4 #include "object-store.h"
6 #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
11 * Note that oe_set_size() becomes expensive when the given size is
12 * above this limit. Don't lower it too much.
14 #define OE_SIZE_BITS 31
15 #define OE_DELTA_SIZE_BITS 20
18 * State flags for depth-first search used for analyzing delta cycles.
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).
31 * The size of struct nearly determines pack-objects's memory
32 * consumption. This struct is packed tight for that reason. When you
33 * add or reorder something in this struct, think a bit about this.
37 * idx.oid is filled up before delta searching starts. idx.crc32 is
38 * only valid after the object is written out and will be used for
39 * generating the index. idx.offset will be both gradually set and
40 * used in writing phase (base objects get offset first, then deltas
43 * "size" is the uncompressed object size. Compressed size of the raw
44 * data for an object in a pack is not stored anywhere but is computed
45 * and made available when reverse .idx is made. Note that when a
46 * delta is reused, "size" is the uncompressed _delta_ size, not the
47 * canonical one after the delta has been applied.
49 * "hash" contains a path name hash which is used for sorting the
50 * delta list and also during delta searching. Once prepare_pack()
51 * returns it's no longer needed.
55 * The (in_pack, in_pack_offset) tuple contains the location of the
56 * object in the source pack. in_pack_header_size allows quickly
57 * skipping the header and going straight to the zlib stream.
59 * "type" and "in_pack_type" both describe object type. in_pack_type
60 * may contain a delta type, while type is always the canonical type.
64 * Delta links (delta, delta_child and delta_sibling) are created to
65 * reflect that delta graph from the source pack then updated or added
66 * during delta searching phase when we find better deltas.
68 * delta_child and delta_sibling are last needed in
69 * compute_write_order(). "delta" and "delta_size" must remain valid
70 * at object writing phase in case the delta is not cached.
72 * If a delta is cached in memory and is compressed, delta_data points
73 * to the data and z_delta_size contains the compressed size. If it's
74 * uncompressed [1], z_delta_size must be zero. delta_size is always
75 * the uncompressed size and must be valid even if the delta is not
78 * [1] during try_delta phase we don't bother with compressing because
79 * the delta could be quickly replaced with a better one.
82 struct pack_idx_entry idx
;
83 void *delta_data
; /* cached delta (uncompressed) */
85 uint32_t hash
; /* name hint hash */
86 unsigned size_
:OE_SIZE_BITS
;
87 unsigned size_valid
:1;
88 uint32_t delta_idx
; /* delta base object */
89 uint32_t delta_child_idx
; /* deltified objects who bases me */
90 uint32_t delta_sibling_idx
; /* other deltified objects who
91 * uses the same base as me
93 unsigned delta_size_
:OE_DELTA_SIZE_BITS
; /* delta data size (uncompressed) */
94 unsigned delta_size_valid
:1;
95 unsigned in_pack_idx
:OE_IN_PACK_BITS
; /* already in pack */
96 unsigned z_delta_size
:OE_Z_DELTA_BITS
;
97 unsigned type_valid
:1;
98 unsigned type_
:TYPE_BITS
;
99 unsigned no_try_delta
:1;
100 unsigned in_pack_type
:TYPE_BITS
; /* could be delta */
101 unsigned preferred_base
:1; /*
102 * we do not pack this, but is available
103 * to be used as the base object to delta
106 unsigned tagged
:1; /* near the very tip of refs */
107 unsigned filled
:1; /* assigned write-order */
108 unsigned dfs_state
:OE_DFS_STATE_BITS
;
109 unsigned char in_pack_header_size
;
110 unsigned depth
:OE_DEPTH_BITS
;
113 * pahole results on 64-bit linux (gcc and clang)
115 * size: 80, bit_padding: 20 bits, holes: 8 bits
117 * and on 32-bit (gcc)
119 * size: 76, bit_padding: 20 bits, holes: 8 bits
123 struct packing_data
{
124 struct object_entry
*objects
;
125 uint32_t nr_objects
, nr_alloc
;
130 unsigned int *in_pack_pos
;
133 * Only one of these can be non-NULL and they have different
134 * sizes. if in_pack_by_idx is allocated, oe_in_pack() returns
135 * the pack of an object using in_pack_idx field. If not,
136 * in_pack[] array is used the same way as in_pack_pos[]
138 struct packed_git
**in_pack_by_idx
;
139 struct packed_git
**in_pack
;
141 uintmax_t oe_size_limit
;
144 void prepare_packing_data(struct packing_data
*pdata
);
145 struct object_entry
*packlist_alloc(struct packing_data
*pdata
,
146 const unsigned char *sha1
,
149 struct object_entry
*packlist_find(struct packing_data
*pdata
,
150 const unsigned char *sha1
,
151 uint32_t *index_pos
);
153 static inline uint32_t pack_name_hash(const char *name
)
155 uint32_t c
, hash
= 0;
161 * This effectively just creates a sortable number from the
162 * last sixteen non-whitespace characters. Last characters
163 * count "most", so things that end in ".c" sort together.
165 while ((c
= *name
++) != 0) {
168 hash
= (hash
>> 2) + (c
<< 24);
173 static inline enum object_type
oe_type(const struct object_entry
*e
)
175 return e
->type_valid
? e
->type_
: OBJ_BAD
;
178 static inline void oe_set_type(struct object_entry
*e
,
179 enum object_type type
)
182 BUG("OBJ_ANY cannot be set in pack-objects code");
184 e
->type_valid
= type
>= OBJ_NONE
;
185 e
->type_
= (unsigned)type
;
188 static inline unsigned int oe_in_pack_pos(const struct packing_data
*pack
,
189 const struct object_entry
*e
)
191 return pack
->in_pack_pos
[e
- pack
->objects
];
194 static inline void oe_set_in_pack_pos(const struct packing_data
*pack
,
195 const struct object_entry
*e
,
198 pack
->in_pack_pos
[e
- pack
->objects
] = pos
;
201 static inline struct packed_git
*oe_in_pack(const struct packing_data
*pack
,
202 const struct object_entry
*e
)
204 if (pack
->in_pack_by_idx
)
205 return pack
->in_pack_by_idx
[e
->in_pack_idx
];
207 return pack
->in_pack
[e
- pack
->objects
];
210 void oe_map_new_pack(struct packing_data
*pack
,
211 struct packed_git
*p
);
212 static inline void oe_set_in_pack(struct packing_data
*pack
,
213 struct object_entry
*e
,
214 struct packed_git
*p
)
217 oe_map_new_pack(pack
, p
);
218 if (pack
->in_pack_by_idx
)
219 e
->in_pack_idx
= p
->index
;
221 pack
->in_pack
[e
- pack
->objects
] = p
;
224 static inline struct object_entry
*oe_delta(
225 const struct packing_data
*pack
,
226 const struct object_entry
*e
)
229 return &pack
->objects
[e
->delta_idx
- 1];
233 static inline void oe_set_delta(struct packing_data
*pack
,
234 struct object_entry
*e
,
235 struct object_entry
*delta
)
238 e
->delta_idx
= (delta
- pack
->objects
) + 1;
243 static inline struct object_entry
*oe_delta_child(
244 const struct packing_data
*pack
,
245 const struct object_entry
*e
)
247 if (e
->delta_child_idx
)
248 return &pack
->objects
[e
->delta_child_idx
- 1];
252 static inline void oe_set_delta_child(struct packing_data
*pack
,
253 struct object_entry
*e
,
254 struct object_entry
*delta
)
257 e
->delta_child_idx
= (delta
- pack
->objects
) + 1;
259 e
->delta_child_idx
= 0;
262 static inline struct object_entry
*oe_delta_sibling(
263 const struct packing_data
*pack
,
264 const struct object_entry
*e
)
266 if (e
->delta_sibling_idx
)
267 return &pack
->objects
[e
->delta_sibling_idx
- 1];
271 static inline void oe_set_delta_sibling(struct packing_data
*pack
,
272 struct object_entry
*e
,
273 struct object_entry
*delta
)
276 e
->delta_sibling_idx
= (delta
- pack
->objects
) + 1;
278 e
->delta_sibling_idx
= 0;
281 unsigned long oe_get_size_slow(struct packing_data
*pack
,
282 const struct object_entry
*e
);
283 static inline unsigned long oe_size(struct packing_data
*pack
,
284 const struct object_entry
*e
)
289 return oe_get_size_slow(pack
, e
);
292 static inline int oe_size_less_than(struct packing_data
*pack
,
293 const struct object_entry
*lhs
,
297 return lhs
->size_
< rhs
;
298 if (rhs
< pack
->oe_size_limit
) /* rhs < 2^x <= lhs ? */
300 return oe_get_size_slow(pack
, lhs
) < rhs
;
303 static inline int oe_size_greater_than(struct packing_data
*pack
,
304 const struct object_entry
*lhs
,
308 return lhs
->size_
> rhs
;
309 if (rhs
< pack
->oe_size_limit
) /* rhs < 2^x <= lhs ? */
311 return oe_get_size_slow(pack
, lhs
) > rhs
;
314 static inline void oe_set_size(struct packing_data
*pack
,
315 struct object_entry
*e
,
318 if (size
< pack
->oe_size_limit
) {
323 if (oe_get_size_slow(pack
, e
) != size
)
324 BUG("'size' is supposed to be the object size!");
328 static inline unsigned long oe_delta_size(struct packing_data
*pack
,
329 const struct object_entry
*e
)
331 if (e
->delta_size_valid
)
332 return e
->delta_size_
;
333 return oe_size(pack
, e
);
336 static inline void oe_set_delta_size(struct packing_data
*pack
,
337 struct object_entry
*e
,
340 e
->delta_size_
= size
;
341 e
->delta_size_valid
= e
->delta_size_
== size
;
342 if (!e
->delta_size_valid
&& size
!= oe_size(pack
, e
))
343 BUG("this can only happen in check_object() "
344 "where delta size is the same as entry size");