idr.c 28.0 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8
/*
 * 2002-10-18  written by Jim Houston jim.houston@ccur.com
 *	Copyright (C) 2002 by Concurrent Computer Corporation
 *	Distributed under the GNU GPL license version 2.
 *
 * Modified by George Anzinger to reuse immediately and to use
 * find bit instructions.  Also removed _irq on spinlocks.
 *
N
Nadia Derbey 已提交
9 10
 * Modified by Nadia Derbey to make it RCU safe.
 *
11
 * Small id to pointer translation service.
L
Linus Torvalds 已提交
12
 *
13
 * It uses a radix tree like structure as a sparse array indexed
L
Linus Torvalds 已提交
14
 * by the id to obtain the pointer.  The bitmap makes allocating
15
 * a new id quick.
L
Linus Torvalds 已提交
16 17 18 19 20 21 22 23 24 25
 *
 * You call it to allocate an id (an int) an associate with that id a
 * pointer or what ever, we treat it as a (void *).  You can pass this
 * id to a user for him to pass back at a later time.  You then pass
 * that id to this code and it returns your pointer.
 */

#ifndef TEST                        // to test in user space...
#include <linux/slab.h>
#include <linux/init.h>
26
#include <linux/export.h>
L
Linus Torvalds 已提交
27
#endif
J
Jeff Mahoney 已提交
28
#include <linux/err.h>
L
Linus Torvalds 已提交
29 30
#include <linux/string.h>
#include <linux/idr.h>
31
#include <linux/spinlock.h>
32 33
#include <linux/percpu.h>
#include <linux/hardirq.h>
L
Linus Torvalds 已提交
34

35 36 37 38 39 40 41 42 43
#define MAX_IDR_SHIFT		(sizeof(int) * 8 - 1)
#define MAX_IDR_BIT		(1U << MAX_IDR_SHIFT)

/* Leave the possibility of an incomplete final layer */
#define MAX_IDR_LEVEL ((MAX_IDR_SHIFT + IDR_BITS - 1) / IDR_BITS)

/* Number of id_layer structs to leave in free list */
#define MAX_IDR_FREE (MAX_IDR_LEVEL * 2)

44
static struct kmem_cache *idr_layer_cache;
45 46
static DEFINE_PER_CPU(struct idr_layer *, idr_preload_head);
static DEFINE_PER_CPU(int, idr_preload_cnt);
47
static DEFINE_SPINLOCK(simple_ida_lock);
L
Linus Torvalds 已提交
48

T
Tejun Heo 已提交
49 50 51 52 53 54 55 56
/* the maximum ID which can be allocated given idr->layers */
static int idr_max(int layers)
{
	int bits = min_t(int, layers * IDR_BITS, MAX_IDR_SHIFT);

	return (1 << bits) - 1;
}

T
Tejun Heo 已提交
57 58 59 60 61 62 63 64 65 66
/*
 * Prefix mask for an idr_layer at @layer.  For layer 0, the prefix mask is
 * all bits except for the lower IDR_BITS.  For layer 1, 2 * IDR_BITS, and
 * so on.
 */
static int idr_layer_prefix_mask(int layer)
{
	return ~idr_max(layer + 1);
}

67
static struct idr_layer *get_from_free_list(struct idr *idp)
L
Linus Torvalds 已提交
68 69
{
	struct idr_layer *p;
70
	unsigned long flags;
L
Linus Torvalds 已提交
71

72
	spin_lock_irqsave(&idp->lock, flags);
L
Linus Torvalds 已提交
73 74 75 76 77
	if ((p = idp->id_free)) {
		idp->id_free = p->ary[0];
		idp->id_free_cnt--;
		p->ary[0] = NULL;
	}
78
	spin_unlock_irqrestore(&idp->lock, flags);
L
Linus Torvalds 已提交
79 80 81
	return(p);
}

82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102
/**
 * idr_layer_alloc - allocate a new idr_layer
 * @gfp_mask: allocation mask
 * @layer_idr: optional idr to allocate from
 *
 * If @layer_idr is %NULL, directly allocate one using @gfp_mask or fetch
 * one from the per-cpu preload buffer.  If @layer_idr is not %NULL, fetch
 * an idr_layer from @idr->id_free.
 *
 * @layer_idr is to maintain backward compatibility with the old alloc
 * interface - idr_pre_get() and idr_get_new*() - and will be removed
 * together with per-pool preload buffer.
 */
static struct idr_layer *idr_layer_alloc(gfp_t gfp_mask, struct idr *layer_idr)
{
	struct idr_layer *new;

	/* this is the old path, bypass to get_from_free_list() */
	if (layer_idr)
		return get_from_free_list(layer_idr);

103 104 105 106 107 108 109 110
	/*
	 * Try to allocate directly from kmem_cache.  We want to try this
	 * before preload buffer; otherwise, non-preloading idr_alloc()
	 * users will end up taking advantage of preloading ones.  As the
	 * following is allowed to fail for preloaded cases, suppress
	 * warning this time.
	 */
	new = kmem_cache_zalloc(idr_layer_cache, gfp_mask | __GFP_NOWARN);
111 112 113 114 115 116 117
	if (new)
		return new;

	/*
	 * Try to fetch one from the per-cpu preload buffer if in process
	 * context.  See idr_preload() for details.
	 */
118 119 120 121 122 123 124 125 126 127 128
	if (!in_interrupt()) {
		preempt_disable();
		new = __this_cpu_read(idr_preload_head);
		if (new) {
			__this_cpu_write(idr_preload_head, new->ary[0]);
			__this_cpu_dec(idr_preload_cnt);
			new->ary[0] = NULL;
		}
		preempt_enable();
		if (new)
			return new;
129
	}
130 131 132 133 134 135

	/*
	 * Both failed.  Try kmem_cache again w/o adding __GFP_NOWARN so
	 * that memory allocation failure warning is printed as intended.
	 */
	return kmem_cache_zalloc(idr_layer_cache, gfp_mask);
136 137
}

N
Nadia Derbey 已提交
138 139 140 141 142 143 144 145
static void idr_layer_rcu_free(struct rcu_head *head)
{
	struct idr_layer *layer;

	layer = container_of(head, struct idr_layer, rcu_head);
	kmem_cache_free(idr_layer_cache, layer);
}

T
Tejun Heo 已提交
146
static inline void free_layer(struct idr *idr, struct idr_layer *p)
N
Nadia Derbey 已提交
147
{
148
	if (idr->hint == p)
T
Tejun Heo 已提交
149
		RCU_INIT_POINTER(idr->hint, NULL);
N
Nadia Derbey 已提交
150 151 152
	call_rcu(&p->rcu_head, idr_layer_rcu_free);
}

S
Sonny Rao 已提交
153
/* only called when idp->lock is held */
154
static void __move_to_free_list(struct idr *idp, struct idr_layer *p)
S
Sonny Rao 已提交
155 156 157 158 159 160
{
	p->ary[0] = idp->id_free;
	idp->id_free = p;
	idp->id_free_cnt++;
}

161
static void move_to_free_list(struct idr *idp, struct idr_layer *p)
L
Linus Torvalds 已提交
162
{
163 164
	unsigned long flags;

L
Linus Torvalds 已提交
165 166 167
	/*
	 * Depends on the return element being zeroed.
	 */
168
	spin_lock_irqsave(&idp->lock, flags);
169
	__move_to_free_list(idp, p);
170
	spin_unlock_irqrestore(&idp->lock, flags);
L
Linus Torvalds 已提交
171 172
}

T
Tejun Heo 已提交
173 174 175 176 177
static void idr_mark_full(struct idr_layer **pa, int id)
{
	struct idr_layer *p = pa[0];
	int l = 0;

178
	__set_bit(id & IDR_MASK, p->bitmap);
T
Tejun Heo 已提交
179 180 181 182 183 184
	/*
	 * If this layer is full mark the bit in the layer above to
	 * show that this part of the radix tree is full.  This may
	 * complete the layer above and require walking up the radix
	 * tree.
	 */
185
	while (bitmap_full(p->bitmap, IDR_SIZE)) {
T
Tejun Heo 已提交
186 187 188
		if (!(p = pa[++l]))
			break;
		id = id >> IDR_BITS;
189
		__set_bit((id & IDR_MASK), p->bitmap);
T
Tejun Heo 已提交
190 191 192
	}
}

S
Stephen Hemminger 已提交
193
static int __idr_pre_get(struct idr *idp, gfp_t gfp_mask)
L
Linus Torvalds 已提交
194
{
195
	while (idp->id_free_cnt < MAX_IDR_FREE) {
L
Linus Torvalds 已提交
196
		struct idr_layer *new;
197
		new = kmem_cache_zalloc(idr_layer_cache, gfp_mask);
198
		if (new == NULL)
L
Linus Torvalds 已提交
199
			return (0);
200
		move_to_free_list(idp, new);
L
Linus Torvalds 已提交
201 202 203 204
	}
	return 1;
}

T
Tejun Heo 已提交
205 206 207 208 209
/**
 * sub_alloc - try to allocate an id without growing the tree depth
 * @idp: idr handle
 * @starting_id: id to start search at
 * @pa: idr_layer[MAX_IDR_LEVEL] used as backtrack buffer
210 211
 * @gfp_mask: allocation mask for idr_layer_alloc()
 * @layer_idr: optional idr passed to idr_layer_alloc()
T
Tejun Heo 已提交
212 213 214 215 216 217 218 219 220
 *
 * Allocate an id in range [@starting_id, INT_MAX] from @idp without
 * growing its depth.  Returns
 *
 *  the allocated id >= 0 if successful,
 *  -EAGAIN if the tree needs to grow for allocation to succeed,
 *  -ENOSPC if the id space is exhausted,
 *  -ENOMEM if more idr_layers need to be allocated.
 */
221 222
static int sub_alloc(struct idr *idp, int *starting_id, struct idr_layer **pa,
		     gfp_t gfp_mask, struct idr *layer_idr)
L
Linus Torvalds 已提交
223 224 225
{
	int n, m, sh;
	struct idr_layer *p, *new;
226
	int l, id, oid;
L
Linus Torvalds 已提交
227 228

	id = *starting_id;
229
 restart:
L
Linus Torvalds 已提交
230 231 232 233 234 235 236 237
	p = idp->top;
	l = idp->layers;
	pa[l--] = NULL;
	while (1) {
		/*
		 * We run around this while until we reach the leaf node...
		 */
		n = (id >> (IDR_BITS*l)) & IDR_MASK;
238
		m = find_next_zero_bit(p->bitmap, IDR_SIZE, n);
L
Linus Torvalds 已提交
239 240 241
		if (m == IDR_SIZE) {
			/* no space available go back to previous layer. */
			l++;
242
			oid = id;
243
			id = (id | ((1 << (IDR_BITS * l)) - 1)) + 1;
244 245

			/* if already at the top layer, we need to grow */
246
			if (id > idr_max(idp->layers)) {
L
Linus Torvalds 已提交
247
				*starting_id = id;
T
Tejun Heo 已提交
248
				return -EAGAIN;
L
Linus Torvalds 已提交
249
			}
250 251
			p = pa[l];
			BUG_ON(!p);
252 253 254 255 256 257 258 259 260

			/* If we need to go up one layer, continue the
			 * loop; otherwise, restart from the top.
			 */
			sh = IDR_BITS * (l + 1);
			if (oid >> sh == id >> sh)
				continue;
			else
				goto restart;
L
Linus Torvalds 已提交
261 262 263 264 265
		}
		if (m != n) {
			sh = IDR_BITS*l;
			id = ((id >> sh) ^ n ^ m) << sh;
		}
266
		if ((id >= MAX_IDR_BIT) || (id < 0))
T
Tejun Heo 已提交
267
			return -ENOSPC;
L
Linus Torvalds 已提交
268 269 270 271 272 273
		if (l == 0)
			break;
		/*
		 * Create the layer below if it is missing.
		 */
		if (!p->ary[m]) {
274
			new = idr_layer_alloc(gfp_mask, layer_idr);
275
			if (!new)
T
Tejun Heo 已提交
276
				return -ENOMEM;
277
			new->layer = l-1;
T
Tejun Heo 已提交
278
			new->prefix = id & idr_layer_prefix_mask(new->layer);
N
Nadia Derbey 已提交
279
			rcu_assign_pointer(p->ary[m], new);
L
Linus Torvalds 已提交
280 281 282 283 284
			p->count++;
		}
		pa[l--] = p;
		p = p->ary[m];
	}
T
Tejun Heo 已提交
285 286 287

	pa[l] = p;
	return id;
L
Linus Torvalds 已提交
288 289
}

T
Tejun Heo 已提交
290
static int idr_get_empty_slot(struct idr *idp, int starting_id,
291 292
			      struct idr_layer **pa, gfp_t gfp_mask,
			      struct idr *layer_idr)
L
Linus Torvalds 已提交
293 294 295
{
	struct idr_layer *p, *new;
	int layers, v, id;
296
	unsigned long flags;
297

L
Linus Torvalds 已提交
298 299 300 301 302
	id = starting_id;
build_up:
	p = idp->top;
	layers = idp->layers;
	if (unlikely(!p)) {
303
		if (!(p = idr_layer_alloc(gfp_mask, layer_idr)))
T
Tejun Heo 已提交
304
			return -ENOMEM;
305
		p->layer = 0;
L
Linus Torvalds 已提交
306 307 308 309 310 311
		layers = 1;
	}
	/*
	 * Add a new layer to the top of the tree if the requested
	 * id is larger than the currently allocated space.
	 */
T
Tejun Heo 已提交
312
	while (id > idr_max(layers)) {
L
Linus Torvalds 已提交
313
		layers++;
314 315 316 317 318 319
		if (!p->count) {
			/* special case: if the tree is currently empty,
			 * then we grow the tree by moving the top node
			 * upwards.
			 */
			p->layer++;
T
Tejun Heo 已提交
320
			WARN_ON_ONCE(p->prefix);
L
Linus Torvalds 已提交
321
			continue;
322
		}
323
		if (!(new = idr_layer_alloc(gfp_mask, layer_idr))) {
L
Linus Torvalds 已提交
324 325 326 327
			/*
			 * The allocation failed.  If we built part of
			 * the structure tear it down.
			 */
328
			spin_lock_irqsave(&idp->lock, flags);
L
Linus Torvalds 已提交
329 330 331
			for (new = p; p && p != idp->top; new = p) {
				p = p->ary[0];
				new->ary[0] = NULL;
332 333
				new->count = 0;
				bitmap_clear(new->bitmap, 0, IDR_SIZE);
334
				__move_to_free_list(idp, new);
L
Linus Torvalds 已提交
335
			}
336
			spin_unlock_irqrestore(&idp->lock, flags);
T
Tejun Heo 已提交
337
			return -ENOMEM;
L
Linus Torvalds 已提交
338 339 340
		}
		new->ary[0] = p;
		new->count = 1;
341
		new->layer = layers-1;
T
Tejun Heo 已提交
342
		new->prefix = id & idr_layer_prefix_mask(new->layer);
343 344
		if (bitmap_full(p->bitmap, IDR_SIZE))
			__set_bit(0, new->bitmap);
L
Linus Torvalds 已提交
345 346
		p = new;
	}
N
Nadia Derbey 已提交
347
	rcu_assign_pointer(idp->top, p);
L
Linus Torvalds 已提交
348
	idp->layers = layers;
349
	v = sub_alloc(idp, &id, pa, gfp_mask, layer_idr);
T
Tejun Heo 已提交
350
	if (v == -EAGAIN)
L
Linus Torvalds 已提交
351 352 353 354
		goto build_up;
	return(v);
}

T
Tejun Heo 已提交
355 356 357 358
/*
 * @id and @pa are from a successful allocation from idr_get_empty_slot().
 * Install the user pointer @ptr and mark the slot full.
 */
T
Tejun Heo 已提交
359 360
static void idr_fill_slot(struct idr *idr, void *ptr, int id,
			  struct idr_layer **pa)
T
Tejun Heo 已提交
361
{
T
Tejun Heo 已提交
362 363 364
	/* update hint used for lookup, cleared from free_layer() */
	rcu_assign_pointer(idr->hint, pa[0]);

T
Tejun Heo 已提交
365 366 367
	rcu_assign_pointer(pa[0]->ary[id & IDR_MASK], (struct idr_layer *)ptr);
	pa[0]->count++;
	idr_mark_full(pa, id);
T
Tejun Heo 已提交
368 369
}

L
Linus Torvalds 已提交
370

371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453
/**
 * idr_preload - preload for idr_alloc()
 * @gfp_mask: allocation mask to use for preloading
 *
 * Preload per-cpu layer buffer for idr_alloc().  Can only be used from
 * process context and each idr_preload() invocation should be matched with
 * idr_preload_end().  Note that preemption is disabled while preloaded.
 *
 * The first idr_alloc() in the preloaded section can be treated as if it
 * were invoked with @gfp_mask used for preloading.  This allows using more
 * permissive allocation masks for idrs protected by spinlocks.
 *
 * For example, if idr_alloc() below fails, the failure can be treated as
 * if idr_alloc() were called with GFP_KERNEL rather than GFP_NOWAIT.
 *
 *	idr_preload(GFP_KERNEL);
 *	spin_lock(lock);
 *
 *	id = idr_alloc(idr, ptr, start, end, GFP_NOWAIT);
 *
 *	spin_unlock(lock);
 *	idr_preload_end();
 *	if (id < 0)
 *		error;
 */
void idr_preload(gfp_t gfp_mask)
{
	/*
	 * Consuming preload buffer from non-process context breaks preload
	 * allocation guarantee.  Disallow usage from those contexts.
	 */
	WARN_ON_ONCE(in_interrupt());
	might_sleep_if(gfp_mask & __GFP_WAIT);

	preempt_disable();

	/*
	 * idr_alloc() is likely to succeed w/o full idr_layer buffer and
	 * return value from idr_alloc() needs to be checked for failure
	 * anyway.  Silently give up if allocation fails.  The caller can
	 * treat failures from idr_alloc() as if idr_alloc() were called
	 * with @gfp_mask which should be enough.
	 */
	while (__this_cpu_read(idr_preload_cnt) < MAX_IDR_FREE) {
		struct idr_layer *new;

		preempt_enable();
		new = kmem_cache_zalloc(idr_layer_cache, gfp_mask);
		preempt_disable();
		if (!new)
			break;

		/* link the new one to per-cpu preload list */
		new->ary[0] = __this_cpu_read(idr_preload_head);
		__this_cpu_write(idr_preload_head, new);
		__this_cpu_inc(idr_preload_cnt);
	}
}
EXPORT_SYMBOL(idr_preload);

/**
 * idr_alloc - allocate new idr entry
 * @idr: the (initialized) idr
 * @ptr: pointer to be associated with the new id
 * @start: the minimum id (inclusive)
 * @end: the maximum id (exclusive, <= 0 for max)
 * @gfp_mask: memory allocation flags
 *
 * Allocate an id in [start, end) and associate it with @ptr.  If no ID is
 * available in the specified range, returns -ENOSPC.  On memory allocation
 * failure, returns -ENOMEM.
 *
 * Note that @end is treated as max when <= 0.  This is to always allow
 * using @start + N as @end as long as N is inside integer range.
 *
 * The user is responsible for exclusively synchronizing all operations
 * which may modify @idr.  However, read-only accesses such as idr_find()
 * or iteration can be performed under RCU read lock provided the user
 * destroys @ptr in RCU-safe way after removal from idr.
 */
int idr_alloc(struct idr *idr, void *ptr, int start, int end, gfp_t gfp_mask)
{
	int max = end > 0 ? end - 1 : INT_MAX;	/* inclusive upper limit */
T
Tejun Heo 已提交
454
	struct idr_layer *pa[MAX_IDR_LEVEL + 1];
455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471
	int id;

	might_sleep_if(gfp_mask & __GFP_WAIT);

	/* sanity checks */
	if (WARN_ON_ONCE(start < 0))
		return -EINVAL;
	if (unlikely(max < start))
		return -ENOSPC;

	/* allocate id */
	id = idr_get_empty_slot(idr, start, pa, gfp_mask, NULL);
	if (unlikely(id < 0))
		return id;
	if (unlikely(id > max))
		return -ENOSPC;

T
Tejun Heo 已提交
472
	idr_fill_slot(idr, ptr, id, pa);
473 474 475 476
	return id;
}
EXPORT_SYMBOL_GPL(idr_alloc);

J
Jeff Layton 已提交
477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503
/**
 * idr_alloc_cyclic - allocate new idr entry in a cyclical fashion
 * @idr: the (initialized) idr
 * @ptr: pointer to be associated with the new id
 * @start: the minimum id (inclusive)
 * @end: the maximum id (exclusive, <= 0 for max)
 * @gfp_mask: memory allocation flags
 *
 * Essentially the same as idr_alloc, but prefers to allocate progressively
 * higher ids if it can. If the "cur" counter wraps, then it will start again
 * at the "start" end of the range and allocate one that has already been used.
 */
int idr_alloc_cyclic(struct idr *idr, void *ptr, int start, int end,
			gfp_t gfp_mask)
{
	int id;

	id = idr_alloc(idr, ptr, max(start, idr->cur), end, gfp_mask);
	if (id == -ENOSPC)
		id = idr_alloc(idr, ptr, start, end, gfp_mask);

	if (likely(id >= 0))
		idr->cur = id + 1;
	return id;
}
EXPORT_SYMBOL(idr_alloc_cyclic);

L
Linus Torvalds 已提交
504 505
static void idr_remove_warning(int id)
{
506
	WARN(1, "idr_remove called for id=%d which is not allocated.\n", id);
L
Linus Torvalds 已提交
507 508 509 510 511
}

static void sub_remove(struct idr *idp, int shift, int id)
{
	struct idr_layer *p = idp->top;
T
Tejun Heo 已提交
512
	struct idr_layer **pa[MAX_IDR_LEVEL + 1];
L
Linus Torvalds 已提交
513
	struct idr_layer ***paa = &pa[0];
N
Nadia Derbey 已提交
514
	struct idr_layer *to_free;
L
Linus Torvalds 已提交
515 516 517 518 519 520 521
	int n;

	*paa = NULL;
	*++paa = &idp->top;

	while ((shift > 0) && p) {
		n = (id >> shift) & IDR_MASK;
522
		__clear_bit(n, p->bitmap);
L
Linus Torvalds 已提交
523 524 525 526 527
		*++paa = &p->ary[n];
		p = p->ary[n];
		shift -= IDR_BITS;
	}
	n = id & IDR_MASK;
528 529
	if (likely(p != NULL && test_bit(n, p->bitmap))) {
		__clear_bit(n, p->bitmap);
530
		RCU_INIT_POINTER(p->ary[n], NULL);
N
Nadia Derbey 已提交
531
		to_free = NULL;
L
Linus Torvalds 已提交
532
		while(*paa && ! --((**paa)->count)){
N
Nadia Derbey 已提交
533
			if (to_free)
T
Tejun Heo 已提交
534
				free_layer(idp, to_free);
N
Nadia Derbey 已提交
535
			to_free = **paa;
L
Linus Torvalds 已提交
536 537
			**paa-- = NULL;
		}
538
		if (!*paa)
L
Linus Torvalds 已提交
539
			idp->layers = 0;
N
Nadia Derbey 已提交
540
		if (to_free)
T
Tejun Heo 已提交
541
			free_layer(idp, to_free);
542
	} else
L
Linus Torvalds 已提交
543 544 545 546
		idr_remove_warning(id);
}

/**
547
 * idr_remove - remove the given id and free its slot
548 549
 * @idp: idr handle
 * @id: unique key
L
Linus Torvalds 已提交
550 551 552 553
 */
void idr_remove(struct idr *idp, int id)
{
	struct idr_layer *p;
N
Nadia Derbey 已提交
554
	struct idr_layer *to_free;
L
Linus Torvalds 已提交
555

556
	if (id < 0)
557
		return;
L
Linus Torvalds 已提交
558

559 560 561 562 563
	if (id > idr_max(idp->layers)) {
		idr_remove_warning(id);
		return;
	}

L
Linus Torvalds 已提交
564
	sub_remove(idp, (idp->layers - 1) * IDR_BITS, id);
565
	if (idp->top && idp->top->count == 1 && (idp->layers > 1) &&
N
Nadia Derbey 已提交
566 567 568 569 570 571 572 573
	    idp->top->ary[0]) {
		/*
		 * Single child at leftmost slot: we can shrink the tree.
		 * This level is not needed anymore since when layers are
		 * inserted, they are inserted at the top of the existing
		 * tree.
		 */
		to_free = idp->top;
L
Linus Torvalds 已提交
574
		p = idp->top->ary[0];
N
Nadia Derbey 已提交
575
		rcu_assign_pointer(idp->top, p);
L
Linus Torvalds 已提交
576
		--idp->layers;
577 578
		to_free->count = 0;
		bitmap_clear(to_free->bitmap, 0, IDR_SIZE);
T
Tejun Heo 已提交
579
		free_layer(idp, to_free);
L
Linus Torvalds 已提交
580 581 582 583
	}
}
EXPORT_SYMBOL(idr_remove);

S
Stephen Hemminger 已提交
584
static void __idr_remove_all(struct idr *idp)
K
Kristian Hoegsberg 已提交
585
{
586
	int n, id, max;
587
	int bt_mask;
K
Kristian Hoegsberg 已提交
588
	struct idr_layer *p;
T
Tejun Heo 已提交
589
	struct idr_layer *pa[MAX_IDR_LEVEL + 1];
K
Kristian Hoegsberg 已提交
590 591 592 593
	struct idr_layer **paa = &pa[0];

	n = idp->layers * IDR_BITS;
	p = idp->top;
594
	RCU_INIT_POINTER(idp->top, NULL);
T
Tejun Heo 已提交
595
	max = idr_max(idp->layers);
K
Kristian Hoegsberg 已提交
596 597

	id = 0;
T
Tejun Heo 已提交
598
	while (id >= 0 && id <= max) {
K
Kristian Hoegsberg 已提交
599 600 601 602 603 604
		while (n > IDR_BITS && p) {
			n -= IDR_BITS;
			*paa++ = p;
			p = p->ary[(id >> n) & IDR_MASK];
		}

605
		bt_mask = id;
K
Kristian Hoegsberg 已提交
606
		id += 1 << n;
607 608
		/* Get the highest bit that the above add changed from 0->1. */
		while (n < fls(id ^ bt_mask)) {
N
Nadia Derbey 已提交
609
			if (p)
T
Tejun Heo 已提交
610
				free_layer(idp, p);
K
Kristian Hoegsberg 已提交
611 612 613 614 615 616 617
			n += IDR_BITS;
			p = *--paa;
		}
	}
	idp->layers = 0;
}

A
Andrew Morton 已提交
618 619
/**
 * idr_destroy - release all cached layers within an idr tree
N
Naohiro Aota 已提交
620
 * @idp: idr handle
621 622 623 624 625 626 627 628 629
 *
 * Free all id mappings and all idp_layers.  After this function, @idp is
 * completely unused and can be freed / recycled.  The caller is
 * responsible for ensuring that no one else accesses @idp during or after
 * idr_destroy().
 *
 * A typical clean-up sequence for objects stored in an idr tree will use
 * idr_for_each() to free all objects, if necessay, then idr_destroy() to
 * free up the id mappings and cached idr_layers.
A
Andrew Morton 已提交
630 631 632
 */
void idr_destroy(struct idr *idp)
{
T
Tejun Heo 已提交
633
	__idr_remove_all(idp);
634

A
Andrew Morton 已提交
635
	while (idp->id_free_cnt) {
636
		struct idr_layer *p = get_from_free_list(idp);
A
Andrew Morton 已提交
637 638 639 640 641
		kmem_cache_free(idr_layer_cache, p);
	}
}
EXPORT_SYMBOL(idr_destroy);

T
Tejun Heo 已提交
642
void *idr_find_slowpath(struct idr *idp, int id)
L
Linus Torvalds 已提交
643 644 645 646
{
	int n;
	struct idr_layer *p;

647
	if (id < 0)
648 649
		return NULL;

650
	p = rcu_dereference_raw(idp->top);
651 652 653
	if (!p)
		return NULL;
	n = (p->layer+1) * IDR_BITS;
L
Linus Torvalds 已提交
654

T
Tejun Heo 已提交
655
	if (id > idr_max(p->layer + 1))
L
Linus Torvalds 已提交
656
		return NULL;
657
	BUG_ON(n == 0);
L
Linus Torvalds 已提交
658 659 660

	while (n > 0 && p) {
		n -= IDR_BITS;
661
		BUG_ON(n != p->layer*IDR_BITS);
662
		p = rcu_dereference_raw(p->ary[(id >> n) & IDR_MASK]);
L
Linus Torvalds 已提交
663 664 665
	}
	return((void *)p);
}
T
Tejun Heo 已提交
666
EXPORT_SYMBOL(idr_find_slowpath);
L
Linus Torvalds 已提交
667

K
Kristian Hoegsberg 已提交
668 669 670 671 672 673 674 675 676 677 678 679 680 681
/**
 * idr_for_each - iterate through all stored pointers
 * @idp: idr handle
 * @fn: function to be called for each pointer
 * @data: data passed back to callback function
 *
 * Iterate over the pointers registered with the given idr.  The
 * callback function will be called for each pointer currently
 * registered, passing the id, the pointer and the data pointer passed
 * to this function.  It is not safe to modify the idr tree while in
 * the callback, so functions such as idr_get_new and idr_remove are
 * not allowed.
 *
 * We check the return of @fn each time. If it returns anything other
682
 * than %0, we break out and return that value.
K
Kristian Hoegsberg 已提交
683 684 685 686 687 688 689 690
 *
 * The caller must serialize idr_for_each() vs idr_get_new() and idr_remove().
 */
int idr_for_each(struct idr *idp,
		 int (*fn)(int id, void *p, void *data), void *data)
{
	int n, id, max, error = 0;
	struct idr_layer *p;
T
Tejun Heo 已提交
691
	struct idr_layer *pa[MAX_IDR_LEVEL + 1];
K
Kristian Hoegsberg 已提交
692 693 694
	struct idr_layer **paa = &pa[0];

	n = idp->layers * IDR_BITS;
695
	p = rcu_dereference_raw(idp->top);
T
Tejun Heo 已提交
696
	max = idr_max(idp->layers);
K
Kristian Hoegsberg 已提交
697 698

	id = 0;
T
Tejun Heo 已提交
699
	while (id >= 0 && id <= max) {
K
Kristian Hoegsberg 已提交
700 701 702
		while (n > 0 && p) {
			n -= IDR_BITS;
			*paa++ = p;
703
			p = rcu_dereference_raw(p->ary[(id >> n) & IDR_MASK]);
K
Kristian Hoegsberg 已提交
704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722
		}

		if (p) {
			error = fn(id, (void *)p, data);
			if (error)
				break;
		}

		id += 1 << n;
		while (n < fls(id)) {
			n += IDR_BITS;
			p = *--paa;
		}
	}

	return error;
}
EXPORT_SYMBOL(idr_for_each);

K
KAMEZAWA Hiroyuki 已提交
723 724 725
/**
 * idr_get_next - lookup next object of id to given id.
 * @idp: idr handle
N
Naohiro Aota 已提交
726
 * @nextidp:  pointer to lookup key
K
KAMEZAWA Hiroyuki 已提交
727 728
 *
 * Returns pointer to registered object with id, which is next number to
729 730
 * given id. After being looked up, *@nextidp will be updated for the next
 * iteration.
731 732 733
 *
 * This function can be called under rcu_read_lock(), given that the leaf
 * pointers lifetimes are correctly managed.
K
KAMEZAWA Hiroyuki 已提交
734 735 736
 */
void *idr_get_next(struct idr *idp, int *nextidp)
{
T
Tejun Heo 已提交
737
	struct idr_layer *p, *pa[MAX_IDR_LEVEL + 1];
K
KAMEZAWA Hiroyuki 已提交
738 739 740 741 742
	struct idr_layer **paa = &pa[0];
	int id = *nextidp;
	int n, max;

	/* find first ent */
743
	p = rcu_dereference_raw(idp->top);
K
KAMEZAWA Hiroyuki 已提交
744 745
	if (!p)
		return NULL;
746
	n = (p->layer + 1) * IDR_BITS;
T
Tejun Heo 已提交
747
	max = idr_max(p->layer + 1);
K
KAMEZAWA Hiroyuki 已提交
748

T
Tejun Heo 已提交
749
	while (id >= 0 && id <= max) {
K
KAMEZAWA Hiroyuki 已提交
750 751 752
		while (n > 0 && p) {
			n -= IDR_BITS;
			*paa++ = p;
753
			p = rcu_dereference_raw(p->ary[(id >> n) & IDR_MASK]);
K
KAMEZAWA Hiroyuki 已提交
754 755 756 757 758 759 760
		}

		if (p) {
			*nextidp = id;
			return p;
		}

761 762 763 764 765 766 767 768
		/*
		 * Proceed to the next layer at the current level.  Unlike
		 * idr_for_each(), @id isn't guaranteed to be aligned to
		 * layer boundary at this point and adding 1 << n may
		 * incorrectly skip IDs.  Make sure we jump to the
		 * beginning of the next layer using round_up().
		 */
		id = round_up(id + 1, 1 << n);
K
KAMEZAWA Hiroyuki 已提交
769 770 771 772 773 774 775
		while (n < fls(id)) {
			n += IDR_BITS;
			p = *--paa;
		}
	}
	return NULL;
}
B
Ben Hutchings 已提交
776
EXPORT_SYMBOL(idr_get_next);
K
KAMEZAWA Hiroyuki 已提交
777 778


J
Jeff Mahoney 已提交
779 780 781 782 783 784 785
/**
 * idr_replace - replace pointer for given id
 * @idp: idr handle
 * @ptr: pointer you want associated with the id
 * @id: lookup key
 *
 * Replace the pointer registered with an id and return the old value.
786 787
 * A %-ENOENT return indicates that @id was not found.
 * A %-EINVAL return indicates that @id was not within valid constraints.
J
Jeff Mahoney 已提交
788
 *
N
Nadia Derbey 已提交
789
 * The caller must serialize with writers.
J
Jeff Mahoney 已提交
790 791 792 793 794 795
 */
void *idr_replace(struct idr *idp, void *ptr, int id)
{
	int n;
	struct idr_layer *p, *old_p;

796
	if (id < 0)
797 798
		return ERR_PTR(-EINVAL);

J
Jeff Mahoney 已提交
799
	p = idp->top;
800
	if (!p)
801
		return ERR_PTR(-ENOENT);
802

803
	if (id > idr_max(p->layer + 1))
804
		return ERR_PTR(-ENOENT);
J
Jeff Mahoney 已提交
805

806
	n = p->layer * IDR_BITS;
J
Jeff Mahoney 已提交
807 808 809 810 811 812
	while ((n > 0) && p) {
		p = p->ary[(id >> n) & IDR_MASK];
		n -= IDR_BITS;
	}

	n = id & IDR_MASK;
813
	if (unlikely(p == NULL || !test_bit(n, p->bitmap)))
J
Jeff Mahoney 已提交
814 815 816
		return ERR_PTR(-ENOENT);

	old_p = p->ary[n];
N
Nadia Derbey 已提交
817
	rcu_assign_pointer(p->ary[n], ptr);
J
Jeff Mahoney 已提交
818 819 820 821 822

	return old_p;
}
EXPORT_SYMBOL(idr_replace);

823
void __init idr_init_cache(void)
L
Linus Torvalds 已提交
824
{
825
	idr_layer_cache = kmem_cache_create("idr_layer_cache",
826
				sizeof(struct idr_layer), 0, SLAB_PANIC, NULL);
L
Linus Torvalds 已提交
827 828 829 830 831 832 833 834 835 836 837 838 839 840 841
}

/**
 * idr_init - initialize idr handle
 * @idp:	idr handle
 *
 * This function is use to set up the handle (@idp) that you will pass
 * to the rest of the functions.
 */
void idr_init(struct idr *idp)
{
	memset(idp, 0, sizeof(struct idr));
	spin_lock_init(&idp->lock);
}
EXPORT_SYMBOL(idr_init);
842

843 844 845 846 847 848 849 850 851 852
static int idr_has_entry(int id, void *p, void *data)
{
	return 1;
}

bool idr_is_empty(struct idr *idp)
{
	return !idr_for_each(idp, idr_has_entry, NULL);
}
EXPORT_SYMBOL(idr_is_empty);
853

854 855
/**
 * DOC: IDA description
856 857
 * IDA - IDR based ID allocator
 *
858
 * This is id allocator without id -> pointer translation.  Memory
859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890
 * usage is much lower than full blown idr because each id only
 * occupies a bit.  ida uses a custom leaf node which contains
 * IDA_BITMAP_BITS slots.
 *
 * 2007-04-25  written by Tejun Heo <htejun@gmail.com>
 */

static void free_bitmap(struct ida *ida, struct ida_bitmap *bitmap)
{
	unsigned long flags;

	if (!ida->free_bitmap) {
		spin_lock_irqsave(&ida->idr.lock, flags);
		if (!ida->free_bitmap) {
			ida->free_bitmap = bitmap;
			bitmap = NULL;
		}
		spin_unlock_irqrestore(&ida->idr.lock, flags);
	}

	kfree(bitmap);
}

/**
 * ida_pre_get - reserve resources for ida allocation
 * @ida:	ida handle
 * @gfp_mask:	memory allocation flag
 *
 * This function should be called prior to locking and calling the
 * following function.  It preallocates enough memory to satisfy the
 * worst possible allocation.
 *
891 892
 * If the system is REALLY out of memory this function returns %0,
 * otherwise %1.
893 894 895 896
 */
int ida_pre_get(struct ida *ida, gfp_t gfp_mask)
{
	/* allocate idr_layers */
897
	if (!__idr_pre_get(&ida->idr, gfp_mask))
898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917
		return 0;

	/* allocate free_bitmap */
	if (!ida->free_bitmap) {
		struct ida_bitmap *bitmap;

		bitmap = kmalloc(sizeof(struct ida_bitmap), gfp_mask);
		if (!bitmap)
			return 0;

		free_bitmap(ida, bitmap);
	}

	return 1;
}
EXPORT_SYMBOL(ida_pre_get);

/**
 * ida_get_new_above - allocate new ID above or equal to a start id
 * @ida:	ida handle
N
Naohiro Aota 已提交
918
 * @starting_id: id to start search at
919 920
 * @p_id:	pointer to the allocated handle
 *
921 922
 * Allocate new ID above or equal to @starting_id.  It should be called
 * with any required locks.
923
 *
924
 * If memory is required, it will return %-EAGAIN, you should unlock
925
 * and go back to the ida_pre_get() call.  If the ida is full, it will
926
 * return %-ENOSPC.
927
 *
928
 * @p_id returns a value in the range @starting_id ... %0x7fffffff.
929 930 931
 */
int ida_get_new_above(struct ida *ida, int starting_id, int *p_id)
{
T
Tejun Heo 已提交
932
	struct idr_layer *pa[MAX_IDR_LEVEL + 1];
933 934 935 936 937 938 939 940
	struct ida_bitmap *bitmap;
	unsigned long flags;
	int idr_id = starting_id / IDA_BITMAP_BITS;
	int offset = starting_id % IDA_BITMAP_BITS;
	int t, id;

 restart:
	/* get vacant slot */
941
	t = idr_get_empty_slot(&ida->idr, idr_id, pa, 0, &ida->idr);
N
Nadia Derbey 已提交
942
	if (t < 0)
T
Tejun Heo 已提交
943
		return t == -ENOMEM ? -EAGAIN : t;
944

945
	if (t * IDA_BITMAP_BITS >= MAX_IDR_BIT)
946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963
		return -ENOSPC;

	if (t != idr_id)
		offset = 0;
	idr_id = t;

	/* if bitmap isn't there, create a new one */
	bitmap = (void *)pa[0]->ary[idr_id & IDR_MASK];
	if (!bitmap) {
		spin_lock_irqsave(&ida->idr.lock, flags);
		bitmap = ida->free_bitmap;
		ida->free_bitmap = NULL;
		spin_unlock_irqrestore(&ida->idr.lock, flags);

		if (!bitmap)
			return -EAGAIN;

		memset(bitmap, 0, sizeof(struct ida_bitmap));
N
Nadia Derbey 已提交
964 965
		rcu_assign_pointer(pa[0]->ary[idr_id & IDR_MASK],
				(void *)bitmap);
966 967 968 969 970 971 972 973 974 975 976 977 978
		pa[0]->count++;
	}

	/* lookup for empty slot */
	t = find_next_zero_bit(bitmap->bitmap, IDA_BITMAP_BITS, offset);
	if (t == IDA_BITMAP_BITS) {
		/* no empty slot after offset, continue to the next chunk */
		idr_id++;
		offset = 0;
		goto restart;
	}

	id = idr_id * IDA_BITMAP_BITS + t;
979
	if (id >= MAX_IDR_BIT)
980 981 982 983 984 985 986 987 988 989 990 991 992 993
		return -ENOSPC;

	__set_bit(t, bitmap->bitmap);
	if (++bitmap->nr_busy == IDA_BITMAP_BITS)
		idr_mark_full(pa, idr_id);

	*p_id = id;

	/* Each leaf node can handle nearly a thousand slots and the
	 * whole idea of ida is to have small memory foot print.
	 * Throw away extra resources one by one after each successful
	 * allocation.
	 */
	if (ida->idr.id_free_cnt || ida->free_bitmap) {
994
		struct idr_layer *p = get_from_free_list(&ida->idr);
995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016
		if (p)
			kmem_cache_free(idr_layer_cache, p);
	}

	return 0;
}
EXPORT_SYMBOL(ida_get_new_above);

/**
 * ida_remove - remove the given ID
 * @ida:	ida handle
 * @id:		ID to free
 */
void ida_remove(struct ida *ida, int id)
{
	struct idr_layer *p = ida->idr.top;
	int shift = (ida->idr.layers - 1) * IDR_BITS;
	int idr_id = id / IDA_BITMAP_BITS;
	int offset = id % IDA_BITMAP_BITS;
	int n;
	struct ida_bitmap *bitmap;

1017 1018 1019
	if (idr_id > idr_max(ida->idr.layers))
		goto err;

1020 1021 1022
	/* clear full bits while looking up the leaf idr_layer */
	while ((shift > 0) && p) {
		n = (idr_id >> shift) & IDR_MASK;
1023
		__clear_bit(n, p->bitmap);
1024 1025 1026 1027 1028 1029 1030 1031
		p = p->ary[n];
		shift -= IDR_BITS;
	}

	if (p == NULL)
		goto err;

	n = idr_id & IDR_MASK;
1032
	__clear_bit(n, p->bitmap);
1033 1034

	bitmap = (void *)p->ary[n];
1035
	if (!bitmap || !test_bit(offset, bitmap->bitmap))
1036 1037 1038 1039 1040
		goto err;

	/* update bitmap and remove it if empty */
	__clear_bit(offset, bitmap->bitmap);
	if (--bitmap->nr_busy == 0) {
1041
		__set_bit(n, p->bitmap);	/* to please idr_remove() */
1042 1043 1044 1045 1046 1047 1048
		idr_remove(&ida->idr, idr_id);
		free_bitmap(ida, bitmap);
	}

	return;

 err:
1049
	WARN(1, "ida_remove called for id=%d which is not allocated.\n", id);
1050 1051 1052 1053 1054
}
EXPORT_SYMBOL(ida_remove);

/**
 * ida_destroy - release all cached layers within an ida tree
N
Naohiro Aota 已提交
1055
 * @ida:		ida handle
1056 1057 1058 1059 1060 1061 1062 1063
 */
void ida_destroy(struct ida *ida)
{
	idr_destroy(&ida->idr);
	kfree(ida->free_bitmap);
}
EXPORT_SYMBOL(ida_destroy);

1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080
/**
 * ida_simple_get - get a new id.
 * @ida: the (initialized) ida.
 * @start: the minimum id (inclusive, < 0x8000000)
 * @end: the maximum id (exclusive, < 0x8000000 or 0)
 * @gfp_mask: memory allocation flags
 *
 * Allocates an id in the range start <= id < end, or returns -ENOSPC.
 * On memory allocation failure, returns -ENOMEM.
 *
 * Use ida_simple_remove() to get rid of an id.
 */
int ida_simple_get(struct ida *ida, unsigned int start, unsigned int end,
		   gfp_t gfp_mask)
{
	int ret, id;
	unsigned int max;
1081
	unsigned long flags;
1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096

	BUG_ON((int)start < 0);
	BUG_ON((int)end < 0);

	if (end == 0)
		max = 0x80000000;
	else {
		BUG_ON(end < start);
		max = end - 1;
	}

again:
	if (!ida_pre_get(ida, gfp_mask))
		return -ENOMEM;

1097
	spin_lock_irqsave(&simple_ida_lock, flags);
1098 1099 1100 1101 1102 1103 1104 1105 1106
	ret = ida_get_new_above(ida, start, &id);
	if (!ret) {
		if (id > max) {
			ida_remove(ida, id);
			ret = -ENOSPC;
		} else {
			ret = id;
		}
	}
1107
	spin_unlock_irqrestore(&simple_ida_lock, flags);
1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122

	if (unlikely(ret == -EAGAIN))
		goto again;

	return ret;
}
EXPORT_SYMBOL(ida_simple_get);

/**
 * ida_simple_remove - remove an allocated id.
 * @ida: the (initialized) ida.
 * @id: the id returned by ida_simple_get.
 */
void ida_simple_remove(struct ida *ida, unsigned int id)
{
1123 1124
	unsigned long flags;

1125
	BUG_ON((int)id < 0);
1126
	spin_lock_irqsave(&simple_ida_lock, flags);
1127
	ida_remove(ida, id);
1128
	spin_unlock_irqrestore(&simple_ida_lock, flags);
1129 1130 1131
}
EXPORT_SYMBOL(ida_simple_remove);

1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145
/**
 * ida_init - initialize ida handle
 * @ida:	ida handle
 *
 * This function is use to set up the handle (@ida) that you will pass
 * to the rest of the functions.
 */
void ida_init(struct ida *ida)
{
	memset(ida, 0, sizeof(struct ida));
	idr_init(&ida->idr);

}
EXPORT_SYMBOL(ida_init);