Commit | Line | Data |
---|---|---|
6387a3c4 A |
1 | // SPDX-License-Identifier: MIT |
2 | /* | |
3 | * Copyright © 2021 Intel Corporation | |
4 | */ | |
5 | ||
6 | #include <linux/kmemleak.h> | |
7 | #include <linux/module.h> | |
8 | #include <linux/sizes.h> | |
9 | ||
10 | #include <drm/drm_buddy.h> | |
11 | ||
12 | static struct kmem_cache *slab_blocks; | |
13 | ||
14 | static struct drm_buddy_block *drm_block_alloc(struct drm_buddy *mm, | |
15 | struct drm_buddy_block *parent, | |
16 | unsigned int order, | |
17 | u64 offset) | |
18 | { | |
19 | struct drm_buddy_block *block; | |
20 | ||
21 | BUG_ON(order > DRM_BUDDY_MAX_ORDER); | |
22 | ||
23 | block = kmem_cache_zalloc(slab_blocks, GFP_KERNEL); | |
24 | if (!block) | |
25 | return NULL; | |
26 | ||
27 | block->header = offset; | |
28 | block->header |= order; | |
29 | block->parent = parent; | |
30 | ||
31 | BUG_ON(block->header & DRM_BUDDY_HEADER_UNUSED); | |
32 | return block; | |
33 | } | |
34 | ||
35 | static void drm_block_free(struct drm_buddy *mm, | |
36 | struct drm_buddy_block *block) | |
37 | { | |
38 | kmem_cache_free(slab_blocks, block); | |
39 | } | |
40 | ||
5640e816 APS |
41 | static void list_insert_sorted(struct drm_buddy *mm, |
42 | struct drm_buddy_block *block) | |
43 | { | |
44 | struct drm_buddy_block *node; | |
45 | struct list_head *head; | |
46 | ||
47 | head = &mm->free_list[drm_buddy_block_order(block)]; | |
48 | if (list_empty(head)) { | |
49 | list_add(&block->link, head); | |
50 | return; | |
51 | } | |
52 | ||
53 | list_for_each_entry(node, head, link) | |
54 | if (drm_buddy_block_offset(block) < drm_buddy_block_offset(node)) | |
55 | break; | |
56 | ||
57 | __list_add(&block->link, node->link.prev, &node->link); | |
58 | } | |
59 | ||
96950929 APS |
60 | static void clear_reset(struct drm_buddy_block *block) |
61 | { | |
62 | block->header &= ~DRM_BUDDY_HEADER_CLEAR; | |
63 | } | |
64 | ||
65 | static void mark_cleared(struct drm_buddy_block *block) | |
66 | { | |
67 | block->header |= DRM_BUDDY_HEADER_CLEAR; | |
68 | } | |
69 | ||
6387a3c4 A |
70 | static void mark_allocated(struct drm_buddy_block *block) |
71 | { | |
72 | block->header &= ~DRM_BUDDY_HEADER_STATE; | |
73 | block->header |= DRM_BUDDY_ALLOCATED; | |
74 | ||
75 | list_del(&block->link); | |
76 | } | |
77 | ||
78 | static void mark_free(struct drm_buddy *mm, | |
79 | struct drm_buddy_block *block) | |
80 | { | |
81 | block->header &= ~DRM_BUDDY_HEADER_STATE; | |
82 | block->header |= DRM_BUDDY_FREE; | |
83 | ||
5640e816 | 84 | list_insert_sorted(mm, block); |
6387a3c4 A |
85 | } |
86 | ||
87 | static void mark_split(struct drm_buddy_block *block) | |
88 | { | |
89 | block->header &= ~DRM_BUDDY_HEADER_STATE; | |
90 | block->header |= DRM_BUDDY_SPLIT; | |
91 | ||
92 | list_del(&block->link); | |
93 | } | |
94 | ||
96950929 APS |
95 | static inline bool overlaps(u64 s1, u64 e1, u64 s2, u64 e2) |
96 | { | |
97 | return s1 <= e2 && e1 >= s2; | |
98 | } | |
99 | ||
100 | static inline bool contains(u64 s1, u64 e1, u64 s2, u64 e2) | |
101 | { | |
102 | return s1 <= s2 && e1 >= e2; | |
103 | } | |
104 | ||
105 | static struct drm_buddy_block * | |
106 | __get_buddy(struct drm_buddy_block *block) | |
107 | { | |
108 | struct drm_buddy_block *parent; | |
109 | ||
110 | parent = block->parent; | |
111 | if (!parent) | |
112 | return NULL; | |
113 | ||
114 | if (parent->left == block) | |
115 | return parent->right; | |
116 | ||
117 | return parent->left; | |
118 | } | |
119 | ||
120 | static unsigned int __drm_buddy_free(struct drm_buddy *mm, | |
121 | struct drm_buddy_block *block, | |
122 | bool force_merge) | |
123 | { | |
124 | struct drm_buddy_block *parent; | |
125 | unsigned int order; | |
126 | ||
127 | while ((parent = block->parent)) { | |
128 | struct drm_buddy_block *buddy; | |
129 | ||
130 | buddy = __get_buddy(block); | |
131 | ||
132 | if (!drm_buddy_block_is_free(buddy)) | |
133 | break; | |
134 | ||
135 | if (!force_merge) { | |
136 | /* | |
137 | * Check the block and its buddy clear state and exit | |
138 | * the loop if they both have the dissimilar state. | |
139 | */ | |
140 | if (drm_buddy_block_is_clear(block) != | |
141 | drm_buddy_block_is_clear(buddy)) | |
142 | break; | |
143 | ||
144 | if (drm_buddy_block_is_clear(block)) | |
145 | mark_cleared(parent); | |
146 | } | |
147 | ||
148 | list_del(&buddy->link); | |
149 | if (force_merge && drm_buddy_block_is_clear(buddy)) | |
150 | mm->clear_avail -= drm_buddy_block_size(mm, buddy); | |
151 | ||
152 | drm_block_free(mm, block); | |
153 | drm_block_free(mm, buddy); | |
154 | ||
155 | block = parent; | |
156 | } | |
157 | ||
158 | order = drm_buddy_block_order(block); | |
159 | mark_free(mm, block); | |
160 | ||
161 | return order; | |
162 | } | |
163 | ||
164 | static int __force_merge(struct drm_buddy *mm, | |
165 | u64 start, | |
166 | u64 end, | |
167 | unsigned int min_order) | |
168 | { | |
169 | unsigned int order; | |
170 | int i; | |
171 | ||
172 | if (!min_order) | |
173 | return -ENOMEM; | |
174 | ||
175 | if (min_order > mm->max_order) | |
176 | return -EINVAL; | |
177 | ||
178 | for (i = min_order - 1; i >= 0; i--) { | |
179 | struct drm_buddy_block *block, *prev; | |
180 | ||
181 | list_for_each_entry_safe_reverse(block, prev, &mm->free_list[i], link) { | |
182 | struct drm_buddy_block *buddy; | |
183 | u64 block_start, block_end; | |
184 | ||
185 | if (!block->parent) | |
186 | continue; | |
187 | ||
188 | block_start = drm_buddy_block_offset(block); | |
189 | block_end = block_start + drm_buddy_block_size(mm, block) - 1; | |
190 | ||
191 | if (!contains(start, end, block_start, block_end)) | |
192 | continue; | |
193 | ||
194 | buddy = __get_buddy(block); | |
195 | if (!drm_buddy_block_is_free(buddy)) | |
196 | continue; | |
197 | ||
198 | WARN_ON(drm_buddy_block_is_clear(block) == | |
199 | drm_buddy_block_is_clear(buddy)); | |
200 | ||
201 | /* | |
202 | * If the prev block is same as buddy, don't access the | |
203 | * block in the next iteration as we would free the | |
204 | * buddy block as part of the free function. | |
205 | */ | |
206 | if (prev == buddy) | |
207 | prev = list_prev_entry(prev, link); | |
208 | ||
209 | list_del(&block->link); | |
210 | if (drm_buddy_block_is_clear(block)) | |
211 | mm->clear_avail -= drm_buddy_block_size(mm, block); | |
212 | ||
213 | order = __drm_buddy_free(mm, block, true); | |
214 | if (order >= min_order) | |
215 | return 0; | |
216 | } | |
217 | } | |
218 | ||
219 | return -ENOMEM; | |
220 | } | |
221 | ||
6387a3c4 A |
222 | /** |
223 | * drm_buddy_init - init memory manager | |
224 | * | |
225 | * @mm: DRM buddy manager to initialize | |
226 | * @size: size in bytes to manage | |
227 | * @chunk_size: minimum page size in bytes for our allocations | |
228 | * | |
229 | * Initializes the memory manager and its resources. | |
230 | * | |
231 | * Returns: | |
232 | * 0 on success, error code on failure. | |
233 | */ | |
234 | int drm_buddy_init(struct drm_buddy *mm, u64 size, u64 chunk_size) | |
235 | { | |
236 | unsigned int i; | |
237 | u64 offset; | |
238 | ||
239 | if (size < chunk_size) | |
240 | return -EINVAL; | |
241 | ||
242 | if (chunk_size < PAGE_SIZE) | |
243 | return -EINVAL; | |
244 | ||
245 | if (!is_power_of_2(chunk_size)) | |
246 | return -EINVAL; | |
247 | ||
248 | size = round_down(size, chunk_size); | |
249 | ||
250 | mm->size = size; | |
251 | mm->avail = size; | |
bb21700b | 252 | mm->clear_avail = 0; |
6387a3c4 A |
253 | mm->chunk_size = chunk_size; |
254 | mm->max_order = ilog2(size) - ilog2(chunk_size); | |
255 | ||
256 | BUG_ON(mm->max_order > DRM_BUDDY_MAX_ORDER); | |
257 | ||
258 | mm->free_list = kmalloc_array(mm->max_order + 1, | |
259 | sizeof(struct list_head), | |
260 | GFP_KERNEL); | |
261 | if (!mm->free_list) | |
262 | return -ENOMEM; | |
263 | ||
264 | for (i = 0; i <= mm->max_order; ++i) | |
265 | INIT_LIST_HEAD(&mm->free_list[i]); | |
266 | ||
267 | mm->n_roots = hweight64(size); | |
268 | ||
269 | mm->roots = kmalloc_array(mm->n_roots, | |
270 | sizeof(struct drm_buddy_block *), | |
271 | GFP_KERNEL); | |
272 | if (!mm->roots) | |
273 | goto out_free_list; | |
274 | ||
275 | offset = 0; | |
276 | i = 0; | |
277 | ||
278 | /* | |
279 | * Split into power-of-two blocks, in case we are given a size that is | |
280 | * not itself a power-of-two. | |
281 | */ | |
282 | do { | |
283 | struct drm_buddy_block *root; | |
284 | unsigned int order; | |
285 | u64 root_size; | |
286 | ||
4453545b DG |
287 | order = ilog2(size) - ilog2(chunk_size); |
288 | root_size = chunk_size << order; | |
6387a3c4 A |
289 | |
290 | root = drm_block_alloc(mm, NULL, order, offset); | |
291 | if (!root) | |
292 | goto out_free_roots; | |
293 | ||
294 | mark_free(mm, root); | |
295 | ||
296 | BUG_ON(i > mm->max_order); | |
297 | BUG_ON(drm_buddy_block_size(mm, root) < chunk_size); | |
298 | ||
299 | mm->roots[i] = root; | |
300 | ||
301 | offset += root_size; | |
302 | size -= root_size; | |
303 | i++; | |
304 | } while (size); | |
305 | ||
306 | return 0; | |
307 | ||
308 | out_free_roots: | |
309 | while (i--) | |
310 | drm_block_free(mm, mm->roots[i]); | |
311 | kfree(mm->roots); | |
312 | out_free_list: | |
313 | kfree(mm->free_list); | |
314 | return -ENOMEM; | |
315 | } | |
316 | EXPORT_SYMBOL(drm_buddy_init); | |
317 | ||
318 | /** | |
319 | * drm_buddy_fini - tear down the memory manager | |
320 | * | |
321 | * @mm: DRM buddy manager to free | |
322 | * | |
323 | * Cleanup memory manager resources and the freelist | |
324 | */ | |
325 | void drm_buddy_fini(struct drm_buddy *mm) | |
326 | { | |
96950929 APS |
327 | u64 root_size, size; |
328 | unsigned int order; | |
6387a3c4 A |
329 | int i; |
330 | ||
96950929 APS |
331 | size = mm->size; |
332 | ||
6387a3c4 | 333 | for (i = 0; i < mm->n_roots; ++i) { |
96950929 APS |
334 | order = ilog2(size) - ilog2(mm->chunk_size); |
335 | __force_merge(mm, 0, size, order); | |
336 | ||
6387a3c4 A |
337 | WARN_ON(!drm_buddy_block_is_free(mm->roots[i])); |
338 | drm_block_free(mm, mm->roots[i]); | |
96950929 APS |
339 | |
340 | root_size = mm->chunk_size << order; | |
341 | size -= root_size; | |
6387a3c4 A |
342 | } |
343 | ||
344 | WARN_ON(mm->avail != mm->size); | |
345 | ||
346 | kfree(mm->roots); | |
347 | kfree(mm->free_list); | |
348 | } | |
349 | EXPORT_SYMBOL(drm_buddy_fini); | |
350 | ||
351 | static int split_block(struct drm_buddy *mm, | |
352 | struct drm_buddy_block *block) | |
353 | { | |
354 | unsigned int block_order = drm_buddy_block_order(block) - 1; | |
355 | u64 offset = drm_buddy_block_offset(block); | |
356 | ||
357 | BUG_ON(!drm_buddy_block_is_free(block)); | |
358 | BUG_ON(!drm_buddy_block_order(block)); | |
359 | ||
360 | block->left = drm_block_alloc(mm, block, block_order, offset); | |
361 | if (!block->left) | |
362 | return -ENOMEM; | |
363 | ||
364 | block->right = drm_block_alloc(mm, block, block_order, | |
365 | offset + (mm->chunk_size << block_order)); | |
366 | if (!block->right) { | |
367 | drm_block_free(mm, block->left); | |
368 | return -ENOMEM; | |
369 | } | |
370 | ||
371 | mark_free(mm, block->left); | |
372 | mark_free(mm, block->right); | |
373 | ||
96950929 APS |
374 | if (drm_buddy_block_is_clear(block)) { |
375 | mark_cleared(block->left); | |
376 | mark_cleared(block->right); | |
377 | clear_reset(block); | |
378 | } | |
379 | ||
6387a3c4 A |
380 | mark_split(block); |
381 | ||
382 | return 0; | |
383 | } | |
384 | ||
92937f17 A |
385 | /** |
386 | * drm_get_buddy - get buddy address | |
387 | * | |
388 | * @block: DRM buddy block | |
389 | * | |
390 | * Returns the corresponding buddy block for @block, or NULL | |
391 | * if this is a root block and can't be merged further. | |
392 | * Requires some kind of locking to protect against | |
393 | * any concurrent allocate and free operations. | |
394 | */ | |
395 | struct drm_buddy_block * | |
396 | drm_get_buddy(struct drm_buddy_block *block) | |
397 | { | |
398 | return __get_buddy(block); | |
399 | } | |
400 | EXPORT_SYMBOL(drm_get_buddy); | |
401 | ||
6387a3c4 A |
402 | /** |
403 | * drm_buddy_free_block - free a block | |
404 | * | |
405 | * @mm: DRM buddy manager | |
406 | * @block: block to be freed | |
407 | */ | |
408 | void drm_buddy_free_block(struct drm_buddy *mm, | |
409 | struct drm_buddy_block *block) | |
410 | { | |
411 | BUG_ON(!drm_buddy_block_is_allocated(block)); | |
412 | mm->avail += drm_buddy_block_size(mm, block); | |
96950929 APS |
413 | if (drm_buddy_block_is_clear(block)) |
414 | mm->clear_avail += drm_buddy_block_size(mm, block); | |
415 | ||
416 | __drm_buddy_free(mm, block, false); | |
6387a3c4 A |
417 | } |
418 | EXPORT_SYMBOL(drm_buddy_free_block); | |
419 | ||
96950929 APS |
420 | static void __drm_buddy_free_list(struct drm_buddy *mm, |
421 | struct list_head *objects, | |
422 | bool mark_clear, | |
423 | bool mark_dirty) | |
6387a3c4 A |
424 | { |
425 | struct drm_buddy_block *block, *on; | |
426 | ||
96950929 APS |
427 | WARN_ON(mark_dirty && mark_clear); |
428 | ||
6387a3c4 | 429 | list_for_each_entry_safe(block, on, objects, link) { |
96950929 APS |
430 | if (mark_clear) |
431 | mark_cleared(block); | |
432 | else if (mark_dirty) | |
433 | clear_reset(block); | |
6387a3c4 A |
434 | drm_buddy_free_block(mm, block); |
435 | cond_resched(); | |
436 | } | |
437 | INIT_LIST_HEAD(objects); | |
438 | } | |
6387a3c4 | 439 | |
96950929 APS |
440 | static void drm_buddy_free_list_internal(struct drm_buddy *mm, |
441 | struct list_head *objects) | |
afea229f | 442 | { |
96950929 APS |
443 | /* |
444 | * Don't touch the clear/dirty bit, since allocation is still internal | |
445 | * at this point. For example we might have just failed part of the | |
446 | * allocation. | |
447 | */ | |
448 | __drm_buddy_free_list(mm, objects, false, false); | |
afea229f A |
449 | } |
450 | ||
96950929 APS |
451 | /** |
452 | * drm_buddy_free_list - free blocks | |
453 | * | |
454 | * @mm: DRM buddy manager | |
455 | * @objects: input list head to free blocks | |
456 | * @flags: optional flags like DRM_BUDDY_CLEARED | |
457 | */ | |
458 | void drm_buddy_free_list(struct drm_buddy *mm, | |
459 | struct list_head *objects, | |
460 | unsigned int flags) | |
afea229f | 461 | { |
96950929 APS |
462 | bool mark_clear = flags & DRM_BUDDY_CLEARED; |
463 | ||
464 | __drm_buddy_free_list(mm, objects, mark_clear, !mark_clear); | |
465 | } | |
466 | EXPORT_SYMBOL(drm_buddy_free_list); | |
467 | ||
468 | static bool block_incompatible(struct drm_buddy_block *block, unsigned int flags) | |
469 | { | |
470 | bool needs_clear = flags & DRM_BUDDY_CLEAR_ALLOCATION; | |
471 | ||
472 | return needs_clear != drm_buddy_block_is_clear(block); | |
afea229f A |
473 | } |
474 | ||
475 | static struct drm_buddy_block * | |
96950929 APS |
476 | __alloc_range_bias(struct drm_buddy *mm, |
477 | u64 start, u64 end, | |
478 | unsigned int order, | |
479 | unsigned long flags, | |
480 | bool fallback) | |
afea229f | 481 | { |
f41900e4 | 482 | u64 req_size = mm->chunk_size << order; |
afea229f A |
483 | struct drm_buddy_block *block; |
484 | struct drm_buddy_block *buddy; | |
485 | LIST_HEAD(dfs); | |
486 | int err; | |
487 | int i; | |
488 | ||
489 | end = end - 1; | |
490 | ||
491 | for (i = 0; i < mm->n_roots; ++i) | |
492 | list_add_tail(&mm->roots[i]->tmp_link, &dfs); | |
493 | ||
494 | do { | |
495 | u64 block_start; | |
496 | u64 block_end; | |
497 | ||
498 | block = list_first_entry_or_null(&dfs, | |
499 | struct drm_buddy_block, | |
500 | tmp_link); | |
501 | if (!block) | |
502 | break; | |
503 | ||
504 | list_del(&block->tmp_link); | |
505 | ||
506 | if (drm_buddy_block_order(block) < order) | |
507 | continue; | |
508 | ||
509 | block_start = drm_buddy_block_offset(block); | |
510 | block_end = block_start + drm_buddy_block_size(mm, block) - 1; | |
511 | ||
512 | if (!overlaps(start, end, block_start, block_end)) | |
513 | continue; | |
514 | ||
515 | if (drm_buddy_block_is_allocated(block)) | |
516 | continue; | |
517 | ||
f41900e4 MA |
518 | if (block_start < start || block_end > end) { |
519 | u64 adjusted_start = max(block_start, start); | |
520 | u64 adjusted_end = min(block_end, end); | |
521 | ||
522 | if (round_down(adjusted_end + 1, req_size) <= | |
523 | round_up(adjusted_start, req_size)) | |
524 | continue; | |
525 | } | |
526 | ||
afea229f A |
527 | if (contains(start, end, block_start, block_end) && |
528 | order == drm_buddy_block_order(block)) { | |
96950929 APS |
529 | if (!fallback && block_incompatible(block, flags)) |
530 | continue; | |
531 | ||
afea229f A |
532 | /* |
533 | * Find the free block within the range. | |
534 | */ | |
535 | if (drm_buddy_block_is_free(block)) | |
536 | return block; | |
537 | ||
538 | continue; | |
539 | } | |
540 | ||
541 | if (!drm_buddy_block_is_split(block)) { | |
542 | err = split_block(mm, block); | |
543 | if (unlikely(err)) | |
544 | goto err_undo; | |
545 | } | |
546 | ||
547 | list_add(&block->right->tmp_link, &dfs); | |
548 | list_add(&block->left->tmp_link, &dfs); | |
549 | } while (1); | |
550 | ||
551 | return ERR_PTR(-ENOSPC); | |
552 | ||
553 | err_undo: | |
554 | /* | |
555 | * We really don't want to leave around a bunch of split blocks, since | |
556 | * bigger is better, so make sure we merge everything back before we | |
557 | * free the allocated blocks. | |
558 | */ | |
92937f17 | 559 | buddy = __get_buddy(block); |
afea229f A |
560 | if (buddy && |
561 | (drm_buddy_block_is_free(block) && | |
562 | drm_buddy_block_is_free(buddy))) | |
96950929 | 563 | __drm_buddy_free(mm, block, false); |
afea229f A |
564 | return ERR_PTR(err); |
565 | } | |
566 | ||
476e4063 | 567 | static struct drm_buddy_block * |
96950929 APS |
568 | __drm_buddy_alloc_range_bias(struct drm_buddy *mm, |
569 | u64 start, u64 end, | |
570 | unsigned int order, | |
571 | unsigned long flags) | |
572 | { | |
573 | struct drm_buddy_block *block; | |
574 | bool fallback = false; | |
575 | ||
576 | block = __alloc_range_bias(mm, start, end, order, | |
577 | flags, fallback); | |
bb21700b | 578 | if (IS_ERR(block)) |
96950929 APS |
579 | return __alloc_range_bias(mm, start, end, order, |
580 | flags, !fallback); | |
581 | ||
582 | return block; | |
583 | } | |
584 | ||
585 | static struct drm_buddy_block * | |
586 | get_maxblock(struct drm_buddy *mm, unsigned int order, | |
587 | unsigned long flags) | |
476e4063 | 588 | { |
96950929 | 589 | struct drm_buddy_block *max_block = NULL, *block = NULL; |
5640e816 | 590 | unsigned int i; |
476e4063 | 591 | |
5640e816 | 592 | for (i = order; i <= mm->max_order; ++i) { |
96950929 APS |
593 | struct drm_buddy_block *tmp_block; |
594 | ||
595 | list_for_each_entry_reverse(tmp_block, &mm->free_list[i], link) { | |
596 | if (block_incompatible(tmp_block, flags)) | |
5640e816 | 597 | continue; |
5640e816 | 598 | |
96950929 APS |
599 | block = tmp_block; |
600 | break; | |
601 | } | |
602 | ||
603 | if (!block) | |
604 | continue; | |
605 | ||
606 | if (!max_block) { | |
607 | max_block = block; | |
608 | continue; | |
609 | } | |
610 | ||
611 | if (drm_buddy_block_offset(block) > | |
612 | drm_buddy_block_offset(max_block)) { | |
613 | max_block = block; | |
5640e816 | 614 | } |
476e4063 A |
615 | } |
616 | ||
617 | return max_block; | |
618 | } | |
619 | ||
afea229f A |
620 | static struct drm_buddy_block * |
621 | alloc_from_freelist(struct drm_buddy *mm, | |
622 | unsigned int order, | |
623 | unsigned long flags) | |
6387a3c4 A |
624 | { |
625 | struct drm_buddy_block *block = NULL; | |
5640e816 | 626 | unsigned int tmp; |
6387a3c4 A |
627 | int err; |
628 | ||
5640e816 | 629 | if (flags & DRM_BUDDY_TOPDOWN_ALLOCATION) { |
96950929 | 630 | block = get_maxblock(mm, order, flags); |
5640e816 APS |
631 | if (block) |
632 | /* Store the obtained block order */ | |
633 | tmp = drm_buddy_block_order(block); | |
634 | } else { | |
96950929 APS |
635 | for (tmp = order; tmp <= mm->max_order; ++tmp) { |
636 | struct drm_buddy_block *tmp_block; | |
637 | ||
638 | list_for_each_entry_reverse(tmp_block, &mm->free_list[tmp], link) { | |
639 | if (block_incompatible(tmp_block, flags)) | |
640 | continue; | |
641 | ||
642 | block = tmp_block; | |
643 | break; | |
644 | } | |
645 | ||
646 | if (block) | |
647 | break; | |
648 | } | |
649 | } | |
650 | ||
651 | if (!block) { | |
652 | /* Fallback method */ | |
5640e816 APS |
653 | for (tmp = order; tmp <= mm->max_order; ++tmp) { |
654 | if (!list_empty(&mm->free_list[tmp])) { | |
655 | block = list_last_entry(&mm->free_list[tmp], | |
656 | struct drm_buddy_block, | |
657 | link); | |
658 | if (block) | |
659 | break; | |
660 | } | |
476e4063 | 661 | } |
6387a3c4 | 662 | |
96950929 APS |
663 | if (!block) |
664 | return ERR_PTR(-ENOSPC); | |
665 | } | |
6387a3c4 A |
666 | |
667 | BUG_ON(!drm_buddy_block_is_free(block)); | |
668 | ||
5640e816 | 669 | while (tmp != order) { |
6387a3c4 A |
670 | err = split_block(mm, block); |
671 | if (unlikely(err)) | |
afea229f | 672 | goto err_undo; |
6387a3c4 | 673 | |
afea229f | 674 | block = block->right; |
5640e816 | 675 | tmp--; |
6387a3c4 | 676 | } |
6387a3c4 A |
677 | return block; |
678 | ||
afea229f | 679 | err_undo: |
5640e816 | 680 | if (tmp != order) |
96950929 | 681 | __drm_buddy_free(mm, block, false); |
6387a3c4 A |
682 | return ERR_PTR(err); |
683 | } | |
6387a3c4 | 684 | |
afea229f A |
685 | static int __alloc_range(struct drm_buddy *mm, |
686 | struct list_head *dfs, | |
687 | u64 start, u64 size, | |
0a1844bf APS |
688 | struct list_head *blocks, |
689 | u64 *total_allocated_on_err) | |
6387a3c4 A |
690 | { |
691 | struct drm_buddy_block *block; | |
692 | struct drm_buddy_block *buddy; | |
0a1844bf | 693 | u64 total_allocated = 0; |
6387a3c4 | 694 | LIST_HEAD(allocated); |
6387a3c4 A |
695 | u64 end; |
696 | int err; | |
6387a3c4 A |
697 | |
698 | end = start + size - 1; | |
699 | ||
700 | do { | |
701 | u64 block_start; | |
702 | u64 block_end; | |
703 | ||
afea229f | 704 | block = list_first_entry_or_null(dfs, |
6387a3c4 A |
705 | struct drm_buddy_block, |
706 | tmp_link); | |
707 | if (!block) | |
708 | break; | |
709 | ||
710 | list_del(&block->tmp_link); | |
711 | ||
712 | block_start = drm_buddy_block_offset(block); | |
713 | block_end = block_start + drm_buddy_block_size(mm, block) - 1; | |
714 | ||
715 | if (!overlaps(start, end, block_start, block_end)) | |
716 | continue; | |
717 | ||
718 | if (drm_buddy_block_is_allocated(block)) { | |
719 | err = -ENOSPC; | |
720 | goto err_free; | |
721 | } | |
722 | ||
723 | if (contains(start, end, block_start, block_end)) { | |
96950929 APS |
724 | if (drm_buddy_block_is_free(block)) { |
725 | mark_allocated(block); | |
726 | total_allocated += drm_buddy_block_size(mm, block); | |
727 | mm->avail -= drm_buddy_block_size(mm, block); | |
728 | if (drm_buddy_block_is_clear(block)) | |
729 | mm->clear_avail -= drm_buddy_block_size(mm, block); | |
730 | list_add_tail(&block->link, &allocated); | |
731 | continue; | |
732 | } else if (!mm->clear_avail) { | |
6387a3c4 A |
733 | err = -ENOSPC; |
734 | goto err_free; | |
735 | } | |
6387a3c4 A |
736 | } |
737 | ||
738 | if (!drm_buddy_block_is_split(block)) { | |
739 | err = split_block(mm, block); | |
740 | if (unlikely(err)) | |
741 | goto err_undo; | |
742 | } | |
743 | ||
afea229f A |
744 | list_add(&block->right->tmp_link, dfs); |
745 | list_add(&block->left->tmp_link, dfs); | |
6387a3c4 A |
746 | } while (1); |
747 | ||
8746c6c9 APS |
748 | if (total_allocated < size) { |
749 | err = -ENOSPC; | |
750 | goto err_free; | |
751 | } | |
752 | ||
02f76a9c APS |
753 | list_splice_tail(&allocated, blocks); |
754 | ||
6387a3c4 A |
755 | return 0; |
756 | ||
757 | err_undo: | |
758 | /* | |
759 | * We really don't want to leave around a bunch of split blocks, since | |
760 | * bigger is better, so make sure we merge everything back before we | |
761 | * free the allocated blocks. | |
762 | */ | |
92937f17 | 763 | buddy = __get_buddy(block); |
6387a3c4 A |
764 | if (buddy && |
765 | (drm_buddy_block_is_free(block) && | |
766 | drm_buddy_block_is_free(buddy))) | |
96950929 | 767 | __drm_buddy_free(mm, block, false); |
6387a3c4 A |
768 | |
769 | err_free: | |
0a1844bf APS |
770 | if (err == -ENOSPC && total_allocated_on_err) { |
771 | list_splice_tail(&allocated, blocks); | |
772 | *total_allocated_on_err = total_allocated; | |
773 | } else { | |
96950929 | 774 | drm_buddy_free_list_internal(mm, &allocated); |
0a1844bf APS |
775 | } |
776 | ||
6387a3c4 A |
777 | return err; |
778 | } | |
afea229f A |
779 | |
780 | static int __drm_buddy_alloc_range(struct drm_buddy *mm, | |
781 | u64 start, | |
782 | u64 size, | |
0a1844bf | 783 | u64 *total_allocated_on_err, |
afea229f A |
784 | struct list_head *blocks) |
785 | { | |
786 | LIST_HEAD(dfs); | |
787 | int i; | |
788 | ||
789 | for (i = 0; i < mm->n_roots; ++i) | |
790 | list_add_tail(&mm->roots[i]->tmp_link, &dfs); | |
791 | ||
0a1844bf APS |
792 | return __alloc_range(mm, &dfs, start, size, |
793 | blocks, total_allocated_on_err); | |
794 | } | |
795 | ||
796 | static int __alloc_contig_try_harder(struct drm_buddy *mm, | |
797 | u64 size, | |
798 | u64 min_block_size, | |
799 | struct list_head *blocks) | |
800 | { | |
801 | u64 rhs_offset, lhs_offset, lhs_size, filled; | |
802 | struct drm_buddy_block *block; | |
803 | struct list_head *list; | |
804 | LIST_HEAD(blocks_lhs); | |
805 | unsigned long pages; | |
806 | unsigned int order; | |
807 | u64 modify_size; | |
808 | int err; | |
809 | ||
810 | modify_size = rounddown_pow_of_two(size); | |
811 | pages = modify_size >> ilog2(mm->chunk_size); | |
812 | order = fls(pages) - 1; | |
813 | if (order == 0) | |
814 | return -ENOSPC; | |
815 | ||
816 | list = &mm->free_list[order]; | |
817 | if (list_empty(list)) | |
818 | return -ENOSPC; | |
819 | ||
820 | list_for_each_entry_reverse(block, list, link) { | |
821 | /* Allocate blocks traversing RHS */ | |
822 | rhs_offset = drm_buddy_block_offset(block); | |
823 | err = __drm_buddy_alloc_range(mm, rhs_offset, size, | |
824 | &filled, blocks); | |
825 | if (!err || err != -ENOSPC) | |
826 | return err; | |
827 | ||
828 | lhs_size = max((size - filled), min_block_size); | |
829 | if (!IS_ALIGNED(lhs_size, min_block_size)) | |
830 | lhs_size = round_up(lhs_size, min_block_size); | |
831 | ||
832 | /* Allocate blocks traversing LHS */ | |
833 | lhs_offset = drm_buddy_block_offset(block) - lhs_size; | |
834 | err = __drm_buddy_alloc_range(mm, lhs_offset, lhs_size, | |
835 | NULL, &blocks_lhs); | |
836 | if (!err) { | |
837 | list_splice(&blocks_lhs, blocks); | |
838 | return 0; | |
839 | } else if (err != -ENOSPC) { | |
96950929 | 840 | drm_buddy_free_list_internal(mm, blocks); |
0a1844bf APS |
841 | return err; |
842 | } | |
843 | /* Free blocks for the next iteration */ | |
96950929 | 844 | drm_buddy_free_list_internal(mm, blocks); |
0a1844bf APS |
845 | } |
846 | ||
847 | return -ENOSPC; | |
afea229f A |
848 | } |
849 | ||
95ee2a8b A |
850 | /** |
851 | * drm_buddy_block_trim - free unused pages | |
852 | * | |
853 | * @mm: DRM buddy manager | |
854 | * @new_size: original size requested | |
855 | * @blocks: Input and output list of allocated blocks. | |
856 | * MUST contain single block as input to be trimmed. | |
857 | * On success will contain the newly allocated blocks | |
858 | * making up the @new_size. Blocks always appear in | |
859 | * ascending order | |
860 | * | |
861 | * For contiguous allocation, we round up the size to the nearest | |
862 | * power of two value, drivers consume *actual* size, so remaining | |
863 | * portions are unused and can be optionally freed with this function | |
864 | * | |
865 | * Returns: | |
866 | * 0 on success, error code on failure. | |
867 | */ | |
868 | int drm_buddy_block_trim(struct drm_buddy *mm, | |
869 | u64 new_size, | |
870 | struct list_head *blocks) | |
871 | { | |
872 | struct drm_buddy_block *parent; | |
873 | struct drm_buddy_block *block; | |
874 | LIST_HEAD(dfs); | |
875 | u64 new_start; | |
876 | int err; | |
877 | ||
878 | if (!list_is_singular(blocks)) | |
879 | return -EINVAL; | |
880 | ||
881 | block = list_first_entry(blocks, | |
882 | struct drm_buddy_block, | |
883 | link); | |
884 | ||
885 | if (WARN_ON(!drm_buddy_block_is_allocated(block))) | |
886 | return -EINVAL; | |
887 | ||
888 | if (new_size > drm_buddy_block_size(mm, block)) | |
889 | return -EINVAL; | |
890 | ||
891 | if (!new_size || !IS_ALIGNED(new_size, mm->chunk_size)) | |
892 | return -EINVAL; | |
893 | ||
894 | if (new_size == drm_buddy_block_size(mm, block)) | |
895 | return 0; | |
896 | ||
897 | list_del(&block->link); | |
898 | mark_free(mm, block); | |
899 | mm->avail += drm_buddy_block_size(mm, block); | |
96950929 APS |
900 | if (drm_buddy_block_is_clear(block)) |
901 | mm->clear_avail += drm_buddy_block_size(mm, block); | |
95ee2a8b A |
902 | |
903 | /* Prevent recursively freeing this node */ | |
904 | parent = block->parent; | |
905 | block->parent = NULL; | |
906 | ||
907 | new_start = drm_buddy_block_offset(block); | |
908 | list_add(&block->tmp_link, &dfs); | |
0a1844bf | 909 | err = __alloc_range(mm, &dfs, new_start, new_size, blocks, NULL); |
95ee2a8b A |
910 | if (err) { |
911 | mark_allocated(block); | |
912 | mm->avail -= drm_buddy_block_size(mm, block); | |
96950929 APS |
913 | if (drm_buddy_block_is_clear(block)) |
914 | mm->clear_avail -= drm_buddy_block_size(mm, block); | |
95ee2a8b A |
915 | list_add(&block->link, blocks); |
916 | } | |
917 | ||
918 | block->parent = parent; | |
919 | return err; | |
920 | } | |
921 | EXPORT_SYMBOL(drm_buddy_block_trim); | |
922 | ||
96950929 APS |
923 | static struct drm_buddy_block * |
924 | __drm_buddy_alloc_blocks(struct drm_buddy *mm, | |
925 | u64 start, u64 end, | |
926 | unsigned int order, | |
927 | unsigned long flags) | |
928 | { | |
929 | if (flags & DRM_BUDDY_RANGE_ALLOCATION) | |
930 | /* Allocate traversing within the range */ | |
931 | return __drm_buddy_alloc_range_bias(mm, start, end, | |
932 | order, flags); | |
933 | else | |
934 | /* Allocate from freelist */ | |
935 | return alloc_from_freelist(mm, order, flags); | |
936 | } | |
937 | ||
afea229f A |
938 | /** |
939 | * drm_buddy_alloc_blocks - allocate power-of-two blocks | |
940 | * | |
941 | * @mm: DRM buddy manager to allocate from | |
942 | * @start: start of the allowed range for this block | |
943 | * @end: end of the allowed range for this block | |
96950929 | 944 | * @size: size of the allocation in bytes |
0a1844bf | 945 | * @min_block_size: alignment of the allocation |
afea229f A |
946 | * @blocks: output list head to add allocated blocks |
947 | * @flags: DRM_BUDDY_*_ALLOCATION flags | |
948 | * | |
949 | * alloc_range_bias() called on range limitations, which traverses | |
950 | * the tree and returns the desired block. | |
951 | * | |
952 | * alloc_from_freelist() called when *no* range restrictions | |
953 | * are enforced, which picks the block from the freelist. | |
954 | * | |
955 | * Returns: | |
956 | * 0 on success, error code on failure. | |
957 | */ | |
958 | int drm_buddy_alloc_blocks(struct drm_buddy *mm, | |
959 | u64 start, u64 end, u64 size, | |
0a1844bf | 960 | u64 min_block_size, |
afea229f A |
961 | struct list_head *blocks, |
962 | unsigned long flags) | |
963 | { | |
964 | struct drm_buddy_block *block = NULL; | |
0a1844bf | 965 | u64 original_size, original_min_size; |
afea229f | 966 | unsigned int min_order, order; |
afea229f | 967 | LIST_HEAD(allocated); |
0a1844bf | 968 | unsigned long pages; |
afea229f A |
969 | int err; |
970 | ||
971 | if (size < mm->chunk_size) | |
972 | return -EINVAL; | |
973 | ||
0a1844bf | 974 | if (min_block_size < mm->chunk_size) |
afea229f A |
975 | return -EINVAL; |
976 | ||
0a1844bf | 977 | if (!is_power_of_2(min_block_size)) |
afea229f A |
978 | return -EINVAL; |
979 | ||
980 | if (!IS_ALIGNED(start | end | size, mm->chunk_size)) | |
981 | return -EINVAL; | |
982 | ||
983 | if (end > mm->size) | |
984 | return -EINVAL; | |
985 | ||
986 | if (range_overflows(start, size, mm->size)) | |
987 | return -EINVAL; | |
988 | ||
989 | /* Actual range allocation */ | |
2986314a MA |
990 | if (start + size == end) { |
991 | if (!IS_ALIGNED(start | end, min_block_size)) | |
992 | return -EINVAL; | |
993 | ||
0a1844bf | 994 | return __drm_buddy_alloc_range(mm, start, size, NULL, blocks); |
2986314a | 995 | } |
0a1844bf APS |
996 | |
997 | original_size = size; | |
998 | original_min_size = min_block_size; | |
999 | ||
1000 | /* Roundup the size to power of 2 */ | |
1001 | if (flags & DRM_BUDDY_CONTIGUOUS_ALLOCATION) { | |
1002 | size = roundup_pow_of_two(size); | |
1003 | min_block_size = size; | |
1004 | /* Align size value to min_block_size */ | |
1005 | } else if (!IS_ALIGNED(size, min_block_size)) { | |
1006 | size = round_up(size, min_block_size); | |
1007 | } | |
5f778760 | 1008 | |
afea229f A |
1009 | pages = size >> ilog2(mm->chunk_size); |
1010 | order = fls(pages) - 1; | |
0a1844bf | 1011 | min_order = ilog2(min_block_size) - ilog2(mm->chunk_size); |
afea229f A |
1012 | |
1013 | do { | |
1014 | order = min(order, (unsigned int)fls(pages) - 1); | |
1015 | BUG_ON(order > mm->max_order); | |
1016 | BUG_ON(order < min_order); | |
1017 | ||
1018 | do { | |
96950929 APS |
1019 | block = __drm_buddy_alloc_blocks(mm, start, |
1020 | end, | |
1021 | order, | |
1022 | flags); | |
afea229f A |
1023 | if (!IS_ERR(block)) |
1024 | break; | |
1025 | ||
1026 | if (order-- == min_order) { | |
96950929 APS |
1027 | /* Try allocation through force merge method */ |
1028 | if (mm->clear_avail && | |
1029 | !__force_merge(mm, start, end, min_order)) { | |
1030 | block = __drm_buddy_alloc_blocks(mm, start, | |
1031 | end, | |
1032 | min_order, | |
1033 | flags); | |
1034 | if (!IS_ERR(block)) { | |
1035 | order = min_order; | |
1036 | break; | |
1037 | } | |
1038 | } | |
1039 | ||
1040 | /* | |
1041 | * Try contiguous block allocation through | |
1042 | * try harder method. | |
1043 | */ | |
0a1844bf APS |
1044 | if (flags & DRM_BUDDY_CONTIGUOUS_ALLOCATION && |
1045 | !(flags & DRM_BUDDY_RANGE_ALLOCATION)) | |
0a1844bf APS |
1046 | return __alloc_contig_try_harder(mm, |
1047 | original_size, | |
1048 | original_min_size, | |
1049 | blocks); | |
afea229f A |
1050 | err = -ENOSPC; |
1051 | goto err_free; | |
1052 | } | |
1053 | } while (1); | |
1054 | ||
1055 | mark_allocated(block); | |
1056 | mm->avail -= drm_buddy_block_size(mm, block); | |
96950929 APS |
1057 | if (drm_buddy_block_is_clear(block)) |
1058 | mm->clear_avail -= drm_buddy_block_size(mm, block); | |
afea229f A |
1059 | kmemleak_update_trace(block); |
1060 | list_add_tail(&block->link, &allocated); | |
1061 | ||
1062 | pages -= BIT(order); | |
1063 | ||
1064 | if (!pages) | |
1065 | break; | |
1066 | } while (1); | |
1067 | ||
0a1844bf APS |
1068 | /* Trim the allocated block to the required size */ |
1069 | if (original_size != size) { | |
1070 | struct list_head *trim_list; | |
1071 | LIST_HEAD(temp); | |
1072 | u64 trim_size; | |
1073 | ||
1074 | trim_list = &allocated; | |
1075 | trim_size = original_size; | |
1076 | ||
1077 | if (!list_is_singular(&allocated)) { | |
1078 | block = list_last_entry(&allocated, typeof(*block), link); | |
1079 | list_move(&block->link, &temp); | |
1080 | trim_list = &temp; | |
1081 | trim_size = drm_buddy_block_size(mm, block) - | |
1082 | (size - original_size); | |
1083 | } | |
1084 | ||
1085 | drm_buddy_block_trim(mm, | |
1086 | trim_size, | |
1087 | trim_list); | |
1088 | ||
1089 | if (!list_empty(&temp)) | |
1090 | list_splice_tail(trim_list, &allocated); | |
1091 | } | |
1092 | ||
afea229f A |
1093 | list_splice_tail(&allocated, blocks); |
1094 | return 0; | |
1095 | ||
1096 | err_free: | |
96950929 | 1097 | drm_buddy_free_list_internal(mm, &allocated); |
afea229f A |
1098 | return err; |
1099 | } | |
1100 | EXPORT_SYMBOL(drm_buddy_alloc_blocks); | |
6387a3c4 A |
1101 | |
1102 | /** | |
1103 | * drm_buddy_block_print - print block information | |
1104 | * | |
1105 | * @mm: DRM buddy manager | |
1106 | * @block: DRM buddy block | |
1107 | * @p: DRM printer to use | |
1108 | */ | |
1109 | void drm_buddy_block_print(struct drm_buddy *mm, | |
1110 | struct drm_buddy_block *block, | |
1111 | struct drm_printer *p) | |
1112 | { | |
1113 | u64 start = drm_buddy_block_offset(block); | |
1114 | u64 size = drm_buddy_block_size(mm, block); | |
1115 | ||
1116 | drm_printf(p, "%#018llx-%#018llx: %llu\n", start, start + size, size); | |
1117 | } | |
1118 | EXPORT_SYMBOL(drm_buddy_block_print); | |
1119 | ||
1120 | /** | |
1121 | * drm_buddy_print - print allocator state | |
1122 | * | |
1123 | * @mm: DRM buddy manager | |
1124 | * @p: DRM printer to use | |
1125 | */ | |
1126 | void drm_buddy_print(struct drm_buddy *mm, struct drm_printer *p) | |
1127 | { | |
1128 | int order; | |
1129 | ||
96950929 APS |
1130 | drm_printf(p, "chunk_size: %lluKiB, total: %lluMiB, free: %lluMiB, clear_free: %lluMiB\n", |
1131 | mm->chunk_size >> 10, mm->size >> 20, mm->avail >> 20, mm->clear_avail >> 20); | |
6387a3c4 A |
1132 | |
1133 | for (order = mm->max_order; order >= 0; order--) { | |
1134 | struct drm_buddy_block *block; | |
1135 | u64 count = 0, free; | |
1136 | ||
1137 | list_for_each_entry(block, &mm->free_list[order], link) { | |
1138 | BUG_ON(!drm_buddy_block_is_free(block)); | |
1139 | count++; | |
1140 | } | |
1141 | ||
cd11589b | 1142 | drm_printf(p, "order-%2d ", order); |
6387a3c4 A |
1143 | |
1144 | free = count * (mm->chunk_size << order); | |
1145 | if (free < SZ_1M) | |
cd11589b | 1146 | drm_printf(p, "free: %8llu KiB", free >> 10); |
6387a3c4 | 1147 | else |
cd11589b | 1148 | drm_printf(p, "free: %8llu MiB", free >> 20); |
6387a3c4 | 1149 | |
cd11589b | 1150 | drm_printf(p, ", blocks: %llu\n", count); |
6387a3c4 A |
1151 | } |
1152 | } | |
1153 | EXPORT_SYMBOL(drm_buddy_print); | |
1154 | ||
1155 | static void drm_buddy_module_exit(void) | |
1156 | { | |
1157 | kmem_cache_destroy(slab_blocks); | |
1158 | } | |
1159 | ||
1160 | static int __init drm_buddy_module_init(void) | |
1161 | { | |
1162 | slab_blocks = KMEM_CACHE(drm_buddy_block, 0); | |
1163 | if (!slab_blocks) | |
1164 | return -ENOMEM; | |
1165 | ||
1166 | return 0; | |
1167 | } | |
1168 | ||
1169 | module_init(drm_buddy_module_init); | |
1170 | module_exit(drm_buddy_module_exit); | |
1171 | ||
1172 | MODULE_DESCRIPTION("DRM Buddy Allocator"); | |
1173 | MODULE_LICENSE("Dual MIT/GPL"); |