Merge branch 'kvm-updates/2.6.29' of git://git.kernel.org/pub/scm/linux/kernel/git/avi/kvm

* 'kvm-updates/2.6.29' of git://git.kernel.org/pub/scm/linux/kernel/git/avi/kvm: (140 commits)
  KVM: MMU: handle large host sptes on invlpg/resync
  KVM: Add locking to virtual i8259 interrupt controller
  KVM: MMU: Don't treat a global pte as such if cr4.pge is cleared
  MAINTAINERS: Maintainership changes for kvm/ia64
  KVM: ia64: Fix kvm_arch_vcpu_ioctl_[gs]et_regs()
  KVM: x86: Rework user space NMI injection as KVM_CAP_USER_NMI
  KVM: VMX: Fix pending NMI-vs.-IRQ race for user space irqchip
  KVM: fix handling of ACK from shared guest IRQ
  KVM: MMU: check for present pdptr shadow page in walk_shadow
  KVM: Consolidate userspace memory capability reporting into common code
  KVM: Advertise the bug in memory region destruction as fixed
  KVM: use cpumask_var_t for cpus_hardware_enabled
  KVM: use modern cpumask primitives, no cpumask_t on stack
  KVM: Extract core of kvm_flush_remote_tlbs/kvm_reload_remote_mmus
  KVM: set owner of cpu and vm file operations
  anon_inodes: use fops->owner for module refcount
  x86: KVM guest: kvm_get_tsc_khz: return khz, not lpj
  KVM: MMU: prepopulate the shadow on invlpg
  KVM: MMU: skip global pgtables on sync due to cr3 switch
  KVM: MMU: collapse remote TLB flushes on root sync
  ...

MAINTAINERS

@@ -2542,8 +2542,6 @@ W:	http://kvm.qumranet.com
 S:	Supported
 
 KERNEL VIRTUAL MACHINE For Itanium (KVM/IA64)
-P:	Anthony Xu
-M:	anthony.xu@intel.com
 P:	Xiantao Zhang
 M:	xiantao.zhang@intel.com
 L:	kvm-ia64@vger.kernel.org

arch/ia64/include/asm/kvm.h

@@ -166,8 +166,6 @@ struct saved_vpd {
 };
 
 struct kvm_regs {
-	char *saved_guest;
-	char *saved_stack;
 	struct saved_vpd vpd;
 	/*Arch-regs*/
 	int mp_state;
@@ -200,6 +198,10 @@ struct kvm_regs {
 	unsigned long fp_psr;       /*used for lazy float register */
 	unsigned long saved_gp;
 	/*for phycial  emulation */
+
+	union context saved_guest;
+
+	unsigned long reserved[64];	/* for future use */
 };
 
 struct kvm_sregs {

arch/ia64/include/asm/kvm_host.h

@@ -23,17 +23,6 @@
 #ifndef __ASM_KVM_HOST_H
 #define __ASM_KVM_HOST_H
 
-
-#include <linux/types.h>
-#include <linux/mm.h>
-#include <linux/kvm.h>
-#include <linux/kvm_para.h>
-#include <linux/kvm_types.h>
-
-#include <asm/pal.h>
-#include <asm/sal.h>
-
-#define KVM_MAX_VCPUS 4
 #define KVM_MEMORY_SLOTS 32
 /* memory slots that does not exposed to userspace */
 #define KVM_PRIVATE_MEM_SLOTS 4
@@ -50,70 +39,132 @@
 #define EXIT_REASON_EXTERNAL_INTERRUPT	6
 #define EXIT_REASON_IPI			7
 #define EXIT_REASON_PTC_G		8
+#define EXIT_REASON_DEBUG		20
 
 /*Define vmm address space and vm data space.*/
-#define KVM_VMM_SIZE (16UL<<20)
+#define KVM_VMM_SIZE (__IA64_UL_CONST(16)<<20)
 #define KVM_VMM_SHIFT 24
-#define KVM_VMM_BASE 0xD000000000000000UL
-#define VMM_SIZE (8UL<<20)
+#define KVM_VMM_BASE 0xD000000000000000
+#define VMM_SIZE (__IA64_UL_CONST(8)<<20)
 
 /*
  * Define vm_buffer, used by PAL Services, base address.
- * Note: vmbuffer is in the VMM-BLOCK, the size must be < 8M
+ * Note: vm_buffer is in the VMM-BLOCK, the size must be < 8M
  */
 #define KVM_VM_BUFFER_BASE (KVM_VMM_BASE + VMM_SIZE)
-#define KVM_VM_BUFFER_SIZE (8UL<<20)
-
-/*Define Virtual machine data layout.*/
-#define KVM_VM_DATA_SHIFT  24
-#define KVM_VM_DATA_SIZE (1UL << KVM_VM_DATA_SHIFT)
-#define KVM_VM_DATA_BASE (KVM_VMM_BASE + KVM_VMM_SIZE)
-
-
-#define KVM_P2M_BASE    KVM_VM_DATA_BASE
-#define KVM_P2M_OFS     0
-#define KVM_P2M_SIZE    (8UL << 20)
-
-#define KVM_VHPT_BASE   (KVM_P2M_BASE + KVM_P2M_SIZE)
-#define KVM_VHPT_OFS    KVM_P2M_SIZE
-#define KVM_VHPT_BLOCK_SIZE   (2UL << 20)
-#define VHPT_SHIFT      18
-#define VHPT_SIZE       (1UL << VHPT_SHIFT)
-#define VHPT_NUM_ENTRIES (1<<(VHPT_SHIFT-5))
-
-#define KVM_VTLB_BASE   (KVM_VHPT_BASE+KVM_VHPT_BLOCK_SIZE)
-#define KVM_VTLB_OFS    (KVM_VHPT_OFS+KVM_VHPT_BLOCK_SIZE)
-#define KVM_VTLB_BLOCK_SIZE   (1UL<<20)
-#define VTLB_SHIFT      17
-#define VTLB_SIZE       (1UL<<VTLB_SHIFT)
-#define VTLB_NUM_ENTRIES (1<<(VTLB_SHIFT-5))
-
-#define KVM_VPD_BASE   (KVM_VTLB_BASE+KVM_VTLB_BLOCK_SIZE)
-#define KVM_VPD_OFS    (KVM_VTLB_OFS+KVM_VTLB_BLOCK_SIZE)
-#define KVM_VPD_BLOCK_SIZE   (2UL<<20)
-#define VPD_SHIFT       16
-#define VPD_SIZE        (1UL<<VPD_SHIFT)
-
-#define KVM_VCPU_BASE   (KVM_VPD_BASE+KVM_VPD_BLOCK_SIZE)
-#define KVM_VCPU_OFS    (KVM_VPD_OFS+KVM_VPD_BLOCK_SIZE)
-#define KVM_VCPU_BLOCK_SIZE   (2UL<<20)
-#define VCPU_SHIFT 18
-#define VCPU_SIZE (1UL<<VCPU_SHIFT)
-#define MAX_VCPU_NUM KVM_VCPU_BLOCK_SIZE/VCPU_SIZE
-
-#define KVM_VM_BASE     (KVM_VCPU_BASE+KVM_VCPU_BLOCK_SIZE)
-#define KVM_VM_OFS      (KVM_VCPU_OFS+KVM_VCPU_BLOCK_SIZE)
-#define KVM_VM_BLOCK_SIZE     (1UL<<19)
-
-#define KVM_MEM_DIRTY_LOG_BASE (KVM_VM_BASE+KVM_VM_BLOCK_SIZE)
-#define KVM_MEM_DIRTY_LOG_OFS  (KVM_VM_OFS+KVM_VM_BLOCK_SIZE)
-#define KVM_MEM_DIRTY_LOG_SIZE (1UL<<19)
-
-/* Get vpd, vhpt, tlb, vcpu, base*/
-#define VPD_ADDR(n) (KVM_VPD_BASE+n*VPD_SIZE)
-#define VHPT_ADDR(n) (KVM_VHPT_BASE+n*VHPT_SIZE)
-#define VTLB_ADDR(n) (KVM_VTLB_BASE+n*VTLB_SIZE)
-#define VCPU_ADDR(n) (KVM_VCPU_BASE+n*VCPU_SIZE)
+#define KVM_VM_BUFFER_SIZE (__IA64_UL_CONST(8)<<20)
+
+/*
+ * kvm guest's data area looks as follow:
+ *
+ *            +----------------------+	-------	KVM_VM_DATA_SIZE
+ *	      |	    vcpu[n]'s data   |	 |     ___________________KVM_STK_OFFSET
+ *     	      |			     |	 |    /			  |
+ *     	      |	       ..........    |	 |   /vcpu's struct&stack |
+ *     	      |	       ..........    |	 |  /---------------------|---- 0
+ *	      |	    vcpu[5]'s data   |	 | /	   vpd		  |
+ *	      |	    vcpu[4]'s data   |	 |/-----------------------|
+ *	      |	    vcpu[3]'s data   |	 /	   vtlb		  |
+ *	      |	    vcpu[2]'s data   |	/|------------------------|
+ *	      |	    vcpu[1]'s data   |/  |	   vhpt		  |
+ *	      |	    vcpu[0]'s data   |____________________________|
+ *            +----------------------+	 |
+ *	      |	   memory dirty log  |	 |
+ *            +----------------------+	 |
+ *	      |	   vm's data struct  |	 |
+ *            +----------------------+	 |
+ *	      |			     |	 |
+ *	      |			     |	 |
+ *	      |			     |	 |
+ *	      |			     |	 |
+ *	      |			     |	 |
+ *	      |			     |	 |
+ *	      |			     |	 |
+ *	      |	  vm's p2m table  |	 |
+ *	      |			     |	 |
+ *            |			     |	 |
+ *	      |			     |	 |  |
+ * vm's data->|			     |   |  |
+ *	      +----------------------+ ------- 0
+ * To support large memory, needs to increase the size of p2m.
+ * To support more vcpus, needs to ensure it has enough space to
+ * hold vcpus' data.
+ */
+
+#define KVM_VM_DATA_SHIFT	26
+#define KVM_VM_DATA_SIZE	(__IA64_UL_CONST(1) << KVM_VM_DATA_SHIFT)
+#define KVM_VM_DATA_BASE	(KVM_VMM_BASE + KVM_VM_DATA_SIZE)
+
+#define KVM_P2M_BASE		KVM_VM_DATA_BASE
+#define KVM_P2M_SIZE		(__IA64_UL_CONST(24) << 20)
+
+#define VHPT_SHIFT		16
+#define VHPT_SIZE		(__IA64_UL_CONST(1) << VHPT_SHIFT)
+#define VHPT_NUM_ENTRIES	(__IA64_UL_CONST(1) << (VHPT_SHIFT-5))
+
+#define VTLB_SHIFT		16
+#define VTLB_SIZE		(__IA64_UL_CONST(1) << VTLB_SHIFT)
+#define VTLB_NUM_ENTRIES	(1UL << (VHPT_SHIFT-5))
+
+#define VPD_SHIFT		16
+#define VPD_SIZE		(__IA64_UL_CONST(1) << VPD_SHIFT)
+
+#define VCPU_STRUCT_SHIFT	16
+#define VCPU_STRUCT_SIZE	(__IA64_UL_CONST(1) << VCPU_STRUCT_SHIFT)
+
+#define KVM_STK_OFFSET		VCPU_STRUCT_SIZE
+
+#define KVM_VM_STRUCT_SHIFT	19
+#define KVM_VM_STRUCT_SIZE	(__IA64_UL_CONST(1) << KVM_VM_STRUCT_SHIFT)
+
+#define KVM_MEM_DIRY_LOG_SHIFT	19
+#define KVM_MEM_DIRTY_LOG_SIZE (__IA64_UL_CONST(1) << KVM_MEM_DIRY_LOG_SHIFT)
+
+#ifndef __ASSEMBLY__
+
+/*Define the max vcpus and memory for Guests.*/
+#define KVM_MAX_VCPUS	(KVM_VM_DATA_SIZE - KVM_P2M_SIZE - KVM_VM_STRUCT_SIZE -\
+			KVM_MEM_DIRTY_LOG_SIZE) / sizeof(struct kvm_vcpu_data)
+#define KVM_MAX_MEM_SIZE (KVM_P2M_SIZE >> 3 << PAGE_SHIFT)
+
+#define VMM_LOG_LEN 256
+
+#include <linux/types.h>
+#include <linux/mm.h>
+#include <linux/kvm.h>
+#include <linux/kvm_para.h>
+#include <linux/kvm_types.h>
+
+#include <asm/pal.h>
+#include <asm/sal.h>
+#include <asm/page.h>
+
+struct kvm_vcpu_data {
+	char vcpu_vhpt[VHPT_SIZE];
+	char vcpu_vtlb[VTLB_SIZE];
+	char vcpu_vpd[VPD_SIZE];
+	char vcpu_struct[VCPU_STRUCT_SIZE];
+};
+
+struct kvm_vm_data {
+	char kvm_p2m[KVM_P2M_SIZE];
+	char kvm_vm_struct[KVM_VM_STRUCT_SIZE];
+	char kvm_mem_dirty_log[KVM_MEM_DIRTY_LOG_SIZE];
+	struct kvm_vcpu_data vcpu_data[KVM_MAX_VCPUS];
+};
+
+#define VCPU_BASE(n)	KVM_VM_DATA_BASE + \
+				offsetof(struct kvm_vm_data, vcpu_data[n])
+#define VM_BASE		KVM_VM_DATA_BASE + \
+				offsetof(struct kvm_vm_data, kvm_vm_struct)
+#define KVM_MEM_DIRTY_LOG_BASE	KVM_VM_DATA_BASE + \
+				offsetof(struct kvm_vm_data, kvm_mem_dirty_log)
+
+#define VHPT_BASE(n) (VCPU_BASE(n) + offsetof(struct kvm_vcpu_data, vcpu_vhpt))
+#define VTLB_BASE(n) (VCPU_BASE(n) + offsetof(struct kvm_vcpu_data, vcpu_vtlb))
+#define VPD_BASE(n)  (VCPU_BASE(n) + offsetof(struct kvm_vcpu_data, vcpu_vpd))
+#define VCPU_STRUCT_BASE(n)	(VCPU_BASE(n) + \
+				offsetof(struct kvm_vcpu_data, vcpu_struct))
 
 /*IO section definitions*/
 #define IOREQ_READ      1
@@ -389,6 +440,7 @@ struct kvm_vcpu_arch {
 
 	unsigned long opcode;
 	unsigned long cause;
+	char log_buf[VMM_LOG_LEN];
 	union context host;
 	union context guest;
 };
@@ -403,14 +455,13 @@ struct kvm_sal_data {
 };
 
 struct kvm_arch {
+	spinlock_t dirty_log_lock;
+
 	unsigned long	vm_base;
 	unsigned long	metaphysical_rr0;
 	unsigned long	metaphysical_rr4;
 	unsigned long	vmm_init_rr;
-	unsigned long	vhpt_base;
-	unsigned long	vtlb_base;
-	unsigned long 	vpd_base;
-	spinlock_t dirty_log_lock;
+
 	struct kvm_ioapic *vioapic;
 	struct kvm_vm_stat stat;
 	struct kvm_sal_data rdv_sal_data;
@@ -512,7 +563,7 @@ struct kvm_pt_regs {
 
 static inline struct kvm_pt_regs *vcpu_regs(struct kvm_vcpu *v)
 {
-	return (struct kvm_pt_regs *) ((unsigned long) v + IA64_STK_OFFSET) - 1;
+	return (struct kvm_pt_regs *) ((unsigned long) v + KVM_STK_OFFSET) - 1;
 }
 
 typedef int kvm_vmm_entry(void);
@@ -531,5 +582,6 @@ int kvm_pal_emul(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run);
 void kvm_sal_emul(struct kvm_vcpu *vcpu);
 
 static inline void kvm_inject_nmi(struct kvm_vcpu *vcpu) {}
+#endif /* __ASSEMBLY__*/
 
 #endif

arch/ia64/kvm/Makefile

@@ -60,7 +60,7 @@ obj-$(CONFIG_KVM) += kvm.o
 
 CFLAGS_vcpu.o += -mfixed-range=f2-f5,f12-f127
 kvm-intel-objs = vmm.o vmm_ivt.o trampoline.o vcpu.o optvfault.o mmio.o \
-	vtlb.o process.o
+	vtlb.o process.o kvm_lib.o
 #Add link memcpy and memset to avoid possible structure assignment error
 kvm-intel-objs += memcpy.o memset.o
 obj-$(CONFIG_KVM_INTEL) += kvm-intel.o

arch/ia64/kvm/asm-offsets.c

@@ -24,19 +24,10 @@
 
 #include <linux/autoconf.h>
 #include <linux/kvm_host.h>
+#include <linux/kbuild.h>
 
 #include "vcpu.h"
 
-#define task_struct kvm_vcpu
-
-#define DEFINE(sym, val) \
-	asm volatile("\n->" #sym " (%0) " #val : : "i" (val))
-
-#define BLANK() asm volatile("\n->" : :)
-
-#define OFFSET(_sym, _str, _mem) \
-    DEFINE(_sym, offsetof(_str, _mem));
-
 void foo(void)
 {
 	DEFINE(VMM_TASK_SIZE, sizeof(struct kvm_vcpu));

arch/ia64/kvm/kvm-ia64.c

@@ -180,7 +180,6 @@ int kvm_dev_ioctl_check_extension(long ext)
 
 	switch (ext) {
 	case KVM_CAP_IRQCHIP:
-	case KVM_CAP_USER_MEMORY:
 	case KVM_CAP_MP_STATE:
 
 		r = 1;
@@ -439,7 +438,6 @@ int kvm_emulate_halt(struct kvm_vcpu *vcpu)
 		expires = div64_u64(itc_diff, cyc_per_usec);
 		kt = ktime_set(0, 1000 * expires);
 
-		down_read(&vcpu->kvm->slots_lock);
 		vcpu->arch.ht_active = 1;
 		hrtimer_start(p_ht, kt, HRTIMER_MODE_ABS);
 
@@ -452,7 +450,6 @@ int kvm_emulate_halt(struct kvm_vcpu *vcpu)
 			if (vcpu->arch.mp_state == KVM_MP_STATE_HALTED)
 				vcpu->arch.mp_state =
 					KVM_MP_STATE_RUNNABLE;
-		up_read(&vcpu->kvm->slots_lock);
 
 		if (vcpu->arch.mp_state != KVM_MP_STATE_RUNNABLE)
 			return -EINTR;
@@ -476,6 +473,13 @@ static int handle_external_interrupt(struct kvm_vcpu *vcpu,
 	return 1;
 }
 
+static int handle_vcpu_debug(struct kvm_vcpu *vcpu,
+				struct kvm_run *kvm_run)
+{
+	printk("VMM: %s", vcpu->arch.log_buf);
+	return 1;
+}
+
 static int (*kvm_vti_exit_handlers[])(struct kvm_vcpu *vcpu,
 		struct kvm_run *kvm_run) = {
 	[EXIT_REASON_VM_PANIC]              = handle_vm_error,
@@ -487,6 +491,7 @@ static int (*kvm_vti_exit_handlers[])(struct kvm_vcpu *vcpu,
 	[EXIT_REASON_EXTERNAL_INTERRUPT]    = handle_external_interrupt,
 	[EXIT_REASON_IPI]		    = handle_ipi,
 	[EXIT_REASON_PTC_G]		    = handle_global_purge,
+	[EXIT_REASON_DEBUG]		    = handle_vcpu_debug,
 
 };
 
@@ -698,27 +703,24 @@ out:
 	return r;
 }
 
-/*
- * Allocate 16M memory for every vm to hold its specific data.
- * Its memory map is defined in kvm_host.h.
- */
 static struct kvm *kvm_alloc_kvm(void)
 {
 
 	struct kvm *kvm;
 	uint64_t  vm_base;
 
+	BUG_ON(sizeof(struct kvm) > KVM_VM_STRUCT_SIZE);
+
 	vm_base = __get_free_pages(GFP_KERNEL, get_order(KVM_VM_DATA_SIZE));
 
 	if (!vm_base)
 		return ERR_PTR(-ENOMEM);
-	printk(KERN_DEBUG"kvm: VM data's base Address:0x%lx\n", vm_base);
 
-	/* Zero all pages before use! */
 	memset((void *)vm_base, 0, KVM_VM_DATA_SIZE);
-
-	kvm = (struct kvm *)(vm_base + KVM_VM_OFS);
+	kvm = (struct kvm *)(vm_base +
+			offsetof(struct kvm_vm_data, kvm_vm_struct));
 	kvm->arch.vm_base = vm_base;
+	printk(KERN_DEBUG"kvm: vm's data area:0x%lx\n", vm_base);
 
 	return kvm;
 }
@@ -760,21 +762,12 @@ static void kvm_build_io_pmt(struct kvm *kvm)
 
 static void kvm_init_vm(struct kvm *kvm)
 {
-	long vm_base;
-
 	BUG_ON(!kvm);
 
 	kvm->arch.metaphysical_rr0 = GUEST_PHYSICAL_RR0;
 	kvm->arch.metaphysical_rr4 = GUEST_PHYSICAL_RR4;
 	kvm->arch.vmm_init_rr = VMM_INIT_RR;
 
-	vm_base = kvm->arch.vm_base;
-	if (vm_base) {
-		kvm->arch.vhpt_base = vm_base + KVM_VHPT_OFS;
-		kvm->arch.vtlb_base = vm_base + KVM_VTLB_OFS;
-		kvm->arch.vpd_base  = vm_base + KVM_VPD_OFS;
-	}
-
 	/*
 	 *Fill P2M entries for MMIO/IO ranges
 	 */
@@ -838,9 +831,8 @@ static int kvm_vm_ioctl_set_irqchip(struct kvm *kvm, struct kvm_irqchip *chip)
 
 int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
-	int i;
 	struct vpd *vpd = to_host(vcpu->kvm, vcpu->arch.vpd);
-	int r;
+	int i;
 
 	vcpu_load(vcpu);
 
@@ -857,18 +849,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 
 	vpd->vpr = regs->vpd.vpr;
 
-	r = -EFAULT;
-	r = copy_from_user(&vcpu->arch.guest, regs->saved_guest,
-						sizeof(union context));
-	if (r)
-		goto out;
-	r = copy_from_user(vcpu + 1, regs->saved_stack +
-			sizeof(struct kvm_vcpu),
-			IA64_STK_OFFSET - sizeof(struct kvm_vcpu));
-	if (r)
-		goto out;
-	vcpu->arch.exit_data =
-		((struct kvm_vcpu *)(regs->saved_stack))->arch.exit_data;
+	memcpy(&vcpu->arch.guest, &regs->saved_guest, sizeof(union context));
 
 	RESTORE_REGS(mp_state);
 	RESTORE_REGS(vmm_rr);
@@ -902,9 +883,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	set_bit(KVM_REQ_RESUME, &vcpu->requests);
 
 	vcpu_put(vcpu);
-	r = 0;
-out:
-	return r;
+
+	return 0;
 }
 
 long kvm_arch_vm_ioctl(struct file *filp,
@@ -1166,10 +1146,11 @@ int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
 		/*Set entry address for first run.*/
 		regs->cr_iip = PALE_RESET_ENTRY;
 
-		/*Initilize itc offset for vcpus*/
+		/*Initialize itc offset for vcpus*/
 		itc_offset = 0UL - ia64_getreg(_IA64_REG_AR_ITC);
-		for (i = 0; i < MAX_VCPU_NUM; i++) {
-			v = (struct kvm_vcpu *)((char *)vcpu + VCPU_SIZE * i);
+		for (i = 0; i < KVM_MAX_VCPUS; i++) {
+			v = (struct kvm_vcpu *)((char *)vcpu +
+					sizeof(struct kvm_vcpu_data) * i);
 			v->arch.itc_offset = itc_offset;
 			v->arch.last_itc = 0;
 		}
@@ -1183,7 +1164,7 @@ int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
 	vcpu->arch.apic->vcpu = vcpu;
 
 	p_ctx->gr[1] = 0;
-	p_ctx->gr[12] = (unsigned long)((char *)vmm_vcpu + IA64_STK_OFFSET);
+	p_ctx->gr[12] = (unsigned long)((char *)vmm_vcpu + KVM_STK_OFFSET);
 	p_ctx->gr[13] = (unsigned long)vmm_vcpu;
 	p_ctx->psr = 0x1008522000UL;
 	p_ctx->ar[40] = FPSR_DEFAULT; /*fpsr*/
@@ -1218,12 +1199,12 @@ int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
 	vcpu->arch.hlt_timer.function = hlt_timer_fn;
 
 	vcpu->arch.last_run_cpu = -1;
-	vcpu->arch.vpd = (struct vpd *)VPD_ADDR(vcpu->vcpu_id);
+	vcpu->arch.vpd = (struct vpd *)VPD_BASE(vcpu->vcpu_id);
 	vcpu->arch.vsa_base = kvm_vsa_base;
 	vcpu->arch.__gp = kvm_vmm_gp;
 	vcpu->arch.dirty_log_lock_pa = __pa(&kvm->arch.dirty_log_lock);
-	vcpu->arch.vhpt.hash = (struct thash_data *)VHPT_ADDR(vcpu->vcpu_id);
-	vcpu->arch.vtlb.hash = (struct thash_data *)VTLB_ADDR(vcpu->vcpu_id);
+	vcpu->arch.vhpt.hash = (struct thash_data *)VHPT_BASE(vcpu->vcpu_id);
+	vcpu->arch.vtlb.hash = (struct thash_data *)VTLB_BASE(vcpu->vcpu_id);
 	init_ptce_info(vcpu);
 
 	r = 0;
@@ -1273,12 +1254,22 @@ struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm,
 	int r;
 	int cpu;
 
+	BUG_ON(sizeof(struct kvm_vcpu) > VCPU_STRUCT_SIZE/2);
+
+	r = -EINVAL;
+	if (id >= KVM_MAX_VCPUS) {
+		printk(KERN_ERR"kvm: Can't configure vcpus > %ld",
+				KVM_MAX_VCPUS);
+		goto fail;
+	}
+
 	r = -ENOMEM;
 	if (!vm_base) {
 		printk(KERN_ERR"kvm: Create vcpu[%d] error!\n", id);
 		goto fail;
 	}
-	vcpu = (struct kvm_vcpu *)(vm_base + KVM_VCPU_OFS + VCPU_SIZE * id);
+	vcpu = (struct kvm_vcpu *)(vm_base + offsetof(struct kvm_vm_data,
+					vcpu_data[id].vcpu_struct));
 	vcpu->kvm = kvm;
 
 	cpu = get_cpu();
@@ -1374,9 +1365,9 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
 
 int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
-	int i;
-	int r;
 	struct vpd *vpd = to_host(vcpu->kvm, vcpu->arch.vpd);
+	int i;
+
 	vcpu_load(vcpu);
 
 	for (i = 0; i < 16; i++) {
@@ -1391,14 +1382,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	regs->vpd.vpsr = vpd->vpsr;
 	regs->vpd.vpr = vpd->vpr;
 
-	r = -EFAULT;
-	r = copy_to_user(regs->saved_guest, &vcpu->arch.guest,
-					sizeof(union context));
-	if (r)
-		goto out;
-	r = copy_to_user(regs->saved_stack, (void *)vcpu, IA64_STK_OFFSET);
-	if (r)
-		goto out;
+	memcpy(&regs->saved_guest, &vcpu->arch.guest, sizeof(union context));
+
 	SAVE_REGS(mp_state);
 	SAVE_REGS(vmm_rr);
 	memcpy(regs->itrs, vcpu->arch.itrs, sizeof(struct thash_data) * NITRS);
@@ -1426,10 +1411,9 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	SAVE_REGS(metaphysical_saved_rr4);
 	SAVE_REGS(fp_psr);
 	SAVE_REGS(saved_gp);
+
 	vcpu_put(vcpu);
-	r = 0;
-out:
-	return r;
+	return 0;
 }
 
 void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu)
@@ -1457,6 +1441,9 @@ int kvm_arch_set_memory_region(struct kvm *kvm,
 	struct kvm_memory_slot *memslot = &kvm->memslots[mem->slot];
 	unsigned long base_gfn = memslot->base_gfn;
 
+	if (base_gfn + npages > (KVM_MAX_MEM_SIZE >> PAGE_SHIFT))
+		return -ENOMEM;
+
 	for (i = 0; i < npages; i++) {
 		pfn = gfn_to_pfn(kvm, base_gfn + i);
 		if (!kvm_is_mmio_pfn(pfn)) {
@@ -1631,8 +1618,8 @@ static int kvm_ia64_sync_dirty_log(struct kvm *kvm,
 	struct kvm_memory_slot *memslot;
 	int r, i;
 	long n, base;
-	unsigned long *dirty_bitmap = (unsigned long *)((void *)kvm - KVM_VM_OFS
-					+ KVM_MEM_DIRTY_LOG_OFS);
+	unsigned long *dirty_bitmap = (unsigned long *)(kvm->arch.vm_base +
+			offsetof(struct kvm_vm_data, kvm_mem_dirty_log));
 
 	r = -EINVAL;
 	if (log->slot >= KVM_MEMORY_SLOTS)

arch/ia64/kvm/kvm_lib.c

@@ -0,0 +1,15 @@
+/*
+ * kvm_lib.c: Compile some libraries for kvm-intel module.
+ *
+ *	Just include kernel's library, and disable symbols export.
+ * 	Copyright (C) 2008, Intel Corporation.
+ *  	Xiantao Zhang  (xiantao.zhang@intel.com)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+#undef CONFIG_MODULES
+#include "../../../lib/vsprintf.c"
+#include "../../../lib/ctype.c"

arch/ia64/kvm/kvm_minstate.h

@@ -24,6 +24,8 @@
 #include <asm/asmmacro.h>
 #include <asm/types.h>
 #include <asm/kregs.h>
+#include <asm/kvm_host.h>
+
 #include "asm-offsets.h"
 
 #define KVM_MINSTATE_START_SAVE_MIN	     					\
@@ -33,7 +35,7 @@
 	addl r22 = VMM_RBS_OFFSET,r1;            /* compute base of RBS */	\
 	;;									\
 	lfetch.fault.excl.nt1 [r22];						\
-	addl r1 = IA64_STK_OFFSET-VMM_PT_REGS_SIZE,r1;  /* compute base of memory stack */  \
+	addl r1 = KVM_STK_OFFSET-VMM_PT_REGS_SIZE, r1;  \
 	mov r23 = ar.bspstore;			/* save ar.bspstore */          \
 	;;									\
 	mov ar.bspstore = r22;				/* switch to kernel RBS */\

arch/ia64/kvm/misc.h

@@ -27,7 +27,8 @@
  */
 static inline uint64_t *kvm_host_get_pmt(struct kvm *kvm)
 {
-	return (uint64_t *)(kvm->arch.vm_base + KVM_P2M_OFS);
+	return (uint64_t *)(kvm->arch.vm_base +
+				offsetof(struct kvm_vm_data, kvm_p2m));
 }
 
 static inline void kvm_set_pmt_entry(struct kvm *kvm, gfn_t gfn,

arch/ia64/kvm/mmio.c

@@ -66,31 +66,25 @@ void lsapic_write(struct kvm_vcpu *v, unsigned long addr,
 
 	switch (addr) {
 	case PIB_OFST_INTA:
-		/*panic_domain(NULL, "Undefined write on PIB INTA\n");*/
-		panic_vm(v);
+		panic_vm(v, "Undefined write on PIB INTA\n");
 		break;
 	case PIB_OFST_XTP:
 		if (length == 1) {
 			vlsapic_write_xtp(v, val);
 		} else {
-			/*panic_domain(NULL,
-			"Undefined write on PIB XTP\n");*/
-			panic_vm(v);
+			panic_vm(v, "Undefined write on PIB XTP\n");
 		}
 		break;
 	default:
 		if (PIB_LOW_HALF(addr)) {
-			/*lower half */
+			/*Lower half */
 			if (length != 8)
-				/*panic_domain(NULL,
-				"Can't LHF write with size %ld!\n",
-				length);*/
-				panic_vm(v);
+				panic_vm(v, "Can't LHF write with size %ld!\n",
+						length);
 			else
 				vlsapic_write_ipi(v, addr, val);
-		} else {   /*	upper half
-				printk("IPI-UHF write %lx\n",addr);*/
-			panic_vm(v);
+		} else {   /*Upper half */
+			panic_vm(v, "IPI-UHF write %lx\n", addr);
 		}
 		break;
 	}
@@ -108,22 +102,18 @@ unsigned long lsapic_read(struct kvm_vcpu *v, unsigned long addr,
 		if (length == 1) /* 1 byte load */
 			; /* There is no i8259, there is no INTA access*/
 		else
-			/*panic_domain(NULL,"Undefined read on PIB INTA\n"); */
-			panic_vm(v);
+			panic_vm(v, "Undefined read on PIB INTA\n");
 
 		break;
 	case PIB_OFST_XTP:
 		if (length == 1) {
 			result = VLSAPIC_XTP(v);
-			/* printk("read xtp %lx\n", result); */
 		} else {
-			/*panic_domain(NULL,
-			"Undefined read on PIB XTP\n");*/
-			panic_vm(v);
+			panic_vm(v, "Undefined read on PIB XTP\n");
 		}
 		break;
 	default:
-		panic_vm(v);
+		panic_vm(v, "Undefined addr access for lsapic!\n");
 		break;
 	}
 	return result;
@@ -162,7 +152,7 @@ static void mmio_access(struct kvm_vcpu *vcpu, u64 src_pa, u64 *dest,
 			/* it's necessary to ensure zero extending */
 			*dest = p->u.ioreq.data & (~0UL >> (64-(s*8)));
 	} else
-		panic_vm(vcpu);
+		panic_vm(vcpu, "Unhandled mmio access returned!\n");
 out:
 	local_irq_restore(psr);
 	return ;
@@ -324,7 +314,9 @@ void emulate_io_inst(struct kvm_vcpu *vcpu, u64 padr, u64 ma)
 		return;
 	} else {
 		inst_type = -1;
-		panic_vm(vcpu);
+		panic_vm(vcpu, "Unsupported MMIO access instruction! \
+				Bunld[0]=0x%lx, Bundle[1]=0x%lx\n",
+				bundle.i64[0], bundle.i64[1]);
 	}
 
 	size = 1 << size;
@@ -335,7 +327,7 @@ void emulate_io_inst(struct kvm_vcpu *vcpu, u64 padr, u64 ma)
 		if (inst_type == SL_INTEGER)
 			vcpu_set_gr(vcpu, inst.M1.r1, data, 0);
 		else
-			panic_vm(vcpu);
+			panic_vm(vcpu, "Unsupported instruction type!\n");
 
 	}
 	vcpu_increment_iip(vcpu);

arch/ia64/kvm/process.c

@@ -527,7 +527,8 @@ void reflect_interruption(u64 ifa, u64 isr, u64 iim,
 	vector = vec2off[vec];
 
 	if (!(vpsr & IA64_PSR_IC) && (vector != IA64_DATA_NESTED_TLB_VECTOR)) {
-		panic_vm(vcpu);
+		panic_vm(vcpu, "Interruption with vector :0x%lx occurs "
+						"with psr.ic = 0\n", vector);
 		return;
 	}
 
@@ -586,7 +587,7 @@ static void set_pal_call_result(struct kvm_vcpu *vcpu)
 		vcpu_set_gr(vcpu, 10, p->u.pal_data.ret.v1, 0);
 		vcpu_set_gr(vcpu, 11, p->u.pal_data.ret.v2, 0);
 	} else
-		panic_vm(vcpu);
+		panic_vm(vcpu, "Mis-set for exit reason!\n");
 }
 
 static void set_sal_call_data(struct kvm_vcpu *vcpu)
@@ -614,7 +615,7 @@ static void set_sal_call_result(struct kvm_vcpu *vcpu)
 		vcpu_set_gr(vcpu, 10, p->u.sal_data.ret.r10, 0);
 		vcpu_set_gr(vcpu, 11, p->u.sal_data.ret.r11, 0);
 	} else
-		panic_vm(vcpu);
+		panic_vm(vcpu, "Mis-set for exit reason!\n");
 }
 
 void  kvm_ia64_handle_break(unsigned long ifa, struct kvm_pt_regs *regs,
@@ -680,7 +681,7 @@ static void generate_exirq(struct kvm_vcpu *vcpu)
 	vpsr = VCPU(vcpu, vpsr);
 	isr = vpsr & IA64_PSR_RI;
 	if (!(vpsr & IA64_PSR_IC))
-		panic_vm(vcpu);
+		panic_vm(vcpu, "Trying to inject one IRQ with psr.ic=0\n");
 	reflect_interruption(0, isr, 0, 12, regs); /* EXT IRQ */
 }
 
@@ -941,8 +942,20 @@ static void vcpu_do_resume(struct kvm_vcpu *vcpu)
 	ia64_set_pta(vcpu->arch.vhpt.pta.val);
 }
 
+static void vmm_sanity_check(struct kvm_vcpu *vcpu)
+{
+	struct exit_ctl_data *p = &vcpu->arch.exit_data;
+
+	if (!vmm_sanity && p->exit_reason != EXIT_REASON_DEBUG) {
+		panic_vm(vcpu, "Failed to do vmm sanity check,"
+			"it maybe caused by crashed vmm!!\n\n");
+	}
+}
+
 static void kvm_do_resume_op(struct kvm_vcpu *vcpu)
 {
+	vmm_sanity_check(vcpu); /*Guarantee vcpu runing on healthy vmm!*/
+
 	if (test_and_clear_bit(KVM_REQ_RESUME, &vcpu->requests)) {
 		vcpu_do_resume(vcpu);
 		return;
@@ -968,3 +981,11 @@ void vmm_transition(struct kvm_vcpu *vcpu)
 						1, 0, 0, 0, 0, 0);
 	kvm_do_resume_op(vcpu);
 }
+
+void vmm_panic_handler(u64 vec)
+{
+	struct kvm_vcpu *vcpu = current_vcpu;
+	vmm_sanity = 0;
+	panic_vm(vcpu, "Unexpected interruption occurs in VMM, vector:0x%lx\n",
+			vec2off[vec]);
+}

arch/ia64/kvm/vcpu.c

@@ -816,8 +816,9 @@ static void vcpu_set_itc(struct kvm_vcpu *vcpu, u64 val)
 	unsigned long vitv = VCPU(vcpu, itv);
 
 	if (vcpu->vcpu_id == 0) {
-		for (i = 0; i < MAX_VCPU_NUM; i++) {
-			v = (struct kvm_vcpu *)((char *)vcpu + VCPU_SIZE * i);
+		for (i = 0; i < KVM_MAX_VCPUS; i++) {
+			v = (struct kvm_vcpu *)((char *)vcpu +
+					sizeof(struct kvm_vcpu_data) * i);
 			VMX(v, itc_offset) = itc_offset;
 			VMX(v, last_itc) = 0;
 		}
@@ -1650,7 +1651,8 @@ void vcpu_set_psr(struct kvm_vcpu *vcpu, unsigned long val)
 	 * Otherwise panic
 	 */
 	if (val & (IA64_PSR_PK | IA64_PSR_IS | IA64_PSR_VM))
-		panic_vm(vcpu);
+		panic_vm(vcpu, "Only support guests with vpsr.pk =0 \
+				& vpsr.is=0\n");
 
 	/*
 	 * For those IA64_PSR bits: id/da/dd/ss/ed/ia
@@ -2103,7 +2105,7 @@ void kvm_init_all_rr(struct kvm_vcpu *vcpu)
 
 	if (is_physical_mode(vcpu)) {
 		if (vcpu->arch.mode_flags & GUEST_PHY_EMUL)
-			panic_vm(vcpu);
+			panic_vm(vcpu, "Machine Status conflicts!\n");
 
 		ia64_set_rr((VRN0 << VRN_SHIFT), vcpu->arch.metaphysical_rr0);
 		ia64_dv_serialize_data();
@@ -2152,10 +2154,70 @@ int vmm_entry(void)
 	return 0;
 }
 
-void panic_vm(struct kvm_vcpu *v)
-{
+static void kvm_show_registers(struct kvm_pt_regs *regs)
+{
+	unsigned long ip = regs->cr_iip + ia64_psr(regs)->ri;
+
+	struct kvm_vcpu *vcpu = current_vcpu;
+	if (vcpu != NULL)
+		printk("vcpu 0x%p vcpu %d\n",
+		       vcpu, vcpu->vcpu_id);
+
+	printk("psr : %016lx ifs : %016lx ip  : [<%016lx>]\n",
+	       regs->cr_ipsr, regs->cr_ifs, ip);
+
+	printk("unat: %016lx pfs : %016lx rsc : %016lx\n",
+	       regs->ar_unat, regs->ar_pfs, regs->ar_rsc);
+	printk("rnat: %016lx bspstore: %016lx pr  : %016lx\n",
+	       regs->ar_rnat, regs->ar_bspstore, regs->pr);
+	printk("ldrs: %016lx ccv : %016lx fpsr: %016lx\n",
+	       regs->loadrs, regs->ar_ccv, regs->ar_fpsr);
+	printk("csd : %016lx ssd : %016lx\n", regs->ar_csd, regs->ar_ssd);
+	printk("b0  : %016lx b6  : %016lx b7  : %016lx\n", regs->b0,
+							regs->b6, regs->b7);
+	printk("f6  : %05lx%016lx f7  : %05lx%016lx\n",
+	       regs->f6.u.bits[1], regs->f6.u.bits[0],
+	       regs->f7.u.bits[1], regs->f7.u.bits[0]);
+	printk("f8  : %05lx%016lx f9  : %05lx%016lx\n",
+	       regs->f8.u.bits[1], regs->f8.u.bits[0],
+	       regs->f9.u.bits[1], regs->f9.u.bits[0]);
+	printk("f10 : %05lx%016lx f11 : %05lx%016lx\n",
+	       regs->f10.u.bits[1], regs->f10.u.bits[0],
+	       regs->f11.u.bits[1], regs->f11.u.bits[0]);
+
+	printk("r1  : %016lx r2  : %016lx r3  : %016lx\n", regs->r1,
+							regs->r2, regs->r3);
+	printk("r8  : %016lx r9  : %016lx r10 : %016lx\n", regs->r8,
+							regs->r9, regs->r10);
+	printk("r11 : %016lx r12 : %016lx r13 : %016lx\n", regs->r11,
+							regs->r12, regs->r13);
+	printk("r14 : %016lx r15 : %016lx r16 : %016lx\n", regs->r14,
+							regs->r15, regs->r16);
+	printk("r17 : %016lx r18 : %016lx r19 : %016lx\n", regs->r17,
+							regs->r18, regs->r19);
+	printk("r20 : %016lx r21 : %016lx r22 : %016lx\n", regs->r20,
+							regs->r21, regs->r22);
+	printk("r23 : %016lx r24 : %016lx r25 : %016lx\n", regs->r23,
+							regs->r24, regs->r25);
+	printk("r26 : %016lx r27 : %016lx r28 : %016lx\n", regs->r26,
+							regs->r27, regs->r28);
+	printk("r29 : %016lx r30 : %016lx r31 : %016lx\n", regs->r29,
+							regs->r30, regs->r31);
+
+}
+
+void panic_vm(struct kvm_vcpu *v, const char *fmt, ...)
+{
+	va_list args;
+	char buf[256];
+
+	struct kvm_pt_regs *regs = vcpu_regs(v);
 	struct exit_ctl_data *p = &v->arch.exit_data;
-
+	va_start(args, fmt);
+	vsnprintf(buf, sizeof(buf), fmt, args);
+	va_end(args);
+	printk(buf);
+	kvm_show_registers(regs);
 	p->exit_reason = EXIT_REASON_VM_PANIC;
 	vmm_transition(v);
 	/*Never to return*/

arch/ia64/kvm/vcpu.h

@@ -737,9 +737,12 @@ void kvm_init_vtlb(struct kvm_vcpu *v);
 void kvm_init_vhpt(struct kvm_vcpu *v);
 void thash_init(struct thash_cb *hcb, u64 sz);
 
-void panic_vm(struct kvm_vcpu *v);
+void panic_vm(struct kvm_vcpu *v, const char *fmt, ...);
 
 extern u64 ia64_call_vsa(u64 proc, u64 arg1, u64 arg2, u64 arg3,
 		u64 arg4, u64 arg5, u64 arg6, u64 arg7);
+
+extern long vmm_sanity;
+
 #endif
 #endif	/* __VCPU_H__ */

arch/ia64/kvm/vmm.c

@@ -20,6 +20,7 @@
  */
 
 
+#include<linux/kernel.h>
 #include<linux/module.h>
 #include<asm/fpswa.h>
 
@@ -31,6 +32,8 @@ MODULE_LICENSE("GPL");
 extern char kvm_ia64_ivt;
 extern fpswa_interface_t *vmm_fpswa_interface;
 
+long vmm_sanity = 1;
+
 struct kvm_vmm_info vmm_info = {
 	.module	     = THIS_MODULE,
 	.vmm_entry   = vmm_entry,
@@ -62,5 +65,31 @@ void vmm_spin_unlock(spinlock_t *lock)
 {
 	_vmm_raw_spin_unlock(lock);
 }
+
+static void vcpu_debug_exit(struct kvm_vcpu *vcpu)
+{
+	struct exit_ctl_data *p = &vcpu->arch.exit_data;
+	long psr;
+
+	local_irq_save(psr);
+	p->exit_reason = EXIT_REASON_DEBUG;
+	vmm_transition(vcpu);
+	local_irq_restore(psr);
+}
+
+asmlinkage int printk(const char *fmt, ...)
+{
+	struct kvm_vcpu *vcpu = current_vcpu;
+	va_list args;
+	int r;
+
+	memset(vcpu->arch.log_buf, 0, VMM_LOG_LEN);
+	va_start(args, fmt);
+	r = vsnprintf(vcpu->arch.log_buf, VMM_LOG_LEN, fmt, args);
+	va_end(args);
+	vcpu_debug_exit(vcpu);
+	return r;
+}
+
 module_init(kvm_vmm_init)
 module_exit(kvm_vmm_exit)

arch/ia64/kvm/vtlb.c

@@ -183,8 +183,8 @@ void mark_pages_dirty(struct kvm_vcpu *v, u64 pte, u64 ps)
 	u64 i, dirty_pages = 1;
 	u64 base_gfn = (pte&_PAGE_PPN_MASK) >> PAGE_SHIFT;
 	spinlock_t *lock = __kvm_va(v->arch.dirty_log_lock_pa);
-	void *dirty_bitmap = (void *)v - (KVM_VCPU_OFS + v->vcpu_id * VCPU_SIZE)
-						+ KVM_MEM_DIRTY_LOG_OFS;
+	void *dirty_bitmap = (void *)KVM_MEM_DIRTY_LOG_BASE;
+
 	dirty_pages <<= ps <= PAGE_SHIFT ? 0 : ps - PAGE_SHIFT;
 
 	vmm_spin_lock(lock);

arch/powerpc/include/asm/disassemble.h

@@ -0,0 +1,80 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Copyright IBM Corp. 2008
+ *
+ * Authors: Hollis Blanchard <hollisb@us.ibm.com>
+ */
+
+#ifndef __ASM_PPC_DISASSEMBLE_H__
+#define __ASM_PPC_DISASSEMBLE_H__
+
+#include <linux/types.h>
+
+static inline unsigned int get_op(u32 inst)
+{
+	return inst >> 26;
+}
+
+static inline unsigned int get_xop(u32 inst)
+{
+	return (inst >> 1) & 0x3ff;
+}
+
+static inline unsigned int get_sprn(u32 inst)
+{
+	return ((inst >> 16) & 0x1f) | ((inst >> 6) & 0x3e0);
+}
+
+static inline unsigned int get_dcrn(u32 inst)
+{
+	return ((inst >> 16) & 0x1f) | ((inst >> 6) & 0x3e0);
+}
+
+static inline unsigned int get_rt(u32 inst)
+{
+	return (inst >> 21) & 0x1f;
+}
+
+static inline unsigned int get_rs(u32 inst)
+{
+	return (inst >> 21) & 0x1f;
+}
+
+static inline unsigned int get_ra(u32 inst)
+{
+	return (inst >> 16) & 0x1f;
+}
+
+static inline unsigned int get_rb(u32 inst)
+{
+	return (inst >> 11) & 0x1f;
+}
+
+static inline unsigned int get_rc(u32 inst)
+{
+	return inst & 0x1;
+}
+
+static inline unsigned int get_ws(u32 inst)
+{
+	return (inst >> 11) & 0x1f;
+}
+
+static inline unsigned int get_d(u32 inst)
+{
+	return inst & 0xffff;
+}
+
+#endif /* __ASM_PPC_DISASSEMBLE_H__ */

arch/powerpc/include/asm/kvm_44x.h

@@ -0,0 +1,61 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Copyright IBM Corp. 2008
+ *
+ * Authors: Hollis Blanchard <hollisb@us.ibm.com>
+ */
+
+#ifndef __ASM_44X_H__
+#define __ASM_44X_H__
+
+#include <linux/kvm_host.h>
+
+#define PPC44x_TLB_SIZE 64
+
+/* If the guest is expecting it, this can be as large as we like; we'd just
+ * need to find some way of advertising it. */
+#define KVM44x_GUEST_TLB_SIZE 64
+
+struct kvmppc_44x_shadow_ref {
+	struct page *page;
+	u16 gtlb_index;
+	u8 writeable;
+	u8 tid;
+};
+
+struct kvmppc_vcpu_44x {
+	/* Unmodified copy of the guest's TLB. */
+	struct kvmppc_44x_tlbe guest_tlb[KVM44x_GUEST_TLB_SIZE];
+
+	/* References to guest pages in the hardware TLB. */
+	struct kvmppc_44x_shadow_ref shadow_refs[PPC44x_TLB_SIZE];
+
+	/* State of the shadow TLB at guest context switch time. */
+	struct kvmppc_44x_tlbe shadow_tlb[PPC44x_TLB_SIZE];
+	u8 shadow_tlb_mod[PPC44x_TLB_SIZE];
+
+	struct kvm_vcpu vcpu;
+};
+
+static inline struct kvmppc_vcpu_44x *to_44x(struct kvm_vcpu *vcpu)
+{
+	return container_of(vcpu, struct kvmppc_vcpu_44x, vcpu);
+}
+
+void kvmppc_set_pid(struct kvm_vcpu *vcpu, u32 new_pid);
+void kvmppc_44x_tlb_put(struct kvm_vcpu *vcpu);
+void kvmppc_44x_tlb_load(struct kvm_vcpu *vcpu);
+
+#endif /* __ASM_44X_H__ */

arch/powerpc/include/asm/kvm_host.h

@@ -64,27 +64,58 @@ struct kvm_vcpu_stat {
 	u32 halt_wakeup;
 };
 
-struct tlbe {
+struct kvmppc_44x_tlbe {
 	u32 tid; /* Only the low 8 bits are used. */
 	u32 word0;
 	u32 word1;
 	u32 word2;
 };
 
-struct kvm_arch {
+enum kvm_exit_types {
+	MMIO_EXITS,
+	DCR_EXITS,
+	SIGNAL_EXITS,
+	ITLB_REAL_MISS_EXITS,
+	ITLB_VIRT_MISS_EXITS,
+	DTLB_REAL_MISS_EXITS,
+	DTLB_VIRT_MISS_EXITS,
+	SYSCALL_EXITS,
+	ISI_EXITS,
+	DSI_EXITS,
+	EMULATED_INST_EXITS,
+	EMULATED_MTMSRWE_EXITS,
+	EMULATED_WRTEE_EXITS,
+	EMULATED_MTSPR_EXITS,
+	EMULATED_MFSPR_EXITS,
+	EMULATED_MTMSR_EXITS,
+	EMULATED_MFMSR_EXITS,
+	EMULATED_TLBSX_EXITS,
+	EMULATED_TLBWE_EXITS,
+	EMULATED_RFI_EXITS,
+	DEC_EXITS,
+	EXT_INTR_EXITS,
+	HALT_WAKEUP,
+	USR_PR_INST,
+	FP_UNAVAIL,
+	DEBUG_EXITS,
+	TIMEINGUEST,
+	__NUMBER_OF_KVM_EXIT_TYPES
 };
 
-struct kvm_vcpu_arch {
-	/* Unmodified copy of the guest's TLB. */
-	struct tlbe guest_tlb[PPC44x_TLB_SIZE];
-	/* TLB that's actually used when the guest is running. */
-	struct tlbe shadow_tlb[PPC44x_TLB_SIZE];
-	/* Pages which are referenced in the shadow TLB. */
-	struct page *shadow_pages[PPC44x_TLB_SIZE];
+/* allow access to big endian 32bit upper/lower parts and 64bit var */
+struct kvmppc_exit_timing {
+	union {
+		u64 tv64;
+		struct {
+			u32 tbu, tbl;
+		} tv32;
+	};
+};
 
-	/* Track which TLB entries we've modified in the current exit. */
-	u8 shadow_tlb_mod[PPC44x_TLB_SIZE];
+struct kvm_arch {
+};
 
+struct kvm_vcpu_arch {
 	u32 host_stack;
 	u32 host_pid;
 	u32 host_dbcr0;
@@ -94,32 +125,32 @@ struct kvm_vcpu_arch {
 	u32 host_msr;
 
 	u64 fpr[32];
-	u32 gpr[32];
+	ulong gpr[32];
 
-	u32 pc;
+	ulong pc;
 	u32 cr;
-	u32 ctr;
-	u32 lr;
-	u32 xer;
+	ulong ctr;
+	ulong lr;
+	ulong xer;
 
-	u32 msr;
+	ulong msr;
 	u32 mmucr;
-	u32 sprg0;
-	u32 sprg1;
-	u32 sprg2;
-	u32 sprg3;
-	u32 sprg4;
-	u32 sprg5;
-	u32 sprg6;
-	u32 sprg7;
-	u32 srr0;
-	u32 srr1;
-	u32 csrr0;
-	u32 csrr1;
-	u32 dsrr0;
-	u32 dsrr1;
-	u32 dear;
-	u32 esr;
+	ulong sprg0;
+	ulong sprg1;
+	ulong sprg2;
+	ulong sprg3;
+	ulong sprg4;
+	ulong sprg5;
+	ulong sprg6;
+	ulong sprg7;
+	ulong srr0;
+	ulong srr1;
+	ulong csrr0;
+	ulong csrr1;
+	ulong dsrr0;
+	ulong dsrr1;
+	ulong dear;
+	ulong esr;
 	u32 dec;
 	u32 decar;
 	u32 tbl;
@@ -127,7 +158,7 @@ struct kvm_vcpu_arch {
 	u32 tcr;
 	u32 tsr;
 	u32 ivor[16];
-	u32 ivpr;
+	ulong ivpr;
 	u32 pir;
 
 	u32 shadow_pid;
@@ -140,9 +171,22 @@ struct kvm_vcpu_arch {
 	u32 dbcr0;
 	u32 dbcr1;
 
+#ifdef CONFIG_KVM_EXIT_TIMING
+	struct kvmppc_exit_timing timing_exit;
+	struct kvmppc_exit_timing timing_last_enter;
+	u32 last_exit_type;
+	u32 timing_count_type[__NUMBER_OF_KVM_EXIT_TYPES];
+	u64 timing_sum_duration[__NUMBER_OF_KVM_EXIT_TYPES];
+	u64 timing_sum_quad_duration[__NUMBER_OF_KVM_EXIT_TYPES];
+	u64 timing_min_duration[__NUMBER_OF_KVM_EXIT_TYPES];
+	u64 timing_max_duration[__NUMBER_OF_KVM_EXIT_TYPES];
+	u64 timing_last_exit;
+	struct dentry *debugfs_exit_timing;
+#endif
+
 	u32 last_inst;
-	u32 fault_dear;
-	u32 fault_esr;
+	ulong fault_dear;
+	ulong fault_esr;
 	gpa_t paddr_accessed;
 
 	u8 io_gpr; /* GPR used as IO source/target */

arch/powerpc/include/asm/kvm_ppc.h

@@ -29,11 +29,6 @@
 #include <linux/kvm_types.h>
 #include <linux/kvm_host.h>
 
-struct kvm_tlb {
-	struct tlbe guest_tlb[PPC44x_TLB_SIZE];
-	struct tlbe shadow_tlb[PPC44x_TLB_SIZE];
-};
-
 enum emulation_result {
 	EMULATE_DONE,         /* no further processing */
 	EMULATE_DO_MMIO,      /* kvm_run filled with MMIO request */
@@ -41,9 +36,6 @@ enum emulation_result {
 	EMULATE_FAIL,         /* can't emulate this instruction */
 };
 
-extern const unsigned char exception_priority[];
-extern const unsigned char priority_exception[];
-
 extern int __kvmppc_vcpu_run(struct kvm_run *kvm_run, struct kvm_vcpu *vcpu);
 extern char kvmppc_handlers_start[];
 extern unsigned long kvmppc_handler_len;
@@ -58,51 +50,44 @@ extern int kvmppc_handle_store(struct kvm_run *run, struct kvm_vcpu *vcpu,
 extern int kvmppc_emulate_instruction(struct kvm_run *run,
                                       struct kvm_vcpu *vcpu);
 extern int kvmppc_emulate_mmio(struct kvm_run *run, struct kvm_vcpu *vcpu);
+extern void kvmppc_emulate_dec(struct kvm_vcpu *vcpu);
 
-extern void kvmppc_mmu_map(struct kvm_vcpu *vcpu, u64 gvaddr, gfn_t gfn,
-                           u64 asid, u32 flags);
-extern void kvmppc_mmu_invalidate(struct kvm_vcpu *vcpu, gva_t eaddr,
-                                  gva_t eend, u32 asid);
+extern void kvmppc_mmu_map(struct kvm_vcpu *vcpu, u64 gvaddr, gpa_t gpaddr,
+                           u64 asid, u32 flags, u32 max_bytes,
+                           unsigned int gtlb_idx);
 extern void kvmppc_mmu_priv_switch(struct kvm_vcpu *vcpu, int usermode);
 extern void kvmppc_mmu_switch_pid(struct kvm_vcpu *vcpu, u32 pid);
 
-/* XXX Book E specific */
-extern void kvmppc_tlbe_set_modified(struct kvm_vcpu *vcpu, unsigned int i);
-
-extern void kvmppc_check_and_deliver_interrupts(struct kvm_vcpu *vcpu);
-
-static inline void kvmppc_queue_exception(struct kvm_vcpu *vcpu, int exception)
-{
-	unsigned int priority = exception_priority[exception];
-	set_bit(priority, &vcpu->arch.pending_exceptions);
-}
-
-static inline void kvmppc_clear_exception(struct kvm_vcpu *vcpu, int exception)
-{
-	unsigned int priority = exception_priority[exception];
-	clear_bit(priority, &vcpu->arch.pending_exceptions);
-}
-
-/* Helper function for "full" MSR writes. No need to call this if only EE is
- * changing. */
-static inline void kvmppc_set_msr(struct kvm_vcpu *vcpu, u32 new_msr)
-{
-	if ((new_msr & MSR_PR) != (vcpu->arch.msr & MSR_PR))
-		kvmppc_mmu_priv_switch(vcpu, new_msr & MSR_PR);
-
-	vcpu->arch.msr = new_msr;
-
-	if (vcpu->arch.msr & MSR_WE)
-		kvm_vcpu_block(vcpu);
-}
-
-static inline void kvmppc_set_pid(struct kvm_vcpu *vcpu, u32 new_pid)
-{
-	if (vcpu->arch.pid != new_pid) {
-		vcpu->arch.pid = new_pid;
-		vcpu->arch.swap_pid = 1;
-	}
-}
+/* Core-specific hooks */
+
+extern struct kvm_vcpu *kvmppc_core_vcpu_create(struct kvm *kvm,
+                                                unsigned int id);
+extern void kvmppc_core_vcpu_free(struct kvm_vcpu *vcpu);
+extern int kvmppc_core_vcpu_setup(struct kvm_vcpu *vcpu);
+extern int kvmppc_core_check_processor_compat(void);
+extern int kvmppc_core_vcpu_translate(struct kvm_vcpu *vcpu,
+                                      struct kvm_translation *tr);
+
+extern void kvmppc_core_vcpu_load(struct kvm_vcpu *vcpu, int cpu);
+extern void kvmppc_core_vcpu_put(struct kvm_vcpu *vcpu);
+
+extern void kvmppc_core_load_guest_debugstate(struct kvm_vcpu *vcpu);
+extern void kvmppc_core_load_host_debugstate(struct kvm_vcpu *vcpu);
+
+extern void kvmppc_core_deliver_interrupts(struct kvm_vcpu *vcpu);
+extern int kvmppc_core_pending_dec(struct kvm_vcpu *vcpu);
+extern void kvmppc_core_queue_program(struct kvm_vcpu *vcpu);
+extern void kvmppc_core_queue_dec(struct kvm_vcpu *vcpu);
+extern void kvmppc_core_queue_external(struct kvm_vcpu *vcpu,
+                                       struct kvm_interrupt *irq);
+
+extern int kvmppc_core_emulate_op(struct kvm_run *run, struct kvm_vcpu *vcpu,
+                                  unsigned int op, int *advance);
+extern int kvmppc_core_emulate_mtspr(struct kvm_vcpu *vcpu, int sprn, int rs);
+extern int kvmppc_core_emulate_mfspr(struct kvm_vcpu *vcpu, int sprn, int rt);
+
+extern int kvmppc_booke_init(void);
+extern void kvmppc_booke_exit(void);
 
 extern void kvmppc_core_destroy_mmu(struct kvm_vcpu *vcpu);
 

arch/powerpc/include/asm/mmu-44x.h

@@ -56,6 +56,7 @@
 #ifndef __ASSEMBLY__
 
 extern unsigned int tlb_44x_hwater;
+extern unsigned int tlb_44x_index;
 
 typedef struct {
 	unsigned int	id;

arch/powerpc/kernel/asm-offsets.c

@@ -23,9 +23,6 @@
 #include <linux/mm.h>
 #include <linux/suspend.h>
 #include <linux/hrtimer.h>
-#ifdef CONFIG_KVM
-#include <linux/kvm_host.h>
-#endif
 #ifdef CONFIG_PPC64
 #include <linux/time.h>
 #include <linux/hardirq.h>
@@ -51,6 +48,9 @@
 #ifdef CONFIG_PPC_ISERIES
 #include <asm/iseries/alpaca.h>
 #endif
+#ifdef CONFIG_KVM
+#include <asm/kvm_44x.h>
+#endif
 
 #if defined(CONFIG_BOOKE) || defined(CONFIG_40x)
 #include "head_booke.h"
@@ -357,12 +357,10 @@ int main(void)
 	DEFINE(PTE_SIZE, sizeof(pte_t));
 
 #ifdef CONFIG_KVM
-	DEFINE(TLBE_BYTES, sizeof(struct tlbe));
+	DEFINE(TLBE_BYTES, sizeof(struct kvmppc_44x_tlbe));
 
 	DEFINE(VCPU_HOST_STACK, offsetof(struct kvm_vcpu, arch.host_stack));
 	DEFINE(VCPU_HOST_PID, offsetof(struct kvm_vcpu, arch.host_pid));
-	DEFINE(VCPU_SHADOW_TLB, offsetof(struct kvm_vcpu, arch.shadow_tlb));
-	DEFINE(VCPU_SHADOW_MOD, offsetof(struct kvm_vcpu, arch.shadow_tlb_mod));
 	DEFINE(VCPU_GPRS, offsetof(struct kvm_vcpu, arch.gpr));
 	DEFINE(VCPU_LR, offsetof(struct kvm_vcpu, arch.lr));
 	DEFINE(VCPU_CR, offsetof(struct kvm_vcpu, arch.cr));
@@ -385,5 +383,16 @@ int main(void)
 	DEFINE(PTE_T_LOG2, PTE_T_LOG2);
 #endif
 
+#ifdef CONFIG_KVM_EXIT_TIMING
+	DEFINE(VCPU_TIMING_EXIT_TBU, offsetof(struct kvm_vcpu,
+						arch.timing_exit.tv32.tbu));
+	DEFINE(VCPU_TIMING_EXIT_TBL, offsetof(struct kvm_vcpu,
+						arch.timing_exit.tv32.tbl));
+	DEFINE(VCPU_TIMING_LAST_ENTER_TBU, offsetof(struct kvm_vcpu,
+					arch.timing_last_enter.tv32.tbu));
+	DEFINE(VCPU_TIMING_LAST_ENTER_TBL, offsetof(struct kvm_vcpu,
+					arch.timing_last_enter.tv32.tbl));
+#endif
+
 	return 0;
 }

arch/powerpc/kvm/44x.c

@@ -0,0 +1,228 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Copyright IBM Corp. 2008
+ *
+ * Authors: Hollis Blanchard <hollisb@us.ibm.com>
+ */
+
+#include <linux/kvm_host.h>
+#include <linux/err.h>
+
+#include <asm/reg.h>
+#include <asm/cputable.h>
+#include <asm/tlbflush.h>
+#include <asm/kvm_44x.h>
+#include <asm/kvm_ppc.h>
+
+#include "44x_tlb.h"
+
+/* Note: clearing MSR[DE] just means that the debug interrupt will not be
+ * delivered *immediately*. Instead, it simply sets the appropriate DBSR bits.
+ * If those DBSR bits are still set when MSR[DE] is re-enabled, the interrupt
+ * will be delivered as an "imprecise debug event" (which is indicated by
+ * DBSR[IDE].
+ */
+static void kvm44x_disable_debug_interrupts(void)
+{
+	mtmsr(mfmsr() & ~MSR_DE);
+}
+
+void kvmppc_core_load_host_debugstate(struct kvm_vcpu *vcpu)
+{
+	kvm44x_disable_debug_interrupts();
+
+	mtspr(SPRN_IAC1, vcpu->arch.host_iac[0]);
+	mtspr(SPRN_IAC2, vcpu->arch.host_iac[1]);
+	mtspr(SPRN_IAC3, vcpu->arch.host_iac[2]);
+	mtspr(SPRN_IAC4, vcpu->arch.host_iac[3]);
+	mtspr(SPRN_DBCR1, vcpu->arch.host_dbcr1);
+	mtspr(SPRN_DBCR2, vcpu->arch.host_dbcr2);
+	mtspr(SPRN_DBCR0, vcpu->arch.host_dbcr0);
+	mtmsr(vcpu->arch.host_msr);
+}
+
+void kvmppc_core_load_guest_debugstate(struct kvm_vcpu *vcpu)
+{
+	struct kvm_guest_debug *dbg = &vcpu->guest_debug;
+	u32 dbcr0 = 0;
+
+	vcpu->arch.host_msr = mfmsr();
+	kvm44x_disable_debug_interrupts();
+
+	/* Save host debug register state. */
+	vcpu->arch.host_iac[0] = mfspr(SPRN_IAC1);
+	vcpu->arch.host_iac[1] = mfspr(SPRN_IAC2);
+	vcpu->arch.host_iac[2] = mfspr(SPRN_IAC3);
+	vcpu->arch.host_iac[3] = mfspr(SPRN_IAC4);
+	vcpu->arch.host_dbcr0 = mfspr(SPRN_DBCR0);
+	vcpu->arch.host_dbcr1 = mfspr(SPRN_DBCR1);
+	vcpu->arch.host_dbcr2 = mfspr(SPRN_DBCR2);
+
+	/* set registers up for guest */
+
+	if (dbg->bp[0]) {
+		mtspr(SPRN_IAC1, dbg->bp[0]);
+		dbcr0 |= DBCR0_IAC1 | DBCR0_IDM;
+	}
+	if (dbg->bp[1]) {
+		mtspr(SPRN_IAC2, dbg->bp[1]);
+		dbcr0 |= DBCR0_IAC2 | DBCR0_IDM;
+	}
+	if (dbg->bp[2]) {
+		mtspr(SPRN_IAC3, dbg->bp[2]);
+		dbcr0 |= DBCR0_IAC3 | DBCR0_IDM;
+	}
+	if (dbg->bp[3]) {
+		mtspr(SPRN_IAC4, dbg->bp[3]);
+		dbcr0 |= DBCR0_IAC4 | DBCR0_IDM;
+	}
+
+	mtspr(SPRN_DBCR0, dbcr0);
+	mtspr(SPRN_DBCR1, 0);
+	mtspr(SPRN_DBCR2, 0);
+}
+
+void kvmppc_core_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
+{
+	kvmppc_44x_tlb_load(vcpu);
+}
+
+void kvmppc_core_vcpu_put(struct kvm_vcpu *vcpu)
+{
+	kvmppc_44x_tlb_put(vcpu);
+}
+
+int kvmppc_core_check_processor_compat(void)
+{
+	int r;
+
+	if (strcmp(cur_cpu_spec->platform, "ppc440") == 0)
+		r = 0;
+	else
+		r = -ENOTSUPP;
+
+	return r;
+}
+
+int kvmppc_core_vcpu_setup(struct kvm_vcpu *vcpu)
+{
+	struct kvmppc_vcpu_44x *vcpu_44x = to_44x(vcpu);
+	struct kvmppc_44x_tlbe *tlbe = &vcpu_44x->guest_tlb[0];
+	int i;
+
+	tlbe->tid = 0;
+	tlbe->word0 = PPC44x_TLB_16M | PPC44x_TLB_VALID;
+	tlbe->word1 = 0;
+	tlbe->word2 = PPC44x_TLB_SX | PPC44x_TLB_SW | PPC44x_TLB_SR;
+
+	tlbe++;
+	tlbe->tid = 0;
+	tlbe->word0 = 0xef600000 | PPC44x_TLB_4K | PPC44x_TLB_VALID;
+	tlbe->word1 = 0xef600000;
+	tlbe->word2 = PPC44x_TLB_SX | PPC44x_TLB_SW | PPC44x_TLB_SR
+	              | PPC44x_TLB_I | PPC44x_TLB_G;
+
+	/* Since the guest can directly access the timebase, it must know the
+	 * real timebase frequency. Accordingly, it must see the state of
+	 * CCR1[TCS]. */
+	vcpu->arch.ccr1 = mfspr(SPRN_CCR1);
+
+	for (i = 0; i < ARRAY_SIZE(vcpu_44x->shadow_refs); i++)
+		vcpu_44x->shadow_refs[i].gtlb_index = -1;
+
+	return 0;
+}
+
+/* 'linear_address' is actually an encoding of AS|PID|EADDR . */
+int kvmppc_core_vcpu_translate(struct kvm_vcpu *vcpu,
+                               struct kvm_translation *tr)
+{
+	struct kvmppc_vcpu_44x *vcpu_44x = to_44x(vcpu);
+	struct kvmppc_44x_tlbe *gtlbe;
+	int index;
+	gva_t eaddr;
+	u8 pid;
+	u8 as;
+
+	eaddr = tr->linear_address;
+	pid = (tr->linear_address >> 32) & 0xff;
+	as = (tr->linear_address >> 40) & 0x1;
+
+	index = kvmppc_44x_tlb_index(vcpu, eaddr, pid, as);
+	if (index == -1) {
+		tr->valid = 0;
+		return 0;
+	}
+
+	gtlbe = &vcpu_44x->guest_tlb[index];
+
+	tr->physical_address = tlb_xlate(gtlbe, eaddr);
+	/* XXX what does "writeable" and "usermode" even mean? */
+	tr->valid = 1;
+
+	return 0;
+}
+
+struct kvm_vcpu *kvmppc_core_vcpu_create(struct kvm *kvm, unsigned int id)
+{
+	struct kvmppc_vcpu_44x *vcpu_44x;
+	struct kvm_vcpu *vcpu;
+	int err;
+
+	vcpu_44x = kmem_cache_zalloc(kvm_vcpu_cache, GFP_KERNEL);
+	if (!vcpu_44x) {
+		err = -ENOMEM;
+		goto out;
+	}
+
+	vcpu = &vcpu_44x->vcpu;
+	err = kvm_vcpu_init(vcpu, kvm, id);
+	if (err)
+		goto free_vcpu;
+
+	return vcpu;
+
+free_vcpu:
+	kmem_cache_free(kvm_vcpu_cache, vcpu_44x);
+out:
+	return ERR_PTR(err);
+}
+
+void kvmppc_core_vcpu_free(struct kvm_vcpu *vcpu)
+{
+	struct kvmppc_vcpu_44x *vcpu_44x = to_44x(vcpu);
+
+	kvm_vcpu_uninit(vcpu);
+	kmem_cache_free(kvm_vcpu_cache, vcpu_44x);
+}
+
+static int kvmppc_44x_init(void)
+{
+	int r;
+
+	r = kvmppc_booke_init();
+	if (r)
+		return r;
+
+	return kvm_init(NULL, sizeof(struct kvmppc_vcpu_44x), THIS_MODULE);
+}
+
+static void kvmppc_44x_exit(void)
+{
+	kvmppc_booke_exit();
+}
+
+module_init(kvmppc_44x_init);
+module_exit(kvmppc_44x_exit);

arch/powerpc/kvm/44x_emulate.c

@@ -0,0 +1,371 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Copyright IBM Corp. 2008
+ *
+ * Authors: Hollis Blanchard <hollisb@us.ibm.com>
+ */
+
+#include <asm/kvm_ppc.h>
+#include <asm/dcr.h>
+#include <asm/dcr-regs.h>
+#include <asm/disassemble.h>
+#include <asm/kvm_44x.h>
+#include "timing.h"
+
+#include "booke.h"
+#include "44x_tlb.h"
+
+#define OP_RFI      19
+
+#define XOP_RFI     50
+#define XOP_MFMSR   83
+#define XOP_WRTEE   131
+#define XOP_MTMSR   146
+#define XOP_WRTEEI  163
+#define XOP_MFDCR   323
+#define XOP_MTDCR   451
+#define XOP_TLBSX   914
+#define XOP_ICCCI   966
+#define XOP_TLBWE   978
+
+static void kvmppc_emul_rfi(struct kvm_vcpu *vcpu)
+{
+	vcpu->arch.pc = vcpu->arch.srr0;
+	kvmppc_set_msr(vcpu, vcpu->arch.srr1);
+}
+
+int kvmppc_core_emulate_op(struct kvm_run *run, struct kvm_vcpu *vcpu,
+                           unsigned int inst, int *advance)
+{
+	int emulated = EMULATE_DONE;
+	int dcrn;
+	int ra;
+	int rb;
+	int rc;
+	int rs;
+	int rt;
+	int ws;
+
+	switch (get_op(inst)) {
+	case OP_RFI:
+		switch (get_xop(inst)) {
+		case XOP_RFI:
+			kvmppc_emul_rfi(vcpu);
+			kvmppc_set_exit_type(vcpu, EMULATED_RFI_EXITS);
+			*advance = 0;
+			break;
+
+		default:
+			emulated = EMULATE_FAIL;
+			break;
+		}
+		break;
+
+	case 31:
+		switch (get_xop(inst)) {
+
+		case XOP_MFMSR:
+			rt = get_rt(inst);
+			vcpu->arch.gpr[rt] = vcpu->arch.msr;
+			kvmppc_set_exit_type(vcpu, EMULATED_MFMSR_EXITS);
+			break;
+
+		case XOP_MTMSR:
+			rs = get_rs(inst);
+			kvmppc_set_exit_type(vcpu, EMULATED_MTMSR_EXITS);
+			kvmppc_set_msr(vcpu, vcpu->arch.gpr[rs]);
+			break;
+
+		case XOP_WRTEE:
+			rs = get_rs(inst);
+			vcpu->arch.msr = (vcpu->arch.msr & ~MSR_EE)
+							 | (vcpu->arch.gpr[rs] & MSR_EE);
+			kvmppc_set_exit_type(vcpu, EMULATED_WRTEE_EXITS);
+			break;
+
+		case XOP_WRTEEI:
+			vcpu->arch.msr = (vcpu->arch.msr & ~MSR_EE)
+							 | (inst & MSR_EE);
+			kvmppc_set_exit_type(vcpu, EMULATED_WRTEE_EXITS);
+			break;
+
+		case XOP_MFDCR:
+			dcrn = get_dcrn(inst);
+			rt = get_rt(inst);
+
+			/* The guest may access CPR0 registers to determine the timebase
+			 * frequency, and it must know the real host frequency because it
+			 * can directly access the timebase registers.
+			 *
+			 * It would be possible to emulate those accesses in userspace,
+			 * but userspace can really only figure out the end frequency.
+			 * We could decompose that into the factors that compute it, but
+			 * that's tricky math, and it's easier to just report the real
+			 * CPR0 values.
+			 */
+			switch (dcrn) {
+			case DCRN_CPR0_CONFIG_ADDR:
+				vcpu->arch.gpr[rt] = vcpu->arch.cpr0_cfgaddr;
+				break;
+			case DCRN_CPR0_CONFIG_DATA:
+				local_irq_disable();
+				mtdcr(DCRN_CPR0_CONFIG_ADDR,
+					  vcpu->arch.cpr0_cfgaddr);
+				vcpu->arch.gpr[rt] = mfdcr(DCRN_CPR0_CONFIG_DATA);
+				local_irq_enable();
+				break;
+			default:
+				run->dcr.dcrn = dcrn;
+				run->dcr.data =  0;
+				run->dcr.is_write = 0;
+				vcpu->arch.io_gpr = rt;
+				vcpu->arch.dcr_needed = 1;
+				kvmppc_account_exit(vcpu, DCR_EXITS);
+				emulated = EMULATE_DO_DCR;
+			}
+
+			break;
+
+		case XOP_MTDCR:
+			dcrn = get_dcrn(inst);
+			rs = get_rs(inst);
+
+			/* emulate some access in kernel */
+			switch (dcrn) {
+			case DCRN_CPR0_CONFIG_ADDR:
+				vcpu->arch.cpr0_cfgaddr = vcpu->arch.gpr[rs];
+				break;
+			default:
+				run->dcr.dcrn = dcrn;
+				run->dcr.data = vcpu->arch.gpr[rs];
+				run->dcr.is_write = 1;
+				vcpu->arch.dcr_needed = 1;
+				kvmppc_account_exit(vcpu, DCR_EXITS);
+				emulated = EMULATE_DO_DCR;
+			}
+
+			break;
+
+		case XOP_TLBWE:
+			ra = get_ra(inst);
+			rs = get_rs(inst);
+			ws = get_ws(inst);
+			emulated = kvmppc_44x_emul_tlbwe(vcpu, ra, rs, ws);
+			break;
+
+		case XOP_TLBSX:
+			rt = get_rt(inst);
+			ra = get_ra(inst);
+			rb = get_rb(inst);
+			rc = get_rc(inst);
+			emulated = kvmppc_44x_emul_tlbsx(vcpu, rt, ra, rb, rc);
+			break;
+
+		case XOP_ICCCI:
+			break;
+
+		default:
+			emulated = EMULATE_FAIL;
+		}
+
+		break;
+
+	default:
+		emulated = EMULATE_FAIL;
+	}
+
+	return emulated;
+}
+
+int kvmppc_core_emulate_mtspr(struct kvm_vcpu *vcpu, int sprn, int rs)
+{
+	switch (sprn) {
+	case SPRN_MMUCR:
+		vcpu->arch.mmucr = vcpu->arch.gpr[rs]; break;
+	case SPRN_PID:
+		kvmppc_set_pid(vcpu, vcpu->arch.gpr[rs]); break;
+	case SPRN_CCR0:
+		vcpu->arch.ccr0 = vcpu->arch.gpr[rs]; break;
+	case SPRN_CCR1:
+		vcpu->arch.ccr1 = vcpu->arch.gpr[rs]; break;
+	case SPRN_DEAR:
+		vcpu->arch.dear = vcpu->arch.gpr[rs]; break;
+	case SPRN_ESR:
+		vcpu->arch.esr = vcpu->arch.gpr[rs]; break;
+	case SPRN_DBCR0:
+		vcpu->arch.dbcr0 = vcpu->arch.gpr[rs]; break;
+	case SPRN_DBCR1:
+		vcpu->arch.dbcr1 = vcpu->arch.gpr[rs]; break;
+	case SPRN_TSR:
+		vcpu->arch.tsr &= ~vcpu->arch.gpr[rs]; break;
+	case SPRN_TCR:
+		vcpu->arch.tcr = vcpu->arch.gpr[rs];
+		kvmppc_emulate_dec(vcpu);
+		break;
+
+	/* Note: SPRG4-7 are user-readable. These values are
+	 * loaded into the real SPRGs when resuming the
+	 * guest. */
+	case SPRN_SPRG4:
+		vcpu->arch.sprg4 = vcpu->arch.gpr[rs]; break;
+	case SPRN_SPRG5:
+		vcpu->arch.sprg5 = vcpu->arch.gpr[rs]; break;
+	case SPRN_SPRG6:
+		vcpu->arch.sprg6 = vcpu->arch.gpr[rs]; break;
+	case SPRN_SPRG7:
+		vcpu->arch.sprg7 = vcpu->arch.gpr[rs]; break;
+
+	case SPRN_IVPR:
+		vcpu->arch.ivpr = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR0:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_CRITICAL] = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR1:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_MACHINE_CHECK] = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR2:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_DATA_STORAGE] = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR3:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_INST_STORAGE] = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR4:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_EXTERNAL] = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR5:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_ALIGNMENT] = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR6:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_PROGRAM] = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR7:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_FP_UNAVAIL] = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR8:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_SYSCALL] = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR9:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_AP_UNAVAIL] = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR10:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_DECREMENTER] = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR11:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_FIT] = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR12:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_WATCHDOG] = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR13:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_DTLB_MISS] = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR14:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_ITLB_MISS] = vcpu->arch.gpr[rs];
+		break;
+	case SPRN_IVOR15:
+		vcpu->arch.ivor[BOOKE_IRQPRIO_DEBUG] = vcpu->arch.gpr[rs];
+		break;
+
+	default:
+		return EMULATE_FAIL;
+	}
+
+	kvmppc_set_exit_type(vcpu, EMULATED_MTSPR_EXITS);
+	return EMULATE_DONE;
+}
+
+int kvmppc_core_emulate_mfspr(struct kvm_vcpu *vcpu, int sprn, int rt)
+{
+	switch (sprn) {
+	/* 440 */
+	case SPRN_MMUCR:
+		vcpu->arch.gpr[rt] = vcpu->arch.mmucr; break;
+	case SPRN_CCR0:
+		vcpu->arch.gpr[rt] = vcpu->arch.ccr0; break;
+	case SPRN_CCR1:
+		vcpu->arch.gpr[rt] = vcpu->arch.ccr1; break;
+
+	/* Book E */
+	case SPRN_PID:
+		vcpu->arch.gpr[rt] = vcpu->arch.pid; break;
+	case SPRN_IVPR:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivpr; break;
+	case SPRN_DEAR:
+		vcpu->arch.gpr[rt] = vcpu->arch.dear; break;
+	case SPRN_ESR:
+		vcpu->arch.gpr[rt] = vcpu->arch.esr; break;
+	case SPRN_DBCR0:
+		vcpu->arch.gpr[rt] = vcpu->arch.dbcr0; break;
+	case SPRN_DBCR1:
+		vcpu->arch.gpr[rt] = vcpu->arch.dbcr1; break;
+
+	case SPRN_IVOR0:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_CRITICAL];
+		break;
+	case SPRN_IVOR1:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_MACHINE_CHECK];
+		break;
+	case SPRN_IVOR2:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_DATA_STORAGE];
+		break;
+	case SPRN_IVOR3:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_INST_STORAGE];
+		break;
+	case SPRN_IVOR4:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_EXTERNAL];
+		break;
+	case SPRN_IVOR5:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_ALIGNMENT];
+		break;
+	case SPRN_IVOR6:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_PROGRAM];
+		break;
+	case SPRN_IVOR7:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_FP_UNAVAIL];
+		break;
+	case SPRN_IVOR8:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_SYSCALL];
+		break;
+	case SPRN_IVOR9:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_AP_UNAVAIL];
+		break;
+	case SPRN_IVOR10:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_DECREMENTER];
+		break;
+	case SPRN_IVOR11:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_FIT];
+		break;
+	case SPRN_IVOR12:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_WATCHDOG];
+		break;
+	case SPRN_IVOR13:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_DTLB_MISS];
+		break;
+	case SPRN_IVOR14:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_ITLB_MISS];
+		break;
+	case SPRN_IVOR15:
+		vcpu->arch.gpr[rt] = vcpu->arch.ivor[BOOKE_IRQPRIO_DEBUG];
+		break;
+
+	default:
+		return EMULATE_FAIL;
+	}
+
+	kvmppc_set_exit_type(vcpu, EMULATED_MFSPR_EXITS);
+	return EMULATE_DONE;
+}
+

arch/powerpc/kvm/44x_tlb.c

@@ -22,20 +22,103 @@
 #include <linux/kvm.h>
 #include <linux/kvm_host.h>
 #include <linux/highmem.h>
+
+#include <asm/tlbflush.h>
 #include <asm/mmu-44x.h>
 #include <asm/kvm_ppc.h>
+#include <asm/kvm_44x.h>
+#include "timing.h"
 
 #include "44x_tlb.h"
 
+#ifndef PPC44x_TLBE_SIZE
+#define PPC44x_TLBE_SIZE	PPC44x_TLB_4K
+#endif
+
+#define PAGE_SIZE_4K (1<<12)
+#define PAGE_MASK_4K (~(PAGE_SIZE_4K - 1))
+
+#define PPC44x_TLB_UATTR_MASK \
+	(PPC44x_TLB_U0|PPC44x_TLB_U1|PPC44x_TLB_U2|PPC44x_TLB_U3)
 #define PPC44x_TLB_USER_PERM_MASK (PPC44x_TLB_UX|PPC44x_TLB_UR|PPC44x_TLB_UW)
 #define PPC44x_TLB_SUPER_PERM_MASK (PPC44x_TLB_SX|PPC44x_TLB_SR|PPC44x_TLB_SW)
 
-static unsigned int kvmppc_tlb_44x_pos;
+#ifdef DEBUG
+void kvmppc_dump_tlbs(struct kvm_vcpu *vcpu)
+{
+	struct kvmppc_44x_tlbe *tlbe;
+	int i;
+
+	printk("vcpu %d TLB dump:\n", vcpu->vcpu_id);
+	printk("| %2s | %3s | %8s | %8s | %8s |\n",
+			"nr", "tid", "word0", "word1", "word2");
+
+	for (i = 0; i < ARRAY_SIZE(vcpu_44x->guest_tlb); i++) {
+		tlbe = &vcpu_44x->guest_tlb[i];
+		if (tlbe->word0 & PPC44x_TLB_VALID)
+			printk(" G%2d |  %02X | %08X | %08X | %08X |\n",
+			       i, tlbe->tid, tlbe->word0, tlbe->word1,
+			       tlbe->word2);
+	}
+}
+#endif
+
+static inline void kvmppc_44x_tlbie(unsigned int index)
+{
+	/* 0 <= index < 64, so the V bit is clear and we can use the index as
+	 * word0. */
+	asm volatile(
+		"tlbwe %[index], %[index], 0\n"
+	:
+	: [index] "r"(index)
+	);
+}
+
+static inline void kvmppc_44x_tlbre(unsigned int index,
+                                    struct kvmppc_44x_tlbe *tlbe)
+{
+	asm volatile(
+		"tlbre %[word0], %[index], 0\n"
+		"mfspr %[tid], %[sprn_mmucr]\n"
+		"andi. %[tid], %[tid], 0xff\n"
+		"tlbre %[word1], %[index], 1\n"
+		"tlbre %[word2], %[index], 2\n"
+		: [word0] "=r"(tlbe->word0),
+		  [word1] "=r"(tlbe->word1),
+		  [word2] "=r"(tlbe->word2),
+		  [tid]   "=r"(tlbe->tid)
+		: [index] "r"(index),
+		  [sprn_mmucr] "i"(SPRN_MMUCR)
+		: "cc"
+	);
+}
+
+static inline void kvmppc_44x_tlbwe(unsigned int index,
+                                    struct kvmppc_44x_tlbe *stlbe)
+{
+	unsigned long tmp;
+
+	asm volatile(
+		"mfspr %[tmp], %[sprn_mmucr]\n"
+		"rlwimi %[tmp], %[tid], 0, 0xff\n"
+		"mtspr %[sprn_mmucr], %[tmp]\n"
+		"tlbwe %[word0], %[index], 0\n"
+		"tlbwe %[word1], %[index], 1\n"
+		"tlbwe %[word2], %[index], 2\n"
+		: [tmp]   "=&r"(tmp)
+		: [word0] "r"(stlbe->word0),
+		  [word1] "r"(stlbe->word1),
+		  [word2] "r"(stlbe->word2),
+		  [tid]   "r"(stlbe->tid),
+		  [index] "r"(index),
+		  [sprn_mmucr] "i"(SPRN_MMUCR)
+	);
+}
 
 static u32 kvmppc_44x_tlb_shadow_attrib(u32 attrib, int usermode)
 {
-	/* Mask off reserved bits. */
-	attrib &= PPC44x_TLB_PERM_MASK|PPC44x_TLB_ATTR_MASK;
+	/* We only care about the guest's permission and user bits. */
+	attrib &= PPC44x_TLB_PERM_MASK|PPC44x_TLB_UATTR_MASK;
 
 	if (!usermode) {
 		/* Guest is in supervisor mode, so we need to translate guest
@@ -47,18 +130,60 @@ static u32 kvmppc_44x_tlb_shadow_attrib(u32 attrib, int usermode)
 	/* Make sure host can always access this memory. */
 	attrib |= PPC44x_TLB_SX|PPC44x_TLB_SR|PPC44x_TLB_SW;
 
+	/* WIMGE = 0b00100 */
+	attrib |= PPC44x_TLB_M;
+
 	return attrib;
 }
 
+/* Load shadow TLB back into hardware. */
+void kvmppc_44x_tlb_load(struct kvm_vcpu *vcpu)
+{
+	struct kvmppc_vcpu_44x *vcpu_44x = to_44x(vcpu);
+	int i;
+
+	for (i = 0; i <= tlb_44x_hwater; i++) {
+		struct kvmppc_44x_tlbe *stlbe = &vcpu_44x->shadow_tlb[i];
+
+		if (get_tlb_v(stlbe) && get_tlb_ts(stlbe))
+			kvmppc_44x_tlbwe(i, stlbe);
+	}
+}
+
+static void kvmppc_44x_tlbe_set_modified(struct kvmppc_vcpu_44x *vcpu_44x,
+                                         unsigned int i)
+{
+	vcpu_44x->shadow_tlb_mod[i] = 1;
+}
+
+/* Save hardware TLB to the vcpu, and invalidate all guest mappings. */
+void kvmppc_44x_tlb_put(struct kvm_vcpu *vcpu)
+{
+	struct kvmppc_vcpu_44x *vcpu_44x = to_44x(vcpu);
+	int i;
+
+	for (i = 0; i <= tlb_44x_hwater; i++) {
+		struct kvmppc_44x_tlbe *stlbe = &vcpu_44x->shadow_tlb[i];
+
+		if (vcpu_44x->shadow_tlb_mod[i])
+			kvmppc_44x_tlbre(i, stlbe);
+
+		if (get_tlb_v(stlbe) && get_tlb_ts(stlbe))
+			kvmppc_44x_tlbie(i);
+	}
+}
+
+
 /* Search the guest TLB for a matching entry. */
 int kvmppc_44x_tlb_index(struct kvm_vcpu *vcpu, gva_t eaddr, unsigned int pid,
                          unsigned int as)
 {
+	struct kvmppc_vcpu_44x *vcpu_44x = to_44x(vcpu);
 	int i;
 
 	/* XXX Replace loop with fancy data structures. */
-	for (i = 0; i < PPC44x_TLB_SIZE; i++) {
-		struct tlbe *tlbe = &vcpu->arch.guest_tlb[i];
+	for (i = 0; i < ARRAY_SIZE(vcpu_44x->guest_tlb); i++) {
+		struct kvmppc_44x_tlbe *tlbe = &vcpu_44x->guest_tlb[i];
 		unsigned int tid;
 
 		if (eaddr < get_tlb_eaddr(tlbe))
@@ -83,78 +208,89 @@ int kvmppc_44x_tlb_index(struct kvm_vcpu *vcpu, gva_t eaddr, unsigned int pid,
 	return -1;
 }
 
-struct tlbe *kvmppc_44x_itlb_search(struct kvm_vcpu *vcpu, gva_t eaddr)
+int kvmppc_44x_itlb_index(struct kvm_vcpu *vcpu, gva_t eaddr)
 {
 	unsigned int as = !!(vcpu->arch.msr & MSR_IS);
-	unsigned int index;
 
-	index = kvmppc_44x_tlb_index(vcpu, eaddr, vcpu->arch.pid, as);
-	if (index == -1)
-		return NULL;
-	return &vcpu->arch.guest_tlb[index];
+	return kvmppc_44x_tlb_index(vcpu, eaddr, vcpu->arch.pid, as);
 }
 
-struct tlbe *kvmppc_44x_dtlb_search(struct kvm_vcpu *vcpu, gva_t eaddr)
+int kvmppc_44x_dtlb_index(struct kvm_vcpu *vcpu, gva_t eaddr)
 {
 	unsigned int as = !!(vcpu->arch.msr & MSR_DS);
-	unsigned int index;
 
-	index = kvmppc_44x_tlb_index(vcpu, eaddr, vcpu->arch.pid, as);
-	if (index == -1)
-		return NULL;
-	return &vcpu->arch.guest_tlb[index];
+	return kvmppc_44x_tlb_index(vcpu, eaddr, vcpu->arch.pid, as);
 }
 
-static int kvmppc_44x_tlbe_is_writable(struct tlbe *tlbe)
+static void kvmppc_44x_shadow_release(struct kvmppc_vcpu_44x *vcpu_44x,
+                                      unsigned int stlb_index)
 {
-	return tlbe->word2 & (PPC44x_TLB_SW|PPC44x_TLB_UW);
-}
+	struct kvmppc_44x_shadow_ref *ref = &vcpu_44x->shadow_refs[stlb_index];
 
-static void kvmppc_44x_shadow_release(struct kvm_vcpu *vcpu,
-                                      unsigned int index)
-{
-	struct tlbe *stlbe = &vcpu->arch.shadow_tlb[index];
-	struct page *page = vcpu->arch.shadow_pages[index];
+	if (!ref->page)
+		return;
 
-	if (get_tlb_v(stlbe)) {
-		if (kvmppc_44x_tlbe_is_writable(stlbe))
-			kvm_release_page_dirty(page);
-		else
-			kvm_release_page_clean(page);
-	}
+	/* Discard from the TLB. */
+	/* Note: we could actually invalidate a host mapping, if the host overwrote
+	 * this TLB entry since we inserted a guest mapping. */
+	kvmppc_44x_tlbie(stlb_index);
+
+	/* Now release the page. */
+	if (ref->writeable)
+		kvm_release_page_dirty(ref->page);
+	else
+		kvm_release_page_clean(ref->page);
+
+	ref->page = NULL;
+
+	/* XXX set tlb_44x_index to stlb_index? */
+
+	KVMTRACE_1D(STLB_INVAL, &vcpu_44x->vcpu, stlb_index, handler);
 }
 
 void kvmppc_core_destroy_mmu(struct kvm_vcpu *vcpu)
 {
+	struct kvmppc_vcpu_44x *vcpu_44x = to_44x(vcpu);
 	int i;
 
 	for (i = 0; i <= tlb_44x_hwater; i++)
-		kvmppc_44x_shadow_release(vcpu, i);
-}
-
-void kvmppc_tlbe_set_modified(struct kvm_vcpu *vcpu, unsigned int i)
-{
-    vcpu->arch.shadow_tlb_mod[i] = 1;
+		kvmppc_44x_shadow_release(vcpu_44x, i);
 }
 
-/* Caller must ensure that the specified guest TLB entry is safe to insert into
- * the shadow TLB. */
-void kvmppc_mmu_map(struct kvm_vcpu *vcpu, u64 gvaddr, gfn_t gfn, u64 asid,
-                    u32 flags)
+/**
+ * kvmppc_mmu_map -- create a host mapping for guest memory
+ *
+ * If the guest wanted a larger page than the host supports, only the first
+ * host page is mapped here and the rest are demand faulted.
+ *
+ * If the guest wanted a smaller page than the host page size, we map only the
+ * guest-size page (i.e. not a full host page mapping).
+ *
+ * Caller must ensure that the specified guest TLB entry is safe to insert into
+ * the shadow TLB.
+ */
+void kvmppc_mmu_map(struct kvm_vcpu *vcpu, u64 gvaddr, gpa_t gpaddr, u64 asid,
+                    u32 flags, u32 max_bytes, unsigned int gtlb_index)
 {
+	struct kvmppc_44x_tlbe stlbe;
+	struct kvmppc_vcpu_44x *vcpu_44x = to_44x(vcpu);
+	struct kvmppc_44x_shadow_ref *ref;
 	struct page *new_page;
-	struct tlbe *stlbe;
 	hpa_t hpaddr;
+	gfn_t gfn;
 	unsigned int victim;
 
-	/* Future optimization: don't overwrite the TLB entry containing the
-	 * current PC (or stack?). */
-	victim = kvmppc_tlb_44x_pos++;
-	if (kvmppc_tlb_44x_pos > tlb_44x_hwater)
-		kvmppc_tlb_44x_pos = 0;
-	stlbe = &vcpu->arch.shadow_tlb[victim];
+	/* Select TLB entry to clobber. Indirectly guard against races with the TLB
+	 * miss handler by disabling interrupts. */
+	local_irq_disable();
+	victim = ++tlb_44x_index;
+	if (victim > tlb_44x_hwater)
+		victim = 0;
+	tlb_44x_index = victim;
+	local_irq_enable();
 
 	/* Get reference to new page. */
+	gfn = gpaddr >> PAGE_SHIFT;
 	new_page = gfn_to_page(vcpu->kvm, gfn);
 	if (is_error_page(new_page)) {
 		printk(KERN_ERR "Couldn't get guest page for gfn %lx!\n", gfn);
@@ -163,10 +299,8 @@ void kvmppc_mmu_map(struct kvm_vcpu *vcpu, u64 gvaddr, gfn_t gfn, u64 asid,
 	}
 	hpaddr = page_to_phys(new_page);
 
-	/* Drop reference to old page. */
-	kvmppc_44x_shadow_release(vcpu, victim);
-
-	vcpu->arch.shadow_pages[victim] = new_page;
+	/* Invalidate any previous shadow mappings. */
+	kvmppc_44x_shadow_release(vcpu_44x, victim);
 
 	/* XXX Make sure (va, size) doesn't overlap any other
 	 * entries. 440x6 user manual says the result would be
@@ -174,78 +308,193 @@ void kvmppc_mmu_map(struct kvm_vcpu *vcpu, u64 gvaddr, gfn_t gfn, u64 asid,
 
 	/* XXX what about AS? */
 
-	stlbe->tid = !(asid & 0xff);
-
 	/* Force TS=1 for all guest mappings. */
-	/* For now we hardcode 4KB mappings, but it will be important to
-	 * use host large pages in the future. */
-	stlbe->word0 = (gvaddr & PAGE_MASK) | PPC44x_TLB_VALID | PPC44x_TLB_TS
-	               | PPC44x_TLB_4K;
-	stlbe->word1 = (hpaddr & 0xfffffc00) | ((hpaddr >> 32) & 0xf);
-	stlbe->word2 = kvmppc_44x_tlb_shadow_attrib(flags,
-	                                            vcpu->arch.msr & MSR_PR);
-	kvmppc_tlbe_set_modified(vcpu, victim);
+	stlbe.word0 = PPC44x_TLB_VALID | PPC44x_TLB_TS;
+
+	if (max_bytes >= PAGE_SIZE) {
+		/* Guest mapping is larger than or equal to host page size. We can use
+		 * a "native" host mapping. */
+		stlbe.word0 |= (gvaddr & PAGE_MASK) | PPC44x_TLBE_SIZE;
+	} else {
+		/* Guest mapping is smaller than host page size. We must restrict the
+		 * size of the mapping to be at most the smaller of the two, but for
+		 * simplicity we fall back to a 4K mapping (this is probably what the
+		 * guest is using anyways). */
+		stlbe.word0 |= (gvaddr & PAGE_MASK_4K) | PPC44x_TLB_4K;
+
+		/* 'hpaddr' is a host page, which is larger than the mapping we're
+		 * inserting here. To compensate, we must add the in-page offset to the
+		 * sub-page. */
+		hpaddr |= gpaddr & (PAGE_MASK ^ PAGE_MASK_4K);
+	}
 
-	KVMTRACE_5D(STLB_WRITE, vcpu, victim,
-			stlbe->tid, stlbe->word0, stlbe->word1, stlbe->word2,
-			handler);
+	stlbe.word1 = (hpaddr & 0xfffffc00) | ((hpaddr >> 32) & 0xf);
+	stlbe.word2 = kvmppc_44x_tlb_shadow_attrib(flags,
+	                                            vcpu->arch.msr & MSR_PR);
+	stlbe.tid = !(asid & 0xff);
+
+	/* Keep track of the reference so we can properly release it later. */
+	ref = &vcpu_44x->shadow_refs[victim];
+	ref->page = new_page;
+	ref->gtlb_index = gtlb_index;
+	ref->writeable = !!(stlbe.word2 & PPC44x_TLB_UW);
+	ref->tid = stlbe.tid;
+
+	/* Insert shadow mapping into hardware TLB. */
+	kvmppc_44x_tlbe_set_modified(vcpu_44x, victim);
+	kvmppc_44x_tlbwe(victim, &stlbe);
+	KVMTRACE_5D(STLB_WRITE, vcpu, victim, stlbe.tid, stlbe.word0, stlbe.word1,
+	            stlbe.word2, handler);
 }
 
-void kvmppc_mmu_invalidate(struct kvm_vcpu *vcpu, gva_t eaddr,
-                           gva_t eend, u32 asid)
+/* For a particular guest TLB entry, invalidate the corresponding host TLB
+ * mappings and release the host pages. */
+static void kvmppc_44x_invalidate(struct kvm_vcpu *vcpu,
+                                  unsigned int gtlb_index)
 {
-	unsigned int pid = !(asid & 0xff);
+	struct kvmppc_vcpu_44x *vcpu_44x = to_44x(vcpu);
 	int i;
 
-	/* XXX Replace loop with fancy data structures. */
-	for (i = 0; i <= tlb_44x_hwater; i++) {
-		struct tlbe *stlbe = &vcpu->arch.shadow_tlb[i];
-		unsigned int tid;
+	for (i = 0; i < ARRAY_SIZE(vcpu_44x->shadow_refs); i++) {
+		struct kvmppc_44x_shadow_ref *ref = &vcpu_44x->shadow_refs[i];
+		if (ref->gtlb_index == gtlb_index)
+			kvmppc_44x_shadow_release(vcpu_44x, i);
+	}
+}
 
-		if (!get_tlb_v(stlbe))
-			continue;
+void kvmppc_mmu_priv_switch(struct kvm_vcpu *vcpu, int usermode)
+{
+	vcpu->arch.shadow_pid = !usermode;
+}
 
-		if (eend < get_tlb_eaddr(stlbe))
-			continue;
+void kvmppc_set_pid(struct kvm_vcpu *vcpu, u32 new_pid)
+{
+	struct kvmppc_vcpu_44x *vcpu_44x = to_44x(vcpu);
+	int i;
 
-		if (eaddr > get_tlb_end(stlbe))
-			continue;
+	if (unlikely(vcpu->arch.pid == new_pid))
+		return;
 
-		tid = get_tlb_tid(stlbe);
-		if (tid && (tid != pid))
-			continue;
+	vcpu->arch.pid = new_pid;
 
-		kvmppc_44x_shadow_release(vcpu, i);
-		stlbe->word0 = 0;
-		kvmppc_tlbe_set_modified(vcpu, i);
-		KVMTRACE_5D(STLB_INVAL, vcpu, i,
-				stlbe->tid, stlbe->word0, stlbe->word1,
-				stlbe->word2, handler);
+	/* Guest userspace runs with TID=0 mappings and PID=0, to make sure it
+	 * can't access guest kernel mappings (TID=1). When we switch to a new
+	 * guest PID, which will also use host PID=0, we must discard the old guest
+	 * userspace mappings. */
+	for (i = 0; i < ARRAY_SIZE(vcpu_44x->shadow_refs); i++) {
+		struct kvmppc_44x_shadow_ref *ref = &vcpu_44x->shadow_refs[i];
+
+		if (ref->tid == 0)
+			kvmppc_44x_shadow_release(vcpu_44x, i);
 	}
 }
 
-/* Invalidate all mappings on the privilege switch after PID has been changed.
- * The guest always runs with PID=1, so we must clear the entire TLB when
- * switching address spaces. */
-void kvmppc_mmu_priv_switch(struct kvm_vcpu *vcpu, int usermode)
+static int tlbe_is_host_safe(const struct kvm_vcpu *vcpu,
+                             const struct kvmppc_44x_tlbe *tlbe)
 {
-	int i;
+	gpa_t gpa;
 
-	if (vcpu->arch.swap_pid) {
-		/* XXX Replace loop with fancy data structures. */
-		for (i = 0; i <= tlb_44x_hwater; i++) {
-			struct tlbe *stlbe = &vcpu->arch.shadow_tlb[i];
-
-			/* Future optimization: clear only userspace mappings. */
-			kvmppc_44x_shadow_release(vcpu, i);
-			stlbe->word0 = 0;
-			kvmppc_tlbe_set_modified(vcpu, i);
-			KVMTRACE_5D(STLB_INVAL, vcpu, i,
-			            stlbe->tid, stlbe->word0, stlbe->word1,
-			            stlbe->word2, handler);
-		}
-		vcpu->arch.swap_pid = 0;
+	if (!get_tlb_v(tlbe))
+		return 0;
+
+	/* Does it match current guest AS? */
+	/* XXX what about IS != DS? */
+	if (get_tlb_ts(tlbe) != !!(vcpu->arch.msr & MSR_IS))
+		return 0;
+
+	gpa = get_tlb_raddr(tlbe);
+	if (!gfn_to_memslot(vcpu->kvm, gpa >> PAGE_SHIFT))
+		/* Mapping is not for RAM. */
+		return 0;
+
+	return 1;
+}
+
+int kvmppc_44x_emul_tlbwe(struct kvm_vcpu *vcpu, u8 ra, u8 rs, u8 ws)
+{
+	struct kvmppc_vcpu_44x *vcpu_44x = to_44x(vcpu);
+	struct kvmppc_44x_tlbe *tlbe;
+	unsigned int gtlb_index;
+
+	gtlb_index = vcpu->arch.gpr[ra];
+	if (gtlb_index > KVM44x_GUEST_TLB_SIZE) {
+		printk("%s: index %d\n", __func__, gtlb_index);
+		kvmppc_dump_vcpu(vcpu);
+		return EMULATE_FAIL;
 	}
 
-	vcpu->arch.shadow_pid = !usermode;
+	tlbe = &vcpu_44x->guest_tlb[gtlb_index];
+
+	/* Invalidate shadow mappings for the about-to-be-clobbered TLB entry. */
+	if (tlbe->word0 & PPC44x_TLB_VALID)
+		kvmppc_44x_invalidate(vcpu, gtlb_index);
+
+	switch (ws) {
+	case PPC44x_TLB_PAGEID:
+		tlbe->tid = get_mmucr_stid(vcpu);
+		tlbe->word0 = vcpu->arch.gpr[rs];
+		break;
+
+	case PPC44x_TLB_XLAT:
+		tlbe->word1 = vcpu->arch.gpr[rs];
+		break;
+
+	case PPC44x_TLB_ATTRIB:
+		tlbe->word2 = vcpu->arch.gpr[rs];
+		break;
+
+	default:
+		return EMULATE_FAIL;
+	}
+
+	if (tlbe_is_host_safe(vcpu, tlbe)) {
+		u64 asid;
+		gva_t eaddr;
+		gpa_t gpaddr;
+		u32 flags;
+		u32 bytes;
+
+		eaddr = get_tlb_eaddr(tlbe);
+		gpaddr = get_tlb_raddr(tlbe);
+
+		/* Use the advertised page size to mask effective and real addrs. */
+		bytes = get_tlb_bytes(tlbe);
+		eaddr &= ~(bytes - 1);
+		gpaddr &= ~(bytes - 1);
+
+		asid = (tlbe->word0 & PPC44x_TLB_TS) | tlbe->tid;
+		flags = tlbe->word2 & 0xffff;
+
+		kvmppc_mmu_map(vcpu, eaddr, gpaddr, asid, flags, bytes, gtlb_index);
+	}
+
+	KVMTRACE_5D(GTLB_WRITE, vcpu, gtlb_index, tlbe->tid, tlbe->word0,
+	            tlbe->word1, tlbe->word2, handler);
+
+	kvmppc_set_exit_type(vcpu, EMULATED_TLBWE_EXITS);
+	return EMULATE_DONE;
+}
+
+int kvmppc_44x_emul_tlbsx(struct kvm_vcpu *vcpu, u8 rt, u8 ra, u8 rb, u8 rc)
+{
+	u32 ea;
+	int gtlb_index;
+	unsigned int as = get_mmucr_sts(vcpu);
+	unsigned int pid = get_mmucr_stid(vcpu);
+
+	ea = vcpu->arch.gpr[rb];
+	if (ra)
+		ea += vcpu->arch.gpr[ra];
+
+	gtlb_index = kvmppc_44x_tlb_index(vcpu, ea, pid, as);
+	if (rc) {
+		if (gtlb_index < 0)
+			vcpu->arch.cr &= ~0x20000000;
+		else
+			vcpu->arch.cr |= 0x20000000;
+	}
+	vcpu->arch.gpr[rt] = gtlb_index;
+
+	kvmppc_set_exit_type(vcpu, EMULATED_TLBSX_EXITS);
+	return EMULATE_DONE;
 }

arch/powerpc/kvm/44x_tlb.h

@@ -25,48 +25,52 @@
 
 extern int kvmppc_44x_tlb_index(struct kvm_vcpu *vcpu, gva_t eaddr,
                                 unsigned int pid, unsigned int as);
-extern struct tlbe *kvmppc_44x_dtlb_search(struct kvm_vcpu *vcpu, gva_t eaddr);
-extern struct tlbe *kvmppc_44x_itlb_search(struct kvm_vcpu *vcpu, gva_t eaddr);
+extern int kvmppc_44x_dtlb_index(struct kvm_vcpu *vcpu, gva_t eaddr);
+extern int kvmppc_44x_itlb_index(struct kvm_vcpu *vcpu, gva_t eaddr);
+
+extern int kvmppc_44x_emul_tlbsx(struct kvm_vcpu *vcpu, u8 rt, u8 ra, u8 rb,
+                                 u8 rc);
+extern int kvmppc_44x_emul_tlbwe(struct kvm_vcpu *vcpu, u8 ra, u8 rs, u8 ws);
 
 /* TLB helper functions */
-static inline unsigned int get_tlb_size(const struct tlbe *tlbe)
+static inline unsigned int get_tlb_size(const struct kvmppc_44x_tlbe *tlbe)
 {
 	return (tlbe->word0 >> 4) & 0xf;
 }
 
-static inline gva_t get_tlb_eaddr(const struct tlbe *tlbe)
+static inline gva_t get_tlb_eaddr(const struct kvmppc_44x_tlbe *tlbe)
 {
 	return tlbe->word0 & 0xfffffc00;
 }
 
-static inline gva_t get_tlb_bytes(const struct tlbe *tlbe)
+static inline gva_t get_tlb_bytes(const struct kvmppc_44x_tlbe *tlbe)
 {
 	unsigned int pgsize = get_tlb_size(tlbe);
 	return 1 << 10 << (pgsize << 1);
 }
 
-static inline gva_t get_tlb_end(const struct tlbe *tlbe)
+static inline gva_t get_tlb_end(const struct kvmppc_44x_tlbe *tlbe)
 {
 	return get_tlb_eaddr(tlbe) + get_tlb_bytes(tlbe) - 1;
 }
 
-static inline u64 get_tlb_raddr(const struct tlbe *tlbe)
+static inline u64 get_tlb_raddr(const struct kvmppc_44x_tlbe *tlbe)
 {
 	u64 word1 = tlbe->word1;
 	return ((word1 & 0xf) << 32) | (word1 & 0xfffffc00);
 }
 
-static inline unsigned int get_tlb_tid(const struct tlbe *tlbe)
+static inline unsigned int get_tlb_tid(const struct kvmppc_44x_tlbe *tlbe)
 {
 	return tlbe->tid & 0xff;
 }
 
-static inline unsigned int get_tlb_ts(const struct tlbe *tlbe)
+static inline unsigned int get_tlb_ts(const struct kvmppc_44x_tlbe *tlbe)
 {
 	return (tlbe->word0 >> 8) & 0x1;
 }
 
-static inline unsigned int get_tlb_v(const struct tlbe *tlbe)
+static inline unsigned int get_tlb_v(const struct kvmppc_44x_tlbe *tlbe)
 {
 	return (tlbe->word0 >> 9) & 0x1;
 }
@@ -81,7 +85,7 @@ static inline unsigned int get_mmucr_sts(const struct kvm_vcpu *vcpu)
 	return (vcpu->arch.mmucr >> 16) & 0x1;
 }
 
-static inline gpa_t tlb_xlate(struct tlbe *tlbe, gva_t eaddr)
+static inline gpa_t tlb_xlate(struct kvmppc_44x_tlbe *tlbe, gva_t eaddr)
 {
 	unsigned int pgmask = get_tlb_bytes(tlbe) - 1;
 

arch/powerpc/kvm/Kconfig

@@ -15,27 +15,33 @@ menuconfig VIRTUALIZATION
 if VIRTUALIZATION
 
 config KVM
-	bool "Kernel-based Virtual Machine (KVM) support"
-	depends on 44x && EXPERIMENTAL
+	bool
 	select PREEMPT_NOTIFIERS
 	select ANON_INODES
-	# We can only run on Book E hosts so far
-	select KVM_BOOKE_HOST
+
+config KVM_440
+	bool "KVM support for PowerPC 440 processors"
+	depends on EXPERIMENTAL && 44x
+	select KVM
 	---help---
-	  Support hosting virtualized guest machines. You will also
-	  need to select one or more of the processor modules below.
+	  Support running unmodified 440 guest kernels in virtual machines on
+	  440 host processors.
 
 	  This module provides access to the hardware capabilities through
 	  a character device node named /dev/kvm.
 
 	  If unsure, say N.
 
-config KVM_BOOKE_HOST
-	bool "KVM host support for Book E PowerPC processors"
-	depends on KVM && 44x
+config KVM_EXIT_TIMING
+	bool "Detailed exit timing"
+	depends on KVM
 	---help---
-	  Provides host support for KVM on Book E PowerPC processors. Currently
-	  this works on 440 processors only.
+	  Calculate elapsed time for every exit/enter cycle. A per-vcpu
+	  report is available in debugfs kvm/vm#_vcpu#_timing.
+	  The overhead is relatively small, however it is not recommended for
+	  production environments.
+
+	  If unsure, say N.
 
 config KVM_TRACE
 	bool "KVM trace support"

arch/powerpc/kvm/Makefile

@@ -8,10 +8,16 @@ common-objs-y = $(addprefix ../../../virt/kvm/, kvm_main.o coalesced_mmio.o)
 
 common-objs-$(CONFIG_KVM_TRACE)  += $(addprefix ../../../virt/kvm/, kvm_trace.o)
 
-kvm-objs := $(common-objs-y) powerpc.o emulate.o booke_guest.o
+kvm-objs := $(common-objs-y) powerpc.o emulate.o
+obj-$(CONFIG_KVM_EXIT_TIMING) += timing.o
 obj-$(CONFIG_KVM) += kvm.o
 
 AFLAGS_booke_interrupts.o := -I$(obj)
 
-kvm-booke-host-objs := booke_host.o booke_interrupts.o 44x_tlb.o
-obj-$(CONFIG_KVM_BOOKE_HOST) += kvm-booke-host.o
+kvm-440-objs := \
+	booke.o \
+	booke_interrupts.o \
+	44x.o \
+	44x_tlb.o \
+	44x_emulate.o
+obj-$(CONFIG_KVM_440) += kvm-440.o

arch/powerpc/kvm/booke_guest.c → arch/powerpc/kvm/booke.c

@@ -24,21 +24,26 @@
 #include <linux/module.h>
 #include <linux/vmalloc.h>
 #include <linux/fs.h>
+
 #include <asm/cputable.h>
 #include <asm/uaccess.h>
 #include <asm/kvm_ppc.h>
+#include "timing.h"
+#include <asm/cacheflush.h>
+#include <asm/kvm_44x.h>
 
+#include "booke.h"
 #include "44x_tlb.h"
 
+unsigned long kvmppc_booke_handlers;
+
 #define VM_STAT(x) offsetof(struct kvm, stat.x), KVM_STAT_VM
 #define VCPU_STAT(x) offsetof(struct kvm_vcpu, stat.x), KVM_STAT_VCPU
 
 struct kvm_stats_debugfs_item debugfs_entries[] = {
-	{ "exits",      VCPU_STAT(sum_exits) },
 	{ "mmio",       VCPU_STAT(mmio_exits) },
 	{ "dcr",        VCPU_STAT(dcr_exits) },
 	{ "sig",        VCPU_STAT(signal_exits) },
-	{ "light",      VCPU_STAT(light_exits) },
 	{ "itlb_r",     VCPU_STAT(itlb_real_miss_exits) },
 	{ "itlb_v",     VCPU_STAT(itlb_virt_miss_exits) },
 	{ "dtlb_r",     VCPU_STAT(dtlb_real_miss_exits) },
@@ -53,103 +58,19 @@ struct kvm_stats_debugfs_item debugfs_entries[] = {
 	{ NULL }
 };
 
-static const u32 interrupt_msr_mask[16] = {
-	[BOOKE_INTERRUPT_CRITICAL]      = MSR_ME,
-	[BOOKE_INTERRUPT_MACHINE_CHECK] = 0,
-	[BOOKE_INTERRUPT_DATA_STORAGE]  = MSR_CE|MSR_ME|MSR_DE,
-	[BOOKE_INTERRUPT_INST_STORAGE]  = MSR_CE|MSR_ME|MSR_DE,
-	[BOOKE_INTERRUPT_EXTERNAL]      = MSR_CE|MSR_ME|MSR_DE,
-	[BOOKE_INTERRUPT_ALIGNMENT]     = MSR_CE|MSR_ME|MSR_DE,
-	[BOOKE_INTERRUPT_PROGRAM]       = MSR_CE|MSR_ME|MSR_DE,
-	[BOOKE_INTERRUPT_FP_UNAVAIL]    = MSR_CE|MSR_ME|MSR_DE,
-	[BOOKE_INTERRUPT_SYSCALL]       = MSR_CE|MSR_ME|MSR_DE,
-	[BOOKE_INTERRUPT_AP_UNAVAIL]    = MSR_CE|MSR_ME|MSR_DE,
-	[BOOKE_INTERRUPT_DECREMENTER]   = MSR_CE|MSR_ME|MSR_DE,
-	[BOOKE_INTERRUPT_FIT]           = MSR_CE|MSR_ME|MSR_DE,
-	[BOOKE_INTERRUPT_WATCHDOG]      = MSR_ME,
-	[BOOKE_INTERRUPT_DTLB_MISS]     = MSR_CE|MSR_ME|MSR_DE,
-	[BOOKE_INTERRUPT_ITLB_MISS]     = MSR_CE|MSR_ME|MSR_DE,
-	[BOOKE_INTERRUPT_DEBUG]         = MSR_ME,
-};
-
-const unsigned char exception_priority[] = {
-	[BOOKE_INTERRUPT_DATA_STORAGE] = 0,
-	[BOOKE_INTERRUPT_INST_STORAGE] = 1,
-	[BOOKE_INTERRUPT_ALIGNMENT] = 2,
-	[BOOKE_INTERRUPT_PROGRAM] = 3,
-	[BOOKE_INTERRUPT_FP_UNAVAIL] = 4,
-	[BOOKE_INTERRUPT_SYSCALL] = 5,
-	[BOOKE_INTERRUPT_AP_UNAVAIL] = 6,
-	[BOOKE_INTERRUPT_DTLB_MISS] = 7,
-	[BOOKE_INTERRUPT_ITLB_MISS] = 8,
-	[BOOKE_INTERRUPT_MACHINE_CHECK] = 9,
-	[BOOKE_INTERRUPT_DEBUG] = 10,
-	[BOOKE_INTERRUPT_CRITICAL] = 11,
-	[BOOKE_INTERRUPT_WATCHDOG] = 12,
-	[BOOKE_INTERRUPT_EXTERNAL] = 13,
-	[BOOKE_INTERRUPT_FIT] = 14,
-	[BOOKE_INTERRUPT_DECREMENTER] = 15,
-};
-
-const unsigned char priority_exception[] = {
-	BOOKE_INTERRUPT_DATA_STORAGE,
-	BOOKE_INTERRUPT_INST_STORAGE,
-	BOOKE_INTERRUPT_ALIGNMENT,
-	BOOKE_INTERRUPT_PROGRAM,
-	BOOKE_INTERRUPT_FP_UNAVAIL,
-	BOOKE_INTERRUPT_SYSCALL,
-	BOOKE_INTERRUPT_AP_UNAVAIL,
-	BOOKE_INTERRUPT_DTLB_MISS,
-	BOOKE_INTERRUPT_ITLB_MISS,
-	BOOKE_INTERRUPT_MACHINE_CHECK,
-	BOOKE_INTERRUPT_DEBUG,
-	BOOKE_INTERRUPT_CRITICAL,
-	BOOKE_INTERRUPT_WATCHDOG,
-	BOOKE_INTERRUPT_EXTERNAL,
-	BOOKE_INTERRUPT_FIT,
-	BOOKE_INTERRUPT_DECREMENTER,
-};
-
-
-void kvmppc_dump_tlbs(struct kvm_vcpu *vcpu)
-{
-	struct tlbe *tlbe;
-	int i;
-
-	printk("vcpu %d TLB dump:\n", vcpu->vcpu_id);
-	printk("| %2s | %3s | %8s | %8s | %8s |\n",
-			"nr", "tid", "word0", "word1", "word2");
-
-	for (i = 0; i < PPC44x_TLB_SIZE; i++) {
-		tlbe = &vcpu->arch.guest_tlb[i];
-		if (tlbe->word0 & PPC44x_TLB_VALID)
-			printk(" G%2d |  %02X | %08X | %08X | %08X |\n",
-			       i, tlbe->tid, tlbe->word0, tlbe->word1,
-			       tlbe->word2);
-	}
-
-	for (i = 0; i < PPC44x_TLB_SIZE; i++) {
-		tlbe = &vcpu->arch.shadow_tlb[i];
-		if (tlbe->word0 & PPC44x_TLB_VALID)
-			printk(" S%2d | %02X | %08X | %08X | %08X |\n",
-			       i, tlbe->tid, tlbe->word0, tlbe->word1,
-			       tlbe->word2);
-	}
-}
-
 /* TODO: use vcpu_printf() */
 void kvmppc_dump_vcpu(struct kvm_vcpu *vcpu)
 {
 	int i;
 
-	printk("pc:   %08x msr:  %08x\n", vcpu->arch.pc, vcpu->arch.msr);
-	printk("lr:   %08x ctr:  %08x\n", vcpu->arch.lr, vcpu->arch.ctr);
-	printk("srr0: %08x srr1: %08x\n", vcpu->arch.srr0, vcpu->arch.srr1);
+	printk("pc:   %08lx msr:  %08lx\n", vcpu->arch.pc, vcpu->arch.msr);
+	printk("lr:   %08lx ctr:  %08lx\n", vcpu->arch.lr, vcpu->arch.ctr);
+	printk("srr0: %08lx srr1: %08lx\n", vcpu->arch.srr0, vcpu->arch.srr1);
 
 	printk("exceptions: %08lx\n", vcpu->arch.pending_exceptions);
 
 	for (i = 0; i < 32; i += 4) {
-		printk("gpr%02d: %08x %08x %08x %08x\n", i,
+		printk("gpr%02d: %08lx %08lx %08lx %08lx\n", i,
 		       vcpu->arch.gpr[i],
 		       vcpu->arch.gpr[i+1],
 		       vcpu->arch.gpr[i+2],
@@ -157,69 +78,96 @@ void kvmppc_dump_vcpu(struct kvm_vcpu *vcpu)
 	}
 }
 
-/* Check if we are ready to deliver the interrupt */
-static int kvmppc_can_deliver_interrupt(struct kvm_vcpu *vcpu, int interrupt)
+static void kvmppc_booke_queue_irqprio(struct kvm_vcpu *vcpu,
+                                       unsigned int priority)
 {
-	int r;
+	set_bit(priority, &vcpu->arch.pending_exceptions);
+}
 
-	switch (interrupt) {
-	case BOOKE_INTERRUPT_CRITICAL:
-		r = vcpu->arch.msr & MSR_CE;
-		break;
-	case BOOKE_INTERRUPT_MACHINE_CHECK:
-		r = vcpu->arch.msr & MSR_ME;
-		break;
-	case BOOKE_INTERRUPT_EXTERNAL:
-		r = vcpu->arch.msr & MSR_EE;
+void kvmppc_core_queue_program(struct kvm_vcpu *vcpu)
+{
+	kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_PROGRAM);
+}
+
+void kvmppc_core_queue_dec(struct kvm_vcpu *vcpu)
+{
+	kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_DECREMENTER);
+}
+
+int kvmppc_core_pending_dec(struct kvm_vcpu *vcpu)
+{
+	return test_bit(BOOKE_IRQPRIO_DECREMENTER, &vcpu->arch.pending_exceptions);
+}
+
+void kvmppc_core_queue_external(struct kvm_vcpu *vcpu,
+                                struct kvm_interrupt *irq)
+{
+	kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_EXTERNAL);
+}
+
+/* Deliver the interrupt of the corresponding priority, if possible. */
+static int kvmppc_booke_irqprio_deliver(struct kvm_vcpu *vcpu,
+                                        unsigned int priority)
+{
+	int allowed = 0;
+	ulong msr_mask;
+
+	switch (priority) {
+	case BOOKE_IRQPRIO_PROGRAM:
+	case BOOKE_IRQPRIO_DTLB_MISS:
+	case BOOKE_IRQPRIO_ITLB_MISS:
+	case BOOKE_IRQPRIO_SYSCALL:
+	case BOOKE_IRQPRIO_DATA_STORAGE:
+	case BOOKE_IRQPRIO_INST_STORAGE:
+	case BOOKE_IRQPRIO_FP_UNAVAIL:
+	case BOOKE_IRQPRIO_AP_UNAVAIL:
+	case BOOKE_IRQPRIO_ALIGNMENT:
+		allowed = 1;
+		msr_mask = MSR_CE|MSR_ME|MSR_DE;
 		break;
-	case BOOKE_INTERRUPT_DECREMENTER:
-		r = vcpu->arch.msr & MSR_EE;
+	case BOOKE_IRQPRIO_CRITICAL:
+	case BOOKE_IRQPRIO_WATCHDOG:
+		allowed = vcpu->arch.msr & MSR_CE;
+		msr_mask = MSR_ME;
 		break;
-	case BOOKE_INTERRUPT_FIT:
-		r = vcpu->arch.msr & MSR_EE;
+	case BOOKE_IRQPRIO_MACHINE_CHECK:
+		allowed = vcpu->arch.msr & MSR_ME;
+		msr_mask = 0;
 		break;
-	case BOOKE_INTERRUPT_WATCHDOG:
-		r = vcpu->arch.msr & MSR_CE;
+	case BOOKE_IRQPRIO_EXTERNAL:
+	case BOOKE_IRQPRIO_DECREMENTER:
+	case BOOKE_IRQPRIO_FIT:
+		allowed = vcpu->arch.msr & MSR_EE;
+		msr_mask = MSR_CE|MSR_ME|MSR_DE;
 		break;
-	case BOOKE_INTERRUPT_DEBUG:
-		r = vcpu->arch.msr & MSR_DE;
+	case BOOKE_IRQPRIO_DEBUG:
+		allowed = vcpu->arch.msr & MSR_DE;
+		msr_mask = MSR_ME;
 		break;
-	default:
-		r = 1;
 	}
 
-	return r;
-}
+	if (allowed) {
+		vcpu->arch.srr0 = vcpu->arch.pc;
+		vcpu->arch.srr1 = vcpu->arch.msr;
+		vcpu->arch.pc = vcpu->arch.ivpr | vcpu->arch.ivor[priority];
+		kvmppc_set_msr(vcpu, vcpu->arch.msr & msr_mask);
 
-static void kvmppc_deliver_interrupt(struct kvm_vcpu *vcpu, int interrupt)
-{
-	switch (interrupt) {
-	case BOOKE_INTERRUPT_DECREMENTER:
-		vcpu->arch.tsr |= TSR_DIS;
-		break;
+		clear_bit(priority, &vcpu->arch.pending_exceptions);
 	}
 
-	vcpu->arch.srr0 = vcpu->arch.pc;
-	vcpu->arch.srr1 = vcpu->arch.msr;
-	vcpu->arch.pc = vcpu->arch.ivpr | vcpu->arch.ivor[interrupt];
-	kvmppc_set_msr(vcpu, vcpu->arch.msr & interrupt_msr_mask[interrupt]);
+	return allowed;
 }
 
 /* Check pending exceptions and deliver one, if possible. */
-void kvmppc_check_and_deliver_interrupts(struct kvm_vcpu *vcpu)
+void kvmppc_core_deliver_interrupts(struct kvm_vcpu *vcpu)
 {
 	unsigned long *pending = &vcpu->arch.pending_exceptions;
-	unsigned int exception;
 	unsigned int priority;
 
-	priority = find_first_bit(pending, BITS_PER_BYTE * sizeof(*pending));
+	priority = __ffs(*pending);
 	while (priority <= BOOKE_MAX_INTERRUPT) {
-		exception = priority_exception[priority];
-		if (kvmppc_can_deliver_interrupt(vcpu, exception)) {
-			kvmppc_clear_exception(vcpu, exception);
-			kvmppc_deliver_interrupt(vcpu, exception);
+		if (kvmppc_booke_irqprio_deliver(vcpu, priority))
 			break;
-		}
 
 		priority = find_next_bit(pending,
 		                         BITS_PER_BYTE * sizeof(*pending),
@@ -238,6 +186,9 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 	enum emulation_result er;
 	int r = RESUME_HOST;
 
+	/* update before a new last_exit_type is rewritten */
+	kvmppc_update_timing_stats(vcpu);
+
 	local_irq_enable();
 
 	run->exit_reason = KVM_EXIT_UNKNOWN;
@@ -251,21 +202,19 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 		break;
 
 	case BOOKE_INTERRUPT_EXTERNAL:
+		kvmppc_account_exit(vcpu, EXT_INTR_EXITS);
+		if (need_resched())
+			cond_resched();
+		r = RESUME_GUEST;
+		break;
+
 	case BOOKE_INTERRUPT_DECREMENTER:
 		/* Since we switched IVPR back to the host's value, the host
 		 * handled this interrupt the moment we enabled interrupts.
 		 * Now we just offer it a chance to reschedule the guest. */
-
-		/* XXX At this point the TLB still holds our shadow TLB, so if
-		 * we do reschedule the host will fault over it. Perhaps we
-		 * should politely restore the host's entries to minimize
-		 * misses before ceding control. */
+		kvmppc_account_exit(vcpu, DEC_EXITS);
 		if (need_resched())
 			cond_resched();
-		if (exit_nr == BOOKE_INTERRUPT_DECREMENTER)
-			vcpu->stat.dec_exits++;
-		else
-			vcpu->stat.ext_intr_exits++;
 		r = RESUME_GUEST;
 		break;
 
@@ -274,17 +223,19 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 			/* Program traps generated by user-level software must be handled
 			 * by the guest kernel. */
 			vcpu->arch.esr = vcpu->arch.fault_esr;
-			kvmppc_queue_exception(vcpu, BOOKE_INTERRUPT_PROGRAM);
+			kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_PROGRAM);
 			r = RESUME_GUEST;
+			kvmppc_account_exit(vcpu, USR_PR_INST);
 			break;
 		}
 
 		er = kvmppc_emulate_instruction(run, vcpu);
 		switch (er) {
 		case EMULATE_DONE:
+			/* don't overwrite subtypes, just account kvm_stats */
+			kvmppc_account_exit_stat(vcpu, EMULATED_INST_EXITS);
 			/* Future optimization: only reload non-volatiles if
 			 * they were actually modified by emulation. */
-			vcpu->stat.emulated_inst_exits++;
 			r = RESUME_GUEST_NV;
 			break;
 		case EMULATE_DO_DCR:
@@ -293,7 +244,7 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 			break;
 		case EMULATE_FAIL:
 			/* XXX Deliver Program interrupt to guest. */
-			printk(KERN_CRIT "%s: emulation at %x failed (%08x)\n",
+			printk(KERN_CRIT "%s: emulation at %lx failed (%08x)\n",
 			       __func__, vcpu->arch.pc, vcpu->arch.last_inst);
 			/* For debugging, encode the failing instruction and
 			 * report it to userspace. */
@@ -307,48 +258,53 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 		break;
 
 	case BOOKE_INTERRUPT_FP_UNAVAIL:
-		kvmppc_queue_exception(vcpu, exit_nr);
+		kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_FP_UNAVAIL);
+		kvmppc_account_exit(vcpu, FP_UNAVAIL);
 		r = RESUME_GUEST;
 		break;
 
 	case BOOKE_INTERRUPT_DATA_STORAGE:
 		vcpu->arch.dear = vcpu->arch.fault_dear;
 		vcpu->arch.esr = vcpu->arch.fault_esr;
-		kvmppc_queue_exception(vcpu, exit_nr);
-		vcpu->stat.dsi_exits++;
+		kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_DATA_STORAGE);
+		kvmppc_account_exit(vcpu, DSI_EXITS);
 		r = RESUME_GUEST;
 		break;
 
 	case BOOKE_INTERRUPT_INST_STORAGE:
 		vcpu->arch.esr = vcpu->arch.fault_esr;
-		kvmppc_queue_exception(vcpu, exit_nr);
-		vcpu->stat.isi_exits++;
+		kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_INST_STORAGE);
+		kvmppc_account_exit(vcpu, ISI_EXITS);
 		r = RESUME_GUEST;
 		break;
 
 	case BOOKE_INTERRUPT_SYSCALL:
-		kvmppc_queue_exception(vcpu, exit_nr);
-		vcpu->stat.syscall_exits++;
+		kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_SYSCALL);
+		kvmppc_account_exit(vcpu, SYSCALL_EXITS);
 		r = RESUME_GUEST;
 		break;
 
+	/* XXX move to a 440-specific file. */
 	case BOOKE_INTERRUPT_DTLB_MISS: {
-		struct tlbe *gtlbe;
+		struct kvmppc_vcpu_44x *vcpu_44x = to_44x(vcpu);
+		struct kvmppc_44x_tlbe *gtlbe;
 		unsigned long eaddr = vcpu->arch.fault_dear;
+		int gtlb_index;
 		gfn_t gfn;
 
 		/* Check the guest TLB. */
-		gtlbe = kvmppc_44x_dtlb_search(vcpu, eaddr);
-		if (!gtlbe) {
+		gtlb_index = kvmppc_44x_dtlb_index(vcpu, eaddr);
+		if (gtlb_index < 0) {
 			/* The guest didn't have a mapping for it. */
-			kvmppc_queue_exception(vcpu, exit_nr);
+			kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_DTLB_MISS);
 			vcpu->arch.dear = vcpu->arch.fault_dear;
 			vcpu->arch.esr = vcpu->arch.fault_esr;
-			vcpu->stat.dtlb_real_miss_exits++;
+			kvmppc_account_exit(vcpu, DTLB_REAL_MISS_EXITS);
 			r = RESUME_GUEST;
 			break;
 		}
 
+		gtlbe = &vcpu_44x->guest_tlb[gtlb_index];
 		vcpu->arch.paddr_accessed = tlb_xlate(gtlbe, eaddr);
 		gfn = vcpu->arch.paddr_accessed >> PAGE_SHIFT;
 
@@ -359,38 +315,45 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 			 * b) the guest used a large mapping which we're faking
 			 * Either way, we need to satisfy the fault without
 			 * invoking the guest. */
-			kvmppc_mmu_map(vcpu, eaddr, gfn, gtlbe->tid,
-			               gtlbe->word2);
-			vcpu->stat.dtlb_virt_miss_exits++;
+			kvmppc_mmu_map(vcpu, eaddr, vcpu->arch.paddr_accessed, gtlbe->tid,
+			               gtlbe->word2, get_tlb_bytes(gtlbe), gtlb_index);
+			kvmppc_account_exit(vcpu, DTLB_VIRT_MISS_EXITS);
 			r = RESUME_GUEST;
 		} else {
 			/* Guest has mapped and accessed a page which is not
 			 * actually RAM. */
 			r = kvmppc_emulate_mmio(run, vcpu);
+			kvmppc_account_exit(vcpu, MMIO_EXITS);
 		}
 
 		break;
 	}
 
+	/* XXX move to a 440-specific file. */
 	case BOOKE_INTERRUPT_ITLB_MISS: {
-		struct tlbe *gtlbe;
+		struct kvmppc_vcpu_44x *vcpu_44x = to_44x(vcpu);
+		struct kvmppc_44x_tlbe *gtlbe;
 		unsigned long eaddr = vcpu->arch.pc;
+		gpa_t gpaddr;
 		gfn_t gfn;
+		int gtlb_index;
 
 		r = RESUME_GUEST;
 
 		/* Check the guest TLB. */
-		gtlbe = kvmppc_44x_itlb_search(vcpu, eaddr);
-		if (!gtlbe) {
+		gtlb_index = kvmppc_44x_itlb_index(vcpu, eaddr);
+		if (gtlb_index < 0) {
 			/* The guest didn't have a mapping for it. */
-			kvmppc_queue_exception(vcpu, exit_nr);
-			vcpu->stat.itlb_real_miss_exits++;
+			kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_ITLB_MISS);
+			kvmppc_account_exit(vcpu, ITLB_REAL_MISS_EXITS);
 			break;
 		}
 
-		vcpu->stat.itlb_virt_miss_exits++;
+		kvmppc_account_exit(vcpu, ITLB_VIRT_MISS_EXITS);
 
-		gfn = tlb_xlate(gtlbe, eaddr) >> PAGE_SHIFT;
+		gtlbe = &vcpu_44x->guest_tlb[gtlb_index];
+		gpaddr = tlb_xlate(gtlbe, eaddr);
+		gfn = gpaddr >> PAGE_SHIFT;
 
 		if (kvm_is_visible_gfn(vcpu->kvm, gfn)) {
 			/* The guest TLB had a mapping, but the shadow TLB
@@ -399,12 +362,11 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 			 * b) the guest used a large mapping which we're faking
 			 * Either way, we need to satisfy the fault without
 			 * invoking the guest. */
-			kvmppc_mmu_map(vcpu, eaddr, gfn, gtlbe->tid,
-			               gtlbe->word2);
+			kvmppc_mmu_map(vcpu, eaddr, gpaddr, gtlbe->tid,
+			               gtlbe->word2, get_tlb_bytes(gtlbe), gtlb_index);
 		} else {
 			/* Guest mapped and leaped at non-RAM! */
-			kvmppc_queue_exception(vcpu,
-			                       BOOKE_INTERRUPT_MACHINE_CHECK);
+			kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_MACHINE_CHECK);
 		}
 
 		break;
@@ -421,6 +383,7 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 		mtspr(SPRN_DBSR, dbsr);
 
 		run->exit_reason = KVM_EXIT_DEBUG;
+		kvmppc_account_exit(vcpu, DEBUG_EXITS);
 		r = RESUME_HOST;
 		break;
 	}
@@ -432,10 +395,8 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 
 	local_irq_disable();
 
-	kvmppc_check_and_deliver_interrupts(vcpu);
+	kvmppc_core_deliver_interrupts(vcpu);
 
-	/* Do some exit accounting. */
-	vcpu->stat.sum_exits++;
 	if (!(r & RESUME_HOST)) {
 		/* To avoid clobbering exit_reason, only check for signals if
 		 * we aren't already exiting to userspace for some other
@@ -443,22 +404,7 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 		if (signal_pending(current)) {
 			run->exit_reason = KVM_EXIT_INTR;
 			r = (-EINTR << 2) | RESUME_HOST | (r & RESUME_FLAG_NV);
-
-			vcpu->stat.signal_exits++;
-		} else {
-			vcpu->stat.light_exits++;
-		}
-	} else {
-		switch (run->exit_reason) {
-		case KVM_EXIT_MMIO:
-			vcpu->stat.mmio_exits++;
-			break;
-		case KVM_EXIT_DCR:
-			vcpu->stat.dcr_exits++;
-			break;
-		case KVM_EXIT_INTR:
-			vcpu->stat.signal_exits++;
-			break;
+			kvmppc_account_exit(vcpu, SIGNAL_EXITS);
 		}
 	}
 
@@ -468,20 +414,6 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 /* Initial guest state: 16MB mapping 0 -> 0, PC = 0, MSR = 0, R1 = 16MB */
 int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
 {
-	struct tlbe *tlbe = &vcpu->arch.guest_tlb[0];
-
-	tlbe->tid = 0;
-	tlbe->word0 = PPC44x_TLB_16M | PPC44x_TLB_VALID;
-	tlbe->word1 = 0;
-	tlbe->word2 = PPC44x_TLB_SX | PPC44x_TLB_SW | PPC44x_TLB_SR;
-
-	tlbe++;
-	tlbe->tid = 0;
-	tlbe->word0 = 0xef600000 | PPC44x_TLB_4K | PPC44x_TLB_VALID;
-	tlbe->word1 = 0xef600000;
-	tlbe->word2 = PPC44x_TLB_SX | PPC44x_TLB_SW | PPC44x_TLB_SR
-	              | PPC44x_TLB_I | PPC44x_TLB_G;
-
 	vcpu->arch.pc = 0;
 	vcpu->arch.msr = 0;
 	vcpu->arch.gpr[1] = (16<<20) - 8; /* -8 for the callee-save LR slot */
@@ -492,12 +424,9 @@ int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
 	 * before it's programmed its own IVPR. */
 	vcpu->arch.ivpr = 0x55550000;
 
-	/* Since the guest can directly access the timebase, it must know the
-	 * real timebase frequency. Accordingly, it must see the state of
-	 * CCR1[TCS]. */
-	vcpu->arch.ccr1 = mfspr(SPRN_CCR1);
+	kvmppc_init_timing_stats(vcpu);
 
-	return 0;
+	return kvmppc_core_vcpu_setup(vcpu);
 }
 
 int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
@@ -536,7 +465,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	vcpu->arch.ctr = regs->ctr;
 	vcpu->arch.lr = regs->lr;
 	vcpu->arch.xer = regs->xer;
-	vcpu->arch.msr = regs->msr;
+	kvmppc_set_msr(vcpu, regs->msr);
 	vcpu->arch.srr0 = regs->srr0;
 	vcpu->arch.srr1 = regs->srr1;
 	vcpu->arch.sprg0 = regs->sprg0;
@@ -575,31 +504,62 @@ int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 	return -ENOTSUPP;
 }
 
-/* 'linear_address' is actually an encoding of AS|PID|EADDR . */
 int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
                                   struct kvm_translation *tr)
 {
-	struct tlbe *gtlbe;
-	int index;
-	gva_t eaddr;
-	u8 pid;
-	u8 as;
-
-	eaddr = tr->linear_address;
-	pid = (tr->linear_address >> 32) & 0xff;
-	as = (tr->linear_address >> 40) & 0x1;
-
-	index = kvmppc_44x_tlb_index(vcpu, eaddr, pid, as);
-	if (index == -1) {
-		tr->valid = 0;
-		return 0;
-	}
+	return kvmppc_core_vcpu_translate(vcpu, tr);
+}
 
-	gtlbe = &vcpu->arch.guest_tlb[index];
+int kvmppc_booke_init(void)
+{
+	unsigned long ivor[16];
+	unsigned long max_ivor = 0;
+	int i;
 
-	tr->physical_address = tlb_xlate(gtlbe, eaddr);
-	/* XXX what does "writeable" and "usermode" even mean? */
-	tr->valid = 1;
+	/* We install our own exception handlers by hijacking IVPR. IVPR must
+	 * be 16-bit aligned, so we need a 64KB allocation. */
+	kvmppc_booke_handlers = __get_free_pages(GFP_KERNEL | __GFP_ZERO,
+	                                         VCPU_SIZE_ORDER);
+	if (!kvmppc_booke_handlers)
+		return -ENOMEM;
+
+	/* XXX make sure our handlers are smaller than Linux's */
+
+	/* Copy our interrupt handlers to match host IVORs. That way we don't
+	 * have to swap the IVORs on every guest/host transition. */
+	ivor[0] = mfspr(SPRN_IVOR0);
+	ivor[1] = mfspr(SPRN_IVOR1);
+	ivor[2] = mfspr(SPRN_IVOR2);
+	ivor[3] = mfspr(SPRN_IVOR3);
+	ivor[4] = mfspr(SPRN_IVOR4);
+	ivor[5] = mfspr(SPRN_IVOR5);
+	ivor[6] = mfspr(SPRN_IVOR6);
+	ivor[7] = mfspr(SPRN_IVOR7);
+	ivor[8] = mfspr(SPRN_IVOR8);
+	ivor[9] = mfspr(SPRN_IVOR9);
+	ivor[10] = mfspr(SPRN_IVOR10);
+	ivor[11] = mfspr(SPRN_IVOR11);
+	ivor[12] = mfspr(SPRN_IVOR12);
+	ivor[13] = mfspr(SPRN_IVOR13);
+	ivor[14] = mfspr(SPRN_IVOR14);
+	ivor[15] = mfspr(SPRN_IVOR15);
+
+	for (i = 0; i < 16; i++) {
+		if (ivor[i] > max_ivor)
+			max_ivor = ivor[i];
+
+		memcpy((void *)kvmppc_booke_handlers + ivor[i],
+		       kvmppc_handlers_start + i * kvmppc_handler_len,
+		       kvmppc_handler_len);
+	}
+	flush_icache_range(kvmppc_booke_handlers,
+	                   kvmppc_booke_handlers + max_ivor + kvmppc_handler_len);
 
 	return 0;
 }
+
+void __exit kvmppc_booke_exit(void)
+{
+	free_pages(kvmppc_booke_handlers, VCPU_SIZE_ORDER);
+	kvm_exit();
+}

arch/powerpc/kvm/booke.h

@@ -0,0 +1,60 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Copyright IBM Corp. 2008
+ *
+ * Authors: Hollis Blanchard <hollisb@us.ibm.com>
+ */
+
+#ifndef __KVM_BOOKE_H__
+#define __KVM_BOOKE_H__
+
+#include <linux/types.h>
+#include <linux/kvm_host.h>
+#include "timing.h"
+
+/* interrupt priortity ordering */
+#define BOOKE_IRQPRIO_DATA_STORAGE 0
+#define BOOKE_IRQPRIO_INST_STORAGE 1
+#define BOOKE_IRQPRIO_ALIGNMENT 2
+#define BOOKE_IRQPRIO_PROGRAM 3
+#define BOOKE_IRQPRIO_FP_UNAVAIL 4
+#define BOOKE_IRQPRIO_SYSCALL 5
+#define BOOKE_IRQPRIO_AP_UNAVAIL 6
+#define BOOKE_IRQPRIO_DTLB_MISS 7
+#define BOOKE_IRQPRIO_ITLB_MISS 8
+#define BOOKE_IRQPRIO_MACHINE_CHECK 9
+#define BOOKE_IRQPRIO_DEBUG 10
+#define BOOKE_IRQPRIO_CRITICAL 11
+#define BOOKE_IRQPRIO_WATCHDOG 12
+#define BOOKE_IRQPRIO_EXTERNAL 13
+#define BOOKE_IRQPRIO_FIT 14
+#define BOOKE_IRQPRIO_DECREMENTER 15
+
+/* Helper function for "full" MSR writes. No need to call this if only EE is
+ * changing. */
+static inline void kvmppc_set_msr(struct kvm_vcpu *vcpu, u32 new_msr)
+{
+	if ((new_msr & MSR_PR) != (vcpu->arch.msr & MSR_PR))
+		kvmppc_mmu_priv_switch(vcpu, new_msr & MSR_PR);
+
+	vcpu->arch.msr = new_msr;
+
+	if (vcpu->arch.msr & MSR_WE) {
+		kvm_vcpu_block(vcpu);
+		kvmppc_set_exit_type(vcpu, EMULATED_MTMSRWE_EXITS);
+	};
+}
+
+#endif /* __KVM_BOOKE_H__ */

arch/powerpc/kvm/booke_host.c

@@ -1,83 +0,0 @@
-/*
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License, version 2, as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
- *
- * Copyright IBM Corp. 2008
- *
- * Authors: Hollis Blanchard <hollisb@us.ibm.com>
- */
-
-#include <linux/errno.h>
-#include <linux/kvm_host.h>
-#include <linux/module.h>
-#include <asm/cacheflush.h>
-#include <asm/kvm_ppc.h>
-
-unsigned long kvmppc_booke_handlers;
-
-static int kvmppc_booke_init(void)
-{
-	unsigned long ivor[16];
-	unsigned long max_ivor = 0;
-	int i;
-
-	/* We install our own exception handlers by hijacking IVPR. IVPR must
-	 * be 16-bit aligned, so we need a 64KB allocation. */
-	kvmppc_booke_handlers = __get_free_pages(GFP_KERNEL | __GFP_ZERO,
-	                                         VCPU_SIZE_ORDER);
-	if (!kvmppc_booke_handlers)
-		return -ENOMEM;
-
-	/* XXX make sure our handlers are smaller than Linux's */
-
-	/* Copy our interrupt handlers to match host IVORs. That way we don't
-	 * have to swap the IVORs on every guest/host transition. */
-	ivor[0] = mfspr(SPRN_IVOR0);
-	ivor[1] = mfspr(SPRN_IVOR1);
-	ivor[2] = mfspr(SPRN_IVOR2);
-	ivor[3] = mfspr(SPRN_IVOR3);
-	ivor[4] = mfspr(SPRN_IVOR4);
-	ivor[5] = mfspr(SPRN_IVOR5);
-	ivor[6] = mfspr(SPRN_IVOR6);
-	ivor[7] = mfspr(SPRN_IVOR7);
-	ivor[8] = mfspr(SPRN_IVOR8);
-	ivor[9] = mfspr(SPRN_IVOR9);
-	ivor[10] = mfspr(SPRN_IVOR10);
-	ivor[11] = mfspr(SPRN_IVOR11);
-	ivor[12] = mfspr(SPRN_IVOR12);
-	ivor[13] = mfspr(SPRN_IVOR13);
-	ivor[14] = mfspr(SPRN_IVOR14);
-	ivor[15] = mfspr(SPRN_IVOR15);
-
-	for (i = 0; i < 16; i++) {
-		if (ivor[i] > max_ivor)
-			max_ivor = ivor[i];
-
-		memcpy((void *)kvmppc_booke_handlers + ivor[i],
-		       kvmppc_handlers_start + i * kvmppc_handler_len,
-		       kvmppc_handler_len);
-	}
-	flush_icache_range(kvmppc_booke_handlers,
-	                   kvmppc_booke_handlers + max_ivor + kvmppc_handler_len);
-
-	return kvm_init(NULL, sizeof(struct kvm_vcpu), THIS_MODULE);
-}
-
-static void __exit kvmppc_booke_exit(void)
-{
-	free_pages(kvmppc_booke_handlers, VCPU_SIZE_ORDER);
-	kvm_exit();
-}
-
-module_init(kvmppc_booke_init)
-module_exit(kvmppc_booke_exit)

arch/powerpc/kvm/booke_interrupts.S

@@ -107,6 +107,18 @@ _GLOBAL(kvmppc_resume_host)
 	li	r6, 1
 	slw	r6, r6, r5
 
+#ifdef CONFIG_KVM_EXIT_TIMING
+	/* save exit time */
+1:
+	mfspr	r7, SPRN_TBRU
+	mfspr	r8, SPRN_TBRL
+	mfspr	r9, SPRN_TBRU
+	cmpw	r9, r7
+	bne	1b
+	stw	r8, VCPU_TIMING_EXIT_TBL(r4)
+	stw	r9, VCPU_TIMING_EXIT_TBU(r4)
+#endif
+
 	/* Save the faulting instruction and all GPRs for emulation. */
 	andi.	r7, r6, NEED_INST_MASK
 	beq	..skip_inst_copy
@@ -335,54 +347,6 @@ lightweight_exit:
 	lwz	r3, VCPU_SHADOW_PID(r4)
 	mtspr	SPRN_PID, r3
 
-	/* Prevent all asynchronous TLB updates. */
-	mfmsr	r5
-	lis	r6, (MSR_EE|MSR_CE|MSR_ME|MSR_DE)@h
-	ori	r6, r6, (MSR_EE|MSR_CE|MSR_ME|MSR_DE)@l
-	andc	r6, r5, r6
-	mtmsr	r6
-
-	/* Load the guest mappings, leaving the host's "pinned" kernel mappings
-	 * in place. */
-	mfspr	r10, SPRN_MMUCR			/* Save host MMUCR. */
-	li	r5, PPC44x_TLB_SIZE
-	lis	r5, tlb_44x_hwater@ha
-	lwz	r5, tlb_44x_hwater@l(r5)
-	mtctr	r5
-	addi	r9, r4, VCPU_SHADOW_TLB
-	addi	r5, r4, VCPU_SHADOW_MOD
-	li	r3, 0
-1:
-	lbzx	r7, r3, r5
-	cmpwi	r7, 0
-	beq	3f
-
-	/* Load guest entry. */
-	mulli	r11, r3, TLBE_BYTES
-	add	r11, r11, r9
-	lwz	r7, 0(r11)
-	mtspr	SPRN_MMUCR, r7
-	lwz	r7, 4(r11)
-	tlbwe	r7, r3, PPC44x_TLB_PAGEID
-	lwz	r7, 8(r11)
-	tlbwe	r7, r3, PPC44x_TLB_XLAT
-	lwz	r7, 12(r11)
-	tlbwe	r7, r3, PPC44x_TLB_ATTRIB
-3:
-	addi	r3, r3, 1                       /* Increment index. */
-	bdnz	1b
-
-	mtspr	SPRN_MMUCR, r10			/* Restore host MMUCR. */
-
-	/* Clear bitmap of modified TLB entries */
-	li	r5, PPC44x_TLB_SIZE>>2
-	mtctr	r5
-	addi	r5, r4, VCPU_SHADOW_MOD - 4
-	li	r6, 0
-1:
-	stwu	r6, 4(r5)
-	bdnz	1b
-
 	iccci	0, 0 /* XXX hack */
 
 	/* Load some guest volatiles. */
@@ -423,6 +387,18 @@ lightweight_exit:
 	lwz	r3, VCPU_SPRG7(r4)
 	mtspr	SPRN_SPRG7, r3
 
+#ifdef CONFIG_KVM_EXIT_TIMING
+	/* save enter time */
+1:
+	mfspr	r6, SPRN_TBRU
+	mfspr	r7, SPRN_TBRL
+	mfspr	r8, SPRN_TBRU
+	cmpw	r8, r6
+	bne	1b
+	stw	r7, VCPU_TIMING_LAST_ENTER_TBL(r4)
+	stw	r8, VCPU_TIMING_LAST_ENTER_TBU(r4)
+#endif
+
 	/* Finish loading guest volatiles and jump to guest. */
 	lwz	r3, VCPU_CTR(r4)
 	mtctr	r3

arch/powerpc/kvm/emulate.c

@@ -23,161 +23,14 @@
 #include <linux/string.h>
 #include <linux/kvm_host.h>
 
-#include <asm/dcr.h>
-#include <asm/dcr-regs.h>
+#include <asm/reg.h>
 #include <asm/time.h>
 #include <asm/byteorder.h>
 #include <asm/kvm_ppc.h>
+#include <asm/disassemble.h>
+#include "timing.h"
 
-#include "44x_tlb.h"
-
-/* Instruction decoding */
-static inline unsigned int get_op(u32 inst)
-{
-	return inst >> 26;
-}
-
-static inline unsigned int get_xop(u32 inst)
-{
-	return (inst >> 1) & 0x3ff;
-}
-
-static inline unsigned int get_sprn(u32 inst)
-{
-	return ((inst >> 16) & 0x1f) | ((inst >> 6) & 0x3e0);
-}
-
-static inline unsigned int get_dcrn(u32 inst)
-{
-	return ((inst >> 16) & 0x1f) | ((inst >> 6) & 0x3e0);
-}
-
-static inline unsigned int get_rt(u32 inst)
-{
-	return (inst >> 21) & 0x1f;
-}
-
-static inline unsigned int get_rs(u32 inst)
-{
-	return (inst >> 21) & 0x1f;
-}
-
-static inline unsigned int get_ra(u32 inst)
-{
-	return (inst >> 16) & 0x1f;
-}
-
-static inline unsigned int get_rb(u32 inst)
-{
-	return (inst >> 11) & 0x1f;
-}
-
-static inline unsigned int get_rc(u32 inst)
-{
-	return inst & 0x1;
-}
-
-static inline unsigned int get_ws(u32 inst)
-{
-	return (inst >> 11) & 0x1f;
-}
-
-static inline unsigned int get_d(u32 inst)
-{
-	return inst & 0xffff;
-}
-
-static int tlbe_is_host_safe(const struct kvm_vcpu *vcpu,
-                             const struct tlbe *tlbe)
-{
-	gpa_t gpa;
-
-	if (!get_tlb_v(tlbe))
-		return 0;
-
-	/* Does it match current guest AS? */
-	/* XXX what about IS != DS? */
-	if (get_tlb_ts(tlbe) != !!(vcpu->arch.msr & MSR_IS))
-		return 0;
-
-	gpa = get_tlb_raddr(tlbe);
-	if (!gfn_to_memslot(vcpu->kvm, gpa >> PAGE_SHIFT))
-		/* Mapping is not for RAM. */
-		return 0;
-
-	return 1;
-}
-
-static int kvmppc_emul_tlbwe(struct kvm_vcpu *vcpu, u32 inst)
-{
-	u64 eaddr;
-	u64 raddr;
-	u64 asid;
-	u32 flags;
-	struct tlbe *tlbe;
-	unsigned int ra;
-	unsigned int rs;
-	unsigned int ws;
-	unsigned int index;
-
-	ra = get_ra(inst);
-	rs = get_rs(inst);
-	ws = get_ws(inst);
-
-	index = vcpu->arch.gpr[ra];
-	if (index > PPC44x_TLB_SIZE) {
-		printk("%s: index %d\n", __func__, index);
-		kvmppc_dump_vcpu(vcpu);
-		return EMULATE_FAIL;
-	}
-
-	tlbe = &vcpu->arch.guest_tlb[index];
-
-	/* Invalidate shadow mappings for the about-to-be-clobbered TLBE. */
-	if (tlbe->word0 & PPC44x_TLB_VALID) {
-		eaddr = get_tlb_eaddr(tlbe);
-		asid = (tlbe->word0 & PPC44x_TLB_TS) | tlbe->tid;
-		kvmppc_mmu_invalidate(vcpu, eaddr, get_tlb_end(tlbe), asid);
-	}
-
-	switch (ws) {
-	case PPC44x_TLB_PAGEID:
-		tlbe->tid = vcpu->arch.mmucr & 0xff;
-		tlbe->word0 = vcpu->arch.gpr[rs];
-		break;
-
-	case PPC44x_TLB_XLAT:
-		tlbe->word1 = vcpu->arch.gpr[rs];
-		break;
-
-	case PPC44x_TLB_ATTRIB:
-		tlbe->word2 = vcpu->arch.gpr[rs];
-		break;
-
-	default:
-		return EMULATE_FAIL;
-	}
-
-	if (tlbe_is_host_safe(vcpu, tlbe)) {
-		eaddr = get_tlb_eaddr(tlbe);
-		raddr = get_tlb_raddr(tlbe);
-		asid = (tlbe->word0 & PPC44x_TLB_TS) | tlbe->tid;
-		flags = tlbe->word2 & 0xffff;
-
-		/* Create a 4KB mapping on the host. If the guest wanted a
-		 * large page, only the first 4KB is mapped here and the rest
-		 * are mapped on the fly. */
-		kvmppc_mmu_map(vcpu, eaddr, raddr >> PAGE_SHIFT, asid, flags);
-	}
-
-	KVMTRACE_5D(GTLB_WRITE, vcpu, index,
-			tlbe->tid, tlbe->word0, tlbe->word1, tlbe->word2,
-			handler);
-
-	return EMULATE_DONE;
-}
-
-static void kvmppc_emulate_dec(struct kvm_vcpu *vcpu)
+void kvmppc_emulate_dec(struct kvm_vcpu *vcpu)
 {
 	if (vcpu->arch.tcr & TCR_DIE) {
 		/* The decrementer ticks at the same rate as the timebase, so
@@ -193,12 +46,6 @@ static void kvmppc_emulate_dec(struct kvm_vcpu *vcpu)
 	}
 }
 
-static void kvmppc_emul_rfi(struct kvm_vcpu *vcpu)
-{
-	vcpu->arch.pc = vcpu->arch.srr0;
-	kvmppc_set_msr(vcpu, vcpu->arch.srr1);
-}
-
 /* XXX to do:
  * lhax
  * lhaux
@@ -213,40 +60,30 @@ static void kvmppc_emul_rfi(struct kvm_vcpu *vcpu)
  *
  * XXX is_bigendian should depend on MMU mapping or MSR[LE]
  */
+/* XXX Should probably auto-generate instruction decoding for a particular core
+ * from opcode tables in the future. */
 int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu)
 {
 	u32 inst = vcpu->arch.last_inst;
 	u32 ea;
 	int ra;
 	int rb;
-	int rc;
 	int rs;
 	int rt;
 	int sprn;
-	int dcrn;
 	enum emulation_result emulated = EMULATE_DONE;
 	int advance = 1;
 
+	/* this default type might be overwritten by subcategories */
+	kvmppc_set_exit_type(vcpu, EMULATED_INST_EXITS);
+
 	switch (get_op(inst)) {
-	case 3:                                                 /* trap */
-		printk("trap!\n");
-		kvmppc_queue_exception(vcpu, BOOKE_INTERRUPT_PROGRAM);
+	case 3:                                             /* trap */
+		vcpu->arch.esr |= ESR_PTR;
+		kvmppc_core_queue_program(vcpu);
 		advance = 0;
 		break;
 
-	case 19:
-		switch (get_xop(inst)) {
-		case 50:                                        /* rfi */
-			kvmppc_emul_rfi(vcpu);
-			advance = 0;
-			break;
-
-		default:
-			emulated = EMULATE_FAIL;
-			break;
-		}
-		break;
-
 	case 31:
 		switch (get_xop(inst)) {
 
@@ -255,27 +92,11 @@ int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu)
 			emulated = kvmppc_handle_load(run, vcpu, rt, 4, 1);
 			break;
 
-		case 83:                                        /* mfmsr */
-			rt = get_rt(inst);
-			vcpu->arch.gpr[rt] = vcpu->arch.msr;
-			break;
-
 		case 87:                                        /* lbzx */
 			rt = get_rt(inst);
 			emulated = kvmppc_handle_load(run, vcpu, rt, 1, 1);
 			break;
 
-		case 131:                                       /* wrtee */
-			rs = get_rs(inst);
-			vcpu->arch.msr = (vcpu->arch.msr & ~MSR_EE)
-			                 | (vcpu->arch.gpr[rs] & MSR_EE);
-			break;
-
-		case 146:                                       /* mtmsr */
-			rs = get_rs(inst);
-			kvmppc_set_msr(vcpu, vcpu->arch.gpr[rs]);
-			break;
-
 		case 151:                                       /* stwx */
 			rs = get_rs(inst);
 			emulated = kvmppc_handle_store(run, vcpu,
@@ -283,11 +104,6 @@ int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu)
 			                               4, 1);
 			break;
 
-		case 163:                                       /* wrteei */
-			vcpu->arch.msr = (vcpu->arch.msr & ~MSR_EE)
-			                 | (inst & MSR_EE);
-			break;
-
 		case 215:                                       /* stbx */
 			rs = get_rs(inst);
 			emulated = kvmppc_handle_store(run, vcpu,
@@ -328,42 +144,6 @@ int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu)
 			vcpu->arch.gpr[ra] = ea;
 			break;
 
-		case 323:                                       /* mfdcr */
-			dcrn = get_dcrn(inst);
-			rt = get_rt(inst);
-
-			/* The guest may access CPR0 registers to determine the timebase
-			 * frequency, and it must know the real host frequency because it
-			 * can directly access the timebase registers.
-			 *
-			 * It would be possible to emulate those accesses in userspace,
-			 * but userspace can really only figure out the end frequency.
-			 * We could decompose that into the factors that compute it, but
-			 * that's tricky math, and it's easier to just report the real
-			 * CPR0 values.
-			 */
-			switch (dcrn) {
-			case DCRN_CPR0_CONFIG_ADDR:
-				vcpu->arch.gpr[rt] = vcpu->arch.cpr0_cfgaddr;
-				break;
-			case DCRN_CPR0_CONFIG_DATA:
-				local_irq_disable();
-				mtdcr(DCRN_CPR0_CONFIG_ADDR,
-				      vcpu->arch.cpr0_cfgaddr);
-				vcpu->arch.gpr[rt] = mfdcr(DCRN_CPR0_CONFIG_DATA);
-				local_irq_enable();
-				break;
-			default:
-				run->dcr.dcrn = dcrn;
-				run->dcr.data =  0;
-				run->dcr.is_write = 0;
-				vcpu->arch.io_gpr = rt;
-				vcpu->arch.dcr_needed = 1;
-				emulated = EMULATE_DO_DCR;
-			}
-
-			break;
-
 		case 339:                                       /* mfspr */
 			sprn = get_sprn(inst);
 			rt = get_rt(inst);
@@ -373,26 +153,8 @@ int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu)
 				vcpu->arch.gpr[rt] = vcpu->arch.srr0; break;
 			case SPRN_SRR1:
 				vcpu->arch.gpr[rt] = vcpu->arch.srr1; break;
-			case SPRN_MMUCR:
-				vcpu->arch.gpr[rt] = vcpu->arch.mmucr; break;
-			case SPRN_PID:
-				vcpu->arch.gpr[rt] = vcpu->arch.pid; break;
-			case SPRN_IVPR:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivpr; break;
-			case SPRN_CCR0:
-				vcpu->arch.gpr[rt] = vcpu->arch.ccr0; break;
-			case SPRN_CCR1:
-				vcpu->arch.gpr[rt] = vcpu->arch.ccr1; break;
 			case SPRN_PVR:
 				vcpu->arch.gpr[rt] = vcpu->arch.pvr; break;
-			case SPRN_DEAR:
-				vcpu->arch.gpr[rt] = vcpu->arch.dear; break;
-			case SPRN_ESR:
-				vcpu->arch.gpr[rt] = vcpu->arch.esr; break;
-			case SPRN_DBCR0:
-				vcpu->arch.gpr[rt] = vcpu->arch.dbcr0; break;
-			case SPRN_DBCR1:
-				vcpu->arch.gpr[rt] = vcpu->arch.dbcr1; break;
 
 			/* Note: mftb and TBRL/TBWL are user-accessible, so
 			 * the guest can always access the real TB anyways.
@@ -413,42 +175,12 @@ int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu)
 			/* Note: SPRG4-7 are user-readable, so we don't get
 			 * a trap. */
 
-			case SPRN_IVOR0:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[0]; break;
-			case SPRN_IVOR1:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[1]; break;
-			case SPRN_IVOR2:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[2]; break;
-			case SPRN_IVOR3:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[3]; break;
-			case SPRN_IVOR4:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[4]; break;
-			case SPRN_IVOR5:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[5]; break;
-			case SPRN_IVOR6:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[6]; break;
-			case SPRN_IVOR7:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[7]; break;
-			case SPRN_IVOR8:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[8]; break;
-			case SPRN_IVOR9:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[9]; break;
-			case SPRN_IVOR10:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[10]; break;
-			case SPRN_IVOR11:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[11]; break;
-			case SPRN_IVOR12:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[12]; break;
-			case SPRN_IVOR13:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[13]; break;
-			case SPRN_IVOR14:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[14]; break;
-			case SPRN_IVOR15:
-				vcpu->arch.gpr[rt] = vcpu->arch.ivor[15]; break;
-
 			default:
-				printk("mfspr: unknown spr %x\n", sprn);
-				vcpu->arch.gpr[rt] = 0;
+				emulated = kvmppc_core_emulate_mfspr(vcpu, sprn, rt);
+				if (emulated == EMULATE_FAIL) {
+					printk("mfspr: unknown spr %x\n", sprn);
+					vcpu->arch.gpr[rt] = 0;
+				}
 				break;
 			}
 			break;
@@ -478,25 +210,6 @@ int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu)
 			vcpu->arch.gpr[ra] = ea;
 			break;
 
-		case 451:                                       /* mtdcr */
-			dcrn = get_dcrn(inst);
-			rs = get_rs(inst);
-
-			/* emulate some access in kernel */
-			switch (dcrn) {
-			case DCRN_CPR0_CONFIG_ADDR:
-				vcpu->arch.cpr0_cfgaddr = vcpu->arch.gpr[rs];
-				break;
-			default:
-				run->dcr.dcrn = dcrn;
-				run->dcr.data = vcpu->arch.gpr[rs];
-				run->dcr.is_write = 1;
-				vcpu->arch.dcr_needed = 1;
-				emulated = EMULATE_DO_DCR;
-			}
-
-			break;
-
 		case 467:                                       /* mtspr */
 			sprn = get_sprn(inst);
 			rs = get_rs(inst);
@@ -505,22 +218,6 @@ int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu)
 				vcpu->arch.srr0 = vcpu->arch.gpr[rs]; break;
 			case SPRN_SRR1:
 				vcpu->arch.srr1 = vcpu->arch.gpr[rs]; break;
-			case SPRN_MMUCR:
-				vcpu->arch.mmucr = vcpu->arch.gpr[rs]; break;
-			case SPRN_PID:
-				kvmppc_set_pid(vcpu, vcpu->arch.gpr[rs]); break;
-			case SPRN_CCR0:
-				vcpu->arch.ccr0 = vcpu->arch.gpr[rs]; break;
-			case SPRN_CCR1:
-				vcpu->arch.ccr1 = vcpu->arch.gpr[rs]; break;
-			case SPRN_DEAR:
-				vcpu->arch.dear = vcpu->arch.gpr[rs]; break;
-			case SPRN_ESR:
-				vcpu->arch.esr = vcpu->arch.gpr[rs]; break;
-			case SPRN_DBCR0:
-				vcpu->arch.dbcr0 = vcpu->arch.gpr[rs]; break;
-			case SPRN_DBCR1:
-				vcpu->arch.dbcr1 = vcpu->arch.gpr[rs]; break;
 
 			/* XXX We need to context-switch the timebase for
 			 * watchdog and FIT. */
@@ -532,14 +229,6 @@ int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu)
 				kvmppc_emulate_dec(vcpu);
 				break;
 
-			case SPRN_TSR:
-				vcpu->arch.tsr &= ~vcpu->arch.gpr[rs]; break;
-
-			case SPRN_TCR:
-				vcpu->arch.tcr = vcpu->arch.gpr[rs];
-				kvmppc_emulate_dec(vcpu);
-				break;
-
 			case SPRN_SPRG0:
 				vcpu->arch.sprg0 = vcpu->arch.gpr[rs]; break;
 			case SPRN_SPRG1:
@@ -549,56 +238,10 @@ int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu)
 			case SPRN_SPRG3:
 				vcpu->arch.sprg3 = vcpu->arch.gpr[rs]; break;
 
-			/* Note: SPRG4-7 are user-readable. These values are
-			 * loaded into the real SPRGs when resuming the
-			 * guest. */
-			case SPRN_SPRG4:
-				vcpu->arch.sprg4 = vcpu->arch.gpr[rs]; break;
-			case SPRN_SPRG5:
-				vcpu->arch.sprg5 = vcpu->arch.gpr[rs]; break;
-			case SPRN_SPRG6:
-				vcpu->arch.sprg6 = vcpu->arch.gpr[rs]; break;
-			case SPRN_SPRG7:
-				vcpu->arch.sprg7 = vcpu->arch.gpr[rs]; break;
-
-			case SPRN_IVPR:
-				vcpu->arch.ivpr = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR0:
-				vcpu->arch.ivor[0] = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR1:
-				vcpu->arch.ivor[1] = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR2:
-				vcpu->arch.ivor[2] = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR3:
-				vcpu->arch.ivor[3] = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR4:
-				vcpu->arch.ivor[4] = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR5:
-				vcpu->arch.ivor[5] = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR6:
-				vcpu->arch.ivor[6] = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR7:
-				vcpu->arch.ivor[7] = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR8:
-				vcpu->arch.ivor[8] = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR9:
-				vcpu->arch.ivor[9] = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR10:
-				vcpu->arch.ivor[10] = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR11:
-				vcpu->arch.ivor[11] = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR12:
-				vcpu->arch.ivor[12] = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR13:
-				vcpu->arch.ivor[13] = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR14:
-				vcpu->arch.ivor[14] = vcpu->arch.gpr[rs]; break;
-			case SPRN_IVOR15:
-				vcpu->arch.ivor[15] = vcpu->arch.gpr[rs]; break;
-
 			default:
-				printk("mtspr: unknown spr %x\n", sprn);
-				emulated = EMULATE_FAIL;
+				emulated = kvmppc_core_emulate_mtspr(vcpu, sprn, rs);
+				if (emulated == EMULATE_FAIL)
+					printk("mtspr: unknown spr %x\n", sprn);
 				break;
 			}
 			break;
@@ -629,36 +272,6 @@ int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu)
 			                               4, 0);
 			break;
 
-		case 978:                                       /* tlbwe */
-			emulated = kvmppc_emul_tlbwe(vcpu, inst);
-			break;
-
-		case 914:       {                               /* tlbsx */
-			int index;
-			unsigned int as = get_mmucr_sts(vcpu);
-			unsigned int pid = get_mmucr_stid(vcpu);
-
-			rt = get_rt(inst);
-			ra = get_ra(inst);
-			rb = get_rb(inst);
-			rc = get_rc(inst);
-
-			ea = vcpu->arch.gpr[rb];
-			if (ra)
-				ea += vcpu->arch.gpr[ra];
-
-			index = kvmppc_44x_tlb_index(vcpu, ea, pid, as);
-			if (rc) {
-				if (index < 0)
-					vcpu->arch.cr &= ~0x20000000;
-				else
-					vcpu->arch.cr |= 0x20000000;
-			}
-			vcpu->arch.gpr[rt] = index;
-
-			}
-			break;
-
 		case 790:                                       /* lhbrx */
 			rt = get_rt(inst);
 			emulated = kvmppc_handle_load(run, vcpu, rt, 2, 0);
@@ -674,14 +287,9 @@ int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu)
 			                               2, 0);
 			break;
 
-		case 966:                                       /* iccci */
-			break;
-
 		default:
-			printk("unknown: op %d xop %d\n", get_op(inst),
-				get_xop(inst));
+			/* Attempt core-specific emulation below. */
 			emulated = EMULATE_FAIL;
-			break;
 		}
 		break;
 
@@ -764,12 +372,19 @@ int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu)
 		break;
 
 	default:
-		printk("unknown op %d\n", get_op(inst));
 		emulated = EMULATE_FAIL;
-		break;
 	}
 
-	KVMTRACE_3D(PPC_INSTR, vcpu, inst, vcpu->arch.pc, emulated, entryexit);
+	if (emulated == EMULATE_FAIL) {
+		emulated = kvmppc_core_emulate_op(run, vcpu, inst, &advance);
+		if (emulated == EMULATE_FAIL) {
+			advance = 0;
+			printk(KERN_ERR "Couldn't emulate instruction 0x%08x "
+			       "(op %d xop %d)\n", inst, get_op(inst), get_xop(inst));
+		}
+	}
+
+	KVMTRACE_3D(PPC_INSTR, vcpu, inst, (int)vcpu->arch.pc, emulated, entryexit);
 
 	if (advance)
 		vcpu->arch.pc += 4; /* Advance past emulated instruction. */

arch/powerpc/kvm/powerpc.c

@@ -28,9 +28,9 @@
 #include <asm/uaccess.h>
 #include <asm/kvm_ppc.h>
 #include <asm/tlbflush.h>
+#include "timing.h"
 #include "../mm/mmu_decl.h"
 
-
 gfn_t unalias_gfn(struct kvm *kvm, gfn_t gfn)
 {
 	return gfn;
@@ -99,14 +99,7 @@ void kvm_arch_hardware_unsetup(void)
 
 void kvm_arch_check_processor_compat(void *rtn)
 {
-	int r;
-
-	if (strcmp(cur_cpu_spec->platform, "ppc440") == 0)
-		r = 0;
-	else
-		r = -ENOTSUPP;
-
-	*(int *)rtn = r;
+	*(int *)rtn = kvmppc_core_check_processor_compat();
 }
 
 struct kvm *kvm_arch_create_vm(void)
@@ -144,9 +137,6 @@ int kvm_dev_ioctl_check_extension(long ext)
 	int r;
 
 	switch (ext) {
-	case KVM_CAP_USER_MEMORY:
-		r = 1;
-		break;
 	case KVM_CAP_COALESCED_MMIO:
 		r = KVM_COALESCED_MMIO_PAGE_OFFSET;
 		break;
@@ -179,30 +169,15 @@ void kvm_arch_flush_shadow(struct kvm *kvm)
 struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, unsigned int id)
 {
 	struct kvm_vcpu *vcpu;
-	int err;
-
-	vcpu = kmem_cache_zalloc(kvm_vcpu_cache, GFP_KERNEL);
-	if (!vcpu) {
-		err = -ENOMEM;
-		goto out;
-	}
-
-	err = kvm_vcpu_init(vcpu, kvm, id);
-	if (err)
-		goto free_vcpu;
-
+	vcpu = kvmppc_core_vcpu_create(kvm, id);
+	kvmppc_create_vcpu_debugfs(vcpu, id);
 	return vcpu;
-
-free_vcpu:
-	kmem_cache_free(kvm_vcpu_cache, vcpu);
-out:
-	return ERR_PTR(err);
 }
 
 void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu)
 {
-	kvm_vcpu_uninit(vcpu);
-	kmem_cache_free(kvm_vcpu_cache, vcpu);
+	kvmppc_remove_vcpu_debugfs(vcpu);
+	kvmppc_core_vcpu_free(vcpu);
 }
 
 void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
@@ -212,16 +187,14 @@ void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
 
 int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu)
 {
-	unsigned int priority = exception_priority[BOOKE_INTERRUPT_DECREMENTER];
-
-	return test_bit(priority, &vcpu->arch.pending_exceptions);
+	return kvmppc_core_pending_dec(vcpu);
 }
 
 static void kvmppc_decrementer_func(unsigned long data)
 {
 	struct kvm_vcpu *vcpu = (struct kvm_vcpu *)data;
 
-	kvmppc_queue_exception(vcpu, BOOKE_INTERRUPT_DECREMENTER);
+	kvmppc_core_queue_dec(vcpu);
 
 	if (waitqueue_active(&vcpu->wq)) {
 		wake_up_interruptible(&vcpu->wq);
@@ -242,96 +215,25 @@ void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu)
 	kvmppc_core_destroy_mmu(vcpu);
 }
 
-/* Note: clearing MSR[DE] just means that the debug interrupt will not be
- * delivered *immediately*. Instead, it simply sets the appropriate DBSR bits.
- * If those DBSR bits are still set when MSR[DE] is re-enabled, the interrupt
- * will be delivered as an "imprecise debug event" (which is indicated by
- * DBSR[IDE].
- */
-static void kvmppc_disable_debug_interrupts(void)
-{
-	mtmsr(mfmsr() & ~MSR_DE);
-}
-
-static void kvmppc_restore_host_debug_state(struct kvm_vcpu *vcpu)
-{
-	kvmppc_disable_debug_interrupts();
-
-	mtspr(SPRN_IAC1, vcpu->arch.host_iac[0]);
-	mtspr(SPRN_IAC2, vcpu->arch.host_iac[1]);
-	mtspr(SPRN_IAC3, vcpu->arch.host_iac[2]);
-	mtspr(SPRN_IAC4, vcpu->arch.host_iac[3]);
-	mtspr(SPRN_DBCR1, vcpu->arch.host_dbcr1);
-	mtspr(SPRN_DBCR2, vcpu->arch.host_dbcr2);
-	mtspr(SPRN_DBCR0, vcpu->arch.host_dbcr0);
-	mtmsr(vcpu->arch.host_msr);
-}
-
-static void kvmppc_load_guest_debug_registers(struct kvm_vcpu *vcpu)
-{
-	struct kvm_guest_debug *dbg = &vcpu->guest_debug;
-	u32 dbcr0 = 0;
-
-	vcpu->arch.host_msr = mfmsr();
-	kvmppc_disable_debug_interrupts();
-
-	/* Save host debug register state. */
-	vcpu->arch.host_iac[0] = mfspr(SPRN_IAC1);
-	vcpu->arch.host_iac[1] = mfspr(SPRN_IAC2);
-	vcpu->arch.host_iac[2] = mfspr(SPRN_IAC3);
-	vcpu->arch.host_iac[3] = mfspr(SPRN_IAC4);
-	vcpu->arch.host_dbcr0 = mfspr(SPRN_DBCR0);
-	vcpu->arch.host_dbcr1 = mfspr(SPRN_DBCR1);
-	vcpu->arch.host_dbcr2 = mfspr(SPRN_DBCR2);
-
-	/* set registers up for guest */
-
-	if (dbg->bp[0]) {
-		mtspr(SPRN_IAC1, dbg->bp[0]);
-		dbcr0 |= DBCR0_IAC1 | DBCR0_IDM;
-	}
-	if (dbg->bp[1]) {
-		mtspr(SPRN_IAC2, dbg->bp[1]);
-		dbcr0 |= DBCR0_IAC2 | DBCR0_IDM;
-	}
-	if (dbg->bp[2]) {
-		mtspr(SPRN_IAC3, dbg->bp[2]);
-		dbcr0 |= DBCR0_IAC3 | DBCR0_IDM;
-	}
-	if (dbg->bp[3]) {
-		mtspr(SPRN_IAC4, dbg->bp[3]);
-		dbcr0 |= DBCR0_IAC4 | DBCR0_IDM;
-	}
-
-	mtspr(SPRN_DBCR0, dbcr0);
-	mtspr(SPRN_DBCR1, 0);
-	mtspr(SPRN_DBCR2, 0);
-}
-
 void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
 {
-	int i;
-
 	if (vcpu->guest_debug.enabled)
-		kvmppc_load_guest_debug_registers(vcpu);
+		kvmppc_core_load_guest_debugstate(vcpu);
 
-	/* Mark every guest entry in the shadow TLB entry modified, so that they
-	 * will all be reloaded on the next vcpu run (instead of being
-	 * demand-faulted). */
-	for (i = 0; i <= tlb_44x_hwater; i++)
-		kvmppc_tlbe_set_modified(vcpu, i);
+	kvmppc_core_vcpu_load(vcpu, cpu);
 }
 
 void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
 {
 	if (vcpu->guest_debug.enabled)
-		kvmppc_restore_host_debug_state(vcpu);
+		kvmppc_core_load_host_debugstate(vcpu);
 
 	/* Don't leave guest TLB entries resident when being de-scheduled. */
 	/* XXX It would be nice to differentiate between heavyweight exit and
 	 * sched_out here, since we could avoid the TLB flush for heavyweight
 	 * exits. */
 	_tlbil_all();
+	kvmppc_core_vcpu_put(vcpu);
 }
 
 int kvm_arch_vcpu_ioctl_debug_guest(struct kvm_vcpu *vcpu,
@@ -355,14 +257,14 @@ int kvm_arch_vcpu_ioctl_debug_guest(struct kvm_vcpu *vcpu,
 static void kvmppc_complete_dcr_load(struct kvm_vcpu *vcpu,
                                      struct kvm_run *run)
 {
-	u32 *gpr = &vcpu->arch.gpr[vcpu->arch.io_gpr];
+	ulong *gpr = &vcpu->arch.gpr[vcpu->arch.io_gpr];
 	*gpr = run->dcr.data;
 }
 
 static void kvmppc_complete_mmio_load(struct kvm_vcpu *vcpu,
                                       struct kvm_run *run)
 {
-	u32 *gpr = &vcpu->arch.gpr[vcpu->arch.io_gpr];
+	ulong *gpr = &vcpu->arch.gpr[vcpu->arch.io_gpr];
 
 	if (run->mmio.len > sizeof(*gpr)) {
 		printk(KERN_ERR "bad MMIO length: %d\n", run->mmio.len);
@@ -460,7 +362,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 		vcpu->arch.dcr_needed = 0;
 	}
 
-	kvmppc_check_and_deliver_interrupts(vcpu);
+	kvmppc_core_deliver_interrupts(vcpu);
 
 	local_irq_disable();
 	kvm_guest_enter();
@@ -478,7 +380,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 
 int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, struct kvm_interrupt *irq)
 {
-	kvmppc_queue_exception(vcpu, BOOKE_INTERRUPT_EXTERNAL);
+	kvmppc_core_queue_external(vcpu, irq);
 
 	if (waitqueue_active(&vcpu->wq)) {
 		wake_up_interruptible(&vcpu->wq);

arch/powerpc/kvm/timing.c

@@ -0,0 +1,239 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Copyright IBM Corp. 2008
+ *
+ * Authors: Hollis Blanchard <hollisb@us.ibm.com>
+ *          Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>
+ */
+
+#include <linux/kvm_host.h>
+#include <linux/fs.h>
+#include <linux/seq_file.h>
+#include <linux/debugfs.h>
+#include <linux/uaccess.h>
+
+#include <asm/time.h>
+#include <asm-generic/div64.h>
+
+#include "timing.h"
+
+void kvmppc_init_timing_stats(struct kvm_vcpu *vcpu)
+{
+	int i;
+
+	/* pause guest execution to avoid concurrent updates */
+	local_irq_disable();
+	mutex_lock(&vcpu->mutex);
+
+	vcpu->arch.last_exit_type = 0xDEAD;
+	for (i = 0; i < __NUMBER_OF_KVM_EXIT_TYPES; i++) {
+		vcpu->arch.timing_count_type[i] = 0;
+		vcpu->arch.timing_max_duration[i] = 0;
+		vcpu->arch.timing_min_duration[i] = 0xFFFFFFFF;
+		vcpu->arch.timing_sum_duration[i] = 0;
+		vcpu->arch.timing_sum_quad_duration[i] = 0;
+	}
+	vcpu->arch.timing_last_exit = 0;
+	vcpu->arch.timing_exit.tv64 = 0;
+	vcpu->arch.timing_last_enter.tv64 = 0;
+
+	mutex_unlock(&vcpu->mutex);
+	local_irq_enable();
+}
+
+static void add_exit_timing(struct kvm_vcpu *vcpu, u64 duration, int type)
+{
+	u64 old;
+
+	do_div(duration, tb_ticks_per_usec);
+	if (unlikely(duration > 0xFFFFFFFF)) {
+		printk(KERN_ERR"%s - duration too big -> overflow"
+			" duration %lld type %d exit #%d\n",
+			__func__, duration, type,
+			vcpu->arch.timing_count_type[type]);
+		return;
+	}
+
+	vcpu->arch.timing_count_type[type]++;
+
+	/* sum */
+	old = vcpu->arch.timing_sum_duration[type];
+	vcpu->arch.timing_sum_duration[type] += duration;
+	if (unlikely(old > vcpu->arch.timing_sum_duration[type])) {
+		printk(KERN_ERR"%s - wrap adding sum of durations"
+			" old %lld new %lld type %d exit # of type %d\n",
+			__func__, old, vcpu->arch.timing_sum_duration[type],
+			type, vcpu->arch.timing_count_type[type]);
+	}
+
+	/* square sum */
+	old = vcpu->arch.timing_sum_quad_duration[type];
+	vcpu->arch.timing_sum_quad_duration[type] += (duration*duration);
+	if (unlikely(old > vcpu->arch.timing_sum_quad_duration[type])) {
+		printk(KERN_ERR"%s - wrap adding sum of squared durations"
+			" old %lld new %lld type %d exit # of type %d\n",
+			__func__, old,
+			vcpu->arch.timing_sum_quad_duration[type],
+			type, vcpu->arch.timing_count_type[type]);
+	}
+
+	/* set min/max */
+	if (unlikely(duration < vcpu->arch.timing_min_duration[type]))
+		vcpu->arch.timing_min_duration[type] = duration;
+	if (unlikely(duration > vcpu->arch.timing_max_duration[type]))
+		vcpu->arch.timing_max_duration[type] = duration;
+}
+
+void kvmppc_update_timing_stats(struct kvm_vcpu *vcpu)
+{
+	u64 exit = vcpu->arch.timing_last_exit;
+	u64 enter = vcpu->arch.timing_last_enter.tv64;
+
+	/* save exit time, used next exit when the reenter time is known */
+	vcpu->arch.timing_last_exit = vcpu->arch.timing_exit.tv64;
+
+	if (unlikely(vcpu->arch.last_exit_type == 0xDEAD || exit == 0))
+		return; /* skip incomplete cycle (e.g. after reset) */
+
+	/* update statistics for average and standard deviation */
+	add_exit_timing(vcpu, (enter - exit), vcpu->arch.last_exit_type);
+	/* enter -> timing_last_exit is time spent in guest - log this too */
+	add_exit_timing(vcpu, (vcpu->arch.timing_last_exit - enter),
+			TIMEINGUEST);
+}
+
+static const char *kvm_exit_names[__NUMBER_OF_KVM_EXIT_TYPES] = {
+	[MMIO_EXITS] =              "MMIO",
+	[DCR_EXITS] =               "DCR",
+	[SIGNAL_EXITS] =            "SIGNAL",
+	[ITLB_REAL_MISS_EXITS] =    "ITLBREAL",
+	[ITLB_VIRT_MISS_EXITS] =    "ITLBVIRT",
+	[DTLB_REAL_MISS_EXITS] =    "DTLBREAL",
+	[DTLB_VIRT_MISS_EXITS] =    "DTLBVIRT",
+	[SYSCALL_EXITS] =           "SYSCALL",
+	[ISI_EXITS] =               "ISI",
+	[DSI_EXITS] =               "DSI",
+	[EMULATED_INST_EXITS] =     "EMULINST",
+	[EMULATED_MTMSRWE_EXITS] =  "EMUL_WAIT",
+	[EMULATED_WRTEE_EXITS] =    "EMUL_WRTEE",
+	[EMULATED_MTSPR_EXITS] =    "EMUL_MTSPR",
+	[EMULATED_MFSPR_EXITS] =    "EMUL_MFSPR",
+	[EMULATED_MTMSR_EXITS] =    "EMUL_MTMSR",
+	[EMULATED_MFMSR_EXITS] =    "EMUL_MFMSR",
+	[EMULATED_TLBSX_EXITS] =    "EMUL_TLBSX",
+	[EMULATED_TLBWE_EXITS] =    "EMUL_TLBWE",
+	[EMULATED_RFI_EXITS] =      "EMUL_RFI",
+	[DEC_EXITS] =               "DEC",
+	[EXT_INTR_EXITS] =          "EXTINT",
+	[HALT_WAKEUP] =             "HALT",
+	[USR_PR_INST] =             "USR_PR_INST",
+	[FP_UNAVAIL] =              "FP_UNAVAIL",
+	[DEBUG_EXITS] =             "DEBUG",
+	[TIMEINGUEST] =             "TIMEINGUEST"
+};
+
+static int kvmppc_exit_timing_show(struct seq_file *m, void *private)
+{
+	struct kvm_vcpu *vcpu = m->private;
+	int i;
+
+	seq_printf(m, "%s", "type	count	min	max	sum	sum_squared\n");
+
+	for (i = 0; i < __NUMBER_OF_KVM_EXIT_TYPES; i++) {
+		seq_printf(m, "%12s	%10d	%10lld	%10lld	%20lld	%20lld\n",
+			kvm_exit_names[i],
+			vcpu->arch.timing_count_type[i],
+			vcpu->arch.timing_min_duration[i],
+			vcpu->arch.timing_max_duration[i],
+			vcpu->arch.timing_sum_duration[i],
+			vcpu->arch.timing_sum_quad_duration[i]);
+	}
+	return 0;
+}
+
+/* Write 'c' to clear the timing statistics. */
+static ssize_t kvmppc_exit_timing_write(struct file *file,
+				       const char __user *user_buf,
+				       size_t count, loff_t *ppos)
+{
+	int err = -EINVAL;
+	char c;
+
+	if (count > 1) {
+		goto done;
+	}
+
+	if (get_user(c, user_buf)) {
+		err = -EFAULT;
+		goto done;
+	}
+
+	if (c == 'c') {
+		struct seq_file *seqf = (struct seq_file *)file->private_data;
+		struct kvm_vcpu *vcpu = seqf->private;
+		/* Write does not affect our buffers previously generated with
+		 * show. seq_file is locked here to prevent races of init with
+		 * a show call */
+		mutex_lock(&seqf->lock);
+		kvmppc_init_timing_stats(vcpu);
+		mutex_unlock(&seqf->lock);
+		err = count;
+	}
+
+done:
+	return err;
+}
+
+static int kvmppc_exit_timing_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, kvmppc_exit_timing_show, inode->i_private);
+}
+
+static struct file_operations kvmppc_exit_timing_fops = {
+	.owner   = THIS_MODULE,
+	.open    = kvmppc_exit_timing_open,
+	.read    = seq_read,
+	.write   = kvmppc_exit_timing_write,
+	.llseek  = seq_lseek,
+	.release = single_release,
+};
+
+void kvmppc_create_vcpu_debugfs(struct kvm_vcpu *vcpu, unsigned int id)
+{
+	static char dbg_fname[50];
+	struct dentry *debugfs_file;
+
+	snprintf(dbg_fname, sizeof(dbg_fname), "vm%u_vcpu%u_timing",
+		 current->pid, id);
+	debugfs_file = debugfs_create_file(dbg_fname, 0666,
+					kvm_debugfs_dir, vcpu,
+					&kvmppc_exit_timing_fops);
+
+	if (!debugfs_file) {
+		printk(KERN_ERR"%s: error creating debugfs file %s\n",
+			__func__, dbg_fname);
+		return;
+	}
+
+	vcpu->arch.debugfs_exit_timing = debugfs_file;
+}
+
+void kvmppc_remove_vcpu_debugfs(struct kvm_vcpu *vcpu)
+{
+	if (vcpu->arch.debugfs_exit_timing) {
+		debugfs_remove(vcpu->arch.debugfs_exit_timing);
+		vcpu->arch.debugfs_exit_timing = NULL;
+	}
+}

arch/powerpc/kvm/timing.h

@@ -0,0 +1,102 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Copyright IBM Corp. 2008
+ *
+ * Authors: Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>
+ */
+
+#ifndef __POWERPC_KVM_EXITTIMING_H__
+#define __POWERPC_KVM_EXITTIMING_H__
+
+#include <linux/kvm_host.h>
+#include <asm/kvm_host.h>
+
+#ifdef CONFIG_KVM_EXIT_TIMING
+void kvmppc_init_timing_stats(struct kvm_vcpu *vcpu);
+void kvmppc_update_timing_stats(struct kvm_vcpu *vcpu);
+void kvmppc_create_vcpu_debugfs(struct kvm_vcpu *vcpu, unsigned int id);
+void kvmppc_remove_vcpu_debugfs(struct kvm_vcpu *vcpu);
+
+static inline void kvmppc_set_exit_type(struct kvm_vcpu *vcpu, int type)
+{
+	vcpu->arch.last_exit_type = type;
+}
+
+#else
+/* if exit timing is not configured there is no need to build the c file */
+static inline void kvmppc_init_timing_stats(struct kvm_vcpu *vcpu) {}
+static inline void kvmppc_update_timing_stats(struct kvm_vcpu *vcpu) {}
+static inline void kvmppc_create_vcpu_debugfs(struct kvm_vcpu *vcpu,
+						unsigned int id) {}
+static inline void kvmppc_remove_vcpu_debugfs(struct kvm_vcpu *vcpu) {}
+static inline void kvmppc_set_exit_type(struct kvm_vcpu *vcpu, int type) {}
+#endif /* CONFIG_KVM_EXIT_TIMING */
+
+/* account the exit in kvm_stats */
+static inline void kvmppc_account_exit_stat(struct kvm_vcpu *vcpu, int type)
+{
+	/* type has to be known at build time for optimization */
+	BUILD_BUG_ON(__builtin_constant_p(type));
+	switch (type) {
+	case EXT_INTR_EXITS:
+		vcpu->stat.ext_intr_exits++;
+		break;
+	case DEC_EXITS:
+		vcpu->stat.dec_exits++;
+		break;
+	case EMULATED_INST_EXITS:
+		vcpu->stat.emulated_inst_exits++;
+		break;
+	case DCR_EXITS:
+		vcpu->stat.dcr_exits++;
+		break;
+	case DSI_EXITS:
+		vcpu->stat.dsi_exits++;
+		break;
+	case ISI_EXITS:
+		vcpu->stat.isi_exits++;
+		break;
+	case SYSCALL_EXITS:
+		vcpu->stat.syscall_exits++;
+		break;
+	case DTLB_REAL_MISS_EXITS:
+		vcpu->stat.dtlb_real_miss_exits++;
+		break;
+	case DTLB_VIRT_MISS_EXITS:
+		vcpu->stat.dtlb_virt_miss_exits++;
+		break;
+	case MMIO_EXITS:
+		vcpu->stat.mmio_exits++;
+		break;
+	case ITLB_REAL_MISS_EXITS:
+		vcpu->stat.itlb_real_miss_exits++;
+		break;
+	case ITLB_VIRT_MISS_EXITS:
+		vcpu->stat.itlb_virt_miss_exits++;
+		break;
+	case SIGNAL_EXITS:
+		vcpu->stat.signal_exits++;
+		break;
+	}
+}
+
+/* wrapper to set exit time and account for it in kvm_stats */
+static inline void kvmppc_account_exit(struct kvm_vcpu *vcpu, int type)
+{
+	kvmppc_set_exit_type(vcpu, type);
+	kvmppc_account_exit_stat(vcpu, type);
+}
+
+#endif /* __POWERPC_KVM_EXITTIMING_H__ */

arch/s390/kvm/kvm-s390.c

@@ -113,8 +113,6 @@ long kvm_arch_dev_ioctl(struct file *filp,
 int kvm_dev_ioctl_check_extension(long ext)
 {
 	switch (ext) {
-	case KVM_CAP_USER_MEMORY:
-		return 1;
 	default:
 		return 0;
 	}
@@ -185,8 +183,6 @@ struct kvm *kvm_arch_create_vm(void)
 	debug_register_view(kvm->arch.dbf, &debug_sprintf_view);
 	VM_EVENT(kvm, 3, "%s", "vm created");
 
-	try_module_get(THIS_MODULE);
-
 	return kvm;
 out_nodbf:
 	free_page((unsigned long)(kvm->arch.sca));
@@ -196,13 +192,33 @@ out_nokvm:
 	return ERR_PTR(rc);
 }
 
+void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
+{
+	VCPU_EVENT(vcpu, 3, "%s", "free cpu");
+	free_page((unsigned long)(vcpu->arch.sie_block));
+	kvm_vcpu_uninit(vcpu);
+	kfree(vcpu);
+}
+
+static void kvm_free_vcpus(struct kvm *kvm)
+{
+	unsigned int i;
+
+	for (i = 0; i < KVM_MAX_VCPUS; ++i) {
+		if (kvm->vcpus[i]) {
+			kvm_arch_vcpu_destroy(kvm->vcpus[i]);
+			kvm->vcpus[i] = NULL;
+		}
+	}
+}
+
 void kvm_arch_destroy_vm(struct kvm *kvm)
 {
-	debug_unregister(kvm->arch.dbf);
+	kvm_free_vcpus(kvm);
 	kvm_free_physmem(kvm);
 	free_page((unsigned long)(kvm->arch.sca));
+	debug_unregister(kvm->arch.dbf);
 	kfree(kvm);
-	module_put(THIS_MODULE);
 }
 
 /* Section: vcpu related */
@@ -213,8 +229,7 @@ int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
 
 void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu)
 {
-	/* kvm common code refers to this, but does'nt call it */
-	BUG();
+	/* Nothing todo */
 }
 
 void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
@@ -308,8 +323,6 @@ struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm,
 	VM_EVENT(kvm, 3, "create cpu %d at %p, sie block at %p", id, vcpu,
 		 vcpu->arch.sie_block);
 
-	try_module_get(THIS_MODULE);
-
 	return vcpu;
 out_free_cpu:
 	kfree(vcpu);
@@ -317,14 +330,6 @@ out_nomem:
 	return ERR_PTR(rc);
 }
 
-void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
-{
-	VCPU_EVENT(vcpu, 3, "%s", "destroy cpu");
-	free_page((unsigned long)(vcpu->arch.sie_block));
-	kfree(vcpu);
-	module_put(THIS_MODULE);
-}
-
 int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu)
 {
 	/* kvm common code refers to this, but never calls it */

arch/x86/include/asm/kvm_host.h

@@ -21,6 +21,7 @@
 
 #include <asm/pvclock-abi.h>
 #include <asm/desc.h>
+#include <asm/mtrr.h>
 
 #define KVM_MAX_VCPUS 16
 #define KVM_MEMORY_SLOTS 32
@@ -86,6 +87,7 @@
 #define KVM_MIN_FREE_MMU_PAGES 5
 #define KVM_REFILL_PAGES 25
 #define KVM_MAX_CPUID_ENTRIES 40
+#define KVM_NR_FIXED_MTRR_REGION 88
 #define KVM_NR_VAR_MTRR 8
 
 extern spinlock_t kvm_lock;
@@ -180,6 +182,8 @@ struct kvm_mmu_page {
 	struct list_head link;
 	struct hlist_node hash_link;
 
+	struct list_head oos_link;
+
 	/*
 	 * The following two entries are used to key the shadow page in the
 	 * hash table.
@@ -190,13 +194,16 @@ struct kvm_mmu_page {
 	u64 *spt;
 	/* hold the gfn of each spte inside spt */
 	gfn_t *gfns;
-	unsigned long slot_bitmap; /* One bit set per slot which has memory
-				    * in this shadow page.
-				    */
+	/*
+	 * One bit set per slot which has memory
+	 * in this shadow page.
+	 */
+	DECLARE_BITMAP(slot_bitmap, KVM_MEMORY_SLOTS + KVM_PRIVATE_MEM_SLOTS);
 	int multimapped;         /* More than one parent_pte? */
 	int root_count;          /* Currently serving as active root */
 	bool unsync;
-	bool unsync_children;
+	bool global;
+	unsigned int unsync_children;
 	union {
 		u64 *parent_pte;               /* !multimapped */
 		struct hlist_head parent_ptes; /* multimapped, kvm_pte_chain */
@@ -327,8 +334,10 @@ struct kvm_vcpu_arch {
 
 	bool nmi_pending;
 	bool nmi_injected;
+	bool nmi_window_open;
 
-	u64 mtrr[0x100];
+	struct mtrr_state_type mtrr_state;
+	u32 pat;
 };
 
 struct kvm_mem_alias {
@@ -350,11 +359,13 @@ struct kvm_arch{
 	 */
 	struct list_head active_mmu_pages;
 	struct list_head assigned_dev_head;
+	struct list_head oos_global_pages;
 	struct dmar_domain *intel_iommu_domain;
 	struct kvm_pic *vpic;
 	struct kvm_ioapic *vioapic;
 	struct kvm_pit *vpit;
 	struct hlist_head irq_ack_notifier_list;
+	int vapics_in_nmi_mode;
 
 	int round_robin_prev_vcpu;
 	unsigned int tss_addr;
@@ -378,6 +389,7 @@ struct kvm_vm_stat {
 	u32 mmu_recycled;
 	u32 mmu_cache_miss;
 	u32 mmu_unsync;
+	u32 mmu_unsync_global;
 	u32 remote_tlb_flush;
 	u32 lpages;
 };
@@ -397,6 +409,7 @@ struct kvm_vcpu_stat {
 	u32 halt_exits;
 	u32 halt_wakeup;
 	u32 request_irq_exits;
+	u32 request_nmi_exits;
 	u32 irq_exits;
 	u32 host_state_reload;
 	u32 efer_reload;
@@ -405,6 +418,7 @@ struct kvm_vcpu_stat {
 	u32 insn_emulation_fail;
 	u32 hypercalls;
 	u32 irq_injections;
+	u32 nmi_injections;
 };
 
 struct descriptor_table {
@@ -477,6 +491,7 @@ struct kvm_x86_ops {
 
 	int (*set_tss_addr)(struct kvm *kvm, unsigned int addr);
 	int (*get_tdp_level)(void);
+	int (*get_mt_mask_shift)(void);
 };
 
 extern struct kvm_x86_ops *kvm_x86_ops;
@@ -490,7 +505,7 @@ int kvm_mmu_setup(struct kvm_vcpu *vcpu);
 void kvm_mmu_set_nonpresent_ptes(u64 trap_pte, u64 notrap_pte);
 void kvm_mmu_set_base_ptes(u64 base_pte);
 void kvm_mmu_set_mask_ptes(u64 user_mask, u64 accessed_mask,
-		u64 dirty_mask, u64 nx_mask, u64 x_mask);
+		u64 dirty_mask, u64 nx_mask, u64 x_mask, u64 mt_mask);
 
 int kvm_mmu_reset_context(struct kvm_vcpu *vcpu);
 void kvm_mmu_slot_remove_write_access(struct kvm *kvm, int slot);
@@ -587,12 +602,14 @@ unsigned long segment_base(u16 selector);
 
 void kvm_mmu_flush_tlb(struct kvm_vcpu *vcpu);
 void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
-		       const u8 *new, int bytes);
+		       const u8 *new, int bytes,
+		       bool guest_initiated);
 int kvm_mmu_unprotect_page_virt(struct kvm_vcpu *vcpu, gva_t gva);
 void __kvm_mmu_free_some_pages(struct kvm_vcpu *vcpu);
 int kvm_mmu_load(struct kvm_vcpu *vcpu);
 void kvm_mmu_unload(struct kvm_vcpu *vcpu);
 void kvm_mmu_sync_roots(struct kvm_vcpu *vcpu);
+void kvm_mmu_sync_global(struct kvm_vcpu *vcpu);
 
 int kvm_emulate_hypercall(struct kvm_vcpu *vcpu);
 
@@ -607,6 +624,8 @@ void kvm_disable_tdp(void);
 int load_pdptrs(struct kvm_vcpu *vcpu, unsigned long cr3);
 int complete_pio(struct kvm_vcpu *vcpu);
 
+struct kvm_memory_slot *gfn_to_memslot_unaliased(struct kvm *kvm, gfn_t gfn);
+
 static inline struct kvm_mmu_page *page_header(hpa_t shadow_page)
 {
 	struct page *page = pfn_to_page(shadow_page >> PAGE_SHIFT);
@@ -702,18 +721,6 @@ static inline void kvm_inject_gp(struct kvm_vcpu *vcpu, u32 error_code)
 	kvm_queue_exception_e(vcpu, GP_VECTOR, error_code);
 }
 
-#define ASM_VMX_VMCLEAR_RAX       ".byte 0x66, 0x0f, 0xc7, 0x30"
-#define ASM_VMX_VMLAUNCH          ".byte 0x0f, 0x01, 0xc2"
-#define ASM_VMX_VMRESUME          ".byte 0x0f, 0x01, 0xc3"
-#define ASM_VMX_VMPTRLD_RAX       ".byte 0x0f, 0xc7, 0x30"
-#define ASM_VMX_VMREAD_RDX_RAX    ".byte 0x0f, 0x78, 0xd0"
-#define ASM_VMX_VMWRITE_RAX_RDX   ".byte 0x0f, 0x79, 0xd0"
-#define ASM_VMX_VMWRITE_RSP_RDX   ".byte 0x0f, 0x79, 0xd4"
-#define ASM_VMX_VMXOFF            ".byte 0x0f, 0x01, 0xc4"
-#define ASM_VMX_VMXON_RAX         ".byte 0xf3, 0x0f, 0xc7, 0x30"
-#define ASM_VMX_INVEPT		  ".byte 0x66, 0x0f, 0x38, 0x80, 0x08"
-#define ASM_VMX_INVVPID		  ".byte 0x66, 0x0f, 0x38, 0x81, 0x08"
-
 #define MSR_IA32_TIME_STAMP_COUNTER		0x010
 
 #define TSS_IOPB_BASE_OFFSET 0x66

arch/x86/include/asm/kvm_x86_emulate.h

@@ -123,6 +123,7 @@ struct decode_cache {
 	u8 ad_bytes;
 	u8 rex_prefix;
 	struct operand src;
+	struct operand src2;
 	struct operand dst;
 	bool has_seg_override;
 	u8 seg_override;
@@ -146,22 +147,18 @@ struct x86_emulate_ctxt {
 	/* Register state before/after emulation. */
 	struct kvm_vcpu *vcpu;
 
-	/* Linear faulting address (if emulating a page-faulting instruction) */
 	unsigned long eflags;
-
 	/* Emulated execution mode, represented by an X86EMUL_MODE value. */
 	int mode;
-
 	u32 cs_base;
 
 	/* decode cache */
-
 	struct decode_cache decode;
 };
 
 /* Repeat String Operation Prefix */
-#define REPE_PREFIX  1
-#define REPNE_PREFIX    2
+#define REPE_PREFIX	1
+#define REPNE_PREFIX	2
 
 /* Execution mode, passed to the emulator. */
 #define X86EMUL_MODE_REAL     0	/* Real mode.             */
@@ -170,7 +167,7 @@ struct x86_emulate_ctxt {
 #define X86EMUL_MODE_PROT64   8	/* 64-bit (long) mode.    */
 
 /* Host execution mode. */
-#if defined(__i386__)
+#if defined(CONFIG_X86_32)
 #define X86EMUL_MODE_HOST X86EMUL_MODE_PROT32
 #elif defined(CONFIG_X86_64)
 #define X86EMUL_MODE_HOST X86EMUL_MODE_PROT64

arch/x86/include/asm/mtrr.h

@@ -57,6 +57,31 @@ struct mtrr_gentry {
 };
 #endif /* !__i386__ */
 
+struct mtrr_var_range {
+	u32 base_lo;
+	u32 base_hi;
+	u32 mask_lo;
+	u32 mask_hi;
+};
+
+/* In the Intel processor's MTRR interface, the MTRR type is always held in
+   an 8 bit field: */
+typedef u8 mtrr_type;
+
+#define MTRR_NUM_FIXED_RANGES 88
+#define MTRR_MAX_VAR_RANGES 256
+
+struct mtrr_state_type {
+	struct mtrr_var_range var_ranges[MTRR_MAX_VAR_RANGES];
+	mtrr_type fixed_ranges[MTRR_NUM_FIXED_RANGES];
+	unsigned char enabled;
+	unsigned char have_fixed;
+	mtrr_type def_type;
+};
+
+#define MTRRphysBase_MSR(reg) (0x200 + 2 * (reg))
+#define MTRRphysMask_MSR(reg) (0x200 + 2 * (reg) + 1)
+
 /*  These are the various ioctls  */
 #define MTRRIOC_ADD_ENTRY        _IOW(MTRR_IOCTL_BASE,  0, struct mtrr_sentry)
 #define MTRRIOC_SET_ENTRY        _IOW(MTRR_IOCTL_BASE,  1, struct mtrr_sentry)

arch/x86/include/asm/virtext.h

@@ -0,0 +1,132 @@
+/* CPU virtualization extensions handling
+ *
+ * This should carry the code for handling CPU virtualization extensions
+ * that needs to live in the kernel core.
+ *
+ * Author: Eduardo Habkost <ehabkost@redhat.com>
+ *
+ * Copyright (C) 2008, Red Hat Inc.
+ *
+ * Contains code from KVM, Copyright (C) 2006 Qumranet, Inc.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.  See
+ * the COPYING file in the top-level directory.
+ */
+#ifndef _ASM_X86_VIRTEX_H
+#define _ASM_X86_VIRTEX_H
+
+#include <asm/processor.h>
+#include <asm/system.h>
+
+#include <asm/vmx.h>
+#include <asm/svm.h>
+
+/*
+ * VMX functions:
+ */
+
+static inline int cpu_has_vmx(void)
+{
+	unsigned long ecx = cpuid_ecx(1);
+	return test_bit(5, &ecx); /* CPUID.1:ECX.VMX[bit 5] -> VT */
+}
+
+
+/** Disable VMX on the current CPU
+ *
+ * vmxoff causes a undefined-opcode exception if vmxon was not run
+ * on the CPU previously. Only call this function if you know VMX
+ * is enabled.
+ */
+static inline void cpu_vmxoff(void)
+{
+	asm volatile (ASM_VMX_VMXOFF : : : "cc");
+	write_cr4(read_cr4() & ~X86_CR4_VMXE);
+}
+
+static inline int cpu_vmx_enabled(void)
+{
+	return read_cr4() & X86_CR4_VMXE;
+}
+
+/** Disable VMX if it is enabled on the current CPU
+ *
+ * You shouldn't call this if cpu_has_vmx() returns 0.
+ */
+static inline void __cpu_emergency_vmxoff(void)
+{
+	if (cpu_vmx_enabled())
+		cpu_vmxoff();
+}
+
+/** Disable VMX if it is supported and enabled on the current CPU
+ */
+static inline void cpu_emergency_vmxoff(void)
+{
+	if (cpu_has_vmx())
+		__cpu_emergency_vmxoff();
+}
+
+
+
+
+/*
+ * SVM functions:
+ */
+
+/** Check if the CPU has SVM support
+ *
+ * You can use the 'msg' arg to get a message describing the problem,
+ * if the function returns zero. Simply pass NULL if you are not interested
+ * on the messages; gcc should take care of not generating code for
+ * the messages on this case.
+ */
+static inline int cpu_has_svm(const char **msg)
+{
+	uint32_t eax, ebx, ecx, edx;
+
+	if (boot_cpu_data.x86_vendor != X86_VENDOR_AMD) {
+		if (msg)
+			*msg = "not amd";
+		return 0;
+	}
+
+	cpuid(0x80000000, &eax, &ebx, &ecx, &edx);
+	if (eax < SVM_CPUID_FUNC) {
+		if (msg)
+			*msg = "can't execute cpuid_8000000a";
+		return 0;
+	}
+
+	cpuid(0x80000001, &eax, &ebx, &ecx, &edx);
+	if (!(ecx & (1 << SVM_CPUID_FEATURE_SHIFT))) {
+		if (msg)
+			*msg = "svm not available";
+		return 0;
+	}
+	return 1;
+}
+
+
+/** Disable SVM on the current CPU
+ *
+ * You should call this only if cpu_has_svm() returned true.
+ */
+static inline void cpu_svm_disable(void)
+{
+	uint64_t efer;
+
+	wrmsrl(MSR_VM_HSAVE_PA, 0);
+	rdmsrl(MSR_EFER, efer);
+	wrmsrl(MSR_EFER, efer & ~MSR_EFER_SVME_MASK);
+}
+
+/** Makes sure SVM is disabled, if it is supported on the CPU
+ */
+static inline void cpu_emergency_svm_disable(void)
+{
+	if (cpu_has_svm(NULL))
+		cpu_svm_disable();
+}
+
+#endif /* _ASM_X86_VIRTEX_H */

arch/x86/kvm/vmx.h → arch/x86/include/asm/vmx.h

@@ -63,10 +63,13 @@
 
 #define VM_EXIT_HOST_ADDR_SPACE_SIZE            0x00000200
 #define VM_EXIT_ACK_INTR_ON_EXIT                0x00008000
+#define VM_EXIT_SAVE_IA32_PAT			0x00040000
+#define VM_EXIT_LOAD_IA32_PAT			0x00080000
 
 #define VM_ENTRY_IA32E_MODE                     0x00000200
 #define VM_ENTRY_SMM                            0x00000400
 #define VM_ENTRY_DEACT_DUAL_MONITOR             0x00000800
+#define VM_ENTRY_LOAD_IA32_PAT			0x00004000
 
 /* VMCS Encodings */
 enum vmcs_field {
@@ -112,6 +115,8 @@ enum vmcs_field {
 	VMCS_LINK_POINTER_HIGH          = 0x00002801,
 	GUEST_IA32_DEBUGCTL             = 0x00002802,
 	GUEST_IA32_DEBUGCTL_HIGH        = 0x00002803,
+	GUEST_IA32_PAT			= 0x00002804,
+	GUEST_IA32_PAT_HIGH		= 0x00002805,
 	GUEST_PDPTR0                    = 0x0000280a,
 	GUEST_PDPTR0_HIGH               = 0x0000280b,
 	GUEST_PDPTR1                    = 0x0000280c,
@@ -120,6 +125,8 @@ enum vmcs_field {
 	GUEST_PDPTR2_HIGH               = 0x0000280f,
 	GUEST_PDPTR3                    = 0x00002810,
 	GUEST_PDPTR3_HIGH               = 0x00002811,
+	HOST_IA32_PAT			= 0x00002c00,
+	HOST_IA32_PAT_HIGH		= 0x00002c01,
 	PIN_BASED_VM_EXEC_CONTROL       = 0x00004000,
 	CPU_BASED_VM_EXEC_CONTROL       = 0x00004002,
 	EXCEPTION_BITMAP                = 0x00004004,
@@ -331,8 +338,9 @@ enum vmcs_field {
 
 #define AR_RESERVD_MASK 0xfffe0f00
 
-#define APIC_ACCESS_PAGE_PRIVATE_MEMSLOT	9
-#define IDENTITY_PAGETABLE_PRIVATE_MEMSLOT	10
+#define TSS_PRIVATE_MEMSLOT			(KVM_MEMORY_SLOTS + 0)
+#define APIC_ACCESS_PAGE_PRIVATE_MEMSLOT	(KVM_MEMORY_SLOTS + 1)
+#define IDENTITY_PAGETABLE_PRIVATE_MEMSLOT	(KVM_MEMORY_SLOTS + 2)
 
 #define VMX_NR_VPIDS				(1 << 16)
 #define VMX_VPID_EXTENT_SINGLE_CONTEXT		1
@@ -356,4 +364,19 @@ enum vmcs_field {
 
 #define VMX_EPT_IDENTITY_PAGETABLE_ADDR		0xfffbc000ul
 
+
+#define ASM_VMX_VMCLEAR_RAX       ".byte 0x66, 0x0f, 0xc7, 0x30"
+#define ASM_VMX_VMLAUNCH          ".byte 0x0f, 0x01, 0xc2"
+#define ASM_VMX_VMRESUME          ".byte 0x0f, 0x01, 0xc3"
+#define ASM_VMX_VMPTRLD_RAX       ".byte 0x0f, 0xc7, 0x30"
+#define ASM_VMX_VMREAD_RDX_RAX    ".byte 0x0f, 0x78, 0xd0"
+#define ASM_VMX_VMWRITE_RAX_RDX   ".byte 0x0f, 0x79, 0xd0"
+#define ASM_VMX_VMWRITE_RSP_RDX   ".byte 0x0f, 0x79, 0xd4"
+#define ASM_VMX_VMXOFF            ".byte 0x0f, 0x01, 0xc4"
+#define ASM_VMX_VMXON_RAX         ".byte 0xf3, 0x0f, 0xc7, 0x30"
+#define ASM_VMX_INVEPT		  ".byte 0x66, 0x0f, 0x38, 0x80, 0x08"
+#define ASM_VMX_INVVPID		  ".byte 0x66, 0x0f, 0x38, 0x81, 0x08"
+
+
+
 #endif

arch/x86/kernel/cpu/mtrr/generic.c

@@ -14,14 +14,6 @@
 #include <asm/pat.h>
 #include "mtrr.h"
 
-struct mtrr_state {
-	struct mtrr_var_range var_ranges[MAX_VAR_RANGES];
-	mtrr_type fixed_ranges[NUM_FIXED_RANGES];
-	unsigned char enabled;
-	unsigned char have_fixed;
-	mtrr_type def_type;
-};
-
 struct fixed_range_block {
 	int base_msr; /* start address of an MTRR block */
 	int ranges;   /* number of MTRRs in this block  */
@@ -35,10 +27,12 @@ static struct fixed_range_block fixed_range_blocks[] = {
 };
 
 static unsigned long smp_changes_mask;
-static struct mtrr_state mtrr_state = {};
 static int mtrr_state_set;
 u64 mtrr_tom2;
 
+struct mtrr_state_type mtrr_state = {};
+EXPORT_SYMBOL_GPL(mtrr_state);
+
 #undef MODULE_PARAM_PREFIX
 #define MODULE_PARAM_PREFIX "mtrr."
 

arch/x86/kernel/cpu/mtrr/main.c

@@ -49,7 +49,7 @@
 
 u32 num_var_ranges = 0;
 
-unsigned int mtrr_usage_table[MAX_VAR_RANGES];
+unsigned int mtrr_usage_table[MTRR_MAX_VAR_RANGES];
 static DEFINE_MUTEX(mtrr_mutex);
 
 u64 size_or_mask, size_and_mask;
@@ -574,7 +574,7 @@ struct mtrr_value {
 	unsigned long	lsize;
 };
 
-static struct mtrr_value mtrr_state[MAX_VAR_RANGES];
+static struct mtrr_value mtrr_state[MTRR_MAX_VAR_RANGES];
 
 static int mtrr_save(struct sys_device * sysdev, pm_message_t state)
 {

arch/x86/kernel/cpu/mtrr/mtrr.h

@@ -8,11 +8,6 @@
 #define MTRRcap_MSR     0x0fe
 #define MTRRdefType_MSR 0x2ff
 
-#define MTRRphysBase_MSR(reg) (0x200 + 2 * (reg))
-#define MTRRphysMask_MSR(reg) (0x200 + 2 * (reg) + 1)
-
-#define NUM_FIXED_RANGES 88
-#define MAX_VAR_RANGES 256
 #define MTRRfix64K_00000_MSR 0x250
 #define MTRRfix16K_80000_MSR 0x258
 #define MTRRfix16K_A0000_MSR 0x259
@@ -29,11 +24,7 @@
 #define MTRR_CHANGE_MASK_VARIABLE  0x02
 #define MTRR_CHANGE_MASK_DEFTYPE   0x04
 
-/* In the Intel processor's MTRR interface, the MTRR type is always held in
-   an 8 bit field: */
-typedef u8 mtrr_type;
-
-extern unsigned int mtrr_usage_table[MAX_VAR_RANGES];
+extern unsigned int mtrr_usage_table[MTRR_MAX_VAR_RANGES];
 
 struct mtrr_ops {
 	u32	vendor;
@@ -70,13 +61,6 @@ struct set_mtrr_context {
 	u32 ccr3;
 };
 
-struct mtrr_var_range {
-	u32 base_lo;
-	u32 base_hi;
-	u32 mask_lo;
-	u32 mask_hi;
-};
-
 void set_mtrr_done(struct set_mtrr_context *ctxt);
 void set_mtrr_cache_disable(struct set_mtrr_context *ctxt);
 void set_mtrr_prepare_save(struct set_mtrr_context *ctxt);

arch/x86/kernel/crash.c

@@ -26,6 +26,7 @@
 #include <linux/kdebug.h>
 #include <asm/smp.h>
 #include <asm/reboot.h>
+#include <asm/virtext.h>
 
 #include <mach_ipi.h>
 
@@ -49,6 +50,15 @@ static void kdump_nmi_callback(int cpu, struct die_args *args)
 #endif
 	crash_save_cpu(regs, cpu);
 
+	/* Disable VMX or SVM if needed.
+	 *
+	 * We need to disable virtualization on all CPUs.
+	 * Having VMX or SVM enabled on any CPU may break rebooting
+	 * after the kdump kernel has finished its task.
+	 */
+	cpu_emergency_vmxoff();
+	cpu_emergency_svm_disable();
+
 	disable_local_APIC();
 }
 
@@ -80,6 +90,14 @@ void native_machine_crash_shutdown(struct pt_regs *regs)
 	local_irq_disable();
 
 	kdump_nmi_shootdown_cpus();
+
+	/* Booting kdump kernel with VMX or SVM enabled won't work,
+	 * because (among other limitations) we can't disable paging
+	 * with the virt flags.
+	 */
+	cpu_emergency_vmxoff();
+	cpu_emergency_svm_disable();
+
 	lapic_shutdown();
 #if defined(CONFIG_X86_IO_APIC)
 	disable_IO_APIC();

arch/x86/kernel/kvmclock.c

@@ -89,17 +89,17 @@ static cycle_t kvm_clock_read(void)
  */
 static unsigned long kvm_get_tsc_khz(void)
 {
-	return preset_lpj;
+	struct pvclock_vcpu_time_info *src;
+	src = &per_cpu(hv_clock, 0);
+	return pvclock_tsc_khz(src);
 }
 
 static void kvm_get_preset_lpj(void)
 {
-	struct pvclock_vcpu_time_info *src;
 	unsigned long khz;
 	u64 lpj;
 
-	src = &per_cpu(hv_clock, 0);
-	khz = pvclock_tsc_khz(src);
+	khz = kvm_get_tsc_khz();
 
 	lpj = ((u64)khz * 1000);
 	do_div(lpj, HZ);
@@ -194,5 +194,7 @@ void __init kvmclock_init(void)
 #endif
 		kvm_get_preset_lpj();
 		clocksource_register(&kvm_clock);
+		pv_info.paravirt_enabled = 1;
+		pv_info.name = "KVM";
 	}
 }

arch/x86/kernel/reboot.c

@@ -12,6 +12,7 @@
 #include <asm/proto.h>
 #include <asm/reboot_fixups.h>
 #include <asm/reboot.h>
+#include <asm/virtext.h>
 
 #ifdef CONFIG_X86_32
 # include <linux/dmi.h>
@@ -39,6 +40,12 @@ int reboot_force;
 static int reboot_cpu = -1;
 #endif
 
+/* This is set if we need to go through the 'emergency' path.
+ * When machine_emergency_restart() is called, we may be on
+ * an inconsistent state and won't be able to do a clean cleanup
+ */
+static int reboot_emergency;
+
 /* This is set by the PCI code if either type 1 or type 2 PCI is detected */
 bool port_cf9_safe = false;
 
@@ -368,6 +375,48 @@ static inline void kb_wait(void)
 	}
 }
 
+static void vmxoff_nmi(int cpu, struct die_args *args)
+{
+	cpu_emergency_vmxoff();
+}
+
+/* Use NMIs as IPIs to tell all CPUs to disable virtualization
+ */
+static void emergency_vmx_disable_all(void)
+{
+	/* Just make sure we won't change CPUs while doing this */
+	local_irq_disable();
+
+	/* We need to disable VMX on all CPUs before rebooting, otherwise
+	 * we risk hanging up the machine, because the CPU ignore INIT
+	 * signals when VMX is enabled.
+	 *
+	 * We can't take any locks and we may be on an inconsistent
+	 * state, so we use NMIs as IPIs to tell the other CPUs to disable
+	 * VMX and halt.
+	 *
+	 * For safety, we will avoid running the nmi_shootdown_cpus()
+	 * stuff unnecessarily, but we don't have a way to check
+	 * if other CPUs have VMX enabled. So we will call it only if the
+	 * CPU we are running on has VMX enabled.
+	 *
+	 * We will miss cases where VMX is not enabled on all CPUs. This
+	 * shouldn't do much harm because KVM always enable VMX on all
+	 * CPUs anyway. But we can miss it on the small window where KVM
+	 * is still enabling VMX.
+	 */
+	if (cpu_has_vmx() && cpu_vmx_enabled()) {
+		/* Disable VMX on this CPU.
+		 */
+		cpu_vmxoff();
+
+		/* Halt and disable VMX on the other CPUs */
+		nmi_shootdown_cpus(vmxoff_nmi);
+
+	}
+}
+
+
 void __attribute__((weak)) mach_reboot_fixups(void)
 {
 }
@@ -376,6 +425,9 @@ static void native_machine_emergency_restart(void)
 {
 	int i;
 
+	if (reboot_emergency)
+		emergency_vmx_disable_all();
+
 	/* Tell the BIOS if we want cold or warm reboot */
 	*((unsigned short *)__va(0x472)) = reboot_mode;
 
@@ -482,13 +534,19 @@ void native_machine_shutdown(void)
 #endif
 }
 
+static void __machine_emergency_restart(int emergency)
+{
+	reboot_emergency = emergency;
+	machine_ops.emergency_restart();
+}
+
 static void native_machine_restart(char *__unused)
 {
 	printk("machine restart\n");
 
 	if (!reboot_force)
 		machine_shutdown();
-	machine_emergency_restart();
+	__machine_emergency_restart(0);
 }
 
 static void native_machine_halt(void)
@@ -532,7 +590,7 @@ void machine_shutdown(void)
 
 void machine_emergency_restart(void)
 {
-	machine_ops.emergency_restart();
+	__machine_emergency_restart(1);
 }
 
 void machine_restart(char *cmd)

arch/x86/kvm/i8254.c

@@ -603,10 +603,29 @@ void kvm_free_pit(struct kvm *kvm)
 
 static void __inject_pit_timer_intr(struct kvm *kvm)
 {
+	struct kvm_vcpu *vcpu;
+	int i;
+
 	mutex_lock(&kvm->lock);
 	kvm_set_irq(kvm, kvm->arch.vpit->irq_source_id, 0, 1);
 	kvm_set_irq(kvm, kvm->arch.vpit->irq_source_id, 0, 0);
 	mutex_unlock(&kvm->lock);
+
+	/*
+	 * Provides NMI watchdog support via Virtual Wire mode.
+	 * The route is: PIT -> PIC -> LVT0 in NMI mode.
+	 *
+	 * Note: Our Virtual Wire implementation is simplified, only
+	 * propagating PIT interrupts to all VCPUs when they have set
+	 * LVT0 to NMI delivery. Other PIC interrupts are just sent to
+	 * VCPU0, and only if its LVT0 is in EXTINT mode.
+	 */
+	if (kvm->arch.vapics_in_nmi_mode > 0)
+		for (i = 0; i < KVM_MAX_VCPUS; ++i) {
+			vcpu = kvm->vcpus[i];
+			if (vcpu)
+				kvm_apic_nmi_wd_deliver(vcpu);
+		}
 }
 
 void kvm_inject_pit_timer_irqs(struct kvm_vcpu *vcpu)

arch/x86/kvm/i8259.c

@@ -26,10 +26,40 @@
  *   Port from Qemu.
  */
 #include <linux/mm.h>
+#include <linux/bitops.h>
 #include "irq.h"
 
 #include <linux/kvm_host.h>
 
+static void pic_lock(struct kvm_pic *s)
+{
+	spin_lock(&s->lock);
+}
+
+static void pic_unlock(struct kvm_pic *s)
+{
+	struct kvm *kvm = s->kvm;
+	unsigned acks = s->pending_acks;
+	bool wakeup = s->wakeup_needed;
+	struct kvm_vcpu *vcpu;
+
+	s->pending_acks = 0;
+	s->wakeup_needed = false;
+
+	spin_unlock(&s->lock);
+
+	while (acks) {
+		kvm_notify_acked_irq(kvm, __ffs(acks));
+		acks &= acks - 1;
+	}
+
+	if (wakeup) {
+		vcpu = s->kvm->vcpus[0];
+		if (vcpu)
+			kvm_vcpu_kick(vcpu);
+	}
+}
+
 static void pic_clear_isr(struct kvm_kpic_state *s, int irq)
 {
 	s->isr &= ~(1 << irq);
@@ -136,17 +166,21 @@ static void pic_update_irq(struct kvm_pic *s)
 
 void kvm_pic_update_irq(struct kvm_pic *s)
 {
+	pic_lock(s);
 	pic_update_irq(s);
+	pic_unlock(s);
 }
 
 void kvm_pic_set_irq(void *opaque, int irq, int level)
 {
 	struct kvm_pic *s = opaque;
 
+	pic_lock(s);
 	if (irq >= 0 && irq < PIC_NUM_PINS) {
 		pic_set_irq1(&s->pics[irq >> 3], irq & 7, level);
 		pic_update_irq(s);
 	}
+	pic_unlock(s);
 }
 
 /*
@@ -172,6 +206,7 @@ int kvm_pic_read_irq(struct kvm *kvm)
 	int irq, irq2, intno;
 	struct kvm_pic *s = pic_irqchip(kvm);
 
+	pic_lock(s);
 	irq = pic_get_irq(&s->pics[0]);
 	if (irq >= 0) {
 		pic_intack(&s->pics[0], irq);
@@ -196,6 +231,7 @@ int kvm_pic_read_irq(struct kvm *kvm)
 		intno = s->pics[0].irq_base + irq;
 	}
 	pic_update_irq(s);
+	pic_unlock(s);
 	kvm_notify_acked_irq(kvm, irq);
 
 	return intno;
@@ -203,7 +239,7 @@ int kvm_pic_read_irq(struct kvm *kvm)
 
 void kvm_pic_reset(struct kvm_kpic_state *s)
 {
-	int irq, irqbase;
+	int irq, irqbase, n;
 	struct kvm *kvm = s->pics_state->irq_request_opaque;
 	struct kvm_vcpu *vcpu0 = kvm->vcpus[0];
 
@@ -214,8 +250,10 @@ void kvm_pic_reset(struct kvm_kpic_state *s)
 
 	for (irq = 0; irq < PIC_NUM_PINS/2; irq++) {
 		if (vcpu0 && kvm_apic_accept_pic_intr(vcpu0))
-			if (s->irr & (1 << irq) || s->isr & (1 << irq))
-				kvm_notify_acked_irq(kvm, irq+irqbase);
+			if (s->irr & (1 << irq) || s->isr & (1 << irq)) {
+				n = irq + irqbase;
+				s->pics_state->pending_acks |= 1 << n;
+			}
 	}
 	s->last_irr = 0;
 	s->irr = 0;
@@ -406,6 +444,7 @@ static void picdev_write(struct kvm_io_device *this,
 			printk(KERN_ERR "PIC: non byte write\n");
 		return;
 	}
+	pic_lock(s);
 	switch (addr) {
 	case 0x20:
 	case 0x21:
@@ -418,6 +457,7 @@ static void picdev_write(struct kvm_io_device *this,
 		elcr_ioport_write(&s->pics[addr & 1], addr, data);
 		break;
 	}
+	pic_unlock(s);
 }
 
 static void picdev_read(struct kvm_io_device *this,
@@ -431,6 +471,7 @@ static void picdev_read(struct kvm_io_device *this,
 			printk(KERN_ERR "PIC: non byte read\n");
 		return;
 	}
+	pic_lock(s);
 	switch (addr) {
 	case 0x20:
 	case 0x21:
@@ -444,6 +485,7 @@ static void picdev_read(struct kvm_io_device *this,
 		break;
 	}
 	*(unsigned char *)val = data;
+	pic_unlock(s);
 }
 
 /*
@@ -459,7 +501,7 @@ static void pic_irq_request(void *opaque, int level)
 	s->output = level;
 	if (vcpu && level && (s->pics[0].isr_ack & (1 << irq))) {
 		s->pics[0].isr_ack &= ~(1 << irq);
-		kvm_vcpu_kick(vcpu);
+		s->wakeup_needed = true;
 	}
 }
 
@@ -469,6 +511,8 @@ struct kvm_pic *kvm_create_pic(struct kvm *kvm)
 	s = kzalloc(sizeof(struct kvm_pic), GFP_KERNEL);
 	if (!s)
 		return NULL;
+	spin_lock_init(&s->lock);
+	s->kvm = kvm;
 	s->pics[0].elcr_mask = 0xf8;
 	s->pics[1].elcr_mask = 0xde;
 	s->irq_request = pic_irq_request;

arch/x86/kvm/irq.h

@@ -25,6 +25,7 @@
 #include <linux/mm_types.h>
 #include <linux/hrtimer.h>
 #include <linux/kvm_host.h>
+#include <linux/spinlock.h>
 
 #include "iodev.h"
 #include "ioapic.h"
@@ -59,6 +60,10 @@ struct kvm_kpic_state {
 };
 
 struct kvm_pic {
+	spinlock_t lock;
+	bool wakeup_needed;
+	unsigned pending_acks;
+	struct kvm *kvm;
 	struct kvm_kpic_state pics[2]; /* 0 is master pic, 1 is slave pic */
 	irq_request_func *irq_request;
 	void *irq_request_opaque;
@@ -87,6 +92,7 @@ void kvm_pic_reset(struct kvm_kpic_state *s);
 void kvm_timer_intr_post(struct kvm_vcpu *vcpu, int vec);
 void kvm_inject_pending_timer_irqs(struct kvm_vcpu *vcpu);
 void kvm_inject_apic_timer_irqs(struct kvm_vcpu *vcpu);
+void kvm_apic_nmi_wd_deliver(struct kvm_vcpu *vcpu);
 void __kvm_migrate_apic_timer(struct kvm_vcpu *vcpu);
 void __kvm_migrate_pit_timer(struct kvm_vcpu *vcpu);
 void __kvm_migrate_timers(struct kvm_vcpu *vcpu);

arch/x86/kvm/kvm_svm.h

@@ -7,7 +7,7 @@
 #include <linux/kvm_host.h>
 #include <asm/msr.h>
 
-#include "svm.h"
+#include <asm/svm.h>
 
 static const u32 host_save_user_msrs[] = {
 #ifdef CONFIG_X86_64

arch/x86/kvm/lapic.c

@@ -130,6 +130,11 @@ static inline int apic_lvtt_period(struct kvm_lapic *apic)
 	return apic_get_reg(apic, APIC_LVTT) & APIC_LVT_TIMER_PERIODIC;
 }
 
+static inline int apic_lvt_nmi_mode(u32 lvt_val)
+{
+	return (lvt_val & (APIC_MODE_MASK | APIC_LVT_MASKED)) == APIC_DM_NMI;
+}
+
 static unsigned int apic_lvt_mask[APIC_LVT_NUM] = {
 	LVT_MASK | APIC_LVT_TIMER_PERIODIC,	/* LVTT */
 	LVT_MASK | APIC_MODE_MASK,	/* LVTTHMR */
@@ -354,6 +359,7 @@ static int __apic_accept_irq(struct kvm_lapic *apic, int delivery_mode,
 
 	case APIC_DM_NMI:
 		kvm_inject_nmi(vcpu);
+		kvm_vcpu_kick(vcpu);
 		break;
 
 	case APIC_DM_INIT:
@@ -380,6 +386,14 @@ static int __apic_accept_irq(struct kvm_lapic *apic, int delivery_mode,
 		}
 		break;
 
+	case APIC_DM_EXTINT:
+		/*
+		 * Should only be called by kvm_apic_local_deliver() with LVT0,
+		 * before NMI watchdog was enabled. Already handled by
+		 * kvm_apic_accept_pic_intr().
+		 */
+		break;
+
 	default:
 		printk(KERN_ERR "TODO: unsupported delivery mode %x\n",
 		       delivery_mode);
@@ -663,6 +677,20 @@ static void start_apic_timer(struct kvm_lapic *apic)
 					apic->timer.period)));
 }
 
+static void apic_manage_nmi_watchdog(struct kvm_lapic *apic, u32 lvt0_val)
+{
+	int nmi_wd_enabled = apic_lvt_nmi_mode(apic_get_reg(apic, APIC_LVT0));
+
+	if (apic_lvt_nmi_mode(lvt0_val)) {
+		if (!nmi_wd_enabled) {
+			apic_debug("Receive NMI setting on APIC_LVT0 "
+				   "for cpu %d\n", apic->vcpu->vcpu_id);
+			apic->vcpu->kvm->arch.vapics_in_nmi_mode++;
+		}
+	} else if (nmi_wd_enabled)
+		apic->vcpu->kvm->arch.vapics_in_nmi_mode--;
+}
+
 static void apic_mmio_write(struct kvm_io_device *this,
 			    gpa_t address, int len, const void *data)
 {
@@ -743,10 +771,11 @@ static void apic_mmio_write(struct kvm_io_device *this,
 		apic_set_reg(apic, APIC_ICR2, val & 0xff000000);
 		break;
 
+	case APIC_LVT0:
+		apic_manage_nmi_watchdog(apic, val);
 	case APIC_LVTT:
 	case APIC_LVTTHMR:
 	case APIC_LVTPC:
-	case APIC_LVT0:
 	case APIC_LVT1:
 	case APIC_LVTERR:
 		/* TODO: Check vector */
@@ -961,12 +990,26 @@ int apic_has_pending_timer(struct kvm_vcpu *vcpu)
 	return 0;
 }
 
-static int __inject_apic_timer_irq(struct kvm_lapic *apic)
+static int kvm_apic_local_deliver(struct kvm_lapic *apic, int lvt_type)
+{
+	u32 reg = apic_get_reg(apic, lvt_type);
+	int vector, mode, trig_mode;
+
+	if (apic_hw_enabled(apic) && !(reg & APIC_LVT_MASKED)) {
+		vector = reg & APIC_VECTOR_MASK;
+		mode = reg & APIC_MODE_MASK;
+		trig_mode = reg & APIC_LVT_LEVEL_TRIGGER;
+		return __apic_accept_irq(apic, mode, vector, 1, trig_mode);
+	}
+	return 0;
+}
+
+void kvm_apic_nmi_wd_deliver(struct kvm_vcpu *vcpu)
 {
-	int vector;
+	struct kvm_lapic *apic = vcpu->arch.apic;
 
-	vector = apic_lvt_vector(apic, APIC_LVTT);
-	return __apic_accept_irq(apic, APIC_DM_FIXED, vector, 1, 0);
+	if (apic)
+		kvm_apic_local_deliver(apic, APIC_LVT0);
 }
 
 static enum hrtimer_restart apic_timer_fn(struct hrtimer *data)
@@ -1061,9 +1104,8 @@ void kvm_inject_apic_timer_irqs(struct kvm_vcpu *vcpu)
 {
 	struct kvm_lapic *apic = vcpu->arch.apic;
 
-	if (apic && apic_lvt_enabled(apic, APIC_LVTT) &&
-		atomic_read(&apic->timer.pending) > 0) {
-		if (__inject_apic_timer_irq(apic))
+	if (apic && atomic_read(&apic->timer.pending) > 0) {
+		if (kvm_apic_local_deliver(apic, APIC_LVTT))
 			atomic_dec(&apic->timer.pending);
 	}
 }

arch/x86/kvm/mmu.c

@@ -17,7 +17,6 @@
  *
  */
 
-#include "vmx.h"
 #include "mmu.h"
 
 #include <linux/kvm_host.h>
@@ -33,6 +32,7 @@
 #include <asm/page.h>
 #include <asm/cmpxchg.h>
 #include <asm/io.h>
+#include <asm/vmx.h>
 
 /*
  * When setting this variable to true it enables Two-Dimensional-Paging
@@ -168,6 +168,7 @@ static u64 __read_mostly shadow_x_mask;	/* mutual exclusive with nx_mask */
 static u64 __read_mostly shadow_user_mask;
 static u64 __read_mostly shadow_accessed_mask;
 static u64 __read_mostly shadow_dirty_mask;
+static u64 __read_mostly shadow_mt_mask;
 
 void kvm_mmu_set_nonpresent_ptes(u64 trap_pte, u64 notrap_pte)
 {
@@ -183,13 +184,14 @@ void kvm_mmu_set_base_ptes(u64 base_pte)
 EXPORT_SYMBOL_GPL(kvm_mmu_set_base_ptes);
 
 void kvm_mmu_set_mask_ptes(u64 user_mask, u64 accessed_mask,
-		u64 dirty_mask, u64 nx_mask, u64 x_mask)
+		u64 dirty_mask, u64 nx_mask, u64 x_mask, u64 mt_mask)
 {
 	shadow_user_mask = user_mask;
 	shadow_accessed_mask = accessed_mask;
 	shadow_dirty_mask = dirty_mask;
 	shadow_nx_mask = nx_mask;
 	shadow_x_mask = x_mask;
+	shadow_mt_mask = mt_mask;
 }
 EXPORT_SYMBOL_GPL(kvm_mmu_set_mask_ptes);
 
@@ -384,7 +386,9 @@ static void account_shadowed(struct kvm *kvm, gfn_t gfn)
 {
 	int *write_count;
 
-	write_count = slot_largepage_idx(gfn, gfn_to_memslot(kvm, gfn));
+	gfn = unalias_gfn(kvm, gfn);
+	write_count = slot_largepage_idx(gfn,
+					 gfn_to_memslot_unaliased(kvm, gfn));
 	*write_count += 1;
 }
 
@@ -392,16 +396,20 @@ static void unaccount_shadowed(struct kvm *kvm, gfn_t gfn)
 {
 	int *write_count;
 
-	write_count = slot_largepage_idx(gfn, gfn_to_memslot(kvm, gfn));
+	gfn = unalias_gfn(kvm, gfn);
+	write_count = slot_largepage_idx(gfn,
+					 gfn_to_memslot_unaliased(kvm, gfn));
 	*write_count -= 1;
 	WARN_ON(*write_count < 0);
 }
 
 static int has_wrprotected_page(struct kvm *kvm, gfn_t gfn)
 {
-	struct kvm_memory_slot *slot = gfn_to_memslot(kvm, gfn);
+	struct kvm_memory_slot *slot;
 	int *largepage_idx;
 
+	gfn = unalias_gfn(kvm, gfn);
+	slot = gfn_to_memslot_unaliased(kvm, gfn);
 	if (slot) {
 		largepage_idx = slot_largepage_idx(gfn, slot);
 		return *largepage_idx;
@@ -613,7 +621,7 @@ static u64 *rmap_next(struct kvm *kvm, unsigned long *rmapp, u64 *spte)
 	return NULL;
 }
 
-static void rmap_write_protect(struct kvm *kvm, u64 gfn)
+static int rmap_write_protect(struct kvm *kvm, u64 gfn)
 {
 	unsigned long *rmapp;
 	u64 *spte;
@@ -659,8 +667,7 @@ static void rmap_write_protect(struct kvm *kvm, u64 gfn)
 		spte = rmap_next(kvm, rmapp, spte);
 	}
 
-	if (write_protected)
-		kvm_flush_remote_tlbs(kvm);
+	return write_protected;
 }
 
 static int kvm_unmap_rmapp(struct kvm *kvm, unsigned long *rmapp)
@@ -786,9 +793,11 @@ static struct kvm_mmu_page *kvm_mmu_alloc_page(struct kvm_vcpu *vcpu,
 	sp->gfns = mmu_memory_cache_alloc(&vcpu->arch.mmu_page_cache, PAGE_SIZE);
 	set_page_private(virt_to_page(sp->spt), (unsigned long)sp);
 	list_add(&sp->link, &vcpu->kvm->arch.active_mmu_pages);
+	INIT_LIST_HEAD(&sp->oos_link);
 	ASSERT(is_empty_shadow_page(sp->spt));
-	sp->slot_bitmap = 0;
+	bitmap_zero(sp->slot_bitmap, KVM_MEMORY_SLOTS + KVM_PRIVATE_MEM_SLOTS);
 	sp->multimapped = 0;
+	sp->global = 1;
 	sp->parent_pte = parent_pte;
 	--vcpu->kvm->arch.n_free_mmu_pages;
 	return sp;
@@ -900,8 +909,9 @@ static void kvm_mmu_update_unsync_bitmap(u64 *spte)
 	struct kvm_mmu_page *sp = page_header(__pa(spte));
 
 	index = spte - sp->spt;
-	__set_bit(index, sp->unsync_child_bitmap);
-	sp->unsync_children = 1;
+	if (!__test_and_set_bit(index, sp->unsync_child_bitmap))
+		sp->unsync_children++;
+	WARN_ON(!sp->unsync_children);
 }
 
 static void kvm_mmu_update_parents_unsync(struct kvm_mmu_page *sp)
@@ -928,7 +938,6 @@ static void kvm_mmu_update_parents_unsync(struct kvm_mmu_page *sp)
 
 static int unsync_walk_fn(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
 {
-	sp->unsync_children = 1;
 	kvm_mmu_update_parents_unsync(sp);
 	return 1;
 }
@@ -959,38 +968,66 @@ static void nonpaging_invlpg(struct kvm_vcpu *vcpu, gva_t gva)
 {
 }
 
+#define KVM_PAGE_ARRAY_NR 16
+
+struct kvm_mmu_pages {
+	struct mmu_page_and_offset {
+		struct kvm_mmu_page *sp;
+		unsigned int idx;
+	} page[KVM_PAGE_ARRAY_NR];
+	unsigned int nr;
+};
+
 #define for_each_unsync_children(bitmap, idx)		\
 	for (idx = find_first_bit(bitmap, 512);		\
 	     idx < 512;					\
 	     idx = find_next_bit(bitmap, 512, idx+1))
 
-static int mmu_unsync_walk(struct kvm_mmu_page *sp,
-			   struct kvm_unsync_walk *walker)
+int mmu_pages_add(struct kvm_mmu_pages *pvec, struct kvm_mmu_page *sp,
+		   int idx)
 {
-	int i, ret;
+	int i;
 
-	if (!sp->unsync_children)
-		return 0;
+	if (sp->unsync)
+		for (i=0; i < pvec->nr; i++)
+			if (pvec->page[i].sp == sp)
+				return 0;
+
+	pvec->page[pvec->nr].sp = sp;
+	pvec->page[pvec->nr].idx = idx;
+	pvec->nr++;
+	return (pvec->nr == KVM_PAGE_ARRAY_NR);
+}
+
+static int __mmu_unsync_walk(struct kvm_mmu_page *sp,
+			   struct kvm_mmu_pages *pvec)
+{
+	int i, ret, nr_unsync_leaf = 0;
 
 	for_each_unsync_children(sp->unsync_child_bitmap, i) {
 		u64 ent = sp->spt[i];
 
-		if (is_shadow_present_pte(ent)) {
+		if (is_shadow_present_pte(ent) && !is_large_pte(ent)) {
 			struct kvm_mmu_page *child;
 			child = page_header(ent & PT64_BASE_ADDR_MASK);
 
 			if (child->unsync_children) {
-				ret = mmu_unsync_walk(child, walker);
-				if (ret)
+				if (mmu_pages_add(pvec, child, i))
+					return -ENOSPC;
+
+				ret = __mmu_unsync_walk(child, pvec);
+				if (!ret)
+					__clear_bit(i, sp->unsync_child_bitmap);
+				else if (ret > 0)
+					nr_unsync_leaf += ret;
+				else
 					return ret;
-				__clear_bit(i, sp->unsync_child_bitmap);
 			}
 
 			if (child->unsync) {
-				ret = walker->entry(child, walker);
-				__clear_bit(i, sp->unsync_child_bitmap);
-				if (ret)
-					return ret;
+				nr_unsync_leaf++;
+				if (mmu_pages_add(pvec, child, i))
+					return -ENOSPC;
 			}
 		}
 	}
@@ -998,7 +1035,17 @@ static int mmu_unsync_walk(struct kvm_mmu_page *sp,
 	if (find_first_bit(sp->unsync_child_bitmap, 512) == 512)
 		sp->unsync_children = 0;
 
-	return 0;
+	return nr_unsync_leaf;
+}
+
+static int mmu_unsync_walk(struct kvm_mmu_page *sp,
+			   struct kvm_mmu_pages *pvec)
+{
+	if (!sp->unsync_children)
+		return 0;
+
+	mmu_pages_add(pvec, sp, 0);
+	return __mmu_unsync_walk(sp, pvec);
 }
 
 static struct kvm_mmu_page *kvm_mmu_lookup_page(struct kvm *kvm, gfn_t gfn)
@@ -1021,10 +1068,18 @@ static struct kvm_mmu_page *kvm_mmu_lookup_page(struct kvm *kvm, gfn_t gfn)
 	return NULL;
 }
 
+static void kvm_unlink_unsync_global(struct kvm *kvm, struct kvm_mmu_page *sp)
+{
+	list_del(&sp->oos_link);
+	--kvm->stat.mmu_unsync_global;
+}
+
 static void kvm_unlink_unsync_page(struct kvm *kvm, struct kvm_mmu_page *sp)
 {
 	WARN_ON(!sp->unsync);
 	sp->unsync = 0;
+	if (sp->global)
+		kvm_unlink_unsync_global(kvm, sp);
 	--kvm->stat.mmu_unsync;
 }
 
@@ -1037,7 +1092,8 @@ static int kvm_sync_page(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
 		return 1;
 	}
 
-	rmap_write_protect(vcpu->kvm, sp->gfn);
+	if (rmap_write_protect(vcpu->kvm, sp->gfn))
+		kvm_flush_remote_tlbs(vcpu->kvm);
 	kvm_unlink_unsync_page(vcpu->kvm, sp);
 	if (vcpu->arch.mmu.sync_page(vcpu, sp)) {
 		kvm_mmu_zap_page(vcpu->kvm, sp);
@@ -1048,30 +1104,89 @@ static int kvm_sync_page(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
 	return 0;
 }
 
-struct sync_walker {
-	struct kvm_vcpu *vcpu;
-	struct kvm_unsync_walk walker;
+struct mmu_page_path {
+	struct kvm_mmu_page *parent[PT64_ROOT_LEVEL-1];
+	unsigned int idx[PT64_ROOT_LEVEL-1];
 };
 
-static int mmu_sync_fn(struct kvm_mmu_page *sp, struct kvm_unsync_walk *walk)
+#define for_each_sp(pvec, sp, parents, i)			\
+		for (i = mmu_pages_next(&pvec, &parents, -1),	\
+			sp = pvec.page[i].sp;			\
+			i < pvec.nr && ({ sp = pvec.page[i].sp; 1;});	\
+			i = mmu_pages_next(&pvec, &parents, i))
+
+int mmu_pages_next(struct kvm_mmu_pages *pvec, struct mmu_page_path *parents,
+		   int i)
 {
-	struct sync_walker *sync_walk = container_of(walk, struct sync_walker,
-						     walker);
-	struct kvm_vcpu *vcpu = sync_walk->vcpu;
+	int n;
 
-	kvm_sync_page(vcpu, sp);
-	return (need_resched() || spin_needbreak(&vcpu->kvm->mmu_lock));
+	for (n = i+1; n < pvec->nr; n++) {
+		struct kvm_mmu_page *sp = pvec->page[n].sp;
+
+		if (sp->role.level == PT_PAGE_TABLE_LEVEL) {
+			parents->idx[0] = pvec->page[n].idx;
+			return n;
+		}
+
+		parents->parent[sp->role.level-2] = sp;
+		parents->idx[sp->role.level-1] = pvec->page[n].idx;
+	}
+
+	return n;
 }
 
-static void mmu_sync_children(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
+void mmu_pages_clear_parents(struct mmu_page_path *parents)
 {
-	struct sync_walker walker = {
-		.walker = { .entry = mmu_sync_fn, },
-		.vcpu = vcpu,
-	};
+	struct kvm_mmu_page *sp;
+	unsigned int level = 0;
+
+	do {
+		unsigned int idx = parents->idx[level];
+
+		sp = parents->parent[level];
+		if (!sp)
+			return;
+
+		--sp->unsync_children;
+		WARN_ON((int)sp->unsync_children < 0);
+		__clear_bit(idx, sp->unsync_child_bitmap);
+		level++;
+	} while (level < PT64_ROOT_LEVEL-1 && !sp->unsync_children);
+}
+
+static void kvm_mmu_pages_init(struct kvm_mmu_page *parent,
+			       struct mmu_page_path *parents,
+			       struct kvm_mmu_pages *pvec)
+{
+	parents->parent[parent->role.level-1] = NULL;
+	pvec->nr = 0;
+}
+
+static void mmu_sync_children(struct kvm_vcpu *vcpu,
+			      struct kvm_mmu_page *parent)
+{
+	int i;
+	struct kvm_mmu_page *sp;
+	struct mmu_page_path parents;
+	struct kvm_mmu_pages pages;
+
+	kvm_mmu_pages_init(parent, &parents, &pages);
+	while (mmu_unsync_walk(parent, &pages)) {
+		int protected = 0;
 
-	while (mmu_unsync_walk(sp, &walker.walker))
+		for_each_sp(pages, sp, parents, i)
+			protected |= rmap_write_protect(vcpu->kvm, sp->gfn);
+
+		if (protected)
+			kvm_flush_remote_tlbs(vcpu->kvm);
+
+		for_each_sp(pages, sp, parents, i) {
+			kvm_sync_page(vcpu, sp);
+			mmu_pages_clear_parents(&parents);
+		}
 		cond_resched_lock(&vcpu->kvm->mmu_lock);
+		kvm_mmu_pages_init(parent, &parents, &pages);
+	}
 }
 
 static struct kvm_mmu_page *kvm_mmu_get_page(struct kvm_vcpu *vcpu,
@@ -1129,7 +1244,8 @@ static struct kvm_mmu_page *kvm_mmu_get_page(struct kvm_vcpu *vcpu,
 	sp->role = role;
 	hlist_add_head(&sp->hash_link, bucket);
 	if (!metaphysical) {
-		rmap_write_protect(vcpu->kvm, gfn);
+		if (rmap_write_protect(vcpu->kvm, gfn))
+			kvm_flush_remote_tlbs(vcpu->kvm);
 		account_shadowed(vcpu->kvm, gfn);
 	}
 	if (shadow_trap_nonpresent_pte != shadow_notrap_nonpresent_pte)
@@ -1153,6 +1269,8 @@ static int walk_shadow(struct kvm_shadow_walk *walker,
 	if (level == PT32E_ROOT_LEVEL) {
 		shadow_addr = vcpu->arch.mmu.pae_root[(addr >> 30) & 3];
 		shadow_addr &= PT64_BASE_ADDR_MASK;
+		if (!shadow_addr)
+			return 1;
 		--level;
 	}
 
@@ -1237,33 +1355,29 @@ static void kvm_mmu_unlink_parents(struct kvm *kvm, struct kvm_mmu_page *sp)
 	}
 }
 
-struct zap_walker {
-	struct kvm_unsync_walk walker;
-	struct kvm *kvm;
-	int zapped;
-};
-
-static int mmu_zap_fn(struct kvm_mmu_page *sp, struct kvm_unsync_walk *walk)
+static int mmu_zap_unsync_children(struct kvm *kvm,
+				   struct kvm_mmu_page *parent)
 {
-	struct zap_walker *zap_walk = container_of(walk, struct zap_walker,
-						     walker);
-	kvm_mmu_zap_page(zap_walk->kvm, sp);
-	zap_walk->zapped = 1;
-	return 0;
-}
+	int i, zapped = 0;
+	struct mmu_page_path parents;
+	struct kvm_mmu_pages pages;
 
-static int mmu_zap_unsync_children(struct kvm *kvm, struct kvm_mmu_page *sp)
-{
-	struct zap_walker walker = {
-		.walker = { .entry = mmu_zap_fn, },
-		.kvm = kvm,
-		.zapped = 0,
-	};
-
-	if (sp->role.level == PT_PAGE_TABLE_LEVEL)
+	if (parent->role.level == PT_PAGE_TABLE_LEVEL)
 		return 0;
-	mmu_unsync_walk(sp, &walker.walker);
-	return walker.zapped;
+
+	kvm_mmu_pages_init(parent, &parents, &pages);
+	while (mmu_unsync_walk(parent, &pages)) {
+		struct kvm_mmu_page *sp;
+
+		for_each_sp(pages, sp, parents, i) {
+			kvm_mmu_zap_page(kvm, sp);
+			mmu_pages_clear_parents(&parents);
+		}
+		zapped += pages.nr;
+		kvm_mmu_pages_init(parent, &parents, &pages);
+	}
+
+	return zapped;
 }
 
 static int kvm_mmu_zap_page(struct kvm *kvm, struct kvm_mmu_page *sp)
@@ -1362,7 +1476,7 @@ static void page_header_update_slot(struct kvm *kvm, void *pte, gfn_t gfn)
 	int slot = memslot_id(kvm, gfn_to_memslot(kvm, gfn));
 	struct kvm_mmu_page *sp = page_header(__pa(pte));
 
-	__set_bit(slot, &sp->slot_bitmap);
+	__set_bit(slot, sp->slot_bitmap);
 }
 
 static void mmu_convert_notrap(struct kvm_mmu_page *sp)
@@ -1393,6 +1507,110 @@ struct page *gva_to_page(struct kvm_vcpu *vcpu, gva_t gva)
 	return page;
 }
 
+/*
+ * The function is based on mtrr_type_lookup() in
+ * arch/x86/kernel/cpu/mtrr/generic.c
+ */
+static int get_mtrr_type(struct mtrr_state_type *mtrr_state,
+			 u64 start, u64 end)
+{
+	int i;
+	u64 base, mask;
+	u8 prev_match, curr_match;
+	int num_var_ranges = KVM_NR_VAR_MTRR;
+
+	if (!mtrr_state->enabled)
+		return 0xFF;
+
+	/* Make end inclusive end, instead of exclusive */
+	end--;
+
+	/* Look in fixed ranges. Just return the type as per start */
+	if (mtrr_state->have_fixed && (start < 0x100000)) {
+		int idx;
+
+		if (start < 0x80000) {
+			idx = 0;
+			idx += (start >> 16);
+			return mtrr_state->fixed_ranges[idx];
+		} else if (start < 0xC0000) {
+			idx = 1 * 8;
+			idx += ((start - 0x80000) >> 14);
+			return mtrr_state->fixed_ranges[idx];
+		} else if (start < 0x1000000) {
+			idx = 3 * 8;
+			idx += ((start - 0xC0000) >> 12);
+			return mtrr_state->fixed_ranges[idx];
+		}
+	}
+
+	/*
+	 * Look in variable ranges
+	 * Look of multiple ranges matching this address and pick type
+	 * as per MTRR precedence
+	 */
+	if (!(mtrr_state->enabled & 2))
+		return mtrr_state->def_type;
+
+	prev_match = 0xFF;
+	for (i = 0; i < num_var_ranges; ++i) {
+		unsigned short start_state, end_state;
+
+		if (!(mtrr_state->var_ranges[i].mask_lo & (1 << 11)))
+			continue;
+
+		base = (((u64)mtrr_state->var_ranges[i].base_hi) << 32) +
+		       (mtrr_state->var_ranges[i].base_lo & PAGE_MASK);
+		mask = (((u64)mtrr_state->var_ranges[i].mask_hi) << 32) +
+		       (mtrr_state->var_ranges[i].mask_lo & PAGE_MASK);
+
+		start_state = ((start & mask) == (base & mask));
+		end_state = ((end & mask) == (base & mask));
+		if (start_state != end_state)
+			return 0xFE;
+
+		if ((start & mask) != (base & mask))
+			continue;
+
+		curr_match = mtrr_state->var_ranges[i].base_lo & 0xff;
+		if (prev_match == 0xFF) {
+			prev_match = curr_match;
+			continue;
+		}
+
+		if (prev_match == MTRR_TYPE_UNCACHABLE ||
+		    curr_match == MTRR_TYPE_UNCACHABLE)
+			return MTRR_TYPE_UNCACHABLE;
+
+		if ((prev_match == MTRR_TYPE_WRBACK &&
+		     curr_match == MTRR_TYPE_WRTHROUGH) ||
+		    (prev_match == MTRR_TYPE_WRTHROUGH &&
+		     curr_match == MTRR_TYPE_WRBACK)) {
+			prev_match = MTRR_TYPE_WRTHROUGH;
+			curr_match = MTRR_TYPE_WRTHROUGH;
+		}
+
+		if (prev_match != curr_match)
+			return MTRR_TYPE_UNCACHABLE;
+	}
+
+	if (prev_match != 0xFF)
+		return prev_match;
+
+	return mtrr_state->def_type;
+}
+
+static u8 get_memory_type(struct kvm_vcpu *vcpu, gfn_t gfn)
+{
+	u8 mtrr;
+
+	mtrr = get_mtrr_type(&vcpu->arch.mtrr_state, gfn << PAGE_SHIFT,
+			     (gfn << PAGE_SHIFT) + PAGE_SIZE);
+	if (mtrr == 0xfe || mtrr == 0xff)
+		mtrr = MTRR_TYPE_WRBACK;
+	return mtrr;
+}
+
 static int kvm_unsync_page(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
 {
 	unsigned index;
@@ -1409,9 +1627,15 @@ static int kvm_unsync_page(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
 		if (s->role.word != sp->role.word)
 			return 1;
 	}
-	kvm_mmu_mark_parents_unsync(vcpu, sp);
 	++vcpu->kvm->stat.mmu_unsync;
 	sp->unsync = 1;
+
+	if (sp->global) {
+		list_add(&sp->oos_link, &vcpu->kvm->arch.oos_global_pages);
+		++vcpu->kvm->stat.mmu_unsync_global;
+	} else
+		kvm_mmu_mark_parents_unsync(vcpu, sp);
+
 	mmu_convert_notrap(sp);
 	return 0;
 }
@@ -1437,11 +1661,24 @@ static int mmu_need_write_protect(struct kvm_vcpu *vcpu, gfn_t gfn,
 static int set_spte(struct kvm_vcpu *vcpu, u64 *shadow_pte,
 		    unsigned pte_access, int user_fault,
 		    int write_fault, int dirty, int largepage,
-		    gfn_t gfn, pfn_t pfn, bool speculative,
+		    int global, gfn_t gfn, pfn_t pfn, bool speculative,
 		    bool can_unsync)
 {
 	u64 spte;
 	int ret = 0;
+	u64 mt_mask = shadow_mt_mask;
+	struct kvm_mmu_page *sp = page_header(__pa(shadow_pte));
+
+	if (!(vcpu->arch.cr4 & X86_CR4_PGE))
+		global = 0;
+	if (!global && sp->global) {
+		sp->global = 0;
+		if (sp->unsync) {
+			kvm_unlink_unsync_global(vcpu->kvm, sp);
+			kvm_mmu_mark_parents_unsync(vcpu, sp);
+		}
+	}
+
 	/*
 	 * We don't set the accessed bit, since we sometimes want to see
 	 * whether the guest actually used the pte (in order to detect
@@ -1460,6 +1697,11 @@ static int set_spte(struct kvm_vcpu *vcpu, u64 *shadow_pte,
 		spte |= shadow_user_mask;
 	if (largepage)
 		spte |= PT_PAGE_SIZE_MASK;
+	if (mt_mask) {
+		mt_mask = get_memory_type(vcpu, gfn) <<
+			  kvm_x86_ops->get_mt_mask_shift();
+		spte |= mt_mask;
+	}
 
 	spte |= (u64)pfn << PAGE_SHIFT;
 
@@ -1474,6 +1716,15 @@ static int set_spte(struct kvm_vcpu *vcpu, u64 *shadow_pte,
 
 		spte |= PT_WRITABLE_MASK;
 
+		/*
+		 * Optimization: for pte sync, if spte was writable the hash
+		 * lookup is unnecessary (and expensive). Write protection
+		 * is responsibility of mmu_get_page / kvm_sync_page.
+		 * Same reasoning can be applied to dirty page accounting.
+		 */
+		if (!can_unsync && is_writeble_pte(*shadow_pte))
+			goto set_pte;
+
 		if (mmu_need_write_protect(vcpu, gfn, can_unsync)) {
 			pgprintk("%s: found shadow page for %lx, marking ro\n",
 				 __func__, gfn);
@@ -1495,8 +1746,8 @@ set_pte:
 static void mmu_set_spte(struct kvm_vcpu *vcpu, u64 *shadow_pte,
 			 unsigned pt_access, unsigned pte_access,
 			 int user_fault, int write_fault, int dirty,
-			 int *ptwrite, int largepage, gfn_t gfn,
-			 pfn_t pfn, bool speculative)
+			 int *ptwrite, int largepage, int global,
+			 gfn_t gfn, pfn_t pfn, bool speculative)
 {
 	int was_rmapped = 0;
 	int was_writeble = is_writeble_pte(*shadow_pte);
@@ -1529,7 +1780,7 @@ static void mmu_set_spte(struct kvm_vcpu *vcpu, u64 *shadow_pte,
 		}
 	}
 	if (set_spte(vcpu, shadow_pte, pte_access, user_fault, write_fault,
-		      dirty, largepage, gfn, pfn, speculative, true)) {
+		      dirty, largepage, global, gfn, pfn, speculative, true)) {
 		if (write_fault)
 			*ptwrite = 1;
 		kvm_x86_ops->tlb_flush(vcpu);
@@ -1586,7 +1837,7 @@ static int direct_map_entry(struct kvm_shadow_walk *_walk,
 	    || (walk->largepage && level == PT_DIRECTORY_LEVEL)) {
 		mmu_set_spte(vcpu, sptep, ACC_ALL, ACC_ALL,
 			     0, walk->write, 1, &walk->pt_write,
-			     walk->largepage, gfn, walk->pfn, false);
+			     walk->largepage, 0, gfn, walk->pfn, false);
 		++vcpu->stat.pf_fixed;
 		return 1;
 	}
@@ -1773,6 +2024,15 @@ static void mmu_sync_roots(struct kvm_vcpu *vcpu)
 	}
 }
 
+static void mmu_sync_global(struct kvm_vcpu *vcpu)
+{
+	struct kvm *kvm = vcpu->kvm;
+	struct kvm_mmu_page *sp, *n;
+
+	list_for_each_entry_safe(sp, n, &kvm->arch.oos_global_pages, oos_link)
+		kvm_sync_page(vcpu, sp);
+}
+
 void kvm_mmu_sync_roots(struct kvm_vcpu *vcpu)
 {
 	spin_lock(&vcpu->kvm->mmu_lock);
@@ -1780,6 +2040,13 @@ void kvm_mmu_sync_roots(struct kvm_vcpu *vcpu)
 	spin_unlock(&vcpu->kvm->mmu_lock);
 }
 
+void kvm_mmu_sync_global(struct kvm_vcpu *vcpu)
+{
+	spin_lock(&vcpu->kvm->mmu_lock);
+	mmu_sync_global(vcpu);
+	spin_unlock(&vcpu->kvm->mmu_lock);
+}
+
 static gpa_t nonpaging_gva_to_gpa(struct kvm_vcpu *vcpu, gva_t vaddr)
 {
 	return vaddr;
@@ -2178,7 +2445,8 @@ static void kvm_mmu_access_page(struct kvm_vcpu *vcpu, gfn_t gfn)
 }
 
 void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
-		       const u8 *new, int bytes)
+		       const u8 *new, int bytes,
+		       bool guest_initiated)
 {
 	gfn_t gfn = gpa >> PAGE_SHIFT;
 	struct kvm_mmu_page *sp;
@@ -2204,15 +2472,17 @@ void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
 	kvm_mmu_free_some_pages(vcpu);
 	++vcpu->kvm->stat.mmu_pte_write;
 	kvm_mmu_audit(vcpu, "pre pte write");
-	if (gfn == vcpu->arch.last_pt_write_gfn
-	    && !last_updated_pte_accessed(vcpu)) {
-		++vcpu->arch.last_pt_write_count;
-		if (vcpu->arch.last_pt_write_count >= 3)
-			flooded = 1;
-	} else {
-		vcpu->arch.last_pt_write_gfn = gfn;
-		vcpu->arch.last_pt_write_count = 1;
-		vcpu->arch.last_pte_updated = NULL;
+	if (guest_initiated) {
+		if (gfn == vcpu->arch.last_pt_write_gfn
+		    && !last_updated_pte_accessed(vcpu)) {
+			++vcpu->arch.last_pt_write_count;
+			if (vcpu->arch.last_pt_write_count >= 3)
+				flooded = 1;
+		} else {
+			vcpu->arch.last_pt_write_gfn = gfn;
+			vcpu->arch.last_pt_write_count = 1;
+			vcpu->arch.last_pte_updated = NULL;
+		}
 	}
 	index = kvm_page_table_hashfn(gfn);
 	bucket = &vcpu->kvm->arch.mmu_page_hash[index];
@@ -2352,9 +2622,7 @@ EXPORT_SYMBOL_GPL(kvm_mmu_page_fault);
 
 void kvm_mmu_invlpg(struct kvm_vcpu *vcpu, gva_t gva)
 {
-	spin_lock(&vcpu->kvm->mmu_lock);
 	vcpu->arch.mmu.invlpg(vcpu, gva);
-	spin_unlock(&vcpu->kvm->mmu_lock);
 	kvm_mmu_flush_tlb(vcpu);
 	++vcpu->stat.invlpg;
 }
@@ -2451,7 +2719,7 @@ void kvm_mmu_slot_remove_write_access(struct kvm *kvm, int slot)
 		int i;
 		u64 *pt;
 
-		if (!test_bit(slot, &sp->slot_bitmap))
+		if (!test_bit(slot, sp->slot_bitmap))
 			continue;
 
 		pt = sp->spt;
@@ -2860,8 +3128,8 @@ static void audit_write_protection(struct kvm_vcpu *vcpu)
 		if (sp->role.metaphysical)
 			continue;
 
-		slot = gfn_to_memslot(vcpu->kvm, sp->gfn);
 		gfn = unalias_gfn(vcpu->kvm, sp->gfn);
+		slot = gfn_to_memslot_unaliased(vcpu->kvm, sp->gfn);
 		rmapp = &slot->rmap[gfn - slot->base_gfn];
 		if (*rmapp)
 			printk(KERN_ERR "%s: (%s) shadow page has writable"

arch/x86/kvm/paging_tmpl.h

@@ -82,6 +82,7 @@ struct shadow_walker {
 	int *ptwrite;
 	pfn_t pfn;
 	u64 *sptep;
+	gpa_t pte_gpa;
 };
 
 static gfn_t gpte_to_gfn(pt_element_t gpte)
@@ -222,7 +223,7 @@ walk:
 		if (ret)
 			goto walk;
 		pte |= PT_DIRTY_MASK;
-		kvm_mmu_pte_write(vcpu, pte_gpa, (u8 *)&pte, sizeof(pte));
+		kvm_mmu_pte_write(vcpu, pte_gpa, (u8 *)&pte, sizeof(pte), 0);
 		walker->ptes[walker->level - 1] = pte;
 	}
 
@@ -274,7 +275,8 @@ static void FNAME(update_pte)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *page,
 		return;
 	kvm_get_pfn(pfn);
 	mmu_set_spte(vcpu, spte, page->role.access, pte_access, 0, 0,
-		     gpte & PT_DIRTY_MASK, NULL, largepage, gpte_to_gfn(gpte),
+		     gpte & PT_DIRTY_MASK, NULL, largepage,
+		     gpte & PT_GLOBAL_MASK, gpte_to_gfn(gpte),
 		     pfn, true);
 }
 
@@ -301,8 +303,9 @@ static int FNAME(shadow_walk_entry)(struct kvm_shadow_walk *_sw,
 		mmu_set_spte(vcpu, sptep, access, gw->pte_access & access,
 			     sw->user_fault, sw->write_fault,
 			     gw->ptes[gw->level-1] & PT_DIRTY_MASK,
-			     sw->ptwrite, sw->largepage, gw->gfn, sw->pfn,
-			     false);
+			     sw->ptwrite, sw->largepage,
+			     gw->ptes[gw->level-1] & PT_GLOBAL_MASK,
+			     gw->gfn, sw->pfn, false);
 		sw->sptep = sptep;
 		return 1;
 	}
@@ -466,10 +469,22 @@ static int FNAME(shadow_invlpg_entry)(struct kvm_shadow_walk *_sw,
 				      struct kvm_vcpu *vcpu, u64 addr,
 				      u64 *sptep, int level)
 {
+	struct shadow_walker *sw =
+		container_of(_sw, struct shadow_walker, walker);
 
-	if (level == PT_PAGE_TABLE_LEVEL) {
-		if (is_shadow_present_pte(*sptep))
+	/* FIXME: properly handle invlpg on large guest pages */
+	if (level == PT_PAGE_TABLE_LEVEL ||
+	    ((level == PT_DIRECTORY_LEVEL) && is_large_pte(*sptep))) {
+		struct kvm_mmu_page *sp = page_header(__pa(sptep));
+
+		sw->pte_gpa = (sp->gfn << PAGE_SHIFT);
+		sw->pte_gpa += (sptep - sp->spt) * sizeof(pt_element_t);
+
+		if (is_shadow_present_pte(*sptep)) {
 			rmap_remove(vcpu->kvm, sptep);
+			if (is_large_pte(*sptep))
+				--vcpu->kvm->stat.lpages;
+		}
 		set_shadow_pte(sptep, shadow_trap_nonpresent_pte);
 		return 1;
 	}
@@ -480,11 +495,26 @@ static int FNAME(shadow_invlpg_entry)(struct kvm_shadow_walk *_sw,
 
 static void FNAME(invlpg)(struct kvm_vcpu *vcpu, gva_t gva)
 {
+	pt_element_t gpte;
 	struct shadow_walker walker = {
 		.walker = { .entry = FNAME(shadow_invlpg_entry), },
+		.pte_gpa = -1,
 	};
 
+	spin_lock(&vcpu->kvm->mmu_lock);
 	walk_shadow(&walker.walker, vcpu, gva);
+	spin_unlock(&vcpu->kvm->mmu_lock);
+	if (walker.pte_gpa == -1)
+		return;
+	if (kvm_read_guest_atomic(vcpu->kvm, walker.pte_gpa, &gpte,
+				  sizeof(pt_element_t)))
+		return;
+	if (is_present_pte(gpte) && (gpte & PT_ACCESSED_MASK)) {
+		if (mmu_topup_memory_caches(vcpu))
+			return;
+		kvm_mmu_pte_write(vcpu, walker.pte_gpa, (const u8 *)&gpte,
+				  sizeof(pt_element_t), 0);
+	}
 }
 
 static gpa_t FNAME(gva_to_gpa)(struct kvm_vcpu *vcpu, gva_t vaddr)
@@ -580,7 +610,7 @@ static int FNAME(sync_page)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
 		nr_present++;
 		pte_access = sp->role.access & FNAME(gpte_access)(vcpu, gpte);
 		set_spte(vcpu, &sp->spt[i], pte_access, 0, 0,
-			 is_dirty_pte(gpte), 0, gfn,
+			 is_dirty_pte(gpte), 0, gpte & PT_GLOBAL_MASK, gfn,
 			 spte_to_pfn(sp->spt[i]), true, false);
 	}
 

arch/x86/kvm/svm.c

@@ -28,6 +28,8 @@
 
 #include <asm/desc.h>
 
+#include <asm/virtext.h>
+
 #define __ex(x) __kvm_handle_fault_on_reboot(x)
 
 MODULE_AUTHOR("Qumranet");
@@ -245,34 +247,19 @@ static void skip_emulated_instruction(struct kvm_vcpu *vcpu)
 
 static int has_svm(void)
 {
-	uint32_t eax, ebx, ecx, edx;
-
-	if (boot_cpu_data.x86_vendor != X86_VENDOR_AMD) {
-		printk(KERN_INFO "has_svm: not amd\n");
-		return 0;
-	}
+	const char *msg;
 
-	cpuid(0x80000000, &eax, &ebx, &ecx, &edx);
-	if (eax < SVM_CPUID_FUNC) {
-		printk(KERN_INFO "has_svm: can't execute cpuid_8000000a\n");
+	if (!cpu_has_svm(&msg)) {
+		printk(KERN_INFO "has_svn: %s\n", msg);
 		return 0;
 	}
 
-	cpuid(0x80000001, &eax, &ebx, &ecx, &edx);
-	if (!(ecx & (1 << SVM_CPUID_FEATURE_SHIFT))) {
-		printk(KERN_DEBUG "has_svm: svm not available\n");
-		return 0;
-	}
 	return 1;
 }
 
 static void svm_hardware_disable(void *garbage)
 {
-	uint64_t efer;
-
-	wrmsrl(MSR_VM_HSAVE_PA, 0);
-	rdmsrl(MSR_EFER, efer);
-	wrmsrl(MSR_EFER, efer & ~MSR_EFER_SVME_MASK);
+	cpu_svm_disable();
 }
 
 static void svm_hardware_enable(void *garbage)
@@ -772,6 +759,22 @@ static void svm_get_segment(struct kvm_vcpu *vcpu,
 	var->l = (s->attrib >> SVM_SELECTOR_L_SHIFT) & 1;
 	var->db = (s->attrib >> SVM_SELECTOR_DB_SHIFT) & 1;
 	var->g = (s->attrib >> SVM_SELECTOR_G_SHIFT) & 1;
+
+	/*
+	 * SVM always stores 0 for the 'G' bit in the CS selector in
+	 * the VMCB on a VMEXIT. This hurts cross-vendor migration:
+	 * Intel's VMENTRY has a check on the 'G' bit.
+	 */
+	if (seg == VCPU_SREG_CS)
+		var->g = s->limit > 0xfffff;
+
+	/*
+	 * Work around a bug where the busy flag in the tr selector
+	 * isn't exposed
+	 */
+	if (seg == VCPU_SREG_TR)
+		var->type |= 0x2;
+
 	var->unusable = !var->present;
 }
 
@@ -1099,6 +1102,7 @@ static int io_interception(struct vcpu_svm *svm, struct kvm_run *kvm_run)
 	rep = (io_info & SVM_IOIO_REP_MASK) != 0;
 	down = (svm->vmcb->save.rflags & X86_EFLAGS_DF) != 0;
 
+	skip_emulated_instruction(&svm->vcpu);
 	return kvm_emulate_pio(&svm->vcpu, kvm_run, in, size, port);
 }
 
@@ -1912,6 +1916,11 @@ static int get_npt_level(void)
 #endif
 }
 
+static int svm_get_mt_mask_shift(void)
+{
+	return 0;
+}
+
 static struct kvm_x86_ops svm_x86_ops = {
 	.cpu_has_kvm_support = has_svm,
 	.disabled_by_bios = is_disabled,
@@ -1967,6 +1976,7 @@ static struct kvm_x86_ops svm_x86_ops = {
 
 	.set_tss_addr = svm_set_tss_addr,
 	.get_tdp_level = get_npt_level,
+	.get_mt_mask_shift = svm_get_mt_mask_shift,
 };
 
 static int __init svm_init(void)

arch/x86/kvm/vmx.c

@@ -16,7 +16,6 @@
  */
 
 #include "irq.h"
-#include "vmx.h"
 #include "mmu.h"
 
 #include <linux/kvm_host.h>
@@ -31,6 +30,8 @@
 
 #include <asm/io.h>
 #include <asm/desc.h>
+#include <asm/vmx.h>
+#include <asm/virtext.h>
 
 #define __ex(x) __kvm_handle_fault_on_reboot(x)
 
@@ -90,6 +91,11 @@ struct vcpu_vmx {
 	} rmode;
 	int vpid;
 	bool emulation_required;
+
+	/* Support for vnmi-less CPUs */
+	int soft_vnmi_blocked;
+	ktime_t entry_time;
+	s64 vnmi_blocked_time;
 };
 
 static inline struct vcpu_vmx *to_vmx(struct kvm_vcpu *vcpu)
@@ -122,7 +128,7 @@ static struct vmcs_config {
 	u32 vmentry_ctrl;
 } vmcs_config;
 
-struct vmx_capability {
+static struct vmx_capability {
 	u32 ept;
 	u32 vpid;
 } vmx_capability;
@@ -957,6 +963,13 @@ static int vmx_set_msr(struct kvm_vcpu *vcpu, u32 msr_index, u64 data)
 		pr_unimpl(vcpu, "unimplemented perfctr wrmsr: 0x%x data 0x%llx\n", msr_index, data);
 
 		break;
+	case MSR_IA32_CR_PAT:
+		if (vmcs_config.vmentry_ctrl & VM_ENTRY_LOAD_IA32_PAT) {
+			vmcs_write64(GUEST_IA32_PAT, data);
+			vcpu->arch.pat = data;
+			break;
+		}
+		/* Otherwise falls through to kvm_set_msr_common */
 	default:
 		vmx_load_host_state(vmx);
 		msr = find_msr_entry(vmx, msr_index);
@@ -1032,8 +1045,7 @@ static int vmx_get_irq(struct kvm_vcpu *vcpu)
 
 static __init int cpu_has_kvm_support(void)
 {
-	unsigned long ecx = cpuid_ecx(1);
-	return test_bit(5, &ecx); /* CPUID.1:ECX.VMX[bit 5] -> VT */
+	return cpu_has_vmx();
 }
 
 static __init int vmx_disabled_by_bios(void)
@@ -1079,13 +1091,22 @@ static void vmclear_local_vcpus(void)
 		__vcpu_clear(vmx);
 }
 
-static void hardware_disable(void *garbage)
+
+/* Just like cpu_vmxoff(), but with the __kvm_handle_fault_on_reboot()
+ * tricks.
+ */
+static void kvm_cpu_vmxoff(void)
 {
-	vmclear_local_vcpus();
 	asm volatile (__ex(ASM_VMX_VMXOFF) : : : "cc");
 	write_cr4(read_cr4() & ~X86_CR4_VMXE);
 }
 
+static void hardware_disable(void *garbage)
+{
+	vmclear_local_vcpus();
+	kvm_cpu_vmxoff();
+}
+
 static __init int adjust_vmx_controls(u32 ctl_min, u32 ctl_opt,
 				      u32 msr, u32 *result)
 {
@@ -1176,12 +1197,13 @@ static __init int setup_vmcs_config(struct vmcs_config *vmcs_conf)
 #ifdef CONFIG_X86_64
 	min |= VM_EXIT_HOST_ADDR_SPACE_SIZE;
 #endif
-	opt = 0;
+	opt = VM_EXIT_SAVE_IA32_PAT | VM_EXIT_LOAD_IA32_PAT;
 	if (adjust_vmx_controls(min, opt, MSR_IA32_VMX_EXIT_CTLS,
 				&_vmexit_control) < 0)
 		return -EIO;
 
-	min = opt = 0;
+	min = 0;
+	opt = VM_ENTRY_LOAD_IA32_PAT;
 	if (adjust_vmx_controls(min, opt, MSR_IA32_VMX_ENTRY_CTLS,
 				&_vmentry_control) < 0)
 		return -EIO;
@@ -2087,8 +2109,9 @@ static void vmx_disable_intercept_for_msr(struct page *msr_bitmap, u32 msr)
  */
 static int vmx_vcpu_setup(struct vcpu_vmx *vmx)
 {
-	u32 host_sysenter_cs;
+	u32 host_sysenter_cs, msr_low, msr_high;
 	u32 junk;
+	u64 host_pat;
 	unsigned long a;
 	struct descriptor_table dt;
 	int i;
@@ -2176,6 +2199,20 @@ static int vmx_vcpu_setup(struct vcpu_vmx *vmx)
 	rdmsrl(MSR_IA32_SYSENTER_EIP, a);
 	vmcs_writel(HOST_IA32_SYSENTER_EIP, a);   /* 22.2.3 */
 
+	if (vmcs_config.vmexit_ctrl & VM_EXIT_LOAD_IA32_PAT) {
+		rdmsr(MSR_IA32_CR_PAT, msr_low, msr_high);
+		host_pat = msr_low | ((u64) msr_high << 32);
+		vmcs_write64(HOST_IA32_PAT, host_pat);
+	}
+	if (vmcs_config.vmentry_ctrl & VM_ENTRY_LOAD_IA32_PAT) {
+		rdmsr(MSR_IA32_CR_PAT, msr_low, msr_high);
+		host_pat = msr_low | ((u64) msr_high << 32);
+		/* Write the default value follow host pat */
+		vmcs_write64(GUEST_IA32_PAT, host_pat);
+		/* Keep arch.pat sync with GUEST_IA32_PAT */
+		vmx->vcpu.arch.pat = host_pat;
+	}
+
 	for (i = 0; i < NR_VMX_MSR; ++i) {
 		u32 index = vmx_msr_index[i];
 		u32 data_low, data_high;
@@ -2230,6 +2267,8 @@ static int vmx_vcpu_reset(struct kvm_vcpu *vcpu)
 
 	vmx->vcpu.arch.rmode.active = 0;
 
+	vmx->soft_vnmi_blocked = 0;
+
 	vmx->vcpu.arch.regs[VCPU_REGS_RDX] = get_rdx_init_val();
 	kvm_set_cr8(&vmx->vcpu, 0);
 	msr = 0xfee00000 | MSR_IA32_APICBASE_ENABLE;
@@ -2335,6 +2374,29 @@ out:
 	return ret;
 }
 
+static void enable_irq_window(struct kvm_vcpu *vcpu)
+{
+	u32 cpu_based_vm_exec_control;
+
+	cpu_based_vm_exec_control = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL);
+	cpu_based_vm_exec_control |= CPU_BASED_VIRTUAL_INTR_PENDING;
+	vmcs_write32(CPU_BASED_VM_EXEC_CONTROL, cpu_based_vm_exec_control);
+}
+
+static void enable_nmi_window(struct kvm_vcpu *vcpu)
+{
+	u32 cpu_based_vm_exec_control;
+
+	if (!cpu_has_virtual_nmis()) {
+		enable_irq_window(vcpu);
+		return;
+	}
+
+	cpu_based_vm_exec_control = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL);
+	cpu_based_vm_exec_control |= CPU_BASED_VIRTUAL_NMI_PENDING;
+	vmcs_write32(CPU_BASED_VM_EXEC_CONTROL, cpu_based_vm_exec_control);
+}
+
 static void vmx_inject_irq(struct kvm_vcpu *vcpu, int irq)
 {
 	struct vcpu_vmx *vmx = to_vmx(vcpu);
@@ -2358,10 +2420,54 @@ static void vmx_inject_irq(struct kvm_vcpu *vcpu, int irq)
 
 static void vmx_inject_nmi(struct kvm_vcpu *vcpu)
 {
+	struct vcpu_vmx *vmx = to_vmx(vcpu);
+
+	if (!cpu_has_virtual_nmis()) {
+		/*
+		 * Tracking the NMI-blocked state in software is built upon
+		 * finding the next open IRQ window. This, in turn, depends on
+		 * well-behaving guests: They have to keep IRQs disabled at
+		 * least as long as the NMI handler runs. Otherwise we may
+		 * cause NMI nesting, maybe breaking the guest. But as this is
+		 * highly unlikely, we can live with the residual risk.
+		 */
+		vmx->soft_vnmi_blocked = 1;
+		vmx->vnmi_blocked_time = 0;
+	}
+
+	++vcpu->stat.nmi_injections;
+	if (vcpu->arch.rmode.active) {
+		vmx->rmode.irq.pending = true;
+		vmx->rmode.irq.vector = NMI_VECTOR;
+		vmx->rmode.irq.rip = kvm_rip_read(vcpu);
+		vmcs_write32(VM_ENTRY_INTR_INFO_FIELD,
+			     NMI_VECTOR | INTR_TYPE_SOFT_INTR |
+			     INTR_INFO_VALID_MASK);
+		vmcs_write32(VM_ENTRY_INSTRUCTION_LEN, 1);
+		kvm_rip_write(vcpu, vmx->rmode.irq.rip - 1);
+		return;
+	}
 	vmcs_write32(VM_ENTRY_INTR_INFO_FIELD,
 			INTR_TYPE_NMI_INTR | INTR_INFO_VALID_MASK | NMI_VECTOR);
 }
 
+static void vmx_update_window_states(struct kvm_vcpu *vcpu)
+{
+	u32 guest_intr = vmcs_read32(GUEST_INTERRUPTIBILITY_INFO);
+
+	vcpu->arch.nmi_window_open =
+		!(guest_intr & (GUEST_INTR_STATE_STI |
+				GUEST_INTR_STATE_MOV_SS |
+				GUEST_INTR_STATE_NMI));
+	if (!cpu_has_virtual_nmis() && to_vmx(vcpu)->soft_vnmi_blocked)
+		vcpu->arch.nmi_window_open = 0;
+
+	vcpu->arch.interrupt_window_open =
+		((vmcs_readl(GUEST_RFLAGS) & X86_EFLAGS_IF) &&
+		 !(guest_intr & (GUEST_INTR_STATE_STI |
+				 GUEST_INTR_STATE_MOV_SS)));
+}
+
 static void kvm_do_inject_irq(struct kvm_vcpu *vcpu)
 {
 	int word_index = __ffs(vcpu->arch.irq_summary);
@@ -2374,40 +2480,49 @@ static void kvm_do_inject_irq(struct kvm_vcpu *vcpu)
 	kvm_queue_interrupt(vcpu, irq);
 }
 
-
 static void do_interrupt_requests(struct kvm_vcpu *vcpu,
 				       struct kvm_run *kvm_run)
 {
-	u32 cpu_based_vm_exec_control;
-
-	vcpu->arch.interrupt_window_open =
-		((vmcs_readl(GUEST_RFLAGS) & X86_EFLAGS_IF) &&
-		 (vmcs_read32(GUEST_INTERRUPTIBILITY_INFO) & 3) == 0);
+	vmx_update_window_states(vcpu);
 
-	if (vcpu->arch.interrupt_window_open &&
-	    vcpu->arch.irq_summary && !vcpu->arch.interrupt.pending)
-		kvm_do_inject_irq(vcpu);
+	if (vcpu->arch.nmi_pending && !vcpu->arch.nmi_injected) {
+		if (vcpu->arch.interrupt.pending) {
+			enable_nmi_window(vcpu);
+		} else if (vcpu->arch.nmi_window_open) {
+			vcpu->arch.nmi_pending = false;
+			vcpu->arch.nmi_injected = true;
+		} else {
+			enable_nmi_window(vcpu);
+			return;
+		}
+	}
+	if (vcpu->arch.nmi_injected) {
+		vmx_inject_nmi(vcpu);
+		if (vcpu->arch.nmi_pending)
+			enable_nmi_window(vcpu);
+		else if (vcpu->arch.irq_summary
+			 || kvm_run->request_interrupt_window)
+			enable_irq_window(vcpu);
+		return;
+	}
 
-	if (vcpu->arch.interrupt_window_open && vcpu->arch.interrupt.pending)
-		vmx_inject_irq(vcpu, vcpu->arch.interrupt.nr);
+	if (vcpu->arch.interrupt_window_open) {
+		if (vcpu->arch.irq_summary && !vcpu->arch.interrupt.pending)
+			kvm_do_inject_irq(vcpu);
 
-	cpu_based_vm_exec_control = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL);
+		if (vcpu->arch.interrupt.pending)
+			vmx_inject_irq(vcpu, vcpu->arch.interrupt.nr);
+	}
 	if (!vcpu->arch.interrupt_window_open &&
 	    (vcpu->arch.irq_summary || 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);
+		enable_irq_window(vcpu);
 }
 
 static int vmx_set_tss_addr(struct kvm *kvm, unsigned int addr)
 {
 	int ret;
 	struct kvm_userspace_memory_region tss_mem = {
-		.slot = 8,
+		.slot = TSS_PRIVATE_MEMSLOT,
 		.guest_phys_addr = addr,
 		.memory_size = PAGE_SIZE * 3,
 		.flags = 0,
@@ -2492,7 +2607,7 @@ static int handle_exception(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 		set_bit(irq / BITS_PER_LONG, &vcpu->arch.irq_summary);
 	}
 
-	if ((intr_info & INTR_INFO_INTR_TYPE_MASK) == 0x200) /* nmi */
+	if ((intr_info & INTR_INFO_INTR_TYPE_MASK) == INTR_TYPE_NMI_INTR)
 		return 1;  /* already handled by vmx_vcpu_run() */
 
 	if (is_no_device(intr_info)) {
@@ -2581,6 +2696,7 @@ static int handle_io(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	rep = (exit_qualification & 32) != 0;
 	port = exit_qualification >> 16;
 
+	skip_emulated_instruction(vcpu);
 	return kvm_emulate_pio(vcpu, kvm_run, in, size, port);
 }
 
@@ -2767,6 +2883,7 @@ static int handle_interrupt_window(struct kvm_vcpu *vcpu,
 	vmcs_write32(CPU_BASED_VM_EXEC_CONTROL, cpu_based_vm_exec_control);
 
 	KVMTRACE_0D(PEND_INTR, vcpu, handler);
+	++vcpu->stat.irq_window_exits;
 
 	/*
 	 * If the user space waits to inject interrupts, exit as soon as
@@ -2775,7 +2892,6 @@ static int handle_interrupt_window(struct kvm_vcpu *vcpu,
 	if (kvm_run->request_interrupt_window &&
 	    !vcpu->arch.irq_summary) {
 		kvm_run->exit_reason = KVM_EXIT_IRQ_WINDOW_OPEN;
-		++vcpu->stat.irq_window_exits;
 		return 0;
 	}
 	return 1;
@@ -2832,6 +2948,7 @@ static int handle_apic_access(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 
 static int handle_task_switch(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 {
+	struct vcpu_vmx *vmx = to_vmx(vcpu);
 	unsigned long exit_qualification;
 	u16 tss_selector;
 	int reason;
@@ -2839,6 +2956,15 @@ static int handle_task_switch(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	exit_qualification = vmcs_readl(EXIT_QUALIFICATION);
 
 	reason = (u32)exit_qualification >> 30;
+	if (reason == TASK_SWITCH_GATE && vmx->vcpu.arch.nmi_injected &&
+	    (vmx->idt_vectoring_info & VECTORING_INFO_VALID_MASK) &&
+	    (vmx->idt_vectoring_info & VECTORING_INFO_TYPE_MASK)
+	    == INTR_TYPE_NMI_INTR) {
+		vcpu->arch.nmi_injected = false;
+		if (cpu_has_virtual_nmis())
+			vmcs_set_bits(GUEST_INTERRUPTIBILITY_INFO,
+				      GUEST_INTR_STATE_NMI);
+	}
 	tss_selector = exit_qualification;
 
 	return kvm_task_switch(vcpu, tss_selector, reason);
@@ -2927,16 +3053,12 @@ static void handle_invalid_guest_state(struct kvm_vcpu *vcpu,
 	while (!guest_state_valid(vcpu)) {
 		err = emulate_instruction(vcpu, kvm_run, 0, 0, 0);
 
-		switch (err) {
-			case EMULATE_DONE:
-				break;
-			case EMULATE_DO_MMIO:
-				kvm_report_emulation_failure(vcpu, "mmio");
-				/* TODO: Handle MMIO */
-				return;
-			default:
-				kvm_report_emulation_failure(vcpu, "emulation failure");
-				return;
+		if (err == EMULATE_DO_MMIO)
+			break;
+
+		if (err != EMULATE_DONE) {
+			kvm_report_emulation_failure(vcpu, "emulation failure");
+			return;
 		}
 
 		if (signal_pending(current))
@@ -2948,8 +3070,10 @@ static void handle_invalid_guest_state(struct kvm_vcpu *vcpu,
 	local_irq_disable();
 	preempt_disable();
 
-	/* Guest state should be valid now, no more emulation should be needed */
-	vmx->emulation_required = 0;
+	/* Guest state should be valid now except if we need to
+	 * emulate an MMIO */
+	if (guest_state_valid(vcpu))
+		vmx->emulation_required = 0;
 }
 
 /*
@@ -2996,6 +3120,11 @@ static int kvm_handle_exit(struct kvm_run *kvm_run, struct kvm_vcpu *vcpu)
 	KVMTRACE_3D(VMEXIT, vcpu, exit_reason, (u32)kvm_rip_read(vcpu),
 		    (u32)((u64)kvm_rip_read(vcpu) >> 32), entryexit);
 
+	/* If we need to emulate an MMIO from handle_invalid_guest_state
+	 * we just return 0 */
+	if (vmx->emulation_required && emulate_invalid_guest_state)
+		return 0;
+
 	/* Access CR3 don't cause VMExit in paging mode, so we need
 	 * to sync with guest real CR3. */
 	if (vm_need_ept() && is_paging(vcpu)) {
@@ -3012,9 +3141,32 @@ static int kvm_handle_exit(struct kvm_run *kvm_run, struct kvm_vcpu *vcpu)
 
 	if ((vectoring_info & VECTORING_INFO_VALID_MASK) &&
 			(exit_reason != EXIT_REASON_EXCEPTION_NMI &&
-			exit_reason != EXIT_REASON_EPT_VIOLATION))
-		printk(KERN_WARNING "%s: unexpected, valid vectoring info and "
-		       "exit reason is 0x%x\n", __func__, exit_reason);
+			exit_reason != EXIT_REASON_EPT_VIOLATION &&
+			exit_reason != EXIT_REASON_TASK_SWITCH))
+		printk(KERN_WARNING "%s: unexpected, valid vectoring info "
+		       "(0x%x) and exit reason is 0x%x\n",
+		       __func__, vectoring_info, exit_reason);
+
+	if (unlikely(!cpu_has_virtual_nmis() && vmx->soft_vnmi_blocked)) {
+		if (vcpu->arch.interrupt_window_open) {
+			vmx->soft_vnmi_blocked = 0;
+			vcpu->arch.nmi_window_open = 1;
+		} else if (vmx->vnmi_blocked_time > 1000000000LL &&
+			   vcpu->arch.nmi_pending) {
+			/*
+			 * This CPU don't support us in finding the end of an
+			 * NMI-blocked window if the guest runs with IRQs
+			 * disabled. So we pull the trigger after 1 s of
+			 * futile waiting, but inform the user about this.
+			 */
+			printk(KERN_WARNING "%s: Breaking out of NMI-blocked "
+			       "state on VCPU %d after 1 s timeout\n",
+			       __func__, vcpu->vcpu_id);
+			vmx->soft_vnmi_blocked = 0;
+			vmx->vcpu.arch.nmi_window_open = 1;
+		}
+	}
+
 	if (exit_reason < kvm_vmx_max_exit_handlers
 	    && kvm_vmx_exit_handlers[exit_reason])
 		return kvm_vmx_exit_handlers[exit_reason](vcpu, kvm_run);
@@ -3042,51 +3194,6 @@ static void update_tpr_threshold(struct kvm_vcpu *vcpu)
 	vmcs_write32(TPR_THRESHOLD, (max_irr > tpr) ? tpr >> 4 : max_irr >> 4);
 }
 
-static void enable_irq_window(struct kvm_vcpu *vcpu)
-{
-	u32 cpu_based_vm_exec_control;
-
-	cpu_based_vm_exec_control = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL);
-	cpu_based_vm_exec_control |= CPU_BASED_VIRTUAL_INTR_PENDING;
-	vmcs_write32(CPU_BASED_VM_EXEC_CONTROL, cpu_based_vm_exec_control);
-}
-
-static void enable_nmi_window(struct kvm_vcpu *vcpu)
-{
-	u32 cpu_based_vm_exec_control;
-
-	if (!cpu_has_virtual_nmis())
-		return;
-
-	cpu_based_vm_exec_control = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL);
-	cpu_based_vm_exec_control |= CPU_BASED_VIRTUAL_NMI_PENDING;
-	vmcs_write32(CPU_BASED_VM_EXEC_CONTROL, cpu_based_vm_exec_control);
-}
-
-static int vmx_nmi_enabled(struct kvm_vcpu *vcpu)
-{
-	u32 guest_intr = vmcs_read32(GUEST_INTERRUPTIBILITY_INFO);
-	return !(guest_intr & (GUEST_INTR_STATE_NMI |
-			       GUEST_INTR_STATE_MOV_SS |
-			       GUEST_INTR_STATE_STI));
-}
-
-static int vmx_irq_enabled(struct kvm_vcpu *vcpu)
-{
-	u32 guest_intr = vmcs_read32(GUEST_INTERRUPTIBILITY_INFO);
-	return (!(guest_intr & (GUEST_INTR_STATE_MOV_SS |
-			       GUEST_INTR_STATE_STI)) &&
-		(vmcs_readl(GUEST_RFLAGS) & X86_EFLAGS_IF));
-}
-
-static void enable_intr_window(struct kvm_vcpu *vcpu)
-{
-	if (vcpu->arch.nmi_pending)
-		enable_nmi_window(vcpu);
-	else if (kvm_cpu_has_interrupt(vcpu))
-		enable_irq_window(vcpu);
-}
-
 static void vmx_complete_interrupts(struct vcpu_vmx *vmx)
 {
 	u32 exit_intr_info;
@@ -3109,7 +3216,9 @@ static void vmx_complete_interrupts(struct vcpu_vmx *vmx)
 		if (unblock_nmi && vector != DF_VECTOR)
 			vmcs_set_bits(GUEST_INTERRUPTIBILITY_INFO,
 				      GUEST_INTR_STATE_NMI);
-	}
+	} else if (unlikely(vmx->soft_vnmi_blocked))
+		vmx->vnmi_blocked_time +=
+			ktime_to_ns(ktime_sub(ktime_get(), vmx->entry_time));
 
 	idt_vectoring_info = vmx->idt_vectoring_info;
 	idtv_info_valid = idt_vectoring_info & VECTORING_INFO_VALID_MASK;
@@ -3147,26 +3256,29 @@ static void vmx_intr_assist(struct kvm_vcpu *vcpu)
 {
 	update_tpr_threshold(vcpu);
 
-	if (cpu_has_virtual_nmis()) {
-		if (vcpu->arch.nmi_pending && !vcpu->arch.nmi_injected) {
-			if (vcpu->arch.interrupt.pending) {
-				enable_nmi_window(vcpu);
-			} else if (vmx_nmi_enabled(vcpu)) {
-				vcpu->arch.nmi_pending = false;
-				vcpu->arch.nmi_injected = true;
-			} else {
-				enable_intr_window(vcpu);
-				return;
-			}
-		}
-		if (vcpu->arch.nmi_injected) {
-			vmx_inject_nmi(vcpu);
-			enable_intr_window(vcpu);
+	vmx_update_window_states(vcpu);
+
+	if (vcpu->arch.nmi_pending && !vcpu->arch.nmi_injected) {
+		if (vcpu->arch.interrupt.pending) {
+			enable_nmi_window(vcpu);
+		} else if (vcpu->arch.nmi_window_open) {
+			vcpu->arch.nmi_pending = false;
+			vcpu->arch.nmi_injected = true;
+		} else {
+			enable_nmi_window(vcpu);
 			return;
 		}
 	}
+	if (vcpu->arch.nmi_injected) {
+		vmx_inject_nmi(vcpu);
+		if (vcpu->arch.nmi_pending)
+			enable_nmi_window(vcpu);
+		else if (kvm_cpu_has_interrupt(vcpu))
+			enable_irq_window(vcpu);
+		return;
+	}
 	if (!vcpu->arch.interrupt.pending && kvm_cpu_has_interrupt(vcpu)) {
-		if (vmx_irq_enabled(vcpu))
+		if (vcpu->arch.interrupt_window_open)
 			kvm_queue_interrupt(vcpu, kvm_cpu_get_interrupt(vcpu));
 		else
 			enable_irq_window(vcpu);
@@ -3174,6 +3286,8 @@ static void vmx_intr_assist(struct kvm_vcpu *vcpu)
 	if (vcpu->arch.interrupt.pending) {
 		vmx_inject_irq(vcpu, vcpu->arch.interrupt.nr);
 		kvm_timer_intr_post(vcpu, vcpu->arch.interrupt.nr);
+		if (kvm_cpu_has_interrupt(vcpu))
+			enable_irq_window(vcpu);
 	}
 }
 
@@ -3213,6 +3327,10 @@ static void vmx_vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	struct vcpu_vmx *vmx = to_vmx(vcpu);
 	u32 intr_info;
 
+	/* Record the guest's net vcpu time for enforced NMI injections. */
+	if (unlikely(!cpu_has_virtual_nmis() && vmx->soft_vnmi_blocked))
+		vmx->entry_time = ktime_get();
+
 	/* Handle invalid guest state instead of entering VMX */
 	if (vmx->emulation_required && emulate_invalid_guest_state) {
 		handle_invalid_guest_state(vcpu, kvm_run);
@@ -3327,9 +3445,7 @@ static void vmx_vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	if (vmx->rmode.irq.pending)
 		fixup_rmode_irq(vmx);
 
-	vcpu->arch.interrupt_window_open =
-		(vmcs_read32(GUEST_INTERRUPTIBILITY_INFO) &
-		 (GUEST_INTR_STATE_STI | GUEST_INTR_STATE_MOV_SS)) == 0;
+	vmx_update_window_states(vcpu);
 
 	asm("mov %0, %%ds; mov %0, %%es" : : "r"(__USER_DS));
 	vmx->launched = 1;
@@ -3337,7 +3453,7 @@ static void vmx_vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	intr_info = vmcs_read32(VM_EXIT_INTR_INFO);
 
 	/* We need to handle NMIs before interrupts are enabled */
-	if ((intr_info & INTR_INFO_INTR_TYPE_MASK) == 0x200 &&
+	if ((intr_info & INTR_INFO_INTR_TYPE_MASK) == INTR_TYPE_NMI_INTR &&
 	    (intr_info & INTR_INFO_VALID_MASK)) {
 		KVMTRACE_0D(NMI, vcpu, handler);
 		asm("int $2");
@@ -3455,6 +3571,11 @@ static int get_ept_level(void)
 	return VMX_EPT_DEFAULT_GAW + 1;
 }
 
+static int vmx_get_mt_mask_shift(void)
+{
+	return VMX_EPT_MT_EPTE_SHIFT;
+}
+
 static struct kvm_x86_ops vmx_x86_ops = {
 	.cpu_has_kvm_support = cpu_has_kvm_support,
 	.disabled_by_bios = vmx_disabled_by_bios,
@@ -3510,6 +3631,7 @@ static struct kvm_x86_ops vmx_x86_ops = {
 
 	.set_tss_addr = vmx_set_tss_addr,
 	.get_tdp_level = get_ept_level,
+	.get_mt_mask_shift = vmx_get_mt_mask_shift,
 };
 
 static int __init vmx_init(void)
@@ -3566,10 +3688,10 @@ static int __init vmx_init(void)
 		bypass_guest_pf = 0;
 		kvm_mmu_set_base_ptes(VMX_EPT_READABLE_MASK |
 			VMX_EPT_WRITABLE_MASK |
-			VMX_EPT_DEFAULT_MT << VMX_EPT_MT_EPTE_SHIFT |
 			VMX_EPT_IGMT_BIT);
 		kvm_mmu_set_mask_ptes(0ull, 0ull, 0ull, 0ull,
-				VMX_EPT_EXECUTABLE_MASK);
+				VMX_EPT_EXECUTABLE_MASK,
+				VMX_EPT_DEFAULT_MT << VMX_EPT_MT_EPTE_SHIFT);
 		kvm_enable_tdp();
 	} else
 		kvm_disable_tdp();

arch/x86/kvm/x86.c

@@ -39,6 +39,7 @@
 #include <asm/uaccess.h>
 #include <asm/msr.h>
 #include <asm/desc.h>
+#include <asm/mtrr.h>
 
 #define MAX_IO_MSRS 256
 #define CR0_RESERVED_BITS						\
@@ -86,6 +87,7 @@ struct kvm_stats_debugfs_item debugfs_entries[] = {
 	{ "halt_wakeup", VCPU_STAT(halt_wakeup) },
 	{ "hypercalls", VCPU_STAT(hypercalls) },
 	{ "request_irq", VCPU_STAT(request_irq_exits) },
+	{ "request_nmi", VCPU_STAT(request_nmi_exits) },
 	{ "irq_exits", VCPU_STAT(irq_exits) },
 	{ "host_state_reload", VCPU_STAT(host_state_reload) },
 	{ "efer_reload", VCPU_STAT(efer_reload) },
@@ -93,6 +95,7 @@ struct kvm_stats_debugfs_item debugfs_entries[] = {
 	{ "insn_emulation", VCPU_STAT(insn_emulation) },
 	{ "insn_emulation_fail", VCPU_STAT(insn_emulation_fail) },
 	{ "irq_injections", VCPU_STAT(irq_injections) },
+	{ "nmi_injections", VCPU_STAT(nmi_injections) },
 	{ "mmu_shadow_zapped", VM_STAT(mmu_shadow_zapped) },
 	{ "mmu_pte_write", VM_STAT(mmu_pte_write) },
 	{ "mmu_pte_updated", VM_STAT(mmu_pte_updated) },
@@ -101,6 +104,7 @@ struct kvm_stats_debugfs_item debugfs_entries[] = {
 	{ "mmu_recycled", VM_STAT(mmu_recycled) },
 	{ "mmu_cache_miss", VM_STAT(mmu_cache_miss) },
 	{ "mmu_unsync", VM_STAT(mmu_unsync) },
+	{ "mmu_unsync_global", VM_STAT(mmu_unsync_global) },
 	{ "remote_tlb_flush", VM_STAT(remote_tlb_flush) },
 	{ "largepages", VM_STAT(lpages) },
 	{ NULL }
@@ -312,6 +316,7 @@ void kvm_set_cr0(struct kvm_vcpu *vcpu, unsigned long cr0)
 	kvm_x86_ops->set_cr0(vcpu, cr0);
 	vcpu->arch.cr0 = cr0;
 
+	kvm_mmu_sync_global(vcpu);
 	kvm_mmu_reset_context(vcpu);
 	return;
 }
@@ -355,6 +360,7 @@ void kvm_set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4)
 	}
 	kvm_x86_ops->set_cr4(vcpu, cr4);
 	vcpu->arch.cr4 = cr4;
+	kvm_mmu_sync_global(vcpu);
 	kvm_mmu_reset_context(vcpu);
 }
 EXPORT_SYMBOL_GPL(kvm_set_cr4);
@@ -449,7 +455,7 @@ static u32 msrs_to_save[] = {
 	MSR_CSTAR, MSR_KERNEL_GS_BASE, MSR_SYSCALL_MASK, MSR_LSTAR,
 #endif
 	MSR_IA32_TIME_STAMP_COUNTER, MSR_KVM_SYSTEM_TIME, MSR_KVM_WALL_CLOCK,
-	MSR_IA32_PERF_STATUS,
+	MSR_IA32_PERF_STATUS, MSR_IA32_CR_PAT
 };
 
 static unsigned num_msrs_to_save;
@@ -648,10 +654,38 @@ static bool msr_mtrr_valid(unsigned msr)
 
 static int set_msr_mtrr(struct kvm_vcpu *vcpu, u32 msr, u64 data)
 {
+	u64 *p = (u64 *)&vcpu->arch.mtrr_state.fixed_ranges;
+
 	if (!msr_mtrr_valid(msr))
 		return 1;
 
-	vcpu->arch.mtrr[msr - 0x200] = data;
+	if (msr == MSR_MTRRdefType) {
+		vcpu->arch.mtrr_state.def_type = data;
+		vcpu->arch.mtrr_state.enabled = (data & 0xc00) >> 10;
+	} else if (msr == MSR_MTRRfix64K_00000)
+		p[0] = data;
+	else if (msr == MSR_MTRRfix16K_80000 || msr == MSR_MTRRfix16K_A0000)
+		p[1 + msr - MSR_MTRRfix16K_80000] = data;
+	else if (msr >= MSR_MTRRfix4K_C0000 && msr <= MSR_MTRRfix4K_F8000)
+		p[3 + msr - MSR_MTRRfix4K_C0000] = data;
+	else if (msr == MSR_IA32_CR_PAT)
+		vcpu->arch.pat = data;
+	else {	/* Variable MTRRs */
+		int idx, is_mtrr_mask;
+		u64 *pt;
+
+		idx = (msr - 0x200) / 2;
+		is_mtrr_mask = msr - 0x200 - 2 * idx;
+		if (!is_mtrr_mask)
+			pt =
+			  (u64 *)&vcpu->arch.mtrr_state.var_ranges[idx].base_lo;
+		else
+			pt =
+			  (u64 *)&vcpu->arch.mtrr_state.var_ranges[idx].mask_lo;
+		*pt = data;
+	}
+
+	kvm_mmu_reset_context(vcpu);
 	return 0;
 }
 
@@ -747,10 +781,37 @@ int kvm_get_msr(struct kvm_vcpu *vcpu, u32 msr_index, u64 *pdata)
 
 static int get_msr_mtrr(struct kvm_vcpu *vcpu, u32 msr, u64 *pdata)
 {
+	u64 *p = (u64 *)&vcpu->arch.mtrr_state.fixed_ranges;
+
 	if (!msr_mtrr_valid(msr))
 		return 1;
 
-	*pdata = vcpu->arch.mtrr[msr - 0x200];
+	if (msr == MSR_MTRRdefType)
+		*pdata = vcpu->arch.mtrr_state.def_type +
+			 (vcpu->arch.mtrr_state.enabled << 10);
+	else if (msr == MSR_MTRRfix64K_00000)
+		*pdata = p[0];
+	else if (msr == MSR_MTRRfix16K_80000 || msr == MSR_MTRRfix16K_A0000)
+		*pdata = p[1 + msr - MSR_MTRRfix16K_80000];
+	else if (msr >= MSR_MTRRfix4K_C0000 && msr <= MSR_MTRRfix4K_F8000)
+		*pdata = p[3 + msr - MSR_MTRRfix4K_C0000];
+	else if (msr == MSR_IA32_CR_PAT)
+		*pdata = vcpu->arch.pat;
+	else {	/* Variable MTRRs */
+		int idx, is_mtrr_mask;
+		u64 *pt;
+
+		idx = (msr - 0x200) / 2;
+		is_mtrr_mask = msr - 0x200 - 2 * idx;
+		if (!is_mtrr_mask)
+			pt =
+			  (u64 *)&vcpu->arch.mtrr_state.var_ranges[idx].base_lo;
+		else
+			pt =
+			  (u64 *)&vcpu->arch.mtrr_state.var_ranges[idx].mask_lo;
+		*pdata = *pt;
+	}
+
 	return 0;
 }
 
@@ -903,7 +964,6 @@ int kvm_dev_ioctl_check_extension(long ext)
 	case KVM_CAP_IRQCHIP:
 	case KVM_CAP_HLT:
 	case KVM_CAP_MMU_SHADOW_CACHE_CONTROL:
-	case KVM_CAP_USER_MEMORY:
 	case KVM_CAP_SET_TSS_ADDR:
 	case KVM_CAP_EXT_CPUID:
 	case KVM_CAP_CLOCKSOURCE:
@@ -1188,6 +1248,7 @@ static void do_cpuid_ent(struct kvm_cpuid_entry2 *entry, u32 function,
 		int t, times = entry->eax & 0xff;
 
 		entry->flags |= KVM_CPUID_FLAG_STATEFUL_FUNC;
+		entry->flags |= KVM_CPUID_FLAG_STATE_READ_NEXT;
 		for (t = 1; t < times && *nent < maxnent; ++t) {
 			do_cpuid_1_ent(&entry[t], function, 0);
 			entry[t].flags |= KVM_CPUID_FLAG_STATEFUL_FUNC;
@@ -1218,7 +1279,7 @@ static void do_cpuid_ent(struct kvm_cpuid_entry2 *entry, u32 function,
 		entry->flags |= KVM_CPUID_FLAG_SIGNIFCANT_INDEX;
 		/* read more entries until level_type is zero */
 		for (i = 1; *nent < maxnent; ++i) {
-			level_type = entry[i - 1].ecx & 0xff;
+			level_type = entry[i - 1].ecx & 0xff00;
 			if (!level_type)
 				break;
 			do_cpuid_1_ent(&entry[i], function, i);
@@ -1318,6 +1379,15 @@ static int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu,
 	return 0;
 }
 
+static int kvm_vcpu_ioctl_nmi(struct kvm_vcpu *vcpu)
+{
+	vcpu_load(vcpu);
+	kvm_inject_nmi(vcpu);
+	vcpu_put(vcpu);
+
+	return 0;
+}
+
 static int vcpu_ioctl_tpr_access_reporting(struct kvm_vcpu *vcpu,
 					   struct kvm_tpr_access_ctl *tac)
 {
@@ -1377,6 +1447,13 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		r = 0;
 		break;
 	}
+	case KVM_NMI: {
+		r = kvm_vcpu_ioctl_nmi(vcpu);
+		if (r)
+			goto out;
+		r = 0;
+		break;
+	}
 	case KVM_SET_CPUID: {
 		struct kvm_cpuid __user *cpuid_arg = argp;
 		struct kvm_cpuid cpuid;
@@ -1968,7 +2045,7 @@ int emulator_write_phys(struct kvm_vcpu *vcpu, gpa_t gpa,
 	ret = kvm_write_guest(vcpu->kvm, gpa, val, bytes);
 	if (ret < 0)
 		return 0;
-	kvm_mmu_pte_write(vcpu, gpa, val, bytes);
+	kvm_mmu_pte_write(vcpu, gpa, val, bytes, 1);
 	return 1;
 }
 
@@ -2404,8 +2481,6 @@ int kvm_emulate_pio(struct kvm_vcpu *vcpu, struct kvm_run *run, int in,
 	val = kvm_register_read(vcpu, VCPU_REGS_RAX);
 	memcpy(vcpu->arch.pio_data, &val, 4);
 
-	kvm_x86_ops->skip_emulated_instruction(vcpu);
-
 	pio_dev = vcpu_find_pio_dev(vcpu, port, size, !in);
 	if (pio_dev) {
 		kernel_pio(pio_dev, vcpu, vcpu->arch.pio_data);
@@ -2541,7 +2616,7 @@ int kvm_arch_init(void *opaque)
 	kvm_mmu_set_nonpresent_ptes(0ull, 0ull);
 	kvm_mmu_set_base_ptes(PT_PRESENT_MASK);
 	kvm_mmu_set_mask_ptes(PT_USER_MASK, PT_ACCESSED_MASK,
-			PT_DIRTY_MASK, PT64_NX_MASK, 0);
+			PT_DIRTY_MASK, PT64_NX_MASK, 0, 0);
 	return 0;
 
 out:
@@ -2729,7 +2804,7 @@ static int move_to_next_stateful_cpuid_entry(struct kvm_vcpu *vcpu, int i)
 
 	e->flags &= ~KVM_CPUID_FLAG_STATE_READ_NEXT;
 	/* when no next entry is found, the current entry[i] is reselected */
-	for (j = i + 1; j == i; j = (j + 1) % nent) {
+	for (j = i + 1; ; j = (j + 1) % nent) {
 		struct kvm_cpuid_entry2 *ej = &vcpu->arch.cpuid_entries[j];
 		if (ej->function == e->function) {
 			ej->flags |= KVM_CPUID_FLAG_STATE_READ_NEXT;
@@ -2973,7 +3048,7 @@ static int __vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 		pr_debug("vcpu %d received sipi with vector # %x\n",
 			 vcpu->vcpu_id, vcpu->arch.sipi_vector);
 		kvm_lapic_reset(vcpu);
-		r = kvm_x86_ops->vcpu_reset(vcpu);
+		r = kvm_arch_vcpu_reset(vcpu);
 		if (r)
 			return r;
 		vcpu->arch.mp_state = KVM_MP_STATE_RUNNABLE;
@@ -3275,9 +3350,9 @@ static void seg_desct_to_kvm_desct(struct desc_struct *seg_desc, u16 selector,
 	kvm_desct->padding = 0;
 }
 
-static void get_segment_descritptor_dtable(struct kvm_vcpu *vcpu,
-					   u16 selector,
-					   struct descriptor_table *dtable)
+static void get_segment_descriptor_dtable(struct kvm_vcpu *vcpu,
+					  u16 selector,
+					  struct descriptor_table *dtable)
 {
 	if (selector & 1 << 2) {
 		struct kvm_segment kvm_seg;
@@ -3302,7 +3377,7 @@ static int load_guest_segment_descriptor(struct kvm_vcpu *vcpu, u16 selector,
 	struct descriptor_table dtable;
 	u16 index = selector >> 3;
 
-	get_segment_descritptor_dtable(vcpu, selector, &dtable);
+	get_segment_descriptor_dtable(vcpu, selector, &dtable);
 
 	if (dtable.limit < index * 8 + 7) {
 		kvm_queue_exception_e(vcpu, GP_VECTOR, selector & 0xfffc);
@@ -3321,7 +3396,7 @@ static int save_guest_segment_descriptor(struct kvm_vcpu *vcpu, u16 selector,
 	struct descriptor_table dtable;
 	u16 index = selector >> 3;
 
-	get_segment_descritptor_dtable(vcpu, selector, &dtable);
+	get_segment_descriptor_dtable(vcpu, selector, &dtable);
 
 	if (dtable.limit < index * 8 + 7)
 		return 1;
@@ -3900,6 +3975,7 @@ int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
 	/* We do fxsave: this must be aligned. */
 	BUG_ON((unsigned long)&vcpu->arch.host_fx_image & 0xF);
 
+	vcpu->arch.mtrr_state.have_fixed = 1;
 	vcpu_load(vcpu);
 	r = kvm_arch_vcpu_reset(vcpu);
 	if (r == 0)
@@ -3925,6 +4001,9 @@ void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
 
 int kvm_arch_vcpu_reset(struct kvm_vcpu *vcpu)
 {
+	vcpu->arch.nmi_pending = false;
+	vcpu->arch.nmi_injected = false;
+
 	return kvm_x86_ops->vcpu_reset(vcpu);
 }
 
@@ -4012,6 +4091,7 @@ struct  kvm *kvm_arch_create_vm(void)
 		return ERR_PTR(-ENOMEM);
 
 	INIT_LIST_HEAD(&kvm->arch.active_mmu_pages);
+	INIT_LIST_HEAD(&kvm->arch.oos_global_pages);
 	INIT_LIST_HEAD(&kvm->arch.assigned_dev_head);
 
 	/* Reserve bit 0 of irq_sources_bitmap for userspace irq source */
@@ -4048,8 +4128,8 @@ static void kvm_free_vcpus(struct kvm *kvm)
 
 void kvm_arch_destroy_vm(struct kvm *kvm)
 {
-	kvm_iommu_unmap_guest(kvm);
 	kvm_free_all_assigned_devices(kvm);
+	kvm_iommu_unmap_guest(kvm);
 	kvm_free_pit(kvm);
 	kfree(kvm->arch.vpic);
 	kfree(kvm->arch.vioapic);
@@ -4127,7 +4207,8 @@ void kvm_arch_flush_shadow(struct kvm *kvm)
 int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu)
 {
 	return vcpu->arch.mp_state == KVM_MP_STATE_RUNNABLE
-	       || vcpu->arch.mp_state == KVM_MP_STATE_SIPI_RECEIVED;
+	       || vcpu->arch.mp_state == KVM_MP_STATE_SIPI_RECEIVED
+	       || vcpu->arch.nmi_pending;
 }
 
 static void vcpu_kick_intr(void *info)

arch/x86/kvm/x86_emulate.c

@@ -58,6 +58,7 @@
 #define SrcMem32    (4<<4)	/* Memory operand (32-bit). */
 #define SrcImm      (5<<4)	/* Immediate operand. */
 #define SrcImmByte  (6<<4)	/* 8-bit sign-extended immediate operand. */
+#define SrcOne      (7<<4)	/* Implied '1' */
 #define SrcMask     (7<<4)
 /* Generic ModRM decode. */
 #define ModRM       (1<<7)
@@ -70,17 +71,23 @@
 #define Group       (1<<14)     /* Bits 3:5 of modrm byte extend opcode */
 #define GroupDual   (1<<15)     /* Alternate decoding of mod == 3 */
 #define GroupMask   0xff        /* Group number stored in bits 0:7 */
+/* Source 2 operand type */
+#define Src2None    (0<<29)
+#define Src2CL      (1<<29)
+#define Src2ImmByte (2<<29)
+#define Src2One     (3<<29)
+#define Src2Mask    (7<<29)
 
 enum {
 	Group1_80, Group1_81, Group1_82, Group1_83,
 	Group1A, Group3_Byte, Group3, Group4, Group5, Group7,
 };
 
-static u16 opcode_table[256] = {
+static u32 opcode_table[256] = {
 	/* 0x00 - 0x07 */
 	ByteOp | DstMem | SrcReg | ModRM, DstMem | SrcReg | ModRM,
 	ByteOp | DstReg | SrcMem | ModRM, DstReg | SrcMem | ModRM,
-	0, 0, 0, 0,
+	ByteOp | DstAcc | SrcImm, DstAcc | SrcImm, 0, 0,
 	/* 0x08 - 0x0F */
 	ByteOp | DstMem | SrcReg | ModRM, DstMem | SrcReg | ModRM,
 	ByteOp | DstReg | SrcMem | ModRM, DstReg | SrcMem | ModRM,
@@ -195,7 +202,7 @@ static u16 opcode_table[256] = {
 	ImplicitOps, ImplicitOps, Group | Group4, Group | Group5,
 };
 
-static u16 twobyte_table[256] = {
+static u32 twobyte_table[256] = {
 	/* 0x00 - 0x0F */
 	0, Group | GroupDual | Group7, 0, 0, 0, 0, ImplicitOps, 0,
 	ImplicitOps, ImplicitOps, 0, 0, 0, ImplicitOps | ModRM, 0, 0,
@@ -230,9 +237,14 @@ static u16 twobyte_table[256] = {
 	/* 0x90 - 0x9F */
 	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
 	/* 0xA0 - 0xA7 */
-	0, 0, 0, DstMem | SrcReg | ModRM | BitOp, 0, 0, 0, 0,
+	0, 0, 0, DstMem | SrcReg | ModRM | BitOp,
+	DstMem | SrcReg | Src2ImmByte | ModRM,
+	DstMem | SrcReg | Src2CL | ModRM, 0, 0,
 	/* 0xA8 - 0xAF */
-	0, 0, 0, DstMem | SrcReg | ModRM | BitOp, 0, 0, ModRM, 0,
+	0, 0, 0, DstMem | SrcReg | ModRM | BitOp,
+	DstMem | SrcReg | Src2ImmByte | ModRM,
+	DstMem | SrcReg | Src2CL | ModRM,
+	ModRM, 0,
 	/* 0xB0 - 0xB7 */
 	ByteOp | DstMem | SrcReg | ModRM, DstMem | SrcReg | ModRM, 0,
 	    DstMem | SrcReg | ModRM | BitOp,
@@ -253,7 +265,7 @@ static u16 twobyte_table[256] = {
 	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
 };
 
-static u16 group_table[] = {
+static u32 group_table[] = {
 	[Group1_80*8] =
 	ByteOp | DstMem | SrcImm | ModRM, ByteOp | DstMem | SrcImm | ModRM,
 	ByteOp | DstMem | SrcImm | ModRM, ByteOp | DstMem | SrcImm | ModRM,
@@ -297,9 +309,9 @@ static u16 group_table[] = {
 	SrcMem16 | ModRM | Mov, SrcMem | ModRM | ByteOp,
 };
 
-static u16 group2_table[] = {
+static u32 group2_table[] = {
 	[Group7*8] =
-	SrcNone | ModRM, 0, 0, 0,
+	SrcNone | ModRM, 0, 0, SrcNone | ModRM,
 	SrcNone | ModRM | DstMem | Mov, 0,
 	SrcMem16 | ModRM | Mov, 0,
 };
@@ -359,49 +371,48 @@ static u16 group2_table[] = {
 	"andl %"_msk",%"_LO32 _tmp"; "		\
 	"orl  %"_LO32 _tmp",%"_sav"; "
 
+#ifdef CONFIG_X86_64
+#define ON64(x) x
+#else
+#define ON64(x)
+#endif
+
+#define ____emulate_2op(_op, _src, _dst, _eflags, _x, _y, _suffix)	\
+	do {								\
+		__asm__ __volatile__ (					\
+			_PRE_EFLAGS("0", "4", "2")			\
+			_op _suffix " %"_x"3,%1; "			\
+			_POST_EFLAGS("0", "4", "2")			\
+			: "=m" (_eflags), "=m" ((_dst).val),		\
+			  "=&r" (_tmp)					\
+			: _y ((_src).val), "i" (EFLAGS_MASK));		\
+	} while (0)
+
+
 /* Raw emulation: instruction has two explicit operands. */
 #define __emulate_2op_nobyte(_op,_src,_dst,_eflags,_wx,_wy,_lx,_ly,_qx,_qy) \
-	do { 								    \
-		unsigned long _tmp;					    \
-									    \
-		switch ((_dst).bytes) {					    \
-		case 2:							    \
-			__asm__ __volatile__ (				    \
-				_PRE_EFLAGS("0", "4", "2")		    \
-				_op"w %"_wx"3,%1; "			    \
-				_POST_EFLAGS("0", "4", "2")		    \
-				: "=m" (_eflags), "=m" ((_dst).val),        \
-				  "=&r" (_tmp)				    \
-				: _wy ((_src).val), "i" (EFLAGS_MASK));     \
-			break;						    \
-		case 4:							    \
-			__asm__ __volatile__ (				    \
-				_PRE_EFLAGS("0", "4", "2")		    \
-				_op"l %"_lx"3,%1; "			    \
-				_POST_EFLAGS("0", "4", "2")		    \
-				: "=m" (_eflags), "=m" ((_dst).val),	    \
-				  "=&r" (_tmp)				    \
-				: _ly ((_src).val), "i" (EFLAGS_MASK));     \
-			break;						    \
-		case 8:							    \
-			__emulate_2op_8byte(_op, _src, _dst,		    \
-					    _eflags, _qx, _qy);		    \
-			break;						    \
-		}							    \
+	do {								\
+		unsigned long _tmp;					\
+									\
+		switch ((_dst).bytes) {					\
+		case 2:							\
+			____emulate_2op(_op,_src,_dst,_eflags,_wx,_wy,"w"); \
+			break;						\
+		case 4:							\
+			____emulate_2op(_op,_src,_dst,_eflags,_lx,_ly,"l"); \
+			break;						\
+		case 8:							\
+			ON64(____emulate_2op(_op,_src,_dst,_eflags,_qx,_qy,"q")); \
+			break;						\
+		}							\
 	} while (0)
 
 #define __emulate_2op(_op,_src,_dst,_eflags,_bx,_by,_wx,_wy,_lx,_ly,_qx,_qy) \
 	do {								     \
-		unsigned long __tmp;					     \
+		unsigned long _tmp;					     \
 		switch ((_dst).bytes) {				             \
 		case 1:							     \
-			__asm__ __volatile__ (				     \
-				_PRE_EFLAGS("0", "4", "2")		     \
-				_op"b %"_bx"3,%1; "			     \
-				_POST_EFLAGS("0", "4", "2")		     \
-				: "=m" (_eflags), "=m" ((_dst).val),	     \
-				  "=&r" (__tmp)				     \
-				: _by ((_src).val), "i" (EFLAGS_MASK));      \
+			____emulate_2op(_op,_src,_dst,_eflags,_bx,_by,"b");  \
 			break;						     \
 		default:						     \
 			__emulate_2op_nobyte(_op, _src, _dst, _eflags,	     \
@@ -425,71 +436,68 @@ static u16 group2_table[] = {
 	__emulate_2op_nobyte(_op, _src, _dst, _eflags,			\
 			     "w", "r", _LO32, "r", "", "r")
 
-/* Instruction has only one explicit operand (no source operand). */
-#define emulate_1op(_op, _dst, _eflags)                                    \
-	do {								\
-		unsigned long _tmp;					\
-									\
-		switch ((_dst).bytes) {				        \
-		case 1:							\
-			__asm__ __volatile__ (				\
-				_PRE_EFLAGS("0", "3", "2")		\
-				_op"b %1; "				\
-				_POST_EFLAGS("0", "3", "2")		\
-				: "=m" (_eflags), "=m" ((_dst).val),	\
-				  "=&r" (_tmp)				\
-				: "i" (EFLAGS_MASK));			\
-			break;						\
-		case 2:							\
-			__asm__ __volatile__ (				\
-				_PRE_EFLAGS("0", "3", "2")		\
-				_op"w %1; "				\
-				_POST_EFLAGS("0", "3", "2")		\
-				: "=m" (_eflags), "=m" ((_dst).val),	\
-				  "=&r" (_tmp)				\
-				: "i" (EFLAGS_MASK));			\
-			break;						\
-		case 4:							\
-			__asm__ __volatile__ (				\
-				_PRE_EFLAGS("0", "3", "2")		\
-				_op"l %1; "				\
-				_POST_EFLAGS("0", "3", "2")		\
-				: "=m" (_eflags), "=m" ((_dst).val),	\
-				  "=&r" (_tmp)				\
-				: "i" (EFLAGS_MASK));			\
-			break;						\
-		case 8:							\
-			__emulate_1op_8byte(_op, _dst, _eflags);	\
-			break;						\
-		}							\
+/* Instruction has three operands and one operand is stored in ECX register */
+#define __emulate_2op_cl(_op, _cl, _src, _dst, _eflags, _suffix, _type) 	\
+	do {									\
+		unsigned long _tmp;						\
+		_type _clv  = (_cl).val;  					\
+		_type _srcv = (_src).val;    					\
+		_type _dstv = (_dst).val;					\
+										\
+		__asm__ __volatile__ (						\
+			_PRE_EFLAGS("0", "5", "2")				\
+			_op _suffix " %4,%1 \n"					\
+			_POST_EFLAGS("0", "5", "2")				\
+			: "=m" (_eflags), "+r" (_dstv), "=&r" (_tmp)		\
+			: "c" (_clv) , "r" (_srcv), "i" (EFLAGS_MASK)		\
+			); 							\
+										\
+		(_cl).val  = (unsigned long) _clv;				\
+		(_src).val = (unsigned long) _srcv;				\
+		(_dst).val = (unsigned long) _dstv;				\
 	} while (0)
 
-/* Emulate an instruction with quadword operands (x86/64 only). */
-#if defined(CONFIG_X86_64)
-#define __emulate_2op_8byte(_op, _src, _dst, _eflags, _qx, _qy)           \
-	do {								  \
-		__asm__ __volatile__ (					  \
-			_PRE_EFLAGS("0", "4", "2")			  \
-			_op"q %"_qx"3,%1; "				  \
-			_POST_EFLAGS("0", "4", "2")			  \
-			: "=m" (_eflags), "=m" ((_dst).val), "=&r" (_tmp) \
-			: _qy ((_src).val), "i" (EFLAGS_MASK));		\
+#define emulate_2op_cl(_op, _cl, _src, _dst, _eflags)				\
+	do {									\
+		switch ((_dst).bytes) {						\
+		case 2:								\
+			__emulate_2op_cl(_op, _cl, _src, _dst, _eflags,  	\
+						"w", unsigned short);         	\
+			break;							\
+		case 4: 							\
+			__emulate_2op_cl(_op, _cl, _src, _dst, _eflags,  	\
+						"l", unsigned int);           	\
+			break;							\
+		case 8:								\
+			ON64(__emulate_2op_cl(_op, _cl, _src, _dst, _eflags,	\
+						"q", unsigned long));  		\
+			break;							\
+		}								\
 	} while (0)
 
-#define __emulate_1op_8byte(_op, _dst, _eflags)                           \
-	do {								  \
-		__asm__ __volatile__ (					  \
-			_PRE_EFLAGS("0", "3", "2")			  \
-			_op"q %1; "					  \
-			_POST_EFLAGS("0", "3", "2")			  \
-			: "=m" (_eflags), "=m" ((_dst).val), "=&r" (_tmp) \
-			: "i" (EFLAGS_MASK));				  \
+#define __emulate_1op(_op, _dst, _eflags, _suffix)			\
+	do {								\
+		unsigned long _tmp;					\
+									\
+		__asm__ __volatile__ (					\
+			_PRE_EFLAGS("0", "3", "2")			\
+			_op _suffix " %1; "				\
+			_POST_EFLAGS("0", "3", "2")			\
+			: "=m" (_eflags), "+m" ((_dst).val),		\
+			  "=&r" (_tmp)					\
+			: "i" (EFLAGS_MASK));				\
 	} while (0)
 
-#elif defined(__i386__)
-#define __emulate_2op_8byte(_op, _src, _dst, _eflags, _qx, _qy)
-#define __emulate_1op_8byte(_op, _dst, _eflags)
-#endif				/* __i386__ */
+/* Instruction has only one explicit operand (no source operand). */
+#define emulate_1op(_op, _dst, _eflags)                                    \
+	do {								\
+		switch ((_dst).bytes) {				        \
+		case 1:	__emulate_1op(_op, _dst, _eflags, "b"); break;	\
+		case 2:	__emulate_1op(_op, _dst, _eflags, "w"); break;	\
+		case 4:	__emulate_1op(_op, _dst, _eflags, "l"); break;	\
+		case 8:	ON64(__emulate_1op(_op, _dst, _eflags, "q")); break; \
+		}							\
+	} while (0)
 
 /* Fetch next part of the instruction being emulated. */
 #define insn_fetch(_type, _size, _eip)                                  \
@@ -1041,6 +1049,33 @@ done_prefixes:
 		c->src.bytes = 1;
 		c->src.val = insn_fetch(s8, 1, c->eip);
 		break;
+	case SrcOne:
+		c->src.bytes = 1;
+		c->src.val = 1;
+		break;
+	}
+
+	/*
+	 * Decode and fetch the second source operand: register, memory
+	 * or immediate.
+	 */
+	switch (c->d & Src2Mask) {
+	case Src2None:
+		break;
+	case Src2CL:
+		c->src2.bytes = 1;
+		c->src2.val = c->regs[VCPU_REGS_RCX] & 0x8;
+		break;
+	case Src2ImmByte:
+		c->src2.type = OP_IMM;
+		c->src2.ptr = (unsigned long *)c->eip;
+		c->src2.bytes = 1;
+		c->src2.val = insn_fetch(u8, 1, c->eip);
+		break;
+	case Src2One:
+		c->src2.bytes = 1;
+		c->src2.val = 1;
+		break;
 	}
 
 	/* Decode and fetch the destination operand: register or memory. */
@@ -1100,20 +1135,33 @@ static inline void emulate_push(struct x86_emulate_ctxt *ctxt)
 					       c->regs[VCPU_REGS_RSP]);
 }
 
-static inline int emulate_grp1a(struct x86_emulate_ctxt *ctxt,
-				struct x86_emulate_ops *ops)
+static int emulate_pop(struct x86_emulate_ctxt *ctxt,
+		       struct x86_emulate_ops *ops)
 {
 	struct decode_cache *c = &ctxt->decode;
 	int rc;
 
-	rc = ops->read_std(register_address(c, ss_base(ctxt),
-					    c->regs[VCPU_REGS_RSP]),
-			   &c->dst.val, c->dst.bytes, ctxt->vcpu);
+	rc = ops->read_emulated(register_address(c, ss_base(ctxt),
+						 c->regs[VCPU_REGS_RSP]),
+				&c->src.val, c->src.bytes, ctxt->vcpu);
 	if (rc != 0)
 		return rc;
 
-	register_address_increment(c, &c->regs[VCPU_REGS_RSP], c->dst.bytes);
+	register_address_increment(c, &c->regs[VCPU_REGS_RSP], c->src.bytes);
+	return rc;
+}
+
+static inline int emulate_grp1a(struct x86_emulate_ctxt *ctxt,
+				struct x86_emulate_ops *ops)
+{
+	struct decode_cache *c = &ctxt->decode;
+	int rc;
 
+	c->src.bytes = c->dst.bytes;
+	rc = emulate_pop(ctxt, ops);
+	if (rc != 0)
+		return rc;
+	c->dst.val = c->src.val;
 	return 0;
 }
 
@@ -1415,24 +1463,15 @@ special_insn:
 		emulate_1op("dec", c->dst, ctxt->eflags);
 		break;
 	case 0x50 ... 0x57:  /* push reg */
-		c->dst.type  = OP_MEM;
-		c->dst.bytes = c->op_bytes;
-		c->dst.val = c->src.val;
-		register_address_increment(c, &c->regs[VCPU_REGS_RSP],
-					   -c->op_bytes);
-		c->dst.ptr = (void *) register_address(
-			c, ss_base(ctxt), c->regs[VCPU_REGS_RSP]);
+		emulate_push(ctxt);
 		break;
 	case 0x58 ... 0x5f: /* pop reg */
 	pop_instruction:
-		if ((rc = ops->read_std(register_address(c, ss_base(ctxt),
-			c->regs[VCPU_REGS_RSP]), c->dst.ptr,
-			c->op_bytes, ctxt->vcpu)) != 0)
+		c->src.bytes = c->op_bytes;
+		rc = emulate_pop(ctxt, ops);
+		if (rc != 0)
 			goto done;
-
-		register_address_increment(c, &c->regs[VCPU_REGS_RSP],
-					   c->op_bytes);
-		c->dst.type = OP_NONE;	/* Disable writeback. */
+		c->dst.val = c->src.val;
 		break;
 	case 0x63:		/* movsxd */
 		if (ctxt->mode != X86EMUL_MODE_PROT64)
@@ -1591,7 +1630,9 @@ special_insn:
 		emulate_push(ctxt);
 		break;
 	case 0x9d: /* popf */
+		c->dst.type = OP_REG;
 		c->dst.ptr = (unsigned long *) &ctxt->eflags;
+		c->dst.bytes = c->op_bytes;
 		goto pop_instruction;
 	case 0xa0 ... 0xa1:	/* mov */
 		c->dst.ptr = (unsigned long *)&c->regs[VCPU_REGS_RAX];
@@ -1689,7 +1730,9 @@ special_insn:
 		emulate_grp2(ctxt);
 		break;
 	case 0xc3: /* ret */
+		c->dst.type = OP_REG;
 		c->dst.ptr = &c->eip;
+		c->dst.bytes = c->op_bytes;
 		goto pop_instruction;
 	case 0xc6 ... 0xc7:	/* mov (sole member of Grp11) */
 	mov:
@@ -1778,7 +1821,7 @@ special_insn:
 			c->eip = saved_eip;
 			goto cannot_emulate;
 		}
-		return 0;
+		break;
 	case 0xf4:              /* hlt */
 		ctxt->vcpu->arch.halt_request = 1;
 		break;
@@ -1999,12 +2042,20 @@ twobyte_insn:
 		c->src.val &= (c->dst.bytes << 3) - 1;
 		emulate_2op_SrcV_nobyte("bt", c->src, c->dst, ctxt->eflags);
 		break;
+	case 0xa4: /* shld imm8, r, r/m */
+	case 0xa5: /* shld cl, r, r/m */
+		emulate_2op_cl("shld", c->src2, c->src, c->dst, ctxt->eflags);
+		break;
 	case 0xab:
 	      bts:		/* bts */
 		/* only subword offset */
 		c->src.val &= (c->dst.bytes << 3) - 1;
 		emulate_2op_SrcV_nobyte("bts", c->src, c->dst, ctxt->eflags);
 		break;
+	case 0xac: /* shrd imm8, r, r/m */
+	case 0xad: /* shrd cl, r, r/m */
+		emulate_2op_cl("shrd", c->src2, c->src, c->dst, ctxt->eflags);
+		break;
 	case 0xae:              /* clflush */
 		break;
 	case 0xb0 ... 0xb1:	/* cmpxchg */

fs/anon_inodes.c

@@ -79,9 +79,12 @@ int anon_inode_getfd(const char *name, const struct file_operations *fops,
 	if (IS_ERR(anon_inode_inode))
 		return -ENODEV;
 
+	if (fops->owner && !try_module_get(fops->owner))
+		return -ENOENT;
+
 	error = get_unused_fd_flags(flags);
 	if (error < 0)
-		return error;
+		goto err_module;
 	fd = error;
 
 	/*
@@ -128,6 +131,8 @@ err_dput:
 	dput(dentry);
 err_put_unused_fd:
 	put_unused_fd(fd);
+err_module:
+	module_put(fops->owner);
 	return error;
 }
 EXPORT_SYMBOL_GPL(anon_inode_getfd);

include/linux/kvm.h

@@ -83,6 +83,7 @@ struct kvm_irqchip {
 #define KVM_EXIT_S390_SIEIC       13
 #define KVM_EXIT_S390_RESET       14
 #define KVM_EXIT_DCR              15
+#define KVM_EXIT_NMI              16
 
 /* for KVM_RUN, returned by mmap(vcpu_fd, offset=0) */
 struct kvm_run {
@@ -387,6 +388,14 @@ struct kvm_trace_rec {
 #define KVM_CAP_DEVICE_ASSIGNMENT 17
 #endif
 #define KVM_CAP_IOMMU 18
+#if defined(CONFIG_X86)
+#define KVM_CAP_DEVICE_MSI 20
+#endif
+/* Bug in KVM_SET_USER_MEMORY_REGION fixed: */
+#define KVM_CAP_DESTROY_MEMORY_REGION_WORKS 21
+#if defined(CONFIG_X86)
+#define KVM_CAP_USER_NMI 22
+#endif
 
 /*
  * ioctls for VM fds
@@ -458,6 +467,8 @@ struct kvm_trace_rec {
 #define KVM_S390_INITIAL_RESET    _IO(KVMIO,  0x97)
 #define KVM_GET_MP_STATE          _IOR(KVMIO,  0x98, struct kvm_mp_state)
 #define KVM_SET_MP_STATE          _IOW(KVMIO,  0x99, struct kvm_mp_state)
+/* Available with KVM_CAP_NMI */
+#define KVM_NMI                   _IO(KVMIO,  0x9a)
 
 #define KVM_TRC_INJ_VIRQ         (KVM_TRC_HANDLER + 0x02)
 #define KVM_TRC_REDELIVER_EVT    (KVM_TRC_HANDLER + 0x03)
@@ -500,10 +511,17 @@ struct kvm_assigned_irq {
 	__u32 guest_irq;
 	__u32 flags;
 	union {
+		struct {
+			__u32 addr_lo;
+			__u32 addr_hi;
+			__u32 data;
+		} guest_msi;
 		__u32 reserved[12];
 	};
 };
 
 #define KVM_DEV_ASSIGN_ENABLE_IOMMU	(1 << 0)
 
+#define KVM_DEV_IRQ_ASSIGN_ENABLE_MSI	(1 << 0)
+
 #endif

include/linux/kvm_host.h

@@ -16,6 +16,7 @@
 #include <linux/mm.h>
 #include <linux/preempt.h>
 #include <linux/marker.h>
+#include <linux/msi.h>
 #include <asm/signal.h>
 
 #include <linux/kvm.h>
@@ -306,8 +307,14 @@ struct kvm_assigned_dev_kernel {
 	int host_busnr;
 	int host_devfn;
 	int host_irq;
+	bool host_irq_disabled;
 	int guest_irq;
-	int irq_requested;
+	struct msi_msg guest_msi;
+#define KVM_ASSIGNED_DEV_GUEST_INTX	(1 << 0)
+#define KVM_ASSIGNED_DEV_GUEST_MSI	(1 << 1)
+#define KVM_ASSIGNED_DEV_HOST_INTX	(1 << 8)
+#define KVM_ASSIGNED_DEV_HOST_MSI	(1 << 9)
+	unsigned long irq_requested_type;
 	int irq_source_id;
 	struct pci_dev *dev;
 	struct kvm *kvm;
@@ -316,8 +323,7 @@ void kvm_set_irq(struct kvm *kvm, int irq_source_id, int irq, int level);
 void kvm_notify_acked_irq(struct kvm *kvm, unsigned gsi);
 void kvm_register_irq_ack_notifier(struct kvm *kvm,
 				   struct kvm_irq_ack_notifier *kian);
-void kvm_unregister_irq_ack_notifier(struct kvm *kvm,
-				     struct kvm_irq_ack_notifier *kian);
+void kvm_unregister_irq_ack_notifier(struct kvm_irq_ack_notifier *kian);
 int kvm_request_irq_source_id(struct kvm *kvm);
 void kvm_free_irq_source_id(struct kvm *kvm, int irq_source_id);
 

virt/kvm/ioapic.c

@@ -150,10 +150,11 @@ static int ioapic_inj_irq(struct kvm_ioapic *ioapic,
 static void ioapic_inj_nmi(struct kvm_vcpu *vcpu)
 {
 	kvm_inject_nmi(vcpu);
+	kvm_vcpu_kick(vcpu);
 }
 
-static u32 ioapic_get_delivery_bitmask(struct kvm_ioapic *ioapic, u8 dest,
-				       u8 dest_mode)
+u32 kvm_ioapic_get_delivery_bitmask(struct kvm_ioapic *ioapic, u8 dest,
+				    u8 dest_mode)
 {
 	u32 mask = 0;
 	int i;
@@ -207,7 +208,8 @@ static int ioapic_deliver(struct kvm_ioapic *ioapic, int irq)
 		     "vector=%x trig_mode=%x\n",
 		     dest, dest_mode, delivery_mode, vector, trig_mode);
 
-	deliver_bitmask = ioapic_get_delivery_bitmask(ioapic, dest, dest_mode);
+	deliver_bitmask = kvm_ioapic_get_delivery_bitmask(ioapic, dest,
+							  dest_mode);
 	if (!deliver_bitmask) {
 		ioapic_debug("no target on destination\n");
 		return 0;

virt/kvm/ioapic.h

@@ -85,5 +85,7 @@ void kvm_ioapic_update_eoi(struct kvm *kvm, int vector, int trigger_mode);
 int kvm_ioapic_init(struct kvm *kvm);
 void kvm_ioapic_set_irq(struct kvm_ioapic *ioapic, int irq, int level);
 void kvm_ioapic_reset(struct kvm_ioapic *ioapic);
+u32 kvm_ioapic_get_delivery_bitmask(struct kvm_ioapic *ioapic, u8 dest,
+				u8 dest_mode);
 
 #endif

virt/kvm/irq_comm.c

@@ -61,10 +61,9 @@ void kvm_register_irq_ack_notifier(struct kvm *kvm,
 	hlist_add_head(&kian->link, &kvm->arch.irq_ack_notifier_list);
 }
 
-void kvm_unregister_irq_ack_notifier(struct kvm *kvm,
-				     struct kvm_irq_ack_notifier *kian)
+void kvm_unregister_irq_ack_notifier(struct kvm_irq_ack_notifier *kian)
 {
-	hlist_del(&kian->link);
+	hlist_del_init(&kian->link);
 }
 
 /* The caller must hold kvm->lock mutex */
@@ -73,11 +72,15 @@ int kvm_request_irq_source_id(struct kvm *kvm)
 	unsigned long *bitmap = &kvm->arch.irq_sources_bitmap;
 	int irq_source_id = find_first_zero_bit(bitmap,
 				sizeof(kvm->arch.irq_sources_bitmap));
+
 	if (irq_source_id >= sizeof(kvm->arch.irq_sources_bitmap)) {
 		printk(KERN_WARNING "kvm: exhaust allocatable IRQ sources!\n");
-		irq_source_id = -EFAULT;
-	} else
-		set_bit(irq_source_id, bitmap);
+		return -EFAULT;
+	}
+
+	ASSERT(irq_source_id != KVM_USERSPACE_IRQ_SOURCE_ID);
+	set_bit(irq_source_id, bitmap);
+
 	return irq_source_id;
 }
 
@@ -85,7 +88,9 @@ void kvm_free_irq_source_id(struct kvm *kvm, int irq_source_id)
 {
 	int i;
 
-	if (irq_source_id <= 0 ||
+	ASSERT(irq_source_id != KVM_USERSPACE_IRQ_SOURCE_ID);
+
+	if (irq_source_id < 0 ||
 	    irq_source_id >= sizeof(kvm->arch.irq_sources_bitmap)) {
 		printk(KERN_ERR "kvm: IRQ source ID out of range!\n");
 		return;

virt/kvm/kvm_main.c

@@ -47,6 +47,10 @@
 #include <asm/uaccess.h>
 #include <asm/pgtable.h>
 
+#ifdef CONFIG_X86
+#include <asm/msidef.h>
+#endif
+
 #ifdef KVM_COALESCED_MMIO_PAGE_OFFSET
 #include "coalesced_mmio.h"
 #endif
@@ -60,10 +64,13 @@
 MODULE_AUTHOR("Qumranet");
 MODULE_LICENSE("GPL");
 
+static int msi2intx = 1;
+module_param(msi2intx, bool, 0);
+
 DEFINE_SPINLOCK(kvm_lock);
 LIST_HEAD(vm_list);
 
-static cpumask_t cpus_hardware_enabled;
+static cpumask_var_t cpus_hardware_enabled;
 
 struct kmem_cache *kvm_vcpu_cache;
 EXPORT_SYMBOL_GPL(kvm_vcpu_cache);
@@ -75,9 +82,60 @@ struct dentry *kvm_debugfs_dir;
 static long kvm_vcpu_ioctl(struct file *file, unsigned int ioctl,
 			   unsigned long arg);
 
-bool kvm_rebooting;
+static bool kvm_rebooting;
 
 #ifdef KVM_CAP_DEVICE_ASSIGNMENT
+
+#ifdef CONFIG_X86
+static void assigned_device_msi_dispatch(struct kvm_assigned_dev_kernel *dev)
+{
+	int vcpu_id;
+	struct kvm_vcpu *vcpu;
+	struct kvm_ioapic *ioapic = ioapic_irqchip(dev->kvm);
+	int dest_id = (dev->guest_msi.address_lo & MSI_ADDR_DEST_ID_MASK)
+			>> MSI_ADDR_DEST_ID_SHIFT;
+	int vector = (dev->guest_msi.data & MSI_DATA_VECTOR_MASK)
+			>> MSI_DATA_VECTOR_SHIFT;
+	int dest_mode = test_bit(MSI_ADDR_DEST_MODE_SHIFT,
+				(unsigned long *)&dev->guest_msi.address_lo);
+	int trig_mode = test_bit(MSI_DATA_TRIGGER_SHIFT,
+				(unsigned long *)&dev->guest_msi.data);
+	int delivery_mode = test_bit(MSI_DATA_DELIVERY_MODE_SHIFT,
+				(unsigned long *)&dev->guest_msi.data);
+	u32 deliver_bitmask;
+
+	BUG_ON(!ioapic);
+
+	deliver_bitmask = kvm_ioapic_get_delivery_bitmask(ioapic,
+				dest_id, dest_mode);
+	/* IOAPIC delivery mode value is the same as MSI here */
+	switch (delivery_mode) {
+	case IOAPIC_LOWEST_PRIORITY:
+		vcpu = kvm_get_lowest_prio_vcpu(ioapic->kvm, vector,
+				deliver_bitmask);
+		if (vcpu != NULL)
+			kvm_apic_set_irq(vcpu, vector, trig_mode);
+		else
+			printk(KERN_INFO "kvm: null lowest priority vcpu!\n");
+		break;
+	case IOAPIC_FIXED:
+		for (vcpu_id = 0; deliver_bitmask != 0; vcpu_id++) {
+			if (!(deliver_bitmask & (1 << vcpu_id)))
+				continue;
+			deliver_bitmask &= ~(1 << vcpu_id);
+			vcpu = ioapic->kvm->vcpus[vcpu_id];
+			if (vcpu)
+				kvm_apic_set_irq(vcpu, vector, trig_mode);
+		}
+		break;
+	default:
+		printk(KERN_INFO "kvm: unsupported MSI delivery mode\n");
+	}
+}
+#else
+static void assigned_device_msi_dispatch(struct kvm_assigned_dev_kernel *dev) {}
+#endif
+
 static struct kvm_assigned_dev_kernel *kvm_find_assigned_dev(struct list_head *head,
 						      int assigned_dev_id)
 {
@@ -104,9 +162,16 @@ static void kvm_assigned_dev_interrupt_work_handler(struct work_struct *work)
 	 * finer-grained lock, update this
 	 */
 	mutex_lock(&assigned_dev->kvm->lock);
-	kvm_set_irq(assigned_dev->kvm,
-		    assigned_dev->irq_source_id,
-		    assigned_dev->guest_irq, 1);
+	if (assigned_dev->irq_requested_type & KVM_ASSIGNED_DEV_GUEST_INTX)
+		kvm_set_irq(assigned_dev->kvm,
+			    assigned_dev->irq_source_id,
+			    assigned_dev->guest_irq, 1);
+	else if (assigned_dev->irq_requested_type &
+				KVM_ASSIGNED_DEV_GUEST_MSI) {
+		assigned_device_msi_dispatch(assigned_dev);
+		enable_irq(assigned_dev->host_irq);
+		assigned_dev->host_irq_disabled = false;
+	}
 	mutex_unlock(&assigned_dev->kvm->lock);
 	kvm_put_kvm(assigned_dev->kvm);
 }
@@ -117,8 +182,12 @@ static irqreturn_t kvm_assigned_dev_intr(int irq, void *dev_id)
 		(struct kvm_assigned_dev_kernel *) dev_id;
 
 	kvm_get_kvm(assigned_dev->kvm);
+
 	schedule_work(&assigned_dev->interrupt_work);
+
 	disable_irq_nosync(irq);
+	assigned_dev->host_irq_disabled = true;
+
 	return IRQ_HANDLED;
 }
 
@@ -132,19 +201,32 @@ static void kvm_assigned_dev_ack_irq(struct kvm_irq_ack_notifier *kian)
 
 	dev = container_of(kian, struct kvm_assigned_dev_kernel,
 			   ack_notifier);
+
 	kvm_set_irq(dev->kvm, dev->irq_source_id, dev->guest_irq, 0);
-	enable_irq(dev->host_irq);
+
+	/* The guest irq may be shared so this ack may be
+	 * from another device.
+	 */
+	if (dev->host_irq_disabled) {
+		enable_irq(dev->host_irq);
+		dev->host_irq_disabled = false;
+	}
 }
 
-static void kvm_free_assigned_device(struct kvm *kvm,
-				     struct kvm_assigned_dev_kernel
-				     *assigned_dev)
+static void kvm_free_assigned_irq(struct kvm *kvm,
+				  struct kvm_assigned_dev_kernel *assigned_dev)
 {
-	if (irqchip_in_kernel(kvm) && assigned_dev->irq_requested)
-		free_irq(assigned_dev->host_irq, (void *)assigned_dev);
+	if (!irqchip_in_kernel(kvm))
+		return;
+
+	kvm_unregister_irq_ack_notifier(&assigned_dev->ack_notifier);
 
-	kvm_unregister_irq_ack_notifier(kvm, &assigned_dev->ack_notifier);
-	kvm_free_irq_source_id(kvm, assigned_dev->irq_source_id);
+	if (assigned_dev->irq_source_id != -1)
+		kvm_free_irq_source_id(kvm, assigned_dev->irq_source_id);
+	assigned_dev->irq_source_id = -1;
+
+	if (!assigned_dev->irq_requested_type)
+		return;
 
 	if (cancel_work_sync(&assigned_dev->interrupt_work))
 		/* We had pending work. That means we will have to take
@@ -152,6 +234,23 @@ static void kvm_free_assigned_device(struct kvm *kvm,
 		 */
 		kvm_put_kvm(kvm);
 
+	free_irq(assigned_dev->host_irq, (void *)assigned_dev);
+
+	if (assigned_dev->irq_requested_type & KVM_ASSIGNED_DEV_HOST_MSI)
+		pci_disable_msi(assigned_dev->dev);
+
+	assigned_dev->irq_requested_type = 0;
+}
+
+
+static void kvm_free_assigned_device(struct kvm *kvm,
+				     struct kvm_assigned_dev_kernel
+				     *assigned_dev)
+{
+	kvm_free_assigned_irq(kvm, assigned_dev);
+
+	pci_reset_function(assigned_dev->dev);
+
 	pci_release_regions(assigned_dev->dev);
 	pci_disable_device(assigned_dev->dev);
 	pci_dev_put(assigned_dev->dev);
@@ -174,6 +273,95 @@ void kvm_free_all_assigned_devices(struct kvm *kvm)
 	}
 }
 
+static int assigned_device_update_intx(struct kvm *kvm,
+			struct kvm_assigned_dev_kernel *adev,
+			struct kvm_assigned_irq *airq)
+{
+	adev->guest_irq = airq->guest_irq;
+	adev->ack_notifier.gsi = airq->guest_irq;
+
+	if (adev->irq_requested_type & KVM_ASSIGNED_DEV_HOST_INTX)
+		return 0;
+
+	if (irqchip_in_kernel(kvm)) {
+		if (!msi2intx &&
+		    adev->irq_requested_type & KVM_ASSIGNED_DEV_HOST_MSI) {
+			free_irq(adev->host_irq, (void *)kvm);
+			pci_disable_msi(adev->dev);
+		}
+
+		if (!capable(CAP_SYS_RAWIO))
+			return -EPERM;
+
+		if (airq->host_irq)
+			adev->host_irq = airq->host_irq;
+		else
+			adev->host_irq = adev->dev->irq;
+
+		/* Even though this is PCI, we don't want to use shared
+		 * interrupts. Sharing host devices with guest-assigned devices
+		 * on the same interrupt line is not a happy situation: there
+		 * are going to be long delays in accepting, acking, etc.
+		 */
+		if (request_irq(adev->host_irq, kvm_assigned_dev_intr,
+				0, "kvm_assigned_intx_device", (void *)adev))
+			return -EIO;
+	}
+
+	adev->irq_requested_type = KVM_ASSIGNED_DEV_GUEST_INTX |
+				   KVM_ASSIGNED_DEV_HOST_INTX;
+	return 0;
+}
+
+#ifdef CONFIG_X86
+static int assigned_device_update_msi(struct kvm *kvm,
+			struct kvm_assigned_dev_kernel *adev,
+			struct kvm_assigned_irq *airq)
+{
+	int r;
+
+	if (airq->flags & KVM_DEV_IRQ_ASSIGN_ENABLE_MSI) {
+		/* x86 don't care upper address of guest msi message addr */
+		adev->irq_requested_type |= KVM_ASSIGNED_DEV_GUEST_MSI;
+		adev->irq_requested_type &= ~KVM_ASSIGNED_DEV_GUEST_INTX;
+		adev->guest_msi.address_lo = airq->guest_msi.addr_lo;
+		adev->guest_msi.data = airq->guest_msi.data;
+		adev->ack_notifier.gsi = -1;
+	} else if (msi2intx) {
+		adev->irq_requested_type |= KVM_ASSIGNED_DEV_GUEST_INTX;
+		adev->irq_requested_type &= ~KVM_ASSIGNED_DEV_GUEST_MSI;
+		adev->guest_irq = airq->guest_irq;
+		adev->ack_notifier.gsi = airq->guest_irq;
+	}
+
+	if (adev->irq_requested_type & KVM_ASSIGNED_DEV_HOST_MSI)
+		return 0;
+
+	if (irqchip_in_kernel(kvm)) {
+		if (!msi2intx) {
+			if (adev->irq_requested_type &
+					KVM_ASSIGNED_DEV_HOST_INTX)
+				free_irq(adev->host_irq, (void *)adev);
+
+			r = pci_enable_msi(adev->dev);
+			if (r)
+				return r;
+		}
+
+		adev->host_irq = adev->dev->irq;
+		if (request_irq(adev->host_irq, kvm_assigned_dev_intr, 0,
+				"kvm_assigned_msi_device", (void *)adev))
+			return -EIO;
+	}
+
+	if (!msi2intx)
+		adev->irq_requested_type = KVM_ASSIGNED_DEV_GUEST_MSI;
+
+	adev->irq_requested_type |= KVM_ASSIGNED_DEV_HOST_MSI;
+	return 0;
+}
+#endif
+
 static int kvm_vm_ioctl_assign_irq(struct kvm *kvm,
 				   struct kvm_assigned_irq
 				   *assigned_irq)
@@ -190,49 +378,68 @@ static int kvm_vm_ioctl_assign_irq(struct kvm *kvm,
 		return -EINVAL;
 	}
 
-	if (match->irq_requested) {
-		match->guest_irq = assigned_irq->guest_irq;
-		match->ack_notifier.gsi = assigned_irq->guest_irq;
-		mutex_unlock(&kvm->lock);
-		return 0;
-	}
+	if (!match->irq_requested_type) {
+		INIT_WORK(&match->interrupt_work,
+				kvm_assigned_dev_interrupt_work_handler);
+		if (irqchip_in_kernel(kvm)) {
+			/* Register ack nofitier */
+			match->ack_notifier.gsi = -1;
+			match->ack_notifier.irq_acked =
+					kvm_assigned_dev_ack_irq;
+			kvm_register_irq_ack_notifier(kvm,
+					&match->ack_notifier);
+
+			/* Request IRQ source ID */
+			r = kvm_request_irq_source_id(kvm);
+			if (r < 0)
+				goto out_release;
+			else
+				match->irq_source_id = r;
 
-	INIT_WORK(&match->interrupt_work,
-		  kvm_assigned_dev_interrupt_work_handler);
+#ifdef CONFIG_X86
+			/* Determine host device irq type, we can know the
+			 * result from dev->msi_enabled */
+			if (msi2intx)
+				pci_enable_msi(match->dev);
+#endif
+		}
+	}
 
-	if (irqchip_in_kernel(kvm)) {
-		if (!capable(CAP_SYS_RAWIO)) {
-			r = -EPERM;
+	if ((!msi2intx &&
+	     (assigned_irq->flags & KVM_DEV_IRQ_ASSIGN_ENABLE_MSI)) ||
+	    (msi2intx && match->dev->msi_enabled)) {
+#ifdef CONFIG_X86
+		r = assigned_device_update_msi(kvm, match, assigned_irq);
+		if (r) {
+			printk(KERN_WARNING "kvm: failed to enable "
+					"MSI device!\n");
 			goto out_release;
 		}
-
-		if (assigned_irq->host_irq)
-			match->host_irq = assigned_irq->host_irq;
-		else
-			match->host_irq = match->dev->irq;
-		match->guest_irq = assigned_irq->guest_irq;
-		match->ack_notifier.gsi = assigned_irq->guest_irq;
-		match->ack_notifier.irq_acked = kvm_assigned_dev_ack_irq;
-		kvm_register_irq_ack_notifier(kvm, &match->ack_notifier);
-		r = kvm_request_irq_source_id(kvm);
-		if (r < 0)
+#else
+		r = -ENOTTY;
+#endif
+	} else if (assigned_irq->host_irq == 0 && match->dev->irq == 0) {
+		/* Host device IRQ 0 means don't support INTx */
+		if (!msi2intx) {
+			printk(KERN_WARNING
+			       "kvm: wait device to enable MSI!\n");
+			r = 0;
+		} else {
+			printk(KERN_WARNING
+			       "kvm: failed to enable MSI device!\n");
+			r = -ENOTTY;
 			goto out_release;
-		else
-			match->irq_source_id = r;
-
-		/* Even though this is PCI, we don't want to use shared
-		 * interrupts. Sharing host devices with guest-assigned devices
-		 * on the same interrupt line is not a happy situation: there
-		 * are going to be long delays in accepting, acking, etc.
-		 */
-		if (request_irq(match->host_irq, kvm_assigned_dev_intr, 0,
-				"kvm_assigned_device", (void *)match)) {
-			r = -EIO;
+		}
+	} else {
+		/* Non-sharing INTx mode */
+		r = assigned_device_update_intx(kvm, match, assigned_irq);
+		if (r) {
+			printk(KERN_WARNING "kvm: failed to enable "
+					"INTx device!\n");
 			goto out_release;
 		}
 	}
 
-	match->irq_requested = true;
 	mutex_unlock(&kvm->lock);
 	return r;
 out_release:
@@ -283,11 +490,14 @@ static int kvm_vm_ioctl_assign_device(struct kvm *kvm,
 		       __func__);
 		goto out_disable;
 	}
+
+	pci_reset_function(dev);
+
 	match->assigned_dev_id = assigned_dev->assigned_dev_id;
 	match->host_busnr = assigned_dev->busnr;
 	match->host_devfn = assigned_dev->devfn;
 	match->dev = dev;
-
+	match->irq_source_id = -1;
 	match->kvm = kvm;
 
 	list_add(&match->list, &kvm->arch.assigned_dev_head);
@@ -355,57 +565,48 @@ static void ack_flush(void *_completed)
 {
 }
 
-void kvm_flush_remote_tlbs(struct kvm *kvm)
+static bool make_all_cpus_request(struct kvm *kvm, unsigned int req)
 {
 	int i, cpu, me;
-	cpumask_t cpus;
+	cpumask_var_t cpus;
+	bool called = true;
 	struct kvm_vcpu *vcpu;
 
+	if (alloc_cpumask_var(&cpus, GFP_ATOMIC))
+		cpumask_clear(cpus);
+
 	me = get_cpu();
-	cpus_clear(cpus);
 	for (i = 0; i < KVM_MAX_VCPUS; ++i) {
 		vcpu = kvm->vcpus[i];
 		if (!vcpu)
 			continue;
-		if (test_and_set_bit(KVM_REQ_TLB_FLUSH, &vcpu->requests))
+		if (test_and_set_bit(req, &vcpu->requests))
 			continue;
 		cpu = vcpu->cpu;
-		if (cpu != -1 && cpu != me)
-			cpu_set(cpu, cpus);
+		if (cpus != NULL && cpu != -1 && cpu != me)
+			cpumask_set_cpu(cpu, cpus);
 	}
-	if (cpus_empty(cpus))
-		goto out;
-	++kvm->stat.remote_tlb_flush;
-	smp_call_function_mask(cpus, ack_flush, NULL, 1);
-out:
+	if (unlikely(cpus == NULL))
+		smp_call_function_many(cpu_online_mask, ack_flush, NULL, 1);
+	else if (!cpumask_empty(cpus))
+		smp_call_function_many(cpus, ack_flush, NULL, 1);
+	else
+		called = false;
 	put_cpu();
+	free_cpumask_var(cpus);
+	return called;
 }
 
-void kvm_reload_remote_mmus(struct kvm *kvm)
+void kvm_flush_remote_tlbs(struct kvm *kvm)
 {
-	int i, cpu, me;
-	cpumask_t cpus;
-	struct kvm_vcpu *vcpu;
-
-	me = get_cpu();
-	cpus_clear(cpus);
-	for (i = 0; i < KVM_MAX_VCPUS; ++i) {
-		vcpu = kvm->vcpus[i];
-		if (!vcpu)
-			continue;
-		if (test_and_set_bit(KVM_REQ_MMU_RELOAD, &vcpu->requests))
-			continue;
-		cpu = vcpu->cpu;
-		if (cpu != -1 && cpu != me)
-			cpu_set(cpu, cpus);
-	}
-	if (cpus_empty(cpus))
-		goto out;
-	smp_call_function_mask(cpus, ack_flush, NULL, 1);
-out:
-	put_cpu();
+	if (make_all_cpus_request(kvm, KVM_REQ_TLB_FLUSH))
+		++kvm->stat.remote_tlb_flush;
 }
 
+void kvm_reload_remote_mmus(struct kvm *kvm)
+{
+	make_all_cpus_request(kvm, KVM_REQ_MMU_RELOAD);
+}
 
 int kvm_vcpu_init(struct kvm_vcpu *vcpu, struct kvm *kvm, unsigned id)
 {
@@ -710,6 +911,8 @@ int __kvm_set_memory_region(struct kvm *kvm,
 		goto out;
 	if (mem->guest_phys_addr & (PAGE_SIZE - 1))
 		goto out;
+	if (user_alloc && (mem->userspace_addr & (PAGE_SIZE - 1)))
+		goto out;
 	if (mem->slot >= KVM_MEMORY_SLOTS + KVM_PRIVATE_MEM_SLOTS)
 		goto out;
 	if (mem->guest_phys_addr + mem->memory_size < mem->guest_phys_addr)
@@ -821,7 +1024,10 @@ int __kvm_set_memory_region(struct kvm *kvm,
 		goto out_free;
 	}
 
-	kvm_free_physmem_slot(&old, &new);
+	kvm_free_physmem_slot(&old, npages ? &new : NULL);
+	/* Slot deletion case: we have to update the current slot */
+	if (!npages)
+		*memslot = old;
 #ifdef CONFIG_DMAR
 	/* map the pages in iommu page table */
 	r = kvm_iommu_map_pages(kvm, base_gfn, npages);
@@ -918,7 +1124,7 @@ int kvm_is_error_hva(unsigned long addr)
 }
 EXPORT_SYMBOL_GPL(kvm_is_error_hva);
 
-static struct kvm_memory_slot *__gfn_to_memslot(struct kvm *kvm, gfn_t gfn)
+struct kvm_memory_slot *gfn_to_memslot_unaliased(struct kvm *kvm, gfn_t gfn)
 {
 	int i;
 
@@ -931,11 +1137,12 @@ static struct kvm_memory_slot *__gfn_to_memslot(struct kvm *kvm, gfn_t gfn)
 	}
 	return NULL;
 }
+EXPORT_SYMBOL_GPL(gfn_to_memslot_unaliased);
 
 struct kvm_memory_slot *gfn_to_memslot(struct kvm *kvm, gfn_t gfn)
 {
 	gfn = unalias_gfn(kvm, gfn);
-	return __gfn_to_memslot(kvm, gfn);
+	return gfn_to_memslot_unaliased(kvm, gfn);
 }
 
 int kvm_is_visible_gfn(struct kvm *kvm, gfn_t gfn)
@@ -959,7 +1166,7 @@ unsigned long gfn_to_hva(struct kvm *kvm, gfn_t gfn)
 	struct kvm_memory_slot *slot;
 
 	gfn = unalias_gfn(kvm, gfn);
-	slot = __gfn_to_memslot(kvm, gfn);
+	slot = gfn_to_memslot_unaliased(kvm, gfn);
 	if (!slot)
 		return bad_hva();
 	return (slot->userspace_addr + (gfn - slot->base_gfn) * PAGE_SIZE);
@@ -1210,7 +1417,7 @@ void mark_page_dirty(struct kvm *kvm, gfn_t gfn)
 	struct kvm_memory_slot *memslot;
 
 	gfn = unalias_gfn(kvm, gfn);
-	memslot = __gfn_to_memslot(kvm, gfn);
+	memslot = gfn_to_memslot_unaliased(kvm, gfn);
 	if (memslot && memslot->dirty_bitmap) {
 		unsigned long rel_gfn = gfn - memslot->base_gfn;
 
@@ -1295,7 +1502,7 @@ static int kvm_vcpu_release(struct inode *inode, struct file *filp)
 	return 0;
 }
 
-static const struct file_operations kvm_vcpu_fops = {
+static struct file_operations kvm_vcpu_fops = {
 	.release        = kvm_vcpu_release,
 	.unlocked_ioctl = kvm_vcpu_ioctl,
 	.compat_ioctl   = kvm_vcpu_ioctl,
@@ -1689,7 +1896,7 @@ static int kvm_vm_mmap(struct file *file, struct vm_area_struct *vma)
 	return 0;
 }
 
-static const struct file_operations kvm_vm_fops = {
+static struct file_operations kvm_vm_fops = {
 	.release        = kvm_vm_release,
 	.unlocked_ioctl = kvm_vm_ioctl,
 	.compat_ioctl   = kvm_vm_ioctl,
@@ -1711,6 +1918,18 @@ static int kvm_dev_ioctl_create_vm(void)
 	return fd;
 }
 
+static long kvm_dev_ioctl_check_extension_generic(long arg)
+{
+	switch (arg) {
+	case KVM_CAP_USER_MEMORY:
+	case KVM_CAP_DESTROY_MEMORY_REGION_WORKS:
+		return 1;
+	default:
+		break;
+	}
+	return kvm_dev_ioctl_check_extension(arg);
+}
+
 static long kvm_dev_ioctl(struct file *filp,
 			  unsigned int ioctl, unsigned long arg)
 {
@@ -1730,7 +1949,7 @@ static long kvm_dev_ioctl(struct file *filp,
 		r = kvm_dev_ioctl_create_vm();
 		break;
 	case KVM_CHECK_EXTENSION:
-		r = kvm_dev_ioctl_check_extension(arg);
+		r = kvm_dev_ioctl_check_extension_generic(arg);
 		break;
 	case KVM_GET_VCPU_MMAP_SIZE:
 		r = -EINVAL;
@@ -1771,9 +1990,9 @@ static void hardware_enable(void *junk)
 {
 	int cpu = raw_smp_processor_id();
 
-	if (cpu_isset(cpu, cpus_hardware_enabled))
+	if (cpumask_test_cpu(cpu, cpus_hardware_enabled))
 		return;
-	cpu_set(cpu, cpus_hardware_enabled);
+	cpumask_set_cpu(cpu, cpus_hardware_enabled);
 	kvm_arch_hardware_enable(NULL);
 }
 
@@ -1781,9 +2000,9 @@ static void hardware_disable(void *junk)
 {
 	int cpu = raw_smp_processor_id();
 
-	if (!cpu_isset(cpu, cpus_hardware_enabled))
+	if (!cpumask_test_cpu(cpu, cpus_hardware_enabled))
 		return;
-	cpu_clear(cpu, cpus_hardware_enabled);
+	cpumask_clear_cpu(cpu, cpus_hardware_enabled);
 	kvm_arch_hardware_disable(NULL);
 }
 
@@ -2017,9 +2236,14 @@ int kvm_init(void *opaque, unsigned int vcpu_size,
 
 	bad_pfn = page_to_pfn(bad_page);
 
+	if (!alloc_cpumask_var(&cpus_hardware_enabled, GFP_KERNEL)) {
+		r = -ENOMEM;
+		goto out_free_0;
+	}
+
 	r = kvm_arch_hardware_setup();
 	if (r < 0)
-		goto out_free_0;
+		goto out_free_0a;
 
 	for_each_online_cpu(cpu) {
 		smp_call_function_single(cpu,
@@ -2053,6 +2277,8 @@ int kvm_init(void *opaque, unsigned int vcpu_size,
 	}
 
 	kvm_chardev_ops.owner = module;
+	kvm_vm_fops.owner = module;
+	kvm_vcpu_fops.owner = module;
 
 	r = misc_register(&kvm_dev);
 	if (r) {
@@ -2062,6 +2288,9 @@ int kvm_init(void *opaque, unsigned int vcpu_size,
 
 	kvm_preempt_ops.sched_in = kvm_sched_in;
 	kvm_preempt_ops.sched_out = kvm_sched_out;
+#ifndef CONFIG_X86
+	msi2intx = 0;
+#endif
 
 	return 0;
 
@@ -2078,6 +2307,8 @@ out_free_2:
 	on_each_cpu(hardware_disable, NULL, 1);
 out_free_1:
 	kvm_arch_hardware_unsetup();
+out_free_0a:
+	free_cpumask_var(cpus_hardware_enabled);
 out_free_0:
 	__free_page(bad_page);
 out:
@@ -2101,6 +2332,7 @@ void kvm_exit(void)
 	kvm_arch_hardware_unsetup();
 	kvm_arch_exit();
 	kvm_exit_debug();
+	free_cpumask_var(cpus_hardware_enabled);
 	__free_page(bad_page);
 }
 EXPORT_SYMBOL_GPL(kvm_exit);

virt/kvm/kvm_trace.c

@@ -252,6 +252,7 @@ void kvm_trace_cleanup(void)
 			struct kvm_trace_probe *p = &kvm_trace_probes[i];
 			marker_probe_unregister(p->name, p->probe_func, p);
 		}
+		marker_synchronize_unregister();
 
 		relay_close(kt->rchan);
 		debugfs_remove(kt->lost_file);