read-cache.c 62.5 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 "tempfile.h"
9
#include "lockfile.h"
10
#include "cache-tree.h"
11
#include "refs.h"
12
#include "dir.h"
13 14
#include "tree.h"
#include "commit.h"
15
#include "blob.h"
16
#include "resolve-undo.h"
17 18
#include "strbuf.h"
#include "varint.h"
Duy Nguyen's avatar
Duy Nguyen committed
19
#include "split-index.h"
20
#include "utf8.h"
21

22 23 24 25
/* Mask for the name length in ce_flags in the on-disk index */

#define CE_NAMEMASK  (0x0fff)

26 27 28 29 30 31 32 33 34 35 36
/* 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" */
37
#define CACHE_EXT_RESOLVE_UNDO 0x52455543 /* "REUC" */
Duy Nguyen's avatar
Duy Nguyen committed
38
#define CACHE_EXT_LINK 0x6c696e6b	  /* "link" */
39
#define CACHE_EXT_UNTRACKED 0x554E5452	  /* "UNTR" */
Duy Nguyen's avatar
Duy Nguyen committed
40 41

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

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

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

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

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

67 68 69 70 71 72 73
void rename_index_entry_at(struct index_state *istate, int nr, const char *new_name)
{
	struct cache_entry *old = istate->cache[nr], *new;
	int namelen = strlen(new_name);

	new = xmalloc(cache_entry_size(namelen));
	copy_cache_entry(new, old);
74
	new->ce_flags &= ~CE_HASHED;
75
	new->ce_namelen = namelen;
Duy Nguyen's avatar
Duy Nguyen committed
76
	new->index = 0;
77 78
	memcpy(new->name, new_name, namelen + 1);

79
	cache_tree_invalidate_path(istate, old->name);
80
	untracked_cache_remove_from_index(istate, old->name);
81 82 83 84
	remove_index_entry_at(istate, nr);
	add_index_entry(istate, new, ADD_CACHE_OK_TO_ADD|ADD_CACHE_OK_TO_REPLACE);
}

85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139
void fill_stat_data(struct stat_data *sd, struct stat *st)
{
	sd->sd_ctime.sec = (unsigned int)st->st_ctime;
	sd->sd_mtime.sec = (unsigned int)st->st_mtime;
	sd->sd_ctime.nsec = ST_CTIME_NSEC(*st);
	sd->sd_mtime.nsec = ST_MTIME_NSEC(*st);
	sd->sd_dev = st->st_dev;
	sd->sd_ino = st->st_ino;
	sd->sd_uid = st->st_uid;
	sd->sd_gid = st->st_gid;
	sd->sd_size = st->st_size;
}

int match_stat_data(const struct stat_data *sd, struct stat *st)
{
	int changed = 0;

	if (sd->sd_mtime.sec != (unsigned int)st->st_mtime)
		changed |= MTIME_CHANGED;
	if (trust_ctime && check_stat &&
	    sd->sd_ctime.sec != (unsigned int)st->st_ctime)
		changed |= CTIME_CHANGED;

#ifdef USE_NSEC
	if (check_stat && sd->sd_mtime.nsec != ST_MTIME_NSEC(*st))
		changed |= MTIME_CHANGED;
	if (trust_ctime && check_stat &&
	    sd->sd_ctime.nsec != ST_CTIME_NSEC(*st))
		changed |= CTIME_CHANGED;
#endif

	if (check_stat) {
		if (sd->sd_uid != (unsigned int) st->st_uid ||
			sd->sd_gid != (unsigned int) st->st_gid)
			changed |= OWNER_CHANGED;
		if (sd->sd_ino != (unsigned int) st->st_ino)
			changed |= INODE_CHANGED;
	}

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

	if (sd->sd_size != (unsigned int) st->st_size)
		changed |= DATA_CHANGED;

	return changed;
}

140 141 142 143 144 145 146
/*
 * 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)
{
147
	fill_stat_data(&ce->ce_stat_data, st);
Junio C Hamano's avatar
Junio C Hamano committed
148 149

	if (assume_unchanged)
150
		ce->ce_flags |= CE_VALID;
151 152 153

	if (S_ISREG(st->st_mode))
		ce_mark_uptodate(ce);
154 155
}

156
static int ce_compare_data(const struct cache_entry *ce, struct stat *st)
Junio C Hamano's avatar
Junio C Hamano committed
157 158
{
	int match = -1;
159 160 161 162 163 164 165 166
	static int cloexec = O_CLOEXEC;
	int fd = open(ce->name, O_RDONLY | cloexec);

	if ((cloexec & O_CLOEXEC) && fd < 0 && errno == EINVAL) {
		/* Try again w/o O_CLOEXEC: the kernel might not support it */
		cloexec &= ~O_CLOEXEC;
		fd = open(ce->name, O_RDONLY | cloexec);
	}
Junio C Hamano's avatar
Junio C Hamano committed
167 168 169

	if (fd >= 0) {
		unsigned char sha1[20];
170
		if (!index_fd(sha1, fd, st, OBJ_BLOB, ce->name, 0))
171
			match = hashcmp(sha1, ce->oid.hash);
172
		/* index_fd() closed the file descriptor already */
Junio C Hamano's avatar
Junio C Hamano committed
173 174 175 176
	}
	return match;
}

177
static int ce_compare_link(const struct cache_entry *ce, size_t expected_size)
Junio C Hamano's avatar
Junio C Hamano committed
178 179 180 181
{
	int match = -1;
	void *buffer;
	unsigned long size;
182
	enum object_type type;
183
	struct strbuf sb = STRBUF_INIT;
Junio C Hamano's avatar
Junio C Hamano committed
184

185
	if (strbuf_readlink(&sb, ce->name, expected_size))
Junio C Hamano's avatar
Junio C Hamano committed
186
		return -1;
187

188
	buffer = read_sha1_file(ce->oid.hash, &type, &size);
189 190 191 192
	if (buffer) {
		if (size == sb.len)
			match = memcmp(buffer, sb.buf, size);
		free(buffer);
Junio C Hamano's avatar
Junio C Hamano committed
193
	}
194
	strbuf_release(&sb);
Junio C Hamano's avatar
Junio C Hamano committed
195 196 197
	return match;
}

198
static int ce_compare_gitlink(const struct cache_entry *ce)
199 200 201 202 203
{
	unsigned char sha1[20];

	/*
	 * We don't actually require that the .git directory
Martin Waitz's avatar
Martin Waitz committed
204
	 * under GITLINK directory be a valid git directory. It
205 206 207 208 209 210 211
	 * 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;
212
	return hashcmp(sha1, ce->oid.hash);
213 214
}

215
static int ce_modified_check_fs(const struct cache_entry *ce, struct stat *st)
Junio C Hamano's avatar
Junio C Hamano committed
216 217 218 219 220 221 222
{
	switch (st->st_mode & S_IFMT) {
	case S_IFREG:
		if (ce_compare_data(ce, st))
			return DATA_CHANGED;
		break;
	case S_IFLNK:
223
		if (ce_compare_link(ce, xsize_t(st->st_size)))
Junio C Hamano's avatar
Junio C Hamano committed
224 225
			return DATA_CHANGED;
		break;
226
	case S_IFDIR:
227
		if (S_ISGITLINK(ce->ce_mode))
228
			return ce_compare_gitlink(ce) ? DATA_CHANGED : 0;
Junio C Hamano's avatar
Junio C Hamano committed
229 230 231 232 233 234
	default:
		return TYPE_CHANGED;
	}
	return 0;
}

235
static int ce_match_stat_basic(const struct cache_entry *ce, struct stat *st)
236 237 238
{
	unsigned int changed = 0;

239 240 241 242
	if (ce->ce_flags & CE_REMOVE)
		return MODE_CHANGED | DATA_CHANGED | TYPE_CHANGED;

	switch (ce->ce_mode & S_IFMT) {
243 244
	case S_IFREG:
		changed |= !S_ISREG(st->st_mode) ? TYPE_CHANGED : 0;
Junio C Hamano's avatar
Junio C Hamano committed
245 246 247 248
		/* We consider only the owner x bit to be relevant for
		 * "mode changes"
		 */
		if (trust_executable_bit &&
249
		    (0100 & (ce->ce_mode ^ st->st_mode)))
250
			changed |= MODE_CHANGED;
251 252
		break;
	case S_IFLNK:
253 254 255
		if (!S_ISLNK(st->st_mode) &&
		    (has_symlinks || !S_ISREG(st->st_mode)))
			changed |= TYPE_CHANGED;
256
		break;
Martin Waitz's avatar
Martin Waitz committed
257
	case S_IFGITLINK:
258
		/* We ignore most of the st_xxx fields for gitlinks */
259 260 261 262
		if (!S_ISDIR(st->st_mode))
			changed |= TYPE_CHANGED;
		else if (ce_compare_gitlink(ce))
			changed |= DATA_CHANGED;
263
		return changed;
264
	default:
265
		die("internal error: ce_mode is %o", ce->ce_mode);
266
	}
267

268
	changed |= match_stat_data(&ce->ce_stat_data, st);
269

270
	/* Racily smudged entry? */
271
	if (!ce->ce_stat_data.sd_size) {
272
		if (!is_empty_blob_sha1(ce->oid.hash))
273 274 275
			changed |= DATA_CHANGED;
	}

Junio C Hamano's avatar
Junio C Hamano committed
276 277 278
	return changed;
}

279 280
static int is_racy_stat(const struct index_state *istate,
			const struct stat_data *sd)
281
{
282
	return (istate->timestamp.sec &&
283 284
#ifdef USE_NSEC
		 /* nanosecond timestamped files can also be racy! */
285 286 287
		(istate->timestamp.sec < sd->sd_mtime.sec ||
		 (istate->timestamp.sec == sd->sd_mtime.sec &&
		  istate->timestamp.nsec <= sd->sd_mtime.nsec))
288
#else
289
		istate->timestamp.sec <= sd->sd_mtime.sec
290
#endif
291 292 293 294 295 296 297 298
		);
}

static int is_racy_timestamp(const struct index_state *istate,
			     const struct cache_entry *ce)
{
	return (!S_ISGITLINK(ce->ce_mode) &&
		is_racy_stat(istate, &ce->ce_stat_data));
299 300
}

301 302 303 304 305 306
int match_stat_data_racy(const struct index_state *istate,
			 const struct stat_data *sd, struct stat *st)
{
	if (is_racy_stat(istate, sd))
		return MTIME_CHANGED;
	return match_stat_data(sd, st);
307 308
}

309
int ie_match_stat(const struct index_state *istate,
310
		  const struct cache_entry *ce, struct stat *st,
311
		  unsigned int options)
Junio C Hamano's avatar
Junio C Hamano committed
312
{
Junio C Hamano's avatar
Junio C Hamano committed
313
	unsigned int changed;
314
	int ignore_valid = options & CE_MATCH_IGNORE_VALID;
315
	int ignore_skip_worktree = options & CE_MATCH_IGNORE_SKIP_WORKTREE;
316
	int assume_racy_is_modified = options & CE_MATCH_RACY_IS_DIRTY;
Junio C Hamano's avatar
Junio C Hamano committed
317 318 319 320

	/*
	 * If it's marked as always valid in the index, it's
	 * valid whatever the checked-out copy says.
321 322
	 *
	 * skip-worktree has the same effect with higher precedence
Junio C Hamano's avatar
Junio C Hamano committed
323
	 */
324 325
	if (!ignore_skip_worktree && ce_skip_worktree(ce))
		return 0;
326
	if (!ignore_valid && (ce->ce_flags & CE_VALID))
Junio C Hamano's avatar
Junio C Hamano committed
327 328
		return 0;

329 330 331 332 333
	/*
	 * Intent-to-add entries have not been added, so the index entry
	 * by definition never matches what is in the work tree until it
	 * actually gets added.
	 */
334
	if (ce_intent_to_add(ce))
335 336
		return DATA_CHANGED | TYPE_CHANGED | MODE_CHANGED;

Junio C Hamano's avatar
Junio C Hamano committed
337
	changed = ce_match_stat_basic(ce, st);
Junio C Hamano's avatar
Junio C Hamano committed
338

Junio C Hamano's avatar
Junio C Hamano committed
339 340 341 342 343 344 345 346 347 348 349 350 351 352
	/*
	 * 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
353
	 * carefully than others.
Junio C Hamano's avatar
Junio C Hamano committed
354
	 */
355
	if (!changed && is_racy_timestamp(istate, ce)) {
356 357 358 359 360
		if (assume_racy_is_modified)
			changed |= DATA_CHANGED;
		else
			changed |= ce_modified_check_fs(ce, st);
	}
361

Junio C Hamano's avatar
Junio C Hamano committed
362
	return changed;
363 364
}

365
int ie_modified(const struct index_state *istate,
366 367
		const struct cache_entry *ce,
		struct stat *st, unsigned int options)
368
{
Junio C Hamano's avatar
Junio C Hamano committed
369
	int changed, changed_fs;
370 371

	changed = ie_match_stat(istate, ce, st, options);
372 373 374 375 376 377 378 379 380
	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;

381 382 383 384 385 386 387 388 389 390 391 392 393
	/*
	 * Immediately after read-tree or update-index --cacheinfo,
	 * the length field is zero, as we have never even read the
	 * lstat(2) information once, and we cannot trust DATA_CHANGED
	 * returned by ie_match_stat() which in turn was returned by
	 * ce_match_stat_basic() to signal that the filesize of the
	 * blob changed.  We have to actually go to the filesystem to
	 * see if the contents match, and if so, should answer "unchanged".
	 *
	 * The logic does not apply to gitlinks, as ce_match_stat_basic()
	 * already has checked the actual HEAD from the filesystem in the
	 * subproject.  If ie_match_stat() already said it is different,
	 * then we know it is.
394
	 */
395
	if ((changed & DATA_CHANGED) &&
396
	    (S_ISGITLINK(ce->ce_mode) || ce->ce_stat_data.sd_size != 0))
397 398
		return changed;

Junio C Hamano's avatar
Junio C Hamano committed
399 400 401
	changed_fs = ce_modified_check_fs(ce, st);
	if (changed_fs)
		return changed | changed_fs;
402 403 404
	return 0;
}

405 406 407 408 409 410 411 412 413 414 415 416
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];
417
	if (!c1 && S_ISDIR(mode1))
418
		c1 = '/';
419
	if (!c2 && S_ISDIR(mode2))
420 421 422 423
		c2 = '/';
	return (c1 < c2) ? -1 : (c1 > c2) ? 1 : 0;
}

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
/*
 * 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;
}

459
int name_compare(const char *name1, size_t len1, const char *name2, size_t len2)
460
{
461 462
	size_t min_len = (len1 < len2) ? len1 : len2;
	int cmp = memcmp(name1, name2, min_len);
463 464 465 466 467 468
	if (cmp)
		return cmp;
	if (len1 < len2)
		return -1;
	if (len1 > len2)
		return 1;
469 470 471 472 473 474 475 476 477 478
	return 0;
}

int cache_name_stage_compare(const char *name1, int len1, int stage1, const char *name2, int len2, int stage2)
{
	int cmp;

	cmp = name_compare(name1, len1, name2, len2);
	if (cmp)
		return cmp;
Junio C Hamano's avatar
Junio C Hamano committed
479

480
	if (stage1 < stage2)
481
		return -1;
482
	if (stage1 > stage2)
483
		return 1;
484 485 486
	return 0;
}

487
static int index_name_stage_pos(const struct index_state *istate, const char *name, int namelen, int stage)
488 489 490 491
{
	int first, last;

	first = 0;
492
	last = istate->cache_nr;
493 494
	while (last > first) {
		int next = (last + first) >> 1;
495
		struct cache_entry *ce = istate->cache[next];
496
		int cmp = cache_name_stage_compare(name, namelen, stage, ce->name, ce_namelen(ce), ce_stage(ce));
497
		if (!cmp)
498
			return next;
499 500 501 502 503 504
		if (cmp < 0) {
			last = next;
			continue;
		}
		first = next+1;
	}
505
	return -first-1;
506 507
}

508 509 510 511 512
int index_name_pos(const struct index_state *istate, const char *name, int namelen)
{
	return index_name_stage_pos(istate, name, namelen, 0);
}

513
/* Remove entry, return true if there are more entries to go.. */
514
int remove_index_entry_at(struct index_state *istate, int pos)
515
{
516 517
	struct cache_entry *ce = istate->cache[pos];

518
	record_resolve_undo(istate, ce);
519
	remove_name_hash(istate, ce);
520
	save_or_free_index_entry(istate, ce);
521
	istate->cache_changed |= CE_ENTRY_REMOVED;
522 523
	istate->cache_nr--;
	if (pos >= istate->cache_nr)
524
		return 0;
525 526 527
	memmove(istate->cache + pos,
		istate->cache + pos + 1,
		(istate->cache_nr - pos) * sizeof(struct cache_entry *));
528 529 530
	return 1;
}

531
/*
Ondřej Bílka's avatar
Ondřej Bílka committed
532
 * Remove all cache entries marked for removal, that is where
533 534 535 536 537 538 539 540 541
 * CE_REMOVE is set in ce_flags.  This is much more effective than
 * calling remove_index_entry_at() for each entry to be removed.
 */
void remove_marked_cache_entries(struct index_state *istate)
{
	struct cache_entry **ce_array = istate->cache;
	unsigned int i, j;

	for (i = j = 0; i < istate->cache_nr; i++) {
542
		if (ce_array[i]->ce_flags & CE_REMOVE) {
543
			remove_name_hash(istate, ce_array[i]);
544
			save_or_free_index_entry(istate, ce_array[i]);
545
		}
546 547 548
		else
			ce_array[j++] = ce_array[i];
	}
549 550
	if (j == istate->cache_nr)
		return;
551
	istate->cache_changed |= CE_ENTRY_REMOVED;
552 553 554
	istate->cache_nr = j;
}

555
int remove_file_from_index(struct index_state *istate, const char *path)
556
{
557
	int pos = index_name_pos(istate, path, strlen(path));
558 559
	if (pos < 0)
		pos = -pos-1;
560
	cache_tree_invalidate_path(istate, path);
561
	untracked_cache_remove_from_index(istate, path);
562 563
	while (pos < istate->cache_nr && !strcmp(istate->cache[pos]->name, path))
		remove_index_entry_at(istate, pos);
564 565 566
	return 0;
}

567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594
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;
}

595 596 597 598 599 600 601 602 603 604 605 606 607 608 609
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
 */
610 611 612
static struct cache_entry *create_alias_ce(struct index_state *istate,
					   struct cache_entry *ce,
					   struct cache_entry *alias)
613 614 615 616 617 618 619 620 621 622 623 624
{
	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);
625
	save_or_free_index_entry(istate, ce);
626 627 628
	return new;
}

629
void set_object_name_for_intent_to_add_entry(struct cache_entry *ce)
630 631 632 633
{
	unsigned char sha1[20];
	if (write_sha1_file("", 0, blob_type, sha1))
		die("cannot create an empty blob in the object database");
634
	hashcpy(ce->oid.hash, sha1);
635 636
}

637
int add_to_index(struct index_state *istate, const char *path, struct stat *st, int flags)
Johannes Schindelin's avatar
Johannes Schindelin committed
638
{
639
	int size, namelen, was_same;
640
	mode_t st_mode = st->st_mode;
641
	struct cache_entry *ce, *alias;
642
	unsigned ce_option = CE_MATCH_IGNORE_VALID|CE_MATCH_IGNORE_SKIP_WORKTREE|CE_MATCH_RACY_IS_DIRTY;
643 644
	int verbose = flags & (ADD_CACHE_VERBOSE | ADD_CACHE_PRETEND);
	int pretend = flags & ADD_CACHE_PRETEND;
645 646 647
	int intent_only = flags & ADD_CACHE_INTENT;
	int add_option = (ADD_CACHE_OK_TO_ADD|ADD_CACHE_OK_TO_REPLACE|
			  (intent_only ? ADD_CACHE_NEW_ONLY : 0));
Johannes Schindelin's avatar
Johannes Schindelin committed
648

649
	if (!S_ISREG(st_mode) && !S_ISLNK(st_mode) && !S_ISDIR(st_mode))
650
		return error("%s: can only add regular files, symbolic links or git-directories", path);
Johannes Schindelin's avatar
Johannes Schindelin committed
651 652

	namelen = strlen(path);
653
	if (S_ISDIR(st_mode)) {
654 655 656
		while (namelen && path[namelen-1] == '/')
			namelen--;
	}
Johannes Schindelin's avatar
Johannes Schindelin committed
657 658 659
	size = cache_entry_size(namelen);
	ce = xcalloc(1, size);
	memcpy(ce->name, path, namelen);
660
	ce->ce_namelen = namelen;
661 662
	if (!intent_only)
		fill_stat_cache_info(ce, st);
663 664
	else
		ce->ce_flags |= CE_INTENT_TO_ADD;
Johannes Schindelin's avatar
Johannes Schindelin committed
665

666 667

	if (trust_executable_bit && has_symlinks) {
668
		ce->ce_mode = create_ce_mode(st_mode);
669
	} else {
670 671
		/* 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
672
		 */
673
		struct cache_entry *ent;
674
		int pos = index_name_pos_also_unmerged(istate, path, namelen);
675

676
		ent = (0 <= pos) ? istate->cache[pos] : NULL;
677
		ce->ce_mode = ce_mode_from_stat(ent, st_mode);
Johannes Schindelin's avatar
Johannes Schindelin committed
678 679
	}

680 681 682 683 684 685
	/* When core.ignorecase=true, determine if a directory of the same name but differing
	 * case already exists within the Git repository.  If it does, ensure the directory
	 * case of the file being added to the repository matches (is folded into) the existing
	 * entry's directory case.
	 */
	if (ignore_case) {
686
		adjust_dirname_case(istate, ce->name);
687 688
	}

689
	alias = index_file_exists(istate, ce->name, ce_namelen(ce), ignore_case);
690
	if (alias && !ce_stage(alias) && !ie_match_stat(istate, alias, st, ce_option)) {
691
		/* Nothing changed, really */
692 693
		if (!S_ISGITLINK(alias->ce_mode))
			ce_mark_uptodate(alias);
694
		alias->ce_flags |= CE_ADDED;
695 696

		free(ce);
697 698
		return 0;
	}
699
	if (!intent_only) {
700
		if (index_path(ce->oid.hash, path, st, HASH_WRITE_OBJECT)) {
701
			free(ce);
702
			return error("unable to index file %s", path);
703
		}
704
	} else
705
		set_object_name_for_intent_to_add_entry(ce);
706

707
	if (ignore_case && alias && different_name(ce, alias))
708
		ce = create_alias_ce(istate, ce, alias);
709
	ce->ce_flags |= CE_ADDED;
710

Junio C Hamano's avatar
Junio C Hamano committed
711
	/* It was suspected to be racily clean, but it turns out to be Ok */
712 713
	was_same = (alias &&
		    !ce_stage(alias) &&
714
		    !oidcmp(&alias->oid, &ce->oid) &&
715 716 717
		    ce->ce_mode == alias->ce_mode);

	if (pretend)
718 719 720 721 722
		free(ce);
	else if (add_index_entry(istate, ce, add_option)) {
		free(ce);
		return error("unable to add %s to index", path);
	}
723
	if (verbose && !was_same)
Johannes Schindelin's avatar
Johannes Schindelin committed
724 725 726 727
		printf("add '%s'\n", path);
	return 0;
}

728
int add_file_to_index(struct index_state *istate, const char *path, int flags)
729 730 731
{
	struct stat st;
	if (lstat(path, &st))
732
		die_errno("unable to stat '%s'", path);
733
	return add_to_index(istate, path, &st, flags);
734 735
}

736 737
struct cache_entry *make_cache_entry(unsigned int mode,
		const unsigned char *sha1, const char *path, int stage,
738
		unsigned int refresh_options)
739 740
{
	int size, len;
741
	struct cache_entry *ce, *ret;
742

743 744
	if (!verify_path(path)) {
		error("Invalid path '%s'", path);
745
		return NULL;
746
	}
747 748 749 750 751

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

752
	hashcpy(ce->oid.hash, sha1);
753
	memcpy(ce->name, path, len);
754 755
	ce->ce_flags = create_ce_flags(stage);
	ce->ce_namelen = len;
756 757
	ce->ce_mode = create_ce_mode(mode);

758
	ret = refresh_cache_entry(ce, refresh_options);
Stefan Beller's avatar
Stefan Beller committed
759
	if (ret != ce)
760
		free(ce);
Stefan Beller's avatar
Stefan Beller committed
761
	return ret;
762 763
}

764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792
/*
 * Chmod an index entry with either +x or -x.
 *
 * Returns -1 if the chmod for the particular cache entry failed (if it's
 * not a regular file), -2 if an invalid flip argument is passed in, 0
 * otherwise.
 */
int chmod_index_entry(struct index_state *istate, struct cache_entry *ce,
		      char flip)
{
	if (!S_ISREG(ce->ce_mode))
		return -1;
	switch (flip) {
	case '+':
		ce->ce_mode |= 0111;
		break;
	case '-':
		ce->ce_mode &= ~0111;
		break;
	default:
		return -2;
	}
	cache_tree_invalidate_path(istate, ce->name);
	ce->ce_flags |= CE_UPDATE_IN_BASE;
	istate->cache_changed |= CE_ENTRY_CHANGED;

	return 0;
}

793
int ce_same_name(const struct cache_entry *a, const struct cache_entry *b)
794 795 796 797 798
{
	int len = ce_namelen(a);
	return ce_namelen(b) == len && !memcmp(a->name, b->name, len);
}

799 800 801 802 803 804 805 806 807 808 809 810 811 812 813
/*
 * 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.
	 */
814

815
	/* "." is not allowed */
816
	if (*rest == '\0' || is_dir_sep(*rest))
817 818
		return 0;

819
	switch (*rest) {
820 821 822 823 824
	/*
	 * ".git" followed by  NUL or slash is bad. This
	 * shares the path end test with the ".." case.
	 */
	case 'g':
825 826
	case 'G':
		if (rest[1] != 'i' && rest[1] != 'I')
827
			break;
828
		if (rest[2] != 't' && rest[2] != 'T')
829 830 831 832
			break;
		rest += 2;
	/* fallthrough */
	case '.':
833
		if (rest[1] == '\0' || is_dir_sep(rest[1]))
834 835 836 837 838 839 840 841 842
			return 0;
	}
	return 1;
}

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

843 844 845
	if (has_dos_drive_prefix(path))
		return 0;

846 847 848 849
	goto inside;
	for (;;) {
		if (!c)
			return 1;
850
		if (is_dir_sep(c)) {
851
inside:
852 853
			if (protect_hfs && is_hfs_dotgit(path))
				return 0;
854 855
			if (protect_ntfs && is_ntfs_dotgit(path))
				return 0;
856
			c = *path++;
857 858 859
			if ((c == '.' && !verify_dotfile(path)) ||
			    is_dir_sep(c) || c == '\0')
				return 0;
860 861 862 863 864
		}
		c = *path++;
	}
}

865 866 867
/*
 * Do we have another file that has the beginning components being a
 * proper superset of the name we're trying to add?
868
 */
869 870
static int has_file_name(struct index_state *istate,
			 const struct cache_entry *ce, int pos, int ok_to_replace)
871
{
872 873
	int retval = 0;
	int len = ce_namelen(ce);
874
	int stage = ce_stage(ce);
875
	const char *name = ce->name;
876

877 878
	while (pos < istate->cache_nr) {
		struct cache_entry *p = istate->cache[pos++];
879

880
		if (len >= ce_namelen(p))
881
			break;
882 883
		if (memcmp(name, p->name, len))
			break;
884 885
		if (ce_stage(p) != stage)
			continue;
886 887
		if (p->name[len] != '/')
			continue;
888
		if (p->ce_flags & CE_REMOVE)
889
			continue;
890 891 892
		retval = -1;
		if (!ok_to_replace)
			break;
893
		remove_index_entry_at(istate, --pos);
894
	}
895 896
	return retval;
}
897

898 899 900 901
/*
 * Do we have another file with a pathname that is a proper
 * subset of the name we're trying to add?
 */
902 903
static int has_dir_name(struct index_state *istate,
			const struct cache_entry *ce, int pos, int ok_to_replace)
904 905
{
	int retval = 0;
906
	int stage = ce_stage(ce);
907 908
	const char *name = ce->name;
	const char *slash = name + ce_namelen(ce);
909

910 911
	for (;;) {
		int len;
912

913 914 915 916 917 918 919
		for (;;) {
			if (*--slash == '/')
				break;
			if (slash <= ce->name)
				return retval;
		}
		len = slash - name;
920

921
		pos = index_name_stage_pos(istate, name, len, stage);
922
		if (pos >= 0) {
923 924 925 926 927 928 929 930
			/*
			 * 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.
			 */
931
			if (!(istate->cache[pos]->ce_flags & CE_REMOVE)) {
932 933 934
				retval = -1;
				if (!ok_to_replace)
					break;
935
				remove_index_entry_at(istate, pos);
936 937
				continue;
			}
938
		}
939 940
		else
			pos = -pos-1;
941 942 943 944

		/*
		 * Trivial optimization: if we find an entry that
		 * already matches the sub-directory, then we know
945
		 * we're ok, and we can exit.
946
		 */
947 948
		while (pos < istate->cache_nr) {
			struct cache_entry *p = istate->cache[pos];
949 950 951 952
			if ((ce_namelen(p) <= len) ||
			    (p->name[len] != '/') ||
			    memcmp(p->name, name, len))
				break; /* not our subdirectory */
953 954 955
			if (ce_stage(p) == stage && !(p->ce_flags & CE_REMOVE))
				/*
				 * p is at the same stage as our entry, and
956 957 958 959 960 961
				 * is a subdirectory of what we are looking
				 * at, so we cannot have conflicts at our
				 * level or anything shorter.
				 */
				return retval;
			pos++;
962
		}
963
	}
964 965 966 967 968 969 970
	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
971
 *
972 973 974 975
 * 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.
 */
976 977 978
static int check_file_directory_conflict(struct index_state *istate,
					 const struct cache_entry *ce,
					 int pos, int ok_to_replace)
979
{
980 981 982 983 984
	int retval;

	/*
	 * When ce is an "I am going away" entry, we allow it to be added
	 */
985
	if (ce->ce_flags & CE_REMOVE)
986 987
		return 0;

988 989 990
	/*
	 * We check if the path is a sub-path of a subsequent pathname
	 * first, since removing those will not change the position
991
	 * in the array.
992
	 */
993
	retval = has_file_name(istate, ce, pos, ok_to_replace);
994

995 996 997 998
	/*
	 * Then check if the path might have a clashing sub-directory
	 * before it.
	 */
999
	return retval + has_dir_name(istate, ce, pos, ok_to_replace);
1000 1001
}

1002
static int add_index_entry_with_check(struct index_state *istate, struct cache_entry *ce, int option)
1003 1004
{
	int pos;
1005 1006
	int ok_to_add = option & ADD_CACHE_OK_TO_ADD;
	int ok_to_replace = option & ADD_CACHE_OK_TO_REPLACE;
1007
	int skip_df_check = option & ADD_CACHE_SKIP_DFCHECK;
1008
	int new_only = option & ADD_CACHE_NEW_ONLY;
Junio C Hamano's avatar
Junio C Hamano committed
1009

1010 1011
	if (!(option & ADD_CACHE_KEEP_CACHE_TREE))
		cache_tree_invalidate_path(istate, ce->name);
1012
	pos = index_name_stage_pos(istate, ce->name, ce_namelen(ce), ce_stage(ce));
1013

Junio C Hamano's avatar
Junio C Hamano committed
1014
	/* existing match? Just replace it. */
1015
	if (pos >= 0) {
1016 1017
		if (!new_only)
			replace_index_entry(istate, pos, ce);
1018 1019
		return 0;
	}
1020
	pos = -pos-1;
1021

1022 1023
	if (!(option & ADD_CACHE_KEEP_CACHE_TREE))
		untracked_cache_add_to_index(istate, ce->name);
1024

1025 1026 1027 1028
	/*
	 * Inserting a merged entry ("stage 0") into the index
	 * will always replace all non-merged entries..
	 */
1029 1030
	if (pos < istate->cache_nr && ce_stage(ce) == 0) {
		while (ce_same_name(istate->cache[pos], ce)) {
1031
			ok_to_add = 1;
1032
			if (!remove_index_entry_at(istate, pos))
1033 1034 1035 1036
				break;
		}
	}

1037 1038
	if (!ok_to_add)
		return -1;
1039
	if (!verify_path(ce->name))
1040
		return error("Invalid path '%s'", ce->name);
1041

Junio C Hamano's avatar
Junio C Hamano committed
1042
	if (!skip_df_check &&
1043
	    check_file_directory_conflict(istate, ce, pos, ok_to_replace)) {
1044
		if (!ok_to_replace)
1045 1046
			return error("'%s' appears as both a file and as a directory",
				     ce->name);
1047
		pos = index_name_stage_pos(istate, ce->name, ce_namelen(ce), ce_stage(ce));
1048 1049
		pos = -pos-1;
	}
1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065
	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;
	}
1066

1067
	/* Make sure the array is big enough .. */
1068
	ALLOC_GROW(istate->cache, istate->cache_nr + 1, istate->cache_alloc);
1069 1070

	/* Add it in.. */
1071
	istate->cache_nr++;
1072
	if (istate->cache_nr > pos + 1)
1073 1074 1075
		memmove(istate->cache + pos + 1,
			istate->cache + pos,
			(istate->cache_nr - pos - 1) * sizeof(ce));
1076
	set_index_entry(istate, pos, ce);