1c588184b28658f9fdaee458019b9c81e5e8f252
1#ifndef PACK_OBJECTS_H
2#define PACK_OBJECTS_H
3#include "object-store.h"
5#define OE_DFS_STATE_BITS 2
7#define OE_DEPTH_BITS 12
8#define OE_IN_PACK_BITS 10
9#define OE_Z_DELTA_BITS 20
10/*
11 * Note that oe_set_size() becomes expensive when the given size is
12 * above this limit. Don't lower it too much.
13 */
14#define OE_SIZE_BITS 31
15#define OE_DELTA_SIZE_BITS 20
16/*
18 * State flags for depth-first search used for analyzing delta cycles.
19 *
20 * The depth is measured in delta-links to the base (so if A is a delta
21 * against B, then A has a depth of 1, and B a depth of 0).
22 */
23enum dfs_state {
24 DFS_NONE = 0,
25 DFS_ACTIVE,
26 DFS_DONE,
27 DFS_NUM_STATES
28};
29/*
31 * basic object info
32 * -----------------
33 * idx.oid is filled up before delta searching starts. idx.crc32 is
34 * only valid after the object is written out and will be used for
35 * generating the index. idx.offset will be both gradually set and
36 * used in writing phase (base objects get offset first, then deltas
37 * refer to them)
38 *
39 * "size" is the uncompressed object size. Compressed size of the raw
40 * data for an object in a pack is not stored anywhere but is computed
41 * and made available when reverse .idx is made. Note that when a
42 * delta is reused, "size" is the uncompressed _delta_ size, not the
43 * canonical one after the delta has been applied.
44 *
45 * "hash" contains a path name hash which is used for sorting the
46 * delta list and also during delta searching. Once prepare_pack()
47 * returns it's no longer needed.
48 *
49 * source pack info
50 * ----------------
51 * The (in_pack, in_pack_offset) tuple contains the location of the
52 * object in the source pack. in_pack_header_size allows quickly
53 * skipping the header and going straight to the zlib stream.
54 *
55 * "type" and "in_pack_type" both describe object type. in_pack_type
56 * may contain a delta type, while type is always the canonical type.
57 *
58 * deltas
59 * ------
60 * Delta links (delta, delta_child and delta_sibling) are created to
61 * reflect that delta graph from the source pack then updated or added
62 * during delta searching phase when we find better deltas.
63 *
64 * delta_child and delta_sibling are last needed in
65 * compute_write_order(). "delta" and "delta_size" must remain valid
66 * at object writing phase in case the delta is not cached.
67 *
68 * If a delta is cached in memory and is compressed, delta_data points
69 * to the data and z_delta_size contains the compressed size. If it's
70 * uncompressed [1], z_delta_size must be zero. delta_size is always
71 * the uncompressed size and must be valid even if the delta is not
72 * cached.
73 *
74 * [1] during try_delta phase we don't bother with compressing because
75 * the delta could be quickly replaced with a better one.
76 */
77struct object_entry {
78 struct pack_idx_entry idx;
79 unsigned size_:OE_SIZE_BITS;
80 unsigned size_valid:1;
81 unsigned in_pack_idx:OE_IN_PACK_BITS; /* already in pack */
82 off_t in_pack_offset;
83 uint32_t delta_idx; /* delta base object */
84 uint32_t delta_child_idx; /* deltified objects who bases me */
85 uint32_t delta_sibling_idx; /* other deltified objects who
86 * uses the same base as me
87 */
88 void *delta_data; /* cached delta (uncompressed) */
89 unsigned delta_size_:OE_DELTA_SIZE_BITS; /* delta data size (uncompressed) */
90 unsigned delta_size_valid:1;
91 unsigned z_delta_size:OE_Z_DELTA_BITS;
92 unsigned type_:TYPE_BITS;
93 unsigned in_pack_type:TYPE_BITS; /* could be delta */
94 unsigned type_valid:1;
95 uint32_t hash; /* name hint hash */
96 unsigned char in_pack_header_size;
97 unsigned preferred_base:1; /*
98 * we do not pack this, but is available
99 * to be used as the base object to delta
100 * objects against.
101 */
102 unsigned no_try_delta:1;
103 unsigned tagged:1; /* near the very tip of refs */
104 unsigned filled:1; /* assigned write-order */
105 unsigned dfs_state:OE_DFS_STATE_BITS;
106 unsigned depth:OE_DEPTH_BITS;
107};
108struct packing_data {
110 struct object_entry *objects;
111 uint32_t nr_objects, nr_alloc;
112 int32_t *index;
114 uint32_t index_size;
115 unsigned int *in_pack_pos;
117 /*
119 * Only one of these can be non-NULL and they have different
120 * sizes. if in_pack_by_idx is allocated, oe_in_pack() returns
121 * the pack of an object using in_pack_idx field. If not,
122 * in_pack[] array is used the same way as in_pack_pos[]
123 */
124 struct packed_git **in_pack_by_idx;
125 struct packed_git **in_pack;
126 uintmax_t oe_size_limit;
128};
129void prepare_packing_data(struct packing_data *pdata);
131struct object_entry *packlist_alloc(struct packing_data *pdata,
132 const unsigned char *sha1,
133 uint32_t index_pos);
134struct object_entry *packlist_find(struct packing_data *pdata,
136 const unsigned char *sha1,
137 uint32_t *index_pos);
138static inline uint32_t pack_name_hash(const char *name)
140{
141 uint32_t c, hash = 0;
142 if (!name)
144 return 0;
145 /*
147 * This effectively just creates a sortable number from the
148 * last sixteen non-whitespace characters. Last characters
149 * count "most", so things that end in ".c" sort together.
150 */
151 while ((c = *name++) != 0) {
152 if (isspace(c))
153 continue;
154 hash = (hash >> 2) + (c << 24);
155 }
156 return hash;
157}
158static inline enum object_type oe_type(const struct object_entry *e)
160{
161 return e->type_valid ? e->type_ : OBJ_BAD;
162}
163static inline void oe_set_type(struct object_entry *e,
165 enum object_type type)
166{
167 if (type >= OBJ_ANY)
168 BUG("OBJ_ANY cannot be set in pack-objects code");
169 e->type_valid = type >= OBJ_NONE;
171 e->type_ = (unsigned)type;
172}
173static inline unsigned int oe_in_pack_pos(const struct packing_data *pack,
175 const struct object_entry *e)
176{
177 return pack->in_pack_pos[e - pack->objects];
178}
179static inline void oe_set_in_pack_pos(const struct packing_data *pack,
181 const struct object_entry *e,
182 unsigned int pos)
183{
184 pack->in_pack_pos[e - pack->objects] = pos;
185}
186static inline struct packed_git *oe_in_pack(const struct packing_data *pack,
188 const struct object_entry *e)
189{
190 if (pack->in_pack_by_idx)
191 return pack->in_pack_by_idx[e->in_pack_idx];
192 else
193 return pack->in_pack[e - pack->objects];
194}
195void oe_map_new_pack(struct packing_data *pack,
197 struct packed_git *p);
198static inline void oe_set_in_pack(struct packing_data *pack,
199 struct object_entry *e,
200 struct packed_git *p)
201{
202 if (!p->index)
203 oe_map_new_pack(pack, p);
204 if (pack->in_pack_by_idx)
205 e->in_pack_idx = p->index;
206 else
207 pack->in_pack[e - pack->objects] = p;
208}
209static inline struct object_entry *oe_delta(
211 const struct packing_data *pack,
212 const struct object_entry *e)
213{
214 if (e->delta_idx)
215 return &pack->objects[e->delta_idx - 1];
216 return NULL;
217}
218static inline void oe_set_delta(struct packing_data *pack,
220 struct object_entry *e,
221 struct object_entry *delta)
222{
223 if (delta)
224 e->delta_idx = (delta - pack->objects) + 1;
225 else
226 e->delta_idx = 0;
227}
228static inline struct object_entry *oe_delta_child(
230 const struct packing_data *pack,
231 const struct object_entry *e)
232{
233 if (e->delta_child_idx)
234 return &pack->objects[e->delta_child_idx - 1];
235 return NULL;
236}
237static inline void oe_set_delta_child(struct packing_data *pack,
239 struct object_entry *e,
240 struct object_entry *delta)
241{
242 if (delta)
243 e->delta_child_idx = (delta - pack->objects) + 1;
244 else
245 e->delta_child_idx = 0;
246}
247static inline struct object_entry *oe_delta_sibling(
249 const struct packing_data *pack,
250 const struct object_entry *e)
251{
252 if (e->delta_sibling_idx)
253 return &pack->objects[e->delta_sibling_idx - 1];
254 return NULL;
255}
256static inline void oe_set_delta_sibling(struct packing_data *pack,
258 struct object_entry *e,
259 struct object_entry *delta)
260{
261 if (delta)
262 e->delta_sibling_idx = (delta - pack->objects) + 1;
263 else
264 e->delta_sibling_idx = 0;
265}
266unsigned long oe_get_size_slow(struct packing_data *pack,
268 const struct object_entry *e);
269static inline unsigned long oe_size(struct packing_data *pack,
270 const struct object_entry *e)
271{
272 if (e->size_valid)
273 return e->size_;
274 return oe_get_size_slow(pack, e);
276}
277static inline int oe_size_less_than(struct packing_data *pack,
279 const struct object_entry *lhs,
280 unsigned long rhs)
281{
282 if (lhs->size_valid)
283 return lhs->size_ < rhs;
284 if (rhs < pack->oe_size_limit) /* rhs < 2^x <= lhs ? */
285 return 0;
286 return oe_get_size_slow(pack, lhs) < rhs;
287}
288static inline int oe_size_greater_than(struct packing_data *pack,
290 const struct object_entry *lhs,
291 unsigned long rhs)
292{
293 if (lhs->size_valid)
294 return lhs->size_ > rhs;
295 if (rhs < pack->oe_size_limit) /* rhs < 2^x <= lhs ? */
296 return 1;
297 return oe_get_size_slow(pack, lhs) > rhs;
298}
299static inline void oe_set_size(struct packing_data *pack,
301 struct object_entry *e,
302 unsigned long size)
303{
304 if (size < pack->oe_size_limit) {
305 e->size_ = size;
306 e->size_valid = 1;
307 } else {
308 e->size_valid = 0;
309 if (oe_get_size_slow(pack, e) != size)
310 BUG("'size' is supposed to be the object size!");
311 }
312}
313static inline unsigned long oe_delta_size(struct packing_data *pack,
315 const struct object_entry *e)
316{
317 if (e->delta_size_valid)
318 return e->delta_size_;
319 return oe_size(pack, e);
320}
321static inline void oe_set_delta_size(struct packing_data *pack,
323 struct object_entry *e,
324 unsigned long size)
325{
326 e->delta_size_ = size;
327 e->delta_size_valid = e->delta_size_ == size;
328 if (!e->delta_size_valid && size != oe_size(pack, e))
329 BUG("this can only happen in check_object() "
330 "where delta size is the same as entry size");
331}
332#endif