]>
Commit | Line | Data |
---|---|---|
355885d5 MK |
1 | #include "cache.h" |
2 | #include "object.h" | |
3 | #include "blob.h" | |
4 | #include "tree.h" | |
5 | #include "tree-walk.h" | |
6 | #include "commit.h" | |
7 | #include "tag.h" | |
8 | #include "fsck.h" | |
cec097be | 9 | #include "refs.h" |
a18fcc9f | 10 | #include "utf8.h" |
cd94c6f9 | 11 | #include "sha1-array.h" |
355885d5 | 12 | |
f50c4407 | 13 | #define FSCK_FATAL -1 |
f27d05b1 | 14 | #define FSCK_INFO -2 |
f50c4407 | 15 | |
c99ba492 | 16 | #define FOREACH_MSG_ID(FUNC) \ |
f50c4407 JS |
17 | /* fatal errors */ \ |
18 | FUNC(NUL_IN_HEADER, FATAL) \ | |
19 | FUNC(UNTERMINATED_HEADER, FATAL) \ | |
c99ba492 JS |
20 | /* errors */ \ |
21 | FUNC(BAD_DATE, ERROR) \ | |
22 | FUNC(BAD_DATE_OVERFLOW, ERROR) \ | |
23 | FUNC(BAD_EMAIL, ERROR) \ | |
24 | FUNC(BAD_NAME, ERROR) \ | |
25 | FUNC(BAD_OBJECT_SHA1, ERROR) \ | |
26 | FUNC(BAD_PARENT_SHA1, ERROR) \ | |
27 | FUNC(BAD_TAG_OBJECT, ERROR) \ | |
28 | FUNC(BAD_TIMEZONE, ERROR) \ | |
29 | FUNC(BAD_TREE, ERROR) \ | |
30 | FUNC(BAD_TREE_SHA1, ERROR) \ | |
31 | FUNC(BAD_TYPE, ERROR) \ | |
32 | FUNC(DUPLICATE_ENTRIES, ERROR) \ | |
33 | FUNC(MISSING_AUTHOR, ERROR) \ | |
34 | FUNC(MISSING_COMMITTER, ERROR) \ | |
35 | FUNC(MISSING_EMAIL, ERROR) \ | |
36 | FUNC(MISSING_GRAFT, ERROR) \ | |
37 | FUNC(MISSING_NAME_BEFORE_EMAIL, ERROR) \ | |
38 | FUNC(MISSING_OBJECT, ERROR) \ | |
39 | FUNC(MISSING_PARENT, ERROR) \ | |
40 | FUNC(MISSING_SPACE_BEFORE_DATE, ERROR) \ | |
41 | FUNC(MISSING_SPACE_BEFORE_EMAIL, ERROR) \ | |
42 | FUNC(MISSING_TAG, ERROR) \ | |
43 | FUNC(MISSING_TAG_ENTRY, ERROR) \ | |
44 | FUNC(MISSING_TAG_OBJECT, ERROR) \ | |
45 | FUNC(MISSING_TREE, ERROR) \ | |
46 | FUNC(MISSING_TYPE, ERROR) \ | |
47 | FUNC(MISSING_TYPE_ENTRY, ERROR) \ | |
c9ad147f | 48 | FUNC(MULTIPLE_AUTHORS, ERROR) \ |
c99ba492 JS |
49 | FUNC(TAG_OBJECT_NOT_TAG, ERROR) \ |
50 | FUNC(TREE_NOT_SORTED, ERROR) \ | |
51 | FUNC(UNKNOWN_TYPE, ERROR) \ | |
c99ba492 JS |
52 | FUNC(ZERO_PADDED_DATE, ERROR) \ |
53 | /* warnings */ \ | |
54 | FUNC(BAD_FILEMODE, WARN) \ | |
c99ba492 JS |
55 | FUNC(EMPTY_NAME, WARN) \ |
56 | FUNC(FULL_PATHNAME, WARN) \ | |
57 | FUNC(HAS_DOT, WARN) \ | |
58 | FUNC(HAS_DOTDOT, WARN) \ | |
59 | FUNC(HAS_DOTGIT, WARN) \ | |
c99ba492 | 60 | FUNC(NULL_SHA1, WARN) \ |
f27d05b1 JS |
61 | FUNC(ZERO_PADDED_FILEMODE, WARN) \ |
62 | /* infos (reported as warnings, but ignored by default) */ \ | |
63 | FUNC(BAD_TAG_NAME, INFO) \ | |
64 | FUNC(MISSING_TAGGER_ENTRY, INFO) | |
c99ba492 JS |
65 | |
66 | #define MSG_ID(id, msg_type) FSCK_MSG_##id, | |
67 | enum fsck_msg_id { | |
68 | FOREACH_MSG_ID(MSG_ID) | |
69 | FSCK_MSG_MAX | |
70 | }; | |
71 | #undef MSG_ID | |
72 | ||
f417eed8 JS |
73 | #define STR(x) #x |
74 | #define MSG_ID(id, msg_type) { STR(id), NULL, FSCK_##msg_type }, | |
c99ba492 | 75 | static struct { |
f417eed8 JS |
76 | const char *id_string; |
77 | const char *downcased; | |
c99ba492 JS |
78 | int msg_type; |
79 | } msg_id_info[FSCK_MSG_MAX + 1] = { | |
80 | FOREACH_MSG_ID(MSG_ID) | |
f417eed8 | 81 | { NULL, NULL, -1 } |
c99ba492 JS |
82 | }; |
83 | #undef MSG_ID | |
84 | ||
f417eed8 JS |
85 | static int parse_msg_id(const char *text) |
86 | { | |
87 | int i; | |
88 | ||
89 | if (!msg_id_info[0].downcased) { | |
90 | /* convert id_string to lower case, without underscores. */ | |
91 | for (i = 0; i < FSCK_MSG_MAX; i++) { | |
92 | const char *p = msg_id_info[i].id_string; | |
93 | int len = strlen(p); | |
94 | char *q = xmalloc(len); | |
95 | ||
96 | msg_id_info[i].downcased = q; | |
97 | while (*p) | |
98 | if (*p == '_') | |
99 | p++; | |
100 | else | |
101 | *(q)++ = tolower(*(p)++); | |
102 | *q = '\0'; | |
103 | } | |
104 | } | |
105 | ||
106 | for (i = 0; i < FSCK_MSG_MAX; i++) | |
107 | if (!strcmp(text, msg_id_info[i].downcased)) | |
108 | return i; | |
109 | ||
110 | return -1; | |
111 | } | |
112 | ||
c99ba492 JS |
113 | static int fsck_msg_type(enum fsck_msg_id msg_id, |
114 | struct fsck_options *options) | |
115 | { | |
116 | int msg_type; | |
117 | ||
0282f4dc JS |
118 | assert(msg_id >= 0 && msg_id < FSCK_MSG_MAX); |
119 | ||
120 | if (options->msg_type) | |
121 | msg_type = options->msg_type[msg_id]; | |
122 | else { | |
123 | msg_type = msg_id_info[msg_id].msg_type; | |
124 | if (options->strict && msg_type == FSCK_WARN) | |
125 | msg_type = FSCK_ERROR; | |
126 | } | |
c99ba492 JS |
127 | |
128 | return msg_type; | |
129 | } | |
130 | ||
cd94c6f9 JS |
131 | static void init_skiplist(struct fsck_options *options, const char *path) |
132 | { | |
133 | static struct sha1_array skiplist = SHA1_ARRAY_INIT; | |
134 | int sorted, fd; | |
135 | char buffer[41]; | |
136 | unsigned char sha1[20]; | |
137 | ||
138 | if (options->skiplist) | |
139 | sorted = options->skiplist->sorted; | |
140 | else { | |
141 | sorted = 1; | |
142 | options->skiplist = &skiplist; | |
143 | } | |
144 | ||
145 | fd = open(path, O_RDONLY); | |
146 | if (fd < 0) | |
147 | die("Could not open skip list: %s", path); | |
148 | for (;;) { | |
149 | int result = read_in_full(fd, buffer, sizeof(buffer)); | |
150 | if (result < 0) | |
151 | die_errno("Could not read '%s'", path); | |
152 | if (!result) | |
153 | break; | |
154 | if (get_sha1_hex(buffer, sha1) || buffer[40] != '\n') | |
155 | die("Invalid SHA-1: %s", buffer); | |
156 | sha1_array_append(&skiplist, sha1); | |
157 | if (sorted && skiplist.nr > 1 && | |
158 | hashcmp(skiplist.sha1[skiplist.nr - 2], | |
159 | sha1) > 0) | |
160 | sorted = 0; | |
161 | } | |
162 | close(fd); | |
163 | ||
164 | if (sorted) | |
165 | skiplist.sorted = 1; | |
166 | } | |
167 | ||
0282f4dc JS |
168 | static int parse_msg_type(const char *str) |
169 | { | |
170 | if (!strcmp(str, "error")) | |
171 | return FSCK_ERROR; | |
172 | else if (!strcmp(str, "warn")) | |
173 | return FSCK_WARN; | |
efaba7cc JS |
174 | else if (!strcmp(str, "ignore")) |
175 | return FSCK_IGNORE; | |
0282f4dc JS |
176 | else |
177 | die("Unknown fsck message type: '%s'", str); | |
178 | } | |
179 | ||
5d477a33 JS |
180 | int is_valid_msg_type(const char *msg_id, const char *msg_type) |
181 | { | |
182 | if (parse_msg_id(msg_id) < 0) | |
183 | return 0; | |
184 | parse_msg_type(msg_type); | |
185 | return 1; | |
186 | } | |
187 | ||
0282f4dc JS |
188 | void fsck_set_msg_type(struct fsck_options *options, |
189 | const char *msg_id, const char *msg_type) | |
190 | { | |
191 | int id = parse_msg_id(msg_id), type; | |
192 | ||
193 | if (id < 0) | |
194 | die("Unhandled message id: %s", msg_id); | |
195 | type = parse_msg_type(msg_type); | |
196 | ||
f50c4407 JS |
197 | if (type != FSCK_ERROR && msg_id_info[id].msg_type == FSCK_FATAL) |
198 | die("Cannot demote %s to %s", msg_id, msg_type); | |
199 | ||
0282f4dc JS |
200 | if (!options->msg_type) { |
201 | int i; | |
b32fa95f JK |
202 | int *msg_type; |
203 | ALLOC_ARRAY(msg_type, FSCK_MSG_MAX); | |
0282f4dc JS |
204 | for (i = 0; i < FSCK_MSG_MAX; i++) |
205 | msg_type[i] = fsck_msg_type(i, options); | |
206 | options->msg_type = msg_type; | |
207 | } | |
208 | ||
209 | options->msg_type[id] = type; | |
210 | } | |
211 | ||
212 | void fsck_set_msg_types(struct fsck_options *options, const char *values) | |
213 | { | |
214 | char *buf = xstrdup(values), *to_free = buf; | |
215 | int done = 0; | |
216 | ||
217 | while (!done) { | |
218 | int len = strcspn(buf, " ,|"), equal; | |
219 | ||
220 | done = !buf[len]; | |
221 | if (!len) { | |
222 | buf++; | |
223 | continue; | |
224 | } | |
225 | buf[len] = '\0'; | |
226 | ||
227 | for (equal = 0; | |
228 | equal < len && buf[equal] != '=' && buf[equal] != ':'; | |
229 | equal++) | |
230 | buf[equal] = tolower(buf[equal]); | |
231 | buf[equal] = '\0'; | |
232 | ||
cd94c6f9 JS |
233 | if (!strcmp(buf, "skiplist")) { |
234 | if (equal == len) | |
235 | die("skiplist requires a path"); | |
236 | init_skiplist(options, buf + equal + 1); | |
237 | buf += len + 1; | |
238 | continue; | |
239 | } | |
240 | ||
0282f4dc JS |
241 | if (equal == len) |
242 | die("Missing '=': '%s'", buf); | |
355885d5 | 243 | |
0282f4dc JS |
244 | fsck_set_msg_type(options, buf, buf + equal + 1); |
245 | buf += len + 1; | |
246 | } | |
247 | free(to_free); | |
248 | } | |
249 | ||
71ab8fa8 JS |
250 | static void append_msg_id(struct strbuf *sb, const char *msg_id) |
251 | { | |
252 | for (;;) { | |
253 | char c = *(msg_id)++; | |
254 | ||
255 | if (!c) | |
256 | break; | |
257 | if (c != '_') | |
258 | strbuf_addch(sb, tolower(c)); | |
259 | else { | |
260 | assert(*msg_id); | |
261 | strbuf_addch(sb, *(msg_id)++); | |
262 | } | |
263 | } | |
264 | ||
265 | strbuf_addstr(sb, ": "); | |
266 | } | |
267 | ||
c99ba492 JS |
268 | __attribute__((format (printf, 4, 5))) |
269 | static int report(struct fsck_options *options, struct object *object, | |
270 | enum fsck_msg_id id, const char *fmt, ...) | |
271 | { | |
272 | va_list ap; | |
273 | struct strbuf sb = STRBUF_INIT; | |
274 | int msg_type = fsck_msg_type(id, options), result; | |
275 | ||
efaba7cc JS |
276 | if (msg_type == FSCK_IGNORE) |
277 | return 0; | |
278 | ||
cd94c6f9 | 279 | if (options->skiplist && object && |
ed1c9977 | 280 | sha1_array_lookup(options->skiplist, object->oid.hash) >= 0) |
cd94c6f9 JS |
281 | return 0; |
282 | ||
f50c4407 JS |
283 | if (msg_type == FSCK_FATAL) |
284 | msg_type = FSCK_ERROR; | |
f27d05b1 JS |
285 | else if (msg_type == FSCK_INFO) |
286 | msg_type = FSCK_WARN; | |
f50c4407 | 287 | |
71ab8fa8 JS |
288 | append_msg_id(&sb, msg_id_info[id].id_string); |
289 | ||
c99ba492 JS |
290 | va_start(ap, fmt); |
291 | strbuf_vaddf(&sb, fmt, ap); | |
292 | result = options->error_func(object, msg_type, sb.buf); | |
293 | strbuf_release(&sb); | |
294 | va_end(ap); | |
295 | ||
296 | return result; | |
297 | } | |
298 | ||
22410549 | 299 | static int fsck_walk_tree(struct tree *tree, void *data, struct fsck_options *options) |
355885d5 MK |
300 | { |
301 | struct tree_desc desc; | |
302 | struct name_entry entry; | |
303 | int res = 0; | |
304 | ||
305 | if (parse_tree(tree)) | |
306 | return -1; | |
307 | ||
308 | init_tree_desc(&desc, tree->buffer, tree->size); | |
309 | while (tree_entry(&desc, &entry)) { | |
310 | int result; | |
311 | ||
312 | if (S_ISGITLINK(entry.mode)) | |
313 | continue; | |
314 | if (S_ISDIR(entry.mode)) | |
22410549 | 315 | result = options->walk(&lookup_tree(entry.sha1)->object, OBJ_TREE, data, options); |
355885d5 | 316 | else if (S_ISREG(entry.mode) || S_ISLNK(entry.mode)) |
22410549 | 317 | result = options->walk(&lookup_blob(entry.sha1)->object, OBJ_BLOB, data, options); |
355885d5 | 318 | else { |
82247e9b | 319 | result = error("in tree %s: entry %s has bad mode %.6o", |
f2fd0760 | 320 | oid_to_hex(&tree->object.oid), entry.path, entry.mode); |
355885d5 MK |
321 | } |
322 | if (result < 0) | |
323 | return result; | |
324 | if (!res) | |
325 | res = result; | |
326 | } | |
327 | return res; | |
328 | } | |
329 | ||
22410549 | 330 | static int fsck_walk_commit(struct commit *commit, void *data, struct fsck_options *options) |
355885d5 MK |
331 | { |
332 | struct commit_list *parents; | |
333 | int res; | |
334 | int result; | |
335 | ||
336 | if (parse_commit(commit)) | |
337 | return -1; | |
338 | ||
22410549 | 339 | result = options->walk((struct object *)commit->tree, OBJ_TREE, data, options); |
355885d5 MK |
340 | if (result < 0) |
341 | return result; | |
342 | res = result; | |
343 | ||
344 | parents = commit->parents; | |
345 | while (parents) { | |
22410549 | 346 | result = options->walk((struct object *)parents->item, OBJ_COMMIT, data, options); |
355885d5 MK |
347 | if (result < 0) |
348 | return result; | |
349 | if (!res) | |
350 | res = result; | |
351 | parents = parents->next; | |
352 | } | |
353 | return res; | |
354 | } | |
355 | ||
22410549 | 356 | static int fsck_walk_tag(struct tag *tag, void *data, struct fsck_options *options) |
355885d5 MK |
357 | { |
358 | if (parse_tag(tag)) | |
359 | return -1; | |
22410549 | 360 | return options->walk(tag->tagged, OBJ_ANY, data, options); |
355885d5 MK |
361 | } |
362 | ||
22410549 | 363 | int fsck_walk(struct object *obj, void *data, struct fsck_options *options) |
355885d5 MK |
364 | { |
365 | if (!obj) | |
366 | return -1; | |
367 | switch (obj->type) { | |
368 | case OBJ_BLOB: | |
369 | return 0; | |
370 | case OBJ_TREE: | |
22410549 | 371 | return fsck_walk_tree((struct tree *)obj, data, options); |
355885d5 | 372 | case OBJ_COMMIT: |
22410549 | 373 | return fsck_walk_commit((struct commit *)obj, data, options); |
355885d5 | 374 | case OBJ_TAG: |
22410549 | 375 | return fsck_walk_tag((struct tag *)obj, data, options); |
355885d5 | 376 | default: |
f2fd0760 | 377 | error("Unknown object type for %s", oid_to_hex(&obj->oid)); |
355885d5 MK |
378 | return -1; |
379 | } | |
380 | } | |
ba002f3b MK |
381 | |
382 | /* | |
383 | * The entries in a tree are ordered in the _path_ order, | |
384 | * which means that a directory entry is ordered by adding | |
385 | * a slash to the end of it. | |
386 | * | |
387 | * So a directory called "a" is ordered _after_ a file | |
388 | * called "a.c", because "a/" sorts after "a.c". | |
389 | */ | |
390 | #define TREE_UNORDERED (-1) | |
391 | #define TREE_HAS_DUPS (-2) | |
392 | ||
393 | static int verify_ordered(unsigned mode1, const char *name1, unsigned mode2, const char *name2) | |
394 | { | |
395 | int len1 = strlen(name1); | |
396 | int len2 = strlen(name2); | |
397 | int len = len1 < len2 ? len1 : len2; | |
398 | unsigned char c1, c2; | |
399 | int cmp; | |
400 | ||
401 | cmp = memcmp(name1, name2, len); | |
402 | if (cmp < 0) | |
403 | return 0; | |
404 | if (cmp > 0) | |
405 | return TREE_UNORDERED; | |
406 | ||
407 | /* | |
408 | * Ok, the first <len> characters are the same. | |
409 | * Now we need to order the next one, but turn | |
410 | * a '\0' into a '/' for a directory entry. | |
411 | */ | |
412 | c1 = name1[len]; | |
413 | c2 = name2[len]; | |
414 | if (!c1 && !c2) | |
415 | /* | |
416 | * git-write-tree used to write out a nonsense tree that has | |
417 | * entries with the same name, one blob and one tree. Make | |
418 | * sure we do not have duplicate entries. | |
419 | */ | |
420 | return TREE_HAS_DUPS; | |
421 | if (!c1 && S_ISDIR(mode1)) | |
422 | c1 = '/'; | |
423 | if (!c2 && S_ISDIR(mode2)) | |
424 | c2 = '/'; | |
425 | return c1 < c2 ? 0 : TREE_UNORDERED; | |
426 | } | |
427 | ||
22410549 | 428 | static int fsck_tree(struct tree *item, struct fsck_options *options) |
ba002f3b MK |
429 | { |
430 | int retval; | |
c479d14a | 431 | int has_null_sha1 = 0; |
ba002f3b MK |
432 | int has_full_path = 0; |
433 | int has_empty_name = 0; | |
5d34a435 JK |
434 | int has_dot = 0; |
435 | int has_dotdot = 0; | |
5c17f512 | 436 | int has_dotgit = 0; |
ba002f3b MK |
437 | int has_zero_pad = 0; |
438 | int has_bad_modes = 0; | |
439 | int has_dup_entries = 0; | |
440 | int not_properly_sorted = 0; | |
441 | struct tree_desc desc; | |
442 | unsigned o_mode; | |
443 | const char *o_name; | |
ba002f3b MK |
444 | |
445 | init_tree_desc(&desc, item->buffer, item->size); | |
446 | ||
447 | o_mode = 0; | |
448 | o_name = NULL; | |
ba002f3b MK |
449 | |
450 | while (desc.size) { | |
451 | unsigned mode; | |
452 | const char *name; | |
c479d14a | 453 | const unsigned char *sha1; |
ba002f3b | 454 | |
c479d14a | 455 | sha1 = tree_entry_extract(&desc, &name, &mode); |
ba002f3b | 456 | |
effd12ec HS |
457 | has_null_sha1 |= is_null_sha1(sha1); |
458 | has_full_path |= !!strchr(name, '/'); | |
459 | has_empty_name |= !*name; | |
460 | has_dot |= !strcmp(name, "."); | |
461 | has_dotdot |= !strcmp(name, ".."); | |
5e519fb8 JH |
462 | has_dotgit |= (!strcmp(name, ".git") || |
463 | is_hfs_dotgit(name) || | |
464 | is_ntfs_dotgit(name)); | |
ba002f3b MK |
465 | has_zero_pad |= *(char *)desc.buffer == '0'; |
466 | update_tree_entry(&desc); | |
467 | ||
468 | switch (mode) { | |
469 | /* | |
470 | * Standard modes.. | |
471 | */ | |
472 | case S_IFREG | 0755: | |
473 | case S_IFREG | 0644: | |
474 | case S_IFLNK: | |
475 | case S_IFDIR: | |
476 | case S_IFGITLINK: | |
477 | break; | |
478 | /* | |
479 | * This is nonstandard, but we had a few of these | |
480 | * early on when we honored the full set of mode | |
481 | * bits.. | |
482 | */ | |
483 | case S_IFREG | 0664: | |
22410549 | 484 | if (!options->strict) |
ba002f3b MK |
485 | break; |
486 | default: | |
487 | has_bad_modes = 1; | |
488 | } | |
489 | ||
490 | if (o_name) { | |
491 | switch (verify_ordered(o_mode, o_name, mode, name)) { | |
492 | case TREE_UNORDERED: | |
493 | not_properly_sorted = 1; | |
494 | break; | |
495 | case TREE_HAS_DUPS: | |
496 | has_dup_entries = 1; | |
497 | break; | |
498 | default: | |
499 | break; | |
500 | } | |
501 | } | |
502 | ||
503 | o_mode = mode; | |
504 | o_name = name; | |
ba002f3b MK |
505 | } |
506 | ||
507 | retval = 0; | |
c479d14a | 508 | if (has_null_sha1) |
c99ba492 | 509 | retval += report(options, &item->object, FSCK_MSG_NULL_SHA1, "contains entries pointing to null sha1"); |
ba002f3b | 510 | if (has_full_path) |
c99ba492 | 511 | retval += report(options, &item->object, FSCK_MSG_FULL_PATHNAME, "contains full pathnames"); |
ba002f3b | 512 | if (has_empty_name) |
c99ba492 | 513 | retval += report(options, &item->object, FSCK_MSG_EMPTY_NAME, "contains empty pathname"); |
5d34a435 | 514 | if (has_dot) |
c99ba492 | 515 | retval += report(options, &item->object, FSCK_MSG_HAS_DOT, "contains '.'"); |
5d34a435 | 516 | if (has_dotdot) |
c99ba492 | 517 | retval += report(options, &item->object, FSCK_MSG_HAS_DOTDOT, "contains '..'"); |
5c17f512 | 518 | if (has_dotgit) |
c99ba492 | 519 | retval += report(options, &item->object, FSCK_MSG_HAS_DOTGIT, "contains '.git'"); |
ba002f3b | 520 | if (has_zero_pad) |
c99ba492 | 521 | retval += report(options, &item->object, FSCK_MSG_ZERO_PADDED_FILEMODE, "contains zero-padded file modes"); |
ba002f3b | 522 | if (has_bad_modes) |
c99ba492 | 523 | retval += report(options, &item->object, FSCK_MSG_BAD_FILEMODE, "contains bad file modes"); |
ba002f3b | 524 | if (has_dup_entries) |
c99ba492 | 525 | retval += report(options, &item->object, FSCK_MSG_DUPLICATE_ENTRIES, "contains duplicate file entries"); |
ba002f3b | 526 | if (not_properly_sorted) |
c99ba492 | 527 | retval += report(options, &item->object, FSCK_MSG_TREE_NOT_SORTED, "not properly sorted"); |
ba002f3b MK |
528 | return retval; |
529 | } | |
530 | ||
84d18c0b | 531 | static int verify_headers(const void *data, unsigned long size, |
b2f44feb | 532 | struct object *obj, struct fsck_options *options) |
4d0d8975 JS |
533 | { |
534 | const char *buffer = (const char *)data; | |
535 | unsigned long i; | |
536 | ||
537 | for (i = 0; i < size; i++) { | |
538 | switch (buffer[i]) { | |
539 | case '\0': | |
c99ba492 JS |
540 | return report(options, obj, |
541 | FSCK_MSG_NUL_IN_HEADER, | |
542 | "unterminated header: NUL at offset %ld", i); | |
4d0d8975 JS |
543 | case '\n': |
544 | if (i + 1 < size && buffer[i + 1] == '\n') | |
545 | return 0; | |
546 | } | |
547 | } | |
548 | ||
84d18c0b JH |
549 | /* |
550 | * We did not find double-LF that separates the header | |
551 | * and the body. Not having a body is not a crime but | |
552 | * we do want to see the terminating LF for the last header | |
553 | * line. | |
554 | */ | |
555 | if (size && buffer[size - 1] == '\n') | |
556 | return 0; | |
557 | ||
c99ba492 JS |
558 | return report(options, obj, |
559 | FSCK_MSG_UNTERMINATED_HEADER, "unterminated header"); | |
4d0d8975 JS |
560 | } |
561 | ||
22410549 | 562 | static int fsck_ident(const char **ident, struct object *obj, struct fsck_options *options) |
daae1922 | 563 | { |
e6826e33 | 564 | const char *p = *ident; |
d4b8de04 JK |
565 | char *end; |
566 | ||
e6826e33 JS |
567 | *ident = strchrnul(*ident, '\n'); |
568 | if (**ident == '\n') | |
569 | (*ident)++; | |
570 | ||
571 | if (*p == '<') | |
c99ba492 | 572 | return report(options, obj, FSCK_MSG_MISSING_NAME_BEFORE_EMAIL, "invalid author/committer line - missing space before email"); |
e6826e33 JS |
573 | p += strcspn(p, "<>\n"); |
574 | if (*p == '>') | |
c99ba492 | 575 | return report(options, obj, FSCK_MSG_BAD_NAME, "invalid author/committer line - bad name"); |
e6826e33 | 576 | if (*p != '<') |
c99ba492 | 577 | return report(options, obj, FSCK_MSG_MISSING_EMAIL, "invalid author/committer line - missing email"); |
e6826e33 | 578 | if (p[-1] != ' ') |
c99ba492 | 579 | return report(options, obj, FSCK_MSG_MISSING_SPACE_BEFORE_EMAIL, "invalid author/committer line - missing space before email"); |
e6826e33 JS |
580 | p++; |
581 | p += strcspn(p, "<>\n"); | |
582 | if (*p != '>') | |
c99ba492 | 583 | return report(options, obj, FSCK_MSG_BAD_EMAIL, "invalid author/committer line - bad email"); |
e6826e33 JS |
584 | p++; |
585 | if (*p != ' ') | |
c99ba492 | 586 | return report(options, obj, FSCK_MSG_MISSING_SPACE_BEFORE_DATE, "invalid author/committer line - missing space before date"); |
e6826e33 JS |
587 | p++; |
588 | if (*p == '0' && p[1] != ' ') | |
c99ba492 | 589 | return report(options, obj, FSCK_MSG_ZERO_PADDED_DATE, "invalid author/committer line - zero-padded date"); |
e6826e33 | 590 | if (date_overflows(strtoul(p, &end, 10))) |
c99ba492 | 591 | return report(options, obj, FSCK_MSG_BAD_DATE_OVERFLOW, "invalid author/committer line - date causes integer overflow"); |
e6826e33 | 592 | if ((end == p || *end != ' ')) |
c99ba492 | 593 | return report(options, obj, FSCK_MSG_BAD_DATE, "invalid author/committer line - bad date"); |
e6826e33 JS |
594 | p = end + 1; |
595 | if ((*p != '+' && *p != '-') || | |
596 | !isdigit(p[1]) || | |
597 | !isdigit(p[2]) || | |
598 | !isdigit(p[3]) || | |
599 | !isdigit(p[4]) || | |
600 | (p[5] != '\n')) | |
c99ba492 | 601 | return report(options, obj, FSCK_MSG_BAD_TIMEZONE, "invalid author/committer line - bad time zone"); |
e6826e33 | 602 | p += 6; |
daae1922 JN |
603 | return 0; |
604 | } | |
605 | ||
bc6b8fc1 | 606 | static int fsck_commit_buffer(struct commit *commit, const char *buffer, |
22410549 | 607 | unsigned long size, struct fsck_options *options) |
ba002f3b | 608 | { |
ba002f3b MK |
609 | unsigned char tree_sha1[20], sha1[20]; |
610 | struct commit_graft *graft; | |
c9ad147f | 611 | unsigned parent_count, parent_line_count = 0, author_count; |
daae1922 | 612 | int err; |
ba002f3b | 613 | |
b2f44feb | 614 | if (verify_headers(buffer, size, &commit->object, options)) |
4d0d8975 JS |
615 | return -1; |
616 | ||
cf4fff57 | 617 | if (!skip_prefix(buffer, "tree ", &buffer)) |
c99ba492 | 618 | return report(options, &commit->object, FSCK_MSG_MISSING_TREE, "invalid format - expected 'tree' line"); |
b3584761 JS |
619 | if (get_sha1_hex(buffer, tree_sha1) || buffer[40] != '\n') { |
620 | err = report(options, &commit->object, FSCK_MSG_BAD_TREE_SHA1, "invalid 'tree' line format - bad sha1"); | |
621 | if (err) | |
622 | return err; | |
623 | } | |
2d820a61 | 624 | buffer += 41; |
cf4fff57 | 625 | while (skip_prefix(buffer, "parent ", &buffer)) { |
b3584761 JS |
626 | if (get_sha1_hex(buffer, sha1) || buffer[40] != '\n') { |
627 | err = report(options, &commit->object, FSCK_MSG_BAD_PARENT_SHA1, "invalid 'parent' line format - bad sha1"); | |
628 | if (err) | |
629 | return err; | |
630 | } | |
2d820a61 | 631 | buffer += 41; |
9d02150c | 632 | parent_line_count++; |
ba002f3b | 633 | } |
ed1c9977 | 634 | graft = lookup_commit_graft(commit->object.oid.hash); |
9d02150c | 635 | parent_count = commit_list_count(commit->parents); |
ba002f3b | 636 | if (graft) { |
9d02150c | 637 | if (graft->nr_parent == -1 && !parent_count) |
ba002f3b | 638 | ; /* shallow commit */ |
b3584761 JS |
639 | else if (graft->nr_parent != parent_count) { |
640 | err = report(options, &commit->object, FSCK_MSG_MISSING_GRAFT, "graft objects missing"); | |
641 | if (err) | |
642 | return err; | |
643 | } | |
ba002f3b | 644 | } else { |
b3584761 JS |
645 | if (parent_count != parent_line_count) { |
646 | err = report(options, &commit->object, FSCK_MSG_MISSING_PARENT, "parent objects missing"); | |
647 | if (err) | |
648 | return err; | |
649 | } | |
ba002f3b | 650 | } |
c9ad147f JS |
651 | author_count = 0; |
652 | while (skip_prefix(buffer, "author ", &buffer)) { | |
653 | author_count++; | |
654 | err = fsck_ident(&buffer, &commit->object, options); | |
655 | if (err) | |
656 | return err; | |
ba002f3b | 657 | } |
c9ad147f JS |
658 | if (author_count < 1) |
659 | err = report(options, &commit->object, FSCK_MSG_MISSING_AUTHOR, "invalid format - expected 'author' line"); | |
660 | else if (author_count > 1) | |
661 | err = report(options, &commit->object, FSCK_MSG_MULTIPLE_AUTHORS, "invalid format - multiple 'author' lines"); | |
daae1922 JN |
662 | if (err) |
663 | return err; | |
cf4fff57 | 664 | if (!skip_prefix(buffer, "committer ", &buffer)) |
c99ba492 | 665 | return report(options, &commit->object, FSCK_MSG_MISSING_COMMITTER, "invalid format - expected 'committer' line"); |
22410549 | 666 | err = fsck_ident(&buffer, &commit->object, options); |
daae1922 JN |
667 | if (err) |
668 | return err; | |
ba002f3b | 669 | if (!commit->tree) |
c99ba492 | 670 | return report(options, &commit->object, FSCK_MSG_BAD_TREE, "could not load commit's tree %s", sha1_to_hex(tree_sha1)); |
ba002f3b MK |
671 | |
672 | return 0; | |
673 | } | |
674 | ||
90a398bb | 675 | static int fsck_commit(struct commit *commit, const char *data, |
22410549 | 676 | unsigned long size, struct fsck_options *options) |
bc6b8fc1 | 677 | { |
90a398bb | 678 | const char *buffer = data ? data : get_commit_buffer(commit, &size); |
22410549 | 679 | int ret = fsck_commit_buffer(commit, buffer, size, options); |
90a398bb JS |
680 | if (!data) |
681 | unuse_commit_buffer(commit, buffer); | |
bc6b8fc1 JK |
682 | return ret; |
683 | } | |
684 | ||
cec097be | 685 | static int fsck_tag_buffer(struct tag *tag, const char *data, |
22410549 | 686 | unsigned long size, struct fsck_options *options) |
cec097be JS |
687 | { |
688 | unsigned char sha1[20]; | |
689 | int ret = 0; | |
690 | const char *buffer; | |
691 | char *to_free = NULL, *eol; | |
692 | struct strbuf sb = STRBUF_INIT; | |
693 | ||
694 | if (data) | |
695 | buffer = data; | |
696 | else { | |
697 | enum object_type type; | |
698 | ||
699 | buffer = to_free = | |
ed1c9977 | 700 | read_sha1_file(tag->object.oid.hash, &type, &size); |
cec097be | 701 | if (!buffer) |
c99ba492 JS |
702 | return report(options, &tag->object, |
703 | FSCK_MSG_MISSING_TAG_OBJECT, | |
cec097be JS |
704 | "cannot read tag object"); |
705 | ||
706 | if (type != OBJ_TAG) { | |
c99ba492 JS |
707 | ret = report(options, &tag->object, |
708 | FSCK_MSG_TAG_OBJECT_NOT_TAG, | |
cec097be JS |
709 | "expected tag got %s", |
710 | typename(type)); | |
711 | goto done; | |
712 | } | |
713 | } | |
714 | ||
8a272f29 RS |
715 | ret = verify_headers(buffer, size, &tag->object, options); |
716 | if (ret) | |
cec097be JS |
717 | goto done; |
718 | ||
719 | if (!skip_prefix(buffer, "object ", &buffer)) { | |
c99ba492 | 720 | ret = report(options, &tag->object, FSCK_MSG_MISSING_OBJECT, "invalid format - expected 'object' line"); |
cec097be JS |
721 | goto done; |
722 | } | |
723 | if (get_sha1_hex(buffer, sha1) || buffer[40] != '\n') { | |
c99ba492 | 724 | ret = report(options, &tag->object, FSCK_MSG_BAD_OBJECT_SHA1, "invalid 'object' line format - bad sha1"); |
7d7d5b05 JS |
725 | if (ret) |
726 | goto done; | |
cec097be JS |
727 | } |
728 | buffer += 41; | |
729 | ||
730 | if (!skip_prefix(buffer, "type ", &buffer)) { | |
c99ba492 | 731 | ret = report(options, &tag->object, FSCK_MSG_MISSING_TYPE_ENTRY, "invalid format - expected 'type' line"); |
cec097be JS |
732 | goto done; |
733 | } | |
734 | eol = strchr(buffer, '\n'); | |
735 | if (!eol) { | |
c99ba492 | 736 | ret = report(options, &tag->object, FSCK_MSG_MISSING_TYPE, "invalid format - unexpected end after 'type' line"); |
cec097be JS |
737 | goto done; |
738 | } | |
739 | if (type_from_string_gently(buffer, eol - buffer, 1) < 0) | |
c99ba492 | 740 | ret = report(options, &tag->object, FSCK_MSG_BAD_TYPE, "invalid 'type' value"); |
cec097be JS |
741 | if (ret) |
742 | goto done; | |
743 | buffer = eol + 1; | |
744 | ||
745 | if (!skip_prefix(buffer, "tag ", &buffer)) { | |
c99ba492 | 746 | ret = report(options, &tag->object, FSCK_MSG_MISSING_TAG_ENTRY, "invalid format - expected 'tag' line"); |
cec097be JS |
747 | goto done; |
748 | } | |
749 | eol = strchr(buffer, '\n'); | |
750 | if (!eol) { | |
c99ba492 | 751 | ret = report(options, &tag->object, FSCK_MSG_MISSING_TAG, "invalid format - unexpected end after 'type' line"); |
cec097be JS |
752 | goto done; |
753 | } | |
754 | strbuf_addf(&sb, "refs/tags/%.*s", (int)(eol - buffer), buffer); | |
f27d05b1 JS |
755 | if (check_refname_format(sb.buf, 0)) { |
756 | ret = report(options, &tag->object, FSCK_MSG_BAD_TAG_NAME, | |
c99ba492 | 757 | "invalid 'tag' name: %.*s", |
7add4419 | 758 | (int)(eol - buffer), buffer); |
f27d05b1 JS |
759 | if (ret) |
760 | goto done; | |
761 | } | |
cec097be JS |
762 | buffer = eol + 1; |
763 | ||
f27d05b1 | 764 | if (!skip_prefix(buffer, "tagger ", &buffer)) { |
cec097be | 765 | /* early tags do not contain 'tagger' lines; warn only */ |
f27d05b1 JS |
766 | ret = report(options, &tag->object, FSCK_MSG_MISSING_TAGGER_ENTRY, "invalid format - expected 'tagger' line"); |
767 | if (ret) | |
768 | goto done; | |
769 | } | |
cec097be | 770 | else |
22410549 | 771 | ret = fsck_ident(&buffer, &tag->object, options); |
cec097be JS |
772 | |
773 | done: | |
774 | strbuf_release(&sb); | |
775 | free(to_free); | |
776 | return ret; | |
777 | } | |
778 | ||
90a398bb | 779 | static int fsck_tag(struct tag *tag, const char *data, |
22410549 | 780 | unsigned long size, struct fsck_options *options) |
ba002f3b MK |
781 | { |
782 | struct object *tagged = tag->tagged; | |
783 | ||
784 | if (!tagged) | |
c99ba492 | 785 | return report(options, &tag->object, FSCK_MSG_BAD_TAG_OBJECT, "could not load tagged object"); |
cec097be | 786 | |
22410549 | 787 | return fsck_tag_buffer(tag, data, size, options); |
ba002f3b MK |
788 | } |
789 | ||
90a398bb | 790 | int fsck_object(struct object *obj, void *data, unsigned long size, |
22410549 | 791 | struct fsck_options *options) |
ba002f3b MK |
792 | { |
793 | if (!obj) | |
c99ba492 | 794 | return report(options, obj, FSCK_MSG_BAD_OBJECT_SHA1, "no valid object to fsck"); |
ba002f3b MK |
795 | |
796 | if (obj->type == OBJ_BLOB) | |
797 | return 0; | |
798 | if (obj->type == OBJ_TREE) | |
22410549 | 799 | return fsck_tree((struct tree *) obj, options); |
ba002f3b | 800 | if (obj->type == OBJ_COMMIT) |
90a398bb | 801 | return fsck_commit((struct commit *) obj, (const char *) data, |
22410549 | 802 | size, options); |
ba002f3b | 803 | if (obj->type == OBJ_TAG) |
90a398bb | 804 | return fsck_tag((struct tag *) obj, (const char *) data, |
22410549 | 805 | size, options); |
ba002f3b | 806 | |
c99ba492 | 807 | return report(options, obj, FSCK_MSG_UNKNOWN_TYPE, "unknown type '%d' (internal fsck error)", |
ba002f3b MK |
808 | obj->type); |
809 | } | |
d6ffc8d7 | 810 | |
c99ba492 | 811 | int fsck_error_function(struct object *obj, int msg_type, const char *message) |
d6ffc8d7 | 812 | { |
0282f4dc | 813 | if (msg_type == FSCK_WARN) { |
f2fd0760 | 814 | warning("object %s: %s", oid_to_hex(&obj->oid), message); |
0282f4dc JS |
815 | return 0; |
816 | } | |
f2fd0760 | 817 | error("object %s: %s", oid_to_hex(&obj->oid), message); |
d6ffc8d7 MK |
818 | return 1; |
819 | } |