read-cache.c 34.7 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

Junio C Hamano's avatar
Junio C Hamano committed
26 27 28
static void set_index_entry(struct index_state *istate, int nr, struct cache_entry *ce)
{
	istate->cache[nr] = ce;
29
	add_name_hash(istate, ce);
Junio C Hamano's avatar
Junio C Hamano committed
30 31
}

32 33 34 35
static void replace_index_entry(struct index_state *istate, int nr, struct cache_entry *ce)
{
	struct cache_entry *old = istate->cache[nr];

36
	remove_name_hash(old);
37
	set_index_entry(istate, nr, ce);
38 39 40
	istate->cache_changed = 1;
}

41 42 43 44 45 46 47
/*
 * 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)
{
48 49 50 51 52 53 54
	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
55 56

	if (assume_unchanged)
57
		ce->ce_flags |= CE_VALID;
58 59 60

	if (S_ISREG(st->st_mode))
		ce_mark_uptodate(ce);
61 62
}

Junio C Hamano's avatar
Junio C Hamano committed
63 64 65 66 67 68 69
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];
70
		if (!index_fd(sha1, fd, st, 0, OBJ_BLOB, ce->name))
71
			match = hashcmp(sha1, ce->sha1);
72
		/* index_fd() closed the file descriptor already */
Junio C Hamano's avatar
Junio C Hamano committed
73 74 75 76
	}
	return match;
}

77
static int ce_compare_link(struct cache_entry *ce, size_t expected_size)
Junio C Hamano's avatar
Junio C Hamano committed
78 79 80 81 82
{
	int match = -1;
	char *target;
	void *buffer;
	unsigned long size;
83
	enum object_type type;
Junio C Hamano's avatar
Junio C Hamano committed
84 85 86 87 88 89 90 91
	int len;

	target = xmalloc(expected_size);
	len = readlink(ce->name, target, expected_size);
	if (len != expected_size) {
		free(target);
		return -1;
	}
92
	buffer = read_sha1_file(ce->sha1, &type, &size);
Junio C Hamano's avatar
Junio C Hamano committed
93 94 95 96 97 98 99 100 101 102 103
	if (!buffer) {
		free(target);
		return -1;
	}
	if (size == expected_size)
		match = memcmp(buffer, target, size);
	free(buffer);
	free(target);
	return match;
}

104 105 106 107 108 109
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
110
	 * under GITLINK directory be a valid git directory. It
111 112 113 114 115 116 117 118 119 120
	 * 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
121 122 123 124 125 126 127 128
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:
129
		if (ce_compare_link(ce, xsize_t(st->st_size)))
Junio C Hamano's avatar
Junio C Hamano committed
130 131
			return DATA_CHANGED;
		break;
132
	case S_IFDIR:
133
		if (S_ISGITLINK(ce->ce_mode))
134
			return 0;
Junio C Hamano's avatar
Junio C Hamano committed
135 136 137 138 139 140
	default:
		return TYPE_CHANGED;
	}
	return 0;
}

Junio C Hamano's avatar
Junio C Hamano committed
141
static int ce_match_stat_basic(struct cache_entry *ce, struct stat *st)
142 143 144
{
	unsigned int changed = 0;

145 146 147 148
	if (ce->ce_flags & CE_REMOVE)
		return MODE_CHANGED | DATA_CHANGED | TYPE_CHANGED;

	switch (ce->ce_mode & S_IFMT) {
149 150
	case S_IFREG:
		changed |= !S_ISREG(st->st_mode) ? TYPE_CHANGED : 0;
Junio C Hamano's avatar
Junio C Hamano committed
151 152 153 154
		/* We consider only the owner x bit to be relevant for
		 * "mode changes"
		 */
		if (trust_executable_bit &&
155
		    (0100 & (ce->ce_mode ^ st->st_mode)))
156
			changed |= MODE_CHANGED;
157 158
		break;
	case S_IFLNK:
159 160 161
		if (!S_ISLNK(st->st_mode) &&
		    (has_symlinks || !S_ISREG(st->st_mode)))
			changed |= TYPE_CHANGED;
162
		break;
Martin Waitz's avatar
Martin Waitz committed
163
	case S_IFGITLINK:
164 165 166 167
		if (!S_ISDIR(st->st_mode))
			changed |= TYPE_CHANGED;
		else if (ce_compare_gitlink(ce))
			changed |= DATA_CHANGED;
168
		return changed;
169
	default:
170
		die("internal error: ce_mode is %o", ce->ce_mode);
171
	}
172
	if (ce->ce_mtime != (unsigned int) st->st_mtime)
173
		changed |= MTIME_CHANGED;
174
	if (ce->ce_ctime != (unsigned int) st->st_ctime)
175
		changed |= CTIME_CHANGED;
176

177 178
	if (ce->ce_uid != (unsigned int) st->st_uid ||
	    ce->ce_gid != (unsigned int) st->st_gid)
179
		changed |= OWNER_CHANGED;
180
	if (ce->ce_ino != (unsigned int) st->st_ino)
181
		changed |= INODE_CHANGED;
182 183 184 185 186 187 188

#ifdef USE_STDEV
	/*
	 * st_dev breaks on network filesystems where different
	 * clients will have different views of what "device"
	 * the filesystem is on
	 */
189
	if (ce->ce_dev != (unsigned int) st->st_dev)
190 191 192
		changed |= INODE_CHANGED;
#endif

193
	if (ce->ce_size != (unsigned int) st->st_size)
194
		changed |= DATA_CHANGED;
195

Junio C Hamano's avatar
Junio C Hamano committed
196 197 198
	return changed;
}

199
static int is_racy_timestamp(const struct index_state *istate, struct cache_entry *ce)
200
{
201 202
	return (!S_ISGITLINK(ce->ce_mode) &&
		istate->timestamp &&
203 204 205
		((unsigned int)istate->timestamp) <= ce->ce_mtime);
}

206
int ie_match_stat(const struct index_state *istate,
207 208
		  struct cache_entry *ce, struct stat *st,
		  unsigned int options)
Junio C Hamano's avatar
Junio C Hamano committed
209
{
Junio C Hamano's avatar
Junio C Hamano committed
210
	unsigned int changed;
211 212
	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
213 214 215 216 217

	/*
	 * If it's marked as always valid in the index, it's
	 * valid whatever the checked-out copy says.
	 */
218
	if (!ignore_valid && (ce->ce_flags & CE_VALID))
Junio C Hamano's avatar
Junio C Hamano committed
219 220 221
		return 0;

	changed = ce_match_stat_basic(ce, st);
Junio C Hamano's avatar
Junio C Hamano committed
222

Junio C Hamano's avatar
Junio C Hamano committed
223 224 225 226 227 228 229 230 231 232 233 234 235 236
	/*
	 * 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
237
	 * carefully than others.
Junio C Hamano's avatar
Junio C Hamano committed
238
	 */
239
	if (!changed && is_racy_timestamp(istate, ce)) {
240 241 242 243 244
		if (assume_racy_is_modified)
			changed |= DATA_CHANGED;
		else
			changed |= ce_modified_check_fs(ce, st);
	}
245

Junio C Hamano's avatar
Junio C Hamano committed
246
	return changed;
247 248
}

249
int ie_modified(const struct index_state *istate,
250
		struct cache_entry *ce, struct stat *st, unsigned int options)
251
{
Junio C Hamano's avatar
Junio C Hamano committed
252
	int changed, changed_fs;
253 254

	changed = ie_match_stat(istate, ce, st, options);
255 256 257 258 259 260 261 262 263 264 265 266 267
	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.
	 */
268
	if ((changed & DATA_CHANGED) && ce->ce_size != 0)
269 270
		return changed;

Junio C Hamano's avatar
Junio C Hamano committed
271 272 273
	changed_fs = ce_modified_check_fs(ce, st);
	if (changed_fs)
		return changed | changed_fs;
274 275 276
	return 0;
}

277 278 279 280 281 282 283 284 285 286 287 288
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];
289
	if (!c1 && S_ISDIR(mode1))
290
		c1 = '/';
291
	if (!c2 && S_ISDIR(mode2))
292 293 294 295
		c2 = '/';
	return (c1 < c2) ? -1 : (c1 > c2) ? 1 : 0;
}

296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330
/*
 * df_name_compare() is identical to base_name_compare(), except it
 * compares conflicting directory/file entries as equal. Note that
 * while a directory name compares as equal to a regular file, they
 * then individually compare _differently_ to a filename that has
 * a dot after the basename (because '\0' < '.' < '/').
 *
 * This is used by routines that want to traverse the git namespace
 * but then handle conflicting entries together when possible.
 */
int df_name_compare(const char *name1, int len1, int mode1,
		    const char *name2, int len2, int mode2)
{
	int len = len1 < len2 ? len1 : len2, cmp;
	unsigned char c1, c2;

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

331
int cache_name_compare(const char *name1, int flags1, const char *name2, int flags2)
332
{
333 334
	int len1 = flags1 & CE_NAMEMASK;
	int len2 = flags2 & CE_NAMEMASK;
335 336 337 338 339 340 341 342 343 344
	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
345

346 347 348
	/* Compare stages  */
	flags1 &= CE_STAGEMASK;
	flags2 &= CE_STAGEMASK;
Junio C Hamano's avatar
Junio C Hamano committed
349

350 351 352 353
	if (flags1 < flags2)
		return -1;
	if (flags1 > flags2)
		return 1;
354 355 356
	return 0;
}

357
int index_name_pos(const struct index_state *istate, const char *name, int namelen)
358 359 360 361
{
	int first, last;

	first = 0;
362
	last = istate->cache_nr;
363 364
	while (last > first) {
		int next = (last + first) >> 1;
365
		struct cache_entry *ce = istate->cache[next];
366
		int cmp = cache_name_compare(name, namelen, ce->name, ce->ce_flags);
367
		if (!cmp)
368
			return next;
369 370 371 372 373 374
		if (cmp < 0) {
			last = next;
			continue;
		}
		first = next+1;
	}
375
	return -first-1;
376 377
}

378
/* Remove entry, return true if there are more entries to go.. */
379
int remove_index_entry_at(struct index_state *istate, int pos)
380
{
381 382
	struct cache_entry *ce = istate->cache[pos];

383
	remove_name_hash(ce);
384 385 386
	istate->cache_changed = 1;
	istate->cache_nr--;
	if (pos >= istate->cache_nr)
387
		return 0;
388 389 390
	memmove(istate->cache + pos,
		istate->cache + pos + 1,
		(istate->cache_nr - pos) * sizeof(struct cache_entry *));
391 392 393
	return 1;
}

394
int remove_file_from_index(struct index_state *istate, const char *path)
395
{
396
	int pos = index_name_pos(istate, path, strlen(path));
397 398
	if (pos < 0)
		pos = -pos-1;
Junio C Hamano's avatar
Junio C Hamano committed
399
	cache_tree_invalidate_path(istate->cache_tree, path);
400 401
	while (pos < istate->cache_nr && !strcmp(istate->cache[pos]->name, path))
		remove_index_entry_at(istate, pos);
402 403 404
	return 0;
}

405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432
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;
}

433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464
static int different_name(struct cache_entry *ce, struct cache_entry *alias)
{
	int len = ce_namelen(ce);
	return ce_namelen(alias) != len || memcmp(ce->name, alias->name, len);
}

/*
 * If we add a filename that aliases in the cache, we will use the
 * name that we already have - but we don't want to update the same
 * alias twice, because that implies that there were actually two
 * different files with aliasing names!
 *
 * So we use the CE_ADDED flag to verify that the alias was an old
 * one before we accept it as
 */
static struct cache_entry *create_alias_ce(struct cache_entry *ce, struct cache_entry *alias)
{
	int len;
	struct cache_entry *new;

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

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

465
int add_to_index(struct index_state *istate, const char *path, struct stat *st, int verbose)
Johannes Schindelin's avatar
Johannes Schindelin committed
466
{
467
	int size, namelen;
468
	mode_t st_mode = st->st_mode;
469
	struct cache_entry *ce, *alias;
470
	unsigned ce_option = CE_MATCH_IGNORE_VALID|CE_MATCH_RACY_IS_DIRTY;
Johannes Schindelin's avatar
Johannes Schindelin committed
471

472
	if (!S_ISREG(st_mode) && !S_ISLNK(st_mode) && !S_ISDIR(st_mode))
473
		die("%s: can only add regular files, symbolic links or git-directories", path);
Johannes Schindelin's avatar
Johannes Schindelin committed
474 475

	namelen = strlen(path);
476
	if (S_ISDIR(st_mode)) {
477 478 479
		while (namelen && path[namelen-1] == '/')
			namelen--;
	}
Johannes Schindelin's avatar
Johannes Schindelin committed
480 481 482
	size = cache_entry_size(namelen);
	ce = xcalloc(1, size);
	memcpy(ce->name, path, namelen);
483
	ce->ce_flags = namelen;
484
	fill_stat_cache_info(ce, st);
Johannes Schindelin's avatar
Johannes Schindelin committed
485

486
	if (trust_executable_bit && has_symlinks)
487
		ce->ce_mode = create_ce_mode(st_mode);
488
	else {
489 490
		/* 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
491
		 */
492
		struct cache_entry *ent;
493
		int pos = index_name_pos_also_unmerged(istate, path, namelen);
494

495
		ent = (0 <= pos) ? istate->cache[pos] : NULL;
496
		ce->ce_mode = ce_mode_from_stat(ent, st_mode);
Johannes Schindelin's avatar
Johannes Schindelin committed
497 498
	}

499
	alias = index_name_exists(istate, ce->name, ce_namelen(ce), ignore_case);
500
	if (alias && !ce_stage(alias) && !ie_match_stat(istate, alias, st, ce_option)) {
501 502
		/* Nothing changed, really */
		free(ce);
503
		ce_mark_uptodate(alias);
504
		alias->ce_flags |= CE_ADDED;
505 506
		return 0;
	}
507
	if (index_path(ce->sha1, path, st, 1))
Johannes Schindelin's avatar
Johannes Schindelin committed
508
		die("unable to index file %s", path);
509 510 511
	if (ignore_case && alias && different_name(ce, alias))
		ce = create_alias_ce(ce, alias);
	ce->ce_flags |= CE_ADDED;
512
	if (add_index_entry(istate, ce, ADD_CACHE_OK_TO_ADD|ADD_CACHE_OK_TO_REPLACE))
Johannes Schindelin's avatar
Johannes Schindelin committed
513 514 515 516 517 518
		die("unable to add %s to index",path);
	if (verbose)
		printf("add '%s'\n", path);
	return 0;
}

519 520 521 522 523 524 525 526
int add_file_to_index(struct index_state *istate, const char *path, int verbose)
{
	struct stat st;
	if (lstat(path, &st))
		die("%s: unable to stat (%s)", path, strerror(errno));
	return add_to_index(istate, path, &st, verbose);
}

527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551
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;
}

552
int ce_same_name(struct cache_entry *a, struct cache_entry *b)
553 554 555 556 557
{
	int len = ce_namelen(a);
	return ce_namelen(b) == len && !memcmp(a->name, b->name, len);
}

558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577
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;
578 579
		if (!matchlen)
			return 1;
580 581 582 583
	}
	return 0;
}

584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647
/*
 * 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++;
	}
}

648 649 650
/*
 * Do we have another file that has the beginning components being a
 * proper superset of the name we're trying to add?
651
 */
652 653
static int has_file_name(struct index_state *istate,
			 const struct cache_entry *ce, int pos, int ok_to_replace)
654
{
655 656
	int retval = 0;
	int len = ce_namelen(ce);
657
	int stage = ce_stage(ce);
658
	const char *name = ce->name;
659

660 661
	while (pos < istate->cache_nr) {
		struct cache_entry *p = istate->cache[pos++];
662

663
		if (len >= ce_namelen(p))
664
			break;
665 666
		if (memcmp(name, p->name, len))
			break;
667 668
		if (ce_stage(p) != stage)
			continue;
669 670
		if (p->name[len] != '/')
			continue;
671
		if (p->ce_flags & CE_REMOVE)
672
			continue;
673 674 675
		retval = -1;
		if (!ok_to_replace)
			break;
676
		remove_index_entry_at(istate, --pos);
677
	}
678 679
	return retval;
}
680

681 682 683 684
/*
 * Do we have another file with a pathname that is a proper
 * subset of the name we're trying to add?
 */
685 686
static int has_dir_name(struct index_state *istate,
			const struct cache_entry *ce, int pos, int ok_to_replace)
687 688
{
	int retval = 0;
689
	int stage = ce_stage(ce);
690 691
	const char *name = ce->name;
	const char *slash = name + ce_namelen(ce);
692

693 694
	for (;;) {
		int len;
695

696 697 698 699 700 701 702
		for (;;) {
			if (*--slash == '/')
				break;
			if (slash <= ce->name)
				return retval;
		}
		len = slash - name;
703

704
		pos = index_name_pos(istate, name, create_ce_flags(len, stage));
705
		if (pos >= 0) {
706 707 708 709 710 711 712 713
			/*
			 * 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.
			 */
714
			if (!(istate->cache[pos]->ce_flags & CE_REMOVE)) {
715 716 717
				retval = -1;
				if (!ok_to_replace)
					break;
718
				remove_index_entry_at(istate, pos);
719 720
				continue;
			}
721
		}
722 723
		else
			pos = -pos-1;
724 725 726 727

		/*
		 * Trivial optimization: if we find an entry that
		 * already matches the sub-directory, then we know
728
		 * we're ok, and we can exit.
729
		 */
730 731
		while (pos < istate->cache_nr) {
			struct cache_entry *p = istate->cache[pos];
732 733 734 735
			if ((ce_namelen(p) <= len) ||
			    (p->name[len] != '/') ||
			    memcmp(p->name, name, len))
				break; /* not our subdirectory */
736 737 738
			if (ce_stage(p) == stage && !(p->ce_flags & CE_REMOVE))
				/*
				 * p is at the same stage as our entry, and
739 740 741 742 743 744
				 * is a subdirectory of what we are looking
				 * at, so we cannot have conflicts at our
				 * level or anything shorter.
				 */
				return retval;
			pos++;
745
		}
746
	}
747 748 749 750 751 752 753
	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
754
 *
755 756 757 758
 * 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.
 */
759 760 761
static int check_file_directory_conflict(struct index_state *istate,
					 const struct cache_entry *ce,
					 int pos, int ok_to_replace)
762
{
763 764 765 766 767
	int retval;

	/*
	 * When ce is an "I am going away" entry, we allow it to be added
	 */
768
	if (ce->ce_flags & CE_REMOVE)
769 770
		return 0;

771 772 773
	/*
	 * We check if the path is a sub-path of a subsequent pathname
	 * first, since removing those will not change the position
774
	 * in the array.
775
	 */
776
	retval = has_file_name(istate, ce, pos, ok_to_replace);
777

778 779 780 781
	/*
	 * Then check if the path might have a clashing sub-directory
	 * before it.
	 */
782
	return retval + has_dir_name(istate, ce, pos, ok_to_replace);
783 784
}

785
static int add_index_entry_with_check(struct index_state *istate, struct cache_entry *ce, int option)
786 787
{
	int pos;
788 789
	int ok_to_add = option & ADD_CACHE_OK_TO_ADD;
	int ok_to_replace = option & ADD_CACHE_OK_TO_REPLACE;
790
	int skip_df_check = option & ADD_CACHE_SKIP_DFCHECK;
Junio C Hamano's avatar
Junio C Hamano committed
791

Junio C Hamano's avatar
Junio C Hamano committed
792
	cache_tree_invalidate_path(istate->cache_tree, ce->name);
793
	pos = index_name_pos(istate, ce->name, ce->ce_flags);
794

Junio C Hamano's avatar
Junio C Hamano committed
795
	/* existing match? Just replace it. */
796
	if (pos >= 0) {
797
		replace_index_entry(istate, pos, ce);
798 799
		return 0;
	}
800
	pos = -pos-1;
801

802 803 804 805
	/*
	 * Inserting a merged entry ("stage 0") into the index
	 * will always replace all non-merged entries..
	 */
806 807
	if (pos < istate->cache_nr && ce_stage(ce) == 0) {
		while (ce_same_name(istate->cache[pos], ce)) {
808
			ok_to_add = 1;
809
			if (!remove_index_entry_at(istate, pos))
810 811 812 813
				break;
		}
	}

814 815
	if (!ok_to_add)
		return -1;
816 817
	if (!verify_path(ce->name))
		return -1;
818

Junio C Hamano's avatar
Junio C Hamano committed
819
	if (!skip_df_check &&
820
	    check_file_directory_conflict(istate, ce, pos, ok_to_replace)) {
821
		if (!ok_to_replace)
822 823
			return error("'%s' appears as both a file and as a directory",
				     ce->name);
824
		pos = index_name_pos(istate, ce->name, ce->ce_flags);
825 826
		pos = -pos-1;
	}
827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842
	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;
	}
843

844
	/* Make sure the array is big enough .. */
845 846 847 848
	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 *));
849 850 851
	}

	/* Add it in.. */
852
	istate->cache_nr++;
853
	if (istate->cache_nr > pos + 1)
854 855 856
		memmove(istate->cache + pos + 1,
			istate->cache + pos,
			(istate->cache_nr - pos - 1) * sizeof(ce));
857
	set_index_entry(istate, pos, ce);
858
	istate->cache_changed = 1;
859 860 861
	return 0;
}

862 863 864 865 866 867 868 869 870 871 872
/*
 * "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.
 */
873
static struct cache_entry *refresh_cache_ent(struct index_state *istate,
874 875
					     struct cache_entry *ce,
					     unsigned int options, int *err)
876 877 878 879
{
	struct stat st;
	struct cache_entry *updated;
	int changed, size;
880
	int ignore_valid = options & CE_MATCH_IGNORE_VALID;
881

882 883 884
	if (ce_uptodate(ce))
		return ce;

885
	if (lstat(ce->name, &st) < 0) {
886 887
		if (err)
			*err = errno;
888 889
		return NULL;
	}
890

891
	changed = ie_match_stat(istate, ce, &st, options);
892
	if (!changed) {
893 894 895 896 897 898 899 900
		/*
		 * 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 &&
901
		    !(ce->ce_flags & CE_VALID))
902
			; /* mark this one VALID again */
903 904 905 906 907 908 909
		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);
910
			return ce;
911
		}
912 913
	}

914
	if (ie_modified(istate, ce, &st, options)) {
915 916
		if (err)
			*err = EINVAL;
917 918
		return NULL;
	}
919 920 921 922 923

	size = ce_size(ce);
	updated = xmalloc(size);
	memcpy(updated, ce, size);
	fill_stat_cache_info(updated, &st);
924 925 926 927 928
	/*
	 * 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.
929
	 */
930
	if (!ignore_valid && assume_unchanged &&
931 932
	    !(ce->ce_flags & CE_VALID))
		updated->ce_flags &= ~CE_VALID;
933 934 935 936

	return updated;
}

937
int refresh_index(struct index_state *istate, unsigned int flags, const char **pathspec, char *seen)
938 939 940 941 942 943 944
{
	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;
945
	unsigned int options = really ? CE_MATCH_IGNORE_VALID : 0;
946

947
	for (i = 0; i < istate->cache_nr; i++) {
948
		struct cache_entry *ce, *new;
949 950
		int cache_errno = 0;

951
		ce = istate->cache[i];
952
		if (ce_stage(ce)) {
953 954
			while ((i < istate->cache_nr) &&
			       ! strcmp(istate->cache[i]->name, ce->name))
955 956 957 958 959 960 961 962 963
				i++;
			i--;
			if (allow_unmerged)
				continue;
			printf("%s: needs merge\n", ce->name);
			has_errors = 1;
			continue;
		}

964 965 966
		if (pathspec && !match_pathspec(pathspec, ce->name, strlen(ce->name), 0, seen))
			continue;

967
		new = refresh_cache_ent(istate, ce, options, &cache_errno);
968
		if (new == ce)
969
			continue;
970 971
		if (!new) {
			if (not_new && cache_errno == ENOENT)
972
				continue;
973
			if (really && cache_errno == EINVAL) {
974 975 976
				/* If we are doing --really-refresh that
				 * means the index is not valid anymore.
				 */
977
				ce->ce_flags &= ~CE_VALID;
978
				istate->cache_changed = 1;
979 980 981 982 983 984 985
			}
			if (quiet)
				continue;
			printf("%s: needs update\n", ce->name);
			has_errors = 1;
			continue;
		}
986 987

		replace_index_entry(istate, i, new);
988 989 990 991
	}
	return has_errors;
}

992 993
struct cache_entry *refresh_cache_entry(struct cache_entry *ce, int really)
{
994
	return refresh_cache_ent(&the_index, ce, really, NULL);
995 996
}

997
static int verify_hdr(struct cache_header *hdr, unsigned long size)
998 999
{
	SHA_CTX c;
1000
	unsigned char sha1[20];
1001

1002
	if (hdr->hdr_signature != htonl(CACHE_SIGNATURE))
1003
		return error("bad signature");
1004 1005
	if (hdr->hdr_version != htonl(2))
		return error("bad index version");
1006
	SHA1_Init(&c);
1007
	SHA1_Update(&c, hdr, size - 20);
1008
	SHA1_Final(sha1, &c);
1009
	if (hashcmp(sha1, (unsigned char *)hdr + size - 20))
1010
		return error("bad index file sha1 signature");
1011 1012 1013
	return 0;
}

1014 1015
static int read_index_extension(struct index_state *istate,
				const char *ext, void *data, unsigned long sz)
1016 1017 1018
{
	switch (CACHE_EXT(ext)) {
	case CACHE_EXT_TREE:
1019
		istate->cache_tree = cache_tree_read(data, sz);
1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030
		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;
}

1031
int read_index(struct index_state *istate)
1032
{
1033
	return read_index_from(istate, get_index_file());
1034 1035
}

1036 1037
static void convert_from_disk(struct ondisk_cache_entry *ondisk, struct cache_entry *ce)
{
1038 1039
	size_t len;

1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050
	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);
1051 1052 1053 1054 1055 1056 1057 1058 1059

	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);
1060 1061
}

1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075
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;
}

1076
/* remember to discard_cache() before reading a different cache! */
1077
int read_index_from(struct index_state *istate, const char *path)
1078 1079 1080
{
	int fd, i;
	struct stat st;
1081
	unsigned long src_offset, dst_offset;
1082
	struct cache_header *hdr;
1083 1084
	void *mmap;
	size_t mmap_size;
1085 1086

	errno = EBUSY;
1087
	if (istate->alloc)
1088
		return istate->cache_nr;
1089

1090
	errno = ENOENT;
1091
	istate->timestamp = 0;
1092
	fd = open(path, O_RDONLY);
1093 1094 1095 1096 1097
	if (fd < 0) {
		if (errno == ENOENT)
			return 0;
		die("index file open failed (%s)", strerror(errno));
	}
1098

1099
	if (fstat(fd, &st))
1100
		die("cannot stat the open index (%s)", strerror(errno));
1101 1102

	errno = EINVAL;
1103 1104
	mmap_size = xsize_t(st.st_size);
	if (mmap_size < sizeof(struct cache_header) + 20)
1105 1106
		die("index file smaller than expected");

1107
	mmap = xmmap(NULL, mmap_size, PROT_READ | PROT_WRITE, MAP_PRIVATE, fd, 0);
1108
	close(fd);
1109 1110
	if (mmap == MAP_FAILED)
		die("unable to map index file");
1111

1112 1113
	hdr = mmap;
	if (verify_hdr(hdr, mmap_size) < 0)
1114 1115
		goto unmap;

1116 1117 1118
	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 *));
1119

1120 1121 1122 1123 1124 1125
	/*
	 * 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
	 */
1126
	istate->alloc = xmalloc(estimate_cache_size(mmap_size, istate->cache_nr));
1127 1128 1129

	src_offset = sizeof(*hdr);
	dst_offset = 0;
1130
	for (i = 0; i < istate->cache_nr; i++) {
1131
		struct ondisk_cache_entry *disk_ce;