read-cache.c 33.6 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1 2 3 4 5
/*
 * GIT - The information manager from hell
 *
 * Copyright (C) Linus Torvalds, 2005
 */
6
#define NO_THE_INDEX_COMPATIBILITY_MACROS
7
#include "cache.h"
8
#include "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
{
	void **pos;
40
	unsigned int hash;
41

42 43 44 45 46
	if (ce->ce_flags & CE_HASHED)
		return;
	ce->ce_flags |= CE_HASHED;
	ce->next = NULL;
	hash = hash_name(ce->name, ce_namelen(ce));
47 48 49 50 51 52 53
	pos = insert_hash(hash, ce, &istate->name_hash);
	if (pos) {
		ce->next = *pos;
		*pos = ce;
	}
}

Junio C Hamano's avatar
Junio C Hamano committed
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)
{
67
	ce->ce_flags &= ~CE_UNHASHED;
Junio C Hamano's avatar
Junio C Hamano committed
68 69 70 71 72
	istate->cache[nr] = ce;
	if (istate->name_hash_initialized)
		hash_index_entry(istate, ce);
}

73 74 75 76
static void replace_index_entry(struct index_state *istate, int nr, struct cache_entry *ce)
{
	struct cache_entry *old = istate->cache[nr];

77
	remove_index_entry(old);
78
	set_index_entry(istate, nr, ce);
79 80 81 82 83 84
	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
85 86 87 88
	struct cache_entry *ce;

	lazy_init_name_hash(istate);
	ce = lookup_hash(hash, &istate->name_hash);
89 90 91 92 93 94 95 96 97 98 99

	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;
}

100 101 102 103 104 105 106
/*
 * 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)
{
107 108 109 110 111 112 113
	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
114 115

	if (assume_unchanged)
116
		ce->ce_flags |= CE_VALID;
117 118 119

	if (S_ISREG(st->st_mode))
		ce_mark_uptodate(ce);
120 121
}

Junio C Hamano's avatar
Junio C Hamano committed
122 123 124 125 126 127 128
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];
129
		if (!index_fd(sha1, fd, st, 0, OBJ_BLOB, ce->name))
130
			match = hashcmp(sha1, ce->sha1);
131
		/* index_fd() closed the file descriptor already */
Junio C Hamano's avatar
Junio C Hamano committed
132 133 134 135
	}
	return match;
}

136
static int ce_compare_link(struct cache_entry *ce, size_t expected_size)
Junio C Hamano's avatar
Junio C Hamano committed
137 138 139 140 141
{
	int match = -1;
	char *target;
	void *buffer;
	unsigned long size;
142
	enum object_type type;
Junio C Hamano's avatar
Junio C Hamano committed
143 144 145 146 147 148 149 150
	int len;

	target = xmalloc(expected_size);
	len = readlink(ce->name, target, expected_size);
	if (len != expected_size) {
		free(target);
		return -1;
	}
151
	buffer = read_sha1_file(ce->sha1, &type, &size);
Junio C Hamano's avatar
Junio C Hamano committed
152 153 154 155 156 157 158 159 160 161 162
	if (!buffer) {
		free(target);
		return -1;
	}
	if (size == expected_size)
		match = memcmp(buffer, target, size);
	free(buffer);
	free(target);
	return match;
}

163 164 165 166 167 168
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
169
	 * under GITLINK directory be a valid git directory. It
170 171 172 173 174 175 176 177 178 179
	 * 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
180 181 182 183 184 185 186 187
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:
188
		if (ce_compare_link(ce, xsize_t(st->st_size)))
Junio C Hamano's avatar
Junio C Hamano committed
189 190
			return DATA_CHANGED;
		break;
191
	case S_IFDIR:
192
		if (S_ISGITLINK(ce->ce_mode))
193
			return 0;
Junio C Hamano's avatar
Junio C Hamano committed
194 195 196 197 198 199
	default:
		return TYPE_CHANGED;
	}
	return 0;
}

Junio C Hamano's avatar
Junio C Hamano committed
200
static int ce_match_stat_basic(struct cache_entry *ce, struct stat *st)
201 202 203
{
	unsigned int changed = 0;

204 205 206 207
	if (ce->ce_flags & CE_REMOVE)
		return MODE_CHANGED | DATA_CHANGED | TYPE_CHANGED;

	switch (ce->ce_mode & S_IFMT) {
208 209
	case S_IFREG:
		changed |= !S_ISREG(st->st_mode) ? TYPE_CHANGED : 0;
Junio C Hamano's avatar
Junio C Hamano committed
210 211 212 213
		/* We consider only the owner x bit to be relevant for
		 * "mode changes"
		 */
		if (trust_executable_bit &&
214
		    (0100 & (ce->ce_mode ^ st->st_mode)))
215
			changed |= MODE_CHANGED;
216 217
		break;
	case S_IFLNK:
218 219 220
		if (!S_ISLNK(st->st_mode) &&
		    (has_symlinks || !S_ISREG(st->st_mode)))
			changed |= TYPE_CHANGED;
221
		break;
Martin Waitz's avatar
Martin Waitz committed
222
	case S_IFGITLINK:
223 224 225 226
		if (!S_ISDIR(st->st_mode))
			changed |= TYPE_CHANGED;
		else if (ce_compare_gitlink(ce))
			changed |= DATA_CHANGED;
227
		return changed;
228
	default:
229
		die("internal error: ce_mode is %o", ce->ce_mode);
230
	}
231
	if (ce->ce_mtime != (unsigned int) st->st_mtime)
232
		changed |= MTIME_CHANGED;
233
	if (ce->ce_ctime != (unsigned int) st->st_ctime)
234
		changed |= CTIME_CHANGED;
235

236 237
	if (ce->ce_uid != (unsigned int) st->st_uid ||
	    ce->ce_gid != (unsigned int) st->st_gid)
238
		changed |= OWNER_CHANGED;
239
	if (ce->ce_ino != (unsigned int) st->st_ino)
240
		changed |= INODE_CHANGED;
241 242 243 244 245 246 247

#ifdef USE_STDEV
	/*
	 * st_dev breaks on network filesystems where different
	 * clients will have different views of what "device"
	 * the filesystem is on
	 */
248
	if (ce->ce_dev != (unsigned int) st->st_dev)
249 250 251
		changed |= INODE_CHANGED;
#endif

252
	if (ce->ce_size != (unsigned int) st->st_size)
253
		changed |= DATA_CHANGED;
254

Junio C Hamano's avatar
Junio C Hamano committed
255 256 257
	return changed;
}

258 259 260 261 262 263
static int is_racy_timestamp(struct index_state *istate, struct cache_entry *ce)
{
	return (istate->timestamp &&
		((unsigned int)istate->timestamp) <= ce->ce_mtime);
}

264
int ie_match_stat(struct index_state *istate,
265 266
		  struct cache_entry *ce, struct stat *st,
		  unsigned int options)
Junio C Hamano's avatar
Junio C Hamano committed
267
{
Junio C Hamano's avatar
Junio C Hamano committed
268
	unsigned int changed;
269 270
	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
271 272 273 274 275

	/*
	 * If it's marked as always valid in the index, it's
	 * valid whatever the checked-out copy says.
	 */
276
	if (!ignore_valid && (ce->ce_flags & CE_VALID))
Junio C Hamano's avatar
Junio C Hamano committed
277 278 279
		return 0;

	changed = ce_match_stat_basic(ce, st);
Junio C Hamano's avatar
Junio C Hamano committed
280

Junio C Hamano's avatar
Junio C Hamano committed
281 282 283 284 285 286 287 288 289 290 291 292 293 294
	/*
	 * 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
295
	 * carefully than others.
Junio C Hamano's avatar
Junio C Hamano committed
296
	 */
297
	if (!changed && is_racy_timestamp(istate, ce)) {
298 299 300 301 302
		if (assume_racy_is_modified)
			changed |= DATA_CHANGED;
		else
			changed |= ce_modified_check_fs(ce, st);
	}
303

Junio C Hamano's avatar
Junio C Hamano committed
304
	return changed;
305 306
}

307
int ie_modified(struct index_state *istate,
308
		struct cache_entry *ce, struct stat *st, unsigned int options)
309
{
Junio C Hamano's avatar
Junio C Hamano committed
310
	int changed, changed_fs;
311 312

	changed = ie_match_stat(istate, ce, st, options);
313 314 315 316 317 318 319 320 321 322 323 324 325
	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.
	 */
326
	if ((changed & DATA_CHANGED) && ce->ce_size != 0)
327 328
		return changed;

Junio C Hamano's avatar
Junio C Hamano committed
329 330 331
	changed_fs = ce_modified_check_fs(ce, st);
	if (changed_fs)
		return changed | changed_fs;
332 333 334
	return 0;
}

335 336 337 338 339 340 341 342 343 344 345 346
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];
347
	if (!c1 && S_ISDIR(mode1))
348
		c1 = '/';
349
	if (!c2 && S_ISDIR(mode2))
350 351 352 353
		c2 = '/';
	return (c1 < c2) ? -1 : (c1 > c2) ? 1 : 0;
}

354
int cache_name_compare(const char *name1, int flags1, const char *name2, int flags2)
355
{
356 357
	int len1 = flags1 & CE_NAMEMASK;
	int len2 = flags2 & CE_NAMEMASK;
358 359 360 361 362 363 364 365 366 367
	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
368

369 370 371
	/* Compare stages  */
	flags1 &= CE_STAGEMASK;
	flags2 &= CE_STAGEMASK;
Junio C Hamano's avatar
Junio C Hamano committed
372

373 374 375 376
	if (flags1 < flags2)
		return -1;
	if (flags1 > flags2)
		return 1;
377 378 379
	return 0;
}

380
int index_name_pos(struct index_state *istate, const char *name, int namelen)
381 382 383 384
{
	int first, last;

	first = 0;
385
	last = istate->cache_nr;
386 387
	while (last > first) {
		int next = (last + first) >> 1;
388
		struct cache_entry *ce = istate->cache[next];
389
		int cmp = cache_name_compare(name, namelen, ce->name, ce->ce_flags);
390
		if (!cmp)
391
			return next;
392 393 394 395 396 397
		if (cmp < 0) {
			last = next;
			continue;
		}
		first = next+1;
	}
398
	return -first-1;
399 400
}

401
/* Remove entry, return true if there are more entries to go.. */
402
int remove_index_entry_at(struct index_state *istate, int pos)
403
{
404 405
	struct cache_entry *ce = istate->cache[pos];

406
	remove_index_entry(ce);
407 408 409
	istate->cache_changed = 1;
	istate->cache_nr--;
	if (pos >= istate->cache_nr)
410
		return 0;
411 412 413
	memmove(istate->cache + pos,
		istate->cache + pos + 1,
		(istate->cache_nr - pos) * sizeof(struct cache_entry *));
414 415 416
	return 1;
}

417
int remove_file_from_index(struct index_state *istate, const char *path)
418
{
419
	int pos = index_name_pos(istate, path, strlen(path));
420 421
	if (pos < 0)
		pos = -pos-1;
Junio C Hamano's avatar
Junio C Hamano committed
422
	cache_tree_invalidate_path(istate->cache_tree, path);
423 424
	while (pos < istate->cache_nr && !strcmp(istate->cache[pos]->name, path))
		remove_index_entry_at(istate, pos);
425 426 427
	return 0;
}

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
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;
}

456
int add_file_to_index(struct index_state *istate, const char *path, int verbose)
Johannes Schindelin's avatar
Johannes Schindelin committed
457
{
458
	int size, namelen, pos;
Johannes Schindelin's avatar
Johannes Schindelin committed
459 460
	struct stat st;
	struct cache_entry *ce;
461
	unsigned ce_option = CE_MATCH_IGNORE_VALID|CE_MATCH_RACY_IS_DIRTY;
Johannes Schindelin's avatar
Johannes Schindelin committed
462 463 464 465

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

466 467
	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
468 469

	namelen = strlen(path);
470 471 472 473
	if (S_ISDIR(st.st_mode)) {
		while (namelen && path[namelen-1] == '/')
			namelen--;
	}
Johannes Schindelin's avatar
Johannes Schindelin committed
474 475 476
	size = cache_entry_size(namelen);
	ce = xcalloc(1, size);
	memcpy(ce->name, path, namelen);
477
	ce->ce_flags = namelen;
Johannes Schindelin's avatar
Johannes Schindelin committed
478 479
	fill_stat_cache_info(ce, &st);

480
	if (trust_executable_bit && has_symlinks)
481 482
		ce->ce_mode = create_ce_mode(st.st_mode);
	else {
483 484
		/* 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
485
		 */
486
		struct cache_entry *ent;
487
		int pos = index_name_pos_also_unmerged(istate, path, namelen);
488

489
		ent = (0 <= pos) ? istate->cache[pos] : NULL;
490
		ce->ce_mode = ce_mode_from_stat(ent, st.st_mode);
Johannes Schindelin's avatar
Johannes Schindelin committed
491 492
	}

493 494 495
	pos = index_name_pos(istate, ce->name, namelen);
	if (0 <= pos &&
	    !ce_stage(istate->cache[pos]) &&
496
	    !ie_match_stat(istate, istate->cache[pos], &st, ce_option)) {
497 498
		/* Nothing changed, really */
		free(ce);
499
		ce_mark_uptodate(istate->cache[pos]);
500 501 502
		return 0;
	}

Johannes Schindelin's avatar
Johannes Schindelin committed
503 504
	if (index_path(ce->sha1, path, &st, 1))
		die("unable to index file %s", path);
505
	if (add_index_entry(istate, ce, ADD_CACHE_OK_TO_ADD|ADD_CACHE_OK_TO_REPLACE))
Johannes Schindelin's avatar
Johannes Schindelin committed
506 507 508 509 510 511
		die("unable to add %s to index",path);
	if (verbose)
		printf("add '%s'\n", path);
	return 0;
}

512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536
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;
}

537
int ce_same_name(struct cache_entry *a, struct cache_entry *b)
538 539 540 541 542
{
	int len = ce_namelen(a);
	return ce_namelen(b) == len && !memcmp(a->name, b->name, len);
}

543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562
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;
563 564
		if (!matchlen)
			return 1;
565 566 567 568
	}
	return 0;
}

569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632
/*
 * 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++;
	}
}

633 634 635
/*
 * Do we have another file that has the beginning components being a
 * proper superset of the name we're trying to add?
636
 */
637 638
static int has_file_name(struct index_state *istate,
			 const struct cache_entry *ce, int pos, int ok_to_replace)
639
{
640 641
	int retval = 0;
	int len = ce_namelen(ce);
642
	int stage = ce_stage(ce);
643
	const char *name = ce->name;
644

645 646
	while (pos < istate->cache_nr) {
		struct cache_entry *p = istate->cache[pos++];
647

648
		if (len >= ce_namelen(p))
649
			break;
650 651
		if (memcmp(name, p->name, len))
			break;
652 653
		if (ce_stage(p) != stage)
			continue;
654 655
		if (p->name[len] != '/')
			continue;
656
		if (p->ce_flags & CE_REMOVE)
657
			continue;
658 659 660
		retval = -1;
		if (!ok_to_replace)
			break;
661
		remove_index_entry_at(istate, --pos);
662
	}
663 664
	return retval;
}
665

666 667 668 669
/*
 * Do we have another file with a pathname that is a proper
 * subset of the name we're trying to add?
 */
670 671
static int has_dir_name(struct index_state *istate,
			const struct cache_entry *ce, int pos, int ok_to_replace)
672 673
{
	int retval = 0;
674
	int stage = ce_stage(ce);
675 676
	const char *name = ce->name;
	const char *slash = name + ce_namelen(ce);
677

678 679
	for (;;) {
		int len;
680

681 682 683 684 685 686 687
		for (;;) {
			if (*--slash == '/')
				break;
			if (slash <= ce->name)
				return retval;
		}
		len = slash - name;
688

689
		pos = index_name_pos(istate, name, create_ce_flags(len, stage));
690
		if (pos >= 0) {
691 692 693 694 695 696 697 698
			/*
			 * 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.
			 */
699
			if (!(istate->cache[pos]->ce_flags & CE_REMOVE)) {
700 701 702
				retval = -1;
				if (!ok_to_replace)
					break;
703
				remove_index_entry_at(istate, pos);
704 705
				continue;
			}
706
		}
707 708
		else
			pos = -pos-1;
709 710 711 712

		/*
		 * Trivial optimization: if we find an entry that
		 * already matches the sub-directory, then we know
713
		 * we're ok, and we can exit.
714
		 */
715 716
		while (pos < istate->cache_nr) {
			struct cache_entry *p = istate->cache[pos];
717 718 719 720
			if ((ce_namelen(p) <= len) ||
			    (p->name[len] != '/') ||
			    memcmp(p->name, name, len))
				break; /* not our subdirectory */
721 722 723
			if (ce_stage(p) == stage && !(p->ce_flags & CE_REMOVE))
				/*
				 * p is at the same stage as our entry, and
724 725 726 727 728 729
				 * is a subdirectory of what we are looking
				 * at, so we cannot have conflicts at our
				 * level or anything shorter.
				 */
				return retval;
			pos++;
730
		}
731
	}
732 733 734 735 736 737 738
	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
739
 *
740 741 742 743
 * 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.
 */
744 745 746
static int check_file_directory_conflict(struct index_state *istate,
					 const struct cache_entry *ce,
					 int pos, int ok_to_replace)
747
{
748 749 750 751 752
	int retval;

	/*
	 * When ce is an "I am going away" entry, we allow it to be added
	 */
753
	if (ce->ce_flags & CE_REMOVE)
754 755
		return 0;

756 757 758
	/*
	 * We check if the path is a sub-path of a subsequent pathname
	 * first, since removing those will not change the position
759
	 * in the array.
760
	 */
761
	retval = has_file_name(istate, ce, pos, ok_to_replace);
762

763 764 765 766
	/*
	 * Then check if the path might have a clashing sub-directory
	 * before it.
	 */
767
	return retval + has_dir_name(istate, ce, pos, ok_to_replace);
768 769
}

770
static int add_index_entry_with_check(struct index_state *istate, struct cache_entry *ce, int option)
771 772
{
	int pos;
773 774
	int ok_to_add = option & ADD_CACHE_OK_TO_ADD;
	int ok_to_replace = option & ADD_CACHE_OK_TO_REPLACE;
775
	int skip_df_check = option & ADD_CACHE_SKIP_DFCHECK;
Junio C Hamano's avatar
Junio C Hamano committed
776

Junio C Hamano's avatar
Junio C Hamano committed
777
	cache_tree_invalidate_path(istate->cache_tree, ce->name);
778
	pos = index_name_pos(istate, ce->name, ce->ce_flags);
779

Junio C Hamano's avatar
Junio C Hamano committed
780
	/* existing match? Just replace it. */
781
	if (pos >= 0) {
782
		replace_index_entry(istate, pos, ce);
783 784
		return 0;
	}
785
	pos = -pos-1;
786

787 788 789 790
	/*
	 * Inserting a merged entry ("stage 0") into the index
	 * will always replace all non-merged entries..
	 */
791 792
	if (pos < istate->cache_nr && ce_stage(ce) == 0) {
		while (ce_same_name(istate->cache[pos], ce)) {
793
			ok_to_add = 1;
794
			if (!remove_index_entry_at(istate, pos))
795 796 797 798
				break;
		}
	}

799 800
	if (!ok_to_add)
		return -1;
801 802
	if (!verify_path(ce->name))
		return -1;
803

Junio C Hamano's avatar
Junio C Hamano committed
804
	if (!skip_df_check &&
805
	    check_file_directory_conflict(istate, ce, pos, ok_to_replace)) {
806
		if (!ok_to_replace)
807 808
			return error("'%s' appears as both a file and as a directory",
				     ce->name);
809
		pos = index_name_pos(istate, ce->name, ce->ce_flags);
810 811
		pos = -pos-1;
	}
812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827
	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;
	}
828

829
	/* Make sure the array is big enough .. */
830 831 832 833
	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 *));
834 835 836
	}

	/* Add it in.. */
837
	istate->cache_nr++;
838
	if (istate->cache_nr > pos + 1)
839 840 841
		memmove(istate->cache + pos + 1,
			istate->cache + pos,
			(istate->cache_nr - pos - 1) * sizeof(ce));
842
	set_index_entry(istate, pos, ce);
843
	istate->cache_changed = 1;
844 845 846
	return 0;
}

847 848 849 850 851 852 853 854 855 856 857
/*
 * "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.
 */
858
static struct cache_entry *refresh_cache_ent(struct index_state *istate,
859 860
					     struct cache_entry *ce,
					     unsigned int options, int *err)
861 862 863 864
{
	struct stat st;
	struct cache_entry *updated;
	int changed, size;
865
	int ignore_valid = options & CE_MATCH_IGNORE_VALID;
866

867 868 869
	if (ce_uptodate(ce))
		return ce;

870
	if (lstat(ce->name, &st) < 0) {
871 872
		if (err)
			*err = errno;
873 874
		return NULL;
	}
875

876
	changed = ie_match_stat(istate, ce, &st, options);
877
	if (!changed) {
878 879 880 881 882 883 884 885
		/*
		 * 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 &&
886
		    !(ce->ce_flags & CE_VALID))
887
			; /* mark this one VALID again */
888 889 890 891 892 893 894
		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);
895
			return ce;
896
		}
897 898
	}

899
	if (ie_modified(istate, ce, &st, options)) {
900 901
		if (err)
			*err = EINVAL;
902 903
		return NULL;
	}
904 905 906 907 908

	size = ce_size(ce);
	updated = xmalloc(size);
	memcpy(updated, ce, size);
	fill_stat_cache_info(updated, &st);
909 910 911 912 913
	/*
	 * 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.
914
	 */
915
	if (!ignore_valid && assume_unchanged &&
916 917
	    !(ce->ce_flags & CE_VALID))
		updated->ce_flags &= ~CE_VALID;
918 919 920 921

	return updated;
}

922
int refresh_index(struct index_state *istate, unsigned int flags, const char **pathspec, char *seen)
923 924 925 926 927 928 929
{
	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;
930
	unsigned int options = really ? CE_MATCH_IGNORE_VALID : 0;
931

932
	for (i = 0; i < istate->cache_nr; i++) {
933
		struct cache_entry *ce, *new;
934 935
		int cache_errno = 0;

936
		ce = istate->cache[i];
937
		if (ce_stage(ce)) {
938 939
			while ((i < istate->cache_nr) &&
			       ! strcmp(istate->cache[i]->name, ce->name))
940 941 942 943 944 945 946 947 948
				i++;
			i--;
			if (allow_unmerged)
				continue;
			printf("%s: needs merge\n", ce->name);
			has_errors = 1;
			continue;
		}

949 950 951
		if (pathspec && !match_pathspec(pathspec, ce->name, strlen(ce->name), 0, seen))
			continue;

952
		new = refresh_cache_ent(istate, ce, options, &cache_errno);
953
		if (new == ce)
954
			continue;
955 956
		if (!new) {
			if (not_new && cache_errno == ENOENT)
957
				continue;
958
			if (really && cache_errno == EINVAL) {
959 960 961
				/* If we are doing --really-refresh that
				 * means the index is not valid anymore.
				 */
962
				ce->ce_flags &= ~CE_VALID;
963
				istate->cache_changed = 1;
964 965 966 967 968 969 970
			}
			if (quiet)
				continue;
			printf("%s: needs update\n", ce->name);
			has_errors = 1;
			continue;
		}
971 972

		replace_index_entry(istate, i, new);
973 974 975 976
	}
	return has_errors;
}

977 978
struct cache_entry *refresh_cache_entry(struct cache_entry *ce, int really)
{
979
	return refresh_cache_ent(&the_index, ce, really, NULL);
980 981
}

982
static int verify_hdr(struct cache_header *hdr, unsigned long size)
983 984
{
	SHA_CTX c;
985
	unsigned char sha1[20];
986

987
	if (hdr->hdr_signature != htonl(CACHE_SIGNATURE))
988
		return error("bad signature");
989 990
	if (hdr->hdr_version != htonl(2))
		return error("bad index version");
991
	SHA1_Init(&c);
992
	SHA1_Update(&c, hdr, size - 20);
993
	SHA1_Final(sha1, &c);
994
	if (hashcmp(sha1, (unsigned char *)hdr + size - 20))
995
		return error("bad index file sha1 signature");
996 997 998
	return 0;
}

999 1000
static int read_index_extension(struct index_state *istate,
				const char *ext, void *data, unsigned long sz)
1001 1002 1003
{
	switch (CACHE_EXT(ext)) {
	case CACHE_EXT_TREE:
1004
		istate->cache_tree = cache_tree_read(data, sz);
1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015
		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;
}

1016
int read_index(struct index_state *istate)
1017
{
1018
	return read_index_from(istate, get_index_file());
1019 1020
}

1021 1022
static void convert_from_disk(struct ondisk_cache_entry *ondisk, struct cache_entry *ce)
{
1023 1024
	size_t len;

1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035
	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);
1036 1037 1038 1039 1040 1041 1042 1043 1044

	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);
1045 1046
}

1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060
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;
}

1061
/* remember to discard_cache() before reading a different cache! */
1062
int read_index_from(struct index_state *istate, const char *path)
1063 1064 1065
{
	int fd, i;
	struct stat st;
1066
	unsigned long src_offset, dst_offset;
1067
	struct cache_header *hdr;
1068 1069
	void *mmap;
	size_t mmap_size;
1070 1071

	errno = EBUSY;
1072
	if (istate->alloc)
1073
		return istate->cache_nr;
1074

1075
	errno = ENOENT;
1076
	istate->timestamp = 0;
1077
	fd = open(path, O_RDONLY);
1078 1079 1080 1081 1082
	if (fd < 0) {
		if (errno == ENOENT)
			return 0;
		die("index file open failed (%s)", strerror(errno));
	}
1083

1084
	if (fstat(fd, &st))
1085
		die("cannot stat the open index (%s)", strerror(errno));
1086 1087

	errno = EINVAL;
1088 1089
	mmap_size = xsize_t(st.st_size);
	if (mmap_size < sizeof(struct cache_header) + 20)
1090 1091
		die("index file smaller than expected");

1092
	mmap = xmmap(NULL, mmap_size, PROT_READ | PROT_WRITE, MAP_PRIVATE, fd, 0);
1093
	close(fd);
1094 1095
	if (mmap == MAP_FAILED)
		die("unable to map index file");
1096

1097 1098
	hdr = mmap;
	if (verify_hdr(hdr, mmap_size) < 0)
1099 1100
		goto unmap;

1101 1102 1103
	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 *));
1104

1105 1106 1107 1108 1109 1110
	/*
	 * 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
	 */
1111
	istate->alloc = xmalloc(estimate_cache_size(mmap_size, istate->cache_nr));
1112 1113 1114

	src_offset = sizeof(*hdr);
	dst_offset = 0;
1115
	for (i = 0; i < istate->cache_nr; i++) {
1116
		struct ondisk_cache_entry *disk_ce;
1117 1118
		struct cache_entry *ce;

1119 1120 1121
		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);
1122
		set_index_entry(istate, i, ce);
1123 1124 1125

		src_offset += ondisk_ce_size(ce);
		dst_offset += ce_size(ce);
1126
	}
1127
	istate->timestamp = st.st_mtime;
1128
	while (src_offset <= mmap_size - 20 - 8) {
1129 1130 1131 1132 1133 1134 1135
		/* 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;
1136
		memcpy(&extsize, (char *)mmap + src_offset + 4, 4);
1137
		extsize = ntohl(extsize);
1138
		if (read_index_extension(istate,
1139 1140
					 (const char *) mmap + src_offset,
					 (char *) mmap + src_offset + 8,
1141
					 extsize) < 0)
1142
			goto unmap;
1143 1144
		src_offset += 8;
		src_offset += extsize;
1145
	}
1146
	munmap(mmap, mmap_size);
1147
	return istate->cache_nr;
1148 1149

unmap:
1150
	munmap(mmap, mmap_size);
1151
	errno = EINVAL;
1152
	die("index file corrupt");
1153 1154
}

1155
int discard_index(struct index_state *istate)