]>
Commit | Line | Data |
---|---|---|
801c135c AB |
1 | /* |
2 | * Copyright (c) International Business Machines Corp., 2006 | |
3 | * | |
4 | * This program is free software; you can redistribute it and/or modify | |
5 | * it under the terms of the GNU General Public License as published by | |
6 | * the Free Software Foundation; either version 2 of the License, or | |
7 | * (at your option) any later version. | |
8 | * | |
9 | * This program is distributed in the hope that it will be useful, | |
10 | * but WITHOUT ANY WARRANTY; without even the implied warranty of | |
11 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See | |
12 | * the GNU General Public License for more details. | |
13 | * | |
14 | * You should have received a copy of the GNU General Public License | |
15 | * along with this program; if not, write to the Free Software | |
16 | * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA | |
17 | * | |
18 | * Authors: Artem Bityutskiy (Битюцкий Артём), Thomas Gleixner | |
19 | */ | |
20 | ||
21 | /* | |
22 | * UBI wear-leveling unit. | |
23 | * | |
24 | * This unit is responsible for wear-leveling. It works in terms of physical | |
25 | * eraseblocks and erase counters and knows nothing about logical eraseblocks, | |
26 | * volumes, etc. From this unit's perspective all physical eraseblocks are of | |
27 | * two types - used and free. Used physical eraseblocks are those that were | |
28 | * "get" by the 'ubi_wl_get_peb()' function, and free physical eraseblocks are | |
29 | * those that were put by the 'ubi_wl_put_peb()' function. | |
30 | * | |
31 | * Physical eraseblocks returned by 'ubi_wl_get_peb()' have only erase counter | |
32 | * header. The rest of the physical eraseblock contains only 0xFF bytes. | |
33 | * | |
34 | * When physical eraseblocks are returned to the WL unit by means of the | |
35 | * 'ubi_wl_put_peb()' function, they are scheduled for erasure. The erasure is | |
36 | * done asynchronously in context of the per-UBI device background thread, | |
37 | * which is also managed by the WL unit. | |
38 | * | |
39 | * The wear-leveling is ensured by means of moving the contents of used | |
40 | * physical eraseblocks with low erase counter to free physical eraseblocks | |
41 | * with high erase counter. | |
42 | * | |
43 | * The 'ubi_wl_get_peb()' function accepts data type hints which help to pick | |
44 | * an "optimal" physical eraseblock. For example, when it is known that the | |
45 | * physical eraseblock will be "put" soon because it contains short-term data, | |
46 | * the WL unit may pick a free physical eraseblock with low erase counter, and | |
47 | * so forth. | |
48 | * | |
49 | * If the WL unit fails to erase a physical eraseblock, it marks it as bad. | |
50 | * | |
51 | * This unit is also responsible for scrubbing. If a bit-flip is detected in a | |
52 | * physical eraseblock, it has to be moved. Technically this is the same as | |
53 | * moving it for wear-leveling reasons. | |
54 | * | |
55 | * As it was said, for the UBI unit all physical eraseblocks are either "free" | |
56 | * or "used". Free eraseblock are kept in the @wl->free RB-tree, while used | |
57 | * eraseblocks are kept in a set of different RB-trees: @wl->used, | |
58 | * @wl->prot.pnum, @wl->prot.aec, and @wl->scrub. | |
59 | * | |
60 | * Note, in this implementation, we keep a small in-RAM object for each physical | |
61 | * eraseblock. This is surely not a scalable solution. But it appears to be good | |
62 | * enough for moderately large flashes and it is simple. In future, one may | |
63 | * re-work this unit and make it more scalable. | |
64 | * | |
65 | * At the moment this unit does not utilize the sequence number, which was | |
66 | * introduced relatively recently. But it would be wise to do this because the | |
67 | * sequence number of a logical eraseblock characterizes how old is it. For | |
68 | * example, when we move a PEB with low erase counter, and we need to pick the | |
69 | * target PEB, we pick a PEB with the highest EC if our PEB is "old" and we | |
70 | * pick target PEB with an average EC if our PEB is not very "old". This is a | |
71 | * room for future re-works of the WL unit. | |
72 | * | |
73 | * FIXME: looks too complex, should be simplified (later). | |
74 | */ | |
75 | ||
76 | #include <linux/slab.h> | |
77 | #include <linux/crc32.h> | |
78 | #include <linux/freezer.h> | |
79 | #include <linux/kthread.h> | |
80 | #include "ubi.h" | |
81 | ||
82 | /* Number of physical eraseblocks reserved for wear-leveling purposes */ | |
83 | #define WL_RESERVED_PEBS 1 | |
84 | ||
85 | /* | |
86 | * How many erase cycles are short term, unknown, and long term physical | |
87 | * eraseblocks protected. | |
88 | */ | |
89 | #define ST_PROTECTION 16 | |
90 | #define U_PROTECTION 10 | |
91 | #define LT_PROTECTION 4 | |
92 | ||
93 | /* | |
94 | * Maximum difference between two erase counters. If this threshold is | |
95 | * exceeded, the WL unit starts moving data from used physical eraseblocks with | |
96 | * low erase counter to free physical eraseblocks with high erase counter. | |
97 | */ | |
98 | #define UBI_WL_THRESHOLD CONFIG_MTD_UBI_WL_THRESHOLD | |
99 | ||
100 | /* | |
101 | * When a physical eraseblock is moved, the WL unit has to pick the target | |
102 | * physical eraseblock to move to. The simplest way would be just to pick the | |
103 | * one with the highest erase counter. But in certain workloads this could lead | |
104 | * to an unlimited wear of one or few physical eraseblock. Indeed, imagine a | |
105 | * situation when the picked physical eraseblock is constantly erased after the | |
106 | * data is written to it. So, we have a constant which limits the highest erase | |
107 | * counter of the free physical eraseblock to pick. Namely, the WL unit does | |
108 | * not pick eraseblocks with erase counter greater then the lowest erase | |
109 | * counter plus %WL_FREE_MAX_DIFF. | |
110 | */ | |
111 | #define WL_FREE_MAX_DIFF (2*UBI_WL_THRESHOLD) | |
112 | ||
113 | /* | |
114 | * Maximum number of consecutive background thread failures which is enough to | |
115 | * switch to read-only mode. | |
116 | */ | |
117 | #define WL_MAX_FAILURES 32 | |
118 | ||
801c135c AB |
119 | /** |
120 | * struct ubi_wl_prot_entry - PEB protection entry. | |
121 | * @rb_pnum: link in the @wl->prot.pnum RB-tree | |
122 | * @rb_aec: link in the @wl->prot.aec RB-tree | |
123 | * @abs_ec: the absolute erase counter value when the protection ends | |
124 | * @e: the wear-leveling entry of the physical eraseblock under protection | |
125 | * | |
126 | * When the WL unit returns a physical eraseblock, the physical eraseblock is | |
127 | * protected from being moved for some "time". For this reason, the physical | |
128 | * eraseblock is not directly moved from the @wl->free tree to the @wl->used | |
129 | * tree. There is one more tree in between where this physical eraseblock is | |
130 | * temporarily stored (@wl->prot). | |
131 | * | |
132 | * All this protection stuff is needed because: | |
133 | * o we don't want to move physical eraseblocks just after we have given them | |
134 | * to the user; instead, we first want to let users fill them up with data; | |
135 | * | |
136 | * o there is a chance that the user will put the physical eraseblock very | |
137 | * soon, so it makes sense not to move it for some time, but wait; this is | |
138 | * especially important in case of "short term" physical eraseblocks. | |
139 | * | |
140 | * Physical eraseblocks stay protected only for limited time. But the "time" is | |
141 | * measured in erase cycles in this case. This is implemented with help of the | |
142 | * absolute erase counter (@wl->abs_ec). When it reaches certain value, the | |
143 | * physical eraseblocks are moved from the protection trees (@wl->prot.*) to | |
144 | * the @wl->used tree. | |
145 | * | |
146 | * Protected physical eraseblocks are searched by physical eraseblock number | |
147 | * (when they are put) and by the absolute erase counter (to check if it is | |
148 | * time to move them to the @wl->used tree). So there are actually 2 RB-trees | |
149 | * storing the protected physical eraseblocks: @wl->prot.pnum and | |
150 | * @wl->prot.aec. They are referred to as the "protection" trees. The | |
151 | * first one is indexed by the physical eraseblock number. The second one is | |
152 | * indexed by the absolute erase counter. Both trees store | |
153 | * &struct ubi_wl_prot_entry objects. | |
154 | * | |
155 | * Each physical eraseblock has 2 main states: free and used. The former state | |
156 | * corresponds to the @wl->free tree. The latter state is split up on several | |
157 | * sub-states: | |
158 | * o the WL movement is allowed (@wl->used tree); | |
159 | * o the WL movement is temporarily prohibited (@wl->prot.pnum and | |
160 | * @wl->prot.aec trees); | |
161 | * o scrubbing is needed (@wl->scrub tree). | |
162 | * | |
163 | * Depending on the sub-state, wear-leveling entries of the used physical | |
164 | * eraseblocks may be kept in one of those trees. | |
165 | */ | |
166 | struct ubi_wl_prot_entry { | |
167 | struct rb_node rb_pnum; | |
168 | struct rb_node rb_aec; | |
169 | unsigned long long abs_ec; | |
170 | struct ubi_wl_entry *e; | |
171 | }; | |
172 | ||
173 | /** | |
174 | * struct ubi_work - UBI work description data structure. | |
175 | * @list: a link in the list of pending works | |
176 | * @func: worker function | |
177 | * @priv: private data of the worker function | |
178 | * | |
179 | * @e: physical eraseblock to erase | |
180 | * @torture: if the physical eraseblock has to be tortured | |
181 | * | |
182 | * The @func pointer points to the worker function. If the @cancel argument is | |
183 | * not zero, the worker has to free the resources and exit immediately. The | |
184 | * worker has to return zero in case of success and a negative error code in | |
185 | * case of failure. | |
186 | */ | |
187 | struct ubi_work { | |
188 | struct list_head list; | |
189 | int (*func)(struct ubi_device *ubi, struct ubi_work *wrk, int cancel); | |
190 | /* The below fields are only relevant to erasure works */ | |
191 | struct ubi_wl_entry *e; | |
192 | int torture; | |
193 | }; | |
194 | ||
195 | #ifdef CONFIG_MTD_UBI_DEBUG_PARANOID | |
e88d6e10 | 196 | static int paranoid_check_ec(struct ubi_device *ubi, int pnum, int ec); |
801c135c AB |
197 | static int paranoid_check_in_wl_tree(struct ubi_wl_entry *e, |
198 | struct rb_root *root); | |
199 | #else | |
200 | #define paranoid_check_ec(ubi, pnum, ec) 0 | |
201 | #define paranoid_check_in_wl_tree(e, root) | |
202 | #endif | |
203 | ||
801c135c AB |
204 | /** |
205 | * wl_tree_add - add a wear-leveling entry to a WL RB-tree. | |
206 | * @e: the wear-leveling entry to add | |
207 | * @root: the root of the tree | |
208 | * | |
209 | * Note, we use (erase counter, physical eraseblock number) pairs as keys in | |
210 | * the @ubi->used and @ubi->free RB-trees. | |
211 | */ | |
212 | static void wl_tree_add(struct ubi_wl_entry *e, struct rb_root *root) | |
213 | { | |
214 | struct rb_node **p, *parent = NULL; | |
215 | ||
216 | p = &root->rb_node; | |
217 | while (*p) { | |
218 | struct ubi_wl_entry *e1; | |
219 | ||
220 | parent = *p; | |
221 | e1 = rb_entry(parent, struct ubi_wl_entry, rb); | |
222 | ||
223 | if (e->ec < e1->ec) | |
224 | p = &(*p)->rb_left; | |
225 | else if (e->ec > e1->ec) | |
226 | p = &(*p)->rb_right; | |
227 | else { | |
228 | ubi_assert(e->pnum != e1->pnum); | |
229 | if (e->pnum < e1->pnum) | |
230 | p = &(*p)->rb_left; | |
231 | else | |
232 | p = &(*p)->rb_right; | |
233 | } | |
234 | } | |
235 | ||
236 | rb_link_node(&e->rb, parent, p); | |
237 | rb_insert_color(&e->rb, root); | |
238 | } | |
239 | ||
801c135c AB |
240 | /** |
241 | * do_work - do one pending work. | |
242 | * @ubi: UBI device description object | |
243 | * | |
244 | * This function returns zero in case of success and a negative error code in | |
245 | * case of failure. | |
246 | */ | |
247 | static int do_work(struct ubi_device *ubi) | |
248 | { | |
249 | int err; | |
250 | struct ubi_work *wrk; | |
251 | ||
43f9b25a AB |
252 | cond_resched(); |
253 | ||
593dd33c AB |
254 | /* |
255 | * @ubi->work_sem is used to synchronize with the workers. Workers take | |
256 | * it in read mode, so many of them may be doing works at a time. But | |
257 | * the queue flush code has to be sure the whole queue of works is | |
258 | * done, and it takes the mutex in write mode. | |
259 | */ | |
260 | down_read(&ubi->work_sem); | |
801c135c | 261 | spin_lock(&ubi->wl_lock); |
801c135c AB |
262 | if (list_empty(&ubi->works)) { |
263 | spin_unlock(&ubi->wl_lock); | |
593dd33c | 264 | up_read(&ubi->work_sem); |
801c135c AB |
265 | return 0; |
266 | } | |
267 | ||
268 | wrk = list_entry(ubi->works.next, struct ubi_work, list); | |
269 | list_del(&wrk->list); | |
16f557ec AB |
270 | ubi->works_count -= 1; |
271 | ubi_assert(ubi->works_count >= 0); | |
801c135c AB |
272 | spin_unlock(&ubi->wl_lock); |
273 | ||
274 | /* | |
275 | * Call the worker function. Do not touch the work structure | |
276 | * after this call as it will have been freed or reused by that | |
277 | * time by the worker function. | |
278 | */ | |
279 | err = wrk->func(ubi, wrk, 0); | |
280 | if (err) | |
281 | ubi_err("work failed with error code %d", err); | |
593dd33c | 282 | up_read(&ubi->work_sem); |
16f557ec | 283 | |
801c135c AB |
284 | return err; |
285 | } | |
286 | ||
287 | /** | |
288 | * produce_free_peb - produce a free physical eraseblock. | |
289 | * @ubi: UBI device description object | |
290 | * | |
291 | * This function tries to make a free PEB by means of synchronous execution of | |
292 | * pending works. This may be needed if, for example the background thread is | |
293 | * disabled. Returns zero in case of success and a negative error code in case | |
294 | * of failure. | |
295 | */ | |
296 | static int produce_free_peb(struct ubi_device *ubi) | |
297 | { | |
298 | int err; | |
299 | ||
300 | spin_lock(&ubi->wl_lock); | |
5abde384 | 301 | while (!ubi->free.rb_node) { |
801c135c AB |
302 | spin_unlock(&ubi->wl_lock); |
303 | ||
304 | dbg_wl("do one work synchronously"); | |
305 | err = do_work(ubi); | |
306 | if (err) | |
307 | return err; | |
308 | ||
309 | spin_lock(&ubi->wl_lock); | |
310 | } | |
311 | spin_unlock(&ubi->wl_lock); | |
312 | ||
313 | return 0; | |
314 | } | |
315 | ||
316 | /** | |
317 | * in_wl_tree - check if wear-leveling entry is present in a WL RB-tree. | |
318 | * @e: the wear-leveling entry to check | |
319 | * @root: the root of the tree | |
320 | * | |
321 | * This function returns non-zero if @e is in the @root RB-tree and zero if it | |
322 | * is not. | |
323 | */ | |
324 | static int in_wl_tree(struct ubi_wl_entry *e, struct rb_root *root) | |
325 | { | |
326 | struct rb_node *p; | |
327 | ||
328 | p = root->rb_node; | |
329 | while (p) { | |
330 | struct ubi_wl_entry *e1; | |
331 | ||
332 | e1 = rb_entry(p, struct ubi_wl_entry, rb); | |
333 | ||
334 | if (e->pnum == e1->pnum) { | |
335 | ubi_assert(e == e1); | |
336 | return 1; | |
337 | } | |
338 | ||
339 | if (e->ec < e1->ec) | |
340 | p = p->rb_left; | |
341 | else if (e->ec > e1->ec) | |
342 | p = p->rb_right; | |
343 | else { | |
344 | ubi_assert(e->pnum != e1->pnum); | |
345 | if (e->pnum < e1->pnum) | |
346 | p = p->rb_left; | |
347 | else | |
348 | p = p->rb_right; | |
349 | } | |
350 | } | |
351 | ||
352 | return 0; | |
353 | } | |
354 | ||
355 | /** | |
356 | * prot_tree_add - add physical eraseblock to protection trees. | |
357 | * @ubi: UBI device description object | |
358 | * @e: the physical eraseblock to add | |
359 | * @pe: protection entry object to use | |
360 | * @abs_ec: absolute erase counter value when this physical eraseblock has | |
361 | * to be removed from the protection trees. | |
362 | * | |
363 | * @wl->lock has to be locked. | |
364 | */ | |
365 | static void prot_tree_add(struct ubi_device *ubi, struct ubi_wl_entry *e, | |
366 | struct ubi_wl_prot_entry *pe, int abs_ec) | |
367 | { | |
368 | struct rb_node **p, *parent = NULL; | |
369 | struct ubi_wl_prot_entry *pe1; | |
370 | ||
371 | pe->e = e; | |
372 | pe->abs_ec = ubi->abs_ec + abs_ec; | |
373 | ||
374 | p = &ubi->prot.pnum.rb_node; | |
375 | while (*p) { | |
376 | parent = *p; | |
377 | pe1 = rb_entry(parent, struct ubi_wl_prot_entry, rb_pnum); | |
378 | ||
379 | if (e->pnum < pe1->e->pnum) | |
380 | p = &(*p)->rb_left; | |
381 | else | |
382 | p = &(*p)->rb_right; | |
383 | } | |
384 | rb_link_node(&pe->rb_pnum, parent, p); | |
385 | rb_insert_color(&pe->rb_pnum, &ubi->prot.pnum); | |
386 | ||
387 | p = &ubi->prot.aec.rb_node; | |
388 | parent = NULL; | |
389 | while (*p) { | |
390 | parent = *p; | |
391 | pe1 = rb_entry(parent, struct ubi_wl_prot_entry, rb_aec); | |
392 | ||
393 | if (pe->abs_ec < pe1->abs_ec) | |
394 | p = &(*p)->rb_left; | |
395 | else | |
396 | p = &(*p)->rb_right; | |
397 | } | |
398 | rb_link_node(&pe->rb_aec, parent, p); | |
399 | rb_insert_color(&pe->rb_aec, &ubi->prot.aec); | |
400 | } | |
401 | ||
402 | /** | |
403 | * find_wl_entry - find wear-leveling entry closest to certain erase counter. | |
404 | * @root: the RB-tree where to look for | |
405 | * @max: highest possible erase counter | |
406 | * | |
407 | * This function looks for a wear leveling entry with erase counter closest to | |
408 | * @max and less then @max. | |
409 | */ | |
410 | static struct ubi_wl_entry *find_wl_entry(struct rb_root *root, int max) | |
411 | { | |
412 | struct rb_node *p; | |
413 | struct ubi_wl_entry *e; | |
414 | ||
415 | e = rb_entry(rb_first(root), struct ubi_wl_entry, rb); | |
416 | max += e->ec; | |
417 | ||
418 | p = root->rb_node; | |
419 | while (p) { | |
420 | struct ubi_wl_entry *e1; | |
421 | ||
422 | e1 = rb_entry(p, struct ubi_wl_entry, rb); | |
423 | if (e1->ec >= max) | |
424 | p = p->rb_left; | |
425 | else { | |
426 | p = p->rb_right; | |
427 | e = e1; | |
428 | } | |
429 | } | |
430 | ||
431 | return e; | |
432 | } | |
433 | ||
434 | /** | |
435 | * ubi_wl_get_peb - get a physical eraseblock. | |
436 | * @ubi: UBI device description object | |
437 | * @dtype: type of data which will be stored in this physical eraseblock | |
438 | * | |
439 | * This function returns a physical eraseblock in case of success and a | |
440 | * negative error code in case of failure. Might sleep. | |
441 | */ | |
442 | int ubi_wl_get_peb(struct ubi_device *ubi, int dtype) | |
443 | { | |
444 | int err, protect, medium_ec; | |
445 | struct ubi_wl_entry *e, *first, *last; | |
446 | struct ubi_wl_prot_entry *pe; | |
447 | ||
448 | ubi_assert(dtype == UBI_LONGTERM || dtype == UBI_SHORTTERM || | |
449 | dtype == UBI_UNKNOWN); | |
450 | ||
33818bbb | 451 | pe = kmalloc(sizeof(struct ubi_wl_prot_entry), GFP_NOFS); |
801c135c AB |
452 | if (!pe) |
453 | return -ENOMEM; | |
454 | ||
455 | retry: | |
456 | spin_lock(&ubi->wl_lock); | |
5abde384 | 457 | if (!ubi->free.rb_node) { |
801c135c AB |
458 | if (ubi->works_count == 0) { |
459 | ubi_assert(list_empty(&ubi->works)); | |
460 | ubi_err("no free eraseblocks"); | |
461 | spin_unlock(&ubi->wl_lock); | |
462 | kfree(pe); | |
463 | return -ENOSPC; | |
464 | } | |
465 | spin_unlock(&ubi->wl_lock); | |
466 | ||
467 | err = produce_free_peb(ubi); | |
468 | if (err < 0) { | |
469 | kfree(pe); | |
470 | return err; | |
471 | } | |
472 | goto retry; | |
473 | } | |
474 | ||
475 | switch (dtype) { | |
476 | case UBI_LONGTERM: | |
477 | /* | |
478 | * For long term data we pick a physical eraseblock | |
479 | * with high erase counter. But the highest erase | |
480 | * counter we can pick is bounded by the the lowest | |
481 | * erase counter plus %WL_FREE_MAX_DIFF. | |
482 | */ | |
483 | e = find_wl_entry(&ubi->free, WL_FREE_MAX_DIFF); | |
484 | protect = LT_PROTECTION; | |
485 | break; | |
486 | case UBI_UNKNOWN: | |
487 | /* | |
488 | * For unknown data we pick a physical eraseblock with | |
489 | * medium erase counter. But we by no means can pick a | |
490 | * physical eraseblock with erase counter greater or | |
491 | * equivalent than the lowest erase counter plus | |
492 | * %WL_FREE_MAX_DIFF. | |
493 | */ | |
494 | first = rb_entry(rb_first(&ubi->free), | |
495 | struct ubi_wl_entry, rb); | |
496 | last = rb_entry(rb_last(&ubi->free), | |
497 | struct ubi_wl_entry, rb); | |
498 | ||
499 | if (last->ec - first->ec < WL_FREE_MAX_DIFF) | |
500 | e = rb_entry(ubi->free.rb_node, | |
501 | struct ubi_wl_entry, rb); | |
502 | else { | |
503 | medium_ec = (first->ec + WL_FREE_MAX_DIFF)/2; | |
504 | e = find_wl_entry(&ubi->free, medium_ec); | |
505 | } | |
506 | protect = U_PROTECTION; | |
507 | break; | |
508 | case UBI_SHORTTERM: | |
509 | /* | |
510 | * For short term data we pick a physical eraseblock | |
511 | * with the lowest erase counter as we expect it will | |
512 | * be erased soon. | |
513 | */ | |
514 | e = rb_entry(rb_first(&ubi->free), | |
515 | struct ubi_wl_entry, rb); | |
516 | protect = ST_PROTECTION; | |
517 | break; | |
518 | default: | |
519 | protect = 0; | |
520 | e = NULL; | |
521 | BUG(); | |
522 | } | |
523 | ||
524 | /* | |
525 | * Move the physical eraseblock to the protection trees where it will | |
526 | * be protected from being moved for some time. | |
527 | */ | |
5abde384 AB |
528 | paranoid_check_in_wl_tree(e, &ubi->free); |
529 | rb_erase(&e->rb, &ubi->free); | |
801c135c AB |
530 | prot_tree_add(ubi, e, pe, protect); |
531 | ||
532 | dbg_wl("PEB %d EC %d, protection %d", e->pnum, e->ec, protect); | |
533 | spin_unlock(&ubi->wl_lock); | |
534 | ||
535 | return e->pnum; | |
536 | } | |
537 | ||
538 | /** | |
539 | * prot_tree_del - remove a physical eraseblock from the protection trees | |
540 | * @ubi: UBI device description object | |
541 | * @pnum: the physical eraseblock to remove | |
43f9b25a AB |
542 | * |
543 | * This function returns PEB @pnum from the protection trees and returns zero | |
544 | * in case of success and %-ENODEV if the PEB was not found in the protection | |
545 | * trees. | |
801c135c | 546 | */ |
43f9b25a | 547 | static int prot_tree_del(struct ubi_device *ubi, int pnum) |
801c135c AB |
548 | { |
549 | struct rb_node *p; | |
550 | struct ubi_wl_prot_entry *pe = NULL; | |
551 | ||
552 | p = ubi->prot.pnum.rb_node; | |
553 | while (p) { | |
554 | ||
555 | pe = rb_entry(p, struct ubi_wl_prot_entry, rb_pnum); | |
556 | ||
557 | if (pnum == pe->e->pnum) | |
43f9b25a | 558 | goto found; |
801c135c AB |
559 | |
560 | if (pnum < pe->e->pnum) | |
561 | p = p->rb_left; | |
562 | else | |
563 | p = p->rb_right; | |
564 | } | |
565 | ||
43f9b25a AB |
566 | return -ENODEV; |
567 | ||
568 | found: | |
801c135c AB |
569 | ubi_assert(pe->e->pnum == pnum); |
570 | rb_erase(&pe->rb_aec, &ubi->prot.aec); | |
571 | rb_erase(&pe->rb_pnum, &ubi->prot.pnum); | |
572 | kfree(pe); | |
43f9b25a | 573 | return 0; |
801c135c AB |
574 | } |
575 | ||
576 | /** | |
577 | * sync_erase - synchronously erase a physical eraseblock. | |
578 | * @ubi: UBI device description object | |
579 | * @e: the the physical eraseblock to erase | |
580 | * @torture: if the physical eraseblock has to be tortured | |
581 | * | |
582 | * This function returns zero in case of success and a negative error code in | |
583 | * case of failure. | |
584 | */ | |
585 | static int sync_erase(struct ubi_device *ubi, struct ubi_wl_entry *e, int torture) | |
586 | { | |
587 | int err; | |
588 | struct ubi_ec_hdr *ec_hdr; | |
589 | unsigned long long ec = e->ec; | |
590 | ||
591 | dbg_wl("erase PEB %d, old EC %llu", e->pnum, ec); | |
592 | ||
593 | err = paranoid_check_ec(ubi, e->pnum, e->ec); | |
594 | if (err > 0) | |
595 | return -EINVAL; | |
596 | ||
33818bbb | 597 | ec_hdr = kzalloc(ubi->ec_hdr_alsize, GFP_NOFS); |
801c135c AB |
598 | if (!ec_hdr) |
599 | return -ENOMEM; | |
600 | ||
601 | err = ubi_io_sync_erase(ubi, e->pnum, torture); | |
602 | if (err < 0) | |
603 | goto out_free; | |
604 | ||
605 | ec += err; | |
606 | if (ec > UBI_MAX_ERASECOUNTER) { | |
607 | /* | |
608 | * Erase counter overflow. Upgrade UBI and use 64-bit | |
609 | * erase counters internally. | |
610 | */ | |
611 | ubi_err("erase counter overflow at PEB %d, EC %llu", | |
612 | e->pnum, ec); | |
613 | err = -EINVAL; | |
614 | goto out_free; | |
615 | } | |
616 | ||
617 | dbg_wl("erased PEB %d, new EC %llu", e->pnum, ec); | |
618 | ||
3261ebd7 | 619 | ec_hdr->ec = cpu_to_be64(ec); |
801c135c AB |
620 | |
621 | err = ubi_io_write_ec_hdr(ubi, e->pnum, ec_hdr); | |
622 | if (err) | |
623 | goto out_free; | |
624 | ||
625 | e->ec = ec; | |
626 | spin_lock(&ubi->wl_lock); | |
627 | if (e->ec > ubi->max_ec) | |
628 | ubi->max_ec = e->ec; | |
629 | spin_unlock(&ubi->wl_lock); | |
630 | ||
631 | out_free: | |
632 | kfree(ec_hdr); | |
633 | return err; | |
634 | } | |
635 | ||
636 | /** | |
637 | * check_protection_over - check if it is time to stop protecting some | |
638 | * physical eraseblocks. | |
639 | * @ubi: UBI device description object | |
640 | * | |
641 | * This function is called after each erase operation, when the absolute erase | |
642 | * counter is incremented, to check if some physical eraseblock have not to be | |
643 | * protected any longer. These physical eraseblocks are moved from the | |
644 | * protection trees to the used tree. | |
645 | */ | |
646 | static void check_protection_over(struct ubi_device *ubi) | |
647 | { | |
648 | struct ubi_wl_prot_entry *pe; | |
649 | ||
650 | /* | |
651 | * There may be several protected physical eraseblock to remove, | |
652 | * process them all. | |
653 | */ | |
654 | while (1) { | |
655 | spin_lock(&ubi->wl_lock); | |
5abde384 | 656 | if (!ubi->prot.aec.rb_node) { |
801c135c AB |
657 | spin_unlock(&ubi->wl_lock); |
658 | break; | |
659 | } | |
660 | ||
661 | pe = rb_entry(rb_first(&ubi->prot.aec), | |
662 | struct ubi_wl_prot_entry, rb_aec); | |
663 | ||
664 | if (pe->abs_ec > ubi->abs_ec) { | |
665 | spin_unlock(&ubi->wl_lock); | |
666 | break; | |
667 | } | |
668 | ||
669 | dbg_wl("PEB %d protection over, abs_ec %llu, PEB abs_ec %llu", | |
670 | pe->e->pnum, ubi->abs_ec, pe->abs_ec); | |
671 | rb_erase(&pe->rb_aec, &ubi->prot.aec); | |
672 | rb_erase(&pe->rb_pnum, &ubi->prot.pnum); | |
5abde384 | 673 | wl_tree_add(pe->e, &ubi->used); |
801c135c AB |
674 | spin_unlock(&ubi->wl_lock); |
675 | ||
676 | kfree(pe); | |
677 | cond_resched(); | |
678 | } | |
679 | } | |
680 | ||
681 | /** | |
682 | * schedule_ubi_work - schedule a work. | |
683 | * @ubi: UBI device description object | |
684 | * @wrk: the work to schedule | |
685 | * | |
686 | * This function enqueues a work defined by @wrk to the tail of the pending | |
687 | * works list. | |
688 | */ | |
689 | static void schedule_ubi_work(struct ubi_device *ubi, struct ubi_work *wrk) | |
690 | { | |
691 | spin_lock(&ubi->wl_lock); | |
692 | list_add_tail(&wrk->list, &ubi->works); | |
693 | ubi_assert(ubi->works_count >= 0); | |
694 | ubi->works_count += 1; | |
695 | if (ubi->thread_enabled) | |
696 | wake_up_process(ubi->bgt_thread); | |
697 | spin_unlock(&ubi->wl_lock); | |
698 | } | |
699 | ||
700 | static int erase_worker(struct ubi_device *ubi, struct ubi_work *wl_wrk, | |
701 | int cancel); | |
702 | ||
703 | /** | |
704 | * schedule_erase - schedule an erase work. | |
705 | * @ubi: UBI device description object | |
706 | * @e: the WL entry of the physical eraseblock to erase | |
707 | * @torture: if the physical eraseblock has to be tortured | |
708 | * | |
709 | * This function returns zero in case of success and a %-ENOMEM in case of | |
710 | * failure. | |
711 | */ | |
712 | static int schedule_erase(struct ubi_device *ubi, struct ubi_wl_entry *e, | |
713 | int torture) | |
714 | { | |
715 | struct ubi_work *wl_wrk; | |
716 | ||
717 | dbg_wl("schedule erasure of PEB %d, EC %d, torture %d", | |
718 | e->pnum, e->ec, torture); | |
719 | ||
33818bbb | 720 | wl_wrk = kmalloc(sizeof(struct ubi_work), GFP_NOFS); |
801c135c AB |
721 | if (!wl_wrk) |
722 | return -ENOMEM; | |
723 | ||
724 | wl_wrk->func = &erase_worker; | |
725 | wl_wrk->e = e; | |
726 | wl_wrk->torture = torture; | |
727 | ||
728 | schedule_ubi_work(ubi, wl_wrk); | |
729 | return 0; | |
730 | } | |
731 | ||
732 | /** | |
733 | * wear_leveling_worker - wear-leveling worker function. | |
734 | * @ubi: UBI device description object | |
735 | * @wrk: the work object | |
736 | * @cancel: non-zero if the worker has to free memory and exit | |
737 | * | |
738 | * This function copies a more worn out physical eraseblock to a less worn out | |
739 | * one. Returns zero in case of success and a negative error code in case of | |
740 | * failure. | |
741 | */ | |
742 | static int wear_leveling_worker(struct ubi_device *ubi, struct ubi_work *wrk, | |
743 | int cancel) | |
744 | { | |
43f9b25a | 745 | int err, put = 0, scrubbing = 0, protect = 0; |
c18a8418 | 746 | struct ubi_wl_prot_entry *uninitialized_var(pe); |
801c135c AB |
747 | struct ubi_wl_entry *e1, *e2; |
748 | struct ubi_vid_hdr *vid_hdr; | |
749 | ||
750 | kfree(wrk); | |
751 | ||
752 | if (cancel) | |
753 | return 0; | |
754 | ||
33818bbb | 755 | vid_hdr = ubi_zalloc_vid_hdr(ubi, GFP_NOFS); |
801c135c AB |
756 | if (!vid_hdr) |
757 | return -ENOMEM; | |
758 | ||
43f9b25a | 759 | mutex_lock(&ubi->move_mutex); |
801c135c | 760 | spin_lock(&ubi->wl_lock); |
43f9b25a AB |
761 | ubi_assert(!ubi->move_from && !ubi->move_to); |
762 | ubi_assert(!ubi->move_to_put); | |
801c135c | 763 | |
43f9b25a | 764 | if (!ubi->free.rb_node || |
5abde384 | 765 | (!ubi->used.rb_node && !ubi->scrub.rb_node)) { |
801c135c | 766 | /* |
43f9b25a AB |
767 | * No free physical eraseblocks? Well, they must be waiting in |
768 | * the queue to be erased. Cancel movement - it will be | |
769 | * triggered again when a free physical eraseblock appears. | |
801c135c AB |
770 | * |
771 | * No used physical eraseblocks? They must be temporarily | |
772 | * protected from being moved. They will be moved to the | |
773 | * @ubi->used tree later and the wear-leveling will be | |
774 | * triggered again. | |
775 | */ | |
776 | dbg_wl("cancel WL, a list is empty: free %d, used %d", | |
5abde384 | 777 | !ubi->free.rb_node, !ubi->used.rb_node); |
43f9b25a | 778 | goto out_cancel; |
801c135c AB |
779 | } |
780 | ||
5abde384 | 781 | if (!ubi->scrub.rb_node) { |
801c135c AB |
782 | /* |
783 | * Now pick the least worn-out used physical eraseblock and a | |
784 | * highly worn-out free physical eraseblock. If the erase | |
785 | * counters differ much enough, start wear-leveling. | |
786 | */ | |
787 | e1 = rb_entry(rb_first(&ubi->used), struct ubi_wl_entry, rb); | |
788 | e2 = find_wl_entry(&ubi->free, WL_FREE_MAX_DIFF); | |
789 | ||
790 | if (!(e2->ec - e1->ec >= UBI_WL_THRESHOLD)) { | |
791 | dbg_wl("no WL needed: min used EC %d, max free EC %d", | |
792 | e1->ec, e2->ec); | |
43f9b25a | 793 | goto out_cancel; |
801c135c | 794 | } |
5abde384 AB |
795 | paranoid_check_in_wl_tree(e1, &ubi->used); |
796 | rb_erase(&e1->rb, &ubi->used); | |
801c135c AB |
797 | dbg_wl("move PEB %d EC %d to PEB %d EC %d", |
798 | e1->pnum, e1->ec, e2->pnum, e2->ec); | |
799 | } else { | |
43f9b25a AB |
800 | /* Perform scrubbing */ |
801 | scrubbing = 1; | |
801c135c AB |
802 | e1 = rb_entry(rb_first(&ubi->scrub), struct ubi_wl_entry, rb); |
803 | e2 = find_wl_entry(&ubi->free, WL_FREE_MAX_DIFF); | |
5abde384 | 804 | paranoid_check_in_wl_tree(e1, &ubi->scrub); |
d2c46855 | 805 | rb_erase(&e1->rb, &ubi->scrub); |
801c135c AB |
806 | dbg_wl("scrub PEB %d to PEB %d", e1->pnum, e2->pnum); |
807 | } | |
808 | ||
5abde384 AB |
809 | paranoid_check_in_wl_tree(e2, &ubi->free); |
810 | rb_erase(&e2->rb, &ubi->free); | |
801c135c AB |
811 | ubi->move_from = e1; |
812 | ubi->move_to = e2; | |
813 | spin_unlock(&ubi->wl_lock); | |
814 | ||
815 | /* | |
816 | * Now we are going to copy physical eraseblock @e1->pnum to @e2->pnum. | |
817 | * We so far do not know which logical eraseblock our physical | |
818 | * eraseblock (@e1) belongs to. We have to read the volume identifier | |
819 | * header first. | |
43f9b25a AB |
820 | * |
821 | * Note, we are protected from this PEB being unmapped and erased. The | |
822 | * 'ubi_wl_put_peb()' would wait for moving to be finished if the PEB | |
823 | * which is being moved was unmapped. | |
801c135c AB |
824 | */ |
825 | ||
826 | err = ubi_io_read_vid_hdr(ubi, e1->pnum, vid_hdr, 0); | |
827 | if (err && err != UBI_IO_BITFLIPS) { | |
828 | if (err == UBI_IO_PEB_FREE) { | |
829 | /* | |
830 | * We are trying to move PEB without a VID header. UBI | |
831 | * always write VID headers shortly after the PEB was | |
832 | * given, so we have a situation when it did not have | |
833 | * chance to write it down because it was preempted. | |
834 | * Just re-schedule the work, so that next time it will | |
835 | * likely have the VID header in place. | |
836 | */ | |
837 | dbg_wl("PEB %d has no VID header", e1->pnum); | |
43f9b25a | 838 | goto out_not_moved; |
801c135c | 839 | } |
43f9b25a AB |
840 | |
841 | ubi_err("error %d while reading VID header from PEB %d", | |
842 | err, e1->pnum); | |
843 | if (err > 0) | |
844 | err = -EIO; | |
845 | goto out_error; | |
801c135c AB |
846 | } |
847 | ||
848 | err = ubi_eba_copy_leb(ubi, e1->pnum, e2->pnum, vid_hdr); | |
849 | if (err) { | |
43f9b25a AB |
850 | |
851 | if (err < 0) | |
852 | goto out_error; | |
853 | if (err == 1) | |
854 | goto out_not_moved; | |
855 | ||
856 | /* | |
857 | * For some reason the LEB was not moved - it might be because | |
858 | * the volume is being deleted. We should prevent this PEB from | |
859 | * being selected for wear-levelling movement for some "time", | |
860 | * so put it to the protection tree. | |
861 | */ | |
862 | ||
863 | dbg_wl("cancelled moving PEB %d", e1->pnum); | |
864 | pe = kmalloc(sizeof(struct ubi_wl_prot_entry), GFP_NOFS); | |
865 | if (!pe) { | |
866 | err = -ENOMEM; | |
867 | goto out_error; | |
868 | } | |
869 | ||
870 | protect = 1; | |
801c135c AB |
871 | } |
872 | ||
873 | ubi_free_vid_hdr(ubi, vid_hdr); | |
874 | spin_lock(&ubi->wl_lock); | |
43f9b25a AB |
875 | if (protect) |
876 | prot_tree_add(ubi, e1, pe, protect); | |
801c135c | 877 | if (!ubi->move_to_put) |
5abde384 | 878 | wl_tree_add(e2, &ubi->used); |
801c135c AB |
879 | else |
880 | put = 1; | |
881 | ubi->move_from = ubi->move_to = NULL; | |
43f9b25a | 882 | ubi->move_to_put = ubi->wl_scheduled = 0; |
801c135c AB |
883 | spin_unlock(&ubi->wl_lock); |
884 | ||
885 | if (put) { | |
886 | /* | |
887 | * Well, the target PEB was put meanwhile, schedule it for | |
888 | * erasure. | |
889 | */ | |
890 | dbg_wl("PEB %d was put meanwhile, erase", e2->pnum); | |
891 | err = schedule_erase(ubi, e2, 0); | |
43f9b25a AB |
892 | if (err) |
893 | goto out_error; | |
801c135c AB |
894 | } |
895 | ||
43f9b25a AB |
896 | if (!protect) { |
897 | err = schedule_erase(ubi, e1, 0); | |
898 | if (err) | |
899 | goto out_error; | |
801c135c AB |
900 | } |
901 | ||
43f9b25a | 902 | |
801c135c | 903 | dbg_wl("done"); |
43f9b25a AB |
904 | mutex_unlock(&ubi->move_mutex); |
905 | return 0; | |
801c135c AB |
906 | |
907 | /* | |
43f9b25a AB |
908 | * For some reasons the LEB was not moved, might be an error, might be |
909 | * something else. @e1 was not changed, so return it back. @e2 might | |
910 | * be changed, schedule it for erasure. | |
801c135c | 911 | */ |
43f9b25a | 912 | out_not_moved: |
801c135c AB |
913 | ubi_free_vid_hdr(ubi, vid_hdr); |
914 | spin_lock(&ubi->wl_lock); | |
43f9b25a AB |
915 | if (scrubbing) |
916 | wl_tree_add(e1, &ubi->scrub); | |
801c135c | 917 | else |
5abde384 | 918 | wl_tree_add(e1, &ubi->used); |
801c135c | 919 | ubi->move_from = ubi->move_to = NULL; |
43f9b25a | 920 | ubi->move_to_put = ubi->wl_scheduled = 0; |
801c135c AB |
921 | spin_unlock(&ubi->wl_lock); |
922 | ||
801c135c | 923 | err = schedule_erase(ubi, e2, 0); |
43f9b25a AB |
924 | if (err) |
925 | goto out_error; | |
926 | ||
927 | mutex_unlock(&ubi->move_mutex); | |
928 | return 0; | |
929 | ||
930 | out_error: | |
931 | ubi_err("error %d while moving PEB %d to PEB %d", | |
932 | err, e1->pnum, e2->pnum); | |
801c135c | 933 | |
43f9b25a AB |
934 | ubi_free_vid_hdr(ubi, vid_hdr); |
935 | spin_lock(&ubi->wl_lock); | |
936 | ubi->move_from = ubi->move_to = NULL; | |
937 | ubi->move_to_put = ubi->wl_scheduled = 0; | |
938 | spin_unlock(&ubi->wl_lock); | |
939 | ||
940 | kmem_cache_free(ubi_wl_entry_slab, e1); | |
941 | kmem_cache_free(ubi_wl_entry_slab, e2); | |
942 | ubi_ro_mode(ubi); | |
943 | ||
944 | mutex_unlock(&ubi->move_mutex); | |
801c135c | 945 | return err; |
43f9b25a AB |
946 | |
947 | out_cancel: | |
948 | ubi->wl_scheduled = 0; | |
949 | spin_unlock(&ubi->wl_lock); | |
950 | mutex_unlock(&ubi->move_mutex); | |
951 | ubi_free_vid_hdr(ubi, vid_hdr); | |
952 | return 0; | |
801c135c AB |
953 | } |
954 | ||
955 | /** | |
956 | * ensure_wear_leveling - schedule wear-leveling if it is needed. | |
957 | * @ubi: UBI device description object | |
958 | * | |
959 | * This function checks if it is time to start wear-leveling and schedules it | |
960 | * if yes. This function returns zero in case of success and a negative error | |
961 | * code in case of failure. | |
962 | */ | |
963 | static int ensure_wear_leveling(struct ubi_device *ubi) | |
964 | { | |
965 | int err = 0; | |
966 | struct ubi_wl_entry *e1; | |
967 | struct ubi_wl_entry *e2; | |
968 | struct ubi_work *wrk; | |
969 | ||
970 | spin_lock(&ubi->wl_lock); | |
971 | if (ubi->wl_scheduled) | |
972 | /* Wear-leveling is already in the work queue */ | |
973 | goto out_unlock; | |
974 | ||
975 | /* | |
976 | * If the ubi->scrub tree is not empty, scrubbing is needed, and the | |
977 | * the WL worker has to be scheduled anyway. | |
978 | */ | |
5abde384 AB |
979 | if (!ubi->scrub.rb_node) { |
980 | if (!ubi->used.rb_node || !ubi->free.rb_node) | |
801c135c AB |
981 | /* No physical eraseblocks - no deal */ |
982 | goto out_unlock; | |
983 | ||
984 | /* | |
985 | * We schedule wear-leveling only if the difference between the | |
986 | * lowest erase counter of used physical eraseblocks and a high | |
987 | * erase counter of free physical eraseblocks is greater then | |
988 | * %UBI_WL_THRESHOLD. | |
989 | */ | |
990 | e1 = rb_entry(rb_first(&ubi->used), struct ubi_wl_entry, rb); | |
991 | e2 = find_wl_entry(&ubi->free, WL_FREE_MAX_DIFF); | |
992 | ||
993 | if (!(e2->ec - e1->ec >= UBI_WL_THRESHOLD)) | |
994 | goto out_unlock; | |
995 | dbg_wl("schedule wear-leveling"); | |
996 | } else | |
997 | dbg_wl("schedule scrubbing"); | |
998 | ||
999 | ubi->wl_scheduled = 1; | |
1000 | spin_unlock(&ubi->wl_lock); | |
1001 | ||
33818bbb | 1002 | wrk = kmalloc(sizeof(struct ubi_work), GFP_NOFS); |
801c135c AB |
1003 | if (!wrk) { |
1004 | err = -ENOMEM; | |
1005 | goto out_cancel; | |
1006 | } | |
1007 | ||
1008 | wrk->func = &wear_leveling_worker; | |
1009 | schedule_ubi_work(ubi, wrk); | |
1010 | return err; | |
1011 | ||
1012 | out_cancel: | |
1013 | spin_lock(&ubi->wl_lock); | |
1014 | ubi->wl_scheduled = 0; | |
1015 | out_unlock: | |
1016 | spin_unlock(&ubi->wl_lock); | |
1017 | return err; | |
1018 | } | |
1019 | ||
1020 | /** | |
1021 | * erase_worker - physical eraseblock erase worker function. | |
1022 | * @ubi: UBI device description object | |
1023 | * @wl_wrk: the work object | |
1024 | * @cancel: non-zero if the worker has to free memory and exit | |
1025 | * | |
1026 | * This function erases a physical eraseblock and perform torture testing if | |
1027 | * needed. It also takes care about marking the physical eraseblock bad if | |
1028 | * needed. Returns zero in case of success and a negative error code in case of | |
1029 | * failure. | |
1030 | */ | |
1031 | static int erase_worker(struct ubi_device *ubi, struct ubi_work *wl_wrk, | |
1032 | int cancel) | |
1033 | { | |
801c135c | 1034 | struct ubi_wl_entry *e = wl_wrk->e; |
784c1454 | 1035 | int pnum = e->pnum, err, need; |
801c135c AB |
1036 | |
1037 | if (cancel) { | |
1038 | dbg_wl("cancel erasure of PEB %d EC %d", pnum, e->ec); | |
1039 | kfree(wl_wrk); | |
06b68ba1 | 1040 | kmem_cache_free(ubi_wl_entry_slab, e); |
801c135c AB |
1041 | return 0; |
1042 | } | |
1043 | ||
1044 | dbg_wl("erase PEB %d EC %d", pnum, e->ec); | |
1045 | ||
1046 | err = sync_erase(ubi, e, wl_wrk->torture); | |
1047 | if (!err) { | |
1048 | /* Fine, we've erased it successfully */ | |
1049 | kfree(wl_wrk); | |
1050 | ||
1051 | spin_lock(&ubi->wl_lock); | |
1052 | ubi->abs_ec += 1; | |
5abde384 | 1053 | wl_tree_add(e, &ubi->free); |
801c135c AB |
1054 | spin_unlock(&ubi->wl_lock); |
1055 | ||
1056 | /* | |
1057 | * One more erase operation has happened, take care about protected | |
1058 | * physical eraseblocks. | |
1059 | */ | |
1060 | check_protection_over(ubi); | |
1061 | ||
1062 | /* And take care about wear-leveling */ | |
1063 | err = ensure_wear_leveling(ubi); | |
1064 | return err; | |
1065 | } | |
1066 | ||
8d2d4011 | 1067 | ubi_err("failed to erase PEB %d, error %d", pnum, err); |
801c135c | 1068 | kfree(wl_wrk); |
06b68ba1 | 1069 | kmem_cache_free(ubi_wl_entry_slab, e); |
801c135c | 1070 | |
784c1454 AB |
1071 | if (err == -EINTR || err == -ENOMEM || err == -EAGAIN || |
1072 | err == -EBUSY) { | |
1073 | int err1; | |
1074 | ||
1075 | /* Re-schedule the LEB for erasure */ | |
1076 | err1 = schedule_erase(ubi, e, 0); | |
1077 | if (err1) { | |
1078 | err = err1; | |
1079 | goto out_ro; | |
1080 | } | |
1081 | return err; | |
1082 | } else if (err != -EIO) { | |
801c135c AB |
1083 | /* |
1084 | * If this is not %-EIO, we have no idea what to do. Scheduling | |
1085 | * this physical eraseblock for erasure again would cause | |
1086 | * errors again and again. Well, lets switch to RO mode. | |
1087 | */ | |
784c1454 | 1088 | goto out_ro; |
801c135c AB |
1089 | } |
1090 | ||
1091 | /* It is %-EIO, the PEB went bad */ | |
1092 | ||
1093 | if (!ubi->bad_allowed) { | |
1094 | ubi_err("bad physical eraseblock %d detected", pnum); | |
784c1454 AB |
1095 | goto out_ro; |
1096 | } | |
801c135c | 1097 | |
784c1454 AB |
1098 | spin_lock(&ubi->volumes_lock); |
1099 | need = ubi->beb_rsvd_level - ubi->beb_rsvd_pebs + 1; | |
1100 | if (need > 0) { | |
1101 | need = ubi->avail_pebs >= need ? need : ubi->avail_pebs; | |
1102 | ubi->avail_pebs -= need; | |
1103 | ubi->rsvd_pebs += need; | |
1104 | ubi->beb_rsvd_pebs += need; | |
1105 | if (need > 0) | |
1106 | ubi_msg("reserve more %d PEBs", need); | |
1107 | } | |
801c135c | 1108 | |
784c1454 | 1109 | if (ubi->beb_rsvd_pebs == 0) { |
801c135c | 1110 | spin_unlock(&ubi->volumes_lock); |
784c1454 AB |
1111 | ubi_err("no reserved physical eraseblocks"); |
1112 | goto out_ro; | |
1113 | } | |
801c135c | 1114 | |
784c1454 AB |
1115 | spin_unlock(&ubi->volumes_lock); |
1116 | ubi_msg("mark PEB %d as bad", pnum); | |
801c135c | 1117 | |
784c1454 AB |
1118 | err = ubi_io_mark_bad(ubi, pnum); |
1119 | if (err) | |
1120 | goto out_ro; | |
1121 | ||
1122 | spin_lock(&ubi->volumes_lock); | |
1123 | ubi->beb_rsvd_pebs -= 1; | |
1124 | ubi->bad_peb_count += 1; | |
1125 | ubi->good_peb_count -= 1; | |
1126 | ubi_calculate_reserved(ubi); | |
1127 | if (ubi->beb_rsvd_pebs == 0) | |
1128 | ubi_warn("last PEB from the reserved pool was used"); | |
1129 | spin_unlock(&ubi->volumes_lock); | |
1130 | ||
1131 | return err; | |
801c135c | 1132 | |
784c1454 AB |
1133 | out_ro: |
1134 | ubi_ro_mode(ubi); | |
801c135c AB |
1135 | return err; |
1136 | } | |
1137 | ||
1138 | /** | |
43f9b25a | 1139 | * ubi_wl_put_peb - return a physical eraseblock to the wear-leveling unit. |
801c135c AB |
1140 | * @ubi: UBI device description object |
1141 | * @pnum: physical eraseblock to return | |
1142 | * @torture: if this physical eraseblock has to be tortured | |
1143 | * | |
1144 | * This function is called to return physical eraseblock @pnum to the pool of | |
1145 | * free physical eraseblocks. The @torture flag has to be set if an I/O error | |
1146 | * occurred to this @pnum and it has to be tested. This function returns zero | |
43f9b25a | 1147 | * in case of success, and a negative error code in case of failure. |
801c135c AB |
1148 | */ |
1149 | int ubi_wl_put_peb(struct ubi_device *ubi, int pnum, int torture) | |
1150 | { | |
1151 | int err; | |
1152 | struct ubi_wl_entry *e; | |
1153 | ||
1154 | dbg_wl("PEB %d", pnum); | |
1155 | ubi_assert(pnum >= 0); | |
1156 | ubi_assert(pnum < ubi->peb_count); | |
1157 | ||
43f9b25a | 1158 | retry: |
801c135c | 1159 | spin_lock(&ubi->wl_lock); |
801c135c AB |
1160 | e = ubi->lookuptbl[pnum]; |
1161 | if (e == ubi->move_from) { | |
1162 | /* | |
1163 | * User is putting the physical eraseblock which was selected to | |
1164 | * be moved. It will be scheduled for erasure in the | |
1165 | * wear-leveling worker. | |
1166 | */ | |
43f9b25a | 1167 | dbg_wl("PEB %d is being moved, wait", pnum); |
801c135c | 1168 | spin_unlock(&ubi->wl_lock); |
43f9b25a AB |
1169 | |
1170 | /* Wait for the WL worker by taking the @ubi->move_mutex */ | |
1171 | mutex_lock(&ubi->move_mutex); | |
1172 | mutex_unlock(&ubi->move_mutex); | |
1173 | goto retry; | |
801c135c AB |
1174 | } else if (e == ubi->move_to) { |
1175 | /* | |
1176 | * User is putting the physical eraseblock which was selected | |
1177 | * as the target the data is moved to. It may happen if the EBA | |
d2c46855 AB |
1178 | * unit already re-mapped the LEB in 'ubi_eba_copy_leb()' but |
1179 | * the WL unit has not put the PEB to the "used" tree yet, but | |
1180 | * it is about to do this. So we just set a flag which will | |
1181 | * tell the WL worker that the PEB is not needed anymore and | |
593dd33c | 1182 | * should be scheduled for erasure. |
801c135c AB |
1183 | */ |
1184 | dbg_wl("PEB %d is the target of data moving", pnum); | |
1185 | ubi_assert(!ubi->move_to_put); | |
1186 | ubi->move_to_put = 1; | |
1187 | spin_unlock(&ubi->wl_lock); | |
1188 | return 0; | |
1189 | } else { | |
5abde384 AB |
1190 | if (in_wl_tree(e, &ubi->used)) { |
1191 | paranoid_check_in_wl_tree(e, &ubi->used); | |
1192 | rb_erase(&e->rb, &ubi->used); | |
1193 | } else if (in_wl_tree(e, &ubi->scrub)) { | |
1194 | paranoid_check_in_wl_tree(e, &ubi->scrub); | |
1195 | rb_erase(&e->rb, &ubi->scrub); | |
43f9b25a AB |
1196 | } else { |
1197 | err = prot_tree_del(ubi, e->pnum); | |
1198 | if (err) { | |
1199 | ubi_err("PEB %d not found", pnum); | |
1200 | ubi_ro_mode(ubi); | |
1201 | spin_unlock(&ubi->wl_lock); | |
1202 | return err; | |
1203 | } | |
1204 | } | |
801c135c AB |
1205 | } |
1206 | spin_unlock(&ubi->wl_lock); | |
1207 | ||
1208 | err = schedule_erase(ubi, e, torture); | |
1209 | if (err) { | |
1210 | spin_lock(&ubi->wl_lock); | |
5abde384 | 1211 | wl_tree_add(e, &ubi->used); |
801c135c AB |
1212 | spin_unlock(&ubi->wl_lock); |
1213 | } | |
1214 | ||
1215 | return err; | |
1216 | } | |
1217 | ||
1218 | /** | |
1219 | * ubi_wl_scrub_peb - schedule a physical eraseblock for scrubbing. | |
1220 | * @ubi: UBI device description object | |
1221 | * @pnum: the physical eraseblock to schedule | |
1222 | * | |
1223 | * If a bit-flip in a physical eraseblock is detected, this physical eraseblock | |
1224 | * needs scrubbing. This function schedules a physical eraseblock for | |
1225 | * scrubbing which is done in background. This function returns zero in case of | |
1226 | * success and a negative error code in case of failure. | |
1227 | */ | |
1228 | int ubi_wl_scrub_peb(struct ubi_device *ubi, int pnum) | |
1229 | { | |
1230 | struct ubi_wl_entry *e; | |
1231 | ||
1232 | ubi_msg("schedule PEB %d for scrubbing", pnum); | |
1233 | ||
1234 | retry: | |
1235 | spin_lock(&ubi->wl_lock); | |
1236 | e = ubi->lookuptbl[pnum]; | |
1237 | if (e == ubi->move_from || in_wl_tree(e, &ubi->scrub)) { | |
1238 | spin_unlock(&ubi->wl_lock); | |
1239 | return 0; | |
1240 | } | |
1241 | ||
1242 | if (e == ubi->move_to) { | |
1243 | /* | |
1244 | * This physical eraseblock was used to move data to. The data | |
1245 | * was moved but the PEB was not yet inserted to the proper | |
1246 | * tree. We should just wait a little and let the WL worker | |
1247 | * proceed. | |
1248 | */ | |
1249 | spin_unlock(&ubi->wl_lock); | |
1250 | dbg_wl("the PEB %d is not in proper tree, retry", pnum); | |
1251 | yield(); | |
1252 | goto retry; | |
1253 | } | |
1254 | ||
5abde384 AB |
1255 | if (in_wl_tree(e, &ubi->used)) { |
1256 | paranoid_check_in_wl_tree(e, &ubi->used); | |
1257 | rb_erase(&e->rb, &ubi->used); | |
43f9b25a AB |
1258 | } else { |
1259 | int err; | |
1260 | ||
1261 | err = prot_tree_del(ubi, e->pnum); | |
1262 | if (err) { | |
1263 | ubi_err("PEB %d not found", pnum); | |
1264 | ubi_ro_mode(ubi); | |
1265 | spin_unlock(&ubi->wl_lock); | |
1266 | return err; | |
1267 | } | |
1268 | } | |
801c135c | 1269 | |
5abde384 | 1270 | wl_tree_add(e, &ubi->scrub); |
801c135c AB |
1271 | spin_unlock(&ubi->wl_lock); |
1272 | ||
1273 | /* | |
1274 | * Technically scrubbing is the same as wear-leveling, so it is done | |
1275 | * by the WL worker. | |
1276 | */ | |
1277 | return ensure_wear_leveling(ubi); | |
1278 | } | |
1279 | ||
1280 | /** | |
1281 | * ubi_wl_flush - flush all pending works. | |
1282 | * @ubi: UBI device description object | |
1283 | * | |
1284 | * This function returns zero in case of success and a negative error code in | |
1285 | * case of failure. | |
1286 | */ | |
1287 | int ubi_wl_flush(struct ubi_device *ubi) | |
1288 | { | |
593dd33c | 1289 | int err; |
801c135c AB |
1290 | |
1291 | /* | |
1292 | * Erase while the pending works queue is not empty, but not more then | |
1293 | * the number of currently pending works. | |
1294 | */ | |
593dd33c AB |
1295 | dbg_wl("flush (%d pending works)", ubi->works_count); |
1296 | while (ubi->works_count) { | |
1297 | err = do_work(ubi); | |
1298 | if (err) | |
1299 | return err; | |
1300 | } | |
1301 | ||
1302 | /* | |
1303 | * Make sure all the works which have been done in parallel are | |
1304 | * finished. | |
1305 | */ | |
1306 | down_write(&ubi->work_sem); | |
1307 | up_write(&ubi->work_sem); | |
1308 | ||
1309 | /* | |
1310 | * And in case last was the WL worker and it cancelled the LEB | |
1311 | * movement, flush again. | |
1312 | */ | |
1313 | while (ubi->works_count) { | |
1314 | dbg_wl("flush more (%d pending works)", ubi->works_count); | |
801c135c AB |
1315 | err = do_work(ubi); |
1316 | if (err) | |
1317 | return err; | |
1318 | } | |
1319 | ||
1320 | return 0; | |
1321 | } | |
1322 | ||
1323 | /** | |
1324 | * tree_destroy - destroy an RB-tree. | |
1325 | * @root: the root of the tree to destroy | |
1326 | */ | |
1327 | static void tree_destroy(struct rb_root *root) | |
1328 | { | |
1329 | struct rb_node *rb; | |
1330 | struct ubi_wl_entry *e; | |
1331 | ||
1332 | rb = root->rb_node; | |
1333 | while (rb) { | |
1334 | if (rb->rb_left) | |
1335 | rb = rb->rb_left; | |
1336 | else if (rb->rb_right) | |
1337 | rb = rb->rb_right; | |
1338 | else { | |
1339 | e = rb_entry(rb, struct ubi_wl_entry, rb); | |
1340 | ||
1341 | rb = rb_parent(rb); | |
1342 | if (rb) { | |
1343 | if (rb->rb_left == &e->rb) | |
1344 | rb->rb_left = NULL; | |
1345 | else | |
1346 | rb->rb_right = NULL; | |
1347 | } | |
1348 | ||
06b68ba1 | 1349 | kmem_cache_free(ubi_wl_entry_slab, e); |
801c135c AB |
1350 | } |
1351 | } | |
1352 | } | |
1353 | ||
1354 | /** | |
1355 | * ubi_thread - UBI background thread. | |
1356 | * @u: the UBI device description object pointer | |
1357 | */ | |
cdfa788a | 1358 | int ubi_thread(void *u) |
801c135c AB |
1359 | { |
1360 | int failures = 0; | |
1361 | struct ubi_device *ubi = u; | |
1362 | ||
1363 | ubi_msg("background thread \"%s\" started, PID %d", | |
ba25f9dc | 1364 | ubi->bgt_name, task_pid_nr(current)); |
801c135c | 1365 | |
83144186 | 1366 | set_freezable(); |
801c135c AB |
1367 | for (;;) { |
1368 | int err; | |
1369 | ||
1370 | if (kthread_should_stop()) | |
1371 | goto out; | |
1372 | ||
1373 | if (try_to_freeze()) | |
1374 | continue; | |
1375 | ||
1376 | spin_lock(&ubi->wl_lock); | |
1377 | if (list_empty(&ubi->works) || ubi->ro_mode || | |
1378 | !ubi->thread_enabled) { | |
1379 | set_current_state(TASK_INTERRUPTIBLE); | |
1380 | spin_unlock(&ubi->wl_lock); | |
1381 | schedule(); | |
1382 | continue; | |
1383 | } | |
1384 | spin_unlock(&ubi->wl_lock); | |
1385 | ||
1386 | err = do_work(ubi); | |
1387 | if (err) { | |
1388 | ubi_err("%s: work failed with error code %d", | |
1389 | ubi->bgt_name, err); | |
1390 | if (failures++ > WL_MAX_FAILURES) { | |
1391 | /* | |
1392 | * Too many failures, disable the thread and | |
1393 | * switch to read-only mode. | |
1394 | */ | |
1395 | ubi_msg("%s: %d consecutive failures", | |
1396 | ubi->bgt_name, WL_MAX_FAILURES); | |
1397 | ubi_ro_mode(ubi); | |
1398 | break; | |
1399 | } | |
1400 | } else | |
1401 | failures = 0; | |
1402 | ||
1403 | cond_resched(); | |
1404 | } | |
1405 | ||
1406 | out: | |
1407 | dbg_wl("background thread \"%s\" is killed", ubi->bgt_name); | |
1408 | return 0; | |
1409 | } | |
1410 | ||
1411 | /** | |
1412 | * cancel_pending - cancel all pending works. | |
1413 | * @ubi: UBI device description object | |
1414 | */ | |
1415 | static void cancel_pending(struct ubi_device *ubi) | |
1416 | { | |
1417 | while (!list_empty(&ubi->works)) { | |
1418 | struct ubi_work *wrk; | |
1419 | ||
1420 | wrk = list_entry(ubi->works.next, struct ubi_work, list); | |
1421 | list_del(&wrk->list); | |
1422 | wrk->func(ubi, wrk, 1); | |
1423 | ubi->works_count -= 1; | |
1424 | ubi_assert(ubi->works_count >= 0); | |
1425 | } | |
1426 | } | |
1427 | ||
1428 | /** | |
1429 | * ubi_wl_init_scan - initialize the wear-leveling unit using scanning | |
1430 | * information. | |
1431 | * @ubi: UBI device description object | |
1432 | * @si: scanning information | |
1433 | * | |
1434 | * This function returns zero in case of success, and a negative error code in | |
1435 | * case of failure. | |
1436 | */ | |
1437 | int ubi_wl_init_scan(struct ubi_device *ubi, struct ubi_scan_info *si) | |
1438 | { | |
1439 | int err; | |
1440 | struct rb_node *rb1, *rb2; | |
1441 | struct ubi_scan_volume *sv; | |
1442 | struct ubi_scan_leb *seb, *tmp; | |
1443 | struct ubi_wl_entry *e; | |
1444 | ||
1445 | ||
1446 | ubi->used = ubi->free = ubi->scrub = RB_ROOT; | |
1447 | ubi->prot.pnum = ubi->prot.aec = RB_ROOT; | |
1448 | spin_lock_init(&ubi->wl_lock); | |
43f9b25a | 1449 | mutex_init(&ubi->move_mutex); |
593dd33c | 1450 | init_rwsem(&ubi->work_sem); |
801c135c AB |
1451 | ubi->max_ec = si->max_ec; |
1452 | INIT_LIST_HEAD(&ubi->works); | |
1453 | ||
1454 | sprintf(ubi->bgt_name, UBI_BGT_NAME_PATTERN, ubi->ubi_num); | |
1455 | ||
801c135c AB |
1456 | err = -ENOMEM; |
1457 | ubi->lookuptbl = kzalloc(ubi->peb_count * sizeof(void *), GFP_KERNEL); | |
1458 | if (!ubi->lookuptbl) | |
cdfa788a | 1459 | return err; |
801c135c AB |
1460 | |
1461 | list_for_each_entry_safe(seb, tmp, &si->erase, u.list) { | |
1462 | cond_resched(); | |
1463 | ||
06b68ba1 | 1464 | e = kmem_cache_alloc(ubi_wl_entry_slab, GFP_KERNEL); |
801c135c AB |
1465 | if (!e) |
1466 | goto out_free; | |
1467 | ||
1468 | e->pnum = seb->pnum; | |
1469 | e->ec = seb->ec; | |
1470 | ubi->lookuptbl[e->pnum] = e; | |
1471 | if (schedule_erase(ubi, e, 0)) { | |
06b68ba1 | 1472 | kmem_cache_free(ubi_wl_entry_slab, e); |
801c135c AB |
1473 | goto out_free; |
1474 | } | |
1475 | } | |
1476 | ||
1477 | list_for_each_entry(seb, &si->free, u.list) { | |
1478 | cond_resched(); | |
1479 | ||
06b68ba1 | 1480 | e = kmem_cache_alloc(ubi_wl_entry_slab, GFP_KERNEL); |
801c135c AB |
1481 | if (!e) |
1482 | goto out_free; | |
1483 | ||
1484 | e->pnum = seb->pnum; | |
1485 | e->ec = seb->ec; | |
1486 | ubi_assert(e->ec >= 0); | |
5abde384 | 1487 | wl_tree_add(e, &ubi->free); |
801c135c AB |
1488 | ubi->lookuptbl[e->pnum] = e; |
1489 | } | |
1490 | ||
1491 | list_for_each_entry(seb, &si->corr, u.list) { | |
1492 | cond_resched(); | |
1493 | ||
06b68ba1 | 1494 | e = kmem_cache_alloc(ubi_wl_entry_slab, GFP_KERNEL); |
801c135c AB |
1495 | if (!e) |
1496 | goto out_free; | |
1497 | ||
1498 | e->pnum = seb->pnum; | |
1499 | e->ec = seb->ec; | |
1500 | ubi->lookuptbl[e->pnum] = e; | |
1501 | if (schedule_erase(ubi, e, 0)) { | |
06b68ba1 | 1502 | kmem_cache_free(ubi_wl_entry_slab, e); |
801c135c AB |
1503 | goto out_free; |
1504 | } | |
1505 | } | |
1506 | ||
1507 | ubi_rb_for_each_entry(rb1, sv, &si->volumes, rb) { | |
1508 | ubi_rb_for_each_entry(rb2, seb, &sv->root, u.rb) { | |
1509 | cond_resched(); | |
1510 | ||
06b68ba1 | 1511 | e = kmem_cache_alloc(ubi_wl_entry_slab, GFP_KERNEL); |
801c135c AB |
1512 | if (!e) |
1513 | goto out_free; | |
1514 | ||
1515 | e->pnum = seb->pnum; | |
1516 | e->ec = seb->ec; | |
1517 | ubi->lookuptbl[e->pnum] = e; | |
1518 | if (!seb->scrub) { | |
1519 | dbg_wl("add PEB %d EC %d to the used tree", | |
1520 | e->pnum, e->ec); | |
5abde384 | 1521 | wl_tree_add(e, &ubi->used); |
801c135c AB |
1522 | } else { |
1523 | dbg_wl("add PEB %d EC %d to the scrub tree", | |
1524 | e->pnum, e->ec); | |
5abde384 | 1525 | wl_tree_add(e, &ubi->scrub); |
801c135c AB |
1526 | } |
1527 | } | |
1528 | } | |
1529 | ||
5abde384 | 1530 | if (ubi->avail_pebs < WL_RESERVED_PEBS) { |
801c135c AB |
1531 | ubi_err("no enough physical eraseblocks (%d, need %d)", |
1532 | ubi->avail_pebs, WL_RESERVED_PEBS); | |
1533 | goto out_free; | |
1534 | } | |
1535 | ubi->avail_pebs -= WL_RESERVED_PEBS; | |
1536 | ubi->rsvd_pebs += WL_RESERVED_PEBS; | |
1537 | ||
1538 | /* Schedule wear-leveling if needed */ | |
1539 | err = ensure_wear_leveling(ubi); | |
1540 | if (err) | |
1541 | goto out_free; | |
1542 | ||
1543 | return 0; | |
1544 | ||
1545 | out_free: | |
1546 | cancel_pending(ubi); | |
1547 | tree_destroy(&ubi->used); | |
1548 | tree_destroy(&ubi->free); | |
1549 | tree_destroy(&ubi->scrub); | |
1550 | kfree(ubi->lookuptbl); | |
801c135c AB |
1551 | return err; |
1552 | } | |
1553 | ||
1554 | /** | |
1555 | * protection_trees_destroy - destroy the protection RB-trees. | |
1556 | * @ubi: UBI device description object | |
1557 | */ | |
1558 | static void protection_trees_destroy(struct ubi_device *ubi) | |
1559 | { | |
1560 | struct rb_node *rb; | |
1561 | struct ubi_wl_prot_entry *pe; | |
1562 | ||
1563 | rb = ubi->prot.aec.rb_node; | |
1564 | while (rb) { | |
1565 | if (rb->rb_left) | |
1566 | rb = rb->rb_left; | |
1567 | else if (rb->rb_right) | |
1568 | rb = rb->rb_right; | |
1569 | else { | |
1570 | pe = rb_entry(rb, struct ubi_wl_prot_entry, rb_aec); | |
1571 | ||
1572 | rb = rb_parent(rb); | |
1573 | if (rb) { | |
1574 | if (rb->rb_left == &pe->rb_aec) | |
1575 | rb->rb_left = NULL; | |
1576 | else | |
1577 | rb->rb_right = NULL; | |
1578 | } | |
1579 | ||
06b68ba1 | 1580 | kmem_cache_free(ubi_wl_entry_slab, pe->e); |
801c135c AB |
1581 | kfree(pe); |
1582 | } | |
1583 | } | |
1584 | } | |
1585 | ||
1586 | /** | |
1587 | * ubi_wl_close - close the wear-leveling unit. | |
1588 | * @ubi: UBI device description object | |
1589 | */ | |
1590 | void ubi_wl_close(struct ubi_device *ubi) | |
1591 | { | |
801c135c AB |
1592 | dbg_wl("close the UBI wear-leveling unit"); |
1593 | ||
1594 | cancel_pending(ubi); | |
1595 | protection_trees_destroy(ubi); | |
1596 | tree_destroy(&ubi->used); | |
1597 | tree_destroy(&ubi->free); | |
1598 | tree_destroy(&ubi->scrub); | |
1599 | kfree(ubi->lookuptbl); | |
801c135c AB |
1600 | } |
1601 | ||
1602 | #ifdef CONFIG_MTD_UBI_DEBUG_PARANOID | |
1603 | ||
1604 | /** | |
1605 | * paranoid_check_ec - make sure that the erase counter of a physical eraseblock | |
1606 | * is correct. | |
1607 | * @ubi: UBI device description object | |
1608 | * @pnum: the physical eraseblock number to check | |
1609 | * @ec: the erase counter to check | |
1610 | * | |
1611 | * This function returns zero if the erase counter of physical eraseblock @pnum | |
1612 | * is equivalent to @ec, %1 if not, and a negative error code if an error | |
1613 | * occurred. | |
1614 | */ | |
e88d6e10 | 1615 | static int paranoid_check_ec(struct ubi_device *ubi, int pnum, int ec) |
801c135c AB |
1616 | { |
1617 | int err; | |
1618 | long long read_ec; | |
1619 | struct ubi_ec_hdr *ec_hdr; | |
1620 | ||
33818bbb | 1621 | ec_hdr = kzalloc(ubi->ec_hdr_alsize, GFP_NOFS); |
801c135c AB |
1622 | if (!ec_hdr) |
1623 | return -ENOMEM; | |
1624 | ||
1625 | err = ubi_io_read_ec_hdr(ubi, pnum, ec_hdr, 0); | |
1626 | if (err && err != UBI_IO_BITFLIPS) { | |
1627 | /* The header does not have to exist */ | |
1628 | err = 0; | |
1629 | goto out_free; | |
1630 | } | |
1631 | ||
3261ebd7 | 1632 | read_ec = be64_to_cpu(ec_hdr->ec); |
801c135c AB |
1633 | if (ec != read_ec) { |
1634 | ubi_err("paranoid check failed for PEB %d", pnum); | |
1635 | ubi_err("read EC is %lld, should be %d", read_ec, ec); | |
1636 | ubi_dbg_dump_stack(); | |
1637 | err = 1; | |
1638 | } else | |
1639 | err = 0; | |
1640 | ||
1641 | out_free: | |
1642 | kfree(ec_hdr); | |
1643 | return err; | |
1644 | } | |
1645 | ||
1646 | /** | |
1647 | * paranoid_check_in_wl_tree - make sure that a wear-leveling entry is present | |
1648 | * in a WL RB-tree. | |
1649 | * @e: the wear-leveling entry to check | |
1650 | * @root: the root of the tree | |
1651 | * | |
1652 | * This function returns zero if @e is in the @root RB-tree and %1 if it | |
1653 | * is not. | |
1654 | */ | |
1655 | static int paranoid_check_in_wl_tree(struct ubi_wl_entry *e, | |
1656 | struct rb_root *root) | |
1657 | { | |
1658 | if (in_wl_tree(e, root)) | |
1659 | return 0; | |
1660 | ||
1661 | ubi_err("paranoid check failed for PEB %d, EC %d, RB-tree %p ", | |
1662 | e->pnum, e->ec, root); | |
1663 | ubi_dbg_dump_stack(); | |
1664 | return 1; | |
1665 | } | |
1666 | ||
1667 | #endif /* CONFIG_MTD_UBI_DEBUG_PARANOID */ |