X-Git-Url: https://git.octo.it/?p=git.git;a=blobdiff_plain;f=pack-redundant.c;h=cd81f5a66eae2a4c1fdab4086b896d6a06dcf687;hp=db3dcde22a3b290bd60762fc0ad81165c02f1fbb;hb=162f41292167a800432fc6bbacfcd9f93a90b0c8;hpb=9bc0f32c77b755b8d77d215cc47fafff53cad6f5

diff --git a/pack-redundant.c b/pack-redundant.c
index db3dcde2..cd81f5a6 100644
--- a/pack-redundant.c
+++ b/pack-redundant.c
@@ -8,50 +8,76 @@
 
 #include "cache.h"
 
+#define BLKSIZE 512
+
 static const char pack_redundant_usage[] =
-"git-pack-redundant [ -v ]  < -a | <.pack filename> ...>";
+"git-pack-redundant [ --verbose ] [ --alt-odb ] < --all | <.pack filename> ...>";
 
-int all = 0, verbose = 0;
+static int load_all_packs = 0, verbose = 0, alt_odb = 0;
 
 struct llist_item {
 	struct llist_item *next;
-	char *sha1;
+	unsigned char *sha1;
 };
-struct llist {
+static struct llist {
 	struct llist_item *front;
 	struct llist_item *back;
 	size_t size;
-} *all_objects;
+} *all_objects; /* all objects which must be present in local packfiles */
 
-struct pack_list {
+static struct pack_list {
 	struct pack_list *next;
 	struct packed_git *pack;
 	struct llist *unique_objects;
 	struct llist *all_objects;
-} *pack_list;
+} *local_packs = NULL, *altodb_packs = NULL;
 
 struct pll {
 	struct pll *next;
 	struct pack_list *pl;
 };
 
-inline void llist_free(struct llist *list)
+static struct llist_item *free_nodes = NULL;
+
+static inline void llist_item_put(struct llist_item *item)
+{
+	item->next = free_nodes;
+	free_nodes = item;
+}
+
+static inline struct llist_item *llist_item_get(void)
+{
+	struct llist_item *new;
+	if ( free_nodes ) {
+		new = free_nodes;
+		free_nodes = free_nodes->next;
+	} else {
+		int i = 1;
+		new = xmalloc(sizeof(struct llist_item) * BLKSIZE);
+		for(;i < BLKSIZE; i++) {
+			llist_item_put(&new[i]);
+		}
+	}
+	return new;
+}
+
+static void llist_free(struct llist *list)
 {
 	while((list->back = list->front)) {
 		list->front = list->front->next;
-		free(list->back);
+		llist_item_put(list->back);
 	}
 	free(list);
 }
 
-inline void llist_init(struct llist **list)
+static inline void llist_init(struct llist **list)
 {
 	*list = xmalloc(sizeof(struct llist));
 	(*list)->front = (*list)->back = NULL;
 	(*list)->size = 0;
 }
 
-struct llist * llist_copy(struct llist *list)
+static struct llist * llist_copy(struct llist *list)
 {
 	struct llist *ret;
 	struct llist_item *new, *old, *prev;
@@ -61,13 +87,13 @@ struct llist * llist_copy(struct llist *list)
 	if ((ret->size = list->size) == 0)
 		return ret;
 
-	new = ret->front = xmalloc(sizeof(struct llist_item));
+	new = ret->front = llist_item_get();
 	new->sha1 = list->front->sha1;
 
 	old = list->front->next;
 	while (old) {
 		prev = new;
-		new = xmalloc(sizeof(struct llist_item));
+		new = llist_item_get();
 		prev->next = new;
 		new->sha1 = old->sha1;
 		old = old->next;
@@ -78,10 +104,11 @@ struct llist * llist_copy(struct llist *list)
 	return ret;
 }
 
-inline struct llist_item * llist_insert(struct llist *list,
-					struct llist_item *after, char *sha1)
+static inline struct llist_item * llist_insert(struct llist *list,
+					       struct llist_item *after,
+					       unsigned char *sha1)
 {
-	struct llist_item *new = xmalloc(sizeof(struct llist_item));
+	struct llist_item *new = llist_item_get();
 	new->sha1 = sha1;
 	new->next = NULL;
 
@@ -101,13 +128,12 @@ inline struct llist_item * llist_insert(struct llist *list,
 	return new;
 }
 
-inline struct llist_item * llist_insert_back(struct llist *list, char *sha1)
+static inline struct llist_item *llist_insert_back(struct llist *list, unsigned char *sha1)
 {
 	return llist_insert(list, list->back, sha1);
 }
 
-inline struct llist_item * llist_insert_sorted_unique(struct llist *list,
-					char *sha1, struct llist_item *hint)
+static inline struct llist_item *llist_insert_sorted_unique(struct llist *list, unsigned char *sha1, struct llist_item *hint)
 {
 	struct llist_item *prev = NULL, *l;
 
@@ -127,38 +153,8 @@ inline struct llist_item * llist_insert_sorted_unique(struct llist *list,
 	return llist_insert_back(list, sha1);
 }
 
-/* computes A\B */
-struct llist * llist_sorted_difference(struct llist_item *A,
-				       struct llist_item *B)
-{
-	struct llist *ret;
-	llist_init(&ret);
-
-	while (A != NULL && B != NULL) {
-		int cmp = memcmp(A->sha1, B->sha1, 20);
-		if (!cmp) {
-			A = A->next;
-			B = B->next;
-			continue;
-		}
-		if(cmp > 0) { /* we'll never find this B */
-			B = B->next;
-			continue;
-		}
-		/* A has the object, B doesn't */
-		llist_insert_back(ret, A->sha1);
-		A = A->next;
-	}
-	while (A != NULL) {
-		llist_insert_back(ret, A->sha1);
-		A = A->next;
-	}
-	return ret;
-}
-
 /* returns a pointer to an item in front of sha1 */
-inline struct llist_item * llist_sorted_remove(struct llist *list, char *sha1,
-					       struct llist_item *hint)
+static inline struct llist_item * llist_sorted_remove(struct llist *list, const unsigned char *sha1, struct llist_item *hint)
 {
 	struct llist_item *prev, *l;
 
@@ -181,7 +177,7 @@ redo_from_start:
 				prev->next = l->next;
 			if (l == list->back)
 				list->back = prev;
-			free(l);
+			llist_item_put(l);
 			list->size--;
 			return prev;
 		}
@@ -191,7 +187,22 @@ redo_from_start:
 	return prev;
 }
 
-inline struct pack_list * pack_list_insert(struct pack_list **pl,
+/* computes A\B */
+static void llist_sorted_difference_inplace(struct llist *A,
+				     struct llist *B)
+{
+	struct llist_item *hint, *b;
+
+	hint = NULL;
+	b = B->front;
+
+	while (b) {
+		hint = llist_sorted_remove(A, b->sha1, hint);
+		b = b->next;
+	}
+}
+
+static inline struct pack_list * pack_list_insert(struct pack_list **pl,
 					   struct pack_list *entry)
 {
 	struct pack_list *p = xmalloc(sizeof(struct pack_list));
@@ -201,10 +212,21 @@ inline struct pack_list * pack_list_insert(struct pack_list **pl,
 	return p;
 }
 
-struct pack_list * pack_list_difference(struct pack_list *A,
-					struct pack_list *B)
+static inline size_t pack_list_size(struct pack_list *pl)
 {
-	struct pack_list *ret, *pl;
+	size_t ret = 0;
+	while(pl) {
+		ret++;
+		pl = pl->next;
+	}
+	return ret;
+}
+
+static struct pack_list * pack_list_difference(const struct pack_list *A,
+					       const struct pack_list *B)
+{
+	struct pack_list *ret;
+	const struct pack_list *pl;
 
 	if (A == NULL)
 		return NULL;
@@ -221,7 +243,7 @@ struct pack_list * pack_list_difference(struct pack_list *A,
 	return ret;
 }
 
-void cmp_two_packs(struct pack_list *p1, struct pack_list *p2)
+static void cmp_two_packs(struct pack_list *p1, struct pack_list *p2)
 {
 	int p1_off, p2_off;
 	void *p1_base, *p2_base;
@@ -253,56 +275,68 @@ void cmp_two_packs(struct pack_list *p1, struct pack_list *p2)
 	}
 }
 
+static void pll_free(struct pll *l)
+{
+	struct pll *old;
+	struct pack_list *opl;
+
+	while (l) {
+		old = l;
+		while (l->pl) {
+			opl = l->pl;
+			l->pl = opl->next;
+			free(opl);
+		}
+		l = l->next;
+		free(old);
+	}
+}
+
 /* all the permutations have to be free()d at the same time,
  * since they refer to each other
  */
-struct pll * get_all_permutations(struct pack_list *list)
+static struct pll * get_permutations(struct pack_list *list, int n)
 {
-	struct pll *subset, *pll, *new_pll = NULL; /*silence warning*/
+	struct pll *subset, *ret = NULL, *new_pll = NULL, *pll;
 
-	if (list == NULL)
+	if (list == NULL || pack_list_size(list) < n || n == 0)
 		return NULL;
 
-	if (list->next == NULL) {
-		new_pll = xmalloc(sizeof(struct pll));
-		new_pll->next = NULL;
-		new_pll->pl = list;
-		return new_pll;
+	if (n == 1) {
+		while (list) {
+			new_pll = xmalloc(sizeof(pll));
+			new_pll->pl = NULL;
+			pack_list_insert(&new_pll->pl, list);
+			new_pll->next = ret;
+			ret = new_pll;
+			list = list->next;
+		}
+		return ret;
 	}
 
-	pll = subset = get_all_permutations(list->next);
-	while (pll) {
-		new_pll = xmalloc(sizeof(struct pll));
-		new_pll->next = pll->next;
-		pll->next = new_pll;
-
-		new_pll->pl = xmalloc(sizeof(struct pack_list));
-		memcpy(new_pll->pl, list, sizeof(struct pack_list));
-		new_pll->pl->next = pll->pl;
-
-		pll = new_pll->next;
+	while (list->next) {
+		subset = get_permutations(list->next, n - 1);
+		while (subset) {
+			new_pll = xmalloc(sizeof(pll));
+			new_pll->pl = subset->pl;
+			pack_list_insert(&new_pll->pl, list);
+			new_pll->next = ret;
+			ret = new_pll;
+			subset = subset->next;
+		}
+		list = list->next;
 	}
-	/* add ourself to the end */
-	new_pll->next = xmalloc(sizeof(struct pll));
-	new_pll->next->pl = xmalloc(sizeof(struct pack_list));
-	new_pll->next->next = NULL;
-	memcpy(new_pll->next->pl, list, sizeof(struct pack_list));
-	new_pll->next->pl->next = NULL;
-
-	return subset;
+	return ret;
 }
 
-int is_superset(struct pack_list *pl, struct llist *list)
+static int is_superset(struct pack_list *pl, struct llist *list)
 {
-	struct llist *diff, *old;
+	struct llist *diff;
 
 	diff = llist_copy(list);
 
 	while (pl) {
-		old = diff;
-		diff = llist_sorted_difference(diff->front,
-					       pl->all_objects->front);
-		llist_free(old);
+		llist_sorted_difference_inplace(diff, pl->all_objects);
 		if (diff->size == 0) { /* we're done */
 			llist_free(diff);
 			return 1;
@@ -313,7 +347,7 @@ int is_superset(struct pack_list *pl, struct llist *list)
 	return 0;
 }
 
-size_t sizeof_union(struct packed_git *p1, struct packed_git *p2)
+static size_t sizeof_union(struct packed_git *p1, struct packed_git *p2)
 {
 	size_t ret = 0;
 	int p1_off, p2_off;
@@ -344,10 +378,14 @@ size_t sizeof_union(struct packed_git *p1, struct packed_git *p2)
 }
 
 /* another O(n^2) function ... */
-size_t get_pack_redundancy(struct pack_list *pl)
+static size_t get_pack_redundancy(struct pack_list *pl)
 {
 	struct pack_list *subset;
 	size_t ret = 0;
+
+	if (pl == NULL)
+		return 0;
+
 	while ((subset = pl->next)) {
 		while(subset) {
 			ret += sizeof_union(pl->pack, subset->pack);
@@ -358,7 +396,7 @@ size_t get_pack_redundancy(struct pack_list *pl)
 	return ret;
 }
 
-inline size_t pack_set_bytecount(struct pack_list *pl)
+static inline size_t pack_set_bytecount(struct pack_list *pl)
 {
 	size_t ret = 0;
 	while (pl) {
@@ -369,15 +407,16 @@ inline size_t pack_set_bytecount(struct pack_list *pl)
 	return ret;
 }
 
-void minimize(struct pack_list **min)
+static void minimize(struct pack_list **min)
 {
 	struct pack_list *pl, *unique = NULL,
 		*non_unique = NULL, *min_perm = NULL;
 	struct pll *perm, *perm_all, *perm_ok = NULL, *new_perm;
-	struct llist *missing, *old;
+	struct llist *missing;
 	size_t min_perm_size = (size_t)-1, perm_size;
+	int n;
 
-	pl = pack_list;
+	pl = local_packs;
 	while (pl) {
 		if(pl->unique_objects->size)
 			pack_list_insert(&unique, pl);
@@ -389,30 +428,32 @@ void minimize(struct pack_list **min)
 	missing = llist_copy(all_objects);
 	pl = unique;
 	while (pl) {
-		old = missing;
-		missing = llist_sorted_difference(missing->front,
-						  pl->all_objects->front);
-		llist_free(old);
+		llist_sorted_difference_inplace(missing, pl->all_objects);
 		pl = pl->next;
 	}
 
+	/* return if there are no objects missing from the unique set */
 	if (missing->size == 0) {
 		*min = unique;
 		return;
 	}
 
 	/* find the permutations which contain all missing objects */
-	perm_all = perm = get_all_permutations(non_unique);
-	while (perm) {
-		if (is_superset(perm->pl, missing)) {
-			new_perm = xmalloc(sizeof(struct pll));
-			new_perm->pl = perm->pl;
-			new_perm->next = perm_ok;
-			perm_ok = new_perm;
+	for (n = 1; n <= pack_list_size(non_unique) && !perm_ok; n++) {
+		perm_all = perm = get_permutations(non_unique, n);
+		while (perm) {
+			if (is_superset(perm->pl, missing)) {
+				new_perm = xmalloc(sizeof(struct pll));
+				memcpy(new_perm, perm, sizeof(struct pll));
+				new_perm->next = perm_ok;
+				perm_ok = new_perm;
+			}
+			perm = perm->next;
 		}
-		perm = perm->next;
+		if (perm_ok)
+			break;
+		pll_free(perm_all);
 	}
-	
 	if (perm_ok == NULL)
 		die("Internal error: No complete sets found!\n");
 
@@ -435,16 +476,14 @@ void minimize(struct pack_list **min)
 	}
 }
 
-void load_all_objects()
+static void load_all_objects(void)
 {
-	struct pack_list *pl = pack_list;
+	struct pack_list *pl = local_packs;
 	struct llist_item *hint, *l;
-	int i;
 
 	llist_init(&all_objects);
 
 	while (pl) {
-		i = 0;
 		hint = NULL;
 		l = pl->all_objects->front;
 		while (l) {
@@ -454,26 +493,52 @@ void load_all_objects()
 		}
 		pl = pl->next;
 	}
+	/* remove objects present in remote packs */
+	pl = altodb_packs;
+	while (pl) {
+		llist_sorted_difference_inplace(all_objects, pl->all_objects);
+		pl = pl->next;
+	}
 }
 
 /* this scales like O(n^2) */
-void cmp_packs()
+static void cmp_local_packs(void)
 {
-	struct pack_list *subset, *curr = pack_list;
+	struct pack_list *subset, *pl = local_packs;
 
-	while ((subset = curr)) {
+	while ((subset = pl)) {
 		while((subset = subset->next))
-			cmp_two_packs(curr, subset);
-		curr = curr->next;
+			cmp_two_packs(pl, subset);
+		pl = pl->next;
 	}
 }
 
-struct pack_list * add_pack(struct packed_git *p)
+static void scan_alt_odb_packs(void)
+{
+	struct pack_list *local, *alt;
+
+	alt = altodb_packs;
+	while (alt) {
+		local = local_packs;
+		while (local) {
+			llist_sorted_difference_inplace(local->unique_objects,
+							alt->all_objects);
+			local = local->next;
+		}
+		llist_sorted_difference_inplace(all_objects, alt->all_objects);
+		alt = alt->next;
+	}
+}
+
+static struct pack_list * add_pack(struct packed_git *p)
 {
 	struct pack_list l;
 	size_t off;
 	void *base;
 
+	if (!p->pack_local && !(alt_odb || verbose))
+		return NULL;
+
 	l.pack = p;
 	llist_init(&l.all_objects);
 
@@ -481,14 +546,17 @@ struct pack_list * add_pack(struct packed_git *p)
 	base = (void *)p->index_base;
 	while (off <= p->index_size - 3 * 20) {
 		llist_insert_back(l.all_objects, base + off);
-		off+=24;
+		off += 24;
 	}
 	/* this list will be pruned in cmp_two_packs later */
 	l.unique_objects = llist_copy(l.all_objects);
-	return pack_list_insert(&pack_list, &l);
+	if (p->pack_local)
+		return pack_list_insert(&local_packs, &l);
+	else
+		return pack_list_insert(&altodb_packs, &l);
 }
 
-struct pack_list * add_pack_file(char *filename)
+static struct pack_list * add_pack_file(char *filename)
 {
 	struct packed_git *p = packed_git;
 
@@ -497,20 +565,18 @@ struct pack_list * add_pack_file(char *filename)
 
 	while (p) {
 		if (strstr(p->pack_name, filename))
-			/* this will silently ignore packs in alt-odb */
 			return add_pack(p);
 		p = p->next;
 	}
 	die("Filename %s not found in packed_git\n", filename);
 }
 
-void load_all()
+static void load_all(void)
 {
 	struct packed_git *p = packed_git;
 
 	while (p) {
-		if (p->pack_local) /* ignore alt-odb for now */
-			add_pack(p);
+		add_pack(p);
 		p = p->next;
 	}
 }
@@ -519,6 +585,11 @@ int main(int argc, char **argv)
 {
 	int i;
 	struct pack_list *min, *red, *pl;
+	struct llist *ignore;
+	unsigned char *sha1;
+	char buf[42]; /* 40 byte sha1 + \n + \0 */
+
+	setup_git_directory();
 
 	for (i = 1; i < argc; i++) {
 		const char *arg = argv[i];
@@ -526,14 +597,18 @@ int main(int argc, char **argv)
 			i++;
 			break;
 		}
-		if(!strcmp(arg, "-a")) {
-			all = 1;
+		if(!strcmp(arg, "--all")) {
+			load_all_packs = 1;
 			continue;
 		}
-		if(!strcmp(arg, "-v")) {
+		if(!strcmp(arg, "--verbose")) {
 			verbose = 1;
 			continue;
 		}
+		if(!strcmp(arg, "--alt-odb")) {
+			alt_odb = 1;
+			continue;
+		}
 		if(*arg == '-')
 			usage(pack_redundant_usage);
 		else
@@ -542,38 +617,67 @@ int main(int argc, char **argv)
 
 	prepare_packed_git();
 
-	if(all)
+	if (load_all_packs)
 		load_all();
 	else
 		while (*(argv + i) != NULL)
 			add_pack_file(*(argv + i++));
 
-	if (pack_list == NULL)
+	if (local_packs == NULL)
 		die("Zero packs found!\n");
 
-	cmp_packs();
-
 	load_all_objects();
 
+	cmp_local_packs();
+	if (alt_odb)
+		scan_alt_odb_packs();
+
+	/* ignore objects given on stdin */
+	llist_init(&ignore);
+	if (!isatty(0)) {
+		while (fgets(buf, sizeof(buf), stdin)) {
+			sha1 = xmalloc(20);
+			if (get_sha1_hex(buf, sha1))
+				die("Bad sha1 on stdin: %s", buf);
+			llist_insert_sorted_unique(ignore, sha1, NULL);
+		}
+	}
+	llist_sorted_difference_inplace(all_objects, ignore);
+	pl = local_packs;
+	while (pl) {
+		llist_sorted_difference_inplace(pl->unique_objects, ignore);
+		pl = pl->next;
+	}
+
 	minimize(&min);
+
 	if (verbose) {
+		fprintf(stderr, "There are %lu packs available in alt-odbs.\n",
+			(unsigned long)pack_list_size(altodb_packs));
 		fprintf(stderr, "The smallest (bytewise) set of packs is:\n");
 		pl = min;
 		while (pl) {
 			fprintf(stderr, "\t%s\n", pl->pack->pack_name);
 			pl = pl->next;
 		}
-		fprintf(stderr, "containing %ld duplicate objects "
-				"with a total size of %ldkb.\n",
-			get_pack_redundancy(min), pack_set_bytecount(min)/1024);
+		fprintf(stderr, "containing %lu duplicate objects "
+				"with a total size of %lukb.\n",
+			(unsigned long)get_pack_redundancy(min),
+			(unsigned long)pack_set_bytecount(min)/1024);
+		fprintf(stderr, "A total of %lu unique objects were considered.\n",
+			(unsigned long)all_objects->size);
 		fprintf(stderr, "Redundant packs (with indexes):\n");
 	}
-	pl = red = pack_list_difference(pack_list, min);
+	pl = red = pack_list_difference(local_packs, min);
 	while (pl) {
 		printf("%s\n%s\n",
-		       sha1_pack_index_name(pl->pack->sha1), pl->pack->pack_name);
+		       sha1_pack_index_name(pl->pack->sha1),
+		       pl->pack->pack_name);
 		pl = pl->next;
 	}
+	if (verbose)
+		fprintf(stderr, "%luMB of redundant packs in total.\n",
+			(unsigned long)pack_set_bytecount(red)/(1024*1024));
 
 	return 0;
 }