[net-next-2.6.git] / mm / backing-dev.c


#include <linux/wait.h>
#include <linux/backing-dev.h>
#include <linux/fs.h>
#include <linux/pagemap.h>
#include <linux/sched.h>
#include <linux/module.h>
#include <linux/writeback.h>
#include <linux/device.h>

void default_unplug_io_fn(struct backing_dev_info *bdi, struct page *page)
{
}
EXPORT_SYMBOL(default_unplug_io_fn);

struct backing_dev_info default_backing_dev_info = {
	.ra_pages	= VM_MAX_READAHEAD * 1024 / PAGE_CACHE_SIZE,
	.state		= 0,
	.capabilities	= BDI_CAP_MAP_COPY,
	.unplug_io_fn	= default_unplug_io_fn,
};
EXPORT_SYMBOL_GPL(default_backing_dev_info);

static struct class *bdi_class;

#ifdef CONFIG_DEBUG_FS
#include <linux/debugfs.h>
#include <linux/seq_file.h>

static struct dentry *bdi_debug_root;

static void bdi_debug_init(void)
{
	bdi_debug_root = debugfs_create_dir("bdi", NULL);
}

static int bdi_debug_stats_show(struct seq_file *m, void *v)
{
	struct backing_dev_info *bdi = m->private;
	unsigned long background_thresh;
	unsigned long dirty_thresh;
	unsigned long bdi_thresh;

	get_dirty_limits(&background_thresh, &dirty_thresh, &bdi_thresh, bdi);

#define K(x) ((x) << (PAGE_SHIFT - 10))
	seq_printf(m,
		   "BdiWriteback:     %8lu kB\n"
		   "BdiReclaimable:   %8lu kB\n"
		   "BdiDirtyThresh:   %8lu kB\n"
		   "DirtyThresh:      %8lu kB\n"
		   "BackgroundThresh: %8lu kB\n",
		   (unsigned long) K(bdi_stat(bdi, BDI_WRITEBACK)),
		   (unsigned long) K(bdi_stat(bdi, BDI_RECLAIMABLE)),
		   K(bdi_thresh),
		   K(dirty_thresh),
		   K(background_thresh));
#undef K

	return 0;
}

static int bdi_debug_stats_open(struct inode *inode, struct file *file)
{
	return single_open(file, bdi_debug_stats_show, inode->i_private);
}

static const struct file_operations bdi_debug_stats_fops = {
	.open		= bdi_debug_stats_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= single_release,
};

static void bdi_debug_register(struct backing_dev_info *bdi, const char *name)
{
	bdi->debug_dir = debugfs_create_dir(name, bdi_debug_root);
	bdi->debug_stats = debugfs_create_file("stats", 0444, bdi->debug_dir,
					       bdi, &bdi_debug_stats_fops);
}

static void bdi_debug_unregister(struct backing_dev_info *bdi)
{
	debugfs_remove(bdi->debug_stats);
	debugfs_remove(bdi->debug_dir);
}
#else
static inline void bdi_debug_init(void)
{
}
static inline void bdi_debug_register(struct backing_dev_info *bdi,
				      const char *name)
{
}
static inline void bdi_debug_unregister(struct backing_dev_info *bdi)
{
}
#endif

static ssize_t read_ahead_kb_store(struct device *dev,
				  struct device_attribute *attr,
				  const char *buf, size_t count)
{
	struct backing_dev_info *bdi = dev_get_drvdata(dev);
	char *end;
	unsigned long read_ahead_kb;
	ssize_t ret = -EINVAL;

	read_ahead_kb = simple_strtoul(buf, &end, 10);
	if (*buf && (end[0] == '\0' || (end[0] == '\n' && end[1] == '\0'))) {
		bdi->ra_pages = read_ahead_kb >> (PAGE_SHIFT - 10);
		ret = count;
	}
	return ret;
}

#define K(pages) ((pages) << (PAGE_SHIFT - 10))

#define BDI_SHOW(name, expr)						\
static ssize_t name##_show(struct device *dev,				\
			   struct device_attribute *attr, char *page)	\
{									\
	struct backing_dev_info *bdi = dev_get_drvdata(dev);		\
									\
	return snprintf(page, PAGE_SIZE-1, "%lld\n", (long long)expr);	\
}

BDI_SHOW(read_ahead_kb, K(bdi->ra_pages))

static ssize_t min_ratio_store(struct device *dev,
		struct device_attribute *attr, const char *buf, size_t count)
{
	struct backing_dev_info *bdi = dev_get_drvdata(dev);
	char *end;
	unsigned int ratio;
	ssize_t ret = -EINVAL;

	ratio = simple_strtoul(buf, &end, 10);
	if (*buf && (end[0] == '\0' || (end[0] == '\n' && end[1] == '\0'))) {
		ret = bdi_set_min_ratio(bdi, ratio);
		if (!ret)
			ret = count;
	}
	return ret;
}
BDI_SHOW(min_ratio, bdi->min_ratio)

static ssize_t max_ratio_store(struct device *dev,
		struct device_attribute *attr, const char *buf, size_t count)
{
	struct backing_dev_info *bdi = dev_get_drvdata(dev);
	char *end;
	unsigned int ratio;
	ssize_t ret = -EINVAL;

	ratio = simple_strtoul(buf, &end, 10);
	if (*buf && (end[0] == '\0' || (end[0] == '\n' && end[1] == '\0'))) {
		ret = bdi_set_max_ratio(bdi, ratio);
		if (!ret)
			ret = count;
	}
	return ret;
}
BDI_SHOW(max_ratio, bdi->max_ratio)

#define __ATTR_RW(attr) __ATTR(attr, 0644, attr##_show, attr##_store)

static struct device_attribute bdi_dev_attrs[] = {
	__ATTR_RW(read_ahead_kb),
	__ATTR_RW(min_ratio),
	__ATTR_RW(max_ratio),
	__ATTR_NULL,
};

static __init int bdi_class_init(void)
{
	bdi_class = class_create(THIS_MODULE, "bdi");
	bdi_class->dev_attrs = bdi_dev_attrs;
	bdi_debug_init();
	return 0;
}
postcore_initcall(bdi_class_init);

static int __init default_bdi_init(void)
{
	int err;

	err = bdi_init(&default_backing_dev_info);
	if (!err)
		bdi_register(&default_backing_dev_info, NULL, "default");

	return err;
}
subsys_initcall(default_bdi_init);

int bdi_register(struct backing_dev_info *bdi, struct device *parent,
		const char *fmt, ...)
{
	va_list args;
	int ret = 0;
	struct device *dev;

	if (bdi->dev)	/* The driver needs to use separate queues per device */
		goto exit;

	va_start(args, fmt);
	dev = device_create_vargs(bdi_class, parent, MKDEV(0, 0), bdi, fmt, args);
	va_end(args);
	if (IS_ERR(dev)) {
		ret = PTR_ERR(dev);
		goto exit;
	}

	bdi->dev = dev;
	bdi_debug_register(bdi, dev_name(dev));

exit:
	return ret;
}
EXPORT_SYMBOL(bdi_register);

int bdi_register_dev(struct backing_dev_info *bdi, dev_t dev)
{
	return bdi_register(bdi, NULL, "%u:%u", MAJOR(dev), MINOR(dev));
}
EXPORT_SYMBOL(bdi_register_dev);

void bdi_unregister(struct backing_dev_info *bdi)
{
	if (bdi->dev) {
		bdi_debug_unregister(bdi);
		device_unregister(bdi->dev);
		bdi->dev = NULL;
	}
}
EXPORT_SYMBOL(bdi_unregister);

int bdi_init(struct backing_dev_info *bdi)
{
	int i;
	int err;

	bdi->dev = NULL;

	bdi->min_ratio = 0;
	bdi->max_ratio = 100;
	bdi->max_prop_frac = PROP_FRAC_BASE;

	for (i = 0; i < NR_BDI_STAT_ITEMS; i++) {
		err = percpu_counter_init(&bdi->bdi_stat[i], 0);
		if (err)
			goto err;
	}

	bdi->dirty_exceeded = 0;
	err = prop_local_init_percpu(&bdi->completions);

	if (err) {
err:
		while (i--)
			percpu_counter_destroy(&bdi->bdi_stat[i]);
	}

	return err;
}
EXPORT_SYMBOL(bdi_init);

void bdi_destroy(struct backing_dev_info *bdi)
{
	int i;

	bdi_unregister(bdi);

	for (i = 0; i < NR_BDI_STAT_ITEMS; i++)
		percpu_counter_destroy(&bdi->bdi_stat[i]);

	prop_local_destroy_percpu(&bdi->completions);
}
EXPORT_SYMBOL(bdi_destroy);

static wait_queue_head_t congestion_wqh[2] = {
		__WAIT_QUEUE_HEAD_INITIALIZER(congestion_wqh[0]),
		__WAIT_QUEUE_HEAD_INITIALIZER(congestion_wqh[1])
	};


void clear_bdi_congested(struct backing_dev_info *bdi, int sync)
{
	enum bdi_state bit;
	wait_queue_head_t *wqh = &congestion_wqh[sync];

	bit = sync ? BDI_sync_congested : BDI_async_congested;
	clear_bit(bit, &bdi->state);
	smp_mb__after_clear_bit();
	if (waitqueue_active(wqh))
		wake_up(wqh);
}
EXPORT_SYMBOL(clear_bdi_congested);

void set_bdi_congested(struct backing_dev_info *bdi, int sync)
{
	enum bdi_state bit;

	bit = sync ? BDI_sync_congested : BDI_async_congested;
	set_bit(bit, &bdi->state);
}
EXPORT_SYMBOL(set_bdi_congested);

/**
 * congestion_wait - wait for a backing_dev to become uncongested
 * @rw: READ or WRITE
 * @timeout: timeout in jiffies
 *
 * Waits for up to @timeout jiffies for a backing_dev (any backing_dev) to exit
 * write congestion.  If no backing_devs are congested then just wait for the
 * next write to be completed.
 */
long congestion_wait(int rw, long timeout)
{
	long ret;
	DEFINE_WAIT(wait);
	wait_queue_head_t *wqh = &congestion_wqh[rw];

	prepare_to_wait(wqh, &wait, TASK_UNINTERRUPTIBLE);
	ret = io_schedule_timeout(timeout);
	finish_wait(wqh, &wait);
	return ret;
}
EXPORT_SYMBOL(congestion_wait);
Commit	Line	Data
3fcfab16 AM	1
	2	#include <linux/wait.h>
	3	#include <linux/backing-dev.h>
	4	#include <linux/fs.h>
26160158	5	#include <linux/pagemap.h>
3fcfab16 AM	6	#include <linux/sched.h>
3fcfab16 AM	7	#include <linux/module.h>
cf0ca9fe PZ	8	#include <linux/writeback.h>
	9	#include <linux/device.h>
	10
26160158 JA	11	void default_unplug_io_fn(struct backing_dev_info bdi, struct page page)
	12	{
	13	}
	14	EXPORT_SYMBOL(default_unplug_io_fn);
	15
	16	struct backing_dev_info default_backing_dev_info = {
	17	.ra_pages = VM_MAX_READAHEAD * 1024 / PAGE_CACHE_SIZE,
	18	.state = 0,
	19	.capabilities = BDI_CAP_MAP_COPY,
	20	.unplug_io_fn = default_unplug_io_fn,
	21	};
	22	EXPORT_SYMBOL_GPL(default_backing_dev_info);
cf0ca9fe PZ	23
	24	static struct class *bdi_class;
	25
76f1418b MS	26	#ifdef CONFIG_DEBUG_FS
	27	#include <linux/debugfs.h>
	28	#include <linux/seq_file.h>
	29
	30	static struct dentry *bdi_debug_root;
	31
	32	static void bdi_debug_init(void)
	33	{
	34	bdi_debug_root = debugfs_create_dir("bdi", NULL);
	35	}
	36
	37	static int bdi_debug_stats_show(struct seq_file m, void v)
	38	{
	39	struct backing_dev_info *bdi = m->private;
364aeb28 DR	40	unsigned long background_thresh;
	41	unsigned long dirty_thresh;
	42	unsigned long bdi_thresh;
76f1418b MS	43
	44	get_dirty_limits(&background_thresh, &dirty_thresh, &bdi_thresh, bdi);
	45
	46	#define K(x) ((x) << (PAGE_SHIFT - 10))
	47	seq_printf(m,
	48	"BdiWriteback: %8lu kB\n"
	49	"BdiReclaimable: %8lu kB\n"
	50	"BdiDirtyThresh: %8lu kB\n"
	51	"DirtyThresh: %8lu kB\n"
	52	"BackgroundThresh: %8lu kB\n",
	53	(unsigned long) K(bdi_stat(bdi, BDI_WRITEBACK)),
	54	(unsigned long) K(bdi_stat(bdi, BDI_RECLAIMABLE)),
	55	K(bdi_thresh),
	56	K(dirty_thresh),
	57	K(background_thresh));
	58	#undef K
	59
	60	return 0;
	61	}
	62
	63	static int bdi_debug_stats_open(struct inode inode, struct file file)
	64	{
	65	return single_open(file, bdi_debug_stats_show, inode->i_private);
	66	}
	67
	68	static const struct file_operations bdi_debug_stats_fops = {
	69	.open = bdi_debug_stats_open,
	70	.read = seq_read,
	71	.llseek = seq_lseek,
	72	.release = single_release,
	73	};
	74
	75	static void bdi_debug_register(struct backing_dev_info bdi, const char name)
	76	{
	77	bdi->debug_dir = debugfs_create_dir(name, bdi_debug_root);
	78	bdi->debug_stats = debugfs_create_file("stats", 0444, bdi->debug_dir,
	79	bdi, &bdi_debug_stats_fops);
	80	}
	81
	82	static void bdi_debug_unregister(struct backing_dev_info *bdi)
	83	{
	84	debugfs_remove(bdi->debug_stats);
	85	debugfs_remove(bdi->debug_dir);
	86	}
	87	#else
	88	static inline void bdi_debug_init(void)
	89	{
	90	}
	91	static inline void bdi_debug_register(struct backing_dev_info *bdi,
	92	const char *name)
	93	{
	94	}
	95	static inline void bdi_debug_unregister(struct backing_dev_info *bdi)
	96	{
	97	}
	98	#endif
	99
cf0ca9fe PZ	100	static ssize_t read_ahead_kb_store(struct device *dev,
	101	struct device_attribute *attr,
	102	const char *buf, size_t count)
	103	{
	104	struct backing_dev_info *bdi = dev_get_drvdata(dev);
	105	char *end;
	106	unsigned long read_ahead_kb;
	107	ssize_t ret = -EINVAL;
	108
	109	read_ahead_kb = simple_strtoul(buf, &end, 10);
	110	if (*buf && (end[0] == '\0' \|\| (end[0] == '\n' && end[1] == '\0'))) {
	111	bdi->ra_pages = read_ahead_kb >> (PAGE_SHIFT - 10);
	112	ret = count;
	113	}
	114	return ret;
	115	}
	116
	117	#define K(pages) ((pages) << (PAGE_SHIFT - 10))
	118
	119	#define BDI_SHOW(name, expr) \
	120	static ssize_t name##_show(struct device *dev, \
	121	struct device_attribute attr, char page) \
	122	{ \
	123	struct backing_dev_info *bdi = dev_get_drvdata(dev); \
	124	\
	125	return snprintf(page, PAGE_SIZE-1, "%lld\n", (long long)expr); \
	126	}
	127
	128	BDI_SHOW(read_ahead_kb, K(bdi->ra_pages))
	129
189d3c4a PZ	130	static ssize_t min_ratio_store(struct device *dev,
	131	struct device_attribute attr, const char buf, size_t count)
	132	{
	133	struct backing_dev_info *bdi = dev_get_drvdata(dev);
	134	char *end;
	135	unsigned int ratio;
	136	ssize_t ret = -EINVAL;
	137
	138	ratio = simple_strtoul(buf, &end, 10);
	139	if (*buf && (end[0] == '\0' \|\| (end[0] == '\n' && end[1] == '\0'))) {
	140	ret = bdi_set_min_ratio(bdi, ratio);
	141	if (!ret)
	142	ret = count;
	143	}
	144	return ret;
	145	}
	146	BDI_SHOW(min_ratio, bdi->min_ratio)
	147
a42dde04 PZ	148	static ssize_t max_ratio_store(struct device *dev,
	149	struct device_attribute attr, const char buf, size_t count)
	150	{
	151	struct backing_dev_info *bdi = dev_get_drvdata(dev);
	152	char *end;
	153	unsigned int ratio;
	154	ssize_t ret = -EINVAL;
	155
	156	ratio = simple_strtoul(buf, &end, 10);
	157	if (*buf && (end[0] == '\0' \|\| (end[0] == '\n' && end[1] == '\0'))) {
	158	ret = bdi_set_max_ratio(bdi, ratio);
	159	if (!ret)
	160	ret = count;
	161	}
	162	return ret;
	163	}
	164	BDI_SHOW(max_ratio, bdi->max_ratio)
	165
cf0ca9fe PZ	166	#define __ATTR_RW(attr) __ATTR(attr, 0644, attr##_show, attr##_store)
	167
	168	static struct device_attribute bdi_dev_attrs[] = {
	169	__ATTR_RW(read_ahead_kb),
189d3c4a	170	__ATTR_RW(min_ratio),
a42dde04	171	__ATTR_RW(max_ratio),
cf0ca9fe PZ	172	__ATTR_NULL,
	173	};
	174
	175	static __init int bdi_class_init(void)
	176	{
	177	bdi_class = class_create(THIS_MODULE, "bdi");
	178	bdi_class->dev_attrs = bdi_dev_attrs;
76f1418b	179	bdi_debug_init();
cf0ca9fe PZ	180	return 0;
cf0ca9fe PZ	181	}
76f1418b	182	postcore_initcall(bdi_class_init);
cf0ca9fe	183
26160158 JA	184	static int __init default_bdi_init(void)
	185	{
	186	int err;
	187
	188	err = bdi_init(&default_backing_dev_info);
	189	if (!err)
	190	bdi_register(&default_backing_dev_info, NULL, "default");
	191
	192	return err;
	193	}
	194	subsys_initcall(default_bdi_init);
	195
cf0ca9fe PZ	196	int bdi_register(struct backing_dev_info bdi, struct device parent,
	197	const char *fmt, ...)
	198	{
cf0ca9fe PZ	199	va_list args;
	200	int ret = 0;
	201	struct device *dev;
	202
69fc208b	203	if (bdi->dev) /* The driver needs to use separate queues per device */
f1d0b063 KS	204	goto exit;
f1d0b063 KS	205
cf0ca9fe	206	va_start(args, fmt);
19051c50	207	dev = device_create_vargs(bdi_class, parent, MKDEV(0, 0), bdi, fmt, args);
cf0ca9fe	208	va_end(args);
cf0ca9fe PZ	209	if (IS_ERR(dev)) {
	210	ret = PTR_ERR(dev);
	211	goto exit;
	212	}
	213
	214	bdi->dev = dev;
19051c50	215	bdi_debug_register(bdi, dev_name(dev));
cf0ca9fe PZ	216
cf0ca9fe PZ	217	exit:
cf0ca9fe PZ	218	return ret;
	219	}
	220	EXPORT_SYMBOL(bdi_register);
	221
	222	int bdi_register_dev(struct backing_dev_info *bdi, dev_t dev)
	223	{
	224	return bdi_register(bdi, NULL, "%u:%u", MAJOR(dev), MINOR(dev));
	225	}
	226	EXPORT_SYMBOL(bdi_register_dev);
	227
	228	void bdi_unregister(struct backing_dev_info *bdi)
	229	{
	230	if (bdi->dev) {
76f1418b	231	bdi_debug_unregister(bdi);
cf0ca9fe PZ	232	device_unregister(bdi->dev);
	233	bdi->dev = NULL;
	234	}
	235	}
	236	EXPORT_SYMBOL(bdi_unregister);
3fcfab16	237
b2e8fb6e PZ	238	int bdi_init(struct backing_dev_info *bdi)
b2e8fb6e PZ	239	{
4b01a0b1	240	int i;
b2e8fb6e PZ	241	int err;
b2e8fb6e PZ	242
cf0ca9fe PZ	243	bdi->dev = NULL;
cf0ca9fe PZ	244
189d3c4a	245	bdi->min_ratio = 0;
a42dde04 PZ	246	bdi->max_ratio = 100;
a42dde04 PZ	247	bdi->max_prop_frac = PROP_FRAC_BASE;
189d3c4a	248
b2e8fb6e	249	for (i = 0; i < NR_BDI_STAT_ITEMS; i++) {
ea319518	250	err = percpu_counter_init(&bdi->bdi_stat[i], 0);
04fbfdc1 PZ	251	if (err)
	252	goto err;
	253	}
	254
	255	bdi->dirty_exceeded = 0;
	256	err = prop_local_init_percpu(&bdi->completions);
	257
	258	if (err) {
	259	err:
4b01a0b1	260	while (i--)
04fbfdc1	261	percpu_counter_destroy(&bdi->bdi_stat[i]);
b2e8fb6e PZ	262	}
	263
	264	return err;
	265	}
	266	EXPORT_SYMBOL(bdi_init);
	267
	268	void bdi_destroy(struct backing_dev_info *bdi)
	269	{
	270	int i;
	271
cf0ca9fe PZ	272	bdi_unregister(bdi);
cf0ca9fe PZ	273
b2e8fb6e PZ	274	for (i = 0; i < NR_BDI_STAT_ITEMS; i++)
b2e8fb6e PZ	275	percpu_counter_destroy(&bdi->bdi_stat[i]);
04fbfdc1 PZ	276
04fbfdc1 PZ	277	prop_local_destroy_percpu(&bdi->completions);
b2e8fb6e PZ	278	}
	279	EXPORT_SYMBOL(bdi_destroy);
	280
3fcfab16 AM	281	static wait_queue_head_t congestion_wqh[2] = {
	282	__WAIT_QUEUE_HEAD_INITIALIZER(congestion_wqh[0]),
	283	__WAIT_QUEUE_HEAD_INITIALIZER(congestion_wqh[1])
	284	};
	285
	286
1faa16d2	287	void clear_bdi_congested(struct backing_dev_info *bdi, int sync)
3fcfab16 AM	288	{
3fcfab16 AM	289	enum bdi_state bit;
1faa16d2	290	wait_queue_head_t *wqh = &congestion_wqh[sync];
3fcfab16	291
1faa16d2	292	bit = sync ? BDI_sync_congested : BDI_async_congested;
3fcfab16 AM	293	clear_bit(bit, &bdi->state);
	294	smp_mb__after_clear_bit();
	295	if (waitqueue_active(wqh))
	296	wake_up(wqh);
	297	}
	298	EXPORT_SYMBOL(clear_bdi_congested);
	299
1faa16d2	300	void set_bdi_congested(struct backing_dev_info *bdi, int sync)
3fcfab16 AM	301	{
	302	enum bdi_state bit;
	303
1faa16d2	304	bit = sync ? BDI_sync_congested : BDI_async_congested;
3fcfab16 AM	305	set_bit(bit, &bdi->state);
	306	}
	307	EXPORT_SYMBOL(set_bdi_congested);
	308
	309	/**
	310	* congestion_wait - wait for a backing_dev to become uncongested
	311	* @rw: READ or WRITE
	312	* @timeout: timeout in jiffies
	313	*
	314	* Waits for up to @timeout jiffies for a backing_dev (any backing_dev) to exit
	315	* write congestion. If no backing_devs are congested then just wait for the
	316	* next write to be completed.
	317	*/
	318	long congestion_wait(int rw, long timeout)
	319	{
	320	long ret;
	321	DEFINE_WAIT(wait);
	322	wait_queue_head_t *wqh = &congestion_wqh[rw];
	323
	324	prepare_to_wait(wqh, &wait, TASK_UNINTERRUPTIBLE);
	325	ret = io_schedule_timeout(timeout);
	326	finish_wait(wqh, &wait);
	327	return ret;
	328	}
	329	EXPORT_SYMBOL(congestion_wait);
04fbfdc1	330