index-pack: use streaming interface on large blobs (most of the time)

unpack_raw_entry() will not allocate and return decompressed blobs if
they are larger than core.bigFileThreshold. sha1_object() may not be
called on those objects because there's no actual content.

sha1_object() is called later on those objects, where we can safely
use get_data_from_pack() to retrieve blob content for checking.
However we always do that when we definitely need the blob
content. And we often don't.

There are two cases when we may need object content. The first case is
when we find an in-repo blob with the same SHA-1. We need to do
collision test, byte-on-byte. If this test is on, the blob must be
loaded on memory (i.e. no streaming). Normally (e.g. in
fetch/pull/clone) this does not happen because git avoid to send
objects that client already has.

The other case is when --strict is specified and the object in
question is not a blob, which can't happen in reality becase we deal
with large _blobs_ here.

Note: --verify (or git-verify-pack) a pack from current repository
will trigger collision test on every object in the pack, which
effectively disables this patch. This could be easily worked around by
setting GIT_DIR to an imaginary place with no packs.

Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

builtin/index-pack.c

@@ -392,9 +392,10 @@ static int is_delta_type(enum object_type type)
 static void *unpack_entry_data(unsigned long offset, unsigned long size,
 			       enum object_type type, unsigned char *sha1)
 {
+	static char fixed_buf[8192];
 	int status;
 	git_zstream stream;
-	void *buf = xmalloc(size);
+	void *buf;
 	git_SHA_CTX c;
 	char hdr[32];
 	int hdrlen;
@@ -405,11 +406,15 @@ static void *unpack_entry_data(unsigned long offset, unsigned long size,
 		git_SHA1_Update(&c, hdr, hdrlen);
 	} else
 		sha1 = NULL;
+	if (type == OBJ_BLOB && size > big_file_threshold)
+		buf = fixed_buf;
+	else
+		buf = xmalloc(size);
 
 	memset(&stream, 0, sizeof(stream));
 	git_inflate_init(&stream);
 	stream.next_out = buf;
-	stream.avail_out = size;
+	stream.avail_out = buf == fixed_buf ? sizeof(fixed_buf) : size;
 
 	do {
 		unsigned char *last_out = stream.next_out;
@@ -419,13 +424,17 @@ static void *unpack_entry_data(unsigned long offset, unsigned long size,
 		use(input_len - stream.avail_in);
 		if (sha1)
 			git_SHA1_Update(&c, last_out, stream.next_out - last_out);
+		if (buf == fixed_buf) {
+			stream.next_out = buf;
+			stream.avail_out = sizeof(fixed_buf);
+		}
 	} while (status == Z_OK);
 	if (stream.total_out != size || status != Z_STREAM_END)
 		bad_object(offset, _("inflate returned %d"), status);
 	git_inflate_end(&stream);
 	if (sha1)
 		git_SHA1_Final(sha1, &c);
-	return buf;
+	return buf == fixed_buf ? NULL : buf;
 }
 
 static void *unpack_raw_entry(struct object_entry *obj,
@@ -591,14 +600,21 @@ static void find_delta_children(const union delta_base *base,
 	*last_index = last;
 }
 
-static void sha1_object(const void *data, unsigned long size,
-			enum object_type type, const unsigned char *sha1)
+static void sha1_object(const void *data, struct object_entry *obj_entry,
+			unsigned long size, enum object_type type,
+			const unsigned char *sha1)
 {
+	void *new_data = NULL;
+
+	assert(data || obj_entry);
+
 	read_lock();
 	if (has_sha1_file(sha1)) {
 		void *has_data;
 		enum object_type has_type;
 		unsigned long has_size;
+		if (!data)
+			data = new_data = get_data_from_pack(obj_entry);
 		has_data = read_sha1_file(sha1, &has_type, &has_size);
 		read_unlock();
 		if (!has_data)
@@ -623,6 +639,9 @@ static void sha1_object(const void *data, unsigned long size,
 			int eaten;
 			void *buf = (void *) data;
 
+			if (!buf)
+				buf = new_data = get_data_from_pack(obj_entry);
+
 			/*
 			 * we do not need to free the memory here, as the
 			 * buf is deleted by the caller.
@@ -647,6 +666,8 @@ static void sha1_object(const void *data, unsigned long size,
 		}
 		read_unlock();
 	}
+
+	free(new_data);
 }
 
 /*
@@ -730,7 +751,7 @@ static void resolve_delta(struct object_entry *delta_obj,
 		bad_object(delta_obj->idx.offset, _("failed to apply delta"));
 	hash_sha1_file(result->data, result->size,
 		       typename(delta_obj->real_type), delta_obj->idx.sha1);
-	sha1_object(result->data, result->size, delta_obj->real_type,
+	sha1_object(result->data, NULL, result->size, delta_obj->real_type,
 		    delta_obj->idx.sha1);
 	counter_lock();
 	nr_resolved_deltas++;
@@ -860,7 +881,7 @@ static void *threaded_second_pass(void *data)
  */
 static void parse_pack_objects(unsigned char *sha1)
 {
-	int i;
+	int i, nr_delays = 0;
 	struct delta_entry *delta = deltas;
 	struct stat st;
 
@@ -876,8 +897,12 @@ static void parse_pack_objects(unsigned char *sha1)
 			nr_deltas++;
 			delta->obj_no = i;
 			delta++;
+		} else if (!data) {
+			/* large blobs, check later */
+			obj->real_type = OBJ_BAD;
+			nr_delays++;
 		} else
-			sha1_object(data, obj->size, obj->type, obj->idx.sha1);
+			sha1_object(data, NULL, obj->size, obj->type, obj->idx.sha1);
 		free(data);
 		display_progress(progress, i+1);
 	}
@@ -897,6 +922,17 @@ static void parse_pack_objects(unsigned char *sha1)
 	if (S_ISREG(st.st_mode) &&
 			lseek(input_fd, 0, SEEK_CUR) - input_len != st.st_size)
 		die(_("pack has junk at the end"));
+
+	for (i = 0; i < nr_objects; i++) {
+		struct object_entry *obj = &objects[i];
+		if (obj->real_type != OBJ_BAD)
+			continue;
+		obj->real_type = obj->type;
+		sha1_object(NULL, obj, obj->size, obj->type, obj->idx.sha1);
+		nr_delays--;
+	}
+	if (nr_delays)
+		die(_("confusion beyond insanity in parse_pack_objects()"));
 }
 
 /*

t/t1050-large.sh

@@ -130,6 +130,11 @@ test_expect_success 'git-show a large file' '
 
 '
 
+test_expect_success 'index-pack' '
+	git clone file://"`pwd`"/.git foo &&
+	GIT_DIR=non-existent git index-pack --strict --verify foo/.git/objects/pack/*.pack
+'
+
 test_expect_success 'repack' '
 	git repack -ad
 '