[PATCH 0/8] Move closer to upstream

[PATCH 0/8] Move closer to upstream

[Glauber Costa]

Hi,

This is another step at getting us closer to qemu upstream. I'm moving
our code to kvm-all.c and kvm.c, and relying on the fact that we use
USE_KVM, while upstream qemu uses CONFIG_KVM, to keep us isolated.

The goal is to slowly reduce that isolation. To demonstrate what I aim
for, the last patches of the series shares code for breakpoint handling.
next in my radar are ioctl functions and cpuid trimming.

Have fun

[PATCH 1/8] Provide KVMState definition outside CONFIG_KVM

[Glauber Costa]

This will allow qemu-kvm to refer to the structure, since qemu-kvm relies
on USE_KVM, not CONFIG_KVM. This should be sent to qemu.git as well

Signed-off-by: Glauber Costa <glommer@redhat.com>
---
 kvm.h |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/kvm.h b/kvm.h
index d9723d7..7648c49 100644
--- a/kvm.h
+++ b/kvm.h
@@ -18,6 +18,9 @@
 #include "sys-queue.h"
 #include "libkvm-all.h"
 
+struct KVMState;
+typedef struct KVMState KVMState;
+
 #ifdef KVM_UPSTREAM
 
 #ifdef CONFIG_KVM
@@ -65,9 +68,6 @@ int kvm_update_guest_debug(CPUState *env, unsigned long reinject_trap);
 
 /* internal API */
 
-struct KVMState;
-typedef struct KVMState KVMState;
-
 int kvm_ioctl(KVMState *s, int type, ...);
 
 int kvm_vm_ioctl(KVMState *s, int type, ...);
-- 
1.6.2.2

Re: [PATCH 1/8] Provide KVMState definition outside CONFIG_KVM

[Avi Kivity]

On 07/08/2009 04:08 PM, Glauber Costa wrote:
> This will allow qemu-kvm to refer to the structure, since qemu-kvm relies
> on USE_KVM, not CONFIG_KVM. This should be sent to qemu.git as well
>
> Signed-off-by: Glauber Costa<glommer@redhat.com>
> ---
>   kvm.h |    6 +++---
>   1 files changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/kvm.h b/kvm.h
> index d9723d7..7648c49 100644
> --- a/kvm.h
> +++ b/kvm.h
> @@ -18,6 +18,9 @@
>   #include "sys-queue.h"
>   #include "libkvm-all.h"
>
> +struct KVMState;
> +typedef struct KVMState KVMState;
> +
>   #ifdef KVM_UPSTREAM
>
>   #ifdef CONFIG_KVM
> @@ -65,9 +68,6 @@ int kvm_update_guest_debug(CPUState *env, unsigned long reinject_trap);
>
>   /* internal API */
>
> -struct KVMState;
> -typedef struct KVMState KVMState;
> -
>   int kvm_ioctl(KVMState *s, int type, ...);
>
>   int kvm_vm_ioctl(KVMState *s, int type, ...);
>    

Not really needed - we can duplicate these two lines where we use them.  
Since they'll later go away, we avoid some (minimal) churn.


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

[PATCH 3/8] put qemu-kvm-x86.c into kvm.c

[Glauber Costa]

big plan to start sharing code with upstream.

Signed-off-by: Glauber Costa <glommer@redhat.com>
---
 Makefile.target   |    4 +-
 qemu-kvm-x86.c    | 1590 -----------------------------------------------------
 target-i386/kvm.c | 1582 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 1584 insertions(+), 1592 deletions(-)
 delete mode 100644 qemu-kvm-x86.c

diff --git a/Makefile.target b/Makefile.target
index 5a54b78..aa79141 100644
--- a/Makefile.target
+++ b/Makefile.target
@@ -174,7 +174,7 @@ libobj-y += op_helper.o helper.o
 ifeq ($(TARGET_ARCH), i386)
 libobj-y += helper.o
 ifeq ($(USE_KVM), 1)
-libobj-y += qemu-kvm-x86.o kvm-tpr-opt.o
+libobj-y += kvm.o kvm-tpr-opt.o
 libobj-y += qemu-kvm-helper.o
 endif
 endif
@@ -182,7 +182,7 @@ endif
 ifeq ($(TARGET_ARCH), x86_64)
 libobj-y += helper.o
 ifeq ($(USE_KVM), 1)
-libobj-y += qemu-kvm-x86.o kvm-tpr-opt.o
+libobj-y += kvm.o kvm-tpr-opt.o
 libobj-y += qemu-kvm-helper.o
 endif
 endif
diff --git a/qemu-kvm-x86.c b/qemu-kvm-x86.c
deleted file mode 100644
index 8ba3b6e..0000000
--- a/qemu-kvm-x86.c
+++ /dev/null
@@ -1,1590 +0,0 @@
-/*
- * qemu/kvm integration, x86 specific code
- *
- * Copyright (C) 2006-2008 Qumranet Technologies
- *
- * Licensed under the terms of the GNU GPL version 2 or higher.
- */
-
-#include "config.h"
-#include "config-host.h"
-
-#include <string.h>
-#include "hw/hw.h"
-#include "gdbstub.h"
-#include <sys/io.h>
-
-#include "qemu-kvm.h"
-#include "libkvm-all.h"
-#include "libkvm.h"
-#include <pthread.h>
-#include <sys/utsname.h>
-#include <linux/kvm_para.h>
-#include <sys/ioctl.h>
-
-#include "kvm.h"
-#include "hw/pc.h"
-
-#define MSR_IA32_TSC		0x10
-
-static struct kvm_msr_list *kvm_msr_list;
-extern unsigned int kvm_shadow_memory;
-static int kvm_has_msr_star;
-static int kvm_has_vm_hsave_pa;
-
-static int lm_capable_kernel;
-
-int kvm_set_tss_addr(kvm_context_t kvm, unsigned long addr)
-{
-#ifdef KVM_CAP_SET_TSS_ADDR
-	int r;
-
-	r = ioctl(kvm->fd, KVM_CHECK_EXTENSION, KVM_CAP_SET_TSS_ADDR);
-	if (r > 0) {
-		r = ioctl(kvm->vm_fd, KVM_SET_TSS_ADDR, addr);
-		if (r == -1) {
-			fprintf(stderr, "kvm_set_tss_addr: %m\n");
-			return -errno;
-		}
-		return 0;
-	}
-#endif
-	return -ENOSYS;
-}
-
-static int kvm_init_tss(kvm_context_t kvm)
-{
-#ifdef KVM_CAP_SET_TSS_ADDR
-	int r;
-
-	r = ioctl(kvm->fd, KVM_CHECK_EXTENSION, KVM_CAP_SET_TSS_ADDR);
-	if (r > 0) {
-		/*
-		 * this address is 3 pages before the bios, and the bios should present
-		 * as unavaible memory
-		 */
-		r = kvm_set_tss_addr(kvm, 0xfffbd000);
-		if (r < 0) {
-			fprintf(stderr, "kvm_init_tss: unable to set tss addr\n");
-			return r;
-		}
-
-	}
-#endif
-	return 0;
-}
-
-static int kvm_create_pit(kvm_context_t kvm)
-{
-#ifdef KVM_CAP_PIT
-	int r;
-
-	kvm->pit_in_kernel = 0;
-	if (!kvm->no_pit_creation) {
-		r = ioctl(kvm->fd, KVM_CHECK_EXTENSION, KVM_CAP_PIT);
-		if (r > 0) {
-			r = ioctl(kvm->vm_fd, KVM_CREATE_PIT);
-			if (r >= 0)
-				kvm->pit_in_kernel = 1;
-			else {
-				fprintf(stderr, "Create kernel PIC irqchip failed\n");
-				return r;
-			}
-		}
-	}
-#endif
-	return 0;
-}
-
-int kvm_arch_create(kvm_context_t kvm, unsigned long phys_mem_bytes,
-			void **vm_mem)
-{
-	int r = 0;
-
-	r = kvm_init_tss(kvm);
-	if (r < 0)
-		return r;
-
-	r = kvm_create_pit(kvm);
-	if (r < 0)
-		return r;
-
-	r = kvm_init_coalesced_mmio(kvm);
-	if (r < 0)
-		return r;
-
-	return 0;
-}
-
-#ifdef KVM_EXIT_TPR_ACCESS
-
-static int kvm_handle_tpr_access(kvm_vcpu_context_t vcpu)
-{
-	struct kvm_run *run = vcpu->run;
-	kvm_tpr_access_report(cpu_single_env,
-                         run->tpr_access.rip,
-                         run->tpr_access.is_write);
-    return 0;
-}
-
-
-int kvm_enable_vapic(kvm_vcpu_context_t vcpu, uint64_t vapic)
-{
-	int r;
-	struct kvm_vapic_addr va = {
-		.vapic_addr = vapic,
-	};
-
-	r = ioctl(vcpu->fd, KVM_SET_VAPIC_ADDR, &va);
-	if (r == -1) {
-		r = -errno;
-		perror("kvm_enable_vapic");
-		return r;
-	}
-	return 0;
-}
-
-#endif
-
-int kvm_arch_run(kvm_vcpu_context_t vcpu)
-{
-	int r = 0;
-	struct kvm_run *run = vcpu->run;
-
-
-	switch (run->exit_reason) {
-#ifdef KVM_EXIT_SET_TPR
-		case KVM_EXIT_SET_TPR:
-			break;
-#endif
-#ifdef KVM_EXIT_TPR_ACCESS
-		case KVM_EXIT_TPR_ACCESS:
-			r = kvm_handle_tpr_access(vcpu);
-			break;
-#endif
-		default:
-			r = 1;
-			break;
-	}
-
-	return r;
-}
-
-#define MAX_ALIAS_SLOTS 4
-static struct {
-	uint64_t start;
-	uint64_t len;
-} kvm_aliases[MAX_ALIAS_SLOTS];
-
-static int get_alias_slot(uint64_t start)
-{
-	int i;
-
-	for (i=0; i<MAX_ALIAS_SLOTS; i++)
-		if (kvm_aliases[i].start == start)
-			return i;
-	return -1;
-}
-static int get_free_alias_slot(void)
-{
-        int i;
-
-        for (i=0; i<MAX_ALIAS_SLOTS; i++)
-                if (kvm_aliases[i].len == 0)
-                        return i;
-        return -1;
-}
-
-static void register_alias(int slot, uint64_t start, uint64_t len)
-{
-	kvm_aliases[slot].start = start;
-	kvm_aliases[slot].len   = len;
-}
-
-int kvm_create_memory_alias(kvm_context_t kvm,
-			    uint64_t phys_start,
-			    uint64_t len,
-			    uint64_t target_phys)
-{
-	struct kvm_memory_alias alias = {
-		.flags = 0,
-		.guest_phys_addr = phys_start,
-		.memory_size = len,
-		.target_phys_addr = target_phys,
-	};
-	int fd = kvm->vm_fd;
-	int r;
-	int slot;
-
-	slot = get_alias_slot(phys_start);
-	if (slot < 0)
-		slot = get_free_alias_slot();
-	if (slot < 0)
-		return -EBUSY;
-	alias.slot = slot;
-
-	r = ioctl(fd, KVM_SET_MEMORY_ALIAS, &alias);
-	if (r == -1)
-	    return -errno;
-
-	register_alias(slot, phys_start, len);
-	return 0;
-}
-
-int kvm_destroy_memory_alias(kvm_context_t kvm, uint64_t phys_start)
-{
-	return kvm_create_memory_alias(kvm, phys_start, 0, 0);
-}
-
-#ifdef KVM_CAP_IRQCHIP
-
-int kvm_get_lapic(kvm_vcpu_context_t vcpu, struct kvm_lapic_state *s)
-{
-	int r;
-	if (!kvm_irqchip_in_kernel(vcpu->kvm))
-		return 0;
-	r = ioctl(vcpu->fd, KVM_GET_LAPIC, s);
-	if (r == -1) {
-		r = -errno;
-		perror("kvm_get_lapic");
-	}
-	return r;
-}
-
-int kvm_set_lapic(kvm_vcpu_context_t vcpu, struct kvm_lapic_state *s)
-{
-	int r;
-	if (!kvm_irqchip_in_kernel(vcpu->kvm))
-		return 0;
-	r = ioctl(vcpu->fd, KVM_SET_LAPIC, s);
-	if (r == -1) {
-		r = -errno;
-		perror("kvm_set_lapic");
-	}
-	return r;
-}
-
-#endif
-
-#ifdef KVM_CAP_PIT
-
-int kvm_get_pit(kvm_context_t kvm, struct kvm_pit_state *s)
-{
-	int r;
-	if (!kvm->pit_in_kernel)
-		return 0;
-	r = ioctl(kvm->vm_fd, KVM_GET_PIT, s);
-	if (r == -1) {
-		r = -errno;
-		perror("kvm_get_pit");
-	}
-	return r;
-}
-
-int kvm_set_pit(kvm_context_t kvm, struct kvm_pit_state *s)
-{
-	int r;
-	if (!kvm->pit_in_kernel)
-		return 0;
-	r = ioctl(kvm->vm_fd, KVM_SET_PIT, s);
-	if (r == -1) {
-		r = -errno;
-		perror("kvm_set_pit");
-	}
-	return r;
-}
-
-#endif
-
-void kvm_show_code(kvm_vcpu_context_t vcpu)
-{
-#define SHOW_CODE_LEN 50
-	int fd = vcpu->fd;
-	struct kvm_regs regs;
-	struct kvm_sregs sregs;
-	int r, n;
-	int back_offset;
-	unsigned char code;
-	char code_str[SHOW_CODE_LEN * 3 + 1];
-	unsigned long rip;
-	kvm_context_t kvm = vcpu->kvm;
-
-	r = ioctl(fd, KVM_GET_SREGS, &sregs);
-	if (r == -1) {
-		perror("KVM_GET_SREGS");
-		return;
-	}
-	r = ioctl(fd, KVM_GET_REGS, &regs);
-	if (r == -1) {
-		perror("KVM_GET_REGS");
-		return;
-	}
-	rip = sregs.cs.base + regs.rip;
-	back_offset = regs.rip;
-	if (back_offset > 20)
-	    back_offset = 20;
-	*code_str = 0;
-	for (n = -back_offset; n < SHOW_CODE_LEN-back_offset; ++n) {
-		if (n == 0)
-			strcat(code_str, " -->");
-		r = kvm_mmio_read(kvm->opaque, rip + n, &code, 1);
-		if (r < 0) {
-			strcat(code_str, " xx");
-			continue;
-		}
-		sprintf(code_str + strlen(code_str), " %02x", code);
-	}
-	fprintf(stderr, "code:%s\n", code_str);
-}
-
-
-/*
- * Returns available msr list.  User must free.
- */
-struct kvm_msr_list *kvm_get_msr_list(kvm_context_t kvm)
-{
-	struct kvm_msr_list sizer, *msrs;
-	int r, e;
-
-	sizer.nmsrs = 0;
-	r = ioctl(kvm->fd, KVM_GET_MSR_INDEX_LIST, &sizer);
-	if (r == -1 && errno != E2BIG)
-		return NULL;
-	/* Old kernel modules had a bug and could write beyond the provided
-	   memory. Allocate at least a safe amount of 1K. */
-	msrs = malloc(MAX(1024, sizeof(*msrs) +
-				sizer.nmsrs * sizeof(*msrs->indices)));
-	if (!msrs) {
-		errno = ENOMEM;
-		return NULL;
-	}
-	msrs->nmsrs = sizer.nmsrs;
-	r = ioctl(kvm->fd, KVM_GET_MSR_INDEX_LIST, msrs);
-	if (r == -1) {
-		e = errno;
-		free(msrs);
-		errno = e;
-		return NULL;
-	}
-	return msrs;
-}
-
-int kvm_get_msrs(kvm_vcpu_context_t vcpu, struct kvm_msr_entry *msrs, int n)
-{
-    struct kvm_msrs *kmsrs = malloc(sizeof *kmsrs + n * sizeof *msrs);
-    int r, e;
-
-    if (!kmsrs) {
-	errno = ENOMEM;
-	return -1;
-    }
-    kmsrs->nmsrs = n;
-    memcpy(kmsrs->entries, msrs, n * sizeof *msrs);
-    r = ioctl(vcpu->fd, KVM_GET_MSRS, kmsrs);
-    e = errno;
-    memcpy(msrs, kmsrs->entries, n * sizeof *msrs);
-    free(kmsrs);
-    errno = e;
-    return r;
-}
-
-int kvm_set_msrs(kvm_vcpu_context_t vcpu, struct kvm_msr_entry *msrs, int n)
-{
-    struct kvm_msrs *kmsrs = malloc(sizeof *kmsrs + n * sizeof *msrs);
-    int r, e;
-
-    if (!kmsrs) {
-	errno = ENOMEM;
-	return -1;
-    }
-    kmsrs->nmsrs = n;
-    memcpy(kmsrs->entries, msrs, n * sizeof *msrs);
-    r = ioctl(vcpu->fd, KVM_SET_MSRS, kmsrs);
-    e = errno;
-    free(kmsrs);
-    errno = e;
-    return r;
-}
-
-static void print_seg(FILE *file, const char *name, struct kvm_segment *seg)
-{
-	fprintf(stderr,
-		"%s %04x (%08llx/%08x p %d dpl %d db %d s %d type %x l %d"
-		" g %d avl %d)\n",
-		name, seg->selector, seg->base, seg->limit, seg->present,
-		seg->dpl, seg->db, seg->s, seg->type, seg->l, seg->g,
-		seg->avl);
-}
-
-static void print_dt(FILE *file, const char *name, struct kvm_dtable *dt)
-{
-	fprintf(stderr, "%s %llx/%x\n", name, dt->base, dt->limit);
-}
-
-void kvm_show_regs(kvm_vcpu_context_t vcpu)
-{
-	int fd = vcpu->fd;
-	struct kvm_regs regs;
-	struct kvm_sregs sregs;
-	int r;
-
-	r = ioctl(fd, KVM_GET_REGS, &regs);
-	if (r == -1) {
-		perror("KVM_GET_REGS");
-		return;
-	}
-	fprintf(stderr,
-		"rax %016llx rbx %016llx rcx %016llx rdx %016llx\n"
-		"rsi %016llx rdi %016llx rsp %016llx rbp %016llx\n"
-		"r8  %016llx r9  %016llx r10 %016llx r11 %016llx\n"
-		"r12 %016llx r13 %016llx r14 %016llx r15 %016llx\n"
-		"rip %016llx rflags %08llx\n",
-		regs.rax, regs.rbx, regs.rcx, regs.rdx,
-		regs.rsi, regs.rdi, regs.rsp, regs.rbp,
-		regs.r8,  regs.r9,  regs.r10, regs.r11,
-		regs.r12, regs.r13, regs.r14, regs.r15,
-		regs.rip, regs.rflags);
-	r = ioctl(fd, KVM_GET_SREGS, &sregs);
-	if (r == -1) {
-		perror("KVM_GET_SREGS");
-		return;
-	}
-	print_seg(stderr, "cs", &sregs.cs);
-	print_seg(stderr, "ds", &sregs.ds);
-	print_seg(stderr, "es", &sregs.es);
-	print_seg(stderr, "ss", &sregs.ss);
-	print_seg(stderr, "fs", &sregs.fs);
-	print_seg(stderr, "gs", &sregs.gs);
-	print_seg(stderr, "tr", &sregs.tr);
-	print_seg(stderr, "ldt", &sregs.ldt);
-	print_dt(stderr, "gdt", &sregs.gdt);
-	print_dt(stderr, "idt", &sregs.idt);
-	fprintf(stderr, "cr0 %llx cr2 %llx cr3 %llx cr4 %llx cr8 %llx"
-		" efer %llx\n",
-		sregs.cr0, sregs.cr2, sregs.cr3, sregs.cr4, sregs.cr8,
-		sregs.efer);
-}
-
-uint64_t kvm_get_apic_base(kvm_vcpu_context_t vcpu)
-{
-	return vcpu->run->apic_base;
-}
-
-void kvm_set_cr8(kvm_vcpu_context_t vcpu, uint64_t cr8)
-{
-	vcpu->run->cr8 = cr8;
-}
-
-__u64 kvm_get_cr8(kvm_vcpu_context_t vcpu)
-{
-	return vcpu->run->cr8;
-}
-
-int kvm_setup_cpuid(kvm_vcpu_context_t vcpu, int nent,
-		    struct kvm_cpuid_entry *entries)
-{
-	struct kvm_cpuid *cpuid;
-	int r;
-
-	cpuid = malloc(sizeof(*cpuid) + nent * sizeof(*entries));
-	if (!cpuid)
-		return -ENOMEM;
-
-	cpuid->nent = nent;
-	memcpy(cpuid->entries, entries, nent * sizeof(*entries));
-	r = ioctl(vcpu->fd, KVM_SET_CPUID, cpuid);
-
-	free(cpuid);
-	return r;
-}
-
-int kvm_setup_cpuid2(kvm_vcpu_context_t vcpu, int nent,
-		     struct kvm_cpuid_entry2 *entries)
-{
-	struct kvm_cpuid2 *cpuid;
-	int r;
-
-	cpuid = malloc(sizeof(*cpuid) + nent * sizeof(*entries));
-	if (!cpuid)
-		return -ENOMEM;
-
-	cpuid->nent = nent;
-	memcpy(cpuid->entries, entries, nent * sizeof(*entries));
-	r = ioctl(vcpu->fd, KVM_SET_CPUID2, cpuid);
-	if (r == -1) {
-		fprintf(stderr, "kvm_setup_cpuid2: %m\n");
-		return -errno;
-	}
-	free(cpuid);
-	return r;
-}
-
-int kvm_set_shadow_pages(kvm_context_t kvm, unsigned int nrshadow_pages)
-{
-#ifdef KVM_CAP_MMU_SHADOW_CACHE_CONTROL
-	int r;
-
-	r = ioctl(kvm->fd, KVM_CHECK_EXTENSION,
-		  KVM_CAP_MMU_SHADOW_CACHE_CONTROL);
-	if (r > 0) {
-		r = ioctl(kvm->vm_fd, KVM_SET_NR_MMU_PAGES, nrshadow_pages);
-		if (r == -1) {
-			fprintf(stderr, "kvm_set_shadow_pages: %m\n");
-			return -errno;
-		}
-		return 0;
-	}
-#endif
-	return -1;
-}
-
-int kvm_get_shadow_pages(kvm_context_t kvm, unsigned int *nrshadow_pages)
-{
-#ifdef KVM_CAP_MMU_SHADOW_CACHE_CONTROL
-	int r;
-
-	r = ioctl(kvm->fd, KVM_CHECK_EXTENSION,
-		  KVM_CAP_MMU_SHADOW_CACHE_CONTROL);
-	if (r > 0) {
-		*nrshadow_pages = ioctl(kvm->vm_fd, KVM_GET_NR_MMU_PAGES);
-		return 0;
-	}
-#endif
-	return -1;
-}
-
-#ifdef KVM_CAP_VAPIC
-
-static int tpr_access_reporting(kvm_vcpu_context_t vcpu, int enabled)
-{
-	int r;
-	struct kvm_tpr_access_ctl tac = {
-		.enabled = enabled,
-	};
-
-	r = ioctl(vcpu->kvm->fd, KVM_CHECK_EXTENSION, KVM_CAP_VAPIC);
-	if (r == -1 || r == 0)
-		return -ENOSYS;
-	r = ioctl(vcpu->fd, KVM_TPR_ACCESS_REPORTING, &tac);
-	if (r == -1) {
-		r = -errno;
-		perror("KVM_TPR_ACCESS_REPORTING");
-		return r;
-	}
-	return 0;
-}
-
-int kvm_enable_tpr_access_reporting(kvm_vcpu_context_t vcpu)
-{
-	return tpr_access_reporting(vcpu, 1);
-}
-
-int kvm_disable_tpr_access_reporting(kvm_vcpu_context_t vcpu)
-{
-	return tpr_access_reporting(vcpu, 0);
-}
-
-#endif
-
-#ifdef KVM_CAP_EXT_CPUID
-
-static struct kvm_cpuid2 *try_get_cpuid(kvm_context_t kvm, int max)
-{
-	struct kvm_cpuid2 *cpuid;
-	int r, size;
-
-	size = sizeof(*cpuid) + max * sizeof(*cpuid->entries);
-	cpuid = (struct kvm_cpuid2 *)malloc(size);
-	cpuid->nent = max;
-	r = ioctl(kvm->fd, KVM_GET_SUPPORTED_CPUID, cpuid);
-	if (r == -1)
-		r = -errno;
-	else if (r == 0 && cpuid->nent >= max)
-		r = -E2BIG;
-	if (r < 0) {
-		if (r == -E2BIG) {
-			free(cpuid);
-			return NULL;
-		} else {
-			fprintf(stderr, "KVM_GET_SUPPORTED_CPUID failed: %s\n",
-				strerror(-r));
-			exit(1);
-		}
-	}
-	return cpuid;
-}
-
-#define R_EAX 0
-#define R_ECX 1
-#define R_EDX 2
-#define R_EBX 3
-#define R_ESP 4
-#define R_EBP 5
-#define R_ESI 6
-#define R_EDI 7
-
-uint32_t kvm_get_supported_cpuid(kvm_context_t kvm, uint32_t function, int reg)
-{
-	struct kvm_cpuid2 *cpuid;
-	int i, max;
-	uint32_t ret = 0;
-	uint32_t cpuid_1_edx;
-
-	if (!kvm_check_extension(kvm, KVM_CAP_EXT_CPUID)) {
-		return -1U;
-	}
-
-	max = 1;
-	while ((cpuid = try_get_cpuid(kvm, max)) == NULL) {
-		max *= 2;
-	}
-
-	for (i = 0; i < cpuid->nent; ++i) {
-		if (cpuid->entries[i].function == function) {
-			switch (reg) {
-			case R_EAX:
-				ret = cpuid->entries[i].eax;
-				break;
-			case R_EBX:
-				ret = cpuid->entries[i].ebx;
-				break;
-			case R_ECX:
-				ret = cpuid->entries[i].ecx;
-				break;
-			case R_EDX:
-				ret = cpuid->entries[i].edx;
-                                if (function == 1) {
-                                    /* kvm misreports the following features
-                                     */
-                                    ret |= 1 << 12; /* MTRR */
-                                    ret |= 1 << 16; /* PAT */
-                                    ret |= 1 << 7;  /* MCE */
-                                    ret |= 1 << 14; /* MCA */
-                                }
-
-				/* On Intel, kvm returns cpuid according to
-				 * the Intel spec, so add missing bits
-				 * according to the AMD spec:
-				 */
-				if (function == 0x80000001) {
-					cpuid_1_edx = kvm_get_supported_cpuid(kvm, 1, R_EDX);
-					ret |= cpuid_1_edx & 0xdfeff7ff;
-				}
-				break;
-			}
-		}
-	}
-
-	free(cpuid);
-
-	return ret;
-}
-
-#else
-
-uint32_t kvm_get_supported_cpuid(kvm_context_t kvm, uint32_t function, int reg)
-{
-	return -1U;
-}
-
-#endif
-int kvm_qemu_create_memory_alias(uint64_t phys_start,
-                                 uint64_t len,
-                                 uint64_t target_phys)
-{
-    return kvm_create_memory_alias(kvm_context, phys_start, len, target_phys);
-}
-
-int kvm_qemu_destroy_memory_alias(uint64_t phys_start)
-{
-	return kvm_destroy_memory_alias(kvm_context, phys_start);
-}
-
-int kvm_arch_qemu_create_context(void)
-{
-    int i;
-    struct utsname utsname;
-
-    uname(&utsname);
-    lm_capable_kernel = strcmp(utsname.machine, "x86_64") == 0;
-
-    if (kvm_shadow_memory)
-        kvm_set_shadow_pages(kvm_context, kvm_shadow_memory);
-
-    kvm_msr_list = kvm_get_msr_list(kvm_context);
-    if (!kvm_msr_list)
-		return -1;
-    for (i = 0; i < kvm_msr_list->nmsrs; ++i) {
-	if (kvm_msr_list->indices[i] == MSR_STAR)
-	    kvm_has_msr_star = 1;
-        if (kvm_msr_list->indices[i] == MSR_VM_HSAVE_PA)
-            kvm_has_vm_hsave_pa = 1;
-    }
-
-    return 0;
-}
-
-static void set_msr_entry(struct kvm_msr_entry *entry, uint32_t index,
-                          uint64_t data)
-{
-    entry->index = index;
-    entry->data  = data;
-}
-
-/* returns 0 on success, non-0 on failure */
-static int get_msr_entry(struct kvm_msr_entry *entry, CPUState *env)
-{
-        switch (entry->index) {
-        case MSR_IA32_SYSENTER_CS:
-            env->sysenter_cs  = entry->data;
-            break;
-        case MSR_IA32_SYSENTER_ESP:
-            env->sysenter_esp = entry->data;
-            break;
-        case MSR_IA32_SYSENTER_EIP:
-            env->sysenter_eip = entry->data;
-            break;
-        case MSR_STAR:
-            env->star         = entry->data;
-            break;
-#ifdef TARGET_X86_64
-        case MSR_CSTAR:
-            env->cstar        = entry->data;
-            break;
-        case MSR_KERNELGSBASE:
-            env->kernelgsbase = entry->data;
-            break;
-        case MSR_FMASK:
-            env->fmask        = entry->data;
-            break;
-        case MSR_LSTAR:
-            env->lstar        = entry->data;
-            break;
-#endif
-        case MSR_IA32_TSC:
-            env->tsc          = entry->data;
-            break;
-        case MSR_VM_HSAVE_PA:
-            env->vm_hsave     = entry->data;
-            break;
-        default:
-            printf("Warning unknown msr index 0x%x\n", entry->index);
-            return 1;
-        }
-        return 0;
-}
-
-#ifdef TARGET_X86_64
-#define MSR_COUNT 9
-#else
-#define MSR_COUNT 5
-#endif
-
-static void set_v8086_seg(struct kvm_segment *lhs, const SegmentCache *rhs)
-{
-    lhs->selector = rhs->selector;
-    lhs->base = rhs->base;
-    lhs->limit = rhs->limit;
-    lhs->type = 3;
-    lhs->present = 1;
-    lhs->dpl = 3;
-    lhs->db = 0;
-    lhs->s = 1;
-    lhs->l = 0;
-    lhs->g = 0;
-    lhs->avl = 0;
-    lhs->unusable = 0;
-}
-
-static void set_seg(struct kvm_segment *lhs, const SegmentCache *rhs)
-{
-    unsigned flags = rhs->flags;
-    lhs->selector = rhs->selector;
-    lhs->base = rhs->base;
-    lhs->limit = rhs->limit;
-    lhs->type = (flags >> DESC_TYPE_SHIFT) & 15;
-    lhs->present = (flags & DESC_P_MASK) != 0;
-    lhs->dpl = rhs->selector & 3;
-    lhs->db = (flags >> DESC_B_SHIFT) & 1;
-    lhs->s = (flags & DESC_S_MASK) != 0;
-    lhs->l = (flags >> DESC_L_SHIFT) & 1;
-    lhs->g = (flags & DESC_G_MASK) != 0;
-    lhs->avl = (flags & DESC_AVL_MASK) != 0;
-    lhs->unusable = 0;
-}
-
-static void get_seg(SegmentCache *lhs, const struct kvm_segment *rhs)
-{
-    lhs->selector = rhs->selector;
-    lhs->base = rhs->base;
-    lhs->limit = rhs->limit;
-    lhs->flags =
-	(rhs->type << DESC_TYPE_SHIFT)
-	| (rhs->present * DESC_P_MASK)
-	| (rhs->dpl << DESC_DPL_SHIFT)
-	| (rhs->db << DESC_B_SHIFT)
-	| (rhs->s * DESC_S_MASK)
-	| (rhs->l << DESC_L_SHIFT)
-	| (rhs->g * DESC_G_MASK)
-	| (rhs->avl * DESC_AVL_MASK);
-}
-
-void kvm_arch_load_regs(CPUState *env)
-{
-    struct kvm_regs regs;
-    struct kvm_fpu fpu;
-    struct kvm_sregs sregs;
-    struct kvm_msr_entry msrs[MSR_COUNT];
-    int rc, n, i;
-
-    regs.rax = env->regs[R_EAX];
-    regs.rbx = env->regs[R_EBX];
-    regs.rcx = env->regs[R_ECX];
-    regs.rdx = env->regs[R_EDX];
-    regs.rsi = env->regs[R_ESI];
-    regs.rdi = env->regs[R_EDI];
-    regs.rsp = env->regs[R_ESP];
-    regs.rbp = env->regs[R_EBP];
-#ifdef TARGET_X86_64
-    regs.r8 = env->regs[8];
-    regs.r9 = env->regs[9];
-    regs.r10 = env->regs[10];
-    regs.r11 = env->regs[11];
-    regs.r12 = env->regs[12];
-    regs.r13 = env->regs[13];
-    regs.r14 = env->regs[14];
-    regs.r15 = env->regs[15];
-#endif
-
-    regs.rflags = env->eflags;
-    regs.rip = env->eip;
-
-    kvm_set_regs(env->kvm_cpu_state.vcpu_ctx, &regs);
-
-    memset(&fpu, 0, sizeof fpu);
-    fpu.fsw = env->fpus & ~(7 << 11);
-    fpu.fsw |= (env->fpstt & 7) << 11;
-    fpu.fcw = env->fpuc;
-    for (i = 0; i < 8; ++i)
-	fpu.ftwx |= (!env->fptags[i]) << i;
-    memcpy(fpu.fpr, env->fpregs, sizeof env->fpregs);
-    memcpy(fpu.xmm, env->xmm_regs, sizeof env->xmm_regs);
-    fpu.mxcsr = env->mxcsr;
-    kvm_set_fpu(env->kvm_cpu_state.vcpu_ctx, &fpu);
-
-    memcpy(sregs.interrupt_bitmap, env->interrupt_bitmap, sizeof(sregs.interrupt_bitmap));
-
-    if ((env->eflags & VM_MASK)) {
-	    set_v8086_seg(&sregs.cs, &env->segs[R_CS]);
-	    set_v8086_seg(&sregs.ds, &env->segs[R_DS]);
-	    set_v8086_seg(&sregs.es, &env->segs[R_ES]);
-	    set_v8086_seg(&sregs.fs, &env->segs[R_FS]);
-	    set_v8086_seg(&sregs.gs, &env->segs[R_GS]);
-	    set_v8086_seg(&sregs.ss, &env->segs[R_SS]);
-    } else {
-	    set_seg(&sregs.cs, &env->segs[R_CS]);
-	    set_seg(&sregs.ds, &env->segs[R_DS]);
-	    set_seg(&sregs.es, &env->segs[R_ES]);
-	    set_seg(&sregs.fs, &env->segs[R_FS]);
-	    set_seg(&sregs.gs, &env->segs[R_GS]);
-	    set_seg(&sregs.ss, &env->segs[R_SS]);
-
-	    if (env->cr[0] & CR0_PE_MASK) {
-		/* force ss cpl to cs cpl */
-		sregs.ss.selector = (sregs.ss.selector & ~3) |
-			(sregs.cs.selector & 3);
-		sregs.ss.dpl = sregs.ss.selector & 3;
-	    }
-    }
-
-    set_seg(&sregs.tr, &env->tr);
-    set_seg(&sregs.ldt, &env->ldt);
-
-    sregs.idt.limit = env->idt.limit;
-    sregs.idt.base = env->idt.base;
-    sregs.gdt.limit = env->gdt.limit;
-    sregs.gdt.base = env->gdt.base;
-
-    sregs.cr0 = env->cr[0];
-    sregs.cr2 = env->cr[2];
-    sregs.cr3 = env->cr[3];
-    sregs.cr4 = env->cr[4];
-
-    sregs.cr8 = cpu_get_apic_tpr(env);
-    sregs.apic_base = cpu_get_apic_base(env);
-
-    sregs.efer = env->efer;
-
-    kvm_set_sregs(env->kvm_cpu_state.vcpu_ctx, &sregs);
-
-    /* msrs */
-    n = 0;
-    set_msr_entry(&msrs[n++], MSR_IA32_SYSENTER_CS,  env->sysenter_cs);
-    set_msr_entry(&msrs[n++], MSR_IA32_SYSENTER_ESP, env->sysenter_esp);
-    set_msr_entry(&msrs[n++], MSR_IA32_SYSENTER_EIP, env->sysenter_eip);
-    if (kvm_has_msr_star)
-	set_msr_entry(&msrs[n++], MSR_STAR,              env->star);
-    if (kvm_has_vm_hsave_pa)
-        set_msr_entry(&msrs[n++], MSR_VM_HSAVE_PA, env->vm_hsave);
-#ifdef TARGET_X86_64
-    if (lm_capable_kernel) {
-        set_msr_entry(&msrs[n++], MSR_CSTAR,             env->cstar);
-        set_msr_entry(&msrs[n++], MSR_KERNELGSBASE,      env->kernelgsbase);
-        set_msr_entry(&msrs[n++], MSR_FMASK,             env->fmask);
-        set_msr_entry(&msrs[n++], MSR_LSTAR  ,           env->lstar);
-    }
-#endif
-
-    rc = kvm_set_msrs(env->kvm_cpu_state.vcpu_ctx, msrs, n);
-    if (rc == -1)
-        perror("kvm_set_msrs FAILED");
-}
-
-void kvm_load_tsc(CPUState *env)
-{
-    int rc;
-    struct kvm_msr_entry msr;
-
-    set_msr_entry(&msr, MSR_IA32_TSC, env->tsc);
-
-    rc = kvm_set_msrs(env->kvm_cpu_state.vcpu_ctx, &msr, 1);
-    if (rc == -1)
-        perror("kvm_set_tsc FAILED.\n");
-}
-
-void kvm_arch_save_mpstate(CPUState *env)
-{
-#ifdef KVM_CAP_MP_STATE
-    int r;
-    struct kvm_mp_state mp_state;
-
-    r = kvm_get_mpstate(env->kvm_cpu_state.vcpu_ctx, &mp_state);
-    if (r < 0)
-        env->mp_state = -1;
-    else
-        env->mp_state = mp_state.mp_state;
-#endif
-}
-
-void kvm_arch_load_mpstate(CPUState *env)
-{
-#ifdef KVM_CAP_MP_STATE
-    struct kvm_mp_state mp_state = { .mp_state = env->mp_state };
-
-    /*
-     * -1 indicates that the host did not support GET_MP_STATE ioctl,
-     *  so don't touch it.
-     */
-    if (env->mp_state != -1)
-        kvm_set_mpstate(env->kvm_cpu_state.vcpu_ctx, &mp_state);
-#endif
-}
-
-void kvm_arch_save_regs(CPUState *env)
-{
-    struct kvm_regs regs;
-    struct kvm_fpu fpu;
-    struct kvm_sregs sregs;
-    struct kvm_msr_entry msrs[MSR_COUNT];
-    uint32_t hflags;
-    uint32_t i, n, rc;
-
-    kvm_get_regs(env->kvm_cpu_state.vcpu_ctx, &regs);
-
-    env->regs[R_EAX] = regs.rax;
-    env->regs[R_EBX] = regs.rbx;
-    env->regs[R_ECX] = regs.rcx;
-    env->regs[R_EDX] = regs.rdx;
-    env->regs[R_ESI] = regs.rsi;
-    env->regs[R_EDI] = regs.rdi;
-    env->regs[R_ESP] = regs.rsp;
-    env->regs[R_EBP] = regs.rbp;
-#ifdef TARGET_X86_64
-    env->regs[8] = regs.r8;
-    env->regs[9] = regs.r9;
-    env->regs[10] = regs.r10;
-    env->regs[11] = regs.r11;
-    env->regs[12] = regs.r12;
-    env->regs[13] = regs.r13;
-    env->regs[14] = regs.r14;
-    env->regs[15] = regs.r15;
-#endif
-
-    env->eflags = regs.rflags;
-    env->eip = regs.rip;
-
-    kvm_get_fpu(env->kvm_cpu_state.vcpu_ctx, &fpu);
-    env->fpstt = (fpu.fsw >> 11) & 7;
-    env->fpus = fpu.fsw;
-    env->fpuc = fpu.fcw;
-    for (i = 0; i < 8; ++i)
-	env->fptags[i] = !((fpu.ftwx >> i) & 1);
-    memcpy(env->fpregs, fpu.fpr, sizeof env->fpregs);
-    memcpy(env->xmm_regs, fpu.xmm, sizeof env->xmm_regs);
-    env->mxcsr = fpu.mxcsr;
-
-    kvm_get_sregs(env->kvm_cpu_state.vcpu_ctx, &sregs);
-
-    memcpy(env->interrupt_bitmap, sregs.interrupt_bitmap, sizeof(env->interrupt_bitmap));
-
-    get_seg(&env->segs[R_CS], &sregs.cs);
-    get_seg(&env->segs[R_DS], &sregs.ds);
-    get_seg(&env->segs[R_ES], &sregs.es);
-    get_seg(&env->segs[R_FS], &sregs.fs);
-    get_seg(&env->segs[R_GS], &sregs.gs);
-    get_seg(&env->segs[R_SS], &sregs.ss);
-
-    get_seg(&env->tr, &sregs.tr);
-    get_seg(&env->ldt, &sregs.ldt);
-
-    env->idt.limit = sregs.idt.limit;
-    env->idt.base = sregs.idt.base;
-    env->gdt.limit = sregs.gdt.limit;
-    env->gdt.base = sregs.gdt.base;
-
-    env->cr[0] = sregs.cr0;
-    env->cr[2] = sregs.cr2;
-    env->cr[3] = sregs.cr3;
-    env->cr[4] = sregs.cr4;
-
-    cpu_set_apic_base(env, sregs.apic_base);
-
-    env->efer = sregs.efer;
-    //cpu_set_apic_tpr(env, sregs.cr8);
-
-#define HFLAG_COPY_MASK ~( \
-			HF_CPL_MASK | HF_PE_MASK | HF_MP_MASK | HF_EM_MASK | \
-			HF_TS_MASK | HF_TF_MASK | HF_VM_MASK | HF_IOPL_MASK | \
-			HF_OSFXSR_MASK | HF_LMA_MASK | HF_CS32_MASK | \
-			HF_SS32_MASK | HF_CS64_MASK | HF_ADDSEG_MASK)
-
-
-
-    hflags = (env->segs[R_CS].flags >> DESC_DPL_SHIFT) & HF_CPL_MASK;
-    hflags |= (env->cr[0] & CR0_PE_MASK) << (HF_PE_SHIFT - CR0_PE_SHIFT);
-    hflags |= (env->cr[0] << (HF_MP_SHIFT - CR0_MP_SHIFT)) &
-	    (HF_MP_MASK | HF_EM_MASK | HF_TS_MASK);
-    hflags |= (env->eflags & (HF_TF_MASK | HF_VM_MASK | HF_IOPL_MASK));
-    hflags |= (env->cr[4] & CR4_OSFXSR_MASK) <<
-	    (HF_OSFXSR_SHIFT - CR4_OSFXSR_SHIFT);
-
-    if (env->efer & MSR_EFER_LMA) {
-        hflags |= HF_LMA_MASK;
-    }
-
-    if ((hflags & HF_LMA_MASK) && (env->segs[R_CS].flags & DESC_L_MASK)) {
-        hflags |= HF_CS32_MASK | HF_SS32_MASK | HF_CS64_MASK;
-    } else {
-        hflags |= (env->segs[R_CS].flags & DESC_B_MASK) >>
-		(DESC_B_SHIFT - HF_CS32_SHIFT);
-        hflags |= (env->segs[R_SS].flags & DESC_B_MASK) >>
-		(DESC_B_SHIFT - HF_SS32_SHIFT);
-        if (!(env->cr[0] & CR0_PE_MASK) ||
-                   (env->eflags & VM_MASK) ||
-                   !(hflags & HF_CS32_MASK)) {
-                hflags |= HF_ADDSEG_MASK;
-            } else {
-                hflags |= ((env->segs[R_DS].base |
-                                env->segs[R_ES].base |
-                                env->segs[R_SS].base) != 0) <<
-                    HF_ADDSEG_SHIFT;
-            }
-    }
-    env->hflags = (env->hflags & HFLAG_COPY_MASK) | hflags;
-
-    /* msrs */
-    n = 0;
-    msrs[n++].index = MSR_IA32_SYSENTER_CS;
-    msrs[n++].index = MSR_IA32_SYSENTER_ESP;
-    msrs[n++].index = MSR_IA32_SYSENTER_EIP;
-    if (kvm_has_msr_star)
-	msrs[n++].index = MSR_STAR;
-    msrs[n++].index = MSR_IA32_TSC;
-    if (kvm_has_vm_hsave_pa)
-        msrs[n++].index = MSR_VM_HSAVE_PA;
-#ifdef TARGET_X86_64
-    if (lm_capable_kernel) {
-        msrs[n++].index = MSR_CSTAR;
-        msrs[n++].index = MSR_KERNELGSBASE;
-        msrs[n++].index = MSR_FMASK;
-        msrs[n++].index = MSR_LSTAR;
-    }
-#endif
-    rc = kvm_get_msrs(env->kvm_cpu_state.vcpu_ctx, msrs, n);
-    if (rc == -1) {
-        perror("kvm_get_msrs FAILED");
-    }
-    else {
-        n = rc; /* actual number of MSRs */
-        for (i=0 ; i<n; i++) {
-            if (get_msr_entry(&msrs[i], env))
-                return;
-        }
-    }
-}
-
-static void do_cpuid_ent(struct kvm_cpuid_entry2 *e, uint32_t function,
-                         uint32_t count, CPUState *env)
-{
-    env->regs[R_EAX] = function;
-    env->regs[R_ECX] = count;
-    qemu_kvm_cpuid_on_env(env);
-    e->function = function;
-    e->flags = 0;
-    e->index = 0;
-    e->eax = env->regs[R_EAX];
-    e->ebx = env->regs[R_EBX];
-    e->ecx = env->regs[R_ECX];
-    e->edx = env->regs[R_EDX];
-}
-
-struct kvm_para_features {
-	int cap;
-	int feature;
-} para_features[] = {
-#ifdef KVM_CAP_CLOCKSOURCE
-	{ KVM_CAP_CLOCKSOURCE, KVM_FEATURE_CLOCKSOURCE },
-#endif
-#ifdef KVM_CAP_NOP_IO_DELAY
-	{ KVM_CAP_NOP_IO_DELAY, KVM_FEATURE_NOP_IO_DELAY },
-#endif
-#ifdef KVM_CAP_PV_MMU
-	{ KVM_CAP_PV_MMU, KVM_FEATURE_MMU_OP },
-#endif
-#ifdef KVM_CAP_CR3_CACHE
-	{ KVM_CAP_CR3_CACHE, KVM_FEATURE_CR3_CACHE },
-#endif
-	{ -1, -1 }
-};
-
-static int get_para_features(kvm_context_t kvm_context)
-{
-	int i, features = 0;
-
-	for (i = 0; i < ARRAY_SIZE(para_features)-1; i++) {
-		if (kvm_check_extension(kvm_context, para_features[i].cap))
-			features |= (1 << para_features[i].feature);
-	}
-
-	return features;
-}
-
-static void kvm_trim_features(uint32_t *features, uint32_t supported)
-{
-    int i;
-    uint32_t mask;
-
-    for (i = 0; i < 32; ++i) {
-        mask = 1U << i;
-        if ((*features & mask) && !(supported & mask)) {
-            *features &= ~mask;
-        }
-    }
-}
-
-int kvm_arch_qemu_init_env(CPUState *cenv)
-{
-    struct kvm_cpuid_entry2 cpuid_ent[100];
-#ifdef KVM_CPUID_SIGNATURE
-    struct kvm_cpuid_entry2 *pv_ent;
-    uint32_t signature[3];
-#endif
-    int cpuid_nent = 0;
-    CPUState copy;
-    uint32_t i, j, limit;
-
-    qemu_kvm_load_lapic(cenv);
-
-
-#ifdef KVM_CPUID_SIGNATURE
-    /* Paravirtualization CPUIDs */
-    memcpy(signature, "KVMKVMKVM\0\0\0", 12);
-    pv_ent = &cpuid_ent[cpuid_nent++];
-    memset(pv_ent, 0, sizeof(*pv_ent));
-    pv_ent->function = KVM_CPUID_SIGNATURE;
-    pv_ent->eax = 0;
-    pv_ent->ebx = signature[0];
-    pv_ent->ecx = signature[1];
-    pv_ent->edx = signature[2];
-
-    pv_ent = &cpuid_ent[cpuid_nent++];
-    memset(pv_ent, 0, sizeof(*pv_ent));
-    pv_ent->function = KVM_CPUID_FEATURES;
-    pv_ent->eax = get_para_features(kvm_context);
-#endif
-
-    kvm_trim_features(&cenv->cpuid_features,
-                      kvm_arch_get_supported_cpuid(cenv, 1, R_EDX));
-
-    /* prevent the hypervisor bit from being cleared by the kernel */
-    i = cenv->cpuid_ext_features & CPUID_EXT_HYPERVISOR;
-    kvm_trim_features(&cenv->cpuid_ext_features,
-                      kvm_arch_get_supported_cpuid(cenv, 1, R_ECX));
-    cenv->cpuid_ext_features |= i;
-
-    kvm_trim_features(&cenv->cpuid_ext2_features,
-                      kvm_arch_get_supported_cpuid(cenv, 0x80000001, R_EDX));
-    kvm_trim_features(&cenv->cpuid_ext3_features,
-                      kvm_arch_get_supported_cpuid(cenv, 0x80000001, R_ECX));
-
-    copy = *cenv;
-
-    copy.regs[R_EAX] = 0;
-    qemu_kvm_cpuid_on_env(&copy);
-    limit = copy.regs[R_EAX];
-
-    for (i = 0; i <= limit; ++i) {
-        if (i == 4 || i == 0xb || i == 0xd) {
-            for (j = 0; ; ++j) {
-                do_cpuid_ent(&cpuid_ent[cpuid_nent], i, j, &copy);
-
-                cpuid_ent[cpuid_nent].flags = KVM_CPUID_FLAG_SIGNIFCANT_INDEX;
-                cpuid_ent[cpuid_nent].index = j;
-
-                cpuid_nent++;
-
-                if (i == 4 && copy.regs[R_EAX] == 0)
-                    break;
-                if (i == 0xb && !(copy.regs[R_ECX] & 0xff00))
-                    break;
-                if (i == 0xd && copy.regs[R_EAX] == 0)
-                    break;
-            }
-        } else
-            do_cpuid_ent(&cpuid_ent[cpuid_nent++], i, 0, &copy);
-    }
-
-    copy.regs[R_EAX] = 0x80000000;
-    qemu_kvm_cpuid_on_env(&copy);
-    limit = copy.regs[R_EAX];
-
-    for (i = 0x80000000; i <= limit; ++i)
-	do_cpuid_ent(&cpuid_ent[cpuid_nent++], i, 0, &copy);
-
-    kvm_setup_cpuid2(cenv->kvm_cpu_state.vcpu_ctx, cpuid_nent, cpuid_ent);
-
-    return 0;
-}
-
-int kvm_arch_halt(void *opaque, kvm_vcpu_context_t vcpu)
-{
-    CPUState *env = cpu_single_env;
-
-    if (!((env->interrupt_request & CPU_INTERRUPT_HARD) &&
-	  (env->eflags & IF_MASK)) &&
-	!(env->interrupt_request & CPU_INTERRUPT_NMI)) {
-            env->halted = 1;
-    }
-    return 1;
-}
-
-void kvm_arch_pre_kvm_run(void *opaque, CPUState *env)
-{
-    if (!kvm_irqchip_in_kernel(kvm_context))
-	kvm_set_cr8(env->kvm_cpu_state.vcpu_ctx, cpu_get_apic_tpr(env));
-}
-
-void kvm_arch_post_kvm_run(void *opaque, CPUState *env)
-{
-    cpu_single_env = env;
-
-    env->eflags = kvm_get_interrupt_flag(env->kvm_cpu_state.vcpu_ctx)
-	? env->eflags | IF_MASK : env->eflags & ~IF_MASK;
-
-    cpu_set_apic_tpr(env, kvm_get_cr8(env->kvm_cpu_state.vcpu_ctx));
-    cpu_set_apic_base(env, kvm_get_apic_base(env->kvm_cpu_state.vcpu_ctx));
-}
-
-int kvm_arch_has_work(CPUState *env)
-{
-    if (((env->interrupt_request & CPU_INTERRUPT_HARD) &&
-	 (env->eflags & IF_MASK)) ||
-	(env->interrupt_request & CPU_INTERRUPT_NMI))
-	return 1;
-    return 0;
-}
-
-int kvm_arch_try_push_interrupts(void *opaque)
-{
-    CPUState *env = cpu_single_env;
-    int r, irq;
-
-    if (kvm_is_ready_for_interrupt_injection(env->kvm_cpu_state.vcpu_ctx) &&
-        (env->interrupt_request & CPU_INTERRUPT_HARD) &&
-        (env->eflags & IF_MASK)) {
-            env->interrupt_request &= ~CPU_INTERRUPT_HARD;
-	    irq = cpu_get_pic_interrupt(env);
-	    if (irq >= 0) {
-		r = kvm_inject_irq(env->kvm_cpu_state.vcpu_ctx, irq);
-		if (r < 0)
-		    printf("cpu %d fail inject %x\n", env->cpu_index, irq);
-	    }
-    }
-
-    return (env->interrupt_request & CPU_INTERRUPT_HARD) != 0;
-}
-
-#ifdef KVM_CAP_USER_NMI
-void kvm_arch_push_nmi(void *opaque)
-{
-    CPUState *env = cpu_single_env;
-    int r;
-
-    if (likely(!(env->interrupt_request & CPU_INTERRUPT_NMI)))
-        return;
-
-    env->interrupt_request &= ~CPU_INTERRUPT_NMI;
-    r = kvm_inject_nmi(env->kvm_cpu_state.vcpu_ctx);
-    if (r < 0)
-        printf("cpu %d fail inject NMI\n", env->cpu_index);
-}
-#endif /* KVM_CAP_USER_NMI */
-
-void kvm_arch_update_regs_for_sipi(CPUState *env)
-{
-    SegmentCache cs = env->segs[R_CS];
-
-    kvm_arch_save_regs(env);
-    env->segs[R_CS] = cs;
-    env->eip = 0;
-    kvm_arch_load_regs(env);
-}
-
-void kvm_arch_cpu_reset(CPUState *env)
-{
-    kvm_arch_load_regs(env);
-    if (!cpu_is_bsp(env)) {
-	if (kvm_irqchip_in_kernel(kvm_context)) {
-#ifdef KVM_CAP_MP_STATE
-	    kvm_reset_mpstate(env->kvm_cpu_state.vcpu_ctx);
-#endif
-	} else {
-	    env->interrupt_request &= ~CPU_INTERRUPT_HARD;
-	    env->halted = 1;
-	}
-    }
-}
-
-int kvm_arch_insert_sw_breakpoint(CPUState *env, struct kvm_sw_breakpoint *bp)
-{
-    uint8_t int3 = 0xcc;
-
-    if (cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&bp->saved_insn, 1, 0) ||
-        cpu_memory_rw_debug(env, bp->pc, &int3, 1, 1))
-        return -EINVAL;
-    return 0;
-}
-
-int kvm_arch_remove_sw_breakpoint(CPUState *env, struct kvm_sw_breakpoint *bp)
-{
-    uint8_t int3;
-
-    if (cpu_memory_rw_debug(env, bp->pc, &int3, 1, 0) || int3 != 0xcc ||
-        cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&bp->saved_insn, 1, 1))
-        return -EINVAL;
-    return 0;
-}
-
-#ifdef KVM_CAP_SET_GUEST_DEBUG
-static struct {
-    target_ulong addr;
-    int len;
-    int type;
-} hw_breakpoint[4];
-
-static int nb_hw_breakpoint;
-
-static int find_hw_breakpoint(target_ulong addr, int len, int type)
-{
-    int n;
-
-    for (n = 0; n < nb_hw_breakpoint; n++)
-	if (hw_breakpoint[n].addr == addr && hw_breakpoint[n].type == type &&
-	    (hw_breakpoint[n].len == len || len == -1))
-	    return n;
-    return -1;
-}
-
-int kvm_arch_insert_hw_breakpoint(target_ulong addr,
-                                  target_ulong len, int type)
-{
-    switch (type) {
-    case GDB_BREAKPOINT_HW:
-	len = 1;
-	break;
-    case GDB_WATCHPOINT_WRITE:
-    case GDB_WATCHPOINT_ACCESS:
-	switch (len) {
-	case 1:
-	    break;
-	case 2:
-	case 4:
-	case 8:
-	    if (addr & (len - 1))
-		return -EINVAL;
-	    break;
-	default:
-	    return -EINVAL;
-	}
-	break;
-    default:
-	return -ENOSYS;
-    }
-
-    if (nb_hw_breakpoint == 4)
-        return -ENOBUFS;
-
-    if (find_hw_breakpoint(addr, len, type) >= 0)
-        return -EEXIST;
-
-    hw_breakpoint[nb_hw_breakpoint].addr = addr;
-    hw_breakpoint[nb_hw_breakpoint].len = len;
-    hw_breakpoint[nb_hw_breakpoint].type = type;
-    nb_hw_breakpoint++;
-
-    return 0;
-}
-
-int kvm_arch_remove_hw_breakpoint(target_ulong addr,
-                                  target_ulong len, int type)
-{
-    int n;
-
-    n = find_hw_breakpoint(addr, (type == GDB_BREAKPOINT_HW) ? 1 : len, type);
-    if (n < 0)
-        return -ENOENT;
-
-    nb_hw_breakpoint--;
-    hw_breakpoint[n] = hw_breakpoint[nb_hw_breakpoint];
-
-    return 0;
-}
-
-void kvm_arch_remove_all_hw_breakpoints(void)
-{
-    nb_hw_breakpoint = 0;
-}
-
-static CPUWatchpoint hw_watchpoint;
-
-int kvm_arch_debug(struct kvm_debug_exit_arch *arch_info)
-{
-    int handle = 0;
-    int n;
-
-    if (arch_info->exception == 1) {
-	if (arch_info->dr6 & (1 << 14)) {
-	    if (cpu_single_env->singlestep_enabled)
-		handle = 1;
-	} else {
-	    for (n = 0; n < 4; n++)
-		if (arch_info->dr6 & (1 << n))
-		    switch ((arch_info->dr7 >> (16 + n*4)) & 0x3) {
-		    case 0x0:
-			handle = 1;
-			break;
-		    case 0x1:
-			handle = 1;
-			cpu_single_env->watchpoint_hit = &hw_watchpoint;
-			hw_watchpoint.vaddr = hw_breakpoint[n].addr;
-			hw_watchpoint.flags = BP_MEM_WRITE;
-			break;
-		    case 0x3:
-			handle = 1;
-			cpu_single_env->watchpoint_hit = &hw_watchpoint;
-			hw_watchpoint.vaddr = hw_breakpoint[n].addr;
-			hw_watchpoint.flags = BP_MEM_ACCESS;
-			break;
-		    }
-	}
-    } else if (kvm_find_sw_breakpoint(arch_info->pc))
-	handle = 1;
-
-    if (!handle)
-	kvm_update_guest_debug(cpu_single_env,
-			(arch_info->exception == 1) ?
-			KVM_GUESTDBG_INJECT_DB : KVM_GUESTDBG_INJECT_BP);
-
-    return handle;
-}
-
-void kvm_arch_update_guest_debug(CPUState *env, struct kvm_guest_debug *dbg)
-{
-    const uint8_t type_code[] = {
-	[GDB_BREAKPOINT_HW] = 0x0,
-	[GDB_WATCHPOINT_WRITE] = 0x1,
-	[GDB_WATCHPOINT_ACCESS] = 0x3
-    };
-    const uint8_t len_code[] = {
-	[1] = 0x0, [2] = 0x1, [4] = 0x3, [8] = 0x2
-    };
-    int n;
-
-    if (!TAILQ_EMPTY(&kvm_sw_breakpoints))
-	dbg->control |= KVM_GUESTDBG_ENABLE | KVM_GUESTDBG_USE_SW_BP;
-
-    if (nb_hw_breakpoint > 0) {
-	dbg->control |= KVM_GUESTDBG_ENABLE | KVM_GUESTDBG_USE_HW_BP;
-	dbg->arch.debugreg[7] = 0x0600;
-	for (n = 0; n < nb_hw_breakpoint; n++) {
-	    dbg->arch.debugreg[n] = hw_breakpoint[n].addr;
-	    dbg->arch.debugreg[7] |= (2 << (n * 2)) |
-		(type_code[hw_breakpoint[n].type] << (16 + n*4)) |
-		(len_code[hw_breakpoint[n].len] << (18 + n*4));
-	}
-    }
-}
-#endif
-
-void kvm_arch_do_ioperm(void *_data)
-{
-    struct ioperm_data *data = _data;
-    ioperm(data->start_port, data->num, data->turn_on);
-}
-
-/*
- * Setup x86 specific IRQ routing
- */
-int kvm_arch_init_irq_routing(void)
-{
-    int i, r;
-
-    if (kvm_irqchip && kvm_has_gsi_routing(kvm_context)) {
-        kvm_clear_gsi_routes(kvm_context);
-        for (i = 0; i < 8; ++i) {
-            if (i == 2)
-                continue;
-            r = kvm_add_irq_route(kvm_context, i, KVM_IRQCHIP_PIC_MASTER, i);
-            if (r < 0)
-                return r;
-        }
-        for (i = 8; i < 16; ++i) {
-            r = kvm_add_irq_route(kvm_context, i, KVM_IRQCHIP_PIC_SLAVE, i - 8);
-            if (r < 0)
-                return r;
-        }
-        for (i = 0; i < 24; ++i) {
-            r = kvm_add_irq_route(kvm_context, i, KVM_IRQCHIP_IOAPIC, i);
-            if (r < 0)
-                return r;
-        }
-        kvm_commit_irq_routes(kvm_context);
-    }
-    return 0;
-}
-
-uint32_t kvm_arch_get_supported_cpuid(CPUState *env, uint32_t function,
-                                      int reg)
-{
-    return kvm_get_supported_cpuid(kvm_context, function, reg);
-}
-
-void kvm_arch_process_irqchip_events(CPUState *env)
-{
-    kvm_arch_save_regs(env);
-    if (env->interrupt_request & CPU_INTERRUPT_INIT)
-        do_cpu_init(env);
-    if (env->interrupt_request & CPU_INTERRUPT_SIPI)
-        do_cpu_sipi(env);
-    kvm_arch_load_regs(env);
-}
diff --git a/target-i386/kvm.c b/target-i386/kvm.c
index 0af2f55..4c3948c 100644
--- a/target-i386/kvm.c
+++ b/target-i386/kvm.c
@@ -14,15 +14,28 @@
 
 #include <sys/types.h>
 #include <sys/ioctl.h>
+#include <sys/io.h>
 #include <sys/mman.h>
+#include <sys/utsname.h>
+#include <pthread.h>
 
 #include <linux/kvm.h>
+#include <linux/kvm_para.h>
 
 #include "qemu-common.h"
 #include "sysemu.h"
 #include "kvm.h"
 #include "cpu.h"
 #include "gdbstub.h"
+#include "hw/hw.h"
+#include "hw/pc.h"
+
+
+#include "qemu-kvm.h"
+#include "libkvm-all.h"
+#include "libkvm.h"
+
+#ifdef CONFIG_KVM
 
 //#define DEBUG_KVM
 
@@ -959,3 +972,1572 @@ void kvm_arch_update_guest_debug(CPUState *env, struct kvm_guest_debug *dbg)
     }
 }
 #endif /* KVM_CAP_SET_GUEST_DEBUG */
+
+#endif /* CONFIG_KVM */
+#ifdef USE_KVM
+
+#define MSR_IA32_TSC		0x10
+
+static struct kvm_msr_list *kvm_msr_list;
+extern unsigned int kvm_shadow_memory;
+static int kvm_has_msr_star;
+static int kvm_has_vm_hsave_pa;
+
+static int lm_capable_kernel;
+
+int kvm_set_tss_addr(kvm_context_t kvm, unsigned long addr)
+{
+#ifdef KVM_CAP_SET_TSS_ADDR
+	int r;
+
+	r = ioctl(kvm->fd, KVM_CHECK_EXTENSION, KVM_CAP_SET_TSS_ADDR);
+	if (r > 0) {
+		r = ioctl(kvm->vm_fd, KVM_SET_TSS_ADDR, addr);
+		if (r == -1) {
+			fprintf(stderr, "kvm_set_tss_addr: %m\n");
+			return -errno;
+		}
+		return 0;
+	}
+#endif
+	return -ENOSYS;
+}
+
+static int kvm_init_tss(kvm_context_t kvm)
+{
+#ifdef KVM_CAP_SET_TSS_ADDR
+	int r;
+
+	r = ioctl(kvm->fd, KVM_CHECK_EXTENSION, KVM_CAP_SET_TSS_ADDR);
+	if (r > 0) {
+		/*
+		 * this address is 3 pages before the bios, and the bios should present
+		 * as unavaible memory
+		 */
+		r = kvm_set_tss_addr(kvm, 0xfffbd000);
+		if (r < 0) {
+			fprintf(stderr, "kvm_init_tss: unable to set tss addr\n");
+			return r;
+		}
+
+	}
+#endif
+	return 0;
+}
+
+static int kvm_create_pit(kvm_context_t kvm)
+{
+#ifdef KVM_CAP_PIT
+	int r;
+
+	kvm->pit_in_kernel = 0;
+	if (!kvm->no_pit_creation) {
+		r = ioctl(kvm->fd, KVM_CHECK_EXTENSION, KVM_CAP_PIT);
+		if (r > 0) {
+			r = ioctl(kvm->vm_fd, KVM_CREATE_PIT);
+			if (r >= 0)
+				kvm->pit_in_kernel = 1;
+			else {
+				fprintf(stderr, "Create kernel PIC irqchip failed\n");
+				return r;
+			}
+		}
+	}
+#endif
+	return 0;
+}
+
+int kvm_arch_create(kvm_context_t kvm, unsigned long phys_mem_bytes,
+			void **vm_mem)
+{
+	int r = 0;
+
+	r = kvm_init_tss(kvm);
+	if (r < 0)
+		return r;
+
+	r = kvm_create_pit(kvm);
+	if (r < 0)
+		return r;
+
+	r = kvm_init_coalesced_mmio(kvm);
+	if (r < 0)
+		return r;
+
+	return 0;
+}
+
+#ifdef KVM_EXIT_TPR_ACCESS
+
+static int kvm_handle_tpr_access(kvm_vcpu_context_t vcpu)
+{
+	struct kvm_run *run = vcpu->run;
+	kvm_tpr_access_report(cpu_single_env,
+                         run->tpr_access.rip,
+                         run->tpr_access.is_write);
+    return 0;
+}
+
+
+int kvm_enable_vapic(kvm_vcpu_context_t vcpu, uint64_t vapic)
+{
+	int r;
+	struct kvm_vapic_addr va = {
+		.vapic_addr = vapic,
+	};
+
+	r = ioctl(vcpu->fd, KVM_SET_VAPIC_ADDR, &va);
+	if (r == -1) {
+		r = -errno;
+		perror("kvm_enable_vapic");
+		return r;
+	}
+	return 0;
+}
+
+#endif
+
+int kvm_arch_run(kvm_vcpu_context_t vcpu)
+{
+	int r = 0;
+	struct kvm_run *run = vcpu->run;
+
+
+	switch (run->exit_reason) {
+#ifdef KVM_EXIT_SET_TPR
+		case KVM_EXIT_SET_TPR:
+			break;
+#endif
+#ifdef KVM_EXIT_TPR_ACCESS
+		case KVM_EXIT_TPR_ACCESS:
+			r = kvm_handle_tpr_access(vcpu);
+			break;
+#endif
+		default:
+			r = 1;
+			break;
+	}
+
+	return r;
+}
+
+#define MAX_ALIAS_SLOTS 4
+static struct {
+	uint64_t start;
+	uint64_t len;
+} kvm_aliases[MAX_ALIAS_SLOTS];
+
+static int get_alias_slot(uint64_t start)
+{
+	int i;
+
+	for (i=0; i<MAX_ALIAS_SLOTS; i++)
+		if (kvm_aliases[i].start == start)
+			return i;
+	return -1;
+}
+static int get_free_alias_slot(void)
+{
+        int i;
+
+        for (i=0; i<MAX_ALIAS_SLOTS; i++)
+                if (kvm_aliases[i].len == 0)
+                        return i;
+        return -1;
+}
+
+static void register_alias(int slot, uint64_t start, uint64_t len)
+{
+	kvm_aliases[slot].start = start;
+	kvm_aliases[slot].len   = len;
+}
+
+int kvm_create_memory_alias(kvm_context_t kvm,
+			    uint64_t phys_start,
+			    uint64_t len,
+			    uint64_t target_phys)
+{
+	struct kvm_memory_alias alias = {
+		.flags = 0,
+		.guest_phys_addr = phys_start,
+		.memory_size = len,
+		.target_phys_addr = target_phys,
+	};
+	int fd = kvm->vm_fd;
+	int r;
+	int slot;
+
+	slot = get_alias_slot(phys_start);
+	if (slot < 0)
+		slot = get_free_alias_slot();
+	if (slot < 0)
+		return -EBUSY;
+	alias.slot = slot;
+
+	r = ioctl(fd, KVM_SET_MEMORY_ALIAS, &alias);
+	if (r == -1)
+	    return -errno;
+
+	register_alias(slot, phys_start, len);
+	return 0;
+}
+
+int kvm_destroy_memory_alias(kvm_context_t kvm, uint64_t phys_start)
+{
+	return kvm_create_memory_alias(kvm, phys_start, 0, 0);
+}
+
+#ifdef KVM_CAP_IRQCHIP
+
+int kvm_get_lapic(kvm_vcpu_context_t vcpu, struct kvm_lapic_state *s)
+{
+	int r;
+	if (!kvm_irqchip_in_kernel(vcpu->kvm))
+		return 0;
+	r = ioctl(vcpu->fd, KVM_GET_LAPIC, s);
+	if (r == -1) {
+		r = -errno;
+		perror("kvm_get_lapic");
+	}
+	return r;
+}
+
+int kvm_set_lapic(kvm_vcpu_context_t vcpu, struct kvm_lapic_state *s)
+{
+	int r;
+	if (!kvm_irqchip_in_kernel(vcpu->kvm))
+		return 0;
+	r = ioctl(vcpu->fd, KVM_SET_LAPIC, s);
+	if (r == -1) {
+		r = -errno;
+		perror("kvm_set_lapic");
+	}
+	return r;
+}
+
+#endif
+
+#ifdef KVM_CAP_PIT
+
+int kvm_get_pit(kvm_context_t kvm, struct kvm_pit_state *s)
+{
+	int r;
+	if (!kvm->pit_in_kernel)
+		return 0;
+	r = ioctl(kvm->vm_fd, KVM_GET_PIT, s);
+	if (r == -1) {
+		r = -errno;
+		perror("kvm_get_pit");
+	}
+	return r;
+}
+
+int kvm_set_pit(kvm_context_t kvm, struct kvm_pit_state *s)
+{
+	int r;
+	if (!kvm->pit_in_kernel)
+		return 0;
+	r = ioctl(kvm->vm_fd, KVM_SET_PIT, s);
+	if (r == -1) {
+		r = -errno;
+		perror("kvm_set_pit");
+	}
+	return r;
+}
+
+#endif
+
+void kvm_show_code(kvm_vcpu_context_t vcpu)
+{
+#define SHOW_CODE_LEN 50
+	int fd = vcpu->fd;
+	struct kvm_regs regs;
+	struct kvm_sregs sregs;
+	int r, n;
+	int back_offset;
+	unsigned char code;
+	char code_str[SHOW_CODE_LEN * 3 + 1];
+	unsigned long rip;
+	kvm_context_t kvm = vcpu->kvm;
+
+	r = ioctl(fd, KVM_GET_SREGS, &sregs);
+	if (r == -1) {
+		perror("KVM_GET_SREGS");
+		return;
+	}
+	r = ioctl(fd, KVM_GET_REGS, &regs);
+	if (r == -1) {
+		perror("KVM_GET_REGS");
+		return;
+	}
+	rip = sregs.cs.base + regs.rip;
+	back_offset = regs.rip;
+	if (back_offset > 20)
+	    back_offset = 20;
+	*code_str = 0;
+	for (n = -back_offset; n < SHOW_CODE_LEN-back_offset; ++n) {
+		if (n == 0)
+			strcat(code_str, " -->");
+		r = kvm_mmio_read(kvm->opaque, rip + n, &code, 1);
+		if (r < 0) {
+			strcat(code_str, " xx");
+			continue;
+		}
+		sprintf(code_str + strlen(code_str), " %02x", code);
+	}
+	fprintf(stderr, "code:%s\n", code_str);
+}
+
+
+/*
+ * Returns available msr list.  User must free.
+ */
+struct kvm_msr_list *kvm_get_msr_list(kvm_context_t kvm)
+{
+	struct kvm_msr_list sizer, *msrs;
+	int r, e;
+
+	sizer.nmsrs = 0;
+	r = ioctl(kvm->fd, KVM_GET_MSR_INDEX_LIST, &sizer);
+	if (r == -1 && errno != E2BIG)
+		return NULL;
+	/* Old kernel modules had a bug and could write beyond the provided
+	   memory. Allocate at least a safe amount of 1K. */
+	msrs = malloc(MAX(1024, sizeof(*msrs) +
+				sizer.nmsrs * sizeof(*msrs->indices)));
+	if (!msrs) {
+		errno = ENOMEM;
+		return NULL;
+	}
+	msrs->nmsrs = sizer.nmsrs;
+	r = ioctl(kvm->fd, KVM_GET_MSR_INDEX_LIST, msrs);
+	if (r == -1) {
+		e = errno;
+		free(msrs);
+		errno = e;
+		return NULL;
+	}
+	return msrs;
+}
+
+int kvm_get_msrs(kvm_vcpu_context_t vcpu, struct kvm_msr_entry *msrs, int n)
+{
+    struct kvm_msrs *kmsrs = malloc(sizeof *kmsrs + n * sizeof *msrs);
+    int r, e;
+
+    if (!kmsrs) {
+	errno = ENOMEM;
+	return -1;
+    }
+    kmsrs->nmsrs = n;
+    memcpy(kmsrs->entries, msrs, n * sizeof *msrs);
+    r = ioctl(vcpu->fd, KVM_GET_MSRS, kmsrs);
+    e = errno;
+    memcpy(msrs, kmsrs->entries, n * sizeof *msrs);
+    free(kmsrs);
+    errno = e;
+    return r;
+}
+
+int kvm_set_msrs(kvm_vcpu_context_t vcpu, struct kvm_msr_entry *msrs, int n)
+{
+    struct kvm_msrs *kmsrs = malloc(sizeof *kmsrs + n * sizeof *msrs);
+    int r, e;
+
+    if (!kmsrs) {
+	errno = ENOMEM;
+	return -1;
+    }
+    kmsrs->nmsrs = n;
+    memcpy(kmsrs->entries, msrs, n * sizeof *msrs);
+    r = ioctl(vcpu->fd, KVM_SET_MSRS, kmsrs);
+    e = errno;
+    free(kmsrs);
+    errno = e;
+    return r;
+}
+
+static void print_seg(FILE *file, const char *name, struct kvm_segment *seg)
+{
+	fprintf(stderr,
+		"%s %04x (%08llx/%08x p %d dpl %d db %d s %d type %x l %d"
+		" g %d avl %d)\n",
+		name, seg->selector, seg->base, seg->limit, seg->present,
+		seg->dpl, seg->db, seg->s, seg->type, seg->l, seg->g,
+		seg->avl);
+}
+
+static void print_dt(FILE *file, const char *name, struct kvm_dtable *dt)
+{
+	fprintf(stderr, "%s %llx/%x\n", name, dt->base, dt->limit);
+}
+
+void kvm_show_regs(kvm_vcpu_context_t vcpu)
+{
+	int fd = vcpu->fd;
+	struct kvm_regs regs;
+	struct kvm_sregs sregs;
+	int r;
+
+	r = ioctl(fd, KVM_GET_REGS, &regs);
+	if (r == -1) {
+		perror("KVM_GET_REGS");
+		return;
+	}
+	fprintf(stderr,
+		"rax %016llx rbx %016llx rcx %016llx rdx %016llx\n"
+		"rsi %016llx rdi %016llx rsp %016llx rbp %016llx\n"
+		"r8  %016llx r9  %016llx r10 %016llx r11 %016llx\n"
+		"r12 %016llx r13 %016llx r14 %016llx r15 %016llx\n"
+		"rip %016llx rflags %08llx\n",
+		regs.rax, regs.rbx, regs.rcx, regs.rdx,
+		regs.rsi, regs.rdi, regs.rsp, regs.rbp,
+		regs.r8,  regs.r9,  regs.r10, regs.r11,
+		regs.r12, regs.r13, regs.r14, regs.r15,
+		regs.rip, regs.rflags);
+	r = ioctl(fd, KVM_GET_SREGS, &sregs);
+	if (r == -1) {
+		perror("KVM_GET_SREGS");
+		return;
+	}
+	print_seg(stderr, "cs", &sregs.cs);
+	print_seg(stderr, "ds", &sregs.ds);
+	print_seg(stderr, "es", &sregs.es);
+	print_seg(stderr, "ss", &sregs.ss);
+	print_seg(stderr, "fs", &sregs.fs);
+	print_seg(stderr, "gs", &sregs.gs);
+	print_seg(stderr, "tr", &sregs.tr);
+	print_seg(stderr, "ldt", &sregs.ldt);
+	print_dt(stderr, "gdt", &sregs.gdt);
+	print_dt(stderr, "idt", &sregs.idt);
+	fprintf(stderr, "cr0 %llx cr2 %llx cr3 %llx cr4 %llx cr8 %llx"
+		" efer %llx\n",
+		sregs.cr0, sregs.cr2, sregs.cr3, sregs.cr4, sregs.cr8,
+		sregs.efer);
+}
+
+uint64_t kvm_get_apic_base(kvm_vcpu_context_t vcpu)
+{
+	return vcpu->run->apic_base;
+}
+
+void kvm_set_cr8(kvm_vcpu_context_t vcpu, uint64_t cr8)
+{
+	vcpu->run->cr8 = cr8;
+}
+
+__u64 kvm_get_cr8(kvm_vcpu_context_t vcpu)
+{
+	return vcpu->run->cr8;
+}
+
+int kvm_setup_cpuid(kvm_vcpu_context_t vcpu, int nent,
+		    struct kvm_cpuid_entry *entries)
+{
+	struct kvm_cpuid *cpuid;
+	int r;
+
+	cpuid = malloc(sizeof(*cpuid) + nent * sizeof(*entries));
+	if (!cpuid)
+		return -ENOMEM;
+
+	cpuid->nent = nent;
+	memcpy(cpuid->entries, entries, nent * sizeof(*entries));
+	r = ioctl(vcpu->fd, KVM_SET_CPUID, cpuid);
+
+	free(cpuid);
+	return r;
+}
+
+int kvm_setup_cpuid2(kvm_vcpu_context_t vcpu, int nent,
+		     struct kvm_cpuid_entry2 *entries)
+{
+	struct kvm_cpuid2 *cpuid;
+	int r;
+
+	cpuid = malloc(sizeof(*cpuid) + nent * sizeof(*entries));
+	if (!cpuid)
+		return -ENOMEM;
+
+	cpuid->nent = nent;
+	memcpy(cpuid->entries, entries, nent * sizeof(*entries));
+	r = ioctl(vcpu->fd, KVM_SET_CPUID2, cpuid);
+	if (r == -1) {
+		fprintf(stderr, "kvm_setup_cpuid2: %m\n");
+		return -errno;
+	}
+	free(cpuid);
+	return r;
+}
+
+int kvm_set_shadow_pages(kvm_context_t kvm, unsigned int nrshadow_pages)
+{
+#ifdef KVM_CAP_MMU_SHADOW_CACHE_CONTROL
+	int r;
+
+	r = ioctl(kvm->fd, KVM_CHECK_EXTENSION,
+		  KVM_CAP_MMU_SHADOW_CACHE_CONTROL);
+	if (r > 0) {
+		r = ioctl(kvm->vm_fd, KVM_SET_NR_MMU_PAGES, nrshadow_pages);
+		if (r == -1) {
+			fprintf(stderr, "kvm_set_shadow_pages: %m\n");
+			return -errno;
+		}
+		return 0;
+	}
+#endif
+	return -1;
+}
+
+int kvm_get_shadow_pages(kvm_context_t kvm, unsigned int *nrshadow_pages)
+{
+#ifdef KVM_CAP_MMU_SHADOW_CACHE_CONTROL
+	int r;
+
+	r = ioctl(kvm->fd, KVM_CHECK_EXTENSION,
+		  KVM_CAP_MMU_SHADOW_CACHE_CONTROL);
+	if (r > 0) {
+		*nrshadow_pages = ioctl(kvm->vm_fd, KVM_GET_NR_MMU_PAGES);
+		return 0;
+	}
+#endif
+	return -1;
+}
+
+#ifdef KVM_CAP_VAPIC
+
+static int tpr_access_reporting(kvm_vcpu_context_t vcpu, int enabled)
+{
+	int r;
+	struct kvm_tpr_access_ctl tac = {
+		.enabled = enabled,
+	};
+
+	r = ioctl(vcpu->kvm->fd, KVM_CHECK_EXTENSION, KVM_CAP_VAPIC);
+	if (r == -1 || r == 0)
+		return -ENOSYS;
+	r = ioctl(vcpu->fd, KVM_TPR_ACCESS_REPORTING, &tac);
+	if (r == -1) {
+		r = -errno;
+		perror("KVM_TPR_ACCESS_REPORTING");
+		return r;
+	}
+	return 0;
+}
+
+int kvm_enable_tpr_access_reporting(kvm_vcpu_context_t vcpu)
+{
+	return tpr_access_reporting(vcpu, 1);
+}
+
+int kvm_disable_tpr_access_reporting(kvm_vcpu_context_t vcpu)
+{
+	return tpr_access_reporting(vcpu, 0);
+}
+
+#endif
+
+#ifdef KVM_CAP_EXT_CPUID
+
+static struct kvm_cpuid2 *try_get_cpuid(kvm_context_t kvm, int max)
+{
+	struct kvm_cpuid2 *cpuid;
+	int r, size;
+
+	size = sizeof(*cpuid) + max * sizeof(*cpuid->entries);
+	cpuid = (struct kvm_cpuid2 *)malloc(size);
+	cpuid->nent = max;
+	r = ioctl(kvm->fd, KVM_GET_SUPPORTED_CPUID, cpuid);
+	if (r == -1)
+		r = -errno;
+	else if (r == 0 && cpuid->nent >= max)
+		r = -E2BIG;
+	if (r < 0) {
+		if (r == -E2BIG) {
+			free(cpuid);
+			return NULL;
+		} else {
+			fprintf(stderr, "KVM_GET_SUPPORTED_CPUID failed: %s\n",
+				strerror(-r));
+			exit(1);
+		}
+	}
+	return cpuid;
+}
+
+#define R_EAX 0
+#define R_ECX 1
+#define R_EDX 2
+#define R_EBX 3
+#define R_ESP 4
+#define R_EBP 5
+#define R_ESI 6
+#define R_EDI 7
+
+uint32_t kvm_get_supported_cpuid(kvm_context_t kvm, uint32_t function, int reg)
+{
+	struct kvm_cpuid2 *cpuid;
+	int i, max;
+	uint32_t ret = 0;
+	uint32_t cpuid_1_edx;
+
+	if (!kvm_check_extension(kvm, KVM_CAP_EXT_CPUID)) {
+		return -1U;
+	}
+
+	max = 1;
+	while ((cpuid = try_get_cpuid(kvm, max)) == NULL) {
+		max *= 2;
+	}
+
+	for (i = 0; i < cpuid->nent; ++i) {
+		if (cpuid->entries[i].function == function) {
+			switch (reg) {
+			case R_EAX:
+				ret = cpuid->entries[i].eax;
+				break;
+			case R_EBX:
+				ret = cpuid->entries[i].ebx;
+				break;
+			case R_ECX:
+				ret = cpuid->entries[i].ecx;
+				break;
+			case R_EDX:
+				ret = cpuid->entries[i].edx;
+                                if (function == 1) {
+                                    /* kvm misreports the following features
+                                     */
+                                    ret |= 1 << 12; /* MTRR */
+                                    ret |= 1 << 16; /* PAT */
+                                    ret |= 1 << 7;  /* MCE */
+                                    ret |= 1 << 14; /* MCA */
+                                }
+
+				/* On Intel, kvm returns cpuid according to
+				 * the Intel spec, so add missing bits
+				 * according to the AMD spec:
+				 */
+				if (function == 0x80000001) {
+					cpuid_1_edx = kvm_get_supported_cpuid(kvm, 1, R_EDX);
+					ret |= cpuid_1_edx & 0xdfeff7ff;
+				}
+				break;
+			}
+		}
+	}
+
+	free(cpuid);
+
+	return ret;
+}
+
+#else
+
+uint32_t kvm_get_supported_cpuid(kvm_context_t kvm, uint32_t function, int reg)
+{
+	return -1U;
+}
+
+#endif
+int kvm_qemu_create_memory_alias(uint64_t phys_start,
+                                 uint64_t len,
+                                 uint64_t target_phys)
+{
+    return kvm_create_memory_alias(kvm_context, phys_start, len, target_phys);
+}
+
+int kvm_qemu_destroy_memory_alias(uint64_t phys_start)
+{
+	return kvm_destroy_memory_alias(kvm_context, phys_start);
+}
+
+int kvm_arch_qemu_create_context(void)
+{
+    int i;
+    struct utsname utsname;
+
+    uname(&utsname);
+    lm_capable_kernel = strcmp(utsname.machine, "x86_64") == 0;
+
+    if (kvm_shadow_memory)
+        kvm_set_shadow_pages(kvm_context, kvm_shadow_memory);
+
+    kvm_msr_list = kvm_get_msr_list(kvm_context);
+    if (!kvm_msr_list)
+		return -1;
+    for (i = 0; i < kvm_msr_list->nmsrs; ++i) {
+	if (kvm_msr_list->indices[i] == MSR_STAR)
+	    kvm_has_msr_star = 1;
+        if (kvm_msr_list->indices[i] == MSR_VM_HSAVE_PA)
+            kvm_has_vm_hsave_pa = 1;
+    }
+
+    return 0;
+}
+
+static void set_msr_entry(struct kvm_msr_entry *entry, uint32_t index,
+                          uint64_t data)
+{
+    entry->index = index;
+    entry->data  = data;
+}
+
+/* returns 0 on success, non-0 on failure */
+static int get_msr_entry(struct kvm_msr_entry *entry, CPUState *env)
+{
+        switch (entry->index) {
+        case MSR_IA32_SYSENTER_CS:
+            env->sysenter_cs  = entry->data;
+            break;
+        case MSR_IA32_SYSENTER_ESP:
+            env->sysenter_esp = entry->data;
+            break;
+        case MSR_IA32_SYSENTER_EIP:
+            env->sysenter_eip = entry->data;
+            break;
+        case MSR_STAR:
+            env->star         = entry->data;
+            break;
+#ifdef TARGET_X86_64
+        case MSR_CSTAR:
+            env->cstar        = entry->data;
+            break;
+        case MSR_KERNELGSBASE:
+            env->kernelgsbase = entry->data;
+            break;
+        case MSR_FMASK:
+            env->fmask        = entry->data;
+            break;
+        case MSR_LSTAR:
+            env->lstar        = entry->data;
+            break;
+#endif
+        case MSR_IA32_TSC:
+            env->tsc          = entry->data;
+            break;
+        case MSR_VM_HSAVE_PA:
+            env->vm_hsave     = entry->data;
+            break;
+        default:
+            printf("Warning unknown msr index 0x%x\n", entry->index);
+            return 1;
+        }
+        return 0;
+}
+
+#ifdef TARGET_X86_64
+#define MSR_COUNT 9
+#else
+#define MSR_COUNT 5
+#endif
+
+static void set_v8086_seg(struct kvm_segment *lhs, const SegmentCache *rhs)
+{
+    lhs->selector = rhs->selector;
+    lhs->base = rhs->base;
+    lhs->limit = rhs->limit;
+    lhs->type = 3;
+    lhs->present = 1;
+    lhs->dpl = 3;
+    lhs->db = 0;
+    lhs->s = 1;
+    lhs->l = 0;
+    lhs->g = 0;
+    lhs->avl = 0;
+    lhs->unusable = 0;
+}
+
+static void set_seg(struct kvm_segment *lhs, const SegmentCache *rhs)
+{
+    unsigned flags = rhs->flags;
+    lhs->selector = rhs->selector;
+    lhs->base = rhs->base;
+    lhs->limit = rhs->limit;
+    lhs->type = (flags >> DESC_TYPE_SHIFT) & 15;
+    lhs->present = (flags & DESC_P_MASK) != 0;
+    lhs->dpl = rhs->selector & 3;
+    lhs->db = (flags >> DESC_B_SHIFT) & 1;
+    lhs->s = (flags & DESC_S_MASK) != 0;
+    lhs->l = (flags >> DESC_L_SHIFT) & 1;
+    lhs->g = (flags & DESC_G_MASK) != 0;
+    lhs->avl = (flags & DESC_AVL_MASK) != 0;
+    lhs->unusable = 0;
+}
+
+static void get_seg(SegmentCache *lhs, const struct kvm_segment *rhs)
+{
+    lhs->selector = rhs->selector;
+    lhs->base = rhs->base;
+    lhs->limit = rhs->limit;
+    lhs->flags =
+	(rhs->type << DESC_TYPE_SHIFT)
+	| (rhs->present * DESC_P_MASK)
+	| (rhs->dpl << DESC_DPL_SHIFT)
+	| (rhs->db << DESC_B_SHIFT)
+	| (rhs->s * DESC_S_MASK)
+	| (rhs->l << DESC_L_SHIFT)
+	| (rhs->g * DESC_G_MASK)
+	| (rhs->avl * DESC_AVL_MASK);
+}
+
+void kvm_arch_load_regs(CPUState *env)
+{
+    struct kvm_regs regs;
+    struct kvm_fpu fpu;
+    struct kvm_sregs sregs;
+    struct kvm_msr_entry msrs[MSR_COUNT];
+    int rc, n, i;
+
+    regs.rax = env->regs[R_EAX];
+    regs.rbx = env->regs[R_EBX];
+    regs.rcx = env->regs[R_ECX];
+    regs.rdx = env->regs[R_EDX];
+    regs.rsi = env->regs[R_ESI];
+    regs.rdi = env->regs[R_EDI];
+    regs.rsp = env->regs[R_ESP];
+    regs.rbp = env->regs[R_EBP];
+#ifdef TARGET_X86_64
+    regs.r8 = env->regs[8];
+    regs.r9 = env->regs[9];
+    regs.r10 = env->regs[10];
+    regs.r11 = env->regs[11];
+    regs.r12 = env->regs[12];
+    regs.r13 = env->regs[13];
+    regs.r14 = env->regs[14];
+    regs.r15 = env->regs[15];
+#endif
+
+    regs.rflags = env->eflags;
+    regs.rip = env->eip;
+
+    kvm_set_regs(env->kvm_cpu_state.vcpu_ctx, &regs);
+
+    memset(&fpu, 0, sizeof fpu);
+    fpu.fsw = env->fpus & ~(7 << 11);
+    fpu.fsw |= (env->fpstt & 7) << 11;
+    fpu.fcw = env->fpuc;
+    for (i = 0; i < 8; ++i)
+	fpu.ftwx |= (!env->fptags[i]) << i;
+    memcpy(fpu.fpr, env->fpregs, sizeof env->fpregs);
+    memcpy(fpu.xmm, env->xmm_regs, sizeof env->xmm_regs);
+    fpu.mxcsr = env->mxcsr;
+    kvm_set_fpu(env->kvm_cpu_state.vcpu_ctx, &fpu);
+
+    memcpy(sregs.interrupt_bitmap, env->interrupt_bitmap, sizeof(sregs.interrupt_bitmap));
+
+    if ((env->eflags & VM_MASK)) {
+	    set_v8086_seg(&sregs.cs, &env->segs[R_CS]);
+	    set_v8086_seg(&sregs.ds, &env->segs[R_DS]);
+	    set_v8086_seg(&sregs.es, &env->segs[R_ES]);
+	    set_v8086_seg(&sregs.fs, &env->segs[R_FS]);
+	    set_v8086_seg(&sregs.gs, &env->segs[R_GS]);
+	    set_v8086_seg(&sregs.ss, &env->segs[R_SS]);
+    } else {
+	    set_seg(&sregs.cs, &env->segs[R_CS]);
+	    set_seg(&sregs.ds, &env->segs[R_DS]);
+	    set_seg(&sregs.es, &env->segs[R_ES]);
+	    set_seg(&sregs.fs, &env->segs[R_FS]);
+	    set_seg(&sregs.gs, &env->segs[R_GS]);
+	    set_seg(&sregs.ss, &env->segs[R_SS]);
+
+	    if (env->cr[0] & CR0_PE_MASK) {
+		/* force ss cpl to cs cpl */
+		sregs.ss.selector = (sregs.ss.selector & ~3) |
+			(sregs.cs.selector & 3);
+		sregs.ss.dpl = sregs.ss.selector & 3;
+	    }
+    }
+
+    set_seg(&sregs.tr, &env->tr);
+    set_seg(&sregs.ldt, &env->ldt);
+
+    sregs.idt.limit = env->idt.limit;
+    sregs.idt.base = env->idt.base;
+    sregs.gdt.limit = env->gdt.limit;
+    sregs.gdt.base = env->gdt.base;
+
+    sregs.cr0 = env->cr[0];
+    sregs.cr2 = env->cr[2];
+    sregs.cr3 = env->cr[3];
+    sregs.cr4 = env->cr[4];
+
+    sregs.cr8 = cpu_get_apic_tpr(env);
+    sregs.apic_base = cpu_get_apic_base(env);
+
+    sregs.efer = env->efer;
+
+    kvm_set_sregs(env->kvm_cpu_state.vcpu_ctx, &sregs);
+
+    /* msrs */
+    n = 0;
+    set_msr_entry(&msrs[n++], MSR_IA32_SYSENTER_CS,  env->sysenter_cs);
+    set_msr_entry(&msrs[n++], MSR_IA32_SYSENTER_ESP, env->sysenter_esp);
+    set_msr_entry(&msrs[n++], MSR_IA32_SYSENTER_EIP, env->sysenter_eip);
+    if (kvm_has_msr_star)
+	set_msr_entry(&msrs[n++], MSR_STAR,              env->star);
+    if (kvm_has_vm_hsave_pa)
+        set_msr_entry(&msrs[n++], MSR_VM_HSAVE_PA, env->vm_hsave);
+#ifdef TARGET_X86_64
+    if (lm_capable_kernel) {
+        set_msr_entry(&msrs[n++], MSR_CSTAR,             env->cstar);
+        set_msr_entry(&msrs[n++], MSR_KERNELGSBASE,      env->kernelgsbase);
+        set_msr_entry(&msrs[n++], MSR_FMASK,             env->fmask);
+        set_msr_entry(&msrs[n++], MSR_LSTAR  ,           env->lstar);
+    }
+#endif
+
+    rc = kvm_set_msrs(env->kvm_cpu_state.vcpu_ctx, msrs, n);
+    if (rc == -1)
+        perror("kvm_set_msrs FAILED");
+}
+
+void kvm_load_tsc(CPUState *env)
+{
+    int rc;
+    struct kvm_msr_entry msr;
+
+    set_msr_entry(&msr, MSR_IA32_TSC, env->tsc);
+
+    rc = kvm_set_msrs(env->kvm_cpu_state.vcpu_ctx, &msr, 1);
+    if (rc == -1)
+        perror("kvm_set_tsc FAILED.\n");
+}
+
+void kvm_arch_save_mpstate(CPUState *env)
+{
+#ifdef KVM_CAP_MP_STATE
+    int r;
+    struct kvm_mp_state mp_state;
+
+    r = kvm_get_mpstate(env->kvm_cpu_state.vcpu_ctx, &mp_state);
+    if (r < 0)
+        env->mp_state = -1;
+    else
+        env->mp_state = mp_state.mp_state;
+#endif
+}
+
+void kvm_arch_load_mpstate(CPUState *env)
+{
+#ifdef KVM_CAP_MP_STATE
+    struct kvm_mp_state mp_state = { .mp_state = env->mp_state };
+
+    /*
+     * -1 indicates that the host did not support GET_MP_STATE ioctl,
+     *  so don't touch it.
+     */
+    if (env->mp_state != -1)
+        kvm_set_mpstate(env->kvm_cpu_state.vcpu_ctx, &mp_state);
+#endif
+}
+
+void kvm_arch_save_regs(CPUState *env)
+{
+    struct kvm_regs regs;
+    struct kvm_fpu fpu;
+    struct kvm_sregs sregs;
+    struct kvm_msr_entry msrs[MSR_COUNT];
+    uint32_t hflags;
+    uint32_t i, n, rc;
+
+    kvm_get_regs(env->kvm_cpu_state.vcpu_ctx, &regs);
+
+    env->regs[R_EAX] = regs.rax;
+    env->regs[R_EBX] = regs.rbx;
+    env->regs[R_ECX] = regs.rcx;
+    env->regs[R_EDX] = regs.rdx;
+    env->regs[R_ESI] = regs.rsi;
+    env->regs[R_EDI] = regs.rdi;
+    env->regs[R_ESP] = regs.rsp;
+    env->regs[R_EBP] = regs.rbp;
+#ifdef TARGET_X86_64
+    env->regs[8] = regs.r8;
+    env->regs[9] = regs.r9;
+    env->regs[10] = regs.r10;
+    env->regs[11] = regs.r11;
+    env->regs[12] = regs.r12;
+    env->regs[13] = regs.r13;
+    env->regs[14] = regs.r14;
+    env->regs[15] = regs.r15;
+#endif
+
+    env->eflags = regs.rflags;
+    env->eip = regs.rip;
+
+    kvm_get_fpu(env->kvm_cpu_state.vcpu_ctx, &fpu);
+    env->fpstt = (fpu.fsw >> 11) & 7;
+    env->fpus = fpu.fsw;
+    env->fpuc = fpu.fcw;
+    for (i = 0; i < 8; ++i)
+	env->fptags[i] = !((fpu.ftwx >> i) & 1);
+    memcpy(env->fpregs, fpu.fpr, sizeof env->fpregs);
+    memcpy(env->xmm_regs, fpu.xmm, sizeof env->xmm_regs);
+    env->mxcsr = fpu.mxcsr;
+
+    kvm_get_sregs(env->kvm_cpu_state.vcpu_ctx, &sregs);
+
+    memcpy(env->interrupt_bitmap, sregs.interrupt_bitmap, sizeof(env->interrupt_bitmap));
+
+    get_seg(&env->segs[R_CS], &sregs.cs);
+    get_seg(&env->segs[R_DS], &sregs.ds);
+    get_seg(&env->segs[R_ES], &sregs.es);
+    get_seg(&env->segs[R_FS], &sregs.fs);
+    get_seg(&env->segs[R_GS], &sregs.gs);
+    get_seg(&env->segs[R_SS], &sregs.ss);
+
+    get_seg(&env->tr, &sregs.tr);
+    get_seg(&env->ldt, &sregs.ldt);
+
+    env->idt.limit = sregs.idt.limit;
+    env->idt.base = sregs.idt.base;
+    env->gdt.limit = sregs.gdt.limit;
+    env->gdt.base = sregs.gdt.base;
+
+    env->cr[0] = sregs.cr0;
+    env->cr[2] = sregs.cr2;
+    env->cr[3] = sregs.cr3;
+    env->cr[4] = sregs.cr4;
+
+    cpu_set_apic_base(env, sregs.apic_base);
+
+    env->efer = sregs.efer;
+    //cpu_set_apic_tpr(env, sregs.cr8);
+
+#define HFLAG_COPY_MASK ~( \
+			HF_CPL_MASK | HF_PE_MASK | HF_MP_MASK | HF_EM_MASK | \
+			HF_TS_MASK | HF_TF_MASK | HF_VM_MASK | HF_IOPL_MASK | \
+			HF_OSFXSR_MASK | HF_LMA_MASK | HF_CS32_MASK | \
+			HF_SS32_MASK | HF_CS64_MASK | HF_ADDSEG_MASK)
+
+
+
+    hflags = (env->segs[R_CS].flags >> DESC_DPL_SHIFT) & HF_CPL_MASK;
+    hflags |= (env->cr[0] & CR0_PE_MASK) << (HF_PE_SHIFT - CR0_PE_SHIFT);
+    hflags |= (env->cr[0] << (HF_MP_SHIFT - CR0_MP_SHIFT)) &
+	    (HF_MP_MASK | HF_EM_MASK | HF_TS_MASK);
+    hflags |= (env->eflags & (HF_TF_MASK | HF_VM_MASK | HF_IOPL_MASK));
+    hflags |= (env->cr[4] & CR4_OSFXSR_MASK) <<
+	    (HF_OSFXSR_SHIFT - CR4_OSFXSR_SHIFT);
+
+    if (env->efer & MSR_EFER_LMA) {
+        hflags |= HF_LMA_MASK;
+    }
+
+    if ((hflags & HF_LMA_MASK) && (env->segs[R_CS].flags & DESC_L_MASK)) {
+        hflags |= HF_CS32_MASK | HF_SS32_MASK | HF_CS64_MASK;
+    } else {
+        hflags |= (env->segs[R_CS].flags & DESC_B_MASK) >>
+		(DESC_B_SHIFT - HF_CS32_SHIFT);
+        hflags |= (env->segs[R_SS].flags & DESC_B_MASK) >>
+		(DESC_B_SHIFT - HF_SS32_SHIFT);
+        if (!(env->cr[0] & CR0_PE_MASK) ||
+                   (env->eflags & VM_MASK) ||
+                   !(hflags & HF_CS32_MASK)) {
+                hflags |= HF_ADDSEG_MASK;
+            } else {
+                hflags |= ((env->segs[R_DS].base |
+                                env->segs[R_ES].base |
+                                env->segs[R_SS].base) != 0) <<
+                    HF_ADDSEG_SHIFT;
+            }
+    }
+    env->hflags = (env->hflags & HFLAG_COPY_MASK) | hflags;
+
+    /* msrs */
+    n = 0;
+    msrs[n++].index = MSR_IA32_SYSENTER_CS;
+    msrs[n++].index = MSR_IA32_SYSENTER_ESP;
+    msrs[n++].index = MSR_IA32_SYSENTER_EIP;
+    if (kvm_has_msr_star)
+	msrs[n++].index = MSR_STAR;
+    msrs[n++].index = MSR_IA32_TSC;
+    if (kvm_has_vm_hsave_pa)
+        msrs[n++].index = MSR_VM_HSAVE_PA;
+#ifdef TARGET_X86_64
+    if (lm_capable_kernel) {
+        msrs[n++].index = MSR_CSTAR;
+        msrs[n++].index = MSR_KERNELGSBASE;
+        msrs[n++].index = MSR_FMASK;
+        msrs[n++].index = MSR_LSTAR;
+    }
+#endif
+    rc = kvm_get_msrs(env->kvm_cpu_state.vcpu_ctx, msrs, n);
+    if (rc == -1) {
+        perror("kvm_get_msrs FAILED");
+    }
+    else {
+        n = rc; /* actual number of MSRs */
+        for (i=0 ; i<n; i++) {
+            if (get_msr_entry(&msrs[i], env))
+                return;
+        }
+    }
+}
+
+static void do_cpuid_ent(struct kvm_cpuid_entry2 *e, uint32_t function,
+                         uint32_t count, CPUState *env)
+{
+    env->regs[R_EAX] = function;
+    env->regs[R_ECX] = count;
+    qemu_kvm_cpuid_on_env(env);
+    e->function = function;
+    e->flags = 0;
+    e->index = 0;
+    e->eax = env->regs[R_EAX];
+    e->ebx = env->regs[R_EBX];
+    e->ecx = env->regs[R_ECX];
+    e->edx = env->regs[R_EDX];
+}
+
+struct kvm_para_features {
+	int cap;
+	int feature;
+} para_features[] = {
+#ifdef KVM_CAP_CLOCKSOURCE
+	{ KVM_CAP_CLOCKSOURCE, KVM_FEATURE_CLOCKSOURCE },
+#endif
+#ifdef KVM_CAP_NOP_IO_DELAY
+	{ KVM_CAP_NOP_IO_DELAY, KVM_FEATURE_NOP_IO_DELAY },
+#endif
+#ifdef KVM_CAP_PV_MMU
+	{ KVM_CAP_PV_MMU, KVM_FEATURE_MMU_OP },
+#endif
+#ifdef KVM_CAP_CR3_CACHE
+	{ KVM_CAP_CR3_CACHE, KVM_FEATURE_CR3_CACHE },
+#endif
+	{ -1, -1 }
+};
+
+static int get_para_features(kvm_context_t kvm_context)
+{
+	int i, features = 0;
+
+	for (i = 0; i < ARRAY_SIZE(para_features)-1; i++) {
+		if (kvm_check_extension(kvm_context, para_features[i].cap))
+			features |= (1 << para_features[i].feature);
+	}
+
+	return features;
+}
+
+static void kvm_trim_features(uint32_t *features, uint32_t supported)
+{
+    int i;
+    uint32_t mask;
+
+    for (i = 0; i < 32; ++i) {
+        mask = 1U << i;
+        if ((*features & mask) && !(supported & mask)) {
+            *features &= ~mask;
+        }
+    }
+}
+
+int kvm_arch_qemu_init_env(CPUState *cenv)
+{
+    struct kvm_cpuid_entry2 cpuid_ent[100];
+#ifdef KVM_CPUID_SIGNATURE
+    struct kvm_cpuid_entry2 *pv_ent;
+    uint32_t signature[3];
+#endif
+    int cpuid_nent = 0;
+    CPUState copy;
+    uint32_t i, j, limit;
+
+    qemu_kvm_load_lapic(cenv);
+
+
+#ifdef KVM_CPUID_SIGNATURE
+    /* Paravirtualization CPUIDs */
+    memcpy(signature, "KVMKVMKVM\0\0\0", 12);
+    pv_ent = &cpuid_ent[cpuid_nent++];
+    memset(pv_ent, 0, sizeof(*pv_ent));
+    pv_ent->function = KVM_CPUID_SIGNATURE;
+    pv_ent->eax = 0;
+    pv_ent->ebx = signature[0];
+    pv_ent->ecx = signature[1];
+    pv_ent->edx = signature[2];
+
+    pv_ent = &cpuid_ent[cpuid_nent++];
+    memset(pv_ent, 0, sizeof(*pv_ent));
+    pv_ent->function = KVM_CPUID_FEATURES;
+    pv_ent->eax = get_para_features(kvm_context);
+#endif
+
+    kvm_trim_features(&cenv->cpuid_features,
+                      kvm_arch_get_supported_cpuid(cenv, 1, R_EDX));
+
+    /* prevent the hypervisor bit from being cleared by the kernel */
+    i = cenv->cpuid_ext_features & CPUID_EXT_HYPERVISOR;
+    kvm_trim_features(&cenv->cpuid_ext_features,
+                      kvm_arch_get_supported_cpuid(cenv, 1, R_ECX));
+    cenv->cpuid_ext_features |= i;
+
+    kvm_trim_features(&cenv->cpuid_ext2_features,
+                      kvm_arch_get_supported_cpuid(cenv, 0x80000001, R_EDX));
+    kvm_trim_features(&cenv->cpuid_ext3_features,
+                      kvm_arch_get_supported_cpuid(cenv, 0x80000001, R_ECX));
+
+    copy = *cenv;
+
+    copy.regs[R_EAX] = 0;
+    qemu_kvm_cpuid_on_env(&copy);
+    limit = copy.regs[R_EAX];
+
+    for (i = 0; i <= limit; ++i) {
+        if (i == 4 || i == 0xb || i == 0xd) {
+            for (j = 0; ; ++j) {
+                do_cpuid_ent(&cpuid_ent[cpuid_nent], i, j, &copy);
+
+                cpuid_ent[cpuid_nent].flags = KVM_CPUID_FLAG_SIGNIFCANT_INDEX;
+                cpuid_ent[cpuid_nent].index = j;
+
+                cpuid_nent++;
+
+                if (i == 4 && copy.regs[R_EAX] == 0)
+                    break;
+                if (i == 0xb && !(copy.regs[R_ECX] & 0xff00))
+                    break;
+                if (i == 0xd && copy.regs[R_EAX] == 0)
+                    break;
+            }
+        } else
+            do_cpuid_ent(&cpuid_ent[cpuid_nent++], i, 0, &copy);
+    }
+
+    copy.regs[R_EAX] = 0x80000000;
+    qemu_kvm_cpuid_on_env(&copy);
+    limit = copy.regs[R_EAX];
+
+    for (i = 0x80000000; i <= limit; ++i)
+	do_cpuid_ent(&cpuid_ent[cpuid_nent++], i, 0, &copy);
+
+    kvm_setup_cpuid2(cenv->kvm_cpu_state.vcpu_ctx, cpuid_nent, cpuid_ent);
+
+    return 0;
+}
+
+int kvm_arch_halt(void *opaque, kvm_vcpu_context_t vcpu)
+{
+    CPUState *env = cpu_single_env;
+
+    if (!((env->interrupt_request & CPU_INTERRUPT_HARD) &&
+	  (env->eflags & IF_MASK)) &&
+	!(env->interrupt_request & CPU_INTERRUPT_NMI)) {
+            env->halted = 1;
+    }
+    return 1;
+}
+
+void kvm_arch_pre_kvm_run(void *opaque, CPUState *env)
+{
+    if (!kvm_irqchip_in_kernel(kvm_context))
+	kvm_set_cr8(env->kvm_cpu_state.vcpu_ctx, cpu_get_apic_tpr(env));
+}
+
+void kvm_arch_post_kvm_run(void *opaque, CPUState *env)
+{
+    cpu_single_env = env;
+
+    env->eflags = kvm_get_interrupt_flag(env->kvm_cpu_state.vcpu_ctx)
+	? env->eflags | IF_MASK : env->eflags & ~IF_MASK;
+
+    cpu_set_apic_tpr(env, kvm_get_cr8(env->kvm_cpu_state.vcpu_ctx));
+    cpu_set_apic_base(env, kvm_get_apic_base(env->kvm_cpu_state.vcpu_ctx));
+}
+
+int kvm_arch_has_work(CPUState *env)
+{
+    if (((env->interrupt_request & CPU_INTERRUPT_HARD) &&
+	 (env->eflags & IF_MASK)) ||
+	(env->interrupt_request & CPU_INTERRUPT_NMI))
+	return 1;
+    return 0;
+}
+
+int kvm_arch_try_push_interrupts(void *opaque)
+{
+    CPUState *env = cpu_single_env;
+    int r, irq;
+
+    if (kvm_is_ready_for_interrupt_injection(env->kvm_cpu_state.vcpu_ctx) &&
+        (env->interrupt_request & CPU_INTERRUPT_HARD) &&
+        (env->eflags & IF_MASK)) {
+            env->interrupt_request &= ~CPU_INTERRUPT_HARD;
+	    irq = cpu_get_pic_interrupt(env);
+	    if (irq >= 0) {
+		r = kvm_inject_irq(env->kvm_cpu_state.vcpu_ctx, irq);
+		if (r < 0)
+		    printf("cpu %d fail inject %x\n", env->cpu_index, irq);
+	    }
+    }
+
+    return (env->interrupt_request & CPU_INTERRUPT_HARD) != 0;
+}
+
+#ifdef KVM_CAP_USER_NMI
+void kvm_arch_push_nmi(void *opaque)
+{
+    CPUState *env = cpu_single_env;
+    int r;
+
+    if (likely(!(env->interrupt_request & CPU_INTERRUPT_NMI)))
+        return;
+
+    env->interrupt_request &= ~CPU_INTERRUPT_NMI;
+    r = kvm_inject_nmi(env->kvm_cpu_state.vcpu_ctx);
+    if (r < 0)
+        printf("cpu %d fail inject NMI\n", env->cpu_index);
+}
+#endif /* KVM_CAP_USER_NMI */
+
+void kvm_arch_update_regs_for_sipi(CPUState *env)
+{
+    SegmentCache cs = env->segs[R_CS];
+
+    kvm_arch_save_regs(env);
+    env->segs[R_CS] = cs;
+    env->eip = 0;
+    kvm_arch_load_regs(env);
+}
+
+void kvm_arch_cpu_reset(CPUState *env)
+{
+    kvm_arch_load_regs(env);
+    if (!cpu_is_bsp(env)) {
+	if (kvm_irqchip_in_kernel(kvm_context)) {
+#ifdef KVM_CAP_MP_STATE
+	    kvm_reset_mpstate(env->kvm_cpu_state.vcpu_ctx);
+#endif
+	} else {
+	    env->interrupt_request &= ~CPU_INTERRUPT_HARD;
+	    env->halted = 1;
+	}
+    }
+}
+
+int kvm_arch_insert_sw_breakpoint(CPUState *env, struct kvm_sw_breakpoint *bp)
+{
+    uint8_t int3 = 0xcc;
+
+    if (cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&bp->saved_insn, 1, 0) ||
+        cpu_memory_rw_debug(env, bp->pc, &int3, 1, 1))
+        return -EINVAL;
+    return 0;
+}
+
+int kvm_arch_remove_sw_breakpoint(CPUState *env, struct kvm_sw_breakpoint *bp)
+{
+    uint8_t int3;
+
+    if (cpu_memory_rw_debug(env, bp->pc, &int3, 1, 0) || int3 != 0xcc ||
+        cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&bp->saved_insn, 1, 1))
+        return -EINVAL;
+    return 0;
+}
+
+#ifdef KVM_CAP_SET_GUEST_DEBUG
+static struct {
+    target_ulong addr;
+    int len;
+    int type;
+} hw_breakpoint[4];
+
+static int nb_hw_breakpoint;
+
+static int find_hw_breakpoint(target_ulong addr, int len, int type)
+{
+    int n;
+
+    for (n = 0; n < nb_hw_breakpoint; n++)
+	if (hw_breakpoint[n].addr == addr && hw_breakpoint[n].type == type &&
+	    (hw_breakpoint[n].len == len || len == -1))
+	    return n;
+    return -1;
+}
+
+int kvm_arch_insert_hw_breakpoint(target_ulong addr,
+                                  target_ulong len, int type)
+{
+    switch (type) {
+    case GDB_BREAKPOINT_HW:
+	len = 1;
+	break;
+    case GDB_WATCHPOINT_WRITE:
+    case GDB_WATCHPOINT_ACCESS:
+	switch (len) {
+	case 1:
+	    break;
+	case 2:
+	case 4:
+	case 8:
+	    if (addr & (len - 1))
+		return -EINVAL;
+	    break;
+	default:
+	    return -EINVAL;
+	}
+	break;
+    default:
+	return -ENOSYS;
+    }
+
+    if (nb_hw_breakpoint == 4)
+        return -ENOBUFS;
+
+    if (find_hw_breakpoint(addr, len, type) >= 0)
+        return -EEXIST;
+
+    hw_breakpoint[nb_hw_breakpoint].addr = addr;
+    hw_breakpoint[nb_hw_breakpoint].len = len;
+    hw_breakpoint[nb_hw_breakpoint].type = type;
+    nb_hw_breakpoint++;
+
+    return 0;
+}
+
+int kvm_arch_remove_hw_breakpoint(target_ulong addr,
+                                  target_ulong len, int type)
+{
+    int n;
+
+    n = find_hw_breakpoint(addr, (type == GDB_BREAKPOINT_HW) ? 1 : len, type);
+    if (n < 0)
+        return -ENOENT;
+
+    nb_hw_breakpoint--;
+    hw_breakpoint[n] = hw_breakpoint[nb_hw_breakpoint];
+
+    return 0;
+}
+
+void kvm_arch_remove_all_hw_breakpoints(void)
+{
+    nb_hw_breakpoint = 0;
+}
+
+static CPUWatchpoint hw_watchpoint;
+
+int kvm_arch_debug(struct kvm_debug_exit_arch *arch_info)
+{
+    int handle = 0;
+    int n;
+
+    if (arch_info->exception == 1) {
+	if (arch_info->dr6 & (1 << 14)) {
+	    if (cpu_single_env->singlestep_enabled)
+		handle = 1;
+	} else {
+	    for (n = 0; n < 4; n++)
+		if (arch_info->dr6 & (1 << n))
+		    switch ((arch_info->dr7 >> (16 + n*4)) & 0x3) {
+		    case 0x0:
+			handle = 1;
+			break;
+		    case 0x1:
+			handle = 1;
+			cpu_single_env->watchpoint_hit = &hw_watchpoint;
+			hw_watchpoint.vaddr = hw_breakpoint[n].addr;
+			hw_watchpoint.flags = BP_MEM_WRITE;
+			break;
+		    case 0x3:
+			handle = 1;
+			cpu_single_env->watchpoint_hit = &hw_watchpoint;
+			hw_watchpoint.vaddr = hw_breakpoint[n].addr;
+			hw_watchpoint.flags = BP_MEM_ACCESS;
+			break;
+		    }
+	}
+    } else if (kvm_find_sw_breakpoint(arch_info->pc))
+	handle = 1;
+
+    if (!handle)
+	kvm_update_guest_debug(cpu_single_env,
+			(arch_info->exception == 1) ?
+			KVM_GUESTDBG_INJECT_DB : KVM_GUESTDBG_INJECT_BP);
+
+    return handle;
+}
+
+void kvm_arch_update_guest_debug(CPUState *env, struct kvm_guest_debug *dbg)
+{
+    const uint8_t type_code[] = {
+	[GDB_BREAKPOINT_HW] = 0x0,
+	[GDB_WATCHPOINT_WRITE] = 0x1,
+	[GDB_WATCHPOINT_ACCESS] = 0x3
+    };
+    const uint8_t len_code[] = {
+	[1] = 0x0, [2] = 0x1, [4] = 0x3, [8] = 0x2
+    };
+    int n;
+
+    if (!TAILQ_EMPTY(&kvm_sw_breakpoints))
+	dbg->control |= KVM_GUESTDBG_ENABLE | KVM_GUESTDBG_USE_SW_BP;
+
+    if (nb_hw_breakpoint > 0) {
+	dbg->control |= KVM_GUESTDBG_ENABLE | KVM_GUESTDBG_USE_HW_BP;
+	dbg->arch.debugreg[7] = 0x0600;
+	for (n = 0; n < nb_hw_breakpoint; n++) {
+	    dbg->arch.debugreg[n] = hw_breakpoint[n].addr;
+	    dbg->arch.debugreg[7] |= (2 << (n * 2)) |
+		(type_code[hw_breakpoint[n].type] << (16 + n*4)) |
+		(len_code[hw_breakpoint[n].len] << (18 + n*4));
+	}
+    }
+}
+#endif
+
+void kvm_arch_do_ioperm(void *_data)
+{
+    struct ioperm_data *data = _data;
+    ioperm(data->start_port, data->num, data->turn_on);
+}
+
+/*
+ * Setup x86 specific IRQ routing
+ */
+int kvm_arch_init_irq_routing(void)
+{
+    int i, r;
+
+    if (kvm_irqchip && kvm_has_gsi_routing(kvm_context)) {
+        kvm_clear_gsi_routes(kvm_context);
+        for (i = 0; i < 8; ++i) {
+            if (i == 2)
+                continue;
+            r = kvm_add_irq_route(kvm_context, i, KVM_IRQCHIP_PIC_MASTER, i);
+            if (r < 0)
+                return r;
+        }
+        for (i = 8; i < 16; ++i) {
+            r = kvm_add_irq_route(kvm_context, i, KVM_IRQCHIP_PIC_SLAVE, i - 8);
+            if (r < 0)
+                return r;
+        }
+        for (i = 0; i < 24; ++i) {
+            r = kvm_add_irq_route(kvm_context, i, KVM_IRQCHIP_IOAPIC, i);
+            if (r < 0)
+                return r;
+        }
+        kvm_commit_irq_routes(kvm_context);
+    }
+    return 0;
+}
+
+uint32_t kvm_arch_get_supported_cpuid(CPUState *env, uint32_t function,
+                                      int reg)
+{
+    return kvm_get_supported_cpuid(kvm_context, function, reg);
+}
+
+void kvm_arch_process_irqchip_events(CPUState *env)
+{
+    kvm_arch_save_regs(env);
+    if (env->interrupt_request & CPU_INTERRUPT_INIT)
+        do_cpu_init(env);
+    if (env->interrupt_request & CPU_INTERRUPT_SIPI)
+        do_cpu_sipi(env);
+    kvm_arch_load_regs(env);
+}
+
+#endif /* USE_KVM */
-- 
1.6.2.2

[PATCH 4/8] replace malloc with qemu_malloc

[Glauber Costa]

This patch replaces both malloc and malloc+memset sequences
with qemu_malloc and qemu_mallocz. Target is upstream integration

Signed-off-by: Glauber Costa <glommer@redhat.com>
---
 kvm-all.c         |   26 +++++---------------------
 target-i386/kvm.c |   31 ++++++++-----------------------
 2 files changed, 13 insertions(+), 44 deletions(-)

diff --git a/kvm-all.c b/kvm-all.c
index fd68f4c..034ae52 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -1469,10 +1469,7 @@ kvm_context_t kvm_init(void *opaque)
 	}
 	kvm_abi = r;
 	kvm_page_size = getpagesize();
-	kvm = malloc(sizeof(*kvm));
-	if (kvm == NULL)
-		goto out_close;
-	memset(kvm, 0, sizeof(*kvm));
+	kvm = qemu_mallocz(sizeof(*kvm));
 	kvm->fd = fd;
 	kvm->vm_fd = -1;
 	kvm->opaque = opaque;
@@ -1486,10 +1483,7 @@ kvm_context_t kvm_init(void *opaque)
 
 		/* Round up so we can search ints using ffs */
 		gsi_bits = ALIGN(gsi_count, 32);
-		kvm->used_gsi_bitmap = malloc(gsi_bits / 8);
-		if (!kvm->used_gsi_bitmap)
-			goto out_close;
-		memset(kvm->used_gsi_bitmap, 0, gsi_bits / 8);
+		kvm->used_gsi_bitmap = qemu_mallocz(gsi_bits / 8);
 		kvm->max_gsi = gsi_bits;
 
 		/* Mark any over-allocated bits as already in use */
@@ -1529,12 +1523,7 @@ kvm_vcpu_context_t kvm_create_vcpu(kvm_context_t kvm, int id)
 {
 	long mmap_size;
 	int r;
-	kvm_vcpu_context_t vcpu_ctx = malloc(sizeof(struct kvm_vcpu_context));
-
-	if (!vcpu_ctx) {
-		errno = ENOMEM;
-		return NULL;
-	}
+	kvm_vcpu_context_t vcpu_ctx = qemu_malloc(sizeof(struct kvm_vcpu_context));
 
 	vcpu_ctx->kvm = kvm;
 	vcpu_ctx->id = id;
@@ -1581,10 +1570,7 @@ int kvm_create_vm(kvm_context_t kvm)
 	int fd = kvm->fd;
 
 #ifdef KVM_CAP_IRQ_ROUTING
-	kvm->irq_routes = malloc(sizeof(*kvm->irq_routes));
-	if (!kvm->irq_routes)
-		return -ENOMEM;
-	memset(kvm->irq_routes, 0, sizeof(*kvm->irq_routes));
+	kvm->irq_routes = qemu_mallocz(sizeof(*kvm->irq_routes));
 	kvm->nr_allocated_irq_routes = 0;
 #endif
 
@@ -2189,9 +2175,7 @@ int kvm_set_signal_mask(kvm_vcpu_context_t vcpu, const sigset_t *sigset)
 			r = -errno;
 		return r;
 	}
-	sigmask = malloc(sizeof(*sigmask) + sizeof(*sigset));
-	if (!sigmask)
-		return -ENOMEM;
+	sigmask = qemu_malloc(sizeof(*sigmask) + sizeof(*sigset));
 
 	sigmask->len = 8;
 	memcpy(sigmask->sigset, sigset, sizeof(*sigset));
diff --git a/target-i386/kvm.c b/target-i386/kvm.c
index 4c3948c..ab324f6 100644
--- a/target-i386/kvm.c
+++ b/target-i386/kvm.c
@@ -1303,12 +1303,9 @@ struct kvm_msr_list *kvm_get_msr_list(kvm_context_t kvm)
 		return NULL;
 	/* Old kernel modules had a bug and could write beyond the provided
 	   memory. Allocate at least a safe amount of 1K. */
-	msrs = malloc(MAX(1024, sizeof(*msrs) +
-				sizer.nmsrs * sizeof(*msrs->indices)));
-	if (!msrs) {
-		errno = ENOMEM;
-		return NULL;
-	}
+	msrs = qemu_malloc(MAX(1024, sizeof(*msrs) +
+				       sizer.nmsrs * sizeof(*msrs->indices)));
+
 	msrs->nmsrs = sizer.nmsrs;
 	r = ioctl(kvm->fd, KVM_GET_MSR_INDEX_LIST, msrs);
 	if (r == -1) {
@@ -1322,13 +1319,9 @@ struct kvm_msr_list *kvm_get_msr_list(kvm_context_t kvm)
 
 int kvm_get_msrs(kvm_vcpu_context_t vcpu, struct kvm_msr_entry *msrs, int n)
 {
-    struct kvm_msrs *kmsrs = malloc(sizeof *kmsrs + n * sizeof *msrs);
+    struct kvm_msrs *kmsrs = qemu_malloc(sizeof *kmsrs + n * sizeof *msrs);
     int r, e;
 
-    if (!kmsrs) {
-	errno = ENOMEM;
-	return -1;
-    }
     kmsrs->nmsrs = n;
     memcpy(kmsrs->entries, msrs, n * sizeof *msrs);
     r = ioctl(vcpu->fd, KVM_GET_MSRS, kmsrs);
@@ -1341,13 +1334,9 @@ int kvm_get_msrs(kvm_vcpu_context_t vcpu, struct kvm_msr_entry *msrs, int n)
 
 int kvm_set_msrs(kvm_vcpu_context_t vcpu, struct kvm_msr_entry *msrs, int n)
 {
-    struct kvm_msrs *kmsrs = malloc(sizeof *kmsrs + n * sizeof *msrs);
+    struct kvm_msrs *kmsrs = qemu_malloc(sizeof *kmsrs + n * sizeof *msrs);
     int r, e;
 
-    if (!kmsrs) {
-	errno = ENOMEM;
-	return -1;
-    }
     kmsrs->nmsrs = n;
     memcpy(kmsrs->entries, msrs, n * sizeof *msrs);
     r = ioctl(vcpu->fd, KVM_SET_MSRS, kmsrs);
@@ -1437,9 +1426,7 @@ int kvm_setup_cpuid(kvm_vcpu_context_t vcpu, int nent,
 	struct kvm_cpuid *cpuid;
 	int r;
 
-	cpuid = malloc(sizeof(*cpuid) + nent * sizeof(*entries));
-	if (!cpuid)
-		return -ENOMEM;
+	cpuid = qemu_malloc(sizeof(*cpuid) + nent * sizeof(*entries));
 
 	cpuid->nent = nent;
 	memcpy(cpuid->entries, entries, nent * sizeof(*entries));
@@ -1455,9 +1442,7 @@ int kvm_setup_cpuid2(kvm_vcpu_context_t vcpu, int nent,
 	struct kvm_cpuid2 *cpuid;
 	int r;
 
-	cpuid = malloc(sizeof(*cpuid) + nent * sizeof(*entries));
-	if (!cpuid)
-		return -ENOMEM;
+	cpuid = qemu_malloc(sizeof(*cpuid) + nent * sizeof(*entries));
 
 	cpuid->nent = nent;
 	memcpy(cpuid->entries, entries, nent * sizeof(*entries));
@@ -1545,7 +1530,7 @@ static struct kvm_cpuid2 *try_get_cpuid(kvm_context_t kvm, int max)
 	int r, size;
 
 	size = sizeof(*cpuid) + max * sizeof(*cpuid->entries);
-	cpuid = (struct kvm_cpuid2 *)malloc(size);
+	cpuid = qemu_malloc(size);
 	cpuid->nent = max;
 	r = ioctl(kvm->fd, KVM_GET_SUPPORTED_CPUID, cpuid);
 	if (r == -1)
-- 
1.6.2.2

[PATCH 5/8] move kvm_context inside KVMState

[Glauber Costa]

To make transition smooth, we still keep a global variable kvm_context
pointing to its position inside the global KVMState. This way we don't
need to hurry about changing all callers.

kvm_init() and kvm_finalize are changed, though, since they have now to
deal with the creation/destruction of a global KVMState

Signed-off-by: Glauber Costa <glommer@redhat.com>
---
 kvm-all.c    |   98 +++++++++++++++++++++++++--------------------------------
 kvm.h        |    3 +-
 libkvm-all.h |    3 +-
 3 files changed, 47 insertions(+), 57 deletions(-)

diff --git a/kvm-all.c b/kvm-all.c
index 034ae52..15bd429 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -35,23 +35,6 @@
 #include "libkvm-all.h"
 #include "libkvm.h"
 
-
-#ifdef CONFIG_KVM
-
-
-/* KVM uses PAGE_SIZE in it's definition of COALESCED_MMIO_MAX */
-#define PAGE_SIZE TARGET_PAGE_SIZE
-
-//#define DEBUG_KVM
-
-#ifdef DEBUG_KVM
-#define dprintf(fmt, ...) \
-    do { fprintf(stderr, fmt, ## __VA_ARGS__); } while (0)
-#else
-#define dprintf(fmt, ...) \
-    do { } while (0)
-#endif
-
 typedef struct KVMSlot
 {
     target_phys_addr_t start_addr;
@@ -63,8 +46,6 @@ typedef struct KVMSlot
 
 typedef struct kvm_dirty_log KVMDirtyLog;
 
-int kvm_allowed = 0;
-
 struct KVMState
 {
     KVMSlot slots[32];
@@ -76,9 +57,28 @@ struct KVMState
 #ifdef KVM_CAP_SET_GUEST_DEBUG
     struct kvm_sw_breakpoint_head kvm_sw_breakpoints;
 #endif
+    struct kvm_context kvm_context;
 };
 
 static KVMState *kvm_state;
+kvm_context_t kvm_context;
+
+#ifdef CONFIG_KVM
+
+/* KVM uses PAGE_SIZE in it's definition of COALESCED_MMIO_MAX */
+#define PAGE_SIZE TARGET_PAGE_SIZE
+
+//#define DEBUG_KVM
+
+#ifdef DEBUG_KVM
+#define dprintf(fmt, ...) \
+    do { fprintf(stderr, fmt, ## __VA_ARGS__); } while (0)
+#else
+#define dprintf(fmt, ...) \
+    do { } while (0)
+#endif
+
+int kvm_allowed = 0;
 
 static KVMSlot *kvm_alloc_slot(KVMState *s)
 {
@@ -1065,7 +1065,6 @@ int kvm_irqchip = 1;
 int kvm_pit = 1;
 int kvm_pit_reinject = 1;
 int kvm_nested = 0;
-kvm_context_t kvm_context;
 
 pthread_mutex_t qemu_mutex = PTHREAD_MUTEX_INITIALIZER;
 pthread_cond_t qemu_vcpu_cond = PTHREAD_COND_INITIALIZER;
@@ -1439,16 +1438,15 @@ int kvm_dirty_pages_log_reset(kvm_context_t kvm)
 }
 
 
-kvm_context_t kvm_init(void *opaque)
+int kvm_qemu_init()
 {
 	int fd;
-	kvm_context_t kvm;
 	int r, gsi_count;
 
 	fd = open("/dev/kvm", O_RDWR);
 	if (fd == -1) {
 		perror("open /dev/kvm");
-		return NULL;
+		return -1;
 	}
 	r = ioctl(fd, KVM_GET_API_VERSION, 0);
 	if (r == -1) {
@@ -1469,35 +1467,37 @@ kvm_context_t kvm_init(void *opaque)
 	}
 	kvm_abi = r;
 	kvm_page_size = getpagesize();
-	kvm = qemu_mallocz(sizeof(*kvm));
-	kvm->fd = fd;
-	kvm->vm_fd = -1;
-	kvm->opaque = opaque;
-	kvm->dirty_pages_log_all = 0;
-	kvm->no_irqchip_creation = 0;
-	kvm->no_pit_creation = 0;
-
-	gsi_count = kvm_get_gsi_count(kvm);
+	kvm_state = qemu_mallocz(sizeof(*kvm_state));
+	kvm_context = &kvm_state->kvm_context;
+	kvm_context->fd = fd;
+	kvm_context->vm_fd = -1;
+	kvm_context->opaque = cpu_single_env;
+	kvm_context->dirty_pages_log_all = 0;
+	kvm_context->no_irqchip_creation = 0;
+	kvm_context->no_pit_creation = 0;
+
+	gsi_count = kvm_get_gsi_count(kvm_context);
 	if (gsi_count > 0) {
 		int gsi_bits, i;
 
 		/* Round up so we can search ints using ffs */
 		gsi_bits = ALIGN(gsi_count, 32);
-		kvm->used_gsi_bitmap = qemu_mallocz(gsi_bits / 8);
-		kvm->max_gsi = gsi_bits;
+		kvm_context->used_gsi_bitmap = qemu_mallocz(gsi_bits / 8);
+		kvm_context->max_gsi = gsi_bits;
 
 		/* Mark any over-allocated bits as already in use */
 		for (i = gsi_count; i < gsi_bits; i++)
-			set_gsi(kvm, i);
+			set_gsi(kvm_context, i);
 	}
 
-	return kvm;
- out_close:
+    pthread_mutex_lock(&qemu_mutex);
+	return 0;
+out_close:
 	close(fd);
-	return NULL;
+	return -1;
 }
 
-void kvm_finalize(kvm_context_t kvm)
+void kvm_finalize(KVMState *kvm_state)
 {
 	/* FIXME
 	if (kvm->vcpu_fd[0] != -1)
@@ -1505,8 +1505,8 @@ void kvm_finalize(kvm_context_t kvm)
 	if (kvm->vm_fd != -1)
 		close(kvm->vm_fd);
 	*/
-	close(kvm->fd);
-	free(kvm);
+	close(kvm_state->kvm_context.fd);
+	free(kvm_state);
 }
 
 void kvm_disable_irqchip_creation(kvm_context_t kvm)
@@ -3220,18 +3220,6 @@ int kvm_main_loop(void)
     return 0;
 }
 
-int kvm_qemu_init()
-{
-    /* Try to initialize kvm */
-    kvm_context = kvm_init(cpu_single_env);
-    if (!kvm_context) {
-      	return -1;
-    }
-    pthread_mutex_lock(&qemu_mutex);
-
-    return 0;
-}
-
 #ifdef TARGET_I386
 static int destroy_region_works = 0;
 #endif
@@ -3255,12 +3243,12 @@ int kvm_qemu_create_context(void)
         kvm_disable_pit_creation(kvm_context);
     }
     if (kvm_create(kvm_context, 0, NULL) < 0) {
-       kvm_finalize(kvm_context);
+       kvm_finalize(kvm_state);
        return -1;
     }
     r = kvm_arch_qemu_create_context();
     if(r <0)
-       kvm_finalize(kvm_context);
+       kvm_finalize(kvm_state);
     if (kvm_pit && !kvm_pit_reinject) {
         if (kvm_reinject_control(kvm_context, 0)) {
             fprintf(stderr, "failure to disable in-kernel PIT reinjection\n");
diff --git a/kvm.h b/kvm.h
index 7648c49..31ebce2 100644
--- a/kvm.h
+++ b/kvm.h
@@ -16,11 +16,12 @@
 
 #include "config.h"
 #include "sys-queue.h"
-#include "libkvm-all.h"
 
 struct KVMState;
 typedef struct KVMState KVMState;
 
+#include "libkvm-all.h"
+
 #ifdef KVM_UPSTREAM
 
 #ifdef CONFIG_KVM
diff --git a/libkvm-all.h b/libkvm-all.h
index e16646c..7857532 100644
--- a/libkvm-all.h
+++ b/libkvm-all.h
@@ -21,6 +21,7 @@
 
 #include <signal.h>
 
+
 /* FIXME: share this number with kvm */
 /* FIXME: or dynamically alloc/realloc regions */
 #ifdef __s390__
@@ -189,7 +190,7 @@ kvm_context_t kvm_init(void *opaque);
  *
  * \param kvm Pointer to the kvm_context that is to be freed
  */
-void kvm_finalize(kvm_context_t kvm);
+void kvm_finalize(KVMState *s);
 
 /*!
  * \brief Disable the in-kernel IRQCHIP creation
-- 
1.6.2.2

[PATCH 6/8] provide env->kvm_fd

[Glauber Costa]

qemu upstream puts kvm information on env. Do that too, since it will
allow us to use CPUState in cpu-specific functions, instead of kvm-specific
types.

Signed-off-by: Glauber Costa <glommer@redhat.com>
---
 kvm-all.c    |   12 ++++++++----
 libkvm-all.h |   12 ------------
 2 files changed, 8 insertions(+), 16 deletions(-)

diff --git a/kvm-all.c b/kvm-all.c
index 15bd429..b404f76 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -1519,21 +1519,25 @@ void kvm_disable_pit_creation(kvm_context_t kvm)
 	kvm->no_pit_creation = 1;
 }
 
-kvm_vcpu_context_t kvm_create_vcpu(kvm_context_t kvm, int id)
+static kvm_vcpu_context_t kvm_create_vcpu(kvm_context_t kvm, CPUState *env)
 {
 	long mmap_size;
 	int r;
 	kvm_vcpu_context_t vcpu_ctx = qemu_malloc(sizeof(struct kvm_vcpu_context));
 
 	vcpu_ctx->kvm = kvm;
-	vcpu_ctx->id = id;
+	vcpu_ctx->id = env->cpu_index;
 
-	r = ioctl(kvm->vm_fd, KVM_CREATE_VCPU, id);
+	r = ioctl(kvm->vm_fd, KVM_CREATE_VCPU, env->cpu_index);
 	if (r == -1) {
 		fprintf(stderr, "kvm_create_vcpu: %m\n");
 		goto err;
 	}
 	vcpu_ctx->fd = r;
+
+    env->kvm_fd = r;
+    env->kvm_state = kvm_state;
+
 	mmap_size = ioctl(kvm->fd, KVM_GET_VCPU_MMAP_SIZE, 0);
 	if (mmap_size == -1) {
 		fprintf(stderr, "get vcpu mmap size: %m\n");
@@ -3029,7 +3033,7 @@ static void *ap_main_loop(void *_env)
     env->thread_id = kvm_get_thread_id();
     sigfillset(&signals);
     sigprocmask(SIG_BLOCK, &signals, NULL);
-    env->kvm_cpu_state.vcpu_ctx = kvm_create_vcpu(kvm_context, env->cpu_index);
+    env->kvm_cpu_state.vcpu_ctx = kvm_create_vcpu(kvm_context, env);
 
 #ifdef USE_KVM_DEVICE_ASSIGNMENT
     /* do ioperm for io ports of assigned devices */
diff --git a/libkvm-all.h b/libkvm-all.h
index 7857532..9c018eb 100644
--- a/libkvm-all.h
+++ b/libkvm-all.h
@@ -234,18 +234,6 @@ int kvm_check_extension(kvm_context_t kvm, int ext);
 void kvm_create_irqchip(kvm_context_t kvm);
 
 /*!
- * \brief Create a new virtual cpu
- *
- * This creates a new virtual cpu (the first vcpu is created by kvm_create()).
- * Should be called from a thread dedicated to the vcpu.
- *
- * \param kvm kvm context
- * \param slot vcpu number (> 0)
- * \return 0 on success, -errno on failure
- */
-kvm_vcpu_context_t kvm_create_vcpu(kvm_context_t kvm, int id);
-
-/*!
  * \brief Start the VCPU
  *
  * This starts the VCPU and virtualization is started.\n
-- 
1.6.2.2

[PATCH 7/8] use kvm_upstream sw_breakpoints structure

[Glauber Costa]

Signed-off-by: Glauber Costa <glommer@redhat.com>
---
 kvm-all.c         |   27 ++++++++++++++++++---------
 qemu-kvm.h        |    6 +++---
 target-i386/kvm.c |    4 ++--
 3 files changed, 23 insertions(+), 14 deletions(-)

diff --git a/kvm-all.c b/kvm-all.c
index b404f76..6f92874 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -1476,6 +1476,10 @@ int kvm_qemu_init()
 	kvm_context->no_irqchip_creation = 0;
 	kvm_context->no_pit_creation = 0;
 
+#ifdef KVM_CAP_SET_GUEST_DEBUG
+    TAILQ_INIT(&kvm_state->kvm_sw_breakpoints);
+#endif
+
 	gsi_count = kvm_get_gsi_count(kvm_context);
 	if (gsi_count > 0) {
 		int gsi_bits, i;
@@ -3434,14 +3438,13 @@ int kvm_qemu_init_env(CPUState *cenv)
 }
 
 #ifdef KVM_CAP_SET_GUEST_DEBUG
-struct kvm_sw_breakpoint_head kvm_sw_breakpoints =
-    TAILQ_HEAD_INITIALIZER(kvm_sw_breakpoints);
 
-struct kvm_sw_breakpoint *kvm_find_sw_breakpoint(target_ulong pc)
+struct kvm_sw_breakpoint *kvm_find_sw_breakpoint(CPUState *env,
+                                                 target_ulong pc)
 {
     struct kvm_sw_breakpoint *bp;
 
-    TAILQ_FOREACH(bp, &kvm_sw_breakpoints, entry) {
+    TAILQ_FOREACH(bp, &env->kvm_state->kvm_sw_breakpoints, entry) {
 	if (bp->pc == pc)
 	    return bp;
     }
@@ -3476,6 +3479,11 @@ int kvm_update_guest_debug(CPUState *env, unsigned long reinject_trap)
     return data.err;
 }
 
+int kvm_sw_breakpoints_active(CPUState *env)
+{
+    return !TAILQ_EMPTY(&env->kvm_state->kvm_sw_breakpoints);
+}
+
 int kvm_insert_breakpoint(CPUState *current_env, target_ulong addr,
                           target_ulong len, int type)
 {
@@ -3484,7 +3492,7 @@ int kvm_insert_breakpoint(CPUState *current_env, target_ulong addr,
     int err;
 
     if (type == GDB_BREAKPOINT_SW) {
-	bp = kvm_find_sw_breakpoint(addr);
+	bp = kvm_find_sw_breakpoint(current_env, addr);
 	if (bp) {
 	    bp->use_count++;
 	    return 0;
@@ -3502,7 +3510,8 @@ int kvm_insert_breakpoint(CPUState *current_env, target_ulong addr,
 	    return err;
 	}
 
-	TAILQ_INSERT_HEAD(&kvm_sw_breakpoints, bp, entry);
+    TAILQ_INSERT_HEAD(&current_env->kvm_state->kvm_sw_breakpoints,
+                      bp, entry);
     } else {
 	err = kvm_arch_insert_hw_breakpoint(addr, len, type);
 	if (err)
@@ -3525,7 +3534,7 @@ int kvm_remove_breakpoint(CPUState *current_env, target_ulong addr,
     int err;
 
     if (type == GDB_BREAKPOINT_SW) {
-	bp = kvm_find_sw_breakpoint(addr);
+	bp = kvm_find_sw_breakpoint(current_env, addr);
 	if (!bp)
 	    return -ENOENT;
 
@@ -3538,7 +3547,7 @@ int kvm_remove_breakpoint(CPUState *current_env, target_ulong addr,
 	if (err)
 	    return err;
 
-	TAILQ_REMOVE(&kvm_sw_breakpoints, bp, entry);
+	TAILQ_REMOVE(&current_env->kvm_state->kvm_sw_breakpoints, bp, entry);
 	qemu_free(bp);
     } else {
 	err = kvm_arch_remove_hw_breakpoint(addr, len, type);
@@ -3559,7 +3568,7 @@ void kvm_remove_all_breakpoints(CPUState *current_env)
     struct kvm_sw_breakpoint *bp, *next;
     CPUState *env;
 
-    TAILQ_FOREACH_SAFE(bp, &kvm_sw_breakpoints, entry, next) {
+    TAILQ_FOREACH_SAFE(bp, &current_env->kvm_state->kvm_sw_breakpoints, entry, next) {
         if (kvm_arch_remove_sw_breakpoint(current_env, bp) != 0) {
             /* Try harder to find a CPU that currently sees the breakpoint. */
             for (env = first_cpu; env != NULL; env = env->next_cpu) {
diff --git a/qemu-kvm.h b/qemu-kvm.h
index 4c185fd..bce80a2 100644
--- a/qemu-kvm.h
+++ b/qemu-kvm.h
@@ -88,12 +88,12 @@ struct kvm_sw_breakpoint {
     int use_count;
     TAILQ_ENTRY(kvm_sw_breakpoint) entry;
 };
-TAILQ_HEAD(kvm_sw_breakpoint_head, kvm_sw_breakpoint);
 
-extern struct kvm_sw_breakpoint_head kvm_sw_breakpoints;
+TAILQ_HEAD(kvm_sw_breakpoint_head, kvm_sw_breakpoint);
 
 int kvm_arch_debug(struct kvm_debug_exit_arch *arch_info);
-struct kvm_sw_breakpoint *kvm_find_sw_breakpoint(target_ulong pc);
+int kvm_sw_breakpoints_active(CPUState *env);
+struct kvm_sw_breakpoint *kvm_find_sw_breakpoint(CPUState *env, target_ulong pc);
 int kvm_arch_insert_sw_breakpoint(CPUState *current_env,
                                   struct kvm_sw_breakpoint *bp);
 int kvm_arch_remove_sw_breakpoint(CPUState *current_env,
diff --git a/target-i386/kvm.c b/target-i386/kvm.c
index ab324f6..66da1ba 100644
--- a/target-i386/kvm.c
+++ b/target-i386/kvm.c
@@ -2433,7 +2433,7 @@ int kvm_arch_debug(struct kvm_debug_exit_arch *arch_info)
 			break;
 		    }
 	}
-    } else if (kvm_find_sw_breakpoint(arch_info->pc))
+    } else if (kvm_find_sw_breakpoint(cpu_single_env, arch_info->pc))
 	handle = 1;
 
     if (!handle)
@@ -2456,7 +2456,7 @@ void kvm_arch_update_guest_debug(CPUState *env, struct kvm_guest_debug *dbg)
     };
     int n;
 
-    if (!TAILQ_EMPTY(&kvm_sw_breakpoints))
+    if (kvm_sw_breakpoints_active(env))
 	dbg->control |= KVM_GUESTDBG_ENABLE | KVM_GUESTDBG_USE_SW_BP;
 
     if (nb_hw_breakpoint > 0) {
-- 
1.6.2.2

[PATCH 8/8] use upstream code for breakpoint handling

[Glauber Costa]

---
 kvm-all.c         |  422 +++++++++++++++-----------------------------
 target-i386/kvm.c |  514 ++++++++++++++++++-----------------------------------
 2 files changed, 314 insertions(+), 622 deletions(-)

diff --git a/kvm-all.c b/kvm-all.c
index 6f92874..a364d34 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -63,6 +63,148 @@ struct KVMState
 static KVMState *kvm_state;
 kvm_context_t kvm_context;
 
+#ifdef KVM_CAP_SET_GUEST_DEBUG
+struct kvm_sw_breakpoint *kvm_find_sw_breakpoint(CPUState *env,
+                                                 target_ulong pc)
+{
+    struct kvm_sw_breakpoint *bp;
+
+    TAILQ_FOREACH(bp, &env->kvm_state->kvm_sw_breakpoints, entry) {
+        if (bp->pc == pc)
+            return bp;
+    }
+    return NULL;
+}
+
+int kvm_sw_breakpoints_active(CPUState *env)
+{
+    return !TAILQ_EMPTY(&env->kvm_state->kvm_sw_breakpoints);
+}
+
+int kvm_insert_breakpoint(CPUState *current_env, target_ulong addr,
+                          target_ulong len, int type)
+{
+    struct kvm_sw_breakpoint *bp;
+    CPUState *env;
+    int err;
+
+    if (type == GDB_BREAKPOINT_SW) {
+        bp = kvm_find_sw_breakpoint(current_env, addr);
+        if (bp) {
+            bp->use_count++;
+            return 0;
+        }
+
+        bp = qemu_malloc(sizeof(struct kvm_sw_breakpoint));
+        if (!bp)
+            return -ENOMEM;
+
+        bp->pc = addr;
+        bp->use_count = 1;
+        err = kvm_arch_insert_sw_breakpoint(current_env, bp);
+        if (err) {
+            free(bp);
+            return err;
+        }
+
+        TAILQ_INSERT_HEAD(&current_env->kvm_state->kvm_sw_breakpoints,
+                          bp, entry);
+    } else {
+        err = kvm_arch_insert_hw_breakpoint(addr, len, type);
+        if (err)
+            return err;
+    }
+
+    for (env = first_cpu; env != NULL; env = env->next_cpu) {
+        err = kvm_update_guest_debug(env, 0);
+        if (err)
+            return err;
+    }
+    return 0;
+}
+
+int kvm_remove_breakpoint(CPUState *current_env, target_ulong addr,
+                          target_ulong len, int type)
+{
+    struct kvm_sw_breakpoint *bp;
+    CPUState *env;
+    int err;
+
+    if (type == GDB_BREAKPOINT_SW) {
+        bp = kvm_find_sw_breakpoint(current_env, addr);
+        if (!bp)
+            return -ENOENT;
+
+        if (bp->use_count > 1) {
+            bp->use_count--;
+            return 0;
+        }
+
+        err = kvm_arch_remove_sw_breakpoint(current_env, bp);
+        if (err)
+            return err;
+
+        TAILQ_REMOVE(&current_env->kvm_state->kvm_sw_breakpoints, bp, entry);
+        qemu_free(bp);
+    } else {
+        err = kvm_arch_remove_hw_breakpoint(addr, len, type);
+        if (err)
+            return err;
+    }
+
+    for (env = first_cpu; env != NULL; env = env->next_cpu) {
+        err = kvm_update_guest_debug(env, 0);
+        if (err)
+            return err;
+    }
+    return 0;
+}
+
+void kvm_remove_all_breakpoints(CPUState *current_env)
+{
+    struct kvm_sw_breakpoint *bp, *next;
+    KVMState *s = current_env->kvm_state;
+    CPUState *env;
+
+    TAILQ_FOREACH_SAFE(bp, &s->kvm_sw_breakpoints, entry, next) {
+        if (kvm_arch_remove_sw_breakpoint(current_env, bp) != 0) {
+            /* Try harder to find a CPU that currently sees the breakpoint. */
+            for (env = first_cpu; env != NULL; env = env->next_cpu) {
+                if (kvm_arch_remove_sw_breakpoint(env, bp) == 0)
+                    break;
+            }
+        }
+    }
+    kvm_arch_remove_all_hw_breakpoints();
+
+    for (env = first_cpu; env != NULL; env = env->next_cpu)
+        kvm_update_guest_debug(env, 0);
+}
+
+#else /* !KVM_CAP_SET_GUEST_DEBUG */
+
+int kvm_update_guest_debug(CPUState *env, unsigned long reinject_trap)
+{
+    return -EINVAL;
+}
+
+int kvm_insert_breakpoint(CPUState *current_env, target_ulong addr,
+                          target_ulong len, int type)
+{
+    return -EINVAL;
+}
+
+int kvm_remove_breakpoint(CPUState *current_env, target_ulong addr,
+                          target_ulong len, int type)
+{
+    return -EINVAL;
+}
+
+void kvm_remove_all_breakpoints(CPUState *current_env)
+{
+}
+#endif /* !KVM_CAP_SET_GUEST_DEBUG */
+
 #ifdef CONFIG_KVM
 
 /* KVM uses PAGE_SIZE in it's definition of COALESCED_MMIO_MAX */
@@ -885,24 +1027,6 @@ void kvm_setup_guest_memory(void *start, size_t size)
     }
 }
 
-#ifdef KVM_CAP_SET_GUEST_DEBUG
-struct kvm_sw_breakpoint *kvm_find_sw_breakpoint(CPUState *env,
-                                                 target_ulong pc)
-{
-    struct kvm_sw_breakpoint *bp;
-
-    TAILQ_FOREACH(bp, &env->kvm_state->kvm_sw_breakpoints, entry) {
-        if (bp->pc == pc)
-            return bp;
-    }
-    return NULL;
-}
-
-int kvm_sw_breakpoints_active(CPUState *env)
-{
-    return !TAILQ_EMPTY(&env->kvm_state->kvm_sw_breakpoints);
-}
-
 int kvm_update_guest_debug(CPUState *env, unsigned long reinject_trap)
 {
     struct kvm_guest_debug dbg;
@@ -917,129 +1041,6 @@ int kvm_update_guest_debug(CPUState *env, unsigned long reinject_trap)
     return kvm_vcpu_ioctl(env, KVM_SET_GUEST_DEBUG, &dbg);
 }
 
-int kvm_insert_breakpoint(CPUState *current_env, target_ulong addr,
-                          target_ulong len, int type)
-{
-    struct kvm_sw_breakpoint *bp;
-    CPUState *env;
-    int err;
-
-    if (type == GDB_BREAKPOINT_SW) {
-        bp = kvm_find_sw_breakpoint(current_env, addr);
-        if (bp) {
-            bp->use_count++;
-            return 0;
-        }
-
-        bp = qemu_malloc(sizeof(struct kvm_sw_breakpoint));
-        if (!bp)
-            return -ENOMEM;
-
-        bp->pc = addr;
-        bp->use_count = 1;
-        err = kvm_arch_insert_sw_breakpoint(current_env, bp);
-        if (err) {
-            free(bp);
-            return err;
-        }
-
-        TAILQ_INSERT_HEAD(&current_env->kvm_state->kvm_sw_breakpoints,
-                          bp, entry);
-    } else {
-        err = kvm_arch_insert_hw_breakpoint(addr, len, type);
-        if (err)
-            return err;
-    }
-
-    for (env = first_cpu; env != NULL; env = env->next_cpu) {
-        err = kvm_update_guest_debug(env, 0);
-        if (err)
-            return err;
-    }
-    return 0;
-}
-
-int kvm_remove_breakpoint(CPUState *current_env, target_ulong addr,
-                          target_ulong len, int type)
-{
-    struct kvm_sw_breakpoint *bp;
-    CPUState *env;
-    int err;
-
-    if (type == GDB_BREAKPOINT_SW) {
-        bp = kvm_find_sw_breakpoint(current_env, addr);
-        if (!bp)
-            return -ENOENT;
-
-        if (bp->use_count > 1) {
-            bp->use_count--;
-            return 0;
-        }
-
-        err = kvm_arch_remove_sw_breakpoint(current_env, bp);
-        if (err)
-            return err;
-
-        TAILQ_REMOVE(&current_env->kvm_state->kvm_sw_breakpoints, bp, entry);
-        qemu_free(bp);
-    } else {
-        err = kvm_arch_remove_hw_breakpoint(addr, len, type);
-        if (err)
-            return err;
-    }
-
-    for (env = first_cpu; env != NULL; env = env->next_cpu) {
-        err = kvm_update_guest_debug(env, 0);
-        if (err)
-            return err;
-    }
-    return 0;
-}
-
-void kvm_remove_all_breakpoints(CPUState *current_env)
-{
-    struct kvm_sw_breakpoint *bp, *next;
-    KVMState *s = current_env->kvm_state;
-    CPUState *env;
-
-    TAILQ_FOREACH_SAFE(bp, &s->kvm_sw_breakpoints, entry, next) {
-        if (kvm_arch_remove_sw_breakpoint(current_env, bp) != 0) {
-            /* Try harder to find a CPU that currently sees the breakpoint. */
-            for (env = first_cpu; env != NULL; env = env->next_cpu) {
-                if (kvm_arch_remove_sw_breakpoint(env, bp) == 0)
-                    break;
-            }
-        }
-    }
-    kvm_arch_remove_all_hw_breakpoints();
-
-    for (env = first_cpu; env != NULL; env = env->next_cpu)
-        kvm_update_guest_debug(env, 0);
-}
-
-#else /* !KVM_CAP_SET_GUEST_DEBUG */
-
-int kvm_update_guest_debug(CPUState *env, unsigned long reinject_trap)
-{
-    return -EINVAL;
-}
-
-int kvm_insert_breakpoint(CPUState *current_env, target_ulong addr,
-                          target_ulong len, int type)
-{
-    return -EINVAL;
-}
-
-int kvm_remove_breakpoint(CPUState *current_env, target_ulong addr,
-                          target_ulong len, int type)
-{
-    return -EINVAL;
-}
-
-void kvm_remove_all_breakpoints(CPUState *current_env)
-{
-}
-#endif /* !KVM_CAP_SET_GUEST_DEBUG */
 #endif /* CONFIG_KVM */
 
 #ifdef USE_KVM
@@ -3439,18 +3440,6 @@ int kvm_qemu_init_env(CPUState *cenv)
 
 #ifdef KVM_CAP_SET_GUEST_DEBUG
 
-struct kvm_sw_breakpoint *kvm_find_sw_breakpoint(CPUState *env,
-                                                 target_ulong pc)
-{
-    struct kvm_sw_breakpoint *bp;
-
-    TAILQ_FOREACH(bp, &env->kvm_state->kvm_sw_breakpoints, entry) {
-	if (bp->pc == pc)
-	    return bp;
-    }
-    return NULL;
-}
-
 struct kvm_set_guest_debug_data {
     struct kvm_guest_debug dbg;
     int err;
@@ -3478,133 +3467,6 @@ int kvm_update_guest_debug(CPUState *env, unsigned long reinject_trap)
     on_vcpu(env, kvm_invoke_set_guest_debug, &data);
     return data.err;
 }
-
-int kvm_sw_breakpoints_active(CPUState *env)
-{
-    return !TAILQ_EMPTY(&env->kvm_state->kvm_sw_breakpoints);
-}
-
-int kvm_insert_breakpoint(CPUState *current_env, target_ulong addr,
-                          target_ulong len, int type)
-{
-    struct kvm_sw_breakpoint *bp;
-    CPUState *env;
-    int err;
-
-    if (type == GDB_BREAKPOINT_SW) {
-	bp = kvm_find_sw_breakpoint(current_env, addr);
-	if (bp) {
-	    bp->use_count++;
-	    return 0;
-	}
-
-	bp = qemu_malloc(sizeof(struct kvm_sw_breakpoint));
-	if (!bp)
-	    return -ENOMEM;
-
-	bp->pc = addr;
-	bp->use_count = 1;
-	err = kvm_arch_insert_sw_breakpoint(current_env, bp);
-	if (err) {
-	    free(bp);
-	    return err;
-	}
-
-    TAILQ_INSERT_HEAD(&current_env->kvm_state->kvm_sw_breakpoints,
-                      bp, entry);
-    } else {
-	err = kvm_arch_insert_hw_breakpoint(addr, len, type);
-	if (err)
-	    return err;
-    }
-
-    for (env = first_cpu; env != NULL; env = env->next_cpu) {
-	err = kvm_update_guest_debug(env, 0);
-	if (err)
-	    return err;
-    }
-    return 0;
-}
-
-int kvm_remove_breakpoint(CPUState *current_env, target_ulong addr,
-                          target_ulong len, int type)
-{
-    struct kvm_sw_breakpoint *bp;
-    CPUState *env;
-    int err;
-
-    if (type == GDB_BREAKPOINT_SW) {
-	bp = kvm_find_sw_breakpoint(current_env, addr);
-	if (!bp)
-	    return -ENOENT;
-
-	if (bp->use_count > 1) {
-	    bp->use_count--;
-	    return 0;
-	}
-
-	err = kvm_arch_remove_sw_breakpoint(current_env, bp);
-	if (err)
-	    return err;
-
-	TAILQ_REMOVE(&current_env->kvm_state->kvm_sw_breakpoints, bp, entry);
-	qemu_free(bp);
-    } else {
-	err = kvm_arch_remove_hw_breakpoint(addr, len, type);
-	if (err)
-	    return err;
-    }
-
-    for (env = first_cpu; env != NULL; env = env->next_cpu) {
-	err = kvm_update_guest_debug(env, 0);
-	if (err)
-	    return err;
-    }
-    return 0;
-}
-
-void kvm_remove_all_breakpoints(CPUState *current_env)
-{
-    struct kvm_sw_breakpoint *bp, *next;
-    CPUState *env;
-
-    TAILQ_FOREACH_SAFE(bp, &current_env->kvm_state->kvm_sw_breakpoints, entry, next) {
-        if (kvm_arch_remove_sw_breakpoint(current_env, bp) != 0) {
-            /* Try harder to find a CPU that currently sees the breakpoint. */
-            for (env = first_cpu; env != NULL; env = env->next_cpu) {
-                if (kvm_arch_remove_sw_breakpoint(env, bp) == 0)
-                    break;
-            }
-        }
-    }
-    kvm_arch_remove_all_hw_breakpoints();
-
-    for (env = first_cpu; env != NULL; env = env->next_cpu)
-	kvm_update_guest_debug(env, 0);
-}
-
-#else /* !KVM_CAP_SET_GUEST_DEBUG */
-
-int kvm_update_guest_debug(CPUState *env, unsigned long reinject_trap)
-{
-    return -EINVAL;
-}
-
-int kvm_insert_breakpoint(CPUState *current_env, target_ulong addr,
-                          target_ulong len, int type)
-{
-    return -EINVAL;
-}
-
-int kvm_remove_breakpoint(CPUState *current_env, target_ulong addr,
-                          target_ulong len, int type)
-{
-    return -EINVAL;
-}
-
-void kvm_remove_all_breakpoints(CPUState *current_env)
-{
-}
 #endif /* !KVM_CAP_SET_GUEST_DEBUG */
 
 /*
diff --git a/target-i386/kvm.c b/target-i386/kvm.c
index 66da1ba..54fc9dc 100644
--- a/target-i386/kvm.c
+++ b/target-i386/kvm.c
@@ -35,6 +35,178 @@
 #include "libkvm-all.h"
 #include "libkvm.h"
 
+#ifdef KVM_CAP_SET_GUEST_DEBUG
+int kvm_arch_insert_sw_breakpoint(CPUState *env, struct kvm_sw_breakpoint *bp)
+{
+    const static uint8_t int3 = 0xcc;
+
+    if (cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&bp->saved_insn, 1, 0) ||
+        cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&int3, 1, 1))
+        return -EINVAL;
+    return 0;
+}
+
+int kvm_arch_remove_sw_breakpoint(CPUState *env, struct kvm_sw_breakpoint *bp)
+{
+    uint8_t int3;
+
+    if (cpu_memory_rw_debug(env, bp->pc, &int3, 1, 0) || int3 != 0xcc ||
+        cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&bp->saved_insn, 1, 1))
+        return -EINVAL;
+    return 0;
+}
+
+static struct {
+    target_ulong addr;
+    int len;
+    int type;
+} hw_breakpoint[4];
+
+static int nb_hw_breakpoint;
+
+static int find_hw_breakpoint(target_ulong addr, int len, int type)
+{
+    int n;
+
+    for (n = 0; n < nb_hw_breakpoint; n++)
+        if (hw_breakpoint[n].addr == addr && hw_breakpoint[n].type == type &&
+            (hw_breakpoint[n].len == len || len == -1))
+            return n;
+    return -1;
+}
+
+int kvm_arch_insert_hw_breakpoint(target_ulong addr,
+                                  target_ulong len, int type)
+{
+    switch (type) {
+    case GDB_BREAKPOINT_HW:
+        len = 1;
+        break;
+    case GDB_WATCHPOINT_WRITE:
+    case GDB_WATCHPOINT_ACCESS:
+        switch (len) {
+        case 1:
+            break;
+        case 2:
+        case 4:
+        case 8:
+            if (addr & (len - 1))
+                return -EINVAL;
+            break;
+        default:
+            return -EINVAL;
+        }
+        break;
+    default:
+        return -ENOSYS;
+    }
+
+    if (nb_hw_breakpoint == 4)
+        return -ENOBUFS;
+
+    if (find_hw_breakpoint(addr, len, type) >= 0)
+        return -EEXIST;
+
+    hw_breakpoint[nb_hw_breakpoint].addr = addr;
+    hw_breakpoint[nb_hw_breakpoint].len = len;
+    hw_breakpoint[nb_hw_breakpoint].type = type;
+    nb_hw_breakpoint++;
+
+    return 0;
+}
+
+int kvm_arch_remove_hw_breakpoint(target_ulong addr,
+                                  target_ulong len, int type)
+{
+    int n;
+
+    n = find_hw_breakpoint(addr, (type == GDB_BREAKPOINT_HW) ? 1 : len, type);
+    if (n < 0)
+        return -ENOENT;
+
+    nb_hw_breakpoint--;
+    hw_breakpoint[n] = hw_breakpoint[nb_hw_breakpoint];
+
+    return 0;
+}
+
+void kvm_arch_remove_all_hw_breakpoints(void)
+{
+    nb_hw_breakpoint = 0;
+}
+
+static CPUWatchpoint hw_watchpoint;
+
+int kvm_arch_debug(struct kvm_debug_exit_arch *arch_info)
+{
+    int handle = 0;
+    int n;
+
+    if (arch_info->exception == 1) {
+        if (arch_info->dr6 & (1 << 14)) {
+            if (cpu_single_env->singlestep_enabled)
+                handle = 1;
+        } else {
+            for (n = 0; n < 4; n++)
+                if (arch_info->dr6 & (1 << n))
+                    switch ((arch_info->dr7 >> (16 + n*4)) & 0x3) {
+                    case 0x0:
+                        handle = 1;
+                        break;
+                    case 0x1:
+                        handle = 1;
+                        cpu_single_env->watchpoint_hit = &hw_watchpoint;
+                        hw_watchpoint.vaddr = hw_breakpoint[n].addr;
+                        hw_watchpoint.flags = BP_MEM_WRITE;
+                        break;
+                    case 0x3:
+                        handle = 1;
+                        cpu_single_env->watchpoint_hit = &hw_watchpoint;
+                        hw_watchpoint.vaddr = hw_breakpoint[n].addr;
+                        hw_watchpoint.flags = BP_MEM_ACCESS;
+                        break;
+                    }
+        }
+    } else if (kvm_find_sw_breakpoint(cpu_single_env, arch_info->pc))
+        handle = 1;
+
+    if (!handle)
+        kvm_update_guest_debug(cpu_single_env,
+                        (arch_info->exception == 1) ?
+                        KVM_GUESTDBG_INJECT_DB : KVM_GUESTDBG_INJECT_BP);
+
+    return handle;
+}
+
+void kvm_arch_update_guest_debug(CPUState *env, struct kvm_guest_debug *dbg)
+{
+    const uint8_t type_code[] = {
+        [GDB_BREAKPOINT_HW] = 0x0,
+        [GDB_WATCHPOINT_WRITE] = 0x1,
+        [GDB_WATCHPOINT_ACCESS] = 0x3
+    };
+    const uint8_t len_code[] = {
+        [1] = 0x0, [2] = 0x1, [4] = 0x3, [8] = 0x2
+    };
+    int n;
+
+    if (kvm_sw_breakpoints_active(env))
+        dbg->control |= KVM_GUESTDBG_ENABLE | KVM_GUESTDBG_USE_SW_BP;
+
+    if (nb_hw_breakpoint > 0) {
+        dbg->control |= KVM_GUESTDBG_ENABLE | KVM_GUESTDBG_USE_HW_BP;
+        dbg->arch.debugreg[7] = 0x0600;
+        for (n = 0; n < nb_hw_breakpoint; n++) {
+            dbg->arch.debugreg[n] = hw_breakpoint[n].addr;
+            dbg->arch.debugreg[7] |= (2 << (n * 2)) |
+                (type_code[hw_breakpoint[n].type] << (16 + n*4)) |
+                (len_code[hw_breakpoint[n].len] << (18 + n*4));
+        }
+    }
+}
+#endif /* KVM_CAP_SET_GUEST_DEBUG */
+
+
 #ifdef CONFIG_KVM
 
 //#define DEBUG_KVM
@@ -802,177 +974,6 @@ int kvm_arch_handle_exit(CPUState *env, struct kvm_run *run)
     return ret;
 }
 
-#ifdef KVM_CAP_SET_GUEST_DEBUG
-int kvm_arch_insert_sw_breakpoint(CPUState *env, struct kvm_sw_breakpoint *bp)
-{
-    const static uint8_t int3 = 0xcc;
-
-    if (cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&bp->saved_insn, 1, 0) ||
-        cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&int3, 1, 1))
-        return -EINVAL;
-    return 0;
-}
-
-int kvm_arch_remove_sw_breakpoint(CPUState *env, struct kvm_sw_breakpoint *bp)
-{
-    uint8_t int3;
-
-    if (cpu_memory_rw_debug(env, bp->pc, &int3, 1, 0) || int3 != 0xcc ||
-        cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&bp->saved_insn, 1, 1))
-        return -EINVAL;
-    return 0;
-}
-
-static struct {
-    target_ulong addr;
-    int len;
-    int type;
-} hw_breakpoint[4];
-
-static int nb_hw_breakpoint;
-
-static int find_hw_breakpoint(target_ulong addr, int len, int type)
-{
-    int n;
-
-    for (n = 0; n < nb_hw_breakpoint; n++)
-        if (hw_breakpoint[n].addr == addr && hw_breakpoint[n].type == type &&
-            (hw_breakpoint[n].len == len || len == -1))
-            return n;
-    return -1;
-}
-
-int kvm_arch_insert_hw_breakpoint(target_ulong addr,
-                                  target_ulong len, int type)
-{
-    switch (type) {
-    case GDB_BREAKPOINT_HW:
-        len = 1;
-        break;
-    case GDB_WATCHPOINT_WRITE:
-    case GDB_WATCHPOINT_ACCESS:
-        switch (len) {
-        case 1:
-            break;
-        case 2:
-        case 4:
-        case 8:
-            if (addr & (len - 1))
-                return -EINVAL;
-            break;
-        default:
-            return -EINVAL;
-        }
-        break;
-    default:
-        return -ENOSYS;
-    }
-
-    if (nb_hw_breakpoint == 4)
-        return -ENOBUFS;
-
-    if (find_hw_breakpoint(addr, len, type) >= 0)
-        return -EEXIST;
-
-    hw_breakpoint[nb_hw_breakpoint].addr = addr;
-    hw_breakpoint[nb_hw_breakpoint].len = len;
-    hw_breakpoint[nb_hw_breakpoint].type = type;
-    nb_hw_breakpoint++;
-
-    return 0;
-}
-
-int kvm_arch_remove_hw_breakpoint(target_ulong addr,
-                                  target_ulong len, int type)
-{
-    int n;
-
-    n = find_hw_breakpoint(addr, (type == GDB_BREAKPOINT_HW) ? 1 : len, type);
-    if (n < 0)
-        return -ENOENT;
-
-    nb_hw_breakpoint--;
-    hw_breakpoint[n] = hw_breakpoint[nb_hw_breakpoint];
-
-    return 0;
-}
-
-void kvm_arch_remove_all_hw_breakpoints(void)
-{
-    nb_hw_breakpoint = 0;
-}
-
-static CPUWatchpoint hw_watchpoint;
-
-int kvm_arch_debug(struct kvm_debug_exit_arch *arch_info)
-{
-    int handle = 0;
-    int n;
-
-    if (arch_info->exception == 1) {
-        if (arch_info->dr6 & (1 << 14)) {
-            if (cpu_single_env->singlestep_enabled)
-                handle = 1;
-        } else {
-            for (n = 0; n < 4; n++)
-                if (arch_info->dr6 & (1 << n))
-                    switch ((arch_info->dr7 >> (16 + n*4)) & 0x3) {
-                    case 0x0:
-                        handle = 1;
-                        break;
-                    case 0x1:
-                        handle = 1;
-                        cpu_single_env->watchpoint_hit = &hw_watchpoint;
-                        hw_watchpoint.vaddr = hw_breakpoint[n].addr;
-                        hw_watchpoint.flags = BP_MEM_WRITE;
-                        break;
-                    case 0x3:
-                        handle = 1;
-                        cpu_single_env->watchpoint_hit = &hw_watchpoint;
-                        hw_watchpoint.vaddr = hw_breakpoint[n].addr;
-                        hw_watchpoint.flags = BP_MEM_ACCESS;
-                        break;
-                    }
-        }
-    } else if (kvm_find_sw_breakpoint(cpu_single_env, arch_info->pc))
-        handle = 1;
-
-    if (!handle)
-        kvm_update_guest_debug(cpu_single_env,
-                        (arch_info->exception == 1) ?
-                        KVM_GUESTDBG_INJECT_DB : KVM_GUESTDBG_INJECT_BP);
-
-    return handle;
-}
-
-void kvm_arch_update_guest_debug(CPUState *env, struct kvm_guest_debug *dbg)
-{
-    const uint8_t type_code[] = {
-        [GDB_BREAKPOINT_HW] = 0x0,
-        [GDB_WATCHPOINT_WRITE] = 0x1,
-        [GDB_WATCHPOINT_ACCESS] = 0x3
-    };
-    const uint8_t len_code[] = {
-        [1] = 0x0, [2] = 0x1, [4] = 0x3, [8] = 0x2
-    };
-    int n;
-
-    if (kvm_sw_breakpoints_active(env))
-        dbg->control |= KVM_GUESTDBG_ENABLE | KVM_GUESTDBG_USE_SW_BP;
-
-    if (nb_hw_breakpoint > 0) {
-        dbg->control |= KVM_GUESTDBG_ENABLE | KVM_GUESTDBG_USE_HW_BP;
-        dbg->arch.debugreg[7] = 0x0600;
-        for (n = 0; n < nb_hw_breakpoint; n++) {
-            dbg->arch.debugreg[n] = hw_breakpoint[n].addr;
-            dbg->arch.debugreg[7] |= (2 << (n * 2)) |
-                (type_code[hw_breakpoint[n].type] << (16 + n*4)) |
-                (len_code[hw_breakpoint[n].len] << (18 + n*4));
-        }
-    }
-}
-#endif /* KVM_CAP_SET_GUEST_DEBUG */
-
 #endif /* CONFIG_KVM */
 #ifdef USE_KVM
 
@@ -2301,177 +2302,6 @@ void kvm_arch_cpu_reset(CPUState *env)
     }
 }
 
-int kvm_arch_insert_sw_breakpoint(CPUState *env, struct kvm_sw_breakpoint *bp)
-{
-    uint8_t int3 = 0xcc;
-
-    if (cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&bp->saved_insn, 1, 0) ||
-        cpu_memory_rw_debug(env, bp->pc, &int3, 1, 1))
-        return -EINVAL;
-    return 0;
-}
-
-int kvm_arch_remove_sw_breakpoint(CPUState *env, struct kvm_sw_breakpoint *bp)
-{
-    uint8_t int3;
-
-    if (cpu_memory_rw_debug(env, bp->pc, &int3, 1, 0) || int3 != 0xcc ||
-        cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&bp->saved_insn, 1, 1))
-        return -EINVAL;
-    return 0;
-}
-
-#ifdef KVM_CAP_SET_GUEST_DEBUG
-static struct {
-    target_ulong addr;
-    int len;
-    int type;
-} hw_breakpoint[4];
-
-static int nb_hw_breakpoint;
-
-static int find_hw_breakpoint(target_ulong addr, int len, int type)
-{
-    int n;
-
-    for (n = 0; n < nb_hw_breakpoint; n++)
-	if (hw_breakpoint[n].addr == addr && hw_breakpoint[n].type == type &&
-	    (hw_breakpoint[n].len == len || len == -1))
-	    return n;
-    return -1;
-}
-
-int kvm_arch_insert_hw_breakpoint(target_ulong addr,
-                                  target_ulong len, int type)
-{
-    switch (type) {
-    case GDB_BREAKPOINT_HW:
-	len = 1;
-	break;
-    case GDB_WATCHPOINT_WRITE:
-    case GDB_WATCHPOINT_ACCESS:
-	switch (len) {
-	case 1:
-	    break;
-	case 2:
-	case 4:
-	case 8:
-	    if (addr & (len - 1))
-		return -EINVAL;
-	    break;
-	default:
-	    return -EINVAL;
-	}
-	break;
-    default:
-	return -ENOSYS;
-    }
-
-    if (nb_hw_breakpoint == 4)
-        return -ENOBUFS;
-
-    if (find_hw_breakpoint(addr, len, type) >= 0)
-        return -EEXIST;
-
-    hw_breakpoint[nb_hw_breakpoint].addr = addr;
-    hw_breakpoint[nb_hw_breakpoint].len = len;
-    hw_breakpoint[nb_hw_breakpoint].type = type;
-    nb_hw_breakpoint++;
-
-    return 0;
-}
-
-int kvm_arch_remove_hw_breakpoint(target_ulong addr,
-                                  target_ulong len, int type)
-{
-    int n;
-
-    n = find_hw_breakpoint(addr, (type == GDB_BREAKPOINT_HW) ? 1 : len, type);
-    if (n < 0)
-        return -ENOENT;
-
-    nb_hw_breakpoint--;
-    hw_breakpoint[n] = hw_breakpoint[nb_hw_breakpoint];
-
-    return 0;
-}
-
-void kvm_arch_remove_all_hw_breakpoints(void)
-{
-    nb_hw_breakpoint = 0;
-}
-
-static CPUWatchpoint hw_watchpoint;
-
-int kvm_arch_debug(struct kvm_debug_exit_arch *arch_info)
-{
-    int handle = 0;
-    int n;
-
-    if (arch_info->exception == 1) {
-	if (arch_info->dr6 & (1 << 14)) {
-	    if (cpu_single_env->singlestep_enabled)
-		handle = 1;
-	} else {
-	    for (n = 0; n < 4; n++)
-		if (arch_info->dr6 & (1 << n))
-		    switch ((arch_info->dr7 >> (16 + n*4)) & 0x3) {
-		    case 0x0:
-			handle = 1;
-			break;
-		    case 0x1:
-			handle = 1;
-			cpu_single_env->watchpoint_hit = &hw_watchpoint;
-			hw_watchpoint.vaddr = hw_breakpoint[n].addr;
-			hw_watchpoint.flags = BP_MEM_WRITE;
-			break;
-		    case 0x3:
-			handle = 1;
-			cpu_single_env->watchpoint_hit = &hw_watchpoint;
-			hw_watchpoint.vaddr = hw_breakpoint[n].addr;
-			hw_watchpoint.flags = BP_MEM_ACCESS;
-			break;
-		    }
-	}
-    } else if (kvm_find_sw_breakpoint(cpu_single_env, arch_info->pc))
-	handle = 1;
-
-    if (!handle)
-	kvm_update_guest_debug(cpu_single_env,
-			(arch_info->exception == 1) ?
-			KVM_GUESTDBG_INJECT_DB : KVM_GUESTDBG_INJECT_BP);
-
-    return handle;
-}
-
-void kvm_arch_update_guest_debug(CPUState *env, struct kvm_guest_debug *dbg)
-{
-    const uint8_t type_code[] = {
-	[GDB_BREAKPOINT_HW] = 0x0,
-	[GDB_WATCHPOINT_WRITE] = 0x1,
-	[GDB_WATCHPOINT_ACCESS] = 0x3
-    };
-    const uint8_t len_code[] = {
-	[1] = 0x0, [2] = 0x1, [4] = 0x3, [8] = 0x2
-    };
-    int n;
-
-    if (kvm_sw_breakpoints_active(env))
-	dbg->control |= KVM_GUESTDBG_ENABLE | KVM_GUESTDBG_USE_SW_BP;
-
-    if (nb_hw_breakpoint > 0) {
-	dbg->control |= KVM_GUESTDBG_ENABLE | KVM_GUESTDBG_USE_HW_BP;
-	dbg->arch.debugreg[7] = 0x0600;
-	for (n = 0; n < nb_hw_breakpoint; n++) {
-	    dbg->arch.debugreg[n] = hw_breakpoint[n].addr;
-	    dbg->arch.debugreg[7] |= (2 << (n * 2)) |
-		(type_code[hw_breakpoint[n].type] << (16 + n*4)) |
-		(len_code[hw_breakpoint[n].len] << (18 + n*4));
-	}
-    }
-}
-#endif
-
 void kvm_arch_do_ioperm(void *_data)
 {
     struct ioperm_data *data = _data;
-- 
1.6.2.2

Re: [PATCH 7/8] use kvm_upstream sw_breakpoints structure

[Avi Kivity]

On 07/08/2009 04:09 PM, Glauber Costa wrote:
> Signed-off-by: Glauber Costa<glommer@redhat.com>
> ---
>   kvm-all.c         |   27 ++++++++++++++++++---------
>   qemu-kvm.h        |    6 +++---
>   target-i386/kvm.c |    4 ++--
>   3 files changed, 23 insertions(+), 14 deletions(-)
>
>    

Did you test the functionality to ensure we do not regress?

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

Re: [PATCH 7/8] use kvm_upstream sw_breakpoints structure

[Glauber Costa]

On Wed, Jul 08, 2009 at 04:27:27PM +0300, Avi Kivity wrote:
> On 07/08/2009 04:09 PM, Glauber Costa wrote:
>> Signed-off-by: Glauber Costa<glommer@redhat.com>
>> ---
>>   kvm-all.c         |   27 ++++++++++++++++++---------
>>   qemu-kvm.h        |    6 +++---
>>   target-i386/kvm.c |    4 ++--
>>   3 files changed, 23 insertions(+), 14 deletions(-)
>>
>>    
>
> Did you test the functionality to ensure we do not regress?
just briefly. To be honest, I'd feel much more confortable
if Jan could give it a try.

Re: [PATCH 7/8] use kvm_upstream sw_breakpoints structure

[Avi Kivity]

On 07/08/2009 04:39 PM, Glauber Costa wrote:
> On Wed, Jul 08, 2009 at 04:27:27PM +0300, Avi Kivity wrote:
>    
>> On 07/08/2009 04:09 PM, Glauber Costa wrote:
>>      
>>> Signed-off-by: Glauber Costa<glommer@redhat.com>
>>> ---
>>>    kvm-all.c         |   27 ++++++++++++++++++---------
>>>    qemu-kvm.h        |    6 +++---
>>>    target-i386/kvm.c |    4 ++--
>>>    3 files changed, 23 insertions(+), 14 deletions(-)
>>>
>>>
>>>        
>> Did you test the functionality to ensure we do not regress?
>>      
> just briefly. To be honest, I'd feel much more confortable
> if Jan could give it a try.
>    

A Jan try (+review) will definitely help.

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

Re: [PATCH 7/8] use kvm_upstream sw_breakpoints structure

[Jan Kiszka]

[-- Attachment #1: Type: text/plain, Size: 743 bytes --]

Avi Kivity wrote:
> On 07/08/2009 04:39 PM, Glauber Costa wrote:
>> On Wed, Jul 08, 2009 at 04:27:27PM +0300, Avi Kivity wrote:
>>   
>>> On 07/08/2009 04:09 PM, Glauber Costa wrote:
>>>     
>>>> Signed-off-by: Glauber Costa<glommer@redhat.com>
>>>> ---
>>>>    kvm-all.c         |   27 ++++++++++++++++++---------
>>>>    qemu-kvm.h        |    6 +++---
>>>>    target-i386/kvm.c |    4 ++--
>>>>    3 files changed, 23 insertions(+), 14 deletions(-)
>>>>
>>>>
>>>>        
>>> Did you test the functionality to ensure we do not regress?
>>>      
>> just briefly. To be honest, I'd feel much more confortable
>> if Jan could give it a try.
>>    
> 
> A Jan try (+review) will definitely help.
> 

Queued.

Jan


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 257 bytes --]

Re: [PATCH 7/8] use kvm_upstream sw_breakpoints structure

[Jan Kiszka]

[-- Attachment #1: Type: text/plain, Size: 1317 bytes --]

Jan Kiszka wrote:
> Avi Kivity wrote:
>> On 07/08/2009 04:39 PM, Glauber Costa wrote:
>>> On Wed, Jul 08, 2009 at 04:27:27PM +0300, Avi Kivity wrote:
>>>   
>>>> On 07/08/2009 04:09 PM, Glauber Costa wrote:
>>>>     
>>>>> Signed-off-by: Glauber Costa<glommer@redhat.com>
>>>>> ---
>>>>>    kvm-all.c         |   27 ++++++++++++++++++---------
>>>>>    qemu-kvm.h        |    6 +++---
>>>>>    target-i386/kvm.c |    4 ++--
>>>>>    3 files changed, 23 insertions(+), 14 deletions(-)
>>>>>
>>>>>
>>>>>        
>>>> Did you test the functionality to ensure we do not regress?
>>>>      
>>> just briefly. To be honest, I'd feel much more confortable
>>> if Jan could give it a try.
>>>    
>> A Jan try (+review) will definitely help.
>>
> 
> Queued.
> 

Deferred until v2. My first impression is that too much upstream code is
moved or touched.

Glauber, if you want to use some function that is currently under
KVM_UPSTREAM, don't move it, just drop the #ifdef around it. And when
done, have a look at the diff between upstream and qemu-kvm to avoid
unneeded variations.

Another question: What prevents using CONFIG_KVM also for qemu-kvm? I
would rather mask out yet unused upstream code via the well-known
KVM_UPSTREAM and drop the easily confusable USE_KVM defines.

Jan


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 257 bytes --]

Re: [PATCH 7/8] use kvm_upstream sw_breakpoints structure

[Avi Kivity]

On 07/08/2009 10:44 PM, Jan Kiszka wrote:
>
> Deferred until v2. My first impression is that too much upstream code is
> moved or touched.
>
> Glauber, if you want to use some function that is currently under
> KVM_UPSTREAM, don't move it, just drop the #ifdef around it. And when
> done, have a look at the diff between upstream and qemu-kvm to avoid
> unneeded variations.
>    

Yes.  I though that's what we'd agreed, but forgot to verify it.  The 
only difference should be #endif/#ifdef pairs.

> Another question: What prevents using CONFIG_KVM also for qemu-kvm? I
> would rather mask out yet unused upstream code via the well-known
> KVM_UPSTREAM and drop the easily confusable USE_KVM defines.
>    

Good idea, would simplify ./configure as well.

-- 
I have a truly marvellous patch that fixes the bug which this
signature is too narrow to contain.

Re: [PATCH 6/8] provide env->kvm_fd

[Gleb Natapov]

On Wed, Jul 08, 2009 at 09:09:00AM -0400, Glauber Costa wrote:
> qemu upstream puts kvm information on env. Do that too, since it will
> allow us to use CPUState in cpu-specific functions, instead of kvm-specific
> types.
> 
> Signed-off-by: Glauber Costa <glommer@redhat.com>
> ---
>  kvm-all.c    |   12 ++++++++----
>  libkvm-all.h |   12 ------------
>  2 files changed, 8 insertions(+), 16 deletions(-)
> 
> diff --git a/kvm-all.c b/kvm-all.c
> index 15bd429..b404f76 100644
> --- a/kvm-all.c
> +++ b/kvm-all.c
> @@ -1519,21 +1519,25 @@ void kvm_disable_pit_creation(kvm_context_t kvm)
>  	kvm->no_pit_creation = 1;
>  }
>  
> -kvm_vcpu_context_t kvm_create_vcpu(kvm_context_t kvm, int id)
> +static kvm_vcpu_context_t kvm_create_vcpu(kvm_context_t kvm, CPUState *env)
>  {
>  	long mmap_size;
>  	int r;
>  	kvm_vcpu_context_t vcpu_ctx = qemu_malloc(sizeof(struct kvm_vcpu_context));
>  
>  	vcpu_ctx->kvm = kvm;
> -	vcpu_ctx->id = id;
> +	vcpu_ctx->id = env->cpu_index;
>  
> -	r = ioctl(kvm->vm_fd, KVM_CREATE_VCPU, id);
> +	r = ioctl(kvm->vm_fd, KVM_CREATE_VCPU, env->cpu_index);
This just reverts all my work to separate apic_id from env->cpu_index.
We want to bring good things from qemu-kvm to upstream not bad things
from upstream to qemu-kvm :) If anything you should use env->cpuid_apic_id
instead of env->cpu_index here.


--
			Gleb.

Re: [PATCH 6/8] provide env->kvm_fd

[Glauber Costa]

On Wed, Jul 08, 2009 at 04:32:37PM +0300, Gleb Natapov wrote:
> On Wed, Jul 08, 2009 at 09:09:00AM -0400, Glauber Costa wrote:
> > qemu upstream puts kvm information on env. Do that too, since it will
> > allow us to use CPUState in cpu-specific functions, instead of kvm-specific
> > types.
> > 
> > Signed-off-by: Glauber Costa <glommer@redhat.com>
> > ---
> >  kvm-all.c    |   12 ++++++++----
> >  libkvm-all.h |   12 ------------
> >  2 files changed, 8 insertions(+), 16 deletions(-)
> > 
> > diff --git a/kvm-all.c b/kvm-all.c
> > index 15bd429..b404f76 100644
> > --- a/kvm-all.c
> > +++ b/kvm-all.c
> > @@ -1519,21 +1519,25 @@ void kvm_disable_pit_creation(kvm_context_t kvm)
> >  	kvm->no_pit_creation = 1;
> >  }
> >  
> > -kvm_vcpu_context_t kvm_create_vcpu(kvm_context_t kvm, int id)
> > +static kvm_vcpu_context_t kvm_create_vcpu(kvm_context_t kvm, CPUState *env)
> >  {
> >  	long mmap_size;
> >  	int r;
> >  	kvm_vcpu_context_t vcpu_ctx = qemu_malloc(sizeof(struct kvm_vcpu_context));
> >  
> >  	vcpu_ctx->kvm = kvm;
> > -	vcpu_ctx->id = id;
> > +	vcpu_ctx->id = env->cpu_index;
> >  
> > -	r = ioctl(kvm->vm_fd, KVM_CREATE_VCPU, id);
> > +	r = ioctl(kvm->vm_fd, KVM_CREATE_VCPU, env->cpu_index);
> This just reverts all my work to separate apic_id from env->cpu_index.
> We want to bring good things from qemu-kvm to upstream not bad things
> from upstream to qemu-kvm :) If anything you should use env->cpuid_apic_id
> instead of env->cpu_index here.
cpuid_apic_id is fine.
my bad here.

Re: [PATCH 6/8] provide env->kvm_fd

[Gleb Natapov]

On Wed, Jul 08, 2009 at 10:44:01AM -0300, Glauber Costa wrote:
> On Wed, Jul 08, 2009 at 04:32:37PM +0300, Gleb Natapov wrote:
> > On Wed, Jul 08, 2009 at 09:09:00AM -0400, Glauber Costa wrote:
> > > qemu upstream puts kvm information on env. Do that too, since it will
> > > allow us to use CPUState in cpu-specific functions, instead of kvm-specific
> > > types.
> > > 
> > > Signed-off-by: Glauber Costa <glommer@redhat.com>
> > > ---
> > >  kvm-all.c    |   12 ++++++++----
> > >  libkvm-all.h |   12 ------------
> > >  2 files changed, 8 insertions(+), 16 deletions(-)
> > > 
> > > diff --git a/kvm-all.c b/kvm-all.c
> > > index 15bd429..b404f76 100644
> > > --- a/kvm-all.c
> > > +++ b/kvm-all.c
> > > @@ -1519,21 +1519,25 @@ void kvm_disable_pit_creation(kvm_context_t kvm)
> > >  	kvm->no_pit_creation = 1;
> > >  }
> > >  
> > > -kvm_vcpu_context_t kvm_create_vcpu(kvm_context_t kvm, int id)
> > > +static kvm_vcpu_context_t kvm_create_vcpu(kvm_context_t kvm, CPUState *env)
> > >  {
> > >  	long mmap_size;
> > >  	int r;
> > >  	kvm_vcpu_context_t vcpu_ctx = qemu_malloc(sizeof(struct kvm_vcpu_context));
> > >  
> > >  	vcpu_ctx->kvm = kvm;
> > > -	vcpu_ctx->id = id;
> > > +	vcpu_ctx->id = env->cpu_index;
> > >  
> > > -	r = ioctl(kvm->vm_fd, KVM_CREATE_VCPU, id);
> > > +	r = ioctl(kvm->vm_fd, KVM_CREATE_VCPU, env->cpu_index);
> > This just reverts all my work to separate apic_id from env->cpu_index.
> > We want to bring good things from qemu-kvm to upstream not bad things
> > from upstream to qemu-kvm :) If anything you should use env->cpuid_apic_id
> > instead of env->cpu_index here.
> cpuid_apic_id is fine.
> my bad here.
Actually I prefer to use id here and pass env->cpuid_apic_id to the
function.

--
			Gleb.

Re: [PATCH 3/8] put qemu-kvm-x86.c into kvm.c

[Avi Kivity]

On 07/08/2009 04:08 PM, Glauber Costa wrote:
> big plan to start sharing code with upstream.
>
>    

What's the big plan re qemu-kvm-ia64.c?

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

Re: [PATCH 3/8] put qemu-kvm-x86.c into kvm.c

[Gleb Natapov]

On Wed, Jul 08, 2009 at 09:08:57AM -0400, Glauber Costa wrote:
> big plan to start sharing code with upstream.
> 
It is hard enough (for me at least) to distinguish between upstream code
and qemu-kvm when I have to look at the code. Having them separated in
different files is very helpful. Looking at the filename in a grep result 
is enough to know if I should care about the result or not.

Can you simple include qemu-kvm-x86.c into kvm.c?

--
			Gleb.

Re: [PATCH 3/8] put qemu-kvm-x86.c into kvm.c

[Glauber Costa]

On Wed, Jul 08, 2009 at 04:28:51PM +0300, Gleb Natapov wrote:
> On Wed, Jul 08, 2009 at 09:08:57AM -0400, Glauber Costa wrote:
> > big plan to start sharing code with upstream.
> > 
> It is hard enough (for me at least) to distinguish between upstream code
> and qemu-kvm when I have to look at the code. Having them separated in
> different files is very helpful. Looking at the filename in a grep result 
> is enough to know if I should care about the result or not.
> 
> Can you simple include qemu-kvm-x86.c into kvm.c?
I could, but I believe this would make sharing, which is the ultimate goal,
a little bit cumbersome. Anyway, the goal is not to have it last long.

Re: [PATCH 2/8] move qemu-kvm.c to kvm-all.c

[Avi Kivity]

On 07/08/2009 04:08 PM, Glauber Costa wrote:
> Big plan to start sharing code with upstream.
>
>    

This patch is very brittle - if qemu-kvm.c changes, it won't apply.  I 
suggest to simply add #include "qemu-kvm.c" and remove it from the 
makefile.  After I apply it, I'll perform the #include and remove 
qemu-kvm.c.

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

Re: [PATCH 2/8] move qemu-kvm.c to kvm-all.c

[Glauber Costa]

On Wed, Jul 08, 2009 at 04:23:39PM +0300, Avi Kivity wrote:
> On 07/08/2009 04:08 PM, Glauber Costa wrote:
>> Big plan to start sharing code with upstream.
>>
>>    
>
> This patch is very brittle - if qemu-kvm.c changes, it won't apply.  I  
> suggest to simply add #include "qemu-kvm.c" and remove it from the  
> makefile.  After I apply it, I'll perform the #include and remove  
> qemu-kvm.c.
ok.

I'll do the same with kvm.c

Re: [PATCH 0/8] Move closer to upstream

[Avi Kivity]

On 07/08/2009 04:08 PM, Glauber Costa wrote:
> Hi,
>
> This is another step at getting us closer to qemu upstream. I'm moving
> our code to kvm-all.c and kvm.c, and relying on the fact that we use
> USE_KVM, while upstream qemu uses CONFIG_KVM, to keep us isolated.
>
> The goal is to slowly reduce that isolation. To demonstrate what I aim
> for, the last patches of the series shares code for breakpoint handling.
> next in my radar are ioctl functions and cpuid trimming.
>
>    

Overall this approach looks good.

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