git/list[1] front-page[2] threads[3] people[4] search[5] about
 

[PATCH v2 14/35] fast-import: make hash-size independent

From
brian m. carlson <sandals@crustytoothpaste.net>
Date
Feb 19, 2019, 00:05 UTC
Message-ID
<20190219000526.476553-15-sandals@crustytoothpaste.net>
In-Reply-To
<20190219000526.476553-1-sandals@crustytoothpaste.net>

Replace several uses of GIT_SHA1_HEXSZ and 40-based constants with references to the_hash_algo. Update the note handling code here to compute path sizes based on GIT_MAX_RAWSZ as well.

Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net>
---
 fast-import.c | 45 +++++++++++++++++++++++++++++----------------
 1 file changed, 29 insertions(+), 16 deletions(-)
diff --git a/fast-import.c b/fast-import.c
index 7c9a10a77b..464db71c75 100644
--- a/fast-import.c
+++ b/fast-import.c
@@ -29,6 +29,13 @@
  */
 #define NO_DELTA S_ISUID
 
+/*
+ * The amount of additional space required in order to write an object into the
+ * current pack. This is the hash lengths at the end of the pack, plus the
+ * length of one object ID.
+ */
+#define PACK_SIZE_THRESHOLD (the_hash_algo->rawsz * 3)
+
 struct object_entry {
 	struct pack_idx_entry idx;
 	struct object_entry *next;
@@ -949,8 +956,9 @@ static int store_object(
 	git_deflate_end(&s);
 
 	/* Determine if we should auto-checkpoint. */
-	if ((max_packsize && (pack_size + 60 + s.total_out) > max_packsize)
-		|| (pack_size + 60 + s.total_out) < pack_size) {
+	if ((max_packsize
+		&& (pack_size + PACK_SIZE_THRESHOLD + s.total_out) > max_packsize)
+		|| (pack_size + PACK_SIZE_THRESHOLD + s.total_out) < pack_size) {
 
 		/* This new object needs to *not* have the current pack_id. */
 		e->pack_id = pack_id + 1;
@@ -1045,8 +1053,9 @@ static void stream_blob(uintmax_t len, struct object_id *oidout, uintmax_t mark)
 	int status = Z_OK;
 
 	/* Determine if we should auto-checkpoint. */
-	if ((max_packsize && (pack_size + 60 + len) > max_packsize)
-		|| (pack_size + 60 + len) < pack_size)
+	if ((max_packsize
+		&& (pack_size + PACK_SIZE_THRESHOLD + len) > max_packsize)
+		|| (pack_size + PACK_SIZE_THRESHOLD + len) < pack_size)
 		cycle_packfile();
 
 	hashfile_checkpoint(pack_file, &checkpoint);
@@ -1241,7 +1250,7 @@ static void load_tree(struct tree_entry *root)
 		c += e->name->str_len + 1;
 		hashcpy(e->versions[0].oid.hash, (unsigned char *)c);
 		hashcpy(e->versions[1].oid.hash, (unsigned char *)c);
-		c += GIT_SHA1_RAWSZ;
+		c += the_hash_algo->rawsz;
 	}
 	free(buf);
 }
@@ -1288,7 +1297,7 @@ static void mktree(struct tree_content *t, int v, struct strbuf *b)
 		strbuf_addf(b, "%o %s%c",
 			(unsigned int)(e->versions[v].mode & ~NO_DELTA),
 			e->name->str_dat, '\0');
-		strbuf_add(b, e->versions[v].oid.hash, GIT_SHA1_RAWSZ);
+		strbuf_add(b, e->versions[v].oid.hash, the_hash_algo->rawsz);
 	}
 }
 
@@ -2047,7 +2056,9 @@ static uintmax_t do_change_note_fanout(
 	unsigned int i, tmp_hex_oid_len, tmp_fullpath_len;
 	uintmax_t num_notes = 0;
 	struct object_id oid;
-	char realpath[60];
+	/* hex oid + '/' between each pair of hex digits + NUL */
+	char realpath[GIT_MAX_HEXSZ + ((GIT_MAX_HEXSZ / 2) - 1) + 1];
+	const unsigned hexsz = the_hash_algo->hexsz;
 
 	if (!root->tree)
 		load_tree(root);
@@ -2067,7 +2078,7 @@ static uintmax_t do_change_note_fanout(
 		 * of 2 chars.
 		 */
 		if (!e->versions[1].mode ||
-		    tmp_hex_oid_len > GIT_SHA1_HEXSZ ||
+		    tmp_hex_oid_len > hexsz ||
 		    e->name->str_len % 2)
 			continue;
 
@@ -2081,7 +2092,7 @@ static uintmax_t do_change_note_fanout(
 		tmp_fullpath_len += e->name->str_len;
 		fullpath[tmp_fullpath_len] = '\0';
 
-		if (tmp_hex_oid_len == GIT_SHA1_HEXSZ && !get_oid_hex(hex_oid, &oid)) {
+		if (tmp_hex_oid_len == hexsz && !get_oid_hex(hex_oid, &oid)) {
 			/* This is a note entry */
 			if (fanout == 0xff) {
 				/* Counting mode, no rename */
@@ -2352,7 +2363,7 @@ static void note_change_n(const char *p, struct branch *b, unsigned char *old_fa
 	struct object_entry *oe;
 	struct branch *s;
 	struct object_id oid, commit_oid;
-	char path[60];
+	char path[GIT_MAX_RAWSZ * 3];
 	uint16_t inline_data = 0;
 	unsigned char new_fanout;
 
@@ -2405,7 +2416,7 @@ static void note_change_n(const char *p, struct branch *b, unsigned char *old_fa
 		char *buf = read_object_with_reference(&commit_oid,
 						       commit_type, &size,
 						       &commit_oid);
-		if (!buf || size < 46)
+		if (!buf || size < the_hash_algo->hexsz + 6)
 			die("Not a valid commit: %s", p);
 		free(buf);
 	} else
@@ -2456,7 +2467,7 @@ static void file_change_deleteall(struct branch *b)
 
 static void parse_from_commit(struct branch *b, char *buf, unsigned long size)
 {
-	if (!buf || size < GIT_SHA1_HEXSZ + 6)
+	if (!buf || size < the_hash_algo->hexsz + 6)
 		die("Not a valid commit: %s", oid_to_hex(&b->oid));
 	if (memcmp("tree ", buf, 5)
 		|| get_oid_hex(buf + 5, &b->branch_tree.versions[1].oid))
@@ -2555,7 +2566,7 @@ static struct hash_list *parse_merge(unsigned int *count)
 			char *buf = read_object_with_reference(&n->oid,
 							       commit_type,
 							       &size, &n->oid);
-			if (!buf || size < 46)
+			if (!buf || size < the_hash_algo->hexsz + 6)
 				die("Not a valid commit: %s", from);
 			free(buf);
 		} else
@@ -2842,7 +2853,7 @@ static void parse_get_mark(const char *p)
 		die("Unknown mark: %s", command_buf.buf);
 
 	xsnprintf(output, sizeof(output), "%s\n", oid_to_hex(&oe->idx.oid));
-	cat_blob_write(output, GIT_SHA1_HEXSZ + 1);
+	cat_blob_write(output, the_hash_algo->hexsz + 1);
 }
 
 static void parse_cat_blob(const char *p)
@@ -2872,6 +2883,8 @@ static struct object_entry *dereference(struct object_entry *oe,
 {
 	unsigned long size;
 	char *buf = NULL;
+	const unsigned hexsz = the_hash_algo->hexsz;
+
 	if (!oe) {
 		enum object_type type = oid_object_info(the_repository, oid,
 							NULL);
@@ -2905,12 +2918,12 @@ static struct object_entry *dereference(struct object_entry *oe,
 	/* Peel one layer. */
 	switch (oe->type) {
 	case OBJ_TAG:
-		if (size < GIT_SHA1_HEXSZ + strlen("object ") ||
+		if (size < hexsz + strlen("object ") ||
 		    get_oid_hex(buf + strlen("object "), oid))
 			die("Invalid SHA1 in tag: %s", command_buf.buf);
 		break;
 	case OBJ_COMMIT:
-		if (size < GIT_SHA1_HEXSZ + strlen("tree ") ||
+		if (size < hexsz + strlen("tree ") ||
 		    get_oid_hex(buf + strlen("tree "), oid))
 			die("Invalid SHA1 in commit: %s", command_buf.buf);
 	}
Previous: brian m. carlsonNext: brian m. carlson
Message 11 of 36 in “Hash function transition part 16”
  1. 00/35 Hash function transition part 16brian m. carlson, Feb 19, 2019
  2. 01/35 t/lib-submodule-update: use appropriate length constantbrian m. carlson, Feb 19, 2019
  3. 02/35 khash: move oid hash table definitionbrian m. carlson, Feb 19, 2019
  4. 03/35 pack-bitmap: make bitmap header handling hash agnosticbrian m. carlson, Feb 19, 2019
  5. 06/35 pack-bitmap: switch hard-coded constants to the_hash_algobrian m. carlson, Feb 19, 2019
  6. 05/35 pack-bitmap: replace sha1_to_hexbrian m. carlson, Feb 19, 2019
  7. 07/35 pack-bitmap: switch hash tables to use struct object_idbrian m. carlson, Feb 19, 2019
  8. 09/35 notes-merge: switch to use the_hash_algobrian m. carlson, Feb 19, 2019
  9. 08/35 submodule: avoid hard-coded constantsbrian m. carlson, Feb 19, 2019
  10. 11/35 notes: replace sha1_to_hexbrian m. carlson, Feb 19, 2019
  11. 14/35 fast-import: make hash-size independentbrian m. carlson, Feb 19, 2019
  12. 17/35 builtin/pull: make hash-size independentbrian m. carlson, Feb 19, 2019
  13. 15/35 fast-import: replace sha1_to_hexbrian m. carlson, Feb 19, 2019
  14. 12/35 object-store: rename and expand packed_git's sha1 memberbrian m. carlson, Feb 19, 2019
  15. 19/35 http-backend: allow 64-character hex namesbrian m. carlson, Feb 19, 2019
  16. 18/35 http-push: convert to use the_hash_algobrian m. carlson, Feb 19, 2019
  17. 21/35 http-walker: replace sha1_to_hexbrian m. carlson, Feb 19, 2019
  18. 22/35 http: replace hard-coded constant with the_hash_algobrian m. carlson, Feb 19, 2019
  19. 20/35 http-push: remove remaining uses of sha1_to_hexbrian m. carlson, Feb 19, 2019
  20. 23/35 http: compute hash of downloaded objects using the_hash_algobrian m. carlson, Feb 19, 2019
  21. 24/35 http: replace sha1_to_hexbrian m. carlson, Feb 19, 2019
  22. 25/35 remote-curl: make hash size independentbrian m. carlson, Feb 19, 2019
  23. 26/35 hash: add a function to lookup hash algorithm by lengthbrian m. carlson, Feb 19, 2019
  24. 30/35 refspec: make hash size independentbrian m. carlson, Feb 19, 2019
  25. 27/35 get-tar-commit-id: parse comment recordbrian m. carlson, Feb 19, 2019
  26. 29/35 archive: convert struct archiver_args to object_idbrian m. carlson, Feb 19, 2019
  27. 34/35 Git.pm: make hash size independentbrian m. carlson, Feb 19, 2019
  28. 33/35 read-cache: read data in a hash-independent waybrian m. carlson, Feb 19, 2019
  29. 31/35 builtin/difftool: use parse_oid_hexbrian m. carlson, Feb 19, 2019
  30. 35/35 gitweb: make hash size independentbrian m. carlson, Feb 19, 2019
  31. 32/35 dir: make untracked cache extension hash size independentbrian m. carlson, Feb 19, 2019
  32. 16/35 builtin/am: make hash size independentbrian m. carlson, Feb 19, 2019
  33. 28/35 builtin/get-tar-commit-id: make hash size independentbrian m. carlson, Feb 19, 2019
  34. 13/35 builtin/name-rev: make hash-size independentbrian m. carlson, Feb 19, 2019
  35. 04/35 pack-bitmap: convert struct stored_bitmap to object_idbrian m. carlson, Feb 19, 2019
  36. 10/35 notes: make hash size independentbrian m. carlson, Feb 19, 2019

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.