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