From efec43c028c52bc252afdaa586b46f83186d3eaa Mon Sep 17 00:00:00 2001 From: Junio C Hamano Date: Mon, 5 Mar 2007 00:21:24 -0800 Subject: [PATCH 1/3] fsck: fix broken loose object check. When "git fsck" without --full found a loose object missing because it was broken, it mistakenly thought it was not parsed because we found it in one of the packs. Back when this code was written, we did not have a way to explicitly check if we have the object in pack, but we do now. Signed-off-by: Junio C Hamano --- builtin-fsck.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/builtin-fsck.c b/builtin-fsck.c index 6abf498d2b..abdd0f2c18 100644 --- a/builtin-fsck.c +++ b/builtin-fsck.c @@ -67,7 +67,7 @@ static void check_reachable_object(struct object *obj) * do a full fsck */ if (!obj->parsed) { - if (has_sha1_file(obj->sha1)) + if (has_sha1_pack(obj->sha1, NULL)) return; /* it is in pack - forget about it */ printf("missing %s %s\n", typename(obj->type), sha1_to_hex(obj->sha1)); return; From 7efbff7531af4281487d54c1dc1401308d988e33 Mon Sep 17 00:00:00 2001 From: Junio C Hamano Date: Mon, 5 Mar 2007 00:21:37 -0800 Subject: [PATCH 2/3] unpack_sha1_file(): detect corrupt loose object files. We did not detect broken loose object files, either when underlying inflate() signalled the breakage, nor inflate() finished and we had garbage trailing at the end. We do better now. We also make unpack_sha1_file() a static function to sha1_file.c, since it is not used by anybody outside. Signed-off-by: Junio C Hamano --- cache.h | 1 - sha1_file.c | 27 +++++++++++++++++++-------- 2 files changed, 19 insertions(+), 9 deletions(-) diff --git a/cache.h b/cache.h index c291163e6d..4b5a7541a8 100644 --- a/cache.h +++ b/cache.h @@ -281,7 +281,6 @@ char *enter_repo(char *path, int strict); /* Read and unpack a sha1 file into memory, write memory to a sha1 file */ extern int sha1_object_info(const unsigned char *, unsigned long *); -extern void * unpack_sha1_file(void *map, unsigned long mapsize, enum object_type *type, unsigned long *size); extern void * read_sha1_file(const unsigned char *sha1, enum object_type *type, unsigned long *size); extern int hash_sha1_file(void *buf, unsigned long len, const char *type, unsigned char *sha1); extern int write_sha1_file(void *buf, unsigned long len, const char *type, unsigned char *return_sha1); diff --git a/sha1_file.c b/sha1_file.c index 6d0a72ed09..ac6b5e00b6 100644 --- a/sha1_file.c +++ b/sha1_file.c @@ -956,11 +956,12 @@ static int unpack_sha1_header(z_stream *stream, unsigned char *map, unsigned lon return 0; } -static void *unpack_sha1_rest(z_stream *stream, void *buffer, unsigned long size) +static void *unpack_sha1_rest(z_stream *stream, void *buffer, unsigned long size, const unsigned char *sha1) { int bytes = strlen(buffer) + 1; unsigned char *buf = xmalloc(1+size); unsigned long n; + int status = Z_OK; n = stream->total_out - bytes; if (n > size) @@ -970,12 +971,22 @@ static void *unpack_sha1_rest(z_stream *stream, void *buffer, unsigned long size if (bytes < size) { stream->next_out = buf + bytes; stream->avail_out = size - bytes; - while (inflate(stream, Z_FINISH) == Z_OK) - /* nothing */; + while (status == Z_OK) + status = inflate(stream, Z_FINISH); } buf[size] = 0; - inflateEnd(stream); - return buf; + if ((status == Z_OK || status == Z_STREAM_END) && !stream->avail_in) { + inflateEnd(stream); + return buf; + } + + if (status < 0) + error("corrupt loose object '%s'", sha1_to_hex(sha1)); + else if (stream->avail_in) + error("garbage at end of loose object '%s'", + sha1_to_hex(sha1)); + free(buf); + return NULL; } /* @@ -1029,7 +1040,7 @@ static int parse_sha1_header(const char *hdr, unsigned long *sizep) return *hdr ? -1 : type_from_string(type); } -void * unpack_sha1_file(void *map, unsigned long mapsize, enum object_type *type, unsigned long *size) +static void *unpack_sha1_file(void *map, unsigned long mapsize, enum object_type *type, unsigned long *size, const unsigned char *sha1) { int ret; z_stream stream; @@ -1039,7 +1050,7 @@ void * unpack_sha1_file(void *map, unsigned long mapsize, enum object_type *type if (ret < Z_OK || (*type = parse_sha1_header(hdr, size)) < 0) return NULL; - return unpack_sha1_rest(&stream, hdr, *size); + return unpack_sha1_rest(&stream, hdr, *size, sha1); } static unsigned long get_delta_base(struct packed_git *p, @@ -1555,7 +1566,7 @@ void *read_sha1_file(const unsigned char *sha1, enum object_type *type, return buf; map = map_sha1_file(sha1, &mapsize); if (map) { - buf = unpack_sha1_file(map, mapsize, type, size); + buf = unpack_sha1_file(map, mapsize, type, size, sha1); munmap(map, mapsize); return buf; } From e2b4f63512357d83f1f11bafadb22063e5d56621 Mon Sep 17 00:00:00 2001 From: Junio C Hamano Date: Mon, 5 Mar 2007 00:22:06 -0800 Subject: [PATCH 3/3] fsck: exit with non-zero status upon errors git-fsck always exited with status 0, which was a bit sloppy. This makes it exit with a non-zero status when errors are found. The error code is an OR'ed result of: 1 if corrupted objects are found. 2 if objects that are ought to be reachable are missing or corrupt. For example, it would exit with 1 in a repository with an unreachable corrupt object. If a tree object of the HEAD commit is corrupt, you would get 3. Signed-off-by: Junio C Hamano --- builtin-fsck.c | 26 +++++++++++++++++++++----- 1 file changed, 21 insertions(+), 5 deletions(-) diff --git a/builtin-fsck.c b/builtin-fsck.c index abdd0f2c18..4d03378c1b 100644 --- a/builtin-fsck.c +++ b/builtin-fsck.c @@ -18,6 +18,9 @@ static int check_full; static int check_strict; static int keep_cache_objects; static unsigned char head_sha1[20]; +static int errors_found; +#define ERROR_OBJECT 01 +#define ERROR_REACHABLE 02 #ifdef NO_D_INO_IN_DIRENT #define SORT_DIRENT 0 @@ -40,6 +43,7 @@ static int objerror(struct object *obj, const char *err, ...) { va_list params; va_start(params, err); + errors_found |= ERROR_OBJECT; objreport(obj, "error", err, params); va_end(params); return -1; @@ -70,6 +74,7 @@ static void check_reachable_object(struct object *obj) if (has_sha1_pack(obj->sha1, NULL)) return; /* it is in pack - forget about it */ printf("missing %s %s\n", typename(obj->type), sha1_to_hex(obj->sha1)); + errors_found |= ERROR_REACHABLE; return; } @@ -88,6 +93,7 @@ static void check_reachable_object(struct object *obj) typename(obj->type), sha1_to_hex(obj->sha1)); printf(" to %7s %s\n", typename(ref->type), sha1_to_hex(ref->sha1)); + errors_found |= ERROR_REACHABLE; } } } @@ -346,8 +352,11 @@ static int fsck_tag(struct tag *tag) static int fsck_sha1(unsigned char *sha1) { struct object *obj = parse_object(sha1); - if (!obj) - return error("%s: object corrupt or missing", sha1_to_hex(sha1)); + if (!obj) { + errors_found |= ERROR_OBJECT; + return error("%s: object corrupt or missing", + sha1_to_hex(sha1)); + } if (obj->flags & SEEN) return 0; obj->flags |= SEEN; @@ -359,8 +368,10 @@ static int fsck_sha1(unsigned char *sha1) return fsck_commit((struct commit *) obj); if (obj->type == OBJ_TAG) return fsck_tag((struct tag *) obj); + /* By now, parse_object() would've returned NULL instead. */ - return objerror(obj, "unknown type '%d' (internal fsck error)", obj->type); + return objerror(obj, "unknown type '%d' (internal fsck error)", + obj->type); } /* @@ -576,11 +587,16 @@ static int fsck_cache_tree(struct cache_tree *it) return err; } +static const char fsck_usage[] = +"git-fsck [--tags] [--root] [[--unreachable] [--cache] [--full] " +"[--strict] *]"; + int cmd_fsck(int argc, char **argv, const char *prefix) { int i, heads; track_object_refs = 1; + errors_found = 0; for (i = 1; i < argc; i++) { const char *arg = argv[i]; @@ -610,7 +626,7 @@ int cmd_fsck(int argc, char **argv, const char *prefix) continue; } if (*arg == '-') - usage("git-fsck [--tags] [--root] [[--unreachable] [--cache] [--full] [--strict] *]"); + usage(fsck_usage); } fsck_head_link(); @@ -690,5 +706,5 @@ int cmd_fsck(int argc, char **argv, const char *prefix) } check_connectivity(); - return 0; + return errors_found; }