hash.h

   1 #ifndef HASH_H
   2 #define HASH_H
   3
   4 #include "git-compat-util.h"
   5 #include "repository.h"
   6
   7 #if defined(SHA1_APPLE)
   8 #include <CommonCrypto/CommonDigest.h>
   9 #elif defined(SHA1_OPENSSL)
  10 #include <openssl/sha.h>
  11 #elif defined(SHA1_DC)
  12 #include "sha1dc_git.h"
  13 #else /* SHA1_BLK */
  14 #include "block-sha1/sha1.h"
  15 #endif
  16
  17 #if defined(SHA256_NETTLE)
  18 #include "sha256/nettle.h"
  19 #elif defined(SHA256_GCRYPT)
  20 #define SHA256_NEEDS_CLONE_HELPER
  21 #include "sha256/gcrypt.h"
  22 #elif defined(SHA256_OPENSSL)
  23 #include <openssl/sha.h>
  24 #else
  25 #include "sha256/block/sha256.h"
  26 #endif
  27
  28 #ifndef platform_SHA_CTX
  29 /*
  30  * platform's underlying implementation of SHA-1; could be OpenSSL,
  31  * blk_SHA, Apple CommonCrypto, etc...  Note that the relevant
  32  * SHA-1 header may have already defined platform_SHA_CTX for our
  33  * own implementations like block-sha1, so we list
  34  * the default for OpenSSL compatible SHA-1 implementations here.
  35  */
  36 #define platform_SHA_CTX        SHA_CTX
  37 #define platform_SHA1_Init      SHA1_Init
  38 #define platform_SHA1_Update    SHA1_Update
  39 #define platform_SHA1_Final     SHA1_Final
  40 #endif
  41
  42 #define git_SHA_CTX             platform_SHA_CTX
  43 #define git_SHA1_Init           platform_SHA1_Init
  44 #define git_SHA1_Update         platform_SHA1_Update
  45 #define git_SHA1_Final          platform_SHA1_Final
  46
  47 #ifndef platform_SHA256_CTX
  48 #define platform_SHA256_CTX     SHA256_CTX
  49 #define platform_SHA256_Init    SHA256_Init
  50 #define platform_SHA256_Update  SHA256_Update
  51 #define platform_SHA256_Final   SHA256_Final
  52 #endif
  53
  54 #define git_SHA256_CTX          platform_SHA256_CTX
  55 #define git_SHA256_Init         platform_SHA256_Init
  56 #define git_SHA256_Update       platform_SHA256_Update
  57 #define git_SHA256_Final        platform_SHA256_Final
  58
  59 #ifdef platform_SHA256_Clone
  60 #define git_SHA256_Clone        platform_SHA256_Clone
  61 #endif
  62
  63 #ifdef SHA1_MAX_BLOCK_SIZE
  64 #include "compat/sha1-chunked.h"
  65 #undef git_SHA1_Update
  66 #define git_SHA1_Update         git_SHA1_Update_Chunked
  67 #endif
  68
  69 static inline void git_SHA1_Clone(git_SHA_CTX *dst, const git_SHA_CTX *src)
  70 {
  71         memcpy(dst, src, sizeof(*dst));
  72 }
  73
  74 #ifndef SHA256_NEEDS_CLONE_HELPER
  75 static inline void git_SHA256_Clone(git_SHA256_CTX *dst, const git_SHA256_CTX *src)
  76 {
  77         memcpy(dst, src, sizeof(*dst));
  78 }
  79 #endif
  80
  81 /*
  82  * Note that these constants are suitable for indexing the hash_algos array and
  83  * comparing against each other, but are otherwise arbitrary, so they should not
  84  * be exposed to the user or serialized to disk.  To know whether a
  85  * git_hash_algo struct points to some usable hash function, test the format_id
  86  * field for being non-zero.  Use the name field for user-visible situations and
  87  * the format_id field for fixed-length fields on disk.
  88  */
  89 /* An unknown hash function. */
  90 #define GIT_HASH_UNKNOWN 0
  91 /* SHA-1 */
  92 #define GIT_HASH_SHA1 1
  93 /* SHA-256  */
  94 #define GIT_HASH_SHA256 2
  95 /* Number of algorithms supported (including unknown). */
  96 #define GIT_HASH_NALGOS (GIT_HASH_SHA256 + 1)
  97
  98 /* "sha1", big-endian */
  99 #define GIT_SHA1_FORMAT_ID 0x73686131
 100
 101 /* The length in bytes and in hex digits of an object name (SHA-1 value). */
 102 #define GIT_SHA1_RAWSZ 20
 103 #define GIT_SHA1_HEXSZ (2 * GIT_SHA1_RAWSZ)
 104 /* The block size of SHA-1. */
 105 #define GIT_SHA1_BLKSZ 64
 106
 107 /* "s256", big-endian */
 108 #define GIT_SHA256_FORMAT_ID 0x73323536
 109
 110 /* The length in bytes and in hex digits of an object name (SHA-256 value). */
 111 #define GIT_SHA256_RAWSZ 32
 112 #define GIT_SHA256_HEXSZ (2 * GIT_SHA256_RAWSZ)
 113 /* The block size of SHA-256. */
 114 #define GIT_SHA256_BLKSZ 64
 115
 116 /* The length in byte and in hex digits of the largest possible hash value. */
 117 #define GIT_MAX_RAWSZ GIT_SHA256_RAWSZ
 118 #define GIT_MAX_HEXSZ GIT_SHA256_HEXSZ
 119 /* The largest possible block size for any supported hash. */
 120 #define GIT_MAX_BLKSZ GIT_SHA256_BLKSZ
 121
 122 struct object_id {
 123         unsigned char hash[GIT_MAX_RAWSZ];
 124         int algo;       /* XXX requires 4-byte alignment */
 125 };
 126
 127 /* A suitably aligned type for stack allocations of hash contexts. */
 128 union git_hash_ctx {
 129         git_SHA_CTX sha1;
 130         git_SHA256_CTX sha256;
 131 };
 132 typedef union git_hash_ctx git_hash_ctx;
 133
 134 typedef void (*git_hash_init_fn)(git_hash_ctx *ctx);
 135 typedef void (*git_hash_clone_fn)(git_hash_ctx *dst, const git_hash_ctx *src);
 136 typedef void (*git_hash_update_fn)(git_hash_ctx *ctx, const void *in, size_t len);
 137 typedef void (*git_hash_final_fn)(unsigned char *hash, git_hash_ctx *ctx);
 138 typedef void (*git_hash_final_oid_fn)(struct object_id *oid, git_hash_ctx *ctx);
 139
 140 struct git_hash_algo {
 141         /*
 142          * The name of the algorithm, as appears in the config file and in
 143          * messages.
 144          */
 145         const char *name;
 146
 147         /* A four-byte version identifier, used in pack indices. */
 148         uint32_t format_id;
 149
 150         /* The length of the hash in binary. */
 151         size_t rawsz;
 152
 153         /* The length of the hash in hex characters. */
 154         size_t hexsz;
 155
 156         /* The block size of the hash. */
 157         size_t blksz;
 158
 159         /* The hash initialization function. */
 160         git_hash_init_fn init_fn;
 161
 162         /* The hash context cloning function. */
 163         git_hash_clone_fn clone_fn;
 164
 165         /* The hash update function. */
 166         git_hash_update_fn update_fn;
 167
 168         /* The hash finalization function. */
 169         git_hash_final_fn final_fn;
 170
 171         /* The hash finalization function for object IDs. */
 172         git_hash_final_oid_fn final_oid_fn;
 173
 174         /* The OID of the empty tree. */
 175         const struct object_id *empty_tree;
 176
 177         /* The OID of the empty blob. */
 178         const struct object_id *empty_blob;
 179
 180         /* The all-zeros OID. */
 181         const struct object_id *null_oid;
 182 };
 183 extern const struct git_hash_algo hash_algos[GIT_HASH_NALGOS];
 184
 185 /*
 186  * Return a GIT_HASH_* constant based on the name.  Returns GIT_HASH_UNKNOWN if
 187  * the name doesn't match a known algorithm.
 188  */
 189 int hash_algo_by_name(const char *name);
 190 /* Identical, except based on the format ID. */
 191 int hash_algo_by_id(uint32_t format_id);
 192 /* Identical, except based on the length. */
 193 int hash_algo_by_length(int len);
 194 /* Identical, except for a pointer to struct git_hash_algo. */
 195 static inline int hash_algo_by_ptr(const struct git_hash_algo *p)
 196 {
 197         return p - hash_algos;
 198 }
 199
 200 #define the_hash_algo the_repository->hash_algo
 201
 202 const struct object_id *null_oid(void);
 203
 204 static inline int hashcmp_algop(const unsigned char *sha1, const unsigned char *sha2, const struct git_hash_algo *algop)
 205 {
 206         /*
 207          * Teach the compiler that there are only two possibilities of hash size
 208          * here, so that it can optimize for this case as much as possible.
 209          */
 210         if (algop->rawsz == GIT_MAX_RAWSZ)
 211                 return memcmp(sha1, sha2, GIT_MAX_RAWSZ);
 212         return memcmp(sha1, sha2, GIT_SHA1_RAWSZ);
 213 }
 214
 215 static inline int hashcmp(const unsigned char *sha1, const unsigned char *sha2)
 216 {
 217         return hashcmp_algop(sha1, sha2, the_hash_algo);
 218 }
 219
 220 static inline int oidcmp(const struct object_id *oid1, const struct object_id *oid2)
 221 {
 222         const struct git_hash_algo *algop;
 223         if (!oid1->algo)
 224                 algop = the_hash_algo;
 225         else
 226                 algop = &hash_algos[oid1->algo];
 227         return hashcmp_algop(oid1->hash, oid2->hash, algop);
 228 }
 229
 230 static inline int hasheq_algop(const unsigned char *sha1, const unsigned char *sha2, const struct git_hash_algo *algop)
 231 {
 232         /*
 233          * We write this here instead of deferring to hashcmp so that the
 234          * compiler can properly inline it and avoid calling memcmp.
 235          */
 236         if (algop->rawsz == GIT_MAX_RAWSZ)
 237                 return !memcmp(sha1, sha2, GIT_MAX_RAWSZ);
 238         return !memcmp(sha1, sha2, GIT_SHA1_RAWSZ);
 239 }
 240
 241 static inline int hasheq(const unsigned char *sha1, const unsigned char *sha2)
 242 {
 243         return hasheq_algop(sha1, sha2, the_hash_algo);
 244 }
 245
 246 static inline int oideq(const struct object_id *oid1, const struct object_id *oid2)
 247 {
 248         const struct git_hash_algo *algop;
 249         if (!oid1->algo)
 250                 algop = the_hash_algo;
 251         else
 252                 algop = &hash_algos[oid1->algo];
 253         return hasheq_algop(oid1->hash, oid2->hash, algop);
 254 }
 255
 256 static inline int is_null_oid(const struct object_id *oid)
 257 {
 258         return oideq(oid, null_oid());
 259 }
 260
 261 static inline void hashcpy(unsigned char *sha_dst, const unsigned char *sha_src)
 262 {
 263         memcpy(sha_dst, sha_src, the_hash_algo->rawsz);
 264 }
 265
 266 static inline void oidcpy(struct object_id *dst, const struct object_id *src)
 267 {
 268         memcpy(dst->hash, src->hash, GIT_MAX_RAWSZ);
 269         dst->algo = src->algo;
 270 }
 271
 272 /* Like oidcpy() but zero-pads the unused bytes in dst's hash array. */
 273 static inline void oidcpy_with_padding(struct object_id *dst,
 274                                        const struct object_id *src)
 275 {
 276         size_t hashsz;
 277
 278         if (!src->algo)
 279                 hashsz = the_hash_algo->rawsz;
 280         else
 281                 hashsz = hash_algos[src->algo].rawsz;
 282
 283         memcpy(dst->hash, src->hash, hashsz);
 284         memset(dst->hash + hashsz, 0, GIT_MAX_RAWSZ - hashsz);
 285         dst->algo = src->algo;
 286 }
 287
 288 static inline struct object_id *oiddup(const struct object_id *src)
 289 {
 290         struct object_id *dst = xmalloc(sizeof(struct object_id));
 291         oidcpy(dst, src);
 292         return dst;
 293 }
 294
 295 static inline void hashclr(unsigned char *hash)
 296 {
 297         memset(hash, 0, the_hash_algo->rawsz);
 298 }
 299
 300 static inline void oidclr(struct object_id *oid)
 301 {
 302         memset(oid->hash, 0, GIT_MAX_RAWSZ);
 303         oid->algo = hash_algo_by_ptr(the_hash_algo);
 304 }
 305
 306 static inline void oidread(struct object_id *oid, const unsigned char *hash)
 307 {
 308         memcpy(oid->hash, hash, the_hash_algo->rawsz);
 309         oid->algo = hash_algo_by_ptr(the_hash_algo);
 310 }
 311
 312 static inline int is_empty_blob_sha1(const unsigned char *sha1)
 313 {
 314         return hasheq(sha1, the_hash_algo->empty_blob->hash);
 315 }
 316
 317 static inline int is_empty_blob_oid(const struct object_id *oid)
 318 {
 319         return oideq(oid, the_hash_algo->empty_blob);
 320 }
 321
 322 static inline int is_empty_tree_sha1(const unsigned char *sha1)
 323 {
 324         return hasheq(sha1, the_hash_algo->empty_tree->hash);
 325 }
 326
 327 static inline int is_empty_tree_oid(const struct object_id *oid)
 328 {
 329         return oideq(oid, the_hash_algo->empty_tree);
 330 }
 331
 332 static inline void oid_set_algo(struct object_id *oid, const struct git_hash_algo *algop)
 333 {
 334         oid->algo = hash_algo_by_ptr(algop);
 335 }
 336
 337 const char *empty_tree_oid_hex(void);
 338 const char *empty_blob_oid_hex(void);
 339
 340 #endif