Merge tag 'trace-v7.3-2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-tracegrafted

Pull tracing fixes from Steven Rostedt:

 - Fix error output of boot instance creation failure

   Currently if a boot instance creation fails, instead of printing out
   the name of the instance that failed, it prints "(null)". That is
   because it prints "cur_str" that had already been processed by
   strsep(). Print the saved name instead.

   While at it, print the error code of the failure.

 - Fix use-after-free for same named historgrams

   Histograms can be named so that they can be used in multiple events.
   But if the named histogram has a variable attached, the second event
   that uses the named histogram which duplicates it and needs to free
   the original after duplication leaves the old variable in place and
   still visible. If another histogram uses than variable, it will use
   the stale one which will try to reference the freed duplicate
   histogram and crash the kernel.

   Free the duplicate variables along with the duplicated histogram
   data.

 - Check return value of kthread_run() in event self test

   The events self tests uses a kthread for testing but does not check
   if it succeeded in creating a kthread. If the kthread creation were
   to fail, the code will still try to call kthread_stop() on the error
   returned.

 - Fix race between reading trace_pipe and updating subbuffer size

   If a user is reading the trace_pipe file at the same time they update
   the ring buffer sub-buffer size, can cause the trace_pipe read to
   read stale data. Add trace_access_lock() around updating the ring
   buffer sub-buffer size.

 - Fix eventfs_inode on failure path in creation of the events directory

   In the creation of the "events" directory, if after allocating the
   eventfs_inode a failure is detected, it calls cleanup_ei() which
   calls free_ei(). The free_ei() will test if eventfs_inode being freed
   has no children. It is a bug if it does. But on the failure case of
   the creation of the "events" directory, the children lists have not
   yet been initialized and the free will trigger a warning because
   list_empty() on an uninitialized list returns false.

   Move the initialization into init_ei() where it makes more sense and
   makes sure that a created eventfs_inode has its lists initialized
   upon creation.

 - Check return value of kthread_run() in ftrace direct sample code

   The sample code that shows how to use the ftrace direct calls does
   not test the return of kthread_run() to see if it succeeds. Return a
   failure if the kthread_run() doesn't succeed.

 - Clear user events state on fork in case of alloc failure

   On fork, the child gets a pointer to the parent's user events state.
   It makes a copy of it then updates the child's pointer to it. But if
   the allocation fails, the duplication function leaves the child with
   a pointer to its parent's descriptor. When the child cleans up its
   data, it will free the parent's descriptor while the parent is still
   using it.

   In the duplication function, set the child's user_event_mm to NULL
   before testing if the allocation succeeded, and when it exits it will
   not free the parent's descriptor.

 - Fix retry exhaustion in simple ring buffer reader swap

   simple_ring_buffer_swap_reader_page() starts with retry set to 8 and
   post-decrements it only after a failed link replacement. On the final
   attempt, a successful replacement leaves retry at zero, while a
   failed replacement leaves it at -1.

   But the check for success expects the retry value to be non-zero and
   exits with an error on zero. This is the opposite result. Fix it.

 - Fail nicely when the remote swap_reader_page() returns an error

   Currently, if the swap_reader_page() of a remote buffer fails, it
   triggers a WARN_ON_ONCE() and continues normally. Instead, have it
   exit with an error and a pr_warn() print instead of a full WARNING.

* tag 'trace-v7.3-2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
  ring-buffer: Stop remote reader update when page swap fails
  tracing: Fix retry exhaustion in simple ring buffer reader swap
  tracing/user_events: Clear copied tracing state before fork duplication
  samples/ftrace: Fix kthread_stop() on ERR_PTR in ftrace-direct-multi-modify
  samples/ftrace: Fix kthread_stop() on ERR_PTR in ftrace-direct-modify
  eventfs: Initialize ei->children and ei->list in init_ei()
  tracing: Fix use-after-free in trace_pipe read on sub-buffer order change
  tracing: Fix crash passing ERR_PTR to kthread_stop()
  tracing: Fix use-after-free with same-name named triggers
  tracing: Fix logged instance name on creation failure

7 files changed, 4589 insertions, 0 deletions

diff --git a/samples/vfio-mdev/Makefile b/samples/vfio-mdev/Makefile
new file mode 100644
index 000000000..10d179c4f
--- /dev/null
+++ b/samples/vfio-mdev/Makefile
@@ -0,0 +1,5 @@
+# SPDX-License-Identifier: GPL-2.0-only
+obj-$(CONFIG_SAMPLE_VFIO_MDEV_MTTY) += mtty.o
+obj-$(CONFIG_SAMPLE_VFIO_MDEV_MDPY) += mdpy.o
+obj-$(CONFIG_SAMPLE_VFIO_MDEV_MDPY_FB) += mdpy-fb.o
+obj-$(CONFIG_SAMPLE_VFIO_MDEV_MBOCHS) += mbochs.o
diff --git a/samples/vfio-mdev/README.rst b/samples/vfio-mdev/README.rst
new file mode 100644
index 000000000..b52eb3773
--- /dev/null
+++ b/samples/vfio-mdev/README.rst
@@ -0,0 +1,100 @@
+Using the mtty vfio-mdev sample code
+====================================
+
+mtty is a sample vfio-mdev driver that demonstrates how to use the mediated
+device framework.
+
+The sample driver creates an mdev device that simulates a serial port over a PCI
+card.
+
+1. Build and load the mtty.ko module.
+
+   This step creates a dummy device, /sys/devices/virtual/mtty/mtty/
+
+   Files in this device directory in sysfs are similar to the following::
+
+     # tree /sys/devices/virtual/mtty/mtty/
+        /sys/devices/virtual/mtty/mtty/
+        |-- mdev_supported_types
+        |   |-- mtty-1
+        |   |   |-- available_instances
+        |   |   |-- create
+        |   |   |-- device_api
+        |   |   |-- devices
+        |   |   `-- name
+        |   `-- mtty-2
+        |       |-- available_instances
+        |       |-- create
+        |       |-- device_api
+        |       |-- devices
+        |       `-- name
+        |-- mtty_dev
+        |   `-- sample_mtty_dev
+        |-- power
+        |   |-- autosuspend_delay_ms
+        |   |-- control
+        |   |-- runtime_active_time
+        |   |-- runtime_status
+        |   `-- runtime_suspended_time
+        |-- subsystem -> ../../../../class/mtty
+        `-- uevent
+
+2. Create a mediated device by using the dummy device that you created in the
+   previous step::
+
+     # echo "83b8f4f2-509f-382f-3c1e-e6bfe0fa1001" >	\
+              /sys/devices/virtual/mtty/mtty/mdev_supported_types/mtty-2/create
+
+3. Add parameters to qemu-kvm::
+
+     -device vfio-pci,\
+      sysfsdev=/sys/bus/mdev/devices/83b8f4f2-509f-382f-3c1e-e6bfe0fa1001
+
+4. Boot the VM.
+
+   In the Linux guest VM, with no hardware on the host, the device appears
+   as  follows::
+
+     # lspci -s 00:05.0 -xxvv
+     00:05.0 Serial controller: Device 4348:3253 (rev 10) (prog-if 02 [16550])
+             Subsystem: Device 4348:3253
+             Physical Slot: 5
+             Control: I/O+ Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr-
+     Stepping- SERR- FastB2B- DisINTx-
+             Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort-
+     <TAbort- <MAbort- >SERR- <PERR- INTx-
+             Interrupt: pin A routed to IRQ 10
+             Region 0: I/O ports at c150 [size=8]
+             Region 1: I/O ports at c158 [size=8]
+             Kernel driver in use: serial
+     00: 48 43 53 32 01 00 00 02 10 02 00 07 00 00 00 00
+     10: 51 c1 00 00 59 c1 00 00 00 00 00 00 00 00 00 00
+     20: 00 00 00 00 00 00 00 00 00 00 00 00 48 43 53 32
+     30: 00 00 00 00 00 00 00 00 00 00 00 00 0a 01 00 00
+
+     In the Linux guest VM, dmesg output for the device is as follows:
+
+     serial 0000:00:05.0: PCI INT A -> Link[LNKA] -> GSI 10 (level, high) -> IRQ 10
+     0000:00:05.0: ttyS1 at I/O 0xc150 (irq = 10) is a 16550A
+     0000:00:05.0: ttyS2 at I/O 0xc158 (irq = 10) is a 16550A
+
+
+5. In the Linux guest VM, check the serial ports::
+
+     # setserial -g /dev/ttyS*
+     /dev/ttyS0, UART: 16550A, Port: 0x03f8, IRQ: 4
+     /dev/ttyS1, UART: 16550A, Port: 0xc150, IRQ: 10
+     /dev/ttyS2, UART: 16550A, Port: 0xc158, IRQ: 10
+
+6. Using minicom or any terminal emulation program, open port /dev/ttyS1 or
+   /dev/ttyS2 with hardware flow control disabled.
+
+7. Type data on the minicom terminal or send data to the terminal emulation
+   program and read the data.
+
+   Data is loop backed from hosts mtty driver.
+
+8. Destroy the mediated device that you created::
+
+     # echo 1 > /sys/bus/mdev/devices/83b8f4f2-509f-382f-3c1e-e6bfe0fa1001/remove
+
diff --git a/samples/vfio-mdev/mbochs.c b/samples/vfio-mdev/mbochs.c
new file mode 100644
index 000000000..64ea19253
--- /dev/null
+++ b/samples/vfio-mdev/mbochs.c
@@ -0,0 +1,1451 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Mediated virtual PCI display host device driver
+ *
+ * Emulate enough of qemu stdvga to make bochs-drm.ko happy.  That is
+ * basically the vram memory bar and the bochs dispi interface vbe
+ * registers in the mmio register bar.	Specifically it does *not*
+ * include any legacy vga stuff.  Device looks a lot like "qemu -device
+ * secondary-vga".
+ *
+ *   (c) Gerd Hoffmann <kraxel@redhat.com>
+ *
+ * based on mtty driver which is:
+ *   Copyright (c) 2016, NVIDIA CORPORATION. All rights reserved.
+ *	 Author: Neo Jia <cjia@nvidia.com>
+ *		 Kirti Wankhede <kwankhede@nvidia.com>
+ *
+ * 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.
+ */
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/vmalloc.h>
+#include <linux/cdev.h>
+#include <linux/vfio.h>
+#include <linux/iommu.h>
+#include <linux/sysfs.h>
+#include <linux/mdev.h>
+#include <linux/pci.h>
+#include <linux/dma-buf.h>
+#include <linux/highmem.h>
+#include <drm/drm_fourcc.h>
+#include <drm/drm_rect.h>
+#include <drm/drm_modeset_lock.h>
+#include <drm/drm_property.h>
+#include <drm/drm_plane.h>
+
+
+#define VBE_DISPI_INDEX_ID		0x0
+#define VBE_DISPI_INDEX_XRES		0x1
+#define VBE_DISPI_INDEX_YRES		0x2
+#define VBE_DISPI_INDEX_BPP		0x3
+#define VBE_DISPI_INDEX_ENABLE		0x4
+#define VBE_DISPI_INDEX_BANK		0x5
+#define VBE_DISPI_INDEX_VIRT_WIDTH	0x6
+#define VBE_DISPI_INDEX_VIRT_HEIGHT	0x7
+#define VBE_DISPI_INDEX_X_OFFSET	0x8
+#define VBE_DISPI_INDEX_Y_OFFSET	0x9
+#define VBE_DISPI_INDEX_VIDEO_MEMORY_64K 0xa
+#define VBE_DISPI_INDEX_COUNT		0xb
+
+#define VBE_DISPI_ID0			0xB0C0
+#define VBE_DISPI_ID1			0xB0C1
+#define VBE_DISPI_ID2			0xB0C2
+#define VBE_DISPI_ID3			0xB0C3
+#define VBE_DISPI_ID4			0xB0C4
+#define VBE_DISPI_ID5			0xB0C5
+
+#define VBE_DISPI_DISABLED		0x00
+#define VBE_DISPI_ENABLED		0x01
+#define VBE_DISPI_GETCAPS		0x02
+#define VBE_DISPI_8BIT_DAC		0x20
+#define VBE_DISPI_LFB_ENABLED		0x40
+#define VBE_DISPI_NOCLEARMEM		0x80
+
+
+#define MBOCHS_NAME		  "mbochs"
+#define MBOCHS_CLASS_NAME	  "mbochs"
+
+#define MBOCHS_EDID_REGION_INDEX  VFIO_PCI_NUM_REGIONS
+#define MBOCHS_NUM_REGIONS        (MBOCHS_EDID_REGION_INDEX+1)
+
+#define MBOCHS_CONFIG_SPACE_SIZE  0xff
+#define MBOCHS_MMIO_BAR_OFFSET	  PAGE_SIZE
+#define MBOCHS_MMIO_BAR_SIZE	  PAGE_SIZE
+#define MBOCHS_EDID_OFFSET	  (MBOCHS_MMIO_BAR_OFFSET +	\
+				   MBOCHS_MMIO_BAR_SIZE)
+#define MBOCHS_EDID_SIZE	  PAGE_SIZE
+#define MBOCHS_MEMORY_BAR_OFFSET  (MBOCHS_EDID_OFFSET + \
+				   MBOCHS_EDID_SIZE)
+
+#define MBOCHS_EDID_BLOB_OFFSET   (MBOCHS_EDID_SIZE/2)
+
+#define STORE_LE16(addr, val)	(*(u16 *)addr = val)
+#define STORE_LE32(addr, val)	(*(u32 *)addr = val)
+
+
+MODULE_DESCRIPTION("Mediated virtual PCI display host device driver");
+MODULE_LICENSE("GPL v2");
+
+static int max_mbytes = 256;
+module_param_named(count, max_mbytes, int, 0444);
+MODULE_PARM_DESC(mem, "megabytes available to " MBOCHS_NAME " devices");
+
+
+#define MBOCHS_TYPE_1 "small"
+#define MBOCHS_TYPE_2 "medium"
+#define MBOCHS_TYPE_3 "large"
+
+static struct mbochs_type {
+	struct mdev_type type;
+	u32 mbytes;
+	u32 max_x;
+	u32 max_y;
+} mbochs_types[] = {
+	{
+		.type.sysfs_name	= MBOCHS_TYPE_1,
+		.type.pretty_name	= MBOCHS_CLASS_NAME "-" MBOCHS_TYPE_1,
+		.mbytes = 4,
+		.max_x  = 800,
+		.max_y  = 600,
+	}, {
+		.type.sysfs_name	= MBOCHS_TYPE_2,
+		.type.pretty_name	= MBOCHS_CLASS_NAME "-" MBOCHS_TYPE_2,
+		.mbytes = 16,
+		.max_x  = 1920,
+		.max_y  = 1440,
+	}, {
+		.type.sysfs_name	= MBOCHS_TYPE_3,
+		.type.pretty_name	= MBOCHS_CLASS_NAME "-" MBOCHS_TYPE_3,
+		.mbytes = 64,
+		.max_x  = 0,
+		.max_y  = 0,
+	},
+};
+
+static struct mdev_type *mbochs_mdev_types[] = {
+	&mbochs_types[0].type,
+	&mbochs_types[1].type,
+	&mbochs_types[2].type,
+};
+
+static dev_t		mbochs_devt;
+static const struct class mbochs_class = {
+	.name = MBOCHS_CLASS_NAME,
+};
+static struct cdev	mbochs_cdev;
+static struct device	mbochs_dev;
+static struct mdev_parent mbochs_parent;
+static atomic_t mbochs_avail_mbytes;
+static const struct vfio_device_ops mbochs_dev_ops;
+
+struct mbochs_mode {
+	u32 drm_format;
+	u32 bytepp;
+	u32 width;
+	u32 height;
+	u32 stride;
+	u32 __pad;
+	u64 offset;
+	u64 size;
+};
+
+struct mbochs_dmabuf {
+	struct mbochs_mode mode;
+	u32 id;
+	struct page **pages;
+	pgoff_t pagecount;
+	struct dma_buf *buf;
+	struct mdev_state *mdev_state;
+	struct list_head next;
+	bool unlinked;
+};
+
+/* State of each mdev device */
+struct mdev_state {
+	struct vfio_device vdev;
+	u8 *vconfig;
+	u64 bar_mask[3];
+	u32 memory_bar_mask;
+	struct mutex ops_lock;
+	struct mdev_device *mdev;
+
+	const struct mbochs_type *type;
+	u16 vbe[VBE_DISPI_INDEX_COUNT];
+	u64 memsize;
+	struct page **pages;
+	pgoff_t pagecount;
+	struct vfio_region_gfx_edid edid_regs;
+	u8 edid_blob[0x400];
+
+	struct list_head dmabufs;
+	u32 active_id;
+	u32 next_id;
+};
+
+static const char *vbe_name_list[VBE_DISPI_INDEX_COUNT] = {
+	[VBE_DISPI_INDEX_ID]               = "id",
+	[VBE_DISPI_INDEX_XRES]             = "xres",
+	[VBE_DISPI_INDEX_YRES]             = "yres",
+	[VBE_DISPI_INDEX_BPP]              = "bpp",
+	[VBE_DISPI_INDEX_ENABLE]           = "enable",
+	[VBE_DISPI_INDEX_BANK]             = "bank",
+	[VBE_DISPI_INDEX_VIRT_WIDTH]       = "virt-width",
+	[VBE_DISPI_INDEX_VIRT_HEIGHT]      = "virt-height",
+	[VBE_DISPI_INDEX_X_OFFSET]         = "x-offset",
+	[VBE_DISPI_INDEX_Y_OFFSET]         = "y-offset",
+	[VBE_DISPI_INDEX_VIDEO_MEMORY_64K] = "video-mem",
+};
+
+static const char *vbe_name(u32 index)
+{
+	if (index < ARRAY_SIZE(vbe_name_list))
+		return vbe_name_list[index];
+	return "(invalid)";
+}
+
+static struct page *__mbochs_get_page(struct mdev_state *mdev_state,
+				      pgoff_t pgoff);
+static struct page *mbochs_get_page(struct mdev_state *mdev_state,
+				    pgoff_t pgoff);
+
+static void mbochs_create_config_space(struct mdev_state *mdev_state)
+{
+	STORE_LE16((u16 *) &mdev_state->vconfig[PCI_VENDOR_ID],
+		   0x1234);
+	STORE_LE16((u16 *) &mdev_state->vconfig[PCI_DEVICE_ID],
+		   0x1111);
+	STORE_LE16((u16 *) &mdev_state->vconfig[PCI_SUBSYSTEM_VENDOR_ID],
+		   PCI_SUBVENDOR_ID_REDHAT_QUMRANET);
+	STORE_LE16((u16 *) &mdev_state->vconfig[PCI_SUBSYSTEM_ID],
+		   PCI_SUBDEVICE_ID_QEMU);
+
+	STORE_LE16((u16 *) &mdev_state->vconfig[PCI_COMMAND],
+		   PCI_COMMAND_IO | PCI_COMMAND_MEMORY);
+	STORE_LE16((u16 *) &mdev_state->vconfig[PCI_CLASS_DEVICE],
+		   PCI_CLASS_DISPLAY_OTHER);
+	mdev_state->vconfig[PCI_CLASS_REVISION] =  0x01;
+
+	STORE_LE32((u32 *) &mdev_state->vconfig[PCI_BASE_ADDRESS_0],
+		   PCI_BASE_ADDRESS_SPACE_MEMORY |
+		   PCI_BASE_ADDRESS_MEM_TYPE_32	 |
+		   PCI_BASE_ADDRESS_MEM_PREFETCH);
+	mdev_state->bar_mask[0] = ~(mdev_state->memsize) + 1;
+
+	STORE_LE32((u32 *) &mdev_state->vconfig[PCI_BASE_ADDRESS_2],
+		   PCI_BASE_ADDRESS_SPACE_MEMORY |
+		   PCI_BASE_ADDRESS_MEM_TYPE_32);
+	mdev_state->bar_mask[2] = ~(MBOCHS_MMIO_BAR_SIZE) + 1;
+}
+
+static int mbochs_check_framebuffer(struct mdev_state *mdev_state,
+				    struct mbochs_mode *mode)
+{
+	struct device *dev = mdev_dev(mdev_state->mdev);
+	u16 *vbe = mdev_state->vbe;
+	u32 virt_width;
+
+	WARN_ON(!mutex_is_locked(&mdev_state->ops_lock));
+
+	if (!(vbe[VBE_DISPI_INDEX_ENABLE] & VBE_DISPI_ENABLED))
+		goto nofb;
+
+	memset(mode, 0, sizeof(*mode));
+	switch (vbe[VBE_DISPI_INDEX_BPP]) {
+	case 32:
+		mode->drm_format = DRM_FORMAT_XRGB8888;
+		mode->bytepp = 4;
+		break;
+	default:
+		dev_info_ratelimited(dev, "%s: bpp %d not supported\n",
+				     __func__, vbe[VBE_DISPI_INDEX_BPP]);
+		goto nofb;
+	}
+
+	mode->width  = vbe[VBE_DISPI_INDEX_XRES];
+	mode->height = vbe[VBE_DISPI_INDEX_YRES];
+	virt_width  = vbe[VBE_DISPI_INDEX_VIRT_WIDTH];
+	if (virt_width < mode->width)
+		virt_width = mode->width;
+	mode->stride = virt_width * mode->bytepp;
+	mode->size   = (u64)mode->stride * mode->height;
+	mode->offset = ((u64)vbe[VBE_DISPI_INDEX_X_OFFSET] * mode->bytepp +
+		       (u64)vbe[VBE_DISPI_INDEX_Y_OFFSET] * mode->stride);
+
+	if (mode->width < 64 || mode->height < 64) {
+		dev_info_ratelimited(dev, "%s: invalid resolution %dx%d\n",
+				     __func__, mode->width, mode->height);
+		goto nofb;
+	}
+	if (mode->offset + mode->size > mdev_state->memsize) {
+		dev_info_ratelimited(dev, "%s: framebuffer memory overflow\n",
+				     __func__);
+		goto nofb;
+	}
+
+	return 0;
+
+nofb:
+	memset(mode, 0, sizeof(*mode));
+	return -EINVAL;
+}
+
+static bool mbochs_modes_equal(struct mbochs_mode *mode1,
+			       struct mbochs_mode *mode2)
+{
+	return memcmp(mode1, mode2, sizeof(struct mbochs_mode)) == 0;
+}
+
+static void handle_pci_cfg_write(struct mdev_state *mdev_state, u16 offset,
+				 char *buf, u32 count)
+{
+	struct device *dev = mdev_dev(mdev_state->mdev);
+	int index = (offset - PCI_BASE_ADDRESS_0) / 0x04;
+	u32 cfg_addr;
+
+	switch (offset) {
+	case PCI_BASE_ADDRESS_0:
+	case PCI_BASE_ADDRESS_2:
+		cfg_addr = *(u32 *)buf;
+
+		if (cfg_addr == 0xffffffff) {
+			cfg_addr = (cfg_addr & mdev_state->bar_mask[index]);
+		} else {
+			cfg_addr &= PCI_BASE_ADDRESS_MEM_MASK;
+			if (cfg_addr)
+				dev_info(dev, "BAR #%d @ 0x%x\n",
+					 index, cfg_addr);
+		}
+
+		cfg_addr |= (mdev_state->vconfig[offset] &
+			     ~PCI_BASE_ADDRESS_MEM_MASK);
+		STORE_LE32(&mdev_state->vconfig[offset], cfg_addr);
+		break;
+	}
+}
+
+static void handle_mmio_write(struct mdev_state *mdev_state, u16 offset,
+			      char *buf, u32 count)
+{
+	struct device *dev = mdev_dev(mdev_state->mdev);
+	int index;
+	u16 reg16;
+
+	switch (offset) {
+	case 0x400 ... 0x41f: /* vga ioports remapped */
+		goto unhandled;
+	case 0x500 ... 0x515: /* bochs dispi interface */
+		if (count != 2)
+			goto unhandled;
+		index = (offset - 0x500) / 2;
+		reg16 = *(u16 *)buf;
+		if (index < ARRAY_SIZE(mdev_state->vbe))
+			mdev_state->vbe[index] = reg16;
+		dev_dbg(dev, "%s: vbe write %d = %d (%s)\n",
+			__func__, index, reg16, vbe_name(index));
+		break;
+	case 0x600 ... 0x607: /* qemu extended regs */
+		goto unhandled;
+	default:
+unhandled:
+		dev_dbg(dev, "%s: @0x%03x, count %d (unhandled)\n",
+			__func__, offset, count);
+		break;
+	}
+}
+
+static void handle_mmio_read(struct mdev_state *mdev_state, u16 offset,
+			     char *buf, u32 count)
+{
+	struct device *dev = mdev_dev(mdev_state->mdev);
+	struct vfio_region_gfx_edid *edid;
+	u16 reg16 = 0;
+	int index;
+
+	switch (offset) {
+	case 0x000 ... 0x3ff: /* edid block */
+		edid = &mdev_state->edid_regs;
+		if (edid->link_state != VFIO_DEVICE_GFX_LINK_STATE_UP ||
+		    offset >= edid->edid_size) {
+			memset(buf, 0, count);
+			break;
+		}
+		memcpy(buf, mdev_state->edid_blob + offset, count);
+		break;
+	case 0x500 ... 0x515: /* bochs dispi interface */
+		if (count != 2)
+			goto unhandled;
+		index = (offset - 0x500) / 2;
+		if (index < ARRAY_SIZE(mdev_state->vbe))
+			reg16 = mdev_state->vbe[index];
+		dev_dbg(dev, "%s: vbe read %d = %d (%s)\n",
+			__func__, index, reg16, vbe_name(index));
+		*(u16 *)buf = reg16;
+		break;
+	default:
+unhandled:
+		dev_dbg(dev, "%s: @0x%03x, count %d (unhandled)\n",
+			__func__, offset, count);
+		memset(buf, 0, count);
+		break;
+	}
+}
+
+static void handle_edid_regs(struct mdev_state *mdev_state, u16 offset,
+			     char *buf, u32 count, bool is_write)
+{
+	char *regs = (void *)&mdev_state->edid_regs;
+
+	if (offset + count > sizeof(mdev_state->edid_regs))
+		return;
+	if (count != 4)
+		return;
+	if (offset % 4)
+		return;
+
+	if (is_write) {
+		switch (offset) {
+		case offsetof(struct vfio_region_gfx_edid, link_state):
+		case offsetof(struct vfio_region_gfx_edid, edid_size):
+			memcpy(regs + offset, buf, count);
+			break;
+		default:
+			/* read-only regs */
+			break;
+		}
+	} else {
+		memcpy(buf, regs + offset, count);
+	}
+}
+
+static void handle_edid_blob(struct mdev_state *mdev_state, u16 offset,
+			     char *buf, u32 count, bool is_write)
+{
+	if (offset + count > mdev_state->edid_regs.edid_max_size)
+		return;
+	if (is_write)
+		memcpy(mdev_state->edid_blob + offset, buf, count);
+	else
+		memcpy(buf, mdev_state->edid_blob + offset, count);
+}
+
+static ssize_t mdev_access(struct mdev_state *mdev_state, char *buf,
+			   size_t count, loff_t pos, bool is_write)
+{
+	struct page *pg;
+	loff_t poff;
+	char *map;
+	int ret = 0;
+
+	mutex_lock(&mdev_state->ops_lock);
+
+	if (pos < MBOCHS_CONFIG_SPACE_SIZE) {
+		if (is_write)
+			handle_pci_cfg_write(mdev_state, pos, buf, count);
+		else
+			memcpy(buf, (mdev_state->vconfig + pos), count);
+
+	} else if (pos >= MBOCHS_MMIO_BAR_OFFSET &&
+		   pos + count <= (MBOCHS_MMIO_BAR_OFFSET +
+				   MBOCHS_MMIO_BAR_SIZE)) {
+		pos -= MBOCHS_MMIO_BAR_OFFSET;
+		if (is_write)
+			handle_mmio_write(mdev_state, pos, buf, count);
+		else
+			handle_mmio_read(mdev_state, pos, buf, count);
+
+	} else if (pos >= MBOCHS_EDID_OFFSET &&
+		   pos + count <= (MBOCHS_EDID_OFFSET +
+				   MBOCHS_EDID_SIZE)) {
+		pos -= MBOCHS_EDID_OFFSET;
+		if (pos < MBOCHS_EDID_BLOB_OFFSET) {
+			handle_edid_regs(mdev_state, pos, buf, count, is_write);
+		} else {
+			pos -= MBOCHS_EDID_BLOB_OFFSET;
+			handle_edid_blob(mdev_state, pos, buf, count, is_write);
+		}
+
+	} else if (pos >= MBOCHS_MEMORY_BAR_OFFSET &&
+		   pos + count <=
+		   MBOCHS_MEMORY_BAR_OFFSET + mdev_state->memsize) {
+		pos -= MBOCHS_MMIO_BAR_OFFSET;
+		poff = pos & ~PAGE_MASK;
+		pg = __mbochs_get_page(mdev_state, pos >> PAGE_SHIFT);
+		map = kmap(pg);
+		if (is_write)
+			memcpy(map + poff, buf, count);
+		else
+			memcpy(buf, map + poff, count);
+		kunmap(pg);
+		put_page(pg);
+
+	} else {
+		dev_dbg(mdev_state->vdev.dev, "%s: %s @0x%llx (unhandled)\n",
+			__func__, is_write ? "WR" : "RD", pos);
+		ret = -1;
+		goto accessfailed;
+	}
+
+	ret = count;
+
+
+accessfailed:
+	mutex_unlock(&mdev_state->ops_lock);
+
+	return ret;
+}
+
+static int mbochs_reset(struct mdev_state *mdev_state)
+{
+	u32 size64k = mdev_state->memsize / (64 * 1024);
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(mdev_state->vbe); i++)
+		mdev_state->vbe[i] = 0;
+	mdev_state->vbe[VBE_DISPI_INDEX_ID] = VBE_DISPI_ID5;
+	mdev_state->vbe[VBE_DISPI_INDEX_VIDEO_MEMORY_64K] = size64k;
+	return 0;
+}
+
+static int mbochs_init_dev(struct vfio_device *vdev)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	struct mdev_device *mdev = to_mdev_device(vdev->dev);
+	struct mbochs_type *type =
+		container_of(mdev->type, struct mbochs_type, type);
+	int avail_mbytes = atomic_read(&mbochs_avail_mbytes);
+	int ret = -ENOMEM;
+
+	do {
+		if (avail_mbytes < type->mbytes)
+			return -ENOSPC;
+	} while (!atomic_try_cmpxchg(&mbochs_avail_mbytes, &avail_mbytes,
+				     avail_mbytes - type->mbytes));
+
+	mdev_state->vconfig = kzalloc(MBOCHS_CONFIG_SPACE_SIZE, GFP_KERNEL);
+	if (!mdev_state->vconfig)
+		goto err_avail;
+
+	mdev_state->memsize = type->mbytes * 1024 * 1024;
+	mdev_state->pagecount = mdev_state->memsize >> PAGE_SHIFT;
+	mdev_state->pages = kcalloc(mdev_state->pagecount,
+				    sizeof(struct page *),
+				    GFP_KERNEL);
+	if (!mdev_state->pages)
+		goto err_vconfig;
+
+	mutex_init(&mdev_state->ops_lock);
+	mdev_state->mdev = mdev;
+	INIT_LIST_HEAD(&mdev_state->dmabufs);
+	mdev_state->next_id = 1;
+
+	mdev_state->type = type;
+	mdev_state->edid_regs.max_xres = type->max_x;
+	mdev_state->edid_regs.max_yres = type->max_y;
+	mdev_state->edid_regs.edid_offset = MBOCHS_EDID_BLOB_OFFSET;
+	mdev_state->edid_regs.edid_max_size = sizeof(mdev_state->edid_blob);
+	mbochs_create_config_space(mdev_state);
+	mbochs_reset(mdev_state);
+
+	dev_info(vdev->dev, "%s: %s, %d MB, %ld pages\n", __func__,
+		 type->type.pretty_name, type->mbytes, mdev_state->pagecount);
+	return 0;
+
+err_vconfig:
+	kfree(mdev_state->vconfig);
+err_avail:
+	atomic_add(type->mbytes, &mbochs_avail_mbytes);
+	return ret;
+}
+
+static int mbochs_probe(struct mdev_device *mdev)
+{
+	struct mdev_state *mdev_state;
+	int ret = -ENOMEM;
+
+	mdev_state = vfio_alloc_device(mdev_state, vdev, &mdev->dev,
+				       &mbochs_dev_ops);
+	if (IS_ERR(mdev_state))
+		return PTR_ERR(mdev_state);
+
+	ret = vfio_register_emulated_iommu_dev(&mdev_state->vdev);
+	if (ret)
+		goto err_put_vdev;
+	dev_set_drvdata(&mdev->dev, mdev_state);
+	return 0;
+
+err_put_vdev:
+	vfio_put_device(&mdev_state->vdev);
+	return ret;
+}
+
+static void mbochs_release_dev(struct vfio_device *vdev)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+
+	atomic_add(mdev_state->type->mbytes, &mbochs_avail_mbytes);
+	kfree(mdev_state->pages);
+	kfree(mdev_state->vconfig);
+}
+
+static void mbochs_remove(struct mdev_device *mdev)
+{
+	struct mdev_state *mdev_state = dev_get_drvdata(&mdev->dev);
+
+	vfio_unregister_group_dev(&mdev_state->vdev);
+	vfio_put_device(&mdev_state->vdev);
+}
+
+static ssize_t mbochs_read(struct vfio_device *vdev, char __user *buf,
+			   size_t count, loff_t *ppos)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	unsigned int done = 0;
+	int ret;
+
+	while (count) {
+		size_t filled;
+
+		if (count >= 4 && !(*ppos % 4)) {
+			u32 val;
+
+			ret =  mdev_access(mdev_state, (char *)&val, sizeof(val),
+					   *ppos, false);
+			if (ret <= 0)
+				goto read_err;
+
+			if (copy_to_user(buf, &val, sizeof(val)))
+				goto read_err;
+
+			filled = 4;
+		} else if (count >= 2 && !(*ppos % 2)) {
+			u16 val;
+
+			ret = mdev_access(mdev_state, (char *)&val, sizeof(val),
+					  *ppos, false);
+			if (ret <= 0)
+				goto read_err;
+
+			if (copy_to_user(buf, &val, sizeof(val)))
+				goto read_err;
+
+			filled = 2;
+		} else {
+			u8 val;
+
+			ret = mdev_access(mdev_state, (char *)&val, sizeof(val),
+					  *ppos, false);
+			if (ret <= 0)
+				goto read_err;
+
+			if (copy_to_user(buf, &val, sizeof(val)))
+				goto read_err;
+
+			filled = 1;
+		}
+
+		count -= filled;
+		done += filled;
+		*ppos += filled;
+		buf += filled;
+	}
+
+	return done;
+
+read_err:
+	return -EFAULT;
+}
+
+static ssize_t mbochs_write(struct vfio_device *vdev, const char __user *buf,
+			    size_t count, loff_t *ppos)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	unsigned int done = 0;
+	int ret;
+
+	while (count) {
+		size_t filled;
+
+		if (count >= 4 && !(*ppos % 4)) {
+			u32 val;
+
+			if (copy_from_user(&val, buf, sizeof(val)))
+				goto write_err;
+
+			ret = mdev_access(mdev_state, (char *)&val, sizeof(val),
+					  *ppos, true);
+			if (ret <= 0)
+				goto write_err;
+
+			filled = 4;
+		} else if (count >= 2 && !(*ppos % 2)) {
+			u16 val;
+
+			if (copy_from_user(&val, buf, sizeof(val)))
+				goto write_err;
+
+			ret = mdev_access(mdev_state, (char *)&val, sizeof(val),
+					  *ppos, true);
+			if (ret <= 0)
+				goto write_err;
+
+			filled = 2;
+		} else {
+			u8 val;
+
+			if (copy_from_user(&val, buf, sizeof(val)))
+				goto write_err;
+
+			ret = mdev_access(mdev_state, (char *)&val, sizeof(val),
+					  *ppos, true);
+			if (ret <= 0)
+				goto write_err;
+
+			filled = 1;
+		}
+		count -= filled;
+		done += filled;
+		*ppos += filled;
+		buf += filled;
+	}
+
+	return done;
+write_err:
+	return -EFAULT;
+}
+
+static struct page *__mbochs_get_page(struct mdev_state *mdev_state,
+				      pgoff_t pgoff)
+{
+	WARN_ON(!mutex_is_locked(&mdev_state->ops_lock));
+
+	if (!mdev_state->pages[pgoff]) {
+		mdev_state->pages[pgoff] =
+			alloc_pages(GFP_HIGHUSER | __GFP_ZERO, 0);
+		if (!mdev_state->pages[pgoff])
+			return NULL;
+	}
+
+	get_page(mdev_state->pages[pgoff]);
+	return mdev_state->pages[pgoff];
+}
+
+static struct page *mbochs_get_page(struct mdev_state *mdev_state,
+				    pgoff_t pgoff)
+{
+	struct page *page;
+
+	if (WARN_ON(pgoff >= mdev_state->pagecount))
+		return NULL;
+
+	mutex_lock(&mdev_state->ops_lock);
+	page = __mbochs_get_page(mdev_state, pgoff);
+	mutex_unlock(&mdev_state->ops_lock);
+
+	return page;
+}
+
+static void mbochs_put_pages(struct mdev_state *mdev_state)
+{
+	struct device *dev = mdev_dev(mdev_state->mdev);
+	int i, count = 0;
+
+	WARN_ON(!mutex_is_locked(&mdev_state->ops_lock));
+
+	for (i = 0; i < mdev_state->pagecount; i++) {
+		if (!mdev_state->pages[i])
+			continue;
+		put_page(mdev_state->pages[i]);
+		mdev_state->pages[i] = NULL;
+		count++;
+	}
+	dev_dbg(dev, "%s: %d pages released\n", __func__, count);
+}
+
+static vm_fault_t mbochs_region_vm_fault(struct vm_fault *vmf)
+{
+	struct vm_area_struct *vma = vmf->vma;
+	struct mdev_state *mdev_state = vma->vm_private_data;
+	pgoff_t page_offset = (vmf->address - vma->vm_start) >> PAGE_SHIFT;
+
+	if (page_offset >= mdev_state->pagecount)
+		return VM_FAULT_SIGBUS;
+
+	vmf->page = mbochs_get_page(mdev_state, page_offset);
+	if (!vmf->page)
+		return VM_FAULT_SIGBUS;
+
+	return 0;
+}
+
+static const struct vm_operations_struct mbochs_region_vm_ops = {
+	.fault = mbochs_region_vm_fault,
+};
+
+static int mbochs_mmap(struct vfio_device *vdev, struct vm_area_struct *vma)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+
+	if (vma->vm_pgoff != MBOCHS_MEMORY_BAR_OFFSET >> PAGE_SHIFT)
+		return -EINVAL;
+	if (vma->vm_end < vma->vm_start)
+		return -EINVAL;
+	if (vma->vm_end - vma->vm_start > mdev_state->memsize)
+		return -EINVAL;
+	if ((vma->vm_flags & VM_SHARED) == 0)
+		return -EINVAL;
+
+	vma->vm_ops = &mbochs_region_vm_ops;
+	vma->vm_private_data = mdev_state;
+	return 0;
+}
+
+static vm_fault_t mbochs_dmabuf_vm_fault(struct vm_fault *vmf)
+{
+	struct vm_area_struct *vma = vmf->vma;
+	struct mbochs_dmabuf *dmabuf = vma->vm_private_data;
+
+	if (WARN_ON(vmf->pgoff >= dmabuf->pagecount))
+		return VM_FAULT_SIGBUS;
+
+	vmf->page = dmabuf->pages[vmf->pgoff];
+	get_page(vmf->page);
+	return 0;
+}
+
+static const struct vm_operations_struct mbochs_dmabuf_vm_ops = {
+	.fault = mbochs_dmabuf_vm_fault,
+};
+
+static int mbochs_mmap_dmabuf(struct dma_buf *buf, struct vm_area_struct *vma)
+{
+	struct mbochs_dmabuf *dmabuf = buf->priv;
+	struct device *dev = mdev_dev(dmabuf->mdev_state->mdev);
+
+	dev_dbg(dev, "%s: %d\n", __func__, dmabuf->id);
+
+	if ((vma->vm_flags & VM_SHARED) == 0)
+		return -EINVAL;
+
+	vma->vm_ops = &mbochs_dmabuf_vm_ops;
+	vma->vm_private_data = dmabuf;
+	return 0;
+}
+
+static void mbochs_print_dmabuf(struct mbochs_dmabuf *dmabuf,
+				const char *prefix)
+{
+	struct device *dev = mdev_dev(dmabuf->mdev_state->mdev);
+	u32 fourcc = dmabuf->mode.drm_format;
+
+	dev_dbg(dev, "%s/%d: %c%c%c%c, %dx%d, stride %d, off 0x%llx, size 0x%llx, pages %ld\n",
+		prefix, dmabuf->id,
+		fourcc ? ((fourcc >>  0) & 0xff) : '-',
+		fourcc ? ((fourcc >>  8) & 0xff) : '-',
+		fourcc ? ((fourcc >> 16) & 0xff) : '-',
+		fourcc ? ((fourcc >> 24) & 0xff) : '-',
+		dmabuf->mode.width, dmabuf->mode.height, dmabuf->mode.stride,
+		dmabuf->mode.offset, dmabuf->mode.size, dmabuf->pagecount);
+}
+
+static struct sg_table *mbochs_map_dmabuf(struct dma_buf_attachment *at,
+					  enum dma_data_direction direction)
+{
+	struct mbochs_dmabuf *dmabuf = at->dmabuf->priv;
+	struct device *dev = mdev_dev(dmabuf->mdev_state->mdev);
+	struct sg_table *sg;
+
+	dev_dbg(dev, "%s: %d\n", __func__, dmabuf->id);
+
+	sg = kzalloc(sizeof(*sg), GFP_KERNEL);
+	if (!sg)
+		goto err1;
+	if (sg_alloc_table_from_pages(sg, dmabuf->pages, dmabuf->pagecount,
+				      0, dmabuf->mode.size, GFP_KERNEL) < 0)
+		goto err2;
+	if (dma_map_sgtable(at->dev, sg, direction, 0))
+		goto err3;
+
+	return sg;
+
+err3:
+	sg_free_table(sg);
+err2:
+	kfree(sg);
+err1:
+	return ERR_PTR(-ENOMEM);
+}
+
+static void mbochs_unmap_dmabuf(struct dma_buf_attachment *at,
+				struct sg_table *sg,
+				enum dma_data_direction direction)
+{
+	struct mbochs_dmabuf *dmabuf = at->dmabuf->priv;
+	struct device *dev = mdev_dev(dmabuf->mdev_state->mdev);
+
+	dev_dbg(dev, "%s: %d\n", __func__, dmabuf->id);
+
+	dma_unmap_sgtable(at->dev, sg, direction, 0);
+	sg_free_table(sg);
+	kfree(sg);
+}
+
+static void mbochs_release_dmabuf(struct dma_buf *buf)
+{
+	struct mbochs_dmabuf *dmabuf = buf->priv;
+	struct mdev_state *mdev_state = dmabuf->mdev_state;
+	struct device *dev = mdev_dev(mdev_state->mdev);
+	pgoff_t pg;
+
+	dev_dbg(dev, "%s: %d\n", __func__, dmabuf->id);
+
+	for (pg = 0; pg < dmabuf->pagecount; pg++)
+		put_page(dmabuf->pages[pg]);
+
+	mutex_lock(&mdev_state->ops_lock);
+	dmabuf->buf = NULL;
+	if (dmabuf->unlinked)
+		kfree(dmabuf);
+	mutex_unlock(&mdev_state->ops_lock);
+}
+
+static struct dma_buf_ops mbochs_dmabuf_ops = {
+	.map_dma_buf	  = mbochs_map_dmabuf,
+	.unmap_dma_buf	  = mbochs_unmap_dmabuf,
+	.release	  = mbochs_release_dmabuf,
+	.mmap		  = mbochs_mmap_dmabuf,
+};
+
+static struct mbochs_dmabuf *mbochs_dmabuf_alloc(struct mdev_state *mdev_state,
+						 struct mbochs_mode *mode)
+{
+	struct mbochs_dmabuf *dmabuf;
+	pgoff_t page_offset, pg;
+
+	WARN_ON(!mutex_is_locked(&mdev_state->ops_lock));
+
+	dmabuf = kzalloc(sizeof(struct mbochs_dmabuf), GFP_KERNEL);
+	if (!dmabuf)
+		return NULL;
+
+	dmabuf->mode = *mode;
+	dmabuf->id = mdev_state->next_id++;
+	dmabuf->pagecount = DIV_ROUND_UP(mode->size, PAGE_SIZE);
+	dmabuf->pages = kcalloc(dmabuf->pagecount, sizeof(struct page *),
+				GFP_KERNEL);
+	if (!dmabuf->pages)
+		goto err_free_dmabuf;
+
+	page_offset = dmabuf->mode.offset >> PAGE_SHIFT;
+	for (pg = 0; pg < dmabuf->pagecount; pg++) {
+		dmabuf->pages[pg] = __mbochs_get_page(mdev_state,
+						      page_offset + pg);
+		if (!dmabuf->pages[pg])
+			goto err_free_pages;
+	}
+
+	dmabuf->mdev_state = mdev_state;
+	list_add(&dmabuf->next, &mdev_state->dmabufs);
+
+	mbochs_print_dmabuf(dmabuf, __func__);
+	return dmabuf;
+
+err_free_pages:
+	while (pg > 0)
+		put_page(dmabuf->pages[--pg]);
+	kfree(dmabuf->pages);
+err_free_dmabuf:
+	kfree(dmabuf);
+	return NULL;
+}
+
+static struct mbochs_dmabuf *
+mbochs_dmabuf_find_by_mode(struct mdev_state *mdev_state,
+			   struct mbochs_mode *mode)
+{
+	struct mbochs_dmabuf *dmabuf;
+
+	WARN_ON(!mutex_is_locked(&mdev_state->ops_lock));
+
+	list_for_each_entry(dmabuf, &mdev_state->dmabufs, next)
+		if (mbochs_modes_equal(&dmabuf->mode, mode))
+			return dmabuf;
+
+	return NULL;
+}
+
+static struct mbochs_dmabuf *
+mbochs_dmabuf_find_by_id(struct mdev_state *mdev_state, u32 id)
+{
+	struct mbochs_dmabuf *dmabuf;
+
+	WARN_ON(!mutex_is_locked(&mdev_state->ops_lock));
+
+	list_for_each_entry(dmabuf, &mdev_state->dmabufs, next)
+		if (dmabuf->id == id)
+			return dmabuf;
+
+	return NULL;
+}
+
+static int mbochs_dmabuf_export(struct mbochs_dmabuf *dmabuf)
+{
+	struct mdev_state *mdev_state = dmabuf->mdev_state;
+	struct device *dev = mdev_state->vdev.dev;
+	DEFINE_DMA_BUF_EXPORT_INFO(exp_info);
+	struct dma_buf *buf;
+
+	WARN_ON(!mutex_is_locked(&mdev_state->ops_lock));
+
+	if (!IS_ALIGNED(dmabuf->mode.offset, PAGE_SIZE)) {
+		dev_info_ratelimited(dev, "%s: framebuffer not page-aligned\n",
+				     __func__);
+		return -EINVAL;
+	}
+
+	exp_info.ops = &mbochs_dmabuf_ops;
+	exp_info.size = dmabuf->mode.size;
+	exp_info.priv = dmabuf;
+
+	buf = dma_buf_export(&exp_info);
+	if (IS_ERR(buf)) {
+		dev_info_ratelimited(dev, "%s: dma_buf_export failed: %ld\n",
+				     __func__, PTR_ERR(buf));
+		return PTR_ERR(buf);
+	}
+
+	dmabuf->buf = buf;
+	dev_dbg(dev, "%s: %d\n", __func__, dmabuf->id);
+	return 0;
+}
+
+static int mbochs_ioctl_get_region_info(struct vfio_device *vdev,
+					struct vfio_region_info *region_info,
+					struct vfio_info_cap *caps)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+
+	if (region_info->index >= MBOCHS_NUM_REGIONS)
+		return -EINVAL;
+
+	switch (region_info->index) {
+	case VFIO_PCI_CONFIG_REGION_INDEX:
+		region_info->offset = 0;
+		region_info->size   = MBOCHS_CONFIG_SPACE_SIZE;
+		region_info->flags  = (VFIO_REGION_INFO_FLAG_READ |
+				       VFIO_REGION_INFO_FLAG_WRITE);
+		break;
+	case VFIO_PCI_BAR0_REGION_INDEX:
+		region_info->offset = MBOCHS_MEMORY_BAR_OFFSET;
+		region_info->size   = mdev_state->memsize;
+		region_info->flags  = (VFIO_REGION_INFO_FLAG_READ  |
+				       VFIO_REGION_INFO_FLAG_WRITE |
+				       VFIO_REGION_INFO_FLAG_MMAP);
+		break;
+	case VFIO_PCI_BAR2_REGION_INDEX:
+		region_info->offset = MBOCHS_MMIO_BAR_OFFSET;
+		region_info->size   = MBOCHS_MMIO_BAR_SIZE;
+		region_info->flags  = (VFIO_REGION_INFO_FLAG_READ  |
+				       VFIO_REGION_INFO_FLAG_WRITE);
+		break;
+	case MBOCHS_EDID_REGION_INDEX: {
+		struct vfio_region_info_cap_type cap_type = {
+			.header.id = VFIO_REGION_INFO_CAP_TYPE,
+			.header.version = 1,
+			.type = VFIO_REGION_TYPE_GFX,
+			.subtype = VFIO_REGION_SUBTYPE_GFX_EDID,
+		};
+
+		region_info->offset = MBOCHS_EDID_OFFSET;
+		region_info->size = MBOCHS_EDID_SIZE;
+		region_info->flags = (VFIO_REGION_INFO_FLAG_READ |
+				      VFIO_REGION_INFO_FLAG_WRITE |
+				      VFIO_REGION_INFO_FLAG_CAPS);
+
+		return vfio_info_add_capability(caps, &cap_type.header,
+						sizeof(cap_type));
+	}
+	default:
+		region_info->size   = 0;
+		region_info->offset = 0;
+		region_info->flags  = 0;
+	}
+
+	return 0;
+}
+
+static int mbochs_get_irq_info(struct vfio_irq_info *irq_info)
+{
+	irq_info->count = 0;
+	return 0;
+}
+
+static int mbochs_get_device_info(struct vfio_device_info *dev_info)
+{
+	dev_info->flags = VFIO_DEVICE_FLAGS_PCI;
+	dev_info->num_regions = MBOCHS_NUM_REGIONS;
+	dev_info->num_irqs = VFIO_PCI_NUM_IRQS;
+	return 0;
+}
+
+static int mbochs_query_gfx_plane(struct mdev_state *mdev_state,
+				  struct vfio_device_gfx_plane_info *plane)
+{
+	struct mbochs_dmabuf *dmabuf;
+	struct mbochs_mode mode;
+	int ret;
+
+	if (plane->flags & VFIO_GFX_PLANE_TYPE_PROBE) {
+		if (plane->flags == (VFIO_GFX_PLANE_TYPE_PROBE |
+				     VFIO_GFX_PLANE_TYPE_DMABUF))
+			return 0;
+		return -EINVAL;
+	}
+
+	if (plane->flags != VFIO_GFX_PLANE_TYPE_DMABUF)
+		return -EINVAL;
+
+	plane->drm_format_mod = 0;
+	plane->x_pos	      = 0;
+	plane->y_pos	      = 0;
+	plane->x_hot	      = 0;
+	plane->y_hot	      = 0;
+
+	mutex_lock(&mdev_state->ops_lock);
+
+	ret = -EINVAL;
+	if (plane->drm_plane_type == DRM_PLANE_TYPE_PRIMARY)
+		ret = mbochs_check_framebuffer(mdev_state, &mode);
+	if (ret < 0) {
+		plane->drm_format     = 0;
+		plane->width	      = 0;
+		plane->height	      = 0;
+		plane->stride	      = 0;
+		plane->size	      = 0;
+		plane->dmabuf_id      = 0;
+		goto done;
+	}
+
+	dmabuf = mbochs_dmabuf_find_by_mode(mdev_state, &mode);
+	if (!dmabuf)
+		mbochs_dmabuf_alloc(mdev_state, &mode);
+	if (!dmabuf) {
+		mutex_unlock(&mdev_state->ops_lock);
+		return -ENOMEM;
+	}
+
+	plane->drm_format     = dmabuf->mode.drm_format;
+	plane->width	      = dmabuf->mode.width;
+	plane->height	      = dmabuf->mode.height;
+	plane->stride	      = dmabuf->mode.stride;
+	plane->size	      = dmabuf->mode.size;
+	plane->dmabuf_id      = dmabuf->id;
+
+done:
+	if (plane->drm_plane_type == DRM_PLANE_TYPE_PRIMARY &&
+	    mdev_state->active_id != plane->dmabuf_id) {
+		dev_dbg(mdev_state->vdev.dev, "%s: primary: %d => %d\n",
+			__func__, mdev_state->active_id, plane->dmabuf_id);
+		mdev_state->active_id = plane->dmabuf_id;
+	}
+	mutex_unlock(&mdev_state->ops_lock);
+	return 0;
+}
+
+static int mbochs_get_gfx_dmabuf(struct mdev_state *mdev_state, u32 id)
+{
+	struct mbochs_dmabuf *dmabuf;
+
+	mutex_lock(&mdev_state->ops_lock);
+
+	dmabuf = mbochs_dmabuf_find_by_id(mdev_state, id);
+	if (!dmabuf) {
+		mutex_unlock(&mdev_state->ops_lock);
+		return -ENOENT;
+	}
+
+	if (!dmabuf->buf)
+		mbochs_dmabuf_export(dmabuf);
+
+	mutex_unlock(&mdev_state->ops_lock);
+
+	if (!dmabuf->buf)
+		return -EINVAL;
+
+	return dma_buf_fd(dmabuf->buf, 0);
+}
+
+static long mbochs_ioctl(struct vfio_device *vdev, unsigned int cmd,
+			 unsigned long arg)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	int ret = 0;
+	unsigned long minsz;
+
+	switch (cmd) {
+	case VFIO_DEVICE_GET_INFO:
+	{
+		struct vfio_device_info info;
+
+		minsz = offsetofend(struct vfio_device_info, num_irqs);
+
+		if (copy_from_user(&info, (void __user *)arg, minsz))
+			return -EFAULT;
+
+		if (info.argsz < minsz)
+			return -EINVAL;
+
+		ret = mbochs_get_device_info(&info);
+		if (ret)
+			return ret;
+
+		if (copy_to_user((void __user *)arg, &info, minsz))
+			return -EFAULT;
+
+		return 0;
+	}
+
+	case VFIO_DEVICE_GET_IRQ_INFO:
+	{
+		struct vfio_irq_info info;
+
+		minsz = offsetofend(struct vfio_irq_info, count);
+
+		if (copy_from_user(&info, (void __user *)arg, minsz))
+			return -EFAULT;
+
+		if ((info.argsz < minsz) ||
+		    (info.index >= VFIO_PCI_NUM_IRQS))
+			return -EINVAL;
+
+		ret = mbochs_get_irq_info(&info);
+		if (ret)
+			return ret;
+
+		if (copy_to_user((void __user *)arg, &info, minsz))
+			return -EFAULT;
+
+		return 0;
+	}
+
+	case VFIO_DEVICE_QUERY_GFX_PLANE:
+	{
+		struct vfio_device_gfx_plane_info plane = {};
+
+		minsz = offsetofend(struct vfio_device_gfx_plane_info,
+				    region_index);
+
+		if (copy_from_user(&plane, (void __user *)arg, minsz))
+			return -EFAULT;
+
+		if (plane.argsz < minsz)
+			return -EINVAL;
+
+		ret = mbochs_query_gfx_plane(mdev_state, &plane);
+		if (ret)
+			return ret;
+
+		if (copy_to_user((void __user *)arg, &plane, minsz))
+			return -EFAULT;
+
+		return 0;
+	}
+
+	case VFIO_DEVICE_GET_GFX_DMABUF:
+	{
+		u32 dmabuf_id;
+
+		if (get_user(dmabuf_id, (__u32 __user *)arg))
+			return -EFAULT;
+
+		return mbochs_get_gfx_dmabuf(mdev_state, dmabuf_id);
+	}
+
+	case VFIO_DEVICE_SET_IRQS:
+		return -EINVAL;
+
+	case VFIO_DEVICE_RESET:
+		return mbochs_reset(mdev_state);
+	}
+	return -ENOTTY;
+}
+
+static void mbochs_close_device(struct vfio_device *vdev)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	struct mbochs_dmabuf *dmabuf, *tmp;
+
+	mutex_lock(&mdev_state->ops_lock);
+
+	list_for_each_entry_safe(dmabuf, tmp, &mdev_state->dmabufs, next) {
+		list_del(&dmabuf->next);
+		if (dmabuf->buf) {
+			/* free in mbochs_release_dmabuf() */
+			dmabuf->unlinked = true;
+		} else {
+			kfree(dmabuf);
+		}
+	}
+	mbochs_put_pages(mdev_state);
+
+	mutex_unlock(&mdev_state->ops_lock);
+}
+
+static ssize_t
+memory_show(struct device *dev, struct device_attribute *attr,
+	    char *buf)
+{
+	struct mdev_state *mdev_state = dev_get_drvdata(dev);
+
+	return sprintf(buf, "%d MB\n", mdev_state->type->mbytes);
+}
+static DEVICE_ATTR_RO(memory);
+
+static struct attribute *mdev_dev_attrs[] = {
+	&dev_attr_memory.attr,
+	NULL,
+};
+
+static const struct attribute_group mdev_dev_group = {
+	.name  = "vendor",
+	.attrs = mdev_dev_attrs,
+};
+
+static const struct attribute_group *mdev_dev_groups[] = {
+	&mdev_dev_group,
+	NULL,
+};
+
+static ssize_t mbochs_show_description(struct mdev_type *mtype, char *buf)
+{
+	struct mbochs_type *type =
+		container_of(mtype, struct mbochs_type, type);
+
+	return sprintf(buf, "virtual display, %d MB video memory\n",
+		       type ? type->mbytes  : 0);
+}
+
+static unsigned int mbochs_get_available(struct mdev_type *mtype)
+{
+	struct mbochs_type *type =
+		container_of(mtype, struct mbochs_type, type);
+
+	return atomic_read(&mbochs_avail_mbytes) / type->mbytes;
+}
+
+static const struct vfio_device_ops mbochs_dev_ops = {
+	.close_device = mbochs_close_device,
+	.init = mbochs_init_dev,
+	.release = mbochs_release_dev,
+	.read = mbochs_read,
+	.write = mbochs_write,
+	.ioctl = mbochs_ioctl,
+	.get_region_info_caps = mbochs_ioctl_get_region_info,
+	.mmap = mbochs_mmap,
+	.bind_iommufd	= vfio_iommufd_emulated_bind,
+	.unbind_iommufd	= vfio_iommufd_emulated_unbind,
+	.attach_ioas	= vfio_iommufd_emulated_attach_ioas,
+	.detach_ioas	= vfio_iommufd_emulated_detach_ioas,
+};
+
+static struct mdev_driver mbochs_driver = {
+	.device_api = VFIO_DEVICE_API_PCI_STRING,
+	.driver = {
+		.name = "mbochs",
+		.owner = THIS_MODULE,
+		.mod_name = KBUILD_MODNAME,
+		.dev_groups = mdev_dev_groups,
+	},
+	.probe = mbochs_probe,
+	.remove	= mbochs_remove,
+	.get_available = mbochs_get_available,
+	.show_description = mbochs_show_description,
+};
+
+static const struct file_operations vd_fops = {
+	.owner		= THIS_MODULE,
+};
+
+static void mbochs_device_release(struct device *dev)
+{
+	/* nothing */
+}
+
+static int __init mbochs_dev_init(void)
+{
+	int ret = 0;
+
+	atomic_set(&mbochs_avail_mbytes, max_mbytes);
+
+	ret = alloc_chrdev_region(&mbochs_devt, 0, MINORMASK + 1, MBOCHS_NAME);
+	if (ret < 0) {
+		pr_err("Error: failed to register mbochs_dev, err: %d\n", ret);
+		return ret;
+	}
+	cdev_init(&mbochs_cdev, &vd_fops);
+	cdev_add(&mbochs_cdev, mbochs_devt, MINORMASK + 1);
+	pr_info("%s: major %d\n", __func__, MAJOR(mbochs_devt));
+
+	ret = mdev_register_driver(&mbochs_driver);
+	if (ret)
+		goto err_cdev;
+
+	ret = class_register(&mbochs_class);
+	if (ret)
+		goto err_driver;
+	mbochs_dev.class = &mbochs_class;
+	mbochs_dev.release = mbochs_device_release;
+	dev_set_name(&mbochs_dev, "%s", MBOCHS_NAME);
+
+	ret = device_register(&mbochs_dev);
+	if (ret)
+		goto err_put;
+
+	ret = mdev_register_parent(&mbochs_parent, &mbochs_dev, &mbochs_driver,
+				   mbochs_mdev_types,
+				   ARRAY_SIZE(mbochs_mdev_types));
+	if (ret)
+		goto err_device;
+
+	return 0;
+
+err_device:
+	device_del(&mbochs_dev);
+err_put:
+	put_device(&mbochs_dev);
+	class_unregister(&mbochs_class);
+err_driver:
+	mdev_unregister_driver(&mbochs_driver);
+err_cdev:
+	cdev_del(&mbochs_cdev);
+	unregister_chrdev_region(mbochs_devt, MINORMASK + 1);
+	return ret;
+}
+
+static void __exit mbochs_dev_exit(void)
+{
+	mbochs_dev.bus = NULL;
+	mdev_unregister_parent(&mbochs_parent);
+
+	device_unregister(&mbochs_dev);
+	mdev_unregister_driver(&mbochs_driver);
+	cdev_del(&mbochs_cdev);
+	unregister_chrdev_region(mbochs_devt, MINORMASK + 1);
+	class_unregister(&mbochs_class);
+}
+
+MODULE_IMPORT_NS("DMA_BUF");
+module_init(mbochs_dev_init)
+module_exit(mbochs_dev_exit)
diff --git a/samples/vfio-mdev/mdpy-defs.h b/samples/vfio-mdev/mdpy-defs.h
new file mode 100644
index 000000000..961c55ec3
--- /dev/null
+++ b/samples/vfio-mdev/mdpy-defs.h
@@ -0,0 +1,22 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Simple pci display device.
+ *
+ * Framebuffer memory is pci bar 0.
+ * Configuration (read-only) is in pci config space.
+ * Format field uses drm fourcc codes.
+ * ATM only DRM_FORMAT_XRGB8888 is supported.
+ */
+
+/* pci ids */
+#define MDPY_PCI_VENDOR_ID	PCI_VENDOR_ID_REDHAT
+#define MDPY_PCI_DEVICE_ID	0x000f
+#define MDPY_PCI_SUBVENDOR_ID	PCI_SUBVENDOR_ID_REDHAT_QUMRANET
+#define MDPY_PCI_SUBDEVICE_ID	PCI_SUBDEVICE_ID_QEMU
+
+/* pci cfg space offsets for fb config (dword) */
+#define MDPY_VENDORCAP_OFFSET   0x40
+#define MDPY_VENDORCAP_SIZE     0x10
+#define MDPY_FORMAT_OFFSET	(MDPY_VENDORCAP_OFFSET + 0x04)
+#define MDPY_WIDTH_OFFSET	(MDPY_VENDORCAP_OFFSET + 0x08)
+#define MDPY_HEIGHT_OFFSET	(MDPY_VENDORCAP_OFFSET + 0x0c)
diff --git a/samples/vfio-mdev/mdpy-fb.c b/samples/vfio-mdev/mdpy-fb.c
new file mode 100644
index 000000000..149af7f59
--- /dev/null
+++ b/samples/vfio-mdev/mdpy-fb.c
@@ -0,0 +1,233 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Framebuffer driver for mdpy (mediated virtual pci display device).
+ *
+ * See mdpy-defs.h for device specs
+ *
+ *   (c) Gerd Hoffmann <kraxel@redhat.com>
+ *
+ * Using some code snippets from simplefb and cirrusfb.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope 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.
+ */
+#include <linux/errno.h>
+#include <linux/fb.h>
+#include <linux/io.h>
+#include <linux/pci.h>
+#include <linux/module.h>
+#include <drm/drm_fourcc.h>
+#include "mdpy-defs.h"
+
+static const struct fb_fix_screeninfo mdpy_fb_fix = {
+	.id		= "mdpy-fb",
+	.type		= FB_TYPE_PACKED_PIXELS,
+	.visual		= FB_VISUAL_TRUECOLOR,
+	.accel		= FB_ACCEL_NONE,
+};
+
+static const struct fb_var_screeninfo mdpy_fb_var = {
+	.height		= -1,
+	.width		= -1,
+	.activate	= FB_ACTIVATE_NOW,
+	.vmode		= FB_VMODE_NONINTERLACED,
+
+	.bits_per_pixel = 32,
+	.transp.offset	= 24,
+	.red.offset	= 16,
+	.green.offset	= 8,
+	.blue.offset	= 0,
+	.transp.length	= 8,
+	.red.length	= 8,
+	.green.length	= 8,
+	.blue.length	= 8,
+};
+
+#define PSEUDO_PALETTE_SIZE 16
+
+struct mdpy_fb_par {
+	u32 palette[PSEUDO_PALETTE_SIZE];
+};
+
+static int mdpy_fb_setcolreg(u_int regno, u_int red, u_int green, u_int blue,
+			      u_int transp, struct fb_info *info)
+{
+	u32 *pal = info->pseudo_palette;
+	u32 cr = red >> (16 - info->var.red.length);
+	u32 cg = green >> (16 - info->var.green.length);
+	u32 cb = blue >> (16 - info->var.blue.length);
+	u32 value, mask;
+
+	if (regno >= PSEUDO_PALETTE_SIZE)
+		return -EINVAL;
+
+	value = (cr << info->var.red.offset) |
+		(cg << info->var.green.offset) |
+		(cb << info->var.blue.offset);
+	if (info->var.transp.length > 0) {
+		mask = (1 << info->var.transp.length) - 1;
+		mask <<= info->var.transp.offset;
+		value |= mask;
+	}
+	pal[regno] = value;
+
+	return 0;
+}
+
+static void mdpy_fb_destroy(struct fb_info *info)
+{
+	if (info->screen_base)
+		iounmap(info->screen_base);
+}
+
+static const struct fb_ops mdpy_fb_ops = {
+	.owner		= THIS_MODULE,
+	FB_DEFAULT_IOMEM_OPS,
+	.fb_destroy	= mdpy_fb_destroy,
+	.fb_setcolreg	= mdpy_fb_setcolreg,
+};
+
+static int mdpy_fb_probe(struct pci_dev *pdev,
+			 const struct pci_device_id *ent)
+{
+	struct fb_info *info;
+	struct mdpy_fb_par *par;
+	u32 format, width, height;
+	int ret;
+
+	ret = pci_enable_device(pdev);
+	if (ret < 0)
+		return ret;
+
+	ret = pci_request_regions(pdev, "mdpy-fb");
+	if (ret < 0)
+		goto err_disable_dev;
+
+	pci_read_config_dword(pdev, MDPY_FORMAT_OFFSET, &format);
+	pci_read_config_dword(pdev, MDPY_WIDTH_OFFSET,	&width);
+	pci_read_config_dword(pdev, MDPY_HEIGHT_OFFSET, &height);
+	if (format != DRM_FORMAT_XRGB8888) {
+		pci_err(pdev, "format mismatch (0x%x != 0x%x)\n",
+			format, DRM_FORMAT_XRGB8888);
+		ret = -EINVAL;
+		goto err_release_regions;
+	}
+	if (width < 100	 || width > 10000) {
+		pci_err(pdev, "width (%d) out of range\n", width);
+		ret = -EINVAL;
+		goto err_release_regions;
+	}
+	if (height < 100 || height > 10000) {
+		pci_err(pdev, "height (%d) out of range\n", height);
+		ret = -EINVAL;
+		goto err_release_regions;
+	}
+	pci_info(pdev, "mdpy found: %dx%d framebuffer\n",
+		 width, height);
+
+	info = framebuffer_alloc(sizeof(struct mdpy_fb_par), &pdev->dev);
+	if (!info) {
+		ret = -ENOMEM;
+		goto err_release_regions;
+	}
+	pci_set_drvdata(pdev, info);
+	par = info->par;
+
+	info->fix = mdpy_fb_fix;
+	info->fix.smem_start = pci_resource_start(pdev, 0);
+	info->fix.smem_len = pci_resource_len(pdev, 0);
+	info->fix.line_length = width * 4;
+
+	info->var = mdpy_fb_var;
+	info->var.xres = width;
+	info->var.yres = height;
+	info->var.xres_virtual = width;
+	info->var.yres_virtual = height;
+
+	info->screen_size = info->fix.smem_len;
+	info->screen_base = ioremap(info->fix.smem_start,
+				    info->screen_size);
+	if (!info->screen_base) {
+		pci_err(pdev, "ioremap(pcibar) failed\n");
+		ret = -EIO;
+		goto err_release_fb;
+	}
+
+	info->fbops = &mdpy_fb_ops;
+	info->pseudo_palette = par->palette;
+
+	ret = register_framebuffer(info);
+	if (ret < 0) {
+		pci_err(pdev, "mdpy-fb device register failed: %d\n", ret);
+		goto err_unmap;
+	}
+
+	pci_info(pdev, "fb%d registered\n", info->node);
+	return 0;
+
+err_unmap:
+	iounmap(info->screen_base);
+
+err_release_fb:
+	framebuffer_release(info);
+
+err_release_regions:
+	pci_release_regions(pdev);
+
+err_disable_dev:
+	pci_disable_device(pdev);
+
+	return ret;
+}
+
+static void mdpy_fb_remove(struct pci_dev *pdev)
+{
+	struct fb_info *info = pci_get_drvdata(pdev);
+
+	unregister_framebuffer(info);
+	iounmap(info->screen_base);
+	framebuffer_release(info);
+	pci_release_regions(pdev);
+	pci_disable_device(pdev);
+}
+
+static struct pci_device_id mdpy_fb_pci_table[] = {
+	{
+		.vendor	   = MDPY_PCI_VENDOR_ID,
+		.device	   = MDPY_PCI_DEVICE_ID,
+		.subvendor = MDPY_PCI_SUBVENDOR_ID,
+		.subdevice = MDPY_PCI_SUBDEVICE_ID,
+	}, {
+		/* end of list */
+	}
+};
+
+static struct pci_driver mdpy_fb_pci_driver = {
+	.name		= "mdpy-fb",
+	.id_table	= mdpy_fb_pci_table,
+	.probe		= mdpy_fb_probe,
+	.remove		= mdpy_fb_remove,
+};
+
+static int __init mdpy_fb_init(void)
+{
+	int ret;
+
+	ret = pci_register_driver(&mdpy_fb_pci_driver);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+module_init(mdpy_fb_init);
+
+MODULE_DEVICE_TABLE(pci, mdpy_fb_pci_table);
+MODULE_DESCRIPTION("Framebuffer driver for mdpy (mediated virtual pci display device)");
+MODULE_LICENSE("GPL v2");
diff --git a/samples/vfio-mdev/mdpy.c b/samples/vfio-mdev/mdpy.c
new file mode 100644
index 000000000..0759bd68e
--- /dev/null
+++ b/samples/vfio-mdev/mdpy.c
@@ -0,0 +1,743 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Mediated virtual PCI display host device driver
+ *
+ * See mdpy-defs.h for device specs
+ *
+ *   (c) Gerd Hoffmann <kraxel@redhat.com>
+ *
+ * based on mtty driver which is:
+ *   Copyright (c) 2016, NVIDIA CORPORATION. All rights reserved.
+ *	 Author: Neo Jia <cjia@nvidia.com>
+ *		 Kirti Wankhede <kwankhede@nvidia.com>
+ *
+ * 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.
+ */
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/vmalloc.h>
+#include <linux/cdev.h>
+#include <linux/vfio.h>
+#include <linux/iommu.h>
+#include <linux/sysfs.h>
+#include <linux/mdev.h>
+#include <linux/pci.h>
+#include <drm/drm_fourcc.h>
+#include "mdpy-defs.h"
+
+#define MDPY_NAME		"mdpy"
+#define MDPY_CLASS_NAME		"mdpy"
+
+#define MDPY_CONFIG_SPACE_SIZE	0xff
+#define MDPY_MEMORY_BAR_OFFSET	PAGE_SIZE
+#define MDPY_DISPLAY_REGION	16
+
+#define STORE_LE16(addr, val)	(*(u16 *)addr = val)
+#define STORE_LE32(addr, val)	(*(u32 *)addr = val)
+
+
+MODULE_DESCRIPTION("Mediated virtual PCI display host device driver");
+MODULE_LICENSE("GPL v2");
+
+#define MDPY_TYPE_1 "vga"
+#define MDPY_TYPE_2 "xga"
+#define MDPY_TYPE_3 "hd"
+
+static struct mdpy_type {
+	struct mdev_type type;
+	u32 format;
+	u32 bytepp;
+	u32 width;
+	u32 height;
+} mdpy_types[] = {
+	{
+		.type.sysfs_name 	= MDPY_TYPE_1,
+		.type.pretty_name	= MDPY_CLASS_NAME "-" MDPY_TYPE_1,
+		.format = DRM_FORMAT_XRGB8888,
+		.bytepp = 4,
+		.width	= 640,
+		.height = 480,
+	}, {
+		.type.sysfs_name 	= MDPY_TYPE_2,
+		.type.pretty_name	= MDPY_CLASS_NAME "-" MDPY_TYPE_2,
+		.format = DRM_FORMAT_XRGB8888,
+		.bytepp = 4,
+		.width	= 1024,
+		.height = 768,
+	}, {
+		.type.sysfs_name 	= MDPY_TYPE_3,
+		.type.pretty_name	= MDPY_CLASS_NAME "-" MDPY_TYPE_3,
+		.format = DRM_FORMAT_XRGB8888,
+		.bytepp = 4,
+		.width	= 1920,
+		.height = 1080,
+	},
+};
+
+static struct mdev_type *mdpy_mdev_types[] = {
+	&mdpy_types[0].type,
+	&mdpy_types[1].type,
+	&mdpy_types[2].type,
+};
+
+static dev_t		mdpy_devt;
+static const struct class mdpy_class = {
+	.name = MDPY_CLASS_NAME,
+};
+static struct cdev	mdpy_cdev;
+static struct device	mdpy_dev;
+static struct mdev_parent mdpy_parent;
+static const struct vfio_device_ops mdpy_dev_ops;
+
+/* State of each mdev device */
+struct mdev_state {
+	struct vfio_device vdev;
+	u8 *vconfig;
+	u32 bar_mask;
+	struct mutex ops_lock;
+	struct mdev_device *mdev;
+	struct vfio_device_info dev_info;
+
+	const struct mdpy_type *type;
+	u32 memsize;
+	void *memblk;
+};
+
+static void mdpy_create_config_space(struct mdev_state *mdev_state)
+{
+	STORE_LE16((u16 *) &mdev_state->vconfig[PCI_VENDOR_ID],
+		   MDPY_PCI_VENDOR_ID);
+	STORE_LE16((u16 *) &mdev_state->vconfig[PCI_DEVICE_ID],
+		   MDPY_PCI_DEVICE_ID);
+	STORE_LE16((u16 *) &mdev_state->vconfig[PCI_SUBSYSTEM_VENDOR_ID],
+		   MDPY_PCI_SUBVENDOR_ID);
+	STORE_LE16((u16 *) &mdev_state->vconfig[PCI_SUBSYSTEM_ID],
+		   MDPY_PCI_SUBDEVICE_ID);
+
+	STORE_LE16((u16 *) &mdev_state->vconfig[PCI_COMMAND],
+		   PCI_COMMAND_IO | PCI_COMMAND_MEMORY);
+	STORE_LE16((u16 *) &mdev_state->vconfig[PCI_STATUS],
+		   PCI_STATUS_CAP_LIST);
+	STORE_LE16((u16 *) &mdev_state->vconfig[PCI_CLASS_DEVICE],
+		   PCI_CLASS_DISPLAY_OTHER);
+	mdev_state->vconfig[PCI_CLASS_REVISION] =  0x01;
+
+	STORE_LE32((u32 *) &mdev_state->vconfig[PCI_BASE_ADDRESS_0],
+		   PCI_BASE_ADDRESS_SPACE_MEMORY |
+		   PCI_BASE_ADDRESS_MEM_TYPE_32	 |
+		   PCI_BASE_ADDRESS_MEM_PREFETCH);
+	mdev_state->bar_mask = ~(mdev_state->memsize) + 1;
+
+	/* vendor specific capability for the config registers */
+	mdev_state->vconfig[PCI_CAPABILITY_LIST]       = MDPY_VENDORCAP_OFFSET;
+	mdev_state->vconfig[MDPY_VENDORCAP_OFFSET + 0] = 0x09; /* vendor cap */
+	mdev_state->vconfig[MDPY_VENDORCAP_OFFSET + 1] = 0x00; /* next ptr */
+	mdev_state->vconfig[MDPY_VENDORCAP_OFFSET + 2] = MDPY_VENDORCAP_SIZE;
+	STORE_LE32((u32 *) &mdev_state->vconfig[MDPY_FORMAT_OFFSET],
+		   mdev_state->type->format);
+	STORE_LE32((u32 *) &mdev_state->vconfig[MDPY_WIDTH_OFFSET],
+		   mdev_state->type->width);
+	STORE_LE32((u32 *) &mdev_state->vconfig[MDPY_HEIGHT_OFFSET],
+		   mdev_state->type->height);
+}
+
+static void handle_pci_cfg_write(struct mdev_state *mdev_state, u16 offset,
+				 char *buf, u32 count)
+{
+	struct device *dev = mdev_dev(mdev_state->mdev);
+	u32 cfg_addr;
+
+	switch (offset) {
+	case PCI_BASE_ADDRESS_0:
+		cfg_addr = *(u32 *)buf;
+
+		if (cfg_addr == 0xffffffff) {
+			cfg_addr = (cfg_addr & mdev_state->bar_mask);
+		} else {
+			cfg_addr &= PCI_BASE_ADDRESS_MEM_MASK;
+			if (cfg_addr)
+				dev_info(dev, "BAR0 @ 0x%x\n", cfg_addr);
+		}
+
+		cfg_addr |= (mdev_state->vconfig[offset] &
+			     ~PCI_BASE_ADDRESS_MEM_MASK);
+		STORE_LE32(&mdev_state->vconfig[offset], cfg_addr);
+		break;
+	}
+}
+
+static ssize_t mdev_access(struct mdev_state *mdev_state, char *buf,
+			   size_t count, loff_t pos, bool is_write)
+{
+	int ret = 0;
+
+	mutex_lock(&mdev_state->ops_lock);
+
+	if (pos < MDPY_CONFIG_SPACE_SIZE) {
+		if (is_write)
+			handle_pci_cfg_write(mdev_state, pos, buf, count);
+		else
+			memcpy(buf, (mdev_state->vconfig + pos), count);
+
+	} else if ((pos >= MDPY_MEMORY_BAR_OFFSET) &&
+		   (pos + count <=
+		    MDPY_MEMORY_BAR_OFFSET + mdev_state->memsize)) {
+		pos -= MDPY_MEMORY_BAR_OFFSET;
+		if (is_write)
+			memcpy(mdev_state->memblk, buf, count);
+		else
+			memcpy(buf, mdev_state->memblk, count);
+
+	} else {
+		dev_info(mdev_state->vdev.dev,
+			 "%s: %s @0x%llx (unhandled)\n", __func__,
+			 is_write ? "WR" : "RD", pos);
+		ret = -1;
+		goto accessfailed;
+	}
+
+	ret = count;
+
+
+accessfailed:
+	mutex_unlock(&mdev_state->ops_lock);
+
+	return ret;
+}
+
+static int mdpy_reset(struct mdev_state *mdev_state)
+{
+	u32 stride, i;
+
+	/* initialize with gray gradient */
+	stride = mdev_state->type->width * mdev_state->type->bytepp;
+	for (i = 0; i < mdev_state->type->height; i++)
+		memset(mdev_state->memblk + i * stride,
+		       i * 255 / mdev_state->type->height,
+		       stride);
+	return 0;
+}
+
+static int mdpy_init_dev(struct vfio_device *vdev)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	struct mdev_device *mdev = to_mdev_device(vdev->dev);
+	const struct mdpy_type *type =
+		container_of(mdev->type, struct mdpy_type, type);
+	u32 fbsize;
+	int ret = -ENOMEM;
+
+	mdev_state->vconfig = kzalloc(MDPY_CONFIG_SPACE_SIZE, GFP_KERNEL);
+	if (!mdev_state->vconfig)
+		return ret;
+
+	fbsize = roundup_pow_of_two(type->width * type->height * type->bytepp);
+
+	mdev_state->memblk = vmalloc_user(fbsize);
+	if (!mdev_state->memblk)
+		goto out_vconfig;
+
+	mutex_init(&mdev_state->ops_lock);
+	mdev_state->mdev = mdev;
+	mdev_state->type = type;
+	mdev_state->memsize = fbsize;
+	mdpy_create_config_space(mdev_state);
+	mdpy_reset(mdev_state);
+
+	dev_info(vdev->dev, "%s: %s (%dx%d)\n", __func__, type->type.pretty_name,
+		 type->width, type->height);
+	return 0;
+
+out_vconfig:
+	kfree(mdev_state->vconfig);
+	return ret;
+}
+
+static int mdpy_probe(struct mdev_device *mdev)
+{
+	struct mdev_state *mdev_state;
+	int ret;
+
+	mdev_state = vfio_alloc_device(mdev_state, vdev, &mdev->dev,
+				       &mdpy_dev_ops);
+	if (IS_ERR(mdev_state))
+		return PTR_ERR(mdev_state);
+
+	ret = vfio_register_emulated_iommu_dev(&mdev_state->vdev);
+	if (ret)
+		goto err_put_vdev;
+	dev_set_drvdata(&mdev->dev, mdev_state);
+	return 0;
+
+err_put_vdev:
+	vfio_put_device(&mdev_state->vdev);
+	return ret;
+}
+
+static void mdpy_release_dev(struct vfio_device *vdev)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+
+	vfree(mdev_state->memblk);
+	kfree(mdev_state->vconfig);
+}
+
+static void mdpy_remove(struct mdev_device *mdev)
+{
+	struct mdev_state *mdev_state = dev_get_drvdata(&mdev->dev);
+
+	dev_info(&mdev->dev, "%s\n", __func__);
+
+	vfio_unregister_group_dev(&mdev_state->vdev);
+	vfio_put_device(&mdev_state->vdev);
+}
+
+static ssize_t mdpy_read(struct vfio_device *vdev, char __user *buf,
+			 size_t count, loff_t *ppos)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	unsigned int done = 0;
+	int ret;
+
+	while (count) {
+		size_t filled;
+
+		if (count >= 4 && !(*ppos % 4)) {
+			u32 val;
+
+			ret = mdev_access(mdev_state, (char *)&val, sizeof(val),
+					  *ppos, false);
+			if (ret <= 0)
+				goto read_err;
+
+			if (copy_to_user(buf, &val, sizeof(val)))
+				goto read_err;
+
+			filled = 4;
+		} else if (count >= 2 && !(*ppos % 2)) {
+			u16 val;
+
+			ret = mdev_access(mdev_state, (char *)&val, sizeof(val),
+					  *ppos, false);
+			if (ret <= 0)
+				goto read_err;
+
+			if (copy_to_user(buf, &val, sizeof(val)))
+				goto read_err;
+
+			filled = 2;
+		} else {
+			u8 val;
+
+			ret = mdev_access(mdev_state, (char *)&val, sizeof(val),
+					  *ppos, false);
+			if (ret <= 0)
+				goto read_err;
+
+			if (copy_to_user(buf, &val, sizeof(val)))
+				goto read_err;
+
+			filled = 1;
+		}
+
+		count -= filled;
+		done += filled;
+		*ppos += filled;
+		buf += filled;
+	}
+
+	return done;
+
+read_err:
+	return -EFAULT;
+}
+
+static ssize_t mdpy_write(struct vfio_device *vdev, const char __user *buf,
+			  size_t count, loff_t *ppos)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	unsigned int done = 0;
+	int ret;
+
+	while (count) {
+		size_t filled;
+
+		if (count >= 4 && !(*ppos % 4)) {
+			u32 val;
+
+			if (copy_from_user(&val, buf, sizeof(val)))
+				goto write_err;
+
+			ret = mdev_access(mdev_state, (char *)&val, sizeof(val),
+					  *ppos, true);
+			if (ret <= 0)
+				goto write_err;
+
+			filled = 4;
+		} else if (count >= 2 && !(*ppos % 2)) {
+			u16 val;
+
+			if (copy_from_user(&val, buf, sizeof(val)))
+				goto write_err;
+
+			ret = mdev_access(mdev_state, (char *)&val, sizeof(val),
+					  *ppos, true);
+			if (ret <= 0)
+				goto write_err;
+
+			filled = 2;
+		} else {
+			u8 val;
+
+			if (copy_from_user(&val, buf, sizeof(val)))
+				goto write_err;
+
+			ret = mdev_access(mdev_state, (char *)&val, sizeof(val),
+					  *ppos, true);
+			if (ret <= 0)
+				goto write_err;
+
+			filled = 1;
+		}
+		count -= filled;
+		done += filled;
+		*ppos += filled;
+		buf += filled;
+	}
+
+	return done;
+write_err:
+	return -EFAULT;
+}
+
+static int mdpy_mmap(struct vfio_device *vdev, struct vm_area_struct *vma)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+
+	if (vma->vm_pgoff != MDPY_MEMORY_BAR_OFFSET >> PAGE_SHIFT)
+		return -EINVAL;
+	if (vma->vm_end < vma->vm_start)
+		return -EINVAL;
+	if (vma->vm_end - vma->vm_start > mdev_state->memsize)
+		return -EINVAL;
+	if ((vma->vm_flags & VM_SHARED) == 0)
+		return -EINVAL;
+
+	return remap_vmalloc_range(vma, mdev_state->memblk, 0);
+}
+
+static int mdpy_ioctl_get_region_info(struct vfio_device *vdev,
+				      struct vfio_region_info *region_info,
+				      struct vfio_info_cap *caps)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+
+	if (region_info->index >= VFIO_PCI_NUM_REGIONS &&
+	    region_info->index != MDPY_DISPLAY_REGION)
+		return -EINVAL;
+
+	switch (region_info->index) {
+	case VFIO_PCI_CONFIG_REGION_INDEX:
+		region_info->offset = 0;
+		region_info->size   = MDPY_CONFIG_SPACE_SIZE;
+		region_info->flags  = (VFIO_REGION_INFO_FLAG_READ |
+				       VFIO_REGION_INFO_FLAG_WRITE);
+		break;
+	case VFIO_PCI_BAR0_REGION_INDEX:
+	case MDPY_DISPLAY_REGION:
+		region_info->offset = MDPY_MEMORY_BAR_OFFSET;
+		region_info->size   = mdev_state->memsize;
+		region_info->flags  = (VFIO_REGION_INFO_FLAG_READ  |
+				       VFIO_REGION_INFO_FLAG_WRITE |
+				       VFIO_REGION_INFO_FLAG_MMAP);
+		break;
+	default:
+		region_info->size   = 0;
+		region_info->offset = 0;
+		region_info->flags  = 0;
+	}
+
+	return 0;
+}
+
+static int mdpy_get_irq_info(struct vfio_irq_info *irq_info)
+{
+	irq_info->count = 0;
+	return 0;
+}
+
+static int mdpy_get_device_info(struct vfio_device_info *dev_info)
+{
+	dev_info->flags = VFIO_DEVICE_FLAGS_PCI;
+	dev_info->num_regions = VFIO_PCI_NUM_REGIONS;
+	dev_info->num_irqs = VFIO_PCI_NUM_IRQS;
+	return 0;
+}
+
+static int mdpy_query_gfx_plane(struct mdev_state *mdev_state,
+				struct vfio_device_gfx_plane_info *plane)
+{
+	if (plane->flags & VFIO_GFX_PLANE_TYPE_PROBE) {
+		if (plane->flags == (VFIO_GFX_PLANE_TYPE_PROBE |
+				     VFIO_GFX_PLANE_TYPE_REGION))
+			return 0;
+		return -EINVAL;
+	}
+
+	if (plane->flags != VFIO_GFX_PLANE_TYPE_REGION)
+		return -EINVAL;
+
+	plane->drm_format     = mdev_state->type->format;
+	plane->width	      = mdev_state->type->width;
+	plane->height	      = mdev_state->type->height;
+	plane->stride	      = (mdev_state->type->width *
+				 mdev_state->type->bytepp);
+	plane->size	      = mdev_state->memsize;
+	plane->region_index   = MDPY_DISPLAY_REGION;
+
+	/* unused */
+	plane->drm_format_mod = 0;
+	plane->x_pos	      = 0;
+	plane->y_pos	      = 0;
+	plane->x_hot	      = 0;
+	plane->y_hot	      = 0;
+
+	return 0;
+}
+
+static long mdpy_ioctl(struct vfio_device *vdev, unsigned int cmd,
+		       unsigned long arg)
+{
+	int ret = 0;
+	unsigned long minsz;
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+
+	switch (cmd) {
+	case VFIO_DEVICE_GET_INFO:
+	{
+		struct vfio_device_info info;
+
+		minsz = offsetofend(struct vfio_device_info, num_irqs);
+
+		if (copy_from_user(&info, (void __user *)arg, minsz))
+			return -EFAULT;
+
+		if (info.argsz < minsz)
+			return -EINVAL;
+
+		ret = mdpy_get_device_info(&info);
+		if (ret)
+			return ret;
+
+		memcpy(&mdev_state->dev_info, &info, sizeof(info));
+
+		if (copy_to_user((void __user *)arg, &info, minsz))
+			return -EFAULT;
+
+		return 0;
+	}
+
+	case VFIO_DEVICE_GET_IRQ_INFO:
+	{
+		struct vfio_irq_info info;
+
+		minsz = offsetofend(struct vfio_irq_info, count);
+
+		if (copy_from_user(&info, (void __user *)arg, minsz))
+			return -EFAULT;
+
+		if ((info.argsz < minsz) ||
+		    (info.index >= mdev_state->dev_info.num_irqs))
+			return -EINVAL;
+
+		ret = mdpy_get_irq_info(&info);
+		if (ret)
+			return ret;
+
+		if (copy_to_user((void __user *)arg, &info, minsz))
+			return -EFAULT;
+
+		return 0;
+	}
+
+	case VFIO_DEVICE_QUERY_GFX_PLANE:
+	{
+		struct vfio_device_gfx_plane_info plane = {};
+
+		minsz = offsetofend(struct vfio_device_gfx_plane_info,
+				    region_index);
+
+		if (copy_from_user(&plane, (void __user *)arg, minsz))
+			return -EFAULT;
+
+		if (plane.argsz < minsz)
+			return -EINVAL;
+
+		ret = mdpy_query_gfx_plane(mdev_state, &plane);
+		if (ret)
+			return ret;
+
+		if (copy_to_user((void __user *)arg, &plane, minsz))
+			return -EFAULT;
+
+		return 0;
+	}
+
+	case VFIO_DEVICE_SET_IRQS:
+		return -EINVAL;
+
+	case VFIO_DEVICE_RESET:
+		return mdpy_reset(mdev_state);
+	}
+	return -ENOTTY;
+}
+
+static ssize_t
+resolution_show(struct device *dev, struct device_attribute *attr,
+		char *buf)
+{
+	struct mdev_state *mdev_state = dev_get_drvdata(dev);
+
+	return sprintf(buf, "%dx%d\n",
+		       mdev_state->type->width,
+		       mdev_state->type->height);
+}
+static DEVICE_ATTR_RO(resolution);
+
+static struct attribute *mdev_dev_attrs[] = {
+	&dev_attr_resolution.attr,
+	NULL,
+};
+
+static const struct attribute_group mdev_dev_group = {
+	.name  = "vendor",
+	.attrs = mdev_dev_attrs,
+};
+
+static const struct attribute_group *mdev_dev_groups[] = {
+	&mdev_dev_group,
+	NULL,
+};
+
+static ssize_t mdpy_show_description(struct mdev_type *mtype, char *buf)
+{
+	struct mdpy_type *type = container_of(mtype, struct mdpy_type, type);
+
+	return sprintf(buf, "virtual display, %dx%d framebuffer\n",
+		       type->width, type->height);
+}
+
+static const struct vfio_device_ops mdpy_dev_ops = {
+	.init = mdpy_init_dev,
+	.release = mdpy_release_dev,
+	.read = mdpy_read,
+	.write = mdpy_write,
+	.ioctl = mdpy_ioctl,
+	.get_region_info_caps = mdpy_ioctl_get_region_info,
+	.mmap = mdpy_mmap,
+	.bind_iommufd	= vfio_iommufd_emulated_bind,
+	.unbind_iommufd	= vfio_iommufd_emulated_unbind,
+	.attach_ioas	= vfio_iommufd_emulated_attach_ioas,
+	.detach_ioas	= vfio_iommufd_emulated_detach_ioas,
+};
+
+static struct mdev_driver mdpy_driver = {
+	.device_api = VFIO_DEVICE_API_PCI_STRING,
+	.max_instances = 4,
+	.driver = {
+		.name = "mdpy",
+		.owner = THIS_MODULE,
+		.mod_name = KBUILD_MODNAME,
+		.dev_groups = mdev_dev_groups,
+	},
+	.probe = mdpy_probe,
+	.remove	= mdpy_remove,
+	.show_description = mdpy_show_description,
+};
+
+static const struct file_operations vd_fops = {
+	.owner		= THIS_MODULE,
+};
+
+static void mdpy_device_release(struct device *dev)
+{
+	/* nothing */
+}
+
+static int __init mdpy_dev_init(void)
+{
+	int ret = 0;
+
+	ret = alloc_chrdev_region(&mdpy_devt, 0, MINORMASK + 1, MDPY_NAME);
+	if (ret < 0) {
+		pr_err("Error: failed to register mdpy_dev, err: %d\n", ret);
+		return ret;
+	}
+	cdev_init(&mdpy_cdev, &vd_fops);
+	cdev_add(&mdpy_cdev, mdpy_devt, MINORMASK + 1);
+	pr_info("%s: major %d\n", __func__, MAJOR(mdpy_devt));
+
+	ret = mdev_register_driver(&mdpy_driver);
+	if (ret)
+		goto err_cdev;
+
+	ret = class_register(&mdpy_class);
+	if (ret)
+		goto err_driver;
+	mdpy_dev.class = &mdpy_class;
+	mdpy_dev.release = mdpy_device_release;
+	dev_set_name(&mdpy_dev, "%s", MDPY_NAME);
+
+	ret = device_register(&mdpy_dev);
+	if (ret)
+		goto err_put;
+
+	ret = mdev_register_parent(&mdpy_parent, &mdpy_dev, &mdpy_driver,
+				   mdpy_mdev_types,
+				   ARRAY_SIZE(mdpy_mdev_types));
+	if (ret)
+		goto err_device;
+
+	return 0;
+
+err_device:
+	device_del(&mdpy_dev);
+err_put:
+	put_device(&mdpy_dev);
+	class_unregister(&mdpy_class);
+err_driver:
+	mdev_unregister_driver(&mdpy_driver);
+err_cdev:
+	cdev_del(&mdpy_cdev);
+	unregister_chrdev_region(mdpy_devt, MINORMASK + 1);
+	return ret;
+}
+
+static void __exit mdpy_dev_exit(void)
+{
+	mdpy_dev.bus = NULL;
+	mdev_unregister_parent(&mdpy_parent);
+
+	device_unregister(&mdpy_dev);
+	mdev_unregister_driver(&mdpy_driver);
+	cdev_del(&mdpy_cdev);
+	unregister_chrdev_region(mdpy_devt, MINORMASK + 1);
+	class_unregister(&mdpy_class);
+}
+
+module_param_named(count, mdpy_driver.max_instances, int, 0444);
+MODULE_PARM_DESC(count, "number of " MDPY_NAME " devices");
+
+module_init(mdpy_dev_init)
+module_exit(mdpy_dev_exit)
diff --git a/samples/vfio-mdev/mtty.c b/samples/vfio-mdev/mtty.c
new file mode 100644
index 000000000..69b6d9def
--- /dev/null
+++ b/samples/vfio-mdev/mtty.c
@@ -0,0 +1,2035 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Mediated virtual PCI serial host device driver
+ *
+ * Copyright (c) 2016, NVIDIA CORPORATION. All rights reserved.
+ *     Author: Neo Jia <cjia@nvidia.com>
+ *             Kirti Wankhede <kwankhede@nvidia.com>
+ *
+ * Sample driver that creates mdev device that simulates serial port over PCI
+ * card.
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/fs.h>
+#include <linux/poll.h>
+#include <linux/slab.h>
+#include <linux/cdev.h>
+#include <linux/sched.h>
+#include <linux/wait.h>
+#include <linux/vfio.h>
+#include <linux/iommu.h>
+#include <linux/sysfs.h>
+#include <linux/ctype.h>
+#include <linux/file.h>
+#include <linux/mdev.h>
+#include <linux/pci.h>
+#include <linux/serial.h>
+#include <uapi/linux/serial_reg.h>
+#include <linux/eventfd.h>
+#include <linux/anon_inodes.h>
+
+/*
+ * #defines
+ */
+
+#define VERSION_STRING  "0.1"
+#define DRIVER_AUTHOR   "NVIDIA Corporation"
+
+#define MTTY_CLASS_NAME "mtty"
+
+#define MTTY_NAME       "mtty"
+
+#define MTTY_STRING_LEN		16
+
+#define MTTY_CONFIG_SPACE_SIZE  0xff
+#define MTTY_IO_BAR_SIZE        0x8
+#define MTTY_MMIO_BAR_SIZE      0x100000
+
+#define STORE_LE16(addr, val)   (*(u16 *)addr = val)
+#define STORE_LE32(addr, val)   (*(u32 *)addr = val)
+
+#define MAX_FIFO_SIZE   16
+
+#define CIRCULAR_BUF_INC_IDX(idx)    (idx = (idx + 1) & (MAX_FIFO_SIZE - 1))
+
+#define MTTY_VFIO_PCI_OFFSET_SHIFT   40
+
+#define MTTY_VFIO_PCI_OFFSET_TO_INDEX(off)   (off >> MTTY_VFIO_PCI_OFFSET_SHIFT)
+#define MTTY_VFIO_PCI_INDEX_TO_OFFSET(index) \
+				((u64)(index) << MTTY_VFIO_PCI_OFFSET_SHIFT)
+#define MTTY_VFIO_PCI_OFFSET_MASK    \
+				(((u64)(1) << MTTY_VFIO_PCI_OFFSET_SHIFT) - 1)
+#define MAX_MTTYS	24
+
+/*
+ * Global Structures
+ */
+
+static const struct class mtty_class = {
+	.name	= MTTY_CLASS_NAME
+};
+
+static struct mtty_dev {
+	dev_t		vd_devt;
+	struct cdev	vd_cdev;
+	struct idr	vd_idr;
+	struct device	dev;
+	struct mdev_parent parent;
+} mtty_dev;
+
+struct mdev_region_info {
+	u64 start;
+	u64 phys_start;
+	u32 size;
+	u64 vfio_offset;
+};
+
+#if defined(DEBUG_REGS)
+static const char *wr_reg[] = {
+	"TX",
+	"IER",
+	"FCR",
+	"LCR",
+	"MCR",
+	"LSR",
+	"MSR",
+	"SCR"
+};
+
+static const char *rd_reg[] = {
+	"RX",
+	"IER",
+	"IIR",
+	"LCR",
+	"MCR",
+	"LSR",
+	"MSR",
+	"SCR"
+};
+#endif
+
+/* loop back buffer */
+struct rxtx {
+	u8 fifo[MAX_FIFO_SIZE];
+	u8 head, tail;
+	u8 count;
+};
+
+struct serial_port {
+	u8 uart_reg[8];         /* 8 registers */
+	struct rxtx rxtx;       /* loop back buffer */
+	bool dlab;
+	bool overrun;
+	u16 divisor;
+	u8 fcr;                 /* FIFO control register */
+	u8 max_fifo_size;
+	u8 intr_trigger_level;  /* interrupt trigger level */
+};
+
+struct mtty_data {
+	u64 magic;
+#define MTTY_MAGIC 0x7e9d09898c3e2c4e /* Nothing clever, just random */
+	u32 major_ver;
+#define MTTY_MAJOR_VER 1
+	u32 minor_ver;
+#define MTTY_MINOR_VER 0
+	u32 nr_ports;
+	u32 flags;
+	struct serial_port ports[2];
+};
+
+struct mdev_state;
+
+struct mtty_migration_file {
+	struct file *filp;
+	struct mutex lock;
+	struct mdev_state *mdev_state;
+	struct mtty_data data;
+	ssize_t filled_size;
+	u8 disabled:1;
+};
+
+/* State of each mdev device */
+struct mdev_state {
+	struct vfio_device vdev;
+	struct eventfd_ctx *intx_evtfd;
+	struct eventfd_ctx *msi_evtfd;
+	int irq_index;
+	u8 *vconfig;
+	struct mutex ops_lock;
+	struct mdev_device *mdev;
+	struct mdev_region_info region_info[VFIO_PCI_NUM_REGIONS];
+	u32 bar_mask[VFIO_PCI_NUM_REGIONS];
+	struct list_head next;
+	struct serial_port s[2];
+	struct mutex rxtx_lock;
+	struct vfio_device_info dev_info;
+	int nr_ports;
+	enum vfio_device_mig_state state;
+	struct mutex state_mutex;
+	struct mutex reset_mutex;
+	struct mtty_migration_file *saving_migf;
+	struct mtty_migration_file *resuming_migf;
+	u8 deferred_reset:1;
+	u8 intx_mask:1;
+};
+
+static struct mtty_type {
+	struct mdev_type type;
+	int nr_ports;
+} mtty_types[2] = {
+	{ .nr_ports = 1, .type.sysfs_name = "1",
+	  .type.pretty_name = "Single port serial" },
+	{ .nr_ports = 2, .type.sysfs_name = "2",
+	  .type.pretty_name = "Dual port serial" },
+};
+
+static struct mdev_type *mtty_mdev_types[] = {
+	&mtty_types[0].type,
+	&mtty_types[1].type,
+};
+
+static atomic_t mdev_avail_ports = ATOMIC_INIT(MAX_MTTYS);
+
+static const struct file_operations vd_fops = {
+	.owner          = THIS_MODULE,
+};
+
+static const struct vfio_device_ops mtty_dev_ops;
+
+/* Helper functions */
+
+static void dump_buffer(u8 *buf, uint32_t count)
+{
+#if defined(DEBUG)
+	int i;
+
+	pr_info("Buffer:\n");
+	for (i = 0; i < count; i++) {
+		pr_info("%2x ", *(buf + i));
+		if ((i + 1) % 16 == 0)
+			pr_info("\n");
+	}
+#endif
+}
+
+static bool is_intx(struct mdev_state *mdev_state)
+{
+	return mdev_state->irq_index == VFIO_PCI_INTX_IRQ_INDEX;
+}
+
+static bool is_msi(struct mdev_state *mdev_state)
+{
+	return mdev_state->irq_index == VFIO_PCI_MSI_IRQ_INDEX;
+}
+
+static bool is_noirq(struct mdev_state *mdev_state)
+{
+	return !is_intx(mdev_state) && !is_msi(mdev_state);
+}
+
+static void mtty_trigger_interrupt(struct mdev_state *mdev_state)
+{
+	lockdep_assert_held(&mdev_state->ops_lock);
+
+	if (is_msi(mdev_state)) {
+		if (mdev_state->msi_evtfd)
+			eventfd_signal(mdev_state->msi_evtfd);
+	} else if (is_intx(mdev_state)) {
+		if (mdev_state->intx_evtfd && !mdev_state->intx_mask) {
+			eventfd_signal(mdev_state->intx_evtfd);
+			mdev_state->intx_mask = true;
+		}
+	}
+}
+
+static void mtty_create_config_space(struct mdev_state *mdev_state)
+{
+	/* PCI dev ID */
+	STORE_LE32((u32 *) &mdev_state->vconfig[0x0], 0x32534348);
+
+	/* Control: I/O+, Mem-, BusMaster- */
+	STORE_LE16((u16 *) &mdev_state->vconfig[0x4], 0x0001);
+
+	/* Status: capabilities list absent */
+	STORE_LE16((u16 *) &mdev_state->vconfig[0x6], 0x0200);
+
+	/* Rev ID */
+	mdev_state->vconfig[0x8] =  0x10;
+
+	/* programming interface class : 16550-compatible serial controller */
+	mdev_state->vconfig[0x9] =  0x02;
+
+	/* Sub class : 00 */
+	mdev_state->vconfig[0xa] =  0x00;
+
+	/* Base class : Simple Communication controllers */
+	mdev_state->vconfig[0xb] =  0x07;
+
+	/* base address registers */
+	/* BAR0: IO space */
+	STORE_LE32((u32 *) &mdev_state->vconfig[0x10], 0x000001);
+	mdev_state->bar_mask[0] = ~(MTTY_IO_BAR_SIZE) + 1;
+
+	if (mdev_state->nr_ports == 2) {
+		/* BAR1: IO space */
+		STORE_LE32((u32 *) &mdev_state->vconfig[0x14], 0x000001);
+		mdev_state->bar_mask[1] = ~(MTTY_IO_BAR_SIZE) + 1;
+	}
+
+	/* Subsystem ID */
+	STORE_LE32((u32 *) &mdev_state->vconfig[0x2c], 0x32534348);
+
+	mdev_state->vconfig[0x34] =  0x00;   /* Cap Ptr */
+	mdev_state->vconfig[0x3d] =  0x01;   /* interrupt pin (INTA#) */
+
+	/* Vendor specific data */
+	mdev_state->vconfig[0x40] =  0x23;
+	mdev_state->vconfig[0x43] =  0x80;
+	mdev_state->vconfig[0x44] =  0x23;
+	mdev_state->vconfig[0x48] =  0x23;
+	mdev_state->vconfig[0x4c] =  0x23;
+
+	mdev_state->vconfig[0x60] =  0x50;
+	mdev_state->vconfig[0x61] =  0x43;
+	mdev_state->vconfig[0x62] =  0x49;
+	mdev_state->vconfig[0x63] =  0x20;
+	mdev_state->vconfig[0x64] =  0x53;
+	mdev_state->vconfig[0x65] =  0x65;
+	mdev_state->vconfig[0x66] =  0x72;
+	mdev_state->vconfig[0x67] =  0x69;
+	mdev_state->vconfig[0x68] =  0x61;
+	mdev_state->vconfig[0x69] =  0x6c;
+	mdev_state->vconfig[0x6a] =  0x2f;
+	mdev_state->vconfig[0x6b] =  0x55;
+	mdev_state->vconfig[0x6c] =  0x41;
+	mdev_state->vconfig[0x6d] =  0x52;
+	mdev_state->vconfig[0x6e] =  0x54;
+}
+
+static void handle_pci_cfg_write(struct mdev_state *mdev_state, u16 offset,
+				 u8 *buf, u32 count)
+{
+	u32 cfg_addr, bar_mask, bar_index = 0;
+
+	switch (offset) {
+	case 0x04: /* device control */
+	case 0x06: /* device status */
+		/* do nothing */
+		break;
+	case 0x3c:  /* interrupt line */
+		mdev_state->vconfig[0x3c] = buf[0];
+		break;
+	case 0x3d:
+		/*
+		 * Interrupt Pin is hardwired to INTA.
+		 * This field is write protected by hardware
+		 */
+		break;
+	case 0x10:  /* BAR0 */
+	case 0x14:  /* BAR1 */
+		if (offset == 0x10)
+			bar_index = 0;
+		else if (offset == 0x14)
+			bar_index = 1;
+
+		if ((mdev_state->nr_ports == 1) && (bar_index == 1)) {
+			STORE_LE32(&mdev_state->vconfig[offset], 0);
+			break;
+		}
+
+		cfg_addr = *(u32 *)buf;
+		pr_info("BAR%d addr 0x%x\n", bar_index, cfg_addr);
+
+		if (cfg_addr == 0xffffffff) {
+			bar_mask = mdev_state->bar_mask[bar_index];
+			cfg_addr = (cfg_addr & bar_mask);
+		}
+
+		cfg_addr |= (mdev_state->vconfig[offset] & 0x3ul);
+		STORE_LE32(&mdev_state->vconfig[offset], cfg_addr);
+		break;
+	case 0x18:  /* BAR2 */
+	case 0x1c:  /* BAR3 */
+	case 0x20:  /* BAR4 */
+		STORE_LE32(&mdev_state->vconfig[offset], 0);
+		break;
+	default:
+		pr_info("PCI config write @0x%x of %d bytes not handled\n",
+			offset, count);
+		break;
+	}
+}
+
+static void handle_bar_write(unsigned int index, struct mdev_state *mdev_state,
+				u16 offset, u8 *buf, u32 count)
+{
+	u8 data = *buf;
+
+	/* Handle data written by guest */
+	switch (offset) {
+	case UART_TX:
+		/* if DLAB set, data is LSB of divisor */
+		if (mdev_state->s[index].dlab) {
+			mdev_state->s[index].divisor |= data;
+			break;
+		}
+
+		mutex_lock(&mdev_state->rxtx_lock);
+
+		/* save in TX buffer */
+		if (mdev_state->s[index].rxtx.count <
+				mdev_state->s[index].max_fifo_size) {
+			mdev_state->s[index].rxtx.fifo[
+					mdev_state->s[index].rxtx.head] = data;
+			mdev_state->s[index].rxtx.count++;
+			CIRCULAR_BUF_INC_IDX(mdev_state->s[index].rxtx.head);
+			mdev_state->s[index].overrun = false;
+
+			/*
+			 * Trigger interrupt if receive data interrupt is
+			 * enabled and fifo reached trigger level
+			 */
+			if ((mdev_state->s[index].uart_reg[UART_IER] &
+						UART_IER_RDI) &&
+			   (mdev_state->s[index].rxtx.count ==
+				    mdev_state->s[index].intr_trigger_level)) {
+				/* trigger interrupt */
+#if defined(DEBUG_INTR)
+				pr_err("Serial port %d: Fifo level trigger\n",
+					index);
+#endif
+				mtty_trigger_interrupt(mdev_state);
+			}
+		} else {
+#if defined(DEBUG_INTR)
+			pr_err("Serial port %d: Buffer Overflow\n", index);
+#endif
+			mdev_state->s[index].overrun = true;
+
+			/*
+			 * Trigger interrupt if receiver line status interrupt
+			 * is enabled
+			 */
+			if (mdev_state->s[index].uart_reg[UART_IER] &
+								UART_IER_RLSI)
+				mtty_trigger_interrupt(mdev_state);
+		}
+		mutex_unlock(&mdev_state->rxtx_lock);
+		break;
+
+	case UART_IER:
+		/* if DLAB set, data is MSB of divisor */
+		if (mdev_state->s[index].dlab)
+			mdev_state->s[index].divisor |= (u16)data << 8;
+		else {
+			mdev_state->s[index].uart_reg[offset] = data;
+			mutex_lock(&mdev_state->rxtx_lock);
+			if ((data & UART_IER_THRI) &&
+			    (mdev_state->s[index].rxtx.head ==
+					mdev_state->s[index].rxtx.tail)) {
+#if defined(DEBUG_INTR)
+				pr_err("Serial port %d: IER_THRI write\n",
+					index);
+#endif
+				mtty_trigger_interrupt(mdev_state);
+			}
+
+			mutex_unlock(&mdev_state->rxtx_lock);
+		}
+
+		break;
+
+	case UART_FCR:
+		mdev_state->s[index].fcr = data;
+
+		mutex_lock(&mdev_state->rxtx_lock);
+		if (data & (UART_FCR_CLEAR_RCVR | UART_FCR_CLEAR_XMIT)) {
+			/* clear loop back FIFO */
+			mdev_state->s[index].rxtx.count = 0;
+			mdev_state->s[index].rxtx.head = 0;
+			mdev_state->s[index].rxtx.tail = 0;
+		}
+		mutex_unlock(&mdev_state->rxtx_lock);
+
+		switch (data & UART_FCR_TRIGGER_MASK) {
+		case UART_FCR_TRIGGER_1:
+			mdev_state->s[index].intr_trigger_level = 1;
+			break;
+
+		case UART_FCR_TRIGGER_4:
+			mdev_state->s[index].intr_trigger_level = 4;
+			break;
+
+		case UART_FCR_TRIGGER_8:
+			mdev_state->s[index].intr_trigger_level = 8;
+			break;
+
+		case UART_FCR_TRIGGER_14:
+			mdev_state->s[index].intr_trigger_level = 14;
+			break;
+		}
+
+		/*
+		 * Set trigger level to 1 otherwise or  implement timer with
+		 * timeout of 4 characters and on expiring that timer set
+		 * Recevice data timeout in IIR register
+		 */
+		mdev_state->s[index].intr_trigger_level = 1;
+		if (data & UART_FCR_ENABLE_FIFO)
+			mdev_state->s[index].max_fifo_size = MAX_FIFO_SIZE;
+		else {
+			mdev_state->s[index].max_fifo_size = 1;
+			mdev_state->s[index].intr_trigger_level = 1;
+		}
+
+		break;
+
+	case UART_LCR:
+		if (data & UART_LCR_DLAB) {
+			mdev_state->s[index].dlab = true;
+			mdev_state->s[index].divisor = 0;
+		} else
+			mdev_state->s[index].dlab = false;
+
+		mdev_state->s[index].uart_reg[offset] = data;
+		break;
+
+	case UART_MCR:
+		mdev_state->s[index].uart_reg[offset] = data;
+
+		if ((mdev_state->s[index].uart_reg[UART_IER] & UART_IER_MSI) &&
+				(data & UART_MCR_OUT2)) {
+#if defined(DEBUG_INTR)
+			pr_err("Serial port %d: MCR_OUT2 write\n", index);
+#endif
+			mtty_trigger_interrupt(mdev_state);
+		}
+
+		if ((mdev_state->s[index].uart_reg[UART_IER] & UART_IER_MSI) &&
+				(data & (UART_MCR_RTS | UART_MCR_DTR))) {
+#if defined(DEBUG_INTR)
+			pr_err("Serial port %d: MCR RTS/DTR write\n", index);
+#endif
+			mtty_trigger_interrupt(mdev_state);
+		}
+		break;
+
+	case UART_LSR:
+	case UART_MSR:
+		/* do nothing */
+		break;
+
+	case UART_SCR:
+		mdev_state->s[index].uart_reg[offset] = data;
+		break;
+
+	default:
+		break;
+	}
+}
+
+static void handle_bar_read(unsigned int index, struct mdev_state *mdev_state,
+			    u16 offset, u8 *buf, u32 count)
+{
+	/* Handle read requests by guest */
+	switch (offset) {
+	case UART_RX:
+		/* if DLAB set, data is LSB of divisor */
+		if (mdev_state->s[index].dlab) {
+			*buf  = (u8)mdev_state->s[index].divisor;
+			break;
+		}
+
+		mutex_lock(&mdev_state->rxtx_lock);
+		/* return data in tx buffer */
+		if (mdev_state->s[index].rxtx.head !=
+				 mdev_state->s[index].rxtx.tail) {
+			*buf = mdev_state->s[index].rxtx.fifo[
+						mdev_state->s[index].rxtx.tail];
+			mdev_state->s[index].rxtx.count--;
+			CIRCULAR_BUF_INC_IDX(mdev_state->s[index].rxtx.tail);
+		}
+
+		if (mdev_state->s[index].rxtx.head ==
+				mdev_state->s[index].rxtx.tail) {
+		/*
+		 *  Trigger interrupt if tx buffer empty interrupt is
+		 *  enabled and fifo is empty
+		 */
+#if defined(DEBUG_INTR)
+			pr_err("Serial port %d: Buffer Empty\n", index);
+#endif
+			if (mdev_state->s[index].uart_reg[UART_IER] &
+							 UART_IER_THRI)
+				mtty_trigger_interrupt(mdev_state);
+		}
+		mutex_unlock(&mdev_state->rxtx_lock);
+
+		break;
+
+	case UART_IER:
+		if (mdev_state->s[index].dlab) {
+			*buf = (u8)(mdev_state->s[index].divisor >> 8);
+			break;
+		}
+		*buf = mdev_state->s[index].uart_reg[offset] & 0x0f;
+		break;
+
+	case UART_IIR:
+	{
+		u8 ier = mdev_state->s[index].uart_reg[UART_IER];
+		*buf = 0;
+
+		mutex_lock(&mdev_state->rxtx_lock);
+		/* Interrupt priority 1: Parity, overrun, framing or break */
+		if ((ier & UART_IER_RLSI) && mdev_state->s[index].overrun)
+			*buf |= UART_IIR_RLSI;
+
+		/* Interrupt priority 2: Fifo trigger level reached */
+		if ((ier & UART_IER_RDI) &&
+		    (mdev_state->s[index].rxtx.count >=
+		      mdev_state->s[index].intr_trigger_level))
+			*buf |= UART_IIR_RDI;
+
+		/* Interrupt priotiry 3: transmitter holding register empty */
+		if ((ier & UART_IER_THRI) &&
+		    (mdev_state->s[index].rxtx.head ==
+				mdev_state->s[index].rxtx.tail))
+			*buf |= UART_IIR_THRI;
+
+		/* Interrupt priotiry 4: Modem status: CTS, DSR, RI or DCD  */
+		if ((ier & UART_IER_MSI) &&
+		    (mdev_state->s[index].uart_reg[UART_MCR] &
+				 (UART_MCR_RTS | UART_MCR_DTR)))
+			*buf |= UART_IIR_MSI;
+
+		/* bit0: 0=> interrupt pending, 1=> no interrupt is pending */
+		if (*buf == 0)
+			*buf = UART_IIR_NO_INT;
+
+		/* set bit 6 & 7 to be 16550 compatible */
+		*buf |= 0xC0;
+		mutex_unlock(&mdev_state->rxtx_lock);
+	}
+	break;
+
+	case UART_LCR:
+	case UART_MCR:
+		*buf = mdev_state->s[index].uart_reg[offset];
+		break;
+
+	case UART_LSR:
+	{
+		u8 lsr = 0;
+
+		mutex_lock(&mdev_state->rxtx_lock);
+		/* at least one char in FIFO */
+		if (mdev_state->s[index].rxtx.head !=
+				 mdev_state->s[index].rxtx.tail)
+			lsr |= UART_LSR_DR;
+
+		/* if FIFO overrun */
+		if (mdev_state->s[index].overrun)
+			lsr |= UART_LSR_OE;
+
+		/* transmit FIFO empty and tramsitter empty */
+		if (mdev_state->s[index].rxtx.head ==
+				 mdev_state->s[index].rxtx.tail)
+			lsr |= UART_LSR_TEMT | UART_LSR_THRE;
+
+		mutex_unlock(&mdev_state->rxtx_lock);
+		*buf = lsr;
+		break;
+	}
+	case UART_MSR:
+		*buf = UART_MSR_DSR | UART_MSR_DDSR | UART_MSR_DCD;
+
+		mutex_lock(&mdev_state->rxtx_lock);
+		/* if AFE is 1 and FIFO have space, set CTS bit */
+		if (mdev_state->s[index].uart_reg[UART_MCR] &
+						 UART_MCR_AFE) {
+			if (mdev_state->s[index].rxtx.count <
+					mdev_state->s[index].max_fifo_size)
+				*buf |= UART_MSR_CTS | UART_MSR_DCTS;
+		} else
+			*buf |= UART_MSR_CTS | UART_MSR_DCTS;
+		mutex_unlock(&mdev_state->rxtx_lock);
+
+		break;
+
+	case UART_SCR:
+		*buf = mdev_state->s[index].uart_reg[offset];
+		break;
+
+	default:
+		break;
+	}
+}
+
+static void mdev_read_base(struct mdev_state *mdev_state)
+{
+	int index, pos;
+	u32 start_lo, start_hi;
+	u32 mem_type;
+
+	pos = PCI_BASE_ADDRESS_0;
+
+	for (index = 0; index <= VFIO_PCI_BAR5_REGION_INDEX; index++) {
+
+		if (!mdev_state->region_info[index].size)
+			continue;
+
+		start_lo = (*(u32 *)(mdev_state->vconfig + pos)) &
+			PCI_BASE_ADDRESS_MEM_MASK;
+		mem_type = (*(u32 *)(mdev_state->vconfig + pos)) &
+			PCI_BASE_ADDRESS_MEM_TYPE_MASK;
+
+		switch (mem_type) {
+		case PCI_BASE_ADDRESS_MEM_TYPE_64:
+			start_hi = (*(u32 *)(mdev_state->vconfig + pos + 4));
+			pos += 4;
+			break;
+		case PCI_BASE_ADDRESS_MEM_TYPE_32:
+		case PCI_BASE_ADDRESS_MEM_TYPE_1M:
+			/* 1M mem BAR treated as 32-bit BAR */
+		default:
+			/* mem unknown type treated as 32-bit BAR */
+			start_hi = 0;
+			break;
+		}
+		pos += 4;
+		mdev_state->region_info[index].start = ((u64)start_hi << 32) |
+							start_lo;
+	}
+}
+
+static ssize_t mdev_access(struct mdev_state *mdev_state, u8 *buf, size_t count,
+			   loff_t pos, bool is_write)
+{
+	unsigned int index;
+	loff_t offset;
+	int ret = 0;
+
+	if (!buf)
+		return -EINVAL;
+
+	mutex_lock(&mdev_state->ops_lock);
+
+	index = MTTY_VFIO_PCI_OFFSET_TO_INDEX(pos);
+	offset = pos & MTTY_VFIO_PCI_OFFSET_MASK;
+	switch (index) {
+	case VFIO_PCI_CONFIG_REGION_INDEX:
+
+#if defined(DEBUG)
+		pr_info("%s: PCI config space %s at offset 0x%llx\n",
+			 __func__, is_write ? "write" : "read", offset);
+#endif
+		if (is_write) {
+			dump_buffer(buf, count);
+			handle_pci_cfg_write(mdev_state, offset, buf, count);
+		} else {
+			memcpy(buf, (mdev_state->vconfig + offset), count);
+			dump_buffer(buf, count);
+		}
+
+		break;
+
+	case VFIO_PCI_BAR0_REGION_INDEX ... VFIO_PCI_BAR5_REGION_INDEX:
+		if (!mdev_state->region_info[index].start)
+			mdev_read_base(mdev_state);
+
+		if (is_write) {
+			dump_buffer(buf, count);
+
+#if defined(DEBUG_REGS)
+			pr_info("%s: BAR%d  WR @0x%llx %s val:0x%02x dlab:%d\n",
+				__func__, index, offset, wr_reg[offset],
+				*buf, mdev_state->s[index].dlab);
+#endif
+			handle_bar_write(index, mdev_state, offset, buf, count);
+		} else {
+			handle_bar_read(index, mdev_state, offset, buf, count);
+			dump_buffer(buf, count);
+
+#if defined(DEBUG_REGS)
+			pr_info("%s: BAR%d  RD @0x%llx %s val:0x%02x dlab:%d\n",
+				__func__, index, offset, rd_reg[offset],
+				*buf, mdev_state->s[index].dlab);
+#endif
+		}
+		break;
+
+	default:
+		ret = -1;
+		goto accessfailed;
+	}
+
+	ret = count;
+
+
+accessfailed:
+	mutex_unlock(&mdev_state->ops_lock);
+
+	return ret;
+}
+
+static size_t mtty_data_size(struct mdev_state *mdev_state)
+{
+	return offsetof(struct mtty_data, ports) +
+		(mdev_state->nr_ports * sizeof(struct serial_port));
+}
+
+static void mtty_disable_file(struct mtty_migration_file *migf)
+{
+	mutex_lock(&migf->lock);
+	migf->disabled = true;
+	migf->filled_size = 0;
+	migf->filp->f_pos = 0;
+	mutex_unlock(&migf->lock);
+}
+
+static void mtty_disable_files(struct mdev_state *mdev_state)
+{
+	if (mdev_state->saving_migf) {
+		mtty_disable_file(mdev_state->saving_migf);
+		fput(mdev_state->saving_migf->filp);
+		mdev_state->saving_migf = NULL;
+	}
+
+	if (mdev_state->resuming_migf) {
+		mtty_disable_file(mdev_state->resuming_migf);
+		fput(mdev_state->resuming_migf->filp);
+		mdev_state->resuming_migf = NULL;
+	}
+}
+
+static void mtty_state_mutex_unlock(struct mdev_state *mdev_state)
+{
+again:
+	mutex_lock(&mdev_state->reset_mutex);
+	if (mdev_state->deferred_reset) {
+		mdev_state->deferred_reset = false;
+		mutex_unlock(&mdev_state->reset_mutex);
+		mdev_state->state = VFIO_DEVICE_STATE_RUNNING;
+		mtty_disable_files(mdev_state);
+		goto again;
+	}
+	mutex_unlock(&mdev_state->state_mutex);
+	mutex_unlock(&mdev_state->reset_mutex);
+}
+
+static int mtty_release_migf(struct inode *inode, struct file *filp)
+{
+	struct mtty_migration_file *migf = filp->private_data;
+
+	mtty_disable_file(migf);
+	mutex_destroy(&migf->lock);
+	kfree(migf);
+
+	return 0;
+}
+
+static long mtty_precopy_ioctl(struct file *filp, unsigned int cmd,
+			       unsigned long arg)
+{
+	struct mtty_migration_file *migf = filp->private_data;
+	struct mdev_state *mdev_state = migf->mdev_state;
+	loff_t *pos = &filp->f_pos;
+	struct vfio_precopy_info info = {};
+	int ret;
+
+	ret = vfio_check_precopy_ioctl(&mdev_state->vdev, cmd, arg, &info);
+	if (ret)
+		return ret;
+
+	mutex_lock(&mdev_state->state_mutex);
+	if (mdev_state->state != VFIO_DEVICE_STATE_PRE_COPY &&
+	    mdev_state->state != VFIO_DEVICE_STATE_PRE_COPY_P2P) {
+		ret = -EINVAL;
+		goto unlock;
+	}
+
+	mutex_lock(&migf->lock);
+
+	if (migf->disabled) {
+		mutex_unlock(&migf->lock);
+		ret = -ENODEV;
+		goto unlock;
+	}
+
+	if (*pos > migf->filled_size) {
+		mutex_unlock(&migf->lock);
+		ret = -EINVAL;
+		goto unlock;
+	}
+
+	info.dirty_bytes = 0;
+	info.initial_bytes = migf->filled_size - *pos;
+	mutex_unlock(&migf->lock);
+
+	ret = copy_to_user((void __user *)arg, &info,
+		offsetofend(struct vfio_precopy_info, dirty_bytes)) ? -EFAULT : 0;
+unlock:
+	mtty_state_mutex_unlock(mdev_state);
+	return ret;
+}
+
+static ssize_t mtty_save_read(struct file *filp, char __user *buf,
+			      size_t len, loff_t *pos)
+{
+	struct mtty_migration_file *migf = filp->private_data;
+	ssize_t ret = 0;
+
+	if (pos)
+		return -ESPIPE;
+
+	pos = &filp->f_pos;
+
+	mutex_lock(&migf->lock);
+
+	dev_dbg(migf->mdev_state->vdev.dev, "%s ask %zu\n", __func__, len);
+
+	if (migf->disabled) {
+		ret = -ENODEV;
+		goto out_unlock;
+	}
+
+	if (*pos > migf->filled_size) {
+		ret = -EINVAL;
+		goto out_unlock;
+	}
+
+	len = min_t(size_t, migf->filled_size - *pos, len);
+	if (len) {
+		if (copy_to_user(buf, (void *)&migf->data + *pos, len)) {
+			ret = -EFAULT;
+			goto out_unlock;
+		}
+		*pos += len;
+		ret = len;
+	}
+out_unlock:
+	dev_dbg(migf->mdev_state->vdev.dev, "%s read %zu\n", __func__, ret);
+	mutex_unlock(&migf->lock);
+	return ret;
+}
+
+static const struct file_operations mtty_save_fops = {
+	.owner = THIS_MODULE,
+	.read = mtty_save_read,
+	.unlocked_ioctl = mtty_precopy_ioctl,
+	.compat_ioctl = compat_ptr_ioctl,
+	.release = mtty_release_migf,
+};
+
+static void mtty_save_state(struct mdev_state *mdev_state)
+{
+	struct mtty_migration_file *migf = mdev_state->saving_migf;
+	int i;
+
+	mutex_lock(&migf->lock);
+	for (i = 0; i < mdev_state->nr_ports; i++) {
+		memcpy(&migf->data.ports[i],
+			&mdev_state->s[i], sizeof(struct serial_port));
+		migf->filled_size += sizeof(struct serial_port);
+	}
+	dev_dbg(mdev_state->vdev.dev,
+		"%s filled to %zu\n", __func__, migf->filled_size);
+	mutex_unlock(&migf->lock);
+}
+
+static int mtty_load_state(struct mdev_state *mdev_state)
+{
+	struct mtty_migration_file *migf = mdev_state->resuming_migf;
+	int i;
+
+	mutex_lock(&migf->lock);
+	/* magic and version already tested by resume write fn */
+	if (migf->filled_size < mtty_data_size(mdev_state)) {
+		dev_dbg(mdev_state->vdev.dev, "%s expected %zu, got %zu\n",
+			__func__, mtty_data_size(mdev_state),
+			migf->filled_size);
+		mutex_unlock(&migf->lock);
+		return -EINVAL;
+	}
+
+	for (i = 0; i < mdev_state->nr_ports; i++)
+		memcpy(&mdev_state->s[i],
+		       &migf->data.ports[i], sizeof(struct serial_port));
+
+	mutex_unlock(&migf->lock);
+	return 0;
+}
+
+static struct mtty_migration_file *
+mtty_save_device_data(struct mdev_state *mdev_state,
+		      enum vfio_device_mig_state state)
+{
+	struct mtty_migration_file *migf = mdev_state->saving_migf;
+	struct mtty_migration_file *ret = NULL;
+
+	if (migf) {
+		if (state == VFIO_DEVICE_STATE_STOP_COPY)
+			goto fill_data;
+		return ret;
+	}
+
+	migf = kzalloc(sizeof(*migf), GFP_KERNEL_ACCOUNT);
+	if (!migf)
+		return ERR_PTR(-ENOMEM);
+
+	migf->filp = anon_inode_getfile("mtty_mig", &mtty_save_fops,
+					migf, O_RDONLY);
+	if (IS_ERR(migf->filp)) {
+		int rc = PTR_ERR(migf->filp);
+
+		kfree(migf);
+		return ERR_PTR(rc);
+	}
+
+	stream_open(migf->filp->f_inode, migf->filp);
+	mutex_init(&migf->lock);
+	migf->mdev_state = mdev_state;
+
+	migf->data.magic = MTTY_MAGIC;
+	migf->data.major_ver = MTTY_MAJOR_VER;
+	migf->data.minor_ver = MTTY_MINOR_VER;
+	migf->data.nr_ports = mdev_state->nr_ports;
+
+	migf->filled_size = offsetof(struct mtty_data, ports);
+
+	dev_dbg(mdev_state->vdev.dev, "%s filled header to %zu\n",
+		__func__, migf->filled_size);
+
+	ret = mdev_state->saving_migf = migf;
+
+fill_data:
+	if (state == VFIO_DEVICE_STATE_STOP_COPY)
+		mtty_save_state(mdev_state);
+
+	return ret;
+}
+
+static ssize_t mtty_resume_write(struct file *filp, const char __user *buf,
+				 size_t len, loff_t *pos)
+{
+	struct mtty_migration_file *migf = filp->private_data;
+	struct mdev_state *mdev_state = migf->mdev_state;
+	loff_t requested_length;
+	ssize_t ret = 0;
+
+	if (pos)
+		return -ESPIPE;
+
+	pos = &filp->f_pos;
+
+	if (*pos < 0 ||
+	    check_add_overflow((loff_t)len, *pos, &requested_length))
+		return -EINVAL;
+
+	if (requested_length > mtty_data_size(mdev_state))
+		return -ENOMEM;
+
+	mutex_lock(&migf->lock);
+
+	if (migf->disabled) {
+		ret = -ENODEV;
+		goto out_unlock;
+	}
+
+	if (copy_from_user((void *)&migf->data + *pos, buf, len)) {
+		ret = -EFAULT;
+		goto out_unlock;
+	}
+
+	*pos += len;
+	ret = len;
+
+	dev_dbg(migf->mdev_state->vdev.dev, "%s received %zu, total %zu\n",
+		__func__, len, migf->filled_size + len);
+
+	if (migf->filled_size < offsetof(struct mtty_data, ports) &&
+	    migf->filled_size + len >= offsetof(struct mtty_data, ports)) {
+		if (migf->data.magic != MTTY_MAGIC || migf->data.flags ||
+		    migf->data.major_ver != MTTY_MAJOR_VER ||
+		    migf->data.minor_ver != MTTY_MINOR_VER ||
+		    migf->data.nr_ports != mdev_state->nr_ports) {
+			dev_dbg(migf->mdev_state->vdev.dev,
+				"%s failed validation\n", __func__);
+			ret = -EFAULT;
+		} else {
+			dev_dbg(migf->mdev_state->vdev.dev,
+				"%s header validated\n", __func__);
+		}
+	}
+
+	migf->filled_size += len;
+
+out_unlock:
+	mutex_unlock(&migf->lock);
+	return ret;
+}
+
+static const struct file_operations mtty_resume_fops = {
+	.owner = THIS_MODULE,
+	.write = mtty_resume_write,
+	.release = mtty_release_migf,
+};
+
+static struct mtty_migration_file *
+mtty_resume_device_data(struct mdev_state *mdev_state)
+{
+	struct mtty_migration_file *migf;
+	int ret;
+
+	migf = kzalloc(sizeof(*migf), GFP_KERNEL_ACCOUNT);
+	if (!migf)
+		return ERR_PTR(-ENOMEM);
+
+	migf->filp = anon_inode_getfile("mtty_mig", &mtty_resume_fops,
+					migf, O_WRONLY);
+	if (IS_ERR(migf->filp)) {
+		ret = PTR_ERR(migf->filp);
+		kfree(migf);
+		return ERR_PTR(ret);
+	}
+
+	stream_open(migf->filp->f_inode, migf->filp);
+	mutex_init(&migf->lock);
+	migf->mdev_state = mdev_state;
+
+	mdev_state->resuming_migf = migf;
+
+	return migf;
+}
+
+static struct file *mtty_step_state(struct mdev_state *mdev_state,
+				     enum vfio_device_mig_state new)
+{
+	enum vfio_device_mig_state cur = mdev_state->state;
+
+	dev_dbg(mdev_state->vdev.dev, "%s: %d -> %d\n", __func__, cur, new);
+
+	/*
+	 * The following state transitions are no-op considering
+	 * mtty does not do DMA nor require any explicit start/stop.
+	 *
+	 *         RUNNING -> RUNNING_P2P
+	 *         RUNNING_P2P -> RUNNING
+	 *         RUNNING_P2P -> STOP
+	 *         PRE_COPY -> PRE_COPY_P2P
+	 *         PRE_COPY_P2P -> PRE_COPY
+	 *         STOP -> RUNNING_P2P
+	 */
+	if ((cur == VFIO_DEVICE_STATE_RUNNING &&
+	     new == VFIO_DEVICE_STATE_RUNNING_P2P) ||
+	    (cur == VFIO_DEVICE_STATE_RUNNING_P2P &&
+	     (new == VFIO_DEVICE_STATE_RUNNING ||
+	      new == VFIO_DEVICE_STATE_STOP)) ||
+	    (cur == VFIO_DEVICE_STATE_PRE_COPY &&
+	     new == VFIO_DEVICE_STATE_PRE_COPY_P2P) ||
+	    (cur == VFIO_DEVICE_STATE_PRE_COPY_P2P &&
+	     new == VFIO_DEVICE_STATE_PRE_COPY) ||
+	    (cur == VFIO_DEVICE_STATE_STOP &&
+	     new == VFIO_DEVICE_STATE_RUNNING_P2P))
+		return NULL;
+
+	/*
+	 * The following state transitions simply close migration files,
+	 * with the exception of RESUMING -> STOP, which needs to load
+	 * the state first.
+	 *
+	 *         RESUMING -> STOP
+	 *         PRE_COPY -> RUNNING
+	 *         PRE_COPY_P2P -> RUNNING_P2P
+	 *         STOP_COPY -> STOP
+	 */
+	if (cur == VFIO_DEVICE_STATE_RESUMING &&
+	    new == VFIO_DEVICE_STATE_STOP) {
+		int ret;
+
+		ret = mtty_load_state(mdev_state);
+		if (ret)
+			return ERR_PTR(ret);
+		mtty_disable_files(mdev_state);
+		return NULL;
+	}
+
+	if ((cur == VFIO_DEVICE_STATE_PRE_COPY &&
+	     new == VFIO_DEVICE_STATE_RUNNING) ||
+	    (cur == VFIO_DEVICE_STATE_PRE_COPY_P2P &&
+	     new == VFIO_DEVICE_STATE_RUNNING_P2P) ||
+	    (cur == VFIO_DEVICE_STATE_STOP_COPY &&
+	     new == VFIO_DEVICE_STATE_STOP)) {
+		mtty_disable_files(mdev_state);
+		return NULL;
+	}
+
+	/*
+	 * The following state transitions return migration files.
+	 *
+	 *         RUNNING -> PRE_COPY
+	 *         RUNNING_P2P -> PRE_COPY_P2P
+	 *         STOP -> STOP_COPY
+	 *         STOP -> RESUMING
+	 *         PRE_COPY_P2P -> STOP_COPY
+	 */
+	if ((cur == VFIO_DEVICE_STATE_RUNNING &&
+	     new == VFIO_DEVICE_STATE_PRE_COPY) ||
+	    (cur == VFIO_DEVICE_STATE_RUNNING_P2P &&
+	     new == VFIO_DEVICE_STATE_PRE_COPY_P2P) ||
+	    (cur == VFIO_DEVICE_STATE_STOP &&
+	     new == VFIO_DEVICE_STATE_STOP_COPY) ||
+	    (cur == VFIO_DEVICE_STATE_PRE_COPY_P2P &&
+	     new == VFIO_DEVICE_STATE_STOP_COPY)) {
+		struct mtty_migration_file *migf;
+
+		migf = mtty_save_device_data(mdev_state, new);
+		if (IS_ERR(migf))
+			return ERR_CAST(migf);
+
+		if (migf) {
+			get_file(migf->filp);
+
+			return migf->filp;
+		}
+		return NULL;
+	}
+
+	if (cur == VFIO_DEVICE_STATE_STOP &&
+	    new == VFIO_DEVICE_STATE_RESUMING) {
+		struct mtty_migration_file *migf;
+
+		migf = mtty_resume_device_data(mdev_state);
+		if (IS_ERR(migf))
+			return ERR_CAST(migf);
+
+		get_file(migf->filp);
+
+		return migf->filp;
+	}
+
+	/* vfio_mig_get_next_state() does not use arcs other than the above */
+	WARN_ON(true);
+	return ERR_PTR(-EINVAL);
+}
+
+static struct file *mtty_set_state(struct vfio_device *vdev,
+				   enum vfio_device_mig_state new_state)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	struct file *ret = NULL;
+
+	dev_dbg(vdev->dev, "%s -> %d\n", __func__, new_state);
+
+	mutex_lock(&mdev_state->state_mutex);
+	while (mdev_state->state != new_state) {
+		enum vfio_device_mig_state next_state;
+		int rc = vfio_mig_get_next_state(vdev, mdev_state->state,
+						 new_state, &next_state);
+		if (rc) {
+			ret = ERR_PTR(rc);
+			break;
+		}
+
+		ret = mtty_step_state(mdev_state, next_state);
+		if (IS_ERR(ret))
+			break;
+
+		mdev_state->state = next_state;
+
+		if (WARN_ON(ret && new_state != next_state)) {
+			fput(ret);
+			ret = ERR_PTR(-EINVAL);
+			break;
+		}
+	}
+	mtty_state_mutex_unlock(mdev_state);
+	return ret;
+}
+
+static int mtty_get_state(struct vfio_device *vdev,
+			  enum vfio_device_mig_state *current_state)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+
+	mutex_lock(&mdev_state->state_mutex);
+	*current_state = mdev_state->state;
+	mtty_state_mutex_unlock(mdev_state);
+	return 0;
+}
+
+static int mtty_get_data_size(struct vfio_device *vdev,
+			      unsigned long *stop_copy_length)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+
+	*stop_copy_length = mtty_data_size(mdev_state);
+	return 0;
+}
+
+static const struct vfio_migration_ops mtty_migration_ops = {
+	.migration_set_state = mtty_set_state,
+	.migration_get_state = mtty_get_state,
+	.migration_get_data_size = mtty_get_data_size,
+};
+
+static int mtty_log_start(struct vfio_device *vdev,
+			  struct rb_root_cached *ranges,
+			  u32 nnodes, u64 *page_size)
+{
+	return 0;
+}
+
+static int mtty_log_stop(struct vfio_device *vdev)
+{
+	return 0;
+}
+
+static int mtty_log_read_and_clear(struct vfio_device *vdev,
+				   unsigned long iova, unsigned long length,
+				   struct iova_bitmap *dirty)
+{
+	return 0;
+}
+
+static const struct vfio_log_ops mtty_log_ops = {
+	.log_start = mtty_log_start,
+	.log_stop = mtty_log_stop,
+	.log_read_and_clear = mtty_log_read_and_clear,
+};
+
+static int mtty_init_dev(struct vfio_device *vdev)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	struct mdev_device *mdev = to_mdev_device(vdev->dev);
+	struct mtty_type *type =
+		container_of(mdev->type, struct mtty_type, type);
+	int avail_ports = atomic_read(&mdev_avail_ports);
+	int ret;
+
+	do {
+		if (avail_ports < type->nr_ports)
+			return -ENOSPC;
+	} while (!atomic_try_cmpxchg(&mdev_avail_ports,
+				     &avail_ports,
+				     avail_ports - type->nr_ports));
+
+	mdev_state->nr_ports = type->nr_ports;
+	mdev_state->irq_index = -1;
+	mdev_state->s[0].max_fifo_size = MAX_FIFO_SIZE;
+	mdev_state->s[1].max_fifo_size = MAX_FIFO_SIZE;
+	mutex_init(&mdev_state->rxtx_lock);
+
+	mdev_state->vconfig = kzalloc(MTTY_CONFIG_SPACE_SIZE, GFP_KERNEL);
+	if (!mdev_state->vconfig) {
+		ret = -ENOMEM;
+		goto err_nr_ports;
+	}
+
+	mutex_init(&mdev_state->ops_lock);
+	mdev_state->mdev = mdev;
+	mtty_create_config_space(mdev_state);
+
+	mutex_init(&mdev_state->state_mutex);
+	mutex_init(&mdev_state->reset_mutex);
+	vdev->migration_flags = VFIO_MIGRATION_STOP_COPY |
+				VFIO_MIGRATION_P2P |
+				VFIO_MIGRATION_PRE_COPY;
+	vdev->mig_ops = &mtty_migration_ops;
+	vdev->log_ops = &mtty_log_ops;
+	mdev_state->state = VFIO_DEVICE_STATE_RUNNING;
+
+	return 0;
+
+err_nr_ports:
+	atomic_add(type->nr_ports, &mdev_avail_ports);
+	return ret;
+}
+
+static int mtty_probe(struct mdev_device *mdev)
+{
+	struct mdev_state *mdev_state;
+	int ret;
+
+	mdev_state = vfio_alloc_device(mdev_state, vdev, &mdev->dev,
+				       &mtty_dev_ops);
+	if (IS_ERR(mdev_state))
+		return PTR_ERR(mdev_state);
+
+	ret = vfio_register_emulated_iommu_dev(&mdev_state->vdev);
+	if (ret)
+		goto err_put_vdev;
+	dev_set_drvdata(&mdev->dev, mdev_state);
+	return 0;
+
+err_put_vdev:
+	vfio_put_device(&mdev_state->vdev);
+	return ret;
+}
+
+static void mtty_release_dev(struct vfio_device *vdev)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+
+	mutex_destroy(&mdev_state->reset_mutex);
+	mutex_destroy(&mdev_state->state_mutex);
+	atomic_add(mdev_state->nr_ports, &mdev_avail_ports);
+	kfree(mdev_state->vconfig);
+}
+
+static void mtty_remove(struct mdev_device *mdev)
+{
+	struct mdev_state *mdev_state = dev_get_drvdata(&mdev->dev);
+
+	vfio_unregister_group_dev(&mdev_state->vdev);
+	vfio_put_device(&mdev_state->vdev);
+}
+
+static int mtty_reset(struct mdev_state *mdev_state)
+{
+	pr_info("%s: called\n", __func__);
+
+	mutex_lock(&mdev_state->reset_mutex);
+	mdev_state->deferred_reset = true;
+	if (!mutex_trylock(&mdev_state->state_mutex)) {
+		mutex_unlock(&mdev_state->reset_mutex);
+		return 0;
+	}
+	mutex_unlock(&mdev_state->reset_mutex);
+	mtty_state_mutex_unlock(mdev_state);
+
+	return 0;
+}
+
+static ssize_t mtty_read(struct vfio_device *vdev, char __user *buf,
+			 size_t count, loff_t *ppos)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	unsigned int done = 0;
+	int ret;
+
+	while (count) {
+		size_t filled;
+
+		if (count >= 4 && !(*ppos % 4)) {
+			u32 val;
+
+			ret =  mdev_access(mdev_state, (u8 *)&val, sizeof(val),
+					   *ppos, false);
+			if (ret <= 0)
+				goto read_err;
+
+			if (copy_to_user(buf, &val, sizeof(val)))
+				goto read_err;
+
+			filled = 4;
+		} else if (count >= 2 && !(*ppos % 2)) {
+			u16 val;
+
+			ret = mdev_access(mdev_state, (u8 *)&val, sizeof(val),
+					  *ppos, false);
+			if (ret <= 0)
+				goto read_err;
+
+			if (copy_to_user(buf, &val, sizeof(val)))
+				goto read_err;
+
+			filled = 2;
+		} else {
+			u8 val;
+
+			ret = mdev_access(mdev_state, (u8 *)&val, sizeof(val),
+					  *ppos, false);
+			if (ret <= 0)
+				goto read_err;
+
+			if (copy_to_user(buf, &val, sizeof(val)))
+				goto read_err;
+
+			filled = 1;
+		}
+
+		count -= filled;
+		done += filled;
+		*ppos += filled;
+		buf += filled;
+	}
+
+	return done;
+
+read_err:
+	return -EFAULT;
+}
+
+static ssize_t mtty_write(struct vfio_device *vdev, const char __user *buf,
+		   size_t count, loff_t *ppos)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	unsigned int done = 0;
+	int ret;
+
+	while (count) {
+		size_t filled;
+
+		if (count >= 4 && !(*ppos % 4)) {
+			u32 val;
+
+			if (copy_from_user(&val, buf, sizeof(val)))
+				goto write_err;
+
+			ret = mdev_access(mdev_state, (u8 *)&val, sizeof(val),
+					  *ppos, true);
+			if (ret <= 0)
+				goto write_err;
+
+			filled = 4;
+		} else if (count >= 2 && !(*ppos % 2)) {
+			u16 val;
+
+			if (copy_from_user(&val, buf, sizeof(val)))
+				goto write_err;
+
+			ret = mdev_access(mdev_state, (u8 *)&val, sizeof(val),
+					  *ppos, true);
+			if (ret <= 0)
+				goto write_err;
+
+			filled = 2;
+		} else {
+			u8 val;
+
+			if (copy_from_user(&val, buf, sizeof(val)))
+				goto write_err;
+
+			ret = mdev_access(mdev_state, (u8 *)&val, sizeof(val),
+					  *ppos, true);
+			if (ret <= 0)
+				goto write_err;
+
+			filled = 1;
+		}
+		count -= filled;
+		done += filled;
+		*ppos += filled;
+		buf += filled;
+	}
+
+	return done;
+write_err:
+	return -EFAULT;
+}
+
+static void mtty_disable_intx(struct mdev_state *mdev_state)
+{
+	if (mdev_state->intx_evtfd) {
+		eventfd_ctx_put(mdev_state->intx_evtfd);
+		mdev_state->intx_evtfd = NULL;
+		mdev_state->intx_mask = false;
+		mdev_state->irq_index = -1;
+	}
+}
+
+static void mtty_disable_msi(struct mdev_state *mdev_state)
+{
+	if (mdev_state->msi_evtfd) {
+		eventfd_ctx_put(mdev_state->msi_evtfd);
+		mdev_state->msi_evtfd = NULL;
+		mdev_state->irq_index = -1;
+	}
+}
+
+static int mtty_set_irqs(struct mdev_state *mdev_state, uint32_t flags,
+			 unsigned int index, unsigned int start,
+			 unsigned int count, void *data)
+{
+	int ret = 0;
+
+	mutex_lock(&mdev_state->ops_lock);
+	switch (index) {
+	case VFIO_PCI_INTX_IRQ_INDEX:
+		switch (flags & VFIO_IRQ_SET_ACTION_TYPE_MASK) {
+		case VFIO_IRQ_SET_ACTION_MASK:
+			if (!is_intx(mdev_state) || start != 0 || count != 1) {
+				ret = -EINVAL;
+				break;
+			}
+
+			if (flags & VFIO_IRQ_SET_DATA_NONE) {
+				mdev_state->intx_mask = true;
+			} else if (flags & VFIO_IRQ_SET_DATA_BOOL) {
+				uint8_t mask = *(uint8_t *)data;
+
+				if (mask)
+					mdev_state->intx_mask = true;
+			} else if (flags &  VFIO_IRQ_SET_DATA_EVENTFD) {
+				ret = -ENOTTY; /* No support for mask fd */
+			}
+			break;
+		case VFIO_IRQ_SET_ACTION_UNMASK:
+			if (!is_intx(mdev_state) || start != 0 || count != 1) {
+				ret = -EINVAL;
+				break;
+			}
+
+			if (flags & VFIO_IRQ_SET_DATA_NONE) {
+				mdev_state->intx_mask = false;
+			} else if (flags & VFIO_IRQ_SET_DATA_BOOL) {
+				uint8_t mask = *(uint8_t *)data;
+
+				if (mask)
+					mdev_state->intx_mask = false;
+			} else if (flags &  VFIO_IRQ_SET_DATA_EVENTFD) {
+				ret = -ENOTTY; /* No support for unmask fd */
+			}
+			break;
+		case VFIO_IRQ_SET_ACTION_TRIGGER:
+			if (is_intx(mdev_state) && !count &&
+			    (flags & VFIO_IRQ_SET_DATA_NONE)) {
+				mtty_disable_intx(mdev_state);
+				break;
+			}
+
+			if (!(is_intx(mdev_state) || is_noirq(mdev_state)) ||
+			    start != 0 || count != 1) {
+				ret = -EINVAL;
+				break;
+			}
+
+			if (flags & VFIO_IRQ_SET_DATA_EVENTFD) {
+				int fd = *(int *)data;
+				struct eventfd_ctx *evt;
+
+				mtty_disable_intx(mdev_state);
+
+				if (fd < 0)
+					break;
+
+				evt = eventfd_ctx_fdget(fd);
+				if (IS_ERR(evt)) {
+					ret = PTR_ERR(evt);
+					break;
+				}
+				mdev_state->intx_evtfd = evt;
+				mdev_state->irq_index = index;
+				break;
+			}
+
+			if (!is_intx(mdev_state)) {
+				ret = -EINVAL;
+				break;
+			}
+
+			if (flags & VFIO_IRQ_SET_DATA_NONE) {
+				mtty_trigger_interrupt(mdev_state);
+			} else if (flags & VFIO_IRQ_SET_DATA_BOOL) {
+				uint8_t trigger = *(uint8_t *)data;
+
+				if (trigger)
+					mtty_trigger_interrupt(mdev_state);
+			}
+			break;
+		}
+		break;
+	case VFIO_PCI_MSI_IRQ_INDEX:
+		switch (flags & VFIO_IRQ_SET_ACTION_TYPE_MASK) {
+		case VFIO_IRQ_SET_ACTION_MASK:
+		case VFIO_IRQ_SET_ACTION_UNMASK:
+			ret = -ENOTTY;
+			break;
+		case VFIO_IRQ_SET_ACTION_TRIGGER:
+			if (is_msi(mdev_state) && !count &&
+			    (flags & VFIO_IRQ_SET_DATA_NONE)) {
+				mtty_disable_msi(mdev_state);
+				break;
+			}
+
+			if (!(is_msi(mdev_state) || is_noirq(mdev_state)) ||
+			    start != 0 || count != 1) {
+				ret = -EINVAL;
+				break;
+			}
+
+			if (flags & VFIO_IRQ_SET_DATA_EVENTFD) {
+				int fd = *(int *)data;
+				struct eventfd_ctx *evt;
+
+				mtty_disable_msi(mdev_state);
+
+				if (fd < 0)
+					break;
+
+				evt = eventfd_ctx_fdget(fd);
+				if (IS_ERR(evt)) {
+					ret = PTR_ERR(evt);
+					break;
+				}
+				mdev_state->msi_evtfd = evt;
+				mdev_state->irq_index = index;
+				break;
+			}
+
+			if (!is_msi(mdev_state)) {
+				ret = -EINVAL;
+				break;
+			}
+
+			if (flags & VFIO_IRQ_SET_DATA_NONE) {
+				mtty_trigger_interrupt(mdev_state);
+			} else if (flags & VFIO_IRQ_SET_DATA_BOOL) {
+				uint8_t trigger = *(uint8_t *)data;
+
+				if (trigger)
+					mtty_trigger_interrupt(mdev_state);
+			}
+			break;
+		}
+		break;
+	case VFIO_PCI_MSIX_IRQ_INDEX:
+		dev_dbg(mdev_state->vdev.dev, "%s: MSIX_IRQ\n", __func__);
+		ret = -ENOTTY;
+		break;
+	case VFIO_PCI_ERR_IRQ_INDEX:
+		dev_dbg(mdev_state->vdev.dev, "%s: ERR_IRQ\n", __func__);
+		ret = -ENOTTY;
+		break;
+	case VFIO_PCI_REQ_IRQ_INDEX:
+		dev_dbg(mdev_state->vdev.dev, "%s: REQ_IRQ\n", __func__);
+		ret = -ENOTTY;
+		break;
+	}
+
+	mutex_unlock(&mdev_state->ops_lock);
+	return ret;
+}
+
+static int mtty_ioctl_get_region_info(struct vfio_device *vdev,
+				      struct vfio_region_info *region_info,
+				      struct vfio_info_cap *caps)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	unsigned int size = 0;
+	u32 bar_index;
+
+	bar_index = region_info->index;
+	if (bar_index >= VFIO_PCI_NUM_REGIONS)
+		return -EINVAL;
+
+	mutex_lock(&mdev_state->ops_lock);
+
+	switch (bar_index) {
+	case VFIO_PCI_CONFIG_REGION_INDEX:
+		size = MTTY_CONFIG_SPACE_SIZE;
+		break;
+	case VFIO_PCI_BAR0_REGION_INDEX:
+		size = MTTY_IO_BAR_SIZE;
+		break;
+	case VFIO_PCI_BAR1_REGION_INDEX:
+		if (mdev_state->nr_ports == 2)
+			size = MTTY_IO_BAR_SIZE;
+		break;
+	default:
+		size = 0;
+		break;
+	}
+
+	mdev_state->region_info[bar_index].size = size;
+	mdev_state->region_info[bar_index].vfio_offset =
+		MTTY_VFIO_PCI_INDEX_TO_OFFSET(bar_index);
+
+	region_info->size = size;
+	region_info->offset = MTTY_VFIO_PCI_INDEX_TO_OFFSET(bar_index);
+	region_info->flags = VFIO_REGION_INFO_FLAG_READ |
+		VFIO_REGION_INFO_FLAG_WRITE;
+	mutex_unlock(&mdev_state->ops_lock);
+	return 0;
+}
+
+static int mtty_get_irq_info(struct vfio_irq_info *irq_info)
+{
+	if (irq_info->index != VFIO_PCI_INTX_IRQ_INDEX &&
+	    irq_info->index != VFIO_PCI_MSI_IRQ_INDEX)
+		return -EINVAL;
+
+	irq_info->flags = VFIO_IRQ_INFO_EVENTFD;
+	irq_info->count = 1;
+
+	if (irq_info->index == VFIO_PCI_INTX_IRQ_INDEX)
+		irq_info->flags |= VFIO_IRQ_INFO_MASKABLE |
+				   VFIO_IRQ_INFO_AUTOMASKED;
+	else
+		irq_info->flags |= VFIO_IRQ_INFO_NORESIZE;
+
+	return 0;
+}
+
+static int mtty_get_device_info(struct vfio_device_info *dev_info)
+{
+	dev_info->flags = VFIO_DEVICE_FLAGS_PCI;
+	dev_info->num_regions = VFIO_PCI_NUM_REGIONS;
+	dev_info->num_irqs = VFIO_PCI_NUM_IRQS;
+
+	return 0;
+}
+
+static long mtty_ioctl(struct vfio_device *vdev, unsigned int cmd,
+			unsigned long arg)
+{
+	struct mdev_state *mdev_state =
+		container_of(vdev, struct mdev_state, vdev);
+	int ret = 0;
+	unsigned long minsz;
+
+	switch (cmd) {
+	case VFIO_DEVICE_GET_INFO:
+	{
+		struct vfio_device_info info;
+
+		minsz = offsetofend(struct vfio_device_info, num_irqs);
+
+		if (copy_from_user(&info, (void __user *)arg, minsz))
+			return -EFAULT;
+
+		if (info.argsz < minsz)
+			return -EINVAL;
+
+		ret = mtty_get_device_info(&info);
+		if (ret)
+			return ret;
+
+		memcpy(&mdev_state->dev_info, &info, sizeof(info));
+
+		if (copy_to_user((void __user *)arg, &info, minsz))
+			return -EFAULT;
+
+		return 0;
+	}
+
+	case VFIO_DEVICE_GET_IRQ_INFO:
+	{
+		struct vfio_irq_info info;
+
+		minsz = offsetofend(struct vfio_irq_info, count);
+
+		if (copy_from_user(&info, (void __user *)arg, minsz))
+			return -EFAULT;
+
+		if ((info.argsz < minsz) ||
+		    (info.index >= mdev_state->dev_info.num_irqs))
+			return -EINVAL;
+
+		ret = mtty_get_irq_info(&info);
+		if (ret)
+			return ret;
+
+		if (copy_to_user((void __user *)arg, &info, minsz))
+			return -EFAULT;
+
+		return 0;
+	}
+	case VFIO_DEVICE_SET_IRQS:
+	{
+		struct vfio_irq_set hdr;
+		u8 *data = NULL, *ptr = NULL;
+		size_t data_size = 0;
+
+		minsz = offsetofend(struct vfio_irq_set, count);
+
+		if (copy_from_user(&hdr, (void __user *)arg, minsz))
+			return -EFAULT;
+
+		ret = vfio_set_irqs_validate_and_prepare(&hdr,
+						mdev_state->dev_info.num_irqs,
+						VFIO_PCI_NUM_IRQS,
+						&data_size);
+		if (ret)
+			return ret;
+
+		if (data_size) {
+			ptr = data = memdup_user((void __user *)(arg + minsz),
+						 data_size);
+			if (IS_ERR(data))
+				return PTR_ERR(data);
+		}
+
+		ret = mtty_set_irqs(mdev_state, hdr.flags, hdr.index, hdr.start,
+				    hdr.count, data);
+
+		kfree(ptr);
+		return ret;
+	}
+	case VFIO_DEVICE_RESET:
+		return mtty_reset(mdev_state);
+	}
+	return -ENOTTY;
+}
+
+static ssize_t
+sample_mdev_dev_show(struct device *dev, struct device_attribute *attr,
+		     char *buf)
+{
+	return sprintf(buf, "This is MDEV %s\n", dev_name(dev));
+}
+
+static DEVICE_ATTR_RO(sample_mdev_dev);
+
+static struct attribute *mdev_dev_attrs[] = {
+	&dev_attr_sample_mdev_dev.attr,
+	NULL,
+};
+
+static const struct attribute_group mdev_dev_group = {
+	.name  = "vendor",
+	.attrs = mdev_dev_attrs,
+};
+
+static const struct attribute_group *mdev_dev_groups[] = {
+	&mdev_dev_group,
+	NULL,
+};
+
+static unsigned int mtty_get_available(struct mdev_type *mtype)
+{
+	struct mtty_type *type = container_of(mtype, struct mtty_type, type);
+
+	return atomic_read(&mdev_avail_ports) / type->nr_ports;
+}
+
+static void mtty_close(struct vfio_device *vdev)
+{
+	struct mdev_state *mdev_state =
+				container_of(vdev, struct mdev_state, vdev);
+
+	mtty_disable_files(mdev_state);
+	mtty_disable_intx(mdev_state);
+	mtty_disable_msi(mdev_state);
+}
+
+static const struct vfio_device_ops mtty_dev_ops = {
+	.name = "vfio-mtty",
+	.init = mtty_init_dev,
+	.release = mtty_release_dev,
+	.read = mtty_read,
+	.write = mtty_write,
+	.ioctl = mtty_ioctl,
+	.get_region_info_caps = mtty_ioctl_get_region_info,
+	.bind_iommufd	= vfio_iommufd_emulated_bind,
+	.unbind_iommufd	= vfio_iommufd_emulated_unbind,
+	.attach_ioas	= vfio_iommufd_emulated_attach_ioas,
+	.detach_ioas	= vfio_iommufd_emulated_detach_ioas,
+	.close_device	= mtty_close,
+};
+
+static struct mdev_driver mtty_driver = {
+	.device_api = VFIO_DEVICE_API_PCI_STRING,
+	.driver = {
+		.name = "mtty",
+		.owner = THIS_MODULE,
+		.mod_name = KBUILD_MODNAME,
+		.dev_groups = mdev_dev_groups,
+	},
+	.probe = mtty_probe,
+	.remove	= mtty_remove,
+	.get_available = mtty_get_available,
+};
+
+static void mtty_device_release(struct device *dev)
+{
+	dev_dbg(dev, "mtty: released\n");
+}
+
+static int __init mtty_dev_init(void)
+{
+	int ret = 0;
+
+	pr_info("mtty_dev: %s\n", __func__);
+
+	memset(&mtty_dev, 0, sizeof(mtty_dev));
+
+	idr_init(&mtty_dev.vd_idr);
+
+	ret = alloc_chrdev_region(&mtty_dev.vd_devt, 0, MINORMASK + 1,
+				  MTTY_NAME);
+
+	if (ret < 0) {
+		pr_err("Error: failed to register mtty_dev, err:%d\n", ret);
+		return ret;
+	}
+
+	cdev_init(&mtty_dev.vd_cdev, &vd_fops);
+	cdev_add(&mtty_dev.vd_cdev, mtty_dev.vd_devt, MINORMASK + 1);
+
+	pr_info("major_number:%d\n", MAJOR(mtty_dev.vd_devt));
+
+	ret = mdev_register_driver(&mtty_driver);
+	if (ret)
+		goto err_cdev;
+
+	ret = class_register(&mtty_class);
+
+	if (ret) {
+		pr_err("Error: failed to register mtty_dev class\n");
+		goto err_driver;
+	}
+
+	mtty_dev.dev.class = &mtty_class;
+	mtty_dev.dev.release = mtty_device_release;
+	dev_set_name(&mtty_dev.dev, "%s", MTTY_NAME);
+
+	ret = device_register(&mtty_dev.dev);
+	if (ret)
+		goto err_put;
+
+	ret = mdev_register_parent(&mtty_dev.parent, &mtty_dev.dev,
+				   &mtty_driver, mtty_mdev_types,
+				   ARRAY_SIZE(mtty_mdev_types));
+	if (ret)
+		goto err_device;
+	return 0;
+
+err_device:
+	device_del(&mtty_dev.dev);
+err_put:
+	put_device(&mtty_dev.dev);
+	class_unregister(&mtty_class);
+err_driver:
+	mdev_unregister_driver(&mtty_driver);
+err_cdev:
+	cdev_del(&mtty_dev.vd_cdev);
+	unregister_chrdev_region(mtty_dev.vd_devt, MINORMASK + 1);
+	return ret;
+}
+
+static void __exit mtty_dev_exit(void)
+{
+	mtty_dev.dev.bus = NULL;
+	mdev_unregister_parent(&mtty_dev.parent);
+
+	device_unregister(&mtty_dev.dev);
+	idr_destroy(&mtty_dev.vd_idr);
+	mdev_unregister_driver(&mtty_driver);
+	cdev_del(&mtty_dev.vd_cdev);
+	unregister_chrdev_region(mtty_dev.vd_devt, MINORMASK + 1);
+	class_unregister(&mtty_class);
+	pr_info("mtty_dev: Unloaded!\n");
+}
+
+module_init(mtty_dev_init)
+module_exit(mtty_dev_exit)
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("Test driver that simulate serial port over PCI");
+MODULE_VERSION(VERSION_STRING);
+MODULE_AUTHOR(DRIVER_AUTHOR);