Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm

Pull kvm fixes from Paolo Bonzini:
 "A bit late... I got sidetracked by back-from-vacation routines and
  conferences. But most of these patches are already a few weeks old and
  things look more calm on the mailing list than what this pull request
  would suggest.

  x86:

   - missing TLB flush

   - nested virtualization fixes for SMM (secure boot on nested
     hypervisor) and other nested SVM fixes

   - syscall fuzzing fixes

   - live migration fix for AMD SEV

   - mirror VMs now work for SEV-ES too

   - fixes for reset

   - possible out-of-bounds access in IOAPIC emulation

   - fix enlightened VMCS on Windows 2022

  ARM:

   - Add missing FORCE target when building the EL2 object

   - Fix a PMU probe regression on some platforms

  Generic:

   - KCSAN fixes

  selftests:

   - random fixes, mostly for clang compilation"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (43 commits)
  selftests: KVM: Explicitly use movq to read xmm registers
  selftests: KVM: Call ucall_init when setting up in rseq_test
  KVM: Remove tlbs_dirty
  KVM: X86: Synchronize the shadow pagetable before link it
  KVM: X86: Fix missed remote tlb flush in rmap_write_protect()
  KVM: x86: nSVM: don't copy virt_ext from vmcb12
  KVM: x86: nSVM: test eax for 4K alignment for GP errata workaround
  KVM: x86: selftests: test simultaneous uses of V_IRQ from L1 and L0
  KVM: x86: nSVM: restore int_vector in svm_clear_vintr
  kvm: x86: Add AMD PMU MSRs to msrs_to_save_all[]
  KVM: x86: nVMX: re-evaluate emulation_required on nested VM exit
  KVM: x86: nVMX: don't fail nested VM entry on invalid guest state if !from_vmentry
  KVM: x86: VMX: synthesize invalid VM exit when emulating invalid guest state
  KVM: x86: nSVM: refactor svm_leave_smm and smm_enter_smm
  KVM: x86: SVM: call KVM_REQ_GET_NESTED_STATE_PAGES on exit from SMM mode
  KVM: x86: reset pdptrs_from_userspace when exiting smm
  KVM: x86: nSVM: restore the L1 host state prior to resuming nested guest on SMM exit
  KVM: nVMX: Filter out all unsupported controls when eVMCS was activated
  KVM: KVM: Use cpumask_available() to check for NULL cpumask when kicking vCPUs
  KVM: Clean up benign vcpu->cpu data races when kicking vCPUs
  ...

tools/testing/selftests/kvm/.gitignore

@@ -24,6 +24,7 @@
 /x86_64/smm_test
 /x86_64/state_test
 /x86_64/svm_vmcall_test
+/x86_64/svm_int_ctl_test
 /x86_64/sync_regs_test
 /x86_64/tsc_msrs_test
 /x86_64/userspace_msr_exit_test

tools/testing/selftests/kvm/Makefile

@@ -56,6 +56,7 @@ TEST_GEN_PROGS_x86_64 += x86_64/smm_test
 TEST_GEN_PROGS_x86_64 += x86_64/state_test
 TEST_GEN_PROGS_x86_64 += x86_64/vmx_preemption_timer_test
 TEST_GEN_PROGS_x86_64 += x86_64/svm_vmcall_test
+TEST_GEN_PROGS_x86_64 += x86_64/svm_int_ctl_test
 TEST_GEN_PROGS_x86_64 += x86_64/sync_regs_test
 TEST_GEN_PROGS_x86_64 += x86_64/userspace_msr_exit_test
 TEST_GEN_PROGS_x86_64 += x86_64/vmx_apic_access_test

tools/testing/selftests/kvm/access_tracking_perf_test.c

@@ -371,9 +371,7 @@ static void help(char *name)
 	printf(" -v: specify the number of vCPUs to run.\n");
 	printf(" -o: Overlap guest memory accesses instead of partitioning\n"
 	       "     them into a separate region of memory for each vCPU.\n");
-	printf(" -s: specify the type of memory that should be used to\n"
-	       "     back the guest data region.\n\n");
-	backing_src_help();
+	backing_src_help("-s");
 	puts("");
 	exit(0);
 }
@@ -381,7 +379,7 @@ static void help(char *name)
 int main(int argc, char *argv[])
 {
 	struct test_params params = {
-		.backing_src = VM_MEM_SRC_ANONYMOUS,
+		.backing_src = DEFAULT_VM_MEM_SRC,
 		.vcpu_memory_bytes = DEFAULT_PER_VCPU_MEM_SIZE,
 		.vcpus = 1,
 	};

tools/testing/selftests/kvm/demand_paging_test.c

@@ -179,7 +179,7 @@ static void *uffd_handler_thread_fn(void *arg)
 			return NULL;
 		}
 
-		if (!pollfd[0].revents & POLLIN)
+		if (!(pollfd[0].revents & POLLIN))
 			continue;
 
 		r = read(uffd, &msg, sizeof(msg));
@@ -416,7 +416,7 @@ static void help(char *name)
 {
 	puts("");
 	printf("usage: %s [-h] [-m vm_mode] [-u uffd_mode] [-d uffd_delay_usec]\n"
-	       "          [-b memory] [-t type] [-v vcpus] [-o]\n", name);
+	       "          [-b memory] [-s type] [-v vcpus] [-o]\n", name);
 	guest_modes_help();
 	printf(" -u: use userfaultfd to handle vCPU page faults. Mode is a\n"
 	       "     UFFD registration mode: 'MISSING' or 'MINOR'.\n");
@@ -426,8 +426,7 @@ static void help(char *name)
 	printf(" -b: specify the size of the memory region which should be\n"
 	       "     demand paged by each vCPU. e.g. 10M or 3G.\n"
 	       "     Default: 1G\n");
-	printf(" -t: The type of backing memory to use. Default: anonymous\n");
-	backing_src_help();
+	backing_src_help("-s");
 	printf(" -v: specify the number of vCPUs to run.\n");
 	printf(" -o: Overlap guest memory accesses instead of partitioning\n"
 	       "     them into a separate region of memory for each vCPU.\n");
@@ -439,14 +438,14 @@ int main(int argc, char *argv[])
 {
 	int max_vcpus = kvm_check_cap(KVM_CAP_MAX_VCPUS);
 	struct test_params p = {
-		.src_type = VM_MEM_SRC_ANONYMOUS,
+		.src_type = DEFAULT_VM_MEM_SRC,
 		.partition_vcpu_memory_access = true,
 	};
 	int opt;
 
 	guest_modes_append_default();
 
-	while ((opt = getopt(argc, argv, "hm:u:d:b:t:v:o")) != -1) {
+	while ((opt = getopt(argc, argv, "hm:u:d:b:s:v:o")) != -1) {
 		switch (opt) {
 		case 'm':
 			guest_modes_cmdline(optarg);
@@ -465,7 +464,7 @@ int main(int argc, char *argv[])
 		case 'b':
 			guest_percpu_mem_size = parse_size(optarg);
 			break;
-		case 't':
+		case 's':
 			p.src_type = parse_backing_src_type(optarg);
 			break;
 		case 'v':
@@ -485,7 +484,7 @@ int main(int argc, char *argv[])
 
 	if (p.uffd_mode == UFFDIO_REGISTER_MODE_MINOR &&
 	    !backing_src_is_shared(p.src_type)) {
-		TEST_FAIL("userfaultfd MINOR mode requires shared memory; pick a different -t");
+		TEST_FAIL("userfaultfd MINOR mode requires shared memory; pick a different -s");
 	}
 
 	for_each_guest_mode(run_test, &p);

tools/testing/selftests/kvm/dirty_log_perf_test.c

@@ -118,42 +118,64 @@ static inline void disable_dirty_logging(struct kvm_vm *vm, int slots)
 	toggle_dirty_logging(vm, slots, false);
 }
 
-static void get_dirty_log(struct kvm_vm *vm, int slots, unsigned long *bitmap,
-			  uint64_t nr_pages)
+static void get_dirty_log(struct kvm_vm *vm, unsigned long *bitmaps[], int slots)
 {
-	uint64_t slot_pages = nr_pages / slots;
 	int i;
 
 	for (i = 0; i < slots; i++) {
 		int slot = PERF_TEST_MEM_SLOT_INDEX + i;
-		unsigned long *slot_bitmap = bitmap + i * slot_pages;
 
-		kvm_vm_get_dirty_log(vm, slot, slot_bitmap);
+		kvm_vm_get_dirty_log(vm, slot, bitmaps[i]);
 	}
 }
 
-static void clear_dirty_log(struct kvm_vm *vm, int slots, unsigned long *bitmap,
-			    uint64_t nr_pages)
+static void clear_dirty_log(struct kvm_vm *vm, unsigned long *bitmaps[],
+			    int slots, uint64_t pages_per_slot)
 {
-	uint64_t slot_pages = nr_pages / slots;
 	int i;
 
 	for (i = 0; i < slots; i++) {
 		int slot = PERF_TEST_MEM_SLOT_INDEX + i;
-		unsigned long *slot_bitmap = bitmap + i * slot_pages;
 
-		kvm_vm_clear_dirty_log(vm, slot, slot_bitmap, 0, slot_pages);
+		kvm_vm_clear_dirty_log(vm, slot, bitmaps[i], 0, pages_per_slot);
 	}
 }
 
+static unsigned long **alloc_bitmaps(int slots, uint64_t pages_per_slot)
+{
+	unsigned long **bitmaps;
+	int i;
+
+	bitmaps = malloc(slots * sizeof(bitmaps[0]));
+	TEST_ASSERT(bitmaps, "Failed to allocate bitmaps array.");
+
+	for (i = 0; i < slots; i++) {
+		bitmaps[i] = bitmap_zalloc(pages_per_slot);
+		TEST_ASSERT(bitmaps[i], "Failed to allocate slot bitmap.");
+	}
+
+	return bitmaps;
+}
+
+static void free_bitmaps(unsigned long *bitmaps[], int slots)
+{
+	int i;
+
+	for (i = 0; i < slots; i++)
+		free(bitmaps[i]);
+
+	free(bitmaps);
+}
+
 static void run_test(enum vm_guest_mode mode, void *arg)
 {
 	struct test_params *p = arg;
 	pthread_t *vcpu_threads;
 	struct kvm_vm *vm;
-	unsigned long *bmap;
+	unsigned long **bitmaps;
 	uint64_t guest_num_pages;
 	uint64_t host_num_pages;
+	uint64_t pages_per_slot;
 	int vcpu_id;
 	struct timespec start;
 	struct timespec ts_diff;
@@ -171,7 +193,9 @@ static void run_test(enum vm_guest_mode mode, void *arg)
 	guest_num_pages = (nr_vcpus * guest_percpu_mem_size) >> vm_get_page_shift(vm);
 	guest_num_pages = vm_adjust_num_guest_pages(mode, guest_num_pages);
 	host_num_pages = vm_num_host_pages(mode, guest_num_pages);
-	bmap = bitmap_zalloc(host_num_pages);
+	pages_per_slot = host_num_pages / p->slots;
+
+	bitmaps = alloc_bitmaps(p->slots, pages_per_slot);
 
 	if (dirty_log_manual_caps) {
 		cap.cap = KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2;
@@ -239,7 +263,7 @@ static void run_test(enum vm_guest_mode mode, void *arg)
 			iteration, ts_diff.tv_sec, ts_diff.tv_nsec);
 
 		clock_gettime(CLOCK_MONOTONIC, &start);
-		get_dirty_log(vm, p->slots, bmap, host_num_pages);
+		get_dirty_log(vm, bitmaps, p->slots);
 		ts_diff = timespec_elapsed(start);
 		get_dirty_log_total = timespec_add(get_dirty_log_total,
 						   ts_diff);
@@ -248,7 +272,7 @@ static void run_test(enum vm_guest_mode mode, void *arg)
 
 		if (dirty_log_manual_caps) {
 			clock_gettime(CLOCK_MONOTONIC, &start);
-			clear_dirty_log(vm, p->slots, bmap, host_num_pages);
+			clear_dirty_log(vm, bitmaps, p->slots, pages_per_slot);
 			ts_diff = timespec_elapsed(start);
 			clear_dirty_log_total = timespec_add(clear_dirty_log_total,
 							     ts_diff);
@@ -281,7 +305,7 @@ static void run_test(enum vm_guest_mode mode, void *arg)
 			clear_dirty_log_total.tv_nsec, avg.tv_sec, avg.tv_nsec);
 	}
 
-	free(bmap);
+	free_bitmaps(bitmaps, p->slots);
 	free(vcpu_threads);
 	perf_test_destroy_vm(vm);
 }
@@ -308,11 +332,9 @@ static void help(char *name)
 	printf(" -v: specify the number of vCPUs to run.\n");
 	printf(" -o: Overlap guest memory accesses instead of partitioning\n"
 	       "     them into a separate region of memory for each vCPU.\n");
-	printf(" -s: specify the type of memory that should be used to\n"
-	       "     back the guest data region.\n\n");
+	backing_src_help("-s");
 	printf(" -x: Split the memory region into this number of memslots.\n"
-	       "     (default: 1)");
-	backing_src_help();
+	       "     (default: 1)\n");
 	puts("");
 	exit(0);
 }
@@ -324,7 +346,7 @@ int main(int argc, char *argv[])
 		.iterations = TEST_HOST_LOOP_N,
 		.wr_fract = 1,
 		.partition_vcpu_memory_access = true,
-		.backing_src = VM_MEM_SRC_ANONYMOUS,
+		.backing_src = DEFAULT_VM_MEM_SRC,
 		.slots = 1,
 	};
 	int opt;

tools/testing/selftests/kvm/include/test_util.h

@@ -90,6 +90,8 @@ enum vm_mem_backing_src_type {
 	NUM_SRC_TYPES,
 };
 
+#define DEFAULT_VM_MEM_SRC VM_MEM_SRC_ANONYMOUS
+
 struct vm_mem_backing_src_alias {
 	const char *name;
 	uint32_t flag;
@@ -102,7 +104,7 @@ size_t get_trans_hugepagesz(void);
 size_t get_def_hugetlb_pagesz(void);
 const struct vm_mem_backing_src_alias *vm_mem_backing_src_alias(uint32_t i);
 size_t get_backing_src_pagesz(uint32_t i);
-void backing_src_help(void);
+void backing_src_help(const char *flag);
 enum vm_mem_backing_src_type parse_backing_src_type(const char *type_name);
 long get_run_delay(void);
 

tools/testing/selftests/kvm/include/x86_64/processor.h

@@ -312,37 +312,37 @@ static inline void set_xmm(int n, unsigned long val)
 	}
 }
 
-typedef unsigned long v1di __attribute__ ((vector_size (8)));
+#define GET_XMM(__xmm)							\
+({									\
+	unsigned long __val;						\
+	asm volatile("movq %%"#__xmm", %0" : "=r"(__val) : : #__xmm);	\
+	__val;								\
+})
+
 static inline unsigned long get_xmm(int n)
 {
 	assert(n >= 0 && n <= 7);
 
-	register v1di xmm0 __asm__("%xmm0");
-	register v1di xmm1 __asm__("%xmm1");
-	register v1di xmm2 __asm__("%xmm2");
-	register v1di xmm3 __asm__("%xmm3");
-	register v1di xmm4 __asm__("%xmm4");
-	register v1di xmm5 __asm__("%xmm5");
-	register v1di xmm6 __asm__("%xmm6");
-	register v1di xmm7 __asm__("%xmm7");
 	switch (n) {
 	case 0:
-		return (unsigned long)xmm0;
+		return GET_XMM(xmm0);
 	case 1:
-		return (unsigned long)xmm1;
+		return GET_XMM(xmm1);
 	case 2:
-		return (unsigned long)xmm2;
+		return GET_XMM(xmm2);
 	case 3:
-		return (unsigned long)xmm3;
+		return GET_XMM(xmm3);
 	case 4:
-		return (unsigned long)xmm4;
+		return GET_XMM(xmm4);
 	case 5:
-		return (unsigned long)xmm5;
+		return GET_XMM(xmm5);
 	case 6:
-		return (unsigned long)xmm6;
+		return GET_XMM(xmm6);
 	case 7:
-		return (unsigned long)xmm7;
+		return GET_XMM(xmm7);
 	}
+
+	/* never reached */
 	return 0;
 }
 

tools/testing/selftests/kvm/kvm_page_table_test.c

@@ -456,10 +456,7 @@ static void help(char *name)
 	       "     (default: 1G)\n");
 	printf(" -v: specify the number of vCPUs to run\n"
 	       "     (default: 1)\n");
-	printf(" -s: specify the type of memory that should be used to\n"
-	       "     back the guest data region.\n"
-	       "     (default: anonymous)\n\n");
-	backing_src_help();
+	backing_src_help("-s");
 	puts("");
 }
 
@@ -468,7 +465,7 @@ int main(int argc, char *argv[])
 	int max_vcpus = kvm_check_cap(KVM_CAP_MAX_VCPUS);
 	struct test_params p = {
 		.test_mem_size = DEFAULT_TEST_MEM_SIZE,
-		.src_type = VM_MEM_SRC_ANONYMOUS,
+		.src_type = DEFAULT_VM_MEM_SRC,
 	};
 	int opt;
 

tools/testing/selftests/kvm/lib/test_util.c

@@ -283,13 +283,22 @@ size_t get_backing_src_pagesz(uint32_t i)
 	}
 }
 
-void backing_src_help(void)
+static void print_available_backing_src_types(const char *prefix)
 {
 	int i;
 
-	printf("Available backing src types:\n");
+	printf("%sAvailable backing src types:\n", prefix);
+
 	for (i = 0; i < NUM_SRC_TYPES; i++)
-		printf("\t%s\n", vm_mem_backing_src_alias(i)->name);
+		printf("%s    %s\n", prefix, vm_mem_backing_src_alias(i)->name);
+}
+
+void backing_src_help(const char *flag)
+{
+	printf(" %s: specify the type of memory that should be used to\n"
+	       "     back the guest data region. (default: %s)\n",
+	       flag, vm_mem_backing_src_alias(DEFAULT_VM_MEM_SRC)->name);
+	print_available_backing_src_types("     ");
 }
 
 enum vm_mem_backing_src_type parse_backing_src_type(const char *type_name)
@@ -300,7 +309,7 @@ enum vm_mem_backing_src_type parse_backing_src_type(const char *type_name)
 		if (!strcmp(type_name, vm_mem_backing_src_alias(i)->name))
 			return i;
 
-	backing_src_help();
+	print_available_backing_src_types("");
 	TEST_FAIL("Unknown backing src type: %s", type_name);
 	return -1;
 }

tools/testing/selftests/kvm/rseq_test.c

@@ -180,6 +180,7 @@ int main(int argc, char *argv[])
 	 * CPU affinity.
 	 */
 	vm = vm_create_default(VCPU_ID, 0, guest_code);
+	ucall_init(vm, NULL);
 
 	pthread_create(&migration_thread, NULL, migration_worker, 0);
 

tools/testing/selftests/kvm/steal_time.c

@@ -116,12 +116,12 @@ struct st_time {
 	uint64_t st_time;
 };
 
-static int64_t smccc(uint32_t func, uint32_t arg)
+static int64_t smccc(uint32_t func, uint64_t arg)
 {
 	unsigned long ret;
 
 	asm volatile(
-		"mov	x0, %1\n"
+		"mov	w0, %w1\n"
 		"mov	x1, %2\n"
 		"hvc	#0\n"
 		"mov	%0, x0\n"

tools/testing/selftests/kvm/x86_64/svm_int_ctl_test.c

@@ -0,0 +1,128 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * svm_int_ctl_test
+ *
+ * Copyright (C) 2021, Red Hat, Inc.
+ *
+ * Nested SVM testing: test simultaneous use of V_IRQ from L1 and L0.
+ */
+
+#include "test_util.h"
+#include "kvm_util.h"
+#include "processor.h"
+#include "svm_util.h"
+#include "apic.h"
+
+#define VCPU_ID		0
+
+static struct kvm_vm *vm;
+
+bool vintr_irq_called;
+bool intr_irq_called;
+
+#define VINTR_IRQ_NUMBER 0x20
+#define INTR_IRQ_NUMBER 0x30
+
+static void vintr_irq_handler(struct ex_regs *regs)
+{
+	vintr_irq_called = true;
+}
+
+static void intr_irq_handler(struct ex_regs *regs)
+{
+	x2apic_write_reg(APIC_EOI, 0x00);
+	intr_irq_called = true;
+}
+
+static void l2_guest_code(struct svm_test_data *svm)
+{
+	/* This code raises interrupt INTR_IRQ_NUMBER in the L1's LAPIC,
+	 * and since L1 didn't enable virtual interrupt masking,
+	 * L2 should receive it and not L1.
+	 *
+	 * L2 also has virtual interrupt 'VINTR_IRQ_NUMBER' pending in V_IRQ
+	 * so it should also receive it after the following 'sti'.
+	 */
+	x2apic_write_reg(APIC_ICR,
+		APIC_DEST_SELF | APIC_INT_ASSERT | INTR_IRQ_NUMBER);
+
+	__asm__ __volatile__(
+		"sti\n"
+		"nop\n"
+	);
+
+	GUEST_ASSERT(vintr_irq_called);
+	GUEST_ASSERT(intr_irq_called);
+
+	__asm__ __volatile__(
+		"vmcall\n"
+	);
+}
+
+static void l1_guest_code(struct svm_test_data *svm)
+{
+	#define L2_GUEST_STACK_SIZE 64
+	unsigned long l2_guest_stack[L2_GUEST_STACK_SIZE];
+	struct vmcb *vmcb = svm->vmcb;
+
+	x2apic_enable();
+
+	/* Prepare for L2 execution. */
+	generic_svm_setup(svm, l2_guest_code,
+			  &l2_guest_stack[L2_GUEST_STACK_SIZE]);
+
+	/* No virtual interrupt masking */
+	vmcb->control.int_ctl &= ~V_INTR_MASKING_MASK;
+
+	/* No intercepts for real and virtual interrupts */
+	vmcb->control.intercept &= ~(1ULL << INTERCEPT_INTR | INTERCEPT_VINTR);
+
+	/* Make a virtual interrupt VINTR_IRQ_NUMBER pending */
+	vmcb->control.int_ctl |= V_IRQ_MASK | (0x1 << V_INTR_PRIO_SHIFT);
+	vmcb->control.int_vector = VINTR_IRQ_NUMBER;
+
+	run_guest(vmcb, svm->vmcb_gpa);
+	GUEST_ASSERT(vmcb->control.exit_code == SVM_EXIT_VMMCALL);
+	GUEST_DONE();
+}
+
+int main(int argc, char *argv[])
+{
+	vm_vaddr_t svm_gva;
+
+	nested_svm_check_supported();
+
+	vm = vm_create_default(VCPU_ID, 0, (void *) l1_guest_code);
+
+	vm_init_descriptor_tables(vm);
+	vcpu_init_descriptor_tables(vm, VCPU_ID);
+
+	vm_install_exception_handler(vm, VINTR_IRQ_NUMBER, vintr_irq_handler);
+	vm_install_exception_handler(vm, INTR_IRQ_NUMBER, intr_irq_handler);
+
+	vcpu_alloc_svm(vm, &svm_gva);
+	vcpu_args_set(vm, VCPU_ID, 1, svm_gva);
+
+	struct kvm_run *run = vcpu_state(vm, VCPU_ID);
+	struct ucall uc;
+
+	vcpu_run(vm, VCPU_ID);
+	TEST_ASSERT(run->exit_reason == KVM_EXIT_IO,
+		    "Got exit_reason other than KVM_EXIT_IO: %u (%s)\n",
+		    run->exit_reason,
+		    exit_reason_str(run->exit_reason));
+
+	switch (get_ucall(vm, VCPU_ID, &uc)) {
+	case UCALL_ABORT:
+		TEST_FAIL("%s", (const char *)uc.args[0]);
+		break;
+		/* NOT REACHED */
+	case UCALL_DONE:
+		goto done;
+	default:
+		TEST_FAIL("Unknown ucall 0x%lx.", uc.cmd);
+	}
+done:
+	kvm_vm_free(vm);
+	return 0;
+}