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