read-cache.c 60.8 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 "cache-tree.h"
9
#include "refs.h"
10
#include "dir.h"
11 12
#include "tree.h"
#include "commit.h"
13
#include "blob.h"
14
#include "resolve-undo.h"
15 16
#include "strbuf.h"
#include "varint.h"
Duy Nguyen's avatar
Duy Nguyen committed
17
#include "split-index.h"
18
#include "sigchain.h"
19

20 21
static struct cache_entry *refresh_cache_entry(struct cache_entry *ce,
					       unsigned int options);
22

23 24 25 26
/* Mask for the name length in ce_flags in the on-disk index */

#define CE_NAMEMASK  (0x0fff)

27 28 29 30 31 32 33 34 35 36 37
/* 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" */
38
#define CACHE_EXT_RESOLVE_UNDO 0x52455543 /* "REUC" */
Duy Nguyen's avatar
Duy Nguyen committed
39 40 41
#define CACHE_EXT_LINK 0x6c696e6b	  /* "link" */

/* changes that can be kept in $GIT_DIR/index (basically all extensions) */
42
#define EXTMASK (RESOLVE_UNDO_CHANGED | CACHE_TREE_CHANGED | \
43 44
		 CE_ENTRY_ADDED | CE_ENTRY_REMOVED | CE_ENTRY_CHANGED | \
		 SPLIT_INDEX_ORDERED)
45

46
struct index_state the_index;
47
static const char *alternate_index_output;
48

Junio C Hamano's avatar
Junio C Hamano committed
49 50 51
static void set_index_entry(struct index_state *istate, int nr, struct cache_entry *ce)
{
	istate->cache[nr] = ce;
52
	add_name_hash(istate, ce);
Junio C Hamano's avatar
Junio C Hamano committed
53 54
}

55 56 57 58
static void replace_index_entry(struct index_state *istate, int nr, struct cache_entry *ce)
{
	struct cache_entry *old = istate->cache[nr];

59
	replace_index_entry_in_base(istate, old, ce);
60
	remove_name_hash(istate, old);
61
	free(old);
62
	set_index_entry(istate, nr, ce);
63
	ce->ce_flags |= CE_UPDATE_IN_BASE;
64
	istate->cache_changed |= CE_ENTRY_CHANGED;
65 66
}

67 68 69 70 71 72 73
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);
74
	new->ce_flags &= ~CE_HASHED;
75
	new->ce_namelen = namelen;
Duy Nguyen's avatar
Duy Nguyen committed
76
	new->index = 0;
77 78
	memcpy(new->name, new_name, namelen + 1);

79
	cache_tree_invalidate_path(istate, old->name);
80 81 82 83
	remove_index_entry_at(istate, nr);
	add_index_entry(istate, new, ADD_CACHE_OK_TO_ADD|ADD_CACHE_OK_TO_REPLACE);
}

84 85 86 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
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;
}

139 140 141 142 143 144 145
/*
 * 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)
{
146
	fill_stat_data(&ce->ce_stat_data, st);
Junio C Hamano's avatar
Junio C Hamano committed
147 148

	if (assume_unchanged)
149
		ce->ce_flags |= CE_VALID;
150 151 152

	if (S_ISREG(st->st_mode))
		ce_mark_uptodate(ce);
153 154
}

155
static int ce_compare_data(const struct cache_entry *ce, struct stat *st)
Junio C Hamano's avatar
Junio C Hamano committed
156 157 158 159 160 161
{
	int match = -1;
	int fd = open(ce->name, O_RDONLY);

	if (fd >= 0) {
		unsigned char sha1[20];
162
		if (!index_fd(sha1, fd, st, OBJ_BLOB, ce->name, 0))
163
			match = hashcmp(sha1, ce->sha1);
164
		/* index_fd() closed the file descriptor already */
Junio C Hamano's avatar
Junio C Hamano committed
165 166 167 168
	}
	return match;
}

169
static int ce_compare_link(const struct cache_entry *ce, size_t expected_size)
Junio C Hamano's avatar
Junio C Hamano committed
170 171 172 173
{
	int match = -1;
	void *buffer;
	unsigned long size;
174
	enum object_type type;
175
	struct strbuf sb = STRBUF_INIT;
Junio C Hamano's avatar
Junio C Hamano committed
176

177
	if (strbuf_readlink(&sb, ce->name, expected_size))
Junio C Hamano's avatar
Junio C Hamano committed
178
		return -1;
179

180
	buffer = read_sha1_file(ce->sha1, &type, &size);
181 182 183 184
	if (buffer) {
		if (size == sb.len)
			match = memcmp(buffer, sb.buf, size);
		free(buffer);
Junio C Hamano's avatar
Junio C Hamano committed
185
	}
186
	strbuf_release(&sb);
Junio C Hamano's avatar
Junio C Hamano committed
187 188 189
	return match;
}

190
static int ce_compare_gitlink(const struct cache_entry *ce)
191 192 193 194 195
{
	unsigned char sha1[20];

	/*
	 * We don't actually require that the .git directory
Martin Waitz's avatar
Martin Waitz committed
196
	 * under GITLINK directory be a valid git directory. It
197 198 199 200 201 202 203 204 205 206
	 * 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);
}

207
static int ce_modified_check_fs(const struct cache_entry *ce, struct stat *st)
Junio C Hamano's avatar
Junio C Hamano committed
208 209 210 211 212 213 214
{
	switch (st->st_mode & S_IFMT) {
	case S_IFREG:
		if (ce_compare_data(ce, st))
			return DATA_CHANGED;
		break;
	case S_IFLNK:
215
		if (ce_compare_link(ce, xsize_t(st->st_size)))
Junio C Hamano's avatar
Junio C Hamano committed
216 217
			return DATA_CHANGED;
		break;
218
	case S_IFDIR:
219
		if (S_ISGITLINK(ce->ce_mode))
220
			return ce_compare_gitlink(ce) ? DATA_CHANGED : 0;
Junio C Hamano's avatar
Junio C Hamano committed
221 222 223 224 225 226
	default:
		return TYPE_CHANGED;
	}
	return 0;
}

227
static int ce_match_stat_basic(const struct cache_entry *ce, struct stat *st)
228 229 230
{
	unsigned int changed = 0;

231 232 233 234
	if (ce->ce_flags & CE_REMOVE)
		return MODE_CHANGED | DATA_CHANGED | TYPE_CHANGED;

	switch (ce->ce_mode & S_IFMT) {
235 236
	case S_IFREG:
		changed |= !S_ISREG(st->st_mode) ? TYPE_CHANGED : 0;
Junio C Hamano's avatar
Junio C Hamano committed
237 238 239 240
		/* We consider only the owner x bit to be relevant for
		 * "mode changes"
		 */
		if (trust_executable_bit &&
241
		    (0100 & (ce->ce_mode ^ st->st_mode)))
242
			changed |= MODE_CHANGED;
243 244
		break;
	case S_IFLNK:
245 246 247
		if (!S_ISLNK(st->st_mode) &&
		    (has_symlinks || !S_ISREG(st->st_mode)))
			changed |= TYPE_CHANGED;
248
		break;
Martin Waitz's avatar
Martin Waitz committed
249
	case S_IFGITLINK:
250
		/* We ignore most of the st_xxx fields for gitlinks */
251 252 253 254
		if (!S_ISDIR(st->st_mode))
			changed |= TYPE_CHANGED;
		else if (ce_compare_gitlink(ce))
			changed |= DATA_CHANGED;
255
		return changed;
256
	default:
257
		die("internal error: ce_mode is %o", ce->ce_mode);
258
	}
259

260
	changed |= match_stat_data(&ce->ce_stat_data, st);
261

262
	/* Racily smudged entry? */
263
	if (!ce->ce_stat_data.sd_size) {
264 265 266 267
		if (!is_empty_blob_sha1(ce->sha1))
			changed |= DATA_CHANGED;
	}

Junio C Hamano's avatar
Junio C Hamano committed
268 269 270
	return changed;
}

271 272
static int is_racy_timestamp(const struct index_state *istate,
			     const struct cache_entry *ce)
273
{
274
	return (!S_ISGITLINK(ce->ce_mode) &&
275 276 277
		istate->timestamp.sec &&
#ifdef USE_NSEC
		 /* nanosecond timestamped files can also be racy! */
278 279 280
		(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))
281
#else
282
		istate->timestamp.sec <= ce->ce_stat_data.sd_mtime.sec
283 284
#endif
		 );
285 286
}

287
int ie_match_stat(const struct index_state *istate,
288
		  const struct cache_entry *ce, struct stat *st,
289
		  unsigned int options)
Junio C Hamano's avatar
Junio C Hamano committed
290
{
Junio C Hamano's avatar
Junio C Hamano committed
291
	unsigned int changed;
292
	int ignore_valid = options & CE_MATCH_IGNORE_VALID;
293
	int ignore_skip_worktree = options & CE_MATCH_IGNORE_SKIP_WORKTREE;
294
	int assume_racy_is_modified = options & CE_MATCH_RACY_IS_DIRTY;
Junio C Hamano's avatar
Junio C Hamano committed
295 296 297 298

	/*
	 * If it's marked as always valid in the index, it's
	 * valid whatever the checked-out copy says.
299 300
	 *
	 * skip-worktree has the same effect with higher precedence
Junio C Hamano's avatar
Junio C Hamano committed
301
	 */
302 303
	if (!ignore_skip_worktree && ce_skip_worktree(ce))
		return 0;
304
	if (!ignore_valid && (ce->ce_flags & CE_VALID))
Junio C Hamano's avatar
Junio C Hamano committed
305 306
		return 0;

307 308 309 310 311 312 313 314
	/*
	 * 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
315
	changed = ce_match_stat_basic(ce, st);
Junio C Hamano's avatar
Junio C Hamano committed
316

Junio C Hamano's avatar
Junio C Hamano committed
317 318 319 320 321 322 323 324 325 326 327 328 329 330
	/*
	 * 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
331
	 * carefully than others.
Junio C Hamano's avatar
Junio C Hamano committed
332
	 */
333
	if (!changed && is_racy_timestamp(istate, ce)) {
334 335 336 337 338
		if (assume_racy_is_modified)
			changed |= DATA_CHANGED;
		else
			changed |= ce_modified_check_fs(ce, st);
	}
339

Junio C Hamano's avatar
Junio C Hamano committed
340
	return changed;
341 342
}

343
int ie_modified(const struct index_state *istate,
344 345
		const struct cache_entry *ce,
		struct stat *st, unsigned int options)
346
{
Junio C Hamano's avatar
Junio C Hamano committed
347
	int changed, changed_fs;
348 349

	changed = ie_match_stat(istate, ce, st, options);
350 351 352 353 354 355 356 357 358
	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;

359 360 361 362 363 364 365 366 367 368 369 370 371
	/*
	 * 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.
372
	 */
373
	if ((changed & DATA_CHANGED) &&
374
	    (S_ISGITLINK(ce->ce_mode) || ce->ce_stat_data.sd_size != 0))
375 376
		return changed;

Junio C Hamano's avatar
Junio C Hamano committed
377 378 379
	changed_fs = ce_modified_check_fs(ce, st);
	if (changed_fs)
		return changed | changed_fs;
380 381 382
	return 0;
}

383 384 385 386 387 388 389 390 391 392 393 394
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];
395
	if (!c1 && S_ISDIR(mode1))
396
		c1 = '/';
397
	if (!c2 && S_ISDIR(mode2))
398 399 400 401
		c2 = '/';
	return (c1 < c2) ? -1 : (c1 > c2) ? 1 : 0;
}

402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436
/*
 * 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;
}

437
int name_compare(const char *name1, size_t len1, const char *name2, size_t len2)
438
{
439 440
	size_t min_len = (len1 < len2) ? len1 : len2;
	int cmp = memcmp(name1, name2, min_len);
441 442 443 444 445 446
	if (cmp)
		return cmp;
	if (len1 < len2)
		return -1;
	if (len1 > len2)
		return 1;
447 448 449 450 451 452 453 454 455 456
	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
457

458
	if (stage1 < stage2)
459
		return -1;
460
	if (stage1 > stage2)
461
		return 1;
462 463 464
	return 0;
}

465
static int index_name_stage_pos(const struct index_state *istate, const char *name, int namelen, int stage)
466 467 468 469
{
	int first, last;

	first = 0;
470
	last = istate->cache_nr;
471 472
	while (last > first) {
		int next = (last + first) >> 1;
473
		struct cache_entry *ce = istate->cache[next];
474
		int cmp = cache_name_stage_compare(name, namelen, stage, ce->name, ce_namelen(ce), ce_stage(ce));
475
		if (!cmp)
476
			return next;
477 478 479 480 481 482
		if (cmp < 0) {
			last = next;
			continue;
		}
		first = next+1;
	}
483
	return -first-1;
484 485
}

486 487 488 489 490
int index_name_pos(const struct index_state *istate, const char *name, int namelen)
{
	return index_name_stage_pos(istate, name, namelen, 0);
}

491
/* Remove entry, return true if there are more entries to go.. */
492
int remove_index_entry_at(struct index_state *istate, int pos)
493
{
494 495
	struct cache_entry *ce = istate->cache[pos];

496
	record_resolve_undo(istate, ce);
497
	remove_name_hash(istate, ce);
498
	save_or_free_index_entry(istate, ce);
499
	istate->cache_changed |= CE_ENTRY_REMOVED;
500 501
	istate->cache_nr--;
	if (pos >= istate->cache_nr)
502
		return 0;
503 504 505
	memmove(istate->cache + pos,
		istate->cache + pos + 1,
		(istate->cache_nr - pos) * sizeof(struct cache_entry *));
506 507 508
	return 1;
}

509
/*
Ondřej Bílka's avatar
Ondřej Bílka committed
510
 * Remove all cache entries marked for removal, that is where
511 512 513 514 515 516 517 518 519
 * 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++) {
520
		if (ce_array[i]->ce_flags & CE_REMOVE) {
521
			remove_name_hash(istate, ce_array[i]);
522
			save_or_free_index_entry(istate, ce_array[i]);
523
		}
524 525 526
		else
			ce_array[j++] = ce_array[i];
	}
527 528
	if (j == istate->cache_nr)
		return;
529
	istate->cache_changed |= CE_ENTRY_REMOVED;
530 531 532
	istate->cache_nr = j;
}

533
int remove_file_from_index(struct index_state *istate, const char *path)
534
{
535
	int pos = index_name_pos(istate, path, strlen(path));
536 537
	if (pos < 0)
		pos = -pos-1;
538
	cache_tree_invalidate_path(istate, path);
539 540
	while (pos < istate->cache_nr && !strcmp(istate->cache[pos]->name, path))
		remove_index_entry_at(istate, pos);
541 542 543
	return 0;
}

544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571
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;
}

572 573 574 575 576 577 578 579 580 581 582 583 584 585 586
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
 */
587 588 589
static struct cache_entry *create_alias_ce(struct index_state *istate,
					   struct cache_entry *ce,
					   struct cache_entry *alias)
590 591 592 593 594 595 596 597 598 599 600 601
{
	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);
602
	save_or_free_index_entry(istate, ce);
603 604 605
	return new;
}

606
void set_object_name_for_intent_to_add_entry(struct cache_entry *ce)
607 608 609 610 611 612 613
{
	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);
}

614
int add_to_index(struct index_state *istate, const char *path, struct stat *st, int flags)
Johannes Schindelin's avatar
Johannes Schindelin committed
615
{
616
	int size, namelen, was_same;
617
	mode_t st_mode = st->st_mode;
618
	struct cache_entry *ce, *alias;
619
	unsigned ce_option = CE_MATCH_IGNORE_VALID|CE_MATCH_IGNORE_SKIP_WORKTREE|CE_MATCH_RACY_IS_DIRTY;
620 621
	int verbose = flags & (ADD_CACHE_VERBOSE | ADD_CACHE_PRETEND);
	int pretend = flags & ADD_CACHE_PRETEND;
622 623 624
	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
625

626
	if (!S_ISREG(st_mode) && !S_ISLNK(st_mode) && !S_ISDIR(st_mode))
627
		return error("%s: can only add regular files, symbolic links or git-directories", path);
Johannes Schindelin's avatar
Johannes Schindelin committed
628 629

	namelen = strlen(path);
630
	if (S_ISDIR(st_mode)) {
631 632 633
		while (namelen && path[namelen-1] == '/')
			namelen--;
	}
Johannes Schindelin's avatar
Johannes Schindelin committed
634 635 636
	size = cache_entry_size(namelen);
	ce = xcalloc(1, size);
	memcpy(ce->name, path, namelen);
637
	ce->ce_namelen = namelen;
638 639
	if (!intent_only)
		fill_stat_cache_info(ce, st);
640 641
	else
		ce->ce_flags |= CE_INTENT_TO_ADD;
Johannes Schindelin's avatar
Johannes Schindelin committed
642

643
	if (trust_executable_bit && has_symlinks)
644
		ce->ce_mode = create_ce_mode(st_mode);
645
	else {
646 647
		/* 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
648
		 */
649
		struct cache_entry *ent;
650
		int pos = index_name_pos_also_unmerged(istate, path, namelen);
651

652
		ent = (0 <= pos) ? istate->cache[pos] : NULL;
653
		ce->ce_mode = ce_mode_from_stat(ent, st_mode);
Johannes Schindelin's avatar
Johannes Schindelin committed
654 655
	}

656 657 658 659 660 661 662 663 664 665 666 667 668 669
	/* 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;
670
				foundce = index_dir_exists(istate, ce->name, ptr - ce->name - 1);
671 672 673 674 675 676 677 678
				if (foundce) {
					memcpy((void *)startPtr, foundce->name + (startPtr - ce->name), ptr - startPtr);
					startPtr = ptr;
				}
			}
		}
	}

679
	alias = index_file_exists(istate, ce->name, ce_namelen(ce), ignore_case);
680
	if (alias && !ce_stage(alias) && !ie_match_stat(istate, alias, st, ce_option)) {
681 682
		/* Nothing changed, really */
		free(ce);
683 684
		if (!S_ISGITLINK(alias->ce_mode))
			ce_mark_uptodate(alias);
685
		alias->ce_flags |= CE_ADDED;
686 687
		return 0;
	}
688
	if (!intent_only) {
689
		if (index_path(ce->sha1, path, st, HASH_WRITE_OBJECT))
690 691
			return error("unable to index file %s", path);
	} else
692
		set_object_name_for_intent_to_add_entry(ce);
693

694
	if (ignore_case && alias && different_name(ce, alias))
695
		ce = create_alias_ce(istate, ce, alias);
696
	ce->ce_flags |= CE_ADDED;
697

Junio C Hamano's avatar
Junio C Hamano committed
698
	/* It was suspected to be racily clean, but it turns out to be Ok */
699 700 701 702 703 704 705
	was_same = (alias &&
		    !ce_stage(alias) &&
		    !hashcmp(alias->sha1, ce->sha1) &&
		    ce->ce_mode == alias->ce_mode);

	if (pretend)
		;
706
	else if (add_index_entry(istate, ce, add_option))
707
		return error("unable to add %s to index",path);
708
	if (verbose && !was_same)
Johannes Schindelin's avatar
Johannes Schindelin committed
709 710 711 712
		printf("add '%s'\n", path);
	return 0;
}

713
int add_file_to_index(struct index_state *istate, const char *path, int flags)
714 715 716
{
	struct stat st;
	if (lstat(path, &st))
717
		die_errno("unable to stat '%s'", path);
718
	return add_to_index(istate, path, &st, flags);
719 720
}

721 722
struct cache_entry *make_cache_entry(unsigned int mode,
		const unsigned char *sha1, const char *path, int stage,
723
		unsigned int refresh_options)
724 725 726 727
{
	int size, len;
	struct cache_entry *ce;

728 729
	if (!verify_path(path)) {
		error("Invalid path '%s'", path);
730
		return NULL;
731
	}
732 733 734 735 736 737 738

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

	hashcpy(ce->sha1, sha1);
	memcpy(ce->name, path, len);
739 740
	ce->ce_flags = create_ce_flags(stage);
	ce->ce_namelen = len;
741 742
	ce->ce_mode = create_ce_mode(mode);

743
	return refresh_cache_entry(ce, refresh_options);
744 745
}

746
int ce_same_name(const struct cache_entry *a, const struct cache_entry *b)
747 748 749 750 751
{
	int len = ce_namelen(a);
	return ce_namelen(b) == len && !memcmp(a->name, b->name, len);
}

752 753 754 755 756 757 758 759 760 761 762 763 764 765 766
/*
 * 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.
	 */
767

768
	/* "." is not allowed */
769
	if (*rest == '\0' || is_dir_sep(*rest))
770 771
		return 0;

772
	switch (*rest) {
773 774 775 776 777 778 779 780 781 782 783 784
	/*
	 * ".git" followed by  NUL or slash is bad. This
	 * shares the path end test with the ".." case.
	 */
	case 'g':
		if (rest[1] != 'i')
			break;
		if (rest[2] != 't')
			break;
		rest += 2;
	/* fallthrough */
	case '.':
785
		if (rest[1] == '\0' || is_dir_sep(rest[1]))
786 787 788 789 790 791 792 793 794
			return 0;
	}
	return 1;
}

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

795 796 797
	if (has_dos_drive_prefix(path))
		return 0;

798 799 800 801
	goto inside;
	for (;;) {
		if (!c)
			return 1;
802
		if (is_dir_sep(c)) {
803 804
inside:
			c = *path++;
805 806 807
			if ((c == '.' && !verify_dotfile(path)) ||
			    is_dir_sep(c) || c == '\0')
				return 0;
808 809 810 811 812
		}
		c = *path++;
	}
}

813 814 815
/*
 * Do we have another file that has the beginning components being a
 * proper superset of the name we're trying to add?
816
 */
817 818
static int has_file_name(struct index_state *istate,
			 const struct cache_entry *ce, int pos, int ok_to_replace)
819
{
820 821
	int retval = 0;
	int len = ce_namelen(ce);
822
	int stage = ce_stage(ce);
823
	const char *name = ce->name;
824

825 826
	while (pos < istate->cache_nr) {
		struct cache_entry *p = istate->cache[pos++];
827

828
		if (len >= ce_namelen(p))
829
			break;
830 831
		if (memcmp(name, p->name, len))
			break;
832 833
		if (ce_stage(p) != stage)
			continue;
834 835
		if (p->name[len] != '/')
			continue;
836
		if (p->ce_flags & CE_REMOVE)
837
			continue;
838 839 840
		retval = -1;
		if (!ok_to_replace)
			break;
841
		remove_index_entry_at(istate, --pos);
842
	}
843 844
	return retval;
}
845

846 847 848 849
/*
 * Do we have another file with a pathname that is a proper
 * subset of the name we're trying to add?
 */
850 851
static int has_dir_name(struct index_state *istate,
			const struct cache_entry *ce, int pos, int ok_to_replace)
852 853
{
	int retval = 0;
854
	int stage = ce_stage(ce);
855 856
	const char *name = ce->name;
	const char *slash = name + ce_namelen(ce);
857

858 859
	for (;;) {
		int len;
860

861 862 863 864 865 866 867
		for (;;) {
			if (*--slash == '/')
				break;
			if (slash <= ce->name)
				return retval;
		}
		len = slash - name;
868

869
		pos = index_name_stage_pos(istate, name, len, stage);
870
		if (pos >= 0) {
871 872 873 874 875 876 877 878
			/*
			 * 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.
			 */
879
			if (!(istate->cache[pos]->ce_flags & CE_REMOVE)) {
880 881 882
				retval = -1;
				if (!ok_to_replace)
					break;
883
				remove_index_entry_at(istate, pos);
884 885
				continue;
			}
886
		}
887 888
		else
			pos = -pos-1;
889 890 891 892

		/*
		 * Trivial optimization: if we find an entry that
		 * already matches the sub-directory, then we know
893
		 * we're ok, and we can exit.
894
		 */
895 896
		while (pos < istate->cache_nr) {
			struct cache_entry *p = istate->cache[pos];
897 898 899 900
			if ((ce_namelen(p) <= len) ||
			    (p->name[len] != '/') ||
			    memcmp(p->name, name, len))
				break; /* not our subdirectory */
901 902 903
			if (ce_stage(p) == stage && !(p->ce_flags & CE_REMOVE))
				/*
				 * p is at the same stage as our entry, and
904 905 906 907 908 909
				 * is a subdirectory of what we are looking
				 * at, so we cannot have conflicts at our
				 * level or anything shorter.
				 */
				return retval;
			pos++;
910
		}
911
	}
912 913 914 915 916 917 918
	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
919
 *
920 921 922 923
 * 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.
 */
924 925 926
static int check_file_directory_conflict(struct index_state *istate,
					 const struct cache_entry *ce,
					 int pos, int ok_to_replace)
927
{
928 929 930 931 932
	int retval;

	/*
	 * When ce is an "I am going away" entry, we allow it to be added
	 */
933
	if (ce->ce_flags & CE_REMOVE)
934 935
		return 0;

936 937 938
	/*
	 * We check if the path is a sub-path of a subsequent pathname
	 * first, since removing those will not change the position
939
	 * in the array.
940
	 */
941
	retval = has_file_name(istate, ce, pos, ok_to_replace);
942

943 944 945 946
	/*
	 * Then check if the path might have a clashing sub-directory
	 * before it.
	 */
947
	return retval + has_dir_name(istate, ce, pos, ok_to_replace);
948 949
}

950
static int add_index_entry_with_check(struct index_state *istate, struct cache_entry *ce, int option)
951 952
{
	int pos;
953 954
	int ok_to_add = option & ADD_CACHE_OK_TO_ADD;
	int ok_to_replace = option & ADD_CACHE_OK_TO_REPLACE;
955
	int skip_df_check = option & ADD_CACHE_SKIP_DFCHECK;
956
	int new_only = option & ADD_CACHE_NEW_ONLY;
Junio C Hamano's avatar
Junio C Hamano committed
957

958 959
	if (!(option & ADD_CACHE_KEEP_CACHE_TREE))
		cache_tree_invalidate_path(istate, ce->name);
960
	pos = index_name_stage_pos(istate, ce->name, ce_namelen(ce), ce_stage(ce));
961

Junio C Hamano's avatar
Junio C Hamano committed
962
	/* existing match? Just replace it. */
963
	if (pos >= 0) {
964 965
		if (!new_only)
			replace_index_entry(istate, pos, ce);
966 967
		return 0;
	}
968
	pos = -pos-1;
969

970 971 972 973
	/*
	 * Inserting a merged entry ("stage 0") into the index
	 * will always replace all non-merged entries..
	 */
974 975
	if (pos < istate->cache_nr && ce_stage(ce) == 0) {
		while (ce_same_name(istate->cache[pos], ce)) {
976
			ok_to_add = 1;
977
			if (!remove_index_entry_at(istate, pos))
978 979 980 981
				break;
		}
	}

982 983
	if (!ok_to_add)
		return -1;
984
	if (!verify_path(ce->name))
985
		return error("Invalid path '%s'", ce->name);
986

Junio C Hamano's avatar
Junio C Hamano committed
987
	if (!skip_df_check &&
988
	    check_file_directory_conflict(istate, ce, pos, ok_to_replace)) {
989
		if (!ok_to_replace)
990 991
			return error("'%s' appears as both a file and as a directory",
				     ce->name);
992
		pos = index_name_stage_pos(istate, ce->name, ce_namelen(ce), ce_stage(ce));
993 994
		pos = -pos-1;
	}
995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010
	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;
	}
1011

1012
	/* Make sure the array is big enough .. */
1013
	ALLOC_GROW(istate->cache, istate->cache_nr + 1, istate->cache_alloc);
1014 1015

	/* Add it in.. */
1016
	istate->cache_nr++;
1017
	if (istate->cache_nr > pos + 1)
1018 1019 1020
		memmove(istate->cache + pos + 1,
			istate->cache + pos,
			(istate->cache_nr - pos - 1) * sizeof(ce));
1021
	set_index_entry(istate, pos, ce);
1022
	istate->cache_changed |= CE_ENTRY_ADDED;
1023 1024 1025
	return 0;
}

1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036
/*
 * "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.
 */
1037
static struct cache_entry *refresh_cache_ent(struct index_state *istate,
1038
					     struct cache_entry *ce,
1039 1040
					     unsigned int options, int *err,
					     int *changed_ret)
1041 1042 1043 1044
{
	struct stat st;
	struct cache_entry *updated;
	int changed, size;
1045
	int refresh = options & CE_MATCH_REFRESH;
1046
	int ignore_valid = options & CE_MATCH_IGNORE_VALID;
1047
	int ignore_skip_worktree = options & CE_MATCH_IGNORE_SKIP_WORKTREE;
1048
	int ignore_missing = options & CE_MATCH_IGNORE_MISSING;
1049

1050
	if (!refresh || ce_uptodate(ce))
1051 1052
		return ce;

1053
	/*
1054 1055 1056
	 * CE_VALID or CE_SKIP_WORKTREE means the user promised us
	 * that the change to the work tree does not matter and told
	 * us not to worry.
1057
	 */
1058 1059 1060 1061
	if (!ignore_skip_worktree && ce_skip_worktree(ce)) {
		ce_mark_uptodate(ce);
		return ce;
	}
1062 1063 1064 1065 1066
	if (!ignore_valid && (ce->ce_flags & CE_VALID)) {
		ce_mark_uptodate(ce);
		return ce;
	}

1067 1068 1069 1070 1071 1072 1073 1074
	if (has_symlink_leading_path(ce->name, ce_namelen(ce))) {
		if (ignore_missing)
			return ce;
		if (err)
			*err = ENOENT;
		return NULL;
	}

1075
	if (lstat(ce->name, &st) < 0) {
1076 1077
		if (ignore_missing && errno == ENOENT)
			return ce;
1078 1079
		if (err)
			*err = errno;
1080 1081
		return NULL;
	}
1082

1083
	changed = ie_match_stat(istate, ce, &st, options);
1084 1085
	if (changed_ret)
		*changed_ret = changed;
1086
	if (!changed) {
1087 1088 1089 1090 1091 1092 1093 1094
		/*
		 * The path is unchanged.  If we were told to ignore
		 * valid bit, then we did the actual stat check and
		 * found that the entry is unmodified.  If the entry
		 * is not marked VALID, this is the place to mark it
		 * valid again, under "assume unchanged" mode.
		 */
		if (ignore_valid && assume_unchanged &&
1095
		    !(ce->ce_flags & CE_VALID))
1096
			; /* mark this one VALID again */
1097 1098 1099 1100 1101 1102
		else {
			/*
			 * We do not mark the index itself "modified"
			 * because CE_UPTODATE flag is in-core only;
			 * we are not going to write this change out.
			 */
1103 1104
			if (!S_ISGITLINK(ce->ce_mode))
				ce_mark_uptodate(ce);