[kvm-unit-tests PATCH v2] x86: Add RDTSC test

[kvm-unit-tests PATCH v2] x86: Add RDTSC test

[Aaron Lewis]

Verify that the difference between a guest RDTSC instruction and the
IA32_TIME_STAMP_COUNTER MSR value stored in the VMCS12's VM-exit
MSR-store list is less than 750 cycles, 99.9% of the time.

Signed-off-by: Aaron Lewis <aaronlewis@google.com>
Reviewed-by: Jim Mattson <jmattson@google.com>
---
 x86/unittests.cfg |  6 ++++
 x86/vmx.h         |  1 +
 x86/vmx_tests.c   | 91 +++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 98 insertions(+)

diff --git a/x86/unittests.cfg b/x86/unittests.cfg
index b4865ac..5291d96 100644
--- a/x86/unittests.cfg
+++ b/x86/unittests.cfg
@@ -284,6 +284,12 @@ extra_params = -cpu host,+vmx -append vmx_vmcs_shadow_test
 arch = x86_64
 groups = vmx
 
+[vmx_rdtsc_vmexit_diff_test]
+file = vmx.flat
+extra_params = -cpu host,+vmx -append rdtsc_vmexit_diff_test
+arch = x86_64
+groups = vmx
+
 [debug]
 file = debug.flat
 arch = x86_64
diff --git a/x86/vmx.h b/x86/vmx.h
index 8496be7..21ba953 100644
--- a/x86/vmx.h
+++ b/x86/vmx.h
@@ -420,6 +420,7 @@ enum Ctrl1 {
 	CPU_SHADOW_VMCS		= 1ul << 14,
 	CPU_RDSEED		= 1ul << 16,
 	CPU_PML                 = 1ul << 17,
+	CPU_USE_TSC_SCALING	= 1ul << 25,
 };
 
 enum Intr_type {
diff --git a/x86/vmx_tests.c b/x86/vmx_tests.c
index 1d8932f..fcf71e7 100644
--- a/x86/vmx_tests.c
+++ b/x86/vmx_tests.c
@@ -8790,7 +8790,97 @@ static void vmx_vmcs_shadow_test(void)
 	enter_guest();
 }
 
+/*
+ * This test monitors the difference between a guest RDTSC instruction
+ * and the IA32_TIME_STAMP_COUNTER MSR value stored in the VMCS12
+ * VM-exit MSR-store list when taking a VM-exit on the instruction
+ * following RDTSC.
+ */
+#define RDTSC_DIFF_ITERS 100000
+#define RDTSC_DIFF_FAILS 100
+#define HOST_RDTSC_LIMIT 750
+
+/*
+ * Set 'use TSC offsetting' and set the guest offset to the
+ * inverse of the host's current TSC value, so that the guest starts running
+ * with an effective TSC value of 0.
+ */
+static void reset_guest_tsc_to_zero(void)
+{
+	TEST_ASSERT_MSG(ctrl_cpu_rev[0].clr & CPU_USE_TSC_OFFSET,
+			"Expected support for 'use TSC offsetting'");
+
+	vmcs_set_bits(CPU_EXEC_CTRL0, CPU_USE_TSC_OFFSET);
+	vmcs_write(TSC_OFFSET, -rdtsc());
+}
+
+static void rdtsc_vmexit_diff_test_guest(void)
+{
+	int i;
+
+	for (i = 0; i < RDTSC_DIFF_ITERS; i++)
+		/* Ensure rdtsc is the last instruction before the vmcall. */
+		asm volatile("rdtsc; vmcall" : : : "eax", "edx");
+}
 
+/*
+ * This function only considers the "use TSC offsetting" VM-execution
+ * control.  It does not handle "use TSC scaling" (because the latter
+ * isn't available to the host today.)
+ */
+static unsigned long long host_time_to_guest_time(unsigned long long t)
+{
+	TEST_ASSERT((vmcs_read(CPU_EXEC_CTRL1) & CPU_USE_TSC_SCALING) == 0);
+
+	if (vmcs_read(CPU_EXEC_CTRL0) & CPU_USE_TSC_OFFSET)
+		t += vmcs_read(TSC_OFFSET);
+
+	return t;
+}
+
+static unsigned long long rdtsc_vmexit_diff_test_iteration(void)
+{
+	unsigned long long guest_tsc, host_to_guest_tsc;
+
+	enter_guest();
+	skip_exit_vmcall();
+	guest_tsc = (u32) regs.rax + (regs.rdx << 32);
+	host_to_guest_tsc = host_time_to_guest_time(exit_msr_store[0].value);
+
+	return host_to_guest_tsc - guest_tsc;
+}
+
+static void rdtsc_vmexit_diff_test(void)
+{
+	int fail = 0;
+	int i;
+
+	test_set_guest(rdtsc_vmexit_diff_test_guest);
+
+	reset_guest_tsc_to_zero();
+
+	/*
+	 * Set up the VMCS12 VM-exit MSR-store list to store just one
+	 * MSR: IA32_TIME_STAMP_COUNTER. Note that the value stored is
+	 * in the host time domain (i.e., it is not adjusted according
+	 * to the TSC multiplier and TSC offset fields in the VMCS12,
+	 * as a guest RDTSC would be.)
+	 */
+	exit_msr_store = alloc_page();
+	exit_msr_store[0].index = MSR_IA32_TSC;
+	vmcs_write(EXI_MSR_ST_CNT, 1);
+	vmcs_write(EXIT_MSR_ST_ADDR, virt_to_phys(exit_msr_store));
+
+	for (i = 0; i < RDTSC_DIFF_ITERS; i++) {
+		if (rdtsc_vmexit_diff_test_iteration() >= HOST_RDTSC_LIMIT)
+			fail++;
+	}
+
+	enter_guest();
+
+	report("RDTSC to VM-exit delta too high in %d of %d iterations",
+	       fail < RDTSC_DIFF_FAILS, fail, RDTSC_DIFF_ITERS);
+}
 
 static int invalid_msr_init(struct vmcs *vmcs)
 {
@@ -9056,5 +9146,6 @@ struct vmx_test vmx_tests[] = {
 	/* Atomic MSR switch tests. */
 	TEST(atomic_switch_max_msrs_test),
 	TEST(atomic_switch_overflow_msrs_test),
+	TEST(rdtsc_vmexit_diff_test),
 	{ NULL, NULL, NULL, NULL, NULL, {0} },
 };
-- 
2.24.0.432.g9d3f5f5b63-goog

Re: [kvm-unit-tests PATCH v2] x86: Add RDTSC test

[Liran Alon]

> On 26 Nov 2019, at 23:44, Aaron Lewis <aaronlewis@google.com> wrote:
> 
> Verify that the difference between a guest RDTSC instruction and the
> IA32_TIME_STAMP_COUNTER MSR value stored in the VMCS12's VM-exit
> MSR-store list is less than 750 cycles, 99.9% of the time.

It will help if commit message would also reference the KVM commit which fixes the issue tested by this test.
i.e. 662f1d1d1931 ("KVM: nVMX: Add support for capturing highest observable L2 TSC”)

> 
> Signed-off-by: Aaron Lewis <aaronlewis@google.com>
> Reviewed-by: Jim Mattson <jmattson@google.com>
> ---
> x86/unittests.cfg |  6 ++++
> x86/vmx.h         |  1 +
> x86/vmx_tests.c   | 91 +++++++++++++++++++++++++++++++++++++++++++++++
> 3 files changed, 98 insertions(+)
> 
> diff --git a/x86/unittests.cfg b/x86/unittests.cfg
> index b4865ac..5291d96 100644
> --- a/x86/unittests.cfg
> +++ b/x86/unittests.cfg
> @@ -284,6 +284,12 @@ extra_params = -cpu host,+vmx -append vmx_vmcs_shadow_test
> arch = x86_64
> groups = vmx
> 
> +[vmx_rdtsc_vmexit_diff_test]
> +file = vmx.flat
> +extra_params = -cpu host,+vmx -append rdtsc_vmexit_diff_test
> +arch = x86_64
> +groups = vmx
> +

I think we are missing some clear guidance on when a VMX unit-test should have it’s own test-section in x86/unittests.cfg.

I believe the guidance should be that all VMX tests are suppose to be run by [vmx] except those that have special requirements
on execution environment (e.g. vmx_smp*) or destroy execution environment after they run (e.g. vmx_init_signal_test) or
require special timeout if they fail (e.g. vmx_apic_passthrough_tpr_threshold_test).
These tests should both be removed from [vmx] by "-append -test_name” and have their own section which runs them.

Being concrete to this patch, I think it shouldn’t have it’s own section.
For example, it will cause the test to run twice: Both as part of [vmx] and as part of [vmx_rdtsc_vmexit_diff_test].

And I can submit a separate patches to:
1) Rename vmx_eoi_bitmap_ioapic_scan & vmx_apic_passthrough_thread to prefix with vmx_smp*
    (It actually seems to me that currently there are no vmx_smp* tests at all…)
    and create a [vmx_smp] section for running them.
2) Remove vmx_hlt_with_rvi_test and vmx_apicv_test sections.

Does anyone think differently?

> [debug]
> file = debug.flat
> arch = x86_64
> diff --git a/x86/vmx.h b/x86/vmx.h
> index 8496be7..21ba953 100644
> --- a/x86/vmx.h
> +++ b/x86/vmx.h
> @@ -420,6 +420,7 @@ enum Ctrl1 {
> 	CPU_SHADOW_VMCS		= 1ul << 14,
> 	CPU_RDSEED		= 1ul << 16,
> 	CPU_PML                 = 1ul << 17,
> +	CPU_USE_TSC_SCALING	= 1ul << 25,
> };
> 
> enum Intr_type {
> diff --git a/x86/vmx_tests.c b/x86/vmx_tests.c
> index 1d8932f..fcf71e7 100644
> --- a/x86/vmx_tests.c
> +++ b/x86/vmx_tests.c
> @@ -8790,7 +8790,97 @@ static void vmx_vmcs_shadow_test(void)
> 	enter_guest();
> }
> 
> +/*
> + * This test monitors the difference between a guest RDTSC instruction
> + * and the IA32_TIME_STAMP_COUNTER MSR value stored in the VMCS12
> + * VM-exit MSR-store list when taking a VM-exit on the instruction
> + * following RDTSC.
> + */
> +#define RDTSC_DIFF_ITERS 100000
> +#define RDTSC_DIFF_FAILS 100
> +#define HOST_RDTSC_LIMIT 750

Nit: I suggest to rename HOST_RDTSC_LIMIT to HOST_CAPTURED_GUEST_TSC_DIFF_THRESHOLD.

> +
> +/*
> + * Set 'use TSC offsetting' and set the guest offset to the
> + * inverse of the host's current TSC value, so that the guest starts running
> + * with an effective TSC value of 0.
> + */
> +static void reset_guest_tsc_to_zero(void)
> +{
> +	TEST_ASSERT_MSG(ctrl_cpu_rev[0].clr & CPU_USE_TSC_OFFSET,
> +			"Expected support for 'use TSC offsetting'");
> +
> +	vmcs_set_bits(CPU_EXEC_CTRL0, CPU_USE_TSC_OFFSET);
> +	vmcs_write(TSC_OFFSET, -rdtsc());
> +}
> +
> +static void rdtsc_vmexit_diff_test_guest(void)
> +{
> +	int i;
> +
> +	for (i = 0; i < RDTSC_DIFF_ITERS; i++)
> +		/* Ensure rdtsc is the last instruction before the vmcall. */
> +		asm volatile("rdtsc; vmcall" : : : "eax", "edx");
> +}
> 
> +/*
> + * This function only considers the "use TSC offsetting" VM-execution
> + * control.  It does not handle "use TSC scaling" (because the latter
> + * isn't available to the host today.)
> + */
> +static unsigned long long host_time_to_guest_time(unsigned long long t)
> +{
> +	TEST_ASSERT((vmcs_read(CPU_EXEC_CTRL1) & CPU_USE_TSC_SCALING) == 0);

It’s problematic to vmcs_read(CPU_EXEC_CTRL1) when test runs on CPU that doesn’t support
secondary VM-execution control. As this will cause VMfailInvalid (i.e. Clear CF,PF,AF,SF,OF and set ZF).

What’s worse is that vmcs_read() today doesn’t assert that RFLAGS.ZF==0 after executing VMREAD.
Maybe we should submit a separate patch for that as-well…

Anyway, you can just change your assert condition to:
TEST_ASSERT(!(ctrl_cpu_rev[0].clr & CPU_SECONDARY) || !(vmcs_read(CPU_EXEC_CTRL1) & CPU_USE_TSC_SCALING));

> +
> +	if (vmcs_read(CPU_EXEC_CTRL0) & CPU_USE_TSC_OFFSET)
> +		t += vmcs_read(TSC_OFFSET);
> +
> +	return t;
> +}
> +
> +static unsigned long long rdtsc_vmexit_diff_test_iteration(void)
> +{
> +	unsigned long long guest_tsc, host_to_guest_tsc;
> +
> +	enter_guest();
> +	skip_exit_vmcall();
> +	guest_tsc = (u32) regs.rax + (regs.rdx << 32);
> +	host_to_guest_tsc = host_time_to_guest_time(exit_msr_store[0].value);
> +
> +	return host_to_guest_tsc - guest_tsc;
> +}
> +
> +static void rdtsc_vmexit_diff_test(void)
> +{
> +	int fail = 0;
> +	int i;
> +
> +	test_set_guest(rdtsc_vmexit_diff_test_guest);
> +
> +	reset_guest_tsc_to_zero();
> +
> +	/*
> +	 * Set up the VMCS12 VM-exit MSR-store list to store just one
> +	 * MSR: IA32_TIME_STAMP_COUNTER. Note that the value stored is
> +	 * in the host time domain (i.e., it is not adjusted according
> +	 * to the TSC multiplier and TSC offset fields in the VMCS12,
> +	 * as a guest RDTSC would be.)
> +	 */
> +	exit_msr_store = alloc_page();
> +	exit_msr_store[0].index = MSR_IA32_TSC;
> +	vmcs_write(EXI_MSR_ST_CNT, 1);
> +	vmcs_write(EXIT_MSR_ST_ADDR, virt_to_phys(exit_msr_store));
> +
> +	for (i = 0; i < RDTSC_DIFF_ITERS; i++) {
> +		if (rdtsc_vmexit_diff_test_iteration() >= HOST_RDTSC_LIMIT)
> +			fail++;
> +	}
> +
> +	enter_guest();
> +
> +	report("RDTSC to VM-exit delta too high in %d of %d iterations",
> +	       fail < RDTSC_DIFF_FAILS, fail, RDTSC_DIFF_ITERS);
> +}
> 
> static int invalid_msr_init(struct vmcs *vmcs)
> {
> @@ -9056,5 +9146,6 @@ struct vmx_test vmx_tests[] = {
> 	/* Atomic MSR switch tests. */
> 	TEST(atomic_switch_max_msrs_test),
> 	TEST(atomic_switch_overflow_msrs_test),
> +	TEST(rdtsc_vmexit_diff_test),
> 	{ NULL, NULL, NULL, NULL, NULL, {0} },
> };
> -- 
> 2.24.0.432.g9d3f5f5b63-goog
>