2019-09-20 23:03:16

by Suleiman Souhlal

[permalink] [raw]
Subject: [RFC 0/2] kvm: Use host timekeeping in guest.

This RFC is to try to solve the following problem:

We have some applications that are currently running in their
own namespace, that still talk to other processes on the
machine, using IPC, and expect to run on the same machine.

We want to move them into a virtual machine, for the usual
benefits of virtualization.

However, some of these programs use CLOCK_MONOTONIC and
CLOCK_BOOTTIME timestamps, as part of their protocol, when talking
to the host.

Generally speaking, we have multiple event sources, for example
sensors, input devices, display controller vsync, etc and we would
like to rely on them in the guest for various scenarios.

As a specific example, we are trying to run some wayland clients
(in the guest) who talk to the server (in the host), and the server
gives input events based on host time. Additionally, there are also
vsync events that the clients use for timing their rendering.

Another use case we have are timestamps from IIO sensors and cameras.
There are applications that need to determine how the timestamps
relate to the current time and the only way to get current time is
clock_gettime(), which would return a value from a different time
domain than the timestamps.

In this case, it is not feasible to change these programs, due to
the number of the places we would have to change.

We spent some time thinking about this, and the best solution we
could come up with was the following:

Make the guest kernel return the same CLOCK_MONOTONIC and
CLOCK_GETTIME timestamps as the host.

To do that, I am changing kvmclock to request to the host to copy
its timekeeping parameters (mult, base, cycle_last, etc), so that
the guest timekeeper can use the same values, so that time can
be synchronized between the guest and the host.

Any suggestions or feedback would be highly appreciated.

Suleiman Souhlal (2):
kvm: Mechanism to copy host timekeeping parameters into guest.
x86/kvmclock: Use host timekeeping.

arch/x86/Kconfig | 9 ++
arch/x86/include/asm/kvm_host.h | 3 +
arch/x86/include/asm/kvmclock.h | 2 +
arch/x86/include/asm/pvclock-abi.h | 27 ++++++
arch/x86/include/uapi/asm/kvm_para.h | 1 +
arch/x86/kernel/kvmclock.c | 127 ++++++++++++++++++++++++++-
arch/x86/kvm/x86.c | 121 +++++++++++++++++++++++++
kernel/time/timekeeping.c | 21 +++++
8 files changed, 307 insertions(+), 4 deletions(-)

--
2.23.0.237.gc6a4ce50a0-goog


2019-09-20 23:22:38

by Suleiman Souhlal

[permalink] [raw]
Subject: [RFC 1/2] kvm: Mechanism to copy host timekeeping parameters into guest.

This is used to synchronize time between host and guest.
The guest can request the (guest) physical address it wants the
data in through the MSR_KVM_TIMEKEEPER_EN MSR.

We maintain a shadow copy of the timekeeper that gets updated
whenever the timekeeper gets updated, and then copied into the
guest.

It currently assumes the host timekeeper is "tsc".

Signed-off-by: Suleiman Souhlal <[email protected]>
---
arch/x86/include/asm/kvm_host.h | 3 +
arch/x86/include/asm/pvclock-abi.h | 27 ++++++
arch/x86/include/uapi/asm/kvm_para.h | 1 +
arch/x86/kvm/x86.c | 121 +++++++++++++++++++++++++++
4 files changed, 152 insertions(+)

diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index bdc16b0aa7c6..b1b4c3a80b8d 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -666,7 +666,10 @@ struct kvm_vcpu_arch {
struct pvclock_vcpu_time_info hv_clock;
unsigned int hw_tsc_khz;
struct gfn_to_hva_cache pv_time;
+ struct gfn_to_hva_cache pv_timekeeper_g2h;
+ struct pvclock_timekeeper pv_timekeeper;
bool pv_time_enabled;
+ bool pv_timekeeper_enabled;
/* set guest stopped flag in pvclock flags field */
bool pvclock_set_guest_stopped_request;

diff --git a/arch/x86/include/asm/pvclock-abi.h b/arch/x86/include/asm/pvclock-abi.h
index 1436226efe3e..2809008b9b26 100644
--- a/arch/x86/include/asm/pvclock-abi.h
+++ b/arch/x86/include/asm/pvclock-abi.h
@@ -40,6 +40,33 @@ struct pvclock_wall_clock {
u32 nsec;
} __attribute__((__packed__));

+struct pvclock_read_base {
+ u64 mask;
+ u64 cycle_last;
+ u32 mult;
+ u32 shift;
+ u64 xtime_nsec;
+ u64 base;
+} __attribute__((__packed__));
+
+struct pvclock_timekeeper {
+ u64 gen;
+ u64 flags;
+ struct pvclock_read_base tkr_mono;
+ struct pvclock_read_base tkr_raw;
+ u64 xtime_sec;
+ u64 ktime_sec;
+ u64 wall_to_monotonic_sec;
+ u64 wall_to_monotonic_nsec;
+ u64 offs_real;
+ u64 offs_boot;
+ u64 offs_tai;
+ u64 raw_sec;
+ u64 tsc_offset;
+} __attribute__((__packed__));
+
+#define PVCLOCK_TIMEKEEPER_ENABLED (1 << 0)
+
#define PVCLOCK_TSC_STABLE_BIT (1 << 0)
#define PVCLOCK_GUEST_STOPPED (1 << 1)
/* PVCLOCK_COUNTS_FROM_ZERO broke ABI and can't be used anymore. */
diff --git a/arch/x86/include/uapi/asm/kvm_para.h b/arch/x86/include/uapi/asm/kvm_para.h
index 2a8e0b6b9805..3ebb1d87db3a 100644
--- a/arch/x86/include/uapi/asm/kvm_para.h
+++ b/arch/x86/include/uapi/asm/kvm_para.h
@@ -50,6 +50,7 @@
#define MSR_KVM_STEAL_TIME 0x4b564d03
#define MSR_KVM_PV_EOI_EN 0x4b564d04
#define MSR_KVM_POLL_CONTROL 0x4b564d05
+#define MSR_KVM_TIMEKEEPER_EN 0x4b564d06

struct kvm_steal_time {
__u64 steal;
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 91602d310a3f..06a940a74005 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -157,6 +157,8 @@ module_param(force_emulation_prefix, bool, S_IRUGO);
int __read_mostly pi_inject_timer = -1;
module_param(pi_inject_timer, bint, S_IRUGO | S_IWUSR);

+static atomic_t pv_timekeepers_nr;
+
#define KVM_NR_SHARED_MSRS 16

struct kvm_shared_msrs_global {
@@ -2621,6 +2623,16 @@ int kvm_set_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info)

break;
}
+ case MSR_KVM_TIMEKEEPER_EN:
+ if (kvm_gfn_to_hva_cache_init(vcpu->kvm,
+ &vcpu->arch.pv_timekeeper_g2h, data,
+ sizeof(struct pvclock_timekeeper)))
+ vcpu->arch.pv_timekeeper_enabled = false;
+ else {
+ vcpu->arch.pv_timekeeper_enabled = true;
+ atomic_inc(&pv_timekeepers_nr);
+ }
+ break;
case MSR_KVM_ASYNC_PF_EN:
if (kvm_pv_enable_async_pf(vcpu, data))
return 1;
@@ -6965,6 +6977,109 @@ static struct perf_guest_info_callbacks kvm_guest_cbs = {
.handle_intel_pt_intr = kvm_handle_intel_pt_intr,
};

+static DEFINE_SPINLOCK(shadow_pvtk_lock);
+static struct pvclock_timekeeper shadow_pvtk;
+
+static void
+pvclock_copy_read_base(struct pvclock_read_base *pvtkr,
+ struct tk_read_base *tkr)
+{
+ pvtkr->cycle_last = tkr->cycle_last;
+ pvtkr->mult = tkr->mult;
+ pvtkr->shift = tkr->shift;
+ pvtkr->mask = tkr->mask;
+ pvtkr->xtime_nsec = tkr->xtime_nsec;
+ pvtkr->base = tkr->base;
+}
+
+static void
+kvm_copy_into_pvtk(struct kvm_vcpu *vcpu)
+{
+ struct pvclock_timekeeper *pvtk;
+ unsigned long flags;
+
+ if (!vcpu->arch.pv_timekeeper_enabled)
+ return;
+
+ pvtk = &vcpu->arch.pv_timekeeper;
+ if (pvclock_gtod_data.clock.vclock_mode == VCLOCK_TSC) {
+ pvtk->flags |= PVCLOCK_TIMEKEEPER_ENABLED;
+ spin_lock_irqsave(&shadow_pvtk_lock, flags);
+ pvtk->tkr_mono = shadow_pvtk.tkr_mono;
+ pvtk->tkr_raw = shadow_pvtk.tkr_raw;
+
+ pvtk->xtime_sec = shadow_pvtk.xtime_sec;
+ pvtk->ktime_sec = shadow_pvtk.ktime_sec;
+ pvtk->wall_to_monotonic_sec =
+ shadow_pvtk.wall_to_monotonic_sec;
+ pvtk->wall_to_monotonic_nsec =
+ shadow_pvtk.wall_to_monotonic_nsec;
+ pvtk->offs_real = shadow_pvtk.offs_real;
+ pvtk->offs_boot = shadow_pvtk.offs_boot;
+ pvtk->offs_tai = shadow_pvtk.offs_tai;
+ pvtk->raw_sec = shadow_pvtk.raw_sec;
+ spin_unlock_irqrestore(&shadow_pvtk_lock, flags);
+
+ pvtk->tsc_offset = kvm_x86_ops->read_l1_tsc_offset(vcpu);
+ } else
+ pvtk->flags &= ~PVCLOCK_TIMEKEEPER_ENABLED;
+
+ BUILD_BUG_ON(offsetof(struct pvclock_timekeeper, gen) != 0);
+
+ /*
+ * Make the gen count odd to indicate we are in the process of
+ * updating.
+ */
+ vcpu->arch.pv_timekeeper.gen++;
+ vcpu->arch.pv_timekeeper.gen |= 1;
+
+ /*
+ * See comment in kvm_guest_time_update() for why we have to do
+ * multiple writes.
+ */
+ kvm_write_guest_cached(vcpu->kvm, &vcpu->arch.pv_timekeeper_g2h,
+ &vcpu->arch.pv_timekeeper, sizeof(vcpu->arch.pv_timekeeper.gen));
+
+ smp_wmb();
+
+ kvm_write_guest_cached(vcpu->kvm, &vcpu->arch.pv_timekeeper_g2h,
+ &vcpu->arch.pv_timekeeper, sizeof(vcpu->arch.pv_timekeeper));
+
+ smp_wmb();
+
+ vcpu->arch.pv_timekeeper.gen++;
+
+ kvm_write_guest_cached(vcpu->kvm, &vcpu->arch.pv_timekeeper_g2h,
+ &vcpu->arch.pv_timekeeper, sizeof(vcpu->arch.pv_timekeeper.gen));
+}
+
+static void
+update_shadow_pvtk(struct timekeeper *tk)
+{
+ struct pvclock_timekeeper *pvtk;
+ unsigned long flags;
+
+ pvtk = &shadow_pvtk;
+
+ if (atomic_read(&pv_timekeepers_nr) == 0 ||
+ pvclock_gtod_data.clock.vclock_mode != VCLOCK_TSC)
+ return;
+
+ spin_lock_irqsave(&shadow_pvtk_lock, flags);
+ pvclock_copy_read_base(&pvtk->tkr_mono, &tk->tkr_mono);
+ pvclock_copy_read_base(&pvtk->tkr_raw, &tk->tkr_raw);
+
+ pvtk->xtime_sec = tk->xtime_sec;
+ pvtk->ktime_sec = tk->ktime_sec;
+ pvtk->wall_to_monotonic_sec = tk->wall_to_monotonic.tv_sec;
+ pvtk->wall_to_monotonic_nsec = tk->wall_to_monotonic.tv_nsec;
+ pvtk->offs_real = tk->offs_real;
+ pvtk->offs_boot = tk->offs_boot;
+ pvtk->offs_tai = tk->offs_tai;
+ pvtk->raw_sec = tk->raw_sec;
+ spin_unlock_irqrestore(&shadow_pvtk_lock, flags);
+}
+
#ifdef CONFIG_X86_64
static void pvclock_gtod_update_fn(struct work_struct *work)
{
@@ -6993,6 +7108,7 @@ static int pvclock_gtod_notify(struct notifier_block *nb, unsigned long unused,
struct timekeeper *tk = priv;

update_pvclock_gtod(tk);
+ update_shadow_pvtk(tk);

/* disable master clock if host does not trust, or does not
* use, TSC based clocksource.
@@ -7809,6 +7925,8 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)

bool req_immediate_exit = false;

+ kvm_copy_into_pvtk(vcpu);
+
if (kvm_request_pending(vcpu)) {
if (kvm_check_request(KVM_REQ_GET_VMCS12_PAGES, vcpu))
kvm_x86_ops->get_vmcs12_pages(vcpu);
@@ -8891,6 +9009,9 @@ void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu)

kvmclock_reset(vcpu);

+ if (vcpu->arch.pv_timekeeper_enabled)
+ atomic_dec(&pv_timekeepers_nr);
+
kvm_x86_ops->vcpu_free(vcpu);
free_cpumask_var(wbinvd_dirty_mask);
}
--
2.23.0.237.gc6a4ce50a0-goog