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 These commands are now in maintenance mode and no new features will
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 support for Linux Kernel TLS data-path. The Linux Kernel data-path
1162 improves application performance by removing data copies and providing
1163 applications with zero-copy system calls such as sendfile and splice.
1167 * The SSL option SSL_OP_CLEANSE_PLAINTEXT is introduced. If that
1168 option is set, openssl cleanses (zeroize) plaintext bytes from
1169 internal buffers after delivering them to the application. Note,
1170 the application is still responsible for cleansing other copies
1171 (e.g.: data received by SSL_read(3)).
1175 * `PKCS12_parse` now maintains the order of the parsed certificates
1176 when outputting them via `*ca` (rather than reversing it).
1183 ### Changes between 1.1.1e and 1.1.1f [xx XXX xxxx]
1185 ### Changes between 1.1.1d and 1.1.1e [17 Mar 2020]
1187 * Properly detect EOF while reading in libssl. Previously if we hit an EOF
1188 while reading in libssl then we would report an error back to the
1189 application (SSL_ERROR_SYSCALL) but errno would be 0. We now add
1190 an error to the stack (which means we instead return SSL_ERROR_SSL) and
1191 therefore give a hint as to what went wrong.
1195 * Check that ed25519 and ed448 are allowed by the security level. Previously
1196 signature algorithms not using an MD were not being checked that they were
1197 allowed by the security level.
1201 * Fixed SSL_get_servername() behaviour. The behaviour of SSL_get_servername()
1202 was not quite right. The behaviour was not consistent between resumption
1203 and normal handshakes, and also not quite consistent with historical
1204 behaviour. The behaviour in various scenarios has been clarified and
1205 it has been updated to make it match historical behaviour as closely as
1210 * *[VMS only]* The header files that the VMS compilers include automatically,
1211 `__DECC_INCLUDE_PROLOGUE.H` and `__DECC_INCLUDE_EPILOGUE.H`, use pragmas
1212 that the C++ compiler doesn't understand. This is a shortcoming in the
1213 compiler, but can be worked around with `__cplusplus` guards.
1215 C++ applications that use OpenSSL libraries must be compiled using the
1216 qualifier `/NAMES=(AS_IS,SHORTENED)` to be able to use all the OpenSSL
1217 functions. Otherwise, only functions with symbols of less than 31
1218 characters can be used, as the linker will not be able to successfully
1219 resolve symbols with longer names.
1223 * Added a new method to gather entropy on VMS, based on SYS$GET_ENTROPY.
1224 The presence of this system service is determined at run-time.
1228 * Added newline escaping functionality to a filename when using openssl dgst.
1229 This output format is to replicate the output format found in the `*sum`
1230 checksum programs. This aims to preserve backward compatibility.
1232 *Matt Eaton, Richard Levitte, and Paul Dale*
1234 * Print all values for a PKCS#12 attribute with 'openssl pkcs12', not just
1239 ### Changes between 1.1.1c and 1.1.1d [10 Sep 2019]
1241 * Fixed a fork protection issue. OpenSSL 1.1.1 introduced a rewritten random
1242 number generator (RNG). This was intended to include protection in the
1243 event of a fork() system call in order to ensure that the parent and child
1244 processes did not share the same RNG state. However this protection was not
1245 being used in the default case.
1247 A partial mitigation for this issue is that the output from a high
1248 precision timer is mixed into the RNG state so the likelihood of a parent
1249 and child process sharing state is significantly reduced.
1251 If an application already calls OPENSSL_init_crypto() explicitly using
1252 OPENSSL_INIT_ATFORK then this problem does not occur at all.
1255 *Matthias St. Pierre*
1257 * For built-in EC curves, ensure an EC_GROUP built from the curve name is
1258 used even when parsing explicit parameters, when loading a encoded key
1259 or calling `EC_GROUP_new_from_ecpkparameters()`/
1260 `EC_GROUP_new_from_ecparameters()`.
1261 This prevents bypass of security hardening and performance gains,
1262 especially for curves with specialized EC_METHODs.
1263 By default, if a key encoded with explicit parameters is loaded and later
1264 encoded, the output is still encoded with explicit parameters, even if
1265 internally a "named" EC_GROUP is used for computation.
1269 * Compute ECC cofactors if not provided during EC_GROUP construction. Before
1270 this change, EC_GROUP_set_generator would accept order and/or cofactor as
1271 NULL. After this change, only the cofactor parameter can be NULL. It also
1272 does some minimal sanity checks on the passed order.
1277 * Fixed a padding oracle in PKCS7_dataDecode and CMS_decrypt_set1_pkey.
1278 An attack is simple, if the first CMS_recipientInfo is valid but the
1279 second CMS_recipientInfo is chosen ciphertext. If the second
1280 recipientInfo decodes to PKCS #1 v1.5 form plaintext, the correct
1281 encryption key will be replaced by garbage, and the message cannot be
1282 decoded, but if the RSA decryption fails, the correct encryption key is
1283 used and the recipient will not notice the attack.
1284 As a work around for this potential attack the length of the decrypted
1285 key must be equal to the cipher default key length, in case the
1286 certifiate is not given and all recipientInfo are tried out.
1287 The old behaviour can be re-enabled in the CMS code by setting the
1288 CMS_DEBUG_DECRYPT flag.
1293 * Early start up entropy quality from the DEVRANDOM seed source has been
1294 improved for older Linux systems. The RAND subsystem will wait for
1295 /dev/random to be producing output before seeding from /dev/urandom.
1296 The seeded state is stored for future library initialisations using
1297 a system global shared memory segment. The shared memory identifier
1298 can be configured by defining OPENSSL_RAND_SEED_DEVRANDOM_SHM_ID to
1299 the desired value. The default identifier is 114.
1303 * Correct the extended master secret constant on EBCDIC systems. Without this
1304 fix TLS connections between an EBCDIC system and a non-EBCDIC system that
1305 negotiate EMS will fail. Unfortunately this also means that TLS connections
1306 between EBCDIC systems with this fix, and EBCDIC systems without this
1307 fix will fail if they negotiate EMS.
1311 * Use Windows installation paths in the mingw builds
1313 Mingw isn't a POSIX environment per se, which means that Windows
1314 paths should be used for installation.
1319 * Changed DH_check to accept parameters with order q and 2q subgroups.
1320 With order 2q subgroups the bit 0 of the private key is not secret
1321 but DH_generate_key works around that by clearing bit 0 of the
1322 private key for those. This avoids leaking bit 0 of the private key.
1326 * Significantly reduce secure memory usage by the randomness pools.
1330 * Revert the DEVRANDOM_WAIT feature for Linux systems
1332 The DEVRANDOM_WAIT feature added a select() call to wait for the
1333 /dev/random device to become readable before reading from the
1334 /dev/urandom device.
1336 It turned out that this change had negative side effects on
1337 performance which were not acceptable. After some discussion it
1338 was decided to revert this feature and leave it up to the OS
1339 resp. the platform maintainer to ensure a proper initialization
1340 during early boot time.
1342 *Matthias St. Pierre*
1344 ### Changes between 1.1.1b and 1.1.1c [28 May 2019]
1346 * Add build tests for C++. These are generated files that only do one
1347 thing, to include one public OpenSSL head file each. This tests that
1348 the public header files can be usefully included in a C++ application.
1350 This test isn't enabled by default. It can be enabled with the option
1351 'enable-buildtest-c++'.
1355 * Enable SHA3 pre-hashing for ECDSA and DSA.
1359 * Change the default RSA, DSA and DH size to 2048 bit instead of 1024.
1360 This changes the size when using the genpkey app when no size is given. It
1361 fixes an omission in earlier changes that changed all RSA, DSA and DH
1362 generation apps to use 2048 bits by default.
1366 * Reorganize the manual pages to consistently have RETURN VALUES,
1367 EXAMPLES, SEE ALSO and HISTORY come in that order, and adjust
1368 util/fix-doc-nits accordingly.
1370 *Paul Yang, Joshua Lock*
1372 * Add the missing accessor EVP_PKEY_get0_engine()
1376 * Have apps like 's_client' and 's_server' output the signature scheme
1377 along with other cipher suite parameters when debugging.
1381 * Make OPENSSL_config() error agnostic again.
1385 * Do the error handling in RSA decryption constant time.
1389 * Prevent over long nonces in ChaCha20-Poly1305.
1391 ChaCha20-Poly1305 is an AEAD cipher, and requires a unique nonce input
1392 for every encryption operation. RFC 7539 specifies that the nonce value
1393 (IV) should be 96 bits (12 bytes). OpenSSL allows a variable nonce length
1394 and front pads the nonce with 0 bytes if it is less than 12
1395 bytes. However it also incorrectly allows a nonce to be set of up to 16
1396 bytes. In this case only the last 12 bytes are significant and any
1397 additional leading bytes are ignored.
1399 It is a requirement of using this cipher that nonce values are
1400 unique. Messages encrypted using a reused nonce value are susceptible to
1401 serious confidentiality and integrity attacks. If an application changes
1402 the default nonce length to be longer than 12 bytes and then makes a
1403 change to the leading bytes of the nonce expecting the new value to be a
1404 new unique nonce then such an application could inadvertently encrypt
1405 messages with a reused nonce.
1407 Additionally the ignored bytes in a long nonce are not covered by the
1408 integrity guarantee of this cipher. Any application that relies on the
1409 integrity of these ignored leading bytes of a long nonce may be further
1410 affected. Any OpenSSL internal use of this cipher, including in SSL/TLS,
1411 is safe because no such use sets such a long nonce value. However user
1412 applications that use this cipher directly and set a non-default nonce
1413 length to be longer than 12 bytes may be vulnerable.
1415 This issue was reported to OpenSSL on 16th of March 2019 by Joran Dirk
1421 * Add DEVRANDOM_WAIT feature for Linux systems
1423 On older Linux systems where the getrandom() system call is not available,
1424 OpenSSL normally uses the /dev/urandom device for seeding its CSPRNG.
1425 Contrary to getrandom(), the /dev/urandom device will not block during
1426 early boot when the kernel CSPRNG has not been seeded yet.
1428 To mitigate this known weakness, use select() to wait for /dev/random to
1429 become readable before reading from /dev/urandom.
1431 * Ensure that SM2 only uses SM3 as digest algorithm
1435 ### Changes between 1.1.1a and 1.1.1b [26 Feb 2019]
1437 * Change the info callback signals for the start and end of a post-handshake
1438 message exchange in TLSv1.3. In 1.1.1/1.1.1a we used SSL_CB_HANDSHAKE_START
1439 and SSL_CB_HANDSHAKE_DONE. Experience has shown that many applications get
1440 confused by this and assume that a TLSv1.2 renegotiation has started. This
1441 can break KeyUpdate handling. Instead we no longer signal the start and end
1442 of a post handshake message exchange (although the messages themselves are
1443 still signalled). This could break some applications that were expecting
1444 the old signals. However without this KeyUpdate is not usable for many
1449 ### Changes between 1.1.1 and 1.1.1a [20 Nov 2018]
1451 * Timing vulnerability in DSA signature generation
1453 The OpenSSL DSA signature algorithm has been shown to be vulnerable to a
1454 timing side channel attack. An attacker could use variations in the signing
1455 algorithm to recover the private key.
1457 This issue was reported to OpenSSL on 16th October 2018 by Samuel Weiser.
1462 * Timing vulnerability in ECDSA signature generation
1464 The OpenSSL ECDSA signature algorithm has been shown to be vulnerable to a
1465 timing side channel attack. An attacker could use variations in the signing
1466 algorithm to recover the private key.
1468 This issue was reported to OpenSSL on 25th October 2018 by Samuel Weiser.
1473 * Fixed the issue that RAND_add()/RAND_seed() silently discards random input
1474 if its length exceeds 4096 bytes. The limit has been raised to a buffer size
1475 of two gigabytes and the error handling improved.
1477 This issue was reported to OpenSSL by Dr. Falko Strenzke. It has been
1478 categorized as a normal bug, not a security issue, because the DRBG reseeds
1479 automatically and is fully functional even without additional randomness
1480 provided by the application.
1482 ### Changes between 1.1.0i and 1.1.1 [11 Sep 2018]
1484 * Add a new ClientHello callback. Provides a callback interface that gives
1485 the application the ability to adjust the nascent SSL object at the
1486 earliest stage of ClientHello processing, immediately after extensions have
1487 been collected but before they have been processed. In particular, this
1488 callback can adjust the supported TLS versions in response to the contents
1493 * Add SM2 base algorithm support.
1497 * s390x assembly pack: add (improved) hardware-support for the following
1498 cryptographic primitives: sha3, shake, aes-gcm, aes-ccm, aes-ctr, aes-ofb,
1499 aes-cfb/cfb8, aes-ecb.
1503 * Make EVP_PKEY_asn1_new() a bit stricter about its input. A NULL pem_str
1504 parameter is no longer accepted, as it leads to a corrupt table. NULL
1505 pem_str is reserved for alias entries only.
1509 * Use the new ec_scalar_mul_ladder scaffold to implement a specialized ladder
1510 step for prime curves. The new implementation is based on formulae from
1511 differential addition-and-doubling in homogeneous projective coordinates
1512 from Izu-Takagi "A fast parallel elliptic curve multiplication resistant
1513 against side channel attacks" and Brier-Joye "Weierstrass Elliptic Curves
1514 and Side-Channel Attacks" Eq. (8) for y-coordinate recovery, modified
1515 to work in projective coordinates.
1517 *Billy Bob Brumley, Nicola Tuveri*
1519 * Change generating and checking of primes so that the error rate of not
1520 being prime depends on the intended use based on the size of the input.
1521 For larger primes this will result in more rounds of Miller-Rabin.
1522 The maximal error rate for primes with more than 1080 bits is lowered
1525 *Kurt Roeckx, Annie Yousar*
1527 * Increase the number of Miller-Rabin rounds for DSA key generating to 64.
1531 * The 'tsget' script is renamed to 'tsget.pl', to avoid confusion when
1532 moving between systems, and to avoid confusion when a Windows build is
1533 done with mingw vs with MSVC. For POSIX installs, there's still a
1534 symlink or copy named 'tsget' to avoid that confusion as well.
1538 * Revert blinding in ECDSA sign and instead make problematic addition
1539 length-invariant. Switch even to fixed-length Montgomery multiplication.
1543 * Use the new ec_scalar_mul_ladder scaffold to implement a specialized ladder
1544 step for binary curves. The new implementation is based on formulae from
1545 differential addition-and-doubling in mixed Lopez-Dahab projective
1546 coordinates, modified to independently blind the operands.
1548 *Billy Bob Brumley, Sohaib ul Hassan, Nicola Tuveri*
1550 * Add a scaffold to optionally enhance the Montgomery ladder implementation
1551 for `ec_scalar_mul_ladder` (formerly `ec_mul_consttime`) allowing
1552 EC_METHODs to implement their own specialized "ladder step", to take
1553 advantage of more favorable coordinate systems or more efficient
1554 differential addition-and-doubling algorithms.
1556 *Billy Bob Brumley, Sohaib ul Hassan, Nicola Tuveri*
1558 * Modified the random device based seed sources to keep the relevant
1559 file descriptors open rather than reopening them on each access.
1560 This allows such sources to operate in a chroot() jail without
1561 the associated device nodes being available. This behaviour can be
1562 controlled using RAND_keep_random_devices_open().
1566 * Numerous side-channel attack mitigations have been applied. This may have
1567 performance impacts for some algorithms for the benefit of improved
1568 security. Specific changes are noted in this change log by their respective
1573 * AIX shared library support overhaul. Switch to AIX "natural" way of
1574 handling shared libraries, which means collecting shared objects of
1575 different versions and bitnesses in one common archive. This allows to
1576 mitigate conflict between 1.0 and 1.1 side-by-side installations. It
1577 doesn't affect the way 3rd party applications are linked, only how
1578 multi-version installation is managed.
1582 * Make ec_group_do_inverse_ord() more robust and available to other
1583 EC cryptosystems, so that irrespective of BN_FLG_CONSTTIME, SCA
1584 mitigations are applied to the fallback BN_mod_inverse().
1585 When using this function rather than BN_mod_inverse() directly, new
1586 EC cryptosystem implementations are then safer-by-default.
1590 * Add coordinate blinding for EC_POINT and implement projective
1591 coordinate blinding for generic prime curves as a countermeasure to
1592 chosen point SCA attacks.
1594 *Sohaib ul Hassan, Nicola Tuveri, Billy Bob Brumley*
1596 * Add blinding to ECDSA and DSA signatures to protect against side channel
1597 attacks discovered by Keegan Ryan (NCC Group).
1601 * Enforce checking in the pkeyutl command line app to ensure that the input
1602 length does not exceed the maximum supported digest length when performing
1603 a sign, verify or verifyrecover operation.
1607 * SSL_MODE_AUTO_RETRY is enabled by default. Applications that use blocking
1608 I/O in combination with something like select() or poll() will hang. This
1609 can be turned off again using SSL_CTX_clear_mode().
1610 Many applications do not properly handle non-application data records, and
1611 TLS 1.3 sends more of such records. Setting SSL_MODE_AUTO_RETRY works
1612 around the problems in those applications, but can also break some.
1613 It's recommended to read the manpages about SSL_read(), SSL_write(),
1614 SSL_get_error(), SSL_shutdown(), SSL_CTX_set_mode() and
1615 SSL_CTX_set_read_ahead() again.
1619 * When unlocking a pass phrase protected PEM file or PKCS#8 container, we
1620 now allow empty (zero character) pass phrases.
1624 * Apply blinding to binary field modular inversion and remove patent
1625 pending (OPENSSL_SUN_GF2M_DIV) BN_GF2m_mod_div implementation.
1629 * Deprecate ec2_mult.c and unify scalar multiplication code paths for
1630 binary and prime elliptic curves.
1634 * Remove ECDSA nonce padding: EC_POINT_mul is now responsible for
1635 constant time fixed point multiplication.
1639 * Revise elliptic curve scalar multiplication with timing attack
1640 defenses: ec_wNAF_mul redirects to a constant time implementation
1641 when computing fixed point and variable point multiplication (which
1642 in OpenSSL are mostly used with secret scalars in keygen, sign,
1643 ECDH derive operations).
1644 *Billy Bob Brumley, Nicola Tuveri, Cesar Pereida GarcÃa,
1647 * Updated CONTRIBUTING
1651 * Updated DRBG / RAND to request nonce and additional low entropy
1652 randomness from the system.
1654 *Matthias St. Pierre*
1656 * Updated 'openssl rehash' to use OpenSSL consistent default.
1660 * Moved the load of the ssl_conf module to libcrypto, which helps
1661 loading engines that libssl uses before libssl is initialised.
1665 * Added EVP_PKEY_sign() and EVP_PKEY_verify() for EdDSA
1669 * Fixed X509_NAME_ENTRY_set to get multi-valued RDNs right in all cases.
1671 *Ingo Schwarze, Rich Salz*
1673 * Added output of accepting IP address and port for 'openssl s_server'
1677 * Added a new API for TLSv1.3 ciphersuites:
1678 SSL_CTX_set_ciphersuites()
1679 SSL_set_ciphersuites()
1683 * Memory allocation failures consistently add an error to the error
1688 * Don't use OPENSSL_ENGINES and OPENSSL_CONF environment values
1689 in libcrypto when run as setuid/setgid.
1693 * Load any config file by default when libssl is used.
1697 * Added new public header file <openssl/rand_drbg.h> and documentation
1698 for the RAND_DRBG API. See manual page RAND_DRBG(7) for an overview.
1700 *Matthias St. Pierre*
1702 * QNX support removed (cannot find contributors to get their approval
1703 for the license change).
1707 * TLSv1.3 replay protection for early data has been implemented. See the
1708 SSL_read_early_data() man page for further details.
1712 * Separated TLSv1.3 ciphersuite configuration out from TLSv1.2 ciphersuite
1713 configuration. TLSv1.3 ciphersuites are not compatible with TLSv1.2 and
1714 below. Similarly TLSv1.2 ciphersuites are not compatible with TLSv1.3.
1715 In order to avoid issues where legacy TLSv1.2 ciphersuite configuration
1716 would otherwise inadvertently disable all TLSv1.3 ciphersuites the
1717 configuration has been separated out. See the ciphers man page or the
1718 SSL_CTX_set_ciphersuites() man page for more information.
1722 * On POSIX (BSD, Linux, ...) systems the ocsp(1) command running
1723 in responder mode now supports the new "-multi" option, which
1724 spawns the specified number of child processes to handle OCSP
1725 requests. The "-timeout" option now also limits the OCSP
1726 responder's patience to wait to receive the full client request
1727 on a newly accepted connection. Child processes are respawned
1728 as needed, and the CA index file is automatically reloaded
1729 when changed. This makes it possible to run the "ocsp" responder
1730 as a long-running service, making the OpenSSL CA somewhat more
1731 feature-complete. In this mode, most diagnostic messages logged
1732 after entering the event loop are logged via syslog(3) rather than
1737 * Added support for X448 and Ed448. Heavily based on original work by
1742 * Extend OSSL_STORE with capabilities to search and to narrow the set of
1743 objects loaded. This adds the functions OSSL_STORE_expect() and
1744 OSSL_STORE_find() as well as needed tools to construct searches and
1745 get the search data out of them.
1749 * Support for TLSv1.3 added. Note that users upgrading from an earlier
1750 version of OpenSSL should review their configuration settings to ensure
1751 that they are still appropriate for TLSv1.3. For further information see:
1752 <https://wiki.openssl.org/index.php/TLS1.3>
1756 * Grand redesign of the OpenSSL random generator
1758 The default RAND method now utilizes an AES-CTR DRBG according to
1759 NIST standard SP 800-90Ar1. The new random generator is essentially
1760 a port of the default random generator from the OpenSSL FIPS 2.0
1761 object module. It is a hybrid deterministic random bit generator
1762 using an AES-CTR bit stream and which seeds and reseeds itself
1763 automatically using trusted system entropy sources.
1765 Some of its new features are:
1766 - Support for multiple DRBG instances with seed chaining.
1767 - The default RAND method makes use of a DRBG.
1768 - There is a public and private DRBG instance.
1769 - The DRBG instances are fork-safe.
1770 - Keep all global DRBG instances on the secure heap if it is enabled.
1771 - The public and private DRBG instance are per thread for lock free
1774 *Paul Dale, Benjamin Kaduk, Kurt Roeckx, Rich Salz, Matthias St. Pierre*
1776 * Changed Configure so it only says what it does and doesn't dump
1777 so much data. Instead, ./configdata.pm should be used as a script
1778 to display all sorts of configuration data.
1782 * Added processing of "make variables" to Configure.
1786 * Added SHA512/224 and SHA512/256 algorithm support.
1790 * The last traces of Netware support, first removed in 1.1.0, have
1795 * Get rid of Makefile.shared, and in the process, make the processing
1796 of certain files (rc.obj, or the .def/.map/.opt files produced from
1797 the ordinal files) more visible and hopefully easier to trace and
1798 debug (or make silent).
1802 * Make it possible to have environment variable assignments as
1803 arguments to config / Configure.
1807 * Add multi-prime RSA (RFC 8017) support.
1811 * Add SM3 implemented according to GB/T 32905-2016
1812 *Jack Lloyd <jack.lloyd@ribose.com>,*
1813 *Ronald Tse <ronald.tse@ribose.com>,*
1814 *Erick Borsboom <erick.borsboom@ribose.com>*
1816 * Add 'Maximum Fragment Length' TLS extension negotiation and support
1817 as documented in RFC6066.
1818 Based on a patch from Tomasz Moń
1820 *Filipe Raimundo da Silva*
1822 * Add SM4 implemented according to GB/T 32907-2016.
1823 *Jack Lloyd <jack.lloyd@ribose.com>,*
1824 *Ronald Tse <ronald.tse@ribose.com>,*
1825 *Erick Borsboom <erick.borsboom@ribose.com>*
1827 * Reimplement -newreq-nodes and ERR_error_string_n; the
1828 original author does not agree with the license change.
1832 * Add ARIA AEAD TLS support.
1836 * Some macro definitions to support VS6 have been removed. Visual
1837 Studio 6 has not worked since 1.1.0
1841 * Add ERR_clear_last_mark(), to allow callers to clear the last mark
1842 without clearing the errors.
1846 * Add "atfork" functions. If building on a system that without
1847 pthreads, see doc/man3/OPENSSL_fork_prepare.pod for application
1848 requirements. The RAND facility now uses/requires this.
1856 * The UI API becomes a permanent and integral part of libcrypto, i.e.
1857 not possible to disable entirely. However, it's still possible to
1858 disable the console reading UI method, UI_OpenSSL() (use UI_null()
1861 To disable, configure with 'no-ui-console'. 'no-ui' is still
1862 possible to use as an alias. Check at compile time with the
1863 macro OPENSSL_NO_UI_CONSOLE. The macro OPENSSL_NO_UI is still
1864 possible to check and is an alias for OPENSSL_NO_UI_CONSOLE.
1868 * Add a STORE module, which implements a uniform and URI based reader of
1869 stores that can contain keys, certificates, CRLs and numerous other
1870 objects. The main API is loosely based on a few stdio functions,
1871 and includes OSSL_STORE_open, OSSL_STORE_load, OSSL_STORE_eof,
1872 OSSL_STORE_error and OSSL_STORE_close.
1873 The implementation uses backends called "loaders" to implement arbitrary
1874 URI schemes. There is one built in "loader" for the 'file' scheme.
1878 * Add devcrypto engine. This has been implemented against cryptodev-linux,
1879 then adjusted to work on FreeBSD 8.4 as well.
1880 Enable by configuring with 'enable-devcryptoeng'. This is done by default
1881 on BSD implementations, as cryptodev.h is assumed to exist on all of them.
1885 * Module names can prefixed with OSSL_ or OPENSSL_. This affects
1886 util/mkerr.pl, which is adapted to allow those prefixes, leading to
1887 error code calls like this:
1889 OSSL_FOOerr(OSSL_FOO_F_SOMETHING, OSSL_FOO_R_WHATEVER);
1891 With this change, we claim the namespaces OSSL and OPENSSL in a manner
1892 that can be encoded in C. For the foreseeable future, this will only
1895 *Richard Levitte and Tim Hudson*
1897 * Removed BSD cryptodev engine.
1901 * Add a build target 'build_all_generated', to build all generated files
1902 and only that. This can be used to prepare everything that requires
1903 things like perl for a system that lacks perl and then move everything
1904 to that system and do the rest of the build there.
1908 * In the UI interface, make it possible to duplicate the user data. This
1909 can be used by engines that need to retain the data for a longer time
1910 than just the call where this user data is passed.
1914 * Ignore the '-named_curve auto' value for compatibility of applications
1917 *Tomas Mraz <tmraz@fedoraproject.org>*
1919 * Fragmented SSL/TLS alerts are no longer accepted. An alert message is 2
1920 bytes long. In theory it is permissible in SSLv3 - TLSv1.2 to fragment such
1921 alerts across multiple records (some of which could be empty). In practice
1922 it make no sense to send an empty alert record, or to fragment one. TLSv1.3
1923 prohibits this altogether and other libraries (BoringSSL, NSS) do not
1924 support this at all. Supporting it adds significant complexity to the
1925 record layer, and its removal is unlikely to cause interoperability
1930 * Add the ASN.1 types INT32, UINT32, INT64, UINT64 and variants prefixed
1931 with Z. These are meant to replace LONG and ZLONG and to be size safe.
1932 The use of LONG and ZLONG is discouraged and scheduled for deprecation
1937 * Add the 'z' and 'j' modifiers to BIO_printf() et al formatting string,
1938 'z' is to be used for [s]size_t, and 'j' - with [u]int64_t.
1940 *Richard Levitte, Andy Polyakov*
1942 * Add EC_KEY_get0_engine(), which does for EC_KEY what RSA_get0_engine()
1947 * Have 'config' recognise 64-bit mingw and choose 'mingw64' as the target
1948 platform rather than 'mingw'.
1952 * The functions X509_STORE_add_cert and X509_STORE_add_crl return
1953 success if they are asked to add an object which already exists
1954 in the store. This change cascades to other functions which load
1955 certificates and CRLs.
1959 * x86_64 assembly pack: annotate code with DWARF CFI directives to
1960 facilitate stack unwinding even from assembly subroutines.
1964 * Remove VAX C specific definitions of OPENSSL_EXPORT, OPENSSL_EXTERN.
1965 Also remove OPENSSL_GLOBAL entirely, as it became a no-op.
1969 * Remove the VMS-specific reimplementation of gmtime from crypto/o_times.c.
1970 VMS C's RTL has a fully up to date gmtime() and gmtime_r() since V7.1,
1971 which is the minimum version we support.
1975 * Certificate time validation (X509_cmp_time) enforces stricter
1976 compliance with RFC 5280. Fractional seconds and timezone offsets
1977 are no longer allowed.
1981 * Add support for ARIA
1985 * s_client will now send the Server Name Indication (SNI) extension by
1986 default unless the new "-noservername" option is used. The server name is
1987 based on the host provided to the "-connect" option unless overridden by
1988 using "-servername".
1992 * Add support for SipHash
1996 * OpenSSL now fails if it receives an unrecognised record type in TLS1.0
1997 or TLS1.1. Previously this only happened in SSLv3 and TLS1.2. This is to
1998 prevent issues where no progress is being made and the peer continually
1999 sends unrecognised record types, using up resources processing them.
2003 * 'openssl passwd' can now produce SHA256 and SHA512 based output,
2004 using the algorithm defined in
2005 <https://www.akkadia.org/drepper/SHA-crypt.txt>
2009 * Heartbeat support has been removed; the ABI is changed for now.
2011 *Richard Levitte, Rich Salz*
2013 * Support for SSL_OP_NO_ENCRYPT_THEN_MAC in SSL_CONF_cmd.
2017 * The RSA "null" method, which was partially supported to avoid patent
2018 issues, has been replaced to always returns NULL.
2025 ### Changes between 1.1.0k and 1.1.0l [10 Sep 2019]
2027 * For built-in EC curves, ensure an EC_GROUP built from the curve name is
2028 used even when parsing explicit parameters, when loading a encoded key
2029 or calling `EC_GROUP_new_from_ecpkparameters()`/
2030 `EC_GROUP_new_from_ecparameters()`.
2031 This prevents bypass of security hardening and performance gains,
2032 especially for curves with specialized EC_METHODs.
2033 By default, if a key encoded with explicit parameters is loaded and later
2034 encoded, the output is still encoded with explicit parameters, even if
2035 internally a "named" EC_GROUP is used for computation.
2039 * Compute ECC cofactors if not provided during EC_GROUP construction. Before
2040 this change, EC_GROUP_set_generator would accept order and/or cofactor as
2041 NULL. After this change, only the cofactor parameter can be NULL. It also
2042 does some minimal sanity checks on the passed order.
2047 * Fixed a padding oracle in PKCS7_dataDecode and CMS_decrypt_set1_pkey.
2048 An attack is simple, if the first CMS_recipientInfo is valid but the
2049 second CMS_recipientInfo is chosen ciphertext. If the second
2050 recipientInfo decodes to PKCS #1 v1.5 form plaintext, the correct
2051 encryption key will be replaced by garbage, and the message cannot be
2052 decoded, but if the RSA decryption fails, the correct encryption key is
2053 used and the recipient will not notice the attack.
2054 As a work around for this potential attack the length of the decrypted
2055 key must be equal to the cipher default key length, in case the
2056 certifiate is not given and all recipientInfo are tried out.
2057 The old behaviour can be re-enabled in the CMS code by setting the
2058 CMS_DEBUG_DECRYPT flag.
2063 * Use Windows installation paths in the mingw builds
2065 Mingw isn't a POSIX environment per se, which means that Windows
2066 paths should be used for installation.
2071 ### Changes between 1.1.0j and 1.1.0k [28 May 2019]
2073 * Change the default RSA, DSA and DH size to 2048 bit instead of 1024.
2074 This changes the size when using the genpkey app when no size is given. It
2075 fixes an omission in earlier changes that changed all RSA, DSA and DH
2076 generation apps to use 2048 bits by default.
2080 * Prevent over long nonces in ChaCha20-Poly1305.
2082 ChaCha20-Poly1305 is an AEAD cipher, and requires a unique nonce input
2083 for every encryption operation. RFC 7539 specifies that the nonce value
2084 (IV) should be 96 bits (12 bytes). OpenSSL allows a variable nonce length
2085 and front pads the nonce with 0 bytes if it is less than 12
2086 bytes. However it also incorrectly allows a nonce to be set of up to 16
2087 bytes. In this case only the last 12 bytes are significant and any
2088 additional leading bytes are ignored.
2090 It is a requirement of using this cipher that nonce values are
2091 unique. Messages encrypted using a reused nonce value are susceptible to
2092 serious confidentiality and integrity attacks. If an application changes
2093 the default nonce length to be longer than 12 bytes and then makes a
2094 change to the leading bytes of the nonce expecting the new value to be a
2095 new unique nonce then such an application could inadvertently encrypt
2096 messages with a reused nonce.
2098 Additionally the ignored bytes in a long nonce are not covered by the
2099 integrity guarantee of this cipher. Any application that relies on the
2100 integrity of these ignored leading bytes of a long nonce may be further
2101 affected. Any OpenSSL internal use of this cipher, including in SSL/TLS,
2102 is safe because no such use sets such a long nonce value. However user
2103 applications that use this cipher directly and set a non-default nonce
2104 length to be longer than 12 bytes may be vulnerable.
2106 This issue was reported to OpenSSL on 16th of March 2019 by Joran Dirk
2112 * Added SCA hardening for modular field inversion in EC_GROUP through
2113 a new dedicated field_inv() pointer in EC_METHOD.
2114 This also addresses a leakage affecting conversions from projective
2115 to affine coordinates.
2117 *Billy Bob Brumley, Nicola Tuveri*
2119 * Fix a use after free bug in d2i_X509_PUBKEY when overwriting a
2120 re-used X509_PUBKEY object if the second PUBKEY is malformed.
2124 * Move strictness check from EVP_PKEY_asn1_new() to EVP_PKEY_asn1_add0().
2128 * Remove the 'dist' target and add a tarball building script. The
2129 'dist' target has fallen out of use, and it shouldn't be
2130 necessary to configure just to create a source distribution.
2134 ### Changes between 1.1.0i and 1.1.0j [20 Nov 2018]
2136 * Timing vulnerability in DSA signature generation
2138 The OpenSSL DSA signature algorithm has been shown to be vulnerable to a
2139 timing side channel attack. An attacker could use variations in the signing
2140 algorithm to recover the private key.
2142 This issue was reported to OpenSSL on 16th October 2018 by Samuel Weiser.
2147 * Timing vulnerability in ECDSA signature generation
2149 The OpenSSL ECDSA signature algorithm has been shown to be vulnerable to a
2150 timing side channel attack. An attacker could use variations in the signing
2151 algorithm to recover the private key.
2153 This issue was reported to OpenSSL on 25th October 2018 by Samuel Weiser.
2158 * Add coordinate blinding for EC_POINT and implement projective
2159 coordinate blinding for generic prime curves as a countermeasure to
2160 chosen point SCA attacks.
2162 *Sohaib ul Hassan, Nicola Tuveri, Billy Bob Brumley*
2164 ### Changes between 1.1.0h and 1.1.0i [14 Aug 2018]
2166 * Client DoS due to large DH parameter
2168 During key agreement in a TLS handshake using a DH(E) based ciphersuite a
2169 malicious server can send a very large prime value to the client. This will
2170 cause the client to spend an unreasonably long period of time generating a
2171 key for this prime resulting in a hang until the client has finished. This
2172 could be exploited in a Denial Of Service attack.
2174 This issue was reported to OpenSSL on 5th June 2018 by Guido Vranken
2179 * Cache timing vulnerability in RSA Key Generation
2181 The OpenSSL RSA Key generation algorithm has been shown to be vulnerable to
2182 a cache timing side channel attack. An attacker with sufficient access to
2183 mount cache timing attacks during the RSA key generation process could
2184 recover the private key.
2186 This issue was reported to OpenSSL on 4th April 2018 by Alejandro Cabrera
2187 Aldaya, Billy Brumley, Cesar Pereida Garcia and Luis Manuel Alvarez Tapia.
2192 * Make EVP_PKEY_asn1_new() a bit stricter about its input. A NULL pem_str
2193 parameter is no longer accepted, as it leads to a corrupt table. NULL
2194 pem_str is reserved for alias entries only.
2198 * Revert blinding in ECDSA sign and instead make problematic addition
2199 length-invariant. Switch even to fixed-length Montgomery multiplication.
2203 * Change generating and checking of primes so that the error rate of not
2204 being prime depends on the intended use based on the size of the input.
2205 For larger primes this will result in more rounds of Miller-Rabin.
2206 The maximal error rate for primes with more than 1080 bits is lowered
2209 *Kurt Roeckx, Annie Yousar*
2211 * Increase the number of Miller-Rabin rounds for DSA key generating to 64.
2215 * Add blinding to ECDSA and DSA signatures to protect against side channel
2216 attacks discovered by Keegan Ryan (NCC Group).
2220 * When unlocking a pass phrase protected PEM file or PKCS#8 container, we
2221 now allow empty (zero character) pass phrases.
2225 * Certificate time validation (X509_cmp_time) enforces stricter
2226 compliance with RFC 5280. Fractional seconds and timezone offsets
2227 are no longer allowed.
2231 * Fixed a text canonicalisation bug in CMS
2233 Where a CMS detached signature is used with text content the text goes
2234 through a canonicalisation process first prior to signing or verifying a
2235 signature. This process strips trailing space at the end of lines, converts
2236 line terminators to CRLF and removes additional trailing line terminators
2237 at the end of a file. A bug in the canonicalisation process meant that
2238 some characters, such as form-feed, were incorrectly treated as whitespace
2239 and removed. This is contrary to the specification (RFC5485). This fix
2240 could mean that detached text data signed with an earlier version of
2241 OpenSSL 1.1.0 may fail to verify using the fixed version, or text data
2242 signed with a fixed OpenSSL may fail to verify with an earlier version of
2243 OpenSSL 1.1.0. A workaround is to only verify the canonicalised text data
2244 and use the "-binary" flag (for the "cms" command line application) or set
2245 the SMIME_BINARY/PKCS7_BINARY/CMS_BINARY flags (if using CMS_verify()).
2249 ### Changes between 1.1.0g and 1.1.0h [27 Mar 2018]
2251 * Constructed ASN.1 types with a recursive definition could exceed the stack
2253 Constructed ASN.1 types with a recursive definition (such as can be found
2254 in PKCS7) could eventually exceed the stack given malicious input with
2255 excessive recursion. This could result in a Denial Of Service attack. There
2256 are no such structures used within SSL/TLS that come from untrusted sources
2257 so this is considered safe.
2259 This issue was reported to OpenSSL on 4th January 2018 by the OSS-fuzz
2265 * Incorrect CRYPTO_memcmp on HP-UX PA-RISC
2267 Because of an implementation bug the PA-RISC CRYPTO_memcmp function is
2268 effectively reduced to only comparing the least significant bit of each
2269 byte. This allows an attacker to forge messages that would be considered as
2270 authenticated in an amount of tries lower than that guaranteed by the
2271 security claims of the scheme. The module can only be compiled by the
2272 HP-UX assembler, so that only HP-UX PA-RISC targets are affected.
2274 This issue was reported to OpenSSL on 2nd March 2018 by Peter Waltenberg
2280 * Add a build target 'build_all_generated', to build all generated files
2281 and only that. This can be used to prepare everything that requires
2282 things like perl for a system that lacks perl and then move everything
2283 to that system and do the rest of the build there.
2287 * Backport SSL_OP_NO_RENGOTIATION
2289 OpenSSL 1.0.2 and below had the ability to disable renegotiation using the
2290 (undocumented) SSL3_FLAGS_NO_RENEGOTIATE_CIPHERS flag. Due to the opacity
2291 changes this is no longer possible in 1.1.0. Therefore the new
2292 SSL_OP_NO_RENEGOTIATION option from 1.1.1-dev has been backported to
2293 1.1.0 to provide equivalent functionality.
2295 Note that if an application built against 1.1.0h headers (or above) is run
2296 using an older version of 1.1.0 (prior to 1.1.0h) then the option will be
2297 accepted but nothing will happen, i.e. renegotiation will not be prevented.
2301 * Removed the OS390-Unix config target. It relied on a script that doesn't
2306 * rsaz_1024_mul_avx2 overflow bug on x86_64
2308 There is an overflow bug in the AVX2 Montgomery multiplication procedure
2309 used in exponentiation with 1024-bit moduli. No EC algorithms are affected.
2310 Analysis suggests that attacks against RSA and DSA as a result of this
2311 defect would be very difficult to perform and are not believed likely.
2312 Attacks against DH1024 are considered just feasible, because most of the
2313 work necessary to deduce information about a private key may be performed
2314 offline. The amount of resources required for such an attack would be
2315 significant. However, for an attack on TLS to be meaningful, the server
2316 would have to share the DH1024 private key among multiple clients, which is
2317 no longer an option since CVE-2016-0701.
2319 This only affects processors that support the AVX2 but not ADX extensions
2320 like Intel Haswell (4th generation).
2322 This issue was reported to OpenSSL by David Benjamin (Google). The issue
2323 was originally found via the OSS-Fuzz project.
2328 ### Changes between 1.1.0f and 1.1.0g [2 Nov 2017]
2330 * bn_sqrx8x_internal carry bug on x86_64
2332 There is a carry propagating bug in the x86_64 Montgomery squaring
2333 procedure. No EC algorithms are affected. Analysis suggests that attacks
2334 against RSA and DSA as a result of this defect would be very difficult to
2335 perform and are not believed likely. Attacks against DH are considered just
2336 feasible (although very difficult) because most of the work necessary to
2337 deduce information about a private key may be performed offline. The amount
2338 of resources required for such an attack would be very significant and
2339 likely only accessible to a limited number of attackers. An attacker would
2340 additionally need online access to an unpatched system using the target
2341 private key in a scenario with persistent DH parameters and a private
2342 key that is shared between multiple clients.
2344 This only affects processors that support the BMI1, BMI2 and ADX extensions
2345 like Intel Broadwell (5th generation) and later or AMD Ryzen.
2347 This issue was reported to OpenSSL by the OSS-Fuzz project.
2352 * Malformed X.509 IPAddressFamily could cause OOB read
2354 If an X.509 certificate has a malformed IPAddressFamily extension,
2355 OpenSSL could do a one-byte buffer overread. The most likely result
2356 would be an erroneous display of the certificate in text format.
2358 This issue was reported to OpenSSL by the OSS-Fuzz project.
2363 ### Changes between 1.1.0e and 1.1.0f [25 May 2017]
2365 * Have 'config' recognise 64-bit mingw and choose 'mingw64' as the target
2366 platform rather than 'mingw'.
2370 * Remove the VMS-specific reimplementation of gmtime from crypto/o_times.c.
2371 VMS C's RTL has a fully up to date gmtime() and gmtime_r() since V7.1,
2372 which is the minimum version we support.
2376 ### Changes between 1.1.0d and 1.1.0e [16 Feb 2017]
2378 * Encrypt-Then-Mac renegotiation crash
2380 During a renegotiation handshake if the Encrypt-Then-Mac extension is
2381 negotiated where it was not in the original handshake (or vice-versa) then
2382 this can cause OpenSSL to crash (dependant on ciphersuite). Both clients
2383 and servers are affected.
2385 This issue was reported to OpenSSL by Joe Orton (Red Hat).
2390 ### Changes between 1.1.0c and 1.1.0d [26 Jan 2017]
2392 * Truncated packet could crash via OOB read
2394 If one side of an SSL/TLS path is running on a 32-bit host and a specific
2395 cipher is being used, then a truncated packet can cause that host to
2396 perform an out-of-bounds read, usually resulting in a crash.
2398 This issue was reported to OpenSSL by Robert Święcki of Google.
2403 * Bad (EC)DHE parameters cause a client crash
2405 If a malicious server supplies bad parameters for a DHE or ECDHE key
2406 exchange then this can result in the client attempting to dereference a
2407 NULL pointer leading to a client crash. This could be exploited in a Denial
2410 This issue was reported to OpenSSL by Guido Vranken.
2415 * BN_mod_exp may produce incorrect results on x86_64
2417 There is a carry propagating bug in the x86_64 Montgomery squaring
2418 procedure. No EC algorithms are affected. Analysis suggests that attacks
2419 against RSA and DSA as a result of this defect would be very difficult to
2420 perform and are not believed likely. Attacks against DH are considered just
2421 feasible (although very difficult) because most of the work necessary to
2422 deduce information about a private key may be performed offline. The amount
2423 of resources required for such an attack would be very significant and
2424 likely only accessible to a limited number of attackers. An attacker would
2425 additionally need online access to an unpatched system using the target
2426 private key in a scenario with persistent DH parameters and a private
2427 key that is shared between multiple clients. For example this can occur by
2428 default in OpenSSL DHE based SSL/TLS ciphersuites. Note: This issue is very
2429 similar to CVE-2015-3193 but must be treated as a separate problem.
2431 This issue was reported to OpenSSL by the OSS-Fuzz project.
2436 ### Changes between 1.1.0b and 1.1.0c [10 Nov 2016]
2438 * ChaCha20/Poly1305 heap-buffer-overflow
2440 TLS connections using `*-CHACHA20-POLY1305` ciphersuites are susceptible to
2441 a DoS attack by corrupting larger payloads. This can result in an OpenSSL
2442 crash. This issue is not considered to be exploitable beyond a DoS.
2444 This issue was reported to OpenSSL by Robert Święcki (Google Security Team)
2449 * CMS Null dereference
2451 Applications parsing invalid CMS structures can crash with a NULL pointer
2452 dereference. This is caused by a bug in the handling of the ASN.1 CHOICE
2453 type in OpenSSL 1.1.0 which can result in a NULL value being passed to the
2454 structure callback if an attempt is made to free certain invalid encodings.
2455 Only CHOICE structures using a callback which do not handle NULL value are
2458 This issue was reported to OpenSSL by Tyler Nighswander of ForAllSecure.
2463 * Montgomery multiplication may produce incorrect results
2465 There is a carry propagating bug in the Broadwell-specific Montgomery
2466 multiplication procedure that handles input lengths divisible by, but
2467 longer than 256 bits. Analysis suggests that attacks against RSA, DSA
2468 and DH private keys are impossible. This is because the subroutine in
2469 question is not used in operations with the private key itself and an input
2470 of the attacker's direct choice. Otherwise the bug can manifest itself as
2471 transient authentication and key negotiation failures or reproducible
2472 erroneous outcome of public-key operations with specially crafted input.
2473 Among EC algorithms only Brainpool P-512 curves are affected and one
2474 presumably can attack ECDH key negotiation. Impact was not analyzed in
2475 detail, because pre-requisites for attack are considered unlikely. Namely
2476 multiple clients have to choose the curve in question and the server has to
2477 share the private key among them, neither of which is default behaviour.
2478 Even then only clients that chose the curve will be affected.
2480 This issue was publicly reported as transient failures and was not
2481 initially recognized as a security issue. Thanks to Richard Morgan for
2482 providing reproducible case.
2487 * Removed automatic addition of RPATH in shared libraries and executables,
2488 as this was a remainder from OpenSSL 1.0.x and isn't needed any more.
2492 ### Changes between 1.1.0a and 1.1.0b [26 Sep 2016]
2494 * Fix Use After Free for large message sizes
2496 The patch applied to address CVE-2016-6307 resulted in an issue where if a
2497 message larger than approx 16k is received then the underlying buffer to
2498 store the incoming message is reallocated and moved. Unfortunately a
2499 dangling pointer to the old location is left which results in an attempt to
2500 write to the previously freed location. This is likely to result in a
2501 crash, however it could potentially lead to execution of arbitrary code.
2503 This issue only affects OpenSSL 1.1.0a.
2505 This issue was reported to OpenSSL by Robert Święcki.
2510 ### Changes between 1.1.0 and 1.1.0a [22 Sep 2016]
2512 * OCSP Status Request extension unbounded memory growth
2514 A malicious client can send an excessively large OCSP Status Request
2515 extension. If that client continually requests renegotiation, sending a
2516 large OCSP Status Request extension each time, then there will be unbounded
2517 memory growth on the server. This will eventually lead to a Denial Of
2518 Service attack through memory exhaustion. Servers with a default
2519 configuration are vulnerable even if they do not support OCSP. Builds using
2520 the "no-ocsp" build time option are not affected.
2522 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
2527 * SSL_peek() hang on empty record
2529 OpenSSL 1.1.0 SSL/TLS will hang during a call to SSL_peek() if the peer
2530 sends an empty record. This could be exploited by a malicious peer in a
2531 Denial Of Service attack.
2533 This issue was reported to OpenSSL by Alex Gaynor.
2538 * Excessive allocation of memory in tls_get_message_header() and
2539 dtls1_preprocess_fragment()
2541 A (D)TLS message includes 3 bytes for its length in the header for the
2542 message. This would allow for messages up to 16Mb in length. Messages of
2543 this length are excessive and OpenSSL includes a check to ensure that a
2544 peer is sending reasonably sized messages in order to avoid too much memory
2545 being consumed to service a connection. A flaw in the logic of version
2546 1.1.0 means that memory for the message is allocated too early, prior to
2547 the excessive message length check. Due to way memory is allocated in
2548 OpenSSL this could mean an attacker could force up to 21Mb to be allocated
2549 to service a connection. This could lead to a Denial of Service through
2550 memory exhaustion. However, the excessive message length check still takes
2551 place, and this would cause the connection to immediately fail. Assuming
2552 that the application calls SSL_free() on the failed connection in a timely
2553 manner then the 21Mb of allocated memory will then be immediately freed
2554 again. Therefore the excessive memory allocation will be transitory in
2555 nature. This then means that there is only a security impact if:
2557 1) The application does not call SSL_free() in a timely manner in the event
2558 that the connection fails
2560 2) The application is working in a constrained environment where there is
2561 very little free memory
2563 3) The attacker initiates multiple connection attempts such that there are
2564 multiple connections in a state where memory has been allocated for the
2565 connection; SSL_free() has not yet been called; and there is insufficient
2566 memory to service the multiple requests.
2568 Except in the instance of (1) above any Denial Of Service is likely to be
2569 transitory because as soon as the connection fails the memory is
2570 subsequently freed again in the SSL_free() call. However there is an
2571 increased risk during this period of application crashes due to the lack of
2572 memory - which would then mean a more serious Denial of Service.
2574 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
2575 (CVE-2016-6307 and CVE-2016-6308)
2579 * solaris-x86-cc, i.e. 32-bit configuration with vendor compiler,
2580 had to be removed. Primary reason is that vendor assembler can't
2581 assemble our modules with -KPIC flag. As result it, assembly
2582 support, was not even available as option. But its lack means
2583 lack of side-channel resistant code, which is incompatible with
2584 security by todays standards. Fortunately gcc is readily available
2585 prepackaged option, which we firmly point at...
2589 ### Changes between 1.0.2h and 1.1.0 [25 Aug 2016]
2591 * Windows command-line tool supports UTF-8 opt-in option for arguments
2592 and console input. Setting OPENSSL_WIN32_UTF8 environment variable
2593 (to any value) allows Windows user to access PKCS#12 file generated
2594 with Windows CryptoAPI and protected with non-ASCII password, as well
2595 as files generated under UTF-8 locale on Linux also protected with
2600 * To mitigate the SWEET32 attack [CVE-2016-2183][], 3DES cipher suites
2601 have been disabled by default and removed from DEFAULT, just like RC4.
2602 See the RC4 item below to re-enable both.
2606 * The method for finding the storage location for the Windows RAND seed file
2607 has changed. First we check %RANDFILE%. If that is not set then we check
2608 the directories %HOME%, %USERPROFILE% and %SYSTEMROOT% in that order. If
2609 all else fails we fall back to C:\.
2613 * The EVP_EncryptUpdate() function has had its return type changed from void
2614 to int. A return of 0 indicates and error while a return of 1 indicates
2619 * The flags RSA_FLAG_NO_CONSTTIME, DSA_FLAG_NO_EXP_CONSTTIME and
2620 DH_FLAG_NO_EXP_CONSTTIME which previously provided the ability to switch
2621 off the constant time implementation for RSA, DSA and DH have been made
2622 no-ops and deprecated.
2626 * Windows RAND implementation was simplified to only get entropy by
2627 calling CryptGenRandom(). Various other RAND-related tickets
2630 *Joseph Wylie Yandle, Rich Salz*
2632 * The stack and lhash API's were renamed to start with `OPENSSL_SK_`
2633 and `OPENSSL_LH_`, respectively. The old names are available
2634 with API compatibility. They new names are now completely documented.
2638 * Unify TYPE_up_ref(obj) methods signature.
2639 SSL_CTX_up_ref(), SSL_up_ref(), X509_up_ref(), EVP_PKEY_up_ref(),
2640 X509_CRL_up_ref(), X509_OBJECT_up_ref_count() methods are now returning an
2641 int (instead of void) like all others TYPE_up_ref() methods.
2642 So now these methods also check the return value of CRYPTO_atomic_add(),
2643 and the validity of object reference counter.
2645 *fdasilvayy@gmail.com*
2647 * With Windows Visual Studio builds, the .pdb files are installed
2648 alongside the installed libraries and executables. For a static
2649 library installation, ossl_static.pdb is the associate compiler
2650 generated .pdb file to be used when linking programs.
2654 * Remove openssl.spec. Packaging files belong with the packagers.
2658 * Automatic Darwin/OSX configuration has had a refresh, it will now
2659 recognise x86_64 architectures automatically. You can still decide
2660 to build for a different bitness with the environment variable
2661 KERNEL_BITS (can be 32 or 64), for example:
2663 KERNEL_BITS=32 ./config
2667 * Change default algorithms in pkcs8 utility to use PKCS#5 v2.0,
2668 256 bit AES and HMAC with SHA256.
2672 * Remove support for MIPS o32 ABI on IRIX (and IRIX only).
2676 * Triple-DES ciphers have been moved from HIGH to MEDIUM.
2680 * To enable users to have their own config files and build file templates,
2681 Configure looks in the directory indicated by the environment variable
2682 OPENSSL_LOCAL_CONFIG_DIR as well as the in-source Configurations/
2683 directory. On VMS, OPENSSL_LOCAL_CONFIG_DIR is expected to be a logical
2684 name and is used as is.
2688 * The following datatypes were made opaque: X509_OBJECT, X509_STORE_CTX,
2689 X509_STORE, X509_LOOKUP, and X509_LOOKUP_METHOD. The unused type
2690 X509_CERT_FILE_CTX was removed.
2694 * "shared" builds are now the default. To create only static libraries use
2695 the "no-shared" Configure option.
2699 * Remove the no-aes, no-hmac, no-rsa, no-sha and no-md5 Configure options.
2700 All of these option have not worked for some while and are fundamental
2705 * Make various cleanup routines no-ops and mark them as deprecated. Most
2706 global cleanup functions are no longer required because they are handled
2707 via auto-deinit (see OPENSSL_init_crypto and OPENSSL_init_ssl man pages).
2708 Explicitly de-initing can cause problems (e.g. where a library that uses
2709 OpenSSL de-inits, but an application is still using it). The affected
2710 functions are CONF_modules_free(), ENGINE_cleanup(), OBJ_cleanup(),
2711 EVP_cleanup(), BIO_sock_cleanup(), CRYPTO_cleanup_all_ex_data(),
2712 RAND_cleanup(), SSL_COMP_free_compression_methods(), ERR_free_strings() and
2713 COMP_zlib_cleanup().
2717 * --strict-warnings no longer enables runtime debugging options
2718 such as REF_DEBUG. Instead, debug options are automatically
2719 enabled with '--debug' builds.
2721 *Andy Polyakov, Emilia Käsper*
2723 * Made DH and DH_METHOD opaque. The structures for managing DH objects
2724 have been moved out of the public header files. New functions for managing
2725 these have been added.
2729 * Made RSA and RSA_METHOD opaque. The structures for managing RSA
2730 objects have been moved out of the public header files. New
2731 functions for managing these have been added.
2735 * Made DSA and DSA_METHOD opaque. The structures for managing DSA objects
2736 have been moved out of the public header files. New functions for managing
2737 these have been added.
2741 * Made BIO and BIO_METHOD opaque. The structures for managing BIOs have been
2742 moved out of the public header files. New functions for managing these
2747 * Removed no-rijndael as a config option. Rijndael is an old name for AES.
2751 * Removed the mk1mf build scripts.
2755 * Headers are now wrapped, if necessary, with OPENSSL_NO_xxx, so
2756 it is always safe to #include a header now.
2760 * Removed the aged BC-32 config and all its supporting scripts
2764 * Removed support for Ultrix, Netware, and OS/2.
2768 * Add support for HKDF.
2770 *Alessandro Ghedini*
2772 * Add support for blake2b and blake2s
2776 * Added support for "pipelining". Ciphers that have the
2777 EVP_CIPH_FLAG_PIPELINE flag set have a capability to process multiple
2778 encryptions/decryptions simultaneously. There are currently no built-in
2779 ciphers with this property but the expectation is that engines will be able
2780 to offer it to significantly improve throughput. Support has been extended
2781 into libssl so that multiple records for a single connection can be
2782 processed in one go (for >=TLS 1.1).
2786 * Added the AFALG engine. This is an async capable engine which is able to
2787 offload work to the Linux kernel. In this initial version it only supports
2788 AES128-CBC. The kernel must be version 4.1.0 or greater.
2792 * OpenSSL now uses a new threading API. It is no longer necessary to
2793 set locking callbacks to use OpenSSL in a multi-threaded environment. There
2794 are two supported threading models: pthreads and windows threads. It is
2795 also possible to configure OpenSSL at compile time for "no-threads". The
2796 old threading API should no longer be used. The functions have been
2797 replaced with "no-op" compatibility macros.
2799 *Alessandro Ghedini, Matt Caswell*
2801 * Modify behavior of ALPN to invoke callback after SNI/servername
2802 callback, such that updates to the SSL_CTX affect ALPN.
2806 * Add SSL_CIPHER queries for authentication and key-exchange.
2810 * Changes to the DEFAULT cipherlist:
2811 - Prefer (EC)DHE handshakes over plain RSA.
2812 - Prefer AEAD ciphers over legacy ciphers.
2813 - Prefer ECDSA over RSA when both certificates are available.
2814 - Prefer TLSv1.2 ciphers/PRF.
2815 - Remove DSS, SEED, IDEA, CAMELLIA, and AES-CCM from the
2820 * Change the ECC default curve list to be this, in order: x25519,
2821 secp256r1, secp521r1, secp384r1.
2825 * RC4 based libssl ciphersuites are now classed as "weak" ciphers and are
2826 disabled by default. They can be re-enabled using the
2827 enable-weak-ssl-ciphers option to Configure.
2831 * If the server has ALPN configured, but supports no protocols that the
2832 client advertises, send a fatal "no_application_protocol" alert.
2833 This behaviour is SHALL in RFC 7301, though it isn't universally
2834 implemented by other servers.
2838 * Add X25519 support.
2839 Add ASN.1 and EVP_PKEY methods for X25519. This includes support
2840 for public and private key encoding using the format documented in
2841 draft-ietf-curdle-pkix-02. The corresponding EVP_PKEY method supports
2842 key generation and key derivation.
2844 TLS support complies with draft-ietf-tls-rfc4492bis-08 and uses
2849 * Deprecate SRP_VBASE_get_by_user.
2850 SRP_VBASE_get_by_user had inconsistent memory management behaviour.
2851 In order to fix an unavoidable memory leak [CVE-2016-0798][],
2852 SRP_VBASE_get_by_user was changed to ignore the "fake user" SRP
2853 seed, even if the seed is configured.
2855 Users should use SRP_VBASE_get1_by_user instead. Note that in
2856 SRP_VBASE_get1_by_user, caller must free the returned value. Note
2857 also that even though configuring the SRP seed attempts to hide
2858 invalid usernames by continuing the handshake with fake
2859 credentials, this behaviour is not constant time and no strong
2860 guarantees are made that the handshake is indistinguishable from
2861 that of a valid user.
2865 * Configuration change; it's now possible to build dynamic engines
2866 without having to build shared libraries and vice versa. This
2867 only applies to the engines in engines/, those in crypto/engine/
2868 will always be built into libcrypto (i.e. "static").
2870 Building dynamic engines is enabled by default; to disable, use
2871 the configuration option "disable-dynamic-engine".
2873 The only requirements for building dynamic engines are the
2874 presence of the DSO module and building with position independent
2875 code, so they will also automatically be disabled if configuring
2876 with "disable-dso" or "disable-pic".
2878 The macros OPENSSL_NO_STATIC_ENGINE and OPENSSL_NO_DYNAMIC_ENGINE
2879 are also taken away from openssl/opensslconf.h, as they are
2884 * Configuration change; if there is a known flag to compile
2885 position independent code, it will always be applied on the
2886 libcrypto and libssl object files, and never on the application
2887 object files. This means other libraries that use routines from
2888 libcrypto / libssl can be made into shared libraries regardless
2889 of how OpenSSL was configured.
2891 If this isn't desirable, the configuration options "disable-pic"
2892 or "no-pic" can be used to disable the use of PIC. This will
2893 also disable building shared libraries and dynamic engines.
2897 * Removed JPAKE code. It was experimental and has no wide use.
2901 * The INSTALL_PREFIX Makefile variable has been renamed to
2902 DESTDIR. That makes for less confusion on what this variable
2903 is for. Also, the configuration option --install_prefix is
2908 * Heartbeat for TLS has been removed and is disabled by default
2909 for DTLS; configure with enable-heartbeats. Code that uses the
2910 old #define's might need to be updated.
2912 *Emilia Käsper, Rich Salz*
2914 * Rename REF_CHECK to REF_DEBUG.
2918 * New "unified" build system
2920 The "unified" build system is aimed to be a common system for all
2921 platforms we support. With it comes new support for VMS.
2923 This system builds supports building in a different directory tree
2924 than the source tree. It produces one Makefile (for unix family
2925 or lookalikes), or one descrip.mms (for VMS).
2927 The source of information to make the Makefile / descrip.mms is
2928 small files called 'build.info', holding the necessary
2929 information for each directory with source to compile, and a
2930 template in Configurations, like unix-Makefile.tmpl or
2933 With this change, the library names were also renamed on Windows
2934 and on VMS. They now have names that are closer to the standard
2935 on Unix, and include the major version number, and in certain
2936 cases, the architecture they are built for. See "Notes on shared
2937 libraries" in INSTALL.
2939 We rely heavily on the perl module Text::Template.
2943 * Added support for auto-initialisation and de-initialisation of the library.
2944 OpenSSL no longer requires explicit init or deinit routines to be called,
2945 except in certain circumstances. See the OPENSSL_init_crypto() and
2946 OPENSSL_init_ssl() man pages for further information.
2950 * The arguments to the DTLSv1_listen function have changed. Specifically the
2951 "peer" argument is now expected to be a BIO_ADDR object.
2953 * Rewrite of BIO networking library. The BIO library lacked consistent
2954 support of IPv6, and adding it required some more extensive
2955 modifications. This introduces the BIO_ADDR and BIO_ADDRINFO types,
2956 which hold all types of addresses and chains of address information.
2957 It also introduces a new API, with functions like BIO_socket,
2958 BIO_connect, BIO_listen, BIO_lookup and a rewrite of BIO_accept.
2959 The source/sink BIOs BIO_s_connect, BIO_s_accept and BIO_s_datagram
2960 have been adapted accordingly.
2964 * RSA_padding_check_PKCS1_type_1 now accepts inputs with and without
2969 * CRIME protection: disable compression by default, even if OpenSSL is
2970 compiled with zlib enabled. Applications can still enable compression
2971 by calling SSL_CTX_clear_options(ctx, SSL_OP_NO_COMPRESSION), or by
2972 using the SSL_CONF library to configure compression.
2976 * The signature of the session callback configured with
2977 SSL_CTX_sess_set_get_cb was changed. The read-only input buffer
2978 was explicitly marked as `const unsigned char*` instead of
2983 * Always DPURIFY. Remove the use of uninitialized memory in the
2984 RNG, and other conditional uses of DPURIFY. This makes -DPURIFY a no-op.
2988 * Removed many obsolete configuration items, including
2989 DES_PTR, DES_RISC1, DES_RISC2, DES_INT
2990 MD2_CHAR, MD2_INT, MD2_LONG
2992 IDEA_SHORT, IDEA_LONG
2993 RC2_SHORT, RC2_LONG, RC4_LONG, RC4_CHUNK, RC4_INDEX
2995 *Rich Salz, with advice from Andy Polyakov*
2997 * Many BN internals have been moved to an internal header file.
2999 *Rich Salz with help from Andy Polyakov*
3001 * Configuration and writing out the results from it has changed.
3002 Files such as Makefile include/openssl/opensslconf.h and are now
3003 produced through general templates, such as Makefile.in and
3004 crypto/opensslconf.h.in and some help from the perl module
3007 Also, the center of configuration information is no longer
3008 Makefile. Instead, Configure produces a perl module in
3009 configdata.pm which holds most of the config data (in the hash
3010 table %config), the target data that comes from the target
3011 configuration in one of the `Configurations/*.conf` files (in
3016 * To clarify their intended purposes, the Configure options
3017 --prefix and --openssldir change their semantics, and become more
3018 straightforward and less interdependent.
3020 --prefix shall be used exclusively to give the location INSTALLTOP
3021 where programs, scripts, libraries, include files and manuals are
3022 going to be installed. The default is now /usr/local.
3024 --openssldir shall be used exclusively to give the default
3025 location OPENSSLDIR where certificates, private keys, CRLs are
3026 managed. This is also where the default openssl.cnf gets
3028 If the directory given with this option is a relative path, the
3029 values of both the --prefix value and the --openssldir value will
3030 be combined to become OPENSSLDIR.
3031 The default for --openssldir is INSTALLTOP/ssl.
3033 Anyone who uses --openssldir to specify where OpenSSL is to be
3034 installed MUST change to use --prefix instead.
3038 * The GOST engine was out of date and therefore it has been removed. An up
3039 to date GOST engine is now being maintained in an external repository.
3040 See: <https://wiki.openssl.org/index.php/Binaries>. Libssl still retains
3041 support for GOST ciphersuites (these are only activated if a GOST engine
3046 * EGD is no longer supported by default; use enable-egd when
3049 *Ben Kaduk and Rich Salz*
3051 * The distribution now has Makefile.in files, which are used to
3052 create Makefile's when Configure is run. *Configure must be run
3053 before trying to build now.*
3057 * The return value for SSL_CIPHER_description() for error conditions
3062 * Support for RFC6698/RFC7671 DANE TLSA peer authentication.
3064 Obtaining and performing DNSSEC validation of TLSA records is
3065 the application's responsibility. The application provides
3066 the TLSA records of its choice to OpenSSL, and these are then
3067 used to authenticate the peer.
3069 The TLSA records need not even come from DNS. They can, for
3070 example, be used to implement local end-entity certificate or
3071 trust-anchor "pinning", where the "pin" data takes the form
3072 of TLSA records, which can augment or replace verification
3073 based on the usual WebPKI public certification authorities.
3077 * Revert default OPENSSL_NO_DEPRECATED setting. Instead OpenSSL
3078 continues to support deprecated interfaces in default builds.
3079 However, applications are strongly advised to compile their
3080 source files with -DOPENSSL_API_COMPAT=0x10100000L, which hides
3081 the declarations of all interfaces deprecated in 0.9.8, 1.0.0
3082 or the 1.1.0 releases.
3084 In environments in which all applications have been ported to
3085 not use any deprecated interfaces OpenSSL's Configure script
3086 should be used with the --api=1.1.0 option to entirely remove
3087 support for the deprecated features from the library and
3088 unconditionally disable them in the installed headers.
3089 Essentially the same effect can be achieved with the "no-deprecated"
3090 argument to Configure, except that this will always restrict
3091 the build to just the latest API, rather than a fixed API
3094 As applications are ported to future revisions of the API,
3095 they should update their compile-time OPENSSL_API_COMPAT define
3096 accordingly, but in most cases should be able to continue to
3097 compile with later releases.
3099 The OPENSSL_API_COMPAT versions for 1.0.0, and 0.9.8 are
3100 0x10000000L and 0x00908000L, respectively. However those
3101 versions did not support the OPENSSL_API_COMPAT feature, and
3102 so applications are not typically tested for explicit support
3103 of just the undeprecated features of either release.
3107 * Add support for setting the minimum and maximum supported protocol.
3108 It can bet set via the SSL_set_min_proto_version() and
3109 SSL_set_max_proto_version(), or via the SSL_CONF's MinProtocol and
3110 MaxProtocol. It's recommended to use the new APIs to disable
3111 protocols instead of disabling individual protocols using
3112 SSL_set_options() or SSL_CONF's Protocol. This change also
3113 removes support for disabling TLS 1.2 in the OpenSSL TLS
3114 client at compile time by defining OPENSSL_NO_TLS1_2_CLIENT.
3118 * Support for ChaCha20 and Poly1305 added to libcrypto and libssl.
3122 * New EC_KEY_METHOD, this replaces the older ECDSA_METHOD and ECDH_METHOD
3123 and integrates ECDSA and ECDH functionality into EC. Implementations can
3124 now redirect key generation and no longer need to convert to or from
3127 Note: the ecdsa.h and ecdh.h headers are now no longer needed and just
3128 include the ec.h header file instead.
3132 * Remove support for all 40 and 56 bit ciphers. This includes all the export
3133 ciphers who are no longer supported and drops support the ephemeral RSA key
3134 exchange. The LOW ciphers currently doesn't have any ciphers in it.
3138 * Made EVP_MD_CTX, EVP_MD, EVP_CIPHER_CTX, EVP_CIPHER and HMAC_CTX
3139 opaque. For HMAC_CTX, the following constructors and destructors
3142 HMAC_CTX *HMAC_CTX_new(void);
3143 void HMAC_CTX_free(HMAC_CTX *ctx);
3145 For EVP_MD and EVP_CIPHER, complete APIs to create, fill and
3146 destroy such methods has been added. See EVP_MD_meth_new(3) and
3147 EVP_CIPHER_meth_new(3) for documentation.
3150 1) `EVP_MD_CTX_cleanup()`, `EVP_CIPHER_CTX_cleanup()` and
3151 `HMAC_CTX_cleanup()` were removed. `HMAC_CTX_reset()` and
3152 `EVP_MD_CTX_reset()` should be called instead to reinitialise
3153 an already created structure.
3154 2) For consistency with the majority of our object creators and
3155 destructors, `EVP_MD_CTX_(create|destroy)` were renamed to
3156 `EVP_MD_CTX_(new|free)`. The old names are retained as macros
3157 for deprecated builds.
3161 * Added ASYNC support. Libcrypto now includes the async sub-library to enable
3162 cryptographic operations to be performed asynchronously as long as an
3163 asynchronous capable engine is used. See the ASYNC_start_job() man page for
3164 further details. Libssl has also had this capability integrated with the
3165 introduction of the new mode SSL_MODE_ASYNC and associated error
3166 SSL_ERROR_WANT_ASYNC. See the SSL_CTX_set_mode() and SSL_get_error() man
3167 pages. This work was developed in partnership with Intel Corp.
3171 * SSL_{CTX_}set_ecdh_auto() has been removed and ECDH is support is
3172 always enabled now. If you want to disable the support you should
3173 exclude it using the list of supported ciphers. This also means that the
3174 "-no_ecdhe" option has been removed from s_server.
3178 * SSL_{CTX}_set_tmp_ecdh() which can set 1 EC curve now internally calls
3179 SSL_{CTX_}set1_curves() which can set a list.
3183 * Remove support for SSL_{CTX_}set_tmp_ecdh_callback(). You should set the
3184 curve you want to support using SSL_{CTX_}set1_curves().
3188 * State machine rewrite. The state machine code has been significantly
3189 refactored in order to remove much duplication of code and solve issues
3190 with the old code (see [ssl/statem/README.md](ssl/statem/README.md) for
3191 further details). This change does have some associated API changes.
3192 Notably the SSL_state() function has been removed and replaced by
3193 SSL_get_state which now returns an "OSSL_HANDSHAKE_STATE" instead of an int.
3194 SSL_set_state() has been removed altogether. The previous handshake states
3195 defined in ssl.h and ssl3.h have also been removed.
3199 * All instances of the string "ssleay" in the public API were replaced
3200 with OpenSSL (case-matching; e.g., OPENSSL_VERSION for #define's)
3201 Some error codes related to internal RSA_eay API's were renamed.
3205 * The demo files in crypto/threads were moved to demo/threads.
3209 * Removed obsolete engines: 4758cca, aep, atalla, cswift, nuron, gmp,
3212 *Matt Caswell, Rich Salz*
3214 * New ASN.1 embed macro.
3216 New ASN.1 macro ASN1_EMBED. This is the same as ASN1_SIMPLE except the
3217 structure is not allocated: it is part of the parent. That is instead of
3225 This reduces memory fragmentation and make it impossible to accidentally
3226 set a mandatory field to NULL.
3228 This currently only works for some fields specifically a SEQUENCE, CHOICE,
3229 or ASN1_STRING type which is part of a parent SEQUENCE. Since it is
3230 equivalent to ASN1_SIMPLE it cannot be tagged, OPTIONAL, SET OF or
3235 * Remove EVP_CHECK_DES_KEY, a compile-time option that never compiled.
3239 * Removed DES and RC4 ciphersuites from DEFAULT. Also removed RC2 although
3240 in 1.0.2 EXPORT was already removed and the only RC2 ciphersuite is also
3241 an EXPORT one. COMPLEMENTOFDEFAULT has been updated accordingly to add
3242 DES and RC4 ciphersuites.
3246 * Rewrite EVP_DecodeUpdate (base64 decoding) to fix several bugs.
3247 This changes the decoding behaviour for some invalid messages,
3248 though the change is mostly in the more lenient direction, and
3249 legacy behaviour is preserved as much as possible.
3253 * Fix no-stdio build.
3254 *David Woodhouse <David.Woodhouse@intel.com> and also*
3255 *Ivan Nestlerode <ivan.nestlerode@sonos.com>*
3257 * New testing framework
3258 The testing framework has been largely rewritten and is now using
3259 perl and the perl modules Test::Harness and an extended variant of
3260 Test::More called OpenSSL::Test to do its work. All test scripts in
3261 test/ have been rewritten into test recipes, and all direct calls to
3262 executables in test/Makefile have become individual recipes using the
3263 simplified testing OpenSSL::Test::Simple.
3265 For documentation on our testing modules, do:
3267 perldoc test/testlib/OpenSSL/Test/Simple.pm
3268 perldoc test/testlib/OpenSSL/Test.pm
3272 * Revamped memory debug; only -DCRYPTO_MDEBUG and -DCRYPTO_MDEBUG_ABORT
3273 are used; the latter aborts on memory leaks (usually checked on exit).
3274 Some undocumented "set malloc, etc., hooks" functions were removed
3275 and others were changed. All are now documented.
3279 * In DSA_generate_parameters_ex, if the provided seed is too short,
3282 *Rich Salz and Ismo Puustinen <ismo.puustinen@intel.com>*
3284 * Rewrite PSK to support ECDHE_PSK, DHE_PSK and RSA_PSK. Add ciphersuites
3285 from RFC4279, RFC4785, RFC5487, RFC5489.
3287 Thanks to Christian J. Dietrich and Giuseppe D'Angelo for the
3288 original RSA_PSK patch.
3292 * Dropped support for the SSL3_FLAGS_DELAY_CLIENT_FINISHED flag. This SSLeay
3293 era flag was never set throughout the codebase (only read). Also removed
3294 SSL3_FLAGS_POP_BUFFER which was only used if
3295 SSL3_FLAGS_DELAY_CLIENT_FINISHED was also set.
3299 * Changed the default name options in the "ca", "crl", "req" and "x509"
3300 to be "oneline" instead of "compat".
3304 * Remove SSL_OP_TLS_BLOCK_PADDING_BUG. This is SSLeay legacy, we're
3305 not aware of clients that still exhibit this bug, and the workaround
3306 hasn't been working properly for a while.
3310 * The return type of BIO_number_read() and BIO_number_written() as well as
3311 the corresponding num_read and num_write members in the BIO structure has
3312 changed from unsigned long to uint64_t. On platforms where an unsigned
3313 long is 32 bits (e.g. Windows) these counters could overflow if >4Gb is
3318 * Given the pervasive nature of TLS extensions it is inadvisable to run
3319 OpenSSL without support for them. It also means that maintaining
3320 the OPENSSL_NO_TLSEXT option within the code is very invasive (and probably
3321 not well tested). Therefore the OPENSSL_NO_TLSEXT option has been removed.
3325 * Removed support for the two export grade static DH ciphersuites
3326 EXP-DH-RSA-DES-CBC-SHA and EXP-DH-DSS-DES-CBC-SHA. These two ciphersuites
3327 were newly added (along with a number of other static DH ciphersuites) to
3328 1.0.2. However the two export ones have *never* worked since they were
3329 introduced. It seems strange in any case to be adding new export
3330 ciphersuites, and given "logjam" it also does not seem correct to fix them.
3334 * Version negotiation has been rewritten. In particular SSLv23_method(),
3335 SSLv23_client_method() and SSLv23_server_method() have been deprecated,
3336 and turned into macros which simply call the new preferred function names
3337 TLS_method(), TLS_client_method() and TLS_server_method(). All new code
3338 should use the new names instead. Also as part of this change the ssl23.h
3339 header file has been removed.
3343 * Support for Kerberos ciphersuites in TLS (RFC2712) has been removed. This
3344 code and the associated standard is no longer considered fit-for-purpose.
3348 * RT2547 was closed. When generating a private key, try to make the
3349 output file readable only by the owner. This behavior change might
3350 be noticeable when interacting with other software.
3352 * Documented all exdata functions. Added CRYPTO_free_ex_index.
3357 * Added HTTP GET support to the ocsp command.
3361 * Changed default digest for the dgst and enc commands from MD5 to
3366 * RAND_pseudo_bytes has been deprecated. Users should use RAND_bytes instead.
3370 * Added support for TLS extended master secret from
3371 draft-ietf-tls-session-hash-03.txt. Thanks for Alfredo Pironti for an
3372 initial patch which was a great help during development.
3376 * All libssl internal structures have been removed from the public header
3377 files, and the OPENSSL_NO_SSL_INTERN option has been removed (since it is
3378 now redundant). Users should not attempt to access internal structures
3379 directly. Instead they should use the provided API functions.
3383 * config has been changed so that by default OPENSSL_NO_DEPRECATED is used.
3384 Access to deprecated functions can be re-enabled by running config with
3385 "enable-deprecated". In addition applications wishing to use deprecated
3386 functions must define OPENSSL_USE_DEPRECATED. Note that this new behaviour
3387 will, by default, disable some transitive includes that previously existed
3388 in the header files (e.g. ec.h will no longer, by default, include bn.h)
3392 * Added support for OCB mode. OpenSSL has been granted a patent license
3393 compatible with the OpenSSL license for use of OCB. Details are available
3394 at <https://www.openssl.org/source/OCB-patent-grant-OpenSSL.pdf>. Support
3395 for OCB can be removed by calling config with no-ocb.
3399 * SSLv2 support has been removed. It still supports receiving a SSLv2
3400 compatible client hello.
3404 * Increased the minimal RSA keysize from 256 to 512 bits [Rich Salz],
3405 done while fixing the error code for the key-too-small case.
3407 *Annie Yousar <a.yousar@informatik.hu-berlin.de>*
3409 * CA.sh has been removed; use CA.pl instead.
3413 * Removed old DES API.
3417 * Remove various unsupported platforms:
3423 Sinix/ReliantUNIX RM400
3428 16-bit platforms such as WIN16
3432 * Clean up OPENSSL_NO_xxx #define's
3433 - Use setbuf() and remove OPENSSL_NO_SETVBUF_IONBF
3434 - Rename OPENSSL_SYSNAME_xxx to OPENSSL_SYS_xxx
3435 - OPENSSL_NO_EC{DH,DSA} merged into OPENSSL_NO_EC
3436 - OPENSSL_NO_RIPEMD160, OPENSSL_NO_RIPEMD merged into OPENSSL_NO_RMD160
3437 - OPENSSL_NO_FP_API merged into OPENSSL_NO_STDIO
3438 - Remove OPENSSL_NO_BIO OPENSSL_NO_BUFFER OPENSSL_NO_CHAIN_VERIFY
3439 OPENSSL_NO_EVP OPENSSL_NO_FIPS_ERR OPENSSL_NO_HASH_COMP
3440 OPENSSL_NO_LHASH OPENSSL_NO_OBJECT OPENSSL_NO_SPEED OPENSSL_NO_STACK
3441 OPENSSL_NO_X509 OPENSSL_NO_X509_VERIFY
3442 - Remove MS_STATIC; it's a relic from platforms <32 bits.
3446 * Cleaned up dead code
3447 Remove all but one '#ifdef undef' which is to be looked at.
3451 * Clean up calling of xxx_free routines.
3452 Just like free(), fix most of the xxx_free routines to accept
3453 NULL. Remove the non-null checks from callers. Save much code.
3457 * Add secure heap for storage of private keys (when possible).
3458 Add BIO_s_secmem(), CBIGNUM, etc.
3459 Contributed by Akamai Technologies under our Corporate CLA.
3463 * Experimental support for a new, fast, unbiased prime candidate generator,
3464 bn_probable_prime_dh_coprime(). Not currently used by any prime generator.
3466 *Felix Laurie von Massenbach <felix@erbridge.co.uk>*
3468 * New output format NSS in the sess_id command line tool. This allows
3469 exporting the session id and the master key in NSS keylog format.
3471 *Martin Kaiser <martin@kaiser.cx>*
3473 * Harmonize version and its documentation. -f flag is used to display
3476 *mancha <mancha1@zoho.com>*
3478 * Fix eckey_priv_encode so it immediately returns an error upon a failure
3479 in i2d_ECPrivateKey. Thanks to Ted Unangst for feedback on this issue.
3481 *mancha <mancha1@zoho.com>*
3483 * Fix some double frees. These are not thought to be exploitable.
3485 *mancha <mancha1@zoho.com>*
3487 * A missing bounds check in the handling of the TLS heartbeat extension
3488 can be used to reveal up to 64k of memory to a connected client or
3491 Thanks for Neel Mehta of Google Security for discovering this bug and to
3492 Adam Langley <agl@chromium.org> and Bodo Moeller <bmoeller@acm.org> for
3493 preparing the fix [CVE-2014-0160][]
3495 *Adam Langley, Bodo Moeller*
3497 * Fix for the attack described in the paper "Recovering OpenSSL
3498 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
3499 by Yuval Yarom and Naomi Benger. Details can be obtained from:
3500 <http://eprint.iacr.org/2014/140>
3502 Thanks to Yuval Yarom and Naomi Benger for discovering this
3503 flaw and to Yuval Yarom for supplying a fix [CVE-2014-0076][]
3505 *Yuval Yarom and Naomi Benger*
3507 * Use algorithm specific chains in SSL_CTX_use_certificate_chain_file():
3508 this fixes a limitation in previous versions of OpenSSL.
3512 * Experimental encrypt-then-mac support.
3514 Experimental support for encrypt then mac from
3515 draft-gutmann-tls-encrypt-then-mac-02.txt
3517 To enable it set the appropriate extension number (0x42 for the test
3518 server) using e.g. -DTLSEXT_TYPE_encrypt_then_mac=0x42
3520 For non-compliant peers (i.e. just about everything) this should have no
3523 WARNING: EXPERIMENTAL, SUBJECT TO CHANGE.
3527 * Add EVP support for key wrapping algorithms, to avoid problems with
3528 existing code the flag EVP_CIPHER_CTX_WRAP_ALLOW has to be set in
3529 the EVP_CIPHER_CTX or an error is returned. Add AES and DES3 wrap
3530 algorithms and include tests cases.
3534 * Extend CMS code to support RSA-PSS signatures and RSA-OAEP for
3539 * Extended RSA OAEP support via EVP_PKEY API. Options to specify digest,
3540 MGF1 digest and OAEP label.
3544 * Make openssl verify return errors.
3546 *Chris Palmer <palmer@google.com> and Ben Laurie*
3548 * New function ASN1_TIME_diff to calculate the difference between two
3549 ASN1_TIME structures or one structure and the current time.
3553 * Update fips_test_suite to support multiple command line options. New
3554 test to induce all self test errors in sequence and check expected
3559 * Add FIPS_{rsa,dsa,ecdsa}_{sign,verify} functions which digest and
3560 sign or verify all in one operation.
3564 * Add fips_algvs: a multicall fips utility incorporating all the algorithm
3565 test programs and fips_test_suite. Includes functionality to parse
3566 the minimal script output of fipsalgest.pl directly.
3570 * Add authorisation parameter to FIPS_module_mode_set().
3574 * Add FIPS selftest for ECDH algorithm using P-224 and B-233 curves.
3578 * Use separate DRBG fields for internal and external flags. New function
3579 FIPS_drbg_health_check() to perform on demand health checking. Add
3580 generation tests to fips_test_suite with reduced health check interval to
3581 demonstrate periodic health checking. Add "nodh" option to
3582 fips_test_suite to skip very slow DH test.
3586 * New function FIPS_get_cipherbynid() to lookup FIPS supported ciphers
3591 * More extensive health check for DRBG checking many more failure modes.
3592 New function FIPS_selftest_drbg_all() to handle every possible DRBG
3593 combination: call this in fips_test_suite.
3597 * Add support for canonical generation of DSA parameter 'g'. See
3600 * Add support for HMAC DRBG from SP800-90. Update DRBG algorithm test and
3601 POST to handle HMAC cases.
3605 * Add functions FIPS_module_version() and FIPS_module_version_text()
3606 to return numerical and string versions of the FIPS module number.
3610 * Rename FIPS_mode_set and FIPS_mode to FIPS_module_mode_set and
3611 FIPS_module_mode. FIPS_mode and FIPS_mode_set will be implemented
3612 outside the validated module in the FIPS capable OpenSSL.
3616 * Minor change to DRBG entropy callback semantics. In some cases
3617 there is no multiple of the block length between min_len and
3618 max_len. Allow the callback to return more than max_len bytes
3619 of entropy but discard any extra: it is the callback's responsibility
3620 to ensure that the extra data discarded does not impact the
3621 requested amount of entropy.
3625 * Add PRNG security strength checks to RSA, DSA and ECDSA using
3626 information in FIPS186-3, SP800-57 and SP800-131A.
3630 * CCM support via EVP. Interface is very similar to GCM case except we
3631 must supply all data in one chunk (i.e. no update, final) and the
3632 message length must be supplied if AAD is used. Add algorithm test
3637 * Initial version of POST overhaul. Add POST callback to allow the status
3638 of POST to be monitored and/or failures induced. Modify fips_test_suite
3639 to use callback. Always run all selftests even if one fails.
3643 * XTS support including algorithm test driver in the fips_gcmtest program.
3644 Note: this does increase the maximum key length from 32 to 64 bytes but
3645 there should be no binary compatibility issues as existing applications
3646 will never use XTS mode.
3650 * Extensive reorganisation of FIPS PRNG behaviour. Remove all dependencies
3651 to OpenSSL RAND code and replace with a tiny FIPS RAND API which also
3652 performs algorithm blocking for unapproved PRNG types. Also do not
3653 set PRNG type in FIPS_mode_set(): leave this to the application.
3654 Add default OpenSSL DRBG handling: sets up FIPS PRNG and seeds with
3655 the standard OpenSSL PRNG: set additional data to a date time vector.
3659 * Rename old X9.31 PRNG functions of the form `FIPS_rand*` to `FIPS_x931*`.
3660 This shouldn't present any incompatibility problems because applications
3661 shouldn't be using these directly and any that are will need to rethink
3662 anyway as the X9.31 PRNG is now deprecated by FIPS 140-2
3666 * Extensive self tests and health checking required by SP800-90 DRBG.
3667 Remove strength parameter from FIPS_drbg_instantiate and always
3668 instantiate at maximum supported strength.
3672 * Add ECDH code to fips module and fips_ecdhvs for primitives only testing.
3676 * New algorithm test program fips_dhvs to handle DH primitives only testing.
3680 * New function DH_compute_key_padded() to compute a DH key and pad with
3681 leading zeroes if needed: this complies with SP800-56A et al.
3685 * Initial implementation of SP800-90 DRBGs for Hash and CTR. Not used by
3686 anything, incomplete, subject to change and largely untested at present.
3690 * Modify fipscanisteronly build option to only build the necessary object
3691 files by filtering FIPS_EX_OBJ through a perl script in crypto/Makefile.
3695 * Add experimental option FIPSSYMS to give all symbols in
3696 fipscanister.o and FIPS or fips prefix. This will avoid
3697 conflicts with future versions of OpenSSL. Add perl script
3698 util/fipsas.pl to preprocess assembly language source files
3699 and rename any affected symbols.
3703 * Add selftest checks and algorithm block of non-fips algorithms in
3704 FIPS mode. Remove DES2 from selftests.
3708 * Add ECDSA code to fips module. Add tiny fips_ecdsa_check to just
3709 return internal method without any ENGINE dependencies. Add new
3710 tiny fips sign and verify functions.
3714 * New build option no-ec2m to disable characteristic 2 code.
3718 * New build option "fipscanisteronly". This only builds fipscanister.o
3719 and (currently) associated fips utilities. Uses the file Makefile.fips
3720 instead of Makefile.org as the prototype.
3724 * Add some FIPS mode restrictions to GCM. Add internal IV generator.
3725 Update fips_gcmtest to use IV generator.
3729 * Initial, experimental EVP support for AES-GCM. AAD can be input by
3730 setting output buffer to NULL. The `*Final` function must be
3731 called although it will not retrieve any additional data. The tag
3732 can be set or retrieved with a ctrl. The IV length is by default 12
3733 bytes (96 bits) but can be set to an alternative value. If the IV
3734 length exceeds the maximum IV length (currently 16 bytes) it cannot be
3739 * New flag in ciphers: EVP_CIPH_FLAG_CUSTOM_CIPHER. This means the
3740 underlying do_cipher function handles all cipher semantics itself
3741 including padding and finalisation. This is useful if (for example)
3742 an ENGINE cipher handles block padding itself. The behaviour of
3743 do_cipher is subtly changed if this flag is set: the return value
3744 is the number of characters written to the output buffer (zero is
3745 no longer an error code) or a negative error code. Also if the
3746 input buffer is NULL and length 0 finalisation should be performed.
3750 * If a candidate issuer certificate is already part of the constructed
3751 path ignore it: new debug notification X509_V_ERR_PATH_LOOP for this case.
3755 * Improve forward-security support: add functions
3757 void SSL_CTX_set_not_resumable_session_callback(
3758 SSL_CTX *ctx, int (*cb)(SSL *ssl, int is_forward_secure))
3759 void SSL_set_not_resumable_session_callback(
3760 SSL *ssl, int (*cb)(SSL *ssl, int is_forward_secure))
3762 for use by SSL/TLS servers; the callback function will be called whenever a
3763 new session is created, and gets to decide whether the session may be
3764 cached to make it resumable (return 0) or not (return 1). (As by the
3765 SSL/TLS protocol specifications, the session_id sent by the server will be
3766 empty to indicate that the session is not resumable; also, the server will
3767 not generate RFC 4507 (RFC 5077) session tickets.)
3769 A simple reasonable callback implementation is to return is_forward_secure.
3770 This parameter will be set to 1 or 0 depending on the ciphersuite selected
3771 by the SSL/TLS server library, indicating whether it can provide forward
3774 *Emilia Käsper <emilia.kasper@esat.kuleuven.be> (Google)*
3776 * New -verify_name option in command line utilities to set verification
3781 * Initial CMAC implementation. WARNING: EXPERIMENTAL, API MAY CHANGE.
3782 Add CMAC pkey methods.
3786 * Experimental renegotiation in s_server -www mode. If the client
3787 browses /reneg connection is renegotiated. If /renegcert it is
3788 renegotiated requesting a certificate.
3792 * Add an "external" session cache for debugging purposes to s_server. This
3793 should help trace issues which normally are only apparent in deployed
3794 multi-process servers.
3798 * Extensive audit of libcrypto with DEBUG_UNUSED. Fix many cases where
3799 return value is ignored. NB. The functions RAND_add(), RAND_seed(),
3800 BIO_set_cipher() and some obscure PEM functions were changed so they
3801 can now return an error. The RAND changes required a change to the
3802 RAND_METHOD structure.
3806 * New macro `__owur` for "OpenSSL Warn Unused Result". This makes use of
3807 a gcc attribute to warn if the result of a function is ignored. This
3808 is enable if DEBUG_UNUSED is set. Add to several functions in evp.h
3809 whose return value is often ignored.
3813 * New -noct, -requestct, -requirect and -ctlogfile options for s_client.
3814 These allow SCTs (signed certificate timestamps) to be requested and
3815 validated when establishing a connection.
3817 *Rob Percival <robpercival@google.com>*
3822 ### Changes between 1.0.2s and 1.0.2t [10 Sep 2019]
3824 * For built-in EC curves, ensure an EC_GROUP built from the curve name is
3825 used even when parsing explicit parameters, when loading a encoded key
3826 or calling `EC_GROUP_new_from_ecpkparameters()`/
3827 `EC_GROUP_new_from_ecparameters()`.
3828 This prevents bypass of security hardening and performance gains,
3829 especially for curves with specialized EC_METHODs.
3830 By default, if a key encoded with explicit parameters is loaded and later
3831 encoded, the output is still encoded with explicit parameters, even if
3832 internally a "named" EC_GROUP is used for computation.
3836 * Compute ECC cofactors if not provided during EC_GROUP construction. Before
3837 this change, EC_GROUP_set_generator would accept order and/or cofactor as
3838 NULL. After this change, only the cofactor parameter can be NULL. It also
3839 does some minimal sanity checks on the passed order.
3844 * Fixed a padding oracle in PKCS7_dataDecode and CMS_decrypt_set1_pkey.
3845 An attack is simple, if the first CMS_recipientInfo is valid but the
3846 second CMS_recipientInfo is chosen ciphertext. If the second
3847 recipientInfo decodes to PKCS #1 v1.5 form plaintext, the correct
3848 encryption key will be replaced by garbage, and the message cannot be
3849 decoded, but if the RSA decryption fails, the correct encryption key is
3850 used and the recipient will not notice the attack.
3851 As a work around for this potential attack the length of the decrypted
3852 key must be equal to the cipher default key length, in case the
3853 certifiate is not given and all recipientInfo are tried out.
3854 The old behaviour can be re-enabled in the CMS code by setting the
3855 CMS_DEBUG_DECRYPT flag.
3860 * Document issue with installation paths in diverse Windows builds
3862 '/usr/local/ssl' is an unsafe prefix for location to install OpenSSL
3863 binaries and run-time config file.
3868 ### Changes between 1.0.2r and 1.0.2s [28 May 2019]
3870 * Change the default RSA, DSA and DH size to 2048 bit instead of 1024.
3871 This changes the size when using the genpkey app when no size is given. It
3872 fixes an omission in earlier changes that changed all RSA, DSA and DH
3873 generation apps to use 2048 bits by default.
3877 * Add FIPS support for Android Arm 64-bit
3879 Support for Android Arm 64-bit was added to the OpenSSL FIPS Object
3880 Module in Version 2.0.10. For some reason, the corresponding target
3881 'android64-aarch64' was missing OpenSSL 1.0.2, whence it could not be
3882 built with FIPS support on Android Arm 64-bit. This omission has been
3885 *Matthias St. Pierre*
3887 ### Changes between 1.0.2q and 1.0.2r [26 Feb 2019]
3889 * 0-byte record padding oracle
3891 If an application encounters a fatal protocol error and then calls
3892 SSL_shutdown() twice (once to send a close_notify, and once to receive one)
3893 then OpenSSL can respond differently to the calling application if a 0 byte
3894 record is received with invalid padding compared to if a 0 byte record is
3895 received with an invalid MAC. If the application then behaves differently
3896 based on that in a way that is detectable to the remote peer, then this
3897 amounts to a padding oracle that could be used to decrypt data.
3899 In order for this to be exploitable "non-stitched" ciphersuites must be in
3900 use. Stitched ciphersuites are optimised implementations of certain
3901 commonly used ciphersuites. Also the application must call SSL_shutdown()
3902 twice even if a protocol error has occurred (applications should not do
3903 this but some do anyway).
3905 This issue was discovered by Juraj Somorovsky, Robert Merget and Nimrod
3906 Aviram, with additional investigation by Steven Collison and Andrew
3907 Hourselt. It was reported to OpenSSL on 10th December 2018.
3912 * Move strictness check from EVP_PKEY_asn1_new() to EVP_PKEY_asn1_add0().
3916 ### Changes between 1.0.2p and 1.0.2q [20 Nov 2018]
3918 * Microarchitecture timing vulnerability in ECC scalar multiplication
3920 OpenSSL ECC scalar multiplication, used in e.g. ECDSA and ECDH, has been
3921 shown to be vulnerable to a microarchitecture timing side channel attack.
3922 An attacker with sufficient access to mount local timing attacks during
3923 ECDSA signature generation could recover the private key.
3925 This issue was reported to OpenSSL on 26th October 2018 by Alejandro
3926 Cabrera Aldaya, Billy Brumley, Sohaib ul Hassan, Cesar Pereida Garcia and
3932 * Timing vulnerability in DSA signature generation
3934 The OpenSSL DSA signature algorithm has been shown to be vulnerable to a
3935 timing side channel attack. An attacker could use variations in the signing
3936 algorithm to recover the private key.
3938 This issue was reported to OpenSSL on 16th October 2018 by Samuel Weiser.
3943 * Resolve a compatibility issue in EC_GROUP handling with the FIPS Object
3944 Module, accidentally introduced while backporting security fixes from the
3945 development branch and hindering the use of ECC in FIPS mode.
3949 ### Changes between 1.0.2o and 1.0.2p [14 Aug 2018]
3951 * Client DoS due to large DH parameter
3953 During key agreement in a TLS handshake using a DH(E) based ciphersuite a
3954 malicious server can send a very large prime value to the client. This will
3955 cause the client to spend an unreasonably long period of time generating a
3956 key for this prime resulting in a hang until the client has finished. This
3957 could be exploited in a Denial Of Service attack.
3959 This issue was reported to OpenSSL on 5th June 2018 by Guido Vranken
3964 * Cache timing vulnerability in RSA Key Generation
3966 The OpenSSL RSA Key generation algorithm has been shown to be vulnerable to
3967 a cache timing side channel attack. An attacker with sufficient access to
3968 mount cache timing attacks during the RSA key generation process could
3969 recover the private key.
3971 This issue was reported to OpenSSL on 4th April 2018 by Alejandro Cabrera
3972 Aldaya, Billy Brumley, Cesar Pereida Garcia and Luis Manuel Alvarez Tapia.
3977 * Make EVP_PKEY_asn1_new() a bit stricter about its input. A NULL pem_str
3978 parameter is no longer accepted, as it leads to a corrupt table. NULL
3979 pem_str is reserved for alias entries only.
3983 * Revert blinding in ECDSA sign and instead make problematic addition
3984 length-invariant. Switch even to fixed-length Montgomery multiplication.
3988 * Change generating and checking of primes so that the error rate of not
3989 being prime depends on the intended use based on the size of the input.
3990 For larger primes this will result in more rounds of Miller-Rabin.
3991 The maximal error rate for primes with more than 1080 bits is lowered
3994 *Kurt Roeckx, Annie Yousar*
3996 * Increase the number of Miller-Rabin rounds for DSA key generating to 64.
4000 * Add blinding to ECDSA and DSA signatures to protect against side channel
4001 attacks discovered by Keegan Ryan (NCC Group).
4005 * When unlocking a pass phrase protected PEM file or PKCS#8 container, we
4006 now allow empty (zero character) pass phrases.
4010 * Certificate time validation (X509_cmp_time) enforces stricter
4011 compliance with RFC 5280. Fractional seconds and timezone offsets
4012 are no longer allowed.
4016 ### Changes between 1.0.2n and 1.0.2o [27 Mar 2018]
4018 * Constructed ASN.1 types with a recursive definition could exceed the stack
4020 Constructed ASN.1 types with a recursive definition (such as can be found
4021 in PKCS7) could eventually exceed the stack given malicious input with
4022 excessive recursion. This could result in a Denial Of Service attack. There
4023 are no such structures used within SSL/TLS that come from untrusted sources
4024 so this is considered safe.
4026 This issue was reported to OpenSSL on 4th January 2018 by the OSS-fuzz
4032 ### Changes between 1.0.2m and 1.0.2n [7 Dec 2017]
4034 * Read/write after SSL object in error state
4036 OpenSSL 1.0.2 (starting from version 1.0.2b) introduced an "error state"
4037 mechanism. The intent was that if a fatal error occurred during a handshake
4038 then OpenSSL would move into the error state and would immediately fail if
4039 you attempted to continue the handshake. This works as designed for the
4040 explicit handshake functions (SSL_do_handshake(), SSL_accept() and
4041 SSL_connect()), however due to a bug it does not work correctly if
4042 SSL_read() or SSL_write() is called directly. In that scenario, if the
4043 handshake fails then a fatal error will be returned in the initial function
4044 call. If SSL_read()/SSL_write() is subsequently called by the application
4045 for the same SSL object then it will succeed and the data is passed without
4046 being decrypted/encrypted directly from the SSL/TLS record layer.
4048 In order to exploit this issue an application bug would have to be present
4049 that resulted in a call to SSL_read()/SSL_write() being issued after having
4050 already received a fatal error.
4052 This issue was reported to OpenSSL by David Benjamin (Google).
4057 * rsaz_1024_mul_avx2 overflow bug on x86_64
4059 There is an overflow bug in the AVX2 Montgomery multiplication procedure
4060 used in exponentiation with 1024-bit moduli. No EC algorithms are affected.
4061 Analysis suggests that attacks against RSA and DSA as a result of this
4062 defect would be very difficult to perform and are not believed likely.
4063 Attacks against DH1024 are considered just feasible, because most of the
4064 work necessary to deduce information about a private key may be performed
4065 offline. The amount of resources required for such an attack would be
4066 significant. However, for an attack on TLS to be meaningful, the server
4067 would have to share the DH1024 private key among multiple clients, which is
4068 no longer an option since CVE-2016-0701.
4070 This only affects processors that support the AVX2 but not ADX extensions
4071 like Intel Haswell (4th generation).
4073 This issue was reported to OpenSSL by David Benjamin (Google). The issue
4074 was originally found via the OSS-Fuzz project.
4079 ### Changes between 1.0.2l and 1.0.2m [2 Nov 2017]
4081 * bn_sqrx8x_internal carry bug on x86_64
4083 There is a carry propagating bug in the x86_64 Montgomery squaring
4084 procedure. No EC algorithms are affected. Analysis suggests that attacks
4085 against RSA and DSA as a result of this defect would be very difficult to
4086 perform and are not believed likely. Attacks against DH are considered just
4087 feasible (although very difficult) because most of the work necessary to
4088 deduce information about a private key may be performed offline. The amount
4089 of resources required for such an attack would be very significant and
4090 likely only accessible to a limited number of attackers. An attacker would
4091 additionally need online access to an unpatched system using the target
4092 private key in a scenario with persistent DH parameters and a private
4093 key that is shared between multiple clients.
4095 This only affects processors that support the BMI1, BMI2 and ADX extensions
4096 like Intel Broadwell (5th generation) and later or AMD Ryzen.
4098 This issue was reported to OpenSSL by the OSS-Fuzz project.
4103 * Malformed X.509 IPAddressFamily could cause OOB read
4105 If an X.509 certificate has a malformed IPAddressFamily extension,
4106 OpenSSL could do a one-byte buffer overread. The most likely result
4107 would be an erroneous display of the certificate in text format.
4109 This issue was reported to OpenSSL by the OSS-Fuzz project.
4114 ### Changes between 1.0.2k and 1.0.2l [25 May 2017]
4116 * Have 'config' recognise 64-bit mingw and choose 'mingw64' as the target
4117 platform rather than 'mingw'.
4121 ### Changes between 1.0.2j and 1.0.2k [26 Jan 2017]
4123 * Truncated packet could crash via OOB read
4125 If one side of an SSL/TLS path is running on a 32-bit host and a specific
4126 cipher is being used, then a truncated packet can cause that host to
4127 perform an out-of-bounds read, usually resulting in a crash.
4129 This issue was reported to OpenSSL by Robert Święcki of Google.
4134 * BN_mod_exp may produce incorrect results on x86_64
4136 There is a carry propagating bug in the x86_64 Montgomery squaring
4137 procedure. No EC algorithms are affected. Analysis suggests that attacks
4138 against RSA and DSA as a result of this defect would be very difficult to
4139 perform and are not believed likely. Attacks against DH are considered just
4140 feasible (although very difficult) because most of the work necessary to
4141 deduce information about a private key may be performed offline. The amount
4142 of resources required for such an attack would be very significant and
4143 likely only accessible to a limited number of attackers. An attacker would
4144 additionally need online access to an unpatched system using the target
4145 private key in a scenario with persistent DH parameters and a private
4146 key that is shared between multiple clients. For example this can occur by
4147 default in OpenSSL DHE based SSL/TLS ciphersuites. Note: This issue is very
4148 similar to CVE-2015-3193 but must be treated as a separate problem.
4150 This issue was reported to OpenSSL by the OSS-Fuzz project.
4155 * Montgomery multiplication may produce incorrect results
4157 There is a carry propagating bug in the Broadwell-specific Montgomery
4158 multiplication procedure that handles input lengths divisible by, but
4159 longer than 256 bits. Analysis suggests that attacks against RSA, DSA
4160 and DH private keys are impossible. This is because the subroutine in
4161 question is not used in operations with the private key itself and an input
4162 of the attacker's direct choice. Otherwise the bug can manifest itself as
4163 transient authentication and key negotiation failures or reproducible
4164 erroneous outcome of public-key operations with specially crafted input.
4165 Among EC algorithms only Brainpool P-512 curves are affected and one
4166 presumably can attack ECDH key negotiation. Impact was not analyzed in
4167 detail, because pre-requisites for attack are considered unlikely. Namely
4168 multiple clients have to choose the curve in question and the server has to
4169 share the private key among them, neither of which is default behaviour.
4170 Even then only clients that chose the curve will be affected.
4172 This issue was publicly reported as transient failures and was not
4173 initially recognized as a security issue. Thanks to Richard Morgan for
4174 providing reproducible case.
4179 * OpenSSL now fails if it receives an unrecognised record type in TLS1.0
4180 or TLS1.1. Previously this only happened in SSLv3 and TLS1.2. This is to
4181 prevent issues where no progress is being made and the peer continually
4182 sends unrecognised record types, using up resources processing them.
4186 ### Changes between 1.0.2i and 1.0.2j [26 Sep 2016]
4188 * Missing CRL sanity check
4190 A bug fix which included a CRL sanity check was added to OpenSSL 1.1.0
4191 but was omitted from OpenSSL 1.0.2i. As a result any attempt to use
4192 CRLs in OpenSSL 1.0.2i will crash with a null pointer exception.
4194 This issue only affects the OpenSSL 1.0.2i
4199 ### Changes between 1.0.2h and 1.0.2i [22 Sep 2016]
4201 * OCSP Status Request extension unbounded memory growth
4203 A malicious client can send an excessively large OCSP Status Request
4204 extension. If that client continually requests renegotiation, sending a
4205 large OCSP Status Request extension each time, then there will be unbounded
4206 memory growth on the server. This will eventually lead to a Denial Of
4207 Service attack through memory exhaustion. Servers with a default
4208 configuration are vulnerable even if they do not support OCSP. Builds using
4209 the "no-ocsp" build time option are not affected.
4211 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4216 * In order to mitigate the SWEET32 attack, the DES ciphers were moved from
4219 This issue was reported to OpenSSL Karthikeyan Bhargavan and Gaetan
4225 * OOB write in MDC2_Update()
4227 An overflow can occur in MDC2_Update() either if called directly or
4228 through the EVP_DigestUpdate() function using MDC2. If an attacker
4229 is able to supply very large amounts of input data after a previous
4230 call to EVP_EncryptUpdate() with a partial block then a length check
4231 can overflow resulting in a heap corruption.
4233 The amount of data needed is comparable to SIZE_MAX which is impractical
4236 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4241 * Malformed SHA512 ticket DoS
4243 If a server uses SHA512 for TLS session ticket HMAC it is vulnerable to a
4244 DoS attack where a malformed ticket will result in an OOB read which will
4247 The use of SHA512 in TLS session tickets is comparatively rare as it requires
4248 a custom server callback and ticket lookup mechanism.
4250 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4255 * OOB write in BN_bn2dec()
4257 The function BN_bn2dec() does not check the return value of BN_div_word().
4258 This can cause an OOB write if an application uses this function with an
4259 overly large BIGNUM. This could be a problem if an overly large certificate
4260 or CRL is printed out from an untrusted source. TLS is not affected because
4261 record limits will reject an oversized certificate before it is parsed.
4263 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4268 * OOB read in TS_OBJ_print_bio()
4270 The function TS_OBJ_print_bio() misuses OBJ_obj2txt(): the return value is
4271 the total length the OID text representation would use and not the amount
4272 of data written. This will result in OOB reads when large OIDs are
4275 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4280 * Pointer arithmetic undefined behaviour
4282 Avoid some undefined pointer arithmetic
4284 A common idiom in the codebase is to check limits in the following manner:
4287 Where "p" points to some malloc'd data of SIZE bytes and
4290 "len" here could be from some externally supplied data (e.g. from a TLS
4293 The rules of C pointer arithmetic are such that "p + len" is only well
4294 defined where len <= SIZE. Therefore the above idiom is actually
4295 undefined behaviour.
4297 For example this could cause problems if some malloc implementation
4298 provides an address for "p" such that "p + len" actually overflows for
4299 values of len that are too big and therefore p + len < limit.
4301 This issue was reported to OpenSSL by Guido Vranken
4306 * Constant time flag not preserved in DSA signing
4308 Operations in the DSA signing algorithm should run in constant time in
4309 order to avoid side channel attacks. A flaw in the OpenSSL DSA
4310 implementation means that a non-constant time codepath is followed for
4311 certain operations. This has been demonstrated through a cache-timing
4312 attack to be sufficient for an attacker to recover the private DSA key.
4314 This issue was reported by César Pereida (Aalto University), Billy Brumley
4315 (Tampere University of Technology), and Yuval Yarom (The University of
4316 Adelaide and NICTA).
4321 * DTLS buffered message DoS
4323 In a DTLS connection where handshake messages are delivered out-of-order
4324 those messages that OpenSSL is not yet ready to process will be buffered
4325 for later use. Under certain circumstances, a flaw in the logic means that
4326 those messages do not get removed from the buffer even though the handshake
4327 has been completed. An attacker could force up to approx. 15 messages to
4328 remain in the buffer when they are no longer required. These messages will
4329 be cleared when the DTLS connection is closed. The default maximum size for
4330 a message is 100k. Therefore the attacker could force an additional 1500k
4331 to be consumed per connection. By opening many simulataneous connections an
4332 attacker could cause a DoS attack through memory exhaustion.
4334 This issue was reported to OpenSSL by Quan Luo.
4339 * DTLS replay protection DoS
4341 A flaw in the DTLS replay attack protection mechanism means that records
4342 that arrive for future epochs update the replay protection "window" before
4343 the MAC for the record has been validated. This could be exploited by an
4344 attacker by sending a record for the next epoch (which does not have to
4345 decrypt or have a valid MAC), with a very large sequence number. This means
4346 that all subsequent legitimate packets are dropped causing a denial of
4347 service for a specific DTLS connection.
4349 This issue was reported to OpenSSL by the OCAP audit team.
4354 * Certificate message OOB reads
4356 In OpenSSL 1.0.2 and earlier some missing message length checks can result
4357 in OOB reads of up to 2 bytes beyond an allocated buffer. There is a
4358 theoretical DoS risk but this has not been observed in practice on common
4361 The messages affected are client certificate, client certificate request
4362 and server certificate. As a result the attack can only be performed
4363 against a client or a server which enables client authentication.
4365 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4370 ### Changes between 1.0.2g and 1.0.2h [3 May 2016]
4372 * Prevent padding oracle in AES-NI CBC MAC check
4374 A MITM attacker can use a padding oracle attack to decrypt traffic
4375 when the connection uses an AES CBC cipher and the server support
4378 This issue was introduced as part of the fix for Lucky 13 padding
4379 attack [CVE-2013-0169][]. The padding check was rewritten to be in
4380 constant time by making sure that always the same bytes are read and
4381 compared against either the MAC or padding bytes. But it no longer
4382 checked that there was enough data to have both the MAC and padding
4385 This issue was reported by Juraj Somorovsky using TLS-Attacker.
4390 * Fix EVP_EncodeUpdate overflow
4392 An overflow can occur in the EVP_EncodeUpdate() function which is used for
4393 Base64 encoding of binary data. If an attacker is able to supply very large
4394 amounts of input data then a length check can overflow resulting in a heap
4397 Internally to OpenSSL the EVP_EncodeUpdate() function is primarily used by
4398 the `PEM_write_bio*` family of functions. These are mainly used within the
4399 OpenSSL command line applications, so any application which processes data
4400 from an untrusted source and outputs it as a PEM file should be considered
4401 vulnerable to this issue. User applications that call these APIs directly
4402 with large amounts of untrusted data may also be vulnerable.
4404 This issue was reported by Guido Vranken.
4409 * Fix EVP_EncryptUpdate overflow
4411 An overflow can occur in the EVP_EncryptUpdate() function. If an attacker
4412 is able to supply very large amounts of input data after a previous call to
4413 EVP_EncryptUpdate() with a partial block then a length check can overflow
4414 resulting in a heap corruption. Following an analysis of all OpenSSL
4415 internal usage of the EVP_EncryptUpdate() function all usage is one of two
4416 forms. The first form is where the EVP_EncryptUpdate() call is known to be
4417 the first called function after an EVP_EncryptInit(), and therefore that
4418 specific call must be safe. The second form is where the length passed to
4419 EVP_EncryptUpdate() can be seen from the code to be some small value and
4420 therefore there is no possibility of an overflow. Since all instances are
4421 one of these two forms, it is believed that there can be no overflows in
4422 internal code due to this problem. It should be noted that
4423 EVP_DecryptUpdate() can call EVP_EncryptUpdate() in certain code paths.
4424 Also EVP_CipherUpdate() is a synonym for EVP_EncryptUpdate(). All instances
4425 of these calls have also been analysed too and it is believed there are no
4426 instances in internal usage where an overflow could occur.
4428 This issue was reported by Guido Vranken.
4433 * Prevent ASN.1 BIO excessive memory allocation
4435 When ASN.1 data is read from a BIO using functions such as d2i_CMS_bio()
4436 a short invalid encoding can cause allocation of large amounts of memory
4437 potentially consuming excessive resources or exhausting memory.
4439 Any application parsing untrusted data through d2i BIO functions is
4440 affected. The memory based functions such as d2i_X509() are *not* affected.
4441 Since the memory based functions are used by the TLS library, TLS
4442 applications are not affected.
4444 This issue was reported by Brian Carpenter.
4451 ASN1 Strings that are over 1024 bytes can cause an overread in applications
4452 using the X509_NAME_oneline() function on EBCDIC systems. This could result
4453 in arbitrary stack data being returned in the buffer.
4455 This issue was reported by Guido Vranken.
4460 * Modify behavior of ALPN to invoke callback after SNI/servername
4461 callback, such that updates to the SSL_CTX affect ALPN.
4465 * Remove LOW from the DEFAULT cipher list. This removes singles DES from the
4470 * Only remove the SSLv2 methods with the no-ssl2-method option. When the
4471 methods are enabled and ssl2 is disabled the methods return NULL.
4475 ### Changes between 1.0.2f and 1.0.2g [1 Mar 2016]
4477 * Disable weak ciphers in SSLv3 and up in default builds of OpenSSL.
4478 Builds that are not configured with "enable-weak-ssl-ciphers" will not
4479 provide any "EXPORT" or "LOW" strength ciphers.
4483 * Disable SSLv2 default build, default negotiation and weak ciphers. SSLv2
4484 is by default disabled at build-time. Builds that are not configured with
4485 "enable-ssl2" will not support SSLv2. Even if "enable-ssl2" is used,
4486 users who want to negotiate SSLv2 via the version-flexible SSLv23_method()
4487 will need to explicitly call either of:
4489 SSL_CTX_clear_options(ctx, SSL_OP_NO_SSLv2);
4491 SSL_clear_options(ssl, SSL_OP_NO_SSLv2);
4493 as appropriate. Even if either of those is used, or the application
4494 explicitly uses the version-specific SSLv2_method() or its client and
4495 server variants, SSLv2 ciphers vulnerable to exhaustive search key
4496 recovery have been removed. Specifically, the SSLv2 40-bit EXPORT
4497 ciphers, and SSLv2 56-bit DES are no longer available.
4502 * Fix a double-free in DSA code
4504 A double free bug was discovered when OpenSSL parses malformed DSA private
4505 keys and could lead to a DoS attack or memory corruption for applications
4506 that receive DSA private keys from untrusted sources. This scenario is
4509 This issue was reported to OpenSSL by Adam Langley(Google/BoringSSL) using
4515 * Disable SRP fake user seed to address a server memory leak.
4517 Add a new method SRP_VBASE_get1_by_user that handles the seed properly.
4519 SRP_VBASE_get_by_user had inconsistent memory management behaviour.
4520 In order to fix an unavoidable memory leak, SRP_VBASE_get_by_user
4521 was changed to ignore the "fake user" SRP seed, even if the seed
4524 Users should use SRP_VBASE_get1_by_user instead. Note that in
4525 SRP_VBASE_get1_by_user, caller must free the returned value. Note
4526 also that even though configuring the SRP seed attempts to hide
4527 invalid usernames by continuing the handshake with fake
4528 credentials, this behaviour is not constant time and no strong
4529 guarantees are made that the handshake is indistinguishable from
4530 that of a valid user.
4535 * Fix BN_hex2bn/BN_dec2bn NULL pointer deref/heap corruption
4537 In the BN_hex2bn function the number of hex digits is calculated using an
4538 int value `i`. Later `bn_expand` is called with a value of `i * 4`. For
4539 large values of `i` this can result in `bn_expand` not allocating any
4540 memory because `i * 4` is negative. This can leave the internal BIGNUM data
4541 field as NULL leading to a subsequent NULL ptr deref. For very large values
4542 of `i`, the calculation `i * 4` could be a positive value smaller than `i`.
4543 In this case memory is allocated to the internal BIGNUM data field, but it
4544 is insufficiently sized leading to heap corruption. A similar issue exists
4545 in BN_dec2bn. This could have security consequences if BN_hex2bn/BN_dec2bn
4546 is ever called by user applications with very large untrusted hex/dec data.
4547 This is anticipated to be a rare occurrence.
4549 All OpenSSL internal usage of these functions use data that is not expected
4550 to be untrusted, e.g. config file data or application command line
4551 arguments. If user developed applications generate config file data based
4552 on untrusted data then it is possible that this could also lead to security
4553 consequences. This is also anticipated to be rare.
4555 This issue was reported to OpenSSL by Guido Vranken.
4560 * Fix memory issues in `BIO_*printf` functions
4562 The internal `fmtstr` function used in processing a "%s" format string in
4563 the `BIO_*printf` functions could overflow while calculating the length of a
4564 string and cause an OOB read when printing very long strings.
4566 Additionally the internal `doapr_outch` function can attempt to write to an
4567 OOB memory location (at an offset from the NULL pointer) in the event of a
4568 memory allocation failure. In 1.0.2 and below this could be caused where
4569 the size of a buffer to be allocated is greater than INT_MAX. E.g. this
4570 could be in processing a very long "%s" format string. Memory leaks can
4573 The first issue may mask the second issue dependent on compiler behaviour.
4574 These problems could enable attacks where large amounts of untrusted data
4575 is passed to the `BIO_*printf` functions. If applications use these functions
4576 in this way then they could be vulnerable. OpenSSL itself uses these
4577 functions when printing out human-readable dumps of ASN.1 data. Therefore
4578 applications that print this data could be vulnerable if the data is from
4579 untrusted sources. OpenSSL command line applications could also be
4580 vulnerable where they print out ASN.1 data, or if untrusted data is passed
4581 as command line arguments.
4583 Libssl is not considered directly vulnerable. Additionally certificates etc
4584 received via remote connections via libssl are also unlikely to be able to
4585 trigger these issues because of message size limits enforced within libssl.
4587 This issue was reported to OpenSSL Guido Vranken.
4592 * Side channel attack on modular exponentiation
4594 A side-channel attack was found which makes use of cache-bank conflicts on
4595 the Intel Sandy-Bridge microarchitecture which could lead to the recovery
4596 of RSA keys. The ability to exploit this issue is limited as it relies on
4597 an attacker who has control of code in a thread running on the same
4598 hyper-threaded core as the victim thread which is performing decryptions.
4600 This issue was reported to OpenSSL by Yuval Yarom, The University of
4601 Adelaide and NICTA, Daniel Genkin, Technion and Tel Aviv University, and
4602 Nadia Heninger, University of Pennsylvania with more information at
4603 <http://cachebleed.info>.
4608 * Change the req app to generate a 2048-bit RSA/DSA key by default,
4609 if no keysize is specified with default_bits. This fixes an
4610 omission in an earlier change that changed all RSA/DSA key generation
4611 apps to use 2048 bits by default.
4615 ### Changes between 1.0.2e and 1.0.2f [28 Jan 2016]
4617 * DH small subgroups
4619 Historically OpenSSL only ever generated DH parameters based on "safe"
4620 primes. More recently (in version 1.0.2) support was provided for
4621 generating X9.42 style parameter files such as those required for RFC 5114
4622 support. The primes used in such files may not be "safe". Where an
4623 application is using DH configured with parameters based on primes that are
4624 not "safe" then an attacker could use this fact to find a peer's private
4625 DH exponent. This attack requires that the attacker complete multiple
4626 handshakes in which the peer uses the same private DH exponent. For example
4627 this could be used to discover a TLS server's private DH exponent if it's
4628 reusing the private DH exponent or it's using a static DH ciphersuite.
4630 OpenSSL provides the option SSL_OP_SINGLE_DH_USE for ephemeral DH (DHE) in
4631 TLS. It is not on by default. If the option is not set then the server
4632 reuses the same private DH exponent for the life of the server process and
4633 would be vulnerable to this attack. It is believed that many popular
4634 applications do set this option and would therefore not be at risk.
4636 The fix for this issue adds an additional check where a "q" parameter is
4637 available (as is the case in X9.42 based parameters). This detects the
4638 only known attack, and is the only possible defense for static DH
4639 ciphersuites. This could have some performance impact.
4641 Additionally the SSL_OP_SINGLE_DH_USE option has been switched on by
4642 default and cannot be disabled. This could have some performance impact.
4644 This issue was reported to OpenSSL by Antonio Sanso (Adobe).
4649 * SSLv2 doesn't block disabled ciphers
4651 A malicious client can negotiate SSLv2 ciphers that have been disabled on
4652 the server and complete SSLv2 handshakes even if all SSLv2 ciphers have
4653 been disabled, provided that the SSLv2 protocol was not also disabled via
4656 This issue was reported to OpenSSL on 26th December 2015 by Nimrod Aviram
4657 and Sebastian Schinzel.
4662 ### Changes between 1.0.2d and 1.0.2e [3 Dec 2015]
4664 * BN_mod_exp may produce incorrect results on x86_64
4666 There is a carry propagating bug in the x86_64 Montgomery squaring
4667 procedure. No EC algorithms are affected. Analysis suggests that attacks
4668 against RSA and DSA as a result of this defect would be very difficult to
4669 perform and are not believed likely. Attacks against DH are considered just
4670 feasible (although very difficult) because most of the work necessary to
4671 deduce information about a private key may be performed offline. The amount
4672 of resources required for such an attack would be very significant and
4673 likely only accessible to a limited number of attackers. An attacker would
4674 additionally need online access to an unpatched system using the target
4675 private key in a scenario with persistent DH parameters and a private
4676 key that is shared between multiple clients. For example this can occur by
4677 default in OpenSSL DHE based SSL/TLS ciphersuites.
4679 This issue was reported to OpenSSL by Hanno Böck.
4684 * Certificate verify crash with missing PSS parameter
4686 The signature verification routines will crash with a NULL pointer
4687 dereference if presented with an ASN.1 signature using the RSA PSS
4688 algorithm and absent mask generation function parameter. Since these
4689 routines are used to verify certificate signature algorithms this can be
4690 used to crash any certificate verification operation and exploited in a
4691 DoS attack. Any application which performs certificate verification is
4692 vulnerable including OpenSSL clients and servers which enable client
4695 This issue was reported to OpenSSL by Loïc Jonas Etienne (Qnective AG).
4700 * X509_ATTRIBUTE memory leak
4702 When presented with a malformed X509_ATTRIBUTE structure OpenSSL will leak
4703 memory. This structure is used by the PKCS#7 and CMS routines so any
4704 application which reads PKCS#7 or CMS data from untrusted sources is
4705 affected. SSL/TLS is not affected.
4707 This issue was reported to OpenSSL by Adam Langley (Google/BoringSSL) using
4713 * Rewrite EVP_DecodeUpdate (base64 decoding) to fix several bugs.
4714 This changes the decoding behaviour for some invalid messages,
4715 though the change is mostly in the more lenient direction, and
4716 legacy behaviour is preserved as much as possible.
4720 * In DSA_generate_parameters_ex, if the provided seed is too short,
4723 *Rich Salz and Ismo Puustinen <ismo.puustinen@intel.com>*
4725 ### Changes between 1.0.2c and 1.0.2d [9 Jul 2015]
4727 * Alternate chains certificate forgery
4729 During certificate verification, OpenSSL will attempt to find an
4730 alternative certificate chain if the first attempt to build such a chain
4731 fails. An error in the implementation of this logic can mean that an
4732 attacker could cause certain checks on untrusted certificates to be
4733 bypassed, such as the CA flag, enabling them to use a valid leaf
4734 certificate to act as a CA and "issue" an invalid certificate.
4736 This issue was reported to OpenSSL by Adam Langley/David Benjamin
4741 ### Changes between 1.0.2b and 1.0.2c [12 Jun 2015]
4743 * Fix HMAC ABI incompatibility. The previous version introduced an ABI
4744 incompatibility in the handling of HMAC. The previous ABI has now been
4749 ### Changes between 1.0.2a and 1.0.2b [11 Jun 2015]
4751 * Malformed ECParameters causes infinite loop
4753 When processing an ECParameters structure OpenSSL enters an infinite loop
4754 if the curve specified is over a specially malformed binary polynomial
4757 This can be used to perform denial of service against any
4758 system which processes public keys, certificate requests or
4759 certificates. This includes TLS clients and TLS servers with
4760 client authentication enabled.
4762 This issue was reported to OpenSSL by Joseph Barr-Pixton.
4767 * Exploitable out-of-bounds read in X509_cmp_time
4769 X509_cmp_time does not properly check the length of the ASN1_TIME
4770 string and can read a few bytes out of bounds. In addition,
4771 X509_cmp_time accepts an arbitrary number of fractional seconds in the
4774 An attacker can use this to craft malformed certificates and CRLs of
4775 various sizes and potentially cause a segmentation fault, resulting in
4776 a DoS on applications that verify certificates or CRLs. TLS clients
4777 that verify CRLs are affected. TLS clients and servers with client
4778 authentication enabled may be affected if they use custom verification
4781 This issue was reported to OpenSSL by Robert Swiecki (Google), and
4782 independently by Hanno Böck.
4787 * PKCS7 crash with missing EnvelopedContent
4789 The PKCS#7 parsing code does not handle missing inner EncryptedContent
4790 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs
4791 with missing content and trigger a NULL pointer dereference on parsing.
4793 Applications that decrypt PKCS#7 data or otherwise parse PKCS#7
4794 structures from untrusted sources are affected. OpenSSL clients and
4795 servers are not affected.
4797 This issue was reported to OpenSSL by Michal Zalewski (Google).
4802 * CMS verify infinite loop with unknown hash function
4804 When verifying a signedData message the CMS code can enter an infinite loop
4805 if presented with an unknown hash function OID. This can be used to perform
4806 denial of service against any system which verifies signedData messages using
4808 This issue was reported to OpenSSL by Johannes Bauer.
4813 * Race condition handling NewSessionTicket
4815 If a NewSessionTicket is received by a multi-threaded client when attempting to
4816 reuse a previous ticket then a race condition can occur potentially leading to
4817 a double free of the ticket data.
4822 * Only support 256-bit or stronger elliptic curves with the
4823 'ecdh_auto' setting (server) or by default (client). Of supported
4824 curves, prefer P-256 (both).
4828 ### Changes between 1.0.2 and 1.0.2a [19 Mar 2015]
4830 * ClientHello sigalgs DoS fix
4832 If a client connects to an OpenSSL 1.0.2 server and renegotiates with an
4833 invalid signature algorithms extension a NULL pointer dereference will
4834 occur. This can be exploited in a DoS attack against the server.
4836 This issue was was reported to OpenSSL by David Ramos of Stanford
4840 *Stephen Henson and Matt Caswell*
4842 * Multiblock corrupted pointer fix
4844 OpenSSL 1.0.2 introduced the "multiblock" performance improvement. This
4845 feature only applies on 64 bit x86 architecture platforms that support AES
4846 NI instructions. A defect in the implementation of "multiblock" can cause
4847 OpenSSL's internal write buffer to become incorrectly set to NULL when
4848 using non-blocking IO. Typically, when the user application is using a
4849 socket BIO for writing, this will only result in a failed connection.
4850 However if some other BIO is used then it is likely that a segmentation
4851 fault will be triggered, thus enabling a potential DoS attack.
4853 This issue was reported to OpenSSL by Daniel Danner and Rainer Mueller.
4858 * Segmentation fault in DTLSv1_listen fix
4860 The DTLSv1_listen function is intended to be stateless and processes the
4861 initial ClientHello from many peers. It is common for user code to loop
4862 over the call to DTLSv1_listen until a valid ClientHello is received with
4863 an associated cookie. A defect in the implementation of DTLSv1_listen means
4864 that state is preserved in the SSL object from one invocation to the next
4865 that can lead to a segmentation fault. Errors processing the initial
4866 ClientHello can trigger this scenario. An example of such an error could be
4867 that a DTLS1.0 only client is attempting to connect to a DTLS1.2 only
4870 This issue was reported to OpenSSL by Per Allansson.
4875 * Segmentation fault in ASN1_TYPE_cmp fix
4877 The function ASN1_TYPE_cmp will crash with an invalid read if an attempt is
4878 made to compare ASN.1 boolean types. Since ASN1_TYPE_cmp is used to check
4879 certificate signature algorithm consistency this can be used to crash any
4880 certificate verification operation and exploited in a DoS attack. Any
4881 application which performs certificate verification is vulnerable including
4882 OpenSSL clients and servers which enable client authentication.
4887 * Segmentation fault for invalid PSS parameters fix
4889 The signature verification routines will crash with a NULL pointer
4890 dereference if presented with an ASN.1 signature using the RSA PSS
4891 algorithm and invalid parameters. Since these routines are used to verify
4892 certificate signature algorithms this can be used to crash any
4893 certificate verification operation and exploited in a DoS attack. Any
4894 application which performs certificate verification is vulnerable including
4895 OpenSSL clients and servers which enable client authentication.
4897 This issue was was reported to OpenSSL by Brian Carpenter.
4902 * ASN.1 structure reuse memory corruption fix
4904 Reusing a structure in ASN.1 parsing may allow an attacker to cause
4905 memory corruption via an invalid write. Such reuse is and has been
4906 strongly discouraged and is believed to be rare.
4908 Applications that parse structures containing CHOICE or ANY DEFINED BY
4909 components may be affected. Certificate parsing (d2i_X509 and related
4910 functions) are however not affected. OpenSSL clients and servers are
4916 * PKCS7 NULL pointer dereferences fix
4918 The PKCS#7 parsing code does not handle missing outer ContentInfo
4919 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs with
4920 missing content and trigger a NULL pointer dereference on parsing.
4922 Applications that verify PKCS#7 signatures, decrypt PKCS#7 data or
4923 otherwise parse PKCS#7 structures from untrusted sources are
4924 affected. OpenSSL clients and servers are not affected.
4926 This issue was reported to OpenSSL by Michal Zalewski (Google).
4931 * DoS via reachable assert in SSLv2 servers fix
4933 A malicious client can trigger an OPENSSL_assert (i.e., an abort) in
4934 servers that both support SSLv2 and enable export cipher suites by sending
4935 a specially crafted SSLv2 CLIENT-MASTER-KEY message.
4937 This issue was discovered by Sean Burford (Google) and Emilia Käsper
4938 (OpenSSL development team).
4943 * Empty CKE with client auth and DHE fix
4945 If client auth is used then a server can seg fault in the event of a DHE
4946 ciphersuite being selected and a zero length ClientKeyExchange message
4947 being sent by the client. This could be exploited in a DoS attack.
4952 * Handshake with unseeded PRNG fix
4954 Under certain conditions an OpenSSL 1.0.2 client can complete a handshake
4955 with an unseeded PRNG. The conditions are:
4956 - The client is on a platform where the PRNG has not been seeded
4957 automatically, and the user has not seeded manually
4958 - A protocol specific client method version has been used (i.e. not
4959 SSL_client_methodv23)
4960 - A ciphersuite is used that does not require additional random data from
4961 the PRNG beyond the initial ClientHello client random (e.g. PSK-RC4-SHA).
4963 If the handshake succeeds then the client random that has been used will
4964 have been generated from a PRNG with insufficient entropy and therefore the
4965 output may be predictable.
4967 For example using the following command with an unseeded openssl will
4968 succeed on an unpatched platform:
4970 openssl s_client -psk 1a2b3c4d -tls1_2 -cipher PSK-RC4-SHA
4975 * Use After Free following d2i_ECPrivatekey error fix
4977 A malformed EC private key file consumed via the d2i_ECPrivateKey function
4978 could cause a use after free condition. This, in turn, could cause a double
4979 free in several private key parsing functions (such as d2i_PrivateKey
4980 or EVP_PKCS82PKEY) and could lead to a DoS attack or memory corruption
4981 for applications that receive EC private keys from untrusted
4982 sources. This scenario is considered rare.
4984 This issue was discovered by the BoringSSL project and fixed in their
4990 * X509_to_X509_REQ NULL pointer deref fix
4992 The function X509_to_X509_REQ will crash with a NULL pointer dereference if
4993 the certificate key is invalid. This function is rarely used in practice.
4995 This issue was discovered by Brian Carpenter.
5000 * Removed the export ciphers from the DEFAULT ciphers
5004 ### Changes between 1.0.1l and 1.0.2 [22 Jan 2015]
5006 * Facilitate "universal" ARM builds targeting range of ARM ISAs, e.g.
5007 ARMv5 through ARMv8, as opposite to "locking" it to single one.
5008 So far those who have to target multiple platforms would compromise
5009 and argue that binary targeting say ARMv5 would still execute on
5010 ARMv8. "Universal" build resolves this compromise by providing
5011 near-optimal performance even on newer platforms.
5015 * Accelerated NIST P-256 elliptic curve implementation for x86_64
5016 (other platforms pending).
5018 *Shay Gueron & Vlad Krasnov (Intel Corp), Andy Polyakov*
5020 * Add support for the SignedCertificateTimestampList certificate and
5021 OCSP response extensions from RFC6962.
5025 * Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
5026 for corner cases. (Certain input points at infinity could lead to
5027 bogus results, with non-infinity inputs mapped to infinity too.)
5031 * Initial support for PowerISA 2.0.7, first implemented in POWER8.
5032 This covers AES, SHA256/512 and GHASH. "Initial" means that most
5033 common cases are optimized and there still is room for further
5034 improvements. Vector Permutation AES for Altivec is also added.
5038 * Add support for little-endian ppc64 Linux target.
5040 *Marcelo Cerri (IBM)*
5042 * Initial support for AMRv8 ISA crypto extensions. This covers AES,
5043 SHA1, SHA256 and GHASH. "Initial" means that most common cases
5044 are optimized and there still is room for further improvements.
5045 Both 32- and 64-bit modes are supported.
5047 *Andy Polyakov, Ard Biesheuvel (Linaro)*
5049 * Improved ARMv7 NEON support.
5053 * Support for SPARC Architecture 2011 crypto extensions, first
5054 implemented in SPARC T4. This covers AES, DES, Camellia, SHA1,
5055 SHA256/512, MD5, GHASH and modular exponentiation.
5057 *Andy Polyakov, David Miller*
5059 * Accelerated modular exponentiation for Intel processors, a.k.a.
5062 *Shay Gueron & Vlad Krasnov (Intel Corp)*
5064 * Support for new and upcoming Intel processors, including AVX2,
5065 BMI and SHA ISA extensions. This includes additional "stitched"
5066 implementations, AESNI-SHA256 and GCM, and multi-buffer support
5069 This work was sponsored by Intel Corp.
5073 * Support for DTLS 1.2. This adds two sets of DTLS methods: DTLS_*_method()
5074 supports both DTLS 1.2 and 1.0 and should use whatever version the peer
5075 supports and DTLSv1_2_*_method() which supports DTLS 1.2 only.
5079 * Use algorithm specific chains in SSL_CTX_use_certificate_chain_file():
5080 this fixes a limitation in previous versions of OpenSSL.
5084 * Extended RSA OAEP support via EVP_PKEY API. Options to specify digest,
5085 MGF1 digest and OAEP label.
5089 * Add EVP support for key wrapping algorithms, to avoid problems with
5090 existing code the flag EVP_CIPHER_CTX_WRAP_ALLOW has to be set in
5091 the EVP_CIPHER_CTX or an error is returned. Add AES and DES3 wrap
5092 algorithms and include tests cases.
5096 * Add functions to allocate and set the fields of an ECDSA_METHOD
5099 *Douglas E. Engert, Steve Henson*
5101 * New functions OPENSSL_gmtime_diff and ASN1_TIME_diff to find the
5102 difference in days and seconds between two tm or ASN1_TIME structures.
5106 * Add -rev test option to s_server to just reverse order of characters
5107 received by client and send back to server. Also prints an abbreviated
5108 summary of the connection parameters.
5112 * New option -brief for s_client and s_server to print out a brief summary
5113 of connection parameters.
5117 * Add callbacks for arbitrary TLS extensions.
5119 *Trevor Perrin <trevp@trevp.net> and Ben Laurie*
5121 * New option -crl_download in several openssl utilities to download CRLs
5122 from CRLDP extension in certificates.
5126 * New options -CRL and -CRLform for s_client and s_server for CRLs.
5130 * New function X509_CRL_diff to generate a delta CRL from the difference
5131 of two full CRLs. Add support to "crl" utility.
5135 * New functions to set lookup_crls function and to retrieve
5136 X509_STORE from X509_STORE_CTX.
5140 * Print out deprecated issuer and subject unique ID fields in
5145 * Extend OCSP I/O functions so they can be used for simple general purpose
5146 HTTP as well as OCSP. New wrapper function which can be used to download
5147 CRLs using the OCSP API.
5151 * Delegate command line handling in s_client/s_server to SSL_CONF APIs.
5155 * `SSL_CONF*` functions. These provide a common framework for application
5156 configuration using configuration files or command lines.
5160 * SSL/TLS tracing code. This parses out SSL/TLS records using the
5161 message callback and prints the results. Needs compile time option
5162 "enable-ssl-trace". New options to s_client and s_server to enable
5167 * New ctrl and macro to retrieve supported points extensions.
5168 Print out extension in s_server and s_client.
5172 * New functions to retrieve certificate signature and signature
5177 * Add functions to retrieve and manipulate the raw cipherlist sent by a
5182 * New Suite B modes for TLS code. These use and enforce the requirements
5183 of RFC6460: restrict ciphersuites, only permit Suite B algorithms and
5184 only use Suite B curves. The Suite B modes can be set by using the
5185 strings "SUITEB128", "SUITEB192" or "SUITEB128ONLY" for the cipherstring.
5189 * New chain verification flags for Suite B levels of security. Check
5190 algorithms are acceptable when flags are set in X509_verify_cert.
5194 * Make tls1_check_chain return a set of flags indicating checks passed
5195 by a certificate chain. Add additional tests to handle client
5196 certificates: checks for matching certificate type and issuer name
5201 * If an attempt is made to use a signature algorithm not in the peer
5202 preference list abort the handshake. If client has no suitable
5203 signature algorithms in response to a certificate request do not
5204 use the certificate.
5208 * If server EC tmp key is not in client preference list abort handshake.
5212 * Add support for certificate stores in CERT structure. This makes it
5213 possible to have different stores per SSL structure or one store in
5214 the parent SSL_CTX. Include distinct stores for certificate chain
5215 verification and chain building. New ctrl SSL_CTRL_BUILD_CERT_CHAIN
5216 to build and store a certificate chain in CERT structure: returning
5217 an error if the chain cannot be built: this will allow applications
5218 to test if a chain is correctly configured.
5220 Note: if the CERT based stores are not set then the parent SSL_CTX
5221 store is used to retain compatibility with existing behaviour.
5225 * New function ssl_set_client_disabled to set a ciphersuite disabled
5226 mask based on the current session, check mask when sending client
5227 hello and checking the requested ciphersuite.
5231 * New ctrls to retrieve and set certificate types in a certificate
5232 request message. Print out received values in s_client. If certificate
5233 types is not set with custom values set sensible values based on
5234 supported signature algorithms.
5238 * Support for distinct client and server supported signature algorithms.
5242 * Add certificate callback. If set this is called whenever a certificate
5243 is required by client or server. An application can decide which
5244 certificate chain to present based on arbitrary criteria: for example
5245 supported signature algorithms. Add very simple example to s_server.
5246 This fixes many of the problems and restrictions of the existing client
5247 certificate callback: for example you can now clear an existing
5248 certificate and specify the whole chain.
5252 * Add new "valid_flags" field to CERT_PKEY structure which determines what
5253 the certificate can be used for (if anything). Set valid_flags field
5254 in new tls1_check_chain function. Simplify ssl_set_cert_masks which used
5255 to have similar checks in it.
5257 Add new "cert_flags" field to CERT structure and include a "strict mode".
5258 This enforces some TLS certificate requirements (such as only permitting
5259 certificate signature algorithms contained in the supported algorithms
5260 extension) which some implementations ignore: this option should be used
5261 with caution as it could cause interoperability issues.
5265 * Update and tidy signature algorithm extension processing. Work out
5266 shared signature algorithms based on preferences and peer algorithms
5267 and print them out in s_client and s_server. Abort handshake if no
5268 shared signature algorithms.
5272 * Add new functions to allow customised supported signature algorithms
5273 for SSL and SSL_CTX structures. Add options to s_client and s_server
5278 * New function SSL_certs_clear() to delete all references to certificates
5279 from an SSL structure. Before this once a certificate had been added
5280 it couldn't be removed.
5284 * Integrate hostname, email address and IP address checking with certificate
5285 verification. New verify options supporting checking in openssl utility.
5289 * Fixes and wildcard matching support to hostname and email checking
5290 functions. Add manual page.
5292 *Florian Weimer (Red Hat Product Security Team)*
5294 * New functions to check a hostname email or IP address against a
5295 certificate. Add options x509 utility to print results of checks against
5300 * Fix OCSP checking.
5302 *Rob Stradling <rob.stradling@comodo.com> and Ben Laurie*
5304 * Initial experimental support for explicitly trusted non-root CAs.
5305 OpenSSL still tries to build a complete chain to a root but if an
5306 intermediate CA has a trust setting included that is used. The first
5307 setting is used: whether to trust (e.g., -addtrust option to the x509
5312 * Add -trusted_first option which attempts to find certificates in the
5313 trusted store even if an untrusted chain is also supplied.
5317 * MIPS assembly pack updates: support for MIPS32r2 and SmartMIPS ASE,
5318 platform support for Linux and Android.
5322 * Support for linux-x32, ILP32 environment in x86_64 framework.
5326 * Experimental multi-implementation support for FIPS capable OpenSSL.
5327 When in FIPS mode the approved implementations are used as normal,
5328 when not in FIPS mode the internal unapproved versions are used instead.
5329 This means that the FIPS capable OpenSSL isn't forced to use the
5330 (often lower performance) FIPS implementations outside FIPS mode.
5334 * Transparently support X9.42 DH parameters when calling
5335 PEM_read_bio_DHparameters. This means existing applications can handle
5336 the new parameter format automatically.
5340 * Initial experimental support for X9.42 DH parameter format: mainly
5341 to support use of 'q' parameter for RFC5114 parameters.
5345 * Add DH parameters from RFC5114 including test data to dhtest.
5349 * Support for automatic EC temporary key parameter selection. If enabled
5350 the most preferred EC parameters are automatically used instead of
5351 hardcoded fixed parameters. Now a server just has to call:
5352 SSL_CTX_set_ecdh_auto(ctx, 1) and the server will automatically
5353 support ECDH and use the most appropriate parameters.
5357 * Enhance and tidy EC curve and point format TLS extension code. Use
5358 static structures instead of allocation if default values are used.
5359 New ctrls to set curves we wish to support and to retrieve shared curves.
5360 Print out shared curves in s_server. New options to s_server and s_client
5361 to set list of supported curves.
5365 * New ctrls to retrieve supported signature algorithms and
5366 supported curve values as an array of NIDs. Extend openssl utility
5367 to print out received values.
5371 * Add new APIs EC_curve_nist2nid and EC_curve_nid2nist which convert
5372 between NIDs and the more common NIST names such as "P-256". Enhance
5373 ecparam utility and ECC method to recognise the NIST names for curves.
5377 * Enhance SSL/TLS certificate chain handling to support different
5378 chains for each certificate instead of one chain in the parent SSL_CTX.
5382 * Support for fixed DH ciphersuite client authentication: where both
5383 server and client use DH certificates with common parameters.
5387 * Support for fixed DH ciphersuites: those requiring DH server
5392 * New function i2d_re_X509_tbs for re-encoding the TBS portion of
5394 Note: Related 1.0.2-beta specific macros X509_get_cert_info,
5395 X509_CINF_set_modified, X509_CINF_get_issuer, X509_CINF_get_extensions and
5396 X509_CINF_get_signature were reverted post internal team review.
5401 ### Changes between 1.0.1t and 1.0.1u [22 Sep 2016]
5403 * OCSP Status Request extension unbounded memory growth
5405 A malicious client can send an excessively large OCSP Status Request
5406 extension. If that client continually requests renegotiation, sending a
5407 large OCSP Status Request extension each time, then there will be unbounded
5408 memory growth on the server. This will eventually lead to a Denial Of
5409 Service attack through memory exhaustion. Servers with a default
5410 configuration are vulnerable even if they do not support OCSP. Builds using
5411 the "no-ocsp" build time option are not affected.
5413 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5418 * In order to mitigate the SWEET32 attack, the DES ciphers were moved from
5421 This issue was reported to OpenSSL Karthikeyan Bhargavan and Gaetan
5427 * OOB write in MDC2_Update()
5429 An overflow can occur in MDC2_Update() either if called directly or
5430 through the EVP_DigestUpdate() function using MDC2. If an attacker
5431 is able to supply very large amounts of input data after a previous
5432 call to EVP_EncryptUpdate() with a partial block then a length check
5433 can overflow resulting in a heap corruption.
5435 The amount of data needed is comparable to SIZE_MAX which is impractical
5438 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5443 * Malformed SHA512 ticket DoS
5445 If a server uses SHA512 for TLS session ticket HMAC it is vulnerable to a
5446 DoS attack where a malformed ticket will result in an OOB read which will
5449 The use of SHA512 in TLS session tickets is comparatively rare as it requires
5450 a custom server callback and ticket lookup mechanism.
5452 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5457 * OOB write in BN_bn2dec()
5459 The function BN_bn2dec() does not check the return value of BN_div_word().
5460 This can cause an OOB write if an application uses this function with an
5461 overly large BIGNUM. This could be a problem if an overly large certificate
5462 or CRL is printed out from an untrusted source. TLS is not affected because
5463 record limits will reject an oversized certificate before it is parsed.
5465 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5470 * OOB read in TS_OBJ_print_bio()
5472 The function TS_OBJ_print_bio() misuses OBJ_obj2txt(): the return value is
5473 the total length the OID text representation would use and not the amount
5474 of data written. This will result in OOB reads when large OIDs are
5477 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5482 * Pointer arithmetic undefined behaviour
5484 Avoid some undefined pointer arithmetic
5486 A common idiom in the codebase is to check limits in the following manner:
5489 Where "p" points to some malloc'd data of SIZE bytes and
5492 "len" here could be from some externally supplied data (e.g. from a TLS
5495 The rules of C pointer arithmetic are such that "p + len" is only well
5496 defined where len <= SIZE. Therefore the above idiom is actually
5497 undefined behaviour.
5499 For example this could cause problems if some malloc implementation
5500 provides an address for "p" such that "p + len" actually overflows for
5501 values of len that are too big and therefore p + len < limit.
5503 This issue was reported to OpenSSL by Guido Vranken
5508 * Constant time flag not preserved in DSA signing
5510 Operations in the DSA signing algorithm should run in constant time in
5511 order to avoid side channel attacks. A flaw in the OpenSSL DSA
5512 implementation means that a non-constant time codepath is followed for
5513 certain operations. This has been demonstrated through a cache-timing
5514 attack to be sufficient for an attacker to recover the private DSA key.
5516 This issue was reported by César Pereida (Aalto University), Billy Brumley
5517 (Tampere University of Technology), and Yuval Yarom (The University of
5518 Adelaide and NICTA).
5523 * DTLS buffered message DoS
5525 In a DTLS connection where handshake messages are delivered out-of-order
5526 those messages that OpenSSL is not yet ready to process will be buffered
5527 for later use. Under certain circumstances, a flaw in the logic means that
5528 those messages do not get removed from the buffer even though the handshake
5529 has been completed. An attacker could force up to approx. 15 messages to
5530 remain in the buffer when they are no longer required. These messages will
5531 be cleared when the DTLS connection is closed. The default maximum size for
5532 a message is 100k. Therefore the attacker could force an additional 1500k
5533 to be consumed per connection. By opening many simulataneous connections an
5534 attacker could cause a DoS attack through memory exhaustion.
5536 This issue was reported to OpenSSL by Quan Luo.
5541 * DTLS replay protection DoS
5543 A flaw in the DTLS replay attack protection mechanism means that records
5544 that arrive for future epochs update the replay protection "window" before
5545 the MAC for the record has been validated. This could be exploited by an
5546 attacker by sending a record for the next epoch (which does not have to
5547 decrypt or have a valid MAC), with a very large sequence number. This means
5548 that all subsequent legitimate packets are dropped causing a denial of
5549 service for a specific DTLS connection.
5551 This issue was reported to OpenSSL by the OCAP audit team.
5556 * Certificate message OOB reads
5558 In OpenSSL 1.0.2 and earlier some missing message length checks can result
5559 in OOB reads of up to 2 bytes beyond an allocated buffer. There is a
5560 theoretical DoS risk but this has not been observed in practice on common
5563 The messages affected are client certificate, client certificate request
5564 and server certificate. As a result the attack can only be performed
5565 against a client or a server which enables client authentication.
5567 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5572 ### Changes between 1.0.1s and 1.0.1t [3 May 2016]
5574 * Prevent padding oracle in AES-NI CBC MAC check
5576 A MITM attacker can use a padding oracle attack to decrypt traffic
5577 when the connection uses an AES CBC cipher and the server support
5580 This issue was introduced as part of the fix for Lucky 13 padding
5581 attack [CVE-2013-0169][]. The padding check was rewritten to be in
5582 constant time by making sure that always the same bytes are read and
5583 compared against either the MAC or padding bytes. But it no longer
5584 checked that there was enough data to have both the MAC and padding
5587 This issue was reported by Juraj Somorovsky using TLS-Attacker.
5592 * Fix EVP_EncodeUpdate overflow
5594 An overflow can occur in the EVP_EncodeUpdate() function which is used for
5595 Base64 encoding of binary data. If an attacker is able to supply very large
5596 amounts of input data then a length check can overflow resulting in a heap
5599 Internally to OpenSSL the EVP_EncodeUpdate() function is primarly used by
5600 the `PEM_write_bio*` family of functions. These are mainly used within the
5601 OpenSSL command line applications, so any application which processes data
5602 from an untrusted source and outputs it as a PEM file should be considered
5603 vulnerable to this issue. User applications that call these APIs directly
5604 with large amounts of untrusted data may also be vulnerable.
5606 This issue was reported by Guido Vranken.
5611 * Fix EVP_EncryptUpdate overflow
5613 An overflow can occur in the EVP_EncryptUpdate() function. If an attacker
5614 is able to supply very large amounts of input data after a previous call to
5615 EVP_EncryptUpdate() with a partial block then a length check can overflow
5616 resulting in a heap corruption. Following an analysis of all OpenSSL
5617 internal usage of the EVP_EncryptUpdate() function all usage is one of two
5618 forms. The first form is where the EVP_EncryptUpdate() call is known to be
5619 the first called function after an EVP_EncryptInit(), and therefore that
5620 specific call must be safe. The second form is where the length passed to
5621 EVP_EncryptUpdate() can be seen from the code to be some small value and
5622 therefore there is no possibility of an overflow. Since all instances are
5623 one of these two forms, it is believed that there can be no overflows in
5624 internal code due to this problem. It should be noted that
5625 EVP_DecryptUpdate() can call EVP_EncryptUpdate() in certain code paths.
5626 Also EVP_CipherUpdate() is a synonym for EVP_EncryptUpdate(). All instances
5627 of these calls have also been analysed too and it is believed there are no
5628 instances in internal usage where an overflow could occur.
5630 This issue was reported by Guido Vranken.
5635 * Prevent ASN.1 BIO excessive memory allocation
5637 When ASN.1 data is read from a BIO using functions such as d2i_CMS_bio()
5638 a short invalid encoding can casuse allocation of large amounts of memory
5639 potentially consuming excessive resources or exhausting memory.
5641 Any application parsing untrusted data through d2i BIO functions is
5642 affected. The memory based functions such as d2i_X509() are *not* affected.
5643 Since the memory based functions are used by the TLS library, TLS
5644 applications are not affected.
5646 This issue was reported by Brian Carpenter.
5653 ASN1 Strings that are over 1024 bytes can cause an overread in applications
5654 using the X509_NAME_oneline() function on EBCDIC systems. This could result
5655 in arbitrary stack data being returned in the buffer.
5657 This issue was reported by Guido Vranken.
5662 * Modify behavior of ALPN to invoke callback after SNI/servername
5663 callback, such that updates to the SSL_CTX affect ALPN.
5667 * Remove LOW from the DEFAULT cipher list. This removes singles DES from the
5672 * Only remove the SSLv2 methods with the no-ssl2-method option. When the
5673 methods are enabled and ssl2 is disabled the methods return NULL.
5677 ### Changes between 1.0.1r and 1.0.1s [1 Mar 2016]
5679 * Disable weak ciphers in SSLv3 and up in default builds of OpenSSL.
5680 Builds that are not configured with "enable-weak-ssl-ciphers" will not
5681 provide any "EXPORT" or "LOW" strength ciphers.
5685 * Disable SSLv2 default build, default negotiation and weak ciphers. SSLv2
5686 is by default disabled at build-time. Builds that are not configured with
5687 "enable-ssl2" will not support SSLv2. Even if "enable-ssl2" is used,
5688 users who want to negotiate SSLv2 via the version-flexible SSLv23_method()
5689 will need to explicitly call either of:
5691 SSL_CTX_clear_options(ctx, SSL_OP_NO_SSLv2);
5693 SSL_clear_options(ssl, SSL_OP_NO_SSLv2);
5695 as appropriate. Even if either of those is used, or the application
5696 explicitly uses the version-specific SSLv2_method() or its client and
5697 server variants, SSLv2 ciphers vulnerable to exhaustive search key
5698 recovery have been removed. Specifically, the SSLv2 40-bit EXPORT
5699 ciphers, and SSLv2 56-bit DES are no longer available.
5704 * Fix a double-free in DSA code
5706 A double free bug was discovered when OpenSSL parses malformed DSA private
5707 keys and could lead to a DoS attack or memory corruption for applications
5708 that receive DSA private keys from untrusted sources. This scenario is
5711 This issue was reported to OpenSSL by Adam Langley(Google/BoringSSL) using
5717 * Disable SRP fake user seed to address a server memory leak.
5719 Add a new method SRP_VBASE_get1_by_user that handles the seed properly.
5721 SRP_VBASE_get_by_user had inconsistent memory management behaviour.
5722 In order to fix an unavoidable memory leak, SRP_VBASE_get_by_user
5723 was changed to ignore the "fake user" SRP seed, even if the seed
5726 Users should use SRP_VBASE_get1_by_user instead. Note that in
5727 SRP_VBASE_get1_by_user, caller must free the returned value. Note
5728 also that even though configuring the SRP seed attempts to hide
5729 invalid usernames by continuing the handshake with fake
5730 credentials, this behaviour is not constant time and no strong
5731 guarantees are made that the handshake is indistinguishable from
5732 that of a valid user.
5737 * Fix BN_hex2bn/BN_dec2bn NULL pointer deref/heap corruption
5739 In the BN_hex2bn function the number of hex digits is calculated using an
5740 int value `i`. Later `bn_expand` is called with a value of `i * 4`. For
5741 large values of `i` this can result in `bn_expand` not allocating any
5742 memory because `i * 4` is negative. This can leave the internal BIGNUM data
5743 field as NULL leading to a subsequent NULL ptr deref. For very large values
5744 of `i`, the calculation `i * 4` could be a positive value smaller than `i`.
5745 In this case memory is allocated to the internal BIGNUM data field, but it
5746 is insufficiently sized leading to heap corruption. A similar issue exists
5747 in BN_dec2bn. This could have security consequences if BN_hex2bn/BN_dec2bn
5748 is ever called by user applications with very large untrusted hex/dec data.
5749 This is anticipated to be a rare occurrence.
5751 All OpenSSL internal usage of these functions use data that is not expected
5752 to be untrusted, e.g. config file data or application command line
5753 arguments. If user developed applications generate config file data based
5754 on untrusted data then it is possible that this could also lead to security
5755 consequences. This is also anticipated to be rare.
5757 This issue was reported to OpenSSL by Guido Vranken.
5762 * Fix memory issues in `BIO_*printf` functions
5764 The internal `fmtstr` function used in processing a "%s" format string in
5765 the `BIO_*printf` functions could overflow while calculating the length of a
5766 string and cause an OOB read when printing very long strings.
5768 Additionally the internal `doapr_outch` function can attempt to write to an
5769 OOB memory location (at an offset from the NULL pointer) in the event of a
5770 memory allocation failure. In 1.0.2 and below this could be caused where
5771 the size of a buffer to be allocated is greater than INT_MAX. E.g. this
5772 could be in processing a very long "%s" format string. Memory leaks can
5775 The first issue may mask the second issue dependent on compiler behaviour.
5776 These problems could enable attacks where large amounts of untrusted data
5777 is passed to the `BIO_*printf` functions. If applications use these functions
5778 in this way then they could be vulnerable. OpenSSL itself uses these
5779 functions when printing out human-readable dumps of ASN.1 data. Therefore
5780 applications that print this data could be vulnerable if the data is from
5781 untrusted sources. OpenSSL command line applications could also be
5782 vulnerable where they print out ASN.1 data, or if untrusted data is passed
5783 as command line arguments.
5785 Libssl is not considered directly vulnerable. Additionally certificates etc
5786 received via remote connections via libssl are also unlikely to be able to
5787 trigger these issues because of message size limits enforced within libssl.
5789 This issue was reported to OpenSSL Guido Vranken.
5794 * Side channel attack on modular exponentiation
5796 A side-channel attack was found which makes use of cache-bank conflicts on
5797 the Intel Sandy-Bridge microarchitecture which could lead to the recovery
5798 of RSA keys. The ability to exploit this issue is limited as it relies on
5799 an attacker who has control of code in a thread running on the same
5800 hyper-threaded core as the victim thread which is performing decryptions.
5802 This issue was reported to OpenSSL by Yuval Yarom, The University of
5803 Adelaide and NICTA, Daniel Genkin, Technion and Tel Aviv University, and
5804 Nadia Heninger, University of Pennsylvania with more information at
5805 <http://cachebleed.info>.
5810 * Change the req app to generate a 2048-bit RSA/DSA key by default,
5811 if no keysize is specified with default_bits. This fixes an
5812 omission in an earlier change that changed all RSA/DSA key generation
5813 apps to use 2048 bits by default.
5817 ### Changes between 1.0.1q and 1.0.1r [28 Jan 2016]
5819 * Protection for DH small subgroup attacks
5821 As a precautionary measure the SSL_OP_SINGLE_DH_USE option has been
5822 switched on by default and cannot be disabled. This could have some
5827 * SSLv2 doesn't block disabled ciphers
5829 A malicious client can negotiate SSLv2 ciphers that have been disabled on
5830 the server and complete SSLv2 handshakes even if all SSLv2 ciphers have
5831 been disabled, provided that the SSLv2 protocol was not also disabled via
5834 This issue was reported to OpenSSL on 26th December 2015 by Nimrod Aviram
5835 and Sebastian Schinzel.
5840 * Reject DH handshakes with parameters shorter than 1024 bits.
5844 ### Changes between 1.0.1p and 1.0.1q [3 Dec 2015]
5846 * Certificate verify crash with missing PSS parameter
5848 The signature verification routines will crash with a NULL pointer
5849 dereference if presented with an ASN.1 signature using the RSA PSS
5850 algorithm and absent mask generation function parameter. Since these
5851 routines are used to verify certificate signature algorithms this can be
5852 used to crash any certificate verification operation and exploited in a
5853 DoS attack. Any application which performs certificate verification is
5854 vulnerable including OpenSSL clients and servers which enable client
5857 This issue was reported to OpenSSL by Loïc Jonas Etienne (Qnective AG).
5862 * X509_ATTRIBUTE memory leak
5864 When presented with a malformed X509_ATTRIBUTE structure OpenSSL will leak
5865 memory. This structure is used by the PKCS#7 and CMS routines so any
5866 application which reads PKCS#7 or CMS data from untrusted sources is
5867 affected. SSL/TLS is not affected.
5869 This issue was reported to OpenSSL by Adam Langley (Google/BoringSSL) using
5875 * Rewrite EVP_DecodeUpdate (base64 decoding) to fix several bugs.
5876 This changes the decoding behaviour for some invalid messages,
5877 though the change is mostly in the more lenient direction, and
5878 legacy behaviour is preserved as much as possible.
5882 * In DSA_generate_parameters_ex, if the provided seed is too short,
5883 use a random seed, as already documented.
5885 *Rich Salz and Ismo Puustinen <ismo.puustinen@intel.com>*
5887 ### Changes between 1.0.1o and 1.0.1p [9 Jul 2015]
5889 * Alternate chains certificate forgery
5891 During certificate verfification, OpenSSL will attempt to find an
5892 alternative certificate chain if the first attempt to build such a chain
5893 fails. An error in the implementation of this logic can mean that an
5894 attacker could cause certain checks on untrusted certificates to be
5895 bypassed, such as the CA flag, enabling them to use a valid leaf
5896 certificate to act as a CA and "issue" an invalid certificate.
5898 This issue was reported to OpenSSL by Adam Langley/David Benjamin
5904 * Race condition handling PSK identify hint
5906 If PSK identity hints are received by a multi-threaded client then
5907 the values are wrongly updated in the parent SSL_CTX structure. This can
5908 result in a race condition potentially leading to a double free of the
5914 ### Changes between 1.0.1n and 1.0.1o [12 Jun 2015]
5916 * Fix HMAC ABI incompatibility. The previous version introduced an ABI
5917 incompatibility in the handling of HMAC. The previous ABI has now been
5920 ### Changes between 1.0.1m and 1.0.1n [11 Jun 2015]
5922 * Malformed ECParameters causes infinite loop
5924 When processing an ECParameters structure OpenSSL enters an infinite loop
5925 if the curve specified is over a specially malformed binary polynomial
5928 This can be used to perform denial of service against any
5929 system which processes public keys, certificate requests or
5930 certificates. This includes TLS clients and TLS servers with
5931 client authentication enabled.
5933 This issue was reported to OpenSSL by Joseph Barr-Pixton.
5938 * Exploitable out-of-bounds read in X509_cmp_time
5940 X509_cmp_time does not properly check the length of the ASN1_TIME
5941 string and can read a few bytes out of bounds. In addition,
5942 X509_cmp_time accepts an arbitrary number of fractional seconds in the
5945 An attacker can use this to craft malformed certificates and CRLs of
5946 various sizes and potentially cause a segmentation fault, resulting in
5947 a DoS on applications that verify certificates or CRLs. TLS clients
5948 that verify CRLs are affected. TLS clients and servers with client
5949 authentication enabled may be affected if they use custom verification
5952 This issue was reported to OpenSSL by Robert Swiecki (Google), and
5953 independently by Hanno Böck.
5958 * PKCS7 crash with missing EnvelopedContent
5960 The PKCS#7 parsing code does not handle missing inner EncryptedContent
5961 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs
5962 with missing content and trigger a NULL pointer dereference on parsing.
5964 Applications that decrypt PKCS#7 data or otherwise parse PKCS#7
5965 structures from untrusted sources are affected. OpenSSL clients and
5966 servers are not affected.
5968 This issue was reported to OpenSSL by Michal Zalewski (Google).
5973 * CMS verify infinite loop with unknown hash function
5975 When verifying a signedData message the CMS code can enter an infinite loop
5976 if presented with an unknown hash function OID. This can be used to perform
5977 denial of service against any system which verifies signedData messages using
5979 This issue was reported to OpenSSL by Johannes Bauer.
5984 * Race condition handling NewSessionTicket
5986 If a NewSessionTicket is received by a multi-threaded client when attempting to
5987 reuse a previous ticket then a race condition can occur potentially leading to
5988 a double free of the ticket data.
5993 * Reject DH handshakes with parameters shorter than 768 bits.
5995 *Kurt Roeckx and Emilia Kasper*
5997 * dhparam: generate 2048-bit parameters by default.
5999 *Kurt Roeckx and Emilia Kasper*
6001 ### Changes between 1.0.1l and 1.0.1m [19 Mar 2015]
6003 * Segmentation fault in ASN1_TYPE_cmp fix
6005 The function ASN1_TYPE_cmp will crash with an invalid read if an attempt is
6006 made to compare ASN.1 boolean types. Since ASN1_TYPE_cmp is used to check
6007 certificate signature algorithm consistency this can be used to crash any
6008 certificate verification operation and exploited in a DoS attack. Any
6009 application which performs certificate verification is vulnerable including
6010 OpenSSL clients and servers which enable client authentication.
6015 * ASN.1 structure reuse memory corruption fix
6017 Reusing a structure in ASN.1 parsing may allow an attacker to cause
6018 memory corruption via an invalid write. Such reuse is and has been
6019 strongly discouraged and is believed to be rare.
6021 Applications that parse structures containing CHOICE or ANY DEFINED BY
6022 components may be affected. Certificate parsing (d2i_X509 and related
6023 functions) are however not affected. OpenSSL clients and servers are
6029 * PKCS7 NULL pointer dereferences fix
6031 The PKCS#7 parsing code does not handle missing outer ContentInfo
6032 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs with
6033 missing content and trigger a NULL pointer dereference on parsing.
6035 Applications that verify PKCS#7 signatures, decrypt PKCS#7 data or
6036 otherwise parse PKCS#7 structures from untrusted sources are
6037 affected. OpenSSL clients and servers are not affected.
6039 This issue was reported to OpenSSL by Michal Zalewski (Google).
6044 * DoS via reachable assert in SSLv2 servers fix
6046 A malicious client can trigger an OPENSSL_assert (i.e., an abort) in
6047 servers that both support SSLv2 and enable export cipher suites by sending
6048 a specially crafted SSLv2 CLIENT-MASTER-KEY message.
6050 This issue was discovered by Sean Burford (Google) and Emilia Käsper
6051 (OpenSSL development team).
6056 * Use After Free following d2i_ECPrivatekey error fix
6058 A malformed EC private key file consumed via the d2i_ECPrivateKey function
6059 could cause a use after free condition. This, in turn, could cause a double
6060 free in several private key parsing functions (such as d2i_PrivateKey
6061 or EVP_PKCS82PKEY) and could lead to a DoS attack or memory corruption
6062 for applications that receive EC private keys from untrusted
6063 sources. This scenario is considered rare.
6065 This issue was discovered by the BoringSSL project and fixed in their
6071 * X509_to_X509_REQ NULL pointer deref fix
6073 The function X509_to_X509_REQ will crash with a NULL pointer dereference if
6074 the certificate key is invalid. This function is rarely used in practice.
6076 This issue was discovered by Brian Carpenter.
6081 * Removed the export ciphers from the DEFAULT ciphers
6085 ### Changes between 1.0.1k and 1.0.1l [15 Jan 2015]
6087 * Build fixes for the Windows and OpenVMS platforms
6089 *Matt Caswell and Richard Levitte*
6091 ### Changes between 1.0.1j and 1.0.1k [8 Jan 2015]
6093 * Fix DTLS segmentation fault in dtls1_get_record. A carefully crafted DTLS
6094 message can cause a segmentation fault in OpenSSL due to a NULL pointer
6095 dereference. This could lead to a Denial Of Service attack. Thanks to
6096 Markus Stenberg of Cisco Systems, Inc. for reporting this issue.
6101 * Fix DTLS memory leak in dtls1_buffer_record. A memory leak can occur in the
6102 dtls1_buffer_record function under certain conditions. In particular this
6103 could occur if an attacker sent repeated DTLS records with the same
6104 sequence number but for the next epoch. The memory leak could be exploited
6105 by an attacker in a Denial of Service attack through memory exhaustion.
6106 Thanks to Chris Mueller for reporting this issue.
6111 * Fix issue where no-ssl3 configuration sets method to NULL. When openssl is
6112 built with the no-ssl3 option and a SSL v3 ClientHello is received the ssl
6113 method would be set to NULL which could later result in a NULL pointer
6114 dereference. Thanks to Frank Schmirler for reporting this issue.
6119 * Abort handshake if server key exchange message is omitted for ephemeral
6122 Thanks to Karthikeyan Bhargavan of the PROSECCO team at INRIA for
6123 reporting this issue.
6128 * Remove non-export ephemeral RSA code on client and server. This code
6129 violated the TLS standard by allowing the use of temporary RSA keys in
6130 non-export ciphersuites and could be used by a server to effectively
6131 downgrade the RSA key length used to a value smaller than the server
6132 certificate. Thanks for Karthikeyan Bhargavan of the PROSECCO team at
6133 INRIA or reporting this issue.
6138 * Fixed issue where DH client certificates are accepted without verification.
6139 An OpenSSL server will accept a DH certificate for client authentication
6140 without the certificate verify message. This effectively allows a client to
6141 authenticate without the use of a private key. This only affects servers
6142 which trust a client certificate authority which issues certificates
6143 containing DH keys: these are extremely rare and hardly ever encountered.
6144 Thanks for Karthikeyan Bhargavan of the PROSECCO team at INRIA or reporting
6150 * Ensure that the session ID context of an SSL is updated when its
6151 SSL_CTX is updated via SSL_set_SSL_CTX.
6153 The session ID context is typically set from the parent SSL_CTX,
6154 and can vary with the CTX.
6158 * Fix various certificate fingerprint issues.
6160 By using non-DER or invalid encodings outside the signed portion of a
6161 certificate the fingerprint can be changed without breaking the signature.
6162 Although no details of the signed portion of the certificate can be changed
6163 this can cause problems with some applications: e.g. those using the
6164 certificate fingerprint for blacklists.
6166 1. Reject signatures with non zero unused bits.
6168 If the BIT STRING containing the signature has non zero unused bits reject
6169 the signature. All current signature algorithms require zero unused bits.
6171 2. Check certificate algorithm consistency.
6173 Check the AlgorithmIdentifier inside TBS matches the one in the
6174 certificate signature. NB: this will result in signature failure
6175 errors for some broken certificates.
6177 Thanks to Konrad Kraszewski from Google for reporting this issue.
6179 3. Check DSA/ECDSA signatures use DER.
6181 Re-encode DSA/ECDSA signatures and compare with the original received
6182 signature. Return an error if there is a mismatch.
6184 This will reject various cases including garbage after signature
6185 (thanks to Antti Karjalainen and Tuomo Untinen from the Codenomicon CROSS
6186 program for discovering this case) and use of BER or invalid ASN.1 INTEGERs
6187 (negative or with leading zeroes).
6189 Further analysis was conducted and fixes were developed by Stephen Henson
6190 of the OpenSSL core team.
6196 * Correct Bignum squaring. Bignum squaring (BN_sqr) may produce incorrect
6197 results on some platforms, including x86_64. This bug occurs at random
6198 with a very low probability, and is not known to be exploitable in any
6199 way, though its exact impact is difficult to determine. Thanks to Pieter
6200 Wuille (Blockstream) who reported this issue and also suggested an initial
6201 fix. Further analysis was conducted by the OpenSSL development team and
6202 Adam Langley of Google. The final fix was developed by Andy Polyakov of
6203 the OpenSSL core team.
6208 * Do not resume sessions on the server if the negotiated protocol
6209 version does not match the session's version. Resuming with a different
6210 version, while not strictly forbidden by the RFC, is of questionable
6211 sanity and breaks all known clients.
6213 *David Benjamin, Emilia Käsper*
6215 * Tighten handling of the ChangeCipherSpec (CCS) message: reject
6216 early CCS messages during renegotiation. (Note that because
6217 renegotiation is encrypted, this early CCS was not exploitable.)
6221 * Tighten client-side session ticket handling during renegotiation:
6222 ensure that the client only accepts a session ticket if the server sends
6223 the extension anew in the ServerHello. Previously, a TLS client would
6224 reuse the old extension state and thus accept a session ticket if one was
6225 announced in the initial ServerHello.
6227 Similarly, ensure that the client requires a session ticket if one
6228 was advertised in the ServerHello. Previously, a TLS client would
6229 ignore a missing NewSessionTicket message.
6233 ### Changes between 1.0.1i and 1.0.1j [15 Oct 2014]
6237 A flaw in the DTLS SRTP extension parsing code allows an attacker, who
6238 sends a carefully crafted handshake message, to cause OpenSSL to fail
6239 to free up to 64k of memory causing a memory leak. This could be
6240 exploited in a Denial Of Service attack. This issue affects OpenSSL
6241 1.0.1 server implementations for both SSL/TLS and DTLS regardless of
6242 whether SRTP is used or configured. Implementations of OpenSSL that
6243 have been compiled with OPENSSL_NO_SRTP defined are not affected.
6245 The fix was developed by the OpenSSL team.
6250 * Session Ticket Memory Leak.
6252 When an OpenSSL SSL/TLS/DTLS server receives a session ticket the
6253 integrity of that ticket is first verified. In the event of a session
6254 ticket integrity check failing, OpenSSL will fail to free memory
6255 causing a memory leak. By sending a large number of invalid session
6256 tickets an attacker could exploit this issue in a Denial Of Service
6262 * Build option no-ssl3 is incomplete.
6264 When OpenSSL is configured with "no-ssl3" as a build option, servers
6265 could accept and complete a SSL 3.0 handshake, and clients could be
6266 configured to send them.
6269 *Akamai and the OpenSSL team*
6271 * Add support for TLS_FALLBACK_SCSV.
6272 Client applications doing fallback retries should call
6273 SSL_set_mode(s, SSL_MODE_SEND_FALLBACK_SCSV).
6276 *Adam Langley, Bodo Moeller*
6278 * Add additional DigestInfo checks.
6280 Re-encode DigestInto in DER and check against the original when
6281 verifying RSA signature: this will reject any improperly encoded
6282 DigestInfo structures.
6284 Note: this is a precautionary measure and no attacks are currently known.
6288 ### Changes between 1.0.1h and 1.0.1i [6 Aug 2014]
6290 * Fix SRP buffer overrun vulnerability. Invalid parameters passed to the
6291 SRP code can be overrun an internal buffer. Add sanity check that
6292 g, A, B < N to SRP code.
6294 Thanks to Sean Devlin and Watson Ladd of Cryptography Services, NCC
6295 Group for discovering this issue.
6300 * A flaw in the OpenSSL SSL/TLS server code causes the server to negotiate
6301 TLS 1.0 instead of higher protocol versions when the ClientHello message
6302 is badly fragmented. This allows a man-in-the-middle attacker to force a
6303 downgrade to TLS 1.0 even if both the server and the client support a
6304 higher protocol version, by modifying the client's TLS records.
6306 Thanks to David Benjamin and Adam Langley (Google) for discovering and
6307 researching this issue.
6312 * OpenSSL DTLS clients enabling anonymous (EC)DH ciphersuites are subject
6313 to a denial of service attack. A malicious server can crash the client
6314 with a null pointer dereference (read) by specifying an anonymous (EC)DH
6315 ciphersuite and sending carefully crafted handshake messages.
6317 Thanks to Felix Gröbert (Google) for discovering and researching this
6323 * By sending carefully crafted DTLS packets an attacker could cause openssl
6324 to leak memory. This can be exploited through a Denial of Service attack.
6325 Thanks to Adam Langley for discovering and researching this issue.
6330 * An attacker can force openssl to consume large amounts of memory whilst
6331 processing DTLS handshake messages. This can be exploited through a
6332 Denial of Service attack.
6333 Thanks to Adam Langley for discovering and researching this issue.
6338 * An attacker can force an error condition which causes openssl to crash
6339 whilst processing DTLS packets due to memory being freed twice. This
6340 can be exploited through a Denial of Service attack.
6341 Thanks to Adam Langley and Wan-Teh Chang for discovering and researching
6347 * If a multithreaded client connects to a malicious server using a resumed
6348 session and the server sends an ec point format extension it could write
6349 up to 255 bytes to freed memory.
6351 Thanks to Gabor Tyukasz (LogMeIn Inc) for discovering and researching this
6357 * A malicious server can crash an OpenSSL client with a null pointer
6358 dereference (read) by specifying an SRP ciphersuite even though it was not
6359 properly negotiated with the client. This can be exploited through a
6360 Denial of Service attack.
6362 Thanks to Joonas Kuorilehto and Riku Hietamäki (Codenomicon) for
6363 discovering and researching this issue.
6368 * A flaw in OBJ_obj2txt may cause pretty printing functions such as
6369 X509_name_oneline, X509_name_print_ex et al. to leak some information
6370 from the stack. Applications may be affected if they echo pretty printing
6371 output to the attacker.
6373 Thanks to Ivan Fratric (Google) for discovering this issue.
6376 *Emilia Käsper, and Steve Henson*
6378 * Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
6379 for corner cases. (Certain input points at infinity could lead to
6380 bogus results, with non-infinity inputs mapped to infinity too.)
6384 ### Changes between 1.0.1g and 1.0.1h [5 Jun 2014]
6386 * Fix for SSL/TLS MITM flaw. An attacker using a carefully crafted
6387 handshake can force the use of weak keying material in OpenSSL
6388 SSL/TLS clients and servers.
6390 Thanks to KIKUCHI Masashi (Lepidum Co. Ltd.) for discovering and
6391 researching this issue. [CVE-2014-0224][]
6393 *KIKUCHI Masashi, Steve Henson*
6395 * Fix DTLS recursion flaw. By sending an invalid DTLS handshake to an
6396 OpenSSL DTLS client the code can be made to recurse eventually crashing
6399 Thanks to Imre Rad (Search-Lab Ltd.) for discovering this issue.
6402 *Imre Rad, Steve Henson*
6404 * Fix DTLS invalid fragment vulnerability. A buffer overrun attack can
6405 be triggered by sending invalid DTLS fragments to an OpenSSL DTLS
6406 client or server. This is potentially exploitable to run arbitrary
6407 code on a vulnerable client or server.
6409 Thanks to Jüri Aedla for reporting this issue. [CVE-2014-0195][]
6411 *Jüri Aedla, Steve Henson*
6413 * Fix bug in TLS code where clients enable anonymous ECDH ciphersuites
6414 are subject to a denial of service attack.
6416 Thanks to Felix Gröbert and Ivan Fratric at Google for discovering
6417 this issue. [CVE-2014-3470][]
6419 *Felix Gröbert, Ivan Fratric, Steve Henson*
6421 * Harmonize version and its documentation. -f flag is used to display
6424 *mancha <mancha1@zoho.com>*
6426 * Fix eckey_priv_encode so it immediately returns an error upon a failure
6427 in i2d_ECPrivateKey.
6429 *mancha <mancha1@zoho.com>*
6431 * Fix some double frees. These are not thought to be exploitable.
6433 *mancha <mancha1@zoho.com>*
6435 ### Changes between 1.0.1f and 1.0.1g [7 Apr 2014]
6437 * A missing bounds check in the handling of the TLS heartbeat extension
6438 can be used to reveal up to 64k of memory to a connected client or
6441 Thanks for Neel Mehta of Google Security for discovering this bug and to
6442 Adam Langley <agl@chromium.org> and Bodo Moeller <bmoeller@acm.org> for
6443 preparing the fix [CVE-2014-0160][]
6445 *Adam Langley, Bodo Moeller*
6447 * Fix for the attack described in the paper "Recovering OpenSSL
6448 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
6449 by Yuval Yarom and Naomi Benger. Details can be obtained from:
6450 <http://eprint.iacr.org/2014/140>
6452 Thanks to Yuval Yarom and Naomi Benger for discovering this
6453 flaw and to Yuval Yarom for supplying a fix [CVE-2014-0076][]
6455 *Yuval Yarom and Naomi Benger*
6457 * TLS pad extension: draft-agl-tls-padding-03
6459 Workaround for the "TLS hang bug" (see FAQ and PR#2771): if the
6460 TLS client Hello record length value would otherwise be > 255 and
6461 less that 512 pad with a dummy extension containing zeroes so it
6462 is at least 512 bytes long.
6464 *Adam Langley, Steve Henson*
6466 ### Changes between 1.0.1e and 1.0.1f [6 Jan 2014]
6468 * Fix for TLS record tampering bug. A carefully crafted invalid
6469 handshake could crash OpenSSL with a NULL pointer exception.
6470 Thanks to Anton Johansson for reporting this issues.
6473 * Keep original DTLS digest and encryption contexts in retransmission
6474 structures so we can use the previous session parameters if they need
6475 to be resent. [CVE-2013-6450][]
6479 * Add option SSL_OP_SAFARI_ECDHE_ECDSA_BUG (part of SSL_OP_ALL) which
6480 avoids preferring ECDHE-ECDSA ciphers when the client appears to be
6481 Safari on OS X. Safari on OS X 10.8..10.8.3 advertises support for
6482 several ECDHE-ECDSA ciphers, but fails to negotiate them. The bug
6483 is fixed in OS X 10.8.4, but Apple have ruled out both hot fixing
6484 10.8..10.8.3 and forcing users to upgrade to 10.8.4 or newer.
6486 *Rob Stradling, Adam Langley*
6488 ### Changes between 1.0.1d and 1.0.1e [11 Feb 2013]
6490 * Correct fix for CVE-2013-0169. The original didn't work on AES-NI
6491 supporting platforms or when small records were transferred.
6493 *Andy Polyakov, Steve Henson*
6495 ### Changes between 1.0.1c and 1.0.1d [5 Feb 2013]
6497 * Make the decoding of SSLv3, TLS and DTLS CBC records constant time.
6499 This addresses the flaw in CBC record processing discovered by
6500 Nadhem Alfardan and Kenny Paterson. Details of this attack can be found
6501 at: <http://www.isg.rhul.ac.uk/tls/>
6503 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
6504 Security Group at Royal Holloway, University of London
6505 (www.isg.rhul.ac.uk) for discovering this flaw and Adam Langley and
6506 Emilia Käsper for the initial patch.
6509 *Emilia Käsper, Adam Langley, Ben Laurie, Andy Polyakov, Steve Henson*
6511 * Fix flaw in AESNI handling of TLS 1.2 and 1.1 records for CBC mode
6512 ciphersuites which can be exploited in a denial of service attack.
6513 Thanks go to and to Adam Langley <agl@chromium.org> for discovering
6514 and detecting this bug and to Wolfgang Ettlinger
6515 <wolfgang.ettlinger@gmail.com> for independently discovering this issue.
6520 * Return an error when checking OCSP signatures when key is NULL.
6521 This fixes a DoS attack. [CVE-2013-0166][]
6525 * Make openssl verify return errors.
6527 *Chris Palmer <palmer@google.com> and Ben Laurie*
6529 * Call OCSP Stapling callback after ciphersuite has been chosen, so
6530 the right response is stapled. Also change SSL_get_certificate()
6531 so it returns the certificate actually sent.
6532 See <http://rt.openssl.org/Ticket/Display.html?id=2836>.
6534 *Rob Stradling <rob.stradling@comodo.com>*
6536 * Fix possible deadlock when decoding public keys.
6540 * Don't use TLS 1.0 record version number in initial client hello
6545 ### Changes between 1.0.1b and 1.0.1c [10 May 2012]
6547 * Sanity check record length before skipping explicit IV in TLS
6548 1.2, 1.1 and DTLS to fix DoS attack.
6550 Thanks to Codenomicon for discovering this issue using Fuzz-o-Matic
6551 fuzzing as a service testing platform.
6556 * Initialise tkeylen properly when encrypting CMS messages.
6557 Thanks to Solar Designer of Openwall for reporting this issue.
6561 * In FIPS mode don't try to use composite ciphers as they are not
6566 ### Changes between 1.0.1a and 1.0.1b [26 Apr 2012]
6568 * OpenSSL 1.0.0 sets SSL_OP_ALL to 0x80000FFFL and OpenSSL 1.0.1 and
6569 1.0.1a set SSL_OP_NO_TLSv1_1 to 0x00000400L which would unfortunately
6570 mean any application compiled against OpenSSL 1.0.0 headers setting
6571 SSL_OP_ALL would also set SSL_OP_NO_TLSv1_1, unintentionally disabling
6572 TLS 1.1 also. Fix this by changing the value of SSL_OP_NO_TLSv1_1 to
6573 0x10000000L Any application which was previously compiled against
6574 OpenSSL 1.0.1 or 1.0.1a headers and which cares about SSL_OP_NO_TLSv1_1
6575 will need to be recompiled as a result. Letting be results in
6576 inability to disable specifically TLS 1.1 and in client context,
6577 in unlike event, limit maximum offered version to TLS 1.0 [see below].
6581 * In order to ensure interoperability SSL_OP_NO_protocolX does not
6582 disable just protocol X, but all protocols above X *if* there are
6583 protocols *below* X still enabled. In more practical terms it means
6584 that if application wants to disable TLS1.0 in favor of TLS1.1 and
6585 above, it's not sufficient to pass `SSL_OP_NO_TLSv1`, one has to pass
6586 `SSL_OP_NO_TLSv1|SSL_OP_NO_SSLv3|SSL_OP_NO_SSLv2`. This applies to
6591 ### Changes between 1.0.1 and 1.0.1a [19 Apr 2012]
6593 * Check for potentially exploitable overflows in asn1_d2i_read_bio
6594 BUF_mem_grow and BUF_mem_grow_clean. Refuse attempts to shrink buffer
6595 in CRYPTO_realloc_clean.
6597 Thanks to Tavis Ormandy, Google Security Team, for discovering this
6598 issue and to Adam Langley <agl@chromium.org> for fixing it.
6601 *Adam Langley (Google), Tavis Ormandy, Google Security Team*
6603 * Don't allow TLS 1.2 SHA-256 ciphersuites in TLS 1.0, 1.1 connections.
6607 * Workarounds for some broken servers that "hang" if a client hello
6608 record length exceeds 255 bytes.
6610 1. Do not use record version number > TLS 1.0 in initial client
6611 hello: some (but not all) hanging servers will now work.
6612 2. If we set OPENSSL_MAX_TLS1_2_CIPHER_LENGTH this will truncate
6613 the number of ciphers sent in the client hello. This should be
6614 set to an even number, such as 50, for example by passing:
6615 -DOPENSSL_MAX_TLS1_2_CIPHER_LENGTH=50 to config or Configure.
6616 Most broken servers should now work.
6617 3. If all else fails setting OPENSSL_NO_TLS1_2_CLIENT will disable
6618 TLS 1.2 client support entirely.
6622 * Fix SEGV in Vector Permutation AES module observed in OpenSSH.
6626 ### Changes between 1.0.0h and 1.0.1 [14 Mar 2012]
6628 * Add compatibility with old MDC2 signatures which use an ASN1 OCTET
6629 STRING form instead of a DigestInfo.
6633 * The format used for MDC2 RSA signatures is inconsistent between EVP
6634 and the RSA_sign/RSA_verify functions. This was made more apparent when
6635 OpenSSL used RSA_sign/RSA_verify for some RSA signatures in particular
6636 those which went through EVP_PKEY_METHOD in 1.0.0 and later. Detect
6637 the correct format in RSA_verify so both forms transparently work.
6641 * Some servers which support TLS 1.0 can choke if we initially indicate
6642 support for TLS 1.2 and later renegotiate using TLS 1.0 in the RSA
6643 encrypted premaster secret. As a workaround use the maximum permitted
6644 client version in client hello, this should keep such servers happy
6645 and still work with previous versions of OpenSSL.
6649 * Add support for TLS/DTLS heartbeats.
6651 *Robin Seggelmann <seggelmann@fh-muenster.de>*
6653 * Add support for SCTP.
6655 *Robin Seggelmann <seggelmann@fh-muenster.de>*
6657 * Improved PRNG seeding for VOS.
6659 *Paul Green <Paul.Green@stratus.com>*
6661 * Extensive assembler packs updates, most notably:
6663 - x86[_64]: AES-NI, PCLMULQDQ, RDRAND support;
6664 - x86[_64]: SSSE3 support (SHA1, vector-permutation AES);
6665 - x86_64: bit-sliced AES implementation;
6666 - ARM: NEON support, contemporary platforms optimizations;
6667 - s390x: z196 support;
6668 - `*`: GHASH and GF(2^m) multiplication implementations;
6672 * Make TLS-SRP code conformant with RFC 5054 API cleanup
6673 (removal of unnecessary code)
6675 *Peter Sylvester <peter.sylvester@edelweb.fr>*
6677 * Add TLS key material exporter from RFC 5705.
6681 * Add DTLS-SRTP negotiation from RFC 5764.
6685 * Add Next Protocol Negotiation,
6686 <http://tools.ietf.org/html/draft-agl-tls-nextprotoneg-00>. Can be
6687 disabled with a no-npn flag to config or Configure. Code donated
6690 *Adam Langley <agl@google.com> and Ben Laurie*
6692 * Add optional 64-bit optimized implementations of elliptic curves NIST-P224,
6693 NIST-P256, NIST-P521, with constant-time single point multiplication on
6694 typical inputs. Compiler support for the nonstandard type `__uint128_t` is
6695 required to use this (present in gcc 4.4 and later, for 64-bit builds).
6696 Code made available under Apache License version 2.0.
6698 Specify "enable-ec_nistp_64_gcc_128" on the Configure (or config) command
6699 line to include this in your build of OpenSSL, and run "make depend" (or
6700 "make update"). This enables the following EC_METHODs:
6702 EC_GFp_nistp224_method()
6703 EC_GFp_nistp256_method()
6704 EC_GFp_nistp521_method()
6706 EC_GROUP_new_by_curve_name() will automatically use these (while
6707 EC_GROUP_new_curve_GFp() currently prefers the more flexible
6710 *Emilia Käsper, Adam Langley, Bodo Moeller (Google)*
6712 * Use type ossl_ssize_t instead of ssize_t which isn't available on
6713 all platforms. Move ssize_t definition from e_os.h to the public
6714 header file e_os2.h as it now appears in public header file cms.h
6718 * New -sigopt option to the ca, req and x509 utilities. Additional
6719 signature parameters can be passed using this option and in
6724 * Add RSA PSS signing function. This will generate and set the
6725 appropriate AlgorithmIdentifiers for PSS based on those in the
6726 corresponding EVP_MD_CTX structure. No application support yet.
6730 * Support for companion algorithm specific ASN1 signing routines.
6731 New function ASN1_item_sign_ctx() signs a pre-initialised
6732 EVP_MD_CTX structure and sets AlgorithmIdentifiers based on
6733 the appropriate parameters.
6737 * Add new algorithm specific ASN1 verification initialisation function
6738 to EVP_PKEY_ASN1_METHOD: this is not in EVP_PKEY_METHOD since the ASN1
6739 handling will be the same no matter what EVP_PKEY_METHOD is used.
6740 Add a PSS handler to support verification of PSS signatures: checked
6741 against a number of sample certificates.
6745 * Add signature printing for PSS. Add PSS OIDs.
6747 *Steve Henson, Martin Kaiser <lists@kaiser.cx>*
6749 * Add algorithm specific signature printing. An individual ASN1 method
6750 can now print out signatures instead of the standard hex dump.
6752 More complex signatures (e.g. PSS) can print out more meaningful
6753 information. Include DSA version that prints out the signature
6758 * Password based recipient info support for CMS library: implementing
6763 * Split password based encryption into PBES2 and PBKDF2 functions. This
6764 neatly separates the code into cipher and PBE sections and is required
6765 for some algorithms that split PBES2 into separate pieces (such as
6766 password based CMS).
6770 * Session-handling fixes:
6771 - Fix handling of connections that are resuming with a session ID,
6772 but also support Session Tickets.
6773 - Fix a bug that suppressed issuing of a new ticket if the client
6774 presented a ticket with an expired session.
6775 - Try to set the ticket lifetime hint to something reasonable.
6776 - Make tickets shorter by excluding irrelevant information.
6777 - On the client side, don't ignore renewed tickets.
6779 *Adam Langley, Bodo Moeller (Google)*
6781 * Fix PSK session representation.
6785 * Add RC4-MD5 and AESNI-SHA1 "stitched" implementations.
6787 This work was sponsored by Intel.
6791 * Add GCM support to TLS library. Some custom code is needed to split
6792 the IV between the fixed (from PRF) and explicit (from TLS record)
6793 portions. This adds all GCM ciphersuites supported by RFC5288 and
6794 RFC5289. Generalise some `AES*` cipherstrings to include GCM and
6795 add a special AESGCM string for GCM only.
6799 * Expand range of ctrls for AES GCM. Permit setting invocation
6800 field on decrypt and retrieval of invocation field only on encrypt.
6804 * Add HMAC ECC ciphersuites from RFC5289. Include SHA384 PRF support.
6805 As required by RFC5289 these ciphersuites cannot be used if for
6806 versions of TLS earlier than 1.2.
6810 * For FIPS capable OpenSSL interpret a NULL default public key method
6811 as unset and return the appropriate default but do *not* set the default.
6812 This means we can return the appropriate method in applications that
6813 switch between FIPS and non-FIPS modes.
6817 * Redirect HMAC and CMAC operations to FIPS module in FIPS mode. If an
6818 ENGINE is used then we cannot handle that in the FIPS module so we
6819 keep original code iff non-FIPS operations are allowed.
6823 * Add -attime option to openssl utilities.
6825 *Peter Eckersley <pde@eff.org>, Ben Laurie and Steve Henson*
6827 * Redirect DSA and DH operations to FIPS module in FIPS mode.
6831 * Redirect ECDSA and ECDH operations to FIPS module in FIPS mode. Also use
6832 FIPS EC methods unconditionally for now.
6836 * New build option no-ec2m to disable characteristic 2 code.
6840 * Backport libcrypto audit of return value checking from 1.1.0-dev; not
6841 all cases can be covered as some introduce binary incompatibilities.
6845 * Redirect RSA operations to FIPS module including keygen,
6846 encrypt, decrypt, sign and verify. Block use of non FIPS RSA methods.
6850 * Add similar low level API blocking to ciphers.
6854 * Low level digest APIs are not approved in FIPS mode: any attempt
6855 to use these will cause a fatal error. Applications that *really* want
6856 to use them can use the `private_*` version instead.
6860 * Redirect cipher operations to FIPS module for FIPS builds.
6864 * Redirect digest operations to FIPS module for FIPS builds.
6868 * Update build system to add "fips" flag which will link in fipscanister.o
6869 for static and shared library builds embedding a signature if needed.
6873 * Output TLS supported curves in preference order instead of numerical
6874 order. This is currently hardcoded for the highest order curves first.
6875 This should be configurable so applications can judge speed vs strength.
6879 * Add TLS v1.2 server support for client authentication.
6883 * Add support for FIPS mode in ssl library: disable SSLv3, non-FIPS ciphers
6888 * Functions FIPS_mode_set() and FIPS_mode() which call the underlying
6889 FIPS modules versions.
6893 * Add TLS v1.2 client side support for client authentication. Keep cache
6894 of handshake records longer as we don't know the hash algorithm to use
6895 until after the certificate request message is received.
6899 * Initial TLS v1.2 client support. Add a default signature algorithms
6900 extension including all the algorithms we support. Parse new signature
6901 format in client key exchange. Relax some ECC signing restrictions for
6902 TLS v1.2 as indicated in RFC5246.
6906 * Add server support for TLS v1.2 signature algorithms extension. Switch
6907 to new signature format when needed using client digest preference.
6908 All server ciphersuites should now work correctly in TLS v1.2. No client
6909 support yet and no support for client certificates.
6913 * Initial TLS v1.2 support. Add new SHA256 digest to ssl code, switch
6914 to SHA256 for PRF when using TLS v1.2 and later. Add new SHA256 based
6915 ciphersuites. At present only RSA key exchange ciphersuites work with
6916 TLS v1.2. Add new option for TLS v1.2 replacing the old and obsolete
6917 SSL_OP_PKCS1_CHECK flags with SSL_OP_NO_TLSv1_2. New TLSv1.2 methods
6918 and version checking.
6922 * New option OPENSSL_NO_SSL_INTERN. If an application can be compiled
6923 with this defined it will not be affected by any changes to ssl internal
6924 structures. Add several utility functions to allow openssl application
6925 to work with OPENSSL_NO_SSL_INTERN defined.
6929 * A long standing patch to add support for SRP from EdelWeb (Peter
6930 Sylvester and Christophe Renou) was integrated.
6931 *Christophe Renou <christophe.renou@edelweb.fr>, Peter Sylvester
6932 <peter.sylvester@edelweb.fr>, Tom Wu <tjw@cs.stanford.edu>, and
6935 * Add functions to copy EVP_PKEY_METHOD and retrieve flags and id.
6939 * Permit abbreviated handshakes when renegotiating using the function
6940 SSL_renegotiate_abbreviated().
6942 *Robin Seggelmann <seggelmann@fh-muenster.de>*
6944 * Add call to ENGINE_register_all_complete() to
6945 ENGINE_load_builtin_engines(), so some implementations get used
6946 automatically instead of needing explicit application support.
6950 * Add support for TLS key exporter as described in RFC5705.
6952 *Robin Seggelmann <seggelmann@fh-muenster.de>, Steve Henson*
6954 * Initial TLSv1.1 support. Since TLSv1.1 is very similar to TLS v1.0 only
6955 a few changes are required:
6957 Add SSL_OP_NO_TLSv1_1 flag.
6958 Add TLSv1_1 methods.
6959 Update version checking logic to handle version 1.1.
6960 Add explicit IV handling (ported from DTLS code).
6961 Add command line options to s_client/s_server.
6968 ### Changes between 1.0.0s and 1.0.0t [3 Dec 2015]
6970 * X509_ATTRIBUTE memory leak
6972 When presented with a malformed X509_ATTRIBUTE structure OpenSSL will leak
6973 memory. This structure is used by the PKCS#7 and CMS routines so any
6974 application which reads PKCS#7 or CMS data from untrusted sources is
6975 affected. SSL/TLS is not affected.
6977 This issue was reported to OpenSSL by Adam Langley (Google/BoringSSL) using
6983 * Race condition handling PSK identify hint
6985 If PSK identity hints are received by a multi-threaded client then
6986 the values are wrongly updated in the parent SSL_CTX structure. This can
6987 result in a race condition potentially leading to a double free of the
6993 ### Changes between 1.0.0r and 1.0.0s [11 Jun 2015]
6995 * Malformed ECParameters causes infinite loop
6997 When processing an ECParameters structure OpenSSL enters an infinite loop
6998 if the curve specified is over a specially malformed binary polynomial
7001 This can be used to perform denial of service against any
7002 system which processes public keys, certificate requests or
7003 certificates. This includes TLS clients and TLS servers with
7004 client authentication enabled.
7006 This issue was reported to OpenSSL by Joseph Barr-Pixton.
7011 * Exploitable out-of-bounds read in X509_cmp_time
7013 X509_cmp_time does not properly check the length of the ASN1_TIME
7014 string and can read a few bytes out of bounds. In addition,
7015 X509_cmp_time accepts an arbitrary number of fractional seconds in the
7018 An attacker can use this to craft malformed certificates and CRLs of
7019 various sizes and potentially cause a segmentation fault, resulting in
7020 a DoS on applications that verify certificates or CRLs. TLS clients
7021 that verify CRLs are affected. TLS clients and servers with client
7022 authentication enabled may be affected if they use custom verification
7025 This issue was reported to OpenSSL by Robert Swiecki (Google), and
7026 independently by Hanno Böck.
7031 * PKCS7 crash with missing EnvelopedContent
7033 The PKCS#7 parsing code does not handle missing inner EncryptedContent
7034 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs
7035 with missing content and trigger a NULL pointer dereference on parsing.
7037 Applications that decrypt PKCS#7 data or otherwise parse PKCS#7
7038 structures from untrusted sources are affected. OpenSSL clients and
7039 servers are not affected.
7041 This issue was reported to OpenSSL by Michal Zalewski (Google).
7046 * CMS verify infinite loop with unknown hash function
7048 When verifying a signedData message the CMS code can enter an infinite loop
7049 if presented with an unknown hash function OID. This can be used to perform
7050 denial of service against any system which verifies signedData messages using
7052 This issue was reported to OpenSSL by Johannes Bauer.
7057 * Race condition handling NewSessionTicket
7059 If a NewSessionTicket is received by a multi-threaded client when attempting to
7060 reuse a previous ticket then a race condition can occur potentially leading to
7061 a double free of the ticket data.
7066 ### Changes between 1.0.0q and 1.0.0r [19 Mar 2015]
7068 * Segmentation fault in ASN1_TYPE_cmp fix
7070 The function ASN1_TYPE_cmp will crash with an invalid read if an attempt is
7071 made to compare ASN.1 boolean types. Since ASN1_TYPE_cmp is used to check
7072 certificate signature algorithm consistency this can be used to crash any
7073 certificate verification operation and exploited in a DoS attack. Any
7074 application which performs certificate verification is vulnerable including
7075 OpenSSL clients and servers which enable client authentication.
7080 * ASN.1 structure reuse memory corruption fix
7082 Reusing a structure in ASN.1 parsing may allow an attacker to cause
7083 memory corruption via an invalid write. Such reuse is and has been
7084 strongly discouraged and is believed to be rare.
7086 Applications that parse structures containing CHOICE or ANY DEFINED BY
7087 components may be affected. Certificate parsing (d2i_X509 and related
7088 functions) are however not affected. OpenSSL clients and servers are
7094 * PKCS7 NULL pointer dereferences fix
7096 The PKCS#7 parsing code does not handle missing outer ContentInfo
7097 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs with
7098 missing content and trigger a NULL pointer dereference on parsing.
7100 Applications that verify PKCS#7 signatures, decrypt PKCS#7 data or
7101 otherwise parse PKCS#7 structures from untrusted sources are
7102 affected. OpenSSL clients and servers are not affected.
7104 This issue was reported to OpenSSL by Michal Zalewski (Google).
7109 * DoS via reachable assert in SSLv2 servers fix
7111 A malicious client can trigger an OPENSSL_assert (i.e., an abort) in
7112 servers that both support SSLv2 and enable export cipher suites by sending
7113 a specially crafted SSLv2 CLIENT-MASTER-KEY message.
7115 This issue was discovered by Sean Burford (Google) and Emilia Käsper
7116 (OpenSSL development team).
7121 * Use After Free following d2i_ECPrivatekey error fix
7123 A malformed EC private key file consumed via the d2i_ECPrivateKey function
7124 could cause a use after free condition. This, in turn, could cause a double
7125 free in several private key parsing functions (such as d2i_PrivateKey
7126 or EVP_PKCS82PKEY) and could lead to a DoS attack or memory corruption
7127 for applications that receive EC private keys from untrusted
7128 sources. This scenario is considered rare.
7130 This issue was discovered by the BoringSSL project and fixed in their
7136 * X509_to_X509_REQ NULL pointer deref fix
7138 The function X509_to_X509_REQ will crash with a NULL pointer dereference if
7139 the certificate key is invalid. This function is rarely used in practice.
7141 This issue was discovered by Brian Carpenter.
7146 * Removed the export ciphers from the DEFAULT ciphers
7150 ### Changes between 1.0.0p and 1.0.0q [15 Jan 2015]
7152 * Build fixes for the Windows and OpenVMS platforms
7154 *Matt Caswell and Richard Levitte*
7156 ### Changes between 1.0.0o and 1.0.0p [8 Jan 2015]
7158 * Fix DTLS segmentation fault in dtls1_get_record. A carefully crafted DTLS
7159 message can cause a segmentation fault in OpenSSL due to a NULL pointer
7160 dereference. This could lead to a Denial Of Service attack. Thanks to
7161 Markus Stenberg of Cisco Systems, Inc. for reporting this issue.
7166 * Fix DTLS memory leak in dtls1_buffer_record. A memory leak can occur in the
7167 dtls1_buffer_record function under certain conditions. In particular this
7168 could occur if an attacker sent repeated DTLS records with the same
7169 sequence number but for the next epoch. The memory leak could be exploited
7170 by an attacker in a Denial of Service attack through memory exhaustion.
7171 Thanks to Chris Mueller for reporting this issue.
7176 * Fix issue where no-ssl3 configuration sets method to NULL. When openssl is
7177 built with the no-ssl3 option and a SSL v3 ClientHello is received the ssl
7178 method would be set to NULL which could later result in a NULL pointer
7179 dereference. Thanks to Frank Schmirler for reporting this issue.
7184 * Abort handshake if server key exchange message is omitted for ephemeral
7187 Thanks to Karthikeyan Bhargavan of the PROSECCO team at INRIA for
7188 reporting this issue.
7193 * Remove non-export ephemeral RSA code on client and server. This code
7194 violated the TLS standard by allowing the use of temporary RSA keys in
7195 non-export ciphersuites and could be used by a server to effectively
7196 downgrade the RSA key length used to a value smaller than the server
7197 certificate. Thanks for Karthikeyan Bhargavan of the PROSECCO team at
7198 INRIA or reporting this issue.
7203 * Fixed issue where DH client certificates are accepted without verification.
7204 An OpenSSL server will accept a DH certificate for client authentication
7205 without the certificate verify message. This effectively allows a client to
7206 authenticate without the use of a private key. This only affects servers
7207 which trust a client certificate authority which issues certificates
7208 containing DH keys: these are extremely rare and hardly ever encountered.
7209 Thanks for Karthikeyan Bhargavan of the PROSECCO team at INRIA or reporting
7215 * Correct Bignum squaring. Bignum squaring (BN_sqr) may produce incorrect
7216 results on some platforms, including x86_64. This bug occurs at random
7217 with a very low probability, and is not known to be exploitable in any
7218 way, though its exact impact is difficult to determine. Thanks to Pieter
7219 Wuille (Blockstream) who reported this issue and also suggested an initial
7220 fix. Further analysis was conducted by the OpenSSL development team and
7221 Adam Langley of Google. The final fix was developed by Andy Polyakov of
7222 the OpenSSL core team.
7227 * Fix various certificate fingerprint issues.
7229 By using non-DER or invalid encodings outside the signed portion of a
7230 certificate the fingerprint can be changed without breaking the signature.
7231 Although no details of the signed portion of the certificate can be changed
7232 this can cause problems with some applications: e.g. those using the
7233 certificate fingerprint for blacklists.
7235 1. Reject signatures with non zero unused bits.
7237 If the BIT STRING containing the signature has non zero unused bits reject
7238 the signature. All current signature algorithms require zero unused bits.
7240 2. Check certificate algorithm consistency.
7242 Check the AlgorithmIdentifier inside TBS matches the one in the
7243 certificate signature. NB: this will result in signature failure
7244 errors for some broken certificates.
7246 Thanks to Konrad Kraszewski from Google for reporting this issue.
7248 3. Check DSA/ECDSA signatures use DER.
7250 Reencode DSA/ECDSA signatures and compare with the original received
7251 signature. Return an error if there is a mismatch.
7253 This will reject various cases including garbage after signature
7254 (thanks to Antti Karjalainen and Tuomo Untinen from the Codenomicon CROSS
7255 program for discovering this case) and use of BER or invalid ASN.1 INTEGERs
7256 (negative or with leading zeroes).
7258 Further analysis was conducted and fixes were developed by Stephen Henson
7259 of the OpenSSL core team.
7265 ### Changes between 1.0.0n and 1.0.0o [15 Oct 2014]
7267 * Session Ticket Memory Leak.
7269 When an OpenSSL SSL/TLS/DTLS server receives a session ticket the
7270 integrity of that ticket is first verified. In the event of a session
7271 ticket integrity check failing, OpenSSL will fail to free memory
7272 causing a memory leak. By sending a large number of invalid session
7273 tickets an attacker could exploit this issue in a Denial Of Service
7279 * Build option no-ssl3 is incomplete.
7281 When OpenSSL is configured with "no-ssl3" as a build option, servers
7282 could accept and complete a SSL 3.0 handshake, and clients could be
7283 configured to send them.
7286 *Akamai and the OpenSSL team*
7288 * Add support for TLS_FALLBACK_SCSV.
7289 Client applications doing fallback retries should call
7290 SSL_set_mode(s, SSL_MODE_SEND_FALLBACK_SCSV).
7293 *Adam Langley, Bodo Moeller*
7295 * Add additional DigestInfo checks.
7297 Reencode DigestInto in DER and check against the original when
7298 verifying RSA signature: this will reject any improperly encoded
7299 DigestInfo structures.
7301 Note: this is a precautionary measure and no attacks are currently known.
7305 ### Changes between 1.0.0m and 1.0.0n [6 Aug 2014]
7307 * OpenSSL DTLS clients enabling anonymous (EC)DH ciphersuites are subject
7308 to a denial of service attack. A malicious server can crash the client
7309 with a null pointer dereference (read) by specifying an anonymous (EC)DH
7310 ciphersuite and sending carefully crafted handshake messages.
7312 Thanks to Felix Gröbert (Google) for discovering and researching this
7318 * By sending carefully crafted DTLS packets an attacker could cause openssl
7319 to leak memory. This can be exploited through a Denial of Service attack.
7320 Thanks to Adam Langley for discovering and researching this issue.
7325 * An attacker can force openssl to consume large amounts of memory whilst
7326 processing DTLS handshake messages. This can be exploited through a
7327 Denial of Service attack.
7328 Thanks to Adam Langley for discovering and researching this issue.
7333 * An attacker can force an error condition which causes openssl to crash
7334 whilst processing DTLS packets due to memory being freed twice. This
7335 can be exploited through a Denial of Service attack.
7336 Thanks to Adam Langley and Wan-Teh Chang for discovering and researching
7342 * If a multithreaded client connects to a malicious server using a resumed
7343 session and the server sends an ec point format extension it could write
7344 up to 255 bytes to freed memory.
7346 Thanks to Gabor Tyukasz (LogMeIn Inc) for discovering and researching this
7352 * A flaw in OBJ_obj2txt may cause pretty printing functions such as
7353 X509_name_oneline, X509_name_print_ex et al. to leak some information
7354 from the stack. Applications may be affected if they echo pretty printing
7355 output to the attacker.
7357 Thanks to Ivan Fratric (Google) for discovering this issue.
7360 *Emilia Käsper, and Steve Henson*
7362 * Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
7363 for corner cases. (Certain input points at infinity could lead to
7364 bogus results, with non-infinity inputs mapped to infinity too.)
7368 ### Changes between 1.0.0l and 1.0.0m [5 Jun 2014]
7370 * Fix for SSL/TLS MITM flaw. An attacker using a carefully crafted
7371 handshake can force the use of weak keying material in OpenSSL
7372 SSL/TLS clients and servers.
7374 Thanks to KIKUCHI Masashi (Lepidum Co. Ltd.) for discovering and
7375 researching this issue. [CVE-2014-0224][]
7377 *KIKUCHI Masashi, Steve Henson*
7379 * Fix DTLS recursion flaw. By sending an invalid DTLS handshake to an
7380 OpenSSL DTLS client the code can be made to recurse eventually crashing
7383 Thanks to Imre Rad (Search-Lab Ltd.) for discovering this issue.
7386 *Imre Rad, Steve Henson*
7388 * Fix DTLS invalid fragment vulnerability. A buffer overrun attack can
7389 be triggered by sending invalid DTLS fragments to an OpenSSL DTLS
7390 client or server. This is potentially exploitable to run arbitrary
7391 code on a vulnerable client or server.
7393 Thanks to Jüri Aedla for reporting this issue. [CVE-2014-0195][]
7395 *Jüri Aedla, Steve Henson*
7397 * Fix bug in TLS code where clients enable anonymous ECDH ciphersuites
7398 are subject to a denial of service attack.
7400 Thanks to Felix Gröbert and Ivan Fratric at Google for discovering
7401 this issue. [CVE-2014-3470][]
7403 *Felix Gröbert, Ivan Fratric, Steve Henson*
7405 * Harmonize version and its documentation. -f flag is used to display
7408 *mancha <mancha1@zoho.com>*
7410 * Fix eckey_priv_encode so it immediately returns an error upon a failure
7411 in i2d_ECPrivateKey.
7413 *mancha <mancha1@zoho.com>*
7415 * Fix some double frees. These are not thought to be exploitable.
7417 *mancha <mancha1@zoho.com>*
7419 * Fix for the attack described in the paper "Recovering OpenSSL
7420 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
7421 by Yuval Yarom and Naomi Benger. Details can be obtained from:
7422 <http://eprint.iacr.org/2014/140>
7424 Thanks to Yuval Yarom and Naomi Benger for discovering this
7425 flaw and to Yuval Yarom for supplying a fix [CVE-2014-0076][]
7427 *Yuval Yarom and Naomi Benger*
7429 ### Changes between 1.0.0k and 1.0.0l [6 Jan 2014]
7431 * Keep original DTLS digest and encryption contexts in retransmission
7432 structures so we can use the previous session parameters if they need
7433 to be resent. [CVE-2013-6450][]
7437 * Add option SSL_OP_SAFARI_ECDHE_ECDSA_BUG (part of SSL_OP_ALL) which
7438 avoids preferring ECDHE-ECDSA ciphers when the client appears to be
7439 Safari on OS X. Safari on OS X 10.8..10.8.3 advertises support for
7440 several ECDHE-ECDSA ciphers, but fails to negotiate them. The bug
7441 is fixed in OS X 10.8.4, but Apple have ruled out both hot fixing
7442 10.8..10.8.3 and forcing users to upgrade to 10.8.4 or newer.
7444 *Rob Stradling, Adam Langley*
7446 ### Changes between 1.0.0j and 1.0.0k [5 Feb 2013]
7448 * Make the decoding of SSLv3, TLS and DTLS CBC records constant time.
7450 This addresses the flaw in CBC record processing discovered by
7451 Nadhem Alfardan and Kenny Paterson. Details of this attack can be found
7452 at: <http://www.isg.rhul.ac.uk/tls/>
7454 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
7455 Security Group at Royal Holloway, University of London
7456 (www.isg.rhul.ac.uk) for discovering this flaw and Adam Langley and
7457 Emilia Käsper for the initial patch.
7460 *Emilia Käsper, Adam Langley, Ben Laurie, Andy Polyakov, Steve Henson*
7462 * Return an error when checking OCSP signatures when key is NULL.
7463 This fixes a DoS attack. [CVE-2013-0166][]
7467 * Call OCSP Stapling callback after ciphersuite has been chosen, so
7468 the right response is stapled. Also change SSL_get_certificate()
7469 so it returns the certificate actually sent.
7470 See <http://rt.openssl.org/Ticket/Display.html?id=2836>.
7471 (This is a backport)
7473 *Rob Stradling <rob.stradling@comodo.com>*
7475 * Fix possible deadlock when decoding public keys.
7479 ### Changes between 1.0.0i and 1.0.0j [10 May 2012]
7481 [NB: OpenSSL 1.0.0i and later 1.0.0 patch levels were released after
7484 * Sanity check record length before skipping explicit IV in DTLS
7487 Thanks to Codenomicon for discovering this issue using Fuzz-o-Matic
7488 fuzzing as a service testing platform.
7493 * Initialise tkeylen properly when encrypting CMS messages.
7494 Thanks to Solar Designer of Openwall for reporting this issue.
7498 ### Changes between 1.0.0h and 1.0.0i [19 Apr 2012]
7500 * Check for potentially exploitable overflows in asn1_d2i_read_bio
7501 BUF_mem_grow and BUF_mem_grow_clean. Refuse attempts to shrink buffer
7502 in CRYPTO_realloc_clean.
7504 Thanks to Tavis Ormandy, Google Security Team, for discovering this
7505 issue and to Adam Langley <agl@chromium.org> for fixing it.
7508 *Adam Langley (Google), Tavis Ormandy, Google Security Team*
7510 ### Changes between 1.0.0g and 1.0.0h [12 Mar 2012]
7512 * Fix MMA (Bleichenbacher's attack on PKCS #1 v1.5 RSA padding) weakness
7513 in CMS and PKCS7 code. When RSA decryption fails use a random key for
7514 content decryption and always return the same error. Note: this attack
7515 needs on average 2^20 messages so it only affects automated senders. The
7516 old behaviour can be re-enabled in the CMS code by setting the
7517 CMS_DEBUG_DECRYPT flag: this is useful for debugging and testing where
7518 an MMA defence is not necessary.
7519 Thanks to Ivan Nestlerode <inestlerode@us.ibm.com> for discovering
7520 this issue. [CVE-2012-0884][]
7524 * Fix CVE-2011-4619: make sure we really are receiving a
7525 client hello before rejecting multiple SGC restarts. Thanks to
7526 Ivan Nestlerode <inestlerode@us.ibm.com> for discovering this bug.
7530 ### Changes between 1.0.0f and 1.0.0g [18 Jan 2012]
7532 * Fix for DTLS DoS issue introduced by fix for CVE-2011-4109.
7533 Thanks to Antonio Martin, Enterprise Secure Access Research and
7534 Development, Cisco Systems, Inc. for discovering this bug and
7535 preparing a fix. [CVE-2012-0050][]
7539 ### Changes between 1.0.0e and 1.0.0f [4 Jan 2012]
7541 * Nadhem Alfardan and Kenny Paterson have discovered an extension
7542 of the Vaudenay padding oracle attack on CBC mode encryption
7543 which enables an efficient plaintext recovery attack against
7544 the OpenSSL implementation of DTLS. Their attack exploits timing
7545 differences arising during decryption processing. A research
7546 paper describing this attack can be found at:
7547 <http://www.isg.rhul.ac.uk/~kp/dtls.pdf>
7548 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
7549 Security Group at Royal Holloway, University of London
7550 (www.isg.rhul.ac.uk) for discovering this flaw and to Robin Seggelmann
7551 <seggelmann@fh-muenster.de> and Michael Tuexen <tuexen@fh-muenster.de>
7552 for preparing the fix. [CVE-2011-4108][]
7554 *Robin Seggelmann, Michael Tuexen*
7556 * Clear bytes used for block padding of SSL 3.0 records.
7559 *Adam Langley (Google)*
7561 * Only allow one SGC handshake restart for SSL/TLS. Thanks to George
7562 Kadianakis <desnacked@gmail.com> for discovering this issue and
7563 Adam Langley for preparing the fix. [CVE-2011-4619][]
7565 *Adam Langley (Google)*
7567 * Check parameters are not NULL in GOST ENGINE. [CVE-2012-0027][]
7569 *Andrey Kulikov <amdeich@gmail.com>*
7571 * Prevent malformed RFC3779 data triggering an assertion failure.
7572 Thanks to Andrew Chi, BBN Technologies, for discovering the flaw
7573 and Rob Austein <sra@hactrn.net> for fixing it. [CVE-2011-4577][]
7575 *Rob Austein <sra@hactrn.net>*
7577 * Improved PRNG seeding for VOS.
7579 *Paul Green <Paul.Green@stratus.com>*
7581 * Fix ssl_ciph.c set-up race.
7583 *Adam Langley (Google)*
7585 * Fix spurious failures in ecdsatest.c.
7587 *Emilia Käsper (Google)*
7589 * Fix the BIO_f_buffer() implementation (which was mixing different
7590 interpretations of the `..._len` fields).
7592 *Adam Langley (Google)*
7594 * Fix handling of BN_BLINDING: now BN_BLINDING_invert_ex (rather than
7595 BN_BLINDING_invert_ex) calls BN_BLINDING_update, ensuring that concurrent
7596 threads won't reuse the same blinding coefficients.
7598 This also avoids the need to obtain the CRYPTO_LOCK_RSA_BLINDING
7599 lock to call BN_BLINDING_invert_ex, and avoids one use of
7600 BN_BLINDING_update for each BN_BLINDING structure (previously,
7601 the last update always remained unused).
7603 *Emilia Käsper (Google)*
7605 * In ssl3_clear, preserve s3->init_extra along with s3->rbuf.
7607 *Bob Buckholz (Google)*
7609 ### Changes between 1.0.0d and 1.0.0e [6 Sep 2011]
7611 * Fix bug where CRLs with nextUpdate in the past are sometimes accepted
7612 by initialising X509_STORE_CTX properly. [CVE-2011-3207][]
7614 *Kaspar Brand <ossl@velox.ch>*
7616 * Fix SSL memory handling for (EC)DH ciphersuites, in particular
7617 for multi-threaded use of ECDH. [CVE-2011-3210][]
7619 *Adam Langley (Google)*
7621 * Fix x509_name_ex_d2i memory leak on bad inputs.
7625 * Remove hard coded ecdsaWithSHA1 signature tests in ssl code and check
7626 signature public key algorithm by using OID xref utilities instead.
7627 Before this you could only use some ECC ciphersuites with SHA1 only.
7631 * Add protection against ECDSA timing attacks as mentioned in the paper
7632 by Billy Bob Brumley and Nicola Tuveri, see:
7633 <http://eprint.iacr.org/2011/232.pdf>
7635 *Billy Bob Brumley and Nicola Tuveri*
7637 ### Changes between 1.0.0c and 1.0.0d [8 Feb 2011]
7639 * Fix parsing of OCSP stapling ClientHello extension. CVE-2011-0014
7641 *Neel Mehta, Adam Langley, Bodo Moeller (Google)*
7643 * Fix bug in string printing code: if *any* escaping is enabled we must
7644 escape the escape character (backslash) or the resulting string is
7649 ### Changes between 1.0.0b and 1.0.0c [2 Dec 2010]
7651 * Disable code workaround for ancient and obsolete Netscape browsers
7652 and servers: an attacker can use it in a ciphersuite downgrade attack.
7653 Thanks to Martin Rex for discovering this bug. CVE-2010-4180
7657 * Fixed J-PAKE implementation error, originally discovered by
7658 Sebastien Martini, further info and confirmation from Stefan
7659 Arentz and Feng Hao. Note that this fix is a security fix. CVE-2010-4252
7663 ### Changes between 1.0.0a and 1.0.0b [16 Nov 2010]
7665 * Fix extension code to avoid race conditions which can result in a buffer
7666 overrun vulnerability: resumed sessions must not be modified as they can
7667 be shared by multiple threads. CVE-2010-3864
7671 * Fix WIN32 build system to correctly link an ENGINE directory into
7676 ### Changes between 1.0.0 and 1.0.0a [01 Jun 2010]
7678 * Check return value of int_rsa_verify in pkey_rsa_verifyrecover
7681 *Steve Henson, Peter-Michael Hager <hager@dortmund.net>*
7683 ### Changes between 0.9.8n and 1.0.0 [29 Mar 2010]
7685 * Add "missing" function EVP_CIPHER_CTX_copy(). This copies a cipher
7686 context. The operation can be customised via the ctrl mechanism in
7687 case ENGINEs want to include additional functionality.
7691 * Tolerate yet another broken PKCS#8 key format: private key value negative.
7695 * Add new -subject_hash_old and -issuer_hash_old options to x509 utility to
7696 output hashes compatible with older versions of OpenSSL.
7698 *Willy Weisz <weisz@vcpc.univie.ac.at>*
7700 * Fix compression algorithm handling: if resuming a session use the
7701 compression algorithm of the resumed session instead of determining
7702 it from client hello again. Don't allow server to change algorithm.
7706 * Add load_crls() function to apps tidying load_certs() too. Add option
7707 to verify utility to allow additional CRLs to be included.
7711 * Update OCSP request code to permit adding custom headers to the request:
7712 some responders need this.
7716 * The function EVP_PKEY_sign() returns <=0 on error: check return code
7719 *Julia Lawall <julia@diku.dk>*
7721 * Update verify callback code in apps/s_cb.c and apps/verify.c, it
7722 needlessly dereferenced structures, used obsolete functions and
7723 didn't handle all updated verify codes correctly.
7727 * Disable MD2 in the default configuration.
7731 * In BIO_pop() and BIO_push() use the ctrl argument (which was NULL) to
7732 indicate the initial BIO being pushed or popped. This makes it possible
7733 to determine whether the BIO is the one explicitly called or as a result
7734 of the ctrl being passed down the chain. Fix BIO_pop() and SSL BIOs so
7735 it handles reference counts correctly and doesn't zero out the I/O bio
7736 when it is not being explicitly popped. WARNING: applications which
7737 included workarounds for the old buggy behaviour will need to be modified
7738 or they could free up already freed BIOs.
7742 * Extend the uni2asc/asc2uni => OPENSSL_uni2asc/OPENSSL_asc2uni
7743 renaming to all platforms (within the 0.9.8 branch, this was
7744 done conditionally on Netware platforms to avoid a name clash).
7746 *Guenter <lists@gknw.net>*
7748 * Add ECDHE and PSK support to DTLS.
7750 *Michael Tuexen <tuexen@fh-muenster.de>*
7752 * Add CHECKED_STACK_OF macro to safestack.h, otherwise safestack can't
7757 * Add "missing" function EVP_MD_flags() (without this the only way to
7758 retrieve a digest flags is by accessing the structure directly. Update
7759 `EVP_MD_do_all*()` and `EVP_CIPHER_do_all*()` to include the name a digest
7760 or cipher is registered as in the "from" argument. Print out all
7761 registered digests in the dgst usage message instead of manually
7762 attempting to work them out.
7766 * If no SSLv2 ciphers are used don't use an SSLv2 compatible client hello:
7767 this allows the use of compression and extensions. Change default cipher
7768 string to remove SSLv2 ciphersuites. This effectively avoids ancient SSLv2
7769 by default unless an application cipher string requests it.
7773 * Alter match criteria in PKCS12_parse(). It used to try to use local
7774 key ids to find matching certificates and keys but some PKCS#12 files
7775 don't follow the (somewhat unwritten) rules and this strategy fails.
7776 Now just gather all certificates together and the first private key
7777 then look for the first certificate that matches the key.
7781 * Support use of registered digest and cipher names for dgst and cipher
7782 commands instead of having to add each one as a special case. So now
7789 openssl dgst -sha256 foo
7791 and this works for ENGINE based algorithms too.
7795 * Update Gost ENGINE to support parameter files.
7797 *Victor B. Wagner <vitus@cryptocom.ru>*
7799 * Support GeneralizedTime in ca utility.
7801 *Oliver Martin <oliver@volatilevoid.net>, Steve Henson*
7803 * Enhance the hash format used for certificate directory links. The new
7804 form uses the canonical encoding (meaning equivalent names will work
7805 even if they aren't identical) and uses SHA1 instead of MD5. This form
7806 is incompatible with the older format and as a result c_rehash should
7807 be used to rebuild symbolic links.
7811 * Make PKCS#8 the default write format for private keys, replacing the
7812 traditional format. This form is standardised, more secure and doesn't
7813 include an implicit MD5 dependency.
7817 * Add a $gcc_devteam_warn option to Configure. The idea is that any code
7818 committed to OpenSSL should pass this lot as a minimum.
7822 * Add session ticket override functionality for use by EAP-FAST.
7824 *Jouni Malinen <j@w1.fi>*
7826 * Modify HMAC functions to return a value. Since these can be implemented
7827 in an ENGINE errors can occur.
7831 * Type-checked OBJ_bsearch_ex.
7835 * Type-checked OBJ_bsearch. Also some constification necessitated
7836 by type-checking. Still to come: TXT_DB, bsearch(?),
7837 OBJ_bsearch_ex, qsort, CRYPTO_EX_DATA, ASN1_VALUE, ASN1_STRING,
7842 * New function OPENSSL_gmtime_adj() to add a specific number of days and
7843 seconds to a tm structure directly, instead of going through OS
7844 specific date routines. This avoids any issues with OS routines such
7845 as the year 2038 bug. New `*_adj()` functions for ASN1 time structures
7846 and X509_time_adj_ex() to cover the extended range. The existing
7847 X509_time_adj() is still usable and will no longer have any date issues.
7851 * Delta CRL support. New use deltas option which will attempt to locate
7852 and search any appropriate delta CRLs available.
7854 This work was sponsored by Google.
7858 * Support for CRLs partitioned by reason code. Reorganise CRL processing
7859 code and add additional score elements. Validate alternate CRL paths
7860 as part of the CRL checking and indicate a new error "CRL path validation
7861 error" in this case. Applications wanting additional details can use
7862 the verify callback and check the new "parent" field. If this is not
7863 NULL CRL path validation is taking place. Existing applications won't
7864 see this because it requires extended CRL support which is off by
7867 This work was sponsored by Google.
7871 * Support for freshest CRL extension.
7873 This work was sponsored by Google.
7877 * Initial indirect CRL support. Currently only supported in the CRLs
7878 passed directly and not via lookup. Process certificate issuer
7879 CRL entry extension and lookup CRL entries by bother issuer name
7880 and serial number. Check and process CRL issuer entry in IDP extension.
7882 This work was sponsored by Google.
7886 * Add support for distinct certificate and CRL paths. The CRL issuer
7887 certificate is validated separately in this case. Only enabled if
7888 an extended CRL support flag is set: this flag will enable additional
7889 CRL functionality in future.
7891 This work was sponsored by Google.
7895 * Add support for policy mappings extension.
7897 This work was sponsored by Google.
7901 * Fixes to pathlength constraint, self issued certificate handling,
7902 policy processing to align with RFC3280 and PKITS tests.
7904 This work was sponsored by Google.
7908 * Support for name constraints certificate extension. DN, email, DNS
7909 and URI types are currently supported.
7911 This work was sponsored by Google.
7915 * To cater for systems that provide a pointer-based thread ID rather
7916 than numeric, deprecate the current numeric thread ID mechanism and
7917 replace it with a structure and associated callback type. This
7918 mechanism allows a numeric "hash" to be extracted from a thread ID in
7919 either case, and on platforms where pointers are larger than 'long',
7920 mixing is done to help ensure the numeric 'hash' is usable even if it
7921 can't be guaranteed unique. The default mechanism is to use "&errno"
7922 as a pointer-based thread ID to distinguish between threads.
7924 Applications that want to provide their own thread IDs should now use
7925 CRYPTO_THREADID_set_callback() to register a callback that will call
7926 either CRYPTO_THREADID_set_numeric() or CRYPTO_THREADID_set_pointer().
7928 Note that ERR_remove_state() is now deprecated, because it is tied
7929 to the assumption that thread IDs are numeric. ERR_remove_state(0)
7930 to free the current thread's error state should be replaced by
7931 ERR_remove_thread_state(NULL).
7933 (This new approach replaces the functions CRYPTO_set_idptr_callback(),
7934 CRYPTO_get_idptr_callback(), and CRYPTO_thread_idptr() that existed in
7935 OpenSSL 0.9.9-dev between June 2006 and August 2008. Also, if an
7936 application was previously providing a numeric thread callback that
7937 was inappropriate for distinguishing threads, then uniqueness might
7938 have been obtained with &errno that happened immediately in the
7939 intermediate development versions of OpenSSL; this is no longer the
7940 case, the numeric thread callback will now override the automatic use
7943 *Geoff Thorpe, with help from Bodo Moeller*
7945 * Initial support for different CRL issuing certificates. This covers a
7946 simple case where the self issued certificates in the chain exist and
7947 the real CRL issuer is higher in the existing chain.
7949 This work was sponsored by Google.
7953 * Removed effectively defunct crypto/store from the build.
7957 * Revamp of STACK to provide stronger type-checking. Still to come:
7958 TXT_DB, bsearch(?), OBJ_bsearch, qsort, CRYPTO_EX_DATA, ASN1_VALUE,
7959 ASN1_STRING, CONF_VALUE.
7963 * Add a new SSL_MODE_RELEASE_BUFFERS mode flag to release unused buffer
7964 RAM on SSL connections. This option can save about 34k per idle SSL.
7968 * Revamp of LHASH to provide stronger type-checking. Still to come:
7969 STACK, TXT_DB, bsearch, qsort.
7973 * Initial support for Cryptographic Message Syntax (aka CMS) based
7974 on RFC3850, RFC3851 and RFC3852. New cms directory and cms utility,
7975 support for data, signedData, compressedData, digestedData and
7976 encryptedData, envelopedData types included. Scripts to check against
7977 RFC4134 examples draft and interop and consistency checks of many
7978 content types and variants.
7982 * Add options to enc utility to support use of zlib compression BIO.
7986 * Extend mk1mf to support importing of options and assembly language
7987 files from Configure script, currently only included in VC-WIN32.
7988 The assembly language rules can now optionally generate the source
7989 files from the associated perl scripts.
7993 * Implement remaining functionality needed to support GOST ciphersuites.
7994 Interop testing has been performed using CryptoPro implementations.
7996 *Victor B. Wagner <vitus@cryptocom.ru>*
7998 * s390x assembler pack.
8002 * ARMv4 assembler pack. ARMv4 refers to v4 and later ISA, not CPU
8007 * Implement Opaque PRF Input TLS extension as specified in
8008 draft-rescorla-tls-opaque-prf-input-00.txt. Since this is not an
8009 official specification yet and no extension type assignment by
8010 IANA exists, this extension (for now) will have to be explicitly
8011 enabled when building OpenSSL by providing the extension number
8012 to use. For example, specify an option
8014 -DTLSEXT_TYPE_opaque_prf_input=0x9527
8016 to the "config" or "Configure" script to enable the extension,
8017 assuming extension number 0x9527 (which is a completely arbitrary
8018 and unofficial assignment based on the MD5 hash of the Internet
8019 Draft). Note that by doing so, you potentially lose
8020 interoperability with other TLS implementations since these might
8021 be using the same extension number for other purposes.
8023 SSL_set_tlsext_opaque_prf_input(ssl, src, len) is used to set the
8024 opaque PRF input value to use in the handshake. This will create
8025 an internal copy of the length-'len' string at 'src', and will
8026 return non-zero for success.
8028 To get more control and flexibility, provide a callback function
8031 SSL_CTX_set_tlsext_opaque_prf_input_callback(ctx, cb)
8032 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg(ctx, arg)
8036 int (*cb)(SSL *, void *peerinput, size_t len, void *arg);
8039 Callback function 'cb' will be called in handshakes, and is
8040 expected to use SSL_set_tlsext_opaque_prf_input() as appropriate.
8041 Argument 'arg' is for application purposes (the value as given to
8042 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg() will directly
8043 be provided to the callback function). The callback function
8044 has to return non-zero to report success: usually 1 to use opaque
8045 PRF input just if possible, or 2 to enforce use of the opaque PRF
8046 input. In the latter case, the library will abort the handshake
8047 if opaque PRF input is not successfully negotiated.
8049 Arguments 'peerinput' and 'len' given to the callback function
8050 will always be NULL and 0 in the case of a client. A server will
8051 see the client's opaque PRF input through these variables if
8052 available (NULL and 0 otherwise). Note that if the server
8053 provides an opaque PRF input, the length must be the same as the
8054 length of the client's opaque PRF input.
8056 Note that the callback function will only be called when creating
8057 a new session (session resumption can resume whatever was
8058 previously negotiated), and will not be called in SSL 2.0
8059 handshakes; thus, SSL_CTX_set_options(ctx, SSL_OP_NO_SSLv2) or
8060 SSL_set_options(ssl, SSL_OP_NO_SSLv2) is especially recommended
8061 for applications that need to enforce opaque PRF input.
8065 * Update ssl code to support digests other than SHA1+MD5 for handshake
8068 *Victor B. Wagner <vitus@cryptocom.ru>*
8070 * Add RFC4507 support to OpenSSL. This includes the corrections in
8071 RFC4507bis. The encrypted ticket format is an encrypted encoded
8072 SSL_SESSION structure, that way new session features are automatically
8075 If a client application caches session in an SSL_SESSION structure
8076 support is transparent because tickets are now stored in the encoded
8079 The SSL_CTX structure automatically generates keys for ticket
8080 protection in servers so again support should be possible
8081 with no application modification.
8083 If a client or server wishes to disable RFC4507 support then the option
8084 SSL_OP_NO_TICKET can be set.
8086 Add a TLS extension debugging callback to allow the contents of any client
8087 or server extensions to be examined.
8089 This work was sponsored by Google.
8093 * Final changes to avoid use of pointer pointer casts in OpenSSL.
8094 OpenSSL should now compile cleanly on gcc 4.2
8096 *Peter Hartley <pdh@utter.chaos.org.uk>, Steve Henson*
8098 * Update SSL library to use new EVP_PKEY MAC API. Include generic MAC
8099 support including streaming MAC support: this is required for GOST
8100 ciphersuite support.
8102 *Victor B. Wagner <vitus@cryptocom.ru>, Steve Henson*
8104 * Add option -stream to use PKCS#7 streaming in smime utility. New
8105 function i2d_PKCS7_bio_stream() and PEM_write_PKCS7_bio_stream()
8106 to output in BER and PEM format.
8110 * Experimental support for use of HMAC via EVP_PKEY interface. This
8111 allows HMAC to be handled via the `EVP_DigestSign*()` interface. The
8112 EVP_PKEY "key" in this case is the HMAC key, potentially allowing
8113 ENGINE support for HMAC keys which are unextractable. New -mac and
8114 -macopt options to dgst utility.
8118 * New option -sigopt to dgst utility. Update dgst to use
8119 `EVP_Digest{Sign,Verify}*`. These two changes make it possible to use
8120 alternative signing parameters such as X9.31 or PSS in the dgst
8125 * Change ssl_cipher_apply_rule(), the internal function that does
8126 the work each time a ciphersuite string requests enabling
8127 ("foo+bar"), moving ("+foo+bar"), disabling ("-foo+bar", or
8128 removing ("!foo+bar") a class of ciphersuites: Now it maintains
8129 the order of disabled ciphersuites such that those ciphersuites
8130 that most recently went from enabled to disabled not only stay
8131 in order with respect to each other, but also have higher priority
8132 than other disabled ciphersuites the next time ciphersuites are
8135 This means that you can now say, e.g., "PSK:-PSK:HIGH" to enable
8136 the same ciphersuites as with "HIGH" alone, but in a specific
8137 order where the PSK ciphersuites come first (since they are the
8138 most recently disabled ciphersuites when "HIGH" is parsed).
8140 Also, change ssl_create_cipher_list() (using this new
8141 functionality) such that between otherwise identical
8142 ciphersuites, ephemeral ECDH is preferred over ephemeral DH in
8147 * Change ssl_create_cipher_list() so that it automatically
8148 arranges the ciphersuites in reasonable order before starting
8149 to process the rule string. Thus, the definition for "DEFAULT"
8150 (SSL_DEFAULT_CIPHER_LIST) now is just "ALL:!aNULL:!eNULL", but
8151 remains equivalent to "AES:ALL:!aNULL:!eNULL:+aECDH:+kRSA:+RC4:@STRENGTH".
8152 This makes it much easier to arrive at a reasonable default order
8153 in applications for which anonymous ciphers are OK (meaning
8154 that you can't actually use DEFAULT).
8156 *Bodo Moeller; suggested by Victor Duchovni*
8158 * Split the SSL/TLS algorithm mask (as used for ciphersuite string
8159 processing) into multiple integers instead of setting
8160 "SSL_MKEY_MASK" bits, "SSL_AUTH_MASK" bits, "SSL_ENC_MASK",
8161 "SSL_MAC_MASK", and "SSL_SSL_MASK" bits all in a single integer.
8162 (These masks as well as the individual bit definitions are hidden
8163 away into the non-exported interface ssl/ssl_locl.h, so this
8164 change to the definition of the SSL_CIPHER structure shouldn't
8165 affect applications.) This give us more bits for each of these
8166 categories, so there is no longer a need to coagulate AES128 and
8167 AES256 into a single algorithm bit, and to coagulate Camellia128
8168 and Camellia256 into a single algorithm bit, which has led to all
8171 Thus, among other things, the kludge introduced in 0.9.7m and
8172 0.9.8e for masking out AES256 independently of AES128 or masking
8173 out Camellia256 independently of AES256 is not needed here in 0.9.9.
8175 With the change, we also introduce new ciphersuite aliases that
8176 so far were missing: "AES128", "AES256", "CAMELLIA128", and
8181 * Add support for dsa-with-SHA224 and dsa-with-SHA256.
8182 Use the leftmost N bytes of the signature input if the input is
8183 larger than the prime q (with N being the size in bytes of q).
8187 * Very *very* experimental PKCS#7 streaming encoder support. Nothing uses
8188 it yet and it is largely untested.
8192 * Add support for the ecdsa-with-SHA224/256/384/512 signature types.
8196 * Initial incomplete changes to avoid need for function casts in OpenSSL
8197 some compilers (gcc 4.2 and later) reject their use. Safestack is
8198 reimplemented. Update ASN1 to avoid use of legacy functions.
8202 * Win32/64 targets are linked with Winsock2.
8206 * Add an X509_CRL_METHOD structure to allow CRL processing to be redirected
8207 to external functions. This can be used to increase CRL handling
8208 efficiency especially when CRLs are very large by (for example) storing
8209 the CRL revoked certificates in a database.
8213 * Overhaul of by_dir code. Add support for dynamic loading of CRLs so
8214 new CRLs added to a directory can be used. New command line option
8215 -verify_return_error to s_client and s_server. This causes real errors
8216 to be returned by the verify callback instead of carrying on no matter
8217 what. This reflects the way a "real world" verify callback would behave.
8221 * GOST engine, supporting several GOST algorithms and public key formats.
8222 Kindly donated by Cryptocom.
8226 * Partial support for Issuing Distribution Point CRL extension. CRLs
8227 partitioned by DP are handled but no indirect CRL or reason partitioning
8228 (yet). Complete overhaul of CRL handling: now the most suitable CRL is
8229 selected via a scoring technique which handles IDP and AKID in CRLs.
8233 * New X509_STORE_CTX callbacks lookup_crls() and lookup_certs() which
8234 will ultimately be used for all verify operations: this will remove the
8235 X509_STORE dependency on certificate verification and allow alternative
8236 lookup methods. X509_STORE based implementations of these two callbacks.
8240 * Allow multiple CRLs to exist in an X509_STORE with matching issuer names.
8241 Modify get_crl() to find a valid (unexpired) CRL if possible.
8245 * New function X509_CRL_match() to check if two CRLs are identical. Normally
8246 this would be called X509_CRL_cmp() but that name is already used by
8247 a function that just compares CRL issuer names. Cache several CRL
8248 extensions in X509_CRL structure and cache CRLDP in X509.
8252 * Store a "canonical" representation of X509_NAME structure (ASN1 Name)
8253 this maps equivalent X509_NAME structures into a consistent structure.
8254 Name comparison can then be performed rapidly using memcmp().
8258 * Non-blocking OCSP request processing. Add -timeout option to ocsp
8263 * Allow digests to supply their own micalg string for S/MIME type using
8264 the ctrl EVP_MD_CTRL_MICALG.
8268 * During PKCS7 signing pass the PKCS7 SignerInfo structure to the
8269 EVP_PKEY_METHOD before and after signing via the EVP_PKEY_CTRL_PKCS7_SIGN
8270 ctrl. It can then customise the structure before and/or after signing
8275 * New function OBJ_add_sigid() to allow application defined signature OIDs
8276 to be added to OpenSSLs internal tables. New function OBJ_sigid_free()
8277 to free up any added signature OIDs.
8281 * New functions EVP_CIPHER_do_all(), EVP_CIPHER_do_all_sorted(),
8282 EVP_MD_do_all() and EVP_MD_do_all_sorted() to enumerate internal
8283 digest and cipher tables. New options added to openssl utility:
8284 list-message-digest-algorithms and list-cipher-algorithms.
8288 * Change the array representation of binary polynomials: the list
8289 of degrees of non-zero coefficients is now terminated with -1.
8290 Previously it was terminated with 0, which was also part of the
8291 value; thus, the array representation was not applicable to
8292 polynomials where t^0 has coefficient zero. This change makes
8293 the array representation useful in a more general context.
8297 * Various modifications and fixes to SSL/TLS cipher string
8298 handling. For ECC, the code now distinguishes between fixed ECDH
8299 with RSA certificates on the one hand and with ECDSA certificates
8300 on the other hand, since these are separate ciphersuites. The
8301 unused code for Fortezza ciphersuites has been removed.
8303 For consistency with EDH, ephemeral ECDH is now called "EECDH"
8304 (not "ECDHE"). For consistency with the code for DH
8305 certificates, use of ECDH certificates is now considered ECDH
8306 authentication, not RSA or ECDSA authentication (the latter is
8307 merely the CA's signing algorithm and not actively used in the
8310 The temporary ciphersuite alias "ECCdraft" is no longer
8311 available, and ECC ciphersuites are no longer excluded from "ALL"
8312 and "DEFAULT". The following aliases now exist for RFC 4492
8313 ciphersuites, most of these by analogy with the DH case:
8315 kECDHr - ECDH cert, signed with RSA
8316 kECDHe - ECDH cert, signed with ECDSA
8317 kECDH - ECDH cert (signed with either RSA or ECDSA)
8318 kEECDH - ephemeral ECDH
8319 ECDH - ECDH cert or ephemeral ECDH
8325 AECDH - anonymous ECDH
8326 EECDH - non-anonymous ephemeral ECDH (equivalent to "kEECDH:-AECDH")
8330 * Add additional S/MIME capabilities for AES and GOST ciphers if supported.
8331 Use correct micalg parameters depending on digest(s) in signed message.
8335 * Add engine support for EVP_PKEY_ASN1_METHOD. Add functions to process
8336 an ENGINE asn1 method. Support ENGINE lookups in the ASN1 code.
8340 * Initial engine support for EVP_PKEY_METHOD. New functions to permit
8341 an engine to register a method. Add ENGINE lookups for methods and
8342 functional reference processing.
8346 * New functions `EVP_Digest{Sign,Verify)*`. These are enhanced versions of
8347 `EVP_{Sign,Verify}*` which allow an application to customise the signature
8352 * New -resign option to smime utility. This adds one or more signers
8353 to an existing PKCS#7 signedData structure. Also -md option to use an
8354 alternative message digest algorithm for signing.
8358 * Tidy up PKCS#7 routines and add new functions to make it easier to
8359 create PKCS7 structures containing multiple signers. Update smime
8360 application to support multiple signers.
8364 * New -macalg option to pkcs12 utility to allow setting of an alternative
8369 * Initial support for PKCS#5 v2.0 PRFs other than default SHA1 HMAC.
8370 Reorganize PBE internals to lookup from a static table using NIDs,
8371 add support for HMAC PBE OID translation. Add a EVP_CIPHER ctrl:
8372 EVP_CTRL_PBE_PRF_NID this allows a cipher to specify an alternative
8373 PRF which will be automatically used with PBES2.
8377 * Replace the algorithm specific calls to generate keys in "req" with the
8382 * Update PKCS#7 enveloped data routines to use new API. This is now
8383 supported by any public key method supporting the encrypt operation. A
8384 ctrl is added to allow the public key algorithm to examine or modify
8385 the PKCS#7 RecipientInfo structure if it needs to: for RSA this is
8390 * Add a ctrl to asn1 method to allow a public key algorithm to express
8391 a default digest type to use. In most cases this will be SHA1 but some
8392 algorithms (such as GOST) need to specify an alternative digest. The
8393 return value indicates how strong the preference is 1 means optional and
8394 2 is mandatory (that is it is the only supported type). Modify
8395 ASN1_item_sign() to accept a NULL digest argument to indicate it should
8396 use the default md. Update openssl utilities to use the default digest
8397 type for signing if it is not explicitly indicated.
8401 * Use OID cross reference table in ASN1_sign() and ASN1_verify(). New
8402 EVP_MD flag EVP_MD_FLAG_PKEY_METHOD_SIGNATURE. This uses the relevant
8403 signing method from the key type. This effectively removes the link
8404 between digests and public key types.
8408 * Add an OID cross reference table and utility functions. Its purpose is to
8409 translate between signature OIDs such as SHA1WithrsaEncryption and SHA1,
8410 rsaEncryption. This will allow some of the algorithm specific hackery
8411 needed to use the correct OID to be removed.
8415 * Remove algorithm specific dependencies when setting PKCS7_SIGNER_INFO
8416 structures for PKCS7_sign(). They are now set up by the relevant public
8421 * Add provisional EC pkey method with support for ECDSA and ECDH.
8425 * Add support for key derivation (agreement) in the API, DH method and
8430 * Add DSA pkey method and DH pkey methods, extend DH ASN1 method to support
8431 public and private key formats. As a side effect these add additional
8432 command line functionality not previously available: DSA signatures can be
8433 generated and verified using pkeyutl and DH key support and generation in
8440 *Oliver Tappe <zooey@hirschkaefer.de>*
8442 * New make target "install_html_docs" installs HTML renditions of the
8445 *Oliver Tappe <zooey@hirschkaefer.de>*
8447 * New utility "genpkey" this is analogous to "genrsa" etc except it can
8448 generate keys for any algorithm. Extend and update EVP_PKEY_METHOD to
8449 support key and parameter generation and add initial key generation
8450 functionality for RSA.
8454 * Add functions for main EVP_PKEY_method operations. The undocumented
8455 functions `EVP_PKEY_{encrypt,decrypt}` have been renamed to
8456 `EVP_PKEY_{encrypt,decrypt}_old`.
8460 * Initial definitions for EVP_PKEY_METHOD. This will be a high level public
8461 key API, doesn't do much yet.
8465 * New function EVP_PKEY_asn1_get0_info() to retrieve information about
8466 public key algorithms. New option to openssl utility:
8467 "list-public-key-algorithms" to print out info.
8471 * Implement the Supported Elliptic Curves Extension for
8472 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
8476 * Don't free up OIDs in OBJ_cleanup() if they are in use by EVP_MD or
8477 EVP_CIPHER structures to avoid later problems in EVP_cleanup().
8481 * New utilities pkey and pkeyparam. These are similar to algorithm specific
8482 utilities such as rsa, dsa, dsaparam etc except they process any key
8487 * Transfer public key printing routines to EVP_PKEY_ASN1_METHOD. New
8488 functions EVP_PKEY_print_public(), EVP_PKEY_print_private(),
8489 EVP_PKEY_print_param() to print public key data from an EVP_PKEY
8494 * Initial support for pluggable public key ASN1.
8495 De-spaghettify the public key ASN1 handling. Move public and private
8496 key ASN1 handling to a new EVP_PKEY_ASN1_METHOD structure. Relocate
8497 algorithm specific handling to a single module within the relevant
8498 algorithm directory. Add functions to allow (near) opaque processing
8499 of public and private key structures.
8503 * Implement the Supported Point Formats Extension for
8504 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
8508 * Add initial support for RFC 4279 PSK TLS ciphersuites. Add members
8509 for the psk identity [hint] and the psk callback functions to the
8510 SSL_SESSION, SSL and SSL_CTX structure.
8513 PSK-RC4-SHA, PSK-3DES-EDE-CBC-SHA, PSK-AES128-CBC-SHA,
8517 SSL_CTX_use_psk_identity_hint
8518 SSL_get_psk_identity_hint
8519 SSL_get_psk_identity
8520 SSL_use_psk_identity_hint
8522 *Mika Kousa and Pasi Eronen of Nokia Corporation*
8524 * Add RFC 3161 compliant time stamp request creation, response generation
8525 and response verification functionality.
8527 *Zoltán Glózik <zglozik@opentsa.org>, The OpenTSA Project*
8529 * Add initial support for TLS extensions, specifically for the server_name
8530 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
8531 have new members for a host name. The SSL data structure has an
8532 additional member `SSL_CTX *initial_ctx` so that new sessions can be
8533 stored in that context to allow for session resumption, even after the
8534 SSL has been switched to a new SSL_CTX in reaction to a client's
8535 server_name extension.
8537 New functions (subject to change):
8539 SSL_get_servername()
8540 SSL_get_servername_type()
8543 New CTRL codes and macros (subject to change):
8545 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
8546 - SSL_CTX_set_tlsext_servername_callback()
8547 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
8548 - SSL_CTX_set_tlsext_servername_arg()
8549 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
8551 openssl s_client has a new '-servername ...' option.
8553 openssl s_server has new options '-servername_host ...', '-cert2 ...',
8554 '-key2 ...', '-servername_fatal' (subject to change). This allows
8555 testing the HostName extension for a specific single host name ('-cert'
8556 and '-key' remain fallbacks for handshakes without HostName
8557 negotiation). If the unrecognized_name alert has to be sent, this by
8558 default is a warning; it becomes fatal with the '-servername_fatal'
8561 *Peter Sylvester, Remy Allais, Christophe Renou*
8563 * Whirlpool hash implementation is added.
8567 * BIGNUM code on 64-bit SPARCv9 targets is switched from bn(64,64) to
8568 bn(64,32). Because of instruction set limitations it doesn't have
8569 any negative impact on performance. This was done mostly in order
8570 to make it possible to share assembler modules, such as bn_mul_mont
8571 implementations, between 32- and 64-bit builds without hassle.
8575 * Move code previously exiled into file crypto/ec/ec2_smpt.c
8576 to ec2_smpl.c, and no longer require the OPENSSL_EC_BIN_PT_COMP
8581 * New candidate for BIGNUM assembler implementation, bn_mul_mont,
8582 dedicated Montgomery multiplication procedure, is introduced.
8583 BN_MONT_CTX is modified to allow bn_mul_mont to reach for higher
8584 "64-bit" performance on certain 32-bit targets.
8588 * New option SSL_OP_NO_COMP to disable use of compression selectively
8589 in SSL structures. New SSL ctrl to set maximum send fragment size.
8590 Save memory by setting the I/O buffer sizes dynamically instead of
8591 using the maximum available value.
8595 * New option -V for 'openssl ciphers'. This prints the ciphersuite code
8596 in addition to the text details.
8600 * Very, very preliminary EXPERIMENTAL support for printing of general
8601 ASN1 structures. This currently produces rather ugly output and doesn't
8602 handle several customised structures at all.
8606 * Integrated support for PVK file format and some related formats such
8607 as MS PUBLICKEYBLOB and PRIVATEKEYBLOB. Command line switches to support
8608 these in the 'rsa' and 'dsa' utilities.
8612 * Support for PKCS#1 RSAPublicKey format on rsa utility command line.
8616 * Remove the ancient ASN1_METHOD code. This was only ever used in one
8617 place for the (very old) "NETSCAPE" format certificates which are now
8618 handled using new ASN1 code equivalents.
8622 * Let the TLSv1_method() etc. functions return a 'const' SSL_METHOD
8623 pointer and make the SSL_METHOD parameter in SSL_CTX_new,
8624 SSL_CTX_set_ssl_version and SSL_set_ssl_method 'const'.
8628 * Modify CRL distribution points extension code to print out previously
8629 unsupported fields. Enhance extension setting code to allow setting of
8634 * Add print and set support for Issuing Distribution Point CRL extension.
8638 * Change 'Configure' script to enable Camellia by default.
8645 ### Changes between 0.9.8m and 0.9.8n [24 Mar 2010]
8647 * When rejecting SSL/TLS records due to an incorrect version number, never
8648 update s->server with a new major version number. As of
8649 - OpenSSL 0.9.8m if 'short' is a 16-bit type,
8650 - OpenSSL 0.9.8f if 'short' is longer than 16 bits,
8651 the previous behavior could result in a read attempt at NULL when
8652 receiving specific incorrect SSL/TLS records once record payload
8653 protection is active. [CVE-2010-0740][]
8655 *Bodo Moeller, Adam Langley <agl@chromium.org>*
8657 * Fix for CVE-2010-0433 where some kerberos enabled versions of OpenSSL
8658 could be crashed if the relevant tables were not present (e.g. chrooted).
8660 *Tomas Hoger <thoger@redhat.com>*
8662 ### Changes between 0.9.8l and 0.9.8m [25 Feb 2010]
8664 * Always check bn_wexpand() return values for failure. [CVE-2009-3245][]
8666 *Martin Olsson, Neel Mehta*
8668 * Fix X509_STORE locking: Every 'objs' access requires a lock (to
8669 accommodate for stack sorting, always a write lock!).
8673 * On some versions of WIN32 Heap32Next is very slow. This can cause
8674 excessive delays in the RAND_poll(): over a minute. As a workaround
8675 include a time check in the inner Heap32Next loop too.
8679 * The code that handled flushing of data in SSL/TLS originally used the
8680 BIO_CTRL_INFO ctrl to see if any data was pending first. This caused
8681 the problem outlined in PR#1949. The fix suggested there however can
8682 trigger problems with buggy BIO_CTRL_WPENDING (e.g. some versions
8683 of Apache). So instead simplify the code to flush unconditionally.
8684 This should be fine since flushing with no data to flush is a no op.
8688 * Handle TLS versions 2.0 and later properly and correctly use the
8689 highest version of TLS/SSL supported. Although TLS >= 2.0 is some way
8690 off ancient servers have a habit of sticking around for a while...
8694 * Modify compression code so it frees up structures without using the
8695 ex_data callbacks. This works around a problem where some applications
8696 call CRYPTO_cleanup_all_ex_data() before application exit (e.g. when
8697 restarting) then use compression (e.g. SSL with compression) later.
8698 This results in significant per-connection memory leaks and
8699 has caused some security issues including CVE-2008-1678 and
8704 * Constify crypto/cast (i.e., <openssl/cast.h>): a CAST_KEY doesn't
8705 change when encrypting or decrypting.
8709 * Add option SSL_OP_LEGACY_SERVER_CONNECT which will allow clients to
8710 connect and renegotiate with servers which do not support RI.
8711 Until RI is more widely deployed this option is enabled by default.
8715 * Add "missing" ssl ctrls to clear options and mode.
8719 * If client attempts to renegotiate and doesn't support RI respond with
8720 a no_renegotiation alert as required by RFC5746. Some renegotiating
8721 TLS clients will continue a connection gracefully when they receive
8722 the alert. Unfortunately OpenSSL mishandled this alert and would hang
8723 waiting for a server hello which it will never receive. Now we treat a
8724 received no_renegotiation alert as a fatal error. This is because
8725 applications requesting a renegotiation might well expect it to succeed
8726 and would have no code in place to handle the server denying it so the
8727 only safe thing to do is to terminate the connection.
8731 * Add ctrl macro SSL_get_secure_renegotiation_support() which returns 1 if
8732 peer supports secure renegotiation and 0 otherwise. Print out peer
8733 renegotiation support in s_client/s_server.
8737 * Replace the highly broken and deprecated SPKAC certification method with
8738 the updated NID creation version. This should correctly handle UTF8.
8742 * Implement RFC5746. Re-enable renegotiation but require the extension
8743 as needed. Unfortunately, SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION
8744 turns out to be a bad idea. It has been replaced by
8745 SSL_OP_ALLOW_UNSAFE_LEGACY_RENEGOTIATION which can be set with
8746 SSL_CTX_set_options(). This is really not recommended unless you
8747 know what you are doing.
8749 *Eric Rescorla <ekr@networkresonance.com>, Ben Laurie, Steve Henson*
8751 * Fixes to stateless session resumption handling. Use initial_ctx when
8752 issuing and attempting to decrypt tickets in case it has changed during
8753 servername handling. Use a non-zero length session ID when attempting
8754 stateless session resumption: this makes it possible to determine if
8755 a resumption has occurred immediately after receiving server hello
8756 (several places in OpenSSL subtly assume this) instead of later in
8761 * The functions ENGINE_ctrl(), OPENSSL_isservice(),
8762 CMS_get1_RecipientRequest() and RAND_bytes() can return <=0 on error
8763 fixes for a few places where the return code is not checked
8766 *Julia Lawall <julia@diku.dk>*
8768 * Add --strict-warnings option to Configure script to include devteam
8769 warnings in other configurations.
8773 * Add support for --libdir option and LIBDIR variable in makefiles. This
8774 makes it possible to install openssl libraries in locations which
8775 have names other than "lib", for example "/usr/lib64" which some
8778 *Steve Henson, based on patch from Jeremy Utley*
8780 * Don't allow the use of leading 0x80 in OIDs. This is a violation of
8781 X690 8.9.12 and can produce some misleading textual output of OIDs.
8783 *Steve Henson, reported by Dan Kaminsky*
8785 * Delete MD2 from algorithm tables. This follows the recommendation in
8786 several standards that it is not used in new applications due to
8787 several cryptographic weaknesses. For binary compatibility reasons
8788 the MD2 API is still compiled in by default.
8792 * Add compression id to {d2i,i2d}_SSL_SESSION so it is correctly saved
8797 * Rename uni2asc and asc2uni functions to OPENSSL_uni2asc and
8798 OPENSSL_asc2uni conditionally on Netware platforms to avoid a name
8801 *Guenter <lists@gknw.net>*
8803 * Fix the server certificate chain building code to use X509_verify_cert(),
8804 it used to have an ad-hoc builder which was unable to cope with anything
8805 other than a simple chain.
8807 *David Woodhouse <dwmw2@infradead.org>, Steve Henson*
8809 * Don't check self signed certificate signatures in X509_verify_cert()
8810 by default (a flag can override this): it just wastes time without
8811 adding any security. As a useful side effect self signed root CAs
8812 with non-FIPS digests are now usable in FIPS mode.
8816 * In dtls1_process_out_of_seq_message() the check if the current message
8817 is already buffered was missing. For every new message was memory
8818 allocated, allowing an attacker to perform an denial of service attack
8819 with sending out of seq handshake messages until there is no memory
8820 left. Additionally every future message was buffered, even if the
8821 sequence number made no sense and would be part of another handshake.
8822 So only messages with sequence numbers less than 10 in advance will be
8823 buffered. [CVE-2009-1378][]
8825 *Robin Seggelmann, discovered by Daniel Mentz*
8827 * Records are buffered if they arrive with a future epoch to be
8828 processed after finishing the corresponding handshake. There is
8829 currently no limitation to this buffer allowing an attacker to perform
8830 a DOS attack with sending records with future epochs until there is no
8831 memory left. This patch adds the pqueue_size() function to determine
8832 the size of a buffer and limits the record buffer to 100 entries.
8835 *Robin Seggelmann, discovered by Daniel Mentz*
8837 * Keep a copy of frag->msg_header.frag_len so it can be used after the
8838 parent structure is freed. [CVE-2009-1379][]
8842 * Handle non-blocking I/O properly in SSL_shutdown() call.
8844 *Darryl Miles <darryl-mailinglists@netbauds.net>*
8846 * Add `2.5.4.*` OIDs
8848 *Ilya O. <vrghost@gmail.com>*
8850 ### Changes between 0.9.8k and 0.9.8l [5 Nov 2009]
8852 * Disable renegotiation completely - this fixes a severe security
8853 problem [CVE-2009-3555][] at the cost of breaking all
8854 renegotiation. Renegotiation can be re-enabled by setting
8855 SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION in s3->flags at
8856 run-time. This is really not recommended unless you know what
8861 ### Changes between 0.9.8j and 0.9.8k [25 Mar 2009]
8863 * Don't set val to NULL when freeing up structures, it is freed up by
8864 underlying code. If `sizeof(void *) > sizeof(long)` this can result in
8865 zeroing past the valid field. [CVE-2009-0789][]
8867 *Paolo Ganci <Paolo.Ganci@AdNovum.CH>*
8869 * Fix bug where return value of CMS_SignerInfo_verify_content() was not
8870 checked correctly. This would allow some invalid signed attributes to
8871 appear to verify correctly. [CVE-2009-0591][]
8873 *Ivan Nestlerode <inestlerode@us.ibm.com>*
8875 * Reject UniversalString and BMPString types with invalid lengths. This
8876 prevents a crash in ASN1_STRING_print_ex() which assumes the strings have
8877 a legal length. [CVE-2009-0590][]
8881 * Set S/MIME signing as the default purpose rather than setting it
8882 unconditionally. This allows applications to override it at the store
8887 * Permit restricted recursion of ASN1 strings. This is needed in practice
8888 to handle some structures.
8892 * Improve efficiency of mem_gets: don't search whole buffer each time
8895 *Jeremy Shapiro <jnshapir@us.ibm.com>*
8897 * New -hex option for openssl rand.
8901 * Print out UTF8String and NumericString when parsing ASN1.
8905 * Support NumericString type for name components.
8909 * Allow CC in the environment to override the automatically chosen
8910 compiler. Note that nothing is done to ensure flags work with the
8915 ### Changes between 0.9.8i and 0.9.8j [07 Jan 2009]
8917 * Properly check EVP_VerifyFinal() and similar return values
8920 *Ben Laurie, Bodo Moeller, Google Security Team*
8922 * Enable TLS extensions by default.
8926 * Allow the CHIL engine to be loaded, whether the application is
8927 multithreaded or not. (This does not release the developer from the
8928 obligation to set up the dynamic locking callbacks.)
8930 *Sander Temme <sander@temme.net>*
8932 * Use correct exit code if there is an error in dgst command.
8934 *Steve Henson; problem pointed out by Roland Dirlewanger*
8936 * Tweak Configure so that you need to say "experimental-jpake" to enable
8937 JPAKE, and need to use -DOPENSSL_EXPERIMENTAL_JPAKE in applications.
8941 * Add experimental JPAKE support, including demo authentication in
8942 s_client and s_server.
8946 * Set the comparison function in v3_addr_canonize().
8948 *Rob Austein <sra@hactrn.net>*
8950 * Add support for XMPP STARTTLS in s_client.
8952 *Philip Paeps <philip@freebsd.org>*
8954 * Change the server-side SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG behavior
8955 to ensure that even with this option, only ciphersuites in the
8956 server's preference list will be accepted. (Note that the option
8957 applies only when resuming a session, so the earlier behavior was
8958 just about the algorithm choice for symmetric cryptography.)
8962 ### Changes between 0.9.8h and 0.9.8i [15 Sep 2008]
8964 * Fix NULL pointer dereference if a DTLS server received
8965 ChangeCipherSpec as first record [CVE-2009-1386][].
8969 * Fix a state transition in s3_srvr.c and d1_srvr.c
8970 (was using SSL3_ST_CW_CLNT_HELLO_B, should be `..._ST_SW_SRVR_...`).
8974 * The fix in 0.9.8c that supposedly got rid of unsafe
8975 double-checked locking was incomplete for RSA blinding,
8976 addressing just one layer of what turns out to have been
8977 doubly unsafe triple-checked locking.
8979 So now fix this for real by retiring the MONT_HELPER macro
8980 in crypto/rsa/rsa_eay.c.
8982 *Bodo Moeller; problem pointed out by Marius Schilder*
8984 * Various precautionary measures:
8986 - Avoid size_t integer overflow in HASH_UPDATE (md32_common.h).
8988 - Avoid a buffer overflow in d2i_SSL_SESSION() (ssl_asn1.c).
8989 (NB: This would require knowledge of the secret session ticket key
8990 to exploit, in which case you'd be SOL either way.)
8992 - Change bn_nist.c so that it will properly handle input BIGNUMs
8993 outside the expected range.
8995 - Enforce the 'num' check in BN_div() (bn_div.c) for non-BN_DEBUG
8998 *Neel Mehta, Bodo Moeller*
9000 * Allow engines to be "soft loaded" - i.e. optionally don't die if
9001 the load fails. Useful for distros.
9003 *Ben Laurie and the FreeBSD team*
9005 * Add support for Local Machine Keyset attribute in PKCS#12 files.
9009 * Fix BN_GF2m_mod_arr() top-bit cleanup code.
9013 * Expand ENGINE to support engine supplied SSL client certificate functions.
9015 This work was sponsored by Logica.
9019 * Add CryptoAPI ENGINE to support use of RSA and DSA keys held in Windows
9020 keystores. Support for SSL/TLS client authentication too.
9021 Not compiled unless enable-capieng specified to Configure.
9023 This work was sponsored by Logica.
9027 * Fix bug in X509_ATTRIBUTE creation: don't set attribute using
9028 ASN1_TYPE_set1 if MBSTRING flag set. This bug would crash certain
9029 attribute creation routines such as certificate requests and PKCS#12
9034 ### Changes between 0.9.8g and 0.9.8h [28 May 2008]
9036 * Fix flaw if 'Server Key exchange message' is omitted from a TLS
9037 handshake which could lead to a client crash as found using the
9038 Codenomicon TLS test suite [CVE-2008-1672][]
9040 *Steve Henson, Mark Cox*
9042 * Fix double free in TLS server name extensions which could lead to
9043 a remote crash found by Codenomicon TLS test suite [CVE-2008-0891][]
9047 * Clear error queue in SSL_CTX_use_certificate_chain_file()
9049 Clear the error queue to ensure that error entries left from
9050 older function calls do not interfere with the correct operation.
9052 *Lutz Jaenicke, Erik de Castro Lopo*
9054 * Remove root CA certificates of commercial CAs:
9056 The OpenSSL project does not recommend any specific CA and does not
9057 have any policy with respect to including or excluding any CA.
9058 Therefore it does not make any sense to ship an arbitrary selection
9059 of root CA certificates with the OpenSSL software.
9063 * RSA OAEP patches to fix two separate invalid memory reads.
9064 The first one involves inputs when 'lzero' is greater than
9065 'SHA_DIGEST_LENGTH' (it would read about SHA_DIGEST_LENGTH bytes
9066 before the beginning of from). The second one involves inputs where
9067 the 'db' section contains nothing but zeroes (there is a one-byte
9068 invalid read after the end of 'db').
9070 *Ivan Nestlerode <inestlerode@us.ibm.com>*
9072 * Partial backport from 0.9.9-dev:
9074 Introduce bn_mul_mont (dedicated Montgomery multiplication
9075 procedure) as a candidate for BIGNUM assembler implementation.
9076 While 0.9.9-dev uses assembler for various architectures, only
9077 x86_64 is available by default here in the 0.9.8 branch, and
9078 32-bit x86 is available through a compile-time setting.
9080 To try the 32-bit x86 assembler implementation, use Configure
9081 option "enable-montasm" (which exists only for this backport).
9083 As "enable-montasm" for 32-bit x86 disclaims code stability
9084 anyway, in this constellation we activate additional code
9085 backported from 0.9.9-dev for further performance improvements,
9086 namely BN_from_montgomery_word. (To enable this otherwise,
9087 e.g. x86_64, try `-DMONT_FROM_WORD___NON_DEFAULT_0_9_8_BUILD`.)
9089 *Andy Polyakov (backport partially by Bodo Moeller)*
9091 * Add TLS session ticket callback. This allows an application to set
9092 TLS ticket cipher and HMAC keys rather than relying on hardcoded fixed
9093 values. This is useful for key rollover for example where several key
9094 sets may exist with different names.
9098 * Reverse ENGINE-internal logic for caching default ENGINE handles.
9099 This was broken until now in 0.9.8 releases, such that the only way
9100 a registered ENGINE could be used (assuming it initialises
9101 successfully on the host) was to explicitly set it as the default
9102 for the relevant algorithms. This is in contradiction with 0.9.7
9103 behaviour and the documentation. With this fix, when an ENGINE is
9104 registered into a given algorithm's table of implementations, the
9105 'uptodate' flag is reset so that auto-discovery will be used next
9106 time a new context for that algorithm attempts to select an
9109 *Ian Lister (tweaked by Geoff Thorpe)*
9111 * Backport of CMS code to OpenSSL 0.9.8. This differs from the 0.9.9
9112 implementation in the following ways:
9114 Lack of EVP_PKEY_ASN1_METHOD means algorithm parameters have to be
9117 Lack of BER streaming support means one pass streaming processing is
9118 only supported if data is detached: setting the streaming flag is
9119 ignored for embedded content.
9121 CMS support is disabled by default and must be explicitly enabled
9122 with the enable-cms configuration option.
9126 * Update the GMP engine glue to do direct copies between BIGNUM and
9127 mpz_t when openssl and GMP use the same limb size. Otherwise the
9128 existing "conversion via a text string export" trick is still used.
9130 *Paul Sheer <paulsheer@gmail.com>*
9132 * Zlib compression BIO. This is a filter BIO which compressed and
9133 uncompresses any data passed through it.
9137 * Add AES_wrap_key() and AES_unwrap_key() functions to implement
9138 RFC3394 compatible AES key wrapping.
9142 * Add utility functions to handle ASN1 structures. ASN1_STRING_set0():
9143 sets string data without copying. X509_ALGOR_set0() and
9144 X509_ALGOR_get0(): set and retrieve X509_ALGOR (AlgorithmIdentifier)
9145 data. Attribute function X509at_get0_data_by_OBJ(): retrieves data
9146 from an X509_ATTRIBUTE structure optionally checking it occurs only
9147 once. ASN1_TYPE_set1(): set and ASN1_TYPE structure copying supplied
9152 * Fix BN flag handling in RSA_eay_mod_exp() and BN_MONT_CTX_set()
9153 to get the expected BN_FLG_CONSTTIME behavior.
9155 *Bodo Moeller (Google)*
9159 - fixed wrong usage of ioctlsocket() when build for LIBC BSD sockets
9160 - fixed do_tests.pl to run the test suite with CLIB builds too (CLIB_OPT)
9161 - added some more tests to do_tests.pl
9162 - fixed RunningProcess usage so that it works with newer LIBC NDKs too
9163 - removed usage of BN_LLONG for CLIB builds to avoid runtime dependency
9164 - added new Configure targets netware-clib-bsdsock, netware-clib-gcc,
9165 netware-clib-bsdsock-gcc, netware-libc-bsdsock-gcc
9166 - various changes to netware.pl to enable gcc-cross builds on Win32
9168 - changed crypto/bio/b_sock.c to work with macro functions (CLIB BSD)
9169 - various changes to fix missing prototype warnings
9170 - fixed x86nasm.pl to create correct asm files for NASM COFF output
9171 - added AES, WHIRLPOOL and CPUID assembler code to build files
9172 - added missing AES assembler make rules to mk1mf.pl
9173 - fixed order of includes in apps/ocsp.c so that e_os.h settings apply
9175 *Guenter Knauf <eflash@gmx.net>*
9177 * Implement certificate status request TLS extension defined in RFC3546.
9178 A client can set the appropriate parameters and receive the encoded
9179 OCSP response via a callback. A server can query the supplied parameters
9180 and set the encoded OCSP response in the callback. Add simplified examples
9181 to s_client and s_server.
9185 ### Changes between 0.9.8f and 0.9.8g [19 Oct 2007]
9188 + Binary incompatibility of ssl_ctx_st structure
9189 + DTLS interoperation with non-compliant servers
9190 + Don't call get_session_cb() without proposed session
9191 + Fix ia64 assembler code
9193 *Andy Polyakov, Steve Henson*
9195 ### Changes between 0.9.8e and 0.9.8f [11 Oct 2007]
9197 * DTLS Handshake overhaul. There were longstanding issues with
9198 OpenSSL DTLS implementation, which were making it impossible for
9199 RFC 4347 compliant client to communicate with OpenSSL server.
9200 Unfortunately just fixing these incompatibilities would "cut off"
9201 pre-0.9.8f clients. To allow for hassle free upgrade post-0.9.8e
9202 server keeps tolerating non RFC compliant syntax. The opposite is
9203 not true, 0.9.8f client can not communicate with earlier server.
9204 This update even addresses CVE-2007-4995.
9208 * Changes to avoid need for function casts in OpenSSL: some compilers
9209 (gcc 4.2 and later) reject their use.
9210 *Kurt Roeckx <kurt@roeckx.be>, Peter Hartley <pdh@utter.chaos.org.uk>,
9213 * Add RFC4507 support to OpenSSL. This includes the corrections in
9214 RFC4507bis. The encrypted ticket format is an encrypted encoded
9215 SSL_SESSION structure, that way new session features are automatically
9218 If a client application caches session in an SSL_SESSION structure
9219 support is transparent because tickets are now stored in the encoded
9222 The SSL_CTX structure automatically generates keys for ticket
9223 protection in servers so again support should be possible
9224 with no application modification.
9226 If a client or server wishes to disable RFC4507 support then the option
9227 SSL_OP_NO_TICKET can be set.
9229 Add a TLS extension debugging callback to allow the contents of any client
9230 or server extensions to be examined.
9232 This work was sponsored by Google.
9236 * Add initial support for TLS extensions, specifically for the server_name
9237 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
9238 have new members for a host name. The SSL data structure has an
9239 additional member `SSL_CTX *initial_ctx` so that new sessions can be
9240 stored in that context to allow for session resumption, even after the
9241 SSL has been switched to a new SSL_CTX in reaction to a client's
9242 server_name extension.
9244 New functions (subject to change):
9246 SSL_get_servername()
9247 SSL_get_servername_type()
9250 New CTRL codes and macros (subject to change):
9252 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
9253 - SSL_CTX_set_tlsext_servername_callback()
9254 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
9255 - SSL_CTX_set_tlsext_servername_arg()
9256 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
9258 openssl s_client has a new '-servername ...' option.
9260 openssl s_server has new options '-servername_host ...', '-cert2 ...',
9261 '-key2 ...', '-servername_fatal' (subject to change). This allows
9262 testing the HostName extension for a specific single host name ('-cert'
9263 and '-key' remain fallbacks for handshakes without HostName
9264 negotiation). If the unrecognized_name alert has to be sent, this by
9265 default is a warning; it becomes fatal with the '-servername_fatal'
9268 *Peter Sylvester, Remy Allais, Christophe Renou, Steve Henson*
9270 * Add AES and SSE2 assembly language support to VC++ build.
9274 * Mitigate attack on final subtraction in Montgomery reduction.
9278 * Fix crypto/ec/ec_mult.c to work properly with scalars of value 0
9279 (which previously caused an internal error).
9283 * Squeeze another 10% out of IGE mode when in != out.
9287 * AES IGE mode speedup.
9289 *Dean Gaudet (Google)*
9291 * Add the Korean symmetric 128-bit cipher SEED (see
9292 <http://www.kisa.or.kr/kisa/seed/jsp/seed_eng.jsp>) and
9293 add SEED ciphersuites from RFC 4162:
9295 TLS_RSA_WITH_SEED_CBC_SHA = "SEED-SHA"
9296 TLS_DHE_DSS_WITH_SEED_CBC_SHA = "DHE-DSS-SEED-SHA"
9297 TLS_DHE_RSA_WITH_SEED_CBC_SHA = "DHE-RSA-SEED-SHA"
9298 TLS_DH_anon_WITH_SEED_CBC_SHA = "ADH-SEED-SHA"
9300 To minimize changes between patchlevels in the OpenSSL 0.9.8
9301 series, SEED remains excluded from compilation unless OpenSSL
9302 is configured with 'enable-seed'.
9304 *KISA, Bodo Moeller*
9306 * Mitigate branch prediction attacks, which can be practical if a
9307 single processor is shared, allowing a spy process to extract
9308 information. For detailed background information, see
9309 <http://eprint.iacr.org/2007/039> (O. Aciicmez, S. Gueron,
9310 J.-P. Seifert, "New Branch Prediction Vulnerabilities in OpenSSL
9311 and Necessary Software Countermeasures"). The core of the change
9312 are new versions BN_div_no_branch() and
9313 BN_mod_inverse_no_branch() of BN_div() and BN_mod_inverse(),
9314 respectively, which are slower, but avoid the security-relevant
9315 conditional branches. These are automatically called by BN_div()
9316 and BN_mod_inverse() if the flag BN_FLG_CONSTTIME is set for one
9317 of the input BIGNUMs. Also, BN_is_bit_set() has been changed to
9318 remove a conditional branch.
9320 BN_FLG_CONSTTIME is the new name for the previous
9321 BN_FLG_EXP_CONSTTIME flag, since it now affects more than just
9322 modular exponentiation. (Since OpenSSL 0.9.7h, setting this flag
9323 in the exponent causes BN_mod_exp_mont() to use the alternative
9324 implementation in BN_mod_exp_mont_consttime().) The old name
9325 remains as a deprecated alias.
9327 Similarly, RSA_FLAG_NO_EXP_CONSTTIME is replaced by a more general
9328 RSA_FLAG_NO_CONSTTIME flag since the RSA implementation now uses
9329 constant-time implementations for more than just exponentiation.
9330 Here too the old name is kept as a deprecated alias.
9332 BN_BLINDING_new() will now use BN_dup() for the modulus so that
9333 the BN_BLINDING structure gets an independent copy of the
9334 modulus. This means that the previous `BIGNUM *m` argument to
9335 BN_BLINDING_new() and to BN_BLINDING_create_param() now
9336 essentially becomes `const BIGNUM *m`, although we can't actually
9337 change this in the header file before 0.9.9. It allows
9338 RSA_setup_blinding() to use BN_with_flags() on the modulus to
9339 enable BN_FLG_CONSTTIME.
9341 *Matthew D Wood (Intel Corp)*
9343 * In the SSL/TLS server implementation, be strict about session ID
9344 context matching (which matters if an application uses a single
9345 external cache for different purposes). Previously,
9346 out-of-context reuse was forbidden only if SSL_VERIFY_PEER was
9347 set. This did ensure strict client verification, but meant that,
9348 with applications using a single external cache for quite
9349 different requirements, clients could circumvent ciphersuite
9350 restrictions for a given session ID context by starting a session
9351 in a different context.
9355 * Include "!eNULL" in SSL_DEFAULT_CIPHER_LIST to make sure that
9356 a ciphersuite string such as "DEFAULT:RSA" cannot enable
9357 authentication-only ciphersuites.
9361 * Update the SSL_get_shared_ciphers() fix CVE-2006-3738 which was
9362 not complete and could lead to a possible single byte overflow
9363 [CVE-2007-5135][] [Ben Laurie]
9365 ### Changes between 0.9.8d and 0.9.8e [23 Feb 2007]
9367 * Since AES128 and AES256 (and similarly Camellia128 and
9368 Camellia256) share a single mask bit in the logic of
9369 ssl/ssl_ciph.c, the code for masking out disabled ciphers needs a
9370 kludge to work properly if AES128 is available and AES256 isn't
9371 (or if Camellia128 is available and Camellia256 isn't).
9375 * Fix the BIT STRING encoding generated by crypto/ec/ec_asn1.c
9376 (within i2d_ECPrivateKey, i2d_ECPKParameters, i2d_ECParameters):
9377 When a point or a seed is encoded in a BIT STRING, we need to
9378 prevent the removal of trailing zero bits to get the proper DER
9379 encoding. (By default, crypto/asn1/a_bitstr.c assumes the case
9380 of a NamedBitList, for which trailing 0 bits need to be removed.)
9384 * Have SSL/TLS server implementation tolerate "mismatched" record
9385 protocol version while receiving ClientHello even if the
9386 ClientHello is fragmented. (The server can't insist on the
9387 particular protocol version it has chosen before the ServerHello
9388 message has informed the client about his choice.)
9392 * Add RFC 3779 support.
9394 *Rob Austein for ARIN, Ben Laurie*
9396 * Load error codes if they are not already present instead of using a
9397 static variable. This allows them to be cleanly unloaded and reloaded.
9398 Improve header file function name parsing.
9402 * extend SMTP and IMAP protocol emulation in s_client to use EHLO
9403 or CAPABILITY handshake as required by RFCs.
9407 ### Changes between 0.9.8c and 0.9.8d [28 Sep 2006]
9409 * Introduce limits to prevent malicious keys being able to
9410 cause a denial of service. [CVE-2006-2940][]
9412 *Steve Henson, Bodo Moeller*
9414 * Fix ASN.1 parsing of certain invalid structures that can result
9415 in a denial of service. [CVE-2006-2937][] [Steve Henson]
9417 * Fix buffer overflow in SSL_get_shared_ciphers() function.
9418 [CVE-2006-3738][] [Tavis Ormandy and Will Drewry, Google Security Team]
9420 * Fix SSL client code which could crash if connecting to a
9421 malicious SSLv2 server. [CVE-2006-4343][]
9423 *Tavis Ormandy and Will Drewry, Google Security Team*
9425 * Since 0.9.8b, ciphersuite strings naming explicit ciphersuites
9426 match only those. Before that, "AES256-SHA" would be interpreted
9427 as a pattern and match "AES128-SHA" too (since AES128-SHA got
9428 the same strength classification in 0.9.7h) as we currently only
9429 have a single AES bit in the ciphersuite description bitmap.
9430 That change, however, also applied to ciphersuite strings such as
9431 "RC4-MD5" that intentionally matched multiple ciphersuites --
9432 namely, SSL 2.0 ciphersuites in addition to the more common ones
9433 from SSL 3.0/TLS 1.0.
9435 So we change the selection algorithm again: Naming an explicit
9436 ciphersuite selects this one ciphersuite, and any other similar
9437 ciphersuite (same bitmap) from *other* protocol versions.
9438 Thus, "RC4-MD5" again will properly select both the SSL 2.0
9439 ciphersuite and the SSL 3.0/TLS 1.0 ciphersuite.
9441 Since SSL 2.0 does not have any ciphersuites for which the
9442 128/256 bit distinction would be relevant, this works for now.
9443 The proper fix will be to use different bits for AES128 and
9444 AES256, which would have avoided the problems from the beginning;
9445 however, bits are scarce, so we can only do this in a new release
9446 (not just a patchlevel) when we can change the SSL_CIPHER
9447 definition to split the single 'unsigned long mask' bitmap into
9448 multiple values to extend the available space.
9452 ### Changes between 0.9.8b and 0.9.8c [05 Sep 2006]
9454 * Avoid PKCS #1 v1.5 signature attack discovered by Daniel Bleichenbacher
9455 [CVE-2006-4339][] [Ben Laurie and Google Security Team]
9457 * Add AES IGE and biIGE modes.
9461 * Change the Unix randomness entropy gathering to use poll() when
9462 possible instead of select(), since the latter has some
9463 undesirable limitations.
9465 *Darryl Miles via Richard Levitte and Bodo Moeller*
9467 * Disable "ECCdraft" ciphersuites more thoroughly. Now special
9468 treatment in ssl/ssl_ciph.s makes sure that these ciphersuites
9469 cannot be implicitly activated as part of, e.g., the "AES" alias.
9470 However, please upgrade to OpenSSL 0.9.9[-dev] for
9471 non-experimental use of the ECC ciphersuites to get TLS extension
9472 support, which is required for curve and point format negotiation
9473 to avoid potential handshake problems.
9477 * Disable rogue ciphersuites:
9479 - SSLv2 0x08 0x00 0x80 ("RC4-64-MD5")
9480 - SSLv3/TLSv1 0x00 0x61 ("EXP1024-RC2-CBC-MD5")
9481 - SSLv3/TLSv1 0x00 0x60 ("EXP1024-RC4-MD5")
9483 The latter two were purportedly from
9484 draft-ietf-tls-56-bit-ciphersuites-0[01].txt, but do not really
9487 Also deactivate the remaining ciphersuites from
9488 draft-ietf-tls-56-bit-ciphersuites-01.txt. These are just as
9489 unofficial, and the ID has long expired.
9493 * Fix RSA blinding Heisenbug (problems sometimes occurred on
9494 dual-core machines) and other potential thread-safety issues.
9498 * Add the symmetric cipher Camellia (128-bit, 192-bit, 256-bit key
9499 versions), which is now available for royalty-free use
9500 (see <http://info.isl.ntt.co.jp/crypt/eng/info/chiteki.html>).
9501 Also, add Camellia TLS ciphersuites from RFC 4132.
9503 To minimize changes between patchlevels in the OpenSSL 0.9.8
9504 series, Camellia remains excluded from compilation unless OpenSSL
9505 is configured with 'enable-camellia'.
9509 * Disable the padding bug check when compression is in use. The padding
9510 bug check assumes the first packet is of even length, this is not
9511 necessarily true if compression is enabled and can result in false
9512 positives causing handshake failure. The actual bug test is ancient
9513 code so it is hoped that implementations will either have fixed it by
9514 now or any which still have the bug do not support compression.
9518 ### Changes between 0.9.8a and 0.9.8b [04 May 2006]
9520 * When applying a cipher rule check to see if string match is an explicit
9521 cipher suite and only match that one cipher suite if it is.
9525 * Link in manifests for VC++ if needed.
9527 *Austin Ziegler <halostatue@gmail.com>*
9529 * Update support for ECC-based TLS ciphersuites according to
9530 draft-ietf-tls-ecc-12.txt with proposed changes (but without
9531 TLS extensions, which are supported starting with the 0.9.9
9532 branch, not in the OpenSSL 0.9.8 branch).
9536 * New functions EVP_CIPHER_CTX_new() and EVP_CIPHER_CTX_free() to support
9537 opaque EVP_CIPHER_CTX handling.
9541 * Fixes and enhancements to zlib compression code. We now only use
9542 "zlib1.dll" and use the default `__cdecl` calling convention on Win32
9543 to conform with the standards mentioned here:
9544 <http://www.zlib.net/DLL_FAQ.txt>
9545 Static zlib linking now works on Windows and the new --with-zlib-include
9546 --with-zlib-lib options to Configure can be used to supply the location
9547 of the headers and library. Gracefully handle case where zlib library
9552 * Several fixes and enhancements to the OID generation code. The old code
9553 sometimes allowed invalid OIDs (1.X for X >= 40 for example), couldn't
9554 handle numbers larger than ULONG_MAX, truncated printing and had a
9555 non standard OBJ_obj2txt() behaviour.
9559 * Add support for building of engines under engine/ as shared libraries
9560 under VC++ build system.
9564 * Corrected the numerous bugs in the Win32 path splitter in DSO.
9565 Hopefully, we will not see any false combination of paths any more.
9569 ### Changes between 0.9.8 and 0.9.8a [11 Oct 2005]
9571 * Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING
9572 (part of SSL_OP_ALL). This option used to disable the
9573 countermeasure against man-in-the-middle protocol-version
9574 rollback in the SSL 2.0 server implementation, which is a bad
9575 idea. [CVE-2005-2969][]
9577 *Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center
9578 for Information Security, National Institute of Advanced Industrial
9579 Science and Technology [AIST], Japan)*
9581 * Add two function to clear and return the verify parameter flags.
9585 * Keep cipherlists sorted in the source instead of sorting them at
9586 runtime, thus removing the need for a lock.
9590 * Avoid some small subgroup attacks in Diffie-Hellman.
9592 *Nick Mathewson and Ben Laurie*
9594 * Add functions for well-known primes.
9598 * Extended Windows CE support.
9600 *Satoshi Nakamura and Andy Polyakov*
9602 * Initialize SSL_METHOD structures at compile time instead of during
9603 runtime, thus removing the need for a lock.
9607 * Make PKCS7_decrypt() work even if no certificate is supplied by
9608 attempting to decrypt each encrypted key in turn. Add support to
9613 ### Changes between 0.9.7h and 0.9.8 [05 Jul 2005]
9615 [NB: OpenSSL 0.9.7i and later 0.9.7 patch levels were released after
9618 * Add libcrypto.pc and libssl.pc for those who feel they need them.
9622 * Change CA.sh and CA.pl so they don't bundle the CSR and the private
9623 key into the same file any more.
9627 * Add initial support for Win64, both IA64 and AMD64/x64 flavors.
9631 * Add -utf8 command line and config file option to 'ca'.
9633 *Stefan <stf@udoma.org*
9635 * Removed the macro des_crypt(), as it seems to conflict with some
9636 libraries. Use DES_crypt().
9640 * Correct naming of the 'chil' and '4758cca' ENGINEs. This
9641 involves renaming the source and generated shared-libs for
9642 both. The engines will accept the corrected or legacy ids
9643 ('ncipher' and '4758_cca' respectively) when binding. NB,
9644 this only applies when building 'shared'.
9646 *Corinna Vinschen <vinschen@redhat.com> and Geoff Thorpe*
9648 * Add attribute functions to EVP_PKEY structure. Modify
9649 PKCS12_create() to recognize a CSP name attribute and
9650 use it. Make -CSP option work again in pkcs12 utility.
9654 * Add new functionality to the bn blinding code:
9655 - automatic re-creation of the BN_BLINDING parameters after
9656 a fixed number of uses (currently 32)
9657 - add new function for parameter creation
9658 - introduce flags to control the update behaviour of the
9659 BN_BLINDING parameters
9660 - hide BN_BLINDING structure
9661 Add a second BN_BLINDING slot to the RSA structure to improve
9662 performance when a single RSA object is shared among several
9667 * Add support for DTLS.
9669 *Nagendra Modadugu <nagendra@cs.stanford.edu> and Ben Laurie*
9671 * Add support for DER encoded private keys (SSL_FILETYPE_ASN1)
9672 to SSL_CTX_use_PrivateKey_file() and SSL_use_PrivateKey_file()
9676 * Remove buggy and incomplete DH cert support from
9677 ssl/ssl_rsa.c and ssl/s3_both.c
9681 * Use SHA-1 instead of MD5 as the default digest algorithm for
9682 the apps/openssl applications.
9686 * Compile clean with "-Wall -Wmissing-prototypes
9687 -Wstrict-prototypes -Wmissing-declarations -Werror". Currently
9688 DEBUG_SAFESTACK must also be set.
9692 * Change ./Configure so that certain algorithms can be disabled by default.
9693 The new counterpiece to "no-xxx" is "enable-xxx".
9695 The patented RC5 and MDC2 algorithms will now be disabled unless
9696 "enable-rc5" and "enable-mdc2", respectively, are specified.
9698 (IDEA remains enabled despite being patented. This is because IDEA
9699 is frequently required for interoperability, and there is no license
9700 fee for non-commercial use. As before, "no-idea" can be used to
9701 avoid this algorithm.)
9705 * Add processing of proxy certificates (see RFC 3820). This work was
9706 sponsored by KTH (The Royal Institute of Technology in Stockholm) and
9707 EGEE (Enabling Grids for E-science in Europe).
9711 * RC4 performance overhaul on modern architectures/implementations, such
9712 as Intel P4, IA-64 and AMD64.
9716 * New utility extract-section.pl. This can be used specify an alternative
9717 section number in a pod file instead of having to treat each file as
9718 a separate case in Makefile. This can be done by adding two lines to the
9721 =for comment openssl_section:XXX
9723 The blank line is mandatory.
9727 * New arguments -certform, -keyform and -pass for s_client and s_server
9728 to allow alternative format key and certificate files and passphrase
9733 * New structure X509_VERIFY_PARAM which combines current verify parameters,
9734 update associated structures and add various utility functions.
9736 Add new policy related verify parameters, include policy checking in
9737 standard verify code. Enhance 'smime' application with extra parameters
9738 to support policy checking and print out.
9742 * Add a new engine to support VIA PadLock ACE extensions in the VIA C3
9743 Nehemiah processors. These extensions support AES encryption in hardware
9744 as well as RNG (though RNG support is currently disabled).
9746 *Michal Ludvig <michal@logix.cz>, with help from Andy Polyakov*
9748 * Deprecate `BN_[get|set]_params()` functions (they were ignored internally).
9752 * New FIPS 180-2 algorithms, SHA-224/-256/-384/-512 are implemented.
9754 *Andy Polyakov and a number of other people*
9756 * Improved PowerPC platform support. Most notably BIGNUM assembler
9757 implementation contributed by IBM.
9759 *Suresh Chari, Peter Waltenberg, Andy Polyakov*
9761 * The new 'RSA_generate_key_ex' function now takes a BIGNUM for the public
9762 exponent rather than 'unsigned long'. There is a corresponding change to
9763 the new 'rsa_keygen' element of the RSA_METHOD structure.
9765 *Jelte Jansen, Geoff Thorpe*
9767 * Functionality for creating the initial serial number file is now
9768 moved from CA.pl to the 'ca' utility with a new option -create_serial.
9770 (Before OpenSSL 0.9.7e, CA.pl used to initialize the serial
9771 number file to 1, which is bound to cause problems. To avoid
9772 the problems while respecting compatibility between different 0.9.7
9773 patchlevels, 0.9.7e employed 'openssl x509 -next_serial' in
9774 CA.pl for serial number initialization. With the new release 0.9.8,
9775 we can fix the problem directly in the 'ca' utility.)
9779 * Reduced header interdependencies by declaring more opaque objects in
9780 ossl_typ.h. As a consequence, including some headers (eg. engine.h) will
9781 give fewer recursive includes, which could break lazy source code - so
9782 this change is covered by the OPENSSL_NO_DEPRECATED symbol. As always,
9783 developers should define this symbol when building and using openssl to
9784 ensure they track the recommended behaviour, interfaces, [etc], but
9785 backwards-compatible behaviour prevails when this isn't defined.
9789 * New function X509_POLICY_NODE_print() which prints out policy nodes.
9793 * Add new EVP function EVP_CIPHER_CTX_rand_key and associated functionality.
9794 This will generate a random key of the appropriate length based on the
9795 cipher context. The EVP_CIPHER can provide its own random key generation
9796 routine to support keys of a specific form. This is used in the des and
9797 3des routines to generate a key of the correct parity. Update S/MIME
9798 code to use new functions and hence generate correct parity DES keys.
9799 Add EVP_CHECK_DES_KEY #define to return an error if the key is not
9800 valid (weak or incorrect parity).
9804 * Add a local set of CRLs that can be used by X509_verify_cert() as well
9805 as looking them up. This is useful when the verified structure may contain
9806 CRLs, for example PKCS#7 signedData. Modify PKCS7_verify() to use any CRLs
9807 present unless the new PKCS7_NO_CRL flag is asserted.
9811 * Extend ASN1 oid configuration module. It now additionally accepts the
9814 shortName = some long name, 1.2.3.4
9818 * Reimplemented the BN_CTX implementation. There is now no more static
9819 limitation on the number of variables it can handle nor the depth of the
9820 "stack" handling for BN_CTX_start()/BN_CTX_end() pairs. The stack
9821 information can now expand as required, and rather than having a single
9822 static array of bignums, BN_CTX now uses a linked-list of such arrays
9823 allowing it to expand on demand whilst maintaining the usefulness of
9824 BN_CTX's "bundling".
9828 * Add a missing BN_CTX parameter to the 'rsa_mod_exp' callback in RSA_METHOD
9829 to allow all RSA operations to function using a single BN_CTX.
9833 * Preliminary support for certificate policy evaluation and checking. This
9834 is initially intended to pass the tests outlined in "Conformance Testing
9835 of Relying Party Client Certificate Path Processing Logic" v1.07.
9839 * bn_dup_expand() has been deprecated, it was introduced in 0.9.7 and
9840 remained unused and not that useful. A variety of other little bignum
9841 tweaks and fixes have also been made continuing on from the audit (see
9846 * Constify all or almost all d2i, c2i, s2i and r2i functions, along with
9847 associated ASN1, EVP and SSL functions and old ASN1 macros.
9851 * BN_zero() only needs to set 'top' and 'neg' to zero for correct results,
9852 and this should never fail. So the return value from the use of
9853 BN_set_word() (which can fail due to needless expansion) is now deprecated;
9854 if OPENSSL_NO_DEPRECATED is defined, BN_zero() is a void macro.
9858 * BN_CTX_get() should return zero-valued bignums, providing the same
9859 initialised value as BN_new().
9861 *Geoff Thorpe, suggested by Ulf Möller*
9863 * Support for inhibitAnyPolicy certificate extension.
9867 * An audit of the BIGNUM code is underway, for which debugging code is
9868 enabled when BN_DEBUG is defined. This makes stricter enforcements on what
9869 is considered valid when processing BIGNUMs, and causes execution to
9870 assert() when a problem is discovered. If BN_DEBUG_RAND is defined,
9871 further steps are taken to deliberately pollute unused data in BIGNUM
9872 structures to try and expose faulty code further on. For now, openssl will
9873 (in its default mode of operation) continue to tolerate the inconsistent
9874 forms that it has tolerated in the past, but authors and packagers should
9875 consider trying openssl and their own applications when compiled with
9876 these debugging symbols defined. It will help highlight potential bugs in
9877 their own code, and will improve the test coverage for OpenSSL itself. At
9878 some point, these tighter rules will become openssl's default to improve
9879 maintainability, though the assert()s and other overheads will remain only
9880 in debugging configurations. See bn.h for more details.
9882 *Geoff Thorpe, Nils Larsch, Ulf Möller*
9884 * BN_CTX_init() has been deprecated, as BN_CTX is an opaque structure
9885 that can only be obtained through BN_CTX_new() (which implicitly
9886 initialises it). The presence of this function only made it possible
9887 to overwrite an existing structure (and cause memory leaks).
9891 * Because of the callback-based approach for implementing LHASH as a
9892 template type, lh_insert() adds opaque objects to hash-tables and
9893 lh_doall() or lh_doall_arg() are typically used with a destructor callback
9894 to clean up those corresponding objects before destroying the hash table
9895 (and losing the object pointers). So some over-zealous constifications in
9896 LHASH have been relaxed so that lh_insert() does not take (nor store) the
9897 objects as "const" and the `lh_doall[_arg]` callback wrappers are not
9898 prototyped to have "const" restrictions on the object pointers they are
9899 given (and so aren't required to cast them away any more).
9903 * The tmdiff.h API was so ugly and minimal that our own timing utility
9904 (speed) prefers to use its own implementation. The two implementations
9905 haven't been consolidated as yet (volunteers?) but the tmdiff API has had
9906 its object type properly exposed (MS_TM) instead of casting to/from
9907 `char *`. This may still change yet if someone realises MS_TM and
9909 aren't necessarily the greatest nomenclatures - but this is what was used
9910 internally to the implementation so I've used that for now.
9914 * Ensure that deprecated functions do not get compiled when
9915 OPENSSL_NO_DEPRECATED is defined. Some "openssl" subcommands and a few of
9916 the self-tests were still using deprecated key-generation functions so
9917 these have been updated also.
9921 * Reorganise PKCS#7 code to separate the digest location functionality
9922 into PKCS7_find_digest(), digest addition into PKCS7_bio_add_digest().
9923 New function PKCS7_set_digest() to set the digest type for PKCS#7
9924 digestedData type. Add additional code to correctly generate the
9925 digestedData type and add support for this type in PKCS7 initialization
9930 * New function PKCS7_set0_type_other() this initializes a PKCS7
9931 structure of type "other".
9935 * Fix prime generation loop in crypto/bn/bn_prime.pl by making
9936 sure the loop does correctly stop and breaking ("division by zero")
9937 modulus operations are not performed. The (pre-generated) prime
9938 table crypto/bn/bn_prime.h was already correct, but it could not be
9939 re-generated on some platforms because of the "division by zero"
9940 situation in the script.
9942 *Ralf S. Engelschall*
9944 * Update support for ECC-based TLS ciphersuites according to
9945 draft-ietf-tls-ecc-03.txt: the KDF1 key derivation function with
9946 SHA-1 now is only used for "small" curves (where the
9947 representation of a field element takes up to 24 bytes); for
9948 larger curves, the field element resulting from ECDH is directly
9949 used as premaster secret.
9951 *Douglas Stebila (Sun Microsystems Laboratories)*
9953 * Add code for kP+lQ timings to crypto/ec/ectest.c, and add SEC2
9954 curve secp160r1 to the tests.
9956 *Douglas Stebila (Sun Microsystems Laboratories)*
9958 * Add the possibility to load symbols globally with DSO.
9960 *Götz Babin-Ebell <babin-ebell@trustcenter.de> via Richard Levitte*
9962 * Add the functions ERR_set_mark() and ERR_pop_to_mark() for better
9963 control of the error stack.
9967 * Add support for STORE in ENGINE.
9971 * Add the STORE type. The intention is to provide a common interface
9972 to certificate and key stores, be they simple file-based stores, or
9973 HSM-type store, or LDAP stores, or...
9974 NOTE: The code is currently UNTESTED and isn't really used anywhere.
9978 * Add a generic structure called OPENSSL_ITEM. This can be used to
9979 pass a list of arguments to any function as well as provide a way
9980 for a function to pass data back to the caller.
9984 * Add the functions BUF_strndup() and BUF_memdup(). BUF_strndup()
9985 works like BUF_strdup() but can be used to duplicate a portion of
9986 a string. The copy gets NUL-terminated. BUF_memdup() duplicates
9991 * Add the function sk_find_ex() which works like sk_find(), but will
9992 return an index to an element even if an exact match couldn't be
9993 found. The index is guaranteed to point at the element where the
9994 searched-for key would be inserted to preserve sorting order.
9998 * Add the function OBJ_bsearch_ex() which works like OBJ_bsearch() but
9999 takes an extra flags argument for optional functionality. Currently,
10000 the following flags are defined:
10002 OBJ_BSEARCH_VALUE_ON_NOMATCH
10003 This one gets OBJ_bsearch_ex() to return a pointer to the first
10004 element where the comparing function returns a negative or zero
10007 OBJ_BSEARCH_FIRST_VALUE_ON_MATCH
10008 This one gets OBJ_bsearch_ex() to return a pointer to the first
10009 element where the comparing function returns zero. This is useful
10010 if there are more than one element where the comparing function
10015 * Make it possible to create self-signed certificates with 'openssl ca'
10016 in such a way that the self-signed certificate becomes part of the
10017 CA database and uses the same mechanisms for serial number generation
10018 as all other certificate signing. The new flag '-selfsign' enables
10019 this functionality. Adapt CA.sh and CA.pl.in.
10023 * Add functionality to check the public key of a certificate request
10024 against a given private. This is useful to check that a certificate
10025 request can be signed by that key (self-signing).
10029 * Make it possible to have multiple active certificates with the same
10030 subject in the CA index file. This is done only if the keyword
10031 'unique_subject' is set to 'no' in the main CA section (default
10032 if 'CA_default') of the configuration file. The value is saved
10033 with the database itself in a separate index attribute file,
10034 named like the index file with '.attr' appended to the name.
10038 * Generate multi-valued AVAs using '+' notation in config files for
10043 * Support for nameConstraints certificate extension.
10047 * Support for policyConstraints certificate extension.
10051 * Support for policyMappings certificate extension.
10055 * Make sure the default DSA_METHOD implementation only uses its
10056 dsa_mod_exp() and/or bn_mod_exp() handlers if they are non-NULL,
10057 and change its own handlers to be NULL so as to remove unnecessary
10058 indirection. This lets alternative implementations fallback to the
10059 default implementation more easily.
10063 * Support for directoryName in GeneralName related extensions
10068 * Make it possible to link applications using Makefile.shared.
10069 Make that possible even when linking against static libraries!
10073 * Support for single pass processing for S/MIME signing. This now
10074 means that S/MIME signing can be done from a pipe, in addition
10075 cleartext signing (multipart/signed type) is effectively streaming
10076 and the signed data does not need to be all held in memory.
10078 This is done with a new flag PKCS7_STREAM. When this flag is set
10079 PKCS7_sign() only initializes the PKCS7 structure and the actual signing
10080 is done after the data is output (and digests calculated) in
10081 SMIME_write_PKCS7().
10085 * Add full support for -rpath/-R, both in shared libraries and
10086 applications, at least on the platforms where it's known how
10091 * In crypto/ec/ec_mult.c, implement fast point multiplication with
10092 precomputation, based on wNAF splitting: EC_GROUP_precompute_mult()
10093 will now compute a table of multiples of the generator that
10094 makes subsequent invocations of EC_POINTs_mul() or EC_POINT_mul()
10095 faster (notably in the case of a single point multiplication,
10096 scalar * generator).
10098 *Nils Larsch, Bodo Moeller*
10100 * IPv6 support for certificate extensions. The various extensions
10101 which use the IP:a.b.c.d can now take IPv6 addresses using the
10102 formats of RFC1884 2.2 . IPv6 addresses are now also displayed
10107 * Added an ENGINE that implements RSA by performing private key
10108 exponentiations with the GMP library. The conversions to and from
10109 GMP's mpz_t format aren't optimised nor are any montgomery forms
10110 cached, and on x86 it appears OpenSSL's own performance has caught up.
10111 However there are likely to be other architectures where GMP could
10112 provide a boost. This ENGINE is not built in by default, but it can be
10113 specified at Configure time and should be accompanied by the necessary
10114 linker additions, eg;
10115 ./config -DOPENSSL_USE_GMP -lgmp
10119 * "openssl engine" will not display ENGINE/DSO load failure errors when
10120 testing availability of engines with "-t" - the old behaviour is
10121 produced by increasing the feature's verbosity with "-tt".
10125 * ECDSA routines: under certain error conditions uninitialized BN objects
10126 could be freed. Solution: make sure initialization is performed early
10127 enough. (Reported and fix supplied by Nils Larsch <nla@trustcenter.de>
10132 * Key-generation can now be implemented in RSA_METHOD, DSA_METHOD
10133 and DH_METHOD (eg. by ENGINE implementations) to override the normal
10134 software implementations. For DSA and DH, parameter generation can
10135 also be overridden by providing the appropriate method callbacks.
10139 * Change the "progress" mechanism used in key-generation and
10140 primality testing to functions that take a new BN_GENCB pointer in
10141 place of callback/argument pairs. The new API functions have `_ex`
10142 postfixes and the older functions are reimplemented as wrappers for
10143 the new ones. The OPENSSL_NO_DEPRECATED symbol can be used to hide
10144 declarations of the old functions to help (graceful) attempts to
10145 migrate to the new functions. Also, the new key-generation API
10146 functions operate on a caller-supplied key-structure and return
10147 success/failure rather than returning a key or NULL - this is to
10148 help make "keygen" another member function of RSA_METHOD etc.
10150 Example for using the new callback interface:
10152 int (*my_callback)(int a, int b, BN_GENCB *cb) = ...;
10153 void *my_arg = ...;
10156 BN_GENCB_set(&my_cb, my_callback, my_arg);
10158 return BN_is_prime_ex(some_bignum, BN_prime_checks, NULL, &cb);
10159 /* For the meaning of a, b in calls to my_callback(), see the
10160 * documentation of the function that calls the callback.
10161 * cb will point to my_cb; my_arg can be retrieved as cb->arg.
10162 * my_callback should return 1 if it wants BN_is_prime_ex()
10163 * to continue, or 0 to stop.
10168 * Change the ZLIB compression method to be stateful, and make it
10169 available to TLS with the number defined in
10170 draft-ietf-tls-compression-04.txt.
10174 * Add the ASN.1 structures and functions for CertificatePair, which
10175 is defined as follows (according to X.509_4thEditionDraftV6.pdf):
10177 CertificatePair ::= SEQUENCE {
10178 forward [0] Certificate OPTIONAL,
10179 reverse [1] Certificate OPTIONAL,
10180 -- at least one of the pair shall be present -- }
10182 Also implement the PEM functions to read and write certificate
10183 pairs, and defined the PEM tag as "CERTIFICATE PAIR".
10185 This needed to be defined, mostly for the sake of the LDAP
10186 attribute crossCertificatePair, but may prove useful elsewhere as
10191 * Make it possible to inhibit symlinking of shared libraries in
10192 Makefile.shared, for Cygwin's sake.
10196 * Extend the BIGNUM API by creating a function
10197 void BN_set_negative(BIGNUM *a, int neg);
10198 and a macro that behave like
10199 int BN_is_negative(const BIGNUM *a);
10201 to avoid the need to access 'a->neg' directly in applications.
10205 * Implement fast modular reduction for pseudo-Mersenne primes
10206 used in NIST curves (crypto/bn/bn_nist.c, crypto/ec/ecp_nist.c).
10207 EC_GROUP_new_curve_GFp() will now automatically use this
10210 *Nils Larsch <nla@trustcenter.de>*
10212 * Add new lock type (CRYPTO_LOCK_BN).
10216 * Change the ENGINE framework to automatically load engines
10217 dynamically from specific directories unless they could be
10218 found to already be built in or loaded. Move all the
10219 current engines except for the cryptodev one to a new
10220 directory engines/.
10221 The engines in engines/ are built as shared libraries if
10222 the "shared" options was given to ./Configure or ./config.
10223 Otherwise, they are inserted in libcrypto.a.
10224 /usr/local/ssl/engines is the default directory for dynamic
10225 engines, but that can be overridden at configure time through
10226 the usual use of --prefix and/or --openssldir, and at run
10227 time with the environment variable OPENSSL_ENGINES.
10229 *Geoff Thorpe and Richard Levitte*
10231 * Add Makefile.shared, a helper makefile to build shared
10232 libraries. Adapt Makefile.org.
10236 * Add version info to Win32 DLLs.
10238 *Peter 'Luna' Runestig" <peter@runestig.com>*
10240 * Add new 'medium level' PKCS#12 API. Certificates and keys
10241 can be added using this API to created arbitrary PKCS#12
10242 files while avoiding the low level API.
10244 New options to PKCS12_create(), key or cert can be NULL and
10245 will then be omitted from the output file. The encryption
10246 algorithm NIDs can be set to -1 for no encryption, the mac
10247 iteration count can be set to 0 to omit the mac.
10249 Enhance pkcs12 utility by making the -nokeys and -nocerts
10250 options work when creating a PKCS#12 file. New option -nomac
10251 to omit the mac, NONE can be set for an encryption algorithm.
10252 New code is modified to use the enhanced PKCS12_create()
10253 instead of the low level API.
10257 * Extend ASN1 encoder to support indefinite length constructed
10258 encoding. This can output sequences tags and octet strings in
10259 this form. Modify pk7_asn1.c to support indefinite length
10260 encoding. This is experimental and needs additional code to
10261 be useful, such as an ASN1 bio and some enhanced streaming
10264 Extend template encode functionality so that tagging is passed
10265 down to the template encoder.
10269 * Let 'openssl req' fail if an argument to '-newkey' is not
10270 recognized instead of using RSA as a default.
10274 * Add support for ECC-based ciphersuites from draft-ietf-tls-ecc-01.txt.
10275 As these are not official, they are not included in "ALL";
10276 the "ECCdraft" ciphersuite group alias can be used to select them.
10278 *Vipul Gupta and Sumit Gupta (Sun Microsystems Laboratories)*
10280 * Add ECDH engine support.
10282 *Nils Gura and Douglas Stebila (Sun Microsystems Laboratories)*
10284 * Add ECDH in new directory crypto/ecdh/.
10286 *Douglas Stebila (Sun Microsystems Laboratories)*
10288 * Let BN_rand_range() abort with an error after 100 iterations
10289 without success (which indicates a broken PRNG).
10293 * Change BN_mod_sqrt() so that it verifies that the input value
10294 is really the square of the return value. (Previously,
10295 BN_mod_sqrt would show GIGO behaviour.)
10299 * Add named elliptic curves over binary fields from X9.62, SECG,
10300 and WAP/WTLS; add OIDs that were still missing.
10302 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10304 * Extend the EC library for elliptic curves over binary fields
10305 (new files ec2_smpl.c, ec2_smpt.c, ec2_mult.c in crypto/ec/).
10308 EC_GF2m_simple_method
10312 EC_GROUP_new_curve_GF2m
10313 EC_GROUP_set_curve_GF2m
10314 EC_GROUP_get_curve_GF2m
10315 EC_POINT_set_affine_coordinates_GF2m
10316 EC_POINT_get_affine_coordinates_GF2m
10317 EC_POINT_set_compressed_coordinates_GF2m
10319 Point compression for binary fields is disabled by default for
10320 patent reasons (compile with OPENSSL_EC_BIN_PT_COMP defined to
10323 As binary polynomials are represented as BIGNUMs, various members
10324 of the EC_GROUP and EC_POINT data structures can be shared
10325 between the implementations for prime fields and binary fields;
10326 the above `..._GF2m functions` (except for EX_GROUP_new_curve_GF2m)
10327 are essentially identical to their `..._GFp` counterparts.
10328 (For simplicity, the `..._GFp` prefix has been dropped from
10329 various internal method names.)
10331 An internal 'field_div' method (similar to 'field_mul' and
10332 'field_sqr') has been added; this is used only for binary fields.
10334 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10336 * Optionally dispatch EC_POINT_mul(), EC_POINT_precompute_mult()
10337 through methods ('mul', 'precompute_mult').
10339 The generic implementations (now internally called 'ec_wNAF_mul'
10340 and 'ec_wNAF_precomputed_mult') remain the default if these
10341 methods are undefined.
10343 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10345 * New function EC_GROUP_get_degree, which is defined through
10346 EC_METHOD. For curves over prime fields, this returns the bit
10347 length of the modulus.
10349 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10351 * New functions EC_GROUP_dup, EC_POINT_dup.
10352 (These simply call ..._new and ..._copy).
10354 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10356 * Add binary polynomial arithmetic software in crypto/bn/bn_gf2m.c.
10357 Polynomials are represented as BIGNUMs (where the sign bit is not
10358 used) in the following functions [macros]:
10361 BN_GF2m_sub [= BN_GF2m_add]
10362 BN_GF2m_mod [wrapper for BN_GF2m_mod_arr]
10363 BN_GF2m_mod_mul [wrapper for BN_GF2m_mod_mul_arr]
10364 BN_GF2m_mod_sqr [wrapper for BN_GF2m_mod_sqr_arr]
10366 BN_GF2m_mod_exp [wrapper for BN_GF2m_mod_exp_arr]
10367 BN_GF2m_mod_sqrt [wrapper for BN_GF2m_mod_sqrt_arr]
10368 BN_GF2m_mod_solve_quad [wrapper for BN_GF2m_mod_solve_quad_arr]
10369 BN_GF2m_cmp [= BN_ucmp]
10371 (Note that only the 'mod' functions are actually for fields GF(2^m).
10372 BN_GF2m_add() is misnomer, but this is for the sake of consistency.)
10374 For some functions, an the irreducible polynomial defining a
10375 field can be given as an 'unsigned int[]' with strictly
10376 decreasing elements giving the indices of those bits that are set;
10377 i.e., p[] represents the polynomial
10378 f(t) = t^p[0] + t^p[1] + ... + t^p[k]
10380 p[0] > p[1] > ... > p[k] = 0.
10381 This applies to the following functions:
10384 BN_GF2m_mod_mul_arr
10385 BN_GF2m_mod_sqr_arr
10386 BN_GF2m_mod_inv_arr [wrapper for BN_GF2m_mod_inv]
10387 BN_GF2m_mod_div_arr [wrapper for BN_GF2m_mod_div]
10388 BN_GF2m_mod_exp_arr
10389 BN_GF2m_mod_sqrt_arr
10390 BN_GF2m_mod_solve_quad_arr
10394 Conversion can be performed by the following functions:
10399 bntest.c has additional tests for binary polynomial arithmetic.
10401 Two implementations for BN_GF2m_mod_div() are available.
10402 The default algorithm simply uses BN_GF2m_mod_inv() and
10403 BN_GF2m_mod_mul(). The alternative algorithm is compiled in only
10404 if OPENSSL_SUN_GF2M_DIV is defined (patent pending; read the
10405 copyright notice in crypto/bn/bn_gf2m.c before enabling it).
10407 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10409 * Add new error code 'ERR_R_DISABLED' that can be used when some
10410 functionality is disabled at compile-time.
10412 *Douglas Stebila <douglas.stebila@sun.com>*
10414 * Change default behaviour of 'openssl asn1parse' so that more
10415 information is visible when viewing, e.g., a certificate:
10417 Modify asn1_parse2 (crypto/asn1/asn1_par.c) so that in non-'dump'
10418 mode the content of non-printable OCTET STRINGs is output in a
10419 style similar to INTEGERs, but with '[HEX DUMP]' prepended to
10420 avoid the appearance of a printable string.
10422 *Nils Larsch <nla@trustcenter.de>*
10424 * Add 'asn1_flag' and 'asn1_form' member to EC_GROUP with access
10426 EC_GROUP_set_asn1_flag()
10427 EC_GROUP_get_asn1_flag()
10428 EC_GROUP_set_point_conversion_form()
10429 EC_GROUP_get_point_conversion_form()
10430 These control ASN1 encoding details:
10431 - Curves (i.e., groups) are encoded explicitly unless asn1_flag
10432 has been set to OPENSSL_EC_NAMED_CURVE.
10433 - Points are encoded in uncompressed form by default; options for
10434 asn1_for are as for point2oct, namely
10435 POINT_CONVERSION_COMPRESSED
10436 POINT_CONVERSION_UNCOMPRESSED
10437 POINT_CONVERSION_HYBRID
10439 Also add 'seed' and 'seed_len' members to EC_GROUP with access
10441 EC_GROUP_set_seed()
10442 EC_GROUP_get0_seed()
10443 EC_GROUP_get_seed_len()
10444 This is used only for ASN1 purposes (so far).
10446 *Nils Larsch <nla@trustcenter.de>*
10448 * Add 'field_type' member to EC_METHOD, which holds the NID
10449 of the appropriate field type OID. The new function
10450 EC_METHOD_get_field_type() returns this value.
10452 *Nils Larsch <nla@trustcenter.de>*
10455 EC_POINT_point2bn()
10456 EC_POINT_bn2point()
10457 EC_POINT_point2hex()
10458 EC_POINT_hex2point()
10459 providing useful interfaces to EC_POINT_point2oct() and
10460 EC_POINT_oct2point().
10462 *Nils Larsch <nla@trustcenter.de>*
10464 * Change internals of the EC library so that the functions
10465 EC_GROUP_set_generator()
10466 EC_GROUP_get_generator()
10467 EC_GROUP_get_order()
10468 EC_GROUP_get_cofactor()
10469 are implemented directly in crypto/ec/ec_lib.c and not dispatched
10470 to methods, which would lead to unnecessary code duplication when
10471 adding different types of curves.
10473 *Nils Larsch <nla@trustcenter.de> with input by Bodo Moeller*
10475 * Implement compute_wNAF (crypto/ec/ec_mult.c) without BIGNUM
10476 arithmetic, and such that modified wNAFs are generated
10477 (which avoid length expansion in many cases).
10481 * Add a function EC_GROUP_check_discriminant() (defined via
10482 EC_METHOD) that verifies that the curve discriminant is non-zero.
10484 Add a function EC_GROUP_check() that makes some sanity tests
10485 on a EC_GROUP, its generator and order. This includes
10486 EC_GROUP_check_discriminant().
10488 *Nils Larsch <nla@trustcenter.de>*
10490 * Add ECDSA in new directory crypto/ecdsa/.
10492 Add applications 'openssl ecparam' and 'openssl ecdsa'
10493 (these are based on 'openssl dsaparam' and 'openssl dsa').
10495 ECDSA support is also included in various other files across the
10496 library. Most notably,
10497 - 'openssl req' now has a '-newkey ecdsa:file' option;
10498 - EVP_PKCS82PKEY (crypto/evp/evp_pkey.c) now can handle ECDSA;
10499 - X509_PUBKEY_get (crypto/asn1/x_pubkey.c) and
10500 d2i_PublicKey (crypto/asn1/d2i_pu.c) have been modified to make
10501 them suitable for ECDSA where domain parameters must be
10502 extracted before the specific public key;
10503 - ECDSA engine support has been added.
10505 *Nils Larsch <nla@trustcenter.de>*
10507 * Include some named elliptic curves, and add OIDs from X9.62,
10508 SECG, and WAP/WTLS. Each curve can be obtained from the new
10510 EC_GROUP_new_by_curve_name(),
10511 and the list of available named curves can be obtained with
10512 EC_get_builtin_curves().
10513 Also add a 'curve_name' member to EC_GROUP objects, which can be
10515 EC_GROUP_set_curve_name()
10516 EC_GROUP_get_curve_name()
10518 *Nils Larsch <larsch@trustcenter.de, Bodo Moeller*
10520 * Remove a few calls to bn_wexpand() in BN_sqr() (the one in there
10521 was actually never needed) and in BN_mul(). The removal in BN_mul()
10522 required a small change in bn_mul_part_recursive() and the addition
10523 of the functions bn_cmp_part_words(), bn_sub_part_words() and
10524 bn_add_part_words(), which do the same thing as bn_cmp_words(),
10525 bn_sub_words() and bn_add_words() except they take arrays with
10530 ### Changes between 0.9.7l and 0.9.7m [23 Feb 2007]
10532 * Cleanse PEM buffers before freeing them since they may contain
10535 *Benjamin Bennett <ben@psc.edu>*
10537 * Include "!eNULL" in SSL_DEFAULT_CIPHER_LIST to make sure that
10538 a ciphersuite string such as "DEFAULT:RSA" cannot enable
10539 authentication-only ciphersuites.
10543 * Since AES128 and AES256 share a single mask bit in the logic of
10544 ssl/ssl_ciph.c, the code for masking out disabled ciphers needs a
10545 kludge to work properly if AES128 is available and AES256 isn't.
10549 * Expand security boundary to match 1.1.1 module.
10553 * Remove redundant features: hash file source, editing of test vectors
10554 modify fipsld to use external fips_premain.c signature.
10558 * New perl script mkfipsscr.pl to create shell scripts or batch files to
10559 run algorithm test programs.
10563 * Make algorithm test programs more tolerant of whitespace.
10567 * Have SSL/TLS server implementation tolerate "mismatched" record
10568 protocol version while receiving ClientHello even if the
10569 ClientHello is fragmented. (The server can't insist on the
10570 particular protocol version it has chosen before the ServerHello
10571 message has informed the client about his choice.)
10575 * Load error codes if they are not already present instead of using a
10576 static variable. This allows them to be cleanly unloaded and reloaded.
10580 ### Changes between 0.9.7k and 0.9.7l [28 Sep 2006]
10582 * Introduce limits to prevent malicious keys being able to
10583 cause a denial of service. [CVE-2006-2940][]
10585 *Steve Henson, Bodo Moeller*
10587 * Fix ASN.1 parsing of certain invalid structures that can result
10588 in a denial of service. [CVE-2006-2937][] [Steve Henson]
10590 * Fix buffer overflow in SSL_get_shared_ciphers() function.
10591 [CVE-2006-3738][] [Tavis Ormandy and Will Drewry, Google Security Team]
10593 * Fix SSL client code which could crash if connecting to a
10594 malicious SSLv2 server. [CVE-2006-4343][]
10596 *Tavis Ormandy and Will Drewry, Google Security Team*
10598 * Change ciphersuite string processing so that an explicit
10599 ciphersuite selects this one ciphersuite (so that "AES256-SHA"
10600 will no longer include "AES128-SHA"), and any other similar
10601 ciphersuite (same bitmap) from *other* protocol versions (so that
10602 "RC4-MD5" will still include both the SSL 2.0 ciphersuite and the
10603 SSL 3.0/TLS 1.0 ciphersuite). This is a backport combining
10604 changes from 0.9.8b and 0.9.8d.
10608 ### Changes between 0.9.7j and 0.9.7k [05 Sep 2006]
10610 * Avoid PKCS #1 v1.5 signature attack discovered by Daniel Bleichenbacher
10611 [CVE-2006-4339][] [Ben Laurie and Google Security Team]
10613 * Change the Unix randomness entropy gathering to use poll() when
10614 possible instead of select(), since the latter has some
10615 undesirable limitations.
10617 *Darryl Miles via Richard Levitte and Bodo Moeller*
10619 * Disable rogue ciphersuites:
10621 - SSLv2 0x08 0x00 0x80 ("RC4-64-MD5")
10622 - SSLv3/TLSv1 0x00 0x61 ("EXP1024-RC2-CBC-MD5")
10623 - SSLv3/TLSv1 0x00 0x60 ("EXP1024-RC4-MD5")
10625 The latter two were purportedly from
10626 draft-ietf-tls-56-bit-ciphersuites-0[01].txt, but do not really
10629 Also deactivate the remaining ciphersuites from
10630 draft-ietf-tls-56-bit-ciphersuites-01.txt. These are just as
10631 unofficial, and the ID has long expired.
10635 * Fix RSA blinding Heisenbug (problems sometimes occurred on
10636 dual-core machines) and other potential thread-safety issues.
10640 ### Changes between 0.9.7i and 0.9.7j [04 May 2006]
10642 * Adapt fipsld and the build system to link against the validated FIPS
10643 module in FIPS mode.
10647 * Fixes for VC++ 2005 build under Windows.
10651 * Add new Windows build target VC-32-GMAKE for VC++. This uses GNU make
10652 from a Windows bash shell such as MSYS. It is autodetected from the
10653 "config" script when run from a VC++ environment. Modify standard VC++
10654 build to use fipscanister.o from the GNU make build.
10658 ### Changes between 0.9.7h and 0.9.7i [14 Oct 2005]
10660 * Wrapped the definition of EVP_MAX_MD_SIZE in a #ifdef OPENSSL_FIPS.
10661 The value now differs depending on if you build for FIPS or not.
10662 BEWARE! A program linked with a shared FIPSed libcrypto can't be
10663 safely run with a non-FIPSed libcrypto, as it may crash because of
10664 the difference induced by this change.
10668 ### Changes between 0.9.7g and 0.9.7h [11 Oct 2005]
10670 * Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING
10671 (part of SSL_OP_ALL). This option used to disable the
10672 countermeasure against man-in-the-middle protocol-version
10673 rollback in the SSL 2.0 server implementation, which is a bad
10674 idea. [CVE-2005-2969][]
10676 *Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center
10677 for Information Security, National Institute of Advanced Industrial
10678 Science and Technology [AIST, Japan)]*
10680 * Minimal support for X9.31 signatures and PSS padding modes. This is
10681 mainly for FIPS compliance and not fully integrated at this stage.
10685 * For DSA signing, unless DSA_FLAG_NO_EXP_CONSTTIME is set, perform
10686 the exponentiation using a fixed-length exponent. (Otherwise,
10687 the information leaked through timing could expose the secret key
10688 after many signatures; cf. Bleichenbacher's attack on DSA with
10693 * Make a new fixed-window mod_exp implementation the default for
10694 RSA, DSA, and DH private-key operations so that the sequence of
10695 squares and multiplies and the memory access pattern are
10696 independent of the particular secret key. This will mitigate
10697 cache-timing and potential related attacks.
10699 BN_mod_exp_mont_consttime() is the new exponentiation implementation,
10700 and this is automatically used by BN_mod_exp_mont() if the new flag
10701 BN_FLG_EXP_CONSTTIME is set for the exponent. RSA, DSA, and DH
10702 will use this BN flag for private exponents unless the flag
10703 RSA_FLAG_NO_EXP_CONSTTIME, DSA_FLAG_NO_EXP_CONSTTIME, or
10704 DH_FLAG_NO_EXP_CONSTTIME, respectively, is set.
10706 *Matthew D Wood (Intel Corp), with some changes by Bodo Moeller*
10708 * Change the client implementation for SSLv23_method() and
10709 SSLv23_client_method() so that is uses the SSL 3.0/TLS 1.0
10710 Client Hello message format if the SSL_OP_NO_SSLv2 option is set.
10711 (Previously, the SSL 2.0 backwards compatible Client Hello
10712 message format would be used even with SSL_OP_NO_SSLv2.)
10716 * Add support for smime-type MIME parameter in S/MIME messages which some
10721 * New function BN_MONT_CTX_set_locked() to set montgomery parameters in
10722 a threadsafe manner. Modify rsa code to use new function and add calls
10723 to dsa and dh code (which had race conditions before).
10727 * Include the fixed error library code in the C error file definitions
10728 instead of fixing them up at runtime. This keeps the error code
10729 structures constant.
10733 ### Changes between 0.9.7f and 0.9.7g [11 Apr 2005]
10735 [NB: OpenSSL 0.9.7h and later 0.9.7 patch levels were released after
10738 * Fixes for newer kerberos headers. NB: the casts are needed because
10739 the 'length' field is signed on one version and unsigned on another
10740 with no (?) obvious way to tell the difference, without these VC++
10741 complains. Also the "definition" of FAR (blank) is no longer included
10742 nor is the error ENOMEM. KRB5_PRIVATE has to be set to 1 to pick up
10743 some needed definitions.
10747 * Undo Cygwin change.
10751 * Added support for proxy certificates according to RFC 3820.
10752 Because they may be a security thread to unaware applications,
10753 they must be explicitly allowed in run-time. See
10754 docs/HOWTO/proxy_certificates.txt for further information.
10758 ### Changes between 0.9.7e and 0.9.7f [22 Mar 2005]
10760 * Use (SSL_RANDOM_VALUE - 4) bytes of pseudo random data when generating
10761 server and client random values. Previously
10762 (SSL_RANDOM_VALUE - sizeof(time_t)) would be used which would result in
10763 less random data when sizeof(time_t) > 4 (some 64 bit platforms).
10765 This change has negligible security impact because:
10767 1. Server and client random values still have 24 bytes of pseudo random
10770 2. Server and client random values are sent in the clear in the initial
10773 3. The master secret is derived using the premaster secret (48 bytes in
10774 size for static RSA ciphersuites) as well as client server and random
10777 The OpenSSL team would like to thank the UK NISCC for bringing this issue
10780 *Stephen Henson, reported by UK NISCC*
10782 * Use Windows randomness collection on Cygwin.
10786 * Fix hang in EGD/PRNGD query when communication socket is closed
10787 prematurely by EGD/PRNGD.
10789 *Darren Tucker <dtucker@zip.com.au> via Lutz Jänicke, resolves #1014*
10791 * Prompt for pass phrases when appropriate for PKCS12 input format.
10795 * Back-port of selected performance improvements from development
10796 branch, as well as improved support for PowerPC platforms.
10800 * Add lots of checks for memory allocation failure, error codes to indicate
10801 failure and freeing up memory if a failure occurs.
10803 *Nauticus Networks SSL Team <openssl@nauticusnet.com>, Steve Henson*
10805 * Add new -passin argument to dgst.
10809 * Perform some character comparisons of different types in X509_NAME_cmp:
10810 this is needed for some certificates that re-encode DNs into UTF8Strings
10811 (in violation of RFC3280) and can't or won't issue name rollover
10816 * Make an explicit check during certificate validation to see that
10817 the CA setting in each certificate on the chain is correct. As a
10818 side effect always do the following basic checks on extensions,
10819 not just when there's an associated purpose to the check:
10821 - if there is an unhandled critical extension (unless the user
10822 has chosen to ignore this fault)
10823 - if the path length has been exceeded (if one is set at all)
10824 - that certain extensions fit the associated purpose (if one has
10829 ### Changes between 0.9.7d and 0.9.7e [25 Oct 2004]
10831 * Avoid a race condition when CRLs are checked in a multi threaded
10832 environment. This would happen due to the reordering of the revoked
10833 entries during signature checking and serial number lookup. Now the
10834 encoding is cached and the serial number sort performed under a lock.
10835 Add new STACK function sk_is_sorted().
10839 * Add Delta CRL to the extension code.
10843 * Various fixes to s3_pkt.c so alerts are sent properly.
10845 *David Holmes <d.holmes@f5.com>*
10847 * Reduce the chances of duplicate issuer name and serial numbers (in
10848 violation of RFC3280) using the OpenSSL certificate creation utilities.
10849 This is done by creating a random 64 bit value for the initial serial
10850 number when a serial number file is created or when a self signed
10851 certificate is created using 'openssl req -x509'. The initial serial
10852 number file is created using 'openssl x509 -next_serial' in CA.pl
10853 rather than being initialized to 1.
10857 ### Changes between 0.9.7c and 0.9.7d [17 Mar 2004]
10859 * Fix null-pointer assignment in do_change_cipher_spec() revealed
10860 by using the Codenomicon TLS Test Tool [CVE-2004-0079][]
10862 *Joe Orton, Steve Henson*
10864 * Fix flaw in SSL/TLS handshaking when using Kerberos ciphersuites
10867 *Joe Orton, Steve Henson*
10869 * Make it possible to have multiple active certificates with the same
10870 subject in the CA index file. This is done only if the keyword
10871 'unique_subject' is set to 'no' in the main CA section (default
10872 if 'CA_default') of the configuration file. The value is saved
10873 with the database itself in a separate index attribute file,
10874 named like the index file with '.attr' appended to the name.
10878 * X509 verify fixes. Disable broken certificate workarounds when
10879 X509_V_FLAGS_X509_STRICT is set. Check CRL issuer has cRLSign set if
10880 keyUsage extension present. Don't accept CRLs with unhandled critical
10881 extensions: since verify currently doesn't process CRL extensions this
10882 rejects a CRL with *any* critical extensions. Add new verify error codes
10887 * When creating an OCSP nonce use an OCTET STRING inside the extnValue.
10888 A clarification of RFC2560 will require the use of OCTET STRINGs and
10889 some implementations cannot handle the current raw format. Since OpenSSL
10890 copies and compares OCSP nonces as opaque blobs without any attempt at
10891 parsing them this should not create any compatibility issues.
10895 * New md flag EVP_MD_CTX_FLAG_REUSE this allows md_data to be reused when
10896 calling EVP_MD_CTX_copy_ex() to avoid calling OPENSSL_malloc(). Without
10897 this HMAC (and other) operations are several times slower than OpenSSL
10902 * Print out GeneralizedTime and UTCTime in ASN1_STRING_print_ex().
10904 *Peter Sylvester <Peter.Sylvester@EdelWeb.fr>*
10906 * Use the correct content when signing type "other".
10910 ### Changes between 0.9.7b and 0.9.7c [30 Sep 2003]
10912 * Fix various bugs revealed by running the NISCC test suite:
10914 Stop out of bounds reads in the ASN1 code when presented with
10915 invalid tags (CVE-2003-0543 and CVE-2003-0544).
10917 Free up ASN1_TYPE correctly if ANY type is invalid [CVE-2003-0545][].
10919 If verify callback ignores invalid public key errors don't try to check
10920 certificate signature with the NULL public key.
10924 * New -ignore_err option in ocsp application to stop the server
10925 exiting on the first error in a request.
10929 * In ssl3_accept() (ssl/s3_srvr.c) only accept a client certificate
10930 if the server requested one: as stated in TLS 1.0 and SSL 3.0
10935 * In ssl3_get_client_hello() (ssl/s3_srvr.c), tolerate additional
10936 extra data after the compression methods not only for TLS 1.0
10937 but also for SSL 3.0 (as required by the specification).
10939 *Bodo Moeller; problem pointed out by Matthias Loepfe*
10941 * Change X509_certificate_type() to mark the key as exported/exportable
10942 when it's 512 *bits* long, not 512 bytes.
10946 * Change AES_cbc_encrypt() so it outputs exact multiple of
10947 blocks during encryption.
10951 * Various fixes to base64 BIO and non blocking I/O. On write
10952 flushes were not handled properly if the BIO retried. On read
10953 data was not being buffered properly and had various logic bugs.
10954 This also affects blocking I/O when the data being decoded is a
10959 * Various S/MIME bugfixes and compatibility changes:
10960 output correct application/pkcs7 MIME type if
10961 PKCS7_NOOLDMIMETYPE is set. Tolerate some broken signatures.
10962 Output CR+LF for EOL if PKCS7_CRLFEOL is set (this makes opening
10963 of files as .eml work). Correctly handle very long lines in MIME
10968 ### Changes between 0.9.7a and 0.9.7b [10 Apr 2003]
10970 * Countermeasure against the Klima-Pokorny-Rosa extension of
10971 Bleichbacher's attack on PKCS #1 v1.5 padding: treat
10972 a protocol version number mismatch like a decryption error
10973 in ssl3_get_client_key_exchange (ssl/s3_srvr.c).
10977 * Turn on RSA blinding by default in the default implementation
10978 to avoid a timing attack. Applications that don't want it can call
10979 RSA_blinding_off() or use the new flag RSA_FLAG_NO_BLINDING.
10980 They would be ill-advised to do so in most cases.
10982 *Ben Laurie, Steve Henson, Geoff Thorpe, Bodo Moeller*
10984 * Change RSA blinding code so that it works when the PRNG is not
10985 seeded (in this case, the secret RSA exponent is abused as
10986 an unpredictable seed -- if it is not unpredictable, there
10987 is no point in blinding anyway). Make RSA blinding thread-safe
10988 by remembering the creator's thread ID in rsa->blinding and
10989 having all other threads use local one-time blinding factors
10990 (this requires more computation than sharing rsa->blinding, but
10991 avoids excessive locking; and if an RSA object is not shared
10992 between threads, blinding will still be very fast).
10996 * Fixed a typo bug that would cause ENGINE_set_default() to set an
10997 ENGINE as defaults for all supported algorithms irrespective of
10998 the 'flags' parameter. 'flags' is now honoured, so applications
10999 should make sure they are passing it correctly.
11003 * Target "mingw" now allows native Windows code to be generated in
11004 the Cygwin environment as well as with the MinGW compiler.
11008 ### Changes between 0.9.7 and 0.9.7a [19 Feb 2003]
11010 * In ssl3_get_record (ssl/s3_pkt.c), minimize information leaked
11011 via timing by performing a MAC computation even if incorrect
11012 block cipher padding has been found. This is a countermeasure
11013 against active attacks where the attacker has to distinguish
11014 between bad padding and a MAC verification error. [CVE-2003-0078][]
11016 *Bodo Moeller; problem pointed out by Brice Canvel (EPFL),
11017 Alain Hiltgen (UBS), Serge Vaudenay (EPFL), and
11018 Martin Vuagnoux (EPFL, Ilion)*
11020 * Make the no-err option work as intended. The intention with no-err
11021 is not to have the whole error stack handling routines removed from
11022 libcrypto, it's only intended to remove all the function name and
11023 reason texts, thereby removing some of the footprint that may not
11024 be interesting if those errors aren't displayed anyway.
11026 NOTE: it's still possible for any application or module to have its
11027 own set of error texts inserted. The routines are there, just not
11028 used by default when no-err is given.
11032 * Add support for FreeBSD on IA64.
11034 *dirk.meyer@dinoex.sub.org via Richard Levitte, resolves #454*
11036 * Adjust DES_cbc_cksum() so it returns the same value as the MIT
11037 Kerberos function mit_des_cbc_cksum(). Before this change,
11038 the value returned by DES_cbc_cksum() was like the one from
11039 mit_des_cbc_cksum(), except the bytes were swapped.
11041 *Kevin Greaney <Kevin.Greaney@hp.com> and Richard Levitte*
11043 * Allow an application to disable the automatic SSL chain building.
11044 Before this a rather primitive chain build was always performed in
11045 ssl3_output_cert_chain(): an application had no way to send the
11046 correct chain if the automatic operation produced an incorrect result.
11048 Now the chain builder is disabled if either:
11050 1. Extra certificates are added via SSL_CTX_add_extra_chain_cert().
11052 2. The mode flag SSL_MODE_NO_AUTO_CHAIN is set.
11054 The reasoning behind this is that an application would not want the
11055 auto chain building to take place if extra chain certificates are
11056 present and it might also want a means of sending no additional
11057 certificates (for example the chain has two certificates and the
11062 * Add the possibility to build without the ENGINE framework.
11064 *Steven Reddie <smr@essemer.com.au> via Richard Levitte*
11066 * Under Win32 gmtime() can return NULL: check return value in
11067 OPENSSL_gmtime(). Add error code for case where gmtime() fails.
11071 * DSA routines: under certain error conditions uninitialized BN objects
11072 could be freed. Solution: make sure initialization is performed early
11073 enough. (Reported and fix supplied by Ivan D Nestlerode <nestler@MIT.EDU>,
11074 Nils Larsch <nla@trustcenter.de> via PR#459)
11078 * Another fix for SSLv2 session ID handling: the session ID was incorrectly
11079 checked on reconnect on the client side, therefore session resumption
11080 could still fail with a "ssl session id is different" error. This
11081 behaviour is masked when SSL_OP_ALL is used due to
11082 SSL_OP_MICROSOFT_SESS_ID_BUG being set.
11083 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as
11084 followup to PR #377.
11088 * IA-32 assembler support enhancements: unified ELF targets, support
11089 for SCO/Caldera platforms, fix for Cygwin shared build.
11093 * Add support for FreeBSD on sparc64. As a consequence, support for
11094 FreeBSD on non-x86 processors is separate from x86 processors on
11095 the config script, much like the NetBSD support.
11097 *Richard Levitte & Kris Kennaway <kris@obsecurity.org>*
11099 ### Changes between 0.9.6h and 0.9.7 [31 Dec 2002]
11101 [NB: OpenSSL 0.9.6i and later 0.9.6 patch levels were released after
11104 * Fix session ID handling in SSLv2 client code: the SERVER FINISHED
11105 code (06) was taken as the first octet of the session ID and the last
11106 octet was ignored consequently. As a result SSLv2 client side session
11107 caching could not have worked due to the session ID mismatch between
11109 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as
11114 * Change the declaration of needed Kerberos libraries to use EX_LIBS
11115 instead of the special (and badly supported) LIBKRB5. LIBKRB5 is
11120 * The hw_ncipher.c engine requires dynamic locks. Unfortunately, it
11121 seems that in spite of existing for more than a year, many application
11122 author have done nothing to provide the necessary callbacks, which
11123 means that this particular engine will not work properly anywhere.
11124 This is a very unfortunate situation which forces us, in the name
11125 of usability, to give the hw_ncipher.c a static lock, which is part
11127 NOTE: This is for the 0.9.7 series ONLY. This hack will never
11128 appear in 0.9.8 or later. We EXPECT application authors to have
11129 dealt properly with this when 0.9.8 is released (unless we actually
11130 make such changes in the libcrypto locking code that changes will
11131 have to be made anyway).
11135 * In asn1_d2i_read_bio() repeatedly call BIO_read() until all content
11136 octets have been read, EOF or an error occurs. Without this change
11137 some truncated ASN1 structures will not produce an error.
11141 * Disable Heimdal support, since it hasn't been fully implemented.
11142 Still give the possibility to force the use of Heimdal, but with
11143 warnings and a request that patches get sent to openssl-dev.
11147 * Add the VC-CE target, introduce the WINCE sysname, and add
11148 INSTALL.WCE and appropriate conditionals to make it build.
11150 *Steven Reddie <smr@essemer.com.au> via Richard Levitte*
11152 * Change the DLL names for Cygwin to cygcrypto-x.y.z.dll and
11153 cygssl-x.y.z.dll, where x, y and z are the major, minor and
11154 edit numbers of the version.
11156 *Corinna Vinschen <vinschen@redhat.com> and Richard Levitte*
11158 * Introduce safe string copy and catenation functions
11159 (BUF_strlcpy() and BUF_strlcat()).
11161 *Ben Laurie (CHATS) and Richard Levitte*
11163 * Avoid using fixed-size buffers for one-line DNs.
11165 *Ben Laurie (CHATS)*
11167 * Add BUF_MEM_grow_clean() to avoid information leakage when
11168 resizing buffers containing secrets, and use where appropriate.
11170 *Ben Laurie (CHATS)*
11172 * Avoid using fixed size buffers for configuration file location.
11174 *Ben Laurie (CHATS)*
11176 * Avoid filename truncation for various CA files.
11178 *Ben Laurie (CHATS)*
11180 * Use sizeof in preference to magic numbers.
11182 *Ben Laurie (CHATS)*
11184 * Avoid filename truncation in cert requests.
11186 *Ben Laurie (CHATS)*
11188 * Add assertions to check for (supposedly impossible) buffer
11191 *Ben Laurie (CHATS)*
11193 * Don't cache truncated DNS entries in the local cache (this could
11194 potentially lead to a spoofing attack).
11196 *Ben Laurie (CHATS)*
11198 * Fix various buffers to be large enough for hex/decimal
11199 representations in a platform independent manner.
11201 *Ben Laurie (CHATS)*
11203 * Add CRYPTO_realloc_clean() to avoid information leakage when
11204 resizing buffers containing secrets, and use where appropriate.
11206 *Ben Laurie (CHATS)*
11208 * Add BIO_indent() to avoid much slightly worrying code to do
11211 *Ben Laurie (CHATS)*
11213 * Convert sprintf()/BIO_puts() to BIO_printf().
11215 *Ben Laurie (CHATS)*
11217 * buffer_gets() could terminate with the buffer only half
11220 *Ben Laurie (CHATS)*
11222 * Add assertions to prevent user-supplied crypto functions from
11223 overflowing internal buffers by having large block sizes, etc.
11225 *Ben Laurie (CHATS)*
11227 * New OPENSSL_assert() macro (similar to assert(), but enabled
11230 *Ben Laurie (CHATS)*
11232 * Eliminate unused copy of key in RC4.
11234 *Ben Laurie (CHATS)*
11236 * Eliminate unused and incorrectly sized buffers for IV in pem.h.
11238 *Ben Laurie (CHATS)*
11240 * Fix off-by-one error in EGD path.
11242 *Ben Laurie (CHATS)*
11244 * If RANDFILE path is too long, ignore instead of truncating.
11246 *Ben Laurie (CHATS)*
11248 * Eliminate unused and incorrectly sized X.509 structure
11251 *Ben Laurie (CHATS)*
11253 * Eliminate unused and dangerous function knumber().
11255 *Ben Laurie (CHATS)*
11257 * Eliminate unused and dangerous structure, KSSL_ERR.
11259 *Ben Laurie (CHATS)*
11261 * Protect against overlong session ID context length in an encoded
11262 session object. Since these are local, this does not appear to be
11265 *Ben Laurie (CHATS)*
11267 * Change from security patch (see 0.9.6e below) that did not affect
11268 the 0.9.6 release series:
11270 Remote buffer overflow in SSL3 protocol - an attacker could
11271 supply an oversized master key in Kerberos-enabled versions.
11274 *Ben Laurie (CHATS)*
11276 * Change the SSL kerb5 codes to match RFC 2712.
11280 * Make -nameopt work fully for req and add -reqopt switch.
11282 *Michael Bell <michael.bell@rz.hu-berlin.de>, Steve Henson*
11284 * The "block size" for block ciphers in CFB and OFB mode should be 1.
11286 *Steve Henson, reported by Yngve Nysaeter Pettersen <yngve@opera.com>*
11288 * Make sure tests can be performed even if the corresponding algorithms
11289 have been removed entirely. This was also the last step to make
11290 OpenSSL compilable with DJGPP under all reasonable conditions.
11292 *Richard Levitte, Doug Kaufman <dkaufman@rahul.net>*
11294 * Add cipher selection rules COMPLEMENTOFALL and COMPLEMENTOFDEFAULT
11295 to allow version independent disabling of normally unselected ciphers,
11296 which may be activated as a side-effect of selecting a single cipher.
11298 (E.g., cipher list string "RSA" enables ciphersuites that are left
11299 out of "ALL" because they do not provide symmetric encryption.
11300 "RSA:!COMPLEMEMENTOFALL" avoids these unsafe ciphersuites.)
11302 *Lutz Jaenicke, Bodo Moeller*
11304 * Add appropriate support for separate platform-dependent build
11305 directories. The recommended way to make a platform-dependent
11306 build directory is the following (tested on Linux), maybe with
11309 # Place yourself outside of the OpenSSL source tree. In
11310 # this example, the environment variable OPENSSL_SOURCE
11311 # is assumed to contain the absolute OpenSSL source directory.
11312 mkdir -p objtree/"`uname -s`-`uname -r`-`uname -m`"
11313 cd objtree/"`uname -s`-`uname -r`-`uname -m`"
11314 (cd $OPENSSL_SOURCE; find . -type f) | while read F; do
11315 mkdir -p `dirname $F`
11316 ln -s $OPENSSL_SOURCE/$F $F
11319 To be absolutely sure not to disturb the source tree, a "make clean"
11320 is a good thing. If it isn't successful, don't worry about it,
11321 it probably means the source directory is very clean.
11325 * Make sure any ENGINE control commands make local copies of string
11326 pointers passed to them whenever necessary. Otherwise it is possible
11327 the caller may have overwritten (or deallocated) the original string
11328 data when a later ENGINE operation tries to use the stored values.
11330 *Götz Babin-Ebell <babinebell@trustcenter.de>*
11332 * Improve diagnostics in file reading and command-line digests.
11334 *Ben Laurie aided and abetted by Solar Designer <solar@openwall.com>*
11336 * Add AES modes CFB and OFB to the object database. Correct an
11337 error in AES-CFB decryption.
11341 * Remove most calls to EVP_CIPHER_CTX_cleanup() in evp_enc.c, this
11342 allows existing EVP_CIPHER_CTX structures to be reused after
11343 calling `EVP_*Final()`. This behaviour is used by encryption
11344 BIOs and some applications. This has the side effect that
11345 applications must explicitly clean up cipher contexts with
11346 EVP_CIPHER_CTX_cleanup() or they will leak memory.
11350 * Check the values of dna and dnb in bn_mul_recursive before calling
11351 bn_mul_comba (a non zero value means the a or b arrays do not contain
11352 n2 elements) and fallback to bn_mul_normal if either is not zero.
11356 * Fix escaping of non-ASCII characters when using the -subj option
11357 of the "openssl req" command line tool. (Robert Joop <joop@fokus.gmd.de>)
11361 * Make object definitions compliant to LDAP (RFC2256): SN is the short
11362 form for "surname", serialNumber has no short form.
11363 Use "mail" as the short name for "rfc822Mailbox" according to RFC2798;
11364 therefore remove "mail" short name for "internet 7".
11365 The OID for unique identifiers in X509 certificates is
11366 x500UniqueIdentifier, not uniqueIdentifier.
11367 Some more OID additions. (Michael Bell <michael.bell@rz.hu-berlin.de>)
11371 * Add an "init" command to the ENGINE config module and auto initialize
11372 ENGINEs. Without any "init" command the ENGINE will be initialized
11373 after all ctrl commands have been executed on it. If init=1 the
11374 ENGINE is initialized at that point (ctrls before that point are run
11375 on the uninitialized ENGINE and after on the initialized one). If
11376 init=0 then the ENGINE will not be initialized at all.
11380 * Fix the 'app_verify_callback' interface so that the user-defined
11381 argument is actually passed to the callback: In the
11382 SSL_CTX_set_cert_verify_callback() prototype, the callback
11383 declaration has been changed from
11386 int (*cb)(X509_STORE_CTX *,void *);
11387 in ssl_verify_cert_chain (ssl/ssl_cert.c), the call
11388 i=s->ctx->app_verify_callback(&ctx)
11389 has been changed into
11390 i=s->ctx->app_verify_callback(&ctx, s->ctx->app_verify_arg).
11392 To update applications using SSL_CTX_set_cert_verify_callback(),
11393 a dummy argument can be added to their callback functions.
11395 *D. K. Smetters <smetters@parc.xerox.com>*
11397 * Added the '4758cca' ENGINE to support IBM 4758 cards.
11399 *Maurice Gittens <maurice@gittens.nl>, touchups by Geoff Thorpe*
11401 * Add and OPENSSL_LOAD_CONF define which will cause
11402 OpenSSL_add_all_algorithms() to load the openssl.cnf config file.
11403 This allows older applications to transparently support certain
11404 OpenSSL features: such as crypto acceleration and dynamic ENGINE loading.
11405 Two new functions OPENSSL_add_all_algorithms_noconf() which will never
11406 load the config file and OPENSSL_add_all_algorithms_conf() which will
11407 always load it have also been added.
11411 * Add the OFB, CFB and CTR (all with 128 bit feedback) to AES.
11412 Adjust NIDs and EVP layer.
11414 *Stephen Sprunk <stephen@sprunk.org> and Richard Levitte*
11416 * Config modules support in openssl utility.
11418 Most commands now load modules from the config file,
11419 though in a few (such as version) this isn't done
11420 because it couldn't be used for anything.
11422 In the case of ca and req the config file used is
11423 the same as the utility itself: that is the -config
11424 command line option can be used to specify an
11429 * Move default behaviour from OPENSSL_config(). If appname is NULL
11430 use "openssl_conf" if filename is NULL use default openssl config file.
11434 * Add an argument to OPENSSL_config() to allow the use of an alternative
11435 config section name. Add a new flag to tolerate a missing config file
11436 and move code to CONF_modules_load_file().
11440 * Support for crypto accelerator cards from Accelerated Encryption
11441 Processing, www.aep.ie. (Use engine 'aep')
11442 The support was copied from 0.9.6c [engine] and adapted/corrected
11443 to work with the new engine framework.
11445 *AEP Inc. and Richard Levitte*
11447 * Support for SureWare crypto accelerator cards from Baltimore
11448 Technologies. (Use engine 'sureware')
11449 The support was copied from 0.9.6c [engine] and adapted
11450 to work with the new engine framework.
11454 * Have the CHIL engine fork-safe (as defined by nCipher) and actually
11455 make the newer ENGINE framework commands for the CHIL engine work.
11457 *Toomas Kiisk <vix@cyber.ee> and Richard Levitte*
11459 * Make it possible to produce shared libraries on ReliantUNIX.
11461 *Robert Dahlem <Robert.Dahlem@ffm2.siemens.de> via Richard Levitte*
11463 * Add the configuration target debug-linux-ppro.
11464 Make 'openssl rsa' use the general key loading routines
11465 implemented in apps.c, and make those routines able to
11466 handle the key format FORMAT_NETSCAPE and the variant
11469 *Toomas Kiisk <vix@cyber.ee> via Richard Levitte*
11471 * Fix a crashbug and a logic bug in hwcrhk_load_pubkey().
11473 *Toomas Kiisk <vix@cyber.ee> via Richard Levitte*
11475 * Add -keyform to rsautl, and document -engine.
11477 *Richard Levitte, inspired by Toomas Kiisk <vix@cyber.ee>*
11479 * Change BIO_new_file (crypto/bio/bss_file.c) to use new
11480 BIO_R_NO_SUCH_FILE error code rather than the generic
11481 ERR_R_SYS_LIB error code if fopen() fails with ENOENT.
11485 * Add new functions
11486 ERR_peek_last_error
11487 ERR_peek_last_error_line
11488 ERR_peek_last_error_line_data.
11489 These are similar to
11491 ERR_peek_error_line
11492 ERR_peek_error_line_data,
11493 but report on the latest error recorded rather than the first one
11494 still in the error queue.
11496 *Ben Laurie, Bodo Moeller*
11498 * default_algorithms option in ENGINE config module. This allows things
11500 default_algorithms = ALL
11501 default_algorithms = RSA, DSA, RAND, CIPHERS, DIGESTS
11505 * Preliminary ENGINE config module.
11509 * New experimental application configuration code.
11513 * Change the AES code to follow the same name structure as all other
11514 symmetric ciphers, and behave the same way. Move everything to
11515 the directory crypto/aes, thereby obsoleting crypto/rijndael.
11517 *Stephen Sprunk <stephen@sprunk.org> and Richard Levitte*
11519 * SECURITY: remove unsafe setjmp/signal interaction from ui_openssl.c.
11521 *Ben Laurie and Theo de Raadt*
11523 * Add option to output public keys in req command.
11525 *Massimiliano Pala madwolf@openca.org*
11527 * Use wNAFs in EC_POINTs_mul() for improved efficiency
11528 (up to about 10% better than before for P-192 and P-224).
11532 * New functions/macros
11534 SSL_CTX_set_msg_callback(ctx, cb)
11535 SSL_CTX_set_msg_callback_arg(ctx, arg)
11536 SSL_set_msg_callback(ssl, cb)
11537 SSL_set_msg_callback_arg(ssl, arg)
11539 to request calling a callback function
11541 void cb(int write_p, int version, int content_type,
11542 const void *buf, size_t len, SSL *ssl, void *arg)
11544 whenever a protocol message has been completely received
11545 (write_p == 0) or sent (write_p == 1). Here 'version' is the
11546 protocol version according to which the SSL library interprets
11547 the current protocol message (SSL2_VERSION, SSL3_VERSION, or
11548 TLS1_VERSION). 'content_type' is 0 in the case of SSL 2.0, or
11549 the content type as defined in the SSL 3.0/TLS 1.0 protocol
11550 specification (change_cipher_spec(20), alert(21), handshake(22)).
11551 'buf' and 'len' point to the actual message, 'ssl' to the
11552 SSL object, and 'arg' is the application-defined value set by
11553 SSL[_CTX]_set_msg_callback_arg().
11555 'openssl s_client' and 'openssl s_server' have new '-msg' options
11556 to enable a callback that displays all protocol messages.
11560 * Change the shared library support so shared libraries are built as
11561 soon as the corresponding static library is finished, and thereby get
11562 openssl and the test programs linked against the shared library.
11563 This still only happens when the keyword "shard" has been given to
11564 the configuration scripts.
11566 NOTE: shared library support is still an experimental thing, and
11567 backward binary compatibility is still not guaranteed.
11569 *"Maciej W. Rozycki" <macro@ds2.pg.gda.pl> and Richard Levitte*
11571 * Add support for Subject Information Access extension.
11573 *Peter Sylvester <Peter.Sylvester@EdelWeb.fr>*
11575 * Make BUF_MEM_grow() behaviour more consistent: Initialise to zero
11576 additional bytes when new memory had to be allocated, not just
11577 when reusing an existing buffer.
11581 * New command line and configuration option 'utf8' for the req command.
11582 This allows field values to be specified as UTF8 strings.
11586 * Add -multi and -mr options to "openssl speed" - giving multiple parallel
11587 runs for the former and machine-readable output for the latter.
11591 * Add '-noemailDN' option to 'openssl ca'. This prevents inclusion
11592 of the e-mail address in the DN (i.e., it will go into a certificate
11593 extension only). The new configuration file option 'email_in_dn = no'
11594 has the same effect.
11596 *Massimiliano Pala madwolf@openca.org*
11598 * Change all functions with names starting with `des_` to be starting
11599 with `DES_` instead. Add wrappers that are compatible with libdes,
11600 but are named `_ossl_old_des_*`. Finally, add macros that map the
11601 `des_*` symbols to the corresponding `_ossl_old_des_*` if libdes
11602 compatibility is desired. If OpenSSL 0.9.6c compatibility is
11603 desired, the `des_*` symbols will be mapped to `DES_*`, with one
11606 Since we provide two compatibility mappings, the user needs to
11607 define the macro OPENSSL_DES_LIBDES_COMPATIBILITY if libdes
11608 compatibility is desired. The default (i.e., when that macro
11609 isn't defined) is OpenSSL 0.9.6c compatibility.
11611 There are also macros that enable and disable the support of old
11612 des functions altogether. Those are OPENSSL_ENABLE_OLD_DES_SUPPORT
11613 and OPENSSL_DISABLE_OLD_DES_SUPPORT. If none or both of those
11614 are defined, the default will apply: to support the old des routines.
11616 In either case, one must include openssl/des.h to get the correct
11617 definitions. Do not try to just include openssl/des_old.h, that
11620 NOTE: This is a major break of an old API into a new one. Software
11621 authors are encouraged to switch to the `DES_` style functions. Some
11622 time in the future, des_old.h and the libdes compatibility functions
11623 will be disable (i.e. OPENSSL_DISABLE_OLD_DES_SUPPORT will be the
11624 default), and then completely removed.
11628 * Test for certificates which contain unsupported critical extensions.
11629 If such a certificate is found during a verify operation it is
11630 rejected by default: this behaviour can be overridden by either
11631 handling the new error X509_V_ERR_UNHANDLED_CRITICAL_EXTENSION or
11632 by setting the verify flag X509_V_FLAG_IGNORE_CRITICAL. A new function
11633 X509_supported_extension() has also been added which returns 1 if a
11634 particular extension is supported.
11638 * Modify the behaviour of EVP cipher functions in similar way to digests
11639 to retain compatibility with existing code.
11643 * Modify the behaviour of EVP_DigestInit() and EVP_DigestFinal() to retain
11644 compatibility with existing code. In particular the 'ctx' parameter does
11645 not have to be to be initialized before the call to EVP_DigestInit() and
11646 it is tidied up after a call to EVP_DigestFinal(). New function
11647 EVP_DigestFinal_ex() which does not tidy up the ctx. Similarly function
11648 EVP_MD_CTX_copy() changed to not require the destination to be
11649 initialized valid and new function EVP_MD_CTX_copy_ex() added which
11650 requires the destination to be valid.
11652 Modify all the OpenSSL digest calls to use EVP_DigestInit_ex(),
11653 EVP_DigestFinal_ex() and EVP_MD_CTX_copy_ex().
11657 * Change ssl3_get_message (ssl/s3_both.c) and the functions using it
11658 so that complete 'Handshake' protocol structures are kept in memory
11659 instead of overwriting 'msg_type' and 'length' with 'body' data.
11663 * Add an implementation of SSL_add_dir_cert_subjects_to_stack for Win32.
11665 *Massimo Santin via Richard Levitte*
11667 * Major restructuring to the underlying ENGINE code. This includes
11668 reduction of linker bloat, separation of pure "ENGINE" manipulation
11669 (initialisation, etc) from functionality dealing with implementations
11670 of specific crypto interfaces. This change also introduces integrated
11671 support for symmetric ciphers and digest implementations - so ENGINEs
11672 can now accelerate these by providing EVP_CIPHER and EVP_MD
11673 implementations of their own. This is detailed in
11674 [crypto/engine/README.md](crypto/engine/README.md)
11675 as it couldn't be adequately described here. However, there are a few
11676 API changes worth noting - some RSA, DSA, DH, and RAND functions that
11677 were changed in the original introduction of ENGINE code have now
11678 reverted back - the hooking from this code to ENGINE is now a good
11679 deal more passive and at run-time, operations deal directly with
11680 RSA_METHODs, DSA_METHODs (etc) as they did before, rather than
11681 dereferencing through an ENGINE pointer any more. Also, the ENGINE
11682 functions dealing with `BN_MOD_EXP[_CRT]` handlers have been removed -
11683 they were not being used by the framework as there is no concept of a
11684 BIGNUM_METHOD and they could not be generalised to the new
11685 'ENGINE_TABLE' mechanism that underlies the new code. Similarly,
11686 ENGINE_cpy() has been removed as it cannot be consistently defined in
11691 * Change ASN1_GENERALIZEDTIME_check() to allow fractional seconds.
11695 * Change mkdef.pl to sort symbols that get the same entry number,
11696 and make sure the automatically generated functions `ERR_load_*`
11697 become part of libeay.num as well.
11701 * New function SSL_renegotiate_pending(). This returns true once
11702 renegotiation has been requested (either SSL_renegotiate() call
11703 or HelloRequest/ClientHello received from the peer) and becomes
11704 false once a handshake has been completed.
11705 (For servers, SSL_renegotiate() followed by SSL_do_handshake()
11706 sends a HelloRequest, but does not ensure that a handshake takes
11707 place. SSL_renegotiate_pending() is useful for checking if the
11708 client has followed the request.)
11712 * New SSL option SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION.
11713 By default, clients may request session resumption even during
11714 renegotiation (if session ID contexts permit); with this option,
11715 session resumption is possible only in the first handshake.
11717 SSL_OP_ALL is now 0x00000FFFL instead of 0x000FFFFFL. This makes
11718 more bits available for options that should not be part of
11719 SSL_OP_ALL (such as SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION).
11723 * Add some demos for certificate and certificate request creation.
11727 * Make maximum certificate chain size accepted from the peer application
11728 settable (`SSL*_get/set_max_cert_list()`), as proposed by
11729 "Douglas E. Engert" <deengert@anl.gov>.
11733 * Add support for shared libraries for Unixware-7
11734 (Boyd Lynn Gerber <gerberb@zenez.com>).
11738 * Add a "destroy" handler to ENGINEs that allows structural cleanup to
11739 be done prior to destruction. Use this to unload error strings from
11740 ENGINEs that load their own error strings. NB: This adds two new API
11741 functions to "get" and "set" this destroy handler in an ENGINE.
11745 * Alter all existing ENGINE implementations (except "openssl" and
11746 "openbsd") to dynamically instantiate their own error strings. This
11747 makes them more flexible to be built both as statically-linked ENGINEs
11748 and self-contained shared-libraries loadable via the "dynamic" ENGINE.
11749 Also, add stub code to each that makes building them as self-contained
11750 shared-libraries easier (see [README-Engine.md](README-Engine.md)).
11754 * Add a "dynamic" ENGINE that provides a mechanism for binding ENGINE
11755 implementations into applications that are completely implemented in
11756 self-contained shared-libraries. The "dynamic" ENGINE exposes control
11757 commands that can be used to configure what shared-library to load and
11758 to control aspects of the way it is handled. Also, made an update to
11759 the [README-Engine.md](README-Engine.md) file
11760 that brings its information up-to-date and
11761 provides some information and instructions on the "dynamic" ENGINE
11762 (ie. how to use it, how to build "dynamic"-loadable ENGINEs, etc).
11766 * Make it possible to unload ranges of ERR strings with a new
11767 "ERR_unload_strings" function.
11771 * Add a copy() function to EVP_MD.
11775 * Make EVP_MD routines take a context pointer instead of just the
11776 md_data void pointer.
11780 * Add flags to EVP_MD and EVP_MD_CTX. EVP_MD_FLAG_ONESHOT indicates
11781 that the digest can only process a single chunk of data
11782 (typically because it is provided by a piece of
11783 hardware). EVP_MD_CTX_FLAG_ONESHOT indicates that the application
11784 is only going to provide a single chunk of data, and hence the
11785 framework needn't accumulate the data for oneshot drivers.
11789 * As with "ERR", make it possible to replace the underlying "ex_data"
11790 functions. This change also alters the storage and management of global
11791 ex_data state - it's now all inside ex_data.c and all "class" code (eg.
11792 RSA, BIO, SSL_CTX, etc) no longer stores its own STACKS and per-class
11793 index counters. The API functions that use this state have been changed
11794 to take a "class_index" rather than pointers to the class's local STACK
11795 and counter, and there is now an API function to dynamically create new
11796 classes. This centralisation allows us to (a) plug a lot of the
11797 thread-safety problems that existed, and (b) makes it possible to clean
11798 up all allocated state using "CRYPTO_cleanup_all_ex_data()". W.r.t. (b)
11799 such data would previously have always leaked in application code and
11800 workarounds were in place to make the memory debugging turn a blind eye
11801 to it. Application code that doesn't use this new function will still
11802 leak as before, but their memory debugging output will announce it now
11803 rather than letting it slide.
11805 Besides the addition of CRYPTO_cleanup_all_ex_data(), another API change
11806 induced by the "ex_data" overhaul is that X509_STORE_CTX_init() now
11807 has a return value to indicate success or failure.
11811 * Make it possible to replace the underlying "ERR" functions such that the
11812 global state (2 LHASH tables and 2 locks) is only used by the "default"
11813 implementation. This change also adds two functions to "get" and "set"
11814 the implementation prior to it being automatically set the first time
11815 any other ERR function takes place. Ie. an application can call "get",
11816 pass the return value to a module it has just loaded, and that module
11817 can call its own "set" function using that value. This means the
11818 module's "ERR" operations will use (and modify) the error state in the
11819 application and not in its own statically linked copy of OpenSSL code.
11823 * Give DH, DSA, and RSA types their own `*_up_ref()` function to increment
11824 reference counts. This performs normal REF_PRINT/REF_CHECK macros on
11825 the operation, and provides a more encapsulated way for external code
11826 (crypto/evp/ and ssl/) to do this. Also changed the evp and ssl code
11827 to use these functions rather than manually incrementing the counts.
11829 Also rename "DSO_up()" function to more descriptive "DSO_up_ref()".
11833 * Add EVP test program.
11837 * Add symmetric cipher support to ENGINE. Expect the API to change!
11841 * New CRL functions: X509_CRL_set_version(), X509_CRL_set_issuer_name()
11842 X509_CRL_set_lastUpdate(), X509_CRL_set_nextUpdate(), X509_CRL_sort(),
11843 X509_REVOKED_set_serialNumber(), and X509_REVOKED_set_revocationDate().
11844 These allow a CRL to be built without having to access X509_CRL fields
11845 directly. Modify 'ca' application to use new functions.
11849 * Move SSL_OP_TLS_ROLLBACK_BUG out of the SSL_OP_ALL list of recommended
11850 bug workarounds. Rollback attack detection is a security feature.
11851 The problem will only arise on OpenSSL servers when TLSv1 is not
11852 available (sslv3_server_method() or SSL_OP_NO_TLSv1).
11853 Software authors not wanting to support TLSv1 will have special reasons
11854 for their choice and can explicitly enable this option.
11856 *Bodo Moeller, Lutz Jaenicke*
11858 * Rationalise EVP so it can be extended: don't include a union of
11859 cipher/digest structures, add init/cleanup functions for EVP_MD_CTX
11860 (similar to those existing for EVP_CIPHER_CTX).
11865 EVP_MD_CTX_init(&md); /* new function call */
11866 EVP_DigestInit(&md, EVP_sha1());
11867 EVP_DigestUpdate(&md, in, len);
11868 EVP_DigestFinal(&md, out, NULL);
11869 EVP_MD_CTX_cleanup(&md); /* new function call */
11873 * Make DES key schedule conform to the usual scheme, as well as
11874 correcting its structure. This means that calls to DES functions
11875 now have to pass a pointer to a des_key_schedule instead of a
11876 plain des_key_schedule (which was actually always a pointer
11879 des_key_schedule ks;
11881 des_set_key_checked(..., &ks);
11882 des_ncbc_encrypt(..., &ks, ...);
11884 (Note that a later change renames 'des_...' into 'DES_...'.)
11888 * Initial reduction of linker bloat: the use of some functions, such as
11889 PEM causes large amounts of unused functions to be linked in due to
11890 poor organisation. For example pem_all.c contains every PEM function
11891 which has a knock on effect of linking in large amounts of (unused)
11892 ASN1 code. Grouping together similar functions and splitting unrelated
11893 functions prevents this.
11897 * Cleanup of EVP macros.
11901 * Change historical references to `{NID,SN,LN}_des_ede` and ede3 to add the
11902 correct `_ecb suffix`.
11906 * Add initial OCSP responder support to ocsp application. The
11907 revocation information is handled using the text based index
11908 use by the ca application. The responder can either handle
11909 requests generated internally, supplied in files (for example
11910 via a CGI script) or using an internal minimal server.
11914 * Add configuration choices to get zlib compression for TLS.
11918 * Changes to Kerberos SSL for RFC 2712 compliance:
11919 1. Implemented real KerberosWrapper, instead of just using
11920 KRB5 AP_REQ message. [Thanks to Simon Wilkinson <sxw@sxw.org.uk>]
11921 2. Implemented optional authenticator field of KerberosWrapper.
11923 Added openssl-style ASN.1 macros for Kerberos ticket, ap_req,
11924 and authenticator structs; see crypto/krb5/.
11926 Generalized Kerberos calls to support multiple Kerberos libraries.
11927 *Vern Staats <staatsvr@asc.hpc.mil>, Jeffrey Altman <jaltman@columbia.edu>
11928 via Richard Levitte*
11930 * Cause 'openssl speed' to use fully hard-coded DSA keys as it
11931 already does with RSA. testdsa.h now has 'priv_key/pub_key'
11932 values for each of the key sizes rather than having just
11933 parameters (and 'speed' generating keys each time).
11937 * Speed up EVP routines.
11940 pe 8 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes
11941 s-cbc 4408.85k 5560.51k 5778.46k 5862.20k 5825.16k
11942 s-cbc 4389.55k 5571.17k 5792.23k 5846.91k 5832.11k
11943 s-cbc 4394.32k 5575.92k 5807.44k 5848.37k 5841.30k
11945 s-cbc 3482.66k 5069.49k 5496.39k 5614.16k 5639.28k
11946 s-cbc 3480.74k 5068.76k 5510.34k 5609.87k 5635.52k
11947 s-cbc 3483.72k 5067.62k 5504.60k 5708.01k 5724.80k
11950 s-cbc 4660.16k 5650.19k 5807.19k 5827.13k 5783.32k
11952 s-cbc 3624.96k 5258.21k 5530.91k 5624.30k 5628.26k
11956 * Added the OS2-EMX target.
11958 *"Brian Havard" <brianh@kheldar.apana.org.au> and Richard Levitte*
11960 * Rewrite apps to use NCONF routines instead of the old CONF. New functions
11961 to support NCONF routines in extension code. New function CONF_set_nconf()
11962 to allow functions which take an NCONF to also handle the old LHASH
11963 structure: this means that the old CONF compatible routines can be
11964 retained (in particular wrt extensions) without having to duplicate the
11965 code. New function X509V3_add_ext_nconf_sk to add extensions to a stack.
11969 * Enhance the general user interface with mechanisms for inner control
11970 and with possibilities to have yes/no kind of prompts.
11974 * Change all calls to low level digest routines in the library and
11975 applications to use EVP. Add missing calls to HMAC_cleanup() and
11976 don't assume HMAC_CTX can be copied using memcpy().
11978 *Verdon Walker <VWalker@novell.com>, Steve Henson*
11980 * Add the possibility to control engines through control names but with
11981 arbitrary arguments instead of just a string.
11982 Change the key loaders to take a UI_METHOD instead of a callback
11983 function pointer. NOTE: this breaks binary compatibility with earlier
11984 versions of OpenSSL [engine].
11985 Adapt the nCipher code for these new conditions and add a card insertion
11990 * Enhance the general user interface with mechanisms to better support
11991 dialog box interfaces, application-defined prompts, the possibility
11992 to use defaults (for example default passwords from somewhere else)
11993 and interrupts/cancellations.
11997 * Tidy up PKCS#12 attribute handling. Add support for the CSP name
11998 attribute in PKCS#12 files, add new -CSP option to pkcs12 utility.
12002 * Fix a memory leak in 'sk_dup()' in the case reallocation fails. (Also
12003 tidy up some unnecessarily weird code in 'sk_new()').
12005 *Geoff, reported by Diego Tartara <dtartara@novamens.com>*
12007 * Change the key loading routines for ENGINEs to use the same kind
12008 callback (pem_password_cb) as all other routines that need this
12013 * Increase ENTROPY_NEEDED to 32 bytes, as Rijndael can operate with
12014 256 bit (=32 byte) keys. Of course seeding with more entropy bytes
12015 than this minimum value is recommended.
12019 * New random seeder for OpenVMS, using the system process statistics
12020 that are easily reachable.
12024 * Windows apparently can't transparently handle global
12025 variables defined in DLLs. Initialisations such as:
12027 const ASN1_ITEM *it = &ASN1_INTEGER_it;
12029 won't compile. This is used by the any applications that need to
12030 declare their own ASN1 modules. This was fixed by adding the option
12031 EXPORT_VAR_AS_FN to all Win32 platforms, although this isn't strictly
12032 needed for static libraries under Win32.
12036 * New functions X509_PURPOSE_set() and X509_TRUST_set() to handle
12037 setting of purpose and trust fields. New X509_STORE trust and
12038 purpose functions and tidy up setting in other SSL functions.
12042 * Add copies of X509_STORE_CTX fields and callbacks to X509_STORE
12043 structure. These are inherited by X509_STORE_CTX when it is
12044 initialised. This allows various defaults to be set in the
12045 X509_STORE structure (such as flags for CRL checking and custom
12046 purpose or trust settings) for functions which only use X509_STORE_CTX
12047 internally such as S/MIME.
12049 Modify X509_STORE_CTX_purpose_inherit() so it only sets purposes and
12050 trust settings if they are not set in X509_STORE. This allows X509_STORE
12051 purposes and trust (in S/MIME for example) to override any set by default.
12053 Add command line options for CRL checking to smime, s_client and s_server
12058 * Initial CRL based revocation checking. If the CRL checking flag(s)
12059 are set then the CRL is looked up in the X509_STORE structure and
12060 its validity and signature checked, then if the certificate is found
12061 in the CRL the verify fails with a revoked error.
12063 Various new CRL related callbacks added to X509_STORE_CTX structure.
12065 Command line options added to 'verify' application to support this.
12067 This needs some additional work, such as being able to handle multiple
12068 CRLs with different times, extension based lookup (rather than just
12069 by subject name) and ultimately more complete V2 CRL extension
12074 * Add a general user interface API (crypto/ui/). This is designed
12075 to replace things like des_read_password and friends (backward
12076 compatibility functions using this new API are provided).
12077 The purpose is to remove prompting functions from the DES code
12078 section as well as provide for prompting through dialog boxes in
12079 a window system and the like.
12083 * Add "ex_data" support to ENGINE so implementations can add state at a
12084 per-structure level rather than having to store it globally.
12088 * Make it possible for ENGINE structures to be copied when retrieved by
12089 ENGINE_by_id() if the ENGINE specifies a new flag: ENGINE_FLAGS_BY_ID_COPY.
12090 This causes the "original" ENGINE structure to act like a template,
12091 analogous to the RSA vs. RSA_METHOD type of separation. Because of this
12092 operational state can be localised to each ENGINE structure, despite the
12093 fact they all share the same "methods". New ENGINE structures returned in
12094 this case have no functional references and the return value is the single
12095 structural reference. This matches the single structural reference returned
12096 by ENGINE_by_id() normally, when it is incremented on the pre-existing
12101 * Fix ASN1 decoder when decoding type ANY and V_ASN1_OTHER: since this
12102 needs to match any other type at all we need to manually clear the
12107 * Changes to the "openssl engine" utility to include;
12108 - verbosity levels ('-v', '-vv', and '-vvv') that provide information
12109 about an ENGINE's available control commands.
12110 - executing control commands from command line arguments using the
12111 '-pre' and '-post' switches. '-post' is only used if '-t' is
12112 specified and the ENGINE is successfully initialised. The syntax for
12113 the individual commands are colon-separated, for example;
12114 openssl engine chil -pre FORK_CHECK:0 -pre SO_PATH:/lib/test.so
12118 * New dynamic control command support for ENGINEs. ENGINEs can now
12119 declare their own commands (numbers), names (strings), descriptions,
12120 and input types for run-time discovery by calling applications. A
12121 subset of these commands are implicitly classed as "executable"
12122 depending on their input type, and only these can be invoked through
12123 the new string-based API function ENGINE_ctrl_cmd_string(). (Eg. this
12124 can be based on user input, config files, etc). The distinction is
12125 that "executable" commands cannot return anything other than a boolean
12126 result and can only support numeric or string input, whereas some
12127 discoverable commands may only be for direct use through
12128 ENGINE_ctrl(), eg. supporting the exchange of binary data, function
12129 pointers, or other custom uses. The "executable" commands are to
12130 support parameterisations of ENGINE behaviour that can be
12131 unambiguously defined by ENGINEs and used consistently across any
12132 OpenSSL-based application. Commands have been added to all the
12133 existing hardware-supporting ENGINEs, noticeably "SO_PATH" to allow
12134 control over shared-library paths without source code alterations.
12138 * Changed all ENGINE implementations to dynamically allocate their
12139 ENGINEs rather than declaring them statically. Apart from this being
12140 necessary with the removal of the ENGINE_FLAGS_MALLOCED distinction,
12141 this also allows the implementations to compile without using the
12142 internal engine_int.h header.
12146 * Minor adjustment to "rand" code. RAND_get_rand_method() now returns a
12147 'const' value. Any code that should be able to modify a RAND_METHOD
12148 should already have non-const pointers to it (ie. they should only
12149 modify their own ones).
12153 * Made a variety of little tweaks to the ENGINE code.
12154 - "atalla" and "ubsec" string definitions were moved from header files
12155 to C code. "nuron" string definitions were placed in variables
12156 rather than hard-coded - allowing parameterisation of these values
12157 later on via ctrl() commands.
12158 - Removed unused "#if 0"'d code.
12159 - Fixed engine list iteration code so it uses ENGINE_free() to release
12160 structural references.
12161 - Constified the RAND_METHOD element of ENGINE structures.
12162 - Constified various get/set functions as appropriate and added
12163 missing functions (including a catch-all ENGINE_cpy that duplicates
12164 all ENGINE values onto a new ENGINE except reference counts/state).
12165 - Removed NULL parameter checks in get/set functions. Setting a method
12166 or function to NULL is a way of cancelling out a previously set
12167 value. Passing a NULL ENGINE parameter is just plain stupid anyway
12168 and doesn't justify the extra error symbols and code.
12169 - Deprecate the ENGINE_FLAGS_MALLOCED define and move the area for
12170 flags from engine_int.h to engine.h.
12171 - Changed prototypes for ENGINE handler functions (init(), finish(),
12172 ctrl(), key-load functions, etc) to take an (ENGINE*) parameter.
12176 * Implement binary inversion algorithm for BN_mod_inverse in addition
12177 to the algorithm using long division. The binary algorithm can be
12178 used only if the modulus is odd. On 32-bit systems, it is faster
12179 only for relatively small moduli (roughly 20-30% for 128-bit moduli,
12180 roughly 5-15% for 256-bit moduli), so we use it only for moduli
12181 up to 450 bits. In 64-bit environments, the binary algorithm
12182 appears to be advantageous for much longer moduli; here we use it
12183 for moduli up to 2048 bits.
12187 * Rewrite CHOICE field setting in ASN1_item_ex_d2i(). The old code
12188 could not support the combine flag in choice fields.
12192 * Add a 'copy_extensions' option to the 'ca' utility. This copies
12193 extensions from a certificate request to the certificate.
12197 * Allow multiple 'certopt' and 'nameopt' options to be separated
12198 by commas. Add 'namopt' and 'certopt' options to the 'ca' config
12199 file: this allows the display of the certificate about to be
12200 signed to be customised, to allow certain fields to be included
12201 or excluded and extension details. The old system didn't display
12202 multicharacter strings properly, omitted fields not in the policy
12203 and couldn't display additional details such as extensions.
12207 * Function EC_POINTs_mul for multiple scalar multiplication
12208 of an arbitrary number of elliptic curve points
12209 \sum scalars[i]*points[i],
12210 optionally including the generator defined for the EC_GROUP:
12211 scalar*generator + \sum scalars[i]*points[i].
12213 EC_POINT_mul is a simple wrapper function for the typical case
12214 that the point list has just one item (besides the optional
12219 * First EC_METHODs for curves over GF(p):
12221 EC_GFp_simple_method() uses the basic BN_mod_mul and BN_mod_sqr
12222 operations and provides various method functions that can also
12223 operate with faster implementations of modular arithmetic.
12225 EC_GFp_mont_method() reuses most functions that are part of
12226 EC_GFp_simple_method, but uses Montgomery arithmetic.
12228 *Bodo Moeller; point addition and point doubling
12229 implementation directly derived from source code provided by
12230 Lenka Fibikova <fibikova@exp-math.uni-essen.de>*
12232 * Framework for elliptic curves (crypto/ec/ec.h, crypto/ec/ec_lcl.h,
12233 crypto/ec/ec_lib.c):
12235 Curves are EC_GROUP objects (with an optional group generator)
12236 based on EC_METHODs that are built into the library.
12238 Points are EC_POINT objects based on EC_GROUP objects.
12240 Most of the framework would be able to handle curves over arbitrary
12241 finite fields, but as there are no obvious types for fields other
12242 than GF(p), some functions are limited to that for now.
12246 * Add the -HTTP option to s_server. It is similar to -WWW, but requires
12247 that the file contains a complete HTTP response.
12251 * Add the ec directory to mkdef.pl and mkfiles.pl. In mkdef.pl
12252 change the def and num file printf format specifier from "%-40sXXX"
12253 to "%-39s XXX". The latter will always guarantee a space after the
12254 field while the former will cause them to run together if the field
12255 is 40 of more characters long.
12259 * Constify the cipher and digest 'method' functions and structures
12260 and modify related functions to take constant EVP_MD and EVP_CIPHER
12265 * Hide BN_CTX structure details in bn_lcl.h instead of publishing them
12266 in <openssl/bn.h>. Also further increase BN_CTX_NUM to 32.
12270 * Modify `EVP_Digest*()` routines so they now return values. Although the
12271 internal software routines can never fail additional hardware versions
12276 * Clean up crypto/err/err.h and change some error codes to avoid conflicts:
12278 Previously ERR_R_FATAL was too small and coincided with ERR_LIB_PKCS7
12279 (= ERR_R_PKCS7_LIB); it is now 64 instead of 32.
12282 ERR_R_NESTED_ASN1_ERROR
12284 ERR_R_MISSING_ASN1_EOS
12285 were 4 .. 9, conflicting with
12286 ERR_LIB_RSA (= ERR_R_RSA_LIB)
12288 ERR_LIB_PEM (= ERR_R_PEM_LIB).
12289 They are now 58 .. 63 (i.e., just below ERR_R_FATAL).
12291 Add new error code 'ERR_R_INTERNAL_ERROR'.
12295 * Don't overuse locks in crypto/err/err.c: For data retrieval, CRYPTO_r_lock
12300 * New option '-subj arg' for 'openssl req' and 'openssl ca'. This
12301 sets the subject name for a new request or supersedes the
12302 subject name in a given request. Formats that can be parsed are
12303 'CN=Some Name, OU=myOU, C=IT'
12305 'CN=Some Name/OU=myOU/C=IT'.
12307 Add options '-batch' and '-verbose' to 'openssl req'.
12309 *Massimiliano Pala <madwolf@hackmasters.net>*
12311 * Introduce the possibility to access global variables through
12312 functions on platform were that's the best way to handle exporting
12313 global variables in shared libraries. To enable this functionality,
12314 one must configure with "EXPORT_VAR_AS_FN" or defined the C macro
12315 "OPENSSL_EXPORT_VAR_AS_FUNCTION" in crypto/opensslconf.h (the latter
12316 is normally done by Configure or something similar).
12318 To implement a global variable, use the macro OPENSSL_IMPLEMENT_GLOBAL
12319 in the source file (foo.c) like this:
12321 OPENSSL_IMPLEMENT_GLOBAL(int,foo)=1;
12322 OPENSSL_IMPLEMENT_GLOBAL(double,bar);
12324 To declare a global variable, use the macros OPENSSL_DECLARE_GLOBAL
12325 and OPENSSL_GLOBAL_REF in the header file (foo.h) like this:
12327 OPENSSL_DECLARE_GLOBAL(int,foo);
12328 #define foo OPENSSL_GLOBAL_REF(foo)
12329 OPENSSL_DECLARE_GLOBAL(double,bar);
12330 #define bar OPENSSL_GLOBAL_REF(bar)
12332 The #defines are very important, and therefore so is including the
12333 header file everywhere where the defined globals are used.
12335 The macro OPENSSL_EXPORT_VAR_AS_FUNCTION also affects the definition
12336 of ASN.1 items, but that structure is a bit different.
12338 The largest change is in util/mkdef.pl which has been enhanced with
12339 better and easier to understand logic to choose which symbols should
12340 go into the Windows .def files as well as a number of fixes and code
12341 cleanup (among others, algorithm keywords are now sorted
12342 lexicographically to avoid constant rewrites).
12346 * In BN_div() keep a copy of the sign of 'num' before writing the
12347 result to 'rm' because if rm==num the value will be overwritten
12348 and produce the wrong result if 'num' is negative: this caused
12349 problems with BN_mod() and BN_nnmod().
12353 * Function OCSP_request_verify(). This checks the signature on an
12354 OCSP request and verifies the signer certificate. The signer
12355 certificate is just checked for a generic purpose and OCSP request
12360 * Add OCSP_check_validity() function to check the validity of OCSP
12361 responses. OCSP responses are prepared in real time and may only
12362 be a few seconds old. Simply checking that the current time lies
12363 between thisUpdate and nextUpdate max reject otherwise valid responses
12364 caused by either OCSP responder or client clock inaccuracy. Instead
12365 we allow thisUpdate and nextUpdate to fall within a certain period of
12366 the current time. The age of the response can also optionally be
12367 checked. Two new options -validity_period and -status_age added to
12372 * If signature or public key algorithm is unrecognized print out its
12373 OID rather that just UNKNOWN.
12377 * Change OCSP_cert_to_id() to tolerate a NULL subject certificate and
12378 OCSP_cert_id_new() a NULL serialNumber. This allows a partial certificate
12379 ID to be generated from the issuer certificate alone which can then be
12380 passed to OCSP_id_issuer_cmp().
12384 * New compilation option ASN1_ITEM_FUNCTIONS. This causes the new
12385 ASN1 modules to export functions returning ASN1_ITEM pointers
12386 instead of the ASN1_ITEM structures themselves. This adds several
12387 new macros which allow the underlying ASN1 function/structure to
12388 be accessed transparently. As a result code should not use ASN1_ITEM
12389 references directly (such as &X509_it) but instead use the relevant
12390 macros (such as ASN1_ITEM_rptr(X509)). This option is to allow
12391 use of the new ASN1 code on platforms where exporting structures
12392 is problematical (for example in shared libraries) but exporting
12393 functions returning pointers to structures is not.
12397 * Add support for overriding the generation of SSL/TLS session IDs.
12398 These callbacks can be registered either in an SSL_CTX or per SSL.
12399 The purpose of this is to allow applications to control, if they wish,
12400 the arbitrary values chosen for use as session IDs, particularly as it
12401 can be useful for session caching in multiple-server environments. A
12402 command-line switch for testing this (and any client code that wishes
12403 to use such a feature) has been added to "s_server".
12405 *Geoff Thorpe, Lutz Jaenicke*
12407 * Modify mkdef.pl to recognise and parse preprocessor conditionals
12408 of the form `#if defined(...) || defined(...) || ...` and
12409 `#if !defined(...) && !defined(...) && ...`. This also avoids
12410 the growing number of special cases it was previously handling.
12414 * Make all configuration macros available for application by making
12415 sure they are available in opensslconf.h, by giving them names starting
12416 with `OPENSSL_` to avoid conflicts with other packages and by making
12417 sure e_os2.h will cover all platform-specific cases together with
12419 Additionally, it is now possible to define configuration/platform-
12420 specific names (called "system identities"). In the C code, these
12421 are prefixed with `OPENSSL_SYSNAME_`. e_os2.h will create another
12422 macro with the name beginning with `OPENSSL_SYS_`, which is determined
12423 from `OPENSSL_SYSNAME_*` or compiler-specific macros depending on
12428 * New option -set_serial to 'req' and 'x509' this allows the serial
12429 number to use to be specified on the command line. Previously self
12430 signed certificates were hard coded with serial number 0 and the
12431 CA options of 'x509' had to use a serial number in a file which was
12436 * New options to 'ca' utility to support V2 CRL entry extensions.
12437 Currently CRL reason, invalidity date and hold instruction are
12438 supported. Add new CRL extensions to V3 code and some new objects.
12442 * New function EVP_CIPHER_CTX_set_padding() this is used to
12443 disable standard block padding (aka PKCS#5 padding) in the EVP
12444 API, which was previously mandatory. This means that the data is
12445 not padded in any way and so the total length much be a multiple
12446 of the block size, otherwise an error occurs.
12450 * Initial (incomplete) OCSP SSL support.
12454 * New function OCSP_parse_url(). This splits up a URL into its host,
12455 port and path components: primarily to parse OCSP URLs. New -url
12456 option to ocsp utility.
12460 * New nonce behavior. The return value of OCSP_check_nonce() now
12461 reflects the various checks performed. Applications can decide
12462 whether to tolerate certain situations such as an absent nonce
12463 in a response when one was present in a request: the ocsp application
12464 just prints out a warning. New function OCSP_add1_basic_nonce()
12465 this is to allow responders to include a nonce in a response even if
12466 the request is nonce-less.
12470 * Disable stdin buffering in load_cert (apps/apps.c) so that no certs are
12471 skipped when using openssl x509 multiple times on a single input file,
12472 e.g. `(openssl x509 -out cert1; openssl x509 -out cert2) <certs`.
12476 * Make ASN1_UTCTIME_set_string() and ASN1_GENERALIZEDTIME_set_string()
12477 set string type: to handle setting ASN1_TIME structures. Fix ca
12478 utility to correctly initialize revocation date of CRLs.
12482 * New option SSL_OP_CIPHER_SERVER_PREFERENCE allows the server to override
12483 the clients preferred ciphersuites and rather use its own preferences.
12484 Should help to work around M$ SGC (Server Gated Cryptography) bug in
12485 Internet Explorer by ensuring unchanged hash method during stepup.
12486 (Also replaces the broken/deactivated SSL_OP_NON_EXPORT_FIRST option.)
12490 * Make mkdef.pl recognise all DECLARE_ASN1 macros, change rijndael
12491 to aes and add a new 'exist' option to print out symbols that don't
12496 * Additional options to ocsp utility to allow flags to be set and
12497 additional certificates supplied.
12501 * Add the option -VAfile to 'openssl ocsp', so the user can give the
12502 OCSP client a number of certificate to only verify the response
12507 * Update Rijndael code to version 3.0 and change EVP AES ciphers to
12508 handle the new API. Currently only ECB, CBC modes supported. Add new
12511 Add TLS AES ciphersuites as described in RFC3268, "Advanced
12512 Encryption Standard (AES) Ciphersuites for Transport Layer
12513 Security (TLS)". (In beta versions of OpenSSL 0.9.7, these were
12514 not enabled by default and were not part of the "ALL" ciphersuite
12515 alias because they were not yet official; they could be
12516 explicitly requested by specifying the "AESdraft" ciphersuite
12517 group alias. In the final release of OpenSSL 0.9.7, the group
12518 alias is called "AES" and is part of "ALL".)
12520 *Ben Laurie, Steve Henson, Bodo Moeller*
12522 * New function OCSP_copy_nonce() to copy nonce value (if present) from
12523 request to response.
12527 * Functions for OCSP responders. OCSP_request_onereq_count(),
12528 OCSP_request_onereq_get0(), OCSP_onereq_get0_id() and OCSP_id_get0_info()
12529 extract information from a certificate request. OCSP_response_create()
12530 creates a response and optionally adds a basic response structure.
12531 OCSP_basic_add1_status() adds a complete single response to a basic
12532 response and returns the OCSP_SINGLERESP structure just added (to allow
12533 extensions to be included for example). OCSP_basic_add1_cert() adds a
12534 certificate to a basic response and OCSP_basic_sign() signs a basic
12535 response with various flags. New helper functions ASN1_TIME_check()
12536 (checks validity of ASN1_TIME structure) and ASN1_TIME_to_generalizedtime()
12537 (converts ASN1_TIME to GeneralizedTime).
12541 * Various new functions. EVP_Digest() combines EVP_Digest{Init,Update,Final}()
12542 in a single operation. X509_get0_pubkey_bitstr() extracts the public_key
12543 structure from a certificate. X509_pubkey_digest() digests the public_key
12544 contents: this is used in various key identifiers.
12548 * Make sk_sort() tolerate a NULL argument.
12550 *Steve Henson reported by Massimiliano Pala <madwolf@comune.modena.it>*
12552 * New OCSP verify flag OCSP_TRUSTOTHER. When set the "other" certificates
12553 passed by the function are trusted implicitly. If any of them signed the
12554 response then it is assumed to be valid and is not verified.
12558 * In PKCS7_set_type() initialise content_type in PKCS7_ENC_CONTENT
12559 to data. This was previously part of the PKCS7 ASN1 code. This
12560 was causing problems with OpenSSL created PKCS#12 and PKCS#7 structures.
12561 *Steve Henson, reported by Kenneth R. Robinette
12562 <support@securenetterm.com>*
12564 * Add CRYPTO_push_info() and CRYPTO_pop_info() calls to new ASN1
12565 routines: without these tracing memory leaks is very painful.
12566 Fix leaks in PKCS12 and PKCS7 routines.
12570 * Make X509_time_adj() cope with the new behaviour of ASN1_TIME_new().
12571 Previously it initialised the 'type' argument to V_ASN1_UTCTIME which
12572 effectively meant GeneralizedTime would never be used. Now it
12573 is initialised to -1 but X509_time_adj() now has to check the value
12574 and use ASN1_TIME_set() if the value is not V_ASN1_UTCTIME or
12575 V_ASN1_GENERALIZEDTIME, without this it always uses GeneralizedTime.
12576 *Steve Henson, reported by Kenneth R. Robinette
12577 <support@securenetterm.com>*
12579 * Fixes to BN_to_ASN1_INTEGER when bn is zero. This would previously
12580 result in a zero length in the ASN1_INTEGER structure which was
12581 not consistent with the structure when d2i_ASN1_INTEGER() was used
12582 and would cause ASN1_INTEGER_cmp() to fail. Enhance s2i_ASN1_INTEGER()
12583 to cope with hex and negative integers. Fix bug in i2a_ASN1_INTEGER()
12584 where it did not print out a minus for negative ASN1_INTEGER.
12588 * Add summary printout to ocsp utility. The various functions which
12589 convert status values to strings have been renamed to:
12590 OCSP_response_status_str(), OCSP_cert_status_str() and
12591 OCSP_crl_reason_str() and are no longer static. New options
12592 to verify nonce values and to disable verification. OCSP response
12593 printout format cleaned up.
12597 * Add additional OCSP certificate checks. These are those specified
12598 in RFC2560. This consists of two separate checks: the CA of the
12599 certificate being checked must either be the OCSP signer certificate
12600 or the issuer of the OCSP signer certificate. In the latter case the
12601 OCSP signer certificate must contain the OCSP signing extended key
12602 usage. This check is performed by attempting to match the OCSP
12603 signer or the OCSP signer CA to the issuerNameHash and issuerKeyHash
12604 in the OCSP_CERTID structures of the response.
12608 * Initial OCSP certificate verification added to OCSP_basic_verify()
12609 and related routines. This uses the standard OpenSSL certificate
12610 verify routines to perform initial checks (just CA validity) and
12611 to obtain the certificate chain. Then additional checks will be
12612 performed on the chain. Currently the root CA is checked to see
12613 if it is explicitly trusted for OCSP signing. This is used to set
12614 a root CA as a global signing root: that is any certificate that
12615 chains to that CA is an acceptable OCSP signing certificate.
12619 * New '-extfile ...' option to 'openssl ca' for reading X.509v3
12620 extensions from a separate configuration file.
12621 As when reading extensions from the main configuration file,
12622 the '-extensions ...' option may be used for specifying the
12625 *Massimiliano Pala <madwolf@comune.modena.it>*
12627 * New OCSP utility. Allows OCSP requests to be generated or
12628 read. The request can be sent to a responder and the output
12629 parsed, outputted or printed in text form. Not complete yet:
12630 still needs to check the OCSP response validity.
12634 * New subcommands for 'openssl ca':
12635 `openssl ca -status <serial>` prints the status of the cert with
12636 the given serial number (according to the index file).
12637 `openssl ca -updatedb` updates the expiry status of certificates
12640 *Massimiliano Pala <madwolf@comune.modena.it>*
12642 * New '-newreq-nodes' command option to CA.pl. This is like
12643 '-newreq', but calls 'openssl req' with the '-nodes' option
12644 so that the resulting key is not encrypted.
12646 *Damien Miller <djm@mindrot.org>*
12648 * New configuration for the GNU Hurd.
12650 *Jonathan Bartlett <johnnyb@wolfram.com> via Richard Levitte*
12652 * Initial code to implement OCSP basic response verify. This
12653 is currently incomplete. Currently just finds the signer's
12654 certificate and verifies the signature on the response.
12658 * New SSLeay_version code SSLEAY_DIR to determine the compiled-in
12659 value of OPENSSLDIR. This is available via the new '-d' option
12660 to 'openssl version', and is also included in 'openssl version -a'.
12664 * Allowing defining memory allocation callbacks that will be given
12665 file name and line number information in additional arguments
12666 (a `const char*` and an int). The basic functionality remains, as
12667 well as the original possibility to just replace malloc(),
12668 realloc() and free() by functions that do not know about these
12669 additional arguments. To register and find out the current
12670 settings for extended allocation functions, the following
12671 functions are provided:
12673 CRYPTO_set_mem_ex_functions
12674 CRYPTO_set_locked_mem_ex_functions
12675 CRYPTO_get_mem_ex_functions
12676 CRYPTO_get_locked_mem_ex_functions
12678 These work the same way as CRYPTO_set_mem_functions and friends.
12679 `CRYPTO_get_[locked_]mem_functions` now writes 0 where such an
12680 extended allocation function is enabled.
12681 Similarly, `CRYPTO_get_[locked_]mem_ex_functions` writes 0 where
12682 a conventional allocation function is enabled.
12684 *Richard Levitte, Bodo Moeller*
12686 * Finish off removing the remaining LHASH function pointer casts.
12687 There should no longer be any prototype-casting required when using
12688 the LHASH abstraction, and any casts that remain are "bugs". See
12689 the callback types and macros at the head of lhash.h for details
12690 (and "OBJ_cleanup" in crypto/objects/obj_dat.c as an example).
12694 * Add automatic query of EGD sockets in RAND_poll() for the unix variant.
12695 If /dev/[u]random devices are not available or do not return enough
12696 entropy, EGD style sockets (served by EGD or PRNGD) will automatically
12698 The locations /var/run/egd-pool, /dev/egd-pool, /etc/egd-pool, and
12699 /etc/entropy will be queried once each in this sequence, querying stops
12700 when enough entropy was collected without querying more sockets.
12704 * Change the Unix RAND_poll() variant to be able to poll several
12705 random devices, as specified by DEVRANDOM, until a sufficient amount
12706 of data has been collected. We spend at most 10 ms on each file
12707 (select timeout) and read in non-blocking mode. DEVRANDOM now
12708 defaults to the list "/dev/urandom", "/dev/random", "/dev/srandom"
12709 (previously it was just the string "/dev/urandom"), so on typical
12710 platforms the 10 ms delay will never occur.
12711 Also separate out the Unix variant to its own file, rand_unix.c.
12712 For VMS, there's a currently-empty rand_vms.c.
12716 * Move OCSP client related routines to ocsp_cl.c. These
12717 provide utility functions which an application needing
12718 to issue a request to an OCSP responder and analyse the
12719 response will typically need: as opposed to those which an
12720 OCSP responder itself would need which will be added later.
12722 OCSP_request_sign() signs an OCSP request with an API similar
12723 to PKCS7_sign(). OCSP_response_status() returns status of OCSP
12724 response. OCSP_response_get1_basic() extracts basic response
12725 from response. OCSP_resp_find_status(): finds and extracts status
12726 information from an OCSP_CERTID structure (which will be created
12727 when the request structure is built). These are built from lower
12728 level functions which work on OCSP_SINGLERESP structures but
12729 won't normally be used unless the application wishes to examine
12730 extensions in the OCSP response for example.
12732 Replace nonce routines with a pair of functions.
12733 OCSP_request_add1_nonce() adds a nonce value and optionally
12734 generates a random value. OCSP_check_nonce() checks the
12735 validity of the nonce in an OCSP response.
12739 * Change function OCSP_request_add() to OCSP_request_add0_id().
12740 This doesn't copy the supplied OCSP_CERTID and avoids the
12741 need to free up the newly created id. Change return type
12742 to OCSP_ONEREQ to return the internal OCSP_ONEREQ structure.
12743 This can then be used to add extensions to the request.
12744 Deleted OCSP_request_new(), since most of its functionality
12745 is now in OCSP_REQUEST_new() (and the case insensitive name
12746 clash) apart from the ability to set the request name which
12747 will be added elsewhere.
12751 * Update OCSP API. Remove obsolete extensions argument from
12752 various functions. Extensions are now handled using the new
12753 OCSP extension code. New simple OCSP HTTP function which
12754 can be used to send requests and parse the response.
12758 * Fix the PKCS#7 (S/MIME) code to work with new ASN1. Two new
12759 ASN1_ITEM structures help with sign and verify. PKCS7_ATTR_SIGN
12760 uses the special reorder version of SET OF to sort the attributes
12761 and reorder them to match the encoded order. This resolves a long
12762 standing problem: a verify on a PKCS7 structure just after signing
12763 it used to fail because the attribute order did not match the
12764 encoded order. PKCS7_ATTR_VERIFY does not reorder the attributes:
12765 it uses the received order. This is necessary to tolerate some broken
12766 software that does not order SET OF. This is handled by encoding
12767 as a SEQUENCE OF but using implicit tagging (with UNIVERSAL class)
12768 to produce the required SET OF.
12772 * Have mk1mf.pl generate the macros OPENSSL_BUILD_SHLIBCRYPTO and
12773 OPENSSL_BUILD_SHLIBSSL and use them appropriately in the header
12774 files to get correct declarations of the ASN.1 item variables.
12778 * Rewrite of PKCS#12 code to use new ASN1 functionality. Replace many
12779 PKCS#12 macros with real functions. Fix two unrelated ASN1 bugs:
12780 asn1_check_tlen() would sometimes attempt to use 'ctx' when it was
12781 NULL and ASN1_TYPE was not dereferenced properly in asn1_ex_c2i().
12782 New ASN1 macro: DECLARE_ASN1_ITEM() which just declares the relevant
12783 ASN1_ITEM and no wrapper functions.
12787 * New functions or ASN1_item_d2i_fp() and ASN1_item_d2i_bio(). These
12788 replace the old function pointer based I/O routines. Change most of
12789 the `*_d2i_bio()` and `*_d2i_fp()` functions to use these.
12793 * Enhance mkdef.pl to be more accepting about spacing in C preprocessor
12794 lines, recognize more "algorithms" that can be deselected, and make
12795 it complain about algorithm deselection that isn't recognised.
12799 * New ASN1 functions to handle dup, sign, verify, digest, pack and
12800 unpack operations in terms of ASN1_ITEM. Modify existing wrappers
12801 to use new functions. Add NO_ASN1_OLD which can be set to remove
12802 some old style ASN1 functions: this can be used to determine if old
12803 code will still work when these eventually go away.
12807 * New extension functions for OCSP structures, these follow the
12808 same conventions as certificates and CRLs.
12812 * New function X509V3_add1_i2d(). This automatically encodes and
12813 adds an extension. Its behaviour can be customised with various
12814 flags to append, replace or delete. Various wrappers added for
12815 certificates and CRLs.
12819 * Fix to avoid calling the underlying ASN1 print routine when
12820 an extension cannot be parsed. Correct a typo in the
12821 OCSP_SERVICELOC extension. Tidy up print OCSP format.
12825 * Make mkdef.pl parse some of the ASN1 macros and add appropriate
12826 entries for variables.
12830 * Add functionality to apps/openssl.c for detecting locking
12831 problems: As the program is single-threaded, all we have
12832 to do is register a locking callback using an array for
12833 storing which locks are currently held by the program.
12837 * Use a lock around the call to CRYPTO_get_ex_new_index() in
12838 SSL_get_ex_data_X509_STORE_idx(), which is used in
12839 ssl_verify_cert_chain() and thus can be called at any time
12840 during TLS/SSL handshakes so that thread-safety is essential.
12841 Unfortunately, the ex_data design is not at all suited
12842 for multi-threaded use, so it probably should be abolished.
12846 * Added Broadcom "ubsec" ENGINE to OpenSSL.
12848 *Broadcom, tweaked and integrated by Geoff Thorpe*
12850 * Move common extension printing code to new function
12851 X509V3_print_extensions(). Reorganise OCSP print routines and
12852 implement some needed OCSP ASN1 functions. Add OCSP extensions.
12856 * New function X509_signature_print() to remove duplication in some
12861 * Add a special meaning when SET OF and SEQUENCE OF flags are both
12862 set (this was treated exactly the same as SET OF previously). This
12863 is used to reorder the STACK representing the structure to match the
12864 encoding. This will be used to get round a problem where a PKCS7
12865 structure which was signed could not be verified because the STACK
12866 order did not reflect the encoded order.
12870 * Reimplement the OCSP ASN1 module using the new code.
12874 * Update the X509V3 code to permit the use of an ASN1_ITEM structure
12875 for its ASN1 operations. The old style function pointers still exist
12876 for now but they will eventually go away.
12880 * Merge in replacement ASN1 code from the ASN1 branch. This almost
12881 completely replaces the old ASN1 functionality with a table driven
12882 encoder and decoder which interprets an ASN1_ITEM structure describing
12883 the ASN1 module. Compatibility with the existing ASN1 API (i2d,d2i) is
12884 largely maintained. Almost all of the old asn1_mac.h macro based ASN1
12885 has also been converted to the new form.
12889 * Change BN_mod_exp_recp so that negative moduli are tolerated
12890 (the sign is ignored). Similarly, ignore the sign in BN_MONT_CTX_set
12891 so that BN_mod_exp_mont and BN_mod_exp_mont_word work
12892 for negative moduli.
12896 * Fix BN_uadd and BN_usub: Always return non-negative results instead
12897 of not touching the result's sign bit.
12901 * BN_div bugfix: If the result is 0, the sign (res->neg) must not be
12906 * Changed the LHASH code to use prototypes for callbacks, and created
12907 macros to declare and implement thin (optionally static) functions
12908 that provide type-safety and avoid function pointer casting for the
12909 type-specific callbacks.
12913 * Added Kerberos Cipher Suites to be used with TLS, as written in
12915 *Veers Staats <staatsvr@asc.hpc.mil>,
12916 Jeffrey Altman <jaltman@columbia.edu>, via Richard Levitte*
12918 * Reformat the FAQ so the different questions and answers can be divided
12919 in sections depending on the subject.
12923 * Have the zlib compression code load ZLIB.DLL dynamically under
12928 * New function BN_mod_sqrt for computing square roots modulo a prime
12929 (using the probabilistic Tonelli-Shanks algorithm unless
12930 p == 3 (mod 4) or p == 5 (mod 8), which are cases that can
12931 be handled deterministically).
12933 *Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller*
12935 * Make BN_mod_inverse faster by explicitly handling small quotients
12936 in the Euclid loop. (Speed gain about 20% for small moduli [256 or
12937 512 bits], about 30% for larger ones [1024 or 2048 bits].)
12941 * New function BN_kronecker.
12945 * Fix BN_gcd so that it works on negative inputs; the result is
12946 positive unless both parameters are zero.
12947 Previously something reasonably close to an infinite loop was
12948 possible because numbers could be growing instead of shrinking
12949 in the implementation of Euclid's algorithm.
12953 * Fix BN_is_word() and BN_is_one() macros to take into account the
12954 sign of the number in question.
12956 Fix BN_is_word(a,w) to work correctly for w == 0.
12958 The old BN_is_word(a,w) macro is now called BN_abs_is_word(a,w)
12959 because its test if the absolute value of 'a' equals 'w'.
12960 Note that BN_abs_is_word does *not* handle w == 0 reliably;
12961 it exists mostly for use in the implementations of BN_is_zero(),
12962 BN_is_one(), and BN_is_word().
12966 * New function BN_swap.
12970 * Use BN_nnmod instead of BN_mod in crypto/bn/bn_exp.c so that
12971 the exponentiation functions are more likely to produce reasonable
12972 results on negative inputs.
12976 * Change BN_mod_mul so that the result is always non-negative.
12977 Previously, it could be negative if one of the factors was negative;
12978 I don't think anyone really wanted that behaviour.
12982 * Move `BN_mod_...` functions into new file `crypto/bn/bn_mod.c`
12983 (except for exponentiation, which stays in `crypto/bn/bn_exp.c`,
12984 and `BN_mod_mul_reciprocal`, which stays in `crypto/bn/bn_recp.c`)
12985 and add new functions:
12994 BN_mod_lshift1_quick
12996 BN_mod_lshift_quick
12998 These functions always generate non-negative results.
13000 `BN_nnmod` otherwise is `like BN_mod` (if `BN_mod` computes a remainder `r`
13001 such that `|m| < r < 0`, `BN_nnmod` will output `rem + |m|` instead).
13003 `BN_mod_XXX_quick(r, a, [b,] m)` generates the same result as
13004 `BN_mod_XXX(r, a, [b,] m, ctx)`, but requires that `a` [and `b`]
13005 be reduced modulo `m`.
13007 *Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller*
13010 The following entry accidentally appeared in the CHANGES file
13011 distributed with OpenSSL 0.9.7. The modifications described in
13012 it do *not* apply to OpenSSL 0.9.7.
13014 * Remove a few calls to bn_wexpand() in BN_sqr() (the one in there
13015 was actually never needed) and in BN_mul(). The removal in BN_mul()
13016 required a small change in bn_mul_part_recursive() and the addition
13017 of the functions bn_cmp_part_words(), bn_sub_part_words() and
13018 bn_add_part_words(), which do the same thing as bn_cmp_words(),
13019 bn_sub_words() and bn_add_words() except they take arrays with
13025 * In 'openssl passwd', verify passwords read from the terminal
13026 unless the '-salt' option is used (which usually means that
13027 verification would just waste user's time since the resulting
13028 hash is going to be compared with some given password hash)
13029 or the new '-noverify' option is used.
13031 This is an incompatible change, but it does not affect
13032 non-interactive use of 'openssl passwd' (passwords on the command
13033 line, '-stdin' option, '-in ...' option) and thus should not
13034 cause any problems.
13038 * Remove all references to RSAref, since there's no more need for it.
13042 * Make DSO load along a path given through an environment variable
13043 (SHLIB_PATH) with shl_load().
13047 * Constify the ENGINE code as a result of BIGNUM constification.
13048 Also constify the RSA code and most things related to it. In a
13049 few places, most notable in the depth of the ASN.1 code, ugly
13050 casts back to non-const were required (to be solved at a later
13055 * Make it so the openssl application has all engines loaded by default.
13059 * Constify the BIGNUM routines a little more.
13063 * Add the following functions:
13065 ENGINE_load_cswift()
13067 ENGINE_load_atalla()
13068 ENGINE_load_nuron()
13069 ENGINE_load_builtin_engines()
13071 That way, an application can itself choose if external engines that
13072 are built-in in OpenSSL shall ever be used or not. The benefit is
13073 that applications won't have to be linked with libdl or other dso
13074 libraries unless it's really needed.
13076 Changed 'openssl engine' to load all engines on demand.
13077 Changed the engine header files to avoid the duplication of some
13078 declarations (they differed!).
13082 * 'openssl engine' can now list capabilities.
13086 * Better error reporting in 'openssl engine'.
13090 * Never call load_dh_param(NULL) in s_server.
13094 * Add engine application. It can currently list engines by name and
13095 identity, and test if they are actually available.
13099 * Improve RPM specification file by forcing symbolic linking and making
13100 sure the installed documentation is also owned by root.root.
13102 *Damien Miller <djm@mindrot.org>*
13104 * Give the OpenSSL applications more possibilities to make use of
13105 keys (public as well as private) handled by engines.
13109 * Add OCSP code that comes from CertCo.
13113 * Add VMS support for the Rijndael code.
13117 * Added untested support for Nuron crypto accelerator.
13121 * Add support for external cryptographic devices. This code was
13122 previously distributed separately as the "engine" branch.
13124 *Geoff Thorpe, Richard Levitte*
13126 * Rework the filename-translation in the DSO code. It is now possible to
13127 have far greater control over how a "name" is turned into a filename
13128 depending on the operating environment and any oddities about the
13129 different shared library filenames on each system.
13133 * Support threads on FreeBSD-elf in Configure.
13137 * Fix for SHA1 assembly problem with MASM: it produces
13138 warnings about corrupt line number information when assembling
13139 with debugging information. This is caused by the overlapping
13142 *Bernd Matthes <mainbug@celocom.de>, Steve Henson*
13145 NCONF_get_number() has no error checking at all. As a replacement,
13146 NCONF_get_number_e() is defined (`_e` for "error checking") and is
13147 promoted strongly. The old NCONF_get_number is kept around for
13148 binary backward compatibility.
13149 Make it possible for methods to load from something other than a BIO,
13150 by providing a function pointer that is given a name instead of a BIO.
13151 For example, this could be used to load configuration data from an
13156 * Fix for non blocking accept BIOs. Added new I/O special reason
13157 BIO_RR_ACCEPT to cover this case. Previously use of accept BIOs
13158 with non blocking I/O was not possible because no retry code was
13159 implemented. Also added new SSL code SSL_WANT_ACCEPT to cover
13164 * Added the beginnings of Rijndael support.
13168 * Fix for bug in DirectoryString mask setting. Add support for
13169 X509_NAME_print_ex() in 'req' and X509_print_ex() function
13170 to allow certificate printing to more controllable, additional
13171 'certopt' option to 'x509' to allow new printing options to be
13176 * Clean old EAY MD5 hack from e_os.h.
13180 ### Changes between 0.9.6l and 0.9.6m [17 Mar 2004]
13182 * Fix null-pointer assignment in do_change_cipher_spec() revealed
13183 by using the Codenomicon TLS Test Tool [CVE-2004-0079][]
13185 *Joe Orton, Steve Henson*
13187 ### Changes between 0.9.6k and 0.9.6l [04 Nov 2003]
13189 * Fix additional bug revealed by the NISCC test suite:
13191 Stop bug triggering large recursion when presented with
13192 certain ASN.1 tags [CVE-2003-0851][]
13196 ### Changes between 0.9.6j and 0.9.6k [30 Sep 2003]
13198 * Fix various bugs revealed by running the NISCC test suite:
13200 Stop out of bounds reads in the ASN1 code when presented with
13201 invalid tags (CVE-2003-0543 and CVE-2003-0544).
13203 If verify callback ignores invalid public key errors don't try to check
13204 certificate signature with the NULL public key.
13208 * In ssl3_accept() (ssl/s3_srvr.c) only accept a client certificate
13209 if the server requested one: as stated in TLS 1.0 and SSL 3.0
13214 * In ssl3_get_client_hello() (ssl/s3_srvr.c), tolerate additional
13215 extra data after the compression methods not only for TLS 1.0
13216 but also for SSL 3.0 (as required by the specification).
13218 *Bodo Moeller; problem pointed out by Matthias Loepfe*
13220 * Change X509_certificate_type() to mark the key as exported/exportable
13221 when it's 512 *bits* long, not 512 bytes.
13225 ### Changes between 0.9.6i and 0.9.6j [10 Apr 2003]
13227 * Countermeasure against the Klima-Pokorny-Rosa extension of
13228 Bleichbacher's attack on PKCS #1 v1.5 padding: treat
13229 a protocol version number mismatch like a decryption error
13230 in ssl3_get_client_key_exchange (ssl/s3_srvr.c).
13234 * Turn on RSA blinding by default in the default implementation
13235 to avoid a timing attack. Applications that don't want it can call
13236 RSA_blinding_off() or use the new flag RSA_FLAG_NO_BLINDING.
13237 They would be ill-advised to do so in most cases.
13239 *Ben Laurie, Steve Henson, Geoff Thorpe, Bodo Moeller*
13241 * Change RSA blinding code so that it works when the PRNG is not
13242 seeded (in this case, the secret RSA exponent is abused as
13243 an unpredictable seed -- if it is not unpredictable, there
13244 is no point in blinding anyway). Make RSA blinding thread-safe
13245 by remembering the creator's thread ID in rsa->blinding and
13246 having all other threads use local one-time blinding factors
13247 (this requires more computation than sharing rsa->blinding, but
13248 avoids excessive locking; and if an RSA object is not shared
13249 between threads, blinding will still be very fast).
13253 ### Changes between 0.9.6h and 0.9.6i [19 Feb 2003]
13255 * In ssl3_get_record (ssl/s3_pkt.c), minimize information leaked
13256 via timing by performing a MAC computation even if incorrect
13257 block cipher padding has been found. This is a countermeasure
13258 against active attacks where the attacker has to distinguish
13259 between bad padding and a MAC verification error. [CVE-2003-0078][]
13261 *Bodo Moeller; problem pointed out by Brice Canvel (EPFL),
13262 Alain Hiltgen (UBS), Serge Vaudenay (EPFL), and
13263 Martin Vuagnoux (EPFL, Ilion)*
13265 ### Changes between 0.9.6g and 0.9.6h [5 Dec 2002]
13267 * New function OPENSSL_cleanse(), which is used to cleanse a section of
13268 memory from its contents. This is done with a counter that will
13269 place alternating values in each byte. This can be used to solve
13270 two issues: 1) the removal of calls to memset() by highly optimizing
13271 compilers, and 2) cleansing with other values than 0, since those can
13272 be read through on certain media, for example a swap space on disk.
13276 * Bugfix: client side session caching did not work with external caching,
13277 because the session->cipher setting was not restored when reloading
13278 from the external cache. This problem was masked, when
13279 SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG (part of SSL_OP_ALL) was set.
13280 (Found by Steve Haslam <steve@araqnid.ddts.net>.)
13284 * Fix client_certificate (ssl/s2_clnt.c): The permissible total
13285 length of the REQUEST-CERTIFICATE message is 18 .. 34, not 17 .. 33.
13287 *Zeev Lieber <zeev-l@yahoo.com>*
13289 * Undo an undocumented change introduced in 0.9.6e which caused
13290 repeated calls to OpenSSL_add_all_ciphers() and
13291 OpenSSL_add_all_digests() to be ignored, even after calling
13296 * Change the default configuration reader to deal with last line not
13297 being properly terminated.
13301 * Change X509_NAME_cmp() so it applies the special rules on handling
13302 DN values that are of type PrintableString, as well as RDNs of type
13303 emailAddress where the value has the type ia5String.
13305 *stefank@valicert.com via Richard Levitte*
13307 * Add a SSL_SESS_CACHE_NO_INTERNAL_STORE flag to take over half
13308 the job SSL_SESS_CACHE_NO_INTERNAL_LOOKUP was inconsistently
13309 doing, define a new flag (SSL_SESS_CACHE_NO_INTERNAL) to be
13310 the bitwise-OR of the two for use by the majority of applications
13311 wanting this behaviour, and update the docs. The documented
13312 behaviour and actual behaviour were inconsistent and had been
13313 changing anyway, so this is more a bug-fix than a behavioural
13316 *Geoff Thorpe, diagnosed by Nadav Har'El*
13318 * Don't impose a 16-byte length minimum on session IDs in ssl/s3_clnt.c
13319 (the SSL 3.0 and TLS 1.0 specifications allow any length up to 32 bytes).
13323 * Fix initialization code race conditions in
13324 SSLv23_method(), SSLv23_client_method(), SSLv23_server_method(),
13325 SSLv2_method(), SSLv2_client_method(), SSLv2_server_method(),
13326 SSLv3_method(), SSLv3_client_method(), SSLv3_server_method(),
13327 TLSv1_method(), TLSv1_client_method(), TLSv1_server_method(),
13328 ssl2_get_cipher_by_char(),
13329 ssl3_get_cipher_by_char().
13331 *Patrick McCormick <patrick@tellme.com>, Bodo Moeller*
13333 * Reorder cleanup sequence in SSL_CTX_free(): only remove the ex_data after
13334 the cached sessions are flushed, as the remove_cb() might use ex_data
13335 contents. Bug found by Sam Varshavchik <mrsam@courier-mta.com>
13336 (see [openssl.org #212]).
13338 *Geoff Thorpe, Lutz Jaenicke*
13340 * Fix typo in OBJ_txt2obj which incorrectly passed the content
13341 length, instead of the encoding length to d2i_ASN1_OBJECT.
13345 ### Changes between 0.9.6f and 0.9.6g [9 Aug 2002]
13347 * [In 0.9.6g-engine release:]
13348 Fix crypto/engine/vendor_defns/cswift.h for WIN32 (use `_stdcall`).
13350 *Lynn Gazis <lgazis@rainbow.com>*
13352 ### Changes between 0.9.6e and 0.9.6f [8 Aug 2002]
13354 * Fix ASN1 checks. Check for overflow by comparing with LONG_MAX
13355 and get fix the header length calculation.
13356 *Florian Weimer <Weimer@CERT.Uni-Stuttgart.DE>,
13357 Alon Kantor <alonk@checkpoint.com> (and others), Steve Henson*
13359 * Use proper error handling instead of 'assertions' in buffer
13360 overflow checks added in 0.9.6e. This prevents DoS (the
13361 assertions could call abort()).
13363 *Arne Ansper <arne@ats.cyber.ee>, Bodo Moeller*
13365 ### Changes between 0.9.6d and 0.9.6e [30 Jul 2002]
13367 * Add various sanity checks to asn1_get_length() to reject
13368 the ASN1 length bytes if they exceed sizeof(long), will appear
13369 negative or the content length exceeds the length of the
13372 *Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>*
13374 * Fix cipher selection routines: ciphers without encryption had no flags
13375 for the cipher strength set and where therefore not handled correctly
13376 by the selection routines (PR #130).
13380 * Fix EVP_dsa_sha macro.
13385 SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS
13386 for disabling the SSL 3.0/TLS 1.0 CBC vulnerability countermeasure
13387 that was added in OpenSSL 0.9.6d.
13389 As the countermeasure turned out to be incompatible with some
13390 broken SSL implementations, the new option is part of SSL_OP_ALL.
13391 SSL_OP_ALL is usually employed when compatibility with weird SSL
13392 implementations is desired (e.g. '-bugs' option to 's_client' and
13393 's_server'), so the new option is automatically set in many
13398 * Changes in security patch:
13400 Changes marked "(CHATS)" were sponsored by the Defense Advanced
13401 Research Projects Agency (DARPA) and Air Force Research Laboratory,
13402 Air Force Materiel Command, USAF, under agreement number
13405 * Add various sanity checks to asn1_get_length() to reject
13406 the ASN1 length bytes if they exceed sizeof(long), will appear
13407 negative or the content length exceeds the length of the
13408 supplied buffer. [CVE-2002-0659][]
13410 *Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>*
13412 * Assertions for various potential buffer overflows, not known to
13413 happen in practice.
13415 *Ben Laurie (CHATS)*
13417 * Various temporary buffers to hold ASCII versions of integers were
13418 too small for 64 bit platforms. [CVE-2002-0655][]
13419 *Matthew Byng-Maddick <mbm@aldigital.co.uk> and Ben Laurie (CHATS)>*
13421 * Remote buffer overflow in SSL3 protocol - an attacker could
13422 supply an oversized session ID to a client. [CVE-2002-0656][]
13424 *Ben Laurie (CHATS)*
13426 * Remote buffer overflow in SSL2 protocol - an attacker could
13427 supply an oversized client master key. [CVE-2002-0656][]
13429 *Ben Laurie (CHATS)*
13431 ### Changes between 0.9.6c and 0.9.6d [9 May 2002]
13433 * Fix crypto/asn1/a_sign.c so that 'parameters' is omitted (not
13434 encoded as NULL) with id-dsa-with-sha1.
13436 *Nils Larsch <nla@trustcenter.de>; problem pointed out by Bodo Moeller*
13438 * Check various `X509_...()` return values in apps/req.c.
13440 *Nils Larsch <nla@trustcenter.de>*
13442 * Fix BASE64 decode (EVP_DecodeUpdate) for data with CR/LF ended lines:
13443 an end-of-file condition would erroneously be flagged, when the CRLF
13444 was just at the end of a processed block. The bug was discovered when
13445 processing data through a buffering memory BIO handing the data to a
13446 BASE64-decoding BIO. Bug fund and patch submitted by Pavel Tsekov
13447 <ptsekov@syntrex.com> and Nedelcho Stanev.
13451 * Implement a countermeasure against a vulnerability recently found
13452 in CBC ciphersuites in SSL 3.0/TLS 1.0: Send an empty fragment
13453 before application data chunks to avoid the use of known IVs
13454 with data potentially chosen by the attacker.
13458 * Fix length checks in ssl3_get_client_hello().
13462 * TLS/SSL library bugfix: use s->s3->in_read_app_data differently
13463 to prevent ssl3_read_internal() from incorrectly assuming that
13464 ssl3_read_bytes() found application data while handshake
13465 processing was enabled when in fact s->s3->in_read_app_data was
13466 merely automatically cleared during the initial handshake.
13468 *Bodo Moeller; problem pointed out by Arne Ansper <arne@ats.cyber.ee>*
13470 * Fix object definitions for Private and Enterprise: they were not
13471 recognized in their shortname (=lowercase) representation. Extend
13472 obj_dat.pl to issue an error when using undefined keywords instead
13473 of silently ignoring the problem (Svenning Sorensen
13474 <sss@sss.dnsalias.net>).
13478 * Fix DH_generate_parameters() so that it works for 'non-standard'
13479 generators, i.e. generators other than 2 and 5. (Previously, the
13480 code did not properly initialise the 'add' and 'rem' values to
13481 BN_generate_prime().)
13483 In the new general case, we do not insist that 'generator' is
13484 actually a primitive root: This requirement is rather pointless;
13485 a generator of the order-q subgroup is just as good, if not
13490 * Map new X509 verification errors to alerts. Discovered and submitted by
13491 Tom Wu <tom@arcot.com>.
13495 * Fix ssl3_pending() (ssl/s3_lib.c) to prevent SSL_pending() from
13496 returning non-zero before the data has been completely received
13497 when using non-blocking I/O.
13499 *Bodo Moeller; problem pointed out by John Hughes*
13501 * Some of the ciphers missed the strength entry (SSL_LOW etc).
13503 *Ben Laurie, Lutz Jaenicke*
13505 * Fix bug in SSL_clear(): bad sessions were not removed (found by
13506 Yoram Zahavi <YoramZ@gilian.com>).
13510 * Add information about CygWin 1.3 and on, and preserve proper
13511 configuration for the versions before that.
13513 *Corinna Vinschen <vinschen@redhat.com> and Richard Levitte*
13515 * Make removal from session cache (SSL_CTX_remove_session()) more robust:
13516 check whether we deal with a copy of a session and do not delete from
13517 the cache in this case. Problem reported by "Izhar Shoshani Levi"
13518 <izhar@checkpoint.com>.
13522 * Do not store session data into the internal session cache, if it
13523 is never intended to be looked up (SSL_SESS_CACHE_NO_INTERNAL_LOOKUP
13524 flag is set). Proposed by Aslam <aslam@funk.com>.
13528 * Have ASN1_BIT_STRING_set_bit() really clear a bit when the requested
13533 * [In 0.9.6d-engine release:]
13534 Fix a crashbug and a logic bug in hwcrhk_load_pubkey().
13536 *Toomas Kiisk <vix@cyber.ee> via Richard Levitte*
13538 * Add the configuration target linux-s390x.
13540 *Neale Ferguson <Neale.Ferguson@SoftwareAG-USA.com> via Richard Levitte*
13542 * The earlier bugfix for the SSL3_ST_SW_HELLO_REQ_C case of
13543 ssl3_accept (ssl/s3_srvr.c) incorrectly used a local flag
13544 variable as an indication that a ClientHello message has been
13545 received. As the flag value will be lost between multiple
13546 invocations of ssl3_accept when using non-blocking I/O, the
13547 function may not be aware that a handshake has actually taken
13548 place, thus preventing a new session from being added to the
13551 To avoid this problem, we now set s->new_session to 2 instead of
13552 using a local variable.
13554 *Lutz Jaenicke, Bodo Moeller*
13556 * Bugfix: Return -1 from ssl3_get_server_done (ssl3/s3_clnt.c)
13557 if the SSL_R_LENGTH_MISMATCH error is detected.
13559 *Geoff Thorpe, Bodo Moeller*
13561 * New 'shared_ldflag' column in Configure platform table.
13565 * Fix EVP_CIPHER_mode macro.
13567 *"Dan S. Camper" <dan@bti.net>*
13569 * Fix ssl3_read_bytes (ssl/s3_pkt.c): To ignore messages of unknown
13570 type, we must throw them away by setting rr->length to 0.
13572 *D P Chang <dpc@qualys.com>*
13574 ### Changes between 0.9.6b and 0.9.6c [21 dec 2001]
13576 * Fix BN_rand_range bug pointed out by Dominikus Scherkl
13577 <Dominikus.Scherkl@biodata.com>. (The previous implementation
13578 worked incorrectly for those cases where range = `10..._2` and
13579 `3*range` is two bits longer than range.)
13583 * Only add signing time to PKCS7 structures if it is not already
13588 * Fix crypto/objects/objects.h: "ld-ce" should be "id-ce",
13589 OBJ_ld_ce should be OBJ_id_ce.
13590 Also some ip-pda OIDs in crypto/objects/objects.txt were
13591 incorrect (cf. RFC 3039).
13593 *Matt Cooper, Frederic Giudicelli, Bodo Moeller*
13595 * Release CRYPTO_LOCK_DYNLOCK when CRYPTO_destroy_dynlockid()
13596 returns early because it has nothing to do.
13598 *Andy Schneider <andy.schneider@bjss.co.uk>*
13600 * [In 0.9.6c-engine release:]
13601 Fix mutex callback return values in crypto/engine/hw_ncipher.c.
13603 *Andy Schneider <andy.schneider@bjss.co.uk>*
13605 * [In 0.9.6c-engine release:]
13606 Add support for Cryptographic Appliance's keyserver technology.
13607 (Use engine 'keyclient')
13609 *Cryptographic Appliances and Geoff Thorpe*
13611 * Add a configuration entry for OS/390 Unix. The C compiler 'c89'
13612 is called via tools/c89.sh because arguments have to be
13613 rearranged (all '-L' options must appear before the first object
13616 *Richard Shapiro <rshapiro@abinitio.com>*
13618 * [In 0.9.6c-engine release:]
13619 Add support for Broadcom crypto accelerator cards, backported
13622 *Broadcom, Nalin Dahyabhai <nalin@redhat.com>, Mark Cox*
13624 * [In 0.9.6c-engine release:]
13625 Add support for SureWare crypto accelerator cards from
13626 Baltimore Technologies. (Use engine 'sureware')
13628 *Baltimore Technologies and Mark Cox*
13630 * [In 0.9.6c-engine release:]
13631 Add support for crypto accelerator cards from Accelerated
13632 Encryption Processing, www.aep.ie. (Use engine 'aep')
13634 *AEP Inc. and Mark Cox*
13636 * Add a configuration entry for gcc on UnixWare.
13638 *Gary Benson <gbenson@redhat.com>*
13640 * Change ssl/s2_clnt.c and ssl/s2_srvr.c so that received handshake
13641 messages are stored in a single piece (fixed-length part and
13642 variable-length part combined) and fix various bugs found on the way.
13646 * Disable caching in BIO_gethostbyname(), directly use gethostbyname()
13647 instead. BIO_gethostbyname() does not know what timeouts are
13648 appropriate, so entries would stay in cache even when they have
13650 *Bodo Moeller; problem pointed out by Rich Salz <rsalz@zolera.com>*
13652 * Change ssl23_get_client_hello (ssl/s23_srvr.c) behaviour when
13653 faced with a pathologically small ClientHello fragment that does
13654 not contain client_version: Instead of aborting with an error,
13655 simply choose the highest available protocol version (i.e.,
13656 TLS 1.0 unless it is disabled). In practice, ClientHello
13657 messages are never sent like this, but this change gives us
13658 strictly correct behaviour at least for TLS.
13662 * Fix SSL handshake functions and SSL_clear() such that SSL_clear()
13663 never resets s->method to s->ctx->method when called from within
13664 one of the SSL handshake functions.
13666 *Bodo Moeller; problem pointed out by Niko Baric*
13668 * In ssl3_get_client_hello (ssl/s3_srvr.c), generate a fatal alert
13669 (sent using the client's version number) if client_version is
13670 smaller than the protocol version in use. Also change
13671 ssl23_get_client_hello (ssl/s23_srvr.c) to select TLS 1.0 if
13672 the client demanded SSL 3.0 but only TLS 1.0 is enabled; then
13673 the client will at least see that alert.
13677 * Fix ssl3_get_message (ssl/s3_both.c) to handle message fragmentation
13682 * Avoid infinite loop in ssl3_get_message (ssl/s3_both.c) if a
13683 client receives HelloRequest while in a handshake.
13685 *Bodo Moeller; bug noticed by Andy Schneider <andy.schneider@bjss.co.uk>*
13687 * Bugfix in ssl3_accept (ssl/s3_srvr.c): Case SSL3_ST_SW_HELLO_REQ_C
13688 should end in 'break', not 'goto end' which circumvents various
13689 cleanups done in state SSL_ST_OK. But session related stuff
13690 must be disabled for SSL_ST_OK in the case that we just sent a
13693 Also avoid some overhead by not calling ssl_init_wbio_buffer()
13694 before just sending a HelloRequest.
13696 *Bodo Moeller, Eric Rescorla <ekr@rtfm.com>*
13698 * Fix ssl/s3_enc.c, ssl/t1_enc.c and ssl/s3_pkt.c so that we don't
13699 reveal whether illegal block cipher padding was found or a MAC
13700 verification error occurred. (Neither SSLerr() codes nor alerts
13701 are directly visible to potential attackers, but the information
13702 may leak via logfiles.)
13704 Similar changes are not required for the SSL 2.0 implementation
13705 because the number of padding bytes is sent in clear for SSL 2.0,
13706 and the extra bytes are just ignored. However ssl/s2_pkt.c
13707 failed to verify that the purported number of padding bytes is in
13712 * Add OpenUNIX-8 support including shared libraries
13713 (Boyd Lynn Gerber <gerberb@zenez.com>).
13717 * Improve RSA_padding_check_PKCS1_OAEP() check again to avoid
13718 'wristwatch attack' using huge encoding parameters (cf.
13719 James H. Manger's CRYPTO 2001 paper). Note that the
13720 RSA_PKCS1_OAEP_PADDING case of RSA_private_decrypt() does not use
13721 encoding parameters and hence was not vulnerable.
13725 * BN_sqr() bug fix.
13727 *Ulf Möller, reported by Jim Ellis <jim.ellis@cavium.com>*
13729 * Rabin-Miller test analyses assume uniformly distributed witnesses,
13730 so use BN_pseudo_rand_range() instead of using BN_pseudo_rand()
13731 followed by modular reduction.
13733 *Bodo Moeller; pointed out by Adam Young <AYoung1@NCSUS.JNJ.COM>*
13735 * Add BN_pseudo_rand_range() with obvious functionality: BN_rand_range()
13736 equivalent based on BN_pseudo_rand() instead of BN_rand().
13740 * s3_srvr.c: allow sending of large client certificate lists (> 16 kB).
13741 This function was broken, as the check for a new client hello message
13742 to handle SGC did not allow these large messages.
13743 (Tracked down by "Douglas E. Engert" <deengert@anl.gov>.)
13747 * Add alert descriptions for TLSv1 to `SSL_alert_desc_string[_long]()`.
13751 * Fix buggy behaviour of BIO_get_num_renegotiates() and BIO_ctrl()
13752 for BIO_C_GET_WRITE_BUF_SIZE ("Stephen Hinton" <shinton@netopia.com>).
13756 * Rework the configuration and shared library support for Tru64 Unix.
13757 The configuration part makes use of modern compiler features and
13758 still retains old compiler behavior for those that run older versions
13759 of the OS. The shared library support part includes a variant that
13760 uses the RPATH feature, and is available through the special
13761 configuration target "alpha-cc-rpath", which will never be selected
13764 *Tim Mooney <mooney@dogbert.cc.ndsu.NoDak.edu> via Richard Levitte*
13766 * In ssl3_get_key_exchange (ssl/s3_clnt.c), call ssl3_get_message()
13767 with the same message size as in ssl3_get_certificate_request().
13768 Otherwise, if no ServerKeyExchange message occurs, CertificateRequest
13769 messages might inadvertently be reject as too long.
13771 *Petr Lampa <lampa@fee.vutbr.cz>*
13773 * Enhanced support for IA-64 Unix platforms (well, Linux and HP-UX).
13777 * Modified SSL library such that the verify_callback that has been set
13778 specifically for an SSL object with SSL_set_verify() is actually being
13779 used. Before the change, a verify_callback set with this function was
13780 ignored and the verify_callback() set in the SSL_CTX at the time of
13781 the call was used. New function X509_STORE_CTX_set_verify_cb() introduced
13782 to allow the necessary settings.
13786 * Initialize static variable in crypto/dsa/dsa_lib.c and crypto/dh/dh_lib.c
13787 explicitly to NULL, as at least on Solaris 8 this seems not always to be
13788 done automatically (in contradiction to the requirements of the C
13789 standard). This made problems when used from OpenSSH.
13793 * In OpenSSL 0.9.6a and 0.9.6b, crypto/dh/dh_key.c ignored
13794 dh->length and always used
13796 BN_rand_range(priv_key, dh->p).
13798 BN_rand_range() is not necessary for Diffie-Hellman, and this
13799 specific range makes Diffie-Hellman unnecessarily inefficient if
13800 dh->length (recommended exponent length) is much smaller than the
13801 length of dh->p. We could use BN_rand_range() if the order of
13802 the subgroup was stored in the DH structure, but we only have
13807 BN_rand(priv_key, l, ...)
13809 where 'l' is dh->length if this is defined, or BN_num_bits(dh->p)-1
13816 RSA_eay_public_encrypt
13817 RSA_eay_private_decrypt
13818 RSA_eay_private_encrypt (signing)
13819 RSA_eay_public_decrypt (signature verification)
13821 (default implementations for RSA_public_encrypt,
13822 RSA_private_decrypt, RSA_private_encrypt, RSA_public_decrypt),
13823 always reject numbers >= n.
13827 * In crypto/rand/md_rand.c, use a new short-time lock CRYPTO_LOCK_RAND2
13828 to synchronize access to 'locking_thread'. This is necessary on
13829 systems where access to 'locking_thread' (an 'unsigned long'
13830 variable) is not atomic.
13834 * In crypto/rand/md_rand.c, set 'locking_thread' to current thread's ID
13835 *before* setting the 'crypto_lock_rand' flag. The previous code had
13836 a race condition if 0 is a valid thread ID.
13838 *Travis Vitek <vitek@roguewave.com>*
13840 * Add support for shared libraries under Irix.
13842 *Albert Chin-A-Young <china@thewrittenword.com>*
13844 * Add configuration option to build on Linux on both big-endian and
13845 little-endian MIPS.
13847 *Ralf Baechle <ralf@uni-koblenz.de>*
13849 * Add the possibility to create shared libraries on HP-UX.
13853 ### Changes between 0.9.6a and 0.9.6b [9 Jul 2001]
13855 * Change ssleay_rand_bytes (crypto/rand/md_rand.c)
13856 to avoid a SSLeay/OpenSSL PRNG weakness pointed out by
13857 Markku-Juhani O. Saarinen <markku-juhani.saarinen@nokia.com>:
13858 PRNG state recovery was possible based on the output of
13859 one PRNG request appropriately sized to gain knowledge on
13860 'md' followed by enough consecutive 1-byte PRNG requests
13861 to traverse all of 'state'.
13863 1. When updating 'md_local' (the current thread's copy of 'md')
13864 during PRNG output generation, hash all of the previous
13865 'md_local' value, not just the half used for PRNG output.
13867 2. Make the number of bytes from 'state' included into the hash
13868 independent from the number of PRNG bytes requested.
13870 The first measure alone would be sufficient to avoid
13871 Markku-Juhani's attack. (Actually it had never occurred
13872 to me that the half of 'md_local' used for chaining was the
13873 half from which PRNG output bytes were taken -- I had always
13874 assumed that the secret half would be used.) The second
13875 measure makes sure that additional data from 'state' is never
13876 mixed into 'md_local' in small portions; this heuristically
13877 further strengthens the PRNG.
13881 * Fix crypto/bn/asm/mips3.s.
13885 * When only the key is given to "enc", the IV is undefined. Print out
13886 an error message in this case.
13890 * Handle special case when X509_NAME is empty in X509 printing routines.
13894 * In dsa_do_verify (crypto/dsa/dsa_ossl.c), verify that r and s are
13895 positive and less than q.
13899 * Don't change `*pointer` in CRYPTO_add_lock() is add_lock_callback is
13900 used: it isn't thread safe and the add_lock_callback should handle
13903 *Paul Rose <Paul.Rose@bridge.com>*
13905 * Verify that incoming data obeys the block size in
13906 ssl3_enc (ssl/s3_enc.c) and tls1_enc (ssl/t1_enc.c).
13912 *Ulf Möller, Bodo Möller*
13914 * The countermeasure against Bleichbacher's attack on PKCS #1 v1.5
13915 RSA encryption was accidentally removed in s3_srvr.c in OpenSSL 0.9.5
13916 when fixing the server behaviour for backwards-compatible 'client
13917 hello' messages. (Note that the attack is impractical against
13918 SSL 3.0 and TLS 1.0 anyway because length and version checking
13919 means that the probability of guessing a valid ciphertext is
13920 around 2^-40; see section 5 in Bleichenbacher's CRYPTO '98
13923 Before 0.9.5, the countermeasure (hide the error by generating a
13924 random 'decryption result') did not work properly because
13925 ERR_clear_error() was missing, meaning that SSL_get_error() would
13926 detect the supposedly ignored error.
13928 Both problems are now fixed.
13932 * In crypto/bio/bf_buff.c, increase DEFAULT_BUFFER_SIZE to 4096
13933 (previously it was 1024).
13937 * Fix for compatibility mode trust settings: ignore trust settings
13938 unless some valid trust or reject settings are present.
13942 * Fix for blowfish EVP: its a variable length cipher.
13946 * Fix various bugs related to DSA S/MIME verification. Handle missing
13947 parameters in DSA public key structures and return an error in the
13948 DSA routines if parameters are absent.
13952 * In versions up to 0.9.6, RAND_file_name() resorted to file ".rnd"
13953 in the current directory if neither $RANDFILE nor $HOME was set.
13954 RAND_file_name() in 0.9.6a returned NULL in this case. This has
13955 caused some confusion to Windows users who haven't defined $HOME.
13956 Thus RAND_file_name() is changed again: e_os.h can define a
13957 DEFAULT_HOME, which will be used if $HOME is not set.
13958 For Windows, we use "C:"; on other platforms, we still require
13959 environment variables.
13961 * Move 'if (!initialized) RAND_poll()' into regions protected by
13962 CRYPTO_LOCK_RAND. This is not strictly necessary, but avoids
13963 having multiple threads call RAND_poll() concurrently.
13967 * In crypto/rand/md_rand.c, replace 'add_do_not_lock' flag by a
13968 combination of a flag and a thread ID variable.
13969 Otherwise while one thread is in ssleay_rand_bytes (which sets the
13970 flag), *other* threads can enter ssleay_add_bytes without obeying
13971 the CRYPTO_LOCK_RAND lock (and may even illegally release the lock
13972 that they do not hold after the first thread unsets add_do_not_lock).
13976 * Change bctest again: '-x' expressions are not available in all
13977 versions of 'test'.
13981 ### Changes between 0.9.6 and 0.9.6a [5 Apr 2001]
13983 * Fix a couple of memory leaks in PKCS7_dataDecode()
13985 *Steve Henson, reported by Heyun Zheng <hzheng@atdsprint.com>*
13987 * Change Configure and Makefiles to provide EXE_EXT, which will contain
13988 the default extension for executables, if any. Also, make the perl
13989 scripts that use symlink() to test if it really exists and use "cp"
13990 if it doesn't. All this made OpenSSL compilable and installable in
13995 * Fix for asn1_GetSequence() for indefinite length constructed data.
13996 If SEQUENCE is length is indefinite just set c->slen to the total
13997 amount of data available.
13999 *Steve Henson, reported by shige@FreeBSD.org*
14001 *This change does not apply to 0.9.7.*
14003 * Change bctest to avoid here-documents inside command substitution
14004 (workaround for FreeBSD /bin/sh bug).
14005 For compatibility with Ultrix, avoid shell functions (introduced
14006 in the bctest version that searches along $PATH).
14010 * Rename 'des_encrypt' to 'des_encrypt1'. This avoids the clashes
14011 with des_encrypt() defined on some operating systems, like Solaris
14016 * Check the result of RSA-CRT (see D. Boneh, R. DeMillo, R. Lipton:
14017 On the Importance of Eliminating Errors in Cryptographic
14018 Computations, J. Cryptology 14 (2001) 2, 101-119,
14019 <http://theory.stanford.edu/~dabo/papers/faults.ps.gz>).
14023 * MIPS assembler BIGNUM division bug fix.
14027 * Disabled incorrect Alpha assembler code.
14031 * Fix PKCS#7 decode routines so they correctly update the length
14032 after reading an EOC for the EXPLICIT tag.
14036 *This change does not apply to 0.9.7.*
14038 * Fix bug in PKCS#12 key generation routines. This was triggered
14039 if a 3DES key was generated with a 0 initial byte. Include
14040 PKCS12_BROKEN_KEYGEN compilation option to retain the old
14041 (but broken) behaviour.
14045 * Enhance bctest to search for a working bc along $PATH and print
14048 *Tim Rice <tim@multitalents.net> via Richard Levitte*
14050 * Fix memory leaks in err.c: free err_data string if necessary;
14051 don't write to the wrong index in ERR_set_error_data.
14055 * Implement ssl23_peek (analogous to ssl23_read), which previously
14060 * Replace rdtsc with `_emit` statements for VC++ version 5.
14062 *Jeremy Cooper <jeremy@baymoo.org>*
14064 * Make it possible to reuse SSLv2 sessions.
14068 * In copy_email() check for >= 0 as a return value for
14069 X509_NAME_get_index_by_NID() since 0 is a valid index.
14071 *Steve Henson reported by Massimiliano Pala <madwolf@opensca.org>*
14073 * Avoid coredump with unsupported or invalid public keys by checking if
14074 X509_get_pubkey() fails in PKCS7_verify(). Fix memory leak when
14075 PKCS7_verify() fails with non detached data.
14079 * Don't use getenv in library functions when run as setuid/setgid.
14080 New function OPENSSL_issetugid().
14084 * Avoid false positives in memory leak detection code (crypto/mem_dbg.c)
14085 due to incorrect handling of multi-threading:
14087 1. Fix timing glitch in the MemCheck_off() portion of CRYPTO_mem_ctrl().
14089 2. Fix logical glitch in is_MemCheck_on() aka CRYPTO_is_mem_check_on().
14091 3. Count how many times MemCheck_off() has been called so that
14092 nested use can be treated correctly. This also avoids
14093 inband-signalling in the previous code (which relied on the
14094 assumption that thread ID 0 is impossible).
14098 * Add "-rand" option also to s_client and s_server.
14102 * Fix CPU detection on Irix 6.x.
14103 *Kurt Hockenbury <khockenb@stevens-tech.edu> and
14104 "Bruce W. Forsberg" <bruce.forsberg@baesystems.com>*
14106 * Fix X509_NAME bug which produced incorrect encoding if X509_NAME
14111 *This change does not apply to 0.9.7.*
14113 * Use the cached encoding of an X509_NAME structure rather than
14114 copying it. This is apparently the reason for the libsafe "errors"
14115 but the code is actually correct.
14119 * Add new function BN_rand_range(), and fix DSA_sign_setup() to prevent
14120 Bleichenbacher's DSA attack.
14121 Extend BN_[pseudo_]rand: As before, top=1 forces the highest two bits
14122 to be set and top=0 forces the highest bit to be set; top=-1 is new
14123 and leaves the highest bit random.
14125 *Ulf Moeller, Bodo Moeller*
14127 * In the `NCONF_...`-based implementations for `CONF_...` queries
14128 (crypto/conf/conf_lib.c), if the input LHASH is NULL, avoid using
14129 a temporary CONF structure with the data component set to NULL
14130 (which gives segmentation faults in lh_retrieve).
14131 Instead, use NULL for the CONF pointer in CONF_get_string and
14132 CONF_get_number (which may use environment variables) and directly
14133 return NULL from CONF_get_section.
14137 * Fix potential buffer overrun for EBCDIC.
14141 * Tolerate nonRepudiation as being valid for S/MIME signing and certSign
14142 keyUsage if basicConstraints absent for a CA.
14146 * Make SMIME_write_PKCS7() write mail header values with a format that
14147 is more generally accepted (no spaces before the semicolon), since
14148 some programs can't parse those values properly otherwise. Also make
14149 sure BIO's that break lines after each write do not create invalid
14154 * Make the CRL encoding routines work with empty SEQUENCE OF. The
14155 macros previously used would not encode an empty SEQUENCE OF
14156 and break the signature.
14160 *This change does not apply to 0.9.7.*
14162 * Zero the premaster secret after deriving the master secret in
14167 * Add some EVP_add_digest_alias registrations (as found in
14168 OpenSSL_add_all_digests()) to SSL_library_init()
14169 aka OpenSSL_add_ssl_algorithms(). This provides improved
14170 compatibility with peers using X.509 certificates
14171 with unconventional AlgorithmIdentifier OIDs.
14175 * Fix for Irix with NO_ASM.
14177 *"Bruce W. Forsberg" <bruce.forsberg@baesystems.com>*
14179 * ./config script fixes.
14181 *Ulf Moeller, Richard Levitte*
14183 * Fix 'openssl passwd -1'.
14187 * Change PKCS12_key_gen_asc() so it can cope with non null
14188 terminated strings whose length is passed in the passlen
14189 parameter, for example from PEM callbacks. This was done
14190 by adding an extra length parameter to asc2uni().
14192 *Steve Henson, reported by <oddissey@samsung.co.kr>*
14194 * Fix C code generated by 'openssl dsaparam -C': If a BN_bin2bn
14195 call failed, free the DSA structure.
14199 * Fix to uni2asc() to cope with zero length Unicode strings.
14200 These are present in some PKCS#12 files.
14204 * Increase s2->wbuf allocation by one byte in ssl2_new (ssl/s2_lib.c).
14205 Otherwise do_ssl_write (ssl/s2_pkt.c) will write beyond buffer limits
14206 when writing a 32767 byte record.
14208 *Bodo Moeller; problem reported by Eric Day <eday@concentric.net>*
14210 * In `RSA_eay_public_{en,ed}crypt` and RSA_eay_mod_exp (rsa_eay.c),
14211 obtain lock CRYPTO_LOCK_RSA before setting `rsa->_method_mod_{n,p,q}`.
14213 (RSA objects have a reference count access to which is protected
14214 by CRYPTO_LOCK_RSA [see rsa_lib.c, s3_srvr.c, ssl_cert.c, ssl_rsa.c],
14215 so they are meant to be shared between threads.)
14216 *Bodo Moeller, Geoff Thorpe; original patch submitted by
14217 "Reddie, Steven" <Steven.Reddie@ca.com>*
14219 * Fix a deadlock in CRYPTO_mem_leaks().
14223 * Use better test patterns in bntest.
14227 * rand_win.c fix for Borland C.
14231 * BN_rshift bugfix for n == 0.
14235 * Add a 'bctest' script that checks for some known 'bc' bugs
14236 so that 'make test' does not abort just because 'bc' is broken.
14240 * Store verify_result within SSL_SESSION also for client side to
14241 avoid potential security hole. (Re-used sessions on the client side
14242 always resulted in verify_result==X509_V_OK, not using the original
14243 result of the server certificate verification.)
14247 * Fix ssl3_pending: If the record in s->s3->rrec is not of type
14248 SSL3_RT_APPLICATION_DATA, return 0.
14249 Similarly, change ssl2_pending to return 0 if SSL_in_init(s) is true.
14254 Both ssl2_peek and ssl3_peek, which were totally broken in earlier
14255 releases, have been re-implemented by renaming the previous
14256 implementations of ssl2_read and ssl3_read to ssl2_read_internal
14257 and ssl3_read_internal, respectively, and adding 'peek' parameters
14258 to them. The new ssl[23]_{read,peek} functions are calls to
14259 ssl[23]_read_internal with the 'peek' flag set appropriately.
14260 A 'peek' parameter has also been added to ssl3_read_bytes, which
14261 does the actual work for ssl3_read_internal.
14265 * Initialise "ex_data" member of RSA/DSA/DH structures prior to calling
14266 the method-specific "init()" handler. Also clean up ex_data after
14267 calling the method-specific "finish()" handler. Previously, this was
14268 happening the other way round.
14272 * Increase BN_CTX_NUM (the number of BIGNUMs in a BN_CTX) to 16.
14273 The previous value, 12, was not always sufficient for BN_mod_exp().
14277 * Make sure that shared libraries get the internal name engine with
14278 the full version number and not just 0. This should mark the
14279 shared libraries as not backward compatible. Of course, this should
14280 be changed again when we can guarantee backward binary compatibility.
14284 * Fix typo in get_cert_by_subject() in by_dir.c
14286 *Jean-Marc Desperrier <jean-marc.desperrier@certplus.com>*
14288 * Rework the system to generate shared libraries:
14290 - Make note of the expected extension for the shared libraries and
14291 if there is a need for symbolic links from for example libcrypto.so.0
14292 to libcrypto.so.0.9.7. There is extended info in Configure for
14295 - Make as few rebuilds of the shared libraries as possible.
14297 - Still avoid linking the OpenSSL programs with the shared libraries.
14299 - When installing, install the shared libraries separately from the
14304 * Fix SSL_CTX_set_read_ahead macro to actually use its argument.
14306 Copy SSL_CTX's read_ahead flag to SSL object directly in SSL_new
14307 and not in SSL_clear because the latter is also used by the
14308 accept/connect functions; previously, the settings made by
14309 SSL_set_read_ahead would be lost during the handshake.
14311 *Bodo Moeller; problems reported by Anders Gertz <gertz@epact.se>*
14313 * Correct util/mkdef.pl to be selective about disabled algorithms.
14314 Previously, it would create entries for disabled algorithms no
14319 * Added several new manual pages for SSL_* function.
14323 ### Changes between 0.9.5a and 0.9.6 [24 Sep 2000]
14325 * In ssl23_get_client_hello, generate an error message when faced
14326 with an initial SSL 3.0/TLS record that is too small to contain the
14327 first two bytes of the ClientHello message, i.e. client_version.
14328 (Note that this is a pathologic case that probably has never happened
14329 in real life.) The previous approach was to use the version number
14330 from the record header as a substitute; but our protocol choice
14331 should not depend on that one because it is not authenticated
14332 by the Finished messages.
14336 * More robust randomness gathering functions for Windows.
14338 *Jeffrey Altman <jaltman@columbia.edu>*
14340 * For compatibility reasons if the flag X509_V_FLAG_ISSUER_CHECK is
14341 not set then we don't setup the error code for issuer check errors
14342 to avoid possibly overwriting other errors which the callback does
14343 handle. If an application does set the flag then we assume it knows
14344 what it is doing and can handle the new informational codes
14349 * Fix for a nasty bug in ASN1_TYPE handling. ASN1_TYPE is used for
14350 a general "ANY" type, as such it should be able to decode anything
14351 including tagged types. However it didn't check the class so it would
14352 wrongly interpret tagged types in the same way as their universal
14353 counterpart and unknown types were just rejected. Changed so that the
14354 tagged and unknown types are handled in the same way as a SEQUENCE:
14355 that is the encoding is stored intact. There is also a new type
14356 "V_ASN1_OTHER" which is used when the class is not universal, in this
14357 case we have no idea what the actual type is so we just lump them all
14362 * On VMS, stdout may very well lead to a file that is written to
14363 in a record-oriented fashion. That means that every write() will
14364 write a separate record, which will be read separately by the
14365 programs trying to read from it. This can be very confusing.
14367 The solution is to put a BIO filter in the way that will buffer
14368 text until a linefeed is reached, and then write everything a
14369 line at a time, so every record written will be an actual line,
14370 not chunks of lines and not (usually doesn't happen, but I've
14371 seen it once) several lines in one record. BIO_f_linebuffer() is
14374 Currently, it's a VMS-only method, because that's where it has
14375 been tested well enough.
14379 * Remove 'optimized' squaring variant in BN_mod_mul_montgomery,
14380 it can return incorrect results.
14381 (Note: The buggy variant was not enabled in OpenSSL 0.9.5a,
14382 but it was in 0.9.6-beta[12].)
14386 * Disable the check for content being present when verifying detached
14387 signatures in pk7_smime.c. Some versions of Netscape (wrongly)
14388 include zero length content when signing messages.
14392 * New BIO_shutdown_wr macro, which invokes the BIO_C_SHUTDOWN_WR
14393 BIO_ctrl (for BIO pairs).
14397 * Add DSO method for VMS.
14401 * Bug fix: Montgomery multiplication could produce results with the
14406 * Add RPM specification openssl.spec and modify it to build three
14407 packages. The default package contains applications, application
14408 documentation and run-time libraries. The devel package contains
14409 include files, static libraries and function documentation. The
14410 doc package contains the contents of the doc directory. The original
14411 openssl.spec was provided by Damien Miller <djm@mindrot.org>.
14415 * Add a large number of documentation files for many SSL routines.
14417 *Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>*
14419 * Add a configuration entry for Sony News 4.
14421 *NAKAJI Hiroyuki <nakaji@tutrp.tut.ac.jp>*
14423 * Don't set the two most significant bits to one when generating a
14424 random number < q in the DSA library.
14428 * New SSL API mode 'SSL_MODE_AUTO_RETRY'. This disables the default
14429 behaviour that SSL_read may result in SSL_ERROR_WANT_READ (even if
14430 the underlying transport is blocking) if a handshake took place.
14431 (The default behaviour is needed by applications such as s_client
14432 and s_server that use select() to determine when to use SSL_read;
14433 but for applications that know in advance when to expect data, it
14434 just makes things more complicated.)
14438 * Add RAND_egd_bytes(), which gives control over the number of bytes read
14443 * Add a few more EBCDIC conditionals that make `req` and `x509`
14444 work better on such systems.
14446 *Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>*
14448 * Add two demo programs for PKCS12_parse() and PKCS12_create().
14449 Update PKCS12_parse() so it copies the friendlyName and the
14450 keyid to the certificates aux info.
14454 * Fix bug in PKCS7_verify() which caused an infinite loop
14455 if there was more than one signature.
14457 *Sven Uszpelkat <su@celocom.de>*
14459 * Major change in util/mkdef.pl to include extra information
14460 about each symbol, as well as presenting variables as well
14461 as functions. This change means that there's n more need
14462 to rebuild the .num files when some algorithms are excluded.
14466 * Allow the verify time to be set by an application,
14467 rather than always using the current time.
14471 * Phase 2 verify code reorganisation. The certificate
14472 verify code now looks up an issuer certificate by a
14473 number of criteria: subject name, authority key id
14474 and key usage. It also verifies self signed certificates
14475 by the same criteria. The main comparison function is
14476 X509_check_issued() which performs these checks.
14478 Lot of changes were necessary in order to support this
14479 without completely rewriting the lookup code.
14481 Authority and subject key identifier are now cached.
14483 The LHASH 'certs' is X509_STORE has now been replaced
14484 by a STACK_OF(X509_OBJECT). This is mainly because an
14485 LHASH can't store or retrieve multiple objects with
14486 the same hash value.
14488 As a result various functions (which were all internal
14489 use only) have changed to handle the new X509_STORE
14490 structure. This will break anything that messed round
14491 with X509_STORE internally.
14493 The functions X509_STORE_add_cert() now checks for an
14494 exact match, rather than just subject name.
14496 The X509_STORE API doesn't directly support the retrieval
14497 of multiple certificates matching a given criteria, however
14498 this can be worked round by performing a lookup first
14499 (which will fill the cache with candidate certificates)
14500 and then examining the cache for matches. This is probably
14501 the best we can do without throwing out X509_LOOKUP
14502 entirely (maybe later...).
14504 The X509_VERIFY_CTX structure has been enhanced considerably.
14506 All certificate lookup operations now go via a get_issuer()
14507 callback. Although this currently uses an X509_STORE it
14508 can be replaced by custom lookups. This is a simple way
14509 to bypass the X509_STORE hackery necessary to make this
14510 work and makes it possible to use more efficient techniques
14511 in future. A very simple version which uses a simple
14512 STACK for its trusted certificate store is also provided
14513 using X509_STORE_CTX_trusted_stack().
14515 The verify_cb() and verify() callbacks now have equivalents
14516 in the X509_STORE_CTX structure.
14518 X509_STORE_CTX also has a 'flags' field which can be used
14519 to customise the verify behaviour.
14523 * Add new PKCS#7 signing option PKCS7_NOSMIMECAP which
14524 excludes S/MIME capabilities.
14528 * When a certificate request is read in keep a copy of the
14529 original encoding of the signed data and use it when outputting
14530 again. Signatures then use the original encoding rather than
14531 a decoded, encoded version which may cause problems if the
14532 request is improperly encoded.
14536 * For consistency with other BIO_puts implementations, call
14537 buffer_write(b, ...) directly in buffer_puts instead of calling
14540 In BIO_puts, increment b->num_write as in BIO_write.
14542 *Peter.Sylvester@EdelWeb.fr*
14544 * Fix BN_mul_word for the case where the word is 0. (We have to use
14545 BN_zero, we may not return a BIGNUM with an array consisting of
14546 words set to zero.)
14550 * Avoid calling abort() from within the library when problems are
14551 detected, except if preprocessor symbols have been defined
14552 (such as REF_CHECK, BN_DEBUG etc.).
14556 * New openssl application 'rsautl'. This utility can be
14557 used for low level RSA operations. DER public key
14558 BIO/fp routines also added.
14562 * New Configure entry and patches for compiling on QNX 4.
14564 *Andreas Schneider <andreas@ds3.etech.fh-hamburg.de>*
14566 * A demo state-machine implementation was sponsored by
14567 Nuron (<http://www.nuron.com/>) and is now available in
14568 demos/state_machine.
14572 * New options added to the 'dgst' utility for signature
14573 generation and verification.
14577 * Unrecognized PKCS#7 content types are now handled via a
14578 catch all ASN1_TYPE structure. This allows unsupported
14579 types to be stored as a "blob" and an application can
14580 encode and decode it manually.
14584 * Fix various signed/unsigned issues to make a_strex.c
14585 compile under VC++.
14587 *Oscar Jacobsson <oscar.jacobsson@celocom.com>*
14589 * ASN1 fixes. i2d_ASN1_OBJECT was not returning the correct
14590 length if passed a buffer. ASN1_INTEGER_to_BN failed
14591 if passed a NULL BN and its argument was negative.
14593 *Steve Henson, pointed out by Sven Heiberg <sven@tartu.cyber.ee>*
14595 * Modification to PKCS#7 encoding routines to output definite
14596 length encoding. Since currently the whole structures are in
14597 memory there's not real point in using indefinite length
14598 constructed encoding. However if OpenSSL is compiled with
14599 the flag PKCS7_INDEFINITE_ENCODING the old form is used.
14603 * Added BIO_vprintf() and BIO_vsnprintf().
14607 * Added more prefixes to parse for in the strings written
14608 through a logging bio, to cover all the levels that are available
14609 through syslog. The prefixes are now:
14611 PANIC, EMERG, EMR => LOG_EMERG
14612 ALERT, ALR => LOG_ALERT
14613 CRIT, CRI => LOG_CRIT
14614 ERROR, ERR => LOG_ERR
14615 WARNING, WARN, WAR => LOG_WARNING
14616 NOTICE, NOTE, NOT => LOG_NOTICE
14617 INFO, INF => LOG_INFO
14618 DEBUG, DBG => LOG_DEBUG
14620 and as before, if none of those prefixes are present at the
14621 beginning of the string, LOG_ERR is chosen.
14623 On Win32, the `LOG_*` levels are mapped according to this:
14625 LOG_EMERG, LOG_ALERT, LOG_CRIT, LOG_ERR => EVENTLOG_ERROR_TYPE
14626 LOG_WARNING => EVENTLOG_WARNING_TYPE
14627 LOG_NOTICE, LOG_INFO, LOG_DEBUG => EVENTLOG_INFORMATION_TYPE
14631 * Made it possible to reconfigure with just the configuration
14632 argument "reconf" or "reconfigure". The command line arguments
14633 are stored in Makefile.ssl in the variable CONFIGURE_ARGS,
14634 and are retrieved from there when reconfiguring.
14640 *Assar Westerlund <assar@sics.se>, Richard Levitte*
14642 * Add the arguments -CAfile and -CApath to the pkcs12 utility.
14646 * The obj_dat.pl script was messing up the sorting of object
14647 names. The reason was that it compared the quoted version
14648 of strings as a result "OCSP" > "OCSP Signing" because
14649 " > SPACE. Changed script to store unquoted versions of
14650 names and add quotes on output. It was also omitting some
14651 names from the lookup table if they were given a default
14652 value (that is if SN is missing it is given the same
14653 value as LN and vice versa), these are now added on the
14654 grounds that if an object has a name we should be able to
14655 look it up. Finally added warning output when duplicate
14656 short or long names are found.
14660 * Changes needed for Tandem NSK.
14662 *Scott Uroff <scott@xypro.com>*
14664 * Fix SSL 2.0 rollback checking: Due to an off-by-one error in
14665 RSA_padding_check_SSLv23(), special padding was never detected
14666 and thus the SSL 3.0/TLS 1.0 countermeasure against protocol
14667 version rollback attacks was not effective.
14669 In s23_clnt.c, don't use special rollback-attack detection padding
14670 (RSA_SSLV23_PADDING) if SSL 2.0 is the only protocol enabled in the
14671 client; similarly, in s23_srvr.c, don't do the rollback check if
14672 SSL 2.0 is the only protocol enabled in the server.
14676 * Make it possible to get hexdumps of unprintable data with 'openssl
14677 asn1parse'. By implication, the functions ASN1_parse_dump() and
14678 BIO_dump_indent() are added.
14682 * New functions ASN1_STRING_print_ex() and X509_NAME_print_ex()
14683 these print out strings and name structures based on various
14684 flags including RFC2253 support and proper handling of
14685 multibyte characters. Added options to the 'x509' utility
14686 to allow the various flags to be set.
14690 * Various fixes to use ASN1_TIME instead of ASN1_UTCTIME.
14691 Also change the functions X509_cmp_current_time() and
14692 X509_gmtime_adj() work with an ASN1_TIME structure,
14693 this will enable certificates using GeneralizedTime in validity
14694 dates to be checked.
14698 * Make the NEG_PUBKEY_BUG code (which tolerates invalid
14699 negative public key encodings) on by default,
14700 NO_NEG_PUBKEY_BUG can be set to disable it.
14704 * New function c2i_ASN1_OBJECT() which acts on ASN1_OBJECT
14705 content octets. An i2c_ASN1_OBJECT is unnecessary because
14706 the encoding can be trivially obtained from the structure.
14710 * crypto/err.c locking bugfix: Use write locks (`CRYPTO_w_[un]lock`),
14711 not read locks (`CRYPTO_r_[un]lock`).
14715 * A first attempt at creating official support for shared
14716 libraries through configuration. I've kept it so the
14717 default is static libraries only, and the OpenSSL programs
14718 are always statically linked for now, but there are
14719 preparations for dynamic linking in place.
14720 This has been tested on Linux and Tru64.
14724 * Randomness polling function for Win9x, as described in:
14725 Peter Gutmann, Software Generation of Practically Strong
14730 * Fix so PRNG is seeded in req if using an already existing
14735 * New options to smime application. -inform and -outform
14736 allow alternative formats for the S/MIME message including
14737 PEM and DER. The -content option allows the content to be
14738 specified separately. This should allow things like Netscape
14739 form signing output easier to verify.
14743 * Fix the ASN1 encoding of tags using the 'long form'.
14747 * New ASN1 functions, `i2c_*` and `c2i_*` for INTEGER and BIT
14748 STRING types. These convert content octets to and from the
14749 underlying type. The actual tag and length octets are
14750 already assumed to have been read in and checked. These
14751 are needed because all other string types have virtually
14752 identical handling apart from the tag. By having versions
14753 of the ASN1 functions that just operate on content octets
14754 IMPLICIT tagging can be handled properly. It also allows
14755 the ASN1_ENUMERATED code to be cut down because ASN1_ENUMERATED
14756 and ASN1_INTEGER are identical apart from the tag.
14760 * Change the handling of OID objects as follows:
14762 - New object identifiers are inserted in objects.txt, following
14763 the syntax given in [crypto/objects/README.md](crypto/objects/README.md).
14764 - objects.pl is used to process obj_mac.num and create a new
14766 - obj_dat.pl is used to create a new obj_dat.h, using the data in
14769 This is currently kind of a hack, and the perl code in objects.pl
14770 isn't very elegant, but it works as I intended. The simplest way
14771 to check that it worked correctly is to look in obj_dat.h and
14772 check the array nid_objs and make sure the objects haven't moved
14773 around (this is important!). Additions are OK, as well as
14774 consistent name changes.
14778 * Add BSD-style MD5-based passwords to 'openssl passwd' (option '-1').
14782 * Addition of the command line parameter '-rand file' to 'openssl req'.
14783 The given file adds to whatever has already been seeded into the
14784 random pool through the RANDFILE configuration file option or
14785 environment variable, or the default random state file.
14789 * mkstack.pl now sorts each macro group into lexical order.
14790 Previously the output order depended on the order the files
14791 appeared in the directory, resulting in needless rewriting
14796 * Patches to make OpenSSL compile under Win32 again. Mostly
14797 work arounds for the VC++ problem that it treats func() as
14798 func(void). Also stripped out the parts of mkdef.pl that
14799 added extra typesafe functions: these no longer exist.
14803 * Reorganisation of the stack code. The macros are now all
14804 collected in safestack.h . Each macro is defined in terms of
14805 a "stack macro" of the form `SKM_<name>(type, a, b)`. The
14806 DEBUG_SAFESTACK is now handled in terms of function casts,
14807 this has the advantage of retaining type safety without the
14808 use of additional functions. If DEBUG_SAFESTACK is not defined
14809 then the non typesafe macros are used instead. Also modified the
14810 mkstack.pl script to handle the new form. Needs testing to see
14811 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK
14812 the default if no major problems. Similar behaviour for ASN1_SET_OF
14813 and PKCS12_STACK_OF.
14817 * When some versions of IIS use the 'NET' form of private key the
14818 key derivation algorithm is different. Normally MD5(password) is
14819 used as a 128 bit RC4 key. In the modified case
14820 MD5(MD5(password) + "SGCKEYSALT") is used instead. Added some
14821 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same
14822 as the old Netscape_RSA functions except they have an additional
14823 'sgckey' parameter which uses the modified algorithm. Also added
14824 an -sgckey command line option to the rsa utility. Thanks to
14825 Adrian Peck <bertie@ncipher.com> for posting details of the modified
14826 algorithm to openssl-dev.
14830 * The evp_local.h macros were using 'c.##kname' which resulted in
14831 invalid expansion on some systems (SCO 5.0.5 for example).
14832 Corrected to 'c.kname'.
14834 *Phillip Porch <root@theporch.com>*
14836 * New X509_get1_email() and X509_REQ_get1_email() functions that return
14837 a STACK of email addresses from a certificate or request, these look
14838 in the subject name and the subject alternative name extensions and
14839 omit any duplicate addresses.
14843 * Re-implement BN_mod_exp2_mont using independent (and larger) windows.
14844 This makes DSA verification about 2 % faster.
14848 * Increase maximum window size in `BN_mod_exp_...` to 6 bits instead of 5
14849 (meaning that now 2^5 values will be precomputed, which is only 4 KB
14850 plus overhead for 1024 bit moduli).
14851 This makes exponentiations about 0.5 % faster for 1024 bit
14852 exponents (as measured by "openssl speed rsa2048").
14856 * Rename memory handling macros to avoid conflicts with other
14858 Malloc => OPENSSL_malloc
14859 Malloc_locked => OPENSSL_malloc_locked
14860 Realloc => OPENSSL_realloc
14861 Free => OPENSSL_free
14865 * New function BN_mod_exp_mont_word for small bases (roughly 15%
14866 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange).
14870 * CygWin32 support.
14872 *John Jarvie <jjarvie@newsguy.com>*
14874 * The type-safe stack code has been rejigged. It is now only compiled
14875 in when OpenSSL is configured with the DEBUG_SAFESTACK option and
14876 by default all type-specific stack functions are "#define"d back to
14877 standard stack functions. This results in more streamlined output
14878 but retains the type-safety checking possibilities of the original
14883 * The STACK code has been cleaned up, and certain type declarations
14884 that didn't make a lot of sense have been brought in line. This has
14885 also involved a cleanup of sorts in safestack.h to more correctly
14886 map type-safe stack functions onto their plain stack counterparts.
14887 This work has also resulted in a variety of "const"ifications of
14888 lots of the code, especially `_cmp` operations which should normally
14889 be prototyped with "const" parameters anyway.
14893 * When generating bytes for the first time in md_rand.c, 'stir the pool'
14894 by seeding with STATE_SIZE dummy bytes (with zero entropy count).
14895 (The PRNG state consists of two parts, the large pool 'state' and 'md',
14896 where all of 'md' is used each time the PRNG is used, but 'state'
14897 is used only indexed by a cyclic counter. As entropy may not be
14898 well distributed from the beginning, 'md' is important as a
14899 chaining variable. However, the output function chains only half
14900 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains
14901 all of 'md', and seeding with STATE_SIZE dummy bytes will result
14902 in all of 'state' being rewritten, with the new values depending
14903 on virtually all of 'md'. This overcomes the 80 bit limitation.)
14907 * In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when
14908 the handshake is continued after ssl_verify_cert_chain();
14909 otherwise, if SSL_VERIFY_NONE is set, remaining error codes
14910 can lead to 'unexplainable' connection aborts later.
14912 *Bodo Moeller; problem tracked down by Lutz Jaenicke*
14914 * Major EVP API cipher revision.
14915 Add hooks for extra EVP features. This allows various cipher
14916 parameters to be set in the EVP interface. Support added for variable
14917 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and
14918 setting of RC2 and RC5 parameters.
14920 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length
14923 Remove lots of duplicated code from the EVP library. For example *every*
14924 cipher init() function handles the 'iv' in the same way according to the
14925 cipher mode. They also all do nothing if the 'key' parameter is NULL and
14926 for CFB and OFB modes they zero ctx->num.
14928 New functionality allows removal of S/MIME code RC2 hack.
14930 Most of the routines have the same form and so can be declared in terms
14933 By shifting this to the top level EVP_CipherInit() it can be removed from
14934 all individual ciphers. If the cipher wants to handle IVs or keys
14935 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT
14938 Change lots of functions like EVP_EncryptUpdate() to now return a
14939 value: although software versions of the algorithms cannot fail
14940 any installed hardware versions can.
14944 * Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if
14945 this option is set, tolerate broken clients that send the negotiated
14946 protocol version number instead of the requested protocol version
14951 * Call dh_tmp_cb (set by `..._TMP_DH_CB`) with correct 'is_export' flag;
14952 i.e. non-zero for export ciphersuites, zero otherwise.
14953 Previous versions had this flag inverted, inconsistent with
14954 rsa_tmp_cb (..._TMP_RSA_CB).
14956 *Bodo Moeller; problem reported by Amit Chopra*
14958 * Add missing DSA library text string. Work around for some IIS
14959 key files with invalid SEQUENCE encoding.
14963 * Add a document (doc/standards.txt) that list all kinds of standards
14964 and so on that are implemented in OpenSSL.
14968 * Enhance c_rehash script. Old version would mishandle certificates
14969 with the same subject name hash and wouldn't handle CRLs at all.
14970 Added -fingerprint option to crl utility, to support new c_rehash
14975 * Eliminate non-ANSI declarations in crypto.h and stack.h.
14979 * Fix for SSL server purpose checking. Server checking was
14980 rejecting certificates which had extended key usage present
14981 but no ssl client purpose.
14983 *Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>*
14985 * Make PKCS#12 code work with no password. The PKCS#12 spec
14986 is a little unclear about how a blank password is handled.
14987 Since the password in encoded as a BMPString with terminating
14988 double NULL a zero length password would end up as just the
14989 double NULL. However no password at all is different and is
14990 handled differently in the PKCS#12 key generation code. NS
14991 treats a blank password as zero length. MSIE treats it as no
14992 password on export: but it will try both on import. We now do
14993 the same: PKCS12_parse() tries zero length and no password if
14994 the password is set to "" or NULL (NULL is now a valid password:
14995 it wasn't before) as does the pkcs12 application.
14999 * Bugfixes in apps/x509.c: Avoid a memory leak; and don't use
15000 perror when PEM_read_bio_X509_REQ fails, the error message must
15001 be obtained from the error queue.
15005 * Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing
15006 it in ERR_remove_state if appropriate, and change ERR_get_state
15007 accordingly to avoid race conditions (this is necessary because
15008 thread_hash is no longer constant once set).
15012 * Bugfix for linux-elf makefile.one.
15016 * RSA_get_default_method() will now cause a default
15017 RSA_METHOD to be chosen if one doesn't exist already.
15018 Previously this was only set during a call to RSA_new()
15019 or RSA_new_method(NULL) meaning it was possible for
15020 RSA_get_default_method() to return NULL.
15024 * Added native name translation to the existing DSO code
15025 that will convert (if the flag to do so is set) filenames
15026 that are sufficiently small and have no path information
15027 into a canonical native form. Eg. "blah" converted to
15028 "libblah.so" or "blah.dll" etc.
15032 * New function ERR_error_string_n(e, buf, len) which is like
15033 ERR_error_string(e, buf), but writes at most 'len' bytes
15034 including the 0 terminator. For ERR_error_string_n, 'buf'
15037 *Damien Miller <djm@mindrot.org>, Bodo Moeller*
15039 * CONF library reworked to become more general. A new CONF
15040 configuration file reader "class" is implemented as well as a
15041 new functions (`NCONF_*`, for "New CONF") to handle it. The now
15042 old `CONF_*` functions are still there, but are reimplemented to
15043 work in terms of the new functions. Also, a set of functions
15044 to handle the internal storage of the configuration data is
15045 provided to make it easier to write new configuration file
15046 reader "classes" (I can definitely see something reading a
15047 configuration file in XML format, for example), called `_CONF_*`,
15048 or "the configuration storage API"...
15050 The new configuration file reading functions are:
15052 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio,
15053 NCONF_get_section, NCONF_get_string, NCONF_get_numbre
15055 NCONF_default, NCONF_WIN32
15057 NCONF_dump_fp, NCONF_dump_bio
15059 NCONF_default and NCONF_WIN32 are method (or "class") choosers,
15060 NCONF_new creates a new CONF object. This works in the same way
15061 as other interfaces in OpenSSL, like the BIO interface.
15062 `NCONF_dump_*` dump the internal storage of the configuration file,
15063 which is useful for debugging. All other functions take the same
15064 arguments as the old `CONF_*` functions with the exception of the
15065 first that must be a `CONF *` instead of a `LHASH *`.
15067 To make it easier to use the new classes with the old `CONF_*` functions,
15068 the function CONF_set_default_method is provided.
15072 * Add '-tls1' option to 'openssl ciphers', which was already
15073 mentioned in the documentation but had not been implemented.
15074 (This option is not yet really useful because even the additional
15075 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.)
15079 * Initial DSO code added into libcrypto for letting OpenSSL (and
15080 OpenSSL-based applications) load shared libraries and bind to
15081 them in a portable way.
15083 *Geoff Thorpe, with contributions from Richard Levitte*
15085 ### Changes between 0.9.5 and 0.9.5a [1 Apr 2000]
15087 * Make sure _lrotl and _lrotr are only used with MSVC.
15089 * Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status
15090 (the default implementation of RAND_status).
15092 * Rename openssl x509 option '-crlext', which was added in 0.9.5,
15093 to '-clrext' (= clear extensions), as intended and documented.
15094 *Bodo Moeller; inconsistency pointed out by Michael Attili
15095 <attili@amaxo.com>*
15097 * Fix for HMAC. It wasn't zeroing the rest of the block if the key length
15098 was larger than the MD block size.
15100 *Steve Henson, pointed out by Yost William <YostW@tce.com>*
15102 * Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument
15103 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set()
15104 using the passed key: if the passed key was a private key the result
15105 of X509_print(), for example, would be to print out all the private key
15110 * des_quad_cksum() byte order bug fix.
15111 *Ulf Möller, using the problem description in krb4-0.9.7, where
15112 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>*
15114 * Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly
15117 *Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>*
15119 * For easily testing in shell scripts whether some command
15120 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX'
15121 returns with exit code 0 iff no command of the given name is available.
15122 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases,
15123 the output goes to stdout and nothing is printed to stderr.
15124 Additional arguments are always ignored.
15126 Since for each cipher there is a command of the same name,
15127 the 'no-cipher' compilation switches can be tested this way.
15129 ('openssl no-XXX' is not able to detect pseudo-commands such
15130 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.)
15134 * Update test suite so that 'make test' succeeds in 'no-rsa' configuration.
15138 * For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE
15139 is set; it will be thrown away anyway because each handshake creates
15141 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition
15142 to parameters -- in previous versions (since OpenSSL 0.9.3) the
15143 'default key' from SSL_CTX_set_tmp_dh would always be lost, meaning
15144 you effectively got SSL_OP_SINGLE_DH_USE when using this macro.
15148 * New s_client option -ign_eof: EOF at stdin is ignored, and
15149 'Q' and 'R' lose their special meanings (quit/renegotiate).
15150 This is part of what -quiet does; unlike -quiet, -ign_eof
15151 does not suppress any output.
15155 * Add compatibility options to the purpose and trust code. The
15156 purpose X509_PURPOSE_ANY is "any purpose" which automatically
15157 accepts a certificate or CA, this was the previous behaviour,
15158 with all the associated security issues.
15160 X509_TRUST_COMPAT is the old trust behaviour: only and
15161 automatically trust self signed roots in certificate store. A
15162 new trust setting X509_TRUST_DEFAULT is used to specify that
15163 a purpose has no associated trust setting and it should instead
15164 use the value in the default purpose.
15168 * Fix the PKCS#8 DSA private key code so it decodes keys again
15169 and fix a memory leak.
15173 * In util/mkerr.pl (which implements 'make errors'), preserve
15174 reason strings from the previous version of the .c file, as
15175 the default to have only downcase letters (and digits) in
15176 automatically generated reasons codes is not always appropriate.
15180 * In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table
15181 using strerror. Previously, ERR_reason_error_string() returned
15182 library names as reason strings for SYSerr; but SYSerr is a special
15183 case where small numbers are errno values, not library numbers.
15187 * Add '-dsaparam' option to 'openssl dhparam' application. This
15188 converts DSA parameters into DH parameters. (When creating parameters,
15189 DSA_generate_parameters is used.)
15193 * Include 'length' (recommended exponent length) in C code generated
15194 by 'openssl dhparam -C'.
15198 * The second argument to set_label in perlasm was already being used
15199 so couldn't be used as a "file scope" flag. Moved to third argument
15204 * In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes
15205 instead of RAND_bytes for encryption IVs and salts.
15209 * Include RAND_status() into RAND_METHOD instead of implementing
15210 it only for md_rand.c Otherwise replacing the PRNG by calling
15211 RAND_set_rand_method would be impossible.
15215 * Don't let DSA_generate_key() enter an infinite loop if the random
15216 number generation fails.
15220 * New 'rand' application for creating pseudo-random output.
15224 * Added configuration support for Linux/IA64
15226 *Rolf Haberrecker <rolf@suse.de>*
15228 * Assembler module support for Mingw32.
15232 * Shared library support for HPUX (in shlib/).
15234 *Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous*
15236 * Shared library support for Solaris gcc.
15238 *Lutz Behnke <behnke@trustcenter.de>*
15240 ### Changes between 0.9.4 and 0.9.5 [28 Feb 2000]
15242 * PKCS7_encrypt() was adding text MIME headers twice because they
15243 were added manually and by SMIME_crlf_copy().
15247 * In bntest.c don't call BN_rand with zero bits argument.
15249 *Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>*
15251 * BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n]
15252 case was implemented. This caused BN_div_recp() to fail occasionally.
15256 * Add an optional second argument to the set_label() in the perl
15257 assembly language builder. If this argument exists and is set
15258 to 1 it signals that the assembler should use a symbol whose
15259 scope is the entire file, not just the current function. This
15260 is needed with MASM which uses the format label:: for this scope.
15262 *Steve Henson, pointed out by Peter Runestig <peter@runestig.com>*
15264 * Change the ASN1 types so they are typedefs by default. Before
15265 almost all types were #define'd to ASN1_STRING which was causing
15266 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING)
15271 * Change names of new functions to the new get1/get0 naming
15272 convention: After 'get1', the caller owns a reference count
15273 and has to call `..._free`; 'get0' returns a pointer to some
15274 data structure without incrementing reference counters.
15275 (Some of the existing 'get' functions increment a reference
15276 counter, some don't.)
15277 Similarly, 'set1' and 'add1' functions increase reference
15278 counters or duplicate objects.
15282 * Allow for the possibility of temp RSA key generation failure:
15283 the code used to assume it always worked and crashed on failure.
15287 * Fix potential buffer overrun problem in BIO_printf().
15288 *Ulf Möller, using public domain code by Patrick Powell; problem
15289 pointed out by David Sacerdote <das33@cornell.edu>*
15291 * Support EGD <http://www.lothar.com/tech/crypto/>. New functions
15292 RAND_egd() and RAND_status(). In the command line application,
15293 the EGD socket can be specified like a seed file using RANDFILE
15298 * Allow the string CERTIFICATE to be tolerated in PKCS#7 structures.
15299 Some CAs (e.g. Verisign) distribute certificates in this form.
15303 * Remove the SSL_ALLOW_ADH compile option and set the default cipher
15304 list to exclude them. This means that no special compilation option
15305 is needed to use anonymous DH: it just needs to be included in the
15310 * Change the EVP_MD_CTX_type macro so its meaning consistent with
15311 EVP_MD_type. The old functionality is available in a new macro called
15312 EVP_MD_md(). Change code that uses it and update docs.
15316 * `..._ctrl` functions now have corresponding `..._callback_ctrl` functions
15317 where the `void *` argument is replaced by a function pointer argument.
15318 Previously `void *` was abused to point to functions, which works on
15319 many platforms, but is not correct. As these functions are usually
15320 called by macros defined in OpenSSL header files, most source code
15321 should work without changes.
15325 * `<openssl/opensslconf.h>` (which is created by Configure) now contains
15326 sections with information on -D... compiler switches used for
15327 compiling the library so that applications can see them. To enable
15328 one of these sections, a pre-processor symbol `OPENSSL_..._DEFINES`
15329 must be defined. E.g.,
15330 #define OPENSSL_ALGORITHM_DEFINES
15331 #include <openssl/opensslconf.h>
15332 defines all pertinent `NO_<algo>` symbols, such as NO_IDEA, NO_RSA, etc.
15334 *Richard Levitte, Ulf and Bodo Möller*
15336 * Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS
15341 * Change the 'other' type in certificate aux info to a STACK_OF
15342 X509_ALGOR. Although not an AlgorithmIdentifier as such it has
15343 the required ASN1 format: arbitrary types determined by an OID.
15347 * Add some PEM_write_X509_REQ_NEW() functions and a command line
15348 argument to 'req'. This is not because the function is newer or
15349 better than others it just uses the work 'NEW' in the certificate
15350 request header lines. Some software needs this.
15354 * Reorganise password command line arguments: now passwords can be
15355 obtained from various sources. Delete the PEM_cb function and make
15356 it the default behaviour: i.e. if the callback is NULL and the
15357 usrdata argument is not NULL interpret it as a null terminated pass
15358 phrase. If usrdata and the callback are NULL then the pass phrase
15359 is prompted for as usual.
15363 * Add support for the Compaq Atalla crypto accelerator. If it is installed,
15364 the support is automatically enabled. The resulting binaries will
15365 autodetect the card and use it if present.
15367 *Ben Laurie and Compaq Inc.*
15369 * Work around for Netscape hang bug. This sends certificate request
15370 and server done in one record. Since this is perfectly legal in the
15371 SSL/TLS protocol it isn't a "bug" option and is on by default. See
15372 the bugs/SSLv3 entry for more info.
15376 * HP-UX tune-up: new unified configs, HP C compiler bug workaround.
15380 * Add -rand argument to smime and pkcs12 applications and read/write
15385 * New 'passwd' tool for crypt(3) and apr1 password hashes.
15389 * Add command line password options to the remaining applications.
15393 * Bug fix for BN_div_recp() for numerators with an even number of
15398 * More tests in bntest.c, and changed test_bn output.
15402 * ./config recognizes MacOS X now.
15406 * Bug fix for BN_div() when the first words of num and divisor are
15407 equal (it gave wrong results if `(rem=(n1-q*d0)&BN_MASK2) < d0)`.
15411 * Add support for various broken PKCS#8 formats, and command line
15412 options to produce them.
15416 * New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to
15417 get temporary BIGNUMs from a BN_CTX.
15421 * Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont()
15426 * Change the `SSLeay_add_all_*()` functions to `OpenSSL_add_all_*()` and
15427 include a #define from the old name to the new. The original intent
15428 was that statically linked binaries could for example just call
15429 SSLeay_add_all_ciphers() to just add ciphers to the table and not
15430 link with digests. This never worked because SSLeay_add_all_digests()
15431 and SSLeay_add_all_ciphers() were in the same source file so calling
15432 one would link with the other. They are now in separate source files.
15436 * Add a new -notext option to 'ca' and a -pubkey option to 'spkac'.
15440 * Use a less unusual form of the Miller-Rabin primality test (it used
15441 a binary algorithm for exponentiation integrated into the Miller-Rabin
15442 loop, our standard modexp algorithms are faster).
15446 * Support for the EBCDIC character set completed.
15448 *Martin Kraemer <Martin.Kraemer@Mch.SNI.De>*
15450 * Source code cleanups: use const where appropriate, eliminate casts,
15451 use `void *` instead of `char *` in lhash.
15455 * Bugfix: ssl3_send_server_key_exchange was not restartable
15456 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of
15457 this the server could overwrite ephemeral keys that the client
15462 * Turn DSA_is_prime into a macro that calls BN_is_prime,
15463 using 50 iterations of the Rabin-Miller test.
15465 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50
15466 iterations of the Rabin-Miller test as required by the appendix
15467 to FIPS PUB 186[-1]) instead of DSA_is_prime.
15468 As BN_is_prime_fasttest includes trial division, DSA parameter
15469 generation becomes much faster.
15471 This implies a change for the callback functions in DSA_is_prime
15472 and DSA_generate_parameters: The callback function is called once
15473 for each positive witness in the Rabin-Miller test, not just
15474 occasionally in the inner loop; and the parameters to the
15475 callback function now provide an iteration count for the outer
15476 loop rather than for the current invocation of the inner loop.
15477 DSA_generate_parameters additionally can call the callback
15478 function with an 'iteration count' of -1, meaning that a
15479 candidate has passed the trial division test (when q is generated
15480 from an application-provided seed, trial division is skipped).
15484 * New function BN_is_prime_fasttest that optionally does trial
15485 division before starting the Rabin-Miller test and has
15486 an additional BN_CTX * argument (whereas BN_is_prime always
15487 has to allocate at least one BN_CTX).
15488 'callback(1, -1, cb_arg)' is called when a number has passed the
15489 trial division stage.
15493 * Fix for bug in CRL encoding. The validity dates weren't being handled
15498 * New -pkcs12 option to CA.pl script to write out a PKCS#12 file.
15502 * New function BN_pseudo_rand().
15506 * Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable)
15507 bignum version of BN_from_montgomery() with the working code from
15508 SSLeay 0.9.0 (the word based version is faster anyway), and clean up
15513 * Avoid a race condition in s2_clnt.c (function get_server_hello) that
15514 made it impossible to use the same SSL_SESSION data structure in
15515 SSL2 clients in multiple threads.
15519 * The return value of RAND_load_file() no longer counts bytes obtained
15520 by stat(). RAND_load_file(..., -1) is new and uses the complete file
15521 to seed the PRNG (previously an explicit byte count was required).
15523 *Ulf Möller, Bodo Möller*
15525 * Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
15526 used `char *` instead of `void *` and had casts all over the place.
15530 * Make BN_generate_prime() return NULL on error if ret!=NULL.
15534 * Retain source code compatibility for BN_prime_checks macro:
15535 BN_is_prime(..., BN_prime_checks, ...) now uses
15536 BN_prime_checks_for_size to determine the appropriate number of
15537 Rabin-Miller iterations.
15541 * Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
15542 DH_CHECK_P_NOT_SAFE_PRIME.
15543 (Check if this is true? OpenPGP calls them "strong".)
15547 * Merge the functionality of "dh" and "gendh" programs into a new program
15548 "dhparam". The old programs are retained for now but will handle DH keys
15549 (instead of parameters) in future.
15553 * Make the ciphers, s_server and s_client programs check the return values
15554 when a new cipher list is set.
15558 * Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
15559 ciphers. Before when the 56bit ciphers were enabled the sorting was
15562 The syntax for the cipher sorting has been extended to support sorting by
15563 cipher-strength (using the strength_bits hard coded in the tables).
15564 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
15566 Fix a bug in the cipher-command parser: when supplying a cipher command
15567 string with an "undefined" symbol (neither command nor alphanumeric
15568 *A-Za-z0-9*, ssl_set_cipher_list used to hang in an endless loop. Now
15569 an error is flagged.
15571 Due to the strength-sorting extension, the code of the
15572 ssl_create_cipher_list() function was completely rearranged. I hope that
15573 the readability was also increased :-)
15575 *Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>*
15577 * Minor change to 'x509' utility. The -CAcreateserial option now uses 1
15578 for the first serial number and places 2 in the serial number file. This
15579 avoids problems when the root CA is created with serial number zero and
15580 the first user certificate has the same issuer name and serial number
15585 * Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
15586 the new code. Add documentation for this stuff.
15590 * Changes to X509_ATTRIBUTE utilities. These have been renamed from
15591 `X509_*()` to `X509at_*()` on the grounds that they don't handle X509
15592 structures and behave in an analogous way to the X509v3 functions:
15593 they shouldn't be called directly but wrapper functions should be used
15596 So we also now have some wrapper functions that call the X509at functions
15597 when passed certificate requests. (TO DO: similar things can be done with
15598 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
15599 things. Some of these need some d2i or i2d and print functionality
15600 because they handle more complex structures.)
15604 * Add missing #ifndefs that caused missing symbols when building libssl
15605 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
15606 NO_RSA in `ssl/s2*.c`.
15608 *Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller*
15610 * Precautions against using the PRNG uninitialized: RAND_bytes() now
15611 has a return value which indicates the quality of the random data
15612 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
15613 error queue. New function RAND_pseudo_bytes() generates output that is
15614 guaranteed to be unique but not unpredictable. RAND_add is like
15615 RAND_seed, but takes an extra argument for an entropy estimate
15616 (RAND_seed always assumes full entropy).
15620 * Do more iterations of Rabin-Miller probable prime test (specifically,
15621 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
15622 instead of only 2 for all lengths; see BN_prime_checks_for_size definition
15623 in crypto/bn/bn_prime.c for the complete table). This guarantees a
15624 false-positive rate of at most 2^-80 for random input.
15628 * Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
15632 * New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain
15633 in the 0.9.5 release), this returns the chain
15634 from an X509_CTX structure with a dup of the stack and all
15635 the X509 reference counts upped: so the stack will exist
15636 after X509_CTX_cleanup() has been called. Modify pkcs12.c
15639 Also make SSL_SESSION_print() print out the verify return
15644 * Add manpage for the pkcs12 command. Also change the default
15645 behaviour so MAC iteration counts are used unless the new
15646 -nomaciter option is used. This improves file security and
15647 only older versions of MSIE (4.0 for example) need it.
15651 * Honor the no-xxx Configure options when creating .DEF files.
15655 * Add PKCS#10 attributes to field table: challengePassword,
15656 unstructuredName and unstructuredAddress. These are taken from
15657 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
15658 international characters are used.
15660 More changes to X509_ATTRIBUTE code: allow the setting of types
15661 based on strings. Remove the 'loc' parameter when adding
15662 attributes because these will be a SET OF encoding which is sorted
15667 * Initial changes to the 'req' utility to allow request generation
15668 automation. This will allow an application to just generate a template
15669 file containing all the field values and have req construct the
15672 Initial support for X509_ATTRIBUTE handling. Stacks of these are
15673 used all over the place including certificate requests and PKCS#7
15674 structures. They are currently handled manually where necessary with
15675 some primitive wrappers for PKCS#7. The new functions behave in a
15676 manner analogous to the X509 extension functions: they allow
15677 attributes to be looked up by NID and added.
15679 Later something similar to the X509V3 code would be desirable to
15680 automatically handle the encoding, decoding and printing of the
15681 more complex types. The string types like challengePassword can
15682 be handled by the string table functions.
15684 Also modified the multi byte string table handling. Now there is
15685 a 'global mask' which masks out certain types. The table itself
15686 can use the flag STABLE_NO_MASK to ignore the mask setting: this
15687 is useful when for example there is only one permissible type
15688 (as in countryName) and using the mask might result in no valid
15693 * Clean up 'Finished' handling, and add functions SSL_get_finished and
15694 SSL_get_peer_finished to allow applications to obtain the latest
15695 Finished messages sent to the peer or expected from the peer,
15696 respectively. (SSL_get_peer_finished is usually the Finished message
15697 actually received from the peer, otherwise the protocol will be aborted.)
15699 As the Finished message are message digests of the complete handshake
15700 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
15701 be used for external authentication procedures when the authentication
15702 provided by SSL/TLS is not desired or is not enough.
15706 * Enhanced support for Alpha Linux is added. Now ./config checks if
15707 the host supports BWX extension and if Compaq C is present on the
15708 $PATH. Just exploiting of the BWX extension results in 20-30%
15709 performance kick for some algorithms, e.g. DES and RC4 to mention
15710 a couple. Compaq C in turn generates ~20% faster code for MD5 and
15715 * Add support for MS "fast SGC". This is arguably a violation of the
15716 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
15717 weak crypto and after checking the certificate is SGC a second one
15718 with strong crypto. MS SGC stops the first handshake after receiving
15719 the server certificate message and sends a second client hello. Since
15720 a server will typically do all the time consuming operations before
15721 expecting any further messages from the client (server key exchange
15722 is the most expensive) there is little difference between the two.
15724 To get OpenSSL to support MS SGC we have to permit a second client
15725 hello message after we have sent server done. In addition we have to
15726 reset the MAC if we do get this second client hello.
15730 * Add a function 'd2i_AutoPrivateKey()' this will automatically decide
15731 if a DER encoded private key is RSA or DSA traditional format. Changed
15732 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
15733 format DER encoded private key. Newer code should use PKCS#8 format which
15734 has the key type encoded in the ASN1 structure. Added DER private key
15735 support to pkcs8 application.
15739 * SSL 3/TLS 1 servers now don't request certificates when an anonymous
15740 ciphersuites has been selected (as required by the SSL 3/TLS 1
15741 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
15742 is set, we interpret this as a request to violate the specification
15743 (the worst that can happen is a handshake failure, and 'correct'
15744 behaviour would result in a handshake failure anyway).
15748 * In SSL_CTX_add_session, take into account that there might be multiple
15749 SSL_SESSION structures with the same session ID (e.g. when two threads
15750 concurrently obtain them from an external cache).
15751 The internal cache can handle only one SSL_SESSION with a given ID,
15752 so if there's a conflict, we now throw out the old one to achieve
15757 * Add OIDs for idea and blowfish in CBC mode. This will allow both
15758 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
15759 some routines that use cipher OIDs: some ciphers do not have OIDs
15760 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
15765 * Simplify the trust setting structure and code. Now we just have
15766 two sequences of OIDs for trusted and rejected settings. These will
15767 typically have values the same as the extended key usage extension
15768 and any application specific purposes.
15770 The trust checking code now has a default behaviour: it will just
15771 check for an object with the same NID as the passed id. Functions can
15772 be provided to override either the default behaviour or the behaviour
15773 for a given id. SSL client, server and email already have functions
15774 in place for compatibility: they check the NID and also return "trusted"
15775 if the certificate is self signed.
15779 * Add d2i,i2d bio/fp functions for PrivateKey: these convert the
15780 traditional format into an EVP_PKEY structure.
15784 * Add a password callback function PEM_cb() which either prompts for
15785 a password if usr_data is NULL or otherwise assumes it is a null
15786 terminated password. Allow passwords to be passed on command line
15787 environment or config files in a few more utilities.
15791 * Add a bunch of DER and PEM functions to handle PKCS#8 format private
15792 keys. Add some short names for PKCS#8 PBE algorithms and allow them
15793 to be specified on the command line for the pkcs8 and pkcs12 utilities.
15794 Update documentation.
15798 * Support for ASN1 "NULL" type. This could be handled before by using
15799 ASN1_TYPE but there wasn't any function that would try to read a NULL
15800 and produce an error if it couldn't. For compatibility we also have
15801 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
15802 don't allocate anything because they don't need to.
15806 * Initial support for MacOS is now provided. Examine INSTALL.MacOS
15809 *Andy Polyakov, Roy Woods <roy@centicsystems.ca>*
15811 * Rebuild of the memory allocation routines used by OpenSSL code and
15812 possibly others as well. The purpose is to make an interface that
15813 provide hooks so anyone can build a separate set of allocation and
15814 deallocation routines to be used by OpenSSL, for example memory
15815 pool implementations, or something else, which was previously hard
15816 since Malloc(), Realloc() and Free() were defined as macros having
15817 the values malloc, realloc and free, respectively (except for Win32
15818 compilations). The same is provided for memory debugging code.
15819 OpenSSL already comes with functionality to find memory leaks, but
15820 this gives people a chance to debug other memory problems.
15822 With these changes, a new set of functions and macros have appeared:
15824 CRYPTO_set_mem_debug_functions() [F]
15825 CRYPTO_get_mem_debug_functions() [F]
15826 CRYPTO_dbg_set_options() [F]
15827 CRYPTO_dbg_get_options() [F]
15828 CRYPTO_malloc_debug_init() [M]
15830 The memory debug functions are NULL by default, unless the library
15831 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
15832 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
15833 gives the standard debugging functions that come with OpenSSL) or
15834 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
15835 provided by the library user) must be used. When the standard
15836 debugging functions are used, CRYPTO_dbg_set_options can be used to
15837 request additional information:
15838 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
15839 the CRYPTO_MDEBUG_xxx macro when compiling the library.
15841 Also, things like CRYPTO_set_mem_functions will always give the
15842 expected result (the new set of functions is used for allocation
15843 and deallocation) at all times, regardless of platform and compiler
15846 To finish it up, some functions that were never use in any other
15847 way than through macros have a new API and new semantic:
15849 CRYPTO_dbg_malloc()
15850 CRYPTO_dbg_realloc()
15853 All macros of value have retained their old syntax.
15855 *Richard Levitte and Bodo Moeller*
15857 * Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
15858 ordering of SMIMECapabilities wasn't in "strength order" and there
15859 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
15864 * Some ASN1 types with illegal zero length encoding (INTEGER,
15865 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
15867 *Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson*
15869 * Merge in my S/MIME library for OpenSSL. This provides a simple
15870 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
15871 functionality to handle multipart/signed properly) and a utility
15872 called 'smime' to call all this stuff. This is based on code I
15873 originally wrote for Celo who have kindly allowed it to be
15874 included in OpenSSL.
15878 * Add variants des_set_key_checked and des_set_key_unchecked of
15879 des_set_key (aka des_key_sched). Global variable des_check_key
15880 decides which of these is called by des_set_key; this way
15881 des_check_key behaves as it always did, but applications and
15882 the library itself, which was buggy for des_check_key == 1,
15883 have a cleaner way to pick the version they need.
15887 * New function PKCS12_newpass() which changes the password of a
15892 * Modify X509_TRUST and X509_PURPOSE so it also uses a static and
15893 dynamic mix. In both cases the ids can be used as an index into the
15894 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
15895 functions so they accept a list of the field values and the
15896 application doesn't need to directly manipulate the X509_TRUST
15901 * Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
15906 * Modify the way the V3 extension code looks up extensions. This now
15907 works in a similar way to the object code: we have some "standard"
15908 extensions in a static table which is searched with OBJ_bsearch()
15909 and the application can add dynamic ones if needed. The file
15910 crypto/x509v3/ext_dat.h now has the info: this file needs to be
15911 updated whenever a new extension is added to the core code and kept
15912 in ext_nid order. There is a simple program 'tabtest.c' which checks
15913 this. New extensions are not added too often so this file can readily
15914 be maintained manually.
15916 There are two big advantages in doing things this way. The extensions
15917 can be looked up immediately and no longer need to be "added" using
15918 X509V3_add_standard_extensions(): this function now does nothing.
15919 Side note: I get *lots* of email saying the extension code doesn't
15920 work because people forget to call this function.
15921 Also no dynamic allocation is done unless new extensions are added:
15922 so if we don't add custom extensions there is no need to call
15923 X509V3_EXT_cleanup().
15927 * Modify enc utility's salting as follows: make salting the default. Add a
15928 magic header, so unsalted files fail gracefully instead of just decrypting
15929 to garbage. This is because not salting is a big security hole, so people
15930 should be discouraged from doing it.
15934 * Fixes and enhancements to the 'x509' utility. It allowed a message
15935 digest to be passed on the command line but it only used this
15936 parameter when signing a certificate. Modified so all relevant
15937 operations are affected by the digest parameter including the
15938 -fingerprint and -x509toreq options. Also -x509toreq choked if a
15939 DSA key was used because it didn't fix the digest.
15943 * Initial certificate chain verify code. Currently tests the untrusted
15944 certificates for consistency with the verify purpose (which is set
15945 when the X509_STORE_CTX structure is set up) and checks the pathlength.
15947 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
15948 this is because it will reject chains with invalid extensions whereas
15949 every previous version of OpenSSL and SSLeay made no checks at all.
15951 Trust code: checks the root CA for the relevant trust settings. Trust
15952 settings have an initial value consistent with the verify purpose: e.g.
15953 if the verify purpose is for SSL client use it expects the CA to be
15954 trusted for SSL client use. However the default value can be changed to
15955 permit custom trust settings: one example of this would be to only trust
15956 certificates from a specific "secure" set of CAs.
15958 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
15959 which should be used for version portability: especially since the
15960 verify structure is likely to change more often now.
15962 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
15963 to set them. If not set then assume SSL clients will verify SSL servers
15966 Two new options to the verify program: -untrusted allows a set of
15967 untrusted certificates to be passed in and -purpose which sets the
15968 intended purpose of the certificate. If a purpose is set then the
15969 new chain verify code is used to check extension consistency.
15973 * Support for the authority information access extension.
15977 * Modify RSA and DSA PEM read routines to transparently handle
15978 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
15979 public keys in a format compatible with certificate
15980 SubjectPublicKeyInfo structures. Unfortunately there were already
15981 functions called *_PublicKey_* which used various odd formats so
15982 these are retained for compatibility: however the DSA variants were
15983 never in a public release so they have been deleted. Changed dsa/rsa
15984 utilities to handle the new format: note no releases ever handled public
15985 keys so we should be OK.
15987 The primary motivation for this change is to avoid the same fiasco
15988 that dogs private keys: there are several incompatible private key
15989 formats some of which are standard and some OpenSSL specific and
15990 require various evil hacks to allow partial transparent handling and
15991 even then it doesn't work with DER formats. Given the option anything
15992 other than PKCS#8 should be dumped: but the other formats have to
15993 stay in the name of compatibility.
15995 With public keys and the benefit of hindsight one standard format
15996 is used which works with EVP_PKEY, RSA or DSA structures: though
15997 it clearly returns an error if you try to read the wrong kind of key.
15999 Added a -pubkey option to the 'x509' utility to output the public key.
16000 Also rename the `EVP_PKEY_get_*()` to `EVP_PKEY_rget_*()`
16001 (renamed to `EVP_PKEY_get1_*()` in the OpenSSL 0.9.5 release) and add
16002 `EVP_PKEY_rset_*()` functions (renamed to `EVP_PKEY_set1_*()`)
16003 that do the same as the `EVP_PKEY_assign_*()` except they up the
16004 reference count of the added key (they don't "swallow" the
16009 * Fixes to crypto/x509/by_file.c the code to read in certificates and
16010 CRLs would fail if the file contained no certificates or no CRLs:
16011 added a new function to read in both types and return the number
16012 read: this means that if none are read it will be an error. The
16013 DER versions of the certificate and CRL reader would always fail
16014 because it isn't possible to mix certificates and CRLs in DER format
16015 without choking one or the other routine. Changed this to just read
16016 a certificate: this is the best we can do. Also modified the code
16017 in apps/verify.c to take notice of return codes: it was previously
16018 attempting to read in certificates from NULL pointers and ignoring
16019 any errors: this is one reason why the cert and CRL reader seemed
16020 to work. It doesn't check return codes from the default certificate
16021 routines: these may well fail if the certificates aren't installed.
16025 * Code to support otherName option in GeneralName.
16029 * First update to verify code. Change the verify utility
16030 so it warns if it is passed a self signed certificate:
16031 for consistency with the normal behaviour. X509_verify
16032 has been modified to it will now verify a self signed
16033 certificate if *exactly* the same certificate appears
16034 in the store: it was previously impossible to trust a
16035 single self signed certificate. This means that:
16036 openssl verify ss.pem
16037 now gives a warning about a self signed certificate but
16038 openssl verify -CAfile ss.pem ss.pem
16043 * For servers, store verify_result in SSL_SESSION data structure
16044 (and add it to external session representation).
16045 This is needed when client certificate verifications fails,
16046 but an application-provided verification callback (set by
16047 SSL_CTX_set_cert_verify_callback) allows accepting the session
16048 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
16049 but returns 1): When the session is reused, we have to set
16050 ssl->verify_result to the appropriate error code to avoid
16053 *Bodo Moeller, problem pointed out by Lutz Jaenicke*
16055 * Fix a bug in the new PKCS#7 code: it didn't consider the
16056 case in PKCS7_dataInit() where the signed PKCS7 structure
16057 didn't contain any existing data because it was being created.
16059 *Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson*
16061 * Add a salt to the key derivation routines in enc.c. This
16062 forms the first 8 bytes of the encrypted file. Also add a
16063 -S option to allow a salt to be input on the command line.
16067 * New function X509_cmp(). Oddly enough there wasn't a function
16068 to compare two certificates. We do this by working out the SHA1
16069 hash and comparing that. X509_cmp() will be needed by the trust
16074 * SSL_get1_session() is like SSL_get_session(), but increments
16075 the reference count in the SSL_SESSION returned.
16077 *Geoff Thorpe <geoff@eu.c2.net>*
16079 * Fix for 'req': it was adding a null to request attributes.
16080 Also change the X509_LOOKUP and X509_INFO code to handle
16081 certificate auxiliary information.
16085 * Add support for 40 and 64 bit RC2 and RC4 algorithms: document
16090 * Add the possibility to add extra information to the memory leak
16091 detecting output, to form tracebacks, showing from where each
16092 allocation was originated: CRYPTO_push_info("constant string") adds
16093 the string plus current file name and line number to a per-thread
16094 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
16095 is like calling CYRPTO_pop_info() until the stack is empty.
16096 Also updated memory leak detection code to be multi-thread-safe.
16100 * Add options -text and -noout to pkcs7 utility and delete the
16101 encryption options which never did anything. Update docs.
16105 * Add options to some of the utilities to allow the pass phrase
16106 to be included on either the command line (not recommended on
16107 OSes like Unix) or read from the environment. Update the
16108 manpages and fix a few bugs.
16112 * Add a few manpages for some of the openssl commands.
16116 * Fix the -revoke option in ca. It was freeing up memory twice,
16117 leaking and not finding already revoked certificates.
16121 * Extensive changes to support certificate auxiliary information.
16122 This involves the use of X509_CERT_AUX structure and X509_AUX
16123 functions. An X509_AUX function such as PEM_read_X509_AUX()
16124 can still read in a certificate file in the usual way but it
16125 will also read in any additional "auxiliary information". By
16126 doing things this way a fair degree of compatibility can be
16127 retained: existing certificates can have this information added
16128 using the new 'x509' options.
16130 Current auxiliary information includes an "alias" and some trust
16131 settings. The trust settings will ultimately be used in enhanced
16132 certificate chain verification routines: currently a certificate
16133 can only be trusted if it is self signed and then it is trusted
16138 * Fix assembler for Alpha (tested only on DEC OSF not Linux or `*BSD`).
16139 The problem was that one of the replacement routines had not been working
16140 since SSLeay releases. For now the offending routine has been replaced
16141 with non-optimised assembler. Even so, this now gives around 95%
16142 performance improvement for 1024 bit RSA signs.
16146 * Hack to fix PKCS#7 decryption when used with some unorthodox RC2
16147 handling. Most clients have the effective key size in bits equal to
16148 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
16149 A few however don't do this and instead use the size of the decrypted key
16150 to determine the RC2 key length and the AlgorithmIdentifier to determine
16151 the effective key length. In this case the effective key length can still
16152 be 40 bits but the key length can be 168 bits for example. This is fixed
16153 by manually forcing an RC2 key into the EVP_PKEY structure because the
16154 EVP code can't currently handle unusual RC2 key sizes: it always assumes
16155 the key length and effective key length are equal.
16159 * Add a bunch of functions that should simplify the creation of
16160 X509_NAME structures. Now you should be able to do:
16161 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
16162 and have it automatically work out the correct field type and fill in
16163 the structures. The more adventurous can try:
16164 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
16165 and it will (hopefully) work out the correct multibyte encoding.
16169 * Change the 'req' utility to use the new field handling and multibyte
16170 copy routines. Before the DN field creation was handled in an ad hoc
16171 way in req, ca, and x509 which was rather broken and didn't support
16172 BMPStrings or UTF8Strings. Since some software doesn't implement
16173 BMPStrings or UTF8Strings yet, they can be enabled using the config file
16174 using the dirstring_type option. See the new comment in the default
16175 openssl.cnf for more info.
16179 * Make crypto/rand/md_rand.c more robust:
16180 - Assure unique random numbers after fork().
16181 - Make sure that concurrent threads access the global counter and
16182 md serializably so that we never lose entropy in them
16183 or use exactly the same state in multiple threads.
16184 Access to the large state is not always serializable because
16185 the additional locking could be a performance killer, and
16186 md should be large enough anyway.
16190 * New file apps/app_rand.c with commonly needed functionality
16191 for handling the random seed file.
16193 Use the random seed file in some applications that previously did not:
16195 dsaparam -genkey (which also ignored its '-rand' option),
16198 x509 (when signing).
16199 Except on systems with /dev/urandom, it is crucial to have a random
16200 seed file at least for key creation, DSA signing, and for DH exchanges;
16201 for RSA signatures we could do without one.
16203 gendh and gendsa (unlike genrsa) used to read only the first byte
16204 of each file listed in the '-rand' option. The function as previously
16205 found in genrsa is now in app_rand.c and is used by all programs
16206 that support '-rand'.
16210 * In RAND_write_file, use mode 0600 for creating files;
16211 don't just chmod when it may be too late.
16215 * Report an error from X509_STORE_load_locations
16216 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
16220 * New function ASN1_mbstring_copy() this copies a string in either
16221 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
16222 into an ASN1_STRING type. A mask of permissible types is passed
16223 and it chooses the "minimal" type to use or an error if not type
16228 * Add function equivalents to the various macros in asn1.h. The old
16229 macros are retained with an `M_` prefix. Code inside the library can
16230 use the `M_` macros. External code (including the openssl utility)
16231 should *NOT* in order to be "shared library friendly".
16235 * Add various functions that can check a certificate's extensions
16236 to see if it usable for various purposes such as SSL client,
16237 server or S/MIME and CAs of these types. This is currently
16238 VERY EXPERIMENTAL but will ultimately be used for certificate chain
16239 verification. Also added a -purpose flag to x509 utility to
16240 print out all the purposes.
16244 * Add a CRYPTO_EX_DATA to X509 certificate structure and associated
16249 * New `X509V3_{X509,CRL,REVOKED}_get_d2i()` functions. These will search
16250 for, obtain and decode and extension and obtain its critical flag.
16251 This allows all the necessary extension code to be handled in a
16252 single function call.
16256 * RC4 tune-up featuring 30-40% performance improvement on most RISC
16257 platforms. See crypto/rc4/rc4_enc.c for further details.
16261 * New -noout option to asn1parse. This causes no output to be produced
16262 its main use is when combined with -strparse and -out to extract data
16263 from a file (which may not be in ASN.1 format).
16267 * Fix for pkcs12 program. It was hashing an invalid certificate pointer
16268 when producing the local key id.
16270 *Richard Levitte <levitte@stacken.kth.se>*
16272 * New option -dhparam in s_server. This allows a DH parameter file to be
16273 stated explicitly. If it is not stated then it tries the first server
16274 certificate file. The previous behaviour hard coded the filename
16279 * Add -pubin and -pubout options to the rsa and dsa commands. These allow
16280 a public key to be input or output. For example:
16281 openssl rsa -in key.pem -pubout -out pubkey.pem
16282 Also added necessary DSA public key functions to handle this.
16286 * Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
16287 in the message. This was handled by allowing
16288 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
16290 *Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>*
16292 * Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
16293 to the end of the strings whereas this didn't. This would cause problems
16294 if strings read with d2i_ASN1_bytes() were later modified.
16296 *Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>*
16298 * Fix for base64 decode bug. When a base64 bio reads only one line of
16299 data and it contains EOF it will end up returning an error. This is
16300 caused by input 46 bytes long. The cause is due to the way base64
16301 BIOs find the start of base64 encoded data. They do this by trying a
16302 trial decode on each line until they find one that works. When they
16303 do a flag is set and it starts again knowing it can pass all the
16304 data directly through the decoder. Unfortunately it doesn't reset
16305 the context it uses. This means that if EOF is reached an attempt
16306 is made to pass two EOFs through the context and this causes the
16307 resulting error. This can also cause other problems as well. As is
16308 usual with these problems it takes *ages* to find and the fix is
16309 trivial: move one line.
16311 *Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer)*
16313 * Ugly workaround to get s_client and s_server working under Windows. The
16314 old code wouldn't work because it needed to select() on sockets and the
16315 tty (for keypresses and to see if data could be written). Win32 only
16316 supports select() on sockets so we select() with a 1s timeout on the
16317 sockets and then see if any characters are waiting to be read, if none
16318 are present then we retry, we also assume we can always write data to
16319 the tty. This isn't nice because the code then blocks until we've
16320 received a complete line of data and it is effectively polling the
16321 keyboard at 1s intervals: however it's quite a bit better than not
16322 working at all :-) A dedicated Windows application might handle this
16323 with an event loop for example.
16327 * Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
16328 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
16329 will be called when RSA_sign() and RSA_verify() are used. This is useful
16330 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
16331 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
16332 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
16333 This necessitated the support of an extra signature type NID_md5_sha1
16334 for SSL signatures and modifications to the SSL library to use it instead
16335 of calling RSA_public_decrypt() and RSA_private_encrypt().
16339 * Add new -verify -CAfile and -CApath options to the crl program, these
16340 will lookup a CRL issuers certificate and verify the signature in a
16341 similar way to the verify program. Tidy up the crl program so it
16342 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
16343 less strict. It will now permit CRL extensions even if it is not
16344 a V2 CRL: this will allow it to tolerate some broken CRLs.
16348 * Initialize all non-automatic variables each time one of the openssl
16349 sub-programs is started (this is necessary as they may be started
16350 multiple times from the "OpenSSL>" prompt).
16352 *Lennart Bang, Bodo Moeller*
16354 * Preliminary compilation option RSA_NULL which disables RSA crypto without
16355 removing all other RSA functionality (this is what NO_RSA does). This
16356 is so (for example) those in the US can disable those operations covered
16357 by the RSA patent while allowing storage and parsing of RSA keys and RSA
16362 * Non-copying interface to BIO pairs.
16363 (still largely untested)
16367 * New function ASN1_tag2str() to convert an ASN1 tag to a descriptive
16368 ASCII string. This was handled independently in various places before.
16372 * New functions UTF8_getc() and UTF8_putc() that parse and generate
16373 UTF8 strings a character at a time.
16377 * Use client_version from client hello to select the protocol
16378 (s23_srvr.c) and for RSA client key exchange verification
16379 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
16383 * Add various utility functions to handle SPKACs, these were previously
16384 handled by poking round in the structure internals. Added new function
16385 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
16386 print, verify and generate SPKACs. Based on an original idea from
16387 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
16391 * RIPEMD160 is operational on all platforms and is back in 'make test'.
16395 * Allow the config file extension section to be overwritten on the
16396 command line. Based on an original idea from Massimiliano Pala
16397 <madwolf@comune.modena.it>. The new option is called -extensions
16398 and can be applied to ca, req and x509. Also -reqexts to override
16399 the request extensions in req and -crlexts to override the crl extensions
16404 * Add new feature to the SPKAC handling in ca. Now you can include
16405 the same field multiple times by preceding it by "XXXX." for example:
16408 this is the same syntax as used in the req config file.
16412 * Allow certificate extensions to be added to certificate requests. These
16413 are specified in a 'req_extensions' option of the req section of the
16414 config file. They can be printed out with the -text option to req but
16415 are otherwise ignored at present.
16419 * Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
16420 data read consists of only the final block it would not decrypted because
16421 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
16422 A misplaced 'break' also meant the decrypted final block might not be
16423 copied until the next read.
16427 * Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
16428 a few extra parameters to the DH structure: these will be useful if
16429 for example we want the value of 'q' or implement X9.42 DH.
16433 * Initial support for DSA_METHOD. This is based on the RSA_METHOD and
16434 provides hooks that allow the default DSA functions or functions on a
16435 "per key" basis to be replaced. This allows hardware acceleration and
16436 hardware key storage to be handled without major modification to the
16437 library. Also added low level modexp hooks and CRYPTO_EX structure and
16438 associated functions.
16442 * Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
16443 as "read only": it can't be written to and the buffer it points to will
16444 not be freed. Reading from a read only BIO is much more efficient than
16445 a normal memory BIO. This was added because there are several times when
16446 an area of memory needs to be read from a BIO. The previous method was
16447 to create a memory BIO and write the data to it, this results in two
16448 copies of the data and an O(n^2) reading algorithm. There is a new
16449 function BIO_new_mem_buf() which creates a read only memory BIO from
16450 an area of memory. Also modified the PKCS#7 routines to use read only
16455 * Bugfix: ssl23_get_client_hello did not work properly when called in
16456 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
16457 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
16458 but a retry condition occurred while trying to read the rest.
16462 * The PKCS7_ENC_CONTENT_new() function was setting the content type as
16463 NID_pkcs7_encrypted by default: this was wrong since this should almost
16464 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
16465 the encrypted data type: this is a more sensible place to put it and it
16466 allows the PKCS#12 code to be tidied up that duplicated this
16471 * Changed obj_dat.pl script so it takes its input and output files on
16472 the command line. This should avoid shell escape redirection problems
16477 * Initial support for certificate extension requests, these are included
16478 in things like Xenroll certificate requests. Included functions to allow
16479 extensions to be obtained and added.
16483 * -crlf option to s_client and s_server for sending newlines as
16484 CRLF (as required by many protocols).
16488 ### Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
16490 * Install libRSAglue.a when OpenSSL is built with RSAref.
16492 *Ralf S. Engelschall*
16494 * A few more `#ifndef NO_FP_API / #endif` pairs for consistency.
16496 *Andrija Antonijevic <TheAntony2@bigfoot.com>*
16498 * Fix -startdate and -enddate (which was missing) arguments to 'ca'
16503 * New function DSA_dup_DH, which duplicates DSA parameters/keys as
16504 DH parameters/keys (q is lost during that conversion, but the resulting
16505 DH parameters contain its length).
16507 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
16508 much faster than DH_generate_parameters (which creates parameters
16509 where `p = 2*q + 1`), and also the smaller q makes DH computations
16510 much more efficient (160-bit exponentiation instead of 1024-bit
16511 exponentiation); so this provides a convenient way to support DHE
16512 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
16513 utter importance to use
16514 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
16516 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
16517 when such DH parameters are used, because otherwise small subgroup
16518 attacks may become possible!
16522 * Avoid memory leak in i2d_DHparams.
16526 * Allow the -k option to be used more than once in the enc program:
16527 this allows the same encrypted message to be read by multiple recipients.
16531 * New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
16532 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
16533 it will always use the numerical form of the OID, even if it has a short
16538 * Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
16539 method only got called if p,q,dmp1,dmq1,iqmp components were present,
16540 otherwise bn_mod_exp was called. In the case of hardware keys for example
16541 no private key components need be present and it might store extra data
16542 in the RSA structure, which cannot be accessed from bn_mod_exp.
16543 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
16544 private key operations.
16548 * Added support for SPARC Linux.
16552 * pem_password_cb function type incompatibly changed from
16553 typedef int pem_password_cb(char *buf, int size, int rwflag);
16555 ....(char *buf, int size, int rwflag, void *userdata);
16556 so that applications can pass data to their callbacks:
16557 The `PEM[_ASN1]_{read,write}...` functions and macros now take an
16558 additional void * argument, which is just handed through whenever
16559 the password callback is called.
16561 *Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller*
16563 New function SSL_CTX_set_default_passwd_cb_userdata.
16565 Compatibility note: As many C implementations push function arguments
16566 onto the stack in reverse order, the new library version is likely to
16567 interoperate with programs that have been compiled with the old
16568 pem_password_cb definition (PEM_whatever takes some data that
16569 happens to be on the stack as its last argument, and the callback
16570 just ignores this garbage); but there is no guarantee whatsoever that
16573 * The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
16574 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
16575 problems not only on Windows, but also on some Unix platforms.
16576 To avoid problematic command lines, these definitions are now in an
16577 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
16578 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
16582 * MIPS III/IV assembler module is reimplemented.
16586 * More DES library cleanups: remove references to srand/rand and
16587 delete an unused file.
16591 * Add support for the free Netwide assembler (NASM) under Win32,
16592 since not many people have MASM (ml) and it can be hard to obtain.
16593 This is currently experimental but it seems to work OK and pass all
16594 the tests. Check out INSTALL.W32 for info.
16598 * Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
16599 without temporary keys kept an extra copy of the server key,
16600 and connections with temporary keys did not free everything in case
16605 * New function RSA_check_key and new openssl rsa option -check
16606 for verifying the consistency of RSA keys.
16608 *Ulf Moeller, Bodo Moeller*
16610 * Various changes to make Win32 compile work:
16611 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
16612 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
16613 comparison" warnings.
16614 3. Add `sk_<TYPE>_sort` to DEF file generator and do make update.
16618 * Add a debugging option to PKCS#5 v2 key generation function: when
16619 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
16620 derived keys are printed to stderr.
16624 * Copy the flags in ASN1_STRING_dup().
16626 *Roman E. Pavlov <pre@mo.msk.ru>*
16628 * The x509 application mishandled signing requests containing DSA
16629 keys when the signing key was also DSA and the parameters didn't match.
16631 It was supposed to omit the parameters when they matched the signing key:
16632 the verifying software was then supposed to automatically use the CA's
16633 parameters if they were absent from the end user certificate.
16635 Omitting parameters is no longer recommended. The test was also
16636 the wrong way round! This was probably due to unusual behaviour in
16637 EVP_cmp_parameters() which returns 1 if the parameters match.
16638 This meant that parameters were omitted when they *didn't* match and
16639 the certificate was useless. Certificates signed with 'ca' didn't have
16642 *Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>*
16644 * Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
16645 The interface is as follows:
16646 Applications can use
16647 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
16648 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
16649 "off" is now the default.
16650 The library internally uses
16651 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
16652 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
16653 to disable memory-checking temporarily.
16655 Some inconsistent states that previously were possible (and were
16656 even the default) are now avoided.
16658 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
16659 with each memory chunk allocated; this is occasionally more helpful
16660 than just having a counter.
16662 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
16664 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
16669 * Introduce "mode" for SSL structures (with defaults in SSL_CTX),
16670 which largely parallels "options", but is for changing API behaviour,
16671 whereas "options" are about protocol behaviour.
16672 Initial "mode" flags are:
16674 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
16675 a single record has been written.
16676 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
16677 retries use the same buffer location.
16678 (But all of the contents must be
16683 * Bugfix: SSL_set_options ignored its parameter, only SSL_CTX_set_options
16686 * Fix problems with no-hmac etc.
16688 *Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>*
16690 * New functions RSA_get_default_method(), RSA_set_method() and
16691 RSA_get_method(). These allows replacement of RSA_METHODs without having
16692 to mess around with the internals of an RSA structure.
16696 * Fix memory leaks in DSA_do_sign and DSA_is_prime.
16697 Also really enable memory leak checks in openssl.c and in some
16700 *Chad C. Mulligan, Bodo Moeller*
16702 * Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
16703 up the length of negative integers. This has now been simplified to just
16704 store the length when it is first determined and use it later, rather
16705 than trying to keep track of where data is copied and updating it to
16707 *Steve Henson, reported by Brien Wheeler <bwheeler@authentica-security.com>*
16709 * Add a new function PKCS7_signatureVerify. This allows the verification
16710 of a PKCS#7 signature but with the signing certificate passed to the
16711 function itself. This contrasts with PKCS7_dataVerify which assumes the
16712 certificate is present in the PKCS#7 structure. This isn't always the
16713 case: certificates can be omitted from a PKCS#7 structure and be
16714 distributed by "out of band" means (such as a certificate database).
16718 * Complete the `PEM_*` macros with DECLARE_PEM versions to replace the
16719 function prototypes in pem.h, also change util/mkdef.pl to add the
16720 necessary function names.
16724 * mk1mf.pl (used by Windows builds) did not properly read the
16725 options set by Configure in the top level Makefile, and Configure
16726 was not even able to write more than one option correctly.
16727 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
16731 * New functions CONF_load_bio() and CONF_load_fp() to allow a config
16732 file to be loaded from a BIO or FILE pointer. The BIO version will
16733 for example allow memory BIOs to contain config info.
16737 * New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
16738 Whoever hopes to achieve shared-library compatibility across versions
16739 must use this, not the compile-time macro.
16740 (Exercise 0.9.4: Which is the minimum library version required by
16742 Note: All this applies only to multi-threaded programs, others don't
16747 * Add missing case to s3_clnt.c state machine -- one of the new SSL tests
16748 through a BIO pair triggered the default case, i.e.
16749 SSLerr(...,SSL_R_UNKNOWN_STATE).
16753 * New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
16754 can use the SSL library even if none of the specific BIOs is
16759 * Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
16760 for the encoded length.
16762 *Jeon KyoungHo <khjeon@sds.samsung.co.kr>*
16764 * Add initial documentation of the X509V3 functions.
16768 * Add a new pair of functions PEM_write_PKCS8PrivateKey() and
16769 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
16770 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
16771 secure PKCS#8 private key format with a high iteration count.
16775 * Fix determination of Perl interpreter: A perl or perl5
16776 *directory* in $PATH was also accepted as the interpreter.
16778 *Ralf S. Engelschall*
16780 * Fix demos/sign/sign.c: well there wasn't anything strictly speaking
16781 wrong with it but it was very old and did things like calling
16782 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
16783 unusual formatting.
16787 * Fix demos/selfsign.c: it used obsolete and deleted functions, changed
16788 to use the new extension code.
16792 * Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
16793 with macros. This should make it easier to change their form, add extra
16794 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
16799 * Add to configuration table a new entry that can specify an alternative
16800 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
16801 according to Mark Crispin <MRC@Panda.COM>.
16805 * DES CBC did not update the IV. Weird.
16809 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
16810 Changing the behaviour of the former might break existing programs --
16811 where IV updating is needed, des_ncbc_encrypt can be used.
16814 * When bntest is run from "make test" it drives bc to check its
16815 calculations, as well as internally checking them. If an internal check
16816 fails, it needs to cause bc to give a non-zero result or make test carries
16817 on without noticing the failure. Fixed.
16821 * DES library cleanups.
16825 * Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
16826 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
16827 ciphers. NOTE: although the key derivation function has been verified
16828 against some published test vectors it has not been extensively tested
16829 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
16834 * Instead of "mkdir -p", which is not fully portable, use new
16835 Perl script "util/mkdir-p.pl".
16839 * Rewrite the way password based encryption (PBE) is handled. It used to
16840 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
16841 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
16842 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
16843 the 'parameter' field of the AlgorithmIdentifier is passed to the
16844 underlying key generation function so it must do its own ASN1 parsing.
16845 This has also changed the EVP_PBE_CipherInit() function which now has a
16846 'parameter' argument instead of literal salt and iteration count values
16847 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
16851 * Support for PKCS#5 v1.5 compatible password based encryption algorithms
16852 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
16853 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
16854 KEY" because this clashed with PKCS#8 unencrypted string. Since this
16855 value was just used as a "magic string" and not used directly its
16856 value doesn't matter.
16860 * Introduce some semblance of const correctness to BN. Shame C doesn't
16865 * "linux-sparc64" configuration (ultrapenguin).
16867 *Ray Miller <ray.miller@oucs.ox.ac.uk>*
16868 "linux-sparc" configuration.
16870 *Christian Forster <fo@hawo.stw.uni-erlangen.de>*
16872 * config now generates no-xxx options for missing ciphers.
16876 * Support the EBCDIC character set (work in progress).
16877 File ebcdic.c not yet included because it has a different license.
16879 *Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>*
16881 * Support BS2000/OSD-POSIX.
16883 *Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>*
16885 * Make callbacks for key generation use `void *` instead of `char *`.
16889 * Make S/MIME samples compile (not yet tested).
16893 * Additional typesafe stacks.
16897 * New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
16901 ### Changes between 0.9.3 and 0.9.3a [29 May 1999]
16903 * New configuration variant "sco5-gcc".
16905 * Updated some demos.
16907 *Sean O Riordain, Wade Scholine*
16909 * Add missing BIO_free at exit of pkcs12 application.
16913 * Fix memory leak in conf.c.
16917 * Updates for Win32 to assembler version of MD5.
16921 * Set #! path to perl in apps/der_chop to where we found it
16922 instead of using a fixed path.
16926 * SHA library changes for irix64-mips4-cc.
16930 * Improvements for VMS support.
16934 ### Changes between 0.9.2b and 0.9.3 [24 May 1999]
16936 * Bignum library bug fix. IRIX 6 passes "make test" now!
16937 This also avoids the problems with SC4.2 and unpatched SC5.
16939 *Andy Polyakov <appro@fy.chalmers.se>*
16941 * New functions sk_num, sk_value and sk_set to replace the previous macros.
16942 These are required because of the typesafe stack would otherwise break
16943 existing code. If old code used a structure member which used to be STACK
16944 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
16945 sk_num or sk_value it would produce an error because the num, data members
16946 are not present in STACK_OF. Now it just produces a warning. sk_set
16947 replaces the old method of assigning a value to sk_value
16948 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
16949 that does this will no longer work (and should use sk_set instead) but
16950 this could be regarded as a "questionable" behaviour anyway.
16954 * Fix most of the other PKCS#7 bugs. The "experimental" code can now
16955 correctly handle encrypted S/MIME data.
16959 * Change type of various DES function arguments from des_cblock
16960 (which means, in function argument declarations, pointer to char)
16961 to des_cblock * (meaning pointer to array with 8 char elements),
16962 which allows the compiler to do more typechecking; it was like
16963 that back in SSLeay, but with lots of ugly casts.
16965 Introduce new type const_des_cblock.
16969 * Reorganise the PKCS#7 library and get rid of some of the more obvious
16970 problems: find RecipientInfo structure that matches recipient certificate
16971 and initialise the ASN1 structures properly based on passed cipher.
16975 * Belatedly make the BN tests actually check the results.
16979 * Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
16980 to and from BNs: it was completely broken. New compilation option
16981 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
16982 key elements as negative integers.
16986 * Reorganize and speed up MD5.
16988 *Andy Polyakov <appro@fy.chalmers.se>*
16992 *Richard Levitte <richard@levitte.org>*
16994 * New option -out to asn1parse to allow the parsed structure to be
16995 output to a file. This is most useful when combined with the -strparse
16996 option to examine the output of things like OCTET STRINGS.
17000 * Make SSL library a little more fool-proof by not requiring any longer
17001 that `SSL_set_{accept,connect}_state` be called before
17002 `SSL_{accept,connect}` may be used (`SSL_set_..._state` is omitted
17003 in many applications because usually everything *appeared* to work as
17004 intended anyway -- now it really works as intended).
17008 * Move openssl.cnf out of lib/.
17012 * Fix various things to let OpenSSL even pass "egcc -pipe -O2 -Wall
17013 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
17014 -Wmissing-declarations -Wnested-externs -Winline" with EGCS 1.1.2+
17016 *Ralf S. Engelschall*
17018 * Various fixes to the EVP and PKCS#7 code. It may now be able to
17019 handle PKCS#7 enveloped data properly.
17021 *Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve*
17023 * Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
17024 copying pointers. The cert_st handling is changed by this in
17025 various ways (and thus what used to be known as ctx->default_cert
17026 is now called ctx->cert, since we don't resort to `s->ctx->[default_]cert`
17027 any longer when s->cert does not give us what we need).
17028 ssl_cert_instantiate becomes obsolete by this change.
17029 As soon as we've got the new code right (possibly it already is?),
17030 we have solved a couple of bugs of the earlier code where s->cert
17031 was used as if it could not have been shared with other SSL structures.
17033 Note that using the SSL API in certain dirty ways now will result
17034 in different behaviour than observed with earlier library versions:
17035 Changing settings for an `SSL_CTX *ctx` after having done s = SSL_new(ctx)
17036 does not influence s as it used to.
17038 In order to clean up things more thoroughly, inside SSL_SESSION
17039 we don't use CERT any longer, but a new structure SESS_CERT
17040 that holds per-session data (if available); currently, this is
17041 the peer's certificate chain and, for clients, the server's certificate
17042 and temporary key. CERT holds only those values that can have
17043 meaningful defaults in an SSL_CTX.
17047 * New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
17048 from the internal representation. Various PKCS#7 fixes: remove some
17049 evil casts and set the enc_dig_alg field properly based on the signing
17054 * Allow PKCS#12 password to be set from the command line or the
17055 environment. Let 'ca' get its config file name from the environment
17056 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
17061 * Allow certificate policies extension to use an IA5STRING for the
17062 organization field. This is contrary to the PKIX definition but
17063 VeriSign uses it and IE5 only recognises this form. Document 'x509'
17068 * Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
17069 without disallowing inline assembler and the like for non-pedantic builds.
17073 * Support Borland C++ builder.
17075 *Janez Jere <jj@void.si>, modified by Ulf Möller*
17081 * SHA-1 cleanups and performance enhancements.
17083 *Andy Polyakov <appro@fy.chalmers.se>*
17085 * Sparc v8plus assembler for the bignum library.
17087 *Andy Polyakov <appro@fy.chalmers.se>*
17089 * Accept any -xxx and +xxx compiler options in Configure.
17093 * Update HPUX configuration.
17097 * Add missing `sk_<type>_unshift()` function to safestack.h
17099 *Ralf S. Engelschall*
17101 * New function SSL_CTX_use_certificate_chain_file that sets the
17102 "extra_cert"s in addition to the certificate. (This makes sense
17103 only for "PEM" format files, as chains as a whole are not
17108 * Support verify_depth from the SSL API.
17109 x509_vfy.c had what can be considered an off-by-one-error:
17110 Its depth (which was not part of the external interface)
17111 was actually counting the number of certificates in a chain;
17112 now it really counts the depth.
17116 * Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
17117 instead of X509err, which often resulted in confusing error
17118 messages since the error codes are not globally unique
17119 (e.g. an alleged error in ssl3_accept when a certificate
17120 didn't match the private key).
17122 * New function SSL_CTX_set_session_id_context that allows to set a default
17123 value (so that you don't need SSL_set_session_id_context for each
17124 connection using the SSL_CTX).
17128 * OAEP decoding bug fix.
17132 * Support INSTALL_PREFIX for package builders, as proposed by
17137 * New Configure options "threads" and "no-threads". For systems
17138 where the proper compiler options are known (currently Solaris
17139 and Linux), "threads" is the default.
17143 * New script util/mklink.pl as a faster substitute for util/mklink.sh.
17147 * Install various scripts to $(OPENSSLDIR)/misc, not to
17148 $(INSTALLTOP)/bin -- they shouldn't clutter directories
17149 such as /usr/local/bin.
17153 * "make linux-shared" to build shared libraries.
17155 *Niels Poppe <niels@netbox.org>*
17157 * New Configure option `no-<cipher>` (rsa, idea, rc5, ...).
17161 * Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
17162 extension adding in x509 utility.
17166 * Remove NOPROTO sections and error code comments.
17170 * Partial rewrite of the DEF file generator to now parse the ANSI
17175 * New Configure options --prefix=DIR and --openssldir=DIR.
17179 * Complete rewrite of the error code script(s). It is all now handled
17180 by one script at the top level which handles error code gathering,
17181 header rewriting and C source file generation. It should be much better
17182 than the old method: it now uses a modified version of Ulf's parser to
17183 read the ANSI prototypes in all header files (thus the old K&R definitions
17184 aren't needed for error creation any more) and do a better job of
17185 translating function codes into names. The old 'ASN1 error code embedded
17186 in a comment' is no longer necessary and it doesn't use .err files which
17187 have now been deleted. Also the error code call doesn't have to appear all
17188 on one line (which resulted in some large lines...).
17192 * Change #include filenames from `<foo.h>` to `<openssl/foo.h>`.
17196 * Change behaviour of ssl2_read when facing length-0 packets: Don't return
17197 0 (which usually indicates a closed connection), but continue reading.
17201 * Fix some race conditions.
17205 * Add support for CRL distribution points extension. Add Certificate
17206 Policies and CRL distribution points documentation.
17210 * Move the autogenerated header file parts to crypto/opensslconf.h.
17214 * Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
17215 8 of keying material. Merlin has also confirmed interop with this fix
17216 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
17218 *Merlin Hughes <merlin@baltimore.ie>*
17220 * Fix lots of warnings.
17222 *Richard Levitte <levitte@stacken.kth.se>*
17224 * In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
17225 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
17227 *Richard Levitte <levitte@stacken.kth.se>*
17229 * Fix problems with sizeof(long) == 8.
17231 *Andy Polyakov <appro@fy.chalmers.se>*
17233 * Change functions to ANSI C.
17237 * Fix typos in error codes.
17239 *Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller*
17241 * Remove defunct assembler files from Configure.
17245 * SPARC v8 assembler BIGNUM implementation.
17247 *Andy Polyakov <appro@fy.chalmers.se>*
17249 * Support for Certificate Policies extension: both print and set.
17250 Various additions to support the r2i method this uses.
17254 * A lot of constification, and fix a bug in X509_NAME_oneline() that could
17255 return a const string when you are expecting an allocated buffer.
17259 * Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
17260 types DirectoryString and DisplayText.
17264 * Add code to allow r2i extensions to access the configuration database,
17265 add an LHASH database driver and add several ctx helper functions.
17269 * Fix an evil bug in bn_expand2() which caused various BN functions to
17270 fail when they extended the size of a BIGNUM.
17274 * Various utility functions to handle SXNet extension. Modify mkdef.pl to
17275 support typesafe stack.
17279 * Fix typo in SSL_[gs]et_options().
17281 *Nils Frostberg <nils@medcom.se>*
17283 * Delete various functions and files that belonged to the (now obsolete)
17284 old X509V3 handling code.
17288 * New Configure option "rsaref".
17292 * Don't auto-generate pem.h.
17296 * Introduce type-safe ASN.1 SETs.
17300 * Convert various additional casted stacks to type-safe STACK_OF() variants.
17302 *Ben Laurie, Ralf S. Engelschall, Steve Henson*
17304 * Introduce type-safe STACKs. This will almost certainly break lots of code
17305 that links with OpenSSL (well at least cause lots of warnings), but fear
17306 not: the conversion is trivial, and it eliminates loads of evil casts. A
17307 few STACKed things have been converted already. Feel free to convert more.
17308 In the fullness of time, I'll do away with the STACK type altogether.
17312 * Add `openssl ca -revoke <certfile>` facility which revokes a certificate
17313 specified in `<certfile>` by updating the entry in the index.txt file.
17314 This way one no longer has to edit the index.txt file manually for
17315 revoking a certificate. The -revoke option does the gory details now.
17317 *Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall*
17319 * Fix `openssl crl -noout -text` combination where `-noout` killed the
17320 `-text` option at all and this way the `-noout -text` combination was
17321 inconsistent in `openssl crl` with the friends in `openssl x509|rsa|dsa`.
17323 *Ralf S. Engelschall*
17325 * Make sure a corresponding plain text error message exists for the
17326 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
17327 verify callback function determined that a certificate was revoked.
17329 *Ralf S. Engelschall*
17331 * Bugfix: In test/testenc, don't test `openssl <cipher>` for
17332 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
17333 all available ciphers including rc5, which was forgotten until now.
17334 In order to let the testing shell script know which algorithms
17335 are available, a new (up to now undocumented) command
17336 `openssl list-cipher-commands` is used.
17340 * Bugfix: s_client occasionally would sleep in select() when
17341 it should have checked SSL_pending() first.
17345 * New functions DSA_do_sign and DSA_do_verify to provide access to
17346 the raw DSA values prior to ASN.1 encoding.
17350 * Tweaks to Configure
17352 *Niels Poppe <niels@netbox.org>*
17354 * Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
17359 * New variables $(RANLIB) and $(PERL) in the Makefiles.
17363 * New config option to avoid instructions that are illegal on the 80386.
17364 The default code is faster, but requires at least a 486.
17368 * Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
17369 SSL2_SERVER_VERSION (not used at all) macros, which are now the
17370 same as SSL2_VERSION anyway.
17374 * New "-showcerts" option for s_client.
17378 * Still more PKCS#12 integration. Add pkcs12 application to openssl
17379 application. Various cleanups and fixes.
17383 * More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
17384 modify error routines to work internally. Add error codes and PBE init
17385 to library startup routines.
17389 * Further PKCS#12 integration. Added password based encryption, PKCS#8 and
17390 packing functions to asn1 and evp. Changed function names and error
17391 codes along the way.
17395 * PKCS12 integration: and so it begins... First of several patches to
17396 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
17397 objects to objects.h
17401 * Add a new 'indent' option to some X509V3 extension code. Initial ASN1
17402 and display support for Thawte strong extranet extension.
17406 * Add LinuxPPC support.
17408 *Jeff Dubrule <igor@pobox.org>*
17410 * Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
17411 bn_div_words in alpha.s.
17413 *Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie*
17415 * Make sure the RSA OAEP test is skipped under -DRSAref because
17416 OAEP isn't supported when OpenSSL is built with RSAref.
17418 *Ulf Moeller <ulf@fitug.de>*
17420 * Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
17421 so they no longer are missing under -DNOPROTO.
17423 *Soren S. Jorvang <soren@t.dk>*
17425 ### Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
17427 * Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
17428 doesn't work when the session is reused. Coming soon!
17432 * Fix a security hole, that allows sessions to be reused in the wrong
17433 context thus bypassing client cert protection! All software that uses
17434 client certs and session caches in multiple contexts NEEDS PATCHING to
17435 allow session reuse! A fuller solution is in the works.
17437 *Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)*
17439 * Some more source tree cleanups (removed obsolete files
17440 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
17441 permission on "config" script to be executable) and a fix for the INSTALL
17444 *Ulf Moeller <ulf@fitug.de>*
17446 * Remove some legacy and erroneous uses of malloc, free instead of
17449 *Lennart Bang <lob@netstream.se>, with minor changes by Steve*
17451 * Make rsa_oaep_test return non-zero on error.
17453 *Ulf Moeller <ulf@fitug.de>*
17455 * Add support for native Solaris shared libraries. Configure
17456 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
17457 if someone would make that last step automatic.
17459 *Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>*
17461 * ctx_size was not built with the right compiler during "make links". Fixed.
17465 * Change the meaning of 'ALL' in the cipher list. It now means "everything
17466 except NULL ciphers". This means the default cipher list will no longer
17467 enable NULL ciphers. They need to be specifically enabled e.g. with
17468 the string "DEFAULT:eNULL".
17472 * Fix to RSA private encryption routines: if p < q then it would
17473 occasionally produce an invalid result. This will only happen with
17474 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
17478 * Be less restrictive and allow also `perl util/perlpath.pl
17479 /path/to/bin/perl` in addition to `perl util/perlpath.pl /path/to/bin`,
17480 because this way one can also use an interpreter named `perl5` (which is
17481 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
17482 installed as `perl`).
17484 *Matthias Loepfe <Matthias.Loepfe@adnovum.ch>*
17486 * Let util/clean-depend.pl work also with older Perl 5.00x versions.
17488 *Matthias Loepfe <Matthias.Loepfe@adnovum.ch>*
17490 * Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
17491 advapi32.lib to Win32 build and change the pem test comparison
17492 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
17493 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
17494 and crypto/des/ede_cbcm_enc.c.
17498 * DES quad checksum was broken on big-endian architectures. Fixed.
17502 * Comment out two functions in bio.h that aren't implemented. Fix up the
17503 Win32 test batch file so it (might) work again. The Win32 test batch file
17504 is horrible: I feel ill....
17508 * Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
17509 in e_os.h. Audit of header files to check ANSI and non ANSI
17510 sections: 10 functions were absent from non ANSI section and not exported
17511 from Windows DLLs. Fixed up libeay.num for new functions.
17515 * Make `openssl version` output lines consistent.
17517 *Ralf S. Engelschall*
17519 * Fix Win32 symbol export lists for BIO functions: Added
17520 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
17521 to ms/libeay{16,32}.def.
17523 *Ralf S. Engelschall*
17525 * Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
17526 fine under Unix and passes some trivial tests I've now added. But the
17527 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
17528 added to make sure no one expects that this stuff really works in the
17529 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
17530 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
17533 *Ralf S. Engelschall*
17535 * Fix the generation of two part addresses in perl.
17537 *Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie*
17539 * Add config entry for Linux on MIPS.
17541 *John Tobey <jtobey@channel1.com>*
17543 * Make links whenever Configure is run, unless we are on Windoze.
17547 * Permit extensions to be added to CRLs using crl_section in openssl.cnf.
17548 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
17553 * Add a useful kludge to allow package maintainers to specify compiler and
17554 other platforms details on the command line without having to patch the
17555 Configure script every time: One now can use
17556 `perl Configure <id>:<details>`,
17557 i.e. platform ids are allowed to have details appended
17558 to them (separated by colons). This is treated as there would be a static
17559 pre-configured entry in Configure's %table under key `<id>` with value
17560 `<details>` and `perl Configure <id>` is called. So, when you want to
17561 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
17562 assembler stuff you can use `perl Configure "FreeBSD-elf:pgcc:-O6:::"`
17563 now, which overrides the FreeBSD-elf entry on-the-fly.
17565 *Ralf S. Engelschall*
17567 * Disable new TLS1 ciphersuites by default: they aren't official yet.
17571 * Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
17572 on the `perl Configure ...` command line. This way one can compile
17573 OpenSSL libraries with Position Independent Code (PIC) which is needed
17574 for linking it into DSOs.
17576 *Ralf S. Engelschall*
17578 * Remarkably, export ciphers were totally broken and no-one had noticed!
17583 * Cleaned up the LICENSE document: The official contact for any license
17584 questions now is the OpenSSL core team under openssl-core@openssl.org.
17585 And add a paragraph about the dual-license situation to make sure people
17586 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
17587 to the OpenSSL toolkit.
17589 *Ralf S. Engelschall*
17591 * General source tree makefile cleanups: Made `making xxx in yyy...`
17592 display consistent in the source tree and replaced `/bin/rm` by `rm`.
17593 Additionally cleaned up the `make links` target: Remove unnecessary
17594 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
17595 to speed processing and no longer clutter the display with confusing
17596 stuff. Instead only the actually done links are displayed.
17598 *Ralf S. Engelschall*
17600 * Permit null encryption ciphersuites, used for authentication only. It used
17601 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
17602 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
17607 * Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
17608 signed attributes when verifying signatures (this would break them),
17609 the detached data encoding was wrong and public keys obtained using
17610 X509_get_pubkey() weren't freed.
17614 * Add text documentation for the BUFFER functions. Also added a work around
17615 to a Win95 console bug. This was triggered by the password read stuff: the
17616 last character typed gets carried over to the next fread(). If you were
17617 generating a new cert request using 'req' for example then the last
17618 character of the passphrase would be CR which would then enter the first
17623 * Added the new 'Includes OpenSSL Cryptography Software' button as
17624 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
17625 button and can be used by applications based on OpenSSL to show the
17626 relationship to the OpenSSL project.
17628 *Ralf S. Engelschall*
17630 * Remove confusing variables in function signatures in files
17631 ssl/ssl_lib.c and ssl/ssl.h.
17633 *Lennart Bong <lob@kulthea.stacken.kth.se>*
17635 * Don't install bss_file.c under PREFIX/include/
17637 *Lennart Bong <lob@kulthea.stacken.kth.se>*
17639 * Get the Win32 compile working again. Modify mkdef.pl so it can handle
17640 functions that return function pointers and has support for NT specific
17641 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
17642 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
17643 unsigned to signed types: this was killing the Win32 compile.
17647 * Add new certificate file to stack functions,
17648 SSL_add_dir_cert_subjects_to_stack() and
17649 SSL_add_file_cert_subjects_to_stack(). These largely supplant
17650 SSL_load_client_CA_file(), and can be used to add multiple certs easily
17651 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
17652 This means that Apache-SSL and similar packages don't have to mess around
17653 to add as many CAs as they want to the preferred list.
17657 * Experiment with doxygen documentation. Currently only partially applied to
17659 See <http://www.stack.nl/~dimitri/doxygen/index.html>, and run doxygen with
17660 openssl.doxy as the configuration file.
17664 * Get rid of remaining C++-style comments which strict C compilers hate.
17666 *Ralf S. Engelschall, pointed out by Carlos Amengual*
17668 * Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
17669 compiled in by default: it has problems with large keys.
17673 * Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
17674 DH private keys and/or callback functions which directly correspond to
17675 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
17676 is needed for applications which have to configure certificates on a
17677 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
17679 For the RSA certificate situation is makes no difference, but
17680 for the DSA certificate situation this fixes the "no shared cipher"
17681 problem where the OpenSSL cipher selection procedure failed because the
17682 temporary keys were not overtaken from the context and the API provided
17683 no way to reconfigure them.
17684 The new functions now let applications reconfigure the stuff and they
17685 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
17686 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
17687 non-public-API function ssl_cert_instantiate() is used as a helper
17688 function and also to reduce code redundancy inside ssl_rsa.c.
17690 *Ralf S. Engelschall*
17692 * Move s_server -dcert and -dkey options out of the undocumented feature
17693 area because they are useful for the DSA situation and should be
17694 recognized by the users.
17696 *Ralf S. Engelschall*
17698 * Fix the cipher decision scheme for export ciphers: the export bits are
17699 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
17700 SSL_EXP_MASK. So, the original variable has to be used instead of the
17701 already masked variable.
17703 *Richard Levitte <levitte@stacken.kth.se>*
17705 * Fix `port` variable from `int` to `unsigned int` in crypto/bio/b_sock.c
17707 *Richard Levitte <levitte@stacken.kth.se>*
17709 * Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
17710 from `int` to `unsigned int` because it is a length and initialized by
17711 EVP_DigestFinal() which expects an `unsigned int *`.
17713 *Richard Levitte <levitte@stacken.kth.se>*
17715 * Don't hard-code path to Perl interpreter on shebang line of Configure
17716 script. Instead use the usual Shell->Perl transition trick.
17718 *Ralf S. Engelschall*
17720 * Make `openssl x509 -noout -modulus`' functional also for DSA certificates
17721 (in addition to RSA certificates) to match the behaviour of `openssl dsa
17722 -noout -modulus` as it's already the case for `openssl rsa -noout
17723 -modulus`. For RSA the -modulus is the real "modulus" while for DSA
17724 currently the public key is printed (a decision which was already done by
17725 `openssl dsa -modulus` in the past) which serves a similar purpose.
17726 Additionally the NO_RSA no longer completely removes the whole -modulus
17727 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
17730 *Ralf S. Engelschall*
17732 * Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
17733 BIO. See the source (crypto/evp/bio_ok.c) for more info.
17735 *Arne Ansper <arne@ats.cyber.ee>*
17737 * Dump the old yucky req code that tried (and failed) to allow raw OIDs
17738 to be added. Now both 'req' and 'ca' can use new objects defined in the
17743 * Add cool BIO that does syslog (or event log on NT).
17745 *Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie*
17747 * Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
17748 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
17749 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
17750 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
17754 * Add preliminary config info for new extension code.
17758 * Make RSA_NO_PADDING really use no padding.
17760 *Ulf Moeller <ulf@fitug.de>*
17762 * Generate errors when private/public key check is done.
17766 * Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
17767 for some CRL extensions and new objects added.
17771 * Really fix the ASN1 IMPLICIT bug this time... Partial support for private
17772 key usage extension and fuller support for authority key id.
17776 * Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
17777 padding method for RSA, which is recommended for new applications in PKCS
17778 #1 v2.0 (RFC 2437, October 1998).
17779 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
17780 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
17781 against Bleichbacher's attack on RSA.
17782 *Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
17785 * Updates to the new SSL compression code
17787 *Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)*
17789 * Fix so that the version number in the master secret, when passed
17790 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
17791 (because the server will not accept higher), that the version number
17792 is 0x03,0x01, not 0x03,0x00
17794 *Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)*
17796 * Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
17797 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
17798 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
17802 * Support for RAW extensions where an arbitrary extension can be
17803 created by including its DER encoding. See apps/openssl.cnf for
17808 * Make sure latest Perl versions don't interpret some generated C array
17809 code as Perl array code in the crypto/err/err_genc.pl script.
17811 *Lars Weber <3weber@informatik.uni-hamburg.de>*
17813 * Modify ms/do_ms.bat to not generate assembly language makefiles since
17814 not many people have the assembler. Various Win32 compilation fixes and
17815 update to the INSTALL.W32 file with (hopefully) more accurate Win32
17816 build instructions.
17820 * Modify configure script 'Configure' to automatically create crypto/date.h
17821 file under Win32 and also build pem.h from pem.org. New script
17822 util/mkfiles.pl to create the MINFO file on environments that can't do a
17823 'make files': perl util/mkfiles.pl >MINFO should work.
17827 * Major rework of DES function declarations, in the pursuit of correctness
17828 and purity. As a result, many evil casts evaporated, and some weirdness,
17829 too. You may find this causes warnings in your code. Zapping your evil
17830 casts will probably fix them. Mostly.
17834 * Fix for a typo in asn1.h. Bug fix to object creation script
17835 obj_dat.pl. It considered a zero in an object definition to mean
17836 "end of object": none of the objects in objects.h have any zeros
17837 so it wasn't spotted.
17839 *Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>*
17841 * Add support for Triple DES Cipher Block Chaining with Output Feedback
17842 Masking (CBCM). In the absence of test vectors, the best I have been able
17843 to do is check that the decrypt undoes the encrypt, so far. Send me test
17844 vectors if you have them.
17848 * Correct calculation of key length for export ciphers (too much space was
17849 allocated for null ciphers). This has not been tested!
17853 * Modifications to the mkdef.pl for Win32 DEF file creation. The usage
17854 message is now correct (it understands "crypto" and "ssl" on its
17855 command line). There is also now an "update" option. This will update
17856 the util/ssleay.num and util/libeay.num files with any new functions.
17858 perl util/mkdef.pl crypto ssl update
17859 it will update them.
17863 * Overhauled the Perl interface:
17864 - ported BN stuff to OpenSSL's different BN library
17865 - made the perl/ source tree CVS-aware
17866 - renamed the package from SSLeay to OpenSSL (the files still contain
17867 their history because I've copied them in the repository)
17868 - removed obsolete files (the test scripts will be replaced
17869 by better Test::Harness variants in the future)
17871 *Ralf S. Engelschall*
17873 * First cut for a very conservative source tree cleanup:
17874 1. merge various obsolete readme texts into doc/ssleay.txt
17875 where we collect the old documents and readme texts.
17876 2. remove the first part of files where I'm already sure that we no
17877 longer need them because of three reasons: either they are just temporary
17878 files which were left by Eric or they are preserved original files where
17879 I've verified that the diff is also available in the CVS via "cvs diff
17880 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
17881 the crypto/md/ stuff).
17883 *Ralf S. Engelschall*
17885 * More extension code. Incomplete support for subject and issuer alt
17886 name, issuer and authority key id. Change the i2v function parameters
17887 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
17888 what that's for :-) Fix to ASN1 macro which messed up
17889 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
17893 * Preliminary support for ENUMERATED type. This is largely copied from the
17898 * Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
17900 *Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)*
17902 * Make sure `make rehash` target really finds the `openssl` program.
17904 *Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>*
17906 * Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
17907 like to hear about it if this slows down other processors.
17911 * Add CygWin32 platform information to Configure script.
17913 *Alan Batie <batie@aahz.jf.intel.com>*
17915 * Fixed ms/32all.bat script: `no_asm` -> `no-asm`
17917 *Rainer W. Gerling <gerling@mpg-gv.mpg.de>*
17919 * New program nseq to manipulate netscape certificate sequences
17923 * Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
17928 * Fixes to BN code. Previously the default was to define BN_RECURSION
17929 but the BN code had some problems that would cause failures when
17930 doing certificate verification and some other functions.
17932 *Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)*
17934 * Add ASN1 and PEM code to support netscape certificate sequences.
17938 * Add ASN1 and PEM code to support netscape certificate sequences.
17942 * Add several PKIX and private extended key usage OIDs.
17946 * Modify the 'ca' program to handle the new extension code. Modify
17947 openssl.cnf for new extension format, add comments.
17951 * More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
17952 and add a sample to openssl.cnf so req -x509 now adds appropriate
17957 * Continued X509 V3 changes. Add to other makefiles, integrate with the
17958 error code, add initial support to X509_print() and x509 application.
17962 * Takes a deep breath and start adding X509 V3 extension support code. Add
17963 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
17964 stuff is currently isolated and isn't even compiled yet.
17968 * Continuing patches for GeneralizedTime. Fix up certificate and CRL
17969 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
17970 Removed the versions check from X509 routines when loading extensions:
17971 this allows certain broken certificates that don't set the version
17972 properly to be processed.
17976 * Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
17977 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
17978 can still be regenerated with "make depend".
17982 * Spelling mistake in C version of CAST-128.
17984 *Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>*
17986 * Changes to the error generation code. The perl script err-code.pl
17987 now reads in the old error codes and retains the old numbers, only
17988 adding new ones if necessary. It also only changes the .err files if new
17989 codes are added. The makefiles have been modified to only insert errors
17990 when needed (to avoid needlessly modifying header files). This is done
17991 by only inserting errors if the .err file is newer than the auto generated
17992 C file. To rebuild all the error codes from scratch (the old behaviour)
17993 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
17994 or delete all the .err files.
17998 * CAST-128 was incorrectly implemented for short keys. The C version has
17999 been fixed, but is untested. The assembler versions are also fixed, but
18000 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
18001 to regenerate it if needed.
18002 *Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
18003 Hagino <itojun@kame.net>*
18005 * File was opened incorrectly in randfile.c.
18007 *Ulf Möller <ulf@fitug.de>*
18009 * Beginning of support for GeneralizedTime. d2i, i2d, check and print
18010 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
18011 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
18012 al: it's just almost always a UTCTime. Note this patch adds new error
18013 codes so do a "make errors" if there are problems.
18017 * Correct Linux 1 recognition in config.
18019 *Ulf Möller <ulf@fitug.de>*
18021 * Remove pointless MD5 hash when using DSA keys in ca.
18023 *Anonymous <nobody@replay.com>*
18025 * Generate an error if given an empty string as a cert directory. Also
18026 generate an error if handed NULL (previously returned 0 to indicate an
18027 error, but didn't set one).
18029 *Ben Laurie, reported by Anonymous <nobody@replay.com>*
18031 * Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
18035 * Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
18036 parameters. This was causing a warning which killed off the Win32 compile.
18040 * Remove C++ style comments from crypto/bn/bn_local.h.
18042 *Neil Costigan <neil.costigan@celocom.com>*
18044 * The function OBJ_txt2nid was broken. It was supposed to return a nid
18045 based on a text string, looking up short and long names and finally
18046 "dot" format. The "dot" format stuff didn't work. Added new function
18047 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
18048 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
18049 OID is not part of the table.
18053 * Add prototypes to X509 lookup/verify methods, fixing a bug in
18054 X509_LOOKUP_by_alias().
18058 * Sort openssl functions by name.
18062 * Get the gendsa program working (hopefully) and add it to app list. Remove
18063 encryption from sample DSA keys (in case anyone is interested the password
18068 * Make *all* `*_free` functions accept a NULL pointer.
18070 *Frans Heymans <fheymans@isaserver.be>*
18072 * If a DH key is generated in s3_srvr.c, don't blow it by trying to use
18075 *Anonymous <nobody@replay.com>*
18077 * s_server should send the CAfile as acceptable CAs, not its own cert.
18079 *Bodo Moeller <3moeller@informatik.uni-hamburg.de>*
18081 * Don't blow it for numeric -newkey arguments to apps/req.
18083 *Bodo Moeller <3moeller@informatik.uni-hamburg.de>*
18085 * Temp key "for export" tests were wrong in s3_srvr.c.
18087 *Anonymous <nobody@replay.com>*
18089 * Add prototype for temp key callback functions
18090 SSL_CTX_set_tmp_{rsa,dh}_callback().
18094 * Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
18095 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
18099 * X509_name_add_entry() freed the wrong thing after an error.
18101 *Arne Ansper <arne@ats.cyber.ee>*
18103 * rsa_eay.c would attempt to free a NULL context.
18105 *Arne Ansper <arne@ats.cyber.ee>*
18107 * BIO_s_socket() had a broken should_retry() on Windoze.
18109 *Arne Ansper <arne@ats.cyber.ee>*
18111 * BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
18113 *Arne Ansper <arne@ats.cyber.ee>*
18115 * Make sure the already existing X509_STORE->depth variable is initialized
18116 in X509_STORE_new(), but document the fact that this variable is still
18117 unused in the certificate verification process.
18119 *Ralf S. Engelschall*
18121 * Fix the various library and apps files to free up pkeys obtained from
18122 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
18126 * Fix reference counting in X509_PUBKEY_get(). This makes
18127 demos/maurice/example2.c work, amongst others, probably.
18129 *Steve Henson and Ben Laurie*
18131 * First cut of a cleanup for apps/. First the `ssleay` program is now named
18132 `openssl` and second, the shortcut symlinks for the `openssl <command>`
18133 are no longer created. This way we have a single and consistent command
18134 line interface `openssl <command>`, similar to `cvs <command>`.
18136 *Ralf S. Engelschall, Paul Sutton and Ben Laurie*
18138 * ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
18139 BIT STRING wrapper always have zero unused bits.
18143 * Add CA.pl, perl version of CA.sh, add extended key usage OID.
18147 * Make the top-level INSTALL documentation easier to understand.
18151 * Makefiles updated to exit if an error occurs in a sub-directory
18152 make (including if user presses ^C) [Paul Sutton]
18154 * Make Montgomery context stuff explicit in RSA data structure.
18158 * Fix build order of pem and err to allow for generated pem.h.
18162 * Fix renumbering bug in X509_NAME_delete_entry().
18166 * Enhanced the err-ins.pl script so it makes the error library number
18167 global and can add a library name. This is needed for external ASN1 and
18168 other error libraries.
18172 * Fixed sk_insert which never worked properly.
18176 * Fix ASN1 macros so they can handle indefinite length constructed
18177 EXPLICIT tags. Some non standard certificates use these: they can now
18182 * Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
18183 into a single doc/ssleay.txt bundle. This way the information is still
18184 preserved but no longer messes up this directory. Now it's new room for
18185 the new set of documentation files.
18187 *Ralf S. Engelschall*
18189 * SETs were incorrectly DER encoded. This was a major pain, because they
18190 shared code with SEQUENCEs, which aren't coded the same. This means that
18191 almost everything to do with SETs or SEQUENCEs has either changed name or
18192 number of arguments.
18194 *Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>*
18196 * Fix test data to work with the above.
18200 * Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
18201 was already fixed by Eric for 0.9.1 it seems.
18203 *Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>*
18205 * Autodetect FreeBSD3.
18209 * Fix various bugs in Configure. This affects the following platforms:
18213 unixware-2.0-pentium
18218 * Eliminate generated files from CVS. Reorder tests to regenerate files
18219 before they are needed.
18223 * Generate Makefile.ssl from Makefile.org (to keep CVS happy).
18227 ### Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
18229 * Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
18230 changed SSLeay to OpenSSL in version strings.
18232 *Ralf S. Engelschall*
18234 * Some fixups to the top-level documents.
18238 * Fixed the nasty bug where rsaref.h was not found under compile-time
18239 because the symlink to include/ was missing.
18241 *Ralf S. Engelschall*
18243 * Incorporated the popular no-RSA/DSA-only patches
18244 which allow to compile a RSA-free SSLeay.
18246 *Andrew Cooke / Interrader Ldt., Ralf S. Engelschall*
18248 * Fixed nasty rehash problem under `make -f Makefile.ssl links`
18249 when "ssleay" is still not found.
18251 *Ralf S. Engelschall*
18253 * Added more platforms to Configure: Cray T3E, HPUX 11,
18255 *Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>*
18257 * Updated the README file.
18259 *Ralf S. Engelschall*
18261 * Added various .cvsignore files in the CVS repository subdirs
18262 to make a "cvs update" really silent.
18264 *Ralf S. Engelschall*
18266 * Recompiled the error-definition header files and added
18267 missing symbols to the Win32 linker tables.
18269 *Ralf S. Engelschall*
18271 * Cleaned up the top-level documents;
18272 o new files: CHANGES and LICENSE
18273 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
18274 o merged COPYRIGHT into LICENSE
18275 o removed obsolete TODO file
18276 o renamed MICROSOFT to INSTALL.W32
18278 *Ralf S. Engelschall*
18280 * Removed dummy files from the 0.9.1b source tree:
18281 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
18282 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
18283 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
18284 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
18285 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
18287 *Ralf S. Engelschall*
18289 * Added various platform portability fixes.
18293 * The Genesis of the OpenSSL rpject:
18294 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
18295 Young and Tim J. Hudson created while they were working for C2Net until
18298 *The OpenSSL Project*
18300 ### Changes between 0.9.0b and 0.9.1b [not released]
18302 * Updated a few CA certificates under certs/
18306 * Changed some BIGNUM api stuff.
18310 * Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
18311 DGUX x86, Linux Alpha, etc.
18315 * New COMP library [crypto/comp/] for SSL Record Layer Compression:
18316 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
18321 * Add -strparse option to asn1pars program which parses nested
18324 *Dr Stephen Henson <shenson@bigfoot.com>*
18326 * Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
18330 * DSA fix for "ca" program.
18334 * Added "-genkey" option to "dsaparam" program.
18338 * Added RIPE MD160 (rmd160) message digest.
18342 * Added -a (all) option to "ssleay version" command.
18346 * Added PLATFORM define which is the id given to Configure.
18350 * Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
18354 * Extended the ASN.1 parser routines.
18358 * Extended BIO routines to support REUSEADDR, seek, tell, etc.
18362 * Added a BN_CTX to the BN library.
18366 * Fixed the weak key values in DES library
18370 * Changed API in EVP library for cipher aliases.
18374 * Added support for RC2/64bit cipher.
18378 * Converted the lhash library to the crypto/mem.c functions.
18382 * Added more recognized ASN.1 object ids.
18386 * Added more RSA padding checks for SSL/TLS.
18390 * Added BIO proxy/filter functionality.
18394 * Added extra_certs to SSL_CTX which can be used
18395 send extra CA certificates to the client in the CA cert chain sending
18396 process. It can be configured with SSL_CTX_add_extra_chain_cert().
18400 * Now Fortezza is denied in the authentication phase because
18401 this is key exchange mechanism is not supported by SSLeay at all.
18405 * Additional PKCS1 checks.
18409 * Support the string "TLSv1" for all TLS v1 ciphers.
18413 * Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
18414 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
18418 * Fixed a few memory leaks.
18422 * Fixed various code and comment typos.
18426 * A minor bug in ssl/s3_clnt.c where there would always be 4 0
18427 bytes sent in the client random.
18429 *Edward Bishop <ebishop@spyglass.com>*
18433 [CVE-2019-1563]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1563
18434 [CVE-2019-1559]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1559
18435 [CVE-2019-1552]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1552
18436 [CVE-2019-1551]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1551
18437 [CVE-2019-1549]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1549
18438 [CVE-2019-1547]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1547
18439 [CVE-2019-1543]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1543
18440 [CVE-2018-5407]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-5407
18441 [CVE-2018-0739]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0739
18442 [CVE-2018-0737]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0737
18443 [CVE-2018-0735]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0735
18444 [CVE-2018-0734]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0734
18445 [CVE-2018-0733]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0733
18446 [CVE-2018-0732]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0732
18447 [CVE-2017-3738]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3738
18448 [CVE-2017-3737]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3737
18449 [CVE-2017-3736]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3736
18450 [CVE-2017-3735]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3735
18451 [CVE-2017-3733]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3733
18452 [CVE-2017-3732]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3732
18453 [CVE-2017-3731]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3731
18454 [CVE-2017-3730]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3730
18455 [CVE-2016-7055]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-7055
18456 [CVE-2016-7054]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-7054
18457 [CVE-2016-7053]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-7053
18458 [CVE-2016-7052]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-7052
18459 [CVE-2016-6309]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6309
18460 [CVE-2016-6308]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6308
18461 [CVE-2016-6307]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6307
18462 [CVE-2016-6306]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6306
18463 [CVE-2016-6305]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6305
18464 [CVE-2016-6304]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6304
18465 [CVE-2016-6303]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6303
18466 [CVE-2016-6302]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6302
18467 [CVE-2016-2183]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2183
18468 [CVE-2016-2182]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2182
18469 [CVE-2016-2181]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2181
18470 [CVE-2016-2180]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2180
18471 [CVE-2016-2179]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2179
18472 [CVE-2016-2178]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2178
18473 [CVE-2016-2177]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2177
18474 [CVE-2016-2176]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2176
18475 [CVE-2016-2109]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2109
18476 [CVE-2016-2107]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2107
18477 [CVE-2016-2106]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2106
18478 [CVE-2016-2105]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2105
18479 [CVE-2016-0800]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0800
18480 [CVE-2016-0799]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0799
18481 [CVE-2016-0798]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0798
18482 [CVE-2016-0797]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0797
18483 [CVE-2016-0705]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0705
18484 [CVE-2016-0702]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0702
18485 [CVE-2016-0701]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0701
18486 [CVE-2015-3197]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-3197
18487 [CVE-2015-3196]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-3196
18488 [CVE-2015-3195]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-3195
18489 [CVE-2015-3194]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-3194
18490 [CVE-2015-3193]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-3193
18491 [CVE-2015-1793]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1793
18492 [CVE-2015-1792]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1792
18493 [CVE-2015-1791]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1791
18494 [CVE-2015-1790]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1790
18495 [CVE-2015-1789]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1789
18496 [CVE-2015-1788]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1788
18497 [CVE-2015-1787]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1787
18498 [CVE-2015-0293]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0293
18499 [CVE-2015-0291]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0291
18500 [CVE-2015-0290]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0290
18501 [CVE-2015-0289]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0289
18502 [CVE-2015-0288]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0288
18503 [CVE-2015-0287]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0287
18504 [CVE-2015-0286]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0286
18505 [CVE-2015-0285]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0285
18506 [CVE-2015-0209]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0209
18507 [CVE-2015-0208]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0208
18508 [CVE-2015-0207]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0207
18509 [CVE-2015-0206]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0206
18510 [CVE-2015-0205]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0205
18511 [CVE-2015-0204]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0204
18512 [CVE-2014-8275]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-8275
18513 [CVE-2014-5139]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-5139
18514 [CVE-2014-3572]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3572
18515 [CVE-2014-3571]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3571
18516 [CVE-2014-3570]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3570
18517 [CVE-2014-3569]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3569
18518 [CVE-2014-3568]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3568
18519 [CVE-2014-3567]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3567
18520 [CVE-2014-3566]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3566
18521 [CVE-2014-3513]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3513
18522 [CVE-2014-3512]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3512
18523 [CVE-2014-3511]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3511
18524 [CVE-2014-3510]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3510
18525 [CVE-2014-3509]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3509
18526 [CVE-2014-3508]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3508
18527 [CVE-2014-3507]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3507
18528 [CVE-2014-3506]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3506
18529 [CVE-2014-3505]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3505
18530 [CVE-2014-3470]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3470
18531 [CVE-2014-0224]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-0224
18532 [CVE-2014-0221]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-0221
18533 [CVE-2014-0195]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-0195
18534 [CVE-2014-0160]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-0160
18535 [CVE-2014-0076]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-0076
18536 [CVE-2013-6450]: https://www.openssl.org/news/vulnerabilities.html#CVE-2013-6450
18537 [CVE-2013-4353]: https://www.openssl.org/news/vulnerabilities.html#CVE-2013-4353
18538 [CVE-2013-0169]: https://www.openssl.org/news/vulnerabilities.html#CVE-2013-0169
18539 [CVE-2013-0166]: https://www.openssl.org/news/vulnerabilities.html#CVE-2013-0166
18540 [CVE-2012-2686]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-2686
18541 [CVE-2012-2333]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-2333
18542 [CVE-2012-2110]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-2110
18543 [CVE-2012-0884]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-0884
18544 [CVE-2012-0050]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-0050
18545 [CVE-2012-0027]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-0027
18546 [CVE-2011-4619]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-4619
18547 [CVE-2011-4577]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-4577
18548 [CVE-2011-4576]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-4576
18549 [CVE-2011-4109]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-4109
18550 [CVE-2011-4108]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-4108
18551 [CVE-2011-3210]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-3210
18552 [CVE-2011-3207]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-3207
18553 [CVE-2011-0014]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-0014
18554 [CVE-2010-4252]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-4252
18555 [CVE-2010-4180]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-4180
18556 [CVE-2010-3864]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-3864
18557 [CVE-2010-1633]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-1633
18558 [CVE-2010-0740]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-0740
18559 [CVE-2010-0433]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-0433
18560 [CVE-2009-4355]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-4355
18561 [CVE-2009-3555]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-3555
18562 [CVE-2009-3245]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-3245
18563 [CVE-2009-1386]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-1386
18564 [CVE-2009-1379]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-1379
18565 [CVE-2009-1378]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-1378
18566 [CVE-2009-1377]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-1377
18567 [CVE-2009-0789]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-0789
18568 [CVE-2009-0591]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-0591
18569 [CVE-2009-0590]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-0590
18570 [CVE-2008-5077]: https://www.openssl.org/news/vulnerabilities.html#CVE-2008-5077
18571 [CVE-2008-1678]: https://www.openssl.org/news/vulnerabilities.html#CVE-2008-1678
18572 [CVE-2008-1672]: https://www.openssl.org/news/vulnerabilities.html#CVE-2008-1672
18573 [CVE-2008-0891]: https://www.openssl.org/news/vulnerabilities.html#CVE-2008-0891
18574 [CVE-2007-5135]: https://www.openssl.org/news/vulnerabilities.html#CVE-2007-5135
18575 [CVE-2007-4995]: https://www.openssl.org/news/vulnerabilities.html#CVE-2007-4995
18576 [CVE-2006-4343]: https://www.openssl.org/news/vulnerabilities.html#CVE-2006-4343
18577 [CVE-2006-4339]: https://www.openssl.org/news/vulnerabilities.html#CVE-2006-4339
18578 [CVE-2006-3738]: https://www.openssl.org/news/vulnerabilities.html#CVE-2006-3738
18579 [CVE-2006-2940]: https://www.openssl.org/news/vulnerabilities.html#CVE-2006-2940
18580 [CVE-2006-2937]: https://www.openssl.org/news/vulnerabilities.html#CVE-2006-2937
18581 [CVE-2005-2969]: https://www.openssl.org/news/vulnerabilities.html#CVE-2005-2969
18582 [CVE-2004-0112]: https://www.openssl.org/news/vulnerabilities.html#CVE-2004-0112
18583 [CVE-2004-0079]: https://www.openssl.org/news/vulnerabilities.html#CVE-2004-0079
18584 [CVE-2003-0851]: https://www.openssl.org/news/vulnerabilities.html#CVE-2003-0851
18585 [CVE-2003-0545]: https://www.openssl.org/news/vulnerabilities.html#CVE-2003-0545
18586 [CVE-2003-0544]: https://www.openssl.org/news/vulnerabilities.html#CVE-2003-0544
18587 [CVE-2003-0543]: https://www.openssl.org/news/vulnerabilities.html#CVE-2003-0543
18588 [CVE-2003-0078]: https://www.openssl.org/news/vulnerabilities.html#CVE-2003-0078
18589 [CVE-2002-0659]: https://www.openssl.org/news/vulnerabilities.html#CVE-2002-0659
18590 [CVE-2002-0657]: https://www.openssl.org/news/vulnerabilities.html#CVE-2002-0657
18591 [CVE-2002-0656]: https://www.openssl.org/news/vulnerabilities.html#CVE-2002-0656
18592 [CVE-2002-0655]: https://www.openssl.org/news/vulnerabilities.html#CVE-2002-0655