sha1_file.c 41.7 KB
Newer Older
1 2 3 4 5 6 7 8 9
/*
 * GIT - The information manager from hell
 *
 * Copyright (C) Linus Torvalds, 2005
 *
 * This handles basic git sha1 object files - packing, unpacking,
 * creation etc.
 */
#include "cache.h"
10
#include "delta.h"
11
#include "pack.h"
12 13 14 15
#include "blob.h"
#include "commit.h"
#include "tag.h"
#include "tree.h"
16

17 18 19 20 21 22 23 24
#ifndef O_NOATIME
#if defined(__linux__) && (defined(__i386__) || defined(__PPC__))
#define O_NOATIME 01000000
#else
#define O_NOATIME 0
#endif
#endif

25
const unsigned char null_sha1[20];
J
Junio C Hamano 已提交
26

27 28
static unsigned int sha1_file_open_flag = O_NOATIME;

29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
static unsigned hexval(char c)
{
	if (c >= '0' && c <= '9')
		return c - '0';
	if (c >= 'a' && c <= 'f')
		return c - 'a' + 10;
	if (c >= 'A' && c <= 'F')
		return c - 'A' + 10;
	return ~0;
}

int get_sha1_hex(const char *hex, unsigned char *sha1)
{
	int i;
	for (i = 0; i < 20; i++) {
		unsigned int val = (hexval(hex[0]) << 4) | hexval(hex[1]);
		if (val & ~0xff)
			return -1;
		*sha1++ = val;
		hex += 2;
	}
	return 0;
}

53 54 55
int safe_create_leading_directories(char *path)
{
	char *pos = path;
56 57
	struct stat st;

58 59
	if (*pos == '/')
		pos++;
60 61 62 63 64 65

	while (pos) {
		pos = strchr(pos, '/');
		if (!pos)
			break;
		*pos = 0;
66 67 68
		if (!stat(path, &st)) {
			/* path exists */
			if (!S_ISDIR(st.st_mode)) {
69
				*pos = '/';
70
				return -3;
71
			}
72
		}
73 74 75 76
		else if (mkdir(path, 0777)) {
			*pos = '/';
			return -1;
		}
77 78 79 80
		else if (adjust_shared_perm(path)) {
			*pos = '/';
			return -2;
		}
81 82 83 84
		*pos++ = '/';
	}
	return 0;
}
85

86 87
char * sha1_to_hex(const unsigned char *sha1)
{
L
Linus Torvalds 已提交
88 89
	static int bufno;
	static char hexbuffer[4][50];
90
	static const char hex[] = "0123456789abcdef";
L
Linus Torvalds 已提交
91
	char *buffer = hexbuffer[3 & ++bufno], *buf = buffer;
92 93 94 95 96 97 98
	int i;

	for (i = 0; i < 20; i++) {
		unsigned int val = *sha1++;
		*buf++ = hex[val >> 4];
		*buf++ = hex[val & 0xf];
	}
99 100
	*buf = '\0';

101 102 103
	return buffer;
}

104 105 106 107 108 109 110 111 112 113 114 115
static void fill_sha1_path(char *pathbuf, const unsigned char *sha1)
{
	int i;
	for (i = 0; i < 20; i++) {
		static char hex[] = "0123456789abcdef";
		unsigned int val = sha1[i];
		char *pos = pathbuf + i*2 + (i > 0);
		*pos++ = hex[val >> 4];
		*pos = hex[val & 0xf];
	}
}

116 117
/*
 * NOTE! This returns a statically allocated buffer, so you have to be
118
 * careful about using it. Do a "xstrdup()" if you need to save the
119
 * filename.
120 121 122
 *
 * Also note that this returns the location for creating.  Reading
 * SHA1 file can happen from any alternate directory listed in the
J
Junio C Hamano 已提交
123
 * DB_ENVIRONMENT environment variable if it is not found in
124
 * the primary object database.
125 126 127 128 129 130
 */
char *sha1_file_name(const unsigned char *sha1)
{
	static char *name, *base;

	if (!base) {
J
Junio C Hamano 已提交
131
		const char *sha1_file_directory = get_object_directory();
132
		int len = strlen(sha1_file_directory);
133
		base = xmalloc(len + 60);
134 135 136 137 138 139
		memcpy(base, sha1_file_directory, len);
		memset(base+len, 0, 60);
		base[len] = '/';
		base[len+3] = '/';
		name = base + len + 1;
	}
140
	fill_sha1_path(name, sha1);
141 142 143
	return base;
}

144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193
char *sha1_pack_name(const unsigned char *sha1)
{
	static const char hex[] = "0123456789abcdef";
	static char *name, *base, *buf;
	int i;

	if (!base) {
		const char *sha1_file_directory = get_object_directory();
		int len = strlen(sha1_file_directory);
		base = xmalloc(len + 60);
		sprintf(base, "%s/pack/pack-1234567890123456789012345678901234567890.pack", sha1_file_directory);
		name = base + len + 11;
	}

	buf = name;

	for (i = 0; i < 20; i++) {
		unsigned int val = *sha1++;
		*buf++ = hex[val >> 4];
		*buf++ = hex[val & 0xf];
	}
	
	return base;
}

char *sha1_pack_index_name(const unsigned char *sha1)
{
	static const char hex[] = "0123456789abcdef";
	static char *name, *base, *buf;
	int i;

	if (!base) {
		const char *sha1_file_directory = get_object_directory();
		int len = strlen(sha1_file_directory);
		base = xmalloc(len + 60);
		sprintf(base, "%s/pack/pack-1234567890123456789012345678901234567890.idx", sha1_file_directory);
		name = base + len + 11;
	}

	buf = name;

	for (i = 0; i < 20; i++) {
		unsigned int val = *sha1++;
		*buf++ = hex[val >> 4];
		*buf++ = hex[val & 0xf];
	}
	
	return base;
}

194 195
struct alternate_object_database *alt_odb_list;
static struct alternate_object_database **alt_odb_tail;
196

M
Martin Waitz 已提交
197 198
static void read_info_alternates(const char * alternates, int depth);

J
Junio C Hamano 已提交
199 200
/*
 * Prepare alternate object database registry.
201 202 203 204 205
 *
 * The variable alt_odb_list points at the list of struct
 * alternate_object_database.  The elements on this list come from
 * non-empty elements from colon separated ALTERNATE_DB_ENVIRONMENT
 * environment variable, and $GIT_OBJECT_DIRECTORY/info/alternates,
206 207
 * whose contents is similar to that environment variable but can be
 * LF separated.  Its base points at a statically allocated buffer that
208 209 210 211 212
 * contains "/the/directory/corresponding/to/.git/objects/...", while
 * its name points just after the slash at the end of ".git/objects/"
 * in the example above, and has enough space to hold 40-byte hex
 * SHA1, an extra slash for the first level indirection, and the
 * terminating NUL.
J
Junio C Hamano 已提交
213
 */
M
Martin Waitz 已提交
214
static int link_alt_odb_entry(const char * entry, int len, const char * relative_base, int depth)
215
{
M
Martin Waitz 已提交
216
	struct stat st;
217
	const char *objdir = get_object_directory();
M
Martin Waitz 已提交
218 219 220 221 222
	struct alternate_object_database *ent;
	struct alternate_object_database *alt;
	/* 43 = 40-byte + 2 '/' + terminating NUL */
	int pfxlen = len;
	int entlen = pfxlen + 43;
223
	int base_len = -1;
224

M
Martin Waitz 已提交
225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292
	if (*entry != '/' && relative_base) {
		/* Relative alt-odb */
		if (base_len < 0)
			base_len = strlen(relative_base) + 1;
		entlen += base_len;
		pfxlen += base_len;
	}
	ent = xmalloc(sizeof(*ent) + entlen);

	if (*entry != '/' && relative_base) {
		memcpy(ent->base, relative_base, base_len - 1);
		ent->base[base_len - 1] = '/';
		memcpy(ent->base + base_len, entry, len);
	}
	else
		memcpy(ent->base, entry, pfxlen);

	ent->name = ent->base + pfxlen + 1;
	ent->base[pfxlen + 3] = '/';
	ent->base[pfxlen] = ent->base[entlen-1] = 0;

	/* Detect cases where alternate disappeared */
	if (stat(ent->base, &st) || !S_ISDIR(st.st_mode)) {
		error("object directory %s does not exist; "
		      "check .git/objects/info/alternates.",
		      ent->base);
		free(ent);
		return -1;
	}

	/* Prevent the common mistake of listing the same
	 * thing twice, or object directory itself.
	 */
	for (alt = alt_odb_list; alt; alt = alt->next) {
		if (!memcmp(ent->base, alt->base, pfxlen)) {
			free(ent);
			return -1;
		}
	}
	if (!memcmp(ent->base, objdir, pfxlen)) {
		free(ent);
		return -1;
	}

	/* add the alternate entry */
	*alt_odb_tail = ent;
	alt_odb_tail = &(ent->next);
	ent->next = NULL;

	/* recursively add alternates */
	read_info_alternates(ent->base, depth + 1);

	ent->base[pfxlen] = '/';

	return 0;
}

static void link_alt_odb_entries(const char *alt, const char *ep, int sep,
				 const char *relative_base, int depth)
{
	const char *cp, *last;

	if (depth > 5) {
		error("%s: ignoring alternate object stores, nesting too deep.",
				relative_base);
		return;
	}

293
	last = alt;
294 295 296 297 298 299 300 301
	while (last < ep) {
		cp = last;
		if (cp < ep && *cp == '#') {
			while (cp < ep && *cp != sep)
				cp++;
			last = cp + 1;
			continue;
		}
M
Martin Waitz 已提交
302 303
		while (cp < ep && *cp != sep)
			cp++;
304
		if (last != cp) {
M
Martin Waitz 已提交
305 306 307 308 309 310
			if ((*last != '/') && depth) {
				error("%s: ignoring relative alternate object store %s",
						relative_base, last);
			} else {
				link_alt_odb_entry(last, cp - last,
						relative_base, depth);
311
			}
312
		}
313
		while (cp < ep && *cp == sep)
314 315
			cp++;
		last = cp;
316
	}
317 318
}

M
Martin Waitz 已提交
319
static void read_info_alternates(const char * relative_base, int depth)
320
{
321
	char *map;
322
	struct stat st;
M
Martin Waitz 已提交
323 324
	char path[PATH_MAX];
	int fd;
325

M
Martin Waitz 已提交
326
	sprintf(path, "%s/info/alternates", relative_base);
327 328 329 330 331
	fd = open(path, O_RDONLY);
	if (fd < 0)
		return;
	if (fstat(fd, &st) || (st.st_size == 0)) {
		close(fd);
332
		return;
333
	}
334 335 336 337 338
	map = mmap(NULL, st.st_size, PROT_READ, MAP_PRIVATE, fd, 0);
	close(fd);
	if (map == MAP_FAILED)
		return;

M
Martin Waitz 已提交
339 340
	link_alt_odb_entries(map, map + st.st_size, '\n', relative_base, depth);

341
	munmap(map, st.st_size);
342 343
}

M
Martin Waitz 已提交
344 345
void prepare_alt_odb(void)
{
T
Timo Hirvonen 已提交
346
	const char *alt;
M
Martin Waitz 已提交
347 348 349 350 351 352 353 354 355 356 357 358

	alt = getenv(ALTERNATE_DB_ENVIRONMENT);
	if (!alt) alt = "";

	if (alt_odb_tail)
		return;
	alt_odb_tail = &alt_odb_list;
	link_alt_odb_entries(alt, alt + strlen(alt), ':', NULL, 0);

	read_info_alternates(get_object_directory(), 0);
}

359 360 361
static char *find_sha1_file(const unsigned char *sha1, struct stat *st)
{
	char *name = sha1_file_name(sha1);
362
	struct alternate_object_database *alt;
363 364 365

	if (!stat(name, st))
		return name;
366
	prepare_alt_odb();
367 368
	for (alt = alt_odb_list; alt; alt = alt->next) {
		name = alt->name;
369
		fill_sha1_path(name, sha1);
370 371
		if (!stat(alt->base, st))
			return alt->base;
372 373 374 375
	}
	return NULL;
}

376 377 378
#define PACK_MAX_SZ (1<<26)
static int pack_used_ctr;
static unsigned long pack_mapped;
379
struct packed_git *packed_git;
380 381 382 383 384 385 386 387

static int check_packed_git_idx(const char *path, unsigned long *idx_size_,
				void **idx_map_)
{
	void *idx_map;
	unsigned int *index;
	unsigned long idx_size;
	int nr, i;
388
	int fd = open(path, O_RDONLY);
389 390 391 392 393 394 395 396 397 398 399 400 401 402
	struct stat st;
	if (fd < 0)
		return -1;
	if (fstat(fd, &st)) {
		close(fd);
		return -1;
	}
	idx_size = st.st_size;
	idx_map = mmap(NULL, idx_size, PROT_READ, MAP_PRIVATE, fd, 0);
	close(fd);
	if (idx_map == MAP_FAILED)
		return -1;

	index = idx_map;
403 404
	*idx_map_ = idx_map;
	*idx_size_ = idx_size;
405 406

	/* check index map */
407
	if (idx_size < 4*256 + 20 + 20)
408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429
		return error("index file too small");
	nr = 0;
	for (i = 0; i < 256; i++) {
		unsigned int n = ntohl(index[i]);
		if (n < nr)
			return error("non-monotonic index");
		nr = n;
	}

	/*
	 * Total size:
	 *  - 256 index entries 4 bytes each
	 *  - 24-byte entries * nr (20-byte sha1 + 4-byte offset)
	 *  - 20-byte SHA1 of the packfile
	 *  - 20-byte SHA1 file checksum
	 */
	if (idx_size != 4*256 + nr * 24 + 20 + 20)
		return error("wrong index file size");

	return 0;
}

430
static int unuse_one_packed_git(void)
431
{
432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449
	struct packed_git *p, *lru = NULL;

	for (p = packed_git; p; p = p->next) {
		if (p->pack_use_cnt || !p->pack_base)
			continue;
		if (!lru || p->pack_last_used < lru->pack_last_used)
			lru = p;
	}
	if (!lru)
		return 0;
	munmap(lru->pack_base, lru->pack_size);
	lru->pack_base = NULL;
	return 1;
}

void unuse_packed_git(struct packed_git *p)
{
	p->pack_use_cnt--;
450 451
}

452
int use_packed_git(struct packed_git *p)
453
{
454 455
	if (!p->pack_size) {
		struct stat st;
456
		/* We created the struct before we had the pack */
457 458 459 460 461
		stat(p->pack_name, &st);
		if (!S_ISREG(st.st_mode))
			die("packfile %s not a regular file", p->pack_name);
		p->pack_size = st.st_size;
	}
462 463 464 465
	if (!p->pack_base) {
		int fd;
		struct stat st;
		void *map;
466
		struct pack_header *hdr;
467 468

		pack_mapped += p->pack_size;
469 470
		while (PACK_MAX_SZ < pack_mapped && unuse_one_packed_git())
			; /* nothing */
471 472
		fd = open(p->pack_name, O_RDONLY);
		if (fd < 0)
473
			die("packfile %s cannot be opened", p->pack_name);
474 475
		if (fstat(fd, &st)) {
			close(fd);
476
			die("packfile %s cannot be opened", p->pack_name);
477 478
		}
		if (st.st_size != p->pack_size)
479
			die("packfile %s size mismatch.", p->pack_name);
480 481 482
		map = mmap(NULL, p->pack_size, PROT_READ, MAP_PRIVATE, fd, 0);
		close(fd);
		if (map == MAP_FAILED)
483
			die("packfile %s cannot be mapped.", p->pack_name);
484
		p->pack_base = map;
485

486 487 488 489 490 491 492 493 494 495 496
		/* Check if we understand this pack file.  If we don't we're
		 * likely too old to handle it.
		 */
		hdr = map;
		if (hdr->hdr_signature != htonl(PACK_SIGNATURE))
			die("packfile %s isn't actually a pack.", p->pack_name);
		if (!pack_version_ok(hdr->hdr_version))
			die("packfile %s is version %i and not supported"
				" (try upgrading GIT to a newer version)",
				p->pack_name, ntohl(hdr->hdr_version));

497 498 499
		/* Check if the pack file matches with the index file.
		 * this is cheap.
		 */
500 501 502 503
		if (hashcmp((unsigned char *)(p->index_base) +
			    p->index_size - 40,
			    (unsigned char *)p->pack_base +
			    p->pack_size - 20)) {
504
			die("packfile %s does not match index.", p->pack_name);
505
		}
506 507
	}
	p->pack_last_used = pack_used_ctr++;
508
	p->pack_use_cnt++;
509 510 511
	return 0;
}

512
struct packed_git *add_packed_git(char *path, int path_len, int local)
513 514 515 516 517
{
	struct stat st;
	struct packed_git *p;
	unsigned long idx_size;
	void *idx_map;
518
	unsigned char sha1[20];
519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537

	if (check_packed_git_idx(path, &idx_size, &idx_map))
		return NULL;

	/* do we have a corresponding .pack file? */
	strcpy(path + path_len - 4, ".pack");
	if (stat(path, &st) || !S_ISREG(st.st_mode)) {
		munmap(idx_map, idx_size);
		return NULL;
	}
	/* ok, it looks sane as far as we can check without
	 * actually mapping the pack file.
	 */
	p = xmalloc(sizeof(*p) + path_len + 2);
	strcpy(p->pack_name, path);
	p->index_size = idx_size;
	p->pack_size = st.st_size;
	p->index_base = idx_map;
	p->next = NULL;
538
	p->pack_base = NULL;
539
	p->pack_last_used = 0;
540
	p->pack_use_cnt = 0;
541
	p->pack_local = local;
P
Pavel Roskin 已提交
542
	if ((path_len > 44) && !get_sha1_hex(path + path_len - 44, sha1))
543
		hashcpy(p->sha1, sha1);
544 545 546
	return p;
}

547
struct packed_git *parse_pack_index(unsigned char *sha1)
548 549 550 551 552
{
	char *path = sha1_pack_index_name(sha1);
	return parse_pack_index_file(sha1, path);
}

553
struct packed_git *parse_pack_index_file(const unsigned char *sha1, char *idx_path)
554 555 556 557
{
	struct packed_git *p;
	unsigned long idx_size;
	void *idx_map;
558
	char *path;
559

560
	if (check_packed_git_idx(idx_path, &idx_size, &idx_map))
561 562 563 564 565 566 567 568 569 570 571 572 573
		return NULL;

	path = sha1_pack_name(sha1);

	p = xmalloc(sizeof(*p) + strlen(path) + 2);
	strcpy(p->pack_name, path);
	p->index_size = idx_size;
	p->pack_size = 0;
	p->index_base = idx_map;
	p->next = NULL;
	p->pack_base = NULL;
	p->pack_last_used = 0;
	p->pack_use_cnt = 0;
574
	hashcpy(p->sha1, sha1);
575 576 577 578 579 580 581 582 583
	return p;
}

void install_packed_git(struct packed_git *pack)
{
	pack->next = packed_git;
	packed_git = pack;
}

584
static void prepare_packed_git_one(char *objdir, int local)
585 586 587 588 589 590 591 592 593
{
	char path[PATH_MAX];
	int len;
	DIR *dir;
	struct dirent *de;

	sprintf(path, "%s/pack", objdir);
	len = strlen(path);
	dir = opendir(path);
594
	if (!dir) {
J
Junio C Hamano 已提交
595
		if (errno != ENOENT)
596
			error("unable to open object pack directory: %s: %s",
J
Junio C Hamano 已提交
597
			      path, strerror(errno));
598
		return;
599
	}
600 601 602 603 604
	path[len++] = '/';
	while ((de = readdir(dir)) != NULL) {
		int namelen = strlen(de->d_name);
		struct packed_git *p;

605
		if (!has_extension(de->d_name, ".idx"))
606 607 608 609
			continue;

		/* we have .idx.  Is it a file we can map? */
		strcpy(path + len, de->d_name);
610 611 612 613 614 615
		for (p = packed_git; p; p = p->next) {
			if (!memcmp(path, p->pack_name, len + namelen - 4))
				break;
		}
		if (p)
			continue;
616
		p = add_packed_git(path, len + namelen, local);
617 618 619 620 621
		if (!p)
			continue;
		p->next = packed_git;
		packed_git = p;
	}
622
	closedir(dir);
623 624
}

625
static int prepare_packed_git_run_once = 0;
626
void prepare_packed_git(void)
627
{
628
	struct alternate_object_database *alt;
629

630
	if (prepare_packed_git_run_once)
631
		return;
632
	prepare_packed_git_one(get_object_directory(), 1);
633
	prepare_alt_odb();
634
	for (alt = alt_odb_list; alt; alt = alt->next) {
635
		alt->name[-1] = 0;
636
		prepare_packed_git_one(alt->base, 0);
637
		alt->name[-1] = '/';
638
	}
639 640 641 642 643 644 645
	prepare_packed_git_run_once = 1;
}

static void reprepare_packed_git(void)
{
	prepare_packed_git_run_once = 0;
	prepare_packed_git();
646 647
}

648
int check_sha1_signature(const unsigned char *sha1, void *map, unsigned long size, const char *type)
649
{
650
	char header[100];
651 652 653 654
	unsigned char real_sha1[20];
	SHA_CTX c;

	SHA1_Init(&c);
655
	SHA1_Update(&c, header, 1+sprintf(header, "%s %lu", type, size));
656 657
	SHA1_Update(&c, map, size);
	SHA1_Final(real_sha1, &c);
658
	return hashcmp(sha1, real_sha1) ? -1 : 0;
659 660
}

661
void *map_sha1_file(const unsigned char *sha1, unsigned long *size)
662 663 664
{
	struct stat st;
	void *map;
665
	int fd;
666 667 668 669 670
	char *filename = find_sha1_file(sha1, &st);

	if (!filename) {
		return NULL;
	}
671

672
	fd = open(filename, O_RDONLY | sha1_file_open_flag);
673
	if (fd < 0) {
674 675 676 677 678 679 680 681 682 683 684
		/* See if it works without O_NOATIME */
		switch (sha1_file_open_flag) {
		default:
			fd = open(filename, O_RDONLY);
			if (fd >= 0)
				break;
		/* Fallthrough */
		case 0:
			return NULL;
		}

685 686 687
		/* If it failed once, it will probably fail again.
		 * Stop using O_NOATIME
		 */
688
		sha1_file_open_flag = 0;
689 690 691
	}
	map = mmap(NULL, st.st_size, PROT_READ, MAP_PRIVATE, fd, 0);
	close(fd);
P
Pavel Roskin 已提交
692
	if (map == MAP_FAILED)
693 694 695 696 697
		return NULL;
	*size = st.st_size;
	return map;
}

698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713
int legacy_loose_object(unsigned char *map)
{
	unsigned int word;

	/*
	 * Is it a zlib-compressed buffer? If so, the first byte
	 * must be 0x78 (15-bit window size, deflated), and the
	 * first 16-bit word is evenly divisible by 31
	 */
	word = (map[0] << 8) + map[1];
	if (map[0] == 0x78 && !(word % 31))
		return 1;
	else
		return 0;
}

714
unsigned long unpack_object_header_gently(const unsigned char *buf, unsigned long len, enum object_type *type, unsigned long *sizep)
715
{
716
	unsigned shift;
717 718
	unsigned char c;
	unsigned long size;
719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744
	unsigned long used = 0;

	c = buf[used++];
	*type = (c >> 4) & 7;
	size = c & 15;
	shift = 4;
	while (c & 0x80) {
		if (len <= used)
			return 0;
		if (sizeof(long) * 8 <= shift)
			return 0;
		c = buf[used++];
		size += (c & 0x7f) << shift;
		shift += 7;
	}
	*sizep = size;
	return used;
}

static int unpack_sha1_header(z_stream *stream, unsigned char *map, unsigned long mapsize, void *buffer, unsigned long bufsiz)
{
	unsigned long size, used;
	static const char valid_loose_object_type[8] = {
		0, /* OBJ_EXT */
		1, 1, 1, 1, /* "commit", "tree", "blob", "tag" */
		0, /* "delta" and others are invalid in a loose object */
745
	};
746
	enum object_type type;
747

748 749 750 751 752
	/* Get the data stream */
	memset(stream, 0, sizeof(*stream));
	stream->next_in = map;
	stream->avail_in = mapsize;
	stream->next_out = buffer;
753 754
	stream->avail_out = bufsiz;

755
	if (legacy_loose_object(map)) {
756 757 758 759
		inflateInit(stream);
		return inflate(stream, 0);
	}

760 761
	used = unpack_object_header_gently(map, mapsize, &type, &size);
	if (!used || !valid_loose_object_type[type])
762
		return -1;
763 764
	map += used;
	mapsize -= used;
765 766 767 768

	/* Set up the stream for the rest.. */
	stream->next_in = map;
	stream->avail_in = mapsize;
769
	inflateInit(stream);
770 771

	/* And generate the fake traditional header */
772 773
	stream->total_out = 1 + snprintf(buffer, bufsiz, "%s %lu",
					 type_names[type], size);
774
	return 0;
775 776
}

L
Linus Torvalds 已提交
777
static void *unpack_sha1_rest(z_stream *stream, void *buffer, unsigned long size)
778 779
{
	int bytes = strlen(buffer) + 1;
780
	unsigned char *buf = xmalloc(1+size);
781
	unsigned long n;
782

783 784 785 786 787
	n = stream->total_out - bytes;
	if (n > size)
		n = size;
	memcpy(buf, (char *) buffer + bytes, n);
	bytes = n;
788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803
	if (bytes < size) {
		stream->next_out = buf + bytes;
		stream->avail_out = size - bytes;
		while (inflate(stream, Z_FINISH) == Z_OK)
			/* nothing */;
	}
	buf[size] = 0;
	inflateEnd(stream);
	return buf;
}

/*
 * We used to just use "sscanf()", but that's actually way
 * too permissive for what we want to check. So do an anal
 * object header parse by hand.
 */
804
static int parse_sha1_header(char *hdr, char *type, unsigned long *sizep)
805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848
{
	int i;
	unsigned long size;

	/*
	 * The type can be at most ten bytes (including the 
	 * terminating '\0' that we add), and is followed by
	 * a space. 
	 */
	i = 10;
	for (;;) {
		char c = *hdr++;
		if (c == ' ')
			break;
		if (!--i)
			return -1;
		*type++ = c;
	}
	*type = 0;

	/*
	 * The length must follow immediately, and be in canonical
	 * decimal format (ie "010" is not valid).
	 */
	size = *hdr++ - '0';
	if (size > 9)
		return -1;
	if (size) {
		for (;;) {
			unsigned long c = *hdr - '0';
			if (c > 9)
				break;
			hdr++;
			size = size * 10 + c;
		}
	}
	*sizep = size;

	/*
	 * The length must be followed by a zero byte
	 */
	return *hdr ? -1 : 0;
}

849 850
void * unpack_sha1_file(void *map, unsigned long mapsize, char *type, unsigned long *size)
{
851
	int ret;
852
	z_stream stream;
853
	char hdr[8192];
854

855 856
	ret = unpack_sha1_header(&stream, map, mapsize, hdr, sizeof(hdr));
	if (ret < Z_OK || parse_sha1_header(hdr, type, size) < 0)
857 858
		return NULL;

859
	return unpack_sha1_rest(&stream, hdr, *size);
860 861
}

J
Junio C Hamano 已提交
862 863 864 865
/* forward declaration for a mutually recursive function */
static int packed_object_info(struct pack_entry *entry,
			      char *type, unsigned long *sizep);

866 867 868 869
static int packed_delta_info(unsigned char *base_sha1,
			     unsigned long delta_size,
			     unsigned long left,
			     char *type,
J
Junio C Hamano 已提交
870 871
			     unsigned long *sizep,
			     struct packed_git *p)
872
{
J
Junio C Hamano 已提交
873 874
	struct pack_entry base_ent;

875 876 877
	if (left < 20)
		die("truncated pack file");

J
Junio C Hamano 已提交
878 879 880 881 882
	/* The base entry _must_ be in the same pack */
	if (!find_pack_entry_one(base_sha1, &base_ent, p))
		die("failed to find delta-pack base object %s",
		    sha1_to_hex(base_sha1));

883 884 885 886 887
	/* We choose to only get the type of the base object and
	 * ignore potentially corrupt pack file that expects the delta
	 * based on a base with a wrong size.  This saves tons of
	 * inflate() calls.
	 */
888

J
Junio C Hamano 已提交
889
	if (packed_object_info(&base_ent, type, NULL))
890
		die("cannot get info for delta-pack base");
891

892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916
	if (sizep) {
		const unsigned char *data;
		unsigned char delta_head[64];
		unsigned long result_size;
		z_stream stream;
		int st;

		memset(&stream, 0, sizeof(stream));

		data = stream.next_in = base_sha1 + 20;
		stream.avail_in = left - 20;
		stream.next_out = delta_head;
		stream.avail_out = sizeof(delta_head);

		inflateInit(&stream);
		st = inflate(&stream, Z_FINISH);
		inflateEnd(&stream);
		if ((st != Z_STREAM_END) &&
		    stream.total_out != sizeof(delta_head))
			die("delta data unpack-initial failed");

		/* Examine the initial part of the delta to figure out
		 * the result size.
		 */
		data = delta_head;
917 918 919

		/* ignore base size */
		get_delta_hdr_size(&data, delta_head+sizeof(delta_head));
920

921
		/* Read the result size */
922
		result_size = get_delta_hdr_size(&data, delta_head+sizeof(delta_head));
923 924
		*sizep = result_size;
	}
925 926 927
	return 0;
}

928 929 930
static unsigned long unpack_object_header(struct packed_git *p, unsigned long offset,
	enum object_type *type, unsigned long *sizep)
{
931
	unsigned long used;
932

933
	if (p->pack_size <= offset)
934
		die("object offset outside of pack file");
935 936 937 938 939 940 941 942

	used = unpack_object_header_gently((unsigned char *)p->pack_base +
					   offset,
					   p->pack_size - offset, type, sizep);
	if (!used)
		die("object offset outside of pack file");

	return offset + used;
943 944
}

945 946 947 948 949 950 951 952 953 954 955 956
int check_reuse_pack_delta(struct packed_git *p, unsigned long offset,
			   unsigned char *base, unsigned long *sizep,
			   enum object_type *kindp)
{
	unsigned long ptr;
	int status = -1;

	use_packed_git(p);
	ptr = offset;
	ptr = unpack_object_header(p, ptr, kindp, sizep);
	if (*kindp != OBJ_DELTA)
		goto done;
957
	hashcpy(base, (unsigned char *) p->pack_base + ptr);
958 959 960 961 962 963
	status = 0;
 done:
	unuse_packed_git(p);
	return status;
}

964 965 966 967
void packed_object_info_detail(struct pack_entry *e,
			       char *type,
			       unsigned long *size,
			       unsigned long *store_size,
J
Junio C Hamano 已提交
968
			       unsigned int *delta_chain_length,
969 970 971
			       unsigned char *base_sha1)
{
	struct packed_git *p = e->p;
972
	unsigned long offset;
973 974 975 976
	unsigned char *pack;
	enum object_type kind;

	offset = unpack_object_header(p, e->offset, &kind, size);
977
	pack = (unsigned char *) p->pack_base + offset;
978 979 980
	if (kind != OBJ_DELTA)
		*delta_chain_length = 0;
	else {
J
Junio C Hamano 已提交
981
		unsigned int chain_length = 0;
982 983 984
		if (p->pack_size <= offset + 20)
			die("pack file %s records an incomplete delta base",
			    p->pack_name);
985
		hashcpy(base_sha1, pack);
986 987 988 989 990 991 992
		do {
			struct pack_entry base_ent;
			unsigned long junk;

			find_pack_entry_one(pack, &base_ent, p);
			offset = unpack_object_header(p, base_ent.offset,
						      &kind, &junk);
993
			pack = (unsigned char *) p->pack_base + offset;
994 995 996 997 998 999 1000 1001 1002
			chain_length++;
		} while (kind == OBJ_DELTA);
		*delta_chain_length = chain_length;
	}
	switch (kind) {
	case OBJ_COMMIT:
	case OBJ_TREE:
	case OBJ_BLOB:
	case OBJ_TAG:
1003
		strcpy(type, type_names[kind]);
1004 1005
		break;
	default:
1006 1007
		die("corrupted pack file %s containing object of kind %d",
		    p->pack_name, kind);
1008 1009 1010 1011
	}
	*store_size = 0; /* notyet */
}

1012 1013 1014 1015 1016 1017
static int packed_object_info(struct pack_entry *entry,
			      char *type, unsigned long *sizep)
{
	struct packed_git *p = entry->p;
	unsigned long offset, size, left;
	unsigned char *pack;
1018
	enum object_type kind;
1019
	int retval;
1020 1021 1022 1023

	if (use_packed_git(p))
		die("cannot map packed file");

1024
	offset = unpack_object_header(p, entry->offset, &kind, &size);
1025
	pack = (unsigned char *) p->pack_base + offset;
1026 1027 1028 1029
	left = p->pack_size - offset;

	switch (kind) {
	case OBJ_DELTA:
J
Junio C Hamano 已提交
1030
		retval = packed_delta_info(pack, size, left, type, sizep, p);
1031 1032
		unuse_packed_git(p);
		return retval;
1033 1034 1035 1036
	case OBJ_COMMIT:
	case OBJ_TREE:
	case OBJ_BLOB:
	case OBJ_TAG:
1037
		strcpy(type, type_names[kind]);
1038
		break;
1039
	default:
1040 1041
		die("corrupted pack file %s containing object of kind %d",
		    p->pack_name, kind);
1042
	}
1043 1044
	if (sizep)
		*sizep = size;
1045
	unuse_packed_git(p);
1046 1047 1048
	return 0;
}

1049 1050 1051
static void *unpack_compressed_entry(struct packed_git *p,
				    unsigned long offset,
				    unsigned long size)
1052 1053 1054 1055 1056 1057 1058 1059
{
	int st;
	z_stream stream;
	unsigned char *buffer;

	buffer = xmalloc(size + 1);
	buffer[size] = 0;
	memset(&stream, 0, sizeof(stream));
1060 1061
	stream.next_in = (unsigned char*)p->pack_base + offset;
	stream.avail_in = p->pack_size - offset;
1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075
	stream.next_out = buffer;
	stream.avail_out = size;

	inflateInit(&stream);
	st = inflate(&stream, Z_FINISH);
	inflateEnd(&stream);
	if ((st != Z_STREAM_END) || stream.total_out != size) {
		free(buffer);
		return NULL;
	}

	return buffer;
}

1076 1077
static void *unpack_delta_entry(struct packed_git *p,
				unsigned long offset,
1078 1079
				unsigned long delta_size,
				char *type,
1080
				unsigned long *sizep)
1081
{
J
Junio C Hamano 已提交
1082
	struct pack_entry base_ent;
1083 1084
	void *delta_data, *result, *base;
	unsigned long result_size, base_size;
1085
	unsigned char* base_sha1;
1086

1087
	if ((offset + 20) >= p->pack_size)
1088
		die("truncated pack file");
1089 1090

	/* The base entry _must_ be in the same pack */
1091
	base_sha1 = (unsigned char*)p->pack_base + offset;
1092 1093 1094 1095 1096 1097 1098 1099
	if (!find_pack_entry_one(base_sha1, &base_ent, p))
		die("failed to find delta-pack base object %s",
		    sha1_to_hex(base_sha1));
	base = unpack_entry_gently(&base_ent, type, &base_size);
	if (!base)
		die("failed to read delta-pack base object %s",
		    sha1_to_hex(base_sha1));

1100
	delta_data = unpack_compressed_entry(p, offset + 20, delta_size);
1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115
	result = patch_delta(base, base_size,
			     delta_data, delta_size,
			     &result_size);
	if (!result)
		die("failed to apply delta");
	free(delta_data);
	free(base);
	*sizep = result_size;
	return result;
}

static void *unpack_entry(struct pack_entry *entry,
			  char *type, unsigned long *sizep)
{
	struct packed_git *p = entry->p;
1116
	void *retval;
1117 1118 1119

	if (use_packed_git(p))
		die("cannot map packed file");
J
Junio C Hamano 已提交
1120 1121 1122
	retval = unpack_entry_gently(entry, type, sizep);
	unuse_packed_git(p);
	if (!retval)
1123
		die("corrupted pack file %s", p->pack_name);
J
Junio C Hamano 已提交
1124 1125 1126 1127 1128 1129 1130 1131
	return retval;
}

/* The caller is responsible for use_packed_git()/unuse_packed_git() pair */
void *unpack_entry_gently(struct pack_entry *entry,
			  char *type, unsigned long *sizep)
{
	struct packed_git *p = entry->p;
1132
	unsigned long offset, size;
J
Junio C Hamano 已提交
1133
	enum object_type kind;
1134

1135 1136 1137
	offset = unpack_object_header(p, entry->offset, &kind, &size);
	switch (kind) {
	case OBJ_DELTA:
1138
		return unpack_delta_entry(p, offset, size, type, sizep);
1139 1140 1141 1142
	case OBJ_COMMIT:
	case OBJ_TREE:
	case OBJ_BLOB:
	case OBJ_TAG:
1143 1144
		strcpy(type, type_names[kind]);
		*sizep = size;
1145
		return unpack_compressed_entry(p, offset, size);
1146
	default:
J
Junio C Hamano 已提交
1147
		return NULL;
1148 1149 1150
	}
}

1151 1152
int num_packed_objects(const struct packed_git *p)
{
1153
	/* See check_packed_git_idx() */
1154 1155 1156 1157 1158 1159 1160 1161 1162
	return (p->index_size - 20 - 20 - 4*256) / 24;
}

int nth_packed_object_sha1(const struct packed_git *p, int n,
			   unsigned char* sha1)
{
	void *index = p->index_base + 256;
	if (n < 0 || num_packed_objects(p) <= n)
		return -1;
1163
	hashcpy(sha1, (unsigned char *) index + (24 * n) + 4);
1164 1165 1166
	return 0;
}

J
Junio C Hamano 已提交
1167 1168
int find_pack_entry_one(const unsigned char *sha1,
			struct pack_entry *e, struct packed_git *p)
1169
{
1170
	unsigned int *level1_ofs = p->index_base;
1171 1172 1173 1174 1175 1176
	int hi = ntohl(level1_ofs[*sha1]);
	int lo = ((*sha1 == 0x0) ? 0 : ntohl(level1_ofs[*sha1 - 1]));
	void *index = p->index_base + 256;

	do {
		int mi = (lo + hi) / 2;
1177
		int cmp = hashcmp((unsigned char *)index + (24 * mi) + 4, sha1);
1178
		if (!cmp) {
1179
			e->offset = ntohl(*((unsigned int *) ((char *) index + (24 * mi))));
1180
			hashcpy(e->sha1, sha1);
1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197
			e->p = p;
			return 1;
		}
		if (cmp > 0)
			hi = mi;
		else
			lo = mi+1;
	} while (lo < hi);
	return 0;
}

static int find_pack_entry(const unsigned char *sha1, struct pack_entry *e)
{
	struct packed_git *p;
	prepare_packed_git();

	for (p = packed_git; p; p = p->next) {
J
Junio C Hamano 已提交
1198
		if (find_pack_entry_one(sha1, e, p))
1199 1200 1201 1202 1203
			return 1;
	}
	return 0;
}

1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217
struct packed_git *find_sha1_pack(const unsigned char *sha1, 
				  struct packed_git *packs)
{
	struct packed_git *p;
	struct pack_entry e;

	for (p = packs; p; p = p->next) {
		if (find_pack_entry_one(sha1, &e, p))
			return p;
	}
	return NULL;
	
}

1218
int sha1_object_info(const unsigned char *sha1, char *type, unsigned long *sizep)
1219
{
1220
	int status;
1221 1222 1223
	unsigned long mapsize, size;
	void *map;
	z_stream stream;
1224
	char hdr[128];
1225

1226
	map = map_sha1_file(sha1, &mapsize);
1227 1228 1229
	if (!map) {
		struct pack_entry e;

1230 1231 1232 1233 1234 1235
		if (find_pack_entry(sha1, &e))
			return packed_object_info(&e, type, sizep);
		reprepare_packed_git();
		if (find_pack_entry(sha1, &e))
			return packed_object_info(&e, type, sizep);
		return error("unable to find %s", sha1_to_hex(sha1));
1236
	}
1237 1238 1239 1240 1241
	if (unpack_sha1_header(&stream, map, mapsize, hdr, sizeof(hdr)) < 0)
		status = error("unable to unpack %s header",
			       sha1_to_hex(sha1));
	if (parse_sha1_header(hdr, type, &size) < 0)
		status = error("unable to parse %s header", sha1_to_hex(sha1));
1242
	else {
1243
		status = 0;
1244 1245
		if (sizep)
			*sizep = size;
1246 1247 1248 1249 1250 1251
	}
	inflateEnd(&stream);
	munmap(map, mapsize);
	return status;
}

1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262
static void *read_packed_sha1(const unsigned char *sha1, char *type, unsigned long *size)
{
	struct pack_entry e;

	if (!find_pack_entry(sha1, &e)) {
		error("cannot read sha1_file for %s", sha1_to_hex(sha1));
		return NULL;
	}
	return unpack_entry(&e, type, size);
}

1263 1264 1265 1266
void * read_sha1_file(const unsigned char *sha1, char *type, unsigned long *size)
{
	unsigned long mapsize;
	void *map, *buf;
1267
	struct pack_entry e;
1268

1269 1270
	if (find_pack_entry(sha1, &e))
		return read_packed_sha1(sha1, type, size);
1271
	map = map_sha1_file(sha1, &mapsize);
1272 1273 1274 1275 1276
	if (map) {
		buf = unpack_sha1_file(map, mapsize, type, size);
		munmap(map, mapsize);
		return buf;
	}
1277 1278 1279
	reprepare_packed_git();
	if (find_pack_entry(sha1, &e))
		return read_packed_sha1(sha1, type, size);
1280
	return NULL;
1281 1282
}

1283
void *read_object_with_reference(const unsigned char *sha1,
1284
				 const char *required_type,
1285 1286
				 unsigned long *size,
				 unsigned char *actual_sha1_return)
1287 1288 1289 1290
{
	char type[20];
	void *buffer;
	unsigned long isize;
1291
	unsigned char actual_sha1[20];
1292

1293
	hashcpy(actual_sha1, sha1);
1294 1295 1296
	while (1) {
		int ref_length = -1;
		const char *ref_type = NULL;
1297

1298 1299 1300 1301 1302 1303
		buffer = read_sha1_file(actual_sha1, type, &isize);
		if (!buffer)
			return NULL;
		if (!strcmp(type, required_type)) {
			*size = isize;
			if (actual_sha1_return)
1304
				hashcpy(actual_sha1_return, actual_sha1);
1305 1306 1307
			return buffer;
		}
		/* Handle references */
1308
		else if (!strcmp(type, commit_type))
1309
			ref_type = "tree ";
1310
		else if (!strcmp(type, tag_type))
1311 1312 1313 1314 1315 1316
			ref_type = "object ";
		else {
			free(buffer);
			return NULL;
		}
		ref_length = strlen(ref_type);
1317

1318
		if (memcmp(buffer, ref_type, ref_length) ||
1319
		    get_sha1_hex((char *) buffer + ref_length, actual_sha1)) {
1320 1321 1322
			free(buffer);
			return NULL;
		}
1323
		free(buffer);
1324 1325
		/* Now we have the ID of the referred-to object in
		 * actual_sha1.  Check again. */
1326 1327 1328
	}
}

1329 1330 1331 1332 1333 1334
char *write_sha1_file_prepare(void *buf,
			      unsigned long len,
			      const char *type,
			      unsigned char *sha1,
			      unsigned char *hdr,
			      int *hdrlen)
1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349
{
	SHA_CTX c;

	/* Generate the header */
	*hdrlen = sprintf((char *)hdr, "%s %lu", type, len)+1;

	/* Sha1.. */
	SHA1_Init(&c);
	SHA1_Update(&c, hdr, *hdrlen);
	SHA1_Update(&c, buf, len);
	SHA1_Final(sha1, &c);

	return sha1_file_name(sha1);
}

1350 1351 1352 1353 1354 1355
/*
 * Link the tempfile to the final place, possibly creating the
 * last directory level as you do so.
 *
 * Returns the errno on failure, 0 on success.
 */
1356
static int link_temp_to_file(const char *tmpfile, const char *filename)
1357 1358
{
	int ret;
S
Shawn Pearce 已提交
1359
	char *dir;
1360 1361 1362 1363 1364

	if (!link(tmpfile, filename))
		return 0;

	/*
S
Shawn Pearce 已提交
1365
	 * Try to mkdir the last path component if that failed.
1366 1367 1368 1369 1370 1371
	 *
	 * Re-try the "link()" regardless of whether the mkdir
	 * succeeds, since a race might mean that somebody
	 * else succeeded.
	 */
	ret = errno;
S
Shawn Pearce 已提交
1372 1373 1374 1375 1376 1377 1378 1379 1380 1381
	dir = strrchr(filename, '/');
	if (dir) {
		*dir = 0;
		mkdir(filename, 0777);
		if (adjust_shared_perm(filename))
			return -2;
		*dir = '/';
		if (!link(tmpfile, filename))
			return 0;
		ret = errno;
1382 1383 1384 1385 1386 1387 1388
	}
	return ret;
}

/*
 * Move the just written object into its final resting place
 */
1389
int move_temp_to_file(const char *tmpfile, const char *filename)
1390 1391
{
	int ret = link_temp_to_file(tmpfile, filename);
1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405

	/*
	 * Coda hack - coda doesn't like cross-directory links,
	 * so we fall back to a rename, which will mean that it
	 * won't be able to check collisions, but that's not a
	 * big deal.
	 *
	 * The same holds for FAT formatted media.
	 *
	 * When this succeeds, we just return 0. We have nothing
	 * left to unlink.
	 */
	if (ret && ret != EEXIST) {
		if (!rename(tmpfile, filename))
1406
			return 0;
1407
		ret = errno;
1408 1409 1410 1411
	}
	unlink(tmpfile);
	if (ret) {
		if (ret != EEXIST) {
A
Alex Riesen 已提交
1412
			fprintf(stderr, "unable to write sha1 filename %s: %s\n", filename, strerror(ret));
1413 1414 1415 1416 1417 1418 1419 1420
			return -1;
		}
		/* FIXME!!! Collision check here ? */
	}

	return 0;
}

L
Linus Torvalds 已提交
1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434
static int write_buffer(int fd, const void *buf, size_t len)
{
	while (len) {
		ssize_t size;

		size = write(fd, buf, len);
		if (!size)
			return error("file write: disk full");
		if (size < 0) {
			if (errno == EINTR || errno == EAGAIN)
				continue;
			return error("file write error (%s)", strerror(errno));
		}
		len -= size;
1435
		buf = (char *) buf + size;
L
Linus Torvalds 已提交
1436 1437 1438 1439
	}
	return 0;
}

1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482
static int write_binary_header(unsigned char *hdr, enum object_type type, unsigned long len)
{
	int hdr_len;
	unsigned char c;

	c = (type << 4) | (len & 15);
	len >>= 4;
	hdr_len = 1;
	while (len) {
		*hdr++ = c | 0x80;
		hdr_len++;
		c = (len & 0x7f);
		len >>= 7;
	}
	*hdr = c;
	return hdr_len;
}

static void setup_object_header(z_stream *stream, const char *type, unsigned long len)
{
	int obj_type, hdr;

	if (use_legacy_headers) {
		while (deflate(stream, 0) == Z_OK)
			/* nothing */;
		return;
	}
	if (!strcmp(type, blob_type))
		obj_type = OBJ_BLOB;
	else if (!strcmp(type, tree_type))
		obj_type = OBJ_TREE;
	else if (!strcmp(type, commit_type))
		obj_type = OBJ_COMMIT;
	else if (!strcmp(type, tag_type))
		obj_type = OBJ_TAG;
	else
		die("trying to generate bogus object of type '%s'", type);
	hdr = write_binary_header(stream->next_out, obj_type, len);
	stream->total_out = hdr;
	stream->next_out += hdr;
	stream->avail_out -= hdr;
}

1483
int write_sha1_file(void *buf, unsigned long len, const char *type, unsigned char *returnsha1)
1484 1485
{
	int size;
1486
	unsigned char *compressed;
1487 1488
	z_stream stream;
	unsigned char sha1[20];
1489
	char *filename;
1490
	static char tmpfile[PATH_MAX];
1491
	unsigned char hdr[50];
1492
	int fd, hdrlen;
1493

1494 1495 1496 1497
	/* Normally if we have it in the pack then we do not bother writing
	 * it out into .git/objects/??/?{38} file.
	 */
	filename = write_sha1_file_prepare(buf, len, type, sha1, hdr, &hdrlen);
1498
	if (returnsha1)
1499
		hashcpy(returnsha1, sha1);
1500 1501
	if (has_sha1_file(sha1))
		return 0;
1502 1503
	fd = open(filename, O_RDONLY);
	if (fd >= 0) {
1504
		/*
1505 1506
		 * FIXME!!! We might do collision checking here, but we'd
		 * need to uncompress the old file and check it. Later.
1507
		 */
1508
		close(fd);
1509 1510 1511
		return 0;
	}

1512
	if (errno != ENOENT) {
A
Alex Riesen 已提交
1513
		fprintf(stderr, "sha1 file %s: %s\n", filename, strerror(errno));
1514 1515 1516 1517
		return -1;
	}

	snprintf(tmpfile, sizeof(tmpfile), "%s/obj_XXXXXX", get_object_directory());
1518

1519 1520
	fd = mkstemp(tmpfile);
	if (fd < 0) {
A
Alex Riesen 已提交
1521
		fprintf(stderr, "unable to create temporary sha1 filename %s: %s\n", tmpfile, strerror(errno));
1522 1523 1524
		return -1;
	}

1525 1526
	/* Set it up */
	memset(&stream, 0, sizeof(stream));
1527
	deflateInit(&stream, zlib_compression_level);
1528
	size = 8 + deflateBound(&stream, len+hdrlen);
1529
	compressed = xmalloc(size);
1530 1531 1532 1533

	/* Compress it */
	stream.next_out = compressed;
	stream.avail_out = size;
1534 1535 1536 1537

	/* First header.. */
	stream.next_in = hdr;
	stream.avail_in = hdrlen;
1538
	setup_object_header(&stream, type, len);
1539 1540 1541 1542

	/* Then the data itself.. */
	stream.next_in = buf;
	stream.avail_in = len;
1543 1544 1545 1546 1547
	while (deflate(&stream, Z_FINISH) == Z_OK)
		/* nothing */;
	deflateEnd(&stream);
	size = stream.total_out;

L
Linus Torvalds 已提交
1548 1549
	if (write_buffer(fd, compressed, size) < 0)
		die("unable to write sha1 file");
1550
	fchmod(fd, 0444);
1551
	close(fd);
1552
	free(compressed);
1553

1554
	return move_temp_to_file(tmpfile, filename);
1555
}
1556

L
Linus Torvalds 已提交
1557 1558 1559 1560 1561
/*
 * We need to unpack and recompress the object for writing
 * it out to a different file.
 */
static void *repack_object(const unsigned char *sha1, unsigned long *objsize)
D
Daniel Barkalow 已提交
1562
{
L
Linus Torvalds 已提交
1563
	size_t size;
D
Daniel Barkalow 已提交
1564
	z_stream stream;
L
Linus Torvalds 已提交
1565 1566 1567 1568 1569 1570
	unsigned char *unpacked;
	unsigned long len;
	char type[20];
	char hdr[50];
	int hdrlen;
	void *buf;
1571

1572
	/* need to unpack and recompress it by itself */
L
Linus Torvalds 已提交
1573
	unpacked = read_packed_sha1(sha1, type, &len);
D
Daniel Barkalow 已提交
1574

L
Linus Torvalds 已提交
1575
	hdrlen = sprintf(hdr, "%s %lu", type, len) + 1;
D
Daniel Barkalow 已提交
1576

L
Linus Torvalds 已提交
1577 1578
	/* Set it up */
	memset(&stream, 0, sizeof(stream));
1579
	deflateInit(&stream, zlib_compression_level);
L
Linus Torvalds 已提交
1580 1581
	size = deflateBound(&stream, len + hdrlen);
	buf = xmalloc(size);
D
Daniel Barkalow 已提交
1582

L
Linus Torvalds 已提交
1583 1584 1585
	/* Compress it */
	stream.next_out = buf;
	stream.avail_out = size;
D
Daniel Barkalow 已提交
1586

L
Linus Torvalds 已提交
1587 1588 1589 1590 1591
	/* First header.. */
	stream.next_in = (void *)hdr;
	stream.avail_in = hdrlen;
	while (deflate(&stream, 0) == Z_OK)
		/* nothing */;
1592

L
Linus Torvalds 已提交
1593 1594 1595 1596 1597 1598 1599
	/* Then the data itself.. */
	stream.next_in = unpacked;
	stream.avail_in = len;
	while (deflate(&stream, Z_FINISH) == Z_OK)
		/* nothing */;
	deflateEnd(&stream);
	free(unpacked);
1600

L
Linus Torvalds 已提交
1601 1602 1603 1604 1605 1606 1607 1608
	*objsize = stream.total_out;
	return buf;
}

int write_sha1_to_fd(int fd, const unsigned char *sha1)
{
	int retval;
	unsigned long objsize;
1609
	void *buf = map_sha1_file(sha1, &objsize);
L
Linus Torvalds 已提交
1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620

	if (buf) {
		retval = write_buffer(fd, buf, objsize);
		munmap(buf, objsize);
		return retval;
	}

	buf = repack_object(sha1, &objsize);
	retval = write_buffer(fd, buf, objsize);
	free(buf);
	return retval;
D
Daniel Barkalow 已提交
1621 1622
}

1623 1624
int write_sha1_from_fd(const unsigned char *sha1, int fd, char *buffer,
		       size_t bufsize, size_t *bufposn)
1625
{
1626
	char tmpfile[PATH_MAX];
1627 1628 1629
	int local;
	z_stream stream;
	unsigned char real_sha1[20];
1630
	unsigned char discard[4096];
1631 1632 1633
	int ret;
	SHA_CTX c;

1634
	snprintf(tmpfile, sizeof(tmpfile), "%s/obj_XXXXXX", get_object_directory());
1635

1636
	local = mkstemp(tmpfile);
1637
	if (local < 0)
1638 1639
		return error("Couldn't open %s for %s",
			     tmpfile, sha1_to_hex(sha1));
1640 1641 1642 1643 1644 1645 1646 1647 1648

	memset(&stream, 0, sizeof(stream));

	inflateInit(&stream);

	SHA1_Init(&c);

	do {
		ssize_t size;
1649 1650
		if (*bufposn) {
			stream.avail_in = *bufposn;
P
Pavel Roskin 已提交
1651
			stream.next_in = (unsigned char *) buffer;
1652 1653 1654 1655 1656 1657 1658
			do {
				stream.next_out = discard;
				stream.avail_out = sizeof(discard);
				ret = inflate(&stream, Z_SYNC_FLUSH);
				SHA1_Update(&c, discard, sizeof(discard) -
					    stream.avail_out);
			} while (stream.avail_in && ret == Z_OK);
L
Linus Torvalds 已提交
1659 1660
			if (write_buffer(local, buffer, *bufposn - stream.avail_in) < 0)
				die("unable to write sha1 file");
1661 1662 1663 1664 1665 1666 1667
			memmove(buffer, buffer + *bufposn - stream.avail_in,
				stream.avail_in);
			*bufposn = stream.avail_in;
			if (ret != Z_OK)
				break;
		}
		size = read(fd, buffer + *bufposn, bufsize - *bufposn);
1668 1669
		if (size <= 0) {
			close(local);
1670
			unlink(tmpfile);
1671 1672 1673 1674 1675
			if (!size)
				return error("Connection closed?");
			perror("Reading from connection");
			return -1;
		}
1676 1677
		*bufposn += size;
	} while (1);
1678 1679 1680 1681 1682
	inflateEnd(&stream);

	close(local);
	SHA1_Final(real_sha1, &c);
	if (ret != Z_STREAM_END) {
1683
		unlink(tmpfile);
1684 1685
		return error("File %s corrupted", sha1_to_hex(sha1));
	}
1686
	if (hashcmp(sha1, real_sha1)) {
1687
		unlink(tmpfile);
1688
		return error("File %s has bad hash", sha1_to_hex(sha1));
1689
	}
1690 1691

	return move_temp_to_file(tmpfile, sha1_file_name(sha1));
1692 1693
}

1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709
int has_pack_index(const unsigned char *sha1)
{
	struct stat st;
	if (stat(sha1_pack_index_name(sha1), &st))
		return 0;
	return 1;
}

int has_pack_file(const unsigned char *sha1)
{
	struct stat st;
	if (stat(sha1_pack_name(sha1), &st))
		return 0;
	return 1;
}

1710 1711 1712 1713 1714 1715
int has_sha1_pack(const unsigned char *sha1)
{
	struct pack_entry e;
	return find_pack_entry(sha1, &e);
}

1716 1717 1718
int has_sha1_file(const unsigned char *sha1)
{
	struct stat st;
1719 1720
	struct pack_entry e;

1721
	if (find_pack_entry(sha1, &e))
1722
		return 1;
1723
	return find_sha1_file(sha1, &st) ? 1 : 0;
1724
}
J
Junio C Hamano 已提交
1725

1726 1727
/*
 * reads from fd as long as possible into a supplied buffer of size bytes.
1728
 * If necessary the buffer's size is increased using realloc()
1729 1730 1731 1732 1733 1734 1735
 *
 * returns 0 if anything went fine and -1 otherwise
 *
 * NOTE: both buf and size may change, but even when -1 is returned
 * you still have to free() it yourself.
 */
int read_pipe(int fd, char** return_buf, unsigned long* return_size)
1736
{
1737 1738 1739
	char* buf = *return_buf;
	unsigned long size = *return_size;
	int iret;
1740
	unsigned long off = 0;
1741

1742
	do {
1743
		iret = xread(fd, buf + off, size - off);
1744 1745 1746 1747
		if (iret > 0) {
			off += iret;
			if (off == size) {
				size *= 2;
J
Jonas Fonseca 已提交
1748
				buf = xrealloc(buf, size);
1749 1750 1751
			}
		}
	} while (iret > 0);
1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763

	*return_buf = buf;
	*return_size = off;

	if (iret < 0)
		return -1;
	return 0;
}

int index_pipe(unsigned char *sha1, int fd, const char *type, int write_object)
{
	unsigned long size = 4096;
J
Jonas Fonseca 已提交
1764
	char *buf = xmalloc(size);
1765 1766 1767 1768 1769
	int ret;
	unsigned char hdr[50];
	int hdrlen;

	if (read_pipe(fd, &buf, &size)) {
1770 1771 1772
		free(buf);
		return -1;
	}
1773

1774
	if (!type)
1775
		type = blob_type;
1776
	if (write_object)
1777
		ret = write_sha1_file(buf, size, type, sha1);
1778
	else {
1779
		write_sha1_file_prepare(buf, size, type, sha1, hdr, &hdrlen);
1780 1781 1782 1783 1784 1785
		ret = 0;
	}
	free(buf);
	return ret;
}

1786
int index_fd(unsigned char *sha1, int fd, struct stat *st, int write_object, const char *type)
J
Junio C Hamano 已提交
1787 1788
{
	unsigned long size = st->st_size;
1789 1790
	void *buf;
	int ret;
1791 1792
	unsigned char hdr[50];
	int hdrlen;
J
Junio C Hamano 已提交
1793

1794
	buf = "";
J
Junio C Hamano 已提交
1795
	if (size)
1796
		buf = mmap(NULL, size, PROT_READ, MAP_PRIVATE, fd, 0);
J
Junio C Hamano 已提交
1797
	close(fd);
P
Pavel Roskin 已提交
1798
	if (buf == MAP_FAILED)
J
Junio C Hamano 已提交
1799 1800
		return -1;

1801
	if (!type)
1802
		type = blob_type;
1803 1804 1805 1806 1807 1808
	if (write_object)
		ret = write_sha1_file(buf, size, type, sha1);
	else {
		write_sha1_file_prepare(buf, size, type, sha1, hdr, &hdrlen);
		ret = 0;
	}
1809 1810 1811
	if (size)
		munmap(buf, size);
	return ret;
J
Junio C Hamano 已提交
1812
}
1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839

int index_path(unsigned char *sha1, const char *path, struct stat *st, int write_object)
{
	int fd;
	char *target;

	switch (st->st_mode & S_IFMT) {
	case S_IFREG:
		fd = open(path, O_RDONLY);
		if (fd < 0)
			return error("open(\"%s\"): %s", path,
				     strerror(errno));
		if (index_fd(sha1, fd, st, write_object, NULL) < 0)
			return error("%s: failed to insert into database",
				     path);
		break;
	case S_IFLNK:
		target = xmalloc(st->st_size+1);
		if (readlink(path, target, st->st_size+1) != st->st_size) {
			char *errstr = strerror(errno);
			free(target);
			return error("readlink(\"%s\"): %s", path,
			             errstr);
		}
		if (!write_object) {
			unsigned char hdr[50];
			int hdrlen;
1840
			write_sha1_file_prepare(target, st->st_size, blob_type,
1841
						sha1, hdr, &hdrlen);
1842
		} else if (write_sha1_file(target, st->st_size, blob_type, sha1))
1843 1844 1845 1846 1847 1848 1849 1850 1851
			return error("%s: failed to insert into database",
				     path);
		free(target);
		break;
	default:
		return error("%s: unsupported file type", path);
	}
	return 0;
}