/src/wget2/libwget/hashfile.c
Line | Count | Source (jump to first uncovered line) |
1 | | /* |
2 | | * Copyright (c) 2012 Tim Ruehsen |
3 | | * Copyright (c) 2015-2024 Free Software Foundation, Inc. |
4 | | * |
5 | | * This file is part of Wget. |
6 | | * |
7 | | * Wget is free software: you can redistribute it and/or modify |
8 | | * it under the terms of the GNU General Public License as published by |
9 | | * the Free Software Foundation, either version 3 of the License, or |
10 | | * (at your option) any later version. |
11 | | * |
12 | | * Wget is distributed in the hope that it will be useful, |
13 | | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
14 | | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
15 | | * GNU General Public License for more details. |
16 | | * |
17 | | * You should have received a copy of the GNU General Public License |
18 | | * along with Wget. If not, see <https://www.gnu.org/licenses/>. |
19 | | * |
20 | | * |
21 | | * Hash routines |
22 | | * |
23 | | * Changelog |
24 | | * 29.07.2012 Tim Ruehsen created |
25 | | * |
26 | | */ |
27 | | |
28 | | #include <config.h> |
29 | | |
30 | | #include <unistd.h> |
31 | | #include <stdlib.h> |
32 | | #include <fcntl.h> |
33 | | #include <sys/stat.h> |
34 | | #ifdef HAVE_MMAP |
35 | | # include <sys/mman.h> |
36 | | #endif |
37 | | |
38 | | #include <wget.h> |
39 | | #include "private.h" |
40 | | |
41 | | /** |
42 | | * \file |
43 | | * \brief Hashing functions |
44 | | * \ingroup libwget-hash |
45 | | * @{ |
46 | | * |
47 | | */ |
48 | | |
49 | | /** |
50 | | * \param[in] hashname Name of the hashing algorithm (see table below) |
51 | | * \return A constant to be used by libwget hashing functions |
52 | | * |
53 | | * Get the hashing algorithms list item that corresponds to the named hashing algorithm. |
54 | | * |
55 | | * This function returns a constant that uniquely identifies a known supported hashing algorithm |
56 | | * within libwget. All the supported algorithms are listed in the ::wget_digest_algorithm enum. |
57 | | * |
58 | | * Algorithm name | Constant |
59 | | * -------------- | -------- |
60 | | * sha1 or sha-1|WGET_DIGTYPE_SHA1 |
61 | | * sha256 or sha-256|WGET_DIGTYPE_SHA256 |
62 | | * sha512 or sha-512|WGET_DIGTYPE_SHA512 |
63 | | * sha224 or sha-224|WGET_DIGTYPE_SHA224 |
64 | | * sha384 or sha-384|WGET_DIGTYPE_SHA384 |
65 | | * md5|WGET_DIGTYPE_MD5 |
66 | | * md2|WGET_DIGTYPE_MD2 |
67 | | * rmd160|WGET_DIGTYPE_RMD160 |
68 | | */ |
69 | | wget_digest_algorithm wget_hash_get_algorithm(const char *hashname) |
70 | 0 | { |
71 | 0 | if (hashname) { |
72 | 0 | if (*hashname == 's' || *hashname == 'S') { |
73 | 0 | if (!wget_strcasecmp_ascii(hashname, "sha-1") || !wget_strcasecmp_ascii(hashname, "sha1")) |
74 | 0 | return WGET_DIGTYPE_SHA1; |
75 | 0 | else if (!wget_strcasecmp_ascii(hashname, "sha-256") || !wget_strcasecmp_ascii(hashname, "sha256")) |
76 | 0 | return WGET_DIGTYPE_SHA256; |
77 | 0 | else if (!wget_strcasecmp_ascii(hashname, "sha-512") || !wget_strcasecmp_ascii(hashname, "sha512")) |
78 | 0 | return WGET_DIGTYPE_SHA512; |
79 | 0 | else if (!wget_strcasecmp_ascii(hashname, "sha-224") || !wget_strcasecmp_ascii(hashname, "sha224")) |
80 | 0 | return WGET_DIGTYPE_SHA224; |
81 | 0 | else if (!wget_strcasecmp_ascii(hashname, "sha-384") || !wget_strcasecmp_ascii(hashname, "sha384")) |
82 | 0 | return WGET_DIGTYPE_SHA384; |
83 | 0 | } |
84 | 0 | else if (!wget_strcasecmp_ascii(hashname, "md5")) |
85 | 0 | return WGET_DIGTYPE_MD5; |
86 | 0 | else if (!wget_strcasecmp_ascii(hashname, "md2")) |
87 | 0 | return WGET_DIGTYPE_MD2; |
88 | 0 | else if (!wget_strcasecmp_ascii(hashname, "rmd160")) |
89 | 0 | return WGET_DIGTYPE_RMD160; |
90 | 0 | } |
91 | | |
92 | 0 | error_printf(_("Unknown hash type '%s'\n"), hashname); |
93 | 0 | return WGET_DIGTYPE_UNKNOWN; |
94 | 0 | } |
95 | | |
96 | | #if defined WITH_GNUTLS && !defined WITH_LIBNETTLE |
97 | | #include <gnutls/gnutls.h> |
98 | | #ifdef HAVE_GNUTLS_CRYPTO_H |
99 | | # include <gnutls/crypto.h> |
100 | | #endif |
101 | | |
102 | | struct wget_hash_hd_st { |
103 | | gnutls_hash_hd_t |
104 | | dig; |
105 | | }; |
106 | | |
107 | | static const gnutls_digest_algorithm_t |
108 | | algorithms[WGET_DIGTYPE_MAX] = { |
109 | | // [WGET_DIGTYPE_UNKNOWN] = GNUTLS_DIG_UNKNOWN, // both values are 0 |
110 | | [WGET_DIGTYPE_MD2] = GNUTLS_DIG_MD2, |
111 | | [WGET_DIGTYPE_MD5] = GNUTLS_DIG_MD5, |
112 | | [WGET_DIGTYPE_RMD160] = GNUTLS_DIG_RMD160, |
113 | | [WGET_DIGTYPE_SHA1] = GNUTLS_DIG_SHA1, |
114 | | [WGET_DIGTYPE_SHA224] = GNUTLS_DIG_SHA224, |
115 | | [WGET_DIGTYPE_SHA256] = GNUTLS_DIG_SHA256, |
116 | | [WGET_DIGTYPE_SHA384] = GNUTLS_DIG_SHA384, |
117 | | [WGET_DIGTYPE_SHA512] = GNUTLS_DIG_SHA512 |
118 | | }; |
119 | | |
120 | | /** |
121 | | * \param[in] algorithm One of the hashing algorithms returned by wget_hash_get_algorithm() |
122 | | * \param[in] text Input data to hash |
123 | | * \param[in] textlen Length of the input data |
124 | | * \param[in] digest Caller-supplied buffer where the output hash will be placed |
125 | | * \return Zero on success or a negative value on error |
126 | | * |
127 | | * Convenience function to hash the given data in a single call. |
128 | | * |
129 | | * The caller must ensure that the provided output buffer \p digest is large enough |
130 | | * to store the hash. A particular hash algorithm is guaranteed to always generate |
131 | | * the same amount of data (e.g. 512 bits) but different hash algorithms will output |
132 | | * different lengths of data. To get the output length of the chosen algorithm \p algorithm, |
133 | | * call wget_hash_get_len(). |
134 | | * |
135 | | * \note This function's behavior depends on the underlying cryptographic engine libwget was compiled with. |
136 | | */ |
137 | | int wget_hash_fast(wget_digest_algorithm algorithm, const void *text, size_t textlen, void *digest) |
138 | | { |
139 | | if ((unsigned) algorithm >= countof(algorithms)) |
140 | | return WGET_E_INVALID; |
141 | | |
142 | | gnutls_digest_algorithm_t hashtype = algorithms[algorithm]; |
143 | | if (hashtype == GNUTLS_DIG_UNKNOWN) |
144 | | return WGET_E_UNSUPPORTED; |
145 | | |
146 | | if (gnutls_hash_fast(algorithms[algorithm], text, textlen, digest) != 0) |
147 | | return WGET_E_UNKNOWN; |
148 | | |
149 | | return WGET_E_SUCCESS; |
150 | | } |
151 | | |
152 | | /** |
153 | | * \param[in] algorithm One of the hashing algorithms returned by wget_hash_get_algorithm() |
154 | | * \return The length of the output data generated by the algorithm |
155 | | * |
156 | | * Determines the output length of the given hashing algorithm. |
157 | | * |
158 | | * A particular hash algorithm is guaranteed to always generate |
159 | | * the same amount of data (e.g. 512 bits) but different hash algorithms will output |
160 | | * different lengths of data. |
161 | | */ |
162 | | int wget_hash_get_len(wget_digest_algorithm algorithm) |
163 | | { |
164 | | if ((unsigned)algorithm < countof(algorithms)) |
165 | | return gnutls_hash_get_len(algorithms[algorithm]); |
166 | | else |
167 | | return 0; |
168 | | } |
169 | | |
170 | | /** |
171 | | * \param[out] handle Caller-provided pointer to a ::wget_hash_hd structure where the handle to this |
172 | | * hashing primitive will be stored, needed in subsequent calls to wget_hash() |
173 | | * \param[in] algorithm One of the hashing algorithms returned by wget_hash_get_algorithm() |
174 | | * \return Zero on success or a negative value on error |
175 | | * |
176 | | * Initialize the cryptographic engine to compute hashes with the given hashing algorithm, |
177 | | * as well as the hashing algorithm itself. |
178 | | * |
179 | | * After this function returns, wget_hash() might be called as many times as desired. |
180 | | */ |
181 | | int wget_hash_init(wget_hash_hd **handle, wget_digest_algorithm algorithm) |
182 | | { |
183 | | if ((unsigned)algorithm >= countof(algorithms)) |
184 | | return WGET_E_INVALID; |
185 | | |
186 | | gnutls_digest_algorithm_t hashtype = algorithms[algorithm]; |
187 | | if (hashtype == GNUTLS_DIG_UNKNOWN) |
188 | | return WGET_E_UNSUPPORTED; |
189 | | |
190 | | if (!(*handle = wget_malloc(sizeof(struct wget_hash_hd_st)))) |
191 | | return WGET_E_MEMORY; |
192 | | |
193 | | if (gnutls_hash_init(&(*handle)->dig, algorithms[algorithm]) != 0) { |
194 | | xfree(*handle); |
195 | | return WGET_E_UNKNOWN; |
196 | | } |
197 | | |
198 | | return WGET_E_SUCCESS; |
199 | | } |
200 | | |
201 | | /** |
202 | | * \param[in] handle Handle to the hashing primitive returned by a subsequent call to wget_hash_init() |
203 | | * \param[in] text Input data |
204 | | * \param[in] textlen Length of the input data |
205 | | * \return Zero on success or a negative value on error |
206 | | * |
207 | | * Update the digest by adding additional input data to it. This method can be called |
208 | | * as many times as desired. Once finished, call wget_hash_deinit() to complete |
209 | | * the computation and get the resulting hash. |
210 | | */ |
211 | | int wget_hash(wget_hash_hd *handle, const void *text, size_t textlen) |
212 | | { |
213 | | return gnutls_hash(handle->dig, text, textlen) == 0 ? 0 : -1; |
214 | | } |
215 | | |
216 | | /** |
217 | | * \param[in] handle Handle to the hashing primitive returned by a subsequent call to wget_hash_init() |
218 | | * \param[out] digest Caller-supplied buffer where the output hash will be placed. |
219 | | * \return 0 on success, < 0 on failure |
220 | | * |
221 | | * Complete the hash computation by performing final operations, such as padding, |
222 | | * and obtain the final result. The result will be placed in the caller-supplied |
223 | | * buffer \p digest. The caller must ensure that the provided output buffer \p digest |
224 | | * is large enough to store the hash. To get the output length of the chosen algorithm |
225 | | * \p algorithm, call wget_hash_get_len(). |
226 | | */ |
227 | | int wget_hash_deinit(wget_hash_hd **handle, void *digest) |
228 | | { |
229 | | gnutls_hash_deinit((*handle)->dig, digest); |
230 | | |
231 | | xfree(*handle); |
232 | | |
233 | | return WGET_E_SUCCESS; |
234 | | } |
235 | | |
236 | | #elif defined WITH_LIBWOLFCRYPT |
237 | | #define WOLFSSL_SHA224 |
238 | | #define WOLFSSL_SHA384 |
239 | | #define WOLFSSL_SHA512 |
240 | | #define WC_NO_HARDEN |
241 | | #include <wolfssl/options.h> |
242 | | #include <wolfssl/wolfcrypt/hash.h> |
243 | | |
244 | | struct wget_hash_hd_st { |
245 | | wc_HashAlg |
246 | | hash; |
247 | | enum wc_HashType |
248 | | type; |
249 | | }; |
250 | | |
251 | | static const enum wc_HashType |
252 | | algorithms[] = { |
253 | | // [WGET_DIGTYPE_UNKNOWN] = WC_HASH_TYPE_NONE, // both values are 0 |
254 | | // [WGET_DIGTYPE_MD2] = WC_HASH_TYPE_MD2, // not in wc_hashAlg |
255 | | [WGET_DIGTYPE_MD5] = WC_HASH_TYPE_MD5, |
256 | | [WGET_DIGTYPE_SHA1] = WC_HASH_TYPE_SHA, |
257 | | [WGET_DIGTYPE_SHA224] = WC_HASH_TYPE_SHA224, |
258 | | [WGET_DIGTYPE_SHA256] = WC_HASH_TYPE_SHA256, |
259 | | [WGET_DIGTYPE_SHA384] = WC_HASH_TYPE_SHA384, |
260 | | [WGET_DIGTYPE_SHA512] = WC_HASH_TYPE_SHA512, |
261 | | }; |
262 | | |
263 | | int wget_hash_fast(wget_digest_algorithm algorithm, const void *text, size_t textlen, void *digest) |
264 | | { |
265 | | if ((unsigned) algorithm >= countof(algorithms)) |
266 | | return WGET_E_INVALID; |
267 | | |
268 | | enum wc_HashType hashtype = algorithms[algorithm]; |
269 | | if (hashtype == WC_HASH_TYPE_NONE) |
270 | | return WGET_E_UNSUPPORTED; |
271 | | |
272 | | if (wc_Hash(hashtype, text, (unsigned) textlen, digest, wc_HashGetDigestSize(hashtype)) != 0) |
273 | | return WGET_E_UNKNOWN; |
274 | | |
275 | | return WGET_E_SUCCESS; |
276 | | } |
277 | | |
278 | | int wget_hash_get_len(wget_digest_algorithm algorithm) |
279 | | { |
280 | | if ((unsigned) algorithm < countof(algorithms)) { |
281 | | int ret = wc_HashGetDigestSize(algorithms[algorithm]); |
282 | | if (ret > 0) |
283 | | return ret; |
284 | | } |
285 | | |
286 | | return 0; |
287 | | } |
288 | | |
289 | | int wget_hash_init(wget_hash_hd **handle, wget_digest_algorithm algorithm) |
290 | | { |
291 | | if ((unsigned) algorithm >= countof(algorithms)) |
292 | | return WGET_E_INVALID; |
293 | | |
294 | | enum wc_HashType hashtype = algorithms[algorithm]; |
295 | | if (hashtype == WC_HASH_TYPE_NONE) |
296 | | return WGET_E_UNSUPPORTED; |
297 | | |
298 | | if (!(*handle = wget_malloc(sizeof(struct wget_hash_hd_st)))) |
299 | | return WGET_E_MEMORY; |
300 | | |
301 | | if (wc_HashInit(&(*handle)->hash, hashtype) != 0) { |
302 | | xfree(*handle); |
303 | | return WGET_E_UNKNOWN; |
304 | | } |
305 | | |
306 | | (*handle)->type = hashtype; |
307 | | |
308 | | return WGET_E_SUCCESS; |
309 | | } |
310 | | |
311 | | int wget_hash(wget_hash_hd *handle, const void *text, size_t textlen) |
312 | | { |
313 | | if (wc_HashUpdate(&handle->hash, handle->type, text, (unsigned) textlen) == 0) |
314 | | return WGET_E_SUCCESS; |
315 | | |
316 | | return WGET_E_UNKNOWN; |
317 | | } |
318 | | |
319 | | int wget_hash_deinit(wget_hash_hd **handle, void *digest) |
320 | | { |
321 | | int rc = wc_HashFinal(&(*handle)->hash, (*handle)->type, digest); |
322 | | |
323 | | xfree(*handle); |
324 | | |
325 | | return rc == 0 ? rc : WGET_E_UNKNOWN; |
326 | | } |
327 | | |
328 | | #elif defined WITH_LIBCRYPTO |
329 | | #include <openssl/evp.h> |
330 | | |
331 | | typedef const EVP_MD *evp_md_func(void); |
332 | | |
333 | | struct wget_hash_hd_st { |
334 | | EVP_MD_CTX |
335 | | *ctx; |
336 | | }; |
337 | | |
338 | | static evp_md_func * |
339 | | algorithms[] = { |
340 | | // [WGET_DIGTYPE_UNKNOWN] = NULL, |
341 | | // [WGET_DIGTYPE_MD2] = EVP_md2, |
342 | | [WGET_DIGTYPE_MD5] = EVP_md5, |
343 | | [WGET_DIGTYPE_RMD160] = EVP_ripemd160, |
344 | | [WGET_DIGTYPE_SHA1] = EVP_sha1, |
345 | | [WGET_DIGTYPE_SHA224] = EVP_sha224, |
346 | | [WGET_DIGTYPE_SHA256] = EVP_sha256, |
347 | | [WGET_DIGTYPE_SHA384] = EVP_sha384, |
348 | | [WGET_DIGTYPE_SHA512] = EVP_sha512, |
349 | | }; |
350 | | |
351 | | int wget_hash_fast(wget_digest_algorithm algorithm, const void *text, size_t textlen, void *digest) |
352 | | { |
353 | | if ((unsigned) algorithm >= countof(algorithms)) |
354 | | return WGET_E_INVALID; |
355 | | |
356 | | evp_md_func *evp = algorithms[algorithm]; |
357 | | if (!evp) |
358 | | return WGET_E_UNSUPPORTED; |
359 | | |
360 | | if (EVP_Digest(text, textlen, digest, NULL, evp(), NULL) == 0) |
361 | | return WGET_E_UNKNOWN; |
362 | | |
363 | | return WGET_E_SUCCESS; |
364 | | } |
365 | | |
366 | | int wget_hash_get_len(wget_digest_algorithm algorithm) |
367 | | { |
368 | | evp_md_func *evp; |
369 | | |
370 | | if ((unsigned) algorithm >= countof(algorithms) |
371 | | || (evp = algorithms[algorithm]) == NULL) |
372 | | return 0; |
373 | | |
374 | | return EVP_MD_size(evp()); |
375 | | } |
376 | | |
377 | | int wget_hash_init(wget_hash_hd **handle, wget_digest_algorithm algorithm) |
378 | | { |
379 | | evp_md_func *evp; |
380 | | |
381 | | if ((unsigned) algorithm >= countof(algorithms)) |
382 | | return WGET_E_UNSUPPORTED; |
383 | | |
384 | | if ((evp = algorithms[algorithm]) == NULL) |
385 | | return WGET_E_UNSUPPORTED; |
386 | | |
387 | | if (!(*handle = wget_malloc(sizeof(struct wget_hash_hd_st)))) |
388 | | return WGET_E_MEMORY; |
389 | | |
390 | | if (!((*handle)->ctx = EVP_MD_CTX_new())) { |
391 | | xfree(*handle); |
392 | | return WGET_E_UNKNOWN; |
393 | | } |
394 | | |
395 | | if (EVP_DigestInit_ex((*handle)->ctx, evp(), NULL)) |
396 | | return WGET_E_SUCCESS; |
397 | | |
398 | | EVP_MD_CTX_free((*handle)->ctx); |
399 | | xfree(*handle); |
400 | | |
401 | | return WGET_E_UNKNOWN; |
402 | | } |
403 | | |
404 | | int wget_hash(wget_hash_hd *handle, const void *text, size_t textlen) |
405 | | { |
406 | | if (EVP_DigestUpdate(handle->ctx, text, textlen)) |
407 | | return WGET_E_SUCCESS; |
408 | | |
409 | | return WGET_E_INVALID; |
410 | | } |
411 | | |
412 | | int wget_hash_deinit(wget_hash_hd **handle, void *digest) |
413 | | { |
414 | | EVP_DigestFinal_ex((*handle)->ctx, digest, NULL); |
415 | | |
416 | | EVP_MD_CTX_free((*handle)->ctx); |
417 | | xfree(*handle); |
418 | | |
419 | | return WGET_E_SUCCESS; |
420 | | } |
421 | | |
422 | | #elif defined WITH_LIBNETTLE |
423 | | #include <nettle/nettle-meta.h> |
424 | | #include <nettle/md2.h> |
425 | | #include <nettle/md5.h> |
426 | | #include <nettle/ripemd160.h> |
427 | | #include <nettle/sha2.h> |
428 | | |
429 | | struct wget_hash_hd_st { |
430 | | const struct nettle_hash |
431 | | *hash; |
432 | | void |
433 | | *context; |
434 | | }; |
435 | | |
436 | | static const struct nettle_hash * |
437 | | algorithms[WGET_DIGTYPE_MAX] = { |
438 | | // [WGET_DIGTYPE_UNKNOWN] = NULL, |
439 | | [WGET_DIGTYPE_MD2] = &nettle_md2, |
440 | | [WGET_DIGTYPE_MD5] = &nettle_md5, |
441 | | #ifdef RIPEMD160_DIGEST_SIZE |
442 | | [WGET_DIGTYPE_RMD160] = &nettle_ripemd160, |
443 | | #endif |
444 | | [WGET_DIGTYPE_SHA1] = &nettle_sha1, |
445 | | #ifdef SHA224_DIGEST_SIZE |
446 | | [WGET_DIGTYPE_SHA224] = &nettle_sha224, |
447 | | #endif |
448 | | #ifdef SHA256_DIGEST_SIZE |
449 | | [WGET_DIGTYPE_SHA256] = &nettle_sha256, |
450 | | #endif |
451 | | #ifdef SHA384_DIGEST_SIZE |
452 | | [WGET_DIGTYPE_SHA384] = &nettle_sha384, |
453 | | #endif |
454 | | #ifdef SHA512_DIGEST_SIZE |
455 | | [WGET_DIGTYPE_SHA512] = &nettle_sha512, |
456 | | #endif |
457 | | }; |
458 | | |
459 | | int wget_hash_fast(wget_digest_algorithm algorithm, const void *text, size_t textlen, void *digest) |
460 | 326 | { |
461 | 326 | wget_hash_hd *dig; |
462 | 326 | int rc; |
463 | | |
464 | 326 | if ((rc = wget_hash_init(&dig, algorithm)) == 0) { |
465 | 326 | rc = wget_hash(dig, text, textlen); |
466 | 326 | wget_hash_deinit(&dig, digest); |
467 | 326 | } |
468 | | |
469 | 326 | return rc; |
470 | 326 | } |
471 | | |
472 | | int wget_hash_get_len(wget_digest_algorithm algorithm) |
473 | 1.72k | { |
474 | 1.72k | if ((unsigned)algorithm < countof(algorithms)) |
475 | 1.72k | return algorithms[algorithm]->digest_size; |
476 | 0 | else |
477 | 0 | return 0; |
478 | 1.72k | } |
479 | | |
480 | | int wget_hash_init(wget_hash_hd **handle, wget_digest_algorithm algorithm) |
481 | 326 | { |
482 | 326 | if ((unsigned)algorithm >= countof(algorithms)) |
483 | 0 | return WGET_E_INVALID; |
484 | | |
485 | 326 | if (!algorithms[algorithm]) |
486 | 0 | return WGET_E_UNSUPPORTED; |
487 | | |
488 | 326 | wget_hash_hd *h; |
489 | | |
490 | 326 | if (!(h = wget_malloc(sizeof(struct wget_hash_hd_st)))) |
491 | 0 | return WGET_E_MEMORY; |
492 | | |
493 | 326 | h->hash = algorithms[algorithm]; |
494 | | |
495 | 326 | if (!(h->context = wget_malloc(h->hash->context_size))) { |
496 | 0 | xfree(h); |
497 | 0 | return WGET_E_MEMORY; |
498 | 0 | } |
499 | | |
500 | 326 | h->hash->init(h->context); |
501 | 326 | *handle = h; |
502 | | |
503 | 326 | return WGET_E_SUCCESS; |
504 | 326 | } |
505 | | |
506 | | int wget_hash(wget_hash_hd *handle, const void *text, size_t textlen) |
507 | 326 | { |
508 | 326 | handle->hash->update(handle->context, textlen, text); |
509 | 326 | return WGET_E_SUCCESS; |
510 | 326 | } |
511 | | |
512 | | int wget_hash_deinit(wget_hash_hd **handle, void *digest) |
513 | 326 | { |
514 | 326 | (*handle)->hash->digest((*handle)->context, (*handle)->hash->digest_size, digest); |
515 | 326 | xfree((*handle)->context); |
516 | 326 | xfree(*handle); |
517 | 326 | return WGET_E_SUCCESS; |
518 | 326 | } |
519 | | |
520 | | #elif defined WITH_GCRYPT |
521 | | #ifdef HAVE_GCRYPT_H |
522 | | #include <gcrypt.h> |
523 | | #endif |
524 | | |
525 | | struct wget_hash_hd_st { |
526 | | int |
527 | | algorithm; |
528 | | gcry_md_hd_t |
529 | | context; |
530 | | }; |
531 | | |
532 | | static const int algorithms[] = { |
533 | | // [WGET_DIGTYPE_UNKNOWN] = GCRY_MD_NONE, |
534 | | [WGET_DIGTYPE_MD2] = GCRY_MD_MD2, |
535 | | [WGET_DIGTYPE_MD5] = GCRY_MD_MD5, |
536 | | [WGET_DIGTYPE_RMD160] = GCRY_MD_RMD160, |
537 | | [WGET_DIGTYPE_SHA1] = GCRY_MD_SHA1, |
538 | | [WGET_DIGTYPE_SHA224] = GCRY_MD_SHA224, |
539 | | [WGET_DIGTYPE_SHA256] = GCRY_MD_SHA256, |
540 | | [WGET_DIGTYPE_SHA384] = GCRY_MD_SHA384, |
541 | | [WGET_DIGTYPE_SHA512] = GCRY_MD_SHA512 |
542 | | }; |
543 | | |
544 | | int wget_hash_fast(wget_digest_algorithm algorithm, const void *text, size_t textlen, void *digest) |
545 | | { |
546 | | wget_hash_hd *dig; |
547 | | int rc; |
548 | | |
549 | | if ((rc = wget_hash_init(&dig, algorithm)) == 0) { |
550 | | rc = wget_hash(dig, text, textlen); |
551 | | wget_hash_deinit(&dig, digest); |
552 | | } |
553 | | |
554 | | return rc; |
555 | | } |
556 | | |
557 | | int wget_hash_get_len(wget_digest_algorithm algorithm) |
558 | | { |
559 | | if ((unsigned)algorithm < countof(algorithms)) |
560 | | return gcry_md_get_algo_dlen(algorithms[algorithm]); |
561 | | else |
562 | | return 0; |
563 | | } |
564 | | |
565 | | int wget_hash_init(wget_hash_hd **handle, wget_digest_algorithm algorithm) |
566 | | { |
567 | | if ((unsigned)algorithm >= countof(algorithms)) |
568 | | return WGET_E_INVALID; |
569 | | |
570 | | if (!algorithms[algorithm]) |
571 | | return WGET_E_UNSUPPORTED; |
572 | | |
573 | | wget_hash_hd *h; |
574 | | |
575 | | if (!(h = wget_malloc(sizeof(struct wget_hash_hd_st)))) |
576 | | return WGET_E_MEMORY; |
577 | | |
578 | | h->algorithm = algorithms[algorithm]; |
579 | | gcry_md_open(&h->context, h->algorithm, 0); |
580 | | |
581 | | *handle = h; |
582 | | |
583 | | return WGET_E_SUCCESS; |
584 | | } |
585 | | |
586 | | int wget_hash(wget_hash_hd *handle, const void *text, size_t textlen) |
587 | | { |
588 | | gcry_md_write(handle->context, text, textlen); |
589 | | return 0; |
590 | | } |
591 | | |
592 | | int wget_hash_deinit(wget_hash_hd **handle, void *digest) |
593 | | { |
594 | | gcry_md_final((*handle)->context); |
595 | | void *ret = gcry_md_read((*handle)->context, (*handle)->algorithm); |
596 | | memcpy(digest, ret, gcry_md_get_algo_dlen((*handle)->algorithm)); |
597 | | gcry_md_close((*handle)->context); |
598 | | xfree(*handle); |
599 | | |
600 | | return WGET_E_SUCCESS; |
601 | | } |
602 | | |
603 | | #else // use the gnulib functions |
604 | | |
605 | | #pragma GCC diagnostic push |
606 | | #pragma GCC diagnostic ignored "-Wundef" |
607 | | #include "md2.h" |
608 | | #include "md5.h" |
609 | | #include "sha1.h" |
610 | | #include "sha256.h" |
611 | | #include "sha512.h" |
612 | | #pragma GCC diagnostic pop |
613 | | |
614 | | typedef void (*_hash_init_t)(void *); |
615 | | typedef void (*_hash_process_t)(const void *, size_t, void *); |
616 | | typedef void (*_hash_finish_t)(void *, void *); |
617 | | typedef void (*_hash_read_t)(const void *, void *); |
618 | | |
619 | | static struct _algorithm { |
620 | | _hash_init_t init; |
621 | | _hash_process_t process; |
622 | | _hash_finish_t finish; |
623 | | _hash_read_t read; |
624 | | size_t ctx_len; |
625 | | size_t digest_len; |
626 | | } _algorithm[WGET_DIGTYPE_MAX] = { |
627 | | [WGET_DIGTYPE_MD2] = { |
628 | | (_hash_init_t)md2_init_ctx, |
629 | | (_hash_process_t)md2_process_bytes, |
630 | | (_hash_finish_t)md2_finish_ctx, |
631 | | (_hash_read_t)md2_read_ctx, |
632 | | sizeof(struct md2_ctx), |
633 | | MD2_DIGEST_SIZE |
634 | | }, |
635 | | [WGET_DIGTYPE_MD5] = { |
636 | | (_hash_init_t)md5_init_ctx, |
637 | | (_hash_process_t)md5_process_bytes, |
638 | | (_hash_finish_t)md5_finish_ctx, |
639 | | (_hash_read_t)md5_read_ctx, |
640 | | sizeof(struct md5_ctx), |
641 | | MD5_DIGEST_SIZE |
642 | | }, |
643 | | [WGET_DIGTYPE_SHA1] = { |
644 | | (_hash_init_t)sha1_init_ctx, |
645 | | (_hash_process_t)sha1_process_bytes, |
646 | | (_hash_finish_t)sha1_finish_ctx, |
647 | | (_hash_read_t)sha1_read_ctx, |
648 | | sizeof(struct sha1_ctx), |
649 | | SHA1_DIGEST_SIZE |
650 | | }, |
651 | | [WGET_DIGTYPE_SHA224] = { |
652 | | (_hash_init_t)sha224_init_ctx, |
653 | | (_hash_process_t)sha256_process_bytes, // sha256 is intentional |
654 | | (_hash_finish_t)sha224_finish_ctx, |
655 | | (_hash_read_t)sha224_read_ctx, |
656 | | sizeof(struct sha256_ctx), // sha256 is intentional |
657 | | SHA224_DIGEST_SIZE |
658 | | }, |
659 | | [WGET_DIGTYPE_SHA256] = { |
660 | | (_hash_init_t)sha256_init_ctx, |
661 | | (_hash_process_t)sha256_process_bytes, |
662 | | (_hash_finish_t)sha256_finish_ctx, |
663 | | (_hash_read_t)sha256_read_ctx, |
664 | | sizeof(struct sha256_ctx), |
665 | | SHA256_DIGEST_SIZE |
666 | | }, |
667 | | [WGET_DIGTYPE_SHA384] = { |
668 | | (_hash_init_t)sha384_init_ctx, |
669 | | (_hash_process_t)sha512_process_bytes, // sha512 is intentional |
670 | | (_hash_finish_t)sha384_finish_ctx, |
671 | | (_hash_read_t)sha384_read_ctx, |
672 | | sizeof(struct sha512_ctx), // sha512 is intentional |
673 | | SHA384_DIGEST_SIZE |
674 | | }, |
675 | | [WGET_DIGTYPE_SHA512] = { |
676 | | (_hash_init_t)sha512_init_ctx, |
677 | | (_hash_process_t)sha512_process_bytes, |
678 | | (_hash_finish_t)sha512_finish_ctx, |
679 | | (_hash_read_t)sha512_read_ctx, |
680 | | sizeof(struct sha512_ctx), |
681 | | SHA512_DIGEST_SIZE |
682 | | } |
683 | | }; |
684 | | |
685 | | struct wget_hash_hd_st { |
686 | | const struct _algorithm |
687 | | *algorithm; |
688 | | void |
689 | | *context; |
690 | | }; |
691 | | |
692 | | int wget_hash_fast(wget_digest_algorithm algorithm, const void *text, size_t textlen, void *digest) |
693 | | { |
694 | | wget_hash_hd *dig; |
695 | | int rc; |
696 | | |
697 | | if ((rc = wget_hash_init(&dig, algorithm)) == WGET_E_SUCCESS) { |
698 | | rc = wget_hash(dig, text, textlen); |
699 | | wget_hash_deinit(&dig, digest); |
700 | | } |
701 | | |
702 | | return rc; |
703 | | } |
704 | | |
705 | | int wget_hash_get_len(wget_digest_algorithm algorithm) |
706 | | { |
707 | | if ((unsigned)algorithm < countof(_algorithm)) |
708 | | return (int) _algorithm[algorithm].digest_len; |
709 | | else |
710 | | return 0; |
711 | | } |
712 | | |
713 | | int wget_hash_init(wget_hash_hd **handle, wget_digest_algorithm algorithm) |
714 | | { |
715 | | if ((unsigned)algorithm >= countof(_algorithm)) |
716 | | return WGET_E_INVALID; |
717 | | |
718 | | if (!_algorithm[algorithm].ctx_len) |
719 | | return WGET_E_UNSUPPORTED; |
720 | | |
721 | | wget_hash_hd *h; |
722 | | |
723 | | if (!(h = wget_malloc(sizeof(struct wget_hash_hd_st)))) |
724 | | return WGET_E_MEMORY; |
725 | | |
726 | | h->algorithm = &_algorithm[algorithm]; |
727 | | |
728 | | if (!(h->context = wget_malloc(h->algorithm->ctx_len))) { |
729 | | xfree(h); |
730 | | return WGET_E_MEMORY; |
731 | | } |
732 | | |
733 | | h->algorithm->init(h->context); |
734 | | *handle = h; |
735 | | |
736 | | return WGET_E_SUCCESS; |
737 | | } |
738 | | |
739 | | int wget_hash(wget_hash_hd *handle, const void *text, size_t textlen) |
740 | | { |
741 | | handle->algorithm->process(text, textlen, handle->context); |
742 | | return 0; |
743 | | } |
744 | | |
745 | | int wget_hash_deinit(wget_hash_hd **handle, void *digest) |
746 | | { |
747 | | (*handle)->algorithm->finish((*handle)->context, digest); |
748 | | xfree((*handle)->context); |
749 | | xfree(*handle); |
750 | | |
751 | | return WGET_E_SUCCESS; |
752 | | } |
753 | | #endif |
754 | | |
755 | | /** |
756 | | * \param[in] hashname Name of the hashing algorithm. See wget_hash_get_algorithm() |
757 | | * \param[in] fd File descriptor for the target file |
758 | | * \param[out] digest_hex caller-supplied buffer that will contain the resulting hex string |
759 | | * \param[in] digest_hex_size Length of \p digest_hex |
760 | | * \param[in] offset File offset to start hashing at |
761 | | * \param[in] length Number of bytes to hash, starting from \p offset. Zero will hash up to the end of the file |
762 | | * \return 0 on success or -1 in case of failure |
763 | | * |
764 | | * Compute the hash of the contents of the target file and return its hex representation. |
765 | | * |
766 | | * This function will encode the resulting hash in a string of hex digits, and |
767 | | * place that string in the user-supplied buffer \p digest_hex. |
768 | | */ |
769 | | int wget_hash_file_fd(const char *hashname, int fd, char *digest_hex, size_t digest_hex_size, off_t offset, off_t length) |
770 | 0 | { |
771 | 0 | wget_digest_algorithm algorithm; |
772 | 0 | int ret = WGET_E_UNKNOWN; |
773 | 0 | struct stat st; |
774 | |
|
775 | 0 | if (digest_hex_size) |
776 | 0 | *digest_hex=0; |
777 | |
|
778 | 0 | if (fd == -1 || fstat(fd, &st) != 0) |
779 | 0 | return WGET_E_IO; |
780 | | |
781 | 0 | if (length == 0) |
782 | 0 | length = st.st_size; |
783 | |
|
784 | 0 | if (offset + length > st.st_size) |
785 | 0 | return WGET_E_INVALID; |
786 | | |
787 | 0 | debug_printf("%s hashing pos %llu, length %llu...\n", hashname, (unsigned long long)offset, (unsigned long long)length); |
788 | |
|
789 | 0 | if ((algorithm = wget_hash_get_algorithm(hashname)) != WGET_DIGTYPE_UNKNOWN) { |
790 | 0 | unsigned char digest[256]; |
791 | 0 | size_t digestlen = wget_hash_get_len(algorithm); |
792 | |
|
793 | 0 | if (digestlen > sizeof(digest)) { |
794 | 0 | error_printf(_("%s: Unexpected hash len %zu > %zu\n"), __func__, digestlen, sizeof(digest)); |
795 | 0 | return ret; |
796 | 0 | } |
797 | | |
798 | 0 | #ifdef HAVE_MMAP |
799 | 0 | char *buf = mmap(NULL, length, PROT_READ, MAP_PRIVATE, fd, offset); |
800 | |
|
801 | 0 | if (buf != MAP_FAILED) { |
802 | 0 | if (wget_hash_fast(algorithm, buf, length, digest) == 0) { |
803 | 0 | wget_memtohex(digest, digestlen, digest_hex, digest_hex_size); |
804 | 0 | debug_printf(" hash %s", digest_hex); |
805 | 0 | ret = WGET_E_SUCCESS; |
806 | 0 | } |
807 | 0 | munmap(buf, length); |
808 | 0 | } else { |
809 | 0 | #endif |
810 | | // Fallback to read |
811 | 0 | ssize_t nbytes = 0; |
812 | 0 | wget_hash_hd *dig; |
813 | 0 | char tmp[65536]; |
814 | |
|
815 | 0 | if ((ret = wget_hash_init(&dig, algorithm))) { |
816 | 0 | error_printf(_("%s: Hash init failed for type '%s': %s\n"), __func__, hashname, wget_strerror(ret)); |
817 | 0 | return ret; |
818 | 0 | } |
819 | | |
820 | 0 | while (length > 0 && (nbytes = read(fd, tmp, sizeof(tmp))) > 0) { |
821 | 0 | if ((ret = wget_hash(dig, tmp, nbytes))) { |
822 | 0 | error_printf(_("%s: Hash update failed: %s\n"), __func__, wget_strerror(ret)); |
823 | 0 | return ret; |
824 | 0 | } |
825 | | |
826 | 0 | if (nbytes <= length) |
827 | 0 | length -= nbytes; |
828 | 0 | else |
829 | 0 | length = 0; |
830 | 0 | } |
831 | | |
832 | 0 | if ((ret = wget_hash_deinit(&dig, digest))) { |
833 | 0 | error_printf(_("%s: Hash finalization failed: %s\n"), __func__, wget_strerror(ret)); |
834 | 0 | return ret; |
835 | 0 | } |
836 | | |
837 | 0 | if (nbytes < 0) { |
838 | 0 | error_printf(_("%s: Failed to read %llu bytes\n"), __func__, (unsigned long long)length); |
839 | 0 | return WGET_E_IO; |
840 | 0 | } |
841 | | |
842 | 0 | wget_memtohex(digest, digestlen, digest_hex, digest_hex_size); |
843 | 0 | ret = WGET_E_SUCCESS; |
844 | 0 | #ifdef HAVE_MMAP |
845 | 0 | } |
846 | 0 | #endif |
847 | 0 | } |
848 | | |
849 | 0 | return ret; |
850 | 0 | } |
851 | | |
852 | | /** |
853 | | * \param[in] hashname Name of the hashing algorithm. See wget_hash_get_algorithm() |
854 | | * \param[in] fname Target file name |
855 | | * \param[out] digest_hex Caller-supplied buffer that will contain the resulting hex string |
856 | | * \param[in] digest_hex_size Length of \p digest_hex |
857 | | * \param[in] offset File offset to start hashing at |
858 | | * \param[in] length Number of bytes to hash, starting from \p offset. Zero will hash up to the end of the file |
859 | | * \return 0 on success or -1 in case of failure |
860 | | * |
861 | | * Compute the hash of the contents of the target file starting from \p offset and up to \p length bytes |
862 | | * and return its hex representation. |
863 | | * |
864 | | * This function will encode the resulting hash in a string of hex digits, and |
865 | | * place that string in the user-supplied buffer \p digest_hex. |
866 | | */ |
867 | | int wget_hash_file_offset(const char *hashname, const char *fname, char *digest_hex, size_t digest_hex_size, off_t offset, off_t length) |
868 | 0 | { |
869 | 0 | int fd, ret; |
870 | |
|
871 | 0 | if ((fd = open(fname, O_RDONLY|O_BINARY)) == -1) { |
872 | 0 | if (digest_hex_size) |
873 | 0 | *digest_hex=0; |
874 | 0 | return 0; |
875 | 0 | } |
876 | | |
877 | 0 | ret = wget_hash_file_fd(hashname, fd, digest_hex, digest_hex_size, offset, length); |
878 | 0 | close(fd); |
879 | |
|
880 | 0 | return ret; |
881 | 0 | } |
882 | | |
883 | | /** |
884 | | * \param[in] hashname Name of the hashing algorithm. See wget_hash_get_algorithm() |
885 | | * \param[in] fname Target file name |
886 | | * \param[out] digest_hex Caller-supplied buffer that will contain the resulting hex string |
887 | | * \param[in] digest_hex_size Length of \p digest_hex |
888 | | * \return 0 on success or -1 in case of failure |
889 | | * |
890 | | * Compute the hash of the contents of the target file and return its hex representation. |
891 | | * |
892 | | * This function will encode the resulting hash in a string of hex digits, and |
893 | | * place that string in the user-supplied buffer \p digest_hex. |
894 | | */ |
895 | | int wget_hash_file(const char *hashname, const char *fname, char *digest_hex, size_t digest_hex_size) |
896 | 0 | { |
897 | 0 | return wget_hash_file_offset(hashname, fname, digest_hex, digest_hex_size, 0, 0); |
898 | 0 | } |
899 | | |
900 | | /**@}*/ |