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 	int child_has_delta_outfd;
26 	struct got_delta_cache *delta_cache;
27 };
28 
29 struct got_packidx;
30 
31 const struct got_error *got_pack_close(struct got_pack *);
32 
33 const struct got_error *got_pack_parse_offset_delta(off_t *, size_t *,
34     struct got_pack *, off_t, int);
35 const struct got_error *got_pack_parse_ref_delta(struct got_object_id *,
36     struct got_pack *, off_t, int);
37 const struct got_error *got_pack_resolve_delta_chain(struct got_delta_chain *,
38     struct got_packidx *, struct got_pack *, off_t, size_t, int, size_t,
39     unsigned int);
40 const struct got_error *got_pack_parse_object_type_and_size(uint8_t *,
41     uint64_t *, size_t *, struct got_pack *, off_t);
42 
43 #define GOT_PACK_PREFIX		"pack-"
44 #define GOT_PACKFILE_SUFFIX	".pack"
45 #define GOT_PACKIDX_SUFFIX		".idx"
46 #define GOT_PACKFILE_NAMELEN	(strlen(GOT_PACK_PREFIX) + \
47 				SHA1_DIGEST_STRING_LENGTH - 1 + \
48 				strlen(GOT_PACKFILE_SUFFIX))
49 #define GOT_PACKIDX_NAMELEN	(strlen(GOT_PACK_PREFIX) + \
50 				SHA1_DIGEST_STRING_LENGTH - 1 + \
51 				strlen(GOT_PACKIDX_SUFFIX))
52 
53 /* See Documentation/technical/pack-format.txt in Git. */
54 
55 struct got_packidx_trailer {
56 	u_int8_t	packfile_sha1[SHA1_DIGEST_LENGTH];
57 	u_int8_t	packidx_sha1[SHA1_DIGEST_LENGTH];
58 } __attribute__((__packed__));
59 
60 struct got_packidx_object_id {
61 	u_int8_t sha1[SHA1_DIGEST_LENGTH];
62 } __attribute__((__packed__));
63 
64 /* Ignore pack index version 1 which is no longer written by Git. */
65 #define GOT_PACKIDX_VERSION 2
66 
67 struct got_packidx_v2_hdr {
68 	uint32_t	*magic;		/* big endian */
69 #define GOT_PACKIDX_V2_MAGIC 0xff744f63	/* "\377t0c" */
70 	uint32_t	*version;
71 
72 	/* 
73 	 * Each entry N in the fanout table contains the number of objects in
74 	 * the packfile whose SHA1 begins with a byte less than or equal to N.
75 	 * The last entry (index 255) contains the number of objects in the
76 	 * pack file whose first SHA1 byte is <= 0xff, and thus records the
77 	 * total number of objects in the pack file. All pointer variables
78 	 * below point to tables with a corresponding number of entries.
79 	 */
80 	uint32_t	*fanout_table;	/* values are big endian */
81 #define GOT_PACKIDX_V2_FANOUT_TABLE_ITEMS (0xff + 1)
82 
83 	/* Sorted SHA1 checksums for each object in the pack file. */
84 	struct got_packidx_object_id *sorted_ids;
85 
86 	/* CRC32 of the packed representation of each object. */
87 	uint32_t	*crc32;
88 
89 	/* Offset into the pack file for each object. */
90 	uint32_t	*offsets;		/* values are big endian */
91 #define GOT_PACKIDX_OFFSET_VAL_MASK		0x7fffffff
92 #define GOT_PACKIDX_OFFSET_VAL_IS_LARGE_IDX	0x80000000
93 
94 	/* Large offsets table is empty for pack files < 2 GB. */
95 	uint64_t	*large_offsets;		/* values are big endian */
96 
97 	struct got_packidx_trailer *trailer;
98 };
99 
100 struct got_pack_offset_index {
101 	uint32_t offset;
102 	uint32_t idx;
103 };
104 
105 struct got_pack_large_offset_index {
106 	uint64_t offset;
107 	uint32_t idx;
108 };
109 
110 /* An open pack index file. */
111 struct got_packidx {
112 	char *path_packidx; /* actual on-disk path */
113 	int fd;
114 	uint8_t *map;
115 	size_t len;
116 	size_t nlargeobj;
117 	struct got_packidx_v2_hdr hdr; /* convenient pointers into map */
118 	struct got_pack_offset_index *sorted_offsets;
119 	struct got_pack_large_offset_index *sorted_large_offsets;
120 };
121 
122 struct got_packfile_hdr {
123 	uint32_t	signature;
124 #define GOT_PACKFILE_SIGNATURE	0x5041434b	/* 'P' 'A' 'C' 'K' */
125 	uint32_t	version;	/* big endian */
126 #define GOT_PACKFILE_VERSION 2
127 	uint32_t	nobjects;	/* big endian */
128 };
129 
130 struct got_packfile_obj_hdr {
131 	/* 
132 	 * The object size field uses a variable length encoding:
133 	 * size0...sizeN form a 4+7+7+...+7 bit integer, where size0 is the
134 	 * least significant part and sizeN is the most significant part.
135 	 * If the MSB of a size byte is set, an additional size byte follows.
136 	 * Of the 7 remaining bits of size0, the first 3 bits indicate the
137 	 * object's type, and the remaining 4 bits contribute to the size.
138 	 */
139 	uint8_t *size;		/* variable length */
140 #define GOT_PACK_OBJ_SIZE_MORE		0x80
141 #define GOT_PACK_OBJ_SIZE0_TYPE_MASK	0x70 /* See struct got_object->type */
142 #define GOT_PACK_OBJ_SIZE0_TYPE_MASK_SHIFT	4
143 #define GOT_PACK_OBJ_SIZE0_VAL_MASK	0x0f
144 #define GOT_PACK_OBJ_SIZE_VAL_MASK	0x7f
145 };
146 
147 /* If object is not a DELTA type. */
148 struct got_packfile_object_data {
149 	uint8_t *data;	/* compressed */
150 };
151 
152 /* If object is of type	GOT_OBJ_TYPE_REF_DELTA. */
153 struct got_packfile_object_data_ref_delta {
154 	uint8_t sha1[SHA1_DIGEST_LENGTH];
155 	uint8_t *delta_data;		/* compressed */
156 };
157 
158 /* If object is of type GOT_OBJ_TYPE_OFFSET_DELTA. */
159 struct got_packfile_object_data_offset_delta {
160 	/* 
161 	 * This offset is interpreted as a negative offset from
162 	 * the got_packfile_obj_hdr corresponding to this object.
163 	 * The size provided in the header specifies the amount
164 	 * of compressed delta data that follows.
165 	 *
166 	 * This field uses a variable length encoding of N bytes,
167 	 * where the MSB is always set except for the last byte.
168 	 * The value is encoded as a series of N 7 bit integers,
169 	 * which are concatenated, and if N > 1 the value 2^7 +
170 	 * 2^14 + ... + 2^(7 * (n-1)) is added to the result.
171 	 */
172 	uint8_t *offset;	/* variable length */
173 #define GOT_PACK_OBJ_DELTA_OFF_MORE		0x80
174 #define GOT_PACK_OBJ_DELTA_OFF_VAL_MASK		0x7f
175 	uint8_t *delta_data;		/* compressed */
176 };
177 
178 struct got_packfile_obj_data {
179 	union {
180 		struct got_packfile_object_data data;
181 		struct got_packfile_object_data_ref_delta ref_delta;
182 		struct got_packfile_object_data_offset_delta offset_delta;
183 	} __attribute__((__packed__));
184 } __attribute__((__packed__));
185 
186 const struct got_error *got_packidx_init_hdr(struct got_packidx *, int, off_t);
187 const struct got_error *got_packidx_open(struct got_packidx **,
188     int, const char *, int);
189 const struct got_error *got_packidx_close(struct got_packidx *);
190 const struct got_error *got_packidx_get_packfile_path(char **, const char *);
191 off_t got_packidx_get_object_offset(struct got_packidx *, int idx);
192 int got_packidx_get_object_idx(struct got_packidx *, struct got_object_id *);
193 const struct got_error *got_packidx_get_offset_idx(int *, struct got_packidx *,
194     off_t);
195 const struct got_error *got_packidx_get_object_id(struct got_object_id *,
196     struct got_packidx *, int);
197 const struct got_error *got_packidx_match_id_str_prefix(
198     struct got_object_id_queue *, struct got_packidx *, const char *);
199 
200 const struct got_error *got_packfile_open_object(struct got_object **,
201     struct got_pack *, struct got_packidx *, int, struct got_object_id *);
202 const struct got_error *got_pack_get_delta_chain_max_size(uint64_t *,
203     struct got_delta_chain *, struct got_pack *);
204 const struct got_error *got_pack_get_max_delta_object_size(uint64_t *,
205     struct got_object *, struct got_pack *);
206 const struct got_error *got_pack_dump_delta_chain_to_file(size_t *,
207     struct got_delta_chain *, struct got_pack *, FILE *, FILE *, FILE *);
208 const struct got_error *got_pack_dump_delta_chain_to_mem(uint8_t **, size_t *,
209     struct got_delta_chain *, struct got_pack *);
210 const struct got_error *got_packfile_extract_object(struct got_pack *,
211     struct got_object *, FILE *, FILE *, FILE *);
212 const struct got_error *got_packfile_extract_object_to_mem(uint8_t **, size_t *,
213     struct got_object *, struct got_pack *);
214 const struct got_error *got_packfile_extract_raw_delta(uint8_t **, size_t *,
215     size_t *, off_t *, off_t *, struct got_object_id *, uint64_t *, uint64_t *,
216     struct got_pack *, struct got_packidx *, int);
217 struct got_pack *got_repo_get_cached_pack(struct got_repository *,
218     const char *);