Merge tag 'virtio-next-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux

Pull virtio updates from Rusty Russell:
 "Some virtio internal cleanups, a new virtio device "virtio input", and
  a change to allow the legacy virtio balloon.

  Most excitingly, some lguest work! No seriously, I got some cleanup
  patches"

* tag 'virtio-next-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux:
  virtio: drop virtio_device_is_legacy_only
  virtio_pci: support non-legacy balloon devices
  virtio_mmio: support non-legacy balloon devices
  virtio_ccw: support non-legacy balloon devices
  virtio: balloon might not be a legacy device
  virtio_balloon: transitional interface
  virtio_ring: Update weak barriers to use dma_wmb/rmb
  virtio_pci_modern: switch to type-safe io accessors
  virtio_pci_modern: type-safe io accessors
  lguest: handle traps on the "interrupt suppressed" iret instruction.
  virtio: drop a useless config read
  virtio_config: reorder functions
  Add virtio-input driver.
  lguest: suppress interrupts for single insn, not range.
  lguest: simplify lguest_iret
  lguest: rename i386_head.S in the comments
  lguest: explicitly set miscdevice's private_data NULL
  lguest: fix pending interrupt test.

12 files changed, 578 insertions, 105 deletions

diff --git a/drivers/lguest/hypercalls.c b/drivers/lguest/hypercalls.c
index 1219af493c0f..19a32280731d 100644
--- a/drivers/lguest/hypercalls.c
+++ b/drivers/lguest/hypercalls.c
@@ -211,10 +211,9 @@ static void initialize(struct lg_cpu *cpu)
 
 	/*
 	 * The Guest tells us where we're not to deliver interrupts by putting
-	 * the range of addresses into "struct lguest_data".
+	 * the instruction address into "struct lguest_data".
 	 */
-	if (get_user(cpu->lg->noirq_start, &cpu->lg->lguest_data->noirq_start)
-	    || get_user(cpu->lg->noirq_end, &cpu->lg->lguest_data->noirq_end))
+	if (get_user(cpu->lg->noirq_iret, &cpu->lg->lguest_data->noirq_iret))
 		kill_guest(cpu, "bad guest page %p", cpu->lg->lguest_data);
 
 	/*
diff --git a/drivers/lguest/interrupts_and_traps.c b/drivers/lguest/interrupts_and_traps.c
index 70dfcdc29f1f..5e7559be222a 100644
--- a/drivers/lguest/interrupts_and_traps.c
+++ b/drivers/lguest/interrupts_and_traps.c
@@ -56,21 +56,16 @@ static void push_guest_stack(struct lg_cpu *cpu, unsigned long *gstack, u32 val)
 }
 
 /*H:210
- * The set_guest_interrupt() routine actually delivers the interrupt or
- * trap.  The mechanics of delivering traps and interrupts to the Guest are the
- * same, except some traps have an "error code" which gets pushed onto the
- * stack as well: the caller tells us if this is one.
- *
- * "lo" and "hi" are the two parts of the Interrupt Descriptor Table for this
- * interrupt or trap.  It's split into two parts for traditional reasons: gcc
- * on i386 used to be frightened by 64 bit numbers.
+ * The push_guest_interrupt_stack() routine saves Guest state on the stack for
+ * an interrupt or trap.  The mechanics of delivering traps and interrupts to
+ * the Guest are the same, except some traps have an "error code" which gets
+ * pushed onto the stack as well: the caller tells us if this is one.
  *
  * We set up the stack just like the CPU does for a real interrupt, so it's
  * identical for the Guest (and the standard "iret" instruction will undo
  * it).
  */
-static void set_guest_interrupt(struct lg_cpu *cpu, u32 lo, u32 hi,
-				bool has_err)
+static void push_guest_interrupt_stack(struct lg_cpu *cpu, bool has_err)
 {
 	unsigned long gstack, origstack;
 	u32 eflags, ss, irq_enable;
@@ -130,12 +125,28 @@ static void set_guest_interrupt(struct lg_cpu *cpu, u32 lo, u32 hi,
 	if (has_err)
 		push_guest_stack(cpu, &gstack, cpu->regs->errcode);
 
-	/*
-	 * Now we've pushed all the old state, we change the stack, the code
-	 * segment and the address to execute.
-	 */
+	/* Adjust the stack pointer and stack segment. */
 	cpu->regs->ss = ss;
 	cpu->regs->esp = virtstack + (gstack - origstack);
+}
+
+/*
+ * This actually makes the Guest start executing the given interrupt/trap
+ * handler.
+ *
+ * "lo" and "hi" are the two parts of the Interrupt Descriptor Table for this
+ * interrupt or trap.  It's split into two parts for traditional reasons: gcc
+ * on i386 used to be frightened by 64 bit numbers.
+ */
+static void guest_run_interrupt(struct lg_cpu *cpu, u32 lo, u32 hi)
+{
+	/* If we're already in the kernel, we don't change stacks. */
+	if ((cpu->regs->ss&0x3) != GUEST_PL)
+		cpu->regs->ss = cpu->esp1;
+
+	/*
+	 * Set the code segment and the address to execute.
+	 */
 	cpu->regs->cs = (__KERNEL_CS|GUEST_PL);
 	cpu->regs->eip = idt_address(lo, hi);
 
@@ -158,6 +169,24 @@ static void set_guest_interrupt(struct lg_cpu *cpu, u32 lo, u32 hi,
 			kill_guest(cpu, "Disabling interrupts");
 }
 
+/* This restores the eflags word which was pushed on the stack by a trap */
+static void restore_eflags(struct lg_cpu *cpu)
+{
+	/* This is the physical address of the stack. */
+	unsigned long stack_pa = guest_pa(cpu, cpu->regs->esp);
+
+	/*
+	 * Stack looks like this:
+	 * Address	Contents
+	 * esp		EIP
+	 * esp + 4	CS
+	 * esp + 8	EFLAGS
+	 */
+	cpu->regs->eflags = lgread(cpu, stack_pa + 8, u32);
+	cpu->regs->eflags &=
+		~(X86_EFLAGS_TF|X86_EFLAGS_VM|X86_EFLAGS_RF|X86_EFLAGS_NT);
+}
+
 /*H:205
  * Virtual Interrupts.
  *
@@ -200,14 +229,6 @@ void try_deliver_interrupt(struct lg_cpu *cpu, unsigned int irq, bool more)
 
 	BUG_ON(irq >= LGUEST_IRQS);
 
-	/*
-	 * They may be in the middle of an iret, where they asked us never to
-	 * deliver interrupts.
-	 */
-	if (cpu->regs->eip >= cpu->lg->noirq_start &&
-	   (cpu->regs->eip < cpu->lg->noirq_end))
-		return;
-
 	/* If they're halted, interrupts restart them. */
 	if (cpu->halted) {
 		/* Re-enable interrupts. */
@@ -237,12 +258,34 @@ void try_deliver_interrupt(struct lg_cpu *cpu, unsigned int irq, bool more)
 	if (idt_present(idt->a, idt->b)) {
 		/* OK, mark it no longer pending and deliver it. */
 		clear_bit(irq, cpu->irqs_pending);
+
 		/*
-		 * set_guest_interrupt() takes the interrupt descriptor and a
-		 * flag to say whether this interrupt pushes an error code onto
-		 * the stack as well: virtual interrupts never do.
+		 * They may be about to iret, where they asked us never to
+		 * deliver interrupts.  In this case, we can emulate that iret
+		 * then immediately deliver the interrupt.  This is basically
+		 * a noop: the iret would pop the interrupt frame and restore
+		 * eflags, and then we'd set it up again.  So just restore the
+		 * eflags word and jump straight to the handler in this case.
+		 *
+		 * Denys Vlasenko points out that this isn't quite right: if
+		 * the iret was returning to userspace, then that interrupt
+		 * would reset the stack pointer (which the Guest told us
+		 * about via LHCALL_SET_STACK).  But unless the Guest is being
+		 * *really* weird, that will be the same as the current stack
+		 * anyway.
 		 */
-		set_guest_interrupt(cpu, idt->a, idt->b, false);
+		if (cpu->regs->eip == cpu->lg->noirq_iret) {
+			restore_eflags(cpu);
+		} else {
+			/*
+			 * set_guest_interrupt() takes a flag to say whether
+			 * this interrupt pushes an error code onto the stack
+			 * as well: virtual interrupts never do.
+			 */
+			push_guest_interrupt_stack(cpu, false);
+		}
+		/* Actually make Guest cpu jump to handler. */
+		guest_run_interrupt(cpu, idt->a, idt->b);
 	}
 
 	/*
@@ -353,8 +396,9 @@ bool deliver_trap(struct lg_cpu *cpu, unsigned int num)
 	 */
 	if (!idt_present(cpu->arch.idt[num].a, cpu->arch.idt[num].b))
 		return false;
-	set_guest_interrupt(cpu, cpu->arch.idt[num].a,
-			    cpu->arch.idt[num].b, has_err(num));
+	push_guest_interrupt_stack(cpu, has_err(num));
+	guest_run_interrupt(cpu, cpu->arch.idt[num].a,
+			    cpu->arch.idt[num].b);
 	return true;
 }
 
@@ -395,8 +439,9 @@ static bool direct_trap(unsigned int num)
  * The Guest has the ability to turn its interrupt gates into trap gates,
  * if it is careful.  The Host will let trap gates can go directly to the
  * Guest, but the Guest needs the interrupts atomically disabled for an
- * interrupt gate.  It can do this by pointing the trap gate at instructions
- * within noirq_start and noirq_end, where it can safely disable interrupts.
+ * interrupt gate.  The Host could provide a mechanism to register more
+ * "no-interrupt" regions, and the Guest could point the trap gate at
+ * instructions within that region, where it can safely disable interrupts.
  */
 
 /*M:006
diff --git a/drivers/lguest/lg.h b/drivers/lguest/lg.h
index 307e8b39e7d1..ac8ad0461e80 100644
--- a/drivers/lguest/lg.h
+++ b/drivers/lguest/lg.h
@@ -102,7 +102,7 @@ struct lguest {
 
 	struct pgdir pgdirs[4];
 
-	unsigned long noirq_start, noirq_end;
+	unsigned long noirq_iret;
 
 	unsigned int stack_pages;
 	u32 tsc_khz;
diff --git a/drivers/lguest/lguest_user.c b/drivers/lguest/lguest_user.c
index c4c6113eb9a6..30c60687d277 100644
--- a/drivers/lguest/lguest_user.c
+++ b/drivers/lguest/lguest_user.c
@@ -339,6 +339,13 @@ static ssize_t write(struct file *file, const char __user *in,
 	}
 }
 
+static int open(struct inode *inode, struct file *file)
+{
+	file->private_data = NULL;
+
+	return 0;
+}
+
 /*L:060
  * The final piece of interface code is the close() routine.  It reverses
  * everything done in initialize().  This is usually called because the
@@ -409,6 +416,7 @@ static int close(struct inode *inode, struct file *file)
  */
 static const struct file_operations lguest_fops = {
 	.owner	 = THIS_MODULE,
+	.open	 = open,
 	.release = close,
 	.write	 = write,
 	.read	 = read,
diff --git a/drivers/s390/kvm/virtio_ccw.c b/drivers/s390/kvm/virtio_ccw.c
index 71d7802aa8b4..6f1fa1773e76 100644
--- a/drivers/s390/kvm/virtio_ccw.c
+++ b/drivers/s390/kvm/virtio_ccw.c
@@ -1201,13 +1201,9 @@ static int virtio_ccw_online(struct ccw_device *cdev)
 	vcdev->vdev.id.vendor = cdev->id.cu_type;
 	vcdev->vdev.id.device = cdev->id.cu_model;
 
-	if (virtio_device_is_legacy_only(vcdev->vdev.id)) {
-		vcdev->revision = 0;
-	} else {
-		ret = virtio_ccw_set_transport_rev(vcdev);
-		if (ret)
-			goto out_free;
-	}
+	ret = virtio_ccw_set_transport_rev(vcdev);
+	if (ret)
+		goto out_free;
 
 	ret = register_virtio_device(&vcdev->vdev);
 	if (ret) {
diff --git a/drivers/virtio/Kconfig b/drivers/virtio/Kconfig
index b546da5d8ea3..cab9f3f63a38 100644
--- a/drivers/virtio/Kconfig
+++ b/drivers/virtio/Kconfig
@@ -48,6 +48,16 @@ config VIRTIO_BALLOON
 
 	 If unsure, say M.
 
+config VIRTIO_INPUT
+	tristate "Virtio input driver"
+	depends on VIRTIO
+	depends on INPUT
+	---help---
+	 This driver supports virtio input devices such as
+	 keyboards, mice and tablets.
+
+	 If unsure, say M.
+
  config VIRTIO_MMIO
 	tristate "Platform bus driver for memory mapped virtio devices"
 	depends on HAS_IOMEM
diff --git a/drivers/virtio/Makefile b/drivers/virtio/Makefile
index d85565b8ea46..41e30e3dc842 100644
--- a/drivers/virtio/Makefile
+++ b/drivers/virtio/Makefile
@@ -4,3 +4,4 @@ obj-$(CONFIG_VIRTIO_PCI) += virtio_pci.o
 virtio_pci-y := virtio_pci_modern.o virtio_pci_common.o
 virtio_pci-$(CONFIG_VIRTIO_PCI_LEGACY) += virtio_pci_legacy.o
 obj-$(CONFIG_VIRTIO_BALLOON) += virtio_balloon.o
+obj-$(CONFIG_VIRTIO_INPUT) += virtio_input.o
diff --git a/drivers/virtio/virtio.c b/drivers/virtio/virtio.c
index 5ce2aa48fc6e..b1877d73fa56 100644
--- a/drivers/virtio/virtio.c
+++ b/drivers/virtio/virtio.c
@@ -278,12 +278,6 @@ static struct bus_type virtio_bus = {
 	.remove = virtio_dev_remove,
 };
 
-bool virtio_device_is_legacy_only(struct virtio_device_id id)
-{
-	return id.device == VIRTIO_ID_BALLOON;
-}
-EXPORT_SYMBOL_GPL(virtio_device_is_legacy_only);
-
 int register_virtio_driver(struct virtio_driver *driver)
 {
 	/* Catch this early. */
diff --git a/drivers/virtio/virtio_balloon.c b/drivers/virtio/virtio_balloon.c
index 6a356e344f82..82e80e034f25 100644
--- a/drivers/virtio/virtio_balloon.c
+++ b/drivers/virtio/virtio_balloon.c
@@ -214,8 +214,8 @@ static inline void update_stat(struct virtio_balloon *vb, int idx,
 			       u16 tag, u64 val)
 {
 	BUG_ON(idx >= VIRTIO_BALLOON_S_NR);
-	vb->stats[idx].tag = tag;
-	vb->stats[idx].val = val;
+	vb->stats[idx].tag = cpu_to_virtio16(vb->vdev, tag);
+	vb->stats[idx].val = cpu_to_virtio64(vb->vdev, val);
 }
 
 #define pages_to_bytes(x) ((u64)(x) << PAGE_SHIFT)
@@ -283,18 +283,27 @@ static void virtballoon_changed(struct virtio_device *vdev)
 
 static inline s64 towards_target(struct virtio_balloon *vb)
 {
-	__le32 v;
 	s64 target;
+	u32 num_pages;
 
-	virtio_cread(vb->vdev, struct virtio_balloon_config, num_pages, &v);
+	virtio_cread(vb->vdev, struct virtio_balloon_config, num_pages,
+		     &num_pages);
 
-	target = le32_to_cpu(v);
+	/* Legacy balloon config space is LE, unlike all other devices. */
+	if (!virtio_has_feature(vb->vdev, VIRTIO_F_VERSION_1))
+		num_pages = le32_to_cpu((__force __le32)num_pages);
+
+	target = num_pages;
 	return target - vb->num_pages;
 }
 
 static void update_balloon_size(struct virtio_balloon *vb)
 {
-	__le32 actual = cpu_to_le32(vb->num_pages);
+	u32 actual = vb->num_pages;
+
+	/* Legacy balloon config space is LE, unlike all other devices. */
+	if (!virtio_has_feature(vb->vdev, VIRTIO_F_VERSION_1))
+		actual = (__force u32)cpu_to_le32(actual);
 
 	virtio_cwrite(vb->vdev, struct virtio_balloon_config, actual,
 		      &actual);
diff --git a/drivers/virtio/virtio_input.c b/drivers/virtio/virtio_input.c
new file mode 100644
index 000000000000..60e2a1677563
--- /dev/null
+++ b/drivers/virtio/virtio_input.c
@@ -0,0 +1,384 @@
+#include <linux/module.h>
+#include <linux/virtio.h>
+#include <linux/virtio_config.h>
+#include <linux/input.h>
+
+#include <uapi/linux/virtio_ids.h>
+#include <uapi/linux/virtio_input.h>
+
+struct virtio_input {
+	struct virtio_device       *vdev;
+	struct input_dev           *idev;
+	char                       name[64];
+	char                       serial[64];
+	char                       phys[64];
+	struct virtqueue           *evt, *sts;
+	struct virtio_input_event  evts[64];
+	spinlock_t                 lock;
+	bool                       ready;
+};
+
+static void virtinput_queue_evtbuf(struct virtio_input *vi,
+				   struct virtio_input_event *evtbuf)
+{
+	struct scatterlist sg[1];
+
+	sg_init_one(sg, evtbuf, sizeof(*evtbuf));
+	virtqueue_add_inbuf(vi->evt, sg, 1, evtbuf, GFP_ATOMIC);
+}
+
+static void virtinput_recv_events(struct virtqueue *vq)
+{
+	struct virtio_input *vi = vq->vdev->priv;
+	struct virtio_input_event *event;
+	unsigned long flags;
+	unsigned int len;
+
+	spin_lock_irqsave(&vi->lock, flags);
+	if (vi->ready) {
+		while ((event = virtqueue_get_buf(vi->evt, &len)) != NULL) {
+			spin_unlock_irqrestore(&vi->lock, flags);
+			input_event(vi->idev,
+				    le16_to_cpu(event->type),
+				    le16_to_cpu(event->code),
+				    le32_to_cpu(event->value));
+			spin_lock_irqsave(&vi->lock, flags);
+			virtinput_queue_evtbuf(vi, event);
+		}
+		virtqueue_kick(vq);
+	}
+	spin_unlock_irqrestore(&vi->lock, flags);
+}
+
+/*
+ * On error we are losing the status update, which isn't critical as
+ * this is typically used for stuff like keyboard leds.
+ */
+static int virtinput_send_status(struct virtio_input *vi,
+				 u16 type, u16 code, s32 value)
+{
+	struct virtio_input_event *stsbuf;
+	struct scatterlist sg[1];
+	unsigned long flags;
+	int rc;
+
+	stsbuf = kzalloc(sizeof(*stsbuf), GFP_ATOMIC);
+	if (!stsbuf)
+		return -ENOMEM;
+
+	stsbuf->type  = cpu_to_le16(type);
+	stsbuf->code  = cpu_to_le16(code);
+	stsbuf->value = cpu_to_le32(value);
+	sg_init_one(sg, stsbuf, sizeof(*stsbuf));
+
+	spin_lock_irqsave(&vi->lock, flags);
+	if (vi->ready) {
+		rc = virtqueue_add_outbuf(vi->sts, sg, 1, stsbuf, GFP_ATOMIC);
+		virtqueue_kick(vi->sts);
+	} else {
+		rc = -ENODEV;
+	}
+	spin_unlock_irqrestore(&vi->lock, flags);
+
+	if (rc != 0)
+		kfree(stsbuf);
+	return rc;
+}
+
+static void virtinput_recv_status(struct virtqueue *vq)
+{
+	struct virtio_input *vi = vq->vdev->priv;
+	struct virtio_input_event *stsbuf;
+	unsigned long flags;
+	unsigned int len;
+
+	spin_lock_irqsave(&vi->lock, flags);
+	while ((stsbuf = virtqueue_get_buf(vi->sts, &len)) != NULL)
+		kfree(stsbuf);
+	spin_unlock_irqrestore(&vi->lock, flags);
+}
+
+static int virtinput_status(struct input_dev *idev, unsigned int type,
+			    unsigned int code, int value)
+{
+	struct virtio_input *vi = input_get_drvdata(idev);
+
+	return virtinput_send_status(vi, type, code, value);
+}
+
+static u8 virtinput_cfg_select(struct virtio_input *vi,
+			       u8 select, u8 subsel)
+{
+	u8 size;
+
+	virtio_cwrite(vi->vdev, struct virtio_input_config, select, &select);
+	virtio_cwrite(vi->vdev, struct virtio_input_config, subsel, &subsel);
+	virtio_cread(vi->vdev, struct virtio_input_config, size, &size);
+	return size;
+}
+
+static void virtinput_cfg_bits(struct virtio_input *vi, int select, int subsel,
+			       unsigned long *bits, unsigned int bitcount)
+{
+	unsigned int bit;
+	u8 *virtio_bits;
+	u8 bytes;
+
+	bytes = virtinput_cfg_select(vi, select, subsel);
+	if (!bytes)
+		return;
+	if (bitcount > bytes * 8)
+		bitcount = bytes * 8;
+
+	/*
+	 * Bitmap in virtio config space is a simple stream of bytes,
+	 * with the first byte carrying bits 0-7, second bits 8-15 and
+	 * so on.
+	 */
+	virtio_bits = kzalloc(bytes, GFP_KERNEL);
+	if (!virtio_bits)
+		return;
+	virtio_cread_bytes(vi->vdev, offsetof(struct virtio_input_config,
+					      u.bitmap),
+			   virtio_bits, bytes);
+	for (bit = 0; bit < bitcount; bit++) {
+		if (virtio_bits[bit / 8] & (1 << (bit % 8)))
+			__set_bit(bit, bits);
+	}
+	kfree(virtio_bits);
+
+	if (select == VIRTIO_INPUT_CFG_EV_BITS)
+		__set_bit(subsel, vi->idev->evbit);
+}
+
+static void virtinput_cfg_abs(struct virtio_input *vi, int abs)
+{
+	u32 mi, ma, re, fu, fl;
+
+	virtinput_cfg_select(vi, VIRTIO_INPUT_CFG_ABS_INFO, abs);
+	virtio_cread(vi->vdev, struct virtio_input_config, u.abs.min, &mi);
+	virtio_cread(vi->vdev, struct virtio_input_config, u.abs.max, &ma);
+	virtio_cread(vi->vdev, struct virtio_input_config, u.abs.res, &re);
+	virtio_cread(vi->vdev, struct virtio_input_config, u.abs.fuzz, &fu);
+	virtio_cread(vi->vdev, struct virtio_input_config, u.abs.flat, &fl);
+	input_set_abs_params(vi->idev, abs, mi, ma, fu, fl);
+	input_abs_set_res(vi->idev, abs, re);
+}
+
+static int virtinput_init_vqs(struct virtio_input *vi)
+{
+	struct virtqueue *vqs[2];
+	vq_callback_t *cbs[] = { virtinput_recv_events,
+				 virtinput_recv_status };
+	static const char *names[] = { "events", "status" };
+	int err;
+
+	err = vi->vdev->config->find_vqs(vi->vdev, 2, vqs, cbs, names);
+	if (err)
+		return err;
+	vi->evt = vqs[0];
+	vi->sts = vqs[1];
+
+	return 0;
+}
+
+static void virtinput_fill_evt(struct virtio_input *vi)
+{
+	unsigned long flags;
+	int i, size;
+
+	spin_lock_irqsave(&vi->lock, flags);
+	size = virtqueue_get_vring_size(vi->evt);
+	if (size > ARRAY_SIZE(vi->evts))
+		size = ARRAY_SIZE(vi->evts);
+	for (i = 0; i < size; i++)
+		virtinput_queue_evtbuf(vi, &vi->evts[i]);
+	virtqueue_kick(vi->evt);
+	spin_unlock_irqrestore(&vi->lock, flags);
+}
+
+static int virtinput_probe(struct virtio_device *vdev)
+{
+	struct virtio_input *vi;
+	unsigned long flags;
+	size_t size;
+	int abs, err;
+
+	if (!virtio_has_feature(vdev, VIRTIO_F_VERSION_1))
+		return -ENODEV;
+
+	vi = kzalloc(sizeof(*vi), GFP_KERNEL);
+	if (!vi)
+		return -ENOMEM;
+
+	vdev->priv = vi;
+	vi->vdev = vdev;
+	spin_lock_init(&vi->lock);
+
+	err = virtinput_init_vqs(vi);
+	if (err)
+		goto err_init_vq;
+
+	vi->idev = input_allocate_device();
+	if (!vi->idev) {
+		err = -ENOMEM;
+		goto err_input_alloc;
+	}
+	input_set_drvdata(vi->idev, vi);
+
+	size = virtinput_cfg_select(vi, VIRTIO_INPUT_CFG_ID_NAME, 0);
+	virtio_cread_bytes(vi->vdev, offsetof(struct virtio_input_config,
+					      u.string),
+			   vi->name, min(size, sizeof(vi->name)));
+	size = virtinput_cfg_select(vi, VIRTIO_INPUT_CFG_ID_SERIAL, 0);
+	virtio_cread_bytes(vi->vdev, offsetof(struct virtio_input_config,
+					      u.string),
+			   vi->serial, min(size, sizeof(vi->serial)));
+	snprintf(vi->phys, sizeof(vi->phys),
+		 "virtio%d/input0", vdev->index);
+	vi->idev->name = vi->name;
+	vi->idev->phys = vi->phys;
+	vi->idev->uniq = vi->serial;
+
+	size = virtinput_cfg_select(vi, VIRTIO_INPUT_CFG_ID_DEVIDS, 0);
+	if (size >= sizeof(struct virtio_input_devids)) {
+		virtio_cread(vi->vdev, struct virtio_input_config,
+			     u.ids.bustype, &vi->idev->id.bustype);
+		virtio_cread(vi->vdev, struct virtio_input_config,
+			     u.ids.vendor, &vi->idev->id.vendor);
+		virtio_cread(vi->vdev, struct virtio_input_config,
+			     u.ids.product, &vi->idev->id.product);
+		virtio_cread(vi->vdev, struct virtio_input_config,
+			     u.ids.version, &vi->idev->id.version);
+	} else {
+		vi->idev->id.bustype = BUS_VIRTUAL;
+	}
+
+	virtinput_cfg_bits(vi, VIRTIO_INPUT_CFG_PROP_BITS, 0,
+			   vi->idev->propbit, INPUT_PROP_CNT);
+	size = virtinput_cfg_select(vi, VIRTIO_INPUT_CFG_EV_BITS, EV_REP);
+	if (size)
+		__set_bit(EV_REP, vi->idev->evbit);
+
+	vi->idev->dev.parent = &vdev->dev;
+	vi->idev->event = virtinput_status;
+
+	/* device -> kernel */
+	virtinput_cfg_bits(vi, VIRTIO_INPUT_CFG_EV_BITS, EV_KEY,
+			   vi->idev->keybit, KEY_CNT);
+	virtinput_cfg_bits(vi, VIRTIO_INPUT_CFG_EV_BITS, EV_REL,
+			   vi->idev->relbit, REL_CNT);
+	virtinput_cfg_bits(vi, VIRTIO_INPUT_CFG_EV_BITS, EV_ABS,
+			   vi->idev->absbit, ABS_CNT);
+	virtinput_cfg_bits(vi, VIRTIO_INPUT_CFG_EV_BITS, EV_MSC,
+			   vi->idev->mscbit, MSC_CNT);
+	virtinput_cfg_bits(vi, VIRTIO_INPUT_CFG_EV_BITS, EV_SW,
+			   vi->idev->swbit,  SW_CNT);
+
+	/* kernel -> device */
+	virtinput_cfg_bits(vi, VIRTIO_INPUT_CFG_EV_BITS, EV_LED,
+			   vi->idev->ledbit, LED_CNT);
+	virtinput_cfg_bits(vi, VIRTIO_INPUT_CFG_EV_BITS, EV_SND,
+			   vi->idev->sndbit, SND_CNT);
+
+	if (test_bit(EV_ABS, vi->idev->evbit)) {
+		for (abs = 0; abs < ABS_CNT; abs++) {
+			if (!test_bit(abs, vi->idev->absbit))
+				continue;
+			virtinput_cfg_abs(vi, abs);
+		}
+	}
+
+	virtio_device_ready(vdev);
+	vi->ready = true;
+	err = input_register_device(vi->idev);
+	if (err)
+		goto err_input_register;
+
+	virtinput_fill_evt(vi);
+	return 0;
+
+err_input_register:
+	spin_lock_irqsave(&vi->lock, flags);
+	vi->ready = false;
+	spin_unlock_irqrestore(&vi->lock, flags);
+	input_free_device(vi->idev);
+err_input_alloc:
+	vdev->config->del_vqs(vdev);
+err_init_vq:
+	kfree(vi);
+	return err;
+}
+
+static void virtinput_remove(struct virtio_device *vdev)
+{
+	struct virtio_input *vi = vdev->priv;
+	unsigned long flags;
+
+	spin_lock_irqsave(&vi->lock, flags);
+	vi->ready = false;
+	spin_unlock_irqrestore(&vi->lock, flags);
+
+	input_unregister_device(vi->idev);
+	vdev->config->del_vqs(vdev);
+	kfree(vi);
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int virtinput_freeze(struct virtio_device *vdev)
+{
+	struct virtio_input *vi = vdev->priv;
+	unsigned long flags;
+
+	spin_lock_irqsave(&vi->lock, flags);
+	vi->ready = false;
+	spin_unlock_irqrestore(&vi->lock, flags);
+
+	vdev->config->del_vqs(vdev);
+	return 0;
+}
+
+static int virtinput_restore(struct virtio_device *vdev)
+{
+	struct virtio_input *vi = vdev->priv;
+	int err;
+
+	err = virtinput_init_vqs(vi);
+	if (err)
+		return err;
+
+	virtio_device_ready(vdev);
+	vi->ready = true;
+	virtinput_fill_evt(vi);
+	return 0;
+}
+#endif
+
+static unsigned int features[] = {
+	/* none */
+};
+static struct virtio_device_id id_table[] = {
+	{ VIRTIO_ID_INPUT, VIRTIO_DEV_ANY_ID },
+	{ 0 },
+};
+
+static struct virtio_driver virtio_input_driver = {
+	.driver.name         = KBUILD_MODNAME,
+	.driver.owner        = THIS_MODULE,
+	.feature_table       = features,
+	.feature_table_size  = ARRAY_SIZE(features),
+	.id_table            = id_table,
+	.probe               = virtinput_probe,
+	.remove              = virtinput_remove,
+#ifdef CONFIG_PM_SLEEP
+	.freeze	             = virtinput_freeze,
+	.restore             = virtinput_restore,
+#endif
+};
+
+module_virtio_driver(virtio_input_driver);
+MODULE_DEVICE_TABLE(virtio, id_table);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Virtio input device driver");
+MODULE_AUTHOR("Gerd Hoffmann <kraxel@redhat.com>");
diff --git a/drivers/virtio/virtio_mmio.c b/drivers/virtio/virtio_mmio.c
index 6010d7ec0a0f..7a5e60dea6c5 100644
--- a/drivers/virtio/virtio_mmio.c
+++ b/drivers/virtio/virtio_mmio.c
@@ -581,14 +581,6 @@ static int virtio_mmio_probe(struct platform_device *pdev)
 	}
 	vm_dev->vdev.id.vendor = readl(vm_dev->base + VIRTIO_MMIO_VENDOR_ID);
 
-	/* Reject legacy-only IDs for version 2 devices */
-	if (vm_dev->version == 2 &&
-			virtio_device_is_legacy_only(vm_dev->vdev.id)) {
-		dev_err(&pdev->dev, "Version 2 not supported for devices %u!\n",
-				vm_dev->vdev.id.device);
-		return -ENODEV;
-	}
-
 	if (vm_dev->version == 1)
 		writel(PAGE_SIZE, vm_dev->base + VIRTIO_MMIO_GUEST_PAGE_SIZE);
 
diff --git a/drivers/virtio/virtio_pci_modern.c b/drivers/virtio/virtio_pci_modern.c
index 2aa38e59db2e..e88e0997a889 100644
--- a/drivers/virtio/virtio_pci_modern.c
+++ b/drivers/virtio/virtio_pci_modern.c
@@ -20,6 +20,50 @@
 #define VIRTIO_PCI_NO_LEGACY
 #include "virtio_pci_common.h"
 
+/*
+ * Type-safe wrappers for io accesses.
+ * Use these to enforce at compile time the following spec requirement:
+ *
+ * The driver MUST access each field using the “natural” access
+ * method, i.e. 32-bit accesses for 32-bit fields, 16-bit accesses
+ * for 16-bit fields and 8-bit accesses for 8-bit fields.
+ */
+static inline u8 vp_ioread8(u8 __iomem *addr)
+{
+	return ioread8(addr);
+}
+static inline u16 vp_ioread16 (u16 __iomem *addr)
+{
+	return ioread16(addr);
+}
+
+static inline u32 vp_ioread32(u32 __iomem *addr)
+{
+	return ioread32(addr);
+}
+
+static inline void vp_iowrite8(u8 value, u8 __iomem *addr)
+{
+	iowrite8(value, addr);
+}
+
+static inline void vp_iowrite16(u16 value, u16 __iomem *addr)
+{
+	iowrite16(value, addr);
+}
+
+static inline void vp_iowrite32(u32 value, u32 __iomem *addr)
+{
+	iowrite32(value, addr);
+}
+
+static void vp_iowrite64_twopart(u64 val,
+				 __le32 __iomem *lo, __le32 __iomem *hi)
+{
+	vp_iowrite32((u32)val, lo);
+	vp_iowrite32(val >> 32, hi);
+}
+
 static void __iomem *map_capability(struct pci_dev *dev, int off,
 				    size_t minlen,
 				    u32 align,
@@ -94,22 +138,16 @@ static void __iomem *map_capability(struct pci_dev *dev, int off,
 	return p;
 }
 
-static void iowrite64_twopart(u64 val, __le32 __iomem *lo, __le32 __iomem *hi)
-{
-	iowrite32((u32)val, lo);
-	iowrite32(val >> 32, hi);
-}
-
 /* virtio config->get_features() implementation */
 static u64 vp_get_features(struct virtio_device *vdev)
 {
 	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
 	u64 features;
 
-	iowrite32(0, &vp_dev->common->device_feature_select);
-	features = ioread32(&vp_dev->common->device_feature);
-	iowrite32(1, &vp_dev->common->device_feature_select);
-	features |= ((u64)ioread32(&vp_dev->common->device_feature) << 32);
+	vp_iowrite32(0, &vp_dev->common->device_feature_select);
+	features = vp_ioread32(&vp_dev->common->device_feature);
+	vp_iowrite32(1, &vp_dev->common->device_feature_select);
+	features |= ((u64)vp_ioread32(&vp_dev->common->device_feature) << 32);
 
 	return features;
 }
@@ -128,10 +166,10 @@ static int vp_finalize_features(struct virtio_device *vdev)
 		return -EINVAL;
 	}
 
-	iowrite32(0, &vp_dev->common->guest_feature_select);
-	iowrite32((u32)vdev->features, &vp_dev->common->guest_feature);
-	iowrite32(1, &vp_dev->common->guest_feature_select);
-	iowrite32(vdev->features >> 32, &vp_dev->common->guest_feature);
+	vp_iowrite32(0, &vp_dev->common->guest_feature_select);
+	vp_iowrite32((u32)vdev->features, &vp_dev->common->guest_feature);
+	vp_iowrite32(1, &vp_dev->common->guest_feature_select);
+	vp_iowrite32(vdev->features >> 32, &vp_dev->common->guest_feature);
 
 	return 0;
 }
@@ -210,14 +248,14 @@ static void vp_set(struct virtio_device *vdev, unsigned offset,
 static u32 vp_generation(struct virtio_device *vdev)
 {
 	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
-	return ioread8(&vp_dev->common->config_generation);
+	return vp_ioread8(&vp_dev->common->config_generation);
 }
 
 /* config->{get,set}_status() implementations */
 static u8 vp_get_status(struct virtio_device *vdev)
 {
 	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
-	return ioread8(&vp_dev->common->device_status);
+	return vp_ioread8(&vp_dev->common->device_status);
 }
 
 static void vp_set_status(struct virtio_device *vdev, u8 status)
@@ -225,17 +263,17 @@ static void vp_set_status(struct virtio_device *vdev, u8 status)
 	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
 	/* We should never be setting status to 0. */
 	BUG_ON(status == 0);
-	iowrite8(status, &vp_dev->common->device_status);
+	vp_iowrite8(status, &vp_dev->common->device_status);
 }
 
 static void vp_reset(struct virtio_device *vdev)
 {
 	struct virtio_pci_device *vp_dev = to_vp_device(vdev);
 	/* 0 status means a reset. */
-	iowrite8(0, &vp_dev->common->device_status);
+	vp_iowrite8(0, &vp_dev->common->device_status);
 	/* Flush out the status write, and flush in device writes,
 	 * including MSI-X interrupts, if any. */
-	ioread8(&vp_dev->common->device_status);
+	vp_ioread8(&vp_dev->common->device_status);
 	/* Flush pending VQ/configuration callbacks. */
 	vp_synchronize_vectors(vdev);
 }
@@ -243,10 +281,10 @@ static void vp_reset(struct virtio_device *vdev)
 static u16 vp_config_vector(struct virtio_pci_device *vp_dev, u16 vector)
 {
 	/* Setup the vector used for configuration events */
-	iowrite16(vector, &vp_dev->common->msix_config);
+	vp_iowrite16(vector, &vp_dev->common->msix_config);
 	/* Verify we had enough resources to assign the vector */
 	/* Will also flush the write out to device */
-	return ioread16(&vp_dev->common->msix_config);
+	return vp_ioread16(&vp_dev->common->msix_config);
 }
 
 static size_t vring_pci_size(u16 num)
@@ -286,15 +324,15 @@ static struct virtqueue *setup_vq(struct virtio_pci_device *vp_dev,
 	u16 num, off;
 	int err;
 
-	if (index >= ioread16(&cfg->num_queues))
+	if (index >= vp_ioread16(&cfg->num_queues))
 		return ERR_PTR(-ENOENT);
 
 	/* Select the queue we're interested in */
-	iowrite16(index, &cfg->queue_select);
+	vp_iowrite16(index, &cfg->queue_select);
 
 	/* Check if queue is either not available or already active. */
-	num = ioread16(&cfg->queue_size);
-	if (!num || ioread16(&cfg->queue_enable))
+	num = vp_ioread16(&cfg->queue_size);
+	if (!num || vp_ioread16(&cfg->queue_enable))
 		return ERR_PTR(-ENOENT);
 
 	if (num & (num - 1)) {
@@ -303,7 +341,7 @@ static struct virtqueue *setup_vq(struct virtio_pci_device *vp_dev,
 	}
 
 	/* get offset of notification word for this vq */
-	off = ioread16(&cfg->queue_notify_off);
+	off = vp_ioread16(&cfg->queue_notify_off);
 
 	info->num = num;
 	info->msix_vector = msix_vec;
@@ -322,13 +360,13 @@ static struct virtqueue *setup_vq(struct virtio_pci_device *vp_dev,
 	}
 
 	/* activate the queue */
-	iowrite16(num, &cfg->queue_size);
-	iowrite64_twopart(virt_to_phys(info->queue),
-			  &cfg->queue_desc_lo, &cfg->queue_desc_hi);
-	iowrite64_twopart(virt_to_phys(virtqueue_get_avail(vq)),
-			  &cfg->queue_avail_lo, &cfg->queue_avail_hi);
-	iowrite64_twopart(virt_to_phys(virtqueue_get_used(vq)),
-			  &cfg->queue_used_lo, &cfg->queue_used_hi);
+	vp_iowrite16(num, &cfg->queue_size);
+	vp_iowrite64_twopart(virt_to_phys(info->queue),
+			     &cfg->queue_desc_lo, &cfg->queue_desc_hi);
+	vp_iowrite64_twopart(virt_to_phys(virtqueue_get_avail(vq)),
+			     &cfg->queue_avail_lo, &cfg->queue_avail_hi);
+	vp_iowrite64_twopart(virt_to_phys(virtqueue_get_used(vq)),
+			     &cfg->queue_used_lo, &cfg->queue_used_hi);
 
 	if (vp_dev->notify_base) {
 		/* offset should not wrap */
@@ -357,8 +395,8 @@ static struct virtqueue *setup_vq(struct virtio_pci_device *vp_dev,
 	}
 
 	if (msix_vec != VIRTIO_MSI_NO_VECTOR) {
-		iowrite16(msix_vec, &cfg->queue_msix_vector);
-		msix_vec = ioread16(&cfg->queue_msix_vector);
+		vp_iowrite16(msix_vec, &cfg->queue_msix_vector);
+		msix_vec = vp_ioread16(&cfg->queue_msix_vector);
 		if (msix_vec == VIRTIO_MSI_NO_VECTOR) {
 			err = -EBUSY;
 			goto err_assign_vector;
@@ -393,8 +431,8 @@ static int vp_modern_find_vqs(struct virtio_device *vdev, unsigned nvqs,
 	 * this, there's no way to go back except reset.
 	 */
 	list_for_each_entry(vq, &vdev->vqs, list) {
-		iowrite16(vq->index, &vp_dev->common->queue_select);
-		iowrite16(1, &vp_dev->common->queue_enable);
+		vp_iowrite16(vq->index, &vp_dev->common->queue_select);
+		vp_iowrite16(1, &vp_dev->common->queue_enable);
 	}
 
 	return 0;
@@ -405,13 +443,13 @@ static void del_vq(struct virtio_pci_vq_info *info)
 	struct virtqueue *vq = info->vq;
 	struct virtio_pci_device *vp_dev = to_vp_device(vq->vdev);
 
-	iowrite16(vq->index, &vp_dev->common->queue_select);
+	vp_iowrite16(vq->index, &vp_dev->common->queue_select);
 
 	if (vp_dev->msix_enabled) {
-		iowrite16(VIRTIO_MSI_NO_VECTOR,
-			  &vp_dev->common->queue_msix_vector);
+		vp_iowrite16(VIRTIO_MSI_NO_VECTOR,
+			     &vp_dev->common->queue_msix_vector);
 		/* Flush the write out to device */
-		ioread16(&vp_dev->common->queue_msix_vector);
+		vp_ioread16(&vp_dev->common->queue_msix_vector);
 	}
 
 	if (!vp_dev->notify_base)
@@ -577,9 +615,6 @@ int virtio_pci_modern_probe(struct virtio_pci_device *vp_dev)
 	}
 	vp_dev->vdev.id.vendor = pci_dev->subsystem_vendor;
 
-	if (virtio_device_is_legacy_only(vp_dev->vdev.id))
-		return -ENODEV;
-
 	/* check for a common config: if not, use legacy mode (bar 0). */
 	common = virtio_pci_find_capability(pci_dev, VIRTIO_PCI_CAP_COMMON_CFG,
 					    IORESOURCE_IO | IORESOURCE_MEM);