Merge tag 'kvm-x86-misc-6.11' of https://github.com/kvm-x86/linux into HEAD

KVM x86 misc changes for 6.11

 - Add a global struct to consolidate tracking of host values, e.g. EFER, and
   move "shadow_phys_bits" into the structure as "maxphyaddr".

 - Add KVM_CAP_X86_APIC_BUS_CYCLES_NS to allow configuring the effective APIC
   bus frequency, because TDX.

 - Print the name of the APICv/AVIC inhibits in the relevant tracepoint.

 - Clean up KVM's handling of vendor specific emulation to consistently act on
   "compatible with Intel/AMD", versus checking for a specific vendor.

 - Misc cleanups

5 files changed, 232 insertions, 0 deletions

diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile
index b073b818be37..b084ba2262a0 100644
--- a/tools/testing/selftests/kvm/Makefile
+++ b/tools/testing/selftests/kvm/Makefile
@@ -112,6 +112,7 @@ TEST_GEN_PROGS_x86_64 += x86_64/vmx_invalid_nested_guest_state
 TEST_GEN_PROGS_x86_64 += x86_64/vmx_set_nested_state_test
 TEST_GEN_PROGS_x86_64 += x86_64/vmx_tsc_adjust_test
 TEST_GEN_PROGS_x86_64 += x86_64/vmx_nested_tsc_scaling_test
+TEST_GEN_PROGS_x86_64 += x86_64/apic_bus_clock_test
 TEST_GEN_PROGS_x86_64 += x86_64/xapic_ipi_test
 TEST_GEN_PROGS_x86_64 += x86_64/xapic_state_test
 TEST_GEN_PROGS_x86_64 += x86_64/xcr0_cpuid_test
diff --git a/tools/testing/selftests/kvm/include/x86_64/apic.h b/tools/testing/selftests/kvm/include/x86_64/apic.h
index bed316fdecd5..0f268b55fa06 100644
--- a/tools/testing/selftests/kvm/include/x86_64/apic.h
+++ b/tools/testing/selftests/kvm/include/x86_64/apic.h
@@ -60,6 +60,14 @@
 #define		APIC_VECTOR_MASK	0x000FF
 #define	APIC_ICR2	0x310
 #define		SET_APIC_DEST_FIELD(x)	((x) << 24)
+#define APIC_LVTT	0x320
+#define		APIC_LVT_TIMER_ONESHOT		(0 << 17)
+#define		APIC_LVT_TIMER_PERIODIC		(1 << 17)
+#define		APIC_LVT_TIMER_TSCDEADLINE	(2 << 17)
+#define		APIC_LVT_MASKED			(1 << 16)
+#define	APIC_TMICT	0x380
+#define	APIC_TMCCT	0x390
+#define	APIC_TDCR	0x3E0
 
 void apic_disable(void);
 void xapic_enable(void);
diff --git a/tools/testing/selftests/kvm/include/x86_64/processor.h b/tools/testing/selftests/kvm/include/x86_64/processor.h
index c0c7c1fe93f9..a0c1440017bb 100644
--- a/tools/testing/selftests/kvm/include/x86_64/processor.h
+++ b/tools/testing/selftests/kvm/include/x86_64/processor.h
@@ -23,6 +23,7 @@
 
 extern bool host_cpu_is_intel;
 extern bool host_cpu_is_amd;
+extern uint64_t guest_tsc_khz;
 
 /* Forced emulation prefix, used to invoke the emulator unconditionally. */
 #define KVM_FEP "ud2; .byte 'k', 'v', 'm';"
@@ -816,6 +817,23 @@ static inline void cpu_relax(void)
 	asm volatile("rep; nop" ::: "memory");
 }
 
+static inline void udelay(unsigned long usec)
+{
+	uint64_t start, now, cycles;
+
+	GUEST_ASSERT(guest_tsc_khz);
+	cycles = guest_tsc_khz / 1000 * usec;
+
+	/*
+	 * Deliberately don't PAUSE, a.k.a. cpu_relax(), so that the delay is
+	 * as accurate as possible, e.g. doesn't trigger PAUSE-Loop VM-Exits.
+	 */
+	start = rdtsc();
+	do {
+		now = rdtsc();
+	} while (now - start < cycles);
+}
+
 #define ud2()			\
 	__asm__ __volatile__(	\
 		"ud2\n"	\
diff --git a/tools/testing/selftests/kvm/lib/x86_64/processor.c b/tools/testing/selftests/kvm/lib/x86_64/processor.c
index 594b061aef52..153739f2e201 100644
--- a/tools/testing/selftests/kvm/lib/x86_64/processor.c
+++ b/tools/testing/selftests/kvm/lib/x86_64/processor.c
@@ -25,6 +25,7 @@ vm_vaddr_t exception_handlers;
 bool host_cpu_is_amd;
 bool host_cpu_is_intel;
 bool is_forced_emulation_enabled;
+uint64_t guest_tsc_khz;
 
 static void regs_dump(FILE *stream, struct kvm_regs *regs, uint8_t indent)
 {
@@ -616,6 +617,11 @@ void assert_on_unhandled_exception(struct kvm_vcpu *vcpu)
 
 void kvm_arch_vm_post_create(struct kvm_vm *vm)
 {
+	int r;
+
+	TEST_ASSERT(kvm_has_cap(KVM_CAP_GET_TSC_KHZ),
+		    "Require KVM_GET_TSC_KHZ to provide udelay() to guest.");
+
 	vm_create_irqchip(vm);
 	vm_init_descriptor_tables(vm);
 
@@ -628,6 +634,11 @@ void kvm_arch_vm_post_create(struct kvm_vm *vm)
 
 		vm_sev_ioctl(vm, KVM_SEV_INIT2, &init);
 	}
+
+	r = __vm_ioctl(vm, KVM_GET_TSC_KHZ, NULL);
+	TEST_ASSERT(r > 0, "KVM_GET_TSC_KHZ did not provide a valid TSC frequency.");
+	guest_tsc_khz = r;
+	sync_global_to_guest(vm, guest_tsc_khz);
 }
 
 void vcpu_arch_set_entry_point(struct kvm_vcpu *vcpu, void *guest_code)
diff --git a/tools/testing/selftests/kvm/x86_64/apic_bus_clock_test.c b/tools/testing/selftests/kvm/x86_64/apic_bus_clock_test.c
new file mode 100644
index 000000000000..f8916bb34405
--- /dev/null
+++ b/tools/testing/selftests/kvm/x86_64/apic_bus_clock_test.c
@@ -0,0 +1,194 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (c) 2024 Intel Corporation
+ *
+ * Verify KVM correctly emulates the APIC bus frequency when the VMM configures
+ * the frequency via KVM_CAP_X86_APIC_BUS_CYCLES_NS.  Start the APIC timer by
+ * programming TMICT (timer initial count) to the largest value possible (so
+ * that the timer will not expire during the test).  Then, after an arbitrary
+ * amount of time has elapsed, verify TMCCT (timer current count) is within 1%
+ * of the expected value based on the time elapsed, the APIC bus frequency, and
+ * the programmed TDCR (timer divide configuration register).
+ */
+
+#include "apic.h"
+#include "test_util.h"
+
+/*
+ * Possible TDCR values with matching divide count. Used to modify APIC
+ * timer frequency.
+ */
+static const struct {
+	const uint32_t tdcr;
+	const uint32_t divide_count;
+} tdcrs[] = {
+	{0x0, 2},
+	{0x1, 4},
+	{0x2, 8},
+	{0x3, 16},
+	{0x8, 32},
+	{0x9, 64},
+	{0xa, 128},
+	{0xb, 1},
+};
+
+static bool is_x2apic;
+
+static void apic_enable(void)
+{
+	if (is_x2apic)
+		x2apic_enable();
+	else
+		xapic_enable();
+}
+
+static uint32_t apic_read_reg(unsigned int reg)
+{
+	return is_x2apic ? x2apic_read_reg(reg) : xapic_read_reg(reg);
+}
+
+static void apic_write_reg(unsigned int reg, uint32_t val)
+{
+	if (is_x2apic)
+		x2apic_write_reg(reg, val);
+	else
+		xapic_write_reg(reg, val);
+}
+
+static void apic_guest_code(uint64_t apic_hz, uint64_t delay_ms)
+{
+	uint64_t tsc_hz = guest_tsc_khz * 1000;
+	const uint32_t tmict = ~0u;
+	uint64_t tsc0, tsc1, freq;
+	uint32_t tmcct;
+	int i;
+
+	apic_enable();
+
+	/*
+	 * Setup one-shot timer.  The vector does not matter because the
+	 * interrupt should not fire.
+	 */
+	apic_write_reg(APIC_LVTT, APIC_LVT_TIMER_ONESHOT | APIC_LVT_MASKED);
+
+	for (i = 0; i < ARRAY_SIZE(tdcrs); i++) {
+		apic_write_reg(APIC_TDCR, tdcrs[i].tdcr);
+		apic_write_reg(APIC_TMICT, tmict);
+
+		tsc0 = rdtsc();
+		udelay(delay_ms * 1000);
+		tmcct = apic_read_reg(APIC_TMCCT);
+		tsc1 = rdtsc();
+
+		/*
+		 * Stop the timer _after_ reading the current, final count, as
+		 * writing the initial counter also modifies the current count.
+		 */
+		apic_write_reg(APIC_TMICT, 0);
+
+		freq = (tmict - tmcct) * tdcrs[i].divide_count * tsc_hz / (tsc1 - tsc0);
+		/* Check if measured frequency is within 5% of configured frequency. */
+		__GUEST_ASSERT(freq < apic_hz * 105 / 100 && freq > apic_hz * 95 / 100,
+			       "Frequency = %lu (wanted %lu - %lu), bus = %lu, div = %u, tsc = %lu",
+			       freq, apic_hz * 95 / 100, apic_hz * 105 / 100,
+			       apic_hz, tdcrs[i].divide_count, tsc_hz);
+	}
+
+	GUEST_DONE();
+}
+
+static void test_apic_bus_clock(struct kvm_vcpu *vcpu)
+{
+	bool done = false;
+	struct ucall uc;
+
+	while (!done) {
+		vcpu_run(vcpu);
+
+		TEST_ASSERT_KVM_EXIT_REASON(vcpu, KVM_EXIT_IO);
+
+		switch (get_ucall(vcpu, &uc)) {
+		case UCALL_DONE:
+			done = true;
+			break;
+		case UCALL_ABORT:
+			REPORT_GUEST_ASSERT(uc);
+			break;
+		default:
+			TEST_FAIL("Unknown ucall %lu", uc.cmd);
+			break;
+		}
+	}
+}
+
+static void run_apic_bus_clock_test(uint64_t apic_hz, uint64_t delay_ms,
+				    bool x2apic)
+{
+	struct kvm_vcpu *vcpu;
+	struct kvm_vm *vm;
+	int ret;
+
+	is_x2apic = x2apic;
+
+	vm = vm_create(1);
+
+	sync_global_to_guest(vm, is_x2apic);
+
+	vm_enable_cap(vm, KVM_CAP_X86_APIC_BUS_CYCLES_NS,
+		      NSEC_PER_SEC / apic_hz);
+
+	vcpu = vm_vcpu_add(vm, 0, apic_guest_code);
+	vcpu_args_set(vcpu, 2, apic_hz, delay_ms);
+
+	ret = __vm_enable_cap(vm, KVM_CAP_X86_APIC_BUS_CYCLES_NS,
+			      NSEC_PER_SEC / apic_hz);
+	TEST_ASSERT(ret < 0 && errno == EINVAL,
+		    "Setting of APIC bus frequency after vCPU is created should fail.");
+
+	if (!is_x2apic)
+		virt_pg_map(vm, APIC_DEFAULT_GPA, APIC_DEFAULT_GPA);
+
+	test_apic_bus_clock(vcpu);
+	kvm_vm_free(vm);
+}
+
+static void help(char *name)
+{
+	puts("");
+	printf("usage: %s [-h] [-d delay] [-f APIC bus freq]\n", name);
+	puts("");
+	printf("-d: Delay (in msec) guest uses to measure APIC bus frequency.\n");
+	printf("-f: The APIC bus frequency (in MHz) to be configured for the guest.\n");
+	puts("");
+}
+
+int main(int argc, char *argv[])
+{
+	/*
+	 * Arbitrarilty default to 25MHz for the APIC bus frequency, which is
+	 * different enough from the default 1GHz to be interesting.
+	 */
+	uint64_t apic_hz = 25 * 1000 * 1000;
+	uint64_t delay_ms = 100;
+	int opt;
+
+	TEST_REQUIRE(kvm_has_cap(KVM_CAP_X86_APIC_BUS_CYCLES_NS));
+
+	while ((opt = getopt(argc, argv, "d:f:h")) != -1) {
+		switch (opt) {
+		case 'f':
+			apic_hz = atoi_positive("APIC bus frequency", optarg) * 1000 * 1000;
+			break;
+		case 'd':
+			delay_ms = atoi_positive("Delay in milliseconds", optarg);
+			break;
+		case 'h':
+		default:
+			help(argv[0]);
+			exit(KSFT_SKIP);
+		}
+	}
+
+	run_apic_bus_clock_test(apic_hz, delay_ms, false);
+	run_apic_bus_clock_test(apic_hz, delay_ms, true);
+}