Palacios Public Git Repository

To checkout Palacios execute

  git clone http://v3vee.org/palacios/palacios.web/palacios.git
This will give you the master branch. You probably want the devel branch or one of the release branches. To switch to the devel branch, simply execute
  cd palacios
  git checkout --track -b devel origin/devel
The other branches are similar.


added write protect flag to realmode CR0, to enable memory hooks
[palacios.git] / palacios / src / palacios / svm.c
index 7889c11..490e342 100644 (file)
@@ -1,3 +1,23 @@
+/* 
+ * This file is part of the Palacios Virtual Machine Monitor developed
+ * by the V3VEE Project with funding from the United States National 
+ * Science Foundation and the Department of Energy.  
+ *
+ * The V3VEE Project is a joint project between Northwestern University
+ * and the University of New Mexico.  You can find out more at 
+ * http://www.v3vee.org
+ *
+ * Copyright (c) 2008, Jack Lange <jarusl@cs.northwestern.edu> 
+ * Copyright (c) 2008, The V3VEE Project <http://www.v3vee.org> 
+ * All rights reserved.
+ *
+ * Author: Jack Lange <jarusl@cs.northwestern.edu>
+ *
+ * This is free software.  You are permitted to use,
+ * redistribute, and modify it as specified in the file "V3VEE_LICENSE".
+ */
+
+
 #include <palacios/svm.h>
 #include <palacios/vmm.h>
 
 #include <palacios/vmm_debug.h>
 #include <palacios/vm_guest_mem.h>
 
-#include <palacios/vmm_emulate.h>
+#include <palacios/vmm_decoder.h>
+#include <palacios/vmm_string.h>
+#include <palacios/vmm_lowlevel.h>
+#include <palacios/svm_msr.h>
 
+#include <palacios/vmm_rbtree.h>
 
-extern struct vmm_os_hooks * os_hooks;
+#include <palacios/vmm_direct_paging.h>
 
-extern uint_t cpuid_ecx(uint_t op);
-extern uint_t cpuid_edx(uint_t op);
-extern void Get_MSR(uint_t MSR, uint_t * high_byte, uint_t * low_byte); 
-extern void Set_MSR(uint_t MSR, uint_t high_byte, uint_t low_byte);
-extern uint_t launch_svm(vmcb_t * vmcb_addr);
-extern void safe_svm_launch(vmcb_t * vmcb_addr, struct v3_gprs * gprs);
+#include <palacios/vmm_ctrl_regs.h>
+#include <palacios/vmm_config.h>
+#include <palacios/svm_io.h>
 
-extern void STGI();
-extern void CLGI();
 
-extern uint_t Get_CR3();
 
+// This is a global pointer to the host's VMCB
+static void * host_vmcb = NULL;
 
-extern void DisableInts();
+extern void v3_stgi();
+extern void v3_clgi();
+//extern int v3_svm_launch(vmcb_t * vmcb, struct v3_gprs * vm_regs, uint64_t * fs, uint64_t * gs);
+extern int v3_svm_launch(vmcb_t * vmcb, struct v3_gprs * vm_regs, vmcb_t * host_vmcb);
 
-/* Checks machine SVM capability */
-/* Implemented from: AMD Arch Manual 3, sect 15.4 */ 
-int is_svm_capable() {
-  uint_t ret =  cpuid_ecx(CPUID_FEATURE_IDS);
-  uint_t vm_cr_low = 0, vm_cr_high = 0;
 
+static vmcb_t * Allocate_VMCB() {
+    vmcb_t * vmcb_page = (vmcb_t *)V3_VAddr(V3_AllocPages(1));
 
-  if ((ret & CPUID_FEATURE_IDS_ecx_svm_avail) == 0) {
-    PrintDebug("SVM Not Available\n");
-    return 0;
-  } 
+    memset(vmcb_page, 0, 4096);
 
-  Get_MSR(SVM_VM_CR_MSR, &vm_cr_high, &vm_cr_low);
+    return vmcb_page;
+}
 
-  if ((ret & CPUID_SVM_REV_AND_FEATURE_IDS_edx_np) == 1) {
-    PrintDebug("Nested Paging not supported\n");
-  }
 
-  if ((vm_cr_low & SVM_VM_CR_MSR_svmdis) == 0) {
-    return 1;
-  }
 
-  ret = cpuid_edx(CPUID_SVM_REV_AND_FEATURE_IDS);
+static void Init_VMCB_BIOS(vmcb_t * vmcb, struct guest_info *vm_info) {
+    vmcb_ctrl_t * ctrl_area = GET_VMCB_CTRL_AREA(vmcb);
+    vmcb_saved_state_t * guest_state = GET_VMCB_SAVE_STATE_AREA(vmcb);
+    uint_t i;
 
-  if ((ret & CPUID_SVM_REV_AND_FEATURE_IDS_edx_svml) == 0) {
-    PrintDebug("SVM BIOS Disabled, not unlockable\n");
-  } else {
-    PrintDebug("SVM is locked with a key\n");
-  }
 
-  return 0;
-}
+    //
+    guest_state->rsp = 0x00;
+    guest_state->rip = 0xfff0;
 
 
+    guest_state->cpl = 0;
 
-void Init_SVM(struct vmm_ctrl_ops * vmm_ops) {
-  reg_ex_t msr;
-  void * host_state;
+    guest_state->efer |= EFER_MSR_svm_enable;
 
 
-  // Enable SVM on the CPU
-  Get_MSR(EFER_MSR, &(msr.e_reg.high), &(msr.e_reg.low));
-  msr.e_reg.low |= EFER_MSR_svm_enable;
-  Set_MSR(EFER_MSR, 0, msr.e_reg.low);
-  
-  PrintDebug("SVM Enabled\n");
+    guest_state->rflags = 0x00000002; // The reserved bit is always 1
+    ctrl_area->svm_instrs.VMRUN = 1;
+    ctrl_area->svm_instrs.VMMCALL = 1;
+    ctrl_area->svm_instrs.VMLOAD = 1;
+    ctrl_area->svm_instrs.VMSAVE = 1;
+    ctrl_area->svm_instrs.STGI = 1;
+    ctrl_area->svm_instrs.CLGI = 1;
+    ctrl_area->svm_instrs.SKINIT = 1;
+    ctrl_area->svm_instrs.RDTSCP = 1;
+    ctrl_area->svm_instrs.ICEBP = 1;
+    ctrl_area->svm_instrs.WBINVD = 1;
+    ctrl_area->svm_instrs.MONITOR = 1;
+    ctrl_area->svm_instrs.MWAIT_always = 1;
+    ctrl_area->svm_instrs.MWAIT_if_armed = 1;
+    ctrl_area->instrs.INVLPGA = 1;
 
 
-  // Setup the host state save area
-  host_state = os_hooks->allocate_pages(4);
+    ctrl_area->instrs.HLT = 1;
+    // guest_state->cr0 = 0x00000001;    // PE 
   
-  msr.e_reg.high = 0;
-  msr.e_reg.low = (uint_t)host_state;
+    /*
+      ctrl_area->exceptions.de = 1;
+      ctrl_area->exceptions.df = 1;
+      
+      ctrl_area->exceptions.ts = 1;
+      ctrl_area->exceptions.ss = 1;
+      ctrl_area->exceptions.ac = 1;
+      ctrl_area->exceptions.mc = 1;
+      ctrl_area->exceptions.gp = 1;
+      ctrl_area->exceptions.ud = 1;
+      ctrl_area->exceptions.np = 1;
+      ctrl_area->exceptions.of = 1;
+      
+      ctrl_area->exceptions.nmi = 1;
+    */
+    
 
+    ctrl_area->instrs.NMI = 1;
+    ctrl_area->instrs.SMI = 1;
+    ctrl_area->instrs.INIT = 1;
+    ctrl_area->instrs.PAUSE = 1;
+    ctrl_area->instrs.shutdown_evts = 1;
 
-  PrintDebug("Host State being saved at %x\n", (uint_t)host_state);
-  Set_MSR(SVM_VM_HSAVE_PA_MSR, msr.e_reg.high, msr.e_reg.low);
+    vm_info->vm_regs.rdx = 0x00000f00;
 
 
+    guest_state->cr0 = 0x60010010; // Set the WP flag so the memory hooks work in real-mode
 
-  // Setup the SVM specific vmm operations
-  vmm_ops->init_guest = &init_svm_guest;
-  vmm_ops->start_guest = &start_svm_guest;
 
+    guest_state->cs.selector = 0xf000;
+    guest_state->cs.limit = 0xffff;
+    guest_state->cs.base = 0x0000000f0000LL;
+    guest_state->cs.attrib.raw = 0xf3;
 
-  return;
-}
 
+    /* DEBUG FOR RETURN CODE */
+    ctrl_area->exit_code = 1;
 
-int init_svm_guest(struct guest_info *info) {
-  PrintDebug("Allocating VMCB\n");
-  info->vmm_data = (void*)Allocate_VMCB();
 
+    struct vmcb_selector *segregs [] = {&(guest_state->ss), &(guest_state->ds), 
+                                       &(guest_state->es), &(guest_state->fs), 
+                                       &(guest_state->gs), NULL};
 
-  //PrintDebug("Generating Guest nested page tables\n");
-  //  info->page_tables = NULL;
-  //info->page_tables = generate_guest_page_tables_64(&(info->mem_layout), &(info->mem_list));
-  //info->page_tables = generate_guest_page_tables(&(info->mem_layout), &(info->mem_list));
-  //  PrintDebugPageTables(info->page_tables);
+    for ( i = 0; segregs[i] != NULL; i++) {
+       struct vmcb_selector * seg = segregs[i];
+       
+       seg->selector = 0x0000;
+       //    seg->base = seg->selector << 4;
+       seg->base = 0x00000000;
+       seg->attrib.raw = 0xf3;
+       seg->limit = ~0u;
+    }
 
+    guest_state->gdtr.limit = 0x0000ffff;
+    guest_state->gdtr.base = 0x0000000000000000LL;
+    guest_state->idtr.limit = 0x0000ffff;
+    guest_state->idtr.base = 0x0000000000000000LL;
 
-  PrintDebug("Initializing VMCB (addr=%x)\n", info->vmm_data);
-  Init_VMCB_BIOS((vmcb_t*)(info->vmm_data), *info);
-  
+    guest_state->ldtr.selector = 0x0000;
+    guest_state->ldtr.limit = 0x0000ffff;
+    guest_state->ldtr.base = 0x0000000000000000LL;
+    guest_state->tr.selector = 0x0000;
+    guest_state->tr.limit = 0x0000ffff;
+    guest_state->tr.base = 0x0000000000000000LL;
 
-  //  info->rip = 0;
 
-  info->vm_regs.rdi = 0;
-  info->vm_regs.rsi = 0;
-  info->vm_regs.rbp = 0;
-  info->vm_regs.rsp = 0;
-  info->vm_regs.rbx = 0;
-  info->vm_regs.rdx = 0;
-  info->vm_regs.rcx = 0;
-  info->vm_regs.rax = 0;
-  
-  return 0;
-}
+    guest_state->dr6 = 0x00000000ffff0ff0LL;
+    guest_state->dr7 = 0x0000000000000400LL;
 
 
-// can we start a kernel thread here...
-int start_svm_guest(struct guest_info *info) {
+    v3_init_svm_io_map(vm_info);
+    ctrl_area->IOPM_BASE_PA = (addr_t)V3_PAddr(vm_info->io_map.arch_data);
+    ctrl_area->instrs.IOIO_PROT = 1;
 
 
 
-  PrintDebug("Launching SVM VM (vmcb=%x)\n", info->vmm_data);
-  //PrintDebugVMCB((vmcb_t*)(info->vmm_data));
+    v3_init_svm_msr_map(vm_info);
+    ctrl_area->MSRPM_BASE_PA = (addr_t)V3_PAddr(vm_info->msr_map.arch_data);
+    ctrl_area->instrs.MSR_PROT = 1;
 
-  while (1) {
 
-    CLGI();
 
-    //PrintDebug("SVM Launch Args (vmcb=%x), (info=%x), (vm_regs=%x)\n", info->vmm_data,  &(info->vm_regs));
-    //PrintDebug("Launching to RIP: %x\n", info->rip);
-    safe_svm_launch((vmcb_t*)(info->vmm_data), &(info->vm_regs));
-    //launch_svm((vmcb_t*)(info->vmm_data));
-    //PrintDebug("SVM Returned\n");
+    PrintDebug("Exiting on interrupts\n");
+    ctrl_area->guest_ctrl.V_INTR_MASKING = 1;
+    ctrl_area->instrs.INTR = 1;
 
-    
 
-    STGI();
+    if (vm_info->shdw_pg_mode == SHADOW_PAGING) {
+       PrintDebug("Creating initial shadow page table\n");
+       
+       /* JRL: This is a performance killer, and a simplistic solution */
+       /* We need to fix this */
+       ctrl_area->TLB_CONTROL = 1;
+       ctrl_area->guest_ASID = 1;
+       
+       
+       if (v3_init_passthrough_pts(vm_info) == -1) {
+           PrintError("Could not initialize passthrough page tables\n");
+           return ;
+       }
+
 
-     
-    if (handle_svm_exit(info) != 0) {
-      vmcb_saved_state_t * guest_state = GET_VMCB_SAVE_STATE_AREA((vmcb_t*)(info->vmm_data));
-      addr_t host_addr;
-      addr_t linear_addr = 0;
+       vm_info->shdw_pg_state.guest_cr0 = 0x0000000000000010LL;
+       PrintDebug("Created\n");
+       
+       guest_state->cr3 = vm_info->direct_map_pt;
 
-      PrintDebug("SVM ERROR!!\n"); 
-      
+       ctrl_area->cr_reads.cr0 = 1;
+       ctrl_area->cr_writes.cr0 = 1;
+       //ctrl_area->cr_reads.cr4 = 1;
+       ctrl_area->cr_writes.cr4 = 1;
+       ctrl_area->cr_reads.cr3 = 1;
+       ctrl_area->cr_writes.cr3 = 1;
 
-      PrintDebug("RIP: %x\n", guest_state->rip);
+       v3_hook_msr(vm_info, EFER_MSR, 
+                   &v3_handle_efer_read,
+                   &v3_handle_efer_write, 
+                   vm_info);
 
+       ctrl_area->instrs.INVLPG = 1;
 
-      linear_addr = get_addr_linear(info, guest_state->rip, &(info->segments.cs));
+       ctrl_area->exceptions.pf = 1;
 
+       guest_state->g_pat = 0x7040600070406ULL;
 
-      PrintDebug("RIP Linear: %x\n", linear_addr);
+       guest_state->cr0 |= 0x80000000;
 
-      guest_pa_to_host_pa(info, linear_addr, &host_addr);
+    } else if (vm_info->shdw_pg_mode == NESTED_PAGING) {
+       // Flush the TLB on entries/exits
+       ctrl_area->TLB_CONTROL = 1;
+       ctrl_area->guest_ASID = 1;
 
-      PrintDebug("Host Address of rip = 0x%x\n", host_addr);
+       // Enable Nested Paging
+       ctrl_area->NP_ENABLE = 1;
 
-      PrintDebug("Instr (15 bytes) at %x:\n", host_addr);
-      PrintTraceMemDump((char*)host_addr, 15);
+       PrintDebug("NP_Enable at 0x%p\n", (void *)&(ctrl_area->NP_ENABLE));
 
-      break;
+       // Set the Nested Page Table pointer
+       if (v3_init_passthrough_pts(vm_info) == -1) {
+           PrintError("Could not initialize Nested page tables\n");
+           return ;
+       }
+
+       ctrl_area->N_CR3 = vm_info->direct_map_pt;
+
+       guest_state->g_pat = 0x7040600070406ULL;
     }
-  }
-  return 0;
 }
 
 
+static int init_svm_guest(struct guest_info * info, struct v3_vm_config * config_ptr) {
+
 
-vmcb_t * Allocate_VMCB() {
-  vmcb_t * vmcb_page = (vmcb_t*)os_hooks->allocate_pages(1);
+    v3_pre_config_guest(info, config_ptr);
 
+    PrintDebug("Allocating VMCB\n");
+    info->vmm_data = (void*)Allocate_VMCB();
 
-  memset(vmcb_page, 0, 4096);
+    PrintDebug("Initializing VMCB (addr=%p)\n", (void *)info->vmm_data);
+    Init_VMCB_BIOS((vmcb_t*)(info->vmm_data), info);
 
-  return vmcb_page;
+    v3_post_config_guest(info, config_ptr);
+
+    return 0;
 }
 
+static int start_svm_guest(struct guest_info *info) {
+    //    vmcb_saved_state_t * guest_state = GET_VMCB_SAVE_STATE_AREA((vmcb_t*)(info->vmm_data));
+    //  vmcb_ctrl_t * guest_ctrl = GET_VMCB_CTRL_AREA((vmcb_t*)(info->vmm_data));
+    uint_t num_exits = 0;
 
 
-void Init_VMCB(vmcb_t * vmcb, struct guest_info vm_info) {
-  vmcb_ctrl_t * ctrl_area = GET_VMCB_CTRL_AREA(vmcb);
-  vmcb_saved_state_t * guest_state = GET_VMCB_SAVE_STATE_AREA(vmcb);
-  uint_t i;
 
+    PrintDebug("Launching SVM VM (vmcb=%p)\n", (void *)info->vmm_data);
+    //PrintDebugVMCB((vmcb_t*)(info->vmm_data));
+    
+    info->run_state = VM_RUNNING;
+    rdtscll(info->yield_start_cycle);
 
-  guest_state->rsp = vm_info.vm_regs.rsp;
-  guest_state->rip = vm_info.rip;
 
+    while (1) {
+       ullong_t tmp_tsc;
+       
+       // Conditionally yield the CPU if the timeslice has expired
+       v3_yield_cond(info);
 
-  //ctrl_area->instrs.instrs.CR0 = 1;
-  ctrl_area->cr_reads.cr0 = 1;
-  ctrl_area->cr_writes.cr0 = 1;
+       /*
+         PrintDebug("SVM Entry to CS=%p  rip=%p...\n", 
+         (void *)(addr_t)info->segments.cs.base, 
+         (void *)(addr_t)info->rip);
+       */
 
-  guest_state->efer |= EFER_MSR_svm_enable;
-  guest_state->rflags = 0x00000002; // The reserved bit is always 1
-  ctrl_area->svm_instrs.VMRUN = 1;
-  // guest_state->cr0 = 0x00000001;    // PE 
-  ctrl_area->guest_ASID = 1;
+       // disable global interrupts for vm state transition
+       v3_clgi();
 
 
-  ctrl_area->exceptions.de = 1;
-  ctrl_area->exceptions.df = 1;
-  ctrl_area->exceptions.pf = 1;
-  ctrl_area->exceptions.ts = 1;
-  ctrl_area->exceptions.ss = 1;
-  ctrl_area->exceptions.ac = 1;
-  ctrl_area->exceptions.mc = 1;
-  ctrl_area->exceptions.gp = 1;
-  ctrl_area->exceptions.ud = 1;
-  ctrl_area->exceptions.np = 1;
-  ctrl_area->exceptions.of = 1;
-  ctrl_area->exceptions.nmi = 1;
 
-  guest_state->cs.selector = 0x0000;
-  guest_state->cs.limit=~0u;
-  guest_state->cs.base = guest_state->cs.selector<<4;
-  guest_state->cs.attrib.raw = 0xf3;
+       rdtscll(info->time_state.cached_host_tsc);
+       //    guest_ctrl->TSC_OFFSET = info->time_state.guest_tsc - info->time_state.cached_host_tsc;
+       
+       v3_svm_launch((vmcb_t*)V3_PAddr(info->vmm_data), &(info->vm_regs), (vmcb_t *)host_vmcb);
+       
+       rdtscll(tmp_tsc);
+
+       
+       //PrintDebug("SVM Returned\n");
+
+       // reenable global interrupts after vm exit
+       v3_stgi();
 
-  
-  struct vmcb_selector *segregs [] = {&(guest_state->ss), &(guest_state->ds), &(guest_state->es), &(guest_state->fs), &(guest_state->gs), NULL};
-  for ( i = 0; segregs[i] != NULL; i++) {
-    struct vmcb_selector * seg = segregs[i];
-    
-    seg->selector = 0x0000;
-    seg->base = seg->selector << 4;
-    seg->attrib.raw = 0xf3;
-    seg->limit = ~0u;
-  }
-  
-  if (vm_info.io_map.num_ports > 0) {
-    vmm_io_hook_t * iter;
-    addr_t io_port_bitmap;
-    
-    io_port_bitmap = (addr_t)os_hooks->allocate_pages(3);
-    memset((uchar_t*)io_port_bitmap, 0, PAGE_SIZE * 3);
-    
-    ctrl_area->IOPM_BASE_PA = io_port_bitmap;
 
-    //PrintDebug("Setting up IO Map at 0x%x\n", io_port_bitmap);
+       // Conditionally yield the CPU if the timeslice has expired
+       v3_yield_cond(info);
 
-    FOREACH_IO_HOOK(vm_info.io_map, iter) {
-      ushort_t port = iter->port;
-      uchar_t * bitmap = (uchar_t *)io_port_bitmap;
 
-      bitmap += (port / 8);
-      PrintDebug("Setting Bit in block %x\n", bitmap);
-      *bitmap |= 1 << (port % 8);
+       v3_update_time(info, tmp_tsc - info->time_state.cached_host_tsc);
+       num_exits++;
+       
+       if ((num_exits % 5000) == 0) {
+           PrintDebug("SVM Exit number %d\n", num_exits);
+       }
+
+       if (v3_handle_svm_exit(info) != 0) {
+           vmcb_ctrl_t * guest_ctrl = GET_VMCB_CTRL_AREA((vmcb_t*)(info->vmm_data));
+           addr_t host_addr;
+           addr_t linear_addr = 0;
+           
+           info->run_state = VM_ERROR;
+           
+           PrintDebug("SVM ERROR!!\n"); 
+      
+           v3_print_guest_state(info);
+
+           PrintDebug("SVM Exit Code: %p\n", (void *)(addr_t)guest_ctrl->exit_code); 
+      
+           PrintDebug("exit_info1 low = 0x%.8x\n", *(uint_t*)&(guest_ctrl->exit_info1));
+           PrintDebug("exit_info1 high = 0x%.8x\n", *(uint_t *)(((uchar_t *)&(guest_ctrl->exit_info1)) + 4));
+      
+           PrintDebug("exit_info2 low = 0x%.8x\n", *(uint_t*)&(guest_ctrl->exit_info2));
+           PrintDebug("exit_info2 high = 0x%.8x\n", *(uint_t *)(((uchar_t *)&(guest_ctrl->exit_info2)) + 4));
+      
+           linear_addr = get_addr_linear(info, info->rip, &(info->segments.cs));
+
+           if (info->mem_mode == PHYSICAL_MEM) {
+               guest_pa_to_host_va(info, linear_addr, &host_addr);
+           } else if (info->mem_mode == VIRTUAL_MEM) {
+               guest_va_to_host_va(info, linear_addr, &host_addr);
+           }
+
+           PrintDebug("Host Address of rip = 0x%p\n", (void *)host_addr);
+
+           PrintDebug("Instr (15 bytes) at %p:\n", (void *)host_addr);
+           v3_dump_mem((uint8_t *)host_addr, 15);
+
+           break;
+       }
     }
+    return 0;
+}
 
 
-    //PrintDebugMemDump((uchar_t*)io_port_bitmap, PAGE_SIZE *2);
 
-    ctrl_area->instrs.IOIO_PROT = 1;
-  }
 
-  ctrl_area->instrs.INTR = 1;
 
+/* Checks machine SVM capability */
+/* Implemented from: AMD Arch Manual 3, sect 15.4 */ 
+int v3_is_svm_capable() {
+    // Dinda
+    uint_t vm_cr_low = 0, vm_cr_high = 0;
+    addr_t eax = 0, ebx = 0, ecx = 0, edx = 0;
 
+    v3_cpuid(CPUID_EXT_FEATURE_IDS, &eax, &ebx, &ecx, &edx);
+  
+    PrintDebug("CPUID_EXT_FEATURE_IDS_ecx=%p\n", (void *)ecx);
 
-  if (vm_info.page_mode == SHADOW_PAGING) {
-    PrintDebug("Creating initial shadow page table\n");
-    vm_info.shdw_pg_state.shadow_cr3 |= ((addr_t)create_passthrough_pde32_pts(&vm_info) & ~0xfff);
-    PrintDebug("Created\n");
+    if ((ecx & CPUID_EXT_FEATURE_IDS_ecx_svm_avail) == 0) {
+      PrintDebug("SVM Not Available\n");
+      return 0;
+    }  else {
+       v3_get_msr(SVM_VM_CR_MSR, &vm_cr_high, &vm_cr_low);
+       
+       PrintDebug("SVM_VM_CR_MSR = 0x%x 0x%x\n", vm_cr_high, vm_cr_low);
+       
+       if ((vm_cr_low & SVM_VM_CR_MSR_svmdis) == 1) {
+           PrintDebug("SVM is available but is disabled.\n");
+           
+           v3_cpuid(CPUID_SVM_REV_AND_FEATURE_IDS, &eax, &ebx, &ecx, &edx);
+           
+           PrintDebug("CPUID_SVM_REV_AND_FEATURE_IDS_edx=%p\n", (void *)edx);
+           
+           if ((edx & CPUID_SVM_REV_AND_FEATURE_IDS_edx_svml) == 0) {
+               PrintDebug("SVM BIOS Disabled, not unlockable\n");
+           } else {
+               PrintDebug("SVM is locked with a key\n");
+           }
+           return 0;
+
+       } else {
+           PrintDebug("SVM is available and  enabled.\n");
+
+           v3_cpuid(CPUID_SVM_REV_AND_FEATURE_IDS, &eax, &ebx, &ecx, &edx);
+           PrintDebug("CPUID_SVM_REV_AND_FEATURE_IDS_eax=%p\n", (void *)eax);
+           PrintDebug("CPUID_SVM_REV_AND_FEATURE_IDS_ebx=%p\n", (void *)ebx);
+           PrintDebug("CPUID_SVM_REV_AND_FEATURE_IDS_ecx=%p\n", (void *)ecx);
+           PrintDebug("CPUID_SVM_REV_AND_FEATURE_IDS_edx=%p\n", (void *)edx);
+
+
+           if ((edx & CPUID_SVM_REV_AND_FEATURE_IDS_edx_np) == 0) {
+               PrintDebug("SVM Nested Paging not supported\n");
+           } else {
+               PrintDebug("SVM Nested Paging supported\n");
+           }
+
+           return 1;
+       }
+    }
+}
 
-    guest_state->cr3 = vm_info.shdw_pg_state.shadow_cr3;
+static int has_svm_nested_paging() {
+    addr_t eax = 0, ebx = 0, ecx = 0, edx = 0;
 
-    ctrl_area->cr_reads.cr3 = 1;
-    ctrl_area->cr_writes.cr3 = 1;
+    v3_cpuid(CPUID_SVM_REV_AND_FEATURE_IDS, &eax, &ebx, &ecx, &edx);
 
+    //PrintDebug("CPUID_EXT_FEATURE_IDS_edx=0x%x\n", edx);
 
-    ctrl_area->instrs.INVLPG = 1;
-    ctrl_area->instrs.INVLPGA = 1;
+    if ((edx & CPUID_SVM_REV_AND_FEATURE_IDS_edx_np) == 0) {
+       PrintDebug("SVM Nested Paging not supported\n");
+       return 0;
+    } else {
+       PrintDebug("SVM Nested Paging supported\n");
+       return 1;
+    }
+}
 
-    guest_state->g_pat = 0x7040600070406ULL;
 
-    guest_state->cr0 |= 0x80000000;
-  } else if (vm_info.page_mode == NESTED_PAGING) {
-    // Flush the TLB on entries/exits
-    //ctrl_area->TLB_CONTROL = 1;
 
-    // Enable Nested Paging
-    //ctrl_area->NP_ENABLE = 1;
+void v3_init_SVM(struct v3_ctrl_ops * vmm_ops) {
+    reg_ex_t msr;
+    extern v3_cpu_arch_t v3_cpu_type;
 
-    //PrintDebug("NP_Enable at 0x%x\n", &(ctrl_area->NP_ENABLE));
+    // Enable SVM on the CPU
+    v3_get_msr(EFER_MSR, &(msr.e_reg.high), &(msr.e_reg.low));
+    msr.e_reg.low |= EFER_MSR_svm_enable;
+    v3_set_msr(EFER_MSR, 0, msr.e_reg.low);
 
-        // Set the Nested Page Table pointer
-    //    ctrl_area->N_CR3 = ((addr_t)vm_info.page_tables);
-    // ctrl_area->N_CR3 = (addr_t)(vm_info.page_tables);
+    PrintDebug("SVM Enabled\n");
 
-    //   ctrl_area->N_CR3 = Get_CR3();
-    // guest_state->cr3 |= (Get_CR3() & 0xfffff000);
+    // Setup the host state save area
+    host_vmcb = V3_AllocPages(4);
 
-    //    guest_state->g_pat = 0x7040600070406ULL;
-  }
+    /* 64-BIT-ISSUE */
+    //  msr.e_reg.high = 0;
+    //msr.e_reg.low = (uint_t)host_vmcb;
+    msr.r_reg = (addr_t)host_vmcb;
+
+    PrintDebug("Host State being saved at %p\n", (void *)(addr_t)host_vmcb);
+    v3_set_msr(SVM_VM_HSAVE_PA_MSR, msr.e_reg.high, msr.e_reg.low);
 
 
 
+
+    if (has_svm_nested_paging() == 1) {
+       v3_cpu_type = V3_SVM_REV3_CPU;
+    } else {
+       v3_cpu_type = V3_SVM_CPU;
+    }
+
+    // Setup the SVM specific vmm operations
+    vmm_ops->init_guest = &init_svm_guest;
+    vmm_ops->start_guest = &start_svm_guest;
+    vmm_ops->has_nested_paging = &has_svm_nested_paging;
+
+    return;
 }
 
 
 
-void Init_VMCB_BIOS(vmcb_t * vmcb, struct guest_info vm_info) {
-  vmcb_ctrl_t * ctrl_area = GET_VMCB_CTRL_AREA(vmcb);
-  vmcb_saved_state_t * guest_state = GET_VMCB_SAVE_STATE_AREA(vmcb);
-  uint_t i;
 
 
-  guest_state->rsp = vm_info.vm_regs.rsp;
-  // guest_state->rip = vm_info.rip;
-  guest_state->rip = 0xfff0;
 
-  guest_state->cpl = 0;
 
-  //ctrl_area->instrs.instrs.CR0 = 1;
-  ctrl_area->cr_reads.cr0 = 1;
-  ctrl_area->cr_writes.cr0 = 1;
 
-  guest_state->efer |= EFER_MSR_svm_enable;
-  guest_state->rflags = 0x00000002; // The reserved bit is always 1
-  ctrl_area->svm_instrs.VMRUN = 1;
-  ctrl_area->instrs.HLT = 1;
-  // guest_state->cr0 = 0x00000001;    // PE 
-  ctrl_area->guest_ASID = 1;
 
-  ctrl_area->exceptions.de = 1;
-  ctrl_area->exceptions.df = 1;
-  ctrl_area->exceptions.pf = 1;
-  ctrl_area->exceptions.ts = 1;
-  ctrl_area->exceptions.ss = 1;
-  ctrl_area->exceptions.ac = 1;
-  ctrl_area->exceptions.mc = 1;
-  ctrl_area->exceptions.gp = 1;
-  ctrl_area->exceptions.ud = 1;
-  ctrl_area->exceptions.np = 1;
-  ctrl_area->exceptions.of = 1;
-  ctrl_area->exceptions.nmi = 1;
 
-  vm_info.vm_regs.rdx = 0x00000f00;
 
-  guest_state->cr0 = 0x60000010;
 
-  guest_state->cs.selector = 0xf000;
-  guest_state->cs.limit=0xffff;
-  guest_state->cs.base = 0x0000000f0000LL;
-  guest_state->cs.attrib.raw = 0xf3;
 
-  
-  struct vmcb_selector *segregs [] = {&(guest_state->ss), &(guest_state->ds), &(guest_state->es), &(guest_state->fs), &(guest_state->gs), NULL};
-  for ( i = 0; segregs[i] != NULL; i++) {
-    struct vmcb_selector * seg = segregs[i];
-    
-    seg->selector = 0x0000;
-    //    seg->base = seg->selector << 4;
-    seg->base = 0x00000000;
-    seg->attrib.raw = 0xf3;
-    seg->limit = ~0u;
-  }
-  
-  guest_state->gdtr.limit = 0x0000ffff;
-  guest_state->gdtr.base = 0x0000000000000000LL;
-  guest_state->idtr.limit = 0x0000ffff;
-  guest_state->idtr.base = 0x0000000000000000LL;
 
-  guest_state->ldtr.selector = 0x0000;
-  guest_state->ldtr.limit = 0x0000ffff;
-  guest_state->ldtr.base = 0x0000000000000000LL;
-  guest_state->tr.selector = 0x0000;
-  guest_state->tr.limit = 0x0000ffff;
-  guest_state->tr.base = 0x0000000000000000LL;
 
 
-  guest_state->dr6 = 0x00000000ffff0ff0LL;
-  guest_state->dr7 = 0x0000000000000400LL;
 
-  if (vm_info.io_map.num_ports > 0) {
-    vmm_io_hook_t * iter;
-    addr_t io_port_bitmap;
-    
-    io_port_bitmap = (addr_t)os_hooks->allocate_pages(3);
-    memset((uchar_t*)io_port_bitmap, 0, PAGE_SIZE * 3);
-    
-    ctrl_area->IOPM_BASE_PA = io_port_bitmap;
 
-    //PrintDebug("Setting up IO Map at 0x%x\n", io_port_bitmap);
 
-    FOREACH_IO_HOOK(vm_info.io_map, iter) {
-      ushort_t port = iter->port;
-      uchar_t * bitmap = (uchar_t *)io_port_bitmap;
 
-      bitmap += (port / 8);
-      PrintDebug("Setting Bit for port 0x%x\n", port);
-      *bitmap |= 1 << (port % 8);
-    }
 
 
-    //PrintDebugMemDump((uchar_t*)io_port_bitmap, PAGE_SIZE *2);
 
-    ctrl_area->instrs.IOIO_PROT = 1;
-  }
 
 
 
-  PrintDebug("Exiting on interrupts\n");
-  ctrl_area->guest_ctrl.V_INTR_MASKING = 1;
-  ctrl_area->instrs.INTR = 1;
 
 
-  if (vm_info.page_mode == SHADOW_PAGING) {
-    PrintDebug("Creating initial shadow page table\n");
-    vm_info.shdw_pg_state.shadow_cr3 |= ((addr_t)create_passthrough_pde32_pts(&vm_info) & ~0xfff);
-    PrintDebug("Created\n");
 
-    guest_state->cr3 = vm_info.shdw_pg_state.shadow_cr3;
 
-    //PrintDebugPageTables((pde32_t*)(vm_info.shdw_pg_state.shadow_cr3.e_reg.low));
 
-    ctrl_area->cr_reads.cr3 = 1;
-    ctrl_area->cr_writes.cr3 = 1;
 
 
-    ctrl_area->instrs.INVLPG = 1;
-    ctrl_area->instrs.INVLPGA = 1;
 
-    guest_state->g_pat = 0x7040600070406ULL;
 
-    guest_state->cr0 |= 0x80000000;
-  } else if (vm_info.page_mode == NESTED_PAGING) {
-    // Flush the TLB on entries/exits
-    //ctrl_area->TLB_CONTROL = 1;
 
-    // Enable Nested Paging
-    //ctrl_area->NP_ENABLE = 1;
 
-    //PrintDebug("NP_Enable at 0x%x\n", &(ctrl_area->NP_ENABLE));
 
-        // Set the Nested Page Table pointer
-    //    ctrl_area->N_CR3 = ((addr_t)vm_info.page_tables);
-    // ctrl_area->N_CR3 = (addr_t)(vm_info.page_tables);
 
-    //   ctrl_area->N_CR3 = Get_CR3();
-    // guest_state->cr3 |= (Get_CR3() & 0xfffff000);
 
-    //    guest_state->g_pat = 0x7040600070406ULL;
-  }
 
 
 
-}
+
+
+
+
+
+
+
+
+
+#if 0
+/* 
+ * Test VMSAVE/VMLOAD Latency 
+ */
+#define vmsave ".byte 0x0F,0x01,0xDB ; "
+#define vmload ".byte 0x0F,0x01,0xDA ; "
+{
+    uint32_t start_lo, start_hi;
+    uint32_t end_lo, end_hi;
+    uint64_t start, end;
+    
+    __asm__ __volatile__ (
+                         "rdtsc ; "
+                         "movl %%eax, %%esi ; "
+                         "movl %%edx, %%edi ; "
+                         "movq  %%rcx, %%rax ; "
+                         vmsave
+                         "rdtsc ; "
+                         : "=D"(start_hi), "=S"(start_lo), "=a"(end_lo),"=d"(end_hi)
+                         : "c"(host_vmcb), "0"(0), "1"(0), "2"(0), "3"(0)
+                         );
+    
+    start = start_hi;
+    start <<= 32;
+    start += start_lo;
+    
+    end = end_hi;
+    end <<= 32;
+    end += end_lo;
+    
+    PrintDebug("VMSave Cycle Latency: %d\n", (uint32_t)(end - start));
+    
+    __asm__ __volatile__ (
+                         "rdtsc ; "
+                         "movl %%eax, %%esi ; "
+                         "movl %%edx, %%edi ; "
+                         "movq  %%rcx, %%rax ; "
+                         vmload
+                         "rdtsc ; "
+                         : "=D"(start_hi), "=S"(start_lo), "=a"(end_lo),"=d"(end_hi)
+                             : "c"(host_vmcb), "0"(0), "1"(0), "2"(0), "3"(0)
+                             );
+       
+       start = start_hi;
+       start <<= 32;
+       start += start_lo;
+
+       end = end_hi;
+       end <<= 32;
+       end += end_lo;
+
+
+       PrintDebug("VMLoad Cycle Latency: %d\n", (uint32_t)(end - start));
+    }
+    /* End Latency Test */
+
+#endif
+
+
+
+
+
 
 
 #if 0
@@ -551,7 +674,7 @@ void Init_VMCB_pe(vmcb_t *vmcb, struct guest_info vm_info) {
   
 
   ctrl_area->instrs.IOIO_PROT = 1;
-  ctrl_area->IOPM_BASE_PA = (uint_t)os_hooks->allocate_pages(3);
+  ctrl_area->IOPM_BASE_PA = (uint_t)V3_AllocPages(3);
   
   {
     reg_ex_t tmp_reg;
@@ -633,157 +756,3 @@ void Init_VMCB_pe(vmcb_t *vmcb, struct guest_info vm_info) {
 #endif
 
 
-
-
-
-
-
-
-
-
-
-
-
-
-/*
-
-
-void Init_VMCB_Real(vmcb_t * vmcb, struct guest_info vm_info) {
-  vmcb_ctrl_t * ctrl_area = GET_VMCB_CTRL_AREA(vmcb);
-  vmcb_saved_state_t * guest_state = GET_VMCB_SAVE_STATE_AREA(vmcb);
-  uint_t i;
-
-
-  guest_state->rsp = vm_info.vm_regs.rsp;
-  guest_state->rip = vm_info.rip;
-
-
-  guest_state->efer |= EFER_MSR_svm_enable;
-  guest_state->rflags = 0x00000002; // The reserved bit is always 1
-  ctrl_area->svm_instrs.instrs.VMRUN = 1;
-  ctrl_area->guest_ASID = 1;
-  guest_state->cr0 = 0x60000010;
-
-
-  ctrl_area->exceptions.de = 1;
-  ctrl_area->exceptions.df = 1;
-  ctrl_area->exceptions.pf = 1;
-  ctrl_area->exceptions.ts = 1;
-  ctrl_area->exceptions.ss = 1;
-  ctrl_area->exceptions.ac = 1;
-  ctrl_area->exceptions.mc = 1;
-  ctrl_area->exceptions.gp = 1;
-  ctrl_area->exceptions.ud = 1;
-  ctrl_area->exceptions.np = 1;
-  ctrl_area->exceptions.of = 1;
-  ctrl_area->exceptions.nmi = 1;
-
-  guest_state->cs.selector = 0xf000;
-  guest_state->cs.limit=0xffff;
-  guest_state->cs.base =  0xffff0000;
-  guest_state->cs.attrib.raw = 0x9a;
-
-  
-  struct vmcb_selector *segregs [] = {&(guest_state->ss), &(guest_state->ds), &(guest_state->es), &(guest_state->fs), &(guest_state->gs), NULL};
-  for ( i = 0; segregs[i] != NULL; i++) {
-    struct vmcb_selector * seg = segregs[i];
-    
-    seg->selector = 0x0000;
-    seg->base = 0xffff0000;
-    seg->attrib.raw = 0x9b;
-    seg->limit = 0xffff;
-  }
-  
-  // Set GPRs 
-  //
-  //  EDX == 0xfxx
-  //  EAX, EBX, ECX, ESI, EDI, EBP, ESP == 0x0
-  //
-
-  guest_state->gdtr.base = 0;
-  guest_state->gdtr.limit = 0xffff;
-  guest_state->gdtr.attrib.raw = 0x0;
-
-  guest_state->idtr.base = 0;
-  guest_state->idtr.limit = 0xffff;
-  guest_state->idtr.attrib.raw = 0x0;
-
-  guest_state->ldtr.base = 0;
-  guest_state->ldtr.limit = 0xffff;
-  guest_state->ldtr.attrib.raw = 0x82;
-
-  guest_state->tr.base = 0;
-  guest_state->tr.limit = 0xffff;
-  guest_state->tr.attrib.raw = 0x83;
-
-
-
-
-  if (vm_info.io_map.num_ports > 0) {
-    vmm_io_hook_t * iter;
-    addr_t io_port_bitmap;
-    
-    io_port_bitmap = (addr_t)os_hooks->allocate_pages(3);
-    memset((uchar_t*)io_port_bitmap, 0, PAGE_SIZE * 3);
-    
-    ctrl_area->IOPM_BASE_PA = io_port_bitmap;
-
-    //PrintDebug("Setting up IO Map at 0x%x\n", io_port_bitmap);
-
-    FOREACH_IO_HOOK(vm_info.io_map, iter) {
-      ushort_t port = iter->port;
-      uchar_t * bitmap = (uchar_t *)io_port_bitmap;
-
-      bitmap += (port / 8);
-      PrintDebug("Setting Bit in block %x\n", bitmap);
-      *bitmap |= 1 << (port % 8);
-    }
-
-    ctrl_area->instrs.instrs.IOIO_PROT = 1;
-  }
-
-  ctrl_area->instrs.instrs.INTR = 1;
-
-  // also determine if CPU supports nested paging
-
-  if (vm_info.page_mode == SHADOW_PAGING) {
-    PrintDebug("Creating initial shadow page table\n");
-    vm_info.shdw_pg_state.shadow_cr3.e_reg.low |= ((addr_t)create_passthrough_pde32_pts(&vm_info) & ~0xfff);
-    PrintDebug("Created\n");
-
-    guest_state->cr3 = vm_info.shdw_pg_state.shadow_cr3.r_reg;
-
-    ctrl_area->cr_reads.crs.cr3 = 1;
-    ctrl_area->cr_writes.crs.cr3 = 1;
-    ctrl_area->cr_reads.crs.cr0 = 1;
-    ctrl_area->cr_writes.crs.cr0 = 1;
-
-    ctrl_area->instrs.instrs.INVLPG = 1;
-    ctrl_area->instrs.instrs.INVLPGA = 1;
-
-       
-    guest_state->g_pat = 0x7040600070406ULL;
-
-    vm_info.shdw_pg_state.guest_cr0.e_reg.low = guest_state->cr0;
-    guest_state->cr0 |= 0x80000000;
-  } else if (vm_info.page_mode == NESTED_PAGING) {
-    // Flush the TLB on entries/exits
-    //ctrl_area->TLB_CONTROL = 1;
-
-    // Enable Nested Paging
-    //ctrl_area->NP_ENABLE = 1;
-
-    //PrintDebug("NP_Enable at 0x%x\n", &(ctrl_area->NP_ENABLE));
-
-        // Set the Nested Page Table pointer
-    //    ctrl_area->N_CR3 = ((addr_t)vm_info.page_tables);
-    // ctrl_area->N_CR3 = (addr_t)(vm_info.page_tables);
-
-    //   ctrl_area->N_CR3 = Get_CR3();
-    // guest_state->cr3 |= (Get_CR3() & 0xfffff000);
-
-    //    guest_state->g_pat = 0x7040600070406ULL;
-  }
-
-}
-*/