6 #include "list-objects.h"
8 #include "sha1-lookup.h"
9 #include "run-command.h"
14 unsigned char (*sha1)[20];
20 static struct sha1_array good_revs;
21 static struct sha1_array skipped_revs;
23 static const unsigned char *current_bad_sha1;
31 static const char *argv_checkout[] = {"checkout", "-q", NULL, "--", NULL};
32 static const char *argv_show_branch[] = {"show-branch", NULL, NULL};
34 /* bits #0-15 in revision.h */
36 #define COUNTED (1u<<16)
39 * This is a truly stupid algorithm, but it's only
40 * used for bisection, and we just don't care enough.
42 * We care just barely enough to avoid recursing for
45 static int count_distance(struct commit_list *entry)
50 struct commit *commit = entry->item;
51 struct commit_list *p;
53 if (commit->object.flags & (UNINTERESTING | COUNTED))
55 if (!(commit->object.flags & TREESAME))
57 commit->object.flags |= COUNTED;
63 nr += count_distance(p);
72 static void clear_distance(struct commit_list *list)
75 struct commit *commit = list->item;
76 commit->object.flags &= ~COUNTED;
81 #define DEBUG_BISECT 0
83 static inline int weight(struct commit_list *elem)
85 return *((int*)(elem->item->util));
88 static inline void weight_set(struct commit_list *elem, int weight)
90 *((int*)(elem->item->util)) = weight;
93 static int count_interesting_parents(struct commit *commit)
95 struct commit_list *p;
98 for (count = 0, p = commit->parents; p; p = p->next) {
99 if (p->item->object.flags & UNINTERESTING)
106 static inline int halfway(struct commit_list *p, int nr)
109 * Don't short-cut something we are not going to return!
111 if (p->item->object.flags & TREESAME)
116 * 2 and 3 are halfway of 5.
117 * 3 is halfway of 6 but 2 and 4 are not.
119 switch (2 * weight(p) - nr) {
120 case -1: case 0: case 1:
128 #define show_list(a,b,c,d) do { ; } while (0)
130 static void show_list(const char *debug, int counted, int nr,
131 struct commit_list *list)
133 struct commit_list *p;
135 fprintf(stderr, "%s (%d/%d)\n", debug, counted, nr);
137 for (p = list; p; p = p->next) {
138 struct commit_list *pp;
139 struct commit *commit = p->item;
140 unsigned flags = commit->object.flags;
141 enum object_type type;
143 char *buf = read_sha1_file(commit->object.sha1, &type, &size);
146 fprintf(stderr, "%c%c%c ",
147 (flags & TREESAME) ? ' ' : 'T',
148 (flags & UNINTERESTING) ? 'U' : ' ',
149 (flags & COUNTED) ? 'C' : ' ');
151 fprintf(stderr, "%3d", weight(p));
153 fprintf(stderr, "---");
154 fprintf(stderr, " %.*s", 8, sha1_to_hex(commit->object.sha1));
155 for (pp = commit->parents; pp; pp = pp->next)
156 fprintf(stderr, " %.*s", 8,
157 sha1_to_hex(pp->item->object.sha1));
159 sp = strstr(buf, "\n\n");
162 for (ep = sp; *ep && *ep != '\n'; ep++)
164 fprintf(stderr, " %.*s", (int)(ep - sp), sp);
166 fprintf(stderr, "\n");
169 #endif /* DEBUG_BISECT */
171 static struct commit_list *best_bisection(struct commit_list *list, int nr)
173 struct commit_list *p, *best;
174 int best_distance = -1;
177 for (p = list; p; p = p->next) {
179 unsigned flags = p->item->object.flags;
181 if (flags & TREESAME)
183 distance = weight(p);
184 if (nr - distance < distance)
185 distance = nr - distance;
186 if (distance > best_distance) {
188 best_distance = distance;
196 struct commit *commit;
200 static int compare_commit_dist(const void *a_, const void *b_)
202 struct commit_dist *a, *b;
204 a = (struct commit_dist *)a_;
205 b = (struct commit_dist *)b_;
206 if (a->distance != b->distance)
207 return b->distance - a->distance; /* desc sort */
208 return hashcmp(a->commit->object.sha1, b->commit->object.sha1);
211 static struct commit_list *best_bisection_sorted(struct commit_list *list, int nr)
213 struct commit_list *p;
214 struct commit_dist *array = xcalloc(nr, sizeof(*array));
217 for (p = list, cnt = 0; p; p = p->next) {
219 unsigned flags = p->item->object.flags;
221 if (flags & TREESAME)
223 distance = weight(p);
224 if (nr - distance < distance)
225 distance = nr - distance;
226 array[cnt].commit = p->item;
227 array[cnt].distance = distance;
230 qsort(array, cnt, sizeof(*array), compare_commit_dist);
231 for (p = list, i = 0; i < cnt; i++) {
232 struct name_decoration *r = xmalloc(sizeof(*r) + 100);
233 struct object *obj = &(array[i].commit->object);
235 sprintf(r->name, "dist=%d", array[i].distance);
236 r->next = add_decoration(&name_decoration, obj, r);
237 p->item = array[i].commit;
247 * zero or positive weight is the number of interesting commits it can
248 * reach, including itself. Especially, weight = 0 means it does not
249 * reach any tree-changing commits (e.g. just above uninteresting one
250 * but traversal is with pathspec).
252 * weight = -1 means it has one parent and its distance is yet to
255 * weight = -2 means it has more than one parent and its distance is
256 * unknown. After running count_distance() first, they will get zero
257 * or positive distance.
259 static struct commit_list *do_find_bisection(struct commit_list *list,
260 int nr, int *weights,
264 struct commit_list *p;
268 for (n = 0, p = list; p; p = p->next) {
269 struct commit *commit = p->item;
270 unsigned flags = commit->object.flags;
272 p->item->util = &weights[n++];
273 switch (count_interesting_parents(commit)) {
275 if (!(flags & TREESAME)) {
278 show_list("bisection 2 count one",
282 * otherwise, it is known not to reach any
283 * tree-changing commit and gets weight 0.
295 show_list("bisection 2 initialize", counted, nr, list);
298 * If you have only one parent in the resulting set
299 * then you can reach one commit more than that parent
300 * can reach. So we do not have to run the expensive
301 * count_distance() for single strand of pearls.
303 * However, if you have more than one parents, you cannot
304 * just add their distance and one for yourself, since
305 * they usually reach the same ancestor and you would
306 * end up counting them twice that way.
308 * So we will first count distance of merges the usual
309 * way, and then fill the blanks using cheaper algorithm.
311 for (p = list; p; p = p->next) {
312 if (p->item->object.flags & UNINTERESTING)
316 weight_set(p, count_distance(p));
317 clear_distance(list);
319 /* Does it happen to be at exactly half-way? */
320 if (!find_all && halfway(p, nr))
325 show_list("bisection 2 count_distance", counted, nr, list);
327 while (counted < nr) {
328 for (p = list; p; p = p->next) {
329 struct commit_list *q;
330 unsigned flags = p->item->object.flags;
334 for (q = p->item->parents; q; q = q->next) {
335 if (q->item->object.flags & UNINTERESTING)
344 * weight for p is unknown but q is known.
345 * add one for p itself if p is to be counted,
346 * otherwise inherit it from q directly.
348 if (!(flags & TREESAME)) {
349 weight_set(p, weight(q)+1);
351 show_list("bisection 2 count one",
355 weight_set(p, weight(q));
357 /* Does it happen to be at exactly half-way? */
358 if (!find_all && halfway(p, nr))
363 show_list("bisection 2 counted all", counted, nr, list);
366 return best_bisection(list, nr);
368 return best_bisection_sorted(list, nr);
371 struct commit_list *find_bisection(struct commit_list *list,
372 int *reaches, int *all,
376 struct commit_list *p, *best, *next, *last;
379 show_list("bisection 2 entry", 0, 0, list);
382 * Count the number of total and tree-changing items on the
383 * list, while reversing the list.
385 for (nr = on_list = 0, last = NULL, p = list;
388 unsigned flags = p->item->object.flags;
391 if (flags & UNINTERESTING)
395 if (!(flags & TREESAME))
400 show_list("bisection 2 sorted", 0, nr, list);
403 weights = xcalloc(on_list, sizeof(*weights));
405 /* Do the real work of finding bisection commit. */
406 best = do_find_bisection(list, nr, weights, find_all);
410 *reaches = weight(best);
416 static void argv_array_push(struct argv_array *array, const char *string)
418 ALLOC_GROW(array->argv, array->argv_nr + 1, array->argv_alloc);
419 array->argv[array->argv_nr++] = string;
422 static void argv_array_push_sha1(struct argv_array *array,
423 const unsigned char *sha1,
426 struct strbuf buf = STRBUF_INIT;
427 strbuf_addf(&buf, format, sha1_to_hex(sha1));
428 argv_array_push(array, strbuf_detach(&buf, NULL));
431 static void sha1_array_push(struct sha1_array *array,
432 const unsigned char *sha1)
434 ALLOC_GROW(array->sha1, array->sha1_nr + 1, array->sha1_alloc);
435 hashcpy(array->sha1[array->sha1_nr++], sha1);
438 static int register_ref(const char *refname, const unsigned char *sha1,
439 int flags, void *cb_data)
441 if (!strcmp(refname, "bad")) {
442 current_bad_sha1 = sha1;
443 } else if (!prefixcmp(refname, "good-")) {
444 sha1_array_push(&good_revs, sha1);
445 } else if (!prefixcmp(refname, "skip-")) {
446 sha1_array_push(&skipped_revs, sha1);
452 static int read_bisect_refs(void)
454 return for_each_ref_in("refs/bisect/", register_ref, NULL);
457 void read_bisect_paths(struct argv_array *array)
459 struct strbuf str = STRBUF_INIT;
460 const char *filename = git_path("BISECT_NAMES");
461 FILE *fp = fopen(filename, "r");
464 die("Could not open file '%s': %s", filename, strerror(errno));
466 while (strbuf_getline(&str, fp, '\n') != EOF) {
471 quoted = strbuf_detach(&str, NULL);
472 res = sq_dequote_to_argv(quoted, &array->argv,
473 &array->argv_nr, &array->argv_alloc);
475 die("Badly quoted content in file '%s': %s",
479 strbuf_release(&str);
483 static int array_cmp(const void *a, const void *b)
485 return hashcmp(a, b);
488 static void sort_sha1_array(struct sha1_array *array)
490 qsort(array->sha1, array->sha1_nr, sizeof(*array->sha1), array_cmp);
495 static const unsigned char *sha1_access(size_t index, void *table)
497 unsigned char (*array)[20] = table;
501 static int lookup_sha1_array(struct sha1_array *array,
502 const unsigned char *sha1)
505 sort_sha1_array(array);
507 return sha1_pos(sha1, array->sha1, array->sha1_nr, sha1_access);
510 static char *join_sha1_array_hex(struct sha1_array *array, char delim)
512 struct strbuf joined_hexs = STRBUF_INIT;
515 for (i = 0; i < array->sha1_nr; i++) {
516 strbuf_addstr(&joined_hexs, sha1_to_hex(array->sha1[i]));
517 if (i + 1 < array->sha1_nr)
518 strbuf_addch(&joined_hexs, delim);
521 return strbuf_detach(&joined_hexs, NULL);
525 * In this function, passing a not NULL skipped_first is very special.
526 * It means that we want to know if the first commit in the list is
527 * skipped because we will want to test a commit away from it if it is
529 * So if the first commit is skipped, we cannot take the shortcut to
530 * just "return list" when we find the first non skipped commit, we
531 * have to return a fully filtered list.
533 * We use (*skipped_first == -1) to mean "it has been found that the
534 * first commit is not skipped". In this case *skipped_first is set back
535 * to 0 just before the function returns.
537 struct commit_list *filter_skipped(struct commit_list *list,
538 struct commit_list **tried,
543 struct commit_list *filtered = NULL, **f = &filtered;
552 if (!skipped_revs.sha1_nr)
556 struct commit_list *next = list->next;
558 if (0 <= lookup_sha1_array(&skipped_revs,
559 list->item->object.sha1)) {
560 if (skipped_first && !*skipped_first)
562 /* Move current to tried list */
567 if (!skipped_first || !*skipped_first)
569 } else if (skipped_first && !*skipped_first) {
570 /* This means we know it's not skipped */
573 /* Move current to filtered list */
582 if (skipped_first && *skipped_first == -1)
588 static struct commit_list *apply_skip_ratio(struct commit_list *list,
590 int skip_num, int skip_denom)
593 struct commit_list *cur, *previous;
597 index = count * skip_num / skip_denom;
599 for (i = 0; cur; cur = cur->next, i++) {
601 if (hashcmp(cur->item->object.sha1, current_bad_sha1))
613 static struct commit_list *managed_skipped(struct commit_list *list,
614 struct commit_list **tried)
616 int count, skipped_first;
617 int skip_num, skip_denom;
621 if (!skipped_revs.sha1_nr)
624 list = filter_skipped(list, tried, 0, &count, &skipped_first);
629 /* Use alternatively 1/5, 2/5 and 3/5 as skip ratio. */
630 skip_num = count % 3 + 1;
633 return apply_skip_ratio(list, count, skip_num, skip_denom);
636 static void bisect_rev_setup(struct rev_info *revs, const char *prefix,
637 const char *bad_format, const char *good_format,
640 struct argv_array rev_argv = { NULL, 0, 0 };
643 init_revisions(revs, prefix);
645 revs->commit_format = CMIT_FMT_UNSPECIFIED;
647 /* rev_argv.argv[0] will be ignored by setup_revisions */
648 argv_array_push(&rev_argv, xstrdup("bisect_rev_setup"));
649 argv_array_push_sha1(&rev_argv, current_bad_sha1, bad_format);
650 for (i = 0; i < good_revs.sha1_nr; i++)
651 argv_array_push_sha1(&rev_argv, good_revs.sha1[i],
653 argv_array_push(&rev_argv, xstrdup("--"));
655 read_bisect_paths(&rev_argv);
656 argv_array_push(&rev_argv, NULL);
658 setup_revisions(rev_argv.argv_nr, rev_argv.argv, revs, NULL);
661 static void bisect_common(struct rev_info *revs)
663 if (prepare_revision_walk(revs))
664 die("revision walk setup failed");
665 if (revs->tree_objects)
666 mark_edges_uninteresting(revs->commits, revs, NULL);
669 static void exit_if_skipped_commits(struct commit_list *tried,
670 const unsigned char *bad)
675 printf("There are only 'skip'ped commits left to test.\n"
676 "The first bad commit could be any of:\n");
677 print_commit_list(tried, "%s\n", "%s\n");
679 printf("%s\n", sha1_to_hex(bad));
680 printf("We cannot bisect more!\n");
684 static int is_expected_rev(const unsigned char *sha1)
686 const char *filename = git_path("BISECT_EXPECTED_REV");
688 struct strbuf str = STRBUF_INIT;
692 if (stat(filename, &st) || !S_ISREG(st.st_mode))
695 fp = fopen(filename, "r");
699 if (strbuf_getline(&str, fp, '\n') != EOF)
700 res = !strcmp(str.buf, sha1_to_hex(sha1));
702 strbuf_release(&str);
708 static void mark_expected_rev(char *bisect_rev_hex)
710 int len = strlen(bisect_rev_hex);
711 const char *filename = git_path("BISECT_EXPECTED_REV");
712 int fd = open(filename, O_CREAT | O_TRUNC | O_WRONLY, 0600);
715 die("could not create file '%s': %s",
716 filename, strerror(errno));
718 bisect_rev_hex[len] = '\n';
719 write_or_die(fd, bisect_rev_hex, len + 1);
720 bisect_rev_hex[len] = '\0';
723 die("closing file %s: %s", filename, strerror(errno));
726 static int bisect_checkout(char *bisect_rev_hex)
730 mark_expected_rev(bisect_rev_hex);
732 argv_checkout[2] = bisect_rev_hex;
733 res = run_command_v_opt(argv_checkout, RUN_GIT_CMD);
737 argv_show_branch[1] = bisect_rev_hex;
738 return run_command_v_opt(argv_show_branch, RUN_GIT_CMD);
741 static struct commit *get_commit_reference(const unsigned char *sha1)
743 struct commit *r = lookup_commit_reference(sha1);
745 die("Not a valid commit name %s", sha1_to_hex(sha1));
749 static struct commit **get_bad_and_good_commits(int *rev_nr)
751 int len = 1 + good_revs.sha1_nr;
752 struct commit **rev = xmalloc(len * sizeof(*rev));
755 rev[n++] = get_commit_reference(current_bad_sha1);
756 for (i = 0; i < good_revs.sha1_nr; i++)
757 rev[n++] = get_commit_reference(good_revs.sha1[i]);
763 static void handle_bad_merge_base(void)
765 if (is_expected_rev(current_bad_sha1)) {
766 char *bad_hex = sha1_to_hex(current_bad_sha1);
767 char *good_hex = join_sha1_array_hex(&good_revs, ' ');
769 fprintf(stderr, "The merge base %s is bad.\n"
770 "This means the bug has been fixed "
771 "between %s and [%s].\n",
772 bad_hex, bad_hex, good_hex);
777 fprintf(stderr, "Some good revs are not ancestor of the bad rev.\n"
778 "git bisect cannot work properly in this case.\n"
779 "Maybe you mistake good and bad revs?\n");
783 void handle_skipped_merge_base(const unsigned char *mb)
785 char *mb_hex = sha1_to_hex(mb);
786 char *bad_hex = sha1_to_hex(current_bad_sha1);
787 char *good_hex = join_sha1_array_hex(&good_revs, ' ');
789 fprintf(stderr, "Warning: the merge base between %s and [%s] "
791 "So we cannot be sure the first bad commit is "
792 "between %s and %s.\n"
793 "We continue anyway.\n",
794 bad_hex, good_hex, mb_hex, bad_hex);
799 * "check_merge_bases" checks that merge bases are not "bad".
801 * - If one is "bad", it means the user assumed something wrong
802 * and we must exit with a non 0 error code.
803 * - If one is "good", that's good, we have nothing to do.
804 * - If one is "skipped", we can't know but we should warn.
805 * - If we don't know, we should check it out and ask the user to test.
807 static void check_merge_bases(void)
809 struct commit_list *result;
811 struct commit **rev = get_bad_and_good_commits(&rev_nr);
813 result = get_merge_bases_many(rev[0], rev_nr - 1, rev + 1, 0);
815 for (; result; result = result->next) {
816 const unsigned char *mb = result->item->object.sha1;
817 if (!hashcmp(mb, current_bad_sha1)) {
818 handle_bad_merge_base();
819 } else if (0 <= lookup_sha1_array(&good_revs, mb)) {
821 } else if (0 <= lookup_sha1_array(&skipped_revs, mb)) {
822 handle_skipped_merge_base(mb);
824 printf("Bisecting: a merge base must be tested\n");
825 exit(bisect_checkout(sha1_to_hex(mb)));
830 free_commit_list(result);
833 static int check_ancestors(const char *prefix)
835 struct rev_info revs;
836 struct object_array pending_copy;
839 bisect_rev_setup(&revs, prefix, "^%s", "%s", 0);
841 /* Save pending objects, so they can be cleaned up later. */
842 memset(&pending_copy, 0, sizeof(pending_copy));
843 for (i = 0; i < revs.pending.nr; i++)
844 add_object_array(revs.pending.objects[i].item,
845 revs.pending.objects[i].name,
848 bisect_common(&revs);
849 res = (revs.commits != NULL);
851 /* Clean up objects used, as they will be reused. */
852 for (i = 0; i < pending_copy.nr; i++) {
853 struct object *o = pending_copy.objects[i].item;
854 clear_commit_marks((struct commit *)o, ALL_REV_FLAGS);
861 * "check_good_are_ancestors_of_bad" checks that all "good" revs are
862 * ancestor of the "bad" rev.
864 * If that's not the case, we need to check the merge bases.
865 * If a merge base must be tested by the user, its source code will be
866 * checked out to be tested by the user and we will exit.
868 static void check_good_are_ancestors_of_bad(const char *prefix)
870 const char *filename = git_path("BISECT_ANCESTORS_OK");
874 if (!current_bad_sha1)
875 die("a bad revision is needed");
877 /* Check if file BISECT_ANCESTORS_OK exists. */
878 if (!stat(filename, &st) && S_ISREG(st.st_mode))
881 /* Bisecting with no good rev is ok. */
882 if (good_revs.sha1_nr == 0)
885 /* Check if all good revs are ancestor of the bad rev. */
886 if (check_ancestors(prefix))
889 /* Create file BISECT_ANCESTORS_OK. */
890 fd = open(filename, O_CREAT | O_TRUNC | O_WRONLY, 0600);
892 warning("could not create file '%s': %s",
893 filename, strerror(errno));
899 * This does "git diff-tree --pretty COMMIT" without one fork+exec.
901 static void show_diff_tree(const char *prefix, struct commit *commit)
906 init_revisions(&opt, prefix);
907 git_config(git_diff_basic_config, NULL); /* no "diff" UI options */
911 /* This is what "--pretty" does */
912 opt.verbose_header = 1;
913 opt.use_terminator = 0;
914 opt.commit_format = CMIT_FMT_DEFAULT;
917 if (!opt.diffopt.output_format)
918 opt.diffopt.output_format = DIFF_FORMAT_RAW;
920 log_tree_commit(&opt, commit);
924 * We use the convention that exiting with an exit code 10 means that
925 * the bisection process finished successfully.
926 * In this case the calling shell script should exit 0.
928 int bisect_next_all(const char *prefix)
930 struct rev_info revs;
931 struct commit_list *tried;
932 int reaches = 0, all = 0, nr;
933 const unsigned char *bisect_rev;
934 char bisect_rev_hex[41];
936 if (read_bisect_refs())
937 die("reading bisect refs failed");
939 check_good_are_ancestors_of_bad(prefix);
941 bisect_rev_setup(&revs, prefix, "%s", "^%s", 1);
944 bisect_common(&revs);
946 revs.commits = find_bisection(revs.commits, &reaches, &all,
947 !!skipped_revs.sha1_nr);
948 revs.commits = managed_skipped(revs.commits, &tried);
952 * We should exit here only if the "bad"
953 * commit is also a "skip" commit.
955 exit_if_skipped_commits(tried, NULL);
957 printf("%s was both good and bad\n",
958 sha1_to_hex(current_bad_sha1));
962 bisect_rev = revs.commits->item->object.sha1;
963 memcpy(bisect_rev_hex, sha1_to_hex(bisect_rev), 41);
965 if (!hashcmp(bisect_rev, current_bad_sha1)) {
966 exit_if_skipped_commits(tried, current_bad_sha1);
967 printf("%s is first bad commit\n", bisect_rev_hex);
968 show_diff_tree(prefix, revs.commits->item);
969 /* This means the bisection process succeeded. */
973 nr = all - reaches - 1;
974 printf("Bisecting: %d revisions left to test after this "
975 "(roughly %d steps)\n", nr, estimate_bisect_steps(all));
977 return bisect_checkout(bisect_rev_hex);