OMAP3 SRF: MPU/CORE/PD latency modeling
[linux-ginger.git] / fs / ntfs / runlist.c
blob56a9a6d25a2a8cddde3776a408bbd5b7535aa89c
1 /**
2 * runlist.c - NTFS runlist handling code. Part of the Linux-NTFS project.
4 * Copyright (c) 2001-2007 Anton Altaparmakov
5 * Copyright (c) 2002-2005 Richard Russon
7 * This program/include file is free software; you can redistribute it and/or
8 * modify it under the terms of the GNU General Public License as published
9 * by the Free Software Foundation; either version 2 of the License, or
10 * (at your option) any later version.
12 * This program/include file is distributed in the hope that it will be
13 * useful, but WITHOUT ANY WARRANTY; without even the implied warranty
14 * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 * GNU General Public License for more details.
17 * You should have received a copy of the GNU General Public License
18 * along with this program (in the main directory of the Linux-NTFS
19 * distribution in the file COPYING); if not, write to the Free Software
20 * Foundation,Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
23 #include "debug.h"
24 #include "dir.h"
25 #include "endian.h"
26 #include "malloc.h"
27 #include "ntfs.h"
29 /**
30 * ntfs_rl_mm - runlist memmove
32 * It is up to the caller to serialize access to the runlist @base.
34 static inline void ntfs_rl_mm(runlist_element *base, int dst, int src,
35 int size)
37 if (likely((dst != src) && (size > 0)))
38 memmove(base + dst, base + src, size * sizeof(*base));
41 /**
42 * ntfs_rl_mc - runlist memory copy
44 * It is up to the caller to serialize access to the runlists @dstbase and
45 * @srcbase.
47 static inline void ntfs_rl_mc(runlist_element *dstbase, int dst,
48 runlist_element *srcbase, int src, int size)
50 if (likely(size > 0))
51 memcpy(dstbase + dst, srcbase + src, size * sizeof(*dstbase));
54 /**
55 * ntfs_rl_realloc - Reallocate memory for runlists
56 * @rl: original runlist
57 * @old_size: number of runlist elements in the original runlist @rl
58 * @new_size: number of runlist elements we need space for
60 * As the runlists grow, more memory will be required. To prevent the
61 * kernel having to allocate and reallocate large numbers of small bits of
62 * memory, this function returns an entire page of memory.
64 * It is up to the caller to serialize access to the runlist @rl.
66 * N.B. If the new allocation doesn't require a different number of pages in
67 * memory, the function will return the original pointer.
69 * On success, return a pointer to the newly allocated, or recycled, memory.
70 * On error, return -errno. The following error codes are defined:
71 * -ENOMEM - Not enough memory to allocate runlist array.
72 * -EINVAL - Invalid parameters were passed in.
74 static inline runlist_element *ntfs_rl_realloc(runlist_element *rl,
75 int old_size, int new_size)
77 runlist_element *new_rl;
79 old_size = PAGE_ALIGN(old_size * sizeof(*rl));
80 new_size = PAGE_ALIGN(new_size * sizeof(*rl));
81 if (old_size == new_size)
82 return rl;
84 new_rl = ntfs_malloc_nofs(new_size);
85 if (unlikely(!new_rl))
86 return ERR_PTR(-ENOMEM);
88 if (likely(rl != NULL)) {
89 if (unlikely(old_size > new_size))
90 old_size = new_size;
91 memcpy(new_rl, rl, old_size);
92 ntfs_free(rl);
94 return new_rl;
97 /**
98 * ntfs_rl_realloc_nofail - Reallocate memory for runlists
99 * @rl: original runlist
100 * @old_size: number of runlist elements in the original runlist @rl
101 * @new_size: number of runlist elements we need space for
103 * As the runlists grow, more memory will be required. To prevent the
104 * kernel having to allocate and reallocate large numbers of small bits of
105 * memory, this function returns an entire page of memory.
107 * This function guarantees that the allocation will succeed. It will sleep
108 * for as long as it takes to complete the allocation.
110 * It is up to the caller to serialize access to the runlist @rl.
112 * N.B. If the new allocation doesn't require a different number of pages in
113 * memory, the function will return the original pointer.
115 * On success, return a pointer to the newly allocated, or recycled, memory.
116 * On error, return -errno. The following error codes are defined:
117 * -ENOMEM - Not enough memory to allocate runlist array.
118 * -EINVAL - Invalid parameters were passed in.
120 static inline runlist_element *ntfs_rl_realloc_nofail(runlist_element *rl,
121 int old_size, int new_size)
123 runlist_element *new_rl;
125 old_size = PAGE_ALIGN(old_size * sizeof(*rl));
126 new_size = PAGE_ALIGN(new_size * sizeof(*rl));
127 if (old_size == new_size)
128 return rl;
130 new_rl = ntfs_malloc_nofs_nofail(new_size);
131 BUG_ON(!new_rl);
133 if (likely(rl != NULL)) {
134 if (unlikely(old_size > new_size))
135 old_size = new_size;
136 memcpy(new_rl, rl, old_size);
137 ntfs_free(rl);
139 return new_rl;
143 * ntfs_are_rl_mergeable - test if two runlists can be joined together
144 * @dst: original runlist
145 * @src: new runlist to test for mergeability with @dst
147 * Test if two runlists can be joined together. For this, their VCNs and LCNs
148 * must be adjacent.
150 * It is up to the caller to serialize access to the runlists @dst and @src.
152 * Return: true Success, the runlists can be merged.
153 * false Failure, the runlists cannot be merged.
155 static inline bool ntfs_are_rl_mergeable(runlist_element *dst,
156 runlist_element *src)
158 BUG_ON(!dst);
159 BUG_ON(!src);
161 /* We can merge unmapped regions even if they are misaligned. */
162 if ((dst->lcn == LCN_RL_NOT_MAPPED) && (src->lcn == LCN_RL_NOT_MAPPED))
163 return true;
164 /* If the runs are misaligned, we cannot merge them. */
165 if ((dst->vcn + dst->length) != src->vcn)
166 return false;
167 /* If both runs are non-sparse and contiguous, we can merge them. */
168 if ((dst->lcn >= 0) && (src->lcn >= 0) &&
169 ((dst->lcn + dst->length) == src->lcn))
170 return true;
171 /* If we are merging two holes, we can merge them. */
172 if ((dst->lcn == LCN_HOLE) && (src->lcn == LCN_HOLE))
173 return true;
174 /* Cannot merge. */
175 return false;
179 * __ntfs_rl_merge - merge two runlists without testing if they can be merged
180 * @dst: original, destination runlist
181 * @src: new runlist to merge with @dst
183 * Merge the two runlists, writing into the destination runlist @dst. The
184 * caller must make sure the runlists can be merged or this will corrupt the
185 * destination runlist.
187 * It is up to the caller to serialize access to the runlists @dst and @src.
189 static inline void __ntfs_rl_merge(runlist_element *dst, runlist_element *src)
191 dst->length += src->length;
195 * ntfs_rl_append - append a runlist after a given element
196 * @dst: original runlist to be worked on
197 * @dsize: number of elements in @dst (including end marker)
198 * @src: runlist to be inserted into @dst
199 * @ssize: number of elements in @src (excluding end marker)
200 * @loc: append the new runlist @src after this element in @dst
202 * Append the runlist @src after element @loc in @dst. Merge the right end of
203 * the new runlist, if necessary. Adjust the size of the hole before the
204 * appended runlist.
206 * It is up to the caller to serialize access to the runlists @dst and @src.
208 * On success, return a pointer to the new, combined, runlist. Note, both
209 * runlists @dst and @src are deallocated before returning so you cannot use
210 * the pointers for anything any more. (Strictly speaking the returned runlist
211 * may be the same as @dst but this is irrelevant.)
213 * On error, return -errno. Both runlists are left unmodified. The following
214 * error codes are defined:
215 * -ENOMEM - Not enough memory to allocate runlist array.
216 * -EINVAL - Invalid parameters were passed in.
218 static inline runlist_element *ntfs_rl_append(runlist_element *dst,
219 int dsize, runlist_element *src, int ssize, int loc)
221 bool right = false; /* Right end of @src needs merging. */
222 int marker; /* End of the inserted runs. */
224 BUG_ON(!dst);
225 BUG_ON(!src);
227 /* First, check if the right hand end needs merging. */
228 if ((loc + 1) < dsize)
229 right = ntfs_are_rl_mergeable(src + ssize - 1, dst + loc + 1);
231 /* Space required: @dst size + @src size, less one if we merged. */
232 dst = ntfs_rl_realloc(dst, dsize, dsize + ssize - right);
233 if (IS_ERR(dst))
234 return dst;
236 * We are guaranteed to succeed from here so can start modifying the
237 * original runlists.
240 /* First, merge the right hand end, if necessary. */
241 if (right)
242 __ntfs_rl_merge(src + ssize - 1, dst + loc + 1);
244 /* First run after the @src runs that have been inserted. */
245 marker = loc + ssize + 1;
247 /* Move the tail of @dst out of the way, then copy in @src. */
248 ntfs_rl_mm(dst, marker, loc + 1 + right, dsize - (loc + 1 + right));
249 ntfs_rl_mc(dst, loc + 1, src, 0, ssize);
251 /* Adjust the size of the preceding hole. */
252 dst[loc].length = dst[loc + 1].vcn - dst[loc].vcn;
254 /* We may have changed the length of the file, so fix the end marker */
255 if (dst[marker].lcn == LCN_ENOENT)
256 dst[marker].vcn = dst[marker - 1].vcn + dst[marker - 1].length;
258 return dst;
262 * ntfs_rl_insert - insert a runlist into another
263 * @dst: original runlist to be worked on
264 * @dsize: number of elements in @dst (including end marker)
265 * @src: new runlist to be inserted
266 * @ssize: number of elements in @src (excluding end marker)
267 * @loc: insert the new runlist @src before this element in @dst
269 * Insert the runlist @src before element @loc in the runlist @dst. Merge the
270 * left end of the new runlist, if necessary. Adjust the size of the hole
271 * after the inserted runlist.
273 * It is up to the caller to serialize access to the runlists @dst and @src.
275 * On success, return a pointer to the new, combined, runlist. Note, both
276 * runlists @dst and @src are deallocated before returning so you cannot use
277 * the pointers for anything any more. (Strictly speaking the returned runlist
278 * may be the same as @dst but this is irrelevant.)
280 * On error, return -errno. Both runlists are left unmodified. The following
281 * error codes are defined:
282 * -ENOMEM - Not enough memory to allocate runlist array.
283 * -EINVAL - Invalid parameters were passed in.
285 static inline runlist_element *ntfs_rl_insert(runlist_element *dst,
286 int dsize, runlist_element *src, int ssize, int loc)
288 bool left = false; /* Left end of @src needs merging. */
289 bool disc = false; /* Discontinuity between @dst and @src. */
290 int marker; /* End of the inserted runs. */
292 BUG_ON(!dst);
293 BUG_ON(!src);
296 * disc => Discontinuity between the end of @dst and the start of @src.
297 * This means we might need to insert a "not mapped" run.
299 if (loc == 0)
300 disc = (src[0].vcn > 0);
301 else {
302 s64 merged_length;
304 left = ntfs_are_rl_mergeable(dst + loc - 1, src);
306 merged_length = dst[loc - 1].length;
307 if (left)
308 merged_length += src->length;
310 disc = (src[0].vcn > dst[loc - 1].vcn + merged_length);
313 * Space required: @dst size + @src size, less one if we merged, plus
314 * one if there was a discontinuity.
316 dst = ntfs_rl_realloc(dst, dsize, dsize + ssize - left + disc);
317 if (IS_ERR(dst))
318 return dst;
320 * We are guaranteed to succeed from here so can start modifying the
321 * original runlist.
323 if (left)
324 __ntfs_rl_merge(dst + loc - 1, src);
326 * First run after the @src runs that have been inserted.
327 * Nominally, @marker equals @loc + @ssize, i.e. location + number of
328 * runs in @src. However, if @left, then the first run in @src has
329 * been merged with one in @dst. And if @disc, then @dst and @src do
330 * not meet and we need an extra run to fill the gap.
332 marker = loc + ssize - left + disc;
334 /* Move the tail of @dst out of the way, then copy in @src. */
335 ntfs_rl_mm(dst, marker, loc, dsize - loc);
336 ntfs_rl_mc(dst, loc + disc, src, left, ssize - left);
338 /* Adjust the VCN of the first run after the insertion... */
339 dst[marker].vcn = dst[marker - 1].vcn + dst[marker - 1].length;
340 /* ... and the length. */
341 if (dst[marker].lcn == LCN_HOLE || dst[marker].lcn == LCN_RL_NOT_MAPPED)
342 dst[marker].length = dst[marker + 1].vcn - dst[marker].vcn;
344 /* Writing beyond the end of the file and there is a discontinuity. */
345 if (disc) {
346 if (loc > 0) {
347 dst[loc].vcn = dst[loc - 1].vcn + dst[loc - 1].length;
348 dst[loc].length = dst[loc + 1].vcn - dst[loc].vcn;
349 } else {
350 dst[loc].vcn = 0;
351 dst[loc].length = dst[loc + 1].vcn;
353 dst[loc].lcn = LCN_RL_NOT_MAPPED;
355 return dst;
359 * ntfs_rl_replace - overwrite a runlist element with another runlist
360 * @dst: original runlist to be worked on
361 * @dsize: number of elements in @dst (including end marker)
362 * @src: new runlist to be inserted
363 * @ssize: number of elements in @src (excluding end marker)
364 * @loc: index in runlist @dst to overwrite with @src
366 * Replace the runlist element @dst at @loc with @src. Merge the left and
367 * right ends of the inserted runlist, if necessary.
369 * It is up to the caller to serialize access to the runlists @dst and @src.
371 * On success, return a pointer to the new, combined, runlist. Note, both
372 * runlists @dst and @src are deallocated before returning so you cannot use
373 * the pointers for anything any more. (Strictly speaking the returned runlist
374 * may be the same as @dst but this is irrelevant.)
376 * On error, return -errno. Both runlists are left unmodified. The following
377 * error codes are defined:
378 * -ENOMEM - Not enough memory to allocate runlist array.
379 * -EINVAL - Invalid parameters were passed in.
381 static inline runlist_element *ntfs_rl_replace(runlist_element *dst,
382 int dsize, runlist_element *src, int ssize, int loc)
384 signed delta;
385 bool left = false; /* Left end of @src needs merging. */
386 bool right = false; /* Right end of @src needs merging. */
387 int tail; /* Start of tail of @dst. */
388 int marker; /* End of the inserted runs. */
390 BUG_ON(!dst);
391 BUG_ON(!src);
393 /* First, see if the left and right ends need merging. */
394 if ((loc + 1) < dsize)
395 right = ntfs_are_rl_mergeable(src + ssize - 1, dst + loc + 1);
396 if (loc > 0)
397 left = ntfs_are_rl_mergeable(dst + loc - 1, src);
399 * Allocate some space. We will need less if the left, right, or both
400 * ends get merged. The -1 accounts for the run being replaced.
402 delta = ssize - 1 - left - right;
403 if (delta > 0) {
404 dst = ntfs_rl_realloc(dst, dsize, dsize + delta);
405 if (IS_ERR(dst))
406 return dst;
409 * We are guaranteed to succeed from here so can start modifying the
410 * original runlists.
413 /* First, merge the left and right ends, if necessary. */
414 if (right)
415 __ntfs_rl_merge(src + ssize - 1, dst + loc + 1);
416 if (left)
417 __ntfs_rl_merge(dst + loc - 1, src);
419 * Offset of the tail of @dst. This needs to be moved out of the way
420 * to make space for the runs to be copied from @src, i.e. the first
421 * run of the tail of @dst.
422 * Nominally, @tail equals @loc + 1, i.e. location, skipping the
423 * replaced run. However, if @right, then one of @dst's runs is
424 * already merged into @src.
426 tail = loc + right + 1;
428 * First run after the @src runs that have been inserted, i.e. where
429 * the tail of @dst needs to be moved to.
430 * Nominally, @marker equals @loc + @ssize, i.e. location + number of
431 * runs in @src. However, if @left, then the first run in @src has
432 * been merged with one in @dst.
434 marker = loc + ssize - left;
436 /* Move the tail of @dst out of the way, then copy in @src. */
437 ntfs_rl_mm(dst, marker, tail, dsize - tail);
438 ntfs_rl_mc(dst, loc, src, left, ssize - left);
440 /* We may have changed the length of the file, so fix the end marker. */
441 if (dsize - tail > 0 && dst[marker].lcn == LCN_ENOENT)
442 dst[marker].vcn = dst[marker - 1].vcn + dst[marker - 1].length;
443 return dst;
447 * ntfs_rl_split - insert a runlist into the centre of a hole
448 * @dst: original runlist to be worked on
449 * @dsize: number of elements in @dst (including end marker)
450 * @src: new runlist to be inserted
451 * @ssize: number of elements in @src (excluding end marker)
452 * @loc: index in runlist @dst at which to split and insert @src
454 * Split the runlist @dst at @loc into two and insert @new in between the two
455 * fragments. No merging of runlists is necessary. Adjust the size of the
456 * holes either side.
458 * It is up to the caller to serialize access to the runlists @dst and @src.
460 * On success, return a pointer to the new, combined, runlist. Note, both
461 * runlists @dst and @src are deallocated before returning so you cannot use
462 * the pointers for anything any more. (Strictly speaking the returned runlist
463 * may be the same as @dst but this is irrelevant.)
465 * On error, return -errno. Both runlists are left unmodified. The following
466 * error codes are defined:
467 * -ENOMEM - Not enough memory to allocate runlist array.
468 * -EINVAL - Invalid parameters were passed in.
470 static inline runlist_element *ntfs_rl_split(runlist_element *dst, int dsize,
471 runlist_element *src, int ssize, int loc)
473 BUG_ON(!dst);
474 BUG_ON(!src);
476 /* Space required: @dst size + @src size + one new hole. */
477 dst = ntfs_rl_realloc(dst, dsize, dsize + ssize + 1);
478 if (IS_ERR(dst))
479 return dst;
481 * We are guaranteed to succeed from here so can start modifying the
482 * original runlists.
485 /* Move the tail of @dst out of the way, then copy in @src. */
486 ntfs_rl_mm(dst, loc + 1 + ssize, loc, dsize - loc);
487 ntfs_rl_mc(dst, loc + 1, src, 0, ssize);
489 /* Adjust the size of the holes either size of @src. */
490 dst[loc].length = dst[loc+1].vcn - dst[loc].vcn;
491 dst[loc+ssize+1].vcn = dst[loc+ssize].vcn + dst[loc+ssize].length;
492 dst[loc+ssize+1].length = dst[loc+ssize+2].vcn - dst[loc+ssize+1].vcn;
494 return dst;
498 * ntfs_runlists_merge - merge two runlists into one
499 * @drl: original runlist to be worked on
500 * @srl: new runlist to be merged into @drl
502 * First we sanity check the two runlists @srl and @drl to make sure that they
503 * are sensible and can be merged. The runlist @srl must be either after the
504 * runlist @drl or completely within a hole (or unmapped region) in @drl.
506 * It is up to the caller to serialize access to the runlists @drl and @srl.
508 * Merging of runlists is necessary in two cases:
509 * 1. When attribute lists are used and a further extent is being mapped.
510 * 2. When new clusters are allocated to fill a hole or extend a file.
512 * There are four possible ways @srl can be merged. It can:
513 * - be inserted at the beginning of a hole,
514 * - split the hole in two and be inserted between the two fragments,
515 * - be appended at the end of a hole, or it can
516 * - replace the whole hole.
517 * It can also be appended to the end of the runlist, which is just a variant
518 * of the insert case.
520 * On success, return a pointer to the new, combined, runlist. Note, both
521 * runlists @drl and @srl are deallocated before returning so you cannot use
522 * the pointers for anything any more. (Strictly speaking the returned runlist
523 * may be the same as @dst but this is irrelevant.)
525 * On error, return -errno. Both runlists are left unmodified. The following
526 * error codes are defined:
527 * -ENOMEM - Not enough memory to allocate runlist array.
528 * -EINVAL - Invalid parameters were passed in.
529 * -ERANGE - The runlists overlap and cannot be merged.
531 runlist_element *ntfs_runlists_merge(runlist_element *drl,
532 runlist_element *srl)
534 int di, si; /* Current index into @[ds]rl. */
535 int sstart; /* First index with lcn > LCN_RL_NOT_MAPPED. */
536 int dins; /* Index into @drl at which to insert @srl. */
537 int dend, send; /* Last index into @[ds]rl. */
538 int dfinal, sfinal; /* The last index into @[ds]rl with
539 lcn >= LCN_HOLE. */
540 int marker = 0;
541 VCN marker_vcn = 0;
543 #ifdef DEBUG
544 ntfs_debug("dst:");
545 ntfs_debug_dump_runlist(drl);
546 ntfs_debug("src:");
547 ntfs_debug_dump_runlist(srl);
548 #endif
550 /* Check for silly calling... */
551 if (unlikely(!srl))
552 return drl;
553 if (IS_ERR(srl) || IS_ERR(drl))
554 return ERR_PTR(-EINVAL);
556 /* Check for the case where the first mapping is being done now. */
557 if (unlikely(!drl)) {
558 drl = srl;
559 /* Complete the source runlist if necessary. */
560 if (unlikely(drl[0].vcn)) {
561 /* Scan to the end of the source runlist. */
562 for (dend = 0; likely(drl[dend].length); dend++)
564 dend++;
565 drl = ntfs_rl_realloc(drl, dend, dend + 1);
566 if (IS_ERR(drl))
567 return drl;
568 /* Insert start element at the front of the runlist. */
569 ntfs_rl_mm(drl, 1, 0, dend);
570 drl[0].vcn = 0;
571 drl[0].lcn = LCN_RL_NOT_MAPPED;
572 drl[0].length = drl[1].vcn;
574 goto finished;
577 si = di = 0;
579 /* Skip any unmapped start element(s) in the source runlist. */
580 while (srl[si].length && srl[si].lcn < LCN_HOLE)
581 si++;
583 /* Can't have an entirely unmapped source runlist. */
584 BUG_ON(!srl[si].length);
586 /* Record the starting points. */
587 sstart = si;
590 * Skip forward in @drl until we reach the position where @srl needs to
591 * be inserted. If we reach the end of @drl, @srl just needs to be
592 * appended to @drl.
594 for (; drl[di].length; di++) {
595 if (drl[di].vcn + drl[di].length > srl[sstart].vcn)
596 break;
598 dins = di;
600 /* Sanity check for illegal overlaps. */
601 if ((drl[di].vcn == srl[si].vcn) && (drl[di].lcn >= 0) &&
602 (srl[si].lcn >= 0)) {
603 ntfs_error(NULL, "Run lists overlap. Cannot merge!");
604 return ERR_PTR(-ERANGE);
607 /* Scan to the end of both runlists in order to know their sizes. */
608 for (send = si; srl[send].length; send++)
610 for (dend = di; drl[dend].length; dend++)
613 if (srl[send].lcn == LCN_ENOENT)
614 marker_vcn = srl[marker = send].vcn;
616 /* Scan to the last element with lcn >= LCN_HOLE. */
617 for (sfinal = send; sfinal >= 0 && srl[sfinal].lcn < LCN_HOLE; sfinal--)
619 for (dfinal = dend; dfinal >= 0 && drl[dfinal].lcn < LCN_HOLE; dfinal--)
623 bool start;
624 bool finish;
625 int ds = dend + 1; /* Number of elements in drl & srl */
626 int ss = sfinal - sstart + 1;
628 start = ((drl[dins].lcn < LCN_RL_NOT_MAPPED) || /* End of file */
629 (drl[dins].vcn == srl[sstart].vcn)); /* Start of hole */
630 finish = ((drl[dins].lcn >= LCN_RL_NOT_MAPPED) && /* End of file */
631 ((drl[dins].vcn + drl[dins].length) <= /* End of hole */
632 (srl[send - 1].vcn + srl[send - 1].length)));
634 /* Or we will lose an end marker. */
635 if (finish && !drl[dins].length)
636 ss++;
637 if (marker && (drl[dins].vcn + drl[dins].length > srl[send - 1].vcn))
638 finish = false;
639 #if 0
640 ntfs_debug("dfinal = %i, dend = %i", dfinal, dend);
641 ntfs_debug("sstart = %i, sfinal = %i, send = %i", sstart, sfinal, send);
642 ntfs_debug("start = %i, finish = %i", start, finish);
643 ntfs_debug("ds = %i, ss = %i, dins = %i", ds, ss, dins);
644 #endif
645 if (start) {
646 if (finish)
647 drl = ntfs_rl_replace(drl, ds, srl + sstart, ss, dins);
648 else
649 drl = ntfs_rl_insert(drl, ds, srl + sstart, ss, dins);
650 } else {
651 if (finish)
652 drl = ntfs_rl_append(drl, ds, srl + sstart, ss, dins);
653 else
654 drl = ntfs_rl_split(drl, ds, srl + sstart, ss, dins);
656 if (IS_ERR(drl)) {
657 ntfs_error(NULL, "Merge failed.");
658 return drl;
660 ntfs_free(srl);
661 if (marker) {
662 ntfs_debug("Triggering marker code.");
663 for (ds = dend; drl[ds].length; ds++)
665 /* We only need to care if @srl ended after @drl. */
666 if (drl[ds].vcn <= marker_vcn) {
667 int slots = 0;
669 if (drl[ds].vcn == marker_vcn) {
670 ntfs_debug("Old marker = 0x%llx, replacing "
671 "with LCN_ENOENT.",
672 (unsigned long long)
673 drl[ds].lcn);
674 drl[ds].lcn = LCN_ENOENT;
675 goto finished;
678 * We need to create an unmapped runlist element in
679 * @drl or extend an existing one before adding the
680 * ENOENT terminator.
682 if (drl[ds].lcn == LCN_ENOENT) {
683 ds--;
684 slots = 1;
686 if (drl[ds].lcn != LCN_RL_NOT_MAPPED) {
687 /* Add an unmapped runlist element. */
688 if (!slots) {
689 drl = ntfs_rl_realloc_nofail(drl, ds,
690 ds + 2);
691 slots = 2;
693 ds++;
694 /* Need to set vcn if it isn't set already. */
695 if (slots != 1)
696 drl[ds].vcn = drl[ds - 1].vcn +
697 drl[ds - 1].length;
698 drl[ds].lcn = LCN_RL_NOT_MAPPED;
699 /* We now used up a slot. */
700 slots--;
702 drl[ds].length = marker_vcn - drl[ds].vcn;
703 /* Finally add the ENOENT terminator. */
704 ds++;
705 if (!slots)
706 drl = ntfs_rl_realloc_nofail(drl, ds, ds + 1);
707 drl[ds].vcn = marker_vcn;
708 drl[ds].lcn = LCN_ENOENT;
709 drl[ds].length = (s64)0;
714 finished:
715 /* The merge was completed successfully. */
716 ntfs_debug("Merged runlist:");
717 ntfs_debug_dump_runlist(drl);
718 return drl;
722 * ntfs_mapping_pairs_decompress - convert mapping pairs array to runlist
723 * @vol: ntfs volume on which the attribute resides
724 * @attr: attribute record whose mapping pairs array to decompress
725 * @old_rl: optional runlist in which to insert @attr's runlist
727 * It is up to the caller to serialize access to the runlist @old_rl.
729 * Decompress the attribute @attr's mapping pairs array into a runlist. On
730 * success, return the decompressed runlist.
732 * If @old_rl is not NULL, decompressed runlist is inserted into the
733 * appropriate place in @old_rl and the resultant, combined runlist is
734 * returned. The original @old_rl is deallocated.
736 * On error, return -errno. @old_rl is left unmodified in that case.
738 * The following error codes are defined:
739 * -ENOMEM - Not enough memory to allocate runlist array.
740 * -EIO - Corrupt runlist.
741 * -EINVAL - Invalid parameters were passed in.
742 * -ERANGE - The two runlists overlap.
744 * FIXME: For now we take the conceptionally simplest approach of creating the
745 * new runlist disregarding the already existing one and then splicing the
746 * two into one, if that is possible (we check for overlap and discard the new
747 * runlist if overlap present before returning ERR_PTR(-ERANGE)).
749 runlist_element *ntfs_mapping_pairs_decompress(const ntfs_volume *vol,
750 const ATTR_RECORD *attr, runlist_element *old_rl)
752 VCN vcn; /* Current vcn. */
753 LCN lcn; /* Current lcn. */
754 s64 deltaxcn; /* Change in [vl]cn. */
755 runlist_element *rl; /* The output runlist. */
756 u8 *buf; /* Current position in mapping pairs array. */
757 u8 *attr_end; /* End of attribute. */
758 int rlsize; /* Size of runlist buffer. */
759 u16 rlpos; /* Current runlist position in units of
760 runlist_elements. */
761 u8 b; /* Current byte offset in buf. */
763 #ifdef DEBUG
764 /* Make sure attr exists and is non-resident. */
765 if (!attr || !attr->non_resident || sle64_to_cpu(
766 attr->data.non_resident.lowest_vcn) < (VCN)0) {
767 ntfs_error(vol->sb, "Invalid arguments.");
768 return ERR_PTR(-EINVAL);
770 #endif
771 /* Start at vcn = lowest_vcn and lcn 0. */
772 vcn = sle64_to_cpu(attr->data.non_resident.lowest_vcn);
773 lcn = 0;
774 /* Get start of the mapping pairs array. */
775 buf = (u8*)attr + le16_to_cpu(
776 attr->data.non_resident.mapping_pairs_offset);
777 attr_end = (u8*)attr + le32_to_cpu(attr->length);
778 if (unlikely(buf < (u8*)attr || buf > attr_end)) {
779 ntfs_error(vol->sb, "Corrupt attribute.");
780 return ERR_PTR(-EIO);
782 /* If the mapping pairs array is valid but empty, nothing to do. */
783 if (!vcn && !*buf)
784 return old_rl;
785 /* Current position in runlist array. */
786 rlpos = 0;
787 /* Allocate first page and set current runlist size to one page. */
788 rl = ntfs_malloc_nofs(rlsize = PAGE_SIZE);
789 if (unlikely(!rl))
790 return ERR_PTR(-ENOMEM);
791 /* Insert unmapped starting element if necessary. */
792 if (vcn) {
793 rl->vcn = 0;
794 rl->lcn = LCN_RL_NOT_MAPPED;
795 rl->length = vcn;
796 rlpos++;
798 while (buf < attr_end && *buf) {
800 * Allocate more memory if needed, including space for the
801 * not-mapped and terminator elements. ntfs_malloc_nofs()
802 * operates on whole pages only.
804 if (((rlpos + 3) * sizeof(*old_rl)) > rlsize) {
805 runlist_element *rl2;
807 rl2 = ntfs_malloc_nofs(rlsize + (int)PAGE_SIZE);
808 if (unlikely(!rl2)) {
809 ntfs_free(rl);
810 return ERR_PTR(-ENOMEM);
812 memcpy(rl2, rl, rlsize);
813 ntfs_free(rl);
814 rl = rl2;
815 rlsize += PAGE_SIZE;
817 /* Enter the current vcn into the current runlist element. */
818 rl[rlpos].vcn = vcn;
820 * Get the change in vcn, i.e. the run length in clusters.
821 * Doing it this way ensures that we signextend negative values.
822 * A negative run length doesn't make any sense, but hey, I
823 * didn't make up the NTFS specs and Windows NT4 treats the run
824 * length as a signed value so that's how it is...
826 b = *buf & 0xf;
827 if (b) {
828 if (unlikely(buf + b > attr_end))
829 goto io_error;
830 for (deltaxcn = (s8)buf[b--]; b; b--)
831 deltaxcn = (deltaxcn << 8) + buf[b];
832 } else { /* The length entry is compulsory. */
833 ntfs_error(vol->sb, "Missing length entry in mapping "
834 "pairs array.");
835 deltaxcn = (s64)-1;
838 * Assume a negative length to indicate data corruption and
839 * hence clean-up and return NULL.
841 if (unlikely(deltaxcn < 0)) {
842 ntfs_error(vol->sb, "Invalid length in mapping pairs "
843 "array.");
844 goto err_out;
847 * Enter the current run length into the current runlist
848 * element.
850 rl[rlpos].length = deltaxcn;
851 /* Increment the current vcn by the current run length. */
852 vcn += deltaxcn;
854 * There might be no lcn change at all, as is the case for
855 * sparse clusters on NTFS 3.0+, in which case we set the lcn
856 * to LCN_HOLE.
858 if (!(*buf & 0xf0))
859 rl[rlpos].lcn = LCN_HOLE;
860 else {
861 /* Get the lcn change which really can be negative. */
862 u8 b2 = *buf & 0xf;
863 b = b2 + ((*buf >> 4) & 0xf);
864 if (buf + b > attr_end)
865 goto io_error;
866 for (deltaxcn = (s8)buf[b--]; b > b2; b--)
867 deltaxcn = (deltaxcn << 8) + buf[b];
868 /* Change the current lcn to its new value. */
869 lcn += deltaxcn;
870 #ifdef DEBUG
872 * On NTFS 1.2-, apparently can have lcn == -1 to
873 * indicate a hole. But we haven't verified ourselves
874 * whether it is really the lcn or the deltaxcn that is
875 * -1. So if either is found give us a message so we
876 * can investigate it further!
878 if (vol->major_ver < 3) {
879 if (unlikely(deltaxcn == (LCN)-1))
880 ntfs_error(vol->sb, "lcn delta == -1");
881 if (unlikely(lcn == (LCN)-1))
882 ntfs_error(vol->sb, "lcn == -1");
884 #endif
885 /* Check lcn is not below -1. */
886 if (unlikely(lcn < (LCN)-1)) {
887 ntfs_error(vol->sb, "Invalid LCN < -1 in "
888 "mapping pairs array.");
889 goto err_out;
891 /* Enter the current lcn into the runlist element. */
892 rl[rlpos].lcn = lcn;
894 /* Get to the next runlist element. */
895 rlpos++;
896 /* Increment the buffer position to the next mapping pair. */
897 buf += (*buf & 0xf) + ((*buf >> 4) & 0xf) + 1;
899 if (unlikely(buf >= attr_end))
900 goto io_error;
902 * If there is a highest_vcn specified, it must be equal to the final
903 * vcn in the runlist - 1, or something has gone badly wrong.
905 deltaxcn = sle64_to_cpu(attr->data.non_resident.highest_vcn);
906 if (unlikely(deltaxcn && vcn - 1 != deltaxcn)) {
907 mpa_err:
908 ntfs_error(vol->sb, "Corrupt mapping pairs array in "
909 "non-resident attribute.");
910 goto err_out;
912 /* Setup not mapped runlist element if this is the base extent. */
913 if (!attr->data.non_resident.lowest_vcn) {
914 VCN max_cluster;
916 max_cluster = ((sle64_to_cpu(
917 attr->data.non_resident.allocated_size) +
918 vol->cluster_size - 1) >>
919 vol->cluster_size_bits) - 1;
921 * A highest_vcn of zero means this is a single extent
922 * attribute so simply terminate the runlist with LCN_ENOENT).
924 if (deltaxcn) {
926 * If there is a difference between the highest_vcn and
927 * the highest cluster, the runlist is either corrupt
928 * or, more likely, there are more extents following
929 * this one.
931 if (deltaxcn < max_cluster) {
932 ntfs_debug("More extents to follow; deltaxcn "
933 "= 0x%llx, max_cluster = "
934 "0x%llx",
935 (unsigned long long)deltaxcn,
936 (unsigned long long)
937 max_cluster);
938 rl[rlpos].vcn = vcn;
939 vcn += rl[rlpos].length = max_cluster -
940 deltaxcn;
941 rl[rlpos].lcn = LCN_RL_NOT_MAPPED;
942 rlpos++;
943 } else if (unlikely(deltaxcn > max_cluster)) {
944 ntfs_error(vol->sb, "Corrupt attribute. "
945 "deltaxcn = 0x%llx, "
946 "max_cluster = 0x%llx",
947 (unsigned long long)deltaxcn,
948 (unsigned long long)
949 max_cluster);
950 goto mpa_err;
953 rl[rlpos].lcn = LCN_ENOENT;
954 } else /* Not the base extent. There may be more extents to follow. */
955 rl[rlpos].lcn = LCN_RL_NOT_MAPPED;
957 /* Setup terminating runlist element. */
958 rl[rlpos].vcn = vcn;
959 rl[rlpos].length = (s64)0;
960 /* If no existing runlist was specified, we are done. */
961 if (!old_rl) {
962 ntfs_debug("Mapping pairs array successfully decompressed:");
963 ntfs_debug_dump_runlist(rl);
964 return rl;
966 /* Now combine the new and old runlists checking for overlaps. */
967 old_rl = ntfs_runlists_merge(old_rl, rl);
968 if (likely(!IS_ERR(old_rl)))
969 return old_rl;
970 ntfs_free(rl);
971 ntfs_error(vol->sb, "Failed to merge runlists.");
972 return old_rl;
973 io_error:
974 ntfs_error(vol->sb, "Corrupt attribute.");
975 err_out:
976 ntfs_free(rl);
977 return ERR_PTR(-EIO);
981 * ntfs_rl_vcn_to_lcn - convert a vcn into a lcn given a runlist
982 * @rl: runlist to use for conversion
983 * @vcn: vcn to convert
985 * Convert the virtual cluster number @vcn of an attribute into a logical
986 * cluster number (lcn) of a device using the runlist @rl to map vcns to their
987 * corresponding lcns.
989 * It is up to the caller to serialize access to the runlist @rl.
991 * Since lcns must be >= 0, we use negative return codes with special meaning:
993 * Return code Meaning / Description
994 * ==================================================
995 * LCN_HOLE Hole / not allocated on disk.
996 * LCN_RL_NOT_MAPPED This is part of the runlist which has not been
997 * inserted into the runlist yet.
998 * LCN_ENOENT There is no such vcn in the attribute.
1000 * Locking: - The caller must have locked the runlist (for reading or writing).
1001 * - This function does not touch the lock, nor does it modify the
1002 * runlist.
1004 LCN ntfs_rl_vcn_to_lcn(const runlist_element *rl, const VCN vcn)
1006 int i;
1008 BUG_ON(vcn < 0);
1010 * If rl is NULL, assume that we have found an unmapped runlist. The
1011 * caller can then attempt to map it and fail appropriately if
1012 * necessary.
1014 if (unlikely(!rl))
1015 return LCN_RL_NOT_MAPPED;
1017 /* Catch out of lower bounds vcn. */
1018 if (unlikely(vcn < rl[0].vcn))
1019 return LCN_ENOENT;
1021 for (i = 0; likely(rl[i].length); i++) {
1022 if (unlikely(vcn < rl[i+1].vcn)) {
1023 if (likely(rl[i].lcn >= (LCN)0))
1024 return rl[i].lcn + (vcn - rl[i].vcn);
1025 return rl[i].lcn;
1029 * The terminator element is setup to the correct value, i.e. one of
1030 * LCN_HOLE, LCN_RL_NOT_MAPPED, or LCN_ENOENT.
1032 if (likely(rl[i].lcn < (LCN)0))
1033 return rl[i].lcn;
1034 /* Just in case... We could replace this with BUG() some day. */
1035 return LCN_ENOENT;
1038 #ifdef NTFS_RW
1041 * ntfs_rl_find_vcn_nolock - find a vcn in a runlist
1042 * @rl: runlist to search
1043 * @vcn: vcn to find
1045 * Find the virtual cluster number @vcn in the runlist @rl and return the
1046 * address of the runlist element containing the @vcn on success.
1048 * Return NULL if @rl is NULL or @vcn is in an unmapped part/out of bounds of
1049 * the runlist.
1051 * Locking: The runlist must be locked on entry.
1053 runlist_element *ntfs_rl_find_vcn_nolock(runlist_element *rl, const VCN vcn)
1055 BUG_ON(vcn < 0);
1056 if (unlikely(!rl || vcn < rl[0].vcn))
1057 return NULL;
1058 while (likely(rl->length)) {
1059 if (unlikely(vcn < rl[1].vcn)) {
1060 if (likely(rl->lcn >= LCN_HOLE))
1061 return rl;
1062 return NULL;
1064 rl++;
1066 if (likely(rl->lcn == LCN_ENOENT))
1067 return rl;
1068 return NULL;
1072 * ntfs_get_nr_significant_bytes - get number of bytes needed to store a number
1073 * @n: number for which to get the number of bytes for
1075 * Return the number of bytes required to store @n unambiguously as
1076 * a signed number.
1078 * This is used in the context of the mapping pairs array to determine how
1079 * many bytes will be needed in the array to store a given logical cluster
1080 * number (lcn) or a specific run length.
1082 * Return the number of bytes written. This function cannot fail.
1084 static inline int ntfs_get_nr_significant_bytes(const s64 n)
1086 s64 l = n;
1087 int i;
1088 s8 j;
1090 i = 0;
1091 do {
1092 l >>= 8;
1093 i++;
1094 } while (l != 0 && l != -1);
1095 j = (n >> 8 * (i - 1)) & 0xff;
1096 /* If the sign bit is wrong, we need an extra byte. */
1097 if ((n < 0 && j >= 0) || (n > 0 && j < 0))
1098 i++;
1099 return i;
1103 * ntfs_get_size_for_mapping_pairs - get bytes needed for mapping pairs array
1104 * @vol: ntfs volume (needed for the ntfs version)
1105 * @rl: locked runlist to determine the size of the mapping pairs of
1106 * @first_vcn: first vcn which to include in the mapping pairs array
1107 * @last_vcn: last vcn which to include in the mapping pairs array
1109 * Walk the locked runlist @rl and calculate the size in bytes of the mapping
1110 * pairs array corresponding to the runlist @rl, starting at vcn @first_vcn and
1111 * finishing with vcn @last_vcn.
1113 * A @last_vcn of -1 means end of runlist and in that case the size of the
1114 * mapping pairs array corresponding to the runlist starting at vcn @first_vcn
1115 * and finishing at the end of the runlist is determined.
1117 * This for example allows us to allocate a buffer of the right size when
1118 * building the mapping pairs array.
1120 * If @rl is NULL, just return 1 (for the single terminator byte).
1122 * Return the calculated size in bytes on success. On error, return -errno.
1123 * The following error codes are defined:
1124 * -EINVAL - Run list contains unmapped elements. Make sure to only pass
1125 * fully mapped runlists to this function.
1126 * -EIO - The runlist is corrupt.
1128 * Locking: @rl must be locked on entry (either for reading or writing), it
1129 * remains locked throughout, and is left locked upon return.
1131 int ntfs_get_size_for_mapping_pairs(const ntfs_volume *vol,
1132 const runlist_element *rl, const VCN first_vcn,
1133 const VCN last_vcn)
1135 LCN prev_lcn;
1136 int rls;
1137 bool the_end = false;
1139 BUG_ON(first_vcn < 0);
1140 BUG_ON(last_vcn < -1);
1141 BUG_ON(last_vcn >= 0 && first_vcn > last_vcn);
1142 if (!rl) {
1143 BUG_ON(first_vcn);
1144 BUG_ON(last_vcn > 0);
1145 return 1;
1147 /* Skip to runlist element containing @first_vcn. */
1148 while (rl->length && first_vcn >= rl[1].vcn)
1149 rl++;
1150 if (unlikely((!rl->length && first_vcn > rl->vcn) ||
1151 first_vcn < rl->vcn))
1152 return -EINVAL;
1153 prev_lcn = 0;
1154 /* Always need the termining zero byte. */
1155 rls = 1;
1156 /* Do the first partial run if present. */
1157 if (first_vcn > rl->vcn) {
1158 s64 delta, length = rl->length;
1160 /* We know rl->length != 0 already. */
1161 if (unlikely(length < 0 || rl->lcn < LCN_HOLE))
1162 goto err_out;
1164 * If @stop_vcn is given and finishes inside this run, cap the
1165 * run length.
1167 if (unlikely(last_vcn >= 0 && rl[1].vcn > last_vcn)) {
1168 s64 s1 = last_vcn + 1;
1169 if (unlikely(rl[1].vcn > s1))
1170 length = s1 - rl->vcn;
1171 the_end = true;
1173 delta = first_vcn - rl->vcn;
1174 /* Header byte + length. */
1175 rls += 1 + ntfs_get_nr_significant_bytes(length - delta);
1177 * If the logical cluster number (lcn) denotes a hole and we
1178 * are on NTFS 3.0+, we don't store it at all, i.e. we need
1179 * zero space. On earlier NTFS versions we just store the lcn.
1180 * Note: this assumes that on NTFS 1.2-, holes are stored with
1181 * an lcn of -1 and not a delta_lcn of -1 (unless both are -1).
1183 if (likely(rl->lcn >= 0 || vol->major_ver < 3)) {
1184 prev_lcn = rl->lcn;
1185 if (likely(rl->lcn >= 0))
1186 prev_lcn += delta;
1187 /* Change in lcn. */
1188 rls += ntfs_get_nr_significant_bytes(prev_lcn);
1190 /* Go to next runlist element. */
1191 rl++;
1193 /* Do the full runs. */
1194 for (; rl->length && !the_end; rl++) {
1195 s64 length = rl->length;
1197 if (unlikely(length < 0 || rl->lcn < LCN_HOLE))
1198 goto err_out;
1200 * If @stop_vcn is given and finishes inside this run, cap the
1201 * run length.
1203 if (unlikely(last_vcn >= 0 && rl[1].vcn > last_vcn)) {
1204 s64 s1 = last_vcn + 1;
1205 if (unlikely(rl[1].vcn > s1))
1206 length = s1 - rl->vcn;
1207 the_end = true;
1209 /* Header byte + length. */
1210 rls += 1 + ntfs_get_nr_significant_bytes(length);
1212 * If the logical cluster number (lcn) denotes a hole and we
1213 * are on NTFS 3.0+, we don't store it at all, i.e. we need
1214 * zero space. On earlier NTFS versions we just store the lcn.
1215 * Note: this assumes that on NTFS 1.2-, holes are stored with
1216 * an lcn of -1 and not a delta_lcn of -1 (unless both are -1).
1218 if (likely(rl->lcn >= 0 || vol->major_ver < 3)) {
1219 /* Change in lcn. */
1220 rls += ntfs_get_nr_significant_bytes(rl->lcn -
1221 prev_lcn);
1222 prev_lcn = rl->lcn;
1225 return rls;
1226 err_out:
1227 if (rl->lcn == LCN_RL_NOT_MAPPED)
1228 rls = -EINVAL;
1229 else
1230 rls = -EIO;
1231 return rls;
1235 * ntfs_write_significant_bytes - write the significant bytes of a number
1236 * @dst: destination buffer to write to
1237 * @dst_max: pointer to last byte of destination buffer for bounds checking
1238 * @n: number whose significant bytes to write
1240 * Store in @dst, the minimum bytes of the number @n which are required to
1241 * identify @n unambiguously as a signed number, taking care not to exceed
1242 * @dest_max, the maximum position within @dst to which we are allowed to
1243 * write.
1245 * This is used when building the mapping pairs array of a runlist to compress
1246 * a given logical cluster number (lcn) or a specific run length to the minumum
1247 * size possible.
1249 * Return the number of bytes written on success. On error, i.e. the
1250 * destination buffer @dst is too small, return -ENOSPC.
1252 static inline int ntfs_write_significant_bytes(s8 *dst, const s8 *dst_max,
1253 const s64 n)
1255 s64 l = n;
1256 int i;
1257 s8 j;
1259 i = 0;
1260 do {
1261 if (unlikely(dst > dst_max))
1262 goto err_out;
1263 *dst++ = l & 0xffll;
1264 l >>= 8;
1265 i++;
1266 } while (l != 0 && l != -1);
1267 j = (n >> 8 * (i - 1)) & 0xff;
1268 /* If the sign bit is wrong, we need an extra byte. */
1269 if (n < 0 && j >= 0) {
1270 if (unlikely(dst > dst_max))
1271 goto err_out;
1272 i++;
1273 *dst = (s8)-1;
1274 } else if (n > 0 && j < 0) {
1275 if (unlikely(dst > dst_max))
1276 goto err_out;
1277 i++;
1278 *dst = (s8)0;
1280 return i;
1281 err_out:
1282 return -ENOSPC;
1286 * ntfs_mapping_pairs_build - build the mapping pairs array from a runlist
1287 * @vol: ntfs volume (needed for the ntfs version)
1288 * @dst: destination buffer to which to write the mapping pairs array
1289 * @dst_len: size of destination buffer @dst in bytes
1290 * @rl: locked runlist for which to build the mapping pairs array
1291 * @first_vcn: first vcn which to include in the mapping pairs array
1292 * @last_vcn: last vcn which to include in the mapping pairs array
1293 * @stop_vcn: first vcn outside destination buffer on success or -ENOSPC
1295 * Create the mapping pairs array from the locked runlist @rl, starting at vcn
1296 * @first_vcn and finishing with vcn @last_vcn and save the array in @dst.
1297 * @dst_len is the size of @dst in bytes and it should be at least equal to the
1298 * value obtained by calling ntfs_get_size_for_mapping_pairs().
1300 * A @last_vcn of -1 means end of runlist and in that case the mapping pairs
1301 * array corresponding to the runlist starting at vcn @first_vcn and finishing
1302 * at the end of the runlist is created.
1304 * If @rl is NULL, just write a single terminator byte to @dst.
1306 * On success or -ENOSPC error, if @stop_vcn is not NULL, *@stop_vcn is set to
1307 * the first vcn outside the destination buffer. Note that on error, @dst has
1308 * been filled with all the mapping pairs that will fit, thus it can be treated
1309 * as partial success, in that a new attribute extent needs to be created or
1310 * the next extent has to be used and the mapping pairs build has to be
1311 * continued with @first_vcn set to *@stop_vcn.
1313 * Return 0 on success and -errno on error. The following error codes are
1314 * defined:
1315 * -EINVAL - Run list contains unmapped elements. Make sure to only pass
1316 * fully mapped runlists to this function.
1317 * -EIO - The runlist is corrupt.
1318 * -ENOSPC - The destination buffer is too small.
1320 * Locking: @rl must be locked on entry (either for reading or writing), it
1321 * remains locked throughout, and is left locked upon return.
1323 int ntfs_mapping_pairs_build(const ntfs_volume *vol, s8 *dst,
1324 const int dst_len, const runlist_element *rl,
1325 const VCN first_vcn, const VCN last_vcn, VCN *const stop_vcn)
1327 LCN prev_lcn;
1328 s8 *dst_max, *dst_next;
1329 int err = -ENOSPC;
1330 bool the_end = false;
1331 s8 len_len, lcn_len;
1333 BUG_ON(first_vcn < 0);
1334 BUG_ON(last_vcn < -1);
1335 BUG_ON(last_vcn >= 0 && first_vcn > last_vcn);
1336 BUG_ON(dst_len < 1);
1337 if (!rl) {
1338 BUG_ON(first_vcn);
1339 BUG_ON(last_vcn > 0);
1340 if (stop_vcn)
1341 *stop_vcn = 0;
1342 /* Terminator byte. */
1343 *dst = 0;
1344 return 0;
1346 /* Skip to runlist element containing @first_vcn. */
1347 while (rl->length && first_vcn >= rl[1].vcn)
1348 rl++;
1349 if (unlikely((!rl->length && first_vcn > rl->vcn) ||
1350 first_vcn < rl->vcn))
1351 return -EINVAL;
1353 * @dst_max is used for bounds checking in
1354 * ntfs_write_significant_bytes().
1356 dst_max = dst + dst_len - 1;
1357 prev_lcn = 0;
1358 /* Do the first partial run if present. */
1359 if (first_vcn > rl->vcn) {
1360 s64 delta, length = rl->length;
1362 /* We know rl->length != 0 already. */
1363 if (unlikely(length < 0 || rl->lcn < LCN_HOLE))
1364 goto err_out;
1366 * If @stop_vcn is given and finishes inside this run, cap the
1367 * run length.
1369 if (unlikely(last_vcn >= 0 && rl[1].vcn > last_vcn)) {
1370 s64 s1 = last_vcn + 1;
1371 if (unlikely(rl[1].vcn > s1))
1372 length = s1 - rl->vcn;
1373 the_end = true;
1375 delta = first_vcn - rl->vcn;
1376 /* Write length. */
1377 len_len = ntfs_write_significant_bytes(dst + 1, dst_max,
1378 length - delta);
1379 if (unlikely(len_len < 0))
1380 goto size_err;
1382 * If the logical cluster number (lcn) denotes a hole and we
1383 * are on NTFS 3.0+, we don't store it at all, i.e. we need
1384 * zero space. On earlier NTFS versions we just write the lcn
1385 * change. FIXME: Do we need to write the lcn change or just
1386 * the lcn in that case? Not sure as I have never seen this
1387 * case on NT4. - We assume that we just need to write the lcn
1388 * change until someone tells us otherwise... (AIA)
1390 if (likely(rl->lcn >= 0 || vol->major_ver < 3)) {
1391 prev_lcn = rl->lcn;
1392 if (likely(rl->lcn >= 0))
1393 prev_lcn += delta;
1394 /* Write change in lcn. */
1395 lcn_len = ntfs_write_significant_bytes(dst + 1 +
1396 len_len, dst_max, prev_lcn);
1397 if (unlikely(lcn_len < 0))
1398 goto size_err;
1399 } else
1400 lcn_len = 0;
1401 dst_next = dst + len_len + lcn_len + 1;
1402 if (unlikely(dst_next > dst_max))
1403 goto size_err;
1404 /* Update header byte. */
1405 *dst = lcn_len << 4 | len_len;
1406 /* Position at next mapping pairs array element. */
1407 dst = dst_next;
1408 /* Go to next runlist element. */
1409 rl++;
1411 /* Do the full runs. */
1412 for (; rl->length && !the_end; rl++) {
1413 s64 length = rl->length;
1415 if (unlikely(length < 0 || rl->lcn < LCN_HOLE))
1416 goto err_out;
1418 * If @stop_vcn is given and finishes inside this run, cap the
1419 * run length.
1421 if (unlikely(last_vcn >= 0 && rl[1].vcn > last_vcn)) {
1422 s64 s1 = last_vcn + 1;
1423 if (unlikely(rl[1].vcn > s1))
1424 length = s1 - rl->vcn;
1425 the_end = true;
1427 /* Write length. */
1428 len_len = ntfs_write_significant_bytes(dst + 1, dst_max,
1429 length);
1430 if (unlikely(len_len < 0))
1431 goto size_err;
1433 * If the logical cluster number (lcn) denotes a hole and we
1434 * are on NTFS 3.0+, we don't store it at all, i.e. we need
1435 * zero space. On earlier NTFS versions we just write the lcn
1436 * change. FIXME: Do we need to write the lcn change or just
1437 * the lcn in that case? Not sure as I have never seen this
1438 * case on NT4. - We assume that we just need to write the lcn
1439 * change until someone tells us otherwise... (AIA)
1441 if (likely(rl->lcn >= 0 || vol->major_ver < 3)) {
1442 /* Write change in lcn. */
1443 lcn_len = ntfs_write_significant_bytes(dst + 1 +
1444 len_len, dst_max, rl->lcn - prev_lcn);
1445 if (unlikely(lcn_len < 0))
1446 goto size_err;
1447 prev_lcn = rl->lcn;
1448 } else
1449 lcn_len = 0;
1450 dst_next = dst + len_len + lcn_len + 1;
1451 if (unlikely(dst_next > dst_max))
1452 goto size_err;
1453 /* Update header byte. */
1454 *dst = lcn_len << 4 | len_len;
1455 /* Position at next mapping pairs array element. */
1456 dst = dst_next;
1458 /* Success. */
1459 err = 0;
1460 size_err:
1461 /* Set stop vcn. */
1462 if (stop_vcn)
1463 *stop_vcn = rl->vcn;
1464 /* Add terminator byte. */
1465 *dst = 0;
1466 return err;
1467 err_out:
1468 if (rl->lcn == LCN_RL_NOT_MAPPED)
1469 err = -EINVAL;
1470 else
1471 err = -EIO;
1472 return err;
1476 * ntfs_rl_truncate_nolock - truncate a runlist starting at a specified vcn
1477 * @vol: ntfs volume (needed for error output)
1478 * @runlist: runlist to truncate
1479 * @new_length: the new length of the runlist in VCNs
1481 * Truncate the runlist described by @runlist as well as the memory buffer
1482 * holding the runlist elements to a length of @new_length VCNs.
1484 * If @new_length lies within the runlist, the runlist elements with VCNs of
1485 * @new_length and above are discarded. As a special case if @new_length is
1486 * zero, the runlist is discarded and set to NULL.
1488 * If @new_length lies beyond the runlist, a sparse runlist element is added to
1489 * the end of the runlist @runlist or if the last runlist element is a sparse
1490 * one already, this is extended.
1492 * Note, no checking is done for unmapped runlist elements. It is assumed that
1493 * the caller has mapped any elements that need to be mapped already.
1495 * Return 0 on success and -errno on error.
1497 * Locking: The caller must hold @runlist->lock for writing.
1499 int ntfs_rl_truncate_nolock(const ntfs_volume *vol, runlist *const runlist,
1500 const s64 new_length)
1502 runlist_element *rl;
1503 int old_size;
1505 ntfs_debug("Entering for new_length 0x%llx.", (long long)new_length);
1506 BUG_ON(!runlist);
1507 BUG_ON(new_length < 0);
1508 rl = runlist->rl;
1509 if (!new_length) {
1510 ntfs_debug("Freeing runlist.");
1511 runlist->rl = NULL;
1512 if (rl)
1513 ntfs_free(rl);
1514 return 0;
1516 if (unlikely(!rl)) {
1518 * Create a runlist consisting of a sparse runlist element of
1519 * length @new_length followed by a terminator runlist element.
1521 rl = ntfs_malloc_nofs(PAGE_SIZE);
1522 if (unlikely(!rl)) {
1523 ntfs_error(vol->sb, "Not enough memory to allocate "
1524 "runlist element buffer.");
1525 return -ENOMEM;
1527 runlist->rl = rl;
1528 rl[1].length = rl->vcn = 0;
1529 rl->lcn = LCN_HOLE;
1530 rl[1].vcn = rl->length = new_length;
1531 rl[1].lcn = LCN_ENOENT;
1532 return 0;
1534 BUG_ON(new_length < rl->vcn);
1535 /* Find @new_length in the runlist. */
1536 while (likely(rl->length && new_length >= rl[1].vcn))
1537 rl++;
1539 * If not at the end of the runlist we need to shrink it.
1540 * If at the end of the runlist we need to expand it.
1542 if (rl->length) {
1543 runlist_element *trl;
1544 bool is_end;
1546 ntfs_debug("Shrinking runlist.");
1547 /* Determine the runlist size. */
1548 trl = rl + 1;
1549 while (likely(trl->length))
1550 trl++;
1551 old_size = trl - runlist->rl + 1;
1552 /* Truncate the run. */
1553 rl->length = new_length - rl->vcn;
1555 * If a run was partially truncated, make the following runlist
1556 * element a terminator.
1558 is_end = false;
1559 if (rl->length) {
1560 rl++;
1561 if (!rl->length)
1562 is_end = true;
1563 rl->vcn = new_length;
1564 rl->length = 0;
1566 rl->lcn = LCN_ENOENT;
1567 /* Reallocate memory if necessary. */
1568 if (!is_end) {
1569 int new_size = rl - runlist->rl + 1;
1570 rl = ntfs_rl_realloc(runlist->rl, old_size, new_size);
1571 if (IS_ERR(rl))
1572 ntfs_warning(vol->sb, "Failed to shrink "
1573 "runlist buffer. This just "
1574 "wastes a bit of memory "
1575 "temporarily so we ignore it "
1576 "and return success.");
1577 else
1578 runlist->rl = rl;
1580 } else if (likely(/* !rl->length && */ new_length > rl->vcn)) {
1581 ntfs_debug("Expanding runlist.");
1583 * If there is a previous runlist element and it is a sparse
1584 * one, extend it. Otherwise need to add a new, sparse runlist
1585 * element.
1587 if ((rl > runlist->rl) && ((rl - 1)->lcn == LCN_HOLE))
1588 (rl - 1)->length = new_length - (rl - 1)->vcn;
1589 else {
1590 /* Determine the runlist size. */
1591 old_size = rl - runlist->rl + 1;
1592 /* Reallocate memory if necessary. */
1593 rl = ntfs_rl_realloc(runlist->rl, old_size,
1594 old_size + 1);
1595 if (IS_ERR(rl)) {
1596 ntfs_error(vol->sb, "Failed to expand runlist "
1597 "buffer, aborting.");
1598 return PTR_ERR(rl);
1600 runlist->rl = rl;
1602 * Set @rl to the same runlist element in the new
1603 * runlist as before in the old runlist.
1605 rl += old_size - 1;
1606 /* Add a new, sparse runlist element. */
1607 rl->lcn = LCN_HOLE;
1608 rl->length = new_length - rl->vcn;
1609 /* Add a new terminator runlist element. */
1610 rl++;
1611 rl->length = 0;
1613 rl->vcn = new_length;
1614 rl->lcn = LCN_ENOENT;
1615 } else /* if (unlikely(!rl->length && new_length == rl->vcn)) */ {
1616 /* Runlist already has same size as requested. */
1617 rl->lcn = LCN_ENOENT;
1619 ntfs_debug("Done.");
1620 return 0;
1624 * ntfs_rl_punch_nolock - punch a hole into a runlist
1625 * @vol: ntfs volume (needed for error output)
1626 * @runlist: runlist to punch a hole into
1627 * @start: starting VCN of the hole to be created
1628 * @length: size of the hole to be created in units of clusters
1630 * Punch a hole into the runlist @runlist starting at VCN @start and of size
1631 * @length clusters.
1633 * Return 0 on success and -errno on error, in which case @runlist has not been
1634 * modified.
1636 * If @start and/or @start + @length are outside the runlist return error code
1637 * -ENOENT.
1639 * If the runlist contains unmapped or error elements between @start and @start
1640 * + @length return error code -EINVAL.
1642 * Locking: The caller must hold @runlist->lock for writing.
1644 int ntfs_rl_punch_nolock(const ntfs_volume *vol, runlist *const runlist,
1645 const VCN start, const s64 length)
1647 const VCN end = start + length;
1648 s64 delta;
1649 runlist_element *rl, *rl_end, *rl_real_end, *trl;
1650 int old_size;
1651 bool lcn_fixup = false;
1653 ntfs_debug("Entering for start 0x%llx, length 0x%llx.",
1654 (long long)start, (long long)length);
1655 BUG_ON(!runlist);
1656 BUG_ON(start < 0);
1657 BUG_ON(length < 0);
1658 BUG_ON(end < 0);
1659 rl = runlist->rl;
1660 if (unlikely(!rl)) {
1661 if (likely(!start && !length))
1662 return 0;
1663 return -ENOENT;
1665 /* Find @start in the runlist. */
1666 while (likely(rl->length && start >= rl[1].vcn))
1667 rl++;
1668 rl_end = rl;
1669 /* Find @end in the runlist. */
1670 while (likely(rl_end->length && end >= rl_end[1].vcn)) {
1671 /* Verify there are no unmapped or error elements. */
1672 if (unlikely(rl_end->lcn < LCN_HOLE))
1673 return -EINVAL;
1674 rl_end++;
1676 /* Check the last element. */
1677 if (unlikely(rl_end->length && rl_end->lcn < LCN_HOLE))
1678 return -EINVAL;
1679 /* This covers @start being out of bounds, too. */
1680 if (!rl_end->length && end > rl_end->vcn)
1681 return -ENOENT;
1682 if (!length)
1683 return 0;
1684 if (!rl->length)
1685 return -ENOENT;
1686 rl_real_end = rl_end;
1687 /* Determine the runlist size. */
1688 while (likely(rl_real_end->length))
1689 rl_real_end++;
1690 old_size = rl_real_end - runlist->rl + 1;
1691 /* If @start is in a hole simply extend the hole. */
1692 if (rl->lcn == LCN_HOLE) {
1694 * If both @start and @end are in the same sparse run, we are
1695 * done.
1697 if (end <= rl[1].vcn) {
1698 ntfs_debug("Done (requested hole is already sparse).");
1699 return 0;
1701 extend_hole:
1702 /* Extend the hole. */
1703 rl->length = end - rl->vcn;
1704 /* If @end is in a hole, merge it with the current one. */
1705 if (rl_end->lcn == LCN_HOLE) {
1706 rl_end++;
1707 rl->length = rl_end->vcn - rl->vcn;
1709 /* We have done the hole. Now deal with the remaining tail. */
1710 rl++;
1711 /* Cut out all runlist elements up to @end. */
1712 if (rl < rl_end)
1713 memmove(rl, rl_end, (rl_real_end - rl_end + 1) *
1714 sizeof(*rl));
1715 /* Adjust the beginning of the tail if necessary. */
1716 if (end > rl->vcn) {
1717 delta = end - rl->vcn;
1718 rl->vcn = end;
1719 rl->length -= delta;
1720 /* Only adjust the lcn if it is real. */
1721 if (rl->lcn >= 0)
1722 rl->lcn += delta;
1724 shrink_allocation:
1725 /* Reallocate memory if the allocation changed. */
1726 if (rl < rl_end) {
1727 rl = ntfs_rl_realloc(runlist->rl, old_size,
1728 old_size - (rl_end - rl));
1729 if (IS_ERR(rl))
1730 ntfs_warning(vol->sb, "Failed to shrink "
1731 "runlist buffer. This just "
1732 "wastes a bit of memory "
1733 "temporarily so we ignore it "
1734 "and return success.");
1735 else
1736 runlist->rl = rl;
1738 ntfs_debug("Done (extend hole).");
1739 return 0;
1742 * If @start is at the beginning of a run things are easier as there is
1743 * no need to split the first run.
1745 if (start == rl->vcn) {
1747 * @start is at the beginning of a run.
1749 * If the previous run is sparse, extend its hole.
1751 * If @end is not in the same run, switch the run to be sparse
1752 * and extend the newly created hole.
1754 * Thus both of these cases reduce the problem to the above
1755 * case of "@start is in a hole".
1757 if (rl > runlist->rl && (rl - 1)->lcn == LCN_HOLE) {
1758 rl--;
1759 goto extend_hole;
1761 if (end >= rl[1].vcn) {
1762 rl->lcn = LCN_HOLE;
1763 goto extend_hole;
1766 * The final case is when @end is in the same run as @start.
1767 * For this need to split the run into two. One run for the
1768 * sparse region between the beginning of the old run, i.e.
1769 * @start, and @end and one for the remaining non-sparse
1770 * region, i.e. between @end and the end of the old run.
1772 trl = ntfs_rl_realloc(runlist->rl, old_size, old_size + 1);
1773 if (IS_ERR(trl))
1774 goto enomem_out;
1775 old_size++;
1776 if (runlist->rl != trl) {
1777 rl = trl + (rl - runlist->rl);
1778 rl_end = trl + (rl_end - runlist->rl);
1779 rl_real_end = trl + (rl_real_end - runlist->rl);
1780 runlist->rl = trl;
1782 split_end:
1783 /* Shift all the runs up by one. */
1784 memmove(rl + 1, rl, (rl_real_end - rl + 1) * sizeof(*rl));
1785 /* Finally, setup the two split runs. */
1786 rl->lcn = LCN_HOLE;
1787 rl->length = length;
1788 rl++;
1789 rl->vcn += length;
1790 /* Only adjust the lcn if it is real. */
1791 if (rl->lcn >= 0 || lcn_fixup)
1792 rl->lcn += length;
1793 rl->length -= length;
1794 ntfs_debug("Done (split one).");
1795 return 0;
1798 * @start is neither in a hole nor at the beginning of a run.
1800 * If @end is in a hole, things are easier as simply truncating the run
1801 * @start is in to end at @start - 1, deleting all runs after that up
1802 * to @end, and finally extending the beginning of the run @end is in
1803 * to be @start is all that is needed.
1805 if (rl_end->lcn == LCN_HOLE) {
1806 /* Truncate the run containing @start. */
1807 rl->length = start - rl->vcn;
1808 rl++;
1809 /* Cut out all runlist elements up to @end. */
1810 if (rl < rl_end)
1811 memmove(rl, rl_end, (rl_real_end - rl_end + 1) *
1812 sizeof(*rl));
1813 /* Extend the beginning of the run @end is in to be @start. */
1814 rl->vcn = start;
1815 rl->length = rl[1].vcn - start;
1816 goto shrink_allocation;
1819 * If @end is not in a hole there are still two cases to distinguish.
1820 * Either @end is or is not in the same run as @start.
1822 * The second case is easier as it can be reduced to an already solved
1823 * problem by truncating the run @start is in to end at @start - 1.
1824 * Then, if @end is in the next run need to split the run into a sparse
1825 * run followed by a non-sparse run (already covered above) and if @end
1826 * is not in the next run switching it to be sparse, again reduces the
1827 * problem to the already covered case of "@start is in a hole".
1829 if (end >= rl[1].vcn) {
1831 * If @end is not in the next run, reduce the problem to the
1832 * case of "@start is in a hole".
1834 if (rl[1].length && end >= rl[2].vcn) {
1835 /* Truncate the run containing @start. */
1836 rl->length = start - rl->vcn;
1837 rl++;
1838 rl->vcn = start;
1839 rl->lcn = LCN_HOLE;
1840 goto extend_hole;
1842 trl = ntfs_rl_realloc(runlist->rl, old_size, old_size + 1);
1843 if (IS_ERR(trl))
1844 goto enomem_out;
1845 old_size++;
1846 if (runlist->rl != trl) {
1847 rl = trl + (rl - runlist->rl);
1848 rl_end = trl + (rl_end - runlist->rl);
1849 rl_real_end = trl + (rl_real_end - runlist->rl);
1850 runlist->rl = trl;
1852 /* Truncate the run containing @start. */
1853 rl->length = start - rl->vcn;
1854 rl++;
1856 * @end is in the next run, reduce the problem to the case
1857 * where "@start is at the beginning of a run and @end is in
1858 * the same run as @start".
1860 delta = rl->vcn - start;
1861 rl->vcn = start;
1862 if (rl->lcn >= 0) {
1863 rl->lcn -= delta;
1864 /* Need this in case the lcn just became negative. */
1865 lcn_fixup = true;
1867 rl->length += delta;
1868 goto split_end;
1871 * The first case from above, i.e. @end is in the same run as @start.
1872 * We need to split the run into three. One run for the non-sparse
1873 * region between the beginning of the old run and @start, one for the
1874 * sparse region between @start and @end, and one for the remaining
1875 * non-sparse region, i.e. between @end and the end of the old run.
1877 trl = ntfs_rl_realloc(runlist->rl, old_size, old_size + 2);
1878 if (IS_ERR(trl))
1879 goto enomem_out;
1880 old_size += 2;
1881 if (runlist->rl != trl) {
1882 rl = trl + (rl - runlist->rl);
1883 rl_end = trl + (rl_end - runlist->rl);
1884 rl_real_end = trl + (rl_real_end - runlist->rl);
1885 runlist->rl = trl;
1887 /* Shift all the runs up by two. */
1888 memmove(rl + 2, rl, (rl_real_end - rl + 1) * sizeof(*rl));
1889 /* Finally, setup the three split runs. */
1890 rl->length = start - rl->vcn;
1891 rl++;
1892 rl->vcn = start;
1893 rl->lcn = LCN_HOLE;
1894 rl->length = length;
1895 rl++;
1896 delta = end - rl->vcn;
1897 rl->vcn = end;
1898 rl->lcn += delta;
1899 rl->length -= delta;
1900 ntfs_debug("Done (split both).");
1901 return 0;
1902 enomem_out:
1903 ntfs_error(vol->sb, "Not enough memory to extend runlist buffer.");
1904 return -ENOMEM;
1907 #endif /* NTFS_RW */