From: Junio C Hamano Date: Fri, 24 Feb 2006 09:56:38 +0000 (-0800) Subject: Merge branches 'jc/rev-list' and 'jc/pack-thin' into next X-Git-Tag: v1.3.0-rc1~54^2~35 X-Git-Url: http://git.tremily.us/?a=commitdiff_plain;h=7465ef5155cec830a34dcffdd8272d4beb243220;p=git.git Merge branches 'jc/rev-list' and 'jc/pack-thin' into next * jc/rev-list: rev-list --objects: use full pathname to help hashing. * jc/pack-thin: pack-objects: hash basename and direname a bit differently. pack-objects: allow "thin" packs to exceed depth limits pack-objects: use full pathname to help hashing with "thin" pack. --- 7465ef5155cec830a34dcffdd8272d4beb243220 diff --cc pack-objects.c index 8da8aabf9,c5a5e6160,095bcb828..be7a2008c --- a/pack-objects.c +++ b/pack-objects.c @@@@ -3,11 -3,9 -3,10 +3,11 @@@@ #include "delta.h" #include "pack.h" #include "csum-file.h" + #include "diff.h" #include ++#include - static const char pack_usage[] = "git-pack-objects [-q] [--non-empty] [--local] [--incremental] [--window=N] [--depth=N] {--stdout | base-name} < object-list"; + static const char pack_usage[] = "git-pack-objects [-q] [--no-reuse-delta] [--non-empty] [--local] [--incremental] [--window=N] [--depth=N] {--stdout | base-name} < object-list"; struct object_entry { unsigned char sha1[20]; @@@@ -61,138 -28,6 -60,137 +61,138 @@@@ static int nr_objects = 0, nr_alloc = 0 static const char *base_name; static unsigned char pack_file_sha1[20]; static int progress = 1; ++static volatile int progress_update = 0; + + /* + * The object names in objects array are hashed with this hashtable, + * to help looking up the entry by object name. Binary search from + * sorted_by_sha is also possible but this was easier to code and faster. + * This hashtable is built after all the objects are seen. + */ + static int *object_ix = NULL; + static int object_ix_hashsz = 0; + + /* + * Pack index for existing packs give us easy access to the offsets into + * corresponding pack file where each object's data starts, but the entries + * do not store the size of the compressed representation (uncompressed + * size is easily available by examining the pack entry header). We build + * a hashtable of existing packs (pack_revindex), and keep reverse index + * here -- pack index file is sorted by object name mapping to offset; this + * pack_revindex[].revindex array is an ordered list of offsets, so if you + * know the offset of an object, next offset is where its packed + * representation ends. + */ + struct pack_revindex { + struct packed_git *p; + unsigned long *revindex; + } *pack_revindex = NULL; + static int pack_revindex_hashsz = 0; + + /* + * stats + */ + static int written = 0; + static int written_delta = 0; + static int reused = 0; + static int reused_delta = 0; + + static int pack_revindex_ix(struct packed_git *p) + { + unsigned int ui = (unsigned int) p; + int i; + + ui = ui ^ (ui >> 16); /* defeat structure alignment */ + i = (int)(ui % pack_revindex_hashsz); + while (pack_revindex[i].p) { + if (pack_revindex[i].p == p) + return i; + if (++i == pack_revindex_hashsz) + i = 0; + } + return -1 - i; + } + + static void prepare_pack_ix(void) + { + int num; + struct packed_git *p; + for (num = 0, p = packed_git; p; p = p->next) + num++; + if (!num) + return; + pack_revindex_hashsz = num * 11; + pack_revindex = xcalloc(sizeof(*pack_revindex), pack_revindex_hashsz); + for (p = packed_git; p; p = p->next) { + num = pack_revindex_ix(p); + num = - 1 - num; + pack_revindex[num].p = p; + } + /* revindex elements are lazily initialized */ + } + + static int cmp_offset(const void *a_, const void *b_) + { + unsigned long a = *(unsigned long *) a_; + unsigned long b = *(unsigned long *) b_; + if (a < b) + return -1; + else if (a == b) + return 0; + else + return 1; + } + + /* + * Ordered list of offsets of objects in the pack. + */ + static void prepare_pack_revindex(struct pack_revindex *rix) + { + struct packed_git *p = rix->p; + int num_ent = num_packed_objects(p); + int i; + void *index = p->index_base + 256; + + rix->revindex = xmalloc(sizeof(unsigned long) * (num_ent + 1)); + for (i = 0; i < num_ent; i++) { + long hl = *((long *)(index + 24 * i)); + rix->revindex[i] = ntohl(hl); + } + /* This knows the pack format -- the 20-byte trailer + * follows immediately after the last object data. + */ + rix->revindex[num_ent] = p->pack_size - 20; + qsort(rix->revindex, num_ent, sizeof(unsigned long), cmp_offset); + } + + static unsigned long find_packed_object_size(struct packed_git *p, + unsigned long ofs) + { + int num; + int lo, hi; + struct pack_revindex *rix; + unsigned long *revindex; + num = pack_revindex_ix(p); + if (num < 0) + die("internal error: pack revindex uninitialized"); + rix = &pack_revindex[num]; + if (!rix->revindex) + prepare_pack_revindex(rix); + revindex = rix->revindex; + lo = 0; + hi = num_packed_objects(p) + 1; + do { + int mi = (lo + hi) / 2; + if (revindex[mi] == ofs) { + return revindex[mi+1] - ofs; + } + else if (ofs < revindex[mi]) + hi = mi; + else + lo = mi + 1; + } while (lo < hi); + die("internal error: pack revindex corrupt"); + } static void *delta_against(void *buf, unsigned long size, struct object_entry *entry) { @@@@ -335,43 -132,24 -333,22 +335,43 @@@@ static void write_pack_file(void int i; struct sha1file *f; unsigned long offset; - unsigned long mb; struct pack_header hdr; ++ unsigned last_percent = 999; ++ int do_progress = 0; if (!base_name) f = sha1fd(1, ""); -- else - f = sha1create("%s-%s.%s", base_name, sha1_to_hex(object_list_sha1), "pack"); ++ else { + f = sha1create("%s-%s.%s", base_name, + sha1_to_hex(object_list_sha1), "pack"); ++ do_progress = progress; ++ } ++ if (do_progress) ++ fprintf(stderr, "Writing %d objects.\n", nr_result); ++ hdr.hdr_signature = htonl(PACK_SIGNATURE); hdr.hdr_version = htonl(PACK_VERSION); - hdr.hdr_entries = htonl(nr_objects); + hdr.hdr_entries = htonl(nr_result); sha1write(f, &hdr, sizeof(hdr)); offset = sizeof(hdr); -- for (i = 0; i < nr_objects; i++) ++ if (!nr_result) ++ goto done; ++ for (i = 0; i < nr_objects; i++) { offset = write_one(f, objects + i, offset); -- ++ if (do_progress) { ++ unsigned percent = written * 100 / nr_result; ++ if (progress_update || percent != last_percent) { ++ fprintf(stderr, "%4u%% (%u/%u) done\r", ++ percent, written, nr_result); ++ progress_update = 0; ++ last_percent = percent; ++ } ++ } ++ } ++ if (do_progress) ++ fputc('\n', stderr); ++ done: sha1close(f, pack_file_sha1, 1); - mb = offset >> 20; - offset &= 0xfffff; } static void write_index_file(void) @@@@ -415,66 -192,17 -392,125 +415,125 @@@@ sha1close(f, NULL, 1); } - static int add_object_entry(unsigned char *sha1, unsigned int hash) + static int locate_object_entry_hash(const unsigned char *sha1) + { + int i; + unsigned int ui; + memcpy(&ui, sha1, sizeof(unsigned int)); + i = ui % object_ix_hashsz; + while (0 < object_ix[i]) { + if (!memcmp(sha1, objects[object_ix[i]-1].sha1, 20)) + return i; + if (++i == object_ix_hashsz) + i = 0; + } + return -1 - i; + } + + static struct object_entry *locate_object_entry(const unsigned char *sha1) + { + int i; + + if (!object_ix_hashsz) + return NULL; + + i = locate_object_entry_hash(sha1); + if (0 <= i) + return &objects[object_ix[i]-1]; + return NULL; + } + + static void rehash_objects(void) + { + int i; + struct object_entry *oe; + + object_ix_hashsz = nr_objects * 3; + if (object_ix_hashsz < 1024) + object_ix_hashsz = 1024; + object_ix = xrealloc(object_ix, sizeof(int) * object_ix_hashsz); + object_ix = memset(object_ix, 0, sizeof(int) * object_ix_hashsz); + for (i = 0, oe = objects; i < nr_objects; i++, oe++) { + int ix = locate_object_entry_hash(oe->sha1); + if (0 <= ix) + continue; + ix = -1 - ix; + object_ix[ix] = i + 1; + } + } + - static int add_object_entry(const unsigned char *sha1, const char *name, int exclude) ++ struct name_path { ++ struct name_path *up; ++ const char *elem; ++ int len; ++ }; ++ ++ #define DIRBITS 12 ++ ++ static unsigned name_hash(struct name_path *path, const char *name) ++ { ++ struct name_path *p = path; ++ const char *n = name + strlen(name); ++ unsigned hash = 0, name_hash = 0, name_done = 0; ++ ++ if (n != name && n[-1] == '\n') ++ n--; ++ while (name <= --n) { ++ unsigned char c = *n; ++ if (c == '/' && !name_done) { ++ name_hash = hash; ++ name_done = 1; ++ hash = 0; ++ } ++ hash = hash * 11 + c; ++ } ++ if (!name_done) { ++ name_hash = hash; ++ hash = 0; ++ } ++ for (p = path; p; p = p->up) { ++ hash = hash * 11 + '/'; ++ n = p->elem + p->len; ++ while (p->elem <= --n) { ++ unsigned char c = *n; ++ hash = hash * 11 + c; ++ } ++ } ++ /* ++ * Make sure "Makefile" and "t/Makefile" are hashed separately ++ * but close enough. ++ */ ++ hash = (name_hash<up) { ++ fputc('/', stderr); ++ n = p->elem + p->len; ++ while (p->elem <= --n) ++ fputc(*n, stderr); ++ } ++ fprintf(stderr, "\t%08x\n", hash); ++ } ++ return hash; ++ } ++ ++ static int add_object_entry(const unsigned char *sha1, unsigned hash, int exclude) { - unsigned int hash = 0; unsigned int idx = nr_objects; struct object_entry *entry; + struct packed_git *p; + unsigned int found_offset = 0; + struct packed_git *found_pack = NULL; + int ix, status = 0; - if (incremental || local) { - struct packed_git *p; - + if (!exclude) { for (p = packed_git; p; p = p->next) { struct pack_entry e; - if (find_pack_entry_one(sha1, &e, p)) { if (incremental) return 0; @@@@ -487,16 -211,7 -523,9 +546,9 @@@@ } } } + if ((entry = locate_object_entry(sha1)) != NULL) + goto already_added; - while (*name) { - unsigned char c = *name++; - if (isspace(c)) - continue; - hash = hash * 11 + c; - } - if (idx >= nr_alloc) { unsigned int needed = (idx + 1024) * 3 / 2; objects = xrealloc(objects, needed * sizeof(*entry)); @@@@ -507,76 -221,8 -536,79 +559,83 @@@@ memset(entry, 0, sizeof(*entry)); memcpy(entry->sha1, sha1, 20); entry->hash = hash; - nr_objects = idx+1; - return 1; + + if (object_ix_hashsz * 3 <= nr_objects * 4) + rehash_objects(); + else { + ix = locate_object_entry_hash(entry->sha1); + if (0 <= ix) + die("internal error in object hashing."); + object_ix[-1 - ix] = idx + 1; + } + status = 1; + + already_added: ++ if (progress_update) { ++ fprintf(stderr, "Counting objects...%d\r", nr_objects); ++ progress_update = 0; ++ } + if (exclude) + entry->preferred_base = 1; + else { + if (found_pack) { + entry->in_pack = found_pack; + entry->in_pack_offset = found_offset; + } + } + return status; + } + - static void add_pbase_tree(struct tree_desc *tree) ++ static void add_pbase_tree(struct tree_desc *tree, struct name_path *up) + { + while (tree->size) { + const unsigned char *sha1; + const char *name; - unsigned mode; ++ unsigned mode, hash; + unsigned long size; + char type[20]; + + sha1 = tree_entry_extract(tree, &name, &mode); + update_tree_entry(tree); + if (!has_sha1_file(sha1)) + continue; + if (sha1_object_info(sha1, type, &size)) + continue; + - if (!add_object_entry(sha1, name, 1)) ++ hash = name_hash(up, name); ++ if (!add_object_entry(sha1, hash, 1)) + continue; + + if (!strcmp(type, "tree")) { + struct tree_desc sub; + void *elem; ++ struct name_path me; ++ + elem = read_sha1_file(sha1, type, &sub.size); + sub.buf = elem; + if (sub.buf) { - add_pbase_tree(&sub); ++ me.up = up; ++ me.elem = name; ++ me.len = strlen(name); ++ add_pbase_tree(&sub, &me); + free(elem); + } + } + } + } + + static void add_preferred_base(unsigned char *sha1) + { + struct tree_desc tree; + void *elem; ++ + elem = read_object_with_reference(sha1, "tree", &tree.size, NULL); + tree.buf = elem; + if (!tree.buf) + return; - if (add_object_entry(sha1, "", 1)) - add_pbase_tree(&tree); ++ if (add_object_entry(sha1, name_hash(NULL, ""), 1)) ++ add_pbase_tree(&tree, NULL); + free(elem); } static void check_object(struct object_entry *entry) @@@@ -657,15 -250,10 -689,28 +716,28 @@@@ static unsigned int check_delta_limit(s static void get_object_details(void) { int i; - struct object_entry *entry = objects; + struct object_entry *entry; - for (i = 0; i < nr_objects; i++) - check_object(entry++); + prepare_pack_ix(); + for (i = 0, entry = objects; i < nr_objects; i++, entry++) + check_object(entry); - for (i = 0, entry = objects; i < nr_objects; i++, entry++) - if (!entry->delta && entry->delta_child) - entry->delta_limit = - check_delta_limit(entry, 1); ++ ++ if (nr_objects == nr_result) { ++ /* ++ * Depth of objects that depend on the entry -- this ++ * is subtracted from depth-max to break too deep ++ * delta chain because of delta data reusing. ++ * However, we loosen this restriction when we know we ++ * are creating a thin pack -- it will have to be ++ * expanded on the other end anyway, so do not ++ * artificially cut the delta chain and let it go as ++ * deep as it wants. ++ */ ++ for (i = 0, entry = objects; i < nr_objects; i++, entry++) ++ if (!entry->delta && entry->delta_child) ++ entry->delta_limit = ++ check_delta_limit(entry, 1); ++ } } typedef int (*entry_sort_t)(const struct object_entry *, const struct object_entry *); @@@@ -762,27 -326,12 -807,28 +834,27 @@@@ static int try_delta(struct unpacked *c if (cur_entry->type != old_entry->type) return -1; - size = cur_entry->size; - if (size < 50) + /* We do not compute delta to *create* objects we are not + * going to pack. + */ + if (cur_entry->preferred_base) return -1; + + /* If the current object is at pack edge, take the depth the + * objects that depend on the current object into account -- + * otherwise they would become too deep. + */ + if (cur_entry->delta_child) { + if (max_depth <= cur_entry->delta_limit) + return 0; + max_depth -= cur_entry->delta_limit; + } + + size = cur_entry->size; - if (size < 50) - return -1; oldsize = old_entry->size; sizediff = oldsize > size ? oldsize - size : size - oldsize; -- if (sizediff > size / 8) ++ ++ if (size < 50) return -1; if (old_entry->depth >= max_depth) return 0; @@@@ -857,25 -378,10 -895,17 +929,25 @@@@ static void find_deltas(struct object_e char type[10]; int j; -- if (progress && i <= eye_candy) { -- eye_candy -= nr_objects / 20; -- fputc('.', stderr); ++ if (!entry->preferred_base) ++ processed++; ++ ++ if (progress) { ++ unsigned percent = processed * 100 / nr_result; ++ if (percent != last_percent || progress_update) { ++ fprintf(stderr, "%4u%% (%u/%u) done\r", ++ percent, processed, nr_result); ++ progress_update = 0; ++ last_percent = percent; ++ } } + + if (entry->delta) + /* This happens if we decided to reuse existing + * delta from a pack. "!no_reuse_delta &&" is implied. + */ + continue; + free(n->data); n->entry = entry; n->data = read_sha1_file(entry->sha1, type, &size); @@@@ -909,7 -411,10 -936,12 +981,7 @@@@ static void prepare_pack(int window, int depth) { - if (progress) - fprintf(stderr, "Packing %d objects", nr_result); get_object_details(); - -- if (progress) - fprintf(stderr, "Packing %d objects", nr_objects); - fputc('.', stderr); - sorted_by_type = create_sorted_list(type_size_sort); if (window && depth) find_deltas(sorted_by_type, window+1, depth); @@@@ -1032,40 -539,52 -1071,49 +1104,40 @@@@ int main(int argc, char **argv usage(pack_usage); prepare_packed_git(); ++ if (progress) { ++ struct itimerval v; ++ v.it_interval.tv_sec = 1; ++ v.it_interval.tv_usec = 0; ++ v.it_value = v.it_interval; ++ signal(SIGALRM, progress_interval); ++ setitimer(ITIMER_REAL, &v, NULL); fprintf(stderr, "Generating pack...\n"); -- gettimeofday(&prev_tv, NULL); } ++ while (fgets(line, sizeof(line), stdin) != NULL) { - unsigned int hash; - char *p; unsigned char sha1[20]; -- if (progress && (eye_candy <= nr_objects)) { -- fprintf(stderr, "Counting objects...%d\r", nr_objects); -- if (eye_candy && (50 <= eye_candy_incr)) { -- struct timeval tv; -- int time_diff; -- gettimeofday(&tv, NULL); -- time_diff = (tv.tv_sec - prev_tv.tv_sec); -- time_diff <<= 10; -- time_diff += (tv.tv_usec - prev_tv.tv_usec); -- if ((1 << 9) < time_diff) -- eye_candy_incr += 50; -- else if (50 < eye_candy_incr) -- eye_candy_incr -= 50; -- } -- eye_candy += eye_candy_incr; - } + if (line[0] == '-') { + if (get_sha1_hex(line+1, sha1)) + die("expected edge sha1, got garbage:\n %s", + line+1); + add_preferred_base(sha1); + continue; } if (get_sha1_hex(line, sha1)) die("expected sha1, got garbage:\n %s", line); - add_object_entry(sha1, line+40, 0); - hash = 0; - p = line+40; - while (*p) { - unsigned char c = *p++; - if (isspace(c)) - continue; - hash = hash * 11 + c; - } - add_object_entry(sha1, hash); ++ add_object_entry(sha1, name_hash(NULL, line+41), 0); } if (progress) fprintf(stderr, "Done counting %d objects.\n", nr_objects); -- if (non_empty && !nr_objects) ++ sorted_by_sha = create_final_object_list(); ++ if (non_empty && !nr_result) return 0; - sorted_by_sha = create_sorted_list(sha1_sort); - sorted_by_sha = create_final_object_list(); SHA1_Init(&ctx); list = sorted_by_sha; - for (i = 0; i < nr_objects; i++) { + for (i = 0; i < nr_result; i++) { struct object_entry *entry = *list++; SHA1_Update(&ctx, entry->sha1, 20); }