aboutsummaryrefslogtreecommitdiffstats
path: root/arch/x86/kvm/vmx
diff options
context:
space:
mode:
Diffstat (limited to 'arch/x86/kvm/vmx')
-rw-r--r--arch/x86/kvm/vmx/nested.c150
-rw-r--r--arch/x86/kvm/vmx/pmu_intel.c31
-rw-r--r--arch/x86/kvm/vmx/vmcs12.h5
-rw-r--r--arch/x86/kvm/vmx/vmx.c31
-rw-r--r--arch/x86/kvm/vmx/vmx.h1
5 files changed, 131 insertions, 87 deletions
diff --git a/arch/x86/kvm/vmx/nested.c b/arch/x86/kvm/vmx/nested.c
index f1a69117ac0f..46af3a5e9209 100644
--- a/arch/x86/kvm/vmx/nested.c
+++ b/arch/x86/kvm/vmx/nested.c
@@ -1397,7 +1397,7 @@ static int copy_enlightened_to_vmcs12(struct vcpu_vmx *vmx)
}
if (unlikely(!(evmcs->hv_clean_fields &
- HV_VMX_ENLIGHTENED_CLEAN_FIELD_CONTROL_PROC))) {
+ HV_VMX_ENLIGHTENED_CLEAN_FIELD_CONTROL_EXCPN))) {
vmcs12->exception_bitmap = evmcs->exception_bitmap;
}
@@ -1437,7 +1437,7 @@ static int copy_enlightened_to_vmcs12(struct vcpu_vmx *vmx)
}
if (unlikely(!(evmcs->hv_clean_fields &
- HV_VMX_ENLIGHTENED_CLEAN_FIELD_HOST_GRP1))) {
+ HV_VMX_ENLIGHTENED_CLEAN_FIELD_CONTROL_GRP1))) {
vmcs12->pin_based_vm_exec_control =
evmcs->pin_based_vm_exec_control;
vmcs12->vm_exit_controls = evmcs->vm_exit_controls;
@@ -2784,14 +2784,13 @@ static int nested_vmx_check_vmentry_hw(struct kvm_vcpu *vcpu)
: "cc", "memory"
);
- preempt_enable();
-
if (vmx->msr_autoload.host.nr)
vmcs_write32(VM_EXIT_MSR_LOAD_COUNT, vmx->msr_autoload.host.nr);
if (vmx->msr_autoload.guest.nr)
vmcs_write32(VM_ENTRY_MSR_LOAD_COUNT, vmx->msr_autoload.guest.nr);
if (vm_fail) {
+ preempt_enable();
WARN_ON_ONCE(vmcs_read32(VM_INSTRUCTION_ERROR) !=
VMXERR_ENTRY_INVALID_CONTROL_FIELD);
return 1;
@@ -2803,6 +2802,7 @@ static int nested_vmx_check_vmentry_hw(struct kvm_vcpu *vcpu)
local_irq_enable();
if (hw_breakpoint_active())
set_debugreg(__this_cpu_read(cpu_dr7), 7);
+ preempt_enable();
/*
* A non-failing VMEntry means we somehow entered guest mode with
@@ -5226,40 +5226,42 @@ static int vmx_get_nested_state(struct kvm_vcpu *vcpu,
struct vmcs12 *vmcs12;
struct kvm_nested_state kvm_state = {
.flags = 0,
- .format = 0,
+ .format = KVM_STATE_NESTED_FORMAT_VMX,
.size = sizeof(kvm_state),
- .vmx.vmxon_pa = -1ull,
- .vmx.vmcs_pa = -1ull,
+ .hdr.vmx.vmxon_pa = -1ull,
+ .hdr.vmx.vmcs12_pa = -1ull,
};
+ struct kvm_vmx_nested_state_data __user *user_vmx_nested_state =
+ &user_kvm_nested_state->data.vmx[0];
if (!vcpu)
- return kvm_state.size + 2 * VMCS12_SIZE;
+ return kvm_state.size + sizeof(*user_vmx_nested_state);
vmx = to_vmx(vcpu);
vmcs12 = get_vmcs12(vcpu);
- if (nested_vmx_allowed(vcpu) && vmx->nested.enlightened_vmcs_enabled)
- kvm_state.flags |= KVM_STATE_NESTED_EVMCS;
-
if (nested_vmx_allowed(vcpu) &&
(vmx->nested.vmxon || vmx->nested.smm.vmxon)) {
- kvm_state.vmx.vmxon_pa = vmx->nested.vmxon_ptr;
- kvm_state.vmx.vmcs_pa = vmx->nested.current_vmptr;
+ kvm_state.hdr.vmx.vmxon_pa = vmx->nested.vmxon_ptr;
+ kvm_state.hdr.vmx.vmcs12_pa = vmx->nested.current_vmptr;
if (vmx_has_valid_vmcs12(vcpu)) {
- kvm_state.size += VMCS12_SIZE;
+ kvm_state.size += sizeof(user_vmx_nested_state->vmcs12);
+
+ if (vmx->nested.hv_evmcs)
+ kvm_state.flags |= KVM_STATE_NESTED_EVMCS;
if (is_guest_mode(vcpu) &&
nested_cpu_has_shadow_vmcs(vmcs12) &&
vmcs12->vmcs_link_pointer != -1ull)
- kvm_state.size += VMCS12_SIZE;
+ kvm_state.size += sizeof(user_vmx_nested_state->shadow_vmcs12);
}
if (vmx->nested.smm.vmxon)
- kvm_state.vmx.smm.flags |= KVM_STATE_NESTED_SMM_VMXON;
+ kvm_state.hdr.vmx.smm.flags |= KVM_STATE_NESTED_SMM_VMXON;
if (vmx->nested.smm.guest_mode)
- kvm_state.vmx.smm.flags |= KVM_STATE_NESTED_SMM_GUEST_MODE;
+ kvm_state.hdr.vmx.smm.flags |= KVM_STATE_NESTED_SMM_GUEST_MODE;
if (is_guest_mode(vcpu)) {
kvm_state.flags |= KVM_STATE_NESTED_GUEST_MODE;
@@ -5294,16 +5296,19 @@ static int vmx_get_nested_state(struct kvm_vcpu *vcpu,
copy_shadow_to_vmcs12(vmx);
}
+ BUILD_BUG_ON(sizeof(user_vmx_nested_state->vmcs12) < VMCS12_SIZE);
+ BUILD_BUG_ON(sizeof(user_vmx_nested_state->shadow_vmcs12) < VMCS12_SIZE);
+
/*
* Copy over the full allocated size of vmcs12 rather than just the size
* of the struct.
*/
- if (copy_to_user(user_kvm_nested_state->data, vmcs12, VMCS12_SIZE))
+ if (copy_to_user(user_vmx_nested_state->vmcs12, vmcs12, VMCS12_SIZE))
return -EFAULT;
if (nested_cpu_has_shadow_vmcs(vmcs12) &&
vmcs12->vmcs_link_pointer != -1ull) {
- if (copy_to_user(user_kvm_nested_state->data + VMCS12_SIZE,
+ if (copy_to_user(user_vmx_nested_state->shadow_vmcs12,
get_shadow_vmcs12(vcpu), VMCS12_SIZE))
return -EFAULT;
}
@@ -5331,33 +5336,44 @@ static int vmx_set_nested_state(struct kvm_vcpu *vcpu,
struct vcpu_vmx *vmx = to_vmx(vcpu);
struct vmcs12 *vmcs12;
u32 exit_qual;
+ struct kvm_vmx_nested_state_data __user *user_vmx_nested_state =
+ &user_kvm_nested_state->data.vmx[0];
int ret;
- if (kvm_state->format != 0)
+ if (kvm_state->format != KVM_STATE_NESTED_FORMAT_VMX)
return -EINVAL;
- if (!nested_vmx_allowed(vcpu))
- return kvm_state->vmx.vmxon_pa == -1ull ? 0 : -EINVAL;
+ if (kvm_state->hdr.vmx.vmxon_pa == -1ull) {
+ if (kvm_state->hdr.vmx.smm.flags)
+ return -EINVAL;
- if (kvm_state->vmx.vmxon_pa == -1ull) {
- if (kvm_state->vmx.smm.flags)
+ if (kvm_state->hdr.vmx.vmcs12_pa != -1ull)
return -EINVAL;
- if (kvm_state->vmx.vmcs_pa != -1ull)
+ /*
+ * KVM_STATE_NESTED_EVMCS used to signal that KVM should
+ * enable eVMCS capability on vCPU. However, since then
+ * code was changed such that flag signals vmcs12 should
+ * be copied into eVMCS in guest memory.
+ *
+ * To preserve backwards compatability, allow user
+ * to set this flag even when there is no VMXON region.
+ */
+ if (kvm_state->flags & ~KVM_STATE_NESTED_EVMCS)
+ return -EINVAL;
+ } else {
+ if (!nested_vmx_allowed(vcpu))
return -EINVAL;
- vmx_leave_nested(vcpu);
- return 0;
+ if (!page_address_valid(vcpu, kvm_state->hdr.vmx.vmxon_pa))
+ return -EINVAL;
}
- if (!page_address_valid(vcpu, kvm_state->vmx.vmxon_pa))
- return -EINVAL;
-
- if ((kvm_state->vmx.smm.flags & KVM_STATE_NESTED_SMM_GUEST_MODE) &&
+ if ((kvm_state->hdr.vmx.smm.flags & KVM_STATE_NESTED_SMM_GUEST_MODE) &&
(kvm_state->flags & KVM_STATE_NESTED_GUEST_MODE))
return -EINVAL;
- if (kvm_state->vmx.smm.flags &
+ if (kvm_state->hdr.vmx.smm.flags &
~(KVM_STATE_NESTED_SMM_GUEST_MODE | KVM_STATE_NESTED_SMM_VMXON))
return -EINVAL;
@@ -5366,21 +5382,26 @@ static int vmx_set_nested_state(struct kvm_vcpu *vcpu,
* nor can VMLAUNCH/VMRESUME be pending. Outside SMM, SMM flags
* must be zero.
*/
- if (is_smm(vcpu) ? kvm_state->flags : kvm_state->vmx.smm.flags)
+ if (is_smm(vcpu) ?
+ (kvm_state->flags &
+ (KVM_STATE_NESTED_GUEST_MODE | KVM_STATE_NESTED_RUN_PENDING))
+ : kvm_state->hdr.vmx.smm.flags)
return -EINVAL;
- if ((kvm_state->vmx.smm.flags & KVM_STATE_NESTED_SMM_GUEST_MODE) &&
- !(kvm_state->vmx.smm.flags & KVM_STATE_NESTED_SMM_VMXON))
+ if ((kvm_state->hdr.vmx.smm.flags & KVM_STATE_NESTED_SMM_GUEST_MODE) &&
+ !(kvm_state->hdr.vmx.smm.flags & KVM_STATE_NESTED_SMM_VMXON))
return -EINVAL;
+ if ((kvm_state->flags & KVM_STATE_NESTED_EVMCS) &&
+ (!nested_vmx_allowed(vcpu) || !vmx->nested.enlightened_vmcs_enabled))
+ return -EINVAL;
+
vmx_leave_nested(vcpu);
- if (kvm_state->vmx.vmxon_pa == -1ull)
- return 0;
- if (kvm_state->flags & KVM_STATE_NESTED_EVMCS)
- nested_enable_evmcs(vcpu, NULL);
+ if (kvm_state->hdr.vmx.vmxon_pa == -1ull)
+ return 0;
- vmx->nested.vmxon_ptr = kvm_state->vmx.vmxon_pa;
+ vmx->nested.vmxon_ptr = kvm_state->hdr.vmx.vmxon_pa;
ret = enter_vmx_operation(vcpu);
if (ret)
return ret;
@@ -5389,12 +5410,12 @@ static int vmx_set_nested_state(struct kvm_vcpu *vcpu,
if (kvm_state->size < sizeof(*kvm_state) + sizeof(*vmcs12))
return 0;
- if (kvm_state->vmx.vmcs_pa != -1ull) {
- if (kvm_state->vmx.vmcs_pa == kvm_state->vmx.vmxon_pa ||
- !page_address_valid(vcpu, kvm_state->vmx.vmcs_pa))
+ if (kvm_state->hdr.vmx.vmcs12_pa != -1ull) {
+ if (kvm_state->hdr.vmx.vmcs12_pa == kvm_state->hdr.vmx.vmxon_pa ||
+ !page_address_valid(vcpu, kvm_state->hdr.vmx.vmcs12_pa))
return -EINVAL;
- set_current_vmptr(vmx, kvm_state->vmx.vmcs_pa);
+ set_current_vmptr(vmx, kvm_state->hdr.vmx.vmcs12_pa);
} else if (kvm_state->flags & KVM_STATE_NESTED_EVMCS) {
/*
* Sync eVMCS upon entry as we may not have
@@ -5405,16 +5426,16 @@ static int vmx_set_nested_state(struct kvm_vcpu *vcpu,
return -EINVAL;
}
- if (kvm_state->vmx.smm.flags & KVM_STATE_NESTED_SMM_VMXON) {
+ if (kvm_state->hdr.vmx.smm.flags & KVM_STATE_NESTED_SMM_VMXON) {
vmx->nested.smm.vmxon = true;
vmx->nested.vmxon = false;
- if (kvm_state->vmx.smm.flags & KVM_STATE_NESTED_SMM_GUEST_MODE)
+ if (kvm_state->hdr.vmx.smm.flags & KVM_STATE_NESTED_SMM_GUEST_MODE)
vmx->nested.smm.guest_mode = true;
}
vmcs12 = get_vmcs12(vcpu);
- if (copy_from_user(vmcs12, user_kvm_nested_state->data, sizeof(*vmcs12)))
+ if (copy_from_user(vmcs12, user_vmx_nested_state->vmcs12, sizeof(*vmcs12)))
return -EFAULT;
if (vmcs12->hdr.revision_id != VMCS12_REVISION)
@@ -5423,39 +5444,46 @@ static int vmx_set_nested_state(struct kvm_vcpu *vcpu,
if (!(kvm_state->flags & KVM_STATE_NESTED_GUEST_MODE))
return 0;
+ vmx->nested.nested_run_pending =
+ !!(kvm_state->flags & KVM_STATE_NESTED_RUN_PENDING);
+
+ ret = -EINVAL;
if (nested_cpu_has_shadow_vmcs(vmcs12) &&
vmcs12->vmcs_link_pointer != -1ull) {
struct vmcs12 *shadow_vmcs12 = get_shadow_vmcs12(vcpu);
- if (kvm_state->size < sizeof(*kvm_state) + 2 * sizeof(*vmcs12))
- return -EINVAL;
+ if (kvm_state->size <
+ sizeof(*kvm_state) +
+ sizeof(user_vmx_nested_state->vmcs12) + sizeof(*shadow_vmcs12))
+ goto error_guest_mode;
if (copy_from_user(shadow_vmcs12,
- user_kvm_nested_state->data + VMCS12_SIZE,
- sizeof(*vmcs12)))
- return -EFAULT;
+ user_vmx_nested_state->shadow_vmcs12,
+ sizeof(*shadow_vmcs12))) {
+ ret = -EFAULT;
+ goto error_guest_mode;
+ }
if (shadow_vmcs12->hdr.revision_id != VMCS12_REVISION ||
!shadow_vmcs12->hdr.shadow_vmcs)
- return -EINVAL;
+ goto error_guest_mode;
}
if (nested_vmx_check_controls(vcpu, vmcs12) ||
nested_vmx_check_host_state(vcpu, vmcs12) ||
nested_vmx_check_guest_state(vcpu, vmcs12, &exit_qual))
- return -EINVAL;
+ goto error_guest_mode;
vmx->nested.dirty_vmcs12 = true;
- vmx->nested.nested_run_pending =
- !!(kvm_state->flags & KVM_STATE_NESTED_RUN_PENDING);
-
ret = nested_vmx_enter_non_root_mode(vcpu, false);
- if (ret) {
- vmx->nested.nested_run_pending = 0;
- return -EINVAL;
- }
+ if (ret)
+ goto error_guest_mode;
return 0;
+
+error_guest_mode:
+ vmx->nested.nested_run_pending = 0;
+ return ret;
}
void nested_vmx_vcpu_setup(void)
diff --git a/arch/x86/kvm/vmx/pmu_intel.c b/arch/x86/kvm/vmx/pmu_intel.c
index f8502c376b37..68d231d49c7a 100644
--- a/arch/x86/kvm/vmx/pmu_intel.c
+++ b/arch/x86/kvm/vmx/pmu_intel.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0-only
/*
* KVM PMU support for Intel CPUs
*
@@ -6,10 +7,6 @@
* Authors:
* Avi Kivity <avi@redhat.com>
* Gleb Natapov <gleb@redhat.com>
- *
- * This work is licensed under the terms of the GNU GPL, version 2. See
- * the COPYING file in the top-level directory.
- *
*/
#include <linux/types.h>
#include <linux/kvm_host.h>
@@ -126,7 +123,7 @@ static int intel_is_valid_msr_idx(struct kvm_vcpu *vcpu, unsigned idx)
}
static struct kvm_pmc *intel_msr_idx_to_pmc(struct kvm_vcpu *vcpu,
- unsigned idx)
+ unsigned idx, u64 *mask)
{
struct kvm_pmu *pmu = vcpu_to_pmu(vcpu);
bool fixed = idx & (1u << 30);
@@ -138,6 +135,7 @@ static struct kvm_pmc *intel_msr_idx_to_pmc(struct kvm_vcpu *vcpu,
if (fixed && idx >= pmu->nr_arch_fixed_counters)
return NULL;
counters = fixed ? pmu->fixed_counters : pmu->gp_counters;
+ *mask &= pmu->counter_bitmask[fixed ? KVM_PMC_FIXED : KVM_PMC_GP];
return &counters[idx];
}
@@ -183,9 +181,13 @@ static int intel_pmu_get_msr(struct kvm_vcpu *vcpu, u32 msr, u64 *data)
*data = pmu->global_ovf_ctrl;
return 0;
default:
- if ((pmc = get_gp_pmc(pmu, msr, MSR_IA32_PERFCTR0)) ||
- (pmc = get_fixed_pmc(pmu, msr))) {
- *data = pmc_read_counter(pmc);
+ if ((pmc = get_gp_pmc(pmu, msr, MSR_IA32_PERFCTR0))) {
+ u64 val = pmc_read_counter(pmc);
+ *data = val & pmu->counter_bitmask[KVM_PMC_GP];
+ return 0;
+ } else if ((pmc = get_fixed_pmc(pmu, msr))) {
+ u64 val = pmc_read_counter(pmc);
+ *data = val & pmu->counter_bitmask[KVM_PMC_FIXED];
return 0;
} else if ((pmc = get_gp_pmc(pmu, msr, MSR_P6_EVNTSEL0))) {
*data = pmc->eventsel;
@@ -235,11 +237,14 @@ static int intel_pmu_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
}
break;
default:
- if ((pmc = get_gp_pmc(pmu, msr, MSR_IA32_PERFCTR0)) ||
- (pmc = get_fixed_pmc(pmu, msr))) {
- if (!msr_info->host_initiated)
- data = (s64)(s32)data;
- pmc->counter += data - pmc_read_counter(pmc);
+ if ((pmc = get_gp_pmc(pmu, msr, MSR_IA32_PERFCTR0))) {
+ if (msr_info->host_initiated)
+ pmc->counter = data;
+ else
+ pmc->counter = (s32)data;
+ return 0;
+ } else if ((pmc = get_fixed_pmc(pmu, msr))) {
+ pmc->counter = data;
return 0;
} else if ((pmc = get_gp_pmc(pmu, msr, MSR_P6_EVNTSEL0))) {
if (data == pmc->eventsel)
diff --git a/arch/x86/kvm/vmx/vmcs12.h b/arch/x86/kvm/vmx/vmcs12.h
index 3a742428ad17..337718fc8a36 100644
--- a/arch/x86/kvm/vmx/vmcs12.h
+++ b/arch/x86/kvm/vmx/vmcs12.h
@@ -201,9 +201,10 @@ struct __packed vmcs12 {
/*
* VMCS12_SIZE is the number of bytes L1 should allocate for the VMXON region
* and any VMCS region. Although only sizeof(struct vmcs12) are used by the
- * current implementation, 4K are reserved to avoid future complications.
+ * current implementation, 4K are reserved to avoid future complications and
+ * to preserve userspace ABI.
*/
-#define VMCS12_SIZE 0x1000
+#define VMCS12_SIZE KVM_STATE_NESTED_VMX_VMCS_SIZE
/*
* VMCS12_MAX_FIELD_INDEX is the highest index value used in any
diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
index 1ac167614032..d98eac371c0a 100644
--- a/arch/x86/kvm/vmx/vmx.c
+++ b/arch/x86/kvm/vmx/vmx.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0-only
/*
* Kernel-based Virtual Machine driver for Linux
*
@@ -10,10 +11,6 @@
* Authors:
* Avi Kivity <avi@qumranet.com>
* Yaniv Kamay <yaniv@qumranet.com>
- *
- * This work is licensed under the terms of the GNU GPL, version 2. See
- * the COPYING file in the top-level directory.
- *
*/
#include <linux/frame.h>
@@ -114,6 +111,9 @@ static u64 __read_mostly host_xss;
bool __read_mostly enable_pml = 1;
module_param_named(pml, enable_pml, bool, S_IRUGO);
+static bool __read_mostly dump_invalid_vmcs = 0;
+module_param(dump_invalid_vmcs, bool, 0644);
+
#define MSR_BITMAP_MODE_X2APIC 1
#define MSR_BITMAP_MODE_X2APIC_APICV 2
@@ -5607,15 +5607,24 @@ static void vmx_dump_dtsel(char *name, uint32_t limit)
void dump_vmcs(void)
{
- u32 vmentry_ctl = vmcs_read32(VM_ENTRY_CONTROLS);
- u32 vmexit_ctl = vmcs_read32(VM_EXIT_CONTROLS);
- u32 cpu_based_exec_ctrl = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL);
- u32 pin_based_exec_ctrl = vmcs_read32(PIN_BASED_VM_EXEC_CONTROL);
- u32 secondary_exec_control = 0;
- unsigned long cr4 = vmcs_readl(GUEST_CR4);
- u64 efer = vmcs_read64(GUEST_IA32_EFER);
+ u32 vmentry_ctl, vmexit_ctl;
+ u32 cpu_based_exec_ctrl, pin_based_exec_ctrl, secondary_exec_control;
+ unsigned long cr4;
+ u64 efer;
int i, n;
+ if (!dump_invalid_vmcs) {
+ pr_warn_ratelimited("set kvm_intel.dump_invalid_vmcs=1 to dump internal KVM state.\n");
+ return;
+ }
+
+ vmentry_ctl = vmcs_read32(VM_ENTRY_CONTROLS);
+ vmexit_ctl = vmcs_read32(VM_EXIT_CONTROLS);
+ cpu_based_exec_ctrl = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL);
+ pin_based_exec_ctrl = vmcs_read32(PIN_BASED_VM_EXEC_CONTROL);
+ cr4 = vmcs_readl(GUEST_CR4);
+ efer = vmcs_read64(GUEST_IA32_EFER);
+ secondary_exec_control = 0;
if (cpu_has_secondary_exec_ctrls())
secondary_exec_control = vmcs_read32(SECONDARY_VM_EXEC_CONTROL);
diff --git a/arch/x86/kvm/vmx/vmx.h b/arch/x86/kvm/vmx/vmx.h
index 63d37ccce3dc..61128b48c503 100644
--- a/arch/x86/kvm/vmx/vmx.h
+++ b/arch/x86/kvm/vmx/vmx.h
@@ -319,6 +319,7 @@ void vmx_set_nmi_mask(struct kvm_vcpu *vcpu, bool masked);
void vmx_set_virtual_apic_mode(struct kvm_vcpu *vcpu);
struct shared_msr_entry *find_msr_entry(struct vcpu_vmx *vmx, u32 msr);
void pt_update_intercept_for_msr(struct vcpu_vmx *vmx);
+void vmx_update_host_rsp(struct vcpu_vmx *vmx, unsigned long host_rsp);
#define POSTED_INTR_ON 0
#define POSTED_INTR_SN 1
Copyright © 1996 – 2023 Jason A. Donenfeld. All Rights Reverse Engineered.