sha1_file.c 44.1 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 53 54 55 56 57 58 59 60 61 62
signed char hexval_table[256] = {
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 00-07 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 08-0f */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 10-17 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 18-1f */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 20-27 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 28-2f */
	  0,  1,  2,  3,  4,  5,  6,  7,		/* 30-37 */
	  8,  9, -1, -1, -1, -1, -1, -1,		/* 38-3f */
	 -1, 10, 11, 12, 13, 14, 15, -1,		/* 40-47 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 48-4f */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 50-57 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 58-5f */
	 -1, 10, 11, 12, 13, 14, 15, -1,		/* 60-67 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 68-67 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 70-77 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 78-7f */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 80-87 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 88-8f */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 90-97 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* 98-9f */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* a0-a7 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* a8-af */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* b0-b7 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* b8-bf */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* c0-c7 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* c8-cf */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* d0-d7 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* d8-df */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* e0-e7 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* e8-ef */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* f0-f7 */
	 -1, -1, -1, -1, -1, -1, -1, -1,		/* f8-ff */
};
63 64 65 66 67 68 69 70 71 72 73 74 75 76

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;
}

77 78 79
int safe_create_leading_directories(char *path)
{
	char *pos = path;
80 81
	struct stat st;

82 83
	if (*pos == '/')
		pos++;
84 85 86 87 88 89

	while (pos) {
		pos = strchr(pos, '/');
		if (!pos)
			break;
		*pos = 0;
90 91 92
		if (!stat(path, &st)) {
			/* path exists */
			if (!S_ISDIR(st.st_mode)) {
93
				*pos = '/';
94
				return -3;
95
			}
96
		}
97 98 99 100
		else if (mkdir(path, 0777)) {
			*pos = '/';
			return -1;
		}
101 102 103 104
		else if (adjust_shared_perm(path)) {
			*pos = '/';
			return -2;
		}
105 106 107 108
		*pos++ = '/';
	}
	return 0;
}
109

110 111
char * sha1_to_hex(const unsigned char *sha1)
{
L
Linus Torvalds 已提交
112 113
	static int bufno;
	static char hexbuffer[4][50];
114
	static const char hex[] = "0123456789abcdef";
L
Linus Torvalds 已提交
115
	char *buffer = hexbuffer[3 & ++bufno], *buf = buffer;
116 117 118 119 120 121 122
	int i;

	for (i = 0; i < 20; i++) {
		unsigned int val = *sha1++;
		*buf++ = hex[val >> 4];
		*buf++ = hex[val & 0xf];
	}
123 124
	*buf = '\0';

125 126 127
	return buffer;
}

128 129 130 131 132 133 134 135 136 137 138 139
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];
	}
}

140 141
/*
 * NOTE! This returns a statically allocated buffer, so you have to be
142
 * careful about using it. Do a "xstrdup()" if you need to save the
143
 * filename.
144 145 146
 *
 * 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
147
 * DB_ENVIRONMENT environment variable if it is not found in
148
 * the primary object database.
149 150 151 152 153 154
 */
char *sha1_file_name(const unsigned char *sha1)
{
	static char *name, *base;

	if (!base) {
J
Junio C Hamano 已提交
155
		const char *sha1_file_directory = get_object_directory();
156
		int len = strlen(sha1_file_directory);
157
		base = xmalloc(len + 60);
158 159 160 161 162 163
		memcpy(base, sha1_file_directory, len);
		memset(base+len, 0, 60);
		base[len] = '/';
		base[len+3] = '/';
		name = base + len + 1;
	}
164
	fill_sha1_path(name, sha1);
165 166 167
	return base;
}

168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217
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;
}

218 219
struct alternate_object_database *alt_odb_list;
static struct alternate_object_database **alt_odb_tail;
220

M
Martin Waitz 已提交
221 222
static void read_info_alternates(const char * alternates, int depth);

J
Junio C Hamano 已提交
223 224
/*
 * Prepare alternate object database registry.
225 226 227 228 229
 *
 * 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,
230 231
 * whose contents is similar to that environment variable but can be
 * LF separated.  Its base points at a statically allocated buffer that
232 233 234 235 236
 * 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
237
 */
M
Martin Waitz 已提交
238
static int link_alt_odb_entry(const char * entry, int len, const char * relative_base, int depth)
239
{
M
Martin Waitz 已提交
240
	struct stat st;
241
	const char *objdir = get_object_directory();
M
Martin Waitz 已提交
242 243 244 245 246
	struct alternate_object_database *ent;
	struct alternate_object_database *alt;
	/* 43 = 40-byte + 2 '/' + terminating NUL */
	int pfxlen = len;
	int entlen = pfxlen + 43;
247
	int base_len = -1;
248

M
Martin Waitz 已提交
249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316
	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;
	}

317
	last = alt;
318 319 320 321 322 323 324 325
	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
326 327
		while (cp < ep && *cp != sep)
			cp++;
328
		if (last != cp) {
M
Martin Waitz 已提交
329 330 331 332 333 334
			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);
335
			}
336
		}
337
		while (cp < ep && *cp == sep)
338 339
			cp++;
		last = cp;
340
	}
341 342
}

M
Martin Waitz 已提交
343
static void read_info_alternates(const char * relative_base, int depth)
344
{
345
	char *map;
346
	struct stat st;
M
Martin Waitz 已提交
347 348
	char path[PATH_MAX];
	int fd;
349

M
Martin Waitz 已提交
350
	sprintf(path, "%s/info/alternates", relative_base);
351 352 353 354 355
	fd = open(path, O_RDONLY);
	if (fd < 0)
		return;
	if (fstat(fd, &st) || (st.st_size == 0)) {
		close(fd);
356
		return;
357
	}
358 359 360 361 362
	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
363 364
	link_alt_odb_entries(map, map + st.st_size, '\n', relative_base, depth);

365
	munmap(map, st.st_size);
366 367
}

M
Martin Waitz 已提交
368 369
void prepare_alt_odb(void)
{
T
Timo Hirvonen 已提交
370
	const char *alt;
M
Martin Waitz 已提交
371 372 373 374 375 376 377 378 379 380 381 382

	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);
}

383 384 385
static char *find_sha1_file(const unsigned char *sha1, struct stat *st)
{
	char *name = sha1_file_name(sha1);
386
	struct alternate_object_database *alt;
387 388 389

	if (!stat(name, st))
		return name;
390
	prepare_alt_odb();
391 392
	for (alt = alt_odb_list; alt; alt = alt->next) {
		name = alt->name;
393
		fill_sha1_path(name, sha1);
394 395
		if (!stat(alt->base, st))
			return alt->base;
396 397 398 399
	}
	return NULL;
}

400 401
static int pack_used_ctr;
static unsigned long pack_mapped;
402
struct packed_git *packed_git;
403 404 405 406 407 408 409 410

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;
411
	int fd = open(path, O_RDONLY);
412 413 414 415 416 417 418 419 420 421 422 423 424 425
	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;
426 427
	*idx_map_ = idx_map;
	*idx_size_ = idx_size;
428 429

	/* check index map */
430
	if (idx_size < 4*256 + 20 + 20)
431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452
		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;
}

453
static int unuse_one_packed_git(void)
454
{
455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472
	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--;
473 474
}

475
int use_packed_git(struct packed_git *p)
476
{
477 478
	if (!p->pack_size) {
		struct stat st;
479
		/* We created the struct before we had the pack */
480 481 482 483 484
		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;
	}
485 486 487 488
	if (!p->pack_base) {
		int fd;
		struct stat st;
		void *map;
489
		struct pack_header *hdr;
490 491

		pack_mapped += p->pack_size;
492
		while (packed_git_limit < pack_mapped && unuse_one_packed_git())
493
			; /* nothing */
494 495
		fd = open(p->pack_name, O_RDONLY);
		if (fd < 0)
496
			die("packfile %s cannot be opened", p->pack_name);
497 498
		if (fstat(fd, &st)) {
			close(fd);
499
			die("packfile %s cannot be opened", p->pack_name);
500 501
		}
		if (st.st_size != p->pack_size)
502
			die("packfile %s size mismatch.", p->pack_name);
503 504 505
		map = mmap(NULL, p->pack_size, PROT_READ, MAP_PRIVATE, fd, 0);
		close(fd);
		if (map == MAP_FAILED)
506
			die("packfile %s cannot be mapped.", p->pack_name);
507
		p->pack_base = map;
508

509 510 511 512 513 514 515 516 517 518 519
		/* 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));

520 521 522
		/* Check if the pack file matches with the index file.
		 * this is cheap.
		 */
523 524 525 526
		if (hashcmp((unsigned char *)(p->index_base) +
			    p->index_size - 40,
			    (unsigned char *)p->pack_base +
			    p->pack_size - 20)) {
527
			die("packfile %s does not match index.", p->pack_name);
528
		}
529 530
	}
	p->pack_last_used = pack_used_ctr++;
531
	p->pack_use_cnt++;
532 533 534
	return 0;
}

535
struct packed_git *add_packed_git(char *path, int path_len, int local)
536 537 538 539 540
{
	struct stat st;
	struct packed_git *p;
	unsigned long idx_size;
	void *idx_map;
541
	unsigned char sha1[20];
542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560

	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;
561
	p->pack_base = NULL;
562
	p->pack_last_used = 0;
563
	p->pack_use_cnt = 0;
564
	p->pack_local = local;
P
Pavel Roskin 已提交
565
	if ((path_len > 44) && !get_sha1_hex(path + path_len - 44, sha1))
566
		hashcpy(p->sha1, sha1);
567 568 569
	return p;
}

570
struct packed_git *parse_pack_index(unsigned char *sha1)
571 572 573 574 575
{
	char *path = sha1_pack_index_name(sha1);
	return parse_pack_index_file(sha1, path);
}

576
struct packed_git *parse_pack_index_file(const unsigned char *sha1, char *idx_path)
577 578 579 580
{
	struct packed_git *p;
	unsigned long idx_size;
	void *idx_map;
581
	char *path;
582

583
	if (check_packed_git_idx(idx_path, &idx_size, &idx_map))
584 585 586 587 588 589 590 591 592 593 594 595 596
		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;
597
	hashcpy(p->sha1, sha1);
598 599 600 601 602 603 604 605 606
	return p;
}

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

607
static void prepare_packed_git_one(char *objdir, int local)
608 609 610 611 612 613 614 615 616
{
	char path[PATH_MAX];
	int len;
	DIR *dir;
	struct dirent *de;

	sprintf(path, "%s/pack", objdir);
	len = strlen(path);
	dir = opendir(path);
617
	if (!dir) {
J
Junio C Hamano 已提交
618
		if (errno != ENOENT)
619
			error("unable to open object pack directory: %s: %s",
J
Junio C Hamano 已提交
620
			      path, strerror(errno));
621
		return;
622
	}
623 624 625 626 627
	path[len++] = '/';
	while ((de = readdir(dir)) != NULL) {
		int namelen = strlen(de->d_name);
		struct packed_git *p;

628
		if (!has_extension(de->d_name, ".idx"))
629 630 631 632
			continue;

		/* we have .idx.  Is it a file we can map? */
		strcpy(path + len, de->d_name);
633 634 635 636 637 638
		for (p = packed_git; p; p = p->next) {
			if (!memcmp(path, p->pack_name, len + namelen - 4))
				break;
		}
		if (p)
			continue;
639
		p = add_packed_git(path, len + namelen, local);
640 641 642 643 644
		if (!p)
			continue;
		p->next = packed_git;
		packed_git = p;
	}
645
	closedir(dir);
646 647
}

648
static int prepare_packed_git_run_once = 0;
649
void prepare_packed_git(void)
650
{
651
	struct alternate_object_database *alt;
652

653
	if (prepare_packed_git_run_once)
654
		return;
655
	prepare_packed_git_one(get_object_directory(), 1);
656
	prepare_alt_odb();
657
	for (alt = alt_odb_list; alt; alt = alt->next) {
658
		alt->name[-1] = 0;
659
		prepare_packed_git_one(alt->base, 0);
660
		alt->name[-1] = '/';
661
	}
662 663 664
	prepare_packed_git_run_once = 1;
}

665
void reprepare_packed_git(void)
666 667 668
{
	prepare_packed_git_run_once = 0;
	prepare_packed_git();
669 670
}

671
int check_sha1_signature(const unsigned char *sha1, void *map, unsigned long size, const char *type)
672 673
{
	unsigned char real_sha1[20];
674
	hash_sha1_file(map, size, type, real_sha1);
675
	return hashcmp(sha1, real_sha1) ? -1 : 0;
676 677
}

678
void *map_sha1_file(const unsigned char *sha1, unsigned long *size)
679 680 681
{
	struct stat st;
	void *map;
682
	int fd;
683 684 685 686 687
	char *filename = find_sha1_file(sha1, &st);

	if (!filename) {
		return NULL;
	}
688

689
	fd = open(filename, O_RDONLY | sha1_file_open_flag);
690
	if (fd < 0) {
691 692 693 694 695 696 697 698 699 700 701
		/* 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;
		}

702 703 704
		/* If it failed once, it will probably fail again.
		 * Stop using O_NOATIME
		 */
705
		sha1_file_open_flag = 0;
706 707 708
	}
	map = mmap(NULL, st.st_size, PROT_READ, MAP_PRIVATE, fd, 0);
	close(fd);
P
Pavel Roskin 已提交
709
	if (map == MAP_FAILED)
710 711 712 713 714
		return NULL;
	*size = st.st_size;
	return map;
}

715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730
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;
}

731
unsigned long unpack_object_header_gently(const unsigned char *buf, unsigned long len, enum object_type *type, unsigned long *sizep)
732
{
733
	unsigned shift;
734 735
	unsigned char c;
	unsigned long size;
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
	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 */
762
	};
763
	enum object_type type;
764

765 766 767 768 769
	/* Get the data stream */
	memset(stream, 0, sizeof(*stream));
	stream->next_in = map;
	stream->avail_in = mapsize;
	stream->next_out = buffer;
770 771
	stream->avail_out = bufsiz;

772
	if (legacy_loose_object(map)) {
773 774 775 776
		inflateInit(stream);
		return inflate(stream, 0);
	}

777 778
	used = unpack_object_header_gently(map, mapsize, &type, &size);
	if (!used || !valid_loose_object_type[type])
779
		return -1;
780 781
	map += used;
	mapsize -= used;
782 783 784 785

	/* Set up the stream for the rest.. */
	stream->next_in = map;
	stream->avail_in = mapsize;
786
	inflateInit(stream);
787 788

	/* And generate the fake traditional header */
789 790
	stream->total_out = 1 + snprintf(buffer, bufsiz, "%s %lu",
					 type_names[type], size);
791
	return 0;
792 793
}

L
Linus Torvalds 已提交
794
static void *unpack_sha1_rest(z_stream *stream, void *buffer, unsigned long size)
795 796
{
	int bytes = strlen(buffer) + 1;
797
	unsigned char *buf = xmalloc(1+size);
798
	unsigned long n;
799

800 801 802 803 804
	n = stream->total_out - bytes;
	if (n > size)
		n = size;
	memcpy(buf, (char *) buffer + bytes, n);
	bytes = n;
805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820
	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.
 */
821
static int parse_sha1_header(char *hdr, char *type, unsigned long *sizep)
822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865
{
	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;
}

866 867
void * unpack_sha1_file(void *map, unsigned long mapsize, char *type, unsigned long *size)
{
868
	int ret;
869
	z_stream stream;
870
	char hdr[8192];
871

872 873
	ret = unpack_sha1_header(&stream, map, mapsize, hdr, sizeof(hdr));
	if (ret < Z_OK || parse_sha1_header(hdr, type, size) < 0)
874 875
		return NULL;

876
	return unpack_sha1_rest(&stream, hdr, *size);
877 878
}

879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919
static unsigned long get_delta_base(struct packed_git *p,
				    unsigned long offset,
				    enum object_type kind,
				    unsigned long delta_obj_offset,
				    unsigned long *base_obj_offset)
{
	unsigned char *base_info = (unsigned char *) p->pack_base + offset;
	unsigned long base_offset;

	/* there must be at least 20 bytes left regardless of delta type */
	if (p->pack_size <= offset + 20)
		die("truncated pack file");

	if (kind == OBJ_OFS_DELTA) {
		unsigned used = 0;
		unsigned char c = base_info[used++];
		base_offset = c & 127;
		while (c & 128) {
			base_offset += 1;
			if (!base_offset || base_offset & ~(~0UL >> 7))
				die("offset value overflow for delta base object");
			c = base_info[used++];
			base_offset = (base_offset << 7) + (c & 127);
		}
		base_offset = delta_obj_offset - base_offset;
		if (base_offset >= delta_obj_offset)
			die("delta base offset out of bound");
		offset += used;
	} else if (kind == OBJ_REF_DELTA) {
		/* The base entry _must_ be in the same pack */
		base_offset = find_pack_entry_one(base_info, p);
		if (!base_offset)
			die("failed to find delta-pack base object %s",
				sha1_to_hex(base_info));
		offset += 20;
	} else
		die("I am totally screwed");
	*base_obj_offset = base_offset;
	return offset;
}

J
Junio C Hamano 已提交
920
/* forward declaration for a mutually recursive function */
N
Nicolas Pitre 已提交
921
static int packed_object_info(struct packed_git *p, unsigned long offset,
J
Junio C Hamano 已提交
922 923
			      char *type, unsigned long *sizep);

N
Nicolas Pitre 已提交
924 925
static int packed_delta_info(struct packed_git *p,
			     unsigned long offset,
926 927
			     enum object_type kind,
			     unsigned long obj_offset,
928
			     char *type,
N
Nicolas Pitre 已提交
929
			     unsigned long *sizep)
930
{
N
Nicolas Pitre 已提交
931
	unsigned long base_offset;
J
Junio C Hamano 已提交
932

933
	offset = get_delta_base(p, offset, kind, obj_offset, &base_offset);
J
Junio C Hamano 已提交
934

935 936 937 938 939
	/* 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.
	 */
N
Nicolas Pitre 已提交
940
	if (packed_object_info(p, base_offset, type, NULL))
941
		die("cannot get info for delta-pack base");
942

943 944
	if (sizep) {
		const unsigned char *data;
N
Nicolas Pitre 已提交
945
		unsigned char delta_head[20];
946 947 948 949 950 951
		unsigned long result_size;
		z_stream stream;
		int st;

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

N
Nicolas Pitre 已提交
952 953
		stream.next_in = (unsigned char *) p->pack_base + offset;
		stream.avail_in = p->pack_size - offset;
954 955 956 957 958 959 960 961 962 963 964 965 966 967
		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;
968 969 970

		/* ignore base size */
		get_delta_hdr_size(&data, delta_head+sizeof(delta_head));
971

972
		/* Read the result size */
973
		result_size = get_delta_hdr_size(&data, delta_head+sizeof(delta_head));
974 975
		*sizep = result_size;
	}
976 977 978
	return 0;
}

979 980 981
static unsigned long unpack_object_header(struct packed_git *p, unsigned long offset,
	enum object_type *type, unsigned long *sizep)
{
982
	unsigned long used;
983

984
	if (p->pack_size <= offset)
985
		die("object offset outside of pack file");
986 987 988 989 990 991 992 993

	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;
994 995
}

N
Nicolas Pitre 已提交
996 997
void packed_object_info_detail(struct packed_git *p,
			       unsigned long offset,
998 999 1000
			       char *type,
			       unsigned long *size,
			       unsigned long *store_size,
J
Junio C Hamano 已提交
1001
			       unsigned int *delta_chain_length,
1002 1003
			       unsigned char *base_sha1)
{
1004
	unsigned long obj_offset, val;
N
Nicolas Pitre 已提交
1005
	unsigned char *next_sha1;
1006 1007
	enum object_type kind;

N
Nicolas Pitre 已提交
1008
	*delta_chain_length = 0;
1009
	obj_offset = offset;
N
Nicolas Pitre 已提交
1010 1011 1012 1013 1014
	offset = unpack_object_header(p, offset, &kind, size);

	for (;;) {
		switch (kind) {
		default:
1015
			die("pack %s contains unknown object type %d",
N
Nicolas Pitre 已提交
1016 1017 1018 1019 1020 1021 1022 1023
			    p->pack_name, kind);
		case OBJ_COMMIT:
		case OBJ_TREE:
		case OBJ_BLOB:
		case OBJ_TAG:
			strcpy(type, type_names[kind]);
			*store_size = 0; /* notyet */
			return;
1024 1025 1026 1027 1028 1029 1030
		case OBJ_OFS_DELTA:
			get_delta_base(p, offset, kind, obj_offset, &offset);
			if (*delta_chain_length == 0) {
				/* TODO: find base_sha1 as pointed by offset */
			}
			break;
		case OBJ_REF_DELTA:
N
Nicolas Pitre 已提交
1031 1032 1033 1034 1035 1036 1037 1038 1039
			if (p->pack_size <= offset + 20)
				die("pack file %s records an incomplete delta base",
				    p->pack_name);
			next_sha1 = (unsigned char *) p->pack_base + offset;
			if (*delta_chain_length == 0)
				hashcpy(base_sha1, next_sha1);
			offset = find_pack_entry_one(next_sha1, p);
			break;
		}
1040
		obj_offset = offset;
N
Nicolas Pitre 已提交
1041 1042
		offset = unpack_object_header(p, offset, &kind, &val);
		(*delta_chain_length)++;
1043 1044 1045
	}
}

N
Nicolas Pitre 已提交
1046
static int packed_object_info(struct packed_git *p, unsigned long offset,
1047 1048
			      char *type, unsigned long *sizep)
{
1049
	unsigned long size, obj_offset = offset;
1050
	enum object_type kind;
1051

N
Nicolas Pitre 已提交
1052
	offset = unpack_object_header(p, offset, &kind, &size);
1053

1054
	switch (kind) {
1055 1056 1057
	case OBJ_OFS_DELTA:
	case OBJ_REF_DELTA:
		return packed_delta_info(p, offset, kind, obj_offset, type, sizep);
1058 1059 1060 1061
	case OBJ_COMMIT:
	case OBJ_TREE:
	case OBJ_BLOB:
	case OBJ_TAG:
1062
		strcpy(type, type_names[kind]);
1063
		break;
1064
	default:
1065
		die("pack %s contains unknown object type %d",
1066
		    p->pack_name, kind);
1067
	}
1068 1069
	if (sizep)
		*sizep = size;
1070 1071 1072
	return 0;
}

1073 1074 1075
static void *unpack_compressed_entry(struct packed_git *p,
				    unsigned long offset,
				    unsigned long size)
1076 1077 1078 1079 1080 1081 1082 1083
{
	int st;
	z_stream stream;
	unsigned char *buffer;

	buffer = xmalloc(size + 1);
	buffer[size] = 0;
	memset(&stream, 0, sizeof(stream));
1084 1085
	stream.next_in = (unsigned char*)p->pack_base + offset;
	stream.avail_in = p->pack_size - offset;
1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099
	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;
}

1100 1101
static void *unpack_delta_entry(struct packed_git *p,
				unsigned long offset,
1102
				unsigned long delta_size,
1103 1104
				enum object_type kind,
				unsigned long obj_offset,
1105
				char *type,
1106
				unsigned long *sizep)
1107
{
1108
	void *delta_data, *result, *base;
N
Nicolas Pitre 已提交
1109 1110
	unsigned long result_size, base_size, base_offset;

1111
	offset = get_delta_base(p, offset, kind, obj_offset, &base_offset);
1112
	base = unpack_entry(p, base_offset, type, &base_size);
1113
	if (!base)
N
Nicolas Pitre 已提交
1114 1115
		die("failed to read delta base object at %lu from %s",
		    base_offset, p->pack_name);
1116

N
Nicolas Pitre 已提交
1117
	delta_data = unpack_compressed_entry(p, offset, delta_size);
1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128
	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;
}

1129
void *unpack_entry(struct packed_git *p, unsigned long offset,
1130 1131
			  char *type, unsigned long *sizep)
{
1132 1133
	unsigned long size, obj_offset = offset;
	enum object_type kind;
1134
	void *retval;
1135 1136 1137

	if (use_packed_git(p))
		die("cannot map packed file");
N
Nicolas Pitre 已提交
1138
	offset = unpack_object_header(p, offset, &kind, &size);
1139
	switch (kind) {
1140 1141
	case OBJ_OFS_DELTA:
	case OBJ_REF_DELTA:
1142 1143
		retval = unpack_delta_entry(p, offset, size, kind, obj_offset, type, sizep);
		break;
1144 1145 1146 1147
	case OBJ_COMMIT:
	case OBJ_TREE:
	case OBJ_BLOB:
	case OBJ_TAG:
1148 1149
		strcpy(type, type_names[kind]);
		*sizep = size;
1150 1151
		retval = unpack_compressed_entry(p, offset, size);
		break;
1152
	default:
1153
		die("unknown object type %i in %s", kind, p->pack_name);
1154
	}
1155 1156
	unuse_packed_git(p);
	return retval;
1157 1158
}

1159 1160
int num_packed_objects(const struct packed_git *p)
{
1161
	/* See check_packed_git_idx() */
1162 1163 1164 1165 1166 1167 1168 1169 1170
	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;
1171
	hashcpy(sha1, (unsigned char *) index + (24 * n) + 4);
1172 1173 1174
	return 0;
}

N
Nicolas Pitre 已提交
1175 1176
unsigned long find_pack_entry_one(const unsigned char *sha1,
				  struct packed_git *p)
1177
{
1178
	unsigned int *level1_ofs = p->index_base;
1179 1180 1181 1182 1183 1184
	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;
1185
		int cmp = hashcmp((unsigned char *)index + (24 * mi) + 4, sha1);
N
Nicolas Pitre 已提交
1186 1187
		if (!cmp)
			return ntohl(*((unsigned int *) ((char *) index + (24 * mi))));
1188 1189 1190 1191 1192 1193 1194 1195
		if (cmp > 0)
			hi = mi;
		else
			lo = mi+1;
	} while (lo < hi);
	return 0;
}

1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213
static int matches_pack_name(struct packed_git *p, const char *ig)
{
	const char *last_c, *c;

	if (!strcmp(p->pack_name, ig))
		return 0;

	for (c = p->pack_name, last_c = c; *c;)
		if (*c == '/')
			last_c = ++c;
		else
			++c;
	if (!strcmp(last_c, ig))
		return 0;

	return 1;
}

1214
static int find_pack_entry(const unsigned char *sha1, struct pack_entry *e, const char **ignore_packed)
1215 1216
{
	struct packed_git *p;
N
Nicolas Pitre 已提交
1217 1218
	unsigned long offset;

1219 1220 1221
	prepare_packed_git();

	for (p = packed_git; p; p = p->next) {
1222 1223 1224
		if (ignore_packed) {
			const char **ig;
			for (ig = ignore_packed; *ig; ig++)
1225
				if (!matches_pack_name(p, *ig))
1226 1227 1228 1229
					break;
			if (*ig)
				continue;
		}
N
Nicolas Pitre 已提交
1230 1231 1232 1233 1234
		offset = find_pack_entry_one(sha1, p);
		if (offset) {
			e->offset = offset;
			e->p = p;
			hashcpy(e->sha1, sha1);
1235
			return 1;
N
Nicolas Pitre 已提交
1236
		}
1237 1238 1239 1240
	}
	return 0;
}

1241 1242 1243 1244 1245 1246
struct packed_git *find_sha1_pack(const unsigned char *sha1, 
				  struct packed_git *packs)
{
	struct packed_git *p;

	for (p = packs; p; p = p->next) {
N
Nicolas Pitre 已提交
1247
		if (find_pack_entry_one(sha1, p))
1248 1249 1250 1251 1252 1253
			return p;
	}
	return NULL;
	
}

1254
static int sha1_loose_object_info(const unsigned char *sha1, char *type, unsigned long *sizep)
1255
{
1256
	int status;
1257 1258 1259
	unsigned long mapsize, size;
	void *map;
	z_stream stream;
1260
	char hdr[128];
1261

1262
	map = map_sha1_file(sha1, &mapsize);
1263 1264
	if (!map)
		return error("unable to find %s", sha1_to_hex(sha1));
1265 1266 1267 1268 1269
	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));
1270
	else {
1271
		status = 0;
1272 1273
		if (sizep)
			*sizep = size;
1274 1275 1276 1277 1278 1279
	}
	inflateEnd(&stream);
	munmap(map, mapsize);
	return status;
}

1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296
int sha1_object_info(const unsigned char *sha1, char *type, unsigned long *sizep)
{
	int status;
	struct pack_entry e;

	if (!find_pack_entry(sha1, &e, NULL)) {
		reprepare_packed_git();
		if (!find_pack_entry(sha1, &e, NULL))
			return sha1_loose_object_info(sha1, type, sizep);
	}
	if (use_packed_git(e.p))
		die("cannot map packed file");
	status = packed_object_info(e.p, e.offset, type, sizep);
	unuse_packed_git(e.p);
	return status;
}

1297 1298 1299 1300
static void *read_packed_sha1(const unsigned char *sha1, char *type, unsigned long *size)
{
	struct pack_entry e;

1301
	if (!find_pack_entry(sha1, &e, NULL)) {
1302 1303 1304
		error("cannot read sha1_file for %s", sha1_to_hex(sha1));
		return NULL;
	}
1305
	return unpack_entry(e.p, e.offset, type, size);
1306 1307
}

1308 1309 1310 1311
void * read_sha1_file(const unsigned char *sha1, char *type, unsigned long *size)
{
	unsigned long mapsize;
	void *map, *buf;
1312
	struct pack_entry e;
1313

1314
	if (find_pack_entry(sha1, &e, NULL))
1315
		return read_packed_sha1(sha1, type, size);
1316
	map = map_sha1_file(sha1, &mapsize);
1317 1318 1319 1320 1321
	if (map) {
		buf = unpack_sha1_file(map, mapsize, type, size);
		munmap(map, mapsize);
		return buf;
	}
1322
	reprepare_packed_git();
1323
	if (find_pack_entry(sha1, &e, NULL))
1324
		return read_packed_sha1(sha1, type, size);
1325
	return NULL;
1326 1327
}

1328
void *read_object_with_reference(const unsigned char *sha1,
1329
				 const char *required_type,
1330 1331
				 unsigned long *size,
				 unsigned char *actual_sha1_return)
1332 1333 1334 1335
{
	char type[20];
	void *buffer;
	unsigned long isize;
1336
	unsigned char actual_sha1[20];
1337

1338
	hashcpy(actual_sha1, sha1);
1339 1340 1341
	while (1) {
		int ref_length = -1;
		const char *ref_type = NULL;
1342

1343 1344 1345 1346 1347 1348
		buffer = read_sha1_file(actual_sha1, type, &isize);
		if (!buffer)
			return NULL;
		if (!strcmp(type, required_type)) {
			*size = isize;
			if (actual_sha1_return)
1349
				hashcpy(actual_sha1_return, actual_sha1);
1350 1351 1352
			return buffer;
		}
		/* Handle references */
1353
		else if (!strcmp(type, commit_type))
1354
			ref_type = "tree ";
1355
		else if (!strcmp(type, tag_type))
1356 1357 1358 1359 1360 1361
			ref_type = "object ";
		else {
			free(buffer);
			return NULL;
		}
		ref_length = strlen(ref_type);
1362

1363
		if (memcmp(buffer, ref_type, ref_length) ||
1364
		    get_sha1_hex((char *) buffer + ref_length, actual_sha1)) {
1365 1366 1367
			free(buffer);
			return NULL;
		}
1368
		free(buffer);
1369 1370
		/* Now we have the ID of the referred-to object in
		 * actual_sha1.  Check again. */
1371 1372 1373
	}
}

R
Rene Scharfe 已提交
1374 1375 1376
static void write_sha1_file_prepare(void *buf, unsigned long len,
                                    const char *type, unsigned char *sha1,
                                    unsigned char *hdr, int *hdrlen)
1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389
{
	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);
}

1390 1391 1392 1393 1394 1395
/*
 * 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.
 */
1396
static int link_temp_to_file(const char *tmpfile, const char *filename)
1397 1398
{
	int ret;
S
Shawn Pearce 已提交
1399
	char *dir;
1400 1401 1402 1403 1404

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

	/*
S
Shawn Pearce 已提交
1405
	 * Try to mkdir the last path component if that failed.
1406 1407 1408 1409 1410 1411
	 *
	 * 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
1412 1413 1414
	dir = strrchr(filename, '/');
	if (dir) {
		*dir = 0;
1415
		if (!mkdir(filename, 0777) && adjust_shared_perm(filename)) {
1416
			*dir = '/';
S
Shawn Pearce 已提交
1417
			return -2;
1418
		}
S
Shawn Pearce 已提交
1419 1420 1421 1422
		*dir = '/';
		if (!link(tmpfile, filename))
			return 0;
		ret = errno;
1423 1424 1425 1426 1427 1428 1429
	}
	return ret;
}

/*
 * Move the just written object into its final resting place
 */
1430
int move_temp_to_file(const char *tmpfile, const char *filename)
1431 1432
{
	int ret = link_temp_to_file(tmpfile, filename);
1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446

	/*
	 * 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))
1447
			return 0;
1448
		ret = errno;
1449 1450 1451 1452
	}
	unlink(tmpfile);
	if (ret) {
		if (ret != EEXIST) {
1453
			return error("unable to write sha1 filename %s: %s\n", filename, strerror(ret));
1454 1455 1456 1457 1458 1459 1460
		}
		/* FIXME!!! Collision check here ? */
	}

	return 0;
}

L
Linus Torvalds 已提交
1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474
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;
1475
		buf = (char *) buf + size;
L
Linus Torvalds 已提交
1476 1477 1478 1479
	}
	return 0;
}

1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522
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;
}

R
Rene Scharfe 已提交
1523 1524 1525 1526 1527 1528 1529 1530 1531
int hash_sha1_file(void *buf, unsigned long len, const char *type,
                   unsigned char *sha1)
{
	unsigned char hdr[50];
	int hdrlen;
	write_sha1_file_prepare(buf, len, type, sha1, hdr, &hdrlen);
	return 0;
}

1532
int write_sha1_file(void *buf, unsigned long len, const char *type, unsigned char *returnsha1)
1533 1534
{
	int size;
1535
	unsigned char *compressed;
1536 1537
	z_stream stream;
	unsigned char sha1[20];
1538
	char *filename;
1539
	static char tmpfile[PATH_MAX];
1540
	unsigned char hdr[50];
1541
	int fd, hdrlen;
1542

1543 1544 1545
	/* Normally if we have it in the pack then we do not bother writing
	 * it out into .git/objects/??/?{38} file.
	 */
R
Rene Scharfe 已提交
1546 1547
	write_sha1_file_prepare(buf, len, type, sha1, hdr, &hdrlen);
	filename = sha1_file_name(sha1);
1548
	if (returnsha1)
1549
		hashcpy(returnsha1, sha1);
1550 1551
	if (has_sha1_file(sha1))
		return 0;
1552 1553
	fd = open(filename, O_RDONLY);
	if (fd >= 0) {
1554
		/*
1555 1556
		 * FIXME!!! We might do collision checking here, but we'd
		 * need to uncompress the old file and check it. Later.
1557
		 */
1558
		close(fd);
1559 1560 1561
		return 0;
	}

1562
	if (errno != ENOENT) {
1563
		return error("sha1 file %s: %s\n", filename, strerror(errno));
1564 1565 1566
	}

	snprintf(tmpfile, sizeof(tmpfile), "%s/obj_XXXXXX", get_object_directory());
1567

1568 1569
	fd = mkstemp(tmpfile);
	if (fd < 0) {
1570 1571 1572 1573
		if (errno == EPERM)
			return error("insufficient permission for adding an object to repository database %s\n", get_object_directory());
		else
			return error("unable to create temporary sha1 filename %s: %s\n", tmpfile, strerror(errno));
1574 1575
	}

1576 1577
	/* Set it up */
	memset(&stream, 0, sizeof(stream));
1578
	deflateInit(&stream, zlib_compression_level);
1579
	size = 8 + deflateBound(&stream, len+hdrlen);
1580
	compressed = xmalloc(size);
1581 1582 1583 1584

	/* Compress it */
	stream.next_out = compressed;
	stream.avail_out = size;
1585 1586 1587 1588

	/* First header.. */
	stream.next_in = hdr;
	stream.avail_in = hdrlen;
1589
	setup_object_header(&stream, type, len);
1590 1591 1592 1593

	/* Then the data itself.. */
	stream.next_in = buf;
	stream.avail_in = len;
1594 1595 1596 1597 1598
	while (deflate(&stream, Z_FINISH) == Z_OK)
		/* nothing */;
	deflateEnd(&stream);
	size = stream.total_out;

L
Linus Torvalds 已提交
1599 1600
	if (write_buffer(fd, compressed, size) < 0)
		die("unable to write sha1 file");
1601
	fchmod(fd, 0444);
1602
	close(fd);
1603
	free(compressed);
1604

1605
	return move_temp_to_file(tmpfile, filename);
1606
}
1607

L
Linus Torvalds 已提交
1608 1609 1610 1611 1612
/*
 * 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
1613
{
L
Linus Torvalds 已提交
1614
	size_t size;
D
Daniel Barkalow 已提交
1615
	z_stream stream;
L
Linus Torvalds 已提交
1616 1617 1618 1619 1620 1621
	unsigned char *unpacked;
	unsigned long len;
	char type[20];
	char hdr[50];
	int hdrlen;
	void *buf;
1622

1623
	/* need to unpack and recompress it by itself */
L
Linus Torvalds 已提交
1624
	unpacked = read_packed_sha1(sha1, type, &len);
D
Daniel Barkalow 已提交
1625

L
Linus Torvalds 已提交
1626
	hdrlen = sprintf(hdr, "%s %lu", type, len) + 1;
D
Daniel Barkalow 已提交
1627

L
Linus Torvalds 已提交
1628 1629
	/* Set it up */
	memset(&stream, 0, sizeof(stream));
1630
	deflateInit(&stream, zlib_compression_level);
L
Linus Torvalds 已提交
1631 1632
	size = deflateBound(&stream, len + hdrlen);
	buf = xmalloc(size);
D
Daniel Barkalow 已提交
1633

L
Linus Torvalds 已提交
1634 1635 1636
	/* Compress it */
	stream.next_out = buf;
	stream.avail_out = size;
D
Daniel Barkalow 已提交
1637

L
Linus Torvalds 已提交
1638 1639 1640 1641 1642
	/* First header.. */
	stream.next_in = (void *)hdr;
	stream.avail_in = hdrlen;
	while (deflate(&stream, 0) == Z_OK)
		/* nothing */;
1643

L
Linus Torvalds 已提交
1644 1645 1646 1647 1648 1649 1650
	/* Then the data itself.. */
	stream.next_in = unpacked;
	stream.avail_in = len;
	while (deflate(&stream, Z_FINISH) == Z_OK)
		/* nothing */;
	deflateEnd(&stream);
	free(unpacked);
1651

L
Linus Torvalds 已提交
1652 1653 1654 1655 1656 1657 1658 1659
	*objsize = stream.total_out;
	return buf;
}

int write_sha1_to_fd(int fd, const unsigned char *sha1)
{
	int retval;
	unsigned long objsize;
1660
	void *buf = map_sha1_file(sha1, &objsize);
L
Linus Torvalds 已提交
1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671

	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
1672 1673
}

1674 1675
int write_sha1_from_fd(const unsigned char *sha1, int fd, char *buffer,
		       size_t bufsize, size_t *bufposn)
1676
{
1677
	char tmpfile[PATH_MAX];
1678 1679 1680
	int local;
	z_stream stream;
	unsigned char real_sha1[20];
1681
	unsigned char discard[4096];
1682 1683 1684
	int ret;
	SHA_CTX c;

1685
	snprintf(tmpfile, sizeof(tmpfile), "%s/obj_XXXXXX", get_object_directory());
1686

1687
	local = mkstemp(tmpfile);
1688 1689 1690 1691 1692 1693
	if (local < 0) {
		if (errno == EPERM)
			return error("insufficient permission for adding an object to repository database %s\n", get_object_directory());
		else
			return error("unable to create temporary sha1 filename %s: %s\n", tmpfile, strerror(errno));
	}
1694 1695 1696 1697 1698 1699 1700 1701 1702

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

	inflateInit(&stream);

	SHA1_Init(&c);

	do {
		ssize_t size;
1703 1704
		if (*bufposn) {
			stream.avail_in = *bufposn;
P
Pavel Roskin 已提交
1705
			stream.next_in = (unsigned char *) buffer;
1706 1707 1708 1709 1710 1711 1712
			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
1713 1714
			if (write_buffer(local, buffer, *bufposn - stream.avail_in) < 0)
				die("unable to write sha1 file");
1715 1716 1717 1718 1719 1720 1721
			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);
1722 1723
		if (size <= 0) {
			close(local);
1724
			unlink(tmpfile);
1725 1726 1727 1728 1729
			if (!size)
				return error("Connection closed?");
			perror("Reading from connection");
			return -1;
		}
1730 1731
		*bufposn += size;
	} while (1);
1732 1733 1734 1735 1736
	inflateEnd(&stream);

	close(local);
	SHA1_Final(real_sha1, &c);
	if (ret != Z_STREAM_END) {
1737
		unlink(tmpfile);
1738 1739
		return error("File %s corrupted", sha1_to_hex(sha1));
	}
1740
	if (hashcmp(sha1, real_sha1)) {
1741
		unlink(tmpfile);
1742
		return error("File %s has bad hash", sha1_to_hex(sha1));
1743
	}
1744 1745

	return move_temp_to_file(tmpfile, sha1_file_name(sha1));
1746 1747
}

1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763
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;
}

1764
int has_sha1_pack(const unsigned char *sha1, const char **ignore_packed)
1765 1766
{
	struct pack_entry e;
1767
	return find_pack_entry(sha1, &e, ignore_packed);
1768 1769
}

1770 1771 1772
int has_sha1_file(const unsigned char *sha1)
{
	struct stat st;
1773 1774
	struct pack_entry e;

1775
	if (find_pack_entry(sha1, &e, NULL))
1776
		return 1;
1777
	return find_sha1_file(sha1, &st) ? 1 : 0;
1778
}
J
Junio C Hamano 已提交
1779

1780 1781
/*
 * reads from fd as long as possible into a supplied buffer of size bytes.
1782
 * If necessary the buffer's size is increased using realloc()
1783 1784 1785 1786 1787 1788 1789
 *
 * 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)
1790
{
1791 1792 1793
	char* buf = *return_buf;
	unsigned long size = *return_size;
	int iret;
1794
	unsigned long off = 0;
1795

1796
	do {
1797
		iret = xread(fd, buf + off, size - off);
1798 1799 1800 1801
		if (iret > 0) {
			off += iret;
			if (off == size) {
				size *= 2;
J
Jonas Fonseca 已提交
1802
				buf = xrealloc(buf, size);
1803 1804 1805
			}
		}
	} while (iret > 0);
1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817

	*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
1818
	char *buf = xmalloc(size);
1819 1820 1821
	int ret;

	if (read_pipe(fd, &buf, &size)) {
1822 1823 1824
		free(buf);
		return -1;
	}
1825

1826
	if (!type)
1827
		type = blob_type;
1828
	if (write_object)
1829
		ret = write_sha1_file(buf, size, type, sha1);
R
Rene Scharfe 已提交
1830 1831
	else
		ret = hash_sha1_file(buf, size, type, sha1);
1832 1833 1834 1835
	free(buf);
	return ret;
}

1836
int index_fd(unsigned char *sha1, int fd, struct stat *st, int write_object, const char *type)
J
Junio C Hamano 已提交
1837 1838
{
	unsigned long size = st->st_size;
1839 1840
	void *buf;
	int ret;
J
Junio C Hamano 已提交
1841

1842
	buf = "";
J
Junio C Hamano 已提交
1843
	if (size)
1844
		buf = mmap(NULL, size, PROT_READ, MAP_PRIVATE, fd, 0);
J
Junio C Hamano 已提交
1845
	close(fd);
P
Pavel Roskin 已提交
1846
	if (buf == MAP_FAILED)
J
Junio C Hamano 已提交
1847 1848
		return -1;

1849
	if (!type)
1850
		type = blob_type;
1851 1852
	if (write_object)
		ret = write_sha1_file(buf, size, type, sha1);
R
Rene Scharfe 已提交
1853 1854
	else
		ret = hash_sha1_file(buf, size, type, sha1);
1855 1856 1857
	if (size)
		munmap(buf, size);
	return ret;
J
Junio C Hamano 已提交
1858
}
1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882

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);
		}
R
Rene Scharfe 已提交
1883 1884 1885
		if (!write_object)
			hash_sha1_file(target, st->st_size, blob_type, sha1);
		else if (write_sha1_file(target, st->st_size, blob_type, sha1))
1886 1887 1888 1889 1890 1891 1892 1893 1894
			return error("%s: failed to insert into database",
				     path);
		free(target);
		break;
	default:
		return error("%s: unsupported file type", path);
	}
	return 0;
}