1 /*
2  * Copyright (c) 2018, 2019 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 };
25 
26 const struct got_error *got_pack_stop_privsep_child(struct got_pack *);
27 const struct got_error *got_pack_close(struct got_pack *);
28 
29 #define GOT_PACK_PREFIX		"pack-"
30 #define GOT_PACKFILE_SUFFIX	".pack"
31 #define GOT_PACKIDX_SUFFIX		".idx"
32 #define GOT_PACKFILE_NAMELEN	(strlen(GOT_PACK_PREFIX) + \
33 				SHA1_DIGEST_STRING_LENGTH - 1 + \
34 				strlen(GOT_PACKFILE_SUFFIX))
35 #define GOT_PACKIDX_NAMELEN	(strlen(GOT_PACK_PREFIX) + \
36 				SHA1_DIGEST_STRING_LENGTH - 1 + \
37 				strlen(GOT_PACKIDX_SUFFIX))
38 
39 /* See Documentation/technical/pack-format.txt in Git. */
40 
41 struct got_packidx_trailer {
42 	u_int8_t	packfile_sha1[SHA1_DIGEST_LENGTH];
43 	u_int8_t	packidx_sha1[SHA1_DIGEST_LENGTH];
44 } __attribute__((__packed__));
45 
46 struct got_packidx_object_id {
47 	u_int8_t sha1[SHA1_DIGEST_LENGTH];
48 } __attribute__((__packed__));
49 
50 /* Ignore pack index version 1 which is no longer written by Git. */
51 #define GOT_PACKIDX_VERSION 2
52 
53 struct got_packidx_v2_hdr {
54 	uint32_t	*magic;		/* big endian */
55 #define GOT_PACKIDX_V2_MAGIC 0xff744f63	/* "\377t0c" */
56 	uint32_t	*version;
57 
58 	/* 
59 	 * Each entry N in the fanout table contains the number of objects in
60 	 * the packfile whose SHA1 begins with a byte less than or equal to N.
61 	 * The last entry (index 255) contains the number of objects in the
62 	 * pack file whose first SHA1 byte is <= 0xff, and thus records the
63 	 * total number of objects in the pack file. All pointer variables
64 	 * below point to tables with a corresponding number of entries.
65 	 */
66 	uint32_t	*fanout_table;	/* values are big endian */
67 #define GOT_PACKIDX_V2_FANOUT_TABLE_ITEMS (0xff + 1)
68 
69 	/* Sorted SHA1 checksums for each object in the pack file. */
70 	struct got_packidx_object_id *sorted_ids;
71 
72 	/* CRC32 of the packed representation of each object. */
73 	uint32_t	*crc32;
74 
75 	/* Offset into the pack file for each object. */
76 	uint32_t	*offsets;		/* values are big endian */
77 #define GOT_PACKIDX_OFFSET_VAL_MASK		0x7fffffff
78 #define GOT_PACKIDX_OFFSET_VAL_IS_LARGE_IDX	0x80000000
79 
80 	/* Large offsets table is empty for pack files < 2 GB. */
81 	uint64_t	*large_offsets;		/* values are big endian */
82 
83 	struct got_packidx_trailer *trailer;
84 };
85 
86 /* An open pack index file. */
87 struct got_packidx {
88 	char *path_packidx; /* actual on-disk path */
89 	int fd;
90 	uint8_t *map;
91 	size_t len;
92 	struct got_packidx_v2_hdr hdr; /* convenient pointers into map */
93 };
94 
95 struct got_packfile_hdr {
96 	uint32_t	signature;
97 #define GOT_PACKFILE_SIGNATURE	0x5041434b	/* 'P' 'A' 'C' 'K' */
98 	uint32_t	version;	/* big endian */
99 #define GOT_PACKFILE_VERSION 2
100 	uint32_t	nobjects;	/* big endian */
101 };
102 
103 struct got_packfile_obj_hdr {
104 	/* 
105 	 * The object size field uses a variable length encoding:
106 	 * size0...sizeN form a 4+7+7+...+7 bit integer, where size0 is the
107 	 * least significant part and sizeN is the most significant part.
108 	 * If the MSB of a size byte is set, an additional size byte follows.
109 	 * Of the 7 remaining bits of size0, the first 3 bits indicate the
110 	 * object's type, and the remaining 4 bits contribute to the size.
111 	 */
112 	uint8_t *size;		/* variable length */
113 #define GOT_PACK_OBJ_SIZE_MORE		0x80
114 #define GOT_PACK_OBJ_SIZE0_TYPE_MASK	0x70 /* See struct got_object->type */
115 #define GOT_PACK_OBJ_SIZE0_TYPE_MASK_SHIFT	4
116 #define GOT_PACK_OBJ_SIZE0_VAL_MASK	0x0f
117 #define GOT_PACK_OBJ_SIZE_VAL_MASK	0x7f
118 };
119 
120 /* If object is not a DELTA type. */
121 struct got_packfile_object_data {
122 	uint8_t *data;	/* compressed */
123 };
124 
125 /* If object is of type	GOT_OBJ_TYPE_REF_DELTA. */
126 struct got_packfile_object_data_ref_delta {
127 	uint8_t sha1[SHA1_DIGEST_LENGTH];
128 	uint8_t *delta_data;		/* compressed */
129 };
130 
131 /* If object is of type GOT_OBJ_TYPE_OFFSET_DELTA. */
132 struct got_packfile_object_data_offset_delta {
133 	/* 
134 	 * This offset is interpreted as a negative offset from
135 	 * the got_packfile_obj_hdr corresponding to this object.
136 	 * The size provided in the header specifies the amount
137 	 * of compressed delta data that follows.
138 	 *
139 	 * This field uses a variable length encoding of N bytes,
140 	 * where the MSB is always set except for the last byte.
141 	 * The value is encoded as a series of N 7 bit integers,
142 	 * which are concatenated, and if N > 1 the value 2^7 +
143 	 * 2^14 + ... + 2^(7 * (n-1)) is added to the result.
144 	 */
145 	uint8_t *offset;	/* variable length */
146 #define GOT_PACK_OBJ_DELTA_OFF_MORE		0x80
147 #define GOT_PACK_OBJ_DELTA_OFF_VAL_MASK		0x7f
148 	uint8_t *delta_data;		/* compressed */
149 };
150 
151 struct got_packfile_obj_data {
152 	union {
153 		struct got_packfile_object_data data;
154 		struct got_packfile_object_data_ref_delta ref_delta;
155 		struct got_packfile_object_data_offset_delta offset_delta;
156 	} __attribute__((__packed__));
157 } __attribute__((__packed__));
158 
159 const struct got_error *got_packidx_init_hdr(struct got_packidx *, int);
160 const struct got_error *got_packidx_open(struct got_packidx **,
161     const char *, int);
162 const struct got_error *got_packidx_close(struct got_packidx *);
163 int got_packidx_get_object_idx(struct got_packidx *, struct got_object_id *);
164 const struct got_error *got_packidx_match_id_str_prefix(
165     struct got_object_id_queue *, struct got_packidx *, const char *);
166 
167 const struct got_error *got_packfile_open_object(struct got_object **,
168     struct got_pack *, struct got_packidx *, int, struct got_object_id *);
169 const struct got_error *got_pack_get_max_delta_object_size(uint64_t *,
170     struct got_object *);
171 const struct got_error *got_packfile_extract_object(struct got_pack *,
172     struct got_object *, FILE *, FILE *, FILE *);
173 const struct got_error *got_packfile_extract_object_to_mem(uint8_t **, size_t *,
174     struct got_object *, struct got_pack *);
175 struct got_pack *got_repo_get_cached_pack(struct got_repository *, const char *);