summaryrefslogtreecommitdiff
path: root/arch/x86/kvm/vmx.c
diff options
context:
space:
mode:
Diffstat (limited to 'arch/x86/kvm/vmx.c')
-rw-r--r--arch/x86/kvm/vmx.c235
1 files changed, 194 insertions, 41 deletions
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 16bb8e35605e..90b7eee6d0f9 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -27,6 +27,7 @@
#include <linux/mm.h>
#include <linux/highmem.h>
#include <linux/sched.h>
+#include <linux/sched/smt.h>
#include <linux/moduleparam.h>
#include <linux/mod_devicetable.h>
#include <linux/trace_events.h>
@@ -739,7 +740,6 @@ struct vcpu_vmx {
u64 msr_guest_kernel_gs_base;
#endif
- u64 arch_capabilities;
u64 spec_ctrl;
u32 vm_entry_controls_shadow;
@@ -2229,7 +2229,8 @@ static void add_atomic_switch_msr(struct vcpu_vmx *vmx, unsigned msr,
if (!entry_only)
j = find_msr(&m->host, msr);
- if (i == NR_AUTOLOAD_MSRS || j == NR_AUTOLOAD_MSRS) {
+ if ((i < 0 && m->guest.nr == NR_AUTOLOAD_MSRS) ||
+ (j < 0 && m->host.nr == NR_AUTOLOAD_MSRS)) {
printk_once(KERN_WARNING "Not enough msr switch entries. "
"Can't add msr %x\n", msr);
return;
@@ -3491,12 +3492,6 @@ static int vmx_get_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
msr_info->data = to_vmx(vcpu)->spec_ctrl;
break;
- case MSR_IA32_ARCH_CAPABILITIES:
- if (!msr_info->host_initiated &&
- !guest_cpuid_has(vcpu, X86_FEATURE_ARCH_CAPABILITIES))
- return 1;
- msr_info->data = to_vmx(vcpu)->arch_capabilities;
- break;
case MSR_IA32_SYSENTER_CS:
msr_info->data = vmcs_read32(GUEST_SYSENTER_CS);
break;
@@ -3661,11 +3656,6 @@ static int vmx_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
vmx_disable_intercept_for_msr(vmx->vmcs01.msr_bitmap, MSR_IA32_PRED_CMD,
MSR_TYPE_W);
break;
- case MSR_IA32_ARCH_CAPABILITIES:
- if (!msr_info->host_initiated)
- return 1;
- vmx->arch_capabilities = data;
- break;
case MSR_IA32_CR_PAT:
if (vmcs_config.vmentry_ctrl & VM_ENTRY_LOAD_IA32_PAT) {
if (!kvm_mtrr_valid(vcpu, MSR_IA32_CR_PAT, data))
@@ -5927,8 +5917,6 @@ static int vmx_vcpu_setup(struct vcpu_vmx *vmx)
++vmx->nmsrs;
}
- vmx->arch_capabilities = kvm_get_arch_capabilities();
-
vm_exit_controls_init(vmx, vmcs_config.vmexit_ctrl);
/* 22.2.1, 20.8.1 */
@@ -7444,25 +7432,50 @@ static int get_vmx_mem_address(struct kvm_vcpu *vcpu,
/* Addr = segment_base + offset */
/* offset = base + [index * scale] + displacement */
off = exit_qualification; /* holds the displacement */
+ if (addr_size == 1)
+ off = (gva_t)sign_extend64(off, 31);
+ else if (addr_size == 0)
+ off = (gva_t)sign_extend64(off, 15);
if (base_is_valid)
off += kvm_register_read(vcpu, base_reg);
if (index_is_valid)
off += kvm_register_read(vcpu, index_reg)<<scaling;
vmx_get_segment(vcpu, &s, seg_reg);
- *ret = s.base + off;
+ /*
+ * The effective address, i.e. @off, of a memory operand is truncated
+ * based on the address size of the instruction. Note that this is
+ * the *effective address*, i.e. the address prior to accounting for
+ * the segment's base.
+ */
if (addr_size == 1) /* 32 bit */
- *ret &= 0xffffffff;
+ off &= 0xffffffff;
+ else if (addr_size == 0) /* 16 bit */
+ off &= 0xffff;
/* Checks for #GP/#SS exceptions. */
exn = false;
if (is_long_mode(vcpu)) {
+ /*
+ * The virtual/linear address is never truncated in 64-bit
+ * mode, e.g. a 32-bit address size can yield a 64-bit virtual
+ * address when using FS/GS with a non-zero base.
+ */
+ *ret = s.base + off;
+
/* Long mode: #GP(0)/#SS(0) if the memory address is in a
* non-canonical form. This is the only check on the memory
* destination for long mode!
*/
exn = is_noncanonical_address(*ret, vcpu);
} else if (is_protmode(vcpu)) {
+ /*
+ * When not in long mode, the virtual/linear address is
+ * unconditionally truncated to 32 bits regardless of the
+ * address size.
+ */
+ *ret = (s.base + off) & 0xffffffff;
+
/* Protected mode: apply checks for segment validity in the
* following order:
* - segment type check (#GP(0) may be thrown)
@@ -7486,10 +7499,16 @@ static int get_vmx_mem_address(struct kvm_vcpu *vcpu,
/* Protected mode: #GP(0)/#SS(0) if the segment is unusable.
*/
exn = (s.unusable != 0);
- /* Protected mode: #GP(0)/#SS(0) if the memory
- * operand is outside the segment limit.
+
+ /*
+ * Protected mode: #GP(0)/#SS(0) if the memory operand is
+ * outside the segment limit. All CPUs that support VMX ignore
+ * limit checks for flat segments, i.e. segments with base==0,
+ * limit==0xffffffff and of type expand-up data or code.
*/
- exn = exn || (off + sizeof(u64) > s.limit);
+ if (!(s.base == 0 && s.limit == 0xffffffff &&
+ ((s.type & 8) || !(s.type & 4))))
+ exn = exn || (off + sizeof(u64) > s.limit);
}
if (exn) {
kvm_queue_exception_e(vcpu,
@@ -7708,6 +7727,7 @@ static void free_nested(struct vcpu_vmx *vmx)
if (!vmx->nested.vmxon)
return;
+ hrtimer_cancel(&vmx->nested.preemption_timer);
vmx->nested.vmxon = false;
free_vpid(vmx->nested.vpid02);
vmx->nested.posted_intr_nv = -1;
@@ -10119,7 +10139,7 @@ static int vmx_vm_init(struct kvm *kvm)
* Warn upon starting the first VM in a potentially
* insecure environment.
*/
- if (cpu_smt_control == CPU_SMT_ENABLED)
+ if (sched_smt_active())
pr_warn_once(L1TF_MSG_SMT);
if (l1tf_vmx_mitigation == VMENTER_L1D_FLUSH_NEVER)
pr_warn_once(L1TF_MSG_L1D);
@@ -11826,24 +11846,6 @@ static void prepare_vmcs12(struct kvm_vcpu *vcpu, struct vmcs12 *vmcs12,
kvm_clear_interrupt_queue(vcpu);
}
-static void load_vmcs12_mmu_host_state(struct kvm_vcpu *vcpu,
- struct vmcs12 *vmcs12)
-{
- u32 entry_failure_code;
-
- nested_ept_uninit_mmu_context(vcpu);
-
- /*
- * Only PDPTE load can fail as the value of cr3 was checked on entry and
- * couldn't have changed.
- */
- if (nested_vmx_load_cr3(vcpu, vmcs12->host_cr3, false, &entry_failure_code))
- nested_vmx_abort(vcpu, VMX_ABORT_LOAD_HOST_PDPTE_FAIL);
-
- if (!enable_ept)
- vcpu->arch.walk_mmu->inject_page_fault = kvm_inject_page_fault;
-}
-
/*
* A part of what we need to when the nested L2 guest exits and we want to
* run its L1 parent, is to reset L1's guest state to the host state specified
@@ -11857,6 +11859,7 @@ static void load_vmcs12_host_state(struct kvm_vcpu *vcpu,
struct vmcs12 *vmcs12)
{
struct kvm_segment seg;
+ u32 entry_failure_code;
if (vmcs12->vm_exit_controls & VM_EXIT_LOAD_IA32_EFER)
vcpu->arch.efer = vmcs12->host_ia32_efer;
@@ -11883,7 +11886,17 @@ static void load_vmcs12_host_state(struct kvm_vcpu *vcpu,
vcpu->arch.cr4_guest_owned_bits = ~vmcs_readl(CR4_GUEST_HOST_MASK);
vmx_set_cr4(vcpu, vmcs12->host_cr4);
- load_vmcs12_mmu_host_state(vcpu, vmcs12);
+ nested_ept_uninit_mmu_context(vcpu);
+
+ /*
+ * Only PDPTE load can fail as the value of cr3 was checked on entry and
+ * couldn't have changed.
+ */
+ if (nested_vmx_load_cr3(vcpu, vmcs12->host_cr3, false, &entry_failure_code))
+ nested_vmx_abort(vcpu, VMX_ABORT_LOAD_HOST_PDPTE_FAIL);
+
+ if (!enable_ept)
+ vcpu->arch.walk_mmu->inject_page_fault = kvm_inject_page_fault;
if (enable_vpid) {
/*
@@ -11974,6 +11987,140 @@ static void load_vmcs12_host_state(struct kvm_vcpu *vcpu,
nested_vmx_abort(vcpu, VMX_ABORT_LOAD_HOST_MSR_FAIL);
}
+static inline u64 nested_vmx_get_vmcs01_guest_efer(struct vcpu_vmx *vmx)
+{
+ struct shared_msr_entry *efer_msr;
+ unsigned int i;
+
+ if (vm_entry_controls_get(vmx) & VM_ENTRY_LOAD_IA32_EFER)
+ return vmcs_read64(GUEST_IA32_EFER);
+
+ if (cpu_has_load_ia32_efer)
+ return host_efer;
+
+ for (i = 0; i < vmx->msr_autoload.guest.nr; ++i) {
+ if (vmx->msr_autoload.guest.val[i].index == MSR_EFER)
+ return vmx->msr_autoload.guest.val[i].value;
+ }
+
+ efer_msr = find_msr_entry(vmx, MSR_EFER);
+ if (efer_msr)
+ return efer_msr->data;
+
+ return host_efer;
+}
+
+static void nested_vmx_restore_host_state(struct kvm_vcpu *vcpu)
+{
+ struct vmcs12 *vmcs12 = get_vmcs12(vcpu);
+ struct vcpu_vmx *vmx = to_vmx(vcpu);
+ struct vmx_msr_entry g, h;
+ struct msr_data msr;
+ gpa_t gpa;
+ u32 i, j;
+
+ vcpu->arch.pat = vmcs_read64(GUEST_IA32_PAT);
+
+ if (vmcs12->vm_entry_controls & VM_ENTRY_LOAD_DEBUG_CONTROLS) {
+ /*
+ * L1's host DR7 is lost if KVM_GUESTDBG_USE_HW_BP is set
+ * as vmcs01.GUEST_DR7 contains a userspace defined value
+ * and vcpu->arch.dr7 is not squirreled away before the
+ * nested VMENTER (not worth adding a variable in nested_vmx).
+ */
+ if (vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP)
+ kvm_set_dr(vcpu, 7, DR7_FIXED_1);
+ else
+ WARN_ON(kvm_set_dr(vcpu, 7, vmcs_readl(GUEST_DR7)));
+ }
+
+ /*
+ * Note that calling vmx_set_{efer,cr0,cr4} is important as they
+ * handle a variety of side effects to KVM's software model.
+ */
+ vmx_set_efer(vcpu, nested_vmx_get_vmcs01_guest_efer(vmx));
+
+ vcpu->arch.cr0_guest_owned_bits = X86_CR0_TS;
+ vmx_set_cr0(vcpu, vmcs_readl(CR0_READ_SHADOW));
+
+ vcpu->arch.cr4_guest_owned_bits = ~vmcs_readl(CR4_GUEST_HOST_MASK);
+ vmx_set_cr4(vcpu, vmcs_readl(CR4_READ_SHADOW));
+
+ nested_ept_uninit_mmu_context(vcpu);
+ vcpu->arch.cr3 = vmcs_readl(GUEST_CR3);
+ __set_bit(VCPU_EXREG_CR3, (ulong *)&vcpu->arch.regs_avail);
+
+ /*
+ * Use ept_save_pdptrs(vcpu) to load the MMU's cached PDPTRs
+ * from vmcs01 (if necessary). The PDPTRs are not loaded on
+ * VMFail, like everything else we just need to ensure our
+ * software model is up-to-date.
+ */
+ ept_save_pdptrs(vcpu);
+
+ kvm_mmu_reset_context(vcpu);
+
+ if (cpu_has_vmx_msr_bitmap())
+ vmx_update_msr_bitmap(vcpu);
+
+ /*
+ * This nasty bit of open coding is a compromise between blindly
+ * loading L1's MSRs using the exit load lists (incorrect emulation
+ * of VMFail), leaving the nested VM's MSRs in the software model
+ * (incorrect behavior) and snapshotting the modified MSRs (too
+ * expensive since the lists are unbound by hardware). For each
+ * MSR that was (prematurely) loaded from the nested VMEntry load
+ * list, reload it from the exit load list if it exists and differs
+ * from the guest value. The intent is to stuff host state as
+ * silently as possible, not to fully process the exit load list.
+ */
+ msr.host_initiated = false;
+ for (i = 0; i < vmcs12->vm_entry_msr_load_count; i++) {
+ gpa = vmcs12->vm_entry_msr_load_addr + (i * sizeof(g));
+ if (kvm_vcpu_read_guest(vcpu, gpa, &g, sizeof(g))) {
+ pr_debug_ratelimited(
+ "%s read MSR index failed (%u, 0x%08llx)\n",
+ __func__, i, gpa);
+ goto vmabort;
+ }
+
+ for (j = 0; j < vmcs12->vm_exit_msr_load_count; j++) {
+ gpa = vmcs12->vm_exit_msr_load_addr + (j * sizeof(h));
+ if (kvm_vcpu_read_guest(vcpu, gpa, &h, sizeof(h))) {
+ pr_debug_ratelimited(
+ "%s read MSR failed (%u, 0x%08llx)\n",
+ __func__, j, gpa);
+ goto vmabort;
+ }
+ if (h.index != g.index)
+ continue;
+ if (h.value == g.value)
+ break;
+
+ if (nested_vmx_load_msr_check(vcpu, &h)) {
+ pr_debug_ratelimited(
+ "%s check failed (%u, 0x%x, 0x%x)\n",
+ __func__, j, h.index, h.reserved);
+ goto vmabort;
+ }
+
+ msr.index = h.index;
+ msr.data = h.value;
+ if (kvm_set_msr(vcpu, &msr)) {
+ pr_debug_ratelimited(
+ "%s WRMSR failed (%u, 0x%x, 0x%llx)\n",
+ __func__, j, h.index, h.value);
+ goto vmabort;
+ }
+ }
+ }
+
+ return;
+
+vmabort:
+ nested_vmx_abort(vcpu, VMX_ABORT_LOAD_HOST_MSR_FAIL);
+}
+
/*
* Emulate an exit from nested guest (L2) to L1, i.e., prepare to run L1
* and modify vmcs12 to make it see what it would expect to see there if
@@ -12106,7 +12253,13 @@ static void nested_vmx_vmexit(struct kvm_vcpu *vcpu, u32 exit_reason,
*/
nested_vmx_failValid(vcpu, VMXERR_ENTRY_INVALID_CONTROL_FIELD);
- load_vmcs12_mmu_host_state(vcpu, vmcs12);
+ /*
+ * Restore L1's host state to KVM's software model. We're here
+ * because a consistency check was caught by hardware, which
+ * means some amount of guest state has been propagated to KVM's
+ * model and needs to be unwound to the host's state.
+ */
+ nested_vmx_restore_host_state(vcpu);
/*
* The emulated instruction was already skipped in
generated by cgit v1.2.3 (git 2.0.4) at 2025-05-13 17:16:11 +0000