - - dpm_noirq_end(); - - if (error) - break; - - if (s2idle_ops && s2idle_ops->sync) - s2idle_ops->sync(); - if (pm_wakeup_pending()) break; pm_wakeup_clear(false); + + s2idle_enter(); } pm_pr_dbg("resume from suspend-to-idle\n"); + +resume: + dpm_noirq_resume_devices(PMSG_RESUME); + dpm_noirq_end(); } void s2idle_wake(void) -- cgit v1.2.3 From b605c44c30b59990e806f930c37bd288b9d901a5 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Mon, 15 Jul 2019 23:52:18 +0200 Subject: PM: sleep: Drop dpm_noirq_begin() and dpm_noirq_end() Note that after previous changes dpm_noirq_begin() and dpm_noirq_end() each have only one caller, so move the code from them to their respective callers and drop them. Also note that dpm_noirq_resume_devices() and dpm_noirq_suspend_devices() need not be exported any more, so make them both static. This change is not expected to alter functionality. Signed-off-by: Rafael J. Wysocki Acked-by: Thomas Gleixner --- drivers/base/power/main.c | 30 ++++++++++++------------------ include/linux/pm.h | 4 ---- 2 files changed, 12 insertions(+), 22 deletions(-) (limited to 'drivers/base') diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c index f08332fab531..134a8af51511 100644 --- a/drivers/base/power/main.c +++ b/drivers/base/power/main.c @@ -716,7 +716,7 @@ static void async_resume_noirq(void *data, async_cookie_t cookie) put_device(dev); } -void dpm_noirq_resume_devices(pm_message_t state) +static void dpm_noirq_resume_devices(pm_message_t state) { struct device *dev; ktime_t starttime = ktime_get(); @@ -760,13 +760,6 @@ void dpm_noirq_resume_devices(pm_message_t state) trace_suspend_resume(TPS("dpm_resume_noirq"), state.event, false); } -void dpm_noirq_end(void) -{ - resume_device_irqs(); - device_wakeup_disarm_wake_irqs(); - cpuidle_resume(); -} - /** * dpm_resume_noirq - Execute "noirq resume" callbacks for all devices. * @state: PM transition of the system being carried out. @@ -777,7 +770,11 @@ void dpm_noirq_end(void) void dpm_resume_noirq(pm_message_t state) { dpm_noirq_resume_devices(state); - dpm_noirq_end(); + + resume_device_irqs(); + device_wakeup_disarm_wake_irqs(); + + cpuidle_resume(); } static pm_callback_t dpm_subsys_resume_early_cb(struct device *dev, @@ -1357,14 +1354,7 @@ static int device_suspend_noirq(struct device *dev) return __device_suspend_noirq(dev, pm_transition, false); } -void dpm_noirq_begin(void) -{ - cpuidle_pause(); - device_wakeup_arm_wake_irqs(); - suspend_device_irqs(); -} - -int dpm_noirq_suspend_devices(pm_message_t state) +static int dpm_noirq_suspend_devices(pm_message_t state) { ktime_t starttime = ktime_get(); int error = 0; @@ -1421,7 +1411,11 @@ int dpm_suspend_noirq(pm_message_t state) { int ret; - dpm_noirq_begin(); + cpuidle_pause(); + + device_wakeup_arm_wake_irqs(); + suspend_device_irqs(); + ret = dpm_noirq_suspend_devices(state); if (ret) dpm_resume_noirq(resume_event(state)); diff --git a/include/linux/pm.h b/include/linux/pm.h index 3619a870eaa4..4c441be03079 100644 --- a/include/linux/pm.h +++ b/include/linux/pm.h @@ -712,8 +712,6 @@ struct dev_pm_domain { extern void device_pm_lock(void); extern void dpm_resume_start(pm_message_t state); extern void dpm_resume_end(pm_message_t state); -extern void dpm_noirq_resume_devices(pm_message_t state); -extern void dpm_noirq_end(void); extern void dpm_resume_noirq(pm_message_t state); extern void dpm_resume_early(pm_message_t state); extern void dpm_resume(pm_message_t state); @@ -722,8 +720,6 @@ extern void dpm_complete(pm_message_t state); extern void device_pm_unlock(void); extern int dpm_suspend_end(pm_message_t state); extern int dpm_suspend_start(pm_message_t state); -extern void dpm_noirq_begin(void); -extern int dpm_noirq_suspend_devices(pm_message_t state); extern int dpm_suspend_noirq(pm_message_t state); extern int dpm_suspend_late(pm_message_t state); extern int dpm_suspend(pm_message_t state); -- cgit v1.2.3