]>
git.ipfire.org Git - thirdparty/git.git/blob - pack-objects.h
4 #include "object-store.h"
5 #include "thread-utils.h"
8 #define DEFAULT_DELTA_CACHE_SIZE (256 * 1024 * 1024)
10 #define OE_DFS_STATE_BITS 2
11 #define OE_DEPTH_BITS 12
12 #define OE_IN_PACK_BITS 10
13 #define OE_Z_DELTA_BITS 20
15 * Note that oe_set_size() becomes expensive when the given size is
16 * above this limit. Don't lower it too much.
18 #define OE_SIZE_BITS 31
19 #define OE_DELTA_SIZE_BITS 23
22 * State flags for depth-first search used for analyzing delta cycles.
24 * The depth is measured in delta-links to the base (so if A is a delta
25 * against B, then A has a depth of 1, and B a depth of 0).
35 * The size of struct nearly determines pack-objects's memory
36 * consumption. This struct is packed tight for that reason. When you
37 * add or reorder something in this struct, think a bit about this.
41 * idx.oid is filled up before delta searching starts. idx.crc32 is
42 * only valid after the object is written out and will be used for
43 * generating the index. idx.offset will be both gradually set and
44 * used in writing phase (base objects get offset first, then deltas
47 * "size" is the uncompressed object size. Compressed size of the raw
48 * data for an object in a pack is not stored anywhere but is computed
49 * and made available when reverse .idx is made. Note that when a
50 * delta is reused, "size" is the uncompressed _delta_ size, not the
51 * canonical one after the delta has been applied.
53 * "hash" contains a path name hash which is used for sorting the
54 * delta list and also during delta searching. Once prepare_pack()
55 * returns it's no longer needed.
59 * The (in_pack, in_pack_offset) tuple contains the location of the
60 * object in the source pack. in_pack_header_size allows quickly
61 * skipping the header and going straight to the zlib stream.
63 * "type" and "in_pack_type" both describe object type. in_pack_type
64 * may contain a delta type, while type is always the canonical type.
68 * Delta links (delta, delta_child and delta_sibling) are created to
69 * reflect that delta graph from the source pack then updated or added
70 * during delta searching phase when we find better deltas.
72 * delta_child and delta_sibling are last needed in
73 * compute_write_order(). "delta" and "delta_size" must remain valid
74 * at object writing phase in case the delta is not cached.
76 * If a delta is cached in memory and is compressed, delta_data points
77 * to the data and z_delta_size contains the compressed size. If it's
78 * uncompressed [1], z_delta_size must be zero. delta_size is always
79 * the uncompressed size and must be valid even if the delta is not
82 * [1] during try_delta phase we don't bother with compressing because
83 * the delta could be quickly replaced with a better one.
86 struct pack_idx_entry idx
;
87 void *delta_data
; /* cached delta (uncompressed) */
89 uint32_t hash
; /* name hint hash */
90 unsigned size_
:OE_SIZE_BITS
;
91 unsigned size_valid
:1;
92 uint32_t delta_idx
; /* delta base object */
93 uint32_t delta_child_idx
; /* deltified objects who bases me */
94 uint32_t delta_sibling_idx
; /* other deltified objects who
95 * uses the same base as me
97 unsigned delta_size_
:OE_DELTA_SIZE_BITS
; /* delta data size (uncompressed) */
98 unsigned delta_size_valid
:1;
99 unsigned char in_pack_header_size
;
100 unsigned in_pack_idx
:OE_IN_PACK_BITS
; /* already in pack */
101 unsigned z_delta_size
:OE_Z_DELTA_BITS
;
102 unsigned type_valid
:1;
103 unsigned no_try_delta
:1;
104 unsigned type_
:TYPE_BITS
;
105 unsigned in_pack_type
:TYPE_BITS
; /* could be delta */
107 unsigned preferred_base
:1; /*
108 * we do not pack this, but is available
109 * to be used as the base object to delta
112 unsigned tagged
:1; /* near the very tip of refs */
113 unsigned filled
:1; /* assigned write-order */
114 unsigned dfs_state
:OE_DFS_STATE_BITS
;
115 unsigned depth
:OE_DEPTH_BITS
;
116 unsigned ext_base
:1; /* delta_idx points outside packlist */
119 * pahole results on 64-bit linux (gcc and clang)
121 * size: 80, bit_padding: 9 bits
123 * and on 32-bit (gcc)
125 * size: 76, bit_padding: 9 bits
129 struct packing_data
{
130 struct object_entry
*objects
;
131 uint32_t nr_objects
, nr_alloc
;
136 unsigned int *in_pack_pos
;
137 unsigned long *delta_size
;
140 * Only one of these can be non-NULL and they have different
141 * sizes. if in_pack_by_idx is allocated, oe_in_pack() returns
142 * the pack of an object using in_pack_idx field. If not,
143 * in_pack[] array is used the same way as in_pack_pos[]
145 struct packed_git
**in_pack_by_idx
;
146 struct packed_git
**in_pack
;
149 * During packing with multiple threads, protect the in-core
150 * object database from concurrent accesses.
152 pthread_mutex_t odb_lock
;
155 * This list contains entries for bases which we know the other side
156 * has (e.g., via reachability bitmaps), but which aren't in our
159 struct object_entry
*ext_bases
;
160 uint32_t nr_ext
, alloc_ext
;
162 uintmax_t oe_size_limit
;
163 uintmax_t oe_delta_size_limit
;
166 unsigned int *tree_depth
;
167 unsigned char *layer
;
170 void prepare_packing_data(struct packing_data
*pdata
);
172 /* Protect access to object database */
173 static inline void packing_data_lock(struct packing_data
*pdata
)
175 pthread_mutex_lock(&pdata
->odb_lock
);
177 static inline void packing_data_unlock(struct packing_data
*pdata
)
179 pthread_mutex_unlock(&pdata
->odb_lock
);
182 struct object_entry
*packlist_alloc(struct packing_data
*pdata
,
183 const unsigned char *sha1
,
186 struct object_entry
*packlist_find(struct packing_data
*pdata
,
187 const unsigned char *sha1
,
188 uint32_t *index_pos
);
190 static inline uint32_t pack_name_hash(const char *name
)
192 uint32_t c
, hash
= 0;
198 * This effectively just creates a sortable number from the
199 * last sixteen non-whitespace characters. Last characters
200 * count "most", so things that end in ".c" sort together.
202 while ((c
= *name
++) != 0) {
205 hash
= (hash
>> 2) + (c
<< 24);
210 static inline enum object_type
oe_type(const struct object_entry
*e
)
212 return e
->type_valid
? e
->type_
: OBJ_BAD
;
215 static inline void oe_set_type(struct object_entry
*e
,
216 enum object_type type
)
219 BUG("OBJ_ANY cannot be set in pack-objects code");
221 e
->type_valid
= type
>= OBJ_NONE
;
222 e
->type_
= (unsigned)type
;
225 static inline unsigned int oe_in_pack_pos(const struct packing_data
*pack
,
226 const struct object_entry
*e
)
228 return pack
->in_pack_pos
[e
- pack
->objects
];
231 static inline void oe_set_in_pack_pos(const struct packing_data
*pack
,
232 const struct object_entry
*e
,
235 pack
->in_pack_pos
[e
- pack
->objects
] = pos
;
238 static inline struct packed_git
*oe_in_pack(const struct packing_data
*pack
,
239 const struct object_entry
*e
)
241 if (pack
->in_pack_by_idx
)
242 return pack
->in_pack_by_idx
[e
->in_pack_idx
];
244 return pack
->in_pack
[e
- pack
->objects
];
247 void oe_map_new_pack(struct packing_data
*pack
,
248 struct packed_git
*p
);
249 static inline void oe_set_in_pack(struct packing_data
*pack
,
250 struct object_entry
*e
,
251 struct packed_git
*p
)
254 oe_map_new_pack(pack
, p
);
255 if (pack
->in_pack_by_idx
)
256 e
->in_pack_idx
= p
->index
;
258 pack
->in_pack
[e
- pack
->objects
] = p
;
261 static inline struct object_entry
*oe_delta(
262 const struct packing_data
*pack
,
263 const struct object_entry
*e
)
268 return &pack
->ext_bases
[e
->delta_idx
- 1];
270 return &pack
->objects
[e
->delta_idx
- 1];
273 static inline void oe_set_delta(struct packing_data
*pack
,
274 struct object_entry
*e
,
275 struct object_entry
*delta
)
278 e
->delta_idx
= (delta
- pack
->objects
) + 1;
283 void oe_set_delta_ext(struct packing_data
*pack
,
284 struct object_entry
*e
,
285 const unsigned char *sha1
);
287 static inline struct object_entry
*oe_delta_child(
288 const struct packing_data
*pack
,
289 const struct object_entry
*e
)
291 if (e
->delta_child_idx
)
292 return &pack
->objects
[e
->delta_child_idx
- 1];
296 static inline void oe_set_delta_child(struct packing_data
*pack
,
297 struct object_entry
*e
,
298 struct object_entry
*delta
)
301 e
->delta_child_idx
= (delta
- pack
->objects
) + 1;
303 e
->delta_child_idx
= 0;
306 static inline struct object_entry
*oe_delta_sibling(
307 const struct packing_data
*pack
,
308 const struct object_entry
*e
)
310 if (e
->delta_sibling_idx
)
311 return &pack
->objects
[e
->delta_sibling_idx
- 1];
315 static inline void oe_set_delta_sibling(struct packing_data
*pack
,
316 struct object_entry
*e
,
317 struct object_entry
*delta
)
320 e
->delta_sibling_idx
= (delta
- pack
->objects
) + 1;
322 e
->delta_sibling_idx
= 0;
325 unsigned long oe_get_size_slow(struct packing_data
*pack
,
326 const struct object_entry
*e
);
327 static inline unsigned long oe_size(struct packing_data
*pack
,
328 const struct object_entry
*e
)
333 return oe_get_size_slow(pack
, e
);
336 static inline int oe_size_less_than(struct packing_data
*pack
,
337 const struct object_entry
*lhs
,
341 return lhs
->size_
< rhs
;
342 if (rhs
< pack
->oe_size_limit
) /* rhs < 2^x <= lhs ? */
344 return oe_get_size_slow(pack
, lhs
) < rhs
;
347 static inline int oe_size_greater_than(struct packing_data
*pack
,
348 const struct object_entry
*lhs
,
352 return lhs
->size_
> rhs
;
353 if (rhs
< pack
->oe_size_limit
) /* rhs < 2^x <= lhs ? */
355 return oe_get_size_slow(pack
, lhs
) > rhs
;
358 static inline void oe_set_size(struct packing_data
*pack
,
359 struct object_entry
*e
,
362 if (size
< pack
->oe_size_limit
) {
367 if (oe_get_size_slow(pack
, e
) != size
)
368 BUG("'size' is supposed to be the object size!");
372 static inline unsigned long oe_delta_size(struct packing_data
*pack
,
373 const struct object_entry
*e
)
375 if (e
->delta_size_valid
)
376 return e
->delta_size_
;
379 * pack->delta_size[] can't be NULL because oe_set_delta_size()
380 * must have been called when a new delta is saved with
382 * If oe_delta() returns NULL (i.e. default state, which means
383 * delta_size_valid is also false), then the caller must never
384 * call oe_delta_size().
386 return pack
->delta_size
[e
- pack
->objects
];
389 static inline void oe_set_delta_size(struct packing_data
*pack
,
390 struct object_entry
*e
,
393 if (size
< pack
->oe_delta_size_limit
) {
394 e
->delta_size_
= size
;
395 e
->delta_size_valid
= 1;
397 packing_data_lock(pack
);
398 if (!pack
->delta_size
)
399 ALLOC_ARRAY(pack
->delta_size
, pack
->nr_alloc
);
400 packing_data_unlock(pack
);
402 pack
->delta_size
[e
- pack
->objects
] = size
;
403 e
->delta_size_valid
= 0;
407 static inline unsigned int oe_tree_depth(struct packing_data
*pack
,
408 struct object_entry
*e
)
410 if (!pack
->tree_depth
)
412 return pack
->tree_depth
[e
- pack
->objects
];
415 static inline void oe_set_tree_depth(struct packing_data
*pack
,
416 struct object_entry
*e
,
417 unsigned int tree_depth
)
419 if (!pack
->tree_depth
)
420 CALLOC_ARRAY(pack
->tree_depth
, pack
->nr_alloc
);
421 pack
->tree_depth
[e
- pack
->objects
] = tree_depth
;
424 static inline unsigned char oe_layer(struct packing_data
*pack
,
425 struct object_entry
*e
)
429 return pack
->layer
[e
- pack
->objects
];
432 static inline void oe_set_layer(struct packing_data
*pack
,
433 struct object_entry
*e
,
437 CALLOC_ARRAY(pack
->layer
, pack
->nr_alloc
);
438 pack
->layer
[e
- pack
->objects
] = layer
;