vmcs.c

来自「xen虚拟机源代码安装包」· C语言 代码 · 共 1,066 行 · 第 1/3 页

    return -ESRCH;
}

int vmx_add_guest_msr(u32 msr)
{
    struct vcpu *curr = current;
    unsigned int i, msr_count = curr->arch.hvm_vmx.msr_count;
    struct vmx_msr_entry *msr_area = curr->arch.hvm_vmx.msr_area;

    if ( msr_area == NULL )
    {
        if ( (msr_area = alloc_xenheap_page()) == NULL )
            return -ENOMEM;
        curr->arch.hvm_vmx.msr_area = msr_area;
        __vmwrite(VM_EXIT_MSR_STORE_ADDR, virt_to_maddr(msr_area));
        __vmwrite(VM_ENTRY_MSR_LOAD_ADDR, virt_to_maddr(msr_area));
    }

    for ( i = 0; i < msr_count; i++ )
        if ( msr_area[i].index == msr )
            return 0;

    if ( msr_count == (PAGE_SIZE / sizeof(struct vmx_msr_entry)) )
        return -ENOSPC;

    msr_area[msr_count].index = msr;
    msr_area[msr_count].mbz   = 0;
    msr_area[msr_count].data  = 0;
    curr->arch.hvm_vmx.msr_count = ++msr_count;
    __vmwrite(VM_EXIT_MSR_STORE_COUNT, msr_count);
    __vmwrite(VM_ENTRY_MSR_LOAD_COUNT, msr_count);

    return 0;
}

int vmx_add_host_load_msr(u32 msr)
{
    struct vcpu *curr = current;
    unsigned int i, msr_count = curr->arch.hvm_vmx.host_msr_count;
    struct vmx_msr_entry *msr_area = curr->arch.hvm_vmx.host_msr_area;

    if ( msr_area == NULL )
    {
        if ( (msr_area = alloc_xenheap_page()) == NULL )
            return -ENOMEM;
        curr->arch.hvm_vmx.host_msr_area = msr_area;
        __vmwrite(VM_EXIT_MSR_LOAD_ADDR, virt_to_maddr(msr_area));
    }

    for ( i = 0; i < msr_count; i++ )
        if ( msr_area[i].index == msr )
            return 0;

    if ( msr_count == (PAGE_SIZE / sizeof(struct vmx_msr_entry)) )
        return -ENOSPC;

    msr_area[msr_count].index = msr;
    msr_area[msr_count].mbz   = 0;
    rdmsrl(msr, msr_area[msr_count].data);
    curr->arch.hvm_vmx.host_msr_count = ++msr_count;
    __vmwrite(VM_EXIT_MSR_LOAD_COUNT, msr_count);

    return 0;
}

int vmx_create_vmcs(struct vcpu *v)
{
    struct arch_vmx_struct *arch_vmx = &v->arch.hvm_vmx;
    int rc;

    if ( (arch_vmx->vmcs = vmx_alloc_vmcs()) == NULL )
        return -ENOMEM;

    INIT_LIST_HEAD(&arch_vmx->active_list);
    __vmpclear(virt_to_maddr(arch_vmx->vmcs));
    arch_vmx->active_cpu = -1;
    arch_vmx->launched   = 0;

    if ( (rc = construct_vmcs(v)) != 0 )
    {
        vmx_free_vmcs(arch_vmx->vmcs);
        return rc;
    }

    return 0;
}

void vmx_destroy_vmcs(struct vcpu *v)
{
    struct arch_vmx_struct *arch_vmx = &v->arch.hvm_vmx;

    vmx_clear_vmcs(v);

    vmx_free_vmcs(arch_vmx->vmcs);

    free_xenheap_page(v->arch.hvm_vmx.host_msr_area);
    free_xenheap_page(v->arch.hvm_vmx.msr_area);
    free_xenheap_page(v->arch.hvm_vmx.msr_bitmap);
}

void vm_launch_fail(void)
{
    unsigned long error = __vmread(VM_INSTRUCTION_ERROR);
    printk("<vm_launch_fail> error code %lx\n", error);
    domain_crash_synchronous();
}

void vm_resume_fail(void)
{
    unsigned long error = __vmread(VM_INSTRUCTION_ERROR);
    printk("<vm_resume_fail> error code %lx\n", error);
    domain_crash_synchronous();
}

static void wbinvd_ipi(void *info)
{
    wbinvd();
}

void vmx_do_resume(struct vcpu *v)
{
    bool_t debug_state;

    if ( v->arch.hvm_vmx.active_cpu == smp_processor_id() )
    {
        if ( v->arch.hvm_vmx.vmcs != this_cpu(current_vmcs) )
            vmx_load_vmcs(v);
    }
    else
    {
        /*
         * For pass-through domain, guest PCI-E device driver may leverage the
         * "Non-Snoop" I/O, and explicitly WBINVD or CLFLUSH to a RAM space.
         * Since migration may occur before WBINVD or CLFLUSH, we need to
         * maintain data consistency either by:
         *  1: flushing cache (wbinvd) when the guest is scheduled out if
         *     there is no wbinvd exit, or
         *  2: execute wbinvd on all dirty pCPUs when guest wbinvd exits.
         */
        if ( has_arch_pdevs(v->domain) && !cpu_has_wbinvd_exiting )
        {
            int cpu = v->arch.hvm_vmx.active_cpu;
            if ( cpu != -1 )
                on_selected_cpus(cpumask_of_cpu(cpu), wbinvd_ipi, NULL, 1, 1);
        }

        vmx_clear_vmcs(v);
        vmx_load_vmcs(v);
        hvm_migrate_timers(v);
        vmx_set_host_env(v);
        vpid_sync_vcpu_all(v);
    }

    debug_state = v->domain->debugger_attached;
    if ( unlikely(v->arch.hvm_vcpu.debug_state_latch != debug_state) )
    {
        unsigned long intercepts = __vmread(EXCEPTION_BITMAP);
        unsigned long mask = (1U << TRAP_debug) | (1U << TRAP_int3);
        v->arch.hvm_vcpu.debug_state_latch = debug_state;
        if ( debug_state )
            intercepts |= mask;
        else
            intercepts &= ~mask;
        __vmwrite(EXCEPTION_BITMAP, intercepts);
    }

    hvm_do_resume(v);
    reset_stack_and_jump(vmx_asm_do_vmentry);
}

static void vmx_dump_sel(char *name, enum x86_segment seg)
{
    struct segment_register sreg;
    hvm_get_segment_register(current, seg, &sreg);
    printk("%s: sel=0x%04x, attr=0x%05x, limit=0x%08x, base=0x%016llx\n", 
           name, sreg.sel, sreg.attr.bytes, sreg.limit,
           (unsigned long long)sreg.base);
}

static unsigned long vmr(unsigned long field)
{
    int rc;
    unsigned long val;
    val = __vmread_safe(field, &rc);
    return rc ? 0 : val;
}

void vmcs_dump_vcpu(struct vcpu *v)
{
    struct cpu_user_regs *regs = &v->arch.guest_context.user_regs;
    unsigned long long x;

    if ( v == current )
        regs = guest_cpu_user_regs();

    vmx_vmcs_enter(v);

    printk("*** Guest State ***\n");
    printk("CR0: actual=0x%016llx, shadow=0x%016llx, gh_mask=%016llx\n",
           (unsigned long long)vmr(GUEST_CR0),
           (unsigned long long)vmr(CR0_READ_SHADOW), 
           (unsigned long long)vmr(CR0_GUEST_HOST_MASK));
    printk("CR4: actual=0x%016llx, shadow=0x%016llx, gh_mask=%016llx\n",
           (unsigned long long)vmr(GUEST_CR4),
           (unsigned long long)vmr(CR4_READ_SHADOW), 
           (unsigned long long)vmr(CR4_GUEST_HOST_MASK));
    printk("CR3: actual=0x%016llx, target_count=%d\n",
           (unsigned long long)vmr(GUEST_CR3),
           (int)vmr(CR3_TARGET_COUNT));
    printk("     target0=%016llx, target1=%016llx\n",
           (unsigned long long)vmr(CR3_TARGET_VALUE0),
           (unsigned long long)vmr(CR3_TARGET_VALUE1));
    printk("     target2=%016llx, target3=%016llx\n",
           (unsigned long long)vmr(CR3_TARGET_VALUE2),
           (unsigned long long)vmr(CR3_TARGET_VALUE3));
    printk("RSP = 0x%016llx (0x%016llx)  RIP = 0x%016llx (0x%016llx)\n", 
           (unsigned long long)vmr(GUEST_RSP),
           (unsigned long long)regs->esp,
           (unsigned long long)vmr(GUEST_RIP),
           (unsigned long long)regs->eip);
    printk("RFLAGS=0x%016llx (0x%016llx)  DR7 = 0x%016llx\n", 
           (unsigned long long)vmr(GUEST_RFLAGS),
           (unsigned long long)regs->eflags,
           (unsigned long long)vmr(GUEST_DR7));
    printk("Sysenter RSP=%016llx CS:RIP=%04x:%016llx\n",
           (unsigned long long)vmr(GUEST_SYSENTER_ESP),
           (int)vmr(GUEST_SYSENTER_CS),
           (unsigned long long)vmr(GUEST_SYSENTER_EIP));
    vmx_dump_sel("CS", x86_seg_cs);
    vmx_dump_sel("DS", x86_seg_ds);
    vmx_dump_sel("SS", x86_seg_ss);
    vmx_dump_sel("ES", x86_seg_es);
    vmx_dump_sel("FS", x86_seg_fs);
    vmx_dump_sel("GS", x86_seg_gs);
    vmx_dump_sel("GDTR", x86_seg_gdtr);
    vmx_dump_sel("LDTR", x86_seg_ldtr);
    vmx_dump_sel("IDTR", x86_seg_idtr);
    vmx_dump_sel("TR", x86_seg_tr);
    x  = (unsigned long long)vmr(TSC_OFFSET_HIGH) << 32;
    x |= (uint32_t)vmr(TSC_OFFSET);
    printk("TSC Offset = %016llx\n", x);
    x  = (unsigned long long)vmr(GUEST_IA32_DEBUGCTL_HIGH) << 32;
    x |= (uint32_t)vmr(GUEST_IA32_DEBUGCTL);
    printk("DebugCtl=%016llx DebugExceptions=%016llx\n", x,
           (unsigned long long)vmr(GUEST_PENDING_DBG_EXCEPTIONS));
    printk("Interruptibility=%04x ActivityState=%04x\n",
           (int)vmr(GUEST_INTERRUPTIBILITY_INFO),
           (int)vmr(GUEST_ACTIVITY_STATE));

    printk("*** Host State ***\n");
    printk("RSP = 0x%016llx  RIP = 0x%016llx\n", 
           (unsigned long long)vmr(HOST_RSP),
           (unsigned long long)vmr(HOST_RIP));
    printk("CS=%04x DS=%04x ES=%04x FS=%04x GS=%04x SS=%04x TR=%04x\n",
           (uint16_t)vmr(HOST_CS_SELECTOR),
           (uint16_t)vmr(HOST_DS_SELECTOR),
           (uint16_t)vmr(HOST_ES_SELECTOR),
           (uint16_t)vmr(HOST_FS_SELECTOR),
           (uint16_t)vmr(HOST_GS_SELECTOR),
           (uint16_t)vmr(HOST_SS_SELECTOR),
           (uint16_t)vmr(HOST_TR_SELECTOR));
    printk("FSBase=%016llx GSBase=%016llx TRBase=%016llx\n",
           (unsigned long long)vmr(HOST_FS_BASE),
           (unsigned long long)vmr(HOST_GS_BASE),
           (unsigned long long)vmr(HOST_TR_BASE));
    printk("GDTBase=%016llx IDTBase=%016llx\n",
           (unsigned long long)vmr(HOST_GDTR_BASE),
           (unsigned long long)vmr(HOST_IDTR_BASE));
    printk("CR0=%016llx CR3=%016llx CR4=%016llx\n",
           (unsigned long long)vmr(HOST_CR0),
           (unsigned long long)vmr(HOST_CR3),
           (unsigned long long)vmr(HOST_CR4));
    printk("Sysenter RSP=%016llx CS:RIP=%04x:%016llx\n",
           (unsigned long long)vmr(HOST_SYSENTER_ESP),
           (int)vmr(HOST_SYSENTER_CS),
           (unsigned long long)vmr(HOST_SYSENTER_EIP));

    printk("*** Control State ***\n");
    printk("PinBased=%08x CPUBased=%08x SecondaryExec=%08x\n",
           (uint32_t)vmr(PIN_BASED_VM_EXEC_CONTROL),
           (uint32_t)vmr(CPU_BASED_VM_EXEC_CONTROL),
           (uint32_t)vmr(SECONDARY_VM_EXEC_CONTROL));
    printk("EntryControls=%08x ExitControls=%08x\n",
           (uint32_t)vmr(VM_ENTRY_CONTROLS),
           (uint32_t)vmr(VM_EXIT_CONTROLS));
    printk("ExceptionBitmap=%08x\n",
           (uint32_t)vmr(EXCEPTION_BITMAP));
    printk("VMEntry: intr_info=%08x errcode=%08x ilen=%08x\n",
           (uint32_t)vmr(VM_ENTRY_INTR_INFO),
           (uint32_t)vmr(VM_ENTRY_EXCEPTION_ERROR_CODE),
           (uint32_t)vmr(VM_ENTRY_INSTRUCTION_LEN));
    printk("VMExit: intr_info=%08x errcode=%08x ilen=%08x\n",
           (uint32_t)vmr(VM_EXIT_INTR_INFO),
           (uint32_t)vmr(VM_EXIT_INTR_ERROR_CODE),
           (uint32_t)vmr(VM_ENTRY_INSTRUCTION_LEN));
    printk("        reason=%08x qualification=%08x\n",
           (uint32_t)vmr(VM_EXIT_REASON),
           (uint32_t)vmr(EXIT_QUALIFICATION));
    printk("IDTVectoring: info=%08x errcode=%08x\n",
           (uint32_t)vmr(IDT_VECTORING_INFO),
           (uint32_t)vmr(IDT_VECTORING_ERROR_CODE));
    printk("TPR Threshold = 0x%02x\n",
           (uint32_t)vmr(TPR_THRESHOLD));
    printk("EPT pointer = 0x%08x%08x\n",
           (uint32_t)vmr(EPT_POINTER_HIGH), (uint32_t)vmr(EPT_POINTER));
    printk("Virtual processor ID = 0x%04x\n",
           (uint32_t)vmr(VIRTUAL_PROCESSOR_ID));

    vmx_vmcs_exit(v);
}

static void vmcs_dump(unsigned char ch)
{
    struct domain *d;
    struct vcpu *v;
    
    printk("*********** VMCS Areas **************\n");

    rcu_read_lock(&domlist_read_lock);

    for_each_domain ( d )
    {
        if ( !is_hvm_domain(d) )
            continue;
        printk("\n>>> Domain %d <<<\n", d->domain_id);
        for_each_vcpu ( d, v )
        {
            printk("\tVCPU %d\n", v->vcpu_id);
            vmcs_dump_vcpu(v);
        }
    }

    rcu_read_unlock(&domlist_read_lock);

    printk("**************************************\n");
}

void setup_vmcs_dump(void)
{
    register_keyhandler('v', vmcs_dump, "dump Intel's VMCS");
}


/*
 * Local variables:
 * mode: C
 * c-set-style: "BSD"
 * c-basic-offset: 4
 * tab-width: 4
 * indent-tabs-mode: nil
 * End:
 */