Merge branches 'slab/for-5.19/stackdepot' and 'slab/for-5.19/refactor' into slab/for-linus

Documentation/vm/slub.rst

@@ -384,5 +384,69 @@ c) Execute ``slabinfo-gnuplot.sh`` in '-t' mode, passing all of the
       40,60`` range will plot only samples collected between 40th and
       60th seconds).
 
+
+DebugFS files for SLUB
+======================
+
+For more information about current state of SLUB caches with the user tracking
+debug option enabled, debugfs files are available, typically under
+/sys/kernel/debug/slab/<cache>/ (created only for caches with enabled user
+tracking). There are 2 types of these files with the following debug
+information:
+
+1. alloc_traces::
+
+    Prints information about unique allocation traces of the currently
+    allocated objects. The output is sorted by frequency of each trace.
+
+    Information in the output:
+    Number of objects, allocating function, minimal/average/maximal jiffies since alloc,
+    pid range of the allocating processes, cpu mask of allocating cpus, and stack trace.
+
+    Example:::
+
+    1085 populate_error_injection_list+0x97/0x110 age=166678/166680/166682 pid=1 cpus=1::
+	__slab_alloc+0x6d/0x90
+	kmem_cache_alloc_trace+0x2eb/0x300
+	populate_error_injection_list+0x97/0x110
+	init_error_injection+0x1b/0x71
+	do_one_initcall+0x5f/0x2d0
+	kernel_init_freeable+0x26f/0x2d7
+	kernel_init+0xe/0x118
+	ret_from_fork+0x22/0x30
+
+
+2. free_traces::
+
+    Prints information about unique freeing traces of the currently allocated
+    objects. The freeing traces thus come from the previous life-cycle of the
+    objects and are reported as not available for objects allocated for the first
+    time. The output is sorted by frequency of each trace.
+
+    Information in the output:
+    Number of objects, freeing function, minimal/average/maximal jiffies since free,
+    pid range of the freeing processes, cpu mask of freeing cpus, and stack trace.
+
+    Example:::
+
+    1980 <not-available> age=4294912290 pid=0 cpus=0
+    51 acpi_ut_update_ref_count+0x6a6/0x782 age=236886/237027/237772 pid=1 cpus=1
+	kfree+0x2db/0x420
+	acpi_ut_update_ref_count+0x6a6/0x782
+	acpi_ut_update_object_reference+0x1ad/0x234
+	acpi_ut_remove_reference+0x7d/0x84
+	acpi_rs_get_prt_method_data+0x97/0xd6
+	acpi_get_irq_routing_table+0x82/0xc4
+	acpi_pci_irq_find_prt_entry+0x8e/0x2e0
+	acpi_pci_irq_lookup+0x3a/0x1e0
+	acpi_pci_irq_enable+0x77/0x240
+	pcibios_enable_device+0x39/0x40
+	do_pci_enable_device.part.0+0x5d/0xe0
+	pci_enable_device_flags+0xfc/0x120
+	pci_enable_device+0x13/0x20
+	virtio_pci_probe+0x9e/0x170
+	local_pci_probe+0x48/0x80
+	pci_device_probe+0x105/0x1c0
+
 Christoph Lameter, May 30, 2007
 Sergey Senozhatsky, October 23, 2015

include/linux/slub_def.h

@@ -105,7 +105,6 @@ struct kmem_cache {
 	struct kmem_cache_order_objects oo;
 
 	/* Allocation and freeing of slabs */
-	struct kmem_cache_order_objects max;
 	struct kmem_cache_order_objects min;
 	gfp_t allocflags;	/* gfp flags to use on each alloc */
 	int refcount;		/* Refcount for slab cache destroy */

include/linux/stackdepot.h

@@ -20,18 +20,36 @@ depot_stack_handle_t __stack_depot_save(unsigned long *entries,
 					gfp_t gfp_flags, bool can_alloc);
 
 /*
- * Every user of stack depot has to call this during its own init when it's
- * decided that it will be calling stack_depot_save() later.
+ * Every user of stack depot has to call stack_depot_init() during its own init
+ * when it's decided that it will be calling stack_depot_save() later. This is
+ * recommended for e.g. modules initialized later in the boot process, when
+ * slab_is_available() is true.
  *
  * The alternative is to select STACKDEPOT_ALWAYS_INIT to have stack depot
  * enabled as part of mm_init(), for subsystems where it's known at compile time
  * that stack depot will be used.
+ *
+ * Another alternative is to call stack_depot_want_early_init(), when the
+ * decision to use stack depot is taken e.g. when evaluating kernel boot
+ * parameters, which precedes the enablement point in mm_init().
+ *
+ * stack_depot_init() and stack_depot_want_early_init() can be called regardless
+ * of CONFIG_STACKDEPOT and are no-op when disabled. The actual save/fetch/print
+ * functions should only be called from code that makes sure CONFIG_STACKDEPOT
+ * is enabled.
  */
+#ifdef CONFIG_STACKDEPOT
 int stack_depot_init(void);
 
-#ifdef CONFIG_STACKDEPOT_ALWAYS_INIT
-static inline int stack_depot_early_init(void)	{ return stack_depot_init(); }
+void __init stack_depot_want_early_init(void);
+
+/* This is supposed to be called only from mm_init() */
+int __init stack_depot_early_init(void);
 #else
+static inline int stack_depot_init(void) { return 0; }
+
+static inline void stack_depot_want_early_init(void) { }
+
 static inline int stack_depot_early_init(void)	{ return 0; }
 #endif
 

init/Kconfig

@@ -1875,6 +1875,7 @@ config SLUB_DEBUG
 	default y
 	bool "Enable SLUB debugging support" if EXPERT
 	depends on SLUB && SYSFS
+	select STACKDEPOT if STACKTRACE_SUPPORT
 	help
 	  SLUB has extensive debug support features. Disabling these can
 	  result in significant savings in code size. This also disables

lib/Kconfig.debug

@@ -709,6 +709,7 @@ config DEBUG_SLAB
 config SLUB_DEBUG_ON
 	bool "SLUB debugging on by default"
 	depends on SLUB && SLUB_DEBUG
+	select STACKDEPOT_ALWAYS_INIT if STACKTRACE_SUPPORT
 	default n
 	help
 	  Boot with debugging on by default. SLUB boots by default with

lib/stackdepot.c

@@ -66,6 +66,9 @@ struct stack_record {
 	unsigned long entries[];	/* Variable-sized array of entries. */
 };
 
+static bool __stack_depot_want_early_init __initdata = IS_ENABLED(CONFIG_STACKDEPOT_ALWAYS_INIT);
+static bool __stack_depot_early_init_passed __initdata;
+
 static void *stack_slabs[STACK_ALLOC_MAX_SLABS];
 
 static int depot_index;
@@ -162,38 +165,58 @@ static int __init is_stack_depot_disabled(char *str)
 }
 early_param("stack_depot_disable", is_stack_depot_disabled);
 
-/*
- * __ref because of memblock_alloc(), which will not be actually called after
- * the __init code is gone, because at that point slab_is_available() is true
- */
-__ref int stack_depot_init(void)
+void __init stack_depot_want_early_init(void)
+{
+	/* Too late to request early init now */
+	WARN_ON(__stack_depot_early_init_passed);
+
+	__stack_depot_want_early_init = true;
+}
+
+int __init stack_depot_early_init(void)
+{
+	size_t size;
+
+	/* This is supposed to be called only once, from mm_init() */
+	if (WARN_ON(__stack_depot_early_init_passed))
+		return 0;
+
+	__stack_depot_early_init_passed = true;
+
+	if (!__stack_depot_want_early_init || stack_depot_disable)
+		return 0;
+
+	size = (STACK_HASH_SIZE * sizeof(struct stack_record *));
+	pr_info("Stack Depot early init allocating hash table with memblock_alloc, %zu bytes\n",
+		size);
+	stack_table = memblock_alloc(size, SMP_CACHE_BYTES);
+
+	if (!stack_table) {
+		pr_err("Stack Depot hash table allocation failed, disabling\n");
+		stack_depot_disable = true;
+		return -ENOMEM;
+	}
+
+	return 0;
+}
+
+int stack_depot_init(void)
 {
 	static DEFINE_MUTEX(stack_depot_init_mutex);
+	int ret = 0;
 
 	mutex_lock(&stack_depot_init_mutex);
 	if (!stack_depot_disable && !stack_table) {
-		size_t size = (STACK_HASH_SIZE * sizeof(struct stack_record *));
-		int i;
-
-		if (slab_is_available()) {
-			pr_info("Stack Depot allocating hash table with kvmalloc\n");
-			stack_table = kvmalloc(size, GFP_KERNEL);
-		} else {
-			pr_info("Stack Depot allocating hash table with memblock_alloc\n");
-			stack_table = memblock_alloc(size, SMP_CACHE_BYTES);
-		}
-		if (stack_table) {
-			for (i = 0; i < STACK_HASH_SIZE;  i++)
-				stack_table[i] = NULL;
-		} else {
+		pr_info("Stack Depot allocating hash table with kvcalloc\n");
+		stack_table = kvcalloc(STACK_HASH_SIZE, sizeof(struct stack_record *), GFP_KERNEL);
+		if (!stack_table) {
 			pr_err("Stack Depot hash table allocation failed, disabling\n");
 			stack_depot_disable = true;
-			mutex_unlock(&stack_depot_init_mutex);
-			return -ENOMEM;
+			ret = -ENOMEM;
 		}
 	}
 	mutex_unlock(&stack_depot_init_mutex);
-	return 0;
+	return ret;
 }
 EXPORT_SYMBOL_GPL(stack_depot_init);
 

mm/page_owner.c

@@ -45,7 +45,12 @@ static void init_early_allocated_pages(void);
 
 static int __init early_page_owner_param(char *buf)
 {
-	return kstrtobool(buf, &page_owner_enabled);
+	int ret = kstrtobool(buf, &page_owner_enabled);
+
+	if (page_owner_enabled)
+		stack_depot_want_early_init();
+
+	return ret;
 }
 early_param("page_owner", early_page_owner_param);
 
@@ -83,8 +88,6 @@ static __init void init_page_owner(void)
 	if (!page_owner_enabled)
 		return;
 
-	stack_depot_init();
-
 	register_dummy_stack();
 	register_failure_stack();
 	register_early_stack();

mm/slab_common.c

@@ -24,6 +24,7 @@
 #include <asm/tlbflush.h>
 #include <asm/page.h>
 #include <linux/memcontrol.h>
+#include <linux/stackdepot.h>
 
 #define CREATE_TRACE_POINTS
 #include <trace/events/kmem.h>
@@ -314,9 +315,13 @@ kmem_cache_create_usercopy(const char *name,
 	 * If no slub_debug was enabled globally, the static key is not yet
 	 * enabled by setup_slub_debug(). Enable it if the cache is being
 	 * created with any of the debugging flags passed explicitly.
+	 * It's also possible that this is the first cache created with
+	 * SLAB_STORE_USER and we should init stack_depot for it.
 	 */
 	if (flags & SLAB_DEBUG_FLAGS)
 		static_branch_enable(&slub_debug_enabled);
+	if (flags & SLAB_STORE_USER)
+		stack_depot_init();
 #endif
 
 	mutex_lock(&slab_mutex);
@@ -849,6 +854,8 @@ new_kmalloc_cache(int idx, enum kmalloc_cache_type type, slab_flags_t flags)
 			return;
 		}
 		flags |= SLAB_ACCOUNT;
+	} else if (IS_ENABLED(CONFIG_ZONE_DMA) && (type == KMALLOC_DMA)) {
+		flags |= SLAB_CACHE_DMA;
 	}
 
 	kmalloc_caches[type][idx] = create_kmalloc_cache(
@@ -877,7 +884,7 @@ void __init create_kmalloc_caches(slab_flags_t flags)
 	/*
 	 * Including KMALLOC_CGROUP if CONFIG_MEMCG_KMEM defined
 	 */
-	for (type = KMALLOC_NORMAL; type <= KMALLOC_RECLAIM; type++) {
+	for (type = KMALLOC_NORMAL; type < NR_KMALLOC_TYPES; type++) {
 		for (i = KMALLOC_SHIFT_LOW; i <= KMALLOC_SHIFT_HIGH; i++) {
 			if (!kmalloc_caches[type][i])
 				new_kmalloc_cache(i, type, flags);
@@ -898,20 +905,6 @@ void __init create_kmalloc_caches(slab_flags_t flags)
 
 	/* Kmalloc array is now usable */
 	slab_state = UP;
-
-#ifdef CONFIG_ZONE_DMA
-	for (i = 0; i <= KMALLOC_SHIFT_HIGH; i++) {
-		struct kmem_cache *s = kmalloc_caches[KMALLOC_NORMAL][i];
-
-		if (s) {
-			kmalloc_caches[KMALLOC_DMA][i] = create_kmalloc_cache(
-				kmalloc_info[i].name[KMALLOC_DMA],
-				kmalloc_info[i].size,
-				SLAB_CACHE_DMA | flags, 0,
-				kmalloc_info[i].size);
-		}
-	}
-#endif
 }
 #endif /* !CONFIG_SLOB */
 

mm/slub.c

@@ -26,6 +26,7 @@
 #include <linux/cpuset.h>
 #include <linux/mempolicy.h>
 #include <linux/ctype.h>
+#include <linux/stackdepot.h>
 #include <linux/debugobjects.h>
 #include <linux/kallsyms.h>
 #include <linux/kfence.h>
@@ -37,6 +38,7 @@
 #include <linux/memcontrol.h>
 #include <linux/random.h>
 #include <kunit/test.h>
+#include <linux/sort.h>
 
 #include <linux/debugfs.h>
 #include <trace/events/kmem.h>
@@ -264,8 +266,8 @@ static inline bool kmem_cache_has_cpu_partial(struct kmem_cache *s)
 #define TRACK_ADDRS_COUNT 16
 struct track {
 	unsigned long addr;	/* Called from address */
-#ifdef CONFIG_STACKTRACE
-	unsigned long addrs[TRACK_ADDRS_COUNT];	/* Called from address */
+#ifdef CONFIG_STACKDEPOT
+	depot_stack_handle_t handle;
 #endif
 	int cpu;		/* Was running on cpu */
 	int pid;		/* Pid context */
@@ -724,57 +726,51 @@ static struct track *get_track(struct kmem_cache *s, void *object,
 	return kasan_reset_tag(p + alloc);
 }
 
-static void set_track(struct kmem_cache *s, void *object,
+static void noinline set_track(struct kmem_cache *s, void *object,
 			enum track_item alloc, unsigned long addr)
 {
 	struct track *p = get_track(s, object, alloc);
 
-	if (addr) {
-#ifdef CONFIG_STACKTRACE
+#ifdef CONFIG_STACKDEPOT
+	unsigned long entries[TRACK_ADDRS_COUNT];
 	unsigned int nr_entries;
 
-		metadata_access_enable();
-		nr_entries = stack_trace_save(kasan_reset_tag(p->addrs),
-					      TRACK_ADDRS_COUNT, 3);
-		metadata_access_disable();
-
-		if (nr_entries < TRACK_ADDRS_COUNT)
-			p->addrs[nr_entries] = 0;
+	nr_entries = stack_trace_save(entries, ARRAY_SIZE(entries), 3);
+	p->handle = stack_depot_save(entries, nr_entries, GFP_NOWAIT);
 #endif
+
 	p->addr = addr;
 	p->cpu = smp_processor_id();
 	p->pid = current->pid;
 	p->when = jiffies;
-	} else {
-		memset(p, 0, sizeof(struct track));
-	}
 }
 
 static void init_tracking(struct kmem_cache *s, void *object)
 {
+	struct track *p;
+
 	if (!(s->flags & SLAB_STORE_USER))
 		return;
 
-	set_track(s, object, TRACK_FREE, 0UL);
-	set_track(s, object, TRACK_ALLOC, 0UL);
+	p = get_track(s, object, TRACK_ALLOC);
+	memset(p, 0, 2*sizeof(struct track));
 }
 
 static void print_track(const char *s, struct track *t, unsigned long pr_time)
 {
+	depot_stack_handle_t handle __maybe_unused;
+
 	if (!t->addr)
 		return;
 
 	pr_err("%s in %pS age=%lu cpu=%u pid=%d\n",
 	       s, (void *)t->addr, pr_time - t->when, t->cpu, t->pid);
-#ifdef CONFIG_STACKTRACE
-	{
-		int i;
-		for (i = 0; i < TRACK_ADDRS_COUNT; i++)
-			if (t->addrs[i])
-				pr_err("\t%pS\n", (void *)t->addrs[i]);
+#ifdef CONFIG_STACKDEPOT
+	handle = READ_ONCE(t->handle);
+	if (handle)
+		stack_depot_print(handle);
 	else
-				break;
-	}
+		pr_err("object allocation/free stack trace missing\n");
 #endif
 }
 
@@ -1532,6 +1528,8 @@ static int __init setup_slub_debug(char *str)
 			global_slub_debug_changed = true;
 		} else {
 			slab_list_specified = true;
+			if (flags & SLAB_STORE_USER)
+				stack_depot_want_early_init();
 		}
 	}
 
@@ -1549,6 +1547,8 @@ static int __init setup_slub_debug(char *str)
 	}
 out:
 	slub_debug = global_flags;
+	if (slub_debug & SLAB_STORE_USER)
+		stack_depot_want_early_init();
 	if (slub_debug != 0 || slub_debug_string)
 		static_branch_enable(&slub_debug_enabled);
 	else
@@ -4162,8 +4162,6 @@ static int calculate_sizes(struct kmem_cache *s)
 	 */
 	s->oo = oo_make(order, size);
 	s->min = oo_make(get_order(size), size);
-	if (oo_objects(s->oo) > oo_objects(s->max))
-		s->max = s->oo;
 
 	return !!oo_objects(s->oo);
 }
@@ -4341,18 +4339,26 @@ void kmem_obj_info(struct kmem_obj_info *kpp, void *object, struct slab *slab)
 	objp = fixup_red_left(s, objp);
 	trackp = get_track(s, objp, TRACK_ALLOC);
 	kpp->kp_ret = (void *)trackp->addr;
-#ifdef CONFIG_STACKTRACE
-	for (i = 0; i < KS_ADDRS_COUNT && i < TRACK_ADDRS_COUNT; i++) {
-		kpp->kp_stack[i] = (void *)trackp->addrs[i];
-		if (!kpp->kp_stack[i])
-			break;
+#ifdef CONFIG_STACKDEPOT
+	{
+		depot_stack_handle_t handle;
+		unsigned long *entries;
+		unsigned int nr_entries;
+
+		handle = READ_ONCE(trackp->handle);
+		if (handle) {
+			nr_entries = stack_depot_fetch(handle, &entries);
+			for (i = 0; i < KS_ADDRS_COUNT && i < nr_entries; i++)
+				kpp->kp_stack[i] = (void *)entries[i];
 		}
 
 		trackp = get_track(s, objp, TRACK_FREE);
-	for (i = 0; i < KS_ADDRS_COUNT && i < TRACK_ADDRS_COUNT; i++) {
-		kpp->kp_free_stack[i] = (void *)trackp->addrs[i];
-		if (!kpp->kp_free_stack[i])
-			break;
+		handle = READ_ONCE(trackp->handle);
+		if (handle) {
+			nr_entries = stack_depot_fetch(handle, &entries);
+			for (i = 0; i < KS_ADDRS_COUNT && i < nr_entries; i++)
+				kpp->kp_free_stack[i] = (void *)entries[i];
+		}
 	}
 #endif
 #endif
@@ -5054,6 +5060,7 @@ EXPORT_SYMBOL(validate_slab_cache);
  */
 
 struct location {
+	depot_stack_handle_t handle;
 	unsigned long count;
 	unsigned long addr;
 	long long sum_time;
@@ -5106,9 +5113,13 @@ static int add_location(struct loc_track *t, struct kmem_cache *s,
 {
 	long start, end, pos;
 	struct location *l;
-	unsigned long caddr;
+	unsigned long caddr, chandle;
 	unsigned long age = jiffies - track->when;
+	depot_stack_handle_t handle = 0;
 
+#ifdef CONFIG_STACKDEPOT
+	handle = READ_ONCE(track->handle);
+#endif
 	start = -1;
 	end = t->count;
 
@@ -5123,7 +5134,8 @@ static int add_location(struct loc_track *t, struct kmem_cache *s,
 			break;
 
 		caddr = t->loc[pos].addr;
-		if (track->addr == caddr) {
+		chandle = t->loc[pos].handle;
+		if ((track->addr == caddr) && (handle == chandle)) {
 
 			l = &t->loc[pos];
 			l->count++;
@@ -5148,6 +5160,8 @@ static int add_location(struct loc_track *t, struct kmem_cache *s,
 
 		if (track->addr < caddr)
 			end = pos;
+		else if (track->addr == caddr && handle < chandle)
+			end = pos;
 		else
 			start = pos;
 	}
@@ -5170,6 +5184,7 @@ static int add_location(struct loc_track *t, struct kmem_cache *s,
 	l->max_time = age;
 	l->min_pid = track->pid;
 	l->max_pid = track->pid;
+	l->handle = handle;
 	cpumask_clear(to_cpumask(l->cpus));
 	cpumask_set_cpu(track->cpu, to_cpumask(l->cpus));
 	nodes_clear(l->nodes);
@@ -6079,6 +6094,21 @@ static int slab_debugfs_show(struct seq_file *seq, void *v)
 			seq_printf(seq, " nodes=%*pbl",
 				 nodemask_pr_args(&l->nodes));
 
+#ifdef CONFIG_STACKDEPOT
+		{
+			depot_stack_handle_t handle;
+			unsigned long *entries;
+			unsigned int nr_entries, j;
+
+			handle = READ_ONCE(l->handle);
+			if (handle) {
+				nr_entries = stack_depot_fetch(handle, &entries);
+				seq_puts(seq, "\n");
+				for (j = 0; j < nr_entries; j++)
+					seq_printf(seq, "        %pS\n", (void *)entries[j]);
+			}
+		}
+#endif
 		seq_puts(seq, "\n");
 	}
 
@@ -6103,6 +6133,17 @@ static void *slab_debugfs_next(struct seq_file *seq, void *v, loff_t *ppos)
 	return NULL;
 }
 
+static int cmp_loc_by_count(const void *a, const void *b, const void *data)
+{
+	struct location *loc1 = (struct location *)a;
+	struct location *loc2 = (struct location *)b;
+
+	if (loc1->count > loc2->count)
+		return -1;
+	else
+		return 1;
+}
+
 static void *slab_debugfs_start(struct seq_file *seq, loff_t *ppos)
 {
 	struct loc_track *t = seq->private;
@@ -6164,6 +6205,10 @@ static int slab_debug_trace_open(struct inode *inode, struct file *filep)
 		spin_unlock_irqrestore(&n->list_lock, flags);
 	}
 
+	/* Sort locations by count */
+	sort_r(t->loc, t->count, sizeof(struct location),
+		cmp_loc_by_count, NULL, NULL);
+
 	bitmap_free(obj_map);
 	return 0;
 }