Merge branch 'pci/hotplug'

- Simplify SHPC existence/permission checks (Bjorn Helgaas)

  - Remove hotplug sample skeleton driver (Lukas Wunner)

  - Convert pciehp to threaded IRQ handling (Lukas Wunner)

  - Improve pciehp tolerance of missed events and initially unstable links
    (Lukas Wunner)

  - Clear spurious pciehp events on resume (Lukas Wunner)

  - Add pciehp runtime PM support, including for Thunderbolt controllers
    (Lukas Wunner)

  - Support interrupts from pciehp bridges in D3hot (Lukas Wunner)

* pci/hotplug:
  PCI: pciehp: Deduplicate presence check on probe & resume
  PCI: pciehp: Avoid implicit fallthroughs in switch statements
  PCI: Whitelist Thunderbolt ports for runtime D3
  PCI: Whitelist native hotplug ports for runtime D3
  PCI: sysfs: Resume to D0 on function reset
  PCI: pciehp: Resume parent to D0 on config space access
  PCI: pciehp: Resume to D0 on enable/disable
  PCI: pciehp: Support interrupts sent from D3hot
  PCI: pciehp: Obey compulsory command delay after resume
  PCI: pciehp: Clear spurious events earlier on resume
  PCI: portdrv: Deduplicate PM callback iterator
  PCI: pciehp: Avoid slot access during reset
  PCI: pciehp: Always enable occupied slot on probe
  PCI: pciehp: Become resilient to missed events
  PCI: pciehp: Tolerate initially unstable link
  PCI: pciehp: Declare pciehp_enable/disable_slot() static
  PCI: pciehp: Drop enable/disable lock
  PCI: pciehp: Enable/disable exclusively from IRQ thread
  PCI: pciehp: Track enable/disable status
  PCI: pciehp: Publish to user space last on probe
  PCI: hotplug: Demidlayer registration with the core
  PCI: pciehp: Drop slot workqueue
  PCI: pciehp: Handle events synchronously
  PCI: pciehp: Stop blinking on slot enable failure
  PCI: pciehp: Convert to threaded polling
  PCI: pciehp: Convert to threaded IRQ
  PCI: pciehp: Document struct slot and struct controller
  PCI: pciehp: Declare pciehp_unconfigure_device() void
  PCI: pciehp: Drop unnecessary NULL pointer check
  PCI: pciehp: Fix unprotected list iteration in IRQ handler
  PCI: pciehp: Fix use-after-free on unplug
  PCI: hotplug: Don't leak pci_slot on registration failure
  PCI: hotplug: Delete skeleton driver
  PCI: shpchp: Separate existence of SHPC and permission to use it

drivers/pci/hotplug/acpi_pcihp.c

@@ -73,20 +73,6 @@ int acpi_get_hp_hw_control_from_firmware(struct pci_dev *pdev)
 	acpi_handle chandle, handle;
 	struct acpi_buffer string = { ACPI_ALLOCATE_BUFFER, NULL };
 
-	/*
-	 * Per PCI firmware specification, we should run the ACPI _OSC
-	 * method to get control of hotplug hardware before using it. If
-	 * an _OSC is missing, we look for an OSHP to do the same thing.
-	 * To handle different BIOS behavior, we look for _OSC on a root
-	 * bridge preferentially (according to PCI fw spec). Later for
-	 * OSHP within the scope of the hotplug controller and its parents,
-	 * up to the host bridge under which this controller exists.
-	 */
-	if (shpchp_is_native(pdev))
-		return 0;
-
-	/* If _OSC exists, we should not evaluate OSHP */
-
 	/*
 	 * If there's no ACPI host bridge (i.e., ACPI support is compiled
 	 * into the kernel but the hardware platform doesn't support ACPI),
@@ -97,9 +83,25 @@ int acpi_get_hp_hw_control_from_firmware(struct pci_dev *pdev)
 	if (!root)
 		return 0;
 
-	if (root->osc_support_set)
+	/*
-		goto no_control;
+	 * If _OSC exists, it determines whether we're allowed to manage
+	 * the SHPC.  We executed it while enumerating the host bridge.
+	 */
+	if (root->osc_support_set) {
+		if (host->native_shpc_hotplug)
+			return 0;
+		return -ENODEV;
+	}
 
+	/*
+	 * In the absence of _OSC, we're always allowed to manage the SHPC.
+	 * However, if an OSHP method is present, we must execute it so the
+	 * firmware can transfer control to the OS, e.g., direct interrupts
+	 * to the OS instead of to the firmware.
+	 *
+	 * N.B. The PCI Firmware Spec (r3.2, sec 4.8) does not endorse
+	 * searching up the ACPI hierarchy, so the loops below are suspect.
+	 */
 	handle = ACPI_HANDLE(&pdev->dev);
 	if (!handle) {
 		/*
@@ -128,7 +130,7 @@ int acpi_get_hp_hw_control_from_firmware(struct pci_dev *pdev)
 		if (ACPI_FAILURE(status))
 			break;
 	}
-no_control:
+
 	pci_info(pdev, "Cannot get control of SHPC hotplug\n");
 	kfree(string.pointer);
 	return -ENODEV;

drivers/pci/hotplug/acpiphp_core.c

@@ -254,20 +254,6 @@ static int get_adapter_status(struct hotplug_slot *hotplug_slot, u8 *value)
 	return 0;
 }
 
-/**
- * release_slot - free up the memory used by a slot
- * @hotplug_slot: slot to free
- */
-static void release_slot(struct hotplug_slot *hotplug_slot)
-{
-	struct slot *slot = hotplug_slot->private;
-
-	pr_debug("%s - physical_slot = %s\n", __func__, slot_name(slot));
-
-	kfree(slot->hotplug_slot);
-	kfree(slot);
-}
-
 /* callback routine to initialize 'struct slot' for each slot */
 int acpiphp_register_hotplug_slot(struct acpiphp_slot *acpiphp_slot,
 				  unsigned int sun)
@@ -287,7 +273,6 @@ int acpiphp_register_hotplug_slot(struct acpiphp_slot *acpiphp_slot,
 	slot->hotplug_slot->info = &slot->info;
 
 	slot->hotplug_slot->private = slot;
-	slot->hotplug_slot->release = &release_slot;
 	slot->hotplug_slot->ops = &acpi_hotplug_slot_ops;
 
 	slot->acpi_slot = acpiphp_slot;
@@ -324,13 +309,12 @@ error:
 void acpiphp_unregister_hotplug_slot(struct acpiphp_slot *acpiphp_slot)
 {
 	struct slot *slot = acpiphp_slot->slot;
-	int retval = 0;
 
 	pr_info("Slot [%s] unregistered\n", slot_name(slot));
 
-	retval = pci_hp_deregister(slot->hotplug_slot);
+	pci_hp_deregister(slot->hotplug_slot);
-	if (retval)
+	kfree(slot->hotplug_slot);
-		pr_err("pci_hp_deregister failed with error %d\n", retval);
+	kfree(slot);
 }
 
 

drivers/pci/hotplug/cpci_hotplug_core.c

@@ -195,10 +195,8 @@ get_latch_status(struct hotplug_slot *hotplug_slot, u8 *value)
 	return 0;
 }
 
-static void release_slot(struct hotplug_slot *hotplug_slot)
+static void release_slot(struct slot *slot)
 {
-	struct slot *slot = hotplug_slot->private;
-
 	kfree(slot->hotplug_slot->info);
 	kfree(slot->hotplug_slot);
 	pci_dev_put(slot->dev);
@@ -253,7 +251,6 @@ cpci_hp_register_bus(struct pci_bus *bus, u8 first, u8 last)
 		snprintf(name, SLOT_NAME_SIZE, "%02x:%02x", bus->number, i);
 
 		hotplug_slot->private = slot;
-		hotplug_slot->release = &release_slot;
 		hotplug_slot->ops = &cpci_hotplug_slot_ops;
 
 		/*
@@ -308,12 +305,8 @@ cpci_hp_unregister_bus(struct pci_bus *bus)
 			slots--;
 
 			dbg("deregistering slot %s", slot_name(slot));
-			status = pci_hp_deregister(slot->hotplug_slot);
+			pci_hp_deregister(slot->hotplug_slot);
-			if (status) {
+			release_slot(slot);
-				err("pci_hp_deregister failed with error %d",
-				    status);
-				break;
-			}
 		}
 	}
 	up_write(&list_rwsem);
@@ -623,6 +616,7 @@ cleanup_slots(void)
 	list_for_each_entry_safe(slot, tmp, &slot_list, slot_list) {
 		list_del(&slot->slot_list);
 		pci_hp_deregister(slot->hotplug_slot);
+		release_slot(slot);
 	}
 cleanup_null:
 	up_write(&list_rwsem);

drivers/pci/hotplug/cpqphp_core.c

@@ -266,17 +266,6 @@ static void __iomem *get_SMBIOS_entry(void __iomem *smbios_start,
 	return previous;
 }
 
-static void release_slot(struct hotplug_slot *hotplug_slot)
-{
-	struct slot *slot = hotplug_slot->private;
-
-	dbg("%s - physical_slot = %s\n", __func__, slot_name(slot));
-
-	kfree(slot->hotplug_slot->info);
-	kfree(slot->hotplug_slot);
-	kfree(slot);
-}
-
 static int ctrl_slot_cleanup(struct controller *ctrl)
 {
 	struct slot *old_slot, *next_slot;
@@ -285,9 +274,11 @@ static int ctrl_slot_cleanup(struct controller *ctrl)
 	ctrl->slot = NULL;
 
 	while (old_slot) {
-		/* memory will be freed by the release_slot callback */
 		next_slot = old_slot->next;
 		pci_hp_deregister(old_slot->hotplug_slot);
+		kfree(old_slot->hotplug_slot->info);
+		kfree(old_slot->hotplug_slot);
+		kfree(old_slot);
 		old_slot = next_slot;
 	}
 
@@ -678,7 +669,6 @@ static int ctrl_slot_setup(struct controller *ctrl,
 			((read_slot_enable(ctrl) << 2) >> ctrl_slot) & 0x04;
 
 		/* register this slot with the hotplug pci core */
-		hotplug_slot->release = &release_slot;
 		hotplug_slot->private = slot;
 		snprintf(name, SLOT_NAME_SIZE, "%u", slot->number);
 		hotplug_slot->ops = &cpqphp_hotplug_slot_ops;

drivers/pci/hotplug/ibmphp_core.c

@@ -673,7 +673,20 @@ static void free_slots(void)
 
 	list_for_each_entry_safe(slot_cur, next, &ibmphp_slot_head,
 				 ibm_slot_list) {
-		pci_hp_deregister(slot_cur->hotplug_slot);
+		pci_hp_del(slot_cur->hotplug_slot);
+		slot_cur->ctrl = NULL;
+		slot_cur->bus_on = NULL;
+
+		/*
+		 * We don't want to actually remove the resources,
+		 * since ibmphp_free_resources() will do just that.
+		 */
+		ibmphp_unconfigure_card(&slot_cur, -1);
+
+		pci_hp_destroy(slot_cur->hotplug_slot);
+		kfree(slot_cur->hotplug_slot->info);
+		kfree(slot_cur->hotplug_slot);
+		kfree(slot_cur);
 	}
 	debug("%s -- exit\n", __func__);
 }

drivers/pci/hotplug/ibmphp_ebda.c

@@ -699,25 +699,6 @@ static int fillslotinfo(struct hotplug_slot *hotplug_slot)
 	return rc;
 }
 
-static void release_slot(struct hotplug_slot *hotplug_slot)
-{
-	struct slot *slot;
-
-	if (!hotplug_slot || !hotplug_slot->private)
-		return;
-
-	slot = hotplug_slot->private;
-	kfree(slot->hotplug_slot->info);
-	kfree(slot->hotplug_slot);
-	slot->ctrl = NULL;
-	slot->bus_on = NULL;
-
-	/* we don't want to actually remove the resources, since free_resources will do just that */
-	ibmphp_unconfigure_card(&slot, -1);
-
-	kfree(slot);
-}
-
 static struct pci_driver ibmphp_driver;
 
 /*
@@ -941,7 +922,6 @@ static int __init ebda_rsrc_controller(void)
 			tmp_slot->hotplug_slot = hp_slot_ptr;
 
 			hp_slot_ptr->private = tmp_slot;
-			hp_slot_ptr->release = release_slot;
 
 			rc = fillslotinfo(hp_slot_ptr);
 			if (rc)

drivers/pci/hotplug/pci_hotplug_core.c

@@ -396,8 +396,9 @@ static struct hotplug_slot *get_slot_from_name(const char *name)
  * @owner: caller module owner
  * @mod_name: caller module name
  *
- * Registers a hotplug slot with the pci hotplug subsystem, which will allow
+ * Prepares a hotplug slot for in-kernel use and immediately publishes it to
- * userspace interaction to the slot.
+ * user space in one go.  Drivers may alternatively carry out the two steps
+ * separately by invoking pci_hp_initialize() and pci_hp_add().
  *
  * Returns 0 if successful, anything else for an error.
  */
@@ -406,45 +407,91 @@ int __pci_hp_register(struct hotplug_slot *slot, struct pci_bus *bus,
 		      struct module *owner, const char *mod_name)
 {
 	int result;
+
+	result = __pci_hp_initialize(slot, bus, devnr, name, owner, mod_name);
+	if (result)
+		return result;
+
+	result = pci_hp_add(slot);
+	if (result)
+		pci_hp_destroy(slot);
+
+	return result;
+}
+EXPORT_SYMBOL_GPL(__pci_hp_register);
+
+/**
+ * __pci_hp_initialize - prepare hotplug slot for in-kernel use
+ * @slot: pointer to the &struct hotplug_slot to initialize
+ * @bus: bus this slot is on
+ * @devnr: slot number
+ * @name: name registered with kobject core
+ * @owner: caller module owner
+ * @mod_name: caller module name
+ *
+ * Allocate and fill in a PCI slot for use by a hotplug driver.  Once this has
+ * been called, the driver may invoke hotplug_slot_name() to get the slot's
+ * unique name.  The driver must be prepared to handle a ->reset_slot callback
+ * from this point on.
+ *
+ * Returns 0 on success or a negative int on error.
+ */
+int __pci_hp_initialize(struct hotplug_slot *slot, struct pci_bus *bus,
+			int devnr, const char *name, struct module *owner,
+			const char *mod_name)
+{
 	struct pci_slot *pci_slot;
 
 	if (slot == NULL)
 		return -ENODEV;
 	if ((slot->info == NULL) || (slot->ops == NULL))
 		return -EINVAL;
-	if (slot->release == NULL) {
-		dbg("Why are you trying to register a hotplug slot without a proper release function?\n");
-		return -EINVAL;
-	}
 
 	slot->ops->owner = owner;
 	slot->ops->mod_name = mod_name;
 
-	mutex_lock(&pci_hp_mutex);
 	/*
 	 * No problems if we call this interface from both ACPI_PCI_SLOT
 	 * driver and call it here again. If we've already created the
 	 * pci_slot, the interface will simply bump the refcount.
 	 */
 	pci_slot = pci_create_slot(bus, devnr, name, slot);
-	if (IS_ERR(pci_slot)) {
+	if (IS_ERR(pci_slot))
-		result = PTR_ERR(pci_slot);
+		return PTR_ERR(pci_slot);
-		goto out;
-	}
 
 	slot->pci_slot = pci_slot;
 	pci_slot->hotplug = slot;
+	return 0;
+}
+EXPORT_SYMBOL_GPL(__pci_hp_initialize);
 
-	list_add(&slot->slot_list, &pci_hotplug_slot_list);
+/**
+ * pci_hp_add - publish hotplug slot to user space
+ * @slot: pointer to the &struct hotplug_slot to publish
+ *
+ * Make a hotplug slot's sysfs interface available and inform user space of its
+ * addition by sending a uevent.  The hotplug driver must be prepared to handle
+ * all &struct hotplug_slot_ops callbacks from this point on.
+ *
+ * Returns 0 on success or a negative int on error.
+ */
+int pci_hp_add(struct hotplug_slot *slot)
+{
+	struct pci_slot *pci_slot = slot->pci_slot;
+	int result;
 
 	result = fs_add_slot(pci_slot);
-	kobject_uevent(&pci_slot->kobj, KOBJ_ADD);
+	if (result)
-	dbg("Added slot %s to the list\n", name);
-out:
-	mutex_unlock(&pci_hp_mutex);
 		return result;
+
+	kobject_uevent(&pci_slot->kobj, KOBJ_ADD);
+	mutex_lock(&pci_hp_mutex);
+	list_add(&slot->slot_list, &pci_hotplug_slot_list);
+	mutex_unlock(&pci_hp_mutex);
+	dbg("Added slot %s to the list\n", hotplug_slot_name(slot));
+	return 0;
 }
-EXPORT_SYMBOL_GPL(__pci_hp_register);
+EXPORT_SYMBOL_GPL(pci_hp_add);
 
 /**
  * pci_hp_deregister - deregister a hotplug_slot with the PCI hotplug subsystem
@@ -455,35 +502,62 @@ EXPORT_SYMBOL_GPL(__pci_hp_register);
  *
  * Returns 0 if successful, anything else for an error.
  */
-int pci_hp_deregister(struct hotplug_slot *slot)
+void pci_hp_deregister(struct hotplug_slot *slot)
+{
+	pci_hp_del(slot);
+	pci_hp_destroy(slot);
+}
+EXPORT_SYMBOL_GPL(pci_hp_deregister);
+
+/**
+ * pci_hp_del - unpublish hotplug slot from user space
+ * @slot: pointer to the &struct hotplug_slot to unpublish
+ *
+ * Remove a hotplug slot's sysfs interface.
+ *
+ * Returns 0 on success or a negative int on error.
+ */
+void pci_hp_del(struct hotplug_slot *slot)
 {
 	struct hotplug_slot *temp;
-	struct pci_slot *pci_slot;
 
-	if (!slot)
+	if (WARN_ON(!slot))
-		return -ENODEV;
+		return;
 
 	mutex_lock(&pci_hp_mutex);
 	temp = get_slot_from_name(hotplug_slot_name(slot));
-	if (temp != slot) {
+	if (WARN_ON(temp != slot)) {
 		mutex_unlock(&pci_hp_mutex);
-		return -ENODEV;
+		return;
 	}
 
 	list_del(&slot->slot_list);
-
+	mutex_unlock(&pci_hp_mutex);
-	pci_slot = slot->pci_slot;
-	fs_remove_slot(pci_slot);
 	dbg("Removed slot %s from the list\n", hotplug_slot_name(slot));
+	fs_remove_slot(slot->pci_slot);
+}
+EXPORT_SYMBOL_GPL(pci_hp_del);
 
-	slot->release(slot);
+/**
+ * pci_hp_destroy - remove hotplug slot from in-kernel use
+ * @slot: pointer to the &struct hotplug_slot to destroy
+ *
+ * Destroy a PCI slot used by a hotplug driver.  Once this has been called,
+ * the driver may no longer invoke hotplug_slot_name() to get the slot's
+ * unique name.  The driver no longer needs to handle a ->reset_slot callback
+ * from this point on.
+ *
+ * Returns 0 on success or a negative int on error.
+ */
+void pci_hp_destroy(struct hotplug_slot *slot)
+{
+	struct pci_slot *pci_slot = slot->pci_slot;
+
+	slot->pci_slot = NULL;
 	pci_slot->hotplug = NULL;
 	pci_destroy_slot(pci_slot);
-	mutex_unlock(&pci_hp_mutex);
-
-	return 0;
 }
-EXPORT_SYMBOL_GPL(pci_hp_deregister);
+EXPORT_SYMBOL_GPL(pci_hp_destroy);
 
 /**
  * pci_hp_change_slot_info - changes the slot's information structure in the core

drivers/pci/hotplug/pciehp.h

@@ -21,6 +21,7 @@
 #include <linux/delay.h>
 #include <linux/sched/signal.h>		/* signal_pending() */
 #include <linux/mutex.h>
+#include <linux/rwsem.h>
 #include <linux/workqueue.h>
 
 #include "../pcie/portdrv.h"
@@ -57,49 +58,111 @@ do {									\
 	dev_warn(&ctrl->pcie->device, format, ## arg)
 
 #define SLOT_NAME_SIZE 10
+
+/**
+ * struct slot - PCIe hotplug slot
+ * @state: current state machine position
+ * @ctrl: pointer to the slot's controller structure
+ * @hotplug_slot: pointer to the structure registered with the PCI hotplug core
+ * @work: work item to turn the slot on or off after 5 seconds in response to
+ *	an Attention Button press
+ * @lock: protects reads and writes of @state;
+ *	protects scheduling, execution and cancellation of @work
+ */
 struct slot {
 	u8 state;
 	struct controller *ctrl;
 	struct hotplug_slot *hotplug_slot;
-	struct delayed_work work;	/* work for button event */
+	struct delayed_work work;
 	struct mutex lock;
-	struct mutex hotplug_lock;
-	struct workqueue_struct *wq;
-};
-
-struct event_info {
-	u32 event_type;
-	struct slot *p_slot;
-	struct work_struct work;
 };
 
+/**
+ * struct controller - PCIe hotplug controller
+ * @ctrl_lock: serializes writes to the Slot Control register
+ * @pcie: pointer to the controller's PCIe port service device
+ * @reset_lock: prevents access to the Data Link Layer Link Active bit in the
+ *	Link Status register and to the Presence Detect State bit in the Slot
+ *	Status register during a slot reset which may cause them to flap
+ * @slot: pointer to the controller's slot structure
+ * @queue: wait queue to wake up on reception of a Command Completed event,
+ *	used for synchronous writes to the Slot Control register
+ * @slot_cap: cached copy of the Slot Capabilities register
+ * @slot_ctrl: cached copy of the Slot Control register
+ * @poll_thread: thread to poll for slot events if no IRQ is available,
+ *	enabled with pciehp_poll_mode module parameter
+ * @cmd_started: jiffies when the Slot Control register was last written;
+ *	the next write is allowed 1 second later, absent a Command Completed
+ *	interrupt (PCIe r4.0, sec 6.7.3.2)
+ * @cmd_busy: flag set on Slot Control register write, cleared by IRQ handler
+ *	on reception of a Command Completed event
+ * @link_active_reporting: cached copy of Data Link Layer Link Active Reporting
+ *	Capable bit in Link Capabilities register; if this bit is zero, the
+ *	Data Link Layer Link Active bit in the Link Status register will never
+ *	be set and the driver is thus confined to wait 1 second before assuming
+ *	the link to a hotplugged device is up and accessing it
+ * @notification_enabled: whether the IRQ was requested successfully
+ * @power_fault_detected: whether a power fault was detected by the hardware
+ *	that has not yet been cleared by the user
+ * @pending_events: used by the IRQ handler to save events retrieved from the
+ *	Slot Status register for later consumption by the IRQ thread
+ * @request_result: result of last user request submitted to the IRQ thread
+ * @requester: wait queue to wake up on completion of user request,
+ *	used for synchronous slot enable/disable request via sysfs
+ */
 struct controller {
-	struct mutex ctrl_lock;		/* controller lock */
+	struct mutex ctrl_lock;
-	struct pcie_device *pcie;	/* PCI Express port service */
+	struct pcie_device *pcie;
+	struct rw_semaphore reset_lock;
 	struct slot *slot;
-	wait_queue_head_t queue;	/* sleep & wake process */
+	wait_queue_head_t queue;
 	u32 slot_cap;
 	u16 slot_ctrl;
-	struct timer_list poll_timer;
+	struct task_struct *poll_thread;
 	unsigned long cmd_started;	/* jiffies */
 	unsigned int cmd_busy:1;
 	unsigned int link_active_reporting:1;
 	unsigned int notification_enabled:1;
 	unsigned int power_fault_detected;
+	atomic_t pending_events;
+	int request_result;
+	wait_queue_head_t requester;
 };
 
-#define INT_PRESENCE_ON			1
+/**
-#define INT_PRESENCE_OFF		2
+ * DOC: Slot state
-#define INT_POWER_FAULT			3
+ *
-#define INT_BUTTON_PRESS		4
+ * @OFF_STATE: slot is powered off, no subordinate devices are enumerated
-#define INT_LINK_UP			5
+ * @BLINKINGON_STATE: slot will be powered on after the 5 second delay,
-#define INT_LINK_DOWN			6
+ *	green led is blinking
-
+ * @BLINKINGOFF_STATE: slot will be powered off after the 5 second delay,
-#define STATIC_STATE			0
+ *	green led is blinking
+ * @POWERON_STATE: slot is currently powering on
+ * @POWEROFF_STATE: slot is currently powering off
+ * @ON_STATE: slot is powered on, subordinate devices have been enumerated
+ */
+#define OFF_STATE			0
 #define BLINKINGON_STATE		1
 #define BLINKINGOFF_STATE		2
 #define POWERON_STATE			3
 #define POWEROFF_STATE			4
+#define ON_STATE			5
+
+/**
+ * DOC: Flags to request an action from the IRQ thread
+ *
+ * These are stored together with events read from the Slot Status register,
+ * hence must be greater than its 16-bit width.
+ *
+ * %DISABLE_SLOT: Disable the slot in response to a user request via sysfs or
+ *	an Attention Button press after the 5 second delay
+ * %RERUN_ISR: Used by the IRQ handler to inform the IRQ thread that the
+ *	hotplug port was inaccessible when the interrupt occurred, requiring
+ *	that the IRQ handler is rerun by the IRQ thread after it has made the
+ *	hotplug port accessible by runtime resuming its parents to D0
+ */
+#define DISABLE_SLOT		(1 << 16)
+#define RERUN_ISR		(1 << 17)
 
 #define ATTN_BUTTN(ctrl)	((ctrl)->slot_cap & PCI_EXP_SLTCAP_ABP)
 #define POWER_CTRL(ctrl)	((ctrl)->slot_cap & PCI_EXP_SLTCAP_PCP)
@@ -113,15 +176,17 @@ struct controller {
 
 int pciehp_sysfs_enable_slot(struct slot *slot);
 int pciehp_sysfs_disable_slot(struct slot *slot);
-void pciehp_queue_interrupt_event(struct slot *slot, u32 event_type);
+void pciehp_request(struct controller *ctrl, int action);
+void pciehp_handle_button_press(struct slot *slot);
+void pciehp_handle_disable_request(struct slot *slot);
+void pciehp_handle_presence_or_link_change(struct slot *slot, u32 events);
 int pciehp_configure_device(struct slot *p_slot);
-int pciehp_unconfigure_device(struct slot *p_slot);
+void pciehp_unconfigure_device(struct slot *p_slot);
 void pciehp_queue_pushbutton_work(struct work_struct *work);
 struct controller *pcie_init(struct pcie_device *dev);
 int pcie_init_notification(struct controller *ctrl);
-int pciehp_enable_slot(struct slot *p_slot);
+void pcie_shutdown_notification(struct controller *ctrl);
-int pciehp_disable_slot(struct slot *p_slot);
+void pcie_clear_hotplug_events(struct controller *ctrl);
-void pcie_reenable_notification(struct controller *ctrl);
 int pciehp_power_on_slot(struct slot *slot);
 void pciehp_power_off_slot(struct slot *slot);
 void pciehp_get_power_status(struct slot *slot, u8 *status);

drivers/pci/hotplug/pciehp_core.c

@@ -26,11 +26,12 @@
 #include <linux/interrupt.h>
 #include <linux/time.h>
 
+#include "../pci.h"
+
 /* Global variables */
 bool pciehp_debug;
 bool pciehp_poll_mode;
 int pciehp_poll_time;
-static bool pciehp_force;
 
 /*
  * not really modular, but the easiest way to keep compat with existing
@@ -39,11 +40,9 @@ static bool pciehp_force;
 module_param(pciehp_debug, bool, 0644);
 module_param(pciehp_poll_mode, bool, 0644);
 module_param(pciehp_poll_time, int, 0644);
-module_param(pciehp_force, bool, 0644);
 MODULE_PARM_DESC(pciehp_debug, "Debugging mode enabled or not");
 MODULE_PARM_DESC(pciehp_poll_mode, "Using polling mechanism for hot-plug events or not");
 MODULE_PARM_DESC(pciehp_poll_time, "Polling mechanism frequency, in seconds");
-MODULE_PARM_DESC(pciehp_force, "Force pciehp, even if OSHP is missing");
 
 #define PCIE_MODULE_NAME "pciehp"
 
@@ -56,17 +55,6 @@ static int get_latch_status(struct hotplug_slot *slot, u8 *value);
 static int get_adapter_status(struct hotplug_slot *slot, u8 *value);
 static int reset_slot(struct hotplug_slot *slot, int probe);
 
-/**
- * release_slot - free up the memory used by a slot
- * @hotplug_slot: slot to free
- */
-static void release_slot(struct hotplug_slot *hotplug_slot)
-{
-	kfree(hotplug_slot->ops);
-	kfree(hotplug_slot->info);
-	kfree(hotplug_slot);
-}
-
 static int init_slot(struct controller *ctrl)
 {
 	struct slot *slot = ctrl->slot;
@@ -107,15 +95,14 @@ static int init_slot(struct controller *ctrl)
 	/* register this slot with the hotplug pci core */
 	hotplug->info = info;
 	hotplug->private = slot;
-	hotplug->release = &release_slot;
 	hotplug->ops = ops;
 	slot->hotplug_slot = hotplug;
 	snprintf(name, SLOT_NAME_SIZE, "%u", PSN(ctrl));
 
-	retval = pci_hp_register(hotplug,
+	retval = pci_hp_initialize(hotplug,
 				   ctrl->pcie->port->subordinate, 0, name);
 	if (retval)
-		ctrl_err(ctrl, "pci_hp_register failed: error %d\n", retval);
+		ctrl_err(ctrl, "pci_hp_initialize failed: error %d\n", retval);
 out:
 	if (retval) {
 		kfree(ops);
@@ -127,7 +114,12 @@ out:
 
 static void cleanup_slot(struct controller *ctrl)
 {
-	pci_hp_deregister(ctrl->slot->hotplug_slot);
+	struct hotplug_slot *hotplug_slot = ctrl->slot->hotplug_slot;
+
+	pci_hp_destroy(hotplug_slot);
+	kfree(hotplug_slot->ops);
+	kfree(hotplug_slot->info);
+	kfree(hotplug_slot);
 }
 
 /*
@@ -136,8 +128,11 @@ static void cleanup_slot(struct controller *ctrl)
 static int set_attention_status(struct hotplug_slot *hotplug_slot, u8 status)
 {
 	struct slot *slot = hotplug_slot->private;
+	struct pci_dev *pdev = slot->ctrl->pcie->port;
 
+	pci_config_pm_runtime_get(pdev);
 	pciehp_set_attention_status(slot, status);
+	pci_config_pm_runtime_put(pdev);
 	return 0;
 }
 
@@ -160,8 +155,11 @@ static int disable_slot(struct hotplug_slot *hotplug_slot)
 static int get_power_status(struct hotplug_slot *hotplug_slot, u8 *value)
 {
 	struct slot *slot = hotplug_slot->private;
+	struct pci_dev *pdev = slot->ctrl->pcie->port;
 
+	pci_config_pm_runtime_get(pdev);
 	pciehp_get_power_status(slot, value);
+	pci_config_pm_runtime_put(pdev);
 	return 0;
 }
 
@@ -176,16 +174,22 @@ static int get_attention_status(struct hotplug_slot *hotplug_slot, u8 *value)
 static int get_latch_status(struct hotplug_slot *hotplug_slot, u8 *value)
 {
 	struct slot *slot = hotplug_slot->private;
+	struct pci_dev *pdev = slot->ctrl->pcie->port;
 
+	pci_config_pm_runtime_get(pdev);
 	pciehp_get_latch_status(slot, value);
+	pci_config_pm_runtime_put(pdev);
 	return 0;
 }
 
 static int get_adapter_status(struct hotplug_slot *hotplug_slot, u8 *value)
 {
 	struct slot *slot = hotplug_slot->private;
+	struct pci_dev *pdev = slot->ctrl->pcie->port;
 
+	pci_config_pm_runtime_get(pdev);
 	pciehp_get_adapter_status(slot, value);
+	pci_config_pm_runtime_put(pdev);
 	return 0;
 }
 
@@ -196,12 +200,40 @@ static int reset_slot(struct hotplug_slot *hotplug_slot, int probe)
 	return pciehp_reset_slot(slot, probe);
 }
 
+/**
+ * pciehp_check_presence() - synthesize event if presence has changed
+ *
+ * On probe and resume, an explicit presence check is necessary to bring up an
+ * occupied slot or bring down an unoccupied slot.  This can't be triggered by
+ * events in the Slot Status register, they may be stale and are therefore
+ * cleared.  Secondly, sending an interrupt for "events that occur while
+ * interrupt generation is disabled [when] interrupt generation is subsequently
+ * enabled" is optional per PCIe r4.0, sec 6.7.3.4.
+ */
+static void pciehp_check_presence(struct controller *ctrl)
+{
+	struct slot *slot = ctrl->slot;
+	u8 occupied;
+
+	down_read(&ctrl->reset_lock);
+	mutex_lock(&slot->lock);
+
+	pciehp_get_adapter_status(slot, &occupied);
+	if ((occupied && (slot->state == OFF_STATE ||
+			  slot->state == BLINKINGON_STATE)) ||
+	    (!occupied && (slot->state == ON_STATE ||
+			   slot->state == BLINKINGOFF_STATE)))
+		pciehp_request(ctrl, PCI_EXP_SLTSTA_PDC);
+
+	mutex_unlock(&slot->lock);
+	up_read(&ctrl->reset_lock);
+}
+
 static int pciehp_probe(struct pcie_device *dev)
 {
 	int rc;
 	struct controller *ctrl;
 	struct slot *slot;
-	u8 occupied, poweron;
 
 	/* If this is not a "hotplug" service, we have no business here. */
 	if (dev->service != PCIE_PORT_SERVICE_HP)
@@ -238,21 +270,20 @@ static int pciehp_probe(struct pcie_device *dev)
 		goto err_out_free_ctrl_slot;
 	}
 
-	/* Check if slot is occupied */
+	/* Publish to user space */
 	slot = ctrl->slot;
-	pciehp_get_adapter_status(slot, &occupied);
+	rc = pci_hp_add(slot->hotplug_slot);
-	pciehp_get_power_status(slot, &poweron);
+	if (rc) {
-	if (occupied && pciehp_force) {
+		ctrl_err(ctrl, "Publication to user space failed (%d)\n", rc);
-		mutex_lock(&slot->hotplug_lock);
+		goto err_out_shutdown_notification;
-		pciehp_enable_slot(slot);
-		mutex_unlock(&slot->hotplug_lock);
 	}
-	/* If empty slot's power status is on, turn power off */
+
-	if (!occupied && poweron && POWER_CTRL(ctrl))
+	pciehp_check_presence(ctrl);
-		pciehp_power_off_slot(slot);
 
 	return 0;
 
+err_out_shutdown_notification:
+	pcie_shutdown_notification(ctrl);
 err_out_free_ctrl_slot:
 	cleanup_slot(ctrl);
 err_out_release_ctlr:
@@ -264,6 +295,8 @@ static void pciehp_remove(struct pcie_device *dev)
 {
 	struct controller *ctrl = get_service_data(dev);
 
+	pci_hp_del(ctrl->slot->hotplug_slot);
+	pcie_shutdown_notification(ctrl);
 	cleanup_slot(ctrl);
 	pciehp_release_ctrl(ctrl);
 }
@@ -274,27 +307,28 @@ static int pciehp_suspend(struct pcie_device *dev)
 	return 0;
 }
 
+static int pciehp_resume_noirq(struct pcie_device *dev)
+{
+	struct controller *ctrl = get_service_data(dev);
+	struct slot *slot = ctrl->slot;
+
+	/* pci_restore_state() just wrote to the Slot Control register */
+	ctrl->cmd_started = jiffies;
+	ctrl->cmd_busy = true;
+
+	/* clear spurious events from rediscovery of inserted card */
+	if (slot->state == ON_STATE || slot->state == BLINKINGOFF_STATE)
+		pcie_clear_hotplug_events(ctrl);
+
+	return 0;
+}
+
 static int pciehp_resume(struct pcie_device *dev)
 {
-	struct controller *ctrl;
+	struct controller *ctrl = get_service_data(dev);
-	struct slot *slot;
-	u8 status;
 
-	ctrl = get_service_data(dev);
+	pciehp_check_presence(ctrl);
 
-	/* reinitialize the chipset's event detection logic */
-	pcie_reenable_notification(ctrl);
-
-	slot = ctrl->slot;
-
-	/* Check if slot is occupied */
-	pciehp_get_adapter_status(slot, &status);
-	mutex_lock(&slot->hotplug_lock);
-	if (status)
-		pciehp_enable_slot(slot);
-	else
-		pciehp_disable_slot(slot);
-	mutex_unlock(&slot->hotplug_lock);
 	return 0;
 }
 #endif /* PM */
@@ -309,6 +343,7 @@ static struct pcie_port_service_driver hpdriver_portdrv = {
 
 #ifdef	CONFIG_PM
 	.suspend	= pciehp_suspend,
+	.resume_noirq	= pciehp_resume_noirq,
 	.resume		= pciehp_resume,
 #endif	/* PM */
 };

drivers/pci/hotplug/pciehp_ctrl.c

@@ -17,28 +17,11 @@
 #include <linux/kernel.h>
 #include <linux/types.h>
 #include <linux/slab.h>
+#include <linux/pm_runtime.h>
 #include <linux/pci.h>
 #include "../pci.h"
 #include "pciehp.h"
 
-static void interrupt_event_handler(struct work_struct *work);
-
-void pciehp_queue_interrupt_event(struct slot *p_slot, u32 event_type)
-{
-	struct event_info *info;
-
-	info = kmalloc(sizeof(*info), GFP_ATOMIC);
-	if (!info) {
-		ctrl_err(p_slot->ctrl, "dropped event %d (ENOMEM)\n", event_type);
-		return;
-	}
-
-	INIT_WORK(&info->work, interrupt_event_handler);
-	info->event_type = event_type;
-	info->p_slot = p_slot;
-	queue_work(p_slot->wq, &info->work);
-}
-
 /* The following routines constitute the bulk of the
    hotplug controller logic
  */
@@ -119,14 +102,11 @@ err_exit:
  * remove_board - Turns off slot and LEDs
  * @p_slot: slot where board is being removed
  */
-static int remove_board(struct slot *p_slot)
+static void remove_board(struct slot *p_slot)
 {
-	int retval;
 	struct controller *ctrl = p_slot->ctrl;
 
-	retval = pciehp_unconfigure_device(p_slot);
+	pciehp_unconfigure_device(p_slot);
-	if (retval)
-		return retval;
 
 	if (POWER_CTRL(ctrl)) {
 		pciehp_power_off_slot(p_slot);
@@ -141,86 +121,30 @@ static int remove_board(struct slot *p_slot)
 
 	/* turn off Green LED */
 	pciehp_green_led_off(p_slot);
-	return 0;
 }
 
-struct power_work_info {
+static int pciehp_enable_slot(struct slot *slot);
-	struct slot *p_slot;
+static int pciehp_disable_slot(struct slot *slot);
-	struct work_struct work;
-	unsigned int req;
-#define DISABLE_REQ 0
-#define ENABLE_REQ  1
-};
 
-/**
+void pciehp_request(struct controller *ctrl, int action)
- * pciehp_power_thread - handle pushbutton events
- * @work: &struct work_struct describing work to be done
- *
- * Scheduled procedure to handle blocking stuff for the pushbuttons.
- * Handles all pending events and exits.
- */
-static void pciehp_power_thread(struct work_struct *work)
 {
-	struct power_work_info *info =
+	atomic_or(action, &ctrl->pending_events);
-		container_of(work, struct power_work_info, work);
+	if (!pciehp_poll_mode)
-	struct slot *p_slot = info->p_slot;
+		irq_wake_thread(ctrl->pcie->irq, ctrl);
-	int ret;
-
-	switch (info->req) {
-	case DISABLE_REQ:
-		mutex_lock(&p_slot->hotplug_lock);
-		pciehp_disable_slot(p_slot);
-		mutex_unlock(&p_slot->hotplug_lock);
-		mutex_lock(&p_slot->lock);
-		p_slot->state = STATIC_STATE;
-		mutex_unlock(&p_slot->lock);
-		break;
-	case ENABLE_REQ:
-		mutex_lock(&p_slot->hotplug_lock);
-		ret = pciehp_enable_slot(p_slot);
-		mutex_unlock(&p_slot->hotplug_lock);
-		if (ret)
-			pciehp_green_led_off(p_slot);
-		mutex_lock(&p_slot->lock);
-		p_slot->state = STATIC_STATE;
-		mutex_unlock(&p_slot->lock);
-		break;
-	default:
-		break;
-	}
-
-	kfree(info);
-}
-
-static void pciehp_queue_power_work(struct slot *p_slot, int req)
-{
-	struct power_work_info *info;
-
-	p_slot->state = (req == ENABLE_REQ) ? POWERON_STATE : POWEROFF_STATE;
-
-	info = kmalloc(sizeof(*info), GFP_KERNEL);
-	if (!info) {
-		ctrl_err(p_slot->ctrl, "no memory to queue %s request\n",
-			 (req == ENABLE_REQ) ? "poweron" : "poweroff");
-		return;
-	}
-	info->p_slot = p_slot;
-	INIT_WORK(&info->work, pciehp_power_thread);
-	info->req = req;
-	queue_work(p_slot->wq, &info->work);
 }
 
 void pciehp_queue_pushbutton_work(struct work_struct *work)
 {
 	struct slot *p_slot = container_of(work, struct slot, work.work);
+	struct controller *ctrl = p_slot->ctrl;
 
 	mutex_lock(&p_slot->lock);
 	switch (p_slot->state) {
 	case BLINKINGOFF_STATE:
-		pciehp_queue_power_work(p_slot, DISABLE_REQ);
+		pciehp_request(ctrl, DISABLE_SLOT);
 		break;
 	case BLINKINGON_STATE:
-		pciehp_queue_power_work(p_slot, ENABLE_REQ);
+		pciehp_request(ctrl, PCI_EXP_SLTSTA_PDC);
 		break;
 	default:
 		break;
@@ -228,18 +152,15 @@ void pciehp_queue_pushbutton_work(struct work_struct *work)
 	mutex_unlock(&p_slot->lock);
 }
 
-/*
+void pciehp_handle_button_press(struct slot *p_slot)
- * Note: This function must be called with slot->lock held
- */
-static void handle_button_press_event(struct slot *p_slot)
 {
 	struct controller *ctrl = p_slot->ctrl;
-	u8 getstatus;
 
+	mutex_lock(&p_slot->lock);
 	switch (p_slot->state) {
-	case STATIC_STATE:
+	case OFF_STATE:
-		pciehp_get_power_status(p_slot, &getstatus);
+	case ON_STATE:
-		if (getstatus) {
+		if (p_slot->state == ON_STATE) {
 			p_slot->state = BLINKINGOFF_STATE;
 			ctrl_info(ctrl, "Slot(%s): Powering off due to button press\n",
 				  slot_name(p_slot));
@@ -251,7 +172,7 @@ static void handle_button_press_event(struct slot *p_slot)
 		/* blink green LED and turn off amber */
 		pciehp_green_led_blink(p_slot);
 		pciehp_set_attention_status(p_slot, 0);
-		queue_delayed_work(p_slot->wq, &p_slot->work, 5*HZ);
+		schedule_delayed_work(&p_slot->work, 5 * HZ);
 		break;
 	case BLINKINGOFF_STATE:
 	case BLINKINGON_STATE:
@@ -262,118 +183,104 @@ static void handle_button_press_event(struct slot *p_slot)
 		 */
 		ctrl_info(ctrl, "Slot(%s): Button cancel\n", slot_name(p_slot));
 		cancel_delayed_work(&p_slot->work);
-		if (p_slot->state == BLINKINGOFF_STATE)
+		if (p_slot->state == BLINKINGOFF_STATE) {
+			p_slot->state = ON_STATE;
 			pciehp_green_led_on(p_slot);
-		else
+		} else {
+			p_slot->state = OFF_STATE;
 			pciehp_green_led_off(p_slot);
+		}
 		pciehp_set_attention_status(p_slot, 0);
 		ctrl_info(ctrl, "Slot(%s): Action canceled due to button press\n",
 			  slot_name(p_slot));
-		p_slot->state = STATIC_STATE;
-		break;
-	case POWEROFF_STATE:
-	case POWERON_STATE:
-		/*
-		 * Ignore if the slot is on power-on or power-off state;
-		 * this means that the previous attention button action
-		 * to hot-add or hot-remove is undergoing
-		 */
-		ctrl_info(ctrl, "Slot(%s): Button ignored\n",
-			  slot_name(p_slot));
 		break;
 	default:
 		ctrl_err(ctrl, "Slot(%s): Ignoring invalid state %#x\n",
 			 slot_name(p_slot), p_slot->state);
 		break;
 	}
-}
-
-/*
- * Note: This function must be called with slot->lock held
- */
-static void handle_link_event(struct slot *p_slot, u32 event)
-{
-	struct controller *ctrl = p_slot->ctrl;
-
-	switch (p_slot->state) {
-	case BLINKINGON_STATE:
-	case BLINKINGOFF_STATE:
-		cancel_delayed_work(&p_slot->work);
-		/* Fall through */
-	case STATIC_STATE:
-		pciehp_queue_power_work(p_slot, event == INT_LINK_UP ?
-					ENABLE_REQ : DISABLE_REQ);
-		break;
-	case POWERON_STATE:
-		if (event == INT_LINK_UP) {
-			ctrl_info(ctrl, "Slot(%s): Link Up event ignored; already powering on\n",
-				  slot_name(p_slot));
-		} else {
-			ctrl_info(ctrl, "Slot(%s): Link Down event queued; currently getting powered on\n",
-				  slot_name(p_slot));
-			pciehp_queue_power_work(p_slot, DISABLE_REQ);
-		}
-		break;
-	case POWEROFF_STATE:
-		if (event == INT_LINK_UP) {
-			ctrl_info(ctrl, "Slot(%s): Link Up event queued; currently getting powered off\n",
-				  slot_name(p_slot));
-			pciehp_queue_power_work(p_slot, ENABLE_REQ);
-		} else {
-			ctrl_info(ctrl, "Slot(%s): Link Down event ignored; already powering off\n",
-				  slot_name(p_slot));
-		}
-		break;
-	default:
-		ctrl_err(ctrl, "Slot(%s): Ignoring invalid state %#x\n",
-			 slot_name(p_slot), p_slot->state);
-		break;
-	}
-}
-
-static void interrupt_event_handler(struct work_struct *work)
-{
-	struct event_info *info = container_of(work, struct event_info, work);
-	struct slot *p_slot = info->p_slot;
-	struct controller *ctrl = p_slot->ctrl;
-
-	mutex_lock(&p_slot->lock);
-	switch (info->event_type) {
-	case INT_BUTTON_PRESS:
-		handle_button_press_event(p_slot);
-		break;
-	case INT_POWER_FAULT:
-		if (!POWER_CTRL(ctrl))
-			break;
-		pciehp_set_attention_status(p_slot, 1);
-		pciehp_green_led_off(p_slot);
-		break;
-	case INT_PRESENCE_ON:
-		pciehp_queue_power_work(p_slot, ENABLE_REQ);
-		break;
-	case INT_PRESENCE_OFF:
-		/*
-		 * Regardless of surprise capability, we need to
-		 * definitely remove a card that has been pulled out!
-		 */
-		pciehp_queue_power_work(p_slot, DISABLE_REQ);
-		break;
-	case INT_LINK_UP:
-	case INT_LINK_DOWN:
-		handle_link_event(p_slot, info->event_type);
-		break;
-	default:
-		break;
-	}
 	mutex_unlock(&p_slot->lock);
-
-	kfree(info);
 }
 
+void pciehp_handle_disable_request(struct slot *slot)
+{
+	struct controller *ctrl = slot->ctrl;
+
+	mutex_lock(&slot->lock);
+	switch (slot->state) {
+	case BLINKINGON_STATE:
+	case BLINKINGOFF_STATE:
+		cancel_delayed_work(&slot->work);
+		break;
+	}
+	slot->state = POWEROFF_STATE;
+	mutex_unlock(&slot->lock);
+
+	ctrl->request_result = pciehp_disable_slot(slot);
+}
+
+void pciehp_handle_presence_or_link_change(struct slot *slot, u32 events)
+{
+	struct controller *ctrl = slot->ctrl;
+	bool link_active;
+	u8 present;
+
 	/*
- * Note: This function must be called with slot->hotplug_lock held
+	 * If the slot is on and presence or link has changed, turn it off.
+	 * Even if it's occupied again, we cannot assume the card is the same.
 	 */
-int pciehp_enable_slot(struct slot *p_slot)
+	mutex_lock(&slot->lock);
+	switch (slot->state) {
+	case BLINKINGOFF_STATE:
+		cancel_delayed_work(&slot->work);
+		/* fall through */
+	case ON_STATE:
+		slot->state = POWEROFF_STATE;
+		mutex_unlock(&slot->lock);
+		if (events & PCI_EXP_SLTSTA_DLLSC)
+			ctrl_info(ctrl, "Slot(%s): Link Down\n",
+				  slot_name(slot));
+		if (events & PCI_EXP_SLTSTA_PDC)
+			ctrl_info(ctrl, "Slot(%s): Card not present\n",
+				  slot_name(slot));
+		pciehp_disable_slot(slot);
+		break;
+	default:
+		mutex_unlock(&slot->lock);
+		break;
+	}
+
+	/* Turn the slot on if it's occupied or link is up */
+	mutex_lock(&slot->lock);
+	pciehp_get_adapter_status(slot, &present);
+	link_active = pciehp_check_link_active(ctrl);
+	if (!present && !link_active) {
+		mutex_unlock(&slot->lock);
+		return;
+	}
+
+	switch (slot->state) {
+	case BLINKINGON_STATE:
+		cancel_delayed_work(&slot->work);
+		/* fall through */
+	case OFF_STATE:
+		slot->state = POWERON_STATE;
+		mutex_unlock(&slot->lock);
+		if (present)
+			ctrl_info(ctrl, "Slot(%s): Card present\n",
+				  slot_name(slot));
+		if (link_active)
+			ctrl_info(ctrl, "Slot(%s): Link Up\n",
+				  slot_name(slot));
+		ctrl->request_result = pciehp_enable_slot(slot);
+		break;
+	default:
+		mutex_unlock(&slot->lock);
+		break;
+	}
+}
+
+static int __pciehp_enable_slot(struct slot *p_slot)
 {
 	u8 getstatus = 0;
 	struct controller *ctrl = p_slot->ctrl;
@@ -404,17 +311,29 @@ int pciehp_enable_slot(struct slot *p_slot)
 	return board_added(p_slot);
 }
 
-/*
+static int pciehp_enable_slot(struct slot *slot)
- * Note: This function must be called with slot->hotplug_lock held
+{
- */
+	struct controller *ctrl = slot->ctrl;
-int pciehp_disable_slot(struct slot *p_slot)
+	int ret;
+
+	pm_runtime_get_sync(&ctrl->pcie->port->dev);
+	ret = __pciehp_enable_slot(slot);
+	if (ret && ATTN_BUTTN(ctrl))
+		pciehp_green_led_off(slot); /* may be blinking */
+	pm_runtime_put(&ctrl->pcie->port->dev);
+
+	mutex_lock(&slot->lock);
+	slot->state = ret ? OFF_STATE : ON_STATE;
+	mutex_unlock(&slot->lock);
+
+	return ret;
+}
+
+static int __pciehp_disable_slot(struct slot *p_slot)
 {
 	u8 getstatus = 0;
 	struct controller *ctrl = p_slot->ctrl;
 
-	if (!p_slot->ctrl)
-		return 1;
-
 	if (POWER_CTRL(p_slot->ctrl)) {
 		pciehp_get_power_status(p_slot, &getstatus);
 		if (!getstatus) {
@@ -424,32 +343,50 @@ int pciehp_disable_slot(struct slot *p_slot)
 		}
 	}
 
-	return remove_board(p_slot);
+	remove_board(p_slot);
+	return 0;
+}
+
+static int pciehp_disable_slot(struct slot *slot)
+{
+	struct controller *ctrl = slot->ctrl;
+	int ret;
+
+	pm_runtime_get_sync(&ctrl->pcie->port->dev);
+	ret = __pciehp_disable_slot(slot);
+	pm_runtime_put(&ctrl->pcie->port->dev);
+
+	mutex_lock(&slot->lock);
+	slot->state = OFF_STATE;
+	mutex_unlock(&slot->lock);
+
+	return ret;
 }
 
 int pciehp_sysfs_enable_slot(struct slot *p_slot)
 {
-	int retval = -ENODEV;
 	struct controller *ctrl = p_slot->ctrl;
 
 	mutex_lock(&p_slot->lock);
 	switch (p_slot->state) {
 	case BLINKINGON_STATE:
-		cancel_delayed_work(&p_slot->work);
+	case OFF_STATE:
-	case STATIC_STATE:
-		p_slot->state = POWERON_STATE;
 		mutex_unlock(&p_slot->lock);
-		mutex_lock(&p_slot->hotplug_lock);
+		/*
-		retval = pciehp_enable_slot(p_slot);
+		 * The IRQ thread becomes a no-op if the user pulls out the
-		mutex_unlock(&p_slot->hotplug_lock);
+		 * card before the thread wakes up, so initialize to -ENODEV.
-		mutex_lock(&p_slot->lock);
+		 */
-		p_slot->state = STATIC_STATE;
+		ctrl->request_result = -ENODEV;
-		break;
+		pciehp_request(ctrl, PCI_EXP_SLTSTA_PDC);
+		wait_event(ctrl->requester,
+			   !atomic_read(&ctrl->pending_events));
+		return ctrl->request_result;
 	case POWERON_STATE:
 		ctrl_info(ctrl, "Slot(%s): Already in powering on state\n",
 			  slot_name(p_slot));
 		break;
 	case BLINKINGOFF_STATE:
+	case ON_STATE:
 	case POWEROFF_STATE:
 		ctrl_info(ctrl, "Slot(%s): Already enabled\n",
 			  slot_name(p_slot));
@@ -461,32 +398,28 @@ int pciehp_sysfs_enable_slot(struct slot *p_slot)
 	}
 	mutex_unlock(&p_slot->lock);
 
-	return retval;
+	return -ENODEV;
 }
 
 int pciehp_sysfs_disable_slot(struct slot *p_slot)
 {
-	int retval = -ENODEV;
 	struct controller *ctrl = p_slot->ctrl;
 
 	mutex_lock(&p_slot->lock);
 	switch (p_slot->state) {
 	case BLINKINGOFF_STATE:
-		cancel_delayed_work(&p_slot->work);
+	case ON_STATE:
-	case STATIC_STATE:
-		p_slot->state = POWEROFF_STATE;
 		mutex_unlock(&p_slot->lock);
-		mutex_lock(&p_slot->hotplug_lock);
+		pciehp_request(ctrl, DISABLE_SLOT);
-		retval = pciehp_disable_slot(p_slot);
+		wait_event(ctrl->requester,
-		mutex_unlock(&p_slot->hotplug_lock);
+			   !atomic_read(&ctrl->pending_events));
-		mutex_lock(&p_slot->lock);
+		return ctrl->request_result;
-		p_slot->state = STATIC_STATE;
-		break;
 	case POWEROFF_STATE:
 		ctrl_info(ctrl, "Slot(%s): Already in powering off state\n",
 			  slot_name(p_slot));
 		break;
 	case BLINKINGON_STATE:
+	case OFF_STATE:
 	case POWERON_STATE:
 		ctrl_info(ctrl, "Slot(%s): Already disabled\n",
 			  slot_name(p_slot));
@@ -498,5 +431,5 @@ int pciehp_sysfs_disable_slot(struct slot *p_slot)
 	}
 	mutex_unlock(&p_slot->lock);
 
-	return retval;
+	return -ENODEV;
 }

drivers/pci/hotplug/pciehp_hpc.c

@@ -17,8 +17,9 @@
 #include <linux/types.h>
 #include <linux/signal.h>
 #include <linux/jiffies.h>
-#include <linux/timer.h>
+#include <linux/kthread.h>
 #include <linux/pci.h>
+#include <linux/pm_runtime.h>
 #include <linux/interrupt.h>
 #include <linux/time.h>
 #include <linux/slab.h>
@@ -31,47 +32,24 @@ static inline struct pci_dev *ctrl_dev(struct controller *ctrl)
 	return ctrl->pcie->port;
 }
 
-static irqreturn_t pcie_isr(int irq, void *dev_id);
+static irqreturn_t pciehp_isr(int irq, void *dev_id);
-static void start_int_poll_timer(struct controller *ctrl, int sec);
+static irqreturn_t pciehp_ist(int irq, void *dev_id);
-
+static int pciehp_poll(void *data);
-/* This is the interrupt polling timeout function. */
-static void int_poll_timeout(struct timer_list *t)
-{
-	struct controller *ctrl = from_timer(ctrl, t, poll_timer);
-
-	/* Poll for interrupt events.  regs == NULL => polling */
-	pcie_isr(0, ctrl);
-
-	if (!pciehp_poll_time)
-		pciehp_poll_time = 2; /* default polling interval is 2 sec */
-
-	start_int_poll_timer(ctrl, pciehp_poll_time);
-}
-
-/* This function starts the interrupt polling timer. */
-static void start_int_poll_timer(struct controller *ctrl, int sec)
-{
-	/* Clamp to sane value */
-	if ((sec <= 0) || (sec > 60))
-		sec = 2;
-
-	ctrl->poll_timer.expires = jiffies + sec * HZ;
-	add_timer(&ctrl->poll_timer);
-}
 
 static inline int pciehp_request_irq(struct controller *ctrl)
 {
 	int retval, irq = ctrl->pcie->irq;
 
-	/* Install interrupt polling timer. Start with 10 sec delay */
 	if (pciehp_poll_mode) {
-		timer_setup(&ctrl->poll_timer, int_poll_timeout, 0);
+		ctrl->poll_thread = kthread_run(&pciehp_poll, ctrl,
-		start_int_poll_timer(ctrl, 10);
+						"pciehp_poll-%s",
-		return 0;
+						slot_name(ctrl->slot));
+		return PTR_ERR_OR_ZERO(ctrl->poll_thread);
 	}
 
 	/* Installs the interrupt handler */
-	retval = request_irq(irq, pcie_isr, IRQF_SHARED, MY_NAME, ctrl);
+	retval = request_threaded_irq(irq, pciehp_isr, pciehp_ist,
+				      IRQF_SHARED, MY_NAME, ctrl);
 	if (retval)
 		ctrl_err(ctrl, "Cannot get irq %d for the hotplug controller\n",
 			 irq);
@@ -81,7 +59,7 @@ static inline int pciehp_request_irq(struct controller *ctrl)
 static inline void pciehp_free_irq(struct controller *ctrl)
 {
 	if (pciehp_poll_mode)
-		del_timer_sync(&ctrl->poll_timer);
+		kthread_stop(ctrl->poll_thread);
 	else
 		free_irq(ctrl->pcie->irq, ctrl);
 }
@@ -293,6 +271,11 @@ int pciehp_check_link_status(struct controller *ctrl)
 	found = pci_bus_check_dev(ctrl->pcie->port->subordinate,
 					PCI_DEVFN(0, 0));
 
+	/* ignore link or presence changes up to this point */
+	if (found)
+		atomic_and(~(PCI_EXP_SLTSTA_DLLSC | PCI_EXP_SLTSTA_PDC),
+			   &ctrl->pending_events);
+
 	pcie_capability_read_word(pdev, PCI_EXP_LNKSTA, &lnk_status);
 	ctrl_dbg(ctrl, "%s: lnk_status = %x\n", __func__, lnk_status);
 	if ((lnk_status & PCI_EXP_LNKSTA_LT) ||
@@ -339,7 +322,9 @@ int pciehp_get_raw_indicator_status(struct hotplug_slot *hotplug_slot,
 	struct pci_dev *pdev = ctrl_dev(slot->ctrl);
 	u16 slot_ctrl;
 
+	pci_config_pm_runtime_get(pdev);
 	pcie_capability_read_word(pdev, PCI_EXP_SLTCTL, &slot_ctrl);
+	pci_config_pm_runtime_put(pdev);
 	*status = (slot_ctrl & (PCI_EXP_SLTCTL_AIC | PCI_EXP_SLTCTL_PIC)) >> 6;
 	return 0;
 }
@@ -350,7 +335,9 @@ void pciehp_get_attention_status(struct slot *slot, u8 *status)
 	struct pci_dev *pdev = ctrl_dev(ctrl);
 	u16 slot_ctrl;
 
+	pci_config_pm_runtime_get(pdev);
 	pcie_capability_read_word(pdev, PCI_EXP_SLTCTL, &slot_ctrl);
+	pci_config_pm_runtime_put(pdev);
 	ctrl_dbg(ctrl, "%s: SLOTCTRL %x, value read %x\n", __func__,
 		 pci_pcie_cap(ctrl->pcie->port) + PCI_EXP_SLTCTL, slot_ctrl);
 
@@ -425,9 +412,12 @@ int pciehp_set_raw_indicator_status(struct hotplug_slot *hotplug_slot,
 {
 	struct slot *slot = hotplug_slot->private;
 	struct controller *ctrl = slot->ctrl;
+	struct pci_dev *pdev = ctrl_dev(ctrl);
 
+	pci_config_pm_runtime_get(pdev);
 	pcie_write_cmd_nowait(ctrl, status << 6,
 			      PCI_EXP_SLTCTL_AIC | PCI_EXP_SLTCTL_PIC);
+	pci_config_pm_runtime_put(pdev);
 	return 0;
 }
 
@@ -539,20 +529,35 @@ static irqreturn_t pciehp_isr(int irq, void *dev_id)
 {
 	struct controller *ctrl = (struct controller *)dev_id;
 	struct pci_dev *pdev = ctrl_dev(ctrl);
-	struct pci_bus *subordinate = pdev->subordinate;
+	struct device *parent = pdev->dev.parent;
-	struct pci_dev *dev;
-	struct slot *slot = ctrl->slot;
 	u16 status, events;
-	u8 present;
-	bool link;
 
-	/* Interrupts cannot originate from a controller that's asleep */
+	/*
+	 * Interrupts only occur in D3hot or shallower (PCIe r4.0, sec 6.7.3.4).
+	 */
 	if (pdev->current_state == PCI_D3cold)
 		return IRQ_NONE;
 
+	/*
+	 * Keep the port accessible by holding a runtime PM ref on its parent.
+	 * Defer resume of the parent to the IRQ thread if it's suspended.
+	 * Mask the interrupt until then.
+	 */
+	if (parent) {
+		pm_runtime_get_noresume(parent);
+		if (!pm_runtime_active(parent)) {
+			pm_runtime_put(parent);
+			disable_irq_nosync(irq);
+			atomic_or(RERUN_ISR, &ctrl->pending_events);
+			return IRQ_WAKE_THREAD;
+		}
+	}
+
 	pcie_capability_read_word(pdev, PCI_EXP_SLTSTA, &status);
 	if (status == (u16) ~0) {
 		ctrl_info(ctrl, "%s: no response from device\n", __func__);
+		if (parent)
+			pm_runtime_put(parent);
 		return IRQ_NONE;
 	}
 
@@ -571,86 +576,119 @@ static irqreturn_t pciehp_isr(int irq, void *dev_id)
 	if (ctrl->power_fault_detected)
 		events &= ~PCI_EXP_SLTSTA_PFD;
 
-	if (!events)
+	if (!events) {
+		if (parent)
+			pm_runtime_put(parent);
 		return IRQ_NONE;
-
+	}
-	/* Capture link status before clearing interrupts */
-	if (events & PCI_EXP_SLTSTA_DLLSC)
-		link = pciehp_check_link_active(ctrl);
 
 	pcie_capability_write_word(pdev, PCI_EXP_SLTSTA, events);
 	ctrl_dbg(ctrl, "pending interrupts %#06x from Slot Status\n", events);
+	if (parent)
+		pm_runtime_put(parent);
 
-	/* Check Command Complete Interrupt Pending */
+	/*
+	 * Command Completed notifications are not deferred to the
+	 * IRQ thread because it may be waiting for their arrival.
+	 */
 	if (events & PCI_EXP_SLTSTA_CC) {
 		ctrl->cmd_busy = 0;
 		smp_mb();
 		wake_up(&ctrl->queue);
+
+		if (events == PCI_EXP_SLTSTA_CC)
+			return IRQ_HANDLED;
+
+		events &= ~PCI_EXP_SLTSTA_CC;
 	}
 
-	if (subordinate) {
+	if (pdev->ignore_hotplug) {
-		list_for_each_entry(dev, &subordinate->devices, bus_list) {
+		ctrl_dbg(ctrl, "ignoring hotplug event %#06x\n", events);
-			if (dev->ignore_hotplug) {
-				ctrl_dbg(ctrl, "ignoring hotplug event %#06x (%s requested no hotplug)\n",
-					 events, pci_name(dev));
 		return IRQ_HANDLED;
 	}
+
+	/* Save pending events for consumption by IRQ thread. */
+	atomic_or(events, &ctrl->pending_events);
+	return IRQ_WAKE_THREAD;
 }
+
+static irqreturn_t pciehp_ist(int irq, void *dev_id)
+{
+	struct controller *ctrl = (struct controller *)dev_id;
+	struct pci_dev *pdev = ctrl_dev(ctrl);
+	struct slot *slot = ctrl->slot;
+	irqreturn_t ret;
+	u32 events;
+
+	pci_config_pm_runtime_get(pdev);
+
+	/* rerun pciehp_isr() if the port was inaccessible on interrupt */
+	if (atomic_fetch_and(~RERUN_ISR, &ctrl->pending_events) & RERUN_ISR) {
+		ret = pciehp_isr(irq, dev_id);
+		enable_irq(irq);
+		if (ret != IRQ_WAKE_THREAD) {
+			pci_config_pm_runtime_put(pdev);
+			return ret;
+		}
+	}
+
+	synchronize_hardirq(irq);
+	events = atomic_xchg(&ctrl->pending_events, 0);
+	if (!events) {
+		pci_config_pm_runtime_put(pdev);
+		return IRQ_NONE;
 	}
 
 	/* Check Attention Button Pressed */
 	if (events & PCI_EXP_SLTSTA_ABP) {
 		ctrl_info(ctrl, "Slot(%s): Attention button pressed\n",
 			  slot_name(slot));
-		pciehp_queue_interrupt_event(slot, INT_BUTTON_PRESS);
+		pciehp_handle_button_press(slot);
 	}
 
 	/*
-	 * Check Link Status Changed at higher precedence than Presence
+	 * Disable requests have higher priority than Presence Detect Changed
-	 * Detect Changed.  The PDS value may be set to "card present" from
+	 * or Data Link Layer State Changed events.
-	 * out-of-band detection, which may be in conflict with a Link Down
-	 * and cause the wrong event to queue.
 	 */
-	if (events & PCI_EXP_SLTSTA_DLLSC) {
+	down_read(&ctrl->reset_lock);
-		ctrl_info(ctrl, "Slot(%s): Link %s\n", slot_name(slot),
+	if (events & DISABLE_SLOT)
-			  link ? "Up" : "Down");
+		pciehp_handle_disable_request(slot);
-		pciehp_queue_interrupt_event(slot, link ? INT_LINK_UP :
+	else if (events & (PCI_EXP_SLTSTA_PDC | PCI_EXP_SLTSTA_DLLSC))
-					     INT_LINK_DOWN);
+		pciehp_handle_presence_or_link_change(slot, events);
-	} else if (events & PCI_EXP_SLTSTA_PDC) {
+	up_read(&ctrl->reset_lock);
-		present = !!(status & PCI_EXP_SLTSTA_PDS);
-		ctrl_info(ctrl, "Slot(%s): Card %spresent\n", slot_name(slot),
-			  present ? "" : "not ");
-		pciehp_queue_interrupt_event(slot, present ? INT_PRESENCE_ON :
-					     INT_PRESENCE_OFF);
-	}
 
 	/* Check Power Fault Detected */
 	if ((events & PCI_EXP_SLTSTA_PFD) && !ctrl->power_fault_detected) {
 		ctrl->power_fault_detected = 1;
 		ctrl_err(ctrl, "Slot(%s): Power fault\n", slot_name(slot));
-		pciehp_queue_interrupt_event(slot, INT_POWER_FAULT);
+		pciehp_set_attention_status(slot, 1);
+		pciehp_green_led_off(slot);
 	}
 
+	pci_config_pm_runtime_put(pdev);
+	wake_up(&ctrl->requester);
 	return IRQ_HANDLED;
 }
 
-static irqreturn_t pcie_isr(int irq, void *dev_id)
+static int pciehp_poll(void *data)
 {
-	irqreturn_t rc, handled = IRQ_NONE;
+	struct controller *ctrl = data;
 
-	/*
+	schedule_timeout_idle(10 * HZ); /* start with 10 sec delay */
-	 * To guarantee that all interrupt events are serviced, we need to
-	 * re-inspect Slot Status register after clearing what is presumed
-	 * to be the last pending interrupt.
-	 */
-	do {
-		rc = pciehp_isr(irq, dev_id);
-		if (rc == IRQ_HANDLED)
-			handled = IRQ_HANDLED;
-	} while (rc == IRQ_HANDLED);
 
-	/* Return IRQ_HANDLED if we handled one or more events */
+	while (!kthread_should_stop()) {
-	return handled;
+		/* poll for interrupt events or user requests */
+		while (pciehp_isr(IRQ_NOTCONNECTED, ctrl) == IRQ_WAKE_THREAD ||
+		       atomic_read(&ctrl->pending_events))
+			pciehp_ist(IRQ_NOTCONNECTED, ctrl);
+
+		if (pciehp_poll_time <= 0 || pciehp_poll_time > 60)
+			pciehp_poll_time = 2; /* clamp to sane value */
+
+		schedule_timeout_idle(pciehp_poll_time * HZ);
+	}
+
+	return 0;
 }
 
 static void pcie_enable_notification(struct controller *ctrl)
@@ -691,17 +729,6 @@ static void pcie_enable_notification(struct controller *ctrl)
 		 pci_pcie_cap(ctrl->pcie->port) + PCI_EXP_SLTCTL, cmd);
 }
 
-void pcie_reenable_notification(struct controller *ctrl)
-{
-	/*
-	 * Clear both Presence and Data Link Layer Changed to make sure
-	 * those events still fire after we have re-enabled them.
-	 */
-	pcie_capability_write_word(ctrl->pcie->port, PCI_EXP_SLTSTA,
-				   PCI_EXP_SLTSTA_PDC | PCI_EXP_SLTSTA_DLLSC);
-	pcie_enable_notification(ctrl);
-}
-
 static void pcie_disable_notification(struct controller *ctrl)
 {
 	u16 mask;
@@ -715,6 +742,12 @@ static void pcie_disable_notification(struct controller *ctrl)
 		 pci_pcie_cap(ctrl->pcie->port) + PCI_EXP_SLTCTL, 0);
 }
 
+void pcie_clear_hotplug_events(struct controller *ctrl)
+{
+	pcie_capability_write_word(ctrl_dev(ctrl), PCI_EXP_SLTSTA,
+				   PCI_EXP_SLTSTA_PDC | PCI_EXP_SLTSTA_DLLSC);
+}
+
 /*
  * pciehp has a 1:1 bus:slot relationship so we ultimately want a secondary
  * bus reset of the bridge, but at the same time we want to ensure that it is
@@ -732,6 +765,8 @@ int pciehp_reset_slot(struct slot *slot, int probe)
 	if (probe)
 		return 0;
 
+	down_write(&ctrl->reset_lock);
+
 	if (!ATTN_BUTTN(ctrl)) {
 		ctrl_mask |= PCI_EXP_SLTCTL_PDCE;
 		stat_mask |= PCI_EXP_SLTSTA_PDC;
@@ -742,8 +777,6 @@ int pciehp_reset_slot(struct slot *slot, int probe)
 	pcie_write_cmd(ctrl, 0, ctrl_mask);
 	ctrl_dbg(ctrl, "%s: SLOTCTRL %x write cmd %x\n", __func__,
 		 pci_pcie_cap(ctrl->pcie->port) + PCI_EXP_SLTCTL, 0);
-	if (pciehp_poll_mode)
-		del_timer_sync(&ctrl->poll_timer);
 
 	pci_reset_bridge_secondary_bus(ctrl->pcie->port);
 
@@ -751,8 +784,8 @@ int pciehp_reset_slot(struct slot *slot, int probe)
 	pcie_write_cmd_nowait(ctrl, ctrl_mask, ctrl_mask);
 	ctrl_dbg(ctrl, "%s: SLOTCTRL %x write cmd %x\n", __func__,
 		 pci_pcie_cap(ctrl->pcie->port) + PCI_EXP_SLTCTL, ctrl_mask);
-	if (pciehp_poll_mode)
+
-		int_poll_timeout(&ctrl->poll_timer);
+	up_write(&ctrl->reset_lock);
 	return 0;
 }
 
@@ -765,7 +798,7 @@ int pcie_init_notification(struct controller *ctrl)
 	return 0;
 }
 
-static void pcie_shutdown_notification(struct controller *ctrl)
+void pcie_shutdown_notification(struct controller *ctrl)
 {
 	if (ctrl->notification_enabled) {
 		pcie_disable_notification(ctrl);
@@ -776,32 +809,29 @@ static void pcie_shutdown_notification(struct controller *ctrl)
 
 static int pcie_init_slot(struct controller *ctrl)
 {
+	struct pci_bus *subordinate = ctrl_dev(ctrl)->subordinate;
 	struct slot *slot;
 
 	slot = kzalloc(sizeof(*slot), GFP_KERNEL);
 	if (!slot)
 		return -ENOMEM;
 
-	slot->wq = alloc_ordered_workqueue("pciehp-%u", 0, PSN(ctrl));
+	down_read(&pci_bus_sem);
-	if (!slot->wq)
+	slot->state = list_empty(&subordinate->devices) ? OFF_STATE : ON_STATE;
-		goto abort;
+	up_read(&pci_bus_sem);
 
 	slot->ctrl = ctrl;
 	mutex_init(&slot->lock);
-	mutex_init(&slot->hotplug_lock);
 	INIT_DELAYED_WORK(&slot->work, pciehp_queue_pushbutton_work);
 	ctrl->slot = slot;
 	return 0;
-abort:
-	kfree(slot);
-	return -ENOMEM;
 }
 
 static void pcie_cleanup_slot(struct controller *ctrl)
 {
 	struct slot *slot = ctrl->slot;
-	cancel_delayed_work(&slot->work);
+
-	destroy_workqueue(slot->wq);
+	cancel_delayed_work_sync(&slot->work);
 	kfree(slot);
 }
 
@@ -826,6 +856,7 @@ struct controller *pcie_init(struct pcie_device *dev)
 {
 	struct controller *ctrl;
 	u32 slot_cap, link_cap;
+	u8 occupied, poweron;
 	struct pci_dev *pdev = dev->port;
 
 	ctrl = kzalloc(sizeof(*ctrl), GFP_KERNEL);
@@ -847,6 +878,8 @@ struct controller *pcie_init(struct pcie_device *dev)
 
 	ctrl->slot_cap = slot_cap;
 	mutex_init(&ctrl->ctrl_lock);
+	init_rwsem(&ctrl->reset_lock);
+	init_waitqueue_head(&ctrl->requester);
 	init_waitqueue_head(&ctrl->queue);
 	dbg_ctrl(ctrl);
 
@@ -855,16 +888,11 @@ struct controller *pcie_init(struct pcie_device *dev)
 	if (link_cap & PCI_EXP_LNKCAP_DLLLARC)
 		ctrl->link_active_reporting = 1;
 
-	/*
+	/* Clear all remaining event bits in Slot Status register. */
-	 * Clear all remaining event bits in Slot Status register except
-	 * Presence Detect Changed. We want to make sure possible
-	 * hotplug event is triggered when the interrupt is unmasked so
-	 * that we don't lose that event.
-	 */
 	pcie_capability_write_word(pdev, PCI_EXP_SLTSTA,
 		PCI_EXP_SLTSTA_ABP | PCI_EXP_SLTSTA_PFD |
 		PCI_EXP_SLTSTA_MRLSC | PCI_EXP_SLTSTA_CC |
-		PCI_EXP_SLTSTA_DLLSC);
+		PCI_EXP_SLTSTA_DLLSC | PCI_EXP_SLTSTA_PDC);
 
 	ctrl_info(ctrl, "Slot #%d AttnBtn%c PwrCtrl%c MRL%c AttnInd%c PwrInd%c HotPlug%c Surprise%c Interlock%c NoCompl%c LLActRep%c%s\n",
 		(slot_cap & PCI_EXP_SLTCAP_PSN) >> 19,
@@ -883,6 +911,19 @@ struct controller *pcie_init(struct pcie_device *dev)
 	if (pcie_init_slot(ctrl))
 		goto abort_ctrl;
 
+	/*
+	 * If empty slot's power status is on, turn power off.  The IRQ isn't
+	 * requested yet, so avoid triggering a notification with this command.
+	 */
+	if (POWER_CTRL(ctrl)) {
+		pciehp_get_adapter_status(ctrl->slot, &occupied);
+		pciehp_get_power_status(ctrl->slot, &poweron);
+		if (!occupied && poweron) {
+			pcie_disable_notification(ctrl);
+			pciehp_power_off_slot(ctrl->slot);
+		}
+	}
+
 	return ctrl;
 
 abort_ctrl:
@@ -893,7 +934,6 @@ abort:
 
 void pciehp_release_ctrl(struct controller *ctrl)
 {
-	pcie_shutdown_notification(ctrl);
 	pcie_cleanup_slot(ctrl);
 	kfree(ctrl);
 }

drivers/pci/hotplug/pciehp_pci.c

@@ -62,9 +62,8 @@ int pciehp_configure_device(struct slot *p_slot)
 	return ret;
 }
 
-int pciehp_unconfigure_device(struct slot *p_slot)
+void pciehp_unconfigure_device(struct slot *p_slot)
 {
-	int rc = 0;
 	u8 presence = 0;
 	struct pci_dev *dev, *temp;
 	struct pci_bus *parent = p_slot->ctrl->pcie->port->subordinate;
@@ -107,5 +106,4 @@ int pciehp_unconfigure_device(struct slot *p_slot)
 	}
 
 	pci_unlock_rescan_remove();
-	return rc;
 }

drivers/pci/hotplug/pcihp_skeleton.c

@@ -1,348 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0+
-/*
- * PCI Hot Plug Controller Skeleton Driver - 0.3
- *
- * Copyright (C) 2001,2003 Greg Kroah-Hartman (greg@kroah.com)
- * Copyright (C) 2001,2003 IBM Corp.
- *
- * All rights reserved.
- *
- * This driver is to be used as a skeleton driver to show how to interface
- * with the pci hotplug core easily.
- *
- * Send feedback to <greg@kroah.com>
- *
- */
-
-#include <linux/module.h>
-#include <linux/moduleparam.h>
-#include <linux/kernel.h>
-#include <linux/slab.h>
-#include <linux/pci.h>
-#include <linux/pci_hotplug.h>
-#include <linux/init.h>
-
-#define SLOT_NAME_SIZE	10
-struct slot {
-	u8 number;
-	struct hotplug_slot *hotplug_slot;
-	struct list_head slot_list;
-	char name[SLOT_NAME_SIZE];
-};
-
-static LIST_HEAD(slot_list);
-
-#define MY_NAME	"pcihp_skeleton"
-
-#define dbg(format, arg...)					\
-	do {							\
-		if (debug)					\
-			printk(KERN_DEBUG "%s: " format "\n",	\
-				MY_NAME, ## arg);		\
-	} while (0)
-#define err(format, arg...) printk(KERN_ERR "%s: " format "\n", MY_NAME, ## arg)
-#define info(format, arg...) printk(KERN_INFO "%s: " format "\n", MY_NAME, ## arg)
-#define warn(format, arg...) printk(KERN_WARNING "%s: " format "\n", MY_NAME, ## arg)
-
-/* local variables */
-static bool debug;
-static int num_slots;
-
-#define DRIVER_VERSION	"0.3"
-#define DRIVER_AUTHOR	"Greg Kroah-Hartman <greg@kroah.com>"
-#define DRIVER_DESC	"Hot Plug PCI Controller Skeleton Driver"
-
-MODULE_AUTHOR(DRIVER_AUTHOR);
-MODULE_DESCRIPTION(DRIVER_DESC);
-MODULE_LICENSE("GPL");
-module_param(debug, bool, 0644);
-MODULE_PARM_DESC(debug, "Debugging mode enabled or not");
-
-static int enable_slot(struct hotplug_slot *slot);
-static int disable_slot(struct hotplug_slot *slot);
-static int set_attention_status(struct hotplug_slot *slot, u8 value);
-static int hardware_test(struct hotplug_slot *slot, u32 value);
-static int get_power_status(struct hotplug_slot *slot, u8 *value);
-static int get_attention_status(struct hotplug_slot *slot, u8 *value);
-static int get_latch_status(struct hotplug_slot *slot, u8 *value);
-static int get_adapter_status(struct hotplug_slot *slot, u8 *value);
-
-static struct hotplug_slot_ops skel_hotplug_slot_ops = {
-	.enable_slot =		enable_slot,
-	.disable_slot =		disable_slot,
-	.set_attention_status =	set_attention_status,
-	.hardware_test =	hardware_test,
-	.get_power_status =	get_power_status,
-	.get_attention_status =	get_attention_status,
-	.get_latch_status =	get_latch_status,
-	.get_adapter_status =	get_adapter_status,
-};
-
-static int enable_slot(struct hotplug_slot *hotplug_slot)
-{
-	struct slot *slot = hotplug_slot->private;
-	int retval = 0;
-
-	dbg("%s - physical_slot = %s\n", __func__, hotplug_slot->name);
-
-	/*
-	 * Fill in code here to enable the specified slot
-	 */
-
-	return retval;
-}
-
-static int disable_slot(struct hotplug_slot *hotplug_slot)
-{
-	struct slot *slot = hotplug_slot->private;
-	int retval = 0;
-
-	dbg("%s - physical_slot = %s\n", __func__, hotplug_slot->name);
-
-	/*
-	 * Fill in code here to disable the specified slot
-	 */
-
-	return retval;
-}
-
-static int set_attention_status(struct hotplug_slot *hotplug_slot, u8 status)
-{
-	struct slot *slot = hotplug_slot->private;
-	int retval = 0;
-
-	dbg("%s - physical_slot = %s\n", __func__, hotplug_slot->name);
-
-	switch (status) {
-	case 0:
-		/*
-		 * Fill in code here to turn light off
-		 */
-		break;
-
-	case 1:
-	default:
-		/*
-		 * Fill in code here to turn light on
-		 */
-		break;
-	}
-
-	return retval;
-}
-
-static int hardware_test(struct hotplug_slot *hotplug_slot, u32 value)
-{
-	struct slot *slot = hotplug_slot->private;
-	int retval = 0;
-
-	dbg("%s - physical_slot = %s\n", __func__, hotplug_slot->name);
-
-	switch (value) {
-	case 0:
-		/* Specify a test here */
-		break;
-	case 1:
-		/* Specify another test here */
-		break;
-	}
-
-	return retval;
-}
-
-static int get_power_status(struct hotplug_slot *hotplug_slot, u8 *value)
-{
-	struct slot *slot = hotplug_slot->private;
-	int retval = 0;
-
-	dbg("%s - physical_slot = %s\n", __func__, hotplug_slot->name);
-
-	/*
-	 * Fill in logic to get the current power status of the specific
-	 * slot and store it in the *value location.
-	 */
-
-	return retval;
-}
-
-static int get_attention_status(struct hotplug_slot *hotplug_slot, u8 *value)
-{
-	struct slot *slot = hotplug_slot->private;
-	int retval = 0;
-
-	dbg("%s - physical_slot = %s\n", __func__, hotplug_slot->name);
-
-	/*
-	 * Fill in logic to get the current attention status of the specific
-	 * slot and store it in the *value location.
-	 */
-
-	return retval;
-}
-
-static int get_latch_status(struct hotplug_slot *hotplug_slot, u8 *value)
-{
-	struct slot *slot = hotplug_slot->private;
-	int retval = 0;
-
-	dbg("%s - physical_slot = %s\n", __func__, hotplug_slot->name);
-
-	/*
-	 * Fill in logic to get the current latch status of the specific
-	 * slot and store it in the *value location.
-	 */
-
-	return retval;
-}
-
-static int get_adapter_status(struct hotplug_slot *hotplug_slot, u8 *value)
-{
-	struct slot *slot = hotplug_slot->private;
-	int retval = 0;
-
-	dbg("%s - physical_slot = %s\n", __func__, hotplug_slot->name);
-
-	/*
-	 * Fill in logic to get the current adapter status of the specific
-	 * slot and store it in the *value location.
-	 */
-
-	return retval;
-}
-
-static void release_slot(struct hotplug_slot *hotplug_slot)
-{
-	struct slot *slot = hotplug_slot->private;
-
-	dbg("%s - physical_slot = %s\n", __func__, hotplug_slot->name);
-	kfree(slot->hotplug_slot->info);
-	kfree(slot->hotplug_slot);
-	kfree(slot);
-}
-
-static void make_slot_name(struct slot *slot)
-{
-	/*
-	 * Stupid way to make a filename out of the slot name.
-	 * replace this if your hardware provides a better way to name slots.
-	 */
-	snprintf(slot->hotplug_slot->name, SLOT_NAME_SIZE, "%d", slot->number);
-}
-
-/**
- * init_slots - initialize 'struct slot' structures for each slot
- *
- */
-static int __init init_slots(void)
-{
-	struct slot *slot;
-	struct hotplug_slot *hotplug_slot;
-	struct hotplug_slot_info *info;
-	int retval;
-	int i;
-
-	/*
-	 * Create a structure for each slot, and register that slot
-	 * with the pci_hotplug subsystem.
-	 */
-	for (i = 0; i < num_slots; ++i) {
-		slot = kzalloc(sizeof(*slot), GFP_KERNEL);
-		if (!slot) {
-			retval = -ENOMEM;
-			goto error;
-		}
-
-		hotplug_slot = kzalloc(sizeof(*hotplug_slot), GFP_KERNEL);
-		if (!hotplug_slot) {
-			retval = -ENOMEM;
-			goto error_slot;
-		}
-		slot->hotplug_slot = hotplug_slot;
-
-		info = kzalloc(sizeof(*info), GFP_KERNEL);
-		if (!info) {
-			retval = -ENOMEM;
-			goto error_hpslot;
-		}
-		hotplug_slot->info = info;
-
-		slot->number = i;
-
-		hotplug_slot->name = slot->name;
-		hotplug_slot->private = slot;
-		hotplug_slot->release = &release_slot;
-		make_slot_name(slot);
-		hotplug_slot->ops = &skel_hotplug_slot_ops;
-
-		/*
-		 * Initialize the slot info structure with some known
-		 * good values.
-		 */
-		get_power_status(hotplug_slot, &info->power_status);
-		get_attention_status(hotplug_slot, &info->attention_status);
-		get_latch_status(hotplug_slot, &info->latch_status);
-		get_adapter_status(hotplug_slot, &info->adapter_status);
-
-		dbg("registering slot %d\n", i);
-		retval = pci_hp_register(slot->hotplug_slot);
-		if (retval) {
-			err("pci_hp_register failed with error %d\n", retval);
-			goto error_info;
-		}
-
-		/* add slot to our internal list */
-		list_add(&slot->slot_list, &slot_list);
-	}
-
-	return 0;
-error_info:
-	kfree(info);
-error_hpslot:
-	kfree(hotplug_slot);
-error_slot:
-	kfree(slot);
-error:
-	return retval;
-}
-
-static void __exit cleanup_slots(void)
-{
-	struct slot *slot, *next;
-
-	/*
-	 * Unregister all of our slots with the pci_hotplug subsystem.
-	 * Memory will be freed in release_slot() callback after slot's
-	 * lifespan is finished.
-	 */
-	list_for_each_entry_safe(slot, next, &slot_list, slot_list) {
-		list_del(&slot->slot_list);
-		pci_hp_deregister(slot->hotplug_slot);
-	}
-}
-
-static int __init pcihp_skel_init(void)
-{
-	int retval;
-
-	info(DRIVER_DESC " version: " DRIVER_VERSION "\n");
-	/*
-	 * Do specific initialization stuff for your driver here
-	 * like initializing your controller hardware (if any) and
-	 * determining the number of slots you have in the system
-	 * right now.
-	 */
-	num_slots = 5;
-
-	return init_slots();
-}
-
-static void __exit pcihp_skel_exit(void)
-{
-	/*
-	 * Clean everything up.
-	 */
-	cleanup_slots();
-}
-
-module_init(pcihp_skel_init);
-module_exit(pcihp_skel_exit);

drivers/pci/hotplug/pnv_php.c

@@ -538,9 +538,8 @@ static struct hotplug_slot_ops php_slot_ops = {
 	.disable_slot		= pnv_php_disable_slot,
 };
 
-static void pnv_php_release(struct hotplug_slot *slot)
+static void pnv_php_release(struct pnv_php_slot *php_slot)
 {
-	struct pnv_php_slot *php_slot = slot->private;
 	unsigned long flags;
 
 	/* Remove from global or child list */
@@ -596,7 +595,6 @@ static struct pnv_php_slot *pnv_php_alloc_slot(struct device_node *dn)
 	php_slot->power_state_check     = false;
 	php_slot->slot.ops              = &php_slot_ops;
 	php_slot->slot.info             = &php_slot->slot_info;
-	php_slot->slot.release          = pnv_php_release;
 	php_slot->slot.private          = php_slot;
 
 	INIT_LIST_HEAD(&php_slot->children);
@@ -924,6 +922,7 @@ static void pnv_php_unregister_one(struct device_node *dn)
 
 	php_slot->state = PNV_PHP_STATE_OFFLINE;
 	pci_hp_deregister(&php_slot->slot);
+	pnv_php_release(php_slot);
 	pnv_php_put_slot(php_slot);
 }
 

drivers/pci/hotplug/rpaphp_core.c

@@ -404,13 +404,13 @@ static void __exit cleanup_slots(void)
 	/*
 	 * Unregister all of our slots with the pci_hotplug subsystem,
 	 * and free up all memory that we had allocated.
-	 * memory will be freed in release_slot callback.
 	 */
 
 	list_for_each_entry_safe(slot, next, &rpaphp_slot_head,
 				 rpaphp_slot_list) {
 		list_del(&slot->rpaphp_slot_list);
 		pci_hp_deregister(slot->hotplug_slot);
+		dealloc_slot_struct(slot);
 	}
 	return;
 }

drivers/pci/hotplug/rpaphp_slot.c

@@ -19,12 +19,6 @@
 #include "rpaphp.h"
 
 /* free up the memory used by a slot */
-static void rpaphp_release_slot(struct hotplug_slot *hotplug_slot)
-{
-	struct slot *slot = (struct slot *) hotplug_slot->private;
-	dealloc_slot_struct(slot);
-}
-
 void dealloc_slot_struct(struct slot *slot)
 {
 	kfree(slot->hotplug_slot->info);
@@ -56,7 +50,6 @@ struct slot *alloc_slot_struct(struct device_node *dn,
 	slot->power_domain = power_domain;
 	slot->hotplug_slot->private = slot;
 	slot->hotplug_slot->ops = &rpaphp_hotplug_slot_ops;
-	slot->hotplug_slot->release = &rpaphp_release_slot;
 
 	return (slot);
 
@@ -90,10 +83,8 @@ int rpaphp_deregister_slot(struct slot *slot)
 		__func__, slot->name);
 
 	list_del(&slot->rpaphp_slot_list);
-
+	pci_hp_deregister(php_slot);
-	retval = pci_hp_deregister(php_slot);
+	dealloc_slot_struct(slot);
-	if (retval)
-		err("Problem unregistering a slot %s\n", slot->name);
 
 	dbg("%s - Exit: rc[%d]\n", __func__, retval);
 	return retval;

drivers/pci/hotplug/s390_pci_hpc.c

@@ -130,15 +130,6 @@ static int get_adapter_status(struct hotplug_slot *hotplug_slot, u8 *value)
 	return 0;
 }
 
-static void release_slot(struct hotplug_slot *hotplug_slot)
-{
-	struct slot *slot = hotplug_slot->private;
-
-	kfree(slot->hotplug_slot->info);
-	kfree(slot->hotplug_slot);
-	kfree(slot);
-}
-
 static struct hotplug_slot_ops s390_hotplug_slot_ops = {
 	.enable_slot =		enable_slot,
 	.disable_slot =		disable_slot,
@@ -175,7 +166,6 @@ int zpci_init_slot(struct zpci_dev *zdev)
 	hotplug_slot->info = info;
 
 	hotplug_slot->ops = &s390_hotplug_slot_ops;
-	hotplug_slot->release = &release_slot;
 
 	get_power_status(hotplug_slot, &info->power_status);
 	get_adapter_status(hotplug_slot, &info->adapter_status);
@@ -209,5 +199,8 @@ void zpci_exit_slot(struct zpci_dev *zdev)
 			continue;
 		list_del(&slot->slot_list);
 		pci_hp_deregister(slot->hotplug_slot);
+		kfree(slot->hotplug_slot->info);
+		kfree(slot->hotplug_slot);
+		kfree(slot);
 	}
 }

drivers/pci/hotplug/sgi_hotplug.c

@@ -628,7 +628,6 @@ static int sn_hotplug_slot_register(struct pci_bus *pci_bus)
 			goto alloc_err;
 		}
 		bss_hotplug_slot->ops = &sn_hotplug_slot_ops;
-		bss_hotplug_slot->release = &sn_release_slot;
 
 		rc = pci_hp_register(bss_hotplug_slot, pci_bus, device, name);
 		if (rc)
@@ -656,8 +655,10 @@ alloc_err:
 		sn_release_slot(bss_hotplug_slot);
 
 	/* destroy anything else on the list */
-	while ((bss_hotplug_slot = sn_hp_destroy()))
+	while ((bss_hotplug_slot = sn_hp_destroy())) {
 		pci_hp_deregister(bss_hotplug_slot);
+		sn_release_slot(bss_hotplug_slot);
+	}
 
 	return rc;
 }
@@ -703,8 +704,10 @@ static void __exit sn_pci_hotplug_exit(void)
 {
 	struct hotplug_slot *bss_hotplug_slot;
 
-	while ((bss_hotplug_slot = sn_hp_destroy()))
+	while ((bss_hotplug_slot = sn_hp_destroy())) {
 		pci_hp_deregister(bss_hotplug_slot);
+		sn_release_slot(bss_hotplug_slot);
+	}
 
 	if (!list_empty(&sn_hp_list))
 		printk(KERN_ERR "%s: internal list is not empty\n", __FILE__);

drivers/pci/hotplug/shpchp_core.c

@@ -61,22 +61,6 @@ static struct hotplug_slot_ops shpchp_hotplug_slot_ops = {
 	.get_adapter_status =	get_adapter_status,
 };
 
-/**
- * release_slot - free up the memory used by a slot
- * @hotplug_slot: slot to free
- */
-static void release_slot(struct hotplug_slot *hotplug_slot)
-{
-	struct slot *slot = hotplug_slot->private;
-
-	ctrl_dbg(slot->ctrl, "%s: physical_slot = %s\n",
-		 __func__, slot_name(slot));
-
-	kfree(slot->hotplug_slot->info);
-	kfree(slot->hotplug_slot);
-	kfree(slot);
-}
-
 static int init_slots(struct controller *ctrl)
 {
 	struct slot *slot;
@@ -125,7 +109,6 @@ static int init_slots(struct controller *ctrl)
 
 		/* register this slot with the hotplug pci core */
 		hotplug_slot->private = slot;
-		hotplug_slot->release = &release_slot;
 		snprintf(name, SLOT_NAME_SIZE, "%d", slot->number);
 		hotplug_slot->ops = &shpchp_hotplug_slot_ops;
 
@@ -171,6 +154,9 @@ void cleanup_slots(struct controller *ctrl)
 		cancel_delayed_work(&slot->work);
 		destroy_workqueue(slot->wq);
 		pci_hp_deregister(slot->hotplug_slot);
+		kfree(slot->hotplug_slot->info);
+		kfree(slot->hotplug_slot);
+		kfree(slot);
 	}
 }
 
@@ -270,11 +256,30 @@ static int get_adapter_status(struct hotplug_slot *hotplug_slot, u8 *value)
 	return 0;
 }
 
+static bool shpc_capable(struct pci_dev *bridge)
+{
+	/*
+	 * It is assumed that AMD GOLAM chips support SHPC but they do not
+	 * have SHPC capability.
+	 */
+	if (bridge->vendor == PCI_VENDOR_ID_AMD &&
+	    bridge->device == PCI_DEVICE_ID_AMD_GOLAM_7450)
+		return true;
+
+	if (pci_find_capability(bridge, PCI_CAP_ID_SHPC))
+		return true;
+
+	return false;
+}
+
 static int shpc_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 {
 	int rc;
 	struct controller *ctrl;
 
+	if (!shpc_capable(pdev))
+		return -ENODEV;
+
 	if (acpi_get_hp_hw_control_from_firmware(pdev))
 		return -ENODEV;
 
@@ -303,6 +308,7 @@ static int shpc_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (rc)
 		goto err_cleanup_slots;
 
+	pdev->shpc_managed = 1;
 	return 0;
 
 err_cleanup_slots:
@@ -319,6 +325,7 @@ static void shpc_remove(struct pci_dev *dev)
 {
 	struct controller *ctrl = pci_get_drvdata(dev);
 
+	dev->shpc_managed = 0;
 	shpchp_remove_ctrl_files(ctrl);
 	ctrl->hpc_ops->release_ctlr(ctrl);
 	kfree(ctrl);

drivers/pci/pci-acpi.c

@@ -403,24 +403,7 @@ bool pciehp_is_native(struct pci_dev *bridge)
  */
 bool shpchp_is_native(struct pci_dev *bridge)
 {
-	const struct pci_host_bridge *host;
+	return bridge->shpc_managed;
-
-	if (!IS_ENABLED(CONFIG_HOTPLUG_PCI_SHPC))
-		return false;
-
-	/*
-	 * It is assumed that AMD GOLAM chips support SHPC but they do not
-	 * have SHPC capability.
-	 */
-	if (bridge->vendor == PCI_VENDOR_ID_AMD &&
-	    bridge->device == PCI_DEVICE_ID_AMD_GOLAM_7450)
-		return true;
-
-	if (!pci_find_capability(bridge, PCI_CAP_ID_SHPC))
-		return false;
-
-	host = pci_find_host_bridge(bridge->bus);
-	return host->native_shpc_hotplug;
 }
 
 /**

drivers/pci/pci-sysfs.c

@@ -1448,7 +1448,9 @@ static ssize_t reset_store(struct device *dev, struct device_attribute *attr,
 	if (val != 1)
 		return -EINVAL;
 
+	pm_runtime_get_sync(dev);
 	result = pci_reset_function(pdev);
+	pm_runtime_put(dev);
 	if (result < 0)
 		return result;
 

drivers/pci/pci.c

@@ -2289,7 +2289,7 @@ void pci_config_pm_runtime_put(struct pci_dev *pdev)
  * @bridge: Bridge to check
  *
  * This function checks if it is possible to move the bridge to D3.
- * Currently we only allow D3 for recent enough PCIe ports.
+ * Currently we only allow D3 for recent enough PCIe ports and Thunderbolt.
  */
 bool pci_bridge_d3_possible(struct pci_dev *bridge)
 {
@@ -2304,18 +2304,27 @@ bool pci_bridge_d3_possible(struct pci_dev *bridge)
 			return false;
 
 		/*
-		 * Hotplug interrupts cannot be delivered if the link is down,
+		 * Hotplug ports handled by firmware in System Management Mode
-		 * so parents of a hotplug port must stay awake. In addition,
-		 * hotplug ports handled by firmware in System Management Mode
 		 * may not be put into D3 by the OS (Thunderbolt on non-Macs).
-		 * For simplicity, disallow in general for now.
 		 */
-		if (bridge->is_hotplug_bridge)
+		if (bridge->is_hotplug_bridge && !pciehp_is_native(bridge))
 			return false;
 
 		if (pci_bridge_d3_force)
 			return true;
 
+		/* Even the oldest 2010 Thunderbolt controller supports D3. */
+		if (bridge->is_thunderbolt)
+			return true;
+
+		/*
+		 * Hotplug ports handled natively by the OS were not validated
+		 * by vendors for runtime D3 at least until 2018 because there
+		 * was no OS support.
+		 */
+		if (bridge->is_hotplug_bridge)
+			return false;
+
 		/*
 		 * It should be safe to put PCIe ports from 2015 or newer
 		 * to D3.

drivers/pci/pcie/portdrv.h

@@ -50,6 +50,7 @@ struct pcie_port_service_driver {
 	int (*probe) (struct pcie_device *dev);
 	void (*remove) (struct pcie_device *dev);
 	int (*suspend) (struct pcie_device *dev);
+	int (*resume_noirq) (struct pcie_device *dev);
 	int (*resume) (struct pcie_device *dev);
 
 	/* Device driver may resume normal operations */
@@ -82,6 +83,7 @@ extern struct bus_type pcie_port_bus_type;
 int pcie_port_device_register(struct pci_dev *dev);
 #ifdef CONFIG_PM
 int pcie_port_device_suspend(struct device *dev);
+int pcie_port_device_resume_noirq(struct device *dev);
 int pcie_port_device_resume(struct device *dev);
 #endif
 void pcie_port_device_remove(struct pci_dev *dev);

drivers/pci/pcie/portdrv_core.c

@@ -353,14 +353,19 @@ error_disable:
 }
 
 #ifdef CONFIG_PM
-static int suspend_iter(struct device *dev, void *data)
+typedef int (*pcie_pm_callback_t)(struct pcie_device *);
+
+static int pm_iter(struct device *dev, void *data)
 {
 	struct pcie_port_service_driver *service_driver;
+	size_t offset = *(size_t *)data;
+	pcie_pm_callback_t cb;
 
 	if ((dev->bus == &pcie_port_bus_type) && dev->driver) {
 		service_driver = to_service_driver(dev->driver);
-		if (service_driver->suspend)
+		cb = *(pcie_pm_callback_t *)((void *)service_driver + offset);
-			service_driver->suspend(to_pcie_device(dev));
+		if (cb)
+			return cb(to_pcie_device(dev));
 	}
 	return 0;
 }
@@ -371,20 +376,14 @@ static int suspend_iter(struct device *dev, void *data)
  */
 int pcie_port_device_suspend(struct device *dev)
 {
-	return device_for_each_child(dev, NULL, suspend_iter);
+	size_t off = offsetof(struct pcie_port_service_driver, suspend);
+	return device_for_each_child(dev, &off, pm_iter);
 }
 
-static int resume_iter(struct device *dev, void *data)
+int pcie_port_device_resume_noirq(struct device *dev)
 {
-	struct pcie_port_service_driver *service_driver;
+	size_t off = offsetof(struct pcie_port_service_driver, resume_noirq);
-
+	return device_for_each_child(dev, &off, pm_iter);
-	if ((dev->bus == &pcie_port_bus_type) &&
-	    (dev->driver)) {
-		service_driver = to_service_driver(dev->driver);
-		if (service_driver->resume)
-			service_driver->resume(to_pcie_device(dev));
-	}
-	return 0;
 }
 
 /**
@@ -393,7 +392,8 @@ static int resume_iter(struct device *dev, void *data)
  */
 int pcie_port_device_resume(struct device *dev)
 {
-	return device_for_each_child(dev, NULL, resume_iter);
+	size_t off = offsetof(struct pcie_port_service_driver, resume);
+	return device_for_each_child(dev, &off, pm_iter);
 }
 #endif /* PM */
 

drivers/pci/pcie/portdrv_pci.c

@@ -65,10 +65,12 @@ static int pcie_port_runtime_idle(struct device *dev)
 
 static const struct dev_pm_ops pcie_portdrv_pm_ops = {
 	.suspend	= pcie_port_device_suspend,
+	.resume_noirq	= pcie_port_device_resume_noirq,
 	.resume		= pcie_port_device_resume,
 	.freeze		= pcie_port_device_suspend,
 	.thaw		= pcie_port_device_resume,
 	.poweroff	= pcie_port_device_suspend,
+	.restore_noirq	= pcie_port_device_resume_noirq,
 	.restore	= pcie_port_device_resume,
 	.runtime_suspend = pcie_port_runtime_suspend,
 	.runtime_resume	= pcie_port_runtime_resume,

drivers/platform/x86/asus-wmi.c

@@ -858,12 +858,6 @@ static int asus_get_adapter_status(struct hotplug_slot *hotplug_slot,
 	return 0;
 }
 
-static void asus_cleanup_pci_hotplug(struct hotplug_slot *hotplug_slot)
-{
-	kfree(hotplug_slot->info);
-	kfree(hotplug_slot);
-}
-
 static struct hotplug_slot_ops asus_hotplug_slot_ops = {
 	.owner = THIS_MODULE,
 	.get_adapter_status = asus_get_adapter_status,
@@ -905,7 +899,6 @@ static int asus_setup_pci_hotplug(struct asus_wmi *asus)
 		goto error_info;
 
 	asus->hotplug_slot->private = asus;
-	asus->hotplug_slot->release = &asus_cleanup_pci_hotplug;
 	asus->hotplug_slot->ops = &asus_hotplug_slot_ops;
 	asus_get_adapter_status(asus->hotplug_slot,
 				&asus->hotplug_slot->info->adapter_status);
@@ -1051,8 +1044,11 @@ static void asus_wmi_rfkill_exit(struct asus_wmi *asus)
 	 * asus_unregister_rfkill_notifier()
 	 */
 	asus_rfkill_hotplug(asus);
-	if (asus->hotplug_slot)
+	if (asus->hotplug_slot) {
 		pci_hp_deregister(asus->hotplug_slot);
+		kfree(asus->hotplug_slot->info);
+		kfree(asus->hotplug_slot);
+	}
 	if (asus->hotplug_workqueue)
 		destroy_workqueue(asus->hotplug_workqueue);
 

drivers/platform/x86/eeepc-laptop.c

@@ -726,12 +726,6 @@ static int eeepc_get_adapter_status(struct hotplug_slot *hotplug_slot,
 	return 0;
 }
 
-static void eeepc_cleanup_pci_hotplug(struct hotplug_slot *hotplug_slot)
-{
-	kfree(hotplug_slot->info);
-	kfree(hotplug_slot);
-}
-
 static struct hotplug_slot_ops eeepc_hotplug_slot_ops = {
 	.owner = THIS_MODULE,
 	.get_adapter_status = eeepc_get_adapter_status,
@@ -758,7 +752,6 @@ static int eeepc_setup_pci_hotplug(struct eeepc_laptop *eeepc)
 		goto error_info;
 
 	eeepc->hotplug_slot->private = eeepc;
-	eeepc->hotplug_slot->release = &eeepc_cleanup_pci_hotplug;
 	eeepc->hotplug_slot->ops = &eeepc_hotplug_slot_ops;
 	eeepc_get_adapter_status(eeepc->hotplug_slot,
 				 &eeepc->hotplug_slot->info->adapter_status);
@@ -837,8 +830,11 @@ static void eeepc_rfkill_exit(struct eeepc_laptop *eeepc)
 		eeepc->wlan_rfkill = NULL;
 	}
 
-	if (eeepc->hotplug_slot)
+	if (eeepc->hotplug_slot) {
 		pci_hp_deregister(eeepc->hotplug_slot);
+		kfree(eeepc->hotplug_slot->info);
+		kfree(eeepc->hotplug_slot);
+	}
 
 	if (eeepc->bluetooth_rfkill) {
 		rfkill_unregister(eeepc->bluetooth_rfkill);

include/linux/pci.h

@@ -388,6 +388,7 @@ struct pci_dev {
 	unsigned int	is_virtfn:1;
 	unsigned int	reset_fn:1;
 	unsigned int	is_hotplug_bridge:1;
+	unsigned int	shpc_managed:1;		/* SHPC owned by shpchp */
 	unsigned int	is_thunderbolt:1;	/* Thunderbolt controller */
 	unsigned int	__aer_firmware_first_valid:1;
 	unsigned int	__aer_firmware_first:1;

include/linux/pci_hotplug.h

@@ -80,15 +80,12 @@ struct hotplug_slot_info {
  * @ops: pointer to the &struct hotplug_slot_ops to be used for this slot
  * @info: pointer to the &struct hotplug_slot_info for the initial values for
  * this slot.
- * @release: called during pci_hp_deregister to free memory allocated in a
- * hotplug_slot structure.
  * @private: used by the hotplug pci controller driver to store whatever it
  * needs.
  */
 struct hotplug_slot {
 	struct hotplug_slot_ops		*ops;
 	struct hotplug_slot_info	*info;
-	void (*release) (struct hotplug_slot *slot);
 	void				*private;
 
 	/* Variables below this are for use only by the hotplug pci core. */
@@ -104,13 +101,23 @@ static inline const char *hotplug_slot_name(const struct hotplug_slot *slot)
 int __pci_hp_register(struct hotplug_slot *slot, struct pci_bus *pbus, int nr,
 		      const char *name, struct module *owner,
 		      const char *mod_name);
-int pci_hp_deregister(struct hotplug_slot *slot);
+int __pci_hp_initialize(struct hotplug_slot *slot, struct pci_bus *bus, int nr,
+			const char *name, struct module *owner,
+			const char *mod_name);
+int pci_hp_add(struct hotplug_slot *slot);
+
+void pci_hp_del(struct hotplug_slot *slot);
+void pci_hp_destroy(struct hotplug_slot *slot);
+void pci_hp_deregister(struct hotplug_slot *slot);
+
 int __must_check pci_hp_change_slot_info(struct hotplug_slot *slot,
 					 struct hotplug_slot_info *info);
 
 /* use a define to avoid include chaining to get THIS_MODULE & friends */
 #define pci_hp_register(slot, pbus, devnr, name) \
 	__pci_hp_register(slot, pbus, devnr, name, THIS_MODULE, KBUILD_MODNAME)
+#define pci_hp_initialize(slot, bus, nr, name) \
+	__pci_hp_initialize(slot, bus, nr, name, THIS_MODULE, KBUILD_MODNAME)
 
 /* PCI Setting Record (Type 0) */
 struct hpp_type0 {