[mirror_ubuntu-zesty-kernel.git] / mm / slab_common.c

/*
 * Slab allocator functions that are independent of the allocator strategy
 *
 * (C) 2012 Christoph Lameter <cl@linux.com>
 */
#include <linux/slab.h>

#include <linux/mm.h>
#include <linux/poison.h>
#include <linux/interrupt.h>
#include <linux/memory.h>
#include <linux/compiler.h>
#include <linux/module.h>
#include <linux/cpu.h>
#include <linux/uaccess.h>
#include <linux/seq_file.h>
#include <linux/proc_fs.h>
#include <asm/cacheflush.h>
#include <asm/tlbflush.h>
#include <asm/page.h>

#include "slab.h"

enum slab_state slab_state;
LIST_HEAD(slab_caches);
DEFINE_MUTEX(slab_mutex);
struct kmem_cache *kmem_cache;

#ifdef CONFIG_DEBUG_VM
static int kmem_cache_sanity_check(const char *name, size_t size)
{
	struct kmem_cache *s = NULL;

	if (!name || in_interrupt() || size < sizeof(void *) ||
		size > KMALLOC_MAX_SIZE) {
		pr_err("kmem_cache_create(%s) integrity check failed\n", name);
		return -EINVAL;
	}

	list_for_each_entry(s, &slab_caches, list) {
		char tmp;
		int res;

		/*
		 * This happens when the module gets unloaded and doesn't
		 * destroy its slab cache and no-one else reuses the vmalloc
		 * area of the module.  Print a warning.
		 */
		res = probe_kernel_address(s->name, tmp);
		if (res) {
			pr_err("Slab cache with size %d has lost its name\n",
			       s->object_size);
			continue;
		}

		if (!strcmp(s->name, name)) {
			pr_err("%s (%s): Cache name already exists.\n",
			       __func__, name);
			dump_stack();
			s = NULL;
			return -EINVAL;
		}
	}

	WARN_ON(strchr(name, ' '));	/* It confuses parsers */
	return 0;
}
#else
static inline int kmem_cache_sanity_check(const char *name, size_t size)
{
	return 0;
}
#endif

/*
 * kmem_cache_create - Create a cache.
 * @name: A string which is used in /proc/slabinfo to identify this cache.
 * @size: The size of objects to be created in this cache.
 * @align: The required alignment for the objects.
 * @flags: SLAB flags
 * @ctor: A constructor for the objects.
 *
 * Returns a ptr to the cache on success, NULL on failure.
 * Cannot be called within a interrupt, but can be interrupted.
 * The @ctor is run when new pages are allocated by the cache.
 *
 * The flags are
 *
 * %SLAB_POISON - Poison the slab with a known test pattern (a5a5a5a5)
 * to catch references to uninitialised memory.
 *
 * %SLAB_RED_ZONE - Insert `Red' zones around the allocated memory to check
 * for buffer overruns.
 *
 * %SLAB_HWCACHE_ALIGN - Align the objects in this cache to a hardware
 * cacheline.  This can be beneficial if you're counting cycles as closely
 * as davem.
 */

struct kmem_cache *kmem_cache_create(const char *name, size_t size, size_t align,
		unsigned long flags, void (*ctor)(void *))
{
	struct kmem_cache *s = NULL;
	int err = 0;

	get_online_cpus();
	mutex_lock(&slab_mutex);

	if (!kmem_cache_sanity_check(name, size) == 0)
		goto out_locked;

	/*
	 * Some allocators will constraint the set of valid flags to a subset
	 * of all flags. We expect them to define CACHE_CREATE_MASK in this
	 * case, and we'll just provide them with a sanitized version of the
	 * passed flags.
	 */
	flags &= CACHE_CREATE_MASK;

	s = __kmem_cache_alias(name, size, align, flags, ctor);
	if (s)
		goto out_locked;

	s = kmem_cache_zalloc(kmem_cache, GFP_KERNEL);
	if (s) {
		s->object_size = s->size = size;
		s->align = align;
		s->ctor = ctor;
		s->name = kstrdup(name, GFP_KERNEL);
		if (!s->name) {
			kmem_cache_free(kmem_cache, s);
			err = -ENOMEM;
			goto out_locked;
		}

		err = __kmem_cache_create(s, flags);
		if (!err) {

			s->refcount = 1;
			list_add(&s->list, &slab_caches);

		} else {
			kfree(s->name);
			kmem_cache_free(kmem_cache, s);
		}
	} else
		err = -ENOMEM;

out_locked:
	mutex_unlock(&slab_mutex);
	put_online_cpus();

	if (err) {

		if (flags & SLAB_PANIC)
			panic("kmem_cache_create: Failed to create slab '%s'. Error %d\n",
				name, err);
		else {
			printk(KERN_WARNING "kmem_cache_create(%s) failed with error %d",
				name, err);
			dump_stack();
		}

		return NULL;
	}

	return s;
}
EXPORT_SYMBOL(kmem_cache_create);

void kmem_cache_destroy(struct kmem_cache *s)
{
	get_online_cpus();
	mutex_lock(&slab_mutex);
	s->refcount--;
	if (!s->refcount) {
		list_del(&s->list);

		if (!__kmem_cache_shutdown(s)) {
			mutex_unlock(&slab_mutex);
			if (s->flags & SLAB_DESTROY_BY_RCU)
				rcu_barrier();

			kfree(s->name);
			kmem_cache_free(kmem_cache, s);
		} else {
			list_add(&s->list, &slab_caches);
			mutex_unlock(&slab_mutex);
			printk(KERN_ERR "kmem_cache_destroy %s: Slab cache still has objects\n",
				s->name);
			dump_stack();
		}
	} else {
		mutex_unlock(&slab_mutex);
	}
	put_online_cpus();
}
EXPORT_SYMBOL(kmem_cache_destroy);

int slab_is_available(void)
{
	return slab_state >= UP;
}

#ifndef CONFIG_SLOB
/* Create a cache during boot when no slab services are available yet */
void __init create_boot_cache(struct kmem_cache *s, const char *name, size_t size,
		unsigned long flags)
{
	int err;

	s->name = name;
	s->size = s->object_size = size;
	s->align = ARCH_KMALLOC_MINALIGN;
	err = __kmem_cache_create(s, flags);

	if (err)
		panic("Creation of kmalloc slab %s size=%zd failed. Reason %d\n",
					name, size, err);

	s->refcount = -1;	/* Exempt from merging for now */
}

struct kmem_cache *__init create_kmalloc_cache(const char *name, size_t size,
				unsigned long flags)
{
	struct kmem_cache *s = kmem_cache_zalloc(kmem_cache, GFP_NOWAIT);

	if (!s)
		panic("Out of memory when creating slab %s\n", name);

	create_boot_cache(s, name, size, flags);
	list_add(&s->list, &slab_caches);
	s->refcount = 1;
	return s;
}

#endif /* !CONFIG_SLOB */


#ifdef CONFIG_SLABINFO
static void print_slabinfo_header(struct seq_file *m)
{
	/*
	 * Output format version, so at least we can change it
	 * without _too_ many complaints.
	 */
#ifdef CONFIG_DEBUG_SLAB
	seq_puts(m, "slabinfo - version: 2.1 (statistics)\n");
#else
	seq_puts(m, "slabinfo - version: 2.1\n");
#endif
	seq_puts(m, "# name            <active_objs> <num_objs> <objsize> "
		 "<objperslab> <pagesperslab>");
	seq_puts(m, " : tunables <limit> <batchcount> <sharedfactor>");
	seq_puts(m, " : slabdata <active_slabs> <num_slabs> <sharedavail>");
#ifdef CONFIG_DEBUG_SLAB
	seq_puts(m, " : globalstat <listallocs> <maxobjs> <grown> <reaped> "
		 "<error> <maxfreeable> <nodeallocs> <remotefrees> <alienoverflow>");
	seq_puts(m, " : cpustat <allochit> <allocmiss> <freehit> <freemiss>");
#endif
	seq_putc(m, '\n');
}

static void *s_start(struct seq_file *m, loff_t *pos)
{
	loff_t n = *pos;

	mutex_lock(&slab_mutex);
	if (!n)
		print_slabinfo_header(m);

	return seq_list_start(&slab_caches, *pos);
}

static void *s_next(struct seq_file *m, void *p, loff_t *pos)
{
	return seq_list_next(p, &slab_caches, pos);
}

static void s_stop(struct seq_file *m, void *p)
{
	mutex_unlock(&slab_mutex);
}

static int s_show(struct seq_file *m, void *p)
{
	struct kmem_cache *s = list_entry(p, struct kmem_cache, list);
	struct slabinfo sinfo;

	memset(&sinfo, 0, sizeof(sinfo));
	get_slabinfo(s, &sinfo);

	seq_printf(m, "%-17s %6lu %6lu %6u %4u %4d",
		   s->name, sinfo.active_objs, sinfo.num_objs, s->size,
		   sinfo.objects_per_slab, (1 << sinfo.cache_order));

	seq_printf(m, " : tunables %4u %4u %4u",
		   sinfo.limit, sinfo.batchcount, sinfo.shared);
	seq_printf(m, " : slabdata %6lu %6lu %6lu",
		   sinfo.active_slabs, sinfo.num_slabs, sinfo.shared_avail);
	slabinfo_show_stats(m, s);
	seq_putc(m, '\n');
	return 0;
}

/*
 * slabinfo_op - iterator that generates /proc/slabinfo
 *
 * Output layout:
 * cache-name
 * num-active-objs
 * total-objs
 * object size
 * num-active-slabs
 * total-slabs
 * num-pages-per-slab
 * + further values on SMP and with statistics enabled
 */
static const struct seq_operations slabinfo_op = {
	.start = s_start,
	.next = s_next,
	.stop = s_stop,
	.show = s_show,
};

static int slabinfo_open(struct inode *inode, struct file *file)
{
	return seq_open(file, &slabinfo_op);
}

static const struct file_operations proc_slabinfo_operations = {
	.open		= slabinfo_open,
	.read		= seq_read,
	.write          = slabinfo_write,
	.llseek		= seq_lseek,
	.release	= seq_release,
};

static int __init slab_proc_init(void)
{
	proc_create("slabinfo", S_IRUSR, NULL, &proc_slabinfo_operations);
	return 0;
}
module_init(slab_proc_init);
#endif /* CONFIG_SLABINFO */
Commit	Line	Data
039363f3 CL	1	/*
	2	* Slab allocator functions that are independent of the allocator strategy
	3	*
	4	* (C) 2012 Christoph Lameter <cl@linux.com>
	5	*/
	6	#include <linux/slab.h>
	7
	8	#include <linux/mm.h>
	9	#include <linux/poison.h>
	10	#include <linux/interrupt.h>
	11	#include <linux/memory.h>
	12	#include <linux/compiler.h>
	13	#include <linux/module.h>
20cea968 CL	14	#include <linux/cpu.h>
20cea968 CL	15	#include <linux/uaccess.h>
b7454ad3 GC	16	#include <linux/seq_file.h>
b7454ad3 GC	17	#include <linux/proc_fs.h>
039363f3 CL	18	#include <asm/cacheflush.h>
	19	#include <asm/tlbflush.h>
	20	#include <asm/page.h>
	21
97d06609 CL	22	#include "slab.h"
	23
	24	enum slab_state slab_state;
18004c5d CL	25	LIST_HEAD(slab_caches);
18004c5d CL	26	DEFINE_MUTEX(slab_mutex);
9b030cb8	27	struct kmem_cache *kmem_cache;
97d06609	28
77be4b13 SK	29	#ifdef CONFIG_DEBUG_VM
77be4b13 SK	30	static int kmem_cache_sanity_check(const char *name, size_t size)
039363f3 CL	31	{
	32	struct kmem_cache *s = NULL;
	33
039363f3 CL	34	if (!name \|\| in_interrupt() \|\| size < sizeof(void *) \|\|
039363f3 CL	35	size > KMALLOC_MAX_SIZE) {
77be4b13 SK	36	pr_err("kmem_cache_create(%s) integrity check failed\n", name);
77be4b13 SK	37	return -EINVAL;
039363f3	38	}
b920536a	39
20cea968 CL	40	list_for_each_entry(s, &slab_caches, list) {
	41	char tmp;
	42	int res;
	43
	44	/*
	45	* This happens when the module gets unloaded and doesn't
	46	* destroy its slab cache and no-one else reuses the vmalloc
	47	* area of the module. Print a warning.
	48	*/
	49	res = probe_kernel_address(s->name, tmp);
	50	if (res) {
77be4b13	51	pr_err("Slab cache with size %d has lost its name\n",
20cea968 CL	52	s->object_size);
	53	continue;
	54	}
	55
	56	if (!strcmp(s->name, name)) {
77be4b13 SK	57	pr_err("%s (%s): Cache name already exists.\n",
77be4b13 SK	58	__func__, name);
20cea968 CL	59	dump_stack();
20cea968 CL	60	s = NULL;
77be4b13	61	return -EINVAL;
20cea968 CL	62	}
	63	}
	64
	65	WARN_ON(strchr(name, ' ')); /* It confuses parsers */
77be4b13 SK	66	return 0;
	67	}
	68	#else
	69	static inline int kmem_cache_sanity_check(const char *name, size_t size)
	70	{
	71	return 0;
	72	}
20cea968 CL	73	#endif
20cea968 CL	74
77be4b13 SK	75	/*
	76	* kmem_cache_create - Create a cache.
	77	* @name: A string which is used in /proc/slabinfo to identify this cache.
	78	* @size: The size of objects to be created in this cache.
	79	* @align: The required alignment for the objects.
	80	* @flags: SLAB flags
	81	* @ctor: A constructor for the objects.
	82	*
	83	* Returns a ptr to the cache on success, NULL on failure.
	84	* Cannot be called within a interrupt, but can be interrupted.
	85	* The @ctor is run when new pages are allocated by the cache.
	86	*
	87	* The flags are
	88	*
	89	* %SLAB_POISON - Poison the slab with a known test pattern (a5a5a5a5)
	90	* to catch references to uninitialised memory.
	91	*
	92	* %SLAB_RED_ZONE - Insert `Red' zones around the allocated memory to check
	93	* for buffer overruns.
	94	*
	95	* %SLAB_HWCACHE_ALIGN - Align the objects in this cache to a hardware
	96	* cacheline. This can be beneficial if you're counting cycles as closely
	97	* as davem.
	98	*/
	99
	100	struct kmem_cache kmem_cache_create(const char name, size_t size, size_t align,
	101	unsigned long flags, void (ctor)(void ))
	102	{
	103	struct kmem_cache *s = NULL;
686d550d	104	int err = 0;
039363f3	105
77be4b13 SK	106	get_online_cpus();
77be4b13 SK	107	mutex_lock(&slab_mutex);
686d550d CL	108
	109	if (!kmem_cache_sanity_check(name, size) == 0)
	110	goto out_locked;
	111
d8843922 GC	112	/*
	113	* Some allocators will constraint the set of valid flags to a subset
	114	* of all flags. We expect them to define CACHE_CREATE_MASK in this
	115	* case, and we'll just provide them with a sanitized version of the
	116	* passed flags.
	117	*/
	118	flags &= CACHE_CREATE_MASK;
686d550d	119
cbb79694 CL	120	s = __kmem_cache_alias(name, size, align, flags, ctor);
	121	if (s)
	122	goto out_locked;
	123
278b1bb1	124	s = kmem_cache_zalloc(kmem_cache, GFP_KERNEL);
db265eca	125	if (s) {
8a13a4cc CL	126	s->object_size = s->size = size;
	127	s->align = align;
	128	s->ctor = ctor;
	129	s->name = kstrdup(name, GFP_KERNEL);
	130	if (!s->name) {
	131	kmem_cache_free(kmem_cache, s);
	132	err = -ENOMEM;
	133	goto out_locked;
	134	}
	135
	136	err = __kmem_cache_create(s, flags);
cce89f4f	137	if (!err) {
278b1bb1	138
cce89f4f	139	s->refcount = 1;
db265eca	140	list_add(&s->list, &slab_caches);
686d550d	141
cce89f4f	142	} else {
8a13a4cc	143	kfree(s->name);
278b1bb1 CL	144	kmem_cache_free(kmem_cache, s);
278b1bb1 CL	145	}
8a13a4cc	146	} else
278b1bb1	147	err = -ENOMEM;
7c9adf5a	148
686d550d	149	out_locked:
20cea968 CL	150	mutex_unlock(&slab_mutex);
	151	put_online_cpus();
	152
686d550d CL	153	if (err) {
	154
	155	if (flags & SLAB_PANIC)
	156	panic("kmem_cache_create: Failed to create slab '%s'. Error %d\n",
	157	name, err);
	158	else {
	159	printk(KERN_WARNING "kmem_cache_create(%s) failed with error %d",
	160	name, err);
	161	dump_stack();
	162	}
	163
	164	return NULL;
	165	}
039363f3 CL	166
	167	return s;
	168	}
	169	EXPORT_SYMBOL(kmem_cache_create);
97d06609	170
945cf2b6 CL	171	void kmem_cache_destroy(struct kmem_cache *s)
	172	{
	173	get_online_cpus();
	174	mutex_lock(&slab_mutex);
	175	s->refcount--;
	176	if (!s->refcount) {
	177	list_del(&s->list);
	178
	179	if (!__kmem_cache_shutdown(s)) {
210ed9de	180	mutex_unlock(&slab_mutex);
945cf2b6 CL	181	if (s->flags & SLAB_DESTROY_BY_RCU)
	182	rcu_barrier();
	183
db265eca	184	kfree(s->name);
8f4c765c	185	kmem_cache_free(kmem_cache, s);
945cf2b6 CL	186	} else {
945cf2b6 CL	187	list_add(&s->list, &slab_caches);
210ed9de	188	mutex_unlock(&slab_mutex);
945cf2b6 CL	189	printk(KERN_ERR "kmem_cache_destroy %s: Slab cache still has objects\n",
	190	s->name);
	191	dump_stack();
	192	}
210ed9de JK	193	} else {
210ed9de JK	194	mutex_unlock(&slab_mutex);
945cf2b6	195	}
945cf2b6 CL	196	put_online_cpus();
	197	}
	198	EXPORT_SYMBOL(kmem_cache_destroy);
	199
97d06609 CL	200	int slab_is_available(void)
	201	{
	202	return slab_state >= UP;
	203	}
b7454ad3	204
45530c44 CL	205	#ifndef CONFIG_SLOB
	206	/* Create a cache during boot when no slab services are available yet */
	207	void __init create_boot_cache(struct kmem_cache s, const char name, size_t size,
	208	unsigned long flags)
	209	{
	210	int err;
	211
	212	s->name = name;
	213	s->size = s->object_size = size;
	214	s->align = ARCH_KMALLOC_MINALIGN;
	215	err = __kmem_cache_create(s, flags);
	216
	217	if (err)
	218	panic("Creation of kmalloc slab %s size=%zd failed. Reason %d\n",
	219	name, size, err);
	220
	221	s->refcount = -1; /* Exempt from merging for now */
	222	}
	223
	224	struct kmem_cache __init create_kmalloc_cache(const char name, size_t size,
	225	unsigned long flags)
	226	{
	227	struct kmem_cache *s = kmem_cache_zalloc(kmem_cache, GFP_NOWAIT);
	228
	229	if (!s)
	230	panic("Out of memory when creating slab %s\n", name);
	231
	232	create_boot_cache(s, name, size, flags);
	233	list_add(&s->list, &slab_caches);
	234	s->refcount = 1;
	235	return s;
	236	}
	237
	238	#endif /* !CONFIG_SLOB */
	239
	240
b7454ad3	241	#ifdef CONFIG_SLABINFO
bcee6e2a GC	242	static void print_slabinfo_header(struct seq_file *m)
	243	{
	244	/*
	245	* Output format version, so at least we can change it
	246	* without _too_ many complaints.
	247	*/
	248	#ifdef CONFIG_DEBUG_SLAB
	249	seq_puts(m, "slabinfo - version: 2.1 (statistics)\n");
	250	#else
	251	seq_puts(m, "slabinfo - version: 2.1\n");
	252	#endif
	253	seq_puts(m, "# name <active_objs> <num_objs> <objsize> "
	254	"<objperslab> <pagesperslab>");
	255	seq_puts(m, " : tunables <limit> <batchcount> <sharedfactor>");
	256	seq_puts(m, " : slabdata <active_slabs> <num_slabs> <sharedavail>");
	257	#ifdef CONFIG_DEBUG_SLAB
	258	seq_puts(m, " : globalstat <listallocs> <maxobjs> <grown> <reaped> "
	259	"<error> <maxfreeable> <nodeallocs> <remotefrees> <alienoverflow>");
	260	seq_puts(m, " : cpustat <allochit> <allocmiss> <freehit> <freemiss>");
	261	#endif
	262	seq_putc(m, '\n');
	263	}
	264
b7454ad3 GC	265	static void s_start(struct seq_file m, loff_t *pos)
	266	{
	267	loff_t n = *pos;
	268
	269	mutex_lock(&slab_mutex);
	270	if (!n)
	271	print_slabinfo_header(m);
	272
	273	return seq_list_start(&slab_caches, *pos);
	274	}
	275
	276	static void s_next(struct seq_file m, void p, loff_t pos)
	277	{
	278	return seq_list_next(p, &slab_caches, pos);
	279	}
	280
	281	static void s_stop(struct seq_file m, void p)
	282	{
	283	mutex_unlock(&slab_mutex);
	284	}
	285
	286	static int s_show(struct seq_file m, void p)
	287	{
0d7561c6 GC	288	struct kmem_cache *s = list_entry(p, struct kmem_cache, list);
	289	struct slabinfo sinfo;
	290
	291	memset(&sinfo, 0, sizeof(sinfo));
	292	get_slabinfo(s, &sinfo);
	293
	294	seq_printf(m, "%-17s %6lu %6lu %6u %4u %4d",
	295	s->name, sinfo.active_objs, sinfo.num_objs, s->size,
	296	sinfo.objects_per_slab, (1 << sinfo.cache_order));
	297
	298	seq_printf(m, " : tunables %4u %4u %4u",
	299	sinfo.limit, sinfo.batchcount, sinfo.shared);
	300	seq_printf(m, " : slabdata %6lu %6lu %6lu",
	301	sinfo.active_slabs, sinfo.num_slabs, sinfo.shared_avail);
	302	slabinfo_show_stats(m, s);
	303	seq_putc(m, '\n');
	304	return 0;
b7454ad3 GC	305	}
	306
	307	/*
	308	* slabinfo_op - iterator that generates /proc/slabinfo
	309	*
	310	* Output layout:
	311	* cache-name
	312	* num-active-objs
	313	* total-objs
	314	* object size
	315	* num-active-slabs
	316	* total-slabs
	317	* num-pages-per-slab
	318	* + further values on SMP and with statistics enabled
	319	*/
	320	static const struct seq_operations slabinfo_op = {
	321	.start = s_start,
	322	.next = s_next,
	323	.stop = s_stop,
	324	.show = s_show,
	325	};
	326
	327	static int slabinfo_open(struct inode inode, struct file file)
	328	{
	329	return seq_open(file, &slabinfo_op);
	330	}
	331
	332	static const struct file_operations proc_slabinfo_operations = {
	333	.open = slabinfo_open,
	334	.read = seq_read,
	335	.write = slabinfo_write,
	336	.llseek = seq_lseek,
	337	.release = seq_release,
	338	};
	339
	340	static int __init slab_proc_init(void)
	341	{
	342	proc_create("slabinfo", S_IRUSR, NULL, &proc_slabinfo_operations);
	343	return 0;
	344	}
	345	module_init(slab_proc_init);
	346	#endif /* CONFIG_SLABINFO */