From: Don Dutile <ddutile@redhat.com>
Date: Tue, 1 Dec 2009 23:43:24 -0500
Subject: [pci] intel-iommu: add hot (un)plug support
Message-id: <4B15AA1C.7040501@redhat.com>
Patchwork-id: 21634
O-Subject: [RHEL5.5 PATCH V2] 6/9: INTEL-IOMMU: Add PCI Hot (un)Plug support
Bugzilla: 516811 518103
RH-Acked-by: Chris Wright <chrisw@redhat.com>
RH-Acked-by: Prarit Bhargava <prarit@redhat.com>

V2: fixed indenting.

BZ 516727

Backport of following patch from iommu-2.6.32.git tree:

    intel-iommu: Support PCIe hot-plug
    commit 99dcadede42f8898d4c963ef69192ef4b9b76ba8

To support PCIe hot plug in IOMMU, we register a notifier to respond to device
change action.

When the notifier gets BUS_NOTIFY_UNBOUND_DRIVER, it removes the device
from its DMAR domain.

A hot added device will be added into an IOMMU domain when it first does IOMMU
op. So there is no need to add more code for hot add.

Without the patch, after a hot-remove, a hot-added device on the same
slot will not work.

0/9 states build & tests.

Please review and ACK.

- Don
>From 071bfdd509bc6da9958b7fc7fb44e442058726af Mon Sep 17 00:00:00 2001
From: Donald Dutile <ddutile@redhat.com>
Date: Tue, 17 Nov 2009 17:03:56 -0500
Subject: [PATCH 6/9]   INTEL-IOMMU: Add PCI Hot (un)Plug support


diff --git a/drivers/base/dd.c b/drivers/base/dd.c
index bce7b65..669e71d 100644
--- a/drivers/base/dd.c
+++ b/drivers/base/dd.c
@@ -235,6 +235,14 @@ static void __device_release_driver(struct device * dev)
 			drv->remove(dev);
 		devres_release_all(dev);
 		dev->driver = NULL;
+		if (dev->bus) {
+			struct blocking_notifier_head *notifier_head;
+
+			notifier_head = get_notifier_for_bus(dev->bus);
+			if (notifier_head)
+				blocking_notifier_call_chain(notifier_head,
+						BUS_NOTIFY_UNBOUND_DRIVER, dev);
+		}
 		put_driver(drv);
 	}
 }
diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c
index 42a596d..69622a0 100644
--- a/drivers/pci/intel-iommu.c
+++ b/drivers/pci/intel-iommu.c
@@ -3228,6 +3228,33 @@ static int __init init_iommu_sysfs(void)
 }
 #endif	/* CONFIG_PM */
 
+/*
+ * Here we only respond to action of unbound device from driver.
+ *
+ * Added device is not attached to its DMAR domain here yet. That will happen
+ * when mapping the device to iova.
+ */
+static int device_notifier(struct notifier_block *nb,
+				unsigned long action, void *data)
+{
+	struct device *dev = data;
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct dmar_domain *domain;
+
+	domain = find_domain(pdev);
+	if (!domain)
+		return 0;
+
+	if (action == BUS_NOTIFY_UNBOUND_DRIVER && !iommu_pass_through)
+		domain_remove_one_dev_info(domain, pdev);
+
+	return 0;
+}
+
+static struct notifier_block device_nb = {
+	.notifier_call = device_notifier,
+};
+
 int __init intel_iommu_init(void)
 {
 	int ret = 0;
@@ -3268,6 +3295,8 @@ int __init intel_iommu_init(void)
 
 	register_iommu(&intel_iommu_ops);
 
+	bus_register_notifier(&pci_bus_type, &device_nb);
+
 	return 0;
 }
 
diff --git a/include/linux/device.h b/include/linux/device.h
index 9cd3442..774b167 100644
--- a/include/linux/device.h
+++ b/include/linux/device.h
@@ -93,6 +93,8 @@ extern struct blocking_notifier_head *get_notifier_for_bus(struct bus_type *bus)
 #define BUS_NOTIFY_DEL_DEVICE		0x00000002 /* device removed */
 #define BUS_NOTIFY_BOUND_DRIVER		0x00000003 /* driver bound to device */
 #define BUS_NOTIFY_UNBIND_DRIVER	0x00000004 /* about to be unbound */
+#define BUS_NOTIFY_UNBOUND_DRIVER	0x00000005 /* driver is unbound
+						      from the device */
 
 /* driverfs interface for exporting bus attributes */