inotify: fix race
[linux-2.6.22.y-op.git] / mm / slob.c
blob71976c5d40d301373ea9bc3f807553dde819c195
1 /*
2 * SLOB Allocator: Simple List Of Blocks
4 * Matt Mackall <mpm@selenic.com> 12/30/03
6 * How SLOB works:
8 * The core of SLOB is a traditional K&R style heap allocator, with
9 * support for returning aligned objects. The granularity of this
10 * allocator is 8 bytes on x86, though it's perhaps possible to reduce
11 * this to 4 if it's deemed worth the effort. The slob heap is a
12 * singly-linked list of pages from __get_free_page, grown on demand
13 * and allocation from the heap is currently first-fit.
15 * Above this is an implementation of kmalloc/kfree. Blocks returned
16 * from kmalloc are 8-byte aligned and prepended with a 8-byte header.
17 * If kmalloc is asked for objects of PAGE_SIZE or larger, it calls
18 * __get_free_pages directly so that it can return page-aligned blocks
19 * and keeps a linked list of such pages and their orders. These
20 * objects are detected in kfree() by their page alignment.
22 * SLAB is emulated on top of SLOB by simply calling constructors and
23 * destructors for every SLAB allocation. Objects are returned with
24 * the 8-byte alignment unless the SLAB_HWCACHE_ALIGN flag is
25 * set, in which case the low-level allocator will fragment blocks to
26 * create the proper alignment. Again, objects of page-size or greater
27 * are allocated by calling __get_free_pages. As SLAB objects know
28 * their size, no separate size bookkeeping is necessary and there is
29 * essentially no allocation space overhead.
32 #include <linux/slab.h>
33 #include <linux/mm.h>
34 #include <linux/cache.h>
35 #include <linux/init.h>
36 #include <linux/module.h>
37 #include <linux/timer.h>
38 #include <linux/rcupdate.h>
40 struct slob_block {
41 int units;
42 struct slob_block *next;
44 typedef struct slob_block slob_t;
46 #define SLOB_UNIT sizeof(slob_t)
47 #define SLOB_UNITS(size) (((size) + SLOB_UNIT - 1)/SLOB_UNIT)
48 #define SLOB_ALIGN L1_CACHE_BYTES
50 struct bigblock {
51 int order;
52 void *pages;
53 struct bigblock *next;
55 typedef struct bigblock bigblock_t;
58 * struct slob_rcu is inserted at the tail of allocated slob blocks, which
59 * were created with a SLAB_DESTROY_BY_RCU slab. slob_rcu is used to free
60 * the block using call_rcu.
62 struct slob_rcu {
63 struct rcu_head head;
64 int size;
67 static slob_t arena = { .next = &arena, .units = 1 };
68 static slob_t *slobfree = &arena;
69 static bigblock_t *bigblocks;
70 static DEFINE_SPINLOCK(slob_lock);
71 static DEFINE_SPINLOCK(block_lock);
73 static void slob_free(void *b, int size);
74 static void slob_timer_cbk(void);
77 static void *slob_alloc(size_t size, gfp_t gfp, int align)
79 slob_t *prev, *cur, *aligned = 0;
80 int delta = 0, units = SLOB_UNITS(size);
81 unsigned long flags;
83 spin_lock_irqsave(&slob_lock, flags);
84 prev = slobfree;
85 for (cur = prev->next; ; prev = cur, cur = cur->next) {
86 if (align) {
87 aligned = (slob_t *)ALIGN((unsigned long)cur, align);
88 delta = aligned - cur;
90 if (cur->units >= units + delta) { /* room enough? */
91 if (delta) { /* need to fragment head to align? */
92 aligned->units = cur->units - delta;
93 aligned->next = cur->next;
94 cur->next = aligned;
95 cur->units = delta;
96 prev = cur;
97 cur = aligned;
100 if (cur->units == units) /* exact fit? */
101 prev->next = cur->next; /* unlink */
102 else { /* fragment */
103 prev->next = cur + units;
104 prev->next->units = cur->units - units;
105 prev->next->next = cur->next;
106 cur->units = units;
109 slobfree = prev;
110 spin_unlock_irqrestore(&slob_lock, flags);
111 return cur;
113 if (cur == slobfree) {
114 spin_unlock_irqrestore(&slob_lock, flags);
116 if (size == PAGE_SIZE) /* trying to shrink arena? */
117 return 0;
119 cur = (slob_t *)__get_free_page(gfp);
120 if (!cur)
121 return 0;
123 slob_free(cur, PAGE_SIZE);
124 spin_lock_irqsave(&slob_lock, flags);
125 cur = slobfree;
130 static void slob_free(void *block, int size)
132 slob_t *cur, *b = (slob_t *)block;
133 unsigned long flags;
135 if (!block)
136 return;
138 if (size)
139 b->units = SLOB_UNITS(size);
141 /* Find reinsertion point */
142 spin_lock_irqsave(&slob_lock, flags);
143 for (cur = slobfree; !(b > cur && b < cur->next); cur = cur->next)
144 if (cur >= cur->next && (b > cur || b < cur->next))
145 break;
147 if (b + b->units == cur->next) {
148 b->units += cur->next->units;
149 b->next = cur->next->next;
150 } else
151 b->next = cur->next;
153 if (cur + cur->units == b) {
154 cur->units += b->units;
155 cur->next = b->next;
156 } else
157 cur->next = b;
159 slobfree = cur;
161 spin_unlock_irqrestore(&slob_lock, flags);
164 void *__kmalloc(size_t size, gfp_t gfp)
166 slob_t *m;
167 bigblock_t *bb;
168 unsigned long flags;
170 if (size < PAGE_SIZE - SLOB_UNIT) {
171 m = slob_alloc(size + SLOB_UNIT, gfp, 0);
172 return m ? (void *)(m + 1) : 0;
175 bb = slob_alloc(sizeof(bigblock_t), gfp, 0);
176 if (!bb)
177 return 0;
179 bb->order = get_order(size);
180 bb->pages = (void *)__get_free_pages(gfp, bb->order);
182 if (bb->pages) {
183 spin_lock_irqsave(&block_lock, flags);
184 bb->next = bigblocks;
185 bigblocks = bb;
186 spin_unlock_irqrestore(&block_lock, flags);
187 return bb->pages;
190 slob_free(bb, sizeof(bigblock_t));
191 return 0;
193 EXPORT_SYMBOL(__kmalloc);
196 * krealloc - reallocate memory. The contents will remain unchanged.
198 * @p: object to reallocate memory for.
199 * @new_size: how many bytes of memory are required.
200 * @flags: the type of memory to allocate.
202 * The contents of the object pointed to are preserved up to the
203 * lesser of the new and old sizes. If @p is %NULL, krealloc()
204 * behaves exactly like kmalloc(). If @size is 0 and @p is not a
205 * %NULL pointer, the object pointed to is freed.
207 void *krealloc(const void *p, size_t new_size, gfp_t flags)
209 void *ret;
211 if (unlikely(!p))
212 return kmalloc_track_caller(new_size, flags);
214 if (unlikely(!new_size)) {
215 kfree(p);
216 return NULL;
219 ret = kmalloc_track_caller(new_size, flags);
220 if (ret) {
221 memcpy(ret, p, min(new_size, ksize(p)));
222 kfree(p);
224 return ret;
226 EXPORT_SYMBOL(krealloc);
228 void kfree(const void *block)
230 bigblock_t *bb, **last = &bigblocks;
231 unsigned long flags;
233 if (!block)
234 return;
236 if (!((unsigned long)block & (PAGE_SIZE-1))) {
237 /* might be on the big block list */
238 spin_lock_irqsave(&block_lock, flags);
239 for (bb = bigblocks; bb; last = &bb->next, bb = bb->next) {
240 if (bb->pages == block) {
241 *last = bb->next;
242 spin_unlock_irqrestore(&block_lock, flags);
243 free_pages((unsigned long)block, bb->order);
244 slob_free(bb, sizeof(bigblock_t));
245 return;
248 spin_unlock_irqrestore(&block_lock, flags);
251 slob_free((slob_t *)block - 1, 0);
252 return;
255 EXPORT_SYMBOL(kfree);
257 size_t ksize(const void *block)
259 bigblock_t *bb;
260 unsigned long flags;
262 if (!block)
263 return 0;
265 if (!((unsigned long)block & (PAGE_SIZE-1))) {
266 spin_lock_irqsave(&block_lock, flags);
267 for (bb = bigblocks; bb; bb = bb->next)
268 if (bb->pages == block) {
269 spin_unlock_irqrestore(&slob_lock, flags);
270 return PAGE_SIZE << bb->order;
272 spin_unlock_irqrestore(&block_lock, flags);
275 return ((slob_t *)block - 1)->units * SLOB_UNIT;
278 struct kmem_cache {
279 unsigned int size, align;
280 unsigned long flags;
281 const char *name;
282 void (*ctor)(void *, struct kmem_cache *, unsigned long);
285 struct kmem_cache *kmem_cache_create(const char *name, size_t size,
286 size_t align, unsigned long flags,
287 void (*ctor)(void*, struct kmem_cache *, unsigned long),
288 void (*dtor)(void*, struct kmem_cache *, unsigned long))
290 struct kmem_cache *c;
292 c = slob_alloc(sizeof(struct kmem_cache), flags, 0);
294 if (c) {
295 c->name = name;
296 c->size = size;
297 if (flags & SLAB_DESTROY_BY_RCU) {
298 /* leave room for rcu footer at the end of object */
299 c->size += sizeof(struct slob_rcu);
301 c->flags = flags;
302 c->ctor = ctor;
303 /* ignore alignment unless it's forced */
304 c->align = (flags & SLAB_HWCACHE_ALIGN) ? SLOB_ALIGN : 0;
305 if (c->align < align)
306 c->align = align;
307 } else if (flags & SLAB_PANIC)
308 panic("Cannot create slab cache %s\n", name);
310 return c;
312 EXPORT_SYMBOL(kmem_cache_create);
314 void kmem_cache_destroy(struct kmem_cache *c)
316 slob_free(c, sizeof(struct kmem_cache));
318 EXPORT_SYMBOL(kmem_cache_destroy);
320 void *kmem_cache_alloc(struct kmem_cache *c, gfp_t flags)
322 void *b;
324 if (c->size < PAGE_SIZE)
325 b = slob_alloc(c->size, flags, c->align);
326 else
327 b = (void *)__get_free_pages(flags, get_order(c->size));
329 if (c->ctor)
330 c->ctor(b, c, 0);
332 return b;
334 EXPORT_SYMBOL(kmem_cache_alloc);
336 void *kmem_cache_zalloc(struct kmem_cache *c, gfp_t flags)
338 void *ret = kmem_cache_alloc(c, flags);
339 if (ret)
340 memset(ret, 0, c->size);
342 return ret;
344 EXPORT_SYMBOL(kmem_cache_zalloc);
346 static void __kmem_cache_free(void *b, int size)
348 if (size < PAGE_SIZE)
349 slob_free(b, size);
350 else
351 free_pages((unsigned long)b, get_order(size));
354 static void kmem_rcu_free(struct rcu_head *head)
356 struct slob_rcu *slob_rcu = (struct slob_rcu *)head;
357 void *b = (void *)slob_rcu - (slob_rcu->size - sizeof(struct slob_rcu));
359 __kmem_cache_free(b, slob_rcu->size);
362 void kmem_cache_free(struct kmem_cache *c, void *b)
364 if (unlikely(c->flags & SLAB_DESTROY_BY_RCU)) {
365 struct slob_rcu *slob_rcu;
366 slob_rcu = b + (c->size - sizeof(struct slob_rcu));
367 INIT_RCU_HEAD(&slob_rcu->head);
368 slob_rcu->size = c->size;
369 call_rcu(&slob_rcu->head, kmem_rcu_free);
370 } else {
371 __kmem_cache_free(b, c->size);
374 EXPORT_SYMBOL(kmem_cache_free);
376 unsigned int kmem_cache_size(struct kmem_cache *c)
378 return c->size;
380 EXPORT_SYMBOL(kmem_cache_size);
382 const char *kmem_cache_name(struct kmem_cache *c)
384 return c->name;
386 EXPORT_SYMBOL(kmem_cache_name);
388 static struct timer_list slob_timer = TIMER_INITIALIZER(
389 (void (*)(unsigned long))slob_timer_cbk, 0, 0);
391 int kmem_cache_shrink(struct kmem_cache *d)
393 return 0;
395 EXPORT_SYMBOL(kmem_cache_shrink);
397 int kmem_ptr_validate(struct kmem_cache *a, const void *b)
399 return 0;
402 void __init kmem_cache_init(void)
404 slob_timer_cbk();
407 static void slob_timer_cbk(void)
409 void *p = slob_alloc(PAGE_SIZE, 0, PAGE_SIZE-1);
411 if (p)
412 free_page((unsigned long)p);
414 mod_timer(&slob_timer, jiffies + HZ);