[PATCH 00/10] in-kernel APIC v3 (kernel side)

[PATCH 00/10] in-kernel APIC v3 (kernel side)

[Gregory Haskins]

I am pleased to announce v3, which builds upon v2 by adding:

1) Incorporated Dor's feedback
2) Fixed bugs preventing XP+ACPI from booting

At this point, I can boot

A) 32-bit XP w/ACPI enabled
B) 64-bit SLED (2.6.16 based)

both with "level-0" (userspace PIC emulation) and "level-1" (in-kernel PIC
emulation).  The performance with XP under level-1 seems to be a huge
improvement over in-trunk performance.

As before, the last patch (VMX-tpr-shadow) is only included as a preview and
RFC.  Don't apply it or your VMX system fail to launch a guest.

These changes require updates to the usermode code as well, so that drop will
follow shortly.

Regards,
-Greg

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

[PATCH 01/10] KVM: Adds support for in-kernel mmio handlers

[Gregory Haskins]

Signed-off-by: Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
---

 drivers/kvm/kvm.h      |   60 +++++++++++++++++++++++++++++++
 drivers/kvm/kvm_main.c |   94 ++++++++++++++++++++++++++++++++++++++++++------
 2 files changed, 142 insertions(+), 12 deletions(-)

diff --git a/drivers/kvm/kvm.h b/drivers/kvm/kvm.h
index 1bbafba..e32f63a 100644
--- a/drivers/kvm/kvm.h
+++ b/drivers/kvm/kvm.h
@@ -256,6 +256,65 @@ struct kvm_stat {
 	u32 light_exits;
 };
 
+struct kvm_io_device {
+	void (*read)(struct kvm_io_device *this,
+		     gpa_t addr,
+		     int len,
+		     void *val);
+	void (*write)(struct kvm_io_device *this,
+		      gpa_t addr,
+		      int len,
+		      const void *val);
+	int (*in_range)(struct kvm_io_device *this, gpa_t addr);
+	void (*destructor)(struct kvm_io_device *this);
+
+	void             *private;
+};
+
+static inline void kvm_iodevice_read(struct kvm_io_device *dev,
+				     gpa_t addr,
+				     int len,
+				     void *val)
+{
+	dev->read(dev, addr, len, val);
+}
+
+static inline void kvm_iodevice_write(struct kvm_io_device *dev,
+				      gpa_t addr,
+				      int len,
+				      const void *val)
+{
+	dev->write(dev, addr, len, val);
+}
+
+static inline int kvm_iodevice_inrange(struct kvm_io_device *dev, gpa_t addr)
+{
+	return dev->in_range(dev, addr);
+}
+
+static inline void kvm_iodevice_destructor(struct kvm_io_device *dev)
+{
+	dev->destructor(dev);
+}
+
+/*
+ * It would be nice to use something smarter than a linear search, TBD...
+ * Thankfully we dont expect many devices to register (famous last words :),
+ * so until then it will suffice.  At least its abstracted so we can change
+ * in one place.
+ */
+struct kvm_io_bus {
+	int                   dev_count;
+#define NR_IOBUS_DEVS 6
+	struct kvm_io_device *devs[NR_IOBUS_DEVS];
+};
+
+void kvm_io_bus_init(struct kvm_io_bus *bus);
+void kvm_io_bus_destroy(struct kvm_io_bus *bus);
+struct kvm_io_device *kvm_io_bus_find_dev(struct kvm_io_bus *bus, gpa_t addr);
+void kvm_io_bus_register_dev(struct kvm_io_bus *bus,
+			     struct kvm_io_device *dev);
+
 struct kvm_vcpu {
 	struct kvm *kvm;
 	union {
@@ -375,6 +434,7 @@ struct kvm {
 	unsigned long rmap_overflow;
 	struct list_head vm_list;
 	struct file *filp;
+	struct kvm_io_bus mmio_bus;
 };
 
 struct descriptor_table {
diff --git a/drivers/kvm/kvm_main.c b/drivers/kvm/kvm_main.c
index 6aa4b40..d230d57 100644
--- a/drivers/kvm/kvm_main.c
+++ b/drivers/kvm/kvm_main.c
@@ -317,6 +317,7 @@ static struct kvm *kvm_create_vm(void)
 
 	spin_lock_init(&kvm->lock);
 	INIT_LIST_HEAD(&kvm->active_mmu_pages);
+	kvm_io_bus_init(&kvm->mmio_bus);
 	for (i = 0; i < KVM_MAX_VCPUS; ++i) {
 		struct kvm_vcpu *vcpu = &kvm->vcpus[i];
 
@@ -414,6 +415,7 @@ static void kvm_destroy_vm(struct kvm *kvm)
 	spin_lock(&kvm_lock);
 	list_del(&kvm->vm_list);
 	spin_unlock(&kvm_lock);
+	kvm_io_bus_destroy(&kvm->mmio_bus);
 	kvm_free_vcpus(kvm);
 	kvm_free_physmem(kvm);
 	kfree(kvm);
@@ -1037,12 +1039,25 @@ static int emulator_write_std(unsigned long addr,
 	return X86EMUL_UNHANDLEABLE;
 }
 
+static struct kvm_io_device *vcpu_find_mmio_dev(struct kvm_vcpu *vcpu,
+						gpa_t addr)
+{
+	/*
+	 * Note that its important to have this wrapper function because
+	 * in the very near future we will be checking for MMIOs against
+	 * the LAPIC as well as the general MMIO bus
+	 */
+	return kvm_io_bus_find_dev(&vcpu->kvm->mmio_bus, addr);
+}
+
 static int emulator_read_emulated(unsigned long addr,
 				  void *val,
 				  unsigned int bytes,
 				  struct x86_emulate_ctxt *ctxt)
 {
-	struct kvm_vcpu *vcpu = ctxt->vcpu;
+	struct kvm_vcpu      *vcpu = ctxt->vcpu;
+	struct kvm_io_device *mmio_dev;
+	gpa_t                 gpa;
 
 	if (vcpu->mmio_read_completed) {
 		memcpy(val, vcpu->mmio_data, bytes);
@@ -1051,18 +1066,26 @@ static int emulator_read_emulated(unsigned long addr,
 	} else if (emulator_read_std(addr, val, bytes, ctxt)
 		   == X86EMUL_CONTINUE)
 		return X86EMUL_CONTINUE;
-	else {
-		gpa_t gpa = vcpu->mmu.gva_to_gpa(vcpu, addr);
 
-		if (gpa == UNMAPPED_GVA)
-			return X86EMUL_PROPAGATE_FAULT;
-		vcpu->mmio_needed = 1;
-		vcpu->mmio_phys_addr = gpa;
-		vcpu->mmio_size = bytes;
-		vcpu->mmio_is_write = 0;
+	gpa = vcpu->mmu.gva_to_gpa(vcpu, addr);
+	if (gpa == UNMAPPED_GVA)
+		return X86EMUL_PROPAGATE_FAULT;
 
-		return X86EMUL_UNHANDLEABLE;
+	/*
+	 * Is this MMIO handled locally?
+	 */
+	mmio_dev = vcpu_find_mmio_dev(vcpu, gpa);
+	if (mmio_dev) {
+		kvm_iodevice_read(mmio_dev, gpa, bytes, val);
+		return X86EMUL_CONTINUE;
 	}
+
+	vcpu->mmio_needed = 1;
+	vcpu->mmio_phys_addr = gpa;
+	vcpu->mmio_size = bytes;
+	vcpu->mmio_is_write = 0;
+
+	return X86EMUL_UNHANDLEABLE;
 }
 
 static int emulator_write_phys(struct kvm_vcpu *vcpu, gpa_t gpa,
@@ -1090,8 +1113,9 @@ static int emulator_write_emulated(unsigned long addr,
 				   unsigned int bytes,
 				   struct x86_emulate_ctxt *ctxt)
 {
-	struct kvm_vcpu *vcpu = ctxt->vcpu;
-	gpa_t gpa = vcpu->mmu.gva_to_gpa(vcpu, addr);
+	struct kvm_vcpu      *vcpu = ctxt->vcpu;
+	struct kvm_io_device *mmio_dev;
+	gpa_t                 gpa = vcpu->mmu.gva_to_gpa(vcpu, addr);
 
 	if (gpa == UNMAPPED_GVA) {
 		kvm_arch_ops->inject_page_fault(vcpu, addr, 2);
@@ -1101,6 +1125,15 @@ static int emulator_write_emulated(unsigned long addr,
 	if (emulator_write_phys(vcpu, gpa, val, bytes))
 		return X86EMUL_CONTINUE;
 
+	/*
+	 * Is this MMIO handled locally?
+	 */
+	mmio_dev = vcpu_find_mmio_dev(vcpu, gpa);
+	if (mmio_dev) {
+		kvm_iodevice_write(mmio_dev, gpa, bytes, val);
+		return X86EMUL_CONTINUE;
+	}
+
 	vcpu->mmio_needed = 1;
 	vcpu->mmio_phys_addr = gpa;
 	vcpu->mmio_size = bytes;
@@ -2930,6 +2963,43 @@ static int kvm_cpu_hotplug(struct notifier_block *notifier, unsigned long val,
 	return NOTIFY_OK;
 }
 
+void kvm_io_bus_init(struct kvm_io_bus *bus)
+{
+	memset(bus, 0, sizeof(*bus));
+}
+
+void kvm_io_bus_destroy(struct kvm_io_bus *bus)
+{
+	int i;
+
+	for (i = 0; i < bus->dev_count; i++) {
+		struct kvm_io_device *pos = bus->devs[i];
+
+		kvm_iodevice_destructor(pos);
+	}
+}
+
+struct kvm_io_device *kvm_io_bus_find_dev(struct kvm_io_bus *bus, gpa_t addr)
+{
+	int i;
+
+	for (i = 0; i < bus->dev_count; i++) {
+		struct kvm_io_device *pos = bus->devs[i];
+
+		if (pos->in_range(pos, addr))
+			return pos;
+	}
+
+	return NULL;
+}
+
+void kvm_io_bus_register_dev(struct kvm_io_bus *bus, struct kvm_io_device *dev)
+{
+	BUG_ON(bus->dev_count > (NR_IOBUS_DEVS-1));
+
+	bus->devs[bus->dev_count++] = dev;
+}
+
 static struct notifier_block kvm_cpu_notifier = {
 	.notifier_call = kvm_cpu_hotplug,
 	.priority = 20, /* must be > scheduler priority */


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

[PATCH 02/10] KVM: VMX - fix interrupt checking on light-exit

[Gregory Haskins]

Signed-off-by: Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
---

 drivers/kvm/vmx.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/kvm/vmx.c b/drivers/kvm/vmx.c
index 804a623..e696d02 100644
--- a/drivers/kvm/vmx.c
+++ b/drivers/kvm/vmx.c
@@ -1934,13 +1934,13 @@ static int vmx_vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	int r;
 
 preempted:
-	if (!vcpu->mmio_read_completed)
-		do_interrupt_requests(vcpu, kvm_run);
-
 	if (vcpu->guest_debug.enabled)
 		kvm_guest_debug_pre(vcpu);
 
 again:
+	if (!vcpu->mmio_read_completed)
+		do_interrupt_requests(vcpu, kvm_run);
+
 	vmx_save_host_state(vcpu);
 	kvm_load_guest_fpu(vcpu);
 


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

[PATCH 03/10] KVM: Add irqdevice object

[Gregory Haskins]

The current code is geared towards using a user-mode (A)PIC.  This patch adds
an "irqdevice" abstraction, and implements a "userint" model to handle the
duties of the original code.  Later, we can develop other irqdevice models
to handle objects like LAPIC, IOAPIC, i8259, etc, as appropriate

Signed-off-by: Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
---

 drivers/kvm/Makefile    |    2 
 drivers/kvm/irqdevice.h |  176 +++++++++++++++++++++++++++++++++++++
 drivers/kvm/kvm.h       |  107 ++++++++++++++++++++++-
 drivers/kvm/kvm_main.c  |   58 +++++++++---
 drivers/kvm/svm.c       |  158 ++++++++++++++++++++++++---------
 drivers/kvm/userint.c   |  223 +++++++++++++++++++++++++++++++++++++++++++++++
 drivers/kvm/vmx.c       |  157 +++++++++++++++++++++++++--------
 7 files changed, 778 insertions(+), 103 deletions(-)

diff --git a/drivers/kvm/Makefile b/drivers/kvm/Makefile
index c0a789f..540afbc 100644
--- a/drivers/kvm/Makefile
+++ b/drivers/kvm/Makefile
@@ -2,7 +2,7 @@
 # Makefile for Kernel-based Virtual Machine module
 #
 
-kvm-objs := kvm_main.o mmu.o x86_emulate.o
+kvm-objs := kvm_main.o mmu.o x86_emulate.o userint.o
 obj-$(CONFIG_KVM) += kvm.o
 kvm-intel-objs = vmx.o
 obj-$(CONFIG_KVM_INTEL) += kvm-intel.o
diff --git a/drivers/kvm/irqdevice.h b/drivers/kvm/irqdevice.h
new file mode 100644
index 0000000..097d179
--- /dev/null
+++ b/drivers/kvm/irqdevice.h
@@ -0,0 +1,176 @@
+/*
+ * Defines an interface for an abstract interrupt controller.  The model
+ * consists of a unit with an arbitrary number of input lines N (IRQ0-(N-1)),
+ * an arbitrary number of output lines (INTR) (LINT, EXTINT, NMI, etc), and
+ * methods for completing an interrupt-acknowledge cycle (INTA).  A particular
+ * implementation of this model will define various policies, such as
+ * irq-to-vector translation, INTA/auto-EOI policy, etc.
+ *
+ * In addition, the INTR callback mechanism allows the unit to be "wired" to
+ * an interruptible source in a very flexible manner. For instance, an
+ * irqdevice could have its INTR wired to a VCPU (ala LAPIC), or another
+ * interrupt controller (ala cascaded i8259s)
+ *
+ * Copyright (C) 2007 Novell
+ *
+ * Authors:
+ *   Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.  See
+ * the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef __IRQDEVICE_H
+#define __IRQDEVICE_H
+
+struct kvm_irqdevice;
+
+typedef enum {
+	kvm_irqpin_localint,
+	kvm_irqpin_extint,
+	kvm_irqpin_smi,
+	kvm_irqpin_nmi,
+	kvm_irqpin_invalid, /* must always be last */
+} kvm_irqpin_t;
+
+
+struct kvm_irqsink {
+	void (*set_intr)(struct kvm_irqsink *this,
+			 struct kvm_irqdevice *dev,
+			 kvm_irqpin_t pin);
+
+	void *private;
+};
+
+#define KVM_IRQACKDATA_VECTOR_VALID   (1 << 0)
+#define KVM_IRQACKDATA_VECTOR_PENDING (1 << 1)
+
+#define KVM_IRQACK_FLAG_PEEK          (1 << 0)
+
+struct kvm_irqack_data {
+	int flags;
+	int vector;
+};
+
+struct kvm_irqdevice {
+	int  (*ack)(struct kvm_irqdevice *this, int flags,
+		    struct kvm_irqack_data *data);
+	int  (*set_pin)(struct kvm_irqdevice *this, int pin, int level);
+	void (*destructor)(struct kvm_irqdevice *this);
+
+	void               *private;
+	struct kvm_irqsink  sink;
+};
+
+/**
+ * kvm_irqdevice_init - initialize the kvm_irqdevice for use
+ * @dev: The device
+ *
+ * Description: Initialize the kvm_irqdevice for use.  Should be called before
+ *              calling any derived implementation init functions
+ *
+ * Returns: (void)
+ */
+static inline void kvm_irqdevice_init(struct kvm_irqdevice *dev)
+{
+	memset(dev, 0, sizeof(*dev));
+}
+
+/**
+ * kvm_irqdevice_ack - read and ack the highest priority vector from the device
+ * @dev: The device
+ * @flags: Modifies default behavior
+ *           [ KVM_IRQACK_FLAG_PEEK - Dont ack vector, just check status ]
+ * @data: A pointer to a kvm_irqack_data structure to hold the result
+ *
+ * Description: Read the highest priority pending vector from the device,
+ *              potentially invoking auto-EOI depending on device policy
+ *
+ *              Successful return indicates that the *data* structure is valid
+ *
+ *               data.flags -
+ *                  [KVM_IRQACKDATA_VECTOR_VALID - data.vector is valid]
+ *                  [KVM_IRQACKDATA_VECTOR_PENDING - more vectors are pending]
+ *
+ * Returns: (int)
+ *   [-1 = failure]
+ *   [ 0 = success]
+ */
+static inline int kvm_irqdevice_ack(struct kvm_irqdevice *dev, int flags,
+				    struct kvm_irqack_data *data)
+{
+	return dev->ack(dev, flags, data);
+}
+
+/**
+ * kvm_irqdevice_set_pin - allows the caller to assert/deassert an IRQ
+ * @dev: The device
+ * @pin: The input pin to alter
+ * @level: The value to set (1 = assert, 0 = deassert)
+ *
+ * Description: Allows the caller to assert/deassert an IRQ input pin to the
+ *              device according to device policy.
+ *
+ * Returns: (int)
+ *   [-1 = failure]
+ *   [ 0 = success]
+ */
+static inline int kvm_irqdevice_set_pin(struct kvm_irqdevice *dev, int pin,
+				  int level)
+{
+	return dev->set_pin(dev, pin, level);
+}
+
+/**
+ * kvm_irqdevice_register_sink - registers an kvm_irqsink object
+ * @dev: The device
+ * @sink: The sink to register.  Data will be copied so building object from
+ *        transient storage is ok.
+ *
+ * Description: Registers an kvm_irqsink object as an INTR callback
+ *
+ * Returns: (void)
+ */
+static inline void kvm_irqdevice_register_sink(struct kvm_irqdevice *dev,
+					       const struct kvm_irqsink *sink)
+{
+	dev->sink = *sink;
+}
+
+/**
+ * kvm_irqdevice_destructor - destroys an irqdevice
+ * @dev: The device
+ *
+ * Returns: (void)
+ */
+static inline void kvm_irqdevice_destructor(struct kvm_irqdevice *dev)
+{
+	dev->destructor(dev);
+}
+
+/**
+ * kvm_irqdevice_set_intr - invokes a registered INTR callback
+ * @dev: The device
+ * @pin: Identifies the pin to alter -
+ *           [ KVM_IRQPIN_LOCALINT (default) - an vector is pending on this
+ *                                             device]
+ *           [ KVM_IRQPIN_EXTINT - a vector is pending on an external device]
+ *           [ KVM_IRQPIN_SMI - system-management-interrupt pin]
+ *           [ KVM_IRQPIN_NMI - non-maskable-interrupt pin
+ *
+ * Description: Invokes a registered INTR callback (if present).  This
+ *              function is meant to be used privately by a irqdevice
+ *              implementation.
+ *
+ * Returns: (void)
+ */
+static inline void kvm_irqdevice_set_intr(struct kvm_irqdevice *dev,
+					  kvm_irqpin_t pin)
+{
+	struct kvm_irqsink *sink = &dev->sink;
+	if (sink->set_intr)
+		sink->set_intr(sink, dev, pin);
+}
+
+#endif /*  __IRQDEVICE_H */
diff --git a/drivers/kvm/kvm.h b/drivers/kvm/kvm.h
index e32f63a..41e4eaa 100644
--- a/drivers/kvm/kvm.h
+++ b/drivers/kvm/kvm.h
@@ -13,6 +13,7 @@
 #include <linux/mm.h>
 
 #include "vmx.h"
+#include "irqdevice.h"
 #include <linux/kvm.h>
 #include <linux/kvm_para.h>
 
@@ -160,6 +161,11 @@ struct vmcs {
 
 struct kvm_vcpu;
 
+int kvm_user_irqdev_init(struct kvm_irqdevice *dev);
+int kvm_user_irqdev_save(struct kvm_irqdevice *this, void *data);
+int kvm_user_irqdev_restore(struct kvm_irqdevice *this, void *data);
+int kvm_userint_init(struct kvm_vcpu *vcpu);
+
 /*
  * x86 supports 3 paging modes (4-level 64-bit, 3-level 64-bit, and 2-level
  * 32-bit).  The kvm_mmu structure abstracts the details of the current mmu
@@ -315,6 +321,18 @@ struct kvm_io_device *kvm_io_bus_find_dev(struct kvm_io_bus *bus, gpa_t addr);
 void kvm_io_bus_register_dev(struct kvm_io_bus *bus,
 			     struct kvm_io_device *dev);
 
+#define NR_IRQ_WORDS KVM_IRQ_BITMAP_SIZE(unsigned long)
+
+/*
+ * structure for maintaining info for interrupting an executing VCPU
+ */
+struct kvm_vcpu_irq {
+	spinlock_t           lock;
+	struct kvm_irqdevice dev;
+	int                  pending;
+	int                  deferred;
+};
+
 struct kvm_vcpu {
 	struct kvm *kvm;
 	union {
@@ -327,9 +345,7 @@ struct kvm_vcpu {
 	u64 host_tsc;
 	struct kvm_run *run;
 	int interrupt_window_open;
-	unsigned long irq_summary; /* bit vector: 1 per word in irq_pending */
-#define NR_IRQ_WORDS KVM_IRQ_BITMAP_SIZE(unsigned long)
-	unsigned long irq_pending[NR_IRQ_WORDS];
+	struct kvm_vcpu_irq irq;
 	unsigned long regs[NR_VCPU_REGS]; /* for rsp: vcpu_load_rsp_rip() */
 	unsigned long rip;      /* needs vcpu_load_rsp_rip() */
 
@@ -402,6 +418,91 @@ struct kvm_vcpu {
 	struct kvm_cpuid_entry cpuid_entries[KVM_MAX_CPUID_ENTRIES];
 };
 
+/*
+ * Assumes lock already held
+ */
+static inline int __kvm_vcpu_irq_all_pending(struct kvm_vcpu *vcpu)
+{
+	int pending = vcpu->irq.pending;
+
+	if (vcpu->irq.deferred != -1)
+		__set_bit(kvm_irqpin_localint, &pending);
+
+	return pending;
+}
+
+/*
+ * These two functions are helpers for determining if a standard interrupt
+ * is pending to replace the old "if (vcpu->irq_summary)" logic.  If the
+ * caller wants to know about some of the new advanced interrupt types
+ * (SMI, NMI, etc) or to differentiate between localint and extint they will
+ *  have to use the new API
+ */
+static inline int __kvm_vcpu_irq_pending(struct kvm_vcpu *vcpu)
+{
+	int pending = __kvm_vcpu_irq_all_pending(vcpu);
+
+	if (test_bit(kvm_irqpin_localint, &pending) ||
+	    test_bit(kvm_irqpin_extint, &pending))
+		return 1;
+
+	return 0;
+}
+
+static inline int kvm_vcpu_irq_pending(struct kvm_vcpu *vcpu)
+{
+	int ret = 0;
+	int flags;
+
+	spin_lock_irqsave(&vcpu->irq.lock, flags);
+	ret = __kvm_vcpu_irq_pending(vcpu);
+	spin_unlock_irqrestore(&vcpu->irq.lock, flags);
+
+	return ret;
+}
+
+/*
+ * Assumes lock already held
+ */
+static inline int kvm_vcpu_irq_pop(struct kvm_vcpu *vcpu,
+				   struct kvm_irqack_data *data)
+{
+	int ret = 0;
+
+	if (vcpu->irq.deferred != -1) {
+		ret = kvm_irqdevice_ack(&vcpu->irq.dev, KVM_IRQACK_FLAG_PEEK,
+					data);
+		data->flags |= KVM_IRQACKDATA_VECTOR_VALID;
+		data->vector = vcpu->irq.deferred;
+		vcpu->irq.deferred = -1;
+	} else
+		ret = kvm_irqdevice_ack(&vcpu->irq.dev, 0, data);
+
+	/*
+	 * If there are no more interrupts we must clear the status flag
+	 */
+	if (!(data->flags & KVM_IRQACKDATA_VECTOR_PENDING))
+		__clear_bit(kvm_irqpin_localint, &vcpu->irq.pending);
+
+	return ret;
+}
+
+static inline void __kvm_vcpu_irq_push(struct kvm_vcpu *vcpu, int irq)
+{
+	BUG_ON(vcpu->irq.deferred != -1); /* We can only hold one deferred */
+
+	vcpu->irq.deferred = irq;
+}
+
+static inline void kvm_vcpu_irq_push(struct kvm_vcpu *vcpu, int irq)
+{
+	int flags;
+
+	spin_lock_irqsave(&vcpu->irq.lock, flags);
+	__kvm_vcpu_irq_push(vcpu, irq);
+	spin_unlock_irqrestore(&vcpu->irq.lock, flags);
+}
+
 struct kvm_mem_alias {
 	gfn_t base_gfn;
 	unsigned long npages;
diff --git a/drivers/kvm/kvm_main.c b/drivers/kvm/kvm_main.c
index d230d57..3f47b20 100644
--- a/drivers/kvm/kvm_main.c
+++ b/drivers/kvm/kvm_main.c
@@ -322,6 +322,11 @@ static struct kvm *kvm_create_vm(void)
 		struct kvm_vcpu *vcpu = &kvm->vcpus[i];
 
 		mutex_init(&vcpu->mutex);
+
+		memset(&vcpu->irq, 0, sizeof(vcpu->irq));
+		spin_lock_init(&vcpu->irq.lock);
+		vcpu->irq.deferred = -1;
+
 		vcpu->cpu = -1;
 		vcpu->kvm = kvm;
 		vcpu->mmu.root_hpa = INVALID_PAGE;
@@ -389,6 +394,7 @@ static void kvm_free_vcpu(struct kvm_vcpu *vcpu)
 	vcpu_load(vcpu);
 	kvm_mmu_destroy(vcpu);
 	vcpu_put(vcpu);
+	kvm_irqdevice_destructor(&vcpu->irq.dev);
 	kvm_arch_ops->vcpu_free(vcpu);
 	free_page((unsigned long)vcpu->run);
 	vcpu->run = NULL;
@@ -2008,8 +2014,7 @@ static int kvm_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 	sregs->efer = vcpu->shadow_efer;
 	sregs->apic_base = vcpu->apic_base;
 
-	memcpy(sregs->interrupt_bitmap, vcpu->irq_pending,
-	       sizeof sregs->interrupt_bitmap);
+	kvm_user_irqdev_save(&vcpu->irq.dev, &sregs->interrupt_bitmap);
 
 	vcpu_put(vcpu);
 
@@ -2026,7 +2031,6 @@ static int kvm_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 				    struct kvm_sregs *sregs)
 {
 	int mmu_reset_needed = 0;
-	int i;
 	struct descriptor_table dt;
 
 	vcpu_load(vcpu);
@@ -2063,12 +2067,8 @@ static int kvm_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 	if (mmu_reset_needed)
 		kvm_mmu_reset_context(vcpu);
 
-	memcpy(vcpu->irq_pending, sregs->interrupt_bitmap,
-	       sizeof vcpu->irq_pending);
-	vcpu->irq_summary = 0;
-	for (i = 0; i < NR_IRQ_WORDS; ++i)
-		if (vcpu->irq_pending[i])
-			__set_bit(i, &vcpu->irq_summary);
+	kvm_user_irqdev_restore(&vcpu->irq.dev,
+				&sregs->interrupt_bitmap[0]);
 
 	set_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
 	set_segment(vcpu, &sregs->ds, VCPU_SREG_DS);
@@ -2229,14 +2229,8 @@ static int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu,
 {
 	if (irq->irq < 0 || irq->irq >= 256)
 		return -EINVAL;
-	vcpu_load(vcpu);
-
-	set_bit(irq->irq, vcpu->irq_pending);
-	set_bit(irq->irq / BITS_PER_LONG, &vcpu->irq_summary);
 
-	vcpu_put(vcpu);
-
-	return 0;
+	return kvm_irqdevice_set_pin(&vcpu->irq.dev, irq->irq, 1);
 }
 
 static int kvm_vcpu_ioctl_debug_guest(struct kvm_vcpu *vcpu,
@@ -2338,6 +2332,32 @@ out1:
 }
 
 /*
+ * This function will be invoked whenever the vcpu->irq.dev raises its INTR
+ * line
+ */
+static void kvm_vcpu_intr(struct kvm_irqsink *this,
+			  struct kvm_irqdevice *dev,
+			  kvm_irqpin_t pin)
+{
+	struct kvm_vcpu *vcpu = (struct kvm_vcpu*)this->private;
+	unsigned long flags;
+
+	spin_lock_irqsave(&vcpu->irq.lock, flags);
+	__set_bit(pin, &vcpu->irq.pending);
+	spin_unlock_irqrestore(&vcpu->irq.lock, flags);
+}
+
+static void kvm_vcpu_irqsink_init(struct kvm_vcpu *vcpu)
+{
+	struct kvm_irqsink sink = {
+		.set_intr   = kvm_vcpu_intr,
+		.private    = vcpu
+	};
+
+	kvm_irqdevice_register_sink(&vcpu->irq.dev, &sink);
+}
+
+/*
  * Creates some virtual cpus.  Good luck creating more than one.
  */
 static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, int n)
@@ -2384,6 +2404,12 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, int n)
 	if (r < 0)
 		goto out_free_vcpus;
 
+	kvm_irqdevice_init(&vcpu->irq.dev);
+	kvm_vcpu_irqsink_init(vcpu);
+	r = kvm_userint_init(vcpu);
+	if (r < 0)
+		goto out_free_vcpus;
+
 	kvm_arch_ops->vcpu_load(vcpu);
 	r = kvm_mmu_setup(vcpu);
 	if (r >= 0)
diff --git a/drivers/kvm/svm.c b/drivers/kvm/svm.c
index b621403..4c03881 100644
--- a/drivers/kvm/svm.c
+++ b/drivers/kvm/svm.c
@@ -106,24 +106,6 @@ static unsigned get_addr_size(struct kvm_vcpu *vcpu)
 				(cs_attrib & SVM_SELECTOR_DB_MASK) ? 4 : 2;
 }
 
-static inline u8 pop_irq(struct kvm_vcpu *vcpu)
-{
-	int word_index = __ffs(vcpu->irq_summary);
-	int bit_index = __ffs(vcpu->irq_pending[word_index]);
-	int irq = word_index * BITS_PER_LONG + bit_index;
-
-	clear_bit(bit_index, &vcpu->irq_pending[word_index]);
-	if (!vcpu->irq_pending[word_index])
-		clear_bit(word_index, &vcpu->irq_summary);
-	return irq;
-}
-
-static inline void push_irq(struct kvm_vcpu *vcpu, u8 irq)
-{
-	set_bit(irq, vcpu->irq_pending);
-	set_bit(irq / BITS_PER_LONG, &vcpu->irq_summary);
-}
-
 static inline void clgi(void)
 {
 	asm volatile (SVM_CLGI);
@@ -904,7 +886,12 @@ static int pf_interception(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	int r;
 
 	if (is_external_interrupt(exit_int_info))
-		push_irq(vcpu, exit_int_info & SVM_EVTINJ_VEC_MASK);
+		/*
+		 * An exception was taken while we were trying to inject an
+		 * IRQ.  We must defer the injection of the vector until
+		 * the next window.
+		 */
+		kvm_vcpu_irq_push(vcpu, exit_int_info & SVM_EVTINJ_VEC_MASK);
 
 	spin_lock(&vcpu->kvm->lock);
 
@@ -1114,7 +1101,7 @@ static int halt_interception(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 {
 	vcpu->svm->next_rip = vcpu->svm->vmcb->save.rip + 1;
 	skip_emulated_instruction(vcpu);
-	if (vcpu->irq_summary)
+	if (kvm_vcpu_irq_pending(vcpu))
 		return 1;
 
 	kvm_run->exit_reason = KVM_EXIT_HLT;
@@ -1285,7 +1272,7 @@ static int interrupt_window_interception(struct kvm_vcpu *vcpu,
 	 * possible
 	 */
 	if (kvm_run->request_interrupt_window &&
-	    !vcpu->irq_summary) {
+	    !kvm_vcpu_irq_pending(vcpu)) {
 		++vcpu->stat.irq_window_exits;
 		kvm_run->exit_reason = KVM_EXIT_IRQ_WINDOW_OPEN;
 		return 0;
@@ -1384,60 +1371,135 @@ static void pre_svm_run(struct kvm_vcpu *vcpu)
 }
 
 
-static inline void kvm_do_inject_irq(struct kvm_vcpu *vcpu)
-{
-	struct vmcb_control_area *control;
-
-	control = &vcpu->svm->vmcb->control;
-	control->int_vector = pop_irq(vcpu);
-	control->int_ctl &= ~V_INTR_PRIO_MASK;
-	control->int_ctl |= V_IRQ_MASK |
-		((/*control->int_vector >> 4*/ 0xf) << V_INTR_PRIO_SHIFT);
-}
-
 static void kvm_reput_irq(struct kvm_vcpu *vcpu)
 {
 	struct vmcb_control_area *control = &vcpu->svm->vmcb->control;
 
 	if (control->int_ctl & V_IRQ_MASK) {
 		control->int_ctl &= ~V_IRQ_MASK;
-		push_irq(vcpu, control->int_vector);
+		kvm_vcpu_irq_push(vcpu, control->int_vector);
 	}
 
 	vcpu->interrupt_window_open =
 		!(control->int_state & SVM_INTERRUPT_SHADOW_MASK);
 }
 
-static void do_interrupt_requests(struct kvm_vcpu *vcpu,
-				       struct kvm_run *kvm_run)
+static int do_intr_requests(struct kvm_vcpu *vcpu,
+			    struct kvm_run *kvm_run,
+			    kvm_irqpin_t pin)
 {
 	struct vmcb_control_area *control = &vcpu->svm->vmcb->control;
+	int handled = 0;
 
 	vcpu->interrupt_window_open =
 		(!(control->int_state & SVM_INTERRUPT_SHADOW_MASK) &&
 		 (vcpu->svm->vmcb->save.rflags & X86_EFLAGS_IF));
 
-	if (vcpu->interrupt_window_open && vcpu->irq_summary)
+	if (vcpu->interrupt_window_open) {
 		/*
-		 * If interrupts enabled, and not blocked by sti or mov ss. Good.
+		 * If interrupts enabled, and not blocked by sti or mov ss.
+		 * Good.
 		 */
-		kvm_do_inject_irq(vcpu);
+		struct kvm_irqack_data ack;
+		int r = 0;
+
+		memset(&ack, 0, sizeof(ack));
+
+		switch (pin) {
+		case kvm_irqpin_localint:
+			r = kvm_vcpu_irq_pop(vcpu, &ack);
+			break;
+		case kvm_irqpin_extint:
+			printk(KERN_WARNING "KVM: external-interrupts not " \
+			       "handled yet\n");
+			__clear_bit(pin, &vcpu->irq.pending);
+			break;
+		case kvm_irqpin_nmi:
+			/*
+			 * FIXME: Someday we will handle this using the
+			 * specific SVN NMI features.  For now, just inject
+			 * the NMI as a standard interrupt on vector 2
+			 */
+			ack.flags |= KVM_IRQACKDATA_VECTOR_VALID;
+			ack.vector = 2;
+			__clear_bit(pin, &vcpu->irq.pending);
+			break;
+		default:
+			panic("KVM: unknown interrupt pin raised: %d\n", pin);
+			break;
+		}
+
+		BUG_ON(r < 0);
+
+		if (ack.flags & KVM_IRQACKDATA_VECTOR_VALID) {
+			control = &vcpu->svm->vmcb->control;
+			control->int_vector = ack.vector;
+			control->int_ctl &= ~V_INTR_PRIO_MASK;
+			control->int_ctl |= V_IRQ_MASK |
+				((/*control->int_vector >> 4*/ 0xf) <<
+				 V_INTR_PRIO_SHIFT);
+
+			handled = 1;
+		}
+	}
 
 	/*
 	 * Interrupts blocked.  Wait for unblock.
 	 */
 	if (!vcpu->interrupt_window_open &&
-	    (vcpu->irq_summary || kvm_run->request_interrupt_window)) {
+	    (__kvm_vcpu_irq_pending(vcpu) ||
+	     kvm_run->request_interrupt_window))
 		control->intercept |= 1ULL << INTERCEPT_VINTR;
-	} else
-		control->intercept &= ~(1ULL << INTERCEPT_VINTR);
+
+	return handled;
+}
+
+static void clear_pending_controls(struct kvm_vcpu *vcpu)
+{
+	struct vmcb_control_area *control = &vcpu->svm->vmcb->control;
+
+	control->intercept &= ~(1ULL << INTERCEPT_VINTR);
+}
+
+static void do_interrupt_requests(struct kvm_vcpu *vcpu,
+				  struct kvm_run *kvm_run)
+{
+	int pending = __kvm_vcpu_irq_all_pending(vcpu);
+
+	clear_pending_controls(vcpu);
+
+	while (pending) {
+		kvm_irqpin_t pin = __fls(pending);
+
+		switch (pin) {
+		case kvm_irqpin_localint:
+		case kvm_irqpin_extint:
+		case kvm_irqpin_nmi:
+			do_intr_requests(vcpu, kvm_run, pin);
+			break;
+		case kvm_irqpin_smi:
+			/* ignored (for now) */
+			printk(KERN_WARNING "KVM: dropping unhandled SMI\n");
+			__clear_bit(pin, &vcpu->irq.pending);
+			break;
+		case kvm_irqpin_invalid:
+			/* drop */
+			break;
+		default:
+			panic("KVM: unknown interrupt pin raised: %d\n", pin);
+			break;
+		}
+
+		__clear_bit(pin, &pending);
+	}
 }
 
 static void post_kvm_run_save(struct kvm_vcpu *vcpu,
 			      struct kvm_run *kvm_run)
 {
-	kvm_run->ready_for_interrupt_injection = (vcpu->interrupt_window_open &&
-						  vcpu->irq_summary == 0);
+	kvm_run->ready_for_interrupt_injection =
+		(vcpu->interrupt_window_open &&
+		 !kvm_vcpu_irq_pending(vcpu));
 	kvm_run->if_flag = (vcpu->svm->vmcb->save.rflags & X86_EFLAGS_IF) != 0;
 	kvm_run->cr8 = vcpu->cr8;
 	kvm_run->apic_base = vcpu->apic_base;
@@ -1452,7 +1514,7 @@ static void post_kvm_run_save(struct kvm_vcpu *vcpu,
 static int dm_request_for_irq_injection(struct kvm_vcpu *vcpu,
 					  struct kvm_run *kvm_run)
 {
-	return (!vcpu->irq_summary &&
+	return (!kvm_vcpu_irq_pending(vcpu) &&
 		kvm_run->request_interrupt_window &&
 		vcpu->interrupt_window_open &&
 		(vcpu->svm->vmcb->save.rflags & X86_EFLAGS_IF));
@@ -1482,9 +1544,17 @@ static int svm_vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	int r;
 
 again:
+	spin_lock(&vcpu->irq.lock);
+
+	/*
+	 * We must inject interrupts (if any) while the irq_lock
+	 * is held
+	 */
 	if (!vcpu->mmio_read_completed)
 		do_interrupt_requests(vcpu, kvm_run);
 
+	spin_unlock(&vcpu->irq.lock);
+
 	clgi();
 
 	pre_svm_run(vcpu);
diff --git a/drivers/kvm/userint.c b/drivers/kvm/userint.c
new file mode 100644
index 0000000..08d26fa
--- /dev/null
+++ b/drivers/kvm/userint.c
@@ -0,0 +1,223 @@
+/*
+ * User Interrupts IRQ device
+ *
+ * This acts as an extention of an interrupt controller that exists elsewhere
+ * (typically in userspace/QEMU).  Because this PIC is a pseudo device that
+ * is downstream from a real emulated PIC, the "IRQ-to-vector" mapping has
+ * already occured.  Therefore, this PIC has the following unusal properties:
+ *
+ * 1) It has 256 "pins" which are literal vectors (i.e. no translation)
+ * 2) It only supports "auto-EOI" behavior since it is expected that the
+ *    upstream emulated PIC will handle the real EOIs (if applicable)
+ * 3) It only listens to "asserts" on the pins (deasserts are dropped)
+ *    because its an auto-EOI device anyway.
+ *
+ * Copyright (C) 2007 Novell
+ *
+ * bitarray code based on original vcpu->irq_pending code,
+ *     Copyright (C) 2007 Qumranet
+ *
+ * Authors:
+ *   Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.  See
+ * the COPYING file in the top-level directory.
+ *
+ */
+
+#include "kvm.h"
+
+/*
+ *----------------------------------------------------------------------
+ * optimized bitarray object - works like bitarrays in bitops, but uses
+ * a summary field to accelerate lookups.  Assumes external locking
+ *---------------------------------------------------------------------
+ */
+
+struct bitarray {
+	unsigned long summary; /* 1 per word in pending */
+	unsigned long pending[NR_IRQ_WORDS];
+};
+
+static inline int bitarray_pending(struct bitarray *this)
+{
+	return this->summary ? 1 : 0;
+}
+
+static inline int bitarray_findhighest(struct bitarray *this)
+{
+	if (!this->summary)
+		return -1;
+	else {
+		int word_index = __fls(this->summary);
+		int bit_index  = __fls(this->pending[word_index]);
+
+		return word_index * BITS_PER_LONG + bit_index;
+	}
+}
+
+static inline void bitarray_set(struct bitarray *this, int nr)
+{
+	__set_bit(nr, &this->pending);
+	__set_bit(nr / BITS_PER_LONG, &this->summary);
+}
+
+static inline void bitarray_clear(struct bitarray *this, int nr)
+{
+	int word = nr / BITS_PER_LONG;
+
+	__clear_bit(nr, &this->pending);
+	if (!this->pending[word])
+		__clear_bit(word, &this->summary);
+}
+
+static inline int bitarray_test(struct bitarray *this, int nr)
+{
+	return test_bit(nr, &this->pending);
+}
+
+static inline int bitarray_test_and_set(struct bitarray *this, int nr, int val)
+{
+	if (bitarray_test(this, nr) != val) {
+		if (val)
+			bitarray_set(this, nr);
+		else
+			bitarray_clear(this, nr);
+		return 1;
+	}
+
+	return 0;
+}
+
+/*
+ *----------------------------------------------------------------------
+ * userint interface - provides the actual kvm_irqdevice implementation
+ *---------------------------------------------------------------------
+ */
+
+struct kvm_user_irqdev {
+	spinlock_t      lock;
+	atomic_t        ref_count;
+	struct bitarray pending;
+};
+
+static int user_irqdev_ack(struct kvm_irqdevice *this, int flags,
+			   struct kvm_irqack_data *data)
+{
+	struct kvm_user_irqdev *s = (struct kvm_user_irqdev*)this->private;
+
+	spin_lock(&s->lock);
+
+	if (!(flags & KVM_IRQACK_FLAG_PEEK)) {
+		int irq = bitarray_findhighest(&s->pending);
+
+		if (irq > -1) {
+			/*
+			 * Automatically clear the interrupt as the EOI
+			 * mechanism (if any) will take place in userspace
+			 */
+			bitarray_clear(&s->pending, irq);
+
+			data->flags |= KVM_IRQACKDATA_VECTOR_VALID;
+		}
+
+		data->vector = irq;
+	}
+
+	if (bitarray_pending(&s->pending))
+		data->flags |= KVM_IRQACKDATA_VECTOR_PENDING;
+
+	spin_unlock(&s->lock);
+
+	return 0;
+}
+
+static int user_irqdev_set_pin(struct kvm_irqdevice *this, int irq, int level)
+{
+	struct kvm_user_irqdev *s = (struct kvm_user_irqdev*)this->private;
+	int forward = 0;
+
+	spin_lock(&s->lock);
+	forward = bitarray_test_and_set(&s->pending, irq, level);
+	spin_unlock(&s->lock);
+
+	/*
+	 * alert the higher layer software we have changes
+	 */
+	if (forward)
+		kvm_irqdevice_set_intr(this, kvm_irqpin_localint);
+
+	return 0;
+}
+
+static void user_irqdev_destructor(struct kvm_irqdevice *this)
+{
+	struct kvm_user_irqdev *s = (struct kvm_user_irqdev*)this->private;
+
+	if (atomic_dec_and_test(&s->ref_count))
+		kfree(s);
+}
+
+int kvm_user_irqdev_init(struct kvm_irqdevice *irqdev)
+{
+	struct kvm_user_irqdev *s;
+
+	s = kzalloc(sizeof(*s), GFP_KERNEL);
+	if (!s)
+		return -ENOMEM;
+
+	spin_lock_init(&s->lock);
+
+	irqdev->ack         = user_irqdev_ack;
+	irqdev->set_pin     = user_irqdev_set_pin;
+	irqdev->destructor  = user_irqdev_destructor;
+
+	irqdev->private = s;
+	atomic_inc(&s->ref_count);
+
+	return 0;
+}
+
+int kvm_user_irqdev_save(struct kvm_irqdevice *this, void *data)
+{
+	struct kvm_user_irqdev *s = (struct kvm_user_irqdev*)this->private;
+
+	spin_lock(&s->lock);
+	memcpy(data, s->pending.pending, sizeof s->pending.pending);
+	spin_unlock(&s->lock);
+
+	return 0;
+}
+
+int kvm_user_irqdev_restore(struct kvm_irqdevice *this, void *data)
+{
+	struct kvm_user_irqdev *s = (struct kvm_user_irqdev*)this->private;
+	int i;
+	int forward = 0;
+
+	spin_lock(&s->lock);
+
+	/*
+	 * walk the interrupt-bitmap and inject an IRQ for each bit found
+	 */
+	for (i = 0; i < 256; ++i) {
+		int val = test_bit(i, data);
+		forward = bitarray_test_and_set(&s->pending, i, val);
+	}
+
+	spin_unlock(&s->lock);
+
+	/*
+	 * alert the higher layer software we have changes
+	 */
+	if (forward)
+		kvm_irqdevice_set_intr(this, kvm_irqpin_localint);
+
+	return 0;
+}
+
+int kvm_userint_init(struct kvm_vcpu *vcpu)
+{
+	return kvm_user_irqdev_init(&vcpu->irq.dev);
+}
+
diff --git a/drivers/kvm/vmx.c b/drivers/kvm/vmx.c
index e696d02..4d51194 100644
--- a/drivers/kvm/vmx.c
+++ b/drivers/kvm/vmx.c
@@ -1415,52 +1415,118 @@ static void inject_rmode_irq(struct kvm_vcpu *vcpu, int irq)
 	vmcs_writel(GUEST_RSP, (vmcs_readl(GUEST_RSP) & ~0xffff) | (sp - 6));
 }
 
-static void kvm_do_inject_irq(struct kvm_vcpu *vcpu)
+static int do_intr_requests(struct kvm_vcpu *vcpu,
+			    struct kvm_run *kvm_run,
+			    kvm_irqpin_t pin)
 {
-	int word_index = __ffs(vcpu->irq_summary);
-	int bit_index = __ffs(vcpu->irq_pending[word_index]);
-	int irq = word_index * BITS_PER_LONG + bit_index;
-
-	clear_bit(bit_index, &vcpu->irq_pending[word_index]);
-	if (!vcpu->irq_pending[word_index])
-		clear_bit(word_index, &vcpu->irq_summary);
-
-	if (vcpu->rmode.active) {
-		inject_rmode_irq(vcpu, irq);
-		return;
-	}
-	vmcs_write32(VM_ENTRY_INTR_INFO_FIELD,
-			irq | INTR_TYPE_EXT_INTR | INTR_INFO_VALID_MASK);
-}
-
-
-static void do_interrupt_requests(struct kvm_vcpu *vcpu,
-				       struct kvm_run *kvm_run)
-{
-	u32 cpu_based_vm_exec_control;
+	int handled = 0;
 
 	vcpu->interrupt_window_open =
 		((vmcs_readl(GUEST_RFLAGS) & X86_EFLAGS_IF) &&
 		 (vmcs_read32(GUEST_INTERRUPTIBILITY_INFO) & 3) == 0);
 
 	if (vcpu->interrupt_window_open &&
-	    vcpu->irq_summary &&
-	    !(vmcs_read32(VM_ENTRY_INTR_INFO_FIELD) & INTR_INFO_VALID_MASK))
+	    !(vmcs_read32(VM_ENTRY_INTR_INFO_FIELD) & INTR_INFO_VALID_MASK)) {
 		/*
-		 * If interrupts enabled, and not blocked by sti or mov ss. Good.
+		 * If interrupts enabled, and not blocked by sti or mov ss.
+		 * Good.
 		 */
-		kvm_do_inject_irq(vcpu);
+		struct kvm_irqack_data ack;
+		int r = 0;
+
+		memset(&ack, 0, sizeof(ack));
+
+		switch (pin) {
+		case kvm_irqpin_localint:
+			r = kvm_vcpu_irq_pop(vcpu, &ack);
+			break;
+		case kvm_irqpin_extint:
+			printk(KERN_WARNING "KVM: external-interrupts not " \
+			       "handled yet\n");
+			__clear_bit(pin, &vcpu->irq.pending);
+			break;
+		case kvm_irqpin_nmi:
+			/*
+			 * FIXME: Someday we will handle this using the
+			 * specific VMX NMI features.  For now, just inject
+			 * the NMI as a standard interrupt on vector 2
+			 */
+			ack.flags |= KVM_IRQACKDATA_VECTOR_VALID;
+			ack.vector = 2;
+			__clear_bit(pin, &vcpu->irq.pending);
+			break;
+		default:
+			panic("KVM: unknown interrupt pin raised: %d\n", pin);
+			break;
+		}
+
+		BUG_ON(r < 0);
+
+		if (ack.flags & KVM_IRQACKDATA_VECTOR_VALID) {
+			if (vcpu->rmode.active)
+				inject_rmode_irq(vcpu, ack.vector);
+			else
+				vmcs_write32(VM_ENTRY_INTR_INFO_FIELD,
+					     ack.vector |
+					     INTR_TYPE_EXT_INTR |
+					     INTR_INFO_VALID_MASK);
+
+			handled = 1;
+		}
+	}
 
-	cpu_based_vm_exec_control = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL);
 	if (!vcpu->interrupt_window_open &&
-	    (vcpu->irq_summary || kvm_run->request_interrupt_window))
+	    (__kvm_vcpu_irq_pending(vcpu) ||
+	     kvm_run->request_interrupt_window)) {
 		/*
 		 * Interrupts blocked.  Wait for unblock.
 		 */
-		cpu_based_vm_exec_control |= CPU_BASED_VIRTUAL_INTR_PENDING;
-	else
-		cpu_based_vm_exec_control &= ~CPU_BASED_VIRTUAL_INTR_PENDING;
-	vmcs_write32(CPU_BASED_VM_EXEC_CONTROL, cpu_based_vm_exec_control);
+		u32 cbvec = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL);
+		cbvec |= CPU_BASED_VIRTUAL_INTR_PENDING;
+		vmcs_write32(CPU_BASED_VM_EXEC_CONTROL, cbvec);
+	}
+
+	return handled;
+}
+
+static void clear_pending_controls(struct kvm_vcpu *vcpu)
+{
+	u32 cbvec = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL);
+	cbvec &= ~CPU_BASED_VIRTUAL_INTR_PENDING;
+	vmcs_write32(CPU_BASED_VM_EXEC_CONTROL, cbvec);
+}
+
+static void do_interrupt_requests(struct kvm_vcpu *vcpu,
+				  struct kvm_run *kvm_run)
+{
+	int pending = __kvm_vcpu_irq_all_pending(vcpu);
+
+	clear_pending_controls(vcpu);
+
+	while (pending) {
+		kvm_irqpin_t pin = __fls(pending);
+
+		switch (pin) {
+		case kvm_irqpin_localint:
+		case kvm_irqpin_extint:
+		case kvm_irqpin_nmi:
+			do_intr_requests(vcpu, kvm_run, pin);
+			break;
+		case kvm_irqpin_smi:
+			/* ignored (for now) */
+			printk(KERN_WARNING "KVM: dropping unhandled SMI\n");
+			__clear_bit(pin, &vcpu->irq.pending);
+			break;
+		case kvm_irqpin_invalid:
+			/* drop */
+			break;
+		default:
+			panic("KVM: unknown interrupt pin raised: %d\n", pin);
+			break;
+		}
+
+		__clear_bit(pin, &pending);
+	}
 }
 
 static void kvm_guest_debug_pre(struct kvm_vcpu *vcpu)
@@ -1511,9 +1577,13 @@ static int handle_exception(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	}
 
 	if (is_external_interrupt(vect_info)) {
+		/*
+		 * An exception was taken while we were trying to inject an
+		 * IRQ.  We must defer the injection of the vector until
+		 * the next window.
+		 */
 		int irq = vect_info & VECTORING_INFO_VECTOR_MASK;
-		set_bit(irq, vcpu->irq_pending);
-		set_bit(irq / BITS_PER_LONG, &vcpu->irq_summary);
+		kvm_vcpu_irq_push(vcpu, irq);
 	}
 
 	if ((intr_info & INTR_INFO_INTR_TYPE_MASK) == 0x200) { /* nmi */
@@ -1829,8 +1899,9 @@ static void post_kvm_run_save(struct kvm_vcpu *vcpu,
 	kvm_run->if_flag = (vmcs_readl(GUEST_RFLAGS) & X86_EFLAGS_IF) != 0;
 	kvm_run->cr8 = vcpu->cr8;
 	kvm_run->apic_base = vcpu->apic_base;
-	kvm_run->ready_for_interrupt_injection = (vcpu->interrupt_window_open &&
-						  vcpu->irq_summary == 0);
+	kvm_run->ready_for_interrupt_injection =
+		(vcpu->interrupt_window_open &&
+		 !kvm_vcpu_irq_pending(vcpu));
 }
 
 static int handle_interrupt_window(struct kvm_vcpu *vcpu,
@@ -1841,7 +1912,7 @@ static int handle_interrupt_window(struct kvm_vcpu *vcpu,
 	 * possible
 	 */
 	if (kvm_run->request_interrupt_window &&
-	    !vcpu->irq_summary) {
+	    !kvm_vcpu_irq_pending(vcpu)) {
 		kvm_run->exit_reason = KVM_EXIT_IRQ_WINDOW_OPEN;
 		++vcpu->stat.irq_window_exits;
 		return 0;
@@ -1852,7 +1923,7 @@ static int handle_interrupt_window(struct kvm_vcpu *vcpu,
 static int handle_halt(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 {
 	skip_emulated_instruction(vcpu);
-	if (vcpu->irq_summary)
+	if (kvm_vcpu_irq_pending(vcpu))
 		return 1;
 
 	kvm_run->exit_reason = KVM_EXIT_HLT;
@@ -1922,7 +1993,7 @@ static int kvm_handle_exit(struct kvm_run *kvm_run, struct kvm_vcpu *vcpu)
 static int dm_request_for_irq_injection(struct kvm_vcpu *vcpu,
 					  struct kvm_run *kvm_run)
 {
-	return (!vcpu->irq_summary &&
+	return (!kvm_vcpu_irq_pending(vcpu) &&
 		kvm_run->request_interrupt_window &&
 		vcpu->interrupt_window_open &&
 		(vmcs_readl(GUEST_RFLAGS) & X86_EFLAGS_IF));
@@ -1938,9 +2009,17 @@ preempted:
 		kvm_guest_debug_pre(vcpu);
 
 again:
+	spin_lock(&vcpu->irq.lock);
+
+	/*
+	 * We must inject interrupts (if any) while the irq.lock
+	 * is held
+	 */
 	if (!vcpu->mmio_read_completed)
 		do_interrupt_requests(vcpu, kvm_run);
 
+	spin_unlock(&vcpu->irq.lock);
+
 	vmx_save_host_state(vcpu);
 	kvm_load_guest_fpu(vcpu);
 


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

[PATCH 04/10] KVM: Adds ability to preempt an executing VCPU

[Gregory Haskins]

The VCPU executes synchronously w.r.t. userspace today, and therefore
interrupt injection is pretty straight forward.  However, we will soon need
to be able to inject interrupts asynchronous to the execution of the VCPU
due to the introduction of SMP, paravirtualized drivers, and asynchronous
hypercalls.  This patch adds support to the interrupt mechanism to force
a VCPU to VMEXIT when a new interrupt is pending.

Signed-off-by: Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
---

 drivers/kvm/kvm.h      |    2 ++
 drivers/kvm/kvm_main.c |   59 +++++++++++++++++++++++++++++++++++++++++++++++-
 drivers/kvm/svm.c      |   53 +++++++++++++++++++++++++++++++++++++++++--
 drivers/kvm/vmx.c      |   39 ++++++++++++++++++++++++++++++++
 4 files changed, 149 insertions(+), 4 deletions(-)

diff --git a/drivers/kvm/kvm.h b/drivers/kvm/kvm.h
index 41e4eaa..7b5d5e6 100644
--- a/drivers/kvm/kvm.h
+++ b/drivers/kvm/kvm.h
@@ -331,6 +331,8 @@ struct kvm_vcpu_irq {
 	struct kvm_irqdevice dev;
 	int                  pending;
 	int                  deferred;
+	struct task_struct  *task;
+	int                  guest_mode;
 };
 
 struct kvm_vcpu {
diff --git a/drivers/kvm/kvm_main.c b/drivers/kvm/kvm_main.c
index 3f47b20..3304cce 100644
--- a/drivers/kvm/kvm_main.c
+++ b/drivers/kvm/kvm_main.c
@@ -1891,6 +1891,9 @@ static int kvm_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 		kvm_arch_ops->decache_regs(vcpu);
 	}
 
+	vcpu->irq.task = current;
+	smp_wmb();
+
 	r = kvm_arch_ops->run(vcpu, kvm_run);
 
 out:
@@ -2332,6 +2335,20 @@ out1:
 }
 
 /*
+ * This function is invoked whenever we want to interrupt a vcpu that is
+ * currently executing in guest-mode.  It currently is a no-op because
+ * the simple delivery of the IPI to execute this function accomplishes our
+ * goal: To cause a VMEXIT.  We pass the vcpu (which contains the
+ * vcpu->irq.task, etc) for future use
+ */
+static void kvm_vcpu_guest_intr(void *info)
+{
+#ifdef NOT_YET
+	struct kvm_vcpu *vcpu = (struct kvm_vcpu*)info;
+#endif
+}
+
+/*
  * This function will be invoked whenever the vcpu->irq.dev raises its INTR
  * line
  */
@@ -2341,10 +2358,50 @@ static void kvm_vcpu_intr(struct kvm_irqsink *this,
 {
 	struct kvm_vcpu *vcpu = (struct kvm_vcpu*)this->private;
 	unsigned long flags;
+	int direct_ipi = -1;
 
 	spin_lock_irqsave(&vcpu->irq.lock, flags);
-	__set_bit(pin, &vcpu->irq.pending);
+
+	if (!test_bit(pin, &vcpu->irq.pending)) {
+		/*
+		 * Record the change..
+		 */
+		__set_bit(pin, &vcpu->irq.pending);
+
+		/*
+		 * then wake up the vcpu (if necessary)
+		 */
+		if (vcpu->irq.task && (vcpu->irq.task != current)) {
+			if (vcpu->irq.guest_mode) {
+				/*
+				 * If we are in guest mode, we can optimize
+				 * the IPI by executing a function directly
+				 * on the owning processor.
+				 */
+				direct_ipi = task_cpu(vcpu->irq.task);
+				BUG_ON(direct_ipi == smp_processor_id());
+			}
+		}
+	}
+
 	spin_unlock_irqrestore(&vcpu->irq.lock, flags);
+
+	/*
+	 * we can safely send the IPI outside of the lock-scope because the
+	 * irq.pending has already been updated.  This code assumes that
+	 * userspace will not sleep on anything other than HLT instructions.
+	 * HLT is covered in a race-free way because irq.pending was updated
+	 * in the critical section, and handle_halt() which check if any
+	 * interrupts are pending before returning to userspace.
+	 *
+	 * If it turns out that userspace can sleep on conditions other than
+	 * HLT, this code will need to be enhanced to allow the irq.pending
+	 * flags to be exported to userspace
+	 */
+	if (direct_ipi != -1)
+		smp_call_function_single(direct_ipi,
+					 kvm_vcpu_guest_intr,
+					 vcpu, 0, 0);
 }
 
 static void kvm_vcpu_irqsink_init(struct kvm_vcpu *vcpu)
diff --git a/drivers/kvm/svm.c b/drivers/kvm/svm.c
index 4c03881..4898a06 100644
--- a/drivers/kvm/svm.c
+++ b/drivers/kvm/svm.c
@@ -1542,11 +1542,46 @@ static int svm_vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	u16 gs_selector;
 	u16 ldt_selector;
 	int r;
+	unsigned long irq_flags;
 
 again:
+	/*
+	 * We disable interrupts until the next VMEXIT to eliminate a race
+	 * condition for delivery of virtual interrutps.  Note that this is
+	 * probably not as bad as it sounds, as interrupts will still invoke
+	 * a VMEXIT once transitioned to GUEST mode (and thus exit this lock
+	 * scope) even if they are disabled.
+	 *
+	 * FIXME: Do we need to do anything additional to mask IPI/NMIs?
+	 */
+	local_irq_save(irq_flags);
+
 	spin_lock(&vcpu->irq.lock);
 
 	/*
+	 * If there are any signals pending (virtual interrupt related or
+	 * otherwise), don't even bother trying to enter guest mode...
+	 */
+	if (signal_pending(current)) {
+		kvm_run->exit_reason = KVM_EXIT_INTR;
+		spin_unlock(&vcpu->irq.lock);
+		local_irq_restore(irq_flags);
+		r = -EINTR;
+		/*
+		 * FIXME: We probably want to move this whole lock-block below
+		 * the host->guest state loading so we don't restore when
+		 * the system was never saved to begin with
+		 */
+		goto out;
+	}
+
+	/*
+	 * There are optimizations we can make when signaling interrupts
+	 * if we know the VCPU is in GUEST mode, so mark that here
+	 */
+	vcpu->irq.guest_mode = 1;
+
+	/*
 	 * We must inject interrupts (if any) while the irq_lock
 	 * is held
 	 */
@@ -1688,6 +1723,15 @@ again:
 #endif
 		: "cc", "memory" );
 
+	/*
+	 * Signal that we have transitioned back to host mode
+	 */
+	spin_lock(&vcpu->irq.lock);
+	vcpu->irq.guest_mode = 0;
+	spin_unlock(&vcpu->irq.lock);
+
+	local_irq_restore(irq_flags);
+
 	if (vcpu->fpu_active) {
 		fx_save(vcpu->guest_fx_image);
 		fx_restore(vcpu->host_fx_image);
@@ -1734,20 +1778,23 @@ again:
 	if (r > 0) {
 		if (signal_pending(current)) {
 			++vcpu->stat.signal_exits;
-			post_kvm_run_save(vcpu, kvm_run);
 			kvm_run->exit_reason = KVM_EXIT_INTR;
-			return -EINTR;
+			r = -EINTR;
+			goto out;
 		}
 
 		if (dm_request_for_irq_injection(vcpu, kvm_run)) {
 			++vcpu->stat.request_irq_exits;
 			post_kvm_run_save(vcpu, kvm_run);
 			kvm_run->exit_reason = KVM_EXIT_INTR;
-			return -EINTR;
+			r = -EINTR;
+			goto out;
 		}
 		kvm_resched(vcpu);
 		goto again;
 	}
+
+ out:
 	post_kvm_run_save(vcpu, kvm_run);
 	return r;
 }
diff --git a/drivers/kvm/vmx.c b/drivers/kvm/vmx.c
index 4d51194..254c5b7 100644
--- a/drivers/kvm/vmx.c
+++ b/drivers/kvm/vmx.c
@@ -2003,15 +2003,45 @@ static int vmx_vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 {
 	u8 fail;
 	int r;
+	unsigned long irq_flags;
 
 preempted:
 	if (vcpu->guest_debug.enabled)
 		kvm_guest_debug_pre(vcpu);
 
 again:
+	/*
+	 * We disable interrupts until the next VMEXIT to eliminate a race
+	 * condition for delivery of virtual interrutps.  Note that this is
+	 * probably not as bad as it sounds, as interrupts will still invoke
+	 * a VMEXIT once transitioned to GUEST mode (and thus exit this lock
+	 * scope) even if they are disabled.
+	 *
+	 * FIXME: Do we need to do anything additional to mask IPI/NMIs?
+	 */
+	local_irq_save(irq_flags);
+
 	spin_lock(&vcpu->irq.lock);
 
 	/*
+	 * If there are any signals pending (virtual interrupt related or
+	 * otherwise), don't even bother trying to enter guest mode...
+	 */
+	if (signal_pending(current)) {
+		kvm_run->exit_reason = KVM_EXIT_INTR;
+		spin_unlock(&vcpu->irq.lock);
+		local_irq_restore(irq_flags);
+		r = -EINTR;
+		goto out;
+	}
+
+	/*
+	 * There are optimizations we can make when signaling interrupts
+	 * if we know the VCPU is in GUEST mode, so mark that here
+	 */
+	vcpu->irq.guest_mode = 1;
+
+	/*
 	 * We must inject interrupts (if any) while the irq.lock
 	 * is held
 	 */
@@ -2147,6 +2177,15 @@ again:
 		[cr2]"i"(offsetof(struct kvm_vcpu, cr2))
 	      : "cc", "memory" );
 
+	/*
+	 * Signal that we have transitioned back to host mode
+	 */
+	spin_lock(&vcpu->irq.lock);
+	vcpu->irq.guest_mode = 0;
+	spin_unlock(&vcpu->irq.lock);
+
+	local_irq_restore(irq_flags);
+
 	++vcpu->stat.exits;
 
 	vcpu->interrupt_window_open = (vmcs_read32(GUEST_INTERRUPTIBILITY_INFO) & 3) == 0;


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

[PATCH 05/10] KVM: Adds ability to signal userspace using a file-descriptor

[Gregory Haskins]

Signed-off-by: Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
---

 drivers/kvm/kvm.h      |    2 +
 drivers/kvm/kvm_main.c |   82 ++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 84 insertions(+), 0 deletions(-)

diff --git a/drivers/kvm/kvm.h b/drivers/kvm/kvm.h
index 7b5d5e6..1c46830 100644
--- a/drivers/kvm/kvm.h
+++ b/drivers/kvm/kvm.h
@@ -333,6 +333,8 @@ struct kvm_vcpu_irq {
 	int                  deferred;
 	struct task_struct  *task;
 	int                  guest_mode;
+	wait_queue_head_t    wq;
+	int                  usignal;
 };
 
 struct kvm_vcpu {
diff --git a/drivers/kvm/kvm_main.c b/drivers/kvm/kvm_main.c
index 3304cce..9a6d2c5 100644
--- a/drivers/kvm/kvm_main.c
+++ b/drivers/kvm/kvm_main.c
@@ -40,6 +40,7 @@
 #include <linux/file.h>
 #include <linux/fs.h>
 #include <linux/mount.h>
+#include <linux/poll.h>
 
 #include "x86_emulate.h"
 #include "segment_descriptor.h"
@@ -326,6 +327,7 @@ static struct kvm *kvm_create_vm(void)
 		memset(&vcpu->irq, 0, sizeof(vcpu->irq));
 		spin_lock_init(&vcpu->irq.lock);
 		vcpu->irq.deferred = -1;
+		init_waitqueue_head(&vcpu->irq.wq);
 
 		vcpu->cpu = -1;
 		vcpu->kvm = kvm;
@@ -2288,11 +2290,78 @@ static int kvm_vcpu_release(struct inode *inode, struct file *filp)
 	return 0;
 }
 
+static unsigned int kvm_vcpu_poll(struct file *filp, poll_table *wait)
+{
+	struct kvm_vcpu *vcpu = filp->private_data;
+	unsigned int events = 0;
+	unsigned long flags;
+
+	poll_wait(filp, &vcpu->irq.wq, wait);
+
+	spin_lock_irqsave(&vcpu->irq.lock, flags);
+	if (vcpu->irq.usignal)
+		events |= POLLIN;
+	spin_unlock_irqrestore(&vcpu->irq.lock, flags);
+
+	return events;
+}
+
+static ssize_t kvm_vcpu_read(struct file *filp, char __user *buf, size_t count,
+			     loff_t *ppos)
+{
+	struct kvm_vcpu *vcpu = filp->private_data;
+	ssize_t res = -EAGAIN;
+	DECLARE_WAITQUEUE(wait, current);
+	unsigned long flags;
+	int val;
+
+	if (count < sizeof(vcpu->irq.usignal))
+		return -EINVAL;
+
+	spin_lock_irqsave(&vcpu->irq.lock, flags);
+
+	val = vcpu->irq.usignal;
+
+	if (val > 0)
+		res = sizeof(val);
+	else if (!(filp->f_flags & O_NONBLOCK)) {
+		__add_wait_queue(&vcpu->irq.wq, &wait);
+		for (res = 0;;) {
+			set_current_state(TASK_INTERRUPTIBLE);
+			if (val > 0) {
+				res = sizeof(val);
+				break;
+			}
+			if (signal_pending(current)) {
+				res = -ERESTARTSYS;
+				break;
+			}
+			spin_unlock_irqrestore(&vcpu->irq.lock, flags);
+			schedule();
+			spin_lock_irqsave(&vcpu->irq.lock, flags);
+		}
+		__remove_wait_queue(&vcpu->irq.wq, &wait);
+		__set_current_state(TASK_RUNNING);
+	}
+
+	if (res > 0)
+		vcpu->irq.usignal = 0;
+
+	spin_unlock_irqrestore(&vcpu->irq.lock, flags);
+
+	if (res > 0 && put_user(val, (int __user *) buf))
+		return -EFAULT;
+
+	return res;
+}
+
 static struct file_operations kvm_vcpu_fops = {
 	.release        = kvm_vcpu_release,
 	.unlocked_ioctl = kvm_vcpu_ioctl,
 	.compat_ioctl   = kvm_vcpu_ioctl,
 	.mmap           = kvm_vcpu_mmap,
+	.poll           = kvm_vcpu_poll,
+	.read           = kvm_vcpu_read,
 };
 
 /*
@@ -2359,6 +2428,7 @@ static void kvm_vcpu_intr(struct kvm_irqsink *this,
 	struct kvm_vcpu *vcpu = (struct kvm_vcpu*)this->private;
 	unsigned long flags;
 	int direct_ipi = -1;
+	int indirect_sig = 0;
 
 	spin_lock_irqsave(&vcpu->irq.lock, flags);
 
@@ -2380,6 +2450,15 @@ static void kvm_vcpu_intr(struct kvm_irqsink *this,
 				 */
 				direct_ipi = task_cpu(vcpu->irq.task);
 				BUG_ON(direct_ipi == smp_processor_id());
+			} else {
+				/*
+				 * otherwise, we must assume that we could be
+				 * blocked anywhere, including userspace. Send
+				 * a signal to give everyone a chance to get
+				 * notification
+				 */
+				vcpu->irq.usignal++;
+				indirect_sig = 1;
 			}
 		}
 	}
@@ -2402,6 +2481,9 @@ static void kvm_vcpu_intr(struct kvm_irqsink *this,
 		smp_call_function_single(direct_ipi,
 					 kvm_vcpu_guest_intr,
 					 vcpu, 0, 0);
+
+	if (indirect_sig && waitqueue_active(&vcpu->irq.wq))
+		wake_up(&vcpu->irq.wq);
 }
 
 static void kvm_vcpu_irqsink_init(struct kvm_vcpu *vcpu)


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 05/10] KVM: Adds ability to signal userspace using a file-descriptor

[Avi Kivity]

Gregory Haskins wrote:
> Signed-off-by: Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
>   

Please include patch descriptions.

> ---
>
>  drivers/kvm/kvm.h      |    2 +
>  drivers/kvm/kvm_main.c |   82 ++++++++++++++++++++++++++++++++++++++++++++++++
>  2 files changed, 84 insertions(+), 0 deletions(-)
>
> diff --git a/drivers/kvm/kvm.h b/drivers/kvm/kvm.h
> index 7b5d5e6..1c46830 100644
> --- a/drivers/kvm/kvm.h
> +++ b/drivers/kvm/kvm.h
> @@ -333,6 +333,8 @@ struct kvm_vcpu_irq {
>  	int                  deferred;
>  	struct task_struct  *task;
>  	int                  guest_mode;
> +	wait_queue_head_t    wq;
> +	int                  usignal;
>  };
>  
>  struct kvm_vcpu {
> diff --git a/drivers/kvm/kvm_main.c b/drivers/kvm/kvm_main.c
> index 3304cce..9a6d2c5 100644
> --- a/drivers/kvm/kvm_main.c
> +++ b/drivers/kvm/kvm_main.c
> @@ -40,6 +40,7 @@
>  #include <linux/file.h>
>  #include <linux/fs.h>
>  #include <linux/mount.h>
> +#include <linux/poll.h>
>  
>  #include "x86_emulate.h"
>  #include "segment_descriptor.h"
> @@ -326,6 +327,7 @@ static struct kvm *kvm_create_vm(void)
>  		memset(&vcpu->irq, 0, sizeof(vcpu->irq));
>  		spin_lock_init(&vcpu->irq.lock);
>  		vcpu->irq.deferred = -1;
> +		init_waitqueue_head(&vcpu->irq.wq);
>  
>  		vcpu->cpu = -1;
>  		vcpu->kvm = kvm;
> @@ -2288,11 +2290,78 @@ static int kvm_vcpu_release(struct inode *inode, struct file *filp)
>  	return 0;
>  }
>  
> +static unsigned int kvm_vcpu_poll(struct file *filp, poll_table *wait)
> +{
> +	struct kvm_vcpu *vcpu = filp->private_data;
> +	unsigned int events = 0;
> +	unsigned long flags;
> +
> +	poll_wait(filp, &vcpu->irq.wq, wait);
> +
> +	spin_lock_irqsave(&vcpu->irq.lock, flags);
> +	if (vcpu->irq.usignal)
> +		events |= POLLIN;
> +	spin_unlock_irqrestore(&vcpu->irq.lock, flags);
> +
> +	return events;
> +}
> +
> +static ssize_t kvm_vcpu_read(struct file *filp, char __user *buf, size_t count,
> +			     loff_t *ppos)
> +{
>   

Is having a read() (or a write()) actually necessary?

> +
> +	if (indirect_sig && waitqueue_active(&vcpu->irq.wq))
> +		wake_up(&vcpu->irq.wq);
>  }
>  
>   

Did you check that we can actually deliver signals with this?  I think a 
fasync_struct or something like that is necessary, but not sure.

Another implementation option (which I've only thought of now, sorry) is 
to have an ioctl which returns a real eventfd, reducing some code 
duplication.

-- 
error compiling committee.c: too many arguments to function


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 05/10] KVM: Adds ability to signal userspace using a file-descriptor

[Gregory Haskins]

>>> On Sun, May 13, 2007 at  9:00 AM, in message <46470BF7.5080108-atKUWr5tajBWk0Htik3J/w@public.gmane.org>,
Avi Kivity <avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote: 
> Gregory Haskins wrote:
>> Signed- off- by: Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
>>   
> 
> Please include patch descriptions.

Ack.

On that topic: Does anyone know how to retroactively change the patch comment in StGIT?

> 
>> ---
>>
>>  drivers/kvm/kvm.h      |    2 +
>>  drivers/kvm/kvm_main.c |   82 
> ++++++++++++++++++++++++++++++++++++++++++++++++
>>  2 files changed, 84 insertions(+), 0 deletions(- )
>>
>> diff -- git a/drivers/kvm/kvm.h b/drivers/kvm/kvm.h
>> index 7b5d5e6..1c46830 100644
>> ---  a/drivers/kvm/kvm.h
>> +++ b/drivers/kvm/kvm.h
>> @@ - 333,6 +333,8 @@ struct kvm_vcpu_irq {
>>  	int                  deferred;
>>  	struct task_struct  *task;
>>  	int                  guest_mode;
>> +	wait_queue_head_t    wq;
>> +	int                  usignal;
>>  };
>>  
>>  struct kvm_vcpu {
>> diff -- git a/drivers/kvm/kvm_main.c b/drivers/kvm/kvm_main.c
>> index 3304cce..9a6d2c5 100644
>> ---  a/drivers/kvm/kvm_main.c
>> +++ b/drivers/kvm/kvm_main.c
>> @@ - 40,6 +40,7 @@
>>  #include <linux/file.h>
>>  #include <linux/fs.h>
>>  #include <linux/mount.h>
>> +#include <linux/poll.h>
>>  
>>  #include "x86_emulate.h"
>>  #include "segment_descriptor.h"
>> @@ - 326,6 +327,7 @@ static struct kvm *kvm_create_vm(void)
>>  		memset(&vcpu- >irq, 0, sizeof(vcpu- >irq));
>>  		spin_lock_init(&vcpu- >irq.lock);
>>  		vcpu- >irq.deferred = - 1;
>> +		init_waitqueue_head(&vcpu- >irq.wq);
>>  
>>  		vcpu- >cpu = - 1;
>>  		vcpu- >kvm = kvm;
>> @@ - 2288,11 +2290,78 @@ static int kvm_vcpu_release(struct inode *inode, 
> struct file *filp)
>>  	return 0;
>>  }
>>  
>> +static unsigned int kvm_vcpu_poll(struct file *filp, poll_table *wait)
>> +{
>> +	struct kvm_vcpu *vcpu = filp- >private_data;
>> +	unsigned int events = 0;
>> +	unsigned long flags;
>> +
>> +	poll_wait(filp, &vcpu- >irq.wq, wait);
>> +
>> +	spin_lock_irqsave(&vcpu- >irq.lock, flags);
>> +	if (vcpu- >irq.usignal)
>> +		events |= POLLIN;
>> +	spin_unlock_irqrestore(&vcpu- >irq.lock, flags);
>> +
>> +	return events;
>> +}
>> +
>> +static ssize_t kvm_vcpu_read(struct file *filp, char __user *buf, size_t 
> count,
>> +			     loff_t *ppos)
>> +{
>>   
> 
> Is having a read() (or a write()) actually necessary?

Based on what I know: yes.  It could be a case of ignorance, however ;)

Heres why I think its necessary:  You need poll to simply tell you when something is pending.  You can't clear the pending status in poll because you cannot predict the internal access pattern (e.g. I assume it could be polled multiple times by the kernel without returning immediately to userspace).  Therefore, you need a second method to actually clear the pending "signal", which I use the read() method for.  I can be convinced otherwise, but that was my original thinking.

> 
>> +
>> +	if (indirect_sig && waitqueue_active(&vcpu- >irq.wq))
>> +		wake_up(&vcpu- >irq.wq);
>>  }
>>  
>>   
> 
> Did you check that we can actually deliver signals with this?  I think a 
> fasync_struct or something like that is necessary, but not sure.

Actually, my signals *didn't* seem to be working, but they werent working with "send_sig()" either so I just assumed I had a userspace coding problem.  Based on what I read, it seemed like what I did should work if you do a fcntl(F_SETSIG), etc.  But again, it could be ignorance.  I am not familiar with fasync_struct.  If you have any pointers, please forward.

> 
> Another implementation option (which I've only thought of now, sorry) is 
> to have an ioctl which returns a real eventfd, reducing some code 
> duplication.

So based on this, I assume eventfd must be in the kernel already?  Cool.  Even if its not, I like this idea much better than what I did.  There was still an unresolved problem regarding how I was going to expose the signaling mechanism to QEMU without giving away the vcpu_fd from the kvmctl library that this solves nicely.

With this methodology, I can simply provide a function like "kvm_vcpu_get_eventfd()" in the library, and return the eventfd directly to the QEMU process.  Then we dont have to worry about layering violations.  Nice!


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 05/10] KVM: Adds ability to signal userspace using a file-descriptor

[Avi Kivity]

Gregory Haskins wrote:
>>
>> Is having a read() (or a write()) actually necessary?
>>     
>
> Based on what I know: yes.  It could be a case of ignorance, however ;)
>
> Heres why I think its necessary:  You need poll to simply tell you when something is pending.  You can't clear the pending status in poll because you cannot predict the internal access pattern (e.g. I assume it could be polled multiple times by the kernel without returning immediately to userspace).  Therefore, you need a second method to actually clear the pending "signal", which I use the read() method for.  I can be convinced otherwise, but that was my original thinking.
>   

I think you are right, but am cc'ing an expert. Davide, we're using an 
fd to signal something to userspace, but have nothing to actually read() 
or write(). Is a read() or write() avoidable?

>   
>>> +
>>> +	if (indirect_sig && waitqueue_active(&vcpu- >irq.wq))
>>> +		wake_up(&vcpu- >irq.wq);
>>>  }
>>>  
>>>   
>>>       
>> Did you check that we can actually deliver signals with this?  I think a 
>> fasync_struct or something like that is necessary, but not sure.
>>     
>
> Actually, my signals *didn't* seem to be working, but they werent working with "send_sig()" either so I just assumed I had a userspace coding problem.  Based on what I read, it seemed like what I did should work if you do a fcntl(F_SETSIG), etc.  But again, it could be ignorance.  I am not familiar with fasync_struct.  If you have any pointers, please forward.
>
>   

fs/pipe.c. hairy stuff.

>> Another implementation option (which I've only thought of now, sorry) is 
>> to have an ioctl which returns a real eventfd, reducing some code 
>> duplication.
>>     
>
> So based on this, I assume eventfd must be in the kernel already?  Cool.

It is in 2.6.22-rc1. As is the anonymous inodes source which can be used 
to retire kvmfs (which will probably break the record for shortest-lived 
filesystem ever).

>   Even if its not, I like this idea much better than what I did.  There was still an unresolved problem regarding how I was going to expose the signaling mechanism to QEMU without giving away the vcpu_fd from the kvmctl library that this solves nicely.
>
> With this methodology, I can simply provide a function like "kvm_vcpu_get_eventfd()" in the library, and return the eventfd directly to the QEMU process.  Then we dont have to worry about layering violations.  Nice!
>   

I hadn't though of it. Looks like a win from all directions.

It means we need to package eventfd for the external module, but that's 
easily done.

-- 
error compiling committee.c: too many arguments to function


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 05/10] KVM: Adds ability to signal userspace using a file-descriptor

[Gregory Haskins]

>>> On Mon, May 14, 2007 at  8:22 AM, in message <46485488.2010608-atKUWr5tajBWk0Htik3J/w@public.gmane.org>,
Avi Kivity <avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote: 
>> So based on this, I assume eventfd must be in the kernel already?  Cool.
> 
> It is in 2.6.22- rc1. As is the anonymous inodes source which can be used 
> to retire kvmfs (which will probably break the record for shortest- lived 
> filesystem ever).
> 

I just did a search against my kvm.git HEAD and do not see anything related to eventfd.  Does this mean I should pull from linus' tree?  If I do this, will it still work in your tree?  (Sorry...relative git-newbie here).

-Greg  

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 05/10] KVM: Adds ability to signal userspace using a file-descriptor

[Avi Kivity]

Gregory Haskins wrote:
>>>> On Mon, May 14, 2007 at  8:22 AM, in message <46485488.2010608-atKUWr5tajBWk0Htik3J/w@public.gmane.org>,
>>>>         
> Avi Kivity <avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote: 
>   
>>> So based on this, I assume eventfd must be in the kernel already?  Cool.
>>>       
>> It is in 2.6.22- rc1. As is the anonymous inodes source which can be used 
>> to retire kvmfs (which will probably break the record for shortest- lived 
>> filesystem ever).
>>
>>     
>
> I just did a search against my kvm.git HEAD and do not see anything related to eventfd.  Does this mean I should pull from linus' tree?  If I do this, will it still work in your tree?  (Sorry...relative git-newbie here).
>
>   

kvm.git has eventfd merged.  See 
http://git.kernel.org/gitweb.cgi?p=linux/kernel/git/avi/kvm.git;a=blob;f=fs/eventfd.c;h=480e2b3c4166a85be538d6f2c5edc25eace5ec6f;hb=HEAD 
.

You were probably hit by the sync delay between master.kernel.org mand 
the mirrors, or pulled before I pushed.

-- 
error compiling committee.c: too many arguments to function


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 05/10] KVM: Adds ability to signal userspace using a file-descriptor

[Gregory Haskins]

>>> On Mon, May 14, 2007 at 10:42 AM, in message <4648756D.5040001-atKUWr5tajBWk0Htik3J/w@public.gmane.org>,
Avi Kivity <avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote: 
> Gregory Haskins wrote:
>>>>> On Mon, May 14, 2007 at  8:22 AM, in message <46485488.2010608-atKUWr5tajBWk0Htik3J/w@public.gmane.org>,
>>>>>         
>> Avi Kivity <avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote: 
>>   
>>>> So based on this, I assume eventfd must be in the kernel already?  Cool.
>>>>       
>>> It is in 2.6.22-  rc1. As is the anonymous inodes source which can be used 
>>> to retire kvmfs (which will probably break the record for shortest-  lived 
>>> filesystem ever).
>>>
>>>     
>>
>> I just did a search against my kvm.git HEAD and do not see anything related 
> to eventfd.  Does this mean I should pull from linus' tree?  If I do this, 
> will it still work in your tree?  (Sorry...relative git- newbie here).
>>
>>   
> 
> kvm.git has eventfd merged.  See 
> http://git.kernel.org/gitweb.cgi?p=linux/kernel/git/avi/kvm.git;a=blob;f=fs/
> eventfd.c;h=480e2b3c4166a85be538d6f2c5edc25eace5ec6f;hb=HEAD 
> .
> 
> You were probably hit by the sync delay between master.kernel.org mand 
> the mirrors, or pulled before I pushed.


Ah thanks.  I see it now.  I am working on the new design that uses the existing eventfd() and should have a v4 patch for you soon (which includes you other feedback).  Stay tuned.

-Greg

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 05/10] KVM: Adds ability to signal userspace using a file-descriptor

[Davide Libenzi]

On Mon, 14 May 2007, Avi Kivity wrote:

> Gregory Haskins wrote:
> > > 
> > > Is having a read() (or a write()) actually necessary?
> > >     
> > 
> > Based on what I know: yes.  It could be a case of ignorance, however ;)
> > 
> > Heres why I think its necessary:  You need poll to simply tell you when
> > something is pending.  You can't clear the pending status in poll because
> > you cannot predict the internal access pattern (e.g. I assume it could be
> > polled multiple times by the kernel without returning immediately to
> > userspace).  Therefore, you need a second method to actually clear the
> > pending "signal", which I use the read() method for.  I can be convinced
> > otherwise, but that was my original thinking.
> >   
> 
> I think you are right, but am cc'ing an expert. Davide, we're using an fd to
> signal something to userspace, but have nothing to actually read() or write().
> Is a read() or write() avoidable?

I don't know exactly what you want to do, but signalfd signal de-queueing 
competes with the standard Linux signal delivery, if signals are not 
blocked.
So if you don't want to read() the signal, you can just leave the signal 
unblocked, and it'll be delivered to the signal handler.
You can even leave the signal blocked and avoid read(), but poll() on the 
signalfd will always return POLLIN if the sigmask includes the pending 
signal.




- Davide



-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 05/10] KVM: Adds ability to signal userspace using a file-descriptor

[Avi Kivity]

Davide Libenzi wrote:
> On Mon, 14 May 2007, Avi Kivity wrote:
>
>   
>> Gregory Haskins wrote:
>>     
>>>> Is having a read() (or a write()) actually necessary?
>>>>     
>>>>         
>>> Based on what I know: yes.  It could be a case of ignorance, however ;)
>>>
>>> Heres why I think its necessary:  You need poll to simply tell you when
>>> something is pending.  You can't clear the pending status in poll because
>>> you cannot predict the internal access pattern (e.g. I assume it could be
>>> polled multiple times by the kernel without returning immediately to
>>> userspace).  Therefore, you need a second method to actually clear the
>>> pending "signal", which I use the read() method for.  I can be convinced
>>> otherwise, but that was my original thinking.
>>>   
>>>       
>> I think you are right, but am cc'ing an expert. Davide, we're using an fd to
>> signal something to userspace, but have nothing to actually read() or write().
>> Is a read() or write() avoidable?
>>     
>
> I don't know exactly what you want to do, but signalfd signal de-queueing 
> competes with the standard Linux signal delivery, if signals are not 
> blocked.
> So if you don't want to read() the signal, you can just leave the signal 
> unblocked, and it'll be delivered to the signal handler.
> You can even leave the signal blocked and avoid read(), but poll() on the 
> signalfd will always return POLLIN if the sigmask includes the pending 
> signal.
>   

This is not about a real signal.  We have an fd (for a pseudo 
filesystem) which wants to indicate its readiness to select(), but which 
doesn't have any real data to produce.  Is it possible to implement this 
without a read() or a write()?

We're also looking at using an eventfd for this, so this may be moot.

-- 
error compiling committee.c: too many arguments to function


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 05/10] KVM: Adds ability to signal userspace using a file-descriptor

[Davide Libenzi]

On Mon, 14 May 2007, Avi Kivity wrote:

> Davide Libenzi wrote:
> > On Mon, 14 May 2007, Avi Kivity wrote:
> > 
> >   
> > > Gregory Haskins wrote:
> > >     
> > > > > Is having a read() (or a write()) actually necessary?
> > > > >             
> > > > Based on what I know: yes.  It could be a case of ignorance, however ;)
> > > > 
> > > > Heres why I think its necessary:  You need poll to simply tell you when
> > > > something is pending.  You can't clear the pending status in poll
> > > > because
> > > > you cannot predict the internal access pattern (e.g. I assume it could
> > > > be
> > > > polled multiple times by the kernel without returning immediately to
> > > > userspace).  Therefore, you need a second method to actually clear the
> > > > pending "signal", which I use the read() method for.  I can be convinced
> > > > otherwise, but that was my original thinking.
> > > >         
> > > I think you are right, but am cc'ing an expert. Davide, we're using an fd
> > > to
> > > signal something to userspace, but have nothing to actually read() or
> > > write().
> > > Is a read() or write() avoidable?
> > >     
> > 
> > I don't know exactly what you want to do, but signalfd signal de-queueing
> > competes with the standard Linux signal delivery, if signals are not
> > blocked.
> > So if you don't want to read() the signal, you can just leave the signal
> > unblocked, and it'll be delivered to the signal handler.
> > You can even leave the signal blocked and avoid read(), but poll() on the
> > signalfd will always return POLLIN if the sigmask includes the pending
> > signal.
> >   
> 
> This is not about a real signal.  We have an fd (for a pseudo filesystem)
> which wants to indicate its readiness to select(), but which doesn't have any
> real data to produce.  Is it possible to implement this without a read() or a
> write()?
> 
> We're also looking at using an eventfd for this, so this may be moot.

Do you close the signaled fd after receiving the signal/event? If you 
don't close it, eventfd will always return ready (POLLIN). With the 
signalfd, you can leave the signal un-blocked and the signal will be 
delivered through the standard signal delivery, and the ready condition 
will be cleared.



- Davide



-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 05/10] KVM: Adds ability to signal userspace using a file-descriptor

[Avi Kivity]

Davide Libenzi wrote:
> On Mon, 14 May 2007, Avi Kivity wrote:
>
>   
>> Davide Libenzi wrote:
>>     
>>> On Mon, 14 May 2007, Avi Kivity wrote:
>>>
>>>   
>>>       
>>>> Gregory Haskins wrote:
>>>>     
>>>>         
>>>>>> Is having a read() (or a write()) actually necessary?
>>>>>>             
>>>>>>             
>>>>> Based on what I know: yes.  It could be a case of ignorance, however ;)
>>>>>
>>>>> Heres why I think its necessary:  You need poll to simply tell you when
>>>>> something is pending.  You can't clear the pending status in poll
>>>>> because
>>>>> you cannot predict the internal access pattern (e.g. I assume it could
>>>>> be
>>>>> polled multiple times by the kernel without returning immediately to
>>>>> userspace).  Therefore, you need a second method to actually clear the
>>>>> pending "signal", which I use the read() method for.  I can be convinced
>>>>> otherwise, but that was my original thinking.
>>>>>         
>>>>>           
>>>> I think you are right, but am cc'ing an expert. Davide, we're using an fd
>>>> to
>>>> signal something to userspace, but have nothing to actually read() or
>>>> write().
>>>> Is a read() or write() avoidable?
>>>>     
>>>>         
>>> I don't know exactly what you want to do, but signalfd signal de-queueing
>>> competes with the standard Linux signal delivery, if signals are not
>>> blocked.
>>> So if you don't want to read() the signal, you can just leave the signal
>>> unblocked, and it'll be delivered to the signal handler.
>>> You can even leave the signal blocked and avoid read(), but poll() on the
>>> signalfd will always return POLLIN if the sigmask includes the pending
>>> signal.
>>>   
>>>       
>> This is not about a real signal.  We have an fd (for a pseudo filesystem)
>> which wants to indicate its readiness to select(), but which doesn't have any
>> real data to produce.  Is it possible to implement this without a read() or a
>> write()?
>>
>> We're also looking at using an eventfd for this, so this may be moot.
>>     
>
> Do you close the signaled fd after receiving the signal/event? If you 
> don't close it, eventfd will always return ready (POLLIN).
>   

We don't.  Anyway, that's what we thought.  Thanks for the confirmation.


-- 
error compiling committee.c: too many arguments to function


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 05/10] KVM: Adds ability to signal userspace using a file-descriptor

[Gregory Haskins]

>>> On Mon, May 14, 2007 at  1:23 PM, in message <46489B12.8030807-atKUWr5tajBWk0Htik3J/w@public.gmane.org>,
Avi Kivity <avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote: 
>
>> Do you close the signaled fd after receiving the signal/event? If you 
>> don't close it, eventfd will always return ready (POLLIN).
>>   
> 
> We don't.  Anyway, that's what we thought.  Thanks for the confirmation.
> 

And plus I just finished converting to Davide's eventfd, so its moot ;)

On that topic, I could use some advice:

I was originally planning on adding a new ioctl like KVM_VCPU_CREATE_EVENTFD which would allocate a new eventfd and return it.  However, I soon realized that the only method to create an eventfd is sys_eventfd(), which is not exported by the eventfd.h headerfile (presumably this must be a new system call).

So based on that, I figured I would change the model so that the usermode app should call the eventfd open() call on its own, and then they could register the fd with me.  So KVM_VCPU_CREATE_EVENTFD becomes KVM_VCPU_SET_EVENTFD (where -1 "unregisters" it).

Does this sound like a reasonable approach?  If so, how does the usermode app actually open the eventfd today?  Is there a new glibc that I need to get the new system call?  Or can the app use open() somehow?  If open(), what is the path that should be specified?

Conversely, if my first approach was the right one how do I invoke the sys_eventfd()?  Is there a way to invoke system calls in kernel mode?  A better way?

Any advice appreciated.

-Greg




-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 05/10] KVM: Adds ability to signal userspace using a file-descriptor

[Davide Libenzi]

On Mon, 14 May 2007, Gregory Haskins wrote:

> >>> On Mon, May 14, 2007 at  1:23 PM, in message <46489B12.8030807-atKUWr5tajBWk0Htik3J/w@public.gmane.org>,
> Avi Kivity <avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote: 
> >
> >> Do you close the signaled fd after receiving the signal/event? If you 
> >> don't close it, eventfd will always return ready (POLLIN).
> >>   
> > 
> > We don't.  Anyway, that's what we thought.  Thanks for the confirmation.
> > 
> 
> And plus I just finished converting to Davide's eventfd, so its moot ;)
> 
> On that topic, I could use some advice:
> 
> I was originally planning on adding a new ioctl like KVM_VCPU_CREATE_EVENTFD which would allocate a new eventfd and return it.  However, I soon realized that the only method to create an eventfd is sys_eventfd(), which is not exported by the eventfd.h headerfile (presumably this must be a new system call).
> 
> So based on that, I figured I would change the model so that the usermode app should call the eventfd open() call on its own, and then they could register the fd with me.  So KVM_VCPU_CREATE_EVENTFD becomes KVM_VCPU_SET_EVENTFD (where -1 "unregisters" it).
> 
> Does this sound like a reasonable approach?  If so, how does the usermode app actually open the eventfd today?  Is there a new glibc that I need to get the new system call?  Or can the app use open() somehow?  If open(), what is the path that should be specified?
> 
> Conversely, if my first approach was the right one how do I invoke the sys_eventfd()?  Is there a way to invoke system calls in kernel mode?  A better way?
> 
> Any advice appreciated.

The eventfd syscall is defined in include/linux/syscalls.h
>From userspace, till glibc aligns:

#include <sys/syscall.h>

#ifndef __NR_eventfd
#if defined(__x86_64__)
#define __NR_eventfd 283
#elif defined(__i386__)
#define __NR_eventfd 323
#else
#error Cannot detect your architecture!
#endif
#endif

static int eventfd(int count) {

        return syscall(__NR_eventfd, count);
}


If the kernel side receives an fd from userspace, it must use:

file = eventfd_fget(fd);
if (IS_ERR(file))
    ....
eventfd_signal(file, 1);
fput(file);



- Davide



-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 05/10] KVM: Adds ability to signal userspace using a file-descriptor

[Gregory Haskins]

>>> On Mon, May 14, 2007 at  3:12 PM, in message
<Pine.LNX.4.64.0705141207070.19682-GPJ85BhbkB8RepQJljzAVbITYcZ0+W3JAL8bYrjMMd8@public.gmane.org>, Davide Libenzi
<davidel-AhlLAIvw+VEjIGhXcJzhZg@public.gmane.org> wrote: 
> On Mon, 14 May 2007, Gregory Haskins wrote:
> 
>> >>> On Mon, May 14, 2007 at  1:23 PM, in message <46489B12.8030807-atKUWr5tajBWk0Htik3J/w@public.gmane.org>,
>> Avi Kivity <avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote: 
>> >
>> >> Do you close the signaled fd after receiving the signal/event? If you 
>> >> don't close it, eventfd will always return ready (POLLIN).
>> >>   
>> > 
>> > We don't.  Anyway, that's what we thought.  Thanks for the confirmation.
>> > 
>> 
>> And plus I just finished converting to Davide's eventfd, so its moot ;)
>> 
>> On that topic, I could use some advice:
>> 
>> I was originally planning on adding a new ioctl like KVM_VCPU_CREATE_EVENTFD 
> which would allocate a new eventfd and return it.  However, I soon realized 
> that the only method to create an eventfd is sys_eventfd(), which is not 
> exported by the eventfd.h headerfile (presumably this must be a new system 
> call).
>> 
>> So based on that, I figured I would change the model so that the usermode 
> app should call the eventfd open() call on its own, and then they could 
> register the fd with me.  So KVM_VCPU_CREATE_EVENTFD becomes 
> KVM_VCPU_SET_EVENTFD (where - 1 "unregisters" it).
>> 
>> Does this sound like a reasonable approach?  If so, how does the usermode 
> app actually open the eventfd today?  Is there a new glibc that I need to get 
> the new system call?  Or can the app use open() somehow?  If open(), what is 
> the path that should be specified?
>> 
>> Conversely, if my first approach was the right one how do I invoke the 
> sys_eventfd()?  Is there a way to invoke system calls in kernel mode?  A 
> better way?
>> 
>> Any advice appreciated.
> 
> The eventfd syscall is defined in include/linux/syscalls.h
> From userspace, till glibc aligns:
> 
> #include <sys/syscall.h>
> 
> #ifndef __NR_eventfd
> #if defined(__x86_64__)
> #define __NR_eventfd 283
> #elif defined(__i386__)
> #define __NR_eventfd 323
> #else
> #error Cannot detect your architecture!
> #endif
> #endif
> 
> static int eventfd(int count) {
> 
>         return syscall(__NR_eventfd, count);
> }
> 
> 
> If the kernel side receives an fd from userspace, it must use:
> 
> file = eventfd_fget(fd);
> if (IS_ERR(file))
>     ....
> eventfd_signal(file, 1);
> fput(file);
> 
> 
> 
> -  Davide

Thanks Davide,
   That is very helpful.  Is there any reason why we can't export eventfd_signal() and eventfd_fget() to modules?

-Greg

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 05/10] KVM: Adds ability to signal userspace using a file-descriptor

[Davide Libenzi]

On Mon, 14 May 2007, Gregory Haskins wrote:

> Thanks Davide,
>  That is very helpful.  Is there any reason why we can't export 
>  eventfd_signal() and eventfd_fget() to modules?

I'll push a patch for that.



- Davide



-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

[PATCH 06/10] KVM: Add support for in-kernel LAPIC model

[Gregory Haskins]

Signed-off-by: Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
---

 drivers/kvm/Makefile   |    2 
 drivers/kvm/kernint.c  |  149 +++++
 drivers/kvm/kvm.h      |   35 +
 drivers/kvm/kvm_main.c |  179 +++++-
 drivers/kvm/lapic.c    | 1421 ++++++++++++++++++++++++++++++++++++++++++++++++
 drivers/kvm/svm.c      |   13 
 drivers/kvm/userint.c  |    8 
 drivers/kvm/vmx.c      |   16 -
 include/linux/kvm.h    |   16 +
 9 files changed, 1798 insertions(+), 41 deletions(-)

diff --git a/drivers/kvm/Makefile b/drivers/kvm/Makefile
index 540afbc..1aad737 100644
--- a/drivers/kvm/Makefile
+++ b/drivers/kvm/Makefile
@@ -2,7 +2,7 @@
 # Makefile for Kernel-based Virtual Machine module
 #
 
-kvm-objs := kvm_main.o mmu.o x86_emulate.o userint.o
+kvm-objs := kvm_main.o mmu.o x86_emulate.o userint.o lapic.o kernint.o
 obj-$(CONFIG_KVM) += kvm.o
 kvm-intel-objs = vmx.o
 obj-$(CONFIG_KVM_INTEL) += kvm-intel.o
diff --git a/drivers/kvm/kernint.c b/drivers/kvm/kernint.c
new file mode 100644
index 0000000..b5cbcae
--- /dev/null
+++ b/drivers/kvm/kernint.c
@@ -0,0 +1,149 @@
+/*
+ * Kernel Interrupt IRQ device
+ *
+ * Provides a model for connecting in-kernel interrupt resources to a VCPU.
+ *
+ * A typical modern x86 processor has the concept of an internal Local-APIC
+ * and some external signal pins.  The way in which interrupts are injected is
+ * dependent on whether software enables the LAPIC or not.  When enabled,
+ * interrupts are acknowledged through the LAPIC.  Otherwise they are through
+ * an externally connected PIC (typically an i8259 on the BSP)
+ *
+ * Copyright (C) 2007 Novell
+ *
+ * Authors:
+ *   Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.  See
+ * the COPYING file in the top-level directory.
+ *
+ */
+
+#include "kvm.h"
+
+struct kvm_kernint {
+	struct kvm_vcpu              *vcpu;
+	struct kvm_irqdevice         *self_irq;
+	struct kvm_irqdevice         *ext_irq;
+	struct kvm_irqdevice          apic_irq;
+
+};
+
+static struct kvm_irqdevice *get_irq_dev(struct kvm_kernint *s)
+{
+	struct kvm_irqdevice *dev;
+
+	if (kvm_lapic_enabled(s->vcpu))
+		dev = &s->apic_irq;
+	else
+		dev = s->ext_irq;
+
+	if (!dev)
+		kvm_crash_guest(s->vcpu->kvm);
+
+	return dev;
+}
+
+static int kernint_irqdev_ack(struct kvm_irqdevice *this, int flags,
+			      struct kvm_irqack_data *data)
+{
+	struct kvm_kernint *s = (struct kvm_kernint*)this->private;
+
+	return kvm_irqdevice_ack(get_irq_dev(s), flags, data);
+}
+
+static int kernint_irqdev_set_pin(struct kvm_irqdevice *this,
+				  int irq, int level)
+{
+	/* no-op */
+	return 0;
+}
+
+static void kernint_irqdev_destructor(struct kvm_irqdevice *this)
+{
+	struct kvm_kernint *s = (struct kvm_kernint*)this->private;
+
+	kvm_irqdevice_destructor(&s->apic_irq);
+	kvm_lapic_destroy(s->vcpu);
+	kfree(s);
+}
+
+static void kvm_apic_intr(struct kvm_irqsink *this,
+			  struct kvm_irqdevice *dev,
+			  kvm_irqpin_t pin)
+{
+	struct kvm_kernint *s = (struct kvm_kernint*)this->private;
+
+	/*
+	 * If the LAPIC sent us an interrupt it *must* be enabled,
+	 * just forward it on to the CPU
+	 */
+	kvm_irqdevice_set_intr(s->self_irq, pin);
+}
+
+static void kvm_ext_intr(struct kvm_irqsink *this,
+			 struct kvm_irqdevice *dev,
+			 kvm_irqpin_t pin)
+{
+	struct kvm_kernint *s = (struct kvm_kernint*)this->private;
+
+	/*
+	 * If the EXTINT device sent us an interrupt, forward it to the LINT0
+	 * pin of the LAPIC
+	 */
+	if (pin != kvm_irqpin_localint)
+		return;
+
+	/*
+	 * "irq 0" = LINT0, 1 = LINT1
+	 */
+	kvm_irqdevice_set_pin(&s->apic_irq, 0, 1);
+}
+
+int kvm_kernint_init(struct kvm_vcpu *vcpu)
+{
+	struct kvm_irqdevice *irqdev = &vcpu->irq.dev;
+	struct kvm_kernint *s;
+	struct kvm_irqsink apicsink;
+
+	s = kzalloc(sizeof(*s), GFP_KERNEL);
+	if (!s)
+		return -ENOMEM;
+
+	s->vcpu = vcpu;
+
+	/*
+	 * Configure the irqdevice interface
+	 */
+	irqdev->ack         = kernint_irqdev_ack;
+	irqdev->set_pin     = kernint_irqdev_set_pin;
+	irqdev->destructor  = kernint_irqdev_destructor;
+
+	irqdev->private = s;
+	s->self_irq = irqdev;
+
+	/*
+	 * Configure the EXTINT device if this is the BSP processor
+	 */
+	if (!vcpu_slot(vcpu)) {
+		struct kvm_irqsink extsink = {
+			.set_intr   = kvm_ext_intr,
+			.private    = s
+		};
+		s->ext_irq = &vcpu->kvm->isa_irq;
+		kvm_irqdevice_register_sink(s->ext_irq, &extsink);
+	}
+
+	/*
+	 * Configure the LAPIC device
+	 */
+	apicsink.set_intr = kvm_apic_intr;
+	apicsink.private  = s;
+
+	kvm_irqdevice_init(&s->apic_irq);
+	kvm_irqdevice_register_sink(&s->apic_irq, &apicsink);
+	kvm_lapic_init(vcpu, &s->apic_irq, 0);
+
+	return 0;
+}
+
diff --git a/drivers/kvm/kvm.h b/drivers/kvm/kvm.h
index 1c46830..f112bd6 100644
--- a/drivers/kvm/kvm.h
+++ b/drivers/kvm/kvm.h
@@ -165,6 +165,21 @@ int kvm_user_irqdev_init(struct kvm_irqdevice *dev);
 int kvm_user_irqdev_save(struct kvm_irqdevice *this, void *data);
 int kvm_user_irqdev_restore(struct kvm_irqdevice *this, void *data);
 int kvm_userint_init(struct kvm_vcpu *vcpu);
+int kvm_kernint_init(struct kvm_vcpu *vcpu);
+
+#define KVM_LAPIC_OPTION_USERMODE (1 << 0)
+
+int kvm_lapic_init(struct kvm_vcpu *vcpu, struct kvm_irqdevice *dev,
+		   int flags);
+void kvm_lapic_destroy(struct kvm_vcpu *vcpu);
+void kvm_lapic_set_tpr(struct kvm_vcpu *vcpu, u64 cr8);
+u64  kvm_lapic_get_tpr(struct kvm_vcpu *vcpu);
+void kvm_lapic_set_base(struct kvm_vcpu *vcpu, u64 base);
+u64  kvm_lapic_get_base(struct kvm_vcpu *vcpu);
+void kvm_lapic_save(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs);
+void kvm_lapic_restore(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs);
+void kvm_lapic_reset(struct kvm_vcpu *vcpu);
+int  kvm_lapic_enabled(struct kvm_vcpu *vcpu);
 
 /*
  * x86 supports 3 paging modes (4-level 64-bit, 3-level 64-bit, and 2-level
@@ -337,6 +352,11 @@ struct kvm_vcpu_irq {
 	int                  usignal;
 };
 
+struct kvm_lapic {
+	void                 *dev;
+	struct kvm_io_device *mmio;
+};
+
 struct kvm_vcpu {
 	struct kvm *kvm;
 	union {
@@ -350,6 +370,7 @@ struct kvm_vcpu {
 	struct kvm_run *run;
 	int interrupt_window_open;
 	struct kvm_vcpu_irq irq;
+	struct kvm_lapic apic;
 	unsigned long regs[NR_VCPU_REGS]; /* for rsp: vcpu_load_rsp_rip() */
 	unsigned long rip;      /* needs vcpu_load_rsp_rip() */
 
@@ -360,10 +381,8 @@ struct kvm_vcpu {
 	struct page *para_state_page;
 	gpa_t hypercall_gpa;
 	unsigned long cr4;
-	unsigned long cr8;
 	u64 pdptrs[4]; /* pae */
 	u64 shadow_efer;
-	u64 apic_base;
 	u64 ia32_misc_enable_msr;
 	int nmsrs;
 	struct vmx_msr_entry *guest_msrs;
@@ -540,6 +559,8 @@ struct kvm {
 	struct list_head vm_list;
 	struct file *filp;
 	struct kvm_io_bus mmio_bus;
+	int enable_kernel_pic;
+	struct kvm_irqdevice isa_irq;
 };
 
 struct descriptor_table {
@@ -614,6 +635,9 @@ void kvm_exit_arch(void);
 int kvm_mmu_module_init(void);
 void kvm_mmu_module_exit(void);
 
+int kvm_apicbus_send(struct kvm *kvm, int dest, int trig_mode, int level,
+		     int dest_mode, int delivery_mode, int vector);
+
 void kvm_mmu_destroy(struct kvm_vcpu *vcpu);
 int kvm_mmu_create(struct kvm_vcpu *vcpu);
 int kvm_mmu_setup(struct kvm_vcpu *vcpu);
@@ -747,6 +771,13 @@ static inline struct kvm_mmu_page *page_header(hpa_t shadow_page)
 	return (struct kvm_mmu_page *)page_private(page);
 }
 
+static inline int vcpu_slot(struct kvm_vcpu *vcpu)
+{
+	return vcpu - vcpu->kvm->vcpus;
+}
+
+void kvm_crash_guest(struct kvm *kvm);
+
 static inline u16 read_fs(void)
 {
 	u16 seg;
diff --git a/drivers/kvm/kvm_main.c b/drivers/kvm/kvm_main.c
index 9a6d2c5..94d92aa 100644
--- a/drivers/kvm/kvm_main.c
+++ b/drivers/kvm/kvm_main.c
@@ -319,6 +319,7 @@ static struct kvm *kvm_create_vm(void)
 	spin_lock_init(&kvm->lock);
 	INIT_LIST_HEAD(&kvm->active_mmu_pages);
 	kvm_io_bus_init(&kvm->mmio_bus);
+	kvm_irqdevice_init(&kvm->isa_irq);
 	for (i = 0; i < KVM_MAX_VCPUS; ++i) {
 		struct kvm_vcpu *vcpu = &kvm->vcpus[i];
 
@@ -413,6 +414,23 @@ static void kvm_free_vcpus(struct kvm *kvm)
 		kvm_free_vcpu(&kvm->vcpus[i]);
 }
 
+/*
+ * The function kills a guest while there still is a user space processes
+ * with a descriptor to it
+ */
+void kvm_crash_guest(struct kvm *kvm)
+{
+	unsigned int i;
+
+	for (i = 0; i < KVM_MAX_VCPUS; ++i) {
+		/*
+		 * FIXME: in the future it should send IPI to gracefully
+		 * stop the other vCPUs
+		 */
+		kvm_free_vcpu(&kvm->vcpus[i]);
+	}
+}
+
 static int kvm_dev_release(struct inode *inode, struct file *filp)
 {
 	return 0;
@@ -424,6 +442,8 @@ static void kvm_destroy_vm(struct kvm *kvm)
 	list_del(&kvm->vm_list);
 	spin_unlock(&kvm_lock);
 	kvm_io_bus_destroy(&kvm->mmio_bus);
+	if (kvm->enable_kernel_pic)
+		kvm_irqdevice_destructor(&kvm->isa_irq);
 	kvm_free_vcpus(kvm);
 	kvm_free_physmem(kvm);
 	kfree(kvm);
@@ -629,7 +649,7 @@ void set_cr8(struct kvm_vcpu *vcpu, unsigned long cr8)
 		inject_gp(vcpu);
 		return;
 	}
-	vcpu->cr8 = cr8;
+	kvm_lapic_set_tpr(vcpu, cr8);
 }
 EXPORT_SYMBOL_GPL(set_cr8);
 
@@ -930,6 +950,69 @@ out:
 	return r;
 }
 
+static int kvm_vm_ioctl_enable_kernel_pic(struct kvm *kvm, __u32 val)
+{
+	/*
+	 * FIXME: We should not allow this if VCPUs have already been created
+	 */
+	if (kvm->enable_kernel_pic)
+		return -EINVAL;
+
+	/*
+	 * Someday we may offer two levels of in-kernel PIC support:
+	 *
+	 *  level 0 = (default) compatiblity mode (everything in userspace)
+	 *  level 1 = LAPIC in kernel, IOAPIC/i8259 in userspace
+	 *  level 2 = All three in kernel
+	 *
+	 * For now we only support level 0 and 1.  However, you cant set
+	 * level 0
+	 */
+	if (val != 1)
+		return -EINVAL;
+
+	kvm->enable_kernel_pic = val;
+
+	printk(KERN_INFO "KVM: Setting in-kernel PIC level to %d\n", val);
+
+	/*
+	 * installing a user_irqdev model to the kvm->isa_irq device
+	 * creates a level-1 environment, where the userspace completely
+	 * controls the ISA domain interrupts in the IOAPIC/i8259.
+	 * Interrupts come down to the VCPU either as an ISA vector to
+	 * this controller, or as an APIC bus message (or both)
+	 */
+	kvm_user_irqdev_init(&kvm->isa_irq);
+
+	return 0;
+}
+
+static int kvm_vm_ioctl_isa_interrupt(struct kvm *kvm,
+				      struct kvm_interrupt *irq)
+{
+	if (irq->irq < 0 || irq->irq >= 256)
+		return -EINVAL;
+
+	if (!kvm->enable_kernel_pic)
+		return -EINVAL;
+
+	return kvm_irqdevice_set_pin(&kvm->isa_irq, irq->irq, 1);
+}
+
+static int kvm_vm_ioctl_apic_msg(struct kvm *kvm,
+				 struct kvm_apic_msg *msg)
+{
+	if (!kvm->enable_kernel_pic)
+		return -EINVAL;
+
+	msg->delivery_mode = (msg->delivery_mode << 8) & 0xF00;
+
+	kvm_apicbus_send(kvm, msg->dest, msg->trig_mode, 1, msg->dest_mode,
+			 msg->delivery_mode, msg->vector);
+
+	return 0;
+}
+
 static gfn_t unalias_gfn(struct kvm *kvm, gfn_t gfn)
 {
 	int i;
@@ -1050,10 +1133,16 @@ static int emulator_write_std(unsigned long addr,
 static struct kvm_io_device *vcpu_find_mmio_dev(struct kvm_vcpu *vcpu,
 						gpa_t addr)
 {
+	struct kvm_io_device *dev = vcpu->apic.mmio;
+
+	/*
+	 * First check if the LAPIC will snarf this request
+	 */
+	if (dev && dev->in_range(dev, addr))
+		return dev;
+
 	/*
-	 * Note that its important to have this wrapper function because
-	 * in the very near future we will be checking for MMIOs against
-	 * the LAPIC as well as the general MMIO bus
+	 * And then fallback to allow any device to participate
 	 */
 	return kvm_io_bus_find_dev(&vcpu->kvm->mmio_bus, addr);
 }
@@ -1520,7 +1609,7 @@ int kvm_get_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 *pdata)
 		data = 3;
 		break;
 	case MSR_IA32_APICBASE:
-		data = vcpu->apic_base;
+		data = kvm_lapic_get_base(vcpu);
 		break;
 	case MSR_IA32_MISC_ENABLE:
 		data = vcpu->ia32_misc_enable_msr;
@@ -1598,7 +1687,7 @@ int kvm_set_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 data)
 	case 0x200 ... 0x2ff: /* MTRRs */
 		break;
 	case MSR_IA32_APICBASE:
-		vcpu->apic_base = data;
+		kvm_lapic_set_base(vcpu, data);
 		break;
 	case MSR_IA32_MISC_ENABLE:
 		vcpu->ia32_misc_enable_msr = data;
@@ -1862,8 +1951,9 @@ static int kvm_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
 
-	/* re-sync apic's tpr */
-	vcpu->cr8 = kvm_run->cr8;
+	if (!vcpu->kvm->enable_kernel_pic)
+		/* re-sync apic's tpr if the APIC is in userspace */
+		kvm_lapic_set_tpr(vcpu, kvm_run->cr8);
 
 	if (vcpu->pio.cur_count) {
 		r = complete_pio(vcpu);
@@ -2015,11 +2105,12 @@ static int kvm_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 	sregs->cr2 = vcpu->cr2;
 	sregs->cr3 = vcpu->cr3;
 	sregs->cr4 = vcpu->cr4;
-	sregs->cr8 = vcpu->cr8;
 	sregs->efer = vcpu->shadow_efer;
-	sregs->apic_base = vcpu->apic_base;
 
-	kvm_user_irqdev_save(&vcpu->irq.dev, &sregs->interrupt_bitmap);
+	kvm_lapic_save(vcpu, sregs);
+
+	if (!vcpu->kvm->enable_kernel_pic)
+		kvm_user_irqdev_save(&vcpu->irq.dev, &sregs->interrupt_bitmap);
 
 	vcpu_put(vcpu);
 
@@ -2051,14 +2142,10 @@ static int kvm_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 	mmu_reset_needed |= vcpu->cr3 != sregs->cr3;
 	vcpu->cr3 = sregs->cr3;
 
-	vcpu->cr8 = sregs->cr8;
-
 	mmu_reset_needed |= vcpu->shadow_efer != sregs->efer;
 #ifdef CONFIG_X86_64
 	kvm_arch_ops->set_efer(vcpu, sregs->efer);
 #endif
-	vcpu->apic_base = sregs->apic_base;
-
 	kvm_arch_ops->decache_cr4_guest_bits(vcpu);
 
 	mmu_reset_needed |= vcpu->cr0 != sregs->cr0;
@@ -2072,8 +2159,11 @@ static int kvm_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 	if (mmu_reset_needed)
 		kvm_mmu_reset_context(vcpu);
 
-	kvm_user_irqdev_restore(&vcpu->irq.dev,
-				&sregs->interrupt_bitmap[0]);
+	kvm_lapic_restore(vcpu, sregs);
+
+	if (!vcpu->kvm->enable_kernel_pic)
+		kvm_user_irqdev_restore(&vcpu->irq.dev,
+					&sregs->interrupt_bitmap[0]);
 
 	set_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
 	set_segment(vcpu, &sregs->ds, VCPU_SREG_DS);
@@ -2545,7 +2635,12 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, int n)
 
 	kvm_irqdevice_init(&vcpu->irq.dev);
 	kvm_vcpu_irqsink_init(vcpu);
-	r = kvm_userint_init(vcpu);
+
+	if (kvm->enable_kernel_pic)
+		r = kvm_kernint_init(vcpu);
+	else
+		r = kvm_userint_init(vcpu);
+
 	if (r < 0)
 		goto out_free_vcpus;
 
@@ -2667,6 +2762,12 @@ static int kvm_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 	return 0;
 }
 
+static int kvm_vcpu_ioctl_apic_reset(struct kvm_vcpu *vcpu)
+{
+	kvm_lapic_reset(vcpu);
+	return 0;
+}
+
 static long kvm_vcpu_ioctl(struct file *filp,
 			   unsigned int ioctl, unsigned long arg)
 {
@@ -2836,6 +2937,13 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = 0;
 		break;
 	}
+	case KVM_APIC_RESET: {
+		r = kvm_vcpu_ioctl_apic_reset(vcpu);
+		if (r)
+			goto out;
+		r = 0;
+		break;
+	}
 	default:
 		;
 	}
@@ -2889,6 +2997,41 @@ static long kvm_vm_ioctl(struct file *filp,
 			goto out;
 		break;
 	}
+	case KVM_ENABLE_KERNEL_PIC: {
+		__u32 val;
+
+		r = -EFAULT;
+		if (copy_from_user(&val, argp, sizeof val))
+			goto out;
+		r = kvm_vm_ioctl_enable_kernel_pic(kvm, val);
+		if (r)
+			goto out;
+		break;
+	}
+	case KVM_ISA_INTERRUPT: {
+		struct kvm_interrupt irq;
+
+		r = -EFAULT;
+		if (copy_from_user(&irq, argp, sizeof irq))
+			goto out;
+		r = kvm_vm_ioctl_isa_interrupt(kvm, &irq);
+		if (r)
+			goto out;
+		r = 0;
+		break;
+	}
+	case KVM_APIC_MSG: {
+		struct kvm_apic_msg msg;
+
+		r = -EFAULT;
+		if (copy_from_user(&msg, argp, sizeof msg))
+			goto out;
+		r = kvm_vm_ioctl_apic_msg(kvm, &msg);
+		if (r)
+			goto out;
+		r = 0;
+		break;
+	}
 	default:
 		;
 	}
diff --git a/drivers/kvm/lapic.c b/drivers/kvm/lapic.c
new file mode 100644
index 0000000..e7e7bd4
--- /dev/null
+++ b/drivers/kvm/lapic.c
@@ -0,0 +1,1421 @@
+/*
+ * Local APIC virtualization
+ *
+ * Copyright (C) 2006 Qumranet, Inc.
+ * Copyright (C) 2007 Novell
+ *
+ * Authors:
+ *   Dor Laor <dor.laor-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
+ *   Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
+ *
+ * Based on Xen 3.0 code, Copyright (c) 2004, Intel Corporation.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.  See
+ * the COPYING file in the top-level directory.
+ */
+
+#include "kvm.h"
+#include <linux/kvm.h>
+#include <linux/mm.h>
+#include <linux/highmem.h>
+#include <linux/smp.h>
+#include <linux/hrtimer.h>
+#include <asm/processor.h>
+#include <asm/io.h>
+#include <asm/msr.h>
+#include <asm/page.h>
+#include <asm/current.h>
+
+/*XXX remove this definition after GFW enabled */
+#define APIC_NO_BIOS
+
+#define PRId64 "d"
+#define PRIx64 "llx"
+#define PRIu64 "u"
+#define PRIo64 "o"
+
+#define APIC_BUS_CYCLE_NS 1
+
+/* #define apic_debug(fmt,arg...) printk(KERN_WARNING fmt,##arg) */
+#define apic_debug(fmt,arg...)
+
+struct kvm_kern_apic {
+	spinlock_t              lock;
+	atomic_t                ref_count;
+	int                     usermode;
+	u32                     status;
+	u32                     vcpu_id;
+	u64                     base_msr;
+	unsigned long           base_address;
+	struct kvm_io_device    mmio_dev;
+	struct {
+		unsigned long   pending;
+		u32             divide_count;
+		ktime_t         last_update;
+		struct hrtimer  dev;
+	} timer;
+	u32                     err_status;
+	u32                     err_write_count;
+	struct kvm_vcpu         *vcpu;
+	struct kvm_irqdevice    *irq_dev;
+	struct page             *regs_page;
+	void                    *regs;
+};
+
+static __inline__ int find_highest_bit(unsigned long *data, int nr_bits)
+{
+	int length = BITS_TO_LONGS(nr_bits);
+	while (length && !data[--length])
+		continue;
+	return __ffs(data[length]) + (length * BITS_PER_LONG);
+}
+
+#define APIC_LVT_NUM			6
+/* 14 is the version for Xeon and Pentium 8.4.8*/
+#define APIC_VERSION			(0x14UL | ((APIC_LVT_NUM - 1) << 16))
+#define VLOCAL_APIC_MEM_LENGTH		(1 << 12)
+/* followed define is not in apicdef.h */
+#define APIC_SHORT_MASK			0xc0000
+#define APIC_DEST_NOSHORT		0x0
+#define APIC_DEST_MASK			0x800
+#define _APIC_GLOB_DISABLE		0x0
+#define APIC_GLOB_DISABLE_MASK		0x1
+#define APIC_SOFTWARE_DISABLE_MASK	0x2
+#define _APIC_BSP_ACCEPT_PIC		0x3
+#define MAX_APIC_INT_VECTOR             256
+
+#define inject_gp(vcpu) kvm_arch_ops->inject_gp(vcpu, 0);
+
+#define apic_enabled(apic)              \
+	(!((apic)->status &                   \
+	   (APIC_GLOB_DISABLE_MASK | APIC_SOFTWARE_DISABLE_MASK)))
+
+#define apic_global_enabled(apic)       \
+	(!(test_bit(_APIC_GLOB_DISABLE, &(apic)->status)))
+
+#define LVT_MASK \
+	APIC_LVT_MASKED | APIC_SEND_PENDING | APIC_VECTOR_MASK
+
+#define LINT_MASK   \
+	LVT_MASK | APIC_MODE_MASK | APIC_INPUT_POLARITY |\
+	APIC_LVT_REMOTE_IRR | APIC_LVT_LEVEL_TRIGGER
+
+#define KVM_APIC_ID(apic)   \
+	(GET_APIC_ID(apic_get_reg(apic, APIC_ID)))
+
+#define apic_lvt_enabled(apic, lvt_type)    \
+	(!(apic_get_reg(apic, lvt_type) & APIC_LVT_MASKED))
+
+#define apic_lvt_vector(apic, lvt_type)     \
+	(apic_get_reg(apic, lvt_type) & APIC_VECTOR_MASK)
+
+#define apic_lvt_dm(apic, lvt_type)           \
+	(apic_get_reg(apic, lvt_type) & APIC_MODE_MASK)
+
+#define apic_lvtt_period(apic)     \
+	(apic_get_reg(apic, APIC_LVTT) & APIC_LVT_TIMER_PERIODIC)
+
+static inline u32 apic_get_reg(struct kvm_kern_apic *apic, u32 reg)
+{
+	return *((u32 *)(apic->regs + reg));
+}
+
+static inline void apic_set_reg(struct kvm_kern_apic *apic,
+				u32 reg, u32 val)
+{
+	*((u32 *)(apic->regs + reg)) = val;
+}
+
+static unsigned int apic_lvt_mask[APIC_LVT_NUM] =
+{
+	LVT_MASK | APIC_LVT_TIMER_PERIODIC, 	/* LVTT */
+	LVT_MASK | APIC_MODE_MASK, 		/* LVTTHMR */
+	LVT_MASK | APIC_MODE_MASK, 		/* LVTPC */
+	LINT_MASK, LINT_MASK, 			/* LVT0-1 */
+	LVT_MASK 				/* LVTERR */
+};
+
+#define ASSERT(x)  							     \
+	if (!(x)) { 							     \
+		printk(KERN_EMERG "assertion failed %s: %d: %s\n",           \
+		       __FILE__, __LINE__, #x);                              \
+		BUG();                                                       \
+	}
+
+static int apic_find_highest_irr(struct kvm_kern_apic *apic)
+{
+	int result;
+
+	result = find_highest_bit((unsigned long *)(apic->regs + APIC_IRR),
+				  MAX_APIC_INT_VECTOR);
+
+	ASSERT( result == 0 || result >= 16);
+
+	return result;
+}
+
+
+static int apic_find_highest_isr(struct kvm_kern_apic *apic)
+{
+	int result;
+
+	result = find_highest_bit((unsigned long *)(apic->regs + APIC_ISR),
+				  MAX_APIC_INT_VECTOR);
+
+	ASSERT( result == 0 || result >= 16);
+
+	return result;
+}
+
+static void apic_dropref(struct kvm_kern_apic *apic)
+{
+	if (atomic_dec_and_test(&apic->ref_count)) {
+
+		spin_lock_bh(&apic->lock);
+
+		hrtimer_cancel(&apic->timer.dev);
+
+		if (apic->regs_page) {
+			__free_page(apic->regs_page);
+			apic->regs_page = 0;
+		}
+
+		spin_unlock_bh(&apic->lock);
+
+		kfree(apic);
+	}
+}
+
+#if 0
+static void apic_dump_state(struct kvm_kern_apic *apic)
+{
+	u64 *tmp;
+
+	printk(KERN_INFO "%s begin\n", __FUNCTION__);
+
+	printk(KERN_INFO "status = 0x%08x\n", apic->status);
+	printk(KERN_INFO "base_msr=0x%016llx, apicbase = 0x%08lx\n",
+	       apic->base_msr, apic->base_address);
+
+	tmp = (u64*)(apic->regs + APIC_IRR);
+	printk(KERN_INFO "IRR = 0x%016llx 0x%016llx 0x%016llx 0x%016llx\n",
+	       tmp[3], tmp[2], tmp[1], tmp[0]);
+	tmp = (u64*)(apic->regs + APIC_ISR);
+	printk(KERN_INFO "ISR = 0x%016llx 0x%016llx 0x%016llx 0x%016llx\n",
+	       tmp[3], tmp[2], tmp[1], tmp[0]);
+	tmp = (u64*)(apic->regs + APIC_TMR);
+	printk(KERN_INFO "TMR = 0x%016llx 0x%016llx 0x%016llx 0x%016llx\n",
+	       tmp[3], tmp[2], tmp[1], tmp[0]);
+
+	printk(KERN_INFO "APIC_ID=0x%08x\n", apic_get_reg(apic, APIC_ID));
+	printk(KERN_INFO "APIC_TASKPRI=0x%08x\n",
+	       apic_get_reg(apic, APIC_TASKPRI) & 0xff);
+	printk(KERN_INFO "APIC_PROCPRI=0x%08x\n",
+	       apic_get_reg(apic, APIC_PROCPRI));
+
+	printk(KERN_INFO "APIC_DFR=0x%08x\n",
+	       apic_get_reg(apic, APIC_DFR) | 0x0FFFFFFF);
+	printk(KERN_INFO "APIC_LDR=0x%08x\n",
+	       apic_get_reg(apic, APIC_LDR) & APIC_LDR_MASK);
+	printk(KERN_INFO "APIC_SPIV=0x%08x\n",
+	       apic_get_reg(apic, APIC_SPIV) & 0x3ff);
+	printk(KERN_INFO "APIC_ESR=0x%08x\n",
+	       apic_get_reg(apic, APIC_ESR));
+	printk(KERN_INFO "APIC_ICR=0x%08x\n",
+	       apic_get_reg(apic, APIC_ICR) & ~(1 << 12));
+	printk(KERN_INFO "APIC_ICR2=0x%08x\n",
+	       apic_get_reg(apic, APIC_ICR2) & 0xff000000);
+
+	printk(KERN_INFO "APIC_LVTERR=0x%08x\n",
+	       apic_get_reg(apic, APIC_LVTERR));
+	printk(KERN_INFO "APIC_LVT1=0x%08x\n",
+	       apic_get_reg(apic, APIC_LVT1));
+	printk(KERN_INFO "APIC_LVT0=0x%08x\n",
+	       apic_get_reg(apic, APIC_LVT0));
+	printk(KERN_INFO "APIC_LVTPC=0x%08x\n",
+	       apic_get_reg(apic, APIC_LVTPC));
+	printk(KERN_INFO "APIC_LVTTHMR=0x%08x\n",
+	       apic_get_reg(apic, APIC_LVTTHMR));
+	printk(KERN_INFO "APIC_LVTT=0x%08x\n",
+	       apic_get_reg(apic, APIC_LVTT));
+
+	printk(KERN_INFO "APIC_TMICT=0x%08x\n",
+	       apic_get_reg(apic, APIC_TMICT));
+	printk(KERN_INFO "APIC_TDCR=0x%08x\n",
+	       apic_get_reg(apic, APIC_TDCR));
+
+	printk(KERN_INFO "%s end\n", __FUNCTION__);
+}
+#endif
+
+
+static int apic_update_ppr(struct kvm_kern_apic *apic)
+{
+	u32 tpr, isrv, ppr, orig_ppr;
+	int irq;
+	int masked = 0;
+	int forward = 0;
+
+	ppr = apic_get_reg(apic, APIC_PROCPRI);
+	orig_ppr = ppr;
+
+	/*
+	 * Before we change anything, see if the only pending vectors we have
+	 * are anything masked by PPR
+	 */
+	irq = apic_find_highest_irr(apic);
+	if (irq && ((irq & 0xf0) <= ppr))
+		masked = true;
+
+	/*
+	 * Compute the PPR value based on the current settings of TPR/ISR
+	 */
+	tpr = apic_get_reg(apic, APIC_TASKPRI);
+	irq = apic_find_highest_isr(apic);
+	isrv = (irq >> 4) & 0xf;
+
+	if ((tpr >> 4) >= isrv)
+		ppr = tpr & 0xff;
+	else
+		ppr = isrv << 4;  /* low 4 bits of PPR have to be cleared */
+
+	apic_set_reg(apic, APIC_PROCPRI, ppr);
+
+	if (masked) {
+		/*
+		 * If we get here its because there were vectors that
+		 * were masked by PPR.  Check again to see if anything is
+		 * now available
+		 */
+		irq = apic_find_highest_irr(apic);
+		if ((irq & 0xf0) > ppr)
+			forward = 1;
+	}
+
+	apic_debug("%s: ppr 0x%x (old) 0x%x (new), isr 0x%x, isrv 0x%x\n",
+	       __FUNCTION__, orig_ppr, ppr, irq, isrv);
+
+	return forward;
+}
+
+static void apic_set_tpr(struct kvm_kern_apic *apic, u32 tpr)
+{
+	int forward = 0;
+
+	apic_debug("new value = %x\n", tpr);
+
+	apic_set_reg(apic, APIC_TASKPRI, tpr);
+	forward = apic_update_ppr(apic);
+
+	if (forward) {
+		spin_unlock_bh(&apic->lock);
+		kvm_irqdevice_set_intr(apic->irq_dev, kvm_irqpin_localint);
+		spin_lock_bh(&apic->lock);
+	}
+}
+
+static int apic_match_dest(struct kvm_kern_apic *target,
+			   int dest,
+			   int dest_mode,
+			   int delivery_mode)
+{
+	int result = 0;
+
+	spin_lock_bh(&target->lock);
+
+	if (!dest_mode) /* Physical */
+		result = (GET_APIC_ID(apic_get_reg(target, APIC_ID)) == dest);
+	else { /* Logical */
+		u32 ldr = apic_get_reg(target, APIC_LDR);
+
+		/* Flat mode */
+		if (apic_get_reg(target, APIC_DFR) == APIC_DFR_FLAT)
+			result = GET_APIC_LOGICAL_ID(ldr) & dest;
+		else {
+			if ((delivery_mode == APIC_DM_LOWEST) &&
+			    (dest == 0xff)) {
+				printk(KERN_ALERT "Broadcast IPI " \
+				       "with lowest priority "
+				       "delivery mode\n");
+				spin_unlock_bh(&target->lock);
+				kvm_crash_guest(target->vcpu->kvm);
+				return 0;
+			}
+			if (GET_APIC_LOGICAL_ID(ldr) == (dest & 0xf))
+				result = (GET_APIC_LOGICAL_ID(ldr) >> 4) &
+					(dest >> 4);
+			else
+				result = 0;
+		}
+	}
+
+	spin_unlock_bh(&target->lock);
+
+	return result;
+}
+
+/*
+ * Add a pending IRQ into lapic.
+ * Return 1 if successfully added and 0 if discarded.
+ */
+static int __apic_accept_irq(struct kvm_kern_apic *apic,
+			     int delivery_mode,
+			     int vector,
+			     int level,
+			     int trig_mode)
+{
+	kvm_irqpin_t pin = kvm_irqpin_invalid;
+
+	switch (delivery_mode) {
+	case APIC_DM_FIXED:
+	case APIC_DM_LOWEST:
+		if (unlikely(!apic_enabled(apic)))
+			break;
+
+		if (test_and_set_bit(vector, apic->regs + APIC_IRR)
+		    && trig_mode) {
+			apic_debug("level trig mode repeatedly for vector " \
+				   "%d\n", vector);
+			break;
+		}
+
+		if (trig_mode) {
+			apic_debug("level trig mode for vector %d\n", vector);
+			set_bit(vector, apic->regs + APIC_TMR);
+		}
+
+		apic_debug("FIXED/LOWEST interrupt for vector %d\n", vector);
+		pin = kvm_irqpin_localint;
+		break;
+	case APIC_DM_REMRD:
+		printk(KERN_WARNING "%s: Ignore deliver mode %d\n",
+		       __FUNCTION__, delivery_mode);
+		break;
+	case APIC_DM_EXTINT:
+		apic_debug("EXTINT interrupt\n");
+		pin  = kvm_irqpin_extint;
+		break;
+	case APIC_DM_SMI:
+		apic_debug("SMI interrupt\n");
+		pin = kvm_irqpin_smi;
+		break;
+	case APIC_DM_NMI:
+		apic_debug("NMI interrupt\n");
+		pin = kvm_irqpin_nmi;
+		break;
+	case APIC_DM_INIT:
+		apic_debug("INIT interrupt\n");
+		if (level) {
+			spin_unlock_bh(&apic->lock);
+			kvm_lapic_reset(apic->vcpu);
+			spin_lock_bh(&apic->lock);
+		}
+		break;
+	case APIC_DM_STARTUP: /* FIXME: currently no support for SMP */
+	default:
+		printk(KERN_ALERT "TODO: support interrupt type %x\n",
+		       delivery_mode);
+		spin_unlock_bh(&apic->lock);
+		kvm_crash_guest(apic->vcpu->kvm);
+		spin_lock_bh(&apic->lock);
+		break;
+	}
+
+	if (likely(pin != kvm_irqpin_invalid)) {
+		/*
+		 * temp release of the lock to transmit
+		 */
+		spin_unlock_bh(&apic->lock);
+		kvm_irqdevice_set_intr(apic->irq_dev, pin);
+		spin_lock_bh(&apic->lock);
+
+		return 1;
+	} else
+		return 0;
+}
+
+static int apic_accept_irq(struct kvm_kern_apic *apic,
+			   int delivery_mode,
+			   int vector,
+			   int level,
+			   int trig_mode)
+{
+	int ret;
+
+	spin_lock_bh(&apic->lock);
+	ret = __apic_accept_irq(apic, delivery_mode, vector,
+				level, trig_mode);
+	spin_unlock_bh(&apic->lock);
+
+	return ret;
+}
+
+static void apic_set_eoi(struct kvm_kern_apic *apic)
+{
+	int vector = apic_find_highest_isr(apic);
+	int forward;
+
+	/*
+	 * Not every write EOI will has corresponding ISR,
+	 * one example is when Kernel check timer on setup_IO_APIC
+	 */
+	if (!vector)
+		return;
+
+	__clear_bit(vector, apic->regs + APIC_ISR);
+	forward = apic_update_ppr(apic);
+
+	__clear_bit(vector, apic->regs + APIC_TMR);
+
+	if (forward) {
+		spin_unlock_bh(&apic->lock);
+		kvm_irqdevice_set_intr(apic->irq_dev, kvm_irqpin_localint);
+		spin_lock_bh(&apic->lock);
+	}
+}
+
+static int apic_check_vector(struct kvm_kern_apic *apic,u32 dm, u32 vector)
+{
+	if ((dm == APIC_DM_FIXED) && (vector < 16)) {
+		apic->err_status |= 0x40;
+		__apic_accept_irq(apic, APIC_DM_FIXED,
+				  apic_lvt_vector(apic, APIC_LVTERR), 0, 0);
+		apic_debug("%s: check failed "
+		       " dm %x vector %x\n", __FUNCTION__, dm, vector);
+		return 0;
+	}
+	return 1;
+}
+
+int kvm_apicbus_send(struct kvm *kvm, int dest, int trig_mode, int level,
+		     int dest_mode, int delivery_mode, int vector)
+{
+	int i;
+	u32 lpr_map = 0;
+
+	apic_debug("%s: %d %d %d %d %d %d\n", __FUNCTION__,
+		   dest, trig_mode, level, dest_mode, delivery_mode, vector);
+
+	for (i = 0; i < KVM_MAX_VCPUS; ++i) {
+		struct kvm_kern_apic *target;
+		target = kvm->vcpus[i].apic.dev;
+
+		if (!target)
+			continue;
+
+		if (apic_match_dest(target, dest, dest_mode, delivery_mode)) {
+			if (delivery_mode == APIC_DM_LOWEST)
+				__set_bit(target->vcpu_id, &lpr_map);
+			else
+				apic_accept_irq(target, delivery_mode,
+						vector, level, trig_mode);
+		}
+	}
+
+	if (delivery_mode == APIC_DM_LOWEST) {
+		struct kvm_kern_apic *target;
+
+		/* Currently only UP is supported */
+		target = kvm->vcpus[0].apic.dev;
+
+		if (target)
+			apic_accept_irq(target, delivery_mode,
+					vector, level, trig_mode);
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(kvm_apicbus_send);
+
+static void apic_send_ipi(struct kvm_kern_apic *apic)
+{
+	u32 icr_low = apic_get_reg(apic, APIC_ICR);
+	u32 icr_high = apic_get_reg(apic, APIC_ICR2);
+
+	unsigned int dest =          GET_APIC_DEST_FIELD(icr_high);
+	unsigned int short_hand =    icr_low & APIC_SHORT_MASK;
+	unsigned int trig_mode =     icr_low & APIC_INT_LEVELTRIG;
+	unsigned int level =         icr_low & APIC_INT_ASSERT;
+	unsigned int dest_mode =     icr_low & APIC_DEST_MASK;
+	unsigned int delivery_mode = icr_low & APIC_MODE_MASK;
+	unsigned int vector =        icr_low & APIC_VECTOR_MASK;
+
+	apic_debug("icr_high 0x%x, icr_low 0x%x, "
+		 "short_hand 0x%x, dest 0x%x, trig_mode 0x%x, level 0x%x, "
+		 "dest_mode 0x%x, delivery_mode 0x%x, vector 0x%x\n",
+		 icr_high, icr_low, short_hand, dest,
+		 trig_mode, level, dest_mode, delivery_mode, vector);
+
+	/*
+	 * We unlock here because we would enter this function in a lock
+	 * state and we dont want to remain this way while we transmit
+	 */
+	spin_unlock_bh(&apic->lock);
+
+	if (short_hand == APIC_DEST_NOSHORT)
+		/*
+		 * If no short-hand notation is in use, just forward the
+		 * message onto the apicbus and let the bus handle the routing.
+		 */
+		kvm_apicbus_send(apic->vcpu->kvm, dest, trig_mode, level,
+				 dest_mode, delivery_mode, vector);
+	else {
+		/*
+		 * Otherwise we need to consider the short-hand to find the
+		 * correct targets.
+		 */
+		unsigned int i;
+
+		for (i = 0; i < KVM_MAX_VCPUS; ++i) {
+			struct kvm_kern_apic *target;
+			int result = 0;
+
+			target = apic->vcpu->kvm->vcpus[i].apic.dev;
+
+			if (!target)
+				continue;
+
+			switch (short_hand) {
+			case APIC_DEST_SELF:
+				if (target == apic)
+					result = 1;
+				break;
+			case APIC_DEST_ALLINC:
+				result = 1;
+				break;
+
+			case APIC_DEST_ALLBUT:
+				if (target != apic)
+					result = 1;
+				break;
+			}
+
+			if (result)
+				apic_accept_irq(target, delivery_mode,
+						vector, level, trig_mode);
+		}
+	}
+
+	/*
+	 * Relock before returning
+	 */
+	spin_lock_bh(&apic->lock);
+
+}
+
+static u32 apic_get_tmcct(struct kvm_kern_apic *apic)
+{
+	u32 counter_passed;
+	ktime_t passed, now = apic->timer.dev.base->get_time();
+	u32 tmcct = apic_get_reg(apic, APIC_TMCCT);
+
+	ASSERT(apic != NULL);
+
+	if (unlikely(ktime_to_ns(now) <=
+		     ktime_to_ns(apic->timer.last_update))) {
+		/* Wrap around */
+		passed = ktime_add(
+			({ (ktime_t){
+				.tv64 = KTIME_MAX -
+					 (apic->timer.last_update).tv64 };
+			}), now);
+		apic_debug("time elapsed\n");
+	} else
+		passed = ktime_sub(now, apic->timer.last_update);
+
+	counter_passed = ktime_to_ns(passed) /
+		(APIC_BUS_CYCLE_NS * apic->timer.divide_count);
+	tmcct -= counter_passed;
+
+	if (tmcct <= 0) {
+		if (unlikely(!apic_lvtt_period(apic))) {
+			tmcct =  0;
+		} else {
+			do {
+				tmcct += apic_get_reg(apic, APIC_TMICT);
+			} while ( tmcct <= 0 );
+		}
+	}
+
+	apic->timer.last_update = now;
+	apic_set_reg(apic, APIC_TMCCT, tmcct);
+
+	return tmcct;
+}
+
+/*
+ *----------------------------------------------------------------------
+ * MMIO
+ *----------------------------------------------------------------------
+ */
+
+#define align(val, len) (val & ~(len-1))
+
+static int validate_mmio(struct kvm_kern_apic *apic, gpa_t address, int len)
+{
+	/*
+	 * According to IA 32 Manual, all registers should be accessed with
+	 * 32 bits alignment.
+	 */
+	if (align(address, 4) != align(address+(len-1), 4)) {
+		printk(KERN_WARNING "KVM: MMIO request for %d bytes at " \
+		       "0x%lx is not 32 bit aligned.  Injecting #GP\n",
+		       len, address);
+		inject_gp(apic->vcpu);
+		return 0;
+	}
+
+	return 1;
+}
+
+static u32 __apic_read(struct kvm_kern_apic *apic,
+				unsigned int offset)
+{
+	u32 val = 0;
+
+	if (offset > APIC_TDCR)
+		return 0;
+
+	switch (offset) {
+	case APIC_ARBPRI:
+		printk(KERN_WARNING "access local APIC ARBPRI register " \
+		       "which is for P6\n");
+		break;
+
+	case APIC_TMCCT:        /* Timer CCR */
+		val = apic_get_tmcct(apic);
+		break;
+
+	case APIC_ESR:
+		apic->err_write_count = 0;
+		/* fall through */
+	default:
+		val = apic_get_reg(apic, offset);
+		break;
+	}
+
+	return val;
+}
+
+static void apic_mmio_read(struct kvm_io_device *this,
+			   gpa_t address,
+			   int len,
+			   void *data)
+{
+	struct kvm_kern_apic *apic = (struct kvm_kern_apic*)this->private;
+	unsigned int          offset = address - apic->base_address;
+	unsigned char         alignment = offset & 0x3;
+	u32                   val;
+
+	if (!validate_mmio(apic, address, len))
+		return;
+
+	spin_lock_bh(&apic->lock);
+	val = __apic_read(apic, offset & ~0x3);
+	spin_unlock_bh(&apic->lock);
+
+	switch (len) {
+	case 1:
+	case 2:
+	case 4:
+		memcpy(data, (char*)((char*)&val + alignment), len);
+		break;
+	default:
+		printk(KERN_ALERT "Local APIC read with len = %x, " \
+		       "should be 1,2, or 4 instead\n", len);
+		inject_gp(apic->vcpu);
+		break;
+	}
+}
+
+static void apic_mmio_write(struct kvm_io_device *this,
+			    gpa_t address,
+			    int len,
+			    const void *data)
+{
+	struct kvm_kern_apic *apic = (struct kvm_kern_apic*)this->private;
+	unsigned int          offset = address - apic->base_address;
+	unsigned char         alignment = offset & 0x3;
+	u32                   val;
+
+	if (!validate_mmio(apic, address, len))
+		return;
+
+	spin_lock_bh(&apic->lock);
+
+	switch (len) {
+	case 1:
+	case 2: {
+		unsigned int tmp;
+
+		/*
+		 * Some kernels will access with byte/word alignment
+		 */
+		apic_debug("Notice: Local APIC write with len = %x\n", len);
+		tmp = __apic_read(apic, offset & ~0x3);
+		switch (len) {
+		case 1:
+			val = *(u8*)data;
+
+			val = (tmp & ~(0xff << (8*alignment))) |
+			      ((val & 0xff) << (8*alignment));
+			break;
+
+		case 2:
+			if (alignment != 0x0 && alignment != 0x2) {
+				printk(KERN_ALERT "alignment error for apic " \
+				       "with len == 2\n");
+				inject_gp(apic->vcpu);
+			}
+
+			/*
+			 * assumes 16 bit alignment on the pointer.
+			 * Mis-alignment is a host-side issue, however, so
+			 * we crash
+			 */
+			BUG_ON(((long)data & 0x1));
+
+			val = *(u16*)data;
+
+			val = (tmp & ~(0xffff << (8*alignment))) |
+			      ((val & 0xffff) << (8*alignment));
+			break;
+		}
+
+		break;
+	}
+	case 4:
+		memcpy(&val, data, 4);
+		break;
+	default:
+		printk(KERN_ALERT "Local APIC write with len = %x, " \
+		       "should be 1,2, or 4 instead\n", len);
+		inject_gp(apic->vcpu);
+		break;
+	}
+
+	/* too common printing */
+	if (offset != APIC_EOI)
+		apic_debug("%s: offset 0x%x with length 0x%x, and value is " \
+			 "0x%lx\n",
+		       __FUNCTION__, offset, len, val);
+
+	offset &= 0xff0;
+
+	switch (offset) {
+	case APIC_ID:   /* Local APIC ID */
+		apic_set_reg(apic, APIC_ID, val);
+		break;
+
+	case APIC_TASKPRI:
+		apic_set_tpr(apic, val & 0xff);
+		break;
+
+	case APIC_EOI:
+		apic_set_eoi(apic);
+		break;
+
+	case APIC_LDR:
+		apic_set_reg(apic, APIC_LDR, val & APIC_LDR_MASK);
+		break;
+
+	case APIC_DFR:
+		apic_set_reg(apic, APIC_DFR, val | 0x0FFFFFFF);
+		break;
+
+	case APIC_SPIV:
+		apic_set_reg(apic, APIC_SPIV, val & 0x3ff);
+		if (!(val & APIC_SPIV_APIC_ENABLED)) {
+			int i;
+			u32 lvt_val;
+
+			apic->status |= APIC_SOFTWARE_DISABLE_MASK;
+			for (i = 0; i < APIC_LVT_NUM; i++) {
+				lvt_val = apic_get_reg(apic,
+							   APIC_LVTT +
+							   0x10 * i);
+				apic_set_reg(apic, APIC_LVTT + 0x10 * i,
+						 lvt_val | APIC_LVT_MASKED);
+			}
+
+			if ((apic_get_reg(apic, APIC_LVT0) &
+			     APIC_MODE_MASK) == APIC_DM_EXTINT)
+				clear_bit(_APIC_BSP_ACCEPT_PIC, &apic->status);
+		} else {
+			apic->status &= ~APIC_SOFTWARE_DISABLE_MASK;
+			if ((apic_get_reg(apic, APIC_LVT0) &
+			     APIC_MODE_MASK) == APIC_DM_EXTINT)
+				set_bit(_APIC_BSP_ACCEPT_PIC, &apic->status);
+		}
+		break;
+
+	case APIC_ESR:
+		apic->err_write_count = !apic->err_write_count;
+		if (!apic->err_write_count)
+			apic->err_status = 0;
+		break;
+
+	case APIC_ICR:
+		/* No delay here, so we always clear the pending bit*/
+		apic_set_reg(apic, APIC_ICR, val & ~(1 << 12));
+		apic_send_ipi(apic);
+		break;
+
+	case APIC_ICR2:
+		apic_set_reg(apic, APIC_ICR2, val & 0xff000000);
+		break;
+
+	case APIC_LVTT:
+	case APIC_LVTTHMR:
+	case APIC_LVTPC:
+	case APIC_LVT0:
+	case APIC_LVT1:
+	case APIC_LVTERR:
+	{
+		if (apic->status & APIC_SOFTWARE_DISABLE_MASK)
+			val |= APIC_LVT_MASKED;
+
+		val &= apic_lvt_mask[(offset - APIC_LVTT) >> 4];
+		apic_set_reg(apic, offset, val);
+
+		/* On hardware, when write vector less than 0x20 will error */
+		if (!(val & APIC_LVT_MASKED))
+			apic_check_vector(apic, apic_lvt_dm(apic, offset),
+					  apic_lvt_vector(apic, offset));
+		if (!apic->vcpu_id && (offset == APIC_LVT0)) {
+			if ((val & APIC_MODE_MASK) == APIC_DM_EXTINT)
+				if (val & APIC_LVT_MASKED)
+					clear_bit(_APIC_BSP_ACCEPT_PIC,
+						  &apic->status);
+				else
+					set_bit(_APIC_BSP_ACCEPT_PIC,
+						&apic->status);
+			else
+				clear_bit(_APIC_BSP_ACCEPT_PIC,
+					  &apic->status);
+		}
+	}
+		break;
+
+	case APIC_TMICT:
+	{
+		ktime_t now = apic->timer.dev.base->get_time();
+		u32 offset;
+
+		apic_set_reg(apic, APIC_TMICT, val);
+		apic_set_reg(apic, APIC_TMCCT, val);
+		apic->timer.last_update = now;
+		offset = APIC_BUS_CYCLE_NS * apic->timer.divide_count * val;
+
+		/* Make sure the lock ordering is coherent */
+		spin_unlock_bh(&apic->lock);
+		hrtimer_cancel(&apic->timer.dev);
+		hrtimer_start(&apic->timer.dev,
+			      ktime_add_ns(now, offset),
+			      HRTIMER_MODE_ABS);
+
+		apic_debug("%s: bus cycle is %"PRId64"ns, now 0x%016"PRIx64", "
+			 "timer initial count 0x%x, offset 0x%x, "
+			 "expire @ 0x%016"PRIx64".\n", __FUNCTION__,
+			 APIC_BUS_CYCLE_NS, ktime_to_ns(now),
+			 apic_get_reg(apic, APIC_TMICT),
+			 offset, ktime_to_ns(ktime_add_ns(now, offset)));
+	}
+		return;
+
+	case APIC_TDCR:
+	{
+		unsigned int tmp1, tmp2;
+
+		tmp1 = val & 0xf;
+		tmp2 = ((tmp1 & 0x3) | ((tmp1 & 0x8) >> 1)) + 1;
+		apic->timer.divide_count = 0x1 << (tmp2 & 0x7);
+
+		apic_set_reg(apic, APIC_TDCR, val);
+
+		apic_debug("timer divide count is 0x%x\n",
+		       apic->timer.divide_count);
+	}
+		break;
+
+	default:
+		printk(KERN_WARNING "Local APIC Write to read-only register\n");
+		break;
+	}
+
+	spin_unlock_bh(&apic->lock);
+}
+
+static int apic_mmio_range(struct kvm_io_device *this, gpa_t addr)
+{
+	struct kvm_kern_apic *apic = (struct kvm_kern_apic*)this->private;
+	int ret = 0;
+
+	spin_lock_bh(&apic->lock);
+
+	if (apic_global_enabled(apic) &&
+	    (addr >= apic->base_address) &&
+	    (addr < (apic->base_address + VLOCAL_APIC_MEM_LENGTH)))
+		ret = 1;
+
+	spin_unlock_bh(&apic->lock);
+
+	return ret;
+}
+
+static void apic_mmio_destructor(struct kvm_io_device *this)
+{
+	struct kvm_kern_apic *apic = (struct kvm_kern_apic*)this->private;
+
+	apic_dropref(apic);
+}
+
+static void apic_mmio_register(struct kvm_kern_apic *apic)
+{
+	/* Register ourselves with the MMIO subsystem */
+	struct kvm_io_device *dev = &apic->mmio_dev;
+
+	dev->read       = apic_mmio_read;
+	dev->write      = apic_mmio_write;
+	dev->in_range   = apic_mmio_range;
+	dev->destructor = apic_mmio_destructor;
+
+	dev->private = apic;
+	atomic_inc(&apic->ref_count);
+
+	apic->vcpu->apic.mmio = dev;
+}
+
+/*
+ *----------------------------------------------------------------------
+ * LAPIC interface
+ *----------------------------------------------------------------------
+ */
+
+void kvm_lapic_set_tpr(struct kvm_vcpu *vcpu, u64 cr8)
+{
+	struct kvm_kern_apic *apic = (struct kvm_kern_apic*)vcpu->apic.dev;
+
+	spin_lock_bh(&apic->lock);
+	apic_set_tpr(apic, ((cr8 & 0x0f) << 4));
+	spin_unlock_bh(&apic->lock);
+}
+
+u64 kvm_lapic_get_tpr(struct kvm_vcpu *vcpu)
+{
+	struct kvm_kern_apic *apic = (struct kvm_kern_apic*)vcpu->apic.dev;
+	u64 tpr;
+
+	spin_lock_bh(&apic->lock);
+	tpr = (u64)apic_get_reg(apic, APIC_TASKPRI);
+	spin_unlock_bh(&apic->lock);
+
+	return (tpr & 0xf0) >> 4;
+}
+EXPORT_SYMBOL_GPL(kvm_lapic_get_tpr);
+
+void kvm_lapic_set_base(struct kvm_vcpu *vcpu, u64 value)
+{
+	struct kvm_kern_apic *apic = (struct kvm_kern_apic*)vcpu->apic.dev;
+
+	spin_lock_bh(&apic->lock);
+	if (apic->vcpu_id)
+		value &= ~MSR_IA32_APICBASE_BSP;
+
+	apic->base_msr = value;
+	apic->base_address = apic->base_msr & MSR_IA32_APICBASE_BASE;
+
+	/* with FSB delivery interrupt, we can restart APIC functionality */
+	if (!(value & MSR_IA32_APICBASE_ENABLE))
+		set_bit(_APIC_GLOB_DISABLE, &apic->status);
+	else
+		clear_bit(_APIC_GLOB_DISABLE, &apic->status);
+
+	apic_debug("apic base msr is 0x%016"PRIx64", and base address is " \
+		 "0x%lx.\n", apic->base_msr, apic->base_address);
+
+	spin_unlock_bh(&apic->lock);
+}
+
+u64 kvm_lapic_get_base(struct kvm_vcpu *vcpu)
+{
+	struct kvm_kern_apic *apic = (struct kvm_kern_apic*)vcpu->apic.dev;
+	u64 base;
+
+	spin_lock_bh(&apic->lock);
+	base = apic->base_msr;
+	spin_unlock_bh(&apic->lock);
+
+	return base;
+}
+EXPORT_SYMBOL_GPL(kvm_lapic_get_base);
+
+void kvm_lapic_save(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs)
+{
+	/*
+	 * FIXME: This needs to support the entire register set when
+	 * enabled
+	 */
+	sregs->cr8       = kvm_lapic_get_tpr(vcpu);
+	sregs->apic_base = kvm_lapic_get_base(vcpu);
+}
+
+void kvm_lapic_restore(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs)
+{
+	/*
+	 * FIXME: This needs to support the entire register set when
+	 * enabled
+	 */
+	kvm_lapic_set_tpr(vcpu, sregs->cr8);
+	kvm_lapic_set_base(vcpu, sregs->apic_base);
+}
+
+void kvm_lapic_reset(struct kvm_vcpu *vcpu)
+{
+	struct kvm_kern_apic *apic;
+	int i;
+
+	apic_debug("%s\n", __FUNCTION__);
+
+	ASSERT(vcpu);
+	apic = vcpu->apic.dev;
+	ASSERT(apic != NULL);
+
+	/* Stop the timer in case it's a reset to an active apic */
+	hrtimer_cancel(&apic->timer.dev);
+
+	spin_lock_bh(&apic->lock);
+
+	apic_set_reg(apic, APIC_ID, vcpu_slot(vcpu) << 24);
+	apic_set_reg(apic, APIC_LVR, APIC_VERSION);
+
+	for (i = 0; i < APIC_LVT_NUM; i++)
+		apic_set_reg(apic, APIC_LVTT + 0x10 * i, APIC_LVT_MASKED);
+
+	apic_set_reg(apic, APIC_DFR, 0xffffffffU);
+	apic_set_reg(apic, APIC_SPIV, 0xff);
+	apic_set_reg(apic, APIC_TASKPRI, 0);
+	apic_set_reg(apic, APIC_LDR, 0);
+	apic_set_reg(apic, APIC_ESR, 0);
+	apic_set_reg(apic, APIC_ICR, 0);
+	apic_set_reg(apic, APIC_ICR2, 0);
+	apic_set_reg(apic, APIC_TDCR, 0);
+	apic_set_reg(apic, APIC_TMICT, 0);
+	memset((void*)(apic->regs + APIC_IRR), 0, KVM_IRQ_BITMAP_SIZE(u8));
+	memset((void*)(apic->regs + APIC_ISR), 0, KVM_IRQ_BITMAP_SIZE(u8));
+	memset((void*)(apic->regs + APIC_TMR), 0, KVM_IRQ_BITMAP_SIZE(u8));
+
+	apic->base_msr =
+		MSR_IA32_APICBASE_ENABLE |
+		APIC_DEFAULT_PHYS_BASE;
+	if (vcpu_slot(vcpu) == 0)
+		apic->base_msr |= MSR_IA32_APICBASE_BSP;
+	apic->base_address = apic->base_msr & MSR_IA32_APICBASE_BASE;
+
+	apic->timer.divide_count = 0;
+	apic->timer.pending = 0;
+	apic->status = 0;
+
+#ifdef APIC_NO_BIOS
+	/*
+	 * XXX According to mp specification, BIOS will enable LVT0/1,
+	 * remove it after BIOS enabled
+	 */
+	if (!vcpu_slot(vcpu)) {
+		apic_set_reg(apic, APIC_LVT0, APIC_MODE_EXTINT << 8);
+		apic_set_reg(apic, APIC_LVT1, APIC_MODE_NMI << 8);
+		set_bit(_APIC_BSP_ACCEPT_PIC, &apic->status);
+	}
+#endif
+
+	spin_unlock_bh(&apic->lock);
+
+	printk(KERN_INFO  "%s: vcpu=%p, id=%d, base_msr=" \
+	       "0x%016"PRIx64", base_address=0x%0lx.\n", __FUNCTION__, vcpu,
+	       GET_APIC_ID(apic_get_reg(apic, APIC_ID)),
+	       apic->base_msr, apic->base_address);
+}
+
+int kvm_lapic_enabled(struct kvm_vcpu *vcpu)
+{
+	struct kvm_kern_apic *apic = (struct kvm_kern_apic*)vcpu->apic.dev;
+	int ret = 0;
+
+	spin_lock_bh(&apic->lock);
+	if (!apic->usermode)
+		ret = apic_enabled(apic);
+	spin_unlock_bh(&apic->lock);
+
+	return ret;
+}
+
+/*
+ *----------------------------------------------------------------------
+ * timer interface
+ *----------------------------------------------------------------------
+ */
+static int __apic_timer_fn(struct kvm_kern_apic *apic)
+{
+	u32 vector;
+	ktime_t now;
+	int result = 0;
+
+	if (unlikely(!apic_enabled(apic) ||
+		     !apic_lvt_enabled(apic, APIC_LVTT))) {
+		apic_debug("%s: time interrupt although apic is down\n",
+			 __FUNCTION__);
+		return 0;
+	}
+
+	vector                  = apic_lvt_vector(apic, APIC_LVTT);
+	now                     = apic->timer.dev.base->get_time();
+	apic->timer.last_update = now;
+	apic->timer.pending++;
+
+	__apic_accept_irq(apic, APIC_DM_FIXED, vector, 1, 0);
+
+	if (apic_lvtt_period(apic)) {
+		u32 offset;
+		u32 tmict = apic_get_reg(apic, APIC_TMICT);
+
+		apic_set_reg(apic, APIC_TMCCT, tmict);
+		offset = APIC_BUS_CYCLE_NS * apic->timer.divide_count * tmict;
+
+		result = 1;
+		apic->timer.dev.expires = ktime_add_ns(now, offset);
+
+		apic_debug("%s: now 0x%016"PRIx64", expire @ 0x%016"PRIx64", "
+		       "timer initial count 0x%x, timer current count 0x%x.\n",
+		       __FUNCTION__,
+		       ktime_to_ns(now), ktime_add_ns(now, offset),
+		       apic_get_reg(apic, APIC_TMICT),
+	               apic_get_reg(apic, APIC_TMCCT));
+	} else {
+		apic_set_reg(apic, APIC_TMCCT, 0);
+		apic_debug("%s: now 0x%016"PRIx64", "
+		       "timer initial count 0x%x, timer current count 0x%x.\n",
+		       __FUNCTION__,
+		       ktime_to_ns(now), apic_get_reg(apic, APIC_TMICT),
+		       apic_get_reg(apic, APIC_TMCCT));
+	}
+
+	return result;
+}
+
+static enum hrtimer_restart apic_timer_fn(struct hrtimer *data)
+{
+	struct kvm_kern_apic *apic;
+	int restart_timer = 0;
+
+	apic = container_of(data, struct kvm_kern_apic, timer.dev);
+
+	spin_lock_bh(&apic->lock);
+	restart_timer = __apic_timer_fn(apic);
+	spin_unlock_bh(&apic->lock);
+
+	if (restart_timer)
+		return HRTIMER_RESTART;
+	else
+		return HRTIMER_NORESTART;
+}
+
+/*
+ *----------------------------------------------------------------------
+ * IRQDEVICE interface
+ *----------------------------------------------------------------------
+ */
+
+static int apic_irqdev_ack(struct kvm_irqdevice *this, int flags,
+			   struct kvm_irqack_data *data)
+{
+	struct kvm_kern_apic *apic = (struct kvm_kern_apic*)this->private;
+	int irq;
+
+	apic_debug("LAPIC ACK attempt\n");
+
+	spin_lock_bh(&apic->lock);
+
+	if (!apic_enabled(apic))
+		goto out;
+
+	if (!(flags & KVM_IRQACK_FLAG_PEEK)) {
+		irq = apic_find_highest_irr(apic);
+		if ((irq & 0xf0) > apic_get_reg(apic, APIC_PROCPRI)) {
+			BUG_ON (irq < 0x10);
+
+			__set_bit(irq, apic->regs + APIC_ISR);
+			__clear_bit(irq, apic->regs + APIC_IRR);
+			apic_update_ppr(apic);
+
+			/*
+			 * We have to special case the timer interrupt
+			 * because we want the vector to stay pending
+			 * for each tick of the clock, even for a backlog.
+			 * Therefore, if this was a timer vector and we
+			 * still have ticks pending, keep IRR set
+			 */
+			if (irq == apic_lvt_vector(apic, APIC_LVTT)) {
+				BUG_ON(!apic->timer.pending);
+				apic->timer.pending--;
+				if (apic->timer.pending)
+					__set_bit(irq, apic->regs + APIC_IRR);
+			}
+
+			data->flags |= KVM_IRQACKDATA_VECTOR_VALID;
+			data->vector = irq;
+		}
+		else
+			data->vector = -1;
+
+		apic_debug("ACK for vector %d\n", data->vector);
+	}
+
+	/*
+	 * See if there is anything still pending.  Don't forget that we may
+	 * have entered this function with vector=NULL just to check pending
+	 * status
+	 */
+	irq = apic_find_highest_irr(apic);
+	if (irq) {
+		/*
+		 * we check TASKPRI (as opposed to PROCPRI) because we
+		 * want to know the threshold against which the CPU is masking
+		 * interrupts, not the APIC itself.  The PROCPRI register
+		 * factors in the ISR in addition to the TPR.  Therefore, we
+		 * would never see a match here if we looked at PPR since
+		 * we just injected the highest ISR during this call
+		 */
+		if (irq > apic_get_reg(apic, APIC_TASKPRI))
+			data->flags |= KVM_IRQACKDATA_VECTOR_PENDING;
+	}
+
+ out:
+	spin_unlock_bh(&apic->lock);
+
+	return 0;
+}
+
+static int apic_irqdev_set_pin(struct kvm_irqdevice *this, int irq, int level)
+{
+	struct kvm_kern_apic *apic = (struct kvm_kern_apic*)this->private;
+	int lvt = 0;
+
+	spin_lock_bh(&apic->lock);
+
+	if (!apic_enabled(apic)) {
+		/*
+		 * If the LAPIC is disabled, we simply forward the interrupt
+		 * on to the output line
+		 */
+		__apic_accept_irq(apic, APIC_DM_EXTINT, 0, level, 1);
+		goto out;
+	}
+
+	/*
+	 * pin "0" is LINT0, and "1" is LINT1
+	 */
+	BUG_ON(irq > 1);
+
+	switch(irq) {
+	case 0:
+		lvt = APIC_LVT0;
+		break;
+	case 1:
+		lvt = APIC_LVT1;
+		break;
+	}
+
+	if (apic_lvt_enabled(apic, lvt))
+		__apic_accept_irq(apic,
+				  apic_lvt_dm(apic, lvt),
+				  apic_lvt_vector(apic, lvt),
+				  level,
+				  1);
+
+
+ out:
+	spin_unlock_bh(&apic->lock);
+
+	return 0;
+}
+
+static void apic_irqdev_destructor(struct kvm_irqdevice *this)
+{
+	struct kvm_kern_apic *apic = (struct kvm_kern_apic*)this->private;
+
+	apic_dropref(apic);
+}
+
+static void apic_irqdev_register(struct kvm_kern_apic *apic,
+				 struct kvm_irqdevice *dev)
+{
+	dev->ack         = apic_irqdev_ack;
+	dev->set_pin     = apic_irqdev_set_pin;
+	dev->destructor  = apic_irqdev_destructor;
+
+	dev->private = apic;
+	atomic_inc(&apic->ref_count);
+
+	apic->irq_dev = dev;
+}
+
+int kvm_lapic_init(struct kvm_vcpu *vcpu,
+		   struct kvm_irqdevice *irq_dev, int flags)
+{
+	struct kvm_kern_apic *apic = NULL;
+	struct kvm_io_device *mmio_dev = NULL;
+
+	ASSERT(vcpu != NULL);
+	apic_debug("apic_init %d\n", vcpu_slot(vcpu));
+
+	apic = kzalloc(sizeof(*apic), GFP_KERNEL);
+	if (!apic)
+		goto nomem;
+
+	spin_lock_init(&apic->lock);
+	atomic_inc(&apic->ref_count);
+	apic->vcpu_id = vcpu_slot(vcpu);
+
+	apic->regs_page = alloc_page(GFP_KERNEL);
+	if ( apic->regs_page == NULL ) {
+		printk(KERN_ALERT "malloc apic regs error for vcpu %x\n",
+		       vcpu_slot(vcpu));
+		goto nomem;
+	}
+	apic->regs = page_address(apic->regs_page);
+	memset(apic->regs, 0, PAGE_SIZE);
+
+	apic->vcpu = vcpu;
+	vcpu->apic.dev = apic;
+
+	if (!(flags & KVM_LAPIC_OPTION_USERMODE)) {
+		apic_irqdev_register(apic, irq_dev);
+		apic_mmio_register(apic);
+	} else
+		apic->usermode = 1;
+
+	hrtimer_init(&apic->timer.dev, CLOCK_MONOTONIC, HRTIMER_MODE_ABS);
+	apic->timer.dev.function = apic_timer_fn;
+
+	kvm_lapic_reset(vcpu);
+	return 0;
+
+ nomem:
+	if (mmio_dev)
+		kfree(mmio_dev);
+
+	if (apic)
+		apic_dropref(apic);
+
+	return -ENOMEM;
+}
+
+void kvm_lapic_destroy(struct kvm_vcpu *vcpu)
+{
+	struct kvm_kern_apic *apic = vcpu->apic.dev;
+
+	if (vcpu->apic.mmio)
+		kvm_iodevice_destructor(vcpu->apic.mmio);
+
+	apic_dropref(apic);
+}
diff --git a/drivers/kvm/svm.c b/drivers/kvm/svm.c
index 4898a06..0d8eeb2 100644
--- a/drivers/kvm/svm.c
+++ b/drivers/kvm/svm.c
@@ -570,9 +570,6 @@ static int svm_create_vcpu(struct kvm_vcpu *vcpu)
 
 	fx_init(vcpu);
 	vcpu->fpu_active = 1;
-	vcpu->apic_base = 0xfee00000 |
-			/*for vcpu 0*/ MSR_IA32_APICBASE_BSP |
-			MSR_IA32_APICBASE_ENABLE;
 
 	return 0;
 
@@ -1410,9 +1407,9 @@ static int do_intr_requests(struct kvm_vcpu *vcpu,
 			r = kvm_vcpu_irq_pop(vcpu, &ack);
 			break;
 		case kvm_irqpin_extint:
-			printk(KERN_WARNING "KVM: external-interrupts not " \
-			       "handled yet\n");
-			__clear_bit(pin, &vcpu->irq.pending);
+			r = kvm_irqdevice_ack(&vcpu->kvm->isa_irq, 0, &ack);
+			if (!(ack.flags & KVM_IRQACKDATA_VECTOR_PENDING))
+				__clear_bit(pin, &vcpu->irq.pending);
 			break;
 		case kvm_irqpin_nmi:
 			/*
@@ -1501,8 +1498,8 @@ static void post_kvm_run_save(struct kvm_vcpu *vcpu,
 		(vcpu->interrupt_window_open &&
 		 !kvm_vcpu_irq_pending(vcpu));
 	kvm_run->if_flag = (vcpu->svm->vmcb->save.rflags & X86_EFLAGS_IF) != 0;
-	kvm_run->cr8 = vcpu->cr8;
-	kvm_run->apic_base = vcpu->apic_base;
+	kvm_run->cr8 = kvm_lapic_get_tpr(vcpu);
+	kvm_run->apic_base = kvm_lapic_get_base(vcpu);
 }
 
 /*
diff --git a/drivers/kvm/userint.c b/drivers/kvm/userint.c
index 08d26fa..16987e1 100644
--- a/drivers/kvm/userint.c
+++ b/drivers/kvm/userint.c
@@ -218,6 +218,12 @@ int kvm_user_irqdev_restore(struct kvm_irqdevice *this, void *data)
 
 int kvm_userint_init(struct kvm_vcpu *vcpu)
 {
-	return kvm_user_irqdev_init(&vcpu->irq.dev);
+	int ret;
+
+	ret = kvm_user_irqdev_init(&vcpu->irq.dev);
+	if (ret < 0)
+		return ret;
+
+	return kvm_lapic_init(vcpu, NULL, KVM_LAPIC_OPTION_USERMODE);
 }
 
diff --git a/drivers/kvm/vmx.c b/drivers/kvm/vmx.c
index 254c5b7..65ae52d 100644
--- a/drivers/kvm/vmx.c
+++ b/drivers/kvm/vmx.c
@@ -1196,10 +1196,6 @@ static int vmx_vcpu_setup(struct kvm_vcpu *vcpu)
 
 	memset(vcpu->regs, 0, sizeof(vcpu->regs));
 	vcpu->regs[VCPU_REGS_RDX] = get_rdx_init_val();
-	vcpu->cr8 = 0;
-	vcpu->apic_base = 0xfee00000 |
-			/*for vcpu 0*/ MSR_IA32_APICBASE_BSP |
-			MSR_IA32_APICBASE_ENABLE;
 
 	fx_init(vcpu);
 
@@ -1441,9 +1437,9 @@ static int do_intr_requests(struct kvm_vcpu *vcpu,
 			r = kvm_vcpu_irq_pop(vcpu, &ack);
 			break;
 		case kvm_irqpin_extint:
-			printk(KERN_WARNING "KVM: external-interrupts not " \
-			       "handled yet\n");
-			__clear_bit(pin, &vcpu->irq.pending);
+			r = kvm_irqdevice_ack(&vcpu->kvm->isa_irq, 0, &ack);
+			if (!(ack.flags & KVM_IRQACKDATA_VECTOR_PENDING))
+				__clear_bit(pin, &vcpu->irq.pending);
 			break;
 		case kvm_irqpin_nmi:
 			/*
@@ -1800,7 +1796,7 @@ static int handle_cr(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 			return 1;
 		case 8:
 			vcpu_load_rsp_rip(vcpu);
-			vcpu->regs[reg] = vcpu->cr8;
+			vcpu->regs[reg] = kvm_lapic_get_tpr(vcpu);
 			vcpu_put_rsp_rip(vcpu);
 			skip_emulated_instruction(vcpu);
 			return 1;
@@ -1897,8 +1893,8 @@ static void post_kvm_run_save(struct kvm_vcpu *vcpu,
 			      struct kvm_run *kvm_run)
 {
 	kvm_run->if_flag = (vmcs_readl(GUEST_RFLAGS) & X86_EFLAGS_IF) != 0;
-	kvm_run->cr8 = vcpu->cr8;
-	kvm_run->apic_base = vcpu->apic_base;
+	kvm_run->cr8 = kvm_lapic_get_tpr(vcpu);
+	kvm_run->apic_base = kvm_lapic_get_base(vcpu);
 	kvm_run->ready_for_interrupt_injection =
 		(vcpu->interrupt_window_open &&
 		 !kvm_vcpu_irq_pending(vcpu));
diff --git a/include/linux/kvm.h b/include/linux/kvm.h
index e6edca8..a83606b 100644
--- a/include/linux/kvm.h
+++ b/include/linux/kvm.h
@@ -231,6 +231,17 @@ struct kvm_dirty_log {
 	};
 };
 
+/* for KVM_APIC */
+struct kvm_apic_msg {
+	/* in */
+	__u32 dest;
+	__u32 trig_mode;
+	__u32 dest_mode;
+	__u32 delivery_mode;
+	__u32 vector;
+	__u32 padding;
+};
+
 struct kvm_cpuid_entry {
 	__u32 function;
 	__u32 eax;
@@ -282,6 +293,9 @@ struct kvm_signal_mask {
 #define KVM_CREATE_VCPU           _IO(KVMIO,  0x41)
 #define KVM_GET_DIRTY_LOG         _IOW(KVMIO, 0x42, struct kvm_dirty_log)
 #define KVM_SET_MEMORY_ALIAS      _IOW(KVMIO, 0x43, struct kvm_memory_alias)
+#define KVM_ENABLE_KERNEL_PIC     _IOW(KVMIO, 0x44, __u32)
+#define KVM_ISA_INTERRUPT         _IOW(KVMIO, 0x45, struct kvm_interrupt)
+#define KVM_APIC_MSG		  _IOW(KVMIO, 0x46, struct kvm_apic_msg)
 
 /*
  * ioctls for vcpu fds
@@ -300,5 +314,5 @@ struct kvm_signal_mask {
 #define KVM_SET_SIGNAL_MASK       _IOW(KVMIO,  0x8b, struct kvm_signal_mask)
 #define KVM_GET_FPU               _IOR(KVMIO,  0x8c, struct kvm_fpu)
 #define KVM_SET_FPU               _IOW(KVMIO,  0x8d, struct kvm_fpu)
-
+#define KVM_APIC_RESET		  _IO(KVMIO,   0x8e)
 #endif


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

[PATCH 07/10] KVM: Adds support for real NMI injection on VMX processors

[Gregory Haskins]

Signed-off-by: Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
---

 drivers/kvm/vmx.c |   63 +++++++++++++++++++++++++++++++++++++++++++++++++----
 drivers/kvm/vmx.h |    3 +++
 2 files changed, 61 insertions(+), 5 deletions(-)

diff --git a/drivers/kvm/vmx.c b/drivers/kvm/vmx.c
index 65ae52d..609b929 100644
--- a/drivers/kvm/vmx.c
+++ b/drivers/kvm/vmx.c
@@ -1261,7 +1261,14 @@ static int vmx_vcpu_setup(struct kvm_vcpu *vcpu)
 			       PIN_BASED_VM_EXEC_CONTROL,
 			       PIN_BASED_EXT_INTR_MASK   /* 20.6.1 */
 			       | PIN_BASED_NMI_EXITING   /* 20.6.1 */
+			       | PIN_BASED_VIRTUAL_NMI   /* 20.6.1 */
 			);
+
+	if (!(vmcs_read32(PIN_BASED_VM_EXEC_CONTROL) & PIN_BASED_VIRTUAL_NMI))
+		printk(KERN_WARNING "KVM: Warning - Host processor does " \
+		       "not support virtual-NMI injection.  Using IRQ " \
+		       "method\n");
+ 
 	vmcs_write32_fixedbits(MSR_IA32_VMX_PROCBASED_CTLS,
 			       CPU_BASED_VM_EXEC_CONTROL,
 			       CPU_BASED_HLT_EXITING         /* 20.6.2 */
@@ -1411,6 +1418,43 @@ static void inject_rmode_irq(struct kvm_vcpu *vcpu, int irq)
 	vmcs_writel(GUEST_RSP, (vmcs_readl(GUEST_RSP) & ~0xffff) | (sp - 6));
 }
 
+static int do_nmi_requests(struct kvm_vcpu *vcpu)
+{
+	int nmi_window = 0;
+
+	BUG_ON(!(test_bit(kvm_irqpin_nmi, &vcpu->irq.pending)));
+
+	nmi_window =
+		(((vmcs_read32(GUEST_INTERRUPTIBILITY_INFO) & 0xb) == 0)
+		 && (vmcs_read32(VM_ENTRY_INTR_INFO_FIELD)
+		     & INTR_INFO_VALID_MASK));
+
+	if (nmi_window) {
+		if (vcpu->rmode.active)
+			inject_rmode_irq(vcpu, 2);
+		else
+			vmcs_write32(VM_ENTRY_INTR_INFO_FIELD,
+				     2 |
+				     INTR_TYPE_NMI |
+				     INTR_INFO_VALID_MASK);
+
+		__clear_bit(kvm_irqpin_nmi, &vcpu->irq.pending);
+	} else {
+		/*
+		 * NMIs blocked.  Wait for unblock.
+		 */
+		u32 cbvec = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL);
+		cbvec |= CPU_BASED_NMI_EXITING;
+		vmcs_write32(CPU_BASED_VM_EXEC_CONTROL, cbvec);	
+	}
+
+	/*
+	 * nmi_window correctly reflects whether we handled this interrupt
+	 * or not, so just return it as the "handled" indicator
+	 */
+	return nmi_window;
+}
+
 static int do_intr_requests(struct kvm_vcpu *vcpu,
 			    struct kvm_run *kvm_run,
 			    kvm_irqpin_t pin)
@@ -1443,9 +1487,11 @@ static int do_intr_requests(struct kvm_vcpu *vcpu,
 			break;
 		case kvm_irqpin_nmi:
 			/*
-			 * FIXME: Someday we will handle this using the
-			 * specific VMX NMI features.  For now, just inject
-			 * the NMI as a standard interrupt on vector 2
+			 * We should only get here if the processor does
+			 * not support virtual NMIs.  Inject the NMI as a
+			 * standard interrupt on vector 2.  The implication is
+			 * that NMIs are going to be subject to RFLAGS.IF
+			 * masking, unfortunately.
 			 */
 			ack.flags |= KVM_IRQACKDATA_VECTOR_VALID;
 			ack.vector = 2;
@@ -1488,7 +1534,8 @@ static int do_intr_requests(struct kvm_vcpu *vcpu,
 static void clear_pending_controls(struct kvm_vcpu *vcpu)
 {
 	u32 cbvec = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL);
-	cbvec &= ~CPU_BASED_VIRTUAL_INTR_PENDING;
+	cbvec &= ~(CPU_BASED_VIRTUAL_INTR_PENDING
+		   | CPU_BASED_NMI_EXITING);
 	vmcs_write32(CPU_BASED_VM_EXEC_CONTROL, cbvec);
 }
 
@@ -1505,7 +1552,6 @@ static void do_interrupt_requests(struct kvm_vcpu *vcpu,
 		switch (pin) {
 		case kvm_irqpin_localint:
 		case kvm_irqpin_extint:
-		case kvm_irqpin_nmi:
 			do_intr_requests(vcpu, kvm_run, pin);
 			break;
 		case kvm_irqpin_smi:
@@ -1513,6 +1559,13 @@ static void do_interrupt_requests(struct kvm_vcpu *vcpu,
 			printk(KERN_WARNING "KVM: dropping unhandled SMI\n");
 			__clear_bit(pin, &vcpu->irq.pending);
 			break;
+		case kvm_irqpin_nmi:
+			if (vmcs_read32(PIN_BASED_VM_EXEC_CONTROL)
+			    & PIN_BASED_VIRTUAL_NMI)
+				do_nmi_requests(vcpu);
+			else
+				do_intr_requests(vcpu, kvm_run, pin);	
+			break;
 		case kvm_irqpin_invalid:
 			/* drop */
 			break;
diff --git a/drivers/kvm/vmx.h b/drivers/kvm/vmx.h
index d0dc93d..d3fe017 100644
--- a/drivers/kvm/vmx.h
+++ b/drivers/kvm/vmx.h
@@ -35,6 +35,7 @@
 #define CPU_BASED_CR8_LOAD_EXITING      0x00080000
 #define CPU_BASED_CR8_STORE_EXITING     0x00100000
 #define CPU_BASED_TPR_SHADOW            0x00200000
+#define CPU_BASED_NMI_EXITING           0x00400000
 #define CPU_BASED_MOV_DR_EXITING        0x00800000
 #define CPU_BASED_UNCOND_IO_EXITING     0x01000000
 #define CPU_BASED_ACTIVATE_IO_BITMAP    0x02000000
@@ -44,6 +45,7 @@
 
 #define PIN_BASED_EXT_INTR_MASK 0x1
 #define PIN_BASED_NMI_EXITING   0x8
+#define PIN_BASED_VIRTUAL_NMI   0x20
 
 #define VM_EXIT_ACK_INTR_ON_EXIT        0x00008000
 #define VM_EXIT_HOST_ADD_SPACE_SIZE     0x00000200
@@ -221,6 +223,7 @@ enum vmcs_field {
 #define VECTORING_INFO_VALID_MASK       	INTR_INFO_VALID_MASK
 
 #define INTR_TYPE_EXT_INTR              (0 << 8) /* external interrupt */
+#define INTR_TYPE_NMI                   (2 << 8) /* non-maskable interrupt */
 #define INTR_TYPE_EXCEPTION             (3 << 8) /* processor exception */
 
 /*


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

[PATCH 08/10] KVM: Adds basic plumbing to support TPR shadow features

[Gregory Haskins]

Signed-off-by: Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
---

 drivers/kvm/irqdevice.h |    3 +++
 drivers/kvm/kvm.h       |    1 +
 drivers/kvm/lapic.c     |   15 +++++++++++++++
 3 files changed, 19 insertions(+), 0 deletions(-)

diff --git a/drivers/kvm/irqdevice.h b/drivers/kvm/irqdevice.h
index 097d179..173313d 100644
--- a/drivers/kvm/irqdevice.h
+++ b/drivers/kvm/irqdevice.h
@@ -45,12 +45,14 @@ struct kvm_irqsink {
 
 #define KVM_IRQACKDATA_VECTOR_VALID   (1 << 0)
 #define KVM_IRQACKDATA_VECTOR_PENDING (1 << 1)
+#define KVM_IRQACKDATA_NEXT_VALID     (1 << 2)
 
 #define KVM_IRQACK_FLAG_PEEK          (1 << 0)
 
 struct kvm_irqack_data {
 	int flags;
 	int vector;
+	int next;
 };
 
 struct kvm_irqdevice {
@@ -92,6 +94,7 @@ static inline void kvm_irqdevice_init(struct kvm_irqdevice *dev)
  *               data.flags -
  *                  [KVM_IRQACKDATA_VECTOR_VALID - data.vector is valid]
  *                  [KVM_IRQACKDATA_VECTOR_PENDING - more vectors are pending]
+ *                  [KVM_IRQACKDATA_NEXT_VALID - next-vector is valid]
  *
  * Returns: (int)
  *   [-1 = failure]
diff --git a/drivers/kvm/kvm.h b/drivers/kvm/kvm.h
index f112bd6..cf90f3f 100644
--- a/drivers/kvm/kvm.h
+++ b/drivers/kvm/kvm.h
@@ -180,6 +180,7 @@ void kvm_lapic_save(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs);
 void kvm_lapic_restore(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs);
 void kvm_lapic_reset(struct kvm_vcpu *vcpu);
 int  kvm_lapic_enabled(struct kvm_vcpu *vcpu);
+void *kvm_lapic_get_regs(struct kvm_vcpu *vcpu);
 
 /*
  * x86 supports 3 paging modes (4-level 64-bit, 3-level 64-bit, and 2-level
diff --git a/drivers/kvm/lapic.c b/drivers/kvm/lapic.c
index e7e7bd4..264a2e5 100644
--- a/drivers/kvm/lapic.c
+++ b/drivers/kvm/lapic.c
@@ -1149,6 +1149,13 @@ int kvm_lapic_enabled(struct kvm_vcpu *vcpu)
 	return ret;
 }
 
+void *kvm_lapic_get_regs(struct kvm_vcpu *vcpu)
+{
+	struct kvm_kern_apic *apic = (struct kvm_kern_apic*)vcpu->apic.dev;
+	return apic->regs;
+}
+EXPORT_SYMBOL_GPL(kvm_lapic_get_regs);
+
 /*
  *----------------------------------------------------------------------
  * timer interface
@@ -1287,6 +1294,14 @@ static int apic_irqdev_ack(struct kvm_irqdevice *this, int flags,
 		 */
 		if (irq > apic_get_reg(apic, APIC_TASKPRI))
 			data->flags |= KVM_IRQACKDATA_VECTOR_PENDING;
+
+		/*
+		 * We report the next pending vector here so that the system
+		 * can asses TPR thresholds for TPR-shadowing purposes
+		 * (if applicable)
+		 */
+		data->next   = irq;
+		data->flags |= KVM_IRQACKDATA_NEXT_VALID;
 	}
 
  out:


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

[PATCH 09/10] KVM: Add statistics from interrupt subsystem

[Gregory Haskins]

Signed-off-by: Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
---

 drivers/kvm/kvm.h      |    3 +++
 drivers/kvm/kvm_main.c |    5 +++++
 drivers/kvm/vmx.c      |    1 +
 3 files changed, 9 insertions(+), 0 deletions(-)

diff --git a/drivers/kvm/kvm.h b/drivers/kvm/kvm.h
index cf90f3f..a189abf 100644
--- a/drivers/kvm/kvm.h
+++ b/drivers/kvm/kvm.h
@@ -276,6 +276,9 @@ struct kvm_stat {
 	u32 request_irq_exits;
 	u32 irq_exits;
 	u32 light_exits;
+	u32 irq_posted;
+	u32 irq_accepted;
+	u32 guest_intr;
 };
 
 struct kvm_io_device {
diff --git a/drivers/kvm/kvm_main.c b/drivers/kvm/kvm_main.c
index 94d92aa..2ffa78f 100644
--- a/drivers/kvm/kvm_main.c
+++ b/drivers/kvm/kvm_main.c
@@ -73,6 +73,9 @@ static struct kvm_stats_debugfs_item {
 	{ "request_irq", STAT_OFFSET(request_irq_exits) },
 	{ "irq_exits", STAT_OFFSET(irq_exits) },
 	{ "light_exits", STAT_OFFSET(light_exits) },
+	{ "irq_posted", STAT_OFFSET(irq_posted) },
+	{ "irq_accepted", STAT_OFFSET(irq_accepted) },
+	{ "guest_intr", STAT_OFFSET(guest_intr) },
 	{ NULL }
 };
 
@@ -2527,6 +2530,7 @@ static void kvm_vcpu_intr(struct kvm_irqsink *this,
 		 * Record the change..
 		 */
 		__set_bit(pin, &vcpu->irq.pending);
+		++vcpu->stat.irq_posted;
 
 		/*
 		 * then wake up the vcpu (if necessary)
@@ -2540,6 +2544,7 @@ static void kvm_vcpu_intr(struct kvm_irqsink *this,
 				 */
 				direct_ipi = task_cpu(vcpu->irq.task);
 				BUG_ON(direct_ipi == smp_processor_id());
+				++vcpu->stat.guest_intr;
 			} else {
 				/*
 				 * otherwise, we must assume that we could be
diff --git a/drivers/kvm/vmx.c b/drivers/kvm/vmx.c
index 609b929..1fc3845 100644
--- a/drivers/kvm/vmx.c
+++ b/drivers/kvm/vmx.c
@@ -1514,6 +1514,7 @@ static int do_intr_requests(struct kvm_vcpu *vcpu,
 					     INTR_INFO_VALID_MASK);
 
 			handled = 1;
+			++vcpu->stat.irq_accepted;
 		}
 	}
 


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 09/10] KVM: Add statistics from interrupt subsystem

[Avi Kivity]

Gregory Haskins wrote:
>  		 * then wake up the vcpu (if necessary)
> @@ -2540,6 +2544,7 @@ static void kvm_vcpu_intr(struct kvm_irqsink *this,
>  				 */
>  				direct_ipi = task_cpu(vcpu->irq.task);
>  				BUG_ON(direct_ipi == smp_processor_id());
> +				++vcpu->stat.guest_intr;
>  			} else {
>  				/*
>   

Maybe call it guest_preempt to make it a bit more specific.


-- 
error compiling committee.c: too many arguments to function


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

[PATCH 10/10] KVM: Adds support for TPR shadowing under VMX processors

[Gregory Haskins]

Signed-off-by: Gregory Haskins <ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org>
---

 drivers/kvm/vmx.c |   32 ++++++++++++++++++++++++++------
 1 files changed, 26 insertions(+), 6 deletions(-)

diff --git a/drivers/kvm/vmx.c b/drivers/kvm/vmx.c
index 1fc3845..9b41cb8 100644
--- a/drivers/kvm/vmx.c
+++ b/drivers/kvm/vmx.c
@@ -1272,13 +1272,26 @@ static int vmx_vcpu_setup(struct kvm_vcpu *vcpu)
 	vmcs_write32_fixedbits(MSR_IA32_VMX_PROCBASED_CTLS,
 			       CPU_BASED_VM_EXEC_CONTROL,
 			       CPU_BASED_HLT_EXITING         /* 20.6.2 */
-			       | CPU_BASED_CR8_LOAD_EXITING    /* 20.6.2 */
-			       | CPU_BASED_CR8_STORE_EXITING   /* 20.6.2 */
+			       | CPU_BASED_TPR_SHADOW        /* 20.6.2 */
 			       | CPU_BASED_ACTIVATE_IO_BITMAP  /* 20.6.2 */
 			       | CPU_BASED_MOV_DR_EXITING
 			       | CPU_BASED_USE_TSC_OFFSETING   /* 21.3 */
 			);
 
+	if (!(vmcs_read32(CPU_BASED_VM_EXEC_CONTROL) & CPU_BASED_TPR_SHADOW)) {
+		u32 cbvec;
+
+		cbvec  = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL);
+		cbvec |= CPU_BASED_CR8_LOAD_EXITING;        /* 20.6.2 */
+		cbvec |= CPU_BASED_CR8_STORE_EXITING;       /* 20.6.2 */
+		vmcs_write32_fixedbits(MSR_IA32_VMX_PROCBASED_CTLS,
+				       CPU_BASED_VM_EXEC_CONTROL,
+				       cbvec);
+
+		printk(KERN_WARNING "KVM: Warning - Host processor does " \
+		       "not support TPR-shadow\n");
+	}
+
 	vmcs_write32(PAGE_FAULT_ERROR_CODE_MASK, 0);
 	vmcs_write32(PAGE_FAULT_ERROR_CODE_MATCH, 0);
 	vmcs_write32(CR3_TARGET_COUNT, 0);           /* 22.2.1 */
@@ -1351,7 +1364,7 @@ static int vmx_vcpu_setup(struct kvm_vcpu *vcpu)
 	vmcs_write32(VM_ENTRY_INTR_INFO_FIELD, 0);  /* 22.2.1 */
 
 #ifdef CONFIG_X86_64
-	vmcs_writel(VIRTUAL_APIC_PAGE_ADDR, 0);
+	vmcs_writel(VIRTUAL_APIC_PAGE_ADDR, kvm_lapic_get_regs(vcpu));
 	vmcs_writel(TPR_THRESHOLD, 0);
 #endif
 
@@ -1460,6 +1473,9 @@ static int do_intr_requests(struct kvm_vcpu *vcpu,
 			    kvm_irqpin_t pin)
 {
 	int handled = 0;
+	struct kvm_irqack_data ack;
+
+	memset(&ack, 0, sizeof(ack));
 
 	vcpu->interrupt_window_open =
 		((vmcs_readl(GUEST_RFLAGS) & X86_EFLAGS_IF) &&
@@ -1471,11 +1487,8 @@ static int do_intr_requests(struct kvm_vcpu *vcpu,
 		 * If interrupts enabled, and not blocked by sti or mov ss.
 		 * Good.
 		 */
-		struct kvm_irqack_data ack;
 		int r = 0;
 
-		memset(&ack, 0, sizeof(ack));
-
 		switch (pin) {
 		case kvm_irqpin_localint:
 			r = kvm_vcpu_irq_pop(vcpu, &ack);
@@ -1529,6 +1542,13 @@ static int do_intr_requests(struct kvm_vcpu *vcpu,
 		vmcs_write32(CPU_BASED_VM_EXEC_CONTROL, cbvec);
 	}
 
+#ifdef CONFIG_X86_64
+	if (ack.flags & KVM_IRQACKDATA_NEXT_VALID)
+		vmcs_writel(TPR_THRESHOLD, ack.next >> 4);
+	else
+		vmcs_writel(TPR_THRESHOLD, 0);
+#endif
+
 	return handled;
 }
 


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 00/10] in-kernel APIC v3 (kernel side)

[Dor Laor]

>At this point, I can boot
>
>A) 32-bit XP w/ACPI enabled
>B) 64-bit SLED (2.6.16 based)
>
>both with "level-0" (userspace PIC emulation) and "level-1" (in-kernel
PIC
>emulation).  The performance with XP under level-1 seems to be a huge
>improvement over in-trunk performance.

Impressive! Can you tell how much host cpu is consumed by an idle xp
guest w acpi?
Next week I'll try to check it with the pv network if I'll have the
time.


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 00/10] in-kernel APIC v3 (kernel side)

[Gregory Haskins]

>>> On Thu, May 10, 2007 at  9:07 AM, in message
<64F9B87B6B770947A9F8391472E032160BBA66AF-yEcIvxbTEBqsx+V+t5oei8rau4O3wl8o3fe8/T/H7NteoWH0uzbU5w@public.gmane.org>,
"Dor Laor" <dor.laor-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote: 
>> At this point, I can boot
>>
>>A) 32- bit XP w/ACPI enabled
>>B) 64- bit SLED (2.6.16 based)
>>
>>both with "level- 0" (userspace PIC emulation) and "level- 1" (in- kernel
> PIC
>>emulation).  The performance with XP under level- 1 seems to be a huge
>>improvement over in- trunk performance.
> 
> Impressive! Can you tell how much host cpu is consumed by an idle xp
> guest w acpi?
> Next week I'll try to check it with the pv network if I'll have the
> time.

The load-average on my system is about 1 while XP is idling.  qemu seems to be mostly at "0%" but will bounce up to 1% on occasion.  Here is the output of "top -b -p <qemu-pid>" over a few seconds:

top - 09:17:45 up 16:58,  3 users,  load average: 1.02, 0.86, 0.42
Tasks:   1 total,   0 running,   1 sleeping,   0 stopped,   0 zombie
Cpu(s):  0.1%us,  0.2%sy,  0.0%ni, 99.5%id,  0.1%wa,  0.0%hi,  0.1%si,  0.0%st
Mem:   3994704k total,  2018980k used,  1975724k free,    70996k buffers
Swap:  2104472k total,        0k used,  2104472k free,  1148284k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND            
10359 ghaskins  15   0  598m  83m  75m S    0  2.1   1:37.87 qemu-system-x86    


top - 09:17:48 up 16:58,  3 users,  load average: 1.02, 0.86, 0.42
Tasks:   1 total,   1 running,   0 sleeping,   0 stopped,   0 zombie
Cpu(s):  0.0%us,  0.2%sy,  0.0%ni, 99.7%id,  0.1%wa,  0.0%hi,  0.0%si,  0.0%st
Mem:   3994704k total,  2019120k used,  1975584k free,    70996k buffers
Swap:  2104472k total,        0k used,  2104472k free,  1148472k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND            
10359 ghaskins  15   0  598m  83m  75m R    0  2.1   1:37.87 qemu-system-x86    


top - 09:17:51 up 16:58,  3 users,  load average: 1.02, 0.87, 0.42
Tasks:   1 total,   1 running,   0 sleeping,   0 stopped,   0 zombie
Cpu(s):  0.0%us,  0.7%sy,  0.0%ni, 99.1%id,  0.0%wa,  0.0%hi,  0.2%si,  0.0%st
Mem:   3994704k total,  2019120k used,  1975584k free,    70996k buffers
Swap:  2104472k total,        0k used,  2104472k free,  1148560k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND            
10359 ghaskins  15   0  598m  83m  75m R    2  2.1   1:37.94 qemu-system-x86    


top - 09:17:54 up 16:58,  3 users,  load average: 1.02, 0.87, 0.42
Tasks:   1 total,   1 running,   0 sleeping,   0 stopped,   0 zombie
Cpu(s):  0.0%us,  0.2%sy,  0.0%ni, 99.7%id,  0.1%wa,  0.0%hi,  0.0%si,  0.0%st
Mem:   3994704k total,  2019368k used,  1975336k free,    70996k buffers
Swap:  2104472k total,        0k used,  2104472k free,  1148744k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND            
10359 ghaskins  15   0  598m  83m  75m R    0  2.1   1:37.94 qemu-system-x86    


top - 09:17:57 up 16:58,  3 users,  load average: 1.02, 0.87, 0.43
Tasks:   1 total,   1 running,   0 sleeping,   0 stopped,   0 zombie
Cpu(s):  0.0%us,  0.3%sy,  0.0%ni, 99.5%id,  0.2%wa,  0.0%hi,  0.0%si,  0.0%st
Mem:   3994704k total,  2019492k used,  1975212k free,    71000k buffers
Swap:  2104472k total,        0k used,  2104472k free,  1148832k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND            
10359 ghaskins  15   0  598m  83m  75m R    1  2.1   1:37.96 qemu-system-x86    


top - 09:18:00 up 16:59,  3 users,  load average: 1.01, 0.87, 0.43
Tasks:   1 total,   1 running,   0 sleeping,   0 stopped,   0 zombie
Cpu(s):  0.3%us,  0.9%sy,  0.0%ni, 98.4%id,  0.0%wa,  0.0%hi,  0.3%si,  0.0%st
Mem:   3994704k total,  2019988k used,  1974716k free,    71000k buffers
Swap:  2104472k total,        0k used,  2104472k free,  1149016k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND            
10359 ghaskins  15   0  598m  83m  75m R    0  2.1   1:37.96 qemu-system-x86    


top - 09:18:03 up 16:59,  3 users,  load average: 1.01, 0.87, 0.43
Tasks:   1 total,   1 running,   0 sleeping,   0 stopped,   0 zombie
Cpu(s):  0.0%us,  0.3%sy,  0.0%ni, 99.6%id,  0.2%wa,  0.0%hi,  0.0%si,  0.0%st
Mem:   3994704k total,  2019988k used,  1974716k free,    71000k buffers
Swap:  2104472k total,        0k used,  2104472k free,  1149104k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND            
10359 ghaskins  15   0  598m  83m  75m R    0  2.1   1:37.96 qemu-system-x86    


top - 09:18:06 up 16:59,  3 users,  load average: 1.01, 0.87, 0.43
Tasks:   1 total,   1 running,   0 sleeping,   0 stopped,   0 zombie
Cpu(s):  0.5%us,  0.7%sy,  0.0%ni, 98.8%id,  0.0%wa,  0.0%hi,  0.1%si,  0.0%st
Mem:   3994704k total,  2020044k used,  1974660k free,    71000k buffers
Swap:  2104472k total,        0k used,  2104472k free,  1149280k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND            
10359 ghaskins  15   0  598m  83m  75m R    0  2.1   1:37.96 qemu-system-x86    


top - 09:18:09 up 16:59,  3 users,  load average: 1.01, 0.87, 0.43
Tasks:   1 total,   1 running,   0 sleeping,   0 stopped,   0 zombie
Cpu(s):  0.0%us,  0.2%sy,  0.0%ni, 99.8%id,  0.0%wa,  0.0%hi,  0.0%si,  0.0%st
Mem:   3994704k total,  2020044k used,  1974660k free,    71000k buffers
Swap:  2104472k total,        0k used,  2104472k free,  1149376k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND            
10359 ghaskins  15   0  598m  83m  75m R    0  2.1   1:37.96 qemu-system-x86    


Here is the output of kvm_stat:
kvm statistics

 exits        12503136    1888
 guest_intr          0       0
 halt_exits      34478      64
 invlpg              0       0
 io_exits      1182337     305
 irq_accepte     64404      79
 irq_exits       12889       0
 irq_posted     110887     137
 irq_window          0       0
 light_exits  10954910    1519
 mmio_exits     294605       0
 pf_fixed       716870       4
 pf_guest        56142       0
 request_irq       158       0
 signal_exit     29311       0
 tlb_flush       58210      32







-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 00/10] in-kernel APIC v3 (kernel side)

[Dor Laor]

>-----Original Message-----
>From: Gregory Haskins [mailto:ghaskins-Et1tbQHTxzrQT0dZR+AlfA@public.gmane.org]
>Sent: Thursday, May 10, 2007 4:20 PM
>To: kvm-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org; Dor Laor
>Subject: RE: [kvm-devel] [PATCH 00/10] in-kernel APIC v3 (kernel side)
>
>>>> On Thu, May 10, 2007 at  9:07 AM, in message
><64F9B87B6B770947A9F8391472E032160BBA66AF@ehost011-
>8.exch011.intermedia.net>,
>"Dor Laor" <dor.laor-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote:
>>> At this point, I can boot
>>>
>>>A) 32- bit XP w/ACPI enabled
>>>B) 64- bit SLED (2.6.16 based)
>>>
>>>both with "level- 0" (userspace PIC emulation) and "level- 1" (in-
kernel
>> PIC
>>>emulation).  The performance with XP under level- 1 seems to be a
huge
>>>improvement over in- trunk performance.
>>
>> Impressive! Can you tell how much host cpu is consumed by an idle xp
>> guest w acpi?
>> Next week I'll try to check it with the pv network if I'll have the
>> time.
>
>The load-average on my system is about 1 while XP is idling.  qemu
seems to
>be mostly at "0%" but will bounce up to 1% on occasion.  Here is the
output
>of "top -b -p <qemu-pid>" over a few seconds:

Superb results, when I run the old in-kernel apic I got much higher idle
cpu consumtion. I know that's vmenter-vmexit latency was drastically
improved but 0-1% cpu consumption for winxp 32 bit sounds too good to be
true.
Are you sure an ACPI HAL is used in the guest? (can be checked by
looking at the computer node in the device manager).
I noticed the kvm_stat below, there are no mmio exits. You need to
update the statistics.

>
>top - 09:17:45 up 16:58,  3 users,  load average: 1.02, 0.86, 0.42
>Tasks:   1 total,   0 running,   1 sleeping,   0 stopped,   0 zombie
>Cpu(s):  0.1%us,  0.2%sy,  0.0%ni, 99.5%id,  0.1%wa,  0.0%hi,  0.1%si,
>0.0%st
>Mem:   3994704k total,  2018980k used,  1975724k free,    70996k
buffers
>Swap:  2104472k total,        0k used,  2104472k free,  1148284k cached
>
>  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
>10359 ghaskins  15   0  598m  83m  75m S    0  2.1   1:37.87
qemu-system-
>x86
>

...

>
>Here is the output of kvm_stat:
>kvm statistics
>
> exits        12503136    1888
> guest_intr          0       0
> halt_exits      34478      64
> invlpg              0       0
> io_exits      1182337     305
> irq_accepte     64404      79
> irq_exits       12889       0
> irq_posted     110887     137
> irq_window          0       0
> light_exits  10954910    1519
> mmio_exits     294605       0
> pf_fixed       716870       4
> pf_guest        56142       0
> request_irq       158       0
> signal_exit     29311       0
> tlb_flush       58210      32
>
>
>
>
>


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 00/10] in-kernel APIC v3 (kernel side)

[Gregory Haskins]

>>> On Sun, May 13, 2007 at  4:00 AM, in message
<64F9B87B6B770947A9F8391472E032160BC745D3-yEcIvxbTEBqsx+V+t5oei8rau4O3wl8o3fe8/T/H7NteoWH0uzbU5w@public.gmane.org>,
"Dor Laor" <dor.laor-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote: 
> 
> Superb results, when I run the old in- kernel apic I got much higher idle
> cpu consumtion. I know that's vmenter- vmexit latency was drastically
> improved but 0- 1% cpu consumption for winxp 32 bit sounds too good to be
> true.
> Are you sure an ACPI HAL is used in the guest? (can be checked by
> looking at the computer node in the device manager).

I did a clean install and didn't change it over to StandardPC.  The Control Panel says "ACPI Uniprocessor PC" or something like that (recalling from memory right now).  As a third datapoint, I am seeing the TPR changing rapidly (which I would expect 0 changes if the ACPI stuff was disabled).

> I noticed the kvm_stat below, there are no mmio exits. You need to
> update the statistics.

Yeah, this is fixed in my latest code which I will send out soon.  I actually added a "light_mmio" and "apic_mmio" stat so we can tell the difference.  I left mmio_exits to count the ones that go to userspace.



-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 00/10] in-kernel APIC v3 (kernel side)

[Dor Laor]

>>
>> Superb results, when I run the old in- kernel apic I got much higher
idle
>> cpu consumtion. I know that's vmenter- vmexit latency was drastically
>> improved but 0- 1% cpu consumption for winxp 32 bit sounds too good
to be
>> true.
>> Are you sure an ACPI HAL is used in the guest? (can be checked by
>> looking at the computer node in the device manager).
>
>I did a clean install and didn't change it over to StandardPC.  The
Control
>Panel says "ACPI Uniprocessor PC" or something like that (recalling
from
>memory right now).  As a third datapoint, I am seeing the TPR changing
>rapidly (which I would expect 0 changes if the ACPI stuff was
disabled).

Great, I was a little skeptical (sorry for that) since several months
ago when I tested the in-kernel-apic the idle time was much worse. After
talking today with Avi, he said his vmentry-exit latency drastically
improved performance and even the qemu's apic can reach these figures.
I'm sure that with load, you implementation will achieve performance
advantage.

>
>> I noticed the kvm_stat below, there are no mmio exits. You need to
>> update the statistics.
>
>Yeah, this is fixed in my latest code which I will send out soon.  I
>actually added a "light_mmio" and "apic_mmio" stat so we can tell the
>difference.  I left mmio_exits to count the ones that go to userspace.

Thanks, for the ultra-fast responsiveness.


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 00/10] in-kernel APIC v3 (kernel side)

[Gregory Haskins]

>>> On Sun, May 13, 2007 at 10:06 AM, in message
<64F9B87B6B770947A9F8391472E032160BC74642-yEcIvxbTEBqsx+V+t5oei8rau4O3wl8o3fe8/T/H7NteoWH0uzbU5w@public.gmane.org>,
"Dor Laor" <dor.laor-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote: 
>> >
>>> Superb results, when I run the old in-  kernel apic I got much higher
> idle
>>> cpu consumtion. I know that's vmenter-  vmexit latency was drastically
>>> improved but 0-  1% cpu consumption for winxp 32 bit sounds too good
> to be
>>> true.
>>> Are you sure an ACPI HAL is used in the guest? (can be checked by
>>> looking at the computer node in the device manager).
>>
>>I did a clean install and didn't change it over to StandardPC.  The
> Control
>>Panel says "ACPI Uniprocessor PC" or something like that (recalling
> from
>>memory right now).  As a third datapoint, I am seeing the TPR changing
>>rapidly (which I would expect 0 changes if the ACPI stuff was
> disabled).
> 
> Great, I was a little skeptical (sorry for that) 

No problem.

> since several months
> ago when I tested the in- kernel- apic the idle time was much worse. After
> talking today with Avi, he said his vmentry- exit latency drastically
> improved performance and even the qemu's apic can reach these figures.

Yeah.  At first I was psyched when I saw my code running at 0-1%, but then I ran trunk and saw its not all that different :(  The good news is that windows is performing pretty well now, regardless of my patch or not.  Thats good for KVM all around.

> I'm sure that with load, you implementation will achieve performance
> advantage.

In theory, we should be able to take advantage of all the light-exit gains we have made recently since that is the primary difference for windows+ACPI now.  This means the APIC change will have an increasing impact as we improve the VMEXIT times (both in HW and SW).  So even if it doesn't make a huge diff today, it will always be contributing more and more as we push the ball forward in other areas.

But I am not disappointed.  My primary motivation is functionality, not performance. ;)  I need to be able to inject interrupts to a guest from the kernel for the work I am doing.  If I also improve performance, great!



-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 00/10] in-kernel APIC v3 (kernel side)

[Dor Laor]

>
>> since several months
>> ago when I tested the in- kernel- apic the idle time was much worse.
>After
>> talking today with Avi, he said his vmentry- exit latency drastically
>> improved performance and even the qemu's apic can reach these
figures.
>
>Yeah.  At first I was psyched when I saw my code running at 0-1%, but
then
>I ran trunk and saw its not all that different :(  The good news is
that
>windows is performing pretty well now, regardless of my patch or not.
>Thats good for KVM all around.
>
>> I'm sure that with load, you implementation will achieve performance
>> advantage.
>
>In theory, we should be able to take advantage of all the light-exit
gains
>we have made recently since that is the primary difference for
windows+ACPI
>now.  This means the APIC change will have an increasing impact as we
>improve the VMEXIT times (both in HW and SW).  So even if it doesn't
make a
>huge diff today, it will always be contributing more and more as we
push
>the ball forward in other areas.
>
>But I am not disappointed.  My primary motivation is functionality, not
>performance. ;)  I need to be able to inject interrupts to a guest from
the
>kernel for the work I am doing.  If I also improve performance, great!
>

Your timing is good for us too since we have our pv network driver that
now works with qemu's interrupts. This means that we use a signal in
order to inject rx packets to the guest. In-kernel apic should improve
that, I'll soon measure the diff although using NAPI might dissolve the
performance gain.
It's good stuff the apic will be in the kernel, once it's there we need
to consider other components too.

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 00/10] in-kernel APIC v3 (kernel side)

[Avi Kivity]

Gregory Haskins wrote:
> The load-average on my system is about 1 while XP is idling.  qemu seems to be mostly at "0%" but will bounce up to 1% on occasion.  Here is the output of "top -b -p <qemu-pid>" over a few seconds:
>
> top - 09:17:45 up 16:58,  3 users,  load average: 1.02, 0.86, 0.42
> Tasks:   1 total,   0 running,   1 sleeping,   0 stopped,   0 zombie
> Cpu(s):  0.1%us,  0.2%sy,  0.0%ni, 99.5%id,  0.1%wa,  0.0%hi,  0.1%si,  0.0%st
> Mem:   3994704k total,  2018980k used,  1975724k free,    70996k buffers
> Swap:  2104472k total,        0k used,  2104472k free,  1148284k cached
>
>   PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND            
> 10359 ghaskins  15   0  598m  83m  75m S    0  2.1   1:37.87 qemu-system-x86    
>   

A good test is to let Windows boot and idle itself, then compare the 
process cpu time under the TIME+ column with model-0 and model-1.

Since the vast majority of exits in the scenario are hitting the tpr, 
I'd be unsurprised if the time if 50% lower or so.

-- 
error compiling committee.c: too many arguments to function


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 00/10] in-kernel APIC v3 (kernel side)

[Gregory Haskins]

>>> On Sun, May 13, 2007 at  8:38 AM, in message <464706BF.6000808-atKUWr5tajBWk0Htik3J/w@public.gmane.org>,
Avi Kivity <avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote: 
> Gregory Haskins wrote:
>> The load- average on my system is about 1 while XP is idling.  qemu seems to 
> be mostly at "0%" but will bounce up to 1% on occasion.  Here is the output 
> of "top - b - p <qemu- pid>" over a few seconds:
>>
>> top -  09:17:45 up 16:58,  3 users,  load average: 1.02, 0.86, 0.42
>> Tasks:   1 total,   0 running,   1 sleeping,   0 stopped,   0 zombie
>> Cpu(s):  0.1%us,  0.2%sy,  0.0%ni, 99.5%id,  0.1%wa,  0.0%hi,  0.1%si,  
> 0.0%st
>> Mem:   3994704k total,  2018980k used,  1975724k free,    70996k buffers
>> Swap:  2104472k total,        0k used,  2104472k free,  1148284k cached
>>
>>   PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND         
>    
>> 10359 ghaskins  15   0  598m  83m  75m S    0  2.1   1:37.87 qemu- system- x86   
>  
>>   
> 
> A good test is to let Windows boot and idle itself, then compare the 
> process cpu time under the TIME+ column with model- 0 and model- 1.

Will do.  One that that I notice that I can't explain yet is as follows:

When I boot windows + level-1, the point at which windows is running in 16 bit real-mode in the very beginning (before the splashscreen comes up), we seem to take a very large number of exits for instruction emulation.  This ends up being a little storm of activity for about 1 second or so.  I am not really sure why this happens with the new code and not with the old.  It doesnt seem to hurt anything other than extra CPU used.  But its weird nonetheless. 

> 
> Since the vast majority of exits in the scenario are hitting the tpr, 
> I'd be unsurprised if the time if 50% lower or so.

Yeah, the new code essentially converts all those TPR exits to lightweight.  Nothing more, nothing less.  I could be crazy, but my perception is the GUI is *much* more responsive because of it, however.  Windows draw very fast and it actually seems usable.  Whereas trunk+ACPI always feels sluggish.  I don't know of a good benchmark to run to see if there really is an improvement, however.





-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 00/10] in-kernel APIC v3 (kernel side)

[Dor Laor]

>>
>> A good test is to let Windows boot and idle itself, then compare the
>> process cpu time under the TIME+ column with model- 0 and model- 1.
>
>Will do.  One that that I notice that I can't explain yet is as
follows:
>
>When I boot windows + level-1, the point at which windows is running in
16
>bit real-mode in the very beginning (before the splashscreen comes up),
we
>seem to take a very large number of exits for instruction emulation.
This
>ends up being a little storm of activity for about 1 second or so.  I
am
>not really sure why this happens with the new code and not with the
old.
>It doesnt seem to hurt anything other than extra CPU used.  But its
weird
>nonetheless.
>
>>
>> Since the vast majority of exits in the scenario are hitting the tpr,
>> I'd be unsurprised if the time if 50% lower or so.
>
>Yeah, the new code essentially converts all those TPR exits to
lightweight.
>Nothing more, nothing less.  I could be crazy, but my perception is the
GUI
>is *much* more responsive because of it, however.  Windows draw very
fast
>and it actually seems usable.  Whereas trunk+ACPI always feels
sluggish.  I
>don't know of a good benchmark to run to see if there really is an
>improvement, however.
>
- Since light exits are much faster then regular exits the performance
change is noticeable :)
- Don't use usb_device tablet option since it use polling and waists
lots of cpu.
- You can use pcmark05 benchmark.

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 00/10] in-kernel APIC v3 (kernel side)

[Avi Kivity]

Gregory Haskins wrote:

  

>> A good test is to let Windows boot and idle itself, then compare the 
>> process cpu time under the TIME+ column with model- 0 and model- 1.
>>     
>
> Will do.  One that that I notice that I can't explain yet is as follows:
>
> When I boot windows + level-1, the point at which windows is running in 16 bit real-mode in the very beginning (before the splashscreen comes up), we seem to take a very large number of exits for instruction emulation.  This ends up being a little storm of activity for about 1 second or so.  I am not really sure why this happens with the new code and not with the old.  It doesnt seem to hurt anything other than extra CPU used.  But its weird nonetheless. 
>
>   

Very strange. Maybe it is a problem with emulating the apic disabled 
mode. Or maybe the initial state of the apic is different between qemu 
and kvm+apic.

>> Since the vast majority of exits in the scenario are hitting the tpr, 
>> I'd be unsurprised if the time if 50% lower or so.
>>     
>
> Yeah, the new code essentially converts all those TPR exits to lightweight.  Nothing more, nothing less.  I could be crazy, but my perception is the GUI is *much* more responsive because of it, however.  Windows draw very fast and it actually seems usable.  Whereas trunk+ACPI always feels sluggish.  I don't know of a good benchmark to run to see if there really is an improvement, however.
>   

Just the cpu time spent during boot would be a good indication.

-- 
error compiling committee.c: too many arguments to function


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 00/10] in-kernel APIC v3 (kernel side)

[Gregory Haskins]

>>> On Thu, May 10, 2007 at  9:07 AM, in message
<64F9B87B6B770947A9F8391472E032160BBA66AF-yEcIvxbTEBqsx+V+t5oei8rau4O3wl8o3fe8/T/H7NteoWH0uzbU5w@public.gmane.org>,
"Dor Laor" <dor.laor-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote: 
>> At this point, I can boot
>>
>>A) 32- bit XP w/ACPI enabled
>>B) 64- bit SLED (2.6.16 based)
>>
>>both with "level- 0" (userspace PIC emulation) and "level- 1" (in- kernel
> PIC
>>emulation).  The performance with XP under level- 1 seems to be a huge
>>improvement over in- trunk performance.
> 
> Impressive! Can you tell how much host cpu is consumed by an idle xp
> guest w acpi?
> Next week I'll try to check it with the pv network if I'll have the
> time.

And in case it matters:

This is git-HEAD + v3 patches on 

Linux novell1 2.6.21.1-smp-debug #7 SMP Mon May 7 19:16:14 EDT 2007 x86_64 x86_64 x86_64 GNU/Linux

on a dual-socket, dual-core 5130 Xeon (Woodcrest)

Also note that I have the kernel built with pretty much every debug option turned on (lockdep, etc).






-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 00/10] in-kernel APIC v3 (kernel side)

[Avi Kivity]

Gregory Haskins wrote:
> I am pleased to announce v3, which builds upon v2 by adding:
>   

While I haven't reviewed all this yet, it look like we can merge this 
early next week.

What remains to be done:
- boot test mainstream (Linux + Windows) guests on (vmx, svm) x (32, 64) 
host x (32, 64) guest
- boot test non-mainstream guests
- measure Windows APIC HAL performance improvement
- measure Linux performance non-regression
- show newer userspace running older with Linux 2.6.22-rc1 and kvm-24 
userspace running with new code
- add non-intrusive external module support

I'll certainly help with the second item, as I've built a considerable 
collection of guests, and possibly with others as time allows.  I guess 
I can help with the last too.

Mainstream target is 2.6.23.

-- 
error compiling committee.c: too many arguments to function


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 00/10] in-kernel APIC v3 (kernel side)

[Gregory Haskins]

>>> On Sun, May 13, 2007 at  9:10 AM, in message <46470E58.2040208-atKUWr5tajBWk0Htik3J/w@public.gmane.org>,
Avi Kivity <avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote: 
> Gregory Haskins wrote:
>> I am pleased to announce v3, which builds upon v2 by adding:
>>   
> 
> While I haven't reviewed all this yet, it look like we can merge this 
> early next week.

Thats great news, thanks!

> 
> What remains to be done:
> -  boot test mainstream (Linux + Windows) guests on (vmx, svm) x (32, 64) 
> host x (32, 64) guest
> -  boot test non- mainstream guests
> -  measure Windows APIC HAL performance improvement
> -  measure Linux performance non- regression
> -  show newer userspace running older with Linux 2.6.22- rc1 and kvm- 24 
> userspace running with new code
> -  add non- intrusive external module support
> 
> I'll certainly help with the second item, as I've built a considerable 
> collection of guests, and possibly with others as time allows.  I guess 
> I can help with the last too.

I can handle the other items, except for SVN testing.  I currently only have a Merom-T7600 and Woodcrest-5130 Intel setup.  I should have some SVN hardware at some point soon, but not sure when (probably a few weeks out at the earliest).  Any help also testing the SVN side would be appreciated.

> 
> Mainstream target is 2.6.23.




-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/

Re: [PATCH 00/10] in-kernel APIC v3 (kernel side)

[Avi Kivity]

Gregory Haskins wrote:
> I can handle the other items, except for SVN testing.  I currently only have a Merom-T7600 and Woodcrest-5130 Intel setup.  I should have some SVN hardware at some point soon, but not sure when (probably a few weeks out at the earliest).  Any help also testing the SVN side would be appreciated.
>   

I can do that.

-- 
error compiling committee.c: too many arguments to function


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/