read-cache.c 33.8 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 13 14 15 16 17 18 19 20 21 22

/* 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" */
23

24
struct index_state the_index;
25

26 27 28 29 30 31 32 33 34 35 36
static unsigned int hash_name(const char *name, int namelen)
{
	unsigned int hash = 0x123;

	do {
		unsigned char c = *name++;
		hash = hash*101 + c;
	} while (--namelen);
	return hash;
}

Junio C Hamano's avatar
Junio C Hamano committed
37
static void hash_index_entry(struct index_state *istate, struct cache_entry *ce)
38 39 40 41 42 43 44 45 46 47 48
{
	void **pos;
	unsigned int hash = hash_name(ce->name, ce_namelen(ce));

	pos = insert_hash(hash, ce, &istate->name_hash);
	if (pos) {
		ce->next = *pos;
		*pos = ce;
	}
}

Junio C Hamano's avatar
Junio C Hamano committed
49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66
static void lazy_init_name_hash(struct index_state *istate)
{
	int nr;

	if (istate->name_hash_initialized)
		return;
	for (nr = 0; nr < istate->cache_nr; nr++)
		hash_index_entry(istate, istate->cache[nr]);
	istate->name_hash_initialized = 1;
}

static void set_index_entry(struct index_state *istate, int nr, struct cache_entry *ce)
{
	istate->cache[nr] = ce;
	if (istate->name_hash_initialized)
		hash_index_entry(istate, ce);
}

67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94
/*
 * We don't actually *remove* it, we can just mark it invalid so that
 * we won't find it in lookups.
 *
 * Not only would we have to search the lists (simple enough), but
 * we'd also have to rehash other hash buckets in case this makes the
 * hash bucket empty (common). So it's much better to just mark
 * it.
 */
static void remove_hash_entry(struct index_state *istate, struct cache_entry *ce)
{
	ce->ce_flags |= CE_UNHASHED;
}

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

	if (ce != old) {
		remove_hash_entry(istate, old);
		set_index_entry(istate, nr, ce);
	}
	istate->cache_changed = 1;
}

int index_name_exists(struct index_state *istate, const char *name, int namelen)
{
	unsigned int hash = hash_name(name, namelen);
Junio C Hamano's avatar
Junio C Hamano committed
95 96 97 98
	struct cache_entry *ce;

	lazy_init_name_hash(istate);
	ce = lookup_hash(hash, &istate->name_hash);
99 100 101 102 103 104 105 106 107 108 109

	while (ce) {
		if (!(ce->ce_flags & CE_UNHASHED)) {
			if (!cache_name_compare(name, namelen, ce->name, ce->ce_flags))
				return 1;
		}
		ce = ce->next;
	}
	return 0;
}

110 111 112 113 114 115 116
/*
 * 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)
{
117 118 119 120 121 122 123
	ce->ce_ctime = st->st_ctime;
	ce->ce_mtime = st->st_mtime;
	ce->ce_dev = st->st_dev;
	ce->ce_ino = st->st_ino;
	ce->ce_uid = st->st_uid;
	ce->ce_gid = st->st_gid;
	ce->ce_size = st->st_size;
Junio C Hamano's avatar
Junio C Hamano committed
124 125

	if (assume_unchanged)
126
		ce->ce_flags |= CE_VALID;
127 128 129

	if (S_ISREG(st->st_mode))
		ce_mark_uptodate(ce);
130 131
}

Junio C Hamano's avatar
Junio C Hamano committed
132 133 134 135 136 137 138
static int ce_compare_data(struct cache_entry *ce, struct stat *st)
{
	int match = -1;
	int fd = open(ce->name, O_RDONLY);

	if (fd >= 0) {
		unsigned char sha1[20];
139
		if (!index_fd(sha1, fd, st, 0, OBJ_BLOB, ce->name))
140
			match = hashcmp(sha1, ce->sha1);
141
		/* index_fd() closed the file descriptor already */
Junio C Hamano's avatar
Junio C Hamano committed
142 143 144 145
	}
	return match;
}

146
static int ce_compare_link(struct cache_entry *ce, size_t expected_size)
Junio C Hamano's avatar
Junio C Hamano committed
147 148 149 150 151
{
	int match = -1;
	char *target;
	void *buffer;
	unsigned long size;
152
	enum object_type type;
Junio C Hamano's avatar
Junio C Hamano committed
153 154 155 156 157 158 159 160
	int len;

	target = xmalloc(expected_size);
	len = readlink(ce->name, target, expected_size);
	if (len != expected_size) {
		free(target);
		return -1;
	}
161
	buffer = read_sha1_file(ce->sha1, &type, &size);
Junio C Hamano's avatar
Junio C Hamano committed
162 163 164 165 166 167 168 169 170 171 172
	if (!buffer) {
		free(target);
		return -1;
	}
	if (size == expected_size)
		match = memcmp(buffer, target, size);
	free(buffer);
	free(target);
	return match;
}

173 174 175 176 177 178
static int ce_compare_gitlink(struct cache_entry *ce)
{
	unsigned char sha1[20];

	/*
	 * We don't actually require that the .git directory
Martin Waitz's avatar
Martin Waitz committed
179
	 * under GITLINK directory be a valid git directory. It
180 181 182 183 184 185 186 187 188 189
	 * 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);
}

Junio C Hamano's avatar
Junio C Hamano committed
190 191 192 193 194 195 196 197
static int ce_modified_check_fs(struct cache_entry *ce, struct stat *st)
{
	switch (st->st_mode & S_IFMT) {
	case S_IFREG:
		if (ce_compare_data(ce, st))
			return DATA_CHANGED;
		break;
	case S_IFLNK:
198
		if (ce_compare_link(ce, xsize_t(st->st_size)))
Junio C Hamano's avatar
Junio C Hamano committed
199 200
			return DATA_CHANGED;
		break;
201
	case S_IFDIR:
202
		if (S_ISGITLINK(ce->ce_mode))
203
			return 0;
Junio C Hamano's avatar
Junio C Hamano committed
204 205 206 207 208 209
	default:
		return TYPE_CHANGED;
	}
	return 0;
}

Junio C Hamano's avatar
Junio C Hamano committed
210
static int ce_match_stat_basic(struct cache_entry *ce, struct stat *st)
211 212 213
{
	unsigned int changed = 0;

214 215 216 217
	if (ce->ce_flags & CE_REMOVE)
		return MODE_CHANGED | DATA_CHANGED | TYPE_CHANGED;

	switch (ce->ce_mode & S_IFMT) {
218 219
	case S_IFREG:
		changed |= !S_ISREG(st->st_mode) ? TYPE_CHANGED : 0;
Junio C Hamano's avatar
Junio C Hamano committed
220 221 222 223
		/* We consider only the owner x bit to be relevant for
		 * "mode changes"
		 */
		if (trust_executable_bit &&
224
		    (0100 & (ce->ce_mode ^ st->st_mode)))
225
			changed |= MODE_CHANGED;
226 227
		break;
	case S_IFLNK:
228 229 230
		if (!S_ISLNK(st->st_mode) &&
		    (has_symlinks || !S_ISREG(st->st_mode)))
			changed |= TYPE_CHANGED;
231
		break;
Martin Waitz's avatar
Martin Waitz committed
232
	case S_IFGITLINK:
233 234 235 236
		if (!S_ISDIR(st->st_mode))
			changed |= TYPE_CHANGED;
		else if (ce_compare_gitlink(ce))
			changed |= DATA_CHANGED;
237
		return changed;
238
	default:
239
		die("internal error: ce_mode is %o", ce->ce_mode);
240
	}
241
	if (ce->ce_mtime != (unsigned int) st->st_mtime)
242
		changed |= MTIME_CHANGED;
243
	if (ce->ce_ctime != (unsigned int) st->st_ctime)
244
		changed |= CTIME_CHANGED;
245

246 247
	if (ce->ce_uid != (unsigned int) st->st_uid ||
	    ce->ce_gid != (unsigned int) st->st_gid)
248
		changed |= OWNER_CHANGED;
249
	if (ce->ce_ino != (unsigned int) st->st_ino)
250
		changed |= INODE_CHANGED;
251 252 253 254 255 256 257

#ifdef USE_STDEV
	/*
	 * st_dev breaks on network filesystems where different
	 * clients will have different views of what "device"
	 * the filesystem is on
	 */
258
	if (ce->ce_dev != (unsigned int) st->st_dev)
259 260 261
		changed |= INODE_CHANGED;
#endif

262
	if (ce->ce_size != (unsigned int) st->st_size)
263
		changed |= DATA_CHANGED;
264

Junio C Hamano's avatar
Junio C Hamano committed
265 266 267
	return changed;
}

268 269 270 271 272 273
static int is_racy_timestamp(struct index_state *istate, struct cache_entry *ce)
{
	return (istate->timestamp &&
		((unsigned int)istate->timestamp) <= ce->ce_mtime);
}

274
int ie_match_stat(struct index_state *istate,
275 276
		  struct cache_entry *ce, struct stat *st,
		  unsigned int options)
Junio C Hamano's avatar
Junio C Hamano committed
277
{
Junio C Hamano's avatar
Junio C Hamano committed
278
	unsigned int changed;
279 280
	int ignore_valid = options & CE_MATCH_IGNORE_VALID;
	int assume_racy_is_modified = options & CE_MATCH_RACY_IS_DIRTY;
Junio C Hamano's avatar
Junio C Hamano committed
281 282 283 284 285

	/*
	 * If it's marked as always valid in the index, it's
	 * valid whatever the checked-out copy says.
	 */
286
	if (!ignore_valid && (ce->ce_flags & CE_VALID))
Junio C Hamano's avatar
Junio C Hamano committed
287 288 289
		return 0;

	changed = ce_match_stat_basic(ce, st);
Junio C Hamano's avatar
Junio C Hamano committed
290

Junio C Hamano's avatar
Junio C Hamano committed
291 292 293 294 295 296 297 298 299 300 301 302 303 304
	/*
	 * 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
305
	 * carefully than others.
Junio C Hamano's avatar
Junio C Hamano committed
306
	 */
307
	if (!changed && is_racy_timestamp(istate, ce)) {
308 309 310 311 312
		if (assume_racy_is_modified)
			changed |= DATA_CHANGED;
		else
			changed |= ce_modified_check_fs(ce, st);
	}
313

Junio C Hamano's avatar
Junio C Hamano committed
314
	return changed;
315 316
}

317
int ie_modified(struct index_state *istate,
318
		struct cache_entry *ce, struct stat *st, unsigned int options)
319
{
Junio C Hamano's avatar
Junio C Hamano committed
320
	int changed, changed_fs;
321 322

	changed = ie_match_stat(istate, ce, st, options);
323 324 325 326 327 328 329 330 331 332 333 334 335
	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;

	/* Immediately after read-tree or update-index --cacheinfo,
	 * the length field is zero.  For other cases the ce_size
	 * should match the SHA1 recorded in the index entry.
	 */
336
	if ((changed & DATA_CHANGED) && ce->ce_size != 0)
337 338
		return changed;

Junio C Hamano's avatar
Junio C Hamano committed
339 340 341
	changed_fs = ce_modified_check_fs(ce, st);
	if (changed_fs)
		return changed | changed_fs;
342 343 344
	return 0;
}

345 346 347 348 349 350 351 352 353 354 355 356
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];
357
	if (!c1 && S_ISDIR(mode1))
358
		c1 = '/';
359
	if (!c2 && S_ISDIR(mode2))
360 361 362 363
		c2 = '/';
	return (c1 < c2) ? -1 : (c1 > c2) ? 1 : 0;
}

364
int cache_name_compare(const char *name1, int flags1, const char *name2, int flags2)
365
{
366 367
	int len1 = flags1 & CE_NAMEMASK;
	int len2 = flags2 & CE_NAMEMASK;
368 369 370 371 372 373 374 375 376 377
	int len = len1 < len2 ? len1 : len2;
	int cmp;

	cmp = memcmp(name1, name2, len);
	if (cmp)
		return cmp;
	if (len1 < len2)
		return -1;
	if (len1 > len2)
		return 1;
Junio C Hamano's avatar
Junio C Hamano committed
378

379 380 381
	/* Compare stages  */
	flags1 &= CE_STAGEMASK;
	flags2 &= CE_STAGEMASK;
Junio C Hamano's avatar
Junio C Hamano committed
382

383 384 385 386
	if (flags1 < flags2)
		return -1;
	if (flags1 > flags2)
		return 1;
387 388 389
	return 0;
}

390
int index_name_pos(struct index_state *istate, const char *name, int namelen)
391 392 393 394
{
	int first, last;

	first = 0;
395
	last = istate->cache_nr;
396 397
	while (last > first) {
		int next = (last + first) >> 1;
398
		struct cache_entry *ce = istate->cache[next];
399
		int cmp = cache_name_compare(name, namelen, ce->name, ce->ce_flags);
400
		if (!cmp)
401
			return next;
402 403 404 405 406 407
		if (cmp < 0) {
			last = next;
			continue;
		}
		first = next+1;
	}
408
	return -first-1;
409 410
}

411
/* Remove entry, return true if there are more entries to go.. */
412
int remove_index_entry_at(struct index_state *istate, int pos)
413
{
414 415 416
	struct cache_entry *ce = istate->cache[pos];

	remove_hash_entry(istate, ce);
417 418 419
	istate->cache_changed = 1;
	istate->cache_nr--;
	if (pos >= istate->cache_nr)
420
		return 0;
421 422 423
	memmove(istate->cache + pos,
		istate->cache + pos + 1,
		(istate->cache_nr - pos) * sizeof(struct cache_entry *));
424 425 426
	return 1;
}

427
int remove_file_from_index(struct index_state *istate, const char *path)
428
{
429
	int pos = index_name_pos(istate, path, strlen(path));
430 431
	if (pos < 0)
		pos = -pos-1;
Junio C Hamano's avatar
Junio C Hamano committed
432
	cache_tree_invalidate_path(istate->cache_tree, path);
433 434
	while (pos < istate->cache_nr && !strcmp(istate->cache[pos]->name, path))
		remove_index_entry_at(istate, pos);
435 436 437
	return 0;
}

438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465
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;
}

466
int add_file_to_index(struct index_state *istate, const char *path, int verbose)
Johannes Schindelin's avatar
Johannes Schindelin committed
467
{
468
	int size, namelen, pos;
Johannes Schindelin's avatar
Johannes Schindelin committed
469 470
	struct stat st;
	struct cache_entry *ce;
471
	unsigned ce_option = CE_MATCH_IGNORE_VALID|CE_MATCH_RACY_IS_DIRTY;
Johannes Schindelin's avatar
Johannes Schindelin committed
472 473 474 475

	if (lstat(path, &st))
		die("%s: unable to stat (%s)", path, strerror(errno));

476 477
	if (!S_ISREG(st.st_mode) && !S_ISLNK(st.st_mode) && !S_ISDIR(st.st_mode))
		die("%s: can only add regular files, symbolic links or git-directories", path);
Johannes Schindelin's avatar
Johannes Schindelin committed
478 479

	namelen = strlen(path);
480 481 482 483
	if (S_ISDIR(st.st_mode)) {
		while (namelen && path[namelen-1] == '/')
			namelen--;
	}
Johannes Schindelin's avatar
Johannes Schindelin committed
484 485 486
	size = cache_entry_size(namelen);
	ce = xcalloc(1, size);
	memcpy(ce->name, path, namelen);
487
	ce->ce_flags = namelen;
Johannes Schindelin's avatar
Johannes Schindelin committed
488 489
	fill_stat_cache_info(ce, &st);

490
	if (trust_executable_bit && has_symlinks)
491 492
		ce->ce_mode = create_ce_mode(st.st_mode);
	else {
493 494
		/* 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
495
		 */
496
		struct cache_entry *ent;
497
		int pos = index_name_pos_also_unmerged(istate, path, namelen);
498

499
		ent = (0 <= pos) ? istate->cache[pos] : NULL;
500
		ce->ce_mode = ce_mode_from_stat(ent, st.st_mode);
Johannes Schindelin's avatar
Johannes Schindelin committed
501 502
	}

503 504 505
	pos = index_name_pos(istate, ce->name, namelen);
	if (0 <= pos &&
	    !ce_stage(istate->cache[pos]) &&
506
	    !ie_match_stat(istate, istate->cache[pos], &st, ce_option)) {
507 508
		/* Nothing changed, really */
		free(ce);
509
		ce_mark_uptodate(istate->cache[pos]);
510 511 512
		return 0;
	}

Johannes Schindelin's avatar
Johannes Schindelin committed
513 514
	if (index_path(ce->sha1, path, &st, 1))
		die("unable to index file %s", path);
515
	if (add_index_entry(istate, ce, ADD_CACHE_OK_TO_ADD|ADD_CACHE_OK_TO_REPLACE))
Johannes Schindelin's avatar
Johannes Schindelin committed
516 517 518 519 520 521
		die("unable to add %s to index",path);
	if (verbose)
		printf("add '%s'\n", path);
	return 0;
}

522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546
struct cache_entry *make_cache_entry(unsigned int mode,
		const unsigned char *sha1, const char *path, int stage,
		int refresh)
{
	int size, len;
	struct cache_entry *ce;

	if (!verify_path(path))
		return NULL;

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

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

	if (refresh)
		return refresh_cache_entry(ce, 0);

	return ce;
}

547
int ce_same_name(struct cache_entry *a, struct cache_entry *b)
548 549 550 551 552
{
	int len = ce_namelen(a);
	return ce_namelen(b) == len && !memcmp(a->name, b->name, len);
}

553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572
int ce_path_match(const struct cache_entry *ce, const char **pathspec)
{
	const char *match, *name;
	int len;

	if (!pathspec)
		return 1;

	len = ce_namelen(ce);
	name = ce->name;
	while ((match = *pathspec++) != NULL) {
		int matchlen = strlen(match);
		if (matchlen > len)
			continue;
		if (memcmp(name, match, matchlen))
			continue;
		if (matchlen && name[matchlen-1] == '/')
			return 1;
		if (name[matchlen] == '/' || !name[matchlen])
			return 1;
573 574
		if (!matchlen)
			return 1;
575 576 577 578
	}
	return 0;
}

579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642
/*
 * 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.
	 */
	switch (*rest) {
	/* "." is not allowed */
	case '\0': case '/':
		return 0;

	/*
	 * ".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 '.':
		if (rest[1] == '\0' || rest[1] == '/')
			return 0;
	}
	return 1;
}

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

	goto inside;
	for (;;) {
		if (!c)
			return 1;
		if (c == '/') {
inside:
			c = *path++;
			switch (c) {
			default:
				continue;
			case '/': case '\0':
				break;
			case '.':
				if (verify_dotfile(path))
					continue;
			}
			return 0;
		}
		c = *path++;
	}
}

643 644 645
/*
 * Do we have another file that has the beginning components being a
 * proper superset of the name we're trying to add?
646
 */
647 648
static int has_file_name(struct index_state *istate,
			 const struct cache_entry *ce, int pos, int ok_to_replace)
649
{
650 651
	int retval = 0;
	int len = ce_namelen(ce);
652
	int stage = ce_stage(ce);
653
	const char *name = ce->name;
654

655 656
	while (pos < istate->cache_nr) {
		struct cache_entry *p = istate->cache[pos++];
657

658
		if (len >= ce_namelen(p))
659
			break;
660 661
		if (memcmp(name, p->name, len))
			break;
662 663
		if (ce_stage(p) != stage)
			continue;
664 665
		if (p->name[len] != '/')
			continue;
666
		if (p->ce_flags & CE_REMOVE)
667
			continue;
668 669 670
		retval = -1;
		if (!ok_to_replace)
			break;
671
		remove_index_entry_at(istate, --pos);
672
	}
673 674
	return retval;
}
675

676 677 678 679
/*
 * Do we have another file with a pathname that is a proper
 * subset of the name we're trying to add?
 */
680 681
static int has_dir_name(struct index_state *istate,
			const struct cache_entry *ce, int pos, int ok_to_replace)
682 683
{
	int retval = 0;
684
	int stage = ce_stage(ce);
685 686
	const char *name = ce->name;
	const char *slash = name + ce_namelen(ce);
687

688 689
	for (;;) {
		int len;
690

691 692 693 694 695 696 697
		for (;;) {
			if (*--slash == '/')
				break;
			if (slash <= ce->name)
				return retval;
		}
		len = slash - name;
698

699
		pos = index_name_pos(istate, name, create_ce_flags(len, stage));
700
		if (pos >= 0) {
701 702 703 704 705 706 707 708
			/*
			 * 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.
			 */
709
			if (!(istate->cache[pos]->ce_flags & CE_REMOVE)) {
710 711 712
				retval = -1;
				if (!ok_to_replace)
					break;
713
				remove_index_entry_at(istate, pos);
714 715
				continue;
			}
716
		}
717 718
		else
			pos = -pos-1;
719 720 721 722

		/*
		 * Trivial optimization: if we find an entry that
		 * already matches the sub-directory, then we know
723
		 * we're ok, and we can exit.
724
		 */
725 726
		while (pos < istate->cache_nr) {
			struct cache_entry *p = istate->cache[pos];
727 728 729 730
			if ((ce_namelen(p) <= len) ||
			    (p->name[len] != '/') ||
			    memcmp(p->name, name, len))
				break; /* not our subdirectory */
731 732 733
			if (ce_stage(p) == stage && !(p->ce_flags & CE_REMOVE))
				/*
				 * p is at the same stage as our entry, and
734 735 736 737 738 739
				 * is a subdirectory of what we are looking
				 * at, so we cannot have conflicts at our
				 * level or anything shorter.
				 */
				return retval;
			pos++;
740
		}
741
	}
742 743 744 745 746 747 748
	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
749
 *
750 751 752 753
 * 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.
 */
754 755 756
static int check_file_directory_conflict(struct index_state *istate,
					 const struct cache_entry *ce,
					 int pos, int ok_to_replace)
757
{
758 759 760 761 762
	int retval;

	/*
	 * When ce is an "I am going away" entry, we allow it to be added
	 */
763
	if (ce->ce_flags & CE_REMOVE)
764 765
		return 0;

766 767 768
	/*
	 * We check if the path is a sub-path of a subsequent pathname
	 * first, since removing those will not change the position
769
	 * in the array.
770
	 */
771
	retval = has_file_name(istate, ce, pos, ok_to_replace);
772

773 774 775 776
	/*
	 * Then check if the path might have a clashing sub-directory
	 * before it.
	 */
777
	return retval + has_dir_name(istate, ce, pos, ok_to_replace);
778 779
}

780
static int add_index_entry_with_check(struct index_state *istate, struct cache_entry *ce, int option)
781 782
{
	int pos;
783 784
	int ok_to_add = option & ADD_CACHE_OK_TO_ADD;
	int ok_to_replace = option & ADD_CACHE_OK_TO_REPLACE;
785
	int skip_df_check = option & ADD_CACHE_SKIP_DFCHECK;
Junio C Hamano's avatar
Junio C Hamano committed
786

Junio C Hamano's avatar
Junio C Hamano committed
787
	cache_tree_invalidate_path(istate->cache_tree, ce->name);
788
	pos = index_name_pos(istate, ce->name, ce->ce_flags);
789

Junio C Hamano's avatar
Junio C Hamano committed
790
	/* existing match? Just replace it. */
791
	if (pos >= 0) {
792
		replace_index_entry(istate, pos, ce);
793 794
		return 0;
	}
795
	pos = -pos-1;
796

797 798 799 800
	/*
	 * Inserting a merged entry ("stage 0") into the index
	 * will always replace all non-merged entries..
	 */
801 802
	if (pos < istate->cache_nr && ce_stage(ce) == 0) {
		while (ce_same_name(istate->cache[pos], ce)) {
803
			ok_to_add = 1;
804
			if (!remove_index_entry_at(istate, pos))
805 806 807 808
				break;
		}
	}

809 810
	if (!ok_to_add)
		return -1;
811 812
	if (!verify_path(ce->name))
		return -1;
813

Junio C Hamano's avatar
Junio C Hamano committed
814
	if (!skip_df_check &&
815
	    check_file_directory_conflict(istate, ce, pos, ok_to_replace)) {
816
		if (!ok_to_replace)
817 818
			return error("'%s' appears as both a file and as a directory",
				     ce->name);
819
		pos = index_name_pos(istate, ce->name, ce->ce_flags);
820 821
		pos = -pos-1;
	}
822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837
	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;
	}
838

839
	/* Make sure the array is big enough .. */
840 841 842 843
	if (istate->cache_nr == istate->cache_alloc) {
		istate->cache_alloc = alloc_nr(istate->cache_alloc);
		istate->cache = xrealloc(istate->cache,
					istate->cache_alloc * sizeof(struct cache_entry *));
844 845 846
	}

	/* Add it in.. */
847
	istate->cache_nr++;
848
	if (istate->cache_nr > pos + 1)
849 850 851
		memmove(istate->cache + pos + 1,
			istate->cache + pos,
			(istate->cache_nr - pos - 1) * sizeof(ce));
852
	set_index_entry(istate, pos, ce);
853
	istate->cache_changed = 1;
854 855 856
	return 0;
}

857 858 859 860 861 862 863 864 865 866 867
/*
 * "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.
 */
868
static struct cache_entry *refresh_cache_ent(struct index_state *istate,
869 870
					     struct cache_entry *ce,
					     unsigned int options, int *err)
871 872 873 874
{
	struct stat st;
	struct cache_entry *updated;
	int changed, size;
875
	int ignore_valid = options & CE_MATCH_IGNORE_VALID;
876

877 878 879
	if (ce_uptodate(ce))
		return ce;

880
	if (lstat(ce->name, &st) < 0) {
881 882
		if (err)
			*err = errno;
883 884
		return NULL;
	}
885

886
	changed = ie_match_stat(istate, ce, &st, options);
887
	if (!changed) {
888 889 890 891 892 893 894 895
		/*
		 * 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 &&
896
		    !(ce->ce_flags & CE_VALID))
897
			; /* mark this one VALID again */
898 899 900 901 902 903 904
		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.
			 */
			ce_mark_uptodate(ce);
905
			return ce;
906
		}
907 908
	}

909
	if (ie_modified(istate, ce, &st, options)) {
910 911
		if (err)
			*err = EINVAL;
912 913
		return NULL;
	}
914 915 916 917 918

	size = ce_size(ce);
	updated = xmalloc(size);
	memcpy(updated, ce, size);
	fill_stat_cache_info(updated, &st);
919 920 921 922 923
	/*
	 * If ignore_valid is not set, we should leave CE_VALID bit
	 * alone.  Otherwise, paths marked with --no-assume-unchanged
	 * (i.e. things to be edited) will reacquire CE_VALID bit
	 * automatically, which is not really what we want.
924
	 */
925
	if (!ignore_valid && assume_unchanged &&
926 927
	    !(ce->ce_flags & CE_VALID))
		updated->ce_flags &= ~CE_VALID;
928 929 930 931

	return updated;
}

932
int refresh_index(struct index_state *istate, unsigned int flags, const char **pathspec, char *seen)
933 934 935 936 937 938 939
{
	int i;
	int has_errors = 0;
	int really = (flags & REFRESH_REALLY) != 0;
	int allow_unmerged = (flags & REFRESH_UNMERGED) != 0;
	int quiet = (flags & REFRESH_QUIET) != 0;
	int not_new = (flags & REFRESH_IGNORE_MISSING) != 0;
940
	unsigned int options = really ? CE_MATCH_IGNORE_VALID : 0;
941

942
	for (i = 0; i < istate->cache_nr; i++) {
943
		struct cache_entry *ce, *new;
944 945
		int cache_errno = 0;

946
		ce = istate->cache[i];
947
		if (ce_stage(ce)) {
948 949
			while ((i < istate->cache_nr) &&
			       ! strcmp(istate->cache[i]->name, ce->name))
950 951 952 953 954 955 956 957 958
				i++;
			i--;
			if (allow_unmerged)
				continue;
			printf("%s: needs merge\n", ce->name);
			has_errors = 1;
			continue;
		}

959 960 961
		if (pathspec && !match_pathspec(pathspec, ce->name, strlen(ce->name), 0, seen))
			continue;

962
		new = refresh_cache_ent(istate, ce, options, &cache_errno);
963
		if (new == ce)
964
			continue;
965 966
		if (!new) {
			if (not_new && cache_errno == ENOENT)
967
				continue;
968
			if (really && cache_errno == EINVAL) {
969 970 971
				/* If we are doing --really-refresh that
				 * means the index is not valid anymore.
				 */
972
				ce->ce_flags &= ~CE_VALID;
973
				istate->cache_changed = 1;
974 975 976 977 978 979 980
			}
			if (quiet)
				continue;
			printf("%s: needs update\n", ce->name);
			has_errors = 1;
			continue;
		}
981 982

		replace_index_entry(istate, i, new);
983 984 985 986
	}
	return has_errors;
}

987 988
struct cache_entry *refresh_cache_entry(struct cache_entry *ce, int really)
{
989
	return refresh_cache_ent(&the_index, ce, really, NULL);
990 991
}

992
static int verify_hdr(struct cache_header *hdr, unsigned long size)
993 994
{
	SHA_CTX c;
995
	unsigned char sha1[20];
996

997
	if (hdr->hdr_signature != htonl(CACHE_SIGNATURE))
998
		return error("bad signature");
999 1000
	if (hdr->hdr_version != htonl(2))
		return error("bad index version");
1001
	SHA1_Init(&c);
1002
	SHA1_Update(&c, hdr, size - 20);
1003
	SHA1_Final(sha1, &c);
1004
	if (hashcmp(sha1, (unsigned char *)hdr + size - 20))
1005
		return error("bad index file sha1 signature");
1006 1007 1008
	return 0;
}

1009 1010
static int read_index_extension(struct index_state *istate,
				const char *ext, void *data, unsigned long sz)
1011 1012 1013
{
	switch (CACHE_EXT(ext)) {
	case CACHE_EXT_TREE:
1014
		istate->cache_tree = cache_tree_read(data, sz);
1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025
		break;
	default:
		if (*ext < 'A' || 'Z' < *ext)
			return error("index uses %.4s extension, which we do not understand",
				     ext);
		fprintf(stderr, "ignoring %.4s extension\n", ext);
		break;
	}
	return 0;
}

1026
int read_index(struct index_state *istate)
1027
{
1028
	return read_index_from(istate, get_index_file());
1029 1030
}

1031 1032
static void convert_from_disk(struct ondisk_cache_entry *ondisk, struct cache_entry *ce)
{
1033 1034
	size_t len;

1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045
	ce->ce_ctime = ntohl(ondisk->ctime.sec);
	ce->ce_mtime = ntohl(ondisk->mtime.sec);
	ce->ce_dev   = ntohl(ondisk->dev);
	ce->ce_ino   = ntohl(ondisk->ino);
	ce->ce_mode  = ntohl(ondisk->mode);
	ce->ce_uid   = ntohl(ondisk->uid);
	ce->ce_gid   = ntohl(ondisk->gid);
	ce->ce_size  = ntohl(ondisk->size);
	/* On-disk flags are just 16 bits */
	ce->ce_flags = ntohs(ondisk->flags);
	hashcpy(ce->sha1, ondisk->sha1);
1046 1047 1048 1049 1050 1051 1052 1053 1054

	len = ce->ce_flags & CE_NAMEMASK;
	if (len == CE_NAMEMASK)
		len = strlen(ondisk->name);
	/*
	 * NEEDSWORK: If the original index is crafted, this copy could
	 * go unchecked.
	 */
	memcpy(ce->name, ondisk->name, len + 1);
1055 1056
}

1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070
static inline size_t estimate_cache_size(size_t ondisk_size, unsigned int entries)
{
	long per_entry;

	per_entry = sizeof(struct cache_entry) - sizeof(struct ondisk_cache_entry);

	/*
	 * Alignment can cause differences. This should be "alignof", but
	 * since that's a gcc'ism, just use the size of a pointer.
	 */
	per_entry += sizeof(void *);
	return ondisk_size + entries*per_entry;
}

1071
/* remember to discard_cache() before reading a different cache! */
1072
int read_index_from(struct index_state *istate, const char *path)
1073 1074 1075
{
	int fd, i;
	struct stat st;
1076
	unsigned long src_offset, dst_offset;
1077
	struct cache_header *hdr;
1078 1079
	void *mmap;
	size_t mmap_size;
1080 1081

	errno = EBUSY;
1082
	if (istate->alloc)
1083
		return istate->cache_nr;
1084

1085
	errno = ENOENT;
1086
	istate->timestamp = 0;
1087
	fd = open(path, O_RDONLY);
1088 1089 1090 1091 1092
	if (fd < 0) {
		if (errno == ENOENT)
			return 0;
		die("index file open failed (%s)", strerror(errno));
	}
1093

1094
	if (fstat(fd, &st))
1095
		die("cannot stat the open index (%s)", strerror(errno));
1096 1097

	errno = EINVAL;
1098 1099
	mmap_size = xsize_t(st.st_size);
	if (mmap_size < sizeof(struct cache_header) + 20)
1100 1101
		die("index file smaller than expected");

1102
	mmap = xmmap(NULL, mmap_size, PROT_READ | PROT_WRITE, MAP_PRIVATE, fd, 0);
1103
	close(fd);
1104 1105
	if (mmap == MAP_FAILED)
		die("unable to map index file");
1106

1107 1108
	hdr = mmap;
	if (verify_hdr(hdr, mmap_size) < 0)
1109 1110
		goto unmap;

1111 1112 1113
	istate->cache_nr = ntohl(hdr->hdr_entries);
	istate->cache_alloc = alloc_nr(istate->cache_nr);
	istate->cache = xcalloc(istate->cache_alloc, sizeof(struct cache_entry *));
1114

1115 1116 1117 1118 1119 1120
	/*
	 * The disk format is actually larger than the in-memory format,
	 * due to space for nsec etc, so even though the in-memory one
	 * has room for a few  more flags, we can allocate using the same
	 * index size
	 */
1121
	istate->alloc = xmalloc(estimate_cache_size(mmap_size, istate->cache_nr));
1122 1123 1124

	src_offset = sizeof(*hdr);
	dst_offset = 0;
1125
	for (i = 0; i < istate->cache_nr; i++) {
1126
		struct ondisk_cache_entry *disk_ce;
1127 1128
		struct cache_entry *ce;

1129 1130 1131
		disk_ce = (struct ondisk_cache_entry *)((char *)mmap + src_offset);
		ce = (struct cache_entry *)((char *)istate->alloc + dst_offset);
		convert_from_disk(disk_ce, ce);
1132
		set_index_entry(istate, i, ce);
1133 1134 1135

		src_offset += ondisk_ce_size(ce);
		dst_offset += ce_size(ce);
1136
	}
1137
	istate->timestamp = st.st_mtime;
1138
	while (src_offset <= mmap_size - 20 - 8) {
1139 1140 1141 1142 1143 1144 1145
		/* After an array of active_nr index entries,
		 * there can be arbitrary number of extended
		 * sections, each of which is prefixed with
		 * extension name (4-byte) and section length
		 * in 4-byte network byte order.
		 */
		unsigned long extsize;
1146
		memcpy(&extsize, (char *)mmap + src_offset + 4, 4);
1147
		extsize = ntohl(extsize);
1148
		if (read_index_extension(istate,
1149 1150
					 (const char *) mmap + src_offset,
					 (char *) mmap + src_offset + 8,