[linux-2.6-block.git] / fs / orangefs / orangefs-bufmap.c

/*
 * (C) 2001 Clemson University and The University of Chicago
 *
 * See COPYING in top-level directory.
 */
#include "protocol.h"
#include "orangefs-kernel.h"
#include "orangefs-bufmap.h"

struct slot_map {
	int c;
	wait_queue_head_t q;
	int count;
	unsigned long *map;
};

static struct slot_map rw_map = {
	.c = -1,
	.q = __WAIT_QUEUE_HEAD_INITIALIZER(rw_map.q)
};
static struct slot_map readdir_map = {
	.c = -1,
	.q = __WAIT_QUEUE_HEAD_INITIALIZER(readdir_map.q)
};


static void install(struct slot_map *m, int count, unsigned long *map)
{
	spin_lock(&m->q.lock);
	m->c = m->count = count;
	m->map = map;
	wake_up_all_locked(&m->q);
	spin_unlock(&m->q.lock);
}

static void mark_killed(struct slot_map *m)
{
	spin_lock(&m->q.lock);
	m->c -= m->count + 1;
	spin_unlock(&m->q.lock);
}

static void run_down(struct slot_map *m)
{
	DEFINE_WAIT(wait);
	spin_lock(&m->q.lock);
	if (m->c != -1) {
		for (;;) {
			if (likely(list_empty(&wait.task_list)))
				__add_wait_queue_tail(&m->q, &wait);
			set_current_state(TASK_UNINTERRUPTIBLE);

			if (m->c == -1)
				break;

			spin_unlock(&m->q.lock);
			schedule();
			spin_lock(&m->q.lock);
		}
		__remove_wait_queue(&m->q, &wait);
		__set_current_state(TASK_RUNNING);
	}
	m->map = NULL;
	spin_unlock(&m->q.lock);
}

static void put(struct slot_map *m, int slot)
{
	int v;
	spin_lock(&m->q.lock);
	__clear_bit(slot, m->map);
	v = ++m->c;
	if (unlikely(v == 1))	/* no free slots -> one free slot */
		wake_up_locked(&m->q);
	else if (unlikely(v == -1))	/* finished dying */
		wake_up_all_locked(&m->q);
	spin_unlock(&m->q.lock);
}

static int wait_for_free(struct slot_map *m)
{
	long left = slot_timeout_secs * HZ;
	DEFINE_WAIT(wait);

	do {
		long n = left, t;
		if (likely(list_empty(&wait.task_list)))
			__add_wait_queue_tail_exclusive(&m->q, &wait);
		set_current_state(TASK_INTERRUPTIBLE);

		if (m->c > 0)
			break;

		if (m->c < 0) {
			/* we are waiting for map to be installed */
			/* it would better be there soon, or we go away */
			if (n > ORANGEFS_BUFMAP_WAIT_TIMEOUT_SECS * HZ)
				n = ORANGEFS_BUFMAP_WAIT_TIMEOUT_SECS * HZ;
		}
		spin_unlock(&m->q.lock);
		t = schedule_timeout(n);
		spin_lock(&m->q.lock);
		if (unlikely(!t) && n != left && m->c < 0)
			left = t;
		else
			left = t + (left - n);
		if (unlikely(signal_pending(current)))
			left = -EINTR;
	} while (left > 0);

	if (!list_empty(&wait.task_list))
		list_del(&wait.task_list);
	else if (left <= 0 && waitqueue_active(&m->q))
		__wake_up_locked_key(&m->q, TASK_INTERRUPTIBLE, NULL);
	__set_current_state(TASK_RUNNING);

	if (likely(left > 0))
		return 0;

	return left < 0 ? -EINTR : -ETIMEDOUT;
}

static int get(struct slot_map *m)
{
	int res = 0;
	spin_lock(&m->q.lock);
	if (unlikely(m->c <= 0))
		res = wait_for_free(m);
	if (likely(!res)) {
		m->c--;
		res = find_first_zero_bit(m->map, m->count);
		__set_bit(res, m->map);
	}
	spin_unlock(&m->q.lock);
	return res;
}

/* used to describe mapped buffers */
struct orangefs_bufmap_desc {
	void *uaddr;			/* user space address pointer */
	struct page **page_array;	/* array of mapped pages */
	int array_count;		/* size of above arrays */
	struct list_head list_link;
};

static struct orangefs_bufmap {
	int desc_size;
	int desc_shift;
	int desc_count;
	int total_size;
	int page_count;

	struct page **page_array;
	struct orangefs_bufmap_desc *desc_array;

	/* array to track usage of buffer descriptors */
	unsigned long *buffer_index_array;

	/* array to track usage of buffer descriptors for readdir */
#define N DIV_ROUND_UP(ORANGEFS_READDIR_DEFAULT_DESC_COUNT, BITS_PER_LONG)
	unsigned long readdir_index_array[N];
#undef N
} *__orangefs_bufmap;

static DEFINE_SPINLOCK(orangefs_bufmap_lock);

static void
orangefs_bufmap_unmap(struct orangefs_bufmap *bufmap)
{
	int i;

	for (i = 0; i < bufmap->page_count; i++)
		put_page(bufmap->page_array[i]);
}

static void
orangefs_bufmap_free(struct orangefs_bufmap *bufmap)
{
	kfree(bufmap->page_array);
	kfree(bufmap->desc_array);
	kfree(bufmap->buffer_index_array);
	kfree(bufmap);
}

/*
 * XXX: Can the size and shift change while the caller gives up the 
 * XXX: lock between calling this and doing something useful?
 */

int orangefs_bufmap_size_query(void)
{
	struct orangefs_bufmap *bufmap;
	int size = 0;
	spin_lock(&orangefs_bufmap_lock);
	bufmap = __orangefs_bufmap;
	if (bufmap)
		size = bufmap->desc_size;
	spin_unlock(&orangefs_bufmap_lock);
	return size;
}

int orangefs_bufmap_shift_query(void)
{
	struct orangefs_bufmap *bufmap;
	int shift = 0;
	spin_lock(&orangefs_bufmap_lock);
	bufmap = __orangefs_bufmap;
	if (bufmap)
		shift = bufmap->desc_shift;
	spin_unlock(&orangefs_bufmap_lock);
	return shift;
}

static DECLARE_WAIT_QUEUE_HEAD(bufmap_waitq);
static DECLARE_WAIT_QUEUE_HEAD(readdir_waitq);

/*
 * orangefs_get_bufmap_init
 *
 * If bufmap_init is 1, then the shared memory system, including the
 * buffer_index_array, is available.  Otherwise, it is not.
 *
 * returns the value of bufmap_init
 */
int orangefs_get_bufmap_init(void)
{
	return __orangefs_bufmap ? 1 : 0;
}


static struct orangefs_bufmap *
orangefs_bufmap_alloc(struct ORANGEFS_dev_map_desc *user_desc)
{
	struct orangefs_bufmap *bufmap;

	bufmap = kzalloc(sizeof(*bufmap), GFP_KERNEL);
	if (!bufmap)
		goto out;

	bufmap->total_size = user_desc->total_size;
	bufmap->desc_count = user_desc->count;
	bufmap->desc_size = user_desc->size;
	bufmap->desc_shift = ilog2(bufmap->desc_size);

	bufmap->buffer_index_array =
		kzalloc(DIV_ROUND_UP(bufmap->desc_count, BITS_PER_LONG), GFP_KERNEL);
	if (!bufmap->buffer_index_array) {
		gossip_err("orangefs: could not allocate %d buffer indices\n",
				bufmap->desc_count);
		goto out_free_bufmap;
	}

	bufmap->desc_array =
		kcalloc(bufmap->desc_count, sizeof(struct orangefs_bufmap_desc),
			GFP_KERNEL);
	if (!bufmap->desc_array) {
		gossip_err("orangefs: could not allocate %d descriptors\n",
				bufmap->desc_count);
		goto out_free_index_array;
	}

	bufmap->page_count = bufmap->total_size / PAGE_SIZE;

	/* allocate storage to track our page mappings */
	bufmap->page_array =
		kcalloc(bufmap->page_count, sizeof(struct page *), GFP_KERNEL);
	if (!bufmap->page_array)
		goto out_free_desc_array;

	return bufmap;

out_free_desc_array:
	kfree(bufmap->desc_array);
out_free_index_array:
	kfree(bufmap->buffer_index_array);
out_free_bufmap:
	kfree(bufmap);
out:
	return NULL;
}

static int
orangefs_bufmap_map(struct orangefs_bufmap *bufmap,
		struct ORANGEFS_dev_map_desc *user_desc)
{
	int pages_per_desc = bufmap->desc_size / PAGE_SIZE;
	int offset = 0, ret, i;

	/* map the pages */
	ret = get_user_pages_fast((unsigned long)user_desc->ptr,
			     bufmap->page_count, 1, bufmap->page_array);

	if (ret < 0)
		return ret;

	if (ret != bufmap->page_count) {
		gossip_err("orangefs error: asked for %d pages, only got %d.\n",
				bufmap->page_count, ret);

		for (i = 0; i < ret; i++) {
			SetPageError(bufmap->page_array[i]);
			put_page(bufmap->page_array[i]);
		}
		return -ENOMEM;
	}

	/*
	 * ideally we want to get kernel space pointers for each page, but
	 * we can't kmap that many pages at once if highmem is being used.
	 * so instead, we just kmap/kunmap the page address each time the
	 * kaddr is needed.
	 */
	for (i = 0; i < bufmap->page_count; i++)
		flush_dcache_page(bufmap->page_array[i]);

	/* build a list of available descriptors */
	for (offset = 0, i = 0; i < bufmap->desc_count; i++) {
		bufmap->desc_array[i].page_array = &bufmap->page_array[offset];
		bufmap->desc_array[i].array_count = pages_per_desc;
		bufmap->desc_array[i].uaddr =
		    (user_desc->ptr + (i * pages_per_desc * PAGE_SIZE));
		offset += pages_per_desc;
	}

	return 0;
}

/*
 * orangefs_bufmap_initialize()
 *
 * initializes the mapped buffer interface
 *
 * returns 0 on success, -errno on failure
 */
int orangefs_bufmap_initialize(struct ORANGEFS_dev_map_desc *user_desc)
{
	struct orangefs_bufmap *bufmap;
	int ret = -EINVAL;

	gossip_debug(GOSSIP_BUFMAP_DEBUG,
		     "orangefs_bufmap_initialize: called (ptr ("
		     "%p) sz (%d) cnt(%d).\n",
		     user_desc->ptr,
		     user_desc->size,
		     user_desc->count);

	/*
	 * sanity check alignment and size of buffer that caller wants to
	 * work with
	 */
	if (PAGE_ALIGN((unsigned long)user_desc->ptr) !=
	    (unsigned long)user_desc->ptr) {
		gossip_err("orangefs error: memory alignment (front). %p\n",
			   user_desc->ptr);
		goto out;
	}

	if (PAGE_ALIGN(((unsigned long)user_desc->ptr + user_desc->total_size))
	    != (unsigned long)(user_desc->ptr + user_desc->total_size)) {
		gossip_err("orangefs error: memory alignment (back).(%p + %d)\n",
			   user_desc->ptr,
			   user_desc->total_size);
		goto out;
	}

	if (user_desc->total_size != (user_desc->size * user_desc->count)) {
		gossip_err("orangefs error: user provided an oddly sized buffer: (%d, %d, %d)\n",
			   user_desc->total_size,
			   user_desc->size,
			   user_desc->count);
		goto out;
	}

	if ((user_desc->size % PAGE_SIZE) != 0) {
		gossip_err("orangefs error: bufmap size not page size divisible (%d).\n",
			   user_desc->size);
		goto out;
	}

	ret = -ENOMEM;
	bufmap = orangefs_bufmap_alloc(user_desc);
	if (!bufmap)
		goto out;

	ret = orangefs_bufmap_map(bufmap, user_desc);
	if (ret)
		goto out_free_bufmap;


	spin_lock(&orangefs_bufmap_lock);
	if (__orangefs_bufmap) {
		spin_unlock(&orangefs_bufmap_lock);
		gossip_err("orangefs: error: bufmap already initialized.\n");
		ret = -EINVAL;
		goto out_unmap_bufmap;
	}
	__orangefs_bufmap = bufmap;
	install(&rw_map,
		bufmap->desc_count,
		bufmap->buffer_index_array);
	install(&readdir_map,
		ORANGEFS_READDIR_DEFAULT_DESC_COUNT,
		bufmap->readdir_index_array);
	spin_unlock(&orangefs_bufmap_lock);

	gossip_debug(GOSSIP_BUFMAP_DEBUG,
		     "orangefs_bufmap_initialize: exiting normally\n");
	return 0;

out_unmap_bufmap:
	orangefs_bufmap_unmap(bufmap);
out_free_bufmap:
	orangefs_bufmap_free(bufmap);
out:
	return ret;
}

/*
 * orangefs_bufmap_finalize()
 *
 * shuts down the mapped buffer interface and releases any resources
 * associated with it
 *
 * no return value
 */
void orangefs_bufmap_finalize(void)
{
	struct orangefs_bufmap *bufmap = __orangefs_bufmap;
	if (!bufmap)
		return;
	gossip_debug(GOSSIP_BUFMAP_DEBUG, "orangefs_bufmap_finalize: called\n");
	mark_killed(&rw_map);
	mark_killed(&readdir_map);
	gossip_debug(GOSSIP_BUFMAP_DEBUG,
		     "orangefs_bufmap_finalize: exiting normally\n");
}

void orangefs_bufmap_run_down(void)
{
	struct orangefs_bufmap *bufmap = __orangefs_bufmap;
	if (!bufmap)
		return;
	run_down(&rw_map);
	run_down(&readdir_map);
	spin_lock(&orangefs_bufmap_lock);
	__orangefs_bufmap = NULL;
	spin_unlock(&orangefs_bufmap_lock);
	orangefs_bufmap_unmap(bufmap);
	orangefs_bufmap_free(bufmap);
}

/*
 * orangefs_bufmap_get()
 *
 * gets a free mapped buffer descriptor, will sleep until one becomes
 * available if necessary
 *
 * returns slot on success, -errno on failure
 */
int orangefs_bufmap_get(void)
{
	return get(&rw_map);
}

/*
 * orangefs_bufmap_put()
 *
 * returns a mapped buffer descriptor to the collection
 *
 * no return value
 */
void orangefs_bufmap_put(int buffer_index)
{
	put(&rw_map, buffer_index);
}

/*
 * orangefs_readdir_index_get()
 *
 * gets a free descriptor, will sleep until one becomes
 * available if necessary.
 * Although the readdir buffers are not mapped into kernel space
 * we could do that at a later point of time. Regardless, these
 * indices are used by the client-core.
 *
 * returns slot on success, -errno on failure
 */
int orangefs_readdir_index_get(void)
{
	return get(&readdir_map);
}

void orangefs_readdir_index_put(int buffer_index)
{
	put(&readdir_map, buffer_index);
}

/*
 * we've been handed an iovec, we need to copy it to 
 * the shared memory descriptor at "buffer_index".
 */
int orangefs_bufmap_copy_from_iovec(struct iov_iter *iter,
				int buffer_index,
				size_t size)
{
	struct orangefs_bufmap_desc *to;
	int i;

	gossip_debug(GOSSIP_BUFMAP_DEBUG,
		     "%s: buffer_index:%d: size:%zu:\n",
		     __func__, buffer_index, size);

	to = &__orangefs_bufmap->desc_array[buffer_index];
	for (i = 0; size; i++) {
		struct page *page = to->page_array[i];
		size_t n = size;
		if (n > PAGE_SIZE)
			n = PAGE_SIZE;
		n = copy_page_from_iter(page, 0, n, iter);
		if (!n)
			return -EFAULT;
		size -= n;
	}
	return 0;

}

/*
 * we've been handed an iovec, we need to fill it from
 * the shared memory descriptor at "buffer_index".
 */
int orangefs_bufmap_copy_to_iovec(struct iov_iter *iter,
				    int buffer_index,
				    size_t size)
{
	struct orangefs_bufmap_desc *from;
	int i;

	from = &__orangefs_bufmap->desc_array[buffer_index];
	gossip_debug(GOSSIP_BUFMAP_DEBUG,
		     "%s: buffer_index:%d: size:%zu:\n",
		     __func__, buffer_index, size);


	for (i = 0; size; i++) {
		struct page *page = from->page_array[i];
		size_t n = size;
		if (n > PAGE_SIZE)
			n = PAGE_SIZE;
		n = copy_page_to_iter(page, 0, n, iter);
		if (!n)
			return -EFAULT;
		size -= n;
	}
	return 0;
}
Commit	Line	Data
274dcf55 MM	1	/*
	2	* (C) 2001 Clemson University and The University of Chicago
	3	*
	4	* See COPYING in top-level directory.
	5	*/
	6	#include "protocol.h"
575e9461 MM	7	#include "orangefs-kernel.h"
575e9461 MM	8	#include "orangefs-bufmap.h"
274dcf55	9
ea2c9c9f AV	10	struct slot_map {
	11	int c;
	12	wait_queue_head_t q;
	13	int count;
	14	unsigned long *map;
	15	};
	16
	17	static struct slot_map rw_map = {
	18	.c = -1,
	19	.q = __WAIT_QUEUE_HEAD_INITIALIZER(rw_map.q)
	20	};
	21	static struct slot_map readdir_map = {
	22	.c = -1,
	23	.q = __WAIT_QUEUE_HEAD_INITIALIZER(readdir_map.q)
	24	};
	25
	26
	27	static void install(struct slot_map m, int count, unsigned long map)
	28	{
	29	spin_lock(&m->q.lock);
	30	m->c = m->count = count;
	31	m->map = map;
	32	wake_up_all_locked(&m->q);
	33	spin_unlock(&m->q.lock);
	34	}
	35
	36	static void mark_killed(struct slot_map *m)
	37	{
	38	spin_lock(&m->q.lock);
	39	m->c -= m->count + 1;
	40	spin_unlock(&m->q.lock);
	41	}
	42
	43	static void run_down(struct slot_map *m)
	44	{
	45	DEFINE_WAIT(wait);
	46	spin_lock(&m->q.lock);
	47	if (m->c != -1) {
	48	for (;;) {
	49	if (likely(list_empty(&wait.task_list)))
	50	__add_wait_queue_tail(&m->q, &wait);
	51	set_current_state(TASK_UNINTERRUPTIBLE);
	52
	53	if (m->c == -1)
	54	break;
	55
	56	spin_unlock(&m->q.lock);
	57	schedule();
	58	spin_lock(&m->q.lock);
	59	}
	60	__remove_wait_queue(&m->q, &wait);
	61	__set_current_state(TASK_RUNNING);
	62	}
	63	m->map = NULL;
	64	spin_unlock(&m->q.lock);
	65	}
	66
	67	static void put(struct slot_map *m, int slot)
	68	{
	69	int v;
	70	spin_lock(&m->q.lock);
	71	__clear_bit(slot, m->map);
	72	v = ++m->c;
	73	if (unlikely(v == 1)) /* no free slots -> one free slot */
74	wake_up_locked(&m->q);
75	else if (unlikely(v == -1)) /* finished dying */
76	wake_up_all_locked(&m->q);
77	spin_unlock(&m->q.lock);
78	}
79
80	static int wait_for_free(struct slot_map *m)
81	{
82	long left = slot_timeout_secs * HZ;
83	DEFINE_WAIT(wait);
84
85	do {
86	long n = left, t;
87	if (likely(list_empty(&wait.task_list)))
88	__add_wait_queue_tail_exclusive(&m->q, &wait);
89	set_current_state(TASK_INTERRUPTIBLE);
90
91	if (m->c > 0)
92	break;
93
94	if (m->c < 0) {
95	/* we are waiting for map to be installed */
96	/* it would better be there soon, or we go away */
97	if (n > ORANGEFS_BUFMAP_WAIT_TIMEOUT_SECS * HZ)
98	n = ORANGEFS_BUFMAP_WAIT_TIMEOUT_SECS * HZ;
99	}
100	spin_unlock(&m->q.lock);
101	t = schedule_timeout(n);
102	spin_lock(&m->q.lock);
103	if (unlikely(!t) && n != left && m->c < 0)
104	left = t;
105	else
106	left = t + (left - n);
107	if (unlikely(signal_pending(current)))
108	left = -EINTR;
109	} while (left > 0);
110
111	if (!list_empty(&wait.task_list))
112	list_del(&wait.task_list);
113	else if (left <= 0 && waitqueue_active(&m->q))
114	__wake_up_locked_key(&m->q, TASK_INTERRUPTIBLE, NULL);
115	__set_current_state(TASK_RUNNING);
116
117	if (likely(left > 0))
118	return 0;
119
120	return left < 0 ? -EINTR : -ETIMEDOUT;
121	}
122
123	static int get(struct slot_map *m)
124	{
125	int res = 0;
126	spin_lock(&m->q.lock);
127	if (unlikely(m->c <= 0))
128	res = wait_for_free(m);
129	if (likely(!res)) {
130	m->c--;
131	res = find_first_zero_bit(m->map, m->count);
132	__set_bit(res, m->map);
133	}
134	spin_unlock(&m->q.lock);
135	return res;
136	}
274dcf55	137
bf89f584 MB	138	/* used to describe mapped buffers */
	139	struct orangefs_bufmap_desc {
	140	void uaddr; / user space address pointer */
	141	struct page *page_array; / array of mapped pages */
	142	int array_count; /* size of above arrays */
	143	struct list_head list_link;
	144	};
	145
8bb8aefd	146	static struct orangefs_bufmap {
274dcf55 MM	147	int desc_size;
	148	int desc_shift;
	149	int desc_count;
	150	int total_size;
	151	int page_count;
	152
	153	struct page **page_array;
8bb8aefd	154	struct orangefs_bufmap_desc *desc_array;
274dcf55 MM	155
274dcf55 MM	156	/* array to track usage of buffer descriptors */
ea2c9c9f	157	unsigned long *buffer_index_array;
274dcf55 MM	158
274dcf55 MM	159	/* array to track usage of buffer descriptors for readdir */
ea2c9c9f AV	160	#define N DIV_ROUND_UP(ORANGEFS_READDIR_DEFAULT_DESC_COUNT, BITS_PER_LONG)
	161	unsigned long readdir_index_array[N];
	162	#undef N
8bb8aefd	163	} *__orangefs_bufmap;
274dcf55	164
8bb8aefd	165	static DEFINE_SPINLOCK(orangefs_bufmap_lock);
274dcf55 MM	166
274dcf55 MM	167	static void
8bb8aefd	168	orangefs_bufmap_unmap(struct orangefs_bufmap *bufmap)
274dcf55 MM	169	{
	170	int i;
	171
	172	for (i = 0; i < bufmap->page_count; i++)
09cbfeaf	173	put_page(bufmap->page_array[i]);
274dcf55 MM	174	}
	175
	176	static void
8bb8aefd	177	orangefs_bufmap_free(struct orangefs_bufmap *bufmap)
274dcf55 MM	178	{
	179	kfree(bufmap->page_array);
	180	kfree(bufmap->desc_array);
	181	kfree(bufmap->buffer_index_array);
	182	kfree(bufmap);
	183	}
	184
b09d10df MB	185	/*
	186	* XXX: Can the size and shift change while the caller gives up the
	187	* XXX: lock between calling this and doing something useful?
	188	*/
	189
765a75b3	190	int orangefs_bufmap_size_query(void)
274dcf55	191	{
b09d10df MB	192	struct orangefs_bufmap *bufmap;
b09d10df MB	193	int size = 0;
17804184 AV	194	spin_lock(&orangefs_bufmap_lock);
	195	bufmap = __orangefs_bufmap;
	196	if (bufmap)
b09d10df	197	size = bufmap->desc_size;
17804184	198	spin_unlock(&orangefs_bufmap_lock);
274dcf55 MM	199	return size;
	200	}
	201
765a75b3	202	int orangefs_bufmap_shift_query(void)
274dcf55	203	{
b09d10df MB	204	struct orangefs_bufmap *bufmap;
b09d10df MB	205	int shift = 0;
17804184 AV	206	spin_lock(&orangefs_bufmap_lock);
	207	bufmap = __orangefs_bufmap;
	208	if (bufmap)
b09d10df	209	shift = bufmap->desc_shift;
17804184	210	spin_unlock(&orangefs_bufmap_lock);
274dcf55 MM	211	return shift;
	212	}
	213
	214	static DECLARE_WAIT_QUEUE_HEAD(bufmap_waitq);
	215	static DECLARE_WAIT_QUEUE_HEAD(readdir_waitq);
	216
	217	/*
7d221485	218	* orangefs_get_bufmap_init
274dcf55 MM	219	*
	220	* If bufmap_init is 1, then the shared memory system, including the
	221	* buffer_index_array, is available. Otherwise, it is not.
	222	*
	223	* returns the value of bufmap_init
	224	*/
7d221485	225	int orangefs_get_bufmap_init(void)
274dcf55	226	{
8bb8aefd	227	return __orangefs_bufmap ? 1 : 0;
274dcf55 MM	228	}
	229
	230
8bb8aefd YL	231	static struct orangefs_bufmap *
8bb8aefd YL	232	orangefs_bufmap_alloc(struct ORANGEFS_dev_map_desc *user_desc)
274dcf55	233	{
8bb8aefd	234	struct orangefs_bufmap *bufmap;
274dcf55 MM	235
	236	bufmap = kzalloc(sizeof(*bufmap), GFP_KERNEL);
	237	if (!bufmap)
	238	goto out;
	239
274dcf55 MM	240	bufmap->total_size = user_desc->total_size;
	241	bufmap->desc_count = user_desc->count;
	242	bufmap->desc_size = user_desc->size;
	243	bufmap->desc_shift = ilog2(bufmap->desc_size);
	244
274dcf55	245	bufmap->buffer_index_array =
ea2c9c9f	246	kzalloc(DIV_ROUND_UP(bufmap->desc_count, BITS_PER_LONG), GFP_KERNEL);
274dcf55	247	if (!bufmap->buffer_index_array) {
8bb8aefd	248	gossip_err("orangefs: could not allocate %d buffer indices\n",
274dcf55 MM	249	bufmap->desc_count);
	250	goto out_free_bufmap;
	251	}
274dcf55 MM	252
274dcf55 MM	253	bufmap->desc_array =
8bb8aefd	254	kcalloc(bufmap->desc_count, sizeof(struct orangefs_bufmap_desc),
274dcf55 MM	255	GFP_KERNEL);
274dcf55 MM	256	if (!bufmap->desc_array) {
8bb8aefd	257	gossip_err("orangefs: could not allocate %d descriptors\n",
274dcf55 MM	258	bufmap->desc_count);
	259	goto out_free_index_array;
	260	}
	261
	262	bufmap->page_count = bufmap->total_size / PAGE_SIZE;
	263
	264	/* allocate storage to track our page mappings */
	265	bufmap->page_array =
	266	kcalloc(bufmap->page_count, sizeof(struct page *), GFP_KERNEL);
	267	if (!bufmap->page_array)
	268	goto out_free_desc_array;
	269
	270	return bufmap;
	271
	272	out_free_desc_array:
	273	kfree(bufmap->desc_array);
	274	out_free_index_array:
	275	kfree(bufmap->buffer_index_array);
	276	out_free_bufmap:
	277	kfree(bufmap);
	278	out:
	279	return NULL;
	280	}
	281
	282	static int
8bb8aefd YL	283	orangefs_bufmap_map(struct orangefs_bufmap *bufmap,
8bb8aefd YL	284	struct ORANGEFS_dev_map_desc *user_desc)
274dcf55 MM	285	{
	286	int pages_per_desc = bufmap->desc_size / PAGE_SIZE;
	287	int offset = 0, ret, i;
	288
	289	/* map the pages */
16742f2d AV	290	ret = get_user_pages_fast((unsigned long)user_desc->ptr,
16742f2d AV	291	bufmap->page_count, 1, bufmap->page_array);
274dcf55 MM	292
	293	if (ret < 0)
	294	return ret;
	295
	296	if (ret != bufmap->page_count) {
8bb8aefd	297	gossip_err("orangefs error: asked for %d pages, only got %d.\n",
274dcf55 MM	298	bufmap->page_count, ret);
	299
	300	for (i = 0; i < ret; i++) {
	301	SetPageError(bufmap->page_array[i]);
09cbfeaf	302	put_page(bufmap->page_array[i]);
274dcf55 MM	303	}
	304	return -ENOMEM;
	305	}
	306
	307	/*
	308	* ideally we want to get kernel space pointers for each page, but
	309	* we can't kmap that many pages at once if highmem is being used.
	310	* so instead, we just kmap/kunmap the page address each time the
	311	* kaddr is needed.
	312	*/
	313	for (i = 0; i < bufmap->page_count; i++)
	314	flush_dcache_page(bufmap->page_array[i]);
	315
	316	/* build a list of available descriptors */
	317	for (offset = 0, i = 0; i < bufmap->desc_count; i++) {
	318	bufmap->desc_array[i].page_array = &bufmap->page_array[offset];
	319	bufmap->desc_array[i].array_count = pages_per_desc;
	320	bufmap->desc_array[i].uaddr =
	321	(user_desc->ptr + (i * pages_per_desc * PAGE_SIZE));
	322	offset += pages_per_desc;
	323	}
	324
	325	return 0;
	326	}
	327
	328	/*
8bb8aefd	329	* orangefs_bufmap_initialize()
274dcf55 MM	330	*
	331	* initializes the mapped buffer interface
	332	*
	333	* returns 0 on success, -errno on failure
	334	*/
8bb8aefd	335	int orangefs_bufmap_initialize(struct ORANGEFS_dev_map_desc *user_desc)
274dcf55	336	{
8bb8aefd	337	struct orangefs_bufmap *bufmap;
274dcf55 MM	338	int ret = -EINVAL;
	339
	340	gossip_debug(GOSSIP_BUFMAP_DEBUG,
8bb8aefd	341	"orangefs_bufmap_initialize: called (ptr ("
274dcf55 MM	342	"%p) sz (%d) cnt(%d).\n",
	343	user_desc->ptr,
	344	user_desc->size,
	345	user_desc->count);
	346
	347	/*
	348	* sanity check alignment and size of buffer that caller wants to
	349	* work with
	350	*/
	351	if (PAGE_ALIGN((unsigned long)user_desc->ptr) !=
	352	(unsigned long)user_desc->ptr) {
8bb8aefd	353	gossip_err("orangefs error: memory alignment (front). %p\n",
274dcf55 MM	354	user_desc->ptr);
	355	goto out;
	356	}
	357
	358	if (PAGE_ALIGN(((unsigned long)user_desc->ptr + user_desc->total_size))
	359	!= (unsigned long)(user_desc->ptr + user_desc->total_size)) {
8bb8aefd	360	gossip_err("orangefs error: memory alignment (back).(%p + %d)\n",
274dcf55 MM	361	user_desc->ptr,
	362	user_desc->total_size);
	363	goto out;
	364	}
	365
	366	if (user_desc->total_size != (user_desc->size * user_desc->count)) {
8bb8aefd	367	gossip_err("orangefs error: user provided an oddly sized buffer: (%d, %d, %d)\n",
274dcf55 MM	368	user_desc->total_size,
	369	user_desc->size,
	370	user_desc->count);
	371	goto out;
	372	}
	373
	374	if ((user_desc->size % PAGE_SIZE) != 0) {
8bb8aefd	375	gossip_err("orangefs error: bufmap size not page size divisible (%d).\n",
274dcf55 MM	376	user_desc->size);
	377	goto out;
	378	}
	379
	380	ret = -ENOMEM;
8bb8aefd	381	bufmap = orangefs_bufmap_alloc(user_desc);
274dcf55 MM	382	if (!bufmap)
	383	goto out;
	384
8bb8aefd	385	ret = orangefs_bufmap_map(bufmap, user_desc);
274dcf55 MM	386	if (ret)
	387	goto out_free_bufmap;
	388
	389
8bb8aefd YL	390	spin_lock(&orangefs_bufmap_lock);
	391	if (__orangefs_bufmap) {
	392	spin_unlock(&orangefs_bufmap_lock);
	393	gossip_err("orangefs: error: bufmap already initialized.\n");
ea2c9c9f	394	ret = -EINVAL;
274dcf55 MM	395	goto out_unmap_bufmap;
274dcf55 MM	396	}
8bb8aefd	397	__orangefs_bufmap = bufmap;
ea2c9c9f AV	398	install(&rw_map,
	399	bufmap->desc_count,
	400	bufmap->buffer_index_array);
	401	install(&readdir_map,
	402	ORANGEFS_READDIR_DEFAULT_DESC_COUNT,
	403	bufmap->readdir_index_array);
8bb8aefd	404	spin_unlock(&orangefs_bufmap_lock);
274dcf55	405
274dcf55	406	gossip_debug(GOSSIP_BUFMAP_DEBUG,
8bb8aefd	407	"orangefs_bufmap_initialize: exiting normally\n");
274dcf55 MM	408	return 0;
	409
	410	out_unmap_bufmap:
8bb8aefd	411	orangefs_bufmap_unmap(bufmap);
274dcf55	412	out_free_bufmap:
8bb8aefd	413	orangefs_bufmap_free(bufmap);
274dcf55 MM	414	out:
	415	return ret;
	416	}
	417
	418	/*
8bb8aefd	419	* orangefs_bufmap_finalize()
274dcf55 MM	420	*
	421	* shuts down the mapped buffer interface and releases any resources
	422	* associated with it
	423	*
	424	* no return value
	425	*/
8bb8aefd	426	void orangefs_bufmap_finalize(void)
274dcf55	427	{
ea2c9c9f AV	428	struct orangefs_bufmap *bufmap = __orangefs_bufmap;
	429	if (!bufmap)
	430	return;
8bb8aefd	431	gossip_debug(GOSSIP_BUFMAP_DEBUG, "orangefs_bufmap_finalize: called\n");
ea2c9c9f AV	432	mark_killed(&rw_map);
ea2c9c9f AV	433	mark_killed(&readdir_map);
274dcf55	434	gossip_debug(GOSSIP_BUFMAP_DEBUG,
8bb8aefd	435	"orangefs_bufmap_finalize: exiting normally\n");
274dcf55 MM	436	}
274dcf55 MM	437
ea2c9c9f	438	void orangefs_bufmap_run_down(void)
274dcf55	439	{
ea2c9c9f AV	440	struct orangefs_bufmap *bufmap = __orangefs_bufmap;
ea2c9c9f AV	441	if (!bufmap)
274dcf55	442	return;
ea2c9c9f AV	443	run_down(&rw_map);
	444	run_down(&readdir_map);
	445	spin_lock(&orangefs_bufmap_lock);
	446	__orangefs_bufmap = NULL;
	447	spin_unlock(&orangefs_bufmap_lock);
	448	orangefs_bufmap_unmap(bufmap);
	449	orangefs_bufmap_free(bufmap);
274dcf55 MM	450	}
	451
	452	/*
8bb8aefd	453	* orangefs_bufmap_get()
274dcf55 MM	454	*
	455	* gets a free mapped buffer descriptor, will sleep until one becomes
	456	* available if necessary
	457	*
b8a99a8f	458	* returns slot on success, -errno on failure
274dcf55	459	*/
b8a99a8f	460	int orangefs_bufmap_get(void)
274dcf55	461	{
b8a99a8f	462	return get(&rw_map);
274dcf55 MM	463	}
	464
	465	/*
8bb8aefd	466	* orangefs_bufmap_put()
274dcf55 MM	467	*
	468	* returns a mapped buffer descriptor to the collection
	469	*
	470	* no return value
	471	*/
1357d06d	472	void orangefs_bufmap_put(int buffer_index)
274dcf55	473	{
ea2c9c9f	474	put(&rw_map, buffer_index);
274dcf55 MM	475	}
	476
	477	/*
7d221485	478	* orangefs_readdir_index_get()
274dcf55 MM	479	*
	480	* gets a free descriptor, will sleep until one becomes
	481	* available if necessary.
	482	* Although the readdir buffers are not mapped into kernel space
	483	* we could do that at a later point of time. Regardless, these
	484	* indices are used by the client-core.
	485	*
b8a99a8f	486	* returns slot on success, -errno on failure
274dcf55	487	*/
b8a99a8f	488	int orangefs_readdir_index_get(void)
274dcf55	489	{
b8a99a8f	490	return get(&readdir_map);
274dcf55 MM	491	}
274dcf55 MM	492
82d37f19	493	void orangefs_readdir_index_put(int buffer_index)
274dcf55	494	{
ea2c9c9f	495	put(&readdir_map, buffer_index);
274dcf55 MM	496	}
274dcf55 MM	497
b5e376ea MM	498	/*
	499	* we've been handed an iovec, we need to copy it to
	500	* the shared memory descriptor at "buffer_index".
	501	*/
bf6bf606	502	int orangefs_bufmap_copy_from_iovec(struct iov_iter *iter,
54804949 MM	503	int buffer_index,
54804949 MM	504	size_t size)
274dcf55	505	{
bf6bf606	506	struct orangefs_bufmap_desc *to;
4d1c4404	507	int i;
274dcf55 MM	508
274dcf55 MM	509	gossip_debug(GOSSIP_BUFMAP_DEBUG,
34204fde	510	"%s: buffer_index:%d: size:%zu:\n",
4d1c4404	511	__func__, buffer_index, size);
274dcf55	512
bf6bf606	513	to = &__orangefs_bufmap->desc_array[buffer_index];
4d1c4404	514	for (i = 0; size; i++) {
34204fde AV	515	struct page *page = to->page_array[i];
	516	size_t n = size;
	517	if (n > PAGE_SIZE)
	518	n = PAGE_SIZE;
	519	n = copy_page_from_iter(page, 0, n, iter);
	520	if (!n)
	521	return -EFAULT;
	522	size -= n;
274dcf55	523	}
34204fde	524	return 0;
274dcf55	525
274dcf55 MM	526	}
	527
	528	/*
b5e376ea MM	529	* we've been handed an iovec, we need to fill it from
b5e376ea MM	530	* the shared memory descriptor at "buffer_index".
274dcf55	531	*/
bf6bf606	532	int orangefs_bufmap_copy_to_iovec(struct iov_iter *iter,
5c278228 AV	533	int buffer_index,
5c278228 AV	534	size_t size)
274dcf55	535	{
bf6bf606	536	struct orangefs_bufmap_desc *from;
4d1c4404	537	int i;
274dcf55	538
bf6bf606	539	from = &__orangefs_bufmap->desc_array[buffer_index];
274dcf55	540	gossip_debug(GOSSIP_BUFMAP_DEBUG,
5c278228 AV	541	"%s: buffer_index:%d: size:%zu:\n",
5c278228 AV	542	__func__, buffer_index, size);
274dcf55	543
274dcf55	544
5c278228 AV	545	for (i = 0; size; i++) {
	546	struct page *page = from->page_array[i];
	547	size_t n = size;
	548	if (n > PAGE_SIZE)
	549	n = PAGE_SIZE;
	550	n = copy_page_to_iter(page, 0, n, iter);
	551	if (!n)
	552	return -EFAULT;
	553	size -= n;
274dcf55	554	}
5c278228	555	return 0;
274dcf55	556	}