Merge tag 'kvm-x86-selftests-6.15' of https://github.com/kvm-x86/linux into HEAD

KVM selftests changes for 6.15, part 2

 - Fix a variety of flaws, bugs, and false failures/passes dirty_log_test, and
   improve its coverage by collecting all dirty entries on each iteration.

 - Fix a few minor bugs related to handling of stats FDs.

 - Add infrastructure to make vCPU and VM stats FDs available to tests by
   default (open the FDs during VM/vCPU creation).

 - Relax an assertion on the number of HLT exits in the xAPIC IPI test when
   running on a CPU that supports AMD's Idle HLT (which elides interception of
   HLT if a virtual IRQ is pending and unmasked).

 - Misc cleanups and fixes.

Author: bonzini

tools/testing/selftests/kvm/dirty_log_test.c

@@ -46,6 +46,12 @@ struct userspace_mem_region {
 	struct hlist_node slot_node;
 };
 
+struct kvm_binary_stats {
+	int fd;
+	struct kvm_stats_header header;
+	struct kvm_stats_desc *desc;
+};
+
 struct kvm_vcpu {
 	struct list_head list;
 	uint32_t id;
@@ -55,6 +61,7 @@ struct kvm_vcpu {
 #ifdef __x86_64__
 	struct kvm_cpuid2 *cpuid;
 #endif
+	struct kvm_binary_stats stats;
 	struct kvm_dirty_gfn *dirty_gfns;
 	uint32_t fetch_index;
 	uint32_t dirty_gfns_count;
@@ -99,10 +106,7 @@ struct kvm_vm {
 
 	struct kvm_vm_arch arch;
 
-	/* Cache of information for binary stats interface */
-	int stats_fd;
-	struct kvm_stats_header stats_header;
-	struct kvm_stats_desc *stats_desc;
+	struct kvm_binary_stats stats;
 
 	/*
 	 * KVM region slots. These are the default memslots used by page
@@ -531,16 +535,19 @@ void read_stat_data(int stats_fd, struct kvm_stats_header *header,
 		    struct kvm_stats_desc *desc, uint64_t *data,
 		    size_t max_elements);
 
-void __vm_get_stat(struct kvm_vm *vm, const char *stat_name, uint64_t *data,
-		   size_t max_elements);
+void kvm_get_stat(struct kvm_binary_stats *stats, const char *name,
+		  uint64_t *data, size_t max_elements);
 
-static inline uint64_t vm_get_stat(struct kvm_vm *vm, const char *stat_name)
-{
-	uint64_t data;
+#define __get_stat(stats, stat)							\
+({										\
+	uint64_t data;								\
+										\
+	kvm_get_stat(stats, #stat, &data, 1);					\
+	data;									\
+})
 
-	__vm_get_stat(vm, stat_name, &data, 1);
-	return data;
-}
+#define vm_get_stat(vm, stat) __get_stat(&(vm)->stats, stat)
+#define vcpu_get_stat(vcpu, stat) __get_stat(&(vcpu)->stats, stat)
 
 void vm_create_irqchip(struct kvm_vm *vm);
 
@@ -963,6 +970,8 @@ static inline struct kvm_vm *vm_create_shape_with_one_vcpu(struct vm_shape shape
 
 struct kvm_vcpu *vm_recreate_with_one_vcpu(struct kvm_vm *vm);
 
+void kvm_set_files_rlimit(uint32_t nr_vcpus);
+
 void kvm_pin_this_task_to_pcpu(uint32_t pcpu);
 void kvm_print_vcpu_pinning_help(void);
 void kvm_parse_vcpu_pinning(const char *pcpus_string, uint32_t vcpu_to_pcpu[],

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

@@ -200,6 +200,7 @@ struct kvm_x86_cpu_feature {
 #define X86_FEATURE_PAUSEFILTER         KVM_X86_CPU_FEATURE(0x8000000A, 0, EDX, 10)
 #define X86_FEATURE_PFTHRESHOLD         KVM_X86_CPU_FEATURE(0x8000000A, 0, EDX, 12)
 #define	X86_FEATURE_VGIF		KVM_X86_CPU_FEATURE(0x8000000A, 0, EDX, 16)
+#define X86_FEATURE_IDLE_HLT		KVM_X86_CPU_FEATURE(0x8000000A, 0, EDX, 30)
 #define X86_FEATURE_SEV			KVM_X86_CPU_FEATURE(0x8000001F, 0, EAX, 1)
 #define X86_FEATURE_SEV_ES		KVM_X86_CPU_FEATURE(0x8000001F, 0, EAX, 3)
 #define	X86_FEATURE_PERFMON_V2		KVM_X86_CPU_FEATURE(0x80000022, 0, EAX, 0)
@@ -1251,7 +1252,7 @@ void vm_install_exception_handler(struct kvm_vm *vm, int vector,
 	uint64_t ign_error_code;					\
 	uint8_t vector;							\
 									\
-	asm volatile(KVM_ASM_SAFE(insn)					\
+	asm volatile(KVM_ASM_SAFE_FEP(insn)				\
 		     : KVM_ASM_SAFE_OUTPUTS(vector, ign_error_code)	\
 		     : inputs						\
 		     : KVM_ASM_SAFE_CLOBBERS);				\

tools/testing/selftests/kvm/include/x86/processor.h

@@ -10,7 +10,6 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
-#include <sys/resource.h>
 
 #include "test_util.h"
 
@@ -39,36 +38,11 @@ int main(int argc, char *argv[])
 {
 	int kvm_max_vcpu_id = kvm_check_cap(KVM_CAP_MAX_VCPU_ID);
 	int kvm_max_vcpus = kvm_check_cap(KVM_CAP_MAX_VCPUS);
-	/*
-	 * Number of file descriptors reqired, KVM_CAP_MAX_VCPUS for vCPU fds +
-	 * an arbitrary number for everything else.
-	 */
-	int nr_fds_wanted = kvm_max_vcpus + 100;
-	struct rlimit rl;
 
 	pr_info("KVM_CAP_MAX_VCPU_ID: %d\n", kvm_max_vcpu_id);
 	pr_info("KVM_CAP_MAX_VCPUS: %d\n", kvm_max_vcpus);
 
-	/*
-	 * Check that we're allowed to open nr_fds_wanted file descriptors and
-	 * try raising the limits if needed.
-	 */
-	TEST_ASSERT(!getrlimit(RLIMIT_NOFILE, &rl), "getrlimit() failed!");
-
-	if (rl.rlim_cur < nr_fds_wanted) {
-		rl.rlim_cur = nr_fds_wanted;
-		if (rl.rlim_max < nr_fds_wanted) {
-			int old_rlim_max = rl.rlim_max;
-			rl.rlim_max = nr_fds_wanted;
-
-			int r = setrlimit(RLIMIT_NOFILE, &rl);
-			__TEST_REQUIRE(r >= 0,
-				       "RLIMIT_NOFILE hard limit is too low (%d, wanted %d)",
-				       old_rlim_max, nr_fds_wanted);
-		} else {
-			TEST_ASSERT(!setrlimit(RLIMIT_NOFILE, &rl), "setrlimit() failed!");
-		}
-	}
+	kvm_set_files_rlimit(kvm_max_vcpus);
 
 	/*
 	 * Upstream KVM prior to 4.8 does not support KVM_CAP_MAX_VCPU_ID.

tools/testing/selftests/kvm/kvm_create_max_vcpus.c

@@ -12,6 +12,7 @@
 #include <assert.h>
 #include <sched.h>
 #include <sys/mman.h>
+#include <sys/resource.h>
 #include <sys/types.h>
 #include <sys/stat.h>
 #include <unistd.h>
@@ -196,6 +197,11 @@ static void vm_open(struct kvm_vm *vm)
 
 	vm->fd = __kvm_ioctl(vm->kvm_fd, KVM_CREATE_VM, (void *)vm->type);
 	TEST_ASSERT(vm->fd >= 0, KVM_IOCTL_ERROR(KVM_CREATE_VM, vm->fd));
+
+	if (kvm_has_cap(KVM_CAP_BINARY_STATS_FD))
+		vm->stats.fd = vm_get_stats_fd(vm);
+	else
+		vm->stats.fd = -1;
 }
 
 const char *vm_guest_mode_string(uint32_t i)
@@ -406,6 +412,38 @@ static uint64_t vm_nr_pages_required(enum vm_guest_mode mode,
 	return vm_adjust_num_guest_pages(mode, nr_pages);
 }
 
+void kvm_set_files_rlimit(uint32_t nr_vcpus)
+{
+	/*
+	 * Each vCPU will open two file descriptors: the vCPU itself and the
+	 * vCPU's binary stats file descriptor.  Add an arbitrary amount of
+	 * buffer for all other files a test may open.
+	 */
+	int nr_fds_wanted = nr_vcpus * 2 + 100;
+	struct rlimit rl;
+
+	/*
+	 * Check that we're allowed to open nr_fds_wanted file descriptors and
+	 * try raising the limits if needed.
+	 */
+	TEST_ASSERT(!getrlimit(RLIMIT_NOFILE, &rl), "getrlimit() failed!");
+
+	if (rl.rlim_cur < nr_fds_wanted) {
+		rl.rlim_cur = nr_fds_wanted;
+		if (rl.rlim_max < nr_fds_wanted) {
+			int old_rlim_max = rl.rlim_max;
+
+			rl.rlim_max = nr_fds_wanted;
+			__TEST_REQUIRE(setrlimit(RLIMIT_NOFILE, &rl) >= 0,
+				       "RLIMIT_NOFILE hard limit is too low (%d, wanted %d)",
+				       old_rlim_max, nr_fds_wanted);
+		} else {
+			TEST_ASSERT(!setrlimit(RLIMIT_NOFILE, &rl), "setrlimit() failed!");
+		}
+	}
+
+}
+
 struct kvm_vm *__vm_create(struct vm_shape shape, uint32_t nr_runnable_vcpus,
 			   uint64_t nr_extra_pages)
 {
@@ -415,6 +453,8 @@ struct kvm_vm *__vm_create(struct vm_shape shape, uint32_t nr_runnable_vcpus,
 	struct kvm_vm *vm;
 	int i;
 
+	kvm_set_files_rlimit(nr_runnable_vcpus);
+
 	pr_debug("%s: mode='%s' type='%d', pages='%ld'\n", __func__,
 		 vm_guest_mode_string(shape.mode), shape.type, nr_pages);
 
@@ -657,6 +697,23 @@ userspace_mem_region_find(struct kvm_vm *vm, uint64_t start, uint64_t end)
 	return NULL;
 }
 
+static void kvm_stats_release(struct kvm_binary_stats *stats)
+{
+	int ret;
+
+	if (stats->fd < 0)
+		return;
+
+	if (stats->desc) {
+		free(stats->desc);
+		stats->desc = NULL;
+	}
+
+	ret = close(stats->fd);
+	TEST_ASSERT(!ret,  __KVM_SYSCALL_ERROR("close()", ret));
+	stats->fd = -1;
+}
+
 __weak void vcpu_arch_free(struct kvm_vcpu *vcpu)
 {
 
@@ -690,6 +747,8 @@ static void vm_vcpu_rm(struct kvm_vm *vm, struct kvm_vcpu *vcpu)
 	ret = close(vcpu->fd);
 	TEST_ASSERT(!ret,  __KVM_SYSCALL_ERROR("close()", ret));
 
+	kvm_stats_release(&vcpu->stats);
+
 	list_del(&vcpu->list);
 
 	vcpu_arch_free(vcpu);
@@ -709,6 +768,9 @@ void kvm_vm_release(struct kvm_vm *vmp)
 
 	ret = close(vmp->kvm_fd);
 	TEST_ASSERT(!ret,  __KVM_SYSCALL_ERROR("close()", ret));
+
+	/* Free cached stats metadata and close FD */
+	kvm_stats_release(&vmp->stats);
 }
 
 static void __vm_mem_region_delete(struct kvm_vm *vm,
@@ -748,12 +810,6 @@ void kvm_vm_free(struct kvm_vm *vmp)
 	if (vmp == NULL)
 		return;
 
-	/* Free cached stats metadata and close FD */
-	if (vmp->stats_fd) {
-		free(vmp->stats_desc);
-		close(vmp->stats_fd);
-	}
-
 	/* Free userspace_mem_regions. */
 	hash_for_each_safe(vmp->regions.slot_hash, ctr, node, region, slot_node)
 		__vm_mem_region_delete(vmp, region);
@@ -1286,6 +1342,11 @@ struct kvm_vcpu *__vm_vcpu_add(struct kvm_vm *vm, uint32_t vcpu_id)
 	TEST_ASSERT(vcpu->run != MAP_FAILED,
 		    __KVM_SYSCALL_ERROR("mmap()", (int)(unsigned long)MAP_FAILED));
 
+	if (kvm_has_cap(KVM_CAP_BINARY_STATS_FD))
+		vcpu->stats.fd = vcpu_get_stats_fd(vcpu);
+	else
+		vcpu->stats.fd = -1;
+
 	/* Add to linked-list of VCPUs. */
 	list_add(&vcpu->list, &vm->vcpus);
 
@@ -2198,46 +2259,31 @@ void read_stat_data(int stats_fd, struct kvm_stats_header *header,
 		    desc->name, size, ret);
 }
 
-/*
- * Read the data of the named stat
- *
- * Input Args:
- *   vm - the VM for which the stat should be read
- *   stat_name - the name of the stat to read
- *   max_elements - the maximum number of 8-byte values to read into data
- *
- * Output Args:
- *   data - the buffer into which stat data should be read
- *
- * Read the data values of a specified stat from the binary stats interface.
- */
-void __vm_get_stat(struct kvm_vm *vm, const char *stat_name, uint64_t *data,
-		   size_t max_elements)
+void kvm_get_stat(struct kvm_binary_stats *stats, const char *name,
+		  uint64_t *data, size_t max_elements)
 {
 	struct kvm_stats_desc *desc;
 	size_t size_desc;
 	int i;
 
-	if (!vm->stats_fd) {
-		vm->stats_fd = vm_get_stats_fd(vm);
-		read_stats_header(vm->stats_fd, &vm->stats_header);
-		vm->stats_desc = read_stats_descriptors(vm->stats_fd,
-							&vm->stats_header);
+	if (!stats->desc) {
+		read_stats_header(stats->fd, &stats->header);
+		stats->desc = read_stats_descriptors(stats->fd, &stats->header);
 	}
 
-	size_desc = get_stats_descriptor_size(&vm->stats_header);
+	size_desc = get_stats_descriptor_size(&stats->header);
 
-	for (i = 0; i < vm->stats_header.num_desc; ++i) {
-		desc = (void *)vm->stats_desc + (i * size_desc);
+	for (i = 0; i < stats->header.num_desc; ++i) {
+		desc = (void *)stats->desc + (i * size_desc);
 
-		if (strcmp(desc->name, stat_name))
+		if (strcmp(desc->name, name))
 			continue;
 
-		read_stat_data(vm->stats_fd, &vm->stats_header, desc,
-			       data, max_elements);
-
-		break;
+		read_stat_data(stats->fd, &stats->header, desc, data, max_elements);
+		return;
 	}
+
+	TEST_FAIL("Unable to find stat '%s'", name);
 }
 
 __weak void kvm_arch_vm_post_create(struct kvm_vm *vm)

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

@@ -114,7 +114,7 @@ struct uffd_desc *uffd_setup_demand_paging(int uffd_mode, useconds_t delay,
 
 	PER_PAGE_DEBUG("Userfaultfd %s mode, faults resolved with %s\n",
 		       is_minor ? "MINOR" : "MISSING",
-		       is_minor ? "UFFDIO_CONINUE" : "UFFDIO_COPY");
+		       is_minor ? "UFFDIO_CONTINUE" : "UFFDIO_COPY");
 
 	uffd_desc = malloc(sizeof(struct uffd_desc));
 	TEST_ASSERT(uffd_desc, "Failed to malloc uffd descriptor");

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

@@ -41,9 +41,9 @@ struct kvm_page_stats {
 
 static void get_page_stats(struct kvm_vm *vm, struct kvm_page_stats *stats, const char *stage)
 {
-	stats->pages_4k = vm_get_stat(vm, "pages_4k");
-	stats->pages_2m = vm_get_stat(vm, "pages_2m");
-	stats->pages_1g = vm_get_stat(vm, "pages_1g");
+	stats->pages_4k = vm_get_stat(vm, pages_4k);
+	stats->pages_2m = vm_get_stat(vm, pages_2m);
+	stats->pages_1g = vm_get_stat(vm, pages_1g);
 	stats->hugepages = stats->pages_2m + stats->pages_1g;
 
 	pr_debug("\nPage stats after %s: 4K: %ld 2M: %ld 1G: %ld huge: %ld\n",

tools/testing/selftests/kvm/x86/dirty_log_page_splitting_test.c

@@ -73,7 +73,7 @@ static void check_2m_page_count(struct kvm_vm *vm, int expected_pages_2m)
 {
 	int actual_pages_2m;
 
-	actual_pages_2m = vm_get_stat(vm, "pages_2m");
+	actual_pages_2m = vm_get_stat(vm, pages_2m);
 
 	TEST_ASSERT(actual_pages_2m == expected_pages_2m,
 		    "Unexpected 2m page count. Expected %d, got %d",
@@ -84,7 +84,7 @@ static void check_split_count(struct kvm_vm *vm, int expected_splits)
 {
 	int actual_splits;
 
-	actual_splits = vm_get_stat(vm, "nx_lpage_splits");
+	actual_splits = vm_get_stat(vm, nx_lpage_splits);
 
 	TEST_ASSERT(actual_splits == expected_splits,
 		    "Unexpected NX huge page split count. Expected %d, got %d",

tools/testing/selftests/kvm/x86/nx_huge_pages_test.c

@@ -466,6 +466,19 @@ int main(int argc, char *argv[])
 	cancel_join_vcpu_thread(threads[0], params[0].vcpu);
 	cancel_join_vcpu_thread(threads[1], params[1].vcpu);
 
+	/*
+	 * If the host support Idle HLT, i.e. KVM *might* be using Idle HLT,
+	 * then the number of HLT exits may be less than the number of HLTs
+	 * that were executed, as Idle HLT elides the exit if the vCPU has an
+	 * unmasked, pending IRQ (or NMI).
+	 */
+	if (this_cpu_has(X86_FEATURE_IDLE_HLT))
+		TEST_ASSERT(data->hlt_count >= vcpu_get_stat(params[0].vcpu, halt_exits),
+			    "HLT insns = %lu, HLT exits = %lu",
+			    data->hlt_count, vcpu_get_stat(params[0].vcpu, halt_exits));
+	else
+		TEST_ASSERT_EQ(data->hlt_count, vcpu_get_stat(params[0].vcpu, halt_exits));
+
 	fprintf(stderr,
 		"Test successful after running for %d seconds.\n"
 		"Sending vCPU sent %lu IPIs to halting vCPU\n"