read-cache.c 24.4 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
#include "cache.h"
7 8 9 10 11 12 13 14 15 16 17 18 19
#include "cache-tree.h"

/* 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" */
20

21
struct cache_entry **active_cache;
Junio C Hamano's avatar
Junio C Hamano committed
22
static time_t index_file_timestamp;
23
unsigned int active_nr, active_alloc, active_cache_changed;
24

25
struct cache_tree *active_cache_tree;
26

27
int cache_errno;
28

29 30
static void *cache_mmap;
static size_t cache_mmap_size;
31

32 33 34 35 36 37 38 39 40
/*
 * 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)
{
	ce->ce_ctime.sec = htonl(st->st_ctime);
	ce->ce_mtime.sec = htonl(st->st_mtime);
41
#ifdef USE_NSEC
42 43 44 45 46 47 48 49
	ce->ce_ctime.nsec = htonl(st->st_ctim.tv_nsec);
	ce->ce_mtime.nsec = htonl(st->st_mtim.tv_nsec);
#endif
	ce->ce_dev = htonl(st->st_dev);
	ce->ce_ino = htonl(st->st_ino);
	ce->ce_uid = htonl(st->st_uid);
	ce->ce_gid = htonl(st->st_gid);
	ce->ce_size = htonl(st->st_size);
Junio C Hamano's avatar
Junio C Hamano committed
50 51 52

	if (assume_unchanged)
		ce->ce_flags |= htons(CE_VALID);
53 54
}

Junio C Hamano's avatar
Junio C Hamano committed
55 56 57 58 59 60 61 62 63
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];
		if (!index_fd(sha1, fd, st, 0, NULL))
			match = memcmp(sha1, ce->sha1, 20);
64
		/* index_fd() closed the file descriptor already */
Junio C Hamano's avatar
Junio C Hamano committed
65 66 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 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112
	}
	return match;
}

static int ce_compare_link(struct cache_entry *ce, unsigned long expected_size)
{
	int match = -1;
	char *target;
	void *buffer;
	unsigned long size;
	char type[10];
	int len;

	target = xmalloc(expected_size);
	len = readlink(ce->name, target, expected_size);
	if (len != expected_size) {
		free(target);
		return -1;
	}
	buffer = read_sha1_file(ce->sha1, type, &size);
	if (!buffer) {
		free(target);
		return -1;
	}
	if (size == expected_size)
		match = memcmp(buffer, target, size);
	free(buffer);
	free(target);
	return match;
}

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:
		if (ce_compare_link(ce, st->st_size))
			return DATA_CHANGED;
		break;
	default:
		return TYPE_CHANGED;
	}
	return 0;
}

Junio C Hamano's avatar
Junio C Hamano committed
113
static int ce_match_stat_basic(struct cache_entry *ce, struct stat *st)
114 115 116
{
	unsigned int changed = 0;

117 118 119
	switch (ntohl(ce->ce_mode) & S_IFMT) {
	case S_IFREG:
		changed |= !S_ISREG(st->st_mode) ? TYPE_CHANGED : 0;
Junio C Hamano's avatar
Junio C Hamano committed
120 121 122 123 124
		/* We consider only the owner x bit to be relevant for
		 * "mode changes"
		 */
		if (trust_executable_bit &&
		    (0100 & (ntohl(ce->ce_mode) ^ st->st_mode)))
125
			changed |= MODE_CHANGED;
126 127 128 129 130 131 132
		break;
	case S_IFLNK:
		changed |= !S_ISLNK(st->st_mode) ? TYPE_CHANGED : 0;
		break;
	default:
		die("internal error: ce_mode is %o", ntohl(ce->ce_mode));
	}
133
	if (ce->ce_mtime.sec != htonl(st->st_mtime))
134
		changed |= MTIME_CHANGED;
135 136 137
	if (ce->ce_ctime.sec != htonl(st->st_ctime))
		changed |= CTIME_CHANGED;

138
#ifdef USE_NSEC
139 140 141 142 143
	/*
	 * nsec seems unreliable - not all filesystems support it, so
	 * as long as it is in the inode cache you get right nsec
	 * but after it gets flushed, you get zero nsec.
	 */
144
	if (ce->ce_mtime.nsec != htonl(st->st_mtim.tv_nsec))
145
		changed |= MTIME_CHANGED;
146
	if (ce->ce_ctime.nsec != htonl(st->st_ctim.tv_nsec))
147
		changed |= CTIME_CHANGED;
148 149 150 151
#endif	

	if (ce->ce_uid != htonl(st->st_uid) ||
	    ce->ce_gid != htonl(st->st_gid))
152
		changed |= OWNER_CHANGED;
153
	if (ce->ce_ino != htonl(st->st_ino))
154
		changed |= INODE_CHANGED;
155 156 157 158 159 160 161 162 163 164 165

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

166
	if (ce->ce_size != htonl(st->st_size))
167
		changed |= DATA_CHANGED;
168

Junio C Hamano's avatar
Junio C Hamano committed
169 170 171
	return changed;
}

172
int ce_match_stat(struct cache_entry *ce, struct stat *st, int options)
Junio C Hamano's avatar
Junio C Hamano committed
173
{
Junio C Hamano's avatar
Junio C Hamano committed
174
	unsigned int changed;
175 176
	int ignore_valid = options & 01;
	int assume_racy_is_modified = options & 02;
Junio C Hamano's avatar
Junio C Hamano committed
177 178 179 180 181 182 183 184 185

	/*
	 * If it's marked as always valid in the index, it's
	 * valid whatever the checked-out copy says.
	 */
	if (!ignore_valid && (ce->ce_flags & htons(CE_VALID)))
		return 0;

	changed = ce_match_stat_basic(ce, st);
Junio C Hamano's avatar
Junio C Hamano committed
186

Junio C Hamano's avatar
Junio C Hamano committed
187 188 189 190 191 192 193 194 195 196 197 198 199 200
	/*
	 * 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
201
	 * carefully than others.
Junio C Hamano's avatar
Junio C Hamano committed
202 203 204
	 */
	if (!changed &&
	    index_file_timestamp &&
205 206 207 208 209 210
	    index_file_timestamp <= ntohl(ce->ce_mtime.sec)) {
		if (assume_racy_is_modified)
			changed |= DATA_CHANGED;
		else
			changed |= ce_modified_check_fs(ce, st);
	}
211

Junio C Hamano's avatar
Junio C Hamano committed
212
	return changed;
213 214
}

Junio C Hamano's avatar
Junio C Hamano committed
215
int ce_modified(struct cache_entry *ce, struct stat *st, int really)
216
{
Junio C Hamano's avatar
Junio C Hamano committed
217
	int changed, changed_fs;
Junio C Hamano's avatar
Junio C Hamano committed
218
	changed = ce_match_stat(ce, st, really);
219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234
	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.
	 */
	if ((changed & DATA_CHANGED) && ce->ce_size != htonl(0))
		return changed;

Junio C Hamano's avatar
Junio C Hamano committed
235 236 237
	changed_fs = ce_modified_check_fs(ce, st);
	if (changed_fs)
		return changed | changed_fs;
238 239 240
	return 0;
}

241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259
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];
	if (!c1 && S_ISDIR(mode1))
		c1 = '/';
	if (!c2 && S_ISDIR(mode2))
		c2 = '/';
	return (c1 < c2) ? -1 : (c1 > c2) ? 1 : 0;
}

260
int cache_name_compare(const char *name1, int flags1, const char *name2, int flags2)
261
{
262 263
	int len1 = flags1 & CE_NAMEMASK;
	int len2 = flags2 & CE_NAMEMASK;
264 265 266 267 268 269 270 271 272 273
	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
274

275 276 277
	/* Compare stages  */
	flags1 &= CE_STAGEMASK;
	flags2 &= CE_STAGEMASK;
Junio C Hamano's avatar
Junio C Hamano committed
278

279 280 281 282
	if (flags1 < flags2)
		return -1;
	if (flags1 > flags2)
		return 1;
283 284 285 286 287 288 289 290 291 292 293 294
	return 0;
}

int cache_name_pos(const char *name, int namelen)
{
	int first, last;

	first = 0;
	last = active_nr;
	while (last > first) {
		int next = (last + first) >> 1;
		struct cache_entry *ce = active_cache[next];
295
		int cmp = cache_name_compare(name, namelen, ce->name, ntohs(ce->ce_flags));
296
		if (!cmp)
297
			return next;
298 299 300 301 302 303
		if (cmp < 0) {
			last = next;
			continue;
		}
		first = next+1;
	}
304
	return -first-1;
305 306
}

307
/* Remove entry, return true if there are more entries to go.. */
308
int remove_cache_entry_at(int pos)
309
{
310
	active_cache_changed = 1;
311 312 313 314 315 316 317
	active_nr--;
	if (pos >= active_nr)
		return 0;
	memmove(active_cache + pos, active_cache + pos + 1, (active_nr - pos) * sizeof(struct cache_entry *));
	return 1;
}

Junio C Hamano's avatar
Junio C Hamano committed
318
int remove_file_from_cache(const char *path)
319 320
{
	int pos = cache_name_pos(path, strlen(path));
321 322 323
	if (pos < 0)
		pos = -pos-1;
	while (pos < active_nr && !strcmp(active_cache[pos]->name, path))
324
		remove_cache_entry_at(pos);
325 326 327
	return 0;
}

Johannes Schindelin's avatar
Johannes Schindelin committed
328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366
int add_file_to_index(const char *path, int verbose)
{
	int size, namelen;
	struct stat st;
	struct cache_entry *ce;

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

	if (!S_ISREG(st.st_mode) && !S_ISLNK(st.st_mode))
		die("%s: can only add regular files or symbolic links", path);

	namelen = strlen(path);
	size = cache_entry_size(namelen);
	ce = xcalloc(1, size);
	memcpy(ce->name, path, namelen);
	ce->ce_flags = htons(namelen);
	fill_stat_cache_info(ce, &st);

	ce->ce_mode = create_ce_mode(st.st_mode);
	if (!trust_executable_bit) {
		/* If there is an existing entry, pick the mode bits
		 * from it.
		 */
		int pos = cache_name_pos(path, namelen);
		if (pos >= 0)
			ce->ce_mode = active_cache[pos]->ce_mode;
	}

	if (index_path(ce->sha1, path, &st, 1))
		die("unable to index file %s", path);
	if (add_cache_entry(ce, ADD_CACHE_OK_TO_ADD))
		die("unable to add %s to index",path);
	if (verbose)
		printf("add '%s'\n", path);
	cache_tree_invalidate_path(active_cache_tree, path);
	return 0;
}

367
int ce_same_name(struct cache_entry *a, struct cache_entry *b)
368 369 370 371 372
{
	int len = ce_namelen(a);
	return ce_namelen(b) == len && !memcmp(a->name, b->name, len);
}

373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392
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;
393 394
		if (!matchlen)
			return 1;
395 396 397 398
	}
	return 0;
}

399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462
/*
 * 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++;
	}
}

463 464 465
/*
 * Do we have another file that has the beginning components being a
 * proper superset of the name we're trying to add?
466
 */
467
static int has_file_name(const struct cache_entry *ce, int pos, int ok_to_replace)
468
{
469 470
	int retval = 0;
	int len = ce_namelen(ce);
471
	int stage = ce_stage(ce);
472
	const char *name = ce->name;
473

474 475
	while (pos < active_nr) {
		struct cache_entry *p = active_cache[pos++];
476

477
		if (len >= ce_namelen(p))
478
			break;
479 480
		if (memcmp(name, p->name, len))
			break;
481 482
		if (ce_stage(p) != stage)
			continue;
483 484 485 486 487 488
		if (p->name[len] != '/')
			continue;
		retval = -1;
		if (!ok_to_replace)
			break;
		remove_cache_entry_at(--pos);
489
	}
490 491
	return retval;
}
492

493 494 495 496 497 498 499
/*
 * Do we have another file with a pathname that is a proper
 * subset of the name we're trying to add?
 */
static int has_dir_name(const struct cache_entry *ce, int pos, int ok_to_replace)
{
	int retval = 0;
500
	int stage = ce_stage(ce);
501 502
	const char *name = ce->name;
	const char *slash = name + ce_namelen(ce);
503

504 505
	for (;;) {
		int len;
506

507 508 509 510 511 512 513
		for (;;) {
			if (*--slash == '/')
				break;
			if (slash <= ce->name)
				return retval;
		}
		len = slash - name;
514

515
		pos = cache_name_pos(name, ntohs(create_ce_flags(len, stage)));
516 517 518 519
		if (pos >= 0) {
			retval = -1;
			if (ok_to_replace)
				break;
520
			remove_cache_entry_at(pos);
521 522 523 524 525 526
			continue;
		}

		/*
		 * Trivial optimization: if we find an entry that
		 * already matches the sub-directory, then we know
527
		 * we're ok, and we can exit.
528 529
		 */
		pos = -pos-1;
530
		while (pos < active_nr) {
531
			struct cache_entry *p = active_cache[pos];
532 533 534 535 536 537 538 539 540 541 542 543
			if ((ce_namelen(p) <= len) ||
			    (p->name[len] != '/') ||
			    memcmp(p->name, name, len))
				break; /* not our subdirectory */
			if (ce_stage(p) == stage)
				/* p is at the same stage as our entry, and
				 * is a subdirectory of what we are looking
				 * at, so we cannot have conflicts at our
				 * level or anything shorter.
				 */
				return retval;
			pos++;
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
	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.
 * 
 * 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.
 */
static int check_file_directory_conflict(const struct cache_entry *ce, int pos, int ok_to_replace)
{
	/*
	 * We check if the path is a sub-path of a subsequent pathname
	 * first, since removing those will not change the position
	 * in the array
	 */
	int retval = has_file_name(ce, pos, ok_to_replace);
	/*
	 * Then check if the path might have a clashing sub-directory
	 * before it.
	 */
	return retval + has_dir_name(ce, pos, ok_to_replace);
571 572
}

573
int add_cache_entry(struct cache_entry *ce, int option)
574 575
{
	int pos;
576 577
	int ok_to_add = option & ADD_CACHE_OK_TO_ADD;
	int ok_to_replace = option & ADD_CACHE_OK_TO_REPLACE;
578
	int skip_df_check = option & ADD_CACHE_SKIP_DFCHECK;
Junio C Hamano's avatar
Junio C Hamano committed
579

580
	pos = cache_name_pos(ce->name, ntohs(ce->ce_flags));
581

Junio C Hamano's avatar
Junio C Hamano committed
582
	/* existing match? Just replace it. */
583
	if (pos >= 0) {
584
		active_cache_changed = 1;
585
		active_cache[pos] = ce;
586 587
		return 0;
	}
588
	pos = -pos-1;
589

590 591 592 593 594
	/*
	 * Inserting a merged entry ("stage 0") into the index
	 * will always replace all non-merged entries..
	 */
	if (pos < active_nr && ce_stage(ce) == 0) {
595
		while (ce_same_name(active_cache[pos], ce)) {
596
			ok_to_add = 1;
597
			if (!remove_cache_entry_at(pos))
598 599 600 601
				break;
		}
	}

602 603
	if (!ok_to_add)
		return -1;
604 605
	if (!verify_path(ce->name))
		return -1;
606

Junio C Hamano's avatar
Junio C Hamano committed
607 608
	if (!skip_df_check &&
	    check_file_directory_conflict(ce, pos, ok_to_replace)) {
609 610
		if (!ok_to_replace)
			return -1;
611
		pos = cache_name_pos(ce->name, ntohs(ce->ce_flags));
612 613
		pos = -pos-1;
	}
614

615 616 617
	/* Make sure the array is big enough .. */
	if (active_nr == active_alloc) {
		active_alloc = alloc_nr(active_alloc);
618
		active_cache = xrealloc(active_cache, active_alloc * sizeof(struct cache_entry *));
619 620 621 622 623 624 625
	}

	/* Add it in.. */
	active_nr++;
	if (active_nr > pos)
		memmove(active_cache + pos + 1, active_cache + pos, (active_nr - pos - 1) * sizeof(ce));
	active_cache[pos] = ce;
626
	active_cache_changed = 1;
627 628 629
	return 0;
}

630 631 632 633 634 635 636 637 638 639 640
/*
 * "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.
 */
641
struct cache_entry *refresh_cache_entry(struct cache_entry *ce, int really)
642 643 644 645 646
{
	struct stat st;
	struct cache_entry *updated;
	int changed, size;

647 648 649 650
	if (lstat(ce->name, &st) < 0) {
		cache_errno = errno;
		return NULL;
	}
651 652 653 654 655 656 657

	changed = ce_match_stat(ce, &st, really);
	if (!changed) {
		if (really && assume_unchanged &&
		    !(ce->ce_flags & htons(CE_VALID)))
			; /* mark this one VALID again */
		else
658
			return ce;
659 660
	}

661 662 663 664
	if (ce_modified(ce, &st, really)) {
		cache_errno = EINVAL;
		return NULL;
	}
665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706

	size = ce_size(ce);
	updated = xmalloc(size);
	memcpy(updated, ce, size);
	fill_stat_cache_info(updated, &st);

	/* In this case, if really 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.
	 */
	if (!really && assume_unchanged && !(ce->ce_flags & htons(CE_VALID)))
		updated->ce_flags &= ~htons(CE_VALID);

	return updated;
}

int refresh_cache(unsigned int flags)
{
	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;

	for (i = 0; i < active_nr; i++) {
		struct cache_entry *ce, *new;
		ce = active_cache[i];
		if (ce_stage(ce)) {
			while ((i < active_nr) &&
			       ! strcmp(active_cache[i]->name, ce->name))
				i++;
			i--;
			if (allow_unmerged)
				continue;
			printf("%s: needs merge\n", ce->name);
			has_errors = 1;
			continue;
		}

707 708
		new = refresh_cache_entry(ce, really);
		if (new == ce)
709
			continue;
710 711
		if (!new) {
			if (not_new && cache_errno == ENOENT)
712
				continue;
713
			if (really && cache_errno == EINVAL) {
714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734
				/* If we are doing --really-refresh that
				 * means the index is not valid anymore.
				 */
				ce->ce_flags &= ~htons(CE_VALID);
				active_cache_changed = 1;
			}
			if (quiet)
				continue;
			printf("%s: needs update\n", ce->name);
			has_errors = 1;
			continue;
		}
		active_cache_changed = 1;
		/* You can NOT just free active_cache[i] here, since it
		 * might not be necessarily malloc()ed but can also come
		 * from mmap(). */
		active_cache[i] = new;
	}
	return has_errors;
}

735
static int verify_hdr(struct cache_header *hdr, unsigned long size)
736 737
{
	SHA_CTX c;
738
	unsigned char sha1[20];
739

740
	if (hdr->hdr_signature != htonl(CACHE_SIGNATURE))
741
		return error("bad signature");
742 743
	if (hdr->hdr_version != htonl(2))
		return error("bad index version");
744
	SHA1_Init(&c);
745
	SHA1_Update(&c, hdr, size - 20);
746
	SHA1_Final(sha1, &c);
747
	if (memcmp(sha1, (char *) hdr + size - 20, 20))
748
		return error("bad index file sha1 signature");
749 750 751
	return 0;
}

752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768
static int read_index_extension(const char *ext, void *data, unsigned long sz)
{
	switch (CACHE_EXT(ext)) {
	case CACHE_EXT_TREE:
		active_cache_tree = cache_tree_read(data, sz);
		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;
}

int read_cache(void)
769 770 771 772 773 774
{
	return read_cache_from(get_index_file());
}

/* remember to discard_cache() before reading a different cache! */
int read_cache_from(const char *path)
775 776 777
{
	int fd, i;
	struct stat st;
778
	unsigned long offset;
779 780 781
	struct cache_header *hdr;

	errno = EBUSY;
782
	if (cache_mmap)
783 784
		return active_nr;

785
	errno = ENOENT;
Junio C Hamano's avatar
Junio C Hamano committed
786
	index_file_timestamp = 0;
787
	fd = open(path, O_RDONLY);
788 789 790 791 792
	if (fd < 0) {
		if (errno == ENOENT)
			return 0;
		die("index file open failed (%s)", strerror(errno));
	}
793

794
	cache_mmap = MAP_FAILED;
795
	if (!fstat(fd, &st)) {
796
		cache_mmap_size = st.st_size;
797
		errno = EINVAL;
798 799
		if (cache_mmap_size >= sizeof(struct cache_header) + 20)
			cache_mmap = mmap(NULL, cache_mmap_size, PROT_READ | PROT_WRITE, MAP_PRIVATE, fd, 0);
800 801
	}
	close(fd);
802
	if (cache_mmap == MAP_FAILED)
803
		die("index file mmap failed (%s)", strerror(errno));
804

805 806
	hdr = cache_mmap;
	if (verify_hdr(hdr, cache_mmap_size) < 0)
807 808
		goto unmap;

809
	active_nr = ntohl(hdr->hdr_entries);
810
	active_alloc = alloc_nr(active_nr);
811
	active_cache = xcalloc(active_alloc, sizeof(struct cache_entry *));
812 813

	offset = sizeof(*hdr);
814
	for (i = 0; i < active_nr; i++) {
815
		struct cache_entry *ce = (struct cache_entry *) ((char *) cache_mmap + offset);
816 817 818
		offset = offset + ce_size(ce);
		active_cache[i] = ce;
	}
Junio C Hamano's avatar
Junio C Hamano committed
819
	index_file_timestamp = st.st_mtime;
820
	while (offset <= cache_mmap_size - 20 - 8) {
821 822 823 824 825 826 827
		/* 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;
828
		memcpy(&extsize, (char *) cache_mmap + offset + 4, 4);
829
		extsize = ntohl(extsize);
830 831
		if (read_index_extension(((const char *) cache_mmap) + offset,
					 (char *) cache_mmap + offset + 8,
832
					 extsize) < 0)
833 834 835 836
			goto unmap;
		offset += 8;
		offset += extsize;
	}
837 838 839
	return active_nr;

unmap:
840
	munmap(cache_mmap, cache_mmap_size);
841
	errno = EINVAL;
842
	die("index file corrupt");
843 844
}

845
#define WRITE_BUFFER_SIZE 8192
846
static unsigned char write_buffer[WRITE_BUFFER_SIZE];
847 848
static unsigned long write_buffer_len;

849 850 851 852 853 854 855 856 857 858 859 860
static int ce_write_flush(SHA_CTX *context, int fd)
{
	unsigned int buffered = write_buffer_len;
	if (buffered) {
		SHA1_Update(context, write_buffer, buffered);
		if (write(fd, write_buffer, buffered) != buffered)
			return -1;
		write_buffer_len = 0;
	}
	return 0;
}

861
static int ce_write(SHA_CTX *context, int fd, void *data, unsigned int len)
862 863 864 865 866 867 868 869 870
{
	while (len) {
		unsigned int buffered = write_buffer_len;
		unsigned int partial = WRITE_BUFFER_SIZE - buffered;
		if (partial > len)
			partial = len;
		memcpy(write_buffer + buffered, data, partial);
		buffered += partial;
		if (buffered == WRITE_BUFFER_SIZE) {
871 872
			write_buffer_len = buffered;
			if (ce_write_flush(context, fd))
873 874 875 876 877
				return -1;
			buffered = 0;
		}
		write_buffer_len = buffered;
		len -= partial;
878
		data = (char *) data + partial;
879 880 881 882
 	}
 	return 0;
}

883
static int write_index_ext_header(SHA_CTX *context, int fd,
884
				  unsigned int ext, unsigned int sz)
885 886 887
{
	ext = htonl(ext);
	sz = htonl(sz);
David Rientjes's avatar
David Rientjes committed
888 889
	return ((ce_write(context, fd, &ext, 4) < 0) ||
		(ce_write(context, fd, &sz, 4) < 0)) ? -1 : 0;
890 891 892
}

static int ce_flush(SHA_CTX *context, int fd)
893 894
{
	unsigned int left = write_buffer_len;
895

896 897
	if (left) {
		write_buffer_len = 0;
898
		SHA1_Update(context, write_buffer, left);
899
	}
900

901 902 903 904 905 906 907
	/* Flush first if not enough space for SHA1 signature */
	if (left + 20 > WRITE_BUFFER_SIZE) {
		if (write(fd, write_buffer, left) != left)
			return -1;
		left = 0;
	}

908
	/* Append the SHA1 signature at the end */
909
	SHA1_Final(write_buffer + left, context);
910
	left += 20;
David Rientjes's avatar
David Rientjes committed
911
	return (write(fd, write_buffer, left) != left) ? -1 : 0;
912 913
}

Junio C Hamano's avatar
Junio C Hamano committed
914 915 916 917 918 919 920 921 922 923 924 925 926 927 928
static void ce_smudge_racily_clean_entry(struct cache_entry *ce)
{
	/*
	 * The only thing we care about in this function is to smudge the
	 * falsely clean entry due to touch-update-touch race, so we leave
	 * everything else as they are.  We are called for entries whose
	 * ce_mtime match the index file mtime.
	 */
	struct stat st;

	if (lstat(ce->name, &st) < 0)
		return;
	if (ce_match_stat_basic(ce, &st))
		return;
	if (ce_modified_check_fs(ce, &st)) {
929 930 931 932 933 934 935 936 937 938 939
		/* This is "racily clean"; smudge it.  Note that this
		 * is a tricky code.  At first glance, it may appear
		 * that it can break with this sequence:
		 *
		 * $ echo xyzzy >frotz
		 * $ git-update-index --add frotz
		 * $ : >frotz
		 * $ sleep 3
		 * $ echo filfre >nitfol
		 * $ git-update-index --add nitfol
		 *
940
		 * but it does not.  When the second update-index runs,
941 942 943 944 945 946 947 948 949 950 951 952 953
		 * it notices that the entry "frotz" has the same timestamp
		 * as index, and if we were to smudge it by resetting its
		 * size to zero here, then the object name recorded
		 * in index is the 6-byte file but the cached stat information
		 * becomes zero --- which would then match what we would
		 * obtain from the filesystem next time we stat("frotz"). 
		 *
		 * However, the second update-index, before calling
		 * this function, notices that the cached size is 6
		 * bytes and what is on the filesystem is an empty
		 * file, and never calls us, so the cached size information
		 * for "frotz" stays 6 which does not match the filesystem.
		 */
Junio C Hamano's avatar
Junio C Hamano committed
954 955 956 957
		ce->ce_size = htonl(0);
	}
}

958
int write_cache(int newfd, struct cache_entry **cache, int entries)
959 960 961
{
	SHA_CTX c;
	struct cache_header hdr;
962
	int i, removed;
963 964 965 966

	for (i = removed = 0; i < entries; i++)
		if (!cache[i]->ce_mode)
			removed++;
967

968
	hdr.hdr_signature = htonl(CACHE_SIGNATURE);
969
	hdr.hdr_version = htonl(2);
970
	hdr.hdr_entries = htonl(entries - removed);
971 972

	SHA1_Init(&c);
973
	if (ce_write(&c, newfd, &hdr, sizeof(hdr)) < 0)
974 975 976 977
		return -1;

	for (i = 0; i < entries; i++) {
		struct cache_entry *ce = cache[i];
978 979
		if (!ce->ce_mode)
			continue;
980 981
		if (index_file_timestamp &&
		    index_file_timestamp <= ntohl(ce->ce_mtime.sec))
Junio C Hamano's avatar
Junio C Hamano committed
982
			ce_smudge_racily_clean_entry(ce);
983
		if (ce_write(&c, newfd, ce, ce_size(ce)) < 0)
984 985
			return -1;
	}
986

987 988 989 990 991 992 993 994 995 996 997 998 999 1000
	/* Write extension data here */
	if (active_cache_tree) {
		unsigned long sz;
		void *data = cache_tree_write(active_cache_tree, &sz);
		if (data &&
		    !write_index_ext_header(&c, newfd, CACHE_EXT_TREE, sz) &&
		    !ce_write(&c, newfd, data, sz))
			;
		else {
			free(data);
			return -1;
		}
	}
	return ce_flush(&c, newfd);
1001
}