object-store.h

   1 #ifndef OBJECT_STORE_H
   2 #define OBJECT_STORE_H
   3
   4 #include "cache.h"
   5 #include "oidmap.h"
   6 #include "list.h"
   7 #include "oid-array.h"
   8 #include "strbuf.h"
   9 #include "thread-utils.h"
  10 #include "khash.h"
  11 #include "dir.h"
  12
  13 struct object_directory {
  14         struct object_directory *next;
  15
  16         /*
  17          * Used to store the results of readdir(3) calls when we are OK
  18          * sacrificing accuracy due to races for speed. That includes
  19          * object existence with OBJECT_INFO_QUICK, as well as
  20          * our search for unique abbreviated hashes. Don't use it for tasks
  21          * requiring greater accuracy!
  22          *
  23          * Be sure to call odb_load_loose_cache() before using.
  24          */
  25         uint32_t loose_objects_subdir_seen[8]; /* 256 bits */
  26         struct oid_array loose_objects_cache[256];
  27
  28         /*
  29          * Path to the alternative object store. If this is a relative path,
  30          * it is relative to the current working directory.
  31          */
  32         char *path;
  33 };
  34
  35 KHASH_INIT(odb_path_map, const char * /* key: odb_path */,
  36         struct object_directory *, 1, fspathhash, fspatheq);
  37
  38 void prepare_alt_odb(struct repository *r);
  39 char *compute_alternate_path(const char *path, struct strbuf *err);
  40 typedef int alt_odb_fn(struct object_directory *, void *);
  41 int foreach_alt_odb(alt_odb_fn, void*);
  42 typedef void alternate_ref_fn(const struct object_id *oid, void *);
  43 void for_each_alternate_ref(alternate_ref_fn, void *);
  44
  45 /*
  46  * Add the directory to the on-disk alternates file; the new entry will also
  47  * take effect in the current process.
  48  */
  49 void add_to_alternates_file(const char *dir);
  50
  51 /*
  52  * Add the directory to the in-memory list of alternates (along with any
  53  * recursive alternates it points to), but do not modify the on-disk alternates
  54  * file.
  55  */
  56 void add_to_alternates_memory(const char *dir);
  57
  58 /*
  59  * Populate and return the loose object cache array corresponding to the
  60  * given object ID.
  61  */
  62 struct oid_array *odb_loose_cache(struct object_directory *odb,
  63                                   const struct object_id *oid);
  64
  65 /* Empty the loose object cache for the specified object directory. */
  66 void odb_clear_loose_cache(struct object_directory *odb);
  67
  68 struct packed_git {
  69         struct hashmap_entry packmap_ent;
  70         struct packed_git *next;
  71         struct list_head mru;
  72         struct pack_window *windows;
  73         off_t pack_size;
  74         const void *index_data;
  75         size_t index_size;
  76         uint32_t num_objects;
  77         uint32_t num_bad_objects;
  78         uint32_t crc_offset;
  79         unsigned char *bad_object_sha1;
  80         int index_version;
  81         time_t mtime;
  82         int pack_fd;
  83         int index;              /* for builtin/pack-objects.c */
  84         unsigned pack_local:1,
  85                  pack_keep:1,
  86                  pack_keep_in_core:1,
  87                  freshened:1,
  88                  do_not_close:1,
  89                  pack_promisor:1,
  90                  multi_pack_index:1;
  91         unsigned char hash[GIT_MAX_RAWSZ];
  92         struct revindex_entry *revindex;
  93         const uint32_t *revindex_data;
  94         const uint32_t *revindex_map;
  95         size_t revindex_size;
  96         /* something like ".git/objects/pack/xxxxx.pack" */
  97         char pack_name[FLEX_ARRAY]; /* more */
  98 };
  99
 100 struct multi_pack_index;
 101
 102 static inline int pack_map_entry_cmp(const void *unused_cmp_data,
 103                                      const struct hashmap_entry *entry,
 104                                      const struct hashmap_entry *entry2,
 105                                      const void *keydata)
 106 {
 107         const char *key = keydata;
 108         const struct packed_git *pg1, *pg2;
 109
 110         pg1 = container_of(entry, const struct packed_git, packmap_ent);
 111         pg2 = container_of(entry2, const struct packed_git, packmap_ent);
 112
 113         return strcmp(pg1->pack_name, key ? key : pg2->pack_name);
 114 }
 115
 116 struct raw_object_store {
 117         /*
 118          * Set of all object directories; the main directory is first (and
 119          * cannot be NULL after initialization). Subsequent directories are
 120          * alternates.
 121          */
 122         struct object_directory *odb;
 123         struct object_directory **odb_tail;
 124         kh_odb_path_map_t *odb_by_path;
 125
 126         int loaded_alternates;
 127
 128         /*
 129          * A list of alternate object directories loaded from the environment;
 130          * this should not generally need to be accessed directly, but will
 131          * populate the "odb" list when prepare_alt_odb() is run.
 132          */
 133         char *alternate_db;
 134
 135         /*
 136          * Objects that should be substituted by other objects
 137          * (see git-replace(1)).
 138          */
 139         struct oidmap *replace_map;
 140         unsigned replace_map_initialized : 1;
 141         pthread_mutex_t replace_mutex; /* protect object replace functions */
 142
 143         struct commit_graph *commit_graph;
 144         unsigned commit_graph_attempted : 1; /* if loading has been attempted */
 145
 146         /*
 147          * private data
 148          *
 149          * should only be accessed directly by packfile.c and midx.c
 150          */
 151         struct multi_pack_index *multi_pack_index;
 152
 153         /*
 154          * private data
 155          *
 156          * should only be accessed directly by packfile.c
 157          */
 158
 159         struct packed_git *packed_git;
 160         /* A most-recently-used ordered version of the packed_git list. */
 161         struct list_head packed_git_mru;
 162
 163         struct {
 164                 struct packed_git **packs;
 165                 unsigned flags;
 166         } kept_pack_cache;
 167
 168         /*
 169          * A map of packfiles to packed_git structs for tracking which
 170          * packs have been loaded already.
 171          */
 172         struct hashmap pack_map;
 173
 174         /*
 175          * A fast, rough count of the number of objects in the repository.
 176          * These two fields are not meant for direct access. Use
 177          * approximate_object_count() instead.
 178          */
 179         unsigned long approximate_object_count;
 180         unsigned approximate_object_count_valid : 1;
 181
 182         /*
 183          * Whether packed_git has already been populated with this repository's
 184          * packs.
 185          */
 186         unsigned packed_git_initialized : 1;
 187 };
 188
 189 struct raw_object_store *raw_object_store_new(void);
 190 void raw_object_store_clear(struct raw_object_store *o);
 191
 192 /*
 193  * Put in `buf` the name of the file in the local object database that
 194  * would be used to store a loose object with the specified oid.
 195  */
 196 const char *loose_object_path(struct repository *r, struct strbuf *buf,
 197                               const struct object_id *oid);
 198
 199 void *map_loose_object(struct repository *r, const struct object_id *oid,
 200                        unsigned long *size);
 201
 202 void *read_object_file_extended(struct repository *r,
 203                                 const struct object_id *oid,
 204                                 enum object_type *type,
 205                                 unsigned long *size, int lookup_replace);
 206 static inline void *repo_read_object_file(struct repository *r,
 207                                           const struct object_id *oid,
 208                                           enum object_type *type,
 209                                           unsigned long *size)
 210 {
 211         return read_object_file_extended(r, oid, type, size, 1);
 212 }
 213 #ifndef NO_THE_REPOSITORY_COMPATIBILITY_MACROS
 214 #define read_object_file(oid, type, size) repo_read_object_file(the_repository, oid, type, size)
 215 #endif
 216
 217 /* Read and unpack an object file into memory, write memory to an object file */
 218 int oid_object_info(struct repository *r, const struct object_id *, unsigned long *);
 219
 220 int hash_object_file(const struct git_hash_algo *algo, const void *buf,
 221                      unsigned long len, const char *type,
 222                      struct object_id *oid);
 223
 224 int write_object_file(const void *buf, unsigned long len,
 225                       const char *type, struct object_id *oid);
 226
 227 int hash_object_file_literally(const void *buf, unsigned long len,
 228                                const char *type, struct object_id *oid,
 229                                unsigned flags);
 230
 231 /*
 232  * Add an object file to the in-memory object store, without writing it
 233  * to disk.
 234  *
 235  * Callers are responsible for calling write_object_file to record the
 236  * object in persistent storage before writing any other new objects
 237  * that reference it.
 238  */
 239 int pretend_object_file(void *, unsigned long, enum object_type,
 240                         struct object_id *oid);
 241
 242 int force_object_loose(const struct object_id *oid, time_t mtime);
 243
 244 /*
 245  * Open the loose object at path, check its hash, and return the contents,
 246  * type, and size. If the object is a blob, then "contents" may return NULL,
 247  * to allow streaming of large blobs.
 248  *
 249  * Returns 0 on success, negative on error (details may be written to stderr).
 250  */
 251 int read_loose_object(const char *path,
 252                       const struct object_id *expected_oid,
 253                       enum object_type *type,
 254                       unsigned long *size,
 255                       void **contents);
 256
 257 /* Retry packed storage after checking packed and loose storage */
 258 #define HAS_OBJECT_RECHECK_PACKED 1
 259
 260 /*
 261  * Returns 1 if the object exists. This function will not lazily fetch objects
 262  * in a partial clone.
 263  */
 264 int has_object(struct repository *r, const struct object_id *oid,
 265                unsigned flags);
 266
 267 /*
 268  * These macros and functions are deprecated. If checking existence for an
 269  * object that is likely to be missing and/or whose absence is relatively
 270  * inconsequential (or is consequential but the caller is prepared to handle
 271  * it), use has_object(), which has better defaults (no lazy fetch in a partial
 272  * clone and no rechecking of packed storage). In the unlikely event that a
 273  * caller needs to assert existence of an object that it fully expects to
 274  * exist, and wants to trigger a lazy fetch in a partial clone, use
 275  * oid_object_info_extended() with a NULL struct object_info.
 276  *
 277  * These functions can be removed once all callers have migrated to
 278  * has_object() and/or oid_object_info_extended().
 279  */
 280 #ifndef NO_THE_REPOSITORY_COMPATIBILITY_MACROS
 281 #define has_sha1_file_with_flags(sha1, flags) repo_has_sha1_file_with_flags(the_repository, sha1, flags)
 282 #define has_sha1_file(sha1) repo_has_sha1_file(the_repository, sha1)
 283 #endif
 284 int repo_has_object_file(struct repository *r, const struct object_id *oid);
 285 int repo_has_object_file_with_flags(struct repository *r,
 286                                     const struct object_id *oid, int flags);
 287 #ifndef NO_THE_REPOSITORY_COMPATIBILITY_MACROS
 288 #define has_object_file(oid) repo_has_object_file(the_repository, oid)
 289 #define has_object_file_with_flags(oid, flags) repo_has_object_file_with_flags(the_repository, oid, flags)
 290 #endif
 291
 292 /*
 293  * Return true iff an alternate object database has a loose object
 294  * with the specified name.  This function does not respect replace
 295  * references.
 296  */
 297 int has_loose_object_nonlocal(const struct object_id *);
 298
 299 void assert_oid_type(const struct object_id *oid, enum object_type expect);
 300
 301 /*
 302  * Enabling the object read lock allows multiple threads to safely call the
 303  * following functions in parallel: repo_read_object_file(), read_object_file(),
 304  * read_object_file_extended(), read_object_with_reference(), read_object(),
 305  * oid_object_info() and oid_object_info_extended().
 306  *
 307  * obj_read_lock() and obj_read_unlock() may also be used to protect other
 308  * section which cannot execute in parallel with object reading. Since the used
 309  * lock is a recursive mutex, these sections can even contain calls to object
 310  * reading functions. However, beware that in these cases zlib inflation won't
 311  * be performed in parallel, losing performance.
 312  *
 313  * TODO: oid_object_info_extended()'s call stack has a recursive behavior. If
 314  * any of its callees end up calling it, this recursive call won't benefit from
 315  * parallel inflation.
 316  */
 317 void enable_obj_read_lock(void);
 318 void disable_obj_read_lock(void);
 319
 320 extern int obj_read_use_lock;
 321 extern pthread_mutex_t obj_read_mutex;
 322
 323 static inline void obj_read_lock(void)
 324 {
 325         if(obj_read_use_lock)
 326                 pthread_mutex_lock(&obj_read_mutex);
 327 }
 328
 329 static inline void obj_read_unlock(void)
 330 {
 331         if(obj_read_use_lock)
 332                 pthread_mutex_unlock(&obj_read_mutex);
 333 }
 334
 335 struct object_info {
 336         /* Request */
 337         enum object_type *typep;
 338         unsigned long *sizep;
 339         off_t *disk_sizep;
 340         struct object_id *delta_base_oid;
 341         struct strbuf *type_name;
 342         void **contentp;
 343
 344         /* Response */
 345         enum {
 346                 OI_CACHED,
 347                 OI_LOOSE,
 348                 OI_PACKED,
 349                 OI_DBCACHED
 350         } whence;
 351         union {
 352                 /*
 353                  * struct {
 354                  *      ... Nothing to expose in this case
 355                  * } cached;
 356                  * struct {
 357                  *      ... Nothing to expose in this case
 358                  * } loose;
 359                  */
 360                 struct {
 361                         struct packed_git *pack;
 362                         off_t offset;
 363                         unsigned int is_delta;
 364                 } packed;
 365         } u;
 366 };
 367
 368 /*
 369  * Initializer for a "struct object_info" that wants no items. You may
 370  * also memset() the memory to all-zeroes.
 371  */
 372 #define OBJECT_INFO_INIT {NULL}
 373
 374 /* Invoke lookup_replace_object() on the given hash */
 375 #define OBJECT_INFO_LOOKUP_REPLACE 1
 376 /* Allow reading from a loose object file of unknown/bogus type */
 377 #define OBJECT_INFO_ALLOW_UNKNOWN_TYPE 2
 378 /* Do not retry packed storage after checking packed and loose storage */
 379 #define OBJECT_INFO_QUICK 8
 380 /* Do not check loose object */
 381 #define OBJECT_INFO_IGNORE_LOOSE 16
 382 /*
 383  * Do not attempt to fetch the object if missing (even if fetch_is_missing is
 384  * nonzero).
 385  */
 386 #define OBJECT_INFO_SKIP_FETCH_OBJECT 32
 387 /*
 388  * This is meant for bulk prefetching of missing blobs in a partial
 389  * clone. Implies OBJECT_INFO_SKIP_FETCH_OBJECT and OBJECT_INFO_QUICK
 390  */
 391 #define OBJECT_INFO_FOR_PREFETCH (OBJECT_INFO_SKIP_FETCH_OBJECT | OBJECT_INFO_QUICK)
 392
 393 int oid_object_info_extended(struct repository *r,
 394                              const struct object_id *,
 395                              struct object_info *, unsigned flags);
 396
 397 /*
 398  * Iterate over the files in the loose-object parts of the object
 399  * directory "path", triggering the following callbacks:
 400  *
 401  *  - loose_object is called for each loose object we find.
 402  *
 403  *  - loose_cruft is called for any files that do not appear to be
 404  *    loose objects. Note that we only look in the loose object
 405  *    directories "objects/[0-9a-f]{2}/", so we will not report
 406  *    "objects/foobar" as cruft.
 407  *
 408  *  - loose_subdir is called for each top-level hashed subdirectory
 409  *    of the object directory (e.g., "$OBJDIR/f0"). It is called
 410  *    after the objects in the directory are processed.
 411  *
 412  * Any callback that is NULL will be ignored. Callbacks returning non-zero
 413  * will end the iteration.
 414  *
 415  * In the "buf" variant, "path" is a strbuf which will also be used as a
 416  * scratch buffer, but restored to its original contents before
 417  * the function returns.
 418  */
 419 typedef int each_loose_object_fn(const struct object_id *oid,
 420                                  const char *path,
 421                                  void *data);
 422 typedef int each_loose_cruft_fn(const char *basename,
 423                                 const char *path,
 424                                 void *data);
 425 typedef int each_loose_subdir_fn(unsigned int nr,
 426                                  const char *path,
 427                                  void *data);
 428 int for_each_file_in_obj_subdir(unsigned int subdir_nr,
 429                                 struct strbuf *path,
 430                                 each_loose_object_fn obj_cb,
 431                                 each_loose_cruft_fn cruft_cb,
 432                                 each_loose_subdir_fn subdir_cb,
 433                                 void *data);
 434 int for_each_loose_file_in_objdir(const char *path,
 435                                   each_loose_object_fn obj_cb,
 436                                   each_loose_cruft_fn cruft_cb,
 437                                   each_loose_subdir_fn subdir_cb,
 438                                   void *data);
 439 int for_each_loose_file_in_objdir_buf(struct strbuf *path,
 440                                       each_loose_object_fn obj_cb,
 441                                       each_loose_cruft_fn cruft_cb,
 442                                       each_loose_subdir_fn subdir_cb,
 443                                       void *data);
 444
 445 /* Flags for for_each_*_object() below. */
 446 enum for_each_object_flags {
 447         /* Iterate only over local objects, not alternates. */
 448         FOR_EACH_OBJECT_LOCAL_ONLY = (1<<0),
 449
 450         /* Only iterate over packs obtained from the promisor remote. */
 451         FOR_EACH_OBJECT_PROMISOR_ONLY = (1<<1),
 452
 453         /*
 454          * Visit objects within a pack in packfile order rather than .idx order
 455          */
 456         FOR_EACH_OBJECT_PACK_ORDER = (1<<2),
 457 };
 458
 459 /*
 460  * Iterate over all accessible loose objects without respect to
 461  * reachability. By default, this includes both local and alternate objects.
 462  * The order in which objects are visited is unspecified.
 463  *
 464  * Any flags specific to packs are ignored.
 465  */
 466 int for_each_loose_object(each_loose_object_fn, void *,
 467                           enum for_each_object_flags flags);
 468
 469 /*
 470  * Iterate over all accessible packed objects without respect to reachability.
 471  * By default, this includes both local and alternate packs.
 472  *
 473  * Note that some objects may appear twice if they are found in multiple packs.
 474  * Each pack is visited in an unspecified order. By default, objects within a
 475  * pack are visited in pack-idx order (i.e., sorted by oid).
 476  */
 477 typedef int each_packed_object_fn(const struct object_id *oid,
 478                                   struct packed_git *pack,
 479                                   uint32_t pos,
 480                                   void *data);
 481 int for_each_object_in_pack(struct packed_git *p,
 482                             each_packed_object_fn, void *data,
 483                             enum for_each_object_flags flags);
 484 int for_each_packed_object(each_packed_object_fn, void *,
 485                            enum for_each_object_flags flags);
 486
 487 #endif /* OBJECT_STORE_H */