sha1_file.c 40.0 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

J
Junio C Hamano 已提交
25 26
const unsigned char null_sha1[20] = { 0, };

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];
	}
}

116 117 118 119
/*
 * NOTE! This returns a statically allocated buffer, so you have to be
 * careful about using it. Do a "strdup()" if you need to save the
 * 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;
	}
462 463 464 465 466 467
	if (!p->pack_base) {
		int fd;
		struct stat st;
		void *map;

		pack_mapped += p->pack_size;
468 469
		while (PACK_MAX_SZ < pack_mapped && unuse_one_packed_git())
			; /* nothing */
470 471
		fd = open(p->pack_name, O_RDONLY);
		if (fd < 0)
472
			die("packfile %s cannot be opened", p->pack_name);
473 474
		if (fstat(fd, &st)) {
			close(fd);
475
			die("packfile %s cannot be opened", p->pack_name);
476 477
		}
		if (st.st_size != p->pack_size)
478
			die("packfile %s size mismatch.", p->pack_name);
479 480 481
		map = mmap(NULL, p->pack_size, PROT_READ, MAP_PRIVATE, fd, 0);
		close(fd);
		if (map == MAP_FAILED)
482
			die("packfile %s cannot be mapped.", p->pack_name);
483
		p->pack_base = map;
484 485 486 487 488

		/* Check if the pack file matches with the index file.
		 * this is cheap.
		 */
		if (memcmp((char*)(p->index_base) + p->index_size - 40,
489 490 491
			   (char *) p->pack_base + p->pack_size - 20,
			   20)) {

492
			die("packfile %s does not match index.", p->pack_name);
493
		}
494 495
	}
	p->pack_last_used = pack_used_ctr++;
496
	p->pack_use_cnt++;
497 498 499
	return 0;
}

500
struct packed_git *add_packed_git(char *path, int path_len, int local)
501 502 503 504 505
{
	struct stat st;
	struct packed_git *p;
	unsigned long idx_size;
	void *idx_map;
506
	unsigned char sha1[20];
507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525

	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;
526
	p->pack_base = NULL;
527
	p->pack_last_used = 0;
528
	p->pack_use_cnt = 0;
529
	p->pack_local = local;
P
Pavel Roskin 已提交
530
	if ((path_len > 44) && !get_sha1_hex(path + path_len - 44, sha1))
L
Lukas_Sandström 已提交
531
		memcpy(p->sha1, sha1, 20);
532 533 534
	return p;
}

535
struct packed_git *parse_pack_index(unsigned char *sha1)
536 537 538 539 540
{
	char *path = sha1_pack_index_name(sha1);
	return parse_pack_index_file(sha1, path);
}

541
struct packed_git *parse_pack_index_file(const unsigned char *sha1, char *idx_path)
542 543 544 545
{
	struct packed_git *p;
	unsigned long idx_size;
	void *idx_map;
546
	char *path;
547

548
	if (check_packed_git_idx(idx_path, &idx_size, &idx_map))
549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571
		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;
	memcpy(p->sha1, sha1, 20);
	return p;
}

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

572
static void prepare_packed_git_one(char *objdir, int local)
573 574 575 576 577 578 579 580 581
{
	char path[PATH_MAX];
	int len;
	DIR *dir;
	struct dirent *de;

	sprintf(path, "%s/pack", objdir);
	len = strlen(path);
	dir = opendir(path);
582
	if (!dir) {
J
Junio C Hamano 已提交
583
		if (errno != ENOENT)
584
			error("unable to open object pack directory: %s: %s",
J
Junio C Hamano 已提交
585
			      path, strerror(errno));
586
		return;
587
	}
588 589 590 591 592 593 594 595 596 597
	path[len++] = '/';
	while ((de = readdir(dir)) != NULL) {
		int namelen = strlen(de->d_name);
		struct packed_git *p;

		if (strcmp(de->d_name + namelen - 4, ".idx"))
			continue;

		/* we have .idx.  Is it a file we can map? */
		strcpy(path + len, de->d_name);
598 599 600 601 602 603
		for (p = packed_git; p; p = p->next) {
			if (!memcmp(path, p->pack_name, len + namelen - 4))
				break;
		}
		if (p)
			continue;
604
		p = add_packed_git(path, len + namelen, local);
605 606 607 608 609
		if (!p)
			continue;
		p->next = packed_git;
		packed_git = p;
	}
610
	closedir(dir);
611 612
}

613
static int prepare_packed_git_run_once = 0;
614
void prepare_packed_git(void)
615
{
616
	struct alternate_object_database *alt;
617

618
	if (prepare_packed_git_run_once)
619
		return;
620
	prepare_packed_git_one(get_object_directory(), 1);
621
	prepare_alt_odb();
622
	for (alt = alt_odb_list; alt; alt = alt->next) {
623
		alt->name[-1] = 0;
624
		prepare_packed_git_one(alt->base, 0);
625
		alt->name[-1] = '/';
626
	}
627 628 629 630 631 632 633
	prepare_packed_git_run_once = 1;
}

static void reprepare_packed_git(void)
{
	prepare_packed_git_run_once = 0;
	prepare_packed_git();
634 635
}

636
int check_sha1_signature(const unsigned char *sha1, void *map, unsigned long size, const char *type)
637
{
638
	char header[100];
639 640 641 642
	unsigned char real_sha1[20];
	SHA_CTX c;

	SHA1_Init(&c);
643
	SHA1_Update(&c, header, 1+sprintf(header, "%s %lu", type, size));
644 645 646 647 648
	SHA1_Update(&c, map, size);
	SHA1_Final(real_sha1, &c);
	return memcmp(sha1, real_sha1, 20) ? -1 : 0;
}

649
static void *map_sha1_file_internal(const unsigned char *sha1,
D
Daniel Barkalow 已提交
650
				    unsigned long *size)
651 652 653
{
	struct stat st;
	void *map;
654
	int fd;
655 656 657 658 659
	char *filename = find_sha1_file(sha1, &st);

	if (!filename) {
		return NULL;
	}
660

661
	fd = open(filename, O_RDONLY | sha1_file_open_flag);
662
	if (fd < 0) {
663 664 665 666 667 668 669 670 671 672 673
		/* 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;
		}

674 675 676
		/* If it failed once, it will probably fail again.
		 * Stop using O_NOATIME
		 */
677
		sha1_file_open_flag = 0;
678 679 680
	}
	map = mmap(NULL, st.st_size, PROT_READ, MAP_PRIVATE, fd, 0);
	close(fd);
P
Pavel Roskin 已提交
681
	if (map == MAP_FAILED)
682 683 684 685 686
		return NULL;
	*size = st.st_size;
	return map;
}

687
static int unpack_sha1_header(z_stream *stream, void *map, unsigned long mapsize, void *buffer, unsigned long size)
688 689 690 691 692 693 694 695 696 697 698 699
{
	/* Get the data stream */
	memset(stream, 0, sizeof(*stream));
	stream->next_in = map;
	stream->avail_in = mapsize;
	stream->next_out = buffer;
	stream->avail_out = size;

	inflateInit(stream);
	return inflate(stream, 0);
}

L
Linus Torvalds 已提交
700
static void *unpack_sha1_rest(z_stream *stream, void *buffer, unsigned long size)
701 702
{
	int bytes = strlen(buffer) + 1;
703
	unsigned char *buf = xmalloc(1+size);
704

705
	memcpy(buf, (char *) buffer + bytes, stream->total_out - bytes);
706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722
	bytes = stream->total_out - bytes;
	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.
 */
723
static int parse_sha1_header(char *hdr, char *type, unsigned long *sizep)
724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767
{
	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;
}

768 769
void * unpack_sha1_file(void *map, unsigned long mapsize, char *type, unsigned long *size)
{
770
	int ret;
771
	z_stream stream;
772
	char hdr[8192];
773

774 775
	ret = unpack_sha1_header(&stream, map, mapsize, hdr, sizeof(hdr));
	if (ret < Z_OK || parse_sha1_header(hdr, type, size) < 0)
776 777
		return NULL;

778
	return unpack_sha1_rest(&stream, hdr, *size);
779 780
}

J
Junio C Hamano 已提交
781 782 783 784
/* forward declaration for a mutually recursive function */
static int packed_object_info(struct pack_entry *entry,
			      char *type, unsigned long *sizep);

785 786 787 788
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
789 790
			     unsigned long *sizep,
			     struct packed_git *p)
791
{
J
Junio C Hamano 已提交
792 793
	struct pack_entry base_ent;

794 795 796
	if (left < 20)
		die("truncated pack file");

J
Junio C Hamano 已提交
797 798 799 800 801
	/* 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));

802 803 804 805 806
	/* 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.
	 */
807

J
Junio C Hamano 已提交
808
	if (packed_object_info(&base_ent, type, NULL))
809
		die("cannot get info for delta-pack base");
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
	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;
836 837 838

		/* ignore base size */
		get_delta_hdr_size(&data, delta_head+sizeof(delta_head));
839

840
		/* Read the result size */
841
		result_size = get_delta_hdr_size(&data, delta_head+sizeof(delta_head));
842 843
		*sizep = result_size;
	}
844 845 846
	return 0;
}

847 848 849
static unsigned long unpack_object_header(struct packed_git *p, unsigned long offset,
	enum object_type *type, unsigned long *sizep)
{
850
	unsigned shift;
851 852 853 854 855 856
	unsigned char *pack, c;
	unsigned long size;

	if (offset >= p->pack_size)
		die("object offset outside of pack file");

857
	pack =  (unsigned char *) p->pack_base + offset;
858 859 860 861
	c = *pack++;
	offset++;
	*type = (c >> 4) & 7;
	size = c & 15;
862
	shift = 4;
863 864 865 866 867
	while (c & 0x80) {
		if (offset >= p->pack_size)
			die("object offset outside of pack file");
		c = *pack++;
		offset++;
868 869
		size += (c & 0x7f) << shift;
		shift += 7;
870 871 872 873 874
	}
	*sizep = size;
	return offset;
}

875 876 877 878 879 880 881 882 883 884 885 886
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;
887
	memcpy(base, (char *) p->pack_base + ptr, 20);
888 889 890 891 892 893
	status = 0;
 done:
	unuse_packed_git(p);
	return status;
}

894 895 896 897
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
898
			       unsigned int *delta_chain_length,
899 900 901
			       unsigned char *base_sha1)
{
	struct packed_git *p = e->p;
902
	unsigned long offset;
903 904 905 906
	unsigned char *pack;
	enum object_type kind;

	offset = unpack_object_header(p, e->offset, &kind, size);
907
	pack = (unsigned char *) p->pack_base + offset;
908 909 910
	if (kind != OBJ_DELTA)
		*delta_chain_length = 0;
	else {
J
Junio C Hamano 已提交
911
		unsigned int chain_length = 0;
912 913 914
		if (p->pack_size <= offset + 20)
			die("pack file %s records an incomplete delta base",
			    p->pack_name);
915 916 917 918 919 920 921 922
		memcpy(base_sha1, pack, 20);
		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);
923
			pack = (unsigned char *) p->pack_base + offset;
924 925 926 927 928 929
			chain_length++;
		} while (kind == OBJ_DELTA);
		*delta_chain_length = chain_length;
	}
	switch (kind) {
	case OBJ_COMMIT:
930
		strcpy(type, commit_type);
931 932
		break;
	case OBJ_TREE:
933
		strcpy(type, tree_type);
934 935
		break;
	case OBJ_BLOB:
936
		strcpy(type, blob_type);
937 938
		break;
	case OBJ_TAG:
939
		strcpy(type, tag_type);
940 941
		break;
	default:
942 943
		die("corrupted pack file %s containing object of kind %d",
		    p->pack_name, kind);
944 945 946 947
	}
	*store_size = 0; /* notyet */
}

948 949 950 951 952 953
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;
954
	enum object_type kind;
955
	int retval;
956 957 958 959

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

960
	offset = unpack_object_header(p, entry->offset, &kind, &size);
961
	pack = (unsigned char *) p->pack_base + offset;
962 963 964 965
	left = p->pack_size - offset;

	switch (kind) {
	case OBJ_DELTA:
J
Junio C Hamano 已提交
966
		retval = packed_delta_info(pack, size, left, type, sizep, p);
967 968
		unuse_packed_git(p);
		return retval;
969
	case OBJ_COMMIT:
970
		strcpy(type, commit_type);
971
		break;
972
	case OBJ_TREE:
973
		strcpy(type, tree_type);
974
		break;
975
	case OBJ_BLOB:
976
		strcpy(type, blob_type);
977
		break;
978
	case OBJ_TAG:
979
		strcpy(type, tag_type);
980
		break;
981
	default:
982 983
		die("corrupted pack file %s containing object of kind %d",
		    p->pack_name, kind);
984
	}
985 986
	if (sizep)
		*sizep = size;
987
	unuse_packed_git(p);
988 989 990 991 992 993 994 995 996 997
	return 0;
}

/* forward declaration for a mutually recursive function */
static void *unpack_entry(struct pack_entry *, char *, unsigned long *);

static void *unpack_delta_entry(unsigned char *base_sha1,
				unsigned long delta_size,
				unsigned long left,
				char *type,
J
Junio C Hamano 已提交
998 999
				unsigned long *sizep,
				struct packed_git *p)
1000
{
J
Junio C Hamano 已提交
1001
	struct pack_entry base_ent;
1002 1003 1004 1005 1006 1007 1008
	void *data, *delta_data, *result, *base;
	unsigned long data_size, result_size, base_size;
	z_stream stream;
	int st;

	if (left < 20)
		die("truncated pack file");
1009 1010 1011 1012 1013 1014 1015 1016 1017 1018

	/* 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));
	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));

1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052
	data = base_sha1 + 20;
	data_size = left - 20;
	delta_data = xmalloc(delta_size);

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

	stream.next_in = data;
	stream.avail_in = data_size;
	stream.next_out = delta_data;
	stream.avail_out = delta_size;

	inflateInit(&stream);
	st = inflate(&stream, Z_FINISH);
	inflateEnd(&stream);
	if ((st != Z_STREAM_END) || stream.total_out != delta_size)
		die("delta data unpack failed");

	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_non_delta_entry(unsigned char *data,
				    unsigned long size,
				    unsigned long left)
{
	int st;
	z_stream stream;
1053
	unsigned char *buffer;
1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077

	buffer = xmalloc(size + 1);
	buffer[size] = 0;
	memset(&stream, 0, sizeof(stream));
	stream.next_in = data;
	stream.avail_in = left;
	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;
}

static void *unpack_entry(struct pack_entry *entry,
			  char *type, unsigned long *sizep)
{
	struct packed_git *p = entry->p;
1078
	void *retval;
1079 1080 1081

	if (use_packed_git(p))
		die("cannot map packed file");
J
Junio C Hamano 已提交
1082 1083 1084
	retval = unpack_entry_gently(entry, type, sizep);
	unuse_packed_git(p);
	if (!retval)
1085
		die("corrupted pack file %s", p->pack_name);
J
Junio C Hamano 已提交
1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097
	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;
	unsigned long offset, size, left;
	unsigned char *pack;
	enum object_type kind;
	void *retval;
1098

1099
	offset = unpack_object_header(p, entry->offset, &kind, &size);
1100
	pack = (unsigned char *) p->pack_base + offset;
1101 1102 1103
	left = p->pack_size - offset;
	switch (kind) {
	case OBJ_DELTA:
J
Junio C Hamano 已提交
1104
		retval = unpack_delta_entry(pack, size, left, type, sizep, p);
1105
		return retval;
1106
	case OBJ_COMMIT:
1107
		strcpy(type, commit_type);
1108
		break;
1109
	case OBJ_TREE:
1110
		strcpy(type, tree_type);
1111
		break;
1112
	case OBJ_BLOB:
1113
		strcpy(type, blob_type);
1114
		break;
1115
	case OBJ_TAG:
1116
		strcpy(type, tag_type);
1117
		break;
1118
	default:
J
Junio C Hamano 已提交
1119
		return NULL;
1120 1121
	}
	*sizep = size;
1122 1123
	retval = unpack_non_delta_entry(pack, size, left);
	return retval;
1124 1125
}

1126 1127
int num_packed_objects(const struct packed_git *p)
{
1128
	/* See check_packed_git_idx() */
1129 1130 1131 1132 1133 1134 1135 1136 1137
	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;
1138
	memcpy(sha1, (char *) index + (24 * n) + 4, 20);
1139 1140 1141
	return 0;
}

J
Junio C Hamano 已提交
1142 1143
int find_pack_entry_one(const unsigned char *sha1,
			struct pack_entry *e, struct packed_git *p)
1144
{
1145
	unsigned int *level1_ofs = p->index_base;
1146 1147 1148 1149 1150 1151
	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;
1152
		int cmp = memcmp((char *) index + (24 * mi) + 4, sha1, 20);
1153
		if (!cmp) {
1154
			e->offset = ntohl(*((unsigned int *) ((char *) index + (24 * mi))));
1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172
			memcpy(e->sha1, sha1, 20);
			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
1173
		if (find_pack_entry_one(sha1, e, p))
1174 1175 1176 1177 1178
			return 1;
	}
	return 0;
}

1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192
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;
	
}

1193
int sha1_object_info(const unsigned char *sha1, char *type, unsigned long *sizep)
1194
{
1195
	int status;
1196 1197 1198
	unsigned long mapsize, size;
	void *map;
	z_stream stream;
1199
	char hdr[128];
1200

D
Daniel Barkalow 已提交
1201
	map = map_sha1_file_internal(sha1, &mapsize);
1202 1203 1204
	if (!map) {
		struct pack_entry e;

1205 1206 1207 1208 1209 1210
		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));
1211
	}
1212 1213 1214 1215 1216
	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));
1217
	else {
1218
		status = 0;
1219 1220
		if (sizep)
			*sizep = size;
1221 1222 1223 1224 1225 1226
	}
	inflateEnd(&stream);
	munmap(map, mapsize);
	return status;
}

1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237
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);
}

1238 1239 1240 1241
void * read_sha1_file(const unsigned char *sha1, char *type, unsigned long *size)
{
	unsigned long mapsize;
	void *map, *buf;
1242
	struct pack_entry e;
1243

1244 1245
	if (find_pack_entry(sha1, &e))
		return read_packed_sha1(sha1, type, size);
D
Daniel Barkalow 已提交
1246
	map = map_sha1_file_internal(sha1, &mapsize);
1247 1248 1249 1250 1251
	if (map) {
		buf = unpack_sha1_file(map, mapsize, type, size);
		munmap(map, mapsize);
		return buf;
	}
1252 1253 1254
	reprepare_packed_git();
	if (find_pack_entry(sha1, &e))
		return read_packed_sha1(sha1, type, size);
1255
	return NULL;
1256 1257
}

1258
void *read_object_with_reference(const unsigned char *sha1,
1259
				 const char *required_type,
1260 1261
				 unsigned long *size,
				 unsigned char *actual_sha1_return)
1262 1263 1264 1265
{
	char type[20];
	void *buffer;
	unsigned long isize;
1266
	unsigned char actual_sha1[20];
1267

1268 1269 1270 1271
	memcpy(actual_sha1, sha1, 20);
	while (1) {
		int ref_length = -1;
		const char *ref_type = NULL;
1272

1273 1274 1275 1276 1277 1278 1279 1280 1281 1282
		buffer = read_sha1_file(actual_sha1, type, &isize);
		if (!buffer)
			return NULL;
		if (!strcmp(type, required_type)) {
			*size = isize;
			if (actual_sha1_return)
				memcpy(actual_sha1_return, actual_sha1, 20);
			return buffer;
		}
		/* Handle references */
1283
		else if (!strcmp(type, commit_type))
1284
			ref_type = "tree ";
1285
		else if (!strcmp(type, tag_type))
1286 1287 1288 1289 1290 1291
			ref_type = "object ";
		else {
			free(buffer);
			return NULL;
		}
		ref_length = strlen(ref_type);
1292

1293
		if (memcmp(buffer, ref_type, ref_length) ||
1294
		    get_sha1_hex((char *) buffer + ref_length, actual_sha1)) {
1295 1296 1297
			free(buffer);
			return NULL;
		}
1298
		free(buffer);
1299 1300
		/* Now we have the ID of the referred-to object in
		 * actual_sha1.  Check again. */
1301 1302 1303
	}
}

1304 1305 1306 1307 1308 1309
char *write_sha1_file_prepare(void *buf,
			      unsigned long len,
			      const char *type,
			      unsigned char *sha1,
			      unsigned char *hdr,
			      int *hdrlen)
1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324
{
	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);
}

1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351
/*
 * 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.
 */
static int link_temp_to_file(const char *tmpfile, char *filename)
{
	int ret;

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

	/*
	 * Try to mkdir the last path component if that failed
	 * with an ENOENT.
	 *
	 * Re-try the "link()" regardless of whether the mkdir
	 * succeeds, since a race might mean that somebody
	 * else succeeded.
	 */
	ret = errno;
	if (ret == ENOENT) {
		char *dir = strrchr(filename, '/');
		if (dir) {
			*dir = 0;
			mkdir(filename, 0777);
1352 1353
			if (adjust_shared_perm(filename))
				return -2;
1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365
			*dir = '/';
			if (!link(tmpfile, filename))
				return 0;
			ret = errno;
		}
	}
	return ret;
}

/*
 * Move the just written object into its final resting place
 */
1366
int move_temp_to_file(const char *tmpfile, char *filename)
1367 1368
{
	int ret = link_temp_to_file(tmpfile, filename);
1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382

	/*
	 * 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))
1383
			return 0;
1384
		ret = errno;
1385 1386 1387 1388
	}
	unlink(tmpfile);
	if (ret) {
		if (ret != EEXIST) {
A
Alex Riesen 已提交
1389
			fprintf(stderr, "unable to write sha1 filename %s: %s\n", filename, strerror(ret));
1390 1391 1392 1393 1394 1395 1396 1397
			return -1;
		}
		/* FIXME!!! Collision check here ? */
	}

	return 0;
}

L
Linus Torvalds 已提交
1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411
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;
1412
		buf = (char *) buf + size;
L
Linus Torvalds 已提交
1413 1414 1415 1416
	}
	return 0;
}

1417
int write_sha1_file(void *buf, unsigned long len, const char *type, unsigned char *returnsha1)
1418 1419
{
	int size;
1420
	unsigned char *compressed;
1421 1422
	z_stream stream;
	unsigned char sha1[20];
1423
	char *filename;
1424
	static char tmpfile[PATH_MAX];
1425
	unsigned char hdr[50];
1426
	int fd, hdrlen;
1427

1428 1429 1430 1431
	/* 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);
1432 1433
	if (returnsha1)
		memcpy(returnsha1, sha1, 20);
1434 1435
	if (has_sha1_file(sha1))
		return 0;
1436 1437
	fd = open(filename, O_RDONLY);
	if (fd >= 0) {
1438
		/*
1439 1440
		 * FIXME!!! We might do collision checking here, but we'd
		 * need to uncompress the old file and check it. Later.
1441
		 */
1442
		close(fd);
1443 1444 1445
		return 0;
	}

1446
	if (errno != ENOENT) {
A
Alex Riesen 已提交
1447
		fprintf(stderr, "sha1 file %s: %s\n", filename, strerror(errno));
1448 1449 1450 1451
		return -1;
	}

	snprintf(tmpfile, sizeof(tmpfile), "%s/obj_XXXXXX", get_object_directory());
1452

1453 1454
	fd = mkstemp(tmpfile);
	if (fd < 0) {
A
Alex Riesen 已提交
1455
		fprintf(stderr, "unable to create temporary sha1 filename %s: %s\n", tmpfile, strerror(errno));
1456 1457 1458
		return -1;
	}

1459 1460
	/* Set it up */
	memset(&stream, 0, sizeof(stream));
1461
	deflateInit(&stream, zlib_compression_level);
1462
	size = deflateBound(&stream, len+hdrlen);
1463
	compressed = xmalloc(size);
1464 1465 1466 1467

	/* Compress it */
	stream.next_out = compressed;
	stream.avail_out = size;
1468 1469 1470 1471 1472

	/* First header.. */
	stream.next_in = hdr;
	stream.avail_in = hdrlen;
	while (deflate(&stream, 0) == Z_OK)
1473
		/* nothing */;
1474 1475 1476 1477

	/* Then the data itself.. */
	stream.next_in = buf;
	stream.avail_in = len;
1478 1479 1480 1481 1482
	while (deflate(&stream, Z_FINISH) == Z_OK)
		/* nothing */;
	deflateEnd(&stream);
	size = stream.total_out;

L
Linus Torvalds 已提交
1483 1484
	if (write_buffer(fd, compressed, size) < 0)
		die("unable to write sha1 file");
1485
	fchmod(fd, 0444);
1486
	close(fd);
1487
	free(compressed);
1488

1489
	return move_temp_to_file(tmpfile, filename);
1490
}
1491

L
Linus Torvalds 已提交
1492 1493 1494 1495 1496
/*
 * 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
1497
{
L
Linus Torvalds 已提交
1498
	size_t size;
D
Daniel Barkalow 已提交
1499
	z_stream stream;
L
Linus Torvalds 已提交
1500 1501 1502 1503 1504 1505
	unsigned char *unpacked;
	unsigned long len;
	char type[20];
	char hdr[50];
	int hdrlen;
	void *buf;
1506

1507
	/* need to unpack and recompress it by itself */
L
Linus Torvalds 已提交
1508
	unpacked = read_packed_sha1(sha1, type, &len);
D
Daniel Barkalow 已提交
1509

L
Linus Torvalds 已提交
1510
	hdrlen = sprintf(hdr, "%s %lu", type, len) + 1;
D
Daniel Barkalow 已提交
1511

L
Linus Torvalds 已提交
1512 1513
	/* Set it up */
	memset(&stream, 0, sizeof(stream));
1514
	deflateInit(&stream, zlib_compression_level);
L
Linus Torvalds 已提交
1515 1516
	size = deflateBound(&stream, len + hdrlen);
	buf = xmalloc(size);
D
Daniel Barkalow 已提交
1517

L
Linus Torvalds 已提交
1518 1519 1520
	/* Compress it */
	stream.next_out = buf;
	stream.avail_out = size;
D
Daniel Barkalow 已提交
1521

L
Linus Torvalds 已提交
1522 1523 1524 1525 1526
	/* First header.. */
	stream.next_in = (void *)hdr;
	stream.avail_in = hdrlen;
	while (deflate(&stream, 0) == Z_OK)
		/* nothing */;
1527

L
Linus Torvalds 已提交
1528 1529 1530 1531 1532 1533 1534
	/* Then the data itself.. */
	stream.next_in = unpacked;
	stream.avail_in = len;
	while (deflate(&stream, Z_FINISH) == Z_OK)
		/* nothing */;
	deflateEnd(&stream);
	free(unpacked);
1535

L
Linus Torvalds 已提交
1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555
	*objsize = stream.total_out;
	return buf;
}

int write_sha1_to_fd(int fd, const unsigned char *sha1)
{
	int retval;
	unsigned long objsize;
	void *buf = map_sha1_file_internal(sha1, &objsize);

	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
1556 1557
}

1558 1559
int write_sha1_from_fd(const unsigned char *sha1, int fd, char *buffer,
		       size_t bufsize, size_t *bufposn)
1560
{
1561
	char tmpfile[PATH_MAX];
1562 1563 1564
	int local;
	z_stream stream;
	unsigned char real_sha1[20];
1565
	unsigned char discard[4096];
1566 1567 1568
	int ret;
	SHA_CTX c;

1569
	snprintf(tmpfile, sizeof(tmpfile), "%s/obj_XXXXXX", get_object_directory());
1570

1571
	local = mkstemp(tmpfile);
1572
	if (local < 0)
1573 1574
		return error("Couldn't open %s for %s",
			     tmpfile, sha1_to_hex(sha1));
1575 1576 1577 1578 1579 1580 1581 1582 1583

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

	inflateInit(&stream);

	SHA1_Init(&c);

	do {
		ssize_t size;
1584 1585
		if (*bufposn) {
			stream.avail_in = *bufposn;
P
Pavel Roskin 已提交
1586
			stream.next_in = (unsigned char *) buffer;
1587 1588 1589 1590 1591 1592 1593
			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
1594 1595
			if (write_buffer(local, buffer, *bufposn - stream.avail_in) < 0)
				die("unable to write sha1 file");
1596 1597 1598 1599 1600 1601 1602
			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);
1603 1604
		if (size <= 0) {
			close(local);
1605
			unlink(tmpfile);
1606 1607 1608 1609 1610
			if (!size)
				return error("Connection closed?");
			perror("Reading from connection");
			return -1;
		}
1611 1612
		*bufposn += size;
	} while (1);
1613 1614 1615 1616 1617
	inflateEnd(&stream);

	close(local);
	SHA1_Final(real_sha1, &c);
	if (ret != Z_STREAM_END) {
1618
		unlink(tmpfile);
1619 1620 1621
		return error("File %s corrupted", sha1_to_hex(sha1));
	}
	if (memcmp(sha1, real_sha1, 20)) {
1622
		unlink(tmpfile);
1623
		return error("File %s has bad hash", sha1_to_hex(sha1));
1624
	}
1625 1626

	return move_temp_to_file(tmpfile, sha1_file_name(sha1));
1627 1628
}

1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644
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;
}

1645 1646 1647 1648 1649 1650
int has_sha1_pack(const unsigned char *sha1)
{
	struct pack_entry e;
	return find_pack_entry(sha1, &e);
}

1651 1652 1653
int has_sha1_file(const unsigned char *sha1)
{
	struct stat st;
1654 1655
	struct pack_entry e;

1656
	if (find_pack_entry(sha1, &e))
1657
		return 1;
1658
	return find_sha1_file(sha1, &st) ? 1 : 0;
1659
}
J
Junio C Hamano 已提交
1660

1661 1662
/*
 * reads from fd as long as possible into a supplied buffer of size bytes.
1663
 * If necessary the buffer's size is increased using realloc()
1664 1665 1666 1667 1668 1669 1670
 *
 * 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)
1671
{
1672 1673 1674
	char* buf = *return_buf;
	unsigned long size = *return_size;
	int iret;
1675
	unsigned long off = 0;
1676

1677
	do {
1678
		iret = xread(fd, buf + off, size - off);
1679 1680 1681 1682 1683 1684 1685 1686
		if (iret > 0) {
			off += iret;
			if (off == size) {
				size *= 2;
				buf = realloc(buf, size);
			}
		}
	} while (iret > 0);
1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704

	*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;
	char *buf = malloc(size);
	int ret;
	unsigned char hdr[50];
	int hdrlen;

	if (read_pipe(fd, &buf, &size)) {
1705 1706 1707
		free(buf);
		return -1;
	}
1708

1709
	if (!type)
1710
		type = blob_type;
1711
	if (write_object)
1712
		ret = write_sha1_file(buf, size, type, sha1);
1713
	else {
1714
		write_sha1_file_prepare(buf, size, type, sha1, hdr, &hdrlen);
1715 1716 1717 1718 1719 1720
		ret = 0;
	}
	free(buf);
	return ret;
}

1721
int index_fd(unsigned char *sha1, int fd, struct stat *st, int write_object, const char *type)
J
Junio C Hamano 已提交
1722 1723
{
	unsigned long size = st->st_size;
1724 1725
	void *buf;
	int ret;
1726 1727
	unsigned char hdr[50];
	int hdrlen;
J
Junio C Hamano 已提交
1728

1729
	buf = "";
J
Junio C Hamano 已提交
1730
	if (size)
1731
		buf = mmap(NULL, size, PROT_READ, MAP_PRIVATE, fd, 0);
J
Junio C Hamano 已提交
1732
	close(fd);
P
Pavel Roskin 已提交
1733
	if (buf == MAP_FAILED)
J
Junio C Hamano 已提交
1734 1735
		return -1;

1736
	if (!type)
1737
		type = blob_type;
1738 1739 1740 1741 1742 1743
	if (write_object)
		ret = write_sha1_file(buf, size, type, sha1);
	else {
		write_sha1_file_prepare(buf, size, type, sha1, hdr, &hdrlen);
		ret = 0;
	}
1744 1745 1746
	if (size)
		munmap(buf, size);
	return ret;
J
Junio C Hamano 已提交
1747
}
1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774

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;
1775
			write_sha1_file_prepare(target, st->st_size, blob_type,
1776
						sha1, hdr, &hdrlen);
1777
		} else if (write_sha1_file(target, st->st_size, blob_type, sha1))
1778 1779 1780 1781 1782 1783 1784 1785 1786
			return error("%s: failed to insert into database",
				     path);
		free(target);
		break;
	default:
		return error("%s: unsupported file type", path);
	}
	return 0;
}