@matheusbsbs / matheus-bagliano / commits / 9d98354f48

cache.h: allow oid_object_info to handle arbitrary repositories

This involves also adapting oid_object_info_extended and a some internal functions that are used to implement these. It all has to happen in one patch, because of a single recursive chain of calls visits all these functions. oid_object_info_extended is also used in partial clones, which allow fetching missing objects. As this series will not add the repository struct to the transport code and fetch_object(), add a TODO note and omit fetching if a user tries to use a partial clone in a repository other than the_repository. Among the functions modified to handle arbitrary repositories, unpack_entry() is one of them. Note that it still references the globals "delta_base_cache" and "delta_base_cached", but those are safe to be referenced (the former is indexed partly by "struct packed_git *", which is repo-specific, and the latter is only used to limit the size of the former as an optimization). Helped-by: Brandon Williams <bmwill@google.com> Helped-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Jonathan Nieder <jrnieder@gmail.com> Signed-off-by: Stefan Beller <sbeller@google.com> Reviewed-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>

Stefan Beller committed Apr 25, 2018 at 11:21 UTC 9d98354f48997faf8251c566d909957f6ae427d5

4 files changed +54 -52

cache.h

+4 -5

@@ -1192,8 +1192,7 @@ static inline void *read_object_file(const struct object_id *oid, enum object_ty

1192 }

1193

1194 /* Read and unpack an object file into memory, write memory to an object file */

1195 -#define oid_object_info(r, o, f) oid_object_info_##r(o, f)

1196 -int oid_object_info_the_repository(const struct object_id *, unsigned long *);

1195 +int oid_object_info(struct repository *r, const struct object_id *, unsigned long *);

1196

1197 extern int hash_object_file(const void *buf, unsigned long len,

1198 const char *type, struct object_id *oid);

@@ -1675,9 +1674,9 @@ struct object_info {

1674 /* Do not check loose object */

1675 #define OBJECT_INFO_IGNORE_LOOSE 16

1676

1678 -#define oid_object_info_extended(r, oid, oi, flags) \

1679 - oid_object_info_extended_##r(oid, oi, flags)

1680 -int oid_object_info_extended_the_repository(const struct object_id *, struct object_info *, unsigned flags);

1677 +int oid_object_info_extended(struct repository *r,

1678 + const struct object_id *,

1679 + struct object_info *, unsigned flags);

1680

1681 /*

1682 * Set this to 0 to prevent sha1_object_info_extended() from fetching missing

packfile.c

+28 -30

@@ -1104,9 +1104,9 @@ static const unsigned char *get_delta_base_sha1(struct packed_git *p,

1104 return NULL;

1105 }

1106

1107 -#define retry_bad_packed_offset(r, p, o) \

1108 - retry_bad_packed_offset_##r(p, o)

1109 -static int retry_bad_packed_offset_the_repository(struct packed_git *p, off_t obj_offset)

1107 +static int retry_bad_packed_offset(struct repository *r,

1108 + struct packed_git *p,

1109 + off_t obj_offset)

1110 {

1111 int type;

1112 struct revindex_entry *revidx;

@@ -1116,7 +1116,7 @@ static int retry_bad_packed_offset_the_repository(struct packed_git *p, off_t ob

1116 return OBJ_BAD;

1117 nth_packed_object_oid(&oid, p, revidx->nr);

1118 mark_bad_packed_object(p, oid.hash);

1119 - type = oid_object_info(the_repository, &oid, NULL);

1119 + type = oid_object_info(r, &oid, NULL);

1120 if (type <= OBJ_NONE)

1121 return OBJ_BAD;

1122 return type;

@@ -1124,13 +1124,12 @@ static int retry_bad_packed_offset_the_repository(struct packed_git *p, off_t ob

1124

1125 #define POI_STACK_PREALLOC 64

1126

1127 -#define packed_to_object_type(r, p, o, t, w, c) \

1128 - packed_to_object_type_##r(p, o, t, w, c)

1129 -static enum object_type packed_to_object_type_the_repository(struct packed_git *p,

1130 - off_t obj_offset,

1131 - enum object_type type,

1132 - struct pack_window **w_curs,

1133 - off_t curpos)

1127 +static enum object_type packed_to_object_type(struct repository *r,

1128 + struct packed_git *p,

1129 + off_t obj_offset,

1130 + enum object_type type,

1131 + struct pack_window **w_curs,

1132 + off_t curpos)

1133 {

1134 off_t small_poi_stack[POI_STACK_PREALLOC];

1135 off_t *poi_stack = small_poi_stack;

@@ -1157,7 +1156,7 @@ static enum object_type packed_to_object_type_the_repository(struct packed_git *

1156 if (type <= OBJ_NONE) {

1157 /* If getting the base itself fails, we first

1158 * retry the base, otherwise unwind */

1160 - type = retry_bad_packed_offset(the_repository, p, base_offset);

1159 + type = retry_bad_packed_offset(r, p, base_offset);

1160 if (type > OBJ_NONE)

1161 goto out;

1162 goto unwind;

@@ -1185,7 +1184,7 @@ out:

1184 unwind:

1185 while (poi_stack_nr) {

1186 obj_offset = poi_stack[--poi_stack_nr];

1188 - type = retry_bad_packed_offset(the_repository, p, obj_offset);

1187 + type = retry_bad_packed_offset(r, p, obj_offset);

1188 if (type > OBJ_NONE)

1189 goto out;

1190 }

@@ -1272,15 +1271,15 @@ static void detach_delta_base_cache_entry(struct delta_base_cache_entry *ent)

1271 free(ent);

1272 }

1273

1275 -#define cache_or_unpack_entry(r, p, bo, bs, t) cache_or_unpack_entry_##r(p, bo, bs, t)

1276 -static void *cache_or_unpack_entry_the_repository(struct packed_git *p, off_t base_offset,

1277 - unsigned long *base_size, enum object_type *type)

1274 +static void *cache_or_unpack_entry(struct repository *r, struct packed_git *p,

1275 + off_t base_offset, unsigned long *base_size,

1276 + enum object_type *type)

1277 {

1278 struct delta_base_cache_entry *ent;

1279

1280 ent = get_delta_base_cache_entry(p, base_offset);

1281 if (!ent)

1283 - return unpack_entry(the_repository, p, base_offset, type, base_size);

1282 + return unpack_entry(r, p, base_offset, type, base_size);

1283

1284 if (type)

1285 *type = ent->type;

@@ -1334,8 +1333,8 @@ static void add_delta_base_cache(struct packed_git *p, off_t base_offset,

1333 hashmap_add(&delta_base_cache, ent);

1334 }

1335

1337 -int packed_object_info_the_repository(struct packed_git *p, off_t obj_offset,

1338 - struct object_info *oi)

1336 +int packed_object_info(struct repository *r, struct packed_git *p,

1337 + off_t obj_offset, struct object_info *oi)

1338 {

1339 struct pack_window *w_curs = NULL;

1340 unsigned long size;

@@ -1347,7 +1346,7 @@ int packed_object_info_the_repository(struct packed_git *p, off_t obj_offset,

1346 * a "real" type later if the caller is interested.

1347 */

1348 if (oi->contentp) {

1350 - *oi->contentp = cache_or_unpack_entry(the_repository, p, obj_offset, oi->sizep,

1349 + *oi->contentp = cache_or_unpack_entry(r, p, obj_offset, oi->sizep,

1350 &type);

1351 if (!*oi->contentp)

1352 type = OBJ_BAD;

@@ -1381,7 +1380,7 @@ int packed_object_info_the_repository(struct packed_git *p, off_t obj_offset,

1380

1381 if (oi->typep || oi->type_name) {

1382 enum object_type ptot;

1384 - ptot = packed_to_object_type(the_repository, p, obj_offset,

1383 + ptot = packed_to_object_type(r, p, obj_offset,

1384 type, &w_curs, curpos);

1385 if (oi->typep)

1386 *oi->typep = ptot;

@@ -1470,10 +1469,10 @@ struct unpack_entry_stack_ent {

1469 unsigned long size;

1470 };

1471

1473 -#define read_object(r, o, t, s) read_object_##r(o, t, s)

1474 -static void *read_object_the_repository(const struct object_id *oid,

1475 - enum object_type *type,

1476 - unsigned long *size)

1472 +static void *read_object(struct repository *r,

1473 + const struct object_id *oid,

1474 + enum object_type *type,

1475 + unsigned long *size)

1476 {

1477 struct object_info oi = OBJECT_INFO_INIT;

1478 void *content;

@@ -1481,14 +1480,13 @@ static void *read_object_the_repository(const struct object_id *oid,

1480 oi.sizep = size;

1481 oi.contentp = &content;

1482

1484 - if (oid_object_info_extended(the_repository, oid, &oi, 0) < 0)

1483 + if (oid_object_info_extended(r, oid, &oi, 0) < 0)

1484 return NULL;

1485 return content;

1486 }

1487

1489 -void *unpack_entry_the_repository(struct packed_git *p, off_t obj_offset,

1490 - enum object_type *final_type,

1491 - unsigned long *final_size)

1488 +void *unpack_entry(struct repository *r, struct packed_git *p, off_t obj_offset,

1489 + enum object_type *final_type, unsigned long *final_size)

1490 {

1491 struct pack_window *w_curs = NULL;

1492 off_t curpos = obj_offset;

@@ -1618,7 +1616,7 @@ void *unpack_entry_the_repository(struct packed_git *p, off_t obj_offset,

1616 oid_to_hex(&base_oid), (uintmax_t)obj_offset,

1617 p->pack_name);

1618 mark_bad_packed_object(p, base_oid.hash);

1621 - base = read_object(the_repository, &base_oid, &type, &base_size);

1619 + base = read_object(r, &base_oid, &type, &base_size);

1620 external_base = base;

1621 }

1622 }

packfile.h

+4 -4

@@ -115,8 +115,7 @@ extern off_t nth_packed_object_offset(const struct packed_git *, uint32_t n);

115 extern off_t find_pack_entry_one(const unsigned char *sha1, struct packed_git *);

116

117 extern int is_pack_valid(struct packed_git *);

118 -#define unpack_entry(r, p, of, ot, s) unpack_entry_##r(p, of, ot, s)

119 -extern void *unpack_entry_the_repository(struct packed_git *, off_t, enum object_type *, unsigned long *);

118 +extern void *unpack_entry(struct repository *r, struct packed_git *, off_t, enum object_type *, unsigned long *);

119 extern unsigned long unpack_object_header_buffer(const unsigned char *buf, unsigned long len, enum object_type *type, unsigned long *sizep);

120 extern unsigned long get_size_from_delta(struct packed_git *, struct pack_window **, off_t);

121 extern int unpack_object_header(struct packed_git *, struct pack_window **, off_t *, unsigned long *);

@@ -126,8 +125,9 @@ extern void release_pack_memory(size_t);

125 /* global flag to enable extra checks when accessing packed objects */

126 extern int do_check_packed_object_crc;

127

129 -#define packed_object_info(r, p, o, oi) packed_object_info_##r(p, o, oi)

130 -extern int packed_object_info_the_repository(struct packed_git *pack, off_t offset, struct object_info *);

128 +extern int packed_object_info(struct repository *r,

129 + struct packed_git *pack,

130 + off_t offset, struct object_info *);

131

132 extern void mark_bad_packed_object(struct packed_git *p, const unsigned char *sha1);

133 extern const struct packed_git *has_packed_and_bad(const unsigned char *sha1);

sha1_file.c

+18 -13

@@ -1231,7 +1231,8 @@ static int sha1_loose_object_info(struct repository *r,

1231

1232 int fetch_if_missing = 1;

1233

1234 -int oid_object_info_extended_the_repository(const struct object_id *oid, struct object_info *oi, unsigned flags)

1234 +int oid_object_info_extended(struct repository *r, const struct object_id *oid,

1235 + struct object_info *oi, unsigned flags)

1236 {

1237 static struct object_info blank_oi = OBJECT_INFO_INIT;

1238 struct pack_entry e;

@@ -1240,7 +1241,7 @@ int oid_object_info_extended_the_repository(const struct object_id *oid, struct

1241 int already_retried = 0;

1242

1243 if (flags & OBJECT_INFO_LOOKUP_REPLACE)

1243 - real = lookup_replace_object(the_repository, oid);

1244 + real = lookup_replace_object(r, oid);

1245

1246 if (is_null_oid(real))

1247 return -1;

@@ -1269,29 +1270,31 @@ int oid_object_info_extended_the_repository(const struct object_id *oid, struct

1270 }

1271

1272 while (1) {

1272 - if (find_pack_entry(the_repository, real->hash, &e))

1273 + if (find_pack_entry(r, real->hash, &e))

1274 break;

1275

1276 if (flags & OBJECT_INFO_IGNORE_LOOSE)

1277 return -1;

1278

1279 /* Most likely it's a loose object. */

1279 - if (!sha1_loose_object_info(the_repository, real->hash, oi, flags))

1280 + if (!sha1_loose_object_info(r, real->hash, oi, flags))

1281 return 0;

1282

1283 /* Not a loose object; someone else may have just packed it. */

1284 if (!(flags & OBJECT_INFO_QUICK)) {

1284 - reprepare_packed_git(the_repository);

1285 - if (find_pack_entry(the_repository, real->hash, &e))

1285 + reprepare_packed_git(r);

1286 + if (find_pack_entry(r, real->hash, &e))

1287 break;

1288 }

1289

1290 /* Check if it is a missing object */

1291 if (fetch_if_missing && repository_format_partial_clone &&

1291 - !already_retried) {

1292 + !already_retried && r == the_repository) {

1293 /*

1293 - * TODO Investigate haveing fetch_object() return

1294 + * TODO Investigate having fetch_object() return

1295 * TODO error/success and stopping the music here.

1296 + * TODO Pass a repository struct through fetch_object,

1297 + * such that arbitrary repositories work.

1298 */

1299 fetch_object(repository_format_partial_clone, real->hash);

1300 already_retried = 1;

@@ -1307,10 +1310,10 @@ int oid_object_info_extended_the_repository(const struct object_id *oid, struct

1310 * information below, so return early.

1311 */

1312 return 0;

1310 - rtype = packed_object_info(the_repository, e.p, e.offset, oi);

1313 + rtype = packed_object_info(r, e.p, e.offset, oi);

1314 if (rtype < 0) {

1315 mark_bad_packed_object(e.p, real->hash);

1313 - return oid_object_info_extended(the_repository, real, oi, 0);

1316 + return oid_object_info_extended(r, real, oi, 0);

1317 } else if (oi->whence == OI_PACKED) {

1318 oi->u.packed.offset = e.offset;

1319 oi->u.packed.pack = e.p;

@@ -1322,15 +1325,17 @@ int oid_object_info_extended_the_repository(const struct object_id *oid, struct

1325 }

1326

1327 /* returns enum object_type or negative */

1325 -int oid_object_info_the_repository(const struct object_id *oid, unsigned long *sizep)

1328 +int oid_object_info(struct repository *r,

1329 + const struct object_id *oid,

1330 + unsigned long *sizep)

1331 {

1332 enum object_type type;

1333 struct object_info oi = OBJECT_INFO_INIT;

1334

1335 oi.typep = &type;

1336 oi.sizep = sizep;

1332 - if (oid_object_info_extended(the_repository, oid, &oi,

1333 - OBJECT_INFO_LOOKUP_REPLACE) < 0)

1337 + if (oid_object_info_extended(r, oid, &oi,

1338 + OBJECT_INFO_LOOKUP_REPLACE) < 0)

1339 return -1;

1340 return type;

1341 }