[mirror_ubuntu-artful-kernel.git] / drivers / virtio / virtio_pci.c

/*
 * Virtio PCI driver
 *
 * This module allows virtio devices to be used over a virtual PCI device.
 * This can be used with QEMU based VMMs like KVM or Xen.
 *
 * Copyright IBM Corp. 2007
 *
 * Authors:
 *  Anthony Liguori  <aliguori@us.ibm.com>
 *
 * This work is licensed under the terms of the GNU GPL, version 2 or later.
 * See the COPYING file in the top-level directory.
 *
 */

#include <linux/module.h>
#include <linux/list.h>
#include <linux/pci.h>
#include <linux/interrupt.h>
#include <linux/virtio.h>
#include <linux/virtio_config.h>
#include <linux/virtio_ring.h>
#include <linux/virtio_pci.h>
#include <linux/highmem.h>
#include <linux/spinlock.h>

MODULE_AUTHOR("Anthony Liguori <aliguori@us.ibm.com>");
MODULE_DESCRIPTION("virtio-pci");
MODULE_LICENSE("GPL");
MODULE_VERSION("1");

/* Our device structure */
struct virtio_pci_device
{
	struct virtio_device vdev;
	struct pci_dev *pci_dev;

	/* the IO mapping for the PCI config space */
	void __iomem *ioaddr;

	/* a list of queues so we can dispatch IRQs */
	spinlock_t lock;
	struct list_head virtqueues;
};

struct virtio_pci_vq_info
{
	/* the actual virtqueue */
	struct virtqueue *vq;

	/* the number of entries in the queue */
	int num;

	/* the index of the queue */
	int queue_index;

	/* the virtual address of the ring queue */
	void *queue;

	/* the list node for the virtqueues list */
	struct list_head node;
};

/* Qumranet donated their vendor ID for devices 0x1000 thru 0x10FF. */
static struct pci_device_id virtio_pci_id_table[] = {
	{ 0x1af4, PCI_ANY_ID, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 },
	{ 0 },
};

MODULE_DEVICE_TABLE(pci, virtio_pci_id_table);

/* A PCI device has it's own struct device and so does a virtio device so
 * we create a place for the virtio devices to show up in sysfs.  I think it
 * would make more sense for virtio to not insist on having it's own device. */
static struct device virtio_pci_root = {
	.parent		= NULL,
	.bus_id		= "virtio-pci",
};

/* Unique numbering for devices under the kvm root */
static unsigned int dev_index;

/* Convert a generic virtio device to our structure */
static struct virtio_pci_device *to_vp_device(struct virtio_device *vdev)
{
	return container_of(vdev, struct virtio_pci_device, vdev);
}

/* virtio config->get_features() implementation */
static u32 vp_get_features(struct virtio_device *vdev)
{
	struct virtio_pci_device *vp_dev = to_vp_device(vdev);

	/* When someone needs more than 32 feature bits, we'll need to
	 * steal a bit to indicate that the rest are somewhere else. */
	return ioread32(vp_dev->ioaddr + VIRTIO_PCI_HOST_FEATURES);
}

/* virtio config->set_features() implementation */
static void vp_set_features(struct virtio_device *vdev, u32 features)
{
	struct virtio_pci_device *vp_dev = to_vp_device(vdev);

	iowrite32(features, vp_dev->ioaddr + VIRTIO_PCI_GUEST_FEATURES);
}

/* virtio config->get() implementation */
static void vp_get(struct virtio_device *vdev, unsigned offset,
		   void *buf, unsigned len)
{
	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
	void __iomem *ioaddr = vp_dev->ioaddr + VIRTIO_PCI_CONFIG + offset;
	u8 *ptr = buf;
	int i;

	for (i = 0; i < len; i++)
		ptr[i] = ioread8(ioaddr + i);
}

/* the config->set() implementation.  it's symmetric to the config->get()
 * implementation */
static void vp_set(struct virtio_device *vdev, unsigned offset,
		   const void *buf, unsigned len)
{
	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
	void __iomem *ioaddr = vp_dev->ioaddr + VIRTIO_PCI_CONFIG + offset;
	const u8 *ptr = buf;
	int i;

	for (i = 0; i < len; i++)
		iowrite8(ptr[i], ioaddr + i);
}

/* config->{get,set}_status() implementations */
static u8 vp_get_status(struct virtio_device *vdev)
{
	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
	return ioread8(vp_dev->ioaddr + VIRTIO_PCI_STATUS);
}

static void vp_set_status(struct virtio_device *vdev, u8 status)
{
	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
	/* We should never be setting status to 0. */
	BUG_ON(status == 0);
	iowrite8(status, vp_dev->ioaddr + VIRTIO_PCI_STATUS);
}

static void vp_reset(struct virtio_device *vdev)
{
	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
	/* 0 status means a reset. */
	iowrite8(0, vp_dev->ioaddr + VIRTIO_PCI_STATUS);
}

/* the notify function used when creating a virt queue */
static void vp_notify(struct virtqueue *vq)
{
	struct virtio_pci_device *vp_dev = to_vp_device(vq->vdev);
	struct virtio_pci_vq_info *info = vq->priv;

	/* we write the queue's selector into the notification register to
	 * signal the other end */
	iowrite16(info->queue_index, vp_dev->ioaddr + VIRTIO_PCI_QUEUE_NOTIFY);
}

/* A small wrapper to also acknowledge the interrupt when it's handled.
 * I really need an EIO hook for the vring so I can ack the interrupt once we
 * know that we'll be handling the IRQ but before we invoke the callback since
 * the callback may notify the host which results in the host attempting to
 * raise an interrupt that we would then mask once we acknowledged the
 * interrupt. */
static irqreturn_t vp_interrupt(int irq, void *opaque)
{
	struct virtio_pci_device *vp_dev = opaque;
	struct virtio_pci_vq_info *info;
	irqreturn_t ret = IRQ_NONE;
	unsigned long flags;
	u8 isr;

	/* reading the ISR has the effect of also clearing it so it's very
	 * important to save off the value. */
	isr = ioread8(vp_dev->ioaddr + VIRTIO_PCI_ISR);

	/* It's definitely not us if the ISR was not high */
	if (!isr)
		return IRQ_NONE;

	/* Configuration change?  Tell driver if it wants to know. */
	if (isr & VIRTIO_PCI_ISR_CONFIG) {
		struct virtio_driver *drv;
		drv = container_of(vp_dev->vdev.dev.driver,
				   struct virtio_driver, driver);

		if (drv->config_changed)
			drv->config_changed(&vp_dev->vdev);
	}

	spin_lock_irqsave(&vp_dev->lock, flags);
	list_for_each_entry(info, &vp_dev->virtqueues, node) {
		if (vring_interrupt(irq, info->vq) == IRQ_HANDLED)
			ret = IRQ_HANDLED;
	}
	spin_unlock_irqrestore(&vp_dev->lock, flags);

	return ret;
}

/* the config->find_vq() implementation */
static struct virtqueue *vp_find_vq(struct virtio_device *vdev, unsigned index,
				    void (*callback)(struct virtqueue *vq))
{
	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
	struct virtio_pci_vq_info *info;
	struct virtqueue *vq;
	unsigned long flags;
	u16 num;
	int err;

	/* Select the queue we're interested in */
	iowrite16(index, vp_dev->ioaddr + VIRTIO_PCI_QUEUE_SEL);

	/* Check if queue is either not available or already active. */
	num = ioread16(vp_dev->ioaddr + VIRTIO_PCI_QUEUE_NUM);
	if (!num || ioread32(vp_dev->ioaddr + VIRTIO_PCI_QUEUE_PFN))
		return ERR_PTR(-ENOENT);

	/* allocate and fill out our structure the represents an active
	 * queue */
	info = kmalloc(sizeof(struct virtio_pci_vq_info), GFP_KERNEL);
	if (!info)
		return ERR_PTR(-ENOMEM);

	info->queue_index = index;
	info->num = num;

	info->queue = kzalloc(PAGE_ALIGN(vring_size(num,PAGE_SIZE)), GFP_KERNEL);
	if (info->queue == NULL) {
		err = -ENOMEM;
		goto out_info;
	}

	/* activate the queue */
	iowrite32(virt_to_phys(info->queue) >> PAGE_SHIFT,
		  vp_dev->ioaddr + VIRTIO_PCI_QUEUE_PFN);

	/* create the vring */
	vq = vring_new_virtqueue(info->num, vdev, info->queue,
				 vp_notify, callback);
	if (!vq) {
		err = -ENOMEM;
		goto out_activate_queue;
	}

	vq->priv = info;
	info->vq = vq;

	spin_lock_irqsave(&vp_dev->lock, flags);
	list_add(&info->node, &vp_dev->virtqueues);
	spin_unlock_irqrestore(&vp_dev->lock, flags);

	return vq;

out_activate_queue:
	iowrite32(0, vp_dev->ioaddr + VIRTIO_PCI_QUEUE_PFN);
	kfree(info->queue);
out_info:
	kfree(info);
	return ERR_PTR(err);
}

/* the config->del_vq() implementation */
static void vp_del_vq(struct virtqueue *vq)
{
	struct virtio_pci_device *vp_dev = to_vp_device(vq->vdev);
	struct virtio_pci_vq_info *info = vq->priv;
	unsigned long flags;

	spin_lock_irqsave(&vp_dev->lock, flags);
	list_del(&info->node);
	spin_unlock_irqrestore(&vp_dev->lock, flags);

	vring_del_virtqueue(vq);

	/* Select and deactivate the queue */
	iowrite16(info->queue_index, vp_dev->ioaddr + VIRTIO_PCI_QUEUE_SEL);
	iowrite32(0, vp_dev->ioaddr + VIRTIO_PCI_QUEUE_PFN);

	kfree(info->queue);
	kfree(info);
}

static struct virtio_config_ops virtio_pci_config_ops = {
	.get		= vp_get,
	.set		= vp_set,
	.get_status	= vp_get_status,
	.set_status	= vp_set_status,
	.reset		= vp_reset,
	.find_vq	= vp_find_vq,
	.del_vq		= vp_del_vq,
	.get_features	= vp_get_features,
	.set_features	= vp_set_features,
};

/* the PCI probing function */
static int __devinit virtio_pci_probe(struct pci_dev *pci_dev,
				      const struct pci_device_id *id)
{
	struct virtio_pci_device *vp_dev;
	int err;

	/* We only own devices >= 0x1000 and <= 0x103f: leave the rest. */
	if (pci_dev->device < 0x1000 || pci_dev->device > 0x103f)
		return -ENODEV;

	if (pci_dev->revision != VIRTIO_PCI_ABI_VERSION) {
		printk(KERN_ERR "virtio_pci: expected ABI version %d, got %d\n",
		       VIRTIO_PCI_ABI_VERSION, pci_dev->revision);
		return -ENODEV;
	}

	/* allocate our structure and fill it out */
	vp_dev = kzalloc(sizeof(struct virtio_pci_device), GFP_KERNEL);
	if (vp_dev == NULL)
		return -ENOMEM;

	vp_dev->vdev.index = dev_index;
	dev_index++;

	vp_dev->vdev.dev.parent = &virtio_pci_root;
	vp_dev->vdev.config = &virtio_pci_config_ops;
	vp_dev->pci_dev = pci_dev;
	INIT_LIST_HEAD(&vp_dev->virtqueues);
	spin_lock_init(&vp_dev->lock);

	/* enable the device */
	err = pci_enable_device(pci_dev);
	if (err)
		goto out;

	err = pci_request_regions(pci_dev, "virtio-pci");
	if (err)
		goto out_enable_device;

	vp_dev->ioaddr = pci_iomap(pci_dev, 0, 0);
	if (vp_dev->ioaddr == NULL)
		goto out_req_regions;

	pci_set_drvdata(pci_dev, vp_dev);

	/* we use the subsystem vendor/device id as the virtio vendor/device
	 * id.  this allows us to use the same PCI vendor/device id for all
	 * virtio devices and to identify the particular virtio driver by
	 * the subsytem ids */
	vp_dev->vdev.id.vendor = pci_dev->subsystem_vendor;
	vp_dev->vdev.id.device = pci_dev->subsystem_device;

	/* register a handler for the queue with the PCI device's interrupt */
	err = request_irq(vp_dev->pci_dev->irq, vp_interrupt, IRQF_SHARED,
			  vp_dev->vdev.dev.bus_id, vp_dev);
	if (err)
		goto out_set_drvdata;

	/* finally register the virtio device */
	err = register_virtio_device(&vp_dev->vdev);
	if (err)
		goto out_req_irq;

	return 0;

out_req_irq:
	free_irq(pci_dev->irq, vp_dev);
out_set_drvdata:
	pci_set_drvdata(pci_dev, NULL);
	pci_iounmap(pci_dev, vp_dev->ioaddr);
out_req_regions:
	pci_release_regions(pci_dev);
out_enable_device:
	pci_disable_device(pci_dev);
out:
	kfree(vp_dev);
	return err;
}

static void __devexit virtio_pci_remove(struct pci_dev *pci_dev)
{
	struct virtio_pci_device *vp_dev = pci_get_drvdata(pci_dev);

	unregister_virtio_device(&vp_dev->vdev);
	free_irq(pci_dev->irq, vp_dev);
	pci_set_drvdata(pci_dev, NULL);
	pci_iounmap(pci_dev, vp_dev->ioaddr);
	pci_release_regions(pci_dev);
	pci_disable_device(pci_dev);
	kfree(vp_dev);
}

#ifdef CONFIG_PM
static int virtio_pci_suspend(struct pci_dev *pci_dev, pm_message_t state)
{
	pci_save_state(pci_dev);
	pci_set_power_state(pci_dev, PCI_D3hot);
	return 0;
}

static int virtio_pci_resume(struct pci_dev *pci_dev)
{
	pci_restore_state(pci_dev);
	pci_set_power_state(pci_dev, PCI_D0);
	return 0;
}
#endif

static struct pci_driver virtio_pci_driver = {
	.name		= "virtio-pci",
	.id_table	= virtio_pci_id_table,
	.probe		= virtio_pci_probe,
	.remove		= virtio_pci_remove,
#ifdef CONFIG_PM
	.suspend	= virtio_pci_suspend,
	.resume		= virtio_pci_resume,
#endif
};

static int __init virtio_pci_init(void)
{
	int err;

	err = device_register(&virtio_pci_root);
	if (err)
		return err;

	err = pci_register_driver(&virtio_pci_driver);
	if (err)
		device_unregister(&virtio_pci_root);

	return err;
}

module_init(virtio_pci_init);

static void __exit virtio_pci_exit(void)
{
	device_unregister(&virtio_pci_root);
	pci_unregister_driver(&virtio_pci_driver);
}

module_exit(virtio_pci_exit);
Commit	Line	Data
3343660d AL	1	/*
	2	* Virtio PCI driver
	3	*
	4	* This module allows virtio devices to be used over a virtual PCI device.
	5	* This can be used with QEMU based VMMs like KVM or Xen.
	6	*
	7	* Copyright IBM Corp. 2007
	8	*
	9	* Authors:
	10	* Anthony Liguori <aliguori@us.ibm.com>
	11	*
	12	* This work is licensed under the terms of the GNU GPL, version 2 or later.
	13	* See the COPYING file in the top-level directory.
	14	*
	15	*/
	16
	17	#include <linux/module.h>
	18	#include <linux/list.h>
	19	#include <linux/pci.h>
	20	#include <linux/interrupt.h>
	21	#include <linux/virtio.h>
	22	#include <linux/virtio_config.h>
	23	#include <linux/virtio_ring.h>
	24	#include <linux/virtio_pci.h>
	25	#include <linux/highmem.h>
	26	#include <linux/spinlock.h>
	27
	28	MODULE_AUTHOR("Anthony Liguori <aliguori@us.ibm.com>");
	29	MODULE_DESCRIPTION("virtio-pci");
	30	MODULE_LICENSE("GPL");
	31	MODULE_VERSION("1");
	32
	33	/* Our device structure */
	34	struct virtio_pci_device
	35	{
	36	struct virtio_device vdev;
	37	struct pci_dev *pci_dev;
	38
	39	/* the IO mapping for the PCI config space */
97968358	40	void __iomem *ioaddr;
3343660d AL	41
	42	/* a list of queues so we can dispatch IRQs */
	43	spinlock_t lock;
	44	struct list_head virtqueues;
	45	};
	46
	47	struct virtio_pci_vq_info
	48	{
	49	/* the actual virtqueue */
	50	struct virtqueue *vq;
	51
	52	/* the number of entries in the queue */
	53	int num;
	54
	55	/* the index of the queue */
	56	int queue_index;
	57
	58	/* the virtual address of the ring queue */
	59	void *queue;
	60
	61	/* the list node for the virtqueues list */
	62	struct list_head node;
	63	};
	64
	65	/* Qumranet donated their vendor ID for devices 0x1000 thru 0x10FF. */
	66	static struct pci_device_id virtio_pci_id_table[] = {
	67	{ 0x1af4, PCI_ANY_ID, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0 },
	68	{ 0 },
	69	};
	70
	71	MODULE_DEVICE_TABLE(pci, virtio_pci_id_table);
	72
	73	/* A PCI device has it's own struct device and so does a virtio device so
	74	* we create a place for the virtio devices to show up in sysfs. I think it
	75	* would make more sense for virtio to not insist on having it's own device. */
	76	static struct device virtio_pci_root = {
	77	.parent = NULL,
	78	.bus_id = "virtio-pci",
	79	};
	80
	81	/* Unique numbering for devices under the kvm root */
	82	static unsigned int dev_index;
	83
	84	/* Convert a generic virtio device to our structure */
	85	static struct virtio_pci_device to_vp_device(struct virtio_device vdev)
	86	{
	87	return container_of(vdev, struct virtio_pci_device, vdev);
	88	}
	89
c45a6816 RR	90	/* virtio config->get_features() implementation */
	91	static u32 vp_get_features(struct virtio_device *vdev)
	92	{
	93	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
	94
	95	/* When someone needs more than 32 feature bits, we'll need to
	96	* steal a bit to indicate that the rest are somewhere else. */
	97	return ioread32(vp_dev->ioaddr + VIRTIO_PCI_HOST_FEATURES);
	98	}
	99
	100	/* virtio config->set_features() implementation */
	101	static void vp_set_features(struct virtio_device *vdev, u32 features)
3343660d AL	102	{
3343660d AL	103	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
3343660d	104
c45a6816	105	iowrite32(features, vp_dev->ioaddr + VIRTIO_PCI_GUEST_FEATURES);
3343660d AL	106	}
	107
	108	/* virtio config->get() implementation */
	109	static void vp_get(struct virtio_device *vdev, unsigned offset,
	110	void *buf, unsigned len)
	111	{
	112	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
97968358	113	void __iomem *ioaddr = vp_dev->ioaddr + VIRTIO_PCI_CONFIG + offset;
3343660d AL	114	u8 *ptr = buf;
	115	int i;
	116
	117	for (i = 0; i < len; i++)
	118	ptr[i] = ioread8(ioaddr + i);
	119	}
	120
	121	/* the config->set() implementation. it's symmetric to the config->get()
	122	* implementation */
	123	static void vp_set(struct virtio_device *vdev, unsigned offset,
	124	const void *buf, unsigned len)
	125	{
	126	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
97968358	127	void __iomem *ioaddr = vp_dev->ioaddr + VIRTIO_PCI_CONFIG + offset;
3343660d AL	128	const u8 *ptr = buf;
	129	int i;
	130
	131	for (i = 0; i < len; i++)
	132	iowrite8(ptr[i], ioaddr + i);
	133	}
	134
	135	/* config->{get,set}_status() implementations */
	136	static u8 vp_get_status(struct virtio_device *vdev)
	137	{
	138	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
	139	return ioread8(vp_dev->ioaddr + VIRTIO_PCI_STATUS);
	140	}
	141
	142	static void vp_set_status(struct virtio_device *vdev, u8 status)
	143	{
	144	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
	145	/* We should never be setting status to 0. */
	146	BUG_ON(status == 0);
597d56e4	147	iowrite8(status, vp_dev->ioaddr + VIRTIO_PCI_STATUS);
3343660d AL	148	}
	149
	150	static void vp_reset(struct virtio_device *vdev)
	151	{
	152	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
	153	/* 0 status means a reset. */
597d56e4	154	iowrite8(0, vp_dev->ioaddr + VIRTIO_PCI_STATUS);
3343660d AL	155	}
	156
	157	/* the notify function used when creating a virt queue */
	158	static void vp_notify(struct virtqueue *vq)
	159	{
	160	struct virtio_pci_device *vp_dev = to_vp_device(vq->vdev);
	161	struct virtio_pci_vq_info *info = vq->priv;
	162
	163	/* we write the queue's selector into the notification register to
	164	* signal the other end */
	165	iowrite16(info->queue_index, vp_dev->ioaddr + VIRTIO_PCI_QUEUE_NOTIFY);
	166	}
	167
	168	/* A small wrapper to also acknowledge the interrupt when it's handled.
	169	* I really need an EIO hook for the vring so I can ack the interrupt once we
	170	* know that we'll be handling the IRQ but before we invoke the callback since
	171	* the callback may notify the host which results in the host attempting to
	172	* raise an interrupt that we would then mask once we acknowledged the
	173	* interrupt. */
	174	static irqreturn_t vp_interrupt(int irq, void *opaque)
	175	{
	176	struct virtio_pci_device *vp_dev = opaque;
	177	struct virtio_pci_vq_info *info;
	178	irqreturn_t ret = IRQ_NONE;
27ebe308	179	unsigned long flags;
3343660d AL	180	u8 isr;
	181
	182	/* reading the ISR has the effect of also clearing it so it's very
	183	* important to save off the value. */
	184	isr = ioread8(vp_dev->ioaddr + VIRTIO_PCI_ISR);
	185
	186	/* It's definitely not us if the ISR was not high */
	187	if (!isr)
	188	return IRQ_NONE;
	189
	190	/* Configuration change? Tell driver if it wants to know. */
	191	if (isr & VIRTIO_PCI_ISR_CONFIG) {
	192	struct virtio_driver *drv;
	193	drv = container_of(vp_dev->vdev.dev.driver,
	194	struct virtio_driver, driver);
	195
	196	if (drv->config_changed)
	197	drv->config_changed(&vp_dev->vdev);
	198	}
	199
27ebe308	200	spin_lock_irqsave(&vp_dev->lock, flags);
3343660d AL	201	list_for_each_entry(info, &vp_dev->virtqueues, node) {
	202	if (vring_interrupt(irq, info->vq) == IRQ_HANDLED)
	203	ret = IRQ_HANDLED;
	204	}
27ebe308	205	spin_unlock_irqrestore(&vp_dev->lock, flags);
3343660d AL	206
	207	return ret;
	208	}
	209
	210	/* the config->find_vq() implementation */
	211	static struct virtqueue vp_find_vq(struct virtio_device vdev, unsigned index,
	212	void (callback)(struct virtqueue vq))
	213	{
	214	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
	215	struct virtio_pci_vq_info *info;
	216	struct virtqueue *vq;
27ebe308	217	unsigned long flags;
3343660d AL	218	u16 num;
	219	int err;
	220
	221	/* Select the queue we're interested in */
	222	iowrite16(index, vp_dev->ioaddr + VIRTIO_PCI_QUEUE_SEL);
	223
	224	/* Check if queue is either not available or already active. */
	225	num = ioread16(vp_dev->ioaddr + VIRTIO_PCI_QUEUE_NUM);
	226	if (!num \|\| ioread32(vp_dev->ioaddr + VIRTIO_PCI_QUEUE_PFN))
	227	return ERR_PTR(-ENOENT);
	228
	229	/* allocate and fill out our structure the represents an active
	230	* queue */
	231	info = kmalloc(sizeof(struct virtio_pci_vq_info), GFP_KERNEL);
	232	if (!info)
	233	return ERR_PTR(-ENOMEM);
	234
	235	info->queue_index = index;
	236	info->num = num;
	237
	238	info->queue = kzalloc(PAGE_ALIGN(vring_size(num,PAGE_SIZE)), GFP_KERNEL);
	239	if (info->queue == NULL) {
	240	err = -ENOMEM;
	241	goto out_info;
	242	}
	243
	244	/* activate the queue */
	245	iowrite32(virt_to_phys(info->queue) >> PAGE_SHIFT,
	246	vp_dev->ioaddr + VIRTIO_PCI_QUEUE_PFN);
	247
	248	/* create the vring */
	249	vq = vring_new_virtqueue(info->num, vdev, info->queue,
	250	vp_notify, callback);
	251	if (!vq) {
	252	err = -ENOMEM;
	253	goto out_activate_queue;
	254	}
	255
	256	vq->priv = info;
	257	info->vq = vq;
	258
27ebe308	259	spin_lock_irqsave(&vp_dev->lock, flags);
3343660d	260	list_add(&info->node, &vp_dev->virtqueues);
27ebe308	261	spin_unlock_irqrestore(&vp_dev->lock, flags);
3343660d AL	262
	263	return vq;
	264
	265	out_activate_queue:
	266	iowrite32(0, vp_dev->ioaddr + VIRTIO_PCI_QUEUE_PFN);
	267	kfree(info->queue);
	268	out_info:
	269	kfree(info);
	270	return ERR_PTR(err);
	271	}
	272
	273	/* the config->del_vq() implementation */
	274	static void vp_del_vq(struct virtqueue *vq)
	275	{
	276	struct virtio_pci_device *vp_dev = to_vp_device(vq->vdev);
	277	struct virtio_pci_vq_info *info = vq->priv;
27ebe308	278	unsigned long flags;
3343660d	279
27ebe308	280	spin_lock_irqsave(&vp_dev->lock, flags);
3343660d	281	list_del(&info->node);
27ebe308	282	spin_unlock_irqrestore(&vp_dev->lock, flags);
3343660d AL	283
	284	vring_del_virtqueue(vq);
	285
	286	/* Select and deactivate the queue */
	287	iowrite16(info->queue_index, vp_dev->ioaddr + VIRTIO_PCI_QUEUE_SEL);
	288	iowrite32(0, vp_dev->ioaddr + VIRTIO_PCI_QUEUE_PFN);
	289
	290	kfree(info->queue);
	291	kfree(info);
	292	}
	293
	294	static struct virtio_config_ops virtio_pci_config_ops = {
3343660d AL	295	.get = vp_get,
	296	.set = vp_set,
	297	.get_status = vp_get_status,
	298	.set_status = vp_set_status,
	299	.reset = vp_reset,
	300	.find_vq = vp_find_vq,
	301	.del_vq = vp_del_vq,
c45a6816 RR	302	.get_features = vp_get_features,
c45a6816 RR	303	.set_features = vp_set_features,
3343660d AL	304	};
	305
	306	/* the PCI probing function */
	307	static int __devinit virtio_pci_probe(struct pci_dev *pci_dev,
	308	const struct pci_device_id *id)
	309	{
	310	struct virtio_pci_device *vp_dev;
	311	int err;
	312
	313	/* We only own devices >= 0x1000 and <= 0x103f: leave the rest. */
	314	if (pci_dev->device < 0x1000 \|\| pci_dev->device > 0x103f)
	315	return -ENODEV;
	316
55a7c066 AL	317	if (pci_dev->revision != VIRTIO_PCI_ABI_VERSION) {
	318	printk(KERN_ERR "virtio_pci: expected ABI version %d, got %d\n",
	319	VIRTIO_PCI_ABI_VERSION, pci_dev->revision);
	320	return -ENODEV;
	321	}
	322
3343660d AL	323	/* allocate our structure and fill it out */
	324	vp_dev = kzalloc(sizeof(struct virtio_pci_device), GFP_KERNEL);
	325	if (vp_dev == NULL)
	326	return -ENOMEM;
	327
3343660d AL	328	vp_dev->vdev.index = dev_index;
	329	dev_index++;
	330
	331	vp_dev->vdev.dev.parent = &virtio_pci_root;
	332	vp_dev->vdev.config = &virtio_pci_config_ops;
	333	vp_dev->pci_dev = pci_dev;
	334	INIT_LIST_HEAD(&vp_dev->virtqueues);
	335	spin_lock_init(&vp_dev->lock);
	336
	337	/* enable the device */
	338	err = pci_enable_device(pci_dev);
	339	if (err)
	340	goto out;
	341
	342	err = pci_request_regions(pci_dev, "virtio-pci");
	343	if (err)
	344	goto out_enable_device;
	345
	346	vp_dev->ioaddr = pci_iomap(pci_dev, 0, 0);
	347	if (vp_dev->ioaddr == NULL)
	348	goto out_req_regions;
	349
	350	pci_set_drvdata(pci_dev, vp_dev);
	351
	352	/* we use the subsystem vendor/device id as the virtio vendor/device
	353	* id. this allows us to use the same PCI vendor/device id for all
	354	* virtio devices and to identify the particular virtio driver by
	355	* the subsytem ids */
	356	vp_dev->vdev.id.vendor = pci_dev->subsystem_vendor;
	357	vp_dev->vdev.id.device = pci_dev->subsystem_device;
	358
	359	/* register a handler for the queue with the PCI device's interrupt */
	360	err = request_irq(vp_dev->pci_dev->irq, vp_interrupt, IRQF_SHARED,
	361	vp_dev->vdev.dev.bus_id, vp_dev);
	362	if (err)
	363	goto out_set_drvdata;
	364
	365	/* finally register the virtio device */
	366	err = register_virtio_device(&vp_dev->vdev);
	367	if (err)
	368	goto out_req_irq;
	369
	370	return 0;
	371
	372	out_req_irq:
	373	free_irq(pci_dev->irq, vp_dev);
	374	out_set_drvdata:
	375	pci_set_drvdata(pci_dev, NULL);
	376	pci_iounmap(pci_dev, vp_dev->ioaddr);
	377	out_req_regions:
	378	pci_release_regions(pci_dev);
	379	out_enable_device:
	380	pci_disable_device(pci_dev);
	381	out:
	382	kfree(vp_dev);
	383	return err;
	384	}
	385
	386	static void __devexit virtio_pci_remove(struct pci_dev *pci_dev)
	387	{
	388	struct virtio_pci_device *vp_dev = pci_get_drvdata(pci_dev);
	389
bd6c2690	390	unregister_virtio_device(&vp_dev->vdev);
3343660d AL	391	free_irq(pci_dev->irq, vp_dev);
	392	pci_set_drvdata(pci_dev, NULL);
	393	pci_iounmap(pci_dev, vp_dev->ioaddr);
	394	pci_release_regions(pci_dev);
	395	pci_disable_device(pci_dev);
	396	kfree(vp_dev);
	397	}
	398
	399	#ifdef CONFIG_PM
	400	static int virtio_pci_suspend(struct pci_dev *pci_dev, pm_message_t state)
	401	{
	402	pci_save_state(pci_dev);
	403	pci_set_power_state(pci_dev, PCI_D3hot);
	404	return 0;
	405	}
	406
	407	static int virtio_pci_resume(struct pci_dev *pci_dev)
	408	{
	409	pci_restore_state(pci_dev);
	410	pci_set_power_state(pci_dev, PCI_D0);
	411	return 0;
	412	}
	413	#endif
	414
	415	static struct pci_driver virtio_pci_driver = {
	416	.name = "virtio-pci",
	417	.id_table = virtio_pci_id_table,
	418	.probe = virtio_pci_probe,
	419	.remove = virtio_pci_remove,
	420	#ifdef CONFIG_PM
	421	.suspend = virtio_pci_suspend,
	422	.resume = virtio_pci_resume,
	423	#endif
	424	};
	425
	426	static int __init virtio_pci_init(void)
	427	{
	428	int err;
	429
	430	err = device_register(&virtio_pci_root);
	431	if (err)
	432	return err;
	433
	434	err = pci_register_driver(&virtio_pci_driver);
	435	if (err)
	436	device_unregister(&virtio_pci_root);
	437
	438	return err;
	439	}
	440
	441	module_init(virtio_pci_init);
	442
	443	static void __exit virtio_pci_exit(void)
	444	{
	445	device_unregister(&virtio_pci_root);
	446	pci_unregister_driver(&virtio_pci_driver);
	447	}
	448
	449	module_exit(virtio_pci_exit);