From: Makarand Sonare <makarandsonare@google.com>
To: kvm@vger.kernel.org, pshier@google.com, jmattson@google.com
Cc: Makarand Sonare <makarandsonare@google.com>
Subject: [PATCH 1/2 v4] KVM: nVMX: Fix VMX preemption timer migration
Date: Thu, 21 May 2020 21:36:33 -0700 [thread overview]
Message-ID: <20200522043634.79779-2-makarandsonare@google.com> (raw)
In-Reply-To: <20200522043634.79779-1-makarandsonare@google.com>
From: Peter Shier <pshier@google.com>
Add new field to hold preemption timer expiration deadline
appended to struct kvm_vmx_nested_state_hdr. This is to prevent
the first VM-Enter after migration from incorrectly restarting the timer
with the full timer value instead of partially decayed timer value.
KVM_SET_NESTED_STATE restarts timer using migrated state regardless
of whether L1 sets VM_EXIT_SAVE_VMX_PREEMPTION_TIMER.
Fixes: cf8b84f48a593 ("kvm: nVMX: Prepare for checkpointing L2 state")
Signed-off-by: Peter Shier <pshier@google.com>
Signed-off-by: Makarand Sonare <makarandsonare@google.com>
---
Documentation/virt/kvm/api.rst | 5 +++
arch/x86/include/uapi/asm/kvm.h | 4 +++
arch/x86/kvm/vmx/nested.c | 58 ++++++++++++++++++++++++++++-----
arch/x86/kvm/vmx/vmx.h | 2 ++
arch/x86/kvm/x86.c | 1 +
include/uapi/linux/kvm.h | 1 +
6 files changed, 62 insertions(+), 9 deletions(-)
diff --git a/Documentation/virt/kvm/api.rst b/Documentation/virt/kvm/api.rst
index d871dacb984e9..a5ebe88cb7348 100644
--- a/Documentation/virt/kvm/api.rst
+++ b/Documentation/virt/kvm/api.rst
@@ -4326,6 +4326,9 @@ Errors:
#define KVM_STATE_NESTED_RUN_PENDING 0x00000002
#define KVM_STATE_NESTED_EVMCS 0x00000004
+ /* Available with KVM_CAP_NESTED_STATE_PREEMPTION_TIMER */
+ #define KVM_STATE_NESTED_PREEMPTION_TIMER 0x00000010
+
#define KVM_STATE_NESTED_FORMAT_VMX 0
#define KVM_STATE_NESTED_FORMAT_SVM 1
@@ -4335,8 +4338,10 @@ Errors:
#define KVM_STATE_NESTED_VMX_SMM_VMXON 0x00000002
struct kvm_vmx_nested_state_hdr {
+ __u32 flags;
__u64 vmxon_pa;
__u64 vmcs12_pa;
+ __u64 preemption_timer_deadline;
struct {
__u16 flags;
diff --git a/arch/x86/include/uapi/asm/kvm.h b/arch/x86/include/uapi/asm/kvm.h
index 3f3f780c8c650..ae7ccb5e8747d 100644
--- a/arch/x86/include/uapi/asm/kvm.h
+++ b/arch/x86/include/uapi/asm/kvm.h
@@ -391,6 +391,8 @@ struct kvm_sync_regs {
#define KVM_STATE_NESTED_RUN_PENDING 0x00000002
#define KVM_STATE_NESTED_EVMCS 0x00000004
#define KVM_STATE_NESTED_MTF_PENDING 0x00000008
+/* Available with KVM_CAP_NESTED_STATE_PREEMPTION_TIMER */
+#define KVM_STATE_NESTED_PREEMPTION_TIMER 0x00000010
#define KVM_STATE_NESTED_SMM_GUEST_MODE 0x00000001
#define KVM_STATE_NESTED_SMM_VMXON 0x00000002
@@ -403,8 +405,10 @@ struct kvm_vmx_nested_state_data {
};
struct kvm_vmx_nested_state_hdr {
+ __u32 flags;
__u64 vmxon_pa;
__u64 vmcs12_pa;
+ __u64 preemption_timer_deadline;
struct {
__u16 flags;
diff --git a/arch/x86/kvm/vmx/nested.c b/arch/x86/kvm/vmx/nested.c
index 51ebb60e1533a..a50cb99c4ed01 100644
--- a/arch/x86/kvm/vmx/nested.c
+++ b/arch/x86/kvm/vmx/nested.c
@@ -2092,9 +2092,29 @@ static enum hrtimer_restart vmx_preemption_timer_fn(struct hrtimer *timer)
return HRTIMER_NORESTART;
}
-static void vmx_start_preemption_timer(struct kvm_vcpu *vcpu)
+static u64 vmx_calc_preemption_timer_value(struct kvm_vcpu *vcpu)
+{
+ struct vcpu_vmx *vmx = to_vmx(vcpu);
+ struct vmcs12 *vmcs12 = get_vmcs12(vcpu);
+ u64 timer_value = 0;
+
+ u64 l1_scaled_tsc = kvm_read_l1_tsc(vcpu, rdtsc()) >>
+ VMX_MISC_EMULATED_PREEMPTION_TIMER_RATE;
+
+ if (!vmx->nested.has_preemption_timer_deadline) {
+ timer_value = vmcs12->vmx_preemption_timer_value;
+ vmx->nested.preemption_timer_deadline = timer_value +
+ l1_scaled_tsc;
+ vmx->nested.has_preemption_timer_deadline = true;
+ } else if (l1_scaled_tsc < vmx->nested.preemption_timer_deadline)
+ timer_value = vmx->nested.preemption_timer_deadline -
+ l1_scaled_tsc;
+ return timer_value;
+}
+
+static void vmx_start_preemption_timer(struct kvm_vcpu *vcpu,
+ u64 preemption_timeout)
{
- u64 preemption_timeout = get_vmcs12(vcpu)->vmx_preemption_timer_value;
struct vcpu_vmx *vmx = to_vmx(vcpu);
/*
@@ -3353,8 +3373,10 @@ enum nvmx_vmentry_status nested_vmx_enter_non_root_mode(struct kvm_vcpu *vcpu,
* the timer.
*/
vmx->nested.preemption_timer_expired = false;
- if (nested_cpu_has_preemption_timer(vmcs12))
- vmx_start_preemption_timer(vcpu);
+ if (nested_cpu_has_preemption_timer(vmcs12)) {
+ u64 timer_value = vmx_calc_preemption_timer_value(vcpu);
+ vmx_start_preemption_timer(vcpu, timer_value);
+ }
/*
* Note no nested_vmx_succeed or nested_vmx_fail here. At this point
@@ -3462,6 +3484,7 @@ static int nested_vmx_run(struct kvm_vcpu *vcpu, bool launch)
* the nested entry.
*/
vmx->nested.nested_run_pending = 1;
+ vmx->nested.has_preemption_timer_deadline = false;
status = nested_vmx_enter_non_root_mode(vcpu, true);
if (unlikely(status != NVMX_VMENTRY_SUCCESS))
goto vmentry_failed;
@@ -3962,9 +3985,10 @@ static void sync_vmcs02_to_vmcs12(struct kvm_vcpu *vcpu, struct vmcs12 *vmcs12)
vmcs12->guest_activity_state = GUEST_ACTIVITY_ACTIVE;
if (nested_cpu_has_preemption_timer(vmcs12) &&
- vmcs12->vm_exit_controls & VM_EXIT_SAVE_VMX_PREEMPTION_TIMER)
- vmcs12->vmx_preemption_timer_value =
- vmx_get_preemption_timer_value(vcpu);
+ vmcs12->vm_exit_controls & VM_EXIT_SAVE_VMX_PREEMPTION_TIMER &&
+ !vmx->nested.nested_run_pending)
+ vmcs12->vmx_preemption_timer_value =
+ vmx_get_preemption_timer_value(vcpu);
/*
* In some cases (usually, nested EPT), L2 is allowed to change its
@@ -5896,8 +5920,10 @@ static int vmx_get_nested_state(struct kvm_vcpu *vcpu,
.flags = 0,
.format = KVM_STATE_NESTED_FORMAT_VMX,
.size = sizeof(kvm_state),
+ .hdr.vmx.flags = 0,
.hdr.vmx.vmxon_pa = -1ull,
.hdr.vmx.vmcs12_pa = -1ull,
+ .hdr.vmx.preemption_timer_deadline = 0,
};
struct kvm_vmx_nested_state_data __user *user_vmx_nested_state =
&user_kvm_nested_state->data.vmx[0];
@@ -5939,6 +5965,14 @@ static int vmx_get_nested_state(struct kvm_vcpu *vcpu,
if (vmx->nested.mtf_pending)
kvm_state.flags |= KVM_STATE_NESTED_MTF_PENDING;
+
+ if (nested_cpu_has_preemption_timer(vmcs12) &&
+ vmx->nested.has_preemption_timer_deadline) {
+ kvm_state.hdr.vmx.flags |=
+ KVM_STATE_NESTED_PREEMPTION_TIMER;
+ kvm_state.hdr.vmx.preemption_timer_deadline =
+ vmx->nested.preemption_timer_deadline;
+ }
}
}
@@ -5984,7 +6018,6 @@ static int vmx_get_nested_state(struct kvm_vcpu *vcpu,
get_shadow_vmcs12(vcpu), VMCS12_SIZE))
return -EFAULT;
}
-
out:
return kvm_state.size;
}
@@ -6056,7 +6089,8 @@ static int vmx_set_nested_state(struct kvm_vcpu *vcpu,
*/
if (is_smm(vcpu) ?
(kvm_state->flags &
- (KVM_STATE_NESTED_GUEST_MODE | KVM_STATE_NESTED_RUN_PENDING))
+ (KVM_STATE_NESTED_GUEST_MODE | KVM_STATE_NESTED_RUN_PENDING |
+ KVM_STATE_NESTED_PREEMPTION_TIMER))
: kvm_state->hdr.vmx.smm.flags)
return -EINVAL;
@@ -6146,6 +6180,12 @@ static int vmx_set_nested_state(struct kvm_vcpu *vcpu,
goto error_guest_mode;
}
+ if (kvm_state->hdr.vmx.flags & KVM_STATE_NESTED_PREEMPTION_TIMER) {
+ vmx->nested.has_preemption_timer_deadline = true;
+ vmx->nested.preemption_timer_deadline =
+ kvm_state->hdr.vmx.preemption_timer_deadline;
+ }
+
if (nested_vmx_check_controls(vcpu, vmcs12) ||
nested_vmx_check_host_state(vcpu, vmcs12) ||
nested_vmx_check_guest_state(vcpu, vmcs12, &ignored))
diff --git a/arch/x86/kvm/vmx/vmx.h b/arch/x86/kvm/vmx/vmx.h
index 298ddef79d009..672c28f17e497 100644
--- a/arch/x86/kvm/vmx/vmx.h
+++ b/arch/x86/kvm/vmx/vmx.h
@@ -169,6 +169,8 @@ struct nested_vmx {
u16 posted_intr_nv;
struct hrtimer preemption_timer;
+ u64 preemption_timer_deadline;
+ bool has_preemption_timer_deadline;
bool preemption_timer_expired;
/* to migrate it to L2 if VM_ENTRY_LOAD_DEBUG_CONTROLS is off */
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 471fccf7f8501..e73309879255c 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -3418,6 +3418,7 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
case KVM_CAP_MSR_PLATFORM_INFO:
case KVM_CAP_EXCEPTION_PAYLOAD:
case KVM_CAP_SET_GUEST_DEBUG:
+ case KVM_CAP_NESTED_STATE_PREEMPTION_TIMER:
r = 1;
break;
case KVM_CAP_SYNC_REGS:
diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h
index ac9eba0289d1b..0868dce12a715 100644
--- a/include/uapi/linux/kvm.h
+++ b/include/uapi/linux/kvm.h
@@ -1018,6 +1018,7 @@ struct kvm_ppc_resize_hpt {
#define KVM_CAP_S390_PROTECTED 180
#define KVM_CAP_PPC_SECURE_GUEST 181
#define KVM_CAP_HALT_POLL 182
+#define KVM_CAP_NESTED_STATE_PREEMPTION_TIMER 183
#ifdef KVM_CAP_IRQ_ROUTING
--
2.27.0.rc0.183.gde8f92d652-goog
next prev parent reply other threads:[~2020-05-22 4:37 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-05-22 4:36 [PATCH 0/2 v4] Fix VMX preemption timer migration Makarand Sonare
2020-05-22 4:36 ` Makarand Sonare [this message]
2020-05-22 7:19 ` [PATCH 1/2 v4] KVM: nVMX: " Paolo Bonzini
2020-05-22 4:36 ` [PATCH 2/2 v4] KVM: selftests: VMX preemption timer migration test Makarand Sonare
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200522043634.79779-2-makarandsonare@google.com \
--to=makarandsonare@google.com \
--cc=jmattson@google.com \
--cc=kvm@vger.kernel.org \
--cc=pshier@google.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).