4 #include "object-store.h"
5 #include "thread-utils.h"
10 #define DEFAULT_DELTA_CACHE_SIZE (256 * 1024 * 1024)
12 #define OE_DFS_STATE_BITS 2
13 #define OE_DEPTH_BITS 12
14 #define OE_IN_PACK_BITS 10
15 #define OE_Z_DELTA_BITS 20
17 * Note that oe_set_size() becomes expensive when the given size is
18 * above this limit. Don't lower it too much.
20 #define OE_SIZE_BITS 31
21 #define OE_DELTA_SIZE_BITS 23
24 * State flags for depth-first search used for analyzing delta cycles.
26 * The depth is measured in delta-links to the base (so if A is a delta
27 * against B, then A has a depth of 1, and B a depth of 0).
37 * The size of struct nearly determines pack-objects's memory
38 * consumption. This struct is packed tight for that reason. When you
39 * add or reorder something in this struct, think a bit about this.
43 * idx.oid is filled up before delta searching starts. idx.crc32 is
44 * only valid after the object is written out and will be used for
45 * generating the index. idx.offset will be both gradually set and
46 * used in writing phase (base objects get offset first, then deltas
49 * "size" is the uncompressed object size. Compressed size of the raw
50 * data for an object in a pack is not stored anywhere but is computed
51 * and made available when reverse .idx is made. Note that when a
52 * delta is reused, "size" is the uncompressed _delta_ size, not the
53 * canonical one after the delta has been applied.
55 * "hash" contains a path name hash which is used for sorting the
56 * delta list and also during delta searching. Once prepare_pack()
57 * returns it's no longer needed.
61 * The (in_pack, in_pack_offset) tuple contains the location of the
62 * object in the source pack. in_pack_header_size allows quickly
63 * skipping the header and going straight to the zlib stream.
65 * "type" and "in_pack_type" both describe object type. in_pack_type
66 * may contain a delta type, while type is always the canonical type.
70 * Delta links (delta, delta_child and delta_sibling) are created to
71 * reflect that delta graph from the source pack then updated or added
72 * during delta searching phase when we find better deltas.
74 * delta_child and delta_sibling are last needed in
75 * compute_write_order(). "delta" and "delta_size" must remain valid
76 * at object writing phase in case the delta is not cached.
78 * If a delta is cached in memory and is compressed, delta_data points
79 * to the data and z_delta_size contains the compressed size. If it's
80 * uncompressed [1], z_delta_size must be zero. delta_size is always
81 * the uncompressed size and must be valid even if the delta is not
84 * [1] during try_delta phase we don't bother with compressing because
85 * the delta could be quickly replaced with a better one.
88 struct pack_idx_entry idx
;
89 void *delta_data
; /* cached delta (uncompressed) */
91 uint32_t hash
; /* name hint hash */
92 unsigned size_
:OE_SIZE_BITS
;
93 unsigned size_valid
:1;
94 uint32_t delta_idx
; /* delta base object */
95 uint32_t delta_child_idx
; /* deltified objects who bases me */
96 uint32_t delta_sibling_idx
; /* other deltified objects who
97 * uses the same base as me
99 unsigned delta_size_
:OE_DELTA_SIZE_BITS
; /* delta data size (uncompressed) */
100 unsigned delta_size_valid
:1;
101 unsigned char in_pack_header_size
;
102 unsigned in_pack_idx
:OE_IN_PACK_BITS
; /* already in pack */
103 unsigned z_delta_size
:OE_Z_DELTA_BITS
;
104 unsigned type_valid
:1;
105 unsigned no_try_delta
:1;
106 unsigned type_
:TYPE_BITS
;
107 unsigned in_pack_type
:TYPE_BITS
; /* could be delta */
109 unsigned preferred_base
:1; /*
110 * we do not pack this, but is available
111 * to be used as the base object to delta
114 unsigned tagged
:1; /* near the very tip of refs */
115 unsigned filled
:1; /* assigned write-order */
116 unsigned dfs_state
:OE_DFS_STATE_BITS
;
117 unsigned depth
:OE_DEPTH_BITS
;
118 unsigned ext_base
:1; /* delta_idx points outside packlist */
121 * pahole results on 64-bit linux (gcc and clang)
123 * size: 80, bit_padding: 9 bits
125 * and on 32-bit (gcc)
127 * size: 76, bit_padding: 9 bits
131 struct packing_data
{
132 struct repository
*repo
;
133 struct object_entry
*objects
;
134 uint32_t nr_objects
, nr_alloc
;
139 unsigned int *in_pack_pos
;
140 unsigned long *delta_size
;
143 * Only one of these can be non-NULL and they have different
144 * sizes. if in_pack_by_idx is allocated, oe_in_pack() returns
145 * the pack of an object using in_pack_idx field. If not,
146 * in_pack[] array is used the same way as in_pack_pos[]
148 struct packed_git
**in_pack_by_idx
;
149 struct packed_git
**in_pack
;
152 * During packing with multiple threads, protect the in-core
153 * object database from concurrent accesses.
155 pthread_mutex_t odb_lock
;
158 * This list contains entries for bases which we know the other side
159 * has (e.g., via reachability bitmaps), but which aren't in our
162 struct object_entry
*ext_bases
;
163 uint32_t nr_ext
, alloc_ext
;
165 uintmax_t oe_size_limit
;
166 uintmax_t oe_delta_size_limit
;
169 unsigned int *tree_depth
;
170 unsigned char *layer
;
173 void prepare_packing_data(struct repository
*r
, struct packing_data
*pdata
);
175 /* Protect access to object database */
176 static inline void packing_data_lock(struct packing_data
*pdata
)
178 pthread_mutex_lock(&pdata
->odb_lock
);
180 static inline void packing_data_unlock(struct packing_data
*pdata
)
182 pthread_mutex_unlock(&pdata
->odb_lock
);
185 struct object_entry
*packlist_alloc(struct packing_data
*pdata
,
186 const unsigned char *sha1
,
189 struct object_entry
*packlist_find(struct packing_data
*pdata
,
190 const struct object_id
*oid
,
191 uint32_t *index_pos
);
193 static inline uint32_t pack_name_hash(const char *name
)
195 uint32_t c
, hash
= 0;
201 * This effectively just creates a sortable number from the
202 * last sixteen non-whitespace characters. Last characters
203 * count "most", so things that end in ".c" sort together.
205 while ((c
= *name
++) != 0) {
208 hash
= (hash
>> 2) + (c
<< 24);
213 static inline enum object_type
oe_type(const struct object_entry
*e
)
215 return e
->type_valid
? e
->type_
: OBJ_BAD
;
218 static inline void oe_set_type(struct object_entry
*e
,
219 enum object_type type
)
222 BUG("OBJ_ANY cannot be set in pack-objects code");
224 e
->type_valid
= type
>= OBJ_NONE
;
225 e
->type_
= (unsigned)type
;
228 static inline unsigned int oe_in_pack_pos(const struct packing_data
*pack
,
229 const struct object_entry
*e
)
231 return pack
->in_pack_pos
[e
- pack
->objects
];
234 static inline void oe_set_in_pack_pos(const struct packing_data
*pack
,
235 const struct object_entry
*e
,
238 pack
->in_pack_pos
[e
- pack
->objects
] = pos
;
241 static inline struct packed_git
*oe_in_pack(const struct packing_data
*pack
,
242 const struct object_entry
*e
)
244 if (pack
->in_pack_by_idx
)
245 return pack
->in_pack_by_idx
[e
->in_pack_idx
];
247 return pack
->in_pack
[e
- pack
->objects
];
250 void oe_map_new_pack(struct packing_data
*pack
);
252 static inline void oe_set_in_pack(struct packing_data
*pack
,
253 struct object_entry
*e
,
254 struct packed_git
*p
)
256 if (pack
->in_pack_by_idx
) {
258 e
->in_pack_idx
= p
->index
;
262 * We're accessing packs by index, but this pack doesn't have
263 * an index (e.g., because it was added since we created the
264 * in_pack_by_idx array). Bail to oe_map_new_pack(), which
265 * will convert us to using the full in_pack array, and then
266 * fall through to our in_pack handling.
268 oe_map_new_pack(pack
);
270 pack
->in_pack
[e
- pack
->objects
] = p
;
273 static inline struct object_entry
*oe_delta(
274 const struct packing_data
*pack
,
275 const struct object_entry
*e
)
280 return &pack
->ext_bases
[e
->delta_idx
- 1];
282 return &pack
->objects
[e
->delta_idx
- 1];
285 static inline void oe_set_delta(struct packing_data
*pack
,
286 struct object_entry
*e
,
287 struct object_entry
*delta
)
290 e
->delta_idx
= (delta
- pack
->objects
) + 1;
295 void oe_set_delta_ext(struct packing_data
*pack
,
296 struct object_entry
*e
,
297 const unsigned char *sha1
);
299 static inline struct object_entry
*oe_delta_child(
300 const struct packing_data
*pack
,
301 const struct object_entry
*e
)
303 if (e
->delta_child_idx
)
304 return &pack
->objects
[e
->delta_child_idx
- 1];
308 static inline void oe_set_delta_child(struct packing_data
*pack
,
309 struct object_entry
*e
,
310 struct object_entry
*delta
)
313 e
->delta_child_idx
= (delta
- pack
->objects
) + 1;
315 e
->delta_child_idx
= 0;
318 static inline struct object_entry
*oe_delta_sibling(
319 const struct packing_data
*pack
,
320 const struct object_entry
*e
)
322 if (e
->delta_sibling_idx
)
323 return &pack
->objects
[e
->delta_sibling_idx
- 1];
327 static inline void oe_set_delta_sibling(struct packing_data
*pack
,
328 struct object_entry
*e
,
329 struct object_entry
*delta
)
332 e
->delta_sibling_idx
= (delta
- pack
->objects
) + 1;
334 e
->delta_sibling_idx
= 0;
337 unsigned long oe_get_size_slow(struct packing_data
*pack
,
338 const struct object_entry
*e
);
339 static inline unsigned long oe_size(struct packing_data
*pack
,
340 const struct object_entry
*e
)
345 return oe_get_size_slow(pack
, e
);
348 static inline int oe_size_less_than(struct packing_data
*pack
,
349 const struct object_entry
*lhs
,
353 return lhs
->size_
< rhs
;
354 if (rhs
< pack
->oe_size_limit
) /* rhs < 2^x <= lhs ? */
356 return oe_get_size_slow(pack
, lhs
) < rhs
;
359 static inline int oe_size_greater_than(struct packing_data
*pack
,
360 const struct object_entry
*lhs
,
364 return lhs
->size_
> rhs
;
365 if (rhs
< pack
->oe_size_limit
) /* rhs < 2^x <= lhs ? */
367 return oe_get_size_slow(pack
, lhs
) > rhs
;
370 static inline void oe_set_size(struct packing_data
*pack
,
371 struct object_entry
*e
,
374 if (size
< pack
->oe_size_limit
) {
379 if (oe_get_size_slow(pack
, e
) != size
)
380 BUG("'size' is supposed to be the object size!");
384 static inline unsigned long oe_delta_size(struct packing_data
*pack
,
385 const struct object_entry
*e
)
387 if (e
->delta_size_valid
)
388 return e
->delta_size_
;
391 * pack->delta_size[] can't be NULL because oe_set_delta_size()
392 * must have been called when a new delta is saved with
394 * If oe_delta() returns NULL (i.e. default state, which means
395 * delta_size_valid is also false), then the caller must never
396 * call oe_delta_size().
398 return pack
->delta_size
[e
- pack
->objects
];
401 static inline void oe_set_delta_size(struct packing_data
*pack
,
402 struct object_entry
*e
,
405 if (size
< pack
->oe_delta_size_limit
) {
406 e
->delta_size_
= size
;
407 e
->delta_size_valid
= 1;
409 packing_data_lock(pack
);
410 if (!pack
->delta_size
)
411 ALLOC_ARRAY(pack
->delta_size
, pack
->nr_alloc
);
412 packing_data_unlock(pack
);
414 pack
->delta_size
[e
- pack
->objects
] = size
;
415 e
->delta_size_valid
= 0;
419 static inline unsigned int oe_tree_depth(struct packing_data
*pack
,
420 struct object_entry
*e
)
422 if (!pack
->tree_depth
)
424 return pack
->tree_depth
[e
- pack
->objects
];
427 static inline void oe_set_tree_depth(struct packing_data
*pack
,
428 struct object_entry
*e
,
429 unsigned int tree_depth
)
431 if (!pack
->tree_depth
)
432 CALLOC_ARRAY(pack
->tree_depth
, pack
->nr_alloc
);
433 pack
->tree_depth
[e
- pack
->objects
] = tree_depth
;
436 static inline unsigned char oe_layer(struct packing_data
*pack
,
437 struct object_entry
*e
)
441 return pack
->layer
[e
- pack
->objects
];
444 static inline void oe_set_layer(struct packing_data
*pack
,
445 struct object_entry
*e
,
449 CALLOC_ARRAY(pack
->layer
, pack
->nr_alloc
);
450 pack
->layer
[e
- pack
->objects
] = layer
;