]>
Commit | Line | Data |
---|---|---|
4f39cd82 | 1 | #include "cache.h" |
ec2dd32c | 2 | #include "list.h" |
0317f455 | 3 | #include "pack.h" |
031dc927 | 4 | #include "repository.h" |
0abe14f6 JT |
5 | #include "dir.h" |
6 | #include "mergesort.h" | |
7 | #include "packfile.h" | |
7b3aa75d | 8 | #include "delta.h" |
f1d8130b JT |
9 | #include "list.h" |
10 | #include "streaming.h" | |
a2551953 | 11 | #include "sha1-lookup.h" |
498f1f61 JT |
12 | #include "commit.h" |
13 | #include "object.h" | |
14 | #include "tag.h" | |
15 | #include "tree-walk.h" | |
16 | #include "tree.h" | |
0d4a1321 | 17 | #include "object-store.h" |
4f39cd82 JT |
18 | |
19 | char *odb_pack_name(struct strbuf *buf, | |
20 | const unsigned char *sha1, | |
21 | const char *ext) | |
22 | { | |
23 | strbuf_reset(buf); | |
24 | strbuf_addf(buf, "%s/pack/pack-%s.%s", get_object_directory(), | |
25 | sha1_to_hex(sha1), ext); | |
26 | return buf->buf; | |
27 | } | |
28 | ||
29 | char *sha1_pack_name(const unsigned char *sha1) | |
30 | { | |
31 | static struct strbuf buf = STRBUF_INIT; | |
32 | return odb_pack_name(&buf, sha1, "pack"); | |
33 | } | |
34 | ||
35 | char *sha1_pack_index_name(const unsigned char *sha1) | |
36 | { | |
37 | static struct strbuf buf = STRBUF_INIT; | |
38 | return odb_pack_name(&buf, sha1, "idx"); | |
39 | } | |
6d6a80e0 | 40 | |
84f80ad5 JT |
41 | static unsigned int pack_used_ctr; |
42 | static unsigned int pack_mmap_calls; | |
43 | static unsigned int peak_pack_open_windows; | |
44 | static unsigned int pack_open_windows; | |
e65f1862 | 45 | static unsigned int pack_open_fds; |
84f80ad5 JT |
46 | static unsigned int pack_max_fds; |
47 | static size_t peak_pack_mapped; | |
48 | static size_t pack_mapped; | |
8e21176c JT |
49 | |
50 | #define SZ_FMT PRIuMAX | |
51 | static inline uintmax_t sz_fmt(size_t s) { return s; } | |
52 | ||
53 | void pack_report(void) | |
54 | { | |
55 | fprintf(stderr, | |
56 | "pack_report: getpagesize() = %10" SZ_FMT "\n" | |
57 | "pack_report: core.packedGitWindowSize = %10" SZ_FMT "\n" | |
58 | "pack_report: core.packedGitLimit = %10" SZ_FMT "\n", | |
59 | sz_fmt(getpagesize()), | |
60 | sz_fmt(packed_git_window_size), | |
61 | sz_fmt(packed_git_limit)); | |
62 | fprintf(stderr, | |
63 | "pack_report: pack_used_ctr = %10u\n" | |
64 | "pack_report: pack_mmap_calls = %10u\n" | |
65 | "pack_report: pack_open_windows = %10u / %10u\n" | |
66 | "pack_report: pack_mapped = " | |
67 | "%10" SZ_FMT " / %10" SZ_FMT "\n", | |
68 | pack_used_ctr, | |
69 | pack_mmap_calls, | |
70 | pack_open_windows, peak_pack_open_windows, | |
71 | sz_fmt(pack_mapped), sz_fmt(peak_pack_mapped)); | |
72 | } | |
0317f455 JT |
73 | |
74 | /* | |
75 | * Open and mmap the index file at path, perform a couple of | |
76 | * consistency checks, then record its information to p. Return 0 on | |
77 | * success. | |
78 | */ | |
79 | static int check_packed_git_idx(const char *path, struct packed_git *p) | |
80 | { | |
81 | void *idx_map; | |
82 | struct pack_idx_header *hdr; | |
83 | size_t idx_size; | |
84 | uint32_t version, nr, i, *index; | |
85 | int fd = git_open(path); | |
86 | struct stat st; | |
37fec86a | 87 | const unsigned int hashsz = the_hash_algo->rawsz; |
0317f455 JT |
88 | |
89 | if (fd < 0) | |
90 | return -1; | |
91 | if (fstat(fd, &st)) { | |
92 | close(fd); | |
93 | return -1; | |
94 | } | |
95 | idx_size = xsize_t(st.st_size); | |
37fec86a | 96 | if (idx_size < 4 * 256 + hashsz + hashsz) { |
0317f455 JT |
97 | close(fd); |
98 | return error("index file %s is too small", path); | |
99 | } | |
100 | idx_map = xmmap(NULL, idx_size, PROT_READ, MAP_PRIVATE, fd, 0); | |
101 | close(fd); | |
102 | ||
103 | hdr = idx_map; | |
104 | if (hdr->idx_signature == htonl(PACK_IDX_SIGNATURE)) { | |
105 | version = ntohl(hdr->idx_version); | |
106 | if (version < 2 || version > 2) { | |
107 | munmap(idx_map, idx_size); | |
108 | return error("index file %s is version %"PRIu32 | |
109 | " and is not supported by this binary" | |
110 | " (try upgrading GIT to a newer version)", | |
111 | path, version); | |
112 | } | |
113 | } else | |
114 | version = 1; | |
115 | ||
116 | nr = 0; | |
117 | index = idx_map; | |
118 | if (version > 1) | |
119 | index += 2; /* skip index header */ | |
120 | for (i = 0; i < 256; i++) { | |
121 | uint32_t n = ntohl(index[i]); | |
122 | if (n < nr) { | |
123 | munmap(idx_map, idx_size); | |
124 | return error("non-monotonic index %s", path); | |
125 | } | |
126 | nr = n; | |
127 | } | |
128 | ||
129 | if (version == 1) { | |
130 | /* | |
131 | * Total size: | |
132 | * - 256 index entries 4 bytes each | |
37fec86a | 133 | * - 24-byte entries * nr (object ID + 4-byte offset) |
134 | * - hash of the packfile | |
135 | * - file checksum | |
0317f455 | 136 | */ |
37fec86a | 137 | if (idx_size != 4*256 + nr * (hashsz + 4) + hashsz + hashsz) { |
0317f455 JT |
138 | munmap(idx_map, idx_size); |
139 | return error("wrong index v1 file size in %s", path); | |
140 | } | |
141 | } else if (version == 2) { | |
142 | /* | |
143 | * Minimum size: | |
144 | * - 8 bytes of header | |
145 | * - 256 index entries 4 bytes each | |
37fec86a | 146 | * - object ID entry * nr |
0317f455 JT |
147 | * - 4-byte crc entry * nr |
148 | * - 4-byte offset entry * nr | |
37fec86a | 149 | * - hash of the packfile |
150 | * - file checksum | |
0317f455 JT |
151 | * And after the 4-byte offset table might be a |
152 | * variable sized table containing 8-byte entries | |
153 | * for offsets larger than 2^31. | |
154 | */ | |
37fec86a | 155 | unsigned long min_size = 8 + 4*256 + nr*(hashsz + 4 + 4) + hashsz + hashsz; |
0317f455 JT |
156 | unsigned long max_size = min_size; |
157 | if (nr) | |
158 | max_size += (nr - 1)*8; | |
159 | if (idx_size < min_size || idx_size > max_size) { | |
160 | munmap(idx_map, idx_size); | |
161 | return error("wrong index v2 file size in %s", path); | |
162 | } | |
163 | if (idx_size != min_size && | |
164 | /* | |
165 | * make sure we can deal with large pack offsets. | |
166 | * 31-bit signed offset won't be enough, neither | |
167 | * 32-bit unsigned one will be. | |
168 | */ | |
169 | (sizeof(off_t) <= 4)) { | |
170 | munmap(idx_map, idx_size); | |
171 | return error("pack too large for current definition of off_t in %s", path); | |
172 | } | |
173 | } | |
174 | ||
175 | p->index_version = version; | |
176 | p->index_data = idx_map; | |
177 | p->index_size = idx_size; | |
178 | p->num_objects = nr; | |
179 | return 0; | |
180 | } | |
181 | ||
182 | int open_pack_index(struct packed_git *p) | |
183 | { | |
184 | char *idx_name; | |
185 | size_t len; | |
186 | int ret; | |
187 | ||
188 | if (p->index_data) | |
189 | return 0; | |
190 | ||
191 | if (!strip_suffix(p->pack_name, ".pack", &len)) | |
033abf97 | 192 | BUG("pack_name does not end in .pack"); |
0317f455 JT |
193 | idx_name = xstrfmt("%.*s.idx", (int)len, p->pack_name); |
194 | ret = check_packed_git_idx(idx_name, p); | |
195 | free(idx_name); | |
196 | return ret; | |
197 | } | |
198 | ||
fe1ed56f DS |
199 | uint32_t get_pack_fanout(struct packed_git *p, uint32_t value) |
200 | { | |
201 | const uint32_t *level1_ofs = p->index_data; | |
202 | ||
203 | if (!level1_ofs) { | |
204 | if (open_pack_index(p)) | |
205 | return 0; | |
206 | level1_ofs = p->index_data; | |
207 | } | |
208 | ||
209 | if (p->index_version > 1) { | |
210 | level1_ofs += 2; | |
211 | } | |
212 | ||
213 | return ntohl(level1_ofs[value]); | |
214 | } | |
215 | ||
0317f455 JT |
216 | static struct packed_git *alloc_packed_git(int extra) |
217 | { | |
218 | struct packed_git *p = xmalloc(st_add(sizeof(*p), extra)); | |
219 | memset(p, 0, sizeof(*p)); | |
220 | p->pack_fd = -1; | |
221 | return p; | |
222 | } | |
223 | ||
224 | struct packed_git *parse_pack_index(unsigned char *sha1, const char *idx_path) | |
225 | { | |
226 | const char *path = sha1_pack_name(sha1); | |
227 | size_t alloc = st_add(strlen(path), 1); | |
228 | struct packed_git *p = alloc_packed_git(alloc); | |
229 | ||
230 | memcpy(p->pack_name, path, alloc); /* includes NUL */ | |
231 | hashcpy(p->sha1, sha1); | |
232 | if (check_packed_git_idx(idx_path, p)) { | |
233 | free(p); | |
234 | return NULL; | |
235 | } | |
236 | ||
237 | return p; | |
238 | } | |
f0e17e86 JT |
239 | |
240 | static void scan_windows(struct packed_git *p, | |
241 | struct packed_git **lru_p, | |
242 | struct pack_window **lru_w, | |
243 | struct pack_window **lru_l) | |
244 | { | |
245 | struct pack_window *w, *w_l; | |
246 | ||
247 | for (w_l = NULL, w = p->windows; w; w = w->next) { | |
248 | if (!w->inuse_cnt) { | |
249 | if (!*lru_w || w->last_used < (*lru_w)->last_used) { | |
250 | *lru_p = p; | |
251 | *lru_w = w; | |
252 | *lru_l = w_l; | |
253 | } | |
254 | } | |
255 | w_l = w; | |
256 | } | |
257 | } | |
258 | ||
84f80ad5 | 259 | static int unuse_one_window(struct packed_git *current) |
f0e17e86 JT |
260 | { |
261 | struct packed_git *p, *lru_p = NULL; | |
262 | struct pack_window *lru_w = NULL, *lru_l = NULL; | |
263 | ||
264 | if (current) | |
265 | scan_windows(current, &lru_p, &lru_w, &lru_l); | |
a80d72db | 266 | for (p = the_repository->objects->packed_git; p; p = p->next) |
f0e17e86 JT |
267 | scan_windows(p, &lru_p, &lru_w, &lru_l); |
268 | if (lru_p) { | |
269 | munmap(lru_w->base, lru_w->len); | |
270 | pack_mapped -= lru_w->len; | |
271 | if (lru_l) | |
272 | lru_l->next = lru_w->next; | |
273 | else | |
274 | lru_p->windows = lru_w->next; | |
275 | free(lru_w); | |
276 | pack_open_windows--; | |
277 | return 1; | |
278 | } | |
279 | return 0; | |
280 | } | |
281 | ||
282 | void release_pack_memory(size_t need) | |
283 | { | |
284 | size_t cur = pack_mapped; | |
285 | while (need >= (cur - pack_mapped) && unuse_one_window(NULL)) | |
286 | ; /* nothing */ | |
287 | } | |
3836d88a JT |
288 | |
289 | void close_pack_windows(struct packed_git *p) | |
290 | { | |
291 | while (p->windows) { | |
292 | struct pack_window *w = p->windows; | |
293 | ||
294 | if (w->inuse_cnt) | |
295 | die("pack '%s' still has open windows to it", | |
296 | p->pack_name); | |
297 | munmap(w->base, w->len); | |
298 | pack_mapped -= w->len; | |
299 | pack_open_windows--; | |
300 | p->windows = w->next; | |
301 | free(w); | |
302 | } | |
303 | } | |
304 | ||
84f80ad5 | 305 | static int close_pack_fd(struct packed_git *p) |
3836d88a JT |
306 | { |
307 | if (p->pack_fd < 0) | |
308 | return 0; | |
309 | ||
310 | close(p->pack_fd); | |
311 | pack_open_fds--; | |
312 | p->pack_fd = -1; | |
313 | ||
314 | return 1; | |
315 | } | |
316 | ||
317 | void close_pack_index(struct packed_git *p) | |
318 | { | |
319 | if (p->index_data) { | |
320 | munmap((void *)p->index_data, p->index_size); | |
321 | p->index_data = NULL; | |
322 | } | |
323 | } | |
324 | ||
049d51a2 | 325 | void close_pack(struct packed_git *p) |
3836d88a JT |
326 | { |
327 | close_pack_windows(p); | |
328 | close_pack_fd(p); | |
329 | close_pack_index(p); | |
330 | } | |
331 | ||
d0b59866 | 332 | void close_all_packs(struct raw_object_store *o) |
3836d88a JT |
333 | { |
334 | struct packed_git *p; | |
335 | ||
d0b59866 | 336 | for (p = o->packed_git; p; p = p->next) |
3836d88a | 337 | if (p->do_not_close) |
033abf97 | 338 | BUG("want to close pack marked 'do-not-close'"); |
3836d88a JT |
339 | else |
340 | close_pack(p); | |
341 | } | |
84f80ad5 JT |
342 | |
343 | /* | |
344 | * The LRU pack is the one with the oldest MRU window, preferring packs | |
345 | * with no used windows, or the oldest mtime if it has no windows allocated. | |
346 | */ | |
347 | static void find_lru_pack(struct packed_git *p, struct packed_git **lru_p, struct pack_window **mru_w, int *accept_windows_inuse) | |
348 | { | |
349 | struct pack_window *w, *this_mru_w; | |
350 | int has_windows_inuse = 0; | |
351 | ||
352 | /* | |
353 | * Reject this pack if it has windows and the previously selected | |
354 | * one does not. If this pack does not have windows, reject | |
355 | * it if the pack file is newer than the previously selected one. | |
356 | */ | |
357 | if (*lru_p && !*mru_w && (p->windows || p->mtime > (*lru_p)->mtime)) | |
358 | return; | |
359 | ||
360 | for (w = this_mru_w = p->windows; w; w = w->next) { | |
361 | /* | |
362 | * Reject this pack if any of its windows are in use, | |
363 | * but the previously selected pack did not have any | |
364 | * inuse windows. Otherwise, record that this pack | |
365 | * has windows in use. | |
366 | */ | |
367 | if (w->inuse_cnt) { | |
368 | if (*accept_windows_inuse) | |
369 | has_windows_inuse = 1; | |
370 | else | |
371 | return; | |
372 | } | |
373 | ||
374 | if (w->last_used > this_mru_w->last_used) | |
375 | this_mru_w = w; | |
376 | ||
377 | /* | |
378 | * Reject this pack if it has windows that have been | |
379 | * used more recently than the previously selected pack. | |
380 | * If the previously selected pack had windows inuse and | |
381 | * we have not encountered a window in this pack that is | |
382 | * inuse, skip this check since we prefer a pack with no | |
383 | * inuse windows to one that has inuse windows. | |
384 | */ | |
385 | if (*mru_w && *accept_windows_inuse == has_windows_inuse && | |
386 | this_mru_w->last_used > (*mru_w)->last_used) | |
387 | return; | |
388 | } | |
389 | ||
390 | /* | |
391 | * Select this pack. | |
392 | */ | |
393 | *mru_w = this_mru_w; | |
394 | *lru_p = p; | |
395 | *accept_windows_inuse = has_windows_inuse; | |
396 | } | |
397 | ||
398 | static int close_one_pack(void) | |
399 | { | |
400 | struct packed_git *p, *lru_p = NULL; | |
401 | struct pack_window *mru_w = NULL; | |
402 | int accept_windows_inuse = 1; | |
403 | ||
a80d72db | 404 | for (p = the_repository->objects->packed_git; p; p = p->next) { |
84f80ad5 JT |
405 | if (p->pack_fd == -1) |
406 | continue; | |
407 | find_lru_pack(p, &lru_p, &mru_w, &accept_windows_inuse); | |
408 | } | |
409 | ||
410 | if (lru_p) | |
411 | return close_pack_fd(lru_p); | |
412 | ||
413 | return 0; | |
414 | } | |
415 | ||
416 | static unsigned int get_max_fd_limit(void) | |
417 | { | |
418 | #ifdef RLIMIT_NOFILE | |
419 | { | |
420 | struct rlimit lim; | |
421 | ||
422 | if (!getrlimit(RLIMIT_NOFILE, &lim)) | |
423 | return lim.rlim_cur; | |
424 | } | |
425 | #endif | |
426 | ||
427 | #ifdef _SC_OPEN_MAX | |
428 | { | |
429 | long open_max = sysconf(_SC_OPEN_MAX); | |
430 | if (0 < open_max) | |
431 | return open_max; | |
432 | /* | |
433 | * Otherwise, we got -1 for one of the two | |
434 | * reasons: | |
435 | * | |
436 | * (1) sysconf() did not understand _SC_OPEN_MAX | |
437 | * and signaled an error with -1; or | |
438 | * (2) sysconf() said there is no limit. | |
439 | * | |
440 | * We _could_ clear errno before calling sysconf() to | |
441 | * tell these two cases apart and return a huge number | |
442 | * in the latter case to let the caller cap it to a | |
443 | * value that is not so selfish, but letting the | |
444 | * fallback OPEN_MAX codepath take care of these cases | |
445 | * is a lot simpler. | |
446 | */ | |
447 | } | |
448 | #endif | |
449 | ||
450 | #ifdef OPEN_MAX | |
451 | return OPEN_MAX; | |
452 | #else | |
453 | return 1; /* see the caller ;-) */ | |
454 | #endif | |
455 | } | |
456 | ||
457 | /* | |
458 | * Do not call this directly as this leaks p->pack_fd on error return; | |
459 | * call open_packed_git() instead. | |
460 | */ | |
461 | static int open_packed_git_1(struct packed_git *p) | |
462 | { | |
463 | struct stat st; | |
464 | struct pack_header hdr; | |
37fec86a | 465 | unsigned char hash[GIT_MAX_RAWSZ]; |
466 | unsigned char *idx_hash; | |
84f80ad5 | 467 | long fd_flag; |
41dcc4dc | 468 | ssize_t read_result; |
37fec86a | 469 | const unsigned hashsz = the_hash_algo->rawsz; |
84f80ad5 JT |
470 | |
471 | if (!p->index_data && open_pack_index(p)) | |
472 | return error("packfile %s index unavailable", p->pack_name); | |
473 | ||
474 | if (!pack_max_fds) { | |
475 | unsigned int max_fds = get_max_fd_limit(); | |
476 | ||
477 | /* Save 3 for stdin/stdout/stderr, 22 for work */ | |
478 | if (25 < max_fds) | |
479 | pack_max_fds = max_fds - 25; | |
480 | else | |
481 | pack_max_fds = 1; | |
482 | } | |
483 | ||
484 | while (pack_max_fds <= pack_open_fds && close_one_pack()) | |
485 | ; /* nothing */ | |
486 | ||
487 | p->pack_fd = git_open(p->pack_name); | |
488 | if (p->pack_fd < 0 || fstat(p->pack_fd, &st)) | |
489 | return -1; | |
490 | pack_open_fds++; | |
491 | ||
492 | /* If we created the struct before we had the pack we lack size. */ | |
493 | if (!p->pack_size) { | |
494 | if (!S_ISREG(st.st_mode)) | |
495 | return error("packfile %s not a regular file", p->pack_name); | |
496 | p->pack_size = st.st_size; | |
497 | } else if (p->pack_size != st.st_size) | |
498 | return error("packfile %s size changed", p->pack_name); | |
499 | ||
500 | /* We leave these file descriptors open with sliding mmap; | |
501 | * there is no point keeping them open across exec(), though. | |
502 | */ | |
503 | fd_flag = fcntl(p->pack_fd, F_GETFD, 0); | |
504 | if (fd_flag < 0) | |
505 | return error("cannot determine file descriptor flags"); | |
506 | fd_flag |= FD_CLOEXEC; | |
507 | if (fcntl(p->pack_fd, F_SETFD, fd_flag) == -1) | |
508 | return error("cannot set FD_CLOEXEC"); | |
509 | ||
510 | /* Verify we recognize this pack file format. */ | |
41dcc4dc JK |
511 | read_result = read_in_full(p->pack_fd, &hdr, sizeof(hdr)); |
512 | if (read_result < 0) | |
513 | return error_errno("error reading from %s", p->pack_name); | |
514 | if (read_result != sizeof(hdr)) | |
84f80ad5 JT |
515 | return error("file %s is far too short to be a packfile", p->pack_name); |
516 | if (hdr.hdr_signature != htonl(PACK_SIGNATURE)) | |
517 | return error("file %s is not a GIT packfile", p->pack_name); | |
518 | if (!pack_version_ok(hdr.hdr_version)) | |
519 | return error("packfile %s is version %"PRIu32" and not" | |
520 | " supported (try upgrading GIT to a newer version)", | |
521 | p->pack_name, ntohl(hdr.hdr_version)); | |
522 | ||
523 | /* Verify the pack matches its index. */ | |
524 | if (p->num_objects != ntohl(hdr.hdr_entries)) | |
525 | return error("packfile %s claims to have %"PRIu32" objects" | |
526 | " while index indicates %"PRIu32" objects", | |
527 | p->pack_name, ntohl(hdr.hdr_entries), | |
528 | p->num_objects); | |
37fec86a | 529 | if (lseek(p->pack_fd, p->pack_size - hashsz, SEEK_SET) == -1) |
84f80ad5 | 530 | return error("end of packfile %s is unavailable", p->pack_name); |
37fec86a | 531 | read_result = read_in_full(p->pack_fd, hash, hashsz); |
41dcc4dc JK |
532 | if (read_result < 0) |
533 | return error_errno("error reading from %s", p->pack_name); | |
37fec86a | 534 | if (read_result != hashsz) |
84f80ad5 | 535 | return error("packfile %s signature is unavailable", p->pack_name); |
37fec86a | 536 | idx_hash = ((unsigned char *)p->index_data) + p->index_size - hashsz * 2; |
537 | if (hashcmp(hash, idx_hash)) | |
84f80ad5 JT |
538 | return error("packfile %s does not match index", p->pack_name); |
539 | return 0; | |
540 | } | |
541 | ||
a2551953 | 542 | static int open_packed_git(struct packed_git *p) |
84f80ad5 JT |
543 | { |
544 | if (!open_packed_git_1(p)) | |
545 | return 0; | |
546 | close_pack_fd(p); | |
547 | return -1; | |
548 | } | |
549 | ||
550 | static int in_window(struct pack_window *win, off_t offset) | |
551 | { | |
37fec86a | 552 | /* We must promise at least one full hash after the |
84f80ad5 JT |
553 | * offset is available from this window, otherwise the offset |
554 | * is not actually in this window and a different window (which | |
555 | * has that one hash excess) must be used. This is to support | |
556 | * the object header and delta base parsing routines below. | |
557 | */ | |
558 | off_t win_off = win->offset; | |
559 | return win_off <= offset | |
37fec86a | 560 | && (offset + the_hash_algo->rawsz) <= (win_off + win->len); |
84f80ad5 JT |
561 | } |
562 | ||
563 | unsigned char *use_pack(struct packed_git *p, | |
564 | struct pack_window **w_cursor, | |
565 | off_t offset, | |
566 | unsigned long *left) | |
567 | { | |
568 | struct pack_window *win = *w_cursor; | |
569 | ||
570 | /* Since packfiles end in a hash of their content and it's | |
571 | * pointless to ask for an offset into the middle of that | |
572 | * hash, and the in_window function above wouldn't match | |
573 | * don't allow an offset too close to the end of the file. | |
574 | */ | |
575 | if (!p->pack_size && p->pack_fd == -1 && open_packed_git(p)) | |
576 | die("packfile %s cannot be accessed", p->pack_name); | |
37fec86a | 577 | if (offset > (p->pack_size - the_hash_algo->rawsz)) |
84f80ad5 JT |
578 | die("offset beyond end of packfile (truncated pack?)"); |
579 | if (offset < 0) | |
580 | die(_("offset before end of packfile (broken .idx?)")); | |
581 | ||
582 | if (!win || !in_window(win, offset)) { | |
583 | if (win) | |
584 | win->inuse_cnt--; | |
585 | for (win = p->windows; win; win = win->next) { | |
586 | if (in_window(win, offset)) | |
587 | break; | |
588 | } | |
589 | if (!win) { | |
590 | size_t window_align = packed_git_window_size / 2; | |
591 | off_t len; | |
592 | ||
593 | if (p->pack_fd == -1 && open_packed_git(p)) | |
594 | die("packfile %s cannot be accessed", p->pack_name); | |
595 | ||
596 | win = xcalloc(1, sizeof(*win)); | |
597 | win->offset = (offset / window_align) * window_align; | |
598 | len = p->pack_size - win->offset; | |
599 | if (len > packed_git_window_size) | |
600 | len = packed_git_window_size; | |
601 | win->len = (size_t)len; | |
602 | pack_mapped += win->len; | |
603 | while (packed_git_limit < pack_mapped | |
604 | && unuse_one_window(p)) | |
605 | ; /* nothing */ | |
606 | win->base = xmmap(NULL, win->len, | |
607 | PROT_READ, MAP_PRIVATE, | |
608 | p->pack_fd, win->offset); | |
609 | if (win->base == MAP_FAILED) | |
610 | die_errno("packfile %s cannot be mapped", | |
611 | p->pack_name); | |
612 | if (!win->offset && win->len == p->pack_size | |
613 | && !p->do_not_close) | |
614 | close_pack_fd(p); | |
615 | pack_mmap_calls++; | |
616 | pack_open_windows++; | |
617 | if (pack_mapped > peak_pack_mapped) | |
618 | peak_pack_mapped = pack_mapped; | |
619 | if (pack_open_windows > peak_pack_open_windows) | |
620 | peak_pack_open_windows = pack_open_windows; | |
621 | win->next = p->windows; | |
622 | p->windows = win; | |
623 | } | |
624 | } | |
625 | if (win != *w_cursor) { | |
626 | win->last_used = pack_used_ctr++; | |
627 | win->inuse_cnt++; | |
628 | *w_cursor = win; | |
629 | } | |
630 | offset -= win->offset; | |
631 | if (left) | |
632 | *left = win->len - xsize_t(offset); | |
633 | return win->base + offset; | |
634 | } | |
97de1803 JT |
635 | |
636 | void unuse_pack(struct pack_window **w_cursor) | |
637 | { | |
638 | struct pack_window *w = *w_cursor; | |
639 | if (w) { | |
640 | w->inuse_cnt--; | |
641 | *w_cursor = NULL; | |
642 | } | |
643 | } | |
9a428653 JT |
644 | |
645 | static void try_to_free_pack_memory(size_t size) | |
646 | { | |
647 | release_pack_memory(size); | |
648 | } | |
649 | ||
650 | struct packed_git *add_packed_git(const char *path, size_t path_len, int local) | |
651 | { | |
652 | static int have_set_try_to_free_routine; | |
653 | struct stat st; | |
654 | size_t alloc; | |
655 | struct packed_git *p; | |
656 | ||
657 | if (!have_set_try_to_free_routine) { | |
658 | have_set_try_to_free_routine = 1; | |
659 | set_try_to_free_routine(try_to_free_pack_memory); | |
660 | } | |
661 | ||
662 | /* | |
663 | * Make sure a corresponding .pack file exists and that | |
664 | * the index looks sane. | |
665 | */ | |
666 | if (!strip_suffix_mem(path, &path_len, ".idx")) | |
667 | return NULL; | |
668 | ||
669 | /* | |
498f1f61 | 670 | * ".promisor" is long enough to hold any suffix we're adding (and |
9a428653 JT |
671 | * the use xsnprintf double-checks that) |
672 | */ | |
498f1f61 | 673 | alloc = st_add3(path_len, strlen(".promisor"), 1); |
9a428653 JT |
674 | p = alloc_packed_git(alloc); |
675 | memcpy(p->pack_name, path, path_len); | |
676 | ||
677 | xsnprintf(p->pack_name + path_len, alloc - path_len, ".keep"); | |
678 | if (!access(p->pack_name, F_OK)) | |
679 | p->pack_keep = 1; | |
680 | ||
498f1f61 JT |
681 | xsnprintf(p->pack_name + path_len, alloc - path_len, ".promisor"); |
682 | if (!access(p->pack_name, F_OK)) | |
683 | p->pack_promisor = 1; | |
684 | ||
9a428653 JT |
685 | xsnprintf(p->pack_name + path_len, alloc - path_len, ".pack"); |
686 | if (stat(p->pack_name, &st) || !S_ISREG(st.st_mode)) { | |
687 | free(p); | |
688 | return NULL; | |
689 | } | |
690 | ||
691 | /* ok, it looks sane as far as we can check without | |
692 | * actually mapping the pack file. | |
693 | */ | |
694 | p->pack_size = st.st_size; | |
695 | p->pack_local = local; | |
696 | p->mtime = st.st_mtime; | |
37fec86a | 697 | if (path_len < the_hash_algo->hexsz || |
698 | get_sha1_hex(path + path_len - the_hash_algo->hexsz, p->sha1)) | |
9a428653 JT |
699 | hashclr(p->sha1); |
700 | return p; | |
701 | } | |
e65f1862 | 702 | |
5babff16 | 703 | void install_packed_git(struct repository *r, struct packed_git *pack) |
e65f1862 JT |
704 | { |
705 | if (pack->pack_fd != -1) | |
706 | pack_open_fds++; | |
707 | ||
5babff16 SB |
708 | pack->next = r->objects->packed_git; |
709 | r->objects->packed_git = pack; | |
e65f1862 | 710 | } |
0abe14f6 JT |
711 | |
712 | void (*report_garbage)(unsigned seen_bits, const char *path); | |
713 | ||
714 | static void report_helper(const struct string_list *list, | |
715 | int seen_bits, int first, int last) | |
716 | { | |
717 | if (seen_bits == (PACKDIR_FILE_PACK|PACKDIR_FILE_IDX)) | |
718 | return; | |
719 | ||
720 | for (; first < last; first++) | |
721 | report_garbage(seen_bits, list->items[first].string); | |
722 | } | |
723 | ||
724 | static void report_pack_garbage(struct string_list *list) | |
725 | { | |
726 | int i, baselen = -1, first = 0, seen_bits = 0; | |
727 | ||
728 | if (!report_garbage) | |
729 | return; | |
730 | ||
731 | string_list_sort(list); | |
732 | ||
733 | for (i = 0; i < list->nr; i++) { | |
734 | const char *path = list->items[i].string; | |
735 | if (baselen != -1 && | |
736 | strncmp(path, list->items[first].string, baselen)) { | |
737 | report_helper(list, seen_bits, first, i); | |
738 | baselen = -1; | |
739 | seen_bits = 0; | |
740 | } | |
741 | if (baselen == -1) { | |
742 | const char *dot = strrchr(path, '.'); | |
743 | if (!dot) { | |
744 | report_garbage(PACKDIR_FILE_GARBAGE, path); | |
745 | continue; | |
746 | } | |
747 | baselen = dot - path + 1; | |
748 | first = i; | |
749 | } | |
750 | if (!strcmp(path + baselen, "pack")) | |
751 | seen_bits |= 1; | |
752 | else if (!strcmp(path + baselen, "idx")) | |
753 | seen_bits |= 2; | |
754 | } | |
755 | report_helper(list, seen_bits, first, list->nr); | |
756 | } | |
757 | ||
9208e318 DS |
758 | void for_each_file_in_pack_dir(const char *objdir, |
759 | each_file_in_pack_dir_fn fn, | |
760 | void *data) | |
0abe14f6 JT |
761 | { |
762 | struct strbuf path = STRBUF_INIT; | |
763 | size_t dirnamelen; | |
764 | DIR *dir; | |
765 | struct dirent *de; | |
0abe14f6 JT |
766 | |
767 | strbuf_addstr(&path, objdir); | |
768 | strbuf_addstr(&path, "/pack"); | |
769 | dir = opendir(path.buf); | |
770 | if (!dir) { | |
771 | if (errno != ENOENT) | |
772 | error_errno("unable to open object pack directory: %s", | |
773 | path.buf); | |
774 | strbuf_release(&path); | |
775 | return; | |
776 | } | |
777 | strbuf_addch(&path, '/'); | |
778 | dirnamelen = path.len; | |
779 | while ((de = readdir(dir)) != NULL) { | |
0abe14f6 JT |
780 | if (is_dot_or_dotdot(de->d_name)) |
781 | continue; | |
782 | ||
783 | strbuf_setlen(&path, dirnamelen); | |
784 | strbuf_addstr(&path, de->d_name); | |
785 | ||
9208e318 | 786 | fn(path.buf, path.len, de->d_name, data); |
0abe14f6 | 787 | } |
9208e318 | 788 | |
0abe14f6 | 789 | closedir(dir); |
0abe14f6 JT |
790 | strbuf_release(&path); |
791 | } | |
792 | ||
9208e318 DS |
793 | struct prepare_pack_data { |
794 | struct repository *r; | |
795 | struct string_list *garbage; | |
796 | int local; | |
797 | }; | |
798 | ||
799 | static void prepare_pack(const char *full_name, size_t full_name_len, | |
800 | const char *file_name, void *_data) | |
801 | { | |
802 | struct prepare_pack_data *data = (struct prepare_pack_data *)_data; | |
803 | struct packed_git *p; | |
804 | size_t base_len = full_name_len; | |
805 | ||
806 | if (strip_suffix_mem(full_name, &base_len, ".idx")) { | |
807 | /* Don't reopen a pack we already have. */ | |
808 | for (p = data->r->objects->packed_git; p; p = p->next) { | |
809 | size_t len; | |
810 | if (strip_suffix(p->pack_name, ".pack", &len) && | |
811 | len == base_len && | |
812 | !memcmp(p->pack_name, full_name, len)) | |
813 | break; | |
814 | } | |
815 | ||
816 | if (!p) { | |
817 | p = add_packed_git(full_name, full_name_len, data->local); | |
818 | if (p) | |
819 | install_packed_git(data->r, p); | |
820 | } | |
821 | } | |
822 | ||
823 | if (!report_garbage) | |
824 | return; | |
825 | ||
826 | if (ends_with(file_name, ".idx") || | |
827 | ends_with(file_name, ".pack") || | |
828 | ends_with(file_name, ".bitmap") || | |
829 | ends_with(file_name, ".keep") || | |
830 | ends_with(file_name, ".promisor")) | |
831 | string_list_append(data->garbage, full_name); | |
832 | else | |
833 | report_garbage(PACKDIR_FILE_GARBAGE, full_name); | |
834 | } | |
835 | ||
836 | static void prepare_packed_git_one(struct repository *r, char *objdir, int local) | |
837 | { | |
838 | struct prepare_pack_data data; | |
839 | struct string_list garbage = STRING_LIST_INIT_DUP; | |
840 | ||
841 | data.r = r; | |
842 | data.garbage = &garbage; | |
843 | data.local = local; | |
844 | ||
845 | for_each_file_in_pack_dir(objdir, prepare_pack, &data); | |
846 | ||
847 | report_pack_garbage(data.garbage); | |
848 | string_list_clear(data.garbage, 0); | |
849 | } | |
850 | ||
464416a2 | 851 | static void prepare_packed_git(struct repository *r); |
0abe14f6 JT |
852 | /* |
853 | * Give a fast, rough count of the number of objects in the repository. This | |
854 | * ignores loose objects completely. If you have a lot of them, then either | |
855 | * you should repack because your performance will be awful, or they are | |
856 | * all unreachable objects about to be pruned, in which case they're not really | |
857 | * interesting as a measure of repo size in the first place. | |
858 | */ | |
859 | unsigned long approximate_object_count(void) | |
860 | { | |
9a00580d SB |
861 | if (!the_repository->objects->approximate_object_count_valid) { |
862 | unsigned long count; | |
0abe14f6 JT |
863 | struct packed_git *p; |
864 | ||
6fdb4e9f | 865 | prepare_packed_git(the_repository); |
0abe14f6 | 866 | count = 0; |
a80d72db | 867 | for (p = the_repository->objects->packed_git; p; p = p->next) { |
0abe14f6 JT |
868 | if (open_pack_index(p)) |
869 | continue; | |
870 | count += p->num_objects; | |
871 | } | |
9a00580d | 872 | the_repository->objects->approximate_object_count = count; |
0abe14f6 | 873 | } |
9a00580d | 874 | return the_repository->objects->approximate_object_count; |
0abe14f6 JT |
875 | } |
876 | ||
877 | static void *get_next_packed_git(const void *p) | |
878 | { | |
879 | return ((const struct packed_git *)p)->next; | |
880 | } | |
881 | ||
882 | static void set_next_packed_git(void *p, void *next) | |
883 | { | |
884 | ((struct packed_git *)p)->next = next; | |
885 | } | |
886 | ||
887 | static int sort_pack(const void *a_, const void *b_) | |
888 | { | |
889 | const struct packed_git *a = a_; | |
890 | const struct packed_git *b = b_; | |
891 | int st; | |
892 | ||
893 | /* | |
894 | * Local packs tend to contain objects specific to our | |
895 | * variant of the project than remote ones. In addition, | |
896 | * remote ones could be on a network mounted filesystem. | |
897 | * Favor local ones for these reasons. | |
898 | */ | |
899 | st = a->pack_local - b->pack_local; | |
900 | if (st) | |
901 | return -st; | |
902 | ||
903 | /* | |
904 | * Younger packs tend to contain more recent objects, | |
905 | * and more recent objects tend to get accessed more | |
906 | * often. | |
907 | */ | |
908 | if (a->mtime < b->mtime) | |
909 | return 1; | |
910 | else if (a->mtime == b->mtime) | |
911 | return 0; | |
912 | return -1; | |
913 | } | |
914 | ||
c235beac | 915 | static void rearrange_packed_git(struct repository *r) |
0abe14f6 | 916 | { |
c235beac SB |
917 | r->objects->packed_git = llist_mergesort( |
918 | r->objects->packed_git, get_next_packed_git, | |
a80d72db | 919 | set_next_packed_git, sort_pack); |
0abe14f6 JT |
920 | } |
921 | ||
804be796 | 922 | static void prepare_packed_git_mru(struct repository *r) |
0abe14f6 JT |
923 | { |
924 | struct packed_git *p; | |
925 | ||
804be796 | 926 | INIT_LIST_HEAD(&r->objects->packed_git_mru); |
ec2dd32c | 927 | |
804be796 SB |
928 | for (p = r->objects->packed_git; p; p = p->next) |
929 | list_add_tail(&p->mru, &r->objects->packed_git_mru); | |
0abe14f6 JT |
930 | } |
931 | ||
464416a2 | 932 | static void prepare_packed_git(struct repository *r) |
0abe14f6 JT |
933 | { |
934 | struct alternate_object_database *alt; | |
935 | ||
0f90a9f2 | 936 | if (r->objects->packed_git_initialized) |
0abe14f6 | 937 | return; |
0f90a9f2 SB |
938 | prepare_packed_git_one(r, r->objects->objectdir, 1); |
939 | prepare_alt_odb(r); | |
940 | for (alt = r->objects->alt_odb_list; alt; alt = alt->next) | |
941 | prepare_packed_git_one(r, alt->path, 0); | |
942 | rearrange_packed_git(r); | |
943 | prepare_packed_git_mru(r); | |
944 | r->objects->packed_git_initialized = 1; | |
0abe14f6 JT |
945 | } |
946 | ||
4c2a13b4 | 947 | void reprepare_packed_git(struct repository *r) |
0abe14f6 | 948 | { |
4c2a13b4 SB |
949 | r->objects->approximate_object_count_valid = 0; |
950 | r->objects->packed_git_initialized = 0; | |
951 | prepare_packed_git(r); | |
0abe14f6 | 952 | } |
32b42e15 | 953 | |
a80d72db SB |
954 | struct packed_git *get_packed_git(struct repository *r) |
955 | { | |
464416a2 | 956 | prepare_packed_git(r); |
a80d72db SB |
957 | return r->objects->packed_git; |
958 | } | |
959 | ||
960 | struct list_head *get_packed_git_mru(struct repository *r) | |
961 | { | |
464416a2 | 962 | prepare_packed_git(r); |
a80d72db SB |
963 | return &r->objects->packed_git_mru; |
964 | } | |
965 | ||
32b42e15 JT |
966 | unsigned long unpack_object_header_buffer(const unsigned char *buf, |
967 | unsigned long len, enum object_type *type, unsigned long *sizep) | |
968 | { | |
969 | unsigned shift; | |
970 | unsigned long size, c; | |
971 | unsigned long used = 0; | |
972 | ||
973 | c = buf[used++]; | |
974 | *type = (c >> 4) & 7; | |
975 | size = c & 15; | |
976 | shift = 4; | |
977 | while (c & 0x80) { | |
978 | if (len <= used || bitsizeof(long) <= shift) { | |
979 | error("bad object header"); | |
980 | size = used = 0; | |
981 | break; | |
982 | } | |
983 | c = buf[used++]; | |
984 | size += (c & 0x7f) << shift; | |
985 | shift += 7; | |
986 | } | |
987 | *sizep = size; | |
988 | return used; | |
989 | } | |
7b3aa75d JT |
990 | |
991 | unsigned long get_size_from_delta(struct packed_git *p, | |
992 | struct pack_window **w_curs, | |
993 | off_t curpos) | |
994 | { | |
995 | const unsigned char *data; | |
996 | unsigned char delta_head[20], *in; | |
997 | git_zstream stream; | |
998 | int st; | |
999 | ||
1000 | memset(&stream, 0, sizeof(stream)); | |
1001 | stream.next_out = delta_head; | |
1002 | stream.avail_out = sizeof(delta_head); | |
1003 | ||
1004 | git_inflate_init(&stream); | |
1005 | do { | |
1006 | in = use_pack(p, w_curs, curpos, &stream.avail_in); | |
1007 | stream.next_in = in; | |
1008 | st = git_inflate(&stream, Z_FINISH); | |
1009 | curpos += stream.next_in - in; | |
1010 | } while ((st == Z_OK || st == Z_BUF_ERROR) && | |
1011 | stream.total_out < sizeof(delta_head)); | |
1012 | git_inflate_end(&stream); | |
1013 | if ((st != Z_STREAM_END) && stream.total_out != sizeof(delta_head)) { | |
1014 | error("delta data unpack-initial failed"); | |
1015 | return 0; | |
1016 | } | |
1017 | ||
1018 | /* Examine the initial part of the delta to figure out | |
1019 | * the result size. | |
1020 | */ | |
1021 | data = delta_head; | |
1022 | ||
1023 | /* ignore base size */ | |
1024 | get_delta_hdr_size(&data, delta_head+sizeof(delta_head)); | |
1025 | ||
1026 | /* Read the result size */ | |
1027 | return get_delta_hdr_size(&data, delta_head+sizeof(delta_head)); | |
1028 | } | |
3588dd6e JT |
1029 | |
1030 | int unpack_object_header(struct packed_git *p, | |
1031 | struct pack_window **w_curs, | |
1032 | off_t *curpos, | |
1033 | unsigned long *sizep) | |
1034 | { | |
1035 | unsigned char *base; | |
1036 | unsigned long left; | |
1037 | unsigned long used; | |
1038 | enum object_type type; | |
1039 | ||
1040 | /* use_pack() assures us we have [base, base + 20) available | |
1041 | * as a range that we can look at. (Its actually the hash | |
1042 | * size that is assured.) With our object header encoding | |
1043 | * the maximum deflated object size is 2^137, which is just | |
1044 | * insane, so we know won't exceed what we have been given. | |
1045 | */ | |
1046 | base = use_pack(p, w_curs, *curpos, &left); | |
1047 | used = unpack_object_header_buffer(base, left, &type, sizep); | |
1048 | if (!used) { | |
1049 | type = OBJ_BAD; | |
1050 | } else | |
1051 | *curpos += used; | |
1052 | ||
1053 | return type; | |
1054 | } | |
f1d8130b JT |
1055 | |
1056 | void mark_bad_packed_object(struct packed_git *p, const unsigned char *sha1) | |
1057 | { | |
1058 | unsigned i; | |
1059 | for (i = 0; i < p->num_bad_objects; i++) | |
1060 | if (!hashcmp(sha1, p->bad_object_sha1 + GIT_SHA1_RAWSZ * i)) | |
1061 | return; | |
1062 | p->bad_object_sha1 = xrealloc(p->bad_object_sha1, | |
1063 | st_mult(GIT_MAX_RAWSZ, | |
1064 | st_add(p->num_bad_objects, 1))); | |
1065 | hashcpy(p->bad_object_sha1 + GIT_SHA1_RAWSZ * p->num_bad_objects, sha1); | |
1066 | p->num_bad_objects++; | |
1067 | } | |
1068 | ||
1069 | const struct packed_git *has_packed_and_bad(const unsigned char *sha1) | |
1070 | { | |
1071 | struct packed_git *p; | |
1072 | unsigned i; | |
1073 | ||
a80d72db | 1074 | for (p = the_repository->objects->packed_git; p; p = p->next) |
f1d8130b | 1075 | for (i = 0; i < p->num_bad_objects; i++) |
37fec86a | 1076 | if (!hashcmp(sha1, |
1077 | p->bad_object_sha1 + the_hash_algo->rawsz * i)) | |
f1d8130b JT |
1078 | return p; |
1079 | return NULL; | |
1080 | } | |
1081 | ||
1082 | static off_t get_delta_base(struct packed_git *p, | |
1083 | struct pack_window **w_curs, | |
1084 | off_t *curpos, | |
1085 | enum object_type type, | |
1086 | off_t delta_obj_offset) | |
1087 | { | |
1088 | unsigned char *base_info = use_pack(p, w_curs, *curpos, NULL); | |
1089 | off_t base_offset; | |
1090 | ||
1091 | /* use_pack() assured us we have [base_info, base_info + 20) | |
1092 | * as a range that we can look at without walking off the | |
1093 | * end of the mapped window. Its actually the hash size | |
1094 | * that is assured. An OFS_DELTA longer than the hash size | |
1095 | * is stupid, as then a REF_DELTA would be smaller to store. | |
1096 | */ | |
1097 | if (type == OBJ_OFS_DELTA) { | |
1098 | unsigned used = 0; | |
1099 | unsigned char c = base_info[used++]; | |
1100 | base_offset = c & 127; | |
1101 | while (c & 128) { | |
1102 | base_offset += 1; | |
1103 | if (!base_offset || MSB(base_offset, 7)) | |
1104 | return 0; /* overflow */ | |
1105 | c = base_info[used++]; | |
1106 | base_offset = (base_offset << 7) + (c & 127); | |
1107 | } | |
1108 | base_offset = delta_obj_offset - base_offset; | |
1109 | if (base_offset <= 0 || base_offset >= delta_obj_offset) | |
1110 | return 0; /* out of bound */ | |
1111 | *curpos += used; | |
1112 | } else if (type == OBJ_REF_DELTA) { | |
1113 | /* The base entry _must_ be in the same pack */ | |
1114 | base_offset = find_pack_entry_one(base_info, p); | |
37fec86a | 1115 | *curpos += the_hash_algo->rawsz; |
f1d8130b JT |
1116 | } else |
1117 | die("I am totally screwed"); | |
1118 | return base_offset; | |
1119 | } | |
1120 | ||
1121 | /* | |
1122 | * Like get_delta_base above, but we return the sha1 instead of the pack | |
1123 | * offset. This means it is cheaper for REF deltas (we do not have to do | |
1124 | * the final object lookup), but more expensive for OFS deltas (we | |
1125 | * have to load the revidx to convert the offset back into a sha1). | |
1126 | */ | |
1127 | static const unsigned char *get_delta_base_sha1(struct packed_git *p, | |
1128 | struct pack_window **w_curs, | |
1129 | off_t curpos, | |
1130 | enum object_type type, | |
1131 | off_t delta_obj_offset) | |
1132 | { | |
1133 | if (type == OBJ_REF_DELTA) { | |
1134 | unsigned char *base = use_pack(p, w_curs, curpos, NULL); | |
1135 | return base; | |
1136 | } else if (type == OBJ_OFS_DELTA) { | |
1137 | struct revindex_entry *revidx; | |
1138 | off_t base_offset = get_delta_base(p, w_curs, &curpos, | |
1139 | type, delta_obj_offset); | |
1140 | ||
1141 | if (!base_offset) | |
1142 | return NULL; | |
1143 | ||
1144 | revidx = find_pack_revindex(p, base_offset); | |
1145 | if (!revidx) | |
1146 | return NULL; | |
1147 | ||
1148 | return nth_packed_object_sha1(p, revidx->nr); | |
1149 | } else | |
1150 | return NULL; | |
1151 | } | |
1152 | ||
9d98354f SB |
1153 | static int retry_bad_packed_offset(struct repository *r, |
1154 | struct packed_git *p, | |
1155 | off_t obj_offset) | |
f1d8130b JT |
1156 | { |
1157 | int type; | |
1158 | struct revindex_entry *revidx; | |
d169d664 | 1159 | struct object_id oid; |
f1d8130b JT |
1160 | revidx = find_pack_revindex(p, obj_offset); |
1161 | if (!revidx) | |
1162 | return OBJ_BAD; | |
d169d664 | 1163 | nth_packed_object_oid(&oid, p, revidx->nr); |
1164 | mark_bad_packed_object(p, oid.hash); | |
9d98354f | 1165 | type = oid_object_info(r, &oid, NULL); |
f1d8130b JT |
1166 | if (type <= OBJ_NONE) |
1167 | return OBJ_BAD; | |
1168 | return type; | |
1169 | } | |
1170 | ||
1171 | #define POI_STACK_PREALLOC 64 | |
1172 | ||
9d98354f SB |
1173 | static enum object_type packed_to_object_type(struct repository *r, |
1174 | struct packed_git *p, | |
f1d8130b JT |
1175 | off_t obj_offset, |
1176 | enum object_type type, | |
1177 | struct pack_window **w_curs, | |
1178 | off_t curpos) | |
1179 | { | |
1180 | off_t small_poi_stack[POI_STACK_PREALLOC]; | |
1181 | off_t *poi_stack = small_poi_stack; | |
1182 | int poi_stack_nr = 0, poi_stack_alloc = POI_STACK_PREALLOC; | |
1183 | ||
1184 | while (type == OBJ_OFS_DELTA || type == OBJ_REF_DELTA) { | |
1185 | off_t base_offset; | |
1186 | unsigned long size; | |
1187 | /* Push the object we're going to leave behind */ | |
1188 | if (poi_stack_nr >= poi_stack_alloc && poi_stack == small_poi_stack) { | |
1189 | poi_stack_alloc = alloc_nr(poi_stack_nr); | |
1190 | ALLOC_ARRAY(poi_stack, poi_stack_alloc); | |
1191 | memcpy(poi_stack, small_poi_stack, sizeof(off_t)*poi_stack_nr); | |
1192 | } else { | |
1193 | ALLOC_GROW(poi_stack, poi_stack_nr+1, poi_stack_alloc); | |
1194 | } | |
1195 | poi_stack[poi_stack_nr++] = obj_offset; | |
1196 | /* If parsing the base offset fails, just unwind */ | |
1197 | base_offset = get_delta_base(p, w_curs, &curpos, type, obj_offset); | |
1198 | if (!base_offset) | |
1199 | goto unwind; | |
1200 | curpos = obj_offset = base_offset; | |
1201 | type = unpack_object_header(p, w_curs, &curpos, &size); | |
1202 | if (type <= OBJ_NONE) { | |
1203 | /* If getting the base itself fails, we first | |
1204 | * retry the base, otherwise unwind */ | |
9d98354f | 1205 | type = retry_bad_packed_offset(r, p, base_offset); |
f1d8130b JT |
1206 | if (type > OBJ_NONE) |
1207 | goto out; | |
1208 | goto unwind; | |
1209 | } | |
1210 | } | |
1211 | ||
1212 | switch (type) { | |
1213 | case OBJ_BAD: | |
1214 | case OBJ_COMMIT: | |
1215 | case OBJ_TREE: | |
1216 | case OBJ_BLOB: | |
1217 | case OBJ_TAG: | |
1218 | break; | |
1219 | default: | |
1220 | error("unknown object type %i at offset %"PRIuMAX" in %s", | |
1221 | type, (uintmax_t)obj_offset, p->pack_name); | |
1222 | type = OBJ_BAD; | |
1223 | } | |
1224 | ||
1225 | out: | |
1226 | if (poi_stack != small_poi_stack) | |
1227 | free(poi_stack); | |
1228 | return type; | |
1229 | ||
1230 | unwind: | |
1231 | while (poi_stack_nr) { | |
1232 | obj_offset = poi_stack[--poi_stack_nr]; | |
9d98354f | 1233 | type = retry_bad_packed_offset(r, p, obj_offset); |
f1d8130b JT |
1234 | if (type > OBJ_NONE) |
1235 | goto out; | |
1236 | } | |
1237 | type = OBJ_BAD; | |
1238 | goto out; | |
1239 | } | |
1240 | ||
1241 | static struct hashmap delta_base_cache; | |
1242 | static size_t delta_base_cached; | |
1243 | ||
1244 | static LIST_HEAD(delta_base_cache_lru); | |
1245 | ||
1246 | struct delta_base_cache_key { | |
1247 | struct packed_git *p; | |
1248 | off_t base_offset; | |
1249 | }; | |
1250 | ||
1251 | struct delta_base_cache_entry { | |
1252 | struct hashmap hash; | |
1253 | struct delta_base_cache_key key; | |
1254 | struct list_head lru; | |
1255 | void *data; | |
1256 | unsigned long size; | |
1257 | enum object_type type; | |
1258 | }; | |
1259 | ||
1260 | static unsigned int pack_entry_hash(struct packed_git *p, off_t base_offset) | |
1261 | { | |
1262 | unsigned int hash; | |
1263 | ||
1264 | hash = (unsigned int)(intptr_t)p + (unsigned int)base_offset; | |
1265 | hash += (hash >> 8) + (hash >> 16); | |
1266 | return hash; | |
1267 | } | |
1268 | ||
1269 | static struct delta_base_cache_entry * | |
1270 | get_delta_base_cache_entry(struct packed_git *p, off_t base_offset) | |
1271 | { | |
1272 | struct hashmap_entry entry; | |
1273 | struct delta_base_cache_key key; | |
1274 | ||
1275 | if (!delta_base_cache.cmpfn) | |
1276 | return NULL; | |
1277 | ||
1278 | hashmap_entry_init(&entry, pack_entry_hash(p, base_offset)); | |
1279 | key.p = p; | |
1280 | key.base_offset = base_offset; | |
1281 | return hashmap_get(&delta_base_cache, &entry, &key); | |
1282 | } | |
1283 | ||
1284 | static int delta_base_cache_key_eq(const struct delta_base_cache_key *a, | |
1285 | const struct delta_base_cache_key *b) | |
1286 | { | |
1287 | return a->p == b->p && a->base_offset == b->base_offset; | |
1288 | } | |
1289 | ||
1290 | static int delta_base_cache_hash_cmp(const void *unused_cmp_data, | |
1291 | const void *va, const void *vb, | |
1292 | const void *vkey) | |
1293 | { | |
1294 | const struct delta_base_cache_entry *a = va, *b = vb; | |
1295 | const struct delta_base_cache_key *key = vkey; | |
1296 | if (key) | |
1297 | return !delta_base_cache_key_eq(&a->key, key); | |
1298 | else | |
1299 | return !delta_base_cache_key_eq(&a->key, &b->key); | |
1300 | } | |
1301 | ||
1302 | static int in_delta_base_cache(struct packed_git *p, off_t base_offset) | |
1303 | { | |
1304 | return !!get_delta_base_cache_entry(p, base_offset); | |
1305 | } | |
1306 | ||
1307 | /* | |
1308 | * Remove the entry from the cache, but do _not_ free the associated | |
1309 | * entry data. The caller takes ownership of the "data" buffer, and | |
1310 | * should copy out any fields it wants before detaching. | |
1311 | */ | |
1312 | static void detach_delta_base_cache_entry(struct delta_base_cache_entry *ent) | |
1313 | { | |
1314 | hashmap_remove(&delta_base_cache, ent, &ent->key); | |
1315 | list_del(&ent->lru); | |
1316 | delta_base_cached -= ent->size; | |
1317 | free(ent); | |
1318 | } | |
1319 | ||
9d98354f SB |
1320 | static void *cache_or_unpack_entry(struct repository *r, struct packed_git *p, |
1321 | off_t base_offset, unsigned long *base_size, | |
1322 | enum object_type *type) | |
f1d8130b JT |
1323 | { |
1324 | struct delta_base_cache_entry *ent; | |
1325 | ||
1326 | ent = get_delta_base_cache_entry(p, base_offset); | |
1327 | if (!ent) | |
9d98354f | 1328 | return unpack_entry(r, p, base_offset, type, base_size); |
f1d8130b JT |
1329 | |
1330 | if (type) | |
1331 | *type = ent->type; | |
1332 | if (base_size) | |
1333 | *base_size = ent->size; | |
1334 | return xmemdupz(ent->data, ent->size); | |
1335 | } | |
1336 | ||
1337 | static inline void release_delta_base_cache(struct delta_base_cache_entry *ent) | |
1338 | { | |
1339 | free(ent->data); | |
1340 | detach_delta_base_cache_entry(ent); | |
1341 | } | |
1342 | ||
1343 | void clear_delta_base_cache(void) | |
1344 | { | |
1345 | struct list_head *lru, *tmp; | |
1346 | list_for_each_safe(lru, tmp, &delta_base_cache_lru) { | |
1347 | struct delta_base_cache_entry *entry = | |
1348 | list_entry(lru, struct delta_base_cache_entry, lru); | |
1349 | release_delta_base_cache(entry); | |
1350 | } | |
1351 | } | |
1352 | ||
1353 | static void add_delta_base_cache(struct packed_git *p, off_t base_offset, | |
1354 | void *base, unsigned long base_size, enum object_type type) | |
1355 | { | |
1356 | struct delta_base_cache_entry *ent = xmalloc(sizeof(*ent)); | |
1357 | struct list_head *lru, *tmp; | |
1358 | ||
1359 | delta_base_cached += base_size; | |
1360 | ||
1361 | list_for_each_safe(lru, tmp, &delta_base_cache_lru) { | |
1362 | struct delta_base_cache_entry *f = | |
1363 | list_entry(lru, struct delta_base_cache_entry, lru); | |
1364 | if (delta_base_cached <= delta_base_cache_limit) | |
1365 | break; | |
1366 | release_delta_base_cache(f); | |
1367 | } | |
1368 | ||
1369 | ent->key.p = p; | |
1370 | ent->key.base_offset = base_offset; | |
1371 | ent->type = type; | |
1372 | ent->data = base; | |
1373 | ent->size = base_size; | |
1374 | list_add_tail(&ent->lru, &delta_base_cache_lru); | |
1375 | ||
1376 | if (!delta_base_cache.cmpfn) | |
1377 | hashmap_init(&delta_base_cache, delta_base_cache_hash_cmp, NULL, 0); | |
1378 | hashmap_entry_init(ent, pack_entry_hash(p, base_offset)); | |
1379 | hashmap_add(&delta_base_cache, ent); | |
1380 | } | |
1381 | ||
9d98354f SB |
1382 | int packed_object_info(struct repository *r, struct packed_git *p, |
1383 | off_t obj_offset, struct object_info *oi) | |
f1d8130b JT |
1384 | { |
1385 | struct pack_window *w_curs = NULL; | |
1386 | unsigned long size; | |
1387 | off_t curpos = obj_offset; | |
1388 | enum object_type type; | |
1389 | ||
1390 | /* | |
1391 | * We always get the representation type, but only convert it to | |
1392 | * a "real" type later if the caller is interested. | |
1393 | */ | |
1394 | if (oi->contentp) { | |
9d98354f | 1395 | *oi->contentp = cache_or_unpack_entry(r, p, obj_offset, oi->sizep, |
f1d8130b JT |
1396 | &type); |
1397 | if (!*oi->contentp) | |
1398 | type = OBJ_BAD; | |
1399 | } else { | |
1400 | type = unpack_object_header(p, &w_curs, &curpos, &size); | |
1401 | } | |
1402 | ||
1403 | if (!oi->contentp && oi->sizep) { | |
1404 | if (type == OBJ_OFS_DELTA || type == OBJ_REF_DELTA) { | |
1405 | off_t tmp_pos = curpos; | |
1406 | off_t base_offset = get_delta_base(p, &w_curs, &tmp_pos, | |
1407 | type, obj_offset); | |
1408 | if (!base_offset) { | |
1409 | type = OBJ_BAD; | |
1410 | goto out; | |
1411 | } | |
1412 | *oi->sizep = get_size_from_delta(p, &w_curs, tmp_pos); | |
1413 | if (*oi->sizep == 0) { | |
1414 | type = OBJ_BAD; | |
1415 | goto out; | |
1416 | } | |
1417 | } else { | |
1418 | *oi->sizep = size; | |
1419 | } | |
1420 | } | |
1421 | ||
1422 | if (oi->disk_sizep) { | |
1423 | struct revindex_entry *revidx = find_pack_revindex(p, obj_offset); | |
1424 | *oi->disk_sizep = revidx[1].offset - obj_offset; | |
1425 | } | |
1426 | ||
6ca32f47 | 1427 | if (oi->typep || oi->type_name) { |
f1d8130b | 1428 | enum object_type ptot; |
9d98354f | 1429 | ptot = packed_to_object_type(r, p, obj_offset, |
144f4948 | 1430 | type, &w_curs, curpos); |
f1d8130b JT |
1431 | if (oi->typep) |
1432 | *oi->typep = ptot; | |
6ca32f47 | 1433 | if (oi->type_name) { |
debca9d2 | 1434 | const char *tn = type_name(ptot); |
f1d8130b | 1435 | if (tn) |
6ca32f47 | 1436 | strbuf_addstr(oi->type_name, tn); |
f1d8130b JT |
1437 | } |
1438 | if (ptot < 0) { | |
1439 | type = OBJ_BAD; | |
1440 | goto out; | |
1441 | } | |
1442 | } | |
1443 | ||
1444 | if (oi->delta_base_sha1) { | |
1445 | if (type == OBJ_OFS_DELTA || type == OBJ_REF_DELTA) { | |
1446 | const unsigned char *base; | |
1447 | ||
1448 | base = get_delta_base_sha1(p, &w_curs, curpos, | |
1449 | type, obj_offset); | |
1450 | if (!base) { | |
1451 | type = OBJ_BAD; | |
1452 | goto out; | |
1453 | } | |
1454 | ||
1455 | hashcpy(oi->delta_base_sha1, base); | |
1456 | } else | |
1457 | hashclr(oi->delta_base_sha1); | |
1458 | } | |
1459 | ||
1460 | oi->whence = in_delta_base_cache(p, obj_offset) ? OI_DBCACHED : | |
1461 | OI_PACKED; | |
1462 | ||
1463 | out: | |
1464 | unuse_pack(&w_curs); | |
1465 | return type; | |
1466 | } | |
1467 | ||
1468 | static void *unpack_compressed_entry(struct packed_git *p, | |
1469 | struct pack_window **w_curs, | |
1470 | off_t curpos, | |
1471 | unsigned long size) | |
1472 | { | |
1473 | int st; | |
1474 | git_zstream stream; | |
1475 | unsigned char *buffer, *in; | |
1476 | ||
1477 | buffer = xmallocz_gently(size); | |
1478 | if (!buffer) | |
1479 | return NULL; | |
1480 | memset(&stream, 0, sizeof(stream)); | |
1481 | stream.next_out = buffer; | |
1482 | stream.avail_out = size + 1; | |
1483 | ||
1484 | git_inflate_init(&stream); | |
1485 | do { | |
1486 | in = use_pack(p, w_curs, curpos, &stream.avail_in); | |
1487 | stream.next_in = in; | |
1488 | st = git_inflate(&stream, Z_FINISH); | |
1489 | if (!stream.avail_out) | |
1490 | break; /* the payload is larger than it should be */ | |
1491 | curpos += stream.next_in - in; | |
1492 | } while (st == Z_OK || st == Z_BUF_ERROR); | |
1493 | git_inflate_end(&stream); | |
1494 | if ((st != Z_STREAM_END) || stream.total_out != size) { | |
1495 | free(buffer); | |
1496 | return NULL; | |
1497 | } | |
1498 | ||
b611396e JL |
1499 | /* versions of zlib can clobber unconsumed portion of outbuf */ |
1500 | buffer[size] = '\0'; | |
1501 | ||
f1d8130b JT |
1502 | return buffer; |
1503 | } | |
1504 | ||
1505 | static void write_pack_access_log(struct packed_git *p, off_t obj_offset) | |
1506 | { | |
1507 | static struct trace_key pack_access = TRACE_KEY_INIT(PACK_ACCESS); | |
1508 | trace_printf_key(&pack_access, "%s %"PRIuMAX"\n", | |
1509 | p->pack_name, (uintmax_t)obj_offset); | |
1510 | } | |
1511 | ||
1512 | int do_check_packed_object_crc; | |
1513 | ||
1514 | #define UNPACK_ENTRY_STACK_PREALLOC 64 | |
1515 | struct unpack_entry_stack_ent { | |
1516 | off_t obj_offset; | |
1517 | off_t curpos; | |
1518 | unsigned long size; | |
1519 | }; | |
1520 | ||
9d98354f SB |
1521 | static void *read_object(struct repository *r, |
1522 | const struct object_id *oid, | |
1523 | enum object_type *type, | |
f1d8130b JT |
1524 | unsigned long *size) |
1525 | { | |
1526 | struct object_info oi = OBJECT_INFO_INIT; | |
1527 | void *content; | |
1528 | oi.typep = type; | |
1529 | oi.sizep = size; | |
1530 | oi.contentp = &content; | |
1531 | ||
9d98354f | 1532 | if (oid_object_info_extended(r, oid, &oi, 0) < 0) |
f1d8130b JT |
1533 | return NULL; |
1534 | return content; | |
1535 | } | |
1536 | ||
9d98354f | 1537 | void *unpack_entry(struct repository *r, struct packed_git *p, off_t obj_offset, |
f1d8130b JT |
1538 | enum object_type *final_type, unsigned long *final_size) |
1539 | { | |
1540 | struct pack_window *w_curs = NULL; | |
1541 | off_t curpos = obj_offset; | |
1542 | void *data = NULL; | |
1543 | unsigned long size; | |
1544 | enum object_type type; | |
1545 | struct unpack_entry_stack_ent small_delta_stack[UNPACK_ENTRY_STACK_PREALLOC]; | |
1546 | struct unpack_entry_stack_ent *delta_stack = small_delta_stack; | |
1547 | int delta_stack_nr = 0, delta_stack_alloc = UNPACK_ENTRY_STACK_PREALLOC; | |
1548 | int base_from_cache = 0; | |
1549 | ||
1550 | write_pack_access_log(p, obj_offset); | |
1551 | ||
1552 | /* PHASE 1: drill down to the innermost base object */ | |
1553 | for (;;) { | |
1554 | off_t base_offset; | |
1555 | int i; | |
1556 | struct delta_base_cache_entry *ent; | |
1557 | ||
1558 | ent = get_delta_base_cache_entry(p, curpos); | |
1559 | if (ent) { | |
1560 | type = ent->type; | |
1561 | data = ent->data; | |
1562 | size = ent->size; | |
1563 | detach_delta_base_cache_entry(ent); | |
1564 | base_from_cache = 1; | |
1565 | break; | |
1566 | } | |
1567 | ||
1568 | if (do_check_packed_object_crc && p->index_version > 1) { | |
1569 | struct revindex_entry *revidx = find_pack_revindex(p, obj_offset); | |
1570 | off_t len = revidx[1].offset - obj_offset; | |
1571 | if (check_pack_crc(p, &w_curs, obj_offset, len, revidx->nr)) { | |
4310b0c4 | 1572 | struct object_id oid; |
1573 | nth_packed_object_oid(&oid, p, revidx->nr); | |
f1d8130b | 1574 | error("bad packed object CRC for %s", |
4310b0c4 | 1575 | oid_to_hex(&oid)); |
1576 | mark_bad_packed_object(p, oid.hash); | |
f1d8130b JT |
1577 | data = NULL; |
1578 | goto out; | |
1579 | } | |
1580 | } | |
1581 | ||
1582 | type = unpack_object_header(p, &w_curs, &curpos, &size); | |
1583 | if (type != OBJ_OFS_DELTA && type != OBJ_REF_DELTA) | |
1584 | break; | |
1585 | ||
1586 | base_offset = get_delta_base(p, &w_curs, &curpos, type, obj_offset); | |
1587 | if (!base_offset) { | |
1588 | error("failed to validate delta base reference " | |
1589 | "at offset %"PRIuMAX" from %s", | |
1590 | (uintmax_t)curpos, p->pack_name); | |
1591 | /* bail to phase 2, in hopes of recovery */ | |
1592 | data = NULL; | |
1593 | break; | |
1594 | } | |
1595 | ||
1596 | /* push object, proceed to base */ | |
1597 | if (delta_stack_nr >= delta_stack_alloc | |
1598 | && delta_stack == small_delta_stack) { | |
1599 | delta_stack_alloc = alloc_nr(delta_stack_nr); | |
1600 | ALLOC_ARRAY(delta_stack, delta_stack_alloc); | |
1601 | memcpy(delta_stack, small_delta_stack, | |
1602 | sizeof(*delta_stack)*delta_stack_nr); | |
1603 | } else { | |
1604 | ALLOC_GROW(delta_stack, delta_stack_nr+1, delta_stack_alloc); | |
1605 | } | |
1606 | i = delta_stack_nr++; | |
1607 | delta_stack[i].obj_offset = obj_offset; | |
1608 | delta_stack[i].curpos = curpos; | |
1609 | delta_stack[i].size = size; | |
1610 | ||
1611 | curpos = obj_offset = base_offset; | |
1612 | } | |
1613 | ||
1614 | /* PHASE 2: handle the base */ | |
1615 | switch (type) { | |
1616 | case OBJ_OFS_DELTA: | |
1617 | case OBJ_REF_DELTA: | |
1618 | if (data) | |
033abf97 | 1619 | BUG("unpack_entry: left loop at a valid delta"); |
f1d8130b JT |
1620 | break; |
1621 | case OBJ_COMMIT: | |
1622 | case OBJ_TREE: | |
1623 | case OBJ_BLOB: | |
1624 | case OBJ_TAG: | |
1625 | if (!base_from_cache) | |
1626 | data = unpack_compressed_entry(p, &w_curs, curpos, size); | |
1627 | break; | |
1628 | default: | |
1629 | data = NULL; | |
1630 | error("unknown object type %i at offset %"PRIuMAX" in %s", | |
1631 | type, (uintmax_t)obj_offset, p->pack_name); | |
1632 | } | |
1633 | ||
1634 | /* PHASE 3: apply deltas in order */ | |
1635 | ||
1636 | /* invariants: | |
1637 | * 'data' holds the base data, or NULL if there was corruption | |
1638 | */ | |
1639 | while (delta_stack_nr) { | |
1640 | void *delta_data; | |
1641 | void *base = data; | |
1642 | void *external_base = NULL; | |
1643 | unsigned long delta_size, base_size = size; | |
1644 | int i; | |
1645 | ||
1646 | data = NULL; | |
1647 | ||
1648 | if (base) | |
1649 | add_delta_base_cache(p, obj_offset, base, base_size, type); | |
1650 | ||
1651 | if (!base) { | |
1652 | /* | |
1653 | * We're probably in deep shit, but let's try to fetch | |
1654 | * the required base anyway from another pack or loose. | |
1655 | * This is costly but should happen only in the presence | |
1656 | * of a corrupted pack, and is better than failing outright. | |
1657 | */ | |
1658 | struct revindex_entry *revidx; | |
4310b0c4 | 1659 | struct object_id base_oid; |
f1d8130b JT |
1660 | revidx = find_pack_revindex(p, obj_offset); |
1661 | if (revidx) { | |
4310b0c4 | 1662 | nth_packed_object_oid(&base_oid, p, revidx->nr); |
f1d8130b JT |
1663 | error("failed to read delta base object %s" |
1664 | " at offset %"PRIuMAX" from %s", | |
4310b0c4 | 1665 | oid_to_hex(&base_oid), (uintmax_t)obj_offset, |
f1d8130b | 1666 | p->pack_name); |
4310b0c4 | 1667 | mark_bad_packed_object(p, base_oid.hash); |
9d98354f | 1668 | base = read_object(r, &base_oid, &type, &base_size); |
f1d8130b JT |
1669 | external_base = base; |
1670 | } | |
1671 | } | |
1672 | ||
1673 | i = --delta_stack_nr; | |
1674 | obj_offset = delta_stack[i].obj_offset; | |
1675 | curpos = delta_stack[i].curpos; | |
1676 | delta_size = delta_stack[i].size; | |
1677 | ||
1678 | if (!base) | |
1679 | continue; | |
1680 | ||
1681 | delta_data = unpack_compressed_entry(p, &w_curs, curpos, delta_size); | |
1682 | ||
1683 | if (!delta_data) { | |
1684 | error("failed to unpack compressed delta " | |
1685 | "at offset %"PRIuMAX" from %s", | |
1686 | (uintmax_t)curpos, p->pack_name); | |
1687 | data = NULL; | |
1688 | free(external_base); | |
1689 | continue; | |
1690 | } | |
1691 | ||
1692 | data = patch_delta(base, base_size, | |
1693 | delta_data, delta_size, | |
1694 | &size); | |
1695 | ||
1696 | /* | |
1697 | * We could not apply the delta; warn the user, but keep going. | |
1698 | * Our failure will be noticed either in the next iteration of | |
1699 | * the loop, or if this is the final delta, in the caller when | |
1700 | * we return NULL. Those code paths will take care of making | |
1701 | * a more explicit warning and retrying with another copy of | |
1702 | * the object. | |
1703 | */ | |
1704 | if (!data) | |
1705 | error("failed to apply delta"); | |
1706 | ||
1707 | free(delta_data); | |
1708 | free(external_base); | |
1709 | } | |
1710 | ||
1711 | if (final_type) | |
1712 | *final_type = type; | |
1713 | if (final_size) | |
1714 | *final_size = size; | |
1715 | ||
1716 | out: | |
1717 | unuse_pack(&w_curs); | |
1718 | ||
1719 | if (delta_stack != small_delta_stack) | |
1720 | free(delta_stack); | |
1721 | ||
1722 | return data; | |
1723 | } | |
d5a16761 | 1724 | |
3d475f46 DS |
1725 | int bsearch_pack(const struct object_id *oid, const struct packed_git *p, uint32_t *result) |
1726 | { | |
1727 | const unsigned char *index_fanout = p->index_data; | |
1728 | const unsigned char *index_lookup; | |
37fec86a | 1729 | const unsigned int hashsz = the_hash_algo->rawsz; |
3d475f46 DS |
1730 | int index_lookup_width; |
1731 | ||
1732 | if (!index_fanout) | |
1733 | BUG("bsearch_pack called without a valid pack-index"); | |
1734 | ||
1735 | index_lookup = index_fanout + 4 * 256; | |
1736 | if (p->index_version == 1) { | |
37fec86a | 1737 | index_lookup_width = hashsz + 4; |
3d475f46 DS |
1738 | index_lookup += 4; |
1739 | } else { | |
37fec86a | 1740 | index_lookup_width = hashsz; |
3d475f46 DS |
1741 | index_fanout += 8; |
1742 | index_lookup += 8; | |
1743 | } | |
1744 | ||
1745 | return bsearch_hash(oid->hash, (const uint32_t*)index_fanout, | |
1746 | index_lookup, index_lookup_width, result); | |
1747 | } | |
1748 | ||
d5a16761 JT |
1749 | const unsigned char *nth_packed_object_sha1(struct packed_git *p, |
1750 | uint32_t n) | |
1751 | { | |
1752 | const unsigned char *index = p->index_data; | |
37fec86a | 1753 | const unsigned int hashsz = the_hash_algo->rawsz; |
d5a16761 JT |
1754 | if (!index) { |
1755 | if (open_pack_index(p)) | |
1756 | return NULL; | |
1757 | index = p->index_data; | |
1758 | } | |
1759 | if (n >= p->num_objects) | |
1760 | return NULL; | |
1761 | index += 4 * 256; | |
1762 | if (p->index_version == 1) { | |
37fec86a | 1763 | return index + (hashsz + 4) * n + 4; |
d5a16761 JT |
1764 | } else { |
1765 | index += 8; | |
37fec86a | 1766 | return index + hashsz * n; |
d5a16761 JT |
1767 | } |
1768 | } | |
1769 | ||
1770 | const struct object_id *nth_packed_object_oid(struct object_id *oid, | |
1771 | struct packed_git *p, | |
1772 | uint32_t n) | |
1773 | { | |
1774 | const unsigned char *hash = nth_packed_object_sha1(p, n); | |
1775 | if (!hash) | |
1776 | return NULL; | |
1777 | hashcpy(oid->hash, hash); | |
1778 | return oid; | |
1779 | } | |
9e0f45f5 JT |
1780 | |
1781 | void check_pack_index_ptr(const struct packed_git *p, const void *vptr) | |
1782 | { | |
1783 | const unsigned char *ptr = vptr; | |
1784 | const unsigned char *start = p->index_data; | |
1785 | const unsigned char *end = start + p->index_size; | |
1786 | if (ptr < start) | |
1787 | die(_("offset before start of pack index for %s (corrupt index?)"), | |
1788 | p->pack_name); | |
1789 | /* No need to check for underflow; .idx files must be at least 8 bytes */ | |
1790 | if (ptr >= end - 8) | |
1791 | die(_("offset beyond end of pack index for %s (truncated index?)"), | |
1792 | p->pack_name); | |
1793 | } | |
1794 | ||
1795 | off_t nth_packed_object_offset(const struct packed_git *p, uint32_t n) | |
1796 | { | |
1797 | const unsigned char *index = p->index_data; | |
37fec86a | 1798 | const unsigned int hashsz = the_hash_algo->rawsz; |
9e0f45f5 JT |
1799 | index += 4 * 256; |
1800 | if (p->index_version == 1) { | |
37fec86a | 1801 | return ntohl(*((uint32_t *)(index + (hashsz + 4) * n))); |
9e0f45f5 JT |
1802 | } else { |
1803 | uint32_t off; | |
37fec86a | 1804 | index += 8 + p->num_objects * (hashsz + 4); |
9e0f45f5 JT |
1805 | off = ntohl(*((uint32_t *)(index + 4 * n))); |
1806 | if (!(off & 0x80000000)) | |
1807 | return off; | |
1808 | index += p->num_objects * 4 + (off & 0x7fffffff) * 8; | |
1809 | check_pack_index_ptr(p, index); | |
ad622a25 | 1810 | return get_be64(index); |
9e0f45f5 JT |
1811 | } |
1812 | } | |
a2551953 JT |
1813 | |
1814 | off_t find_pack_entry_one(const unsigned char *sha1, | |
1815 | struct packed_git *p) | |
1816 | { | |
a2551953 | 1817 | const unsigned char *index = p->index_data; |
3d475f46 | 1818 | struct object_id oid; |
b4e00f73 | 1819 | uint32_t result; |
a2551953 JT |
1820 | |
1821 | if (!index) { | |
1822 | if (open_pack_index(p)) | |
1823 | return 0; | |
a2551953 JT |
1824 | } |
1825 | ||
3d475f46 DS |
1826 | hashcpy(oid.hash, sha1); |
1827 | if (bsearch_pack(&oid, p, &result)) | |
b4e00f73 | 1828 | return nth_packed_object_offset(p, result); |
a2551953 JT |
1829 | return 0; |
1830 | } | |
1831 | ||
1832 | int is_pack_valid(struct packed_git *p) | |
1833 | { | |
1834 | /* An already open pack is known to be valid. */ | |
1835 | if (p->pack_fd != -1) | |
1836 | return 1; | |
1837 | ||
1838 | /* If the pack has one window completely covering the | |
1839 | * file size, the pack is known to be valid even if | |
1840 | * the descriptor is not currently open. | |
1841 | */ | |
1842 | if (p->windows) { | |
1843 | struct pack_window *w = p->windows; | |
1844 | ||
1845 | if (!w->offset && w->len == p->pack_size) | |
1846 | return 1; | |
1847 | } | |
1848 | ||
1849 | /* Force the pack to open to prove its valid. */ | |
1850 | return !open_packed_git(p); | |
1851 | } | |
d6fe0036 JT |
1852 | |
1853 | struct packed_git *find_sha1_pack(const unsigned char *sha1, | |
1854 | struct packed_git *packs) | |
1855 | { | |
1856 | struct packed_git *p; | |
1857 | ||
1858 | for (p = packs; p; p = p->next) { | |
1859 | if (find_pack_entry_one(sha1, p)) | |
1860 | return p; | |
1861 | } | |
1862 | return NULL; | |
1863 | ||
1864 | } | |
1a1e5d4f | 1865 | |
544443cb | 1866 | static int fill_pack_entry(const struct object_id *oid, |
1a1e5d4f JT |
1867 | struct pack_entry *e, |
1868 | struct packed_git *p) | |
1869 | { | |
1870 | off_t offset; | |
1871 | ||
1872 | if (p->num_bad_objects) { | |
1873 | unsigned i; | |
1874 | for (i = 0; i < p->num_bad_objects; i++) | |
37fec86a | 1875 | if (!hashcmp(oid->hash, |
1876 | p->bad_object_sha1 + the_hash_algo->rawsz * i)) | |
1a1e5d4f JT |
1877 | return 0; |
1878 | } | |
1879 | ||
544443cb | 1880 | offset = find_pack_entry_one(oid->hash, p); |
1a1e5d4f JT |
1881 | if (!offset) |
1882 | return 0; | |
1883 | ||
1884 | /* | |
1885 | * We are about to tell the caller where they can locate the | |
1886 | * requested object. We better make sure the packfile is | |
1887 | * still here and can be accessed before supplying that | |
1888 | * answer, as it may have been deleted since the index was | |
1889 | * loaded! | |
1890 | */ | |
1891 | if (!is_pack_valid(p)) | |
1892 | return 0; | |
1893 | e->offset = offset; | |
1894 | e->p = p; | |
1a1e5d4f JT |
1895 | return 1; |
1896 | } | |
1897 | ||
544443cb | 1898 | int find_pack_entry(struct repository *r, const struct object_id *oid, struct pack_entry *e) |
1a1e5d4f | 1899 | { |
8865859d | 1900 | struct list_head *pos; |
1a1e5d4f | 1901 | |
0a0dd632 SB |
1902 | prepare_packed_git(r); |
1903 | if (!r->objects->packed_git) | |
1a1e5d4f JT |
1904 | return 0; |
1905 | ||
0a0dd632 | 1906 | list_for_each(pos, &r->objects->packed_git_mru) { |
ec2dd32c | 1907 | struct packed_git *p = list_entry(pos, struct packed_git, mru); |
544443cb | 1908 | if (fill_pack_entry(oid, e, p)) { |
0a0dd632 | 1909 | list_move(&p->mru, &r->objects->packed_git_mru); |
1a1e5d4f JT |
1910 | return 1; |
1911 | } | |
1912 | } | |
1913 | return 0; | |
1914 | } | |
150e3001 | 1915 | |
14c3c80c | 1916 | int has_object_pack(const struct object_id *oid) |
150e3001 JT |
1917 | { |
1918 | struct pack_entry e; | |
544443cb | 1919 | return find_pack_entry(the_repository, oid, &e); |
150e3001 | 1920 | } |
f9a8672a JT |
1921 | |
1922 | int has_pack_index(const unsigned char *sha1) | |
1923 | { | |
1924 | struct stat st; | |
1925 | if (stat(sha1_pack_index_name(sha1), &st)) | |
1926 | return 0; | |
1927 | return 1; | |
1928 | } | |
7709f468 | 1929 | |
049d51a2 | 1930 | int for_each_object_in_pack(struct packed_git *p, each_packed_object_fn cb, void *data) |
7709f468 JT |
1931 | { |
1932 | uint32_t i; | |
1933 | int r = 0; | |
1934 | ||
1935 | for (i = 0; i < p->num_objects; i++) { | |
1936 | struct object_id oid; | |
1937 | ||
1938 | if (!nth_packed_object_oid(&oid, p, i)) | |
1939 | return error("unable to get sha1 of object %u in %s", | |
1940 | i, p->pack_name); | |
1941 | ||
1942 | r = cb(&oid, p, i, data); | |
1943 | if (r) | |
1944 | break; | |
1945 | } | |
1946 | return r; | |
1947 | } | |
1948 | ||
1949 | int for_each_packed_object(each_packed_object_fn cb, void *data, unsigned flags) | |
1950 | { | |
1951 | struct packed_git *p; | |
1952 | int r = 0; | |
1953 | int pack_errors = 0; | |
1954 | ||
6fdb4e9f | 1955 | prepare_packed_git(the_repository); |
a80d72db | 1956 | for (p = the_repository->objects->packed_git; p; p = p->next) { |
7709f468 JT |
1957 | if ((flags & FOR_EACH_OBJECT_LOCAL_ONLY) && !p->pack_local) |
1958 | continue; | |
498f1f61 JT |
1959 | if ((flags & FOR_EACH_OBJECT_PROMISOR_ONLY) && |
1960 | !p->pack_promisor) | |
1961 | continue; | |
7709f468 JT |
1962 | if (open_pack_index(p)) { |
1963 | pack_errors = 1; | |
1964 | continue; | |
1965 | } | |
1966 | r = for_each_object_in_pack(p, cb, data); | |
1967 | if (r) | |
1968 | break; | |
1969 | } | |
1970 | return r ? r : pack_errors; | |
1971 | } | |
498f1f61 JT |
1972 | |
1973 | static int add_promisor_object(const struct object_id *oid, | |
1974 | struct packed_git *pack, | |
1975 | uint32_t pos, | |
1976 | void *set_) | |
1977 | { | |
1978 | struct oidset *set = set_; | |
1979 | struct object *obj = parse_object(oid); | |
1980 | if (!obj) | |
1981 | return 1; | |
1982 | ||
1983 | oidset_insert(set, oid); | |
1984 | ||
1985 | /* | |
1986 | * If this is a tree, commit, or tag, the objects it refers | |
a80d72db | 1987 | * to are also promisor objects. (Blobs refer to no objects->) |
498f1f61 JT |
1988 | */ |
1989 | if (obj->type == OBJ_TREE) { | |
1990 | struct tree *tree = (struct tree *)obj; | |
1991 | struct tree_desc desc; | |
1992 | struct name_entry entry; | |
1993 | if (init_tree_desc_gently(&desc, tree->buffer, tree->size)) | |
1994 | /* | |
1995 | * Error messages are given when packs are | |
1996 | * verified, so do not print any here. | |
1997 | */ | |
1998 | return 0; | |
1999 | while (tree_entry_gently(&desc, &entry)) | |
2000 | oidset_insert(set, entry.oid); | |
2001 | } else if (obj->type == OBJ_COMMIT) { | |
2002 | struct commit *commit = (struct commit *) obj; | |
2003 | struct commit_list *parents = commit->parents; | |
2004 | ||
2e27bd77 | 2005 | oidset_insert(set, get_commit_tree_oid(commit)); |
498f1f61 JT |
2006 | for (; parents; parents = parents->next) |
2007 | oidset_insert(set, &parents->item->object.oid); | |
2008 | } else if (obj->type == OBJ_TAG) { | |
2009 | struct tag *tag = (struct tag *) obj; | |
2010 | oidset_insert(set, &tag->tagged->oid); | |
2011 | } | |
2012 | return 0; | |
2013 | } | |
2014 | ||
2015 | int is_promisor_object(const struct object_id *oid) | |
2016 | { | |
2017 | static struct oidset promisor_objects; | |
2018 | static int promisor_objects_prepared; | |
2019 | ||
2020 | if (!promisor_objects_prepared) { | |
2021 | if (repository_format_partial_clone) { | |
2022 | for_each_packed_object(add_promisor_object, | |
2023 | &promisor_objects, | |
2024 | FOR_EACH_OBJECT_PROMISOR_ONLY); | |
2025 | } | |
2026 | promisor_objects_prepared = 1; | |
2027 | } | |
2028 | return oidset_contains(&promisor_objects, oid); | |
2029 | } |