[thirdparty/git.git] / fsck-cache.c

#include "cache.h"

#include <sys/types.h>
#include <dirent.h>

struct needs {
	unsigned char parent[20];
	unsigned char needs[20];
	char tag[10];
};

struct seen {
	unsigned char sha1[20];
	char tag[10];
	unsigned needed;
};

static struct needs *needs;
static struct seen *seen;

static int nr_seen, alloc_seen, nr_needs, alloc_needs;

/*
 * These two functions build up a graph in memory about
 * what objects we've referenced, and found, and types..
 */
static int compare_seen(const void *s1, const void *s2)
{
	return memcmp(s1, s2, 20);
}

static int lookup_seen(unsigned char *sha1, char *tag)
{
	int first = 0, last = nr_seen;

	while (last > first) {
		int next = (last + first) / 2;
		struct seen *s = seen + next;
		int cmp = memcmp(sha1, s->sha1, 20);

		if (cmp < 0) {
			last = next;
			continue;
		}
		if (cmp > 0) {
			first = next+1;
			continue;
		}
		if (strcmp(tag, s->tag))
			break;
		s->needed++;
		return 1;
	}
	return 0;
}

static void check_connectivity(void)
{
	int i;

	/* Sort the "seen" tags for quicker lookup */
	qsort(seen, nr_seen, sizeof(struct seen), compare_seen);

	/* Look up all the requirements, warn about missing objects.. */
	for (i = 0; i < nr_needs; i++) {
		struct needs *n = needs + i;
		char hex[60];

		if (lookup_seen(n->needs, n->tag))
			continue;
		strcpy(hex, sha1_to_hex(n->parent));
		printf("missing %s: %s referenced by %s\n", n->tag, sha1_to_hex(n->needs), hex);
	}

	/* Tell the user about things not referenced.. */
	for (i = 0; i < nr_seen; i++) {
		struct seen *s = seen + i;

		if (s->needed)
			continue;
		printf("unreferenced %s: %s\n", s->tag, sha1_to_hex(s->sha1));
	}
}

static void mark_needs_sha1(unsigned char *parent, const char * tag, unsigned char *child)
{
	struct needs *n;

	if (nr_needs == alloc_needs) {
		alloc_needs = alloc_nr(alloc_needs);
		needs = realloc(needs, alloc_needs*sizeof(struct needs));
	}
	n = needs + nr_needs;
	nr_needs++;
	memcpy(n->parent, parent, 20);
	memcpy(n->needs, child, 20);
	strncpy(n->tag, tag, sizeof(n->tag));
}

static int mark_sha1_seen(unsigned char *sha1, char *tag)
{
	struct seen *s;

	if (nr_seen == alloc_seen) {
		alloc_seen = alloc_nr(alloc_seen);
		seen = realloc(seen, alloc_seen*sizeof(struct seen));
	}
	s = seen + nr_seen;
	memset(s, 0, sizeof(*s));
	nr_seen++;
	memcpy(s->sha1, sha1, 20);
	strncpy(s->tag, tag, sizeof(s->tag));
	
	return 0;
}

static int fsck_tree(unsigned char *sha1, void *data, unsigned long size)
{
	int warn_old_tree = 1;

	while (size) {
		int len = 1+strlen(data);
		unsigned char *file_sha1 = data + len;
		char *path = strchr(data, ' ');
		unsigned int mode;
		if (size < len + 20 || !path || sscanf(data, "%o", &mode) != 1)
			return -1;

		/* Warn about trees that don't do the recursive thing.. */
		if (warn_old_tree && strchr(path, '/')) {
			fprintf(stderr, "warning: fsck-cache: tree %s has full pathnames in it\n", sha1_to_hex(sha1));
			warn_old_tree = 0;
		}

		data += len + 20;
		size -= len + 20;
		mark_needs_sha1(sha1, S_ISDIR(mode) ? "tree" : "blob", file_sha1);
	}
	return 0;
}

static int fsck_commit(unsigned char *sha1, void *data, unsigned long size)
{
	int parents;
	unsigned char tree_sha1[20];
	unsigned char parent_sha1[20];

	if (memcmp(data, "tree ", 5))
		return -1;
	if (get_sha1_hex(data + 5, tree_sha1) < 0)
		return -1;
	mark_needs_sha1(sha1, "tree", tree_sha1);
	data += 5 + 40 + 1;	/* "tree " + <hex sha1> + '\n' */
	parents = 0;
	while (!memcmp(data, "parent ", 7)) {
		if (get_sha1_hex(data + 7, parent_sha1) < 0)
			return -1;
		mark_needs_sha1(sha1, "commit", parent_sha1);
		data += 7 + 40 + 1; 	/* "parent " + <hex sha1> + '\n' */
		parents++;
	}
	if (!parents)
		printf("root: %s\n", sha1_to_hex(sha1));
	return 0;
}

static int fsck_entry(unsigned char *sha1, char *tag, void *data, unsigned long size)
{
	if (!strcmp(tag, "blob")) {
		/* Nothing to check */;
	} else if (!strcmp(tag, "tree")) {
		if (fsck_tree(sha1, data, size) < 0)
			return -1;
	} else if (!strcmp(tag, "commit")) {
		if (fsck_commit(sha1, data, size) < 0)
			return -1;
	} else
		return -1;
	return mark_sha1_seen(sha1, tag);
}

static int fsck_name(char *hex)
{
	unsigned char sha1[20];
	if (!get_sha1_hex(hex, sha1)) {
		unsigned long mapsize;
		void *map = map_sha1_file(sha1, &mapsize);
		if (map) {
			char type[100];
			unsigned long size;
			void *buffer = NULL;
			if (!check_sha1_signature(sha1, map, mapsize))
				buffer = unpack_sha1_file(map, mapsize, type, &size);
			munmap(map, mapsize);
			if (buffer && !fsck_entry(sha1, type, buffer, size))
				return 0;
		}
	}
	return -1;
}

static int fsck_dir(int i, char *path)
{
	DIR *dir = opendir(path);
	struct dirent *de;

	if (!dir) {
		return error("missing sha1 directory '%s'", path);
	}

	while ((de = readdir(dir)) != NULL) {
		char name[100];
		int len = strlen(de->d_name);

		switch (len) {
		case 2:
			if (de->d_name[1] != '.')
				break;
		case 1:
			if (de->d_name[0] != '.')
				break;
			continue;
		case 38:
			sprintf(name, "%02x", i);
			memcpy(name+2, de->d_name, len+1);
			if (!fsck_name(name))
				continue;
		}
		fprintf(stderr, "bad sha1 file: %s/%s\n", path, de->d_name);
	}
	closedir(dir);
	return 0;
}

int main(int argc, char **argv)
{
	int i;
	char *sha1_dir;

	if (argc != 1)
		usage("fsck-cache");
	sha1_dir = getenv(DB_ENVIRONMENT) ? : DEFAULT_DB_ENVIRONMENT;
	for (i = 0; i < 256; i++) {
		static char dir[4096];
		sprintf(dir, "%s/%02x", sha1_dir, i);
		fsck_dir(i, dir);
	}
	check_connectivity();
	return 0;
}
Commit	Line	Data
20222118 LT	1	#include "cache.h"
	2
	3	#include <sys/types.h>
	4	#include <dirent.h>
	5
8ba0bbb2 LT	6	struct needs {
	7	unsigned char parent[20];
	8	unsigned char needs[20];
	9	char tag[10];
	10	};
	11
	12	struct seen {
	13	unsigned char sha1[20];
	14	char tag[10];
	15	unsigned needed;
	16	};
	17
	18	static struct needs *needs;
	19	static struct seen *seen;
	20
	21	static int nr_seen, alloc_seen, nr_needs, alloc_needs;
	22
20222118	23	/*
8ba0bbb2	24	* These two functions build up a graph in memory about
20222118	25	* what objects we've referenced, and found, and types..
20222118	26	*/
8ba0bbb2 LT	27	static int compare_seen(const void s1, const void s2)
	28	{
	29	return memcmp(s1, s2, 20);
	30	}
	31
	32	static int lookup_seen(unsigned char sha1, char tag)
	33	{
	34	int first = 0, last = nr_seen;
	35
	36	while (last > first) {
	37	int next = (last + first) / 2;
	38	struct seen *s = seen + next;
	39	int cmp = memcmp(sha1, s->sha1, 20);
	40
	41	if (cmp < 0) {
	42	last = next;
	43	continue;
	44	}
	45	if (cmp > 0) {
	46	first = next+1;
	47	continue;
	48	}
	49	if (strcmp(tag, s->tag))
	50	break;
	51	s->needed++;
	52	return 1;
	53	}
	54	return 0;
	55	}
	56
	57	static void check_connectivity(void)
	58	{
	59	int i;
	60
	61	/* Sort the "seen" tags for quicker lookup */
	62	qsort(seen, nr_seen, sizeof(struct seen), compare_seen);
	63
	64	/* Look up all the requirements, warn about missing objects.. */
	65	for (i = 0; i < nr_needs; i++) {
	66	struct needs *n = needs + i;
	67	char hex[60];
	68
	69	if (lookup_seen(n->needs, n->tag))
	70	continue;
	71	strcpy(hex, sha1_to_hex(n->parent));
f1257f3a	72	printf("missing %s: %s referenced by %s\n", n->tag, sha1_to_hex(n->needs), hex);
8ba0bbb2 LT	73	}
	74
	75	/* Tell the user about things not referenced.. */
	76	for (i = 0; i < nr_seen; i++) {
	77	struct seen *s = seen + i;
	78
	79	if (s->needed)
	80	continue;
	81	printf("unreferenced %s: %s\n", s->tag, sha1_to_hex(s->sha1));
	82	}
	83	}
	84
	85	static void mark_needs_sha1(unsigned char parent, const char tag, unsigned char *child)
20222118	86	{
8ba0bbb2 LT	87	struct needs *n;
	88
	89	if (nr_needs == alloc_needs) {
	90	alloc_needs = alloc_nr(alloc_needs);
	91	needs = realloc(needs, alloc_needs*sizeof(struct needs));
	92	}
	93	n = needs + nr_needs;
	94	nr_needs++;
	95	memcpy(n->parent, parent, 20);
	96	memcpy(n->needs, child, 20);
	97	strncpy(n->tag, tag, sizeof(n->tag));
20222118 LT	98	}
20222118 LT	99
1ea34e36	100	static int mark_sha1_seen(unsigned char sha1, char tag)
20222118	101	{
8ba0bbb2 LT	102	struct seen *s;
	103
	104	if (nr_seen == alloc_seen) {
	105	alloc_seen = alloc_nr(alloc_seen);
	106	seen = realloc(seen, alloc_seen*sizeof(struct seen));
	107	}
	108	s = seen + nr_seen;
91af81a9	109	memset(s, 0, sizeof(*s));
8ba0bbb2 LT	110	nr_seen++;
	111	memcpy(s->sha1, sha1, 20);
	112	strncpy(s->tag, tag, sizeof(s->tag));
	113
1ea34e36	114	return 0;
20222118 LT	115	}
20222118 LT	116
1ea34e36	117	static int fsck_tree(unsigned char sha1, void data, unsigned long size)
20222118	118	{
4e6616ab LT	119	int warn_old_tree = 1;
4e6616ab LT	120
1ea34e36 LT	121	while (size) {
	122	int len = 1+strlen(data);
	123	unsigned char *file_sha1 = data + len;
	124	char *path = strchr(data, ' ');
f768846e LT	125	unsigned int mode;
f768846e LT	126	if (size < len + 20 \|\| !path \|\| sscanf(data, "%o", &mode) != 1)
1ea34e36	127	return -1;
4e6616ab LT	128
	129	/* Warn about trees that don't do the recursive thing.. */
	130	if (warn_old_tree && strchr(path, '/')) {
	131	fprintf(stderr, "warning: fsck-cache: tree %s has full pathnames in it\n", sha1_to_hex(sha1));
	132	warn_old_tree = 0;
	133	}
	134
1ea34e36 LT	135	data += len + 20;
1ea34e36 LT	136	size -= len + 20;
f768846e	137	mark_needs_sha1(sha1, S_ISDIR(mode) ? "tree" : "blob", file_sha1);
1ea34e36	138	}
59c1e249	139	return 0;
1ea34e36 LT	140	}
	141
	142	static int fsck_commit(unsigned char sha1, void data, unsigned long size)
	143	{
16d4d1ba	144	int parents;
1ea34e36 LT	145	unsigned char tree_sha1[20];
	146	unsigned char parent_sha1[20];
	147
	148	if (memcmp(data, "tree ", 5))
	149	return -1;
	150	if (get_sha1_hex(data + 5, tree_sha1) < 0)
	151	return -1;
	152	mark_needs_sha1(sha1, "tree", tree_sha1);
	153	data += 5 + 40 + 1; /* "tree " + <hex sha1> + '\n' */
16d4d1ba	154	parents = 0;
1ea34e36 LT	155	while (!memcmp(data, "parent ", 7)) {
	156	if (get_sha1_hex(data + 7, parent_sha1) < 0)
	157	return -1;
	158	mark_needs_sha1(sha1, "commit", parent_sha1);
	159	data += 7 + 40 + 1; /* "parent " + <hex sha1> + '\n' */
16d4d1ba	160	parents++;
1ea34e36	161	}
16d4d1ba LT	162	if (!parents)
16d4d1ba LT	163	printf("root: %s\n", sha1_to_hex(sha1));
59c1e249	164	return 0;
20222118 LT	165	}
	166
	167	static int fsck_entry(unsigned char sha1, char tag, void *data, unsigned long size)
	168	{
1ea34e36	169	if (!strcmp(tag, "blob")) {
20222118	170	/* Nothing to check */;
1ea34e36 LT	171	} else if (!strcmp(tag, "tree")) {
	172	if (fsck_tree(sha1, data, size) < 0)
	173	return -1;
	174	} else if (!strcmp(tag, "commit")) {
	175	if (fsck_commit(sha1, data, size) < 0)
	176	return -1;
	177	} else
20222118 LT	178	return -1;
	179	return mark_sha1_seen(sha1, tag);
	180	}
	181
	182	static int fsck_name(char *hex)
	183	{
	184	unsigned char sha1[20];
	185	if (!get_sha1_hex(hex, sha1)) {
	186	unsigned long mapsize;
	187	void *map = map_sha1_file(sha1, &mapsize);
	188	if (map) {
	189	char type[100];
	190	unsigned long size;
	191	void *buffer = NULL;
	192	if (!check_sha1_signature(sha1, map, mapsize))
	193	buffer = unpack_sha1_file(map, mapsize, type, &size);
	194	munmap(map, mapsize);
	195	if (buffer && !fsck_entry(sha1, type, buffer, size))
	196	return 0;
	197	}
	198	}
	199	return -1;
	200	}
	201
	202	static int fsck_dir(int i, char *path)
	203	{
	204	DIR *dir = opendir(path);
	205	struct dirent *de;
	206
	207	if (!dir) {
2de381f9	208	return error("missing sha1 directory '%s'", path);
20222118 LT	209	}
	210
	211	while ((de = readdir(dir)) != NULL) {
	212	char name[100];
	213	int len = strlen(de->d_name);
	214
	215	switch (len) {
	216	case 2:
	217	if (de->d_name[1] != '.')
	218	break;
	219	case 1:
	220	if (de->d_name[0] != '.')
	221	break;
	222	continue;
	223	case 38:
	224	sprintf(name, "%02x", i);
	225	memcpy(name+2, de->d_name, len+1);
	226	if (!fsck_name(name))
	227	continue;
	228	}
	229	fprintf(stderr, "bad sha1 file: %s/%s\n", path, de->d_name);
	230	}
	231	closedir(dir);
	232	return 0;
	233	}
	234
	235	int main(int argc, char **argv)
	236	{
	237	int i;
	238	char *sha1_dir;
	239
	240	if (argc != 1)
	241	usage("fsck-cache");
	242	sha1_dir = getenv(DB_ENVIRONMENT) ? : DEFAULT_DB_ENVIRONMENT;
	243	for (i = 0; i < 256; i++) {
	244	static char dir[4096];
	245	sprintf(dir, "%s/%02x", sha1_dir, i);
	246	fsck_dir(i, dir);
	247	}
8ba0bbb2	248	check_connectivity();
20222118 LT	249	return 0;
20222118 LT	250	}