OrgACRN
/
acrn-hypervisor
mirror of https://github.com/projectacrn/acrn-hypervisor.git
1
0
Fork 0
Code Issues Releases Wiki Activity
acrn-hypervisor/hypervisor/arch/x86/vmx.c

1130 lines
33 KiB
C
Raw Blame History

/*
* Copyright (C) 2018 Intel Corporation. All rights reserved.
*
* SPDX-License-Identifier: BSD-3-Clause
*/
#include <hypervisor.h>
#include <vm0_boot.h>
#include <cpu.h>
#ifdef CONFIG_EFI_STUB
extern struct efi_context* efi_ctx;
#endif
#define REAL_MODE_BSP_INIT_CODE_SEL (0xf000U)
#define REAL_MODE_DATA_SEG_AR (0x0093U)
#define REAL_MODE_CODE_SEG_AR (0x009fU)
#define PROTECTED_MODE_DATA_SEG_AR (0xc093U)
#define PROTECTED_MODE_CODE_SEG_AR (0xc09bU)
#define DR7_INIT_VALUE (0x400UL)
#define LDTR_AR (0x0082U) /* LDT, type must be 2, refer to SDM Vol3 26.3.1.2 */
#define TR_AR (0x008bU) /* TSS (busy), refer to SDM Vol3 26.3.1.2 */
static uint64_t cr0_host_mask;
static uint64_t cr0_always_on_mask;
static uint64_t cr0_always_off_mask;
static uint64_t cr4_host_mask;
static uint64_t cr4_always_on_mask;
static uint64_t cr4_always_off_mask;
void update_msr_bitmap_x2apic_apicv(struct acrn_vcpu *vcpu);
void update_msr_bitmap_x2apic_passthru(struct acrn_vcpu *vcpu);
bool is_vmx_disabled(void)
{
uint64_t msr_val;
/* Read Feature ControL MSR */
msr_val = msr_read(MSR_IA32_FEATURE_CONTROL);
/* Check if feature control is locked and vmx cannot be enabled */
if ((msr_val & MSR_IA32_FEATURE_CONTROL_LOCK) != 0U &&
(msr_val & MSR_IA32_FEATURE_CONTROL_VMX_NO_SMX) == 0U) {
return true;
}
return false;
}
/**
* @pre addr != NULL && addr is 4KB-aligned
* rev[31:0] 32 bits located at vmxon region physical address
* @pre rev[30:0] == VMCS revision && rev[31] == 0
*/
static inline void exec_vmxon(void *addr)
{
/* Turn VMX on, pre-conditions can avoid VMfailInvalid
* here no need check RFLAGS since it will generate #GP or #UD
* except VMsuccess. SDM 30.3
*/
asm volatile (
"vmxon (%%rax)\n"
:
: "a"(addr)
: "cc", "memory");
}
/* Per cpu data to hold the vmxon_region for each pcpu.
* It will be used again when we start a pcpu after the pcpu was down.
* S3 enter/exit will use it.
*/
void exec_vmxon_instr(uint16_t pcpu_id)
{
uint64_t tmp64, vmcs_pa;
uint32_t tmp32;
void *vmxon_region_va = (void *)per_cpu(vmxon_region, pcpu_id);
uint64_t vmxon_region_pa;
struct acrn_vcpu *vcpu = get_ever_run_vcpu(pcpu_id);
/* Initialize vmxon page with revision id from IA32 VMX BASIC MSR */
tmp32 = (uint32_t)msr_read(MSR_IA32_VMX_BASIC);
(void)memcpy_s((uint32_t *) vmxon_region_va, 4U, (void *)&tmp32, 4U);
/* Turn on CR0.NE and CR4.VMXE */
CPU_CR_READ(cr0, &tmp64);
CPU_CR_WRITE(cr0, tmp64 | CR0_NE);
CPU_CR_READ(cr4, &tmp64);
CPU_CR_WRITE(cr4, tmp64 | CR4_VMXE);
/* Read Feature ControL MSR */
tmp64 = msr_read(MSR_IA32_FEATURE_CONTROL);
/* Check if feature control is locked */
if ((tmp64 & MSR_IA32_FEATURE_CONTROL_LOCK) == 0U) {
/* Lock and enable VMX support */
tmp64 |= (MSR_IA32_FEATURE_CONTROL_LOCK |
MSR_IA32_FEATURE_CONTROL_VMX_NO_SMX);
msr_write(MSR_IA32_FEATURE_CONTROL, tmp64);
}
/* Turn ON VMX */
vmxon_region_pa = hva2hpa(vmxon_region_va);
exec_vmxon(&vmxon_region_pa);
vmcs_pa = hva2hpa(vcpu->arch.vmcs);
exec_vmptrld(&vmcs_pa);
}
static inline void exec_vmxoff(void)
{
asm volatile ("vmxoff" : : : "memory");
}
void vmx_off(uint16_t pcpu_id)
{
struct acrn_vcpu *vcpu = get_ever_run_vcpu(pcpu_id);
uint64_t vmcs_pa;
vmcs_pa = hva2hpa(vcpu->arch.vmcs);
exec_vmclear((void *)&vmcs_pa);
exec_vmxoff();
}
/**
* @pre addr != NULL && addr is 4KB-aligned
* @pre addr != VMXON pointer
*/
void exec_vmclear(void *addr)
{
/* pre-conditions can avoid VMfail
* here no need check RFLAGS since it will generate #GP or #UD
* except VMsuccess. SDM 30.3
*/
asm volatile (
"vmclear (%%rax)\n"
:
: "a"(addr)
: "cc", "memory");
}
/**
* @pre addr != NULL && addr is 4KB-aligned
* @pre addr != VMXON pointer
*/
void exec_vmptrld(void *addr)
{
/* pre-conditions can avoid VMfail
* here no need check RFLAGS since it will generate #GP or #UD
* except VMsuccess. SDM 30.3
*/
asm volatile (
"vmptrld (%%rax)\n"
:
: "a"(addr)
: "cc", "memory");
}
uint64_t exec_vmread64(uint32_t field_full)
{
uint64_t value;
asm volatile (
"vmread %%rdx, %%rax "
: "=a" (value)
: "d"(field_full)
: "cc");
return value;
}
uint32_t exec_vmread32(uint32_t field)
{
uint64_t value;
value = exec_vmread64(field);
return (uint32_t)value;
}
uint16_t exec_vmread16(uint32_t field)
{
uint64_t value;
value = exec_vmread64(field);
return (uint16_t)value;
}
void exec_vmwrite64(uint32_t field_full, uint64_t value)
{
asm volatile (
"vmwrite %%rax, %%rdx "
: : "a" (value), "d"(field_full)
: "cc");
}
void exec_vmwrite32(uint32_t field, uint32_t value)
{
exec_vmwrite64(field, (uint64_t)value);
}
void exec_vmwrite16(uint32_t field, uint16_t value)
{
exec_vmwrite64(field, (uint64_t)value);
}
static void init_cr0_cr4_host_mask(void)
{
static bool inited = false;
uint64_t fixed0, fixed1;
if (!inited) {
/* Read the CR0 fixed0 / fixed1 MSR registers */
fixed0 = msr_read(MSR_IA32_VMX_CR0_FIXED0);
fixed1 = msr_read(MSR_IA32_VMX_CR0_FIXED1);
cr0_host_mask = ~(fixed0 ^ fixed1);
/* Add the bit hv wants to trap */
cr0_host_mask |= CR0_TRAP_MASK;
/* CR0 clear PE/PG from always on bits due to "unrestructed
* guest" feature */
cr0_always_on_mask = fixed0 & (~(CR0_PE | CR0_PG));
cr0_always_off_mask = ~fixed1;
/* SDM 2.5
* bit 63:32 of CR0 and CR4 ar reserved and must be written
* zero. We could merge it with always off mask.
*/
cr0_always_off_mask |= 0xFFFFFFFF00000000UL;
/* Read the CR4 fixed0 / fixed1 MSR registers */
fixed0 = msr_read(MSR_IA32_VMX_CR4_FIXED0);
fixed1 = msr_read(MSR_IA32_VMX_CR4_FIXED1);
cr4_host_mask = ~(fixed0 ^ fixed1);
/* Add the bit hv wants to trap */
cr4_host_mask |= CR4_TRAP_MASK;
cr4_always_on_mask = fixed0;
/* Record the bit fixed to 0 for CR4, including reserved bits */
cr4_always_off_mask = ~fixed1;
/* SDM 2.5
* bit 63:32 of CR0 and CR4 ar reserved and must be written
* zero. We could merge it with always off mask.
*/
cr4_always_off_mask |= 0xFFFFFFFF00000000UL;
cr4_always_off_mask |= CR4_RESERVED_MASK;
inited = true;
}
exec_vmwrite(VMX_CR0_MASK, cr0_host_mask);
/* Output CR0 mask value */
pr_dbg("CR0 mask value: 0x%016llx", cr0_host_mask);
exec_vmwrite(VMX_CR4_MASK, cr4_host_mask);
/* Output CR4 mask value */
pr_dbg("CR4 mask value: 0x%016llx", cr4_host_mask);
}
uint64_t vmx_rdmsr_pat(const struct acrn_vcpu *vcpu)
{
/*
* note: if run_ctx->cr0.CD is set, the actual value in guest's
* IA32_PAT MSR is PAT_ALL_UC_VALUE, which may be different from
* the saved value guest_msrs[MSR_IA32_PAT]
*/
return vcpu_get_guest_msr(vcpu, MSR_IA32_PAT);
}
int vmx_wrmsr_pat(struct acrn_vcpu *vcpu, uint64_t value)
{
uint32_t i;
uint64_t field;
for (i = 0U; i < 8U; i++) {
field = (value >> (i * 8U)) & 0xffUL;
if (pat_mem_type_invalid(field) ||
((PAT_FIELD_RSV_BITS & field) != 0UL)) {
pr_err("invalid guest IA32_PAT: 0x%016llx", value);
return -EINVAL;
}
}
vcpu_set_guest_msr(vcpu, MSR_IA32_PAT, value);
/*
* If context->cr0.CD is set, we defer any further requests to write
* guest's IA32_PAT, until the time when guest's CR0.CD is being cleared
*/
if ((vcpu_get_cr0(vcpu) & CR0_CD) == 0UL) {
exec_vmwrite64(VMX_GUEST_IA32_PAT_FULL, value);
}
return 0;
}
static void load_pdptrs(const struct acrn_vcpu *vcpu)
{
uint64_t guest_cr3 = exec_vmread(VMX_GUEST_CR3);
/* TODO: check whether guest cr3 is valid */
uint64_t *guest_cr3_hva = (uint64_t *)gpa2hva(vcpu->vm, guest_cr3);
exec_vmwrite64(VMX_GUEST_PDPTE0_FULL, get_pgentry(guest_cr3_hva + 0UL));
exec_vmwrite64(VMX_GUEST_PDPTE1_FULL, get_pgentry(guest_cr3_hva + 1UL));
exec_vmwrite64(VMX_GUEST_PDPTE2_FULL, get_pgentry(guest_cr3_hva + 2UL));
exec_vmwrite64(VMX_GUEST_PDPTE3_FULL, get_pgentry(guest_cr3_hva + 3UL));
}
static bool is_cr0_write_valid(struct acrn_vcpu *vcpu, uint64_t cr0)
{
/* Shouldn't set always off bit */
if ((cr0 & cr0_always_off_mask) != 0UL) {
return false;
}
/* SDM 25.3 "Changes to instruction behavior in VMX non-root"
*
* We always require "unrestricted guest" control enabled. So
*
* CR0.PG = 1, CR4.PAE = 0 and IA32_EFER.LME = 1 is invalid.
* CR0.PE = 0 and CR0.PG = 1 is invalid.
*/
if (((cr0 & CR0_PG) != 0UL) && !is_pae(vcpu)
&& ((vcpu_get_efer(vcpu) & MSR_IA32_EFER_LME_BIT) != 0UL)) {
return false;
}
if (((cr0 & CR0_PE) == 0UL) && ((cr0 & CR0_PG) != 0UL)) {
return false;
}
/* SDM 6.15 "Exception and Interrupt Refrerence" GP Exception
*
* Loading CR0 regsiter with a set NW flag and a clear CD flag
* is invalid
*/
if (((cr0 & CR0_CD) == 0UL) && ((cr0 & CR0_NW) != 0UL)) {
return false;
}
return true;
}
/*
* Handling of CR0:
* Assume "unrestricted guest" feature is supported by vmx.
* For mode switch, hv only needs to take care of enabling/disabling long mode,
* thanks to "unrestricted guest" feature.
*
* - PE (0) Trapped to track cpu mode.
* Set the value according to the value from guest.
* - MP (1) Flexible to guest
* - EM (2) Flexible to guest
* - TS (3) Flexible to guest
* - ET (4) Flexible to guest
* - NE (5) must always be 1
* - WP (16) Trapped to get if it inhibits supervisor level procedures to
* write into ro-pages.
* - AM (18) Flexible to guest
* - NW (29) Trapped to emulate cache disable situation
* - CD (30) Trapped to emulate cache disable situation
* - PG (31) Trapped to track cpu/paging mode.
* Set the value according to the value from guest.
*/
void vmx_write_cr0(struct acrn_vcpu *vcpu, uint64_t cr0)
{
uint64_t cr0_vmx;
uint32_t entry_ctrls;
bool old_paging_enabled = is_paging_enabled(vcpu);
uint64_t cr0_changed_bits = vcpu_get_cr0(vcpu) ^ cr0;
if (!is_cr0_write_valid(vcpu, cr0)) {
pr_dbg("Invalid cr0 write operation from guest");
vcpu_inject_gp(vcpu, 0U);
return;
}
/* SDM 2.5
* When loading a control register, reserved bit should always set
* to the value previously read.
*/
cr0 &= ~CR0_RESERVED_MASK;
if (!old_paging_enabled && ((cr0 & CR0_PG) != 0UL)) {
if ((vcpu_get_efer(vcpu) & MSR_IA32_EFER_LME_BIT) != 0UL) {
/* Enable long mode */
pr_dbg("VMM: Enable long mode");
entry_ctrls = exec_vmread32(VMX_ENTRY_CONTROLS);
entry_ctrls |= VMX_ENTRY_CTLS_IA32E_MODE;
exec_vmwrite32(VMX_ENTRY_CONTROLS, entry_ctrls);
vcpu_set_efer(vcpu,
vcpu_get_efer(vcpu) | MSR_IA32_EFER_LMA_BIT);
} else if (is_pae(vcpu)) {
/* enabled PAE from paging disabled */
load_pdptrs(vcpu);
} else {
/* do nothing */
}
} else if (old_paging_enabled && ((cr0 & CR0_PG) == 0UL)) {
if ((vcpu_get_efer(vcpu) & MSR_IA32_EFER_LME_BIT) != 0UL) {
/* Disable long mode */
pr_dbg("VMM: Disable long mode");
entry_ctrls = exec_vmread32(VMX_ENTRY_CONTROLS);
entry_ctrls &= ~VMX_ENTRY_CTLS_IA32E_MODE;
exec_vmwrite32(VMX_ENTRY_CONTROLS, entry_ctrls);
vcpu_set_efer(vcpu,
vcpu_get_efer(vcpu) & ~MSR_IA32_EFER_LMA_BIT);
}
} else {
/* do nothing */
}
/* If CR0.CD or CR0.NW get cr0_changed_bits */
if ((cr0_changed_bits & (CR0_CD | CR0_NW)) != 0UL) {
/* No action if only CR0.NW is cr0_changed_bits */
if ((cr0_changed_bits & CR0_CD) != 0UL) {
if ((cr0 & CR0_CD) != 0UL) {
/*
* When the guest requests to set CR0.CD, we don't allow
* guest's CR0.CD to be actually set, instead, we write guest
* IA32_PAT with all-UC entries to emulate the cache
* disabled behavior
*/
exec_vmwrite64(VMX_GUEST_IA32_PAT_FULL, PAT_ALL_UC_VALUE);
if (!iommu_snoop_supported(vcpu->vm)) {
cache_flush_invalidate_all();
}
} else {
/* Restore IA32_PAT to enable cache again */
exec_vmwrite64(VMX_GUEST_IA32_PAT_FULL,
vcpu_get_guest_msr(vcpu, MSR_IA32_PAT));
}
vcpu_make_request(vcpu, ACRN_REQUEST_EPT_FLUSH);
}
}
if ((cr0_changed_bits & (CR0_PG | CR0_WP)) != 0UL) {
vcpu_make_request(vcpu, ACRN_REQUEST_EPT_FLUSH);
}
/* CR0 has no always off bits, except the always on bits, and reserved
* bits, allow to set according to guest.
*/
cr0_vmx = cr0_always_on_mask | cr0;
/* Don't set CD or NW bit to guest */
cr0_vmx &= ~(CR0_CD | CR0_NW);
exec_vmwrite(VMX_GUEST_CR0, cr0_vmx & 0xFFFFFFFFUL);
exec_vmwrite(VMX_CR0_READ_SHADOW, cr0 & 0xFFFFFFFFUL);
/* clear read cache, next time read should from VMCS */
bitmap_clear_lock(CPU_REG_CR0, &vcpu->reg_cached);
pr_dbg("VMM: Try to write %016llx, allow to write 0x%016llx to CR0",
cr0, cr0_vmx);
}
static bool is_cr4_write_valid(struct acrn_vcpu *vcpu, uint64_t cr4)
{
/* Check if guest try to set fixed to 0 bits or reserved bits */
if ((cr4 & cr4_always_off_mask) != 0U) {
return false;
}
/* Do NOT support nested guest */
if ((cr4 & CR4_VMXE) != 0UL) {
return false;
}
/* Do NOT support PCID in guest */
if ((cr4 & CR4_PCIDE) != 0UL) {
return false;
}
if (is_long_mode(vcpu)) {
if ((cr4 & CR4_PAE) == 0UL) {
return false;
}
}
return true;
}
/*
* Handling of CR4:
* Assume "unrestricted guest" feature is supported by vmx.
*
* For CR4, if some feature is not supported by hardware, the corresponding bit
* will be set in cr4_always_off_mask. If guest try to set these bits after
* vmexit, will inject a #GP.
* If a bit for a feature not supported by hardware, which is flexible to guest,
* and write to it do not lead to a VM exit, a #GP should be generated inside
* guest.
*
* - VME (0) Flexible to guest
* - PVI (1) Flexible to guest
* - TSD (2) Flexible to guest
* - DE (3) Flexible to guest
* - PSE (4) Trapped to track paging mode.
* Set the value according to the value from guest.
* - PAE (5) Trapped to track paging mode.
* Set the value according to the value from guest.
* - MCE (6) Flexible to guest
* - PGE (7) Flexible to guest
* - PCE (8) Flexible to guest
* - OSFXSR (9) Flexible to guest
* - OSXMMEXCPT (10) Flexible to guest
* - VMXE (13) Trapped to hide from guest
* - SMXE (14) must always be 0 => must lead to a VM exit
* - PCIDE (17) Trapped to hide from guest
* - OSXSAVE (18) Flexible to guest
* - XSAVE (19) Flexible to guest
* We always keep align with physical cpu. So it's flexible to
* guest
* - SMEP (20) Flexible to guest
* - SMAP (21) Flexible to guest
* - PKE (22) Flexible to guest
*/
void vmx_write_cr4(struct acrn_vcpu *vcpu, uint64_t cr4)
{
uint64_t cr4_vmx;
uint64_t old_cr4 = vcpu_get_cr4(vcpu);
if (!is_cr4_write_valid(vcpu, cr4)) {
pr_dbg("Invalid cr4 write operation from guest");
vcpu_inject_gp(vcpu, 0U);
return;
}
if (((cr4 ^ old_cr4) & (CR4_PGE | CR4_PSE | CR4_PAE |
CR4_SMEP | CR4_SMAP | CR4_PKE)) != 0UL) {
if (((cr4 & CR4_PAE) != 0UL) && is_paging_enabled(vcpu) &&
(is_long_mode(vcpu))) {
load_pdptrs(vcpu);
}
vcpu_make_request(vcpu, ACRN_REQUEST_EPT_FLUSH);
}
/* Aways off bits and reserved bits has been filtered above */
cr4_vmx = cr4_always_on_mask | cr4;
exec_vmwrite(VMX_GUEST_CR4, cr4_vmx & 0xFFFFFFFFUL);
exec_vmwrite(VMX_CR4_READ_SHADOW, cr4 & 0xFFFFFFFFUL);
/* clear read cache, next time read should from VMCS */
bitmap_clear_lock(CPU_REG_CR4, &vcpu->reg_cached);
pr_dbg("VMM: Try to write %016llx, allow to write 0x%016llx to CR4",
cr4, cr4_vmx);
}
/* rip, rsp, ia32_efer and rflags are written to VMCS in start_vcpu */
static void init_guest_vmx(struct acrn_vcpu *vcpu, uint64_t cr0, uint64_t cr3,
uint64_t cr4)
{
struct cpu_context *ctx =
&vcpu->arch.contexts[vcpu->arch.cur_context];
struct ext_context *ectx = &ctx->ext_ctx;
vcpu_set_cr4(vcpu, cr4);
vcpu_set_cr0(vcpu, cr0);
exec_vmwrite(VMX_GUEST_CR3, cr3);
exec_vmwrite(VMX_GUEST_GDTR_BASE, ectx->gdtr.base);
pr_dbg("VMX_GUEST_GDTR_BASE: 0x%016llx", ectx->gdtr.base);
exec_vmwrite32(VMX_GUEST_GDTR_LIMIT, ectx->gdtr.limit);
pr_dbg("VMX_GUEST_GDTR_LIMIT: 0x%016llx", ectx->gdtr.limit);
exec_vmwrite(VMX_GUEST_IDTR_BASE, ectx->idtr.base);
pr_dbg("VMX_GUEST_IDTR_BASE: 0x%016llx", ectx->idtr.base);
exec_vmwrite32(VMX_GUEST_IDTR_LIMIT, ectx->idtr.limit);
pr_dbg("VMX_GUEST_IDTR_LIMIT: 0x%016llx", ectx->idtr.limit);
/* init segment selectors: es, cs, ss, ds, fs, gs, ldtr, tr */
load_segment(ectx->cs, VMX_GUEST_CS);
load_segment(ectx->ss, VMX_GUEST_SS);
load_segment(ectx->ds, VMX_GUEST_DS);
load_segment(ectx->es, VMX_GUEST_ES);
load_segment(ectx->fs, VMX_GUEST_FS);
load_segment(ectx->gs, VMX_GUEST_GS);
load_segment(ectx->tr, VMX_GUEST_TR);
load_segment(ectx->ldtr, VMX_GUEST_LDTR);
/* fixed values */
exec_vmwrite32(VMX_GUEST_IA32_SYSENTER_CS, 0U);
exec_vmwrite(VMX_GUEST_IA32_SYSENTER_ESP, 0UL);
exec_vmwrite(VMX_GUEST_IA32_SYSENTER_EIP, 0UL);
exec_vmwrite(VMX_GUEST_PENDING_DEBUG_EXCEPT, 0UL);
exec_vmwrite(VMX_GUEST_IA32_DEBUGCTL_FULL, 0UL);
exec_vmwrite32(VMX_GUEST_INTERRUPTIBILITY_INFO, 0U);
exec_vmwrite32(VMX_GUEST_ACTIVITY_STATE, 0U);
exec_vmwrite32(VMX_GUEST_SMBASE, 0U);
vcpu_set_guest_msr(vcpu, MSR_IA32_PAT, PAT_POWER_ON_VALUE);
exec_vmwrite(VMX_GUEST_IA32_PAT_FULL, PAT_POWER_ON_VALUE);
exec_vmwrite(VMX_GUEST_DR7, DR7_INIT_VALUE);
}
static void init_guest_state(struct acrn_vcpu *vcpu)
{
struct cpu_context *ctx =
&vcpu->arch.contexts[vcpu->arch.cur_context];
init_guest_vmx(vcpu, ctx->run_ctx.cr0, ctx->ext_ctx.cr3,
ctx->run_ctx.cr4 & ~CR4_VMXE);
}
static void init_host_state(void)
{
uint16_t value16;
uint64_t value64;
uint64_t value;
uint64_t tss_addr;
uint64_t gdt_base;
uint64_t idt_base;
pr_dbg("*********************");
pr_dbg("Initialize host state");
pr_dbg("*********************");
/***************************************************
* 16 - Bit fields
* Move the current ES, CS, SS, DS, FS, GS, TR, LDTR * values to the
* corresponding 16-bit host * segment selection (ES, CS, SS, DS, FS,
* GS), * Task Register (TR), * Local Descriptor Table Register (LDTR)
*
***************************************************/
CPU_SEG_READ(es, &value16);
exec_vmwrite16(VMX_HOST_ES_SEL, value16);
pr_dbg("VMX_HOST_ES_SEL: 0x%hx ", value16);
CPU_SEG_READ(cs, &value16);
exec_vmwrite16(VMX_HOST_CS_SEL, value16);
pr_dbg("VMX_HOST_CS_SEL: 0x%hx ", value16);
CPU_SEG_READ(ss, &value16);
exec_vmwrite16(VMX_HOST_SS_SEL, value16);
pr_dbg("VMX_HOST_SS_SEL: 0x%hx ", value16);
CPU_SEG_READ(ds, &value16);
exec_vmwrite16(VMX_HOST_DS_SEL, value16);
pr_dbg("VMX_HOST_DS_SEL: 0x%hx ", value16);
CPU_SEG_READ(fs, &value16);
exec_vmwrite16(VMX_HOST_FS_SEL, value16);
pr_dbg("VMX_HOST_FS_SEL: 0x%hx ", value16);
CPU_SEG_READ(gs, &value16);
exec_vmwrite16(VMX_HOST_GS_SEL, value16);
pr_dbg("VMX_HOST_GS_SEL: 0x%hx ", value16);
exec_vmwrite16(VMX_HOST_TR_SEL, HOST_GDT_RING0_CPU_TSS_SEL);
pr_dbg("VMX_HOST_TR_SEL: 0x%hx ", HOST_GDT_RING0_CPU_TSS_SEL);
/******************************************************
* 32-bit fields
* Set up the 32 bit host state fields - pg 3418 B.3.3 * Set limit for
* ES, CS, DD, DS, FS, GS, LDTR, Guest TR, * GDTR, and IDTR
******************************************************/
/* TODO: Should guest GDTB point to host GDTB ? */
/* Obtain the current global descriptor table base */
gdt_base = sgdt();
if (((gdt_base >> 47U) & 0x1UL) != 0UL) {
gdt_base |= 0xffff000000000000UL;
}
/* Set up the guest and host GDTB base fields with current GDTB base */
exec_vmwrite(VMX_HOST_GDTR_BASE, gdt_base);
pr_dbg("VMX_HOST_GDTR_BASE: 0x%x ", gdt_base);
tss_addr = hva2hpa((void *)&get_cpu_var(tss));
/* Set up host TR base fields */
exec_vmwrite(VMX_HOST_TR_BASE, tss_addr);
pr_dbg("VMX_HOST_TR_BASE: 0x%016llx ", tss_addr);
/* Obtain the current interrupt descriptor table base */
idt_base = sidt();
/* base */
if (((idt_base >> 47U) & 0x1UL) != 0UL) {
idt_base |= 0xffff000000000000UL;
}
exec_vmwrite(VMX_HOST_IDTR_BASE, idt_base);
pr_dbg("VMX_HOST_IDTR_BASE: 0x%x ", idt_base);
/**************************************************/
/* 64-bit fields */
pr_dbg("64-bit********");
value64 = msr_read(MSR_IA32_PAT);
exec_vmwrite64(VMX_HOST_IA32_PAT_FULL, value64);
pr_dbg("VMX_HOST_IA32_PAT: 0x%016llx ", value64);
value64 = msr_read(MSR_IA32_EFER);
exec_vmwrite64(VMX_HOST_IA32_EFER_FULL, value64);
pr_dbg("VMX_HOST_IA32_EFER: 0x%016llx ",
value64);
/**************************************************/
/* Natural width fields */
pr_dbg("Natural-width********");
/* Set up host CR0 field */
CPU_CR_READ(cr0, &value);
exec_vmwrite(VMX_HOST_CR0, value);
pr_dbg("VMX_HOST_CR0: 0x%016llx ", value);
/* Set up host CR3 field */
CPU_CR_READ(cr3, &value);
exec_vmwrite(VMX_HOST_CR3, value);
pr_dbg("VMX_HOST_CR3: 0x%016llx ", value);
/* Set up host CR4 field */
CPU_CR_READ(cr4, &value);
exec_vmwrite(VMX_HOST_CR4, value);
pr_dbg("VMX_HOST_CR4: 0x%016llx ", value);
/* Set up host and guest FS base address */
value = msr_read(MSR_IA32_FS_BASE);
exec_vmwrite(VMX_HOST_FS_BASE, value);
pr_dbg("VMX_HOST_FS_BASE: 0x%016llx ", value);
value = msr_read(MSR_IA32_GS_BASE);
exec_vmwrite(VMX_HOST_GS_BASE, value);
pr_dbg("VMX_HOST_GS_BASE: 0x%016llx ", value);
/* Set up host instruction pointer on VM Exit */
value64 = (uint64_t)&vm_exit;
pr_dbg("HOST RIP on VMExit %016llx ", value64);
exec_vmwrite(VMX_HOST_RIP, value64);
pr_dbg("vm exit return address = %016llx ", value64);
/* As a type I hypervisor, just init sysenter fields to 0 */
exec_vmwrite32(VMX_HOST_IA32_SYSENTER_CS, 0U);
exec_vmwrite(VMX_HOST_IA32_SYSENTER_ESP, 0UL);
exec_vmwrite(VMX_HOST_IA32_SYSENTER_EIP, 0UL);
}
static uint32_t check_vmx_ctrl(uint32_t msr, uint32_t ctrl_req)
{
uint64_t vmx_msr;
uint32_t vmx_msr_low, vmx_msr_high;
uint32_t ctrl = ctrl_req;
vmx_msr = msr_read(msr);
vmx_msr_low = (uint32_t)vmx_msr;
vmx_msr_high = (uint32_t)(vmx_msr >> 32U);
pr_dbg("VMX_PIN_VM_EXEC_CONTROLS:low=0x%x, high=0x%x\n",
vmx_msr_low, vmx_msr_high);
/* high 32b: must 0 setting
* low 32b: must 1 setting
*/
ctrl &= vmx_msr_high;
ctrl |= vmx_msr_low;
if ((ctrl_req & ~ctrl) != 0U) {
pr_err("VMX ctrl 0x%x not fully enabled: "
"request 0x%x but get 0x%x\n",
msr, ctrl_req, ctrl);
}
return ctrl;
}
static void init_exec_ctrl(struct acrn_vcpu *vcpu)
{
uint32_t value32;
uint64_t value64;
struct acrn_vm *vm = vcpu->vm;
/* Log messages to show initializing VMX execution controls */
pr_dbg("*****************************");
pr_dbg("Initialize execution control ");
pr_dbg("*****************************");
/* Set up VM Execution control to enable Set VM-exits on external
* interrupts preemption timer - pg 2899 24.6.1
*/
/* enable external interrupt VM Exit */
value32 = check_vmx_ctrl(MSR_IA32_VMX_PINBASED_CTLS,
VMX_PINBASED_CTLS_IRQ_EXIT);
if (is_apicv_posted_intr_supported()) {
value32 |= VMX_PINBASED_CTLS_POST_IRQ;
}
exec_vmwrite32(VMX_PIN_VM_EXEC_CONTROLS, value32);
pr_dbg("VMX_PIN_VM_EXEC_CONTROLS: 0x%x ", value32);
/* Set up primary processor based VM execution controls - pg 2900
* 24.6.2. Set up for:
* Enable TSC offsetting
* Enable TSC exiting
* guest access to IO bit-mapped ports causes VM exit
* guest access to MSR causes VM exit
* Activate secondary controls
*/
/* These are bits 1,4-6,8,13-16, and 26, the corresponding bits of
* the IA32_VMX_PROCBASED_CTRLS MSR are always read as 1 --- A.3.2
*/
value32 = check_vmx_ctrl(MSR_IA32_VMX_PROCBASED_CTLS,
VMX_PROCBASED_CTLS_TSC_OFF |
/* VMX_PROCBASED_CTLS_RDTSC | */
VMX_PROCBASED_CTLS_TPR_SHADOW |
VMX_PROCBASED_CTLS_IO_BITMAP |
VMX_PROCBASED_CTLS_MSR_BITMAP |
VMX_PROCBASED_CTLS_SECONDARY);
/*Disable VM_EXIT for CR3 access*/
value32 &= ~(VMX_PROCBASED_CTLS_CR3_LOAD |
VMX_PROCBASED_CTLS_CR3_STORE);
/*
* Disable VM_EXIT for invlpg execution.
*/
value32 &= ~VMX_PROCBASED_CTLS_INVLPG;
exec_vmwrite32(VMX_PROC_VM_EXEC_CONTROLS, value32);
pr_dbg("VMX_PROC_VM_EXEC_CONTROLS: 0x%x ", value32);
/* Set up secondary processor based VM execution controls - pg 2901
* 24.6.2. Set up for: * Enable EPT * Enable RDTSCP * Unrestricted
* guest (optional)
*/
value32 = check_vmx_ctrl(MSR_IA32_VMX_PROCBASED_CTLS2,
VMX_PROCBASED_CTLS2_VAPIC |
VMX_PROCBASED_CTLS2_EPT |
VMX_PROCBASED_CTLS2_RDTSCP |
VMX_PROCBASED_CTLS2_UNRESTRICT|
VMX_PROCBASED_CTLS2_VAPIC_REGS);
if (vcpu->arch.vpid != 0U) {
value32 |= VMX_PROCBASED_CTLS2_VPID;
} else {
value32 &= ~VMX_PROCBASED_CTLS2_VPID;
}
if (is_apicv_intr_delivery_supported()) {
value32 |= VMX_PROCBASED_CTLS2_VIRQ;
} else {
/*
* This field exists only on processors that support
* the 1-setting of the "use TPR shadow"
* VM-execution control.
*
* Set up TPR threshold for virtual interrupt delivery
* - pg 2904 24.6.8
*/
exec_vmwrite32(VMX_TPR_THRESHOLD, 0U);
}
if (cpu_has_cap(X86_FEATURE_OSXSAVE)) {
exec_vmwrite64(VMX_XSS_EXITING_BITMAP_FULL, 0UL);
value32 |= VMX_PROCBASED_CTLS2_XSVE_XRSTR;
}
value32 |= VMX_PROCBASED_CTLS2_WBINVD;
exec_vmwrite32(VMX_PROC_VM_EXEC_CONTROLS2, value32);
pr_dbg("VMX_PROC_VM_EXEC_CONTROLS2: 0x%x ", value32);
/*APIC-v, config APIC-access address*/
value64 = vlapic_apicv_get_apic_access_addr();
exec_vmwrite64(VMX_APIC_ACCESS_ADDR_FULL, value64);
/*APIC-v, config APIC virtualized page address*/
value64 = vlapic_apicv_get_apic_page_addr(vcpu_vlapic(vcpu));
exec_vmwrite64(VMX_VIRTUAL_APIC_PAGE_ADDR_FULL, value64);
if (is_apicv_intr_delivery_supported()) {
/* Disable all EOI VMEXIT by default and
* clear RVI and SVI.
*/
exec_vmwrite64(VMX_EOI_EXIT0_FULL, 0UL);
exec_vmwrite64(VMX_EOI_EXIT1_FULL, 0UL);
exec_vmwrite64(VMX_EOI_EXIT2_FULL, 0UL);
exec_vmwrite64(VMX_EOI_EXIT3_FULL, 0UL);
exec_vmwrite16(VMX_GUEST_INTR_STATUS, 0U);
if (is_apicv_posted_intr_supported()) {
exec_vmwrite16(VMX_POSTED_INTR_VECTOR,
VECTOR_POSTED_INTR);
exec_vmwrite64(VMX_PIR_DESC_ADDR_FULL,
apicv_get_pir_desc_paddr(vcpu));
}
}
/* Load EPTP execution control
* TODO: introduce API to make this data driven based
* on VMX_EPT_VPID_CAP
*/
value64 = hva2hpa(vm->arch_vm.nworld_eptp) | (3UL << 3U) | 6UL;
exec_vmwrite64(VMX_EPT_POINTER_FULL, value64);
pr_dbg("VMX_EPT_POINTER: 0x%016llx ", value64);
/* Set up guest exception mask bitmap setting a bit * causes a VM exit
* on corresponding guest * exception - pg 2902 24.6.3
* enable VM exit on MC only
*/
value32 = (1U << IDT_MC);
exec_vmwrite32(VMX_EXCEPTION_BITMAP, value32);
/* Set up page fault error code mask - second paragraph * pg 2902
* 24.6.3 - guest page fault exception causing * vmexit is governed by
* both VMX_EXCEPTION_BITMAP and * VMX_PF_ERROR_CODE_MASK
*/
exec_vmwrite32(VMX_PF_ERROR_CODE_MASK, 0U);
/* Set up page fault error code match - second paragraph * pg 2902
* 24.6.3 - guest page fault exception causing * vmexit is governed by
* both VMX_EXCEPTION_BITMAP and * VMX_PF_ERROR_CODE_MATCH
*/
exec_vmwrite32(VMX_PF_ERROR_CODE_MATCH, 0U);
/* Set up CR3 target count - An execution of mov to CR3 * by guest
* causes HW to evaluate operand match with * one of N CR3-Target Value
* registers. The CR3 target * count values tells the number of
* target-value regs to evaluate
*/
exec_vmwrite32(VMX_CR3_TARGET_COUNT, 0U);
/* Set up IO bitmap register A and B - pg 2902 24.6.4 */
value64 = hva2hpa(vm->arch_vm.io_bitmap);
exec_vmwrite64(VMX_IO_BITMAP_A_FULL, value64);
pr_dbg("VMX_IO_BITMAP_A: 0x%016llx ", value64);
value64 = hva2hpa((void *)&(vm->arch_vm.io_bitmap[PAGE_SIZE]));
exec_vmwrite64(VMX_IO_BITMAP_B_FULL, value64);
pr_dbg("VMX_IO_BITMAP_B: 0x%016llx ", value64);
init_msr_emulation(vcpu);
/* Set up executive VMCS pointer - pg 2905 24.6.10 */
exec_vmwrite64(VMX_EXECUTIVE_VMCS_PTR_FULL, 0UL);
/* Setup Time stamp counter offset - pg 2902 24.6.5 */
exec_vmwrite64(VMX_TSC_OFFSET_FULL, 0UL);
/* Set up the link pointer */
exec_vmwrite64(VMX_VMS_LINK_PTR_FULL, 0xFFFFFFFFFFFFFFFFUL);
/* Natural-width */
pr_dbg("Natural-width*********");
init_cr0_cr4_host_mask();
/* The CR3 target registers work in concert with VMX_CR3_TARGET_COUNT
* field. Using these registers guest CR3 access can be managed. i.e.,
* if operand does not match one of these register values a VM exit
* would occur
*/
exec_vmwrite(VMX_CR3_TARGET_0, 0UL);
exec_vmwrite(VMX_CR3_TARGET_1, 0UL);
exec_vmwrite(VMX_CR3_TARGET_2, 0UL);
exec_vmwrite(VMX_CR3_TARGET_3, 0UL);
}
static void init_entry_ctrl(const struct acrn_vcpu *vcpu)
{
uint32_t value32;
/* Log messages to show initializing VMX entry controls */
pr_dbg("*************************");
pr_dbg("Initialize Entry control ");
pr_dbg("*************************");
/* Set up VMX entry controls - pg 2908 24.8.1 * Set IA32e guest mode -
* on VM entry processor is in IA32e 64 bitmode * Start guest with host
* IA32_PAT and IA32_EFER
*/
value32 = (VMX_ENTRY_CTLS_LOAD_EFER |
VMX_ENTRY_CTLS_LOAD_PAT);
if (get_vcpu_mode(vcpu) == CPU_MODE_64BIT) {
value32 |= (VMX_ENTRY_CTLS_IA32E_MODE);
}
value32 = check_vmx_ctrl(MSR_IA32_VMX_ENTRY_CTLS, value32);
exec_vmwrite32(VMX_ENTRY_CONTROLS, value32);
pr_dbg("VMX_ENTRY_CONTROLS: 0x%x ", value32);
/* Set up VMX entry MSR load count - pg 2908 24.8.2 Tells the number of
* MSRs on load from memory on VM entry from mem address provided by
* VM-entry MSR load address field
*/
exec_vmwrite32(VMX_ENTRY_MSR_LOAD_COUNT, MSR_AREA_COUNT);
exec_vmwrite64(VMX_ENTRY_MSR_LOAD_ADDR_FULL, (uint64_t)vcpu->arch.msr_area.guest);
/* Set up VM entry interrupt information field pg 2909 24.8.3 */
exec_vmwrite32(VMX_ENTRY_INT_INFO_FIELD, 0U);
/* Set up VM entry exception error code - pg 2910 24.8.3 */
exec_vmwrite32(VMX_ENTRY_EXCEPTION_ERROR_CODE, 0U);
/* Set up VM entry instruction length - pg 2910 24.8.3 */
exec_vmwrite32(VMX_ENTRY_INSTR_LENGTH, 0U);
}
static void init_exit_ctrl(struct acrn_vcpu *vcpu)
{
uint32_t value32;
/* Log messages to show initializing VMX entry controls */
pr_dbg("************************");
pr_dbg("Initialize Exit control ");
pr_dbg("************************");
/* Set up VM exit controls - pg 2907 24.7.1 for: Host address space
* size is 64 bit Set up to acknowledge interrupt on exit, if 1 the HW
* acks the interrupt in VMX non-root and saves the interrupt vector to
* the relevant VM exit field for further processing by Hypervisor
* Enable saving and loading of IA32_PAT and IA32_EFER on VMEXIT Enable
* saving of pre-emption timer on VMEXIT
*/
value32 = check_vmx_ctrl(MSR_IA32_VMX_EXIT_CTLS,
VMX_EXIT_CTLS_ACK_IRQ |
VMX_EXIT_CTLS_SAVE_PAT |
VMX_EXIT_CTLS_LOAD_PAT |
VMX_EXIT_CTLS_LOAD_EFER |
VMX_EXIT_CTLS_SAVE_EFER |
VMX_EXIT_CTLS_HOST_ADDR64);
exec_vmwrite32(VMX_EXIT_CONTROLS, value32);
pr_dbg("VMX_EXIT_CONTROL: 0x%x ", value32);
/* Set up VM exit MSR store and load counts pg 2908 24.7.2 - tells the
* HW number of MSRs to stored to mem and loaded from mem on VM exit.
* The 64 bit VM-exit MSR store and load address fields provide the
* corresponding addresses
*/
exec_vmwrite32(VMX_EXIT_MSR_STORE_COUNT, MSR_AREA_COUNT);
exec_vmwrite32(VMX_EXIT_MSR_LOAD_COUNT, MSR_AREA_COUNT);
exec_vmwrite64(VMX_EXIT_MSR_STORE_ADDR_FULL, (uint64_t)vcpu->arch.msr_area.guest);
exec_vmwrite64(VMX_EXIT_MSR_LOAD_ADDR_FULL, (uint64_t)vcpu->arch.msr_area.host);
}
/**
* @pre vcpu != NULL
*/
void init_vmcs(struct acrn_vcpu *vcpu)
{
uint64_t vmx_rev_id;
uint64_t vmcs_pa;
/* Log message */
pr_dbg("Initializing VMCS");
/* Obtain the VM Rev ID from HW and populate VMCS page with it */
vmx_rev_id = msr_read(MSR_IA32_VMX_BASIC);
(void)memcpy_s(vcpu->arch.vmcs, 4U, (void *)&vmx_rev_id, 4U);
/* Execute VMCLEAR on current VMCS */
vmcs_pa = hva2hpa(vcpu->arch.vmcs);
exec_vmclear((void *)&vmcs_pa);
/* Load VMCS pointer */
exec_vmptrld((void *)&vmcs_pa);
/* Initialize the Virtual Machine Control Structure (VMCS) */
init_host_state();
/* init exec_ctrl needs to run before init_guest_state */
init_exec_ctrl(vcpu);
init_guest_state(vcpu);
init_entry_ctrl(vcpu);
init_exit_ctrl(vcpu);
}
#ifndef CONFIG_PARTITION_MODE
void switch_apicv_mode_x2apic(struct acrn_vcpu *vcpu)
{
uint32_t value32;
value32 = exec_vmread32(VMX_PROC_VM_EXEC_CONTROLS2);
value32 &= ~VMX_PROCBASED_CTLS2_VAPIC;
value32 |= VMX_PROCBASED_CTLS2_VX2APIC;
exec_vmwrite32(VMX_PROC_VM_EXEC_CONTROLS2, value32);
update_msr_bitmap_x2apic_apicv(vcpu);
}
#else
void switch_apicv_mode_x2apic(struct acrn_vcpu *vcpu)
{
uint32_t value32;
if(vcpu->vm->vm_desc->lapic_pt) {
/*
* Disable external interrupt exiting and irq ack
* Disable posted interrupt processing
* update x2apic msr bitmap for pass-thru
* enable inteception only for ICR
* disable pre-emption for TSC DEADLINE MSR
* Disable Register Virtualization and virtual interrupt delivery
* Disable "use TPR shadow"
*/
value32 = exec_vmread32(VMX_PIN_VM_EXEC_CONTROLS);
value32 &= ~VMX_PINBASED_CTLS_IRQ_EXIT;
if (is_apicv_posted_intr_supported()) {
value32 &= ~VMX_PINBASED_CTLS_POST_IRQ;
}
exec_vmwrite32(VMX_PIN_VM_EXEC_CONTROLS, value32);
value32 = exec_vmread32(VMX_EXIT_CONTROLS);
value32 &= ~VMX_EXIT_CTLS_ACK_IRQ;
exec_vmwrite32(VMX_EXIT_CONTROLS, value32);
value32 = exec_vmread32(VMX_PROC_VM_EXEC_CONTROLS);
value32 &= ~VMX_PROCBASED_CTLS_TPR_SHADOW;
exec_vmwrite32(VMX_PROC_VM_EXEC_CONTROLS, value32);
exec_vmwrite32(VMX_TPR_THRESHOLD, 0U);
value32 = exec_vmread32(VMX_PROC_VM_EXEC_CONTROLS2);
value32 &= ~VMX_PROCBASED_CTLS2_VAPIC_REGS;
if (is_apicv_intr_delivery_supported()) {
value32 &= ~VMX_PROCBASED_CTLS2_VIRQ;
}
exec_vmwrite32(VMX_PROC_VM_EXEC_CONTROLS2, value32);
update_msr_bitmap_x2apic_passthru(vcpu);
} else {
value32 = exec_vmread32(VMX_PROC_VM_EXEC_CONTROLS2);
value32 &= ~VMX_PROCBASED_CTLS2_VAPIC;
value32 |= VMX_PROCBASED_CTLS2_VX2APIC;
exec_vmwrite32(VMX_PROC_VM_EXEC_CONTROLS2, value32);
update_msr_bitmap_x2apic_apicv(vcpu);
}
}
#endif
Reference in New Issue View Git Blame Copy Permalink