[PATCH] staging: Android: Add 'vsoc' driver for cuttlefish.

[Alistair Strachan]

From: Greg Hartman <ghartman@xxxxxxxxxx>

The cuttlefish system is a virtual SoC architecture based on QEMU. It
uses the QEMU ivshmem feature to share memory regions between guest and
host with a custom protocol.

Cc: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
Cc: Arve HjÃnnevÃg <arve@xxxxxxxxxxx>
Cc: Todd Kjos <tkjos@xxxxxxxxxxx>
Cc: Martijn Coenen <maco@xxxxxxxxxxx>
Cc: devel@xxxxxxxxxxxxxxxxxxxx
Cc: kernel-team@xxxxxxxxxxx
Signed-off-by: Greg Hartman <ghartman@xxxxxxxxxx>
[astrachan: rebased against 4.16, added TODO, fixed checkpatch issues]
Signed-off-by: Alistair Strachan <astrachan@xxxxxxxxxx>
---
 drivers/staging/android/Kconfig         |    9 +
 drivers/staging/android/Makefile        |    1 +
 drivers/staging/android/TODO            |   10 +
 drivers/staging/android/uapi/vsoc_shm.h |  303 ++++++
 drivers/staging/android/vsoc.c          | 1167 +++++++++++++++++++++++
 5 files changed, 1490 insertions(+)
 create mode 100644 drivers/staging/android/uapi/vsoc_shm.h
 create mode 100644 drivers/staging/android/vsoc.c

diff --git a/drivers/staging/android/Kconfig b/drivers/staging/android/Kconfig
index 71a50b99caff..29d891355f7a 100644
--- a/drivers/staging/android/Kconfig
+++ b/drivers/staging/android/Kconfig
@@ -14,6 +14,15 @@ config ASHMEM
 	  It is, in theory, a good memory allocator for low-memory devices,
 	  because it can discard shared memory units when under memory pressure.
 
+config ANDROID_VSOC
+	tristate "Android Virtual SoC support"
+	default n
+	depends on PCI_MSI
+	---help---
+	  This option adds support for the Virtual SoC driver needed to boot
+	  a 'cuttlefish' Android image inside QEmu. The driver interacts with
+	  a QEmu ivshmem device. If built as a module, it will be called vsoc.
+
 source "drivers/staging/android/ion/Kconfig"
 
 endif # if ANDROID
diff --git a/drivers/staging/android/Makefile b/drivers/staging/android/Makefile
index 7cf1564a49a5..90e6154f11a4 100644
--- a/drivers/staging/android/Makefile
+++ b/drivers/staging/android/Makefile
@@ -3,3 +3,4 @@ ccflags-y += -I$(src)			# needed for trace events
 obj-y					+= ion/
 
 obj-$(CONFIG_ASHMEM)			+= ashmem.o
+obj-$(CONFIG_ANDROID_VSOC)		+= vsoc.o
diff --git a/drivers/staging/android/TODO b/drivers/staging/android/TODO
index 687e0eac85bf..6aab759efa48 100644
--- a/drivers/staging/android/TODO
+++ b/drivers/staging/android/TODO
@@ -11,5 +11,15 @@ ion/
  - Split /dev/ion up into multiple nodes (e.g. /dev/ion/heap0)
  - Better test framework (integration with VGEM was suggested)
 
+vsoc.c, uapi/vsoc_shm.h
+ - The current driver uses the same wait queue for all of the futexes in a
+   region. This will cause false wakeups in regions with a large number of
+   waiting threads. We should eventually use multiple queues and select the
+   queue based on the region.
+ - Add debugfs support for examining the permissions of regions.
+ - Use ioremap_wc instead of ioremap_nocache.
+ - Remove VSOC_WAIT_FOR_INCOMING_INTERRUPT ioctl. This functionality has been
+   superseded by the futex and is there for legacy reasons.
+
 Please send patches to Greg Kroah-Hartman <greg@xxxxxxxxx> and Cc:
 Arve HjÃnnevÃg <arve@xxxxxxxxxxx> and Riley Andrews <riandrews@xxxxxxxxxxx>
diff --git a/drivers/staging/android/uapi/vsoc_shm.h b/drivers/staging/android/uapi/vsoc_shm.h
new file mode 100644
index 000000000000..741b1387c25b
--- /dev/null
+++ b/drivers/staging/android/uapi/vsoc_shm.h
@@ -0,0 +1,303 @@
+/*
+ * Copyright (C) 2017 Google, Inc.
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ */
+
+#ifndef _UAPI_LINUX_VSOC_SHM_H
+#define _UAPI_LINUX_VSOC_SHM_H
+
+#include <linux/types.h>
+
+/**
+ * A permission is a token that permits a receiver to read and/or write an area
+ * of memory within a Vsoc region.
+ *
+ * An fd_scoped permission grants both read and write access, and can be
+ * attached to a file description (see open(2)).
+ * Ownership of the area can then be shared by passing a file descriptor
+ * among processes.
+ *
+ * begin_offset and end_offset define the area of memory that is controlled by
+ * the permission. owner_offset points to a word, also in shared memory, that
+ * controls ownership of the area.
+ *
+ * ownership of the region expires when the associated file description is
+ * released.
+ *
+ * At most one permission can be attached to each file description.
+ *
+ * This is useful when implementing HALs like gralloc that scope and pass
+ * ownership of shared resources via file descriptors.
+ *
+ * The caller is responsibe for doing any fencing.
+ *
+ * The calling process will normally identify a currently free area of
+ * memory. It will construct a proposed fd_scoped_permission_arg structure:
+ *
+ *   begin_offset and end_offset describe the area being claimed
+ *
+ *   owner_offset points to the location in shared memory that indicates the
+ *   owner of the area.
+ *
+ *   owned_value is the value that will be stored in owner_offset iff the
+ *   permission can be granted. It must be different than VSOC_REGION_FREE.
+ *
+ * Two fd_scoped_permission structures are compatible if they vary only by
+ * their owned_value fields.
+ *
+ * The driver ensures that, for any group of simultaneous callers proposing
+ * compatible fd_scoped_permissions, it will accept exactly one of the
+ * propopsals. The other callers will get a failure with errno of EAGAIN.
+ *
+ * A process receiving a file descriptor can identify the region being
+ * granted using the VSOC_GET_FD_SCOPED_PERMISSION ioctl.
+ */
+struct fd_scoped_permission {
+	__u32 begin_offset;
+	__u32 end_offset;
+	__u32 owner_offset;
+	__u32 owned_value;
+};
+
+/*
+ * This value represents a free area of memory. The driver expects to see this
+ * value at owner_offset when creating a permission otherwise it will not do it,
+ * and will write this value back once the permission is no longer needed.
+ */
+#define VSOC_REGION_FREE ((__u32)0)
+
+/**
+ * ioctl argument for VSOC_CREATE_FD_SCOPE_PERMISSION
+ */
+struct fd_scoped_permission_arg {
+	struct fd_scoped_permission perm;
+	__s32 managed_region_fd;
+};
+
+#define VSOC_NODE_FREE ((__u32)0)
+
+/*
+ * Describes a signal table in shared memory. Each non-zero entry in the
+ * table indicates that the receiver should signal the futex at the given
+ * offset. Offsets are relative to the region, not the shared memory window.
+ *
+ * interrupt_signalled_offset is used to reliably signal interrupts across the
+ * vmm boundary. There are two roles: transmitter and receiver. For example,
+ * in the host_to_guest_signal_table the host is the transmitter and the
+ * guest is the receiver. The protocol is as follows:
+ *
+ * 1. The transmitter should convert the offset of the futex to an offset
+ *    in the signal table [0, (1 << num_nodes_lg2))
+ *    The transmitter can choose any appropriate hashing algorithm, including
+ *    hash = futex_offset & ((1 << num_nodes_lg2) - 1)
+ *
+ * 3. The transmitter should atomically compare and swap futex_offset with 0
+ *    at hash. There are 3 possible outcomes
+ *      a. The swap fails because the futex_offset is already in the table.
+ *         The transmitter should stop.
+ *      b. Some other offset is in the table. This is a hash collision. The
+ *         transmitter should move to another table slot and try again. One
+ *         possible algorithm:
+ *         hash = (hash + 1) & ((1 << num_nodes_lg2) - 1)
+ *      c. The swap worked. Continue below.
+ *
+ * 3. The transmitter atomically swaps 1 with the value at the
+ *    interrupt_signalled_offset. There are two outcomes:
+ *      a. The prior value was 1. In this case an interrupt has already been
+ *         posted. The transmitter is done.
+ *      b. The prior value was 0, indicating that the receiver may be sleeping.
+ *         The transmitter will issue an interrupt.
+ *
+ * 4. On waking the receiver immediately exchanges a 0 with the
+ *    interrupt_signalled_offset. If it receives a 0 then this a spurious
+ *    interrupt. That may occasionally happen in the current protocol, but
+ *    should be rare.
+ *
+ * 5. The receiver scans the signal table by atomicaly exchanging 0 at each
+ *    location. If a non-zero offset is returned from the exchange the
+ *    receiver wakes all sleepers at the given offset:
+ *      futex((int*)(region_base + old_value), FUTEX_WAKE, MAX_INT);
+ *
+ * 6. The receiver thread then does a conditional wait, waking immediately
+ *    if the value at interrupt_signalled_offset is non-zero. This catches cases
+ *    here additional  signals were posted while the table was being scanned.
+ *    On the guest the wait is handled via the VSOC_WAIT_FOR_INCOMING_INTERRUPT
+ *    ioctl.
+ */
+struct vsoc_signal_table_layout {
+	/* log_2(Number of signal table entries) */
+	__u32 num_nodes_lg2;
+	/*
+	 * Offset to the first signal table entry relative to the start of the
+	 * region
+	 */
+	__u32 futex_uaddr_table_offset;
+	/*
+	 * Offset to an atomic_t / atomic uint32_t. A non-zero value indicates
+	 * that one or more offsets are currently posted in the table.
+	 * semi-unique access to an entry in the table
+	 */
+	__u32 interrupt_signalled_offset;
+};
+
+#define VSOC_REGION_WHOLE ((__s32)0)
+#define VSOC_DEVICE_NAME_SZ 16
+
+/**
+ * Each HAL would (usually) talk to a single device region
+ * Mulitple entities care about these regions:
+ * - The ivshmem_server will populate the regions in shared memory
+ * - The guest kernel will read the region, create minor device nodes, and
+ *   allow interested parties to register for FUTEX_WAKE events in the region
+ * - HALs will access via the minor device nodes published by the guest kernel
+ * - Host side processes will access the region via the ivshmem_server:
+ *   1. Pass name to ivshmem_server at a UNIX socket
+ *   2. ivshmemserver will reply with 2 fds:
+ *     - host->guest doorbell fd
+ *     - guest->host doorbell fd
+ *     - fd for the shared memory region
+ *     - region offset
+ *   3. Start a futex receiver thread on the doorbell fd pointed at the
+ *      signal_nodes
+ */
+struct vsoc_device_region {
+	__u16 current_version;
+	__u16 min_compatible_version;
+	__u32 region_begin_offset;
+	__u32 region_end_offset;
+	__u32 offset_of_region_data;
+	struct vsoc_signal_table_layout guest_to_host_signal_table;
+	struct vsoc_signal_table_layout host_to_guest_signal_table;
+	/* Name of the device. Must always be terminated with a '\0', so
+	 * the longest supported device name is 15 characters.
+	 */
+	char device_name[VSOC_DEVICE_NAME_SZ];
+	/* There are two ways that permissions to access regions are handled:
+	 *   - When subdivided_by is VSOC_REGION_WHOLE, any process that can
+	 *     open the device node for the region gains complete access to it.
+	 *   - When subdivided is set processes that open the region cannot
+	 *     access it. Access to a sub-region must be established by invoking
+	 *     the VSOC_CREATE_FD_SCOPE_PERMISSION ioctl on the region
+	 *     referenced in subdivided_by, providing a fileinstance
+	 *     (represented by a fd) opened on this region.
+	 */
+	__u32 managed_by;
+};
+
+/*
+ * The vsoc layout descriptor.
+ * The first 4K should be reserved for the shm header and region descriptors.
+ * The regions should be page aligned.
+ */
+
+struct vsoc_shm_layout_descriptor {
+	__u16 major_version;
+	__u16 minor_version;
+
+	/* size of the shm. This may be redundant but nice to have */
+	__u32 size;
+
+	/* number of shared memory regions */
+	__u32 region_count;
+
+	/* The offset to the start of region descriptors */
+	__u32 vsoc_region_desc_offset;
+};
+
+/*
+ * This specifies the current version that should be stored in
+ * vsoc_shm_layout_descriptor.major_version and
+ * vsoc_shm_layout_descriptor.minor_version.
+ * It should be updated only if the vsoc_device_region and
+ * vsoc_shm_layout_descriptor structures have changed.
+ * Versioning within each region is transferred
+ * via the min_compatible_version and current_version fields in
+ * vsoc_device_region. The driver does not consult these fields: they are left
+ * for the HALs and host processes and will change independently of the layout
+ * version.
+ */
+#define CURRENT_VSOC_LAYOUT_MAJOR_VERSION 2
+#define CURRENT_VSOC_LAYOUT_MINOR_VERSION 0
+
+#define VSOC_CREATE_FD_SCOPED_PERMISSION \
+	_IOW(0xF5, 0, struct fd_scoped_permission)
+#define VSOC_GET_FD_SCOPED_PERMISSION _IOR(0xF5, 1, struct fd_scoped_permission)
+
+/*
+ * This is used to signal the host to scan the guest_to_host_signal_table
+ * for new futexes to wake. This sends an interrupt if one is not already
+ * in flight.
+ */
+#define VSOC_MAYBE_SEND_INTERRUPT_TO_HOST _IO(0xF5, 2)
+
+/*
+ * When this returns the guest will scan host_to_guest_signal_table to
+ * check for new futexes to wake.
+ */
+/* TODO(ghartman): Consider moving this to the bottom half */
+#define VSOC_WAIT_FOR_INCOMING_INTERRUPT _IO(0xF5, 3)
+
+/*
+ * Guest HALs will use this to retrieve the region description after
+ * opening their device node.
+ */
+#define VSOC_DESCRIBE_REGION _IOR(0xF5, 4, struct vsoc_device_region)
+
+/*
+ * Wake any threads that may be waiting for a host interrupt on this region.
+ * This is mostly used during shutdown.
+ */
+#define VSOC_SELF_INTERRUPT _IO(0xF5, 5)
+
+/*
+ * This is used to signal the host to scan the guest_to_host_signal_table
+ * for new futexes to wake. This sends an interrupt unconditionally.
+ */
+#define VSOC_SEND_INTERRUPT_TO_HOST _IO(0xF5, 6)
+
+enum wait_types {
+	VSOC_WAIT_UNDEFINED = 0,
+	VSOC_WAIT_IF_EQUAL = 1,
+	VSOC_WAIT_IF_EQUAL_TIMEOUT = 2
+};
+
+/*
+ * Wait for a condition to be true
+ *
+ * Note, this is sized and aligned so the 32 bit and 64 bit layouts are
+ * identical.
+ */
+struct vsoc_cond_wait {
+	/* Input: Offset of the 32 bit word to check */
+	__u32 offset;
+	/* Input: Value that will be compared with the offset */
+	__u32 value;
+	/* Monotonic time to wake at in seconds */
+	__u64 wake_time_sec;
+	/* Input: Monotonic time to wait in nanoseconds */
+	__u32 wake_time_nsec;
+	/* Input: Type of wait */
+	__u32 wait_type;
+	/* Output: Number of times the thread woke before returning. */
+	__u32 wakes;
+	/* Ensure that we're 8-byte aligned and 8 byte length for 32/64 bit
+	 * compatibility.
+	 */
+	__u32 reserved_1;
+};
+
+#define VSOC_COND_WAIT _IOWR(0xF5, 7, struct vsoc_cond_wait)
+
+/* Wake any local threads waiting at the offset given in arg */
+#define VSOC_COND_WAKE _IO(0xF5, 8)
+
+#endif /* _UAPI_LINUX_VSOC_SHM_H */
diff --git a/drivers/staging/android/vsoc.c b/drivers/staging/android/vsoc.c
new file mode 100644
index 000000000000..61c6f73b372a
--- /dev/null
+++ b/drivers/staging/android/vsoc.c
@@ -0,0 +1,1167 @@
+/*
+ * drivers/android/staging/vsoc.c
+ *
+ * Android Virtual System on a Chip (VSoC) driver
+ *
+ * Copyright (C) 2017 Google, Inc.
+ *
+ * Author: ghartman@xxxxxxxxxx
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ *
+ * Based on drivers/char/kvm_ivshmem.c - driver for KVM Inter-VM shared memory
+ *         Copyright 2009 Cam Macdonell <cam@xxxxxxxxxxxxxx>
+ *
+ * Based on cirrusfb.c and 8139cp.c:
+ *   Copyright 1999-2001 Jeff Garzik
+ *   Copyright 2001-2004 Jeff Garzik
+ */
+
+#include <linux/dma-mapping.h>
+#include <linux/freezer.h>
+#include <linux/futex.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/pci.h>
+#include <linux/proc_fs.h>
+#include <linux/sched.h>
+#include <linux/syscalls.h>
+#include <linux/uaccess.h>
+#include <linux/interrupt.h>
+#include <linux/mutex.h>
+#include <linux/cdev.h>
+#include <linux/file.h>
+#include "uapi/vsoc_shm.h"
+
+#define VSOC_DEV_NAME "vsoc"
+
+/*
+ * Description of the ivshmem-doorbell PCI device used by QEmu. These
+ * constants follow docs/specs/ivshmem-spec.txt, which can be found in
+ * the QEmu repository. This was last reconciled with the version that
+ * came out with 2.8
+ */
+
+/*
+ * These constants are determined KVM Inter-VM shared memory device
+ * register offsets
+ */
+enum {
+	INTR_MASK = 0x00,	/* Interrupt Mask */
+	INTR_STATUS = 0x04,	/* Interrupt Status */
+	IV_POSITION = 0x08,	/* VM ID */
+	DOORBELL = 0x0c,	/* Doorbell */
+};
+
+static const int REGISTER_BAR;  /* Equal to 0 */
+static const int MAX_REGISTER_BAR_LEN = 0x100;
+/*
+ * The MSI-x BAR is not used directly.
+ *
+ * static const int MSI_X_BAR = 1;
+ */
+static const int SHARED_MEMORY_BAR = 2;
+
+struct vsoc_region_data {
+	char name[VSOC_DEVICE_NAME_SZ + 1];
+	wait_queue_head_t interrupt_wait_queue;
+	/* TODO(b/73664181): Use multiple futex wait queues */
+	wait_queue_head_t futex_wait_queue;
+	/* Flag indicating that an interrupt has been signalled by the host. */
+	atomic_t *incoming_signalled;
+	/* Flag indicating the guest has signalled the host. */
+	atomic_t *outgoing_signalled;
+	int irq_requested;
+	int device_created;
+};
+
+struct vsoc_device {
+	/* Kernel virtual address of REGISTER_BAR. */
+	void __iomem *regs;
+	/* Physical address of SHARED_MEMORY_BAR. */
+	phys_addr_t shm_phys_start;
+	/* Kernel virtual address of SHARED_MEMORY_BAR. */
+	void *kernel_mapped_shm;
+	/* Size of the entire shared memory window in bytes. */
+	size_t shm_size;
+	/*
+	 * Pointer to the virtual address of the shared memory layout structure.
+	 * This is probably identical to kernel_mapped_shm, but saving this
+	 * here saves a lot of annoying casts.
+	 */
+	struct vsoc_shm_layout_descriptor *layout;
+	/*
+	 * Points to a table of region descriptors in the kernel's virtual
+	 * address space. Calculated from
+	 * vsoc_shm_layout_descriptor.vsoc_region_desc_offset
+	 */
+	struct vsoc_device_region *regions;
+	/* Head of a list of permissions that have been granted. */
+	struct list_head permissions;
+	struct pci_dev *dev;
+	/* Per-region (and therefore per-interrupt) information. */
+	struct vsoc_region_data *regions_data;
+	/*
+	 * Table of msi-x entries. This has to be separated from struct
+	 * vsoc_region_data because the kernel deals with them as an array.
+	 */
+	struct msix_entry *msix_entries;
+	/*
+	 * Flags that indicate what we've initialzied. These are used to do an
+	 * orderly cleanup of the device.
+	 */
+	char enabled_device;
+	char requested_regions;
+	char cdev_added;
+	char msix_enabled;
+	/* Mutex that protectes the permission list */
+	struct mutex mtx;
+	/* Major number assigned by the kernel */
+	int major;
+
+	struct cdev cdev;
+	struct class *class;
+};
+
+static struct vsoc_device vsoc_dev;
+
+/*
+ * TODO(ghartman): Add a /sys filesystem entry that summarizes the permissions.
+ */
+
+struct fd_scoped_permission_node {
+	struct fd_scoped_permission permission;
+	struct list_head list;
+};
+
+struct vsoc_private_data {
+	struct fd_scoped_permission_node *fd_scoped_permission_node;
+};
+
+static long vsoc_ioctl(struct file *, unsigned int, unsigned long);
+static int vsoc_mmap(struct file *, struct vm_area_struct *);
+static int vsoc_open(struct inode *, struct file *);
+static int vsoc_release(struct inode *, struct file *);
+static ssize_t vsoc_read(struct file *, char *, size_t, loff_t *);
+static ssize_t vsoc_write(struct file *, const char *, size_t, loff_t *);
+static loff_t vsoc_lseek(struct file *filp, loff_t offset, int origin);
+static int do_create_fd_scoped_permission(
+	struct vsoc_device_region *region_p,
+	struct fd_scoped_permission_node *np,
+	struct fd_scoped_permission_arg *__user arg);
+static void do_destroy_fd_scoped_permission(
+	struct vsoc_device_region *owner_region_p,
+	struct fd_scoped_permission *perm);
+static long do_vsoc_describe_region(struct file *,
+				    struct vsoc_device_region __user *);
+static ssize_t vsoc_get_area(struct file *filp, __u32 *perm_off);
+
+/**
+ * Validate arguments on entry points to the driver.
+ */
+inline int vsoc_validate_inode(struct inode *inode)
+{
+	if (iminor(inode) >= vsoc_dev.layout->region_count) {
+		dev_err(&vsoc_dev.dev->dev,
+			"describe_region: invalid region %d\n", iminor(inode));
+		return -ENODEV;
+	}
+	return 0;
+}
+
+inline int vsoc_validate_filep(struct file *filp)
+{
+	int ret = vsoc_validate_inode(file_inode(filp));
+
+	if (ret)
+		return ret;
+	if (!filp->private_data) {
+		dev_err(&vsoc_dev.dev->dev,
+			"No private data on fd, region %d\n",
+			iminor(file_inode(filp)));
+		return -EBADFD;
+	}
+	return 0;
+}
+
+/* Converts from shared memory offset to virtual address */
+static inline void *shm_off_to_virtual_addr(__u32 offset)
+{
+	return vsoc_dev.kernel_mapped_shm + offset;
+}
+
+/* Converts from shared memory offset to physical address */
+static inline phys_addr_t shm_off_to_phys_addr(__u32 offset)
+{
+	return vsoc_dev.shm_phys_start + offset;
+}
+
+/**
+ * Convenience functions to obtain the region from the inode or file.
+ * Dangerous to call before validating the inode/file.
+ */
+static inline struct vsoc_device_region *vsoc_region_from_inode(
+	struct inode *inode)
+{
+	return &vsoc_dev.regions[iminor(inode)];
+}
+
+static inline struct vsoc_device_region *vsoc_region_from_filep(
+	struct file *inode)
+{
+	return vsoc_region_from_inode(file_inode(inode));
+}
+
+static inline uint32_t vsoc_device_region_size(struct vsoc_device_region *r)
+{
+	return r->region_end_offset - r->region_begin_offset;
+}
+
+static const struct file_operations vsoc_ops = {
+	.owner = THIS_MODULE,
+	.open = vsoc_open,
+	.mmap = vsoc_mmap,
+	.read = vsoc_read,
+	.unlocked_ioctl = vsoc_ioctl,
+	.compat_ioctl = vsoc_ioctl,
+	.write = vsoc_write,
+	.llseek = vsoc_lseek,
+	.release = vsoc_release,
+};
+
+static struct pci_device_id vsoc_id_table[] = {
+	{0x1af4, 0x1110, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0},
+	{0},
+};
+
+MODULE_DEVICE_TABLE(pci, vsoc_id_table);
+
+static void vsoc_remove_device(struct pci_dev *pdev);
+static int vsoc_probe_device(struct pci_dev *pdev,
+			     const struct pci_device_id *ent);
+
+static struct pci_driver vsoc_pci_driver = {
+	.name = "vsoc",
+	.id_table = vsoc_id_table,
+	.probe = vsoc_probe_device,
+	.remove = vsoc_remove_device,
+};
+
+static int do_create_fd_scoped_permission(
+	struct vsoc_device_region *region_p,
+	struct fd_scoped_permission_node *np,
+	struct fd_scoped_permission_arg *__user arg)
+{
+	struct file *managed_filp;
+	s32 managed_fd;
+	atomic_t *owner_ptr = NULL;
+	struct vsoc_device_region *managed_region_p;
+
+	if (copy_from_user(&np->permission, &arg->perm, sizeof(*np)) ||
+	    copy_from_user(&managed_fd,
+			   &arg->managed_region_fd, sizeof(managed_fd))) {
+		return -EFAULT;
+	}
+	managed_filp = fdget(managed_fd).file;
+	/* Check that it's a valid fd, */
+	if (!managed_filp || vsoc_validate_filep(managed_filp))
+		return -EPERM;
+	/* EEXIST if the given fd already has a permission. */
+	if (((struct vsoc_private_data *)managed_filp->private_data)->
+	    fd_scoped_permission_node)
+		return -EEXIST;
+	managed_region_p = vsoc_region_from_filep(managed_filp);
+	/* Check that the provided region is managed by this one */
+	if (&vsoc_dev.regions[managed_region_p->managed_by] != region_p)
+		return -EPERM;
+	/* The area must be well formed and have non-zero size */
+	if (np->permission.begin_offset >= np->permission.end_offset)
+		return -EINVAL;
+	/* The area must fit in the memory window */
+	if (np->permission.end_offset >
+	    vsoc_device_region_size(managed_region_p))
+		return -ERANGE;
+	/* The area must be in the region data section */
+	if (np->permission.begin_offset <
+	    managed_region_p->offset_of_region_data)
+		return -ERANGE;
+	/* The area must be page aligned */
+	if (!PAGE_ALIGNED(np->permission.begin_offset) ||
+	    !PAGE_ALIGNED(np->permission.end_offset))
+		return -EINVAL;
+	/* The owner flag must reside in the owner memory */
+	if (np->permission.owner_offset + sizeof(np->permission.owner_offset) >
+	    vsoc_device_region_size(region_p))
+		return -ERANGE;
+	/* The owner flag must reside in the data section */
+	if (np->permission.owner_offset < region_p->offset_of_region_data)
+		return -EINVAL;
+	/* Owner offset must be naturally aligned in the window */
+	if (np->permission.owner_offset &
+	    (sizeof(np->permission.owner_offset) - 1))
+		return -EINVAL;
+	/* The owner value must change to claim the memory */
+	if (np->permission.owned_value == VSOC_REGION_FREE)
+		return -EINVAL;
+	owner_ptr =
+	    (atomic_t *)shm_off_to_virtual_addr(region_p->region_begin_offset +
+						np->permission.owner_offset);
+	/* We've already verified that this is in the shared memory window, so
+	 * it should be safe to write to this address.
+	 */
+	if (atomic_cmpxchg(owner_ptr,
+			   VSOC_REGION_FREE,
+			   np->permission.owned_value) != VSOC_REGION_FREE) {
+		return -EBUSY;
+	}
+	((struct vsoc_private_data *)managed_filp->private_data)->
+	    fd_scoped_permission_node = np;
+	/* The file offset needs to be adjusted if the calling
+	 * process did any read/write operations on the fd
+	 * before creating the permission.
+	 */
+	if (managed_filp->f_pos) {
+		if (managed_filp->f_pos > np->permission.end_offset) {
+			/* If the offset is beyond the permission end, set it
+			 * to the end.
+			 */
+			managed_filp->f_pos = np->permission.end_offset;
+		} else {
+			/* If the offset is within the permission interval
+			 * keep it there otherwise reset it to zero.
+			 */
+			if (managed_filp->f_pos < np->permission.begin_offset) {
+				managed_filp->f_pos = 0;
+			} else {
+				managed_filp->f_pos -=
+				    np->permission.begin_offset;
+			}
+		}
+	}
+	return 0;
+}
+
+static void do_destroy_fd_scoped_permission_node(
+	struct vsoc_device_region *owner_region_p,
+	struct fd_scoped_permission_node *node)
+{
+	if (node) {
+		do_destroy_fd_scoped_permission(owner_region_p,
+						&node->permission);
+		mutex_lock(&vsoc_dev.mtx);
+		list_del(&node->list);
+		mutex_unlock(&vsoc_dev.mtx);
+		kfree(node);
+	}
+}
+
+static void do_destroy_fd_scoped_permission(
+		struct vsoc_device_region *owner_region_p,
+		struct fd_scoped_permission *perm)
+{
+	atomic_t *owner_ptr = NULL;
+	int prev = 0;
+
+	if (!perm)
+		return;
+	owner_ptr = (atomic_t *)shm_off_to_virtual_addr(
+		owner_region_p->region_begin_offset + perm->owner_offset);
+	prev = atomic_xchg(owner_ptr, VSOC_REGION_FREE);
+	if (prev != perm->owned_value)
+		dev_err(&vsoc_dev.dev->dev,
+			"%x-%x: owner (%s) %x: expected to be %x was %x",
+			perm->begin_offset, perm->end_offset,
+			owner_region_p->device_name, perm->owner_offset,
+			perm->owned_value, prev);
+}
+
+static long do_vsoc_describe_region(struct file *filp,
+				    struct vsoc_device_region __user *dest)
+{
+	struct vsoc_device_region *region_p;
+	int retval = vsoc_validate_filep(filp);
+
+	if (retval)
+		return retval;
+	region_p = vsoc_region_from_filep(filp);
+	if (copy_to_user(dest, region_p, sizeof(*region_p)))
+		return -EFAULT;
+	return 0;
+}
+
+/**
+ * Implements the inner logic of cond_wait. Copies to and from userspace are
+ * done in the helper function below.
+ */
+static int handle_vsoc_cond_wait(struct file *filp, struct vsoc_cond_wait *arg)
+{
+	DEFINE_WAIT(wait);
+	u32 region_number = iminor(file_inode(filp));
+	struct vsoc_region_data *data = vsoc_dev.regions_data + region_number;
+	struct hrtimer_sleeper timeout, *to = NULL;
+	int ret = 0;
+	struct vsoc_device_region *region_p = vsoc_region_from_filep(filp);
+	atomic_t *address = NULL;
+	struct timespec ts;
+
+	/* Ensure that the offset is aligned */
+	if (arg->offset & (sizeof(uint32_t) - 1))
+		return -EADDRNOTAVAIL;
+	/* Ensure that the offset is within shared memory */
+	if (((uint64_t)arg->offset) + region_p->region_begin_offset +
+	    sizeof(uint32_t) > region_p->region_end_offset)
+		return -E2BIG;
+	address = shm_off_to_virtual_addr(region_p->region_begin_offset +
+					  arg->offset);
+
+	/* Ensure that the type of wait is valid */
+	switch (arg->wait_type) {
+	case VSOC_WAIT_IF_EQUAL:
+		break;
+	case VSOC_WAIT_IF_EQUAL_TIMEOUT:
+		to = &timeout;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	if (to) {
+		/* Copy the user-supplied timesec into the kernel structure.
+		 * We do things this way to flatten differences between 32 bit
+		 * and 64 bit timespecs.
+		 */
+		ts.tv_sec = arg->wake_time_sec;
+		ts.tv_nsec = arg->wake_time_nsec;
+
+		if (!timespec_valid(&ts))
+			return -EINVAL;
+		hrtimer_init_on_stack(&to->timer, CLOCK_MONOTONIC,
+				      HRTIMER_MODE_ABS);
+		hrtimer_set_expires_range_ns(&to->timer, timespec_to_ktime(ts),
+					     current->timer_slack_ns);
+
+		hrtimer_init_sleeper(to, current);
+	}
+
+	while (1) {
+		prepare_to_wait(&data->futex_wait_queue, &wait,
+				TASK_INTERRUPTIBLE);
+		/*
+		 * Check the sentinel value after prepare_to_wait. If the value
+		 * changes after this check the writer will call signal,
+		 * changing the task state from INTERRUPTIBLE to RUNNING. That
+		 * will ensure that schedule() will eventually schedule this
+		 * task.
+		 */
+		if (atomic_read(address) != arg->value) {
+			ret = 0;
+			break;
+		}
+		if (to) {
+			hrtimer_start_expires(&to->timer, HRTIMER_MODE_ABS);
+			if (likely(to->task))
+				freezable_schedule();
+			hrtimer_cancel(&to->timer);
+			if (!to->task) {
+				ret = -ETIMEDOUT;
+				break;
+			}
+		} else {
+			freezable_schedule();
+		}
+		/* Count the number of times that we woke up. This is useful
+		 * for unit testing.
+		 */
+		++arg->wakes;
+		if (signal_pending(current)) {
+			ret = -EINTR;
+			break;
+		}
+	}
+	finish_wait(&data->futex_wait_queue, &wait);
+	if (to)
+		destroy_hrtimer_on_stack(&to->timer);
+	return ret;
+}
+
+/**
+ * Handles the details of copying from/to userspace to ensure that the copies
+ * happen on all of the return paths of cond_wait.
+ */
+static int do_vsoc_cond_wait(struct file *filp,
+			     struct vsoc_cond_wait __user *untrusted_in)
+{
+	struct vsoc_cond_wait arg;
+	int rval = 0;
+
+	if (copy_from_user(&arg, untrusted_in, sizeof(arg)))
+		return -EFAULT;
+	/* wakes is an out parameter. Initialize it to something sensible. */
+	arg.wakes = 0;
+	rval = handle_vsoc_cond_wait(filp, &arg);
+	if (copy_to_user(untrusted_in, &arg, sizeof(arg)))
+		return -EFAULT;
+	return rval;
+}
+
+static int do_vsoc_cond_wake(struct file *filp, uint32_t offset)
+{
+	struct vsoc_device_region *region_p = vsoc_region_from_filep(filp);
+	u32 region_number = iminor(file_inode(filp));
+	struct vsoc_region_data *data = vsoc_dev.regions_data + region_number;
+	/* Ensure that the offset is aligned */
+	if (offset & (sizeof(uint32_t) - 1))
+		return -EADDRNOTAVAIL;
+	/* Ensure that the offset is within shared memory */
+	if (((uint64_t)offset) + region_p->region_begin_offset +
+	    sizeof(uint32_t) > region_p->region_end_offset)
+		return -E2BIG;
+	/*
+	 * TODO(b/73664181): Use multiple futex wait queues.
+	 * We need to wake every sleeper when the condition changes. Typically
+	 * only a single thread will be waiting on the condition, but there
+	 * are exceptions. The worst case is about 10 threads.
+	 */
+	wake_up_interruptible_all(&data->futex_wait_queue);
+	return 0;
+}
+
+static long vsoc_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
+{
+	int rv = 0;
+	struct vsoc_device_region *region_p;
+	u32 reg_num;
+	struct vsoc_region_data *reg_data;
+	int retval = vsoc_validate_filep(filp);
+
+	if (retval)
+		return retval;
+	region_p = vsoc_region_from_filep(filp);
+	reg_num = iminor(file_inode(filp));
+	reg_data = vsoc_dev.regions_data + reg_num;
+	switch (cmd) {
+	case VSOC_CREATE_FD_SCOPED_PERMISSION:
+		{
+			struct fd_scoped_permission_node *node = NULL;
+
+			node = kzalloc(sizeof(*node), GFP_KERNEL);
+			/* We can't allocate memory for the permission */
+			if (!node)
+				return -ENOMEM;
+			INIT_LIST_HEAD(&node->list);
+			rv = do_create_fd_scoped_permission(
+				region_p,
+				node,
+				(struct fd_scoped_permission_arg __user *)arg);
+			if (!rv) {
+				mutex_lock(&vsoc_dev.mtx);
+				list_add(&node->list, &vsoc_dev.permissions);
+				mutex_unlock(&vsoc_dev.mtx);
+			} else {
+				kfree(node);
+				return rv;
+			}
+		}
+		break;
+
+	case VSOC_GET_FD_SCOPED_PERMISSION:
+		{
+			struct fd_scoped_permission_node *node =
+			    ((struct vsoc_private_data *)filp->private_data)->
+			    fd_scoped_permission_node;
+			if (!node)
+				return -ENOENT;
+			if (copy_to_user
+			    ((struct fd_scoped_permission __user *)arg,
+			     &node->permission, sizeof(node->permission)))
+				return -EFAULT;
+		}
+		break;
+
+	case VSOC_MAYBE_SEND_INTERRUPT_TO_HOST:
+		if (!atomic_xchg(
+			    reg_data->outgoing_signalled,
+			    1)) {
+			writel(reg_num, vsoc_dev.regs + DOORBELL);
+			return 0;
+		} else {
+			return -EBUSY;
+		}
+		break;
+
+	case VSOC_SEND_INTERRUPT_TO_HOST:
+		writel(reg_num, vsoc_dev.regs + DOORBELL);
+		return 0;
+
+	case VSOC_WAIT_FOR_INCOMING_INTERRUPT:
+		wait_event_interruptible(
+			reg_data->interrupt_wait_queue,
+			(atomic_read(reg_data->incoming_signalled) != 0));
+		break;
+
+	case VSOC_DESCRIBE_REGION:
+		return do_vsoc_describe_region(
+			filp,
+			(struct vsoc_device_region __user *)arg);
+
+	case VSOC_SELF_INTERRUPT:
+		atomic_set(reg_data->incoming_signalled, 1);
+		wake_up_interruptible(&reg_data->interrupt_wait_queue);
+		break;
+
+	case VSOC_COND_WAIT:
+		return do_vsoc_cond_wait(filp,
+					 (struct vsoc_cond_wait __user *)arg);
+	case VSOC_COND_WAKE:
+		return do_vsoc_cond_wake(filp, arg);
+
+	default:
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static ssize_t vsoc_read(struct file *filp, char *buffer, size_t len,
+			 loff_t *poffset)
+{
+	__u32 area_off;
+	void *area_p;
+	ssize_t area_len;
+	int retval = vsoc_validate_filep(filp);
+
+	if (retval)
+		return retval;
+	area_len = vsoc_get_area(filp, &area_off);
+	area_p = shm_off_to_virtual_addr(area_off);
+	area_p += *poffset;
+	area_len -= *poffset;
+	if (area_len <= 0)
+		return 0;
+	if (area_len < len)
+		len = area_len;
+	if (copy_to_user(buffer, area_p, len))
+		return -EFAULT;
+	*poffset += len;
+	return len;
+}
+
+static loff_t vsoc_lseek(struct file *filp, loff_t offset, int origin)
+{
+	ssize_t area_len = 0;
+	int retval = vsoc_validate_filep(filp);
+
+	if (retval)
+		return retval;
+	area_len = vsoc_get_area(filp, NULL);
+	switch (origin) {
+	case SEEK_SET:
+		break;
+
+	case SEEK_CUR:
+		if (offset > 0 && offset + filp->f_pos < 0)
+			return -EOVERFLOW;
+		offset += filp->f_pos;
+		break;
+
+	case SEEK_END:
+		if (offset > 0 && offset + area_len < 0)
+			return -EOVERFLOW;
+		offset += area_len;
+		break;
+
+	case SEEK_DATA:
+		if (offset >= area_len)
+			return -EINVAL;
+		if (offset < 0)
+			offset = 0;
+		break;
+
+	case SEEK_HOLE:
+		/* Next hole is always the end of the region, unless offset is
+		 * beyond that
+		 */
+		if (offset < area_len)
+			offset = area_len;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	if (offset < 0 || offset > area_len)
+		return -EINVAL;
+	filp->f_pos = offset;
+
+	return offset;
+}
+
+static ssize_t vsoc_write(struct file *filp, const char *buffer,
+			  size_t len, loff_t *poffset)
+{
+	__u32 area_off;
+	void *area_p;
+	ssize_t area_len;
+	int retval = vsoc_validate_filep(filp);
+
+	if (retval)
+		return retval;
+	area_len = vsoc_get_area(filp, &area_off);
+	area_p = shm_off_to_virtual_addr(area_off);
+	area_p += *poffset;
+	area_len -= *poffset;
+	if (area_len <= 0)
+		return 0;
+	if (area_len < len)
+		len = area_len;
+	if (copy_from_user(area_p, buffer, len))
+		return -EFAULT;
+	*poffset += len;
+	return len;
+}
+
+static irqreturn_t vsoc_interrupt(int irq, void *region_data_v)
+{
+	struct vsoc_region_data *region_data =
+	    (struct vsoc_region_data *)region_data_v;
+	int reg_num = region_data - vsoc_dev.regions_data;
+
+	if (unlikely(!region_data))
+		return IRQ_NONE;
+
+	if (unlikely(reg_num < 0 ||
+		     reg_num >= vsoc_dev.layout->region_count)) {
+		dev_err(&vsoc_dev.dev->dev,
+			"invalid irq @%p reg_num=0x%04x\n",
+			region_data, reg_num);
+		return IRQ_NONE;
+	}
+	if (unlikely(vsoc_dev.regions_data + reg_num != region_data)) {
+		dev_err(&vsoc_dev.dev->dev,
+			"irq not aligned @%p reg_num=0x%04x\n",
+			region_data, reg_num);
+		return IRQ_NONE;
+	}
+	wake_up_interruptible(&region_data->interrupt_wait_queue);
+	return IRQ_HANDLED;
+}
+
+static int vsoc_probe_device(struct pci_dev *pdev,
+			     const struct pci_device_id *ent)
+{
+	int result;
+	int i;
+	resource_size_t reg_size;
+	dev_t devt;
+
+	vsoc_dev.dev = pdev;
+	result = pci_enable_device(pdev);
+	if (result) {
+		dev_err(&pdev->dev,
+			"pci_enable_device failed %s: error %d\n",
+			pci_name(pdev), result);
+		return result;
+	}
+	vsoc_dev.enabled_device = 1;
+	result = pci_request_regions(pdev, "vsoc");
+	if (result < 0) {
+		dev_err(&pdev->dev, "pci_request_regions failed\n");
+		vsoc_remove_device(pdev);
+		return -EBUSY;
+	}
+	vsoc_dev.requested_regions = 1;
+	/* Set up the control registers in BAR 0 */
+	reg_size = pci_resource_len(pdev, REGISTER_BAR);
+	if (reg_size > MAX_REGISTER_BAR_LEN)
+		vsoc_dev.regs =
+		    pci_iomap(pdev, REGISTER_BAR, MAX_REGISTER_BAR_LEN);
+	else
+		vsoc_dev.regs = pci_iomap(pdev, REGISTER_BAR, reg_size);
+
+	if (!vsoc_dev.regs) {
+		dev_err(&pdev->dev,
+			"cannot ioremap registers of size %zu\n",
+		       (size_t)reg_size);
+		vsoc_remove_device(pdev);
+		return -EBUSY;
+	}
+
+	/* Map the shared memory in BAR 2 */
+	vsoc_dev.shm_phys_start = pci_resource_start(pdev, SHARED_MEMORY_BAR);
+	vsoc_dev.shm_size = pci_resource_len(pdev, SHARED_MEMORY_BAR);
+
+	dev_info(&pdev->dev, "shared memory @ DMA %p size=0x%zx\n",
+		 (void *)vsoc_dev.shm_phys_start, vsoc_dev.shm_size);
+	/* TODO(ghartman): ioremap_wc should work here */
+	vsoc_dev.kernel_mapped_shm = ioremap_nocache(
+			vsoc_dev.shm_phys_start, vsoc_dev.shm_size);
+	if (!vsoc_dev.kernel_mapped_shm) {
+		dev_err(&vsoc_dev.dev->dev, "cannot iomap region\n");
+		vsoc_remove_device(pdev);
+		return -EBUSY;
+	}
+
+	vsoc_dev.layout =
+	    (struct vsoc_shm_layout_descriptor *)vsoc_dev.kernel_mapped_shm;
+	dev_info(&pdev->dev, "major_version: %d\n",
+		 vsoc_dev.layout->major_version);
+	dev_info(&pdev->dev, "minor_version: %d\n",
+		 vsoc_dev.layout->minor_version);
+	dev_info(&pdev->dev, "size: 0x%x\n", vsoc_dev.layout->size);
+	dev_info(&pdev->dev, "regions: %d\n", vsoc_dev.layout->region_count);
+	if (vsoc_dev.layout->major_version !=
+	    CURRENT_VSOC_LAYOUT_MAJOR_VERSION) {
+		dev_err(&vsoc_dev.dev->dev,
+			"driver supports only major_version %d\n",
+			CURRENT_VSOC_LAYOUT_MAJOR_VERSION);
+		vsoc_remove_device(pdev);
+		return -EBUSY;
+	}
+	result = alloc_chrdev_region(&devt, 0, vsoc_dev.layout->region_count,
+				     VSOC_DEV_NAME);
+	if (result) {
+		dev_err(&vsoc_dev.dev->dev, "alloc_chrdev_region failed\n");
+		vsoc_remove_device(pdev);
+		return -EBUSY;
+	}
+	vsoc_dev.major = MAJOR(devt);
+	cdev_init(&vsoc_dev.cdev, &vsoc_ops);
+	vsoc_dev.cdev.owner = THIS_MODULE;
+	result = cdev_add(&vsoc_dev.cdev, devt, vsoc_dev.layout->region_count);
+	if (result) {
+		dev_err(&vsoc_dev.dev->dev, "cdev_add error\n");
+		vsoc_remove_device(pdev);
+		return -EBUSY;
+	}
+	vsoc_dev.cdev_added = 1;
+	vsoc_dev.class = class_create(THIS_MODULE, VSOC_DEV_NAME);
+	if (!vsoc_dev.class) {
+		dev_err(&vsoc_dev.dev->dev, "class_create failed\n");
+		vsoc_remove_device(pdev);
+		return -EBUSY;
+	}
+	vsoc_dev.regions = (struct vsoc_device_region *)
+		(vsoc_dev.kernel_mapped_shm +
+		 vsoc_dev.layout->vsoc_region_desc_offset);
+	vsoc_dev.msix_entries = kcalloc(
+			vsoc_dev.layout->region_count,
+			sizeof(vsoc_dev.msix_entries[0]), GFP_KERNEL);
+	if (!vsoc_dev.msix_entries) {
+		dev_err(&vsoc_dev.dev->dev,
+			"unable to allocate msix_entries\n");
+		vsoc_remove_device(pdev);
+		return -ENOSPC;
+	}
+	vsoc_dev.regions_data = kcalloc(
+			vsoc_dev.layout->region_count,
+			sizeof(vsoc_dev.regions_data[0]), GFP_KERNEL);
+	if (!vsoc_dev.regions_data) {
+		dev_err(&vsoc_dev.dev->dev,
+			"unable to allocate regions' data\n");
+		vsoc_remove_device(pdev);
+		return -ENOSPC;
+	}
+	for (i = 0; i < vsoc_dev.layout->region_count; ++i)
+		vsoc_dev.msix_entries[i].entry = i;
+
+	result = pci_enable_msix_exact(vsoc_dev.dev, vsoc_dev.msix_entries,
+				       vsoc_dev.layout->region_count);
+	if (result) {
+		dev_info(&pdev->dev, "pci_enable_msix failed: %d\n", result);
+		vsoc_remove_device(pdev);
+		return -ENOSPC;
+	}
+	/* Check that all regions are well formed */
+	for (i = 0; i < vsoc_dev.layout->region_count; ++i) {
+		const struct vsoc_device_region *region = vsoc_dev.regions + i;
+
+		if (!PAGE_ALIGNED(region->region_begin_offset) ||
+		    !PAGE_ALIGNED(region->region_end_offset)) {
+			dev_err(&vsoc_dev.dev->dev,
+				"region %d not aligned (%x:%x)", i,
+				region->region_begin_offset,
+				region->region_end_offset);
+			vsoc_remove_device(pdev);
+			return -EFAULT;
+		}
+		if (region->region_begin_offset >= region->region_end_offset ||
+		    region->region_end_offset > vsoc_dev.shm_size) {
+			dev_err(&vsoc_dev.dev->dev,
+				"region %d offsets are wrong: %x %x %zx",
+				i, region->region_begin_offset,
+				region->region_end_offset, vsoc_dev.shm_size);
+			vsoc_remove_device(pdev);
+			return -EFAULT;
+		}
+		if (region->managed_by >= vsoc_dev.layout->region_count) {
+			dev_err(&vsoc_dev.dev->dev,
+				"region %d has invalid owner: %u",
+				i, region->managed_by);
+			vsoc_remove_device(pdev);
+			return -EFAULT;
+		}
+	}
+	vsoc_dev.msix_enabled = 1;
+	for (i = 0; i < vsoc_dev.layout->region_count; ++i) {
+		const struct vsoc_device_region *region = vsoc_dev.regions + i;
+		size_t name_sz = sizeof(vsoc_dev.regions_data[i].name) - 1;
+		const struct vsoc_signal_table_layout *h_to_g_signal_table =
+			&region->host_to_guest_signal_table;
+		const struct vsoc_signal_table_layout *g_to_h_signal_table =
+			&region->guest_to_host_signal_table;
+
+		vsoc_dev.regions_data[i].name[name_sz] = '\0';
+		memcpy(vsoc_dev.regions_data[i].name, region->device_name,
+		       name_sz);
+		dev_info(&pdev->dev, "region %d name=%s\n",
+			 i, vsoc_dev.regions_data[i].name);
+		init_waitqueue_head(
+				&vsoc_dev.regions_data[i].interrupt_wait_queue);
+		init_waitqueue_head(&vsoc_dev.regions_data[i].futex_wait_queue);
+		vsoc_dev.regions_data[i].incoming_signalled =
+			vsoc_dev.kernel_mapped_shm +
+			region->region_begin_offset +
+			h_to_g_signal_table->interrupt_signalled_offset;
+		vsoc_dev.regions_data[i].outgoing_signalled =
+			vsoc_dev.kernel_mapped_shm +
+			region->region_begin_offset +
+			g_to_h_signal_table->interrupt_signalled_offset;
+
+		result = request_irq(
+				vsoc_dev.msix_entries[i].vector,
+				vsoc_interrupt, 0,
+				vsoc_dev.regions_data[i].name,
+				vsoc_dev.regions_data + i);
+		if (result) {
+			dev_info(&pdev->dev,
+				 "request_irq failed irq=%d vector=%d\n",
+				i, vsoc_dev.msix_entries[i].vector);
+			vsoc_remove_device(pdev);
+			return -ENOSPC;
+		}
+		vsoc_dev.regions_data[i].irq_requested = 1;
+		if (!device_create(vsoc_dev.class, NULL,
+				   MKDEV(vsoc_dev.major, i),
+				   NULL, vsoc_dev.regions_data[i].name)) {
+			dev_err(&vsoc_dev.dev->dev, "device_create failed\n");
+			vsoc_remove_device(pdev);
+			return -EBUSY;
+		}
+		vsoc_dev.regions_data[i].device_created = 1;
+	}
+	return 0;
+}
+
+/*
+ * This should undo all of the allocations in the probe function in reverse
+ * order.
+ *
+ * Notes:
+ *
+ *   The device may have been partially initialized, so double check
+ *   that the allocations happened.
+ *
+ *   This function may be called multiple times, so mark resources as freed
+ *   as they are deallocated.
+ */
+static void vsoc_remove_device(struct pci_dev *pdev)
+{
+	int i;
+	/*
+	 * pdev is the first thing to be set on probe and the last thing
+	 * to be cleared here. If it's NULL then there is no cleanup.
+	 */
+	if (!pdev || !vsoc_dev.dev)
+		return;
+	dev_info(&pdev->dev, "remove_device\n");
+	if (vsoc_dev.regions_data) {
+		for (i = 0; i < vsoc_dev.layout->region_count; ++i) {
+			if (vsoc_dev.regions_data[i].device_created) {
+				device_destroy(vsoc_dev.class,
+					       MKDEV(vsoc_dev.major, i));
+				vsoc_dev.regions_data[i].device_created = 0;
+			}
+			if (vsoc_dev.regions_data[i].irq_requested)
+				free_irq(vsoc_dev.msix_entries[i].vector, NULL);
+			vsoc_dev.regions_data[i].irq_requested = 0;
+		}
+		kfree(vsoc_dev.regions_data);
+		vsoc_dev.regions_data = 0;
+	}
+	if (vsoc_dev.msix_enabled) {
+		pci_disable_msix(pdev);
+		vsoc_dev.msix_enabled = 0;
+	}
+	kfree(vsoc_dev.msix_entries);
+	vsoc_dev.msix_entries = 0;
+	vsoc_dev.regions = 0;
+	if (vsoc_dev.class) {
+		class_destroy(vsoc_dev.class);
+		vsoc_dev.class = 0;
+	}
+	if (vsoc_dev.cdev_added) {
+		cdev_del(&vsoc_dev.cdev);
+		vsoc_dev.cdev_added = 0;
+	}
+	if (vsoc_dev.major && vsoc_dev.layout) {
+		unregister_chrdev_region(MKDEV(vsoc_dev.major, 0),
+					 vsoc_dev.layout->region_count);
+		vsoc_dev.major = 0;
+	}
+	vsoc_dev.layout = 0;
+	if (vsoc_dev.kernel_mapped_shm && pdev) {
+		pci_iounmap(pdev, vsoc_dev.kernel_mapped_shm);
+		vsoc_dev.kernel_mapped_shm = 0;
+	}
+	if (vsoc_dev.regs && pdev) {
+		pci_iounmap(pdev, vsoc_dev.regs);
+		vsoc_dev.regs = 0;
+	}
+	if (vsoc_dev.requested_regions && pdev) {
+		pci_release_regions(pdev);
+		vsoc_dev.requested_regions = 0;
+	}
+	if (vsoc_dev.enabled_device && pdev) {
+		pci_disable_device(pdev);
+		vsoc_dev.enabled_device = 0;
+	}
+	/* Do this last: it indicates that the device is not initialized. */
+	vsoc_dev.dev = NULL;
+}
+
+static void __exit vsoc_cleanup_module(void)
+{
+	vsoc_remove_device(vsoc_dev.dev);
+	pci_unregister_driver(&vsoc_pci_driver);
+}
+
+static int __init vsoc_init_module(void)
+{
+	int err = -ENOMEM;
+
+	INIT_LIST_HEAD(&vsoc_dev.permissions);
+	mutex_init(&vsoc_dev.mtx);
+
+	err = pci_register_driver(&vsoc_pci_driver);
+	if (err < 0)
+		return err;
+	return 0;
+}
+
+static int vsoc_open(struct inode *inode, struct file *filp)
+{
+	/* Can't use vsoc_validate_filep because filp is still incomplete */
+	int ret = vsoc_validate_inode(inode);
+
+	if (ret)
+		return ret;
+	filp->private_data =
+		kzalloc(sizeof(struct vsoc_private_data), GFP_KERNEL);
+	if (!filp->private_data)
+		return -ENOMEM;
+	return 0;
+}
+
+static int vsoc_release(struct inode *inode, struct file *filp)
+{
+	struct vsoc_private_data *private_data = NULL;
+	struct fd_scoped_permission_node *node = NULL;
+	struct vsoc_device_region *owner_region_p = NULL;
+	int retval = vsoc_validate_filep(filp);
+
+	if (retval)
+		return retval;
+	private_data = (struct vsoc_private_data *)filp->private_data;
+	if (!private_data)
+		return 0;
+
+	node = private_data->fd_scoped_permission_node;
+	if (node) {
+		owner_region_p = vsoc_region_from_inode(inode);
+		if (owner_region_p->managed_by != VSOC_REGION_WHOLE) {
+			owner_region_p =
+			    &vsoc_dev.regions[owner_region_p->managed_by];
+		}
+		do_destroy_fd_scoped_permission_node(owner_region_p, node);
+		private_data->fd_scoped_permission_node = NULL;
+	}
+	kfree(private_data);
+	filp->private_data = NULL;
+
+	return 0;
+}
+
+/*
+ * Returns the device relative offset and length of the area specified by the
+ * fd scoped permission. If there is no fd scoped permission set, a default
+ * permission covering the entire region is assumed, unless the region is owned
+ * by another one, in which case the default is a permission with zero size.
+ */
+static ssize_t vsoc_get_area(struct file *filp, __u32 *area_offset)
+{
+	__u32 off = 0;
+	ssize_t length = 0;
+	struct vsoc_device_region *region_p;
+	struct fd_scoped_permission *perm;
+
+	region_p = vsoc_region_from_filep(filp);
+	off = region_p->region_begin_offset;
+	perm = &((struct vsoc_private_data *)filp->private_data)->
+		fd_scoped_permission_node->permission;
+	if (perm) {
+		off += perm->begin_offset;
+		length = perm->end_offset - perm->begin_offset;
+	} else if (region_p->managed_by == VSOC_REGION_WHOLE) {
+		/* No permission set and the regions is not owned by another,
+		 * default to full region access.
+		 */
+		length = vsoc_device_region_size(region_p);
+	} else {
+		/* return zero length, access is denied. */
+		length = 0;
+	}
+	if (area_offset)
+		*area_offset = off;
+	return length;
+}
+
+static int vsoc_mmap(struct file *filp, struct vm_area_struct *vma)
+{
+	unsigned long len = vma->vm_end - vma->vm_start;
+	__u32 area_off;
+	phys_addr_t mem_off;
+	ssize_t area_len;
+	int retval = vsoc_validate_filep(filp);
+
+	if (retval)
+		return retval;
+	area_len = vsoc_get_area(filp, &area_off);
+	/* Add the requested offset */
+	area_off += (vma->vm_pgoff << PAGE_SHIFT);
+	area_len -= (vma->vm_pgoff << PAGE_SHIFT);
+	if (area_len < len)
+		return -EINVAL;
+	vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
+	mem_off = shm_off_to_phys_addr(area_off);
+	if (io_remap_pfn_range(vma, vma->vm_start, mem_off >> PAGE_SHIFT,
+			       len, vma->vm_page_prot))
+		return -EAGAIN;
+	return 0;
+}
+
+module_init(vsoc_init_module);
+module_exit(vsoc_cleanup_module);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Greg Hartman <ghartman@xxxxxxxxxx>");
+MODULE_DESCRIPTION("VSoC interpretation of QEmu's ivshmem device");
+MODULE_VERSION("1.0");
-- 
2.17.0.484.g0c8726318c-goog