4 This is a high-level summary of the most important changes.
5 For a full list of changes, see the [git commit log][log] and
6 pick the appropriate release branch.
8 [log]: https://github.com/openssl/openssl/commits/
13 - [OpenSSL 3.0](#openssl-30)
14 - [OpenSSL 1.1.1](#openssl-111)
15 - [OpenSSL 1.1.0](#openssl-110)
16 - [OpenSSL 1.0.2](#openssl-102)
17 - [OpenSSL 1.0.1](#openssl-101)
18 - [OpenSSL 1.0.0](#openssl-100)
19 - [OpenSSL 0.9.x](#openssl-09x)
24 ### Changes between 1.1.1 and 3.0 [xx XXX xxxx]
26 * Remove the RAND_DRBG API
28 The RAND_DRBG API did not fit well into the new provider concept as
29 implemented by EVP_RAND and EVP_RAND_CTX. The main reason is that the
30 RAND_DRBG API is a mixture of 'front end' and 'back end' API calls
31 and some of its API calls are rather low-level. This holds in particular
32 for the callback mechanism (RAND_DRBG_set_callbacks()).
34 Adding a compatibility layer to continue supporting the RAND_DRBG API as
35 a legacy API for a regular deprecation period turned out to come at the
36 price of complicating the new provider API unnecessarily. Since the
37 RAND_DRBG API exists only since version 1.1.1, it was decided by the OMC
40 *Paul Dale and Matthias St. Pierre*
42 * Allow SSL_set1_host() and SSL_add1_host() to take IP literal addresses
43 as well as actual hostnames.
47 * The 'MinProtocol' and 'MaxProtocol' configuration commands now silently
48 ignore TLS protocol version bounds when configuring DTLS-based contexts, and
49 conversely, silently ignore DTLS protocol version bounds when configuring
50 TLS-based contexts. The commands can be repeated to set bounds of both
51 types. The same applies with the corresponding "min_protocol" and
52 "max_protocol" command-line switches, in case some application uses both TLS
55 SSL_CTX instances that are created for a fixed protocol version (e.g.
56 TLSv1_server_method()) also silently ignore version bounds. Previously
57 attempts to apply bounds to these protocol versions would result in an
58 error. Now only the "version-flexible" SSL_CTX instances are subject to
59 limits in configuration files in command-line options.
63 * Deprecated the `ENGINE` API. Engines should be replaced with providers
68 * Reworked the recorded ERR codes to make better space for system errors.
69 To distinguish them, the macro `ERR_SYSTEM_ERROR()` indicates if the
70 given code is a system error (true) or an OpenSSL error (false).
74 * Reworked the test perl framework to better allow parallel testing.
76 *Nicola Tuveri and David von Oheimb*
78 * Added ciphertext stealing algorithms AES-128-CBC-CTS, AES-192-CBC-CTS and
79 AES-256-CBC-CTS to the providers. CS1, CS2 and CS3 variants are supported.
83 * 'Configure' has been changed to figure out the configuration target if
84 none is given on the command line. Consequently, the 'config' script is
85 now only a mere wrapper. All documentation is changed to only mention
88 *Rich Salz and Richard Levitte*
90 * Added a library context that applications as well as other
91 libraries can use to form a separate context within which libcrypto
92 operations are performed.
94 There are two ways this can be used:
96 - Directly, by passing a library context to functions that take
97 such an argument, such as `EVP_CIPHER_fetch` and similar algorithm
99 - Indirectly, by creating a new library context and then assigning
100 it as the new default, with `OPENSSL_CTX_set0_default`.
102 All public OpenSSL functions that take an `OPENSSL_CTX` pointer,
103 apart from the functions directly related to `OPENSSL_CTX`, accept
104 NULL to indicate that the default library context should be used.
106 Library code that changes the default library context using
107 `OPENSSL_CTX_set0_default` should take care to restore it with a
108 second call before returning to the caller.
112 * Handshake now fails if Extended Master Secret extension is dropped
117 * Dropped interactive mode from the 'openssl' program. From now on,
118 the `openssl` command without arguments is equivalent to `openssl
123 * Renamed EVP_PKEY_cmp() to EVP_PKEY_eq() and
124 EVP_PKEY_cmp_parameters() to EVP_PKEY_parameters_eq().
125 While the old function names have been retained for backward compatibility
126 they should not be used in new developments
127 because their return values are confusing: Unlike other `_cmp()` functions
128 they do not return 0 in case their arguments are equal.
132 * Deprecated EC_METHOD_get_field_type(). Applications should switch to
133 EC_GROUP_get_field_type().
137 * Deprecated EC_GFp_simple_method(), EC_GFp_mont_method(),
138 EC_GF2m_simple_method(), EC_GFp_nist_method(), EC_GFp_nistp224_method()
139 EC_GFp_nistp256_method(), and EC_GFp_nistp521_method().
140 Applications should rely on the library automatically assigning a suitable
141 EC_METHOD internally upon EC_GROUP construction.
145 * Deprecated EC_GROUP_new(), EC_GROUP_method_of(), and EC_POINT_method_of().
146 EC_METHOD is now an internal-only concept and a suitable EC_METHOD is
147 assigned internally without application intervention.
148 Users of EC_GROUP_new() should switch to a different suitable constructor.
152 * Add CAdES-BES signature verification support, mostly derived
153 from ESSCertIDv2 TS (RFC 5816) contribution by Marek Klein.
155 *Filipe Raimundo da Silva*
157 * Add CAdES-BES signature scheme and attributes support (RFC 5126) to CMS API.
161 * Deprecated EC_POINT_make_affine() and EC_POINTs_make_affine(). These
162 functions are not widely used and now OpenSSL automatically perform this
163 conversion when needed.
167 * Deprecated EC_GROUP_precompute_mult(), EC_GROUP_have_precompute_mult(), and
168 EC_KEY_precompute_mult(). These functions are not widely used and
169 applications should instead switch to named curves which OpenSSL has
170 hardcoded lookup tables for.
174 * Deprecated EC_POINTs_mul(). This function is not widely used and applications
175 should instead use the L<EC_POINT_mul(3)> function.
179 * Removed FIPS_mode() and FIPS_mode_set(). These functions are legacy API's
180 that are not applicable to the new provider model. Applications should
181 instead use EVP_default_properties_is_fips_enabled() and
182 EVP_default_properties_enable_fips().
186 * The SSL option SSL_OP_IGNORE_UNEXPECTED_EOF is introduced. If that option
187 is set, an unexpected EOF is ignored, it pretends a close notify was received
188 instead and so the returned error becomes SSL_ERROR_ZERO_RETURN.
192 * Deprecated EC_POINT_set_Jprojective_coordinates_GFp() and
193 EC_POINT_get_Jprojective_coordinates_GFp(). These functions are not widely
194 used and applications should instead use the
195 L<EC_POINT_set_affine_coordinates(3)> and
196 L<EC_POINT_get_affine_coordinates(3)> functions.
200 * Added OSSL_PARAM_BLD to the public interface. This allows OSSL_PARAM
201 arrays to be more easily constructed via a series of utility functions.
202 Create a parameter builder using OSSL_PARAM_BLD_new(), add parameters using
203 the various push functions and finally convert to a passable OSSL_PARAM
204 array using OSSL_PARAM_BLD_to_param().
208 * The security strength of SHA1 and MD5 based signatures in TLS has been
209 reduced. This results in SSL 3, TLS 1.0, TLS 1.1 and DTLS 1.0 no longer
210 working at the default security level of 1 and instead requires security
211 level 0. The security level can be changed either using the cipher string
212 with @SECLEVEL, or calling SSL_CTX_set_security_level().
216 * EVP_PKEY_get0_RSA(), EVP_PKEY_get0_DSA(), EVP_PKEY_get0_DH(), and
217 EVP_PKEY_get0_EC_KEY() can now handle EVP_PKEYs with provider side
218 internal keys, if they correspond to one of those built in types.
222 * Added EVP_PKEY_set_type_by_keymgmt(), to initialise an EVP_PKEY to
223 contain a provider side internal key.
227 * ASN1_verify(), ASN1_digest() and ASN1_sign() have been deprecated.
228 They are old functions that we don't use, and that you could disable with
229 the macro NO_ASN1_OLD. This goes all the way back to OpenSSL 0.9.7.
233 * Project text documents not yet having a proper file name extension
234 (`HACKING`, `LICENSE`, `NOTES*`, `README*`, `VERSION`) have been renamed to
235 `*.md` as far as reasonable, else `*.txt`, for better use with file managers.
239 * The main project documents (README, NEWS, CHANGES, INSTALL, SUPPORT)
240 have been converted to Markdown with the goal to produce documents
241 which not only look pretty when viewed online in the browser, but
242 remain well readable inside a plain text editor.
244 To achieve this goal, a 'minimalistic' Markdown style has been applied
245 which avoids formatting elements that interfere too much with the
246 reading flow in the text file. For example, it
248 * avoids [ATX headings][] and uses [setext headings][] instead
249 (which works for `<h1>` and `<h2>` headings only).
250 * avoids [inline links][] and uses [reference links][] instead.
251 * avoids [fenced code blocks][] and uses [indented code blocks][] instead.
253 [ATX headings]: https://github.github.com/gfm/#atx-headings
254 [setext headings]: https://github.github.com/gfm/#setext-headings
255 [inline links]: https://github.github.com/gfm/#inline-link
256 [reference links]: https://github.github.com/gfm/#reference-link
257 [fenced code blocks]: https://github.github.com/gfm/#fenced-code-blocks
258 [indented code blocks]: https://github.github.com/gfm/#indented-code-blocks
260 *Matthias St. Pierre*
262 * The test suite is changed to preserve results of each test recipe.
263 A new directory test-runs/ with subdirectories named like the
264 test recipes are created in the build tree for this purpose.
268 * Added an implementation of CMP and CRMF (RFC 4210, RFC 4211 RFC 6712).
269 This adds crypto/cmp/, crpyto/crmf/, apps/cmp.c, and test/cmp_*.
270 See L<openssl-cmp(1)> and L<OSSL_CMP_exec_IR_ses(3)> as starting points.
272 *David von Oheimb, Martin Peylo*
274 * Generalized the HTTP client code from crypto/ocsp/ into crpyto/http/.
275 The legacy OCSP-focused and only partly documented API is retained.
276 See L<OSSL_CMP_MSG_http_perform(3)> etc. for details.
280 * Added `util/check-format.pl`, a tool for checking adherence to the
281 OpenSSL coding style <https://www.openssl.org/policies/codingstyle.html>.
282 The checks performed are incomplete and yield some false positives.
283 Still the tool should be useful for detecting most typical glitches.
287 * BIO_do_connect and BIO_do_handshake have been extended:
288 If domain name resolution yields multiple IP addresses all of them are tried
289 after connect() failures.
293 * All of the low level RSA functions have been deprecated including:
295 RSA_new_method, RSA_size, RSA_security_bits, RSA_get0_pss_params,
296 RSA_get_version, RSA_get0_engine, RSA_generate_key_ex,
297 RSA_generate_multi_prime_key, RSA_X931_derive_ex, RSA_X931_generate_key_ex,
298 RSA_check_key, RSA_check_key_ex, RSA_public_encrypt, RSA_private_encrypt,
299 RSA_public_decrypt, RSA_private_decrypt, RSA_set_default_method,
300 RSA_get_default_method, RSA_null_method, RSA_get_method, RSA_set_method,
301 RSA_PKCS1_OpenSSL, RSA_print_fp, RSA_print, RSA_sign, RSA_verify,
302 RSA_sign_ASN1_OCTET_STRING, RSA_verify_ASN1_OCTET_STRING, RSA_blinding_on,
303 RSA_blinding_off, RSA_setup_blinding, RSA_padding_add_PKCS1_type_1,
304 RSA_padding_check_PKCS1_type_1, RSA_padding_add_PKCS1_type_2,
305 RSA_padding_check_PKCS1_type_2, PKCS1_MGF1, RSA_padding_add_PKCS1_OAEP,
306 RSA_padding_check_PKCS1_OAEP, RSA_padding_add_PKCS1_OAEP_mgf1,
307 RSA_padding_check_PKCS1_OAEP_mgf1, RSA_padding_add_SSLv23,
308 RSA_padding_check_SSLv23, RSA_padding_add_none, RSA_padding_check_none,
309 RSA_padding_add_X931, RSA_padding_check_X931, RSA_X931_hash_id,
310 RSA_verify_PKCS1_PSS, RSA_padding_add_PKCS1_PSS, RSA_verify_PKCS1_PSS_mgf1,
311 RSA_padding_add_PKCS1_PSS_mgf1, RSA_set_ex_data, RSA_get_ex_data,
312 RSA_meth_new, RSA_meth_free, RSA_meth_dup, RSA_meth_get0_name,
313 RSA_meth_set1_name, RSA_meth_get_flags, RSA_meth_set_flags,
314 RSA_meth_get0_app_data, RSA_meth_set0_app_data, RSA_meth_get_pub_enc,
315 RSA_meth_set_pub_enc, RSA_meth_get_pub_dec, RSA_meth_set_pub_dec,
316 RSA_meth_get_priv_enc, RSA_meth_set_priv_enc, RSA_meth_get_priv_dec,
317 RSA_meth_set_priv_dec, RSA_meth_get_mod_exp, RSA_meth_set_mod_exp,
318 RSA_meth_get_bn_mod_exp, RSA_meth_set_bn_mod_exp, RSA_meth_get_init,
319 RSA_meth_set_init, RSA_meth_get_finish, RSA_meth_set_finish,
320 RSA_meth_get_sign, RSA_meth_set_sign, RSA_meth_get_verify,
321 RSA_meth_set_verify, RSA_meth_get_keygen, RSA_meth_set_keygen,
322 RSA_meth_get_multi_prime_keygen and RSA_meth_set_multi_prime_keygen.
324 Use of these low level functions has been informally discouraged for a long
325 time. Instead applications should use L<EVP_PKEY_encrypt_init(3)>,
326 L<EVP_PKEY_encrypt(3)>, L<EVP_PKEY_decrypt_init(3)> and
327 L<EVP_PKEY_decrypt(3)>.
331 * X509 certificates signed using SHA1 are no longer allowed at security
333 In TLS/SSL the default security level is 1. It can be set either
334 using the cipher string with @SECLEVEL, or calling
335 SSL_CTX_set_security_level(). If the leaf certificate is signed with SHA-1,
336 a call to SSL_CTX_use_certificate() will fail if the security level is not
338 Outside TLS/SSL, the default security level is -1 (effectively 0). It can
339 be set using X509_VERIFY_PARAM_set_auth_level() or using the -auth_level
344 * The command line utilities dhparam, dsa, gendsa and dsaparam have been
345 modified to use PKEY APIs. These commands are now in maintenance mode
346 and no new features will be added to them.
350 * The command line utility rsautl has been deprecated.
351 Instead use the pkeyutl program.
355 * The command line utilities genrsa and rsa have been modified to use PKEY
356 APIs. They now write PKCS#8 keys by default. These commands are now in
357 maintenance mode and no new features will be added to them.
361 * All of the low level DH functions have been deprecated including:
363 DH_OpenSSL, DH_set_default_method, DH_get_default_method, DH_set_method,
364 DH_new_method, DH_size, DH_security_bits, DH_get_ex_new_index,
365 DH_set_ex_data, DH_get_ex_data, DH_generate_parameters_ex,
366 DH_check_params_ex, DH_check_ex, DH_check_pub_key_ex,
367 DH_check, DH_check_pub_key, DH_generate_key, DH_compute_key,
368 DH_compute_key_padded, DHparams_print_fp, DHparams_print, DH_get_nid,
369 DH_KDF_X9_42, DH_get0_engine, DH_meth_new, DH_meth_free, DH_meth_dup,
370 DH_meth_get0_name, DH_meth_set1_name, DH_meth_get_flags, DH_meth_set_flags,
371 DH_meth_get0_app_data, DH_meth_set0_app_data, DH_meth_get_generate_key,
372 DH_meth_set_generate_key, DH_meth_get_compute_key, DH_meth_set_compute_key,
373 DH_meth_get_bn_mod_exp, DH_meth_set_bn_mod_exp, DH_meth_get_init,
374 DH_meth_set_init, DH_meth_get_finish, DH_meth_set_finish,
375 DH_meth_get_generate_params and DH_meth_set_generate_params.
377 Use of these low level functions has been informally discouraged for a long
378 time. Instead applications should use L<EVP_PKEY_derive_init(3)>
379 and L<EVP_PKEY_derive(3)>.
383 * All of the low level DSA functions have been deprecated including:
385 DSA_do_sign, DSA_do_verify, DSA_OpenSSL, DSA_set_default_method,
386 DSA_get_default_method, DSA_set_method, DSA_get_method,
387 DSA_new_method, DSA_size, DSA_security_bits, DSA_sign_setup, DSA_sign,
388 DSA_verify, DSA_get_ex_new_index, DSA_set_ex_data, DSA_get_ex_data,
389 DSA_generate_parameters_ex, DSA_generate_key, DSA_meth_new, DSA_get0_engine,
390 DSA_meth_free, DSA_meth_dup, DSA_meth_get0_name, DSA_meth_set1_name,
391 DSA_meth_get_flags, DSA_meth_set_flags, DSA_meth_get0_app_data,
392 DSA_meth_set0_app_data, DSA_meth_get_sign, DSA_meth_set_sign,
393 DSA_meth_get_sign_setup, DSA_meth_set_sign_setup, DSA_meth_get_verify,
394 DSA_meth_set_verify, DSA_meth_get_mod_exp, DSA_meth_set_mod_exp,
395 DSA_meth_get_bn_mod_exp, DSA_meth_set_bn_mod_exp, DSA_meth_get_init,
396 DSA_meth_set_init, DSA_meth_get_finish, DSA_meth_set_finish,
397 DSA_meth_get_paramgen, DSA_meth_set_paramgen, DSA_meth_get_keygen and
400 Use of these low level functions has been informally discouraged for a long
401 time. Instead applications should use L<EVP_DigestSignInit_ex(3)>,
402 L<EVP_DigestSignUpdate(3)> and L<EVP_DigestSignFinal(3)>.
406 * Reworked the treatment of EC EVP_PKEYs with the SM2 curve to
407 automatically become EVP_PKEY_SM2 rather than EVP_PKEY_EC.
408 This means that applications don't have to look at the curve NID and
409 `EVP_PKEY_set_alias_type(pkey, EVP_PKEY_SM2)` to get SM2 computations.
410 However, they still can, that EVP_PKEY_set_alias_type() call acts as
411 a no-op when the EVP_PKEY is already of the given type.
413 Parameter and key generation is also reworked to make it possible
414 to generate EVP_PKEY_SM2 parameters and keys without having to go
415 through EVP_PKEY_EC generation and then change the EVP_PKEY type.
416 However, code that does the latter will still work as before.
420 * Deprecated low level ECDH and ECDSA functions. These include:
422 ECDH_compute_key, ECDSA_do_sign, ECDSA_do_sign_ex, ECDSA_do_verify,
423 ECDSA_sign_setup, ECDSA_sign, ECDSA_sign_ex, ECDSA_verify and
426 Use of these low level functions has been informally discouraged for a long
427 time. Instead applications should use the EVP_PKEY_derive(3),
428 EVP_DigestSign(3) and EVP_DigestVerify(3) functions.
432 * Deprecated the EC_KEY_METHOD functions. These include:
434 EC_KEY_METHOD_new, EC_KEY_METHOD_free, EC_KEY_METHOD_set_init,
435 EC_KEY_METHOD_set_keygen, EC_KEY_METHOD_set_compute_key,
436 EC_KEY_METHOD_set_sign, EC_KEY_METHOD_set_verify,
437 EC_KEY_METHOD_get_init, EC_KEY_METHOD_get_keygen,
438 EC_KEY_METHOD_get_compute_key, EC_KEY_METHOD_get_sign and
439 EC_KEY_METHOD_get_verify.
441 Instead applications and extension writers should use the OSSL_PROVIDER APIs.
445 * Deprecated EVP_PKEY_decrypt_old(), please use EVP_PKEY_decrypt_init()
446 and EVP_PKEY_decrypt() instead.
447 Deprecated EVP_PKEY_encrypt_old(), please use EVP_PKEY_encrypt_init()
448 and EVP_PKEY_encrypt() instead.
452 * Enhanced the documentation of EVP_PKEY_size(), EVP_PKEY_bits()
453 and EVP_PKEY_security_bits(). Especially EVP_PKEY_size() needed
454 a new formulation to include all the things it can be used for,
455 as well as words of caution.
459 * The SSL_CTX_set_tlsext_ticket_key_cb(3) function has been deprecated.
460 Instead used the new SSL_CTX_set_tlsext_ticket_key_evp_cb(3) function.
464 * All of the low level HMAC functions have been deprecated including:
466 HMAC, HMAC_size, HMAC_CTX_new, HMAC_CTX_reset, HMAC_CTX_free,
467 HMAC_Init_ex, HMAC_Update, HMAC_Final, HMAC_CTX_copy, HMAC_CTX_set_flags
470 Use of these low level functions has been informally discouraged for a long
471 time. Instead applications should use L<EVP_MAC_CTX_new(3)>,
472 L<EVP_MAC_CTX_free(3)>, L<EVP_MAC_init(3)>, L<EVP_MAC_update(3)>
473 and L<EVP_MAC_final(3)>.
477 * Over two thousand fixes were made to the documentation, including:
478 - Common options (such as -rand/-writerand, TLS version control, etc)
479 were refactored and point to newly-enhanced descriptions in openssl.pod.
480 - Added style conformance for all options (with help from Richard Levitte),
481 documented all reported missing options, added a CI build to check
482 that all options are documented and that no unimplemented options
484 - Documented some internals, such as all use of environment variables.
485 - Addressed all internal broken L<> references.
489 * All of the low level CMAC functions have been deprecated including:
491 CMAC_CTX_new, CMAC_CTX_cleanup, CMAC_CTX_free, CMAC_CTX_get0_cipher_ctx,
492 CMAC_CTX_copy, CMAC_Init, CMAC_Update, CMAC_Final and CMAC_resume.
494 Use of these low level functions has been informally discouraged for a long
495 time. Instead applications should use L<EVP_MAC_CTX_new(3)>,
496 L<EVP_MAC_CTX_free(3)>, L<EVP_MAC_init(3)>, L<EVP_MAC_update(3)>
497 and L<EVP_MAC_final(3)>.
501 * All of the low level MD2, MD4, MD5, MDC2, RIPEMD160, SHA1, SHA224, SHA256,
502 SHA384, SHA512 and Whirlpool digest functions have been deprecated.
505 MD2, MD2_options, MD2_Init, MD2_Update, MD2_Final, MD4, MD4_Init,
506 MD4_Update, MD4_Final, MD4_Transform, MD5, MD5_Init, MD5_Update,
507 MD5_Final, MD5_Transform, MDC2, MDC2_Init, MDC2_Update, MDC2_Final,
508 RIPEMD160, RIPEMD160_Init, RIPEMD160_Update, RIPEMD160_Final,
509 RIPEMD160_Transform, SHA1_Init, SHA1_Update, SHA1_Final, SHA1_Transform,
510 SHA224_Init, SHA224_Update, SHA224_Final, SHA224_Transform, SHA256_Init,
511 SHA256_Update, SHA256_Final, SHA256_Transform, SHA384, SHA384_Init,
512 SHA384_Update, SHA384_Final, SHA512, SHA512_Init, SHA512_Update,
513 SHA512_Final, SHA512_Transform, WHIRLPOOL, WHIRLPOOL_Init,
514 WHIRLPOOL_Update, WHIRLPOOL_BitUpdate and WHIRLPOOL_Final.
516 Use of these low level functions has been informally discouraged
517 for a long time. Applications should use the EVP_DigestInit_ex(3),
518 EVP_DigestUpdate(3) and EVP_DigestFinal_ex(3) functions instead.
522 * Corrected the documentation of the return values from the `EVP_DigestSign*`
523 set of functions. The documentation mentioned negative values for some
524 errors, but this was never the case, so the mention of negative values
527 Code that followed the documentation and thereby check with something
528 like `EVP_DigestSignInit(...) <= 0` will continue to work undisturbed.
532 * All of the low level cipher functions have been deprecated including:
534 AES_options, AES_set_encrypt_key, AES_set_decrypt_key, AES_encrypt,
535 AES_decrypt, AES_ecb_encrypt, AES_cbc_encrypt, AES_cfb128_encrypt,
536 AES_cfb1_encrypt, AES_cfb8_encrypt, AES_ofb128_encrypt,
537 AES_wrap_key, AES_unwrap_key, BF_set_key, BF_encrypt, BF_decrypt,
538 BF_ecb_encrypt, BF_cbc_encrypt, BF_cfb64_encrypt, BF_ofb64_encrypt,
539 BF_options, Camellia_set_key, Camellia_encrypt, Camellia_decrypt,
540 Camellia_ecb_encrypt, Camellia_cbc_encrypt, Camellia_cfb128_encrypt,
541 Camellia_cfb1_encrypt, Camellia_cfb8_encrypt, Camellia_ofb128_encrypt,
542 Camellia_ctr128_encrypt, CAST_set_key, CAST_encrypt, CAST_decrypt,
543 CAST_ecb_encrypt, CAST_cbc_encrypt, CAST_cfb64_encrypt,
544 CAST_ofb64_encrypt, DES_options, DES_encrypt1, DES_encrypt2,
545 DES_encrypt3, DES_decrypt3, DES_cbc_encrypt, DES_ncbc_encrypt,
546 DES_pcbc_encrypt, DES_xcbc_encrypt, DES_cfb_encrypt, DES_cfb64_encrypt,
547 DES_ecb_encrypt, DES_ofb_encrypt, DES_ofb64_encrypt, DES_random_key,
548 DES_set_odd_parity, DES_check_key_parity, DES_is_weak_key, DES_set_key,
549 DES_key_sched, DES_set_key_checked, DES_set_key_unchecked,
550 DES_string_to_key, DES_string_to_2keys, DES_fixup_key_parity,
551 DES_ecb2_encrypt, DES_ede2_cbc_encrypt, DES_ede2_cfb64_encrypt,
552 DES_ede2_ofb64_encrypt, DES_ecb3_encrypt, DES_ede3_cbc_encrypt,
553 DES_ede3_cfb64_encrypt, DES_ede3_cfb_encrypt, DES_ede3_ofb64_encrypt,
554 DES_cbc_cksum, DES_quad_cksum, IDEA_encrypt, IDEA_options,
555 IDEA_ecb_encrypt, IDEA_set_encrypt_key, IDEA_set_decrypt_key,
556 IDEA_cbc_encrypt, IDEA_cfb64_encrypt, IDEA_ofb64_encrypt, RC2_set_key,
557 RC2_encrypt, RC2_decrypt, RC2_ecb_encrypt, RC2_cbc_encrypt,
558 RC2_cfb64_encrypt, RC2_ofb64_encrypt, RC4, RC4_options, RC4_set_key,
559 RC5_32_set_key, RC5_32_encrypt, RC5_32_decrypt, RC5_32_ecb_encrypt,
560 RC5_32_cbc_encrypt, RC5_32_cfb64_encrypt, RC5_32_ofb64_encrypt,
561 SEED_set_key, SEED_encrypt, SEED_decrypt, SEED_ecb_encrypt,
562 SEED_cbc_encrypt, SEED_cfb128_encrypt and SEED_ofb128_encrypt.
564 Use of these low level functions has been informally discouraged for
565 a long time. Applications should use the high level EVP APIs, e.g.
566 EVP_EncryptInit_ex, EVP_EncryptUpdate, EVP_EncryptFinal_ex, and the
567 equivalently named decrypt functions instead.
569 *Matt Caswell and Paul Dale*
571 * Removed include/openssl/opensslconf.h.in and replaced it with
572 include/openssl/configuration.h.in, which differs in not including
573 <openssl/macros.h>. A short header include/openssl/opensslconf.h
574 was added to include both.
576 This allows internal hacks where one might need to modify the set
577 of configured macros, for example this if deprecated symbols are
578 still supposed to be available internally:
580 #include <openssl/configuration.h>
582 #undef OPENSSL_NO_DEPRECATED
583 #define OPENSSL_SUPPRESS_DEPRECATED
585 #include <openssl/macros.h>
587 This should not be used by applications that use the exported
588 symbols, as that will lead to linking errors.
592 * Fixed an overflow bug in the x64_64 Montgomery squaring procedure
593 used in exponentiation with 512-bit moduli. No EC algorithms are
594 affected. Analysis suggests that attacks against 2-prime RSA1024,
595 3-prime RSA1536, and DSA1024 as a result of this defect would be very
596 difficult to perform and are not believed likely. Attacks against DH512
597 are considered just feasible. However, for an attack the target would
598 have to re-use the DH512 private key, which is not recommended anyway.
599 Also applications directly using the low level API BN_mod_exp may be
600 affected if they use BN_FLG_CONSTTIME.
605 * Most memory-debug features have been deprecated, and the functionality
606 replaced with no-ops.
610 * Added documentation for the STACK API. OpenSSL only defines the STACK
611 functions where they are used.
615 * Introduced a new method type and API, OSSL_ENCODER, to
616 represent generic encoders. An implementation is expected to
617 be able to encode an object associated with a given name (such
618 as an algorithm name for an asymmetric key) into forms given by
619 implementation properties.
621 Encoders are primarily used from inside libcrypto, through
622 calls to functions like EVP_PKEY_print_private(),
623 PEM_write_bio_PrivateKey() and similar.
625 Encoders are specified in such a way that they can be made to
626 directly handle the provider side portion of an object, if this
627 provider side part comes from the same provider as the encoder
628 itself, but can also be made to handle objects in parametrized
629 form (as an OSSL_PARAM array of data). This allows a provider to
630 offer generic encoders as a service for any other provider.
634 * Added a .pragma directive to the syntax of configuration files, to
635 allow varying behavior in a supported and predictable manner.
636 Currently added pragma:
640 This allows dollar signs to be a keyword character unless it's
641 followed by a opening brace or parenthesis. This is useful for
642 platforms where dollar signs are commonly used in names, such as
643 volume names and system directory names on VMS.
647 * Added functionality to create an EVP_PKEY from user data. This
648 is effectively the same as creating a RSA, DH or DSA object and
649 then assigning them to an EVP_PKEY, but directly using algorithm
650 agnostic EVP functions. A benefit is that this should be future
651 proof for public key algorithms to come.
655 * Change the interpretation of the '--api' configuration option to
656 mean that this is a desired API compatibility level with no
657 further meaning. The previous interpretation, that this would
658 also mean to remove all deprecated symbols up to and including
659 the given version, no requires that 'no-deprecated' is also used
660 in the configuration.
662 When building applications, the desired API compatibility level
663 can be set with the OPENSSL_API_COMPAT macro like before. For
664 API compatibility version below 3.0, the old style numerical
665 value is valid as before, such as -DOPENSSL_API_COMPAT=0x10100000L.
666 For version 3.0 and on, the value is expected to be the decimal
667 value calculated from the major and minor version like this:
669 MAJOR * 10000 + MINOR * 100
673 -DOPENSSL_API_COMPAT=30000 For 3.0
674 -DOPENSSL_API_COMPAT=30200 For 3.2
676 To hide declarations that are deprecated up to and including the
677 given API compatibility level, -DOPENSSL_NO_DEPRECATED must be
678 given when building the application as well.
682 * Added the X509_LOOKUP_METHOD called X509_LOOKUP_store, to allow
683 access to certificate and CRL stores via URIs and OSSL_STORE
686 This adds the following functions:
688 - X509_LOOKUP_store()
689 - X509_STORE_load_file()
690 - X509_STORE_load_path()
691 - X509_STORE_load_store()
692 - SSL_add_store_cert_subjects_to_stack()
693 - SSL_CTX_set_default_verify_store()
694 - SSL_CTX_load_verify_file()
695 - SSL_CTX_load_verify_dir()
696 - SSL_CTX_load_verify_store()
700 * Added a new method to gather entropy on VMS, based on SYS$GET_ENTROPY.
701 The presence of this system service is determined at run-time.
705 * Added functionality to create an EVP_PKEY context based on data
706 for methods from providers. This takes an algorithm name and a
707 property query string and simply stores them, with the intent
708 that any operation that uses this context will use those strings
709 to fetch the needed methods implicitly, thereby making the port
710 of application written for pre-3.0 OpenSSL easier.
714 * The undocumented function NCONF_WIN32() has been deprecated; for
715 conversion details see the HISTORY section of doc/man5/config.pod
719 * Introduced the new functions EVP_DigestSignInit_ex() and
720 EVP_DigestVerifyInit_ex(). The macros EVP_DigestSignUpdate() and
721 EVP_DigestVerifyUpdate() have been converted to functions. See the man
722 pages for further details.
726 * Over two thousand fixes were made to the documentation, including:
727 adding missing command flags, better style conformance, documentation
730 *Rich Salz, Richard Levitte*
732 * s390x assembly pack: add hardware-support for P-256, P-384, P-521,
733 X25519, X448, Ed25519 and Ed448.
737 * Print all values for a PKCS#12 attribute with 'openssl pkcs12', not just
742 * Deprecated the public definition of ERR_STATE as well as the function
743 ERR_get_state(). This is done in preparation of making ERR_STATE an
748 * Added ERR functionality to give callers access to the stored function
749 names that have replaced the older function code based functions.
751 New functions are ERR_get_error_func(), ERR_peek_error_func(),
752 ERR_peek_last_error_func(), ERR_get_error_data(), ERR_peek_error_data(),
753 ERR_peek_last_error_data(), ERR_get_error_all(), ERR_peek_error_all()
754 and ERR_peek_last_error_all().
756 These functions have become deprecated: ERR_get_error_line_data(),
757 ERR_peek_error_line_data(), ERR_peek_last_error_line_data() and
758 ERR_func_error_string().
762 * Extended testing to be verbose for failing tests only. The make variables
763 VERBOSE_FAILURE or VF can be used to enable this:
765 $ make VF=1 test # Unix
766 $ mms /macro=(VF=1) test ! OpenVMS
767 $ nmake VF=1 test # Windows
771 * For built-in EC curves, ensure an EC_GROUP built from the curve name is
772 used even when parsing explicit parameters, when loading a encoded key
773 or calling `EC_GROUP_new_from_ecpkparameters()`/
774 `EC_GROUP_new_from_ecparameters()`.
775 This prevents bypass of security hardening and performance gains,
776 especially for curves with specialized EC_METHODs.
777 By default, if a key encoded with explicit parameters is loaded and later
778 encoded, the output is still encoded with explicit parameters, even if
779 internally a "named" EC_GROUP is used for computation.
783 * Compute ECC cofactors if not provided during EC_GROUP construction. Before
784 this change, EC_GROUP_set_generator would accept order and/or cofactor as
785 NULL. After this change, only the cofactor parameter can be NULL. It also
786 does some minimal sanity checks on the passed order.
791 * Fixed a padding oracle in PKCS7_dataDecode and CMS_decrypt_set1_pkey.
792 An attack is simple, if the first CMS_recipientInfo is valid but the
793 second CMS_recipientInfo is chosen ciphertext. If the second
794 recipientInfo decodes to PKCS #1 v1.5 form plaintext, the correct
795 encryption key will be replaced by garbage, and the message cannot be
796 decoded, but if the RSA decryption fails, the correct encryption key is
797 used and the recipient will not notice the attack.
798 As a work around for this potential attack the length of the decrypted
799 key must be equal to the cipher default key length, in case the
800 certifiate is not given and all recipientInfo are tried out.
801 The old behaviour can be re-enabled in the CMS code by setting the
802 CMS_DEBUG_DECRYPT flag.
806 * Early start up entropy quality from the DEVRANDOM seed source has been
807 improved for older Linux systems. The RAND subsystem will wait for
808 /dev/random to be producing output before seeding from /dev/urandom.
809 The seeded state is stored for future library initialisations using
810 a system global shared memory segment. The shared memory identifier
811 can be configured by defining OPENSSL_RAND_SEED_DEVRANDOM_SHM_ID to
812 the desired value. The default identifier is 114.
816 * Revised BN_generate_prime_ex to not avoid factors 2..17863 in p-1
817 when primes for RSA keys are computed.
818 Since we previously always generated primes == 2 (mod 3) for RSA keys,
819 the 2-prime and 3-prime RSA modules were easy to distinguish, since
820 `N = p*q = 1 (mod 3)`, but `N = p*q*r = 2 (mod 3)`. Therefore fingerprinting
821 2-prime vs. 3-prime RSA keys was possible by computing N mod 3.
822 This avoids possible fingerprinting of newly generated RSA modules.
826 * Correct the extended master secret constant on EBCDIC systems. Without this
827 fix TLS connections between an EBCDIC system and a non-EBCDIC system that
828 negotiate EMS will fail. Unfortunately this also means that TLS connections
829 between EBCDIC systems with this fix, and EBCDIC systems without this
830 fix will fail if they negotiate EMS.
834 * Changed the library initialisation so that the config file is now loaded
835 by default. This was already the case for libssl. It now occurs for both
836 libcrypto and libssl. Use the OPENSSL_INIT_NO_LOAD_CONFIG option to
837 OPENSSL_init_crypto() to suppress automatic loading of a config file.
841 * Introduced new error raising macros, ERR_raise() and ERR_raise_data(),
842 where the former acts as a replacement for ERR_put_error(), and the
843 latter replaces the combination ERR_put_error()+ERR_add_error_data().
844 ERR_raise_data() adds more flexibility by taking a format string and
845 an arbitrary number of arguments following it, to be processed with
850 * Introduced a new function, OSSL_PROVIDER_available(), which can be used
851 to check if a named provider is loaded and available. When called, it
852 will also activate all fallback providers if such are still present.
856 * Enforce a minimum DH modulus size of 512 bits.
860 * Changed DH parameters to generate the order q subgroup instead of 2q.
861 Previously generated DH parameters are still accepted by DH_check
862 but DH_generate_key works around that by clearing bit 0 of the
863 private key for those. This avoids leaking bit 0 of the private key.
867 * Significantly reduce secure memory usage by the randomness pools.
871 * `{CRYPTO,OPENSSL}_mem_debug_{push,pop}` are now no-ops and have been
876 * A new type, EVP_KEYEXCH, has been introduced to represent key exchange
877 algorithms. An implementation of a key exchange algorithm can be obtained
878 by using the function EVP_KEYEXCH_fetch(). An EVP_KEYEXCH algorithm can be
879 used in a call to EVP_PKEY_derive_init_ex() which works in a similar way to
880 the older EVP_PKEY_derive_init() function. See the man pages for the new
881 functions for further details.
885 * The EVP_PKEY_CTX_set_dh_pad() macro has now been converted to a function.
889 * Removed the function names from error messages and deprecated the
892 * Removed NextStep support and the macro OPENSSL_UNISTD
896 * Removed DES_check_key. Also removed OPENSSL_IMPLEMENT_GLOBAL,
897 OPENSSL_GLOBAL_REF, OPENSSL_DECLARE_GLOBAL.
898 Also removed "export var as function" capability; we do not export
899 variables, only functions.
903 * RC5_32_set_key has been changed to return an int type, with 0 indicating
904 an error and 1 indicating success. In previous versions of OpenSSL this
905 was a void type. If a key was set longer than the maximum possible this
910 * Support SM2 signing and verification schemes with X509 certificate.
914 * Use SHA256 as the default digest for TS query in the ts app.
918 * Change PBKDF2 to conform to SP800-132 instead of the older PKCS5 RFC2898.
919 This checks that the salt length is at least 128 bits, the derived key
920 length is at least 112 bits, and that the iteration count is at least 1000.
921 For backwards compatibility these checks are disabled by default in the
922 default provider, but are enabled by default in the fips provider.
923 To enable or disable these checks use the control
924 EVP_KDF_CTRL_SET_PBKDF2_PKCS5_MODE.
928 * Default cipher lists/suites are now available via a function, the
929 #defines are deprecated.
933 * Add target VC-WIN32-UWP, VC-WIN64A-UWP, VC-WIN32-ARM-UWP and
934 VC-WIN64-ARM-UWP in Windows OneCore target for making building libraries
935 for Windows Store apps easier. Also, the "no-uplink" option has been added.
939 * Join the directories crypto/x509 and crypto/x509v3
943 * Change the default RSA, DSA and DH size to 2048 bit instead of 1024.
944 This changes the size when using the genpkey app when no size is given. It
945 fixes an omission in earlier changes that changed all RSA, DSA and DH
946 generation apps to use 2048 bits by default.
950 * Added command 'openssl kdf' that uses the EVP_KDF API.
954 * Added command 'openssl mac' that uses the EVP_MAC API.
958 * Added OPENSSL_info() to get diverse built-in OpenSSL data, such
959 as default directories. Also added the command 'openssl info'
960 for scripting purposes.
964 * The functions AES_ige_encrypt() and AES_bi_ige_encrypt() have been
965 deprecated. These undocumented functions were never integrated into the EVP
966 layer and implement the AES Infinite Garble Extension (IGE) mode and AES
967 Bi-directional IGE mode. These modes were never formally standardised and
968 usage of these functions is believed to be very small. In particular
969 AES_bi_ige_encrypt() has a known bug. It accepts 2 AES keys, but only one
970 is ever used. The security implications are believed to be minimal, but
971 this issue was never fixed for backwards compatibility reasons. New code
972 should not use these modes.
976 * Add prediction resistance to the DRBG reseeding process.
980 * Limit the number of blocks in a data unit for AES-XTS to 2^20 as
981 mandated by IEEE Std 1619-2018.
985 * Added newline escaping functionality to a filename when using openssl dgst.
986 This output format is to replicate the output format found in the `*sum`
987 checksum programs. This aims to preserve backward compatibility.
989 *Matt Eaton, Richard Levitte, and Paul Dale*
991 * Removed the heartbeat message in DTLS feature, as it has very
992 little usage and doesn't seem to fulfill a valuable purpose.
993 The configuration option is now deprecated.
997 * Changed the output of 'openssl {digestname} < file' to display the
998 digest name in its output.
1002 * Added a new generic trace API which provides support for enabling
1003 instrumentation through trace output. This feature is mainly intended
1004 as an aid for developers and is disabled by default. To utilize it,
1005 OpenSSL needs to be configured with the `enable-trace` option.
1007 If the tracing API is enabled, the application can activate trace output
1008 by registering BIOs as trace channels for a number of tracing and debugging
1011 The 'openssl' application has been expanded to enable any of the types
1012 available via environment variables defined by the user, and serves as
1013 one possible example on how to use this functionality.
1015 *Richard Levitte & Matthias St. Pierre*
1017 * Added build tests for C++. These are generated files that only do one
1018 thing, to include one public OpenSSL head file each. This tests that
1019 the public header files can be usefully included in a C++ application.
1021 This test isn't enabled by default. It can be enabled with the option
1022 'enable-buildtest-c++'.
1026 * Add Single Step KDF (EVP_KDF_SS) to EVP_KDF.
1030 * Add KMAC to EVP_MAC.
1034 * Added property based algorithm implementation selection framework to
1039 * Added SCA hardening for modular field inversion in EC_GROUP through
1040 a new dedicated field_inv() pointer in EC_METHOD.
1041 This also addresses a leakage affecting conversions from projective
1042 to affine coordinates.
1044 *Billy Bob Brumley, Nicola Tuveri*
1046 * Added EVP_KDF, an EVP layer KDF API, to simplify adding KDF and PRF
1047 implementations. This includes an EVP_PKEY to EVP_KDF bridge for
1048 those algorithms that were already supported through the EVP_PKEY API
1049 (scrypt, TLS1 PRF and HKDF). The low-level KDF functions for PBKDF2
1050 and scrypt are now wrappers that call EVP_KDF.
1054 * Build devcrypto engine as a dynamic engine.
1056 *Eneas U de Queiroz*
1058 * Add keyed BLAKE2 to EVP_MAC.
1062 * Fix a bug in the computation of the endpoint-pair shared secret used
1063 by DTLS over SCTP. This breaks interoperability with older versions
1064 of OpenSSL like OpenSSL 1.1.0 and OpenSSL 1.0.2. There is a runtime
1065 switch SSL_MODE_DTLS_SCTP_LABEL_LENGTH_BUG (off by default) enabling
1066 interoperability with such broken implementations. However, enabling
1067 this switch breaks interoperability with correct implementations.
1069 * Fix a use after free bug in d2i_X509_PUBKEY when overwriting a
1070 re-used X509_PUBKEY object if the second PUBKEY is malformed.
1074 * Move strictness check from EVP_PKEY_asn1_new() to EVP_PKEY_asn1_add0().
1078 * Change the license to the Apache License v2.0.
1082 * Switch to a new version scheme using three numbers MAJOR.MINOR.PATCH.
1084 - Major releases (indicated by incrementing the MAJOR release number)
1085 may introduce incompatible API/ABI changes.
1086 - Minor releases (indicated by incrementing the MINOR release number)
1087 may introduce new features but retain API/ABI compatibility.
1088 - Patch releases (indicated by incrementing the PATCH number)
1089 are intended for bug fixes and other improvements of existing
1090 features only (like improving performance or adding documentation)
1091 and retain API/ABI compatibility.
1095 * Add support for RFC5297 SIV mode (siv128), including AES-SIV.
1099 * Remove the 'dist' target and add a tarball building script. The
1100 'dist' target has fallen out of use, and it shouldn't be
1101 necessary to configure just to create a source distribution.
1105 * Recreate the OS390-Unix config target. It no longer relies on a
1106 special script like it did for OpenSSL pre-1.1.0.
1110 * Instead of having the source directories listed in Configure, add
1111 a 'build.info' keyword SUBDIRS to indicate what sub-directories to
1116 * Add GMAC to EVP_MAC.
1120 * Ported the HMAC, CMAC and SipHash EVP_PKEY_METHODs to EVP_MAC.
1124 * Added EVP_MAC, an EVP layer MAC API, to simplify adding MAC
1125 implementations. This includes a generic EVP_PKEY to EVP_MAC bridge,
1126 to facilitate the continued use of MACs through raw private keys in
1127 functionality such as `EVP_DigestSign*` and `EVP_DigestVerify*`.
1131 * Deprecate ECDH_KDF_X9_62() and mark its replacement as internal. Users
1132 should use the EVP interface instead (EVP_PKEY_CTX_set_ecdh_kdf_type).
1136 * Added EVP_PKEY_ECDH_KDF_X9_63 and ecdh_KDF_X9_63() as replacements for
1137 the EVP_PKEY_ECDH_KDF_X9_62 KDF type and ECDH_KDF_X9_62(). The old names
1138 are retained for backwards compatibility.
1142 * AES-XTS mode now enforces that its two keys are different to mitigate
1143 the attacked described in "Efficient Instantiations of Tweakable
1144 Blockciphers and Refinements to Modes OCB and PMAC" by Phillip Rogaway.
1145 Details of this attack can be obtained from:
1146 <http://web.cs.ucdavis.edu/%7Erogaway/papers/offsets.pdf>
1150 * Rename the object files, i.e. give them other names than in previous
1151 versions. Their names now include the name of the final product, as
1152 well as its type mnemonic (bin, lib, shlib).
1156 * Added new option for 'openssl list', '-objects', which will display the
1157 list of built in objects, i.e. OIDs with names.
1161 * Added the options `-crl_lastupdate` and `-crl_nextupdate` to `openssl ca`,
1162 allowing the `lastUpdate` and `nextUpdate` fields in the generated CRL to
1167 * Added support for Linux Kernel TLS data-path. The Linux Kernel data-path
1168 improves application performance by removing data copies and providing
1169 applications with zero-copy system calls such as sendfile and splice.
1173 * The SSL option SSL_OP_CLEANSE_PLAINTEXT is introduced. If that
1174 option is set, openssl cleanses (zeroize) plaintext bytes from
1175 internal buffers after delivering them to the application. Note,
1176 the application is still responsible for cleansing other copies
1177 (e.g.: data received by SSL_read(3)).
1181 * `PKCS12_parse` now maintains the order of the parsed certificates
1182 when outputting them via `*ca` (rather than reversing it).
1189 ### Changes between 1.1.1e and 1.1.1f [xx XXX xxxx]
1191 ### Changes between 1.1.1d and 1.1.1e [17 Mar 2020]
1193 * Properly detect EOF while reading in libssl. Previously if we hit an EOF
1194 while reading in libssl then we would report an error back to the
1195 application (SSL_ERROR_SYSCALL) but errno would be 0. We now add
1196 an error to the stack (which means we instead return SSL_ERROR_SSL) and
1197 therefore give a hint as to what went wrong.
1201 * Check that ed25519 and ed448 are allowed by the security level. Previously
1202 signature algorithms not using an MD were not being checked that they were
1203 allowed by the security level.
1207 * Fixed SSL_get_servername() behaviour. The behaviour of SSL_get_servername()
1208 was not quite right. The behaviour was not consistent between resumption
1209 and normal handshakes, and also not quite consistent with historical
1210 behaviour. The behaviour in various scenarios has been clarified and
1211 it has been updated to make it match historical behaviour as closely as
1216 * *[VMS only]* The header files that the VMS compilers include automatically,
1217 `__DECC_INCLUDE_PROLOGUE.H` and `__DECC_INCLUDE_EPILOGUE.H`, use pragmas
1218 that the C++ compiler doesn't understand. This is a shortcoming in the
1219 compiler, but can be worked around with `__cplusplus` guards.
1221 C++ applications that use OpenSSL libraries must be compiled using the
1222 qualifier `/NAMES=(AS_IS,SHORTENED)` to be able to use all the OpenSSL
1223 functions. Otherwise, only functions with symbols of less than 31
1224 characters can be used, as the linker will not be able to successfully
1225 resolve symbols with longer names.
1229 * Added a new method to gather entropy on VMS, based on SYS$GET_ENTROPY.
1230 The presence of this system service is determined at run-time.
1234 * Added newline escaping functionality to a filename when using openssl dgst.
1235 This output format is to replicate the output format found in the `*sum`
1236 checksum programs. This aims to preserve backward compatibility.
1238 *Matt Eaton, Richard Levitte, and Paul Dale*
1240 * Print all values for a PKCS#12 attribute with 'openssl pkcs12', not just
1245 ### Changes between 1.1.1c and 1.1.1d [10 Sep 2019]
1247 * Fixed a fork protection issue. OpenSSL 1.1.1 introduced a rewritten random
1248 number generator (RNG). This was intended to include protection in the
1249 event of a fork() system call in order to ensure that the parent and child
1250 processes did not share the same RNG state. However this protection was not
1251 being used in the default case.
1253 A partial mitigation for this issue is that the output from a high
1254 precision timer is mixed into the RNG state so the likelihood of a parent
1255 and child process sharing state is significantly reduced.
1257 If an application already calls OPENSSL_init_crypto() explicitly using
1258 OPENSSL_INIT_ATFORK then this problem does not occur at all.
1261 *Matthias St. Pierre*
1263 * For built-in EC curves, ensure an EC_GROUP built from the curve name is
1264 used even when parsing explicit parameters, when loading a encoded key
1265 or calling `EC_GROUP_new_from_ecpkparameters()`/
1266 `EC_GROUP_new_from_ecparameters()`.
1267 This prevents bypass of security hardening and performance gains,
1268 especially for curves with specialized EC_METHODs.
1269 By default, if a key encoded with explicit parameters is loaded and later
1270 encoded, the output is still encoded with explicit parameters, even if
1271 internally a "named" EC_GROUP is used for computation.
1275 * Compute ECC cofactors if not provided during EC_GROUP construction. Before
1276 this change, EC_GROUP_set_generator would accept order and/or cofactor as
1277 NULL. After this change, only the cofactor parameter can be NULL. It also
1278 does some minimal sanity checks on the passed order.
1283 * Fixed a padding oracle in PKCS7_dataDecode and CMS_decrypt_set1_pkey.
1284 An attack is simple, if the first CMS_recipientInfo is valid but the
1285 second CMS_recipientInfo is chosen ciphertext. If the second
1286 recipientInfo decodes to PKCS #1 v1.5 form plaintext, the correct
1287 encryption key will be replaced by garbage, and the message cannot be
1288 decoded, but if the RSA decryption fails, the correct encryption key is
1289 used and the recipient will not notice the attack.
1290 As a work around for this potential attack the length of the decrypted
1291 key must be equal to the cipher default key length, in case the
1292 certifiate is not given and all recipientInfo are tried out.
1293 The old behaviour can be re-enabled in the CMS code by setting the
1294 CMS_DEBUG_DECRYPT flag.
1299 * Early start up entropy quality from the DEVRANDOM seed source has been
1300 improved for older Linux systems. The RAND subsystem will wait for
1301 /dev/random to be producing output before seeding from /dev/urandom.
1302 The seeded state is stored for future library initialisations using
1303 a system global shared memory segment. The shared memory identifier
1304 can be configured by defining OPENSSL_RAND_SEED_DEVRANDOM_SHM_ID to
1305 the desired value. The default identifier is 114.
1309 * Correct the extended master secret constant on EBCDIC systems. Without this
1310 fix TLS connections between an EBCDIC system and a non-EBCDIC system that
1311 negotiate EMS will fail. Unfortunately this also means that TLS connections
1312 between EBCDIC systems with this fix, and EBCDIC systems without this
1313 fix will fail if they negotiate EMS.
1317 * Use Windows installation paths in the mingw builds
1319 Mingw isn't a POSIX environment per se, which means that Windows
1320 paths should be used for installation.
1325 * Changed DH_check to accept parameters with order q and 2q subgroups.
1326 With order 2q subgroups the bit 0 of the private key is not secret
1327 but DH_generate_key works around that by clearing bit 0 of the
1328 private key for those. This avoids leaking bit 0 of the private key.
1332 * Significantly reduce secure memory usage by the randomness pools.
1336 * Revert the DEVRANDOM_WAIT feature for Linux systems
1338 The DEVRANDOM_WAIT feature added a select() call to wait for the
1339 /dev/random device to become readable before reading from the
1340 /dev/urandom device.
1342 It turned out that this change had negative side effects on
1343 performance which were not acceptable. After some discussion it
1344 was decided to revert this feature and leave it up to the OS
1345 resp. the platform maintainer to ensure a proper initialization
1346 during early boot time.
1348 *Matthias St. Pierre*
1350 ### Changes between 1.1.1b and 1.1.1c [28 May 2019]
1352 * Add build tests for C++. These are generated files that only do one
1353 thing, to include one public OpenSSL head file each. This tests that
1354 the public header files can be usefully included in a C++ application.
1356 This test isn't enabled by default. It can be enabled with the option
1357 'enable-buildtest-c++'.
1361 * Enable SHA3 pre-hashing for ECDSA and DSA.
1365 * Change the default RSA, DSA and DH size to 2048 bit instead of 1024.
1366 This changes the size when using the genpkey app when no size is given. It
1367 fixes an omission in earlier changes that changed all RSA, DSA and DH
1368 generation apps to use 2048 bits by default.
1372 * Reorganize the manual pages to consistently have RETURN VALUES,
1373 EXAMPLES, SEE ALSO and HISTORY come in that order, and adjust
1374 util/fix-doc-nits accordingly.
1376 *Paul Yang, Joshua Lock*
1378 * Add the missing accessor EVP_PKEY_get0_engine()
1382 * Have apps like 's_client' and 's_server' output the signature scheme
1383 along with other cipher suite parameters when debugging.
1387 * Make OPENSSL_config() error agnostic again.
1391 * Do the error handling in RSA decryption constant time.
1395 * Prevent over long nonces in ChaCha20-Poly1305.
1397 ChaCha20-Poly1305 is an AEAD cipher, and requires a unique nonce input
1398 for every encryption operation. RFC 7539 specifies that the nonce value
1399 (IV) should be 96 bits (12 bytes). OpenSSL allows a variable nonce length
1400 and front pads the nonce with 0 bytes if it is less than 12
1401 bytes. However it also incorrectly allows a nonce to be set of up to 16
1402 bytes. In this case only the last 12 bytes are significant and any
1403 additional leading bytes are ignored.
1405 It is a requirement of using this cipher that nonce values are
1406 unique. Messages encrypted using a reused nonce value are susceptible to
1407 serious confidentiality and integrity attacks. If an application changes
1408 the default nonce length to be longer than 12 bytes and then makes a
1409 change to the leading bytes of the nonce expecting the new value to be a
1410 new unique nonce then such an application could inadvertently encrypt
1411 messages with a reused nonce.
1413 Additionally the ignored bytes in a long nonce are not covered by the
1414 integrity guarantee of this cipher. Any application that relies on the
1415 integrity of these ignored leading bytes of a long nonce may be further
1416 affected. Any OpenSSL internal use of this cipher, including in SSL/TLS,
1417 is safe because no such use sets such a long nonce value. However user
1418 applications that use this cipher directly and set a non-default nonce
1419 length to be longer than 12 bytes may be vulnerable.
1421 This issue was reported to OpenSSL on 16th of March 2019 by Joran Dirk
1427 * Add DEVRANDOM_WAIT feature for Linux systems
1429 On older Linux systems where the getrandom() system call is not available,
1430 OpenSSL normally uses the /dev/urandom device for seeding its CSPRNG.
1431 Contrary to getrandom(), the /dev/urandom device will not block during
1432 early boot when the kernel CSPRNG has not been seeded yet.
1434 To mitigate this known weakness, use select() to wait for /dev/random to
1435 become readable before reading from /dev/urandom.
1437 * Ensure that SM2 only uses SM3 as digest algorithm
1441 ### Changes between 1.1.1a and 1.1.1b [26 Feb 2019]
1443 * Change the info callback signals for the start and end of a post-handshake
1444 message exchange in TLSv1.3. In 1.1.1/1.1.1a we used SSL_CB_HANDSHAKE_START
1445 and SSL_CB_HANDSHAKE_DONE. Experience has shown that many applications get
1446 confused by this and assume that a TLSv1.2 renegotiation has started. This
1447 can break KeyUpdate handling. Instead we no longer signal the start and end
1448 of a post handshake message exchange (although the messages themselves are
1449 still signalled). This could break some applications that were expecting
1450 the old signals. However without this KeyUpdate is not usable for many
1455 ### Changes between 1.1.1 and 1.1.1a [20 Nov 2018]
1457 * Timing vulnerability in DSA signature generation
1459 The OpenSSL DSA signature algorithm has been shown to be vulnerable to a
1460 timing side channel attack. An attacker could use variations in the signing
1461 algorithm to recover the private key.
1463 This issue was reported to OpenSSL on 16th October 2018 by Samuel Weiser.
1468 * Timing vulnerability in ECDSA signature generation
1470 The OpenSSL ECDSA signature algorithm has been shown to be vulnerable to a
1471 timing side channel attack. An attacker could use variations in the signing
1472 algorithm to recover the private key.
1474 This issue was reported to OpenSSL on 25th October 2018 by Samuel Weiser.
1479 * Fixed the issue that RAND_add()/RAND_seed() silently discards random input
1480 if its length exceeds 4096 bytes. The limit has been raised to a buffer size
1481 of two gigabytes and the error handling improved.
1483 This issue was reported to OpenSSL by Dr. Falko Strenzke. It has been
1484 categorized as a normal bug, not a security issue, because the DRBG reseeds
1485 automatically and is fully functional even without additional randomness
1486 provided by the application.
1488 ### Changes between 1.1.0i and 1.1.1 [11 Sep 2018]
1490 * Add a new ClientHello callback. Provides a callback interface that gives
1491 the application the ability to adjust the nascent SSL object at the
1492 earliest stage of ClientHello processing, immediately after extensions have
1493 been collected but before they have been processed. In particular, this
1494 callback can adjust the supported TLS versions in response to the contents
1499 * Add SM2 base algorithm support.
1503 * s390x assembly pack: add (improved) hardware-support for the following
1504 cryptographic primitives: sha3, shake, aes-gcm, aes-ccm, aes-ctr, aes-ofb,
1505 aes-cfb/cfb8, aes-ecb.
1509 * Make EVP_PKEY_asn1_new() a bit stricter about its input. A NULL pem_str
1510 parameter is no longer accepted, as it leads to a corrupt table. NULL
1511 pem_str is reserved for alias entries only.
1515 * Use the new ec_scalar_mul_ladder scaffold to implement a specialized ladder
1516 step for prime curves. The new implementation is based on formulae from
1517 differential addition-and-doubling in homogeneous projective coordinates
1518 from Izu-Takagi "A fast parallel elliptic curve multiplication resistant
1519 against side channel attacks" and Brier-Joye "Weierstrass Elliptic Curves
1520 and Side-Channel Attacks" Eq. (8) for y-coordinate recovery, modified
1521 to work in projective coordinates.
1523 *Billy Bob Brumley, Nicola Tuveri*
1525 * Change generating and checking of primes so that the error rate of not
1526 being prime depends on the intended use based on the size of the input.
1527 For larger primes this will result in more rounds of Miller-Rabin.
1528 The maximal error rate for primes with more than 1080 bits is lowered
1531 *Kurt Roeckx, Annie Yousar*
1533 * Increase the number of Miller-Rabin rounds for DSA key generating to 64.
1537 * The 'tsget' script is renamed to 'tsget.pl', to avoid confusion when
1538 moving between systems, and to avoid confusion when a Windows build is
1539 done with mingw vs with MSVC. For POSIX installs, there's still a
1540 symlink or copy named 'tsget' to avoid that confusion as well.
1544 * Revert blinding in ECDSA sign and instead make problematic addition
1545 length-invariant. Switch even to fixed-length Montgomery multiplication.
1549 * Use the new ec_scalar_mul_ladder scaffold to implement a specialized ladder
1550 step for binary curves. The new implementation is based on formulae from
1551 differential addition-and-doubling in mixed Lopez-Dahab projective
1552 coordinates, modified to independently blind the operands.
1554 *Billy Bob Brumley, Sohaib ul Hassan, Nicola Tuveri*
1556 * Add a scaffold to optionally enhance the Montgomery ladder implementation
1557 for `ec_scalar_mul_ladder` (formerly `ec_mul_consttime`) allowing
1558 EC_METHODs to implement their own specialized "ladder step", to take
1559 advantage of more favorable coordinate systems or more efficient
1560 differential addition-and-doubling algorithms.
1562 *Billy Bob Brumley, Sohaib ul Hassan, Nicola Tuveri*
1564 * Modified the random device based seed sources to keep the relevant
1565 file descriptors open rather than reopening them on each access.
1566 This allows such sources to operate in a chroot() jail without
1567 the associated device nodes being available. This behaviour can be
1568 controlled using RAND_keep_random_devices_open().
1572 * Numerous side-channel attack mitigations have been applied. This may have
1573 performance impacts for some algorithms for the benefit of improved
1574 security. Specific changes are noted in this change log by their respective
1579 * AIX shared library support overhaul. Switch to AIX "natural" way of
1580 handling shared libraries, which means collecting shared objects of
1581 different versions and bitnesses in one common archive. This allows to
1582 mitigate conflict between 1.0 and 1.1 side-by-side installations. It
1583 doesn't affect the way 3rd party applications are linked, only how
1584 multi-version installation is managed.
1588 * Make ec_group_do_inverse_ord() more robust and available to other
1589 EC cryptosystems, so that irrespective of BN_FLG_CONSTTIME, SCA
1590 mitigations are applied to the fallback BN_mod_inverse().
1591 When using this function rather than BN_mod_inverse() directly, new
1592 EC cryptosystem implementations are then safer-by-default.
1596 * Add coordinate blinding for EC_POINT and implement projective
1597 coordinate blinding for generic prime curves as a countermeasure to
1598 chosen point SCA attacks.
1600 *Sohaib ul Hassan, Nicola Tuveri, Billy Bob Brumley*
1602 * Add blinding to ECDSA and DSA signatures to protect against side channel
1603 attacks discovered by Keegan Ryan (NCC Group).
1607 * Enforce checking in the pkeyutl command line app to ensure that the input
1608 length does not exceed the maximum supported digest length when performing
1609 a sign, verify or verifyrecover operation.
1613 * SSL_MODE_AUTO_RETRY is enabled by default. Applications that use blocking
1614 I/O in combination with something like select() or poll() will hang. This
1615 can be turned off again using SSL_CTX_clear_mode().
1616 Many applications do not properly handle non-application data records, and
1617 TLS 1.3 sends more of such records. Setting SSL_MODE_AUTO_RETRY works
1618 around the problems in those applications, but can also break some.
1619 It's recommended to read the manpages about SSL_read(), SSL_write(),
1620 SSL_get_error(), SSL_shutdown(), SSL_CTX_set_mode() and
1621 SSL_CTX_set_read_ahead() again.
1625 * When unlocking a pass phrase protected PEM file or PKCS#8 container, we
1626 now allow empty (zero character) pass phrases.
1630 * Apply blinding to binary field modular inversion and remove patent
1631 pending (OPENSSL_SUN_GF2M_DIV) BN_GF2m_mod_div implementation.
1635 * Deprecate ec2_mult.c and unify scalar multiplication code paths for
1636 binary and prime elliptic curves.
1640 * Remove ECDSA nonce padding: EC_POINT_mul is now responsible for
1641 constant time fixed point multiplication.
1645 * Revise elliptic curve scalar multiplication with timing attack
1646 defenses: ec_wNAF_mul redirects to a constant time implementation
1647 when computing fixed point and variable point multiplication (which
1648 in OpenSSL are mostly used with secret scalars in keygen, sign,
1649 ECDH derive operations).
1650 *Billy Bob Brumley, Nicola Tuveri, Cesar Pereida GarcÃa,
1653 * Updated CONTRIBUTING
1657 * Updated DRBG / RAND to request nonce and additional low entropy
1658 randomness from the system.
1660 *Matthias St. Pierre*
1662 * Updated 'openssl rehash' to use OpenSSL consistent default.
1666 * Moved the load of the ssl_conf module to libcrypto, which helps
1667 loading engines that libssl uses before libssl is initialised.
1671 * Added EVP_PKEY_sign() and EVP_PKEY_verify() for EdDSA
1675 * Fixed X509_NAME_ENTRY_set to get multi-valued RDNs right in all cases.
1677 *Ingo Schwarze, Rich Salz*
1679 * Added output of accepting IP address and port for 'openssl s_server'
1683 * Added a new API for TLSv1.3 ciphersuites:
1684 SSL_CTX_set_ciphersuites()
1685 SSL_set_ciphersuites()
1689 * Memory allocation failures consistently add an error to the error
1694 * Don't use OPENSSL_ENGINES and OPENSSL_CONF environment values
1695 in libcrypto when run as setuid/setgid.
1699 * Load any config file by default when libssl is used.
1703 * Added new public header file <openssl/rand_drbg.h> and documentation
1704 for the RAND_DRBG API. See manual page RAND_DRBG(7) for an overview.
1706 *Matthias St. Pierre*
1708 * QNX support removed (cannot find contributors to get their approval
1709 for the license change).
1713 * TLSv1.3 replay protection for early data has been implemented. See the
1714 SSL_read_early_data() man page for further details.
1718 * Separated TLSv1.3 ciphersuite configuration out from TLSv1.2 ciphersuite
1719 configuration. TLSv1.3 ciphersuites are not compatible with TLSv1.2 and
1720 below. Similarly TLSv1.2 ciphersuites are not compatible with TLSv1.3.
1721 In order to avoid issues where legacy TLSv1.2 ciphersuite configuration
1722 would otherwise inadvertently disable all TLSv1.3 ciphersuites the
1723 configuration has been separated out. See the ciphers man page or the
1724 SSL_CTX_set_ciphersuites() man page for more information.
1728 * On POSIX (BSD, Linux, ...) systems the ocsp(1) command running
1729 in responder mode now supports the new "-multi" option, which
1730 spawns the specified number of child processes to handle OCSP
1731 requests. The "-timeout" option now also limits the OCSP
1732 responder's patience to wait to receive the full client request
1733 on a newly accepted connection. Child processes are respawned
1734 as needed, and the CA index file is automatically reloaded
1735 when changed. This makes it possible to run the "ocsp" responder
1736 as a long-running service, making the OpenSSL CA somewhat more
1737 feature-complete. In this mode, most diagnostic messages logged
1738 after entering the event loop are logged via syslog(3) rather than
1743 * Added support for X448 and Ed448. Heavily based on original work by
1748 * Extend OSSL_STORE with capabilities to search and to narrow the set of
1749 objects loaded. This adds the functions OSSL_STORE_expect() and
1750 OSSL_STORE_find() as well as needed tools to construct searches and
1751 get the search data out of them.
1755 * Support for TLSv1.3 added. Note that users upgrading from an earlier
1756 version of OpenSSL should review their configuration settings to ensure
1757 that they are still appropriate for TLSv1.3. For further information see:
1758 <https://wiki.openssl.org/index.php/TLS1.3>
1762 * Grand redesign of the OpenSSL random generator
1764 The default RAND method now utilizes an AES-CTR DRBG according to
1765 NIST standard SP 800-90Ar1. The new random generator is essentially
1766 a port of the default random generator from the OpenSSL FIPS 2.0
1767 object module. It is a hybrid deterministic random bit generator
1768 using an AES-CTR bit stream and which seeds and reseeds itself
1769 automatically using trusted system entropy sources.
1771 Some of its new features are:
1772 - Support for multiple DRBG instances with seed chaining.
1773 - The default RAND method makes use of a DRBG.
1774 - There is a public and private DRBG instance.
1775 - The DRBG instances are fork-safe.
1776 - Keep all global DRBG instances on the secure heap if it is enabled.
1777 - The public and private DRBG instance are per thread for lock free
1780 *Paul Dale, Benjamin Kaduk, Kurt Roeckx, Rich Salz, Matthias St. Pierre*
1782 * Changed Configure so it only says what it does and doesn't dump
1783 so much data. Instead, ./configdata.pm should be used as a script
1784 to display all sorts of configuration data.
1788 * Added processing of "make variables" to Configure.
1792 * Added SHA512/224 and SHA512/256 algorithm support.
1796 * The last traces of Netware support, first removed in 1.1.0, have
1801 * Get rid of Makefile.shared, and in the process, make the processing
1802 of certain files (rc.obj, or the .def/.map/.opt files produced from
1803 the ordinal files) more visible and hopefully easier to trace and
1804 debug (or make silent).
1808 * Make it possible to have environment variable assignments as
1809 arguments to config / Configure.
1813 * Add multi-prime RSA (RFC 8017) support.
1817 * Add SM3 implemented according to GB/T 32905-2016
1818 *Jack Lloyd <jack.lloyd@ribose.com>,*
1819 *Ronald Tse <ronald.tse@ribose.com>,*
1820 *Erick Borsboom <erick.borsboom@ribose.com>*
1822 * Add 'Maximum Fragment Length' TLS extension negotiation and support
1823 as documented in RFC6066.
1824 Based on a patch from Tomasz Moń
1826 *Filipe Raimundo da Silva*
1828 * Add SM4 implemented according to GB/T 32907-2016.
1829 *Jack Lloyd <jack.lloyd@ribose.com>,*
1830 *Ronald Tse <ronald.tse@ribose.com>,*
1831 *Erick Borsboom <erick.borsboom@ribose.com>*
1833 * Reimplement -newreq-nodes and ERR_error_string_n; the
1834 original author does not agree with the license change.
1838 * Add ARIA AEAD TLS support.
1842 * Some macro definitions to support VS6 have been removed. Visual
1843 Studio 6 has not worked since 1.1.0
1847 * Add ERR_clear_last_mark(), to allow callers to clear the last mark
1848 without clearing the errors.
1852 * Add "atfork" functions. If building on a system that without
1853 pthreads, see doc/man3/OPENSSL_fork_prepare.pod for application
1854 requirements. The RAND facility now uses/requires this.
1862 * The UI API becomes a permanent and integral part of libcrypto, i.e.
1863 not possible to disable entirely. However, it's still possible to
1864 disable the console reading UI method, UI_OpenSSL() (use UI_null()
1867 To disable, configure with 'no-ui-console'. 'no-ui' is still
1868 possible to use as an alias. Check at compile time with the
1869 macro OPENSSL_NO_UI_CONSOLE. The macro OPENSSL_NO_UI is still
1870 possible to check and is an alias for OPENSSL_NO_UI_CONSOLE.
1874 * Add a STORE module, which implements a uniform and URI based reader of
1875 stores that can contain keys, certificates, CRLs and numerous other
1876 objects. The main API is loosely based on a few stdio functions,
1877 and includes OSSL_STORE_open, OSSL_STORE_load, OSSL_STORE_eof,
1878 OSSL_STORE_error and OSSL_STORE_close.
1879 The implementation uses backends called "loaders" to implement arbitrary
1880 URI schemes. There is one built in "loader" for the 'file' scheme.
1884 * Add devcrypto engine. This has been implemented against cryptodev-linux,
1885 then adjusted to work on FreeBSD 8.4 as well.
1886 Enable by configuring with 'enable-devcryptoeng'. This is done by default
1887 on BSD implementations, as cryptodev.h is assumed to exist on all of them.
1891 * Module names can prefixed with OSSL_ or OPENSSL_. This affects
1892 util/mkerr.pl, which is adapted to allow those prefixes, leading to
1893 error code calls like this:
1895 OSSL_FOOerr(OSSL_FOO_F_SOMETHING, OSSL_FOO_R_WHATEVER);
1897 With this change, we claim the namespaces OSSL and OPENSSL in a manner
1898 that can be encoded in C. For the foreseeable future, this will only
1901 *Richard Levitte and Tim Hudson*
1903 * Removed BSD cryptodev engine.
1907 * Add a build target 'build_all_generated', to build all generated files
1908 and only that. This can be used to prepare everything that requires
1909 things like perl for a system that lacks perl and then move everything
1910 to that system and do the rest of the build there.
1914 * In the UI interface, make it possible to duplicate the user data. This
1915 can be used by engines that need to retain the data for a longer time
1916 than just the call where this user data is passed.
1920 * Ignore the '-named_curve auto' value for compatibility of applications
1923 *Tomas Mraz <tmraz@fedoraproject.org>*
1925 * Fragmented SSL/TLS alerts are no longer accepted. An alert message is 2
1926 bytes long. In theory it is permissible in SSLv3 - TLSv1.2 to fragment such
1927 alerts across multiple records (some of which could be empty). In practice
1928 it make no sense to send an empty alert record, or to fragment one. TLSv1.3
1929 prohibits this altogether and other libraries (BoringSSL, NSS) do not
1930 support this at all. Supporting it adds significant complexity to the
1931 record layer, and its removal is unlikely to cause interoperability
1936 * Add the ASN.1 types INT32, UINT32, INT64, UINT64 and variants prefixed
1937 with Z. These are meant to replace LONG and ZLONG and to be size safe.
1938 The use of LONG and ZLONG is discouraged and scheduled for deprecation
1943 * Add the 'z' and 'j' modifiers to BIO_printf() et al formatting string,
1944 'z' is to be used for [s]size_t, and 'j' - with [u]int64_t.
1946 *Richard Levitte, Andy Polyakov*
1948 * Add EC_KEY_get0_engine(), which does for EC_KEY what RSA_get0_engine()
1953 * Have 'config' recognise 64-bit mingw and choose 'mingw64' as the target
1954 platform rather than 'mingw'.
1958 * The functions X509_STORE_add_cert and X509_STORE_add_crl return
1959 success if they are asked to add an object which already exists
1960 in the store. This change cascades to other functions which load
1961 certificates and CRLs.
1965 * x86_64 assembly pack: annotate code with DWARF CFI directives to
1966 facilitate stack unwinding even from assembly subroutines.
1970 * Remove VAX C specific definitions of OPENSSL_EXPORT, OPENSSL_EXTERN.
1971 Also remove OPENSSL_GLOBAL entirely, as it became a no-op.
1975 * Remove the VMS-specific reimplementation of gmtime from crypto/o_times.c.
1976 VMS C's RTL has a fully up to date gmtime() and gmtime_r() since V7.1,
1977 which is the minimum version we support.
1981 * Certificate time validation (X509_cmp_time) enforces stricter
1982 compliance with RFC 5280. Fractional seconds and timezone offsets
1983 are no longer allowed.
1987 * Add support for ARIA
1991 * s_client will now send the Server Name Indication (SNI) extension by
1992 default unless the new "-noservername" option is used. The server name is
1993 based on the host provided to the "-connect" option unless overridden by
1994 using "-servername".
1998 * Add support for SipHash
2002 * OpenSSL now fails if it receives an unrecognised record type in TLS1.0
2003 or TLS1.1. Previously this only happened in SSLv3 and TLS1.2. This is to
2004 prevent issues where no progress is being made and the peer continually
2005 sends unrecognised record types, using up resources processing them.
2009 * 'openssl passwd' can now produce SHA256 and SHA512 based output,
2010 using the algorithm defined in
2011 <https://www.akkadia.org/drepper/SHA-crypt.txt>
2015 * Heartbeat support has been removed; the ABI is changed for now.
2017 *Richard Levitte, Rich Salz*
2019 * Support for SSL_OP_NO_ENCRYPT_THEN_MAC in SSL_CONF_cmd.
2023 * The RSA "null" method, which was partially supported to avoid patent
2024 issues, has been replaced to always returns NULL.
2031 ### Changes between 1.1.0k and 1.1.0l [10 Sep 2019]
2033 * For built-in EC curves, ensure an EC_GROUP built from the curve name is
2034 used even when parsing explicit parameters, when loading a encoded key
2035 or calling `EC_GROUP_new_from_ecpkparameters()`/
2036 `EC_GROUP_new_from_ecparameters()`.
2037 This prevents bypass of security hardening and performance gains,
2038 especially for curves with specialized EC_METHODs.
2039 By default, if a key encoded with explicit parameters is loaded and later
2040 encoded, the output is still encoded with explicit parameters, even if
2041 internally a "named" EC_GROUP is used for computation.
2045 * Compute ECC cofactors if not provided during EC_GROUP construction. Before
2046 this change, EC_GROUP_set_generator would accept order and/or cofactor as
2047 NULL. After this change, only the cofactor parameter can be NULL. It also
2048 does some minimal sanity checks on the passed order.
2053 * Fixed a padding oracle in PKCS7_dataDecode and CMS_decrypt_set1_pkey.
2054 An attack is simple, if the first CMS_recipientInfo is valid but the
2055 second CMS_recipientInfo is chosen ciphertext. If the second
2056 recipientInfo decodes to PKCS #1 v1.5 form plaintext, the correct
2057 encryption key will be replaced by garbage, and the message cannot be
2058 decoded, but if the RSA decryption fails, the correct encryption key is
2059 used and the recipient will not notice the attack.
2060 As a work around for this potential attack the length of the decrypted
2061 key must be equal to the cipher default key length, in case the
2062 certifiate is not given and all recipientInfo are tried out.
2063 The old behaviour can be re-enabled in the CMS code by setting the
2064 CMS_DEBUG_DECRYPT flag.
2069 * Use Windows installation paths in the mingw builds
2071 Mingw isn't a POSIX environment per se, which means that Windows
2072 paths should be used for installation.
2077 ### Changes between 1.1.0j and 1.1.0k [28 May 2019]
2079 * Change the default RSA, DSA and DH size to 2048 bit instead of 1024.
2080 This changes the size when using the genpkey app when no size is given. It
2081 fixes an omission in earlier changes that changed all RSA, DSA and DH
2082 generation apps to use 2048 bits by default.
2086 * Prevent over long nonces in ChaCha20-Poly1305.
2088 ChaCha20-Poly1305 is an AEAD cipher, and requires a unique nonce input
2089 for every encryption operation. RFC 7539 specifies that the nonce value
2090 (IV) should be 96 bits (12 bytes). OpenSSL allows a variable nonce length
2091 and front pads the nonce with 0 bytes if it is less than 12
2092 bytes. However it also incorrectly allows a nonce to be set of up to 16
2093 bytes. In this case only the last 12 bytes are significant and any
2094 additional leading bytes are ignored.
2096 It is a requirement of using this cipher that nonce values are
2097 unique. Messages encrypted using a reused nonce value are susceptible to
2098 serious confidentiality and integrity attacks. If an application changes
2099 the default nonce length to be longer than 12 bytes and then makes a
2100 change to the leading bytes of the nonce expecting the new value to be a
2101 new unique nonce then such an application could inadvertently encrypt
2102 messages with a reused nonce.
2104 Additionally the ignored bytes in a long nonce are not covered by the
2105 integrity guarantee of this cipher. Any application that relies on the
2106 integrity of these ignored leading bytes of a long nonce may be further
2107 affected. Any OpenSSL internal use of this cipher, including in SSL/TLS,
2108 is safe because no such use sets such a long nonce value. However user
2109 applications that use this cipher directly and set a non-default nonce
2110 length to be longer than 12 bytes may be vulnerable.
2112 This issue was reported to OpenSSL on 16th of March 2019 by Joran Dirk
2118 * Added SCA hardening for modular field inversion in EC_GROUP through
2119 a new dedicated field_inv() pointer in EC_METHOD.
2120 This also addresses a leakage affecting conversions from projective
2121 to affine coordinates.
2123 *Billy Bob Brumley, Nicola Tuveri*
2125 * Fix a use after free bug in d2i_X509_PUBKEY when overwriting a
2126 re-used X509_PUBKEY object if the second PUBKEY is malformed.
2130 * Move strictness check from EVP_PKEY_asn1_new() to EVP_PKEY_asn1_add0().
2134 * Remove the 'dist' target and add a tarball building script. The
2135 'dist' target has fallen out of use, and it shouldn't be
2136 necessary to configure just to create a source distribution.
2140 ### Changes between 1.1.0i and 1.1.0j [20 Nov 2018]
2142 * Timing vulnerability in DSA signature generation
2144 The OpenSSL DSA signature algorithm has been shown to be vulnerable to a
2145 timing side channel attack. An attacker could use variations in the signing
2146 algorithm to recover the private key.
2148 This issue was reported to OpenSSL on 16th October 2018 by Samuel Weiser.
2153 * Timing vulnerability in ECDSA signature generation
2155 The OpenSSL ECDSA signature algorithm has been shown to be vulnerable to a
2156 timing side channel attack. An attacker could use variations in the signing
2157 algorithm to recover the private key.
2159 This issue was reported to OpenSSL on 25th October 2018 by Samuel Weiser.
2164 * Add coordinate blinding for EC_POINT and implement projective
2165 coordinate blinding for generic prime curves as a countermeasure to
2166 chosen point SCA attacks.
2168 *Sohaib ul Hassan, Nicola Tuveri, Billy Bob Brumley*
2170 ### Changes between 1.1.0h and 1.1.0i [14 Aug 2018]
2172 * Client DoS due to large DH parameter
2174 During key agreement in a TLS handshake using a DH(E) based ciphersuite a
2175 malicious server can send a very large prime value to the client. This will
2176 cause the client to spend an unreasonably long period of time generating a
2177 key for this prime resulting in a hang until the client has finished. This
2178 could be exploited in a Denial Of Service attack.
2180 This issue was reported to OpenSSL on 5th June 2018 by Guido Vranken
2185 * Cache timing vulnerability in RSA Key Generation
2187 The OpenSSL RSA Key generation algorithm has been shown to be vulnerable to
2188 a cache timing side channel attack. An attacker with sufficient access to
2189 mount cache timing attacks during the RSA key generation process could
2190 recover the private key.
2192 This issue was reported to OpenSSL on 4th April 2018 by Alejandro Cabrera
2193 Aldaya, Billy Brumley, Cesar Pereida Garcia and Luis Manuel Alvarez Tapia.
2198 * Make EVP_PKEY_asn1_new() a bit stricter about its input. A NULL pem_str
2199 parameter is no longer accepted, as it leads to a corrupt table. NULL
2200 pem_str is reserved for alias entries only.
2204 * Revert blinding in ECDSA sign and instead make problematic addition
2205 length-invariant. Switch even to fixed-length Montgomery multiplication.
2209 * Change generating and checking of primes so that the error rate of not
2210 being prime depends on the intended use based on the size of the input.
2211 For larger primes this will result in more rounds of Miller-Rabin.
2212 The maximal error rate for primes with more than 1080 bits is lowered
2215 *Kurt Roeckx, Annie Yousar*
2217 * Increase the number of Miller-Rabin rounds for DSA key generating to 64.
2221 * Add blinding to ECDSA and DSA signatures to protect against side channel
2222 attacks discovered by Keegan Ryan (NCC Group).
2226 * When unlocking a pass phrase protected PEM file or PKCS#8 container, we
2227 now allow empty (zero character) pass phrases.
2231 * Certificate time validation (X509_cmp_time) enforces stricter
2232 compliance with RFC 5280. Fractional seconds and timezone offsets
2233 are no longer allowed.
2237 * Fixed a text canonicalisation bug in CMS
2239 Where a CMS detached signature is used with text content the text goes
2240 through a canonicalisation process first prior to signing or verifying a
2241 signature. This process strips trailing space at the end of lines, converts
2242 line terminators to CRLF and removes additional trailing line terminators
2243 at the end of a file. A bug in the canonicalisation process meant that
2244 some characters, such as form-feed, were incorrectly treated as whitespace
2245 and removed. This is contrary to the specification (RFC5485). This fix
2246 could mean that detached text data signed with an earlier version of
2247 OpenSSL 1.1.0 may fail to verify using the fixed version, or text data
2248 signed with a fixed OpenSSL may fail to verify with an earlier version of
2249 OpenSSL 1.1.0. A workaround is to only verify the canonicalised text data
2250 and use the "-binary" flag (for the "cms" command line application) or set
2251 the SMIME_BINARY/PKCS7_BINARY/CMS_BINARY flags (if using CMS_verify()).
2255 ### Changes between 1.1.0g and 1.1.0h [27 Mar 2018]
2257 * Constructed ASN.1 types with a recursive definition could exceed the stack
2259 Constructed ASN.1 types with a recursive definition (such as can be found
2260 in PKCS7) could eventually exceed the stack given malicious input with
2261 excessive recursion. This could result in a Denial Of Service attack. There
2262 are no such structures used within SSL/TLS that come from untrusted sources
2263 so this is considered safe.
2265 This issue was reported to OpenSSL on 4th January 2018 by the OSS-fuzz
2271 * Incorrect CRYPTO_memcmp on HP-UX PA-RISC
2273 Because of an implementation bug the PA-RISC CRYPTO_memcmp function is
2274 effectively reduced to only comparing the least significant bit of each
2275 byte. This allows an attacker to forge messages that would be considered as
2276 authenticated in an amount of tries lower than that guaranteed by the
2277 security claims of the scheme. The module can only be compiled by the
2278 HP-UX assembler, so that only HP-UX PA-RISC targets are affected.
2280 This issue was reported to OpenSSL on 2nd March 2018 by Peter Waltenberg
2286 * Add a build target 'build_all_generated', to build all generated files
2287 and only that. This can be used to prepare everything that requires
2288 things like perl for a system that lacks perl and then move everything
2289 to that system and do the rest of the build there.
2293 * Backport SSL_OP_NO_RENGOTIATION
2295 OpenSSL 1.0.2 and below had the ability to disable renegotiation using the
2296 (undocumented) SSL3_FLAGS_NO_RENEGOTIATE_CIPHERS flag. Due to the opacity
2297 changes this is no longer possible in 1.1.0. Therefore the new
2298 SSL_OP_NO_RENEGOTIATION option from 1.1.1-dev has been backported to
2299 1.1.0 to provide equivalent functionality.
2301 Note that if an application built against 1.1.0h headers (or above) is run
2302 using an older version of 1.1.0 (prior to 1.1.0h) then the option will be
2303 accepted but nothing will happen, i.e. renegotiation will not be prevented.
2307 * Removed the OS390-Unix config target. It relied on a script that doesn't
2312 * rsaz_1024_mul_avx2 overflow bug on x86_64
2314 There is an overflow bug in the AVX2 Montgomery multiplication procedure
2315 used in exponentiation with 1024-bit moduli. No EC algorithms are affected.
2316 Analysis suggests that attacks against RSA and DSA as a result of this
2317 defect would be very difficult to perform and are not believed likely.
2318 Attacks against DH1024 are considered just feasible, because most of the
2319 work necessary to deduce information about a private key may be performed
2320 offline. The amount of resources required for such an attack would be
2321 significant. However, for an attack on TLS to be meaningful, the server
2322 would have to share the DH1024 private key among multiple clients, which is
2323 no longer an option since CVE-2016-0701.
2325 This only affects processors that support the AVX2 but not ADX extensions
2326 like Intel Haswell (4th generation).
2328 This issue was reported to OpenSSL by David Benjamin (Google). The issue
2329 was originally found via the OSS-Fuzz project.
2334 ### Changes between 1.1.0f and 1.1.0g [2 Nov 2017]
2336 * bn_sqrx8x_internal carry bug on x86_64
2338 There is a carry propagating bug in the x86_64 Montgomery squaring
2339 procedure. No EC algorithms are affected. Analysis suggests that attacks
2340 against RSA and DSA as a result of this defect would be very difficult to
2341 perform and are not believed likely. Attacks against DH are considered just
2342 feasible (although very difficult) because most of the work necessary to
2343 deduce information about a private key may be performed offline. The amount
2344 of resources required for such an attack would be very significant and
2345 likely only accessible to a limited number of attackers. An attacker would
2346 additionally need online access to an unpatched system using the target
2347 private key in a scenario with persistent DH parameters and a private
2348 key that is shared between multiple clients.
2350 This only affects processors that support the BMI1, BMI2 and ADX extensions
2351 like Intel Broadwell (5th generation) and later or AMD Ryzen.
2353 This issue was reported to OpenSSL by the OSS-Fuzz project.
2358 * Malformed X.509 IPAddressFamily could cause OOB read
2360 If an X.509 certificate has a malformed IPAddressFamily extension,
2361 OpenSSL could do a one-byte buffer overread. The most likely result
2362 would be an erroneous display of the certificate in text format.
2364 This issue was reported to OpenSSL by the OSS-Fuzz project.
2369 ### Changes between 1.1.0e and 1.1.0f [25 May 2017]
2371 * Have 'config' recognise 64-bit mingw and choose 'mingw64' as the target
2372 platform rather than 'mingw'.
2376 * Remove the VMS-specific reimplementation of gmtime from crypto/o_times.c.
2377 VMS C's RTL has a fully up to date gmtime() and gmtime_r() since V7.1,
2378 which is the minimum version we support.
2382 ### Changes between 1.1.0d and 1.1.0e [16 Feb 2017]
2384 * Encrypt-Then-Mac renegotiation crash
2386 During a renegotiation handshake if the Encrypt-Then-Mac extension is
2387 negotiated where it was not in the original handshake (or vice-versa) then
2388 this can cause OpenSSL to crash (dependant on ciphersuite). Both clients
2389 and servers are affected.
2391 This issue was reported to OpenSSL by Joe Orton (Red Hat).
2396 ### Changes between 1.1.0c and 1.1.0d [26 Jan 2017]
2398 * Truncated packet could crash via OOB read
2400 If one side of an SSL/TLS path is running on a 32-bit host and a specific
2401 cipher is being used, then a truncated packet can cause that host to
2402 perform an out-of-bounds read, usually resulting in a crash.
2404 This issue was reported to OpenSSL by Robert Święcki of Google.
2409 * Bad (EC)DHE parameters cause a client crash
2411 If a malicious server supplies bad parameters for a DHE or ECDHE key
2412 exchange then this can result in the client attempting to dereference a
2413 NULL pointer leading to a client crash. This could be exploited in a Denial
2416 This issue was reported to OpenSSL by Guido Vranken.
2421 * BN_mod_exp may produce incorrect results on x86_64
2423 There is a carry propagating bug in the x86_64 Montgomery squaring
2424 procedure. No EC algorithms are affected. Analysis suggests that attacks
2425 against RSA and DSA as a result of this defect would be very difficult to
2426 perform and are not believed likely. Attacks against DH are considered just
2427 feasible (although very difficult) because most of the work necessary to
2428 deduce information about a private key may be performed offline. The amount
2429 of resources required for such an attack would be very significant and
2430 likely only accessible to a limited number of attackers. An attacker would
2431 additionally need online access to an unpatched system using the target
2432 private key in a scenario with persistent DH parameters and a private
2433 key that is shared between multiple clients. For example this can occur by
2434 default in OpenSSL DHE based SSL/TLS ciphersuites. Note: This issue is very
2435 similar to CVE-2015-3193 but must be treated as a separate problem.
2437 This issue was reported to OpenSSL by the OSS-Fuzz project.
2442 ### Changes between 1.1.0b and 1.1.0c [10 Nov 2016]
2444 * ChaCha20/Poly1305 heap-buffer-overflow
2446 TLS connections using `*-CHACHA20-POLY1305` ciphersuites are susceptible to
2447 a DoS attack by corrupting larger payloads. This can result in an OpenSSL
2448 crash. This issue is not considered to be exploitable beyond a DoS.
2450 This issue was reported to OpenSSL by Robert Święcki (Google Security Team)
2455 * CMS Null dereference
2457 Applications parsing invalid CMS structures can crash with a NULL pointer
2458 dereference. This is caused by a bug in the handling of the ASN.1 CHOICE
2459 type in OpenSSL 1.1.0 which can result in a NULL value being passed to the
2460 structure callback if an attempt is made to free certain invalid encodings.
2461 Only CHOICE structures using a callback which do not handle NULL value are
2464 This issue was reported to OpenSSL by Tyler Nighswander of ForAllSecure.
2469 * Montgomery multiplication may produce incorrect results
2471 There is a carry propagating bug in the Broadwell-specific Montgomery
2472 multiplication procedure that handles input lengths divisible by, but
2473 longer than 256 bits. Analysis suggests that attacks against RSA, DSA
2474 and DH private keys are impossible. This is because the subroutine in
2475 question is not used in operations with the private key itself and an input
2476 of the attacker's direct choice. Otherwise the bug can manifest itself as
2477 transient authentication and key negotiation failures or reproducible
2478 erroneous outcome of public-key operations with specially crafted input.
2479 Among EC algorithms only Brainpool P-512 curves are affected and one
2480 presumably can attack ECDH key negotiation. Impact was not analyzed in
2481 detail, because pre-requisites for attack are considered unlikely. Namely
2482 multiple clients have to choose the curve in question and the server has to
2483 share the private key among them, neither of which is default behaviour.
2484 Even then only clients that chose the curve will be affected.
2486 This issue was publicly reported as transient failures and was not
2487 initially recognized as a security issue. Thanks to Richard Morgan for
2488 providing reproducible case.
2493 * Removed automatic addition of RPATH in shared libraries and executables,
2494 as this was a remainder from OpenSSL 1.0.x and isn't needed any more.
2498 ### Changes between 1.1.0a and 1.1.0b [26 Sep 2016]
2500 * Fix Use After Free for large message sizes
2502 The patch applied to address CVE-2016-6307 resulted in an issue where if a
2503 message larger than approx 16k is received then the underlying buffer to
2504 store the incoming message is reallocated and moved. Unfortunately a
2505 dangling pointer to the old location is left which results in an attempt to
2506 write to the previously freed location. This is likely to result in a
2507 crash, however it could potentially lead to execution of arbitrary code.
2509 This issue only affects OpenSSL 1.1.0a.
2511 This issue was reported to OpenSSL by Robert Święcki.
2516 ### Changes between 1.1.0 and 1.1.0a [22 Sep 2016]
2518 * OCSP Status Request extension unbounded memory growth
2520 A malicious client can send an excessively large OCSP Status Request
2521 extension. If that client continually requests renegotiation, sending a
2522 large OCSP Status Request extension each time, then there will be unbounded
2523 memory growth on the server. This will eventually lead to a Denial Of
2524 Service attack through memory exhaustion. Servers with a default
2525 configuration are vulnerable even if they do not support OCSP. Builds using
2526 the "no-ocsp" build time option are not affected.
2528 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
2533 * SSL_peek() hang on empty record
2535 OpenSSL 1.1.0 SSL/TLS will hang during a call to SSL_peek() if the peer
2536 sends an empty record. This could be exploited by a malicious peer in a
2537 Denial Of Service attack.
2539 This issue was reported to OpenSSL by Alex Gaynor.
2544 * Excessive allocation of memory in tls_get_message_header() and
2545 dtls1_preprocess_fragment()
2547 A (D)TLS message includes 3 bytes for its length in the header for the
2548 message. This would allow for messages up to 16Mb in length. Messages of
2549 this length are excessive and OpenSSL includes a check to ensure that a
2550 peer is sending reasonably sized messages in order to avoid too much memory
2551 being consumed to service a connection. A flaw in the logic of version
2552 1.1.0 means that memory for the message is allocated too early, prior to
2553 the excessive message length check. Due to way memory is allocated in
2554 OpenSSL this could mean an attacker could force up to 21Mb to be allocated
2555 to service a connection. This could lead to a Denial of Service through
2556 memory exhaustion. However, the excessive message length check still takes
2557 place, and this would cause the connection to immediately fail. Assuming
2558 that the application calls SSL_free() on the failed connection in a timely
2559 manner then the 21Mb of allocated memory will then be immediately freed
2560 again. Therefore the excessive memory allocation will be transitory in
2561 nature. This then means that there is only a security impact if:
2563 1) The application does not call SSL_free() in a timely manner in the event
2564 that the connection fails
2566 2) The application is working in a constrained environment where there is
2567 very little free memory
2569 3) The attacker initiates multiple connection attempts such that there are
2570 multiple connections in a state where memory has been allocated for the
2571 connection; SSL_free() has not yet been called; and there is insufficient
2572 memory to service the multiple requests.
2574 Except in the instance of (1) above any Denial Of Service is likely to be
2575 transitory because as soon as the connection fails the memory is
2576 subsequently freed again in the SSL_free() call. However there is an
2577 increased risk during this period of application crashes due to the lack of
2578 memory - which would then mean a more serious Denial of Service.
2580 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
2581 (CVE-2016-6307 and CVE-2016-6308)
2585 * solaris-x86-cc, i.e. 32-bit configuration with vendor compiler,
2586 had to be removed. Primary reason is that vendor assembler can't
2587 assemble our modules with -KPIC flag. As result it, assembly
2588 support, was not even available as option. But its lack means
2589 lack of side-channel resistant code, which is incompatible with
2590 security by todays standards. Fortunately gcc is readily available
2591 prepackaged option, which we firmly point at...
2595 ### Changes between 1.0.2h and 1.1.0 [25 Aug 2016]
2597 * Windows command-line tool supports UTF-8 opt-in option for arguments
2598 and console input. Setting OPENSSL_WIN32_UTF8 environment variable
2599 (to any value) allows Windows user to access PKCS#12 file generated
2600 with Windows CryptoAPI and protected with non-ASCII password, as well
2601 as files generated under UTF-8 locale on Linux also protected with
2606 * To mitigate the SWEET32 attack [CVE-2016-2183][], 3DES cipher suites
2607 have been disabled by default and removed from DEFAULT, just like RC4.
2608 See the RC4 item below to re-enable both.
2612 * The method for finding the storage location for the Windows RAND seed file
2613 has changed. First we check %RANDFILE%. If that is not set then we check
2614 the directories %HOME%, %USERPROFILE% and %SYSTEMROOT% in that order. If
2615 all else fails we fall back to C:\.
2619 * The EVP_EncryptUpdate() function has had its return type changed from void
2620 to int. A return of 0 indicates and error while a return of 1 indicates
2625 * The flags RSA_FLAG_NO_CONSTTIME, DSA_FLAG_NO_EXP_CONSTTIME and
2626 DH_FLAG_NO_EXP_CONSTTIME which previously provided the ability to switch
2627 off the constant time implementation for RSA, DSA and DH have been made
2628 no-ops and deprecated.
2632 * Windows RAND implementation was simplified to only get entropy by
2633 calling CryptGenRandom(). Various other RAND-related tickets
2636 *Joseph Wylie Yandle, Rich Salz*
2638 * The stack and lhash API's were renamed to start with `OPENSSL_SK_`
2639 and `OPENSSL_LH_`, respectively. The old names are available
2640 with API compatibility. They new names are now completely documented.
2644 * Unify TYPE_up_ref(obj) methods signature.
2645 SSL_CTX_up_ref(), SSL_up_ref(), X509_up_ref(), EVP_PKEY_up_ref(),
2646 X509_CRL_up_ref(), X509_OBJECT_up_ref_count() methods are now returning an
2647 int (instead of void) like all others TYPE_up_ref() methods.
2648 So now these methods also check the return value of CRYPTO_atomic_add(),
2649 and the validity of object reference counter.
2651 *fdasilvayy@gmail.com*
2653 * With Windows Visual Studio builds, the .pdb files are installed
2654 alongside the installed libraries and executables. For a static
2655 library installation, ossl_static.pdb is the associate compiler
2656 generated .pdb file to be used when linking programs.
2660 * Remove openssl.spec. Packaging files belong with the packagers.
2664 * Automatic Darwin/OSX configuration has had a refresh, it will now
2665 recognise x86_64 architectures automatically. You can still decide
2666 to build for a different bitness with the environment variable
2667 KERNEL_BITS (can be 32 or 64), for example:
2669 KERNEL_BITS=32 ./config
2673 * Change default algorithms in pkcs8 utility to use PKCS#5 v2.0,
2674 256 bit AES and HMAC with SHA256.
2678 * Remove support for MIPS o32 ABI on IRIX (and IRIX only).
2682 * Triple-DES ciphers have been moved from HIGH to MEDIUM.
2686 * To enable users to have their own config files and build file templates,
2687 Configure looks in the directory indicated by the environment variable
2688 OPENSSL_LOCAL_CONFIG_DIR as well as the in-source Configurations/
2689 directory. On VMS, OPENSSL_LOCAL_CONFIG_DIR is expected to be a logical
2690 name and is used as is.
2694 * The following datatypes were made opaque: X509_OBJECT, X509_STORE_CTX,
2695 X509_STORE, X509_LOOKUP, and X509_LOOKUP_METHOD. The unused type
2696 X509_CERT_FILE_CTX was removed.
2700 * "shared" builds are now the default. To create only static libraries use
2701 the "no-shared" Configure option.
2705 * Remove the no-aes, no-hmac, no-rsa, no-sha and no-md5 Configure options.
2706 All of these option have not worked for some while and are fundamental
2711 * Make various cleanup routines no-ops and mark them as deprecated. Most
2712 global cleanup functions are no longer required because they are handled
2713 via auto-deinit (see OPENSSL_init_crypto and OPENSSL_init_ssl man pages).
2714 Explicitly de-initing can cause problems (e.g. where a library that uses
2715 OpenSSL de-inits, but an application is still using it). The affected
2716 functions are CONF_modules_free(), ENGINE_cleanup(), OBJ_cleanup(),
2717 EVP_cleanup(), BIO_sock_cleanup(), CRYPTO_cleanup_all_ex_data(),
2718 RAND_cleanup(), SSL_COMP_free_compression_methods(), ERR_free_strings() and
2719 COMP_zlib_cleanup().
2723 * --strict-warnings no longer enables runtime debugging options
2724 such as REF_DEBUG. Instead, debug options are automatically
2725 enabled with '--debug' builds.
2727 *Andy Polyakov, Emilia Käsper*
2729 * Made DH and DH_METHOD opaque. The structures for managing DH objects
2730 have been moved out of the public header files. New functions for managing
2731 these have been added.
2735 * Made RSA and RSA_METHOD opaque. The structures for managing RSA
2736 objects have been moved out of the public header files. New
2737 functions for managing these have been added.
2741 * Made DSA and DSA_METHOD opaque. The structures for managing DSA objects
2742 have been moved out of the public header files. New functions for managing
2743 these have been added.
2747 * Made BIO and BIO_METHOD opaque. The structures for managing BIOs have been
2748 moved out of the public header files. New functions for managing these
2753 * Removed no-rijndael as a config option. Rijndael is an old name for AES.
2757 * Removed the mk1mf build scripts.
2761 * Headers are now wrapped, if necessary, with OPENSSL_NO_xxx, so
2762 it is always safe to #include a header now.
2766 * Removed the aged BC-32 config and all its supporting scripts
2770 * Removed support for Ultrix, Netware, and OS/2.
2774 * Add support for HKDF.
2776 *Alessandro Ghedini*
2778 * Add support for blake2b and blake2s
2782 * Added support for "pipelining". Ciphers that have the
2783 EVP_CIPH_FLAG_PIPELINE flag set have a capability to process multiple
2784 encryptions/decryptions simultaneously. There are currently no built-in
2785 ciphers with this property but the expectation is that engines will be able
2786 to offer it to significantly improve throughput. Support has been extended
2787 into libssl so that multiple records for a single connection can be
2788 processed in one go (for >=TLS 1.1).
2792 * Added the AFALG engine. This is an async capable engine which is able to
2793 offload work to the Linux kernel. In this initial version it only supports
2794 AES128-CBC. The kernel must be version 4.1.0 or greater.
2798 * OpenSSL now uses a new threading API. It is no longer necessary to
2799 set locking callbacks to use OpenSSL in a multi-threaded environment. There
2800 are two supported threading models: pthreads and windows threads. It is
2801 also possible to configure OpenSSL at compile time for "no-threads". The
2802 old threading API should no longer be used. The functions have been
2803 replaced with "no-op" compatibility macros.
2805 *Alessandro Ghedini, Matt Caswell*
2807 * Modify behavior of ALPN to invoke callback after SNI/servername
2808 callback, such that updates to the SSL_CTX affect ALPN.
2812 * Add SSL_CIPHER queries for authentication and key-exchange.
2816 * Changes to the DEFAULT cipherlist:
2817 - Prefer (EC)DHE handshakes over plain RSA.
2818 - Prefer AEAD ciphers over legacy ciphers.
2819 - Prefer ECDSA over RSA when both certificates are available.
2820 - Prefer TLSv1.2 ciphers/PRF.
2821 - Remove DSS, SEED, IDEA, CAMELLIA, and AES-CCM from the
2826 * Change the ECC default curve list to be this, in order: x25519,
2827 secp256r1, secp521r1, secp384r1.
2831 * RC4 based libssl ciphersuites are now classed as "weak" ciphers and are
2832 disabled by default. They can be re-enabled using the
2833 enable-weak-ssl-ciphers option to Configure.
2837 * If the server has ALPN configured, but supports no protocols that the
2838 client advertises, send a fatal "no_application_protocol" alert.
2839 This behaviour is SHALL in RFC 7301, though it isn't universally
2840 implemented by other servers.
2844 * Add X25519 support.
2845 Add ASN.1 and EVP_PKEY methods for X25519. This includes support
2846 for public and private key encoding using the format documented in
2847 draft-ietf-curdle-pkix-02. The corresponding EVP_PKEY method supports
2848 key generation and key derivation.
2850 TLS support complies with draft-ietf-tls-rfc4492bis-08 and uses
2855 * Deprecate SRP_VBASE_get_by_user.
2856 SRP_VBASE_get_by_user had inconsistent memory management behaviour.
2857 In order to fix an unavoidable memory leak [CVE-2016-0798][],
2858 SRP_VBASE_get_by_user was changed to ignore the "fake user" SRP
2859 seed, even if the seed is configured.
2861 Users should use SRP_VBASE_get1_by_user instead. Note that in
2862 SRP_VBASE_get1_by_user, caller must free the returned value. Note
2863 also that even though configuring the SRP seed attempts to hide
2864 invalid usernames by continuing the handshake with fake
2865 credentials, this behaviour is not constant time and no strong
2866 guarantees are made that the handshake is indistinguishable from
2867 that of a valid user.
2871 * Configuration change; it's now possible to build dynamic engines
2872 without having to build shared libraries and vice versa. This
2873 only applies to the engines in engines/, those in crypto/engine/
2874 will always be built into libcrypto (i.e. "static").
2876 Building dynamic engines is enabled by default; to disable, use
2877 the configuration option "disable-dynamic-engine".
2879 The only requirements for building dynamic engines are the
2880 presence of the DSO module and building with position independent
2881 code, so they will also automatically be disabled if configuring
2882 with "disable-dso" or "disable-pic".
2884 The macros OPENSSL_NO_STATIC_ENGINE and OPENSSL_NO_DYNAMIC_ENGINE
2885 are also taken away from openssl/opensslconf.h, as they are
2890 * Configuration change; if there is a known flag to compile
2891 position independent code, it will always be applied on the
2892 libcrypto and libssl object files, and never on the application
2893 object files. This means other libraries that use routines from
2894 libcrypto / libssl can be made into shared libraries regardless
2895 of how OpenSSL was configured.
2897 If this isn't desirable, the configuration options "disable-pic"
2898 or "no-pic" can be used to disable the use of PIC. This will
2899 also disable building shared libraries and dynamic engines.
2903 * Removed JPAKE code. It was experimental and has no wide use.
2907 * The INSTALL_PREFIX Makefile variable has been renamed to
2908 DESTDIR. That makes for less confusion on what this variable
2909 is for. Also, the configuration option --install_prefix is
2914 * Heartbeat for TLS has been removed and is disabled by default
2915 for DTLS; configure with enable-heartbeats. Code that uses the
2916 old #define's might need to be updated.
2918 *Emilia Käsper, Rich Salz*
2920 * Rename REF_CHECK to REF_DEBUG.
2924 * New "unified" build system
2926 The "unified" build system is aimed to be a common system for all
2927 platforms we support. With it comes new support for VMS.
2929 This system builds supports building in a different directory tree
2930 than the source tree. It produces one Makefile (for unix family
2931 or lookalikes), or one descrip.mms (for VMS).
2933 The source of information to make the Makefile / descrip.mms is
2934 small files called 'build.info', holding the necessary
2935 information for each directory with source to compile, and a
2936 template in Configurations, like unix-Makefile.tmpl or
2939 With this change, the library names were also renamed on Windows
2940 and on VMS. They now have names that are closer to the standard
2941 on Unix, and include the major version number, and in certain
2942 cases, the architecture they are built for. See "Notes on shared
2943 libraries" in INSTALL.
2945 We rely heavily on the perl module Text::Template.
2949 * Added support for auto-initialisation and de-initialisation of the library.
2950 OpenSSL no longer requires explicit init or deinit routines to be called,
2951 except in certain circumstances. See the OPENSSL_init_crypto() and
2952 OPENSSL_init_ssl() man pages for further information.
2956 * The arguments to the DTLSv1_listen function have changed. Specifically the
2957 "peer" argument is now expected to be a BIO_ADDR object.
2959 * Rewrite of BIO networking library. The BIO library lacked consistent
2960 support of IPv6, and adding it required some more extensive
2961 modifications. This introduces the BIO_ADDR and BIO_ADDRINFO types,
2962 which hold all types of addresses and chains of address information.
2963 It also introduces a new API, with functions like BIO_socket,
2964 BIO_connect, BIO_listen, BIO_lookup and a rewrite of BIO_accept.
2965 The source/sink BIOs BIO_s_connect, BIO_s_accept and BIO_s_datagram
2966 have been adapted accordingly.
2970 * RSA_padding_check_PKCS1_type_1 now accepts inputs with and without
2975 * CRIME protection: disable compression by default, even if OpenSSL is
2976 compiled with zlib enabled. Applications can still enable compression
2977 by calling SSL_CTX_clear_options(ctx, SSL_OP_NO_COMPRESSION), or by
2978 using the SSL_CONF library to configure compression.
2982 * The signature of the session callback configured with
2983 SSL_CTX_sess_set_get_cb was changed. The read-only input buffer
2984 was explicitly marked as `const unsigned char*` instead of
2989 * Always DPURIFY. Remove the use of uninitialized memory in the
2990 RNG, and other conditional uses of DPURIFY. This makes -DPURIFY a no-op.
2994 * Removed many obsolete configuration items, including
2995 DES_PTR, DES_RISC1, DES_RISC2, DES_INT
2996 MD2_CHAR, MD2_INT, MD2_LONG
2998 IDEA_SHORT, IDEA_LONG
2999 RC2_SHORT, RC2_LONG, RC4_LONG, RC4_CHUNK, RC4_INDEX
3001 *Rich Salz, with advice from Andy Polyakov*
3003 * Many BN internals have been moved to an internal header file.
3005 *Rich Salz with help from Andy Polyakov*
3007 * Configuration and writing out the results from it has changed.
3008 Files such as Makefile include/openssl/opensslconf.h and are now
3009 produced through general templates, such as Makefile.in and
3010 crypto/opensslconf.h.in and some help from the perl module
3013 Also, the center of configuration information is no longer
3014 Makefile. Instead, Configure produces a perl module in
3015 configdata.pm which holds most of the config data (in the hash
3016 table %config), the target data that comes from the target
3017 configuration in one of the `Configurations/*.conf` files (in
3022 * To clarify their intended purposes, the Configure options
3023 --prefix and --openssldir change their semantics, and become more
3024 straightforward and less interdependent.
3026 --prefix shall be used exclusively to give the location INSTALLTOP
3027 where programs, scripts, libraries, include files and manuals are
3028 going to be installed. The default is now /usr/local.
3030 --openssldir shall be used exclusively to give the default
3031 location OPENSSLDIR where certificates, private keys, CRLs are
3032 managed. This is also where the default openssl.cnf gets
3034 If the directory given with this option is a relative path, the
3035 values of both the --prefix value and the --openssldir value will
3036 be combined to become OPENSSLDIR.
3037 The default for --openssldir is INSTALLTOP/ssl.
3039 Anyone who uses --openssldir to specify where OpenSSL is to be
3040 installed MUST change to use --prefix instead.
3044 * The GOST engine was out of date and therefore it has been removed. An up
3045 to date GOST engine is now being maintained in an external repository.
3046 See: <https://wiki.openssl.org/index.php/Binaries>. Libssl still retains
3047 support for GOST ciphersuites (these are only activated if a GOST engine
3052 * EGD is no longer supported by default; use enable-egd when
3055 *Ben Kaduk and Rich Salz*
3057 * The distribution now has Makefile.in files, which are used to
3058 create Makefile's when Configure is run. *Configure must be run
3059 before trying to build now.*
3063 * The return value for SSL_CIPHER_description() for error conditions
3068 * Support for RFC6698/RFC7671 DANE TLSA peer authentication.
3070 Obtaining and performing DNSSEC validation of TLSA records is
3071 the application's responsibility. The application provides
3072 the TLSA records of its choice to OpenSSL, and these are then
3073 used to authenticate the peer.
3075 The TLSA records need not even come from DNS. They can, for
3076 example, be used to implement local end-entity certificate or
3077 trust-anchor "pinning", where the "pin" data takes the form
3078 of TLSA records, which can augment or replace verification
3079 based on the usual WebPKI public certification authorities.
3083 * Revert default OPENSSL_NO_DEPRECATED setting. Instead OpenSSL
3084 continues to support deprecated interfaces in default builds.
3085 However, applications are strongly advised to compile their
3086 source files with -DOPENSSL_API_COMPAT=0x10100000L, which hides
3087 the declarations of all interfaces deprecated in 0.9.8, 1.0.0
3088 or the 1.1.0 releases.
3090 In environments in which all applications have been ported to
3091 not use any deprecated interfaces OpenSSL's Configure script
3092 should be used with the --api=1.1.0 option to entirely remove
3093 support for the deprecated features from the library and
3094 unconditionally disable them in the installed headers.
3095 Essentially the same effect can be achieved with the "no-deprecated"
3096 argument to Configure, except that this will always restrict
3097 the build to just the latest API, rather than a fixed API
3100 As applications are ported to future revisions of the API,
3101 they should update their compile-time OPENSSL_API_COMPAT define
3102 accordingly, but in most cases should be able to continue to
3103 compile with later releases.
3105 The OPENSSL_API_COMPAT versions for 1.0.0, and 0.9.8 are
3106 0x10000000L and 0x00908000L, respectively. However those
3107 versions did not support the OPENSSL_API_COMPAT feature, and
3108 so applications are not typically tested for explicit support
3109 of just the undeprecated features of either release.
3113 * Add support for setting the minimum and maximum supported protocol.
3114 It can bet set via the SSL_set_min_proto_version() and
3115 SSL_set_max_proto_version(), or via the SSL_CONF's MinProtocol and
3116 MaxProtocol. It's recommended to use the new APIs to disable
3117 protocols instead of disabling individual protocols using
3118 SSL_set_options() or SSL_CONF's Protocol. This change also
3119 removes support for disabling TLS 1.2 in the OpenSSL TLS
3120 client at compile time by defining OPENSSL_NO_TLS1_2_CLIENT.
3124 * Support for ChaCha20 and Poly1305 added to libcrypto and libssl.
3128 * New EC_KEY_METHOD, this replaces the older ECDSA_METHOD and ECDH_METHOD
3129 and integrates ECDSA and ECDH functionality into EC. Implementations can
3130 now redirect key generation and no longer need to convert to or from
3133 Note: the ecdsa.h and ecdh.h headers are now no longer needed and just
3134 include the ec.h header file instead.
3138 * Remove support for all 40 and 56 bit ciphers. This includes all the export
3139 ciphers who are no longer supported and drops support the ephemeral RSA key
3140 exchange. The LOW ciphers currently doesn't have any ciphers in it.
3144 * Made EVP_MD_CTX, EVP_MD, EVP_CIPHER_CTX, EVP_CIPHER and HMAC_CTX
3145 opaque. For HMAC_CTX, the following constructors and destructors
3148 HMAC_CTX *HMAC_CTX_new(void);
3149 void HMAC_CTX_free(HMAC_CTX *ctx);
3151 For EVP_MD and EVP_CIPHER, complete APIs to create, fill and
3152 destroy such methods has been added. See EVP_MD_meth_new(3) and
3153 EVP_CIPHER_meth_new(3) for documentation.
3156 1) `EVP_MD_CTX_cleanup()`, `EVP_CIPHER_CTX_cleanup()` and
3157 `HMAC_CTX_cleanup()` were removed. `HMAC_CTX_reset()` and
3158 `EVP_MD_CTX_reset()` should be called instead to reinitialise
3159 an already created structure.
3160 2) For consistency with the majority of our object creators and
3161 destructors, `EVP_MD_CTX_(create|destroy)` were renamed to
3162 `EVP_MD_CTX_(new|free)`. The old names are retained as macros
3163 for deprecated builds.
3167 * Added ASYNC support. Libcrypto now includes the async sub-library to enable
3168 cryptographic operations to be performed asynchronously as long as an
3169 asynchronous capable engine is used. See the ASYNC_start_job() man page for
3170 further details. Libssl has also had this capability integrated with the
3171 introduction of the new mode SSL_MODE_ASYNC and associated error
3172 SSL_ERROR_WANT_ASYNC. See the SSL_CTX_set_mode() and SSL_get_error() man
3173 pages. This work was developed in partnership with Intel Corp.
3177 * SSL_{CTX_}set_ecdh_auto() has been removed and ECDH is support is
3178 always enabled now. If you want to disable the support you should
3179 exclude it using the list of supported ciphers. This also means that the
3180 "-no_ecdhe" option has been removed from s_server.
3184 * SSL_{CTX}_set_tmp_ecdh() which can set 1 EC curve now internally calls
3185 SSL_{CTX_}set1_curves() which can set a list.
3189 * Remove support for SSL_{CTX_}set_tmp_ecdh_callback(). You should set the
3190 curve you want to support using SSL_{CTX_}set1_curves().
3194 * State machine rewrite. The state machine code has been significantly
3195 refactored in order to remove much duplication of code and solve issues
3196 with the old code (see [ssl/statem/README.md](ssl/statem/README.md) for
3197 further details). This change does have some associated API changes.
3198 Notably the SSL_state() function has been removed and replaced by
3199 SSL_get_state which now returns an "OSSL_HANDSHAKE_STATE" instead of an int.
3200 SSL_set_state() has been removed altogether. The previous handshake states
3201 defined in ssl.h and ssl3.h have also been removed.
3205 * All instances of the string "ssleay" in the public API were replaced
3206 with OpenSSL (case-matching; e.g., OPENSSL_VERSION for #define's)
3207 Some error codes related to internal RSA_eay API's were renamed.
3211 * The demo files in crypto/threads were moved to demo/threads.
3215 * Removed obsolete engines: 4758cca, aep, atalla, cswift, nuron, gmp,
3218 *Matt Caswell, Rich Salz*
3220 * New ASN.1 embed macro.
3222 New ASN.1 macro ASN1_EMBED. This is the same as ASN1_SIMPLE except the
3223 structure is not allocated: it is part of the parent. That is instead of
3231 This reduces memory fragmentation and make it impossible to accidentally
3232 set a mandatory field to NULL.
3234 This currently only works for some fields specifically a SEQUENCE, CHOICE,
3235 or ASN1_STRING type which is part of a parent SEQUENCE. Since it is
3236 equivalent to ASN1_SIMPLE it cannot be tagged, OPTIONAL, SET OF or
3241 * Remove EVP_CHECK_DES_KEY, a compile-time option that never compiled.
3245 * Removed DES and RC4 ciphersuites from DEFAULT. Also removed RC2 although
3246 in 1.0.2 EXPORT was already removed and the only RC2 ciphersuite is also
3247 an EXPORT one. COMPLEMENTOFDEFAULT has been updated accordingly to add
3248 DES and RC4 ciphersuites.
3252 * Rewrite EVP_DecodeUpdate (base64 decoding) to fix several bugs.
3253 This changes the decoding behaviour for some invalid messages,
3254 though the change is mostly in the more lenient direction, and
3255 legacy behaviour is preserved as much as possible.
3259 * Fix no-stdio build.
3260 *David Woodhouse <David.Woodhouse@intel.com> and also*
3261 *Ivan Nestlerode <ivan.nestlerode@sonos.com>*
3263 * New testing framework
3264 The testing framework has been largely rewritten and is now using
3265 perl and the perl modules Test::Harness and an extended variant of
3266 Test::More called OpenSSL::Test to do its work. All test scripts in
3267 test/ have been rewritten into test recipes, and all direct calls to
3268 executables in test/Makefile have become individual recipes using the
3269 simplified testing OpenSSL::Test::Simple.
3271 For documentation on our testing modules, do:
3273 perldoc test/testlib/OpenSSL/Test/Simple.pm
3274 perldoc test/testlib/OpenSSL/Test.pm
3278 * Revamped memory debug; only -DCRYPTO_MDEBUG and -DCRYPTO_MDEBUG_ABORT
3279 are used; the latter aborts on memory leaks (usually checked on exit).
3280 Some undocumented "set malloc, etc., hooks" functions were removed
3281 and others were changed. All are now documented.
3285 * In DSA_generate_parameters_ex, if the provided seed is too short,
3288 *Rich Salz and Ismo Puustinen <ismo.puustinen@intel.com>*
3290 * Rewrite PSK to support ECDHE_PSK, DHE_PSK and RSA_PSK. Add ciphersuites
3291 from RFC4279, RFC4785, RFC5487, RFC5489.
3293 Thanks to Christian J. Dietrich and Giuseppe D'Angelo for the
3294 original RSA_PSK patch.
3298 * Dropped support for the SSL3_FLAGS_DELAY_CLIENT_FINISHED flag. This SSLeay
3299 era flag was never set throughout the codebase (only read). Also removed
3300 SSL3_FLAGS_POP_BUFFER which was only used if
3301 SSL3_FLAGS_DELAY_CLIENT_FINISHED was also set.
3305 * Changed the default name options in the "ca", "crl", "req" and "x509"
3306 to be "oneline" instead of "compat".
3310 * Remove SSL_OP_TLS_BLOCK_PADDING_BUG. This is SSLeay legacy, we're
3311 not aware of clients that still exhibit this bug, and the workaround
3312 hasn't been working properly for a while.
3316 * The return type of BIO_number_read() and BIO_number_written() as well as
3317 the corresponding num_read and num_write members in the BIO structure has
3318 changed from unsigned long to uint64_t. On platforms where an unsigned
3319 long is 32 bits (e.g. Windows) these counters could overflow if >4Gb is
3324 * Given the pervasive nature of TLS extensions it is inadvisable to run
3325 OpenSSL without support for them. It also means that maintaining
3326 the OPENSSL_NO_TLSEXT option within the code is very invasive (and probably
3327 not well tested). Therefore the OPENSSL_NO_TLSEXT option has been removed.
3331 * Removed support for the two export grade static DH ciphersuites
3332 EXP-DH-RSA-DES-CBC-SHA and EXP-DH-DSS-DES-CBC-SHA. These two ciphersuites
3333 were newly added (along with a number of other static DH ciphersuites) to
3334 1.0.2. However the two export ones have *never* worked since they were
3335 introduced. It seems strange in any case to be adding new export
3336 ciphersuites, and given "logjam" it also does not seem correct to fix them.
3340 * Version negotiation has been rewritten. In particular SSLv23_method(),
3341 SSLv23_client_method() and SSLv23_server_method() have been deprecated,
3342 and turned into macros which simply call the new preferred function names
3343 TLS_method(), TLS_client_method() and TLS_server_method(). All new code
3344 should use the new names instead. Also as part of this change the ssl23.h
3345 header file has been removed.
3349 * Support for Kerberos ciphersuites in TLS (RFC2712) has been removed. This
3350 code and the associated standard is no longer considered fit-for-purpose.
3354 * RT2547 was closed. When generating a private key, try to make the
3355 output file readable only by the owner. This behavior change might
3356 be noticeable when interacting with other software.
3358 * Documented all exdata functions. Added CRYPTO_free_ex_index.
3363 * Added HTTP GET support to the ocsp command.
3367 * Changed default digest for the dgst and enc commands from MD5 to
3372 * RAND_pseudo_bytes has been deprecated. Users should use RAND_bytes instead.
3376 * Added support for TLS extended master secret from
3377 draft-ietf-tls-session-hash-03.txt. Thanks for Alfredo Pironti for an
3378 initial patch which was a great help during development.
3382 * All libssl internal structures have been removed from the public header
3383 files, and the OPENSSL_NO_SSL_INTERN option has been removed (since it is
3384 now redundant). Users should not attempt to access internal structures
3385 directly. Instead they should use the provided API functions.
3389 * config has been changed so that by default OPENSSL_NO_DEPRECATED is used.
3390 Access to deprecated functions can be re-enabled by running config with
3391 "enable-deprecated". In addition applications wishing to use deprecated
3392 functions must define OPENSSL_USE_DEPRECATED. Note that this new behaviour
3393 will, by default, disable some transitive includes that previously existed
3394 in the header files (e.g. ec.h will no longer, by default, include bn.h)
3398 * Added support for OCB mode. OpenSSL has been granted a patent license
3399 compatible with the OpenSSL license for use of OCB. Details are available
3400 at <https://www.openssl.org/source/OCB-patent-grant-OpenSSL.pdf>. Support
3401 for OCB can be removed by calling config with no-ocb.
3405 * SSLv2 support has been removed. It still supports receiving a SSLv2
3406 compatible client hello.
3410 * Increased the minimal RSA keysize from 256 to 512 bits [Rich Salz],
3411 done while fixing the error code for the key-too-small case.
3413 *Annie Yousar <a.yousar@informatik.hu-berlin.de>*
3415 * CA.sh has been removed; use CA.pl instead.
3419 * Removed old DES API.
3423 * Remove various unsupported platforms:
3429 Sinix/ReliantUNIX RM400
3434 16-bit platforms such as WIN16
3438 * Clean up OPENSSL_NO_xxx #define's
3439 - Use setbuf() and remove OPENSSL_NO_SETVBUF_IONBF
3440 - Rename OPENSSL_SYSNAME_xxx to OPENSSL_SYS_xxx
3441 - OPENSSL_NO_EC{DH,DSA} merged into OPENSSL_NO_EC
3442 - OPENSSL_NO_RIPEMD160, OPENSSL_NO_RIPEMD merged into OPENSSL_NO_RMD160
3443 - OPENSSL_NO_FP_API merged into OPENSSL_NO_STDIO
3444 - Remove OPENSSL_NO_BIO OPENSSL_NO_BUFFER OPENSSL_NO_CHAIN_VERIFY
3445 OPENSSL_NO_EVP OPENSSL_NO_FIPS_ERR OPENSSL_NO_HASH_COMP
3446 OPENSSL_NO_LHASH OPENSSL_NO_OBJECT OPENSSL_NO_SPEED OPENSSL_NO_STACK
3447 OPENSSL_NO_X509 OPENSSL_NO_X509_VERIFY
3448 - Remove MS_STATIC; it's a relic from platforms <32 bits.
3452 * Cleaned up dead code
3453 Remove all but one '#ifdef undef' which is to be looked at.
3457 * Clean up calling of xxx_free routines.
3458 Just like free(), fix most of the xxx_free routines to accept
3459 NULL. Remove the non-null checks from callers. Save much code.
3463 * Add secure heap for storage of private keys (when possible).
3464 Add BIO_s_secmem(), CBIGNUM, etc.
3465 Contributed by Akamai Technologies under our Corporate CLA.
3469 * Experimental support for a new, fast, unbiased prime candidate generator,
3470 bn_probable_prime_dh_coprime(). Not currently used by any prime generator.
3472 *Felix Laurie von Massenbach <felix@erbridge.co.uk>*
3474 * New output format NSS in the sess_id command line tool. This allows
3475 exporting the session id and the master key in NSS keylog format.
3477 *Martin Kaiser <martin@kaiser.cx>*
3479 * Harmonize version and its documentation. -f flag is used to display
3482 *mancha <mancha1@zoho.com>*
3484 * Fix eckey_priv_encode so it immediately returns an error upon a failure
3485 in i2d_ECPrivateKey. Thanks to Ted Unangst for feedback on this issue.
3487 *mancha <mancha1@zoho.com>*
3489 * Fix some double frees. These are not thought to be exploitable.
3491 *mancha <mancha1@zoho.com>*
3493 * A missing bounds check in the handling of the TLS heartbeat extension
3494 can be used to reveal up to 64k of memory to a connected client or
3497 Thanks for Neel Mehta of Google Security for discovering this bug and to
3498 Adam Langley <agl@chromium.org> and Bodo Moeller <bmoeller@acm.org> for
3499 preparing the fix [CVE-2014-0160][]
3501 *Adam Langley, Bodo Moeller*
3503 * Fix for the attack described in the paper "Recovering OpenSSL
3504 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
3505 by Yuval Yarom and Naomi Benger. Details can be obtained from:
3506 <http://eprint.iacr.org/2014/140>
3508 Thanks to Yuval Yarom and Naomi Benger for discovering this
3509 flaw and to Yuval Yarom for supplying a fix [CVE-2014-0076][]
3511 *Yuval Yarom and Naomi Benger*
3513 * Use algorithm specific chains in SSL_CTX_use_certificate_chain_file():
3514 this fixes a limitation in previous versions of OpenSSL.
3518 * Experimental encrypt-then-mac support.
3520 Experimental support for encrypt then mac from
3521 draft-gutmann-tls-encrypt-then-mac-02.txt
3523 To enable it set the appropriate extension number (0x42 for the test
3524 server) using e.g. -DTLSEXT_TYPE_encrypt_then_mac=0x42
3526 For non-compliant peers (i.e. just about everything) this should have no
3529 WARNING: EXPERIMENTAL, SUBJECT TO CHANGE.
3533 * Add EVP support for key wrapping algorithms, to avoid problems with
3534 existing code the flag EVP_CIPHER_CTX_WRAP_ALLOW has to be set in
3535 the EVP_CIPHER_CTX or an error is returned. Add AES and DES3 wrap
3536 algorithms and include tests cases.
3540 * Extend CMS code to support RSA-PSS signatures and RSA-OAEP for
3545 * Extended RSA OAEP support via EVP_PKEY API. Options to specify digest,
3546 MGF1 digest and OAEP label.
3550 * Make openssl verify return errors.
3552 *Chris Palmer <palmer@google.com> and Ben Laurie*
3554 * New function ASN1_TIME_diff to calculate the difference between two
3555 ASN1_TIME structures or one structure and the current time.
3559 * Update fips_test_suite to support multiple command line options. New
3560 test to induce all self test errors in sequence and check expected
3565 * Add FIPS_{rsa,dsa,ecdsa}_{sign,verify} functions which digest and
3566 sign or verify all in one operation.
3570 * Add fips_algvs: a multicall fips utility incorporating all the algorithm
3571 test programs and fips_test_suite. Includes functionality to parse
3572 the minimal script output of fipsalgest.pl directly.
3576 * Add authorisation parameter to FIPS_module_mode_set().
3580 * Add FIPS selftest for ECDH algorithm using P-224 and B-233 curves.
3584 * Use separate DRBG fields for internal and external flags. New function
3585 FIPS_drbg_health_check() to perform on demand health checking. Add
3586 generation tests to fips_test_suite with reduced health check interval to
3587 demonstrate periodic health checking. Add "nodh" option to
3588 fips_test_suite to skip very slow DH test.
3592 * New function FIPS_get_cipherbynid() to lookup FIPS supported ciphers
3597 * More extensive health check for DRBG checking many more failure modes.
3598 New function FIPS_selftest_drbg_all() to handle every possible DRBG
3599 combination: call this in fips_test_suite.
3603 * Add support for canonical generation of DSA parameter 'g'. See
3606 * Add support for HMAC DRBG from SP800-90. Update DRBG algorithm test and
3607 POST to handle HMAC cases.
3611 * Add functions FIPS_module_version() and FIPS_module_version_text()
3612 to return numerical and string versions of the FIPS module number.
3616 * Rename FIPS_mode_set and FIPS_mode to FIPS_module_mode_set and
3617 FIPS_module_mode. FIPS_mode and FIPS_mode_set will be implemented
3618 outside the validated module in the FIPS capable OpenSSL.
3622 * Minor change to DRBG entropy callback semantics. In some cases
3623 there is no multiple of the block length between min_len and
3624 max_len. Allow the callback to return more than max_len bytes
3625 of entropy but discard any extra: it is the callback's responsibility
3626 to ensure that the extra data discarded does not impact the
3627 requested amount of entropy.
3631 * Add PRNG security strength checks to RSA, DSA and ECDSA using
3632 information in FIPS186-3, SP800-57 and SP800-131A.
3636 * CCM support via EVP. Interface is very similar to GCM case except we
3637 must supply all data in one chunk (i.e. no update, final) and the
3638 message length must be supplied if AAD is used. Add algorithm test
3643 * Initial version of POST overhaul. Add POST callback to allow the status
3644 of POST to be monitored and/or failures induced. Modify fips_test_suite
3645 to use callback. Always run all selftests even if one fails.
3649 * XTS support including algorithm test driver in the fips_gcmtest program.
3650 Note: this does increase the maximum key length from 32 to 64 bytes but
3651 there should be no binary compatibility issues as existing applications
3652 will never use XTS mode.
3656 * Extensive reorganisation of FIPS PRNG behaviour. Remove all dependencies
3657 to OpenSSL RAND code and replace with a tiny FIPS RAND API which also
3658 performs algorithm blocking for unapproved PRNG types. Also do not
3659 set PRNG type in FIPS_mode_set(): leave this to the application.
3660 Add default OpenSSL DRBG handling: sets up FIPS PRNG and seeds with
3661 the standard OpenSSL PRNG: set additional data to a date time vector.
3665 * Rename old X9.31 PRNG functions of the form `FIPS_rand*` to `FIPS_x931*`.
3666 This shouldn't present any incompatibility problems because applications
3667 shouldn't be using these directly and any that are will need to rethink
3668 anyway as the X9.31 PRNG is now deprecated by FIPS 140-2
3672 * Extensive self tests and health checking required by SP800-90 DRBG.
3673 Remove strength parameter from FIPS_drbg_instantiate and always
3674 instantiate at maximum supported strength.
3678 * Add ECDH code to fips module and fips_ecdhvs for primitives only testing.
3682 * New algorithm test program fips_dhvs to handle DH primitives only testing.
3686 * New function DH_compute_key_padded() to compute a DH key and pad with
3687 leading zeroes if needed: this complies with SP800-56A et al.
3691 * Initial implementation of SP800-90 DRBGs for Hash and CTR. Not used by
3692 anything, incomplete, subject to change and largely untested at present.
3696 * Modify fipscanisteronly build option to only build the necessary object
3697 files by filtering FIPS_EX_OBJ through a perl script in crypto/Makefile.
3701 * Add experimental option FIPSSYMS to give all symbols in
3702 fipscanister.o and FIPS or fips prefix. This will avoid
3703 conflicts with future versions of OpenSSL. Add perl script
3704 util/fipsas.pl to preprocess assembly language source files
3705 and rename any affected symbols.
3709 * Add selftest checks and algorithm block of non-fips algorithms in
3710 FIPS mode. Remove DES2 from selftests.
3714 * Add ECDSA code to fips module. Add tiny fips_ecdsa_check to just
3715 return internal method without any ENGINE dependencies. Add new
3716 tiny fips sign and verify functions.
3720 * New build option no-ec2m to disable characteristic 2 code.
3724 * New build option "fipscanisteronly". This only builds fipscanister.o
3725 and (currently) associated fips utilities. Uses the file Makefile.fips
3726 instead of Makefile.org as the prototype.
3730 * Add some FIPS mode restrictions to GCM. Add internal IV generator.
3731 Update fips_gcmtest to use IV generator.
3735 * Initial, experimental EVP support for AES-GCM. AAD can be input by
3736 setting output buffer to NULL. The `*Final` function must be
3737 called although it will not retrieve any additional data. The tag
3738 can be set or retrieved with a ctrl. The IV length is by default 12
3739 bytes (96 bits) but can be set to an alternative value. If the IV
3740 length exceeds the maximum IV length (currently 16 bytes) it cannot be
3745 * New flag in ciphers: EVP_CIPH_FLAG_CUSTOM_CIPHER. This means the
3746 underlying do_cipher function handles all cipher semantics itself
3747 including padding and finalisation. This is useful if (for example)
3748 an ENGINE cipher handles block padding itself. The behaviour of
3749 do_cipher is subtly changed if this flag is set: the return value
3750 is the number of characters written to the output buffer (zero is
3751 no longer an error code) or a negative error code. Also if the
3752 input buffer is NULL and length 0 finalisation should be performed.
3756 * If a candidate issuer certificate is already part of the constructed
3757 path ignore it: new debug notification X509_V_ERR_PATH_LOOP for this case.
3761 * Improve forward-security support: add functions
3763 void SSL_CTX_set_not_resumable_session_callback(
3764 SSL_CTX *ctx, int (*cb)(SSL *ssl, int is_forward_secure))
3765 void SSL_set_not_resumable_session_callback(
3766 SSL *ssl, int (*cb)(SSL *ssl, int is_forward_secure))
3768 for use by SSL/TLS servers; the callback function will be called whenever a
3769 new session is created, and gets to decide whether the session may be
3770 cached to make it resumable (return 0) or not (return 1). (As by the
3771 SSL/TLS protocol specifications, the session_id sent by the server will be
3772 empty to indicate that the session is not resumable; also, the server will
3773 not generate RFC 4507 (RFC 5077) session tickets.)
3775 A simple reasonable callback implementation is to return is_forward_secure.
3776 This parameter will be set to 1 or 0 depending on the ciphersuite selected
3777 by the SSL/TLS server library, indicating whether it can provide forward
3780 *Emilia Käsper <emilia.kasper@esat.kuleuven.be> (Google)*
3782 * New -verify_name option in command line utilities to set verification
3787 * Initial CMAC implementation. WARNING: EXPERIMENTAL, API MAY CHANGE.
3788 Add CMAC pkey methods.
3792 * Experimental renegotiation in s_server -www mode. If the client
3793 browses /reneg connection is renegotiated. If /renegcert it is
3794 renegotiated requesting a certificate.
3798 * Add an "external" session cache for debugging purposes to s_server. This
3799 should help trace issues which normally are only apparent in deployed
3800 multi-process servers.
3804 * Extensive audit of libcrypto with DEBUG_UNUSED. Fix many cases where
3805 return value is ignored. NB. The functions RAND_add(), RAND_seed(),
3806 BIO_set_cipher() and some obscure PEM functions were changed so they
3807 can now return an error. The RAND changes required a change to the
3808 RAND_METHOD structure.
3812 * New macro `__owur` for "OpenSSL Warn Unused Result". This makes use of
3813 a gcc attribute to warn if the result of a function is ignored. This
3814 is enable if DEBUG_UNUSED is set. Add to several functions in evp.h
3815 whose return value is often ignored.
3819 * New -noct, -requestct, -requirect and -ctlogfile options for s_client.
3820 These allow SCTs (signed certificate timestamps) to be requested and
3821 validated when establishing a connection.
3823 *Rob Percival <robpercival@google.com>*
3828 ### Changes between 1.0.2s and 1.0.2t [10 Sep 2019]
3830 * For built-in EC curves, ensure an EC_GROUP built from the curve name is
3831 used even when parsing explicit parameters, when loading a encoded key
3832 or calling `EC_GROUP_new_from_ecpkparameters()`/
3833 `EC_GROUP_new_from_ecparameters()`.
3834 This prevents bypass of security hardening and performance gains,
3835 especially for curves with specialized EC_METHODs.
3836 By default, if a key encoded with explicit parameters is loaded and later
3837 encoded, the output is still encoded with explicit parameters, even if
3838 internally a "named" EC_GROUP is used for computation.
3842 * Compute ECC cofactors if not provided during EC_GROUP construction. Before
3843 this change, EC_GROUP_set_generator would accept order and/or cofactor as
3844 NULL. After this change, only the cofactor parameter can be NULL. It also
3845 does some minimal sanity checks on the passed order.
3850 * Fixed a padding oracle in PKCS7_dataDecode and CMS_decrypt_set1_pkey.
3851 An attack is simple, if the first CMS_recipientInfo is valid but the
3852 second CMS_recipientInfo is chosen ciphertext. If the second
3853 recipientInfo decodes to PKCS #1 v1.5 form plaintext, the correct
3854 encryption key will be replaced by garbage, and the message cannot be
3855 decoded, but if the RSA decryption fails, the correct encryption key is
3856 used and the recipient will not notice the attack.
3857 As a work around for this potential attack the length of the decrypted
3858 key must be equal to the cipher default key length, in case the
3859 certifiate is not given and all recipientInfo are tried out.
3860 The old behaviour can be re-enabled in the CMS code by setting the
3861 CMS_DEBUG_DECRYPT flag.
3866 * Document issue with installation paths in diverse Windows builds
3868 '/usr/local/ssl' is an unsafe prefix for location to install OpenSSL
3869 binaries and run-time config file.
3874 ### Changes between 1.0.2r and 1.0.2s [28 May 2019]
3876 * Change the default RSA, DSA and DH size to 2048 bit instead of 1024.
3877 This changes the size when using the genpkey app when no size is given. It
3878 fixes an omission in earlier changes that changed all RSA, DSA and DH
3879 generation apps to use 2048 bits by default.
3883 * Add FIPS support for Android Arm 64-bit
3885 Support for Android Arm 64-bit was added to the OpenSSL FIPS Object
3886 Module in Version 2.0.10. For some reason, the corresponding target
3887 'android64-aarch64' was missing OpenSSL 1.0.2, whence it could not be
3888 built with FIPS support on Android Arm 64-bit. This omission has been
3891 *Matthias St. Pierre*
3893 ### Changes between 1.0.2q and 1.0.2r [26 Feb 2019]
3895 * 0-byte record padding oracle
3897 If an application encounters a fatal protocol error and then calls
3898 SSL_shutdown() twice (once to send a close_notify, and once to receive one)
3899 then OpenSSL can respond differently to the calling application if a 0 byte
3900 record is received with invalid padding compared to if a 0 byte record is
3901 received with an invalid MAC. If the application then behaves differently
3902 based on that in a way that is detectable to the remote peer, then this
3903 amounts to a padding oracle that could be used to decrypt data.
3905 In order for this to be exploitable "non-stitched" ciphersuites must be in
3906 use. Stitched ciphersuites are optimised implementations of certain
3907 commonly used ciphersuites. Also the application must call SSL_shutdown()
3908 twice even if a protocol error has occurred (applications should not do
3909 this but some do anyway).
3911 This issue was discovered by Juraj Somorovsky, Robert Merget and Nimrod
3912 Aviram, with additional investigation by Steven Collison and Andrew
3913 Hourselt. It was reported to OpenSSL on 10th December 2018.
3918 * Move strictness check from EVP_PKEY_asn1_new() to EVP_PKEY_asn1_add0().
3922 ### Changes between 1.0.2p and 1.0.2q [20 Nov 2018]
3924 * Microarchitecture timing vulnerability in ECC scalar multiplication
3926 OpenSSL ECC scalar multiplication, used in e.g. ECDSA and ECDH, has been
3927 shown to be vulnerable to a microarchitecture timing side channel attack.
3928 An attacker with sufficient access to mount local timing attacks during
3929 ECDSA signature generation could recover the private key.
3931 This issue was reported to OpenSSL on 26th October 2018 by Alejandro
3932 Cabrera Aldaya, Billy Brumley, Sohaib ul Hassan, Cesar Pereida Garcia and
3938 * Timing vulnerability in DSA signature generation
3940 The OpenSSL DSA signature algorithm has been shown to be vulnerable to a
3941 timing side channel attack. An attacker could use variations in the signing
3942 algorithm to recover the private key.
3944 This issue was reported to OpenSSL on 16th October 2018 by Samuel Weiser.
3949 * Resolve a compatibility issue in EC_GROUP handling with the FIPS Object
3950 Module, accidentally introduced while backporting security fixes from the
3951 development branch and hindering the use of ECC in FIPS mode.
3955 ### Changes between 1.0.2o and 1.0.2p [14 Aug 2018]
3957 * Client DoS due to large DH parameter
3959 During key agreement in a TLS handshake using a DH(E) based ciphersuite a
3960 malicious server can send a very large prime value to the client. This will
3961 cause the client to spend an unreasonably long period of time generating a
3962 key for this prime resulting in a hang until the client has finished. This
3963 could be exploited in a Denial Of Service attack.
3965 This issue was reported to OpenSSL on 5th June 2018 by Guido Vranken
3970 * Cache timing vulnerability in RSA Key Generation
3972 The OpenSSL RSA Key generation algorithm has been shown to be vulnerable to
3973 a cache timing side channel attack. An attacker with sufficient access to
3974 mount cache timing attacks during the RSA key generation process could
3975 recover the private key.
3977 This issue was reported to OpenSSL on 4th April 2018 by Alejandro Cabrera
3978 Aldaya, Billy Brumley, Cesar Pereida Garcia and Luis Manuel Alvarez Tapia.
3983 * Make EVP_PKEY_asn1_new() a bit stricter about its input. A NULL pem_str
3984 parameter is no longer accepted, as it leads to a corrupt table. NULL
3985 pem_str is reserved for alias entries only.
3989 * Revert blinding in ECDSA sign and instead make problematic addition
3990 length-invariant. Switch even to fixed-length Montgomery multiplication.
3994 * Change generating and checking of primes so that the error rate of not
3995 being prime depends on the intended use based on the size of the input.
3996 For larger primes this will result in more rounds of Miller-Rabin.
3997 The maximal error rate for primes with more than 1080 bits is lowered
4000 *Kurt Roeckx, Annie Yousar*
4002 * Increase the number of Miller-Rabin rounds for DSA key generating to 64.
4006 * Add blinding to ECDSA and DSA signatures to protect against side channel
4007 attacks discovered by Keegan Ryan (NCC Group).
4011 * When unlocking a pass phrase protected PEM file or PKCS#8 container, we
4012 now allow empty (zero character) pass phrases.
4016 * Certificate time validation (X509_cmp_time) enforces stricter
4017 compliance with RFC 5280. Fractional seconds and timezone offsets
4018 are no longer allowed.
4022 ### Changes between 1.0.2n and 1.0.2o [27 Mar 2018]
4024 * Constructed ASN.1 types with a recursive definition could exceed the stack
4026 Constructed ASN.1 types with a recursive definition (such as can be found
4027 in PKCS7) could eventually exceed the stack given malicious input with
4028 excessive recursion. This could result in a Denial Of Service attack. There
4029 are no such structures used within SSL/TLS that come from untrusted sources
4030 so this is considered safe.
4032 This issue was reported to OpenSSL on 4th January 2018 by the OSS-fuzz
4038 ### Changes between 1.0.2m and 1.0.2n [7 Dec 2017]
4040 * Read/write after SSL object in error state
4042 OpenSSL 1.0.2 (starting from version 1.0.2b) introduced an "error state"
4043 mechanism. The intent was that if a fatal error occurred during a handshake
4044 then OpenSSL would move into the error state and would immediately fail if
4045 you attempted to continue the handshake. This works as designed for the
4046 explicit handshake functions (SSL_do_handshake(), SSL_accept() and
4047 SSL_connect()), however due to a bug it does not work correctly if
4048 SSL_read() or SSL_write() is called directly. In that scenario, if the
4049 handshake fails then a fatal error will be returned in the initial function
4050 call. If SSL_read()/SSL_write() is subsequently called by the application
4051 for the same SSL object then it will succeed and the data is passed without
4052 being decrypted/encrypted directly from the SSL/TLS record layer.
4054 In order to exploit this issue an application bug would have to be present
4055 that resulted in a call to SSL_read()/SSL_write() being issued after having
4056 already received a fatal error.
4058 This issue was reported to OpenSSL by David Benjamin (Google).
4063 * rsaz_1024_mul_avx2 overflow bug on x86_64
4065 There is an overflow bug in the AVX2 Montgomery multiplication procedure
4066 used in exponentiation with 1024-bit moduli. No EC algorithms are affected.
4067 Analysis suggests that attacks against RSA and DSA as a result of this
4068 defect would be very difficult to perform and are not believed likely.
4069 Attacks against DH1024 are considered just feasible, because most of the
4070 work necessary to deduce information about a private key may be performed
4071 offline. The amount of resources required for such an attack would be
4072 significant. However, for an attack on TLS to be meaningful, the server
4073 would have to share the DH1024 private key among multiple clients, which is
4074 no longer an option since CVE-2016-0701.
4076 This only affects processors that support the AVX2 but not ADX extensions
4077 like Intel Haswell (4th generation).
4079 This issue was reported to OpenSSL by David Benjamin (Google). The issue
4080 was originally found via the OSS-Fuzz project.
4085 ### Changes between 1.0.2l and 1.0.2m [2 Nov 2017]
4087 * bn_sqrx8x_internal carry bug on x86_64
4089 There is a carry propagating bug in the x86_64 Montgomery squaring
4090 procedure. No EC algorithms are affected. Analysis suggests that attacks
4091 against RSA and DSA as a result of this defect would be very difficult to
4092 perform and are not believed likely. Attacks against DH are considered just
4093 feasible (although very difficult) because most of the work necessary to
4094 deduce information about a private key may be performed offline. The amount
4095 of resources required for such an attack would be very significant and
4096 likely only accessible to a limited number of attackers. An attacker would
4097 additionally need online access to an unpatched system using the target
4098 private key in a scenario with persistent DH parameters and a private
4099 key that is shared between multiple clients.
4101 This only affects processors that support the BMI1, BMI2 and ADX extensions
4102 like Intel Broadwell (5th generation) and later or AMD Ryzen.
4104 This issue was reported to OpenSSL by the OSS-Fuzz project.
4109 * Malformed X.509 IPAddressFamily could cause OOB read
4111 If an X.509 certificate has a malformed IPAddressFamily extension,
4112 OpenSSL could do a one-byte buffer overread. The most likely result
4113 would be an erroneous display of the certificate in text format.
4115 This issue was reported to OpenSSL by the OSS-Fuzz project.
4120 ### Changes between 1.0.2k and 1.0.2l [25 May 2017]
4122 * Have 'config' recognise 64-bit mingw and choose 'mingw64' as the target
4123 platform rather than 'mingw'.
4127 ### Changes between 1.0.2j and 1.0.2k [26 Jan 2017]
4129 * Truncated packet could crash via OOB read
4131 If one side of an SSL/TLS path is running on a 32-bit host and a specific
4132 cipher is being used, then a truncated packet can cause that host to
4133 perform an out-of-bounds read, usually resulting in a crash.
4135 This issue was reported to OpenSSL by Robert Święcki of Google.
4140 * BN_mod_exp may produce incorrect results on x86_64
4142 There is a carry propagating bug in the x86_64 Montgomery squaring
4143 procedure. No EC algorithms are affected. Analysis suggests that attacks
4144 against RSA and DSA as a result of this defect would be very difficult to
4145 perform and are not believed likely. Attacks against DH are considered just
4146 feasible (although very difficult) because most of the work necessary to
4147 deduce information about a private key may be performed offline. The amount
4148 of resources required for such an attack would be very significant and
4149 likely only accessible to a limited number of attackers. An attacker would
4150 additionally need online access to an unpatched system using the target
4151 private key in a scenario with persistent DH parameters and a private
4152 key that is shared between multiple clients. For example this can occur by
4153 default in OpenSSL DHE based SSL/TLS ciphersuites. Note: This issue is very
4154 similar to CVE-2015-3193 but must be treated as a separate problem.
4156 This issue was reported to OpenSSL by the OSS-Fuzz project.
4161 * Montgomery multiplication may produce incorrect results
4163 There is a carry propagating bug in the Broadwell-specific Montgomery
4164 multiplication procedure that handles input lengths divisible by, but
4165 longer than 256 bits. Analysis suggests that attacks against RSA, DSA
4166 and DH private keys are impossible. This is because the subroutine in
4167 question is not used in operations with the private key itself and an input
4168 of the attacker's direct choice. Otherwise the bug can manifest itself as
4169 transient authentication and key negotiation failures or reproducible
4170 erroneous outcome of public-key operations with specially crafted input.
4171 Among EC algorithms only Brainpool P-512 curves are affected and one
4172 presumably can attack ECDH key negotiation. Impact was not analyzed in
4173 detail, because pre-requisites for attack are considered unlikely. Namely
4174 multiple clients have to choose the curve in question and the server has to
4175 share the private key among them, neither of which is default behaviour.
4176 Even then only clients that chose the curve will be affected.
4178 This issue was publicly reported as transient failures and was not
4179 initially recognized as a security issue. Thanks to Richard Morgan for
4180 providing reproducible case.
4185 * OpenSSL now fails if it receives an unrecognised record type in TLS1.0
4186 or TLS1.1. Previously this only happened in SSLv3 and TLS1.2. This is to
4187 prevent issues where no progress is being made and the peer continually
4188 sends unrecognised record types, using up resources processing them.
4192 ### Changes between 1.0.2i and 1.0.2j [26 Sep 2016]
4194 * Missing CRL sanity check
4196 A bug fix which included a CRL sanity check was added to OpenSSL 1.1.0
4197 but was omitted from OpenSSL 1.0.2i. As a result any attempt to use
4198 CRLs in OpenSSL 1.0.2i will crash with a null pointer exception.
4200 This issue only affects the OpenSSL 1.0.2i
4205 ### Changes between 1.0.2h and 1.0.2i [22 Sep 2016]
4207 * OCSP Status Request extension unbounded memory growth
4209 A malicious client can send an excessively large OCSP Status Request
4210 extension. If that client continually requests renegotiation, sending a
4211 large OCSP Status Request extension each time, then there will be unbounded
4212 memory growth on the server. This will eventually lead to a Denial Of
4213 Service attack through memory exhaustion. Servers with a default
4214 configuration are vulnerable even if they do not support OCSP. Builds using
4215 the "no-ocsp" build time option are not affected.
4217 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4222 * In order to mitigate the SWEET32 attack, the DES ciphers were moved from
4225 This issue was reported to OpenSSL Karthikeyan Bhargavan and Gaetan
4231 * OOB write in MDC2_Update()
4233 An overflow can occur in MDC2_Update() either if called directly or
4234 through the EVP_DigestUpdate() function using MDC2. If an attacker
4235 is able to supply very large amounts of input data after a previous
4236 call to EVP_EncryptUpdate() with a partial block then a length check
4237 can overflow resulting in a heap corruption.
4239 The amount of data needed is comparable to SIZE_MAX which is impractical
4242 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4247 * Malformed SHA512 ticket DoS
4249 If a server uses SHA512 for TLS session ticket HMAC it is vulnerable to a
4250 DoS attack where a malformed ticket will result in an OOB read which will
4253 The use of SHA512 in TLS session tickets is comparatively rare as it requires
4254 a custom server callback and ticket lookup mechanism.
4256 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4261 * OOB write in BN_bn2dec()
4263 The function BN_bn2dec() does not check the return value of BN_div_word().
4264 This can cause an OOB write if an application uses this function with an
4265 overly large BIGNUM. This could be a problem if an overly large certificate
4266 or CRL is printed out from an untrusted source. TLS is not affected because
4267 record limits will reject an oversized certificate before it is parsed.
4269 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4274 * OOB read in TS_OBJ_print_bio()
4276 The function TS_OBJ_print_bio() misuses OBJ_obj2txt(): the return value is
4277 the total length the OID text representation would use and not the amount
4278 of data written. This will result in OOB reads when large OIDs are
4281 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4286 * Pointer arithmetic undefined behaviour
4288 Avoid some undefined pointer arithmetic
4290 A common idiom in the codebase is to check limits in the following manner:
4293 Where "p" points to some malloc'd data of SIZE bytes and
4296 "len" here could be from some externally supplied data (e.g. from a TLS
4299 The rules of C pointer arithmetic are such that "p + len" is only well
4300 defined where len <= SIZE. Therefore the above idiom is actually
4301 undefined behaviour.
4303 For example this could cause problems if some malloc implementation
4304 provides an address for "p" such that "p + len" actually overflows for
4305 values of len that are too big and therefore p + len < limit.
4307 This issue was reported to OpenSSL by Guido Vranken
4312 * Constant time flag not preserved in DSA signing
4314 Operations in the DSA signing algorithm should run in constant time in
4315 order to avoid side channel attacks. A flaw in the OpenSSL DSA
4316 implementation means that a non-constant time codepath is followed for
4317 certain operations. This has been demonstrated through a cache-timing
4318 attack to be sufficient for an attacker to recover the private DSA key.
4320 This issue was reported by César Pereida (Aalto University), Billy Brumley
4321 (Tampere University of Technology), and Yuval Yarom (The University of
4322 Adelaide and NICTA).
4327 * DTLS buffered message DoS
4329 In a DTLS connection where handshake messages are delivered out-of-order
4330 those messages that OpenSSL is not yet ready to process will be buffered
4331 for later use. Under certain circumstances, a flaw in the logic means that
4332 those messages do not get removed from the buffer even though the handshake
4333 has been completed. An attacker could force up to approx. 15 messages to
4334 remain in the buffer when they are no longer required. These messages will
4335 be cleared when the DTLS connection is closed. The default maximum size for
4336 a message is 100k. Therefore the attacker could force an additional 1500k
4337 to be consumed per connection. By opening many simulataneous connections an
4338 attacker could cause a DoS attack through memory exhaustion.
4340 This issue was reported to OpenSSL by Quan Luo.
4345 * DTLS replay protection DoS
4347 A flaw in the DTLS replay attack protection mechanism means that records
4348 that arrive for future epochs update the replay protection "window" before
4349 the MAC for the record has been validated. This could be exploited by an
4350 attacker by sending a record for the next epoch (which does not have to
4351 decrypt or have a valid MAC), with a very large sequence number. This means
4352 that all subsequent legitimate packets are dropped causing a denial of
4353 service for a specific DTLS connection.
4355 This issue was reported to OpenSSL by the OCAP audit team.
4360 * Certificate message OOB reads
4362 In OpenSSL 1.0.2 and earlier some missing message length checks can result
4363 in OOB reads of up to 2 bytes beyond an allocated buffer. There is a
4364 theoretical DoS risk but this has not been observed in practice on common
4367 The messages affected are client certificate, client certificate request
4368 and server certificate. As a result the attack can only be performed
4369 against a client or a server which enables client authentication.
4371 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4376 ### Changes between 1.0.2g and 1.0.2h [3 May 2016]
4378 * Prevent padding oracle in AES-NI CBC MAC check
4380 A MITM attacker can use a padding oracle attack to decrypt traffic
4381 when the connection uses an AES CBC cipher and the server support
4384 This issue was introduced as part of the fix for Lucky 13 padding
4385 attack [CVE-2013-0169][]. The padding check was rewritten to be in
4386 constant time by making sure that always the same bytes are read and
4387 compared against either the MAC or padding bytes. But it no longer
4388 checked that there was enough data to have both the MAC and padding
4391 This issue was reported by Juraj Somorovsky using TLS-Attacker.
4396 * Fix EVP_EncodeUpdate overflow
4398 An overflow can occur in the EVP_EncodeUpdate() function which is used for
4399 Base64 encoding of binary data. If an attacker is able to supply very large
4400 amounts of input data then a length check can overflow resulting in a heap
4403 Internally to OpenSSL the EVP_EncodeUpdate() function is primarily used by
4404 the `PEM_write_bio*` family of functions. These are mainly used within the
4405 OpenSSL command line applications, so any application which processes data
4406 from an untrusted source and outputs it as a PEM file should be considered
4407 vulnerable to this issue. User applications that call these APIs directly
4408 with large amounts of untrusted data may also be vulnerable.
4410 This issue was reported by Guido Vranken.
4415 * Fix EVP_EncryptUpdate overflow
4417 An overflow can occur in the EVP_EncryptUpdate() function. If an attacker
4418 is able to supply very large amounts of input data after a previous call to
4419 EVP_EncryptUpdate() with a partial block then a length check can overflow
4420 resulting in a heap corruption. Following an analysis of all OpenSSL
4421 internal usage of the EVP_EncryptUpdate() function all usage is one of two
4422 forms. The first form is where the EVP_EncryptUpdate() call is known to be
4423 the first called function after an EVP_EncryptInit(), and therefore that
4424 specific call must be safe. The second form is where the length passed to
4425 EVP_EncryptUpdate() can be seen from the code to be some small value and
4426 therefore there is no possibility of an overflow. Since all instances are
4427 one of these two forms, it is believed that there can be no overflows in
4428 internal code due to this problem. It should be noted that
4429 EVP_DecryptUpdate() can call EVP_EncryptUpdate() in certain code paths.
4430 Also EVP_CipherUpdate() is a synonym for EVP_EncryptUpdate(). All instances
4431 of these calls have also been analysed too and it is believed there are no
4432 instances in internal usage where an overflow could occur.
4434 This issue was reported by Guido Vranken.
4439 * Prevent ASN.1 BIO excessive memory allocation
4441 When ASN.1 data is read from a BIO using functions such as d2i_CMS_bio()
4442 a short invalid encoding can cause allocation of large amounts of memory
4443 potentially consuming excessive resources or exhausting memory.
4445 Any application parsing untrusted data through d2i BIO functions is
4446 affected. The memory based functions such as d2i_X509() are *not* affected.
4447 Since the memory based functions are used by the TLS library, TLS
4448 applications are not affected.
4450 This issue was reported by Brian Carpenter.
4457 ASN1 Strings that are over 1024 bytes can cause an overread in applications
4458 using the X509_NAME_oneline() function on EBCDIC systems. This could result
4459 in arbitrary stack data being returned in the buffer.
4461 This issue was reported by Guido Vranken.
4466 * Modify behavior of ALPN to invoke callback after SNI/servername
4467 callback, such that updates to the SSL_CTX affect ALPN.
4471 * Remove LOW from the DEFAULT cipher list. This removes singles DES from the
4476 * Only remove the SSLv2 methods with the no-ssl2-method option. When the
4477 methods are enabled and ssl2 is disabled the methods return NULL.
4481 ### Changes between 1.0.2f and 1.0.2g [1 Mar 2016]
4483 * Disable weak ciphers in SSLv3 and up in default builds of OpenSSL.
4484 Builds that are not configured with "enable-weak-ssl-ciphers" will not
4485 provide any "EXPORT" or "LOW" strength ciphers.
4489 * Disable SSLv2 default build, default negotiation and weak ciphers. SSLv2
4490 is by default disabled at build-time. Builds that are not configured with
4491 "enable-ssl2" will not support SSLv2. Even if "enable-ssl2" is used,
4492 users who want to negotiate SSLv2 via the version-flexible SSLv23_method()
4493 will need to explicitly call either of:
4495 SSL_CTX_clear_options(ctx, SSL_OP_NO_SSLv2);
4497 SSL_clear_options(ssl, SSL_OP_NO_SSLv2);
4499 as appropriate. Even if either of those is used, or the application
4500 explicitly uses the version-specific SSLv2_method() or its client and
4501 server variants, SSLv2 ciphers vulnerable to exhaustive search key
4502 recovery have been removed. Specifically, the SSLv2 40-bit EXPORT
4503 ciphers, and SSLv2 56-bit DES are no longer available.
4508 * Fix a double-free in DSA code
4510 A double free bug was discovered when OpenSSL parses malformed DSA private
4511 keys and could lead to a DoS attack or memory corruption for applications
4512 that receive DSA private keys from untrusted sources. This scenario is
4515 This issue was reported to OpenSSL by Adam Langley(Google/BoringSSL) using
4521 * Disable SRP fake user seed to address a server memory leak.
4523 Add a new method SRP_VBASE_get1_by_user that handles the seed properly.
4525 SRP_VBASE_get_by_user had inconsistent memory management behaviour.
4526 In order to fix an unavoidable memory leak, SRP_VBASE_get_by_user
4527 was changed to ignore the "fake user" SRP seed, even if the seed
4530 Users should use SRP_VBASE_get1_by_user instead. Note that in
4531 SRP_VBASE_get1_by_user, caller must free the returned value. Note
4532 also that even though configuring the SRP seed attempts to hide
4533 invalid usernames by continuing the handshake with fake
4534 credentials, this behaviour is not constant time and no strong
4535 guarantees are made that the handshake is indistinguishable from
4536 that of a valid user.
4541 * Fix BN_hex2bn/BN_dec2bn NULL pointer deref/heap corruption
4543 In the BN_hex2bn function the number of hex digits is calculated using an
4544 int value `i`. Later `bn_expand` is called with a value of `i * 4`. For
4545 large values of `i` this can result in `bn_expand` not allocating any
4546 memory because `i * 4` is negative. This can leave the internal BIGNUM data
4547 field as NULL leading to a subsequent NULL ptr deref. For very large values
4548 of `i`, the calculation `i * 4` could be a positive value smaller than `i`.
4549 In this case memory is allocated to the internal BIGNUM data field, but it
4550 is insufficiently sized leading to heap corruption. A similar issue exists
4551 in BN_dec2bn. This could have security consequences if BN_hex2bn/BN_dec2bn
4552 is ever called by user applications with very large untrusted hex/dec data.
4553 This is anticipated to be a rare occurrence.
4555 All OpenSSL internal usage of these functions use data that is not expected
4556 to be untrusted, e.g. config file data or application command line
4557 arguments. If user developed applications generate config file data based
4558 on untrusted data then it is possible that this could also lead to security
4559 consequences. This is also anticipated to be rare.
4561 This issue was reported to OpenSSL by Guido Vranken.
4566 * Fix memory issues in `BIO_*printf` functions
4568 The internal `fmtstr` function used in processing a "%s" format string in
4569 the `BIO_*printf` functions could overflow while calculating the length of a
4570 string and cause an OOB read when printing very long strings.
4572 Additionally the internal `doapr_outch` function can attempt to write to an
4573 OOB memory location (at an offset from the NULL pointer) in the event of a
4574 memory allocation failure. In 1.0.2 and below this could be caused where
4575 the size of a buffer to be allocated is greater than INT_MAX. E.g. this
4576 could be in processing a very long "%s" format string. Memory leaks can
4579 The first issue may mask the second issue dependent on compiler behaviour.
4580 These problems could enable attacks where large amounts of untrusted data
4581 is passed to the `BIO_*printf` functions. If applications use these functions
4582 in this way then they could be vulnerable. OpenSSL itself uses these
4583 functions when printing out human-readable dumps of ASN.1 data. Therefore
4584 applications that print this data could be vulnerable if the data is from
4585 untrusted sources. OpenSSL command line applications could also be
4586 vulnerable where they print out ASN.1 data, or if untrusted data is passed
4587 as command line arguments.
4589 Libssl is not considered directly vulnerable. Additionally certificates etc
4590 received via remote connections via libssl are also unlikely to be able to
4591 trigger these issues because of message size limits enforced within libssl.
4593 This issue was reported to OpenSSL Guido Vranken.
4598 * Side channel attack on modular exponentiation
4600 A side-channel attack was found which makes use of cache-bank conflicts on
4601 the Intel Sandy-Bridge microarchitecture which could lead to the recovery
4602 of RSA keys. The ability to exploit this issue is limited as it relies on
4603 an attacker who has control of code in a thread running on the same
4604 hyper-threaded core as the victim thread which is performing decryptions.
4606 This issue was reported to OpenSSL by Yuval Yarom, The University of
4607 Adelaide and NICTA, Daniel Genkin, Technion and Tel Aviv University, and
4608 Nadia Heninger, University of Pennsylvania with more information at
4609 <http://cachebleed.info>.
4614 * Change the req app to generate a 2048-bit RSA/DSA key by default,
4615 if no keysize is specified with default_bits. This fixes an
4616 omission in an earlier change that changed all RSA/DSA key generation
4617 apps to use 2048 bits by default.
4621 ### Changes between 1.0.2e and 1.0.2f [28 Jan 2016]
4623 * DH small subgroups
4625 Historically OpenSSL only ever generated DH parameters based on "safe"
4626 primes. More recently (in version 1.0.2) support was provided for
4627 generating X9.42 style parameter files such as those required for RFC 5114
4628 support. The primes used in such files may not be "safe". Where an
4629 application is using DH configured with parameters based on primes that are
4630 not "safe" then an attacker could use this fact to find a peer's private
4631 DH exponent. This attack requires that the attacker complete multiple
4632 handshakes in which the peer uses the same private DH exponent. For example
4633 this could be used to discover a TLS server's private DH exponent if it's
4634 reusing the private DH exponent or it's using a static DH ciphersuite.
4636 OpenSSL provides the option SSL_OP_SINGLE_DH_USE for ephemeral DH (DHE) in
4637 TLS. It is not on by default. If the option is not set then the server
4638 reuses the same private DH exponent for the life of the server process and
4639 would be vulnerable to this attack. It is believed that many popular
4640 applications do set this option and would therefore not be at risk.
4642 The fix for this issue adds an additional check where a "q" parameter is
4643 available (as is the case in X9.42 based parameters). This detects the
4644 only known attack, and is the only possible defense for static DH
4645 ciphersuites. This could have some performance impact.
4647 Additionally the SSL_OP_SINGLE_DH_USE option has been switched on by
4648 default and cannot be disabled. This could have some performance impact.
4650 This issue was reported to OpenSSL by Antonio Sanso (Adobe).
4655 * SSLv2 doesn't block disabled ciphers
4657 A malicious client can negotiate SSLv2 ciphers that have been disabled on
4658 the server and complete SSLv2 handshakes even if all SSLv2 ciphers have
4659 been disabled, provided that the SSLv2 protocol was not also disabled via
4662 This issue was reported to OpenSSL on 26th December 2015 by Nimrod Aviram
4663 and Sebastian Schinzel.
4668 ### Changes between 1.0.2d and 1.0.2e [3 Dec 2015]
4670 * BN_mod_exp may produce incorrect results on x86_64
4672 There is a carry propagating bug in the x86_64 Montgomery squaring
4673 procedure. No EC algorithms are affected. Analysis suggests that attacks
4674 against RSA and DSA as a result of this defect would be very difficult to
4675 perform and are not believed likely. Attacks against DH are considered just
4676 feasible (although very difficult) because most of the work necessary to
4677 deduce information about a private key may be performed offline. The amount
4678 of resources required for such an attack would be very significant and
4679 likely only accessible to a limited number of attackers. An attacker would
4680 additionally need online access to an unpatched system using the target
4681 private key in a scenario with persistent DH parameters and a private
4682 key that is shared between multiple clients. For example this can occur by
4683 default in OpenSSL DHE based SSL/TLS ciphersuites.
4685 This issue was reported to OpenSSL by Hanno Böck.
4690 * Certificate verify crash with missing PSS parameter
4692 The signature verification routines will crash with a NULL pointer
4693 dereference if presented with an ASN.1 signature using the RSA PSS
4694 algorithm and absent mask generation function parameter. Since these
4695 routines are used to verify certificate signature algorithms this can be
4696 used to crash any certificate verification operation and exploited in a
4697 DoS attack. Any application which performs certificate verification is
4698 vulnerable including OpenSSL clients and servers which enable client
4701 This issue was reported to OpenSSL by Loïc Jonas Etienne (Qnective AG).
4706 * X509_ATTRIBUTE memory leak
4708 When presented with a malformed X509_ATTRIBUTE structure OpenSSL will leak
4709 memory. This structure is used by the PKCS#7 and CMS routines so any
4710 application which reads PKCS#7 or CMS data from untrusted sources is
4711 affected. SSL/TLS is not affected.
4713 This issue was reported to OpenSSL by Adam Langley (Google/BoringSSL) using
4719 * Rewrite EVP_DecodeUpdate (base64 decoding) to fix several bugs.
4720 This changes the decoding behaviour for some invalid messages,
4721 though the change is mostly in the more lenient direction, and
4722 legacy behaviour is preserved as much as possible.
4726 * In DSA_generate_parameters_ex, if the provided seed is too short,
4729 *Rich Salz and Ismo Puustinen <ismo.puustinen@intel.com>*
4731 ### Changes between 1.0.2c and 1.0.2d [9 Jul 2015]
4733 * Alternate chains certificate forgery
4735 During certificate verification, OpenSSL will attempt to find an
4736 alternative certificate chain if the first attempt to build such a chain
4737 fails. An error in the implementation of this logic can mean that an
4738 attacker could cause certain checks on untrusted certificates to be
4739 bypassed, such as the CA flag, enabling them to use a valid leaf
4740 certificate to act as a CA and "issue" an invalid certificate.
4742 This issue was reported to OpenSSL by Adam Langley/David Benjamin
4747 ### Changes between 1.0.2b and 1.0.2c [12 Jun 2015]
4749 * Fix HMAC ABI incompatibility. The previous version introduced an ABI
4750 incompatibility in the handling of HMAC. The previous ABI has now been
4755 ### Changes between 1.0.2a and 1.0.2b [11 Jun 2015]
4757 * Malformed ECParameters causes infinite loop
4759 When processing an ECParameters structure OpenSSL enters an infinite loop
4760 if the curve specified is over a specially malformed binary polynomial
4763 This can be used to perform denial of service against any
4764 system which processes public keys, certificate requests or
4765 certificates. This includes TLS clients and TLS servers with
4766 client authentication enabled.
4768 This issue was reported to OpenSSL by Joseph Barr-Pixton.
4773 * Exploitable out-of-bounds read in X509_cmp_time
4775 X509_cmp_time does not properly check the length of the ASN1_TIME
4776 string and can read a few bytes out of bounds. In addition,
4777 X509_cmp_time accepts an arbitrary number of fractional seconds in the
4780 An attacker can use this to craft malformed certificates and CRLs of
4781 various sizes and potentially cause a segmentation fault, resulting in
4782 a DoS on applications that verify certificates or CRLs. TLS clients
4783 that verify CRLs are affected. TLS clients and servers with client
4784 authentication enabled may be affected if they use custom verification
4787 This issue was reported to OpenSSL by Robert Swiecki (Google), and
4788 independently by Hanno Böck.
4793 * PKCS7 crash with missing EnvelopedContent
4795 The PKCS#7 parsing code does not handle missing inner EncryptedContent
4796 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs
4797 with missing content and trigger a NULL pointer dereference on parsing.
4799 Applications that decrypt PKCS#7 data or otherwise parse PKCS#7
4800 structures from untrusted sources are affected. OpenSSL clients and
4801 servers are not affected.
4803 This issue was reported to OpenSSL by Michal Zalewski (Google).
4808 * CMS verify infinite loop with unknown hash function
4810 When verifying a signedData message the CMS code can enter an infinite loop
4811 if presented with an unknown hash function OID. This can be used to perform
4812 denial of service against any system which verifies signedData messages using
4814 This issue was reported to OpenSSL by Johannes Bauer.
4819 * Race condition handling NewSessionTicket
4821 If a NewSessionTicket is received by a multi-threaded client when attempting to
4822 reuse a previous ticket then a race condition can occur potentially leading to
4823 a double free of the ticket data.
4828 * Only support 256-bit or stronger elliptic curves with the
4829 'ecdh_auto' setting (server) or by default (client). Of supported
4830 curves, prefer P-256 (both).
4834 ### Changes between 1.0.2 and 1.0.2a [19 Mar 2015]
4836 * ClientHello sigalgs DoS fix
4838 If a client connects to an OpenSSL 1.0.2 server and renegotiates with an
4839 invalid signature algorithms extension a NULL pointer dereference will
4840 occur. This can be exploited in a DoS attack against the server.
4842 This issue was was reported to OpenSSL by David Ramos of Stanford
4846 *Stephen Henson and Matt Caswell*
4848 * Multiblock corrupted pointer fix
4850 OpenSSL 1.0.2 introduced the "multiblock" performance improvement. This
4851 feature only applies on 64 bit x86 architecture platforms that support AES
4852 NI instructions. A defect in the implementation of "multiblock" can cause
4853 OpenSSL's internal write buffer to become incorrectly set to NULL when
4854 using non-blocking IO. Typically, when the user application is using a
4855 socket BIO for writing, this will only result in a failed connection.
4856 However if some other BIO is used then it is likely that a segmentation
4857 fault will be triggered, thus enabling a potential DoS attack.
4859 This issue was reported to OpenSSL by Daniel Danner and Rainer Mueller.
4864 * Segmentation fault in DTLSv1_listen fix
4866 The DTLSv1_listen function is intended to be stateless and processes the
4867 initial ClientHello from many peers. It is common for user code to loop
4868 over the call to DTLSv1_listen until a valid ClientHello is received with
4869 an associated cookie. A defect in the implementation of DTLSv1_listen means
4870 that state is preserved in the SSL object from one invocation to the next
4871 that can lead to a segmentation fault. Errors processing the initial
4872 ClientHello can trigger this scenario. An example of such an error could be
4873 that a DTLS1.0 only client is attempting to connect to a DTLS1.2 only
4876 This issue was reported to OpenSSL by Per Allansson.
4881 * Segmentation fault in ASN1_TYPE_cmp fix
4883 The function ASN1_TYPE_cmp will crash with an invalid read if an attempt is
4884 made to compare ASN.1 boolean types. Since ASN1_TYPE_cmp is used to check
4885 certificate signature algorithm consistency this can be used to crash any
4886 certificate verification operation and exploited in a DoS attack. Any
4887 application which performs certificate verification is vulnerable including
4888 OpenSSL clients and servers which enable client authentication.
4893 * Segmentation fault for invalid PSS parameters fix
4895 The signature verification routines will crash with a NULL pointer
4896 dereference if presented with an ASN.1 signature using the RSA PSS
4897 algorithm and invalid parameters. Since these routines are used to verify
4898 certificate signature algorithms this can be used to crash any
4899 certificate verification operation and exploited in a DoS attack. Any
4900 application which performs certificate verification is vulnerable including
4901 OpenSSL clients and servers which enable client authentication.
4903 This issue was was reported to OpenSSL by Brian Carpenter.
4908 * ASN.1 structure reuse memory corruption fix
4910 Reusing a structure in ASN.1 parsing may allow an attacker to cause
4911 memory corruption via an invalid write. Such reuse is and has been
4912 strongly discouraged and is believed to be rare.
4914 Applications that parse structures containing CHOICE or ANY DEFINED BY
4915 components may be affected. Certificate parsing (d2i_X509 and related
4916 functions) are however not affected. OpenSSL clients and servers are
4922 * PKCS7 NULL pointer dereferences fix
4924 The PKCS#7 parsing code does not handle missing outer ContentInfo
4925 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs with
4926 missing content and trigger a NULL pointer dereference on parsing.
4928 Applications that verify PKCS#7 signatures, decrypt PKCS#7 data or
4929 otherwise parse PKCS#7 structures from untrusted sources are
4930 affected. OpenSSL clients and servers are not affected.
4932 This issue was reported to OpenSSL by Michal Zalewski (Google).
4937 * DoS via reachable assert in SSLv2 servers fix
4939 A malicious client can trigger an OPENSSL_assert (i.e., an abort) in
4940 servers that both support SSLv2 and enable export cipher suites by sending
4941 a specially crafted SSLv2 CLIENT-MASTER-KEY message.
4943 This issue was discovered by Sean Burford (Google) and Emilia Käsper
4944 (OpenSSL development team).
4949 * Empty CKE with client auth and DHE fix
4951 If client auth is used then a server can seg fault in the event of a DHE
4952 ciphersuite being selected and a zero length ClientKeyExchange message
4953 being sent by the client. This could be exploited in a DoS attack.
4958 * Handshake with unseeded PRNG fix
4960 Under certain conditions an OpenSSL 1.0.2 client can complete a handshake
4961 with an unseeded PRNG. The conditions are:
4962 - The client is on a platform where the PRNG has not been seeded
4963 automatically, and the user has not seeded manually
4964 - A protocol specific client method version has been used (i.e. not
4965 SSL_client_methodv23)
4966 - A ciphersuite is used that does not require additional random data from
4967 the PRNG beyond the initial ClientHello client random (e.g. PSK-RC4-SHA).
4969 If the handshake succeeds then the client random that has been used will
4970 have been generated from a PRNG with insufficient entropy and therefore the
4971 output may be predictable.
4973 For example using the following command with an unseeded openssl will
4974 succeed on an unpatched platform:
4976 openssl s_client -psk 1a2b3c4d -tls1_2 -cipher PSK-RC4-SHA
4981 * Use After Free following d2i_ECPrivatekey error fix
4983 A malformed EC private key file consumed via the d2i_ECPrivateKey function
4984 could cause a use after free condition. This, in turn, could cause a double
4985 free in several private key parsing functions (such as d2i_PrivateKey
4986 or EVP_PKCS82PKEY) and could lead to a DoS attack or memory corruption
4987 for applications that receive EC private keys from untrusted
4988 sources. This scenario is considered rare.
4990 This issue was discovered by the BoringSSL project and fixed in their
4996 * X509_to_X509_REQ NULL pointer deref fix
4998 The function X509_to_X509_REQ will crash with a NULL pointer dereference if
4999 the certificate key is invalid. This function is rarely used in practice.
5001 This issue was discovered by Brian Carpenter.
5006 * Removed the export ciphers from the DEFAULT ciphers
5010 ### Changes between 1.0.1l and 1.0.2 [22 Jan 2015]
5012 * Facilitate "universal" ARM builds targeting range of ARM ISAs, e.g.
5013 ARMv5 through ARMv8, as opposite to "locking" it to single one.
5014 So far those who have to target multiple platforms would compromise
5015 and argue that binary targeting say ARMv5 would still execute on
5016 ARMv8. "Universal" build resolves this compromise by providing
5017 near-optimal performance even on newer platforms.
5021 * Accelerated NIST P-256 elliptic curve implementation for x86_64
5022 (other platforms pending).
5024 *Shay Gueron & Vlad Krasnov (Intel Corp), Andy Polyakov*
5026 * Add support for the SignedCertificateTimestampList certificate and
5027 OCSP response extensions from RFC6962.
5031 * Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
5032 for corner cases. (Certain input points at infinity could lead to
5033 bogus results, with non-infinity inputs mapped to infinity too.)
5037 * Initial support for PowerISA 2.0.7, first implemented in POWER8.
5038 This covers AES, SHA256/512 and GHASH. "Initial" means that most
5039 common cases are optimized and there still is room for further
5040 improvements. Vector Permutation AES for Altivec is also added.
5044 * Add support for little-endian ppc64 Linux target.
5046 *Marcelo Cerri (IBM)*
5048 * Initial support for AMRv8 ISA crypto extensions. This covers AES,
5049 SHA1, SHA256 and GHASH. "Initial" means that most common cases
5050 are optimized and there still is room for further improvements.
5051 Both 32- and 64-bit modes are supported.
5053 *Andy Polyakov, Ard Biesheuvel (Linaro)*
5055 * Improved ARMv7 NEON support.
5059 * Support for SPARC Architecture 2011 crypto extensions, first
5060 implemented in SPARC T4. This covers AES, DES, Camellia, SHA1,
5061 SHA256/512, MD5, GHASH and modular exponentiation.
5063 *Andy Polyakov, David Miller*
5065 * Accelerated modular exponentiation for Intel processors, a.k.a.
5068 *Shay Gueron & Vlad Krasnov (Intel Corp)*
5070 * Support for new and upcoming Intel processors, including AVX2,
5071 BMI and SHA ISA extensions. This includes additional "stitched"
5072 implementations, AESNI-SHA256 and GCM, and multi-buffer support
5075 This work was sponsored by Intel Corp.
5079 * Support for DTLS 1.2. This adds two sets of DTLS methods: DTLS_*_method()
5080 supports both DTLS 1.2 and 1.0 and should use whatever version the peer
5081 supports and DTLSv1_2_*_method() which supports DTLS 1.2 only.
5085 * Use algorithm specific chains in SSL_CTX_use_certificate_chain_file():
5086 this fixes a limitation in previous versions of OpenSSL.
5090 * Extended RSA OAEP support via EVP_PKEY API. Options to specify digest,
5091 MGF1 digest and OAEP label.
5095 * Add EVP support for key wrapping algorithms, to avoid problems with
5096 existing code the flag EVP_CIPHER_CTX_WRAP_ALLOW has to be set in
5097 the EVP_CIPHER_CTX or an error is returned. Add AES and DES3 wrap
5098 algorithms and include tests cases.
5102 * Add functions to allocate and set the fields of an ECDSA_METHOD
5105 *Douglas E. Engert, Steve Henson*
5107 * New functions OPENSSL_gmtime_diff and ASN1_TIME_diff to find the
5108 difference in days and seconds between two tm or ASN1_TIME structures.
5112 * Add -rev test option to s_server to just reverse order of characters
5113 received by client and send back to server. Also prints an abbreviated
5114 summary of the connection parameters.
5118 * New option -brief for s_client and s_server to print out a brief summary
5119 of connection parameters.
5123 * Add callbacks for arbitrary TLS extensions.
5125 *Trevor Perrin <trevp@trevp.net> and Ben Laurie*
5127 * New option -crl_download in several openssl utilities to download CRLs
5128 from CRLDP extension in certificates.
5132 * New options -CRL and -CRLform for s_client and s_server for CRLs.
5136 * New function X509_CRL_diff to generate a delta CRL from the difference
5137 of two full CRLs. Add support to "crl" utility.
5141 * New functions to set lookup_crls function and to retrieve
5142 X509_STORE from X509_STORE_CTX.
5146 * Print out deprecated issuer and subject unique ID fields in
5151 * Extend OCSP I/O functions so they can be used for simple general purpose
5152 HTTP as well as OCSP. New wrapper function which can be used to download
5153 CRLs using the OCSP API.
5157 * Delegate command line handling in s_client/s_server to SSL_CONF APIs.
5161 * `SSL_CONF*` functions. These provide a common framework for application
5162 configuration using configuration files or command lines.
5166 * SSL/TLS tracing code. This parses out SSL/TLS records using the
5167 message callback and prints the results. Needs compile time option
5168 "enable-ssl-trace". New options to s_client and s_server to enable
5173 * New ctrl and macro to retrieve supported points extensions.
5174 Print out extension in s_server and s_client.
5178 * New functions to retrieve certificate signature and signature
5183 * Add functions to retrieve and manipulate the raw cipherlist sent by a
5188 * New Suite B modes for TLS code. These use and enforce the requirements
5189 of RFC6460: restrict ciphersuites, only permit Suite B algorithms and
5190 only use Suite B curves. The Suite B modes can be set by using the
5191 strings "SUITEB128", "SUITEB192" or "SUITEB128ONLY" for the cipherstring.
5195 * New chain verification flags for Suite B levels of security. Check
5196 algorithms are acceptable when flags are set in X509_verify_cert.
5200 * Make tls1_check_chain return a set of flags indicating checks passed
5201 by a certificate chain. Add additional tests to handle client
5202 certificates: checks for matching certificate type and issuer name
5207 * If an attempt is made to use a signature algorithm not in the peer
5208 preference list abort the handshake. If client has no suitable
5209 signature algorithms in response to a certificate request do not
5210 use the certificate.
5214 * If server EC tmp key is not in client preference list abort handshake.
5218 * Add support for certificate stores in CERT structure. This makes it
5219 possible to have different stores per SSL structure or one store in
5220 the parent SSL_CTX. Include distinct stores for certificate chain
5221 verification and chain building. New ctrl SSL_CTRL_BUILD_CERT_CHAIN
5222 to build and store a certificate chain in CERT structure: returning
5223 an error if the chain cannot be built: this will allow applications
5224 to test if a chain is correctly configured.
5226 Note: if the CERT based stores are not set then the parent SSL_CTX
5227 store is used to retain compatibility with existing behaviour.
5231 * New function ssl_set_client_disabled to set a ciphersuite disabled
5232 mask based on the current session, check mask when sending client
5233 hello and checking the requested ciphersuite.
5237 * New ctrls to retrieve and set certificate types in a certificate
5238 request message. Print out received values in s_client. If certificate
5239 types is not set with custom values set sensible values based on
5240 supported signature algorithms.
5244 * Support for distinct client and server supported signature algorithms.
5248 * Add certificate callback. If set this is called whenever a certificate
5249 is required by client or server. An application can decide which
5250 certificate chain to present based on arbitrary criteria: for example
5251 supported signature algorithms. Add very simple example to s_server.
5252 This fixes many of the problems and restrictions of the existing client
5253 certificate callback: for example you can now clear an existing
5254 certificate and specify the whole chain.
5258 * Add new "valid_flags" field to CERT_PKEY structure which determines what
5259 the certificate can be used for (if anything). Set valid_flags field
5260 in new tls1_check_chain function. Simplify ssl_set_cert_masks which used
5261 to have similar checks in it.
5263 Add new "cert_flags" field to CERT structure and include a "strict mode".
5264 This enforces some TLS certificate requirements (such as only permitting
5265 certificate signature algorithms contained in the supported algorithms
5266 extension) which some implementations ignore: this option should be used
5267 with caution as it could cause interoperability issues.
5271 * Update and tidy signature algorithm extension processing. Work out
5272 shared signature algorithms based on preferences and peer algorithms
5273 and print them out in s_client and s_server. Abort handshake if no
5274 shared signature algorithms.
5278 * Add new functions to allow customised supported signature algorithms
5279 for SSL and SSL_CTX structures. Add options to s_client and s_server
5284 * New function SSL_certs_clear() to delete all references to certificates
5285 from an SSL structure. Before this once a certificate had been added
5286 it couldn't be removed.
5290 * Integrate hostname, email address and IP address checking with certificate
5291 verification. New verify options supporting checking in openssl utility.
5295 * Fixes and wildcard matching support to hostname and email checking
5296 functions. Add manual page.
5298 *Florian Weimer (Red Hat Product Security Team)*
5300 * New functions to check a hostname email or IP address against a
5301 certificate. Add options x509 utility to print results of checks against
5306 * Fix OCSP checking.
5308 *Rob Stradling <rob.stradling@comodo.com> and Ben Laurie*
5310 * Initial experimental support for explicitly trusted non-root CAs.
5311 OpenSSL still tries to build a complete chain to a root but if an
5312 intermediate CA has a trust setting included that is used. The first
5313 setting is used: whether to trust (e.g., -addtrust option to the x509
5318 * Add -trusted_first option which attempts to find certificates in the
5319 trusted store even if an untrusted chain is also supplied.
5323 * MIPS assembly pack updates: support for MIPS32r2 and SmartMIPS ASE,
5324 platform support for Linux and Android.
5328 * Support for linux-x32, ILP32 environment in x86_64 framework.
5332 * Experimental multi-implementation support for FIPS capable OpenSSL.
5333 When in FIPS mode the approved implementations are used as normal,
5334 when not in FIPS mode the internal unapproved versions are used instead.
5335 This means that the FIPS capable OpenSSL isn't forced to use the
5336 (often lower performance) FIPS implementations outside FIPS mode.
5340 * Transparently support X9.42 DH parameters when calling
5341 PEM_read_bio_DHparameters. This means existing applications can handle
5342 the new parameter format automatically.
5346 * Initial experimental support for X9.42 DH parameter format: mainly
5347 to support use of 'q' parameter for RFC5114 parameters.
5351 * Add DH parameters from RFC5114 including test data to dhtest.
5355 * Support for automatic EC temporary key parameter selection. If enabled
5356 the most preferred EC parameters are automatically used instead of
5357 hardcoded fixed parameters. Now a server just has to call:
5358 SSL_CTX_set_ecdh_auto(ctx, 1) and the server will automatically
5359 support ECDH and use the most appropriate parameters.
5363 * Enhance and tidy EC curve and point format TLS extension code. Use
5364 static structures instead of allocation if default values are used.
5365 New ctrls to set curves we wish to support and to retrieve shared curves.
5366 Print out shared curves in s_server. New options to s_server and s_client
5367 to set list of supported curves.
5371 * New ctrls to retrieve supported signature algorithms and
5372 supported curve values as an array of NIDs. Extend openssl utility
5373 to print out received values.
5377 * Add new APIs EC_curve_nist2nid and EC_curve_nid2nist which convert
5378 between NIDs and the more common NIST names such as "P-256". Enhance
5379 ecparam utility and ECC method to recognise the NIST names for curves.
5383 * Enhance SSL/TLS certificate chain handling to support different
5384 chains for each certificate instead of one chain in the parent SSL_CTX.
5388 * Support for fixed DH ciphersuite client authentication: where both
5389 server and client use DH certificates with common parameters.
5393 * Support for fixed DH ciphersuites: those requiring DH server
5398 * New function i2d_re_X509_tbs for re-encoding the TBS portion of
5400 Note: Related 1.0.2-beta specific macros X509_get_cert_info,
5401 X509_CINF_set_modified, X509_CINF_get_issuer, X509_CINF_get_extensions and
5402 X509_CINF_get_signature were reverted post internal team review.
5407 ### Changes between 1.0.1t and 1.0.1u [22 Sep 2016]
5409 * OCSP Status Request extension unbounded memory growth
5411 A malicious client can send an excessively large OCSP Status Request
5412 extension. If that client continually requests renegotiation, sending a
5413 large OCSP Status Request extension each time, then there will be unbounded
5414 memory growth on the server. This will eventually lead to a Denial Of
5415 Service attack through memory exhaustion. Servers with a default
5416 configuration are vulnerable even if they do not support OCSP. Builds using
5417 the "no-ocsp" build time option are not affected.
5419 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5424 * In order to mitigate the SWEET32 attack, the DES ciphers were moved from
5427 This issue was reported to OpenSSL Karthikeyan Bhargavan and Gaetan
5433 * OOB write in MDC2_Update()
5435 An overflow can occur in MDC2_Update() either if called directly or
5436 through the EVP_DigestUpdate() function using MDC2. If an attacker
5437 is able to supply very large amounts of input data after a previous
5438 call to EVP_EncryptUpdate() with a partial block then a length check
5439 can overflow resulting in a heap corruption.
5441 The amount of data needed is comparable to SIZE_MAX which is impractical
5444 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5449 * Malformed SHA512 ticket DoS
5451 If a server uses SHA512 for TLS session ticket HMAC it is vulnerable to a
5452 DoS attack where a malformed ticket will result in an OOB read which will
5455 The use of SHA512 in TLS session tickets is comparatively rare as it requires
5456 a custom server callback and ticket lookup mechanism.
5458 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5463 * OOB write in BN_bn2dec()
5465 The function BN_bn2dec() does not check the return value of BN_div_word().
5466 This can cause an OOB write if an application uses this function with an
5467 overly large BIGNUM. This could be a problem if an overly large certificate
5468 or CRL is printed out from an untrusted source. TLS is not affected because
5469 record limits will reject an oversized certificate before it is parsed.
5471 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5476 * OOB read in TS_OBJ_print_bio()
5478 The function TS_OBJ_print_bio() misuses OBJ_obj2txt(): the return value is
5479 the total length the OID text representation would use and not the amount
5480 of data written. This will result in OOB reads when large OIDs are
5483 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5488 * Pointer arithmetic undefined behaviour
5490 Avoid some undefined pointer arithmetic
5492 A common idiom in the codebase is to check limits in the following manner:
5495 Where "p" points to some malloc'd data of SIZE bytes and
5498 "len" here could be from some externally supplied data (e.g. from a TLS
5501 The rules of C pointer arithmetic are such that "p + len" is only well
5502 defined where len <= SIZE. Therefore the above idiom is actually
5503 undefined behaviour.
5505 For example this could cause problems if some malloc implementation
5506 provides an address for "p" such that "p + len" actually overflows for
5507 values of len that are too big and therefore p + len < limit.
5509 This issue was reported to OpenSSL by Guido Vranken
5514 * Constant time flag not preserved in DSA signing
5516 Operations in the DSA signing algorithm should run in constant time in
5517 order to avoid side channel attacks. A flaw in the OpenSSL DSA
5518 implementation means that a non-constant time codepath is followed for
5519 certain operations. This has been demonstrated through a cache-timing
5520 attack to be sufficient for an attacker to recover the private DSA key.
5522 This issue was reported by César Pereida (Aalto University), Billy Brumley
5523 (Tampere University of Technology), and Yuval Yarom (The University of
5524 Adelaide and NICTA).
5529 * DTLS buffered message DoS
5531 In a DTLS connection where handshake messages are delivered out-of-order
5532 those messages that OpenSSL is not yet ready to process will be buffered
5533 for later use. Under certain circumstances, a flaw in the logic means that
5534 those messages do not get removed from the buffer even though the handshake
5535 has been completed. An attacker could force up to approx. 15 messages to
5536 remain in the buffer when they are no longer required. These messages will
5537 be cleared when the DTLS connection is closed. The default maximum size for
5538 a message is 100k. Therefore the attacker could force an additional 1500k
5539 to be consumed per connection. By opening many simulataneous connections an
5540 attacker could cause a DoS attack through memory exhaustion.
5542 This issue was reported to OpenSSL by Quan Luo.
5547 * DTLS replay protection DoS
5549 A flaw in the DTLS replay attack protection mechanism means that records
5550 that arrive for future epochs update the replay protection "window" before
5551 the MAC for the record has been validated. This could be exploited by an
5552 attacker by sending a record for the next epoch (which does not have to
5553 decrypt or have a valid MAC), with a very large sequence number. This means
5554 that all subsequent legitimate packets are dropped causing a denial of
5555 service for a specific DTLS connection.
5557 This issue was reported to OpenSSL by the OCAP audit team.
5562 * Certificate message OOB reads
5564 In OpenSSL 1.0.2 and earlier some missing message length checks can result
5565 in OOB reads of up to 2 bytes beyond an allocated buffer. There is a
5566 theoretical DoS risk but this has not been observed in practice on common
5569 The messages affected are client certificate, client certificate request
5570 and server certificate. As a result the attack can only be performed
5571 against a client or a server which enables client authentication.
5573 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5578 ### Changes between 1.0.1s and 1.0.1t [3 May 2016]
5580 * Prevent padding oracle in AES-NI CBC MAC check
5582 A MITM attacker can use a padding oracle attack to decrypt traffic
5583 when the connection uses an AES CBC cipher and the server support
5586 This issue was introduced as part of the fix for Lucky 13 padding
5587 attack [CVE-2013-0169][]. The padding check was rewritten to be in
5588 constant time by making sure that always the same bytes are read and
5589 compared against either the MAC or padding bytes. But it no longer
5590 checked that there was enough data to have both the MAC and padding
5593 This issue was reported by Juraj Somorovsky using TLS-Attacker.
5598 * Fix EVP_EncodeUpdate overflow
5600 An overflow can occur in the EVP_EncodeUpdate() function which is used for
5601 Base64 encoding of binary data. If an attacker is able to supply very large
5602 amounts of input data then a length check can overflow resulting in a heap
5605 Internally to OpenSSL the EVP_EncodeUpdate() function is primarly used by
5606 the `PEM_write_bio*` family of functions. These are mainly used within the
5607 OpenSSL command line applications, so any application which processes data
5608 from an untrusted source and outputs it as a PEM file should be considered
5609 vulnerable to this issue. User applications that call these APIs directly
5610 with large amounts of untrusted data may also be vulnerable.
5612 This issue was reported by Guido Vranken.
5617 * Fix EVP_EncryptUpdate overflow
5619 An overflow can occur in the EVP_EncryptUpdate() function. If an attacker
5620 is able to supply very large amounts of input data after a previous call to
5621 EVP_EncryptUpdate() with a partial block then a length check can overflow
5622 resulting in a heap corruption. Following an analysis of all OpenSSL
5623 internal usage of the EVP_EncryptUpdate() function all usage is one of two
5624 forms. The first form is where the EVP_EncryptUpdate() call is known to be
5625 the first called function after an EVP_EncryptInit(), and therefore that
5626 specific call must be safe. The second form is where the length passed to
5627 EVP_EncryptUpdate() can be seen from the code to be some small value and
5628 therefore there is no possibility of an overflow. Since all instances are
5629 one of these two forms, it is believed that there can be no overflows in
5630 internal code due to this problem. It should be noted that
5631 EVP_DecryptUpdate() can call EVP_EncryptUpdate() in certain code paths.
5632 Also EVP_CipherUpdate() is a synonym for EVP_EncryptUpdate(). All instances
5633 of these calls have also been analysed too and it is believed there are no
5634 instances in internal usage where an overflow could occur.
5636 This issue was reported by Guido Vranken.
5641 * Prevent ASN.1 BIO excessive memory allocation
5643 When ASN.1 data is read from a BIO using functions such as d2i_CMS_bio()
5644 a short invalid encoding can casuse allocation of large amounts of memory
5645 potentially consuming excessive resources or exhausting memory.
5647 Any application parsing untrusted data through d2i BIO functions is
5648 affected. The memory based functions such as d2i_X509() are *not* affected.
5649 Since the memory based functions are used by the TLS library, TLS
5650 applications are not affected.
5652 This issue was reported by Brian Carpenter.
5659 ASN1 Strings that are over 1024 bytes can cause an overread in applications
5660 using the X509_NAME_oneline() function on EBCDIC systems. This could result
5661 in arbitrary stack data being returned in the buffer.
5663 This issue was reported by Guido Vranken.
5668 * Modify behavior of ALPN to invoke callback after SNI/servername
5669 callback, such that updates to the SSL_CTX affect ALPN.
5673 * Remove LOW from the DEFAULT cipher list. This removes singles DES from the
5678 * Only remove the SSLv2 methods with the no-ssl2-method option. When the
5679 methods are enabled and ssl2 is disabled the methods return NULL.
5683 ### Changes between 1.0.1r and 1.0.1s [1 Mar 2016]
5685 * Disable weak ciphers in SSLv3 and up in default builds of OpenSSL.
5686 Builds that are not configured with "enable-weak-ssl-ciphers" will not
5687 provide any "EXPORT" or "LOW" strength ciphers.
5691 * Disable SSLv2 default build, default negotiation and weak ciphers. SSLv2
5692 is by default disabled at build-time. Builds that are not configured with
5693 "enable-ssl2" will not support SSLv2. Even if "enable-ssl2" is used,
5694 users who want to negotiate SSLv2 via the version-flexible SSLv23_method()
5695 will need to explicitly call either of:
5697 SSL_CTX_clear_options(ctx, SSL_OP_NO_SSLv2);
5699 SSL_clear_options(ssl, SSL_OP_NO_SSLv2);
5701 as appropriate. Even if either of those is used, or the application
5702 explicitly uses the version-specific SSLv2_method() or its client and
5703 server variants, SSLv2 ciphers vulnerable to exhaustive search key
5704 recovery have been removed. Specifically, the SSLv2 40-bit EXPORT
5705 ciphers, and SSLv2 56-bit DES are no longer available.
5710 * Fix a double-free in DSA code
5712 A double free bug was discovered when OpenSSL parses malformed DSA private
5713 keys and could lead to a DoS attack or memory corruption for applications
5714 that receive DSA private keys from untrusted sources. This scenario is
5717 This issue was reported to OpenSSL by Adam Langley(Google/BoringSSL) using
5723 * Disable SRP fake user seed to address a server memory leak.
5725 Add a new method SRP_VBASE_get1_by_user that handles the seed properly.
5727 SRP_VBASE_get_by_user had inconsistent memory management behaviour.
5728 In order to fix an unavoidable memory leak, SRP_VBASE_get_by_user
5729 was changed to ignore the "fake user" SRP seed, even if the seed
5732 Users should use SRP_VBASE_get1_by_user instead. Note that in
5733 SRP_VBASE_get1_by_user, caller must free the returned value. Note
5734 also that even though configuring the SRP seed attempts to hide
5735 invalid usernames by continuing the handshake with fake
5736 credentials, this behaviour is not constant time and no strong
5737 guarantees are made that the handshake is indistinguishable from
5738 that of a valid user.
5743 * Fix BN_hex2bn/BN_dec2bn NULL pointer deref/heap corruption
5745 In the BN_hex2bn function the number of hex digits is calculated using an
5746 int value `i`. Later `bn_expand` is called with a value of `i * 4`. For
5747 large values of `i` this can result in `bn_expand` not allocating any
5748 memory because `i * 4` is negative. This can leave the internal BIGNUM data
5749 field as NULL leading to a subsequent NULL ptr deref. For very large values
5750 of `i`, the calculation `i * 4` could be a positive value smaller than `i`.
5751 In this case memory is allocated to the internal BIGNUM data field, but it
5752 is insufficiently sized leading to heap corruption. A similar issue exists
5753 in BN_dec2bn. This could have security consequences if BN_hex2bn/BN_dec2bn
5754 is ever called by user applications with very large untrusted hex/dec data.
5755 This is anticipated to be a rare occurrence.
5757 All OpenSSL internal usage of these functions use data that is not expected
5758 to be untrusted, e.g. config file data or application command line
5759 arguments. If user developed applications generate config file data based
5760 on untrusted data then it is possible that this could also lead to security
5761 consequences. This is also anticipated to be rare.
5763 This issue was reported to OpenSSL by Guido Vranken.
5768 * Fix memory issues in `BIO_*printf` functions
5770 The internal `fmtstr` function used in processing a "%s" format string in
5771 the `BIO_*printf` functions could overflow while calculating the length of a
5772 string and cause an OOB read when printing very long strings.
5774 Additionally the internal `doapr_outch` function can attempt to write to an
5775 OOB memory location (at an offset from the NULL pointer) in the event of a
5776 memory allocation failure. In 1.0.2 and below this could be caused where
5777 the size of a buffer to be allocated is greater than INT_MAX. E.g. this
5778 could be in processing a very long "%s" format string. Memory leaks can
5781 The first issue may mask the second issue dependent on compiler behaviour.
5782 These problems could enable attacks where large amounts of untrusted data
5783 is passed to the `BIO_*printf` functions. If applications use these functions
5784 in this way then they could be vulnerable. OpenSSL itself uses these
5785 functions when printing out human-readable dumps of ASN.1 data. Therefore
5786 applications that print this data could be vulnerable if the data is from
5787 untrusted sources. OpenSSL command line applications could also be
5788 vulnerable where they print out ASN.1 data, or if untrusted data is passed
5789 as command line arguments.
5791 Libssl is not considered directly vulnerable. Additionally certificates etc
5792 received via remote connections via libssl are also unlikely to be able to
5793 trigger these issues because of message size limits enforced within libssl.
5795 This issue was reported to OpenSSL Guido Vranken.
5800 * Side channel attack on modular exponentiation
5802 A side-channel attack was found which makes use of cache-bank conflicts on
5803 the Intel Sandy-Bridge microarchitecture which could lead to the recovery
5804 of RSA keys. The ability to exploit this issue is limited as it relies on
5805 an attacker who has control of code in a thread running on the same
5806 hyper-threaded core as the victim thread which is performing decryptions.
5808 This issue was reported to OpenSSL by Yuval Yarom, The University of
5809 Adelaide and NICTA, Daniel Genkin, Technion and Tel Aviv University, and
5810 Nadia Heninger, University of Pennsylvania with more information at
5811 <http://cachebleed.info>.
5816 * Change the req app to generate a 2048-bit RSA/DSA key by default,
5817 if no keysize is specified with default_bits. This fixes an
5818 omission in an earlier change that changed all RSA/DSA key generation
5819 apps to use 2048 bits by default.
5823 ### Changes between 1.0.1q and 1.0.1r [28 Jan 2016]
5825 * Protection for DH small subgroup attacks
5827 As a precautionary measure the SSL_OP_SINGLE_DH_USE option has been
5828 switched on by default and cannot be disabled. This could have some
5833 * SSLv2 doesn't block disabled ciphers
5835 A malicious client can negotiate SSLv2 ciphers that have been disabled on
5836 the server and complete SSLv2 handshakes even if all SSLv2 ciphers have
5837 been disabled, provided that the SSLv2 protocol was not also disabled via
5840 This issue was reported to OpenSSL on 26th December 2015 by Nimrod Aviram
5841 and Sebastian Schinzel.
5846 * Reject DH handshakes with parameters shorter than 1024 bits.
5850 ### Changes between 1.0.1p and 1.0.1q [3 Dec 2015]
5852 * Certificate verify crash with missing PSS parameter
5854 The signature verification routines will crash with a NULL pointer
5855 dereference if presented with an ASN.1 signature using the RSA PSS
5856 algorithm and absent mask generation function parameter. Since these
5857 routines are used to verify certificate signature algorithms this can be
5858 used to crash any certificate verification operation and exploited in a
5859 DoS attack. Any application which performs certificate verification is
5860 vulnerable including OpenSSL clients and servers which enable client
5863 This issue was reported to OpenSSL by Loïc Jonas Etienne (Qnective AG).
5868 * X509_ATTRIBUTE memory leak
5870 When presented with a malformed X509_ATTRIBUTE structure OpenSSL will leak
5871 memory. This structure is used by the PKCS#7 and CMS routines so any
5872 application which reads PKCS#7 or CMS data from untrusted sources is
5873 affected. SSL/TLS is not affected.
5875 This issue was reported to OpenSSL by Adam Langley (Google/BoringSSL) using
5881 * Rewrite EVP_DecodeUpdate (base64 decoding) to fix several bugs.
5882 This changes the decoding behaviour for some invalid messages,
5883 though the change is mostly in the more lenient direction, and
5884 legacy behaviour is preserved as much as possible.
5888 * In DSA_generate_parameters_ex, if the provided seed is too short,
5889 use a random seed, as already documented.
5891 *Rich Salz and Ismo Puustinen <ismo.puustinen@intel.com>*
5893 ### Changes between 1.0.1o and 1.0.1p [9 Jul 2015]
5895 * Alternate chains certificate forgery
5897 During certificate verfification, OpenSSL will attempt to find an
5898 alternative certificate chain if the first attempt to build such a chain
5899 fails. An error in the implementation of this logic can mean that an
5900 attacker could cause certain checks on untrusted certificates to be
5901 bypassed, such as the CA flag, enabling them to use a valid leaf
5902 certificate to act as a CA and "issue" an invalid certificate.
5904 This issue was reported to OpenSSL by Adam Langley/David Benjamin
5910 * Race condition handling PSK identify hint
5912 If PSK identity hints are received by a multi-threaded client then
5913 the values are wrongly updated in the parent SSL_CTX structure. This can
5914 result in a race condition potentially leading to a double free of the
5920 ### Changes between 1.0.1n and 1.0.1o [12 Jun 2015]
5922 * Fix HMAC ABI incompatibility. The previous version introduced an ABI
5923 incompatibility in the handling of HMAC. The previous ABI has now been
5926 ### Changes between 1.0.1m and 1.0.1n [11 Jun 2015]
5928 * Malformed ECParameters causes infinite loop
5930 When processing an ECParameters structure OpenSSL enters an infinite loop
5931 if the curve specified is over a specially malformed binary polynomial
5934 This can be used to perform denial of service against any
5935 system which processes public keys, certificate requests or
5936 certificates. This includes TLS clients and TLS servers with
5937 client authentication enabled.
5939 This issue was reported to OpenSSL by Joseph Barr-Pixton.
5944 * Exploitable out-of-bounds read in X509_cmp_time
5946 X509_cmp_time does not properly check the length of the ASN1_TIME
5947 string and can read a few bytes out of bounds. In addition,
5948 X509_cmp_time accepts an arbitrary number of fractional seconds in the
5951 An attacker can use this to craft malformed certificates and CRLs of
5952 various sizes and potentially cause a segmentation fault, resulting in
5953 a DoS on applications that verify certificates or CRLs. TLS clients
5954 that verify CRLs are affected. TLS clients and servers with client
5955 authentication enabled may be affected if they use custom verification
5958 This issue was reported to OpenSSL by Robert Swiecki (Google), and
5959 independently by Hanno Böck.
5964 * PKCS7 crash with missing EnvelopedContent
5966 The PKCS#7 parsing code does not handle missing inner EncryptedContent
5967 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs
5968 with missing content and trigger a NULL pointer dereference on parsing.
5970 Applications that decrypt PKCS#7 data or otherwise parse PKCS#7
5971 structures from untrusted sources are affected. OpenSSL clients and
5972 servers are not affected.
5974 This issue was reported to OpenSSL by Michal Zalewski (Google).
5979 * CMS verify infinite loop with unknown hash function
5981 When verifying a signedData message the CMS code can enter an infinite loop
5982 if presented with an unknown hash function OID. This can be used to perform
5983 denial of service against any system which verifies signedData messages using
5985 This issue was reported to OpenSSL by Johannes Bauer.
5990 * Race condition handling NewSessionTicket
5992 If a NewSessionTicket is received by a multi-threaded client when attempting to
5993 reuse a previous ticket then a race condition can occur potentially leading to
5994 a double free of the ticket data.
5999 * Reject DH handshakes with parameters shorter than 768 bits.
6001 *Kurt Roeckx and Emilia Kasper*
6003 * dhparam: generate 2048-bit parameters by default.
6005 *Kurt Roeckx and Emilia Kasper*
6007 ### Changes between 1.0.1l and 1.0.1m [19 Mar 2015]
6009 * Segmentation fault in ASN1_TYPE_cmp fix
6011 The function ASN1_TYPE_cmp will crash with an invalid read if an attempt is
6012 made to compare ASN.1 boolean types. Since ASN1_TYPE_cmp is used to check
6013 certificate signature algorithm consistency this can be used to crash any
6014 certificate verification operation and exploited in a DoS attack. Any
6015 application which performs certificate verification is vulnerable including
6016 OpenSSL clients and servers which enable client authentication.
6021 * ASN.1 structure reuse memory corruption fix
6023 Reusing a structure in ASN.1 parsing may allow an attacker to cause
6024 memory corruption via an invalid write. Such reuse is and has been
6025 strongly discouraged and is believed to be rare.
6027 Applications that parse structures containing CHOICE or ANY DEFINED BY
6028 components may be affected. Certificate parsing (d2i_X509 and related
6029 functions) are however not affected. OpenSSL clients and servers are
6035 * PKCS7 NULL pointer dereferences fix
6037 The PKCS#7 parsing code does not handle missing outer ContentInfo
6038 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs with
6039 missing content and trigger a NULL pointer dereference on parsing.
6041 Applications that verify PKCS#7 signatures, decrypt PKCS#7 data or
6042 otherwise parse PKCS#7 structures from untrusted sources are
6043 affected. OpenSSL clients and servers are not affected.
6045 This issue was reported to OpenSSL by Michal Zalewski (Google).
6050 * DoS via reachable assert in SSLv2 servers fix
6052 A malicious client can trigger an OPENSSL_assert (i.e., an abort) in
6053 servers that both support SSLv2 and enable export cipher suites by sending
6054 a specially crafted SSLv2 CLIENT-MASTER-KEY message.
6056 This issue was discovered by Sean Burford (Google) and Emilia Käsper
6057 (OpenSSL development team).
6062 * Use After Free following d2i_ECPrivatekey error fix
6064 A malformed EC private key file consumed via the d2i_ECPrivateKey function
6065 could cause a use after free condition. This, in turn, could cause a double
6066 free in several private key parsing functions (such as d2i_PrivateKey
6067 or EVP_PKCS82PKEY) and could lead to a DoS attack or memory corruption
6068 for applications that receive EC private keys from untrusted
6069 sources. This scenario is considered rare.
6071 This issue was discovered by the BoringSSL project and fixed in their
6077 * X509_to_X509_REQ NULL pointer deref fix
6079 The function X509_to_X509_REQ will crash with a NULL pointer dereference if
6080 the certificate key is invalid. This function is rarely used in practice.
6082 This issue was discovered by Brian Carpenter.
6087 * Removed the export ciphers from the DEFAULT ciphers
6091 ### Changes between 1.0.1k and 1.0.1l [15 Jan 2015]
6093 * Build fixes for the Windows and OpenVMS platforms
6095 *Matt Caswell and Richard Levitte*
6097 ### Changes between 1.0.1j and 1.0.1k [8 Jan 2015]
6099 * Fix DTLS segmentation fault in dtls1_get_record. A carefully crafted DTLS
6100 message can cause a segmentation fault in OpenSSL due to a NULL pointer
6101 dereference. This could lead to a Denial Of Service attack. Thanks to
6102 Markus Stenberg of Cisco Systems, Inc. for reporting this issue.
6107 * Fix DTLS memory leak in dtls1_buffer_record. A memory leak can occur in the
6108 dtls1_buffer_record function under certain conditions. In particular this
6109 could occur if an attacker sent repeated DTLS records with the same
6110 sequence number but for the next epoch. The memory leak could be exploited
6111 by an attacker in a Denial of Service attack through memory exhaustion.
6112 Thanks to Chris Mueller for reporting this issue.
6117 * Fix issue where no-ssl3 configuration sets method to NULL. When openssl is
6118 built with the no-ssl3 option and a SSL v3 ClientHello is received the ssl
6119 method would be set to NULL which could later result in a NULL pointer
6120 dereference. Thanks to Frank Schmirler for reporting this issue.
6125 * Abort handshake if server key exchange message is omitted for ephemeral
6128 Thanks to Karthikeyan Bhargavan of the PROSECCO team at INRIA for
6129 reporting this issue.
6134 * Remove non-export ephemeral RSA code on client and server. This code
6135 violated the TLS standard by allowing the use of temporary RSA keys in
6136 non-export ciphersuites and could be used by a server to effectively
6137 downgrade the RSA key length used to a value smaller than the server
6138 certificate. Thanks for Karthikeyan Bhargavan of the PROSECCO team at
6139 INRIA or reporting this issue.
6144 * Fixed issue where DH client certificates are accepted without verification.
6145 An OpenSSL server will accept a DH certificate for client authentication
6146 without the certificate verify message. This effectively allows a client to
6147 authenticate without the use of a private key. This only affects servers
6148 which trust a client certificate authority which issues certificates
6149 containing DH keys: these are extremely rare and hardly ever encountered.
6150 Thanks for Karthikeyan Bhargavan of the PROSECCO team at INRIA or reporting
6156 * Ensure that the session ID context of an SSL is updated when its
6157 SSL_CTX is updated via SSL_set_SSL_CTX.
6159 The session ID context is typically set from the parent SSL_CTX,
6160 and can vary with the CTX.
6164 * Fix various certificate fingerprint issues.
6166 By using non-DER or invalid encodings outside the signed portion of a
6167 certificate the fingerprint can be changed without breaking the signature.
6168 Although no details of the signed portion of the certificate can be changed
6169 this can cause problems with some applications: e.g. those using the
6170 certificate fingerprint for blacklists.
6172 1. Reject signatures with non zero unused bits.
6174 If the BIT STRING containing the signature has non zero unused bits reject
6175 the signature. All current signature algorithms require zero unused bits.
6177 2. Check certificate algorithm consistency.
6179 Check the AlgorithmIdentifier inside TBS matches the one in the
6180 certificate signature. NB: this will result in signature failure
6181 errors for some broken certificates.
6183 Thanks to Konrad Kraszewski from Google for reporting this issue.
6185 3. Check DSA/ECDSA signatures use DER.
6187 Re-encode DSA/ECDSA signatures and compare with the original received
6188 signature. Return an error if there is a mismatch.
6190 This will reject various cases including garbage after signature
6191 (thanks to Antti Karjalainen and Tuomo Untinen from the Codenomicon CROSS
6192 program for discovering this case) and use of BER or invalid ASN.1 INTEGERs
6193 (negative or with leading zeroes).
6195 Further analysis was conducted and fixes were developed by Stephen Henson
6196 of the OpenSSL core team.
6202 * Correct Bignum squaring. Bignum squaring (BN_sqr) may produce incorrect
6203 results on some platforms, including x86_64. This bug occurs at random
6204 with a very low probability, and is not known to be exploitable in any
6205 way, though its exact impact is difficult to determine. Thanks to Pieter
6206 Wuille (Blockstream) who reported this issue and also suggested an initial
6207 fix. Further analysis was conducted by the OpenSSL development team and
6208 Adam Langley of Google. The final fix was developed by Andy Polyakov of
6209 the OpenSSL core team.
6214 * Do not resume sessions on the server if the negotiated protocol
6215 version does not match the session's version. Resuming with a different
6216 version, while not strictly forbidden by the RFC, is of questionable
6217 sanity and breaks all known clients.
6219 *David Benjamin, Emilia Käsper*
6221 * Tighten handling of the ChangeCipherSpec (CCS) message: reject
6222 early CCS messages during renegotiation. (Note that because
6223 renegotiation is encrypted, this early CCS was not exploitable.)
6227 * Tighten client-side session ticket handling during renegotiation:
6228 ensure that the client only accepts a session ticket if the server sends
6229 the extension anew in the ServerHello. Previously, a TLS client would
6230 reuse the old extension state and thus accept a session ticket if one was
6231 announced in the initial ServerHello.
6233 Similarly, ensure that the client requires a session ticket if one
6234 was advertised in the ServerHello. Previously, a TLS client would
6235 ignore a missing NewSessionTicket message.
6239 ### Changes between 1.0.1i and 1.0.1j [15 Oct 2014]
6243 A flaw in the DTLS SRTP extension parsing code allows an attacker, who
6244 sends a carefully crafted handshake message, to cause OpenSSL to fail
6245 to free up to 64k of memory causing a memory leak. This could be
6246 exploited in a Denial Of Service attack. This issue affects OpenSSL
6247 1.0.1 server implementations for both SSL/TLS and DTLS regardless of
6248 whether SRTP is used or configured. Implementations of OpenSSL that
6249 have been compiled with OPENSSL_NO_SRTP defined are not affected.
6251 The fix was developed by the OpenSSL team.
6256 * Session Ticket Memory Leak.
6258 When an OpenSSL SSL/TLS/DTLS server receives a session ticket the
6259 integrity of that ticket is first verified. In the event of a session
6260 ticket integrity check failing, OpenSSL will fail to free memory
6261 causing a memory leak. By sending a large number of invalid session
6262 tickets an attacker could exploit this issue in a Denial Of Service
6268 * Build option no-ssl3 is incomplete.
6270 When OpenSSL is configured with "no-ssl3" as a build option, servers
6271 could accept and complete a SSL 3.0 handshake, and clients could be
6272 configured to send them.
6275 *Akamai and the OpenSSL team*
6277 * Add support for TLS_FALLBACK_SCSV.
6278 Client applications doing fallback retries should call
6279 SSL_set_mode(s, SSL_MODE_SEND_FALLBACK_SCSV).
6282 *Adam Langley, Bodo Moeller*
6284 * Add additional DigestInfo checks.
6286 Re-encode DigestInto in DER and check against the original when
6287 verifying RSA signature: this will reject any improperly encoded
6288 DigestInfo structures.
6290 Note: this is a precautionary measure and no attacks are currently known.
6294 ### Changes between 1.0.1h and 1.0.1i [6 Aug 2014]
6296 * Fix SRP buffer overrun vulnerability. Invalid parameters passed to the
6297 SRP code can be overrun an internal buffer. Add sanity check that
6298 g, A, B < N to SRP code.
6300 Thanks to Sean Devlin and Watson Ladd of Cryptography Services, NCC
6301 Group for discovering this issue.
6306 * A flaw in the OpenSSL SSL/TLS server code causes the server to negotiate
6307 TLS 1.0 instead of higher protocol versions when the ClientHello message
6308 is badly fragmented. This allows a man-in-the-middle attacker to force a
6309 downgrade to TLS 1.0 even if both the server and the client support a
6310 higher protocol version, by modifying the client's TLS records.
6312 Thanks to David Benjamin and Adam Langley (Google) for discovering and
6313 researching this issue.
6318 * OpenSSL DTLS clients enabling anonymous (EC)DH ciphersuites are subject
6319 to a denial of service attack. A malicious server can crash the client
6320 with a null pointer dereference (read) by specifying an anonymous (EC)DH
6321 ciphersuite and sending carefully crafted handshake messages.
6323 Thanks to Felix Gröbert (Google) for discovering and researching this
6329 * By sending carefully crafted DTLS packets an attacker could cause openssl
6330 to leak memory. This can be exploited through a Denial of Service attack.
6331 Thanks to Adam Langley for discovering and researching this issue.
6336 * An attacker can force openssl to consume large amounts of memory whilst
6337 processing DTLS handshake messages. This can be exploited through a
6338 Denial of Service attack.
6339 Thanks to Adam Langley for discovering and researching this issue.
6344 * An attacker can force an error condition which causes openssl to crash
6345 whilst processing DTLS packets due to memory being freed twice. This
6346 can be exploited through a Denial of Service attack.
6347 Thanks to Adam Langley and Wan-Teh Chang for discovering and researching
6353 * If a multithreaded client connects to a malicious server using a resumed
6354 session and the server sends an ec point format extension it could write
6355 up to 255 bytes to freed memory.
6357 Thanks to Gabor Tyukasz (LogMeIn Inc) for discovering and researching this
6363 * A malicious server can crash an OpenSSL client with a null pointer
6364 dereference (read) by specifying an SRP ciphersuite even though it was not
6365 properly negotiated with the client. This can be exploited through a
6366 Denial of Service attack.
6368 Thanks to Joonas Kuorilehto and Riku Hietamäki (Codenomicon) for
6369 discovering and researching this issue.
6374 * A flaw in OBJ_obj2txt may cause pretty printing functions such as
6375 X509_name_oneline, X509_name_print_ex et al. to leak some information
6376 from the stack. Applications may be affected if they echo pretty printing
6377 output to the attacker.
6379 Thanks to Ivan Fratric (Google) for discovering this issue.
6382 *Emilia Käsper, and Steve Henson*
6384 * Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
6385 for corner cases. (Certain input points at infinity could lead to
6386 bogus results, with non-infinity inputs mapped to infinity too.)
6390 ### Changes between 1.0.1g and 1.0.1h [5 Jun 2014]
6392 * Fix for SSL/TLS MITM flaw. An attacker using a carefully crafted
6393 handshake can force the use of weak keying material in OpenSSL
6394 SSL/TLS clients and servers.
6396 Thanks to KIKUCHI Masashi (Lepidum Co. Ltd.) for discovering and
6397 researching this issue. [CVE-2014-0224][]
6399 *KIKUCHI Masashi, Steve Henson*
6401 * Fix DTLS recursion flaw. By sending an invalid DTLS handshake to an
6402 OpenSSL DTLS client the code can be made to recurse eventually crashing
6405 Thanks to Imre Rad (Search-Lab Ltd.) for discovering this issue.
6408 *Imre Rad, Steve Henson*
6410 * Fix DTLS invalid fragment vulnerability. A buffer overrun attack can
6411 be triggered by sending invalid DTLS fragments to an OpenSSL DTLS
6412 client or server. This is potentially exploitable to run arbitrary
6413 code on a vulnerable client or server.
6415 Thanks to Jüri Aedla for reporting this issue. [CVE-2014-0195][]
6417 *Jüri Aedla, Steve Henson*
6419 * Fix bug in TLS code where clients enable anonymous ECDH ciphersuites
6420 are subject to a denial of service attack.
6422 Thanks to Felix Gröbert and Ivan Fratric at Google for discovering
6423 this issue. [CVE-2014-3470][]
6425 *Felix Gröbert, Ivan Fratric, Steve Henson*
6427 * Harmonize version and its documentation. -f flag is used to display
6430 *mancha <mancha1@zoho.com>*
6432 * Fix eckey_priv_encode so it immediately returns an error upon a failure
6433 in i2d_ECPrivateKey.
6435 *mancha <mancha1@zoho.com>*
6437 * Fix some double frees. These are not thought to be exploitable.
6439 *mancha <mancha1@zoho.com>*
6441 ### Changes between 1.0.1f and 1.0.1g [7 Apr 2014]
6443 * A missing bounds check in the handling of the TLS heartbeat extension
6444 can be used to reveal up to 64k of memory to a connected client or
6447 Thanks for Neel Mehta of Google Security for discovering this bug and to
6448 Adam Langley <agl@chromium.org> and Bodo Moeller <bmoeller@acm.org> for
6449 preparing the fix [CVE-2014-0160][]
6451 *Adam Langley, Bodo Moeller*
6453 * Fix for the attack described in the paper "Recovering OpenSSL
6454 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
6455 by Yuval Yarom and Naomi Benger. Details can be obtained from:
6456 <http://eprint.iacr.org/2014/140>
6458 Thanks to Yuval Yarom and Naomi Benger for discovering this
6459 flaw and to Yuval Yarom for supplying a fix [CVE-2014-0076][]
6461 *Yuval Yarom and Naomi Benger*
6463 * TLS pad extension: draft-agl-tls-padding-03
6465 Workaround for the "TLS hang bug" (see FAQ and PR#2771): if the
6466 TLS client Hello record length value would otherwise be > 255 and
6467 less that 512 pad with a dummy extension containing zeroes so it
6468 is at least 512 bytes long.
6470 *Adam Langley, Steve Henson*
6472 ### Changes between 1.0.1e and 1.0.1f [6 Jan 2014]
6474 * Fix for TLS record tampering bug. A carefully crafted invalid
6475 handshake could crash OpenSSL with a NULL pointer exception.
6476 Thanks to Anton Johansson for reporting this issues.
6479 * Keep original DTLS digest and encryption contexts in retransmission
6480 structures so we can use the previous session parameters if they need
6481 to be resent. [CVE-2013-6450][]
6485 * Add option SSL_OP_SAFARI_ECDHE_ECDSA_BUG (part of SSL_OP_ALL) which
6486 avoids preferring ECDHE-ECDSA ciphers when the client appears to be
6487 Safari on OS X. Safari on OS X 10.8..10.8.3 advertises support for
6488 several ECDHE-ECDSA ciphers, but fails to negotiate them. The bug
6489 is fixed in OS X 10.8.4, but Apple have ruled out both hot fixing
6490 10.8..10.8.3 and forcing users to upgrade to 10.8.4 or newer.
6492 *Rob Stradling, Adam Langley*
6494 ### Changes between 1.0.1d and 1.0.1e [11 Feb 2013]
6496 * Correct fix for CVE-2013-0169. The original didn't work on AES-NI
6497 supporting platforms or when small records were transferred.
6499 *Andy Polyakov, Steve Henson*
6501 ### Changes between 1.0.1c and 1.0.1d [5 Feb 2013]
6503 * Make the decoding of SSLv3, TLS and DTLS CBC records constant time.
6505 This addresses the flaw in CBC record processing discovered by
6506 Nadhem Alfardan and Kenny Paterson. Details of this attack can be found
6507 at: <http://www.isg.rhul.ac.uk/tls/>
6509 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
6510 Security Group at Royal Holloway, University of London
6511 (www.isg.rhul.ac.uk) for discovering this flaw and Adam Langley and
6512 Emilia Käsper for the initial patch.
6515 *Emilia Käsper, Adam Langley, Ben Laurie, Andy Polyakov, Steve Henson*
6517 * Fix flaw in AESNI handling of TLS 1.2 and 1.1 records for CBC mode
6518 ciphersuites which can be exploited in a denial of service attack.
6519 Thanks go to and to Adam Langley <agl@chromium.org> for discovering
6520 and detecting this bug and to Wolfgang Ettlinger
6521 <wolfgang.ettlinger@gmail.com> for independently discovering this issue.
6526 * Return an error when checking OCSP signatures when key is NULL.
6527 This fixes a DoS attack. [CVE-2013-0166][]
6531 * Make openssl verify return errors.
6533 *Chris Palmer <palmer@google.com> and Ben Laurie*
6535 * Call OCSP Stapling callback after ciphersuite has been chosen, so
6536 the right response is stapled. Also change SSL_get_certificate()
6537 so it returns the certificate actually sent.
6538 See <http://rt.openssl.org/Ticket/Display.html?id=2836>.
6540 *Rob Stradling <rob.stradling@comodo.com>*
6542 * Fix possible deadlock when decoding public keys.
6546 * Don't use TLS 1.0 record version number in initial client hello
6551 ### Changes between 1.0.1b and 1.0.1c [10 May 2012]
6553 * Sanity check record length before skipping explicit IV in TLS
6554 1.2, 1.1 and DTLS to fix DoS attack.
6556 Thanks to Codenomicon for discovering this issue using Fuzz-o-Matic
6557 fuzzing as a service testing platform.
6562 * Initialise tkeylen properly when encrypting CMS messages.
6563 Thanks to Solar Designer of Openwall for reporting this issue.
6567 * In FIPS mode don't try to use composite ciphers as they are not
6572 ### Changes between 1.0.1a and 1.0.1b [26 Apr 2012]
6574 * OpenSSL 1.0.0 sets SSL_OP_ALL to 0x80000FFFL and OpenSSL 1.0.1 and
6575 1.0.1a set SSL_OP_NO_TLSv1_1 to 0x00000400L which would unfortunately
6576 mean any application compiled against OpenSSL 1.0.0 headers setting
6577 SSL_OP_ALL would also set SSL_OP_NO_TLSv1_1, unintentionally disabling
6578 TLS 1.1 also. Fix this by changing the value of SSL_OP_NO_TLSv1_1 to
6579 0x10000000L Any application which was previously compiled against
6580 OpenSSL 1.0.1 or 1.0.1a headers and which cares about SSL_OP_NO_TLSv1_1
6581 will need to be recompiled as a result. Letting be results in
6582 inability to disable specifically TLS 1.1 and in client context,
6583 in unlike event, limit maximum offered version to TLS 1.0 [see below].
6587 * In order to ensure interoperability SSL_OP_NO_protocolX does not
6588 disable just protocol X, but all protocols above X *if* there are
6589 protocols *below* X still enabled. In more practical terms it means
6590 that if application wants to disable TLS1.0 in favor of TLS1.1 and
6591 above, it's not sufficient to pass `SSL_OP_NO_TLSv1`, one has to pass
6592 `SSL_OP_NO_TLSv1|SSL_OP_NO_SSLv3|SSL_OP_NO_SSLv2`. This applies to
6597 ### Changes between 1.0.1 and 1.0.1a [19 Apr 2012]
6599 * Check for potentially exploitable overflows in asn1_d2i_read_bio
6600 BUF_mem_grow and BUF_mem_grow_clean. Refuse attempts to shrink buffer
6601 in CRYPTO_realloc_clean.
6603 Thanks to Tavis Ormandy, Google Security Team, for discovering this
6604 issue and to Adam Langley <agl@chromium.org> for fixing it.
6607 *Adam Langley (Google), Tavis Ormandy, Google Security Team*
6609 * Don't allow TLS 1.2 SHA-256 ciphersuites in TLS 1.0, 1.1 connections.
6613 * Workarounds for some broken servers that "hang" if a client hello
6614 record length exceeds 255 bytes.
6616 1. Do not use record version number > TLS 1.0 in initial client
6617 hello: some (but not all) hanging servers will now work.
6618 2. If we set OPENSSL_MAX_TLS1_2_CIPHER_LENGTH this will truncate
6619 the number of ciphers sent in the client hello. This should be
6620 set to an even number, such as 50, for example by passing:
6621 -DOPENSSL_MAX_TLS1_2_CIPHER_LENGTH=50 to config or Configure.
6622 Most broken servers should now work.
6623 3. If all else fails setting OPENSSL_NO_TLS1_2_CLIENT will disable
6624 TLS 1.2 client support entirely.
6628 * Fix SEGV in Vector Permutation AES module observed in OpenSSH.
6632 ### Changes between 1.0.0h and 1.0.1 [14 Mar 2012]
6634 * Add compatibility with old MDC2 signatures which use an ASN1 OCTET
6635 STRING form instead of a DigestInfo.
6639 * The format used for MDC2 RSA signatures is inconsistent between EVP
6640 and the RSA_sign/RSA_verify functions. This was made more apparent when
6641 OpenSSL used RSA_sign/RSA_verify for some RSA signatures in particular
6642 those which went through EVP_PKEY_METHOD in 1.0.0 and later. Detect
6643 the correct format in RSA_verify so both forms transparently work.
6647 * Some servers which support TLS 1.0 can choke if we initially indicate
6648 support for TLS 1.2 and later renegotiate using TLS 1.0 in the RSA
6649 encrypted premaster secret. As a workaround use the maximum permitted
6650 client version in client hello, this should keep such servers happy
6651 and still work with previous versions of OpenSSL.
6655 * Add support for TLS/DTLS heartbeats.
6657 *Robin Seggelmann <seggelmann@fh-muenster.de>*
6659 * Add support for SCTP.
6661 *Robin Seggelmann <seggelmann@fh-muenster.de>*
6663 * Improved PRNG seeding for VOS.
6665 *Paul Green <Paul.Green@stratus.com>*
6667 * Extensive assembler packs updates, most notably:
6669 - x86[_64]: AES-NI, PCLMULQDQ, RDRAND support;
6670 - x86[_64]: SSSE3 support (SHA1, vector-permutation AES);
6671 - x86_64: bit-sliced AES implementation;
6672 - ARM: NEON support, contemporary platforms optimizations;
6673 - s390x: z196 support;
6674 - `*`: GHASH and GF(2^m) multiplication implementations;
6678 * Make TLS-SRP code conformant with RFC 5054 API cleanup
6679 (removal of unnecessary code)
6681 *Peter Sylvester <peter.sylvester@edelweb.fr>*
6683 * Add TLS key material exporter from RFC 5705.
6687 * Add DTLS-SRTP negotiation from RFC 5764.
6691 * Add Next Protocol Negotiation,
6692 <http://tools.ietf.org/html/draft-agl-tls-nextprotoneg-00>. Can be
6693 disabled with a no-npn flag to config or Configure. Code donated
6696 *Adam Langley <agl@google.com> and Ben Laurie*
6698 * Add optional 64-bit optimized implementations of elliptic curves NIST-P224,
6699 NIST-P256, NIST-P521, with constant-time single point multiplication on
6700 typical inputs. Compiler support for the nonstandard type `__uint128_t` is
6701 required to use this (present in gcc 4.4 and later, for 64-bit builds).
6702 Code made available under Apache License version 2.0.
6704 Specify "enable-ec_nistp_64_gcc_128" on the Configure (or config) command
6705 line to include this in your build of OpenSSL, and run "make depend" (or
6706 "make update"). This enables the following EC_METHODs:
6708 EC_GFp_nistp224_method()
6709 EC_GFp_nistp256_method()
6710 EC_GFp_nistp521_method()
6712 EC_GROUP_new_by_curve_name() will automatically use these (while
6713 EC_GROUP_new_curve_GFp() currently prefers the more flexible
6716 *Emilia Käsper, Adam Langley, Bodo Moeller (Google)*
6718 * Use type ossl_ssize_t instead of ssize_t which isn't available on
6719 all platforms. Move ssize_t definition from e_os.h to the public
6720 header file e_os2.h as it now appears in public header file cms.h
6724 * New -sigopt option to the ca, req and x509 utilities. Additional
6725 signature parameters can be passed using this option and in
6730 * Add RSA PSS signing function. This will generate and set the
6731 appropriate AlgorithmIdentifiers for PSS based on those in the
6732 corresponding EVP_MD_CTX structure. No application support yet.
6736 * Support for companion algorithm specific ASN1 signing routines.
6737 New function ASN1_item_sign_ctx() signs a pre-initialised
6738 EVP_MD_CTX structure and sets AlgorithmIdentifiers based on
6739 the appropriate parameters.
6743 * Add new algorithm specific ASN1 verification initialisation function
6744 to EVP_PKEY_ASN1_METHOD: this is not in EVP_PKEY_METHOD since the ASN1
6745 handling will be the same no matter what EVP_PKEY_METHOD is used.
6746 Add a PSS handler to support verification of PSS signatures: checked
6747 against a number of sample certificates.
6751 * Add signature printing for PSS. Add PSS OIDs.
6753 *Steve Henson, Martin Kaiser <lists@kaiser.cx>*
6755 * Add algorithm specific signature printing. An individual ASN1 method
6756 can now print out signatures instead of the standard hex dump.
6758 More complex signatures (e.g. PSS) can print out more meaningful
6759 information. Include DSA version that prints out the signature
6764 * Password based recipient info support for CMS library: implementing
6769 * Split password based encryption into PBES2 and PBKDF2 functions. This
6770 neatly separates the code into cipher and PBE sections and is required
6771 for some algorithms that split PBES2 into separate pieces (such as
6772 password based CMS).
6776 * Session-handling fixes:
6777 - Fix handling of connections that are resuming with a session ID,
6778 but also support Session Tickets.
6779 - Fix a bug that suppressed issuing of a new ticket if the client
6780 presented a ticket with an expired session.
6781 - Try to set the ticket lifetime hint to something reasonable.
6782 - Make tickets shorter by excluding irrelevant information.
6783 - On the client side, don't ignore renewed tickets.
6785 *Adam Langley, Bodo Moeller (Google)*
6787 * Fix PSK session representation.
6791 * Add RC4-MD5 and AESNI-SHA1 "stitched" implementations.
6793 This work was sponsored by Intel.
6797 * Add GCM support to TLS library. Some custom code is needed to split
6798 the IV between the fixed (from PRF) and explicit (from TLS record)
6799 portions. This adds all GCM ciphersuites supported by RFC5288 and
6800 RFC5289. Generalise some `AES*` cipherstrings to include GCM and
6801 add a special AESGCM string for GCM only.
6805 * Expand range of ctrls for AES GCM. Permit setting invocation
6806 field on decrypt and retrieval of invocation field only on encrypt.
6810 * Add HMAC ECC ciphersuites from RFC5289. Include SHA384 PRF support.
6811 As required by RFC5289 these ciphersuites cannot be used if for
6812 versions of TLS earlier than 1.2.
6816 * For FIPS capable OpenSSL interpret a NULL default public key method
6817 as unset and return the appropriate default but do *not* set the default.
6818 This means we can return the appropriate method in applications that
6819 switch between FIPS and non-FIPS modes.
6823 * Redirect HMAC and CMAC operations to FIPS module in FIPS mode. If an
6824 ENGINE is used then we cannot handle that in the FIPS module so we
6825 keep original code iff non-FIPS operations are allowed.
6829 * Add -attime option to openssl utilities.
6831 *Peter Eckersley <pde@eff.org>, Ben Laurie and Steve Henson*
6833 * Redirect DSA and DH operations to FIPS module in FIPS mode.
6837 * Redirect ECDSA and ECDH operations to FIPS module in FIPS mode. Also use
6838 FIPS EC methods unconditionally for now.
6842 * New build option no-ec2m to disable characteristic 2 code.
6846 * Backport libcrypto audit of return value checking from 1.1.0-dev; not
6847 all cases can be covered as some introduce binary incompatibilities.
6851 * Redirect RSA operations to FIPS module including keygen,
6852 encrypt, decrypt, sign and verify. Block use of non FIPS RSA methods.
6856 * Add similar low level API blocking to ciphers.
6860 * Low level digest APIs are not approved in FIPS mode: any attempt
6861 to use these will cause a fatal error. Applications that *really* want
6862 to use them can use the `private_*` version instead.
6866 * Redirect cipher operations to FIPS module for FIPS builds.
6870 * Redirect digest operations to FIPS module for FIPS builds.
6874 * Update build system to add "fips" flag which will link in fipscanister.o
6875 for static and shared library builds embedding a signature if needed.
6879 * Output TLS supported curves in preference order instead of numerical
6880 order. This is currently hardcoded for the highest order curves first.
6881 This should be configurable so applications can judge speed vs strength.
6885 * Add TLS v1.2 server support for client authentication.
6889 * Add support for FIPS mode in ssl library: disable SSLv3, non-FIPS ciphers
6894 * Functions FIPS_mode_set() and FIPS_mode() which call the underlying
6895 FIPS modules versions.
6899 * Add TLS v1.2 client side support for client authentication. Keep cache
6900 of handshake records longer as we don't know the hash algorithm to use
6901 until after the certificate request message is received.
6905 * Initial TLS v1.2 client support. Add a default signature algorithms
6906 extension including all the algorithms we support. Parse new signature
6907 format in client key exchange. Relax some ECC signing restrictions for
6908 TLS v1.2 as indicated in RFC5246.
6912 * Add server support for TLS v1.2 signature algorithms extension. Switch
6913 to new signature format when needed using client digest preference.
6914 All server ciphersuites should now work correctly in TLS v1.2. No client
6915 support yet and no support for client certificates.
6919 * Initial TLS v1.2 support. Add new SHA256 digest to ssl code, switch
6920 to SHA256 for PRF when using TLS v1.2 and later. Add new SHA256 based
6921 ciphersuites. At present only RSA key exchange ciphersuites work with
6922 TLS v1.2. Add new option for TLS v1.2 replacing the old and obsolete
6923 SSL_OP_PKCS1_CHECK flags with SSL_OP_NO_TLSv1_2. New TLSv1.2 methods
6924 and version checking.
6928 * New option OPENSSL_NO_SSL_INTERN. If an application can be compiled
6929 with this defined it will not be affected by any changes to ssl internal
6930 structures. Add several utility functions to allow openssl application
6931 to work with OPENSSL_NO_SSL_INTERN defined.
6935 * A long standing patch to add support for SRP from EdelWeb (Peter
6936 Sylvester and Christophe Renou) was integrated.
6937 *Christophe Renou <christophe.renou@edelweb.fr>, Peter Sylvester
6938 <peter.sylvester@edelweb.fr>, Tom Wu <tjw@cs.stanford.edu>, and
6941 * Add functions to copy EVP_PKEY_METHOD and retrieve flags and id.
6945 * Permit abbreviated handshakes when renegotiating using the function
6946 SSL_renegotiate_abbreviated().
6948 *Robin Seggelmann <seggelmann@fh-muenster.de>*
6950 * Add call to ENGINE_register_all_complete() to
6951 ENGINE_load_builtin_engines(), so some implementations get used
6952 automatically instead of needing explicit application support.
6956 * Add support for TLS key exporter as described in RFC5705.
6958 *Robin Seggelmann <seggelmann@fh-muenster.de>, Steve Henson*
6960 * Initial TLSv1.1 support. Since TLSv1.1 is very similar to TLS v1.0 only
6961 a few changes are required:
6963 Add SSL_OP_NO_TLSv1_1 flag.
6964 Add TLSv1_1 methods.
6965 Update version checking logic to handle version 1.1.
6966 Add explicit IV handling (ported from DTLS code).
6967 Add command line options to s_client/s_server.
6974 ### Changes between 1.0.0s and 1.0.0t [3 Dec 2015]
6976 * X509_ATTRIBUTE memory leak
6978 When presented with a malformed X509_ATTRIBUTE structure OpenSSL will leak
6979 memory. This structure is used by the PKCS#7 and CMS routines so any
6980 application which reads PKCS#7 or CMS data from untrusted sources is
6981 affected. SSL/TLS is not affected.
6983 This issue was reported to OpenSSL by Adam Langley (Google/BoringSSL) using
6989 * Race condition handling PSK identify hint
6991 If PSK identity hints are received by a multi-threaded client then
6992 the values are wrongly updated in the parent SSL_CTX structure. This can
6993 result in a race condition potentially leading to a double free of the
6999 ### Changes between 1.0.0r and 1.0.0s [11 Jun 2015]
7001 * Malformed ECParameters causes infinite loop
7003 When processing an ECParameters structure OpenSSL enters an infinite loop
7004 if the curve specified is over a specially malformed binary polynomial
7007 This can be used to perform denial of service against any
7008 system which processes public keys, certificate requests or
7009 certificates. This includes TLS clients and TLS servers with
7010 client authentication enabled.
7012 This issue was reported to OpenSSL by Joseph Barr-Pixton.
7017 * Exploitable out-of-bounds read in X509_cmp_time
7019 X509_cmp_time does not properly check the length of the ASN1_TIME
7020 string and can read a few bytes out of bounds. In addition,
7021 X509_cmp_time accepts an arbitrary number of fractional seconds in the
7024 An attacker can use this to craft malformed certificates and CRLs of
7025 various sizes and potentially cause a segmentation fault, resulting in
7026 a DoS on applications that verify certificates or CRLs. TLS clients
7027 that verify CRLs are affected. TLS clients and servers with client
7028 authentication enabled may be affected if they use custom verification
7031 This issue was reported to OpenSSL by Robert Swiecki (Google), and
7032 independently by Hanno Böck.
7037 * PKCS7 crash with missing EnvelopedContent
7039 The PKCS#7 parsing code does not handle missing inner EncryptedContent
7040 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs
7041 with missing content and trigger a NULL pointer dereference on parsing.
7043 Applications that decrypt PKCS#7 data or otherwise parse PKCS#7
7044 structures from untrusted sources are affected. OpenSSL clients and
7045 servers are not affected.
7047 This issue was reported to OpenSSL by Michal Zalewski (Google).
7052 * CMS verify infinite loop with unknown hash function
7054 When verifying a signedData message the CMS code can enter an infinite loop
7055 if presented with an unknown hash function OID. This can be used to perform
7056 denial of service against any system which verifies signedData messages using
7058 This issue was reported to OpenSSL by Johannes Bauer.
7063 * Race condition handling NewSessionTicket
7065 If a NewSessionTicket is received by a multi-threaded client when attempting to
7066 reuse a previous ticket then a race condition can occur potentially leading to
7067 a double free of the ticket data.
7072 ### Changes between 1.0.0q and 1.0.0r [19 Mar 2015]
7074 * Segmentation fault in ASN1_TYPE_cmp fix
7076 The function ASN1_TYPE_cmp will crash with an invalid read if an attempt is
7077 made to compare ASN.1 boolean types. Since ASN1_TYPE_cmp is used to check
7078 certificate signature algorithm consistency this can be used to crash any
7079 certificate verification operation and exploited in a DoS attack. Any
7080 application which performs certificate verification is vulnerable including
7081 OpenSSL clients and servers which enable client authentication.
7086 * ASN.1 structure reuse memory corruption fix
7088 Reusing a structure in ASN.1 parsing may allow an attacker to cause
7089 memory corruption via an invalid write. Such reuse is and has been
7090 strongly discouraged and is believed to be rare.
7092 Applications that parse structures containing CHOICE or ANY DEFINED BY
7093 components may be affected. Certificate parsing (d2i_X509 and related
7094 functions) are however not affected. OpenSSL clients and servers are
7100 * PKCS7 NULL pointer dereferences fix
7102 The PKCS#7 parsing code does not handle missing outer ContentInfo
7103 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs with
7104 missing content and trigger a NULL pointer dereference on parsing.
7106 Applications that verify PKCS#7 signatures, decrypt PKCS#7 data or
7107 otherwise parse PKCS#7 structures from untrusted sources are
7108 affected. OpenSSL clients and servers are not affected.
7110 This issue was reported to OpenSSL by Michal Zalewski (Google).
7115 * DoS via reachable assert in SSLv2 servers fix
7117 A malicious client can trigger an OPENSSL_assert (i.e., an abort) in
7118 servers that both support SSLv2 and enable export cipher suites by sending
7119 a specially crafted SSLv2 CLIENT-MASTER-KEY message.
7121 This issue was discovered by Sean Burford (Google) and Emilia Käsper
7122 (OpenSSL development team).
7127 * Use After Free following d2i_ECPrivatekey error fix
7129 A malformed EC private key file consumed via the d2i_ECPrivateKey function
7130 could cause a use after free condition. This, in turn, could cause a double
7131 free in several private key parsing functions (such as d2i_PrivateKey
7132 or EVP_PKCS82PKEY) and could lead to a DoS attack or memory corruption
7133 for applications that receive EC private keys from untrusted
7134 sources. This scenario is considered rare.
7136 This issue was discovered by the BoringSSL project and fixed in their
7142 * X509_to_X509_REQ NULL pointer deref fix
7144 The function X509_to_X509_REQ will crash with a NULL pointer dereference if
7145 the certificate key is invalid. This function is rarely used in practice.
7147 This issue was discovered by Brian Carpenter.
7152 * Removed the export ciphers from the DEFAULT ciphers
7156 ### Changes between 1.0.0p and 1.0.0q [15 Jan 2015]
7158 * Build fixes for the Windows and OpenVMS platforms
7160 *Matt Caswell and Richard Levitte*
7162 ### Changes between 1.0.0o and 1.0.0p [8 Jan 2015]
7164 * Fix DTLS segmentation fault in dtls1_get_record. A carefully crafted DTLS
7165 message can cause a segmentation fault in OpenSSL due to a NULL pointer
7166 dereference. This could lead to a Denial Of Service attack. Thanks to
7167 Markus Stenberg of Cisco Systems, Inc. for reporting this issue.
7172 * Fix DTLS memory leak in dtls1_buffer_record. A memory leak can occur in the
7173 dtls1_buffer_record function under certain conditions. In particular this
7174 could occur if an attacker sent repeated DTLS records with the same
7175 sequence number but for the next epoch. The memory leak could be exploited
7176 by an attacker in a Denial of Service attack through memory exhaustion.
7177 Thanks to Chris Mueller for reporting this issue.
7182 * Fix issue where no-ssl3 configuration sets method to NULL. When openssl is
7183 built with the no-ssl3 option and a SSL v3 ClientHello is received the ssl
7184 method would be set to NULL which could later result in a NULL pointer
7185 dereference. Thanks to Frank Schmirler for reporting this issue.
7190 * Abort handshake if server key exchange message is omitted for ephemeral
7193 Thanks to Karthikeyan Bhargavan of the PROSECCO team at INRIA for
7194 reporting this issue.
7199 * Remove non-export ephemeral RSA code on client and server. This code
7200 violated the TLS standard by allowing the use of temporary RSA keys in
7201 non-export ciphersuites and could be used by a server to effectively
7202 downgrade the RSA key length used to a value smaller than the server
7203 certificate. Thanks for Karthikeyan Bhargavan of the PROSECCO team at
7204 INRIA or reporting this issue.
7209 * Fixed issue where DH client certificates are accepted without verification.
7210 An OpenSSL server will accept a DH certificate for client authentication
7211 without the certificate verify message. This effectively allows a client to
7212 authenticate without the use of a private key. This only affects servers
7213 which trust a client certificate authority which issues certificates
7214 containing DH keys: these are extremely rare and hardly ever encountered.
7215 Thanks for Karthikeyan Bhargavan of the PROSECCO team at INRIA or reporting
7221 * Correct Bignum squaring. Bignum squaring (BN_sqr) may produce incorrect
7222 results on some platforms, including x86_64. This bug occurs at random
7223 with a very low probability, and is not known to be exploitable in any
7224 way, though its exact impact is difficult to determine. Thanks to Pieter
7225 Wuille (Blockstream) who reported this issue and also suggested an initial
7226 fix. Further analysis was conducted by the OpenSSL development team and
7227 Adam Langley of Google. The final fix was developed by Andy Polyakov of
7228 the OpenSSL core team.
7233 * Fix various certificate fingerprint issues.
7235 By using non-DER or invalid encodings outside the signed portion of a
7236 certificate the fingerprint can be changed without breaking the signature.
7237 Although no details of the signed portion of the certificate can be changed
7238 this can cause problems with some applications: e.g. those using the
7239 certificate fingerprint for blacklists.
7241 1. Reject signatures with non zero unused bits.
7243 If the BIT STRING containing the signature has non zero unused bits reject
7244 the signature. All current signature algorithms require zero unused bits.
7246 2. Check certificate algorithm consistency.
7248 Check the AlgorithmIdentifier inside TBS matches the one in the
7249 certificate signature. NB: this will result in signature failure
7250 errors for some broken certificates.
7252 Thanks to Konrad Kraszewski from Google for reporting this issue.
7254 3. Check DSA/ECDSA signatures use DER.
7256 Reencode DSA/ECDSA signatures and compare with the original received
7257 signature. Return an error if there is a mismatch.
7259 This will reject various cases including garbage after signature
7260 (thanks to Antti Karjalainen and Tuomo Untinen from the Codenomicon CROSS
7261 program for discovering this case) and use of BER or invalid ASN.1 INTEGERs
7262 (negative or with leading zeroes).
7264 Further analysis was conducted and fixes were developed by Stephen Henson
7265 of the OpenSSL core team.
7271 ### Changes between 1.0.0n and 1.0.0o [15 Oct 2014]
7273 * Session Ticket Memory Leak.
7275 When an OpenSSL SSL/TLS/DTLS server receives a session ticket the
7276 integrity of that ticket is first verified. In the event of a session
7277 ticket integrity check failing, OpenSSL will fail to free memory
7278 causing a memory leak. By sending a large number of invalid session
7279 tickets an attacker could exploit this issue in a Denial Of Service
7285 * Build option no-ssl3 is incomplete.
7287 When OpenSSL is configured with "no-ssl3" as a build option, servers
7288 could accept and complete a SSL 3.0 handshake, and clients could be
7289 configured to send them.
7292 *Akamai and the OpenSSL team*
7294 * Add support for TLS_FALLBACK_SCSV.
7295 Client applications doing fallback retries should call
7296 SSL_set_mode(s, SSL_MODE_SEND_FALLBACK_SCSV).
7299 *Adam Langley, Bodo Moeller*
7301 * Add additional DigestInfo checks.
7303 Reencode DigestInto in DER and check against the original when
7304 verifying RSA signature: this will reject any improperly encoded
7305 DigestInfo structures.
7307 Note: this is a precautionary measure and no attacks are currently known.
7311 ### Changes between 1.0.0m and 1.0.0n [6 Aug 2014]
7313 * OpenSSL DTLS clients enabling anonymous (EC)DH ciphersuites are subject
7314 to a denial of service attack. A malicious server can crash the client
7315 with a null pointer dereference (read) by specifying an anonymous (EC)DH
7316 ciphersuite and sending carefully crafted handshake messages.
7318 Thanks to Felix Gröbert (Google) for discovering and researching this
7324 * By sending carefully crafted DTLS packets an attacker could cause openssl
7325 to leak memory. This can be exploited through a Denial of Service attack.
7326 Thanks to Adam Langley for discovering and researching this issue.
7331 * An attacker can force openssl to consume large amounts of memory whilst
7332 processing DTLS handshake messages. This can be exploited through a
7333 Denial of Service attack.
7334 Thanks to Adam Langley for discovering and researching this issue.
7339 * An attacker can force an error condition which causes openssl to crash
7340 whilst processing DTLS packets due to memory being freed twice. This
7341 can be exploited through a Denial of Service attack.
7342 Thanks to Adam Langley and Wan-Teh Chang for discovering and researching
7348 * If a multithreaded client connects to a malicious server using a resumed
7349 session and the server sends an ec point format extension it could write
7350 up to 255 bytes to freed memory.
7352 Thanks to Gabor Tyukasz (LogMeIn Inc) for discovering and researching this
7358 * A flaw in OBJ_obj2txt may cause pretty printing functions such as
7359 X509_name_oneline, X509_name_print_ex et al. to leak some information
7360 from the stack. Applications may be affected if they echo pretty printing
7361 output to the attacker.
7363 Thanks to Ivan Fratric (Google) for discovering this issue.
7366 *Emilia Käsper, and Steve Henson*
7368 * Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
7369 for corner cases. (Certain input points at infinity could lead to
7370 bogus results, with non-infinity inputs mapped to infinity too.)
7374 ### Changes between 1.0.0l and 1.0.0m [5 Jun 2014]
7376 * Fix for SSL/TLS MITM flaw. An attacker using a carefully crafted
7377 handshake can force the use of weak keying material in OpenSSL
7378 SSL/TLS clients and servers.
7380 Thanks to KIKUCHI Masashi (Lepidum Co. Ltd.) for discovering and
7381 researching this issue. [CVE-2014-0224][]
7383 *KIKUCHI Masashi, Steve Henson*
7385 * Fix DTLS recursion flaw. By sending an invalid DTLS handshake to an
7386 OpenSSL DTLS client the code can be made to recurse eventually crashing
7389 Thanks to Imre Rad (Search-Lab Ltd.) for discovering this issue.
7392 *Imre Rad, Steve Henson*
7394 * Fix DTLS invalid fragment vulnerability. A buffer overrun attack can
7395 be triggered by sending invalid DTLS fragments to an OpenSSL DTLS
7396 client or server. This is potentially exploitable to run arbitrary
7397 code on a vulnerable client or server.
7399 Thanks to Jüri Aedla for reporting this issue. [CVE-2014-0195][]
7401 *Jüri Aedla, Steve Henson*
7403 * Fix bug in TLS code where clients enable anonymous ECDH ciphersuites
7404 are subject to a denial of service attack.
7406 Thanks to Felix Gröbert and Ivan Fratric at Google for discovering
7407 this issue. [CVE-2014-3470][]
7409 *Felix Gröbert, Ivan Fratric, Steve Henson*
7411 * Harmonize version and its documentation. -f flag is used to display
7414 *mancha <mancha1@zoho.com>*
7416 * Fix eckey_priv_encode so it immediately returns an error upon a failure
7417 in i2d_ECPrivateKey.
7419 *mancha <mancha1@zoho.com>*
7421 * Fix some double frees. These are not thought to be exploitable.
7423 *mancha <mancha1@zoho.com>*
7425 * Fix for the attack described in the paper "Recovering OpenSSL
7426 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
7427 by Yuval Yarom and Naomi Benger. Details can be obtained from:
7428 <http://eprint.iacr.org/2014/140>
7430 Thanks to Yuval Yarom and Naomi Benger for discovering this
7431 flaw and to Yuval Yarom for supplying a fix [CVE-2014-0076][]
7433 *Yuval Yarom and Naomi Benger*
7435 ### Changes between 1.0.0k and 1.0.0l [6 Jan 2014]
7437 * Keep original DTLS digest and encryption contexts in retransmission
7438 structures so we can use the previous session parameters if they need
7439 to be resent. [CVE-2013-6450][]
7443 * Add option SSL_OP_SAFARI_ECDHE_ECDSA_BUG (part of SSL_OP_ALL) which
7444 avoids preferring ECDHE-ECDSA ciphers when the client appears to be
7445 Safari on OS X. Safari on OS X 10.8..10.8.3 advertises support for
7446 several ECDHE-ECDSA ciphers, but fails to negotiate them. The bug
7447 is fixed in OS X 10.8.4, but Apple have ruled out both hot fixing
7448 10.8..10.8.3 and forcing users to upgrade to 10.8.4 or newer.
7450 *Rob Stradling, Adam Langley*
7452 ### Changes between 1.0.0j and 1.0.0k [5 Feb 2013]
7454 * Make the decoding of SSLv3, TLS and DTLS CBC records constant time.
7456 This addresses the flaw in CBC record processing discovered by
7457 Nadhem Alfardan and Kenny Paterson. Details of this attack can be found
7458 at: <http://www.isg.rhul.ac.uk/tls/>
7460 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
7461 Security Group at Royal Holloway, University of London
7462 (www.isg.rhul.ac.uk) for discovering this flaw and Adam Langley and
7463 Emilia Käsper for the initial patch.
7466 *Emilia Käsper, Adam Langley, Ben Laurie, Andy Polyakov, Steve Henson*
7468 * Return an error when checking OCSP signatures when key is NULL.
7469 This fixes a DoS attack. [CVE-2013-0166][]
7473 * Call OCSP Stapling callback after ciphersuite has been chosen, so
7474 the right response is stapled. Also change SSL_get_certificate()
7475 so it returns the certificate actually sent.
7476 See <http://rt.openssl.org/Ticket/Display.html?id=2836>.
7477 (This is a backport)
7479 *Rob Stradling <rob.stradling@comodo.com>*
7481 * Fix possible deadlock when decoding public keys.
7485 ### Changes between 1.0.0i and 1.0.0j [10 May 2012]
7487 [NB: OpenSSL 1.0.0i and later 1.0.0 patch levels were released after
7490 * Sanity check record length before skipping explicit IV in DTLS
7493 Thanks to Codenomicon for discovering this issue using Fuzz-o-Matic
7494 fuzzing as a service testing platform.
7499 * Initialise tkeylen properly when encrypting CMS messages.
7500 Thanks to Solar Designer of Openwall for reporting this issue.
7504 ### Changes between 1.0.0h and 1.0.0i [19 Apr 2012]
7506 * Check for potentially exploitable overflows in asn1_d2i_read_bio
7507 BUF_mem_grow and BUF_mem_grow_clean. Refuse attempts to shrink buffer
7508 in CRYPTO_realloc_clean.
7510 Thanks to Tavis Ormandy, Google Security Team, for discovering this
7511 issue and to Adam Langley <agl@chromium.org> for fixing it.
7514 *Adam Langley (Google), Tavis Ormandy, Google Security Team*
7516 ### Changes between 1.0.0g and 1.0.0h [12 Mar 2012]
7518 * Fix MMA (Bleichenbacher's attack on PKCS #1 v1.5 RSA padding) weakness
7519 in CMS and PKCS7 code. When RSA decryption fails use a random key for
7520 content decryption and always return the same error. Note: this attack
7521 needs on average 2^20 messages so it only affects automated senders. The
7522 old behaviour can be re-enabled in the CMS code by setting the
7523 CMS_DEBUG_DECRYPT flag: this is useful for debugging and testing where
7524 an MMA defence is not necessary.
7525 Thanks to Ivan Nestlerode <inestlerode@us.ibm.com> for discovering
7526 this issue. [CVE-2012-0884][]
7530 * Fix CVE-2011-4619: make sure we really are receiving a
7531 client hello before rejecting multiple SGC restarts. Thanks to
7532 Ivan Nestlerode <inestlerode@us.ibm.com> for discovering this bug.
7536 ### Changes between 1.0.0f and 1.0.0g [18 Jan 2012]
7538 * Fix for DTLS DoS issue introduced by fix for CVE-2011-4109.
7539 Thanks to Antonio Martin, Enterprise Secure Access Research and
7540 Development, Cisco Systems, Inc. for discovering this bug and
7541 preparing a fix. [CVE-2012-0050][]
7545 ### Changes between 1.0.0e and 1.0.0f [4 Jan 2012]
7547 * Nadhem Alfardan and Kenny Paterson have discovered an extension
7548 of the Vaudenay padding oracle attack on CBC mode encryption
7549 which enables an efficient plaintext recovery attack against
7550 the OpenSSL implementation of DTLS. Their attack exploits timing
7551 differences arising during decryption processing. A research
7552 paper describing this attack can be found at:
7553 <http://www.isg.rhul.ac.uk/~kp/dtls.pdf>
7554 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
7555 Security Group at Royal Holloway, University of London
7556 (www.isg.rhul.ac.uk) for discovering this flaw and to Robin Seggelmann
7557 <seggelmann@fh-muenster.de> and Michael Tuexen <tuexen@fh-muenster.de>
7558 for preparing the fix. [CVE-2011-4108][]
7560 *Robin Seggelmann, Michael Tuexen*
7562 * Clear bytes used for block padding of SSL 3.0 records.
7565 *Adam Langley (Google)*
7567 * Only allow one SGC handshake restart for SSL/TLS. Thanks to George
7568 Kadianakis <desnacked@gmail.com> for discovering this issue and
7569 Adam Langley for preparing the fix. [CVE-2011-4619][]
7571 *Adam Langley (Google)*
7573 * Check parameters are not NULL in GOST ENGINE. [CVE-2012-0027][]
7575 *Andrey Kulikov <amdeich@gmail.com>*
7577 * Prevent malformed RFC3779 data triggering an assertion failure.
7578 Thanks to Andrew Chi, BBN Technologies, for discovering the flaw
7579 and Rob Austein <sra@hactrn.net> for fixing it. [CVE-2011-4577][]
7581 *Rob Austein <sra@hactrn.net>*
7583 * Improved PRNG seeding for VOS.
7585 *Paul Green <Paul.Green@stratus.com>*
7587 * Fix ssl_ciph.c set-up race.
7589 *Adam Langley (Google)*
7591 * Fix spurious failures in ecdsatest.c.
7593 *Emilia Käsper (Google)*
7595 * Fix the BIO_f_buffer() implementation (which was mixing different
7596 interpretations of the `..._len` fields).
7598 *Adam Langley (Google)*
7600 * Fix handling of BN_BLINDING: now BN_BLINDING_invert_ex (rather than
7601 BN_BLINDING_invert_ex) calls BN_BLINDING_update, ensuring that concurrent
7602 threads won't reuse the same blinding coefficients.
7604 This also avoids the need to obtain the CRYPTO_LOCK_RSA_BLINDING
7605 lock to call BN_BLINDING_invert_ex, and avoids one use of
7606 BN_BLINDING_update for each BN_BLINDING structure (previously,
7607 the last update always remained unused).
7609 *Emilia Käsper (Google)*
7611 * In ssl3_clear, preserve s3->init_extra along with s3->rbuf.
7613 *Bob Buckholz (Google)*
7615 ### Changes between 1.0.0d and 1.0.0e [6 Sep 2011]
7617 * Fix bug where CRLs with nextUpdate in the past are sometimes accepted
7618 by initialising X509_STORE_CTX properly. [CVE-2011-3207][]
7620 *Kaspar Brand <ossl@velox.ch>*
7622 * Fix SSL memory handling for (EC)DH ciphersuites, in particular
7623 for multi-threaded use of ECDH. [CVE-2011-3210][]
7625 *Adam Langley (Google)*
7627 * Fix x509_name_ex_d2i memory leak on bad inputs.
7631 * Remove hard coded ecdsaWithSHA1 signature tests in ssl code and check
7632 signature public key algorithm by using OID xref utilities instead.
7633 Before this you could only use some ECC ciphersuites with SHA1 only.
7637 * Add protection against ECDSA timing attacks as mentioned in the paper
7638 by Billy Bob Brumley and Nicola Tuveri, see:
7639 <http://eprint.iacr.org/2011/232.pdf>
7641 *Billy Bob Brumley and Nicola Tuveri*
7643 ### Changes between 1.0.0c and 1.0.0d [8 Feb 2011]
7645 * Fix parsing of OCSP stapling ClientHello extension. CVE-2011-0014
7647 *Neel Mehta, Adam Langley, Bodo Moeller (Google)*
7649 * Fix bug in string printing code: if *any* escaping is enabled we must
7650 escape the escape character (backslash) or the resulting string is
7655 ### Changes between 1.0.0b and 1.0.0c [2 Dec 2010]
7657 * Disable code workaround for ancient and obsolete Netscape browsers
7658 and servers: an attacker can use it in a ciphersuite downgrade attack.
7659 Thanks to Martin Rex for discovering this bug. CVE-2010-4180
7663 * Fixed J-PAKE implementation error, originally discovered by
7664 Sebastien Martini, further info and confirmation from Stefan
7665 Arentz and Feng Hao. Note that this fix is a security fix. CVE-2010-4252
7669 ### Changes between 1.0.0a and 1.0.0b [16 Nov 2010]
7671 * Fix extension code to avoid race conditions which can result in a buffer
7672 overrun vulnerability: resumed sessions must not be modified as they can
7673 be shared by multiple threads. CVE-2010-3864
7677 * Fix WIN32 build system to correctly link an ENGINE directory into
7682 ### Changes between 1.0.0 and 1.0.0a [01 Jun 2010]
7684 * Check return value of int_rsa_verify in pkey_rsa_verifyrecover
7687 *Steve Henson, Peter-Michael Hager <hager@dortmund.net>*
7689 ### Changes between 0.9.8n and 1.0.0 [29 Mar 2010]
7691 * Add "missing" function EVP_CIPHER_CTX_copy(). This copies a cipher
7692 context. The operation can be customised via the ctrl mechanism in
7693 case ENGINEs want to include additional functionality.
7697 * Tolerate yet another broken PKCS#8 key format: private key value negative.
7701 * Add new -subject_hash_old and -issuer_hash_old options to x509 utility to
7702 output hashes compatible with older versions of OpenSSL.
7704 *Willy Weisz <weisz@vcpc.univie.ac.at>*
7706 * Fix compression algorithm handling: if resuming a session use the
7707 compression algorithm of the resumed session instead of determining
7708 it from client hello again. Don't allow server to change algorithm.
7712 * Add load_crls() function to apps tidying load_certs() too. Add option
7713 to verify utility to allow additional CRLs to be included.
7717 * Update OCSP request code to permit adding custom headers to the request:
7718 some responders need this.
7722 * The function EVP_PKEY_sign() returns <=0 on error: check return code
7725 *Julia Lawall <julia@diku.dk>*
7727 * Update verify callback code in apps/s_cb.c and apps/verify.c, it
7728 needlessly dereferenced structures, used obsolete functions and
7729 didn't handle all updated verify codes correctly.
7733 * Disable MD2 in the default configuration.
7737 * In BIO_pop() and BIO_push() use the ctrl argument (which was NULL) to
7738 indicate the initial BIO being pushed or popped. This makes it possible
7739 to determine whether the BIO is the one explicitly called or as a result
7740 of the ctrl being passed down the chain. Fix BIO_pop() and SSL BIOs so
7741 it handles reference counts correctly and doesn't zero out the I/O bio
7742 when it is not being explicitly popped. WARNING: applications which
7743 included workarounds for the old buggy behaviour will need to be modified
7744 or they could free up already freed BIOs.
7748 * Extend the uni2asc/asc2uni => OPENSSL_uni2asc/OPENSSL_asc2uni
7749 renaming to all platforms (within the 0.9.8 branch, this was
7750 done conditionally on Netware platforms to avoid a name clash).
7752 *Guenter <lists@gknw.net>*
7754 * Add ECDHE and PSK support to DTLS.
7756 *Michael Tuexen <tuexen@fh-muenster.de>*
7758 * Add CHECKED_STACK_OF macro to safestack.h, otherwise safestack can't
7763 * Add "missing" function EVP_MD_flags() (without this the only way to
7764 retrieve a digest flags is by accessing the structure directly. Update
7765 `EVP_MD_do_all*()` and `EVP_CIPHER_do_all*()` to include the name a digest
7766 or cipher is registered as in the "from" argument. Print out all
7767 registered digests in the dgst usage message instead of manually
7768 attempting to work them out.
7772 * If no SSLv2 ciphers are used don't use an SSLv2 compatible client hello:
7773 this allows the use of compression and extensions. Change default cipher
7774 string to remove SSLv2 ciphersuites. This effectively avoids ancient SSLv2
7775 by default unless an application cipher string requests it.
7779 * Alter match criteria in PKCS12_parse(). It used to try to use local
7780 key ids to find matching certificates and keys but some PKCS#12 files
7781 don't follow the (somewhat unwritten) rules and this strategy fails.
7782 Now just gather all certificates together and the first private key
7783 then look for the first certificate that matches the key.
7787 * Support use of registered digest and cipher names for dgst and cipher
7788 commands instead of having to add each one as a special case. So now
7795 openssl dgst -sha256 foo
7797 and this works for ENGINE based algorithms too.
7801 * Update Gost ENGINE to support parameter files.
7803 *Victor B. Wagner <vitus@cryptocom.ru>*
7805 * Support GeneralizedTime in ca utility.
7807 *Oliver Martin <oliver@volatilevoid.net>, Steve Henson*
7809 * Enhance the hash format used for certificate directory links. The new
7810 form uses the canonical encoding (meaning equivalent names will work
7811 even if they aren't identical) and uses SHA1 instead of MD5. This form
7812 is incompatible with the older format and as a result c_rehash should
7813 be used to rebuild symbolic links.
7817 * Make PKCS#8 the default write format for private keys, replacing the
7818 traditional format. This form is standardised, more secure and doesn't
7819 include an implicit MD5 dependency.
7823 * Add a $gcc_devteam_warn option to Configure. The idea is that any code
7824 committed to OpenSSL should pass this lot as a minimum.
7828 * Add session ticket override functionality for use by EAP-FAST.
7830 *Jouni Malinen <j@w1.fi>*
7832 * Modify HMAC functions to return a value. Since these can be implemented
7833 in an ENGINE errors can occur.
7837 * Type-checked OBJ_bsearch_ex.
7841 * Type-checked OBJ_bsearch. Also some constification necessitated
7842 by type-checking. Still to come: TXT_DB, bsearch(?),
7843 OBJ_bsearch_ex, qsort, CRYPTO_EX_DATA, ASN1_VALUE, ASN1_STRING,
7848 * New function OPENSSL_gmtime_adj() to add a specific number of days and
7849 seconds to a tm structure directly, instead of going through OS
7850 specific date routines. This avoids any issues with OS routines such
7851 as the year 2038 bug. New `*_adj()` functions for ASN1 time structures
7852 and X509_time_adj_ex() to cover the extended range. The existing
7853 X509_time_adj() is still usable and will no longer have any date issues.
7857 * Delta CRL support. New use deltas option which will attempt to locate
7858 and search any appropriate delta CRLs available.
7860 This work was sponsored by Google.
7864 * Support for CRLs partitioned by reason code. Reorganise CRL processing
7865 code and add additional score elements. Validate alternate CRL paths
7866 as part of the CRL checking and indicate a new error "CRL path validation
7867 error" in this case. Applications wanting additional details can use
7868 the verify callback and check the new "parent" field. If this is not
7869 NULL CRL path validation is taking place. Existing applications won't
7870 see this because it requires extended CRL support which is off by
7873 This work was sponsored by Google.
7877 * Support for freshest CRL extension.
7879 This work was sponsored by Google.
7883 * Initial indirect CRL support. Currently only supported in the CRLs
7884 passed directly and not via lookup. Process certificate issuer
7885 CRL entry extension and lookup CRL entries by bother issuer name
7886 and serial number. Check and process CRL issuer entry in IDP extension.
7888 This work was sponsored by Google.
7892 * Add support for distinct certificate and CRL paths. The CRL issuer
7893 certificate is validated separately in this case. Only enabled if
7894 an extended CRL support flag is set: this flag will enable additional
7895 CRL functionality in future.
7897 This work was sponsored by Google.
7901 * Add support for policy mappings extension.
7903 This work was sponsored by Google.
7907 * Fixes to pathlength constraint, self issued certificate handling,
7908 policy processing to align with RFC3280 and PKITS tests.
7910 This work was sponsored by Google.
7914 * Support for name constraints certificate extension. DN, email, DNS
7915 and URI types are currently supported.
7917 This work was sponsored by Google.
7921 * To cater for systems that provide a pointer-based thread ID rather
7922 than numeric, deprecate the current numeric thread ID mechanism and
7923 replace it with a structure and associated callback type. This
7924 mechanism allows a numeric "hash" to be extracted from a thread ID in
7925 either case, and on platforms where pointers are larger than 'long',
7926 mixing is done to help ensure the numeric 'hash' is usable even if it
7927 can't be guaranteed unique. The default mechanism is to use "&errno"
7928 as a pointer-based thread ID to distinguish between threads.
7930 Applications that want to provide their own thread IDs should now use
7931 CRYPTO_THREADID_set_callback() to register a callback that will call
7932 either CRYPTO_THREADID_set_numeric() or CRYPTO_THREADID_set_pointer().
7934 Note that ERR_remove_state() is now deprecated, because it is tied
7935 to the assumption that thread IDs are numeric. ERR_remove_state(0)
7936 to free the current thread's error state should be replaced by
7937 ERR_remove_thread_state(NULL).
7939 (This new approach replaces the functions CRYPTO_set_idptr_callback(),
7940 CRYPTO_get_idptr_callback(), and CRYPTO_thread_idptr() that existed in
7941 OpenSSL 0.9.9-dev between June 2006 and August 2008. Also, if an
7942 application was previously providing a numeric thread callback that
7943 was inappropriate for distinguishing threads, then uniqueness might
7944 have been obtained with &errno that happened immediately in the
7945 intermediate development versions of OpenSSL; this is no longer the
7946 case, the numeric thread callback will now override the automatic use
7949 *Geoff Thorpe, with help from Bodo Moeller*
7951 * Initial support for different CRL issuing certificates. This covers a
7952 simple case where the self issued certificates in the chain exist and
7953 the real CRL issuer is higher in the existing chain.
7955 This work was sponsored by Google.
7959 * Removed effectively defunct crypto/store from the build.
7963 * Revamp of STACK to provide stronger type-checking. Still to come:
7964 TXT_DB, bsearch(?), OBJ_bsearch, qsort, CRYPTO_EX_DATA, ASN1_VALUE,
7965 ASN1_STRING, CONF_VALUE.
7969 * Add a new SSL_MODE_RELEASE_BUFFERS mode flag to release unused buffer
7970 RAM on SSL connections. This option can save about 34k per idle SSL.
7974 * Revamp of LHASH to provide stronger type-checking. Still to come:
7975 STACK, TXT_DB, bsearch, qsort.
7979 * Initial support for Cryptographic Message Syntax (aka CMS) based
7980 on RFC3850, RFC3851 and RFC3852. New cms directory and cms utility,
7981 support for data, signedData, compressedData, digestedData and
7982 encryptedData, envelopedData types included. Scripts to check against
7983 RFC4134 examples draft and interop and consistency checks of many
7984 content types and variants.
7988 * Add options to enc utility to support use of zlib compression BIO.
7992 * Extend mk1mf to support importing of options and assembly language
7993 files from Configure script, currently only included in VC-WIN32.
7994 The assembly language rules can now optionally generate the source
7995 files from the associated perl scripts.
7999 * Implement remaining functionality needed to support GOST ciphersuites.
8000 Interop testing has been performed using CryptoPro implementations.
8002 *Victor B. Wagner <vitus@cryptocom.ru>*
8004 * s390x assembler pack.
8008 * ARMv4 assembler pack. ARMv4 refers to v4 and later ISA, not CPU
8013 * Implement Opaque PRF Input TLS extension as specified in
8014 draft-rescorla-tls-opaque-prf-input-00.txt. Since this is not an
8015 official specification yet and no extension type assignment by
8016 IANA exists, this extension (for now) will have to be explicitly
8017 enabled when building OpenSSL by providing the extension number
8018 to use. For example, specify an option
8020 -DTLSEXT_TYPE_opaque_prf_input=0x9527
8022 to the "config" or "Configure" script to enable the extension,
8023 assuming extension number 0x9527 (which is a completely arbitrary
8024 and unofficial assignment based on the MD5 hash of the Internet
8025 Draft). Note that by doing so, you potentially lose
8026 interoperability with other TLS implementations since these might
8027 be using the same extension number for other purposes.
8029 SSL_set_tlsext_opaque_prf_input(ssl, src, len) is used to set the
8030 opaque PRF input value to use in the handshake. This will create
8031 an internal copy of the length-'len' string at 'src', and will
8032 return non-zero for success.
8034 To get more control and flexibility, provide a callback function
8037 SSL_CTX_set_tlsext_opaque_prf_input_callback(ctx, cb)
8038 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg(ctx, arg)
8042 int (*cb)(SSL *, void *peerinput, size_t len, void *arg);
8045 Callback function 'cb' will be called in handshakes, and is
8046 expected to use SSL_set_tlsext_opaque_prf_input() as appropriate.
8047 Argument 'arg' is for application purposes (the value as given to
8048 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg() will directly
8049 be provided to the callback function). The callback function
8050 has to return non-zero to report success: usually 1 to use opaque
8051 PRF input just if possible, or 2 to enforce use of the opaque PRF
8052 input. In the latter case, the library will abort the handshake
8053 if opaque PRF input is not successfully negotiated.
8055 Arguments 'peerinput' and 'len' given to the callback function
8056 will always be NULL and 0 in the case of a client. A server will
8057 see the client's opaque PRF input through these variables if
8058 available (NULL and 0 otherwise). Note that if the server
8059 provides an opaque PRF input, the length must be the same as the
8060 length of the client's opaque PRF input.
8062 Note that the callback function will only be called when creating
8063 a new session (session resumption can resume whatever was
8064 previously negotiated), and will not be called in SSL 2.0
8065 handshakes; thus, SSL_CTX_set_options(ctx, SSL_OP_NO_SSLv2) or
8066 SSL_set_options(ssl, SSL_OP_NO_SSLv2) is especially recommended
8067 for applications that need to enforce opaque PRF input.
8071 * Update ssl code to support digests other than SHA1+MD5 for handshake
8074 *Victor B. Wagner <vitus@cryptocom.ru>*
8076 * Add RFC4507 support to OpenSSL. This includes the corrections in
8077 RFC4507bis. The encrypted ticket format is an encrypted encoded
8078 SSL_SESSION structure, that way new session features are automatically
8081 If a client application caches session in an SSL_SESSION structure
8082 support is transparent because tickets are now stored in the encoded
8085 The SSL_CTX structure automatically generates keys for ticket
8086 protection in servers so again support should be possible
8087 with no application modification.
8089 If a client or server wishes to disable RFC4507 support then the option
8090 SSL_OP_NO_TICKET can be set.
8092 Add a TLS extension debugging callback to allow the contents of any client
8093 or server extensions to be examined.
8095 This work was sponsored by Google.
8099 * Final changes to avoid use of pointer pointer casts in OpenSSL.
8100 OpenSSL should now compile cleanly on gcc 4.2
8102 *Peter Hartley <pdh@utter.chaos.org.uk>, Steve Henson*
8104 * Update SSL library to use new EVP_PKEY MAC API. Include generic MAC
8105 support including streaming MAC support: this is required for GOST
8106 ciphersuite support.
8108 *Victor B. Wagner <vitus@cryptocom.ru>, Steve Henson*
8110 * Add option -stream to use PKCS#7 streaming in smime utility. New
8111 function i2d_PKCS7_bio_stream() and PEM_write_PKCS7_bio_stream()
8112 to output in BER and PEM format.
8116 * Experimental support for use of HMAC via EVP_PKEY interface. This
8117 allows HMAC to be handled via the `EVP_DigestSign*()` interface. The
8118 EVP_PKEY "key" in this case is the HMAC key, potentially allowing
8119 ENGINE support for HMAC keys which are unextractable. New -mac and
8120 -macopt options to dgst utility.
8124 * New option -sigopt to dgst utility. Update dgst to use
8125 `EVP_Digest{Sign,Verify}*`. These two changes make it possible to use
8126 alternative signing parameters such as X9.31 or PSS in the dgst
8131 * Change ssl_cipher_apply_rule(), the internal function that does
8132 the work each time a ciphersuite string requests enabling
8133 ("foo+bar"), moving ("+foo+bar"), disabling ("-foo+bar", or
8134 removing ("!foo+bar") a class of ciphersuites: Now it maintains
8135 the order of disabled ciphersuites such that those ciphersuites
8136 that most recently went from enabled to disabled not only stay
8137 in order with respect to each other, but also have higher priority
8138 than other disabled ciphersuites the next time ciphersuites are
8141 This means that you can now say, e.g., "PSK:-PSK:HIGH" to enable
8142 the same ciphersuites as with "HIGH" alone, but in a specific
8143 order where the PSK ciphersuites come first (since they are the
8144 most recently disabled ciphersuites when "HIGH" is parsed).
8146 Also, change ssl_create_cipher_list() (using this new
8147 functionality) such that between otherwise identical
8148 ciphersuites, ephemeral ECDH is preferred over ephemeral DH in
8153 * Change ssl_create_cipher_list() so that it automatically
8154 arranges the ciphersuites in reasonable order before starting
8155 to process the rule string. Thus, the definition for "DEFAULT"
8156 (SSL_DEFAULT_CIPHER_LIST) now is just "ALL:!aNULL:!eNULL", but
8157 remains equivalent to "AES:ALL:!aNULL:!eNULL:+aECDH:+kRSA:+RC4:@STRENGTH".
8158 This makes it much easier to arrive at a reasonable default order
8159 in applications for which anonymous ciphers are OK (meaning
8160 that you can't actually use DEFAULT).
8162 *Bodo Moeller; suggested by Victor Duchovni*
8164 * Split the SSL/TLS algorithm mask (as used for ciphersuite string
8165 processing) into multiple integers instead of setting
8166 "SSL_MKEY_MASK" bits, "SSL_AUTH_MASK" bits, "SSL_ENC_MASK",
8167 "SSL_MAC_MASK", and "SSL_SSL_MASK" bits all in a single integer.
8168 (These masks as well as the individual bit definitions are hidden
8169 away into the non-exported interface ssl/ssl_locl.h, so this
8170 change to the definition of the SSL_CIPHER structure shouldn't
8171 affect applications.) This give us more bits for each of these
8172 categories, so there is no longer a need to coagulate AES128 and
8173 AES256 into a single algorithm bit, and to coagulate Camellia128
8174 and Camellia256 into a single algorithm bit, which has led to all
8177 Thus, among other things, the kludge introduced in 0.9.7m and
8178 0.9.8e for masking out AES256 independently of AES128 or masking
8179 out Camellia256 independently of AES256 is not needed here in 0.9.9.
8181 With the change, we also introduce new ciphersuite aliases that
8182 so far were missing: "AES128", "AES256", "CAMELLIA128", and
8187 * Add support for dsa-with-SHA224 and dsa-with-SHA256.
8188 Use the leftmost N bytes of the signature input if the input is
8189 larger than the prime q (with N being the size in bytes of q).
8193 * Very *very* experimental PKCS#7 streaming encoder support. Nothing uses
8194 it yet and it is largely untested.
8198 * Add support for the ecdsa-with-SHA224/256/384/512 signature types.
8202 * Initial incomplete changes to avoid need for function casts in OpenSSL
8203 some compilers (gcc 4.2 and later) reject their use. Safestack is
8204 reimplemented. Update ASN1 to avoid use of legacy functions.
8208 * Win32/64 targets are linked with Winsock2.
8212 * Add an X509_CRL_METHOD structure to allow CRL processing to be redirected
8213 to external functions. This can be used to increase CRL handling
8214 efficiency especially when CRLs are very large by (for example) storing
8215 the CRL revoked certificates in a database.
8219 * Overhaul of by_dir code. Add support for dynamic loading of CRLs so
8220 new CRLs added to a directory can be used. New command line option
8221 -verify_return_error to s_client and s_server. This causes real errors
8222 to be returned by the verify callback instead of carrying on no matter
8223 what. This reflects the way a "real world" verify callback would behave.
8227 * GOST engine, supporting several GOST algorithms and public key formats.
8228 Kindly donated by Cryptocom.
8232 * Partial support for Issuing Distribution Point CRL extension. CRLs
8233 partitioned by DP are handled but no indirect CRL or reason partitioning
8234 (yet). Complete overhaul of CRL handling: now the most suitable CRL is
8235 selected via a scoring technique which handles IDP and AKID in CRLs.
8239 * New X509_STORE_CTX callbacks lookup_crls() and lookup_certs() which
8240 will ultimately be used for all verify operations: this will remove the
8241 X509_STORE dependency on certificate verification and allow alternative
8242 lookup methods. X509_STORE based implementations of these two callbacks.
8246 * Allow multiple CRLs to exist in an X509_STORE with matching issuer names.
8247 Modify get_crl() to find a valid (unexpired) CRL if possible.
8251 * New function X509_CRL_match() to check if two CRLs are identical. Normally
8252 this would be called X509_CRL_cmp() but that name is already used by
8253 a function that just compares CRL issuer names. Cache several CRL
8254 extensions in X509_CRL structure and cache CRLDP in X509.
8258 * Store a "canonical" representation of X509_NAME structure (ASN1 Name)
8259 this maps equivalent X509_NAME structures into a consistent structure.
8260 Name comparison can then be performed rapidly using memcmp().
8264 * Non-blocking OCSP request processing. Add -timeout option to ocsp
8269 * Allow digests to supply their own micalg string for S/MIME type using
8270 the ctrl EVP_MD_CTRL_MICALG.
8274 * During PKCS7 signing pass the PKCS7 SignerInfo structure to the
8275 EVP_PKEY_METHOD before and after signing via the EVP_PKEY_CTRL_PKCS7_SIGN
8276 ctrl. It can then customise the structure before and/or after signing
8281 * New function OBJ_add_sigid() to allow application defined signature OIDs
8282 to be added to OpenSSLs internal tables. New function OBJ_sigid_free()
8283 to free up any added signature OIDs.
8287 * New functions EVP_CIPHER_do_all(), EVP_CIPHER_do_all_sorted(),
8288 EVP_MD_do_all() and EVP_MD_do_all_sorted() to enumerate internal
8289 digest and cipher tables. New options added to openssl utility:
8290 list-message-digest-algorithms and list-cipher-algorithms.
8294 * Change the array representation of binary polynomials: the list
8295 of degrees of non-zero coefficients is now terminated with -1.
8296 Previously it was terminated with 0, which was also part of the
8297 value; thus, the array representation was not applicable to
8298 polynomials where t^0 has coefficient zero. This change makes
8299 the array representation useful in a more general context.
8303 * Various modifications and fixes to SSL/TLS cipher string
8304 handling. For ECC, the code now distinguishes between fixed ECDH
8305 with RSA certificates on the one hand and with ECDSA certificates
8306 on the other hand, since these are separate ciphersuites. The
8307 unused code for Fortezza ciphersuites has been removed.
8309 For consistency with EDH, ephemeral ECDH is now called "EECDH"
8310 (not "ECDHE"). For consistency with the code for DH
8311 certificates, use of ECDH certificates is now considered ECDH
8312 authentication, not RSA or ECDSA authentication (the latter is
8313 merely the CA's signing algorithm and not actively used in the
8316 The temporary ciphersuite alias "ECCdraft" is no longer
8317 available, and ECC ciphersuites are no longer excluded from "ALL"
8318 and "DEFAULT". The following aliases now exist for RFC 4492
8319 ciphersuites, most of these by analogy with the DH case:
8321 kECDHr - ECDH cert, signed with RSA
8322 kECDHe - ECDH cert, signed with ECDSA
8323 kECDH - ECDH cert (signed with either RSA or ECDSA)
8324 kEECDH - ephemeral ECDH
8325 ECDH - ECDH cert or ephemeral ECDH
8331 AECDH - anonymous ECDH
8332 EECDH - non-anonymous ephemeral ECDH (equivalent to "kEECDH:-AECDH")
8336 * Add additional S/MIME capabilities for AES and GOST ciphers if supported.
8337 Use correct micalg parameters depending on digest(s) in signed message.
8341 * Add engine support for EVP_PKEY_ASN1_METHOD. Add functions to process
8342 an ENGINE asn1 method. Support ENGINE lookups in the ASN1 code.
8346 * Initial engine support for EVP_PKEY_METHOD. New functions to permit
8347 an engine to register a method. Add ENGINE lookups for methods and
8348 functional reference processing.
8352 * New functions `EVP_Digest{Sign,Verify)*`. These are enhanced versions of
8353 `EVP_{Sign,Verify}*` which allow an application to customise the signature
8358 * New -resign option to smime utility. This adds one or more signers
8359 to an existing PKCS#7 signedData structure. Also -md option to use an
8360 alternative message digest algorithm for signing.
8364 * Tidy up PKCS#7 routines and add new functions to make it easier to
8365 create PKCS7 structures containing multiple signers. Update smime
8366 application to support multiple signers.
8370 * New -macalg option to pkcs12 utility to allow setting of an alternative
8375 * Initial support for PKCS#5 v2.0 PRFs other than default SHA1 HMAC.
8376 Reorganize PBE internals to lookup from a static table using NIDs,
8377 add support for HMAC PBE OID translation. Add a EVP_CIPHER ctrl:
8378 EVP_CTRL_PBE_PRF_NID this allows a cipher to specify an alternative
8379 PRF which will be automatically used with PBES2.
8383 * Replace the algorithm specific calls to generate keys in "req" with the
8388 * Update PKCS#7 enveloped data routines to use new API. This is now
8389 supported by any public key method supporting the encrypt operation. A
8390 ctrl is added to allow the public key algorithm to examine or modify
8391 the PKCS#7 RecipientInfo structure if it needs to: for RSA this is
8396 * Add a ctrl to asn1 method to allow a public key algorithm to express
8397 a default digest type to use. In most cases this will be SHA1 but some
8398 algorithms (such as GOST) need to specify an alternative digest. The
8399 return value indicates how strong the preference is 1 means optional and
8400 2 is mandatory (that is it is the only supported type). Modify
8401 ASN1_item_sign() to accept a NULL digest argument to indicate it should
8402 use the default md. Update openssl utilities to use the default digest
8403 type for signing if it is not explicitly indicated.
8407 * Use OID cross reference table in ASN1_sign() and ASN1_verify(). New
8408 EVP_MD flag EVP_MD_FLAG_PKEY_METHOD_SIGNATURE. This uses the relevant
8409 signing method from the key type. This effectively removes the link
8410 between digests and public key types.
8414 * Add an OID cross reference table and utility functions. Its purpose is to
8415 translate between signature OIDs such as SHA1WithrsaEncryption and SHA1,
8416 rsaEncryption. This will allow some of the algorithm specific hackery
8417 needed to use the correct OID to be removed.
8421 * Remove algorithm specific dependencies when setting PKCS7_SIGNER_INFO
8422 structures for PKCS7_sign(). They are now set up by the relevant public
8427 * Add provisional EC pkey method with support for ECDSA and ECDH.
8431 * Add support for key derivation (agreement) in the API, DH method and
8436 * Add DSA pkey method and DH pkey methods, extend DH ASN1 method to support
8437 public and private key formats. As a side effect these add additional
8438 command line functionality not previously available: DSA signatures can be
8439 generated and verified using pkeyutl and DH key support and generation in
8446 *Oliver Tappe <zooey@hirschkaefer.de>*
8448 * New make target "install_html_docs" installs HTML renditions of the
8451 *Oliver Tappe <zooey@hirschkaefer.de>*
8453 * New utility "genpkey" this is analogous to "genrsa" etc except it can
8454 generate keys for any algorithm. Extend and update EVP_PKEY_METHOD to
8455 support key and parameter generation and add initial key generation
8456 functionality for RSA.
8460 * Add functions for main EVP_PKEY_method operations. The undocumented
8461 functions `EVP_PKEY_{encrypt,decrypt}` have been renamed to
8462 `EVP_PKEY_{encrypt,decrypt}_old`.
8466 * Initial definitions for EVP_PKEY_METHOD. This will be a high level public
8467 key API, doesn't do much yet.
8471 * New function EVP_PKEY_asn1_get0_info() to retrieve information about
8472 public key algorithms. New option to openssl utility:
8473 "list-public-key-algorithms" to print out info.
8477 * Implement the Supported Elliptic Curves Extension for
8478 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
8482 * Don't free up OIDs in OBJ_cleanup() if they are in use by EVP_MD or
8483 EVP_CIPHER structures to avoid later problems in EVP_cleanup().
8487 * New utilities pkey and pkeyparam. These are similar to algorithm specific
8488 utilities such as rsa, dsa, dsaparam etc except they process any key
8493 * Transfer public key printing routines to EVP_PKEY_ASN1_METHOD. New
8494 functions EVP_PKEY_print_public(), EVP_PKEY_print_private(),
8495 EVP_PKEY_print_param() to print public key data from an EVP_PKEY
8500 * Initial support for pluggable public key ASN1.
8501 De-spaghettify the public key ASN1 handling. Move public and private
8502 key ASN1 handling to a new EVP_PKEY_ASN1_METHOD structure. Relocate
8503 algorithm specific handling to a single module within the relevant
8504 algorithm directory. Add functions to allow (near) opaque processing
8505 of public and private key structures.
8509 * Implement the Supported Point Formats Extension for
8510 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
8514 * Add initial support for RFC 4279 PSK TLS ciphersuites. Add members
8515 for the psk identity [hint] and the psk callback functions to the
8516 SSL_SESSION, SSL and SSL_CTX structure.
8519 PSK-RC4-SHA, PSK-3DES-EDE-CBC-SHA, PSK-AES128-CBC-SHA,
8523 SSL_CTX_use_psk_identity_hint
8524 SSL_get_psk_identity_hint
8525 SSL_get_psk_identity
8526 SSL_use_psk_identity_hint
8528 *Mika Kousa and Pasi Eronen of Nokia Corporation*
8530 * Add RFC 3161 compliant time stamp request creation, response generation
8531 and response verification functionality.
8533 *Zoltán Glózik <zglozik@opentsa.org>, The OpenTSA Project*
8535 * Add initial support for TLS extensions, specifically for the server_name
8536 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
8537 have new members for a host name. The SSL data structure has an
8538 additional member `SSL_CTX *initial_ctx` so that new sessions can be
8539 stored in that context to allow for session resumption, even after the
8540 SSL has been switched to a new SSL_CTX in reaction to a client's
8541 server_name extension.
8543 New functions (subject to change):
8545 SSL_get_servername()
8546 SSL_get_servername_type()
8549 New CTRL codes and macros (subject to change):
8551 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
8552 - SSL_CTX_set_tlsext_servername_callback()
8553 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
8554 - SSL_CTX_set_tlsext_servername_arg()
8555 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
8557 openssl s_client has a new '-servername ...' option.
8559 openssl s_server has new options '-servername_host ...', '-cert2 ...',
8560 '-key2 ...', '-servername_fatal' (subject to change). This allows
8561 testing the HostName extension for a specific single host name ('-cert'
8562 and '-key' remain fallbacks for handshakes without HostName
8563 negotiation). If the unrecognized_name alert has to be sent, this by
8564 default is a warning; it becomes fatal with the '-servername_fatal'
8567 *Peter Sylvester, Remy Allais, Christophe Renou*
8569 * Whirlpool hash implementation is added.
8573 * BIGNUM code on 64-bit SPARCv9 targets is switched from bn(64,64) to
8574 bn(64,32). Because of instruction set limitations it doesn't have
8575 any negative impact on performance. This was done mostly in order
8576 to make it possible to share assembler modules, such as bn_mul_mont
8577 implementations, between 32- and 64-bit builds without hassle.
8581 * Move code previously exiled into file crypto/ec/ec2_smpt.c
8582 to ec2_smpl.c, and no longer require the OPENSSL_EC_BIN_PT_COMP
8587 * New candidate for BIGNUM assembler implementation, bn_mul_mont,
8588 dedicated Montgomery multiplication procedure, is introduced.
8589 BN_MONT_CTX is modified to allow bn_mul_mont to reach for higher
8590 "64-bit" performance on certain 32-bit targets.
8594 * New option SSL_OP_NO_COMP to disable use of compression selectively
8595 in SSL structures. New SSL ctrl to set maximum send fragment size.
8596 Save memory by setting the I/O buffer sizes dynamically instead of
8597 using the maximum available value.
8601 * New option -V for 'openssl ciphers'. This prints the ciphersuite code
8602 in addition to the text details.
8606 * Very, very preliminary EXPERIMENTAL support for printing of general
8607 ASN1 structures. This currently produces rather ugly output and doesn't
8608 handle several customised structures at all.
8612 * Integrated support for PVK file format and some related formats such
8613 as MS PUBLICKEYBLOB and PRIVATEKEYBLOB. Command line switches to support
8614 these in the 'rsa' and 'dsa' utilities.
8618 * Support for PKCS#1 RSAPublicKey format on rsa utility command line.
8622 * Remove the ancient ASN1_METHOD code. This was only ever used in one
8623 place for the (very old) "NETSCAPE" format certificates which are now
8624 handled using new ASN1 code equivalents.
8628 * Let the TLSv1_method() etc. functions return a 'const' SSL_METHOD
8629 pointer and make the SSL_METHOD parameter in SSL_CTX_new,
8630 SSL_CTX_set_ssl_version and SSL_set_ssl_method 'const'.
8634 * Modify CRL distribution points extension code to print out previously
8635 unsupported fields. Enhance extension setting code to allow setting of
8640 * Add print and set support for Issuing Distribution Point CRL extension.
8644 * Change 'Configure' script to enable Camellia by default.
8651 ### Changes between 0.9.8m and 0.9.8n [24 Mar 2010]
8653 * When rejecting SSL/TLS records due to an incorrect version number, never
8654 update s->server with a new major version number. As of
8655 - OpenSSL 0.9.8m if 'short' is a 16-bit type,
8656 - OpenSSL 0.9.8f if 'short' is longer than 16 bits,
8657 the previous behavior could result in a read attempt at NULL when
8658 receiving specific incorrect SSL/TLS records once record payload
8659 protection is active. [CVE-2010-0740][]
8661 *Bodo Moeller, Adam Langley <agl@chromium.org>*
8663 * Fix for CVE-2010-0433 where some kerberos enabled versions of OpenSSL
8664 could be crashed if the relevant tables were not present (e.g. chrooted).
8666 *Tomas Hoger <thoger@redhat.com>*
8668 ### Changes between 0.9.8l and 0.9.8m [25 Feb 2010]
8670 * Always check bn_wexpand() return values for failure. [CVE-2009-3245][]
8672 *Martin Olsson, Neel Mehta*
8674 * Fix X509_STORE locking: Every 'objs' access requires a lock (to
8675 accommodate for stack sorting, always a write lock!).
8679 * On some versions of WIN32 Heap32Next is very slow. This can cause
8680 excessive delays in the RAND_poll(): over a minute. As a workaround
8681 include a time check in the inner Heap32Next loop too.
8685 * The code that handled flushing of data in SSL/TLS originally used the
8686 BIO_CTRL_INFO ctrl to see if any data was pending first. This caused
8687 the problem outlined in PR#1949. The fix suggested there however can
8688 trigger problems with buggy BIO_CTRL_WPENDING (e.g. some versions
8689 of Apache). So instead simplify the code to flush unconditionally.
8690 This should be fine since flushing with no data to flush is a no op.
8694 * Handle TLS versions 2.0 and later properly and correctly use the
8695 highest version of TLS/SSL supported. Although TLS >= 2.0 is some way
8696 off ancient servers have a habit of sticking around for a while...
8700 * Modify compression code so it frees up structures without using the
8701 ex_data callbacks. This works around a problem where some applications
8702 call CRYPTO_cleanup_all_ex_data() before application exit (e.g. when
8703 restarting) then use compression (e.g. SSL with compression) later.
8704 This results in significant per-connection memory leaks and
8705 has caused some security issues including CVE-2008-1678 and
8710 * Constify crypto/cast (i.e., <openssl/cast.h>): a CAST_KEY doesn't
8711 change when encrypting or decrypting.
8715 * Add option SSL_OP_LEGACY_SERVER_CONNECT which will allow clients to
8716 connect and renegotiate with servers which do not support RI.
8717 Until RI is more widely deployed this option is enabled by default.
8721 * Add "missing" ssl ctrls to clear options and mode.
8725 * If client attempts to renegotiate and doesn't support RI respond with
8726 a no_renegotiation alert as required by RFC5746. Some renegotiating
8727 TLS clients will continue a connection gracefully when they receive
8728 the alert. Unfortunately OpenSSL mishandled this alert and would hang
8729 waiting for a server hello which it will never receive. Now we treat a
8730 received no_renegotiation alert as a fatal error. This is because
8731 applications requesting a renegotiation might well expect it to succeed
8732 and would have no code in place to handle the server denying it so the
8733 only safe thing to do is to terminate the connection.
8737 * Add ctrl macro SSL_get_secure_renegotiation_support() which returns 1 if
8738 peer supports secure renegotiation and 0 otherwise. Print out peer
8739 renegotiation support in s_client/s_server.
8743 * Replace the highly broken and deprecated SPKAC certification method with
8744 the updated NID creation version. This should correctly handle UTF8.
8748 * Implement RFC5746. Re-enable renegotiation but require the extension
8749 as needed. Unfortunately, SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION
8750 turns out to be a bad idea. It has been replaced by
8751 SSL_OP_ALLOW_UNSAFE_LEGACY_RENEGOTIATION which can be set with
8752 SSL_CTX_set_options(). This is really not recommended unless you
8753 know what you are doing.
8755 *Eric Rescorla <ekr@networkresonance.com>, Ben Laurie, Steve Henson*
8757 * Fixes to stateless session resumption handling. Use initial_ctx when
8758 issuing and attempting to decrypt tickets in case it has changed during
8759 servername handling. Use a non-zero length session ID when attempting
8760 stateless session resumption: this makes it possible to determine if
8761 a resumption has occurred immediately after receiving server hello
8762 (several places in OpenSSL subtly assume this) instead of later in
8767 * The functions ENGINE_ctrl(), OPENSSL_isservice(),
8768 CMS_get1_RecipientRequest() and RAND_bytes() can return <=0 on error
8769 fixes for a few places where the return code is not checked
8772 *Julia Lawall <julia@diku.dk>*
8774 * Add --strict-warnings option to Configure script to include devteam
8775 warnings in other configurations.
8779 * Add support for --libdir option and LIBDIR variable in makefiles. This
8780 makes it possible to install openssl libraries in locations which
8781 have names other than "lib", for example "/usr/lib64" which some
8784 *Steve Henson, based on patch from Jeremy Utley*
8786 * Don't allow the use of leading 0x80 in OIDs. This is a violation of
8787 X690 8.9.12 and can produce some misleading textual output of OIDs.
8789 *Steve Henson, reported by Dan Kaminsky*
8791 * Delete MD2 from algorithm tables. This follows the recommendation in
8792 several standards that it is not used in new applications due to
8793 several cryptographic weaknesses. For binary compatibility reasons
8794 the MD2 API is still compiled in by default.
8798 * Add compression id to {d2i,i2d}_SSL_SESSION so it is correctly saved
8803 * Rename uni2asc and asc2uni functions to OPENSSL_uni2asc and
8804 OPENSSL_asc2uni conditionally on Netware platforms to avoid a name
8807 *Guenter <lists@gknw.net>*
8809 * Fix the server certificate chain building code to use X509_verify_cert(),
8810 it used to have an ad-hoc builder which was unable to cope with anything
8811 other than a simple chain.
8813 *David Woodhouse <dwmw2@infradead.org>, Steve Henson*
8815 * Don't check self signed certificate signatures in X509_verify_cert()
8816 by default (a flag can override this): it just wastes time without
8817 adding any security. As a useful side effect self signed root CAs
8818 with non-FIPS digests are now usable in FIPS mode.
8822 * In dtls1_process_out_of_seq_message() the check if the current message
8823 is already buffered was missing. For every new message was memory
8824 allocated, allowing an attacker to perform an denial of service attack
8825 with sending out of seq handshake messages until there is no memory
8826 left. Additionally every future message was buffered, even if the
8827 sequence number made no sense and would be part of another handshake.
8828 So only messages with sequence numbers less than 10 in advance will be
8829 buffered. [CVE-2009-1378][]
8831 *Robin Seggelmann, discovered by Daniel Mentz*
8833 * Records are buffered if they arrive with a future epoch to be
8834 processed after finishing the corresponding handshake. There is
8835 currently no limitation to this buffer allowing an attacker to perform
8836 a DOS attack with sending records with future epochs until there is no
8837 memory left. This patch adds the pqueue_size() function to determine
8838 the size of a buffer and limits the record buffer to 100 entries.
8841 *Robin Seggelmann, discovered by Daniel Mentz*
8843 * Keep a copy of frag->msg_header.frag_len so it can be used after the
8844 parent structure is freed. [CVE-2009-1379][]
8848 * Handle non-blocking I/O properly in SSL_shutdown() call.
8850 *Darryl Miles <darryl-mailinglists@netbauds.net>*
8852 * Add `2.5.4.*` OIDs
8854 *Ilya O. <vrghost@gmail.com>*
8856 ### Changes between 0.9.8k and 0.9.8l [5 Nov 2009]
8858 * Disable renegotiation completely - this fixes a severe security
8859 problem [CVE-2009-3555][] at the cost of breaking all
8860 renegotiation. Renegotiation can be re-enabled by setting
8861 SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION in s3->flags at
8862 run-time. This is really not recommended unless you know what
8867 ### Changes between 0.9.8j and 0.9.8k [25 Mar 2009]
8869 * Don't set val to NULL when freeing up structures, it is freed up by
8870 underlying code. If `sizeof(void *) > sizeof(long)` this can result in
8871 zeroing past the valid field. [CVE-2009-0789][]
8873 *Paolo Ganci <Paolo.Ganci@AdNovum.CH>*
8875 * Fix bug where return value of CMS_SignerInfo_verify_content() was not
8876 checked correctly. This would allow some invalid signed attributes to
8877 appear to verify correctly. [CVE-2009-0591][]
8879 *Ivan Nestlerode <inestlerode@us.ibm.com>*
8881 * Reject UniversalString and BMPString types with invalid lengths. This
8882 prevents a crash in ASN1_STRING_print_ex() which assumes the strings have
8883 a legal length. [CVE-2009-0590][]
8887 * Set S/MIME signing as the default purpose rather than setting it
8888 unconditionally. This allows applications to override it at the store
8893 * Permit restricted recursion of ASN1 strings. This is needed in practice
8894 to handle some structures.
8898 * Improve efficiency of mem_gets: don't search whole buffer each time
8901 *Jeremy Shapiro <jnshapir@us.ibm.com>*
8903 * New -hex option for openssl rand.
8907 * Print out UTF8String and NumericString when parsing ASN1.
8911 * Support NumericString type for name components.
8915 * Allow CC in the environment to override the automatically chosen
8916 compiler. Note that nothing is done to ensure flags work with the
8921 ### Changes between 0.9.8i and 0.9.8j [07 Jan 2009]
8923 * Properly check EVP_VerifyFinal() and similar return values
8926 *Ben Laurie, Bodo Moeller, Google Security Team*
8928 * Enable TLS extensions by default.
8932 * Allow the CHIL engine to be loaded, whether the application is
8933 multithreaded or not. (This does not release the developer from the
8934 obligation to set up the dynamic locking callbacks.)
8936 *Sander Temme <sander@temme.net>*
8938 * Use correct exit code if there is an error in dgst command.
8940 *Steve Henson; problem pointed out by Roland Dirlewanger*
8942 * Tweak Configure so that you need to say "experimental-jpake" to enable
8943 JPAKE, and need to use -DOPENSSL_EXPERIMENTAL_JPAKE in applications.
8947 * Add experimental JPAKE support, including demo authentication in
8948 s_client and s_server.
8952 * Set the comparison function in v3_addr_canonize().
8954 *Rob Austein <sra@hactrn.net>*
8956 * Add support for XMPP STARTTLS in s_client.
8958 *Philip Paeps <philip@freebsd.org>*
8960 * Change the server-side SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG behavior
8961 to ensure that even with this option, only ciphersuites in the
8962 server's preference list will be accepted. (Note that the option
8963 applies only when resuming a session, so the earlier behavior was
8964 just about the algorithm choice for symmetric cryptography.)
8968 ### Changes between 0.9.8h and 0.9.8i [15 Sep 2008]
8970 * Fix NULL pointer dereference if a DTLS server received
8971 ChangeCipherSpec as first record [CVE-2009-1386][].
8975 * Fix a state transition in s3_srvr.c and d1_srvr.c
8976 (was using SSL3_ST_CW_CLNT_HELLO_B, should be `..._ST_SW_SRVR_...`).
8980 * The fix in 0.9.8c that supposedly got rid of unsafe
8981 double-checked locking was incomplete for RSA blinding,
8982 addressing just one layer of what turns out to have been
8983 doubly unsafe triple-checked locking.
8985 So now fix this for real by retiring the MONT_HELPER macro
8986 in crypto/rsa/rsa_eay.c.
8988 *Bodo Moeller; problem pointed out by Marius Schilder*
8990 * Various precautionary measures:
8992 - Avoid size_t integer overflow in HASH_UPDATE (md32_common.h).
8994 - Avoid a buffer overflow in d2i_SSL_SESSION() (ssl_asn1.c).
8995 (NB: This would require knowledge of the secret session ticket key
8996 to exploit, in which case you'd be SOL either way.)
8998 - Change bn_nist.c so that it will properly handle input BIGNUMs
8999 outside the expected range.
9001 - Enforce the 'num' check in BN_div() (bn_div.c) for non-BN_DEBUG
9004 *Neel Mehta, Bodo Moeller*
9006 * Allow engines to be "soft loaded" - i.e. optionally don't die if
9007 the load fails. Useful for distros.
9009 *Ben Laurie and the FreeBSD team*
9011 * Add support for Local Machine Keyset attribute in PKCS#12 files.
9015 * Fix BN_GF2m_mod_arr() top-bit cleanup code.
9019 * Expand ENGINE to support engine supplied SSL client certificate functions.
9021 This work was sponsored by Logica.
9025 * Add CryptoAPI ENGINE to support use of RSA and DSA keys held in Windows
9026 keystores. Support for SSL/TLS client authentication too.
9027 Not compiled unless enable-capieng specified to Configure.
9029 This work was sponsored by Logica.
9033 * Fix bug in X509_ATTRIBUTE creation: don't set attribute using
9034 ASN1_TYPE_set1 if MBSTRING flag set. This bug would crash certain
9035 attribute creation routines such as certificate requests and PKCS#12
9040 ### Changes between 0.9.8g and 0.9.8h [28 May 2008]
9042 * Fix flaw if 'Server Key exchange message' is omitted from a TLS
9043 handshake which could lead to a client crash as found using the
9044 Codenomicon TLS test suite [CVE-2008-1672][]
9046 *Steve Henson, Mark Cox*
9048 * Fix double free in TLS server name extensions which could lead to
9049 a remote crash found by Codenomicon TLS test suite [CVE-2008-0891][]
9053 * Clear error queue in SSL_CTX_use_certificate_chain_file()
9055 Clear the error queue to ensure that error entries left from
9056 older function calls do not interfere with the correct operation.
9058 *Lutz Jaenicke, Erik de Castro Lopo*
9060 * Remove root CA certificates of commercial CAs:
9062 The OpenSSL project does not recommend any specific CA and does not
9063 have any policy with respect to including or excluding any CA.
9064 Therefore it does not make any sense to ship an arbitrary selection
9065 of root CA certificates with the OpenSSL software.
9069 * RSA OAEP patches to fix two separate invalid memory reads.
9070 The first one involves inputs when 'lzero' is greater than
9071 'SHA_DIGEST_LENGTH' (it would read about SHA_DIGEST_LENGTH bytes
9072 before the beginning of from). The second one involves inputs where
9073 the 'db' section contains nothing but zeroes (there is a one-byte
9074 invalid read after the end of 'db').
9076 *Ivan Nestlerode <inestlerode@us.ibm.com>*
9078 * Partial backport from 0.9.9-dev:
9080 Introduce bn_mul_mont (dedicated Montgomery multiplication
9081 procedure) as a candidate for BIGNUM assembler implementation.
9082 While 0.9.9-dev uses assembler for various architectures, only
9083 x86_64 is available by default here in the 0.9.8 branch, and
9084 32-bit x86 is available through a compile-time setting.
9086 To try the 32-bit x86 assembler implementation, use Configure
9087 option "enable-montasm" (which exists only for this backport).
9089 As "enable-montasm" for 32-bit x86 disclaims code stability
9090 anyway, in this constellation we activate additional code
9091 backported from 0.9.9-dev for further performance improvements,
9092 namely BN_from_montgomery_word. (To enable this otherwise,
9093 e.g. x86_64, try `-DMONT_FROM_WORD___NON_DEFAULT_0_9_8_BUILD`.)
9095 *Andy Polyakov (backport partially by Bodo Moeller)*
9097 * Add TLS session ticket callback. This allows an application to set
9098 TLS ticket cipher and HMAC keys rather than relying on hardcoded fixed
9099 values. This is useful for key rollover for example where several key
9100 sets may exist with different names.
9104 * Reverse ENGINE-internal logic for caching default ENGINE handles.
9105 This was broken until now in 0.9.8 releases, such that the only way
9106 a registered ENGINE could be used (assuming it initialises
9107 successfully on the host) was to explicitly set it as the default
9108 for the relevant algorithms. This is in contradiction with 0.9.7
9109 behaviour and the documentation. With this fix, when an ENGINE is
9110 registered into a given algorithm's table of implementations, the
9111 'uptodate' flag is reset so that auto-discovery will be used next
9112 time a new context for that algorithm attempts to select an
9115 *Ian Lister (tweaked by Geoff Thorpe)*
9117 * Backport of CMS code to OpenSSL 0.9.8. This differs from the 0.9.9
9118 implementation in the following ways:
9120 Lack of EVP_PKEY_ASN1_METHOD means algorithm parameters have to be
9123 Lack of BER streaming support means one pass streaming processing is
9124 only supported if data is detached: setting the streaming flag is
9125 ignored for embedded content.
9127 CMS support is disabled by default and must be explicitly enabled
9128 with the enable-cms configuration option.
9132 * Update the GMP engine glue to do direct copies between BIGNUM and
9133 mpz_t when openssl and GMP use the same limb size. Otherwise the
9134 existing "conversion via a text string export" trick is still used.
9136 *Paul Sheer <paulsheer@gmail.com>*
9138 * Zlib compression BIO. This is a filter BIO which compressed and
9139 uncompresses any data passed through it.
9143 * Add AES_wrap_key() and AES_unwrap_key() functions to implement
9144 RFC3394 compatible AES key wrapping.
9148 * Add utility functions to handle ASN1 structures. ASN1_STRING_set0():
9149 sets string data without copying. X509_ALGOR_set0() and
9150 X509_ALGOR_get0(): set and retrieve X509_ALGOR (AlgorithmIdentifier)
9151 data. Attribute function X509at_get0_data_by_OBJ(): retrieves data
9152 from an X509_ATTRIBUTE structure optionally checking it occurs only
9153 once. ASN1_TYPE_set1(): set and ASN1_TYPE structure copying supplied
9158 * Fix BN flag handling in RSA_eay_mod_exp() and BN_MONT_CTX_set()
9159 to get the expected BN_FLG_CONSTTIME behavior.
9161 *Bodo Moeller (Google)*
9165 - fixed wrong usage of ioctlsocket() when build for LIBC BSD sockets
9166 - fixed do_tests.pl to run the test suite with CLIB builds too (CLIB_OPT)
9167 - added some more tests to do_tests.pl
9168 - fixed RunningProcess usage so that it works with newer LIBC NDKs too
9169 - removed usage of BN_LLONG for CLIB builds to avoid runtime dependency
9170 - added new Configure targets netware-clib-bsdsock, netware-clib-gcc,
9171 netware-clib-bsdsock-gcc, netware-libc-bsdsock-gcc
9172 - various changes to netware.pl to enable gcc-cross builds on Win32
9174 - changed crypto/bio/b_sock.c to work with macro functions (CLIB BSD)
9175 - various changes to fix missing prototype warnings
9176 - fixed x86nasm.pl to create correct asm files for NASM COFF output
9177 - added AES, WHIRLPOOL and CPUID assembler code to build files
9178 - added missing AES assembler make rules to mk1mf.pl
9179 - fixed order of includes in apps/ocsp.c so that e_os.h settings apply
9181 *Guenter Knauf <eflash@gmx.net>*
9183 * Implement certificate status request TLS extension defined in RFC3546.
9184 A client can set the appropriate parameters and receive the encoded
9185 OCSP response via a callback. A server can query the supplied parameters
9186 and set the encoded OCSP response in the callback. Add simplified examples
9187 to s_client and s_server.
9191 ### Changes between 0.9.8f and 0.9.8g [19 Oct 2007]
9194 + Binary incompatibility of ssl_ctx_st structure
9195 + DTLS interoperation with non-compliant servers
9196 + Don't call get_session_cb() without proposed session
9197 + Fix ia64 assembler code
9199 *Andy Polyakov, Steve Henson*
9201 ### Changes between 0.9.8e and 0.9.8f [11 Oct 2007]
9203 * DTLS Handshake overhaul. There were longstanding issues with
9204 OpenSSL DTLS implementation, which were making it impossible for
9205 RFC 4347 compliant client to communicate with OpenSSL server.
9206 Unfortunately just fixing these incompatibilities would "cut off"
9207 pre-0.9.8f clients. To allow for hassle free upgrade post-0.9.8e
9208 server keeps tolerating non RFC compliant syntax. The opposite is
9209 not true, 0.9.8f client can not communicate with earlier server.
9210 This update even addresses CVE-2007-4995.
9214 * Changes to avoid need for function casts in OpenSSL: some compilers
9215 (gcc 4.2 and later) reject their use.
9216 *Kurt Roeckx <kurt@roeckx.be>, Peter Hartley <pdh@utter.chaos.org.uk>,
9219 * Add RFC4507 support to OpenSSL. This includes the corrections in
9220 RFC4507bis. The encrypted ticket format is an encrypted encoded
9221 SSL_SESSION structure, that way new session features are automatically
9224 If a client application caches session in an SSL_SESSION structure
9225 support is transparent because tickets are now stored in the encoded
9228 The SSL_CTX structure automatically generates keys for ticket
9229 protection in servers so again support should be possible
9230 with no application modification.
9232 If a client or server wishes to disable RFC4507 support then the option
9233 SSL_OP_NO_TICKET can be set.
9235 Add a TLS extension debugging callback to allow the contents of any client
9236 or server extensions to be examined.
9238 This work was sponsored by Google.
9242 * Add initial support for TLS extensions, specifically for the server_name
9243 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
9244 have new members for a host name. The SSL data structure has an
9245 additional member `SSL_CTX *initial_ctx` so that new sessions can be
9246 stored in that context to allow for session resumption, even after the
9247 SSL has been switched to a new SSL_CTX in reaction to a client's
9248 server_name extension.
9250 New functions (subject to change):
9252 SSL_get_servername()
9253 SSL_get_servername_type()
9256 New CTRL codes and macros (subject to change):
9258 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
9259 - SSL_CTX_set_tlsext_servername_callback()
9260 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
9261 - SSL_CTX_set_tlsext_servername_arg()
9262 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
9264 openssl s_client has a new '-servername ...' option.
9266 openssl s_server has new options '-servername_host ...', '-cert2 ...',
9267 '-key2 ...', '-servername_fatal' (subject to change). This allows
9268 testing the HostName extension for a specific single host name ('-cert'
9269 and '-key' remain fallbacks for handshakes without HostName
9270 negotiation). If the unrecognized_name alert has to be sent, this by
9271 default is a warning; it becomes fatal with the '-servername_fatal'
9274 *Peter Sylvester, Remy Allais, Christophe Renou, Steve Henson*
9276 * Add AES and SSE2 assembly language support to VC++ build.
9280 * Mitigate attack on final subtraction in Montgomery reduction.
9284 * Fix crypto/ec/ec_mult.c to work properly with scalars of value 0
9285 (which previously caused an internal error).
9289 * Squeeze another 10% out of IGE mode when in != out.
9293 * AES IGE mode speedup.
9295 *Dean Gaudet (Google)*
9297 * Add the Korean symmetric 128-bit cipher SEED (see
9298 <http://www.kisa.or.kr/kisa/seed/jsp/seed_eng.jsp>) and
9299 add SEED ciphersuites from RFC 4162:
9301 TLS_RSA_WITH_SEED_CBC_SHA = "SEED-SHA"
9302 TLS_DHE_DSS_WITH_SEED_CBC_SHA = "DHE-DSS-SEED-SHA"
9303 TLS_DHE_RSA_WITH_SEED_CBC_SHA = "DHE-RSA-SEED-SHA"
9304 TLS_DH_anon_WITH_SEED_CBC_SHA = "ADH-SEED-SHA"
9306 To minimize changes between patchlevels in the OpenSSL 0.9.8
9307 series, SEED remains excluded from compilation unless OpenSSL
9308 is configured with 'enable-seed'.
9310 *KISA, Bodo Moeller*
9312 * Mitigate branch prediction attacks, which can be practical if a
9313 single processor is shared, allowing a spy process to extract
9314 information. For detailed background information, see
9315 <http://eprint.iacr.org/2007/039> (O. Aciicmez, S. Gueron,
9316 J.-P. Seifert, "New Branch Prediction Vulnerabilities in OpenSSL
9317 and Necessary Software Countermeasures"). The core of the change
9318 are new versions BN_div_no_branch() and
9319 BN_mod_inverse_no_branch() of BN_div() and BN_mod_inverse(),
9320 respectively, which are slower, but avoid the security-relevant
9321 conditional branches. These are automatically called by BN_div()
9322 and BN_mod_inverse() if the flag BN_FLG_CONSTTIME is set for one
9323 of the input BIGNUMs. Also, BN_is_bit_set() has been changed to
9324 remove a conditional branch.
9326 BN_FLG_CONSTTIME is the new name for the previous
9327 BN_FLG_EXP_CONSTTIME flag, since it now affects more than just
9328 modular exponentiation. (Since OpenSSL 0.9.7h, setting this flag
9329 in the exponent causes BN_mod_exp_mont() to use the alternative
9330 implementation in BN_mod_exp_mont_consttime().) The old name
9331 remains as a deprecated alias.
9333 Similarly, RSA_FLAG_NO_EXP_CONSTTIME is replaced by a more general
9334 RSA_FLAG_NO_CONSTTIME flag since the RSA implementation now uses
9335 constant-time implementations for more than just exponentiation.
9336 Here too the old name is kept as a deprecated alias.
9338 BN_BLINDING_new() will now use BN_dup() for the modulus so that
9339 the BN_BLINDING structure gets an independent copy of the
9340 modulus. This means that the previous `BIGNUM *m` argument to
9341 BN_BLINDING_new() and to BN_BLINDING_create_param() now
9342 essentially becomes `const BIGNUM *m`, although we can't actually
9343 change this in the header file before 0.9.9. It allows
9344 RSA_setup_blinding() to use BN_with_flags() on the modulus to
9345 enable BN_FLG_CONSTTIME.
9347 *Matthew D Wood (Intel Corp)*
9349 * In the SSL/TLS server implementation, be strict about session ID
9350 context matching (which matters if an application uses a single
9351 external cache for different purposes). Previously,
9352 out-of-context reuse was forbidden only if SSL_VERIFY_PEER was
9353 set. This did ensure strict client verification, but meant that,
9354 with applications using a single external cache for quite
9355 different requirements, clients could circumvent ciphersuite
9356 restrictions for a given session ID context by starting a session
9357 in a different context.
9361 * Include "!eNULL" in SSL_DEFAULT_CIPHER_LIST to make sure that
9362 a ciphersuite string such as "DEFAULT:RSA" cannot enable
9363 authentication-only ciphersuites.
9367 * Update the SSL_get_shared_ciphers() fix CVE-2006-3738 which was
9368 not complete and could lead to a possible single byte overflow
9369 [CVE-2007-5135][] [Ben Laurie]
9371 ### Changes between 0.9.8d and 0.9.8e [23 Feb 2007]
9373 * Since AES128 and AES256 (and similarly Camellia128 and
9374 Camellia256) share a single mask bit in the logic of
9375 ssl/ssl_ciph.c, the code for masking out disabled ciphers needs a
9376 kludge to work properly if AES128 is available and AES256 isn't
9377 (or if Camellia128 is available and Camellia256 isn't).
9381 * Fix the BIT STRING encoding generated by crypto/ec/ec_asn1.c
9382 (within i2d_ECPrivateKey, i2d_ECPKParameters, i2d_ECParameters):
9383 When a point or a seed is encoded in a BIT STRING, we need to
9384 prevent the removal of trailing zero bits to get the proper DER
9385 encoding. (By default, crypto/asn1/a_bitstr.c assumes the case
9386 of a NamedBitList, for which trailing 0 bits need to be removed.)
9390 * Have SSL/TLS server implementation tolerate "mismatched" record
9391 protocol version while receiving ClientHello even if the
9392 ClientHello is fragmented. (The server can't insist on the
9393 particular protocol version it has chosen before the ServerHello
9394 message has informed the client about his choice.)
9398 * Add RFC 3779 support.
9400 *Rob Austein for ARIN, Ben Laurie*
9402 * Load error codes if they are not already present instead of using a
9403 static variable. This allows them to be cleanly unloaded and reloaded.
9404 Improve header file function name parsing.
9408 * extend SMTP and IMAP protocol emulation in s_client to use EHLO
9409 or CAPABILITY handshake as required by RFCs.
9413 ### Changes between 0.9.8c and 0.9.8d [28 Sep 2006]
9415 * Introduce limits to prevent malicious keys being able to
9416 cause a denial of service. [CVE-2006-2940][]
9418 *Steve Henson, Bodo Moeller*
9420 * Fix ASN.1 parsing of certain invalid structures that can result
9421 in a denial of service. [CVE-2006-2937][] [Steve Henson]
9423 * Fix buffer overflow in SSL_get_shared_ciphers() function.
9424 [CVE-2006-3738][] [Tavis Ormandy and Will Drewry, Google Security Team]
9426 * Fix SSL client code which could crash if connecting to a
9427 malicious SSLv2 server. [CVE-2006-4343][]
9429 *Tavis Ormandy and Will Drewry, Google Security Team*
9431 * Since 0.9.8b, ciphersuite strings naming explicit ciphersuites
9432 match only those. Before that, "AES256-SHA" would be interpreted
9433 as a pattern and match "AES128-SHA" too (since AES128-SHA got
9434 the same strength classification in 0.9.7h) as we currently only
9435 have a single AES bit in the ciphersuite description bitmap.
9436 That change, however, also applied to ciphersuite strings such as
9437 "RC4-MD5" that intentionally matched multiple ciphersuites --
9438 namely, SSL 2.0 ciphersuites in addition to the more common ones
9439 from SSL 3.0/TLS 1.0.
9441 So we change the selection algorithm again: Naming an explicit
9442 ciphersuite selects this one ciphersuite, and any other similar
9443 ciphersuite (same bitmap) from *other* protocol versions.
9444 Thus, "RC4-MD5" again will properly select both the SSL 2.0
9445 ciphersuite and the SSL 3.0/TLS 1.0 ciphersuite.
9447 Since SSL 2.0 does not have any ciphersuites for which the
9448 128/256 bit distinction would be relevant, this works for now.
9449 The proper fix will be to use different bits for AES128 and
9450 AES256, which would have avoided the problems from the beginning;
9451 however, bits are scarce, so we can only do this in a new release
9452 (not just a patchlevel) when we can change the SSL_CIPHER
9453 definition to split the single 'unsigned long mask' bitmap into
9454 multiple values to extend the available space.
9458 ### Changes between 0.9.8b and 0.9.8c [05 Sep 2006]
9460 * Avoid PKCS #1 v1.5 signature attack discovered by Daniel Bleichenbacher
9461 [CVE-2006-4339][] [Ben Laurie and Google Security Team]
9463 * Add AES IGE and biIGE modes.
9467 * Change the Unix randomness entropy gathering to use poll() when
9468 possible instead of select(), since the latter has some
9469 undesirable limitations.
9471 *Darryl Miles via Richard Levitte and Bodo Moeller*
9473 * Disable "ECCdraft" ciphersuites more thoroughly. Now special
9474 treatment in ssl/ssl_ciph.s makes sure that these ciphersuites
9475 cannot be implicitly activated as part of, e.g., the "AES" alias.
9476 However, please upgrade to OpenSSL 0.9.9[-dev] for
9477 non-experimental use of the ECC ciphersuites to get TLS extension
9478 support, which is required for curve and point format negotiation
9479 to avoid potential handshake problems.
9483 * Disable rogue ciphersuites:
9485 - SSLv2 0x08 0x00 0x80 ("RC4-64-MD5")
9486 - SSLv3/TLSv1 0x00 0x61 ("EXP1024-RC2-CBC-MD5")
9487 - SSLv3/TLSv1 0x00 0x60 ("EXP1024-RC4-MD5")
9489 The latter two were purportedly from
9490 draft-ietf-tls-56-bit-ciphersuites-0[01].txt, but do not really
9493 Also deactivate the remaining ciphersuites from
9494 draft-ietf-tls-56-bit-ciphersuites-01.txt. These are just as
9495 unofficial, and the ID has long expired.
9499 * Fix RSA blinding Heisenbug (problems sometimes occurred on
9500 dual-core machines) and other potential thread-safety issues.
9504 * Add the symmetric cipher Camellia (128-bit, 192-bit, 256-bit key
9505 versions), which is now available for royalty-free use
9506 (see <http://info.isl.ntt.co.jp/crypt/eng/info/chiteki.html>).
9507 Also, add Camellia TLS ciphersuites from RFC 4132.
9509 To minimize changes between patchlevels in the OpenSSL 0.9.8
9510 series, Camellia remains excluded from compilation unless OpenSSL
9511 is configured with 'enable-camellia'.
9515 * Disable the padding bug check when compression is in use. The padding
9516 bug check assumes the first packet is of even length, this is not
9517 necessarily true if compression is enabled and can result in false
9518 positives causing handshake failure. The actual bug test is ancient
9519 code so it is hoped that implementations will either have fixed it by
9520 now or any which still have the bug do not support compression.
9524 ### Changes between 0.9.8a and 0.9.8b [04 May 2006]
9526 * When applying a cipher rule check to see if string match is an explicit
9527 cipher suite and only match that one cipher suite if it is.
9531 * Link in manifests for VC++ if needed.
9533 *Austin Ziegler <halostatue@gmail.com>*
9535 * Update support for ECC-based TLS ciphersuites according to
9536 draft-ietf-tls-ecc-12.txt with proposed changes (but without
9537 TLS extensions, which are supported starting with the 0.9.9
9538 branch, not in the OpenSSL 0.9.8 branch).
9542 * New functions EVP_CIPHER_CTX_new() and EVP_CIPHER_CTX_free() to support
9543 opaque EVP_CIPHER_CTX handling.
9547 * Fixes and enhancements to zlib compression code. We now only use
9548 "zlib1.dll" and use the default `__cdecl` calling convention on Win32
9549 to conform with the standards mentioned here:
9550 <http://www.zlib.net/DLL_FAQ.txt>
9551 Static zlib linking now works on Windows and the new --with-zlib-include
9552 --with-zlib-lib options to Configure can be used to supply the location
9553 of the headers and library. Gracefully handle case where zlib library
9558 * Several fixes and enhancements to the OID generation code. The old code
9559 sometimes allowed invalid OIDs (1.X for X >= 40 for example), couldn't
9560 handle numbers larger than ULONG_MAX, truncated printing and had a
9561 non standard OBJ_obj2txt() behaviour.
9565 * Add support for building of engines under engine/ as shared libraries
9566 under VC++ build system.
9570 * Corrected the numerous bugs in the Win32 path splitter in DSO.
9571 Hopefully, we will not see any false combination of paths any more.
9575 ### Changes between 0.9.8 and 0.9.8a [11 Oct 2005]
9577 * Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING
9578 (part of SSL_OP_ALL). This option used to disable the
9579 countermeasure against man-in-the-middle protocol-version
9580 rollback in the SSL 2.0 server implementation, which is a bad
9581 idea. [CVE-2005-2969][]
9583 *Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center
9584 for Information Security, National Institute of Advanced Industrial
9585 Science and Technology [AIST], Japan)*
9587 * Add two function to clear and return the verify parameter flags.
9591 * Keep cipherlists sorted in the source instead of sorting them at
9592 runtime, thus removing the need for a lock.
9596 * Avoid some small subgroup attacks in Diffie-Hellman.
9598 *Nick Mathewson and Ben Laurie*
9600 * Add functions for well-known primes.
9604 * Extended Windows CE support.
9606 *Satoshi Nakamura and Andy Polyakov*
9608 * Initialize SSL_METHOD structures at compile time instead of during
9609 runtime, thus removing the need for a lock.
9613 * Make PKCS7_decrypt() work even if no certificate is supplied by
9614 attempting to decrypt each encrypted key in turn. Add support to
9619 ### Changes between 0.9.7h and 0.9.8 [05 Jul 2005]
9621 [NB: OpenSSL 0.9.7i and later 0.9.7 patch levels were released after
9624 * Add libcrypto.pc and libssl.pc for those who feel they need them.
9628 * Change CA.sh and CA.pl so they don't bundle the CSR and the private
9629 key into the same file any more.
9633 * Add initial support for Win64, both IA64 and AMD64/x64 flavors.
9637 * Add -utf8 command line and config file option to 'ca'.
9639 *Stefan <stf@udoma.org*
9641 * Removed the macro des_crypt(), as it seems to conflict with some
9642 libraries. Use DES_crypt().
9646 * Correct naming of the 'chil' and '4758cca' ENGINEs. This
9647 involves renaming the source and generated shared-libs for
9648 both. The engines will accept the corrected or legacy ids
9649 ('ncipher' and '4758_cca' respectively) when binding. NB,
9650 this only applies when building 'shared'.
9652 *Corinna Vinschen <vinschen@redhat.com> and Geoff Thorpe*
9654 * Add attribute functions to EVP_PKEY structure. Modify
9655 PKCS12_create() to recognize a CSP name attribute and
9656 use it. Make -CSP option work again in pkcs12 utility.
9660 * Add new functionality to the bn blinding code:
9661 - automatic re-creation of the BN_BLINDING parameters after
9662 a fixed number of uses (currently 32)
9663 - add new function for parameter creation
9664 - introduce flags to control the update behaviour of the
9665 BN_BLINDING parameters
9666 - hide BN_BLINDING structure
9667 Add a second BN_BLINDING slot to the RSA structure to improve
9668 performance when a single RSA object is shared among several
9673 * Add support for DTLS.
9675 *Nagendra Modadugu <nagendra@cs.stanford.edu> and Ben Laurie*
9677 * Add support for DER encoded private keys (SSL_FILETYPE_ASN1)
9678 to SSL_CTX_use_PrivateKey_file() and SSL_use_PrivateKey_file()
9682 * Remove buggy and incomplete DH cert support from
9683 ssl/ssl_rsa.c and ssl/s3_both.c
9687 * Use SHA-1 instead of MD5 as the default digest algorithm for
9688 the apps/openssl applications.
9692 * Compile clean with "-Wall -Wmissing-prototypes
9693 -Wstrict-prototypes -Wmissing-declarations -Werror". Currently
9694 DEBUG_SAFESTACK must also be set.
9698 * Change ./Configure so that certain algorithms can be disabled by default.
9699 The new counterpiece to "no-xxx" is "enable-xxx".
9701 The patented RC5 and MDC2 algorithms will now be disabled unless
9702 "enable-rc5" and "enable-mdc2", respectively, are specified.
9704 (IDEA remains enabled despite being patented. This is because IDEA
9705 is frequently required for interoperability, and there is no license
9706 fee for non-commercial use. As before, "no-idea" can be used to
9707 avoid this algorithm.)
9711 * Add processing of proxy certificates (see RFC 3820). This work was
9712 sponsored by KTH (The Royal Institute of Technology in Stockholm) and
9713 EGEE (Enabling Grids for E-science in Europe).
9717 * RC4 performance overhaul on modern architectures/implementations, such
9718 as Intel P4, IA-64 and AMD64.
9722 * New utility extract-section.pl. This can be used specify an alternative
9723 section number in a pod file instead of having to treat each file as
9724 a separate case in Makefile. This can be done by adding two lines to the
9727 =for comment openssl_section:XXX
9729 The blank line is mandatory.
9733 * New arguments -certform, -keyform and -pass for s_client and s_server
9734 to allow alternative format key and certificate files and passphrase
9739 * New structure X509_VERIFY_PARAM which combines current verify parameters,
9740 update associated structures and add various utility functions.
9742 Add new policy related verify parameters, include policy checking in
9743 standard verify code. Enhance 'smime' application with extra parameters
9744 to support policy checking and print out.
9748 * Add a new engine to support VIA PadLock ACE extensions in the VIA C3
9749 Nehemiah processors. These extensions support AES encryption in hardware
9750 as well as RNG (though RNG support is currently disabled).
9752 *Michal Ludvig <michal@logix.cz>, with help from Andy Polyakov*
9754 * Deprecate `BN_[get|set]_params()` functions (they were ignored internally).
9758 * New FIPS 180-2 algorithms, SHA-224/-256/-384/-512 are implemented.
9760 *Andy Polyakov and a number of other people*
9762 * Improved PowerPC platform support. Most notably BIGNUM assembler
9763 implementation contributed by IBM.
9765 *Suresh Chari, Peter Waltenberg, Andy Polyakov*
9767 * The new 'RSA_generate_key_ex' function now takes a BIGNUM for the public
9768 exponent rather than 'unsigned long'. There is a corresponding change to
9769 the new 'rsa_keygen' element of the RSA_METHOD structure.
9771 *Jelte Jansen, Geoff Thorpe*
9773 * Functionality for creating the initial serial number file is now
9774 moved from CA.pl to the 'ca' utility with a new option -create_serial.
9776 (Before OpenSSL 0.9.7e, CA.pl used to initialize the serial
9777 number file to 1, which is bound to cause problems. To avoid
9778 the problems while respecting compatibility between different 0.9.7
9779 patchlevels, 0.9.7e employed 'openssl x509 -next_serial' in
9780 CA.pl for serial number initialization. With the new release 0.9.8,
9781 we can fix the problem directly in the 'ca' utility.)
9785 * Reduced header interdependencies by declaring more opaque objects in
9786 ossl_typ.h. As a consequence, including some headers (eg. engine.h) will
9787 give fewer recursive includes, which could break lazy source code - so
9788 this change is covered by the OPENSSL_NO_DEPRECATED symbol. As always,
9789 developers should define this symbol when building and using openssl to
9790 ensure they track the recommended behaviour, interfaces, [etc], but
9791 backwards-compatible behaviour prevails when this isn't defined.
9795 * New function X509_POLICY_NODE_print() which prints out policy nodes.
9799 * Add new EVP function EVP_CIPHER_CTX_rand_key and associated functionality.
9800 This will generate a random key of the appropriate length based on the
9801 cipher context. The EVP_CIPHER can provide its own random key generation
9802 routine to support keys of a specific form. This is used in the des and
9803 3des routines to generate a key of the correct parity. Update S/MIME
9804 code to use new functions and hence generate correct parity DES keys.
9805 Add EVP_CHECK_DES_KEY #define to return an error if the key is not
9806 valid (weak or incorrect parity).
9810 * Add a local set of CRLs that can be used by X509_verify_cert() as well
9811 as looking them up. This is useful when the verified structure may contain
9812 CRLs, for example PKCS#7 signedData. Modify PKCS7_verify() to use any CRLs
9813 present unless the new PKCS7_NO_CRL flag is asserted.
9817 * Extend ASN1 oid configuration module. It now additionally accepts the
9820 shortName = some long name, 1.2.3.4
9824 * Reimplemented the BN_CTX implementation. There is now no more static
9825 limitation on the number of variables it can handle nor the depth of the
9826 "stack" handling for BN_CTX_start()/BN_CTX_end() pairs. The stack
9827 information can now expand as required, and rather than having a single
9828 static array of bignums, BN_CTX now uses a linked-list of such arrays
9829 allowing it to expand on demand whilst maintaining the usefulness of
9830 BN_CTX's "bundling".
9834 * Add a missing BN_CTX parameter to the 'rsa_mod_exp' callback in RSA_METHOD
9835 to allow all RSA operations to function using a single BN_CTX.
9839 * Preliminary support for certificate policy evaluation and checking. This
9840 is initially intended to pass the tests outlined in "Conformance Testing
9841 of Relying Party Client Certificate Path Processing Logic" v1.07.
9845 * bn_dup_expand() has been deprecated, it was introduced in 0.9.7 and
9846 remained unused and not that useful. A variety of other little bignum
9847 tweaks and fixes have also been made continuing on from the audit (see
9852 * Constify all or almost all d2i, c2i, s2i and r2i functions, along with
9853 associated ASN1, EVP and SSL functions and old ASN1 macros.
9857 * BN_zero() only needs to set 'top' and 'neg' to zero for correct results,
9858 and this should never fail. So the return value from the use of
9859 BN_set_word() (which can fail due to needless expansion) is now deprecated;
9860 if OPENSSL_NO_DEPRECATED is defined, BN_zero() is a void macro.
9864 * BN_CTX_get() should return zero-valued bignums, providing the same
9865 initialised value as BN_new().
9867 *Geoff Thorpe, suggested by Ulf Möller*
9869 * Support for inhibitAnyPolicy certificate extension.
9873 * An audit of the BIGNUM code is underway, for which debugging code is
9874 enabled when BN_DEBUG is defined. This makes stricter enforcements on what
9875 is considered valid when processing BIGNUMs, and causes execution to
9876 assert() when a problem is discovered. If BN_DEBUG_RAND is defined,
9877 further steps are taken to deliberately pollute unused data in BIGNUM
9878 structures to try and expose faulty code further on. For now, openssl will
9879 (in its default mode of operation) continue to tolerate the inconsistent
9880 forms that it has tolerated in the past, but authors and packagers should
9881 consider trying openssl and their own applications when compiled with
9882 these debugging symbols defined. It will help highlight potential bugs in
9883 their own code, and will improve the test coverage for OpenSSL itself. At
9884 some point, these tighter rules will become openssl's default to improve
9885 maintainability, though the assert()s and other overheads will remain only
9886 in debugging configurations. See bn.h for more details.
9888 *Geoff Thorpe, Nils Larsch, Ulf Möller*
9890 * BN_CTX_init() has been deprecated, as BN_CTX is an opaque structure
9891 that can only be obtained through BN_CTX_new() (which implicitly
9892 initialises it). The presence of this function only made it possible
9893 to overwrite an existing structure (and cause memory leaks).
9897 * Because of the callback-based approach for implementing LHASH as a
9898 template type, lh_insert() adds opaque objects to hash-tables and
9899 lh_doall() or lh_doall_arg() are typically used with a destructor callback
9900 to clean up those corresponding objects before destroying the hash table
9901 (and losing the object pointers). So some over-zealous constifications in
9902 LHASH have been relaxed so that lh_insert() does not take (nor store) the
9903 objects as "const" and the `lh_doall[_arg]` callback wrappers are not
9904 prototyped to have "const" restrictions on the object pointers they are
9905 given (and so aren't required to cast them away any more).
9909 * The tmdiff.h API was so ugly and minimal that our own timing utility
9910 (speed) prefers to use its own implementation. The two implementations
9911 haven't been consolidated as yet (volunteers?) but the tmdiff API has had
9912 its object type properly exposed (MS_TM) instead of casting to/from
9913 `char *`. This may still change yet if someone realises MS_TM and
9915 aren't necessarily the greatest nomenclatures - but this is what was used
9916 internally to the implementation so I've used that for now.
9920 * Ensure that deprecated functions do not get compiled when
9921 OPENSSL_NO_DEPRECATED is defined. Some "openssl" subcommands and a few of
9922 the self-tests were still using deprecated key-generation functions so
9923 these have been updated also.
9927 * Reorganise PKCS#7 code to separate the digest location functionality
9928 into PKCS7_find_digest(), digest addition into PKCS7_bio_add_digest().
9929 New function PKCS7_set_digest() to set the digest type for PKCS#7
9930 digestedData type. Add additional code to correctly generate the
9931 digestedData type and add support for this type in PKCS7 initialization
9936 * New function PKCS7_set0_type_other() this initializes a PKCS7
9937 structure of type "other".
9941 * Fix prime generation loop in crypto/bn/bn_prime.pl by making
9942 sure the loop does correctly stop and breaking ("division by zero")
9943 modulus operations are not performed. The (pre-generated) prime
9944 table crypto/bn/bn_prime.h was already correct, but it could not be
9945 re-generated on some platforms because of the "division by zero"
9946 situation in the script.
9948 *Ralf S. Engelschall*
9950 * Update support for ECC-based TLS ciphersuites according to
9951 draft-ietf-tls-ecc-03.txt: the KDF1 key derivation function with
9952 SHA-1 now is only used for "small" curves (where the
9953 representation of a field element takes up to 24 bytes); for
9954 larger curves, the field element resulting from ECDH is directly
9955 used as premaster secret.
9957 *Douglas Stebila (Sun Microsystems Laboratories)*
9959 * Add code for kP+lQ timings to crypto/ec/ectest.c, and add SEC2
9960 curve secp160r1 to the tests.
9962 *Douglas Stebila (Sun Microsystems Laboratories)*
9964 * Add the possibility to load symbols globally with DSO.
9966 *Götz Babin-Ebell <babin-ebell@trustcenter.de> via Richard Levitte*
9968 * Add the functions ERR_set_mark() and ERR_pop_to_mark() for better
9969 control of the error stack.
9973 * Add support for STORE in ENGINE.
9977 * Add the STORE type. The intention is to provide a common interface
9978 to certificate and key stores, be they simple file-based stores, or
9979 HSM-type store, or LDAP stores, or...
9980 NOTE: The code is currently UNTESTED and isn't really used anywhere.
9984 * Add a generic structure called OPENSSL_ITEM. This can be used to
9985 pass a list of arguments to any function as well as provide a way
9986 for a function to pass data back to the caller.
9990 * Add the functions BUF_strndup() and BUF_memdup(). BUF_strndup()
9991 works like BUF_strdup() but can be used to duplicate a portion of
9992 a string. The copy gets NUL-terminated. BUF_memdup() duplicates
9997 * Add the function sk_find_ex() which works like sk_find(), but will
9998 return an index to an element even if an exact match couldn't be
9999 found. The index is guaranteed to point at the element where the
10000 searched-for key would be inserted to preserve sorting order.
10004 * Add the function OBJ_bsearch_ex() which works like OBJ_bsearch() but
10005 takes an extra flags argument for optional functionality. Currently,
10006 the following flags are defined:
10008 OBJ_BSEARCH_VALUE_ON_NOMATCH
10009 This one gets OBJ_bsearch_ex() to return a pointer to the first
10010 element where the comparing function returns a negative or zero
10013 OBJ_BSEARCH_FIRST_VALUE_ON_MATCH
10014 This one gets OBJ_bsearch_ex() to return a pointer to the first
10015 element where the comparing function returns zero. This is useful
10016 if there are more than one element where the comparing function
10021 * Make it possible to create self-signed certificates with 'openssl ca'
10022 in such a way that the self-signed certificate becomes part of the
10023 CA database and uses the same mechanisms for serial number generation
10024 as all other certificate signing. The new flag '-selfsign' enables
10025 this functionality. Adapt CA.sh and CA.pl.in.
10029 * Add functionality to check the public key of a certificate request
10030 against a given private. This is useful to check that a certificate
10031 request can be signed by that key (self-signing).
10035 * Make it possible to have multiple active certificates with the same
10036 subject in the CA index file. This is done only if the keyword
10037 'unique_subject' is set to 'no' in the main CA section (default
10038 if 'CA_default') of the configuration file. The value is saved
10039 with the database itself in a separate index attribute file,
10040 named like the index file with '.attr' appended to the name.
10044 * Generate multi-valued AVAs using '+' notation in config files for
10049 * Support for nameConstraints certificate extension.
10053 * Support for policyConstraints certificate extension.
10057 * Support for policyMappings certificate extension.
10061 * Make sure the default DSA_METHOD implementation only uses its
10062 dsa_mod_exp() and/or bn_mod_exp() handlers if they are non-NULL,
10063 and change its own handlers to be NULL so as to remove unnecessary
10064 indirection. This lets alternative implementations fallback to the
10065 default implementation more easily.
10069 * Support for directoryName in GeneralName related extensions
10074 * Make it possible to link applications using Makefile.shared.
10075 Make that possible even when linking against static libraries!
10079 * Support for single pass processing for S/MIME signing. This now
10080 means that S/MIME signing can be done from a pipe, in addition
10081 cleartext signing (multipart/signed type) is effectively streaming
10082 and the signed data does not need to be all held in memory.
10084 This is done with a new flag PKCS7_STREAM. When this flag is set
10085 PKCS7_sign() only initializes the PKCS7 structure and the actual signing
10086 is done after the data is output (and digests calculated) in
10087 SMIME_write_PKCS7().
10091 * Add full support for -rpath/-R, both in shared libraries and
10092 applications, at least on the platforms where it's known how
10097 * In crypto/ec/ec_mult.c, implement fast point multiplication with
10098 precomputation, based on wNAF splitting: EC_GROUP_precompute_mult()
10099 will now compute a table of multiples of the generator that
10100 makes subsequent invocations of EC_POINTs_mul() or EC_POINT_mul()
10101 faster (notably in the case of a single point multiplication,
10102 scalar * generator).
10104 *Nils Larsch, Bodo Moeller*
10106 * IPv6 support for certificate extensions. The various extensions
10107 which use the IP:a.b.c.d can now take IPv6 addresses using the
10108 formats of RFC1884 2.2 . IPv6 addresses are now also displayed
10113 * Added an ENGINE that implements RSA by performing private key
10114 exponentiations with the GMP library. The conversions to and from
10115 GMP's mpz_t format aren't optimised nor are any montgomery forms
10116 cached, and on x86 it appears OpenSSL's own performance has caught up.
10117 However there are likely to be other architectures where GMP could
10118 provide a boost. This ENGINE is not built in by default, but it can be
10119 specified at Configure time and should be accompanied by the necessary
10120 linker additions, eg;
10121 ./config -DOPENSSL_USE_GMP -lgmp
10125 * "openssl engine" will not display ENGINE/DSO load failure errors when
10126 testing availability of engines with "-t" - the old behaviour is
10127 produced by increasing the feature's verbosity with "-tt".
10131 * ECDSA routines: under certain error conditions uninitialized BN objects
10132 could be freed. Solution: make sure initialization is performed early
10133 enough. (Reported and fix supplied by Nils Larsch <nla@trustcenter.de>
10138 * Key-generation can now be implemented in RSA_METHOD, DSA_METHOD
10139 and DH_METHOD (eg. by ENGINE implementations) to override the normal
10140 software implementations. For DSA and DH, parameter generation can
10141 also be overridden by providing the appropriate method callbacks.
10145 * Change the "progress" mechanism used in key-generation and
10146 primality testing to functions that take a new BN_GENCB pointer in
10147 place of callback/argument pairs. The new API functions have `_ex`
10148 postfixes and the older functions are reimplemented as wrappers for
10149 the new ones. The OPENSSL_NO_DEPRECATED symbol can be used to hide
10150 declarations of the old functions to help (graceful) attempts to
10151 migrate to the new functions. Also, the new key-generation API
10152 functions operate on a caller-supplied key-structure and return
10153 success/failure rather than returning a key or NULL - this is to
10154 help make "keygen" another member function of RSA_METHOD etc.
10156 Example for using the new callback interface:
10158 int (*my_callback)(int a, int b, BN_GENCB *cb) = ...;
10159 void *my_arg = ...;
10162 BN_GENCB_set(&my_cb, my_callback, my_arg);
10164 return BN_is_prime_ex(some_bignum, BN_prime_checks, NULL, &cb);
10165 /* For the meaning of a, b in calls to my_callback(), see the
10166 * documentation of the function that calls the callback.
10167 * cb will point to my_cb; my_arg can be retrieved as cb->arg.
10168 * my_callback should return 1 if it wants BN_is_prime_ex()
10169 * to continue, or 0 to stop.
10174 * Change the ZLIB compression method to be stateful, and make it
10175 available to TLS with the number defined in
10176 draft-ietf-tls-compression-04.txt.
10180 * Add the ASN.1 structures and functions for CertificatePair, which
10181 is defined as follows (according to X.509_4thEditionDraftV6.pdf):
10183 CertificatePair ::= SEQUENCE {
10184 forward [0] Certificate OPTIONAL,
10185 reverse [1] Certificate OPTIONAL,
10186 -- at least one of the pair shall be present -- }
10188 Also implement the PEM functions to read and write certificate
10189 pairs, and defined the PEM tag as "CERTIFICATE PAIR".
10191 This needed to be defined, mostly for the sake of the LDAP
10192 attribute crossCertificatePair, but may prove useful elsewhere as
10197 * Make it possible to inhibit symlinking of shared libraries in
10198 Makefile.shared, for Cygwin's sake.
10202 * Extend the BIGNUM API by creating a function
10203 void BN_set_negative(BIGNUM *a, int neg);
10204 and a macro that behave like
10205 int BN_is_negative(const BIGNUM *a);
10207 to avoid the need to access 'a->neg' directly in applications.
10211 * Implement fast modular reduction for pseudo-Mersenne primes
10212 used in NIST curves (crypto/bn/bn_nist.c, crypto/ec/ecp_nist.c).
10213 EC_GROUP_new_curve_GFp() will now automatically use this
10216 *Nils Larsch <nla@trustcenter.de>*
10218 * Add new lock type (CRYPTO_LOCK_BN).
10222 * Change the ENGINE framework to automatically load engines
10223 dynamically from specific directories unless they could be
10224 found to already be built in or loaded. Move all the
10225 current engines except for the cryptodev one to a new
10226 directory engines/.
10227 The engines in engines/ are built as shared libraries if
10228 the "shared" options was given to ./Configure or ./config.
10229 Otherwise, they are inserted in libcrypto.a.
10230 /usr/local/ssl/engines is the default directory for dynamic
10231 engines, but that can be overridden at configure time through
10232 the usual use of --prefix and/or --openssldir, and at run
10233 time with the environment variable OPENSSL_ENGINES.
10235 *Geoff Thorpe and Richard Levitte*
10237 * Add Makefile.shared, a helper makefile to build shared
10238 libraries. Adapt Makefile.org.
10242 * Add version info to Win32 DLLs.
10244 *Peter 'Luna' Runestig" <peter@runestig.com>*
10246 * Add new 'medium level' PKCS#12 API. Certificates and keys
10247 can be added using this API to created arbitrary PKCS#12
10248 files while avoiding the low level API.
10250 New options to PKCS12_create(), key or cert can be NULL and
10251 will then be omitted from the output file. The encryption
10252 algorithm NIDs can be set to -1 for no encryption, the mac
10253 iteration count can be set to 0 to omit the mac.
10255 Enhance pkcs12 utility by making the -nokeys and -nocerts
10256 options work when creating a PKCS#12 file. New option -nomac
10257 to omit the mac, NONE can be set for an encryption algorithm.
10258 New code is modified to use the enhanced PKCS12_create()
10259 instead of the low level API.
10263 * Extend ASN1 encoder to support indefinite length constructed
10264 encoding. This can output sequences tags and octet strings in
10265 this form. Modify pk7_asn1.c to support indefinite length
10266 encoding. This is experimental and needs additional code to
10267 be useful, such as an ASN1 bio and some enhanced streaming
10270 Extend template encode functionality so that tagging is passed
10271 down to the template encoder.
10275 * Let 'openssl req' fail if an argument to '-newkey' is not
10276 recognized instead of using RSA as a default.
10280 * Add support for ECC-based ciphersuites from draft-ietf-tls-ecc-01.txt.
10281 As these are not official, they are not included in "ALL";
10282 the "ECCdraft" ciphersuite group alias can be used to select them.
10284 *Vipul Gupta and Sumit Gupta (Sun Microsystems Laboratories)*
10286 * Add ECDH engine support.
10288 *Nils Gura and Douglas Stebila (Sun Microsystems Laboratories)*
10290 * Add ECDH in new directory crypto/ecdh/.
10292 *Douglas Stebila (Sun Microsystems Laboratories)*
10294 * Let BN_rand_range() abort with an error after 100 iterations
10295 without success (which indicates a broken PRNG).
10299 * Change BN_mod_sqrt() so that it verifies that the input value
10300 is really the square of the return value. (Previously,
10301 BN_mod_sqrt would show GIGO behaviour.)
10305 * Add named elliptic curves over binary fields from X9.62, SECG,
10306 and WAP/WTLS; add OIDs that were still missing.
10308 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10310 * Extend the EC library for elliptic curves over binary fields
10311 (new files ec2_smpl.c, ec2_smpt.c, ec2_mult.c in crypto/ec/).
10314 EC_GF2m_simple_method
10318 EC_GROUP_new_curve_GF2m
10319 EC_GROUP_set_curve_GF2m
10320 EC_GROUP_get_curve_GF2m
10321 EC_POINT_set_affine_coordinates_GF2m
10322 EC_POINT_get_affine_coordinates_GF2m
10323 EC_POINT_set_compressed_coordinates_GF2m
10325 Point compression for binary fields is disabled by default for
10326 patent reasons (compile with OPENSSL_EC_BIN_PT_COMP defined to
10329 As binary polynomials are represented as BIGNUMs, various members
10330 of the EC_GROUP and EC_POINT data structures can be shared
10331 between the implementations for prime fields and binary fields;
10332 the above `..._GF2m functions` (except for EX_GROUP_new_curve_GF2m)
10333 are essentially identical to their `..._GFp` counterparts.
10334 (For simplicity, the `..._GFp` prefix has been dropped from
10335 various internal method names.)
10337 An internal 'field_div' method (similar to 'field_mul' and
10338 'field_sqr') has been added; this is used only for binary fields.
10340 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10342 * Optionally dispatch EC_POINT_mul(), EC_POINT_precompute_mult()
10343 through methods ('mul', 'precompute_mult').
10345 The generic implementations (now internally called 'ec_wNAF_mul'
10346 and 'ec_wNAF_precomputed_mult') remain the default if these
10347 methods are undefined.
10349 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10351 * New function EC_GROUP_get_degree, which is defined through
10352 EC_METHOD. For curves over prime fields, this returns the bit
10353 length of the modulus.
10355 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10357 * New functions EC_GROUP_dup, EC_POINT_dup.
10358 (These simply call ..._new and ..._copy).
10360 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10362 * Add binary polynomial arithmetic software in crypto/bn/bn_gf2m.c.
10363 Polynomials are represented as BIGNUMs (where the sign bit is not
10364 used) in the following functions [macros]:
10367 BN_GF2m_sub [= BN_GF2m_add]
10368 BN_GF2m_mod [wrapper for BN_GF2m_mod_arr]
10369 BN_GF2m_mod_mul [wrapper for BN_GF2m_mod_mul_arr]
10370 BN_GF2m_mod_sqr [wrapper for BN_GF2m_mod_sqr_arr]
10372 BN_GF2m_mod_exp [wrapper for BN_GF2m_mod_exp_arr]
10373 BN_GF2m_mod_sqrt [wrapper for BN_GF2m_mod_sqrt_arr]
10374 BN_GF2m_mod_solve_quad [wrapper for BN_GF2m_mod_solve_quad_arr]
10375 BN_GF2m_cmp [= BN_ucmp]
10377 (Note that only the 'mod' functions are actually for fields GF(2^m).
10378 BN_GF2m_add() is misnomer, but this is for the sake of consistency.)
10380 For some functions, an the irreducible polynomial defining a
10381 field can be given as an 'unsigned int[]' with strictly
10382 decreasing elements giving the indices of those bits that are set;
10383 i.e., p[] represents the polynomial
10384 f(t) = t^p[0] + t^p[1] + ... + t^p[k]
10386 p[0] > p[1] > ... > p[k] = 0.
10387 This applies to the following functions:
10390 BN_GF2m_mod_mul_arr
10391 BN_GF2m_mod_sqr_arr
10392 BN_GF2m_mod_inv_arr [wrapper for BN_GF2m_mod_inv]
10393 BN_GF2m_mod_div_arr [wrapper for BN_GF2m_mod_div]
10394 BN_GF2m_mod_exp_arr
10395 BN_GF2m_mod_sqrt_arr
10396 BN_GF2m_mod_solve_quad_arr
10400 Conversion can be performed by the following functions:
10405 bntest.c has additional tests for binary polynomial arithmetic.
10407 Two implementations for BN_GF2m_mod_div() are available.
10408 The default algorithm simply uses BN_GF2m_mod_inv() and
10409 BN_GF2m_mod_mul(). The alternative algorithm is compiled in only
10410 if OPENSSL_SUN_GF2M_DIV is defined (patent pending; read the
10411 copyright notice in crypto/bn/bn_gf2m.c before enabling it).
10413 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10415 * Add new error code 'ERR_R_DISABLED' that can be used when some
10416 functionality is disabled at compile-time.
10418 *Douglas Stebila <douglas.stebila@sun.com>*
10420 * Change default behaviour of 'openssl asn1parse' so that more
10421 information is visible when viewing, e.g., a certificate:
10423 Modify asn1_parse2 (crypto/asn1/asn1_par.c) so that in non-'dump'
10424 mode the content of non-printable OCTET STRINGs is output in a
10425 style similar to INTEGERs, but with '[HEX DUMP]' prepended to
10426 avoid the appearance of a printable string.
10428 *Nils Larsch <nla@trustcenter.de>*
10430 * Add 'asn1_flag' and 'asn1_form' member to EC_GROUP with access
10432 EC_GROUP_set_asn1_flag()
10433 EC_GROUP_get_asn1_flag()
10434 EC_GROUP_set_point_conversion_form()
10435 EC_GROUP_get_point_conversion_form()
10436 These control ASN1 encoding details:
10437 - Curves (i.e., groups) are encoded explicitly unless asn1_flag
10438 has been set to OPENSSL_EC_NAMED_CURVE.
10439 - Points are encoded in uncompressed form by default; options for
10440 asn1_for are as for point2oct, namely
10441 POINT_CONVERSION_COMPRESSED
10442 POINT_CONVERSION_UNCOMPRESSED
10443 POINT_CONVERSION_HYBRID
10445 Also add 'seed' and 'seed_len' members to EC_GROUP with access
10447 EC_GROUP_set_seed()
10448 EC_GROUP_get0_seed()
10449 EC_GROUP_get_seed_len()
10450 This is used only for ASN1 purposes (so far).
10452 *Nils Larsch <nla@trustcenter.de>*
10454 * Add 'field_type' member to EC_METHOD, which holds the NID
10455 of the appropriate field type OID. The new function
10456 EC_METHOD_get_field_type() returns this value.
10458 *Nils Larsch <nla@trustcenter.de>*
10461 EC_POINT_point2bn()
10462 EC_POINT_bn2point()
10463 EC_POINT_point2hex()
10464 EC_POINT_hex2point()
10465 providing useful interfaces to EC_POINT_point2oct() and
10466 EC_POINT_oct2point().
10468 *Nils Larsch <nla@trustcenter.de>*
10470 * Change internals of the EC library so that the functions
10471 EC_GROUP_set_generator()
10472 EC_GROUP_get_generator()
10473 EC_GROUP_get_order()
10474 EC_GROUP_get_cofactor()
10475 are implemented directly in crypto/ec/ec_lib.c and not dispatched
10476 to methods, which would lead to unnecessary code duplication when
10477 adding different types of curves.
10479 *Nils Larsch <nla@trustcenter.de> with input by Bodo Moeller*
10481 * Implement compute_wNAF (crypto/ec/ec_mult.c) without BIGNUM
10482 arithmetic, and such that modified wNAFs are generated
10483 (which avoid length expansion in many cases).
10487 * Add a function EC_GROUP_check_discriminant() (defined via
10488 EC_METHOD) that verifies that the curve discriminant is non-zero.
10490 Add a function EC_GROUP_check() that makes some sanity tests
10491 on a EC_GROUP, its generator and order. This includes
10492 EC_GROUP_check_discriminant().
10494 *Nils Larsch <nla@trustcenter.de>*
10496 * Add ECDSA in new directory crypto/ecdsa/.
10498 Add applications 'openssl ecparam' and 'openssl ecdsa'
10499 (these are based on 'openssl dsaparam' and 'openssl dsa').
10501 ECDSA support is also included in various other files across the
10502 library. Most notably,
10503 - 'openssl req' now has a '-newkey ecdsa:file' option;
10504 - EVP_PKCS82PKEY (crypto/evp/evp_pkey.c) now can handle ECDSA;
10505 - X509_PUBKEY_get (crypto/asn1/x_pubkey.c) and
10506 d2i_PublicKey (crypto/asn1/d2i_pu.c) have been modified to make
10507 them suitable for ECDSA where domain parameters must be
10508 extracted before the specific public key;
10509 - ECDSA engine support has been added.
10511 *Nils Larsch <nla@trustcenter.de>*
10513 * Include some named elliptic curves, and add OIDs from X9.62,
10514 SECG, and WAP/WTLS. Each curve can be obtained from the new
10516 EC_GROUP_new_by_curve_name(),
10517 and the list of available named curves can be obtained with
10518 EC_get_builtin_curves().
10519 Also add a 'curve_name' member to EC_GROUP objects, which can be
10521 EC_GROUP_set_curve_name()
10522 EC_GROUP_get_curve_name()
10524 *Nils Larsch <larsch@trustcenter.de, Bodo Moeller*
10526 * Remove a few calls to bn_wexpand() in BN_sqr() (the one in there
10527 was actually never needed) and in BN_mul(). The removal in BN_mul()
10528 required a small change in bn_mul_part_recursive() and the addition
10529 of the functions bn_cmp_part_words(), bn_sub_part_words() and
10530 bn_add_part_words(), which do the same thing as bn_cmp_words(),
10531 bn_sub_words() and bn_add_words() except they take arrays with
10536 ### Changes between 0.9.7l and 0.9.7m [23 Feb 2007]
10538 * Cleanse PEM buffers before freeing them since they may contain
10541 *Benjamin Bennett <ben@psc.edu>*
10543 * Include "!eNULL" in SSL_DEFAULT_CIPHER_LIST to make sure that
10544 a ciphersuite string such as "DEFAULT:RSA" cannot enable
10545 authentication-only ciphersuites.
10549 * Since AES128 and AES256 share a single mask bit in the logic of
10550 ssl/ssl_ciph.c, the code for masking out disabled ciphers needs a
10551 kludge to work properly if AES128 is available and AES256 isn't.
10555 * Expand security boundary to match 1.1.1 module.
10559 * Remove redundant features: hash file source, editing of test vectors
10560 modify fipsld to use external fips_premain.c signature.
10564 * New perl script mkfipsscr.pl to create shell scripts or batch files to
10565 run algorithm test programs.
10569 * Make algorithm test programs more tolerant of whitespace.
10573 * Have SSL/TLS server implementation tolerate "mismatched" record
10574 protocol version while receiving ClientHello even if the
10575 ClientHello is fragmented. (The server can't insist on the
10576 particular protocol version it has chosen before the ServerHello
10577 message has informed the client about his choice.)
10581 * Load error codes if they are not already present instead of using a
10582 static variable. This allows them to be cleanly unloaded and reloaded.
10586 ### Changes between 0.9.7k and 0.9.7l [28 Sep 2006]
10588 * Introduce limits to prevent malicious keys being able to
10589 cause a denial of service. [CVE-2006-2940][]
10591 *Steve Henson, Bodo Moeller*
10593 * Fix ASN.1 parsing of certain invalid structures that can result
10594 in a denial of service. [CVE-2006-2937][] [Steve Henson]
10596 * Fix buffer overflow in SSL_get_shared_ciphers() function.
10597 [CVE-2006-3738][] [Tavis Ormandy and Will Drewry, Google Security Team]
10599 * Fix SSL client code which could crash if connecting to a
10600 malicious SSLv2 server. [CVE-2006-4343][]
10602 *Tavis Ormandy and Will Drewry, Google Security Team*
10604 * Change ciphersuite string processing so that an explicit
10605 ciphersuite selects this one ciphersuite (so that "AES256-SHA"
10606 will no longer include "AES128-SHA"), and any other similar
10607 ciphersuite (same bitmap) from *other* protocol versions (so that
10608 "RC4-MD5" will still include both the SSL 2.0 ciphersuite and the
10609 SSL 3.0/TLS 1.0 ciphersuite). This is a backport combining
10610 changes from 0.9.8b and 0.9.8d.
10614 ### Changes between 0.9.7j and 0.9.7k [05 Sep 2006]
10616 * Avoid PKCS #1 v1.5 signature attack discovered by Daniel Bleichenbacher
10617 [CVE-2006-4339][] [Ben Laurie and Google Security Team]
10619 * Change the Unix randomness entropy gathering to use poll() when
10620 possible instead of select(), since the latter has some
10621 undesirable limitations.
10623 *Darryl Miles via Richard Levitte and Bodo Moeller*
10625 * Disable rogue ciphersuites:
10627 - SSLv2 0x08 0x00 0x80 ("RC4-64-MD5")
10628 - SSLv3/TLSv1 0x00 0x61 ("EXP1024-RC2-CBC-MD5")
10629 - SSLv3/TLSv1 0x00 0x60 ("EXP1024-RC4-MD5")
10631 The latter two were purportedly from
10632 draft-ietf-tls-56-bit-ciphersuites-0[01].txt, but do not really
10635 Also deactivate the remaining ciphersuites from
10636 draft-ietf-tls-56-bit-ciphersuites-01.txt. These are just as
10637 unofficial, and the ID has long expired.
10641 * Fix RSA blinding Heisenbug (problems sometimes occurred on
10642 dual-core machines) and other potential thread-safety issues.
10646 ### Changes between 0.9.7i and 0.9.7j [04 May 2006]
10648 * Adapt fipsld and the build system to link against the validated FIPS
10649 module in FIPS mode.
10653 * Fixes for VC++ 2005 build under Windows.
10657 * Add new Windows build target VC-32-GMAKE for VC++. This uses GNU make
10658 from a Windows bash shell such as MSYS. It is autodetected from the
10659 "config" script when run from a VC++ environment. Modify standard VC++
10660 build to use fipscanister.o from the GNU make build.
10664 ### Changes between 0.9.7h and 0.9.7i [14 Oct 2005]
10666 * Wrapped the definition of EVP_MAX_MD_SIZE in a #ifdef OPENSSL_FIPS.
10667 The value now differs depending on if you build for FIPS or not.
10668 BEWARE! A program linked with a shared FIPSed libcrypto can't be
10669 safely run with a non-FIPSed libcrypto, as it may crash because of
10670 the difference induced by this change.
10674 ### Changes between 0.9.7g and 0.9.7h [11 Oct 2005]
10676 * Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING
10677 (part of SSL_OP_ALL). This option used to disable the
10678 countermeasure against man-in-the-middle protocol-version
10679 rollback in the SSL 2.0 server implementation, which is a bad
10680 idea. [CVE-2005-2969][]
10682 *Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center
10683 for Information Security, National Institute of Advanced Industrial
10684 Science and Technology [AIST, Japan)]*
10686 * Minimal support for X9.31 signatures and PSS padding modes. This is
10687 mainly for FIPS compliance and not fully integrated at this stage.
10691 * For DSA signing, unless DSA_FLAG_NO_EXP_CONSTTIME is set, perform
10692 the exponentiation using a fixed-length exponent. (Otherwise,
10693 the information leaked through timing could expose the secret key
10694 after many signatures; cf. Bleichenbacher's attack on DSA with
10699 * Make a new fixed-window mod_exp implementation the default for
10700 RSA, DSA, and DH private-key operations so that the sequence of
10701 squares and multiplies and the memory access pattern are
10702 independent of the particular secret key. This will mitigate
10703 cache-timing and potential related attacks.
10705 BN_mod_exp_mont_consttime() is the new exponentiation implementation,
10706 and this is automatically used by BN_mod_exp_mont() if the new flag
10707 BN_FLG_EXP_CONSTTIME is set for the exponent. RSA, DSA, and DH
10708 will use this BN flag for private exponents unless the flag
10709 RSA_FLAG_NO_EXP_CONSTTIME, DSA_FLAG_NO_EXP_CONSTTIME, or
10710 DH_FLAG_NO_EXP_CONSTTIME, respectively, is set.
10712 *Matthew D Wood (Intel Corp), with some changes by Bodo Moeller*
10714 * Change the client implementation for SSLv23_method() and
10715 SSLv23_client_method() so that is uses the SSL 3.0/TLS 1.0
10716 Client Hello message format if the SSL_OP_NO_SSLv2 option is set.
10717 (Previously, the SSL 2.0 backwards compatible Client Hello
10718 message format would be used even with SSL_OP_NO_SSLv2.)
10722 * Add support for smime-type MIME parameter in S/MIME messages which some
10727 * New function BN_MONT_CTX_set_locked() to set montgomery parameters in
10728 a threadsafe manner. Modify rsa code to use new function and add calls
10729 to dsa and dh code (which had race conditions before).
10733 * Include the fixed error library code in the C error file definitions
10734 instead of fixing them up at runtime. This keeps the error code
10735 structures constant.
10739 ### Changes between 0.9.7f and 0.9.7g [11 Apr 2005]
10741 [NB: OpenSSL 0.9.7h and later 0.9.7 patch levels were released after
10744 * Fixes for newer kerberos headers. NB: the casts are needed because
10745 the 'length' field is signed on one version and unsigned on another
10746 with no (?) obvious way to tell the difference, without these VC++
10747 complains. Also the "definition" of FAR (blank) is no longer included
10748 nor is the error ENOMEM. KRB5_PRIVATE has to be set to 1 to pick up
10749 some needed definitions.
10753 * Undo Cygwin change.
10757 * Added support for proxy certificates according to RFC 3820.
10758 Because they may be a security thread to unaware applications,
10759 they must be explicitly allowed in run-time. See
10760 docs/HOWTO/proxy_certificates.txt for further information.
10764 ### Changes between 0.9.7e and 0.9.7f [22 Mar 2005]
10766 * Use (SSL_RANDOM_VALUE - 4) bytes of pseudo random data when generating
10767 server and client random values. Previously
10768 (SSL_RANDOM_VALUE - sizeof(time_t)) would be used which would result in
10769 less random data when sizeof(time_t) > 4 (some 64 bit platforms).
10771 This change has negligible security impact because:
10773 1. Server and client random values still have 24 bytes of pseudo random
10776 2. Server and client random values are sent in the clear in the initial
10779 3. The master secret is derived using the premaster secret (48 bytes in
10780 size for static RSA ciphersuites) as well as client server and random
10783 The OpenSSL team would like to thank the UK NISCC for bringing this issue
10786 *Stephen Henson, reported by UK NISCC*
10788 * Use Windows randomness collection on Cygwin.
10792 * Fix hang in EGD/PRNGD query when communication socket is closed
10793 prematurely by EGD/PRNGD.
10795 *Darren Tucker <dtucker@zip.com.au> via Lutz Jänicke, resolves #1014*
10797 * Prompt for pass phrases when appropriate for PKCS12 input format.
10801 * Back-port of selected performance improvements from development
10802 branch, as well as improved support for PowerPC platforms.
10806 * Add lots of checks for memory allocation failure, error codes to indicate
10807 failure and freeing up memory if a failure occurs.
10809 *Nauticus Networks SSL Team <openssl@nauticusnet.com>, Steve Henson*
10811 * Add new -passin argument to dgst.
10815 * Perform some character comparisons of different types in X509_NAME_cmp:
10816 this is needed for some certificates that re-encode DNs into UTF8Strings
10817 (in violation of RFC3280) and can't or won't issue name rollover
10822 * Make an explicit check during certificate validation to see that
10823 the CA setting in each certificate on the chain is correct. As a
10824 side effect always do the following basic checks on extensions,
10825 not just when there's an associated purpose to the check:
10827 - if there is an unhandled critical extension (unless the user
10828 has chosen to ignore this fault)
10829 - if the path length has been exceeded (if one is set at all)
10830 - that certain extensions fit the associated purpose (if one has
10835 ### Changes between 0.9.7d and 0.9.7e [25 Oct 2004]
10837 * Avoid a race condition when CRLs are checked in a multi threaded
10838 environment. This would happen due to the reordering of the revoked
10839 entries during signature checking and serial number lookup. Now the
10840 encoding is cached and the serial number sort performed under a lock.
10841 Add new STACK function sk_is_sorted().
10845 * Add Delta CRL to the extension code.
10849 * Various fixes to s3_pkt.c so alerts are sent properly.
10851 *David Holmes <d.holmes@f5.com>*
10853 * Reduce the chances of duplicate issuer name and serial numbers (in
10854 violation of RFC3280) using the OpenSSL certificate creation utilities.
10855 This is done by creating a random 64 bit value for the initial serial
10856 number when a serial number file is created or when a self signed
10857 certificate is created using 'openssl req -x509'. The initial serial
10858 number file is created using 'openssl x509 -next_serial' in CA.pl
10859 rather than being initialized to 1.
10863 ### Changes between 0.9.7c and 0.9.7d [17 Mar 2004]
10865 * Fix null-pointer assignment in do_change_cipher_spec() revealed
10866 by using the Codenomicon TLS Test Tool [CVE-2004-0079][]
10868 *Joe Orton, Steve Henson*
10870 * Fix flaw in SSL/TLS handshaking when using Kerberos ciphersuites
10873 *Joe Orton, Steve Henson*
10875 * Make it possible to have multiple active certificates with the same
10876 subject in the CA index file. This is done only if the keyword
10877 'unique_subject' is set to 'no' in the main CA section (default
10878 if 'CA_default') of the configuration file. The value is saved
10879 with the database itself in a separate index attribute file,
10880 named like the index file with '.attr' appended to the name.
10884 * X509 verify fixes. Disable broken certificate workarounds when
10885 X509_V_FLAGS_X509_STRICT is set. Check CRL issuer has cRLSign set if
10886 keyUsage extension present. Don't accept CRLs with unhandled critical
10887 extensions: since verify currently doesn't process CRL extensions this
10888 rejects a CRL with *any* critical extensions. Add new verify error codes
10893 * When creating an OCSP nonce use an OCTET STRING inside the extnValue.
10894 A clarification of RFC2560 will require the use of OCTET STRINGs and
10895 some implementations cannot handle the current raw format. Since OpenSSL
10896 copies and compares OCSP nonces as opaque blobs without any attempt at
10897 parsing them this should not create any compatibility issues.
10901 * New md flag EVP_MD_CTX_FLAG_REUSE this allows md_data to be reused when
10902 calling EVP_MD_CTX_copy_ex() to avoid calling OPENSSL_malloc(). Without
10903 this HMAC (and other) operations are several times slower than OpenSSL
10908 * Print out GeneralizedTime and UTCTime in ASN1_STRING_print_ex().
10910 *Peter Sylvester <Peter.Sylvester@EdelWeb.fr>*
10912 * Use the correct content when signing type "other".
10916 ### Changes between 0.9.7b and 0.9.7c [30 Sep 2003]
10918 * Fix various bugs revealed by running the NISCC test suite:
10920 Stop out of bounds reads in the ASN1 code when presented with
10921 invalid tags (CVE-2003-0543 and CVE-2003-0544).
10923 Free up ASN1_TYPE correctly if ANY type is invalid [CVE-2003-0545][].
10925 If verify callback ignores invalid public key errors don't try to check
10926 certificate signature with the NULL public key.
10930 * New -ignore_err option in ocsp application to stop the server
10931 exiting on the first error in a request.
10935 * In ssl3_accept() (ssl/s3_srvr.c) only accept a client certificate
10936 if the server requested one: as stated in TLS 1.0 and SSL 3.0
10941 * In ssl3_get_client_hello() (ssl/s3_srvr.c), tolerate additional
10942 extra data after the compression methods not only for TLS 1.0
10943 but also for SSL 3.0 (as required by the specification).
10945 *Bodo Moeller; problem pointed out by Matthias Loepfe*
10947 * Change X509_certificate_type() to mark the key as exported/exportable
10948 when it's 512 *bits* long, not 512 bytes.
10952 * Change AES_cbc_encrypt() so it outputs exact multiple of
10953 blocks during encryption.
10957 * Various fixes to base64 BIO and non blocking I/O. On write
10958 flushes were not handled properly if the BIO retried. On read
10959 data was not being buffered properly and had various logic bugs.
10960 This also affects blocking I/O when the data being decoded is a
10965 * Various S/MIME bugfixes and compatibility changes:
10966 output correct application/pkcs7 MIME type if
10967 PKCS7_NOOLDMIMETYPE is set. Tolerate some broken signatures.
10968 Output CR+LF for EOL if PKCS7_CRLFEOL is set (this makes opening
10969 of files as .eml work). Correctly handle very long lines in MIME
10974 ### Changes between 0.9.7a and 0.9.7b [10 Apr 2003]
10976 * Countermeasure against the Klima-Pokorny-Rosa extension of
10977 Bleichbacher's attack on PKCS #1 v1.5 padding: treat
10978 a protocol version number mismatch like a decryption error
10979 in ssl3_get_client_key_exchange (ssl/s3_srvr.c).
10983 * Turn on RSA blinding by default in the default implementation
10984 to avoid a timing attack. Applications that don't want it can call
10985 RSA_blinding_off() or use the new flag RSA_FLAG_NO_BLINDING.
10986 They would be ill-advised to do so in most cases.
10988 *Ben Laurie, Steve Henson, Geoff Thorpe, Bodo Moeller*
10990 * Change RSA blinding code so that it works when the PRNG is not
10991 seeded (in this case, the secret RSA exponent is abused as
10992 an unpredictable seed -- if it is not unpredictable, there
10993 is no point in blinding anyway). Make RSA blinding thread-safe
10994 by remembering the creator's thread ID in rsa->blinding and
10995 having all other threads use local one-time blinding factors
10996 (this requires more computation than sharing rsa->blinding, but
10997 avoids excessive locking; and if an RSA object is not shared
10998 between threads, blinding will still be very fast).
11002 * Fixed a typo bug that would cause ENGINE_set_default() to set an
11003 ENGINE as defaults for all supported algorithms irrespective of
11004 the 'flags' parameter. 'flags' is now honoured, so applications
11005 should make sure they are passing it correctly.
11009 * Target "mingw" now allows native Windows code to be generated in
11010 the Cygwin environment as well as with the MinGW compiler.
11014 ### Changes between 0.9.7 and 0.9.7a [19 Feb 2003]
11016 * In ssl3_get_record (ssl/s3_pkt.c), minimize information leaked
11017 via timing by performing a MAC computation even if incorrect
11018 block cipher padding has been found. This is a countermeasure
11019 against active attacks where the attacker has to distinguish
11020 between bad padding and a MAC verification error. [CVE-2003-0078][]
11022 *Bodo Moeller; problem pointed out by Brice Canvel (EPFL),
11023 Alain Hiltgen (UBS), Serge Vaudenay (EPFL), and
11024 Martin Vuagnoux (EPFL, Ilion)*
11026 * Make the no-err option work as intended. The intention with no-err
11027 is not to have the whole error stack handling routines removed from
11028 libcrypto, it's only intended to remove all the function name and
11029 reason texts, thereby removing some of the footprint that may not
11030 be interesting if those errors aren't displayed anyway.
11032 NOTE: it's still possible for any application or module to have its
11033 own set of error texts inserted. The routines are there, just not
11034 used by default when no-err is given.
11038 * Add support for FreeBSD on IA64.
11040 *dirk.meyer@dinoex.sub.org via Richard Levitte, resolves #454*
11042 * Adjust DES_cbc_cksum() so it returns the same value as the MIT
11043 Kerberos function mit_des_cbc_cksum(). Before this change,
11044 the value returned by DES_cbc_cksum() was like the one from
11045 mit_des_cbc_cksum(), except the bytes were swapped.
11047 *Kevin Greaney <Kevin.Greaney@hp.com> and Richard Levitte*
11049 * Allow an application to disable the automatic SSL chain building.
11050 Before this a rather primitive chain build was always performed in
11051 ssl3_output_cert_chain(): an application had no way to send the
11052 correct chain if the automatic operation produced an incorrect result.
11054 Now the chain builder is disabled if either:
11056 1. Extra certificates are added via SSL_CTX_add_extra_chain_cert().
11058 2. The mode flag SSL_MODE_NO_AUTO_CHAIN is set.
11060 The reasoning behind this is that an application would not want the
11061 auto chain building to take place if extra chain certificates are
11062 present and it might also want a means of sending no additional
11063 certificates (for example the chain has two certificates and the
11068 * Add the possibility to build without the ENGINE framework.
11070 *Steven Reddie <smr@essemer.com.au> via Richard Levitte*
11072 * Under Win32 gmtime() can return NULL: check return value in
11073 OPENSSL_gmtime(). Add error code for case where gmtime() fails.
11077 * DSA routines: under certain error conditions uninitialized BN objects
11078 could be freed. Solution: make sure initialization is performed early
11079 enough. (Reported and fix supplied by Ivan D Nestlerode <nestler@MIT.EDU>,
11080 Nils Larsch <nla@trustcenter.de> via PR#459)
11084 * Another fix for SSLv2 session ID handling: the session ID was incorrectly
11085 checked on reconnect on the client side, therefore session resumption
11086 could still fail with a "ssl session id is different" error. This
11087 behaviour is masked when SSL_OP_ALL is used due to
11088 SSL_OP_MICROSOFT_SESS_ID_BUG being set.
11089 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as
11090 followup to PR #377.
11094 * IA-32 assembler support enhancements: unified ELF targets, support
11095 for SCO/Caldera platforms, fix for Cygwin shared build.
11099 * Add support for FreeBSD on sparc64. As a consequence, support for
11100 FreeBSD on non-x86 processors is separate from x86 processors on
11101 the config script, much like the NetBSD support.
11103 *Richard Levitte & Kris Kennaway <kris@obsecurity.org>*
11105 ### Changes between 0.9.6h and 0.9.7 [31 Dec 2002]
11107 [NB: OpenSSL 0.9.6i and later 0.9.6 patch levels were released after
11110 * Fix session ID handling in SSLv2 client code: the SERVER FINISHED
11111 code (06) was taken as the first octet of the session ID and the last
11112 octet was ignored consequently. As a result SSLv2 client side session
11113 caching could not have worked due to the session ID mismatch between
11115 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as
11120 * Change the declaration of needed Kerberos libraries to use EX_LIBS
11121 instead of the special (and badly supported) LIBKRB5. LIBKRB5 is
11126 * The hw_ncipher.c engine requires dynamic locks. Unfortunately, it
11127 seems that in spite of existing for more than a year, many application
11128 author have done nothing to provide the necessary callbacks, which
11129 means that this particular engine will not work properly anywhere.
11130 This is a very unfortunate situation which forces us, in the name
11131 of usability, to give the hw_ncipher.c a static lock, which is part
11133 NOTE: This is for the 0.9.7 series ONLY. This hack will never
11134 appear in 0.9.8 or later. We EXPECT application authors to have
11135 dealt properly with this when 0.9.8 is released (unless we actually
11136 make such changes in the libcrypto locking code that changes will
11137 have to be made anyway).
11141 * In asn1_d2i_read_bio() repeatedly call BIO_read() until all content
11142 octets have been read, EOF or an error occurs. Without this change
11143 some truncated ASN1 structures will not produce an error.
11147 * Disable Heimdal support, since it hasn't been fully implemented.
11148 Still give the possibility to force the use of Heimdal, but with
11149 warnings and a request that patches get sent to openssl-dev.
11153 * Add the VC-CE target, introduce the WINCE sysname, and add
11154 INSTALL.WCE and appropriate conditionals to make it build.
11156 *Steven Reddie <smr@essemer.com.au> via Richard Levitte*
11158 * Change the DLL names for Cygwin to cygcrypto-x.y.z.dll and
11159 cygssl-x.y.z.dll, where x, y and z are the major, minor and
11160 edit numbers of the version.
11162 *Corinna Vinschen <vinschen@redhat.com> and Richard Levitte*
11164 * Introduce safe string copy and catenation functions
11165 (BUF_strlcpy() and BUF_strlcat()).
11167 *Ben Laurie (CHATS) and Richard Levitte*
11169 * Avoid using fixed-size buffers for one-line DNs.
11171 *Ben Laurie (CHATS)*
11173 * Add BUF_MEM_grow_clean() to avoid information leakage when
11174 resizing buffers containing secrets, and use where appropriate.
11176 *Ben Laurie (CHATS)*
11178 * Avoid using fixed size buffers for configuration file location.
11180 *Ben Laurie (CHATS)*
11182 * Avoid filename truncation for various CA files.
11184 *Ben Laurie (CHATS)*
11186 * Use sizeof in preference to magic numbers.
11188 *Ben Laurie (CHATS)*
11190 * Avoid filename truncation in cert requests.
11192 *Ben Laurie (CHATS)*
11194 * Add assertions to check for (supposedly impossible) buffer
11197 *Ben Laurie (CHATS)*
11199 * Don't cache truncated DNS entries in the local cache (this could
11200 potentially lead to a spoofing attack).
11202 *Ben Laurie (CHATS)*
11204 * Fix various buffers to be large enough for hex/decimal
11205 representations in a platform independent manner.
11207 *Ben Laurie (CHATS)*
11209 * Add CRYPTO_realloc_clean() to avoid information leakage when
11210 resizing buffers containing secrets, and use where appropriate.
11212 *Ben Laurie (CHATS)*
11214 * Add BIO_indent() to avoid much slightly worrying code to do
11217 *Ben Laurie (CHATS)*
11219 * Convert sprintf()/BIO_puts() to BIO_printf().
11221 *Ben Laurie (CHATS)*
11223 * buffer_gets() could terminate with the buffer only half
11226 *Ben Laurie (CHATS)*
11228 * Add assertions to prevent user-supplied crypto functions from
11229 overflowing internal buffers by having large block sizes, etc.
11231 *Ben Laurie (CHATS)*
11233 * New OPENSSL_assert() macro (similar to assert(), but enabled
11236 *Ben Laurie (CHATS)*
11238 * Eliminate unused copy of key in RC4.
11240 *Ben Laurie (CHATS)*
11242 * Eliminate unused and incorrectly sized buffers for IV in pem.h.
11244 *Ben Laurie (CHATS)*
11246 * Fix off-by-one error in EGD path.
11248 *Ben Laurie (CHATS)*
11250 * If RANDFILE path is too long, ignore instead of truncating.
11252 *Ben Laurie (CHATS)*
11254 * Eliminate unused and incorrectly sized X.509 structure
11257 *Ben Laurie (CHATS)*
11259 * Eliminate unused and dangerous function knumber().
11261 *Ben Laurie (CHATS)*
11263 * Eliminate unused and dangerous structure, KSSL_ERR.
11265 *Ben Laurie (CHATS)*
11267 * Protect against overlong session ID context length in an encoded
11268 session object. Since these are local, this does not appear to be
11271 *Ben Laurie (CHATS)*
11273 * Change from security patch (see 0.9.6e below) that did not affect
11274 the 0.9.6 release series:
11276 Remote buffer overflow in SSL3 protocol - an attacker could
11277 supply an oversized master key in Kerberos-enabled versions.
11280 *Ben Laurie (CHATS)*
11282 * Change the SSL kerb5 codes to match RFC 2712.
11286 * Make -nameopt work fully for req and add -reqopt switch.
11288 *Michael Bell <michael.bell@rz.hu-berlin.de>, Steve Henson*
11290 * The "block size" for block ciphers in CFB and OFB mode should be 1.
11292 *Steve Henson, reported by Yngve Nysaeter Pettersen <yngve@opera.com>*
11294 * Make sure tests can be performed even if the corresponding algorithms
11295 have been removed entirely. This was also the last step to make
11296 OpenSSL compilable with DJGPP under all reasonable conditions.
11298 *Richard Levitte, Doug Kaufman <dkaufman@rahul.net>*
11300 * Add cipher selection rules COMPLEMENTOFALL and COMPLEMENTOFDEFAULT
11301 to allow version independent disabling of normally unselected ciphers,
11302 which may be activated as a side-effect of selecting a single cipher.
11304 (E.g., cipher list string "RSA" enables ciphersuites that are left
11305 out of "ALL" because they do not provide symmetric encryption.
11306 "RSA:!COMPLEMEMENTOFALL" avoids these unsafe ciphersuites.)
11308 *Lutz Jaenicke, Bodo Moeller*
11310 * Add appropriate support for separate platform-dependent build
11311 directories. The recommended way to make a platform-dependent
11312 build directory is the following (tested on Linux), maybe with
11315 # Place yourself outside of the OpenSSL source tree. In
11316 # this example, the environment variable OPENSSL_SOURCE
11317 # is assumed to contain the absolute OpenSSL source directory.
11318 mkdir -p objtree/"`uname -s`-`uname -r`-`uname -m`"
11319 cd objtree/"`uname -s`-`uname -r`-`uname -m`"
11320 (cd $OPENSSL_SOURCE; find . -type f) | while read F; do
11321 mkdir -p `dirname $F`
11322 ln -s $OPENSSL_SOURCE/$F $F
11325 To be absolutely sure not to disturb the source tree, a "make clean"
11326 is a good thing. If it isn't successful, don't worry about it,
11327 it probably means the source directory is very clean.
11331 * Make sure any ENGINE control commands make local copies of string
11332 pointers passed to them whenever necessary. Otherwise it is possible
11333 the caller may have overwritten (or deallocated) the original string
11334 data when a later ENGINE operation tries to use the stored values.
11336 *Götz Babin-Ebell <babinebell@trustcenter.de>*
11338 * Improve diagnostics in file reading and command-line digests.
11340 *Ben Laurie aided and abetted by Solar Designer <solar@openwall.com>*
11342 * Add AES modes CFB and OFB to the object database. Correct an
11343 error in AES-CFB decryption.
11347 * Remove most calls to EVP_CIPHER_CTX_cleanup() in evp_enc.c, this
11348 allows existing EVP_CIPHER_CTX structures to be reused after
11349 calling `EVP_*Final()`. This behaviour is used by encryption
11350 BIOs and some applications. This has the side effect that
11351 applications must explicitly clean up cipher contexts with
11352 EVP_CIPHER_CTX_cleanup() or they will leak memory.
11356 * Check the values of dna and dnb in bn_mul_recursive before calling
11357 bn_mul_comba (a non zero value means the a or b arrays do not contain
11358 n2 elements) and fallback to bn_mul_normal if either is not zero.
11362 * Fix escaping of non-ASCII characters when using the -subj option
11363 of the "openssl req" command line tool. (Robert Joop <joop@fokus.gmd.de>)
11367 * Make object definitions compliant to LDAP (RFC2256): SN is the short
11368 form for "surname", serialNumber has no short form.
11369 Use "mail" as the short name for "rfc822Mailbox" according to RFC2798;
11370 therefore remove "mail" short name for "internet 7".
11371 The OID for unique identifiers in X509 certificates is
11372 x500UniqueIdentifier, not uniqueIdentifier.
11373 Some more OID additions. (Michael Bell <michael.bell@rz.hu-berlin.de>)
11377 * Add an "init" command to the ENGINE config module and auto initialize
11378 ENGINEs. Without any "init" command the ENGINE will be initialized
11379 after all ctrl commands have been executed on it. If init=1 the
11380 ENGINE is initialized at that point (ctrls before that point are run
11381 on the uninitialized ENGINE and after on the initialized one). If
11382 init=0 then the ENGINE will not be initialized at all.
11386 * Fix the 'app_verify_callback' interface so that the user-defined
11387 argument is actually passed to the callback: In the
11388 SSL_CTX_set_cert_verify_callback() prototype, the callback
11389 declaration has been changed from
11392 int (*cb)(X509_STORE_CTX *,void *);
11393 in ssl_verify_cert_chain (ssl/ssl_cert.c), the call
11394 i=s->ctx->app_verify_callback(&ctx)
11395 has been changed into
11396 i=s->ctx->app_verify_callback(&ctx, s->ctx->app_verify_arg).
11398 To update applications using SSL_CTX_set_cert_verify_callback(),
11399 a dummy argument can be added to their callback functions.
11401 *D. K. Smetters <smetters@parc.xerox.com>*
11403 * Added the '4758cca' ENGINE to support IBM 4758 cards.
11405 *Maurice Gittens <maurice@gittens.nl>, touchups by Geoff Thorpe*
11407 * Add and OPENSSL_LOAD_CONF define which will cause
11408 OpenSSL_add_all_algorithms() to load the openssl.cnf config file.
11409 This allows older applications to transparently support certain
11410 OpenSSL features: such as crypto acceleration and dynamic ENGINE loading.
11411 Two new functions OPENSSL_add_all_algorithms_noconf() which will never
11412 load the config file and OPENSSL_add_all_algorithms_conf() which will
11413 always load it have also been added.
11417 * Add the OFB, CFB and CTR (all with 128 bit feedback) to AES.
11418 Adjust NIDs and EVP layer.
11420 *Stephen Sprunk <stephen@sprunk.org> and Richard Levitte*
11422 * Config modules support in openssl utility.
11424 Most commands now load modules from the config file,
11425 though in a few (such as version) this isn't done
11426 because it couldn't be used for anything.
11428 In the case of ca and req the config file used is
11429 the same as the utility itself: that is the -config
11430 command line option can be used to specify an
11435 * Move default behaviour from OPENSSL_config(). If appname is NULL
11436 use "openssl_conf" if filename is NULL use default openssl config file.
11440 * Add an argument to OPENSSL_config() to allow the use of an alternative
11441 config section name. Add a new flag to tolerate a missing config file
11442 and move code to CONF_modules_load_file().
11446 * Support for crypto accelerator cards from Accelerated Encryption
11447 Processing, www.aep.ie. (Use engine 'aep')
11448 The support was copied from 0.9.6c [engine] and adapted/corrected
11449 to work with the new engine framework.
11451 *AEP Inc. and Richard Levitte*
11453 * Support for SureWare crypto accelerator cards from Baltimore
11454 Technologies. (Use engine 'sureware')
11455 The support was copied from 0.9.6c [engine] and adapted
11456 to work with the new engine framework.
11460 * Have the CHIL engine fork-safe (as defined by nCipher) and actually
11461 make the newer ENGINE framework commands for the CHIL engine work.
11463 *Toomas Kiisk <vix@cyber.ee> and Richard Levitte*
11465 * Make it possible to produce shared libraries on ReliantUNIX.
11467 *Robert Dahlem <Robert.Dahlem@ffm2.siemens.de> via Richard Levitte*
11469 * Add the configuration target debug-linux-ppro.
11470 Make 'openssl rsa' use the general key loading routines
11471 implemented in apps.c, and make those routines able to
11472 handle the key format FORMAT_NETSCAPE and the variant
11475 *Toomas Kiisk <vix@cyber.ee> via Richard Levitte*
11477 * Fix a crashbug and a logic bug in hwcrhk_load_pubkey().
11479 *Toomas Kiisk <vix@cyber.ee> via Richard Levitte*
11481 * Add -keyform to rsautl, and document -engine.
11483 *Richard Levitte, inspired by Toomas Kiisk <vix@cyber.ee>*
11485 * Change BIO_new_file (crypto/bio/bss_file.c) to use new
11486 BIO_R_NO_SUCH_FILE error code rather than the generic
11487 ERR_R_SYS_LIB error code if fopen() fails with ENOENT.
11491 * Add new functions
11492 ERR_peek_last_error
11493 ERR_peek_last_error_line
11494 ERR_peek_last_error_line_data.
11495 These are similar to
11497 ERR_peek_error_line
11498 ERR_peek_error_line_data,
11499 but report on the latest error recorded rather than the first one
11500 still in the error queue.
11502 *Ben Laurie, Bodo Moeller*
11504 * default_algorithms option in ENGINE config module. This allows things
11506 default_algorithms = ALL
11507 default_algorithms = RSA, DSA, RAND, CIPHERS, DIGESTS
11511 * Preliminary ENGINE config module.
11515 * New experimental application configuration code.
11519 * Change the AES code to follow the same name structure as all other
11520 symmetric ciphers, and behave the same way. Move everything to
11521 the directory crypto/aes, thereby obsoleting crypto/rijndael.
11523 *Stephen Sprunk <stephen@sprunk.org> and Richard Levitte*
11525 * SECURITY: remove unsafe setjmp/signal interaction from ui_openssl.c.
11527 *Ben Laurie and Theo de Raadt*
11529 * Add option to output public keys in req command.
11531 *Massimiliano Pala madwolf@openca.org*
11533 * Use wNAFs in EC_POINTs_mul() for improved efficiency
11534 (up to about 10% better than before for P-192 and P-224).
11538 * New functions/macros
11540 SSL_CTX_set_msg_callback(ctx, cb)
11541 SSL_CTX_set_msg_callback_arg(ctx, arg)
11542 SSL_set_msg_callback(ssl, cb)
11543 SSL_set_msg_callback_arg(ssl, arg)
11545 to request calling a callback function
11547 void cb(int write_p, int version, int content_type,
11548 const void *buf, size_t len, SSL *ssl, void *arg)
11550 whenever a protocol message has been completely received
11551 (write_p == 0) or sent (write_p == 1). Here 'version' is the
11552 protocol version according to which the SSL library interprets
11553 the current protocol message (SSL2_VERSION, SSL3_VERSION, or
11554 TLS1_VERSION). 'content_type' is 0 in the case of SSL 2.0, or
11555 the content type as defined in the SSL 3.0/TLS 1.0 protocol
11556 specification (change_cipher_spec(20), alert(21), handshake(22)).
11557 'buf' and 'len' point to the actual message, 'ssl' to the
11558 SSL object, and 'arg' is the application-defined value set by
11559 SSL[_CTX]_set_msg_callback_arg().
11561 'openssl s_client' and 'openssl s_server' have new '-msg' options
11562 to enable a callback that displays all protocol messages.
11566 * Change the shared library support so shared libraries are built as
11567 soon as the corresponding static library is finished, and thereby get
11568 openssl and the test programs linked against the shared library.
11569 This still only happens when the keyword "shard" has been given to
11570 the configuration scripts.
11572 NOTE: shared library support is still an experimental thing, and
11573 backward binary compatibility is still not guaranteed.
11575 *"Maciej W. Rozycki" <macro@ds2.pg.gda.pl> and Richard Levitte*
11577 * Add support for Subject Information Access extension.
11579 *Peter Sylvester <Peter.Sylvester@EdelWeb.fr>*
11581 * Make BUF_MEM_grow() behaviour more consistent: Initialise to zero
11582 additional bytes when new memory had to be allocated, not just
11583 when reusing an existing buffer.
11587 * New command line and configuration option 'utf8' for the req command.
11588 This allows field values to be specified as UTF8 strings.
11592 * Add -multi and -mr options to "openssl speed" - giving multiple parallel
11593 runs for the former and machine-readable output for the latter.
11597 * Add '-noemailDN' option to 'openssl ca'. This prevents inclusion
11598 of the e-mail address in the DN (i.e., it will go into a certificate
11599 extension only). The new configuration file option 'email_in_dn = no'
11600 has the same effect.
11602 *Massimiliano Pala madwolf@openca.org*
11604 * Change all functions with names starting with `des_` to be starting
11605 with `DES_` instead. Add wrappers that are compatible with libdes,
11606 but are named `_ossl_old_des_*`. Finally, add macros that map the
11607 `des_*` symbols to the corresponding `_ossl_old_des_*` if libdes
11608 compatibility is desired. If OpenSSL 0.9.6c compatibility is
11609 desired, the `des_*` symbols will be mapped to `DES_*`, with one
11612 Since we provide two compatibility mappings, the user needs to
11613 define the macro OPENSSL_DES_LIBDES_COMPATIBILITY if libdes
11614 compatibility is desired. The default (i.e., when that macro
11615 isn't defined) is OpenSSL 0.9.6c compatibility.
11617 There are also macros that enable and disable the support of old
11618 des functions altogether. Those are OPENSSL_ENABLE_OLD_DES_SUPPORT
11619 and OPENSSL_DISABLE_OLD_DES_SUPPORT. If none or both of those
11620 are defined, the default will apply: to support the old des routines.
11622 In either case, one must include openssl/des.h to get the correct
11623 definitions. Do not try to just include openssl/des_old.h, that
11626 NOTE: This is a major break of an old API into a new one. Software
11627 authors are encouraged to switch to the `DES_` style functions. Some
11628 time in the future, des_old.h and the libdes compatibility functions
11629 will be disable (i.e. OPENSSL_DISABLE_OLD_DES_SUPPORT will be the
11630 default), and then completely removed.
11634 * Test for certificates which contain unsupported critical extensions.
11635 If such a certificate is found during a verify operation it is
11636 rejected by default: this behaviour can be overridden by either
11637 handling the new error X509_V_ERR_UNHANDLED_CRITICAL_EXTENSION or
11638 by setting the verify flag X509_V_FLAG_IGNORE_CRITICAL. A new function
11639 X509_supported_extension() has also been added which returns 1 if a
11640 particular extension is supported.
11644 * Modify the behaviour of EVP cipher functions in similar way to digests
11645 to retain compatibility with existing code.
11649 * Modify the behaviour of EVP_DigestInit() and EVP_DigestFinal() to retain
11650 compatibility with existing code. In particular the 'ctx' parameter does
11651 not have to be to be initialized before the call to EVP_DigestInit() and
11652 it is tidied up after a call to EVP_DigestFinal(). New function
11653 EVP_DigestFinal_ex() which does not tidy up the ctx. Similarly function
11654 EVP_MD_CTX_copy() changed to not require the destination to be
11655 initialized valid and new function EVP_MD_CTX_copy_ex() added which
11656 requires the destination to be valid.
11658 Modify all the OpenSSL digest calls to use EVP_DigestInit_ex(),
11659 EVP_DigestFinal_ex() and EVP_MD_CTX_copy_ex().
11663 * Change ssl3_get_message (ssl/s3_both.c) and the functions using it
11664 so that complete 'Handshake' protocol structures are kept in memory
11665 instead of overwriting 'msg_type' and 'length' with 'body' data.
11669 * Add an implementation of SSL_add_dir_cert_subjects_to_stack for Win32.
11671 *Massimo Santin via Richard Levitte*
11673 * Major restructuring to the underlying ENGINE code. This includes
11674 reduction of linker bloat, separation of pure "ENGINE" manipulation
11675 (initialisation, etc) from functionality dealing with implementations
11676 of specific crypto interfaces. This change also introduces integrated
11677 support for symmetric ciphers and digest implementations - so ENGINEs
11678 can now accelerate these by providing EVP_CIPHER and EVP_MD
11679 implementations of their own. This is detailed in
11680 [crypto/engine/README.md](crypto/engine/README.md)
11681 as it couldn't be adequately described here. However, there are a few
11682 API changes worth noting - some RSA, DSA, DH, and RAND functions that
11683 were changed in the original introduction of ENGINE code have now
11684 reverted back - the hooking from this code to ENGINE is now a good
11685 deal more passive and at run-time, operations deal directly with
11686 RSA_METHODs, DSA_METHODs (etc) as they did before, rather than
11687 dereferencing through an ENGINE pointer any more. Also, the ENGINE
11688 functions dealing with `BN_MOD_EXP[_CRT]` handlers have been removed -
11689 they were not being used by the framework as there is no concept of a
11690 BIGNUM_METHOD and they could not be generalised to the new
11691 'ENGINE_TABLE' mechanism that underlies the new code. Similarly,
11692 ENGINE_cpy() has been removed as it cannot be consistently defined in
11697 * Change ASN1_GENERALIZEDTIME_check() to allow fractional seconds.
11701 * Change mkdef.pl to sort symbols that get the same entry number,
11702 and make sure the automatically generated functions `ERR_load_*`
11703 become part of libeay.num as well.
11707 * New function SSL_renegotiate_pending(). This returns true once
11708 renegotiation has been requested (either SSL_renegotiate() call
11709 or HelloRequest/ClientHello received from the peer) and becomes
11710 false once a handshake has been completed.
11711 (For servers, SSL_renegotiate() followed by SSL_do_handshake()
11712 sends a HelloRequest, but does not ensure that a handshake takes
11713 place. SSL_renegotiate_pending() is useful for checking if the
11714 client has followed the request.)
11718 * New SSL option SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION.
11719 By default, clients may request session resumption even during
11720 renegotiation (if session ID contexts permit); with this option,
11721 session resumption is possible only in the first handshake.
11723 SSL_OP_ALL is now 0x00000FFFL instead of 0x000FFFFFL. This makes
11724 more bits available for options that should not be part of
11725 SSL_OP_ALL (such as SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION).
11729 * Add some demos for certificate and certificate request creation.
11733 * Make maximum certificate chain size accepted from the peer application
11734 settable (`SSL*_get/set_max_cert_list()`), as proposed by
11735 "Douglas E. Engert" <deengert@anl.gov>.
11739 * Add support for shared libraries for Unixware-7
11740 (Boyd Lynn Gerber <gerberb@zenez.com>).
11744 * Add a "destroy" handler to ENGINEs that allows structural cleanup to
11745 be done prior to destruction. Use this to unload error strings from
11746 ENGINEs that load their own error strings. NB: This adds two new API
11747 functions to "get" and "set" this destroy handler in an ENGINE.
11751 * Alter all existing ENGINE implementations (except "openssl" and
11752 "openbsd") to dynamically instantiate their own error strings. This
11753 makes them more flexible to be built both as statically-linked ENGINEs
11754 and self-contained shared-libraries loadable via the "dynamic" ENGINE.
11755 Also, add stub code to each that makes building them as self-contained
11756 shared-libraries easier (see [README-Engine.md](README-Engine.md)).
11760 * Add a "dynamic" ENGINE that provides a mechanism for binding ENGINE
11761 implementations into applications that are completely implemented in
11762 self-contained shared-libraries. The "dynamic" ENGINE exposes control
11763 commands that can be used to configure what shared-library to load and
11764 to control aspects of the way it is handled. Also, made an update to
11765 the [README-Engine.md](README-Engine.md) file
11766 that brings its information up-to-date and
11767 provides some information and instructions on the "dynamic" ENGINE
11768 (ie. how to use it, how to build "dynamic"-loadable ENGINEs, etc).
11772 * Make it possible to unload ranges of ERR strings with a new
11773 "ERR_unload_strings" function.
11777 * Add a copy() function to EVP_MD.
11781 * Make EVP_MD routines take a context pointer instead of just the
11782 md_data void pointer.
11786 * Add flags to EVP_MD and EVP_MD_CTX. EVP_MD_FLAG_ONESHOT indicates
11787 that the digest can only process a single chunk of data
11788 (typically because it is provided by a piece of
11789 hardware). EVP_MD_CTX_FLAG_ONESHOT indicates that the application
11790 is only going to provide a single chunk of data, and hence the
11791 framework needn't accumulate the data for oneshot drivers.
11795 * As with "ERR", make it possible to replace the underlying "ex_data"
11796 functions. This change also alters the storage and management of global
11797 ex_data state - it's now all inside ex_data.c and all "class" code (eg.
11798 RSA, BIO, SSL_CTX, etc) no longer stores its own STACKS and per-class
11799 index counters. The API functions that use this state have been changed
11800 to take a "class_index" rather than pointers to the class's local STACK
11801 and counter, and there is now an API function to dynamically create new
11802 classes. This centralisation allows us to (a) plug a lot of the
11803 thread-safety problems that existed, and (b) makes it possible to clean
11804 up all allocated state using "CRYPTO_cleanup_all_ex_data()". W.r.t. (b)
11805 such data would previously have always leaked in application code and
11806 workarounds were in place to make the memory debugging turn a blind eye
11807 to it. Application code that doesn't use this new function will still
11808 leak as before, but their memory debugging output will announce it now
11809 rather than letting it slide.
11811 Besides the addition of CRYPTO_cleanup_all_ex_data(), another API change
11812 induced by the "ex_data" overhaul is that X509_STORE_CTX_init() now
11813 has a return value to indicate success or failure.
11817 * Make it possible to replace the underlying "ERR" functions such that the
11818 global state (2 LHASH tables and 2 locks) is only used by the "default"
11819 implementation. This change also adds two functions to "get" and "set"
11820 the implementation prior to it being automatically set the first time
11821 any other ERR function takes place. Ie. an application can call "get",
11822 pass the return value to a module it has just loaded, and that module
11823 can call its own "set" function using that value. This means the
11824 module's "ERR" operations will use (and modify) the error state in the
11825 application and not in its own statically linked copy of OpenSSL code.
11829 * Give DH, DSA, and RSA types their own `*_up_ref()` function to increment
11830 reference counts. This performs normal REF_PRINT/REF_CHECK macros on
11831 the operation, and provides a more encapsulated way for external code
11832 (crypto/evp/ and ssl/) to do this. Also changed the evp and ssl code
11833 to use these functions rather than manually incrementing the counts.
11835 Also rename "DSO_up()" function to more descriptive "DSO_up_ref()".
11839 * Add EVP test program.
11843 * Add symmetric cipher support to ENGINE. Expect the API to change!
11847 * New CRL functions: X509_CRL_set_version(), X509_CRL_set_issuer_name()
11848 X509_CRL_set_lastUpdate(), X509_CRL_set_nextUpdate(), X509_CRL_sort(),
11849 X509_REVOKED_set_serialNumber(), and X509_REVOKED_set_revocationDate().
11850 These allow a CRL to be built without having to access X509_CRL fields
11851 directly. Modify 'ca' application to use new functions.
11855 * Move SSL_OP_TLS_ROLLBACK_BUG out of the SSL_OP_ALL list of recommended
11856 bug workarounds. Rollback attack detection is a security feature.
11857 The problem will only arise on OpenSSL servers when TLSv1 is not
11858 available (sslv3_server_method() or SSL_OP_NO_TLSv1).
11859 Software authors not wanting to support TLSv1 will have special reasons
11860 for their choice and can explicitly enable this option.
11862 *Bodo Moeller, Lutz Jaenicke*
11864 * Rationalise EVP so it can be extended: don't include a union of
11865 cipher/digest structures, add init/cleanup functions for EVP_MD_CTX
11866 (similar to those existing for EVP_CIPHER_CTX).
11871 EVP_MD_CTX_init(&md); /* new function call */
11872 EVP_DigestInit(&md, EVP_sha1());
11873 EVP_DigestUpdate(&md, in, len);
11874 EVP_DigestFinal(&md, out, NULL);
11875 EVP_MD_CTX_cleanup(&md); /* new function call */
11879 * Make DES key schedule conform to the usual scheme, as well as
11880 correcting its structure. This means that calls to DES functions
11881 now have to pass a pointer to a des_key_schedule instead of a
11882 plain des_key_schedule (which was actually always a pointer
11885 des_key_schedule ks;
11887 des_set_key_checked(..., &ks);
11888 des_ncbc_encrypt(..., &ks, ...);
11890 (Note that a later change renames 'des_...' into 'DES_...'.)
11894 * Initial reduction of linker bloat: the use of some functions, such as
11895 PEM causes large amounts of unused functions to be linked in due to
11896 poor organisation. For example pem_all.c contains every PEM function
11897 which has a knock on effect of linking in large amounts of (unused)
11898 ASN1 code. Grouping together similar functions and splitting unrelated
11899 functions prevents this.
11903 * Cleanup of EVP macros.
11907 * Change historical references to `{NID,SN,LN}_des_ede` and ede3 to add the
11908 correct `_ecb suffix`.
11912 * Add initial OCSP responder support to ocsp application. The
11913 revocation information is handled using the text based index
11914 use by the ca application. The responder can either handle
11915 requests generated internally, supplied in files (for example
11916 via a CGI script) or using an internal minimal server.
11920 * Add configuration choices to get zlib compression for TLS.
11924 * Changes to Kerberos SSL for RFC 2712 compliance:
11925 1. Implemented real KerberosWrapper, instead of just using
11926 KRB5 AP_REQ message. [Thanks to Simon Wilkinson <sxw@sxw.org.uk>]
11927 2. Implemented optional authenticator field of KerberosWrapper.
11929 Added openssl-style ASN.1 macros for Kerberos ticket, ap_req,
11930 and authenticator structs; see crypto/krb5/.
11932 Generalized Kerberos calls to support multiple Kerberos libraries.
11933 *Vern Staats <staatsvr@asc.hpc.mil>, Jeffrey Altman <jaltman@columbia.edu>
11934 via Richard Levitte*
11936 * Cause 'openssl speed' to use fully hard-coded DSA keys as it
11937 already does with RSA. testdsa.h now has 'priv_key/pub_key'
11938 values for each of the key sizes rather than having just
11939 parameters (and 'speed' generating keys each time).
11943 * Speed up EVP routines.
11946 pe 8 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes
11947 s-cbc 4408.85k 5560.51k 5778.46k 5862.20k 5825.16k
11948 s-cbc 4389.55k 5571.17k 5792.23k 5846.91k 5832.11k
11949 s-cbc 4394.32k 5575.92k 5807.44k 5848.37k 5841.30k
11951 s-cbc 3482.66k 5069.49k 5496.39k 5614.16k 5639.28k
11952 s-cbc 3480.74k 5068.76k 5510.34k 5609.87k 5635.52k
11953 s-cbc 3483.72k 5067.62k 5504.60k 5708.01k 5724.80k
11956 s-cbc 4660.16k 5650.19k 5807.19k 5827.13k 5783.32k
11958 s-cbc 3624.96k 5258.21k 5530.91k 5624.30k 5628.26k
11962 * Added the OS2-EMX target.
11964 *"Brian Havard" <brianh@kheldar.apana.org.au> and Richard Levitte*
11966 * Rewrite apps to use NCONF routines instead of the old CONF. New functions
11967 to support NCONF routines in extension code. New function CONF_set_nconf()
11968 to allow functions which take an NCONF to also handle the old LHASH
11969 structure: this means that the old CONF compatible routines can be
11970 retained (in particular wrt extensions) without having to duplicate the
11971 code. New function X509V3_add_ext_nconf_sk to add extensions to a stack.
11975 * Enhance the general user interface with mechanisms for inner control
11976 and with possibilities to have yes/no kind of prompts.
11980 * Change all calls to low level digest routines in the library and
11981 applications to use EVP. Add missing calls to HMAC_cleanup() and
11982 don't assume HMAC_CTX can be copied using memcpy().
11984 *Verdon Walker <VWalker@novell.com>, Steve Henson*
11986 * Add the possibility to control engines through control names but with
11987 arbitrary arguments instead of just a string.
11988 Change the key loaders to take a UI_METHOD instead of a callback
11989 function pointer. NOTE: this breaks binary compatibility with earlier
11990 versions of OpenSSL [engine].
11991 Adapt the nCipher code for these new conditions and add a card insertion
11996 * Enhance the general user interface with mechanisms to better support
11997 dialog box interfaces, application-defined prompts, the possibility
11998 to use defaults (for example default passwords from somewhere else)
11999 and interrupts/cancellations.
12003 * Tidy up PKCS#12 attribute handling. Add support for the CSP name
12004 attribute in PKCS#12 files, add new -CSP option to pkcs12 utility.
12008 * Fix a memory leak in 'sk_dup()' in the case reallocation fails. (Also
12009 tidy up some unnecessarily weird code in 'sk_new()').
12011 *Geoff, reported by Diego Tartara <dtartara@novamens.com>*
12013 * Change the key loading routines for ENGINEs to use the same kind
12014 callback (pem_password_cb) as all other routines that need this
12019 * Increase ENTROPY_NEEDED to 32 bytes, as Rijndael can operate with
12020 256 bit (=32 byte) keys. Of course seeding with more entropy bytes
12021 than this minimum value is recommended.
12025 * New random seeder for OpenVMS, using the system process statistics
12026 that are easily reachable.
12030 * Windows apparently can't transparently handle global
12031 variables defined in DLLs. Initialisations such as:
12033 const ASN1_ITEM *it = &ASN1_INTEGER_it;
12035 won't compile. This is used by the any applications that need to
12036 declare their own ASN1 modules. This was fixed by adding the option
12037 EXPORT_VAR_AS_FN to all Win32 platforms, although this isn't strictly
12038 needed for static libraries under Win32.
12042 * New functions X509_PURPOSE_set() and X509_TRUST_set() to handle
12043 setting of purpose and trust fields. New X509_STORE trust and
12044 purpose functions and tidy up setting in other SSL functions.
12048 * Add copies of X509_STORE_CTX fields and callbacks to X509_STORE
12049 structure. These are inherited by X509_STORE_CTX when it is
12050 initialised. This allows various defaults to be set in the
12051 X509_STORE structure (such as flags for CRL checking and custom
12052 purpose or trust settings) for functions which only use X509_STORE_CTX
12053 internally such as S/MIME.
12055 Modify X509_STORE_CTX_purpose_inherit() so it only sets purposes and
12056 trust settings if they are not set in X509_STORE. This allows X509_STORE
12057 purposes and trust (in S/MIME for example) to override any set by default.
12059 Add command line options for CRL checking to smime, s_client and s_server
12064 * Initial CRL based revocation checking. If the CRL checking flag(s)
12065 are set then the CRL is looked up in the X509_STORE structure and
12066 its validity and signature checked, then if the certificate is found
12067 in the CRL the verify fails with a revoked error.
12069 Various new CRL related callbacks added to X509_STORE_CTX structure.
12071 Command line options added to 'verify' application to support this.
12073 This needs some additional work, such as being able to handle multiple
12074 CRLs with different times, extension based lookup (rather than just
12075 by subject name) and ultimately more complete V2 CRL extension
12080 * Add a general user interface API (crypto/ui/). This is designed
12081 to replace things like des_read_password and friends (backward
12082 compatibility functions using this new API are provided).
12083 The purpose is to remove prompting functions from the DES code
12084 section as well as provide for prompting through dialog boxes in
12085 a window system and the like.
12089 * Add "ex_data" support to ENGINE so implementations can add state at a
12090 per-structure level rather than having to store it globally.
12094 * Make it possible for ENGINE structures to be copied when retrieved by
12095 ENGINE_by_id() if the ENGINE specifies a new flag: ENGINE_FLAGS_BY_ID_COPY.
12096 This causes the "original" ENGINE structure to act like a template,
12097 analogous to the RSA vs. RSA_METHOD type of separation. Because of this
12098 operational state can be localised to each ENGINE structure, despite the
12099 fact they all share the same "methods". New ENGINE structures returned in
12100 this case have no functional references and the return value is the single
12101 structural reference. This matches the single structural reference returned
12102 by ENGINE_by_id() normally, when it is incremented on the pre-existing
12107 * Fix ASN1 decoder when decoding type ANY and V_ASN1_OTHER: since this
12108 needs to match any other type at all we need to manually clear the
12113 * Changes to the "openssl engine" utility to include;
12114 - verbosity levels ('-v', '-vv', and '-vvv') that provide information
12115 about an ENGINE's available control commands.
12116 - executing control commands from command line arguments using the
12117 '-pre' and '-post' switches. '-post' is only used if '-t' is
12118 specified and the ENGINE is successfully initialised. The syntax for
12119 the individual commands are colon-separated, for example;
12120 openssl engine chil -pre FORK_CHECK:0 -pre SO_PATH:/lib/test.so
12124 * New dynamic control command support for ENGINEs. ENGINEs can now
12125 declare their own commands (numbers), names (strings), descriptions,
12126 and input types for run-time discovery by calling applications. A
12127 subset of these commands are implicitly classed as "executable"
12128 depending on their input type, and only these can be invoked through
12129 the new string-based API function ENGINE_ctrl_cmd_string(). (Eg. this
12130 can be based on user input, config files, etc). The distinction is
12131 that "executable" commands cannot return anything other than a boolean
12132 result and can only support numeric or string input, whereas some
12133 discoverable commands may only be for direct use through
12134 ENGINE_ctrl(), eg. supporting the exchange of binary data, function
12135 pointers, or other custom uses. The "executable" commands are to
12136 support parameterisations of ENGINE behaviour that can be
12137 unambiguously defined by ENGINEs and used consistently across any
12138 OpenSSL-based application. Commands have been added to all the
12139 existing hardware-supporting ENGINEs, noticeably "SO_PATH" to allow
12140 control over shared-library paths without source code alterations.
12144 * Changed all ENGINE implementations to dynamically allocate their
12145 ENGINEs rather than declaring them statically. Apart from this being
12146 necessary with the removal of the ENGINE_FLAGS_MALLOCED distinction,
12147 this also allows the implementations to compile without using the
12148 internal engine_int.h header.
12152 * Minor adjustment to "rand" code. RAND_get_rand_method() now returns a
12153 'const' value. Any code that should be able to modify a RAND_METHOD
12154 should already have non-const pointers to it (ie. they should only
12155 modify their own ones).
12159 * Made a variety of little tweaks to the ENGINE code.
12160 - "atalla" and "ubsec" string definitions were moved from header files
12161 to C code. "nuron" string definitions were placed in variables
12162 rather than hard-coded - allowing parameterisation of these values
12163 later on via ctrl() commands.
12164 - Removed unused "#if 0"'d code.
12165 - Fixed engine list iteration code so it uses ENGINE_free() to release
12166 structural references.
12167 - Constified the RAND_METHOD element of ENGINE structures.
12168 - Constified various get/set functions as appropriate and added
12169 missing functions (including a catch-all ENGINE_cpy that duplicates
12170 all ENGINE values onto a new ENGINE except reference counts/state).
12171 - Removed NULL parameter checks in get/set functions. Setting a method
12172 or function to NULL is a way of cancelling out a previously set
12173 value. Passing a NULL ENGINE parameter is just plain stupid anyway
12174 and doesn't justify the extra error symbols and code.
12175 - Deprecate the ENGINE_FLAGS_MALLOCED define and move the area for
12176 flags from engine_int.h to engine.h.
12177 - Changed prototypes for ENGINE handler functions (init(), finish(),
12178 ctrl(), key-load functions, etc) to take an (ENGINE*) parameter.
12182 * Implement binary inversion algorithm for BN_mod_inverse in addition
12183 to the algorithm using long division. The binary algorithm can be
12184 used only if the modulus is odd. On 32-bit systems, it is faster
12185 only for relatively small moduli (roughly 20-30% for 128-bit moduli,
12186 roughly 5-15% for 256-bit moduli), so we use it only for moduli
12187 up to 450 bits. In 64-bit environments, the binary algorithm
12188 appears to be advantageous for much longer moduli; here we use it
12189 for moduli up to 2048 bits.
12193 * Rewrite CHOICE field setting in ASN1_item_ex_d2i(). The old code
12194 could not support the combine flag in choice fields.
12198 * Add a 'copy_extensions' option to the 'ca' utility. This copies
12199 extensions from a certificate request to the certificate.
12203 * Allow multiple 'certopt' and 'nameopt' options to be separated
12204 by commas. Add 'namopt' and 'certopt' options to the 'ca' config
12205 file: this allows the display of the certificate about to be
12206 signed to be customised, to allow certain fields to be included
12207 or excluded and extension details. The old system didn't display
12208 multicharacter strings properly, omitted fields not in the policy
12209 and couldn't display additional details such as extensions.
12213 * Function EC_POINTs_mul for multiple scalar multiplication
12214 of an arbitrary number of elliptic curve points
12215 \sum scalars[i]*points[i],
12216 optionally including the generator defined for the EC_GROUP:
12217 scalar*generator + \sum scalars[i]*points[i].
12219 EC_POINT_mul is a simple wrapper function for the typical case
12220 that the point list has just one item (besides the optional
12225 * First EC_METHODs for curves over GF(p):
12227 EC_GFp_simple_method() uses the basic BN_mod_mul and BN_mod_sqr
12228 operations and provides various method functions that can also
12229 operate with faster implementations of modular arithmetic.
12231 EC_GFp_mont_method() reuses most functions that are part of
12232 EC_GFp_simple_method, but uses Montgomery arithmetic.
12234 *Bodo Moeller; point addition and point doubling
12235 implementation directly derived from source code provided by
12236 Lenka Fibikova <fibikova@exp-math.uni-essen.de>*
12238 * Framework for elliptic curves (crypto/ec/ec.h, crypto/ec/ec_lcl.h,
12239 crypto/ec/ec_lib.c):
12241 Curves are EC_GROUP objects (with an optional group generator)
12242 based on EC_METHODs that are built into the library.
12244 Points are EC_POINT objects based on EC_GROUP objects.
12246 Most of the framework would be able to handle curves over arbitrary
12247 finite fields, but as there are no obvious types for fields other
12248 than GF(p), some functions are limited to that for now.
12252 * Add the -HTTP option to s_server. It is similar to -WWW, but requires
12253 that the file contains a complete HTTP response.
12257 * Add the ec directory to mkdef.pl and mkfiles.pl. In mkdef.pl
12258 change the def and num file printf format specifier from "%-40sXXX"
12259 to "%-39s XXX". The latter will always guarantee a space after the
12260 field while the former will cause them to run together if the field
12261 is 40 of more characters long.
12265 * Constify the cipher and digest 'method' functions and structures
12266 and modify related functions to take constant EVP_MD and EVP_CIPHER
12271 * Hide BN_CTX structure details in bn_lcl.h instead of publishing them
12272 in <openssl/bn.h>. Also further increase BN_CTX_NUM to 32.
12276 * Modify `EVP_Digest*()` routines so they now return values. Although the
12277 internal software routines can never fail additional hardware versions
12282 * Clean up crypto/err/err.h and change some error codes to avoid conflicts:
12284 Previously ERR_R_FATAL was too small and coincided with ERR_LIB_PKCS7
12285 (= ERR_R_PKCS7_LIB); it is now 64 instead of 32.
12288 ERR_R_NESTED_ASN1_ERROR
12290 ERR_R_MISSING_ASN1_EOS
12291 were 4 .. 9, conflicting with
12292 ERR_LIB_RSA (= ERR_R_RSA_LIB)
12294 ERR_LIB_PEM (= ERR_R_PEM_LIB).
12295 They are now 58 .. 63 (i.e., just below ERR_R_FATAL).
12297 Add new error code 'ERR_R_INTERNAL_ERROR'.
12301 * Don't overuse locks in crypto/err/err.c: For data retrieval, CRYPTO_r_lock
12306 * New option '-subj arg' for 'openssl req' and 'openssl ca'. This
12307 sets the subject name for a new request or supersedes the
12308 subject name in a given request. Formats that can be parsed are
12309 'CN=Some Name, OU=myOU, C=IT'
12311 'CN=Some Name/OU=myOU/C=IT'.
12313 Add options '-batch' and '-verbose' to 'openssl req'.
12315 *Massimiliano Pala <madwolf@hackmasters.net>*
12317 * Introduce the possibility to access global variables through
12318 functions on platform were that's the best way to handle exporting
12319 global variables in shared libraries. To enable this functionality,
12320 one must configure with "EXPORT_VAR_AS_FN" or defined the C macro
12321 "OPENSSL_EXPORT_VAR_AS_FUNCTION" in crypto/opensslconf.h (the latter
12322 is normally done by Configure or something similar).
12324 To implement a global variable, use the macro OPENSSL_IMPLEMENT_GLOBAL
12325 in the source file (foo.c) like this:
12327 OPENSSL_IMPLEMENT_GLOBAL(int,foo)=1;
12328 OPENSSL_IMPLEMENT_GLOBAL(double,bar);
12330 To declare a global variable, use the macros OPENSSL_DECLARE_GLOBAL
12331 and OPENSSL_GLOBAL_REF in the header file (foo.h) like this:
12333 OPENSSL_DECLARE_GLOBAL(int,foo);
12334 #define foo OPENSSL_GLOBAL_REF(foo)
12335 OPENSSL_DECLARE_GLOBAL(double,bar);
12336 #define bar OPENSSL_GLOBAL_REF(bar)
12338 The #defines are very important, and therefore so is including the
12339 header file everywhere where the defined globals are used.
12341 The macro OPENSSL_EXPORT_VAR_AS_FUNCTION also affects the definition
12342 of ASN.1 items, but that structure is a bit different.
12344 The largest change is in util/mkdef.pl which has been enhanced with
12345 better and easier to understand logic to choose which symbols should
12346 go into the Windows .def files as well as a number of fixes and code
12347 cleanup (among others, algorithm keywords are now sorted
12348 lexicographically to avoid constant rewrites).
12352 * In BN_div() keep a copy of the sign of 'num' before writing the
12353 result to 'rm' because if rm==num the value will be overwritten
12354 and produce the wrong result if 'num' is negative: this caused
12355 problems with BN_mod() and BN_nnmod().
12359 * Function OCSP_request_verify(). This checks the signature on an
12360 OCSP request and verifies the signer certificate. The signer
12361 certificate is just checked for a generic purpose and OCSP request
12366 * Add OCSP_check_validity() function to check the validity of OCSP
12367 responses. OCSP responses are prepared in real time and may only
12368 be a few seconds old. Simply checking that the current time lies
12369 between thisUpdate and nextUpdate max reject otherwise valid responses
12370 caused by either OCSP responder or client clock inaccuracy. Instead
12371 we allow thisUpdate and nextUpdate to fall within a certain period of
12372 the current time. The age of the response can also optionally be
12373 checked. Two new options -validity_period and -status_age added to
12378 * If signature or public key algorithm is unrecognized print out its
12379 OID rather that just UNKNOWN.
12383 * Change OCSP_cert_to_id() to tolerate a NULL subject certificate and
12384 OCSP_cert_id_new() a NULL serialNumber. This allows a partial certificate
12385 ID to be generated from the issuer certificate alone which can then be
12386 passed to OCSP_id_issuer_cmp().
12390 * New compilation option ASN1_ITEM_FUNCTIONS. This causes the new
12391 ASN1 modules to export functions returning ASN1_ITEM pointers
12392 instead of the ASN1_ITEM structures themselves. This adds several
12393 new macros which allow the underlying ASN1 function/structure to
12394 be accessed transparently. As a result code should not use ASN1_ITEM
12395 references directly (such as &X509_it) but instead use the relevant
12396 macros (such as ASN1_ITEM_rptr(X509)). This option is to allow
12397 use of the new ASN1 code on platforms where exporting structures
12398 is problematical (for example in shared libraries) but exporting
12399 functions returning pointers to structures is not.
12403 * Add support for overriding the generation of SSL/TLS session IDs.
12404 These callbacks can be registered either in an SSL_CTX or per SSL.
12405 The purpose of this is to allow applications to control, if they wish,
12406 the arbitrary values chosen for use as session IDs, particularly as it
12407 can be useful for session caching in multiple-server environments. A
12408 command-line switch for testing this (and any client code that wishes
12409 to use such a feature) has been added to "s_server".
12411 *Geoff Thorpe, Lutz Jaenicke*
12413 * Modify mkdef.pl to recognise and parse preprocessor conditionals
12414 of the form `#if defined(...) || defined(...) || ...` and
12415 `#if !defined(...) && !defined(...) && ...`. This also avoids
12416 the growing number of special cases it was previously handling.
12420 * Make all configuration macros available for application by making
12421 sure they are available in opensslconf.h, by giving them names starting
12422 with `OPENSSL_` to avoid conflicts with other packages and by making
12423 sure e_os2.h will cover all platform-specific cases together with
12425 Additionally, it is now possible to define configuration/platform-
12426 specific names (called "system identities"). In the C code, these
12427 are prefixed with `OPENSSL_SYSNAME_`. e_os2.h will create another
12428 macro with the name beginning with `OPENSSL_SYS_`, which is determined
12429 from `OPENSSL_SYSNAME_*` or compiler-specific macros depending on
12434 * New option -set_serial to 'req' and 'x509' this allows the serial
12435 number to use to be specified on the command line. Previously self
12436 signed certificates were hard coded with serial number 0 and the
12437 CA options of 'x509' had to use a serial number in a file which was
12442 * New options to 'ca' utility to support V2 CRL entry extensions.
12443 Currently CRL reason, invalidity date and hold instruction are
12444 supported. Add new CRL extensions to V3 code and some new objects.
12448 * New function EVP_CIPHER_CTX_set_padding() this is used to
12449 disable standard block padding (aka PKCS#5 padding) in the EVP
12450 API, which was previously mandatory. This means that the data is
12451 not padded in any way and so the total length much be a multiple
12452 of the block size, otherwise an error occurs.
12456 * Initial (incomplete) OCSP SSL support.
12460 * New function OCSP_parse_url(). This splits up a URL into its host,
12461 port and path components: primarily to parse OCSP URLs. New -url
12462 option to ocsp utility.
12466 * New nonce behavior. The return value of OCSP_check_nonce() now
12467 reflects the various checks performed. Applications can decide
12468 whether to tolerate certain situations such as an absent nonce
12469 in a response when one was present in a request: the ocsp application
12470 just prints out a warning. New function OCSP_add1_basic_nonce()
12471 this is to allow responders to include a nonce in a response even if
12472 the request is nonce-less.
12476 * Disable stdin buffering in load_cert (apps/apps.c) so that no certs are
12477 skipped when using openssl x509 multiple times on a single input file,
12478 e.g. `(openssl x509 -out cert1; openssl x509 -out cert2) <certs`.
12482 * Make ASN1_UTCTIME_set_string() and ASN1_GENERALIZEDTIME_set_string()
12483 set string type: to handle setting ASN1_TIME structures. Fix ca
12484 utility to correctly initialize revocation date of CRLs.
12488 * New option SSL_OP_CIPHER_SERVER_PREFERENCE allows the server to override
12489 the clients preferred ciphersuites and rather use its own preferences.
12490 Should help to work around M$ SGC (Server Gated Cryptography) bug in
12491 Internet Explorer by ensuring unchanged hash method during stepup.
12492 (Also replaces the broken/deactivated SSL_OP_NON_EXPORT_FIRST option.)
12496 * Make mkdef.pl recognise all DECLARE_ASN1 macros, change rijndael
12497 to aes and add a new 'exist' option to print out symbols that don't
12502 * Additional options to ocsp utility to allow flags to be set and
12503 additional certificates supplied.
12507 * Add the option -VAfile to 'openssl ocsp', so the user can give the
12508 OCSP client a number of certificate to only verify the response
12513 * Update Rijndael code to version 3.0 and change EVP AES ciphers to
12514 handle the new API. Currently only ECB, CBC modes supported. Add new
12517 Add TLS AES ciphersuites as described in RFC3268, "Advanced
12518 Encryption Standard (AES) Ciphersuites for Transport Layer
12519 Security (TLS)". (In beta versions of OpenSSL 0.9.7, these were
12520 not enabled by default and were not part of the "ALL" ciphersuite
12521 alias because they were not yet official; they could be
12522 explicitly requested by specifying the "AESdraft" ciphersuite
12523 group alias. In the final release of OpenSSL 0.9.7, the group
12524 alias is called "AES" and is part of "ALL".)
12526 *Ben Laurie, Steve Henson, Bodo Moeller*
12528 * New function OCSP_copy_nonce() to copy nonce value (if present) from
12529 request to response.
12533 * Functions for OCSP responders. OCSP_request_onereq_count(),
12534 OCSP_request_onereq_get0(), OCSP_onereq_get0_id() and OCSP_id_get0_info()
12535 extract information from a certificate request. OCSP_response_create()
12536 creates a response and optionally adds a basic response structure.
12537 OCSP_basic_add1_status() adds a complete single response to a basic
12538 response and returns the OCSP_SINGLERESP structure just added (to allow
12539 extensions to be included for example). OCSP_basic_add1_cert() adds a
12540 certificate to a basic response and OCSP_basic_sign() signs a basic
12541 response with various flags. New helper functions ASN1_TIME_check()
12542 (checks validity of ASN1_TIME structure) and ASN1_TIME_to_generalizedtime()
12543 (converts ASN1_TIME to GeneralizedTime).
12547 * Various new functions. EVP_Digest() combines EVP_Digest{Init,Update,Final}()
12548 in a single operation. X509_get0_pubkey_bitstr() extracts the public_key
12549 structure from a certificate. X509_pubkey_digest() digests the public_key
12550 contents: this is used in various key identifiers.
12554 * Make sk_sort() tolerate a NULL argument.
12556 *Steve Henson reported by Massimiliano Pala <madwolf@comune.modena.it>*
12558 * New OCSP verify flag OCSP_TRUSTOTHER. When set the "other" certificates
12559 passed by the function are trusted implicitly. If any of them signed the
12560 response then it is assumed to be valid and is not verified.
12564 * In PKCS7_set_type() initialise content_type in PKCS7_ENC_CONTENT
12565 to data. This was previously part of the PKCS7 ASN1 code. This
12566 was causing problems with OpenSSL created PKCS#12 and PKCS#7 structures.
12567 *Steve Henson, reported by Kenneth R. Robinette
12568 <support@securenetterm.com>*
12570 * Add CRYPTO_push_info() and CRYPTO_pop_info() calls to new ASN1
12571 routines: without these tracing memory leaks is very painful.
12572 Fix leaks in PKCS12 and PKCS7 routines.
12576 * Make X509_time_adj() cope with the new behaviour of ASN1_TIME_new().
12577 Previously it initialised the 'type' argument to V_ASN1_UTCTIME which
12578 effectively meant GeneralizedTime would never be used. Now it
12579 is initialised to -1 but X509_time_adj() now has to check the value
12580 and use ASN1_TIME_set() if the value is not V_ASN1_UTCTIME or
12581 V_ASN1_GENERALIZEDTIME, without this it always uses GeneralizedTime.
12582 *Steve Henson, reported by Kenneth R. Robinette
12583 <support@securenetterm.com>*
12585 * Fixes to BN_to_ASN1_INTEGER when bn is zero. This would previously
12586 result in a zero length in the ASN1_INTEGER structure which was
12587 not consistent with the structure when d2i_ASN1_INTEGER() was used
12588 and would cause ASN1_INTEGER_cmp() to fail. Enhance s2i_ASN1_INTEGER()
12589 to cope with hex and negative integers. Fix bug in i2a_ASN1_INTEGER()
12590 where it did not print out a minus for negative ASN1_INTEGER.
12594 * Add summary printout to ocsp utility. The various functions which
12595 convert status values to strings have been renamed to:
12596 OCSP_response_status_str(), OCSP_cert_status_str() and
12597 OCSP_crl_reason_str() and are no longer static. New options
12598 to verify nonce values and to disable verification. OCSP response
12599 printout format cleaned up.
12603 * Add additional OCSP certificate checks. These are those specified
12604 in RFC2560. This consists of two separate checks: the CA of the
12605 certificate being checked must either be the OCSP signer certificate
12606 or the issuer of the OCSP signer certificate. In the latter case the
12607 OCSP signer certificate must contain the OCSP signing extended key
12608 usage. This check is performed by attempting to match the OCSP
12609 signer or the OCSP signer CA to the issuerNameHash and issuerKeyHash
12610 in the OCSP_CERTID structures of the response.
12614 * Initial OCSP certificate verification added to OCSP_basic_verify()
12615 and related routines. This uses the standard OpenSSL certificate
12616 verify routines to perform initial checks (just CA validity) and
12617 to obtain the certificate chain. Then additional checks will be
12618 performed on the chain. Currently the root CA is checked to see
12619 if it is explicitly trusted for OCSP signing. This is used to set
12620 a root CA as a global signing root: that is any certificate that
12621 chains to that CA is an acceptable OCSP signing certificate.
12625 * New '-extfile ...' option to 'openssl ca' for reading X.509v3
12626 extensions from a separate configuration file.
12627 As when reading extensions from the main configuration file,
12628 the '-extensions ...' option may be used for specifying the
12631 *Massimiliano Pala <madwolf@comune.modena.it>*
12633 * New OCSP utility. Allows OCSP requests to be generated or
12634 read. The request can be sent to a responder and the output
12635 parsed, outputted or printed in text form. Not complete yet:
12636 still needs to check the OCSP response validity.
12640 * New subcommands for 'openssl ca':
12641 `openssl ca -status <serial>` prints the status of the cert with
12642 the given serial number (according to the index file).
12643 `openssl ca -updatedb` updates the expiry status of certificates
12646 *Massimiliano Pala <madwolf@comune.modena.it>*
12648 * New '-newreq-nodes' command option to CA.pl. This is like
12649 '-newreq', but calls 'openssl req' with the '-nodes' option
12650 so that the resulting key is not encrypted.
12652 *Damien Miller <djm@mindrot.org>*
12654 * New configuration for the GNU Hurd.
12656 *Jonathan Bartlett <johnnyb@wolfram.com> via Richard Levitte*
12658 * Initial code to implement OCSP basic response verify. This
12659 is currently incomplete. Currently just finds the signer's
12660 certificate and verifies the signature on the response.
12664 * New SSLeay_version code SSLEAY_DIR to determine the compiled-in
12665 value of OPENSSLDIR. This is available via the new '-d' option
12666 to 'openssl version', and is also included in 'openssl version -a'.
12670 * Allowing defining memory allocation callbacks that will be given
12671 file name and line number information in additional arguments
12672 (a `const char*` and an int). The basic functionality remains, as
12673 well as the original possibility to just replace malloc(),
12674 realloc() and free() by functions that do not know about these
12675 additional arguments. To register and find out the current
12676 settings for extended allocation functions, the following
12677 functions are provided:
12679 CRYPTO_set_mem_ex_functions
12680 CRYPTO_set_locked_mem_ex_functions
12681 CRYPTO_get_mem_ex_functions
12682 CRYPTO_get_locked_mem_ex_functions
12684 These work the same way as CRYPTO_set_mem_functions and friends.
12685 `CRYPTO_get_[locked_]mem_functions` now writes 0 where such an
12686 extended allocation function is enabled.
12687 Similarly, `CRYPTO_get_[locked_]mem_ex_functions` writes 0 where
12688 a conventional allocation function is enabled.
12690 *Richard Levitte, Bodo Moeller*
12692 * Finish off removing the remaining LHASH function pointer casts.
12693 There should no longer be any prototype-casting required when using
12694 the LHASH abstraction, and any casts that remain are "bugs". See
12695 the callback types and macros at the head of lhash.h for details
12696 (and "OBJ_cleanup" in crypto/objects/obj_dat.c as an example).
12700 * Add automatic query of EGD sockets in RAND_poll() for the unix variant.
12701 If /dev/[u]random devices are not available or do not return enough
12702 entropy, EGD style sockets (served by EGD or PRNGD) will automatically
12704 The locations /var/run/egd-pool, /dev/egd-pool, /etc/egd-pool, and
12705 /etc/entropy will be queried once each in this sequence, querying stops
12706 when enough entropy was collected without querying more sockets.
12710 * Change the Unix RAND_poll() variant to be able to poll several
12711 random devices, as specified by DEVRANDOM, until a sufficient amount
12712 of data has been collected. We spend at most 10 ms on each file
12713 (select timeout) and read in non-blocking mode. DEVRANDOM now
12714 defaults to the list "/dev/urandom", "/dev/random", "/dev/srandom"
12715 (previously it was just the string "/dev/urandom"), so on typical
12716 platforms the 10 ms delay will never occur.
12717 Also separate out the Unix variant to its own file, rand_unix.c.
12718 For VMS, there's a currently-empty rand_vms.c.
12722 * Move OCSP client related routines to ocsp_cl.c. These
12723 provide utility functions which an application needing
12724 to issue a request to an OCSP responder and analyse the
12725 response will typically need: as opposed to those which an
12726 OCSP responder itself would need which will be added later.
12728 OCSP_request_sign() signs an OCSP request with an API similar
12729 to PKCS7_sign(). OCSP_response_status() returns status of OCSP
12730 response. OCSP_response_get1_basic() extracts basic response
12731 from response. OCSP_resp_find_status(): finds and extracts status
12732 information from an OCSP_CERTID structure (which will be created
12733 when the request structure is built). These are built from lower
12734 level functions which work on OCSP_SINGLERESP structures but
12735 won't normally be used unless the application wishes to examine
12736 extensions in the OCSP response for example.
12738 Replace nonce routines with a pair of functions.
12739 OCSP_request_add1_nonce() adds a nonce value and optionally
12740 generates a random value. OCSP_check_nonce() checks the
12741 validity of the nonce in an OCSP response.
12745 * Change function OCSP_request_add() to OCSP_request_add0_id().
12746 This doesn't copy the supplied OCSP_CERTID and avoids the
12747 need to free up the newly created id. Change return type
12748 to OCSP_ONEREQ to return the internal OCSP_ONEREQ structure.
12749 This can then be used to add extensions to the request.
12750 Deleted OCSP_request_new(), since most of its functionality
12751 is now in OCSP_REQUEST_new() (and the case insensitive name
12752 clash) apart from the ability to set the request name which
12753 will be added elsewhere.
12757 * Update OCSP API. Remove obsolete extensions argument from
12758 various functions. Extensions are now handled using the new
12759 OCSP extension code. New simple OCSP HTTP function which
12760 can be used to send requests and parse the response.
12764 * Fix the PKCS#7 (S/MIME) code to work with new ASN1. Two new
12765 ASN1_ITEM structures help with sign and verify. PKCS7_ATTR_SIGN
12766 uses the special reorder version of SET OF to sort the attributes
12767 and reorder them to match the encoded order. This resolves a long
12768 standing problem: a verify on a PKCS7 structure just after signing
12769 it used to fail because the attribute order did not match the
12770 encoded order. PKCS7_ATTR_VERIFY does not reorder the attributes:
12771 it uses the received order. This is necessary to tolerate some broken
12772 software that does not order SET OF. This is handled by encoding
12773 as a SEQUENCE OF but using implicit tagging (with UNIVERSAL class)
12774 to produce the required SET OF.
12778 * Have mk1mf.pl generate the macros OPENSSL_BUILD_SHLIBCRYPTO and
12779 OPENSSL_BUILD_SHLIBSSL and use them appropriately in the header
12780 files to get correct declarations of the ASN.1 item variables.
12784 * Rewrite of PKCS#12 code to use new ASN1 functionality. Replace many
12785 PKCS#12 macros with real functions. Fix two unrelated ASN1 bugs:
12786 asn1_check_tlen() would sometimes attempt to use 'ctx' when it was
12787 NULL and ASN1_TYPE was not dereferenced properly in asn1_ex_c2i().
12788 New ASN1 macro: DECLARE_ASN1_ITEM() which just declares the relevant
12789 ASN1_ITEM and no wrapper functions.
12793 * New functions or ASN1_item_d2i_fp() and ASN1_item_d2i_bio(). These
12794 replace the old function pointer based I/O routines. Change most of
12795 the `*_d2i_bio()` and `*_d2i_fp()` functions to use these.
12799 * Enhance mkdef.pl to be more accepting about spacing in C preprocessor
12800 lines, recognize more "algorithms" that can be deselected, and make
12801 it complain about algorithm deselection that isn't recognised.
12805 * New ASN1 functions to handle dup, sign, verify, digest, pack and
12806 unpack operations in terms of ASN1_ITEM. Modify existing wrappers
12807 to use new functions. Add NO_ASN1_OLD which can be set to remove
12808 some old style ASN1 functions: this can be used to determine if old
12809 code will still work when these eventually go away.
12813 * New extension functions for OCSP structures, these follow the
12814 same conventions as certificates and CRLs.
12818 * New function X509V3_add1_i2d(). This automatically encodes and
12819 adds an extension. Its behaviour can be customised with various
12820 flags to append, replace or delete. Various wrappers added for
12821 certificates and CRLs.
12825 * Fix to avoid calling the underlying ASN1 print routine when
12826 an extension cannot be parsed. Correct a typo in the
12827 OCSP_SERVICELOC extension. Tidy up print OCSP format.
12831 * Make mkdef.pl parse some of the ASN1 macros and add appropriate
12832 entries for variables.
12836 * Add functionality to apps/openssl.c for detecting locking
12837 problems: As the program is single-threaded, all we have
12838 to do is register a locking callback using an array for
12839 storing which locks are currently held by the program.
12843 * Use a lock around the call to CRYPTO_get_ex_new_index() in
12844 SSL_get_ex_data_X509_STORE_idx(), which is used in
12845 ssl_verify_cert_chain() and thus can be called at any time
12846 during TLS/SSL handshakes so that thread-safety is essential.
12847 Unfortunately, the ex_data design is not at all suited
12848 for multi-threaded use, so it probably should be abolished.
12852 * Added Broadcom "ubsec" ENGINE to OpenSSL.
12854 *Broadcom, tweaked and integrated by Geoff Thorpe*
12856 * Move common extension printing code to new function
12857 X509V3_print_extensions(). Reorganise OCSP print routines and
12858 implement some needed OCSP ASN1 functions. Add OCSP extensions.
12862 * New function X509_signature_print() to remove duplication in some
12867 * Add a special meaning when SET OF and SEQUENCE OF flags are both
12868 set (this was treated exactly the same as SET OF previously). This
12869 is used to reorder the STACK representing the structure to match the
12870 encoding. This will be used to get round a problem where a PKCS7
12871 structure which was signed could not be verified because the STACK
12872 order did not reflect the encoded order.
12876 * Reimplement the OCSP ASN1 module using the new code.
12880 * Update the X509V3 code to permit the use of an ASN1_ITEM structure
12881 for its ASN1 operations. The old style function pointers still exist
12882 for now but they will eventually go away.
12886 * Merge in replacement ASN1 code from the ASN1 branch. This almost
12887 completely replaces the old ASN1 functionality with a table driven
12888 encoder and decoder which interprets an ASN1_ITEM structure describing
12889 the ASN1 module. Compatibility with the existing ASN1 API (i2d,d2i) is
12890 largely maintained. Almost all of the old asn1_mac.h macro based ASN1
12891 has also been converted to the new form.
12895 * Change BN_mod_exp_recp so that negative moduli are tolerated
12896 (the sign is ignored). Similarly, ignore the sign in BN_MONT_CTX_set
12897 so that BN_mod_exp_mont and BN_mod_exp_mont_word work
12898 for negative moduli.
12902 * Fix BN_uadd and BN_usub: Always return non-negative results instead
12903 of not touching the result's sign bit.
12907 * BN_div bugfix: If the result is 0, the sign (res->neg) must not be
12912 * Changed the LHASH code to use prototypes for callbacks, and created
12913 macros to declare and implement thin (optionally static) functions
12914 that provide type-safety and avoid function pointer casting for the
12915 type-specific callbacks.
12919 * Added Kerberos Cipher Suites to be used with TLS, as written in
12921 *Veers Staats <staatsvr@asc.hpc.mil>,
12922 Jeffrey Altman <jaltman@columbia.edu>, via Richard Levitte*
12924 * Reformat the FAQ so the different questions and answers can be divided
12925 in sections depending on the subject.
12929 * Have the zlib compression code load ZLIB.DLL dynamically under
12934 * New function BN_mod_sqrt for computing square roots modulo a prime
12935 (using the probabilistic Tonelli-Shanks algorithm unless
12936 p == 3 (mod 4) or p == 5 (mod 8), which are cases that can
12937 be handled deterministically).
12939 *Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller*
12941 * Make BN_mod_inverse faster by explicitly handling small quotients
12942 in the Euclid loop. (Speed gain about 20% for small moduli [256 or
12943 512 bits], about 30% for larger ones [1024 or 2048 bits].)
12947 * New function BN_kronecker.
12951 * Fix BN_gcd so that it works on negative inputs; the result is
12952 positive unless both parameters are zero.
12953 Previously something reasonably close to an infinite loop was
12954 possible because numbers could be growing instead of shrinking
12955 in the implementation of Euclid's algorithm.
12959 * Fix BN_is_word() and BN_is_one() macros to take into account the
12960 sign of the number in question.
12962 Fix BN_is_word(a,w) to work correctly for w == 0.
12964 The old BN_is_word(a,w) macro is now called BN_abs_is_word(a,w)
12965 because its test if the absolute value of 'a' equals 'w'.
12966 Note that BN_abs_is_word does *not* handle w == 0 reliably;
12967 it exists mostly for use in the implementations of BN_is_zero(),
12968 BN_is_one(), and BN_is_word().
12972 * New function BN_swap.
12976 * Use BN_nnmod instead of BN_mod in crypto/bn/bn_exp.c so that
12977 the exponentiation functions are more likely to produce reasonable
12978 results on negative inputs.
12982 * Change BN_mod_mul so that the result is always non-negative.
12983 Previously, it could be negative if one of the factors was negative;
12984 I don't think anyone really wanted that behaviour.
12988 * Move `BN_mod_...` functions into new file `crypto/bn/bn_mod.c`
12989 (except for exponentiation, which stays in `crypto/bn/bn_exp.c`,
12990 and `BN_mod_mul_reciprocal`, which stays in `crypto/bn/bn_recp.c`)
12991 and add new functions:
13000 BN_mod_lshift1_quick
13002 BN_mod_lshift_quick
13004 These functions always generate non-negative results.
13006 `BN_nnmod` otherwise is `like BN_mod` (if `BN_mod` computes a remainder `r`
13007 such that `|m| < r < 0`, `BN_nnmod` will output `rem + |m|` instead).
13009 `BN_mod_XXX_quick(r, a, [b,] m)` generates the same result as
13010 `BN_mod_XXX(r, a, [b,] m, ctx)`, but requires that `a` [and `b`]
13011 be reduced modulo `m`.
13013 *Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller*
13016 The following entry accidentally appeared in the CHANGES file
13017 distributed with OpenSSL 0.9.7. The modifications described in
13018 it do *not* apply to OpenSSL 0.9.7.
13020 * Remove a few calls to bn_wexpand() in BN_sqr() (the one in there
13021 was actually never needed) and in BN_mul(). The removal in BN_mul()
13022 required a small change in bn_mul_part_recursive() and the addition
13023 of the functions bn_cmp_part_words(), bn_sub_part_words() and
13024 bn_add_part_words(), which do the same thing as bn_cmp_words(),
13025 bn_sub_words() and bn_add_words() except they take arrays with
13031 * In 'openssl passwd', verify passwords read from the terminal
13032 unless the '-salt' option is used (which usually means that
13033 verification would just waste user's time since the resulting
13034 hash is going to be compared with some given password hash)
13035 or the new '-noverify' option is used.
13037 This is an incompatible change, but it does not affect
13038 non-interactive use of 'openssl passwd' (passwords on the command
13039 line, '-stdin' option, '-in ...' option) and thus should not
13040 cause any problems.
13044 * Remove all references to RSAref, since there's no more need for it.
13048 * Make DSO load along a path given through an environment variable
13049 (SHLIB_PATH) with shl_load().
13053 * Constify the ENGINE code as a result of BIGNUM constification.
13054 Also constify the RSA code and most things related to it. In a
13055 few places, most notable in the depth of the ASN.1 code, ugly
13056 casts back to non-const were required (to be solved at a later
13061 * Make it so the openssl application has all engines loaded by default.
13065 * Constify the BIGNUM routines a little more.
13069 * Add the following functions:
13071 ENGINE_load_cswift()
13073 ENGINE_load_atalla()
13074 ENGINE_load_nuron()
13075 ENGINE_load_builtin_engines()
13077 That way, an application can itself choose if external engines that
13078 are built-in in OpenSSL shall ever be used or not. The benefit is
13079 that applications won't have to be linked with libdl or other dso
13080 libraries unless it's really needed.
13082 Changed 'openssl engine' to load all engines on demand.
13083 Changed the engine header files to avoid the duplication of some
13084 declarations (they differed!).
13088 * 'openssl engine' can now list capabilities.
13092 * Better error reporting in 'openssl engine'.
13096 * Never call load_dh_param(NULL) in s_server.
13100 * Add engine application. It can currently list engines by name and
13101 identity, and test if they are actually available.
13105 * Improve RPM specification file by forcing symbolic linking and making
13106 sure the installed documentation is also owned by root.root.
13108 *Damien Miller <djm@mindrot.org>*
13110 * Give the OpenSSL applications more possibilities to make use of
13111 keys (public as well as private) handled by engines.
13115 * Add OCSP code that comes from CertCo.
13119 * Add VMS support for the Rijndael code.
13123 * Added untested support for Nuron crypto accelerator.
13127 * Add support for external cryptographic devices. This code was
13128 previously distributed separately as the "engine" branch.
13130 *Geoff Thorpe, Richard Levitte*
13132 * Rework the filename-translation in the DSO code. It is now possible to
13133 have far greater control over how a "name" is turned into a filename
13134 depending on the operating environment and any oddities about the
13135 different shared library filenames on each system.
13139 * Support threads on FreeBSD-elf in Configure.
13143 * Fix for SHA1 assembly problem with MASM: it produces
13144 warnings about corrupt line number information when assembling
13145 with debugging information. This is caused by the overlapping
13148 *Bernd Matthes <mainbug@celocom.de>, Steve Henson*
13151 NCONF_get_number() has no error checking at all. As a replacement,
13152 NCONF_get_number_e() is defined (`_e` for "error checking") and is
13153 promoted strongly. The old NCONF_get_number is kept around for
13154 binary backward compatibility.
13155 Make it possible for methods to load from something other than a BIO,
13156 by providing a function pointer that is given a name instead of a BIO.
13157 For example, this could be used to load configuration data from an
13162 * Fix for non blocking accept BIOs. Added new I/O special reason
13163 BIO_RR_ACCEPT to cover this case. Previously use of accept BIOs
13164 with non blocking I/O was not possible because no retry code was
13165 implemented. Also added new SSL code SSL_WANT_ACCEPT to cover
13170 * Added the beginnings of Rijndael support.
13174 * Fix for bug in DirectoryString mask setting. Add support for
13175 X509_NAME_print_ex() in 'req' and X509_print_ex() function
13176 to allow certificate printing to more controllable, additional
13177 'certopt' option to 'x509' to allow new printing options to be
13182 * Clean old EAY MD5 hack from e_os.h.
13186 ### Changes between 0.9.6l and 0.9.6m [17 Mar 2004]
13188 * Fix null-pointer assignment in do_change_cipher_spec() revealed
13189 by using the Codenomicon TLS Test Tool [CVE-2004-0079][]
13191 *Joe Orton, Steve Henson*
13193 ### Changes between 0.9.6k and 0.9.6l [04 Nov 2003]
13195 * Fix additional bug revealed by the NISCC test suite:
13197 Stop bug triggering large recursion when presented with
13198 certain ASN.1 tags [CVE-2003-0851][]
13202 ### Changes between 0.9.6j and 0.9.6k [30 Sep 2003]
13204 * Fix various bugs revealed by running the NISCC test suite:
13206 Stop out of bounds reads in the ASN1 code when presented with
13207 invalid tags (CVE-2003-0543 and CVE-2003-0544).
13209 If verify callback ignores invalid public key errors don't try to check
13210 certificate signature with the NULL public key.
13214 * In ssl3_accept() (ssl/s3_srvr.c) only accept a client certificate
13215 if the server requested one: as stated in TLS 1.0 and SSL 3.0
13220 * In ssl3_get_client_hello() (ssl/s3_srvr.c), tolerate additional
13221 extra data after the compression methods not only for TLS 1.0
13222 but also for SSL 3.0 (as required by the specification).
13224 *Bodo Moeller; problem pointed out by Matthias Loepfe*
13226 * Change X509_certificate_type() to mark the key as exported/exportable
13227 when it's 512 *bits* long, not 512 bytes.
13231 ### Changes between 0.9.6i and 0.9.6j [10 Apr 2003]
13233 * Countermeasure against the Klima-Pokorny-Rosa extension of
13234 Bleichbacher's attack on PKCS #1 v1.5 padding: treat
13235 a protocol version number mismatch like a decryption error
13236 in ssl3_get_client_key_exchange (ssl/s3_srvr.c).
13240 * Turn on RSA blinding by default in the default implementation
13241 to avoid a timing attack. Applications that don't want it can call
13242 RSA_blinding_off() or use the new flag RSA_FLAG_NO_BLINDING.
13243 They would be ill-advised to do so in most cases.
13245 *Ben Laurie, Steve Henson, Geoff Thorpe, Bodo Moeller*
13247 * Change RSA blinding code so that it works when the PRNG is not
13248 seeded (in this case, the secret RSA exponent is abused as
13249 an unpredictable seed -- if it is not unpredictable, there
13250 is no point in blinding anyway). Make RSA blinding thread-safe
13251 by remembering the creator's thread ID in rsa->blinding and
13252 having all other threads use local one-time blinding factors
13253 (this requires more computation than sharing rsa->blinding, but
13254 avoids excessive locking; and if an RSA object is not shared
13255 between threads, blinding will still be very fast).
13259 ### Changes between 0.9.6h and 0.9.6i [19 Feb 2003]
13261 * In ssl3_get_record (ssl/s3_pkt.c), minimize information leaked
13262 via timing by performing a MAC computation even if incorrect
13263 block cipher padding has been found. This is a countermeasure
13264 against active attacks where the attacker has to distinguish
13265 between bad padding and a MAC verification error. [CVE-2003-0078][]
13267 *Bodo Moeller; problem pointed out by Brice Canvel (EPFL),
13268 Alain Hiltgen (UBS), Serge Vaudenay (EPFL), and
13269 Martin Vuagnoux (EPFL, Ilion)*
13271 ### Changes between 0.9.6g and 0.9.6h [5 Dec 2002]
13273 * New function OPENSSL_cleanse(), which is used to cleanse a section of
13274 memory from its contents. This is done with a counter that will
13275 place alternating values in each byte. This can be used to solve
13276 two issues: 1) the removal of calls to memset() by highly optimizing
13277 compilers, and 2) cleansing with other values than 0, since those can
13278 be read through on certain media, for example a swap space on disk.
13282 * Bugfix: client side session caching did not work with external caching,
13283 because the session->cipher setting was not restored when reloading
13284 from the external cache. This problem was masked, when
13285 SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG (part of SSL_OP_ALL) was set.
13286 (Found by Steve Haslam <steve@araqnid.ddts.net>.)
13290 * Fix client_certificate (ssl/s2_clnt.c): The permissible total
13291 length of the REQUEST-CERTIFICATE message is 18 .. 34, not 17 .. 33.
13293 *Zeev Lieber <zeev-l@yahoo.com>*
13295 * Undo an undocumented change introduced in 0.9.6e which caused
13296 repeated calls to OpenSSL_add_all_ciphers() and
13297 OpenSSL_add_all_digests() to be ignored, even after calling
13302 * Change the default configuration reader to deal with last line not
13303 being properly terminated.
13307 * Change X509_NAME_cmp() so it applies the special rules on handling
13308 DN values that are of type PrintableString, as well as RDNs of type
13309 emailAddress where the value has the type ia5String.
13311 *stefank@valicert.com via Richard Levitte*
13313 * Add a SSL_SESS_CACHE_NO_INTERNAL_STORE flag to take over half
13314 the job SSL_SESS_CACHE_NO_INTERNAL_LOOKUP was inconsistently
13315 doing, define a new flag (SSL_SESS_CACHE_NO_INTERNAL) to be
13316 the bitwise-OR of the two for use by the majority of applications
13317 wanting this behaviour, and update the docs. The documented
13318 behaviour and actual behaviour were inconsistent and had been
13319 changing anyway, so this is more a bug-fix than a behavioural
13322 *Geoff Thorpe, diagnosed by Nadav Har'El*
13324 * Don't impose a 16-byte length minimum on session IDs in ssl/s3_clnt.c
13325 (the SSL 3.0 and TLS 1.0 specifications allow any length up to 32 bytes).
13329 * Fix initialization code race conditions in
13330 SSLv23_method(), SSLv23_client_method(), SSLv23_server_method(),
13331 SSLv2_method(), SSLv2_client_method(), SSLv2_server_method(),
13332 SSLv3_method(), SSLv3_client_method(), SSLv3_server_method(),
13333 TLSv1_method(), TLSv1_client_method(), TLSv1_server_method(),
13334 ssl2_get_cipher_by_char(),
13335 ssl3_get_cipher_by_char().
13337 *Patrick McCormick <patrick@tellme.com>, Bodo Moeller*
13339 * Reorder cleanup sequence in SSL_CTX_free(): only remove the ex_data after
13340 the cached sessions are flushed, as the remove_cb() might use ex_data
13341 contents. Bug found by Sam Varshavchik <mrsam@courier-mta.com>
13342 (see [openssl.org #212]).
13344 *Geoff Thorpe, Lutz Jaenicke*
13346 * Fix typo in OBJ_txt2obj which incorrectly passed the content
13347 length, instead of the encoding length to d2i_ASN1_OBJECT.
13351 ### Changes between 0.9.6f and 0.9.6g [9 Aug 2002]
13353 * [In 0.9.6g-engine release:]
13354 Fix crypto/engine/vendor_defns/cswift.h for WIN32 (use `_stdcall`).
13356 *Lynn Gazis <lgazis@rainbow.com>*
13358 ### Changes between 0.9.6e and 0.9.6f [8 Aug 2002]
13360 * Fix ASN1 checks. Check for overflow by comparing with LONG_MAX
13361 and get fix the header length calculation.
13362 *Florian Weimer <Weimer@CERT.Uni-Stuttgart.DE>,
13363 Alon Kantor <alonk@checkpoint.com> (and others), Steve Henson*
13365 * Use proper error handling instead of 'assertions' in buffer
13366 overflow checks added in 0.9.6e. This prevents DoS (the
13367 assertions could call abort()).
13369 *Arne Ansper <arne@ats.cyber.ee>, Bodo Moeller*
13371 ### Changes between 0.9.6d and 0.9.6e [30 Jul 2002]
13373 * Add various sanity checks to asn1_get_length() to reject
13374 the ASN1 length bytes if they exceed sizeof(long), will appear
13375 negative or the content length exceeds the length of the
13378 *Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>*
13380 * Fix cipher selection routines: ciphers without encryption had no flags
13381 for the cipher strength set and where therefore not handled correctly
13382 by the selection routines (PR #130).
13386 * Fix EVP_dsa_sha macro.
13391 SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS
13392 for disabling the SSL 3.0/TLS 1.0 CBC vulnerability countermeasure
13393 that was added in OpenSSL 0.9.6d.
13395 As the countermeasure turned out to be incompatible with some
13396 broken SSL implementations, the new option is part of SSL_OP_ALL.
13397 SSL_OP_ALL is usually employed when compatibility with weird SSL
13398 implementations is desired (e.g. '-bugs' option to 's_client' and
13399 's_server'), so the new option is automatically set in many
13404 * Changes in security patch:
13406 Changes marked "(CHATS)" were sponsored by the Defense Advanced
13407 Research Projects Agency (DARPA) and Air Force Research Laboratory,
13408 Air Force Materiel Command, USAF, under agreement number
13411 * Add various sanity checks to asn1_get_length() to reject
13412 the ASN1 length bytes if they exceed sizeof(long), will appear
13413 negative or the content length exceeds the length of the
13414 supplied buffer. [CVE-2002-0659][]
13416 *Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>*
13418 * Assertions for various potential buffer overflows, not known to
13419 happen in practice.
13421 *Ben Laurie (CHATS)*
13423 * Various temporary buffers to hold ASCII versions of integers were
13424 too small for 64 bit platforms. [CVE-2002-0655][]
13425 *Matthew Byng-Maddick <mbm@aldigital.co.uk> and Ben Laurie (CHATS)>*
13427 * Remote buffer overflow in SSL3 protocol - an attacker could
13428 supply an oversized session ID to a client. [CVE-2002-0656][]
13430 *Ben Laurie (CHATS)*
13432 * Remote buffer overflow in SSL2 protocol - an attacker could
13433 supply an oversized client master key. [CVE-2002-0656][]
13435 *Ben Laurie (CHATS)*
13437 ### Changes between 0.9.6c and 0.9.6d [9 May 2002]
13439 * Fix crypto/asn1/a_sign.c so that 'parameters' is omitted (not
13440 encoded as NULL) with id-dsa-with-sha1.
13442 *Nils Larsch <nla@trustcenter.de>; problem pointed out by Bodo Moeller*
13444 * Check various `X509_...()` return values in apps/req.c.
13446 *Nils Larsch <nla@trustcenter.de>*
13448 * Fix BASE64 decode (EVP_DecodeUpdate) for data with CR/LF ended lines:
13449 an end-of-file condition would erroneously be flagged, when the CRLF
13450 was just at the end of a processed block. The bug was discovered when
13451 processing data through a buffering memory BIO handing the data to a
13452 BASE64-decoding BIO. Bug fund and patch submitted by Pavel Tsekov
13453 <ptsekov@syntrex.com> and Nedelcho Stanev.
13457 * Implement a countermeasure against a vulnerability recently found
13458 in CBC ciphersuites in SSL 3.0/TLS 1.0: Send an empty fragment
13459 before application data chunks to avoid the use of known IVs
13460 with data potentially chosen by the attacker.
13464 * Fix length checks in ssl3_get_client_hello().
13468 * TLS/SSL library bugfix: use s->s3->in_read_app_data differently
13469 to prevent ssl3_read_internal() from incorrectly assuming that
13470 ssl3_read_bytes() found application data while handshake
13471 processing was enabled when in fact s->s3->in_read_app_data was
13472 merely automatically cleared during the initial handshake.
13474 *Bodo Moeller; problem pointed out by Arne Ansper <arne@ats.cyber.ee>*
13476 * Fix object definitions for Private and Enterprise: they were not
13477 recognized in their shortname (=lowercase) representation. Extend
13478 obj_dat.pl to issue an error when using undefined keywords instead
13479 of silently ignoring the problem (Svenning Sorensen
13480 <sss@sss.dnsalias.net>).
13484 * Fix DH_generate_parameters() so that it works for 'non-standard'
13485 generators, i.e. generators other than 2 and 5. (Previously, the
13486 code did not properly initialise the 'add' and 'rem' values to
13487 BN_generate_prime().)
13489 In the new general case, we do not insist that 'generator' is
13490 actually a primitive root: This requirement is rather pointless;
13491 a generator of the order-q subgroup is just as good, if not
13496 * Map new X509 verification errors to alerts. Discovered and submitted by
13497 Tom Wu <tom@arcot.com>.
13501 * Fix ssl3_pending() (ssl/s3_lib.c) to prevent SSL_pending() from
13502 returning non-zero before the data has been completely received
13503 when using non-blocking I/O.
13505 *Bodo Moeller; problem pointed out by John Hughes*
13507 * Some of the ciphers missed the strength entry (SSL_LOW etc).
13509 *Ben Laurie, Lutz Jaenicke*
13511 * Fix bug in SSL_clear(): bad sessions were not removed (found by
13512 Yoram Zahavi <YoramZ@gilian.com>).
13516 * Add information about CygWin 1.3 and on, and preserve proper
13517 configuration for the versions before that.
13519 *Corinna Vinschen <vinschen@redhat.com> and Richard Levitte*
13521 * Make removal from session cache (SSL_CTX_remove_session()) more robust:
13522 check whether we deal with a copy of a session and do not delete from
13523 the cache in this case. Problem reported by "Izhar Shoshani Levi"
13524 <izhar@checkpoint.com>.
13528 * Do not store session data into the internal session cache, if it
13529 is never intended to be looked up (SSL_SESS_CACHE_NO_INTERNAL_LOOKUP
13530 flag is set). Proposed by Aslam <aslam@funk.com>.
13534 * Have ASN1_BIT_STRING_set_bit() really clear a bit when the requested
13539 * [In 0.9.6d-engine release:]
13540 Fix a crashbug and a logic bug in hwcrhk_load_pubkey().
13542 *Toomas Kiisk <vix@cyber.ee> via Richard Levitte*
13544 * Add the configuration target linux-s390x.
13546 *Neale Ferguson <Neale.Ferguson@SoftwareAG-USA.com> via Richard Levitte*
13548 * The earlier bugfix for the SSL3_ST_SW_HELLO_REQ_C case of
13549 ssl3_accept (ssl/s3_srvr.c) incorrectly used a local flag
13550 variable as an indication that a ClientHello message has been
13551 received. As the flag value will be lost between multiple
13552 invocations of ssl3_accept when using non-blocking I/O, the
13553 function may not be aware that a handshake has actually taken
13554 place, thus preventing a new session from being added to the
13557 To avoid this problem, we now set s->new_session to 2 instead of
13558 using a local variable.
13560 *Lutz Jaenicke, Bodo Moeller*
13562 * Bugfix: Return -1 from ssl3_get_server_done (ssl3/s3_clnt.c)
13563 if the SSL_R_LENGTH_MISMATCH error is detected.
13565 *Geoff Thorpe, Bodo Moeller*
13567 * New 'shared_ldflag' column in Configure platform table.
13571 * Fix EVP_CIPHER_mode macro.
13573 *"Dan S. Camper" <dan@bti.net>*
13575 * Fix ssl3_read_bytes (ssl/s3_pkt.c): To ignore messages of unknown
13576 type, we must throw them away by setting rr->length to 0.
13578 *D P Chang <dpc@qualys.com>*
13580 ### Changes between 0.9.6b and 0.9.6c [21 dec 2001]
13582 * Fix BN_rand_range bug pointed out by Dominikus Scherkl
13583 <Dominikus.Scherkl@biodata.com>. (The previous implementation
13584 worked incorrectly for those cases where range = `10..._2` and
13585 `3*range` is two bits longer than range.)
13589 * Only add signing time to PKCS7 structures if it is not already
13594 * Fix crypto/objects/objects.h: "ld-ce" should be "id-ce",
13595 OBJ_ld_ce should be OBJ_id_ce.
13596 Also some ip-pda OIDs in crypto/objects/objects.txt were
13597 incorrect (cf. RFC 3039).
13599 *Matt Cooper, Frederic Giudicelli, Bodo Moeller*
13601 * Release CRYPTO_LOCK_DYNLOCK when CRYPTO_destroy_dynlockid()
13602 returns early because it has nothing to do.
13604 *Andy Schneider <andy.schneider@bjss.co.uk>*
13606 * [In 0.9.6c-engine release:]
13607 Fix mutex callback return values in crypto/engine/hw_ncipher.c.
13609 *Andy Schneider <andy.schneider@bjss.co.uk>*
13611 * [In 0.9.6c-engine release:]
13612 Add support for Cryptographic Appliance's keyserver technology.
13613 (Use engine 'keyclient')
13615 *Cryptographic Appliances and Geoff Thorpe*
13617 * Add a configuration entry for OS/390 Unix. The C compiler 'c89'
13618 is called via tools/c89.sh because arguments have to be
13619 rearranged (all '-L' options must appear before the first object
13622 *Richard Shapiro <rshapiro@abinitio.com>*
13624 * [In 0.9.6c-engine release:]
13625 Add support for Broadcom crypto accelerator cards, backported
13628 *Broadcom, Nalin Dahyabhai <nalin@redhat.com>, Mark Cox*
13630 * [In 0.9.6c-engine release:]
13631 Add support for SureWare crypto accelerator cards from
13632 Baltimore Technologies. (Use engine 'sureware')
13634 *Baltimore Technologies and Mark Cox*
13636 * [In 0.9.6c-engine release:]
13637 Add support for crypto accelerator cards from Accelerated
13638 Encryption Processing, www.aep.ie. (Use engine 'aep')
13640 *AEP Inc. and Mark Cox*
13642 * Add a configuration entry for gcc on UnixWare.
13644 *Gary Benson <gbenson@redhat.com>*
13646 * Change ssl/s2_clnt.c and ssl/s2_srvr.c so that received handshake
13647 messages are stored in a single piece (fixed-length part and
13648 variable-length part combined) and fix various bugs found on the way.
13652 * Disable caching in BIO_gethostbyname(), directly use gethostbyname()
13653 instead. BIO_gethostbyname() does not know what timeouts are
13654 appropriate, so entries would stay in cache even when they have
13656 *Bodo Moeller; problem pointed out by Rich Salz <rsalz@zolera.com>*
13658 * Change ssl23_get_client_hello (ssl/s23_srvr.c) behaviour when
13659 faced with a pathologically small ClientHello fragment that does
13660 not contain client_version: Instead of aborting with an error,
13661 simply choose the highest available protocol version (i.e.,
13662 TLS 1.0 unless it is disabled). In practice, ClientHello
13663 messages are never sent like this, but this change gives us
13664 strictly correct behaviour at least for TLS.
13668 * Fix SSL handshake functions and SSL_clear() such that SSL_clear()
13669 never resets s->method to s->ctx->method when called from within
13670 one of the SSL handshake functions.
13672 *Bodo Moeller; problem pointed out by Niko Baric*
13674 * In ssl3_get_client_hello (ssl/s3_srvr.c), generate a fatal alert
13675 (sent using the client's version number) if client_version is
13676 smaller than the protocol version in use. Also change
13677 ssl23_get_client_hello (ssl/s23_srvr.c) to select TLS 1.0 if
13678 the client demanded SSL 3.0 but only TLS 1.0 is enabled; then
13679 the client will at least see that alert.
13683 * Fix ssl3_get_message (ssl/s3_both.c) to handle message fragmentation
13688 * Avoid infinite loop in ssl3_get_message (ssl/s3_both.c) if a
13689 client receives HelloRequest while in a handshake.
13691 *Bodo Moeller; bug noticed by Andy Schneider <andy.schneider@bjss.co.uk>*
13693 * Bugfix in ssl3_accept (ssl/s3_srvr.c): Case SSL3_ST_SW_HELLO_REQ_C
13694 should end in 'break', not 'goto end' which circumvents various
13695 cleanups done in state SSL_ST_OK. But session related stuff
13696 must be disabled for SSL_ST_OK in the case that we just sent a
13699 Also avoid some overhead by not calling ssl_init_wbio_buffer()
13700 before just sending a HelloRequest.
13702 *Bodo Moeller, Eric Rescorla <ekr@rtfm.com>*
13704 * Fix ssl/s3_enc.c, ssl/t1_enc.c and ssl/s3_pkt.c so that we don't
13705 reveal whether illegal block cipher padding was found or a MAC
13706 verification error occurred. (Neither SSLerr() codes nor alerts
13707 are directly visible to potential attackers, but the information
13708 may leak via logfiles.)
13710 Similar changes are not required for the SSL 2.0 implementation
13711 because the number of padding bytes is sent in clear for SSL 2.0,
13712 and the extra bytes are just ignored. However ssl/s2_pkt.c
13713 failed to verify that the purported number of padding bytes is in
13718 * Add OpenUNIX-8 support including shared libraries
13719 (Boyd Lynn Gerber <gerberb@zenez.com>).
13723 * Improve RSA_padding_check_PKCS1_OAEP() check again to avoid
13724 'wristwatch attack' using huge encoding parameters (cf.
13725 James H. Manger's CRYPTO 2001 paper). Note that the
13726 RSA_PKCS1_OAEP_PADDING case of RSA_private_decrypt() does not use
13727 encoding parameters and hence was not vulnerable.
13731 * BN_sqr() bug fix.
13733 *Ulf Möller, reported by Jim Ellis <jim.ellis@cavium.com>*
13735 * Rabin-Miller test analyses assume uniformly distributed witnesses,
13736 so use BN_pseudo_rand_range() instead of using BN_pseudo_rand()
13737 followed by modular reduction.
13739 *Bodo Moeller; pointed out by Adam Young <AYoung1@NCSUS.JNJ.COM>*
13741 * Add BN_pseudo_rand_range() with obvious functionality: BN_rand_range()
13742 equivalent based on BN_pseudo_rand() instead of BN_rand().
13746 * s3_srvr.c: allow sending of large client certificate lists (> 16 kB).
13747 This function was broken, as the check for a new client hello message
13748 to handle SGC did not allow these large messages.
13749 (Tracked down by "Douglas E. Engert" <deengert@anl.gov>.)
13753 * Add alert descriptions for TLSv1 to `SSL_alert_desc_string[_long]()`.
13757 * Fix buggy behaviour of BIO_get_num_renegotiates() and BIO_ctrl()
13758 for BIO_C_GET_WRITE_BUF_SIZE ("Stephen Hinton" <shinton@netopia.com>).
13762 * Rework the configuration and shared library support for Tru64 Unix.
13763 The configuration part makes use of modern compiler features and
13764 still retains old compiler behavior for those that run older versions
13765 of the OS. The shared library support part includes a variant that
13766 uses the RPATH feature, and is available through the special
13767 configuration target "alpha-cc-rpath", which will never be selected
13770 *Tim Mooney <mooney@dogbert.cc.ndsu.NoDak.edu> via Richard Levitte*
13772 * In ssl3_get_key_exchange (ssl/s3_clnt.c), call ssl3_get_message()
13773 with the same message size as in ssl3_get_certificate_request().
13774 Otherwise, if no ServerKeyExchange message occurs, CertificateRequest
13775 messages might inadvertently be reject as too long.
13777 *Petr Lampa <lampa@fee.vutbr.cz>*
13779 * Enhanced support for IA-64 Unix platforms (well, Linux and HP-UX).
13783 * Modified SSL library such that the verify_callback that has been set
13784 specifically for an SSL object with SSL_set_verify() is actually being
13785 used. Before the change, a verify_callback set with this function was
13786 ignored and the verify_callback() set in the SSL_CTX at the time of
13787 the call was used. New function X509_STORE_CTX_set_verify_cb() introduced
13788 to allow the necessary settings.
13792 * Initialize static variable in crypto/dsa/dsa_lib.c and crypto/dh/dh_lib.c
13793 explicitly to NULL, as at least on Solaris 8 this seems not always to be
13794 done automatically (in contradiction to the requirements of the C
13795 standard). This made problems when used from OpenSSH.
13799 * In OpenSSL 0.9.6a and 0.9.6b, crypto/dh/dh_key.c ignored
13800 dh->length and always used
13802 BN_rand_range(priv_key, dh->p).
13804 BN_rand_range() is not necessary for Diffie-Hellman, and this
13805 specific range makes Diffie-Hellman unnecessarily inefficient if
13806 dh->length (recommended exponent length) is much smaller than the
13807 length of dh->p. We could use BN_rand_range() if the order of
13808 the subgroup was stored in the DH structure, but we only have
13813 BN_rand(priv_key, l, ...)
13815 where 'l' is dh->length if this is defined, or BN_num_bits(dh->p)-1
13822 RSA_eay_public_encrypt
13823 RSA_eay_private_decrypt
13824 RSA_eay_private_encrypt (signing)
13825 RSA_eay_public_decrypt (signature verification)
13827 (default implementations for RSA_public_encrypt,
13828 RSA_private_decrypt, RSA_private_encrypt, RSA_public_decrypt),
13829 always reject numbers >= n.
13833 * In crypto/rand/md_rand.c, use a new short-time lock CRYPTO_LOCK_RAND2
13834 to synchronize access to 'locking_thread'. This is necessary on
13835 systems where access to 'locking_thread' (an 'unsigned long'
13836 variable) is not atomic.
13840 * In crypto/rand/md_rand.c, set 'locking_thread' to current thread's ID
13841 *before* setting the 'crypto_lock_rand' flag. The previous code had
13842 a race condition if 0 is a valid thread ID.
13844 *Travis Vitek <vitek@roguewave.com>*
13846 * Add support for shared libraries under Irix.
13848 *Albert Chin-A-Young <china@thewrittenword.com>*
13850 * Add configuration option to build on Linux on both big-endian and
13851 little-endian MIPS.
13853 *Ralf Baechle <ralf@uni-koblenz.de>*
13855 * Add the possibility to create shared libraries on HP-UX.
13859 ### Changes between 0.9.6a and 0.9.6b [9 Jul 2001]
13861 * Change ssleay_rand_bytes (crypto/rand/md_rand.c)
13862 to avoid a SSLeay/OpenSSL PRNG weakness pointed out by
13863 Markku-Juhani O. Saarinen <markku-juhani.saarinen@nokia.com>:
13864 PRNG state recovery was possible based on the output of
13865 one PRNG request appropriately sized to gain knowledge on
13866 'md' followed by enough consecutive 1-byte PRNG requests
13867 to traverse all of 'state'.
13869 1. When updating 'md_local' (the current thread's copy of 'md')
13870 during PRNG output generation, hash all of the previous
13871 'md_local' value, not just the half used for PRNG output.
13873 2. Make the number of bytes from 'state' included into the hash
13874 independent from the number of PRNG bytes requested.
13876 The first measure alone would be sufficient to avoid
13877 Markku-Juhani's attack. (Actually it had never occurred
13878 to me that the half of 'md_local' used for chaining was the
13879 half from which PRNG output bytes were taken -- I had always
13880 assumed that the secret half would be used.) The second
13881 measure makes sure that additional data from 'state' is never
13882 mixed into 'md_local' in small portions; this heuristically
13883 further strengthens the PRNG.
13887 * Fix crypto/bn/asm/mips3.s.
13891 * When only the key is given to "enc", the IV is undefined. Print out
13892 an error message in this case.
13896 * Handle special case when X509_NAME is empty in X509 printing routines.
13900 * In dsa_do_verify (crypto/dsa/dsa_ossl.c), verify that r and s are
13901 positive and less than q.
13905 * Don't change `*pointer` in CRYPTO_add_lock() is add_lock_callback is
13906 used: it isn't thread safe and the add_lock_callback should handle
13909 *Paul Rose <Paul.Rose@bridge.com>*
13911 * Verify that incoming data obeys the block size in
13912 ssl3_enc (ssl/s3_enc.c) and tls1_enc (ssl/t1_enc.c).
13918 *Ulf Möller, Bodo Möller*
13920 * The countermeasure against Bleichbacher's attack on PKCS #1 v1.5
13921 RSA encryption was accidentally removed in s3_srvr.c in OpenSSL 0.9.5
13922 when fixing the server behaviour for backwards-compatible 'client
13923 hello' messages. (Note that the attack is impractical against
13924 SSL 3.0 and TLS 1.0 anyway because length and version checking
13925 means that the probability of guessing a valid ciphertext is
13926 around 2^-40; see section 5 in Bleichenbacher's CRYPTO '98
13929 Before 0.9.5, the countermeasure (hide the error by generating a
13930 random 'decryption result') did not work properly because
13931 ERR_clear_error() was missing, meaning that SSL_get_error() would
13932 detect the supposedly ignored error.
13934 Both problems are now fixed.
13938 * In crypto/bio/bf_buff.c, increase DEFAULT_BUFFER_SIZE to 4096
13939 (previously it was 1024).
13943 * Fix for compatibility mode trust settings: ignore trust settings
13944 unless some valid trust or reject settings are present.
13948 * Fix for blowfish EVP: its a variable length cipher.
13952 * Fix various bugs related to DSA S/MIME verification. Handle missing
13953 parameters in DSA public key structures and return an error in the
13954 DSA routines if parameters are absent.
13958 * In versions up to 0.9.6, RAND_file_name() resorted to file ".rnd"
13959 in the current directory if neither $RANDFILE nor $HOME was set.
13960 RAND_file_name() in 0.9.6a returned NULL in this case. This has
13961 caused some confusion to Windows users who haven't defined $HOME.
13962 Thus RAND_file_name() is changed again: e_os.h can define a
13963 DEFAULT_HOME, which will be used if $HOME is not set.
13964 For Windows, we use "C:"; on other platforms, we still require
13965 environment variables.
13967 * Move 'if (!initialized) RAND_poll()' into regions protected by
13968 CRYPTO_LOCK_RAND. This is not strictly necessary, but avoids
13969 having multiple threads call RAND_poll() concurrently.
13973 * In crypto/rand/md_rand.c, replace 'add_do_not_lock' flag by a
13974 combination of a flag and a thread ID variable.
13975 Otherwise while one thread is in ssleay_rand_bytes (which sets the
13976 flag), *other* threads can enter ssleay_add_bytes without obeying
13977 the CRYPTO_LOCK_RAND lock (and may even illegally release the lock
13978 that they do not hold after the first thread unsets add_do_not_lock).
13982 * Change bctest again: '-x' expressions are not available in all
13983 versions of 'test'.
13987 ### Changes between 0.9.6 and 0.9.6a [5 Apr 2001]
13989 * Fix a couple of memory leaks in PKCS7_dataDecode()
13991 *Steve Henson, reported by Heyun Zheng <hzheng@atdsprint.com>*
13993 * Change Configure and Makefiles to provide EXE_EXT, which will contain
13994 the default extension for executables, if any. Also, make the perl
13995 scripts that use symlink() to test if it really exists and use "cp"
13996 if it doesn't. All this made OpenSSL compilable and installable in
14001 * Fix for asn1_GetSequence() for indefinite length constructed data.
14002 If SEQUENCE is length is indefinite just set c->slen to the total
14003 amount of data available.
14005 *Steve Henson, reported by shige@FreeBSD.org*
14007 *This change does not apply to 0.9.7.*
14009 * Change bctest to avoid here-documents inside command substitution
14010 (workaround for FreeBSD /bin/sh bug).
14011 For compatibility with Ultrix, avoid shell functions (introduced
14012 in the bctest version that searches along $PATH).
14016 * Rename 'des_encrypt' to 'des_encrypt1'. This avoids the clashes
14017 with des_encrypt() defined on some operating systems, like Solaris
14022 * Check the result of RSA-CRT (see D. Boneh, R. DeMillo, R. Lipton:
14023 On the Importance of Eliminating Errors in Cryptographic
14024 Computations, J. Cryptology 14 (2001) 2, 101-119,
14025 <http://theory.stanford.edu/~dabo/papers/faults.ps.gz>).
14029 * MIPS assembler BIGNUM division bug fix.
14033 * Disabled incorrect Alpha assembler code.
14037 * Fix PKCS#7 decode routines so they correctly update the length
14038 after reading an EOC for the EXPLICIT tag.
14042 *This change does not apply to 0.9.7.*
14044 * Fix bug in PKCS#12 key generation routines. This was triggered
14045 if a 3DES key was generated with a 0 initial byte. Include
14046 PKCS12_BROKEN_KEYGEN compilation option to retain the old
14047 (but broken) behaviour.
14051 * Enhance bctest to search for a working bc along $PATH and print
14054 *Tim Rice <tim@multitalents.net> via Richard Levitte*
14056 * Fix memory leaks in err.c: free err_data string if necessary;
14057 don't write to the wrong index in ERR_set_error_data.
14061 * Implement ssl23_peek (analogous to ssl23_read), which previously
14066 * Replace rdtsc with `_emit` statements for VC++ version 5.
14068 *Jeremy Cooper <jeremy@baymoo.org>*
14070 * Make it possible to reuse SSLv2 sessions.
14074 * In copy_email() check for >= 0 as a return value for
14075 X509_NAME_get_index_by_NID() since 0 is a valid index.
14077 *Steve Henson reported by Massimiliano Pala <madwolf@opensca.org>*
14079 * Avoid coredump with unsupported or invalid public keys by checking if
14080 X509_get_pubkey() fails in PKCS7_verify(). Fix memory leak when
14081 PKCS7_verify() fails with non detached data.
14085 * Don't use getenv in library functions when run as setuid/setgid.
14086 New function OPENSSL_issetugid().
14090 * Avoid false positives in memory leak detection code (crypto/mem_dbg.c)
14091 due to incorrect handling of multi-threading:
14093 1. Fix timing glitch in the MemCheck_off() portion of CRYPTO_mem_ctrl().
14095 2. Fix logical glitch in is_MemCheck_on() aka CRYPTO_is_mem_check_on().
14097 3. Count how many times MemCheck_off() has been called so that
14098 nested use can be treated correctly. This also avoids
14099 inband-signalling in the previous code (which relied on the
14100 assumption that thread ID 0 is impossible).
14104 * Add "-rand" option also to s_client and s_server.
14108 * Fix CPU detection on Irix 6.x.
14109 *Kurt Hockenbury <khockenb@stevens-tech.edu> and
14110 "Bruce W. Forsberg" <bruce.forsberg@baesystems.com>*
14112 * Fix X509_NAME bug which produced incorrect encoding if X509_NAME
14117 *This change does not apply to 0.9.7.*
14119 * Use the cached encoding of an X509_NAME structure rather than
14120 copying it. This is apparently the reason for the libsafe "errors"
14121 but the code is actually correct.
14125 * Add new function BN_rand_range(), and fix DSA_sign_setup() to prevent
14126 Bleichenbacher's DSA attack.
14127 Extend BN_[pseudo_]rand: As before, top=1 forces the highest two bits
14128 to be set and top=0 forces the highest bit to be set; top=-1 is new
14129 and leaves the highest bit random.
14131 *Ulf Moeller, Bodo Moeller*
14133 * In the `NCONF_...`-based implementations for `CONF_...` queries
14134 (crypto/conf/conf_lib.c), if the input LHASH is NULL, avoid using
14135 a temporary CONF structure with the data component set to NULL
14136 (which gives segmentation faults in lh_retrieve).
14137 Instead, use NULL for the CONF pointer in CONF_get_string and
14138 CONF_get_number (which may use environment variables) and directly
14139 return NULL from CONF_get_section.
14143 * Fix potential buffer overrun for EBCDIC.
14147 * Tolerate nonRepudiation as being valid for S/MIME signing and certSign
14148 keyUsage if basicConstraints absent for a CA.
14152 * Make SMIME_write_PKCS7() write mail header values with a format that
14153 is more generally accepted (no spaces before the semicolon), since
14154 some programs can't parse those values properly otherwise. Also make
14155 sure BIO's that break lines after each write do not create invalid
14160 * Make the CRL encoding routines work with empty SEQUENCE OF. The
14161 macros previously used would not encode an empty SEQUENCE OF
14162 and break the signature.
14166 *This change does not apply to 0.9.7.*
14168 * Zero the premaster secret after deriving the master secret in
14173 * Add some EVP_add_digest_alias registrations (as found in
14174 OpenSSL_add_all_digests()) to SSL_library_init()
14175 aka OpenSSL_add_ssl_algorithms(). This provides improved
14176 compatibility with peers using X.509 certificates
14177 with unconventional AlgorithmIdentifier OIDs.
14181 * Fix for Irix with NO_ASM.
14183 *"Bruce W. Forsberg" <bruce.forsberg@baesystems.com>*
14185 * ./config script fixes.
14187 *Ulf Moeller, Richard Levitte*
14189 * Fix 'openssl passwd -1'.
14193 * Change PKCS12_key_gen_asc() so it can cope with non null
14194 terminated strings whose length is passed in the passlen
14195 parameter, for example from PEM callbacks. This was done
14196 by adding an extra length parameter to asc2uni().
14198 *Steve Henson, reported by <oddissey@samsung.co.kr>*
14200 * Fix C code generated by 'openssl dsaparam -C': If a BN_bin2bn
14201 call failed, free the DSA structure.
14205 * Fix to uni2asc() to cope with zero length Unicode strings.
14206 These are present in some PKCS#12 files.
14210 * Increase s2->wbuf allocation by one byte in ssl2_new (ssl/s2_lib.c).
14211 Otherwise do_ssl_write (ssl/s2_pkt.c) will write beyond buffer limits
14212 when writing a 32767 byte record.
14214 *Bodo Moeller; problem reported by Eric Day <eday@concentric.net>*
14216 * In `RSA_eay_public_{en,ed}crypt` and RSA_eay_mod_exp (rsa_eay.c),
14217 obtain lock CRYPTO_LOCK_RSA before setting `rsa->_method_mod_{n,p,q}`.
14219 (RSA objects have a reference count access to which is protected
14220 by CRYPTO_LOCK_RSA [see rsa_lib.c, s3_srvr.c, ssl_cert.c, ssl_rsa.c],
14221 so they are meant to be shared between threads.)
14222 *Bodo Moeller, Geoff Thorpe; original patch submitted by
14223 "Reddie, Steven" <Steven.Reddie@ca.com>*
14225 * Fix a deadlock in CRYPTO_mem_leaks().
14229 * Use better test patterns in bntest.
14233 * rand_win.c fix for Borland C.
14237 * BN_rshift bugfix for n == 0.
14241 * Add a 'bctest' script that checks for some known 'bc' bugs
14242 so that 'make test' does not abort just because 'bc' is broken.
14246 * Store verify_result within SSL_SESSION also for client side to
14247 avoid potential security hole. (Re-used sessions on the client side
14248 always resulted in verify_result==X509_V_OK, not using the original
14249 result of the server certificate verification.)
14253 * Fix ssl3_pending: If the record in s->s3->rrec is not of type
14254 SSL3_RT_APPLICATION_DATA, return 0.
14255 Similarly, change ssl2_pending to return 0 if SSL_in_init(s) is true.
14260 Both ssl2_peek and ssl3_peek, which were totally broken in earlier
14261 releases, have been re-implemented by renaming the previous
14262 implementations of ssl2_read and ssl3_read to ssl2_read_internal
14263 and ssl3_read_internal, respectively, and adding 'peek' parameters
14264 to them. The new ssl[23]_{read,peek} functions are calls to
14265 ssl[23]_read_internal with the 'peek' flag set appropriately.
14266 A 'peek' parameter has also been added to ssl3_read_bytes, which
14267 does the actual work for ssl3_read_internal.
14271 * Initialise "ex_data" member of RSA/DSA/DH structures prior to calling
14272 the method-specific "init()" handler. Also clean up ex_data after
14273 calling the method-specific "finish()" handler. Previously, this was
14274 happening the other way round.
14278 * Increase BN_CTX_NUM (the number of BIGNUMs in a BN_CTX) to 16.
14279 The previous value, 12, was not always sufficient for BN_mod_exp().
14283 * Make sure that shared libraries get the internal name engine with
14284 the full version number and not just 0. This should mark the
14285 shared libraries as not backward compatible. Of course, this should
14286 be changed again when we can guarantee backward binary compatibility.
14290 * Fix typo in get_cert_by_subject() in by_dir.c
14292 *Jean-Marc Desperrier <jean-marc.desperrier@certplus.com>*
14294 * Rework the system to generate shared libraries:
14296 - Make note of the expected extension for the shared libraries and
14297 if there is a need for symbolic links from for example libcrypto.so.0
14298 to libcrypto.so.0.9.7. There is extended info in Configure for
14301 - Make as few rebuilds of the shared libraries as possible.
14303 - Still avoid linking the OpenSSL programs with the shared libraries.
14305 - When installing, install the shared libraries separately from the
14310 * Fix SSL_CTX_set_read_ahead macro to actually use its argument.
14312 Copy SSL_CTX's read_ahead flag to SSL object directly in SSL_new
14313 and not in SSL_clear because the latter is also used by the
14314 accept/connect functions; previously, the settings made by
14315 SSL_set_read_ahead would be lost during the handshake.
14317 *Bodo Moeller; problems reported by Anders Gertz <gertz@epact.se>*
14319 * Correct util/mkdef.pl to be selective about disabled algorithms.
14320 Previously, it would create entries for disabled algorithms no
14325 * Added several new manual pages for SSL_* function.
14329 ### Changes between 0.9.5a and 0.9.6 [24 Sep 2000]
14331 * In ssl23_get_client_hello, generate an error message when faced
14332 with an initial SSL 3.0/TLS record that is too small to contain the
14333 first two bytes of the ClientHello message, i.e. client_version.
14334 (Note that this is a pathologic case that probably has never happened
14335 in real life.) The previous approach was to use the version number
14336 from the record header as a substitute; but our protocol choice
14337 should not depend on that one because it is not authenticated
14338 by the Finished messages.
14342 * More robust randomness gathering functions for Windows.
14344 *Jeffrey Altman <jaltman@columbia.edu>*
14346 * For compatibility reasons if the flag X509_V_FLAG_ISSUER_CHECK is
14347 not set then we don't setup the error code for issuer check errors
14348 to avoid possibly overwriting other errors which the callback does
14349 handle. If an application does set the flag then we assume it knows
14350 what it is doing and can handle the new informational codes
14355 * Fix for a nasty bug in ASN1_TYPE handling. ASN1_TYPE is used for
14356 a general "ANY" type, as such it should be able to decode anything
14357 including tagged types. However it didn't check the class so it would
14358 wrongly interpret tagged types in the same way as their universal
14359 counterpart and unknown types were just rejected. Changed so that the
14360 tagged and unknown types are handled in the same way as a SEQUENCE:
14361 that is the encoding is stored intact. There is also a new type
14362 "V_ASN1_OTHER" which is used when the class is not universal, in this
14363 case we have no idea what the actual type is so we just lump them all
14368 * On VMS, stdout may very well lead to a file that is written to
14369 in a record-oriented fashion. That means that every write() will
14370 write a separate record, which will be read separately by the
14371 programs trying to read from it. This can be very confusing.
14373 The solution is to put a BIO filter in the way that will buffer
14374 text until a linefeed is reached, and then write everything a
14375 line at a time, so every record written will be an actual line,
14376 not chunks of lines and not (usually doesn't happen, but I've
14377 seen it once) several lines in one record. BIO_f_linebuffer() is
14380 Currently, it's a VMS-only method, because that's where it has
14381 been tested well enough.
14385 * Remove 'optimized' squaring variant in BN_mod_mul_montgomery,
14386 it can return incorrect results.
14387 (Note: The buggy variant was not enabled in OpenSSL 0.9.5a,
14388 but it was in 0.9.6-beta[12].)
14392 * Disable the check for content being present when verifying detached
14393 signatures in pk7_smime.c. Some versions of Netscape (wrongly)
14394 include zero length content when signing messages.
14398 * New BIO_shutdown_wr macro, which invokes the BIO_C_SHUTDOWN_WR
14399 BIO_ctrl (for BIO pairs).
14403 * Add DSO method for VMS.
14407 * Bug fix: Montgomery multiplication could produce results with the
14412 * Add RPM specification openssl.spec and modify it to build three
14413 packages. The default package contains applications, application
14414 documentation and run-time libraries. The devel package contains
14415 include files, static libraries and function documentation. The
14416 doc package contains the contents of the doc directory. The original
14417 openssl.spec was provided by Damien Miller <djm@mindrot.org>.
14421 * Add a large number of documentation files for many SSL routines.
14423 *Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>*
14425 * Add a configuration entry for Sony News 4.
14427 *NAKAJI Hiroyuki <nakaji@tutrp.tut.ac.jp>*
14429 * Don't set the two most significant bits to one when generating a
14430 random number < q in the DSA library.
14434 * New SSL API mode 'SSL_MODE_AUTO_RETRY'. This disables the default
14435 behaviour that SSL_read may result in SSL_ERROR_WANT_READ (even if
14436 the underlying transport is blocking) if a handshake took place.
14437 (The default behaviour is needed by applications such as s_client
14438 and s_server that use select() to determine when to use SSL_read;
14439 but for applications that know in advance when to expect data, it
14440 just makes things more complicated.)
14444 * Add RAND_egd_bytes(), which gives control over the number of bytes read
14449 * Add a few more EBCDIC conditionals that make `req` and `x509`
14450 work better on such systems.
14452 *Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>*
14454 * Add two demo programs for PKCS12_parse() and PKCS12_create().
14455 Update PKCS12_parse() so it copies the friendlyName and the
14456 keyid to the certificates aux info.
14460 * Fix bug in PKCS7_verify() which caused an infinite loop
14461 if there was more than one signature.
14463 *Sven Uszpelkat <su@celocom.de>*
14465 * Major change in util/mkdef.pl to include extra information
14466 about each symbol, as well as presenting variables as well
14467 as functions. This change means that there's n more need
14468 to rebuild the .num files when some algorithms are excluded.
14472 * Allow the verify time to be set by an application,
14473 rather than always using the current time.
14477 * Phase 2 verify code reorganisation. The certificate
14478 verify code now looks up an issuer certificate by a
14479 number of criteria: subject name, authority key id
14480 and key usage. It also verifies self signed certificates
14481 by the same criteria. The main comparison function is
14482 X509_check_issued() which performs these checks.
14484 Lot of changes were necessary in order to support this
14485 without completely rewriting the lookup code.
14487 Authority and subject key identifier are now cached.
14489 The LHASH 'certs' is X509_STORE has now been replaced
14490 by a STACK_OF(X509_OBJECT). This is mainly because an
14491 LHASH can't store or retrieve multiple objects with
14492 the same hash value.
14494 As a result various functions (which were all internal
14495 use only) have changed to handle the new X509_STORE
14496 structure. This will break anything that messed round
14497 with X509_STORE internally.
14499 The functions X509_STORE_add_cert() now checks for an
14500 exact match, rather than just subject name.
14502 The X509_STORE API doesn't directly support the retrieval
14503 of multiple certificates matching a given criteria, however
14504 this can be worked round by performing a lookup first
14505 (which will fill the cache with candidate certificates)
14506 and then examining the cache for matches. This is probably
14507 the best we can do without throwing out X509_LOOKUP
14508 entirely (maybe later...).
14510 The X509_VERIFY_CTX structure has been enhanced considerably.
14512 All certificate lookup operations now go via a get_issuer()
14513 callback. Although this currently uses an X509_STORE it
14514 can be replaced by custom lookups. This is a simple way
14515 to bypass the X509_STORE hackery necessary to make this
14516 work and makes it possible to use more efficient techniques
14517 in future. A very simple version which uses a simple
14518 STACK for its trusted certificate store is also provided
14519 using X509_STORE_CTX_trusted_stack().
14521 The verify_cb() and verify() callbacks now have equivalents
14522 in the X509_STORE_CTX structure.
14524 X509_STORE_CTX also has a 'flags' field which can be used
14525 to customise the verify behaviour.
14529 * Add new PKCS#7 signing option PKCS7_NOSMIMECAP which
14530 excludes S/MIME capabilities.
14534 * When a certificate request is read in keep a copy of the
14535 original encoding of the signed data and use it when outputting
14536 again. Signatures then use the original encoding rather than
14537 a decoded, encoded version which may cause problems if the
14538 request is improperly encoded.
14542 * For consistency with other BIO_puts implementations, call
14543 buffer_write(b, ...) directly in buffer_puts instead of calling
14546 In BIO_puts, increment b->num_write as in BIO_write.
14548 *Peter.Sylvester@EdelWeb.fr*
14550 * Fix BN_mul_word for the case where the word is 0. (We have to use
14551 BN_zero, we may not return a BIGNUM with an array consisting of
14552 words set to zero.)
14556 * Avoid calling abort() from within the library when problems are
14557 detected, except if preprocessor symbols have been defined
14558 (such as REF_CHECK, BN_DEBUG etc.).
14562 * New openssl application 'rsautl'. This utility can be
14563 used for low level RSA operations. DER public key
14564 BIO/fp routines also added.
14568 * New Configure entry and patches for compiling on QNX 4.
14570 *Andreas Schneider <andreas@ds3.etech.fh-hamburg.de>*
14572 * A demo state-machine implementation was sponsored by
14573 Nuron (<http://www.nuron.com/>) and is now available in
14574 demos/state_machine.
14578 * New options added to the 'dgst' utility for signature
14579 generation and verification.
14583 * Unrecognized PKCS#7 content types are now handled via a
14584 catch all ASN1_TYPE structure. This allows unsupported
14585 types to be stored as a "blob" and an application can
14586 encode and decode it manually.
14590 * Fix various signed/unsigned issues to make a_strex.c
14591 compile under VC++.
14593 *Oscar Jacobsson <oscar.jacobsson@celocom.com>*
14595 * ASN1 fixes. i2d_ASN1_OBJECT was not returning the correct
14596 length if passed a buffer. ASN1_INTEGER_to_BN failed
14597 if passed a NULL BN and its argument was negative.
14599 *Steve Henson, pointed out by Sven Heiberg <sven@tartu.cyber.ee>*
14601 * Modification to PKCS#7 encoding routines to output definite
14602 length encoding. Since currently the whole structures are in
14603 memory there's not real point in using indefinite length
14604 constructed encoding. However if OpenSSL is compiled with
14605 the flag PKCS7_INDEFINITE_ENCODING the old form is used.
14609 * Added BIO_vprintf() and BIO_vsnprintf().
14613 * Added more prefixes to parse for in the strings written
14614 through a logging bio, to cover all the levels that are available
14615 through syslog. The prefixes are now:
14617 PANIC, EMERG, EMR => LOG_EMERG
14618 ALERT, ALR => LOG_ALERT
14619 CRIT, CRI => LOG_CRIT
14620 ERROR, ERR => LOG_ERR
14621 WARNING, WARN, WAR => LOG_WARNING
14622 NOTICE, NOTE, NOT => LOG_NOTICE
14623 INFO, INF => LOG_INFO
14624 DEBUG, DBG => LOG_DEBUG
14626 and as before, if none of those prefixes are present at the
14627 beginning of the string, LOG_ERR is chosen.
14629 On Win32, the `LOG_*` levels are mapped according to this:
14631 LOG_EMERG, LOG_ALERT, LOG_CRIT, LOG_ERR => EVENTLOG_ERROR_TYPE
14632 LOG_WARNING => EVENTLOG_WARNING_TYPE
14633 LOG_NOTICE, LOG_INFO, LOG_DEBUG => EVENTLOG_INFORMATION_TYPE
14637 * Made it possible to reconfigure with just the configuration
14638 argument "reconf" or "reconfigure". The command line arguments
14639 are stored in Makefile.ssl in the variable CONFIGURE_ARGS,
14640 and are retrieved from there when reconfiguring.
14646 *Assar Westerlund <assar@sics.se>, Richard Levitte*
14648 * Add the arguments -CAfile and -CApath to the pkcs12 utility.
14652 * The obj_dat.pl script was messing up the sorting of object
14653 names. The reason was that it compared the quoted version
14654 of strings as a result "OCSP" > "OCSP Signing" because
14655 " > SPACE. Changed script to store unquoted versions of
14656 names and add quotes on output. It was also omitting some
14657 names from the lookup table if they were given a default
14658 value (that is if SN is missing it is given the same
14659 value as LN and vice versa), these are now added on the
14660 grounds that if an object has a name we should be able to
14661 look it up. Finally added warning output when duplicate
14662 short or long names are found.
14666 * Changes needed for Tandem NSK.
14668 *Scott Uroff <scott@xypro.com>*
14670 * Fix SSL 2.0 rollback checking: Due to an off-by-one error in
14671 RSA_padding_check_SSLv23(), special padding was never detected
14672 and thus the SSL 3.0/TLS 1.0 countermeasure against protocol
14673 version rollback attacks was not effective.
14675 In s23_clnt.c, don't use special rollback-attack detection padding
14676 (RSA_SSLV23_PADDING) if SSL 2.0 is the only protocol enabled in the
14677 client; similarly, in s23_srvr.c, don't do the rollback check if
14678 SSL 2.0 is the only protocol enabled in the server.
14682 * Make it possible to get hexdumps of unprintable data with 'openssl
14683 asn1parse'. By implication, the functions ASN1_parse_dump() and
14684 BIO_dump_indent() are added.
14688 * New functions ASN1_STRING_print_ex() and X509_NAME_print_ex()
14689 these print out strings and name structures based on various
14690 flags including RFC2253 support and proper handling of
14691 multibyte characters. Added options to the 'x509' utility
14692 to allow the various flags to be set.
14696 * Various fixes to use ASN1_TIME instead of ASN1_UTCTIME.
14697 Also change the functions X509_cmp_current_time() and
14698 X509_gmtime_adj() work with an ASN1_TIME structure,
14699 this will enable certificates using GeneralizedTime in validity
14700 dates to be checked.
14704 * Make the NEG_PUBKEY_BUG code (which tolerates invalid
14705 negative public key encodings) on by default,
14706 NO_NEG_PUBKEY_BUG can be set to disable it.
14710 * New function c2i_ASN1_OBJECT() which acts on ASN1_OBJECT
14711 content octets. An i2c_ASN1_OBJECT is unnecessary because
14712 the encoding can be trivially obtained from the structure.
14716 * crypto/err.c locking bugfix: Use write locks (`CRYPTO_w_[un]lock`),
14717 not read locks (`CRYPTO_r_[un]lock`).
14721 * A first attempt at creating official support for shared
14722 libraries through configuration. I've kept it so the
14723 default is static libraries only, and the OpenSSL programs
14724 are always statically linked for now, but there are
14725 preparations for dynamic linking in place.
14726 This has been tested on Linux and Tru64.
14730 * Randomness polling function for Win9x, as described in:
14731 Peter Gutmann, Software Generation of Practically Strong
14736 * Fix so PRNG is seeded in req if using an already existing
14741 * New options to smime application. -inform and -outform
14742 allow alternative formats for the S/MIME message including
14743 PEM and DER. The -content option allows the content to be
14744 specified separately. This should allow things like Netscape
14745 form signing output easier to verify.
14749 * Fix the ASN1 encoding of tags using the 'long form'.
14753 * New ASN1 functions, `i2c_*` and `c2i_*` for INTEGER and BIT
14754 STRING types. These convert content octets to and from the
14755 underlying type. The actual tag and length octets are
14756 already assumed to have been read in and checked. These
14757 are needed because all other string types have virtually
14758 identical handling apart from the tag. By having versions
14759 of the ASN1 functions that just operate on content octets
14760 IMPLICIT tagging can be handled properly. It also allows
14761 the ASN1_ENUMERATED code to be cut down because ASN1_ENUMERATED
14762 and ASN1_INTEGER are identical apart from the tag.
14766 * Change the handling of OID objects as follows:
14768 - New object identifiers are inserted in objects.txt, following
14769 the syntax given in [crypto/objects/README.md](crypto/objects/README.md).
14770 - objects.pl is used to process obj_mac.num and create a new
14772 - obj_dat.pl is used to create a new obj_dat.h, using the data in
14775 This is currently kind of a hack, and the perl code in objects.pl
14776 isn't very elegant, but it works as I intended. The simplest way
14777 to check that it worked correctly is to look in obj_dat.h and
14778 check the array nid_objs and make sure the objects haven't moved
14779 around (this is important!). Additions are OK, as well as
14780 consistent name changes.
14784 * Add BSD-style MD5-based passwords to 'openssl passwd' (option '-1').
14788 * Addition of the command line parameter '-rand file' to 'openssl req'.
14789 The given file adds to whatever has already been seeded into the
14790 random pool through the RANDFILE configuration file option or
14791 environment variable, or the default random state file.
14795 * mkstack.pl now sorts each macro group into lexical order.
14796 Previously the output order depended on the order the files
14797 appeared in the directory, resulting in needless rewriting
14802 * Patches to make OpenSSL compile under Win32 again. Mostly
14803 work arounds for the VC++ problem that it treats func() as
14804 func(void). Also stripped out the parts of mkdef.pl that
14805 added extra typesafe functions: these no longer exist.
14809 * Reorganisation of the stack code. The macros are now all
14810 collected in safestack.h . Each macro is defined in terms of
14811 a "stack macro" of the form `SKM_<name>(type, a, b)`. The
14812 DEBUG_SAFESTACK is now handled in terms of function casts,
14813 this has the advantage of retaining type safety without the
14814 use of additional functions. If DEBUG_SAFESTACK is not defined
14815 then the non typesafe macros are used instead. Also modified the
14816 mkstack.pl script to handle the new form. Needs testing to see
14817 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK
14818 the default if no major problems. Similar behaviour for ASN1_SET_OF
14819 and PKCS12_STACK_OF.
14823 * When some versions of IIS use the 'NET' form of private key the
14824 key derivation algorithm is different. Normally MD5(password) is
14825 used as a 128 bit RC4 key. In the modified case
14826 MD5(MD5(password) + "SGCKEYSALT") is used instead. Added some
14827 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same
14828 as the old Netscape_RSA functions except they have an additional
14829 'sgckey' parameter which uses the modified algorithm. Also added
14830 an -sgckey command line option to the rsa utility. Thanks to
14831 Adrian Peck <bertie@ncipher.com> for posting details of the modified
14832 algorithm to openssl-dev.
14836 * The evp_local.h macros were using 'c.##kname' which resulted in
14837 invalid expansion on some systems (SCO 5.0.5 for example).
14838 Corrected to 'c.kname'.
14840 *Phillip Porch <root@theporch.com>*
14842 * New X509_get1_email() and X509_REQ_get1_email() functions that return
14843 a STACK of email addresses from a certificate or request, these look
14844 in the subject name and the subject alternative name extensions and
14845 omit any duplicate addresses.
14849 * Re-implement BN_mod_exp2_mont using independent (and larger) windows.
14850 This makes DSA verification about 2 % faster.
14854 * Increase maximum window size in `BN_mod_exp_...` to 6 bits instead of 5
14855 (meaning that now 2^5 values will be precomputed, which is only 4 KB
14856 plus overhead for 1024 bit moduli).
14857 This makes exponentiations about 0.5 % faster for 1024 bit
14858 exponents (as measured by "openssl speed rsa2048").
14862 * Rename memory handling macros to avoid conflicts with other
14864 Malloc => OPENSSL_malloc
14865 Malloc_locked => OPENSSL_malloc_locked
14866 Realloc => OPENSSL_realloc
14867 Free => OPENSSL_free
14871 * New function BN_mod_exp_mont_word for small bases (roughly 15%
14872 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange).
14876 * CygWin32 support.
14878 *John Jarvie <jjarvie@newsguy.com>*
14880 * The type-safe stack code has been rejigged. It is now only compiled
14881 in when OpenSSL is configured with the DEBUG_SAFESTACK option and
14882 by default all type-specific stack functions are "#define"d back to
14883 standard stack functions. This results in more streamlined output
14884 but retains the type-safety checking possibilities of the original
14889 * The STACK code has been cleaned up, and certain type declarations
14890 that didn't make a lot of sense have been brought in line. This has
14891 also involved a cleanup of sorts in safestack.h to more correctly
14892 map type-safe stack functions onto their plain stack counterparts.
14893 This work has also resulted in a variety of "const"ifications of
14894 lots of the code, especially `_cmp` operations which should normally
14895 be prototyped with "const" parameters anyway.
14899 * When generating bytes for the first time in md_rand.c, 'stir the pool'
14900 by seeding with STATE_SIZE dummy bytes (with zero entropy count).
14901 (The PRNG state consists of two parts, the large pool 'state' and 'md',
14902 where all of 'md' is used each time the PRNG is used, but 'state'
14903 is used only indexed by a cyclic counter. As entropy may not be
14904 well distributed from the beginning, 'md' is important as a
14905 chaining variable. However, the output function chains only half
14906 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains
14907 all of 'md', and seeding with STATE_SIZE dummy bytes will result
14908 in all of 'state' being rewritten, with the new values depending
14909 on virtually all of 'md'. This overcomes the 80 bit limitation.)
14913 * In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when
14914 the handshake is continued after ssl_verify_cert_chain();
14915 otherwise, if SSL_VERIFY_NONE is set, remaining error codes
14916 can lead to 'unexplainable' connection aborts later.
14918 *Bodo Moeller; problem tracked down by Lutz Jaenicke*
14920 * Major EVP API cipher revision.
14921 Add hooks for extra EVP features. This allows various cipher
14922 parameters to be set in the EVP interface. Support added for variable
14923 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and
14924 setting of RC2 and RC5 parameters.
14926 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length
14929 Remove lots of duplicated code from the EVP library. For example *every*
14930 cipher init() function handles the 'iv' in the same way according to the
14931 cipher mode. They also all do nothing if the 'key' parameter is NULL and
14932 for CFB and OFB modes they zero ctx->num.
14934 New functionality allows removal of S/MIME code RC2 hack.
14936 Most of the routines have the same form and so can be declared in terms
14939 By shifting this to the top level EVP_CipherInit() it can be removed from
14940 all individual ciphers. If the cipher wants to handle IVs or keys
14941 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT
14944 Change lots of functions like EVP_EncryptUpdate() to now return a
14945 value: although software versions of the algorithms cannot fail
14946 any installed hardware versions can.
14950 * Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if
14951 this option is set, tolerate broken clients that send the negotiated
14952 protocol version number instead of the requested protocol version
14957 * Call dh_tmp_cb (set by `..._TMP_DH_CB`) with correct 'is_export' flag;
14958 i.e. non-zero for export ciphersuites, zero otherwise.
14959 Previous versions had this flag inverted, inconsistent with
14960 rsa_tmp_cb (..._TMP_RSA_CB).
14962 *Bodo Moeller; problem reported by Amit Chopra*
14964 * Add missing DSA library text string. Work around for some IIS
14965 key files with invalid SEQUENCE encoding.
14969 * Add a document (doc/standards.txt) that list all kinds of standards
14970 and so on that are implemented in OpenSSL.
14974 * Enhance c_rehash script. Old version would mishandle certificates
14975 with the same subject name hash and wouldn't handle CRLs at all.
14976 Added -fingerprint option to crl utility, to support new c_rehash
14981 * Eliminate non-ANSI declarations in crypto.h and stack.h.
14985 * Fix for SSL server purpose checking. Server checking was
14986 rejecting certificates which had extended key usage present
14987 but no ssl client purpose.
14989 *Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>*
14991 * Make PKCS#12 code work with no password. The PKCS#12 spec
14992 is a little unclear about how a blank password is handled.
14993 Since the password in encoded as a BMPString with terminating
14994 double NULL a zero length password would end up as just the
14995 double NULL. However no password at all is different and is
14996 handled differently in the PKCS#12 key generation code. NS
14997 treats a blank password as zero length. MSIE treats it as no
14998 password on export: but it will try both on import. We now do
14999 the same: PKCS12_parse() tries zero length and no password if
15000 the password is set to "" or NULL (NULL is now a valid password:
15001 it wasn't before) as does the pkcs12 application.
15005 * Bugfixes in apps/x509.c: Avoid a memory leak; and don't use
15006 perror when PEM_read_bio_X509_REQ fails, the error message must
15007 be obtained from the error queue.
15011 * Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing
15012 it in ERR_remove_state if appropriate, and change ERR_get_state
15013 accordingly to avoid race conditions (this is necessary because
15014 thread_hash is no longer constant once set).
15018 * Bugfix for linux-elf makefile.one.
15022 * RSA_get_default_method() will now cause a default
15023 RSA_METHOD to be chosen if one doesn't exist already.
15024 Previously this was only set during a call to RSA_new()
15025 or RSA_new_method(NULL) meaning it was possible for
15026 RSA_get_default_method() to return NULL.
15030 * Added native name translation to the existing DSO code
15031 that will convert (if the flag to do so is set) filenames
15032 that are sufficiently small and have no path information
15033 into a canonical native form. Eg. "blah" converted to
15034 "libblah.so" or "blah.dll" etc.
15038 * New function ERR_error_string_n(e, buf, len) which is like
15039 ERR_error_string(e, buf), but writes at most 'len' bytes
15040 including the 0 terminator. For ERR_error_string_n, 'buf'
15043 *Damien Miller <djm@mindrot.org>, Bodo Moeller*
15045 * CONF library reworked to become more general. A new CONF
15046 configuration file reader "class" is implemented as well as a
15047 new functions (`NCONF_*`, for "New CONF") to handle it. The now
15048 old `CONF_*` functions are still there, but are reimplemented to
15049 work in terms of the new functions. Also, a set of functions
15050 to handle the internal storage of the configuration data is
15051 provided to make it easier to write new configuration file
15052 reader "classes" (I can definitely see something reading a
15053 configuration file in XML format, for example), called `_CONF_*`,
15054 or "the configuration storage API"...
15056 The new configuration file reading functions are:
15058 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio,
15059 NCONF_get_section, NCONF_get_string, NCONF_get_numbre
15061 NCONF_default, NCONF_WIN32
15063 NCONF_dump_fp, NCONF_dump_bio
15065 NCONF_default and NCONF_WIN32 are method (or "class") choosers,
15066 NCONF_new creates a new CONF object. This works in the same way
15067 as other interfaces in OpenSSL, like the BIO interface.
15068 `NCONF_dump_*` dump the internal storage of the configuration file,
15069 which is useful for debugging. All other functions take the same
15070 arguments as the old `CONF_*` functions with the exception of the
15071 first that must be a `CONF *` instead of a `LHASH *`.
15073 To make it easier to use the new classes with the old `CONF_*` functions,
15074 the function CONF_set_default_method is provided.
15078 * Add '-tls1' option to 'openssl ciphers', which was already
15079 mentioned in the documentation but had not been implemented.
15080 (This option is not yet really useful because even the additional
15081 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.)
15085 * Initial DSO code added into libcrypto for letting OpenSSL (and
15086 OpenSSL-based applications) load shared libraries and bind to
15087 them in a portable way.
15089 *Geoff Thorpe, with contributions from Richard Levitte*
15091 ### Changes between 0.9.5 and 0.9.5a [1 Apr 2000]
15093 * Make sure _lrotl and _lrotr are only used with MSVC.
15095 * Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status
15096 (the default implementation of RAND_status).
15098 * Rename openssl x509 option '-crlext', which was added in 0.9.5,
15099 to '-clrext' (= clear extensions), as intended and documented.
15100 *Bodo Moeller; inconsistency pointed out by Michael Attili
15101 <attili@amaxo.com>*
15103 * Fix for HMAC. It wasn't zeroing the rest of the block if the key length
15104 was larger than the MD block size.
15106 *Steve Henson, pointed out by Yost William <YostW@tce.com>*
15108 * Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument
15109 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set()
15110 using the passed key: if the passed key was a private key the result
15111 of X509_print(), for example, would be to print out all the private key
15116 * des_quad_cksum() byte order bug fix.
15117 *Ulf Möller, using the problem description in krb4-0.9.7, where
15118 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>*
15120 * Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly
15123 *Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>*
15125 * For easily testing in shell scripts whether some command
15126 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX'
15127 returns with exit code 0 iff no command of the given name is available.
15128 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases,
15129 the output goes to stdout and nothing is printed to stderr.
15130 Additional arguments are always ignored.
15132 Since for each cipher there is a command of the same name,
15133 the 'no-cipher' compilation switches can be tested this way.
15135 ('openssl no-XXX' is not able to detect pseudo-commands such
15136 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.)
15140 * Update test suite so that 'make test' succeeds in 'no-rsa' configuration.
15144 * For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE
15145 is set; it will be thrown away anyway because each handshake creates
15147 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition
15148 to parameters -- in previous versions (since OpenSSL 0.9.3) the
15149 'default key' from SSL_CTX_set_tmp_dh would always be lost, meaning
15150 you effectively got SSL_OP_SINGLE_DH_USE when using this macro.
15154 * New s_client option -ign_eof: EOF at stdin is ignored, and
15155 'Q' and 'R' lose their special meanings (quit/renegotiate).
15156 This is part of what -quiet does; unlike -quiet, -ign_eof
15157 does not suppress any output.
15161 * Add compatibility options to the purpose and trust code. The
15162 purpose X509_PURPOSE_ANY is "any purpose" which automatically
15163 accepts a certificate or CA, this was the previous behaviour,
15164 with all the associated security issues.
15166 X509_TRUST_COMPAT is the old trust behaviour: only and
15167 automatically trust self signed roots in certificate store. A
15168 new trust setting X509_TRUST_DEFAULT is used to specify that
15169 a purpose has no associated trust setting and it should instead
15170 use the value in the default purpose.
15174 * Fix the PKCS#8 DSA private key code so it decodes keys again
15175 and fix a memory leak.
15179 * In util/mkerr.pl (which implements 'make errors'), preserve
15180 reason strings from the previous version of the .c file, as
15181 the default to have only downcase letters (and digits) in
15182 automatically generated reasons codes is not always appropriate.
15186 * In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table
15187 using strerror. Previously, ERR_reason_error_string() returned
15188 library names as reason strings for SYSerr; but SYSerr is a special
15189 case where small numbers are errno values, not library numbers.
15193 * Add '-dsaparam' option to 'openssl dhparam' application. This
15194 converts DSA parameters into DH parameters. (When creating parameters,
15195 DSA_generate_parameters is used.)
15199 * Include 'length' (recommended exponent length) in C code generated
15200 by 'openssl dhparam -C'.
15204 * The second argument to set_label in perlasm was already being used
15205 so couldn't be used as a "file scope" flag. Moved to third argument
15210 * In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes
15211 instead of RAND_bytes for encryption IVs and salts.
15215 * Include RAND_status() into RAND_METHOD instead of implementing
15216 it only for md_rand.c Otherwise replacing the PRNG by calling
15217 RAND_set_rand_method would be impossible.
15221 * Don't let DSA_generate_key() enter an infinite loop if the random
15222 number generation fails.
15226 * New 'rand' application for creating pseudo-random output.
15230 * Added configuration support for Linux/IA64
15232 *Rolf Haberrecker <rolf@suse.de>*
15234 * Assembler module support for Mingw32.
15238 * Shared library support for HPUX (in shlib/).
15240 *Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous*
15242 * Shared library support for Solaris gcc.
15244 *Lutz Behnke <behnke@trustcenter.de>*
15246 ### Changes between 0.9.4 and 0.9.5 [28 Feb 2000]
15248 * PKCS7_encrypt() was adding text MIME headers twice because they
15249 were added manually and by SMIME_crlf_copy().
15253 * In bntest.c don't call BN_rand with zero bits argument.
15255 *Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>*
15257 * BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n]
15258 case was implemented. This caused BN_div_recp() to fail occasionally.
15262 * Add an optional second argument to the set_label() in the perl
15263 assembly language builder. If this argument exists and is set
15264 to 1 it signals that the assembler should use a symbol whose
15265 scope is the entire file, not just the current function. This
15266 is needed with MASM which uses the format label:: for this scope.
15268 *Steve Henson, pointed out by Peter Runestig <peter@runestig.com>*
15270 * Change the ASN1 types so they are typedefs by default. Before
15271 almost all types were #define'd to ASN1_STRING which was causing
15272 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING)
15277 * Change names of new functions to the new get1/get0 naming
15278 convention: After 'get1', the caller owns a reference count
15279 and has to call `..._free`; 'get0' returns a pointer to some
15280 data structure without incrementing reference counters.
15281 (Some of the existing 'get' functions increment a reference
15282 counter, some don't.)
15283 Similarly, 'set1' and 'add1' functions increase reference
15284 counters or duplicate objects.
15288 * Allow for the possibility of temp RSA key generation failure:
15289 the code used to assume it always worked and crashed on failure.
15293 * Fix potential buffer overrun problem in BIO_printf().
15294 *Ulf Möller, using public domain code by Patrick Powell; problem
15295 pointed out by David Sacerdote <das33@cornell.edu>*
15297 * Support EGD <http://www.lothar.com/tech/crypto/>. New functions
15298 RAND_egd() and RAND_status(). In the command line application,
15299 the EGD socket can be specified like a seed file using RANDFILE
15304 * Allow the string CERTIFICATE to be tolerated in PKCS#7 structures.
15305 Some CAs (e.g. Verisign) distribute certificates in this form.
15309 * Remove the SSL_ALLOW_ADH compile option and set the default cipher
15310 list to exclude them. This means that no special compilation option
15311 is needed to use anonymous DH: it just needs to be included in the
15316 * Change the EVP_MD_CTX_type macro so its meaning consistent with
15317 EVP_MD_type. The old functionality is available in a new macro called
15318 EVP_MD_md(). Change code that uses it and update docs.
15322 * `..._ctrl` functions now have corresponding `..._callback_ctrl` functions
15323 where the `void *` argument is replaced by a function pointer argument.
15324 Previously `void *` was abused to point to functions, which works on
15325 many platforms, but is not correct. As these functions are usually
15326 called by macros defined in OpenSSL header files, most source code
15327 should work without changes.
15331 * `<openssl/opensslconf.h>` (which is created by Configure) now contains
15332 sections with information on -D... compiler switches used for
15333 compiling the library so that applications can see them. To enable
15334 one of these sections, a pre-processor symbol `OPENSSL_..._DEFINES`
15335 must be defined. E.g.,
15336 #define OPENSSL_ALGORITHM_DEFINES
15337 #include <openssl/opensslconf.h>
15338 defines all pertinent `NO_<algo>` symbols, such as NO_IDEA, NO_RSA, etc.
15340 *Richard Levitte, Ulf and Bodo Möller*
15342 * Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS
15347 * Change the 'other' type in certificate aux info to a STACK_OF
15348 X509_ALGOR. Although not an AlgorithmIdentifier as such it has
15349 the required ASN1 format: arbitrary types determined by an OID.
15353 * Add some PEM_write_X509_REQ_NEW() functions and a command line
15354 argument to 'req'. This is not because the function is newer or
15355 better than others it just uses the work 'NEW' in the certificate
15356 request header lines. Some software needs this.
15360 * Reorganise password command line arguments: now passwords can be
15361 obtained from various sources. Delete the PEM_cb function and make
15362 it the default behaviour: i.e. if the callback is NULL and the
15363 usrdata argument is not NULL interpret it as a null terminated pass
15364 phrase. If usrdata and the callback are NULL then the pass phrase
15365 is prompted for as usual.
15369 * Add support for the Compaq Atalla crypto accelerator. If it is installed,
15370 the support is automatically enabled. The resulting binaries will
15371 autodetect the card and use it if present.
15373 *Ben Laurie and Compaq Inc.*
15375 * Work around for Netscape hang bug. This sends certificate request
15376 and server done in one record. Since this is perfectly legal in the
15377 SSL/TLS protocol it isn't a "bug" option and is on by default. See
15378 the bugs/SSLv3 entry for more info.
15382 * HP-UX tune-up: new unified configs, HP C compiler bug workaround.
15386 * Add -rand argument to smime and pkcs12 applications and read/write
15391 * New 'passwd' tool for crypt(3) and apr1 password hashes.
15395 * Add command line password options to the remaining applications.
15399 * Bug fix for BN_div_recp() for numerators with an even number of
15404 * More tests in bntest.c, and changed test_bn output.
15408 * ./config recognizes MacOS X now.
15412 * Bug fix for BN_div() when the first words of num and divisor are
15413 equal (it gave wrong results if `(rem=(n1-q*d0)&BN_MASK2) < d0)`.
15417 * Add support for various broken PKCS#8 formats, and command line
15418 options to produce them.
15422 * New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to
15423 get temporary BIGNUMs from a BN_CTX.
15427 * Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont()
15432 * Change the `SSLeay_add_all_*()` functions to `OpenSSL_add_all_*()` and
15433 include a #define from the old name to the new. The original intent
15434 was that statically linked binaries could for example just call
15435 SSLeay_add_all_ciphers() to just add ciphers to the table and not
15436 link with digests. This never worked because SSLeay_add_all_digests()
15437 and SSLeay_add_all_ciphers() were in the same source file so calling
15438 one would link with the other. They are now in separate source files.
15442 * Add a new -notext option to 'ca' and a -pubkey option to 'spkac'.
15446 * Use a less unusual form of the Miller-Rabin primality test (it used
15447 a binary algorithm for exponentiation integrated into the Miller-Rabin
15448 loop, our standard modexp algorithms are faster).
15452 * Support for the EBCDIC character set completed.
15454 *Martin Kraemer <Martin.Kraemer@Mch.SNI.De>*
15456 * Source code cleanups: use const where appropriate, eliminate casts,
15457 use `void *` instead of `char *` in lhash.
15461 * Bugfix: ssl3_send_server_key_exchange was not restartable
15462 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of
15463 this the server could overwrite ephemeral keys that the client
15468 * Turn DSA_is_prime into a macro that calls BN_is_prime,
15469 using 50 iterations of the Rabin-Miller test.
15471 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50
15472 iterations of the Rabin-Miller test as required by the appendix
15473 to FIPS PUB 186[-1]) instead of DSA_is_prime.
15474 As BN_is_prime_fasttest includes trial division, DSA parameter
15475 generation becomes much faster.
15477 This implies a change for the callback functions in DSA_is_prime
15478 and DSA_generate_parameters: The callback function is called once
15479 for each positive witness in the Rabin-Miller test, not just
15480 occasionally in the inner loop; and the parameters to the
15481 callback function now provide an iteration count for the outer
15482 loop rather than for the current invocation of the inner loop.
15483 DSA_generate_parameters additionally can call the callback
15484 function with an 'iteration count' of -1, meaning that a
15485 candidate has passed the trial division test (when q is generated
15486 from an application-provided seed, trial division is skipped).
15490 * New function BN_is_prime_fasttest that optionally does trial
15491 division before starting the Rabin-Miller test and has
15492 an additional BN_CTX * argument (whereas BN_is_prime always
15493 has to allocate at least one BN_CTX).
15494 'callback(1, -1, cb_arg)' is called when a number has passed the
15495 trial division stage.
15499 * Fix for bug in CRL encoding. The validity dates weren't being handled
15504 * New -pkcs12 option to CA.pl script to write out a PKCS#12 file.
15508 * New function BN_pseudo_rand().
15512 * Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable)
15513 bignum version of BN_from_montgomery() with the working code from
15514 SSLeay 0.9.0 (the word based version is faster anyway), and clean up
15519 * Avoid a race condition in s2_clnt.c (function get_server_hello) that
15520 made it impossible to use the same SSL_SESSION data structure in
15521 SSL2 clients in multiple threads.
15525 * The return value of RAND_load_file() no longer counts bytes obtained
15526 by stat(). RAND_load_file(..., -1) is new and uses the complete file
15527 to seed the PRNG (previously an explicit byte count was required).
15529 *Ulf Möller, Bodo Möller*
15531 * Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
15532 used `char *` instead of `void *` and had casts all over the place.
15536 * Make BN_generate_prime() return NULL on error if ret!=NULL.
15540 * Retain source code compatibility for BN_prime_checks macro:
15541 BN_is_prime(..., BN_prime_checks, ...) now uses
15542 BN_prime_checks_for_size to determine the appropriate number of
15543 Rabin-Miller iterations.
15547 * Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
15548 DH_CHECK_P_NOT_SAFE_PRIME.
15549 (Check if this is true? OpenPGP calls them "strong".)
15553 * Merge the functionality of "dh" and "gendh" programs into a new program
15554 "dhparam". The old programs are retained for now but will handle DH keys
15555 (instead of parameters) in future.
15559 * Make the ciphers, s_server and s_client programs check the return values
15560 when a new cipher list is set.
15564 * Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
15565 ciphers. Before when the 56bit ciphers were enabled the sorting was
15568 The syntax for the cipher sorting has been extended to support sorting by
15569 cipher-strength (using the strength_bits hard coded in the tables).
15570 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
15572 Fix a bug in the cipher-command parser: when supplying a cipher command
15573 string with an "undefined" symbol (neither command nor alphanumeric
15574 *A-Za-z0-9*, ssl_set_cipher_list used to hang in an endless loop. Now
15575 an error is flagged.
15577 Due to the strength-sorting extension, the code of the
15578 ssl_create_cipher_list() function was completely rearranged. I hope that
15579 the readability was also increased :-)
15581 *Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>*
15583 * Minor change to 'x509' utility. The -CAcreateserial option now uses 1
15584 for the first serial number and places 2 in the serial number file. This
15585 avoids problems when the root CA is created with serial number zero and
15586 the first user certificate has the same issuer name and serial number
15591 * Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
15592 the new code. Add documentation for this stuff.
15596 * Changes to X509_ATTRIBUTE utilities. These have been renamed from
15597 `X509_*()` to `X509at_*()` on the grounds that they don't handle X509
15598 structures and behave in an analogous way to the X509v3 functions:
15599 they shouldn't be called directly but wrapper functions should be used
15602 So we also now have some wrapper functions that call the X509at functions
15603 when passed certificate requests. (TO DO: similar things can be done with
15604 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
15605 things. Some of these need some d2i or i2d and print functionality
15606 because they handle more complex structures.)
15610 * Add missing #ifndefs that caused missing symbols when building libssl
15611 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
15612 NO_RSA in `ssl/s2*.c`.
15614 *Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller*
15616 * Precautions against using the PRNG uninitialized: RAND_bytes() now
15617 has a return value which indicates the quality of the random data
15618 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
15619 error queue. New function RAND_pseudo_bytes() generates output that is
15620 guaranteed to be unique but not unpredictable. RAND_add is like
15621 RAND_seed, but takes an extra argument for an entropy estimate
15622 (RAND_seed always assumes full entropy).
15626 * Do more iterations of Rabin-Miller probable prime test (specifically,
15627 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
15628 instead of only 2 for all lengths; see BN_prime_checks_for_size definition
15629 in crypto/bn/bn_prime.c for the complete table). This guarantees a
15630 false-positive rate of at most 2^-80 for random input.
15634 * Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
15638 * New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain
15639 in the 0.9.5 release), this returns the chain
15640 from an X509_CTX structure with a dup of the stack and all
15641 the X509 reference counts upped: so the stack will exist
15642 after X509_CTX_cleanup() has been called. Modify pkcs12.c
15645 Also make SSL_SESSION_print() print out the verify return
15650 * Add manpage for the pkcs12 command. Also change the default
15651 behaviour so MAC iteration counts are used unless the new
15652 -nomaciter option is used. This improves file security and
15653 only older versions of MSIE (4.0 for example) need it.
15657 * Honor the no-xxx Configure options when creating .DEF files.
15661 * Add PKCS#10 attributes to field table: challengePassword,
15662 unstructuredName and unstructuredAddress. These are taken from
15663 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
15664 international characters are used.
15666 More changes to X509_ATTRIBUTE code: allow the setting of types
15667 based on strings. Remove the 'loc' parameter when adding
15668 attributes because these will be a SET OF encoding which is sorted
15673 * Initial changes to the 'req' utility to allow request generation
15674 automation. This will allow an application to just generate a template
15675 file containing all the field values and have req construct the
15678 Initial support for X509_ATTRIBUTE handling. Stacks of these are
15679 used all over the place including certificate requests and PKCS#7
15680 structures. They are currently handled manually where necessary with
15681 some primitive wrappers for PKCS#7. The new functions behave in a
15682 manner analogous to the X509 extension functions: they allow
15683 attributes to be looked up by NID and added.
15685 Later something similar to the X509V3 code would be desirable to
15686 automatically handle the encoding, decoding and printing of the
15687 more complex types. The string types like challengePassword can
15688 be handled by the string table functions.
15690 Also modified the multi byte string table handling. Now there is
15691 a 'global mask' which masks out certain types. The table itself
15692 can use the flag STABLE_NO_MASK to ignore the mask setting: this
15693 is useful when for example there is only one permissible type
15694 (as in countryName) and using the mask might result in no valid
15699 * Clean up 'Finished' handling, and add functions SSL_get_finished and
15700 SSL_get_peer_finished to allow applications to obtain the latest
15701 Finished messages sent to the peer or expected from the peer,
15702 respectively. (SSL_get_peer_finished is usually the Finished message
15703 actually received from the peer, otherwise the protocol will be aborted.)
15705 As the Finished message are message digests of the complete handshake
15706 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
15707 be used for external authentication procedures when the authentication
15708 provided by SSL/TLS is not desired or is not enough.
15712 * Enhanced support for Alpha Linux is added. Now ./config checks if
15713 the host supports BWX extension and if Compaq C is present on the
15714 $PATH. Just exploiting of the BWX extension results in 20-30%
15715 performance kick for some algorithms, e.g. DES and RC4 to mention
15716 a couple. Compaq C in turn generates ~20% faster code for MD5 and
15721 * Add support for MS "fast SGC". This is arguably a violation of the
15722 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
15723 weak crypto and after checking the certificate is SGC a second one
15724 with strong crypto. MS SGC stops the first handshake after receiving
15725 the server certificate message and sends a second client hello. Since
15726 a server will typically do all the time consuming operations before
15727 expecting any further messages from the client (server key exchange
15728 is the most expensive) there is little difference between the two.
15730 To get OpenSSL to support MS SGC we have to permit a second client
15731 hello message after we have sent server done. In addition we have to
15732 reset the MAC if we do get this second client hello.
15736 * Add a function 'd2i_AutoPrivateKey()' this will automatically decide
15737 if a DER encoded private key is RSA or DSA traditional format. Changed
15738 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
15739 format DER encoded private key. Newer code should use PKCS#8 format which
15740 has the key type encoded in the ASN1 structure. Added DER private key
15741 support to pkcs8 application.
15745 * SSL 3/TLS 1 servers now don't request certificates when an anonymous
15746 ciphersuites has been selected (as required by the SSL 3/TLS 1
15747 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
15748 is set, we interpret this as a request to violate the specification
15749 (the worst that can happen is a handshake failure, and 'correct'
15750 behaviour would result in a handshake failure anyway).
15754 * In SSL_CTX_add_session, take into account that there might be multiple
15755 SSL_SESSION structures with the same session ID (e.g. when two threads
15756 concurrently obtain them from an external cache).
15757 The internal cache can handle only one SSL_SESSION with a given ID,
15758 so if there's a conflict, we now throw out the old one to achieve
15763 * Add OIDs for idea and blowfish in CBC mode. This will allow both
15764 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
15765 some routines that use cipher OIDs: some ciphers do not have OIDs
15766 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
15771 * Simplify the trust setting structure and code. Now we just have
15772 two sequences of OIDs for trusted and rejected settings. These will
15773 typically have values the same as the extended key usage extension
15774 and any application specific purposes.
15776 The trust checking code now has a default behaviour: it will just
15777 check for an object with the same NID as the passed id. Functions can
15778 be provided to override either the default behaviour or the behaviour
15779 for a given id. SSL client, server and email already have functions
15780 in place for compatibility: they check the NID and also return "trusted"
15781 if the certificate is self signed.
15785 * Add d2i,i2d bio/fp functions for PrivateKey: these convert the
15786 traditional format into an EVP_PKEY structure.
15790 * Add a password callback function PEM_cb() which either prompts for
15791 a password if usr_data is NULL or otherwise assumes it is a null
15792 terminated password. Allow passwords to be passed on command line
15793 environment or config files in a few more utilities.
15797 * Add a bunch of DER and PEM functions to handle PKCS#8 format private
15798 keys. Add some short names for PKCS#8 PBE algorithms and allow them
15799 to be specified on the command line for the pkcs8 and pkcs12 utilities.
15800 Update documentation.
15804 * Support for ASN1 "NULL" type. This could be handled before by using
15805 ASN1_TYPE but there wasn't any function that would try to read a NULL
15806 and produce an error if it couldn't. For compatibility we also have
15807 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
15808 don't allocate anything because they don't need to.
15812 * Initial support for MacOS is now provided. Examine INSTALL.MacOS
15815 *Andy Polyakov, Roy Woods <roy@centicsystems.ca>*
15817 * Rebuild of the memory allocation routines used by OpenSSL code and
15818 possibly others as well. The purpose is to make an interface that
15819 provide hooks so anyone can build a separate set of allocation and
15820 deallocation routines to be used by OpenSSL, for example memory
15821 pool implementations, or something else, which was previously hard
15822 since Malloc(), Realloc() and Free() were defined as macros having
15823 the values malloc, realloc and free, respectively (except for Win32
15824 compilations). The same is provided for memory debugging code.
15825 OpenSSL already comes with functionality to find memory leaks, but
15826 this gives people a chance to debug other memory problems.
15828 With these changes, a new set of functions and macros have appeared:
15830 CRYPTO_set_mem_debug_functions() [F]
15831 CRYPTO_get_mem_debug_functions() [F]
15832 CRYPTO_dbg_set_options() [F]
15833 CRYPTO_dbg_get_options() [F]
15834 CRYPTO_malloc_debug_init() [M]
15836 The memory debug functions are NULL by default, unless the library
15837 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
15838 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
15839 gives the standard debugging functions that come with OpenSSL) or
15840 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
15841 provided by the library user) must be used. When the standard
15842 debugging functions are used, CRYPTO_dbg_set_options can be used to
15843 request additional information:
15844 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
15845 the CRYPTO_MDEBUG_xxx macro when compiling the library.
15847 Also, things like CRYPTO_set_mem_functions will always give the
15848 expected result (the new set of functions is used for allocation
15849 and deallocation) at all times, regardless of platform and compiler
15852 To finish it up, some functions that were never use in any other
15853 way than through macros have a new API and new semantic:
15855 CRYPTO_dbg_malloc()
15856 CRYPTO_dbg_realloc()
15859 All macros of value have retained their old syntax.
15861 *Richard Levitte and Bodo Moeller*
15863 * Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
15864 ordering of SMIMECapabilities wasn't in "strength order" and there
15865 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
15870 * Some ASN1 types with illegal zero length encoding (INTEGER,
15871 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
15873 *Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson*
15875 * Merge in my S/MIME library for OpenSSL. This provides a simple
15876 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
15877 functionality to handle multipart/signed properly) and a utility
15878 called 'smime' to call all this stuff. This is based on code I
15879 originally wrote for Celo who have kindly allowed it to be
15880 included in OpenSSL.
15884 * Add variants des_set_key_checked and des_set_key_unchecked of
15885 des_set_key (aka des_key_sched). Global variable des_check_key
15886 decides which of these is called by des_set_key; this way
15887 des_check_key behaves as it always did, but applications and
15888 the library itself, which was buggy for des_check_key == 1,
15889 have a cleaner way to pick the version they need.
15893 * New function PKCS12_newpass() which changes the password of a
15898 * Modify X509_TRUST and X509_PURPOSE so it also uses a static and
15899 dynamic mix. In both cases the ids can be used as an index into the
15900 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
15901 functions so they accept a list of the field values and the
15902 application doesn't need to directly manipulate the X509_TRUST
15907 * Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
15912 * Modify the way the V3 extension code looks up extensions. This now
15913 works in a similar way to the object code: we have some "standard"
15914 extensions in a static table which is searched with OBJ_bsearch()
15915 and the application can add dynamic ones if needed. The file
15916 crypto/x509v3/ext_dat.h now has the info: this file needs to be
15917 updated whenever a new extension is added to the core code and kept
15918 in ext_nid order. There is a simple program 'tabtest.c' which checks
15919 this. New extensions are not added too often so this file can readily
15920 be maintained manually.
15922 There are two big advantages in doing things this way. The extensions
15923 can be looked up immediately and no longer need to be "added" using
15924 X509V3_add_standard_extensions(): this function now does nothing.
15925 Side note: I get *lots* of email saying the extension code doesn't
15926 work because people forget to call this function.
15927 Also no dynamic allocation is done unless new extensions are added:
15928 so if we don't add custom extensions there is no need to call
15929 X509V3_EXT_cleanup().
15933 * Modify enc utility's salting as follows: make salting the default. Add a
15934 magic header, so unsalted files fail gracefully instead of just decrypting
15935 to garbage. This is because not salting is a big security hole, so people
15936 should be discouraged from doing it.
15940 * Fixes and enhancements to the 'x509' utility. It allowed a message
15941 digest to be passed on the command line but it only used this
15942 parameter when signing a certificate. Modified so all relevant
15943 operations are affected by the digest parameter including the
15944 -fingerprint and -x509toreq options. Also -x509toreq choked if a
15945 DSA key was used because it didn't fix the digest.
15949 * Initial certificate chain verify code. Currently tests the untrusted
15950 certificates for consistency with the verify purpose (which is set
15951 when the X509_STORE_CTX structure is set up) and checks the pathlength.
15953 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
15954 this is because it will reject chains with invalid extensions whereas
15955 every previous version of OpenSSL and SSLeay made no checks at all.
15957 Trust code: checks the root CA for the relevant trust settings. Trust
15958 settings have an initial value consistent with the verify purpose: e.g.
15959 if the verify purpose is for SSL client use it expects the CA to be
15960 trusted for SSL client use. However the default value can be changed to
15961 permit custom trust settings: one example of this would be to only trust
15962 certificates from a specific "secure" set of CAs.
15964 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
15965 which should be used for version portability: especially since the
15966 verify structure is likely to change more often now.
15968 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
15969 to set them. If not set then assume SSL clients will verify SSL servers
15972 Two new options to the verify program: -untrusted allows a set of
15973 untrusted certificates to be passed in and -purpose which sets the
15974 intended purpose of the certificate. If a purpose is set then the
15975 new chain verify code is used to check extension consistency.
15979 * Support for the authority information access extension.
15983 * Modify RSA and DSA PEM read routines to transparently handle
15984 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
15985 public keys in a format compatible with certificate
15986 SubjectPublicKeyInfo structures. Unfortunately there were already
15987 functions called *_PublicKey_* which used various odd formats so
15988 these are retained for compatibility: however the DSA variants were
15989 never in a public release so they have been deleted. Changed dsa/rsa
15990 utilities to handle the new format: note no releases ever handled public
15991 keys so we should be OK.
15993 The primary motivation for this change is to avoid the same fiasco
15994 that dogs private keys: there are several incompatible private key
15995 formats some of which are standard and some OpenSSL specific and
15996 require various evil hacks to allow partial transparent handling and
15997 even then it doesn't work with DER formats. Given the option anything
15998 other than PKCS#8 should be dumped: but the other formats have to
15999 stay in the name of compatibility.
16001 With public keys and the benefit of hindsight one standard format
16002 is used which works with EVP_PKEY, RSA or DSA structures: though
16003 it clearly returns an error if you try to read the wrong kind of key.
16005 Added a -pubkey option to the 'x509' utility to output the public key.
16006 Also rename the `EVP_PKEY_get_*()` to `EVP_PKEY_rget_*()`
16007 (renamed to `EVP_PKEY_get1_*()` in the OpenSSL 0.9.5 release) and add
16008 `EVP_PKEY_rset_*()` functions (renamed to `EVP_PKEY_set1_*()`)
16009 that do the same as the `EVP_PKEY_assign_*()` except they up the
16010 reference count of the added key (they don't "swallow" the
16015 * Fixes to crypto/x509/by_file.c the code to read in certificates and
16016 CRLs would fail if the file contained no certificates or no CRLs:
16017 added a new function to read in both types and return the number
16018 read: this means that if none are read it will be an error. The
16019 DER versions of the certificate and CRL reader would always fail
16020 because it isn't possible to mix certificates and CRLs in DER format
16021 without choking one or the other routine. Changed this to just read
16022 a certificate: this is the best we can do. Also modified the code
16023 in apps/verify.c to take notice of return codes: it was previously
16024 attempting to read in certificates from NULL pointers and ignoring
16025 any errors: this is one reason why the cert and CRL reader seemed
16026 to work. It doesn't check return codes from the default certificate
16027 routines: these may well fail if the certificates aren't installed.
16031 * Code to support otherName option in GeneralName.
16035 * First update to verify code. Change the verify utility
16036 so it warns if it is passed a self signed certificate:
16037 for consistency with the normal behaviour. X509_verify
16038 has been modified to it will now verify a self signed
16039 certificate if *exactly* the same certificate appears
16040 in the store: it was previously impossible to trust a
16041 single self signed certificate. This means that:
16042 openssl verify ss.pem
16043 now gives a warning about a self signed certificate but
16044 openssl verify -CAfile ss.pem ss.pem
16049 * For servers, store verify_result in SSL_SESSION data structure
16050 (and add it to external session representation).
16051 This is needed when client certificate verifications fails,
16052 but an application-provided verification callback (set by
16053 SSL_CTX_set_cert_verify_callback) allows accepting the session
16054 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
16055 but returns 1): When the session is reused, we have to set
16056 ssl->verify_result to the appropriate error code to avoid
16059 *Bodo Moeller, problem pointed out by Lutz Jaenicke*
16061 * Fix a bug in the new PKCS#7 code: it didn't consider the
16062 case in PKCS7_dataInit() where the signed PKCS7 structure
16063 didn't contain any existing data because it was being created.
16065 *Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson*
16067 * Add a salt to the key derivation routines in enc.c. This
16068 forms the first 8 bytes of the encrypted file. Also add a
16069 -S option to allow a salt to be input on the command line.
16073 * New function X509_cmp(). Oddly enough there wasn't a function
16074 to compare two certificates. We do this by working out the SHA1
16075 hash and comparing that. X509_cmp() will be needed by the trust
16080 * SSL_get1_session() is like SSL_get_session(), but increments
16081 the reference count in the SSL_SESSION returned.
16083 *Geoff Thorpe <geoff@eu.c2.net>*
16085 * Fix for 'req': it was adding a null to request attributes.
16086 Also change the X509_LOOKUP and X509_INFO code to handle
16087 certificate auxiliary information.
16091 * Add support for 40 and 64 bit RC2 and RC4 algorithms: document
16096 * Add the possibility to add extra information to the memory leak
16097 detecting output, to form tracebacks, showing from where each
16098 allocation was originated: CRYPTO_push_info("constant string") adds
16099 the string plus current file name and line number to a per-thread
16100 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
16101 is like calling CYRPTO_pop_info() until the stack is empty.
16102 Also updated memory leak detection code to be multi-thread-safe.
16106 * Add options -text and -noout to pkcs7 utility and delete the
16107 encryption options which never did anything. Update docs.
16111 * Add options to some of the utilities to allow the pass phrase
16112 to be included on either the command line (not recommended on
16113 OSes like Unix) or read from the environment. Update the
16114 manpages and fix a few bugs.
16118 * Add a few manpages for some of the openssl commands.
16122 * Fix the -revoke option in ca. It was freeing up memory twice,
16123 leaking and not finding already revoked certificates.
16127 * Extensive changes to support certificate auxiliary information.
16128 This involves the use of X509_CERT_AUX structure and X509_AUX
16129 functions. An X509_AUX function such as PEM_read_X509_AUX()
16130 can still read in a certificate file in the usual way but it
16131 will also read in any additional "auxiliary information". By
16132 doing things this way a fair degree of compatibility can be
16133 retained: existing certificates can have this information added
16134 using the new 'x509' options.
16136 Current auxiliary information includes an "alias" and some trust
16137 settings. The trust settings will ultimately be used in enhanced
16138 certificate chain verification routines: currently a certificate
16139 can only be trusted if it is self signed and then it is trusted
16144 * Fix assembler for Alpha (tested only on DEC OSF not Linux or `*BSD`).
16145 The problem was that one of the replacement routines had not been working
16146 since SSLeay releases. For now the offending routine has been replaced
16147 with non-optimised assembler. Even so, this now gives around 95%
16148 performance improvement for 1024 bit RSA signs.
16152 * Hack to fix PKCS#7 decryption when used with some unorthodox RC2
16153 handling. Most clients have the effective key size in bits equal to
16154 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
16155 A few however don't do this and instead use the size of the decrypted key
16156 to determine the RC2 key length and the AlgorithmIdentifier to determine
16157 the effective key length. In this case the effective key length can still
16158 be 40 bits but the key length can be 168 bits for example. This is fixed
16159 by manually forcing an RC2 key into the EVP_PKEY structure because the
16160 EVP code can't currently handle unusual RC2 key sizes: it always assumes
16161 the key length and effective key length are equal.
16165 * Add a bunch of functions that should simplify the creation of
16166 X509_NAME structures. Now you should be able to do:
16167 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
16168 and have it automatically work out the correct field type and fill in
16169 the structures. The more adventurous can try:
16170 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
16171 and it will (hopefully) work out the correct multibyte encoding.
16175 * Change the 'req' utility to use the new field handling and multibyte
16176 copy routines. Before the DN field creation was handled in an ad hoc
16177 way in req, ca, and x509 which was rather broken and didn't support
16178 BMPStrings or UTF8Strings. Since some software doesn't implement
16179 BMPStrings or UTF8Strings yet, they can be enabled using the config file
16180 using the dirstring_type option. See the new comment in the default
16181 openssl.cnf for more info.
16185 * Make crypto/rand/md_rand.c more robust:
16186 - Assure unique random numbers after fork().
16187 - Make sure that concurrent threads access the global counter and
16188 md serializably so that we never lose entropy in them
16189 or use exactly the same state in multiple threads.
16190 Access to the large state is not always serializable because
16191 the additional locking could be a performance killer, and
16192 md should be large enough anyway.
16196 * New file apps/app_rand.c with commonly needed functionality
16197 for handling the random seed file.
16199 Use the random seed file in some applications that previously did not:
16201 dsaparam -genkey (which also ignored its '-rand' option),
16204 x509 (when signing).
16205 Except on systems with /dev/urandom, it is crucial to have a random
16206 seed file at least for key creation, DSA signing, and for DH exchanges;
16207 for RSA signatures we could do without one.
16209 gendh and gendsa (unlike genrsa) used to read only the first byte
16210 of each file listed in the '-rand' option. The function as previously
16211 found in genrsa is now in app_rand.c and is used by all programs
16212 that support '-rand'.
16216 * In RAND_write_file, use mode 0600 for creating files;
16217 don't just chmod when it may be too late.
16221 * Report an error from X509_STORE_load_locations
16222 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
16226 * New function ASN1_mbstring_copy() this copies a string in either
16227 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
16228 into an ASN1_STRING type. A mask of permissible types is passed
16229 and it chooses the "minimal" type to use or an error if not type
16234 * Add function equivalents to the various macros in asn1.h. The old
16235 macros are retained with an `M_` prefix. Code inside the library can
16236 use the `M_` macros. External code (including the openssl utility)
16237 should *NOT* in order to be "shared library friendly".
16241 * Add various functions that can check a certificate's extensions
16242 to see if it usable for various purposes such as SSL client,
16243 server or S/MIME and CAs of these types. This is currently
16244 VERY EXPERIMENTAL but will ultimately be used for certificate chain
16245 verification. Also added a -purpose flag to x509 utility to
16246 print out all the purposes.
16250 * Add a CRYPTO_EX_DATA to X509 certificate structure and associated
16255 * New `X509V3_{X509,CRL,REVOKED}_get_d2i()` functions. These will search
16256 for, obtain and decode and extension and obtain its critical flag.
16257 This allows all the necessary extension code to be handled in a
16258 single function call.
16262 * RC4 tune-up featuring 30-40% performance improvement on most RISC
16263 platforms. See crypto/rc4/rc4_enc.c for further details.
16267 * New -noout option to asn1parse. This causes no output to be produced
16268 its main use is when combined with -strparse and -out to extract data
16269 from a file (which may not be in ASN.1 format).
16273 * Fix for pkcs12 program. It was hashing an invalid certificate pointer
16274 when producing the local key id.
16276 *Richard Levitte <levitte@stacken.kth.se>*
16278 * New option -dhparam in s_server. This allows a DH parameter file to be
16279 stated explicitly. If it is not stated then it tries the first server
16280 certificate file. The previous behaviour hard coded the filename
16285 * Add -pubin and -pubout options to the rsa and dsa commands. These allow
16286 a public key to be input or output. For example:
16287 openssl rsa -in key.pem -pubout -out pubkey.pem
16288 Also added necessary DSA public key functions to handle this.
16292 * Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
16293 in the message. This was handled by allowing
16294 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
16296 *Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>*
16298 * Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
16299 to the end of the strings whereas this didn't. This would cause problems
16300 if strings read with d2i_ASN1_bytes() were later modified.
16302 *Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>*
16304 * Fix for base64 decode bug. When a base64 bio reads only one line of
16305 data and it contains EOF it will end up returning an error. This is
16306 caused by input 46 bytes long. The cause is due to the way base64
16307 BIOs find the start of base64 encoded data. They do this by trying a
16308 trial decode on each line until they find one that works. When they
16309 do a flag is set and it starts again knowing it can pass all the
16310 data directly through the decoder. Unfortunately it doesn't reset
16311 the context it uses. This means that if EOF is reached an attempt
16312 is made to pass two EOFs through the context and this causes the
16313 resulting error. This can also cause other problems as well. As is
16314 usual with these problems it takes *ages* to find and the fix is
16315 trivial: move one line.
16317 *Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer)*
16319 * Ugly workaround to get s_client and s_server working under Windows. The
16320 old code wouldn't work because it needed to select() on sockets and the
16321 tty (for keypresses and to see if data could be written). Win32 only
16322 supports select() on sockets so we select() with a 1s timeout on the
16323 sockets and then see if any characters are waiting to be read, if none
16324 are present then we retry, we also assume we can always write data to
16325 the tty. This isn't nice because the code then blocks until we've
16326 received a complete line of data and it is effectively polling the
16327 keyboard at 1s intervals: however it's quite a bit better than not
16328 working at all :-) A dedicated Windows application might handle this
16329 with an event loop for example.
16333 * Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
16334 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
16335 will be called when RSA_sign() and RSA_verify() are used. This is useful
16336 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
16337 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
16338 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
16339 This necessitated the support of an extra signature type NID_md5_sha1
16340 for SSL signatures and modifications to the SSL library to use it instead
16341 of calling RSA_public_decrypt() and RSA_private_encrypt().
16345 * Add new -verify -CAfile and -CApath options to the crl program, these
16346 will lookup a CRL issuers certificate and verify the signature in a
16347 similar way to the verify program. Tidy up the crl program so it
16348 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
16349 less strict. It will now permit CRL extensions even if it is not
16350 a V2 CRL: this will allow it to tolerate some broken CRLs.
16354 * Initialize all non-automatic variables each time one of the openssl
16355 sub-programs is started (this is necessary as they may be started
16356 multiple times from the "OpenSSL>" prompt).
16358 *Lennart Bang, Bodo Moeller*
16360 * Preliminary compilation option RSA_NULL which disables RSA crypto without
16361 removing all other RSA functionality (this is what NO_RSA does). This
16362 is so (for example) those in the US can disable those operations covered
16363 by the RSA patent while allowing storage and parsing of RSA keys and RSA
16368 * Non-copying interface to BIO pairs.
16369 (still largely untested)
16373 * New function ASN1_tag2str() to convert an ASN1 tag to a descriptive
16374 ASCII string. This was handled independently in various places before.
16378 * New functions UTF8_getc() and UTF8_putc() that parse and generate
16379 UTF8 strings a character at a time.
16383 * Use client_version from client hello to select the protocol
16384 (s23_srvr.c) and for RSA client key exchange verification
16385 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
16389 * Add various utility functions to handle SPKACs, these were previously
16390 handled by poking round in the structure internals. Added new function
16391 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
16392 print, verify and generate SPKACs. Based on an original idea from
16393 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
16397 * RIPEMD160 is operational on all platforms and is back in 'make test'.
16401 * Allow the config file extension section to be overwritten on the
16402 command line. Based on an original idea from Massimiliano Pala
16403 <madwolf@comune.modena.it>. The new option is called -extensions
16404 and can be applied to ca, req and x509. Also -reqexts to override
16405 the request extensions in req and -crlexts to override the crl extensions
16410 * Add new feature to the SPKAC handling in ca. Now you can include
16411 the same field multiple times by preceding it by "XXXX." for example:
16414 this is the same syntax as used in the req config file.
16418 * Allow certificate extensions to be added to certificate requests. These
16419 are specified in a 'req_extensions' option of the req section of the
16420 config file. They can be printed out with the -text option to req but
16421 are otherwise ignored at present.
16425 * Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
16426 data read consists of only the final block it would not decrypted because
16427 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
16428 A misplaced 'break' also meant the decrypted final block might not be
16429 copied until the next read.
16433 * Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
16434 a few extra parameters to the DH structure: these will be useful if
16435 for example we want the value of 'q' or implement X9.42 DH.
16439 * Initial support for DSA_METHOD. This is based on the RSA_METHOD and
16440 provides hooks that allow the default DSA functions or functions on a
16441 "per key" basis to be replaced. This allows hardware acceleration and
16442 hardware key storage to be handled without major modification to the
16443 library. Also added low level modexp hooks and CRYPTO_EX structure and
16444 associated functions.
16448 * Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
16449 as "read only": it can't be written to and the buffer it points to will
16450 not be freed. Reading from a read only BIO is much more efficient than
16451 a normal memory BIO. This was added because there are several times when
16452 an area of memory needs to be read from a BIO. The previous method was
16453 to create a memory BIO and write the data to it, this results in two
16454 copies of the data and an O(n^2) reading algorithm. There is a new
16455 function BIO_new_mem_buf() which creates a read only memory BIO from
16456 an area of memory. Also modified the PKCS#7 routines to use read only
16461 * Bugfix: ssl23_get_client_hello did not work properly when called in
16462 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
16463 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
16464 but a retry condition occurred while trying to read the rest.
16468 * The PKCS7_ENC_CONTENT_new() function was setting the content type as
16469 NID_pkcs7_encrypted by default: this was wrong since this should almost
16470 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
16471 the encrypted data type: this is a more sensible place to put it and it
16472 allows the PKCS#12 code to be tidied up that duplicated this
16477 * Changed obj_dat.pl script so it takes its input and output files on
16478 the command line. This should avoid shell escape redirection problems
16483 * Initial support for certificate extension requests, these are included
16484 in things like Xenroll certificate requests. Included functions to allow
16485 extensions to be obtained and added.
16489 * -crlf option to s_client and s_server for sending newlines as
16490 CRLF (as required by many protocols).
16494 ### Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
16496 * Install libRSAglue.a when OpenSSL is built with RSAref.
16498 *Ralf S. Engelschall*
16500 * A few more `#ifndef NO_FP_API / #endif` pairs for consistency.
16502 *Andrija Antonijevic <TheAntony2@bigfoot.com>*
16504 * Fix -startdate and -enddate (which was missing) arguments to 'ca'
16509 * New function DSA_dup_DH, which duplicates DSA parameters/keys as
16510 DH parameters/keys (q is lost during that conversion, but the resulting
16511 DH parameters contain its length).
16513 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
16514 much faster than DH_generate_parameters (which creates parameters
16515 where `p = 2*q + 1`), and also the smaller q makes DH computations
16516 much more efficient (160-bit exponentiation instead of 1024-bit
16517 exponentiation); so this provides a convenient way to support DHE
16518 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
16519 utter importance to use
16520 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
16522 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
16523 when such DH parameters are used, because otherwise small subgroup
16524 attacks may become possible!
16528 * Avoid memory leak in i2d_DHparams.
16532 * Allow the -k option to be used more than once in the enc program:
16533 this allows the same encrypted message to be read by multiple recipients.
16537 * New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
16538 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
16539 it will always use the numerical form of the OID, even if it has a short
16544 * Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
16545 method only got called if p,q,dmp1,dmq1,iqmp components were present,
16546 otherwise bn_mod_exp was called. In the case of hardware keys for example
16547 no private key components need be present and it might store extra data
16548 in the RSA structure, which cannot be accessed from bn_mod_exp.
16549 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
16550 private key operations.
16554 * Added support for SPARC Linux.
16558 * pem_password_cb function type incompatibly changed from
16559 typedef int pem_password_cb(char *buf, int size, int rwflag);
16561 ....(char *buf, int size, int rwflag, void *userdata);
16562 so that applications can pass data to their callbacks:
16563 The `PEM[_ASN1]_{read,write}...` functions and macros now take an
16564 additional void * argument, which is just handed through whenever
16565 the password callback is called.
16567 *Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller*
16569 New function SSL_CTX_set_default_passwd_cb_userdata.
16571 Compatibility note: As many C implementations push function arguments
16572 onto the stack in reverse order, the new library version is likely to
16573 interoperate with programs that have been compiled with the old
16574 pem_password_cb definition (PEM_whatever takes some data that
16575 happens to be on the stack as its last argument, and the callback
16576 just ignores this garbage); but there is no guarantee whatsoever that
16579 * The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
16580 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
16581 problems not only on Windows, but also on some Unix platforms.
16582 To avoid problematic command lines, these definitions are now in an
16583 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
16584 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
16588 * MIPS III/IV assembler module is reimplemented.
16592 * More DES library cleanups: remove references to srand/rand and
16593 delete an unused file.
16597 * Add support for the free Netwide assembler (NASM) under Win32,
16598 since not many people have MASM (ml) and it can be hard to obtain.
16599 This is currently experimental but it seems to work OK and pass all
16600 the tests. Check out INSTALL.W32 for info.
16604 * Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
16605 without temporary keys kept an extra copy of the server key,
16606 and connections with temporary keys did not free everything in case
16611 * New function RSA_check_key and new openssl rsa option -check
16612 for verifying the consistency of RSA keys.
16614 *Ulf Moeller, Bodo Moeller*
16616 * Various changes to make Win32 compile work:
16617 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
16618 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
16619 comparison" warnings.
16620 3. Add `sk_<TYPE>_sort` to DEF file generator and do make update.
16624 * Add a debugging option to PKCS#5 v2 key generation function: when
16625 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
16626 derived keys are printed to stderr.
16630 * Copy the flags in ASN1_STRING_dup().
16632 *Roman E. Pavlov <pre@mo.msk.ru>*
16634 * The x509 application mishandled signing requests containing DSA
16635 keys when the signing key was also DSA and the parameters didn't match.
16637 It was supposed to omit the parameters when they matched the signing key:
16638 the verifying software was then supposed to automatically use the CA's
16639 parameters if they were absent from the end user certificate.
16641 Omitting parameters is no longer recommended. The test was also
16642 the wrong way round! This was probably due to unusual behaviour in
16643 EVP_cmp_parameters() which returns 1 if the parameters match.
16644 This meant that parameters were omitted when they *didn't* match and
16645 the certificate was useless. Certificates signed with 'ca' didn't have
16648 *Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>*
16650 * Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
16651 The interface is as follows:
16652 Applications can use
16653 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
16654 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
16655 "off" is now the default.
16656 The library internally uses
16657 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
16658 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
16659 to disable memory-checking temporarily.
16661 Some inconsistent states that previously were possible (and were
16662 even the default) are now avoided.
16664 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
16665 with each memory chunk allocated; this is occasionally more helpful
16666 than just having a counter.
16668 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
16670 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
16675 * Introduce "mode" for SSL structures (with defaults in SSL_CTX),
16676 which largely parallels "options", but is for changing API behaviour,
16677 whereas "options" are about protocol behaviour.
16678 Initial "mode" flags are:
16680 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
16681 a single record has been written.
16682 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
16683 retries use the same buffer location.
16684 (But all of the contents must be
16689 * Bugfix: SSL_set_options ignored its parameter, only SSL_CTX_set_options
16692 * Fix problems with no-hmac etc.
16694 *Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>*
16696 * New functions RSA_get_default_method(), RSA_set_method() and
16697 RSA_get_method(). These allows replacement of RSA_METHODs without having
16698 to mess around with the internals of an RSA structure.
16702 * Fix memory leaks in DSA_do_sign and DSA_is_prime.
16703 Also really enable memory leak checks in openssl.c and in some
16706 *Chad C. Mulligan, Bodo Moeller*
16708 * Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
16709 up the length of negative integers. This has now been simplified to just
16710 store the length when it is first determined and use it later, rather
16711 than trying to keep track of where data is copied and updating it to
16713 *Steve Henson, reported by Brien Wheeler <bwheeler@authentica-security.com>*
16715 * Add a new function PKCS7_signatureVerify. This allows the verification
16716 of a PKCS#7 signature but with the signing certificate passed to the
16717 function itself. This contrasts with PKCS7_dataVerify which assumes the
16718 certificate is present in the PKCS#7 structure. This isn't always the
16719 case: certificates can be omitted from a PKCS#7 structure and be
16720 distributed by "out of band" means (such as a certificate database).
16724 * Complete the `PEM_*` macros with DECLARE_PEM versions to replace the
16725 function prototypes in pem.h, also change util/mkdef.pl to add the
16726 necessary function names.
16730 * mk1mf.pl (used by Windows builds) did not properly read the
16731 options set by Configure in the top level Makefile, and Configure
16732 was not even able to write more than one option correctly.
16733 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
16737 * New functions CONF_load_bio() and CONF_load_fp() to allow a config
16738 file to be loaded from a BIO or FILE pointer. The BIO version will
16739 for example allow memory BIOs to contain config info.
16743 * New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
16744 Whoever hopes to achieve shared-library compatibility across versions
16745 must use this, not the compile-time macro.
16746 (Exercise 0.9.4: Which is the minimum library version required by
16748 Note: All this applies only to multi-threaded programs, others don't
16753 * Add missing case to s3_clnt.c state machine -- one of the new SSL tests
16754 through a BIO pair triggered the default case, i.e.
16755 SSLerr(...,SSL_R_UNKNOWN_STATE).
16759 * New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
16760 can use the SSL library even if none of the specific BIOs is
16765 * Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
16766 for the encoded length.
16768 *Jeon KyoungHo <khjeon@sds.samsung.co.kr>*
16770 * Add initial documentation of the X509V3 functions.
16774 * Add a new pair of functions PEM_write_PKCS8PrivateKey() and
16775 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
16776 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
16777 secure PKCS#8 private key format with a high iteration count.
16781 * Fix determination of Perl interpreter: A perl or perl5
16782 *directory* in $PATH was also accepted as the interpreter.
16784 *Ralf S. Engelschall*
16786 * Fix demos/sign/sign.c: well there wasn't anything strictly speaking
16787 wrong with it but it was very old and did things like calling
16788 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
16789 unusual formatting.
16793 * Fix demos/selfsign.c: it used obsolete and deleted functions, changed
16794 to use the new extension code.
16798 * Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
16799 with macros. This should make it easier to change their form, add extra
16800 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
16805 * Add to configuration table a new entry that can specify an alternative
16806 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
16807 according to Mark Crispin <MRC@Panda.COM>.
16811 * DES CBC did not update the IV. Weird.
16815 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
16816 Changing the behaviour of the former might break existing programs --
16817 where IV updating is needed, des_ncbc_encrypt can be used.
16820 * When bntest is run from "make test" it drives bc to check its
16821 calculations, as well as internally checking them. If an internal check
16822 fails, it needs to cause bc to give a non-zero result or make test carries
16823 on without noticing the failure. Fixed.
16827 * DES library cleanups.
16831 * Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
16832 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
16833 ciphers. NOTE: although the key derivation function has been verified
16834 against some published test vectors it has not been extensively tested
16835 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
16840 * Instead of "mkdir -p", which is not fully portable, use new
16841 Perl script "util/mkdir-p.pl".
16845 * Rewrite the way password based encryption (PBE) is handled. It used to
16846 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
16847 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
16848 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
16849 the 'parameter' field of the AlgorithmIdentifier is passed to the
16850 underlying key generation function so it must do its own ASN1 parsing.
16851 This has also changed the EVP_PBE_CipherInit() function which now has a
16852 'parameter' argument instead of literal salt and iteration count values
16853 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
16857 * Support for PKCS#5 v1.5 compatible password based encryption algorithms
16858 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
16859 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
16860 KEY" because this clashed with PKCS#8 unencrypted string. Since this
16861 value was just used as a "magic string" and not used directly its
16862 value doesn't matter.
16866 * Introduce some semblance of const correctness to BN. Shame C doesn't
16871 * "linux-sparc64" configuration (ultrapenguin).
16873 *Ray Miller <ray.miller@oucs.ox.ac.uk>*
16874 "linux-sparc" configuration.
16876 *Christian Forster <fo@hawo.stw.uni-erlangen.de>*
16878 * config now generates no-xxx options for missing ciphers.
16882 * Support the EBCDIC character set (work in progress).
16883 File ebcdic.c not yet included because it has a different license.
16885 *Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>*
16887 * Support BS2000/OSD-POSIX.
16889 *Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>*
16891 * Make callbacks for key generation use `void *` instead of `char *`.
16895 * Make S/MIME samples compile (not yet tested).
16899 * Additional typesafe stacks.
16903 * New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
16907 ### Changes between 0.9.3 and 0.9.3a [29 May 1999]
16909 * New configuration variant "sco5-gcc".
16911 * Updated some demos.
16913 *Sean O Riordain, Wade Scholine*
16915 * Add missing BIO_free at exit of pkcs12 application.
16919 * Fix memory leak in conf.c.
16923 * Updates for Win32 to assembler version of MD5.
16927 * Set #! path to perl in apps/der_chop to where we found it
16928 instead of using a fixed path.
16932 * SHA library changes for irix64-mips4-cc.
16936 * Improvements for VMS support.
16940 ### Changes between 0.9.2b and 0.9.3 [24 May 1999]
16942 * Bignum library bug fix. IRIX 6 passes "make test" now!
16943 This also avoids the problems with SC4.2 and unpatched SC5.
16945 *Andy Polyakov <appro@fy.chalmers.se>*
16947 * New functions sk_num, sk_value and sk_set to replace the previous macros.
16948 These are required because of the typesafe stack would otherwise break
16949 existing code. If old code used a structure member which used to be STACK
16950 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
16951 sk_num or sk_value it would produce an error because the num, data members
16952 are not present in STACK_OF. Now it just produces a warning. sk_set
16953 replaces the old method of assigning a value to sk_value
16954 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
16955 that does this will no longer work (and should use sk_set instead) but
16956 this could be regarded as a "questionable" behaviour anyway.
16960 * Fix most of the other PKCS#7 bugs. The "experimental" code can now
16961 correctly handle encrypted S/MIME data.
16965 * Change type of various DES function arguments from des_cblock
16966 (which means, in function argument declarations, pointer to char)
16967 to des_cblock * (meaning pointer to array with 8 char elements),
16968 which allows the compiler to do more typechecking; it was like
16969 that back in SSLeay, but with lots of ugly casts.
16971 Introduce new type const_des_cblock.
16975 * Reorganise the PKCS#7 library and get rid of some of the more obvious
16976 problems: find RecipientInfo structure that matches recipient certificate
16977 and initialise the ASN1 structures properly based on passed cipher.
16981 * Belatedly make the BN tests actually check the results.
16985 * Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
16986 to and from BNs: it was completely broken. New compilation option
16987 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
16988 key elements as negative integers.
16992 * Reorganize and speed up MD5.
16994 *Andy Polyakov <appro@fy.chalmers.se>*
16998 *Richard Levitte <richard@levitte.org>*
17000 * New option -out to asn1parse to allow the parsed structure to be
17001 output to a file. This is most useful when combined with the -strparse
17002 option to examine the output of things like OCTET STRINGS.
17006 * Make SSL library a little more fool-proof by not requiring any longer
17007 that `SSL_set_{accept,connect}_state` be called before
17008 `SSL_{accept,connect}` may be used (`SSL_set_..._state` is omitted
17009 in many applications because usually everything *appeared* to work as
17010 intended anyway -- now it really works as intended).
17014 * Move openssl.cnf out of lib/.
17018 * Fix various things to let OpenSSL even pass "egcc -pipe -O2 -Wall
17019 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
17020 -Wmissing-declarations -Wnested-externs -Winline" with EGCS 1.1.2+
17022 *Ralf S. Engelschall*
17024 * Various fixes to the EVP and PKCS#7 code. It may now be able to
17025 handle PKCS#7 enveloped data properly.
17027 *Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve*
17029 * Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
17030 copying pointers. The cert_st handling is changed by this in
17031 various ways (and thus what used to be known as ctx->default_cert
17032 is now called ctx->cert, since we don't resort to `s->ctx->[default_]cert`
17033 any longer when s->cert does not give us what we need).
17034 ssl_cert_instantiate becomes obsolete by this change.
17035 As soon as we've got the new code right (possibly it already is?),
17036 we have solved a couple of bugs of the earlier code where s->cert
17037 was used as if it could not have been shared with other SSL structures.
17039 Note that using the SSL API in certain dirty ways now will result
17040 in different behaviour than observed with earlier library versions:
17041 Changing settings for an `SSL_CTX *ctx` after having done s = SSL_new(ctx)
17042 does not influence s as it used to.
17044 In order to clean up things more thoroughly, inside SSL_SESSION
17045 we don't use CERT any longer, but a new structure SESS_CERT
17046 that holds per-session data (if available); currently, this is
17047 the peer's certificate chain and, for clients, the server's certificate
17048 and temporary key. CERT holds only those values that can have
17049 meaningful defaults in an SSL_CTX.
17053 * New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
17054 from the internal representation. Various PKCS#7 fixes: remove some
17055 evil casts and set the enc_dig_alg field properly based on the signing
17060 * Allow PKCS#12 password to be set from the command line or the
17061 environment. Let 'ca' get its config file name from the environment
17062 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
17067 * Allow certificate policies extension to use an IA5STRING for the
17068 organization field. This is contrary to the PKIX definition but
17069 VeriSign uses it and IE5 only recognises this form. Document 'x509'
17074 * Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
17075 without disallowing inline assembler and the like for non-pedantic builds.
17079 * Support Borland C++ builder.
17081 *Janez Jere <jj@void.si>, modified by Ulf Möller*
17087 * SHA-1 cleanups and performance enhancements.
17089 *Andy Polyakov <appro@fy.chalmers.se>*
17091 * Sparc v8plus assembler for the bignum library.
17093 *Andy Polyakov <appro@fy.chalmers.se>*
17095 * Accept any -xxx and +xxx compiler options in Configure.
17099 * Update HPUX configuration.
17103 * Add missing `sk_<type>_unshift()` function to safestack.h
17105 *Ralf S. Engelschall*
17107 * New function SSL_CTX_use_certificate_chain_file that sets the
17108 "extra_cert"s in addition to the certificate. (This makes sense
17109 only for "PEM" format files, as chains as a whole are not
17114 * Support verify_depth from the SSL API.
17115 x509_vfy.c had what can be considered an off-by-one-error:
17116 Its depth (which was not part of the external interface)
17117 was actually counting the number of certificates in a chain;
17118 now it really counts the depth.
17122 * Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
17123 instead of X509err, which often resulted in confusing error
17124 messages since the error codes are not globally unique
17125 (e.g. an alleged error in ssl3_accept when a certificate
17126 didn't match the private key).
17128 * New function SSL_CTX_set_session_id_context that allows to set a default
17129 value (so that you don't need SSL_set_session_id_context for each
17130 connection using the SSL_CTX).
17134 * OAEP decoding bug fix.
17138 * Support INSTALL_PREFIX for package builders, as proposed by
17143 * New Configure options "threads" and "no-threads". For systems
17144 where the proper compiler options are known (currently Solaris
17145 and Linux), "threads" is the default.
17149 * New script util/mklink.pl as a faster substitute for util/mklink.sh.
17153 * Install various scripts to $(OPENSSLDIR)/misc, not to
17154 $(INSTALLTOP)/bin -- they shouldn't clutter directories
17155 such as /usr/local/bin.
17159 * "make linux-shared" to build shared libraries.
17161 *Niels Poppe <niels@netbox.org>*
17163 * New Configure option `no-<cipher>` (rsa, idea, rc5, ...).
17167 * Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
17168 extension adding in x509 utility.
17172 * Remove NOPROTO sections and error code comments.
17176 * Partial rewrite of the DEF file generator to now parse the ANSI
17181 * New Configure options --prefix=DIR and --openssldir=DIR.
17185 * Complete rewrite of the error code script(s). It is all now handled
17186 by one script at the top level which handles error code gathering,
17187 header rewriting and C source file generation. It should be much better
17188 than the old method: it now uses a modified version of Ulf's parser to
17189 read the ANSI prototypes in all header files (thus the old K&R definitions
17190 aren't needed for error creation any more) and do a better job of
17191 translating function codes into names. The old 'ASN1 error code embedded
17192 in a comment' is no longer necessary and it doesn't use .err files which
17193 have now been deleted. Also the error code call doesn't have to appear all
17194 on one line (which resulted in some large lines...).
17198 * Change #include filenames from `<foo.h>` to `<openssl/foo.h>`.
17202 * Change behaviour of ssl2_read when facing length-0 packets: Don't return
17203 0 (which usually indicates a closed connection), but continue reading.
17207 * Fix some race conditions.
17211 * Add support for CRL distribution points extension. Add Certificate
17212 Policies and CRL distribution points documentation.
17216 * Move the autogenerated header file parts to crypto/opensslconf.h.
17220 * Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
17221 8 of keying material. Merlin has also confirmed interop with this fix
17222 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
17224 *Merlin Hughes <merlin@baltimore.ie>*
17226 * Fix lots of warnings.
17228 *Richard Levitte <levitte@stacken.kth.se>*
17230 * In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
17231 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
17233 *Richard Levitte <levitte@stacken.kth.se>*
17235 * Fix problems with sizeof(long) == 8.
17237 *Andy Polyakov <appro@fy.chalmers.se>*
17239 * Change functions to ANSI C.
17243 * Fix typos in error codes.
17245 *Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller*
17247 * Remove defunct assembler files from Configure.
17251 * SPARC v8 assembler BIGNUM implementation.
17253 *Andy Polyakov <appro@fy.chalmers.se>*
17255 * Support for Certificate Policies extension: both print and set.
17256 Various additions to support the r2i method this uses.
17260 * A lot of constification, and fix a bug in X509_NAME_oneline() that could
17261 return a const string when you are expecting an allocated buffer.
17265 * Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
17266 types DirectoryString and DisplayText.
17270 * Add code to allow r2i extensions to access the configuration database,
17271 add an LHASH database driver and add several ctx helper functions.
17275 * Fix an evil bug in bn_expand2() which caused various BN functions to
17276 fail when they extended the size of a BIGNUM.
17280 * Various utility functions to handle SXNet extension. Modify mkdef.pl to
17281 support typesafe stack.
17285 * Fix typo in SSL_[gs]et_options().
17287 *Nils Frostberg <nils@medcom.se>*
17289 * Delete various functions and files that belonged to the (now obsolete)
17290 old X509V3 handling code.
17294 * New Configure option "rsaref".
17298 * Don't auto-generate pem.h.
17302 * Introduce type-safe ASN.1 SETs.
17306 * Convert various additional casted stacks to type-safe STACK_OF() variants.
17308 *Ben Laurie, Ralf S. Engelschall, Steve Henson*
17310 * Introduce type-safe STACKs. This will almost certainly break lots of code
17311 that links with OpenSSL (well at least cause lots of warnings), but fear
17312 not: the conversion is trivial, and it eliminates loads of evil casts. A
17313 few STACKed things have been converted already. Feel free to convert more.
17314 In the fullness of time, I'll do away with the STACK type altogether.
17318 * Add `openssl ca -revoke <certfile>` facility which revokes a certificate
17319 specified in `<certfile>` by updating the entry in the index.txt file.
17320 This way one no longer has to edit the index.txt file manually for
17321 revoking a certificate. The -revoke option does the gory details now.
17323 *Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall*
17325 * Fix `openssl crl -noout -text` combination where `-noout` killed the
17326 `-text` option at all and this way the `-noout -text` combination was
17327 inconsistent in `openssl crl` with the friends in `openssl x509|rsa|dsa`.
17329 *Ralf S. Engelschall*
17331 * Make sure a corresponding plain text error message exists for the
17332 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
17333 verify callback function determined that a certificate was revoked.
17335 *Ralf S. Engelschall*
17337 * Bugfix: In test/testenc, don't test `openssl <cipher>` for
17338 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
17339 all available ciphers including rc5, which was forgotten until now.
17340 In order to let the testing shell script know which algorithms
17341 are available, a new (up to now undocumented) command
17342 `openssl list-cipher-commands` is used.
17346 * Bugfix: s_client occasionally would sleep in select() when
17347 it should have checked SSL_pending() first.
17351 * New functions DSA_do_sign and DSA_do_verify to provide access to
17352 the raw DSA values prior to ASN.1 encoding.
17356 * Tweaks to Configure
17358 *Niels Poppe <niels@netbox.org>*
17360 * Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
17365 * New variables $(RANLIB) and $(PERL) in the Makefiles.
17369 * New config option to avoid instructions that are illegal on the 80386.
17370 The default code is faster, but requires at least a 486.
17374 * Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
17375 SSL2_SERVER_VERSION (not used at all) macros, which are now the
17376 same as SSL2_VERSION anyway.
17380 * New "-showcerts" option for s_client.
17384 * Still more PKCS#12 integration. Add pkcs12 application to openssl
17385 application. Various cleanups and fixes.
17389 * More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
17390 modify error routines to work internally. Add error codes and PBE init
17391 to library startup routines.
17395 * Further PKCS#12 integration. Added password based encryption, PKCS#8 and
17396 packing functions to asn1 and evp. Changed function names and error
17397 codes along the way.
17401 * PKCS12 integration: and so it begins... First of several patches to
17402 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
17403 objects to objects.h
17407 * Add a new 'indent' option to some X509V3 extension code. Initial ASN1
17408 and display support for Thawte strong extranet extension.
17412 * Add LinuxPPC support.
17414 *Jeff Dubrule <igor@pobox.org>*
17416 * Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
17417 bn_div_words in alpha.s.
17419 *Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie*
17421 * Make sure the RSA OAEP test is skipped under -DRSAref because
17422 OAEP isn't supported when OpenSSL is built with RSAref.
17424 *Ulf Moeller <ulf@fitug.de>*
17426 * Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
17427 so they no longer are missing under -DNOPROTO.
17429 *Soren S. Jorvang <soren@t.dk>*
17431 ### Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
17433 * Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
17434 doesn't work when the session is reused. Coming soon!
17438 * Fix a security hole, that allows sessions to be reused in the wrong
17439 context thus bypassing client cert protection! All software that uses
17440 client certs and session caches in multiple contexts NEEDS PATCHING to
17441 allow session reuse! A fuller solution is in the works.
17443 *Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)*
17445 * Some more source tree cleanups (removed obsolete files
17446 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
17447 permission on "config" script to be executable) and a fix for the INSTALL
17450 *Ulf Moeller <ulf@fitug.de>*
17452 * Remove some legacy and erroneous uses of malloc, free instead of
17455 *Lennart Bang <lob@netstream.se>, with minor changes by Steve*
17457 * Make rsa_oaep_test return non-zero on error.
17459 *Ulf Moeller <ulf@fitug.de>*
17461 * Add support for native Solaris shared libraries. Configure
17462 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
17463 if someone would make that last step automatic.
17465 *Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>*
17467 * ctx_size was not built with the right compiler during "make links". Fixed.
17471 * Change the meaning of 'ALL' in the cipher list. It now means "everything
17472 except NULL ciphers". This means the default cipher list will no longer
17473 enable NULL ciphers. They need to be specifically enabled e.g. with
17474 the string "DEFAULT:eNULL".
17478 * Fix to RSA private encryption routines: if p < q then it would
17479 occasionally produce an invalid result. This will only happen with
17480 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
17484 * Be less restrictive and allow also `perl util/perlpath.pl
17485 /path/to/bin/perl` in addition to `perl util/perlpath.pl /path/to/bin`,
17486 because this way one can also use an interpreter named `perl5` (which is
17487 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
17488 installed as `perl`).
17490 *Matthias Loepfe <Matthias.Loepfe@adnovum.ch>*
17492 * Let util/clean-depend.pl work also with older Perl 5.00x versions.
17494 *Matthias Loepfe <Matthias.Loepfe@adnovum.ch>*
17496 * Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
17497 advapi32.lib to Win32 build and change the pem test comparison
17498 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
17499 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
17500 and crypto/des/ede_cbcm_enc.c.
17504 * DES quad checksum was broken on big-endian architectures. Fixed.
17508 * Comment out two functions in bio.h that aren't implemented. Fix up the
17509 Win32 test batch file so it (might) work again. The Win32 test batch file
17510 is horrible: I feel ill....
17514 * Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
17515 in e_os.h. Audit of header files to check ANSI and non ANSI
17516 sections: 10 functions were absent from non ANSI section and not exported
17517 from Windows DLLs. Fixed up libeay.num for new functions.
17521 * Make `openssl version` output lines consistent.
17523 *Ralf S. Engelschall*
17525 * Fix Win32 symbol export lists for BIO functions: Added
17526 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
17527 to ms/libeay{16,32}.def.
17529 *Ralf S. Engelschall*
17531 * Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
17532 fine under Unix and passes some trivial tests I've now added. But the
17533 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
17534 added to make sure no one expects that this stuff really works in the
17535 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
17536 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
17539 *Ralf S. Engelschall*
17541 * Fix the generation of two part addresses in perl.
17543 *Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie*
17545 * Add config entry for Linux on MIPS.
17547 *John Tobey <jtobey@channel1.com>*
17549 * Make links whenever Configure is run, unless we are on Windoze.
17553 * Permit extensions to be added to CRLs using crl_section in openssl.cnf.
17554 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
17559 * Add a useful kludge to allow package maintainers to specify compiler and
17560 other platforms details on the command line without having to patch the
17561 Configure script every time: One now can use
17562 `perl Configure <id>:<details>`,
17563 i.e. platform ids are allowed to have details appended
17564 to them (separated by colons). This is treated as there would be a static
17565 pre-configured entry in Configure's %table under key `<id>` with value
17566 `<details>` and `perl Configure <id>` is called. So, when you want to
17567 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
17568 assembler stuff you can use `perl Configure "FreeBSD-elf:pgcc:-O6:::"`
17569 now, which overrides the FreeBSD-elf entry on-the-fly.
17571 *Ralf S. Engelschall*
17573 * Disable new TLS1 ciphersuites by default: they aren't official yet.
17577 * Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
17578 on the `perl Configure ...` command line. This way one can compile
17579 OpenSSL libraries with Position Independent Code (PIC) which is needed
17580 for linking it into DSOs.
17582 *Ralf S. Engelschall*
17584 * Remarkably, export ciphers were totally broken and no-one had noticed!
17589 * Cleaned up the LICENSE document: The official contact for any license
17590 questions now is the OpenSSL core team under openssl-core@openssl.org.
17591 And add a paragraph about the dual-license situation to make sure people
17592 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
17593 to the OpenSSL toolkit.
17595 *Ralf S. Engelschall*
17597 * General source tree makefile cleanups: Made `making xxx in yyy...`
17598 display consistent in the source tree and replaced `/bin/rm` by `rm`.
17599 Additionally cleaned up the `make links` target: Remove unnecessary
17600 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
17601 to speed processing and no longer clutter the display with confusing
17602 stuff. Instead only the actually done links are displayed.
17604 *Ralf S. Engelschall*
17606 * Permit null encryption ciphersuites, used for authentication only. It used
17607 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
17608 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
17613 * Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
17614 signed attributes when verifying signatures (this would break them),
17615 the detached data encoding was wrong and public keys obtained using
17616 X509_get_pubkey() weren't freed.
17620 * Add text documentation for the BUFFER functions. Also added a work around
17621 to a Win95 console bug. This was triggered by the password read stuff: the
17622 last character typed gets carried over to the next fread(). If you were
17623 generating a new cert request using 'req' for example then the last
17624 character of the passphrase would be CR which would then enter the first
17629 * Added the new 'Includes OpenSSL Cryptography Software' button as
17630 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
17631 button and can be used by applications based on OpenSSL to show the
17632 relationship to the OpenSSL project.
17634 *Ralf S. Engelschall*
17636 * Remove confusing variables in function signatures in files
17637 ssl/ssl_lib.c and ssl/ssl.h.
17639 *Lennart Bong <lob@kulthea.stacken.kth.se>*
17641 * Don't install bss_file.c under PREFIX/include/
17643 *Lennart Bong <lob@kulthea.stacken.kth.se>*
17645 * Get the Win32 compile working again. Modify mkdef.pl so it can handle
17646 functions that return function pointers and has support for NT specific
17647 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
17648 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
17649 unsigned to signed types: this was killing the Win32 compile.
17653 * Add new certificate file to stack functions,
17654 SSL_add_dir_cert_subjects_to_stack() and
17655 SSL_add_file_cert_subjects_to_stack(). These largely supplant
17656 SSL_load_client_CA_file(), and can be used to add multiple certs easily
17657 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
17658 This means that Apache-SSL and similar packages don't have to mess around
17659 to add as many CAs as they want to the preferred list.
17663 * Experiment with doxygen documentation. Currently only partially applied to
17665 See <http://www.stack.nl/~dimitri/doxygen/index.html>, and run doxygen with
17666 openssl.doxy as the configuration file.
17670 * Get rid of remaining C++-style comments which strict C compilers hate.
17672 *Ralf S. Engelschall, pointed out by Carlos Amengual*
17674 * Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
17675 compiled in by default: it has problems with large keys.
17679 * Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
17680 DH private keys and/or callback functions which directly correspond to
17681 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
17682 is needed for applications which have to configure certificates on a
17683 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
17685 For the RSA certificate situation is makes no difference, but
17686 for the DSA certificate situation this fixes the "no shared cipher"
17687 problem where the OpenSSL cipher selection procedure failed because the
17688 temporary keys were not overtaken from the context and the API provided
17689 no way to reconfigure them.
17690 The new functions now let applications reconfigure the stuff and they
17691 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
17692 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
17693 non-public-API function ssl_cert_instantiate() is used as a helper
17694 function and also to reduce code redundancy inside ssl_rsa.c.
17696 *Ralf S. Engelschall*
17698 * Move s_server -dcert and -dkey options out of the undocumented feature
17699 area because they are useful for the DSA situation and should be
17700 recognized by the users.
17702 *Ralf S. Engelschall*
17704 * Fix the cipher decision scheme for export ciphers: the export bits are
17705 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
17706 SSL_EXP_MASK. So, the original variable has to be used instead of the
17707 already masked variable.
17709 *Richard Levitte <levitte@stacken.kth.se>*
17711 * Fix `port` variable from `int` to `unsigned int` in crypto/bio/b_sock.c
17713 *Richard Levitte <levitte@stacken.kth.se>*
17715 * Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
17716 from `int` to `unsigned int` because it is a length and initialized by
17717 EVP_DigestFinal() which expects an `unsigned int *`.
17719 *Richard Levitte <levitte@stacken.kth.se>*
17721 * Don't hard-code path to Perl interpreter on shebang line of Configure
17722 script. Instead use the usual Shell->Perl transition trick.
17724 *Ralf S. Engelschall*
17726 * Make `openssl x509 -noout -modulus`' functional also for DSA certificates
17727 (in addition to RSA certificates) to match the behaviour of `openssl dsa
17728 -noout -modulus` as it's already the case for `openssl rsa -noout
17729 -modulus`. For RSA the -modulus is the real "modulus" while for DSA
17730 currently the public key is printed (a decision which was already done by
17731 `openssl dsa -modulus` in the past) which serves a similar purpose.
17732 Additionally the NO_RSA no longer completely removes the whole -modulus
17733 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
17736 *Ralf S. Engelschall*
17738 * Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
17739 BIO. See the source (crypto/evp/bio_ok.c) for more info.
17741 *Arne Ansper <arne@ats.cyber.ee>*
17743 * Dump the old yucky req code that tried (and failed) to allow raw OIDs
17744 to be added. Now both 'req' and 'ca' can use new objects defined in the
17749 * Add cool BIO that does syslog (or event log on NT).
17751 *Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie*
17753 * Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
17754 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
17755 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
17756 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
17760 * Add preliminary config info for new extension code.
17764 * Make RSA_NO_PADDING really use no padding.
17766 *Ulf Moeller <ulf@fitug.de>*
17768 * Generate errors when private/public key check is done.
17772 * Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
17773 for some CRL extensions and new objects added.
17777 * Really fix the ASN1 IMPLICIT bug this time... Partial support for private
17778 key usage extension and fuller support for authority key id.
17782 * Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
17783 padding method for RSA, which is recommended for new applications in PKCS
17784 #1 v2.0 (RFC 2437, October 1998).
17785 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
17786 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
17787 against Bleichbacher's attack on RSA.
17788 *Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
17791 * Updates to the new SSL compression code
17793 *Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)*
17795 * Fix so that the version number in the master secret, when passed
17796 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
17797 (because the server will not accept higher), that the version number
17798 is 0x03,0x01, not 0x03,0x00
17800 *Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)*
17802 * Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
17803 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
17804 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
17808 * Support for RAW extensions where an arbitrary extension can be
17809 created by including its DER encoding. See apps/openssl.cnf for
17814 * Make sure latest Perl versions don't interpret some generated C array
17815 code as Perl array code in the crypto/err/err_genc.pl script.
17817 *Lars Weber <3weber@informatik.uni-hamburg.de>*
17819 * Modify ms/do_ms.bat to not generate assembly language makefiles since
17820 not many people have the assembler. Various Win32 compilation fixes and
17821 update to the INSTALL.W32 file with (hopefully) more accurate Win32
17822 build instructions.
17826 * Modify configure script 'Configure' to automatically create crypto/date.h
17827 file under Win32 and also build pem.h from pem.org. New script
17828 util/mkfiles.pl to create the MINFO file on environments that can't do a
17829 'make files': perl util/mkfiles.pl >MINFO should work.
17833 * Major rework of DES function declarations, in the pursuit of correctness
17834 and purity. As a result, many evil casts evaporated, and some weirdness,
17835 too. You may find this causes warnings in your code. Zapping your evil
17836 casts will probably fix them. Mostly.
17840 * Fix for a typo in asn1.h. Bug fix to object creation script
17841 obj_dat.pl. It considered a zero in an object definition to mean
17842 "end of object": none of the objects in objects.h have any zeros
17843 so it wasn't spotted.
17845 *Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>*
17847 * Add support for Triple DES Cipher Block Chaining with Output Feedback
17848 Masking (CBCM). In the absence of test vectors, the best I have been able
17849 to do is check that the decrypt undoes the encrypt, so far. Send me test
17850 vectors if you have them.
17854 * Correct calculation of key length for export ciphers (too much space was
17855 allocated for null ciphers). This has not been tested!
17859 * Modifications to the mkdef.pl for Win32 DEF file creation. The usage
17860 message is now correct (it understands "crypto" and "ssl" on its
17861 command line). There is also now an "update" option. This will update
17862 the util/ssleay.num and util/libeay.num files with any new functions.
17864 perl util/mkdef.pl crypto ssl update
17865 it will update them.
17869 * Overhauled the Perl interface:
17870 - ported BN stuff to OpenSSL's different BN library
17871 - made the perl/ source tree CVS-aware
17872 - renamed the package from SSLeay to OpenSSL (the files still contain
17873 their history because I've copied them in the repository)
17874 - removed obsolete files (the test scripts will be replaced
17875 by better Test::Harness variants in the future)
17877 *Ralf S. Engelschall*
17879 * First cut for a very conservative source tree cleanup:
17880 1. merge various obsolete readme texts into doc/ssleay.txt
17881 where we collect the old documents and readme texts.
17882 2. remove the first part of files where I'm already sure that we no
17883 longer need them because of three reasons: either they are just temporary
17884 files which were left by Eric or they are preserved original files where
17885 I've verified that the diff is also available in the CVS via "cvs diff
17886 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
17887 the crypto/md/ stuff).
17889 *Ralf S. Engelschall*
17891 * More extension code. Incomplete support for subject and issuer alt
17892 name, issuer and authority key id. Change the i2v function parameters
17893 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
17894 what that's for :-) Fix to ASN1 macro which messed up
17895 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
17899 * Preliminary support for ENUMERATED type. This is largely copied from the
17904 * Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
17906 *Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)*
17908 * Make sure `make rehash` target really finds the `openssl` program.
17910 *Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>*
17912 * Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
17913 like to hear about it if this slows down other processors.
17917 * Add CygWin32 platform information to Configure script.
17919 *Alan Batie <batie@aahz.jf.intel.com>*
17921 * Fixed ms/32all.bat script: `no_asm` -> `no-asm`
17923 *Rainer W. Gerling <gerling@mpg-gv.mpg.de>*
17925 * New program nseq to manipulate netscape certificate sequences
17929 * Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
17934 * Fixes to BN code. Previously the default was to define BN_RECURSION
17935 but the BN code had some problems that would cause failures when
17936 doing certificate verification and some other functions.
17938 *Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)*
17940 * Add ASN1 and PEM code to support netscape certificate sequences.
17944 * Add ASN1 and PEM code to support netscape certificate sequences.
17948 * Add several PKIX and private extended key usage OIDs.
17952 * Modify the 'ca' program to handle the new extension code. Modify
17953 openssl.cnf for new extension format, add comments.
17957 * More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
17958 and add a sample to openssl.cnf so req -x509 now adds appropriate
17963 * Continued X509 V3 changes. Add to other makefiles, integrate with the
17964 error code, add initial support to X509_print() and x509 application.
17968 * Takes a deep breath and start adding X509 V3 extension support code. Add
17969 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
17970 stuff is currently isolated and isn't even compiled yet.
17974 * Continuing patches for GeneralizedTime. Fix up certificate and CRL
17975 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
17976 Removed the versions check from X509 routines when loading extensions:
17977 this allows certain broken certificates that don't set the version
17978 properly to be processed.
17982 * Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
17983 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
17984 can still be regenerated with "make depend".
17988 * Spelling mistake in C version of CAST-128.
17990 *Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>*
17992 * Changes to the error generation code. The perl script err-code.pl
17993 now reads in the old error codes and retains the old numbers, only
17994 adding new ones if necessary. It also only changes the .err files if new
17995 codes are added. The makefiles have been modified to only insert errors
17996 when needed (to avoid needlessly modifying header files). This is done
17997 by only inserting errors if the .err file is newer than the auto generated
17998 C file. To rebuild all the error codes from scratch (the old behaviour)
17999 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
18000 or delete all the .err files.
18004 * CAST-128 was incorrectly implemented for short keys. The C version has
18005 been fixed, but is untested. The assembler versions are also fixed, but
18006 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
18007 to regenerate it if needed.
18008 *Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
18009 Hagino <itojun@kame.net>*
18011 * File was opened incorrectly in randfile.c.
18013 *Ulf Möller <ulf@fitug.de>*
18015 * Beginning of support for GeneralizedTime. d2i, i2d, check and print
18016 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
18017 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
18018 al: it's just almost always a UTCTime. Note this patch adds new error
18019 codes so do a "make errors" if there are problems.
18023 * Correct Linux 1 recognition in config.
18025 *Ulf Möller <ulf@fitug.de>*
18027 * Remove pointless MD5 hash when using DSA keys in ca.
18029 *Anonymous <nobody@replay.com>*
18031 * Generate an error if given an empty string as a cert directory. Also
18032 generate an error if handed NULL (previously returned 0 to indicate an
18033 error, but didn't set one).
18035 *Ben Laurie, reported by Anonymous <nobody@replay.com>*
18037 * Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
18041 * Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
18042 parameters. This was causing a warning which killed off the Win32 compile.
18046 * Remove C++ style comments from crypto/bn/bn_local.h.
18048 *Neil Costigan <neil.costigan@celocom.com>*
18050 * The function OBJ_txt2nid was broken. It was supposed to return a nid
18051 based on a text string, looking up short and long names and finally
18052 "dot" format. The "dot" format stuff didn't work. Added new function
18053 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
18054 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
18055 OID is not part of the table.
18059 * Add prototypes to X509 lookup/verify methods, fixing a bug in
18060 X509_LOOKUP_by_alias().
18064 * Sort openssl functions by name.
18068 * Get the gendsa program working (hopefully) and add it to app list. Remove
18069 encryption from sample DSA keys (in case anyone is interested the password
18074 * Make *all* `*_free` functions accept a NULL pointer.
18076 *Frans Heymans <fheymans@isaserver.be>*
18078 * If a DH key is generated in s3_srvr.c, don't blow it by trying to use
18081 *Anonymous <nobody@replay.com>*
18083 * s_server should send the CAfile as acceptable CAs, not its own cert.
18085 *Bodo Moeller <3moeller@informatik.uni-hamburg.de>*
18087 * Don't blow it for numeric -newkey arguments to apps/req.
18089 *Bodo Moeller <3moeller@informatik.uni-hamburg.de>*
18091 * Temp key "for export" tests were wrong in s3_srvr.c.
18093 *Anonymous <nobody@replay.com>*
18095 * Add prototype for temp key callback functions
18096 SSL_CTX_set_tmp_{rsa,dh}_callback().
18100 * Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
18101 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
18105 * X509_name_add_entry() freed the wrong thing after an error.
18107 *Arne Ansper <arne@ats.cyber.ee>*
18109 * rsa_eay.c would attempt to free a NULL context.
18111 *Arne Ansper <arne@ats.cyber.ee>*
18113 * BIO_s_socket() had a broken should_retry() on Windoze.
18115 *Arne Ansper <arne@ats.cyber.ee>*
18117 * BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
18119 *Arne Ansper <arne@ats.cyber.ee>*
18121 * Make sure the already existing X509_STORE->depth variable is initialized
18122 in X509_STORE_new(), but document the fact that this variable is still
18123 unused in the certificate verification process.
18125 *Ralf S. Engelschall*
18127 * Fix the various library and apps files to free up pkeys obtained from
18128 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
18132 * Fix reference counting in X509_PUBKEY_get(). This makes
18133 demos/maurice/example2.c work, amongst others, probably.
18135 *Steve Henson and Ben Laurie*
18137 * First cut of a cleanup for apps/. First the `ssleay` program is now named
18138 `openssl` and second, the shortcut symlinks for the `openssl <command>`
18139 are no longer created. This way we have a single and consistent command
18140 line interface `openssl <command>`, similar to `cvs <command>`.
18142 *Ralf S. Engelschall, Paul Sutton and Ben Laurie*
18144 * ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
18145 BIT STRING wrapper always have zero unused bits.
18149 * Add CA.pl, perl version of CA.sh, add extended key usage OID.
18153 * Make the top-level INSTALL documentation easier to understand.
18157 * Makefiles updated to exit if an error occurs in a sub-directory
18158 make (including if user presses ^C) [Paul Sutton]
18160 * Make Montgomery context stuff explicit in RSA data structure.
18164 * Fix build order of pem and err to allow for generated pem.h.
18168 * Fix renumbering bug in X509_NAME_delete_entry().
18172 * Enhanced the err-ins.pl script so it makes the error library number
18173 global and can add a library name. This is needed for external ASN1 and
18174 other error libraries.
18178 * Fixed sk_insert which never worked properly.
18182 * Fix ASN1 macros so they can handle indefinite length constructed
18183 EXPLICIT tags. Some non standard certificates use these: they can now
18188 * Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
18189 into a single doc/ssleay.txt bundle. This way the information is still
18190 preserved but no longer messes up this directory. Now it's new room for
18191 the new set of documentation files.
18193 *Ralf S. Engelschall*
18195 * SETs were incorrectly DER encoded. This was a major pain, because they
18196 shared code with SEQUENCEs, which aren't coded the same. This means that
18197 almost everything to do with SETs or SEQUENCEs has either changed name or
18198 number of arguments.
18200 *Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>*
18202 * Fix test data to work with the above.
18206 * Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
18207 was already fixed by Eric for 0.9.1 it seems.
18209 *Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>*
18211 * Autodetect FreeBSD3.
18215 * Fix various bugs in Configure. This affects the following platforms:
18219 unixware-2.0-pentium
18224 * Eliminate generated files from CVS. Reorder tests to regenerate files
18225 before they are needed.
18229 * Generate Makefile.ssl from Makefile.org (to keep CVS happy).
18233 ### Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
18235 * Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
18236 changed SSLeay to OpenSSL in version strings.
18238 *Ralf S. Engelschall*
18240 * Some fixups to the top-level documents.
18244 * Fixed the nasty bug where rsaref.h was not found under compile-time
18245 because the symlink to include/ was missing.
18247 *Ralf S. Engelschall*
18249 * Incorporated the popular no-RSA/DSA-only patches
18250 which allow to compile a RSA-free SSLeay.
18252 *Andrew Cooke / Interrader Ldt., Ralf S. Engelschall*
18254 * Fixed nasty rehash problem under `make -f Makefile.ssl links`
18255 when "ssleay" is still not found.
18257 *Ralf S. Engelschall*
18259 * Added more platforms to Configure: Cray T3E, HPUX 11,
18261 *Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>*
18263 * Updated the README file.
18265 *Ralf S. Engelschall*
18267 * Added various .cvsignore files in the CVS repository subdirs
18268 to make a "cvs update" really silent.
18270 *Ralf S. Engelschall*
18272 * Recompiled the error-definition header files and added
18273 missing symbols to the Win32 linker tables.
18275 *Ralf S. Engelschall*
18277 * Cleaned up the top-level documents;
18278 o new files: CHANGES and LICENSE
18279 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
18280 o merged COPYRIGHT into LICENSE
18281 o removed obsolete TODO file
18282 o renamed MICROSOFT to INSTALL.W32
18284 *Ralf S. Engelschall*
18286 * Removed dummy files from the 0.9.1b source tree:
18287 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
18288 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
18289 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
18290 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
18291 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
18293 *Ralf S. Engelschall*
18295 * Added various platform portability fixes.
18299 * The Genesis of the OpenSSL rpject:
18300 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
18301 Young and Tim J. Hudson created while they were working for C2Net until
18304 *The OpenSSL Project*
18306 ### Changes between 0.9.0b and 0.9.1b [not released]
18308 * Updated a few CA certificates under certs/
18312 * Changed some BIGNUM api stuff.
18316 * Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
18317 DGUX x86, Linux Alpha, etc.
18321 * New COMP library [crypto/comp/] for SSL Record Layer Compression:
18322 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
18327 * Add -strparse option to asn1pars program which parses nested
18330 *Dr Stephen Henson <shenson@bigfoot.com>*
18332 * Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
18336 * DSA fix for "ca" program.
18340 * Added "-genkey" option to "dsaparam" program.
18344 * Added RIPE MD160 (rmd160) message digest.
18348 * Added -a (all) option to "ssleay version" command.
18352 * Added PLATFORM define which is the id given to Configure.
18356 * Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
18360 * Extended the ASN.1 parser routines.
18364 * Extended BIO routines to support REUSEADDR, seek, tell, etc.
18368 * Added a BN_CTX to the BN library.
18372 * Fixed the weak key values in DES library
18376 * Changed API in EVP library for cipher aliases.
18380 * Added support for RC2/64bit cipher.
18384 * Converted the lhash library to the crypto/mem.c functions.
18388 * Added more recognized ASN.1 object ids.
18392 * Added more RSA padding checks for SSL/TLS.
18396 * Added BIO proxy/filter functionality.
18400 * Added extra_certs to SSL_CTX which can be used
18401 send extra CA certificates to the client in the CA cert chain sending
18402 process. It can be configured with SSL_CTX_add_extra_chain_cert().
18406 * Now Fortezza is denied in the authentication phase because
18407 this is key exchange mechanism is not supported by SSLeay at all.
18411 * Additional PKCS1 checks.
18415 * Support the string "TLSv1" for all TLS v1 ciphers.
18419 * Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
18420 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
18424 * Fixed a few memory leaks.
18428 * Fixed various code and comment typos.
18432 * A minor bug in ssl/s3_clnt.c where there would always be 4 0
18433 bytes sent in the client random.
18435 *Edward Bishop <ebishop@spyglass.com>*
18439 [CVE-2019-1563]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1563
18440 [CVE-2019-1559]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1559
18441 [CVE-2019-1552]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1552
18442 [CVE-2019-1551]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1551
18443 [CVE-2019-1549]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1549
18444 [CVE-2019-1547]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1547
18445 [CVE-2019-1543]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1543
18446 [CVE-2018-5407]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-5407
18447 [CVE-2018-0739]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0739
18448 [CVE-2018-0737]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0737
18449 [CVE-2018-0735]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0735
18450 [CVE-2018-0734]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0734
18451 [CVE-2018-0733]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0733
18452 [CVE-2018-0732]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0732
18453 [CVE-2017-3738]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3738
18454 [CVE-2017-3737]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3737
18455 [CVE-2017-3736]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3736
18456 [CVE-2017-3735]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3735
18457 [CVE-2017-3733]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3733
18458 [CVE-2017-3732]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3732
18459 [CVE-2017-3731]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3731
18460 [CVE-2017-3730]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3730
18461 [CVE-2016-7055]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-7055
18462 [CVE-2016-7054]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-7054
18463 [CVE-2016-7053]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-7053
18464 [CVE-2016-7052]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-7052
18465 [CVE-2016-6309]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6309
18466 [CVE-2016-6308]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6308
18467 [CVE-2016-6307]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6307
18468 [CVE-2016-6306]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6306
18469 [CVE-2016-6305]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6305
18470 [CVE-2016-6304]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6304
18471 [CVE-2016-6303]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6303
18472 [CVE-2016-6302]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6302
18473 [CVE-2016-2183]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2183
18474 [CVE-2016-2182]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2182
18475 [CVE-2016-2181]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2181
18476 [CVE-2016-2180]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2180
18477 [CVE-2016-2179]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2179
18478 [CVE-2016-2178]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2178
18479 [CVE-2016-2177]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2177
18480 [CVE-2016-2176]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2176
18481 [CVE-2016-2109]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2109
18482 [CVE-2016-2107]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2107
18483 [CVE-2016-2106]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2106
18484 [CVE-2016-2105]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2105
18485 [CVE-2016-0800]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0800
18486 [CVE-2016-0799]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0799
18487 [CVE-2016-0798]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0798
18488 [CVE-2016-0797]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0797
18489 [CVE-2016-0705]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0705
18490 [CVE-2016-0702]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0702
18491 [CVE-2016-0701]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0701
18492 [CVE-2015-3197]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-3197
18493 [CVE-2015-3196]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-3196
18494 [CVE-2015-3195]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-3195
18495 [CVE-2015-3194]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-3194
18496 [CVE-2015-3193]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-3193
18497 [CVE-2015-1793]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1793
18498 [CVE-2015-1792]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1792
18499 [CVE-2015-1791]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1791
18500 [CVE-2015-1790]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1790
18501 [CVE-2015-1789]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1789
18502 [CVE-2015-1788]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1788
18503 [CVE-2015-1787]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1787
18504 [CVE-2015-0293]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0293
18505 [CVE-2015-0291]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0291
18506 [CVE-2015-0290]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0290
18507 [CVE-2015-0289]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0289
18508 [CVE-2015-0288]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0288
18509 [CVE-2015-0287]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0287
18510 [CVE-2015-0286]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0286
18511 [CVE-2015-0285]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0285
18512 [CVE-2015-0209]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0209
18513 [CVE-2015-0208]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0208
18514 [CVE-2015-0207]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0207
18515 [CVE-2015-0206]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0206
18516 [CVE-2015-0205]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0205
18517 [CVE-2015-0204]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0204
18518 [CVE-2014-8275]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-8275
18519 [CVE-2014-5139]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-5139
18520 [CVE-2014-3572]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3572
18521 [CVE-2014-3571]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3571
18522 [CVE-2014-3570]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3570
18523 [CVE-2014-3569]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3569
18524 [CVE-2014-3568]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3568
18525 [CVE-2014-3567]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3567
18526 [CVE-2014-3566]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3566
18527 [CVE-2014-3513]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3513
18528 [CVE-2014-3512]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3512
18529 [CVE-2014-3511]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3511
18530 [CVE-2014-3510]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3510
18531 [CVE-2014-3509]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3509
18532 [CVE-2014-3508]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3508
18533 [CVE-2014-3507]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3507
18534 [CVE-2014-3506]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3506
18535 [CVE-2014-3505]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3505
18536 [CVE-2014-3470]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3470
18537 [CVE-2014-0224]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-0224
18538 [CVE-2014-0221]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-0221
18539 [CVE-2014-0195]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-0195
18540 [CVE-2014-0160]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-0160
18541 [CVE-2014-0076]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-0076
18542 [CVE-2013-6450]: https://www.openssl.org/news/vulnerabilities.html#CVE-2013-6450
18543 [CVE-2013-4353]: https://www.openssl.org/news/vulnerabilities.html#CVE-2013-4353
18544 [CVE-2013-0169]: https://www.openssl.org/news/vulnerabilities.html#CVE-2013-0169
18545 [CVE-2013-0166]: https://www.openssl.org/news/vulnerabilities.html#CVE-2013-0166
18546 [CVE-2012-2686]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-2686
18547 [CVE-2012-2333]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-2333
18548 [CVE-2012-2110]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-2110
18549 [CVE-2012-0884]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-0884
18550 [CVE-2012-0050]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-0050
18551 [CVE-2012-0027]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-0027
18552 [CVE-2011-4619]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-4619
18553 [CVE-2011-4577]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-4577
18554 [CVE-2011-4576]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-4576
18555 [CVE-2011-4109]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-4109
18556 [CVE-2011-4108]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-4108
18557 [CVE-2011-3210]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-3210
18558 [CVE-2011-3207]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-3207
18559 [CVE-2011-0014]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-0014
18560 [CVE-2010-4252]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-4252
18561 [CVE-2010-4180]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-4180
18562 [CVE-2010-3864]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-3864
18563 [CVE-2010-1633]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-1633
18564 [CVE-2010-0740]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-0740
18565 [CVE-2010-0433]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-0433
18566 [CVE-2009-4355]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-4355
18567 [CVE-2009-3555]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-3555
18568 [CVE-2009-3245]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-3245
18569 [CVE-2009-1386]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-1386
18570 [CVE-2009-1379]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-1379
18571 [CVE-2009-1378]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-1378
18572 [CVE-2009-1377]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-1377
18573 [CVE-2009-0789]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-0789
18574 [CVE-2009-0591]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-0591
18575 [CVE-2009-0590]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-0590
18576 [CVE-2008-5077]: https://www.openssl.org/news/vulnerabilities.html#CVE-2008-5077
18577 [CVE-2008-1678]: https://www.openssl.org/news/vulnerabilities.html#CVE-2008-1678
18578 [CVE-2008-1672]: https://www.openssl.org/news/vulnerabilities.html#CVE-2008-1672
18579 [CVE-2008-0891]: https://www.openssl.org/news/vulnerabilities.html#CVE-2008-0891
18580 [CVE-2007-5135]: https://www.openssl.org/news/vulnerabilities.html#CVE-2007-5135
18581 [CVE-2007-4995]: https://www.openssl.org/news/vulnerabilities.html#CVE-2007-4995
18582 [CVE-2006-4343]: https://www.openssl.org/news/vulnerabilities.html#CVE-2006-4343
18583 [CVE-2006-4339]: https://www.openssl.org/news/vulnerabilities.html#CVE-2006-4339
18584 [CVE-2006-3738]: https://www.openssl.org/news/vulnerabilities.html#CVE-2006-3738
18585 [CVE-2006-2940]: https://www.openssl.org/news/vulnerabilities.html#CVE-2006-2940
18586 [CVE-2006-2937]: https://www.openssl.org/news/vulnerabilities.html#CVE-2006-2937
18587 [CVE-2005-2969]: https://www.openssl.org/news/vulnerabilities.html#CVE-2005-2969
18588 [CVE-2004-0112]: https://www.openssl.org/news/vulnerabilities.html#CVE-2004-0112
18589 [CVE-2004-0079]: https://www.openssl.org/news/vulnerabilities.html#CVE-2004-0079
18590 [CVE-2003-0851]: https://www.openssl.org/news/vulnerabilities.html#CVE-2003-0851
18591 [CVE-2003-0545]: https://www.openssl.org/news/vulnerabilities.html#CVE-2003-0545
18592 [CVE-2003-0544]: https://www.openssl.org/news/vulnerabilities.html#CVE-2003-0544
18593 [CVE-2003-0543]: https://www.openssl.org/news/vulnerabilities.html#CVE-2003-0543
18594 [CVE-2003-0078]: https://www.openssl.org/news/vulnerabilities.html#CVE-2003-0078
18595 [CVE-2002-0659]: https://www.openssl.org/news/vulnerabilities.html#CVE-2002-0659
18596 [CVE-2002-0657]: https://www.openssl.org/news/vulnerabilities.html#CVE-2002-0657
18597 [CVE-2002-0656]: https://www.openssl.org/news/vulnerabilities.html#CVE-2002-0656
18598 [CVE-2002-0655]: https://www.openssl.org/news/vulnerabilities.html#CVE-2002-0655