13 files changed, 653 insertions, 60 deletions

diff --git a/drivers/xen/Kconfig b/drivers/xen/Kconfig
index 88bca1c42db..c0c490ea7d1 100644
--- a/drivers/xen/Kconfig
+++ b/drivers/xen/Kconfig
@@ -18,6 +18,16 @@ config XEN_SCRUB_PAGES
           secure, but slightly less efficient.
           If in doubt, say yes.
 
+config XEN_DEV_EVTCHN
+        tristate "Xen /dev/xen/evtchn device"
+        depends on XEN
+        default y
+        help
+          The evtchn driver allows a userspace process to triger event
+          channels and to receive notification of an event channel
+          firing.
+          If in doubt, say yes.
+
 config XENFS
         tristate "Xen filesystem"
         depends on XEN
diff --git a/drivers/xen/Makefile b/drivers/xen/Makefile
index f3603a39db5..ec2a39b1e26 100644
--- a/drivers/xen/Makefile
+++ b/drivers/xen/Makefile
@@ -4,5 +4,6 @@ obj-y	+= xenbus/
 obj-$(CONFIG_HOTPLUG_CPU)       += cpu_hotplug.o
 obj-$(CONFIG_XEN_XENCOMM)       += xencomm.o
 obj-$(CONFIG_XEN_BALLOON)       += balloon.o
+obj-$(CONFIG_XEN_DEV_EVTCHN)    += evtchn.o
 obj-$(CONFIG_XENFS)             += xenfs/
-obj-$(CONFIG_XEN_SYS_HYPERVISOR)        += sys-hypervisor.o
+obj-$(CONFIG_XEN_SYS_HYPERVISOR)        += sys-hypervisor.o
\ No newline at end of file
diff --git a/drivers/xen/events.c b/drivers/xen/events.c
index 30963af5dba..1cd2a0e15ae 100644
--- a/drivers/xen/events.c
+++ b/drivers/xen/events.c
@@ -151,6 +151,12 @@ static unsigned int evtchn_from_irq(unsigned irq)
         return info_for_irq(irq)->evtchn;
 }
 
+unsigned irq_from_evtchn(unsigned int evtchn)
+{
+        return evtchn_to_irq[evtchn];
+}
+EXPORT_SYMBOL_GPL(irq_from_evtchn);
+
 static enum ipi_vector ipi_from_irq(unsigned irq)
 {
         struct irq_info *info = info_for_irq(irq);
diff --git a/drivers/xen/evtchn.c b/drivers/xen/evtchn.c
new file mode 100644
index 00000000000..af031950f9b
--- /dev/null
+++ b/drivers/xen/evtchn.c
@@ -0,0 +1,507 @@
+/******************************************************************************
+ * evtchn.c
+ *
+ * Driver for receiving and demuxing event-channel signals.
+ *
+ * Copyright (c) 2004-2005, K A Fraser
+ * Multi-process extensions Copyright (c) 2004, Steven Smith
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2
+ * as published by the Free Software Foundation; or, when distributed
+ * separately from the Linux kernel or incorporated into other
+ * software packages, subject to the following license:
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this source file (the "Software"), to deal in the Software without
+ * restriction, including without limitation the rights to use, copy, modify,
+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
+ * and to permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/slab.h>
+#include <linux/string.h>
+#include <linux/errno.h>
+#include <linux/fs.h>
+#include <linux/errno.h>
+#include <linux/miscdevice.h>
+#include <linux/major.h>
+#include <linux/proc_fs.h>
+#include <linux/stat.h>
+#include <linux/poll.h>
+#include <linux/irq.h>
+#include <linux/init.h>
+#include <linux/gfp.h>
+#include <linux/mutex.h>
+#include <linux/cpu.h>
+#include <xen/events.h>
+#include <xen/evtchn.h>
+#include <asm/xen/hypervisor.h>
+
+struct per_user_data {
+        struct mutex bind_mutex; /* serialize bind/unbind operations */
+
+        /* Notification ring, accessed via /dev/xen/evtchn. */
+#define EVTCHN_RING_SIZE     (PAGE_SIZE / sizeof(evtchn_port_t))
+#define EVTCHN_RING_MASK(_i) ((_i)&(EVTCHN_RING_SIZE-1))
+        evtchn_port_t *ring;
+        unsigned int ring_cons, ring_prod, ring_overflow;
+        struct mutex ring_cons_mutex; /* protect against concurrent readers */
+
+        /* Processes wait on this queue when ring is empty. */
+        wait_queue_head_t evtchn_wait;
+        struct fasync_struct *evtchn_async_queue;
+        const char *name;
+};
+
+/* Who's bound to each port? */
+static struct per_user_data *port_user[NR_EVENT_CHANNELS];
+static DEFINE_SPINLOCK(port_user_lock); /* protects port_user[] and ring_prod */
+
+irqreturn_t evtchn_interrupt(int irq, void *data)
+{
+        unsigned int port = (unsigned long)data;
+        struct per_user_data *u;
+
+        spin_lock(&port_user_lock);
+
+        u = port_user[port];
+
+        disable_irq_nosync(irq);
+
+        if ((u->ring_prod - u->ring_cons) < EVTCHN_RING_SIZE) {
+                u->ring[EVTCHN_RING_MASK(u->ring_prod)] = port;
+                wmb(); /* Ensure ring contents visible */
+                if (u->ring_cons == u->ring_prod++) {
+                        wake_up_interruptible(&u->evtchn_wait);
+                        kill_fasync(&u->evtchn_async_queue,
+                                    SIGIO, POLL_IN);
+                }
+        } else {
+                u->ring_overflow = 1;
+        }
+
+        spin_unlock(&port_user_lock);
+
+        return IRQ_HANDLED;
+}
+
+static ssize_t evtchn_read(struct file *file, char __user *buf,
+                           size_t count, loff_t *ppos)
+{
+        int rc;
+        unsigned int c, p, bytes1 = 0, bytes2 = 0;
+        struct per_user_data *u = file->private_data;
+
+        /* Whole number of ports. */
+        count &= ~(sizeof(evtchn_port_t)-1);
+
+        if (count == 0)
+                return 0;
+
+        if (count > PAGE_SIZE)
+                count = PAGE_SIZE;
+
+        for (;;) {
+                mutex_lock(&u->ring_cons_mutex);
+
+                rc = -EFBIG;
+                if (u->ring_overflow)
+                        goto unlock_out;
+
+                c = u->ring_cons;
+                p = u->ring_prod;
+                if (c != p)
+                        break;
+
+                mutex_unlock(&u->ring_cons_mutex);
+
+                if (file->f_flags & O_NONBLOCK)
+                        return -EAGAIN;
+
+                rc = wait_event_interruptible(u->evtchn_wait,
+                                              u->ring_cons != u->ring_prod);
+                if (rc)
+                        return rc;
+        }
+
+        /* Byte lengths of two chunks. Chunk split (if any) is at ring wrap. */
+        if (((c ^ p) & EVTCHN_RING_SIZE) != 0) {
+                bytes1 = (EVTCHN_RING_SIZE - EVTCHN_RING_MASK(c)) *
+                        sizeof(evtchn_port_t);
+                bytes2 = EVTCHN_RING_MASK(p) * sizeof(evtchn_port_t);
+        } else {
+                bytes1 = (p - c) * sizeof(evtchn_port_t);
+                bytes2 = 0;
+        }
+
+        /* Truncate chunks according to caller's maximum byte count. */
+        if (bytes1 > count) {
+                bytes1 = count;
+                bytes2 = 0;
+        } else if ((bytes1 + bytes2) > count) {
+                bytes2 = count - bytes1;
+        }
+
+        rc = -EFAULT;
+        rmb(); /* Ensure that we see the port before we copy it. */
+        if (copy_to_user(buf, &u->ring[EVTCHN_RING_MASK(c)], bytes1) ||
+            ((bytes2 != 0) &&
+             copy_to_user(&buf[bytes1], &u->ring[0], bytes2)))
+                goto unlock_out;
+
+        u->ring_cons += (bytes1 + bytes2) / sizeof(evtchn_port_t);
+        rc = bytes1 + bytes2;
+
+ unlock_out:
+        mutex_unlock(&u->ring_cons_mutex);
+        return rc;
+}
+
+static ssize_t evtchn_write(struct file *file, const char __user *buf,
+                            size_t count, loff_t *ppos)
+{
+        int rc, i;
+        evtchn_port_t *kbuf = (evtchn_port_t *)__get_free_page(GFP_KERNEL);
+        struct per_user_data *u = file->private_data;
+
+        if (kbuf == NULL)
+                return -ENOMEM;
+
+        /* Whole number of ports. */
+        count &= ~(sizeof(evtchn_port_t)-1);
+
+        rc = 0;
+        if (count == 0)
+                goto out;
+
+        if (count > PAGE_SIZE)
+                count = PAGE_SIZE;
+
+        rc = -EFAULT;
+        if (copy_from_user(kbuf, buf, count) != 0)
+                goto out;
+
+        spin_lock_irq(&port_user_lock);
+        for (i = 0; i < (count/sizeof(evtchn_port_t)); i++)
+                if ((kbuf[i] < NR_EVENT_CHANNELS) && (port_user[kbuf[i]] == u))
+                        enable_irq(irq_from_evtchn(kbuf[i]));
+        spin_unlock_irq(&port_user_lock);
+
+        rc = count;
+
+ out:
+        free_page((unsigned long)kbuf);
+        return rc;
+}
+
+static int evtchn_bind_to_user(struct per_user_data *u, int port)
+{
+        int rc = 0;
+
+        /*
+         * Ports are never reused, so every caller should pass in a
+         * unique port.
+         *
+         * (Locking not necessary because we haven't registered the
+         * interrupt handler yet, and our caller has already
+         * serialized bind operations.)
+         */
+        BUG_ON(port_user[port] != NULL);
+        port_user[port] = u;
+
+        rc = bind_evtchn_to_irqhandler(port, evtchn_interrupt, IRQF_DISABLED,
+                                       u->name, (void *)(unsigned long)port);
+        if (rc >= 0)
+                rc = 0;
+
+        return rc;
+}
+
+static void evtchn_unbind_from_user(struct per_user_data *u, int port)
+{
+        int irq = irq_from_evtchn(port);
+
+        unbind_from_irqhandler(irq, (void *)(unsigned long)port);
+
+        /* make sure we unbind the irq handler before clearing the port */
+        barrier();
+
+        port_user[port] = NULL;
+}
+
+static long evtchn_ioctl(struct file *file,
+                         unsigned int cmd, unsigned long arg)
+{
+        int rc;
+        struct per_user_data *u = file->private_data;
+        void __user *uarg = (void __user *) arg;
+
+        /* Prevent bind from racing with unbind */
+        mutex_lock(&u->bind_mutex);
+
+        switch (cmd) {
+        case IOCTL_EVTCHN_BIND_VIRQ: {
+                struct ioctl_evtchn_bind_virq bind;
+                struct evtchn_bind_virq bind_virq;
+
+                rc = -EFAULT;
+                if (copy_from_user(&bind, uarg, sizeof(bind)))
+                        break;
+
+                bind_virq.virq = bind.virq;
+                bind_virq.vcpu = 0;
+                rc = HYPERVISOR_event_channel_op(EVTCHNOP_bind_virq,
+                                                 &bind_virq);
+                if (rc != 0)
+                        break;
+
+                rc = evtchn_bind_to_user(u, bind_virq.port);
+                if (rc == 0)
+                        rc = bind_virq.port;
+                break;
+        }
+
+        case IOCTL_EVTCHN_BIND_INTERDOMAIN: {
+                struct ioctl_evtchn_bind_interdomain bind;
+                struct evtchn_bind_interdomain bind_interdomain;
+
+                rc = -EFAULT;
+                if (copy_from_user(&bind, uarg, sizeof(bind)))
+                        break;
+
+                bind_interdomain.remote_dom  = bind.remote_domain;
+                bind_interdomain.remote_port = bind.remote_port;
+                rc = HYPERVISOR_event_channel_op(EVTCHNOP_bind_interdomain,
+                                                 &bind_interdomain);
+                if (rc != 0)
+                        break;
+
+                rc = evtchn_bind_to_user(u, bind_interdomain.local_port);
+                if (rc == 0)
+                        rc = bind_interdomain.local_port;
+                break;
+        }
+
+        case IOCTL_EVTCHN_BIND_UNBOUND_PORT: {
+                struct ioctl_evtchn_bind_unbound_port bind;
+                struct evtchn_alloc_unbound alloc_unbound;
+
+                rc = -EFAULT;
+                if (copy_from_user(&bind, uarg, sizeof(bind)))
+                        break;
+
+                alloc_unbound.dom        = DOMID_SELF;
+                alloc_unbound.remote_dom = bind.remote_domain;
+                rc = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound,
+                                                 &alloc_unbound);
+                if (rc != 0)
+                        break;
+
+                rc = evtchn_bind_to_user(u, alloc_unbound.port);
+                if (rc == 0)
+                        rc = alloc_unbound.port;
+                break;
+        }
+
+        case IOCTL_EVTCHN_UNBIND: {
+                struct ioctl_evtchn_unbind unbind;
+
+                rc = -EFAULT;
+                if (copy_from_user(&unbind, uarg, sizeof(unbind)))
+                        break;
+
+                rc = -EINVAL;
+                if (unbind.port >= NR_EVENT_CHANNELS)
+                        break;
+
+                spin_lock_irq(&port_user_lock);
+
+                rc = -ENOTCONN;
+                if (port_user[unbind.port] != u) {
+                        spin_unlock_irq(&port_user_lock);
+                        break;
+                }
+
+                evtchn_unbind_from_user(u, unbind.port);
+
+                spin_unlock_irq(&port_user_lock);
+
+                rc = 0;
+                break;
+        }
+
+        case IOCTL_EVTCHN_NOTIFY: {
+                struct ioctl_evtchn_notify notify;
+
+                rc = -EFAULT;
+                if (copy_from_user(&notify, uarg, sizeof(notify)))
+                        break;
+
+                if (notify.port >= NR_EVENT_CHANNELS) {
+                        rc = -EINVAL;
+                } else if (port_user[notify.port] != u) {
+                        rc = -ENOTCONN;
+                } else {
+                        notify_remote_via_evtchn(notify.port);
+                        rc = 0;
+                }
+                break;
+        }
+
+        case IOCTL_EVTCHN_RESET: {
+                /* Initialise the ring to empty. Clear errors. */
+                mutex_lock(&u->ring_cons_mutex);
+                spin_lock_irq(&port_user_lock);
+                u->ring_cons = u->ring_prod = u->ring_overflow = 0;
+                spin_unlock_irq(&port_user_lock);
+                mutex_unlock(&u->ring_cons_mutex);
+                rc = 0;
+                break;
+        }
+
+        default:
+                rc = -ENOSYS;
+                break;
+        }
+        mutex_unlock(&u->bind_mutex);
+
+        return rc;
+}
+
+static unsigned int evtchn_poll(struct file *file, poll_table *wait)
+{
+        unsigned int mask = POLLOUT | POLLWRNORM;
+        struct per_user_data *u = file->private_data;
+
+        poll_wait(file, &u->evtchn_wait, wait);
+        if (u->ring_cons != u->ring_prod)
+                mask |= POLLIN | POLLRDNORM;
+        if (u->ring_overflow)
+                mask = POLLERR;
+        return mask;
+}
+
+static int evtchn_fasync(int fd, struct file *filp, int on)
+{
+        struct per_user_data *u = filp->private_data;
+        return fasync_helper(fd, filp, on, &u->evtchn_async_queue);
+}
+
+static int evtchn_open(struct inode *inode, struct file *filp)
+{
+        struct per_user_data *u;
+
+        u = kzalloc(sizeof(*u), GFP_KERNEL);
+        if (u == NULL)
+                return -ENOMEM;
+
+        u->name = kasprintf(GFP_KERNEL, "evtchn:%s", current->comm);
+        if (u->name == NULL) {
+                kfree(u);
+                return -ENOMEM;
+        }
+
+        init_waitqueue_head(&u->evtchn_wait);
+
+        u->ring = (evtchn_port_t *)__get_free_page(GFP_KERNEL);
+        if (u->ring == NULL) {
+                kfree(u->name);
+                kfree(u);
+                return -ENOMEM;
+        }
+
+        mutex_init(&u->bind_mutex);
+        mutex_init(&u->ring_cons_mutex);
+
+        filp->private_data = u;
+
+        return 0;
+}
+
+static int evtchn_release(struct inode *inode, struct file *filp)
+{
+        int i;
+        struct per_user_data *u = filp->private_data;
+
+        spin_lock_irq(&port_user_lock);
+
+        free_page((unsigned long)u->ring);
+
+        for (i = 0; i < NR_EVENT_CHANNELS; i++) {
+                if (port_user[i] != u)
+                        continue;
+
+                evtchn_unbind_from_user(port_user[i], i);
+        }
+
+        spin_unlock_irq(&port_user_lock);
+
+        kfree(u->name);
+        kfree(u);
+
+        return 0;
+}
+
+static const struct file_operations evtchn_fops = {
+        .owner   = THIS_MODULE,
+        .read    = evtchn_read,
+        .write   = evtchn_write,
+        .unlocked_ioctl = evtchn_ioctl,
+        .poll    = evtchn_poll,
+        .fasync  = evtchn_fasync,
+        .open    = evtchn_open,
+        .release = evtchn_release,
+};
+
+static struct miscdevice evtchn_miscdev = {
+        .minor        = MISC_DYNAMIC_MINOR,
+        .name         = "evtchn",
+        .fops         = &evtchn_fops,
+};
+static int __init evtchn_init(void)
+{
+        int err;
+
+        if (!xen_domain())
+                return -ENODEV;
+
+        spin_lock_init(&port_user_lock);
+        memset(port_user, 0, sizeof(port_user));
+
+        /* Create '/dev/misc/evtchn'. */
+        err = misc_register(&evtchn_miscdev);
+        if (err != 0) {
+                printk(KERN_ALERT "Could not register /dev/misc/evtchn\n");
+                return err;
+        }
+
+        printk(KERN_INFO "Event-channel device installed.\n");
+
+        return 0;
+}
+
+static void __exit evtchn_cleanup(void)
+{
+        misc_deregister(&evtchn_miscdev);
+}
+
+module_init(evtchn_init);
+module_exit(evtchn_cleanup);
+
+MODULE_LICENSE("GPL");
diff --git a/drivers/xen/manage.c b/drivers/xen/manage.c
index b703dd2c9f1..5269bb4d249 100644
--- a/drivers/xen/manage.c
+++ b/drivers/xen/manage.c
@@ -104,9 +104,8 @@ static void do_suspend(void)
                 goto out;
         }
 
-        printk("suspending xenbus...\n");
-        /* XXX use normal device tree? */
-        xenbus_suspend();
+        printk(KERN_DEBUG "suspending xenstore...\n");
+        xs_suspend();
 
         err = stop_machine(xen_suspend, &cancelled, cpumask_of(0));
         if (err) {
@@ -116,9 +115,9 @@ static void do_suspend(void)
 
         if (!cancelled) {
                 xen_arch_resume();
-                xenbus_resume();
+                xs_resume();
         } else
-                xenbus_suspend_cancel();
+                xs_suspend_cancel();
 
         device_resume(PMSG_RESUME);
 
diff --git a/drivers/xen/xenbus/xenbus_probe.c b/drivers/xen/xenbus/xenbus_probe.c
index 773d1cf2328..d42e25d5968 100644
--- a/drivers/xen/xenbus/xenbus_probe.c
+++ b/drivers/xen/xenbus/xenbus_probe.c
@@ -71,6 +71,9 @@ static int xenbus_probe_frontend(const char *type, const char *name);
 
 static void xenbus_dev_shutdown(struct device *_dev);
 
+static int xenbus_dev_suspend(struct device *dev, pm_message_t state);
+static int xenbus_dev_resume(struct device *dev);
+
 /* If something in array of ids matches this device, return it. */
 static const struct xenbus_device_id *
 match_device(const struct xenbus_device_id *arr, struct xenbus_device *dev)
@@ -188,6 +191,9 @@ static struct xen_bus_type xenbus_frontend = {
                 .remove    = xenbus_dev_remove,
                 .shutdown  = xenbus_dev_shutdown,
                 .dev_attrs = xenbus_dev_attrs,
+
+                .suspend   = xenbus_dev_suspend,
+                .resume    = xenbus_dev_resume,
         },
 };
 
@@ -654,6 +660,7 @@ void xenbus_dev_changed(const char *node, struct xen_bus_type *bus)
 
         kfree(root);
 }
+EXPORT_SYMBOL_GPL(xenbus_dev_changed);
 
 static void frontend_changed(struct xenbus_watch *watch,
                              const char **vec, unsigned int len)
@@ -669,7 +676,7 @@ static struct xenbus_watch fe_watch = {
         .callback = frontend_changed,
 };
 
-static int suspend_dev(struct device *dev, void *data)
+static int xenbus_dev_suspend(struct device *dev, pm_message_t state)
 {
         int err = 0;
         struct xenbus_driver *drv;
@@ -682,35 +689,14 @@ static int suspend_dev(struct device *dev, void *data)
         drv = to_xenbus_driver(dev->driver);
         xdev = container_of(dev, struct xenbus_device, dev);
         if (drv->suspend)
-                err = drv->suspend(xdev);
+                err = drv->suspend(xdev, state);
         if (err)
                 printk(KERN_WARNING
                        "xenbus: suspend %s failed: %i\n", dev_name(dev), err);
         return 0;
 }
 
-static int suspend_cancel_dev(struct device *dev, void *data)
-{
-        int err = 0;
-        struct xenbus_driver *drv;
-        struct xenbus_device *xdev;
-
-        DPRINTK("");
-
-        if (dev->driver == NULL)
-                return 0;
-        drv = to_xenbus_driver(dev->driver);
-        xdev = container_of(dev, struct xenbus_device, dev);
-        if (drv->suspend_cancel)
-                err = drv->suspend_cancel(xdev);
-        if (err)
-                printk(KERN_WARNING
-                       "xenbus: suspend_cancel %s failed: %i\n",
-                       dev_name(dev), err);
-        return 0;
-}
-
-static int resume_dev(struct device *dev, void *data)
+static int xenbus_dev_resume(struct device *dev)
 {
         int err;
         struct xenbus_driver *drv;
@@ -755,33 +741,6 @@ static int resume_dev(struct device *dev, void *data)
         return 0;
 }
 
-void xenbus_suspend(void)
-{
-        DPRINTK("");
-
-        bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, suspend_dev);
-        xenbus_backend_suspend(suspend_dev);
-        xs_suspend();
-}
-EXPORT_SYMBOL_GPL(xenbus_suspend);
-
-void xenbus_resume(void)
-{
-        xb_init_comms();
-        xs_resume();
-        bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, resume_dev);
-        xenbus_backend_resume(resume_dev);
-}
-EXPORT_SYMBOL_GPL(xenbus_resume);
-
-void xenbus_suspend_cancel(void)
-{
-        xs_suspend_cancel();
-        bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, suspend_cancel_dev);
-        xenbus_backend_resume(suspend_cancel_dev);
-}
-EXPORT_SYMBOL_GPL(xenbus_suspend_cancel);
-
 /* A flag to determine if xenstored is 'ready' (i.e. has started) */
 int xenstored_ready = 0;
 
diff --git a/drivers/xen/xenbus/xenbus_xs.c b/drivers/xen/xenbus/xenbus_xs.c
index e325eab4724..eab33f1dbdf 100644
--- a/drivers/xen/xenbus/xenbus_xs.c
+++ b/drivers/xen/xenbus/xenbus_xs.c
@@ -673,6 +673,8 @@ void xs_resume(void)
         struct xenbus_watch *watch;
         char token[sizeof(watch) * 2 + 1];
 
+        xb_init_comms();
+
         mutex_unlock(&xs_state.response_mutex);
         mutex_unlock(&xs_state.request_mutex);
         up_write(&xs_state.transaction_mutex);
diff --git a/drivers/xen/xenfs/super.c b/drivers/xen/xenfs/super.c
index 515741a8e6b..6559e0c752c 100644
--- a/drivers/xen/xenfs/super.c
+++ b/drivers/xen/xenfs/super.c
@@ -20,10 +20,27 @@
 MODULE_DESCRIPTION("Xen filesystem");
 MODULE_LICENSE("GPL");
 
+static ssize_t capabilities_read(struct file *file, char __user *buf,
+                                 size_t size, loff_t *off)
+{
+        char *tmp = "";
+
+        if (xen_initial_domain())
+                tmp = "control_d\n";
+
+        return simple_read_from_buffer(buf, size, off, tmp, strlen(tmp));
+}
+
+static const struct file_operations capabilities_file_ops = {
+        .read = capabilities_read,
+};
+
 static int xenfs_fill_super(struct super_block *sb, void *data, int silent)
 {
         static struct tree_descr xenfs_files[] = {
-                [2] = {"xenbus", &xenbus_file_ops, S_IRUSR|S_IWUSR},
+                [1] = {},
+                { "xenbus", &xenbus_file_ops, S_IRUSR|S_IWUSR },
+                { "capabilities", &capabilities_file_ops, S_IRUGO },
                 {""},
         };
 
diff --git a/include/Kbuild b/include/Kbuild
index d8c3e3cbf41..fe36accd432 100644
--- a/include/Kbuild
+++ b/include/Kbuild
@@ -8,3 +8,4 @@ header-y += mtd/
 header-y += rdma/
 header-y += video/
 header-y += drm/
+header-y += xen/
diff --git a/include/xen/Kbuild b/include/xen/Kbuild
new file mode 100644
index 00000000000..4e65c16a445
--- /dev/null
+++ b/include/xen/Kbuild
@@ -0,0 +1 @@
+header-y += evtchn.h
diff --git a/include/xen/events.h b/include/xen/events.h
index 0d5f1adc036..e68d59a90ca 100644
--- a/include/xen/events.h
+++ b/include/xen/events.h
@@ -53,4 +53,7 @@ bool xen_test_irq_pending(int irq);
    irq will be disabled so it won't deliver an interrupt. */
 void xen_poll_irq(int irq);
 
+/* Determine the IRQ which is bound to an event channel */
+unsigned irq_from_evtchn(unsigned int evtchn);
+
 #endif  /* _XEN_EVENTS_H */
diff --git a/include/xen/evtchn.h b/include/xen/evtchn.h
new file mode 100644
index 00000000000..14e833ee4e0
--- /dev/null
+++ b/include/xen/evtchn.h
@@ -0,0 +1,88 @@
+/******************************************************************************
+ * evtchn.h
+ *
+ * Interface to /dev/xen/evtchn.
+ *
+ * Copyright (c) 2003-2005, K A Fraser
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2
+ * as published by the Free Software Foundation; or, when distributed
+ * separately from the Linux kernel or incorporated into other
+ * software packages, subject to the following license:
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this source file (the "Software"), to deal in the Software without
+ * restriction, including without limitation the rights to use, copy, modify,
+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
+ * and to permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#ifndef __LINUX_PUBLIC_EVTCHN_H__
+#define __LINUX_PUBLIC_EVTCHN_H__
+
+/*
+ * Bind a fresh port to VIRQ @virq.
+ * Return allocated port.
+ */
+#define IOCTL_EVTCHN_BIND_VIRQ                          \
+        _IOC(_IOC_NONE, 'E', 0, sizeof(struct ioctl_evtchn_bind_virq))
+struct ioctl_evtchn_bind_virq {
+        unsigned int virq;
+};
+
+/*
+ * Bind a fresh port to remote <@remote_domain, @remote_port>.
+ * Return allocated port.
+ */
+#define IOCTL_EVTCHN_BIND_INTERDOMAIN                   \
+        _IOC(_IOC_NONE, 'E', 1, sizeof(struct ioctl_evtchn_bind_interdomain))
+struct ioctl_evtchn_bind_interdomain {
+        unsigned int remote_domain, remote_port;
+};
+
+/*
+ * Allocate a fresh port for binding to @remote_domain.
+ * Return allocated port.
+ */
+#define IOCTL_EVTCHN_BIND_UNBOUND_PORT                  \
+        _IOC(_IOC_NONE, 'E', 2, sizeof(struct ioctl_evtchn_bind_unbound_port))
+struct ioctl_evtchn_bind_unbound_port {
+        unsigned int remote_domain;
+};
+
+/*
+ * Unbind previously allocated @port.
+ */
+#define IOCTL_EVTCHN_UNBIND                             \
+        _IOC(_IOC_NONE, 'E', 3, sizeof(struct ioctl_evtchn_unbind))
+struct ioctl_evtchn_unbind {
+        unsigned int port;
+};
+
+/*
+ * Unbind previously allocated @port.
+ */
+#define IOCTL_EVTCHN_NOTIFY                             \
+        _IOC(_IOC_NONE, 'E', 4, sizeof(struct ioctl_evtchn_notify))
+struct ioctl_evtchn_notify {
+        unsigned int port;
+};
+
+/* Clear and reinitialise the event buffer. Clear error condition. */
+#define IOCTL_EVTCHN_RESET                              \
+        _IOC(_IOC_NONE, 'E', 5, 0)
+
+#endif /* __LINUX_PUBLIC_EVTCHN_H__ */
diff --git a/include/xen/xenbus.h b/include/xen/xenbus.h
index f87f9614844..b9763badbd7 100644
--- a/include/xen/xenbus.h
+++ b/include/xen/xenbus.h
@@ -91,8 +91,7 @@ struct xenbus_driver {
         void (*otherend_changed)(struct xenbus_device *dev,
                                  enum xenbus_state backend_state);
         int (*remove)(struct xenbus_device *dev);
-        int (*suspend)(struct xenbus_device *dev);
-        int (*suspend_cancel)(struct xenbus_device *dev);
+        int (*suspend)(struct xenbus_device *dev, pm_message_t state);
         int (*resume)(struct xenbus_device *dev);
         int (*uevent)(struct xenbus_device *, char **, int, char *, int);
         struct device_driver driver;