1 /*
2  * Copyright (c) 2018, 2019, 2020 Stefan Sperling <stsp@openbsd.org>
3  *
4  * Permission to use, copy, modify, and distribute this software for any
5  * purpose with or without fee is hereby granted, provided that the above
6  * copyright notice and this permission notice appear in all copies.
7  *
8  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
9  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
10  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
11  * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
12  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
13  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
14  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
15  */
16 
17 /* An open pack file. */
18 struct got_pack {
19 	char *path_packfile;
20 	int fd;
21 	uint8_t *map;
22 	size_t filesize;
23 	struct got_privsep_child *privsep_child;
24 	int child_has_tempfiles;
25 	struct got_delta_cache *delta_cache;
26 };
27 
28 struct got_packidx;
29 
30 const struct got_error *got_pack_close(struct got_pack *);
31 
32 const struct got_error *got_pack_parse_offset_delta(off_t *, size_t *,
33     struct got_pack *, off_t, int);
34 const struct got_error *got_pack_parse_ref_delta(struct got_object_id *,
35     struct got_pack *, off_t, int);
36 const struct got_error *got_pack_resolve_delta_chain(struct got_delta_chain *,
37     struct got_packidx *, struct got_pack *, off_t, size_t, int, size_t,
38     unsigned int);
39 const struct got_error *got_pack_parse_object_type_and_size(uint8_t *,
40     uint64_t *, size_t *, struct got_pack *, off_t);
41 
42 #define GOT_PACK_PREFIX		"pack-"
43 #define GOT_PACKFILE_SUFFIX	".pack"
44 #define GOT_PACKIDX_SUFFIX		".idx"
45 #define GOT_PACKFILE_NAMELEN	(strlen(GOT_PACK_PREFIX) + \
46 				SHA1_DIGEST_STRING_LENGTH - 1 + \
47 				strlen(GOT_PACKFILE_SUFFIX))
48 #define GOT_PACKIDX_NAMELEN	(strlen(GOT_PACK_PREFIX) + \
49 				SHA1_DIGEST_STRING_LENGTH - 1 + \
50 				strlen(GOT_PACKIDX_SUFFIX))
51 
52 /* See Documentation/technical/pack-format.txt in Git. */
53 
54 struct got_packidx_trailer {
55 	u_int8_t	packfile_sha1[SHA1_DIGEST_LENGTH];
56 	u_int8_t	packidx_sha1[SHA1_DIGEST_LENGTH];
57 } __attribute__((__packed__));
58 
59 struct got_packidx_object_id {
60 	u_int8_t sha1[SHA1_DIGEST_LENGTH];
61 } __attribute__((__packed__));
62 
63 /* Ignore pack index version 1 which is no longer written by Git. */
64 #define GOT_PACKIDX_VERSION 2
65 
66 struct got_packidx_v2_hdr {
67 	uint32_t	*magic;		/* big endian */
68 #define GOT_PACKIDX_V2_MAGIC 0xff744f63	/* "\377t0c" */
69 	uint32_t	*version;
70 
71 	/* 
72 	 * Each entry N in the fanout table contains the number of objects in
73 	 * the packfile whose SHA1 begins with a byte less than or equal to N.
74 	 * The last entry (index 255) contains the number of objects in the
75 	 * pack file whose first SHA1 byte is <= 0xff, and thus records the
76 	 * total number of objects in the pack file. All pointer variables
77 	 * below point to tables with a corresponding number of entries.
78 	 */
79 	uint32_t	*fanout_table;	/* values are big endian */
80 #define GOT_PACKIDX_V2_FANOUT_TABLE_ITEMS (0xff + 1)
81 
82 	/* Sorted SHA1 checksums for each object in the pack file. */
83 	struct got_packidx_object_id *sorted_ids;
84 
85 	/* CRC32 of the packed representation of each object. */
86 	uint32_t	*crc32;
87 
88 	/* Offset into the pack file for each object. */
89 	uint32_t	*offsets;		/* values are big endian */
90 #define GOT_PACKIDX_OFFSET_VAL_MASK		0x7fffffff
91 #define GOT_PACKIDX_OFFSET_VAL_IS_LARGE_IDX	0x80000000
92 
93 	/* Large offsets table is empty for pack files < 2 GB. */
94 	uint64_t	*large_offsets;		/* values are big endian */
95 
96 	struct got_packidx_trailer *trailer;
97 };
98 
99 /* An open pack index file. */
100 struct got_packidx {
101 	char *path_packidx; /* actual on-disk path */
102 	int fd;
103 	uint8_t *map;
104 	size_t len;
105 	size_t nlargeobj;
106 	struct got_packidx_v2_hdr hdr; /* convenient pointers into map */
107 };
108 
109 struct got_packfile_hdr {
110 	uint32_t	signature;
111 #define GOT_PACKFILE_SIGNATURE	0x5041434b	/* 'P' 'A' 'C' 'K' */
112 	uint32_t	version;	/* big endian */
113 #define GOT_PACKFILE_VERSION 2
114 	uint32_t	nobjects;	/* big endian */
115 };
116 
117 struct got_packfile_obj_hdr {
118 	/* 
119 	 * The object size field uses a variable length encoding:
120 	 * size0...sizeN form a 4+7+7+...+7 bit integer, where size0 is the
121 	 * least significant part and sizeN is the most significant part.
122 	 * If the MSB of a size byte is set, an additional size byte follows.
123 	 * Of the 7 remaining bits of size0, the first 3 bits indicate the
124 	 * object's type, and the remaining 4 bits contribute to the size.
125 	 */
126 	uint8_t *size;		/* variable length */
127 #define GOT_PACK_OBJ_SIZE_MORE		0x80
128 #define GOT_PACK_OBJ_SIZE0_TYPE_MASK	0x70 /* See struct got_object->type */
129 #define GOT_PACK_OBJ_SIZE0_TYPE_MASK_SHIFT	4
130 #define GOT_PACK_OBJ_SIZE0_VAL_MASK	0x0f
131 #define GOT_PACK_OBJ_SIZE_VAL_MASK	0x7f
132 };
133 
134 /* If object is not a DELTA type. */
135 struct got_packfile_object_data {
136 	uint8_t *data;	/* compressed */
137 };
138 
139 /* If object is of type	GOT_OBJ_TYPE_REF_DELTA. */
140 struct got_packfile_object_data_ref_delta {
141 	uint8_t sha1[SHA1_DIGEST_LENGTH];
142 	uint8_t *delta_data;		/* compressed */
143 };
144 
145 /* If object is of type GOT_OBJ_TYPE_OFFSET_DELTA. */
146 struct got_packfile_object_data_offset_delta {
147 	/* 
148 	 * This offset is interpreted as a negative offset from
149 	 * the got_packfile_obj_hdr corresponding to this object.
150 	 * The size provided in the header specifies the amount
151 	 * of compressed delta data that follows.
152 	 *
153 	 * This field uses a variable length encoding of N bytes,
154 	 * where the MSB is always set except for the last byte.
155 	 * The value is encoded as a series of N 7 bit integers,
156 	 * which are concatenated, and if N > 1 the value 2^7 +
157 	 * 2^14 + ... + 2^(7 * (n-1)) is added to the result.
158 	 */
159 	uint8_t *offset;	/* variable length */
160 #define GOT_PACK_OBJ_DELTA_OFF_MORE		0x80
161 #define GOT_PACK_OBJ_DELTA_OFF_VAL_MASK		0x7f
162 	uint8_t *delta_data;		/* compressed */
163 };
164 
165 struct got_packfile_obj_data {
166 	union {
167 		struct got_packfile_object_data data;
168 		struct got_packfile_object_data_ref_delta ref_delta;
169 		struct got_packfile_object_data_offset_delta offset_delta;
170 	} __attribute__((__packed__));
171 } __attribute__((__packed__));
172 
173 const struct got_error *got_packidx_init_hdr(struct got_packidx *, int, off_t);
174 const struct got_error *got_packidx_open(struct got_packidx **,
175     int, const char *, int);
176 const struct got_error *got_packidx_close(struct got_packidx *);
177 const struct got_error *got_packidx_get_packfile_path(char **, const char *);
178 off_t got_packidx_get_object_offset(struct got_packidx *, int idx);
179 int got_packidx_get_object_idx(struct got_packidx *, struct got_object_id *);
180 const struct got_error *got_packidx_match_id_str_prefix(
181     struct got_object_id_queue *, struct got_packidx *, const char *);
182 
183 const struct got_error *got_packfile_open_object(struct got_object **,
184     struct got_pack *, struct got_packidx *, int, struct got_object_id *);
185 const struct got_error *got_pack_get_delta_chain_max_size(uint64_t *,
186     struct got_delta_chain *, struct got_pack *);
187 const struct got_error *got_pack_get_max_delta_object_size(uint64_t *,
188     struct got_object *, struct got_pack *);
189 const struct got_error *got_pack_dump_delta_chain_to_file(size_t *,
190     struct got_delta_chain *, struct got_pack *, FILE *, FILE *, FILE *);
191 const struct got_error *got_pack_dump_delta_chain_to_mem(uint8_t **, size_t *,
192     struct got_delta_chain *, struct got_pack *);
193 const struct got_error *got_packfile_extract_object(struct got_pack *,
194     struct got_object *, FILE *, FILE *, FILE *);
195 const struct got_error *got_packfile_extract_object_to_mem(uint8_t **, size_t *,
196     struct got_object *, struct got_pack *);
197 struct got_pack *got_repo_get_cached_pack(struct got_repository *,
198     const char *);