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