* nonzero if already written.
*/
unsigned int depth; /* delta depth */
- unsigned int delta_limit; /* base adjustment for in-pack delta */
unsigned int hash; /* name hint hash */
enum object_type type;
enum object_type in_pack_type; /* could be delta */
* be used as the base objectto delta huge
* objects against.
*/
+ uint32_t crc32; /* crc of raw pack data for this object */
};
/*
static int incremental;
static int allow_ofs_delta;
-static struct object_entry **sorted_by_sha, **sorted_by_type;
+static struct object_entry **sorted_by_sha;
static struct object_entry *objects;
static uint32_t nr_objects, nr_alloc, nr_result;
static const char *base_name;
static void prepare_pack_revindex(struct pack_revindex *rix)
{
struct packed_git *p = rix->p;
- int num_ent = num_packed_objects(p);
+ int num_ent = p->num_objects;
int i;
const char *index = p->index_data;
- index += 4 * 256;
rix->revindex = xmalloc(sizeof(*rix->revindex) * (num_ent + 1));
- for (i = 0; i < num_ent; i++) {
- uint32_t hl = *((uint32_t *)(index + 24 * i));
- rix->revindex[i].offset = ntohl(hl);
- rix->revindex[i].nr = i;
+ index += 4 * 256;
+
+ if (p->index_version > 1) {
+ const uint32_t *off_32 =
+ (uint32_t *)(index + 8 + p->num_objects * (20 + 4));
+ const uint32_t *off_64 = off_32 + p->num_objects;
+ for (i = 0; i < num_ent; i++) {
+ uint32_t off = ntohl(*off_32++);
+ if (!(off & 0x80000000)) {
+ rix->revindex[i].offset = off;
+ } else {
+ rix->revindex[i].offset =
+ ((uint64_t)ntohl(*off_64++)) << 32;
+ rix->revindex[i].offset |=
+ ntohl(*off_64++);
+ }
+ rix->revindex[i].nr = i;
+ }
+ } else {
+ for (i = 0; i < num_ent; i++) {
+ uint32_t hl = *((uint32_t *)(index + 24 * i));
+ rix->revindex[i].offset = ntohl(hl);
+ rix->revindex[i].nr = i;
+ }
}
+
/* This knows the pack format -- the 20-byte trailer
* follows immediately after the last object data.
*/
prepare_pack_revindex(rix);
revindex = rix->revindex;
lo = 0;
- hi = num_packed_objects(p) + 1;
+ hi = p->num_objects + 1;
do {
int mi = (lo + hi) / 2;
if (revindex[mi].offset == ofs) {
die("internal error: pack revindex corrupt");
}
-static off_t find_packed_object_size(struct packed_git *p, off_t ofs)
-{
- struct revindex_entry *entry = find_packed_object(p, ofs);
- return entry[1].offset - ofs;
-}
-
static const unsigned char *find_packed_object_name(struct packed_git *p,
off_t ofs)
{
struct revindex_entry *entry = find_packed_object(p, ofs);
- return ((unsigned char *)p->index_data) + 4 * 256 + 24 * entry->nr + 4;
+ return nth_packed_object_sha1(p, entry->nr);
}
static void *delta_against(void *buf, unsigned long size, struct object_entry *entry)
stream.total_in == len) ? 0 : -1;
}
+static int check_pack_crc(struct packed_git *p, struct pack_window **w_curs,
+ off_t offset, off_t len, unsigned int nr)
+{
+ const uint32_t *index_crc;
+ uint32_t data_crc = crc32(0, Z_NULL, 0);
+
+ do {
+ unsigned int avail;
+ void *data = use_pack(p, w_curs, offset, &avail);
+ if (avail > len)
+ avail = len;
+ data_crc = crc32(data_crc, data, avail);
+ offset += avail;
+ len -= avail;
+ } while (len);
+
+ index_crc = p->index_data;
+ index_crc += 2 + 256 + p->num_objects * (20/4) + nr;
+
+ return data_crc != ntohl(*index_crc);
+}
+
static void copy_pack_data(struct sha1file *f,
struct packed_git *p,
struct pack_window **w_curs,
return check_loose_inflate(map, mapsize, size);
}
-static off_t write_object(struct sha1file *f,
+static unsigned long write_object(struct sha1file *f,
struct object_entry *entry)
{
unsigned long size;
enum object_type obj_type;
int to_reuse = 0;
+ if (!pack_to_stdout)
+ crc32_begin(f);
+
obj_type = entry->type;
if (! entry->in_pack)
to_reuse = 0; /* can't reuse what we don't have */
else {
struct packed_git *p = entry->in_pack;
struct pack_window *w_curs = NULL;
+ struct revindex_entry *revidx;
off_t offset;
if (entry->delta) {
hdrlen += 20;
}
- offset = entry->in_pack_offset + entry->in_pack_header_size;
- datalen = find_packed_object_size(p, entry->in_pack_offset)
- - entry->in_pack_header_size;
- if (!pack_to_stdout && check_pack_inflate(p, &w_curs,
- offset, datalen, entry->size))
- die("corrupt delta in pack %s", sha1_to_hex(entry->sha1));
+ offset = entry->in_pack_offset;
+ revidx = find_packed_object(p, offset);
+ datalen = revidx[1].offset - offset;
+ if (!pack_to_stdout && p->index_version > 1 &&
+ check_pack_crc(p, &w_curs, offset, datalen, revidx->nr))
+ die("bad packed object CRC for %s", sha1_to_hex(entry->sha1));
+ offset += entry->in_pack_header_size;
+ datalen -= entry->in_pack_header_size;
+ if (!pack_to_stdout && p->index_version == 1 &&
+ check_pack_inflate(p, &w_curs, offset, datalen, entry->size))
+ die("corrupt packed object for %s", sha1_to_hex(entry->sha1));
copy_pack_data(f, p, &w_curs, offset, datalen);
unuse_pack(&w_curs);
reused++;
if (entry->delta)
written_delta++;
written++;
+ if (!pack_to_stdout)
+ entry->crc32 = crc32_end(f);
return hdrlen + datalen;
}
struct object_entry *e,
off_t offset)
{
+ unsigned long size;
+
+ /* offset is non zero if object is written already. */
if (e->offset || e->preferred_base)
- /* offset starts from header size and cannot be zero
- * if it is written already.
- */
return offset;
- /* if we are deltified, write out its base object first. */
+
+ /* if we are deltified, write out base object first. */
if (e->delta)
offset = write_one(f, e->delta, offset);
+
e->offset = offset;
- return offset + write_object(f, e);
+ size = write_object(f, e);
+
+ /* make sure off_t is sufficiently large not to wrap */
+ if (offset > offset + size)
+ die("pack too large for current definition of off_t");
+ return offset + size;
}
-static void write_pack_file(void)
+static off_t write_pack_file(void)
{
uint32_t i;
struct sha1file *f;
- off_t offset;
+ off_t offset, last_obj_offset = 0;
struct pack_header hdr;
unsigned last_percent = 999;
int do_progress = progress;
if (!nr_result)
goto done;
for (i = 0; i < nr_objects; i++) {
+ last_obj_offset = offset;
offset = write_one(f, objects + i, offset);
if (do_progress) {
unsigned percent = written * 100 / nr_result;
if (written != nr_result)
die("wrote %u objects while expecting %u", written, nr_result);
sha1close(f, pack_file_sha1, 1);
+
+ return last_obj_offset;
}
-static void write_index_file(void)
+static uint32_t index_default_version = 1;
+static uint32_t index_off32_limit = 0x7fffffff;
+
+static void write_index_file(off_t last_obj_offset)
{
uint32_t i;
struct sha1file *f = sha1create("%s-%s.%s", base_name,
struct object_entry **list = sorted_by_sha;
struct object_entry **last = list + nr_result;
uint32_t array[256];
+ uint32_t index_version;
+
+ /* if last object's offset is >= 2^31 we should use index V2 */
+ index_version = (last_obj_offset >> 31) ? 2 : index_default_version;
+
+ /* index versions 2 and above need a header */
+ if (index_version >= 2) {
+ struct pack_idx_header hdr;
+ hdr.idx_signature = htonl(PACK_IDX_SIGNATURE);
+ hdr.idx_version = htonl(index_version);
+ sha1write(f, &hdr, sizeof(hdr));
+ }
/*
* Write the first-level table (the list is sorted,
list = sorted_by_sha;
for (i = 0; i < nr_result; i++) {
struct object_entry *entry = *list++;
- uint32_t offset = htonl(entry->offset);
- sha1write(f, &offset, 4);
+ if (index_version < 2) {
+ uint32_t offset = htonl(entry->offset);
+ sha1write(f, &offset, 4);
+ }
sha1write(f, entry->sha1, 20);
}
+
+ if (index_version >= 2) {
+ unsigned int nr_large_offset = 0;
+
+ /* write the crc32 table */
+ list = sorted_by_sha;
+ for (i = 0; i < nr_objects; i++) {
+ struct object_entry *entry = *list++;
+ uint32_t crc32_val = htonl(entry->crc32);
+ sha1write(f, &crc32_val, 4);
+ }
+
+ /* write the 32-bit offset table */
+ list = sorted_by_sha;
+ for (i = 0; i < nr_objects; i++) {
+ struct object_entry *entry = *list++;
+ uint32_t offset = (entry->offset <= index_off32_limit) ?
+ entry->offset : (0x80000000 | nr_large_offset++);
+ offset = htonl(offset);
+ sha1write(f, &offset, 4);
+ }
+
+ /* write the large offset table */
+ list = sorted_by_sha;
+ while (nr_large_offset) {
+ struct object_entry *entry = *list++;
+ uint64_t offset = entry->offset;
+ if (offset > index_off32_limit) {
+ uint32_t split[2];
+ split[0] = htonl(offset >> 32);
+ split[1] = htonl(offset & 0xffffffff);
+ sha1write(f, split, 8);
+ nr_large_offset--;
+ }
+ }
+ }
+
sha1write(f, pack_file_sha1, 20);
sha1close(f, NULL, 1);
}
static int add_object_entry(const unsigned char *sha1, unsigned hash, int exclude)
{
- uint32_t idx = nr_objects;
struct object_entry *entry;
- struct packed_git *p;
+ struct packed_git *p, *found_pack = NULL;
off_t found_offset = 0;
- struct packed_git *found_pack = NULL;
- int ix, status = 0;
+ int ix;
+
+ ix = nr_objects ? locate_object_entry_hash(sha1) : -1;
+ if (ix >= 0) {
+ if (exclude) {
+ entry = objects + object_ix[ix] - 1;
+ entry->preferred_base = 1;
+ }
+ return 0;
+ }
if (!exclude) {
for (p = packed_git; p; p = p->next) {
}
}
}
- if ((entry = locate_object_entry(sha1)) != NULL)
- goto already_added;
- if (idx >= nr_alloc) {
- nr_alloc = (idx + 1024) * 3 / 2;
+ if (nr_objects >= nr_alloc) {
+ nr_alloc = (nr_alloc + 1024) * 3 / 2;
objects = xrealloc(objects, nr_alloc * sizeof(*entry));
}
- entry = objects + idx;
- nr_objects = idx + 1;
+
+ entry = objects + nr_objects++;
memset(entry, 0, sizeof(*entry));
hashcpy(entry->sha1, sha1);
entry->hash = hash;
+ if (exclude)
+ entry->preferred_base = 1;
+ if (found_pack) {
+ entry->in_pack = found_pack;
+ entry->in_pack_offset = found_offset;
+ }
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;
+ else
+ object_ix[-1 - ix] = nr_objects;
- already_added:
if (progress_update) {
fprintf(stderr, "Counting objects...%u\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;
+
+ return 1;
}
struct pbase_tree_cache {
const char *fullname)
{
struct name_entry entry;
+ int cmp;
while (tree_entry(tree,&entry)) {
- unsigned long size;
- enum object_type type;
-
- if (tree_entry_len(entry.path, entry.sha1) != cmplen ||
- memcmp(entry.path, name, cmplen) ||
- !has_sha1_file(entry.sha1) ||
- (type = sha1_object_info(entry.sha1, &size)) < 0)
+ cmp = tree_entry_len(entry.path, entry.sha1) != cmplen ? 1 :
+ memcmp(name, entry.path, cmplen);
+ if (cmp > 0)
continue;
+ if (cmp < 0)
+ return;
if (name[cmplen] != '/') {
unsigned hash = name_hash(fullname);
add_object_entry(entry.sha1, hash, 1);
return;
}
- if (type == OBJ_TREE) {
+ if (S_ISDIR(entry.mode)) {
struct tree_desc sub;
struct pbase_tree_cache *tree;
const char *down = name+cmplen+1;
static void add_preferred_base_object(const char *name, unsigned hash)
{
struct pbase_tree *it;
- int cmplen = name_cmp_len(name);
+ int cmplen;
- if (check_pbase_path(hash))
+ if (!num_preferred_base || check_pbase_path(hash))
return;
+ cmplen = name_cmp_len(name);
for (it = pbase_tree; it; it = it->next) {
if (cmplen == 0) {
- hash = name_hash("");
- add_object_entry(it->pcache.sha1, hash, 1);
+ add_object_entry(it->pcache.sha1, 0, 1);
}
else {
struct tree_desc tree;
ofs = c & 127;
while (c & 128) {
ofs += 1;
- if (!ofs || ofs & ~(~0UL >> 7))
+ if (!ofs || MSB(ofs, 7))
die("delta base offset overflow in pack for %s",
sha1_to_hex(entry->sha1));
c = buf[used_0++];
sha1_to_hex(entry->sha1));
}
-static unsigned int check_delta_limit(struct object_entry *me, unsigned int n)
-{
- struct object_entry *child = me->delta_child;
- unsigned int m = n;
- while (child) {
- unsigned int c = check_delta_limit(child, n + 1);
- if (m < c)
- m = c;
- child = child->delta_sibling;
- }
- return m;
-}
-
static void get_object_details(void)
{
uint32_t i;
prepare_pack_ix();
for (i = 0, entry = objects; i < nr_objects; i++, entry++)
check_object(entry);
-
- 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 *);
-
-static entry_sort_t current_sort;
-
-static int sort_comparator(const void *_a, const void *_b)
-{
- struct object_entry *a = *(struct object_entry **)_a;
- struct object_entry *b = *(struct object_entry **)_b;
- return current_sort(a,b);
-}
-
-static struct object_entry **create_sorted_list(entry_sort_t sort)
-{
- struct object_entry **list = xmalloc(nr_objects * sizeof(struct object_entry *));
- uint32_t i;
-
- for (i = 0; i < nr_objects; i++)
- list[i] = objects + i;
- current_sort = sort;
- qsort(list, nr_objects, sizeof(struct object_entry *), sort_comparator);
- return list;
-}
-
-static int sha1_sort(const struct object_entry *a, const struct object_entry *b)
+static int sha1_sort(const void *_a, const void *_b)
{
+ const struct object_entry *a = *(struct object_entry **)_a;
+ const struct object_entry *b = *(struct object_entry **)_b;
return hashcmp(a->sha1, b->sha1);
}
if (!objects[i].preferred_base)
list[j++] = objects + i;
}
- current_sort = sha1_sort;
- qsort(list, nr_result, sizeof(struct object_entry *), sort_comparator);
+ qsort(list, nr_result, sizeof(struct object_entry *), sha1_sort);
return list;
}
-static int type_size_sort(const struct object_entry *a, const struct object_entry *b)
+static int type_size_sort(const void *_a, const void *_b)
{
+ const struct object_entry *a = *(struct object_entry **)_a;
+ const struct object_entry *b = *(struct object_entry **)_b;
+
if (a->type < b->type)
return -1;
if (a->type > b->type)
return -1;
if (a->size > b->size)
return 1;
- return a < b ? -1 : (a > b);
+ return a > b ? -1 : (a < b); /* newest last */
}
struct unpacked {
trg_entry->in_pack_type != OBJ_OFS_DELTA)
return 0;
- /*
- * 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 (trg_entry->delta_child) {
- if (max_depth <= trg_entry->delta_limit)
- return 0;
- max_depth -= trg_entry->delta_limit;
- }
+ /* Let's not bust the allowed depth. */
if (src_entry->depth >= max_depth)
return 0;
return 1;
}
-static void progress_interval(int signum)
+static unsigned int check_delta_limit(struct object_entry *me, unsigned int n)
{
- progress_update = 1;
+ struct object_entry *child = me->delta_child;
+ unsigned int m = n;
+ while (child) {
+ unsigned int c = check_delta_limit(child, n + 1);
+ if (m < c)
+ m = c;
+ child = child->delta_sibling;
+ }
+ return m;
}
static void find_deltas(struct object_entry **list, int window, int depth)
unsigned int array_size = window * sizeof(struct unpacked);
struct unpacked *array;
unsigned last_percent = 999;
+ int max_depth;
if (!nr_objects)
return;
n->data = NULL;
n->entry = entry;
+ /*
+ * 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.
+ */
+ max_depth = depth;
+ if (entry->delta_child) {
+ max_depth -= check_delta_limit(entry, 0);
+ if (max_depth <= 0)
+ goto next;
+ }
+
j = window;
while (--j > 0) {
uint32_t other_idx = idx + j;
m = array + other_idx;
if (!m->entry)
break;
- if (try_delta(n, m, depth) < 0)
+ if (try_delta(n, m, max_depth) < 0)
break;
}
+
/* if we made n a delta, and if n is already at max
* depth, leaving it in the window is pointless. we
* should evict it first.
if (entry->delta && depth <= entry->depth)
continue;
+ next:
idx++;
if (idx >= window)
idx = 0;
static void prepare_pack(int window, int depth)
{
+ struct object_entry **delta_list;
+ uint32_t i;
+
get_object_details();
- sorted_by_type = create_sorted_list(type_size_sort);
- if (window && depth)
- find_deltas(sorted_by_type, window+1, depth);
+
+ if (!window || !depth)
+ return;
+
+ delta_list = xmalloc(nr_objects * sizeof(*delta_list));
+ for (i = 0; i < nr_objects; i++)
+ delta_list[i] = objects + i;
+ qsort(delta_list, nr_objects, sizeof(*delta_list), type_size_sort);
+ find_deltas(delta_list, window+1, depth);
+ free(delta_list);
}
static int reuse_cached_pack(unsigned char *sha1)
return 1;
}
+static void progress_interval(int signum)
+{
+ progress_update = 1;
+}
+
static void setup_progress_signal(void)
{
struct sigaction sa;
static void show_commit(struct commit *commit)
{
- unsigned hash = name_hash("");
- add_preferred_base_object("", hash);
- add_object_entry(commit->object.sha1, hash, 0);
+ add_object_entry(commit->object.sha1, 0, 0);
}
static void show_object(struct object_array_entry *p)
rp_av[1] = "--objects-edge";
continue;
}
+ if (!prefixcmp(arg, "--index-version=")) {
+ char *c;
+ index_default_version = strtoul(arg + 16, &c, 10);
+ if (index_default_version > 2)
+ die("bad %s", arg);
+ if (*c == ',')
+ index_off32_limit = strtoul(c+1, &c, 0);
+ if (*c || index_off32_limit & 0x80000000)
+ die("bad %s", arg);
+ continue;
+ }
usage(pack_usage);
}
if (reuse_cached_pack(object_list_sha1))
;
else {
+ off_t last_obj_offset;
if (nr_result)
prepare_pack(window, depth);
if (progress == 1 && pack_to_stdout) {
signal(SIGALRM, SIG_IGN );
progress_update = 0;
}
- write_pack_file();
+ last_obj_offset = write_pack_file();
if (!pack_to_stdout) {
- write_index_file();
+ write_index_file(last_obj_offset);
puts(sha1_to_hex(object_list_sha1));
}
}