read-cache.c 61.6 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1 2 3 4 5
/*
 * GIT - The information manager from hell
 *
 * Copyright (C) Linus Torvalds, 2005
 */
6
#define NO_THE_INDEX_COMPATIBILITY_MACROS
7
#include "cache.h"
8
#include "lockfile.h"
9
#include "cache-tree.h"
10
#include "refs.h"
11
#include "dir.h"
12 13
#include "tree.h"
#include "commit.h"
14
#include "blob.h"
15
#include "resolve-undo.h"
16 17
#include "strbuf.h"
#include "varint.h"
Duy Nguyen's avatar
Duy Nguyen committed
18
#include "split-index.h"
19
#include "sigchain.h"
20
#include "utf8.h"
21

22 23
static struct cache_entry *refresh_cache_entry(struct cache_entry *ce,
					       unsigned int options);
24

25 26 27 28
/* Mask for the name length in ce_flags in the on-disk index */

#define CE_NAMEMASK  (0x0fff)

29 30 31 32 33 34 35 36 37 38 39
/* Index extensions.
 *
 * The first letter should be 'A'..'Z' for extensions that are not
 * necessary for a correct operation (i.e. optimization data).
 * When new extensions are added that _needs_ to be understood in
 * order to correctly interpret the index file, pick character that
 * is outside the range, to cause the reader to abort.
 */

#define CACHE_EXT(s) ( (s[0]<<24)|(s[1]<<16)|(s[2]<<8)|(s[3]) )
#define CACHE_EXT_TREE 0x54524545	/* "TREE" */
40
#define CACHE_EXT_RESOLVE_UNDO 0x52455543 /* "REUC" */
Duy Nguyen's avatar
Duy Nguyen committed
41
#define CACHE_EXT_LINK 0x6c696e6b	  /* "link" */
42
#define CACHE_EXT_UNTRACKED 0x554E5452	  /* "UNTR" */
Duy Nguyen's avatar
Duy Nguyen committed
43 44

/* changes that can be kept in $GIT_DIR/index (basically all extensions) */
45
#define EXTMASK (RESOLVE_UNDO_CHANGED | CACHE_TREE_CHANGED | \
46 47
		 CE_ENTRY_ADDED | CE_ENTRY_REMOVED | CE_ENTRY_CHANGED | \
		 SPLIT_INDEX_ORDERED)
48

49
struct index_state the_index;
50
static const char *alternate_index_output;
51

Junio C Hamano's avatar
Junio C Hamano committed
52 53 54
static void set_index_entry(struct index_state *istate, int nr, struct cache_entry *ce)
{
	istate->cache[nr] = ce;
55
	add_name_hash(istate, ce);
Junio C Hamano's avatar
Junio C Hamano committed
56 57
}

58 59 60 61
static void replace_index_entry(struct index_state *istate, int nr, struct cache_entry *ce)
{
	struct cache_entry *old = istate->cache[nr];

62
	replace_index_entry_in_base(istate, old, ce);
63
	remove_name_hash(istate, old);
64
	free(old);
65
	set_index_entry(istate, nr, ce);
66
	ce->ce_flags |= CE_UPDATE_IN_BASE;
67
	istate->cache_changed |= CE_ENTRY_CHANGED;
68 69
}

70 71 72 73 74 75 76
void rename_index_entry_at(struct index_state *istate, int nr, const char *new_name)
{
	struct cache_entry *old = istate->cache[nr], *new;
	int namelen = strlen(new_name);

	new = xmalloc(cache_entry_size(namelen));
	copy_cache_entry(new, old);
77
	new->ce_flags &= ~CE_HASHED;
78
	new->ce_namelen = namelen;
Duy Nguyen's avatar
Duy Nguyen committed
79
	new->index = 0;
80 81
	memcpy(new->name, new_name, namelen + 1);

82
	cache_tree_invalidate_path(istate, old->name);
83 84 85 86
	remove_index_entry_at(istate, nr);
	add_index_entry(istate, new, ADD_CACHE_OK_TO_ADD|ADD_CACHE_OK_TO_REPLACE);
}

87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141
void fill_stat_data(struct stat_data *sd, struct stat *st)
{
	sd->sd_ctime.sec = (unsigned int)st->st_ctime;
	sd->sd_mtime.sec = (unsigned int)st->st_mtime;
	sd->sd_ctime.nsec = ST_CTIME_NSEC(*st);
	sd->sd_mtime.nsec = ST_MTIME_NSEC(*st);
	sd->sd_dev = st->st_dev;
	sd->sd_ino = st->st_ino;
	sd->sd_uid = st->st_uid;
	sd->sd_gid = st->st_gid;
	sd->sd_size = st->st_size;
}

int match_stat_data(const struct stat_data *sd, struct stat *st)
{
	int changed = 0;

	if (sd->sd_mtime.sec != (unsigned int)st->st_mtime)
		changed |= MTIME_CHANGED;
	if (trust_ctime && check_stat &&
	    sd->sd_ctime.sec != (unsigned int)st->st_ctime)
		changed |= CTIME_CHANGED;

#ifdef USE_NSEC
	if (check_stat && sd->sd_mtime.nsec != ST_MTIME_NSEC(*st))
		changed |= MTIME_CHANGED;
	if (trust_ctime && check_stat &&
	    sd->sd_ctime.nsec != ST_CTIME_NSEC(*st))
		changed |= CTIME_CHANGED;
#endif

	if (check_stat) {
		if (sd->sd_uid != (unsigned int) st->st_uid ||
			sd->sd_gid != (unsigned int) st->st_gid)
			changed |= OWNER_CHANGED;
		if (sd->sd_ino != (unsigned int) st->st_ino)
			changed |= INODE_CHANGED;
	}

#ifdef USE_STDEV
	/*
	 * st_dev breaks on network filesystems where different
	 * clients will have different views of what "device"
	 * the filesystem is on
	 */
	if (check_stat && sd->sd_dev != (unsigned int) st->st_dev)
			changed |= INODE_CHANGED;
#endif

	if (sd->sd_size != (unsigned int) st->st_size)
		changed |= DATA_CHANGED;

	return changed;
}

142 143 144 145 146 147 148
/*
 * This only updates the "non-critical" parts of the directory
 * cache, ie the parts that aren't tracked by GIT, and only used
 * to validate the cache.
 */
void fill_stat_cache_info(struct cache_entry *ce, struct stat *st)
{
149
	fill_stat_data(&ce->ce_stat_data, st);
Junio C Hamano's avatar
Junio C Hamano committed
150 151

	if (assume_unchanged)
152
		ce->ce_flags |= CE_VALID;
153 154 155

	if (S_ISREG(st->st_mode))
		ce_mark_uptodate(ce);
156 157
}

158
static int ce_compare_data(const struct cache_entry *ce, struct stat *st)
Junio C Hamano's avatar
Junio C Hamano committed
159 160 161 162 163 164
{
	int match = -1;
	int fd = open(ce->name, O_RDONLY);

	if (fd >= 0) {
		unsigned char sha1[20];
165
		if (!index_fd(sha1, fd, st, OBJ_BLOB, ce->name, 0))
166
			match = hashcmp(sha1, ce->sha1);
167
		/* index_fd() closed the file descriptor already */
Junio C Hamano's avatar
Junio C Hamano committed
168 169 170 171
	}
	return match;
}

172
static int ce_compare_link(const struct cache_entry *ce, size_t expected_size)
Junio C Hamano's avatar
Junio C Hamano committed
173 174 175 176
{
	int match = -1;
	void *buffer;
	unsigned long size;
177
	enum object_type type;
178
	struct strbuf sb = STRBUF_INIT;
Junio C Hamano's avatar
Junio C Hamano committed
179

180
	if (strbuf_readlink(&sb, ce->name, expected_size))
Junio C Hamano's avatar
Junio C Hamano committed
181
		return -1;
182

183
	buffer = read_sha1_file(ce->sha1, &type, &size);
184 185 186 187
	if (buffer) {
		if (size == sb.len)
			match = memcmp(buffer, sb.buf, size);
		free(buffer);
Junio C Hamano's avatar
Junio C Hamano committed
188
	}
189
	strbuf_release(&sb);
Junio C Hamano's avatar
Junio C Hamano committed
190 191 192
	return match;
}

193
static int ce_compare_gitlink(const struct cache_entry *ce)
194 195 196 197 198
{
	unsigned char sha1[20];

	/*
	 * We don't actually require that the .git directory
Martin Waitz's avatar
Martin Waitz committed
199
	 * under GITLINK directory be a valid git directory. It
200 201 202 203 204 205 206 207 208 209
	 * might even be missing (in case nobody populated that
	 * sub-project).
	 *
	 * If so, we consider it always to match.
	 */
	if (resolve_gitlink_ref(ce->name, "HEAD", sha1) < 0)
		return 0;
	return hashcmp(sha1, ce->sha1);
}

210
static int ce_modified_check_fs(const struct cache_entry *ce, struct stat *st)
Junio C Hamano's avatar
Junio C Hamano committed
211 212 213 214 215 216 217
{
	switch (st->st_mode & S_IFMT) {
	case S_IFREG:
		if (ce_compare_data(ce, st))
			return DATA_CHANGED;
		break;
	case S_IFLNK:
218
		if (ce_compare_link(ce, xsize_t(st->st_size)))
Junio C Hamano's avatar
Junio C Hamano committed
219 220
			return DATA_CHANGED;
		break;
221
	case S_IFDIR:
222
		if (S_ISGITLINK(ce->ce_mode))
223
			return ce_compare_gitlink(ce) ? DATA_CHANGED : 0;
Junio C Hamano's avatar
Junio C Hamano committed
224 225 226 227 228 229
	default:
		return TYPE_CHANGED;
	}
	return 0;
}

230
static int ce_match_stat_basic(const struct cache_entry *ce, struct stat *st)
231 232 233
{
	unsigned int changed = 0;

234 235 236 237
	if (ce->ce_flags & CE_REMOVE)
		return MODE_CHANGED | DATA_CHANGED | TYPE_CHANGED;

	switch (ce->ce_mode & S_IFMT) {
238 239
	case S_IFREG:
		changed |= !S_ISREG(st->st_mode) ? TYPE_CHANGED : 0;
Junio C Hamano's avatar
Junio C Hamano committed
240 241 242 243
		/* We consider only the owner x bit to be relevant for
		 * "mode changes"
		 */
		if (trust_executable_bit &&
244
		    (0100 & (ce->ce_mode ^ st->st_mode)))
245
			changed |= MODE_CHANGED;
246 247
		break;
	case S_IFLNK:
248 249 250
		if (!S_ISLNK(st->st_mode) &&
		    (has_symlinks || !S_ISREG(st->st_mode)))
			changed |= TYPE_CHANGED;
251
		break;
Martin Waitz's avatar
Martin Waitz committed
252
	case S_IFGITLINK:
253
		/* We ignore most of the st_xxx fields for gitlinks */
254 255 256 257
		if (!S_ISDIR(st->st_mode))
			changed |= TYPE_CHANGED;
		else if (ce_compare_gitlink(ce))
			changed |= DATA_CHANGED;
258
		return changed;
259
	default:
260
		die("internal error: ce_mode is %o", ce->ce_mode);
261
	}
262

263
	changed |= match_stat_data(&ce->ce_stat_data, st);
264

265
	/* Racily smudged entry? */
266
	if (!ce->ce_stat_data.sd_size) {
267 268 269 270
		if (!is_empty_blob_sha1(ce->sha1))
			changed |= DATA_CHANGED;
	}

Junio C Hamano's avatar
Junio C Hamano committed
271 272 273
	return changed;
}

274 275
static int is_racy_timestamp(const struct index_state *istate,
			     const struct cache_entry *ce)
276
{
277
	return (!S_ISGITLINK(ce->ce_mode) &&
278 279 280
		istate->timestamp.sec &&
#ifdef USE_NSEC
		 /* nanosecond timestamped files can also be racy! */
281 282 283
		(istate->timestamp.sec < ce->ce_stat_data.sd_mtime.sec ||
		 (istate->timestamp.sec == ce->ce_stat_data.sd_mtime.sec &&
		  istate->timestamp.nsec <= ce->ce_stat_data.sd_mtime.nsec))
284
#else
285
		istate->timestamp.sec <= ce->ce_stat_data.sd_mtime.sec
286 287
#endif
		 );
288 289
}

290
int ie_match_stat(const struct index_state *istate,
291
		  const struct cache_entry *ce, struct stat *st,
292
		  unsigned int options)
Junio C Hamano's avatar
Junio C Hamano committed
293
{
Junio C Hamano's avatar
Junio C Hamano committed
294
	unsigned int changed;
295
	int ignore_valid = options & CE_MATCH_IGNORE_VALID;
296
	int ignore_skip_worktree = options & CE_MATCH_IGNORE_SKIP_WORKTREE;
297
	int assume_racy_is_modified = options & CE_MATCH_RACY_IS_DIRTY;
Junio C Hamano's avatar
Junio C Hamano committed
298 299 300 301

	/*
	 * If it's marked as always valid in the index, it's
	 * valid whatever the checked-out copy says.
302 303
	 *
	 * skip-worktree has the same effect with higher precedence
Junio C Hamano's avatar
Junio C Hamano committed
304
	 */
305 306
	if (!ignore_skip_worktree && ce_skip_worktree(ce))
		return 0;
307
	if (!ignore_valid && (ce->ce_flags & CE_VALID))
Junio C Hamano's avatar
Junio C Hamano committed
308 309
		return 0;

310 311 312 313 314 315 316 317
	/*
	 * Intent-to-add entries have not been added, so the index entry
	 * by definition never matches what is in the work tree until it
	 * actually gets added.
	 */
	if (ce->ce_flags & CE_INTENT_TO_ADD)
		return DATA_CHANGED | TYPE_CHANGED | MODE_CHANGED;

Junio C Hamano's avatar
Junio C Hamano committed
318
	changed = ce_match_stat_basic(ce, st);
Junio C Hamano's avatar
Junio C Hamano committed
319

Junio C Hamano's avatar
Junio C Hamano committed
320 321 322 323 324 325 326 327 328 329 330 331 332 333
	/*
	 * Within 1 second of this sequence:
	 * 	echo xyzzy >file && git-update-index --add file
	 * running this command:
	 * 	echo frotz >file
	 * would give a falsely clean cache entry.  The mtime and
	 * length match the cache, and other stat fields do not change.
	 *
	 * We could detect this at update-index time (the cache entry
	 * being registered/updated records the same time as "now")
	 * and delay the return from git-update-index, but that would
	 * effectively mean we can make at most one commit per second,
	 * which is not acceptable.  Instead, we check cache entries
	 * whose mtime are the same as the index file timestamp more
Junio C Hamano's avatar
Junio C Hamano committed
334
	 * carefully than others.
Junio C Hamano's avatar
Junio C Hamano committed
335
	 */
336
	if (!changed && is_racy_timestamp(istate, ce)) {
337 338 339 340 341
		if (assume_racy_is_modified)
			changed |= DATA_CHANGED;
		else
			changed |= ce_modified_check_fs(ce, st);
	}
342

Junio C Hamano's avatar
Junio C Hamano committed
343
	return changed;
344 345
}

346
int ie_modified(const struct index_state *istate,
347 348
		const struct cache_entry *ce,
		struct stat *st, unsigned int options)
349
{
Junio C Hamano's avatar
Junio C Hamano committed
350
	int changed, changed_fs;
351 352

	changed = ie_match_stat(istate, ce, st, options);
353 354 355 356 357 358 359 360 361
	if (!changed)
		return 0;
	/*
	 * If the mode or type has changed, there's no point in trying
	 * to refresh the entry - it's not going to match
	 */
	if (changed & (MODE_CHANGED | TYPE_CHANGED))
		return changed;

362 363 364 365 366 367 368 369 370 371 372 373 374
	/*
	 * Immediately after read-tree or update-index --cacheinfo,
	 * the length field is zero, as we have never even read the
	 * lstat(2) information once, and we cannot trust DATA_CHANGED
	 * returned by ie_match_stat() which in turn was returned by
	 * ce_match_stat_basic() to signal that the filesize of the
	 * blob changed.  We have to actually go to the filesystem to
	 * see if the contents match, and if so, should answer "unchanged".
	 *
	 * The logic does not apply to gitlinks, as ce_match_stat_basic()
	 * already has checked the actual HEAD from the filesystem in the
	 * subproject.  If ie_match_stat() already said it is different,
	 * then we know it is.
375
	 */
376
	if ((changed & DATA_CHANGED) &&
377
	    (S_ISGITLINK(ce->ce_mode) || ce->ce_stat_data.sd_size != 0))
378 379
		return changed;

Junio C Hamano's avatar
Junio C Hamano committed
380 381 382
	changed_fs = ce_modified_check_fs(ce, st);
	if (changed_fs)
		return changed | changed_fs;
383 384 385
	return 0;
}

386 387 388 389 390 391 392 393 394 395 396 397
int base_name_compare(const char *name1, int len1, int mode1,
		      const char *name2, int len2, int mode2)
{
	unsigned char c1, c2;
	int len = len1 < len2 ? len1 : len2;
	int cmp;

	cmp = memcmp(name1, name2, len);
	if (cmp)
		return cmp;
	c1 = name1[len];
	c2 = name2[len];
398
	if (!c1 && S_ISDIR(mode1))
399
		c1 = '/';
400
	if (!c2 && S_ISDIR(mode2))
401 402 403 404
		c2 = '/';
	return (c1 < c2) ? -1 : (c1 > c2) ? 1 : 0;
}

405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439
/*
 * df_name_compare() is identical to base_name_compare(), except it
 * compares conflicting directory/file entries as equal. Note that
 * while a directory name compares as equal to a regular file, they
 * then individually compare _differently_ to a filename that has
 * a dot after the basename (because '\0' < '.' < '/').
 *
 * This is used by routines that want to traverse the git namespace
 * but then handle conflicting entries together when possible.
 */
int df_name_compare(const char *name1, int len1, int mode1,
		    const char *name2, int len2, int mode2)
{
	int len = len1 < len2 ? len1 : len2, cmp;
	unsigned char c1, c2;

	cmp = memcmp(name1, name2, len);
	if (cmp)
		return cmp;
	/* Directories and files compare equal (same length, same name) */
	if (len1 == len2)
		return 0;
	c1 = name1[len];
	if (!c1 && S_ISDIR(mode1))
		c1 = '/';
	c2 = name2[len];
	if (!c2 && S_ISDIR(mode2))
		c2 = '/';
	if (c1 == '/' && !c2)
		return 0;
	if (c2 == '/' && !c1)
		return 0;
	return c1 - c2;
}

440
int name_compare(const char *name1, size_t len1, const char *name2, size_t len2)
441
{
442 443
	size_t min_len = (len1 < len2) ? len1 : len2;
	int cmp = memcmp(name1, name2, min_len);
444 445 446 447 448 449
	if (cmp)
		return cmp;
	if (len1 < len2)
		return -1;
	if (len1 > len2)
		return 1;
450 451 452 453 454 455 456 457 458 459
	return 0;
}

int cache_name_stage_compare(const char *name1, int len1, int stage1, const char *name2, int len2, int stage2)
{
	int cmp;

	cmp = name_compare(name1, len1, name2, len2);
	if (cmp)
		return cmp;
Junio C Hamano's avatar
Junio C Hamano committed
460

461
	if (stage1 < stage2)
462
		return -1;
463
	if (stage1 > stage2)
464
		return 1;
465 466 467
	return 0;
}

468
static int index_name_stage_pos(const struct index_state *istate, const char *name, int namelen, int stage)
469 470 471 472
{
	int first, last;

	first = 0;
473
	last = istate->cache_nr;
474 475
	while (last > first) {
		int next = (last + first) >> 1;
476
		struct cache_entry *ce = istate->cache[next];
477
		int cmp = cache_name_stage_compare(name, namelen, stage, ce->name, ce_namelen(ce), ce_stage(ce));
478
		if (!cmp)
479
			return next;
480 481 482 483 484 485
		if (cmp < 0) {
			last = next;
			continue;
		}
		first = next+1;
	}
486
	return -first-1;
487 488
}

489 490 491 492 493
int index_name_pos(const struct index_state *istate, const char *name, int namelen)
{
	return index_name_stage_pos(istate, name, namelen, 0);
}

494
/* Remove entry, return true if there are more entries to go.. */
495
int remove_index_entry_at(struct index_state *istate, int pos)
496
{
497 498
	struct cache_entry *ce = istate->cache[pos];

499
	record_resolve_undo(istate, ce);
500
	remove_name_hash(istate, ce);
501
	save_or_free_index_entry(istate, ce);
502
	istate->cache_changed |= CE_ENTRY_REMOVED;
503 504
	istate->cache_nr--;
	if (pos >= istate->cache_nr)
505
		return 0;
506 507 508
	memmove(istate->cache + pos,
		istate->cache + pos + 1,
		(istate->cache_nr - pos) * sizeof(struct cache_entry *));
509 510 511
	return 1;
}

512
/*
Ondřej Bílka's avatar
Ondřej Bílka committed
513
 * Remove all cache entries marked for removal, that is where
514 515 516 517 518 519 520 521 522
 * CE_REMOVE is set in ce_flags.  This is much more effective than
 * calling remove_index_entry_at() for each entry to be removed.
 */
void remove_marked_cache_entries(struct index_state *istate)
{
	struct cache_entry **ce_array = istate->cache;
	unsigned int i, j;

	for (i = j = 0; i < istate->cache_nr; i++) {
523
		if (ce_array[i]->ce_flags & CE_REMOVE) {
524
			remove_name_hash(istate, ce_array[i]);
525
			save_or_free_index_entry(istate, ce_array[i]);
526
		}
527 528 529
		else
			ce_array[j++] = ce_array[i];
	}
530 531
	if (j == istate->cache_nr)
		return;
532
	istate->cache_changed |= CE_ENTRY_REMOVED;
533 534 535
	istate->cache_nr = j;
}

536
int remove_file_from_index(struct index_state *istate, const char *path)
537
{
538
	int pos = index_name_pos(istate, path, strlen(path));
539 540
	if (pos < 0)
		pos = -pos-1;
541
	cache_tree_invalidate_path(istate, path);
542 543
	while (pos < istate->cache_nr && !strcmp(istate->cache[pos]->name, path))
		remove_index_entry_at(istate, pos);
544 545 546
	return 0;
}

547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574
static int compare_name(struct cache_entry *ce, const char *path, int namelen)
{
	return namelen != ce_namelen(ce) || memcmp(path, ce->name, namelen);
}

static int index_name_pos_also_unmerged(struct index_state *istate,
	const char *path, int namelen)
{
	int pos = index_name_pos(istate, path, namelen);
	struct cache_entry *ce;

	if (pos >= 0)
		return pos;

	/* maybe unmerged? */
	pos = -1 - pos;
	if (pos >= istate->cache_nr ||
			compare_name((ce = istate->cache[pos]), path, namelen))
		return -1;

	/* order of preference: stage 2, 1, 3 */
	if (ce_stage(ce) == 1 && pos + 1 < istate->cache_nr &&
			ce_stage((ce = istate->cache[pos + 1])) == 2 &&
			!compare_name(ce, path, namelen))
		pos++;
	return pos;
}

575 576 577 578 579 580 581 582 583 584 585 586 587 588 589
static int different_name(struct cache_entry *ce, struct cache_entry *alias)
{
	int len = ce_namelen(ce);
	return ce_namelen(alias) != len || memcmp(ce->name, alias->name, len);
}

/*
 * If we add a filename that aliases in the cache, we will use the
 * name that we already have - but we don't want to update the same
 * alias twice, because that implies that there were actually two
 * different files with aliasing names!
 *
 * So we use the CE_ADDED flag to verify that the alias was an old
 * one before we accept it as
 */
590 591 592
static struct cache_entry *create_alias_ce(struct index_state *istate,
					   struct cache_entry *ce,
					   struct cache_entry *alias)
593 594 595 596 597 598 599 600 601 602 603 604
{
	int len;
	struct cache_entry *new;

	if (alias->ce_flags & CE_ADDED)
		die("Will not add file alias '%s' ('%s' already exists in index)", ce->name, alias->name);

	/* Ok, create the new entry using the name of the existing alias */
	len = ce_namelen(alias);
	new = xcalloc(1, cache_entry_size(len));
	memcpy(new->name, alias->name, len);
	copy_cache_entry(new, ce);
605
	save_or_free_index_entry(istate, ce);
606 607 608
	return new;
}

609
void set_object_name_for_intent_to_add_entry(struct cache_entry *ce)
610 611 612 613 614 615 616
{
	unsigned char sha1[20];
	if (write_sha1_file("", 0, blob_type, sha1))
		die("cannot create an empty blob in the object database");
	hashcpy(ce->sha1, sha1);
}

617
int add_to_index(struct index_state *istate, const char *path, struct stat *st, int flags)
Johannes Schindelin's avatar
Johannes Schindelin committed
618
{
619
	int size, namelen, was_same;
620
	mode_t st_mode = st->st_mode;
621
	struct cache_entry *ce, *alias;
622
	unsigned ce_option = CE_MATCH_IGNORE_VALID|CE_MATCH_IGNORE_SKIP_WORKTREE|CE_MATCH_RACY_IS_DIRTY;
623 624
	int verbose = flags & (ADD_CACHE_VERBOSE | ADD_CACHE_PRETEND);
	int pretend = flags & ADD_CACHE_PRETEND;
625 626 627
	int intent_only = flags & ADD_CACHE_INTENT;
	int add_option = (ADD_CACHE_OK_TO_ADD|ADD_CACHE_OK_TO_REPLACE|
			  (intent_only ? ADD_CACHE_NEW_ONLY : 0));
Johannes Schindelin's avatar
Johannes Schindelin committed
628

629
	if (!S_ISREG(st_mode) && !S_ISLNK(st_mode) && !S_ISDIR(st_mode))
630
		return error("%s: can only add regular files, symbolic links or git-directories", path);
Johannes Schindelin's avatar
Johannes Schindelin committed
631 632

	namelen = strlen(path);
633
	if (S_ISDIR(st_mode)) {
634 635 636
		while (namelen && path[namelen-1] == '/')
			namelen--;
	}
Johannes Schindelin's avatar
Johannes Schindelin committed
637 638 639
	size = cache_entry_size(namelen);
	ce = xcalloc(1, size);
	memcpy(ce->name, path, namelen);
640
	ce->ce_namelen = namelen;
641 642
	if (!intent_only)
		fill_stat_cache_info(ce, st);
643 644
	else
		ce->ce_flags |= CE_INTENT_TO_ADD;
Johannes Schindelin's avatar
Johannes Schindelin committed
645

646
	if (trust_executable_bit && has_symlinks)
647
		ce->ce_mode = create_ce_mode(st_mode);
648
	else {
649 650
		/* If there is an existing entry, pick the mode bits and type
		 * from it, otherwise assume unexecutable regular file.
Johannes Schindelin's avatar
Johannes Schindelin committed
651
		 */
652
		struct cache_entry *ent;
653
		int pos = index_name_pos_also_unmerged(istate, path, namelen);
654

655
		ent = (0 <= pos) ? istate->cache[pos] : NULL;
656
		ce->ce_mode = ce_mode_from_stat(ent, st_mode);
Johannes Schindelin's avatar
Johannes Schindelin committed
657 658
	}

659 660 661 662 663 664 665 666 667 668 669 670 671 672
	/* When core.ignorecase=true, determine if a directory of the same name but differing
	 * case already exists within the Git repository.  If it does, ensure the directory
	 * case of the file being added to the repository matches (is folded into) the existing
	 * entry's directory case.
	 */
	if (ignore_case) {
		const char *startPtr = ce->name;
		const char *ptr = startPtr;
		while (*ptr) {
			while (*ptr && *ptr != '/')
				++ptr;
			if (*ptr == '/') {
				struct cache_entry *foundce;
				++ptr;
673
				foundce = index_dir_exists(istate, ce->name, ptr - ce->name - 1);
674 675 676 677 678 679 680 681
				if (foundce) {
					memcpy((void *)startPtr, foundce->name + (startPtr - ce->name), ptr - startPtr);
					startPtr = ptr;
				}
			}
		}
	}

682
	alias = index_file_exists(istate, ce->name, ce_namelen(ce), ignore_case);
683
	if (alias && !ce_stage(alias) && !ie_match_stat(istate, alias, st, ce_option)) {
684 685
		/* Nothing changed, really */
		free(ce);
686 687
		if (!S_ISGITLINK(alias->ce_mode))
			ce_mark_uptodate(alias);
688
		alias->ce_flags |= CE_ADDED;
689 690
		return 0;
	}
691
	if (!intent_only) {
692
		if (index_path(ce->sha1, path, st, HASH_WRITE_OBJECT))
693 694
			return error("unable to index file %s", path);
	} else
695
		set_object_name_for_intent_to_add_entry(ce);
696

697
	if (ignore_case && alias && different_name(ce, alias))
698
		ce = create_alias_ce(istate, ce, alias);
699
	ce->ce_flags |= CE_ADDED;
700

Junio C Hamano's avatar
Junio C Hamano committed
701
	/* It was suspected to be racily clean, but it turns out to be Ok */
702 703 704 705 706 707 708
	was_same = (alias &&
		    !ce_stage(alias) &&
		    !hashcmp(alias->sha1, ce->sha1) &&
		    ce->ce_mode == alias->ce_mode);

	if (pretend)
		;
709
	else if (add_index_entry(istate, ce, add_option))
710
		return error("unable to add %s to index",path);
711
	if (verbose && !was_same)
Johannes Schindelin's avatar
Johannes Schindelin committed
712 713 714 715
		printf("add '%s'\n", path);
	return 0;
}

716
int add_file_to_index(struct index_state *istate, const char *path, int flags)
717 718 719
{
	struct stat st;
	if (lstat(path, &st))
720
		die_errno("unable to stat '%s'", path);
721
	return add_to_index(istate, path, &st, flags);
722 723
}

724 725
struct cache_entry *make_cache_entry(unsigned int mode,
		const unsigned char *sha1, const char *path, int stage,
726
		unsigned int refresh_options)
727 728
{
	int size, len;
729
	struct cache_entry *ce, *ret;
730

731 732
	if (!verify_path(path)) {
		error("Invalid path '%s'", path);
733
		return NULL;
734
	}
735 736 737 738 739 740 741

	len = strlen(path);
	size = cache_entry_size(len);
	ce = xcalloc(1, size);

	hashcpy(ce->sha1, sha1);
	memcpy(ce->name, path, len);
742 743
	ce->ce_flags = create_ce_flags(stage);
	ce->ce_namelen = len;
744 745
	ce->ce_mode = create_ce_mode(mode);

746 747 748 749 750 751 752
	ret = refresh_cache_entry(ce, refresh_options);
	if (!ret) {
		free(ce);
		return NULL;
	} else {
		return ret;
	}
753 754
}

755
int ce_same_name(const struct cache_entry *a, const struct cache_entry *b)
756 757 758 759 760
{
	int len = ce_namelen(a);
	return ce_namelen(b) == len && !memcmp(a->name, b->name, len);
}

761 762 763 764 765 766 767 768 769 770 771 772 773 774 775
/*
 * We fundamentally don't like some paths: we don't want
 * dot or dot-dot anywhere, and for obvious reasons don't
 * want to recurse into ".git" either.
 *
 * Also, we don't want double slashes or slashes at the
 * end that can make pathnames ambiguous.
 */
static int verify_dotfile(const char *rest)
{
	/*
	 * The first character was '.', but that
	 * has already been discarded, we now test
	 * the rest.
	 */
776

777
	/* "." is not allowed */
778
	if (*rest == '\0' || is_dir_sep(*rest))
779 780
		return 0;

781
	switch (*rest) {
782 783 784 785 786
	/*
	 * ".git" followed by  NUL or slash is bad. This
	 * shares the path end test with the ".." case.
	 */
	case 'g':
787 788
	case 'G':
		if (rest[1] != 'i' && rest[1] != 'I')
789
			break;
790
		if (rest[2] != 't' && rest[2] != 'T')
791 792 793 794
			break;
		rest += 2;
	/* fallthrough */
	case '.':
795
		if (rest[1] == '\0' || is_dir_sep(rest[1]))
796 797 798 799 800 801 802 803 804
			return 0;
	}
	return 1;
}

int verify_path(const char *path)
{
	char c;

805 806 807
	if (has_dos_drive_prefix(path))
		return 0;

808 809 810 811
	goto inside;
	for (;;) {
		if (!c)
			return 1;
812
		if (is_dir_sep(c)) {
813
inside:
814 815
			if (protect_hfs && is_hfs_dotgit(path))
				return 0;
816 817
			if (protect_ntfs && is_ntfs_dotgit(path))
				return 0;
818
			c = *path++;
819 820 821
			if ((c == '.' && !verify_dotfile(path)) ||
			    is_dir_sep(c) || c == '\0')
				return 0;
822 823 824 825 826
		}
		c = *path++;
	}
}

827 828 829
/*
 * Do we have another file that has the beginning components being a
 * proper superset of the name we're trying to add?
830
 */
831 832
static int has_file_name(struct index_state *istate,
			 const struct cache_entry *ce, int pos, int ok_to_replace)
833
{
834 835
	int retval = 0;
	int len = ce_namelen(ce);
836
	int stage = ce_stage(ce);
837
	const char *name = ce->name;
838

839 840
	while (pos < istate->cache_nr) {
		struct cache_entry *p = istate->cache[pos++];
841

842
		if (len >= ce_namelen(p))
843
			break;
844 845
		if (memcmp(name, p->name, len))
			break;
846 847
		if (ce_stage(p) != stage)
			continue;
848 849
		if (p->name[len] != '/')
			continue;
850
		if (p->ce_flags & CE_REMOVE)
851
			continue;
852 853 854
		retval = -1;
		if (!ok_to_replace)
			break;
855
		remove_index_entry_at(istate, --pos);
856
	}
857 858
	return retval;
}
859

860 861 862 863
/*
 * Do we have another file with a pathname that is a proper
 * subset of the name we're trying to add?
 */
864 865
static int has_dir_name(struct index_state *istate,
			const struct cache_entry *ce, int pos, int ok_to_replace)
866 867
{
	int retval = 0;
868
	int stage = ce_stage(ce);
869 870
	const char *name = ce->name;
	const char *slash = name + ce_namelen(ce);
871

872 873
	for (;;) {
		int len;
874

875 876 877 878 879 880 881
		for (;;) {
			if (*--slash == '/')
				break;
			if (slash <= ce->name)
				return retval;
		}
		len = slash - name;
882

883
		pos = index_name_stage_pos(istate, name, len, stage);
884
		if (pos >= 0) {
885 886 887 888 889 890 891 892
			/*
			 * Found one, but not so fast.  This could
			 * be a marker that says "I was here, but
			 * I am being removed".  Such an entry is
			 * not a part of the resulting tree, and
			 * it is Ok to have a directory at the same
			 * path.
			 */
893
			if (!(istate->cache[pos]->ce_flags & CE_REMOVE)) {
894 895 896
				retval = -1;
				if (!ok_to_replace)
					break;
897
				remove_index_entry_at(istate, pos);
898 899
				continue;
			}
900
		}
901 902
		else
			pos = -pos-1;
903 904 905 906

		/*
		 * Trivial optimization: if we find an entry that
		 * already matches the sub-directory, then we know
907
		 * we're ok, and we can exit.
908
		 */
909 910
		while (pos < istate->cache_nr) {
			struct cache_entry *p = istate->cache[pos];
911 912 913 914
			if ((ce_namelen(p) <= len) ||
			    (p->name[len] != '/') ||
			    memcmp(p->name, name, len))
				break; /* not our subdirectory */
915 916 917
			if (ce_stage(p) == stage && !(p->ce_flags & CE_REMOVE))
				/*
				 * p is at the same stage as our entry, and
918 919 920 921 922 923
				 * is a subdirectory of what we are looking
				 * at, so we cannot have conflicts at our
				 * level or anything shorter.
				 */
				return retval;
			pos++;
924
		}
925
	}
926 927 928 929 930 931 932
	return retval;
}

/* We may be in a situation where we already have path/file and path
 * is being added, or we already have path and path/file is being
 * added.  Either one would result in a nonsense tree that has path
 * twice when git-write-tree tries to write it out.  Prevent it.
Junio C Hamano's avatar
Junio C Hamano committed
933
 *
934 935 936 937
 * If ok-to-replace is specified, we remove the conflicting entries
 * from the cache so the caller should recompute the insert position.
 * When this happens, we return non-zero.
 */
938 939 940
static int check_file_directory_conflict(struct index_state *istate,
					 const struct cache_entry *ce,
					 int pos, int ok_to_replace)
941
{
942 943 944 945 946
	int retval;

	/*
	 * When ce is an "I am going away" entry, we allow it to be added
	 */
947
	if (ce->ce_flags & CE_REMOVE)
948 949
		return 0;

950 951 952
	/*
	 * We check if the path is a sub-path of a subsequent pathname
	 * first, since removing those will not change the position
953
	 * in the array.
954
	 */
955
	retval = has_file_name(istate, ce, pos, ok_to_replace);
956

957 958 959 960
	/*
	 * Then check if the path might have a clashing sub-directory
	 * before it.
	 */
961
	return retval + has_dir_name(istate, ce, pos, ok_to_replace);
962 963
}

964
static int add_index_entry_with_check(struct index_state *istate, struct cache_entry *ce, int option)
965 966
{
	int pos;
967 968
	int ok_to_add = option & ADD_CACHE_OK_TO_ADD;
	int ok_to_replace = option & ADD_CACHE_OK_TO_REPLACE;
969
	int skip_df_check = option & ADD_CACHE_SKIP_DFCHECK;
970
	int new_only = option & ADD_CACHE_NEW_ONLY;
Junio C Hamano's avatar
Junio C Hamano committed
971

972 973
	if (!(option & ADD_CACHE_KEEP_CACHE_TREE))
		cache_tree_invalidate_path(istate, ce->name);
974
	pos = index_name_stage_pos(istate, ce->name, ce_namelen(ce), ce_stage(ce));
975

Junio C Hamano's avatar
Junio C Hamano committed
976
	/* existing match? Just replace it. */
977
	if (pos >= 0) {
978 979
		if (!new_only)
			replace_index_entry(istate, pos, ce);
980 981
		return 0;
	}
982
	pos = -pos-1;
983

984 985 986 987
	/*
	 * Inserting a merged entry ("stage 0") into the index
	 * will always replace all non-merged entries..
	 */
988 989
	if (pos < istate->cache_nr && ce_stage(ce) == 0) {
		while (ce_same_name(istate->cache[pos], ce)) {
990
			ok_to_add = 1;
991
			if (!remove_index_entry_at(istate, pos))
992 993 994 995
				break;
		}
	}

996 997
	if (!ok_to_add)
		return -1;
998
	if (!verify_path(ce->name))
999
		return error("Invalid path '%s'", ce->name);
1000

Junio C Hamano's avatar
Junio C Hamano committed
1001
	if (!skip_df_check &&
1002
	    check_file_directory_conflict(istate, ce, pos, ok_to_replace)) {
1003
		if (!ok_to_replace)
1004 1005
			return error("'%s' appears as both a file and as a directory",
				     ce->name);
1006
		pos = index_name_stage_pos(istate, ce->name, ce_namelen(ce), ce_stage(ce));
1007 1008
		pos = -pos-1;
	}
1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024
	return pos + 1;
}

int add_index_entry(struct index_state *istate, struct cache_entry *ce, int option)
{
	int pos;

	if (option & ADD_CACHE_JUST_APPEND)
		pos = istate->cache_nr;
	else {
		int ret;
		ret = add_index_entry_with_check(istate, ce, option);
		if (ret <= 0)
			return ret;
		pos = ret - 1;
	}
1025

1026
	/* Make sure the array is big enough .. */
1027
	ALLOC_GROW(istate->cache, istate->cache_nr + 1, istate->cache_alloc);
1028 1029

	/* Add it in.. */
1030
	istate->cache_nr++;
1031
	if (istate->cache_nr > pos + 1)
1032 1033 1034
		memmove(istate->cache + pos + 1,
			istate->cache + pos,
			(istate->cache_nr - pos - 1) * sizeof(ce));
1035
	set_index_entry(istate, pos, ce);
1036
	istate->cache_changed |= CE_ENTRY_ADDED;
1037 1038 1039
	return 0;
}

1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050
/*
 * "refresh" does not calculate a new sha1 file or bring the
 * cache up-to-date for mode/content changes. But what it
 * _does_ do is to "re-match" the stat information of a file
 * with the cache, so that you can refresh the cache for a
 * file that hasn't been changed but where the stat entry is
 * out of date.
 *
 * For example, you'd want to do this after doing a "git-read-tree",
 * to link up the stat cache details with the proper files.
 */
1051
static struct cache_entry *refresh_cache_ent(struct index_state *istate,
1052
					     struct cache_entry *ce,
1053 1054
					     unsigned int options, int *err,
					     int *changed_ret)
1055 1056 1057 1058
{
	struct stat st;
	struct cache_entry *updated;
	int changed, size;
1059