4 #include "object-store.h"
7 #define DEFAULT_DELTA_CACHE_SIZE (256 * 1024 * 1024)
9 #define OE_DFS_STATE_BITS 2
10 #define OE_DEPTH_BITS 12
11 #define OE_IN_PACK_BITS 10
12 #define OE_Z_DELTA_BITS 20
14 * Note that oe_set_size() becomes expensive when the given size is
15 * above this limit. Don't lower it too much.
17 #define OE_SIZE_BITS 31
18 #define OE_DELTA_SIZE_BITS 20
21 * State flags for depth-first search used for analyzing delta cycles.
23 * The depth is measured in delta-links to the base (so if A is a delta
24 * against B, then A has a depth of 1, and B a depth of 0).
34 * The size of struct nearly determines pack-objects's memory
35 * consumption. This struct is packed tight for that reason. When you
36 * add or reorder something in this struct, think a bit about this.
40 * idx.oid is filled up before delta searching starts. idx.crc32 is
41 * only valid after the object is written out and will be used for
42 * generating the index. idx.offset will be both gradually set and
43 * used in writing phase (base objects get offset first, then deltas
46 * "size" is the uncompressed object size. Compressed size of the raw
47 * data for an object in a pack is not stored anywhere but is computed
48 * and made available when reverse .idx is made. Note that when a
49 * delta is reused, "size" is the uncompressed _delta_ size, not the
50 * canonical one after the delta has been applied.
52 * "hash" contains a path name hash which is used for sorting the
53 * delta list and also during delta searching. Once prepare_pack()
54 * returns it's no longer needed.
58 * The (in_pack, in_pack_offset) tuple contains the location of the
59 * object in the source pack. in_pack_header_size allows quickly
60 * skipping the header and going straight to the zlib stream.
62 * "type" and "in_pack_type" both describe object type. in_pack_type
63 * may contain a delta type, while type is always the canonical type.
67 * Delta links (delta, delta_child and delta_sibling) are created to
68 * reflect that delta graph from the source pack then updated or added
69 * during delta searching phase when we find better deltas.
71 * delta_child and delta_sibling are last needed in
72 * compute_write_order(). "delta" and "delta_size" must remain valid
73 * at object writing phase in case the delta is not cached.
75 * If a delta is cached in memory and is compressed, delta_data points
76 * to the data and z_delta_size contains the compressed size. If it's
77 * uncompressed [1], z_delta_size must be zero. delta_size is always
78 * the uncompressed size and must be valid even if the delta is not
81 * [1] during try_delta phase we don't bother with compressing because
82 * the delta could be quickly replaced with a better one.
85 struct pack_idx_entry idx;
86 void *delta_data; /* cached delta (uncompressed) */
88 uint32_t hash; /* name hint hash */
89 unsigned size_:OE_SIZE_BITS;
90 unsigned size_valid:1;
91 uint32_t delta_idx; /* delta base object */
92 uint32_t delta_child_idx; /* deltified objects who bases me */
93 uint32_t delta_sibling_idx; /* other deltified objects who
94 * uses the same base as me
96 unsigned delta_size_:OE_DELTA_SIZE_BITS; /* delta data size (uncompressed) */
97 unsigned delta_size_valid:1;
98 unsigned in_pack_idx:OE_IN_PACK_BITS; /* already in pack */
99 unsigned z_delta_size:OE_Z_DELTA_BITS;
100 unsigned type_valid:1;
101 unsigned type_:TYPE_BITS;
102 unsigned no_try_delta:1;
103 unsigned in_pack_type:TYPE_BITS; /* could be delta */
104 unsigned preferred_base:1; /*
105 * we do not pack this, but is available
106 * to be used as the base object to delta
109 unsigned tagged:1; /* near the very tip of refs */
110 unsigned filled:1; /* assigned write-order */
111 unsigned dfs_state:OE_DFS_STATE_BITS;
112 unsigned char in_pack_header_size;
113 unsigned depth:OE_DEPTH_BITS;
116 * pahole results on 64-bit linux (gcc and clang)
118 * size: 80, bit_padding: 20 bits, holes: 8 bits
120 * and on 32-bit (gcc)
122 * size: 76, bit_padding: 20 bits, holes: 8 bits
126 struct packing_data {
127 struct object_entry *objects;
128 uint32_t nr_objects, nr_alloc;
133 unsigned int *in_pack_pos;
136 * Only one of these can be non-NULL and they have different
137 * sizes. if in_pack_by_idx is allocated, oe_in_pack() returns
138 * the pack of an object using in_pack_idx field. If not,
139 * in_pack[] array is used the same way as in_pack_pos[]
141 struct packed_git **in_pack_by_idx;
142 struct packed_git **in_pack;
144 uintmax_t oe_size_limit;
147 void prepare_packing_data(struct packing_data *pdata);
148 struct object_entry *packlist_alloc(struct packing_data *pdata,
149 const unsigned char *sha1,
152 struct object_entry *packlist_find(struct packing_data *pdata,
153 const unsigned char *sha1,
154 uint32_t *index_pos);
156 static inline uint32_t pack_name_hash(const char *name)
158 uint32_t c, hash = 0;
164 * This effectively just creates a sortable number from the
165 * last sixteen non-whitespace characters. Last characters
166 * count "most", so things that end in ".c" sort together.
168 while ((c = *name++) != 0) {
171 hash = (hash >> 2) + (c << 24);
176 static inline enum object_type oe_type(const struct object_entry *e)
178 return e->type_valid ? e->type_ : OBJ_BAD;
181 static inline void oe_set_type(struct object_entry *e,
182 enum object_type type)
185 BUG("OBJ_ANY cannot be set in pack-objects code");
187 e->type_valid = type >= OBJ_NONE;
188 e->type_ = (unsigned)type;
191 static inline unsigned int oe_in_pack_pos(const struct packing_data *pack,
192 const struct object_entry *e)
194 return pack->in_pack_pos[e - pack->objects];
197 static inline void oe_set_in_pack_pos(const struct packing_data *pack,
198 const struct object_entry *e,
201 pack->in_pack_pos[e - pack->objects] = pos;
204 static inline struct packed_git *oe_in_pack(const struct packing_data *pack,
205 const struct object_entry *e)
207 if (pack->in_pack_by_idx)
208 return pack->in_pack_by_idx[e->in_pack_idx];
210 return pack->in_pack[e - pack->objects];
213 void oe_map_new_pack(struct packing_data *pack,
214 struct packed_git *p);
215 static inline void oe_set_in_pack(struct packing_data *pack,
216 struct object_entry *e,
217 struct packed_git *p)
220 oe_map_new_pack(pack, p);
221 if (pack->in_pack_by_idx)
222 e->in_pack_idx = p->index;
224 pack->in_pack[e - pack->objects] = p;
227 static inline struct object_entry *oe_delta(
228 const struct packing_data *pack,
229 const struct object_entry *e)
232 return &pack->objects[e->delta_idx - 1];
236 static inline void oe_set_delta(struct packing_data *pack,
237 struct object_entry *e,
238 struct object_entry *delta)
241 e->delta_idx = (delta - pack->objects) + 1;
246 static inline struct object_entry *oe_delta_child(
247 const struct packing_data *pack,
248 const struct object_entry *e)
250 if (e->delta_child_idx)
251 return &pack->objects[e->delta_child_idx - 1];
255 static inline void oe_set_delta_child(struct packing_data *pack,
256 struct object_entry *e,
257 struct object_entry *delta)
260 e->delta_child_idx = (delta - pack->objects) + 1;
262 e->delta_child_idx = 0;
265 static inline struct object_entry *oe_delta_sibling(
266 const struct packing_data *pack,
267 const struct object_entry *e)
269 if (e->delta_sibling_idx)
270 return &pack->objects[e->delta_sibling_idx - 1];
274 static inline void oe_set_delta_sibling(struct packing_data *pack,
275 struct object_entry *e,
276 struct object_entry *delta)
279 e->delta_sibling_idx = (delta - pack->objects) + 1;
281 e->delta_sibling_idx = 0;
284 unsigned long oe_get_size_slow(struct packing_data *pack,
285 const struct object_entry *e);
286 static inline unsigned long oe_size(struct packing_data *pack,
287 const struct object_entry *e)
292 return oe_get_size_slow(pack, e);
295 static inline int oe_size_less_than(struct packing_data *pack,
296 const struct object_entry *lhs,
300 return lhs->size_ < rhs;
301 if (rhs < pack->oe_size_limit) /* rhs < 2^x <= lhs ? */
303 return oe_get_size_slow(pack, lhs) < rhs;
306 static inline int oe_size_greater_than(struct packing_data *pack,
307 const struct object_entry *lhs,
311 return lhs->size_ > rhs;
312 if (rhs < pack->oe_size_limit) /* rhs < 2^x <= lhs ? */
314 return oe_get_size_slow(pack, lhs) > rhs;
317 static inline void oe_set_size(struct packing_data *pack,
318 struct object_entry *e,
321 if (size < pack->oe_size_limit) {
326 if (oe_get_size_slow(pack, e) != size)
327 BUG("'size' is supposed to be the object size!");
331 static inline unsigned long oe_delta_size(struct packing_data *pack,
332 const struct object_entry *e)
334 if (e->delta_size_valid)
335 return e->delta_size_;
336 return oe_size(pack, e);
339 static inline void oe_set_delta_size(struct packing_data *pack,
340 struct object_entry *e,
343 e->delta_size_ = size;
344 e->delta_size_valid = e->delta_size_ == size;
345 if (!e->delta_size_valid && size != oe_size(pack, e))
346 BUG("this can only happen in check_object() "
347 "where delta size is the same as entry size");