Merge branch 'pm-sleep'

Merge updates related to system sleep for 5.16-rc1:

 - Clean up hib_wait_io() (Falla Coulibaly).

 - Fix sparse warnings in hibernation-related code (Anders Roxell).

 - Use vzalloc() and kzalloc() instead of their open-coded
   equivalents in hibernation-related code (Cai Huoqing).

 - Prevent user space from crashing the kernel by attempting to
   restore the system state from a swap partition in use (Ye Bin).

 - Do not let "syscore" devices runtime-suspend during system PM
   transitions (Rafael Wysocki).

 - Do not pause cpuidle in the suspend-to-idle path (Rafael Wysocki).

 - Pause cpuidle later and resume it earlier during system PM
   transitions (Rafael Wysocki).

 - Make system suspend code use valid_state() consistently (Rafael
   Wysocki).

 - Add support for enabling wakeup IRQs after invoking the
   ->runtime_suspend() callback and make two drivers use it (Chunfeng
   Yun).

* pm-sleep:
  usb: mtu3: enable wake-up interrupt after runtime_suspend called
  usb: xhci-mtk: enable wake-up interrupt after runtime_suspend called
  PM / wakeirq: support enabling wake-up irq after runtime_suspend called
  PM: suspend: Use valid_state() consistently
  PM: sleep: Pause cpuidle later and resume it earlier during system transitions
  PM: suspend: Do not pause cpuidle in the suspend-to-idle path
  PM: sleep: Do not let "syscore" devices runtime-suspend during system transitions
  PM: hibernate: Get block device exclusively in swsusp_check()
  PM: hibernate: swap: Use vzalloc() and kzalloc()
  PM: hibernate: fix sparse warnings
  Revert "PM: sleep: Do not assume that "mem" is always present"
  PM: hibernate: Remove blk_status_to_errno in hib_wait_io
  PM: sleep: Do not assume that "mem" is always present

drivers/base/power/main.c

@@ -32,7 +32,6 @@
 #include <linux/suspend.h>
 #include <trace/events/power.h>
 #include <linux/cpufreq.h>
-#include <linux/cpuidle.h>
 #include <linux/devfreq.h>
 #include <linux/timer.h>
 
@@ -747,8 +746,6 @@ void dpm_resume_noirq(pm_message_t state)
 
 	resume_device_irqs();
 	device_wakeup_disarm_wake_irqs();
-
-	cpuidle_resume();
 }
 
 /**
@@ -1051,7 +1048,7 @@ static void device_complete(struct device *dev, pm_message_t state)
 	const char *info = NULL;
 
 	if (dev->power.syscore)
-		return;
+		goto out;
 
 	device_lock(dev);
 
@@ -1081,6 +1078,7 @@ static void device_complete(struct device *dev, pm_message_t state)
 
 	device_unlock(dev);
 
+out:
 	pm_runtime_put(dev);
 }
 
@@ -1336,8 +1334,6 @@ int dpm_suspend_noirq(pm_message_t state)
 {
 	int ret;
 
-	cpuidle_pause();
-
 	device_wakeup_arm_wake_irqs();
 	suspend_device_irqs();
 
@@ -1794,9 +1790,6 @@ static int device_prepare(struct device *dev, pm_message_t state)
 	int (*callback)(struct device *) = NULL;
 	int ret = 0;
 
-	if (dev->power.syscore)
-		return 0;
-
 	/*
 	 * If a device's parent goes into runtime suspend at the wrong time,
 	 * it won't be possible to resume the device.  To prevent this we
@@ -1805,6 +1798,9 @@ static int device_prepare(struct device *dev, pm_message_t state)
 	 */
 	pm_runtime_get_noresume(dev);
 
+	if (dev->power.syscore)
+		return 0;
+
 	device_lock(dev);
 
 	dev->power.wakeup_path = false;

drivers/base/power/power.h

@@ -25,8 +25,10 @@ extern u64 pm_runtime_active_time(struct device *dev);
 
 #define WAKE_IRQ_DEDICATED_ALLOCATED	BIT(0)
 #define WAKE_IRQ_DEDICATED_MANAGED	BIT(1)
+#define WAKE_IRQ_DEDICATED_REVERSE	BIT(2)
 #define WAKE_IRQ_DEDICATED_MASK		(WAKE_IRQ_DEDICATED_ALLOCATED | \
-					 WAKE_IRQ_DEDICATED_MANAGED)
+					 WAKE_IRQ_DEDICATED_MANAGED | \
+					 WAKE_IRQ_DEDICATED_REVERSE)
 
 struct wake_irq {
 	struct device *dev;
@@ -39,7 +41,8 @@ extern void dev_pm_arm_wake_irq(struct wake_irq *wirq);
 extern void dev_pm_disarm_wake_irq(struct wake_irq *wirq);
 extern void dev_pm_enable_wake_irq_check(struct device *dev,
 					 bool can_change_status);
-extern void dev_pm_disable_wake_irq_check(struct device *dev);
+extern void dev_pm_disable_wake_irq_check(struct device *dev, bool cond_disable);
+extern void dev_pm_enable_wake_irq_complete(struct device *dev);
 
 #ifdef CONFIG_PM_SLEEP
 

drivers/base/power/runtime.c

@@ -645,6 +645,8 @@ static int rpm_suspend(struct device *dev, int rpmflags)
 	if (retval)
 		goto fail;
 
+	dev_pm_enable_wake_irq_complete(dev);
+
  no_callback:
 	__update_runtime_status(dev, RPM_SUSPENDED);
 	pm_runtime_deactivate_timer(dev);
@@ -690,7 +692,7 @@ static int rpm_suspend(struct device *dev, int rpmflags)
 	return retval;
 
  fail:
-	dev_pm_disable_wake_irq_check(dev);
+	dev_pm_disable_wake_irq_check(dev, true);
 	__update_runtime_status(dev, RPM_ACTIVE);
 	dev->power.deferred_resume = false;
 	wake_up_all(&dev->power.wait_queue);
@@ -873,7 +875,7 @@ static int rpm_resume(struct device *dev, int rpmflags)
 
 	callback = RPM_GET_CALLBACK(dev, runtime_resume);
 
-	dev_pm_disable_wake_irq_check(dev);
+	dev_pm_disable_wake_irq_check(dev, false);
 	retval = rpm_callback(callback, dev);
 	if (retval) {
 		__update_runtime_status(dev, RPM_SUSPENDED);

drivers/base/power/wakeirq.c

@@ -142,24 +142,7 @@ static irqreturn_t handle_threaded_wake_irq(int irq, void *_wirq)
 	return IRQ_HANDLED;
 }
 
-/**
- * dev_pm_set_dedicated_wake_irq - Request a dedicated wake-up interrupt
- * @dev: Device entry
- * @irq: Device wake-up interrupt
- *
- * Unless your hardware has separate wake-up interrupts in addition
- * to the device IO interrupts, you don't need this.
- *
- * Sets up a threaded interrupt handler for a device that has
- * a dedicated wake-up interrupt in addition to the device IO
- * interrupt.
- *
- * The interrupt starts disabled, and needs to be managed for
- * the device by the bus code or the device driver using
- * dev_pm_enable_wake_irq() and dev_pm_disable_wake_irq()
- * functions.
- */
-int dev_pm_set_dedicated_wake_irq(struct device *dev, int irq)
+static int __dev_pm_set_dedicated_wake_irq(struct device *dev, int irq, unsigned int flag)
 {
 	struct wake_irq *wirq;
 	int err;
@@ -197,7 +180,7 @@ int dev_pm_set_dedicated_wake_irq(struct device *dev, int irq)
 	if (err)
 		goto err_free_irq;
 
-	wirq->status = WAKE_IRQ_DEDICATED_ALLOCATED;
+	wirq->status = WAKE_IRQ_DEDICATED_ALLOCATED | flag;
 
 	return err;
 
@@ -210,8 +193,57 @@ int dev_pm_set_dedicated_wake_irq(struct device *dev, int irq)
 
 	return err;
 }
+
+
+/**
+ * dev_pm_set_dedicated_wake_irq - Request a dedicated wake-up interrupt
+ * @dev: Device entry
+ * @irq: Device wake-up interrupt
+ *
+ * Unless your hardware has separate wake-up interrupts in addition
+ * to the device IO interrupts, you don't need this.
+ *
+ * Sets up a threaded interrupt handler for a device that has
+ * a dedicated wake-up interrupt in addition to the device IO
+ * interrupt.
+ *
+ * The interrupt starts disabled, and needs to be managed for
+ * the device by the bus code or the device driver using
+ * dev_pm_enable_wake_irq*() and dev_pm_disable_wake_irq*()
+ * functions.
+ */
+int dev_pm_set_dedicated_wake_irq(struct device *dev, int irq)
+{
+	return __dev_pm_set_dedicated_wake_irq(dev, irq, 0);
+}
 EXPORT_SYMBOL_GPL(dev_pm_set_dedicated_wake_irq);
 
+/**
+ * dev_pm_set_dedicated_wake_irq_reverse - Request a dedicated wake-up interrupt
+ *                                         with reverse enable ordering
+ * @dev: Device entry
+ * @irq: Device wake-up interrupt
+ *
+ * Unless your hardware has separate wake-up interrupts in addition
+ * to the device IO interrupts, you don't need this.
+ *
+ * Sets up a threaded interrupt handler for a device that has a dedicated
+ * wake-up interrupt in addition to the device IO interrupt. It sets
+ * the status of WAKE_IRQ_DEDICATED_REVERSE to tell rpm_suspend()
+ * to enable dedicated wake-up interrupt after running the runtime suspend
+ * callback for @dev.
+ *
+ * The interrupt starts disabled, and needs to be managed for
+ * the device by the bus code or the device driver using
+ * dev_pm_enable_wake_irq*() and dev_pm_disable_wake_irq*()
+ * functions.
+ */
+int dev_pm_set_dedicated_wake_irq_reverse(struct device *dev, int irq)
+{
+	return __dev_pm_set_dedicated_wake_irq(dev, irq, WAKE_IRQ_DEDICATED_REVERSE);
+}
+EXPORT_SYMBOL_GPL(dev_pm_set_dedicated_wake_irq_reverse);
+
 /**
  * dev_pm_enable_wake_irq - Enable device wake-up interrupt
  * @dev: Device
@@ -282,27 +314,54 @@ void dev_pm_enable_wake_irq_check(struct device *dev,
 	return;
 
 enable:
-	enable_irq(wirq->irq);
+	if (!can_change_status || !(wirq->status & WAKE_IRQ_DEDICATED_REVERSE))
+		enable_irq(wirq->irq);
 }
 
 /**
  * dev_pm_disable_wake_irq_check - Checks and disables wake-up interrupt
  * @dev: Device
+ * @cond_disable: if set, also check WAKE_IRQ_DEDICATED_REVERSE
  *
  * Disables wake-up interrupt conditionally based on status.
  * Should be only called from rpm_suspend() and rpm_resume() path.
  */
-void dev_pm_disable_wake_irq_check(struct device *dev)
+void dev_pm_disable_wake_irq_check(struct device *dev, bool cond_disable)
 {
 	struct wake_irq *wirq = dev->power.wakeirq;
 
 	if (!wirq || !(wirq->status & WAKE_IRQ_DEDICATED_MASK))
 		return;
 
+	if (cond_disable && (wirq->status & WAKE_IRQ_DEDICATED_REVERSE))
+		return;
+
 	if (wirq->status & WAKE_IRQ_DEDICATED_MANAGED)
 		disable_irq_nosync(wirq->irq);
 }
 
+/**
+ * dev_pm_enable_wake_irq_complete - enable wake IRQ not enabled before
+ * @dev: Device using the wake IRQ
+ *
+ * Enable wake IRQ conditionally based on status, mainly used if want to
+ * enable wake IRQ after running ->runtime_suspend() which depends on
+ * WAKE_IRQ_DEDICATED_REVERSE.
+ *
+ * Should be only called from rpm_suspend() path.
+ */
+void dev_pm_enable_wake_irq_complete(struct device *dev)
+{
+	struct wake_irq *wirq = dev->power.wakeirq;
+
+	if (!wirq || !(wirq->status & WAKE_IRQ_DEDICATED_MASK))
+		return;
+
+	if (wirq->status & WAKE_IRQ_DEDICATED_MANAGED &&
+	    wirq->status & WAKE_IRQ_DEDICATED_REVERSE)
+		enable_irq(wirq->irq);
+}
+
 /**
  * dev_pm_arm_wake_irq - Arm device wake-up
  * @wirq: Device wake-up interrupt

drivers/usb/host/xhci-mtk.c

@@ -602,7 +602,7 @@ static int xhci_mtk_probe(struct platform_device *pdev)
 		goto dealloc_usb2_hcd;
 
 	if (wakeup_irq > 0) {
-		ret = dev_pm_set_dedicated_wake_irq(dev, wakeup_irq);
+		ret = dev_pm_set_dedicated_wake_irq_reverse(dev, wakeup_irq);
 		if (ret) {
 			dev_err(dev, "set wakeup irq %d failed\n", wakeup_irq);
 			goto dealloc_usb3_hcd;

drivers/usb/mtu3/mtu3_plat.c

@@ -337,7 +337,7 @@ static int mtu3_probe(struct platform_device *pdev)
 		goto comm_init_err;
 
 	if (ssusb->wakeup_irq > 0) {
-		ret = dev_pm_set_dedicated_wake_irq(dev, ssusb->wakeup_irq);
+		ret = dev_pm_set_dedicated_wake_irq_reverse(dev, ssusb->wakeup_irq);
 		if (ret) {
 			dev_err(dev, "failed to set wakeup irq %d\n", ssusb->wakeup_irq);
 			goto comm_exit;

include/linux/pm_wakeirq.h

@@ -17,8 +17,8 @@
 #ifdef CONFIG_PM
 
 extern int dev_pm_set_wake_irq(struct device *dev, int irq);
-extern int dev_pm_set_dedicated_wake_irq(struct device *dev,
-					 int irq);
+extern int dev_pm_set_dedicated_wake_irq(struct device *dev, int irq);
+extern int dev_pm_set_dedicated_wake_irq_reverse(struct device *dev, int irq);
 extern void dev_pm_clear_wake_irq(struct device *dev);
 extern void dev_pm_enable_wake_irq(struct device *dev);
 extern void dev_pm_disable_wake_irq(struct device *dev);
@@ -35,6 +35,11 @@ static inline int dev_pm_set_dedicated_wake_irq(struct device *dev, int irq)
 	return 0;
 }
 
+static inline int dev_pm_set_dedicated_wake_irq_reverse(struct device *dev, int irq)
+{
+	return 0;
+}
+
 static inline void dev_pm_clear_wake_irq(struct device *dev)
 {
 }

kernel/power/hibernate.c

@@ -300,7 +300,7 @@ static int create_image(int platform_mode)
 	if (error || hibernation_test(TEST_PLATFORM))
 		goto Platform_finish;
 
-	error = suspend_disable_secondary_cpus();
+	error = pm_sleep_disable_secondary_cpus();
 	if (error || hibernation_test(TEST_CPUS))
 		goto Enable_cpus;
 
@@ -342,7 +342,7 @@ static int create_image(int platform_mode)
 	local_irq_enable();
 
  Enable_cpus:
-	suspend_enable_secondary_cpus();
+	pm_sleep_enable_secondary_cpus();
 
 	/* Allow architectures to do nosmt-specific post-resume dances */
 	if (!in_suspend)
@@ -466,6 +466,8 @@ static int resume_target_kernel(bool platform_mode)
 	if (error)
 		goto Cleanup;
 
+	cpuidle_pause();
+
 	error = hibernate_resume_nonboot_cpu_disable();
 	if (error)
 		goto Enable_cpus;
@@ -509,7 +511,7 @@ static int resume_target_kernel(bool platform_mode)
 	local_irq_enable();
 
  Enable_cpus:
-	suspend_enable_secondary_cpus();
+	pm_sleep_enable_secondary_cpus();
 
  Cleanup:
 	platform_restore_cleanup(platform_mode);
@@ -587,7 +589,7 @@ int hibernation_platform_enter(void)
 	if (error)
 		goto Platform_finish;
 
-	error = suspend_disable_secondary_cpus();
+	error = pm_sleep_disable_secondary_cpus();
 	if (error)
 		goto Enable_cpus;
 
@@ -609,7 +611,7 @@ int hibernation_platform_enter(void)
 	local_irq_enable();
 
  Enable_cpus:
-	suspend_enable_secondary_cpus();
+	pm_sleep_enable_secondary_cpus();
 
  Platform_finish:
 	hibernation_ops->finish();

kernel/power/power.h

@@ -4,6 +4,8 @@
 #include <linux/utsname.h>
 #include <linux/freezer.h>
 #include <linux/compiler.h>
+#include <linux/cpu.h>
+#include <linux/cpuidle.h>
 
 struct swsusp_info {
 	struct new_utsname	uts;
@@ -310,3 +312,15 @@ extern int pm_wake_lock(const char *buf);
 extern int pm_wake_unlock(const char *buf);
 
 #endif /* !CONFIG_PM_WAKELOCKS */
+
+static inline int pm_sleep_disable_secondary_cpus(void)
+{
+	cpuidle_pause();
+	return suspend_disable_secondary_cpus();
+}
+
+static inline void pm_sleep_enable_secondary_cpus(void)
+{
+	suspend_enable_secondary_cpus();
+	cpuidle_resume();
+}

kernel/power/suspend.c

@@ -97,7 +97,6 @@ static void s2idle_enter(void)
 	raw_spin_unlock_irq(&s2idle_lock);
 
 	cpus_read_lock();
-	cpuidle_resume();
 
 	/* Push all the CPUs into the idle loop. */
 	wake_up_all_idle_cpus();
@@ -105,7 +104,6 @@ static void s2idle_enter(void)
 	swait_event_exclusive(s2idle_wait_head,
 		    s2idle_state == S2IDLE_STATE_WAKE);
 
-	cpuidle_pause();
 	cpus_read_unlock();
 
 	raw_spin_lock_irq(&s2idle_lock);
@@ -162,11 +160,13 @@ EXPORT_SYMBOL_GPL(s2idle_wake);
 static bool valid_state(suspend_state_t state)
 {
 	/*
-	 * PM_SUSPEND_STANDBY and PM_SUSPEND_MEM states need low level
-	 * support and need to be valid to the low level
-	 * implementation, no valid callback implies that none are valid.
+	 * The PM_SUSPEND_STANDBY and PM_SUSPEND_MEM states require low-level
+	 * support and need to be valid to the low-level implementation.
+	 *
+	 * No ->valid() or ->enter() callback implies that none are valid.
 	 */
-	return suspend_ops && suspend_ops->valid && suspend_ops->valid(state);
+	return suspend_ops && suspend_ops->valid && suspend_ops->valid(state) &&
+		suspend_ops->enter;
 }
 
 void __init pm_states_init(void)
@@ -238,7 +238,7 @@ EXPORT_SYMBOL_GPL(suspend_valid_only_mem);
 
 static bool sleep_state_supported(suspend_state_t state)
 {
-	return state == PM_SUSPEND_TO_IDLE || (suspend_ops && suspend_ops->enter);
+	return state == PM_SUSPEND_TO_IDLE || valid_state(state);
 }
 
 static int platform_suspend_prepare(suspend_state_t state)
@@ -422,7 +422,7 @@ static int suspend_enter(suspend_state_t state, bool *wakeup)
 		goto Platform_wake;
 	}
 
-	error = suspend_disable_secondary_cpus();
+	error = pm_sleep_disable_secondary_cpus();
 	if (error || suspend_test(TEST_CPUS))
 		goto Enable_cpus;
 
@@ -452,7 +452,7 @@ static int suspend_enter(suspend_state_t state, bool *wakeup)
 	BUG_ON(irqs_disabled());
 
  Enable_cpus:
-	suspend_enable_secondary_cpus();
+	pm_sleep_enable_secondary_cpus();
 
  Platform_wake:
 	platform_resume_noirq(state);

kernel/power/swap.c

@@ -299,7 +299,7 @@ static int hib_submit_io(int op, int op_flags, pgoff_t page_off, void *addr,
 	return error;
 }
 
-static blk_status_t hib_wait_io(struct hib_bio_batch *hb)
+static int hib_wait_io(struct hib_bio_batch *hb)
 {
 	/*
 	 * We are relying on the behavior of blk_plug that a thread with
@@ -705,22 +705,19 @@ static int save_image_lzo(struct swap_map_handle *handle,
 		goto out_clean;
 	}
 
-	data = vmalloc(array_size(nr_threads, sizeof(*data)));
+	data = vzalloc(array_size(nr_threads, sizeof(*data)));
 	if (!data) {
 		pr_err("Failed to allocate LZO data\n");
 		ret = -ENOMEM;
 		goto out_clean;
 	}
-	for (thr = 0; thr < nr_threads; thr++)
-		memset(&data[thr], 0, offsetof(struct cmp_data, go));
 
-	crc = kmalloc(sizeof(*crc), GFP_KERNEL);
+	crc = kzalloc(sizeof(*crc), GFP_KERNEL);
 	if (!crc) {
 		pr_err("Failed to allocate crc\n");
 		ret = -ENOMEM;
 		goto out_clean;
 	}
-	memset(crc, 0, offsetof(struct crc_data, go));
 
 	/*
 	 * Start the compression threads.
@@ -1198,22 +1195,19 @@ static int load_image_lzo(struct swap_map_handle *handle,
 		goto out_clean;
 	}
 
-	data = vmalloc(array_size(nr_threads, sizeof(*data)));
+	data = vzalloc(array_size(nr_threads, sizeof(*data)));
 	if (!data) {
 		pr_err("Failed to allocate LZO data\n");
 		ret = -ENOMEM;
 		goto out_clean;
 	}
-	for (thr = 0; thr < nr_threads; thr++)
-		memset(&data[thr], 0, offsetof(struct dec_data, go));
 
-	crc = kmalloc(sizeof(*crc), GFP_KERNEL);
+	crc = kzalloc(sizeof(*crc), GFP_KERNEL);
 	if (!crc) {
 		pr_err("Failed to allocate crc\n");
 		ret = -ENOMEM;
 		goto out_clean;
 	}
-	memset(crc, 0, offsetof(struct crc_data, go));
 
 	clean_pages_on_decompress = true;
 
@@ -1521,9 +1515,10 @@ int swsusp_read(unsigned int *flags_p)
 int swsusp_check(void)
 {
 	int error;
+	void *holder;
 
 	hib_resume_bdev = blkdev_get_by_dev(swsusp_resume_device,
-					    FMODE_READ, NULL);
+					    FMODE_READ | FMODE_EXCL, &holder);
 	if (!IS_ERR(hib_resume_bdev)) {
 		set_blocksize(hib_resume_bdev, PAGE_SIZE);
 		clear_page(swsusp_header);
@@ -1545,7 +1540,7 @@ int swsusp_check(void)
 
 put:
 		if (error)
-			blkdev_put(hib_resume_bdev, FMODE_READ);
+			blkdev_put(hib_resume_bdev, FMODE_READ | FMODE_EXCL);
 		else
 			pr_debug("Image signature found, resuming\n");
 	} else {