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.

MAINTAINERS

@@ -10517,6 +10517,12 @@ S:	Maintained
 F:	drivers/vhost/
 F:	include/uapi/linux/vhost.h
 
+VIRTIO INPUT DRIVER
+M:	Gerd Hoffmann <kraxel@redhat.com>
+S:	Maintained
+F:	drivers/virtio/virtio_input.c
+F:	include/uapi/linux/virtio_input.h
+
 VIA RHINE NETWORK DRIVER
 M:	Roger Luethi <rl@hellgate.ch>
 S:	Maintained

arch/x86/include/asm/lguest.h

@@ -20,13 +20,10 @@ extern unsigned long switcher_addr;
 /* Found in switcher.S */
 extern unsigned long default_idt_entries[];
 
-/* Declarations for definitions in lguest_guest.S */
-extern char lguest_noirq_start[], lguest_noirq_end[];
+/* Declarations for definitions in arch/x86/lguest/head_32.S */
+extern char lguest_noirq_iret[];
 extern const char lgstart_cli[], lgend_cli[];
-extern const char lgstart_sti[], lgend_sti[];
-extern const char lgstart_popf[], lgend_popf[];
 extern const char lgstart_pushf[], lgend_pushf[];
-extern const char lgstart_iret[], lgend_iret[];
 
 extern void lguest_iret(void);
 extern void lguest_init(void);

arch/x86/lguest/boot.c

@@ -87,8 +87,7 @@
 
 struct lguest_data lguest_data = {
 	.hcall_status = { [0 ... LHCALL_RING_SIZE-1] = 0xFF },
-	.noirq_start = (u32)lguest_noirq_start,
-	.noirq_end = (u32)lguest_noirq_end,
+	.noirq_iret = (u32)lguest_noirq_iret,
 	.kernel_address = PAGE_OFFSET,
 	.blocked_interrupts = { 1 }, /* Block timer interrupts */
 	.syscall_vec = SYSCALL_VECTOR,
@@ -262,7 +261,7 @@ PV_CALLEE_SAVE_REGS_THUNK(lguest_save_fl);
 PV_CALLEE_SAVE_REGS_THUNK(lguest_irq_disable);
 /*:*/
 
-/* These are in i386_head.S */
+/* These are in head_32.S */
 extern void lg_irq_enable(void);
 extern void lg_restore_fl(unsigned long flags);
 
@@ -1368,7 +1367,7 @@ static void lguest_restart(char *reason)
  * fit comfortably.
  *
  * First we need assembly templates of each of the patchable Guest operations,
- * and these are in i386_head.S.
+ * and these are in head_32.S.
  */
 
 /*G:060 We construct a table from the assembler templates: */

arch/x86/lguest/head_32.S

@@ -84,7 +84,7 @@ ENTRY(lg_irq_enable)
 	 * set lguest_data.irq_pending to X86_EFLAGS_IF.  If it's not zero, we
 	 * jump to send_interrupts, otherwise we're done.
 	 */
-	testl $0, lguest_data+LGUEST_DATA_irq_pending
+	cmpl $0, lguest_data+LGUEST_DATA_irq_pending
 	jnz send_interrupts
 	/*
 	 * One cool thing about x86 is that you can do many things without using
@@ -133,9 +133,8 @@ ENTRY(lg_restore_fl)
 	ret
 /*:*/
 
-/* These demark the EIP range where host should never deliver interrupts. */
-.global lguest_noirq_start
-.global lguest_noirq_end
+/* These demark the EIP where host should never deliver interrupts. */
+.global lguest_noirq_iret
 
 /*M:004
  * When the Host reflects a trap or injects an interrupt into the Guest, it
@@ -168,29 +167,26 @@ ENTRY(lg_restore_fl)
  * So we have to copy eflags from the stack to lguest_data.irq_enabled before
  * we do the "iret".
  *
- * There are two problems with this: firstly, we need to use a register to do
- * the copy and secondly, the whole thing needs to be atomic.  The first
- * problem is easy to solve: push %eax on the stack so we can use it, and then
- * restore it at the end just before the real "iret".
+ * There are two problems with this: firstly, we can't clobber any registers
+ * and secondly, the whole thing needs to be atomic.  The first problem
+ * is solved by using "push memory"/"pop memory" instruction pair for copying.
  *
  * The second is harder: copying eflags to lguest_data.irq_enabled will turn
  * interrupts on before we're finished, so we could be interrupted before we
- * return to userspace or wherever.  Our solution to this is to surround the
- * code with lguest_noirq_start: and lguest_noirq_end: labels.  We tell the
+ * return to userspace or wherever.  Our solution to this is to tell the
  * Host that it is *never* to interrupt us there, even if interrupts seem to be
- * enabled.
+ * enabled. (It's not necessary to protect pop instruction, since
+ * data gets updated only after it completes, so we only need to protect
+ * one instruction, iret).
  */
 ENTRY(lguest_iret)
-	pushl	%eax
-	movl	12(%esp), %eax
-lguest_noirq_start:
+	pushl	2*4(%esp)
 	/*
 	 * Note the %ss: segment prefix here.  Normal data accesses use the
 	 * "ds" segment, but that will have already been restored for whatever
 	 * we're returning to (such as userspace): we can't trust it.  The %ss:
 	 * prefix makes sure we use the stack segment, which is still valid.
 	 */
-	movl	%eax,%ss:lguest_data+LGUEST_DATA_irq_enabled
-	popl	%eax
+	popl	%ss:lguest_data+LGUEST_DATA_irq_enabled
+lguest_noirq_iret:
 	iret
-lguest_noirq_end:

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);
 
 	/*

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

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;

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,

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) {

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

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

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. */

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);

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);
 

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);

include/linux/lguest.h

@@ -61,8 +61,8 @@ struct lguest_data {
 	u32 tsc_khz;
 
 /* Fields initialized by the Guest at boot: */
-	/* Instruction range to suppress interrupts even if enabled */
-	unsigned long noirq_start, noirq_end;
+	/* Instruction to suppress interrupts even if enabled */
+	unsigned long noirq_iret;
 	/* Address above which page tables are all identical. */
 	unsigned long kernel_address;
 	/* The vector to try to use for system calls (0x40 or 0x80). */

include/linux/virtio.h

@@ -108,8 +108,6 @@ struct virtio_device {
 	void *priv;
 };
 
-bool virtio_device_is_legacy_only(struct virtio_device_id id);
-
 static inline struct virtio_device *dev_to_virtio(struct device *_dev)
 {
 	return container_of(_dev, struct virtio_device, dev);

include/linux/virtio_config.h

@@ -298,13 +298,6 @@ static inline __virtio64 cpu_to_virtio64(struct virtio_device *vdev, u64 val)
 		}							\
 	} while(0)
 
-static inline u8 virtio_cread8(struct virtio_device *vdev, unsigned int offset)
-{
-	u8 ret;
-	vdev->config->get(vdev, offset, &ret, sizeof(ret));
-	return ret;
-}
-
 /* Read @count fields, @bytes each. */
 static inline void __virtio_cread_many(struct virtio_device *vdev,
 				       unsigned int offset,
@@ -326,7 +319,6 @@ static inline void __virtio_cread_many(struct virtio_device *vdev,
 	} while (gen != old);
 }
 
-
 static inline void virtio_cread_bytes(struct virtio_device *vdev,
 				      unsigned int offset,
 				      void *buf, size_t len)
@@ -334,6 +326,13 @@ static inline void virtio_cread_bytes(struct virtio_device *vdev,
 	__virtio_cread_many(vdev, offset, buf, len, 1);
 }
 
+static inline u8 virtio_cread8(struct virtio_device *vdev, unsigned int offset)
+{
+	u8 ret;
+	vdev->config->get(vdev, offset, &ret, sizeof(ret));
+	return ret;
+}
+
 static inline void virtio_cwrite8(struct virtio_device *vdev,
 				  unsigned int offset, u8 val)
 {
@@ -374,7 +373,6 @@ static inline u64 virtio_cread64(struct virtio_device *vdev,
 				 unsigned int offset)
 {
 	u64 ret;
-	vdev->config->get(vdev, offset, &ret, sizeof(ret));
 	__virtio_cread_many(vdev, offset, &ret, 1, sizeof(ret));
 	return virtio64_to_cpu(vdev, (__force __virtio64)ret);
 }

include/linux/virtio_ring.h

@@ -21,19 +21,20 @@
  * actually quite cheap.
  */
 
-#ifdef CONFIG_SMP
 static inline void virtio_mb(bool weak_barriers)
 {
+#ifdef CONFIG_SMP
 	if (weak_barriers)
 		smp_mb();
 	else
+#endif
 		mb();
 }
 
 static inline void virtio_rmb(bool weak_barriers)
 {
 	if (weak_barriers)
-		smp_rmb();
+		dma_rmb();
 	else
 		rmb();
 }
@@ -41,26 +42,10 @@ static inline void virtio_rmb(bool weak_barriers)
 static inline void virtio_wmb(bool weak_barriers)
 {
 	if (weak_barriers)
-		smp_wmb();
+		dma_wmb();
 	else
 		wmb();
 }
-#else
-static inline void virtio_mb(bool weak_barriers)
-{
-	mb();
-}
-
-static inline void virtio_rmb(bool weak_barriers)
-{
-	rmb();
-}
-
-static inline void virtio_wmb(bool weak_barriers)
-{
-	wmb();
-}
-#endif
 
 struct virtio_device;
 struct virtqueue;

include/uapi/linux/Kbuild

@@ -431,6 +431,7 @@ header-y += virtio_blk.h
 header-y += virtio_config.h
 header-y += virtio_console.h
 header-y += virtio_ids.h
+header-y += virtio_input.h
 header-y += virtio_net.h
 header-y += virtio_pci.h
 header-y += virtio_ring.h

include/uapi/linux/virtio_balloon.h

@@ -25,6 +25,7 @@
  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  * SUCH DAMAGE. */
+#include <linux/types.h>
 #include <linux/virtio_ids.h>
 #include <linux/virtio_config.h>
 
@@ -38,9 +39,9 @@
 
 struct virtio_balloon_config {
 	/* Number of pages host wants Guest to give up. */
-	__le32 num_pages;
+	__u32 num_pages;
 	/* Number of pages we've actually got in balloon. */
-	__le32 actual;
+	__u32 actual;
 };
 
 #define VIRTIO_BALLOON_S_SWAP_IN  0   /* Amount of memory swapped in */
@@ -51,9 +52,32 @@ struct virtio_balloon_config {
 #define VIRTIO_BALLOON_S_MEMTOT   5   /* Total amount of memory */
 #define VIRTIO_BALLOON_S_NR       6
 
+/*
+ * Memory statistics structure.
+ * Driver fills an array of these structures and passes to device.
+ *
+ * NOTE: fields are laid out in a way that would make compiler add padding
+ * between and after fields, so we have to use compiler-specific attributes to
+ * pack it, to disable this padding. This also often causes compiler to
+ * generate suboptimal code.
+ *
+ * We maintain this statistics structure format for backwards compatibility,
+ * but don't follow this example.
+ *
+ * If implementing a similar structure, do something like the below instead:
+ *     struct virtio_balloon_stat {
+ *         __virtio16 tag;
+ *         __u8 reserved[6];
+ *         __virtio64 val;
+ *     };
+ *
+ * In other words, add explicit reserved fields to align field and
+ * structure boundaries at field size, avoiding compiler padding
+ * without the packed attribute.
+ */
 struct virtio_balloon_stat {
-	__u16 tag;
-	__u64 val;
+	__virtio16 tag;
+	__virtio64 val;
 } __attribute__((packed));
 
 #endif /* _LINUX_VIRTIO_BALLOON_H */

include/uapi/linux/virtio_ids.h

@@ -39,5 +39,6 @@
 #define VIRTIO_ID_9P		9 /* 9p virtio console */
 #define VIRTIO_ID_RPROC_SERIAL 11 /* virtio remoteproc serial link */
 #define VIRTIO_ID_CAIF	       12 /* Virtio caif */
+#define VIRTIO_ID_INPUT        18 /* virtio input */
 
 #endif /* _LINUX_VIRTIO_IDS_H */

include/uapi/linux/virtio_input.h

+#ifndef _LINUX_VIRTIO_INPUT_H
+#define _LINUX_VIRTIO_INPUT_H
+/* This header is BSD licensed so anyone can use the definitions to implement
+ * compatible drivers/servers.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of IBM nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL IBM OR
+ * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF
+ * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
+ * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+ * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
+ * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE. */
+
+#include <linux/types.h>
+
+enum virtio_input_config_select {
+	VIRTIO_INPUT_CFG_UNSET      = 0x00,
+	VIRTIO_INPUT_CFG_ID_NAME    = 0x01,
+	VIRTIO_INPUT_CFG_ID_SERIAL  = 0x02,
+	VIRTIO_INPUT_CFG_ID_DEVIDS  = 0x03,
+	VIRTIO_INPUT_CFG_PROP_BITS  = 0x10,
+	VIRTIO_INPUT_CFG_EV_BITS    = 0x11,
+	VIRTIO_INPUT_CFG_ABS_INFO   = 0x12,
+};
+
+struct virtio_input_absinfo {
+	__u32 min;
+	__u32 max;
+	__u32 fuzz;
+	__u32 flat;
+	__u32 res;
+};
+
+struct virtio_input_devids {
+	__u16 bustype;
+	__u16 vendor;
+	__u16 product;
+	__u16 version;
+};
+
+struct virtio_input_config {
+	__u8    select;
+	__u8    subsel;
+	__u8    size;
+	__u8    reserved[5];
+	union {
+		char string[128];
+		__u8 bitmap[128];
+		struct virtio_input_absinfo abs;
+		struct virtio_input_devids ids;
+	} u;
+};
+
+struct virtio_input_event {
+	__le16 type;
+	__le16 code;
+	__le32 value;
+};
+
+#endif /* _LINUX_VIRTIO_INPUT_H */