[deliverable/linux.git] / arch / x86 / pci / xen.c

/*
 * Xen PCI - handle PCI (INTx) and MSI infrastructure calls for PV, HVM and
 * initial domain support. We also handle the DSDT _PRT callbacks for GSI's
 * used in HVM and initial domain mode (PV does not parse ACPI, so it has no
 * concept of GSIs). Under PV we hook under the pnbbios API for IRQs and
 * 0xcf8 PCI configuration read/write.
 *
 *   Author: Ryan Wilson <hap9@epoch.ncsc.mil>
 *           Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
 *           Stefano Stabellini <stefano.stabellini@eu.citrix.com>
 */
#include <linux/module.h>
#include <linux/init.h>
#include <linux/pci.h>
#include <linux/acpi.h>

#include <linux/io.h>
#include <asm/io_apic.h>
#include <asm/pci_x86.h>

#include <asm/xen/hypervisor.h>

#include <xen/features.h>
#include <xen/events.h>
#include <asm/xen/pci.h>

static int xen_pcifront_enable_irq(struct pci_dev *dev)
{
	int rc;
	int share = 1;
	int pirq;
	u8 gsi;

	rc = pci_read_config_byte(dev, PCI_INTERRUPT_LINE, &gsi);
	if (rc < 0) {
		dev_warn(&dev->dev, "Xen PCI: failed to read interrupt line: %d\n",
			 rc);
		return rc;
	}

	rc = xen_allocate_pirq_gsi(gsi);
	if (rc < 0) {
		dev_warn(&dev->dev, "Xen PCI: failed to allocate a PIRQ for GSI%d: %d\n",
			 gsi, rc);
		return rc;
	}
	pirq = rc;

	if (gsi < NR_IRQS_LEGACY)
		share = 0;

	rc = xen_bind_pirq_gsi_to_irq(gsi, pirq, share, "pcifront");
	if (rc < 0) {
		dev_warn(&dev->dev, "Xen PCI: failed to bind GSI%d (PIRQ%d) to IRQ: %d\n",
			 gsi, pirq, rc);
		return rc;
	}

	dev->irq = rc;
	dev_info(&dev->dev, "Xen PCI mapped GSI%d to IRQ%d\n", gsi, dev->irq);
	return 0;
}

#ifdef CONFIG_ACPI
static int acpi_register_gsi_xen_hvm(struct device *dev, u32 gsi,
				 int trigger, int polarity)
{
	int rc, irq;
	struct physdev_map_pirq map_irq;
	int shareable = 0;
	char *name;

	if (!xen_hvm_domain())
		return -1;

	map_irq.domid = DOMID_SELF;
	map_irq.type = MAP_PIRQ_TYPE_GSI;
	map_irq.index = gsi;
	map_irq.pirq = -1;

	rc = HYPERVISOR_physdev_op(PHYSDEVOP_map_pirq, &map_irq);
	if (rc) {
		printk(KERN_WARNING "xen map irq failed %d\n", rc);
		return -1;
	}

	if (trigger == ACPI_EDGE_SENSITIVE) {
		shareable = 0;
		name = "ioapic-edge";
	} else {
		shareable = 1;
		name = "ioapic-level";
	}

	irq = xen_bind_pirq_gsi_to_irq(gsi, map_irq.pirq, shareable, name);

	printk(KERN_DEBUG "xen: --> irq=%d, pirq=%d\n", irq, map_irq.pirq);

	return irq;
}
#endif

#ifdef CONFIG_XEN_DOM0
static int xen_register_pirq(u32 gsi, int gsi_override, int triggering)
{
	int rc, pirq, irq = -1;
	struct physdev_map_pirq map_irq;
	int shareable = 0;
	char *name;

	if (!xen_pv_domain())
		return -1;

	if (triggering == ACPI_EDGE_SENSITIVE) {
		shareable = 0;
		name = "ioapic-edge";
	} else {
		shareable = 1;
		name = "ioapic-level";
	}
	pirq = xen_allocate_pirq_gsi(gsi);
	if (pirq < 0)
		goto out;

	if (gsi_override >= 0)
		irq = xen_bind_pirq_gsi_to_irq(gsi_override, pirq, shareable, name);
	else
		irq = xen_bind_pirq_gsi_to_irq(gsi, pirq, shareable, name);
	if (irq < 0)
		goto out;

	printk(KERN_DEBUG "xen: --> pirq=%d -> irq=%d (gsi=%d)\n", pirq, irq, gsi);

	map_irq.domid = DOMID_SELF;
	map_irq.type = MAP_PIRQ_TYPE_GSI;
	map_irq.index = gsi;
	map_irq.pirq = pirq;

	rc = HYPERVISOR_physdev_op(PHYSDEVOP_map_pirq, &map_irq);
	if (rc) {
		printk(KERN_WARNING "xen map irq failed %d\n", rc);
		return -1;
	}

out:
	return irq;
}

static int xen_register_gsi(u32 gsi, int gsi_override, int triggering, int polarity)
{
	int rc, irq;
	struct physdev_setup_gsi setup_gsi;

	if (!xen_pv_domain())
		return -1;

	printk(KERN_DEBUG "xen: registering gsi %u triggering %d polarity %d\n",
			gsi, triggering, polarity);

	irq = xen_register_pirq(gsi, gsi_override, triggering);

	setup_gsi.gsi = gsi;
	setup_gsi.triggering = (triggering == ACPI_EDGE_SENSITIVE ? 0 : 1);
	setup_gsi.polarity = (polarity == ACPI_ACTIVE_HIGH ? 0 : 1);

	rc = HYPERVISOR_physdev_op(PHYSDEVOP_setup_gsi, &setup_gsi);
	if (rc == -EEXIST)
		printk(KERN_INFO "Already setup the GSI :%d\n", gsi);
	else if (rc) {
		printk(KERN_ERR "Failed to setup GSI :%d, err_code:%d\n",
				gsi, rc);
	}

	return irq;
}

static int acpi_register_gsi_xen(struct device *dev, u32 gsi,
				 int trigger, int polarity)
{
	return xen_register_gsi(gsi, -1 /* no GSI override */, trigger, polarity);
}
#endif

#if defined(CONFIG_PCI_MSI)
#include <linux/msi.h>
#include <asm/msidef.h>

struct xen_pci_frontend_ops *xen_pci_frontend;
EXPORT_SYMBOL_GPL(xen_pci_frontend);

static int xen_setup_msi_irqs(struct pci_dev *dev, int nvec, int type)
{
	int irq, ret, i;
	struct msi_desc *msidesc;
	int *v;

	v = kzalloc(sizeof(int) * max(1, nvec), GFP_KERNEL);
	if (!v)
		return -ENOMEM;

	if (type == PCI_CAP_ID_MSIX)
		ret = xen_pci_frontend_enable_msix(dev, v, nvec);
	else
		ret = xen_pci_frontend_enable_msi(dev, v);
	if (ret)
		goto error;
	i = 0;
	list_for_each_entry(msidesc, &dev->msi_list, list) {
		irq = xen_bind_pirq_msi_to_irq(dev, msidesc, v[i], 0,
					       (type == PCI_CAP_ID_MSIX) ?
					       "pcifront-msi-x" :
					       "pcifront-msi",
						DOMID_SELF);
		if (irq < 0)
			goto free;
		i++;
	}
	kfree(v);
	return 0;

error:
	dev_err(&dev->dev, "Xen PCI frontend has not registered MSI/MSI-X support!\n");
free:
	kfree(v);
	return ret;
}

#define XEN_PIRQ_MSI_DATA  (MSI_DATA_TRIGGER_EDGE | \
		MSI_DATA_LEVEL_ASSERT | (3 << 8) | MSI_DATA_VECTOR(0))

static void xen_msi_compose_msg(struct pci_dev *pdev, unsigned int pirq,
		struct msi_msg *msg)
{
	/* We set vector == 0 to tell the hypervisor we don't care about it,
	 * but we want a pirq setup instead.
	 * We use the dest_id field to pass the pirq that we want. */
	msg->address_hi = MSI_ADDR_BASE_HI | MSI_ADDR_EXT_DEST_ID(pirq);
	msg->address_lo =
		MSI_ADDR_BASE_LO |
		MSI_ADDR_DEST_MODE_PHYSICAL |
		MSI_ADDR_REDIRECTION_CPU |
		MSI_ADDR_DEST_ID(pirq);

	msg->data = XEN_PIRQ_MSI_DATA;
}

static int xen_hvm_setup_msi_irqs(struct pci_dev *dev, int nvec, int type)
{
	int irq, pirq;
	struct msi_desc *msidesc;
	struct msi_msg msg;

	list_for_each_entry(msidesc, &dev->msi_list, list) {
		__read_msi_msg(msidesc, &msg);
		pirq = MSI_ADDR_EXT_DEST_ID(msg.address_hi) |
			((msg.address_lo >> MSI_ADDR_DEST_ID_SHIFT) & 0xff);
		if (msg.data != XEN_PIRQ_MSI_DATA ||
		    xen_irq_from_pirq(pirq) < 0) {
			pirq = xen_allocate_pirq_msi(dev, msidesc);
			if (pirq < 0)
				goto error;
			xen_msi_compose_msg(dev, pirq, &msg);
			__write_msi_msg(msidesc, &msg);
			dev_dbg(&dev->dev, "xen: msi bound to pirq=%d\n", pirq);
		} else {
			dev_dbg(&dev->dev,
				"xen: msi already bound to pirq=%d\n", pirq);
		}
		irq = xen_bind_pirq_msi_to_irq(dev, msidesc, pirq, 0,
					       (type == PCI_CAP_ID_MSIX) ?
					       "msi-x" : "msi",
					       DOMID_SELF);
		if (irq < 0)
			goto error;
		dev_dbg(&dev->dev,
			"xen: msi --> pirq=%d --> irq=%d\n", pirq, irq);
	}
	return 0;

error:
	dev_err(&dev->dev,
		"Xen PCI frontend has not registered MSI/MSI-X support!\n");
	return -ENODEV;
}

#ifdef CONFIG_XEN_DOM0
static int xen_initdom_setup_msi_irqs(struct pci_dev *dev, int nvec, int type)
{
	int ret = 0;
	struct msi_desc *msidesc;

	list_for_each_entry(msidesc, &dev->msi_list, list) {
		struct physdev_map_pirq map_irq;
		domid_t domid;

		domid = ret = xen_find_device_domain_owner(dev);
		/* N.B. Casting int's -ENODEV to uint16_t results in 0xFFED,
		 * hence check ret value for < 0. */
		if (ret < 0)
			domid = DOMID_SELF;

		memset(&map_irq, 0, sizeof(map_irq));
		map_irq.domid = domid;
		map_irq.type = MAP_PIRQ_TYPE_MSI;
		map_irq.index = -1;
		map_irq.pirq = -1;
		map_irq.bus = dev->bus->number;
		map_irq.devfn = dev->devfn;

		if (type == PCI_CAP_ID_MSIX) {
			int pos;
			u32 table_offset, bir;

			pos = pci_find_capability(dev, PCI_CAP_ID_MSIX);

			pci_read_config_dword(dev, pos + PCI_MSIX_TABLE,
					      &table_offset);
			bir = (u8)(table_offset & PCI_MSIX_FLAGS_BIRMASK);

			map_irq.table_base = pci_resource_start(dev, bir);
			map_irq.entry_nr = msidesc->msi_attrib.entry_nr;
		}

		ret = HYPERVISOR_physdev_op(PHYSDEVOP_map_pirq, &map_irq);
		if (ret) {
			dev_warn(&dev->dev, "xen map irq failed %d for %d domain\n",
				 ret, domid);
			goto out;
		}

		ret = xen_bind_pirq_msi_to_irq(dev, msidesc,
					       map_irq.pirq, map_irq.index,
					       (type == PCI_CAP_ID_MSIX) ?
					       "msi-x" : "msi",
						domid);
		if (ret < 0)
			goto out;
	}
	ret = 0;
out:
	return ret;
}
#endif

static void xen_teardown_msi_irqs(struct pci_dev *dev)
{
	struct msi_desc *msidesc;

	msidesc = list_entry(dev->msi_list.next, struct msi_desc, list);
	if (msidesc->msi_attrib.is_msix)
		xen_pci_frontend_disable_msix(dev);
	else
		xen_pci_frontend_disable_msi(dev);

	/* Free the IRQ's and the msidesc using the generic code. */
	default_teardown_msi_irqs(dev);
}

static void xen_teardown_msi_irq(unsigned int irq)
{
	xen_destroy_irq(irq);
}

#endif

int __init pci_xen_init(void)
{
	if (!xen_pv_domain() || xen_initial_domain())
		return -ENODEV;

	printk(KERN_INFO "PCI: setting up Xen PCI frontend stub\n");

	pcibios_set_cache_line_size();

	pcibios_enable_irq = xen_pcifront_enable_irq;
	pcibios_disable_irq = NULL;

#ifdef CONFIG_ACPI
	/* Keep ACPI out of the picture */
	acpi_noirq = 1;
#endif

#ifdef CONFIG_PCI_MSI
	x86_msi.setup_msi_irqs = xen_setup_msi_irqs;
	x86_msi.teardown_msi_irq = xen_teardown_msi_irq;
	x86_msi.teardown_msi_irqs = xen_teardown_msi_irqs;
#endif
	return 0;
}

int __init pci_xen_hvm_init(void)
{
	if (!xen_feature(XENFEAT_hvm_pirqs))
		return 0;

#ifdef CONFIG_ACPI
	/*
	 * We don't want to change the actual ACPI delivery model,
	 * just how GSIs get registered.
	 */
	__acpi_register_gsi = acpi_register_gsi_xen_hvm;
#endif

#ifdef CONFIG_PCI_MSI
	x86_msi.setup_msi_irqs = xen_hvm_setup_msi_irqs;
	x86_msi.teardown_msi_irq = xen_teardown_msi_irq;
#endif
	return 0;
}

#ifdef CONFIG_XEN_DOM0
static __init void xen_setup_acpi_sci(void)
{
	int rc;
	int trigger, polarity;
	int gsi = acpi_sci_override_gsi;
	int irq = -1;
	int gsi_override = -1;

	if (!gsi)
		return;

	rc = acpi_get_override_irq(gsi, &trigger, &polarity);
	if (rc) {
		printk(KERN_WARNING "xen: acpi_get_override_irq failed for acpi"
				" sci, rc=%d\n", rc);
		return;
	}
	trigger = trigger ? ACPI_LEVEL_SENSITIVE : ACPI_EDGE_SENSITIVE;
	polarity = polarity ? ACPI_ACTIVE_LOW : ACPI_ACTIVE_HIGH;

	printk(KERN_INFO "xen: sci override: global_irq=%d trigger=%d "
			"polarity=%d\n", gsi, trigger, polarity);

	/* Before we bind the GSI to a Linux IRQ, check whether
	 * we need to override it with bus_irq (IRQ) value. Usually for
	 * IRQs below IRQ_LEGACY_IRQ this holds IRQ == GSI, as so:
	 *  ACPI: INT_SRC_OVR (bus 0 bus_irq 9 global_irq 9 low level)
	 * but there are oddballs where the IRQ != GSI:
	 *  ACPI: INT_SRC_OVR (bus 0 bus_irq 9 global_irq 20 low level)
	 * which ends up being: gsi_to_irq[9] == 20
	 * (which is what acpi_gsi_to_irq ends up calling when starting the
	 * the ACPI interpreter and keels over since IRQ 9 has not been
	 * setup as we had setup IRQ 20 for it).
	 */
	/* Check whether the GSI != IRQ */
	if (acpi_gsi_to_irq(gsi, &irq) == 0) {
		if (irq >= 0 && irq != gsi)
			/* Bugger, we MUST have that IRQ. */
			gsi_override = irq;
	}

	gsi = xen_register_gsi(gsi, gsi_override, trigger, polarity);
	printk(KERN_INFO "xen: acpi sci %d\n", gsi);

	return;
}

static int __init pci_xen_initial_domain(void)
{
#ifdef CONFIG_PCI_MSI
	x86_msi.setup_msi_irqs = xen_initdom_setup_msi_irqs;
	x86_msi.teardown_msi_irq = xen_teardown_msi_irq;
#endif
	xen_setup_acpi_sci();
	__acpi_register_gsi = acpi_register_gsi_xen;

	return 0;
}

void __init xen_setup_pirqs(void)
{
	int pirq, irq;

	pci_xen_initial_domain();

	if (0 == nr_ioapics) {
		for (irq = 0; irq < NR_IRQS_LEGACY; irq++) {
			pirq = xen_allocate_pirq_gsi(irq);
			if (WARN(pirq < 0,
				 "Could not allocate PIRQ for legacy interrupt\n"))
				break;
			irq = xen_bind_pirq_gsi_to_irq(irq, pirq, 0, "xt-pic");
		}
		return;
	}

	/* Pre-allocate legacy irqs */
	for (irq = 0; irq < NR_IRQS_LEGACY; irq++) {
		int trigger, polarity;

		if (acpi_get_override_irq(irq, &trigger, &polarity) == -1)
			continue;

		xen_register_pirq(irq, -1 /* no GSI override */,
			trigger ? ACPI_LEVEL_SENSITIVE : ACPI_EDGE_SENSITIVE);
	}
}

struct xen_device_domain_owner {
	domid_t domain;
	struct pci_dev *dev;
	struct list_head list;
};

static DEFINE_SPINLOCK(dev_domain_list_spinlock);
static struct list_head dev_domain_list = LIST_HEAD_INIT(dev_domain_list);

static struct xen_device_domain_owner *find_device(struct pci_dev *dev)
{
	struct xen_device_domain_owner *owner;

	list_for_each_entry(owner, &dev_domain_list, list) {
		if (owner->dev == dev)
			return owner;
	}
	return NULL;
}

int xen_find_device_domain_owner(struct pci_dev *dev)
{
	struct xen_device_domain_owner *owner;
	int domain = -ENODEV;

	spin_lock(&dev_domain_list_spinlock);
	owner = find_device(dev);
	if (owner)
		domain = owner->domain;
	spin_unlock(&dev_domain_list_spinlock);
	return domain;
}
EXPORT_SYMBOL_GPL(xen_find_device_domain_owner);

int xen_register_device_domain_owner(struct pci_dev *dev, uint16_t domain)
{
	struct xen_device_domain_owner *owner;

	owner = kzalloc(sizeof(struct xen_device_domain_owner), GFP_KERNEL);
	if (!owner)
		return -ENODEV;

	spin_lock(&dev_domain_list_spinlock);
	if (find_device(dev)) {
		spin_unlock(&dev_domain_list_spinlock);
		kfree(owner);
		return -EEXIST;
	}
	owner->domain = domain;
	owner->dev = dev;
	list_add_tail(&owner->list, &dev_domain_list);
	spin_unlock(&dev_domain_list_spinlock);
	return 0;
}
EXPORT_SYMBOL_GPL(xen_register_device_domain_owner);

int xen_unregister_device_domain_owner(struct pci_dev *dev)
{
	struct xen_device_domain_owner *owner;

	spin_lock(&dev_domain_list_spinlock);
	owner = find_device(dev);
	if (!owner) {
		spin_unlock(&dev_domain_list_spinlock);
		return -ENODEV;
	}
	list_del(&owner->list);
	spin_unlock(&dev_domain_list_spinlock);
	kfree(owner);
	return 0;
}
EXPORT_SYMBOL_GPL(xen_unregister_device_domain_owner);
#endif
Commit	Line	Data
b5401a96	1	/*
996c34ae KRW	2	* Xen PCI - handle PCI (INTx) and MSI infrastructure calls for PV, HVM and
	3	* initial domain support. We also handle the DSDT _PRT callbacks for GSI's
	4	* used in HVM and initial domain mode (PV does not parse ACPI, so it has no
	5	* concept of GSIs). Under PV we hook under the pnbbios API for IRQs and
	6	* 0xcf8 PCI configuration read/write.
b5401a96 AN	7	*
b5401a96 AN	8	* Author: Ryan Wilson <hap9@epoch.ncsc.mil>
996c34ae KRW	9	* Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
996c34ae KRW	10	* Stefano Stabellini <stefano.stabellini@eu.citrix.com>
b5401a96 AN	11	*/
	12	#include <linux/module.h>
	13	#include <linux/init.h>
	14	#include <linux/pci.h>
	15	#include <linux/acpi.h>
	16
	17	#include <linux/io.h>
0e058e52	18	#include <asm/io_apic.h>
b5401a96 AN	19	#include <asm/pci_x86.h>
	20
	21	#include <asm/xen/hypervisor.h>
	22
3942b740	23	#include <xen/features.h>
b5401a96 AN	24	#include <xen/events.h>
	25	#include <asm/xen/pci.h>
	26
fef6e262 KRW	27	static int xen_pcifront_enable_irq(struct pci_dev *dev)
	28	{
	29	int rc;
	30	int share = 1;
	31	int pirq;
	32	u8 gsi;
	33
	34	rc = pci_read_config_byte(dev, PCI_INTERRUPT_LINE, &gsi);
	35	if (rc < 0) {
	36	dev_warn(&dev->dev, "Xen PCI: failed to read interrupt line: %d\n",
	37	rc);
	38	return rc;
	39	}
	40
	41	rc = xen_allocate_pirq_gsi(gsi);
	42	if (rc < 0) {
	43	dev_warn(&dev->dev, "Xen PCI: failed to allocate a PIRQ for GSI%d: %d\n",
	44	gsi, rc);
	45	return rc;
	46	}
	47	pirq = rc;
	48
	49	if (gsi < NR_IRQS_LEGACY)
	50	share = 0;
	51
	52	rc = xen_bind_pirq_gsi_to_irq(gsi, pirq, share, "pcifront");
	53	if (rc < 0) {
	54	dev_warn(&dev->dev, "Xen PCI: failed to bind GSI%d (PIRQ%d) to IRQ: %d\n",
	55	gsi, pirq, rc);
	56	return rc;
	57	}
	58
	59	dev->irq = rc;
	60	dev_info(&dev->dev, "Xen PCI mapped GSI%d to IRQ%d\n", gsi, dev->irq);
	61	return 0;
	62	}
	63
42a1de56	64	#ifdef CONFIG_ACPI
9a626612 IC	65	static int acpi_register_gsi_xen_hvm(struct device *dev, u32 gsi,
9a626612 IC	66	int trigger, int polarity)
42a1de56 SS	67	{
	68	int rc, irq;
	69	struct physdev_map_pirq map_irq;
	70	int shareable = 0;
	71	char *name;
	72
	73	if (!xen_hvm_domain())
	74	return -1;
	75
	76	map_irq.domid = DOMID_SELF;
	77	map_irq.type = MAP_PIRQ_TYPE_GSI;
	78	map_irq.index = gsi;
	79	map_irq.pirq = -1;
	80
	81	rc = HYPERVISOR_physdev_op(PHYSDEVOP_map_pirq, &map_irq);
	82	if (rc) {
	83	printk(KERN_WARNING "xen map irq failed %d\n", rc);
	84	return -1;
	85	}
	86
9a626612	87	if (trigger == ACPI_EDGE_SENSITIVE) {
42a1de56 SS	88	shareable = 0;
	89	name = "ioapic-edge";
	90	} else {
	91	shareable = 1;
	92	name = "ioapic-level";
	93	}
	94
f4d0635b	95	irq = xen_bind_pirq_gsi_to_irq(gsi, map_irq.pirq, shareable, name);
42a1de56 SS	96
	97	printk(KERN_DEBUG "xen: --> irq=%d, pirq=%d\n", irq, map_irq.pirq);
	98
	99	return irq;
	100	}
	101	#endif
	102
fef6e262 KRW	103	#ifdef CONFIG_XEN_DOM0
	104	static int xen_register_pirq(u32 gsi, int gsi_override, int triggering)
	105	{
	106	int rc, pirq, irq = -1;
	107	struct physdev_map_pirq map_irq;
	108	int shareable = 0;
	109	char *name;
	110
	111	if (!xen_pv_domain())
	112	return -1;
	113
	114	if (triggering == ACPI_EDGE_SENSITIVE) {
	115	shareable = 0;
	116	name = "ioapic-edge";
	117	} else {
	118	shareable = 1;
	119	name = "ioapic-level";
	120	}
	121	pirq = xen_allocate_pirq_gsi(gsi);
	122	if (pirq < 0)
	123	goto out;
	124
	125	if (gsi_override >= 0)
	126	irq = xen_bind_pirq_gsi_to_irq(gsi_override, pirq, shareable, name);
	127	else
	128	irq = xen_bind_pirq_gsi_to_irq(gsi, pirq, shareable, name);
	129	if (irq < 0)
	130	goto out;
	131
	132	printk(KERN_DEBUG "xen: --> pirq=%d -> irq=%d (gsi=%d)\n", pirq, irq, gsi);
	133
	134	map_irq.domid = DOMID_SELF;
	135	map_irq.type = MAP_PIRQ_TYPE_GSI;
	136	map_irq.index = gsi;
	137	map_irq.pirq = pirq;
	138
	139	rc = HYPERVISOR_physdev_op(PHYSDEVOP_map_pirq, &map_irq);
	140	if (rc) {
	141	printk(KERN_WARNING "xen map irq failed %d\n", rc);
	142	return -1;
	143	}
	144
	145	out:
	146	return irq;
	147	}
	148
	149	static int xen_register_gsi(u32 gsi, int gsi_override, int triggering, int polarity)
	150	{
	151	int rc, irq;
	152	struct physdev_setup_gsi setup_gsi;
	153
	154	if (!xen_pv_domain())
	155	return -1;
	156
	157	printk(KERN_DEBUG "xen: registering gsi %u triggering %d polarity %d\n",
	158	gsi, triggering, polarity);
	159
	160	irq = xen_register_pirq(gsi, gsi_override, triggering);
	161
	162	setup_gsi.gsi = gsi;
	163	setup_gsi.triggering = (triggering == ACPI_EDGE_SENSITIVE ? 0 : 1);
	164	setup_gsi.polarity = (polarity == ACPI_ACTIVE_HIGH ? 0 : 1);
	165
	166	rc = HYPERVISOR_physdev_op(PHYSDEVOP_setup_gsi, &setup_gsi);
167	if (rc == -EEXIST)
168	printk(KERN_INFO "Already setup the GSI :%d\n", gsi);
169	else if (rc) {
170	printk(KERN_ERR "Failed to setup GSI :%d, err_code:%d\n",
171	gsi, rc);
172	}
173
174	return irq;
175	}
176
177	static int acpi_register_gsi_xen(struct device *dev, u32 gsi,
178	int trigger, int polarity)
179	{
180	return xen_register_gsi(gsi, -1 /* no GSI override */, trigger, polarity);
181	}
182	#endif
183
b5401a96 AN	184	#if defined(CONFIG_PCI_MSI)
b5401a96 AN	185	#include <linux/msi.h>
809f9267	186	#include <asm/msidef.h>
b5401a96 AN	187
	188	struct xen_pci_frontend_ops *xen_pci_frontend;
	189	EXPORT_SYMBOL_GPL(xen_pci_frontend);
	190
fef6e262 KRW	191	static int xen_setup_msi_irqs(struct pci_dev *dev, int nvec, int type)
	192	{
	193	int irq, ret, i;
	194	struct msi_desc *msidesc;
	195	int *v;
	196
	197	v = kzalloc(sizeof(int) * max(1, nvec), GFP_KERNEL);
	198	if (!v)
	199	return -ENOMEM;
	200
	201	if (type == PCI_CAP_ID_MSIX)
	202	ret = xen_pci_frontend_enable_msix(dev, v, nvec);
	203	else
	204	ret = xen_pci_frontend_enable_msi(dev, v);
	205	if (ret)
	206	goto error;
	207	i = 0;
	208	list_for_each_entry(msidesc, &dev->msi_list, list) {
	209	irq = xen_bind_pirq_msi_to_irq(dev, msidesc, v[i], 0,
	210	(type == PCI_CAP_ID_MSIX) ?
	211	"pcifront-msi-x" :
	212	"pcifront-msi",
	213	DOMID_SELF);
	214	if (irq < 0)
	215	goto free;
	216	i++;
	217	}
	218	kfree(v);
	219	return 0;
	220
	221	error:
	222	dev_err(&dev->dev, "Xen PCI frontend has not registered MSI/MSI-X support!\n");
	223	free:
	224	kfree(v);
	225	return ret;
	226	}
	227
af42b8d1 SS	228	#define XEN_PIRQ_MSI_DATA (MSI_DATA_TRIGGER_EDGE \| \
	229	MSI_DATA_LEVEL_ASSERT \| (3 << 8) \| MSI_DATA_VECTOR(0))
	230
809f9267 SS	231	static void xen_msi_compose_msg(struct pci_dev *pdev, unsigned int pirq,
	232	struct msi_msg *msg)
	233	{
	234	/* We set vector == 0 to tell the hypervisor we don't care about it,
	235	* but we want a pirq setup instead.
	236	* We use the dest_id field to pass the pirq that we want. */
	237	msg->address_hi = MSI_ADDR_BASE_HI \| MSI_ADDR_EXT_DEST_ID(pirq);
	238	msg->address_lo =
	239	MSI_ADDR_BASE_LO \|
	240	MSI_ADDR_DEST_MODE_PHYSICAL \|
	241	MSI_ADDR_REDIRECTION_CPU \|
	242	MSI_ADDR_DEST_ID(pirq);
	243
af42b8d1	244	msg->data = XEN_PIRQ_MSI_DATA;
809f9267 SS	245	}
	246
	247	static int xen_hvm_setup_msi_irqs(struct pci_dev *dev, int nvec, int type)
	248	{
bf480d95	249	int irq, pirq;
809f9267 SS	250	struct msi_desc *msidesc;
	251	struct msi_msg msg;
	252
	253	list_for_each_entry(msidesc, &dev->msi_list, list) {
af42b8d1 SS	254	__read_msi_msg(msidesc, &msg);
	255	pirq = MSI_ADDR_EXT_DEST_ID(msg.address_hi) \|
	256	((msg.address_lo >> MSI_ADDR_DEST_ID_SHIFT) & 0xff);
bf480d95 IC	257	if (msg.data != XEN_PIRQ_MSI_DATA \|\|
	258	xen_irq_from_pirq(pirq) < 0) {
	259	pirq = xen_allocate_pirq_msi(dev, msidesc);
	260	if (pirq < 0)
af42b8d1	261	goto error;
bf480d95 IC	262	xen_msi_compose_msg(dev, pirq, &msg);
	263	__write_msi_msg(msidesc, &msg);
	264	dev_dbg(&dev->dev, "xen: msi bound to pirq=%d\n", pirq);
	265	} else {
	266	dev_dbg(&dev->dev,
	267	"xen: msi already bound to pirq=%d\n", pirq);
af42b8d1	268	}
ca1d8fe9	269	irq = xen_bind_pirq_msi_to_irq(dev, msidesc, pirq, 0,
bf480d95	270	(type == PCI_CAP_ID_MSIX) ?
beafbdc1 KRW	271	"msi-x" : "msi",
beafbdc1 KRW	272	DOMID_SELF);
bf480d95	273	if (irq < 0)
809f9267	274	goto error;
bf480d95 IC	275	dev_dbg(&dev->dev,
bf480d95 IC	276	"xen: msi --> pirq=%d --> irq=%d\n", pirq, irq);
809f9267 SS	277	}
	278	return 0;
	279
809f9267	280	error:
bf480d95 IC	281	dev_err(&dev->dev,
	282	"Xen PCI frontend has not registered MSI/MSI-X support!\n");
	283	return -ENODEV;
809f9267 SS	284	}
809f9267 SS	285
260a7d4c	286	#ifdef CONFIG_XEN_DOM0
f731e3ef QH	287	static int xen_initdom_setup_msi_irqs(struct pci_dev *dev, int nvec, int type)
f731e3ef QH	288	{
71eef7d1	289	int ret = 0;
f731e3ef QH	290	struct msi_desc *msidesc;
	291
	292	list_for_each_entry(msidesc, &dev->msi_list, list) {
71eef7d1	293	struct physdev_map_pirq map_irq;
beafbdc1 KRW	294	domid_t domid;
	295
	296	domid = ret = xen_find_device_domain_owner(dev);
	297	/* N.B. Casting int's -ENODEV to uint16_t results in 0xFFED,
	298	* hence check ret value for < 0. */
	299	if (ret < 0)
	300	domid = DOMID_SELF;
71eef7d1 IC	301
71eef7d1 IC	302	memset(&map_irq, 0, sizeof(map_irq));
beafbdc1	303	map_irq.domid = domid;
71eef7d1 IC	304	map_irq.type = MAP_PIRQ_TYPE_MSI;
	305	map_irq.index = -1;
	306	map_irq.pirq = -1;
	307	map_irq.bus = dev->bus->number;
	308	map_irq.devfn = dev->devfn;
	309
	310	if (type == PCI_CAP_ID_MSIX) {
	311	int pos;
	312	u32 table_offset, bir;
	313
	314	pos = pci_find_capability(dev, PCI_CAP_ID_MSIX);
	315
	316	pci_read_config_dword(dev, pos + PCI_MSIX_TABLE,
	317	&table_offset);
	318	bir = (u8)(table_offset & PCI_MSIX_FLAGS_BIRMASK);
	319
	320	map_irq.table_base = pci_resource_start(dev, bir);
	321	map_irq.entry_nr = msidesc->msi_attrib.entry_nr;
	322	}
	323
	324	ret = HYPERVISOR_physdev_op(PHYSDEVOP_map_pirq, &map_irq);
	325	if (ret) {
beafbdc1 KRW	326	dev_warn(&dev->dev, "xen map irq failed %d for %d domain\n",
beafbdc1 KRW	327	ret, domid);
71eef7d1 IC	328	goto out;
	329	}
	330
	331	ret = xen_bind_pirq_msi_to_irq(dev, msidesc,
	332	map_irq.pirq, map_irq.index,
	333	(type == PCI_CAP_ID_MSIX) ?
beafbdc1 KRW	334	"msi-x" : "msi",
beafbdc1 KRW	335	domid);
71eef7d1 IC	336	if (ret < 0)
71eef7d1 IC	337	goto out;
f731e3ef	338	}
71eef7d1 IC	339	ret = 0;
	340	out:
	341	return ret;
f731e3ef	342	}
b5401a96 AN	343	#endif
b5401a96 AN	344
fef6e262	345	static void xen_teardown_msi_irqs(struct pci_dev *dev)
b5401a96	346	{
fef6e262	347	struct msi_desc *msidesc;
b5401a96	348
fef6e262 KRW	349	msidesc = list_entry(dev->msi_list.next, struct msi_desc, list);
	350	if (msidesc->msi_attrib.is_msix)
	351	xen_pci_frontend_disable_msix(dev);
	352	else
	353	xen_pci_frontend_disable_msi(dev);
b5401a96	354
fef6e262 KRW	355	/* Free the IRQ's and the msidesc using the generic code. */
	356	default_teardown_msi_irqs(dev);
	357	}
f4d0635b	358
fef6e262 KRW	359	static void xen_teardown_msi_irq(unsigned int irq)
	360	{
	361	xen_destroy_irq(irq);
	362	}
b5401a96	363
fef6e262	364	#endif
3f2a230c	365
b5401a96 AN	366	int __init pci_xen_init(void)
	367	{
	368	if (!xen_pv_domain() \|\| xen_initial_domain())
	369	return -ENODEV;
	370
	371	printk(KERN_INFO "PCI: setting up Xen PCI frontend stub\n");
	372
	373	pcibios_set_cache_line_size();
	374
	375	pcibios_enable_irq = xen_pcifront_enable_irq;
	376	pcibios_disable_irq = NULL;
	377
	378	#ifdef CONFIG_ACPI
	379	/* Keep ACPI out of the picture */
	380	acpi_noirq = 1;
	381	#endif
	382
b5401a96 AN	383	#ifdef CONFIG_PCI_MSI
	384	x86_msi.setup_msi_irqs = xen_setup_msi_irqs;
	385	x86_msi.teardown_msi_irq = xen_teardown_msi_irq;
	386	x86_msi.teardown_msi_irqs = xen_teardown_msi_irqs;
	387	#endif
	388	return 0;
	389	}
3942b740 SS	390
	391	int __init pci_xen_hvm_init(void)
	392	{
	393	if (!xen_feature(XENFEAT_hvm_pirqs))
	394	return 0;
	395
	396	#ifdef CONFIG_ACPI
	397	/*
	398	* We don't want to change the actual ACPI delivery model,
	399	* just how GSIs get registered.
	400	*/
	401	__acpi_register_gsi = acpi_register_gsi_xen_hvm;
	402	#endif
809f9267 SS	403
	404	#ifdef CONFIG_PCI_MSI
	405	x86_msi.setup_msi_irqs = xen_hvm_setup_msi_irqs;
	406	x86_msi.teardown_msi_irq = xen_teardown_msi_irq;
	407	#endif
3942b740 SS	408	return 0;
3942b740 SS	409	}
38aa66fc JF	410
38aa66fc JF	411	#ifdef CONFIG_XEN_DOM0
38aa66fc JF	412	static __init void xen_setup_acpi_sci(void)
	413	{
	414	int rc;
	415	int trigger, polarity;
	416	int gsi = acpi_sci_override_gsi;
ee339fe6 KRW	417	int irq = -1;
ee339fe6 KRW	418	int gsi_override = -1;
38aa66fc JF	419
	420	if (!gsi)
	421	return;
	422
	423	rc = acpi_get_override_irq(gsi, &trigger, &polarity);
	424	if (rc) {
	425	printk(KERN_WARNING "xen: acpi_get_override_irq failed for acpi"
	426	" sci, rc=%d\n", rc);
	427	return;
	428	}
	429	trigger = trigger ? ACPI_LEVEL_SENSITIVE : ACPI_EDGE_SENSITIVE;
	430	polarity = polarity ? ACPI_ACTIVE_LOW : ACPI_ACTIVE_HIGH;
996c34ae	431
38aa66fc JF	432	printk(KERN_INFO "xen: sci override: global_irq=%d trigger=%d "
	433	"polarity=%d\n", gsi, trigger, polarity);
	434
ee339fe6 KRW	435	/* Before we bind the GSI to a Linux IRQ, check whether
	436	* we need to override it with bus_irq (IRQ) value. Usually for
	437	* IRQs below IRQ_LEGACY_IRQ this holds IRQ == GSI, as so:
	438	* ACPI: INT_SRC_OVR (bus 0 bus_irq 9 global_irq 9 low level)
	439	* but there are oddballs where the IRQ != GSI:
	440	* ACPI: INT_SRC_OVR (bus 0 bus_irq 9 global_irq 20 low level)
	441	* which ends up being: gsi_to_irq[9] == 20
	442	* (which is what acpi_gsi_to_irq ends up calling when starting the
	443	* the ACPI interpreter and keels over since IRQ 9 has not been
	444	* setup as we had setup IRQ 20 for it).
	445	*/
	446	/* Check whether the GSI != IRQ */
	447	if (acpi_gsi_to_irq(gsi, &irq) == 0) {
	448	if (irq >= 0 && irq != gsi)
	449	/* Bugger, we MUST have that IRQ. */
	450	gsi_override = irq;
	451	}
	452
	453	gsi = xen_register_gsi(gsi, gsi_override, trigger, polarity);
38aa66fc JF	454	printk(KERN_INFO "xen: acpi sci %d\n", gsi);
	455
	456	return;
	457	}
	458
38aa66fc JF	459	static int __init pci_xen_initial_domain(void)
38aa66fc JF	460	{
f731e3ef QH	461	#ifdef CONFIG_PCI_MSI
	462	x86_msi.setup_msi_irqs = xen_initdom_setup_msi_irqs;
	463	x86_msi.teardown_msi_irq = xen_teardown_msi_irq;
	464	#endif
38aa66fc JF	465	xen_setup_acpi_sci();
	466	__acpi_register_gsi = acpi_register_gsi_xen;
	467
	468	return 0;
	469	}
	470
	471	void __init xen_setup_pirqs(void)
	472	{
f4d0635b	473	int pirq, irq;
38aa66fc JF	474
	475	pci_xen_initial_domain();
	476
	477	if (0 == nr_ioapics) {
f4d0635b IC	478	for (irq = 0; irq < NR_IRQS_LEGACY; irq++) {
	479	pirq = xen_allocate_pirq_gsi(irq);
	480	if (WARN(pirq < 0,
	481	"Could not allocate PIRQ for legacy interrupt\n"))
	482	break;
	483	irq = xen_bind_pirq_gsi_to_irq(irq, pirq, 0, "xt-pic");
	484	}
38aa66fc JF	485	return;
	486	}
	487
	488	/* Pre-allocate legacy irqs */
	489	for (irq = 0; irq < NR_IRQS_LEGACY; irq++) {
	490	int trigger, polarity;
	491
	492	if (acpi_get_override_irq(irq, &trigger, &polarity) == -1)
	493	continue;
	494
ee339fe6	495	xen_register_pirq(irq, -1 /* no GSI override */,
38aa66fc JF	496	trigger ? ACPI_LEVEL_SENSITIVE : ACPI_EDGE_SENSITIVE);
	497	}
	498	}
c55fa78b KRW	499
	500	struct xen_device_domain_owner {
	501	domid_t domain;
	502	struct pci_dev *dev;
	503	struct list_head list;
	504	};
	505
	506	static DEFINE_SPINLOCK(dev_domain_list_spinlock);
	507	static struct list_head dev_domain_list = LIST_HEAD_INIT(dev_domain_list);
	508
	509	static struct xen_device_domain_owner find_device(struct pci_dev dev)
	510	{
	511	struct xen_device_domain_owner *owner;
	512
	513	list_for_each_entry(owner, &dev_domain_list, list) {
	514	if (owner->dev == dev)
	515	return owner;
	516	}
	517	return NULL;
	518	}
	519
	520	int xen_find_device_domain_owner(struct pci_dev *dev)
	521	{
	522	struct xen_device_domain_owner *owner;
	523	int domain = -ENODEV;
	524
	525	spin_lock(&dev_domain_list_spinlock);
	526	owner = find_device(dev);
	527	if (owner)
	528	domain = owner->domain;
	529	spin_unlock(&dev_domain_list_spinlock);
	530	return domain;
	531	}
	532	EXPORT_SYMBOL_GPL(xen_find_device_domain_owner);
	533
	534	int xen_register_device_domain_owner(struct pci_dev *dev, uint16_t domain)
	535	{
	536	struct xen_device_domain_owner *owner;
	537
	538	owner = kzalloc(sizeof(struct xen_device_domain_owner), GFP_KERNEL);
	539	if (!owner)
	540	return -ENODEV;
	541
	542	spin_lock(&dev_domain_list_spinlock);
	543	if (find_device(dev)) {
	544	spin_unlock(&dev_domain_list_spinlock);
	545	kfree(owner);
	546	return -EEXIST;
	547	}
	548	owner->domain = domain;
	549	owner->dev = dev;
	550	list_add_tail(&owner->list, &dev_domain_list);
	551	spin_unlock(&dev_domain_list_spinlock);
	552	return 0;
	553	}
	554	EXPORT_SYMBOL_GPL(xen_register_device_domain_owner);
	555
	556	int xen_unregister_device_domain_owner(struct pci_dev *dev)
	557	{
	558	struct xen_device_domain_owner *owner;
	559
	560	spin_lock(&dev_domain_list_spinlock);
	561	owner = find_device(dev);
	562	if (!owner) {
563	spin_unlock(&dev_domain_list_spinlock);
564	return -ENODEV;
565	}
566	list_del(&owner->list);
567	spin_unlock(&dev_domain_list_spinlock);
568	kfree(owner);
569	return 0;
570	}
571	EXPORT_SYMBOL_GPL(xen_unregister_device_domain_owner);
7c1bfd68	572	#endif