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 * 'Configure' has been changed to figure out the configuration target if
27 none is given on the command line. Consequently, the 'config' script is
28 now only a mere wrapper. All documentation is changed to only mention
31 *Rich Salz and Richard Levitte*
33 * Added a library context that applications as well as other
34 libraries can use to form a separate context within which libcrypto
35 operations are performed.
37 There are two ways this can be used:
39 - Directly, by passing a library context to functions that take
40 such an argument, such as `EVP_CIPHER_fetch` and similar algorithm
42 - Indirectly, by creating a new library context and then assigning
43 it as the new default, with `OPENSSL_CTX_set0_default`.
45 All public OpenSSL functions that take an `OPENSSL_CTX` pointer,
46 apart from the functions directly related to `OPENSSL_CTX`, accept
47 NULL to indicate that the default library context should be used.
49 Library code that changes the default library context using
50 `OPENSSL_CTX_set0_default` should take care to restore it with a
51 second call before returning to the caller.
55 * Handshake now fails if Extended Master Secret extension is dropped
60 * Dropped interactive mode from the 'openssl' program. From now on,
61 the `openssl` command without arguments is equivalent to `openssl
66 * Renamed EVP_PKEY_cmp() to EVP_PKEY_eq() and
67 EVP_PKEY_cmp_parameters() to EVP_PKEY_parameters_eq().
68 While the old function names have been retained for backward compatibility
69 they should not be used in new developments
70 because their return values are confusing: Unlike other `_cmp()` functions
71 they do not return 0 in case their arguments are equal.
75 * Deprecated EC_METHOD_get_field_type(). Applications should switch to
76 EC_GROUP_get_field_type().
80 * Deprecated EC_GFp_simple_method(), EC_GFp_mont_method(),
81 EC_GF2m_simple_method(), EC_GFp_nist_method(), EC_GFp_nistp224_method()
82 EC_GFp_nistp256_method(), and EC_GFp_nistp521_method().
83 Applications should rely on the library automatically assigning a suitable
84 EC_METHOD internally upon EC_GROUP construction.
88 * Deprecated EC_GROUP_new(), EC_GROUP_method_of(), and EC_POINT_method_of().
89 EC_METHOD is now an internal-only concept and a suitable EC_METHOD is
90 assigned internally without application intervention.
91 Users of EC_GROUP_new() should switch to a different suitable constructor.
95 * Add CAdES-BES signature verification support, mostly derived
96 from ESSCertIDv2 TS (RFC 5816) contribution by Marek Klein.
98 *Filipe Raimundo da Silva*
100 * Add CAdES-BES signature scheme and attributes support (RFC 5126) to CMS API.
104 * Deprecated EC_POINT_make_affine() and EC_POINTs_make_affine(). These
105 functions are not widely used and now OpenSSL automatically perform this
106 conversion when needed.
110 * Deprecated EC_GROUP_precompute_mult(), EC_GROUP_have_precompute_mult(), and
111 EC_KEY_precompute_mult(). These functions are not widely used and
112 applications should instead switch to named curves which OpenSSL has
113 hardcoded lookup tables for.
117 * Deprecated EC_POINTs_mul(). This function is not widely used and applications
118 should instead use the L<EC_POINT_mul(3)> function.
122 * Removed FIPS_mode() and FIPS_mode_set(). These functions are legacy API's
123 that are not applicable to the new provider model. Applications should
124 instead use EVP_default_properties_is_fips_enabled() and
125 EVP_default_properties_enable_fips().
129 * The SSL option SSL_OP_IGNORE_UNEXPECTED_EOF is introduced. If that option
130 is set, an unexpected EOF is ignored, it pretends a close notify was received
131 instead and so the returned error becomes SSL_ERROR_ZERO_RETURN.
135 * Deprecated EC_POINT_set_Jprojective_coordinates_GFp() and
136 EC_POINT_get_Jprojective_coordinates_GFp(). These functions are not widely
137 used and applications should instead use the
138 L<EC_POINT_set_affine_coordinates(3)> and
139 L<EC_POINT_get_affine_coordinates(3)> functions.
143 * Added OSSL_PARAM_BLD to the public interface. This allows OSSL_PARAM
144 arrays to be more easily constructed via a series of utility functions.
145 Create a parameter builder using OSSL_PARAM_BLD_new(), add parameters using
146 the various push functions and finally convert to a passable OSSL_PARAM
147 array using OSSL_PARAM_BLD_to_param().
151 * The security strength of SHA1 and MD5 based signatures in TLS has been
152 reduced. This results in SSL 3, TLS 1.0, TLS 1.1 and DTLS 1.0 no longer
153 working at the default security level of 1 and instead requires security
154 level 0. The security level can be changed either using the cipher string
155 with @SECLEVEL, or calling SSL_CTX_set_security_level().
159 * EVP_PKEY_get0_RSA(), EVP_PKEY_get0_DSA(), EVP_PKEY_get0_DH(), and
160 EVP_PKEY_get0_EC_KEY() can now handle EVP_PKEYs with provider side
161 internal keys, if they correspond to one of those built in types.
165 * Added EVP_PKEY_set_type_by_keymgmt(), to initialise an EVP_PKEY to
166 contain a provider side internal key.
170 * ASN1_verify(), ASN1_digest() and ASN1_sign() have been deprecated.
171 They are old functions that we don't use, and that you could disable with
172 the macro NO_ASN1_OLD. This goes all the way back to OpenSSL 0.9.7.
176 * Project text documents not yet having a proper file name extension
177 (HACKING, LICENSE, NOTES*, README*, VERSION) have been renamed to *.md
178 as far as reasonable, else to *.txt, for better use with file managers.
182 * The main project documents (README, NEWS, CHANGES, INSTALL, SUPPORT)
183 have been converted to Markdown with the goal to produce documents
184 which not only look pretty when viewed online in the browser, but
185 remain well readable inside a plain text editor.
187 To achieve this goal, a 'minimalistic' Markdown style has been applied
188 which avoids formatting elements that interfere too much with the
189 reading flow in the text file. For example, it
191 * avoids [ATX headings][] and uses [setext headings][] instead
192 (which works for `<h1>` and `<h2>` headings only).
193 * avoids [inline links][] and uses [reference links][] instead.
194 * avoids [fenced code blocks][] and uses [indented code blocks][] instead.
196 [ATX headings]: https://github.github.com/gfm/#atx-headings
197 [setext headings]: https://github.github.com/gfm/#setext-headings
198 [inline links]: https://github.github.com/gfm/#inline-link
199 [reference links]: https://github.github.com/gfm/#reference-link
200 [fenced code blocks]: https://github.github.com/gfm/#fenced-code-blocks
201 [indented code blocks]: https://github.github.com/gfm/#indented-code-blocks
203 *Matthias St. Pierre*
205 * The test suite is changed to preserve results of each test recipe.
206 A new directory test-runs/ with subdirectories named like the
207 test recipes are created in the build tree for this purpose.
211 * Added an implementation of CMP and CRMF (RFC 4210, RFC 4211 RFC 6712).
212 This adds crypto/cmp/, crpyto/crmf/, apps/cmp.c, and test/cmp_*.
213 See L<openssl-cmp(1)> and L<OSSL_CMP_exec_IR_ses(3)> as starting points.
215 *David von Oheimb, Martin Peylo*
217 * Generalized the HTTP client code from crypto/ocsp/ into crpyto/http/.
218 The legacy OCSP-focused and only partly documented API is retained.
219 See L<OSSL_CMP_MSG_http_perform(3)> etc. for details.
223 * BIO_do_connect and BIO_do_handshake have been extended:
224 If domain name resolution yields multiple IP addresses all of them are tried
225 after connect() failures.
229 * All of the low level RSA functions have been deprecated including:
231 RSA_new_method, RSA_size, RSA_security_bits, RSA_get0_pss_params,
232 RSA_get_version, RSA_get0_engine, RSA_generate_key_ex,
233 RSA_generate_multi_prime_key, RSA_X931_derive_ex, RSA_X931_generate_key_ex,
234 RSA_check_key, RSA_check_key_ex, RSA_public_encrypt, RSA_private_encrypt,
235 RSA_public_decrypt, RSA_private_decrypt, RSA_set_default_method,
236 RSA_get_default_method, RSA_null_method, RSA_get_method, RSA_set_method,
237 RSA_PKCS1_OpenSSL, RSA_print_fp, RSA_print, RSA_sign, RSA_verify,
238 RSA_sign_ASN1_OCTET_STRING, RSA_verify_ASN1_OCTET_STRING, RSA_blinding_on,
239 RSA_blinding_off, RSA_setup_blinding, RSA_padding_add_PKCS1_type_1,
240 RSA_padding_check_PKCS1_type_1, RSA_padding_add_PKCS1_type_2,
241 RSA_padding_check_PKCS1_type_2, PKCS1_MGF1, RSA_padding_add_PKCS1_OAEP,
242 RSA_padding_check_PKCS1_OAEP, RSA_padding_add_PKCS1_OAEP_mgf1,
243 RSA_padding_check_PKCS1_OAEP_mgf1, RSA_padding_add_SSLv23,
244 RSA_padding_check_SSLv23, RSA_padding_add_none, RSA_padding_check_none,
245 RSA_padding_add_X931, RSA_padding_check_X931, RSA_X931_hash_id,
246 RSA_verify_PKCS1_PSS, RSA_padding_add_PKCS1_PSS, RSA_verify_PKCS1_PSS_mgf1,
247 RSA_padding_add_PKCS1_PSS_mgf1, RSA_set_ex_data, RSA_get_ex_data,
248 RSA_meth_new, RSA_meth_free, RSA_meth_dup, RSA_meth_get0_name,
249 RSA_meth_set1_name, RSA_meth_get_flags, RSA_meth_set_flags,
250 RSA_meth_get0_app_data, RSA_meth_set0_app_data, RSA_meth_get_pub_enc,
251 RSA_meth_set_pub_enc, RSA_meth_get_pub_dec, RSA_meth_set_pub_dec,
252 RSA_meth_get_priv_enc, RSA_meth_set_priv_enc, RSA_meth_get_priv_dec,
253 RSA_meth_set_priv_dec, RSA_meth_get_mod_exp, RSA_meth_set_mod_exp,
254 RSA_meth_get_bn_mod_exp, RSA_meth_set_bn_mod_exp, RSA_meth_get_init,
255 RSA_meth_set_init, RSA_meth_get_finish, RSA_meth_set_finish,
256 RSA_meth_get_sign, RSA_meth_set_sign, RSA_meth_get_verify,
257 RSA_meth_set_verify, RSA_meth_get_keygen, RSA_meth_set_keygen,
258 RSA_meth_get_multi_prime_keygen and RSA_meth_set_multi_prime_keygen.
260 Use of these low level functions has been informally discouraged for a long
261 time. Instead applications should use L<EVP_PKEY_encrypt_init(3)>,
262 L<EVP_PKEY_encrypt(3)>, L<EVP_PKEY_decrypt_init(3)> and
263 L<EVP_PKEY_decrypt(3)>.
267 * X509 certificates signed using SHA1 are no longer allowed at security
269 In TLS/SSL the default security level is 1. It can be set either
270 using the cipher string with @SECLEVEL, or calling
271 SSL_CTX_set_security_level(). If the leaf certificate is signed with SHA-1,
272 a call to SSL_CTX_use_certificate() will fail if the security level is not
274 Outside TLS/SSL, the default security level is -1 (effectively 0). It can
275 be set using X509_VERIFY_PARAM_set_auth_level() or using the -auth_level
280 * The command line utilities dhparam, dsa, gendsa and dsaparam have been
281 modified to use PKEY APIs. These commands are now in maintenance mode
282 and no new features will be added to them.
286 * The command line utility rsautl has been deprecated.
287 Instead use the pkeyutl program.
291 * The command line utilities genrsa and rsa have been modified to use PKEY
292 APIs These commands are now in maintenance mode and no new features will
297 * All of the low level DH functions have been deprecated including:
299 DH_OpenSSL, DH_set_default_method, DH_get_default_method, DH_set_method,
300 DH_new_method, DH_size, DH_security_bits, DH_get_ex_new_index,
301 DH_set_ex_data, DH_get_ex_data, DH_generate_parameters_ex,
302 DH_check_params_ex, DH_check_ex, DH_check_pub_key_ex,
303 DH_check, DH_check_pub_key, DH_generate_key, DH_compute_key,
304 DH_compute_key_padded, DHparams_print_fp, DHparams_print, DH_get_nid,
305 DH_KDF_X9_42, DH_get0_engine, DH_meth_new, DH_meth_free, DH_meth_dup,
306 DH_meth_get0_name, DH_meth_set1_name, DH_meth_get_flags, DH_meth_set_flags,
307 DH_meth_get0_app_data, DH_meth_set0_app_data, DH_meth_get_generate_key,
308 DH_meth_set_generate_key, DH_meth_get_compute_key, DH_meth_set_compute_key,
309 DH_meth_get_bn_mod_exp, DH_meth_set_bn_mod_exp, DH_meth_get_init,
310 DH_meth_set_init, DH_meth_get_finish, DH_meth_set_finish,
311 DH_meth_get_generate_params and DH_meth_set_generate_params.
313 Use of these low level functions has been informally discouraged for a long
314 time. Instead applications should use L<EVP_PKEY_derive_init(3)>
315 and L<EVP_PKEY_derive(3)>.
319 * All of the low level DSA functions have been deprecated including:
321 DSA_do_sign, DSA_do_verify, DSA_OpenSSL, DSA_set_default_method,
322 DSA_get_default_method, DSA_set_method, DSA_get_method,
323 DSA_new_method, DSA_size, DSA_security_bits, DSA_sign_setup, DSA_sign,
324 DSA_verify, DSA_get_ex_new_index, DSA_set_ex_data, DSA_get_ex_data,
325 DSA_generate_parameters_ex, DSA_generate_key, DSA_meth_new, DSA_get0_engine,
326 DSA_meth_free, DSA_meth_dup, DSA_meth_get0_name, DSA_meth_set1_name,
327 DSA_meth_get_flags, DSA_meth_set_flags, DSA_meth_get0_app_data,
328 DSA_meth_set0_app_data, DSA_meth_get_sign, DSA_meth_set_sign,
329 DSA_meth_get_sign_setup, DSA_meth_set_sign_setup, DSA_meth_get_verify,
330 DSA_meth_set_verify, DSA_meth_get_mod_exp, DSA_meth_set_mod_exp,
331 DSA_meth_get_bn_mod_exp, DSA_meth_set_bn_mod_exp, DSA_meth_get_init,
332 DSA_meth_set_init, DSA_meth_get_finish, DSA_meth_set_finish,
333 DSA_meth_get_paramgen, DSA_meth_set_paramgen, DSA_meth_get_keygen and
336 Use of these low level functions has been informally discouraged for a long
337 time. Instead applications should use L<EVP_DigestSignInit_ex(3)>,
338 L<EVP_DigestSignUpdate(3)> and L<EVP_DigestSignFinal(3)>.
342 * Reworked the treatment of EC EVP_PKEYs with the SM2 curve to
343 automatically become EVP_PKEY_SM2 rather than EVP_PKEY_EC.
344 This means that applications don't have to look at the curve NID and
345 `EVP_PKEY_set_alias_type(pkey, EVP_PKEY_SM2)` to get SM2 computations.
346 However, they still can, that EVP_PKEY_set_alias_type() call acts as
347 a no-op when the EVP_PKEY is already of the given type.
349 Parameter and key generation is also reworked to make it possible
350 to generate EVP_PKEY_SM2 parameters and keys without having to go
351 through EVP_PKEY_EC generation and then change the EVP_PKEY type.
352 However, code that does the latter will still work as before.
356 * Deprecated low level ECDH and ECDSA functions. These include:
358 ECDH_compute_key, ECDSA_do_sign, ECDSA_do_sign_ex, ECDSA_do_verify,
359 ECDSA_sign_setup, ECDSA_sign, ECDSA_sign_ex, ECDSA_verify and
362 Use of these low level functions has been informally discouraged for a long
363 time. Instead applications should use the EVP_PKEY_derive(3),
364 EVP_DigestSign(3) and EVP_DigestVerify(3) functions.
368 * Deprecated the EC_KEY_METHOD functions. These include:
370 EC_KEY_METHOD_new, EC_KEY_METHOD_free, EC_KEY_METHOD_set_init,
371 EC_KEY_METHOD_set_keygen, EC_KEY_METHOD_set_compute_key,
372 EC_KEY_METHOD_set_sign, EC_KEY_METHOD_set_verify,
373 EC_KEY_METHOD_get_init, EC_KEY_METHOD_get_keygen,
374 EC_KEY_METHOD_get_compute_key, EC_KEY_METHOD_get_sign and
375 EC_KEY_METHOD_get_verify.
377 Instead applications and extension writers should use the OSSL_PROVIDER APIs.
381 * Deprecated EVP_PKEY_decrypt_old(), please use EVP_PKEY_decrypt_init()
382 and EVP_PKEY_decrypt() instead.
383 Deprecated EVP_PKEY_encrypt_old(), please use EVP_PKEY_encrypt_init()
384 and EVP_PKEY_encrypt() instead.
388 * Enhanced the documentation of EVP_PKEY_size(), EVP_PKEY_bits()
389 and EVP_PKEY_security_bits(). Especially EVP_PKEY_size() needed
390 a new formulation to include all the things it can be used for,
391 as well as words of caution.
395 * The SSL_CTX_set_tlsext_ticket_key_cb(3) function has been deprecated.
396 Instead used the new SSL_CTX_set_tlsext_ticket_key_evp_cb(3) function.
400 * All of the low level HMAC functions have been deprecated including:
402 HMAC, HMAC_size, HMAC_CTX_new, HMAC_CTX_reset, HMAC_CTX_free,
403 HMAC_Init_ex, HMAC_Update, HMAC_Final, HMAC_CTX_copy, HMAC_CTX_set_flags
406 Use of these low level functions has been informally discouraged for a long
407 time. Instead applications should use L<EVP_MAC_new_ctx(3)>,
408 L<EVP_MAC_free_ctx(3)>, L<EVP_MAC_init(3)>, L<EVP_MAC_update(3)>
409 and L<EVP_MAC_final(3)>.
413 * Over two thousand fixes were made to the documentation, including:
414 - Common options (such as -rand/-writerand, TLS version control, etc)
415 were refactored and point to newly-enhanced descriptions in openssl.pod.
416 - Added style conformance for all options (with help from Richard Levitte),
417 documented all reported missing options, added a CI build to check
418 that all options are documented and that no unimplemented options
420 - Documented some internals, such as all use of environment variables.
421 - Addressed all internal broken L<> references.
425 * All of the low level CMAC functions have been deprecated including:
427 CMAC_CTX_new, CMAC_CTX_cleanup, CMAC_CTX_free, CMAC_CTX_get0_cipher_ctx,
428 CMAC_CTX_copy, CMAC_Init, CMAC_Update, CMAC_Final and CMAC_resume.
430 Use of these low level functions has been informally discouraged for a long
431 time. Instead applications should use L<EVP_MAC_new_ctx(3)>,
432 L<EVP_MAC_free_ctx(3)>, L<EVP_MAC_init(3)>, L<EVP_MAC_update(3)>
433 and L<EVP_MAC_final(3)>.
437 * All of the low level MD2, MD4, MD5, MDC2, RIPEMD160, SHA1, SHA224, SHA256,
438 SHA384, SHA512 and Whirlpool digest functions have been deprecated.
441 MD2, MD2_options, MD2_Init, MD2_Update, MD2_Final, MD4, MD4_Init,
442 MD4_Update, MD4_Final, MD4_Transform, MD5, MD5_Init, MD5_Update,
443 MD5_Final, MD5_Transform, MDC2, MDC2_Init, MDC2_Update, MDC2_Final,
444 RIPEMD160, RIPEMD160_Init, RIPEMD160_Update, RIPEMD160_Final,
445 RIPEMD160_Transform, SHA1_Init, SHA1_Update, SHA1_Final, SHA1_Transform,
446 SHA224_Init, SHA224_Update, SHA224_Final, SHA224_Transform, SHA256_Init,
447 SHA256_Update, SHA256_Final, SHA256_Transform, SHA384, SHA384_Init,
448 SHA384_Update, SHA384_Final, SHA512, SHA512_Init, SHA512_Update,
449 SHA512_Final, SHA512_Transform, WHIRLPOOL, WHIRLPOOL_Init,
450 WHIRLPOOL_Update, WHIRLPOOL_BitUpdate and WHIRLPOOL_Final.
452 Use of these low level functions has been informally discouraged
453 for a long time. Applications should use the EVP_DigestInit_ex(3),
454 EVP_DigestUpdate(3) and EVP_DigestFinal_ex(3) functions instead.
458 * Corrected the documentation of the return values from the `EVP_DigestSign*`
459 set of functions. The documentation mentioned negative values for some
460 errors, but this was never the case, so the mention of negative values
463 Code that followed the documentation and thereby check with something
464 like `EVP_DigestSignInit(...) <= 0` will continue to work undisturbed.
468 * All of the low level cipher functions have been deprecated including:
470 AES_options, AES_set_encrypt_key, AES_set_decrypt_key, AES_encrypt,
471 AES_decrypt, AES_ecb_encrypt, AES_cbc_encrypt, AES_cfb128_encrypt,
472 AES_cfb1_encrypt, AES_cfb8_encrypt, AES_ofb128_encrypt,
473 AES_wrap_key, AES_unwrap_key, BF_set_key, BF_encrypt, BF_decrypt,
474 BF_ecb_encrypt, BF_cbc_encrypt, BF_cfb64_encrypt, BF_ofb64_encrypt,
475 BF_options, Camellia_set_key, Camellia_encrypt, Camellia_decrypt,
476 Camellia_ecb_encrypt, Camellia_cbc_encrypt, Camellia_cfb128_encrypt,
477 Camellia_cfb1_encrypt, Camellia_cfb8_encrypt, Camellia_ofb128_encrypt,
478 Camellia_ctr128_encrypt, CAST_set_key, CAST_encrypt, CAST_decrypt,
479 CAST_ecb_encrypt, CAST_cbc_encrypt, CAST_cfb64_encrypt,
480 CAST_ofb64_encrypt, DES_options, DES_encrypt1, DES_encrypt2,
481 DES_encrypt3, DES_decrypt3, DES_cbc_encrypt, DES_ncbc_encrypt,
482 DES_pcbc_encrypt, DES_xcbc_encrypt, DES_cfb_encrypt, DES_cfb64_encrypt,
483 DES_ecb_encrypt, DES_ofb_encrypt, DES_ofb64_encrypt, DES_random_key,
484 DES_set_odd_parity, DES_check_key_parity, DES_is_weak_key, DES_set_key,
485 DES_key_sched, DES_set_key_checked, DES_set_key_unchecked,
486 DES_string_to_key, DES_string_to_2keys, DES_fixup_key_parity,
487 DES_ecb2_encrypt, DES_ede2_cbc_encrypt, DES_ede2_cfb64_encrypt,
488 DES_ede2_ofb64_encrypt, DES_ecb3_encrypt, DES_ede3_cbc_encrypt,
489 DES_ede3_cfb64_encrypt, DES_ede3_cfb_encrypt, DES_ede3_ofb64_encrypt,
490 DES_cbc_cksum, DES_quad_cksum, IDEA_encrypt, IDEA_options,
491 IDEA_ecb_encrypt, IDEA_set_encrypt_key, IDEA_set_decrypt_key,
492 IDEA_cbc_encrypt, IDEA_cfb64_encrypt, IDEA_ofb64_encrypt, RC2_set_key,
493 RC2_encrypt, RC2_decrypt, RC2_ecb_encrypt, RC2_cbc_encrypt,
494 RC2_cfb64_encrypt, RC2_ofb64_encrypt, RC4, RC4_options, RC4_set_key,
495 RC5_32_set_key, RC5_32_encrypt, RC5_32_decrypt, RC5_32_ecb_encrypt,
496 RC5_32_cbc_encrypt, RC5_32_cfb64_encrypt, RC5_32_ofb64_encrypt,
497 SEED_set_key, SEED_encrypt, SEED_decrypt, SEED_ecb_encrypt,
498 SEED_cbc_encrypt, SEED_cfb128_encrypt and SEED_ofb128_encrypt.
500 Use of these low level functions has been informally discouraged for
501 a long time. Applications should use the high level EVP APIs, e.g.
502 EVP_EncryptInit_ex, EVP_EncryptUpdate, EVP_EncryptFinal_ex, and the
503 equivalently named decrypt functions instead.
505 *Matt Caswell and Paul Dale*
507 * Removed include/openssl/opensslconf.h.in and replaced it with
508 include/openssl/configuration.h.in, which differs in not including
509 <openssl/macros.h>. A short header include/openssl/opensslconf.h
510 was added to include both.
512 This allows internal hacks where one might need to modify the set
513 of configured macros, for example this if deprecated symbols are
514 still supposed to be available internally:
516 #include <openssl/configuration.h>
518 #undef OPENSSL_NO_DEPRECATED
519 #define OPENSSL_SUPPRESS_DEPRECATED
521 #include <openssl/macros.h>
523 This should not be used by applications that use the exported
524 symbols, as that will lead to linking errors.
528 * Fixed an overflow bug in the x64_64 Montgomery squaring procedure
529 used in exponentiation with 512-bit moduli. No EC algorithms are
530 affected. Analysis suggests that attacks against 2-prime RSA1024,
531 3-prime RSA1536, and DSA1024 as a result of this defect would be very
532 difficult to perform and are not believed likely. Attacks against DH512
533 are considered just feasible. However, for an attack the target would
534 have to re-use the DH512 private key, which is not recommended anyway.
535 Also applications directly using the low level API BN_mod_exp may be
536 affected if they use BN_FLG_CONSTTIME.
541 * Most memory-debug features have been deprecated, and the functionality
542 replaced with no-ops.
546 * Added documentation for the STACK API. OpenSSL only defines the STACK
547 functions where they are used.
551 * Introduced a new method type and API, OSSL_SERIALIZER, to
552 represent generic serializers. An implementation is expected to
553 be able to serialize an object associated with a given name (such
554 as an algorithm name for an asymmetric key) into forms given by
555 implementation properties.
557 Serializers are primarily used from inside libcrypto, through
558 calls to functions like EVP_PKEY_print_private(),
559 PEM_write_bio_PrivateKey() and similar.
561 Serializers are specified in such a way that they can be made to
562 directly handle the provider side portion of an object, if this
563 provider side part comes from the same provider as the serializer
564 itself, but can also be made to handle objects in parametrized
565 form (as an OSSL_PARAM array of data). This allows a provider to
566 offer generic serializers as a service for any other provider.
570 * Added a .pragma directive to the syntax of configuration files, to
571 allow varying behavior in a supported and predictable manner.
572 Currently added pragma:
576 This allows dollar signs to be a keyword character unless it's
577 followed by a opening brace or parenthesis. This is useful for
578 platforms where dollar signs are commonly used in names, such as
579 volume names and system directory names on VMS.
583 * Added functionality to create an EVP_PKEY from user data. This
584 is effectively the same as creating a RSA, DH or DSA object and
585 then assigning them to an EVP_PKEY, but directly using algorithm
586 agnostic EVP functions. A benefit is that this should be future
587 proof for public key algorithms to come.
591 * Change the interpretation of the '--api' configuration option to
592 mean that this is a desired API compatibility level with no
593 further meaning. The previous interpretation, that this would
594 also mean to remove all deprecated symbols up to and including
595 the given version, no requires that 'no-deprecated' is also used
596 in the configuration.
598 When building applications, the desired API compatibility level
599 can be set with the OPENSSL_API_COMPAT macro like before. For
600 API compatibility version below 3.0, the old style numerical
601 value is valid as before, such as -DOPENSSL_API_COMPAT=0x10100000L.
602 For version 3.0 and on, the value is expected to be the decimal
603 value calculated from the major and minor version like this:
605 MAJOR * 10000 + MINOR * 100
609 -DOPENSSL_API_COMPAT=30000 For 3.0
610 -DOPENSSL_API_COMPAT=30200 For 3.2
612 To hide declarations that are deprecated up to and including the
613 given API compatibility level, -DOPENSSL_NO_DEPRECATED must be
614 given when building the application as well.
618 * Added the X509_LOOKUP_METHOD called X509_LOOKUP_store, to allow
619 access to certificate and CRL stores via URIs and OSSL_STORE
622 This adds the following functions:
624 - X509_LOOKUP_store()
625 - X509_STORE_load_file()
626 - X509_STORE_load_path()
627 - X509_STORE_load_store()
628 - SSL_add_store_cert_subjects_to_stack()
629 - SSL_CTX_set_default_verify_store()
630 - SSL_CTX_load_verify_file()
631 - SSL_CTX_load_verify_dir()
632 - SSL_CTX_load_verify_store()
636 * Added a new method to gather entropy on VMS, based on SYS$GET_ENTROPY.
637 The presence of this system service is determined at run-time.
641 * Added functionality to create an EVP_PKEY context based on data
642 for methods from providers. This takes an algorithm name and a
643 property query string and simply stores them, with the intent
644 that any operation that uses this context will use those strings
645 to fetch the needed methods implicitly, thereby making the port
646 of application written for pre-3.0 OpenSSL easier.
650 * The undocumented function NCONF_WIN32() has been deprecated; for
651 conversion details see the HISTORY section of doc/man5/config.pod
655 * Introduced the new functions EVP_DigestSignInit_ex() and
656 EVP_DigestVerifyInit_ex(). The macros EVP_DigestSignUpdate() and
657 EVP_DigestVerifyUpdate() have been converted to functions. See the man
658 pages for further details.
662 * Over two thousand fixes were made to the documentation, including:
663 adding missing command flags, better style conformance, documentation
666 *Rich Salz, Richard Levitte*
668 * s390x assembly pack: add hardware-support for P-256, P-384, P-521,
669 X25519, X448, Ed25519 and Ed448.
673 * Print all values for a PKCS#12 attribute with 'openssl pkcs12', not just
678 * Deprecated the public definition of ERR_STATE as well as the function
679 ERR_get_state(). This is done in preparation of making ERR_STATE an
684 * Added ERR functionality to give callers access to the stored function
685 names that have replaced the older function code based functions.
687 New functions are ERR_get_error_func(), ERR_peek_error_func(),
688 ERR_peek_last_error_func(), ERR_get_error_data(), ERR_peek_error_data(),
689 ERR_peek_last_error_data(), ERR_get_error_all(), ERR_peek_error_all()
690 and ERR_peek_last_error_all().
692 These functions have become deprecated: ERR_get_error_line_data(),
693 ERR_peek_error_line_data(), ERR_peek_last_error_line_data() and
694 ERR_func_error_string().
698 * Extended testing to be verbose for failing tests only. The make variables
699 VERBOSE_FAILURE or VF can be used to enable this:
701 $ make VF=1 test # Unix
702 $ mms /macro=(VF=1) test ! OpenVMS
703 $ nmake VF=1 test # Windows
707 * For built-in EC curves, ensure an EC_GROUP built from the curve name is
708 used even when parsing explicit parameters, when loading a serialized key
709 or calling `EC_GROUP_new_from_ecpkparameters()`/
710 `EC_GROUP_new_from_ecparameters()`.
711 This prevents bypass of security hardening and performance gains,
712 especially for curves with specialized EC_METHODs.
713 By default, if a key encoded with explicit parameters is loaded and later
714 serialized, the output is still encoded with explicit parameters, even if
715 internally a "named" EC_GROUP is used for computation.
719 * Compute ECC cofactors if not provided during EC_GROUP construction. Before
720 this change, EC_GROUP_set_generator would accept order and/or cofactor as
721 NULL. After this change, only the cofactor parameter can be NULL. It also
722 does some minimal sanity checks on the passed order.
727 * Fixed a padding oracle in PKCS7_dataDecode and CMS_decrypt_set1_pkey.
728 An attack is simple, if the first CMS_recipientInfo is valid but the
729 second CMS_recipientInfo is chosen ciphertext. If the second
730 recipientInfo decodes to PKCS #1 v1.5 form plaintext, the correct
731 encryption key will be replaced by garbage, and the message cannot be
732 decoded, but if the RSA decryption fails, the correct encryption key is
733 used and the recipient will not notice the attack.
734 As a work around for this potential attack the length of the decrypted
735 key must be equal to the cipher default key length, in case the
736 certifiate is not given and all recipientInfo are tried out.
737 The old behaviour can be re-enabled in the CMS code by setting the
738 CMS_DEBUG_DECRYPT flag.
742 * Early start up entropy quality from the DEVRANDOM seed source has been
743 improved for older Linux systems. The RAND subsystem will wait for
744 /dev/random to be producing output before seeding from /dev/urandom.
745 The seeded state is stored for future library initialisations using
746 a system global shared memory segment. The shared memory identifier
747 can be configured by defining OPENSSL_RAND_SEED_DEVRANDOM_SHM_ID to
748 the desired value. The default identifier is 114.
752 * Revised BN_generate_prime_ex to not avoid factors 2..17863 in p-1
753 when primes for RSA keys are computed.
754 Since we previously always generated primes == 2 (mod 3) for RSA keys,
755 the 2-prime and 3-prime RSA modules were easy to distinguish, since
756 `N = p*q = 1 (mod 3)`, but `N = p*q*r = 2 (mod 3)`. Therefore fingerprinting
757 2-prime vs. 3-prime RSA keys was possible by computing N mod 3.
758 This avoids possible fingerprinting of newly generated RSA modules.
762 * Correct the extended master secret constant on EBCDIC systems. Without this
763 fix TLS connections between an EBCDIC system and a non-EBCDIC system that
764 negotiate EMS will fail. Unfortunately this also means that TLS connections
765 between EBCDIC systems with this fix, and EBCDIC systems without this
766 fix will fail if they negotiate EMS.
770 * Changed the library initialisation so that the config file is now loaded
771 by default. This was already the case for libssl. It now occurs for both
772 libcrypto and libssl. Use the OPENSSL_INIT_NO_LOAD_CONFIG option to
773 OPENSSL_init_crypto() to suppress automatic loading of a config file.
777 * Introduced new error raising macros, ERR_raise() and ERR_raise_data(),
778 where the former acts as a replacement for ERR_put_error(), and the
779 latter replaces the combination ERR_put_error()+ERR_add_error_data().
780 ERR_raise_data() adds more flexibility by taking a format string and
781 an arbitrary number of arguments following it, to be processed with
786 * Introduced a new function, OSSL_PROVIDER_available(), which can be used
787 to check if a named provider is loaded and available. When called, it
788 will also activate all fallback providers if such are still present.
792 * Enforce a minimum DH modulus size of 512 bits.
796 * Changed DH parameters to generate the order q subgroup instead of 2q.
797 Previously generated DH parameters are still accepted by DH_check
798 but DH_generate_key works around that by clearing bit 0 of the
799 private key for those. This avoids leaking bit 0 of the private key.
803 * Significantly reduce secure memory usage by the randomness pools.
807 * `{CRYPTO,OPENSSL}_mem_debug_{push,pop}` are now no-ops and have been
812 * A new type, EVP_KEYEXCH, has been introduced to represent key exchange
813 algorithms. An implementation of a key exchange algorithm can be obtained
814 by using the function EVP_KEYEXCH_fetch(). An EVP_KEYEXCH algorithm can be
815 used in a call to EVP_PKEY_derive_init_ex() which works in a similar way to
816 the older EVP_PKEY_derive_init() function. See the man pages for the new
817 functions for further details.
821 * The EVP_PKEY_CTX_set_dh_pad() macro has now been converted to a function.
825 * Removed the function names from error messages and deprecated the
828 * Removed NextStep support and the macro OPENSSL_UNISTD
832 * Removed DES_check_key. Also removed OPENSSL_IMPLEMENT_GLOBAL,
833 OPENSSL_GLOBAL_REF, OPENSSL_DECLARE_GLOBAL.
834 Also removed "export var as function" capability; we do not export
835 variables, only functions.
839 * RC5_32_set_key has been changed to return an int type, with 0 indicating
840 an error and 1 indicating success. In previous versions of OpenSSL this
841 was a void type. If a key was set longer than the maximum possible this
846 * Support SM2 signing and verification schemes with X509 certificate.
850 * Use SHA256 as the default digest for TS query in the ts app.
854 * Change PBKDF2 to conform to SP800-132 instead of the older PKCS5 RFC2898.
855 This checks that the salt length is at least 128 bits, the derived key
856 length is at least 112 bits, and that the iteration count is at least 1000.
857 For backwards compatibility these checks are disabled by default in the
858 default provider, but are enabled by default in the fips provider.
859 To enable or disable these checks use the control
860 EVP_KDF_CTRL_SET_PBKDF2_PKCS5_MODE.
864 * Default cipher lists/suites are now available via a function, the
865 #defines are deprecated.
869 * Add target VC-WIN32-UWP, VC-WIN64A-UWP, VC-WIN32-ARM-UWP and
870 VC-WIN64-ARM-UWP in Windows OneCore target for making building libraries
871 for Windows Store apps easier. Also, the "no-uplink" option has been added.
875 * Join the directories crypto/x509 and crypto/x509v3
879 * Change the default RSA, DSA and DH size to 2048 bit instead of 1024.
880 This changes the size when using the genpkey app when no size is given. It
881 fixes an omission in earlier changes that changed all RSA, DSA and DH
882 generation apps to use 2048 bits by default.
886 * Added command 'openssl kdf' that uses the EVP_KDF API.
890 * Added command 'openssl mac' that uses the EVP_MAC API.
894 * Added OPENSSL_info() to get diverse built-in OpenSSL data, such
895 as default directories. Also added the command 'openssl info'
896 for scripting purposes.
900 * The functions AES_ige_encrypt() and AES_bi_ige_encrypt() have been
901 deprecated. These undocumented functions were never integrated into the EVP
902 layer and implement the AES Infinite Garble Extension (IGE) mode and AES
903 Bi-directional IGE mode. These modes were never formally standardised and
904 usage of these functions is believed to be very small. In particular
905 AES_bi_ige_encrypt() has a known bug. It accepts 2 AES keys, but only one
906 is ever used. The security implications are believed to be minimal, but
907 this issue was never fixed for backwards compatibility reasons. New code
908 should not use these modes.
912 * Add prediction resistance to the DRBG reseeding process.
916 * Limit the number of blocks in a data unit for AES-XTS to 2^20 as
917 mandated by IEEE Std 1619-2018.
921 * Added newline escaping functionality to a filename when using openssl dgst.
922 This output format is to replicate the output format found in the `*sum`
923 checksum programs. This aims to preserve backward compatibility.
925 *Matt Eaton, Richard Levitte, and Paul Dale*
927 * Removed the heartbeat message in DTLS feature, as it has very
928 little usage and doesn't seem to fulfill a valuable purpose.
929 The configuration option is now deprecated.
933 * Changed the output of 'openssl {digestname} < file' to display the
934 digest name in its output.
938 * Added a new generic trace API which provides support for enabling
939 instrumentation through trace output. This feature is mainly intended
940 as an aid for developers and is disabled by default. To utilize it,
941 OpenSSL needs to be configured with the `enable-trace` option.
943 If the tracing API is enabled, the application can activate trace output
944 by registering BIOs as trace channels for a number of tracing and debugging
947 The 'openssl' application has been expanded to enable any of the types
948 available via environment variables defined by the user, and serves as
949 one possible example on how to use this functionality.
951 *Richard Levitte & Matthias St. Pierre*
953 * Added build tests for C++. These are generated files that only do one
954 thing, to include one public OpenSSL head file each. This tests that
955 the public header files can be usefully included in a C++ application.
957 This test isn't enabled by default. It can be enabled with the option
958 'enable-buildtest-c++'.
962 * Add Single Step KDF (EVP_KDF_SS) to EVP_KDF.
966 * Add KMAC to EVP_MAC.
970 * Added property based algorithm implementation selection framework to
975 * Added SCA hardening for modular field inversion in EC_GROUP through
976 a new dedicated field_inv() pointer in EC_METHOD.
977 This also addresses a leakage affecting conversions from projective
978 to affine coordinates.
980 *Billy Bob Brumley, Nicola Tuveri*
982 * Added EVP_KDF, an EVP layer KDF API, to simplify adding KDF and PRF
983 implementations. This includes an EVP_PKEY to EVP_KDF bridge for
984 those algorithms that were already supported through the EVP_PKEY API
985 (scrypt, TLS1 PRF and HKDF). The low-level KDF functions for PBKDF2
986 and scrypt are now wrappers that call EVP_KDF.
990 * Build devcrypto engine as a dynamic engine.
994 * Add keyed BLAKE2 to EVP_MAC.
998 * Fix a bug in the computation of the endpoint-pair shared secret used
999 by DTLS over SCTP. This breaks interoperability with older versions
1000 of OpenSSL like OpenSSL 1.1.0 and OpenSSL 1.0.2. There is a runtime
1001 switch SSL_MODE_DTLS_SCTP_LABEL_LENGTH_BUG (off by default) enabling
1002 interoperability with such broken implementations. However, enabling
1003 this switch breaks interoperability with correct implementations.
1005 * Fix a use after free bug in d2i_X509_PUBKEY when overwriting a
1006 re-used X509_PUBKEY object if the second PUBKEY is malformed.
1010 * Move strictness check from EVP_PKEY_asn1_new() to EVP_PKEY_asn1_add0().
1014 * Change the license to the Apache License v2.0.
1018 * Switch to a new version scheme using three numbers MAJOR.MINOR.PATCH.
1020 - Major releases (indicated by incrementing the MAJOR release number)
1021 may introduce incompatible API/ABI changes.
1022 - Minor releases (indicated by incrementing the MINOR release number)
1023 may introduce new features but retain API/ABI compatibility.
1024 - Patch releases (indicated by incrementing the PATCH number)
1025 are intended for bug fixes and other improvements of existing
1026 features only (like improving performance or adding documentation)
1027 and retain API/ABI compatibility.
1031 * Add support for RFC5297 SIV mode (siv128), including AES-SIV.
1035 * Remove the 'dist' target and add a tarball building script. The
1036 'dist' target has fallen out of use, and it shouldn't be
1037 necessary to configure just to create a source distribution.
1041 * Recreate the OS390-Unix config target. It no longer relies on a
1042 special script like it did for OpenSSL pre-1.1.0.
1046 * Instead of having the source directories listed in Configure, add
1047 a 'build.info' keyword SUBDIRS to indicate what sub-directories to
1052 * Add GMAC to EVP_MAC.
1056 * Ported the HMAC, CMAC and SipHash EVP_PKEY_METHODs to EVP_MAC.
1060 * Added EVP_MAC, an EVP layer MAC API, to simplify adding MAC
1061 implementations. This includes a generic EVP_PKEY to EVP_MAC bridge,
1062 to facilitate the continued use of MACs through raw private keys in
1063 functionality such as EVP_DigestSign* and EVP_DigestVerify*.
1067 * Deprecate ECDH_KDF_X9_62() and mark its replacement as internal. Users
1068 should use the EVP interface instead (EVP_PKEY_CTX_set_ecdh_kdf_type).
1072 * Added EVP_PKEY_ECDH_KDF_X9_63 and ecdh_KDF_X9_63() as replacements for
1073 the EVP_PKEY_ECDH_KDF_X9_62 KDF type and ECDH_KDF_X9_62(). The old names
1074 are retained for backwards compatibility.
1078 * AES-XTS mode now enforces that its two keys are different to mitigate
1079 the attacked described in "Efficient Instantiations of Tweakable
1080 Blockciphers and Refinements to Modes OCB and PMAC" by Phillip Rogaway.
1081 Details of this attack can be obtained from:
1082 <http://web.cs.ucdavis.edu/%7Erogaway/papers/offsets.pdf>
1086 * Rename the object files, i.e. give them other names than in previous
1087 versions. Their names now include the name of the final product, as
1088 well as its type mnemonic (bin, lib, shlib).
1092 * Added new option for 'openssl list', '-objects', which will display the
1093 list of built in objects, i.e. OIDs with names.
1097 * Added support for Linux Kernel TLS data-path. The Linux Kernel data-path
1098 improves application performance by removing data copies and providing
1099 applications with zero-copy system calls such as sendfile and splice.
1106 ### Changes between 1.1.1e and 1.1.1f [xx XXX xxxx]
1108 ### Changes between 1.1.1d and 1.1.1e [17 Mar 2020]
1110 * Properly detect EOF while reading in libssl. Previously if we hit an EOF
1111 while reading in libssl then we would report an error back to the
1112 application (SSL_ERROR_SYSCALL) but errno would be 0. We now add
1113 an error to the stack (which means we instead return SSL_ERROR_SSL) and
1114 therefore give a hint as to what went wrong.
1118 * Check that ed25519 and ed448 are allowed by the security level. Previously
1119 signature algorithms not using an MD were not being checked that they were
1120 allowed by the security level.
1124 * Fixed SSL_get_servername() behaviour. The behaviour of SSL_get_servername()
1125 was not quite right. The behaviour was not consistent between resumption
1126 and normal handshakes, and also not quite consistent with historical
1127 behaviour. The behaviour in various scenarios has been clarified and
1128 it has been updated to make it match historical behaviour as closely as
1133 * *[VMS only]* The header files that the VMS compilers include automatically,
1134 `__DECC_INCLUDE_PROLOGUE.H` and `__DECC_INCLUDE_EPILOGUE.H`, use pragmas
1135 that the C++ compiler doesn't understand. This is a shortcoming in the
1136 compiler, but can be worked around with `__cplusplus` guards.
1138 C++ applications that use OpenSSL libraries must be compiled using the
1139 qualifier `/NAMES=(AS_IS,SHORTENED)` to be able to use all the OpenSSL
1140 functions. Otherwise, only functions with symbols of less than 31
1141 characters can be used, as the linker will not be able to successfully
1142 resolve symbols with longer names.
1146 * Added a new method to gather entropy on VMS, based on SYS$GET_ENTROPY.
1147 The presence of this system service is determined at run-time.
1151 * Added newline escaping functionality to a filename when using openssl dgst.
1152 This output format is to replicate the output format found in the `*sum`
1153 checksum programs. This aims to preserve backward compatibility.
1155 *Matt Eaton, Richard Levitte, and Paul Dale*
1157 * Print all values for a PKCS#12 attribute with 'openssl pkcs12', not just
1162 ### Changes between 1.1.1c and 1.1.1d [10 Sep 2019]
1164 * Fixed a fork protection issue. OpenSSL 1.1.1 introduced a rewritten random
1165 number generator (RNG). This was intended to include protection in the
1166 event of a fork() system call in order to ensure that the parent and child
1167 processes did not share the same RNG state. However this protection was not
1168 being used in the default case.
1170 A partial mitigation for this issue is that the output from a high
1171 precision timer is mixed into the RNG state so the likelihood of a parent
1172 and child process sharing state is significantly reduced.
1174 If an application already calls OPENSSL_init_crypto() explicitly using
1175 OPENSSL_INIT_ATFORK then this problem does not occur at all.
1178 *Matthias St. Pierre*
1180 * For built-in EC curves, ensure an EC_GROUP built from the curve name is
1181 used even when parsing explicit parameters, when loading a serialized key
1182 or calling `EC_GROUP_new_from_ecpkparameters()`/
1183 `EC_GROUP_new_from_ecparameters()`.
1184 This prevents bypass of security hardening and performance gains,
1185 especially for curves with specialized EC_METHODs.
1186 By default, if a key encoded with explicit parameters is loaded and later
1187 serialized, the output is still encoded with explicit parameters, even if
1188 internally a "named" EC_GROUP is used for computation.
1192 * Compute ECC cofactors if not provided during EC_GROUP construction. Before
1193 this change, EC_GROUP_set_generator would accept order and/or cofactor as
1194 NULL. After this change, only the cofactor parameter can be NULL. It also
1195 does some minimal sanity checks on the passed order.
1200 * Fixed a padding oracle in PKCS7_dataDecode and CMS_decrypt_set1_pkey.
1201 An attack is simple, if the first CMS_recipientInfo is valid but the
1202 second CMS_recipientInfo is chosen ciphertext. If the second
1203 recipientInfo decodes to PKCS #1 v1.5 form plaintext, the correct
1204 encryption key will be replaced by garbage, and the message cannot be
1205 decoded, but if the RSA decryption fails, the correct encryption key is
1206 used and the recipient will not notice the attack.
1207 As a work around for this potential attack the length of the decrypted
1208 key must be equal to the cipher default key length, in case the
1209 certifiate is not given and all recipientInfo are tried out.
1210 The old behaviour can be re-enabled in the CMS code by setting the
1211 CMS_DEBUG_DECRYPT flag.
1216 * Early start up entropy quality from the DEVRANDOM seed source has been
1217 improved for older Linux systems. The RAND subsystem will wait for
1218 /dev/random to be producing output before seeding from /dev/urandom.
1219 The seeded state is stored for future library initialisations using
1220 a system global shared memory segment. The shared memory identifier
1221 can be configured by defining OPENSSL_RAND_SEED_DEVRANDOM_SHM_ID to
1222 the desired value. The default identifier is 114.
1226 * Correct the extended master secret constant on EBCDIC systems. Without this
1227 fix TLS connections between an EBCDIC system and a non-EBCDIC system that
1228 negotiate EMS will fail. Unfortunately this also means that TLS connections
1229 between EBCDIC systems with this fix, and EBCDIC systems without this
1230 fix will fail if they negotiate EMS.
1234 * Use Windows installation paths in the mingw builds
1236 Mingw isn't a POSIX environment per se, which means that Windows
1237 paths should be used for installation.
1242 * Changed DH_check to accept parameters with order q and 2q subgroups.
1243 With order 2q subgroups the bit 0 of the private key is not secret
1244 but DH_generate_key works around that by clearing bit 0 of the
1245 private key for those. This avoids leaking bit 0 of the private key.
1249 * Significantly reduce secure memory usage by the randomness pools.
1253 * Revert the DEVRANDOM_WAIT feature for Linux systems
1255 The DEVRANDOM_WAIT feature added a select() call to wait for the
1256 /dev/random device to become readable before reading from the
1257 /dev/urandom device.
1259 It turned out that this change had negative side effects on
1260 performance which were not acceptable. After some discussion it
1261 was decided to revert this feature and leave it up to the OS
1262 resp. the platform maintainer to ensure a proper initialization
1263 during early boot time.
1265 *Matthias St. Pierre*
1267 ### Changes between 1.1.1b and 1.1.1c [28 May 2019]
1269 * Add build tests for C++. These are generated files that only do one
1270 thing, to include one public OpenSSL head file each. This tests that
1271 the public header files can be usefully included in a C++ application.
1273 This test isn't enabled by default. It can be enabled with the option
1274 'enable-buildtest-c++'.
1278 * Enable SHA3 pre-hashing for ECDSA and DSA.
1282 * Change the default RSA, DSA and DH size to 2048 bit instead of 1024.
1283 This changes the size when using the genpkey app when no size is given. It
1284 fixes an omission in earlier changes that changed all RSA, DSA and DH
1285 generation apps to use 2048 bits by default.
1289 * Reorganize the manual pages to consistently have RETURN VALUES,
1290 EXAMPLES, SEE ALSO and HISTORY come in that order, and adjust
1291 util/fix-doc-nits accordingly.
1293 *Paul Yang, Joshua Lock*
1295 * Add the missing accessor EVP_PKEY_get0_engine()
1299 * Have apps like 's_client' and 's_server' output the signature scheme
1300 along with other cipher suite parameters when debugging.
1304 * Make OPENSSL_config() error agnostic again.
1308 * Do the error handling in RSA decryption constant time.
1312 * Prevent over long nonces in ChaCha20-Poly1305.
1314 ChaCha20-Poly1305 is an AEAD cipher, and requires a unique nonce input
1315 for every encryption operation. RFC 7539 specifies that the nonce value
1316 (IV) should be 96 bits (12 bytes). OpenSSL allows a variable nonce length
1317 and front pads the nonce with 0 bytes if it is less than 12
1318 bytes. However it also incorrectly allows a nonce to be set of up to 16
1319 bytes. In this case only the last 12 bytes are significant and any
1320 additional leading bytes are ignored.
1322 It is a requirement of using this cipher that nonce values are
1323 unique. Messages encrypted using a reused nonce value are susceptible to
1324 serious confidentiality and integrity attacks. If an application changes
1325 the default nonce length to be longer than 12 bytes and then makes a
1326 change to the leading bytes of the nonce expecting the new value to be a
1327 new unique nonce then such an application could inadvertently encrypt
1328 messages with a reused nonce.
1330 Additionally the ignored bytes in a long nonce are not covered by the
1331 integrity guarantee of this cipher. Any application that relies on the
1332 integrity of these ignored leading bytes of a long nonce may be further
1333 affected. Any OpenSSL internal use of this cipher, including in SSL/TLS,
1334 is safe because no such use sets such a long nonce value. However user
1335 applications that use this cipher directly and set a non-default nonce
1336 length to be longer than 12 bytes may be vulnerable.
1338 This issue was reported to OpenSSL on 16th of March 2019 by Joran Dirk
1344 * Add DEVRANDOM_WAIT feature for Linux systems
1346 On older Linux systems where the getrandom() system call is not available,
1347 OpenSSL normally uses the /dev/urandom device for seeding its CSPRNG.
1348 Contrary to getrandom(), the /dev/urandom device will not block during
1349 early boot when the kernel CSPRNG has not been seeded yet.
1351 To mitigate this known weakness, use select() to wait for /dev/random to
1352 become readable before reading from /dev/urandom.
1354 * Ensure that SM2 only uses SM3 as digest algorithm
1358 ### Changes between 1.1.1a and 1.1.1b [26 Feb 2019]
1360 * Change the info callback signals for the start and end of a post-handshake
1361 message exchange in TLSv1.3. In 1.1.1/1.1.1a we used SSL_CB_HANDSHAKE_START
1362 and SSL_CB_HANDSHAKE_DONE. Experience has shown that many applications get
1363 confused by this and assume that a TLSv1.2 renegotiation has started. This
1364 can break KeyUpdate handling. Instead we no longer signal the start and end
1365 of a post handshake message exchange (although the messages themselves are
1366 still signalled). This could break some applications that were expecting
1367 the old signals. However without this KeyUpdate is not usable for many
1372 ### Changes between 1.1.1 and 1.1.1a [20 Nov 2018]
1374 * Timing vulnerability in DSA signature generation
1376 The OpenSSL DSA signature algorithm has been shown to be vulnerable to a
1377 timing side channel attack. An attacker could use variations in the signing
1378 algorithm to recover the private key.
1380 This issue was reported to OpenSSL on 16th October 2018 by Samuel Weiser.
1385 * Timing vulnerability in ECDSA signature generation
1387 The OpenSSL ECDSA signature algorithm has been shown to be vulnerable to a
1388 timing side channel attack. An attacker could use variations in the signing
1389 algorithm to recover the private key.
1391 This issue was reported to OpenSSL on 25th October 2018 by Samuel Weiser.
1396 * Fixed the issue that RAND_add()/RAND_seed() silently discards random input
1397 if its length exceeds 4096 bytes. The limit has been raised to a buffer size
1398 of two gigabytes and the error handling improved.
1400 This issue was reported to OpenSSL by Dr. Falko Strenzke. It has been
1401 categorized as a normal bug, not a security issue, because the DRBG reseeds
1402 automatically and is fully functional even without additional randomness
1403 provided by the application.
1405 ### Changes between 1.1.0i and 1.1.1 [11 Sep 2018]
1407 * Add a new ClientHello callback. Provides a callback interface that gives
1408 the application the ability to adjust the nascent SSL object at the
1409 earliest stage of ClientHello processing, immediately after extensions have
1410 been collected but before they have been processed. In particular, this
1411 callback can adjust the supported TLS versions in response to the contents
1416 * Add SM2 base algorithm support.
1420 * s390x assembly pack: add (improved) hardware-support for the following
1421 cryptographic primitives: sha3, shake, aes-gcm, aes-ccm, aes-ctr, aes-ofb,
1422 aes-cfb/cfb8, aes-ecb.
1426 * Make EVP_PKEY_asn1_new() a bit stricter about its input. A NULL pem_str
1427 parameter is no longer accepted, as it leads to a corrupt table. NULL
1428 pem_str is reserved for alias entries only.
1432 * Use the new ec_scalar_mul_ladder scaffold to implement a specialized ladder
1433 step for prime curves. The new implementation is based on formulae from
1434 differential addition-and-doubling in homogeneous projective coordinates
1435 from Izu-Takagi "A fast parallel elliptic curve multiplication resistant
1436 against side channel attacks" and Brier-Joye "Weierstrass Elliptic Curves
1437 and Side-Channel Attacks" Eq. (8) for y-coordinate recovery, modified
1438 to work in projective coordinates.
1440 *Billy Bob Brumley, Nicola Tuveri*
1442 * Change generating and checking of primes so that the error rate of not
1443 being prime depends on the intended use based on the size of the input.
1444 For larger primes this will result in more rounds of Miller-Rabin.
1445 The maximal error rate for primes with more than 1080 bits is lowered
1448 *Kurt Roeckx, Annie Yousar*
1450 * Increase the number of Miller-Rabin rounds for DSA key generating to 64.
1454 * The 'tsget' script is renamed to 'tsget.pl', to avoid confusion when
1455 moving between systems, and to avoid confusion when a Windows build is
1456 done with mingw vs with MSVC. For POSIX installs, there's still a
1457 symlink or copy named 'tsget' to avoid that confusion as well.
1461 * Revert blinding in ECDSA sign and instead make problematic addition
1462 length-invariant. Switch even to fixed-length Montgomery multiplication.
1466 * Use the new ec_scalar_mul_ladder scaffold to implement a specialized ladder
1467 step for binary curves. The new implementation is based on formulae from
1468 differential addition-and-doubling in mixed Lopez-Dahab projective
1469 coordinates, modified to independently blind the operands.
1471 *Billy Bob Brumley, Sohaib ul Hassan, Nicola Tuveri*
1473 * Add a scaffold to optionally enhance the Montgomery ladder implementation
1474 for `ec_scalar_mul_ladder` (formerly `ec_mul_consttime`) allowing
1475 EC_METHODs to implement their own specialized "ladder step", to take
1476 advantage of more favorable coordinate systems or more efficient
1477 differential addition-and-doubling algorithms.
1479 *Billy Bob Brumley, Sohaib ul Hassan, Nicola Tuveri*
1481 * Modified the random device based seed sources to keep the relevant
1482 file descriptors open rather than reopening them on each access.
1483 This allows such sources to operate in a chroot() jail without
1484 the associated device nodes being available. This behaviour can be
1485 controlled using RAND_keep_random_devices_open().
1489 * Numerous side-channel attack mitigations have been applied. This may have
1490 performance impacts for some algorithms for the benefit of improved
1491 security. Specific changes are noted in this change log by their respective
1496 * AIX shared library support overhaul. Switch to AIX "natural" way of
1497 handling shared libraries, which means collecting shared objects of
1498 different versions and bitnesses in one common archive. This allows to
1499 mitigate conflict between 1.0 and 1.1 side-by-side installations. It
1500 doesn't affect the way 3rd party applications are linked, only how
1501 multi-version installation is managed.
1505 * Make ec_group_do_inverse_ord() more robust and available to other
1506 EC cryptosystems, so that irrespective of BN_FLG_CONSTTIME, SCA
1507 mitigations are applied to the fallback BN_mod_inverse().
1508 When using this function rather than BN_mod_inverse() directly, new
1509 EC cryptosystem implementations are then safer-by-default.
1513 * Add coordinate blinding for EC_POINT and implement projective
1514 coordinate blinding for generic prime curves as a countermeasure to
1515 chosen point SCA attacks.
1517 *Sohaib ul Hassan, Nicola Tuveri, Billy Bob Brumley*
1519 * Add blinding to ECDSA and DSA signatures to protect against side channel
1520 attacks discovered by Keegan Ryan (NCC Group).
1524 * Enforce checking in the pkeyutl command line app to ensure that the input
1525 length does not exceed the maximum supported digest length when performing
1526 a sign, verify or verifyrecover operation.
1530 * SSL_MODE_AUTO_RETRY is enabled by default. Applications that use blocking
1531 I/O in combination with something like select() or poll() will hang. This
1532 can be turned off again using SSL_CTX_clear_mode().
1533 Many applications do not properly handle non-application data records, and
1534 TLS 1.3 sends more of such records. Setting SSL_MODE_AUTO_RETRY works
1535 around the problems in those applications, but can also break some.
1536 It's recommended to read the manpages about SSL_read(), SSL_write(),
1537 SSL_get_error(), SSL_shutdown(), SSL_CTX_set_mode() and
1538 SSL_CTX_set_read_ahead() again.
1542 * When unlocking a pass phrase protected PEM file or PKCS#8 container, we
1543 now allow empty (zero character) pass phrases.
1547 * Apply blinding to binary field modular inversion and remove patent
1548 pending (OPENSSL_SUN_GF2M_DIV) BN_GF2m_mod_div implementation.
1552 * Deprecate ec2_mult.c and unify scalar multiplication code paths for
1553 binary and prime elliptic curves.
1557 * Remove ECDSA nonce padding: EC_POINT_mul is now responsible for
1558 constant time fixed point multiplication.
1562 * Revise elliptic curve scalar multiplication with timing attack
1563 defenses: ec_wNAF_mul redirects to a constant time implementation
1564 when computing fixed point and variable point multiplication (which
1565 in OpenSSL are mostly used with secret scalars in keygen, sign,
1566 ECDH derive operations).
1567 *Billy Bob Brumley, Nicola Tuveri, Cesar Pereida GarcÃa,
1570 * Updated CONTRIBUTING
1574 * Updated DRBG / RAND to request nonce and additional low entropy
1575 randomness from the system.
1577 *Matthias St. Pierre*
1579 * Updated 'openssl rehash' to use OpenSSL consistent default.
1583 * Moved the load of the ssl_conf module to libcrypto, which helps
1584 loading engines that libssl uses before libssl is initialised.
1588 * Added EVP_PKEY_sign() and EVP_PKEY_verify() for EdDSA
1592 * Fixed X509_NAME_ENTRY_set to get multi-valued RDNs right in all cases.
1594 *Ingo Schwarze, Rich Salz*
1596 * Added output of accepting IP address and port for 'openssl s_server'
1600 * Added a new API for TLSv1.3 ciphersuites:
1601 SSL_CTX_set_ciphersuites()
1602 SSL_set_ciphersuites()
1606 * Memory allocation failures consistently add an error to the error
1611 * Don't use OPENSSL_ENGINES and OPENSSL_CONF environment values
1612 in libcrypto when run as setuid/setgid.
1616 * Load any config file by default when libssl is used.
1620 * Added new public header file <openssl/rand_drbg.h> and documentation
1621 for the RAND_DRBG API. See manual page RAND_DRBG(7) for an overview.
1623 *Matthias St. Pierre*
1625 * QNX support removed (cannot find contributors to get their approval
1626 for the license change).
1630 * TLSv1.3 replay protection for early data has been implemented. See the
1631 SSL_read_early_data() man page for further details.
1635 * Separated TLSv1.3 ciphersuite configuration out from TLSv1.2 ciphersuite
1636 configuration. TLSv1.3 ciphersuites are not compatible with TLSv1.2 and
1637 below. Similarly TLSv1.2 ciphersuites are not compatible with TLSv1.3.
1638 In order to avoid issues where legacy TLSv1.2 ciphersuite configuration
1639 would otherwise inadvertently disable all TLSv1.3 ciphersuites the
1640 configuration has been separated out. See the ciphers man page or the
1641 SSL_CTX_set_ciphersuites() man page for more information.
1645 * On POSIX (BSD, Linux, ...) systems the ocsp(1) command running
1646 in responder mode now supports the new "-multi" option, which
1647 spawns the specified number of child processes to handle OCSP
1648 requests. The "-timeout" option now also limits the OCSP
1649 responder's patience to wait to receive the full client request
1650 on a newly accepted connection. Child processes are respawned
1651 as needed, and the CA index file is automatically reloaded
1652 when changed. This makes it possible to run the "ocsp" responder
1653 as a long-running service, making the OpenSSL CA somewhat more
1654 feature-complete. In this mode, most diagnostic messages logged
1655 after entering the event loop are logged via syslog(3) rather than
1660 * Added support for X448 and Ed448. Heavily based on original work by
1665 * Extend OSSL_STORE with capabilities to search and to narrow the set of
1666 objects loaded. This adds the functions OSSL_STORE_expect() and
1667 OSSL_STORE_find() as well as needed tools to construct searches and
1668 get the search data out of them.
1672 * Support for TLSv1.3 added. Note that users upgrading from an earlier
1673 version of OpenSSL should review their configuration settings to ensure
1674 that they are still appropriate for TLSv1.3. For further information see:
1675 <https://wiki.openssl.org/index.php/TLS1.3>
1679 * Grand redesign of the OpenSSL random generator
1681 The default RAND method now utilizes an AES-CTR DRBG according to
1682 NIST standard SP 800-90Ar1. The new random generator is essentially
1683 a port of the default random generator from the OpenSSL FIPS 2.0
1684 object module. It is a hybrid deterministic random bit generator
1685 using an AES-CTR bit stream and which seeds and reseeds itself
1686 automatically using trusted system entropy sources.
1688 Some of its new features are:
1689 - Support for multiple DRBG instances with seed chaining.
1690 - The default RAND method makes use of a DRBG.
1691 - There is a public and private DRBG instance.
1692 - The DRBG instances are fork-safe.
1693 - Keep all global DRBG instances on the secure heap if it is enabled.
1694 - The public and private DRBG instance are per thread for lock free
1697 *Paul Dale, Benjamin Kaduk, Kurt Roeckx, Rich Salz, Matthias St. Pierre*
1699 * Changed Configure so it only says what it does and doesn't dump
1700 so much data. Instead, ./configdata.pm should be used as a script
1701 to display all sorts of configuration data.
1705 * Added processing of "make variables" to Configure.
1709 * Added SHA512/224 and SHA512/256 algorithm support.
1713 * The last traces of Netware support, first removed in 1.1.0, have
1718 * Get rid of Makefile.shared, and in the process, make the processing
1719 of certain files (rc.obj, or the .def/.map/.opt files produced from
1720 the ordinal files) more visible and hopefully easier to trace and
1721 debug (or make silent).
1725 * Make it possible to have environment variable assignments as
1726 arguments to config / Configure.
1730 * Add multi-prime RSA (RFC 8017) support.
1734 * Add SM3 implemented according to GB/T 32905-2016
1735 * Jack Lloyd <jack.lloyd@ribose.com>,
1736 Ronald Tse <ronald.tse@ribose.com>,
1737 Erick Borsboom <erick.borsboom@ribose.com> *
1739 * Add 'Maximum Fragment Length' TLS extension negotiation and support
1740 as documented in RFC6066.
1741 Based on a patch from Tomasz Moń
1743 *Filipe Raimundo da Silva*
1745 * Add SM4 implemented according to GB/T 32907-2016.
1746 * Jack Lloyd <jack.lloyd@ribose.com>,
1747 Ronald Tse <ronald.tse@ribose.com>,
1748 Erick Borsboom <erick.borsboom@ribose.com> *
1750 * Reimplement -newreq-nodes and ERR_error_string_n; the
1751 original author does not agree with the license change.
1755 * Add ARIA AEAD TLS support.
1759 * Some macro definitions to support VS6 have been removed. Visual
1760 Studio 6 has not worked since 1.1.0
1764 * Add ERR_clear_last_mark(), to allow callers to clear the last mark
1765 without clearing the errors.
1769 * Add "atfork" functions. If building on a system that without
1770 pthreads, see doc/man3/OPENSSL_fork_prepare.pod for application
1771 requirements. The RAND facility now uses/requires this.
1779 * The UI API becomes a permanent and integral part of libcrypto, i.e.
1780 not possible to disable entirely. However, it's still possible to
1781 disable the console reading UI method, UI_OpenSSL() (use UI_null()
1784 To disable, configure with 'no-ui-console'. 'no-ui' is still
1785 possible to use as an alias. Check at compile time with the
1786 macro OPENSSL_NO_UI_CONSOLE. The macro OPENSSL_NO_UI is still
1787 possible to check and is an alias for OPENSSL_NO_UI_CONSOLE.
1791 * Add a STORE module, which implements a uniform and URI based reader of
1792 stores that can contain keys, certificates, CRLs and numerous other
1793 objects. The main API is loosely based on a few stdio functions,
1794 and includes OSSL_STORE_open, OSSL_STORE_load, OSSL_STORE_eof,
1795 OSSL_STORE_error and OSSL_STORE_close.
1796 The implementation uses backends called "loaders" to implement arbitrary
1797 URI schemes. There is one built in "loader" for the 'file' scheme.
1801 * Add devcrypto engine. This has been implemented against cryptodev-linux,
1802 then adjusted to work on FreeBSD 8.4 as well.
1803 Enable by configuring with 'enable-devcryptoeng'. This is done by default
1804 on BSD implementations, as cryptodev.h is assumed to exist on all of them.
1808 * Module names can prefixed with OSSL_ or OPENSSL_. This affects
1809 util/mkerr.pl, which is adapted to allow those prefixes, leading to
1810 error code calls like this:
1812 OSSL_FOOerr(OSSL_FOO_F_SOMETHING, OSSL_FOO_R_WHATEVER);
1814 With this change, we claim the namespaces OSSL and OPENSSL in a manner
1815 that can be encoded in C. For the foreseeable future, this will only
1818 *Richard Levitte and Tim Hudson*
1820 * Removed BSD cryptodev engine.
1824 * Add a build target 'build_all_generated', to build all generated files
1825 and only that. This can be used to prepare everything that requires
1826 things like perl for a system that lacks perl and then move everything
1827 to that system and do the rest of the build there.
1831 * In the UI interface, make it possible to duplicate the user data. This
1832 can be used by engines that need to retain the data for a longer time
1833 than just the call where this user data is passed.
1837 * Ignore the '-named_curve auto' value for compatibility of applications
1840 *Tomas Mraz <tmraz@fedoraproject.org>*
1842 * Fragmented SSL/TLS alerts are no longer accepted. An alert message is 2
1843 bytes long. In theory it is permissible in SSLv3 - TLSv1.2 to fragment such
1844 alerts across multiple records (some of which could be empty). In practice
1845 it make no sense to send an empty alert record, or to fragment one. TLSv1.3
1846 prohibits this altogether and other libraries (BoringSSL, NSS) do not
1847 support this at all. Supporting it adds significant complexity to the
1848 record layer, and its removal is unlikely to cause interoperability
1853 * Add the ASN.1 types INT32, UINT32, INT64, UINT64 and variants prefixed
1854 with Z. These are meant to replace LONG and ZLONG and to be size safe.
1855 The use of LONG and ZLONG is discouraged and scheduled for deprecation
1860 * Add the 'z' and 'j' modifiers to BIO_printf() et al formatting string,
1861 'z' is to be used for [s]size_t, and 'j' - with [u]int64_t.
1863 *Richard Levitte, Andy Polyakov*
1865 * Add EC_KEY_get0_engine(), which does for EC_KEY what RSA_get0_engine()
1870 * Have 'config' recognise 64-bit mingw and choose 'mingw64' as the target
1871 platform rather than 'mingw'.
1875 * The functions X509_STORE_add_cert and X509_STORE_add_crl return
1876 success if they are asked to add an object which already exists
1877 in the store. This change cascades to other functions which load
1878 certificates and CRLs.
1882 * x86_64 assembly pack: annotate code with DWARF CFI directives to
1883 facilitate stack unwinding even from assembly subroutines.
1887 * Remove VAX C specific definitions of OPENSSL_EXPORT, OPENSSL_EXTERN.
1888 Also remove OPENSSL_GLOBAL entirely, as it became a no-op.
1892 * Remove the VMS-specific reimplementation of gmtime from crypto/o_times.c.
1893 VMS C's RTL has a fully up to date gmtime() and gmtime_r() since V7.1,
1894 which is the minimum version we support.
1898 * Certificate time validation (X509_cmp_time) enforces stricter
1899 compliance with RFC 5280. Fractional seconds and timezone offsets
1900 are no longer allowed.
1904 * Add support for ARIA
1908 * s_client will now send the Server Name Indication (SNI) extension by
1909 default unless the new "-noservername" option is used. The server name is
1910 based on the host provided to the "-connect" option unless overridden by
1911 using "-servername".
1915 * Add support for SipHash
1919 * OpenSSL now fails if it receives an unrecognised record type in TLS1.0
1920 or TLS1.1. Previously this only happened in SSLv3 and TLS1.2. This is to
1921 prevent issues where no progress is being made and the peer continually
1922 sends unrecognised record types, using up resources processing them.
1926 * 'openssl passwd' can now produce SHA256 and SHA512 based output,
1927 using the algorithm defined in
1928 <https://www.akkadia.org/drepper/SHA-crypt.txt>
1932 * Heartbeat support has been removed; the ABI is changed for now.
1934 *Richard Levitte, Rich Salz*
1936 * Support for SSL_OP_NO_ENCRYPT_THEN_MAC in SSL_CONF_cmd.
1940 * The RSA "null" method, which was partially supported to avoid patent
1941 issues, has been replaced to always returns NULL.
1948 ### Changes between 1.1.0k and 1.1.0l [10 Sep 2019]
1950 * For built-in EC curves, ensure an EC_GROUP built from the curve name is
1951 used even when parsing explicit parameters, when loading a serialized key
1952 or calling `EC_GROUP_new_from_ecpkparameters()`/
1953 `EC_GROUP_new_from_ecparameters()`.
1954 This prevents bypass of security hardening and performance gains,
1955 especially for curves with specialized EC_METHODs.
1956 By default, if a key encoded with explicit parameters is loaded and later
1957 serialized, the output is still encoded with explicit parameters, even if
1958 internally a "named" EC_GROUP is used for computation.
1962 * Compute ECC cofactors if not provided during EC_GROUP construction. Before
1963 this change, EC_GROUP_set_generator would accept order and/or cofactor as
1964 NULL. After this change, only the cofactor parameter can be NULL. It also
1965 does some minimal sanity checks on the passed order.
1970 * Fixed a padding oracle in PKCS7_dataDecode and CMS_decrypt_set1_pkey.
1971 An attack is simple, if the first CMS_recipientInfo is valid but the
1972 second CMS_recipientInfo is chosen ciphertext. If the second
1973 recipientInfo decodes to PKCS #1 v1.5 form plaintext, the correct
1974 encryption key will be replaced by garbage, and the message cannot be
1975 decoded, but if the RSA decryption fails, the correct encryption key is
1976 used and the recipient will not notice the attack.
1977 As a work around for this potential attack the length of the decrypted
1978 key must be equal to the cipher default key length, in case the
1979 certifiate is not given and all recipientInfo are tried out.
1980 The old behaviour can be re-enabled in the CMS code by setting the
1981 CMS_DEBUG_DECRYPT flag.
1986 * Use Windows installation paths in the mingw builds
1988 Mingw isn't a POSIX environment per se, which means that Windows
1989 paths should be used for installation.
1994 ### Changes between 1.1.0j and 1.1.0k [28 May 2019]
1996 * Change the default RSA, DSA and DH size to 2048 bit instead of 1024.
1997 This changes the size when using the genpkey app when no size is given. It
1998 fixes an omission in earlier changes that changed all RSA, DSA and DH
1999 generation apps to use 2048 bits by default.
2003 * Prevent over long nonces in ChaCha20-Poly1305.
2005 ChaCha20-Poly1305 is an AEAD cipher, and requires a unique nonce input
2006 for every encryption operation. RFC 7539 specifies that the nonce value
2007 (IV) should be 96 bits (12 bytes). OpenSSL allows a variable nonce length
2008 and front pads the nonce with 0 bytes if it is less than 12
2009 bytes. However it also incorrectly allows a nonce to be set of up to 16
2010 bytes. In this case only the last 12 bytes are significant and any
2011 additional leading bytes are ignored.
2013 It is a requirement of using this cipher that nonce values are
2014 unique. Messages encrypted using a reused nonce value are susceptible to
2015 serious confidentiality and integrity attacks. If an application changes
2016 the default nonce length to be longer than 12 bytes and then makes a
2017 change to the leading bytes of the nonce expecting the new value to be a
2018 new unique nonce then such an application could inadvertently encrypt
2019 messages with a reused nonce.
2021 Additionally the ignored bytes in a long nonce are not covered by the
2022 integrity guarantee of this cipher. Any application that relies on the
2023 integrity of these ignored leading bytes of a long nonce may be further
2024 affected. Any OpenSSL internal use of this cipher, including in SSL/TLS,
2025 is safe because no such use sets such a long nonce value. However user
2026 applications that use this cipher directly and set a non-default nonce
2027 length to be longer than 12 bytes may be vulnerable.
2029 This issue was reported to OpenSSL on 16th of March 2019 by Joran Dirk
2035 * Added SCA hardening for modular field inversion in EC_GROUP through
2036 a new dedicated field_inv() pointer in EC_METHOD.
2037 This also addresses a leakage affecting conversions from projective
2038 to affine coordinates.
2040 *Billy Bob Brumley, Nicola Tuveri*
2042 * Fix a use after free bug in d2i_X509_PUBKEY when overwriting a
2043 re-used X509_PUBKEY object if the second PUBKEY is malformed.
2047 * Move strictness check from EVP_PKEY_asn1_new() to EVP_PKEY_asn1_add0().
2051 * Remove the 'dist' target and add a tarball building script. The
2052 'dist' target has fallen out of use, and it shouldn't be
2053 necessary to configure just to create a source distribution.
2057 ### Changes between 1.1.0i and 1.1.0j [20 Nov 2018]
2059 * Timing vulnerability in DSA signature generation
2061 The OpenSSL DSA signature algorithm has been shown to be vulnerable to a
2062 timing side channel attack. An attacker could use variations in the signing
2063 algorithm to recover the private key.
2065 This issue was reported to OpenSSL on 16th October 2018 by Samuel Weiser.
2070 * Timing vulnerability in ECDSA signature generation
2072 The OpenSSL ECDSA signature algorithm has been shown to be vulnerable to a
2073 timing side channel attack. An attacker could use variations in the signing
2074 algorithm to recover the private key.
2076 This issue was reported to OpenSSL on 25th October 2018 by Samuel Weiser.
2081 * Add coordinate blinding for EC_POINT and implement projective
2082 coordinate blinding for generic prime curves as a countermeasure to
2083 chosen point SCA attacks.
2085 *Sohaib ul Hassan, Nicola Tuveri, Billy Bob Brumley*
2087 ### Changes between 1.1.0h and 1.1.0i [14 Aug 2018]
2089 * Client DoS due to large DH parameter
2091 During key agreement in a TLS handshake using a DH(E) based ciphersuite a
2092 malicious server can send a very large prime value to the client. This will
2093 cause the client to spend an unreasonably long period of time generating a
2094 key for this prime resulting in a hang until the client has finished. This
2095 could be exploited in a Denial Of Service attack.
2097 This issue was reported to OpenSSL on 5th June 2018 by Guido Vranken
2102 * Cache timing vulnerability in RSA Key Generation
2104 The OpenSSL RSA Key generation algorithm has been shown to be vulnerable to
2105 a cache timing side channel attack. An attacker with sufficient access to
2106 mount cache timing attacks during the RSA key generation process could
2107 recover the private key.
2109 This issue was reported to OpenSSL on 4th April 2018 by Alejandro Cabrera
2110 Aldaya, Billy Brumley, Cesar Pereida Garcia and Luis Manuel Alvarez Tapia.
2115 * Make EVP_PKEY_asn1_new() a bit stricter about its input. A NULL pem_str
2116 parameter is no longer accepted, as it leads to a corrupt table. NULL
2117 pem_str is reserved for alias entries only.
2121 * Revert blinding in ECDSA sign and instead make problematic addition
2122 length-invariant. Switch even to fixed-length Montgomery multiplication.
2126 * Change generating and checking of primes so that the error rate of not
2127 being prime depends on the intended use based on the size of the input.
2128 For larger primes this will result in more rounds of Miller-Rabin.
2129 The maximal error rate for primes with more than 1080 bits is lowered
2132 *Kurt Roeckx, Annie Yousar*
2134 * Increase the number of Miller-Rabin rounds for DSA key generating to 64.
2138 * Add blinding to ECDSA and DSA signatures to protect against side channel
2139 attacks discovered by Keegan Ryan (NCC Group).
2143 * When unlocking a pass phrase protected PEM file or PKCS#8 container, we
2144 now allow empty (zero character) pass phrases.
2148 * Certificate time validation (X509_cmp_time) enforces stricter
2149 compliance with RFC 5280. Fractional seconds and timezone offsets
2150 are no longer allowed.
2154 * Fixed a text canonicalisation bug in CMS
2156 Where a CMS detached signature is used with text content the text goes
2157 through a canonicalisation process first prior to signing or verifying a
2158 signature. This process strips trailing space at the end of lines, converts
2159 line terminators to CRLF and removes additional trailing line terminators
2160 at the end of a file. A bug in the canonicalisation process meant that
2161 some characters, such as form-feed, were incorrectly treated as whitespace
2162 and removed. This is contrary to the specification (RFC5485). This fix
2163 could mean that detached text data signed with an earlier version of
2164 OpenSSL 1.1.0 may fail to verify using the fixed version, or text data
2165 signed with a fixed OpenSSL may fail to verify with an earlier version of
2166 OpenSSL 1.1.0. A workaround is to only verify the canonicalised text data
2167 and use the "-binary" flag (for the "cms" command line application) or set
2168 the SMIME_BINARY/PKCS7_BINARY/CMS_BINARY flags (if using CMS_verify()).
2172 ### Changes between 1.1.0g and 1.1.0h [27 Mar 2018]
2174 * Constructed ASN.1 types with a recursive definition could exceed the stack
2176 Constructed ASN.1 types with a recursive definition (such as can be found
2177 in PKCS7) could eventually exceed the stack given malicious input with
2178 excessive recursion. This could result in a Denial Of Service attack. There
2179 are no such structures used within SSL/TLS that come from untrusted sources
2180 so this is considered safe.
2182 This issue was reported to OpenSSL on 4th January 2018 by the OSS-fuzz
2188 * Incorrect CRYPTO_memcmp on HP-UX PA-RISC
2190 Because of an implementation bug the PA-RISC CRYPTO_memcmp function is
2191 effectively reduced to only comparing the least significant bit of each
2192 byte. This allows an attacker to forge messages that would be considered as
2193 authenticated in an amount of tries lower than that guaranteed by the
2194 security claims of the scheme. The module can only be compiled by the
2195 HP-UX assembler, so that only HP-UX PA-RISC targets are affected.
2197 This issue was reported to OpenSSL on 2nd March 2018 by Peter Waltenberg
2203 * Add a build target 'build_all_generated', to build all generated files
2204 and only that. This can be used to prepare everything that requires
2205 things like perl for a system that lacks perl and then move everything
2206 to that system and do the rest of the build there.
2210 * Backport SSL_OP_NO_RENGOTIATION
2212 OpenSSL 1.0.2 and below had the ability to disable renegotiation using the
2213 (undocumented) SSL3_FLAGS_NO_RENEGOTIATE_CIPHERS flag. Due to the opacity
2214 changes this is no longer possible in 1.1.0. Therefore the new
2215 SSL_OP_NO_RENEGOTIATION option from 1.1.1-dev has been backported to
2216 1.1.0 to provide equivalent functionality.
2218 Note that if an application built against 1.1.0h headers (or above) is run
2219 using an older version of 1.1.0 (prior to 1.1.0h) then the option will be
2220 accepted but nothing will happen, i.e. renegotiation will not be prevented.
2224 * Removed the OS390-Unix config target. It relied on a script that doesn't
2229 * rsaz_1024_mul_avx2 overflow bug on x86_64
2231 There is an overflow bug in the AVX2 Montgomery multiplication procedure
2232 used in exponentiation with 1024-bit moduli. No EC algorithms are affected.
2233 Analysis suggests that attacks against RSA and DSA as a result of this
2234 defect would be very difficult to perform and are not believed likely.
2235 Attacks against DH1024 are considered just feasible, because most of the
2236 work necessary to deduce information about a private key may be performed
2237 offline. The amount of resources required for such an attack would be
2238 significant. However, for an attack on TLS to be meaningful, the server
2239 would have to share the DH1024 private key among multiple clients, which is
2240 no longer an option since CVE-2016-0701.
2242 This only affects processors that support the AVX2 but not ADX extensions
2243 like Intel Haswell (4th generation).
2245 This issue was reported to OpenSSL by David Benjamin (Google). The issue
2246 was originally found via the OSS-Fuzz project.
2251 ### Changes between 1.1.0f and 1.1.0g [2 Nov 2017]
2253 * bn_sqrx8x_internal carry bug on x86_64
2255 There is a carry propagating bug in the x86_64 Montgomery squaring
2256 procedure. No EC algorithms are affected. Analysis suggests that attacks
2257 against RSA and DSA as a result of this defect would be very difficult to
2258 perform and are not believed likely. Attacks against DH are considered just
2259 feasible (although very difficult) because most of the work necessary to
2260 deduce information about a private key may be performed offline. The amount
2261 of resources required for such an attack would be very significant and
2262 likely only accessible to a limited number of attackers. An attacker would
2263 additionally need online access to an unpatched system using the target
2264 private key in a scenario with persistent DH parameters and a private
2265 key that is shared between multiple clients.
2267 This only affects processors that support the BMI1, BMI2 and ADX extensions
2268 like Intel Broadwell (5th generation) and later or AMD Ryzen.
2270 This issue was reported to OpenSSL by the OSS-Fuzz project.
2275 * Malformed X.509 IPAddressFamily could cause OOB read
2277 If an X.509 certificate has a malformed IPAddressFamily extension,
2278 OpenSSL could do a one-byte buffer overread. The most likely result
2279 would be an erroneous display of the certificate in text format.
2281 This issue was reported to OpenSSL by the OSS-Fuzz project.
2286 ### Changes between 1.1.0e and 1.1.0f [25 May 2017]
2288 * Have 'config' recognise 64-bit mingw and choose 'mingw64' as the target
2289 platform rather than 'mingw'.
2293 * Remove the VMS-specific reimplementation of gmtime from crypto/o_times.c.
2294 VMS C's RTL has a fully up to date gmtime() and gmtime_r() since V7.1,
2295 which is the minimum version we support.
2299 ### Changes between 1.1.0d and 1.1.0e [16 Feb 2017]
2301 * Encrypt-Then-Mac renegotiation crash
2303 During a renegotiation handshake if the Encrypt-Then-Mac extension is
2304 negotiated where it was not in the original handshake (or vice-versa) then
2305 this can cause OpenSSL to crash (dependant on ciphersuite). Both clients
2306 and servers are affected.
2308 This issue was reported to OpenSSL by Joe Orton (Red Hat).
2313 ### Changes between 1.1.0c and 1.1.0d [26 Jan 2017]
2315 * Truncated packet could crash via OOB read
2317 If one side of an SSL/TLS path is running on a 32-bit host and a specific
2318 cipher is being used, then a truncated packet can cause that host to
2319 perform an out-of-bounds read, usually resulting in a crash.
2321 This issue was reported to OpenSSL by Robert Święcki of Google.
2326 * Bad (EC)DHE parameters cause a client crash
2328 If a malicious server supplies bad parameters for a DHE or ECDHE key
2329 exchange then this can result in the client attempting to dereference a
2330 NULL pointer leading to a client crash. This could be exploited in a Denial
2333 This issue was reported to OpenSSL by Guido Vranken.
2338 * BN_mod_exp may produce incorrect results on x86_64
2340 There is a carry propagating bug in the x86_64 Montgomery squaring
2341 procedure. No EC algorithms are affected. Analysis suggests that attacks
2342 against RSA and DSA as a result of this defect would be very difficult to
2343 perform and are not believed likely. Attacks against DH are considered just
2344 feasible (although very difficult) because most of the work necessary to
2345 deduce information about a private key may be performed offline. The amount
2346 of resources required for such an attack would be very significant and
2347 likely only accessible to a limited number of attackers. An attacker would
2348 additionally need online access to an unpatched system using the target
2349 private key in a scenario with persistent DH parameters and a private
2350 key that is shared between multiple clients. For example this can occur by
2351 default in OpenSSL DHE based SSL/TLS ciphersuites. Note: This issue is very
2352 similar to CVE-2015-3193 but must be treated as a separate problem.
2354 This issue was reported to OpenSSL by the OSS-Fuzz project.
2359 ### Changes between 1.1.0b and 1.1.0c [10 Nov 2016]
2361 * ChaCha20/Poly1305 heap-buffer-overflow
2363 TLS connections using `*-CHACHA20-POLY1305` ciphersuites are susceptible to
2364 a DoS attack by corrupting larger payloads. This can result in an OpenSSL
2365 crash. This issue is not considered to be exploitable beyond a DoS.
2367 This issue was reported to OpenSSL by Robert Święcki (Google Security Team)
2372 * CMS Null dereference
2374 Applications parsing invalid CMS structures can crash with a NULL pointer
2375 dereference. This is caused by a bug in the handling of the ASN.1 CHOICE
2376 type in OpenSSL 1.1.0 which can result in a NULL value being passed to the
2377 structure callback if an attempt is made to free certain invalid encodings.
2378 Only CHOICE structures using a callback which do not handle NULL value are
2381 This issue was reported to OpenSSL by Tyler Nighswander of ForAllSecure.
2386 * Montgomery multiplication may produce incorrect results
2388 There is a carry propagating bug in the Broadwell-specific Montgomery
2389 multiplication procedure that handles input lengths divisible by, but
2390 longer than 256 bits. Analysis suggests that attacks against RSA, DSA
2391 and DH private keys are impossible. This is because the subroutine in
2392 question is not used in operations with the private key itself and an input
2393 of the attacker's direct choice. Otherwise the bug can manifest itself as
2394 transient authentication and key negotiation failures or reproducible
2395 erroneous outcome of public-key operations with specially crafted input.
2396 Among EC algorithms only Brainpool P-512 curves are affected and one
2397 presumably can attack ECDH key negotiation. Impact was not analyzed in
2398 detail, because pre-requisites for attack are considered unlikely. Namely
2399 multiple clients have to choose the curve in question and the server has to
2400 share the private key among them, neither of which is default behaviour.
2401 Even then only clients that chose the curve will be affected.
2403 This issue was publicly reported as transient failures and was not
2404 initially recognized as a security issue. Thanks to Richard Morgan for
2405 providing reproducible case.
2410 * Removed automatic addition of RPATH in shared libraries and executables,
2411 as this was a remainder from OpenSSL 1.0.x and isn't needed any more.
2415 ### Changes between 1.1.0a and 1.1.0b [26 Sep 2016]
2417 * Fix Use After Free for large message sizes
2419 The patch applied to address CVE-2016-6307 resulted in an issue where if a
2420 message larger than approx 16k is received then the underlying buffer to
2421 store the incoming message is reallocated and moved. Unfortunately a
2422 dangling pointer to the old location is left which results in an attempt to
2423 write to the previously freed location. This is likely to result in a
2424 crash, however it could potentially lead to execution of arbitrary code.
2426 This issue only affects OpenSSL 1.1.0a.
2428 This issue was reported to OpenSSL by Robert Święcki.
2433 ### Changes between 1.1.0 and 1.1.0a [22 Sep 2016]
2435 * OCSP Status Request extension unbounded memory growth
2437 A malicious client can send an excessively large OCSP Status Request
2438 extension. If that client continually requests renegotiation, sending a
2439 large OCSP Status Request extension each time, then there will be unbounded
2440 memory growth on the server. This will eventually lead to a Denial Of
2441 Service attack through memory exhaustion. Servers with a default
2442 configuration are vulnerable even if they do not support OCSP. Builds using
2443 the "no-ocsp" build time option are not affected.
2445 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
2450 * SSL_peek() hang on empty record
2452 OpenSSL 1.1.0 SSL/TLS will hang during a call to SSL_peek() if the peer
2453 sends an empty record. This could be exploited by a malicious peer in a
2454 Denial Of Service attack.
2456 This issue was reported to OpenSSL by Alex Gaynor.
2461 * Excessive allocation of memory in tls_get_message_header() and
2462 dtls1_preprocess_fragment()
2464 A (D)TLS message includes 3 bytes for its length in the header for the
2465 message. This would allow for messages up to 16Mb in length. Messages of
2466 this length are excessive and OpenSSL includes a check to ensure that a
2467 peer is sending reasonably sized messages in order to avoid too much memory
2468 being consumed to service a connection. A flaw in the logic of version
2469 1.1.0 means that memory for the message is allocated too early, prior to
2470 the excessive message length check. Due to way memory is allocated in
2471 OpenSSL this could mean an attacker could force up to 21Mb to be allocated
2472 to service a connection. This could lead to a Denial of Service through
2473 memory exhaustion. However, the excessive message length check still takes
2474 place, and this would cause the connection to immediately fail. Assuming
2475 that the application calls SSL_free() on the failed connection in a timely
2476 manner then the 21Mb of allocated memory will then be immediately freed
2477 again. Therefore the excessive memory allocation will be transitory in
2478 nature. This then means that there is only a security impact if:
2480 1) The application does not call SSL_free() in a timely manner in the event
2481 that the connection fails
2483 2) The application is working in a constrained environment where there is
2484 very little free memory
2486 3) The attacker initiates multiple connection attempts such that there are
2487 multiple connections in a state where memory has been allocated for the
2488 connection; SSL_free() has not yet been called; and there is insufficient
2489 memory to service the multiple requests.
2491 Except in the instance of (1) above any Denial Of Service is likely to be
2492 transitory because as soon as the connection fails the memory is
2493 subsequently freed again in the SSL_free() call. However there is an
2494 increased risk during this period of application crashes due to the lack of
2495 memory - which would then mean a more serious Denial of Service.
2497 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
2498 (CVE-2016-6307 and CVE-2016-6308)
2502 * solaris-x86-cc, i.e. 32-bit configuration with vendor compiler,
2503 had to be removed. Primary reason is that vendor assembler can't
2504 assemble our modules with -KPIC flag. As result it, assembly
2505 support, was not even available as option. But its lack means
2506 lack of side-channel resistant code, which is incompatible with
2507 security by todays standards. Fortunately gcc is readily available
2508 prepackaged option, which we firmly point at...
2512 ### Changes between 1.0.2h and 1.1.0 [25 Aug 2016]
2514 * Windows command-line tool supports UTF-8 opt-in option for arguments
2515 and console input. Setting OPENSSL_WIN32_UTF8 environment variable
2516 (to any value) allows Windows user to access PKCS#12 file generated
2517 with Windows CryptoAPI and protected with non-ASCII password, as well
2518 as files generated under UTF-8 locale on Linux also protected with
2523 * To mitigate the SWEET32 attack [CVE-2016-2183][], 3DES cipher suites
2524 have been disabled by default and removed from DEFAULT, just like RC4.
2525 See the RC4 item below to re-enable both.
2529 * The method for finding the storage location for the Windows RAND seed file
2530 has changed. First we check %RANDFILE%. If that is not set then we check
2531 the directories %HOME%, %USERPROFILE% and %SYSTEMROOT% in that order. If
2532 all else fails we fall back to C:\.
2536 * The EVP_EncryptUpdate() function has had its return type changed from void
2537 to int. A return of 0 indicates and error while a return of 1 indicates
2542 * The flags RSA_FLAG_NO_CONSTTIME, DSA_FLAG_NO_EXP_CONSTTIME and
2543 DH_FLAG_NO_EXP_CONSTTIME which previously provided the ability to switch
2544 off the constant time implementation for RSA, DSA and DH have been made
2545 no-ops and deprecated.
2549 * Windows RAND implementation was simplified to only get entropy by
2550 calling CryptGenRandom(). Various other RAND-related tickets
2553 *Joseph Wylie Yandle, Rich Salz*
2555 * The stack and lhash API's were renamed to start with `OPENSSL_SK_`
2556 and `OPENSSL_LH_`, respectively. The old names are available
2557 with API compatibility. They new names are now completely documented.
2561 * Unify TYPE_up_ref(obj) methods signature.
2562 SSL_CTX_up_ref(), SSL_up_ref(), X509_up_ref(), EVP_PKEY_up_ref(),
2563 X509_CRL_up_ref(), X509_OBJECT_up_ref_count() methods are now returning an
2564 int (instead of void) like all others TYPE_up_ref() methods.
2565 So now these methods also check the return value of CRYPTO_atomic_add(),
2566 and the validity of object reference counter.
2568 *fdasilvayy@gmail.com*
2570 * With Windows Visual Studio builds, the .pdb files are installed
2571 alongside the installed libraries and executables. For a static
2572 library installation, ossl_static.pdb is the associate compiler
2573 generated .pdb file to be used when linking programs.
2577 * Remove openssl.spec. Packaging files belong with the packagers.
2581 * Automatic Darwin/OSX configuration has had a refresh, it will now
2582 recognise x86_64 architectures automatically. You can still decide
2583 to build for a different bitness with the environment variable
2584 KERNEL_BITS (can be 32 or 64), for example:
2586 KERNEL_BITS=32 ./config
2590 * Change default algorithms in pkcs8 utility to use PKCS#5 v2.0,
2591 256 bit AES and HMAC with SHA256.
2595 * Remove support for MIPS o32 ABI on IRIX (and IRIX only).
2599 * Triple-DES ciphers have been moved from HIGH to MEDIUM.
2603 * To enable users to have their own config files and build file templates,
2604 Configure looks in the directory indicated by the environment variable
2605 OPENSSL_LOCAL_CONFIG_DIR as well as the in-source Configurations/
2606 directory. On VMS, OPENSSL_LOCAL_CONFIG_DIR is expected to be a logical
2607 name and is used as is.
2611 * The following datatypes were made opaque: X509_OBJECT, X509_STORE_CTX,
2612 X509_STORE, X509_LOOKUP, and X509_LOOKUP_METHOD. The unused type
2613 X509_CERT_FILE_CTX was removed.
2617 * "shared" builds are now the default. To create only static libraries use
2618 the "no-shared" Configure option.
2622 * Remove the no-aes, no-hmac, no-rsa, no-sha and no-md5 Configure options.
2623 All of these option have not worked for some while and are fundamental
2628 * Make various cleanup routines no-ops and mark them as deprecated. Most
2629 global cleanup functions are no longer required because they are handled
2630 via auto-deinit (see OPENSSL_init_crypto and OPENSSL_init_ssl man pages).
2631 Explicitly de-initing can cause problems (e.g. where a library that uses
2632 OpenSSL de-inits, but an application is still using it). The affected
2633 functions are CONF_modules_free(), ENGINE_cleanup(), OBJ_cleanup(),
2634 EVP_cleanup(), BIO_sock_cleanup(), CRYPTO_cleanup_all_ex_data(),
2635 RAND_cleanup(), SSL_COMP_free_compression_methods(), ERR_free_strings() and
2636 COMP_zlib_cleanup().
2640 * --strict-warnings no longer enables runtime debugging options
2641 such as REF_DEBUG. Instead, debug options are automatically
2642 enabled with '--debug' builds.
2644 *Andy Polyakov, Emilia Käsper*
2646 * Made DH and DH_METHOD opaque. The structures for managing DH objects
2647 have been moved out of the public header files. New functions for managing
2648 these have been added.
2652 * Made RSA and RSA_METHOD opaque. The structures for managing RSA
2653 objects have been moved out of the public header files. New
2654 functions for managing these have been added.
2658 * Made DSA and DSA_METHOD opaque. The structures for managing DSA objects
2659 have been moved out of the public header files. New functions for managing
2660 these have been added.
2664 * Made BIO and BIO_METHOD opaque. The structures for managing BIOs have been
2665 moved out of the public header files. New functions for managing these
2670 * Removed no-rijndael as a config option. Rijndael is an old name for AES.
2674 * Removed the mk1mf build scripts.
2678 * Headers are now wrapped, if necessary, with OPENSSL_NO_xxx, so
2679 it is always safe to #include a header now.
2683 * Removed the aged BC-32 config and all its supporting scripts
2687 * Removed support for Ultrix, Netware, and OS/2.
2691 * Add support for HKDF.
2693 *Alessandro Ghedini*
2695 * Add support for blake2b and blake2s
2699 * Added support for "pipelining". Ciphers that have the
2700 EVP_CIPH_FLAG_PIPELINE flag set have a capability to process multiple
2701 encryptions/decryptions simultaneously. There are currently no built-in
2702 ciphers with this property but the expectation is that engines will be able
2703 to offer it to significantly improve throughput. Support has been extended
2704 into libssl so that multiple records for a single connection can be
2705 processed in one go (for >=TLS 1.1).
2709 * Added the AFALG engine. This is an async capable engine which is able to
2710 offload work to the Linux kernel. In this initial version it only supports
2711 AES128-CBC. The kernel must be version 4.1.0 or greater.
2715 * OpenSSL now uses a new threading API. It is no longer necessary to
2716 set locking callbacks to use OpenSSL in a multi-threaded environment. There
2717 are two supported threading models: pthreads and windows threads. It is
2718 also possible to configure OpenSSL at compile time for "no-threads". The
2719 old threading API should no longer be used. The functions have been
2720 replaced with "no-op" compatibility macros.
2722 *Alessandro Ghedini, Matt Caswell*
2724 * Modify behavior of ALPN to invoke callback after SNI/servername
2725 callback, such that updates to the SSL_CTX affect ALPN.
2729 * Add SSL_CIPHER queries for authentication and key-exchange.
2733 * Changes to the DEFAULT cipherlist:
2734 - Prefer (EC)DHE handshakes over plain RSA.
2735 - Prefer AEAD ciphers over legacy ciphers.
2736 - Prefer ECDSA over RSA when both certificates are available.
2737 - Prefer TLSv1.2 ciphers/PRF.
2738 - Remove DSS, SEED, IDEA, CAMELLIA, and AES-CCM from the
2743 * Change the ECC default curve list to be this, in order: x25519,
2744 secp256r1, secp521r1, secp384r1.
2748 * RC4 based libssl ciphersuites are now classed as "weak" ciphers and are
2749 disabled by default. They can be re-enabled using the
2750 enable-weak-ssl-ciphers option to Configure.
2754 * If the server has ALPN configured, but supports no protocols that the
2755 client advertises, send a fatal "no_application_protocol" alert.
2756 This behaviour is SHALL in RFC 7301, though it isn't universally
2757 implemented by other servers.
2761 * Add X25519 support.
2762 Add ASN.1 and EVP_PKEY methods for X25519. This includes support
2763 for public and private key encoding using the format documented in
2764 draft-ietf-curdle-pkix-02. The corresponding EVP_PKEY method supports
2765 key generation and key derivation.
2767 TLS support complies with draft-ietf-tls-rfc4492bis-08 and uses
2772 * Deprecate SRP_VBASE_get_by_user.
2773 SRP_VBASE_get_by_user had inconsistent memory management behaviour.
2774 In order to fix an unavoidable memory leak [CVE-2016-0798][],
2775 SRP_VBASE_get_by_user was changed to ignore the "fake user" SRP
2776 seed, even if the seed is configured.
2778 Users should use SRP_VBASE_get1_by_user instead. Note that in
2779 SRP_VBASE_get1_by_user, caller must free the returned value. Note
2780 also that even though configuring the SRP seed attempts to hide
2781 invalid usernames by continuing the handshake with fake
2782 credentials, this behaviour is not constant time and no strong
2783 guarantees are made that the handshake is indistinguishable from
2784 that of a valid user.
2788 * Configuration change; it's now possible to build dynamic engines
2789 without having to build shared libraries and vice versa. This
2790 only applies to the engines in engines/, those in crypto/engine/
2791 will always be built into libcrypto (i.e. "static").
2793 Building dynamic engines is enabled by default; to disable, use
2794 the configuration option "disable-dynamic-engine".
2796 The only requirements for building dynamic engines are the
2797 presence of the DSO module and building with position independent
2798 code, so they will also automatically be disabled if configuring
2799 with "disable-dso" or "disable-pic".
2801 The macros OPENSSL_NO_STATIC_ENGINE and OPENSSL_NO_DYNAMIC_ENGINE
2802 are also taken away from openssl/opensslconf.h, as they are
2807 * Configuration change; if there is a known flag to compile
2808 position independent code, it will always be applied on the
2809 libcrypto and libssl object files, and never on the application
2810 object files. This means other libraries that use routines from
2811 libcrypto / libssl can be made into shared libraries regardless
2812 of how OpenSSL was configured.
2814 If this isn't desirable, the configuration options "disable-pic"
2815 or "no-pic" can be used to disable the use of PIC. This will
2816 also disable building shared libraries and dynamic engines.
2820 * Removed JPAKE code. It was experimental and has no wide use.
2824 * The INSTALL_PREFIX Makefile variable has been renamed to
2825 DESTDIR. That makes for less confusion on what this variable
2826 is for. Also, the configuration option --install_prefix is
2831 * Heartbeat for TLS has been removed and is disabled by default
2832 for DTLS; configure with enable-heartbeats. Code that uses the
2833 old #define's might need to be updated.
2835 *Emilia Käsper, Rich Salz*
2837 * Rename REF_CHECK to REF_DEBUG.
2841 * New "unified" build system
2843 The "unified" build system is aimed to be a common system for all
2844 platforms we support. With it comes new support for VMS.
2846 This system builds supports building in a different directory tree
2847 than the source tree. It produces one Makefile (for unix family
2848 or lookalikes), or one descrip.mms (for VMS).
2850 The source of information to make the Makefile / descrip.mms is
2851 small files called 'build.info', holding the necessary
2852 information for each directory with source to compile, and a
2853 template in Configurations, like unix-Makefile.tmpl or
2856 With this change, the library names were also renamed on Windows
2857 and on VMS. They now have names that are closer to the standard
2858 on Unix, and include the major version number, and in certain
2859 cases, the architecture they are built for. See "Notes on shared
2860 libraries" in INSTALL.
2862 We rely heavily on the perl module Text::Template.
2866 * Added support for auto-initialisation and de-initialisation of the library.
2867 OpenSSL no longer requires explicit init or deinit routines to be called,
2868 except in certain circumstances. See the OPENSSL_init_crypto() and
2869 OPENSSL_init_ssl() man pages for further information.
2873 * The arguments to the DTLSv1_listen function have changed. Specifically the
2874 "peer" argument is now expected to be a BIO_ADDR object.
2876 * Rewrite of BIO networking library. The BIO library lacked consistent
2877 support of IPv6, and adding it required some more extensive
2878 modifications. This introduces the BIO_ADDR and BIO_ADDRINFO types,
2879 which hold all types of addresses and chains of address information.
2880 It also introduces a new API, with functions like BIO_socket,
2881 BIO_connect, BIO_listen, BIO_lookup and a rewrite of BIO_accept.
2882 The source/sink BIOs BIO_s_connect, BIO_s_accept and BIO_s_datagram
2883 have been adapted accordingly.
2887 * RSA_padding_check_PKCS1_type_1 now accepts inputs with and without
2892 * CRIME protection: disable compression by default, even if OpenSSL is
2893 compiled with zlib enabled. Applications can still enable compression
2894 by calling SSL_CTX_clear_options(ctx, SSL_OP_NO_COMPRESSION), or by
2895 using the SSL_CONF library to configure compression.
2899 * The signature of the session callback configured with
2900 SSL_CTX_sess_set_get_cb was changed. The read-only input buffer
2901 was explicitly marked as `const unsigned char*` instead of
2906 * Always DPURIFY. Remove the use of uninitialized memory in the
2907 RNG, and other conditional uses of DPURIFY. This makes -DPURIFY a no-op.
2911 * Removed many obsolete configuration items, including
2912 DES_PTR, DES_RISC1, DES_RISC2, DES_INT
2913 MD2_CHAR, MD2_INT, MD2_LONG
2915 IDEA_SHORT, IDEA_LONG
2916 RC2_SHORT, RC2_LONG, RC4_LONG, RC4_CHUNK, RC4_INDEX
2918 *Rich Salz, with advice from Andy Polyakov*
2920 * Many BN internals have been moved to an internal header file.
2922 *Rich Salz with help from Andy Polyakov*
2924 * Configuration and writing out the results from it has changed.
2925 Files such as Makefile include/openssl/opensslconf.h and are now
2926 produced through general templates, such as Makefile.in and
2927 crypto/opensslconf.h.in and some help from the perl module
2930 Also, the center of configuration information is no longer
2931 Makefile. Instead, Configure produces a perl module in
2932 configdata.pm which holds most of the config data (in the hash
2933 table %config), the target data that comes from the target
2934 configuration in one of the `Configurations/*.conf~ files (in
2939 * To clarify their intended purposes, the Configure options
2940 --prefix and --openssldir change their semantics, and become more
2941 straightforward and less interdependent.
2943 --prefix shall be used exclusively to give the location INSTALLTOP
2944 where programs, scripts, libraries, include files and manuals are
2945 going to be installed. The default is now /usr/local.
2947 --openssldir shall be used exclusively to give the default
2948 location OPENSSLDIR where certificates, private keys, CRLs are
2949 managed. This is also where the default openssl.cnf gets
2951 If the directory given with this option is a relative path, the
2952 values of both the --prefix value and the --openssldir value will
2953 be combined to become OPENSSLDIR.
2954 The default for --openssldir is INSTALLTOP/ssl.
2956 Anyone who uses --openssldir to specify where OpenSSL is to be
2957 installed MUST change to use --prefix instead.
2961 * The GOST engine was out of date and therefore it has been removed. An up
2962 to date GOST engine is now being maintained in an external repository.
2963 See: <https://wiki.openssl.org/index.php/Binaries>. Libssl still retains
2964 support for GOST ciphersuites (these are only activated if a GOST engine
2969 * EGD is no longer supported by default; use enable-egd when
2972 *Ben Kaduk and Rich Salz*
2974 * The distribution now has Makefile.in files, which are used to
2975 create Makefile's when Configure is run. *Configure must be run
2976 before trying to build now.*
2980 * The return value for SSL_CIPHER_description() for error conditions
2985 * Support for RFC6698/RFC7671 DANE TLSA peer authentication.
2987 Obtaining and performing DNSSEC validation of TLSA records is
2988 the application's responsibility. The application provides
2989 the TLSA records of its choice to OpenSSL, and these are then
2990 used to authenticate the peer.
2992 The TLSA records need not even come from DNS. They can, for
2993 example, be used to implement local end-entity certificate or
2994 trust-anchor "pinning", where the "pin" data takes the form
2995 of TLSA records, which can augment or replace verification
2996 based on the usual WebPKI public certification authorities.
3000 * Revert default OPENSSL_NO_DEPRECATED setting. Instead OpenSSL
3001 continues to support deprecated interfaces in default builds.
3002 However, applications are strongly advised to compile their
3003 source files with -DOPENSSL_API_COMPAT=0x10100000L, which hides
3004 the declarations of all interfaces deprecated in 0.9.8, 1.0.0
3005 or the 1.1.0 releases.
3007 In environments in which all applications have been ported to
3008 not use any deprecated interfaces OpenSSL's Configure script
3009 should be used with the --api=1.1.0 option to entirely remove
3010 support for the deprecated features from the library and
3011 unconditionally disable them in the installed headers.
3012 Essentially the same effect can be achieved with the "no-deprecated"
3013 argument to Configure, except that this will always restrict
3014 the build to just the latest API, rather than a fixed API
3017 As applications are ported to future revisions of the API,
3018 they should update their compile-time OPENSSL_API_COMPAT define
3019 accordingly, but in most cases should be able to continue to
3020 compile with later releases.
3022 The OPENSSL_API_COMPAT versions for 1.0.0, and 0.9.8 are
3023 0x10000000L and 0x00908000L, respectively. However those
3024 versions did not support the OPENSSL_API_COMPAT feature, and
3025 so applications are not typically tested for explicit support
3026 of just the undeprecated features of either release.
3030 * Add support for setting the minimum and maximum supported protocol.
3031 It can bet set via the SSL_set_min_proto_version() and
3032 SSL_set_max_proto_version(), or via the SSL_CONF's MinProtocol and
3033 MaxProtocol. It's recommended to use the new APIs to disable
3034 protocols instead of disabling individual protocols using
3035 SSL_set_options() or SSL_CONF's Protocol. This change also
3036 removes support for disabling TLS 1.2 in the OpenSSL TLS
3037 client at compile time by defining OPENSSL_NO_TLS1_2_CLIENT.
3041 * Support for ChaCha20 and Poly1305 added to libcrypto and libssl.
3045 * New EC_KEY_METHOD, this replaces the older ECDSA_METHOD and ECDH_METHOD
3046 and integrates ECDSA and ECDH functionality into EC. Implementations can
3047 now redirect key generation and no longer need to convert to or from
3050 Note: the ecdsa.h and ecdh.h headers are now no longer needed and just
3051 include the ec.h header file instead.
3055 * Remove support for all 40 and 56 bit ciphers. This includes all the export
3056 ciphers who are no longer supported and drops support the ephemeral RSA key
3057 exchange. The LOW ciphers currently doesn't have any ciphers in it.
3061 * Made EVP_MD_CTX, EVP_MD, EVP_CIPHER_CTX, EVP_CIPHER and HMAC_CTX
3062 opaque. For HMAC_CTX, the following constructors and destructors
3065 HMAC_CTX *HMAC_CTX_new(void);
3066 void HMAC_CTX_free(HMAC_CTX *ctx);
3068 For EVP_MD and EVP_CIPHER, complete APIs to create, fill and
3069 destroy such methods has been added. See EVP_MD_meth_new(3) and
3070 EVP_CIPHER_meth_new(3) for documentation.
3073 1) EVP_MD_CTX_cleanup(), EVP_CIPHER_CTX_cleanup() and
3074 HMAC_CTX_cleanup() were removed. HMAC_CTX_reset() and
3075 EVP_MD_CTX_reset() should be called instead to reinitialise
3076 an already created structure.
3077 2) For consistency with the majority of our object creators and
3078 destructors, EVP_MD_CTX_(create|destroy) were renamed to
3079 EVP_MD_CTX_(new|free). The old names are retained as macros
3080 for deprecated builds.
3084 * Added ASYNC support. Libcrypto now includes the async sub-library to enable
3085 cryptographic operations to be performed asynchronously as long as an
3086 asynchronous capable engine is used. See the ASYNC_start_job() man page for
3087 further details. Libssl has also had this capability integrated with the
3088 introduction of the new mode SSL_MODE_ASYNC and associated error
3089 SSL_ERROR_WANT_ASYNC. See the SSL_CTX_set_mode() and SSL_get_error() man
3090 pages. This work was developed in partnership with Intel Corp.
3094 * SSL_{CTX_}set_ecdh_auto() has been removed and ECDH is support is
3095 always enabled now. If you want to disable the support you should
3096 exclude it using the list of supported ciphers. This also means that the
3097 "-no_ecdhe" option has been removed from s_server.
3101 * SSL_{CTX}_set_tmp_ecdh() which can set 1 EC curve now internally calls
3102 SSL_{CTX_}set1_curves() which can set a list.
3106 * Remove support for SSL_{CTX_}set_tmp_ecdh_callback(). You should set the
3107 curve you want to support using SSL_{CTX_}set1_curves().
3111 * State machine rewrite. The state machine code has been significantly
3112 refactored in order to remove much duplication of code and solve issues
3113 with the old code (see [ssl/statem/README.md](ssl/statem/README.md) for
3114 further details). This change does have some associated API changes.
3115 Notably the SSL_state() function has been removed and replaced by
3116 SSL_get_state which now returns an "OSSL_HANDSHAKE_STATE" instead of an int.
3117 SSL_set_state() has been removed altogether. The previous handshake states
3118 defined in ssl.h and ssl3.h have also been removed.
3122 * All instances of the string "ssleay" in the public API were replaced
3123 with OpenSSL (case-matching; e.g., OPENSSL_VERSION for #define's)
3124 Some error codes related to internal RSA_eay API's were renamed.
3128 * The demo files in crypto/threads were moved to demo/threads.
3132 * Removed obsolete engines: 4758cca, aep, atalla, cswift, nuron, gmp,
3135 *Matt Caswell, Rich Salz*
3137 * New ASN.1 embed macro.
3139 New ASN.1 macro ASN1_EMBED. This is the same as ASN1_SIMPLE except the
3140 structure is not allocated: it is part of the parent. That is instead of
3148 This reduces memory fragmentation and make it impossible to accidentally
3149 set a mandatory field to NULL.
3151 This currently only works for some fields specifically a SEQUENCE, CHOICE,
3152 or ASN1_STRING type which is part of a parent SEQUENCE. Since it is
3153 equivalent to ASN1_SIMPLE it cannot be tagged, OPTIONAL, SET OF or
3158 * Remove EVP_CHECK_DES_KEY, a compile-time option that never compiled.
3162 * Removed DES and RC4 ciphersuites from DEFAULT. Also removed RC2 although
3163 in 1.0.2 EXPORT was already removed and the only RC2 ciphersuite is also
3164 an EXPORT one. COMPLEMENTOFDEFAULT has been updated accordingly to add
3165 DES and RC4 ciphersuites.
3169 * Rewrite EVP_DecodeUpdate (base64 decoding) to fix several bugs.
3170 This changes the decoding behaviour for some invalid messages,
3171 though the change is mostly in the more lenient direction, and
3172 legacy behaviour is preserved as much as possible.
3176 * Fix no-stdio build.
3177 * David Woodhouse <David.Woodhouse@intel.com> and also
3178 Ivan Nestlerode <ivan.nestlerode@sonos.com> *
3180 * New testing framework
3181 The testing framework has been largely rewritten and is now using
3182 perl and the perl modules Test::Harness and an extended variant of
3183 Test::More called OpenSSL::Test to do its work. All test scripts in
3184 test/ have been rewritten into test recipes, and all direct calls to
3185 executables in test/Makefile have become individual recipes using the
3186 simplified testing OpenSSL::Test::Simple.
3188 For documentation on our testing modules, do:
3190 perldoc test/testlib/OpenSSL/Test/Simple.pm
3191 perldoc test/testlib/OpenSSL/Test.pm
3195 * Revamped memory debug; only -DCRYPTO_MDEBUG and -DCRYPTO_MDEBUG_ABORT
3196 are used; the latter aborts on memory leaks (usually checked on exit).
3197 Some undocumented "set malloc, etc., hooks" functions were removed
3198 and others were changed. All are now documented.
3202 * In DSA_generate_parameters_ex, if the provided seed is too short,
3205 *Rich Salz and Ismo Puustinen <ismo.puustinen@intel.com>*
3207 * Rewrite PSK to support ECDHE_PSK, DHE_PSK and RSA_PSK. Add ciphersuites
3208 from RFC4279, RFC4785, RFC5487, RFC5489.
3210 Thanks to Christian J. Dietrich and Giuseppe D'Angelo for the
3211 original RSA_PSK patch.
3215 * Dropped support for the SSL3_FLAGS_DELAY_CLIENT_FINISHED flag. This SSLeay
3216 era flag was never set throughout the codebase (only read). Also removed
3217 SSL3_FLAGS_POP_BUFFER which was only used if
3218 SSL3_FLAGS_DELAY_CLIENT_FINISHED was also set.
3222 * Changed the default name options in the "ca", "crl", "req" and "x509"
3223 to be "oneline" instead of "compat".
3227 * Remove SSL_OP_TLS_BLOCK_PADDING_BUG. This is SSLeay legacy, we're
3228 not aware of clients that still exhibit this bug, and the workaround
3229 hasn't been working properly for a while.
3233 * The return type of BIO_number_read() and BIO_number_written() as well as
3234 the corresponding num_read and num_write members in the BIO structure has
3235 changed from unsigned long to uint64_t. On platforms where an unsigned
3236 long is 32 bits (e.g. Windows) these counters could overflow if >4Gb is
3241 * Given the pervasive nature of TLS extensions it is inadvisable to run
3242 OpenSSL without support for them. It also means that maintaining
3243 the OPENSSL_NO_TLSEXT option within the code is very invasive (and probably
3244 not well tested). Therefore the OPENSSL_NO_TLSEXT option has been removed.
3248 * Removed support for the two export grade static DH ciphersuites
3249 EXP-DH-RSA-DES-CBC-SHA and EXP-DH-DSS-DES-CBC-SHA. These two ciphersuites
3250 were newly added (along with a number of other static DH ciphersuites) to
3251 1.0.2. However the two export ones have *never* worked since they were
3252 introduced. It seems strange in any case to be adding new export
3253 ciphersuites, and given "logjam" it also does not seem correct to fix them.
3257 * Version negotiation has been rewritten. In particular SSLv23_method(),
3258 SSLv23_client_method() and SSLv23_server_method() have been deprecated,
3259 and turned into macros which simply call the new preferred function names
3260 TLS_method(), TLS_client_method() and TLS_server_method(). All new code
3261 should use the new names instead. Also as part of this change the ssl23.h
3262 header file has been removed.
3266 * Support for Kerberos ciphersuites in TLS (RFC2712) has been removed. This
3267 code and the associated standard is no longer considered fit-for-purpose.
3271 * RT2547 was closed. When generating a private key, try to make the
3272 output file readable only by the owner. This behavior change might
3273 be noticeable when interacting with other software.
3275 * Documented all exdata functions. Added CRYPTO_free_ex_index.
3280 * Added HTTP GET support to the ocsp command.
3284 * Changed default digest for the dgst and enc commands from MD5 to
3289 * RAND_pseudo_bytes has been deprecated. Users should use RAND_bytes instead.
3293 * Added support for TLS extended master secret from
3294 draft-ietf-tls-session-hash-03.txt. Thanks for Alfredo Pironti for an
3295 initial patch which was a great help during development.
3299 * All libssl internal structures have been removed from the public header
3300 files, and the OPENSSL_NO_SSL_INTERN option has been removed (since it is
3301 now redundant). Users should not attempt to access internal structures
3302 directly. Instead they should use the provided API functions.
3306 * config has been changed so that by default OPENSSL_NO_DEPRECATED is used.
3307 Access to deprecated functions can be re-enabled by running config with
3308 "enable-deprecated". In addition applications wishing to use deprecated
3309 functions must define OPENSSL_USE_DEPRECATED. Note that this new behaviour
3310 will, by default, disable some transitive includes that previously existed
3311 in the header files (e.g. ec.h will no longer, by default, include bn.h)
3315 * Added support for OCB mode. OpenSSL has been granted a patent license
3316 compatible with the OpenSSL license for use of OCB. Details are available
3317 at <https://www.openssl.org/source/OCB-patent-grant-OpenSSL.pdf>. Support
3318 for OCB can be removed by calling config with no-ocb.
3322 * SSLv2 support has been removed. It still supports receiving a SSLv2
3323 compatible client hello.
3327 * Increased the minimal RSA keysize from 256 to 512 bits [Rich Salz],
3328 done while fixing the error code for the key-too-small case.
3330 *Annie Yousar <a.yousar@informatik.hu-berlin.de>*
3332 * CA.sh has been removed; use CA.pl instead.
3336 * Removed old DES API.
3340 * Remove various unsupported platforms:
3346 Sinix/ReliantUNIX RM400
3351 16-bit platforms such as WIN16
3355 * Clean up OPENSSL_NO_xxx #define's
3356 - Use setbuf() and remove OPENSSL_NO_SETVBUF_IONBF
3357 - Rename OPENSSL_SYSNAME_xxx to OPENSSL_SYS_xxx
3358 - OPENSSL_NO_EC{DH,DSA} merged into OPENSSL_NO_EC
3359 - OPENSSL_NO_RIPEMD160, OPENSSL_NO_RIPEMD merged into OPENSSL_NO_RMD160
3360 - OPENSSL_NO_FP_API merged into OPENSSL_NO_STDIO
3361 - Remove OPENSSL_NO_BIO OPENSSL_NO_BUFFER OPENSSL_NO_CHAIN_VERIFY
3362 OPENSSL_NO_EVP OPENSSL_NO_FIPS_ERR OPENSSL_NO_HASH_COMP
3363 OPENSSL_NO_LHASH OPENSSL_NO_OBJECT OPENSSL_NO_SPEED OPENSSL_NO_STACK
3364 OPENSSL_NO_X509 OPENSSL_NO_X509_VERIFY
3365 - Remove MS_STATIC; it's a relic from platforms <32 bits.
3369 * Cleaned up dead code
3370 Remove all but one '#ifdef undef' which is to be looked at.
3374 * Clean up calling of xxx_free routines.
3375 Just like free(), fix most of the xxx_free routines to accept
3376 NULL. Remove the non-null checks from callers. Save much code.
3380 * Add secure heap for storage of private keys (when possible).
3381 Add BIO_s_secmem(), CBIGNUM, etc.
3382 Contributed by Akamai Technologies under our Corporate CLA.
3386 * Experimental support for a new, fast, unbiased prime candidate generator,
3387 bn_probable_prime_dh_coprime(). Not currently used by any prime generator.
3389 *Felix Laurie von Massenbach <felix@erbridge.co.uk>*
3391 * New output format NSS in the sess_id command line tool. This allows
3392 exporting the session id and the master key in NSS keylog format.
3394 *Martin Kaiser <martin@kaiser.cx>*
3396 * Harmonize version and its documentation. -f flag is used to display
3399 *mancha <mancha1@zoho.com>*
3401 * Fix eckey_priv_encode so it immediately returns an error upon a failure
3402 in i2d_ECPrivateKey. Thanks to Ted Unangst for feedback on this issue.
3404 *mancha <mancha1@zoho.com>*
3406 * Fix some double frees. These are not thought to be exploitable.
3408 *mancha <mancha1@zoho.com>*
3410 * A missing bounds check in the handling of the TLS heartbeat extension
3411 can be used to reveal up to 64k of memory to a connected client or
3414 Thanks for Neel Mehta of Google Security for discovering this bug and to
3415 Adam Langley <agl@chromium.org> and Bodo Moeller <bmoeller@acm.org> for
3416 preparing the fix [CVE-2014-0160][]
3418 *Adam Langley, Bodo Moeller*
3420 * Fix for the attack described in the paper "Recovering OpenSSL
3421 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
3422 by Yuval Yarom and Naomi Benger. Details can be obtained from:
3423 <http://eprint.iacr.org/2014/140>
3425 Thanks to Yuval Yarom and Naomi Benger for discovering this
3426 flaw and to Yuval Yarom for supplying a fix [CVE-2014-0076][]
3428 *Yuval Yarom and Naomi Benger*
3430 * Use algorithm specific chains in SSL_CTX_use_certificate_chain_file():
3431 this fixes a limitation in previous versions of OpenSSL.
3435 * Experimental encrypt-then-mac support.
3437 Experimental support for encrypt then mac from
3438 draft-gutmann-tls-encrypt-then-mac-02.txt
3440 To enable it set the appropriate extension number (0x42 for the test
3441 server) using e.g. -DTLSEXT_TYPE_encrypt_then_mac=0x42
3443 For non-compliant peers (i.e. just about everything) this should have no
3446 WARNING: EXPERIMENTAL, SUBJECT TO CHANGE.
3450 * Add EVP support for key wrapping algorithms, to avoid problems with
3451 existing code the flag EVP_CIPHER_CTX_WRAP_ALLOW has to be set in
3452 the EVP_CIPHER_CTX or an error is returned. Add AES and DES3 wrap
3453 algorithms and include tests cases.
3457 * Extend CMS code to support RSA-PSS signatures and RSA-OAEP for
3462 * Extended RSA OAEP support via EVP_PKEY API. Options to specify digest,
3463 MGF1 digest and OAEP label.
3467 * Make openssl verify return errors.
3469 *Chris Palmer <palmer@google.com> and Ben Laurie*
3471 * New function ASN1_TIME_diff to calculate the difference between two
3472 ASN1_TIME structures or one structure and the current time.
3476 * Update fips_test_suite to support multiple command line options. New
3477 test to induce all self test errors in sequence and check expected
3482 * Add FIPS_{rsa,dsa,ecdsa}_{sign,verify} functions which digest and
3483 sign or verify all in one operation.
3487 * Add fips_algvs: a multicall fips utility incorporating all the algorithm
3488 test programs and fips_test_suite. Includes functionality to parse
3489 the minimal script output of fipsalgest.pl directly.
3493 * Add authorisation parameter to FIPS_module_mode_set().
3497 * Add FIPS selftest for ECDH algorithm using P-224 and B-233 curves.
3501 * Use separate DRBG fields for internal and external flags. New function
3502 FIPS_drbg_health_check() to perform on demand health checking. Add
3503 generation tests to fips_test_suite with reduced health check interval to
3504 demonstrate periodic health checking. Add "nodh" option to
3505 fips_test_suite to skip very slow DH test.
3509 * New function FIPS_get_cipherbynid() to lookup FIPS supported ciphers
3514 * More extensive health check for DRBG checking many more failure modes.
3515 New function FIPS_selftest_drbg_all() to handle every possible DRBG
3516 combination: call this in fips_test_suite.
3520 * Add support for canonical generation of DSA parameter 'g'. See
3523 * Add support for HMAC DRBG from SP800-90. Update DRBG algorithm test and
3524 POST to handle HMAC cases.
3528 * Add functions FIPS_module_version() and FIPS_module_version_text()
3529 to return numerical and string versions of the FIPS module number.
3533 * Rename FIPS_mode_set and FIPS_mode to FIPS_module_mode_set and
3534 FIPS_module_mode. FIPS_mode and FIPS_mode_set will be implemented
3535 outside the validated module in the FIPS capable OpenSSL.
3539 * Minor change to DRBG entropy callback semantics. In some cases
3540 there is no multiple of the block length between min_len and
3541 max_len. Allow the callback to return more than max_len bytes
3542 of entropy but discard any extra: it is the callback's responsibility
3543 to ensure that the extra data discarded does not impact the
3544 requested amount of entropy.
3548 * Add PRNG security strength checks to RSA, DSA and ECDSA using
3549 information in FIPS186-3, SP800-57 and SP800-131A.
3553 * CCM support via EVP. Interface is very similar to GCM case except we
3554 must supply all data in one chunk (i.e. no update, final) and the
3555 message length must be supplied if AAD is used. Add algorithm test
3560 * Initial version of POST overhaul. Add POST callback to allow the status
3561 of POST to be monitored and/or failures induced. Modify fips_test_suite
3562 to use callback. Always run all selftests even if one fails.
3566 * XTS support including algorithm test driver in the fips_gcmtest program.
3567 Note: this does increase the maximum key length from 32 to 64 bytes but
3568 there should be no binary compatibility issues as existing applications
3569 will never use XTS mode.
3573 * Extensive reorganisation of FIPS PRNG behaviour. Remove all dependencies
3574 to OpenSSL RAND code and replace with a tiny FIPS RAND API which also
3575 performs algorithm blocking for unapproved PRNG types. Also do not
3576 set PRNG type in FIPS_mode_set(): leave this to the application.
3577 Add default OpenSSL DRBG handling: sets up FIPS PRNG and seeds with
3578 the standard OpenSSL PRNG: set additional data to a date time vector.
3582 * Rename old X9.31 PRNG functions of the form FIPS_rand* to FIPS_x931*.
3583 This shouldn't present any incompatibility problems because applications
3584 shouldn't be using these directly and any that are will need to rethink
3585 anyway as the X9.31 PRNG is now deprecated by FIPS 140-2
3589 * Extensive self tests and health checking required by SP800-90 DRBG.
3590 Remove strength parameter from FIPS_drbg_instantiate and always
3591 instantiate at maximum supported strength.
3595 * Add ECDH code to fips module and fips_ecdhvs for primitives only testing.
3599 * New algorithm test program fips_dhvs to handle DH primitives only testing.
3603 * New function DH_compute_key_padded() to compute a DH key and pad with
3604 leading zeroes if needed: this complies with SP800-56A et al.
3608 * Initial implementation of SP800-90 DRBGs for Hash and CTR. Not used by
3609 anything, incomplete, subject to change and largely untested at present.
3613 * Modify fipscanisteronly build option to only build the necessary object
3614 files by filtering FIPS_EX_OBJ through a perl script in crypto/Makefile.
3618 * Add experimental option FIPSSYMS to give all symbols in
3619 fipscanister.o and FIPS or fips prefix. This will avoid
3620 conflicts with future versions of OpenSSL. Add perl script
3621 util/fipsas.pl to preprocess assembly language source files
3622 and rename any affected symbols.
3626 * Add selftest checks and algorithm block of non-fips algorithms in
3627 FIPS mode. Remove DES2 from selftests.
3631 * Add ECDSA code to fips module. Add tiny fips_ecdsa_check to just
3632 return internal method without any ENGINE dependencies. Add new
3633 tiny fips sign and verify functions.
3637 * New build option no-ec2m to disable characteristic 2 code.
3641 * New build option "fipscanisteronly". This only builds fipscanister.o
3642 and (currently) associated fips utilities. Uses the file Makefile.fips
3643 instead of Makefile.org as the prototype.
3647 * Add some FIPS mode restrictions to GCM. Add internal IV generator.
3648 Update fips_gcmtest to use IV generator.
3652 * Initial, experimental EVP support for AES-GCM. AAD can be input by
3653 setting output buffer to NULL. The `*Final` function must be
3654 called although it will not retrieve any additional data. The tag
3655 can be set or retrieved with a ctrl. The IV length is by default 12
3656 bytes (96 bits) but can be set to an alternative value. If the IV
3657 length exceeds the maximum IV length (currently 16 bytes) it cannot be
3662 * New flag in ciphers: EVP_CIPH_FLAG_CUSTOM_CIPHER. This means the
3663 underlying do_cipher function handles all cipher semantics itself
3664 including padding and finalisation. This is useful if (for example)
3665 an ENGINE cipher handles block padding itself. The behaviour of
3666 do_cipher is subtly changed if this flag is set: the return value
3667 is the number of characters written to the output buffer (zero is
3668 no longer an error code) or a negative error code. Also if the
3669 input buffer is NULL and length 0 finalisation should be performed.
3673 * If a candidate issuer certificate is already part of the constructed
3674 path ignore it: new debug notification X509_V_ERR_PATH_LOOP for this case.
3678 * Improve forward-security support: add functions
3680 void SSL_CTX_set_not_resumable_session_callback(
3681 SSL_CTX *ctx, int (*cb)(SSL *ssl, int is_forward_secure))
3682 void SSL_set_not_resumable_session_callback(
3683 SSL *ssl, int (*cb)(SSL *ssl, int is_forward_secure))
3685 for use by SSL/TLS servers; the callback function will be called whenever a
3686 new session is created, and gets to decide whether the session may be
3687 cached to make it resumable (return 0) or not (return 1). (As by the
3688 SSL/TLS protocol specifications, the session_id sent by the server will be
3689 empty to indicate that the session is not resumable; also, the server will
3690 not generate RFC 4507 (RFC 5077) session tickets.)
3692 A simple reasonable callback implementation is to return is_forward_secure.
3693 This parameter will be set to 1 or 0 depending on the ciphersuite selected
3694 by the SSL/TLS server library, indicating whether it can provide forward
3697 *Emilia Käsper <emilia.kasper@esat.kuleuven.be> (Google)*
3699 * New -verify_name option in command line utilities to set verification
3704 * Initial CMAC implementation. WARNING: EXPERIMENTAL, API MAY CHANGE.
3705 Add CMAC pkey methods.
3709 * Experimental renegotiation in s_server -www mode. If the client
3710 browses /reneg connection is renegotiated. If /renegcert it is
3711 renegotiated requesting a certificate.
3715 * Add an "external" session cache for debugging purposes to s_server. This
3716 should help trace issues which normally are only apparent in deployed
3717 multi-process servers.
3721 * Extensive audit of libcrypto with DEBUG_UNUSED. Fix many cases where
3722 return value is ignored. NB. The functions RAND_add(), RAND_seed(),
3723 BIO_set_cipher() and some obscure PEM functions were changed so they
3724 can now return an error. The RAND changes required a change to the
3725 RAND_METHOD structure.
3729 * New macro `__owur` for "OpenSSL Warn Unused Result". This makes use of
3730 a gcc attribute to warn if the result of a function is ignored. This
3731 is enable if DEBUG_UNUSED is set. Add to several functions in evp.h
3732 whose return value is often ignored.
3736 * New -noct, -requestct, -requirect and -ctlogfile options for s_client.
3737 These allow SCTs (signed certificate timestamps) to be requested and
3738 validated when establishing a connection.
3740 *Rob Percival <robpercival@google.com>*
3745 ### Changes between 1.0.2s and 1.0.2t [10 Sep 2019]
3747 * For built-in EC curves, ensure an EC_GROUP built from the curve name is
3748 used even when parsing explicit parameters, when loading a serialized key
3749 or calling `EC_GROUP_new_from_ecpkparameters()`/
3750 `EC_GROUP_new_from_ecparameters()`.
3751 This prevents bypass of security hardening and performance gains,
3752 especially for curves with specialized EC_METHODs.
3753 By default, if a key encoded with explicit parameters is loaded and later
3754 serialized, the output is still encoded with explicit parameters, even if
3755 internally a "named" EC_GROUP is used for computation.
3759 * Compute ECC cofactors if not provided during EC_GROUP construction. Before
3760 this change, EC_GROUP_set_generator would accept order and/or cofactor as
3761 NULL. After this change, only the cofactor parameter can be NULL. It also
3762 does some minimal sanity checks on the passed order.
3767 * Fixed a padding oracle in PKCS7_dataDecode and CMS_decrypt_set1_pkey.
3768 An attack is simple, if the first CMS_recipientInfo is valid but the
3769 second CMS_recipientInfo is chosen ciphertext. If the second
3770 recipientInfo decodes to PKCS #1 v1.5 form plaintext, the correct
3771 encryption key will be replaced by garbage, and the message cannot be
3772 decoded, but if the RSA decryption fails, the correct encryption key is
3773 used and the recipient will not notice the attack.
3774 As a work around for this potential attack the length of the decrypted
3775 key must be equal to the cipher default key length, in case the
3776 certifiate is not given and all recipientInfo are tried out.
3777 The old behaviour can be re-enabled in the CMS code by setting the
3778 CMS_DEBUG_DECRYPT flag.
3783 * Document issue with installation paths in diverse Windows builds
3785 '/usr/local/ssl' is an unsafe prefix for location to install OpenSSL
3786 binaries and run-time config file.
3791 ### Changes between 1.0.2r and 1.0.2s [28 May 2019]
3793 * Change the default RSA, DSA and DH size to 2048 bit instead of 1024.
3794 This changes the size when using the genpkey app when no size is given. It
3795 fixes an omission in earlier changes that changed all RSA, DSA and DH
3796 generation apps to use 2048 bits by default.
3800 * Add FIPS support for Android Arm 64-bit
3802 Support for Android Arm 64-bit was added to the OpenSSL FIPS Object
3803 Module in Version 2.0.10. For some reason, the corresponding target
3804 'android64-aarch64' was missing OpenSSL 1.0.2, whence it could not be
3805 built with FIPS support on Android Arm 64-bit. This omission has been
3808 *Matthias St. Pierre*
3810 ### Changes between 1.0.2q and 1.0.2r [26 Feb 2019]
3812 * 0-byte record padding oracle
3814 If an application encounters a fatal protocol error and then calls
3815 SSL_shutdown() twice (once to send a close_notify, and once to receive one)
3816 then OpenSSL can respond differently to the calling application if a 0 byte
3817 record is received with invalid padding compared to if a 0 byte record is
3818 received with an invalid MAC. If the application then behaves differently
3819 based on that in a way that is detectable to the remote peer, then this
3820 amounts to a padding oracle that could be used to decrypt data.
3822 In order for this to be exploitable "non-stitched" ciphersuites must be in
3823 use. Stitched ciphersuites are optimised implementations of certain
3824 commonly used ciphersuites. Also the application must call SSL_shutdown()
3825 twice even if a protocol error has occurred (applications should not do
3826 this but some do anyway).
3828 This issue was discovered by Juraj Somorovsky, Robert Merget and Nimrod
3829 Aviram, with additional investigation by Steven Collison and Andrew
3830 Hourselt. It was reported to OpenSSL on 10th December 2018.
3835 * Move strictness check from EVP_PKEY_asn1_new() to EVP_PKEY_asn1_add0().
3839 ### Changes between 1.0.2p and 1.0.2q [20 Nov 2018]
3841 * Microarchitecture timing vulnerability in ECC scalar multiplication
3843 OpenSSL ECC scalar multiplication, used in e.g. ECDSA and ECDH, has been
3844 shown to be vulnerable to a microarchitecture timing side channel attack.
3845 An attacker with sufficient access to mount local timing attacks during
3846 ECDSA signature generation could recover the private key.
3848 This issue was reported to OpenSSL on 26th October 2018 by Alejandro
3849 Cabrera Aldaya, Billy Brumley, Sohaib ul Hassan, Cesar Pereida Garcia and
3855 * Timing vulnerability in DSA signature generation
3857 The OpenSSL DSA signature algorithm has been shown to be vulnerable to a
3858 timing side channel attack. An attacker could use variations in the signing
3859 algorithm to recover the private key.
3861 This issue was reported to OpenSSL on 16th October 2018 by Samuel Weiser.
3866 * Resolve a compatibility issue in EC_GROUP handling with the FIPS Object
3867 Module, accidentally introduced while backporting security fixes from the
3868 development branch and hindering the use of ECC in FIPS mode.
3872 ### Changes between 1.0.2o and 1.0.2p [14 Aug 2018]
3874 * Client DoS due to large DH parameter
3876 During key agreement in a TLS handshake using a DH(E) based ciphersuite a
3877 malicious server can send a very large prime value to the client. This will
3878 cause the client to spend an unreasonably long period of time generating a
3879 key for this prime resulting in a hang until the client has finished. This
3880 could be exploited in a Denial Of Service attack.
3882 This issue was reported to OpenSSL on 5th June 2018 by Guido Vranken
3887 * Cache timing vulnerability in RSA Key Generation
3889 The OpenSSL RSA Key generation algorithm has been shown to be vulnerable to
3890 a cache timing side channel attack. An attacker with sufficient access to
3891 mount cache timing attacks during the RSA key generation process could
3892 recover the private key.
3894 This issue was reported to OpenSSL on 4th April 2018 by Alejandro Cabrera
3895 Aldaya, Billy Brumley, Cesar Pereida Garcia and Luis Manuel Alvarez Tapia.
3900 * Make EVP_PKEY_asn1_new() a bit stricter about its input. A NULL pem_str
3901 parameter is no longer accepted, as it leads to a corrupt table. NULL
3902 pem_str is reserved for alias entries only.
3906 * Revert blinding in ECDSA sign and instead make problematic addition
3907 length-invariant. Switch even to fixed-length Montgomery multiplication.
3911 * Change generating and checking of primes so that the error rate of not
3912 being prime depends on the intended use based on the size of the input.
3913 For larger primes this will result in more rounds of Miller-Rabin.
3914 The maximal error rate for primes with more than 1080 bits is lowered
3917 *Kurt Roeckx, Annie Yousar*
3919 * Increase the number of Miller-Rabin rounds for DSA key generating to 64.
3923 * Add blinding to ECDSA and DSA signatures to protect against side channel
3924 attacks discovered by Keegan Ryan (NCC Group).
3928 * When unlocking a pass phrase protected PEM file or PKCS#8 container, we
3929 now allow empty (zero character) pass phrases.
3933 * Certificate time validation (X509_cmp_time) enforces stricter
3934 compliance with RFC 5280. Fractional seconds and timezone offsets
3935 are no longer allowed.
3939 ### Changes between 1.0.2n and 1.0.2o [27 Mar 2018]
3941 * Constructed ASN.1 types with a recursive definition could exceed the stack
3943 Constructed ASN.1 types with a recursive definition (such as can be found
3944 in PKCS7) could eventually exceed the stack given malicious input with
3945 excessive recursion. This could result in a Denial Of Service attack. There
3946 are no such structures used within SSL/TLS that come from untrusted sources
3947 so this is considered safe.
3949 This issue was reported to OpenSSL on 4th January 2018 by the OSS-fuzz
3955 ### Changes between 1.0.2m and 1.0.2n [7 Dec 2017]
3957 * Read/write after SSL object in error state
3959 OpenSSL 1.0.2 (starting from version 1.0.2b) introduced an "error state"
3960 mechanism. The intent was that if a fatal error occurred during a handshake
3961 then OpenSSL would move into the error state and would immediately fail if
3962 you attempted to continue the handshake. This works as designed for the
3963 explicit handshake functions (SSL_do_handshake(), SSL_accept() and
3964 SSL_connect()), however due to a bug it does not work correctly if
3965 SSL_read() or SSL_write() is called directly. In that scenario, if the
3966 handshake fails then a fatal error will be returned in the initial function
3967 call. If SSL_read()/SSL_write() is subsequently called by the application
3968 for the same SSL object then it will succeed and the data is passed without
3969 being decrypted/encrypted directly from the SSL/TLS record layer.
3971 In order to exploit this issue an application bug would have to be present
3972 that resulted in a call to SSL_read()/SSL_write() being issued after having
3973 already received a fatal error.
3975 This issue was reported to OpenSSL by David Benjamin (Google).
3980 * rsaz_1024_mul_avx2 overflow bug on x86_64
3982 There is an overflow bug in the AVX2 Montgomery multiplication procedure
3983 used in exponentiation with 1024-bit moduli. No EC algorithms are affected.
3984 Analysis suggests that attacks against RSA and DSA as a result of this
3985 defect would be very difficult to perform and are not believed likely.
3986 Attacks against DH1024 are considered just feasible, because most of the
3987 work necessary to deduce information about a private key may be performed
3988 offline. The amount of resources required for such an attack would be
3989 significant. However, for an attack on TLS to be meaningful, the server
3990 would have to share the DH1024 private key among multiple clients, which is
3991 no longer an option since CVE-2016-0701.
3993 This only affects processors that support the AVX2 but not ADX extensions
3994 like Intel Haswell (4th generation).
3996 This issue was reported to OpenSSL by David Benjamin (Google). The issue
3997 was originally found via the OSS-Fuzz project.
4002 ### Changes between 1.0.2l and 1.0.2m [2 Nov 2017]
4004 * bn_sqrx8x_internal carry bug on x86_64
4006 There is a carry propagating bug in the x86_64 Montgomery squaring
4007 procedure. No EC algorithms are affected. Analysis suggests that attacks
4008 against RSA and DSA as a result of this defect would be very difficult to
4009 perform and are not believed likely. Attacks against DH are considered just
4010 feasible (although very difficult) because most of the work necessary to
4011 deduce information about a private key may be performed offline. The amount
4012 of resources required for such an attack would be very significant and
4013 likely only accessible to a limited number of attackers. An attacker would
4014 additionally need online access to an unpatched system using the target
4015 private key in a scenario with persistent DH parameters and a private
4016 key that is shared between multiple clients.
4018 This only affects processors that support the BMI1, BMI2 and ADX extensions
4019 like Intel Broadwell (5th generation) and later or AMD Ryzen.
4021 This issue was reported to OpenSSL by the OSS-Fuzz project.
4026 * Malformed X.509 IPAddressFamily could cause OOB read
4028 If an X.509 certificate has a malformed IPAddressFamily extension,
4029 OpenSSL could do a one-byte buffer overread. The most likely result
4030 would be an erroneous display of the certificate in text format.
4032 This issue was reported to OpenSSL by the OSS-Fuzz project.
4037 ### Changes between 1.0.2k and 1.0.2l [25 May 2017]
4039 * Have 'config' recognise 64-bit mingw and choose 'mingw64' as the target
4040 platform rather than 'mingw'.
4044 ### Changes between 1.0.2j and 1.0.2k [26 Jan 2017]
4046 * Truncated packet could crash via OOB read
4048 If one side of an SSL/TLS path is running on a 32-bit host and a specific
4049 cipher is being used, then a truncated packet can cause that host to
4050 perform an out-of-bounds read, usually resulting in a crash.
4052 This issue was reported to OpenSSL by Robert Święcki of Google.
4057 * BN_mod_exp may produce incorrect results on x86_64
4059 There is a carry propagating bug in the x86_64 Montgomery squaring
4060 procedure. No EC algorithms are affected. Analysis suggests that attacks
4061 against RSA and DSA as a result of this defect would be very difficult to
4062 perform and are not believed likely. Attacks against DH are considered just
4063 feasible (although very difficult) because most of the work necessary to
4064 deduce information about a private key may be performed offline. The amount
4065 of resources required for such an attack would be very significant and
4066 likely only accessible to a limited number of attackers. An attacker would
4067 additionally need online access to an unpatched system using the target
4068 private key in a scenario with persistent DH parameters and a private
4069 key that is shared between multiple clients. For example this can occur by
4070 default in OpenSSL DHE based SSL/TLS ciphersuites. Note: This issue is very
4071 similar to CVE-2015-3193 but must be treated as a separate problem.
4073 This issue was reported to OpenSSL by the OSS-Fuzz project.
4078 * Montgomery multiplication may produce incorrect results
4080 There is a carry propagating bug in the Broadwell-specific Montgomery
4081 multiplication procedure that handles input lengths divisible by, but
4082 longer than 256 bits. Analysis suggests that attacks against RSA, DSA
4083 and DH private keys are impossible. This is because the subroutine in
4084 question is not used in operations with the private key itself and an input
4085 of the attacker's direct choice. Otherwise the bug can manifest itself as
4086 transient authentication and key negotiation failures or reproducible
4087 erroneous outcome of public-key operations with specially crafted input.
4088 Among EC algorithms only Brainpool P-512 curves are affected and one
4089 presumably can attack ECDH key negotiation. Impact was not analyzed in
4090 detail, because pre-requisites for attack are considered unlikely. Namely
4091 multiple clients have to choose the curve in question and the server has to
4092 share the private key among them, neither of which is default behaviour.
4093 Even then only clients that chose the curve will be affected.
4095 This issue was publicly reported as transient failures and was not
4096 initially recognized as a security issue. Thanks to Richard Morgan for
4097 providing reproducible case.
4102 * OpenSSL now fails if it receives an unrecognised record type in TLS1.0
4103 or TLS1.1. Previously this only happened in SSLv3 and TLS1.2. This is to
4104 prevent issues where no progress is being made and the peer continually
4105 sends unrecognised record types, using up resources processing them.
4109 ### Changes between 1.0.2i and 1.0.2j [26 Sep 2016]
4111 * Missing CRL sanity check
4113 A bug fix which included a CRL sanity check was added to OpenSSL 1.1.0
4114 but was omitted from OpenSSL 1.0.2i. As a result any attempt to use
4115 CRLs in OpenSSL 1.0.2i will crash with a null pointer exception.
4117 This issue only affects the OpenSSL 1.0.2i
4122 ### Changes between 1.0.2h and 1.0.2i [22 Sep 2016]
4124 * OCSP Status Request extension unbounded memory growth
4126 A malicious client can send an excessively large OCSP Status Request
4127 extension. If that client continually requests renegotiation, sending a
4128 large OCSP Status Request extension each time, then there will be unbounded
4129 memory growth on the server. This will eventually lead to a Denial Of
4130 Service attack through memory exhaustion. Servers with a default
4131 configuration are vulnerable even if they do not support OCSP. Builds using
4132 the "no-ocsp" build time option are not affected.
4134 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4139 * In order to mitigate the SWEET32 attack, the DES ciphers were moved from
4142 This issue was reported to OpenSSL Karthikeyan Bhargavan and Gaetan
4148 * OOB write in MDC2_Update()
4150 An overflow can occur in MDC2_Update() either if called directly or
4151 through the EVP_DigestUpdate() function using MDC2. If an attacker
4152 is able to supply very large amounts of input data after a previous
4153 call to EVP_EncryptUpdate() with a partial block then a length check
4154 can overflow resulting in a heap corruption.
4156 The amount of data needed is comparable to SIZE_MAX which is impractical
4159 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4164 * Malformed SHA512 ticket DoS
4166 If a server uses SHA512 for TLS session ticket HMAC it is vulnerable to a
4167 DoS attack where a malformed ticket will result in an OOB read which will
4170 The use of SHA512 in TLS session tickets is comparatively rare as it requires
4171 a custom server callback and ticket lookup mechanism.
4173 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4178 * OOB write in BN_bn2dec()
4180 The function BN_bn2dec() does not check the return value of BN_div_word().
4181 This can cause an OOB write if an application uses this function with an
4182 overly large BIGNUM. This could be a problem if an overly large certificate
4183 or CRL is printed out from an untrusted source. TLS is not affected because
4184 record limits will reject an oversized certificate before it is parsed.
4186 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4191 * OOB read in TS_OBJ_print_bio()
4193 The function TS_OBJ_print_bio() misuses OBJ_obj2txt(): the return value is
4194 the total length the OID text representation would use and not the amount
4195 of data written. This will result in OOB reads when large OIDs are
4198 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4203 * Pointer arithmetic undefined behaviour
4205 Avoid some undefined pointer arithmetic
4207 A common idiom in the codebase is to check limits in the following manner:
4210 Where "p" points to some malloc'd data of SIZE bytes and
4213 "len" here could be from some externally supplied data (e.g. from a TLS
4216 The rules of C pointer arithmetic are such that "p + len" is only well
4217 defined where len <= SIZE. Therefore the above idiom is actually
4218 undefined behaviour.
4220 For example this could cause problems if some malloc implementation
4221 provides an address for "p" such that "p + len" actually overflows for
4222 values of len that are too big and therefore p + len < limit.
4224 This issue was reported to OpenSSL by Guido Vranken
4229 * Constant time flag not preserved in DSA signing
4231 Operations in the DSA signing algorithm should run in constant time in
4232 order to avoid side channel attacks. A flaw in the OpenSSL DSA
4233 implementation means that a non-constant time codepath is followed for
4234 certain operations. This has been demonstrated through a cache-timing
4235 attack to be sufficient for an attacker to recover the private DSA key.
4237 This issue was reported by César Pereida (Aalto University), Billy Brumley
4238 (Tampere University of Technology), and Yuval Yarom (The University of
4239 Adelaide and NICTA).
4244 * DTLS buffered message DoS
4246 In a DTLS connection where handshake messages are delivered out-of-order
4247 those messages that OpenSSL is not yet ready to process will be buffered
4248 for later use. Under certain circumstances, a flaw in the logic means that
4249 those messages do not get removed from the buffer even though the handshake
4250 has been completed. An attacker could force up to approx. 15 messages to
4251 remain in the buffer when they are no longer required. These messages will
4252 be cleared when the DTLS connection is closed. The default maximum size for
4253 a message is 100k. Therefore the attacker could force an additional 1500k
4254 to be consumed per connection. By opening many simulataneous connections an
4255 attacker could cause a DoS attack through memory exhaustion.
4257 This issue was reported to OpenSSL by Quan Luo.
4262 * DTLS replay protection DoS
4264 A flaw in the DTLS replay attack protection mechanism means that records
4265 that arrive for future epochs update the replay protection "window" before
4266 the MAC for the record has been validated. This could be exploited by an
4267 attacker by sending a record for the next epoch (which does not have to
4268 decrypt or have a valid MAC), with a very large sequence number. This means
4269 that all subsequent legitimate packets are dropped causing a denial of
4270 service for a specific DTLS connection.
4272 This issue was reported to OpenSSL by the OCAP audit team.
4277 * Certificate message OOB reads
4279 In OpenSSL 1.0.2 and earlier some missing message length checks can result
4280 in OOB reads of up to 2 bytes beyond an allocated buffer. There is a
4281 theoretical DoS risk but this has not been observed in practice on common
4284 The messages affected are client certificate, client certificate request
4285 and server certificate. As a result the attack can only be performed
4286 against a client or a server which enables client authentication.
4288 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
4293 ### Changes between 1.0.2g and 1.0.2h [3 May 2016]
4295 * Prevent padding oracle in AES-NI CBC MAC check
4297 A MITM attacker can use a padding oracle attack to decrypt traffic
4298 when the connection uses an AES CBC cipher and the server support
4301 This issue was introduced as part of the fix for Lucky 13 padding
4302 attack [CVE-2013-0169][]. The padding check was rewritten to be in
4303 constant time by making sure that always the same bytes are read and
4304 compared against either the MAC or padding bytes. But it no longer
4305 checked that there was enough data to have both the MAC and padding
4308 This issue was reported by Juraj Somorovsky using TLS-Attacker.
4313 * Fix EVP_EncodeUpdate overflow
4315 An overflow can occur in the EVP_EncodeUpdate() function which is used for
4316 Base64 encoding of binary data. If an attacker is able to supply very large
4317 amounts of input data then a length check can overflow resulting in a heap
4320 Internally to OpenSSL the EVP_EncodeUpdate() function is primarily used by
4321 the `PEM_write_bio*` family of functions. These are mainly used within the
4322 OpenSSL command line applications, so any application which processes data
4323 from an untrusted source and outputs it as a PEM file should be considered
4324 vulnerable to this issue. User applications that call these APIs directly
4325 with large amounts of untrusted data may also be vulnerable.
4327 This issue was reported by Guido Vranken.
4332 * Fix EVP_EncryptUpdate overflow
4334 An overflow can occur in the EVP_EncryptUpdate() function. If an attacker
4335 is able to supply very large amounts of input data after a previous call to
4336 EVP_EncryptUpdate() with a partial block then a length check can overflow
4337 resulting in a heap corruption. Following an analysis of all OpenSSL
4338 internal usage of the EVP_EncryptUpdate() function all usage is one of two
4339 forms. The first form is where the EVP_EncryptUpdate() call is known to be
4340 the first called function after an EVP_EncryptInit(), and therefore that
4341 specific call must be safe. The second form is where the length passed to
4342 EVP_EncryptUpdate() can be seen from the code to be some small value and
4343 therefore there is no possibility of an overflow. Since all instances are
4344 one of these two forms, it is believed that there can be no overflows in
4345 internal code due to this problem. It should be noted that
4346 EVP_DecryptUpdate() can call EVP_EncryptUpdate() in certain code paths.
4347 Also EVP_CipherUpdate() is a synonym for EVP_EncryptUpdate(). All instances
4348 of these calls have also been analysed too and it is believed there are no
4349 instances in internal usage where an overflow could occur.
4351 This issue was reported by Guido Vranken.
4356 * Prevent ASN.1 BIO excessive memory allocation
4358 When ASN.1 data is read from a BIO using functions such as d2i_CMS_bio()
4359 a short invalid encoding can cause allocation of large amounts of memory
4360 potentially consuming excessive resources or exhausting memory.
4362 Any application parsing untrusted data through d2i BIO functions is
4363 affected. The memory based functions such as d2i_X509() are *not* affected.
4364 Since the memory based functions are used by the TLS library, TLS
4365 applications are not affected.
4367 This issue was reported by Brian Carpenter.
4374 ASN1 Strings that are over 1024 bytes can cause an overread in applications
4375 using the X509_NAME_oneline() function on EBCDIC systems. This could result
4376 in arbitrary stack data being returned in the buffer.
4378 This issue was reported by Guido Vranken.
4383 * Modify behavior of ALPN to invoke callback after SNI/servername
4384 callback, such that updates to the SSL_CTX affect ALPN.
4388 * Remove LOW from the DEFAULT cipher list. This removes singles DES from the
4393 * Only remove the SSLv2 methods with the no-ssl2-method option. When the
4394 methods are enabled and ssl2 is disabled the methods return NULL.
4398 ### Changes between 1.0.2f and 1.0.2g [1 Mar 2016]
4400 * Disable weak ciphers in SSLv3 and up in default builds of OpenSSL.
4401 Builds that are not configured with "enable-weak-ssl-ciphers" will not
4402 provide any "EXPORT" or "LOW" strength ciphers.
4406 * Disable SSLv2 default build, default negotiation and weak ciphers. SSLv2
4407 is by default disabled at build-time. Builds that are not configured with
4408 "enable-ssl2" will not support SSLv2. Even if "enable-ssl2" is used,
4409 users who want to negotiate SSLv2 via the version-flexible SSLv23_method()
4410 will need to explicitly call either of:
4412 SSL_CTX_clear_options(ctx, SSL_OP_NO_SSLv2);
4414 SSL_clear_options(ssl, SSL_OP_NO_SSLv2);
4416 as appropriate. Even if either of those is used, or the application
4417 explicitly uses the version-specific SSLv2_method() or its client and
4418 server variants, SSLv2 ciphers vulnerable to exhaustive search key
4419 recovery have been removed. Specifically, the SSLv2 40-bit EXPORT
4420 ciphers, and SSLv2 56-bit DES are no longer available.
4425 * Fix a double-free in DSA code
4427 A double free bug was discovered when OpenSSL parses malformed DSA private
4428 keys and could lead to a DoS attack or memory corruption for applications
4429 that receive DSA private keys from untrusted sources. This scenario is
4432 This issue was reported to OpenSSL by Adam Langley(Google/BoringSSL) using
4438 * Disable SRP fake user seed to address a server memory leak.
4440 Add a new method SRP_VBASE_get1_by_user that handles the seed properly.
4442 SRP_VBASE_get_by_user had inconsistent memory management behaviour.
4443 In order to fix an unavoidable memory leak, SRP_VBASE_get_by_user
4444 was changed to ignore the "fake user" SRP seed, even if the seed
4447 Users should use SRP_VBASE_get1_by_user instead. Note that in
4448 SRP_VBASE_get1_by_user, caller must free the returned value. Note
4449 also that even though configuring the SRP seed attempts to hide
4450 invalid usernames by continuing the handshake with fake
4451 credentials, this behaviour is not constant time and no strong
4452 guarantees are made that the handshake is indistinguishable from
4453 that of a valid user.
4458 * Fix BN_hex2bn/BN_dec2bn NULL pointer deref/heap corruption
4460 In the BN_hex2bn function the number of hex digits is calculated using an
4461 int value |i|. Later |bn_expand| is called with a value of |i * 4|. For
4462 large values of |i| this can result in |bn_expand| not allocating any
4463 memory because |i * 4| is negative. This can leave the internal BIGNUM data
4464 field as NULL leading to a subsequent NULL ptr deref. For very large values
4465 of |i|, the calculation |i * 4| could be a positive value smaller than |i|.
4466 In this case memory is allocated to the internal BIGNUM data field, but it
4467 is insufficiently sized leading to heap corruption. A similar issue exists
4468 in BN_dec2bn. This could have security consequences if BN_hex2bn/BN_dec2bn
4469 is ever called by user applications with very large untrusted hex/dec data.
4470 This is anticipated to be a rare occurrence.
4472 All OpenSSL internal usage of these functions use data that is not expected
4473 to be untrusted, e.g. config file data or application command line
4474 arguments. If user developed applications generate config file data based
4475 on untrusted data then it is possible that this could also lead to security
4476 consequences. This is also anticipated to be rare.
4478 This issue was reported to OpenSSL by Guido Vranken.
4483 * Fix memory issues in `BIO_*printf` functions
4485 The internal |fmtstr| function used in processing a "%s" format string in
4486 the `BIO_*printf` functions could overflow while calculating the length of a
4487 string and cause an OOB read when printing very long strings.
4489 Additionally the internal |doapr_outch| function can attempt to write to an
4490 OOB memory location (at an offset from the NULL pointer) in the event of a
4491 memory allocation failure. In 1.0.2 and below this could be caused where
4492 the size of a buffer to be allocated is greater than INT_MAX. E.g. this
4493 could be in processing a very long "%s" format string. Memory leaks can
4496 The first issue may mask the second issue dependent on compiler behaviour.
4497 These problems could enable attacks where large amounts of untrusted data
4498 is passed to the `BIO_*printf` functions. If applications use these functions
4499 in this way then they could be vulnerable. OpenSSL itself uses these
4500 functions when printing out human-readable dumps of ASN.1 data. Therefore
4501 applications that print this data could be vulnerable if the data is from
4502 untrusted sources. OpenSSL command line applications could also be
4503 vulnerable where they print out ASN.1 data, or if untrusted data is passed
4504 as command line arguments.
4506 Libssl is not considered directly vulnerable. Additionally certificates etc
4507 received via remote connections via libssl are also unlikely to be able to
4508 trigger these issues because of message size limits enforced within libssl.
4510 This issue was reported to OpenSSL Guido Vranken.
4515 * Side channel attack on modular exponentiation
4517 A side-channel attack was found which makes use of cache-bank conflicts on
4518 the Intel Sandy-Bridge microarchitecture which could lead to the recovery
4519 of RSA keys. The ability to exploit this issue is limited as it relies on
4520 an attacker who has control of code in a thread running on the same
4521 hyper-threaded core as the victim thread which is performing decryptions.
4523 This issue was reported to OpenSSL by Yuval Yarom, The University of
4524 Adelaide and NICTA, Daniel Genkin, Technion and Tel Aviv University, and
4525 Nadia Heninger, University of Pennsylvania with more information at
4526 <http://cachebleed.info>.
4531 * Change the req app to generate a 2048-bit RSA/DSA key by default,
4532 if no keysize is specified with default_bits. This fixes an
4533 omission in an earlier change that changed all RSA/DSA key generation
4534 apps to use 2048 bits by default.
4538 ### Changes between 1.0.2e and 1.0.2f [28 Jan 2016]
4540 * DH small subgroups
4542 Historically OpenSSL only ever generated DH parameters based on "safe"
4543 primes. More recently (in version 1.0.2) support was provided for
4544 generating X9.42 style parameter files such as those required for RFC 5114
4545 support. The primes used in such files may not be "safe". Where an
4546 application is using DH configured with parameters based on primes that are
4547 not "safe" then an attacker could use this fact to find a peer's private
4548 DH exponent. This attack requires that the attacker complete multiple
4549 handshakes in which the peer uses the same private DH exponent. For example
4550 this could be used to discover a TLS server's private DH exponent if it's
4551 reusing the private DH exponent or it's using a static DH ciphersuite.
4553 OpenSSL provides the option SSL_OP_SINGLE_DH_USE for ephemeral DH (DHE) in
4554 TLS. It is not on by default. If the option is not set then the server
4555 reuses the same private DH exponent for the life of the server process and
4556 would be vulnerable to this attack. It is believed that many popular
4557 applications do set this option and would therefore not be at risk.
4559 The fix for this issue adds an additional check where a "q" parameter is
4560 available (as is the case in X9.42 based parameters). This detects the
4561 only known attack, and is the only possible defense for static DH
4562 ciphersuites. This could have some performance impact.
4564 Additionally the SSL_OP_SINGLE_DH_USE option has been switched on by
4565 default and cannot be disabled. This could have some performance impact.
4567 This issue was reported to OpenSSL by Antonio Sanso (Adobe).
4572 * SSLv2 doesn't block disabled ciphers
4574 A malicious client can negotiate SSLv2 ciphers that have been disabled on
4575 the server and complete SSLv2 handshakes even if all SSLv2 ciphers have
4576 been disabled, provided that the SSLv2 protocol was not also disabled via
4579 This issue was reported to OpenSSL on 26th December 2015 by Nimrod Aviram
4580 and Sebastian Schinzel.
4585 ### Changes between 1.0.2d and 1.0.2e [3 Dec 2015]
4587 * BN_mod_exp may produce incorrect results on x86_64
4589 There is a carry propagating bug in the x86_64 Montgomery squaring
4590 procedure. No EC algorithms are affected. Analysis suggests that attacks
4591 against RSA and DSA as a result of this defect would be very difficult to
4592 perform and are not believed likely. Attacks against DH are considered just
4593 feasible (although very difficult) because most of the work necessary to
4594 deduce information about a private key may be performed offline. The amount
4595 of resources required for such an attack would be very significant and
4596 likely only accessible to a limited number of attackers. An attacker would
4597 additionally need online access to an unpatched system using the target
4598 private key in a scenario with persistent DH parameters and a private
4599 key that is shared between multiple clients. For example this can occur by
4600 default in OpenSSL DHE based SSL/TLS ciphersuites.
4602 This issue was reported to OpenSSL by Hanno Böck.
4607 * Certificate verify crash with missing PSS parameter
4609 The signature verification routines will crash with a NULL pointer
4610 dereference if presented with an ASN.1 signature using the RSA PSS
4611 algorithm and absent mask generation function parameter. Since these
4612 routines are used to verify certificate signature algorithms this can be
4613 used to crash any certificate verification operation and exploited in a
4614 DoS attack. Any application which performs certificate verification is
4615 vulnerable including OpenSSL clients and servers which enable client
4618 This issue was reported to OpenSSL by Loïc Jonas Etienne (Qnective AG).
4623 * X509_ATTRIBUTE memory leak
4625 When presented with a malformed X509_ATTRIBUTE structure OpenSSL will leak
4626 memory. This structure is used by the PKCS#7 and CMS routines so any
4627 application which reads PKCS#7 or CMS data from untrusted sources is
4628 affected. SSL/TLS is not affected.
4630 This issue was reported to OpenSSL by Adam Langley (Google/BoringSSL) using
4636 * Rewrite EVP_DecodeUpdate (base64 decoding) to fix several bugs.
4637 This changes the decoding behaviour for some invalid messages,
4638 though the change is mostly in the more lenient direction, and
4639 legacy behaviour is preserved as much as possible.
4643 * In DSA_generate_parameters_ex, if the provided seed is too short,
4646 *Rich Salz and Ismo Puustinen <ismo.puustinen@intel.com>*
4648 ### Changes between 1.0.2c and 1.0.2d [9 Jul 2015]
4650 * Alternate chains certificate forgery
4652 During certificate verification, OpenSSL will attempt to find an
4653 alternative certificate chain if the first attempt to build such a chain
4654 fails. An error in the implementation of this logic can mean that an
4655 attacker could cause certain checks on untrusted certificates to be
4656 bypassed, such as the CA flag, enabling them to use a valid leaf
4657 certificate to act as a CA and "issue" an invalid certificate.
4659 This issue was reported to OpenSSL by Adam Langley/David Benjamin
4664 ### Changes between 1.0.2b and 1.0.2c [12 Jun 2015]
4666 * Fix HMAC ABI incompatibility. The previous version introduced an ABI
4667 incompatibility in the handling of HMAC. The previous ABI has now been
4672 ### Changes between 1.0.2a and 1.0.2b [11 Jun 2015]
4674 * Malformed ECParameters causes infinite loop
4676 When processing an ECParameters structure OpenSSL enters an infinite loop
4677 if the curve specified is over a specially malformed binary polynomial
4680 This can be used to perform denial of service against any
4681 system which processes public keys, certificate requests or
4682 certificates. This includes TLS clients and TLS servers with
4683 client authentication enabled.
4685 This issue was reported to OpenSSL by Joseph Barr-Pixton.
4690 * Exploitable out-of-bounds read in X509_cmp_time
4692 X509_cmp_time does not properly check the length of the ASN1_TIME
4693 string and can read a few bytes out of bounds. In addition,
4694 X509_cmp_time accepts an arbitrary number of fractional seconds in the
4697 An attacker can use this to craft malformed certificates and CRLs of
4698 various sizes and potentially cause a segmentation fault, resulting in
4699 a DoS on applications that verify certificates or CRLs. TLS clients
4700 that verify CRLs are affected. TLS clients and servers with client
4701 authentication enabled may be affected if they use custom verification
4704 This issue was reported to OpenSSL by Robert Swiecki (Google), and
4705 independently by Hanno Böck.
4710 * PKCS7 crash with missing EnvelopedContent
4712 The PKCS#7 parsing code does not handle missing inner EncryptedContent
4713 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs
4714 with missing content and trigger a NULL pointer dereference on parsing.
4716 Applications that decrypt PKCS#7 data or otherwise parse PKCS#7
4717 structures from untrusted sources are affected. OpenSSL clients and
4718 servers are not affected.
4720 This issue was reported to OpenSSL by Michal Zalewski (Google).
4725 * CMS verify infinite loop with unknown hash function
4727 When verifying a signedData message the CMS code can enter an infinite loop
4728 if presented with an unknown hash function OID. This can be used to perform
4729 denial of service against any system which verifies signedData messages using
4731 This issue was reported to OpenSSL by Johannes Bauer.
4736 * Race condition handling NewSessionTicket
4738 If a NewSessionTicket is received by a multi-threaded client when attempting to
4739 reuse a previous ticket then a race condition can occur potentially leading to
4740 a double free of the ticket data.
4745 * Only support 256-bit or stronger elliptic curves with the
4746 'ecdh_auto' setting (server) or by default (client). Of supported
4747 curves, prefer P-256 (both).
4751 ### Changes between 1.0.2 and 1.0.2a [19 Mar 2015]
4753 * ClientHello sigalgs DoS fix
4755 If a client connects to an OpenSSL 1.0.2 server and renegotiates with an
4756 invalid signature algorithms extension a NULL pointer dereference will
4757 occur. This can be exploited in a DoS attack against the server.
4759 This issue was was reported to OpenSSL by David Ramos of Stanford
4763 *Stephen Henson and Matt Caswell*
4765 * Multiblock corrupted pointer fix
4767 OpenSSL 1.0.2 introduced the "multiblock" performance improvement. This
4768 feature only applies on 64 bit x86 architecture platforms that support AES
4769 NI instructions. A defect in the implementation of "multiblock" can cause
4770 OpenSSL's internal write buffer to become incorrectly set to NULL when
4771 using non-blocking IO. Typically, when the user application is using a
4772 socket BIO for writing, this will only result in a failed connection.
4773 However if some other BIO is used then it is likely that a segmentation
4774 fault will be triggered, thus enabling a potential DoS attack.
4776 This issue was reported to OpenSSL by Daniel Danner and Rainer Mueller.
4781 * Segmentation fault in DTLSv1_listen fix
4783 The DTLSv1_listen function is intended to be stateless and processes the
4784 initial ClientHello from many peers. It is common for user code to loop
4785 over the call to DTLSv1_listen until a valid ClientHello is received with
4786 an associated cookie. A defect in the implementation of DTLSv1_listen means
4787 that state is preserved in the SSL object from one invocation to the next
4788 that can lead to a segmentation fault. Errors processing the initial
4789 ClientHello can trigger this scenario. An example of such an error could be
4790 that a DTLS1.0 only client is attempting to connect to a DTLS1.2 only
4793 This issue was reported to OpenSSL by Per Allansson.
4798 * Segmentation fault in ASN1_TYPE_cmp fix
4800 The function ASN1_TYPE_cmp will crash with an invalid read if an attempt is
4801 made to compare ASN.1 boolean types. Since ASN1_TYPE_cmp is used to check
4802 certificate signature algorithm consistency this can be used to crash any
4803 certificate verification operation and exploited in a DoS attack. Any
4804 application which performs certificate verification is vulnerable including
4805 OpenSSL clients and servers which enable client authentication.
4810 * Segmentation fault for invalid PSS parameters fix
4812 The signature verification routines will crash with a NULL pointer
4813 dereference if presented with an ASN.1 signature using the RSA PSS
4814 algorithm and invalid parameters. Since these routines are used to verify
4815 certificate signature algorithms this can be used to crash any
4816 certificate verification operation and exploited in a DoS attack. Any
4817 application which performs certificate verification is vulnerable including
4818 OpenSSL clients and servers which enable client authentication.
4820 This issue was was reported to OpenSSL by Brian Carpenter.
4825 * ASN.1 structure reuse memory corruption fix
4827 Reusing a structure in ASN.1 parsing may allow an attacker to cause
4828 memory corruption via an invalid write. Such reuse is and has been
4829 strongly discouraged and is believed to be rare.
4831 Applications that parse structures containing CHOICE or ANY DEFINED BY
4832 components may be affected. Certificate parsing (d2i_X509 and related
4833 functions) are however not affected. OpenSSL clients and servers are
4839 * PKCS7 NULL pointer dereferences fix
4841 The PKCS#7 parsing code does not handle missing outer ContentInfo
4842 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs with
4843 missing content and trigger a NULL pointer dereference on parsing.
4845 Applications that verify PKCS#7 signatures, decrypt PKCS#7 data or
4846 otherwise parse PKCS#7 structures from untrusted sources are
4847 affected. OpenSSL clients and servers are not affected.
4849 This issue was reported to OpenSSL by Michal Zalewski (Google).
4854 * DoS via reachable assert in SSLv2 servers fix
4856 A malicious client can trigger an OPENSSL_assert (i.e., an abort) in
4857 servers that both support SSLv2 and enable export cipher suites by sending
4858 a specially crafted SSLv2 CLIENT-MASTER-KEY message.
4860 This issue was discovered by Sean Burford (Google) and Emilia Käsper
4861 (OpenSSL development team).
4866 * Empty CKE with client auth and DHE fix
4868 If client auth is used then a server can seg fault in the event of a DHE
4869 ciphersuite being selected and a zero length ClientKeyExchange message
4870 being sent by the client. This could be exploited in a DoS attack.
4875 * Handshake with unseeded PRNG fix
4877 Under certain conditions an OpenSSL 1.0.2 client can complete a handshake
4878 with an unseeded PRNG. The conditions are:
4879 - The client is on a platform where the PRNG has not been seeded
4880 automatically, and the user has not seeded manually
4881 - A protocol specific client method version has been used (i.e. not
4882 SSL_client_methodv23)
4883 - A ciphersuite is used that does not require additional random data from
4884 the PRNG beyond the initial ClientHello client random (e.g. PSK-RC4-SHA).
4886 If the handshake succeeds then the client random that has been used will
4887 have been generated from a PRNG with insufficient entropy and therefore the
4888 output may be predictable.
4890 For example using the following command with an unseeded openssl will
4891 succeed on an unpatched platform:
4893 openssl s_client -psk 1a2b3c4d -tls1_2 -cipher PSK-RC4-SHA
4898 * Use After Free following d2i_ECPrivatekey error fix
4900 A malformed EC private key file consumed via the d2i_ECPrivateKey function
4901 could cause a use after free condition. This, in turn, could cause a double
4902 free in several private key parsing functions (such as d2i_PrivateKey
4903 or EVP_PKCS82PKEY) and could lead to a DoS attack or memory corruption
4904 for applications that receive EC private keys from untrusted
4905 sources. This scenario is considered rare.
4907 This issue was discovered by the BoringSSL project and fixed in their
4913 * X509_to_X509_REQ NULL pointer deref fix
4915 The function X509_to_X509_REQ will crash with a NULL pointer dereference if
4916 the certificate key is invalid. This function is rarely used in practice.
4918 This issue was discovered by Brian Carpenter.
4923 * Removed the export ciphers from the DEFAULT ciphers
4927 ### Changes between 1.0.1l and 1.0.2 [22 Jan 2015]
4929 * Facilitate "universal" ARM builds targeting range of ARM ISAs, e.g.
4930 ARMv5 through ARMv8, as opposite to "locking" it to single one.
4931 So far those who have to target multiple platforms would compromise
4932 and argue that binary targeting say ARMv5 would still execute on
4933 ARMv8. "Universal" build resolves this compromise by providing
4934 near-optimal performance even on newer platforms.
4938 * Accelerated NIST P-256 elliptic curve implementation for x86_64
4939 (other platforms pending).
4941 *Shay Gueron & Vlad Krasnov (Intel Corp), Andy Polyakov*
4943 * Add support for the SignedCertificateTimestampList certificate and
4944 OCSP response extensions from RFC6962.
4948 * Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
4949 for corner cases. (Certain input points at infinity could lead to
4950 bogus results, with non-infinity inputs mapped to infinity too.)
4954 * Initial support for PowerISA 2.0.7, first implemented in POWER8.
4955 This covers AES, SHA256/512 and GHASH. "Initial" means that most
4956 common cases are optimized and there still is room for further
4957 improvements. Vector Permutation AES for Altivec is also added.
4961 * Add support for little-endian ppc64 Linux target.
4963 *Marcelo Cerri (IBM)*
4965 * Initial support for AMRv8 ISA crypto extensions. This covers AES,
4966 SHA1, SHA256 and GHASH. "Initial" means that most common cases
4967 are optimized and there still is room for further improvements.
4968 Both 32- and 64-bit modes are supported.
4970 *Andy Polyakov, Ard Biesheuvel (Linaro)*
4972 * Improved ARMv7 NEON support.
4976 * Support for SPARC Architecture 2011 crypto extensions, first
4977 implemented in SPARC T4. This covers AES, DES, Camellia, SHA1,
4978 SHA256/512, MD5, GHASH and modular exponentiation.
4980 *Andy Polyakov, David Miller*
4982 * Accelerated modular exponentiation for Intel processors, a.k.a.
4985 *Shay Gueron & Vlad Krasnov (Intel Corp)*
4987 * Support for new and upcoming Intel processors, including AVX2,
4988 BMI and SHA ISA extensions. This includes additional "stitched"
4989 implementations, AESNI-SHA256 and GCM, and multi-buffer support
4992 This work was sponsored by Intel Corp.
4996 * Support for DTLS 1.2. This adds two sets of DTLS methods: DTLS_*_method()
4997 supports both DTLS 1.2 and 1.0 and should use whatever version the peer
4998 supports and DTLSv1_2_*_method() which supports DTLS 1.2 only.
5002 * Use algorithm specific chains in SSL_CTX_use_certificate_chain_file():
5003 this fixes a limitation in previous versions of OpenSSL.
5007 * Extended RSA OAEP support via EVP_PKEY API. Options to specify digest,
5008 MGF1 digest and OAEP label.
5012 * Add EVP support for key wrapping algorithms, to avoid problems with
5013 existing code the flag EVP_CIPHER_CTX_WRAP_ALLOW has to be set in
5014 the EVP_CIPHER_CTX or an error is returned. Add AES and DES3 wrap
5015 algorithms and include tests cases.
5019 * Add functions to allocate and set the fields of an ECDSA_METHOD
5022 *Douglas E. Engert, Steve Henson*
5024 * New functions OPENSSL_gmtime_diff and ASN1_TIME_diff to find the
5025 difference in days and seconds between two tm or ASN1_TIME structures.
5029 * Add -rev test option to s_server to just reverse order of characters
5030 received by client and send back to server. Also prints an abbreviated
5031 summary of the connection parameters.
5035 * New option -brief for s_client and s_server to print out a brief summary
5036 of connection parameters.
5040 * Add callbacks for arbitrary TLS extensions.
5042 *Trevor Perrin <trevp@trevp.net> and Ben Laurie*
5044 * New option -crl_download in several openssl utilities to download CRLs
5045 from CRLDP extension in certificates.
5049 * New options -CRL and -CRLform for s_client and s_server for CRLs.
5053 * New function X509_CRL_diff to generate a delta CRL from the difference
5054 of two full CRLs. Add support to "crl" utility.
5058 * New functions to set lookup_crls function and to retrieve
5059 X509_STORE from X509_STORE_CTX.
5063 * Print out deprecated issuer and subject unique ID fields in
5068 * Extend OCSP I/O functions so they can be used for simple general purpose
5069 HTTP as well as OCSP. New wrapper function which can be used to download
5070 CRLs using the OCSP API.
5074 * Delegate command line handling in s_client/s_server to SSL_CONF APIs.
5078 * `SSL_CONF*` functions. These provide a common framework for application
5079 configuration using configuration files or command lines.
5083 * SSL/TLS tracing code. This parses out SSL/TLS records using the
5084 message callback and prints the results. Needs compile time option
5085 "enable-ssl-trace". New options to s_client and s_server to enable
5090 * New ctrl and macro to retrieve supported points extensions.
5091 Print out extension in s_server and s_client.
5095 * New functions to retrieve certificate signature and signature
5100 * Add functions to retrieve and manipulate the raw cipherlist sent by a
5105 * New Suite B modes for TLS code. These use and enforce the requirements
5106 of RFC6460: restrict ciphersuites, only permit Suite B algorithms and
5107 only use Suite B curves. The Suite B modes can be set by using the
5108 strings "SUITEB128", "SUITEB192" or "SUITEB128ONLY" for the cipherstring.
5112 * New chain verification flags for Suite B levels of security. Check
5113 algorithms are acceptable when flags are set in X509_verify_cert.
5117 * Make tls1_check_chain return a set of flags indicating checks passed
5118 by a certificate chain. Add additional tests to handle client
5119 certificates: checks for matching certificate type and issuer name
5124 * If an attempt is made to use a signature algorithm not in the peer
5125 preference list abort the handshake. If client has no suitable
5126 signature algorithms in response to a certificate request do not
5127 use the certificate.
5131 * If server EC tmp key is not in client preference list abort handshake.
5135 * Add support for certificate stores in CERT structure. This makes it
5136 possible to have different stores per SSL structure or one store in
5137 the parent SSL_CTX. Include distinct stores for certificate chain
5138 verification and chain building. New ctrl SSL_CTRL_BUILD_CERT_CHAIN
5139 to build and store a certificate chain in CERT structure: returning
5140 an error if the chain cannot be built: this will allow applications
5141 to test if a chain is correctly configured.
5143 Note: if the CERT based stores are not set then the parent SSL_CTX
5144 store is used to retain compatibility with existing behaviour.
5148 * New function ssl_set_client_disabled to set a ciphersuite disabled
5149 mask based on the current session, check mask when sending client
5150 hello and checking the requested ciphersuite.
5154 * New ctrls to retrieve and set certificate types in a certificate
5155 request message. Print out received values in s_client. If certificate
5156 types is not set with custom values set sensible values based on
5157 supported signature algorithms.
5161 * Support for distinct client and server supported signature algorithms.
5165 * Add certificate callback. If set this is called whenever a certificate
5166 is required by client or server. An application can decide which
5167 certificate chain to present based on arbitrary criteria: for example
5168 supported signature algorithms. Add very simple example to s_server.
5169 This fixes many of the problems and restrictions of the existing client
5170 certificate callback: for example you can now clear an existing
5171 certificate and specify the whole chain.
5175 * Add new "valid_flags" field to CERT_PKEY structure which determines what
5176 the certificate can be used for (if anything). Set valid_flags field
5177 in new tls1_check_chain function. Simplify ssl_set_cert_masks which used
5178 to have similar checks in it.
5180 Add new "cert_flags" field to CERT structure and include a "strict mode".
5181 This enforces some TLS certificate requirements (such as only permitting
5182 certificate signature algorithms contained in the supported algorithms
5183 extension) which some implementations ignore: this option should be used
5184 with caution as it could cause interoperability issues.
5188 * Update and tidy signature algorithm extension processing. Work out
5189 shared signature algorithms based on preferences and peer algorithms
5190 and print them out in s_client and s_server. Abort handshake if no
5191 shared signature algorithms.
5195 * Add new functions to allow customised supported signature algorithms
5196 for SSL and SSL_CTX structures. Add options to s_client and s_server
5201 * New function SSL_certs_clear() to delete all references to certificates
5202 from an SSL structure. Before this once a certificate had been added
5203 it couldn't be removed.
5207 * Integrate hostname, email address and IP address checking with certificate
5208 verification. New verify options supporting checking in openssl utility.
5212 * Fixes and wildcard matching support to hostname and email checking
5213 functions. Add manual page.
5215 *Florian Weimer (Red Hat Product Security Team)*
5217 * New functions to check a hostname email or IP address against a
5218 certificate. Add options x509 utility to print results of checks against
5223 * Fix OCSP checking.
5225 *Rob Stradling <rob.stradling@comodo.com> and Ben Laurie*
5227 * Initial experimental support for explicitly trusted non-root CAs.
5228 OpenSSL still tries to build a complete chain to a root but if an
5229 intermediate CA has a trust setting included that is used. The first
5230 setting is used: whether to trust (e.g., -addtrust option to the x509
5235 * Add -trusted_first option which attempts to find certificates in the
5236 trusted store even if an untrusted chain is also supplied.
5240 * MIPS assembly pack updates: support for MIPS32r2 and SmartMIPS ASE,
5241 platform support for Linux and Android.
5245 * Support for linux-x32, ILP32 environment in x86_64 framework.
5249 * Experimental multi-implementation support for FIPS capable OpenSSL.
5250 When in FIPS mode the approved implementations are used as normal,
5251 when not in FIPS mode the internal unapproved versions are used instead.
5252 This means that the FIPS capable OpenSSL isn't forced to use the
5253 (often lower performance) FIPS implementations outside FIPS mode.
5257 * Transparently support X9.42 DH parameters when calling
5258 PEM_read_bio_DHparameters. This means existing applications can handle
5259 the new parameter format automatically.
5263 * Initial experimental support for X9.42 DH parameter format: mainly
5264 to support use of 'q' parameter for RFC5114 parameters.
5268 * Add DH parameters from RFC5114 including test data to dhtest.
5272 * Support for automatic EC temporary key parameter selection. If enabled
5273 the most preferred EC parameters are automatically used instead of
5274 hardcoded fixed parameters. Now a server just has to call:
5275 SSL_CTX_set_ecdh_auto(ctx, 1) and the server will automatically
5276 support ECDH and use the most appropriate parameters.
5280 * Enhance and tidy EC curve and point format TLS extension code. Use
5281 static structures instead of allocation if default values are used.
5282 New ctrls to set curves we wish to support and to retrieve shared curves.
5283 Print out shared curves in s_server. New options to s_server and s_client
5284 to set list of supported curves.
5288 * New ctrls to retrieve supported signature algorithms and
5289 supported curve values as an array of NIDs. Extend openssl utility
5290 to print out received values.
5294 * Add new APIs EC_curve_nist2nid and EC_curve_nid2nist which convert
5295 between NIDs and the more common NIST names such as "P-256". Enhance
5296 ecparam utility and ECC method to recognise the NIST names for curves.
5300 * Enhance SSL/TLS certificate chain handling to support different
5301 chains for each certificate instead of one chain in the parent SSL_CTX.
5305 * Support for fixed DH ciphersuite client authentication: where both
5306 server and client use DH certificates with common parameters.
5310 * Support for fixed DH ciphersuites: those requiring DH server
5315 * New function i2d_re_X509_tbs for re-encoding the TBS portion of
5317 Note: Related 1.0.2-beta specific macros X509_get_cert_info,
5318 X509_CINF_set_modified, X509_CINF_get_issuer, X509_CINF_get_extensions and
5319 X509_CINF_get_signature were reverted post internal team review.
5324 ### Changes between 1.0.1t and 1.0.1u [22 Sep 2016]
5326 * OCSP Status Request extension unbounded memory growth
5328 A malicious client can send an excessively large OCSP Status Request
5329 extension. If that client continually requests renegotiation, sending a
5330 large OCSP Status Request extension each time, then there will be unbounded
5331 memory growth on the server. This will eventually lead to a Denial Of
5332 Service attack through memory exhaustion. Servers with a default
5333 configuration are vulnerable even if they do not support OCSP. Builds using
5334 the "no-ocsp" build time option are not affected.
5336 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5341 * In order to mitigate the SWEET32 attack, the DES ciphers were moved from
5344 This issue was reported to OpenSSL Karthikeyan Bhargavan and Gaetan
5350 * OOB write in MDC2_Update()
5352 An overflow can occur in MDC2_Update() either if called directly or
5353 through the EVP_DigestUpdate() function using MDC2. If an attacker
5354 is able to supply very large amounts of input data after a previous
5355 call to EVP_EncryptUpdate() with a partial block then a length check
5356 can overflow resulting in a heap corruption.
5358 The amount of data needed is comparable to SIZE_MAX which is impractical
5361 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5366 * Malformed SHA512 ticket DoS
5368 If a server uses SHA512 for TLS session ticket HMAC it is vulnerable to a
5369 DoS attack where a malformed ticket will result in an OOB read which will
5372 The use of SHA512 in TLS session tickets is comparatively rare as it requires
5373 a custom server callback and ticket lookup mechanism.
5375 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5380 * OOB write in BN_bn2dec()
5382 The function BN_bn2dec() does not check the return value of BN_div_word().
5383 This can cause an OOB write if an application uses this function with an
5384 overly large BIGNUM. This could be a problem if an overly large certificate
5385 or CRL is printed out from an untrusted source. TLS is not affected because
5386 record limits will reject an oversized certificate before it is parsed.
5388 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5393 * OOB read in TS_OBJ_print_bio()
5395 The function TS_OBJ_print_bio() misuses OBJ_obj2txt(): the return value is
5396 the total length the OID text representation would use and not the amount
5397 of data written. This will result in OOB reads when large OIDs are
5400 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5405 * Pointer arithmetic undefined behaviour
5407 Avoid some undefined pointer arithmetic
5409 A common idiom in the codebase is to check limits in the following manner:
5412 Where "p" points to some malloc'd data of SIZE bytes and
5415 "len" here could be from some externally supplied data (e.g. from a TLS
5418 The rules of C pointer arithmetic are such that "p + len" is only well
5419 defined where len <= SIZE. Therefore the above idiom is actually
5420 undefined behaviour.
5422 For example this could cause problems if some malloc implementation
5423 provides an address for "p" such that "p + len" actually overflows for
5424 values of len that are too big and therefore p + len < limit.
5426 This issue was reported to OpenSSL by Guido Vranken
5431 * Constant time flag not preserved in DSA signing
5433 Operations in the DSA signing algorithm should run in constant time in
5434 order to avoid side channel attacks. A flaw in the OpenSSL DSA
5435 implementation means that a non-constant time codepath is followed for
5436 certain operations. This has been demonstrated through a cache-timing
5437 attack to be sufficient for an attacker to recover the private DSA key.
5439 This issue was reported by César Pereida (Aalto University), Billy Brumley
5440 (Tampere University of Technology), and Yuval Yarom (The University of
5441 Adelaide and NICTA).
5446 * DTLS buffered message DoS
5448 In a DTLS connection where handshake messages are delivered out-of-order
5449 those messages that OpenSSL is not yet ready to process will be buffered
5450 for later use. Under certain circumstances, a flaw in the logic means that
5451 those messages do not get removed from the buffer even though the handshake
5452 has been completed. An attacker could force up to approx. 15 messages to
5453 remain in the buffer when they are no longer required. These messages will
5454 be cleared when the DTLS connection is closed. The default maximum size for
5455 a message is 100k. Therefore the attacker could force an additional 1500k
5456 to be consumed per connection. By opening many simulataneous connections an
5457 attacker could cause a DoS attack through memory exhaustion.
5459 This issue was reported to OpenSSL by Quan Luo.
5464 * DTLS replay protection DoS
5466 A flaw in the DTLS replay attack protection mechanism means that records
5467 that arrive for future epochs update the replay protection "window" before
5468 the MAC for the record has been validated. This could be exploited by an
5469 attacker by sending a record for the next epoch (which does not have to
5470 decrypt or have a valid MAC), with a very large sequence number. This means
5471 that all subsequent legitimate packets are dropped causing a denial of
5472 service for a specific DTLS connection.
5474 This issue was reported to OpenSSL by the OCAP audit team.
5479 * Certificate message OOB reads
5481 In OpenSSL 1.0.2 and earlier some missing message length checks can result
5482 in OOB reads of up to 2 bytes beyond an allocated buffer. There is a
5483 theoretical DoS risk but this has not been observed in practice on common
5486 The messages affected are client certificate, client certificate request
5487 and server certificate. As a result the attack can only be performed
5488 against a client or a server which enables client authentication.
5490 This issue was reported to OpenSSL by Shi Lei (Gear Team, Qihoo 360 Inc.)
5495 ### Changes between 1.0.1s and 1.0.1t [3 May 2016]
5497 * Prevent padding oracle in AES-NI CBC MAC check
5499 A MITM attacker can use a padding oracle attack to decrypt traffic
5500 when the connection uses an AES CBC cipher and the server support
5503 This issue was introduced as part of the fix for Lucky 13 padding
5504 attack [CVE-2013-0169][]. The padding check was rewritten to be in
5505 constant time by making sure that always the same bytes are read and
5506 compared against either the MAC or padding bytes. But it no longer
5507 checked that there was enough data to have both the MAC and padding
5510 This issue was reported by Juraj Somorovsky using TLS-Attacker.
5515 * Fix EVP_EncodeUpdate overflow
5517 An overflow can occur in the EVP_EncodeUpdate() function which is used for
5518 Base64 encoding of binary data. If an attacker is able to supply very large
5519 amounts of input data then a length check can overflow resulting in a heap
5522 Internally to OpenSSL the EVP_EncodeUpdate() function is primarly used by
5523 the `PEM_write_bio*` family of functions. These are mainly used within the
5524 OpenSSL command line applications, so any application which processes data
5525 from an untrusted source and outputs it as a PEM file should be considered
5526 vulnerable to this issue. User applications that call these APIs directly
5527 with large amounts of untrusted data may also be vulnerable.
5529 This issue was reported by Guido Vranken.
5534 * Fix EVP_EncryptUpdate overflow
5536 An overflow can occur in the EVP_EncryptUpdate() function. If an attacker
5537 is able to supply very large amounts of input data after a previous call to
5538 EVP_EncryptUpdate() with a partial block then a length check can overflow
5539 resulting in a heap corruption. Following an analysis of all OpenSSL
5540 internal usage of the EVP_EncryptUpdate() function all usage is one of two
5541 forms. The first form is where the EVP_EncryptUpdate() call is known to be
5542 the first called function after an EVP_EncryptInit(), and therefore that
5543 specific call must be safe. The second form is where the length passed to
5544 EVP_EncryptUpdate() can be seen from the code to be some small value and
5545 therefore there is no possibility of an overflow. Since all instances are
5546 one of these two forms, it is believed that there can be no overflows in
5547 internal code due to this problem. It should be noted that
5548 EVP_DecryptUpdate() can call EVP_EncryptUpdate() in certain code paths.
5549 Also EVP_CipherUpdate() is a synonym for EVP_EncryptUpdate(). All instances
5550 of these calls have also been analysed too and it is believed there are no
5551 instances in internal usage where an overflow could occur.
5553 This issue was reported by Guido Vranken.
5558 * Prevent ASN.1 BIO excessive memory allocation
5560 When ASN.1 data is read from a BIO using functions such as d2i_CMS_bio()
5561 a short invalid encoding can casuse allocation of large amounts of memory
5562 potentially consuming excessive resources or exhausting memory.
5564 Any application parsing untrusted data through d2i BIO functions is
5565 affected. The memory based functions such as d2i_X509() are *not* affected.
5566 Since the memory based functions are used by the TLS library, TLS
5567 applications are not affected.
5569 This issue was reported by Brian Carpenter.
5576 ASN1 Strings that are over 1024 bytes can cause an overread in applications
5577 using the X509_NAME_oneline() function on EBCDIC systems. This could result
5578 in arbitrary stack data being returned in the buffer.
5580 This issue was reported by Guido Vranken.
5585 * Modify behavior of ALPN to invoke callback after SNI/servername
5586 callback, such that updates to the SSL_CTX affect ALPN.
5590 * Remove LOW from the DEFAULT cipher list. This removes singles DES from the
5595 * Only remove the SSLv2 methods with the no-ssl2-method option. When the
5596 methods are enabled and ssl2 is disabled the methods return NULL.
5600 ### Changes between 1.0.1r and 1.0.1s [1 Mar 2016]
5602 * Disable weak ciphers in SSLv3 and up in default builds of OpenSSL.
5603 Builds that are not configured with "enable-weak-ssl-ciphers" will not
5604 provide any "EXPORT" or "LOW" strength ciphers.
5608 * Disable SSLv2 default build, default negotiation and weak ciphers. SSLv2
5609 is by default disabled at build-time. Builds that are not configured with
5610 "enable-ssl2" will not support SSLv2. Even if "enable-ssl2" is used,
5611 users who want to negotiate SSLv2 via the version-flexible SSLv23_method()
5612 will need to explicitly call either of:
5614 SSL_CTX_clear_options(ctx, SSL_OP_NO_SSLv2);
5616 SSL_clear_options(ssl, SSL_OP_NO_SSLv2);
5618 as appropriate. Even if either of those is used, or the application
5619 explicitly uses the version-specific SSLv2_method() or its client and
5620 server variants, SSLv2 ciphers vulnerable to exhaustive search key
5621 recovery have been removed. Specifically, the SSLv2 40-bit EXPORT
5622 ciphers, and SSLv2 56-bit DES are no longer available.
5627 * Fix a double-free in DSA code
5629 A double free bug was discovered when OpenSSL parses malformed DSA private
5630 keys and could lead to a DoS attack or memory corruption for applications
5631 that receive DSA private keys from untrusted sources. This scenario is
5634 This issue was reported to OpenSSL by Adam Langley(Google/BoringSSL) using
5640 * Disable SRP fake user seed to address a server memory leak.
5642 Add a new method SRP_VBASE_get1_by_user that handles the seed properly.
5644 SRP_VBASE_get_by_user had inconsistent memory management behaviour.
5645 In order to fix an unavoidable memory leak, SRP_VBASE_get_by_user
5646 was changed to ignore the "fake user" SRP seed, even if the seed
5649 Users should use SRP_VBASE_get1_by_user instead. Note that in
5650 SRP_VBASE_get1_by_user, caller must free the returned value. Note
5651 also that even though configuring the SRP seed attempts to hide
5652 invalid usernames by continuing the handshake with fake
5653 credentials, this behaviour is not constant time and no strong
5654 guarantees are made that the handshake is indistinguishable from
5655 that of a valid user.
5660 * Fix BN_hex2bn/BN_dec2bn NULL pointer deref/heap corruption
5662 In the BN_hex2bn function the number of hex digits is calculated using an
5663 int value |i|. Later |bn_expand| is called with a value of |i * 4|. For
5664 large values of |i| this can result in |bn_expand| not allocating any
5665 memory because |i * 4| is negative. This can leave the internal BIGNUM data
5666 field as NULL leading to a subsequent NULL ptr deref. For very large values
5667 of |i|, the calculation |i * 4| could be a positive value smaller than |i|.
5668 In this case memory is allocated to the internal BIGNUM data field, but it
5669 is insufficiently sized leading to heap corruption. A similar issue exists
5670 in BN_dec2bn. This could have security consequences if BN_hex2bn/BN_dec2bn
5671 is ever called by user applications with very large untrusted hex/dec data.
5672 This is anticipated to be a rare occurrence.
5674 All OpenSSL internal usage of these functions use data that is not expected
5675 to be untrusted, e.g. config file data or application command line
5676 arguments. If user developed applications generate config file data based
5677 on untrusted data then it is possible that this could also lead to security
5678 consequences. This is also anticipated to be rare.
5680 This issue was reported to OpenSSL by Guido Vranken.
5685 * Fix memory issues in `BIO_*printf` functions
5687 The internal |fmtstr| function used in processing a "%s" format string in
5688 the `BIO_*printf` functions could overflow while calculating the length of a
5689 string and cause an OOB read when printing very long strings.
5691 Additionally the internal |doapr_outch| function can attempt to write to an
5692 OOB memory location (at an offset from the NULL pointer) in the event of a
5693 memory allocation failure. In 1.0.2 and below this could be caused where
5694 the size of a buffer to be allocated is greater than INT_MAX. E.g. this
5695 could be in processing a very long "%s" format string. Memory leaks can
5698 The first issue may mask the second issue dependent on compiler behaviour.
5699 These problems could enable attacks where large amounts of untrusted data
5700 is passed to the `BIO_*printf` functions. If applications use these functions
5701 in this way then they could be vulnerable. OpenSSL itself uses these
5702 functions when printing out human-readable dumps of ASN.1 data. Therefore
5703 applications that print this data could be vulnerable if the data is from
5704 untrusted sources. OpenSSL command line applications could also be
5705 vulnerable where they print out ASN.1 data, or if untrusted data is passed
5706 as command line arguments.
5708 Libssl is not considered directly vulnerable. Additionally certificates etc
5709 received via remote connections via libssl are also unlikely to be able to
5710 trigger these issues because of message size limits enforced within libssl.
5712 This issue was reported to OpenSSL Guido Vranken.
5717 * Side channel attack on modular exponentiation
5719 A side-channel attack was found which makes use of cache-bank conflicts on
5720 the Intel Sandy-Bridge microarchitecture which could lead to the recovery
5721 of RSA keys. The ability to exploit this issue is limited as it relies on
5722 an attacker who has control of code in a thread running on the same
5723 hyper-threaded core as the victim thread which is performing decryptions.
5725 This issue was reported to OpenSSL by Yuval Yarom, The University of
5726 Adelaide and NICTA, Daniel Genkin, Technion and Tel Aviv University, and
5727 Nadia Heninger, University of Pennsylvania with more information at
5728 <http://cachebleed.info>.
5733 * Change the req app to generate a 2048-bit RSA/DSA key by default,
5734 if no keysize is specified with default_bits. This fixes an
5735 omission in an earlier change that changed all RSA/DSA key generation
5736 apps to use 2048 bits by default.
5740 ### Changes between 1.0.1q and 1.0.1r [28 Jan 2016]
5742 * Protection for DH small subgroup attacks
5744 As a precautionary measure the SSL_OP_SINGLE_DH_USE option has been
5745 switched on by default and cannot be disabled. This could have some
5750 * SSLv2 doesn't block disabled ciphers
5752 A malicious client can negotiate SSLv2 ciphers that have been disabled on
5753 the server and complete SSLv2 handshakes even if all SSLv2 ciphers have
5754 been disabled, provided that the SSLv2 protocol was not also disabled via
5757 This issue was reported to OpenSSL on 26th December 2015 by Nimrod Aviram
5758 and Sebastian Schinzel.
5763 * Reject DH handshakes with parameters shorter than 1024 bits.
5767 ### Changes between 1.0.1p and 1.0.1q [3 Dec 2015]
5769 * Certificate verify crash with missing PSS parameter
5771 The signature verification routines will crash with a NULL pointer
5772 dereference if presented with an ASN.1 signature using the RSA PSS
5773 algorithm and absent mask generation function parameter. Since these
5774 routines are used to verify certificate signature algorithms this can be
5775 used to crash any certificate verification operation and exploited in a
5776 DoS attack. Any application which performs certificate verification is
5777 vulnerable including OpenSSL clients and servers which enable client
5780 This issue was reported to OpenSSL by Loïc Jonas Etienne (Qnective AG).
5785 * X509_ATTRIBUTE memory leak
5787 When presented with a malformed X509_ATTRIBUTE structure OpenSSL will leak
5788 memory. This structure is used by the PKCS#7 and CMS routines so any
5789 application which reads PKCS#7 or CMS data from untrusted sources is
5790 affected. SSL/TLS is not affected.
5792 This issue was reported to OpenSSL by Adam Langley (Google/BoringSSL) using
5798 * Rewrite EVP_DecodeUpdate (base64 decoding) to fix several bugs.
5799 This changes the decoding behaviour for some invalid messages,
5800 though the change is mostly in the more lenient direction, and
5801 legacy behaviour is preserved as much as possible.
5805 * In DSA_generate_parameters_ex, if the provided seed is too short,
5806 use a random seed, as already documented.
5808 *Rich Salz and Ismo Puustinen <ismo.puustinen@intel.com>*
5810 ### Changes between 1.0.1o and 1.0.1p [9 Jul 2015]
5812 * Alternate chains certificate forgery
5814 During certificate verfification, OpenSSL will attempt to find an
5815 alternative certificate chain if the first attempt to build such a chain
5816 fails. An error in the implementation of this logic can mean that an
5817 attacker could cause certain checks on untrusted certificates to be
5818 bypassed, such as the CA flag, enabling them to use a valid leaf
5819 certificate to act as a CA and "issue" an invalid certificate.
5821 This issue was reported to OpenSSL by Adam Langley/David Benjamin
5827 * Race condition handling PSK identify hint
5829 If PSK identity hints are received by a multi-threaded client then
5830 the values are wrongly updated in the parent SSL_CTX structure. This can
5831 result in a race condition potentially leading to a double free of the
5837 ### Changes between 1.0.1n and 1.0.1o [12 Jun 2015]
5839 * Fix HMAC ABI incompatibility. The previous version introduced an ABI
5840 incompatibility in the handling of HMAC. The previous ABI has now been
5843 ### Changes between 1.0.1m and 1.0.1n [11 Jun 2015]
5845 * Malformed ECParameters causes infinite loop
5847 When processing an ECParameters structure OpenSSL enters an infinite loop
5848 if the curve specified is over a specially malformed binary polynomial
5851 This can be used to perform denial of service against any
5852 system which processes public keys, certificate requests or
5853 certificates. This includes TLS clients and TLS servers with
5854 client authentication enabled.
5856 This issue was reported to OpenSSL by Joseph Barr-Pixton.
5861 * Exploitable out-of-bounds read in X509_cmp_time
5863 X509_cmp_time does not properly check the length of the ASN1_TIME
5864 string and can read a few bytes out of bounds. In addition,
5865 X509_cmp_time accepts an arbitrary number of fractional seconds in the
5868 An attacker can use this to craft malformed certificates and CRLs of
5869 various sizes and potentially cause a segmentation fault, resulting in
5870 a DoS on applications that verify certificates or CRLs. TLS clients
5871 that verify CRLs are affected. TLS clients and servers with client
5872 authentication enabled may be affected if they use custom verification
5875 This issue was reported to OpenSSL by Robert Swiecki (Google), and
5876 independently by Hanno Böck.
5881 * PKCS7 crash with missing EnvelopedContent
5883 The PKCS#7 parsing code does not handle missing inner EncryptedContent
5884 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs
5885 with missing content and trigger a NULL pointer dereference on parsing.
5887 Applications that decrypt PKCS#7 data or otherwise parse PKCS#7
5888 structures from untrusted sources are affected. OpenSSL clients and
5889 servers are not affected.
5891 This issue was reported to OpenSSL by Michal Zalewski (Google).
5896 * CMS verify infinite loop with unknown hash function
5898 When verifying a signedData message the CMS code can enter an infinite loop
5899 if presented with an unknown hash function OID. This can be used to perform
5900 denial of service against any system which verifies signedData messages using
5902 This issue was reported to OpenSSL by Johannes Bauer.
5907 * Race condition handling NewSessionTicket
5909 If a NewSessionTicket is received by a multi-threaded client when attempting to
5910 reuse a previous ticket then a race condition can occur potentially leading to
5911 a double free of the ticket data.
5916 * Reject DH handshakes with parameters shorter than 768 bits.
5918 *Kurt Roeckx and Emilia Kasper*
5920 * dhparam: generate 2048-bit parameters by default.
5922 *Kurt Roeckx and Emilia Kasper*
5924 ### Changes between 1.0.1l and 1.0.1m [19 Mar 2015]
5926 * Segmentation fault in ASN1_TYPE_cmp fix
5928 The function ASN1_TYPE_cmp will crash with an invalid read if an attempt is
5929 made to compare ASN.1 boolean types. Since ASN1_TYPE_cmp is used to check
5930 certificate signature algorithm consistency this can be used to crash any
5931 certificate verification operation and exploited in a DoS attack. Any
5932 application which performs certificate verification is vulnerable including
5933 OpenSSL clients and servers which enable client authentication.
5938 * ASN.1 structure reuse memory corruption fix
5940 Reusing a structure in ASN.1 parsing may allow an attacker to cause
5941 memory corruption via an invalid write. Such reuse is and has been
5942 strongly discouraged and is believed to be rare.
5944 Applications that parse structures containing CHOICE or ANY DEFINED BY
5945 components may be affected. Certificate parsing (d2i_X509 and related
5946 functions) are however not affected. OpenSSL clients and servers are
5952 * PKCS7 NULL pointer dereferences fix
5954 The PKCS#7 parsing code does not handle missing outer ContentInfo
5955 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs with
5956 missing content and trigger a NULL pointer dereference on parsing.
5958 Applications that verify PKCS#7 signatures, decrypt PKCS#7 data or
5959 otherwise parse PKCS#7 structures from untrusted sources are
5960 affected. OpenSSL clients and servers are not affected.
5962 This issue was reported to OpenSSL by Michal Zalewski (Google).
5967 * DoS via reachable assert in SSLv2 servers fix
5969 A malicious client can trigger an OPENSSL_assert (i.e., an abort) in
5970 servers that both support SSLv2 and enable export cipher suites by sending
5971 a specially crafted SSLv2 CLIENT-MASTER-KEY message.
5973 This issue was discovered by Sean Burford (Google) and Emilia Käsper
5974 (OpenSSL development team).
5979 * Use After Free following d2i_ECPrivatekey error fix
5981 A malformed EC private key file consumed via the d2i_ECPrivateKey function
5982 could cause a use after free condition. This, in turn, could cause a double
5983 free in several private key parsing functions (such as d2i_PrivateKey
5984 or EVP_PKCS82PKEY) and could lead to a DoS attack or memory corruption
5985 for applications that receive EC private keys from untrusted
5986 sources. This scenario is considered rare.
5988 This issue was discovered by the BoringSSL project and fixed in their
5994 * X509_to_X509_REQ NULL pointer deref fix
5996 The function X509_to_X509_REQ will crash with a NULL pointer dereference if
5997 the certificate key is invalid. This function is rarely used in practice.
5999 This issue was discovered by Brian Carpenter.
6004 * Removed the export ciphers from the DEFAULT ciphers
6008 ### Changes between 1.0.1k and 1.0.1l [15 Jan 2015]
6010 * Build fixes for the Windows and OpenVMS platforms
6012 *Matt Caswell and Richard Levitte*
6014 ### Changes between 1.0.1j and 1.0.1k [8 Jan 2015]
6016 * Fix DTLS segmentation fault in dtls1_get_record. A carefully crafted DTLS
6017 message can cause a segmentation fault in OpenSSL due to a NULL pointer
6018 dereference. This could lead to a Denial Of Service attack. Thanks to
6019 Markus Stenberg of Cisco Systems, Inc. for reporting this issue.
6024 * Fix DTLS memory leak in dtls1_buffer_record. A memory leak can occur in the
6025 dtls1_buffer_record function under certain conditions. In particular this
6026 could occur if an attacker sent repeated DTLS records with the same
6027 sequence number but for the next epoch. The memory leak could be exploited
6028 by an attacker in a Denial of Service attack through memory exhaustion.
6029 Thanks to Chris Mueller for reporting this issue.
6034 * Fix issue where no-ssl3 configuration sets method to NULL. When openssl is
6035 built with the no-ssl3 option and a SSL v3 ClientHello is received the ssl
6036 method would be set to NULL which could later result in a NULL pointer
6037 dereference. Thanks to Frank Schmirler for reporting this issue.
6042 * Abort handshake if server key exchange message is omitted for ephemeral
6045 Thanks to Karthikeyan Bhargavan of the PROSECCO team at INRIA for
6046 reporting this issue.
6051 * Remove non-export ephemeral RSA code on client and server. This code
6052 violated the TLS standard by allowing the use of temporary RSA keys in
6053 non-export ciphersuites and could be used by a server to effectively
6054 downgrade the RSA key length used to a value smaller than the server
6055 certificate. Thanks for Karthikeyan Bhargavan of the PROSECCO team at
6056 INRIA or reporting this issue.
6061 * Fixed issue where DH client certificates are accepted without verification.
6062 An OpenSSL server will accept a DH certificate for client authentication
6063 without the certificate verify message. This effectively allows a client to
6064 authenticate without the use of a private key. This only affects servers
6065 which trust a client certificate authority which issues certificates
6066 containing DH keys: these are extremely rare and hardly ever encountered.
6067 Thanks for Karthikeyan Bhargavan of the PROSECCO team at INRIA or reporting
6073 * Ensure that the session ID context of an SSL is updated when its
6074 SSL_CTX is updated via SSL_set_SSL_CTX.
6076 The session ID context is typically set from the parent SSL_CTX,
6077 and can vary with the CTX.
6081 * Fix various certificate fingerprint issues.
6083 By using non-DER or invalid encodings outside the signed portion of a
6084 certificate the fingerprint can be changed without breaking the signature.
6085 Although no details of the signed portion of the certificate can be changed
6086 this can cause problems with some applications: e.g. those using the
6087 certificate fingerprint for blacklists.
6089 1. Reject signatures with non zero unused bits.
6091 If the BIT STRING containing the signature has non zero unused bits reject
6092 the signature. All current signature algorithms require zero unused bits.
6094 2. Check certificate algorithm consistency.
6096 Check the AlgorithmIdentifier inside TBS matches the one in the
6097 certificate signature. NB: this will result in signature failure
6098 errors for some broken certificates.
6100 Thanks to Konrad Kraszewski from Google for reporting this issue.
6102 3. Check DSA/ECDSA signatures use DER.
6104 Re-encode DSA/ECDSA signatures and compare with the original received
6105 signature. Return an error if there is a mismatch.
6107 This will reject various cases including garbage after signature
6108 (thanks to Antti Karjalainen and Tuomo Untinen from the Codenomicon CROSS
6109 program for discovering this case) and use of BER or invalid ASN.1 INTEGERs
6110 (negative or with leading zeroes).
6112 Further analysis was conducted and fixes were developed by Stephen Henson
6113 of the OpenSSL core team.
6119 * Correct Bignum squaring. Bignum squaring (BN_sqr) may produce incorrect
6120 results on some platforms, including x86_64. This bug occurs at random
6121 with a very low probability, and is not known to be exploitable in any
6122 way, though its exact impact is difficult to determine. Thanks to Pieter
6123 Wuille (Blockstream) who reported this issue and also suggested an initial
6124 fix. Further analysis was conducted by the OpenSSL development team and
6125 Adam Langley of Google. The final fix was developed by Andy Polyakov of
6126 the OpenSSL core team.
6131 * Do not resume sessions on the server if the negotiated protocol
6132 version does not match the session's version. Resuming with a different
6133 version, while not strictly forbidden by the RFC, is of questionable
6134 sanity and breaks all known clients.
6136 *David Benjamin, Emilia Käsper*
6138 * Tighten handling of the ChangeCipherSpec (CCS) message: reject
6139 early CCS messages during renegotiation. (Note that because
6140 renegotiation is encrypted, this early CCS was not exploitable.)
6144 * Tighten client-side session ticket handling during renegotiation:
6145 ensure that the client only accepts a session ticket if the server sends
6146 the extension anew in the ServerHello. Previously, a TLS client would
6147 reuse the old extension state and thus accept a session ticket if one was
6148 announced in the initial ServerHello.
6150 Similarly, ensure that the client requires a session ticket if one
6151 was advertised in the ServerHello. Previously, a TLS client would
6152 ignore a missing NewSessionTicket message.
6156 ### Changes between 1.0.1i and 1.0.1j [15 Oct 2014]
6160 A flaw in the DTLS SRTP extension parsing code allows an attacker, who
6161 sends a carefully crafted handshake message, to cause OpenSSL to fail
6162 to free up to 64k of memory causing a memory leak. This could be
6163 exploited in a Denial Of Service attack. This issue affects OpenSSL
6164 1.0.1 server implementations for both SSL/TLS and DTLS regardless of
6165 whether SRTP is used or configured. Implementations of OpenSSL that
6166 have been compiled with OPENSSL_NO_SRTP defined are not affected.
6168 The fix was developed by the OpenSSL team.
6173 * Session Ticket Memory Leak.
6175 When an OpenSSL SSL/TLS/DTLS server receives a session ticket the
6176 integrity of that ticket is first verified. In the event of a session
6177 ticket integrity check failing, OpenSSL will fail to free memory
6178 causing a memory leak. By sending a large number of invalid session
6179 tickets an attacker could exploit this issue in a Denial Of Service
6185 * Build option no-ssl3 is incomplete.
6187 When OpenSSL is configured with "no-ssl3" as a build option, servers
6188 could accept and complete a SSL 3.0 handshake, and clients could be
6189 configured to send them.
6192 *Akamai and the OpenSSL team*
6194 * Add support for TLS_FALLBACK_SCSV.
6195 Client applications doing fallback retries should call
6196 SSL_set_mode(s, SSL_MODE_SEND_FALLBACK_SCSV).
6199 *Adam Langley, Bodo Moeller*
6201 * Add additional DigestInfo checks.
6203 Re-encode DigestInto in DER and check against the original when
6204 verifying RSA signature: this will reject any improperly encoded
6205 DigestInfo structures.
6207 Note: this is a precautionary measure and no attacks are currently known.
6211 ### Changes between 1.0.1h and 1.0.1i [6 Aug 2014]
6213 * Fix SRP buffer overrun vulnerability. Invalid parameters passed to the
6214 SRP code can be overrun an internal buffer. Add sanity check that
6215 g, A, B < N to SRP code.
6217 Thanks to Sean Devlin and Watson Ladd of Cryptography Services, NCC
6218 Group for discovering this issue.
6223 * A flaw in the OpenSSL SSL/TLS server code causes the server to negotiate
6224 TLS 1.0 instead of higher protocol versions when the ClientHello message
6225 is badly fragmented. This allows a man-in-the-middle attacker to force a
6226 downgrade to TLS 1.0 even if both the server and the client support a
6227 higher protocol version, by modifying the client's TLS records.
6229 Thanks to David Benjamin and Adam Langley (Google) for discovering and
6230 researching this issue.
6235 * OpenSSL DTLS clients enabling anonymous (EC)DH ciphersuites are subject
6236 to a denial of service attack. A malicious server can crash the client
6237 with a null pointer dereference (read) by specifying an anonymous (EC)DH
6238 ciphersuite and sending carefully crafted handshake messages.
6240 Thanks to Felix Gröbert (Google) for discovering and researching this
6246 * By sending carefully crafted DTLS packets an attacker could cause openssl
6247 to leak memory. This can be exploited through a Denial of Service attack.
6248 Thanks to Adam Langley for discovering and researching this issue.
6253 * An attacker can force openssl to consume large amounts of memory whilst
6254 processing DTLS handshake messages. This can be exploited through a
6255 Denial of Service attack.
6256 Thanks to Adam Langley for discovering and researching this issue.
6261 * An attacker can force an error condition which causes openssl to crash
6262 whilst processing DTLS packets due to memory being freed twice. This
6263 can be exploited through a Denial of Service attack.
6264 Thanks to Adam Langley and Wan-Teh Chang for discovering and researching
6270 * If a multithreaded client connects to a malicious server using a resumed
6271 session and the server sends an ec point format extension it could write
6272 up to 255 bytes to freed memory.
6274 Thanks to Gabor Tyukasz (LogMeIn Inc) for discovering and researching this
6280 * A malicious server can crash an OpenSSL client with a null pointer
6281 dereference (read) by specifying an SRP ciphersuite even though it was not
6282 properly negotiated with the client. This can be exploited through a
6283 Denial of Service attack.
6285 Thanks to Joonas Kuorilehto and Riku Hietamäki (Codenomicon) for
6286 discovering and researching this issue.
6291 * A flaw in OBJ_obj2txt may cause pretty printing functions such as
6292 X509_name_oneline, X509_name_print_ex et al. to leak some information
6293 from the stack. Applications may be affected if they echo pretty printing
6294 output to the attacker.
6296 Thanks to Ivan Fratric (Google) for discovering this issue.
6299 *Emilia Käsper, and Steve Henson*
6301 * Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
6302 for corner cases. (Certain input points at infinity could lead to
6303 bogus results, with non-infinity inputs mapped to infinity too.)
6307 ### Changes between 1.0.1g and 1.0.1h [5 Jun 2014]
6309 * Fix for SSL/TLS MITM flaw. An attacker using a carefully crafted
6310 handshake can force the use of weak keying material in OpenSSL
6311 SSL/TLS clients and servers.
6313 Thanks to KIKUCHI Masashi (Lepidum Co. Ltd.) for discovering and
6314 researching this issue. [CVE-2014-0224][]
6316 *KIKUCHI Masashi, Steve Henson*
6318 * Fix DTLS recursion flaw. By sending an invalid DTLS handshake to an
6319 OpenSSL DTLS client the code can be made to recurse eventually crashing
6322 Thanks to Imre Rad (Search-Lab Ltd.) for discovering this issue.
6325 *Imre Rad, Steve Henson*
6327 * Fix DTLS invalid fragment vulnerability. A buffer overrun attack can
6328 be triggered by sending invalid DTLS fragments to an OpenSSL DTLS
6329 client or server. This is potentially exploitable to run arbitrary
6330 code on a vulnerable client or server.
6332 Thanks to Jüri Aedla for reporting this issue. [CVE-2014-0195][]
6334 *Jüri Aedla, Steve Henson*
6336 * Fix bug in TLS code where clients enable anonymous ECDH ciphersuites
6337 are subject to a denial of service attack.
6339 Thanks to Felix Gröbert and Ivan Fratric at Google for discovering
6340 this issue. [CVE-2014-3470][]
6342 *Felix Gröbert, Ivan Fratric, Steve Henson*
6344 * Harmonize version and its documentation. -f flag is used to display
6347 *mancha <mancha1@zoho.com>*
6349 * Fix eckey_priv_encode so it immediately returns an error upon a failure
6350 in i2d_ECPrivateKey.
6352 *mancha <mancha1@zoho.com>*
6354 * Fix some double frees. These are not thought to be exploitable.
6356 *mancha <mancha1@zoho.com>*
6358 ### Changes between 1.0.1f and 1.0.1g [7 Apr 2014]
6360 * A missing bounds check in the handling of the TLS heartbeat extension
6361 can be used to reveal up to 64k of memory to a connected client or
6364 Thanks for Neel Mehta of Google Security for discovering this bug and to
6365 Adam Langley <agl@chromium.org> and Bodo Moeller <bmoeller@acm.org> for
6366 preparing the fix [CVE-2014-0160][]
6368 *Adam Langley, Bodo Moeller*
6370 * Fix for the attack described in the paper "Recovering OpenSSL
6371 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
6372 by Yuval Yarom and Naomi Benger. Details can be obtained from:
6373 <http://eprint.iacr.org/2014/140>
6375 Thanks to Yuval Yarom and Naomi Benger for discovering this
6376 flaw and to Yuval Yarom for supplying a fix [CVE-2014-0076][]
6378 *Yuval Yarom and Naomi Benger*
6380 * TLS pad extension: draft-agl-tls-padding-03
6382 Workaround for the "TLS hang bug" (see FAQ and PR#2771): if the
6383 TLS client Hello record length value would otherwise be > 255 and
6384 less that 512 pad with a dummy extension containing zeroes so it
6385 is at least 512 bytes long.
6387 *Adam Langley, Steve Henson*
6389 ### Changes between 1.0.1e and 1.0.1f [6 Jan 2014]
6391 * Fix for TLS record tampering bug. A carefully crafted invalid
6392 handshake could crash OpenSSL with a NULL pointer exception.
6393 Thanks to Anton Johansson for reporting this issues.
6396 * Keep original DTLS digest and encryption contexts in retransmission
6397 structures so we can use the previous session parameters if they need
6398 to be resent. [CVE-2013-6450][]
6402 * Add option SSL_OP_SAFARI_ECDHE_ECDSA_BUG (part of SSL_OP_ALL) which
6403 avoids preferring ECDHE-ECDSA ciphers when the client appears to be
6404 Safari on OS X. Safari on OS X 10.8..10.8.3 advertises support for
6405 several ECDHE-ECDSA ciphers, but fails to negotiate them. The bug
6406 is fixed in OS X 10.8.4, but Apple have ruled out both hot fixing
6407 10.8..10.8.3 and forcing users to upgrade to 10.8.4 or newer.
6409 *Rob Stradling, Adam Langley*
6411 ### Changes between 1.0.1d and 1.0.1e [11 Feb 2013]
6413 * Correct fix for CVE-2013-0169. The original didn't work on AES-NI
6414 supporting platforms or when small records were transferred.
6416 *Andy Polyakov, Steve Henson*
6418 ### Changes between 1.0.1c and 1.0.1d [5 Feb 2013]
6420 * Make the decoding of SSLv3, TLS and DTLS CBC records constant time.
6422 This addresses the flaw in CBC record processing discovered by
6423 Nadhem Alfardan and Kenny Paterson. Details of this attack can be found
6424 at: <http://www.isg.rhul.ac.uk/tls/>
6426 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
6427 Security Group at Royal Holloway, University of London
6428 (www.isg.rhul.ac.uk) for discovering this flaw and Adam Langley and
6429 Emilia Käsper for the initial patch.
6432 *Emilia Käsper, Adam Langley, Ben Laurie, Andy Polyakov, Steve Henson*
6434 * Fix flaw in AESNI handling of TLS 1.2 and 1.1 records for CBC mode
6435 ciphersuites which can be exploited in a denial of service attack.
6436 Thanks go to and to Adam Langley <agl@chromium.org> for discovering
6437 and detecting this bug and to Wolfgang Ettlinger
6438 <wolfgang.ettlinger@gmail.com> for independently discovering this issue.
6443 * Return an error when checking OCSP signatures when key is NULL.
6444 This fixes a DoS attack. [CVE-2013-0166][]
6448 * Make openssl verify return errors.
6450 *Chris Palmer <palmer@google.com> and Ben Laurie*
6452 * Call OCSP Stapling callback after ciphersuite has been chosen, so
6453 the right response is stapled. Also change SSL_get_certificate()
6454 so it returns the certificate actually sent.
6455 See <http://rt.openssl.org/Ticket/Display.html?id=2836>.
6457 *Rob Stradling <rob.stradling@comodo.com>*
6459 * Fix possible deadlock when decoding public keys.
6463 * Don't use TLS 1.0 record version number in initial client hello
6468 ### Changes between 1.0.1b and 1.0.1c [10 May 2012]
6470 * Sanity check record length before skipping explicit IV in TLS
6471 1.2, 1.1 and DTLS to fix DoS attack.
6473 Thanks to Codenomicon for discovering this issue using Fuzz-o-Matic
6474 fuzzing as a service testing platform.
6479 * Initialise tkeylen properly when encrypting CMS messages.
6480 Thanks to Solar Designer of Openwall for reporting this issue.
6484 * In FIPS mode don't try to use composite ciphers as they are not
6489 ### Changes between 1.0.1a and 1.0.1b [26 Apr 2012]
6491 * OpenSSL 1.0.0 sets SSL_OP_ALL to 0x80000FFFL and OpenSSL 1.0.1 and
6492 1.0.1a set SSL_OP_NO_TLSv1_1 to 0x00000400L which would unfortunately
6493 mean any application compiled against OpenSSL 1.0.0 headers setting
6494 SSL_OP_ALL would also set SSL_OP_NO_TLSv1_1, unintentionally disabling
6495 TLS 1.1 also. Fix this by changing the value of SSL_OP_NO_TLSv1_1 to
6496 0x10000000L Any application which was previously compiled against
6497 OpenSSL 1.0.1 or 1.0.1a headers and which cares about SSL_OP_NO_TLSv1_1
6498 will need to be recompiled as a result. Letting be results in
6499 inability to disable specifically TLS 1.1 and in client context,
6500 in unlike event, limit maximum offered version to TLS 1.0 [see below].
6504 * In order to ensure interoperability SSL_OP_NO_protocolX does not
6505 disable just protocol X, but all protocols above X *if* there are
6506 protocols *below* X still enabled. In more practical terms it means
6507 that if application wants to disable TLS1.0 in favor of TLS1.1 and
6508 above, it's not sufficient to pass SSL_OP_NO_TLSv1, one has to pass
6509 SSL_OP_NO_TLSv1|SSL_OP_NO_SSLv3|SSL_OP_NO_SSLv2. This applies to
6514 ### Changes between 1.0.1 and 1.0.1a [19 Apr 2012]
6516 * Check for potentially exploitable overflows in asn1_d2i_read_bio
6517 BUF_mem_grow and BUF_mem_grow_clean. Refuse attempts to shrink buffer
6518 in CRYPTO_realloc_clean.
6520 Thanks to Tavis Ormandy, Google Security Team, for discovering this
6521 issue and to Adam Langley <agl@chromium.org> for fixing it.
6524 *Adam Langley (Google), Tavis Ormandy, Google Security Team*
6526 * Don't allow TLS 1.2 SHA-256 ciphersuites in TLS 1.0, 1.1 connections.
6530 * Workarounds for some broken servers that "hang" if a client hello
6531 record length exceeds 255 bytes.
6533 1. Do not use record version number > TLS 1.0 in initial client
6534 hello: some (but not all) hanging servers will now work.
6535 2. If we set OPENSSL_MAX_TLS1_2_CIPHER_LENGTH this will truncate
6536 the number of ciphers sent in the client hello. This should be
6537 set to an even number, such as 50, for example by passing:
6538 -DOPENSSL_MAX_TLS1_2_CIPHER_LENGTH=50 to config or Configure.
6539 Most broken servers should now work.
6540 3. If all else fails setting OPENSSL_NO_TLS1_2_CLIENT will disable
6541 TLS 1.2 client support entirely.
6545 * Fix SEGV in Vector Permutation AES module observed in OpenSSH.
6549 ### Changes between 1.0.0h and 1.0.1 [14 Mar 2012]
6551 * Add compatibility with old MDC2 signatures which use an ASN1 OCTET
6552 STRING form instead of a DigestInfo.
6556 * The format used for MDC2 RSA signatures is inconsistent between EVP
6557 and the RSA_sign/RSA_verify functions. This was made more apparent when
6558 OpenSSL used RSA_sign/RSA_verify for some RSA signatures in particular
6559 those which went through EVP_PKEY_METHOD in 1.0.0 and later. Detect
6560 the correct format in RSA_verify so both forms transparently work.
6564 * Some servers which support TLS 1.0 can choke if we initially indicate
6565 support for TLS 1.2 and later renegotiate using TLS 1.0 in the RSA
6566 encrypted premaster secret. As a workaround use the maximum permitted
6567 client version in client hello, this should keep such servers happy
6568 and still work with previous versions of OpenSSL.
6572 * Add support for TLS/DTLS heartbeats.
6574 *Robin Seggelmann <seggelmann@fh-muenster.de>*
6576 * Add support for SCTP.
6578 *Robin Seggelmann <seggelmann@fh-muenster.de>*
6580 * Improved PRNG seeding for VOS.
6582 *Paul Green <Paul.Green@stratus.com>*
6584 * Extensive assembler packs updates, most notably:
6586 - x86[_64]: AES-NI, PCLMULQDQ, RDRAND support;
6587 - x86[_64]: SSSE3 support (SHA1, vector-permutation AES);
6588 - x86_64: bit-sliced AES implementation;
6589 - ARM: NEON support, contemporary platforms optimizations;
6590 - s390x: z196 support;
6591 - `*`: GHASH and GF(2^m) multiplication implementations;
6595 * Make TLS-SRP code conformant with RFC 5054 API cleanup
6596 (removal of unnecessary code)
6598 *Peter Sylvester <peter.sylvester@edelweb.fr>*
6600 * Add TLS key material exporter from RFC 5705.
6604 * Add DTLS-SRTP negotiation from RFC 5764.
6608 * Add Next Protocol Negotiation,
6609 <http://tools.ietf.org/html/draft-agl-tls-nextprotoneg-00>. Can be
6610 disabled with a no-npn flag to config or Configure. Code donated
6613 *Adam Langley <agl@google.com> and Ben Laurie*
6615 * Add optional 64-bit optimized implementations of elliptic curves NIST-P224,
6616 NIST-P256, NIST-P521, with constant-time single point multiplication on
6617 typical inputs. Compiler support for the nonstandard type `__uint128_t` is
6618 required to use this (present in gcc 4.4 and later, for 64-bit builds).
6619 Code made available under Apache License version 2.0.
6621 Specify "enable-ec_nistp_64_gcc_128" on the Configure (or config) command
6622 line to include this in your build of OpenSSL, and run "make depend" (or
6623 "make update"). This enables the following EC_METHODs:
6625 EC_GFp_nistp224_method()
6626 EC_GFp_nistp256_method()
6627 EC_GFp_nistp521_method()
6629 EC_GROUP_new_by_curve_name() will automatically use these (while
6630 EC_GROUP_new_curve_GFp() currently prefers the more flexible
6633 *Emilia Käsper, Adam Langley, Bodo Moeller (Google)*
6635 * Use type ossl_ssize_t instead of ssize_t which isn't available on
6636 all platforms. Move ssize_t definition from e_os.h to the public
6637 header file e_os2.h as it now appears in public header file cms.h
6641 * New -sigopt option to the ca, req and x509 utilities. Additional
6642 signature parameters can be passed using this option and in
6647 * Add RSA PSS signing function. This will generate and set the
6648 appropriate AlgorithmIdentifiers for PSS based on those in the
6649 corresponding EVP_MD_CTX structure. No application support yet.
6653 * Support for companion algorithm specific ASN1 signing routines.
6654 New function ASN1_item_sign_ctx() signs a pre-initialised
6655 EVP_MD_CTX structure and sets AlgorithmIdentifiers based on
6656 the appropriate parameters.
6660 * Add new algorithm specific ASN1 verification initialisation function
6661 to EVP_PKEY_ASN1_METHOD: this is not in EVP_PKEY_METHOD since the ASN1
6662 handling will be the same no matter what EVP_PKEY_METHOD is used.
6663 Add a PSS handler to support verification of PSS signatures: checked
6664 against a number of sample certificates.
6668 * Add signature printing for PSS. Add PSS OIDs.
6670 *Steve Henson, Martin Kaiser <lists@kaiser.cx>*
6672 * Add algorithm specific signature printing. An individual ASN1 method
6673 can now print out signatures instead of the standard hex dump.
6675 More complex signatures (e.g. PSS) can print out more meaningful
6676 information. Include DSA version that prints out the signature
6681 * Password based recipient info support for CMS library: implementing
6686 * Split password based encryption into PBES2 and PBKDF2 functions. This
6687 neatly separates the code into cipher and PBE sections and is required
6688 for some algorithms that split PBES2 into separate pieces (such as
6689 password based CMS).
6693 * Session-handling fixes:
6694 - Fix handling of connections that are resuming with a session ID,
6695 but also support Session Tickets.
6696 - Fix a bug that suppressed issuing of a new ticket if the client
6697 presented a ticket with an expired session.
6698 - Try to set the ticket lifetime hint to something reasonable.
6699 - Make tickets shorter by excluding irrelevant information.
6700 - On the client side, don't ignore renewed tickets.
6702 *Adam Langley, Bodo Moeller (Google)*
6704 * Fix PSK session representation.
6708 * Add RC4-MD5 and AESNI-SHA1 "stitched" implementations.
6710 This work was sponsored by Intel.
6714 * Add GCM support to TLS library. Some custom code is needed to split
6715 the IV between the fixed (from PRF) and explicit (from TLS record)
6716 portions. This adds all GCM ciphersuites supported by RFC5288 and
6717 RFC5289. Generalise some `AES*` cipherstrings to include GCM and
6718 add a special AESGCM string for GCM only.
6722 * Expand range of ctrls for AES GCM. Permit setting invocation
6723 field on decrypt and retrieval of invocation field only on encrypt.
6727 * Add HMAC ECC ciphersuites from RFC5289. Include SHA384 PRF support.
6728 As required by RFC5289 these ciphersuites cannot be used if for
6729 versions of TLS earlier than 1.2.
6733 * For FIPS capable OpenSSL interpret a NULL default public key method
6734 as unset and return the appropriate default but do *not* set the default.
6735 This means we can return the appropriate method in applications that
6736 switch between FIPS and non-FIPS modes.
6740 * Redirect HMAC and CMAC operations to FIPS module in FIPS mode. If an
6741 ENGINE is used then we cannot handle that in the FIPS module so we
6742 keep original code iff non-FIPS operations are allowed.
6746 * Add -attime option to openssl utilities.
6748 *Peter Eckersley <pde@eff.org>, Ben Laurie and Steve Henson*
6750 * Redirect DSA and DH operations to FIPS module in FIPS mode.
6754 * Redirect ECDSA and ECDH operations to FIPS module in FIPS mode. Also use
6755 FIPS EC methods unconditionally for now.
6759 * New build option no-ec2m to disable characteristic 2 code.
6763 * Backport libcrypto audit of return value checking from 1.1.0-dev; not
6764 all cases can be covered as some introduce binary incompatibilities.
6768 * Redirect RSA operations to FIPS module including keygen,
6769 encrypt, decrypt, sign and verify. Block use of non FIPS RSA methods.
6773 * Add similar low level API blocking to ciphers.
6777 * Low level digest APIs are not approved in FIPS mode: any attempt
6778 to use these will cause a fatal error. Applications that *really* want
6779 to use them can use the `private_*` version instead.
6783 * Redirect cipher operations to FIPS module for FIPS builds.
6787 * Redirect digest operations to FIPS module for FIPS builds.
6791 * Update build system to add "fips" flag which will link in fipscanister.o
6792 for static and shared library builds embedding a signature if needed.
6796 * Output TLS supported curves in preference order instead of numerical
6797 order. This is currently hardcoded for the highest order curves first.
6798 This should be configurable so applications can judge speed vs strength.
6802 * Add TLS v1.2 server support for client authentication.
6806 * Add support for FIPS mode in ssl library: disable SSLv3, non-FIPS ciphers
6811 * Functions FIPS_mode_set() and FIPS_mode() which call the underlying
6812 FIPS modules versions.
6816 * Add TLS v1.2 client side support for client authentication. Keep cache
6817 of handshake records longer as we don't know the hash algorithm to use
6818 until after the certificate request message is received.
6822 * Initial TLS v1.2 client support. Add a default signature algorithms
6823 extension including all the algorithms we support. Parse new signature
6824 format in client key exchange. Relax some ECC signing restrictions for
6825 TLS v1.2 as indicated in RFC5246.
6829 * Add server support for TLS v1.2 signature algorithms extension. Switch
6830 to new signature format when needed using client digest preference.
6831 All server ciphersuites should now work correctly in TLS v1.2. No client
6832 support yet and no support for client certificates.
6836 * Initial TLS v1.2 support. Add new SHA256 digest to ssl code, switch
6837 to SHA256 for PRF when using TLS v1.2 and later. Add new SHA256 based
6838 ciphersuites. At present only RSA key exchange ciphersuites work with
6839 TLS v1.2. Add new option for TLS v1.2 replacing the old and obsolete
6840 SSL_OP_PKCS1_CHECK flags with SSL_OP_NO_TLSv1_2. New TLSv1.2 methods
6841 and version checking.
6845 * New option OPENSSL_NO_SSL_INTERN. If an application can be compiled
6846 with this defined it will not be affected by any changes to ssl internal
6847 structures. Add several utility functions to allow openssl application
6848 to work with OPENSSL_NO_SSL_INTERN defined.
6852 * A long standing patch to add support for SRP from EdelWeb (Peter
6853 Sylvester and Christophe Renou) was integrated.
6854 *Christophe Renou <christophe.renou@edelweb.fr>, Peter Sylvester
6855 <peter.sylvester@edelweb.fr>, Tom Wu <tjw@cs.stanford.edu>, and
6858 * Add functions to copy EVP_PKEY_METHOD and retrieve flags and id.
6862 * Permit abbreviated handshakes when renegotiating using the function
6863 SSL_renegotiate_abbreviated().
6865 *Robin Seggelmann <seggelmann@fh-muenster.de>*
6867 * Add call to ENGINE_register_all_complete() to
6868 ENGINE_load_builtin_engines(), so some implementations get used
6869 automatically instead of needing explicit application support.
6873 * Add support for TLS key exporter as described in RFC5705.
6875 *Robin Seggelmann <seggelmann@fh-muenster.de>, Steve Henson*
6877 * Initial TLSv1.1 support. Since TLSv1.1 is very similar to TLS v1.0 only
6878 a few changes are required:
6880 Add SSL_OP_NO_TLSv1_1 flag.
6881 Add TLSv1_1 methods.
6882 Update version checking logic to handle version 1.1.
6883 Add explicit IV handling (ported from DTLS code).
6884 Add command line options to s_client/s_server.
6891 ### Changes between 1.0.0s and 1.0.0t [3 Dec 2015]
6893 * X509_ATTRIBUTE memory leak
6895 When presented with a malformed X509_ATTRIBUTE structure OpenSSL will leak
6896 memory. This structure is used by the PKCS#7 and CMS routines so any
6897 application which reads PKCS#7 or CMS data from untrusted sources is
6898 affected. SSL/TLS is not affected.
6900 This issue was reported to OpenSSL by Adam Langley (Google/BoringSSL) using
6906 * Race condition handling PSK identify hint
6908 If PSK identity hints are received by a multi-threaded client then
6909 the values are wrongly updated in the parent SSL_CTX structure. This can
6910 result in a race condition potentially leading to a double free of the
6916 ### Changes between 1.0.0r and 1.0.0s [11 Jun 2015]
6918 * Malformed ECParameters causes infinite loop
6920 When processing an ECParameters structure OpenSSL enters an infinite loop
6921 if the curve specified is over a specially malformed binary polynomial
6924 This can be used to perform denial of service against any
6925 system which processes public keys, certificate requests or
6926 certificates. This includes TLS clients and TLS servers with
6927 client authentication enabled.
6929 This issue was reported to OpenSSL by Joseph Barr-Pixton.
6934 * Exploitable out-of-bounds read in X509_cmp_time
6936 X509_cmp_time does not properly check the length of the ASN1_TIME
6937 string and can read a few bytes out of bounds. In addition,
6938 X509_cmp_time accepts an arbitrary number of fractional seconds in the
6941 An attacker can use this to craft malformed certificates and CRLs of
6942 various sizes and potentially cause a segmentation fault, resulting in
6943 a DoS on applications that verify certificates or CRLs. TLS clients
6944 that verify CRLs are affected. TLS clients and servers with client
6945 authentication enabled may be affected if they use custom verification
6948 This issue was reported to OpenSSL by Robert Swiecki (Google), and
6949 independently by Hanno Böck.
6954 * PKCS7 crash with missing EnvelopedContent
6956 The PKCS#7 parsing code does not handle missing inner EncryptedContent
6957 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs
6958 with missing content and trigger a NULL pointer dereference on parsing.
6960 Applications that decrypt PKCS#7 data or otherwise parse PKCS#7
6961 structures from untrusted sources are affected. OpenSSL clients and
6962 servers are not affected.
6964 This issue was reported to OpenSSL by Michal Zalewski (Google).
6969 * CMS verify infinite loop with unknown hash function
6971 When verifying a signedData message the CMS code can enter an infinite loop
6972 if presented with an unknown hash function OID. This can be used to perform
6973 denial of service against any system which verifies signedData messages using
6975 This issue was reported to OpenSSL by Johannes Bauer.
6980 * Race condition handling NewSessionTicket
6982 If a NewSessionTicket is received by a multi-threaded client when attempting to
6983 reuse a previous ticket then a race condition can occur potentially leading to
6984 a double free of the ticket data.
6989 ### Changes between 1.0.0q and 1.0.0r [19 Mar 2015]
6991 * Segmentation fault in ASN1_TYPE_cmp fix
6993 The function ASN1_TYPE_cmp will crash with an invalid read if an attempt is
6994 made to compare ASN.1 boolean types. Since ASN1_TYPE_cmp is used to check
6995 certificate signature algorithm consistency this can be used to crash any
6996 certificate verification operation and exploited in a DoS attack. Any
6997 application which performs certificate verification is vulnerable including
6998 OpenSSL clients and servers which enable client authentication.
7003 * ASN.1 structure reuse memory corruption fix
7005 Reusing a structure in ASN.1 parsing may allow an attacker to cause
7006 memory corruption via an invalid write. Such reuse is and has been
7007 strongly discouraged and is believed to be rare.
7009 Applications that parse structures containing CHOICE or ANY DEFINED BY
7010 components may be affected. Certificate parsing (d2i_X509 and related
7011 functions) are however not affected. OpenSSL clients and servers are
7017 * PKCS7 NULL pointer dereferences fix
7019 The PKCS#7 parsing code does not handle missing outer ContentInfo
7020 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs with
7021 missing content and trigger a NULL pointer dereference on parsing.
7023 Applications that verify PKCS#7 signatures, decrypt PKCS#7 data or
7024 otherwise parse PKCS#7 structures from untrusted sources are
7025 affected. OpenSSL clients and servers are not affected.
7027 This issue was reported to OpenSSL by Michal Zalewski (Google).
7032 * DoS via reachable assert in SSLv2 servers fix
7034 A malicious client can trigger an OPENSSL_assert (i.e., an abort) in
7035 servers that both support SSLv2 and enable export cipher suites by sending
7036 a specially crafted SSLv2 CLIENT-MASTER-KEY message.
7038 This issue was discovered by Sean Burford (Google) and Emilia Käsper
7039 (OpenSSL development team).
7044 * Use After Free following d2i_ECPrivatekey error fix
7046 A malformed EC private key file consumed via the d2i_ECPrivateKey function
7047 could cause a use after free condition. This, in turn, could cause a double
7048 free in several private key parsing functions (such as d2i_PrivateKey
7049 or EVP_PKCS82PKEY) and could lead to a DoS attack or memory corruption
7050 for applications that receive EC private keys from untrusted
7051 sources. This scenario is considered rare.
7053 This issue was discovered by the BoringSSL project and fixed in their
7059 * X509_to_X509_REQ NULL pointer deref fix
7061 The function X509_to_X509_REQ will crash with a NULL pointer dereference if
7062 the certificate key is invalid. This function is rarely used in practice.
7064 This issue was discovered by Brian Carpenter.
7069 * Removed the export ciphers from the DEFAULT ciphers
7073 ### Changes between 1.0.0p and 1.0.0q [15 Jan 2015]
7075 * Build fixes for the Windows and OpenVMS platforms
7077 *Matt Caswell and Richard Levitte*
7079 ### Changes between 1.0.0o and 1.0.0p [8 Jan 2015]
7081 * Fix DTLS segmentation fault in dtls1_get_record. A carefully crafted DTLS
7082 message can cause a segmentation fault in OpenSSL due to a NULL pointer
7083 dereference. This could lead to a Denial Of Service attack. Thanks to
7084 Markus Stenberg of Cisco Systems, Inc. for reporting this issue.
7089 * Fix DTLS memory leak in dtls1_buffer_record. A memory leak can occur in the
7090 dtls1_buffer_record function under certain conditions. In particular this
7091 could occur if an attacker sent repeated DTLS records with the same
7092 sequence number but for the next epoch. The memory leak could be exploited
7093 by an attacker in a Denial of Service attack through memory exhaustion.
7094 Thanks to Chris Mueller for reporting this issue.
7099 * Fix issue where no-ssl3 configuration sets method to NULL. When openssl is
7100 built with the no-ssl3 option and a SSL v3 ClientHello is received the ssl
7101 method would be set to NULL which could later result in a NULL pointer
7102 dereference. Thanks to Frank Schmirler for reporting this issue.
7107 * Abort handshake if server key exchange message is omitted for ephemeral
7110 Thanks to Karthikeyan Bhargavan of the PROSECCO team at INRIA for
7111 reporting this issue.
7116 * Remove non-export ephemeral RSA code on client and server. This code
7117 violated the TLS standard by allowing the use of temporary RSA keys in
7118 non-export ciphersuites and could be used by a server to effectively
7119 downgrade the RSA key length used to a value smaller than the server
7120 certificate. Thanks for Karthikeyan Bhargavan of the PROSECCO team at
7121 INRIA or reporting this issue.
7126 * Fixed issue where DH client certificates are accepted without verification.
7127 An OpenSSL server will accept a DH certificate for client authentication
7128 without the certificate verify message. This effectively allows a client to
7129 authenticate without the use of a private key. This only affects servers
7130 which trust a client certificate authority which issues certificates
7131 containing DH keys: these are extremely rare and hardly ever encountered.
7132 Thanks for Karthikeyan Bhargavan of the PROSECCO team at INRIA or reporting
7138 * Correct Bignum squaring. Bignum squaring (BN_sqr) may produce incorrect
7139 results on some platforms, including x86_64. This bug occurs at random
7140 with a very low probability, and is not known to be exploitable in any
7141 way, though its exact impact is difficult to determine. Thanks to Pieter
7142 Wuille (Blockstream) who reported this issue and also suggested an initial
7143 fix. Further analysis was conducted by the OpenSSL development team and
7144 Adam Langley of Google. The final fix was developed by Andy Polyakov of
7145 the OpenSSL core team.
7150 * Fix various certificate fingerprint issues.
7152 By using non-DER or invalid encodings outside the signed portion of a
7153 certificate the fingerprint can be changed without breaking the signature.
7154 Although no details of the signed portion of the certificate can be changed
7155 this can cause problems with some applications: e.g. those using the
7156 certificate fingerprint for blacklists.
7158 1. Reject signatures with non zero unused bits.
7160 If the BIT STRING containing the signature has non zero unused bits reject
7161 the signature. All current signature algorithms require zero unused bits.
7163 2. Check certificate algorithm consistency.
7165 Check the AlgorithmIdentifier inside TBS matches the one in the
7166 certificate signature. NB: this will result in signature failure
7167 errors for some broken certificates.
7169 Thanks to Konrad Kraszewski from Google for reporting this issue.
7171 3. Check DSA/ECDSA signatures use DER.
7173 Reencode DSA/ECDSA signatures and compare with the original received
7174 signature. Return an error if there is a mismatch.
7176 This will reject various cases including garbage after signature
7177 (thanks to Antti Karjalainen and Tuomo Untinen from the Codenomicon CROSS
7178 program for discovering this case) and use of BER or invalid ASN.1 INTEGERs
7179 (negative or with leading zeroes).
7181 Further analysis was conducted and fixes were developed by Stephen Henson
7182 of the OpenSSL core team.
7188 ### Changes between 1.0.0n and 1.0.0o [15 Oct 2014]
7190 * Session Ticket Memory Leak.
7192 When an OpenSSL SSL/TLS/DTLS server receives a session ticket the
7193 integrity of that ticket is first verified. In the event of a session
7194 ticket integrity check failing, OpenSSL will fail to free memory
7195 causing a memory leak. By sending a large number of invalid session
7196 tickets an attacker could exploit this issue in a Denial Of Service
7202 * Build option no-ssl3 is incomplete.
7204 When OpenSSL is configured with "no-ssl3" as a build option, servers
7205 could accept and complete a SSL 3.0 handshake, and clients could be
7206 configured to send them.
7209 *Akamai and the OpenSSL team*
7211 * Add support for TLS_FALLBACK_SCSV.
7212 Client applications doing fallback retries should call
7213 SSL_set_mode(s, SSL_MODE_SEND_FALLBACK_SCSV).
7216 *Adam Langley, Bodo Moeller*
7218 * Add additional DigestInfo checks.
7220 Reencode DigestInto in DER and check against the original when
7221 verifying RSA signature: this will reject any improperly encoded
7222 DigestInfo structures.
7224 Note: this is a precautionary measure and no attacks are currently known.
7228 ### Changes between 1.0.0m and 1.0.0n [6 Aug 2014]
7230 * OpenSSL DTLS clients enabling anonymous (EC)DH ciphersuites are subject
7231 to a denial of service attack. A malicious server can crash the client
7232 with a null pointer dereference (read) by specifying an anonymous (EC)DH
7233 ciphersuite and sending carefully crafted handshake messages.
7235 Thanks to Felix Gröbert (Google) for discovering and researching this
7241 * By sending carefully crafted DTLS packets an attacker could cause openssl
7242 to leak memory. This can be exploited through a Denial of Service attack.
7243 Thanks to Adam Langley for discovering and researching this issue.
7248 * An attacker can force openssl to consume large amounts of memory whilst
7249 processing DTLS handshake messages. This can be exploited through a
7250 Denial of Service attack.
7251 Thanks to Adam Langley for discovering and researching this issue.
7256 * An attacker can force an error condition which causes openssl to crash
7257 whilst processing DTLS packets due to memory being freed twice. This
7258 can be exploited through a Denial of Service attack.
7259 Thanks to Adam Langley and Wan-Teh Chang for discovering and researching
7265 * If a multithreaded client connects to a malicious server using a resumed
7266 session and the server sends an ec point format extension it could write
7267 up to 255 bytes to freed memory.
7269 Thanks to Gabor Tyukasz (LogMeIn Inc) for discovering and researching this
7275 * A flaw in OBJ_obj2txt may cause pretty printing functions such as
7276 X509_name_oneline, X509_name_print_ex et al. to leak some information
7277 from the stack. Applications may be affected if they echo pretty printing
7278 output to the attacker.
7280 Thanks to Ivan Fratric (Google) for discovering this issue.
7283 *Emilia Käsper, and Steve Henson*
7285 * Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
7286 for corner cases. (Certain input points at infinity could lead to
7287 bogus results, with non-infinity inputs mapped to infinity too.)
7291 ### Changes between 1.0.0l and 1.0.0m [5 Jun 2014]
7293 * Fix for SSL/TLS MITM flaw. An attacker using a carefully crafted
7294 handshake can force the use of weak keying material in OpenSSL
7295 SSL/TLS clients and servers.
7297 Thanks to KIKUCHI Masashi (Lepidum Co. Ltd.) for discovering and
7298 researching this issue. [CVE-2014-0224][]
7300 *KIKUCHI Masashi, Steve Henson*
7302 * Fix DTLS recursion flaw. By sending an invalid DTLS handshake to an
7303 OpenSSL DTLS client the code can be made to recurse eventually crashing
7306 Thanks to Imre Rad (Search-Lab Ltd.) for discovering this issue.
7309 *Imre Rad, Steve Henson*
7311 * Fix DTLS invalid fragment vulnerability. A buffer overrun attack can
7312 be triggered by sending invalid DTLS fragments to an OpenSSL DTLS
7313 client or server. This is potentially exploitable to run arbitrary
7314 code on a vulnerable client or server.
7316 Thanks to Jüri Aedla for reporting this issue. [CVE-2014-0195][]
7318 *Jüri Aedla, Steve Henson*
7320 * Fix bug in TLS code where clients enable anonymous ECDH ciphersuites
7321 are subject to a denial of service attack.
7323 Thanks to Felix Gröbert and Ivan Fratric at Google for discovering
7324 this issue. [CVE-2014-3470][]
7326 *Felix Gröbert, Ivan Fratric, Steve Henson*
7328 * Harmonize version and its documentation. -f flag is used to display
7331 *mancha <mancha1@zoho.com>*
7333 * Fix eckey_priv_encode so it immediately returns an error upon a failure
7334 in i2d_ECPrivateKey.
7336 *mancha <mancha1@zoho.com>*
7338 * Fix some double frees. These are not thought to be exploitable.
7340 *mancha <mancha1@zoho.com>*
7342 * Fix for the attack described in the paper "Recovering OpenSSL
7343 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
7344 by Yuval Yarom and Naomi Benger. Details can be obtained from:
7345 <http://eprint.iacr.org/2014/140>
7347 Thanks to Yuval Yarom and Naomi Benger for discovering this
7348 flaw and to Yuval Yarom for supplying a fix [CVE-2014-0076][]
7350 *Yuval Yarom and Naomi Benger*
7352 ### Changes between 1.0.0k and 1.0.0l [6 Jan 2014]
7354 * Keep original DTLS digest and encryption contexts in retransmission
7355 structures so we can use the previous session parameters if they need
7356 to be resent. [CVE-2013-6450][]
7360 * Add option SSL_OP_SAFARI_ECDHE_ECDSA_BUG (part of SSL_OP_ALL) which
7361 avoids preferring ECDHE-ECDSA ciphers when the client appears to be
7362 Safari on OS X. Safari on OS X 10.8..10.8.3 advertises support for
7363 several ECDHE-ECDSA ciphers, but fails to negotiate them. The bug
7364 is fixed in OS X 10.8.4, but Apple have ruled out both hot fixing
7365 10.8..10.8.3 and forcing users to upgrade to 10.8.4 or newer.
7367 *Rob Stradling, Adam Langley*
7369 ### Changes between 1.0.0j and 1.0.0k [5 Feb 2013]
7371 * Make the decoding of SSLv3, TLS and DTLS CBC records constant time.
7373 This addresses the flaw in CBC record processing discovered by
7374 Nadhem Alfardan and Kenny Paterson. Details of this attack can be found
7375 at: <http://www.isg.rhul.ac.uk/tls/>
7377 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
7378 Security Group at Royal Holloway, University of London
7379 (www.isg.rhul.ac.uk) for discovering this flaw and Adam Langley and
7380 Emilia Käsper for the initial patch.
7383 *Emilia Käsper, Adam Langley, Ben Laurie, Andy Polyakov, Steve Henson*
7385 * Return an error when checking OCSP signatures when key is NULL.
7386 This fixes a DoS attack. [CVE-2013-0166][]
7390 * Call OCSP Stapling callback after ciphersuite has been chosen, so
7391 the right response is stapled. Also change SSL_get_certificate()
7392 so it returns the certificate actually sent.
7393 See <http://rt.openssl.org/Ticket/Display.html?id=2836>.
7394 (This is a backport)
7396 *Rob Stradling <rob.stradling@comodo.com>*
7398 * Fix possible deadlock when decoding public keys.
7402 ### Changes between 1.0.0i and 1.0.0j [10 May 2012]
7404 [NB: OpenSSL 1.0.0i and later 1.0.0 patch levels were released after
7407 * Sanity check record length before skipping explicit IV in DTLS
7410 Thanks to Codenomicon for discovering this issue using Fuzz-o-Matic
7411 fuzzing as a service testing platform.
7416 * Initialise tkeylen properly when encrypting CMS messages.
7417 Thanks to Solar Designer of Openwall for reporting this issue.
7421 ### Changes between 1.0.0h and 1.0.0i [19 Apr 2012]
7423 * Check for potentially exploitable overflows in asn1_d2i_read_bio
7424 BUF_mem_grow and BUF_mem_grow_clean. Refuse attempts to shrink buffer
7425 in CRYPTO_realloc_clean.
7427 Thanks to Tavis Ormandy, Google Security Team, for discovering this
7428 issue and to Adam Langley <agl@chromium.org> for fixing it.
7431 *Adam Langley (Google), Tavis Ormandy, Google Security Team*
7433 ### Changes between 1.0.0g and 1.0.0h [12 Mar 2012]
7435 * Fix MMA (Bleichenbacher's attack on PKCS #1 v1.5 RSA padding) weakness
7436 in CMS and PKCS7 code. When RSA decryption fails use a random key for
7437 content decryption and always return the same error. Note: this attack
7438 needs on average 2^20 messages so it only affects automated senders. The
7439 old behaviour can be re-enabled in the CMS code by setting the
7440 CMS_DEBUG_DECRYPT flag: this is useful for debugging and testing where
7441 an MMA defence is not necessary.
7442 Thanks to Ivan Nestlerode <inestlerode@us.ibm.com> for discovering
7443 this issue. [CVE-2012-0884][]
7447 * Fix CVE-2011-4619: make sure we really are receiving a
7448 client hello before rejecting multiple SGC restarts. Thanks to
7449 Ivan Nestlerode <inestlerode@us.ibm.com> for discovering this bug.
7453 ### Changes between 1.0.0f and 1.0.0g [18 Jan 2012]
7455 * Fix for DTLS DoS issue introduced by fix for CVE-2011-4109.
7456 Thanks to Antonio Martin, Enterprise Secure Access Research and
7457 Development, Cisco Systems, Inc. for discovering this bug and
7458 preparing a fix. [CVE-2012-0050][]
7462 ### Changes between 1.0.0e and 1.0.0f [4 Jan 2012]
7464 * Nadhem Alfardan and Kenny Paterson have discovered an extension
7465 of the Vaudenay padding oracle attack on CBC mode encryption
7466 which enables an efficient plaintext recovery attack against
7467 the OpenSSL implementation of DTLS. Their attack exploits timing
7468 differences arising during decryption processing. A research
7469 paper describing this attack can be found at:
7470 <http://www.isg.rhul.ac.uk/~kp/dtls.pdf>
7471 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
7472 Security Group at Royal Holloway, University of London
7473 (www.isg.rhul.ac.uk) for discovering this flaw and to Robin Seggelmann
7474 <seggelmann@fh-muenster.de> and Michael Tuexen <tuexen@fh-muenster.de>
7475 for preparing the fix. [CVE-2011-4108][]
7477 *Robin Seggelmann, Michael Tuexen*
7479 * Clear bytes used for block padding of SSL 3.0 records.
7482 *Adam Langley (Google)*
7484 * Only allow one SGC handshake restart for SSL/TLS. Thanks to George
7485 Kadianakis <desnacked@gmail.com> for discovering this issue and
7486 Adam Langley for preparing the fix. [CVE-2011-4619][]
7488 *Adam Langley (Google)*
7490 * Check parameters are not NULL in GOST ENGINE. [CVE-2012-0027][]
7492 *Andrey Kulikov <amdeich@gmail.com>*
7494 * Prevent malformed RFC3779 data triggering an assertion failure.
7495 Thanks to Andrew Chi, BBN Technologies, for discovering the flaw
7496 and Rob Austein <sra@hactrn.net> for fixing it. [CVE-2011-4577][]
7498 *Rob Austein <sra@hactrn.net>*
7500 * Improved PRNG seeding for VOS.
7502 *Paul Green <Paul.Green@stratus.com>*
7504 * Fix ssl_ciph.c set-up race.
7506 *Adam Langley (Google)*
7508 * Fix spurious failures in ecdsatest.c.
7510 *Emilia Käsper (Google)*
7512 * Fix the BIO_f_buffer() implementation (which was mixing different
7513 interpretations of the `..._len` fields).
7515 *Adam Langley (Google)*
7517 * Fix handling of BN_BLINDING: now BN_BLINDING_invert_ex (rather than
7518 BN_BLINDING_invert_ex) calls BN_BLINDING_update, ensuring that concurrent
7519 threads won't reuse the same blinding coefficients.
7521 This also avoids the need to obtain the CRYPTO_LOCK_RSA_BLINDING
7522 lock to call BN_BLINDING_invert_ex, and avoids one use of
7523 BN_BLINDING_update for each BN_BLINDING structure (previously,
7524 the last update always remained unused).
7526 *Emilia Käsper (Google)*
7528 * In ssl3_clear, preserve s3->init_extra along with s3->rbuf.
7530 *Bob Buckholz (Google)*
7532 ### Changes between 1.0.0d and 1.0.0e [6 Sep 2011]
7534 * Fix bug where CRLs with nextUpdate in the past are sometimes accepted
7535 by initialising X509_STORE_CTX properly. [CVE-2011-3207][]
7537 *Kaspar Brand <ossl@velox.ch>*
7539 * Fix SSL memory handling for (EC)DH ciphersuites, in particular
7540 for multi-threaded use of ECDH. [CVE-2011-3210][]
7542 *Adam Langley (Google)*
7544 * Fix x509_name_ex_d2i memory leak on bad inputs.
7548 * Remove hard coded ecdsaWithSHA1 signature tests in ssl code and check
7549 signature public key algorithm by using OID xref utilities instead.
7550 Before this you could only use some ECC ciphersuites with SHA1 only.
7554 * Add protection against ECDSA timing attacks as mentioned in the paper
7555 by Billy Bob Brumley and Nicola Tuveri, see:
7556 <http://eprint.iacr.org/2011/232.pdf>
7558 *Billy Bob Brumley and Nicola Tuveri*
7560 ### Changes between 1.0.0c and 1.0.0d [8 Feb 2011]
7562 * Fix parsing of OCSP stapling ClientHello extension. CVE-2011-0014
7564 *Neel Mehta, Adam Langley, Bodo Moeller (Google)*
7566 * Fix bug in string printing code: if *any* escaping is enabled we must
7567 escape the escape character (backslash) or the resulting string is
7572 ### Changes between 1.0.0b and 1.0.0c [2 Dec 2010]
7574 * Disable code workaround for ancient and obsolete Netscape browsers
7575 and servers: an attacker can use it in a ciphersuite downgrade attack.
7576 Thanks to Martin Rex for discovering this bug. CVE-2010-4180
7580 * Fixed J-PAKE implementation error, originally discovered by
7581 Sebastien Martini, further info and confirmation from Stefan
7582 Arentz and Feng Hao. Note that this fix is a security fix. CVE-2010-4252
7586 ### Changes between 1.0.0a and 1.0.0b [16 Nov 2010]
7588 * Fix extension code to avoid race conditions which can result in a buffer
7589 overrun vulnerability: resumed sessions must not be modified as they can
7590 be shared by multiple threads. CVE-2010-3864
7594 * Fix WIN32 build system to correctly link an ENGINE directory into
7599 ### Changes between 1.0.0 and 1.0.0a [01 Jun 2010]
7601 * Check return value of int_rsa_verify in pkey_rsa_verifyrecover
7604 *Steve Henson, Peter-Michael Hager <hager@dortmund.net>*
7606 ### Changes between 0.9.8n and 1.0.0 [29 Mar 2010]
7608 * Add "missing" function EVP_CIPHER_CTX_copy(). This copies a cipher
7609 context. The operation can be customised via the ctrl mechanism in
7610 case ENGINEs want to include additional functionality.
7614 * Tolerate yet another broken PKCS#8 key format: private key value negative.
7618 * Add new -subject_hash_old and -issuer_hash_old options to x509 utility to
7619 output hashes compatible with older versions of OpenSSL.
7621 *Willy Weisz <weisz@vcpc.univie.ac.at>*
7623 * Fix compression algorithm handling: if resuming a session use the
7624 compression algorithm of the resumed session instead of determining
7625 it from client hello again. Don't allow server to change algorithm.
7629 * Add load_crls() function to apps tidying load_certs() too. Add option
7630 to verify utility to allow additional CRLs to be included.
7634 * Update OCSP request code to permit adding custom headers to the request:
7635 some responders need this.
7639 * The function EVP_PKEY_sign() returns <=0 on error: check return code
7642 *Julia Lawall <julia@diku.dk>*
7644 * Update verify callback code in apps/s_cb.c and apps/verify.c, it
7645 needlessly dereferenced structures, used obsolete functions and
7646 didn't handle all updated verify codes correctly.
7650 * Disable MD2 in the default configuration.
7654 * In BIO_pop() and BIO_push() use the ctrl argument (which was NULL) to
7655 indicate the initial BIO being pushed or popped. This makes it possible
7656 to determine whether the BIO is the one explicitly called or as a result
7657 of the ctrl being passed down the chain. Fix BIO_pop() and SSL BIOs so
7658 it handles reference counts correctly and doesn't zero out the I/O bio
7659 when it is not being explicitly popped. WARNING: applications which
7660 included workarounds for the old buggy behaviour will need to be modified
7661 or they could free up already freed BIOs.
7665 * Extend the uni2asc/asc2uni => OPENSSL_uni2asc/OPENSSL_asc2uni
7666 renaming to all platforms (within the 0.9.8 branch, this was
7667 done conditionally on Netware platforms to avoid a name clash).
7669 *Guenter <lists@gknw.net>*
7671 * Add ECDHE and PSK support to DTLS.
7673 *Michael Tuexen <tuexen@fh-muenster.de>*
7675 * Add CHECKED_STACK_OF macro to safestack.h, otherwise safestack can't
7680 * Add "missing" function EVP_MD_flags() (without this the only way to
7681 retrieve a digest flags is by accessing the structure directly. Update
7682 `EVP_MD_do_all*()` and `EVP_CIPHER_do_all*()` to include the name a digest
7683 or cipher is registered as in the "from" argument. Print out all
7684 registered digests in the dgst usage message instead of manually
7685 attempting to work them out.
7689 * If no SSLv2 ciphers are used don't use an SSLv2 compatible client hello:
7690 this allows the use of compression and extensions. Change default cipher
7691 string to remove SSLv2 ciphersuites. This effectively avoids ancient SSLv2
7692 by default unless an application cipher string requests it.
7696 * Alter match criteria in PKCS12_parse(). It used to try to use local
7697 key ids to find matching certificates and keys but some PKCS#12 files
7698 don't follow the (somewhat unwritten) rules and this strategy fails.
7699 Now just gather all certificates together and the first private key
7700 then look for the first certificate that matches the key.
7704 * Support use of registered digest and cipher names for dgst and cipher
7705 commands instead of having to add each one as a special case. So now
7712 openssl dgst -sha256 foo
7714 and this works for ENGINE based algorithms too.
7718 * Update Gost ENGINE to support parameter files.
7720 *Victor B. Wagner <vitus@cryptocom.ru>*
7722 * Support GeneralizedTime in ca utility.
7724 *Oliver Martin <oliver@volatilevoid.net>, Steve Henson*
7726 * Enhance the hash format used for certificate directory links. The new
7727 form uses the canonical encoding (meaning equivalent names will work
7728 even if they aren't identical) and uses SHA1 instead of MD5. This form
7729 is incompatible with the older format and as a result c_rehash should
7730 be used to rebuild symbolic links.
7734 * Make PKCS#8 the default write format for private keys, replacing the
7735 traditional format. This form is standardised, more secure and doesn't
7736 include an implicit MD5 dependency.
7740 * Add a $gcc_devteam_warn option to Configure. The idea is that any code
7741 committed to OpenSSL should pass this lot as a minimum.
7745 * Add session ticket override functionality for use by EAP-FAST.
7747 *Jouni Malinen <j@w1.fi>*
7749 * Modify HMAC functions to return a value. Since these can be implemented
7750 in an ENGINE errors can occur.
7754 * Type-checked OBJ_bsearch_ex.
7758 * Type-checked OBJ_bsearch. Also some constification necessitated
7759 by type-checking. Still to come: TXT_DB, bsearch(?),
7760 OBJ_bsearch_ex, qsort, CRYPTO_EX_DATA, ASN1_VALUE, ASN1_STRING,
7765 * New function OPENSSL_gmtime_adj() to add a specific number of days and
7766 seconds to a tm structure directly, instead of going through OS
7767 specific date routines. This avoids any issues with OS routines such
7768 as the year 2038 bug. New `*_adj()` functions for ASN1 time structures
7769 and X509_time_adj_ex() to cover the extended range. The existing
7770 X509_time_adj() is still usable and will no longer have any date issues.
7774 * Delta CRL support. New use deltas option which will attempt to locate
7775 and search any appropriate delta CRLs available.
7777 This work was sponsored by Google.
7781 * Support for CRLs partitioned by reason code. Reorganise CRL processing
7782 code and add additional score elements. Validate alternate CRL paths
7783 as part of the CRL checking and indicate a new error "CRL path validation
7784 error" in this case. Applications wanting additional details can use
7785 the verify callback and check the new "parent" field. If this is not
7786 NULL CRL path validation is taking place. Existing applications won't
7787 see this because it requires extended CRL support which is off by
7790 This work was sponsored by Google.
7794 * Support for freshest CRL extension.
7796 This work was sponsored by Google.
7800 * Initial indirect CRL support. Currently only supported in the CRLs
7801 passed directly and not via lookup. Process certificate issuer
7802 CRL entry extension and lookup CRL entries by bother issuer name
7803 and serial number. Check and process CRL issuer entry in IDP extension.
7805 This work was sponsored by Google.
7809 * Add support for distinct certificate and CRL paths. The CRL issuer
7810 certificate is validated separately in this case. Only enabled if
7811 an extended CRL support flag is set: this flag will enable additional
7812 CRL functionality in future.
7814 This work was sponsored by Google.
7818 * Add support for policy mappings extension.
7820 This work was sponsored by Google.
7824 * Fixes to pathlength constraint, self issued certificate handling,
7825 policy processing to align with RFC3280 and PKITS tests.
7827 This work was sponsored by Google.
7831 * Support for name constraints certificate extension. DN, email, DNS
7832 and URI types are currently supported.
7834 This work was sponsored by Google.
7838 * To cater for systems that provide a pointer-based thread ID rather
7839 than numeric, deprecate the current numeric thread ID mechanism and
7840 replace it with a structure and associated callback type. This
7841 mechanism allows a numeric "hash" to be extracted from a thread ID in
7842 either case, and on platforms where pointers are larger than 'long',
7843 mixing is done to help ensure the numeric 'hash' is usable even if it
7844 can't be guaranteed unique. The default mechanism is to use "&errno"
7845 as a pointer-based thread ID to distinguish between threads.
7847 Applications that want to provide their own thread IDs should now use
7848 CRYPTO_THREADID_set_callback() to register a callback that will call
7849 either CRYPTO_THREADID_set_numeric() or CRYPTO_THREADID_set_pointer().
7851 Note that ERR_remove_state() is now deprecated, because it is tied
7852 to the assumption that thread IDs are numeric. ERR_remove_state(0)
7853 to free the current thread's error state should be replaced by
7854 ERR_remove_thread_state(NULL).
7856 (This new approach replaces the functions CRYPTO_set_idptr_callback(),
7857 CRYPTO_get_idptr_callback(), and CRYPTO_thread_idptr() that existed in
7858 OpenSSL 0.9.9-dev between June 2006 and August 2008. Also, if an
7859 application was previously providing a numeric thread callback that
7860 was inappropriate for distinguishing threads, then uniqueness might
7861 have been obtained with &errno that happened immediately in the
7862 intermediate development versions of OpenSSL; this is no longer the
7863 case, the numeric thread callback will now override the automatic use
7866 *Geoff Thorpe, with help from Bodo Moeller*
7868 * Initial support for different CRL issuing certificates. This covers a
7869 simple case where the self issued certificates in the chain exist and
7870 the real CRL issuer is higher in the existing chain.
7872 This work was sponsored by Google.
7876 * Removed effectively defunct crypto/store from the build.
7880 * Revamp of STACK to provide stronger type-checking. Still to come:
7881 TXT_DB, bsearch(?), OBJ_bsearch, qsort, CRYPTO_EX_DATA, ASN1_VALUE,
7882 ASN1_STRING, CONF_VALUE.
7886 * Add a new SSL_MODE_RELEASE_BUFFERS mode flag to release unused buffer
7887 RAM on SSL connections. This option can save about 34k per idle SSL.
7891 * Revamp of LHASH to provide stronger type-checking. Still to come:
7892 STACK, TXT_DB, bsearch, qsort.
7896 * Initial support for Cryptographic Message Syntax (aka CMS) based
7897 on RFC3850, RFC3851 and RFC3852. New cms directory and cms utility,
7898 support for data, signedData, compressedData, digestedData and
7899 encryptedData, envelopedData types included. Scripts to check against
7900 RFC4134 examples draft and interop and consistency checks of many
7901 content types and variants.
7905 * Add options to enc utility to support use of zlib compression BIO.
7909 * Extend mk1mf to support importing of options and assembly language
7910 files from Configure script, currently only included in VC-WIN32.
7911 The assembly language rules can now optionally generate the source
7912 files from the associated perl scripts.
7916 * Implement remaining functionality needed to support GOST ciphersuites.
7917 Interop testing has been performed using CryptoPro implementations.
7919 *Victor B. Wagner <vitus@cryptocom.ru>*
7921 * s390x assembler pack.
7925 * ARMv4 assembler pack. ARMv4 refers to v4 and later ISA, not CPU
7930 * Implement Opaque PRF Input TLS extension as specified in
7931 draft-rescorla-tls-opaque-prf-input-00.txt. Since this is not an
7932 official specification yet and no extension type assignment by
7933 IANA exists, this extension (for now) will have to be explicitly
7934 enabled when building OpenSSL by providing the extension number
7935 to use. For example, specify an option
7937 -DTLSEXT_TYPE_opaque_prf_input=0x9527
7939 to the "config" or "Configure" script to enable the extension,
7940 assuming extension number 0x9527 (which is a completely arbitrary
7941 and unofficial assignment based on the MD5 hash of the Internet
7942 Draft). Note that by doing so, you potentially lose
7943 interoperability with other TLS implementations since these might
7944 be using the same extension number for other purposes.
7946 SSL_set_tlsext_opaque_prf_input(ssl, src, len) is used to set the
7947 opaque PRF input value to use in the handshake. This will create
7948 an internal copy of the length-'len' string at 'src', and will
7949 return non-zero for success.
7951 To get more control and flexibility, provide a callback function
7954 SSL_CTX_set_tlsext_opaque_prf_input_callback(ctx, cb)
7955 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg(ctx, arg)
7959 int (*cb)(SSL *, void *peerinput, size_t len, void *arg);
7962 Callback function 'cb' will be called in handshakes, and is
7963 expected to use SSL_set_tlsext_opaque_prf_input() as appropriate.
7964 Argument 'arg' is for application purposes (the value as given to
7965 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg() will directly
7966 be provided to the callback function). The callback function
7967 has to return non-zero to report success: usually 1 to use opaque
7968 PRF input just if possible, or 2 to enforce use of the opaque PRF
7969 input. In the latter case, the library will abort the handshake
7970 if opaque PRF input is not successfully negotiated.
7972 Arguments 'peerinput' and 'len' given to the callback function
7973 will always be NULL and 0 in the case of a client. A server will
7974 see the client's opaque PRF input through these variables if
7975 available (NULL and 0 otherwise). Note that if the server
7976 provides an opaque PRF input, the length must be the same as the
7977 length of the client's opaque PRF input.
7979 Note that the callback function will only be called when creating
7980 a new session (session resumption can resume whatever was
7981 previously negotiated), and will not be called in SSL 2.0
7982 handshakes; thus, SSL_CTX_set_options(ctx, SSL_OP_NO_SSLv2) or
7983 SSL_set_options(ssl, SSL_OP_NO_SSLv2) is especially recommended
7984 for applications that need to enforce opaque PRF input.
7988 * Update ssl code to support digests other than SHA1+MD5 for handshake
7991 *Victor B. Wagner <vitus@cryptocom.ru>*
7993 * Add RFC4507 support to OpenSSL. This includes the corrections in
7994 RFC4507bis. The encrypted ticket format is an encrypted encoded
7995 SSL_SESSION structure, that way new session features are automatically
7998 If a client application caches session in an SSL_SESSION structure
7999 support is transparent because tickets are now stored in the encoded
8002 The SSL_CTX structure automatically generates keys for ticket
8003 protection in servers so again support should be possible
8004 with no application modification.
8006 If a client or server wishes to disable RFC4507 support then the option
8007 SSL_OP_NO_TICKET can be set.
8009 Add a TLS extension debugging callback to allow the contents of any client
8010 or server extensions to be examined.
8012 This work was sponsored by Google.
8016 * Final changes to avoid use of pointer pointer casts in OpenSSL.
8017 OpenSSL should now compile cleanly on gcc 4.2
8019 *Peter Hartley <pdh@utter.chaos.org.uk>, Steve Henson*
8021 * Update SSL library to use new EVP_PKEY MAC API. Include generic MAC
8022 support including streaming MAC support: this is required for GOST
8023 ciphersuite support.
8025 *Victor B. Wagner <vitus@cryptocom.ru>, Steve Henson*
8027 * Add option -stream to use PKCS#7 streaming in smime utility. New
8028 function i2d_PKCS7_bio_stream() and PEM_write_PKCS7_bio_stream()
8029 to output in BER and PEM format.
8033 * Experimental support for use of HMAC via EVP_PKEY interface. This
8034 allows HMAC to be handled via the `EVP_DigestSign*()` interface. The
8035 EVP_PKEY "key" in this case is the HMAC key, potentially allowing
8036 ENGINE support for HMAC keys which are unextractable. New -mac and
8037 -macopt options to dgst utility.
8041 * New option -sigopt to dgst utility. Update dgst to use
8042 `EVP_Digest{Sign,Verify}*`. These two changes make it possible to use
8043 alternative signing parameters such as X9.31 or PSS in the dgst
8048 * Change ssl_cipher_apply_rule(), the internal function that does
8049 the work each time a ciphersuite string requests enabling
8050 ("foo+bar"), moving ("+foo+bar"), disabling ("-foo+bar", or
8051 removing ("!foo+bar") a class of ciphersuites: Now it maintains
8052 the order of disabled ciphersuites such that those ciphersuites
8053 that most recently went from enabled to disabled not only stay
8054 in order with respect to each other, but also have higher priority
8055 than other disabled ciphersuites the next time ciphersuites are
8058 This means that you can now say, e.g., "PSK:-PSK:HIGH" to enable
8059 the same ciphersuites as with "HIGH" alone, but in a specific
8060 order where the PSK ciphersuites come first (since they are the
8061 most recently disabled ciphersuites when "HIGH" is parsed).
8063 Also, change ssl_create_cipher_list() (using this new
8064 functionality) such that between otherwise identical
8065 ciphersuites, ephemeral ECDH is preferred over ephemeral DH in
8070 * Change ssl_create_cipher_list() so that it automatically
8071 arranges the ciphersuites in reasonable order before starting
8072 to process the rule string. Thus, the definition for "DEFAULT"
8073 (SSL_DEFAULT_CIPHER_LIST) now is just "ALL:!aNULL:!eNULL", but
8074 remains equivalent to "AES:ALL:!aNULL:!eNULL:+aECDH:+kRSA:+RC4:@STRENGTH".
8075 This makes it much easier to arrive at a reasonable default order
8076 in applications for which anonymous ciphers are OK (meaning
8077 that you can't actually use DEFAULT).
8079 *Bodo Moeller; suggested by Victor Duchovni*
8081 * Split the SSL/TLS algorithm mask (as used for ciphersuite string
8082 processing) into multiple integers instead of setting
8083 "SSL_MKEY_MASK" bits, "SSL_AUTH_MASK" bits, "SSL_ENC_MASK",
8084 "SSL_MAC_MASK", and "SSL_SSL_MASK" bits all in a single integer.
8085 (These masks as well as the individual bit definitions are hidden
8086 away into the non-exported interface ssl/ssl_locl.h, so this
8087 change to the definition of the SSL_CIPHER structure shouldn't
8088 affect applications.) This give us more bits for each of these
8089 categories, so there is no longer a need to coagulate AES128 and
8090 AES256 into a single algorithm bit, and to coagulate Camellia128
8091 and Camellia256 into a single algorithm bit, which has led to all
8094 Thus, among other things, the kludge introduced in 0.9.7m and
8095 0.9.8e for masking out AES256 independently of AES128 or masking
8096 out Camellia256 independently of AES256 is not needed here in 0.9.9.
8098 With the change, we also introduce new ciphersuite aliases that
8099 so far were missing: "AES128", "AES256", "CAMELLIA128", and
8104 * Add support for dsa-with-SHA224 and dsa-with-SHA256.
8105 Use the leftmost N bytes of the signature input if the input is
8106 larger than the prime q (with N being the size in bytes of q).
8110 * Very *very* experimental PKCS#7 streaming encoder support. Nothing uses
8111 it yet and it is largely untested.
8115 * Add support for the ecdsa-with-SHA224/256/384/512 signature types.
8119 * Initial incomplete changes to avoid need for function casts in OpenSSL
8120 some compilers (gcc 4.2 and later) reject their use. Safestack is
8121 reimplemented. Update ASN1 to avoid use of legacy functions.
8125 * Win32/64 targets are linked with Winsock2.
8129 * Add an X509_CRL_METHOD structure to allow CRL processing to be redirected
8130 to external functions. This can be used to increase CRL handling
8131 efficiency especially when CRLs are very large by (for example) storing
8132 the CRL revoked certificates in a database.
8136 * Overhaul of by_dir code. Add support for dynamic loading of CRLs so
8137 new CRLs added to a directory can be used. New command line option
8138 -verify_return_error to s_client and s_server. This causes real errors
8139 to be returned by the verify callback instead of carrying on no matter
8140 what. This reflects the way a "real world" verify callback would behave.
8144 * GOST engine, supporting several GOST algorithms and public key formats.
8145 Kindly donated by Cryptocom.
8149 * Partial support for Issuing Distribution Point CRL extension. CRLs
8150 partitioned by DP are handled but no indirect CRL or reason partitioning
8151 (yet). Complete overhaul of CRL handling: now the most suitable CRL is
8152 selected via a scoring technique which handles IDP and AKID in CRLs.
8156 * New X509_STORE_CTX callbacks lookup_crls() and lookup_certs() which
8157 will ultimately be used for all verify operations: this will remove the
8158 X509_STORE dependency on certificate verification and allow alternative
8159 lookup methods. X509_STORE based implementations of these two callbacks.
8163 * Allow multiple CRLs to exist in an X509_STORE with matching issuer names.
8164 Modify get_crl() to find a valid (unexpired) CRL if possible.
8168 * New function X509_CRL_match() to check if two CRLs are identical. Normally
8169 this would be called X509_CRL_cmp() but that name is already used by
8170 a function that just compares CRL issuer names. Cache several CRL
8171 extensions in X509_CRL structure and cache CRLDP in X509.
8175 * Store a "canonical" representation of X509_NAME structure (ASN1 Name)
8176 this maps equivalent X509_NAME structures into a consistent structure.
8177 Name comparison can then be performed rapidly using memcmp().
8181 * Non-blocking OCSP request processing. Add -timeout option to ocsp
8186 * Allow digests to supply their own micalg string for S/MIME type using
8187 the ctrl EVP_MD_CTRL_MICALG.
8191 * During PKCS7 signing pass the PKCS7 SignerInfo structure to the
8192 EVP_PKEY_METHOD before and after signing via the EVP_PKEY_CTRL_PKCS7_SIGN
8193 ctrl. It can then customise the structure before and/or after signing
8198 * New function OBJ_add_sigid() to allow application defined signature OIDs
8199 to be added to OpenSSLs internal tables. New function OBJ_sigid_free()
8200 to free up any added signature OIDs.
8204 * New functions EVP_CIPHER_do_all(), EVP_CIPHER_do_all_sorted(),
8205 EVP_MD_do_all() and EVP_MD_do_all_sorted() to enumerate internal
8206 digest and cipher tables. New options added to openssl utility:
8207 list-message-digest-algorithms and list-cipher-algorithms.
8211 * Change the array representation of binary polynomials: the list
8212 of degrees of non-zero coefficients is now terminated with -1.
8213 Previously it was terminated with 0, which was also part of the
8214 value; thus, the array representation was not applicable to
8215 polynomials where t^0 has coefficient zero. This change makes
8216 the array representation useful in a more general context.
8220 * Various modifications and fixes to SSL/TLS cipher string
8221 handling. For ECC, the code now distinguishes between fixed ECDH
8222 with RSA certificates on the one hand and with ECDSA certificates
8223 on the other hand, since these are separate ciphersuites. The
8224 unused code for Fortezza ciphersuites has been removed.
8226 For consistency with EDH, ephemeral ECDH is now called "EECDH"
8227 (not "ECDHE"). For consistency with the code for DH
8228 certificates, use of ECDH certificates is now considered ECDH
8229 authentication, not RSA or ECDSA authentication (the latter is
8230 merely the CA's signing algorithm and not actively used in the
8233 The temporary ciphersuite alias "ECCdraft" is no longer
8234 available, and ECC ciphersuites are no longer excluded from "ALL"
8235 and "DEFAULT". The following aliases now exist for RFC 4492
8236 ciphersuites, most of these by analogy with the DH case:
8238 kECDHr - ECDH cert, signed with RSA
8239 kECDHe - ECDH cert, signed with ECDSA
8240 kECDH - ECDH cert (signed with either RSA or ECDSA)
8241 kEECDH - ephemeral ECDH
8242 ECDH - ECDH cert or ephemeral ECDH
8248 AECDH - anonymous ECDH
8249 EECDH - non-anonymous ephemeral ECDH (equivalent to "kEECDH:-AECDH")
8253 * Add additional S/MIME capabilities for AES and GOST ciphers if supported.
8254 Use correct micalg parameters depending on digest(s) in signed message.
8258 * Add engine support for EVP_PKEY_ASN1_METHOD. Add functions to process
8259 an ENGINE asn1 method. Support ENGINE lookups in the ASN1 code.
8263 * Initial engine support for EVP_PKEY_METHOD. New functions to permit
8264 an engine to register a method. Add ENGINE lookups for methods and
8265 functional reference processing.
8269 * New functions `EVP_Digest{Sign,Verify)*`. These are enhanced versions of
8270 `EVP_{Sign,Verify}*` which allow an application to customise the signature
8275 * New -resign option to smime utility. This adds one or more signers
8276 to an existing PKCS#7 signedData structure. Also -md option to use an
8277 alternative message digest algorithm for signing.
8281 * Tidy up PKCS#7 routines and add new functions to make it easier to
8282 create PKCS7 structures containing multiple signers. Update smime
8283 application to support multiple signers.
8287 * New -macalg option to pkcs12 utility to allow setting of an alternative
8292 * Initial support for PKCS#5 v2.0 PRFs other than default SHA1 HMAC.
8293 Reorganize PBE internals to lookup from a static table using NIDs,
8294 add support for HMAC PBE OID translation. Add a EVP_CIPHER ctrl:
8295 EVP_CTRL_PBE_PRF_NID this allows a cipher to specify an alternative
8296 PRF which will be automatically used with PBES2.
8300 * Replace the algorithm specific calls to generate keys in "req" with the
8305 * Update PKCS#7 enveloped data routines to use new API. This is now
8306 supported by any public key method supporting the encrypt operation. A
8307 ctrl is added to allow the public key algorithm to examine or modify
8308 the PKCS#7 RecipientInfo structure if it needs to: for RSA this is
8313 * Add a ctrl to asn1 method to allow a public key algorithm to express
8314 a default digest type to use. In most cases this will be SHA1 but some
8315 algorithms (such as GOST) need to specify an alternative digest. The
8316 return value indicates how strong the preference is 1 means optional and
8317 2 is mandatory (that is it is the only supported type). Modify
8318 ASN1_item_sign() to accept a NULL digest argument to indicate it should
8319 use the default md. Update openssl utilities to use the default digest
8320 type for signing if it is not explicitly indicated.
8324 * Use OID cross reference table in ASN1_sign() and ASN1_verify(). New
8325 EVP_MD flag EVP_MD_FLAG_PKEY_METHOD_SIGNATURE. This uses the relevant
8326 signing method from the key type. This effectively removes the link
8327 between digests and public key types.
8331 * Add an OID cross reference table and utility functions. Its purpose is to
8332 translate between signature OIDs such as SHA1WithrsaEncryption and SHA1,
8333 rsaEncryption. This will allow some of the algorithm specific hackery
8334 needed to use the correct OID to be removed.
8338 * Remove algorithm specific dependencies when setting PKCS7_SIGNER_INFO
8339 structures for PKCS7_sign(). They are now set up by the relevant public
8344 * Add provisional EC pkey method with support for ECDSA and ECDH.
8348 * Add support for key derivation (agreement) in the API, DH method and
8353 * Add DSA pkey method and DH pkey methods, extend DH ASN1 method to support
8354 public and private key formats. As a side effect these add additional
8355 command line functionality not previously available: DSA signatures can be
8356 generated and verified using pkeyutl and DH key support and generation in
8363 *Oliver Tappe <zooey@hirschkaefer.de>*
8365 * New make target "install_html_docs" installs HTML renditions of the
8368 *Oliver Tappe <zooey@hirschkaefer.de>*
8370 * New utility "genpkey" this is analogous to "genrsa" etc except it can
8371 generate keys for any algorithm. Extend and update EVP_PKEY_METHOD to
8372 support key and parameter generation and add initial key generation
8373 functionality for RSA.
8377 * Add functions for main EVP_PKEY_method operations. The undocumented
8378 functions `EVP_PKEY_{encrypt,decrypt}` have been renamed to
8379 `EVP_PKEY_{encrypt,decrypt}_old`.
8383 * Initial definitions for EVP_PKEY_METHOD. This will be a high level public
8384 key API, doesn't do much yet.
8388 * New function EVP_PKEY_asn1_get0_info() to retrieve information about
8389 public key algorithms. New option to openssl utility:
8390 "list-public-key-algorithms" to print out info.
8394 * Implement the Supported Elliptic Curves Extension for
8395 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
8399 * Don't free up OIDs in OBJ_cleanup() if they are in use by EVP_MD or
8400 EVP_CIPHER structures to avoid later problems in EVP_cleanup().
8404 * New utilities pkey and pkeyparam. These are similar to algorithm specific
8405 utilities such as rsa, dsa, dsaparam etc except they process any key
8410 * Transfer public key printing routines to EVP_PKEY_ASN1_METHOD. New
8411 functions EVP_PKEY_print_public(), EVP_PKEY_print_private(),
8412 EVP_PKEY_print_param() to print public key data from an EVP_PKEY
8417 * Initial support for pluggable public key ASN1.
8418 De-spaghettify the public key ASN1 handling. Move public and private
8419 key ASN1 handling to a new EVP_PKEY_ASN1_METHOD structure. Relocate
8420 algorithm specific handling to a single module within the relevant
8421 algorithm directory. Add functions to allow (near) opaque processing
8422 of public and private key structures.
8426 * Implement the Supported Point Formats Extension for
8427 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
8431 * Add initial support for RFC 4279 PSK TLS ciphersuites. Add members
8432 for the psk identity [hint] and the psk callback functions to the
8433 SSL_SESSION, SSL and SSL_CTX structure.
8436 PSK-RC4-SHA, PSK-3DES-EDE-CBC-SHA, PSK-AES128-CBC-SHA,
8440 SSL_CTX_use_psk_identity_hint
8441 SSL_get_psk_identity_hint
8442 SSL_get_psk_identity
8443 SSL_use_psk_identity_hint
8445 *Mika Kousa and Pasi Eronen of Nokia Corporation*
8447 * Add RFC 3161 compliant time stamp request creation, response generation
8448 and response verification functionality.
8450 *Zoltán Glózik <zglozik@opentsa.org>, The OpenTSA Project*
8452 * Add initial support for TLS extensions, specifically for the server_name
8453 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
8454 have new members for a host name. The SSL data structure has an
8455 additional member `SSL_CTX *initial_ctx` so that new sessions can be
8456 stored in that context to allow for session resumption, even after the
8457 SSL has been switched to a new SSL_CTX in reaction to a client's
8458 server_name extension.
8460 New functions (subject to change):
8462 SSL_get_servername()
8463 SSL_get_servername_type()
8466 New CTRL codes and macros (subject to change):
8468 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
8469 - SSL_CTX_set_tlsext_servername_callback()
8470 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
8471 - SSL_CTX_set_tlsext_servername_arg()
8472 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
8474 openssl s_client has a new '-servername ...' option.
8476 openssl s_server has new options '-servername_host ...', '-cert2 ...',
8477 '-key2 ...', '-servername_fatal' (subject to change). This allows
8478 testing the HostName extension for a specific single host name ('-cert'
8479 and '-key' remain fallbacks for handshakes without HostName
8480 negotiation). If the unrecognized_name alert has to be sent, this by
8481 default is a warning; it becomes fatal with the '-servername_fatal'
8484 *Peter Sylvester, Remy Allais, Christophe Renou*
8486 * Whirlpool hash implementation is added.
8490 * BIGNUM code on 64-bit SPARCv9 targets is switched from bn(64,64) to
8491 bn(64,32). Because of instruction set limitations it doesn't have
8492 any negative impact on performance. This was done mostly in order
8493 to make it possible to share assembler modules, such as bn_mul_mont
8494 implementations, between 32- and 64-bit builds without hassle.
8498 * Move code previously exiled into file crypto/ec/ec2_smpt.c
8499 to ec2_smpl.c, and no longer require the OPENSSL_EC_BIN_PT_COMP
8504 * New candidate for BIGNUM assembler implementation, bn_mul_mont,
8505 dedicated Montgomery multiplication procedure, is introduced.
8506 BN_MONT_CTX is modified to allow bn_mul_mont to reach for higher
8507 "64-bit" performance on certain 32-bit targets.
8511 * New option SSL_OP_NO_COMP to disable use of compression selectively
8512 in SSL structures. New SSL ctrl to set maximum send fragment size.
8513 Save memory by setting the I/O buffer sizes dynamically instead of
8514 using the maximum available value.
8518 * New option -V for 'openssl ciphers'. This prints the ciphersuite code
8519 in addition to the text details.
8523 * Very, very preliminary EXPERIMENTAL support for printing of general
8524 ASN1 structures. This currently produces rather ugly output and doesn't
8525 handle several customised structures at all.
8529 * Integrated support for PVK file format and some related formats such
8530 as MS PUBLICKEYBLOB and PRIVATEKEYBLOB. Command line switches to support
8531 these in the 'rsa' and 'dsa' utilities.
8535 * Support for PKCS#1 RSAPublicKey format on rsa utility command line.
8539 * Remove the ancient ASN1_METHOD code. This was only ever used in one
8540 place for the (very old) "NETSCAPE" format certificates which are now
8541 handled using new ASN1 code equivalents.
8545 * Let the TLSv1_method() etc. functions return a 'const' SSL_METHOD
8546 pointer and make the SSL_METHOD parameter in SSL_CTX_new,
8547 SSL_CTX_set_ssl_version and SSL_set_ssl_method 'const'.
8551 * Modify CRL distribution points extension code to print out previously
8552 unsupported fields. Enhance extension setting code to allow setting of
8557 * Add print and set support for Issuing Distribution Point CRL extension.
8561 * Change 'Configure' script to enable Camellia by default.
8568 ### Changes between 0.9.8m and 0.9.8n [24 Mar 2010]
8570 * When rejecting SSL/TLS records due to an incorrect version number, never
8571 update s->server with a new major version number. As of
8572 - OpenSSL 0.9.8m if 'short' is a 16-bit type,
8573 - OpenSSL 0.9.8f if 'short' is longer than 16 bits,
8574 the previous behavior could result in a read attempt at NULL when
8575 receiving specific incorrect SSL/TLS records once record payload
8576 protection is active. [CVE-2010-0740][]
8578 *Bodo Moeller, Adam Langley <agl@chromium.org>*
8580 * Fix for CVE-2010-0433 where some kerberos enabled versions of OpenSSL
8581 could be crashed if the relevant tables were not present (e.g. chrooted).
8583 *Tomas Hoger <thoger@redhat.com>*
8585 ### Changes between 0.9.8l and 0.9.8m [25 Feb 2010]
8587 * Always check bn_wexpand() return values for failure. [CVE-2009-3245][]
8589 *Martin Olsson, Neel Mehta*
8591 * Fix X509_STORE locking: Every 'objs' access requires a lock (to
8592 accommodate for stack sorting, always a write lock!).
8596 * On some versions of WIN32 Heap32Next is very slow. This can cause
8597 excessive delays in the RAND_poll(): over a minute. As a workaround
8598 include a time check in the inner Heap32Next loop too.
8602 * The code that handled flushing of data in SSL/TLS originally used the
8603 BIO_CTRL_INFO ctrl to see if any data was pending first. This caused
8604 the problem outlined in PR#1949. The fix suggested there however can
8605 trigger problems with buggy BIO_CTRL_WPENDING (e.g. some versions
8606 of Apache). So instead simplify the code to flush unconditionally.
8607 This should be fine since flushing with no data to flush is a no op.
8611 * Handle TLS versions 2.0 and later properly and correctly use the
8612 highest version of TLS/SSL supported. Although TLS >= 2.0 is some way
8613 off ancient servers have a habit of sticking around for a while...
8617 * Modify compression code so it frees up structures without using the
8618 ex_data callbacks. This works around a problem where some applications
8619 call CRYPTO_cleanup_all_ex_data() before application exit (e.g. when
8620 restarting) then use compression (e.g. SSL with compression) later.
8621 This results in significant per-connection memory leaks and
8622 has caused some security issues including CVE-2008-1678 and
8627 * Constify crypto/cast (i.e., <openssl/cast.h>): a CAST_KEY doesn't
8628 change when encrypting or decrypting.
8632 * Add option SSL_OP_LEGACY_SERVER_CONNECT which will allow clients to
8633 connect and renegotiate with servers which do not support RI.
8634 Until RI is more widely deployed this option is enabled by default.
8638 * Add "missing" ssl ctrls to clear options and mode.
8642 * If client attempts to renegotiate and doesn't support RI respond with
8643 a no_renegotiation alert as required by RFC5746. Some renegotiating
8644 TLS clients will continue a connection gracefully when they receive
8645 the alert. Unfortunately OpenSSL mishandled this alert and would hang
8646 waiting for a server hello which it will never receive. Now we treat a
8647 received no_renegotiation alert as a fatal error. This is because
8648 applications requesting a renegotiation might well expect it to succeed
8649 and would have no code in place to handle the server denying it so the
8650 only safe thing to do is to terminate the connection.
8654 * Add ctrl macro SSL_get_secure_renegotiation_support() which returns 1 if
8655 peer supports secure renegotiation and 0 otherwise. Print out peer
8656 renegotiation support in s_client/s_server.
8660 * Replace the highly broken and deprecated SPKAC certification method with
8661 the updated NID creation version. This should correctly handle UTF8.
8665 * Implement RFC5746. Re-enable renegotiation but require the extension
8666 as needed. Unfortunately, SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION
8667 turns out to be a bad idea. It has been replaced by
8668 SSL_OP_ALLOW_UNSAFE_LEGACY_RENEGOTIATION which can be set with
8669 SSL_CTX_set_options(). This is really not recommended unless you
8670 know what you are doing.
8672 *Eric Rescorla <ekr@networkresonance.com>, Ben Laurie, Steve Henson*
8674 * Fixes to stateless session resumption handling. Use initial_ctx when
8675 issuing and attempting to decrypt tickets in case it has changed during
8676 servername handling. Use a non-zero length session ID when attempting
8677 stateless session resumption: this makes it possible to determine if
8678 a resumption has occurred immediately after receiving server hello
8679 (several places in OpenSSL subtly assume this) instead of later in
8684 * The functions ENGINE_ctrl(), OPENSSL_isservice(),
8685 CMS_get1_RecipientRequest() and RAND_bytes() can return <=0 on error
8686 fixes for a few places where the return code is not checked
8689 *Julia Lawall <julia@diku.dk>*
8691 * Add --strict-warnings option to Configure script to include devteam
8692 warnings in other configurations.
8696 * Add support for --libdir option and LIBDIR variable in makefiles. This
8697 makes it possible to install openssl libraries in locations which
8698 have names other than "lib", for example "/usr/lib64" which some
8701 *Steve Henson, based on patch from Jeremy Utley*
8703 * Don't allow the use of leading 0x80 in OIDs. This is a violation of
8704 X690 8.9.12 and can produce some misleading textual output of OIDs.
8706 *Steve Henson, reported by Dan Kaminsky*
8708 * Delete MD2 from algorithm tables. This follows the recommendation in
8709 several standards that it is not used in new applications due to
8710 several cryptographic weaknesses. For binary compatibility reasons
8711 the MD2 API is still compiled in by default.
8715 * Add compression id to {d2i,i2d}_SSL_SESSION so it is correctly saved
8720 * Rename uni2asc and asc2uni functions to OPENSSL_uni2asc and
8721 OPENSSL_asc2uni conditionally on Netware platforms to avoid a name
8724 *Guenter <lists@gknw.net>*
8726 * Fix the server certificate chain building code to use X509_verify_cert(),
8727 it used to have an ad-hoc builder which was unable to cope with anything
8728 other than a simple chain.
8730 *David Woodhouse <dwmw2@infradead.org>, Steve Henson*
8732 * Don't check self signed certificate signatures in X509_verify_cert()
8733 by default (a flag can override this): it just wastes time without
8734 adding any security. As a useful side effect self signed root CAs
8735 with non-FIPS digests are now usable in FIPS mode.
8739 * In dtls1_process_out_of_seq_message() the check if the current message
8740 is already buffered was missing. For every new message was memory
8741 allocated, allowing an attacker to perform an denial of service attack
8742 with sending out of seq handshake messages until there is no memory
8743 left. Additionally every future message was buffered, even if the
8744 sequence number made no sense and would be part of another handshake.
8745 So only messages with sequence numbers less than 10 in advance will be
8746 buffered. [CVE-2009-1378][]
8748 *Robin Seggelmann, discovered by Daniel Mentz*
8750 * Records are buffered if they arrive with a future epoch to be
8751 processed after finishing the corresponding handshake. There is
8752 currently no limitation to this buffer allowing an attacker to perform
8753 a DOS attack with sending records with future epochs until there is no
8754 memory left. This patch adds the pqueue_size() function to determine
8755 the size of a buffer and limits the record buffer to 100 entries.
8758 *Robin Seggelmann, discovered by Daniel Mentz*
8760 * Keep a copy of frag->msg_header.frag_len so it can be used after the
8761 parent structure is freed. [CVE-2009-1379][]
8765 * Handle non-blocking I/O properly in SSL_shutdown() call.
8767 *Darryl Miles <darryl-mailinglists@netbauds.net>*
8769 * Add `2.5.4.*` OIDs
8771 *Ilya O. <vrghost@gmail.com>*
8773 ### Changes between 0.9.8k and 0.9.8l [5 Nov 2009]
8775 * Disable renegotiation completely - this fixes a severe security
8776 problem [CVE-2009-3555][] at the cost of breaking all
8777 renegotiation. Renegotiation can be re-enabled by setting
8778 SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION in s3->flags at
8779 run-time. This is really not recommended unless you know what
8784 ### Changes between 0.9.8j and 0.9.8k [25 Mar 2009]
8786 * Don't set val to NULL when freeing up structures, it is freed up by
8787 underlying code. If `sizeof(void *) > sizeof(long)` this can result in
8788 zeroing past the valid field. [CVE-2009-0789][]
8790 *Paolo Ganci <Paolo.Ganci@AdNovum.CH>*
8792 * Fix bug where return value of CMS_SignerInfo_verify_content() was not
8793 checked correctly. This would allow some invalid signed attributes to
8794 appear to verify correctly. [CVE-2009-0591][]
8796 *Ivan Nestlerode <inestlerode@us.ibm.com>*
8798 * Reject UniversalString and BMPString types with invalid lengths. This
8799 prevents a crash in ASN1_STRING_print_ex() which assumes the strings have
8800 a legal length. [CVE-2009-0590][]
8804 * Set S/MIME signing as the default purpose rather than setting it
8805 unconditionally. This allows applications to override it at the store
8810 * Permit restricted recursion of ASN1 strings. This is needed in practice
8811 to handle some structures.
8815 * Improve efficiency of mem_gets: don't search whole buffer each time
8818 *Jeremy Shapiro <jnshapir@us.ibm.com>*
8820 * New -hex option for openssl rand.
8824 * Print out UTF8String and NumericString when parsing ASN1.
8828 * Support NumericString type for name components.
8832 * Allow CC in the environment to override the automatically chosen
8833 compiler. Note that nothing is done to ensure flags work with the
8838 ### Changes between 0.9.8i and 0.9.8j [07 Jan 2009]
8840 * Properly check EVP_VerifyFinal() and similar return values
8843 *Ben Laurie, Bodo Moeller, Google Security Team*
8845 * Enable TLS extensions by default.
8849 * Allow the CHIL engine to be loaded, whether the application is
8850 multithreaded or not. (This does not release the developer from the
8851 obligation to set up the dynamic locking callbacks.)
8853 *Sander Temme <sander@temme.net>*
8855 * Use correct exit code if there is an error in dgst command.
8857 *Steve Henson; problem pointed out by Roland Dirlewanger*
8859 * Tweak Configure so that you need to say "experimental-jpake" to enable
8860 JPAKE, and need to use -DOPENSSL_EXPERIMENTAL_JPAKE in applications.
8864 * Add experimental JPAKE support, including demo authentication in
8865 s_client and s_server.
8869 * Set the comparison function in v3_addr_canonize().
8871 *Rob Austein <sra@hactrn.net>*
8873 * Add support for XMPP STARTTLS in s_client.
8875 *Philip Paeps <philip@freebsd.org>*
8877 * Change the server-side SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG behavior
8878 to ensure that even with this option, only ciphersuites in the
8879 server's preference list will be accepted. (Note that the option
8880 applies only when resuming a session, so the earlier behavior was
8881 just about the algorithm choice for symmetric cryptography.)
8885 ### Changes between 0.9.8h and 0.9.8i [15 Sep 2008]
8887 * Fix NULL pointer dereference if a DTLS server received
8888 ChangeCipherSpec as first record [CVE-2009-1386][].
8892 * Fix a state transition in s3_srvr.c and d1_srvr.c
8893 (was using SSL3_ST_CW_CLNT_HELLO_B, should be `..._ST_SW_SRVR_...`).
8897 * The fix in 0.9.8c that supposedly got rid of unsafe
8898 double-checked locking was incomplete for RSA blinding,
8899 addressing just one layer of what turns out to have been
8900 doubly unsafe triple-checked locking.
8902 So now fix this for real by retiring the MONT_HELPER macro
8903 in crypto/rsa/rsa_eay.c.
8905 *Bodo Moeller; problem pointed out by Marius Schilder*
8907 * Various precautionary measures:
8909 - Avoid size_t integer overflow in HASH_UPDATE (md32_common.h).
8911 - Avoid a buffer overflow in d2i_SSL_SESSION() (ssl_asn1.c).
8912 (NB: This would require knowledge of the secret session ticket key
8913 to exploit, in which case you'd be SOL either way.)
8915 - Change bn_nist.c so that it will properly handle input BIGNUMs
8916 outside the expected range.
8918 - Enforce the 'num' check in BN_div() (bn_div.c) for non-BN_DEBUG
8921 *Neel Mehta, Bodo Moeller*
8923 * Allow engines to be "soft loaded" - i.e. optionally don't die if
8924 the load fails. Useful for distros.
8926 *Ben Laurie and the FreeBSD team*
8928 * Add support for Local Machine Keyset attribute in PKCS#12 files.
8932 * Fix BN_GF2m_mod_arr() top-bit cleanup code.
8936 * Expand ENGINE to support engine supplied SSL client certificate functions.
8938 This work was sponsored by Logica.
8942 * Add CryptoAPI ENGINE to support use of RSA and DSA keys held in Windows
8943 keystores. Support for SSL/TLS client authentication too.
8944 Not compiled unless enable-capieng specified to Configure.
8946 This work was sponsored by Logica.
8950 * Fix bug in X509_ATTRIBUTE creation: don't set attribute using
8951 ASN1_TYPE_set1 if MBSTRING flag set. This bug would crash certain
8952 attribute creation routines such as certificate requests and PKCS#12
8957 ### Changes between 0.9.8g and 0.9.8h [28 May 2008]
8959 * Fix flaw if 'Server Key exchange message' is omitted from a TLS
8960 handshake which could lead to a client crash as found using the
8961 Codenomicon TLS test suite [CVE-2008-1672][]
8963 *Steve Henson, Mark Cox*
8965 * Fix double free in TLS server name extensions which could lead to
8966 a remote crash found by Codenomicon TLS test suite [CVE-2008-0891][]
8970 * Clear error queue in SSL_CTX_use_certificate_chain_file()
8972 Clear the error queue to ensure that error entries left from
8973 older function calls do not interfere with the correct operation.
8975 *Lutz Jaenicke, Erik de Castro Lopo*
8977 * Remove root CA certificates of commercial CAs:
8979 The OpenSSL project does not recommend any specific CA and does not
8980 have any policy with respect to including or excluding any CA.
8981 Therefore it does not make any sense to ship an arbitrary selection
8982 of root CA certificates with the OpenSSL software.
8986 * RSA OAEP patches to fix two separate invalid memory reads.
8987 The first one involves inputs when 'lzero' is greater than
8988 'SHA_DIGEST_LENGTH' (it would read about SHA_DIGEST_LENGTH bytes
8989 before the beginning of from). The second one involves inputs where
8990 the 'db' section contains nothing but zeroes (there is a one-byte
8991 invalid read after the end of 'db').
8993 *Ivan Nestlerode <inestlerode@us.ibm.com>*
8995 * Partial backport from 0.9.9-dev:
8997 Introduce bn_mul_mont (dedicated Montgomery multiplication
8998 procedure) as a candidate for BIGNUM assembler implementation.
8999 While 0.9.9-dev uses assembler for various architectures, only
9000 x86_64 is available by default here in the 0.9.8 branch, and
9001 32-bit x86 is available through a compile-time setting.
9003 To try the 32-bit x86 assembler implementation, use Configure
9004 option "enable-montasm" (which exists only for this backport).
9006 As "enable-montasm" for 32-bit x86 disclaims code stability
9007 anyway, in this constellation we activate additional code
9008 backported from 0.9.9-dev for further performance improvements,
9009 namely BN_from_montgomery_word. (To enable this otherwise,
9010 e.g. x86_64, try `-DMONT_FROM_WORD___NON_DEFAULT_0_9_8_BUILD`.)
9012 *Andy Polyakov (backport partially by Bodo Moeller)*
9014 * Add TLS session ticket callback. This allows an application to set
9015 TLS ticket cipher and HMAC keys rather than relying on hardcoded fixed
9016 values. This is useful for key rollover for example where several key
9017 sets may exist with different names.
9021 * Reverse ENGINE-internal logic for caching default ENGINE handles.
9022 This was broken until now in 0.9.8 releases, such that the only way
9023 a registered ENGINE could be used (assuming it initialises
9024 successfully on the host) was to explicitly set it as the default
9025 for the relevant algorithms. This is in contradiction with 0.9.7
9026 behaviour and the documentation. With this fix, when an ENGINE is
9027 registered into a given algorithm's table of implementations, the
9028 'uptodate' flag is reset so that auto-discovery will be used next
9029 time a new context for that algorithm attempts to select an
9032 *Ian Lister (tweaked by Geoff Thorpe)*
9034 * Backport of CMS code to OpenSSL 0.9.8. This differs from the 0.9.9
9035 implementation in the following ways:
9037 Lack of EVP_PKEY_ASN1_METHOD means algorithm parameters have to be
9040 Lack of BER streaming support means one pass streaming processing is
9041 only supported if data is detached: setting the streaming flag is
9042 ignored for embedded content.
9044 CMS support is disabled by default and must be explicitly enabled
9045 with the enable-cms configuration option.
9049 * Update the GMP engine glue to do direct copies between BIGNUM and
9050 mpz_t when openssl and GMP use the same limb size. Otherwise the
9051 existing "conversion via a text string export" trick is still used.
9053 *Paul Sheer <paulsheer@gmail.com>*
9055 * Zlib compression BIO. This is a filter BIO which compressed and
9056 uncompresses any data passed through it.
9060 * Add AES_wrap_key() and AES_unwrap_key() functions to implement
9061 RFC3394 compatible AES key wrapping.
9065 * Add utility functions to handle ASN1 structures. ASN1_STRING_set0():
9066 sets string data without copying. X509_ALGOR_set0() and
9067 X509_ALGOR_get0(): set and retrieve X509_ALGOR (AlgorithmIdentifier)
9068 data. Attribute function X509at_get0_data_by_OBJ(): retrieves data
9069 from an X509_ATTRIBUTE structure optionally checking it occurs only
9070 once. ASN1_TYPE_set1(): set and ASN1_TYPE structure copying supplied
9075 * Fix BN flag handling in RSA_eay_mod_exp() and BN_MONT_CTX_set()
9076 to get the expected BN_FLG_CONSTTIME behavior.
9078 *Bodo Moeller (Google)*
9082 - fixed wrong usage of ioctlsocket() when build for LIBC BSD sockets
9083 - fixed do_tests.pl to run the test suite with CLIB builds too (CLIB_OPT)
9084 - added some more tests to do_tests.pl
9085 - fixed RunningProcess usage so that it works with newer LIBC NDKs too
9086 - removed usage of BN_LLONG for CLIB builds to avoid runtime dependency
9087 - added new Configure targets netware-clib-bsdsock, netware-clib-gcc,
9088 netware-clib-bsdsock-gcc, netware-libc-bsdsock-gcc
9089 - various changes to netware.pl to enable gcc-cross builds on Win32
9091 - changed crypto/bio/b_sock.c to work with macro functions (CLIB BSD)
9092 - various changes to fix missing prototype warnings
9093 - fixed x86nasm.pl to create correct asm files for NASM COFF output
9094 - added AES, WHIRLPOOL and CPUID assembler code to build files
9095 - added missing AES assembler make rules to mk1mf.pl
9096 - fixed order of includes in apps/ocsp.c so that e_os.h settings apply
9098 *Guenter Knauf <eflash@gmx.net>*
9100 * Implement certificate status request TLS extension defined in RFC3546.
9101 A client can set the appropriate parameters and receive the encoded
9102 OCSP response via a callback. A server can query the supplied parameters
9103 and set the encoded OCSP response in the callback. Add simplified examples
9104 to s_client and s_server.
9108 ### Changes between 0.9.8f and 0.9.8g [19 Oct 2007]
9111 + Binary incompatibility of ssl_ctx_st structure
9112 + DTLS interoperation with non-compliant servers
9113 + Don't call get_session_cb() without proposed session
9114 + Fix ia64 assembler code
9116 *Andy Polyakov, Steve Henson*
9118 ### Changes between 0.9.8e and 0.9.8f [11 Oct 2007]
9120 * DTLS Handshake overhaul. There were longstanding issues with
9121 OpenSSL DTLS implementation, which were making it impossible for
9122 RFC 4347 compliant client to communicate with OpenSSL server.
9123 Unfortunately just fixing these incompatibilities would "cut off"
9124 pre-0.9.8f clients. To allow for hassle free upgrade post-0.9.8e
9125 server keeps tolerating non RFC compliant syntax. The opposite is
9126 not true, 0.9.8f client can not communicate with earlier server.
9127 This update even addresses CVE-2007-4995.
9131 * Changes to avoid need for function casts in OpenSSL: some compilers
9132 (gcc 4.2 and later) reject their use.
9133 *Kurt Roeckx <kurt@roeckx.be>, Peter Hartley <pdh@utter.chaos.org.uk>,
9136 * Add RFC4507 support to OpenSSL. This includes the corrections in
9137 RFC4507bis. The encrypted ticket format is an encrypted encoded
9138 SSL_SESSION structure, that way new session features are automatically
9141 If a client application caches session in an SSL_SESSION structure
9142 support is transparent because tickets are now stored in the encoded
9145 The SSL_CTX structure automatically generates keys for ticket
9146 protection in servers so again support should be possible
9147 with no application modification.
9149 If a client or server wishes to disable RFC4507 support then the option
9150 SSL_OP_NO_TICKET can be set.
9152 Add a TLS extension debugging callback to allow the contents of any client
9153 or server extensions to be examined.
9155 This work was sponsored by Google.
9159 * Add initial support for TLS extensions, specifically for the server_name
9160 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
9161 have new members for a host name. The SSL data structure has an
9162 additional member `SSL_CTX *initial_ctx` so that new sessions can be
9163 stored in that context to allow for session resumption, even after the
9164 SSL has been switched to a new SSL_CTX in reaction to a client's
9165 server_name extension.
9167 New functions (subject to change):
9169 SSL_get_servername()
9170 SSL_get_servername_type()
9173 New CTRL codes and macros (subject to change):
9175 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
9176 - SSL_CTX_set_tlsext_servername_callback()
9177 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
9178 - SSL_CTX_set_tlsext_servername_arg()
9179 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
9181 openssl s_client has a new '-servername ...' option.
9183 openssl s_server has new options '-servername_host ...', '-cert2 ...',
9184 '-key2 ...', '-servername_fatal' (subject to change). This allows
9185 testing the HostName extension for a specific single host name ('-cert'
9186 and '-key' remain fallbacks for handshakes without HostName
9187 negotiation). If the unrecognized_name alert has to be sent, this by
9188 default is a warning; it becomes fatal with the '-servername_fatal'
9191 *Peter Sylvester, Remy Allais, Christophe Renou, Steve Henson*
9193 * Add AES and SSE2 assembly language support to VC++ build.
9197 * Mitigate attack on final subtraction in Montgomery reduction.
9201 * Fix crypto/ec/ec_mult.c to work properly with scalars of value 0
9202 (which previously caused an internal error).
9206 * Squeeze another 10% out of IGE mode when in != out.
9210 * AES IGE mode speedup.
9212 *Dean Gaudet (Google)*
9214 * Add the Korean symmetric 128-bit cipher SEED (see
9215 <http://www.kisa.or.kr/kisa/seed/jsp/seed_eng.jsp>) and
9216 add SEED ciphersuites from RFC 4162:
9218 TLS_RSA_WITH_SEED_CBC_SHA = "SEED-SHA"
9219 TLS_DHE_DSS_WITH_SEED_CBC_SHA = "DHE-DSS-SEED-SHA"
9220 TLS_DHE_RSA_WITH_SEED_CBC_SHA = "DHE-RSA-SEED-SHA"
9221 TLS_DH_anon_WITH_SEED_CBC_SHA = "ADH-SEED-SHA"
9223 To minimize changes between patchlevels in the OpenSSL 0.9.8
9224 series, SEED remains excluded from compilation unless OpenSSL
9225 is configured with 'enable-seed'.
9227 *KISA, Bodo Moeller*
9229 * Mitigate branch prediction attacks, which can be practical if a
9230 single processor is shared, allowing a spy process to extract
9231 information. For detailed background information, see
9232 <http://eprint.iacr.org/2007/039> (O. Aciicmez, S. Gueron,
9233 J.-P. Seifert, "New Branch Prediction Vulnerabilities in OpenSSL
9234 and Necessary Software Countermeasures"). The core of the change
9235 are new versions BN_div_no_branch() and
9236 BN_mod_inverse_no_branch() of BN_div() and BN_mod_inverse(),
9237 respectively, which are slower, but avoid the security-relevant
9238 conditional branches. These are automatically called by BN_div()
9239 and BN_mod_inverse() if the flag BN_FLG_CONSTTIME is set for one
9240 of the input BIGNUMs. Also, BN_is_bit_set() has been changed to
9241 remove a conditional branch.
9243 BN_FLG_CONSTTIME is the new name for the previous
9244 BN_FLG_EXP_CONSTTIME flag, since it now affects more than just
9245 modular exponentiation. (Since OpenSSL 0.9.7h, setting this flag
9246 in the exponent causes BN_mod_exp_mont() to use the alternative
9247 implementation in BN_mod_exp_mont_consttime().) The old name
9248 remains as a deprecated alias.
9250 Similarly, RSA_FLAG_NO_EXP_CONSTTIME is replaced by a more general
9251 RSA_FLAG_NO_CONSTTIME flag since the RSA implementation now uses
9252 constant-time implementations for more than just exponentiation.
9253 Here too the old name is kept as a deprecated alias.
9255 BN_BLINDING_new() will now use BN_dup() for the modulus so that
9256 the BN_BLINDING structure gets an independent copy of the
9257 modulus. This means that the previous `BIGNUM *m` argument to
9258 BN_BLINDING_new() and to BN_BLINDING_create_param() now
9259 essentially becomes `const BIGNUM *m`, although we can't actually
9260 change this in the header file before 0.9.9. It allows
9261 RSA_setup_blinding() to use BN_with_flags() on the modulus to
9262 enable BN_FLG_CONSTTIME.
9264 *Matthew D Wood (Intel Corp)*
9266 * In the SSL/TLS server implementation, be strict about session ID
9267 context matching (which matters if an application uses a single
9268 external cache for different purposes). Previously,
9269 out-of-context reuse was forbidden only if SSL_VERIFY_PEER was
9270 set. This did ensure strict client verification, but meant that,
9271 with applications using a single external cache for quite
9272 different requirements, clients could circumvent ciphersuite
9273 restrictions for a given session ID context by starting a session
9274 in a different context.
9278 * Include "!eNULL" in SSL_DEFAULT_CIPHER_LIST to make sure that
9279 a ciphersuite string such as "DEFAULT:RSA" cannot enable
9280 authentication-only ciphersuites.
9284 * Update the SSL_get_shared_ciphers() fix CVE-2006-3738 which was
9285 not complete and could lead to a possible single byte overflow
9286 [CVE-2007-5135][] [Ben Laurie]
9288 ### Changes between 0.9.8d and 0.9.8e [23 Feb 2007]
9290 * Since AES128 and AES256 (and similarly Camellia128 and
9291 Camellia256) share a single mask bit in the logic of
9292 ssl/ssl_ciph.c, the code for masking out disabled ciphers needs a
9293 kludge to work properly if AES128 is available and AES256 isn't
9294 (or if Camellia128 is available and Camellia256 isn't).
9298 * Fix the BIT STRING encoding generated by crypto/ec/ec_asn1.c
9299 (within i2d_ECPrivateKey, i2d_ECPKParameters, i2d_ECParameters):
9300 When a point or a seed is encoded in a BIT STRING, we need to
9301 prevent the removal of trailing zero bits to get the proper DER
9302 encoding. (By default, crypto/asn1/a_bitstr.c assumes the case
9303 of a NamedBitList, for which trailing 0 bits need to be removed.)
9307 * Have SSL/TLS server implementation tolerate "mismatched" record
9308 protocol version while receiving ClientHello even if the
9309 ClientHello is fragmented. (The server can't insist on the
9310 particular protocol version it has chosen before the ServerHello
9311 message has informed the client about his choice.)
9315 * Add RFC 3779 support.
9317 *Rob Austein for ARIN, Ben Laurie*
9319 * Load error codes if they are not already present instead of using a
9320 static variable. This allows them to be cleanly unloaded and reloaded.
9321 Improve header file function name parsing.
9325 * extend SMTP and IMAP protocol emulation in s_client to use EHLO
9326 or CAPABILITY handshake as required by RFCs.
9330 ### Changes between 0.9.8c and 0.9.8d [28 Sep 2006]
9332 * Introduce limits to prevent malicious keys being able to
9333 cause a denial of service. [CVE-2006-2940][]
9335 *Steve Henson, Bodo Moeller*
9337 * Fix ASN.1 parsing of certain invalid structures that can result
9338 in a denial of service. [CVE-2006-2937][] [Steve Henson]
9340 * Fix buffer overflow in SSL_get_shared_ciphers() function.
9341 [CVE-2006-3738][] [Tavis Ormandy and Will Drewry, Google Security Team]
9343 * Fix SSL client code which could crash if connecting to a
9344 malicious SSLv2 server. [CVE-2006-4343][]
9346 *Tavis Ormandy and Will Drewry, Google Security Team*
9348 * Since 0.9.8b, ciphersuite strings naming explicit ciphersuites
9349 match only those. Before that, "AES256-SHA" would be interpreted
9350 as a pattern and match "AES128-SHA" too (since AES128-SHA got
9351 the same strength classification in 0.9.7h) as we currently only
9352 have a single AES bit in the ciphersuite description bitmap.
9353 That change, however, also applied to ciphersuite strings such as
9354 "RC4-MD5" that intentionally matched multiple ciphersuites --
9355 namely, SSL 2.0 ciphersuites in addition to the more common ones
9356 from SSL 3.0/TLS 1.0.
9358 So we change the selection algorithm again: Naming an explicit
9359 ciphersuite selects this one ciphersuite, and any other similar
9360 ciphersuite (same bitmap) from *other* protocol versions.
9361 Thus, "RC4-MD5" again will properly select both the SSL 2.0
9362 ciphersuite and the SSL 3.0/TLS 1.0 ciphersuite.
9364 Since SSL 2.0 does not have any ciphersuites for which the
9365 128/256 bit distinction would be relevant, this works for now.
9366 The proper fix will be to use different bits for AES128 and
9367 AES256, which would have avoided the problems from the beginning;
9368 however, bits are scarce, so we can only do this in a new release
9369 (not just a patchlevel) when we can change the SSL_CIPHER
9370 definition to split the single 'unsigned long mask' bitmap into
9371 multiple values to extend the available space.
9375 ### Changes between 0.9.8b and 0.9.8c [05 Sep 2006]
9377 * Avoid PKCS #1 v1.5 signature attack discovered by Daniel Bleichenbacher
9378 [CVE-2006-4339][] [Ben Laurie and Google Security Team]
9380 * Add AES IGE and biIGE modes.
9384 * Change the Unix randomness entropy gathering to use poll() when
9385 possible instead of select(), since the latter has some
9386 undesirable limitations.
9388 *Darryl Miles via Richard Levitte and Bodo Moeller*
9390 * Disable "ECCdraft" ciphersuites more thoroughly. Now special
9391 treatment in ssl/ssl_ciph.s makes sure that these ciphersuites
9392 cannot be implicitly activated as part of, e.g., the "AES" alias.
9393 However, please upgrade to OpenSSL 0.9.9[-dev] for
9394 non-experimental use of the ECC ciphersuites to get TLS extension
9395 support, which is required for curve and point format negotiation
9396 to avoid potential handshake problems.
9400 * Disable rogue ciphersuites:
9402 - SSLv2 0x08 0x00 0x80 ("RC4-64-MD5")
9403 - SSLv3/TLSv1 0x00 0x61 ("EXP1024-RC2-CBC-MD5")
9404 - SSLv3/TLSv1 0x00 0x60 ("EXP1024-RC4-MD5")
9406 The latter two were purportedly from
9407 draft-ietf-tls-56-bit-ciphersuites-0[01].txt, but do not really
9410 Also deactivate the remaining ciphersuites from
9411 draft-ietf-tls-56-bit-ciphersuites-01.txt. These are just as
9412 unofficial, and the ID has long expired.
9416 * Fix RSA blinding Heisenbug (problems sometimes occurred on
9417 dual-core machines) and other potential thread-safety issues.
9421 * Add the symmetric cipher Camellia (128-bit, 192-bit, 256-bit key
9422 versions), which is now available for royalty-free use
9423 (see <http://info.isl.ntt.co.jp/crypt/eng/info/chiteki.html>).
9424 Also, add Camellia TLS ciphersuites from RFC 4132.
9426 To minimize changes between patchlevels in the OpenSSL 0.9.8
9427 series, Camellia remains excluded from compilation unless OpenSSL
9428 is configured with 'enable-camellia'.
9432 * Disable the padding bug check when compression is in use. The padding
9433 bug check assumes the first packet is of even length, this is not
9434 necessarily true if compression is enabled and can result in false
9435 positives causing handshake failure. The actual bug test is ancient
9436 code so it is hoped that implementations will either have fixed it by
9437 now or any which still have the bug do not support compression.
9441 ### Changes between 0.9.8a and 0.9.8b [04 May 2006]
9443 * When applying a cipher rule check to see if string match is an explicit
9444 cipher suite and only match that one cipher suite if it is.
9448 * Link in manifests for VC++ if needed.
9450 *Austin Ziegler <halostatue@gmail.com>*
9452 * Update support for ECC-based TLS ciphersuites according to
9453 draft-ietf-tls-ecc-12.txt with proposed changes (but without
9454 TLS extensions, which are supported starting with the 0.9.9
9455 branch, not in the OpenSSL 0.9.8 branch).
9459 * New functions EVP_CIPHER_CTX_new() and EVP_CIPHER_CTX_free() to support
9460 opaque EVP_CIPHER_CTX handling.
9464 * Fixes and enhancements to zlib compression code. We now only use
9465 "zlib1.dll" and use the default `__cdecl` calling convention on Win32
9466 to conform with the standards mentioned here:
9467 <http://www.zlib.net/DLL_FAQ.txt>
9468 Static zlib linking now works on Windows and the new --with-zlib-include
9469 --with-zlib-lib options to Configure can be used to supply the location
9470 of the headers and library. Gracefully handle case where zlib library
9475 * Several fixes and enhancements to the OID generation code. The old code
9476 sometimes allowed invalid OIDs (1.X for X >= 40 for example), couldn't
9477 handle numbers larger than ULONG_MAX, truncated printing and had a
9478 non standard OBJ_obj2txt() behaviour.
9482 * Add support for building of engines under engine/ as shared libraries
9483 under VC++ build system.
9487 * Corrected the numerous bugs in the Win32 path splitter in DSO.
9488 Hopefully, we will not see any false combination of paths any more.
9492 ### Changes between 0.9.8 and 0.9.8a [11 Oct 2005]
9494 * Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING
9495 (part of SSL_OP_ALL). This option used to disable the
9496 countermeasure against man-in-the-middle protocol-version
9497 rollback in the SSL 2.0 server implementation, which is a bad
9498 idea. [CVE-2005-2969][]
9500 *Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center
9501 for Information Security, National Institute of Advanced Industrial
9502 Science and Technology [AIST], Japan)*
9504 * Add two function to clear and return the verify parameter flags.
9508 * Keep cipherlists sorted in the source instead of sorting them at
9509 runtime, thus removing the need for a lock.
9513 * Avoid some small subgroup attacks in Diffie-Hellman.
9515 *Nick Mathewson and Ben Laurie*
9517 * Add functions for well-known primes.
9521 * Extended Windows CE support.
9523 *Satoshi Nakamura and Andy Polyakov*
9525 * Initialize SSL_METHOD structures at compile time instead of during
9526 runtime, thus removing the need for a lock.
9530 * Make PKCS7_decrypt() work even if no certificate is supplied by
9531 attempting to decrypt each encrypted key in turn. Add support to
9536 ### Changes between 0.9.7h and 0.9.8 [05 Jul 2005]
9538 [NB: OpenSSL 0.9.7i and later 0.9.7 patch levels were released after
9541 * Add libcrypto.pc and libssl.pc for those who feel they need them.
9545 * Change CA.sh and CA.pl so they don't bundle the CSR and the private
9546 key into the same file any more.
9550 * Add initial support for Win64, both IA64 and AMD64/x64 flavors.
9554 * Add -utf8 command line and config file option to 'ca'.
9556 *Stefan <stf@udoma.org*
9558 * Removed the macro des_crypt(), as it seems to conflict with some
9559 libraries. Use DES_crypt().
9563 * Correct naming of the 'chil' and '4758cca' ENGINEs. This
9564 involves renaming the source and generated shared-libs for
9565 both. The engines will accept the corrected or legacy ids
9566 ('ncipher' and '4758_cca' respectively) when binding. NB,
9567 this only applies when building 'shared'.
9569 *Corinna Vinschen <vinschen@redhat.com> and Geoff Thorpe*
9571 * Add attribute functions to EVP_PKEY structure. Modify
9572 PKCS12_create() to recognize a CSP name attribute and
9573 use it. Make -CSP option work again in pkcs12 utility.
9577 * Add new functionality to the bn blinding code:
9578 - automatic re-creation of the BN_BLINDING parameters after
9579 a fixed number of uses (currently 32)
9580 - add new function for parameter creation
9581 - introduce flags to control the update behaviour of the
9582 BN_BLINDING parameters
9583 - hide BN_BLINDING structure
9584 Add a second BN_BLINDING slot to the RSA structure to improve
9585 performance when a single RSA object is shared among several
9590 * Add support for DTLS.
9592 *Nagendra Modadugu <nagendra@cs.stanford.edu> and Ben Laurie*
9594 * Add support for DER encoded private keys (SSL_FILETYPE_ASN1)
9595 to SSL_CTX_use_PrivateKey_file() and SSL_use_PrivateKey_file()
9599 * Remove buggy and incomplete DH cert support from
9600 ssl/ssl_rsa.c and ssl/s3_both.c
9604 * Use SHA-1 instead of MD5 as the default digest algorithm for
9605 the apps/openssl applications.
9609 * Compile clean with "-Wall -Wmissing-prototypes
9610 -Wstrict-prototypes -Wmissing-declarations -Werror". Currently
9611 DEBUG_SAFESTACK must also be set.
9615 * Change ./Configure so that certain algorithms can be disabled by default.
9616 The new counterpiece to "no-xxx" is "enable-xxx".
9618 The patented RC5 and MDC2 algorithms will now be disabled unless
9619 "enable-rc5" and "enable-mdc2", respectively, are specified.
9621 (IDEA remains enabled despite being patented. This is because IDEA
9622 is frequently required for interoperability, and there is no license
9623 fee for non-commercial use. As before, "no-idea" can be used to
9624 avoid this algorithm.)
9628 * Add processing of proxy certificates (see RFC 3820). This work was
9629 sponsored by KTH (The Royal Institute of Technology in Stockholm) and
9630 EGEE (Enabling Grids for E-science in Europe).
9634 * RC4 performance overhaul on modern architectures/implementations, such
9635 as Intel P4, IA-64 and AMD64.
9639 * New utility extract-section.pl. This can be used specify an alternative
9640 section number in a pod file instead of having to treat each file as
9641 a separate case in Makefile. This can be done by adding two lines to the
9644 =for comment openssl_section:XXX
9646 The blank line is mandatory.
9650 * New arguments -certform, -keyform and -pass for s_client and s_server
9651 to allow alternative format key and certificate files and passphrase
9656 * New structure X509_VERIFY_PARAM which combines current verify parameters,
9657 update associated structures and add various utility functions.
9659 Add new policy related verify parameters, include policy checking in
9660 standard verify code. Enhance 'smime' application with extra parameters
9661 to support policy checking and print out.
9665 * Add a new engine to support VIA PadLock ACE extensions in the VIA C3
9666 Nehemiah processors. These extensions support AES encryption in hardware
9667 as well as RNG (though RNG support is currently disabled).
9669 *Michal Ludvig <michal@logix.cz>, with help from Andy Polyakov*
9671 * Deprecate `BN_[get|set]_params()` functions (they were ignored internally).
9675 * New FIPS 180-2 algorithms, SHA-224/-256/-384/-512 are implemented.
9677 *Andy Polyakov and a number of other people*
9679 * Improved PowerPC platform support. Most notably BIGNUM assembler
9680 implementation contributed by IBM.
9682 *Suresh Chari, Peter Waltenberg, Andy Polyakov*
9684 * The new 'RSA_generate_key_ex' function now takes a BIGNUM for the public
9685 exponent rather than 'unsigned long'. There is a corresponding change to
9686 the new 'rsa_keygen' element of the RSA_METHOD structure.
9688 *Jelte Jansen, Geoff Thorpe*
9690 * Functionality for creating the initial serial number file is now
9691 moved from CA.pl to the 'ca' utility with a new option -create_serial.
9693 (Before OpenSSL 0.9.7e, CA.pl used to initialize the serial
9694 number file to 1, which is bound to cause problems. To avoid
9695 the problems while respecting compatibility between different 0.9.7
9696 patchlevels, 0.9.7e employed 'openssl x509 -next_serial' in
9697 CA.pl for serial number initialization. With the new release 0.9.8,
9698 we can fix the problem directly in the 'ca' utility.)
9702 * Reduced header interdependencies by declaring more opaque objects in
9703 ossl_typ.h. As a consequence, including some headers (eg. engine.h) will
9704 give fewer recursive includes, which could break lazy source code - so
9705 this change is covered by the OPENSSL_NO_DEPRECATED symbol. As always,
9706 developers should define this symbol when building and using openssl to
9707 ensure they track the recommended behaviour, interfaces, [etc], but
9708 backwards-compatible behaviour prevails when this isn't defined.
9712 * New function X509_POLICY_NODE_print() which prints out policy nodes.
9716 * Add new EVP function EVP_CIPHER_CTX_rand_key and associated functionality.
9717 This will generate a random key of the appropriate length based on the
9718 cipher context. The EVP_CIPHER can provide its own random key generation
9719 routine to support keys of a specific form. This is used in the des and
9720 3des routines to generate a key of the correct parity. Update S/MIME
9721 code to use new functions and hence generate correct parity DES keys.
9722 Add EVP_CHECK_DES_KEY #define to return an error if the key is not
9723 valid (weak or incorrect parity).
9727 * Add a local set of CRLs that can be used by X509_verify_cert() as well
9728 as looking them up. This is useful when the verified structure may contain
9729 CRLs, for example PKCS#7 signedData. Modify PKCS7_verify() to use any CRLs
9730 present unless the new PKCS7_NO_CRL flag is asserted.
9734 * Extend ASN1 oid configuration module. It now additionally accepts the
9737 shortName = some long name, 1.2.3.4
9741 * Reimplemented the BN_CTX implementation. There is now no more static
9742 limitation on the number of variables it can handle nor the depth of the
9743 "stack" handling for BN_CTX_start()/BN_CTX_end() pairs. The stack
9744 information can now expand as required, and rather than having a single
9745 static array of bignums, BN_CTX now uses a linked-list of such arrays
9746 allowing it to expand on demand whilst maintaining the usefulness of
9747 BN_CTX's "bundling".
9751 * Add a missing BN_CTX parameter to the 'rsa_mod_exp' callback in RSA_METHOD
9752 to allow all RSA operations to function using a single BN_CTX.
9756 * Preliminary support for certificate policy evaluation and checking. This
9757 is initially intended to pass the tests outlined in "Conformance Testing
9758 of Relying Party Client Certificate Path Processing Logic" v1.07.
9762 * bn_dup_expand() has been deprecated, it was introduced in 0.9.7 and
9763 remained unused and not that useful. A variety of other little bignum
9764 tweaks and fixes have also been made continuing on from the audit (see
9769 * Constify all or almost all d2i, c2i, s2i and r2i functions, along with
9770 associated ASN1, EVP and SSL functions and old ASN1 macros.
9774 * BN_zero() only needs to set 'top' and 'neg' to zero for correct results,
9775 and this should never fail. So the return value from the use of
9776 BN_set_word() (which can fail due to needless expansion) is now deprecated;
9777 if OPENSSL_NO_DEPRECATED is defined, BN_zero() is a void macro.
9781 * BN_CTX_get() should return zero-valued bignums, providing the same
9782 initialised value as BN_new().
9784 *Geoff Thorpe, suggested by Ulf Möller*
9786 * Support for inhibitAnyPolicy certificate extension.
9790 * An audit of the BIGNUM code is underway, for which debugging code is
9791 enabled when BN_DEBUG is defined. This makes stricter enforcements on what
9792 is considered valid when processing BIGNUMs, and causes execution to
9793 assert() when a problem is discovered. If BN_DEBUG_RAND is defined,
9794 further steps are taken to deliberately pollute unused data in BIGNUM
9795 structures to try and expose faulty code further on. For now, openssl will
9796 (in its default mode of operation) continue to tolerate the inconsistent
9797 forms that it has tolerated in the past, but authors and packagers should
9798 consider trying openssl and their own applications when compiled with
9799 these debugging symbols defined. It will help highlight potential bugs in
9800 their own code, and will improve the test coverage for OpenSSL itself. At
9801 some point, these tighter rules will become openssl's default to improve
9802 maintainability, though the assert()s and other overheads will remain only
9803 in debugging configurations. See bn.h for more details.
9805 *Geoff Thorpe, Nils Larsch, Ulf Möller*
9807 * BN_CTX_init() has been deprecated, as BN_CTX is an opaque structure
9808 that can only be obtained through BN_CTX_new() (which implicitly
9809 initialises it). The presence of this function only made it possible
9810 to overwrite an existing structure (and cause memory leaks).
9814 * Because of the callback-based approach for implementing LHASH as a
9815 template type, lh_insert() adds opaque objects to hash-tables and
9816 lh_doall() or lh_doall_arg() are typically used with a destructor callback
9817 to clean up those corresponding objects before destroying the hash table
9818 (and losing the object pointers). So some over-zealous constifications in
9819 LHASH have been relaxed so that lh_insert() does not take (nor store) the
9820 objects as "const" and the `lh_doall[_arg]` callback wrappers are not
9821 prototyped to have "const" restrictions on the object pointers they are
9822 given (and so aren't required to cast them away any more).
9826 * The tmdiff.h API was so ugly and minimal that our own timing utility
9827 (speed) prefers to use its own implementation. The two implementations
9828 haven't been consolidated as yet (volunteers?) but the tmdiff API has had
9829 its object type properly exposed (MS_TM) instead of casting to/from
9830 `char *`. This may still change yet if someone realises MS_TM and
9832 aren't necessarily the greatest nomenclatures - but this is what was used
9833 internally to the implementation so I've used that for now.
9837 * Ensure that deprecated functions do not get compiled when
9838 OPENSSL_NO_DEPRECATED is defined. Some "openssl" subcommands and a few of
9839 the self-tests were still using deprecated key-generation functions so
9840 these have been updated also.
9844 * Reorganise PKCS#7 code to separate the digest location functionality
9845 into PKCS7_find_digest(), digest addition into PKCS7_bio_add_digest().
9846 New function PKCS7_set_digest() to set the digest type for PKCS#7
9847 digestedData type. Add additional code to correctly generate the
9848 digestedData type and add support for this type in PKCS7 initialization
9853 * New function PKCS7_set0_type_other() this initializes a PKCS7
9854 structure of type "other".
9858 * Fix prime generation loop in crypto/bn/bn_prime.pl by making
9859 sure the loop does correctly stop and breaking ("division by zero")
9860 modulus operations are not performed. The (pre-generated) prime
9861 table crypto/bn/bn_prime.h was already correct, but it could not be
9862 re-generated on some platforms because of the "division by zero"
9863 situation in the script.
9865 *Ralf S. Engelschall*
9867 * Update support for ECC-based TLS ciphersuites according to
9868 draft-ietf-tls-ecc-03.txt: the KDF1 key derivation function with
9869 SHA-1 now is only used for "small" curves (where the
9870 representation of a field element takes up to 24 bytes); for
9871 larger curves, the field element resulting from ECDH is directly
9872 used as premaster secret.
9874 *Douglas Stebila (Sun Microsystems Laboratories)*
9876 * Add code for kP+lQ timings to crypto/ec/ectest.c, and add SEC2
9877 curve secp160r1 to the tests.
9879 *Douglas Stebila (Sun Microsystems Laboratories)*
9881 * Add the possibility to load symbols globally with DSO.
9883 *Götz Babin-Ebell <babin-ebell@trustcenter.de> via Richard Levitte*
9885 * Add the functions ERR_set_mark() and ERR_pop_to_mark() for better
9886 control of the error stack.
9890 * Add support for STORE in ENGINE.
9894 * Add the STORE type. The intention is to provide a common interface
9895 to certificate and key stores, be they simple file-based stores, or
9896 HSM-type store, or LDAP stores, or...
9897 NOTE: The code is currently UNTESTED and isn't really used anywhere.
9901 * Add a generic structure called OPENSSL_ITEM. This can be used to
9902 pass a list of arguments to any function as well as provide a way
9903 for a function to pass data back to the caller.
9907 * Add the functions BUF_strndup() and BUF_memdup(). BUF_strndup()
9908 works like BUF_strdup() but can be used to duplicate a portion of
9909 a string. The copy gets NUL-terminated. BUF_memdup() duplicates
9914 * Add the function sk_find_ex() which works like sk_find(), but will
9915 return an index to an element even if an exact match couldn't be
9916 found. The index is guaranteed to point at the element where the
9917 searched-for key would be inserted to preserve sorting order.
9921 * Add the function OBJ_bsearch_ex() which works like OBJ_bsearch() but
9922 takes an extra flags argument for optional functionality. Currently,
9923 the following flags are defined:
9925 OBJ_BSEARCH_VALUE_ON_NOMATCH
9926 This one gets OBJ_bsearch_ex() to return a pointer to the first
9927 element where the comparing function returns a negative or zero
9930 OBJ_BSEARCH_FIRST_VALUE_ON_MATCH
9931 This one gets OBJ_bsearch_ex() to return a pointer to the first
9932 element where the comparing function returns zero. This is useful
9933 if there are more than one element where the comparing function
9938 * Make it possible to create self-signed certificates with 'openssl ca'
9939 in such a way that the self-signed certificate becomes part of the
9940 CA database and uses the same mechanisms for serial number generation
9941 as all other certificate signing. The new flag '-selfsign' enables
9942 this functionality. Adapt CA.sh and CA.pl.in.
9946 * Add functionality to check the public key of a certificate request
9947 against a given private. This is useful to check that a certificate
9948 request can be signed by that key (self-signing).
9952 * Make it possible to have multiple active certificates with the same
9953 subject in the CA index file. This is done only if the keyword
9954 'unique_subject' is set to 'no' in the main CA section (default
9955 if 'CA_default') of the configuration file. The value is saved
9956 with the database itself in a separate index attribute file,
9957 named like the index file with '.attr' appended to the name.
9961 * Generate multi-valued AVAs using '+' notation in config files for
9966 * Support for nameConstraints certificate extension.
9970 * Support for policyConstraints certificate extension.
9974 * Support for policyMappings certificate extension.
9978 * Make sure the default DSA_METHOD implementation only uses its
9979 dsa_mod_exp() and/or bn_mod_exp() handlers if they are non-NULL,
9980 and change its own handlers to be NULL so as to remove unnecessary
9981 indirection. This lets alternative implementations fallback to the
9982 default implementation more easily.
9986 * Support for directoryName in GeneralName related extensions
9991 * Make it possible to link applications using Makefile.shared.
9992 Make that possible even when linking against static libraries!
9996 * Support for single pass processing for S/MIME signing. This now
9997 means that S/MIME signing can be done from a pipe, in addition
9998 cleartext signing (multipart/signed type) is effectively streaming
9999 and the signed data does not need to be all held in memory.
10001 This is done with a new flag PKCS7_STREAM. When this flag is set
10002 PKCS7_sign() only initializes the PKCS7 structure and the actual signing
10003 is done after the data is output (and digests calculated) in
10004 SMIME_write_PKCS7().
10008 * Add full support for -rpath/-R, both in shared libraries and
10009 applications, at least on the platforms where it's known how
10014 * In crypto/ec/ec_mult.c, implement fast point multiplication with
10015 precomputation, based on wNAF splitting: EC_GROUP_precompute_mult()
10016 will now compute a table of multiples of the generator that
10017 makes subsequent invocations of EC_POINTs_mul() or EC_POINT_mul()
10018 faster (notably in the case of a single point multiplication,
10019 scalar * generator).
10021 *Nils Larsch, Bodo Moeller*
10023 * IPv6 support for certificate extensions. The various extensions
10024 which use the IP:a.b.c.d can now take IPv6 addresses using the
10025 formats of RFC1884 2.2 . IPv6 addresses are now also displayed
10030 * Added an ENGINE that implements RSA by performing private key
10031 exponentiations with the GMP library. The conversions to and from
10032 GMP's mpz_t format aren't optimised nor are any montgomery forms
10033 cached, and on x86 it appears OpenSSL's own performance has caught up.
10034 However there are likely to be other architectures where GMP could
10035 provide a boost. This ENGINE is not built in by default, but it can be
10036 specified at Configure time and should be accompanied by the necessary
10037 linker additions, eg;
10038 ./config -DOPENSSL_USE_GMP -lgmp
10042 * "openssl engine" will not display ENGINE/DSO load failure errors when
10043 testing availability of engines with "-t" - the old behaviour is
10044 produced by increasing the feature's verbosity with "-tt".
10048 * ECDSA routines: under certain error conditions uninitialized BN objects
10049 could be freed. Solution: make sure initialization is performed early
10050 enough. (Reported and fix supplied by Nils Larsch <nla@trustcenter.de>
10055 * Key-generation can now be implemented in RSA_METHOD, DSA_METHOD
10056 and DH_METHOD (eg. by ENGINE implementations) to override the normal
10057 software implementations. For DSA and DH, parameter generation can
10058 also be overridden by providing the appropriate method callbacks.
10062 * Change the "progress" mechanism used in key-generation and
10063 primality testing to functions that take a new BN_GENCB pointer in
10064 place of callback/argument pairs. The new API functions have `_ex`
10065 postfixes and the older functions are reimplemented as wrappers for
10066 the new ones. The OPENSSL_NO_DEPRECATED symbol can be used to hide
10067 declarations of the old functions to help (graceful) attempts to
10068 migrate to the new functions. Also, the new key-generation API
10069 functions operate on a caller-supplied key-structure and return
10070 success/failure rather than returning a key or NULL - this is to
10071 help make "keygen" another member function of RSA_METHOD etc.
10073 Example for using the new callback interface:
10075 int (*my_callback)(int a, int b, BN_GENCB *cb) = ...;
10076 void *my_arg = ...;
10079 BN_GENCB_set(&my_cb, my_callback, my_arg);
10081 return BN_is_prime_ex(some_bignum, BN_prime_checks, NULL, &cb);
10082 /* For the meaning of a, b in calls to my_callback(), see the
10083 * documentation of the function that calls the callback.
10084 * cb will point to my_cb; my_arg can be retrieved as cb->arg.
10085 * my_callback should return 1 if it wants BN_is_prime_ex()
10086 * to continue, or 0 to stop.
10091 * Change the ZLIB compression method to be stateful, and make it
10092 available to TLS with the number defined in
10093 draft-ietf-tls-compression-04.txt.
10097 * Add the ASN.1 structures and functions for CertificatePair, which
10098 is defined as follows (according to X.509_4thEditionDraftV6.pdf):
10100 CertificatePair ::= SEQUENCE {
10101 forward [0] Certificate OPTIONAL,
10102 reverse [1] Certificate OPTIONAL,
10103 -- at least one of the pair shall be present -- }
10105 Also implement the PEM functions to read and write certificate
10106 pairs, and defined the PEM tag as "CERTIFICATE PAIR".
10108 This needed to be defined, mostly for the sake of the LDAP
10109 attribute crossCertificatePair, but may prove useful elsewhere as
10114 * Make it possible to inhibit symlinking of shared libraries in
10115 Makefile.shared, for Cygwin's sake.
10119 * Extend the BIGNUM API by creating a function
10120 void BN_set_negative(BIGNUM *a, int neg);
10121 and a macro that behave like
10122 int BN_is_negative(const BIGNUM *a);
10124 to avoid the need to access 'a->neg' directly in applications.
10128 * Implement fast modular reduction for pseudo-Mersenne primes
10129 used in NIST curves (crypto/bn/bn_nist.c, crypto/ec/ecp_nist.c).
10130 EC_GROUP_new_curve_GFp() will now automatically use this
10133 *Nils Larsch <nla@trustcenter.de>*
10135 * Add new lock type (CRYPTO_LOCK_BN).
10139 * Change the ENGINE framework to automatically load engines
10140 dynamically from specific directories unless they could be
10141 found to already be built in or loaded. Move all the
10142 current engines except for the cryptodev one to a new
10143 directory engines/.
10144 The engines in engines/ are built as shared libraries if
10145 the "shared" options was given to ./Configure or ./config.
10146 Otherwise, they are inserted in libcrypto.a.
10147 /usr/local/ssl/engines is the default directory for dynamic
10148 engines, but that can be overridden at configure time through
10149 the usual use of --prefix and/or --openssldir, and at run
10150 time with the environment variable OPENSSL_ENGINES.
10152 *Geoff Thorpe and Richard Levitte*
10154 * Add Makefile.shared, a helper makefile to build shared
10155 libraries. Adapt Makefile.org.
10159 * Add version info to Win32 DLLs.
10161 *Peter 'Luna' Runestig" <peter@runestig.com>*
10163 * Add new 'medium level' PKCS#12 API. Certificates and keys
10164 can be added using this API to created arbitrary PKCS#12
10165 files while avoiding the low level API.
10167 New options to PKCS12_create(), key or cert can be NULL and
10168 will then be omitted from the output file. The encryption
10169 algorithm NIDs can be set to -1 for no encryption, the mac
10170 iteration count can be set to 0 to omit the mac.
10172 Enhance pkcs12 utility by making the -nokeys and -nocerts
10173 options work when creating a PKCS#12 file. New option -nomac
10174 to omit the mac, NONE can be set for an encryption algorithm.
10175 New code is modified to use the enhanced PKCS12_create()
10176 instead of the low level API.
10180 * Extend ASN1 encoder to support indefinite length constructed
10181 encoding. This can output sequences tags and octet strings in
10182 this form. Modify pk7_asn1.c to support indefinite length
10183 encoding. This is experimental and needs additional code to
10184 be useful, such as an ASN1 bio and some enhanced streaming
10187 Extend template encode functionality so that tagging is passed
10188 down to the template encoder.
10192 * Let 'openssl req' fail if an argument to '-newkey' is not
10193 recognized instead of using RSA as a default.
10197 * Add support for ECC-based ciphersuites from draft-ietf-tls-ecc-01.txt.
10198 As these are not official, they are not included in "ALL";
10199 the "ECCdraft" ciphersuite group alias can be used to select them.
10201 *Vipul Gupta and Sumit Gupta (Sun Microsystems Laboratories)*
10203 * Add ECDH engine support.
10205 *Nils Gura and Douglas Stebila (Sun Microsystems Laboratories)*
10207 * Add ECDH in new directory crypto/ecdh/.
10209 *Douglas Stebila (Sun Microsystems Laboratories)*
10211 * Let BN_rand_range() abort with an error after 100 iterations
10212 without success (which indicates a broken PRNG).
10216 * Change BN_mod_sqrt() so that it verifies that the input value
10217 is really the square of the return value. (Previously,
10218 BN_mod_sqrt would show GIGO behaviour.)
10222 * Add named elliptic curves over binary fields from X9.62, SECG,
10223 and WAP/WTLS; add OIDs that were still missing.
10225 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10227 * Extend the EC library for elliptic curves over binary fields
10228 (new files ec2_smpl.c, ec2_smpt.c, ec2_mult.c in crypto/ec/).
10231 EC_GF2m_simple_method
10235 EC_GROUP_new_curve_GF2m
10236 EC_GROUP_set_curve_GF2m
10237 EC_GROUP_get_curve_GF2m
10238 EC_POINT_set_affine_coordinates_GF2m
10239 EC_POINT_get_affine_coordinates_GF2m
10240 EC_POINT_set_compressed_coordinates_GF2m
10242 Point compression for binary fields is disabled by default for
10243 patent reasons (compile with OPENSSL_EC_BIN_PT_COMP defined to
10246 As binary polynomials are represented as BIGNUMs, various members
10247 of the EC_GROUP and EC_POINT data structures can be shared
10248 between the implementations for prime fields and binary fields;
10249 the above `..._GF2m functions` (except for EX_GROUP_new_curve_GF2m)
10250 are essentially identical to their `..._GFp` counterparts.
10251 (For simplicity, the `..._GFp` prefix has been dropped from
10252 various internal method names.)
10254 An internal 'field_div' method (similar to 'field_mul' and
10255 'field_sqr') has been added; this is used only for binary fields.
10257 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10259 * Optionally dispatch EC_POINT_mul(), EC_POINT_precompute_mult()
10260 through methods ('mul', 'precompute_mult').
10262 The generic implementations (now internally called 'ec_wNAF_mul'
10263 and 'ec_wNAF_precomputed_mult') remain the default if these
10264 methods are undefined.
10266 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10268 * New function EC_GROUP_get_degree, which is defined through
10269 EC_METHOD. For curves over prime fields, this returns the bit
10270 length of the modulus.
10272 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10274 * New functions EC_GROUP_dup, EC_POINT_dup.
10275 (These simply call ..._new and ..._copy).
10277 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10279 * Add binary polynomial arithmetic software in crypto/bn/bn_gf2m.c.
10280 Polynomials are represented as BIGNUMs (where the sign bit is not
10281 used) in the following functions [macros]:
10284 BN_GF2m_sub [= BN_GF2m_add]
10285 BN_GF2m_mod [wrapper for BN_GF2m_mod_arr]
10286 BN_GF2m_mod_mul [wrapper for BN_GF2m_mod_mul_arr]
10287 BN_GF2m_mod_sqr [wrapper for BN_GF2m_mod_sqr_arr]
10289 BN_GF2m_mod_exp [wrapper for BN_GF2m_mod_exp_arr]
10290 BN_GF2m_mod_sqrt [wrapper for BN_GF2m_mod_sqrt_arr]
10291 BN_GF2m_mod_solve_quad [wrapper for BN_GF2m_mod_solve_quad_arr]
10292 BN_GF2m_cmp [= BN_ucmp]
10294 (Note that only the 'mod' functions are actually for fields GF(2^m).
10295 BN_GF2m_add() is misnomer, but this is for the sake of consistency.)
10297 For some functions, an the irreducible polynomial defining a
10298 field can be given as an 'unsigned int[]' with strictly
10299 decreasing elements giving the indices of those bits that are set;
10300 i.e., p[] represents the polynomial
10301 f(t) = t^p[0] + t^p[1] + ... + t^p[k]
10303 p[0] > p[1] > ... > p[k] = 0.
10304 This applies to the following functions:
10307 BN_GF2m_mod_mul_arr
10308 BN_GF2m_mod_sqr_arr
10309 BN_GF2m_mod_inv_arr [wrapper for BN_GF2m_mod_inv]
10310 BN_GF2m_mod_div_arr [wrapper for BN_GF2m_mod_div]
10311 BN_GF2m_mod_exp_arr
10312 BN_GF2m_mod_sqrt_arr
10313 BN_GF2m_mod_solve_quad_arr
10317 Conversion can be performed by the following functions:
10322 bntest.c has additional tests for binary polynomial arithmetic.
10324 Two implementations for BN_GF2m_mod_div() are available.
10325 The default algorithm simply uses BN_GF2m_mod_inv() and
10326 BN_GF2m_mod_mul(). The alternative algorithm is compiled in only
10327 if OPENSSL_SUN_GF2M_DIV is defined (patent pending; read the
10328 copyright notice in crypto/bn/bn_gf2m.c before enabling it).
10330 *Sheueling Chang Shantz and Douglas Stebila (Sun Microsystems Laboratories)*
10332 * Add new error code 'ERR_R_DISABLED' that can be used when some
10333 functionality is disabled at compile-time.
10335 *Douglas Stebila <douglas.stebila@sun.com>*
10337 * Change default behaviour of 'openssl asn1parse' so that more
10338 information is visible when viewing, e.g., a certificate:
10340 Modify asn1_parse2 (crypto/asn1/asn1_par.c) so that in non-'dump'
10341 mode the content of non-printable OCTET STRINGs is output in a
10342 style similar to INTEGERs, but with '[HEX DUMP]' prepended to
10343 avoid the appearance of a printable string.
10345 *Nils Larsch <nla@trustcenter.de>*
10347 * Add 'asn1_flag' and 'asn1_form' member to EC_GROUP with access
10349 EC_GROUP_set_asn1_flag()
10350 EC_GROUP_get_asn1_flag()
10351 EC_GROUP_set_point_conversion_form()
10352 EC_GROUP_get_point_conversion_form()
10353 These control ASN1 encoding details:
10354 - Curves (i.e., groups) are encoded explicitly unless asn1_flag
10355 has been set to OPENSSL_EC_NAMED_CURVE.
10356 - Points are encoded in uncompressed form by default; options for
10357 asn1_for are as for point2oct, namely
10358 POINT_CONVERSION_COMPRESSED
10359 POINT_CONVERSION_UNCOMPRESSED
10360 POINT_CONVERSION_HYBRID
10362 Also add 'seed' and 'seed_len' members to EC_GROUP with access
10364 EC_GROUP_set_seed()
10365 EC_GROUP_get0_seed()
10366 EC_GROUP_get_seed_len()
10367 This is used only for ASN1 purposes (so far).
10369 *Nils Larsch <nla@trustcenter.de>*
10371 * Add 'field_type' member to EC_METHOD, which holds the NID
10372 of the appropriate field type OID. The new function
10373 EC_METHOD_get_field_type() returns this value.
10375 *Nils Larsch <nla@trustcenter.de>*
10378 EC_POINT_point2bn()
10379 EC_POINT_bn2point()
10380 EC_POINT_point2hex()
10381 EC_POINT_hex2point()
10382 providing useful interfaces to EC_POINT_point2oct() and
10383 EC_POINT_oct2point().
10385 *Nils Larsch <nla@trustcenter.de>*
10387 * Change internals of the EC library so that the functions
10388 EC_GROUP_set_generator()
10389 EC_GROUP_get_generator()
10390 EC_GROUP_get_order()
10391 EC_GROUP_get_cofactor()
10392 are implemented directly in crypto/ec/ec_lib.c and not dispatched
10393 to methods, which would lead to unnecessary code duplication when
10394 adding different types of curves.
10396 *Nils Larsch <nla@trustcenter.de> with input by Bodo Moeller*
10398 * Implement compute_wNAF (crypto/ec/ec_mult.c) without BIGNUM
10399 arithmetic, and such that modified wNAFs are generated
10400 (which avoid length expansion in many cases).
10404 * Add a function EC_GROUP_check_discriminant() (defined via
10405 EC_METHOD) that verifies that the curve discriminant is non-zero.
10407 Add a function EC_GROUP_check() that makes some sanity tests
10408 on a EC_GROUP, its generator and order. This includes
10409 EC_GROUP_check_discriminant().
10411 *Nils Larsch <nla@trustcenter.de>*
10413 * Add ECDSA in new directory crypto/ecdsa/.
10415 Add applications 'openssl ecparam' and 'openssl ecdsa'
10416 (these are based on 'openssl dsaparam' and 'openssl dsa').
10418 ECDSA support is also included in various other files across the
10419 library. Most notably,
10420 - 'openssl req' now has a '-newkey ecdsa:file' option;
10421 - EVP_PKCS82PKEY (crypto/evp/evp_pkey.c) now can handle ECDSA;
10422 - X509_PUBKEY_get (crypto/asn1/x_pubkey.c) and
10423 d2i_PublicKey (crypto/asn1/d2i_pu.c) have been modified to make
10424 them suitable for ECDSA where domain parameters must be
10425 extracted before the specific public key;
10426 - ECDSA engine support has been added.
10428 *Nils Larsch <nla@trustcenter.de>*
10430 * Include some named elliptic curves, and add OIDs from X9.62,
10431 SECG, and WAP/WTLS. Each curve can be obtained from the new
10433 EC_GROUP_new_by_curve_name(),
10434 and the list of available named curves can be obtained with
10435 EC_get_builtin_curves().
10436 Also add a 'curve_name' member to EC_GROUP objects, which can be
10438 EC_GROUP_set_curve_name()
10439 EC_GROUP_get_curve_name()
10441 *Nils Larsch <larsch@trustcenter.de, Bodo Moeller*
10443 * Remove a few calls to bn_wexpand() in BN_sqr() (the one in there
10444 was actually never needed) and in BN_mul(). The removal in BN_mul()
10445 required a small change in bn_mul_part_recursive() and the addition
10446 of the functions bn_cmp_part_words(), bn_sub_part_words() and
10447 bn_add_part_words(), which do the same thing as bn_cmp_words(),
10448 bn_sub_words() and bn_add_words() except they take arrays with
10453 ### Changes between 0.9.7l and 0.9.7m [23 Feb 2007]
10455 * Cleanse PEM buffers before freeing them since they may contain
10458 *Benjamin Bennett <ben@psc.edu>*
10460 * Include "!eNULL" in SSL_DEFAULT_CIPHER_LIST to make sure that
10461 a ciphersuite string such as "DEFAULT:RSA" cannot enable
10462 authentication-only ciphersuites.
10466 * Since AES128 and AES256 share a single mask bit in the logic of
10467 ssl/ssl_ciph.c, the code for masking out disabled ciphers needs a
10468 kludge to work properly if AES128 is available and AES256 isn't.
10472 * Expand security boundary to match 1.1.1 module.
10476 * Remove redundant features: hash file source, editing of test vectors
10477 modify fipsld to use external fips_premain.c signature.
10481 * New perl script mkfipsscr.pl to create shell scripts or batch files to
10482 run algorithm test programs.
10486 * Make algorithm test programs more tolerant of whitespace.
10490 * Have SSL/TLS server implementation tolerate "mismatched" record
10491 protocol version while receiving ClientHello even if the
10492 ClientHello is fragmented. (The server can't insist on the
10493 particular protocol version it has chosen before the ServerHello
10494 message has informed the client about his choice.)
10498 * Load error codes if they are not already present instead of using a
10499 static variable. This allows them to be cleanly unloaded and reloaded.
10503 ### Changes between 0.9.7k and 0.9.7l [28 Sep 2006]
10505 * Introduce limits to prevent malicious keys being able to
10506 cause a denial of service. [CVE-2006-2940][]
10508 *Steve Henson, Bodo Moeller*
10510 * Fix ASN.1 parsing of certain invalid structures that can result
10511 in a denial of service. [CVE-2006-2937][] [Steve Henson]
10513 * Fix buffer overflow in SSL_get_shared_ciphers() function.
10514 [CVE-2006-3738][] [Tavis Ormandy and Will Drewry, Google Security Team]
10516 * Fix SSL client code which could crash if connecting to a
10517 malicious SSLv2 server. [CVE-2006-4343][]
10519 *Tavis Ormandy and Will Drewry, Google Security Team*
10521 * Change ciphersuite string processing so that an explicit
10522 ciphersuite selects this one ciphersuite (so that "AES256-SHA"
10523 will no longer include "AES128-SHA"), and any other similar
10524 ciphersuite (same bitmap) from *other* protocol versions (so that
10525 "RC4-MD5" will still include both the SSL 2.0 ciphersuite and the
10526 SSL 3.0/TLS 1.0 ciphersuite). This is a backport combining
10527 changes from 0.9.8b and 0.9.8d.
10531 ### Changes between 0.9.7j and 0.9.7k [05 Sep 2006]
10533 * Avoid PKCS #1 v1.5 signature attack discovered by Daniel Bleichenbacher
10534 [CVE-2006-4339][] [Ben Laurie and Google Security Team]
10536 * Change the Unix randomness entropy gathering to use poll() when
10537 possible instead of select(), since the latter has some
10538 undesirable limitations.
10540 *Darryl Miles via Richard Levitte and Bodo Moeller*
10542 * Disable rogue ciphersuites:
10544 - SSLv2 0x08 0x00 0x80 ("RC4-64-MD5")
10545 - SSLv3/TLSv1 0x00 0x61 ("EXP1024-RC2-CBC-MD5")
10546 - SSLv3/TLSv1 0x00 0x60 ("EXP1024-RC4-MD5")
10548 The latter two were purportedly from
10549 draft-ietf-tls-56-bit-ciphersuites-0[01].txt, but do not really
10552 Also deactivate the remaining ciphersuites from
10553 draft-ietf-tls-56-bit-ciphersuites-01.txt. These are just as
10554 unofficial, and the ID has long expired.
10558 * Fix RSA blinding Heisenbug (problems sometimes occurred on
10559 dual-core machines) and other potential thread-safety issues.
10563 ### Changes between 0.9.7i and 0.9.7j [04 May 2006]
10565 * Adapt fipsld and the build system to link against the validated FIPS
10566 module in FIPS mode.
10570 * Fixes for VC++ 2005 build under Windows.
10574 * Add new Windows build target VC-32-GMAKE for VC++. This uses GNU make
10575 from a Windows bash shell such as MSYS. It is autodetected from the
10576 "config" script when run from a VC++ environment. Modify standard VC++
10577 build to use fipscanister.o from the GNU make build.
10581 ### Changes between 0.9.7h and 0.9.7i [14 Oct 2005]
10583 * Wrapped the definition of EVP_MAX_MD_SIZE in a #ifdef OPENSSL_FIPS.
10584 The value now differs depending on if you build for FIPS or not.
10585 BEWARE! A program linked with a shared FIPSed libcrypto can't be
10586 safely run with a non-FIPSed libcrypto, as it may crash because of
10587 the difference induced by this change.
10591 ### Changes between 0.9.7g and 0.9.7h [11 Oct 2005]
10593 * Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING
10594 (part of SSL_OP_ALL). This option used to disable the
10595 countermeasure against man-in-the-middle protocol-version
10596 rollback in the SSL 2.0 server implementation, which is a bad
10597 idea. [CVE-2005-2969][]
10599 *Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center
10600 for Information Security, National Institute of Advanced Industrial
10601 Science and Technology [AIST, Japan)]*
10603 * Minimal support for X9.31 signatures and PSS padding modes. This is
10604 mainly for FIPS compliance and not fully integrated at this stage.
10608 * For DSA signing, unless DSA_FLAG_NO_EXP_CONSTTIME is set, perform
10609 the exponentiation using a fixed-length exponent. (Otherwise,
10610 the information leaked through timing could expose the secret key
10611 after many signatures; cf. Bleichenbacher's attack on DSA with
10616 * Make a new fixed-window mod_exp implementation the default for
10617 RSA, DSA, and DH private-key operations so that the sequence of
10618 squares and multiplies and the memory access pattern are
10619 independent of the particular secret key. This will mitigate
10620 cache-timing and potential related attacks.
10622 BN_mod_exp_mont_consttime() is the new exponentiation implementation,
10623 and this is automatically used by BN_mod_exp_mont() if the new flag
10624 BN_FLG_EXP_CONSTTIME is set for the exponent. RSA, DSA, and DH
10625 will use this BN flag for private exponents unless the flag
10626 RSA_FLAG_NO_EXP_CONSTTIME, DSA_FLAG_NO_EXP_CONSTTIME, or
10627 DH_FLAG_NO_EXP_CONSTTIME, respectively, is set.
10629 *Matthew D Wood (Intel Corp), with some changes by Bodo Moeller*
10631 * Change the client implementation for SSLv23_method() and
10632 SSLv23_client_method() so that is uses the SSL 3.0/TLS 1.0
10633 Client Hello message format if the SSL_OP_NO_SSLv2 option is set.
10634 (Previously, the SSL 2.0 backwards compatible Client Hello
10635 message format would be used even with SSL_OP_NO_SSLv2.)
10639 * Add support for smime-type MIME parameter in S/MIME messages which some
10644 * New function BN_MONT_CTX_set_locked() to set montgomery parameters in
10645 a threadsafe manner. Modify rsa code to use new function and add calls
10646 to dsa and dh code (which had race conditions before).
10650 * Include the fixed error library code in the C error file definitions
10651 instead of fixing them up at runtime. This keeps the error code
10652 structures constant.
10656 ### Changes between 0.9.7f and 0.9.7g [11 Apr 2005]
10658 [NB: OpenSSL 0.9.7h and later 0.9.7 patch levels were released after
10661 * Fixes for newer kerberos headers. NB: the casts are needed because
10662 the 'length' field is signed on one version and unsigned on another
10663 with no (?) obvious way to tell the difference, without these VC++
10664 complains. Also the "definition" of FAR (blank) is no longer included
10665 nor is the error ENOMEM. KRB5_PRIVATE has to be set to 1 to pick up
10666 some needed definitions.
10670 * Undo Cygwin change.
10674 * Added support for proxy certificates according to RFC 3820.
10675 Because they may be a security thread to unaware applications,
10676 they must be explicitly allowed in run-time. See
10677 docs/HOWTO/proxy_certificates.txt for further information.
10681 ### Changes between 0.9.7e and 0.9.7f [22 Mar 2005]
10683 * Use (SSL_RANDOM_VALUE - 4) bytes of pseudo random data when generating
10684 server and client random values. Previously
10685 (SSL_RANDOM_VALUE - sizeof(time_t)) would be used which would result in
10686 less random data when sizeof(time_t) > 4 (some 64 bit platforms).
10688 This change has negligible security impact because:
10690 1. Server and client random values still have 24 bytes of pseudo random
10693 2. Server and client random values are sent in the clear in the initial
10696 3. The master secret is derived using the premaster secret (48 bytes in
10697 size for static RSA ciphersuites) as well as client server and random
10700 The OpenSSL team would like to thank the UK NISCC for bringing this issue
10703 *Stephen Henson, reported by UK NISCC*
10705 * Use Windows randomness collection on Cygwin.
10709 * Fix hang in EGD/PRNGD query when communication socket is closed
10710 prematurely by EGD/PRNGD.
10712 *Darren Tucker <dtucker@zip.com.au> via Lutz Jänicke, resolves #1014*
10714 * Prompt for pass phrases when appropriate for PKCS12 input format.
10718 * Back-port of selected performance improvements from development
10719 branch, as well as improved support for PowerPC platforms.
10723 * Add lots of checks for memory allocation failure, error codes to indicate
10724 failure and freeing up memory if a failure occurs.
10726 *Nauticus Networks SSL Team <openssl@nauticusnet.com>, Steve Henson*
10728 * Add new -passin argument to dgst.
10732 * Perform some character comparisons of different types in X509_NAME_cmp:
10733 this is needed for some certificates that re-encode DNs into UTF8Strings
10734 (in violation of RFC3280) and can't or won't issue name rollover
10739 * Make an explicit check during certificate validation to see that
10740 the CA setting in each certificate on the chain is correct. As a
10741 side effect always do the following basic checks on extensions,
10742 not just when there's an associated purpose to the check:
10744 - if there is an unhandled critical extension (unless the user
10745 has chosen to ignore this fault)
10746 - if the path length has been exceeded (if one is set at all)
10747 - that certain extensions fit the associated purpose (if one has
10752 ### Changes between 0.9.7d and 0.9.7e [25 Oct 2004]
10754 * Avoid a race condition when CRLs are checked in a multi threaded
10755 environment. This would happen due to the reordering of the revoked
10756 entries during signature checking and serial number lookup. Now the
10757 encoding is cached and the serial number sort performed under a lock.
10758 Add new STACK function sk_is_sorted().
10762 * Add Delta CRL to the extension code.
10766 * Various fixes to s3_pkt.c so alerts are sent properly.
10768 *David Holmes <d.holmes@f5.com>*
10770 * Reduce the chances of duplicate issuer name and serial numbers (in
10771 violation of RFC3280) using the OpenSSL certificate creation utilities.
10772 This is done by creating a random 64 bit value for the initial serial
10773 number when a serial number file is created or when a self signed
10774 certificate is created using 'openssl req -x509'. The initial serial
10775 number file is created using 'openssl x509 -next_serial' in CA.pl
10776 rather than being initialized to 1.
10780 ### Changes between 0.9.7c and 0.9.7d [17 Mar 2004]
10782 * Fix null-pointer assignment in do_change_cipher_spec() revealed
10783 by using the Codenomicon TLS Test Tool [CVE-2004-0079][]
10785 *Joe Orton, Steve Henson*
10787 * Fix flaw in SSL/TLS handshaking when using Kerberos ciphersuites
10790 *Joe Orton, Steve Henson*
10792 * Make it possible to have multiple active certificates with the same
10793 subject in the CA index file. This is done only if the keyword
10794 'unique_subject' is set to 'no' in the main CA section (default
10795 if 'CA_default') of the configuration file. The value is saved
10796 with the database itself in a separate index attribute file,
10797 named like the index file with '.attr' appended to the name.
10801 * X509 verify fixes. Disable broken certificate workarounds when
10802 X509_V_FLAGS_X509_STRICT is set. Check CRL issuer has cRLSign set if
10803 keyUsage extension present. Don't accept CRLs with unhandled critical
10804 extensions: since verify currently doesn't process CRL extensions this
10805 rejects a CRL with *any* critical extensions. Add new verify error codes
10810 * When creating an OCSP nonce use an OCTET STRING inside the extnValue.
10811 A clarification of RFC2560 will require the use of OCTET STRINGs and
10812 some implementations cannot handle the current raw format. Since OpenSSL
10813 copies and compares OCSP nonces as opaque blobs without any attempt at
10814 parsing them this should not create any compatibility issues.
10818 * New md flag EVP_MD_CTX_FLAG_REUSE this allows md_data to be reused when
10819 calling EVP_MD_CTX_copy_ex() to avoid calling OPENSSL_malloc(). Without
10820 this HMAC (and other) operations are several times slower than OpenSSL
10825 * Print out GeneralizedTime and UTCTime in ASN1_STRING_print_ex().
10827 *Peter Sylvester <Peter.Sylvester@EdelWeb.fr>*
10829 * Use the correct content when signing type "other".
10833 ### Changes between 0.9.7b and 0.9.7c [30 Sep 2003]
10835 * Fix various bugs revealed by running the NISCC test suite:
10837 Stop out of bounds reads in the ASN1 code when presented with
10838 invalid tags (CVE-2003-0543 and CVE-2003-0544).
10840 Free up ASN1_TYPE correctly if ANY type is invalid [CVE-2003-0545][].
10842 If verify callback ignores invalid public key errors don't try to check
10843 certificate signature with the NULL public key.
10847 * New -ignore_err option in ocsp application to stop the server
10848 exiting on the first error in a request.
10852 * In ssl3_accept() (ssl/s3_srvr.c) only accept a client certificate
10853 if the server requested one: as stated in TLS 1.0 and SSL 3.0
10858 * In ssl3_get_client_hello() (ssl/s3_srvr.c), tolerate additional
10859 extra data after the compression methods not only for TLS 1.0
10860 but also for SSL 3.0 (as required by the specification).
10862 *Bodo Moeller; problem pointed out by Matthias Loepfe*
10864 * Change X509_certificate_type() to mark the key as exported/exportable
10865 when it's 512 *bits* long, not 512 bytes.
10869 * Change AES_cbc_encrypt() so it outputs exact multiple of
10870 blocks during encryption.
10874 * Various fixes to base64 BIO and non blocking I/O. On write
10875 flushes were not handled properly if the BIO retried. On read
10876 data was not being buffered properly and had various logic bugs.
10877 This also affects blocking I/O when the data being decoded is a
10882 * Various S/MIME bugfixes and compatibility changes:
10883 output correct application/pkcs7 MIME type if
10884 PKCS7_NOOLDMIMETYPE is set. Tolerate some broken signatures.
10885 Output CR+LF for EOL if PKCS7_CRLFEOL is set (this makes opening
10886 of files as .eml work). Correctly handle very long lines in MIME
10891 ### Changes between 0.9.7a and 0.9.7b [10 Apr 2003]
10893 * Countermeasure against the Klima-Pokorny-Rosa extension of
10894 Bleichbacher's attack on PKCS #1 v1.5 padding: treat
10895 a protocol version number mismatch like a decryption error
10896 in ssl3_get_client_key_exchange (ssl/s3_srvr.c).
10900 * Turn on RSA blinding by default in the default implementation
10901 to avoid a timing attack. Applications that don't want it can call
10902 RSA_blinding_off() or use the new flag RSA_FLAG_NO_BLINDING.
10903 They would be ill-advised to do so in most cases.
10905 *Ben Laurie, Steve Henson, Geoff Thorpe, Bodo Moeller*
10907 * Change RSA blinding code so that it works when the PRNG is not
10908 seeded (in this case, the secret RSA exponent is abused as
10909 an unpredictable seed -- if it is not unpredictable, there
10910 is no point in blinding anyway). Make RSA blinding thread-safe
10911 by remembering the creator's thread ID in rsa->blinding and
10912 having all other threads use local one-time blinding factors
10913 (this requires more computation than sharing rsa->blinding, but
10914 avoids excessive locking; and if an RSA object is not shared
10915 between threads, blinding will still be very fast).
10919 * Fixed a typo bug that would cause ENGINE_set_default() to set an
10920 ENGINE as defaults for all supported algorithms irrespective of
10921 the 'flags' parameter. 'flags' is now honoured, so applications
10922 should make sure they are passing it correctly.
10926 * Target "mingw" now allows native Windows code to be generated in
10927 the Cygwin environment as well as with the MinGW compiler.
10931 ### Changes between 0.9.7 and 0.9.7a [19 Feb 2003]
10933 * In ssl3_get_record (ssl/s3_pkt.c), minimize information leaked
10934 via timing by performing a MAC computation even if incorrect
10935 block cipher padding has been found. This is a countermeasure
10936 against active attacks where the attacker has to distinguish
10937 between bad padding and a MAC verification error. [CVE-2003-0078][]
10939 *Bodo Moeller; problem pointed out by Brice Canvel (EPFL),
10940 Alain Hiltgen (UBS), Serge Vaudenay (EPFL), and
10941 Martin Vuagnoux (EPFL, Ilion)*
10943 * Make the no-err option work as intended. The intention with no-err
10944 is not to have the whole error stack handling routines removed from
10945 libcrypto, it's only intended to remove all the function name and
10946 reason texts, thereby removing some of the footprint that may not
10947 be interesting if those errors aren't displayed anyway.
10949 NOTE: it's still possible for any application or module to have its
10950 own set of error texts inserted. The routines are there, just not
10951 used by default when no-err is given.
10955 * Add support for FreeBSD on IA64.
10957 *dirk.meyer@dinoex.sub.org via Richard Levitte, resolves #454*
10959 * Adjust DES_cbc_cksum() so it returns the same value as the MIT
10960 Kerberos function mit_des_cbc_cksum(). Before this change,
10961 the value returned by DES_cbc_cksum() was like the one from
10962 mit_des_cbc_cksum(), except the bytes were swapped.
10964 *Kevin Greaney <Kevin.Greaney@hp.com> and Richard Levitte*
10966 * Allow an application to disable the automatic SSL chain building.
10967 Before this a rather primitive chain build was always performed in
10968 ssl3_output_cert_chain(): an application had no way to send the
10969 correct chain if the automatic operation produced an incorrect result.
10971 Now the chain builder is disabled if either:
10973 1. Extra certificates are added via SSL_CTX_add_extra_chain_cert().
10975 2. The mode flag SSL_MODE_NO_AUTO_CHAIN is set.
10977 The reasoning behind this is that an application would not want the
10978 auto chain building to take place if extra chain certificates are
10979 present and it might also want a means of sending no additional
10980 certificates (for example the chain has two certificates and the
10985 * Add the possibility to build without the ENGINE framework.
10987 *Steven Reddie <smr@essemer.com.au> via Richard Levitte*
10989 * Under Win32 gmtime() can return NULL: check return value in
10990 OPENSSL_gmtime(). Add error code for case where gmtime() fails.
10994 * DSA routines: under certain error conditions uninitialized BN objects
10995 could be freed. Solution: make sure initialization is performed early
10996 enough. (Reported and fix supplied by Ivan D Nestlerode <nestler@MIT.EDU>,
10997 Nils Larsch <nla@trustcenter.de> via PR#459)
11001 * Another fix for SSLv2 session ID handling: the session ID was incorrectly
11002 checked on reconnect on the client side, therefore session resumption
11003 could still fail with a "ssl session id is different" error. This
11004 behaviour is masked when SSL_OP_ALL is used due to
11005 SSL_OP_MICROSOFT_SESS_ID_BUG being set.
11006 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as
11007 followup to PR #377.
11011 * IA-32 assembler support enhancements: unified ELF targets, support
11012 for SCO/Caldera platforms, fix for Cygwin shared build.
11016 * Add support for FreeBSD on sparc64. As a consequence, support for
11017 FreeBSD on non-x86 processors is separate from x86 processors on
11018 the config script, much like the NetBSD support.
11020 *Richard Levitte & Kris Kennaway <kris@obsecurity.org>*
11022 ### Changes between 0.9.6h and 0.9.7 [31 Dec 2002]
11024 [NB: OpenSSL 0.9.6i and later 0.9.6 patch levels were released after
11027 * Fix session ID handling in SSLv2 client code: the SERVER FINISHED
11028 code (06) was taken as the first octet of the session ID and the last
11029 octet was ignored consequently. As a result SSLv2 client side session
11030 caching could not have worked due to the session ID mismatch between
11032 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as
11037 * Change the declaration of needed Kerberos libraries to use EX_LIBS
11038 instead of the special (and badly supported) LIBKRB5. LIBKRB5 is
11043 * The hw_ncipher.c engine requires dynamic locks. Unfortunately, it
11044 seems that in spite of existing for more than a year, many application
11045 author have done nothing to provide the necessary callbacks, which
11046 means that this particular engine will not work properly anywhere.
11047 This is a very unfortunate situation which forces us, in the name
11048 of usability, to give the hw_ncipher.c a static lock, which is part
11050 NOTE: This is for the 0.9.7 series ONLY. This hack will never
11051 appear in 0.9.8 or later. We EXPECT application authors to have
11052 dealt properly with this when 0.9.8 is released (unless we actually
11053 make such changes in the libcrypto locking code that changes will
11054 have to be made anyway).
11058 * In asn1_d2i_read_bio() repeatedly call BIO_read() until all content
11059 octets have been read, EOF or an error occurs. Without this change
11060 some truncated ASN1 structures will not produce an error.
11064 * Disable Heimdal support, since it hasn't been fully implemented.
11065 Still give the possibility to force the use of Heimdal, but with
11066 warnings and a request that patches get sent to openssl-dev.
11070 * Add the VC-CE target, introduce the WINCE sysname, and add
11071 INSTALL.WCE and appropriate conditionals to make it build.
11073 *Steven Reddie <smr@essemer.com.au> via Richard Levitte*
11075 * Change the DLL names for Cygwin to cygcrypto-x.y.z.dll and
11076 cygssl-x.y.z.dll, where x, y and z are the major, minor and
11077 edit numbers of the version.
11079 *Corinna Vinschen <vinschen@redhat.com> and Richard Levitte*
11081 * Introduce safe string copy and catenation functions
11082 (BUF_strlcpy() and BUF_strlcat()).
11084 *Ben Laurie (CHATS) and Richard Levitte*
11086 * Avoid using fixed-size buffers for one-line DNs.
11088 *Ben Laurie (CHATS)*
11090 * Add BUF_MEM_grow_clean() to avoid information leakage when
11091 resizing buffers containing secrets, and use where appropriate.
11093 *Ben Laurie (CHATS)*
11095 * Avoid using fixed size buffers for configuration file location.
11097 *Ben Laurie (CHATS)*
11099 * Avoid filename truncation for various CA files.
11101 *Ben Laurie (CHATS)*
11103 * Use sizeof in preference to magic numbers.
11105 *Ben Laurie (CHATS)*
11107 * Avoid filename truncation in cert requests.
11109 *Ben Laurie (CHATS)*
11111 * Add assertions to check for (supposedly impossible) buffer
11114 *Ben Laurie (CHATS)*
11116 * Don't cache truncated DNS entries in the local cache (this could
11117 potentially lead to a spoofing attack).
11119 *Ben Laurie (CHATS)*
11121 * Fix various buffers to be large enough for hex/decimal
11122 representations in a platform independent manner.
11124 *Ben Laurie (CHATS)*
11126 * Add CRYPTO_realloc_clean() to avoid information leakage when
11127 resizing buffers containing secrets, and use where appropriate.
11129 *Ben Laurie (CHATS)*
11131 * Add BIO_indent() to avoid much slightly worrying code to do
11134 *Ben Laurie (CHATS)*
11136 * Convert sprintf()/BIO_puts() to BIO_printf().
11138 *Ben Laurie (CHATS)*
11140 * buffer_gets() could terminate with the buffer only half
11143 *Ben Laurie (CHATS)*
11145 * Add assertions to prevent user-supplied crypto functions from
11146 overflowing internal buffers by having large block sizes, etc.
11148 *Ben Laurie (CHATS)*
11150 * New OPENSSL_assert() macro (similar to assert(), but enabled
11153 *Ben Laurie (CHATS)*
11155 * Eliminate unused copy of key in RC4.
11157 *Ben Laurie (CHATS)*
11159 * Eliminate unused and incorrectly sized buffers for IV in pem.h.
11161 *Ben Laurie (CHATS)*
11163 * Fix off-by-one error in EGD path.
11165 *Ben Laurie (CHATS)*
11167 * If RANDFILE path is too long, ignore instead of truncating.
11169 *Ben Laurie (CHATS)*
11171 * Eliminate unused and incorrectly sized X.509 structure
11174 *Ben Laurie (CHATS)*
11176 * Eliminate unused and dangerous function knumber().
11178 *Ben Laurie (CHATS)*
11180 * Eliminate unused and dangerous structure, KSSL_ERR.
11182 *Ben Laurie (CHATS)*
11184 * Protect against overlong session ID context length in an encoded
11185 session object. Since these are local, this does not appear to be
11188 *Ben Laurie (CHATS)*
11190 * Change from security patch (see 0.9.6e below) that did not affect
11191 the 0.9.6 release series:
11193 Remote buffer overflow in SSL3 protocol - an attacker could
11194 supply an oversized master key in Kerberos-enabled versions.
11197 *Ben Laurie (CHATS)*
11199 * Change the SSL kerb5 codes to match RFC 2712.
11203 * Make -nameopt work fully for req and add -reqopt switch.
11205 *Michael Bell <michael.bell@rz.hu-berlin.de>, Steve Henson*
11207 * The "block size" for block ciphers in CFB and OFB mode should be 1.
11209 *Steve Henson, reported by Yngve Nysaeter Pettersen <yngve@opera.com>*
11211 * Make sure tests can be performed even if the corresponding algorithms
11212 have been removed entirely. This was also the last step to make
11213 OpenSSL compilable with DJGPP under all reasonable conditions.
11215 *Richard Levitte, Doug Kaufman <dkaufman@rahul.net>*
11217 * Add cipher selection rules COMPLEMENTOFALL and COMPLEMENTOFDEFAULT
11218 to allow version independent disabling of normally unselected ciphers,
11219 which may be activated as a side-effect of selecting a single cipher.
11221 (E.g., cipher list string "RSA" enables ciphersuites that are left
11222 out of "ALL" because they do not provide symmetric encryption.
11223 "RSA:!COMPLEMEMENTOFALL" avoids these unsafe ciphersuites.)
11225 *Lutz Jaenicke, Bodo Moeller*
11227 * Add appropriate support for separate platform-dependent build
11228 directories. The recommended way to make a platform-dependent
11229 build directory is the following (tested on Linux), maybe with
11232 # Place yourself outside of the OpenSSL source tree. In
11233 # this example, the environment variable OPENSSL_SOURCE
11234 # is assumed to contain the absolute OpenSSL source directory.
11235 mkdir -p objtree/"`uname -s`-`uname -r`-`uname -m`"
11236 cd objtree/"`uname -s`-`uname -r`-`uname -m`"
11237 (cd $OPENSSL_SOURCE; find . -type f) | while read F; do
11238 mkdir -p `dirname $F`
11239 ln -s $OPENSSL_SOURCE/$F $F
11242 To be absolutely sure not to disturb the source tree, a "make clean"
11243 is a good thing. If it isn't successful, don't worry about it,
11244 it probably means the source directory is very clean.
11248 * Make sure any ENGINE control commands make local copies of string
11249 pointers passed to them whenever necessary. Otherwise it is possible
11250 the caller may have overwritten (or deallocated) the original string
11251 data when a later ENGINE operation tries to use the stored values.
11253 *Götz Babin-Ebell <babinebell@trustcenter.de>*
11255 * Improve diagnostics in file reading and command-line digests.
11257 *Ben Laurie aided and abetted by Solar Designer <solar@openwall.com>*
11259 * Add AES modes CFB and OFB to the object database. Correct an
11260 error in AES-CFB decryption.
11264 * Remove most calls to EVP_CIPHER_CTX_cleanup() in evp_enc.c, this
11265 allows existing EVP_CIPHER_CTX structures to be reused after
11266 calling `EVP_*Final()`. This behaviour is used by encryption
11267 BIOs and some applications. This has the side effect that
11268 applications must explicitly clean up cipher contexts with
11269 EVP_CIPHER_CTX_cleanup() or they will leak memory.
11273 * Check the values of dna and dnb in bn_mul_recursive before calling
11274 bn_mul_comba (a non zero value means the a or b arrays do not contain
11275 n2 elements) and fallback to bn_mul_normal if either is not zero.
11279 * Fix escaping of non-ASCII characters when using the -subj option
11280 of the "openssl req" command line tool. (Robert Joop <joop@fokus.gmd.de>)
11284 * Make object definitions compliant to LDAP (RFC2256): SN is the short
11285 form for "surname", serialNumber has no short form.
11286 Use "mail" as the short name for "rfc822Mailbox" according to RFC2798;
11287 therefore remove "mail" short name for "internet 7".
11288 The OID for unique identifiers in X509 certificates is
11289 x500UniqueIdentifier, not uniqueIdentifier.
11290 Some more OID additions. (Michael Bell <michael.bell@rz.hu-berlin.de>)
11294 * Add an "init" command to the ENGINE config module and auto initialize
11295 ENGINEs. Without any "init" command the ENGINE will be initialized
11296 after all ctrl commands have been executed on it. If init=1 the
11297 ENGINE is initialized at that point (ctrls before that point are run
11298 on the uninitialized ENGINE and after on the initialized one). If
11299 init=0 then the ENGINE will not be initialized at all.
11303 * Fix the 'app_verify_callback' interface so that the user-defined
11304 argument is actually passed to the callback: In the
11305 SSL_CTX_set_cert_verify_callback() prototype, the callback
11306 declaration has been changed from
11309 int (*cb)(X509_STORE_CTX *,void *);
11310 in ssl_verify_cert_chain (ssl/ssl_cert.c), the call
11311 i=s->ctx->app_verify_callback(&ctx)
11312 has been changed into
11313 i=s->ctx->app_verify_callback(&ctx, s->ctx->app_verify_arg).
11315 To update applications using SSL_CTX_set_cert_verify_callback(),
11316 a dummy argument can be added to their callback functions.
11318 *D. K. Smetters <smetters@parc.xerox.com>*
11320 * Added the '4758cca' ENGINE to support IBM 4758 cards.
11322 *Maurice Gittens <maurice@gittens.nl>, touchups by Geoff Thorpe*
11324 * Add and OPENSSL_LOAD_CONF define which will cause
11325 OpenSSL_add_all_algorithms() to load the openssl.cnf config file.
11326 This allows older applications to transparently support certain
11327 OpenSSL features: such as crypto acceleration and dynamic ENGINE loading.
11328 Two new functions OPENSSL_add_all_algorithms_noconf() which will never
11329 load the config file and OPENSSL_add_all_algorithms_conf() which will
11330 always load it have also been added.
11334 * Add the OFB, CFB and CTR (all with 128 bit feedback) to AES.
11335 Adjust NIDs and EVP layer.
11337 *Stephen Sprunk <stephen@sprunk.org> and Richard Levitte*
11339 * Config modules support in openssl utility.
11341 Most commands now load modules from the config file,
11342 though in a few (such as version) this isn't done
11343 because it couldn't be used for anything.
11345 In the case of ca and req the config file used is
11346 the same as the utility itself: that is the -config
11347 command line option can be used to specify an
11352 * Move default behaviour from OPENSSL_config(). If appname is NULL
11353 use "openssl_conf" if filename is NULL use default openssl config file.
11357 * Add an argument to OPENSSL_config() to allow the use of an alternative
11358 config section name. Add a new flag to tolerate a missing config file
11359 and move code to CONF_modules_load_file().
11363 * Support for crypto accelerator cards from Accelerated Encryption
11364 Processing, www.aep.ie. (Use engine 'aep')
11365 The support was copied from 0.9.6c [engine] and adapted/corrected
11366 to work with the new engine framework.
11368 *AEP Inc. and Richard Levitte*
11370 * Support for SureWare crypto accelerator cards from Baltimore
11371 Technologies. (Use engine 'sureware')
11372 The support was copied from 0.9.6c [engine] and adapted
11373 to work with the new engine framework.
11377 * Have the CHIL engine fork-safe (as defined by nCipher) and actually
11378 make the newer ENGINE framework commands for the CHIL engine work.
11380 *Toomas Kiisk <vix@cyber.ee> and Richard Levitte*
11382 * Make it possible to produce shared libraries on ReliantUNIX.
11384 *Robert Dahlem <Robert.Dahlem@ffm2.siemens.de> via Richard Levitte*
11386 * Add the configuration target debug-linux-ppro.
11387 Make 'openssl rsa' use the general key loading routines
11388 implemented in apps.c, and make those routines able to
11389 handle the key format FORMAT_NETSCAPE and the variant
11392 *Toomas Kiisk <vix@cyber.ee> via Richard Levitte*
11394 * Fix a crashbug and a logic bug in hwcrhk_load_pubkey().
11396 *Toomas Kiisk <vix@cyber.ee> via Richard Levitte*
11398 * Add -keyform to rsautl, and document -engine.
11400 *Richard Levitte, inspired by Toomas Kiisk <vix@cyber.ee>*
11402 * Change BIO_new_file (crypto/bio/bss_file.c) to use new
11403 BIO_R_NO_SUCH_FILE error code rather than the generic
11404 ERR_R_SYS_LIB error code if fopen() fails with ENOENT.
11408 * Add new functions
11409 ERR_peek_last_error
11410 ERR_peek_last_error_line
11411 ERR_peek_last_error_line_data.
11412 These are similar to
11414 ERR_peek_error_line
11415 ERR_peek_error_line_data,
11416 but report on the latest error recorded rather than the first one
11417 still in the error queue.
11419 *Ben Laurie, Bodo Moeller*
11421 * default_algorithms option in ENGINE config module. This allows things
11423 default_algorithms = ALL
11424 default_algorithms = RSA, DSA, RAND, CIPHERS, DIGESTS
11428 * Preliminary ENGINE config module.
11432 * New experimental application configuration code.
11436 * Change the AES code to follow the same name structure as all other
11437 symmetric ciphers, and behave the same way. Move everything to
11438 the directory crypto/aes, thereby obsoleting crypto/rijndael.
11440 *Stephen Sprunk <stephen@sprunk.org> and Richard Levitte*
11442 * SECURITY: remove unsafe setjmp/signal interaction from ui_openssl.c.
11444 *Ben Laurie and Theo de Raadt*
11446 * Add option to output public keys in req command.
11448 *Massimiliano Pala madwolf@openca.org*
11450 * Use wNAFs in EC_POINTs_mul() for improved efficiency
11451 (up to about 10% better than before for P-192 and P-224).
11455 * New functions/macros
11457 SSL_CTX_set_msg_callback(ctx, cb)
11458 SSL_CTX_set_msg_callback_arg(ctx, arg)
11459 SSL_set_msg_callback(ssl, cb)
11460 SSL_set_msg_callback_arg(ssl, arg)
11462 to request calling a callback function
11464 void cb(int write_p, int version, int content_type,
11465 const void *buf, size_t len, SSL *ssl, void *arg)
11467 whenever a protocol message has been completely received
11468 (write_p == 0) or sent (write_p == 1). Here 'version' is the
11469 protocol version according to which the SSL library interprets
11470 the current protocol message (SSL2_VERSION, SSL3_VERSION, or
11471 TLS1_VERSION). 'content_type' is 0 in the case of SSL 2.0, or
11472 the content type as defined in the SSL 3.0/TLS 1.0 protocol
11473 specification (change_cipher_spec(20), alert(21), handshake(22)).
11474 'buf' and 'len' point to the actual message, 'ssl' to the
11475 SSL object, and 'arg' is the application-defined value set by
11476 SSL[_CTX]_set_msg_callback_arg().
11478 'openssl s_client' and 'openssl s_server' have new '-msg' options
11479 to enable a callback that displays all protocol messages.
11483 * Change the shared library support so shared libraries are built as
11484 soon as the corresponding static library is finished, and thereby get
11485 openssl and the test programs linked against the shared library.
11486 This still only happens when the keyword "shard" has been given to
11487 the configuration scripts.
11489 NOTE: shared library support is still an experimental thing, and
11490 backward binary compatibility is still not guaranteed.
11492 *"Maciej W. Rozycki" <macro@ds2.pg.gda.pl> and Richard Levitte*
11494 * Add support for Subject Information Access extension.
11496 *Peter Sylvester <Peter.Sylvester@EdelWeb.fr>*
11498 * Make BUF_MEM_grow() behaviour more consistent: Initialise to zero
11499 additional bytes when new memory had to be allocated, not just
11500 when reusing an existing buffer.
11504 * New command line and configuration option 'utf8' for the req command.
11505 This allows field values to be specified as UTF8 strings.
11509 * Add -multi and -mr options to "openssl speed" - giving multiple parallel
11510 runs for the former and machine-readable output for the latter.
11514 * Add '-noemailDN' option to 'openssl ca'. This prevents inclusion
11515 of the e-mail address in the DN (i.e., it will go into a certificate
11516 extension only). The new configuration file option 'email_in_dn = no'
11517 has the same effect.
11519 *Massimiliano Pala madwolf@openca.org*
11521 * Change all functions with names starting with `des_` to be starting
11522 with `DES_` instead. Add wrappers that are compatible with libdes,
11523 but are named `_ossl_old_des_*`. Finally, add macros that map the
11524 `des_*` symbols to the corresponding `_ossl_old_des_*` if libdes
11525 compatibility is desired. If OpenSSL 0.9.6c compatibility is
11526 desired, the `des_*` symbols will be mapped to `DES_*`, with one
11529 Since we provide two compatibility mappings, the user needs to
11530 define the macro OPENSSL_DES_LIBDES_COMPATIBILITY if libdes
11531 compatibility is desired. The default (i.e., when that macro
11532 isn't defined) is OpenSSL 0.9.6c compatibility.
11534 There are also macros that enable and disable the support of old
11535 des functions altogether. Those are OPENSSL_ENABLE_OLD_DES_SUPPORT
11536 and OPENSSL_DISABLE_OLD_DES_SUPPORT. If none or both of those
11537 are defined, the default will apply: to support the old des routines.
11539 In either case, one must include openssl/des.h to get the correct
11540 definitions. Do not try to just include openssl/des_old.h, that
11543 NOTE: This is a major break of an old API into a new one. Software
11544 authors are encouraged to switch to the `DES_` style functions. Some
11545 time in the future, des_old.h and the libdes compatibility functions
11546 will be disable (i.e. OPENSSL_DISABLE_OLD_DES_SUPPORT will be the
11547 default), and then completely removed.
11551 * Test for certificates which contain unsupported critical extensions.
11552 If such a certificate is found during a verify operation it is
11553 rejected by default: this behaviour can be overridden by either
11554 handling the new error X509_V_ERR_UNHANDLED_CRITICAL_EXTENSION or
11555 by setting the verify flag X509_V_FLAG_IGNORE_CRITICAL. A new function
11556 X509_supported_extension() has also been added which returns 1 if a
11557 particular extension is supported.
11561 * Modify the behaviour of EVP cipher functions in similar way to digests
11562 to retain compatibility with existing code.
11566 * Modify the behaviour of EVP_DigestInit() and EVP_DigestFinal() to retain
11567 compatibility with existing code. In particular the 'ctx' parameter does
11568 not have to be to be initialized before the call to EVP_DigestInit() and
11569 it is tidied up after a call to EVP_DigestFinal(). New function
11570 EVP_DigestFinal_ex() which does not tidy up the ctx. Similarly function
11571 EVP_MD_CTX_copy() changed to not require the destination to be
11572 initialized valid and new function EVP_MD_CTX_copy_ex() added which
11573 requires the destination to be valid.
11575 Modify all the OpenSSL digest calls to use EVP_DigestInit_ex(),
11576 EVP_DigestFinal_ex() and EVP_MD_CTX_copy_ex().
11580 * Change ssl3_get_message (ssl/s3_both.c) and the functions using it
11581 so that complete 'Handshake' protocol structures are kept in memory
11582 instead of overwriting 'msg_type' and 'length' with 'body' data.
11586 * Add an implementation of SSL_add_dir_cert_subjects_to_stack for Win32.
11588 *Massimo Santin via Richard Levitte*
11590 * Major restructuring to the underlying ENGINE code. This includes
11591 reduction of linker bloat, separation of pure "ENGINE" manipulation
11592 (initialisation, etc) from functionality dealing with implementations
11593 of specific crypto interfaces. This change also introduces integrated
11594 support for symmetric ciphers and digest implementations - so ENGINEs
11595 can now accelerate these by providing EVP_CIPHER and EVP_MD
11596 implementations of their own. This is detailed in
11597 [crypto/engine/README.md](crypto/engine/README.md)
11598 as it couldn't be adequately described here. However, there are a few
11599 API changes worth noting - some RSA, DSA, DH, and RAND functions that
11600 were changed in the original introduction of ENGINE code have now
11601 reverted back - the hooking from this code to ENGINE is now a good
11602 deal more passive and at run-time, operations deal directly with
11603 RSA_METHODs, DSA_METHODs (etc) as they did before, rather than
11604 dereferencing through an ENGINE pointer any more. Also, the ENGINE
11605 functions dealing with `BN_MOD_EXP[_CRT]` handlers have been removed -
11606 they were not being used by the framework as there is no concept of a
11607 BIGNUM_METHOD and they could not be generalised to the new
11608 'ENGINE_TABLE' mechanism that underlies the new code. Similarly,
11609 ENGINE_cpy() has been removed as it cannot be consistently defined in
11614 * Change ASN1_GENERALIZEDTIME_check() to allow fractional seconds.
11618 * Change mkdef.pl to sort symbols that get the same entry number,
11619 and make sure the automatically generated functions `ERR_load_*`
11620 become part of libeay.num as well.
11624 * New function SSL_renegotiate_pending(). This returns true once
11625 renegotiation has been requested (either SSL_renegotiate() call
11626 or HelloRequest/ClientHello received from the peer) and becomes
11627 false once a handshake has been completed.
11628 (For servers, SSL_renegotiate() followed by SSL_do_handshake()
11629 sends a HelloRequest, but does not ensure that a handshake takes
11630 place. SSL_renegotiate_pending() is useful for checking if the
11631 client has followed the request.)
11635 * New SSL option SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION.
11636 By default, clients may request session resumption even during
11637 renegotiation (if session ID contexts permit); with this option,
11638 session resumption is possible only in the first handshake.
11640 SSL_OP_ALL is now 0x00000FFFL instead of 0x000FFFFFL. This makes
11641 more bits available for options that should not be part of
11642 SSL_OP_ALL (such as SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION).
11646 * Add some demos for certificate and certificate request creation.
11650 * Make maximum certificate chain size accepted from the peer application
11651 settable (`SSL*_get/set_max_cert_list()`), as proposed by
11652 "Douglas E. Engert" <deengert@anl.gov>.
11656 * Add support for shared libraries for Unixware-7
11657 (Boyd Lynn Gerber <gerberb@zenez.com>).
11661 * Add a "destroy" handler to ENGINEs that allows structural cleanup to
11662 be done prior to destruction. Use this to unload error strings from
11663 ENGINEs that load their own error strings. NB: This adds two new API
11664 functions to "get" and "set" this destroy handler in an ENGINE.
11668 * Alter all existing ENGINE implementations (except "openssl" and
11669 "openbsd") to dynamically instantiate their own error strings. This
11670 makes them more flexible to be built both as statically-linked ENGINEs
11671 and self-contained shared-libraries loadable via the "dynamic" ENGINE.
11672 Also, add stub code to each that makes building them as self-contained
11673 shared-libraries easier (see [README-Engine.md](README-Engine.md)).
11677 * Add a "dynamic" ENGINE that provides a mechanism for binding ENGINE
11678 implementations into applications that are completely implemented in
11679 self-contained shared-libraries. The "dynamic" ENGINE exposes control
11680 commands that can be used to configure what shared-library to load and
11681 to control aspects of the way it is handled. Also, made an update to
11682 the [README-Engine.md](README-Engine.md) file
11683 that brings its information up-to-date and
11684 provides some information and instructions on the "dynamic" ENGINE
11685 (ie. how to use it, how to build "dynamic"-loadable ENGINEs, etc).
11689 * Make it possible to unload ranges of ERR strings with a new
11690 "ERR_unload_strings" function.
11694 * Add a copy() function to EVP_MD.
11698 * Make EVP_MD routines take a context pointer instead of just the
11699 md_data void pointer.
11703 * Add flags to EVP_MD and EVP_MD_CTX. EVP_MD_FLAG_ONESHOT indicates
11704 that the digest can only process a single chunk of data
11705 (typically because it is provided by a piece of
11706 hardware). EVP_MD_CTX_FLAG_ONESHOT indicates that the application
11707 is only going to provide a single chunk of data, and hence the
11708 framework needn't accumulate the data for oneshot drivers.
11712 * As with "ERR", make it possible to replace the underlying "ex_data"
11713 functions. This change also alters the storage and management of global
11714 ex_data state - it's now all inside ex_data.c and all "class" code (eg.
11715 RSA, BIO, SSL_CTX, etc) no longer stores its own STACKS and per-class
11716 index counters. The API functions that use this state have been changed
11717 to take a "class_index" rather than pointers to the class's local STACK
11718 and counter, and there is now an API function to dynamically create new
11719 classes. This centralisation allows us to (a) plug a lot of the
11720 thread-safety problems that existed, and (b) makes it possible to clean
11721 up all allocated state using "CRYPTO_cleanup_all_ex_data()". W.r.t. (b)
11722 such data would previously have always leaked in application code and
11723 workarounds were in place to make the memory debugging turn a blind eye
11724 to it. Application code that doesn't use this new function will still
11725 leak as before, but their memory debugging output will announce it now
11726 rather than letting it slide.
11728 Besides the addition of CRYPTO_cleanup_all_ex_data(), another API change
11729 induced by the "ex_data" overhaul is that X509_STORE_CTX_init() now
11730 has a return value to indicate success or failure.
11734 * Make it possible to replace the underlying "ERR" functions such that the
11735 global state (2 LHASH tables and 2 locks) is only used by the "default"
11736 implementation. This change also adds two functions to "get" and "set"
11737 the implementation prior to it being automatically set the first time
11738 any other ERR function takes place. Ie. an application can call "get",
11739 pass the return value to a module it has just loaded, and that module
11740 can call its own "set" function using that value. This means the
11741 module's "ERR" operations will use (and modify) the error state in the
11742 application and not in its own statically linked copy of OpenSSL code.
11746 * Give DH, DSA, and RSA types their own `*_up_ref()` function to increment
11747 reference counts. This performs normal REF_PRINT/REF_CHECK macros on
11748 the operation, and provides a more encapsulated way for external code
11749 (crypto/evp/ and ssl/) to do this. Also changed the evp and ssl code
11750 to use these functions rather than manually incrementing the counts.
11752 Also rename "DSO_up()" function to more descriptive "DSO_up_ref()".
11756 * Add EVP test program.
11760 * Add symmetric cipher support to ENGINE. Expect the API to change!
11764 * New CRL functions: X509_CRL_set_version(), X509_CRL_set_issuer_name()
11765 X509_CRL_set_lastUpdate(), X509_CRL_set_nextUpdate(), X509_CRL_sort(),
11766 X509_REVOKED_set_serialNumber(), and X509_REVOKED_set_revocationDate().
11767 These allow a CRL to be built without having to access X509_CRL fields
11768 directly. Modify 'ca' application to use new functions.
11772 * Move SSL_OP_TLS_ROLLBACK_BUG out of the SSL_OP_ALL list of recommended
11773 bug workarounds. Rollback attack detection is a security feature.
11774 The problem will only arise on OpenSSL servers when TLSv1 is not
11775 available (sslv3_server_method() or SSL_OP_NO_TLSv1).
11776 Software authors not wanting to support TLSv1 will have special reasons
11777 for their choice and can explicitly enable this option.
11779 *Bodo Moeller, Lutz Jaenicke*
11781 * Rationalise EVP so it can be extended: don't include a union of
11782 cipher/digest structures, add init/cleanup functions for EVP_MD_CTX
11783 (similar to those existing for EVP_CIPHER_CTX).
11788 EVP_MD_CTX_init(&md); /* new function call */
11789 EVP_DigestInit(&md, EVP_sha1());
11790 EVP_DigestUpdate(&md, in, len);
11791 EVP_DigestFinal(&md, out, NULL);
11792 EVP_MD_CTX_cleanup(&md); /* new function call */
11796 * Make DES key schedule conform to the usual scheme, as well as
11797 correcting its structure. This means that calls to DES functions
11798 now have to pass a pointer to a des_key_schedule instead of a
11799 plain des_key_schedule (which was actually always a pointer
11802 des_key_schedule ks;
11804 des_set_key_checked(..., &ks);
11805 des_ncbc_encrypt(..., &ks, ...);
11807 (Note that a later change renames 'des_...' into 'DES_...'.)
11811 * Initial reduction of linker bloat: the use of some functions, such as
11812 PEM causes large amounts of unused functions to be linked in due to
11813 poor organisation. For example pem_all.c contains every PEM function
11814 which has a knock on effect of linking in large amounts of (unused)
11815 ASN1 code. Grouping together similar functions and splitting unrelated
11816 functions prevents this.
11820 * Cleanup of EVP macros.
11824 * Change historical references to `{NID,SN,LN}_des_ede` and ede3 to add the
11825 correct `_ecb suffix`.
11829 * Add initial OCSP responder support to ocsp application. The
11830 revocation information is handled using the text based index
11831 use by the ca application. The responder can either handle
11832 requests generated internally, supplied in files (for example
11833 via a CGI script) or using an internal minimal server.
11837 * Add configuration choices to get zlib compression for TLS.
11841 * Changes to Kerberos SSL for RFC 2712 compliance:
11842 1. Implemented real KerberosWrapper, instead of just using
11843 KRB5 AP_REQ message. [Thanks to Simon Wilkinson <sxw@sxw.org.uk>]
11844 2. Implemented optional authenticator field of KerberosWrapper.
11846 Added openssl-style ASN.1 macros for Kerberos ticket, ap_req,
11847 and authenticator structs; see crypto/krb5/.
11849 Generalized Kerberos calls to support multiple Kerberos libraries.
11850 *Vern Staats <staatsvr@asc.hpc.mil>, Jeffrey Altman <jaltman@columbia.edu>
11851 via Richard Levitte*
11853 * Cause 'openssl speed' to use fully hard-coded DSA keys as it
11854 already does with RSA. testdsa.h now has 'priv_key/pub_key'
11855 values for each of the key sizes rather than having just
11856 parameters (and 'speed' generating keys each time).
11860 * Speed up EVP routines.
11863 pe 8 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes
11864 s-cbc 4408.85k 5560.51k 5778.46k 5862.20k 5825.16k
11865 s-cbc 4389.55k 5571.17k 5792.23k 5846.91k 5832.11k
11866 s-cbc 4394.32k 5575.92k 5807.44k 5848.37k 5841.30k
11868 s-cbc 3482.66k 5069.49k 5496.39k 5614.16k 5639.28k
11869 s-cbc 3480.74k 5068.76k 5510.34k 5609.87k 5635.52k
11870 s-cbc 3483.72k 5067.62k 5504.60k 5708.01k 5724.80k
11873 s-cbc 4660.16k 5650.19k 5807.19k 5827.13k 5783.32k
11875 s-cbc 3624.96k 5258.21k 5530.91k 5624.30k 5628.26k
11879 * Added the OS2-EMX target.
11881 *"Brian Havard" <brianh@kheldar.apana.org.au> and Richard Levitte*
11883 * Rewrite apps to use NCONF routines instead of the old CONF. New functions
11884 to support NCONF routines in extension code. New function CONF_set_nconf()
11885 to allow functions which take an NCONF to also handle the old LHASH
11886 structure: this means that the old CONF compatible routines can be
11887 retained (in particular wrt extensions) without having to duplicate the
11888 code. New function X509V3_add_ext_nconf_sk to add extensions to a stack.
11892 * Enhance the general user interface with mechanisms for inner control
11893 and with possibilities to have yes/no kind of prompts.
11897 * Change all calls to low level digest routines in the library and
11898 applications to use EVP. Add missing calls to HMAC_cleanup() and
11899 don't assume HMAC_CTX can be copied using memcpy().
11901 *Verdon Walker <VWalker@novell.com>, Steve Henson*
11903 * Add the possibility to control engines through control names but with
11904 arbitrary arguments instead of just a string.
11905 Change the key loaders to take a UI_METHOD instead of a callback
11906 function pointer. NOTE: this breaks binary compatibility with earlier
11907 versions of OpenSSL [engine].
11908 Adapt the nCipher code for these new conditions and add a card insertion
11913 * Enhance the general user interface with mechanisms to better support
11914 dialog box interfaces, application-defined prompts, the possibility
11915 to use defaults (for example default passwords from somewhere else)
11916 and interrupts/cancellations.
11920 * Tidy up PKCS#12 attribute handling. Add support for the CSP name
11921 attribute in PKCS#12 files, add new -CSP option to pkcs12 utility.
11925 * Fix a memory leak in 'sk_dup()' in the case reallocation fails. (Also
11926 tidy up some unnecessarily weird code in 'sk_new()').
11928 *Geoff, reported by Diego Tartara <dtartara@novamens.com>*
11930 * Change the key loading routines for ENGINEs to use the same kind
11931 callback (pem_password_cb) as all other routines that need this
11936 * Increase ENTROPY_NEEDED to 32 bytes, as Rijndael can operate with
11937 256 bit (=32 byte) keys. Of course seeding with more entropy bytes
11938 than this minimum value is recommended.
11942 * New random seeder for OpenVMS, using the system process statistics
11943 that are easily reachable.
11947 * Windows apparently can't transparently handle global
11948 variables defined in DLLs. Initialisations such as:
11950 const ASN1_ITEM *it = &ASN1_INTEGER_it;
11952 won't compile. This is used by the any applications that need to
11953 declare their own ASN1 modules. This was fixed by adding the option
11954 EXPORT_VAR_AS_FN to all Win32 platforms, although this isn't strictly
11955 needed for static libraries under Win32.
11959 * New functions X509_PURPOSE_set() and X509_TRUST_set() to handle
11960 setting of purpose and trust fields. New X509_STORE trust and
11961 purpose functions and tidy up setting in other SSL functions.
11965 * Add copies of X509_STORE_CTX fields and callbacks to X509_STORE
11966 structure. These are inherited by X509_STORE_CTX when it is
11967 initialised. This allows various defaults to be set in the
11968 X509_STORE structure (such as flags for CRL checking and custom
11969 purpose or trust settings) for functions which only use X509_STORE_CTX
11970 internally such as S/MIME.
11972 Modify X509_STORE_CTX_purpose_inherit() so it only sets purposes and
11973 trust settings if they are not set in X509_STORE. This allows X509_STORE
11974 purposes and trust (in S/MIME for example) to override any set by default.
11976 Add command line options for CRL checking to smime, s_client and s_server
11981 * Initial CRL based revocation checking. If the CRL checking flag(s)
11982 are set then the CRL is looked up in the X509_STORE structure and
11983 its validity and signature checked, then if the certificate is found
11984 in the CRL the verify fails with a revoked error.
11986 Various new CRL related callbacks added to X509_STORE_CTX structure.
11988 Command line options added to 'verify' application to support this.
11990 This needs some additional work, such as being able to handle multiple
11991 CRLs with different times, extension based lookup (rather than just
11992 by subject name) and ultimately more complete V2 CRL extension
11997 * Add a general user interface API (crypto/ui/). This is designed
11998 to replace things like des_read_password and friends (backward
11999 compatibility functions using this new API are provided).
12000 The purpose is to remove prompting functions from the DES code
12001 section as well as provide for prompting through dialog boxes in
12002 a window system and the like.
12006 * Add "ex_data" support to ENGINE so implementations can add state at a
12007 per-structure level rather than having to store it globally.
12011 * Make it possible for ENGINE structures to be copied when retrieved by
12012 ENGINE_by_id() if the ENGINE specifies a new flag: ENGINE_FLAGS_BY_ID_COPY.
12013 This causes the "original" ENGINE structure to act like a template,
12014 analogous to the RSA vs. RSA_METHOD type of separation. Because of this
12015 operational state can be localised to each ENGINE structure, despite the
12016 fact they all share the same "methods". New ENGINE structures returned in
12017 this case have no functional references and the return value is the single
12018 structural reference. This matches the single structural reference returned
12019 by ENGINE_by_id() normally, when it is incremented on the pre-existing
12024 * Fix ASN1 decoder when decoding type ANY and V_ASN1_OTHER: since this
12025 needs to match any other type at all we need to manually clear the
12030 * Changes to the "openssl engine" utility to include;
12031 - verbosity levels ('-v', '-vv', and '-vvv') that provide information
12032 about an ENGINE's available control commands.
12033 - executing control commands from command line arguments using the
12034 '-pre' and '-post' switches. '-post' is only used if '-t' is
12035 specified and the ENGINE is successfully initialised. The syntax for
12036 the individual commands are colon-separated, for example;
12037 openssl engine chil -pre FORK_CHECK:0 -pre SO_PATH:/lib/test.so
12041 * New dynamic control command support for ENGINEs. ENGINEs can now
12042 declare their own commands (numbers), names (strings), descriptions,
12043 and input types for run-time discovery by calling applications. A
12044 subset of these commands are implicitly classed as "executable"
12045 depending on their input type, and only these can be invoked through
12046 the new string-based API function ENGINE_ctrl_cmd_string(). (Eg. this
12047 can be based on user input, config files, etc). The distinction is
12048 that "executable" commands cannot return anything other than a boolean
12049 result and can only support numeric or string input, whereas some
12050 discoverable commands may only be for direct use through
12051 ENGINE_ctrl(), eg. supporting the exchange of binary data, function
12052 pointers, or other custom uses. The "executable" commands are to
12053 support parameterisations of ENGINE behaviour that can be
12054 unambiguously defined by ENGINEs and used consistently across any
12055 OpenSSL-based application. Commands have been added to all the
12056 existing hardware-supporting ENGINEs, noticeably "SO_PATH" to allow
12057 control over shared-library paths without source code alterations.
12061 * Changed all ENGINE implementations to dynamically allocate their
12062 ENGINEs rather than declaring them statically. Apart from this being
12063 necessary with the removal of the ENGINE_FLAGS_MALLOCED distinction,
12064 this also allows the implementations to compile without using the
12065 internal engine_int.h header.
12069 * Minor adjustment to "rand" code. RAND_get_rand_method() now returns a
12070 'const' value. Any code that should be able to modify a RAND_METHOD
12071 should already have non-const pointers to it (ie. they should only
12072 modify their own ones).
12076 * Made a variety of little tweaks to the ENGINE code.
12077 - "atalla" and "ubsec" string definitions were moved from header files
12078 to C code. "nuron" string definitions were placed in variables
12079 rather than hard-coded - allowing parameterisation of these values
12080 later on via ctrl() commands.
12081 - Removed unused "#if 0"'d code.
12082 - Fixed engine list iteration code so it uses ENGINE_free() to release
12083 structural references.
12084 - Constified the RAND_METHOD element of ENGINE structures.
12085 - Constified various get/set functions as appropriate and added
12086 missing functions (including a catch-all ENGINE_cpy that duplicates
12087 all ENGINE values onto a new ENGINE except reference counts/state).
12088 - Removed NULL parameter checks in get/set functions. Setting a method
12089 or function to NULL is a way of cancelling out a previously set
12090 value. Passing a NULL ENGINE parameter is just plain stupid anyway
12091 and doesn't justify the extra error symbols and code.
12092 - Deprecate the ENGINE_FLAGS_MALLOCED define and move the area for
12093 flags from engine_int.h to engine.h.
12094 - Changed prototypes for ENGINE handler functions (init(), finish(),
12095 ctrl(), key-load functions, etc) to take an (ENGINE*) parameter.
12099 * Implement binary inversion algorithm for BN_mod_inverse in addition
12100 to the algorithm using long division. The binary algorithm can be
12101 used only if the modulus is odd. On 32-bit systems, it is faster
12102 only for relatively small moduli (roughly 20-30% for 128-bit moduli,
12103 roughly 5-15% for 256-bit moduli), so we use it only for moduli
12104 up to 450 bits. In 64-bit environments, the binary algorithm
12105 appears to be advantageous for much longer moduli; here we use it
12106 for moduli up to 2048 bits.
12110 * Rewrite CHOICE field setting in ASN1_item_ex_d2i(). The old code
12111 could not support the combine flag in choice fields.
12115 * Add a 'copy_extensions' option to the 'ca' utility. This copies
12116 extensions from a certificate request to the certificate.
12120 * Allow multiple 'certopt' and 'nameopt' options to be separated
12121 by commas. Add 'namopt' and 'certopt' options to the 'ca' config
12122 file: this allows the display of the certificate about to be
12123 signed to be customised, to allow certain fields to be included
12124 or excluded and extension details. The old system didn't display
12125 multicharacter strings properly, omitted fields not in the policy
12126 and couldn't display additional details such as extensions.
12130 * Function EC_POINTs_mul for multiple scalar multiplication
12131 of an arbitrary number of elliptic curve points
12132 \sum scalars[i]*points[i],
12133 optionally including the generator defined for the EC_GROUP:
12134 scalar*generator + \sum scalars[i]*points[i].
12136 EC_POINT_mul is a simple wrapper function for the typical case
12137 that the point list has just one item (besides the optional
12142 * First EC_METHODs for curves over GF(p):
12144 EC_GFp_simple_method() uses the basic BN_mod_mul and BN_mod_sqr
12145 operations and provides various method functions that can also
12146 operate with faster implementations of modular arithmetic.
12148 EC_GFp_mont_method() reuses most functions that are part of
12149 EC_GFp_simple_method, but uses Montgomery arithmetic.
12151 *Bodo Moeller; point addition and point doubling
12152 implementation directly derived from source code provided by
12153 Lenka Fibikova <fibikova@exp-math.uni-essen.de>*
12155 * Framework for elliptic curves (crypto/ec/ec.h, crypto/ec/ec_lcl.h,
12156 crypto/ec/ec_lib.c):
12158 Curves are EC_GROUP objects (with an optional group generator)
12159 based on EC_METHODs that are built into the library.
12161 Points are EC_POINT objects based on EC_GROUP objects.
12163 Most of the framework would be able to handle curves over arbitrary
12164 finite fields, but as there are no obvious types for fields other
12165 than GF(p), some functions are limited to that for now.
12169 * Add the -HTTP option to s_server. It is similar to -WWW, but requires
12170 that the file contains a complete HTTP response.
12174 * Add the ec directory to mkdef.pl and mkfiles.pl. In mkdef.pl
12175 change the def and num file printf format specifier from "%-40sXXX"
12176 to "%-39s XXX". The latter will always guarantee a space after the
12177 field while the former will cause them to run together if the field
12178 is 40 of more characters long.
12182 * Constify the cipher and digest 'method' functions and structures
12183 and modify related functions to take constant EVP_MD and EVP_CIPHER
12188 * Hide BN_CTX structure details in bn_lcl.h instead of publishing them
12189 in <openssl/bn.h>. Also further increase BN_CTX_NUM to 32.
12193 * Modify `EVP_Digest*()` routines so they now return values. Although the
12194 internal software routines can never fail additional hardware versions
12199 * Clean up crypto/err/err.h and change some error codes to avoid conflicts:
12201 Previously ERR_R_FATAL was too small and coincided with ERR_LIB_PKCS7
12202 (= ERR_R_PKCS7_LIB); it is now 64 instead of 32.
12205 ERR_R_NESTED_ASN1_ERROR
12207 ERR_R_MISSING_ASN1_EOS
12208 were 4 .. 9, conflicting with
12209 ERR_LIB_RSA (= ERR_R_RSA_LIB)
12211 ERR_LIB_PEM (= ERR_R_PEM_LIB).
12212 They are now 58 .. 63 (i.e., just below ERR_R_FATAL).
12214 Add new error code 'ERR_R_INTERNAL_ERROR'.
12218 * Don't overuse locks in crypto/err/err.c: For data retrieval, CRYPTO_r_lock
12223 * New option '-subj arg' for 'openssl req' and 'openssl ca'. This
12224 sets the subject name for a new request or supersedes the
12225 subject name in a given request. Formats that can be parsed are
12226 'CN=Some Name, OU=myOU, C=IT'
12228 'CN=Some Name/OU=myOU/C=IT'.
12230 Add options '-batch' and '-verbose' to 'openssl req'.
12232 *Massimiliano Pala <madwolf@hackmasters.net>*
12234 * Introduce the possibility to access global variables through
12235 functions on platform were that's the best way to handle exporting
12236 global variables in shared libraries. To enable this functionality,
12237 one must configure with "EXPORT_VAR_AS_FN" or defined the C macro
12238 "OPENSSL_EXPORT_VAR_AS_FUNCTION" in crypto/opensslconf.h (the latter
12239 is normally done by Configure or something similar).
12241 To implement a global variable, use the macro OPENSSL_IMPLEMENT_GLOBAL
12242 in the source file (foo.c) like this:
12244 OPENSSL_IMPLEMENT_GLOBAL(int,foo)=1;
12245 OPENSSL_IMPLEMENT_GLOBAL(double,bar);
12247 To declare a global variable, use the macros OPENSSL_DECLARE_GLOBAL
12248 and OPENSSL_GLOBAL_REF in the header file (foo.h) like this:
12250 OPENSSL_DECLARE_GLOBAL(int,foo);
12251 #define foo OPENSSL_GLOBAL_REF(foo)
12252 OPENSSL_DECLARE_GLOBAL(double,bar);
12253 #define bar OPENSSL_GLOBAL_REF(bar)
12255 The #defines are very important, and therefore so is including the
12256 header file everywhere where the defined globals are used.
12258 The macro OPENSSL_EXPORT_VAR_AS_FUNCTION also affects the definition
12259 of ASN.1 items, but that structure is a bit different.
12261 The largest change is in util/mkdef.pl which has been enhanced with
12262 better and easier to understand logic to choose which symbols should
12263 go into the Windows .def files as well as a number of fixes and code
12264 cleanup (among others, algorithm keywords are now sorted
12265 lexicographically to avoid constant rewrites).
12269 * In BN_div() keep a copy of the sign of 'num' before writing the
12270 result to 'rm' because if rm==num the value will be overwritten
12271 and produce the wrong result if 'num' is negative: this caused
12272 problems with BN_mod() and BN_nnmod().
12276 * Function OCSP_request_verify(). This checks the signature on an
12277 OCSP request and verifies the signer certificate. The signer
12278 certificate is just checked for a generic purpose and OCSP request
12283 * Add OCSP_check_validity() function to check the validity of OCSP
12284 responses. OCSP responses are prepared in real time and may only
12285 be a few seconds old. Simply checking that the current time lies
12286 between thisUpdate and nextUpdate max reject otherwise valid responses
12287 caused by either OCSP responder or client clock inaccuracy. Instead
12288 we allow thisUpdate and nextUpdate to fall within a certain period of
12289 the current time. The age of the response can also optionally be
12290 checked. Two new options -validity_period and -status_age added to
12295 * If signature or public key algorithm is unrecognized print out its
12296 OID rather that just UNKNOWN.
12300 * Change OCSP_cert_to_id() to tolerate a NULL subject certificate and
12301 OCSP_cert_id_new() a NULL serialNumber. This allows a partial certificate
12302 ID to be generated from the issuer certificate alone which can then be
12303 passed to OCSP_id_issuer_cmp().
12307 * New compilation option ASN1_ITEM_FUNCTIONS. This causes the new
12308 ASN1 modules to export functions returning ASN1_ITEM pointers
12309 instead of the ASN1_ITEM structures themselves. This adds several
12310 new macros which allow the underlying ASN1 function/structure to
12311 be accessed transparently. As a result code should not use ASN1_ITEM
12312 references directly (such as &X509_it) but instead use the relevant
12313 macros (such as ASN1_ITEM_rptr(X509)). This option is to allow
12314 use of the new ASN1 code on platforms where exporting structures
12315 is problematical (for example in shared libraries) but exporting
12316 functions returning pointers to structures is not.
12320 * Add support for overriding the generation of SSL/TLS session IDs.
12321 These callbacks can be registered either in an SSL_CTX or per SSL.
12322 The purpose of this is to allow applications to control, if they wish,
12323 the arbitrary values chosen for use as session IDs, particularly as it
12324 can be useful for session caching in multiple-server environments. A
12325 command-line switch for testing this (and any client code that wishes
12326 to use such a feature) has been added to "s_server".
12328 *Geoff Thorpe, Lutz Jaenicke*
12330 * Modify mkdef.pl to recognise and parse preprocessor conditionals
12331 of the form '#if defined(...) || defined(...) || ...' and
12332 '#if !defined(...) && !defined(...) && ...'. This also avoids
12333 the growing number of special cases it was previously handling.
12337 * Make all configuration macros available for application by making
12338 sure they are available in opensslconf.h, by giving them names starting
12339 with `OPENSSL_` to avoid conflicts with other packages and by making
12340 sure e_os2.h will cover all platform-specific cases together with
12342 Additionally, it is now possible to define configuration/platform-
12343 specific names (called "system identities"). In the C code, these
12344 are prefixed with `OPENSSL_SYSNAME_`. e_os2.h will create another
12345 macro with the name beginning with `OPENSSL_SYS_`, which is determined
12346 from `OPENSSL_SYSNAME_*` or compiler-specific macros depending on
12351 * New option -set_serial to 'req' and 'x509' this allows the serial
12352 number to use to be specified on the command line. Previously self
12353 signed certificates were hard coded with serial number 0 and the
12354 CA options of 'x509' had to use a serial number in a file which was
12359 * New options to 'ca' utility to support V2 CRL entry extensions.
12360 Currently CRL reason, invalidity date and hold instruction are
12361 supported. Add new CRL extensions to V3 code and some new objects.
12365 * New function EVP_CIPHER_CTX_set_padding() this is used to
12366 disable standard block padding (aka PKCS#5 padding) in the EVP
12367 API, which was previously mandatory. This means that the data is
12368 not padded in any way and so the total length much be a multiple
12369 of the block size, otherwise an error occurs.
12373 * Initial (incomplete) OCSP SSL support.
12377 * New function OCSP_parse_url(). This splits up a URL into its host,
12378 port and path components: primarily to parse OCSP URLs. New -url
12379 option to ocsp utility.
12383 * New nonce behavior. The return value of OCSP_check_nonce() now
12384 reflects the various checks performed. Applications can decide
12385 whether to tolerate certain situations such as an absent nonce
12386 in a response when one was present in a request: the ocsp application
12387 just prints out a warning. New function OCSP_add1_basic_nonce()
12388 this is to allow responders to include a nonce in a response even if
12389 the request is nonce-less.
12393 * Disable stdin buffering in load_cert (apps/apps.c) so that no certs are
12394 skipped when using openssl x509 multiple times on a single input file,
12395 e.g. `(openssl x509 -out cert1; openssl x509 -out cert2) <certs`.
12399 * Make ASN1_UTCTIME_set_string() and ASN1_GENERALIZEDTIME_set_string()
12400 set string type: to handle setting ASN1_TIME structures. Fix ca
12401 utility to correctly initialize revocation date of CRLs.
12405 * New option SSL_OP_CIPHER_SERVER_PREFERENCE allows the server to override
12406 the clients preferred ciphersuites and rather use its own preferences.
12407 Should help to work around M$ SGC (Server Gated Cryptography) bug in
12408 Internet Explorer by ensuring unchanged hash method during stepup.
12409 (Also replaces the broken/deactivated SSL_OP_NON_EXPORT_FIRST option.)
12413 * Make mkdef.pl recognise all DECLARE_ASN1 macros, change rijndael
12414 to aes and add a new 'exist' option to print out symbols that don't
12419 * Additional options to ocsp utility to allow flags to be set and
12420 additional certificates supplied.
12424 * Add the option -VAfile to 'openssl ocsp', so the user can give the
12425 OCSP client a number of certificate to only verify the response
12430 * Update Rijndael code to version 3.0 and change EVP AES ciphers to
12431 handle the new API. Currently only ECB, CBC modes supported. Add new
12434 Add TLS AES ciphersuites as described in RFC3268, "Advanced
12435 Encryption Standard (AES) Ciphersuites for Transport Layer
12436 Security (TLS)". (In beta versions of OpenSSL 0.9.7, these were
12437 not enabled by default and were not part of the "ALL" ciphersuite
12438 alias because they were not yet official; they could be
12439 explicitly requested by specifying the "AESdraft" ciphersuite
12440 group alias. In the final release of OpenSSL 0.9.7, the group
12441 alias is called "AES" and is part of "ALL".)
12443 *Ben Laurie, Steve Henson, Bodo Moeller*
12445 * New function OCSP_copy_nonce() to copy nonce value (if present) from
12446 request to response.
12450 * Functions for OCSP responders. OCSP_request_onereq_count(),
12451 OCSP_request_onereq_get0(), OCSP_onereq_get0_id() and OCSP_id_get0_info()
12452 extract information from a certificate request. OCSP_response_create()
12453 creates a response and optionally adds a basic response structure.
12454 OCSP_basic_add1_status() adds a complete single response to a basic
12455 response and returns the OCSP_SINGLERESP structure just added (to allow
12456 extensions to be included for example). OCSP_basic_add1_cert() adds a
12457 certificate to a basic response and OCSP_basic_sign() signs a basic
12458 response with various flags. New helper functions ASN1_TIME_check()
12459 (checks validity of ASN1_TIME structure) and ASN1_TIME_to_generalizedtime()
12460 (converts ASN1_TIME to GeneralizedTime).
12464 * Various new functions. EVP_Digest() combines EVP_Digest{Init,Update,Final}()
12465 in a single operation. X509_get0_pubkey_bitstr() extracts the public_key
12466 structure from a certificate. X509_pubkey_digest() digests the public_key
12467 contents: this is used in various key identifiers.
12471 * Make sk_sort() tolerate a NULL argument.
12473 *Steve Henson reported by Massimiliano Pala <madwolf@comune.modena.it>*
12475 * New OCSP verify flag OCSP_TRUSTOTHER. When set the "other" certificates
12476 passed by the function are trusted implicitly. If any of them signed the
12477 response then it is assumed to be valid and is not verified.
12481 * In PKCS7_set_type() initialise content_type in PKCS7_ENC_CONTENT
12482 to data. This was previously part of the PKCS7 ASN1 code. This
12483 was causing problems with OpenSSL created PKCS#12 and PKCS#7 structures.
12484 *Steve Henson, reported by Kenneth R. Robinette
12485 <support@securenetterm.com>*
12487 * Add CRYPTO_push_info() and CRYPTO_pop_info() calls to new ASN1
12488 routines: without these tracing memory leaks is very painful.
12489 Fix leaks in PKCS12 and PKCS7 routines.
12493 * Make X509_time_adj() cope with the new behaviour of ASN1_TIME_new().
12494 Previously it initialised the 'type' argument to V_ASN1_UTCTIME which
12495 effectively meant GeneralizedTime would never be used. Now it
12496 is initialised to -1 but X509_time_adj() now has to check the value
12497 and use ASN1_TIME_set() if the value is not V_ASN1_UTCTIME or
12498 V_ASN1_GENERALIZEDTIME, without this it always uses GeneralizedTime.
12499 *Steve Henson, reported by Kenneth R. Robinette
12500 <support@securenetterm.com>*
12502 * Fixes to BN_to_ASN1_INTEGER when bn is zero. This would previously
12503 result in a zero length in the ASN1_INTEGER structure which was
12504 not consistent with the structure when d2i_ASN1_INTEGER() was used
12505 and would cause ASN1_INTEGER_cmp() to fail. Enhance s2i_ASN1_INTEGER()
12506 to cope with hex and negative integers. Fix bug in i2a_ASN1_INTEGER()
12507 where it did not print out a minus for negative ASN1_INTEGER.
12511 * Add summary printout to ocsp utility. The various functions which
12512 convert status values to strings have been renamed to:
12513 OCSP_response_status_str(), OCSP_cert_status_str() and
12514 OCSP_crl_reason_str() and are no longer static. New options
12515 to verify nonce values and to disable verification. OCSP response
12516 printout format cleaned up.
12520 * Add additional OCSP certificate checks. These are those specified
12521 in RFC2560. This consists of two separate checks: the CA of the
12522 certificate being checked must either be the OCSP signer certificate
12523 or the issuer of the OCSP signer certificate. In the latter case the
12524 OCSP signer certificate must contain the OCSP signing extended key
12525 usage. This check is performed by attempting to match the OCSP
12526 signer or the OCSP signer CA to the issuerNameHash and issuerKeyHash
12527 in the OCSP_CERTID structures of the response.
12531 * Initial OCSP certificate verification added to OCSP_basic_verify()
12532 and related routines. This uses the standard OpenSSL certificate
12533 verify routines to perform initial checks (just CA validity) and
12534 to obtain the certificate chain. Then additional checks will be
12535 performed on the chain. Currently the root CA is checked to see
12536 if it is explicitly trusted for OCSP signing. This is used to set
12537 a root CA as a global signing root: that is any certificate that
12538 chains to that CA is an acceptable OCSP signing certificate.
12542 * New '-extfile ...' option to 'openssl ca' for reading X.509v3
12543 extensions from a separate configuration file.
12544 As when reading extensions from the main configuration file,
12545 the '-extensions ...' option may be used for specifying the
12548 *Massimiliano Pala <madwolf@comune.modena.it>*
12550 * New OCSP utility. Allows OCSP requests to be generated or
12551 read. The request can be sent to a responder and the output
12552 parsed, outputted or printed in text form. Not complete yet:
12553 still needs to check the OCSP response validity.
12557 * New subcommands for 'openssl ca':
12558 `openssl ca -status <serial>` prints the status of the cert with
12559 the given serial number (according to the index file).
12560 `openssl ca -updatedb` updates the expiry status of certificates
12563 *Massimiliano Pala <madwolf@comune.modena.it>*
12565 * New '-newreq-nodes' command option to CA.pl. This is like
12566 '-newreq', but calls 'openssl req' with the '-nodes' option
12567 so that the resulting key is not encrypted.
12569 *Damien Miller <djm@mindrot.org>*
12571 * New configuration for the GNU Hurd.
12573 *Jonathan Bartlett <johnnyb@wolfram.com> via Richard Levitte*
12575 * Initial code to implement OCSP basic response verify. This
12576 is currently incomplete. Currently just finds the signer's
12577 certificate and verifies the signature on the response.
12581 * New SSLeay_version code SSLEAY_DIR to determine the compiled-in
12582 value of OPENSSLDIR. This is available via the new '-d' option
12583 to 'openssl version', and is also included in 'openssl version -a'.
12587 * Allowing defining memory allocation callbacks that will be given
12588 file name and line number information in additional arguments
12589 (a `const char*` and an int). The basic functionality remains, as
12590 well as the original possibility to just replace malloc(),
12591 realloc() and free() by functions that do not know about these
12592 additional arguments. To register and find out the current
12593 settings for extended allocation functions, the following
12594 functions are provided:
12596 CRYPTO_set_mem_ex_functions
12597 CRYPTO_set_locked_mem_ex_functions
12598 CRYPTO_get_mem_ex_functions
12599 CRYPTO_get_locked_mem_ex_functions
12601 These work the same way as CRYPTO_set_mem_functions and friends.
12602 `CRYPTO_get_[locked_]mem_functions` now writes 0 where such an
12603 extended allocation function is enabled.
12604 Similarly, `CRYPTO_get_[locked_]mem_ex_functions` writes 0 where
12605 a conventional allocation function is enabled.
12607 *Richard Levitte, Bodo Moeller*
12609 * Finish off removing the remaining LHASH function pointer casts.
12610 There should no longer be any prototype-casting required when using
12611 the LHASH abstraction, and any casts that remain are "bugs". See
12612 the callback types and macros at the head of lhash.h for details
12613 (and "OBJ_cleanup" in crypto/objects/obj_dat.c as an example).
12617 * Add automatic query of EGD sockets in RAND_poll() for the unix variant.
12618 If /dev/[u]random devices are not available or do not return enough
12619 entropy, EGD style sockets (served by EGD or PRNGD) will automatically
12621 The locations /var/run/egd-pool, /dev/egd-pool, /etc/egd-pool, and
12622 /etc/entropy will be queried once each in this sequence, querying stops
12623 when enough entropy was collected without querying more sockets.
12627 * Change the Unix RAND_poll() variant to be able to poll several
12628 random devices, as specified by DEVRANDOM, until a sufficient amount
12629 of data has been collected. We spend at most 10 ms on each file
12630 (select timeout) and read in non-blocking mode. DEVRANDOM now
12631 defaults to the list "/dev/urandom", "/dev/random", "/dev/srandom"
12632 (previously it was just the string "/dev/urandom"), so on typical
12633 platforms the 10 ms delay will never occur.
12634 Also separate out the Unix variant to its own file, rand_unix.c.
12635 For VMS, there's a currently-empty rand_vms.c.
12639 * Move OCSP client related routines to ocsp_cl.c. These
12640 provide utility functions which an application needing
12641 to issue a request to an OCSP responder and analyse the
12642 response will typically need: as opposed to those which an
12643 OCSP responder itself would need which will be added later.
12645 OCSP_request_sign() signs an OCSP request with an API similar
12646 to PKCS7_sign(). OCSP_response_status() returns status of OCSP
12647 response. OCSP_response_get1_basic() extracts basic response
12648 from response. OCSP_resp_find_status(): finds and extracts status
12649 information from an OCSP_CERTID structure (which will be created
12650 when the request structure is built). These are built from lower
12651 level functions which work on OCSP_SINGLERESP structures but
12652 won't normally be used unless the application wishes to examine
12653 extensions in the OCSP response for example.
12655 Replace nonce routines with a pair of functions.
12656 OCSP_request_add1_nonce() adds a nonce value and optionally
12657 generates a random value. OCSP_check_nonce() checks the
12658 validity of the nonce in an OCSP response.
12662 * Change function OCSP_request_add() to OCSP_request_add0_id().
12663 This doesn't copy the supplied OCSP_CERTID and avoids the
12664 need to free up the newly created id. Change return type
12665 to OCSP_ONEREQ to return the internal OCSP_ONEREQ structure.
12666 This can then be used to add extensions to the request.
12667 Deleted OCSP_request_new(), since most of its functionality
12668 is now in OCSP_REQUEST_new() (and the case insensitive name
12669 clash) apart from the ability to set the request name which
12670 will be added elsewhere.
12674 * Update OCSP API. Remove obsolete extensions argument from
12675 various functions. Extensions are now handled using the new
12676 OCSP extension code. New simple OCSP HTTP function which
12677 can be used to send requests and parse the response.
12681 * Fix the PKCS#7 (S/MIME) code to work with new ASN1. Two new
12682 ASN1_ITEM structures help with sign and verify. PKCS7_ATTR_SIGN
12683 uses the special reorder version of SET OF to sort the attributes
12684 and reorder them to match the encoded order. This resolves a long
12685 standing problem: a verify on a PKCS7 structure just after signing
12686 it used to fail because the attribute order did not match the
12687 encoded order. PKCS7_ATTR_VERIFY does not reorder the attributes:
12688 it uses the received order. This is necessary to tolerate some broken
12689 software that does not order SET OF. This is handled by encoding
12690 as a SEQUENCE OF but using implicit tagging (with UNIVERSAL class)
12691 to produce the required SET OF.
12695 * Have mk1mf.pl generate the macros OPENSSL_BUILD_SHLIBCRYPTO and
12696 OPENSSL_BUILD_SHLIBSSL and use them appropriately in the header
12697 files to get correct declarations of the ASN.1 item variables.
12701 * Rewrite of PKCS#12 code to use new ASN1 functionality. Replace many
12702 PKCS#12 macros with real functions. Fix two unrelated ASN1 bugs:
12703 asn1_check_tlen() would sometimes attempt to use 'ctx' when it was
12704 NULL and ASN1_TYPE was not dereferenced properly in asn1_ex_c2i().
12705 New ASN1 macro: DECLARE_ASN1_ITEM() which just declares the relevant
12706 ASN1_ITEM and no wrapper functions.
12710 * New functions or ASN1_item_d2i_fp() and ASN1_item_d2i_bio(). These
12711 replace the old function pointer based I/O routines. Change most of
12712 the `*_d2i_bio()` and `*_d2i_fp()` functions to use these.
12716 * Enhance mkdef.pl to be more accepting about spacing in C preprocessor
12717 lines, recognize more "algorithms" that can be deselected, and make
12718 it complain about algorithm deselection that isn't recognised.
12722 * New ASN1 functions to handle dup, sign, verify, digest, pack and
12723 unpack operations in terms of ASN1_ITEM. Modify existing wrappers
12724 to use new functions. Add NO_ASN1_OLD which can be set to remove
12725 some old style ASN1 functions: this can be used to determine if old
12726 code will still work when these eventually go away.
12730 * New extension functions for OCSP structures, these follow the
12731 same conventions as certificates and CRLs.
12735 * New function X509V3_add1_i2d(). This automatically encodes and
12736 adds an extension. Its behaviour can be customised with various
12737 flags to append, replace or delete. Various wrappers added for
12738 certificates and CRLs.
12742 * Fix to avoid calling the underlying ASN1 print routine when
12743 an extension cannot be parsed. Correct a typo in the
12744 OCSP_SERVICELOC extension. Tidy up print OCSP format.
12748 * Make mkdef.pl parse some of the ASN1 macros and add appropriate
12749 entries for variables.
12753 * Add functionality to apps/openssl.c for detecting locking
12754 problems: As the program is single-threaded, all we have
12755 to do is register a locking callback using an array for
12756 storing which locks are currently held by the program.
12760 * Use a lock around the call to CRYPTO_get_ex_new_index() in
12761 SSL_get_ex_data_X509_STORE_idx(), which is used in
12762 ssl_verify_cert_chain() and thus can be called at any time
12763 during TLS/SSL handshakes so that thread-safety is essential.
12764 Unfortunately, the ex_data design is not at all suited
12765 for multi-threaded use, so it probably should be abolished.
12769 * Added Broadcom "ubsec" ENGINE to OpenSSL.
12771 *Broadcom, tweaked and integrated by Geoff Thorpe*
12773 * Move common extension printing code to new function
12774 X509V3_print_extensions(). Reorganise OCSP print routines and
12775 implement some needed OCSP ASN1 functions. Add OCSP extensions.
12779 * New function X509_signature_print() to remove duplication in some
12784 * Add a special meaning when SET OF and SEQUENCE OF flags are both
12785 set (this was treated exactly the same as SET OF previously). This
12786 is used to reorder the STACK representing the structure to match the
12787 encoding. This will be used to get round a problem where a PKCS7
12788 structure which was signed could not be verified because the STACK
12789 order did not reflect the encoded order.
12793 * Reimplement the OCSP ASN1 module using the new code.
12797 * Update the X509V3 code to permit the use of an ASN1_ITEM structure
12798 for its ASN1 operations. The old style function pointers still exist
12799 for now but they will eventually go away.
12803 * Merge in replacement ASN1 code from the ASN1 branch. This almost
12804 completely replaces the old ASN1 functionality with a table driven
12805 encoder and decoder which interprets an ASN1_ITEM structure describing
12806 the ASN1 module. Compatibility with the existing ASN1 API (i2d,d2i) is
12807 largely maintained. Almost all of the old asn1_mac.h macro based ASN1
12808 has also been converted to the new form.
12812 * Change BN_mod_exp_recp so that negative moduli are tolerated
12813 (the sign is ignored). Similarly, ignore the sign in BN_MONT_CTX_set
12814 so that BN_mod_exp_mont and BN_mod_exp_mont_word work
12815 for negative moduli.
12819 * Fix BN_uadd and BN_usub: Always return non-negative results instead
12820 of not touching the result's sign bit.
12824 * BN_div bugfix: If the result is 0, the sign (res->neg) must not be
12829 * Changed the LHASH code to use prototypes for callbacks, and created
12830 macros to declare and implement thin (optionally static) functions
12831 that provide type-safety and avoid function pointer casting for the
12832 type-specific callbacks.
12836 * Added Kerberos Cipher Suites to be used with TLS, as written in
12838 *Veers Staats <staatsvr@asc.hpc.mil>,
12839 Jeffrey Altman <jaltman@columbia.edu>, via Richard Levitte*
12841 * Reformat the FAQ so the different questions and answers can be divided
12842 in sections depending on the subject.
12846 * Have the zlib compression code load ZLIB.DLL dynamically under
12851 * New function BN_mod_sqrt for computing square roots modulo a prime
12852 (using the probabilistic Tonelli-Shanks algorithm unless
12853 p == 3 (mod 4) or p == 5 (mod 8), which are cases that can
12854 be handled deterministically).
12856 *Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller*
12858 * Make BN_mod_inverse faster by explicitly handling small quotients
12859 in the Euclid loop. (Speed gain about 20% for small moduli [256 or
12860 512 bits], about 30% for larger ones [1024 or 2048 bits].)
12864 * New function BN_kronecker.
12868 * Fix BN_gcd so that it works on negative inputs; the result is
12869 positive unless both parameters are zero.
12870 Previously something reasonably close to an infinite loop was
12871 possible because numbers could be growing instead of shrinking
12872 in the implementation of Euclid's algorithm.
12876 * Fix BN_is_word() and BN_is_one() macros to take into account the
12877 sign of the number in question.
12879 Fix BN_is_word(a,w) to work correctly for w == 0.
12881 The old BN_is_word(a,w) macro is now called BN_abs_is_word(a,w)
12882 because its test if the absolute value of 'a' equals 'w'.
12883 Note that BN_abs_is_word does *not* handle w == 0 reliably;
12884 it exists mostly for use in the implementations of BN_is_zero(),
12885 BN_is_one(), and BN_is_word().
12889 * New function BN_swap.
12893 * Use BN_nnmod instead of BN_mod in crypto/bn/bn_exp.c so that
12894 the exponentiation functions are more likely to produce reasonable
12895 results on negative inputs.
12899 * Change BN_mod_mul so that the result is always non-negative.
12900 Previously, it could be negative if one of the factors was negative;
12901 I don't think anyone really wanted that behaviour.
12905 * Move `BN_mod_...` functions into new file crypto/bn/bn_mod.c
12906 (except for exponentiation, which stays in crypto/bn/bn_exp.c,
12907 and BN_mod_mul_reciprocal, which stays in crypto/bn/bn_recp.c)
12908 and add new functions:
12917 BN_mod_lshift1_quick
12919 BN_mod_lshift_quick
12921 These functions always generate non-negative results.
12923 BN_nnmod otherwise is like BN_mod (if BN_mod computes a remainder r
12924 such that |m| < r < 0, BN_nnmod will output rem + |m| instead).
12926 BN_mod_XXX_quick(r, a, [b,] m) generates the same result as
12927 BN_mod_XXX(r, a, [b,] m, ctx), but requires that a [and b]
12928 be reduced modulo m.
12930 *Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller*
12933 The following entry accidentally appeared in the CHANGES file
12934 distributed with OpenSSL 0.9.7. The modifications described in
12935 it do *not* apply to OpenSSL 0.9.7.
12937 * Remove a few calls to bn_wexpand() in BN_sqr() (the one in there
12938 was actually never needed) and in BN_mul(). The removal in BN_mul()
12939 required a small change in bn_mul_part_recursive() and the addition
12940 of the functions bn_cmp_part_words(), bn_sub_part_words() and
12941 bn_add_part_words(), which do the same thing as bn_cmp_words(),
12942 bn_sub_words() and bn_add_words() except they take arrays with
12948 * In 'openssl passwd', verify passwords read from the terminal
12949 unless the '-salt' option is used (which usually means that
12950 verification would just waste user's time since the resulting
12951 hash is going to be compared with some given password hash)
12952 or the new '-noverify' option is used.
12954 This is an incompatible change, but it does not affect
12955 non-interactive use of 'openssl passwd' (passwords on the command
12956 line, '-stdin' option, '-in ...' option) and thus should not
12957 cause any problems.
12961 * Remove all references to RSAref, since there's no more need for it.
12965 * Make DSO load along a path given through an environment variable
12966 (SHLIB_PATH) with shl_load().
12970 * Constify the ENGINE code as a result of BIGNUM constification.
12971 Also constify the RSA code and most things related to it. In a
12972 few places, most notable in the depth of the ASN.1 code, ugly
12973 casts back to non-const were required (to be solved at a later
12978 * Make it so the openssl application has all engines loaded by default.
12982 * Constify the BIGNUM routines a little more.
12986 * Add the following functions:
12988 ENGINE_load_cswift()
12990 ENGINE_load_atalla()
12991 ENGINE_load_nuron()
12992 ENGINE_load_builtin_engines()
12994 That way, an application can itself choose if external engines that
12995 are built-in in OpenSSL shall ever be used or not. The benefit is
12996 that applications won't have to be linked with libdl or other dso
12997 libraries unless it's really needed.
12999 Changed 'openssl engine' to load all engines on demand.
13000 Changed the engine header files to avoid the duplication of some
13001 declarations (they differed!).
13005 * 'openssl engine' can now list capabilities.
13009 * Better error reporting in 'openssl engine'.
13013 * Never call load_dh_param(NULL) in s_server.
13017 * Add engine application. It can currently list engines by name and
13018 identity, and test if they are actually available.
13022 * Improve RPM specification file by forcing symbolic linking and making
13023 sure the installed documentation is also owned by root.root.
13025 *Damien Miller <djm@mindrot.org>*
13027 * Give the OpenSSL applications more possibilities to make use of
13028 keys (public as well as private) handled by engines.
13032 * Add OCSP code that comes from CertCo.
13036 * Add VMS support for the Rijndael code.
13040 * Added untested support for Nuron crypto accelerator.
13044 * Add support for external cryptographic devices. This code was
13045 previously distributed separately as the "engine" branch.
13047 *Geoff Thorpe, Richard Levitte*
13049 * Rework the filename-translation in the DSO code. It is now possible to
13050 have far greater control over how a "name" is turned into a filename
13051 depending on the operating environment and any oddities about the
13052 different shared library filenames on each system.
13056 * Support threads on FreeBSD-elf in Configure.
13060 * Fix for SHA1 assembly problem with MASM: it produces
13061 warnings about corrupt line number information when assembling
13062 with debugging information. This is caused by the overlapping
13065 *Bernd Matthes <mainbug@celocom.de>, Steve Henson*
13068 NCONF_get_number() has no error checking at all. As a replacement,
13069 NCONF_get_number_e() is defined (`_e` for "error checking") and is
13070 promoted strongly. The old NCONF_get_number is kept around for
13071 binary backward compatibility.
13072 Make it possible for methods to load from something other than a BIO,
13073 by providing a function pointer that is given a name instead of a BIO.
13074 For example, this could be used to load configuration data from an
13079 * Fix for non blocking accept BIOs. Added new I/O special reason
13080 BIO_RR_ACCEPT to cover this case. Previously use of accept BIOs
13081 with non blocking I/O was not possible because no retry code was
13082 implemented. Also added new SSL code SSL_WANT_ACCEPT to cover
13087 * Added the beginnings of Rijndael support.
13091 * Fix for bug in DirectoryString mask setting. Add support for
13092 X509_NAME_print_ex() in 'req' and X509_print_ex() function
13093 to allow certificate printing to more controllable, additional
13094 'certopt' option to 'x509' to allow new printing options to be
13099 * Clean old EAY MD5 hack from e_os.h.
13103 ### Changes between 0.9.6l and 0.9.6m [17 Mar 2004]
13105 * Fix null-pointer assignment in do_change_cipher_spec() revealed
13106 by using the Codenomicon TLS Test Tool [CVE-2004-0079][]
13108 *Joe Orton, Steve Henson*
13110 ### Changes between 0.9.6k and 0.9.6l [04 Nov 2003]
13112 * Fix additional bug revealed by the NISCC test suite:
13114 Stop bug triggering large recursion when presented with
13115 certain ASN.1 tags [CVE-2003-0851][]
13119 ### Changes between 0.9.6j and 0.9.6k [30 Sep 2003]
13121 * Fix various bugs revealed by running the NISCC test suite:
13123 Stop out of bounds reads in the ASN1 code when presented with
13124 invalid tags (CVE-2003-0543 and CVE-2003-0544).
13126 If verify callback ignores invalid public key errors don't try to check
13127 certificate signature with the NULL public key.
13131 * In ssl3_accept() (ssl/s3_srvr.c) only accept a client certificate
13132 if the server requested one: as stated in TLS 1.0 and SSL 3.0
13137 * In ssl3_get_client_hello() (ssl/s3_srvr.c), tolerate additional
13138 extra data after the compression methods not only for TLS 1.0
13139 but also for SSL 3.0 (as required by the specification).
13141 *Bodo Moeller; problem pointed out by Matthias Loepfe*
13143 * Change X509_certificate_type() to mark the key as exported/exportable
13144 when it's 512 *bits* long, not 512 bytes.
13148 ### Changes between 0.9.6i and 0.9.6j [10 Apr 2003]
13150 * Countermeasure against the Klima-Pokorny-Rosa extension of
13151 Bleichbacher's attack on PKCS #1 v1.5 padding: treat
13152 a protocol version number mismatch like a decryption error
13153 in ssl3_get_client_key_exchange (ssl/s3_srvr.c).
13157 * Turn on RSA blinding by default in the default implementation
13158 to avoid a timing attack. Applications that don't want it can call
13159 RSA_blinding_off() or use the new flag RSA_FLAG_NO_BLINDING.
13160 They would be ill-advised to do so in most cases.
13162 *Ben Laurie, Steve Henson, Geoff Thorpe, Bodo Moeller*
13164 * Change RSA blinding code so that it works when the PRNG is not
13165 seeded (in this case, the secret RSA exponent is abused as
13166 an unpredictable seed -- if it is not unpredictable, there
13167 is no point in blinding anyway). Make RSA blinding thread-safe
13168 by remembering the creator's thread ID in rsa->blinding and
13169 having all other threads use local one-time blinding factors
13170 (this requires more computation than sharing rsa->blinding, but
13171 avoids excessive locking; and if an RSA object is not shared
13172 between threads, blinding will still be very fast).
13176 ### Changes between 0.9.6h and 0.9.6i [19 Feb 2003]
13178 * In ssl3_get_record (ssl/s3_pkt.c), minimize information leaked
13179 via timing by performing a MAC computation even if incorrect
13180 block cipher padding has been found. This is a countermeasure
13181 against active attacks where the attacker has to distinguish
13182 between bad padding and a MAC verification error. [CVE-2003-0078][]
13184 *Bodo Moeller; problem pointed out by Brice Canvel (EPFL),
13185 Alain Hiltgen (UBS), Serge Vaudenay (EPFL), and
13186 Martin Vuagnoux (EPFL, Ilion)*
13188 ### Changes between 0.9.6g and 0.9.6h [5 Dec 2002]
13190 * New function OPENSSL_cleanse(), which is used to cleanse a section of
13191 memory from its contents. This is done with a counter that will
13192 place alternating values in each byte. This can be used to solve
13193 two issues: 1) the removal of calls to memset() by highly optimizing
13194 compilers, and 2) cleansing with other values than 0, since those can
13195 be read through on certain media, for example a swap space on disk.
13199 * Bugfix: client side session caching did not work with external caching,
13200 because the session->cipher setting was not restored when reloading
13201 from the external cache. This problem was masked, when
13202 SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG (part of SSL_OP_ALL) was set.
13203 (Found by Steve Haslam <steve@araqnid.ddts.net>.)
13207 * Fix client_certificate (ssl/s2_clnt.c): The permissible total
13208 length of the REQUEST-CERTIFICATE message is 18 .. 34, not 17 .. 33.
13210 *Zeev Lieber <zeev-l@yahoo.com>*
13212 * Undo an undocumented change introduced in 0.9.6e which caused
13213 repeated calls to OpenSSL_add_all_ciphers() and
13214 OpenSSL_add_all_digests() to be ignored, even after calling
13219 * Change the default configuration reader to deal with last line not
13220 being properly terminated.
13224 * Change X509_NAME_cmp() so it applies the special rules on handling
13225 DN values that are of type PrintableString, as well as RDNs of type
13226 emailAddress where the value has the type ia5String.
13228 *stefank@valicert.com via Richard Levitte*
13230 * Add a SSL_SESS_CACHE_NO_INTERNAL_STORE flag to take over half
13231 the job SSL_SESS_CACHE_NO_INTERNAL_LOOKUP was inconsistently
13232 doing, define a new flag (SSL_SESS_CACHE_NO_INTERNAL) to be
13233 the bitwise-OR of the two for use by the majority of applications
13234 wanting this behaviour, and update the docs. The documented
13235 behaviour and actual behaviour were inconsistent and had been
13236 changing anyway, so this is more a bug-fix than a behavioural
13239 *Geoff Thorpe, diagnosed by Nadav Har'El*
13241 * Don't impose a 16-byte length minimum on session IDs in ssl/s3_clnt.c
13242 (the SSL 3.0 and TLS 1.0 specifications allow any length up to 32 bytes).
13246 * Fix initialization code race conditions in
13247 SSLv23_method(), SSLv23_client_method(), SSLv23_server_method(),
13248 SSLv2_method(), SSLv2_client_method(), SSLv2_server_method(),
13249 SSLv3_method(), SSLv3_client_method(), SSLv3_server_method(),
13250 TLSv1_method(), TLSv1_client_method(), TLSv1_server_method(),
13251 ssl2_get_cipher_by_char(),
13252 ssl3_get_cipher_by_char().
13254 *Patrick McCormick <patrick@tellme.com>, Bodo Moeller*
13256 * Reorder cleanup sequence in SSL_CTX_free(): only remove the ex_data after
13257 the cached sessions are flushed, as the remove_cb() might use ex_data
13258 contents. Bug found by Sam Varshavchik <mrsam@courier-mta.com>
13259 (see [openssl.org #212]).
13261 *Geoff Thorpe, Lutz Jaenicke*
13263 * Fix typo in OBJ_txt2obj which incorrectly passed the content
13264 length, instead of the encoding length to d2i_ASN1_OBJECT.
13268 ### Changes between 0.9.6f and 0.9.6g [9 Aug 2002]
13270 * [In 0.9.6g-engine release:]
13271 Fix crypto/engine/vendor_defns/cswift.h for WIN32 (use `_stdcall`).
13273 *Lynn Gazis <lgazis@rainbow.com>*
13275 ### Changes between 0.9.6e and 0.9.6f [8 Aug 2002]
13277 * Fix ASN1 checks. Check for overflow by comparing with LONG_MAX
13278 and get fix the header length calculation.
13279 *Florian Weimer <Weimer@CERT.Uni-Stuttgart.DE>,
13280 Alon Kantor <alonk@checkpoint.com> (and others), Steve Henson*
13282 * Use proper error handling instead of 'assertions' in buffer
13283 overflow checks added in 0.9.6e. This prevents DoS (the
13284 assertions could call abort()).
13286 *Arne Ansper <arne@ats.cyber.ee>, Bodo Moeller*
13288 ### Changes between 0.9.6d and 0.9.6e [30 Jul 2002]
13290 * Add various sanity checks to asn1_get_length() to reject
13291 the ASN1 length bytes if they exceed sizeof(long), will appear
13292 negative or the content length exceeds the length of the
13295 *Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>*
13297 * Fix cipher selection routines: ciphers without encryption had no flags
13298 for the cipher strength set and where therefore not handled correctly
13299 by the selection routines (PR #130).
13303 * Fix EVP_dsa_sha macro.
13308 SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS
13309 for disabling the SSL 3.0/TLS 1.0 CBC vulnerability countermeasure
13310 that was added in OpenSSL 0.9.6d.
13312 As the countermeasure turned out to be incompatible with some
13313 broken SSL implementations, the new option is part of SSL_OP_ALL.
13314 SSL_OP_ALL is usually employed when compatibility with weird SSL
13315 implementations is desired (e.g. '-bugs' option to 's_client' and
13316 's_server'), so the new option is automatically set in many
13321 * Changes in security patch:
13323 Changes marked "(CHATS)" were sponsored by the Defense Advanced
13324 Research Projects Agency (DARPA) and Air Force Research Laboratory,
13325 Air Force Materiel Command, USAF, under agreement number
13328 * Add various sanity checks to asn1_get_length() to reject
13329 the ASN1 length bytes if they exceed sizeof(long), will appear
13330 negative or the content length exceeds the length of the
13331 supplied buffer. [CVE-2002-0659][]
13333 *Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>*
13335 * Assertions for various potential buffer overflows, not known to
13336 happen in practice.
13338 *Ben Laurie (CHATS)*
13340 * Various temporary buffers to hold ASCII versions of integers were
13341 too small for 64 bit platforms. [CVE-2002-0655][]
13342 *Matthew Byng-Maddick <mbm@aldigital.co.uk> and Ben Laurie (CHATS)>*
13344 * Remote buffer overflow in SSL3 protocol - an attacker could
13345 supply an oversized session ID to a client. [CVE-2002-0656][]
13347 *Ben Laurie (CHATS)*
13349 * Remote buffer overflow in SSL2 protocol - an attacker could
13350 supply an oversized client master key. [CVE-2002-0656][]
13352 *Ben Laurie (CHATS)*
13354 ### Changes between 0.9.6c and 0.9.6d [9 May 2002]
13356 * Fix crypto/asn1/a_sign.c so that 'parameters' is omitted (not
13357 encoded as NULL) with id-dsa-with-sha1.
13359 *Nils Larsch <nla@trustcenter.de>; problem pointed out by Bodo Moeller*
13361 * Check various `X509_...()` return values in apps/req.c.
13363 *Nils Larsch <nla@trustcenter.de>*
13365 * Fix BASE64 decode (EVP_DecodeUpdate) for data with CR/LF ended lines:
13366 an end-of-file condition would erroneously be flagged, when the CRLF
13367 was just at the end of a processed block. The bug was discovered when
13368 processing data through a buffering memory BIO handing the data to a
13369 BASE64-decoding BIO. Bug fund and patch submitted by Pavel Tsekov
13370 <ptsekov@syntrex.com> and Nedelcho Stanev.
13374 * Implement a countermeasure against a vulnerability recently found
13375 in CBC ciphersuites in SSL 3.0/TLS 1.0: Send an empty fragment
13376 before application data chunks to avoid the use of known IVs
13377 with data potentially chosen by the attacker.
13381 * Fix length checks in ssl3_get_client_hello().
13385 * TLS/SSL library bugfix: use s->s3->in_read_app_data differently
13386 to prevent ssl3_read_internal() from incorrectly assuming that
13387 ssl3_read_bytes() found application data while handshake
13388 processing was enabled when in fact s->s3->in_read_app_data was
13389 merely automatically cleared during the initial handshake.
13391 *Bodo Moeller; problem pointed out by Arne Ansper <arne@ats.cyber.ee>*
13393 * Fix object definitions for Private and Enterprise: they were not
13394 recognized in their shortname (=lowercase) representation. Extend
13395 obj_dat.pl to issue an error when using undefined keywords instead
13396 of silently ignoring the problem (Svenning Sorensen
13397 <sss@sss.dnsalias.net>).
13401 * Fix DH_generate_parameters() so that it works for 'non-standard'
13402 generators, i.e. generators other than 2 and 5. (Previously, the
13403 code did not properly initialise the 'add' and 'rem' values to
13404 BN_generate_prime().)
13406 In the new general case, we do not insist that 'generator' is
13407 actually a primitive root: This requirement is rather pointless;
13408 a generator of the order-q subgroup is just as good, if not
13413 * Map new X509 verification errors to alerts. Discovered and submitted by
13414 Tom Wu <tom@arcot.com>.
13418 * Fix ssl3_pending() (ssl/s3_lib.c) to prevent SSL_pending() from
13419 returning non-zero before the data has been completely received
13420 when using non-blocking I/O.
13422 *Bodo Moeller; problem pointed out by John Hughes*
13424 * Some of the ciphers missed the strength entry (SSL_LOW etc).
13426 *Ben Laurie, Lutz Jaenicke*
13428 * Fix bug in SSL_clear(): bad sessions were not removed (found by
13429 Yoram Zahavi <YoramZ@gilian.com>).
13433 * Add information about CygWin 1.3 and on, and preserve proper
13434 configuration for the versions before that.
13436 *Corinna Vinschen <vinschen@redhat.com> and Richard Levitte*
13438 * Make removal from session cache (SSL_CTX_remove_session()) more robust:
13439 check whether we deal with a copy of a session and do not delete from
13440 the cache in this case. Problem reported by "Izhar Shoshani Levi"
13441 <izhar@checkpoint.com>.
13445 * Do not store session data into the internal session cache, if it
13446 is never intended to be looked up (SSL_SESS_CACHE_NO_INTERNAL_LOOKUP
13447 flag is set). Proposed by Aslam <aslam@funk.com>.
13451 * Have ASN1_BIT_STRING_set_bit() really clear a bit when the requested
13456 * [In 0.9.6d-engine release:]
13457 Fix a crashbug and a logic bug in hwcrhk_load_pubkey().
13459 *Toomas Kiisk <vix@cyber.ee> via Richard Levitte*
13461 * Add the configuration target linux-s390x.
13463 *Neale Ferguson <Neale.Ferguson@SoftwareAG-USA.com> via Richard Levitte*
13465 * The earlier bugfix for the SSL3_ST_SW_HELLO_REQ_C case of
13466 ssl3_accept (ssl/s3_srvr.c) incorrectly used a local flag
13467 variable as an indication that a ClientHello message has been
13468 received. As the flag value will be lost between multiple
13469 invocations of ssl3_accept when using non-blocking I/O, the
13470 function may not be aware that a handshake has actually taken
13471 place, thus preventing a new session from being added to the
13474 To avoid this problem, we now set s->new_session to 2 instead of
13475 using a local variable.
13477 *Lutz Jaenicke, Bodo Moeller*
13479 * Bugfix: Return -1 from ssl3_get_server_done (ssl3/s3_clnt.c)
13480 if the SSL_R_LENGTH_MISMATCH error is detected.
13482 *Geoff Thorpe, Bodo Moeller*
13484 * New 'shared_ldflag' column in Configure platform table.
13488 * Fix EVP_CIPHER_mode macro.
13490 *"Dan S. Camper" <dan@bti.net>*
13492 * Fix ssl3_read_bytes (ssl/s3_pkt.c): To ignore messages of unknown
13493 type, we must throw them away by setting rr->length to 0.
13495 *D P Chang <dpc@qualys.com>*
13497 ### Changes between 0.9.6b and 0.9.6c [21 dec 2001]
13499 * Fix BN_rand_range bug pointed out by Dominikus Scherkl
13500 <Dominikus.Scherkl@biodata.com>. (The previous implementation
13501 worked incorrectly for those cases where range = `10..._2` and
13502 `3*range` is two bits longer than range.)
13506 * Only add signing time to PKCS7 structures if it is not already
13511 * Fix crypto/objects/objects.h: "ld-ce" should be "id-ce",
13512 OBJ_ld_ce should be OBJ_id_ce.
13513 Also some ip-pda OIDs in crypto/objects/objects.txt were
13514 incorrect (cf. RFC 3039).
13516 *Matt Cooper, Frederic Giudicelli, Bodo Moeller*
13518 * Release CRYPTO_LOCK_DYNLOCK when CRYPTO_destroy_dynlockid()
13519 returns early because it has nothing to do.
13521 *Andy Schneider <andy.schneider@bjss.co.uk>*
13523 * [In 0.9.6c-engine release:]
13524 Fix mutex callback return values in crypto/engine/hw_ncipher.c.
13526 *Andy Schneider <andy.schneider@bjss.co.uk>*
13528 * [In 0.9.6c-engine release:]
13529 Add support for Cryptographic Appliance's keyserver technology.
13530 (Use engine 'keyclient')
13532 *Cryptographic Appliances and Geoff Thorpe*
13534 * Add a configuration entry for OS/390 Unix. The C compiler 'c89'
13535 is called via tools/c89.sh because arguments have to be
13536 rearranged (all '-L' options must appear before the first object
13539 *Richard Shapiro <rshapiro@abinitio.com>*
13541 * [In 0.9.6c-engine release:]
13542 Add support for Broadcom crypto accelerator cards, backported
13545 *Broadcom, Nalin Dahyabhai <nalin@redhat.com>, Mark Cox*
13547 * [In 0.9.6c-engine release:]
13548 Add support for SureWare crypto accelerator cards from
13549 Baltimore Technologies. (Use engine 'sureware')
13551 *Baltimore Technologies and Mark Cox*
13553 * [In 0.9.6c-engine release:]
13554 Add support for crypto accelerator cards from Accelerated
13555 Encryption Processing, www.aep.ie. (Use engine 'aep')
13557 *AEP Inc. and Mark Cox*
13559 * Add a configuration entry for gcc on UnixWare.
13561 *Gary Benson <gbenson@redhat.com>*
13563 * Change ssl/s2_clnt.c and ssl/s2_srvr.c so that received handshake
13564 messages are stored in a single piece (fixed-length part and
13565 variable-length part combined) and fix various bugs found on the way.
13569 * Disable caching in BIO_gethostbyname(), directly use gethostbyname()
13570 instead. BIO_gethostbyname() does not know what timeouts are
13571 appropriate, so entries would stay in cache even when they have
13573 *Bodo Moeller; problem pointed out by Rich Salz <rsalz@zolera.com>*
13575 * Change ssl23_get_client_hello (ssl/s23_srvr.c) behaviour when
13576 faced with a pathologically small ClientHello fragment that does
13577 not contain client_version: Instead of aborting with an error,
13578 simply choose the highest available protocol version (i.e.,
13579 TLS 1.0 unless it is disabled). In practice, ClientHello
13580 messages are never sent like this, but this change gives us
13581 strictly correct behaviour at least for TLS.
13585 * Fix SSL handshake functions and SSL_clear() such that SSL_clear()
13586 never resets s->method to s->ctx->method when called from within
13587 one of the SSL handshake functions.
13589 *Bodo Moeller; problem pointed out by Niko Baric*
13591 * In ssl3_get_client_hello (ssl/s3_srvr.c), generate a fatal alert
13592 (sent using the client's version number) if client_version is
13593 smaller than the protocol version in use. Also change
13594 ssl23_get_client_hello (ssl/s23_srvr.c) to select TLS 1.0 if
13595 the client demanded SSL 3.0 but only TLS 1.0 is enabled; then
13596 the client will at least see that alert.
13600 * Fix ssl3_get_message (ssl/s3_both.c) to handle message fragmentation
13605 * Avoid infinite loop in ssl3_get_message (ssl/s3_both.c) if a
13606 client receives HelloRequest while in a handshake.
13608 *Bodo Moeller; bug noticed by Andy Schneider <andy.schneider@bjss.co.uk>*
13610 * Bugfix in ssl3_accept (ssl/s3_srvr.c): Case SSL3_ST_SW_HELLO_REQ_C
13611 should end in 'break', not 'goto end' which circumvents various
13612 cleanups done in state SSL_ST_OK. But session related stuff
13613 must be disabled for SSL_ST_OK in the case that we just sent a
13616 Also avoid some overhead by not calling ssl_init_wbio_buffer()
13617 before just sending a HelloRequest.
13619 *Bodo Moeller, Eric Rescorla <ekr@rtfm.com>*
13621 * Fix ssl/s3_enc.c, ssl/t1_enc.c and ssl/s3_pkt.c so that we don't
13622 reveal whether illegal block cipher padding was found or a MAC
13623 verification error occurred. (Neither SSLerr() codes nor alerts
13624 are directly visible to potential attackers, but the information
13625 may leak via logfiles.)
13627 Similar changes are not required for the SSL 2.0 implementation
13628 because the number of padding bytes is sent in clear for SSL 2.0,
13629 and the extra bytes are just ignored. However ssl/s2_pkt.c
13630 failed to verify that the purported number of padding bytes is in
13635 * Add OpenUNIX-8 support including shared libraries
13636 (Boyd Lynn Gerber <gerberb@zenez.com>).
13640 * Improve RSA_padding_check_PKCS1_OAEP() check again to avoid
13641 'wristwatch attack' using huge encoding parameters (cf.
13642 James H. Manger's CRYPTO 2001 paper). Note that the
13643 RSA_PKCS1_OAEP_PADDING case of RSA_private_decrypt() does not use
13644 encoding parameters and hence was not vulnerable.
13648 * BN_sqr() bug fix.
13650 *Ulf Möller, reported by Jim Ellis <jim.ellis@cavium.com>*
13652 * Rabin-Miller test analyses assume uniformly distributed witnesses,
13653 so use BN_pseudo_rand_range() instead of using BN_pseudo_rand()
13654 followed by modular reduction.
13656 *Bodo Moeller; pointed out by Adam Young <AYoung1@NCSUS.JNJ.COM>*
13658 * Add BN_pseudo_rand_range() with obvious functionality: BN_rand_range()
13659 equivalent based on BN_pseudo_rand() instead of BN_rand().
13663 * s3_srvr.c: allow sending of large client certificate lists (> 16 kB).
13664 This function was broken, as the check for a new client hello message
13665 to handle SGC did not allow these large messages.
13666 (Tracked down by "Douglas E. Engert" <deengert@anl.gov>.)
13670 * Add alert descriptions for TLSv1 to `SSL_alert_desc_string[_long]()`.
13674 * Fix buggy behaviour of BIO_get_num_renegotiates() and BIO_ctrl()
13675 for BIO_C_GET_WRITE_BUF_SIZE ("Stephen Hinton" <shinton@netopia.com>).
13679 * Rework the configuration and shared library support for Tru64 Unix.
13680 The configuration part makes use of modern compiler features and
13681 still retains old compiler behavior for those that run older versions
13682 of the OS. The shared library support part includes a variant that
13683 uses the RPATH feature, and is available through the special
13684 configuration target "alpha-cc-rpath", which will never be selected
13687 *Tim Mooney <mooney@dogbert.cc.ndsu.NoDak.edu> via Richard Levitte*
13689 * In ssl3_get_key_exchange (ssl/s3_clnt.c), call ssl3_get_message()
13690 with the same message size as in ssl3_get_certificate_request().
13691 Otherwise, if no ServerKeyExchange message occurs, CertificateRequest
13692 messages might inadvertently be reject as too long.
13694 *Petr Lampa <lampa@fee.vutbr.cz>*
13696 * Enhanced support for IA-64 Unix platforms (well, Linux and HP-UX).
13700 * Modified SSL library such that the verify_callback that has been set
13701 specifically for an SSL object with SSL_set_verify() is actually being
13702 used. Before the change, a verify_callback set with this function was
13703 ignored and the verify_callback() set in the SSL_CTX at the time of
13704 the call was used. New function X509_STORE_CTX_set_verify_cb() introduced
13705 to allow the necessary settings.
13709 * Initialize static variable in crypto/dsa/dsa_lib.c and crypto/dh/dh_lib.c
13710 explicitly to NULL, as at least on Solaris 8 this seems not always to be
13711 done automatically (in contradiction to the requirements of the C
13712 standard). This made problems when used from OpenSSH.
13716 * In OpenSSL 0.9.6a and 0.9.6b, crypto/dh/dh_key.c ignored
13717 dh->length and always used
13719 BN_rand_range(priv_key, dh->p).
13721 BN_rand_range() is not necessary for Diffie-Hellman, and this
13722 specific range makes Diffie-Hellman unnecessarily inefficient if
13723 dh->length (recommended exponent length) is much smaller than the
13724 length of dh->p. We could use BN_rand_range() if the order of
13725 the subgroup was stored in the DH structure, but we only have
13730 BN_rand(priv_key, l, ...)
13732 where 'l' is dh->length if this is defined, or BN_num_bits(dh->p)-1
13739 RSA_eay_public_encrypt
13740 RSA_eay_private_decrypt
13741 RSA_eay_private_encrypt (signing)
13742 RSA_eay_public_decrypt (signature verification)
13744 (default implementations for RSA_public_encrypt,
13745 RSA_private_decrypt, RSA_private_encrypt, RSA_public_decrypt),
13746 always reject numbers >= n.
13750 * In crypto/rand/md_rand.c, use a new short-time lock CRYPTO_LOCK_RAND2
13751 to synchronize access to 'locking_thread'. This is necessary on
13752 systems where access to 'locking_thread' (an 'unsigned long'
13753 variable) is not atomic.
13757 * In crypto/rand/md_rand.c, set 'locking_thread' to current thread's ID
13758 *before* setting the 'crypto_lock_rand' flag. The previous code had
13759 a race condition if 0 is a valid thread ID.
13761 *Travis Vitek <vitek@roguewave.com>*
13763 * Add support for shared libraries under Irix.
13765 *Albert Chin-A-Young <china@thewrittenword.com>*
13767 * Add configuration option to build on Linux on both big-endian and
13768 little-endian MIPS.
13770 *Ralf Baechle <ralf@uni-koblenz.de>*
13772 * Add the possibility to create shared libraries on HP-UX.
13776 ### Changes between 0.9.6a and 0.9.6b [9 Jul 2001]
13778 * Change ssleay_rand_bytes (crypto/rand/md_rand.c)
13779 to avoid a SSLeay/OpenSSL PRNG weakness pointed out by
13780 Markku-Juhani O. Saarinen <markku-juhani.saarinen@nokia.com>:
13781 PRNG state recovery was possible based on the output of
13782 one PRNG request appropriately sized to gain knowledge on
13783 'md' followed by enough consecutive 1-byte PRNG requests
13784 to traverse all of 'state'.
13786 1. When updating 'md_local' (the current thread's copy of 'md')
13787 during PRNG output generation, hash all of the previous
13788 'md_local' value, not just the half used for PRNG output.
13790 2. Make the number of bytes from 'state' included into the hash
13791 independent from the number of PRNG bytes requested.
13793 The first measure alone would be sufficient to avoid
13794 Markku-Juhani's attack. (Actually it had never occurred
13795 to me that the half of 'md_local' used for chaining was the
13796 half from which PRNG output bytes were taken -- I had always
13797 assumed that the secret half would be used.) The second
13798 measure makes sure that additional data from 'state' is never
13799 mixed into 'md_local' in small portions; this heuristically
13800 further strengthens the PRNG.
13804 * Fix crypto/bn/asm/mips3.s.
13808 * When only the key is given to "enc", the IV is undefined. Print out
13809 an error message in this case.
13813 * Handle special case when X509_NAME is empty in X509 printing routines.
13817 * In dsa_do_verify (crypto/dsa/dsa_ossl.c), verify that r and s are
13818 positive and less than q.
13822 * Don't change `*pointer` in CRYPTO_add_lock() is add_lock_callback is
13823 used: it isn't thread safe and the add_lock_callback should handle
13826 *Paul Rose <Paul.Rose@bridge.com>*
13828 * Verify that incoming data obeys the block size in
13829 ssl3_enc (ssl/s3_enc.c) and tls1_enc (ssl/t1_enc.c).
13835 *Ulf Möller, Bodo Möller*
13837 * The countermeasure against Bleichbacher's attack on PKCS #1 v1.5
13838 RSA encryption was accidentally removed in s3_srvr.c in OpenSSL 0.9.5
13839 when fixing the server behaviour for backwards-compatible 'client
13840 hello' messages. (Note that the attack is impractical against
13841 SSL 3.0 and TLS 1.0 anyway because length and version checking
13842 means that the probability of guessing a valid ciphertext is
13843 around 2^-40; see section 5 in Bleichenbacher's CRYPTO '98
13846 Before 0.9.5, the countermeasure (hide the error by generating a
13847 random 'decryption result') did not work properly because
13848 ERR_clear_error() was missing, meaning that SSL_get_error() would
13849 detect the supposedly ignored error.
13851 Both problems are now fixed.
13855 * In crypto/bio/bf_buff.c, increase DEFAULT_BUFFER_SIZE to 4096
13856 (previously it was 1024).
13860 * Fix for compatibility mode trust settings: ignore trust settings
13861 unless some valid trust or reject settings are present.
13865 * Fix for blowfish EVP: its a variable length cipher.
13869 * Fix various bugs related to DSA S/MIME verification. Handle missing
13870 parameters in DSA public key structures and return an error in the
13871 DSA routines if parameters are absent.
13875 * In versions up to 0.9.6, RAND_file_name() resorted to file ".rnd"
13876 in the current directory if neither $RANDFILE nor $HOME was set.
13877 RAND_file_name() in 0.9.6a returned NULL in this case. This has
13878 caused some confusion to Windows users who haven't defined $HOME.
13879 Thus RAND_file_name() is changed again: e_os.h can define a
13880 DEFAULT_HOME, which will be used if $HOME is not set.
13881 For Windows, we use "C:"; on other platforms, we still require
13882 environment variables.
13884 * Move 'if (!initialized) RAND_poll()' into regions protected by
13885 CRYPTO_LOCK_RAND. This is not strictly necessary, but avoids
13886 having multiple threads call RAND_poll() concurrently.
13890 * In crypto/rand/md_rand.c, replace 'add_do_not_lock' flag by a
13891 combination of a flag and a thread ID variable.
13892 Otherwise while one thread is in ssleay_rand_bytes (which sets the
13893 flag), *other* threads can enter ssleay_add_bytes without obeying
13894 the CRYPTO_LOCK_RAND lock (and may even illegally release the lock
13895 that they do not hold after the first thread unsets add_do_not_lock).
13899 * Change bctest again: '-x' expressions are not available in all
13900 versions of 'test'.
13904 ### Changes between 0.9.6 and 0.9.6a [5 Apr 2001]
13906 * Fix a couple of memory leaks in PKCS7_dataDecode()
13908 *Steve Henson, reported by Heyun Zheng <hzheng@atdsprint.com>*
13910 * Change Configure and Makefiles to provide EXE_EXT, which will contain
13911 the default extension for executables, if any. Also, make the perl
13912 scripts that use symlink() to test if it really exists and use "cp"
13913 if it doesn't. All this made OpenSSL compilable and installable in
13918 * Fix for asn1_GetSequence() for indefinite length constructed data.
13919 If SEQUENCE is length is indefinite just set c->slen to the total
13920 amount of data available.
13922 *Steve Henson, reported by shige@FreeBSD.org*
13924 *This change does not apply to 0.9.7.*
13926 * Change bctest to avoid here-documents inside command substitution
13927 (workaround for FreeBSD /bin/sh bug).
13928 For compatibility with Ultrix, avoid shell functions (introduced
13929 in the bctest version that searches along $PATH).
13933 * Rename 'des_encrypt' to 'des_encrypt1'. This avoids the clashes
13934 with des_encrypt() defined on some operating systems, like Solaris
13939 * Check the result of RSA-CRT (see D. Boneh, R. DeMillo, R. Lipton:
13940 On the Importance of Eliminating Errors in Cryptographic
13941 Computations, J. Cryptology 14 (2001) 2, 101-119,
13942 <http://theory.stanford.edu/~dabo/papers/faults.ps.gz>).
13946 * MIPS assembler BIGNUM division bug fix.
13950 * Disabled incorrect Alpha assembler code.
13954 * Fix PKCS#7 decode routines so they correctly update the length
13955 after reading an EOC for the EXPLICIT tag.
13959 *This change does not apply to 0.9.7.*
13961 * Fix bug in PKCS#12 key generation routines. This was triggered
13962 if a 3DES key was generated with a 0 initial byte. Include
13963 PKCS12_BROKEN_KEYGEN compilation option to retain the old
13964 (but broken) behaviour.
13968 * Enhance bctest to search for a working bc along $PATH and print
13971 *Tim Rice <tim@multitalents.net> via Richard Levitte*
13973 * Fix memory leaks in err.c: free err_data string if necessary;
13974 don't write to the wrong index in ERR_set_error_data.
13978 * Implement ssl23_peek (analogous to ssl23_read), which previously
13983 * Replace rdtsc with `_emit` statements for VC++ version 5.
13985 *Jeremy Cooper <jeremy@baymoo.org>*
13987 * Make it possible to reuse SSLv2 sessions.
13991 * In copy_email() check for >= 0 as a return value for
13992 X509_NAME_get_index_by_NID() since 0 is a valid index.
13994 *Steve Henson reported by Massimiliano Pala <madwolf@opensca.org>*
13996 * Avoid coredump with unsupported or invalid public keys by checking if
13997 X509_get_pubkey() fails in PKCS7_verify(). Fix memory leak when
13998 PKCS7_verify() fails with non detached data.
14002 * Don't use getenv in library functions when run as setuid/setgid.
14003 New function OPENSSL_issetugid().
14007 * Avoid false positives in memory leak detection code (crypto/mem_dbg.c)
14008 due to incorrect handling of multi-threading:
14010 1. Fix timing glitch in the MemCheck_off() portion of CRYPTO_mem_ctrl().
14012 2. Fix logical glitch in is_MemCheck_on() aka CRYPTO_is_mem_check_on().
14014 3. Count how many times MemCheck_off() has been called so that
14015 nested use can be treated correctly. This also avoids
14016 inband-signalling in the previous code (which relied on the
14017 assumption that thread ID 0 is impossible).
14021 * Add "-rand" option also to s_client and s_server.
14025 * Fix CPU detection on Irix 6.x.
14026 *Kurt Hockenbury <khockenb@stevens-tech.edu> and
14027 "Bruce W. Forsberg" <bruce.forsberg@baesystems.com>*
14029 * Fix X509_NAME bug which produced incorrect encoding if X509_NAME
14034 *This change does not apply to 0.9.7.*
14036 * Use the cached encoding of an X509_NAME structure rather than
14037 copying it. This is apparently the reason for the libsafe "errors"
14038 but the code is actually correct.
14042 * Add new function BN_rand_range(), and fix DSA_sign_setup() to prevent
14043 Bleichenbacher's DSA attack.
14044 Extend BN_[pseudo_]rand: As before, top=1 forces the highest two bits
14045 to be set and top=0 forces the highest bit to be set; top=-1 is new
14046 and leaves the highest bit random.
14048 *Ulf Moeller, Bodo Moeller*
14050 * In the `NCONF_...`-based implementations for `CONF_...` queries
14051 (crypto/conf/conf_lib.c), if the input LHASH is NULL, avoid using
14052 a temporary CONF structure with the data component set to NULL
14053 (which gives segmentation faults in lh_retrieve).
14054 Instead, use NULL for the CONF pointer in CONF_get_string and
14055 CONF_get_number (which may use environment variables) and directly
14056 return NULL from CONF_get_section.
14060 * Fix potential buffer overrun for EBCDIC.
14064 * Tolerate nonRepudiation as being valid for S/MIME signing and certSign
14065 keyUsage if basicConstraints absent for a CA.
14069 * Make SMIME_write_PKCS7() write mail header values with a format that
14070 is more generally accepted (no spaces before the semicolon), since
14071 some programs can't parse those values properly otherwise. Also make
14072 sure BIO's that break lines after each write do not create invalid
14077 * Make the CRL encoding routines work with empty SEQUENCE OF. The
14078 macros previously used would not encode an empty SEQUENCE OF
14079 and break the signature.
14083 *This change does not apply to 0.9.7.*
14085 * Zero the premaster secret after deriving the master secret in
14090 * Add some EVP_add_digest_alias registrations (as found in
14091 OpenSSL_add_all_digests()) to SSL_library_init()
14092 aka OpenSSL_add_ssl_algorithms(). This provides improved
14093 compatibility with peers using X.509 certificates
14094 with unconventional AlgorithmIdentifier OIDs.
14098 * Fix for Irix with NO_ASM.
14100 *"Bruce W. Forsberg" <bruce.forsberg@baesystems.com>*
14102 * ./config script fixes.
14104 *Ulf Moeller, Richard Levitte*
14106 * Fix 'openssl passwd -1'.
14110 * Change PKCS12_key_gen_asc() so it can cope with non null
14111 terminated strings whose length is passed in the passlen
14112 parameter, for example from PEM callbacks. This was done
14113 by adding an extra length parameter to asc2uni().
14115 *Steve Henson, reported by <oddissey@samsung.co.kr>*
14117 * Fix C code generated by 'openssl dsaparam -C': If a BN_bin2bn
14118 call failed, free the DSA structure.
14122 * Fix to uni2asc() to cope with zero length Unicode strings.
14123 These are present in some PKCS#12 files.
14127 * Increase s2->wbuf allocation by one byte in ssl2_new (ssl/s2_lib.c).
14128 Otherwise do_ssl_write (ssl/s2_pkt.c) will write beyond buffer limits
14129 when writing a 32767 byte record.
14131 *Bodo Moeller; problem reported by Eric Day <eday@concentric.net>*
14133 * In `RSA_eay_public_{en,ed}crypt` and RSA_eay_mod_exp (rsa_eay.c),
14134 obtain lock CRYPTO_LOCK_RSA before setting `rsa->_method_mod_{n,p,q}`.
14136 (RSA objects have a reference count access to which is protected
14137 by CRYPTO_LOCK_RSA [see rsa_lib.c, s3_srvr.c, ssl_cert.c, ssl_rsa.c],
14138 so they are meant to be shared between threads.)
14139 *Bodo Moeller, Geoff Thorpe; original patch submitted by
14140 "Reddie, Steven" <Steven.Reddie@ca.com>*
14142 * Fix a deadlock in CRYPTO_mem_leaks().
14146 * Use better test patterns in bntest.
14150 * rand_win.c fix for Borland C.
14154 * BN_rshift bugfix for n == 0.
14158 * Add a 'bctest' script that checks for some known 'bc' bugs
14159 so that 'make test' does not abort just because 'bc' is broken.
14163 * Store verify_result within SSL_SESSION also for client side to
14164 avoid potential security hole. (Re-used sessions on the client side
14165 always resulted in verify_result==X509_V_OK, not using the original
14166 result of the server certificate verification.)
14170 * Fix ssl3_pending: If the record in s->s3->rrec is not of type
14171 SSL3_RT_APPLICATION_DATA, return 0.
14172 Similarly, change ssl2_pending to return 0 if SSL_in_init(s) is true.
14177 Both ssl2_peek and ssl3_peek, which were totally broken in earlier
14178 releases, have been re-implemented by renaming the previous
14179 implementations of ssl2_read and ssl3_read to ssl2_read_internal
14180 and ssl3_read_internal, respectively, and adding 'peek' parameters
14181 to them. The new ssl[23]_{read,peek} functions are calls to
14182 ssl[23]_read_internal with the 'peek' flag set appropriately.
14183 A 'peek' parameter has also been added to ssl3_read_bytes, which
14184 does the actual work for ssl3_read_internal.
14188 * Initialise "ex_data" member of RSA/DSA/DH structures prior to calling
14189 the method-specific "init()" handler. Also clean up ex_data after
14190 calling the method-specific "finish()" handler. Previously, this was
14191 happening the other way round.
14195 * Increase BN_CTX_NUM (the number of BIGNUMs in a BN_CTX) to 16.
14196 The previous value, 12, was not always sufficient for BN_mod_exp().
14200 * Make sure that shared libraries get the internal name engine with
14201 the full version number and not just 0. This should mark the
14202 shared libraries as not backward compatible. Of course, this should
14203 be changed again when we can guarantee backward binary compatibility.
14207 * Fix typo in get_cert_by_subject() in by_dir.c
14209 *Jean-Marc Desperrier <jean-marc.desperrier@certplus.com>*
14211 * Rework the system to generate shared libraries:
14213 - Make note of the expected extension for the shared libraries and
14214 if there is a need for symbolic links from for example libcrypto.so.0
14215 to libcrypto.so.0.9.7. There is extended info in Configure for
14218 - Make as few rebuilds of the shared libraries as possible.
14220 - Still avoid linking the OpenSSL programs with the shared libraries.
14222 - When installing, install the shared libraries separately from the
14227 * Fix SSL_CTX_set_read_ahead macro to actually use its argument.
14229 Copy SSL_CTX's read_ahead flag to SSL object directly in SSL_new
14230 and not in SSL_clear because the latter is also used by the
14231 accept/connect functions; previously, the settings made by
14232 SSL_set_read_ahead would be lost during the handshake.
14234 *Bodo Moeller; problems reported by Anders Gertz <gertz@epact.se>*
14236 * Correct util/mkdef.pl to be selective about disabled algorithms.
14237 Previously, it would create entries for disabled algorithms no
14242 * Added several new manual pages for SSL_* function.
14246 ### Changes between 0.9.5a and 0.9.6 [24 Sep 2000]
14248 * In ssl23_get_client_hello, generate an error message when faced
14249 with an initial SSL 3.0/TLS record that is too small to contain the
14250 first two bytes of the ClientHello message, i.e. client_version.
14251 (Note that this is a pathologic case that probably has never happened
14252 in real life.) The previous approach was to use the version number
14253 from the record header as a substitute; but our protocol choice
14254 should not depend on that one because it is not authenticated
14255 by the Finished messages.
14259 * More robust randomness gathering functions for Windows.
14261 *Jeffrey Altman <jaltman@columbia.edu>*
14263 * For compatibility reasons if the flag X509_V_FLAG_ISSUER_CHECK is
14264 not set then we don't setup the error code for issuer check errors
14265 to avoid possibly overwriting other errors which the callback does
14266 handle. If an application does set the flag then we assume it knows
14267 what it is doing and can handle the new informational codes
14272 * Fix for a nasty bug in ASN1_TYPE handling. ASN1_TYPE is used for
14273 a general "ANY" type, as such it should be able to decode anything
14274 including tagged types. However it didn't check the class so it would
14275 wrongly interpret tagged types in the same way as their universal
14276 counterpart and unknown types were just rejected. Changed so that the
14277 tagged and unknown types are handled in the same way as a SEQUENCE:
14278 that is the encoding is stored intact. There is also a new type
14279 "V_ASN1_OTHER" which is used when the class is not universal, in this
14280 case we have no idea what the actual type is so we just lump them all
14285 * On VMS, stdout may very well lead to a file that is written to
14286 in a record-oriented fashion. That means that every write() will
14287 write a separate record, which will be read separately by the
14288 programs trying to read from it. This can be very confusing.
14290 The solution is to put a BIO filter in the way that will buffer
14291 text until a linefeed is reached, and then write everything a
14292 line at a time, so every record written will be an actual line,
14293 not chunks of lines and not (usually doesn't happen, but I've
14294 seen it once) several lines in one record. BIO_f_linebuffer() is
14297 Currently, it's a VMS-only method, because that's where it has
14298 been tested well enough.
14302 * Remove 'optimized' squaring variant in BN_mod_mul_montgomery,
14303 it can return incorrect results.
14304 (Note: The buggy variant was not enabled in OpenSSL 0.9.5a,
14305 but it was in 0.9.6-beta[12].)
14309 * Disable the check for content being present when verifying detached
14310 signatures in pk7_smime.c. Some versions of Netscape (wrongly)
14311 include zero length content when signing messages.
14315 * New BIO_shutdown_wr macro, which invokes the BIO_C_SHUTDOWN_WR
14316 BIO_ctrl (for BIO pairs).
14320 * Add DSO method for VMS.
14324 * Bug fix: Montgomery multiplication could produce results with the
14329 * Add RPM specification openssl.spec and modify it to build three
14330 packages. The default package contains applications, application
14331 documentation and run-time libraries. The devel package contains
14332 include files, static libraries and function documentation. The
14333 doc package contains the contents of the doc directory. The original
14334 openssl.spec was provided by Damien Miller <djm@mindrot.org>.
14338 * Add a large number of documentation files for many SSL routines.
14340 *Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>*
14342 * Add a configuration entry for Sony News 4.
14344 *NAKAJI Hiroyuki <nakaji@tutrp.tut.ac.jp>*
14346 * Don't set the two most significant bits to one when generating a
14347 random number < q in the DSA library.
14351 * New SSL API mode 'SSL_MODE_AUTO_RETRY'. This disables the default
14352 behaviour that SSL_read may result in SSL_ERROR_WANT_READ (even if
14353 the underlying transport is blocking) if a handshake took place.
14354 (The default behaviour is needed by applications such as s_client
14355 and s_server that use select() to determine when to use SSL_read;
14356 but for applications that know in advance when to expect data, it
14357 just makes things more complicated.)
14361 * Add RAND_egd_bytes(), which gives control over the number of bytes read
14366 * Add a few more EBCDIC conditionals that make `req` and `x509`
14367 work better on such systems.
14369 *Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>*
14371 * Add two demo programs for PKCS12_parse() and PKCS12_create().
14372 Update PKCS12_parse() so it copies the friendlyName and the
14373 keyid to the certificates aux info.
14377 * Fix bug in PKCS7_verify() which caused an infinite loop
14378 if there was more than one signature.
14380 *Sven Uszpelkat <su@celocom.de>*
14382 * Major change in util/mkdef.pl to include extra information
14383 about each symbol, as well as presenting variables as well
14384 as functions. This change means that there's n more need
14385 to rebuild the .num files when some algorithms are excluded.
14389 * Allow the verify time to be set by an application,
14390 rather than always using the current time.
14394 * Phase 2 verify code reorganisation. The certificate
14395 verify code now looks up an issuer certificate by a
14396 number of criteria: subject name, authority key id
14397 and key usage. It also verifies self signed certificates
14398 by the same criteria. The main comparison function is
14399 X509_check_issued() which performs these checks.
14401 Lot of changes were necessary in order to support this
14402 without completely rewriting the lookup code.
14404 Authority and subject key identifier are now cached.
14406 The LHASH 'certs' is X509_STORE has now been replaced
14407 by a STACK_OF(X509_OBJECT). This is mainly because an
14408 LHASH can't store or retrieve multiple objects with
14409 the same hash value.
14411 As a result various functions (which were all internal
14412 use only) have changed to handle the new X509_STORE
14413 structure. This will break anything that messed round
14414 with X509_STORE internally.
14416 The functions X509_STORE_add_cert() now checks for an
14417 exact match, rather than just subject name.
14419 The X509_STORE API doesn't directly support the retrieval
14420 of multiple certificates matching a given criteria, however
14421 this can be worked round by performing a lookup first
14422 (which will fill the cache with candidate certificates)
14423 and then examining the cache for matches. This is probably
14424 the best we can do without throwing out X509_LOOKUP
14425 entirely (maybe later...).
14427 The X509_VERIFY_CTX structure has been enhanced considerably.
14429 All certificate lookup operations now go via a get_issuer()
14430 callback. Although this currently uses an X509_STORE it
14431 can be replaced by custom lookups. This is a simple way
14432 to bypass the X509_STORE hackery necessary to make this
14433 work and makes it possible to use more efficient techniques
14434 in future. A very simple version which uses a simple
14435 STACK for its trusted certificate store is also provided
14436 using X509_STORE_CTX_trusted_stack().
14438 The verify_cb() and verify() callbacks now have equivalents
14439 in the X509_STORE_CTX structure.
14441 X509_STORE_CTX also has a 'flags' field which can be used
14442 to customise the verify behaviour.
14446 * Add new PKCS#7 signing option PKCS7_NOSMIMECAP which
14447 excludes S/MIME capabilities.
14451 * When a certificate request is read in keep a copy of the
14452 original encoding of the signed data and use it when outputting
14453 again. Signatures then use the original encoding rather than
14454 a decoded, encoded version which may cause problems if the
14455 request is improperly encoded.
14459 * For consistency with other BIO_puts implementations, call
14460 buffer_write(b, ...) directly in buffer_puts instead of calling
14463 In BIO_puts, increment b->num_write as in BIO_write.
14465 *Peter.Sylvester@EdelWeb.fr*
14467 * Fix BN_mul_word for the case where the word is 0. (We have to use
14468 BN_zero, we may not return a BIGNUM with an array consisting of
14469 words set to zero.)
14473 * Avoid calling abort() from within the library when problems are
14474 detected, except if preprocessor symbols have been defined
14475 (such as REF_CHECK, BN_DEBUG etc.).
14479 * New openssl application 'rsautl'. This utility can be
14480 used for low level RSA operations. DER public key
14481 BIO/fp routines also added.
14485 * New Configure entry and patches for compiling on QNX 4.
14487 *Andreas Schneider <andreas@ds3.etech.fh-hamburg.de>*
14489 * A demo state-machine implementation was sponsored by
14490 Nuron (<http://www.nuron.com/>) and is now available in
14491 demos/state_machine.
14495 * New options added to the 'dgst' utility for signature
14496 generation and verification.
14500 * Unrecognized PKCS#7 content types are now handled via a
14501 catch all ASN1_TYPE structure. This allows unsupported
14502 types to be stored as a "blob" and an application can
14503 encode and decode it manually.
14507 * Fix various signed/unsigned issues to make a_strex.c
14508 compile under VC++.
14510 *Oscar Jacobsson <oscar.jacobsson@celocom.com>*
14512 * ASN1 fixes. i2d_ASN1_OBJECT was not returning the correct
14513 length if passed a buffer. ASN1_INTEGER_to_BN failed
14514 if passed a NULL BN and its argument was negative.
14516 *Steve Henson, pointed out by Sven Heiberg <sven@tartu.cyber.ee>*
14518 * Modification to PKCS#7 encoding routines to output definite
14519 length encoding. Since currently the whole structures are in
14520 memory there's not real point in using indefinite length
14521 constructed encoding. However if OpenSSL is compiled with
14522 the flag PKCS7_INDEFINITE_ENCODING the old form is used.
14526 * Added BIO_vprintf() and BIO_vsnprintf().
14530 * Added more prefixes to parse for in the strings written
14531 through a logging bio, to cover all the levels that are available
14532 through syslog. The prefixes are now:
14534 PANIC, EMERG, EMR => LOG_EMERG
14535 ALERT, ALR => LOG_ALERT
14536 CRIT, CRI => LOG_CRIT
14537 ERROR, ERR => LOG_ERR
14538 WARNING, WARN, WAR => LOG_WARNING
14539 NOTICE, NOTE, NOT => LOG_NOTICE
14540 INFO, INF => LOG_INFO
14541 DEBUG, DBG => LOG_DEBUG
14543 and as before, if none of those prefixes are present at the
14544 beginning of the string, LOG_ERR is chosen.
14546 On Win32, the `LOG_*` levels are mapped according to this:
14548 LOG_EMERG, LOG_ALERT, LOG_CRIT, LOG_ERR => EVENTLOG_ERROR_TYPE
14549 LOG_WARNING => EVENTLOG_WARNING_TYPE
14550 LOG_NOTICE, LOG_INFO, LOG_DEBUG => EVENTLOG_INFORMATION_TYPE
14554 * Made it possible to reconfigure with just the configuration
14555 argument "reconf" or "reconfigure". The command line arguments
14556 are stored in Makefile.ssl in the variable CONFIGURE_ARGS,
14557 and are retrieved from there when reconfiguring.
14563 *Assar Westerlund <assar@sics.se>, Richard Levitte*
14565 * Add the arguments -CAfile and -CApath to the pkcs12 utility.
14569 * The obj_dat.pl script was messing up the sorting of object
14570 names. The reason was that it compared the quoted version
14571 of strings as a result "OCSP" > "OCSP Signing" because
14572 " > SPACE. Changed script to store unquoted versions of
14573 names and add quotes on output. It was also omitting some
14574 names from the lookup table if they were given a default
14575 value (that is if SN is missing it is given the same
14576 value as LN and vice versa), these are now added on the
14577 grounds that if an object has a name we should be able to
14578 look it up. Finally added warning output when duplicate
14579 short or long names are found.
14583 * Changes needed for Tandem NSK.
14585 *Scott Uroff <scott@xypro.com>*
14587 * Fix SSL 2.0 rollback checking: Due to an off-by-one error in
14588 RSA_padding_check_SSLv23(), special padding was never detected
14589 and thus the SSL 3.0/TLS 1.0 countermeasure against protocol
14590 version rollback attacks was not effective.
14592 In s23_clnt.c, don't use special rollback-attack detection padding
14593 (RSA_SSLV23_PADDING) if SSL 2.0 is the only protocol enabled in the
14594 client; similarly, in s23_srvr.c, don't do the rollback check if
14595 SSL 2.0 is the only protocol enabled in the server.
14599 * Make it possible to get hexdumps of unprintable data with 'openssl
14600 asn1parse'. By implication, the functions ASN1_parse_dump() and
14601 BIO_dump_indent() are added.
14605 * New functions ASN1_STRING_print_ex() and X509_NAME_print_ex()
14606 these print out strings and name structures based on various
14607 flags including RFC2253 support and proper handling of
14608 multibyte characters. Added options to the 'x509' utility
14609 to allow the various flags to be set.
14613 * Various fixes to use ASN1_TIME instead of ASN1_UTCTIME.
14614 Also change the functions X509_cmp_current_time() and
14615 X509_gmtime_adj() work with an ASN1_TIME structure,
14616 this will enable certificates using GeneralizedTime in validity
14617 dates to be checked.
14621 * Make the NEG_PUBKEY_BUG code (which tolerates invalid
14622 negative public key encodings) on by default,
14623 NO_NEG_PUBKEY_BUG can be set to disable it.
14627 * New function c2i_ASN1_OBJECT() which acts on ASN1_OBJECT
14628 content octets. An i2c_ASN1_OBJECT is unnecessary because
14629 the encoding can be trivially obtained from the structure.
14633 * crypto/err.c locking bugfix: Use write locks (`CRYPTO_w_[un]lock`),
14634 not read locks (`CRYPTO_r_[un]lock`).
14638 * A first attempt at creating official support for shared
14639 libraries through configuration. I've kept it so the
14640 default is static libraries only, and the OpenSSL programs
14641 are always statically linked for now, but there are
14642 preparations for dynamic linking in place.
14643 This has been tested on Linux and Tru64.
14647 * Randomness polling function for Win9x, as described in:
14648 Peter Gutmann, Software Generation of Practically Strong
14653 * Fix so PRNG is seeded in req if using an already existing
14658 * New options to smime application. -inform and -outform
14659 allow alternative formats for the S/MIME message including
14660 PEM and DER. The -content option allows the content to be
14661 specified separately. This should allow things like Netscape
14662 form signing output easier to verify.
14666 * Fix the ASN1 encoding of tags using the 'long form'.
14670 * New ASN1 functions, `i2c_*` and `c2i_*` for INTEGER and BIT
14671 STRING types. These convert content octets to and from the
14672 underlying type. The actual tag and length octets are
14673 already assumed to have been read in and checked. These
14674 are needed because all other string types have virtually
14675 identical handling apart from the tag. By having versions
14676 of the ASN1 functions that just operate on content octets
14677 IMPLICIT tagging can be handled properly. It also allows
14678 the ASN1_ENUMERATED code to be cut down because ASN1_ENUMERATED
14679 and ASN1_INTEGER are identical apart from the tag.
14683 * Change the handling of OID objects as follows:
14685 - New object identifiers are inserted in objects.txt, following
14686 the syntax given in objects.README.
14687 - objects.pl is used to process obj_mac.num and create a new
14689 - obj_dat.pl is used to create a new obj_dat.h, using the data in
14692 This is currently kind of a hack, and the perl code in objects.pl
14693 isn't very elegant, but it works as I intended. The simplest way
14694 to check that it worked correctly is to look in obj_dat.h and
14695 check the array nid_objs and make sure the objects haven't moved
14696 around (this is important!). Additions are OK, as well as
14697 consistent name changes.
14701 * Add BSD-style MD5-based passwords to 'openssl passwd' (option '-1').
14705 * Addition of the command line parameter '-rand file' to 'openssl req'.
14706 The given file adds to whatever has already been seeded into the
14707 random pool through the RANDFILE configuration file option or
14708 environment variable, or the default random state file.
14712 * mkstack.pl now sorts each macro group into lexical order.
14713 Previously the output order depended on the order the files
14714 appeared in the directory, resulting in needless rewriting
14719 * Patches to make OpenSSL compile under Win32 again. Mostly
14720 work arounds for the VC++ problem that it treats func() as
14721 func(void). Also stripped out the parts of mkdef.pl that
14722 added extra typesafe functions: these no longer exist.
14726 * Reorganisation of the stack code. The macros are now all
14727 collected in safestack.h . Each macro is defined in terms of
14728 a "stack macro" of the form `SKM_<name>(type, a, b)`. The
14729 DEBUG_SAFESTACK is now handled in terms of function casts,
14730 this has the advantage of retaining type safety without the
14731 use of additional functions. If DEBUG_SAFESTACK is not defined
14732 then the non typesafe macros are used instead. Also modified the
14733 mkstack.pl script to handle the new form. Needs testing to see
14734 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK
14735 the default if no major problems. Similar behaviour for ASN1_SET_OF
14736 and PKCS12_STACK_OF.
14740 * When some versions of IIS use the 'NET' form of private key the
14741 key derivation algorithm is different. Normally MD5(password) is
14742 used as a 128 bit RC4 key. In the modified case
14743 MD5(MD5(password) + "SGCKEYSALT") is used instead. Added some
14744 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same
14745 as the old Netscape_RSA functions except they have an additional
14746 'sgckey' parameter which uses the modified algorithm. Also added
14747 an -sgckey command line option to the rsa utility. Thanks to
14748 Adrian Peck <bertie@ncipher.com> for posting details of the modified
14749 algorithm to openssl-dev.
14753 * The evp_local.h macros were using 'c.##kname' which resulted in
14754 invalid expansion on some systems (SCO 5.0.5 for example).
14755 Corrected to 'c.kname'.
14757 *Phillip Porch <root@theporch.com>*
14759 * New X509_get1_email() and X509_REQ_get1_email() functions that return
14760 a STACK of email addresses from a certificate or request, these look
14761 in the subject name and the subject alternative name extensions and
14762 omit any duplicate addresses.
14766 * Re-implement BN_mod_exp2_mont using independent (and larger) windows.
14767 This makes DSA verification about 2 % faster.
14771 * Increase maximum window size in `BN_mod_exp_...` to 6 bits instead of 5
14772 (meaning that now 2^5 values will be precomputed, which is only 4 KB
14773 plus overhead for 1024 bit moduli).
14774 This makes exponentiations about 0.5 % faster for 1024 bit
14775 exponents (as measured by "openssl speed rsa2048").
14779 * Rename memory handling macros to avoid conflicts with other
14781 Malloc => OPENSSL_malloc
14782 Malloc_locked => OPENSSL_malloc_locked
14783 Realloc => OPENSSL_realloc
14784 Free => OPENSSL_free
14788 * New function BN_mod_exp_mont_word for small bases (roughly 15%
14789 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange).
14793 * CygWin32 support.
14795 *John Jarvie <jjarvie@newsguy.com>*
14797 * The type-safe stack code has been rejigged. It is now only compiled
14798 in when OpenSSL is configured with the DEBUG_SAFESTACK option and
14799 by default all type-specific stack functions are "#define"d back to
14800 standard stack functions. This results in more streamlined output
14801 but retains the type-safety checking possibilities of the original
14806 * The STACK code has been cleaned up, and certain type declarations
14807 that didn't make a lot of sense have been brought in line. This has
14808 also involved a cleanup of sorts in safestack.h to more correctly
14809 map type-safe stack functions onto their plain stack counterparts.
14810 This work has also resulted in a variety of "const"ifications of
14811 lots of the code, especially `_cmp` operations which should normally
14812 be prototyped with "const" parameters anyway.
14816 * When generating bytes for the first time in md_rand.c, 'stir the pool'
14817 by seeding with STATE_SIZE dummy bytes (with zero entropy count).
14818 (The PRNG state consists of two parts, the large pool 'state' and 'md',
14819 where all of 'md' is used each time the PRNG is used, but 'state'
14820 is used only indexed by a cyclic counter. As entropy may not be
14821 well distributed from the beginning, 'md' is important as a
14822 chaining variable. However, the output function chains only half
14823 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains
14824 all of 'md', and seeding with STATE_SIZE dummy bytes will result
14825 in all of 'state' being rewritten, with the new values depending
14826 on virtually all of 'md'. This overcomes the 80 bit limitation.)
14830 * In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when
14831 the handshake is continued after ssl_verify_cert_chain();
14832 otherwise, if SSL_VERIFY_NONE is set, remaining error codes
14833 can lead to 'unexplainable' connection aborts later.
14835 *Bodo Moeller; problem tracked down by Lutz Jaenicke*
14837 * Major EVP API cipher revision.
14838 Add hooks for extra EVP features. This allows various cipher
14839 parameters to be set in the EVP interface. Support added for variable
14840 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and
14841 setting of RC2 and RC5 parameters.
14843 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length
14846 Remove lots of duplicated code from the EVP library. For example *every*
14847 cipher init() function handles the 'iv' in the same way according to the
14848 cipher mode. They also all do nothing if the 'key' parameter is NULL and
14849 for CFB and OFB modes they zero ctx->num.
14851 New functionality allows removal of S/MIME code RC2 hack.
14853 Most of the routines have the same form and so can be declared in terms
14856 By shifting this to the top level EVP_CipherInit() it can be removed from
14857 all individual ciphers. If the cipher wants to handle IVs or keys
14858 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT
14861 Change lots of functions like EVP_EncryptUpdate() to now return a
14862 value: although software versions of the algorithms cannot fail
14863 any installed hardware versions can.
14867 * Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if
14868 this option is set, tolerate broken clients that send the negotiated
14869 protocol version number instead of the requested protocol version
14874 * Call dh_tmp_cb (set by `..._TMP_DH_CB`) with correct 'is_export' flag;
14875 i.e. non-zero for export ciphersuites, zero otherwise.
14876 Previous versions had this flag inverted, inconsistent with
14877 rsa_tmp_cb (..._TMP_RSA_CB).
14879 *Bodo Moeller; problem reported by Amit Chopra*
14881 * Add missing DSA library text string. Work around for some IIS
14882 key files with invalid SEQUENCE encoding.
14886 * Add a document (doc/standards.txt) that list all kinds of standards
14887 and so on that are implemented in OpenSSL.
14891 * Enhance c_rehash script. Old version would mishandle certificates
14892 with the same subject name hash and wouldn't handle CRLs at all.
14893 Added -fingerprint option to crl utility, to support new c_rehash
14898 * Eliminate non-ANSI declarations in crypto.h and stack.h.
14902 * Fix for SSL server purpose checking. Server checking was
14903 rejecting certificates which had extended key usage present
14904 but no ssl client purpose.
14906 *Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>*
14908 * Make PKCS#12 code work with no password. The PKCS#12 spec
14909 is a little unclear about how a blank password is handled.
14910 Since the password in encoded as a BMPString with terminating
14911 double NULL a zero length password would end up as just the
14912 double NULL. However no password at all is different and is
14913 handled differently in the PKCS#12 key generation code. NS
14914 treats a blank password as zero length. MSIE treats it as no
14915 password on export: but it will try both on import. We now do
14916 the same: PKCS12_parse() tries zero length and no password if
14917 the password is set to "" or NULL (NULL is now a valid password:
14918 it wasn't before) as does the pkcs12 application.
14922 * Bugfixes in apps/x509.c: Avoid a memory leak; and don't use
14923 perror when PEM_read_bio_X509_REQ fails, the error message must
14924 be obtained from the error queue.
14928 * Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing
14929 it in ERR_remove_state if appropriate, and change ERR_get_state
14930 accordingly to avoid race conditions (this is necessary because
14931 thread_hash is no longer constant once set).
14935 * Bugfix for linux-elf makefile.one.
14939 * RSA_get_default_method() will now cause a default
14940 RSA_METHOD to be chosen if one doesn't exist already.
14941 Previously this was only set during a call to RSA_new()
14942 or RSA_new_method(NULL) meaning it was possible for
14943 RSA_get_default_method() to return NULL.
14947 * Added native name translation to the existing DSO code
14948 that will convert (if the flag to do so is set) filenames
14949 that are sufficiently small and have no path information
14950 into a canonical native form. Eg. "blah" converted to
14951 "libblah.so" or "blah.dll" etc.
14955 * New function ERR_error_string_n(e, buf, len) which is like
14956 ERR_error_string(e, buf), but writes at most 'len' bytes
14957 including the 0 terminator. For ERR_error_string_n, 'buf'
14960 *Damien Miller <djm@mindrot.org>, Bodo Moeller*
14962 * CONF library reworked to become more general. A new CONF
14963 configuration file reader "class" is implemented as well as a
14964 new functions (`NCONF_*`, for "New CONF") to handle it. The now
14965 old `CONF_*` functions are still there, but are reimplemented to
14966 work in terms of the new functions. Also, a set of functions
14967 to handle the internal storage of the configuration data is
14968 provided to make it easier to write new configuration file
14969 reader "classes" (I can definitely see something reading a
14970 configuration file in XML format, for example), called `_CONF_*`,
14971 or "the configuration storage API"...
14973 The new configuration file reading functions are:
14975 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio,
14976 NCONF_get_section, NCONF_get_string, NCONF_get_numbre
14978 NCONF_default, NCONF_WIN32
14980 NCONF_dump_fp, NCONF_dump_bio
14982 NCONF_default and NCONF_WIN32 are method (or "class") choosers,
14983 NCONF_new creates a new CONF object. This works in the same way
14984 as other interfaces in OpenSSL, like the BIO interface.
14985 `NCONF_dump_*` dump the internal storage of the configuration file,
14986 which is useful for debugging. All other functions take the same
14987 arguments as the old `CONF_*` functions with the exception of the
14988 first that must be a `CONF *` instead of a `LHASH *`.
14990 To make it easier to use the new classes with the old `CONF_*` functions,
14991 the function CONF_set_default_method is provided.
14995 * Add '-tls1' option to 'openssl ciphers', which was already
14996 mentioned in the documentation but had not been implemented.
14997 (This option is not yet really useful because even the additional
14998 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.)
15002 * Initial DSO code added into libcrypto for letting OpenSSL (and
15003 OpenSSL-based applications) load shared libraries and bind to
15004 them in a portable way.
15006 *Geoff Thorpe, with contributions from Richard Levitte*
15008 ### Changes between 0.9.5 and 0.9.5a [1 Apr 2000]
15010 * Make sure _lrotl and _lrotr are only used with MSVC.
15012 * Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status
15013 (the default implementation of RAND_status).
15015 * Rename openssl x509 option '-crlext', which was added in 0.9.5,
15016 to '-clrext' (= clear extensions), as intended and documented.
15017 *Bodo Moeller; inconsistency pointed out by Michael Attili
15018 <attili@amaxo.com>*
15020 * Fix for HMAC. It wasn't zeroing the rest of the block if the key length
15021 was larger than the MD block size.
15023 *Steve Henson, pointed out by Yost William <YostW@tce.com>*
15025 * Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument
15026 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set()
15027 using the passed key: if the passed key was a private key the result
15028 of X509_print(), for example, would be to print out all the private key
15033 * des_quad_cksum() byte order bug fix.
15034 *Ulf Möller, using the problem description in krb4-0.9.7, where
15035 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>*
15037 * Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly
15040 *Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>*
15042 * For easily testing in shell scripts whether some command
15043 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX'
15044 returns with exit code 0 iff no command of the given name is available.
15045 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases,
15046 the output goes to stdout and nothing is printed to stderr.
15047 Additional arguments are always ignored.
15049 Since for each cipher there is a command of the same name,
15050 the 'no-cipher' compilation switches can be tested this way.
15052 ('openssl no-XXX' is not able to detect pseudo-commands such
15053 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.)
15057 * Update test suite so that 'make test' succeeds in 'no-rsa' configuration.
15061 * For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE
15062 is set; it will be thrown away anyway because each handshake creates
15064 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition
15065 to parameters -- in previous versions (since OpenSSL 0.9.3) the
15066 'default key' from SSL_CTX_set_tmp_dh would always be lost, meaning
15067 you effectively got SSL_OP_SINGLE_DH_USE when using this macro.
15071 * New s_client option -ign_eof: EOF at stdin is ignored, and
15072 'Q' and 'R' lose their special meanings (quit/renegotiate).
15073 This is part of what -quiet does; unlike -quiet, -ign_eof
15074 does not suppress any output.
15078 * Add compatibility options to the purpose and trust code. The
15079 purpose X509_PURPOSE_ANY is "any purpose" which automatically
15080 accepts a certificate or CA, this was the previous behaviour,
15081 with all the associated security issues.
15083 X509_TRUST_COMPAT is the old trust behaviour: only and
15084 automatically trust self signed roots in certificate store. A
15085 new trust setting X509_TRUST_DEFAULT is used to specify that
15086 a purpose has no associated trust setting and it should instead
15087 use the value in the default purpose.
15091 * Fix the PKCS#8 DSA private key code so it decodes keys again
15092 and fix a memory leak.
15096 * In util/mkerr.pl (which implements 'make errors'), preserve
15097 reason strings from the previous version of the .c file, as
15098 the default to have only downcase letters (and digits) in
15099 automatically generated reasons codes is not always appropriate.
15103 * In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table
15104 using strerror. Previously, ERR_reason_error_string() returned
15105 library names as reason strings for SYSerr; but SYSerr is a special
15106 case where small numbers are errno values, not library numbers.
15110 * Add '-dsaparam' option to 'openssl dhparam' application. This
15111 converts DSA parameters into DH parameters. (When creating parameters,
15112 DSA_generate_parameters is used.)
15116 * Include 'length' (recommended exponent length) in C code generated
15117 by 'openssl dhparam -C'.
15121 * The second argument to set_label in perlasm was already being used
15122 so couldn't be used as a "file scope" flag. Moved to third argument
15127 * In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes
15128 instead of RAND_bytes for encryption IVs and salts.
15132 * Include RAND_status() into RAND_METHOD instead of implementing
15133 it only for md_rand.c Otherwise replacing the PRNG by calling
15134 RAND_set_rand_method would be impossible.
15138 * Don't let DSA_generate_key() enter an infinite loop if the random
15139 number generation fails.
15143 * New 'rand' application for creating pseudo-random output.
15147 * Added configuration support for Linux/IA64
15149 *Rolf Haberrecker <rolf@suse.de>*
15151 * Assembler module support for Mingw32.
15155 * Shared library support for HPUX (in shlib/).
15157 *Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous*
15159 * Shared library support for Solaris gcc.
15161 *Lutz Behnke <behnke@trustcenter.de>*
15163 ### Changes between 0.9.4 and 0.9.5 [28 Feb 2000]
15165 * PKCS7_encrypt() was adding text MIME headers twice because they
15166 were added manually and by SMIME_crlf_copy().
15170 * In bntest.c don't call BN_rand with zero bits argument.
15172 *Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>*
15174 * BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n]
15175 case was implemented. This caused BN_div_recp() to fail occasionally.
15179 * Add an optional second argument to the set_label() in the perl
15180 assembly language builder. If this argument exists and is set
15181 to 1 it signals that the assembler should use a symbol whose
15182 scope is the entire file, not just the current function. This
15183 is needed with MASM which uses the format label:: for this scope.
15185 *Steve Henson, pointed out by Peter Runestig <peter@runestig.com>*
15187 * Change the ASN1 types so they are typedefs by default. Before
15188 almost all types were #define'd to ASN1_STRING which was causing
15189 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING)
15194 * Change names of new functions to the new get1/get0 naming
15195 convention: After 'get1', the caller owns a reference count
15196 and has to call `..._free`; 'get0' returns a pointer to some
15197 data structure without incrementing reference counters.
15198 (Some of the existing 'get' functions increment a reference
15199 counter, some don't.)
15200 Similarly, 'set1' and 'add1' functions increase reference
15201 counters or duplicate objects.
15205 * Allow for the possibility of temp RSA key generation failure:
15206 the code used to assume it always worked and crashed on failure.
15210 * Fix potential buffer overrun problem in BIO_printf().
15211 *Ulf Möller, using public domain code by Patrick Powell; problem
15212 pointed out by David Sacerdote <das33@cornell.edu>*
15214 * Support EGD <http://www.lothar.com/tech/crypto/>. New functions
15215 RAND_egd() and RAND_status(). In the command line application,
15216 the EGD socket can be specified like a seed file using RANDFILE
15221 * Allow the string CERTIFICATE to be tolerated in PKCS#7 structures.
15222 Some CAs (e.g. Verisign) distribute certificates in this form.
15226 * Remove the SSL_ALLOW_ADH compile option and set the default cipher
15227 list to exclude them. This means that no special compilation option
15228 is needed to use anonymous DH: it just needs to be included in the
15233 * Change the EVP_MD_CTX_type macro so its meaning consistent with
15234 EVP_MD_type. The old functionality is available in a new macro called
15235 EVP_MD_md(). Change code that uses it and update docs.
15239 * `..._ctrl` functions now have corresponding `..._callback_ctrl` functions
15240 where the `void *` argument is replaced by a function pointer argument.
15241 Previously `void *` was abused to point to functions, which works on
15242 many platforms, but is not correct. As these functions are usually
15243 called by macros defined in OpenSSL header files, most source code
15244 should work without changes.
15248 * `<openssl/opensslconf.h>` (which is created by Configure) now contains
15249 sections with information on -D... compiler switches used for
15250 compiling the library so that applications can see them. To enable
15251 one of these sections, a pre-processor symbol `OPENSSL_..._DEFINES`
15252 must be defined. E.g.,
15253 #define OPENSSL_ALGORITHM_DEFINES
15254 #include <openssl/opensslconf.h>
15255 defines all pertinent `NO_<algo>` symbols, such as NO_IDEA, NO_RSA, etc.
15257 *Richard Levitte, Ulf and Bodo Möller*
15259 * Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS
15264 * Change the 'other' type in certificate aux info to a STACK_OF
15265 X509_ALGOR. Although not an AlgorithmIdentifier as such it has
15266 the required ASN1 format: arbitrary types determined by an OID.
15270 * Add some PEM_write_X509_REQ_NEW() functions and a command line
15271 argument to 'req'. This is not because the function is newer or
15272 better than others it just uses the work 'NEW' in the certificate
15273 request header lines. Some software needs this.
15277 * Reorganise password command line arguments: now passwords can be
15278 obtained from various sources. Delete the PEM_cb function and make
15279 it the default behaviour: i.e. if the callback is NULL and the
15280 usrdata argument is not NULL interpret it as a null terminated pass
15281 phrase. If usrdata and the callback are NULL then the pass phrase
15282 is prompted for as usual.
15286 * Add support for the Compaq Atalla crypto accelerator. If it is installed,
15287 the support is automatically enabled. The resulting binaries will
15288 autodetect the card and use it if present.
15290 *Ben Laurie and Compaq Inc.*
15292 * Work around for Netscape hang bug. This sends certificate request
15293 and server done in one record. Since this is perfectly legal in the
15294 SSL/TLS protocol it isn't a "bug" option and is on by default. See
15295 the bugs/SSLv3 entry for more info.
15299 * HP-UX tune-up: new unified configs, HP C compiler bug workaround.
15303 * Add -rand argument to smime and pkcs12 applications and read/write
15308 * New 'passwd' tool for crypt(3) and apr1 password hashes.
15312 * Add command line password options to the remaining applications.
15316 * Bug fix for BN_div_recp() for numerators with an even number of
15321 * More tests in bntest.c, and changed test_bn output.
15325 * ./config recognizes MacOS X now.
15329 * Bug fix for BN_div() when the first words of num and divisor are
15330 equal (it gave wrong results if `(rem=(n1-q*d0)&BN_MASK2) < d0)`.
15334 * Add support for various broken PKCS#8 formats, and command line
15335 options to produce them.
15339 * New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to
15340 get temporary BIGNUMs from a BN_CTX.
15344 * Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont()
15349 * Change the `SSLeay_add_all_*()` functions to `OpenSSL_add_all_*()` and
15350 include a #define from the old name to the new. The original intent
15351 was that statically linked binaries could for example just call
15352 SSLeay_add_all_ciphers() to just add ciphers to the table and not
15353 link with digests. This never worked because SSLeay_add_all_digests()
15354 and SSLeay_add_all_ciphers() were in the same source file so calling
15355 one would link with the other. They are now in separate source files.
15359 * Add a new -notext option to 'ca' and a -pubkey option to 'spkac'.
15363 * Use a less unusual form of the Miller-Rabin primality test (it used
15364 a binary algorithm for exponentiation integrated into the Miller-Rabin
15365 loop, our standard modexp algorithms are faster).
15369 * Support for the EBCDIC character set completed.
15371 *Martin Kraemer <Martin.Kraemer@Mch.SNI.De>*
15373 * Source code cleanups: use const where appropriate, eliminate casts,
15374 use `void *` instead of `char *` in lhash.
15378 * Bugfix: ssl3_send_server_key_exchange was not restartable
15379 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of
15380 this the server could overwrite ephemeral keys that the client
15385 * Turn DSA_is_prime into a macro that calls BN_is_prime,
15386 using 50 iterations of the Rabin-Miller test.
15388 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50
15389 iterations of the Rabin-Miller test as required by the appendix
15390 to FIPS PUB 186[-1]) instead of DSA_is_prime.
15391 As BN_is_prime_fasttest includes trial division, DSA parameter
15392 generation becomes much faster.
15394 This implies a change for the callback functions in DSA_is_prime
15395 and DSA_generate_parameters: The callback function is called once
15396 for each positive witness in the Rabin-Miller test, not just
15397 occasionally in the inner loop; and the parameters to the
15398 callback function now provide an iteration count for the outer
15399 loop rather than for the current invocation of the inner loop.
15400 DSA_generate_parameters additionally can call the callback
15401 function with an 'iteration count' of -1, meaning that a
15402 candidate has passed the trial division test (when q is generated
15403 from an application-provided seed, trial division is skipped).
15407 * New function BN_is_prime_fasttest that optionally does trial
15408 division before starting the Rabin-Miller test and has
15409 an additional BN_CTX * argument (whereas BN_is_prime always
15410 has to allocate at least one BN_CTX).
15411 'callback(1, -1, cb_arg)' is called when a number has passed the
15412 trial division stage.
15416 * Fix for bug in CRL encoding. The validity dates weren't being handled
15421 * New -pkcs12 option to CA.pl script to write out a PKCS#12 file.
15425 * New function BN_pseudo_rand().
15429 * Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable)
15430 bignum version of BN_from_montgomery() with the working code from
15431 SSLeay 0.9.0 (the word based version is faster anyway), and clean up
15436 * Avoid a race condition in s2_clnt.c (function get_server_hello) that
15437 made it impossible to use the same SSL_SESSION data structure in
15438 SSL2 clients in multiple threads.
15442 * The return value of RAND_load_file() no longer counts bytes obtained
15443 by stat(). RAND_load_file(..., -1) is new and uses the complete file
15444 to seed the PRNG (previously an explicit byte count was required).
15446 *Ulf Möller, Bodo Möller*
15448 * Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
15449 used `char *` instead of `void *` and had casts all over the place.
15453 * Make BN_generate_prime() return NULL on error if ret!=NULL.
15457 * Retain source code compatibility for BN_prime_checks macro:
15458 BN_is_prime(..., BN_prime_checks, ...) now uses
15459 BN_prime_checks_for_size to determine the appropriate number of
15460 Rabin-Miller iterations.
15464 * Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
15465 DH_CHECK_P_NOT_SAFE_PRIME.
15466 (Check if this is true? OpenPGP calls them "strong".)
15470 * Merge the functionality of "dh" and "gendh" programs into a new program
15471 "dhparam". The old programs are retained for now but will handle DH keys
15472 (instead of parameters) in future.
15476 * Make the ciphers, s_server and s_client programs check the return values
15477 when a new cipher list is set.
15481 * Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
15482 ciphers. Before when the 56bit ciphers were enabled the sorting was
15485 The syntax for the cipher sorting has been extended to support sorting by
15486 cipher-strength (using the strength_bits hard coded in the tables).
15487 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
15489 Fix a bug in the cipher-command parser: when supplying a cipher command
15490 string with an "undefined" symbol (neither command nor alphanumeric
15491 *A-Za-z0-9*, ssl_set_cipher_list used to hang in an endless loop. Now
15492 an error is flagged.
15494 Due to the strength-sorting extension, the code of the
15495 ssl_create_cipher_list() function was completely rearranged. I hope that
15496 the readability was also increased :-)
15498 *Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>*
15500 * Minor change to 'x509' utility. The -CAcreateserial option now uses 1
15501 for the first serial number and places 2 in the serial number file. This
15502 avoids problems when the root CA is created with serial number zero and
15503 the first user certificate has the same issuer name and serial number
15508 * Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
15509 the new code. Add documentation for this stuff.
15513 * Changes to X509_ATTRIBUTE utilities. These have been renamed from
15514 `X509_*()` to `X509at_*()` on the grounds that they don't handle X509
15515 structures and behave in an analogous way to the X509v3 functions:
15516 they shouldn't be called directly but wrapper functions should be used
15519 So we also now have some wrapper functions that call the X509at functions
15520 when passed certificate requests. (TO DO: similar things can be done with
15521 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
15522 things. Some of these need some d2i or i2d and print functionality
15523 because they handle more complex structures.)
15527 * Add missing #ifndefs that caused missing symbols when building libssl
15528 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
15529 NO_RSA in `ssl/s2*.c`.
15531 *Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller*
15533 * Precautions against using the PRNG uninitialized: RAND_bytes() now
15534 has a return value which indicates the quality of the random data
15535 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
15536 error queue. New function RAND_pseudo_bytes() generates output that is
15537 guaranteed to be unique but not unpredictable. RAND_add is like
15538 RAND_seed, but takes an extra argument for an entropy estimate
15539 (RAND_seed always assumes full entropy).
15543 * Do more iterations of Rabin-Miller probable prime test (specifically,
15544 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
15545 instead of only 2 for all lengths; see BN_prime_checks_for_size definition
15546 in crypto/bn/bn_prime.c for the complete table). This guarantees a
15547 false-positive rate of at most 2^-80 for random input.
15551 * Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
15555 * New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain
15556 in the 0.9.5 release), this returns the chain
15557 from an X509_CTX structure with a dup of the stack and all
15558 the X509 reference counts upped: so the stack will exist
15559 after X509_CTX_cleanup() has been called. Modify pkcs12.c
15562 Also make SSL_SESSION_print() print out the verify return
15567 * Add manpage for the pkcs12 command. Also change the default
15568 behaviour so MAC iteration counts are used unless the new
15569 -nomaciter option is used. This improves file security and
15570 only older versions of MSIE (4.0 for example) need it.
15574 * Honor the no-xxx Configure options when creating .DEF files.
15578 * Add PKCS#10 attributes to field table: challengePassword,
15579 unstructuredName and unstructuredAddress. These are taken from
15580 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
15581 international characters are used.
15583 More changes to X509_ATTRIBUTE code: allow the setting of types
15584 based on strings. Remove the 'loc' parameter when adding
15585 attributes because these will be a SET OF encoding which is sorted
15590 * Initial changes to the 'req' utility to allow request generation
15591 automation. This will allow an application to just generate a template
15592 file containing all the field values and have req construct the
15595 Initial support for X509_ATTRIBUTE handling. Stacks of these are
15596 used all over the place including certificate requests and PKCS#7
15597 structures. They are currently handled manually where necessary with
15598 some primitive wrappers for PKCS#7. The new functions behave in a
15599 manner analogous to the X509 extension functions: they allow
15600 attributes to be looked up by NID and added.
15602 Later something similar to the X509V3 code would be desirable to
15603 automatically handle the encoding, decoding and printing of the
15604 more complex types. The string types like challengePassword can
15605 be handled by the string table functions.
15607 Also modified the multi byte string table handling. Now there is
15608 a 'global mask' which masks out certain types. The table itself
15609 can use the flag STABLE_NO_MASK to ignore the mask setting: this
15610 is useful when for example there is only one permissible type
15611 (as in countryName) and using the mask might result in no valid
15616 * Clean up 'Finished' handling, and add functions SSL_get_finished and
15617 SSL_get_peer_finished to allow applications to obtain the latest
15618 Finished messages sent to the peer or expected from the peer,
15619 respectively. (SSL_get_peer_finished is usually the Finished message
15620 actually received from the peer, otherwise the protocol will be aborted.)
15622 As the Finished message are message digests of the complete handshake
15623 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
15624 be used for external authentication procedures when the authentication
15625 provided by SSL/TLS is not desired or is not enough.
15629 * Enhanced support for Alpha Linux is added. Now ./config checks if
15630 the host supports BWX extension and if Compaq C is present on the
15631 $PATH. Just exploiting of the BWX extension results in 20-30%
15632 performance kick for some algorithms, e.g. DES and RC4 to mention
15633 a couple. Compaq C in turn generates ~20% faster code for MD5 and
15638 * Add support for MS "fast SGC". This is arguably a violation of the
15639 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
15640 weak crypto and after checking the certificate is SGC a second one
15641 with strong crypto. MS SGC stops the first handshake after receiving
15642 the server certificate message and sends a second client hello. Since
15643 a server will typically do all the time consuming operations before
15644 expecting any further messages from the client (server key exchange
15645 is the most expensive) there is little difference between the two.
15647 To get OpenSSL to support MS SGC we have to permit a second client
15648 hello message after we have sent server done. In addition we have to
15649 reset the MAC if we do get this second client hello.
15653 * Add a function 'd2i_AutoPrivateKey()' this will automatically decide
15654 if a DER encoded private key is RSA or DSA traditional format. Changed
15655 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
15656 format DER encoded private key. Newer code should use PKCS#8 format which
15657 has the key type encoded in the ASN1 structure. Added DER private key
15658 support to pkcs8 application.
15662 * SSL 3/TLS 1 servers now don't request certificates when an anonymous
15663 ciphersuites has been selected (as required by the SSL 3/TLS 1
15664 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
15665 is set, we interpret this as a request to violate the specification
15666 (the worst that can happen is a handshake failure, and 'correct'
15667 behaviour would result in a handshake failure anyway).
15671 * In SSL_CTX_add_session, take into account that there might be multiple
15672 SSL_SESSION structures with the same session ID (e.g. when two threads
15673 concurrently obtain them from an external cache).
15674 The internal cache can handle only one SSL_SESSION with a given ID,
15675 so if there's a conflict, we now throw out the old one to achieve
15680 * Add OIDs for idea and blowfish in CBC mode. This will allow both
15681 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
15682 some routines that use cipher OIDs: some ciphers do not have OIDs
15683 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
15688 * Simplify the trust setting structure and code. Now we just have
15689 two sequences of OIDs for trusted and rejected settings. These will
15690 typically have values the same as the extended key usage extension
15691 and any application specific purposes.
15693 The trust checking code now has a default behaviour: it will just
15694 check for an object with the same NID as the passed id. Functions can
15695 be provided to override either the default behaviour or the behaviour
15696 for a given id. SSL client, server and email already have functions
15697 in place for compatibility: they check the NID and also return "trusted"
15698 if the certificate is self signed.
15702 * Add d2i,i2d bio/fp functions for PrivateKey: these convert the
15703 traditional format into an EVP_PKEY structure.
15707 * Add a password callback function PEM_cb() which either prompts for
15708 a password if usr_data is NULL or otherwise assumes it is a null
15709 terminated password. Allow passwords to be passed on command line
15710 environment or config files in a few more utilities.
15714 * Add a bunch of DER and PEM functions to handle PKCS#8 format private
15715 keys. Add some short names for PKCS#8 PBE algorithms and allow them
15716 to be specified on the command line for the pkcs8 and pkcs12 utilities.
15717 Update documentation.
15721 * Support for ASN1 "NULL" type. This could be handled before by using
15722 ASN1_TYPE but there wasn't any function that would try to read a NULL
15723 and produce an error if it couldn't. For compatibility we also have
15724 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
15725 don't allocate anything because they don't need to.
15729 * Initial support for MacOS is now provided. Examine INSTALL.MacOS
15732 *Andy Polyakov, Roy Woods <roy@centicsystems.ca>*
15734 * Rebuild of the memory allocation routines used by OpenSSL code and
15735 possibly others as well. The purpose is to make an interface that
15736 provide hooks so anyone can build a separate set of allocation and
15737 deallocation routines to be used by OpenSSL, for example memory
15738 pool implementations, or something else, which was previously hard
15739 since Malloc(), Realloc() and Free() were defined as macros having
15740 the values malloc, realloc and free, respectively (except for Win32
15741 compilations). The same is provided for memory debugging code.
15742 OpenSSL already comes with functionality to find memory leaks, but
15743 this gives people a chance to debug other memory problems.
15745 With these changes, a new set of functions and macros have appeared:
15747 CRYPTO_set_mem_debug_functions() [F]
15748 CRYPTO_get_mem_debug_functions() [F]
15749 CRYPTO_dbg_set_options() [F]
15750 CRYPTO_dbg_get_options() [F]
15751 CRYPTO_malloc_debug_init() [M]
15753 The memory debug functions are NULL by default, unless the library
15754 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
15755 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
15756 gives the standard debugging functions that come with OpenSSL) or
15757 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
15758 provided by the library user) must be used. When the standard
15759 debugging functions are used, CRYPTO_dbg_set_options can be used to
15760 request additional information:
15761 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
15762 the CRYPTO_MDEBUG_xxx macro when compiling the library.
15764 Also, things like CRYPTO_set_mem_functions will always give the
15765 expected result (the new set of functions is used for allocation
15766 and deallocation) at all times, regardless of platform and compiler
15769 To finish it up, some functions that were never use in any other
15770 way than through macros have a new API and new semantic:
15772 CRYPTO_dbg_malloc()
15773 CRYPTO_dbg_realloc()
15776 All macros of value have retained their old syntax.
15778 *Richard Levitte and Bodo Moeller*
15780 * Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
15781 ordering of SMIMECapabilities wasn't in "strength order" and there
15782 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
15787 * Some ASN1 types with illegal zero length encoding (INTEGER,
15788 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
15790 *Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson*
15792 * Merge in my S/MIME library for OpenSSL. This provides a simple
15793 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
15794 functionality to handle multipart/signed properly) and a utility
15795 called 'smime' to call all this stuff. This is based on code I
15796 originally wrote for Celo who have kindly allowed it to be
15797 included in OpenSSL.
15801 * Add variants des_set_key_checked and des_set_key_unchecked of
15802 des_set_key (aka des_key_sched). Global variable des_check_key
15803 decides which of these is called by des_set_key; this way
15804 des_check_key behaves as it always did, but applications and
15805 the library itself, which was buggy for des_check_key == 1,
15806 have a cleaner way to pick the version they need.
15810 * New function PKCS12_newpass() which changes the password of a
15815 * Modify X509_TRUST and X509_PURPOSE so it also uses a static and
15816 dynamic mix. In both cases the ids can be used as an index into the
15817 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
15818 functions so they accept a list of the field values and the
15819 application doesn't need to directly manipulate the X509_TRUST
15824 * Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
15829 * Modify the way the V3 extension code looks up extensions. This now
15830 works in a similar way to the object code: we have some "standard"
15831 extensions in a static table which is searched with OBJ_bsearch()
15832 and the application can add dynamic ones if needed. The file
15833 crypto/x509v3/ext_dat.h now has the info: this file needs to be
15834 updated whenever a new extension is added to the core code and kept
15835 in ext_nid order. There is a simple program 'tabtest.c' which checks
15836 this. New extensions are not added too often so this file can readily
15837 be maintained manually.
15839 There are two big advantages in doing things this way. The extensions
15840 can be looked up immediately and no longer need to be "added" using
15841 X509V3_add_standard_extensions(): this function now does nothing.
15842 Side note: I get *lots* of email saying the extension code doesn't
15843 work because people forget to call this function.
15844 Also no dynamic allocation is done unless new extensions are added:
15845 so if we don't add custom extensions there is no need to call
15846 X509V3_EXT_cleanup().
15850 * Modify enc utility's salting as follows: make salting the default. Add a
15851 magic header, so unsalted files fail gracefully instead of just decrypting
15852 to garbage. This is because not salting is a big security hole, so people
15853 should be discouraged from doing it.
15857 * Fixes and enhancements to the 'x509' utility. It allowed a message
15858 digest to be passed on the command line but it only used this
15859 parameter when signing a certificate. Modified so all relevant
15860 operations are affected by the digest parameter including the
15861 -fingerprint and -x509toreq options. Also -x509toreq choked if a
15862 DSA key was used because it didn't fix the digest.
15866 * Initial certificate chain verify code. Currently tests the untrusted
15867 certificates for consistency with the verify purpose (which is set
15868 when the X509_STORE_CTX structure is set up) and checks the pathlength.
15870 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
15871 this is because it will reject chains with invalid extensions whereas
15872 every previous version of OpenSSL and SSLeay made no checks at all.
15874 Trust code: checks the root CA for the relevant trust settings. Trust
15875 settings have an initial value consistent with the verify purpose: e.g.
15876 if the verify purpose is for SSL client use it expects the CA to be
15877 trusted for SSL client use. However the default value can be changed to
15878 permit custom trust settings: one example of this would be to only trust
15879 certificates from a specific "secure" set of CAs.
15881 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
15882 which should be used for version portability: especially since the
15883 verify structure is likely to change more often now.
15885 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
15886 to set them. If not set then assume SSL clients will verify SSL servers
15889 Two new options to the verify program: -untrusted allows a set of
15890 untrusted certificates to be passed in and -purpose which sets the
15891 intended purpose of the certificate. If a purpose is set then the
15892 new chain verify code is used to check extension consistency.
15896 * Support for the authority information access extension.
15900 * Modify RSA and DSA PEM read routines to transparently handle
15901 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
15902 public keys in a format compatible with certificate
15903 SubjectPublicKeyInfo structures. Unfortunately there were already
15904 functions called *_PublicKey_* which used various odd formats so
15905 these are retained for compatibility: however the DSA variants were
15906 never in a public release so they have been deleted. Changed dsa/rsa
15907 utilities to handle the new format: note no releases ever handled public
15908 keys so we should be OK.
15910 The primary motivation for this change is to avoid the same fiasco
15911 that dogs private keys: there are several incompatible private key
15912 formats some of which are standard and some OpenSSL specific and
15913 require various evil hacks to allow partial transparent handling and
15914 even then it doesn't work with DER formats. Given the option anything
15915 other than PKCS#8 should be dumped: but the other formats have to
15916 stay in the name of compatibility.
15918 With public keys and the benefit of hindsight one standard format
15919 is used which works with EVP_PKEY, RSA or DSA structures: though
15920 it clearly returns an error if you try to read the wrong kind of key.
15922 Added a -pubkey option to the 'x509' utility to output the public key.
15923 Also rename the `EVP_PKEY_get_*()` to `EVP_PKEY_rget_*()`
15924 (renamed to `EVP_PKEY_get1_*()` in the OpenSSL 0.9.5 release) and add
15925 `EVP_PKEY_rset_*()` functions (renamed to `EVP_PKEY_set1_*()`)
15926 that do the same as the `EVP_PKEY_assign_*()` except they up the
15927 reference count of the added key (they don't "swallow" the
15932 * Fixes to crypto/x509/by_file.c the code to read in certificates and
15933 CRLs would fail if the file contained no certificates or no CRLs:
15934 added a new function to read in both types and return the number
15935 read: this means that if none are read it will be an error. The
15936 DER versions of the certificate and CRL reader would always fail
15937 because it isn't possible to mix certificates and CRLs in DER format
15938 without choking one or the other routine. Changed this to just read
15939 a certificate: this is the best we can do. Also modified the code
15940 in apps/verify.c to take notice of return codes: it was previously
15941 attempting to read in certificates from NULL pointers and ignoring
15942 any errors: this is one reason why the cert and CRL reader seemed
15943 to work. It doesn't check return codes from the default certificate
15944 routines: these may well fail if the certificates aren't installed.
15948 * Code to support otherName option in GeneralName.
15952 * First update to verify code. Change the verify utility
15953 so it warns if it is passed a self signed certificate:
15954 for consistency with the normal behaviour. X509_verify
15955 has been modified to it will now verify a self signed
15956 certificate if *exactly* the same certificate appears
15957 in the store: it was previously impossible to trust a
15958 single self signed certificate. This means that:
15959 openssl verify ss.pem
15960 now gives a warning about a self signed certificate but
15961 openssl verify -CAfile ss.pem ss.pem
15966 * For servers, store verify_result in SSL_SESSION data structure
15967 (and add it to external session representation).
15968 This is needed when client certificate verifications fails,
15969 but an application-provided verification callback (set by
15970 SSL_CTX_set_cert_verify_callback) allows accepting the session
15971 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
15972 but returns 1): When the session is reused, we have to set
15973 ssl->verify_result to the appropriate error code to avoid
15976 *Bodo Moeller, problem pointed out by Lutz Jaenicke*
15978 * Fix a bug in the new PKCS#7 code: it didn't consider the
15979 case in PKCS7_dataInit() where the signed PKCS7 structure
15980 didn't contain any existing data because it was being created.
15982 *Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson*
15984 * Add a salt to the key derivation routines in enc.c. This
15985 forms the first 8 bytes of the encrypted file. Also add a
15986 -S option to allow a salt to be input on the command line.
15990 * New function X509_cmp(). Oddly enough there wasn't a function
15991 to compare two certificates. We do this by working out the SHA1
15992 hash and comparing that. X509_cmp() will be needed by the trust
15997 * SSL_get1_session() is like SSL_get_session(), but increments
15998 the reference count in the SSL_SESSION returned.
16000 *Geoff Thorpe <geoff@eu.c2.net>*
16002 * Fix for 'req': it was adding a null to request attributes.
16003 Also change the X509_LOOKUP and X509_INFO code to handle
16004 certificate auxiliary information.
16008 * Add support for 40 and 64 bit RC2 and RC4 algorithms: document
16013 * Add the possibility to add extra information to the memory leak
16014 detecting output, to form tracebacks, showing from where each
16015 allocation was originated: CRYPTO_push_info("constant string") adds
16016 the string plus current file name and line number to a per-thread
16017 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
16018 is like calling CYRPTO_pop_info() until the stack is empty.
16019 Also updated memory leak detection code to be multi-thread-safe.
16023 * Add options -text and -noout to pkcs7 utility and delete the
16024 encryption options which never did anything. Update docs.
16028 * Add options to some of the utilities to allow the pass phrase
16029 to be included on either the command line (not recommended on
16030 OSes like Unix) or read from the environment. Update the
16031 manpages and fix a few bugs.
16035 * Add a few manpages for some of the openssl commands.
16039 * Fix the -revoke option in ca. It was freeing up memory twice,
16040 leaking and not finding already revoked certificates.
16044 * Extensive changes to support certificate auxiliary information.
16045 This involves the use of X509_CERT_AUX structure and X509_AUX
16046 functions. An X509_AUX function such as PEM_read_X509_AUX()
16047 can still read in a certificate file in the usual way but it
16048 will also read in any additional "auxiliary information". By
16049 doing things this way a fair degree of compatibility can be
16050 retained: existing certificates can have this information added
16051 using the new 'x509' options.
16053 Current auxiliary information includes an "alias" and some trust
16054 settings. The trust settings will ultimately be used in enhanced
16055 certificate chain verification routines: currently a certificate
16056 can only be trusted if it is self signed and then it is trusted
16061 * Fix assembler for Alpha (tested only on DEC OSF not Linux or `*BSD`).
16062 The problem was that one of the replacement routines had not been working
16063 since SSLeay releases. For now the offending routine has been replaced
16064 with non-optimised assembler. Even so, this now gives around 95%
16065 performance improvement for 1024 bit RSA signs.
16069 * Hack to fix PKCS#7 decryption when used with some unorthodox RC2
16070 handling. Most clients have the effective key size in bits equal to
16071 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
16072 A few however don't do this and instead use the size of the decrypted key
16073 to determine the RC2 key length and the AlgorithmIdentifier to determine
16074 the effective key length. In this case the effective key length can still
16075 be 40 bits but the key length can be 168 bits for example. This is fixed
16076 by manually forcing an RC2 key into the EVP_PKEY structure because the
16077 EVP code can't currently handle unusual RC2 key sizes: it always assumes
16078 the key length and effective key length are equal.
16082 * Add a bunch of functions that should simplify the creation of
16083 X509_NAME structures. Now you should be able to do:
16084 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
16085 and have it automatically work out the correct field type and fill in
16086 the structures. The more adventurous can try:
16087 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
16088 and it will (hopefully) work out the correct multibyte encoding.
16092 * Change the 'req' utility to use the new field handling and multibyte
16093 copy routines. Before the DN field creation was handled in an ad hoc
16094 way in req, ca, and x509 which was rather broken and didn't support
16095 BMPStrings or UTF8Strings. Since some software doesn't implement
16096 BMPStrings or UTF8Strings yet, they can be enabled using the config file
16097 using the dirstring_type option. See the new comment in the default
16098 openssl.cnf for more info.
16102 * Make crypto/rand/md_rand.c more robust:
16103 - Assure unique random numbers after fork().
16104 - Make sure that concurrent threads access the global counter and
16105 md serializably so that we never lose entropy in them
16106 or use exactly the same state in multiple threads.
16107 Access to the large state is not always serializable because
16108 the additional locking could be a performance killer, and
16109 md should be large enough anyway.
16113 * New file apps/app_rand.c with commonly needed functionality
16114 for handling the random seed file.
16116 Use the random seed file in some applications that previously did not:
16118 dsaparam -genkey (which also ignored its '-rand' option),
16121 x509 (when signing).
16122 Except on systems with /dev/urandom, it is crucial to have a random
16123 seed file at least for key creation, DSA signing, and for DH exchanges;
16124 for RSA signatures we could do without one.
16126 gendh and gendsa (unlike genrsa) used to read only the first byte
16127 of each file listed in the '-rand' option. The function as previously
16128 found in genrsa is now in app_rand.c and is used by all programs
16129 that support '-rand'.
16133 * In RAND_write_file, use mode 0600 for creating files;
16134 don't just chmod when it may be too late.
16138 * Report an error from X509_STORE_load_locations
16139 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
16143 * New function ASN1_mbstring_copy() this copies a string in either
16144 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
16145 into an ASN1_STRING type. A mask of permissible types is passed
16146 and it chooses the "minimal" type to use or an error if not type
16151 * Add function equivalents to the various macros in asn1.h. The old
16152 macros are retained with an `M_` prefix. Code inside the library can
16153 use the `M_` macros. External code (including the openssl utility)
16154 should *NOT* in order to be "shared library friendly".
16158 * Add various functions that can check a certificate's extensions
16159 to see if it usable for various purposes such as SSL client,
16160 server or S/MIME and CAs of these types. This is currently
16161 VERY EXPERIMENTAL but will ultimately be used for certificate chain
16162 verification. Also added a -purpose flag to x509 utility to
16163 print out all the purposes.
16167 * Add a CRYPTO_EX_DATA to X509 certificate structure and associated
16172 * New `X509V3_{X509,CRL,REVOKED}_get_d2i()` functions. These will search
16173 for, obtain and decode and extension and obtain its critical flag.
16174 This allows all the necessary extension code to be handled in a
16175 single function call.
16179 * RC4 tune-up featuring 30-40% performance improvement on most RISC
16180 platforms. See crypto/rc4/rc4_enc.c for further details.
16184 * New -noout option to asn1parse. This causes no output to be produced
16185 its main use is when combined with -strparse and -out to extract data
16186 from a file (which may not be in ASN.1 format).
16190 * Fix for pkcs12 program. It was hashing an invalid certificate pointer
16191 when producing the local key id.
16193 *Richard Levitte <levitte@stacken.kth.se>*
16195 * New option -dhparam in s_server. This allows a DH parameter file to be
16196 stated explicitly. If it is not stated then it tries the first server
16197 certificate file. The previous behaviour hard coded the filename
16202 * Add -pubin and -pubout options to the rsa and dsa commands. These allow
16203 a public key to be input or output. For example:
16204 openssl rsa -in key.pem -pubout -out pubkey.pem
16205 Also added necessary DSA public key functions to handle this.
16209 * Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
16210 in the message. This was handled by allowing
16211 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
16213 *Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>*
16215 * Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
16216 to the end of the strings whereas this didn't. This would cause problems
16217 if strings read with d2i_ASN1_bytes() were later modified.
16219 *Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>*
16221 * Fix for base64 decode bug. When a base64 bio reads only one line of
16222 data and it contains EOF it will end up returning an error. This is
16223 caused by input 46 bytes long. The cause is due to the way base64
16224 BIOs find the start of base64 encoded data. They do this by trying a
16225 trial decode on each line until they find one that works. When they
16226 do a flag is set and it starts again knowing it can pass all the
16227 data directly through the decoder. Unfortunately it doesn't reset
16228 the context it uses. This means that if EOF is reached an attempt
16229 is made to pass two EOFs through the context and this causes the
16230 resulting error. This can also cause other problems as well. As is
16231 usual with these problems it takes *ages* to find and the fix is
16232 trivial: move one line.
16234 *Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer)*
16236 * Ugly workaround to get s_client and s_server working under Windows. The
16237 old code wouldn't work because it needed to select() on sockets and the
16238 tty (for keypresses and to see if data could be written). Win32 only
16239 supports select() on sockets so we select() with a 1s timeout on the
16240 sockets and then see if any characters are waiting to be read, if none
16241 are present then we retry, we also assume we can always write data to
16242 the tty. This isn't nice because the code then blocks until we've
16243 received a complete line of data and it is effectively polling the
16244 keyboard at 1s intervals: however it's quite a bit better than not
16245 working at all :-) A dedicated Windows application might handle this
16246 with an event loop for example.
16250 * Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
16251 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
16252 will be called when RSA_sign() and RSA_verify() are used. This is useful
16253 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
16254 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
16255 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
16256 This necessitated the support of an extra signature type NID_md5_sha1
16257 for SSL signatures and modifications to the SSL library to use it instead
16258 of calling RSA_public_decrypt() and RSA_private_encrypt().
16262 * Add new -verify -CAfile and -CApath options to the crl program, these
16263 will lookup a CRL issuers certificate and verify the signature in a
16264 similar way to the verify program. Tidy up the crl program so it
16265 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
16266 less strict. It will now permit CRL extensions even if it is not
16267 a V2 CRL: this will allow it to tolerate some broken CRLs.
16271 * Initialize all non-automatic variables each time one of the openssl
16272 sub-programs is started (this is necessary as they may be started
16273 multiple times from the "OpenSSL>" prompt).
16275 *Lennart Bang, Bodo Moeller*
16277 * Preliminary compilation option RSA_NULL which disables RSA crypto without
16278 removing all other RSA functionality (this is what NO_RSA does). This
16279 is so (for example) those in the US can disable those operations covered
16280 by the RSA patent while allowing storage and parsing of RSA keys and RSA
16285 * Non-copying interface to BIO pairs.
16286 (still largely untested)
16290 * New function ASN1_tag2str() to convert an ASN1 tag to a descriptive
16291 ASCII string. This was handled independently in various places before.
16295 * New functions UTF8_getc() and UTF8_putc() that parse and generate
16296 UTF8 strings a character at a time.
16300 * Use client_version from client hello to select the protocol
16301 (s23_srvr.c) and for RSA client key exchange verification
16302 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
16306 * Add various utility functions to handle SPKACs, these were previously
16307 handled by poking round in the structure internals. Added new function
16308 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
16309 print, verify and generate SPKACs. Based on an original idea from
16310 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
16314 * RIPEMD160 is operational on all platforms and is back in 'make test'.
16318 * Allow the config file extension section to be overwritten on the
16319 command line. Based on an original idea from Massimiliano Pala
16320 <madwolf@comune.modena.it>. The new option is called -extensions
16321 and can be applied to ca, req and x509. Also -reqexts to override
16322 the request extensions in req and -crlexts to override the crl extensions
16327 * Add new feature to the SPKAC handling in ca. Now you can include
16328 the same field multiple times by preceding it by "XXXX." for example:
16331 this is the same syntax as used in the req config file.
16335 * Allow certificate extensions to be added to certificate requests. These
16336 are specified in a 'req_extensions' option of the req section of the
16337 config file. They can be printed out with the -text option to req but
16338 are otherwise ignored at present.
16342 * Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
16343 data read consists of only the final block it would not decrypted because
16344 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
16345 A misplaced 'break' also meant the decrypted final block might not be
16346 copied until the next read.
16350 * Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
16351 a few extra parameters to the DH structure: these will be useful if
16352 for example we want the value of 'q' or implement X9.42 DH.
16356 * Initial support for DSA_METHOD. This is based on the RSA_METHOD and
16357 provides hooks that allow the default DSA functions or functions on a
16358 "per key" basis to be replaced. This allows hardware acceleration and
16359 hardware key storage to be handled without major modification to the
16360 library. Also added low level modexp hooks and CRYPTO_EX structure and
16361 associated functions.
16365 * Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
16366 as "read only": it can't be written to and the buffer it points to will
16367 not be freed. Reading from a read only BIO is much more efficient than
16368 a normal memory BIO. This was added because there are several times when
16369 an area of memory needs to be read from a BIO. The previous method was
16370 to create a memory BIO and write the data to it, this results in two
16371 copies of the data and an O(n^2) reading algorithm. There is a new
16372 function BIO_new_mem_buf() which creates a read only memory BIO from
16373 an area of memory. Also modified the PKCS#7 routines to use read only
16378 * Bugfix: ssl23_get_client_hello did not work properly when called in
16379 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
16380 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
16381 but a retry condition occurred while trying to read the rest.
16385 * The PKCS7_ENC_CONTENT_new() function was setting the content type as
16386 NID_pkcs7_encrypted by default: this was wrong since this should almost
16387 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
16388 the encrypted data type: this is a more sensible place to put it and it
16389 allows the PKCS#12 code to be tidied up that duplicated this
16394 * Changed obj_dat.pl script so it takes its input and output files on
16395 the command line. This should avoid shell escape redirection problems
16400 * Initial support for certificate extension requests, these are included
16401 in things like Xenroll certificate requests. Included functions to allow
16402 extensions to be obtained and added.
16406 * -crlf option to s_client and s_server for sending newlines as
16407 CRLF (as required by many protocols).
16411 ### Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
16413 * Install libRSAglue.a when OpenSSL is built with RSAref.
16415 *Ralf S. Engelschall*
16417 * A few more `#ifndef NO_FP_API / #endif` pairs for consistency.
16419 *Andrija Antonijevic <TheAntony2@bigfoot.com>*
16421 * Fix -startdate and -enddate (which was missing) arguments to 'ca'
16426 * New function DSA_dup_DH, which duplicates DSA parameters/keys as
16427 DH parameters/keys (q is lost during that conversion, but the resulting
16428 DH parameters contain its length).
16430 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
16431 much faster than DH_generate_parameters (which creates parameters
16432 where `p = 2*q + 1`), and also the smaller q makes DH computations
16433 much more efficient (160-bit exponentiation instead of 1024-bit
16434 exponentiation); so this provides a convenient way to support DHE
16435 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
16436 utter importance to use
16437 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
16439 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
16440 when such DH parameters are used, because otherwise small subgroup
16441 attacks may become possible!
16445 * Avoid memory leak in i2d_DHparams.
16449 * Allow the -k option to be used more than once in the enc program:
16450 this allows the same encrypted message to be read by multiple recipients.
16454 * New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
16455 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
16456 it will always use the numerical form of the OID, even if it has a short
16461 * Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
16462 method only got called if p,q,dmp1,dmq1,iqmp components were present,
16463 otherwise bn_mod_exp was called. In the case of hardware keys for example
16464 no private key components need be present and it might store extra data
16465 in the RSA structure, which cannot be accessed from bn_mod_exp.
16466 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
16467 private key operations.
16471 * Added support for SPARC Linux.
16475 * pem_password_cb function type incompatibly changed from
16476 typedef int pem_password_cb(char *buf, int size, int rwflag);
16478 ....(char *buf, int size, int rwflag, void *userdata);
16479 so that applications can pass data to their callbacks:
16480 The `PEM[_ASN1]_{read,write}...` functions and macros now take an
16481 additional void * argument, which is just handed through whenever
16482 the password callback is called.
16484 *Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller*
16486 New function SSL_CTX_set_default_passwd_cb_userdata.
16488 Compatibility note: As many C implementations push function arguments
16489 onto the stack in reverse order, the new library version is likely to
16490 interoperate with programs that have been compiled with the old
16491 pem_password_cb definition (PEM_whatever takes some data that
16492 happens to be on the stack as its last argument, and the callback
16493 just ignores this garbage); but there is no guarantee whatsoever that
16496 * The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
16497 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
16498 problems not only on Windows, but also on some Unix platforms.
16499 To avoid problematic command lines, these definitions are now in an
16500 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
16501 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
16505 * MIPS III/IV assembler module is reimplemented.
16509 * More DES library cleanups: remove references to srand/rand and
16510 delete an unused file.
16514 * Add support for the free Netwide assembler (NASM) under Win32,
16515 since not many people have MASM (ml) and it can be hard to obtain.
16516 This is currently experimental but it seems to work OK and pass all
16517 the tests. Check out INSTALL.W32 for info.
16521 * Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
16522 without temporary keys kept an extra copy of the server key,
16523 and connections with temporary keys did not free everything in case
16528 * New function RSA_check_key and new openssl rsa option -check
16529 for verifying the consistency of RSA keys.
16531 *Ulf Moeller, Bodo Moeller*
16533 * Various changes to make Win32 compile work:
16534 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
16535 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
16536 comparison" warnings.
16537 3. Add `sk_<TYPE>_sort` to DEF file generator and do make update.
16541 * Add a debugging option to PKCS#5 v2 key generation function: when
16542 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
16543 derived keys are printed to stderr.
16547 * Copy the flags in ASN1_STRING_dup().
16549 *Roman E. Pavlov <pre@mo.msk.ru>*
16551 * The x509 application mishandled signing requests containing DSA
16552 keys when the signing key was also DSA and the parameters didn't match.
16554 It was supposed to omit the parameters when they matched the signing key:
16555 the verifying software was then supposed to automatically use the CA's
16556 parameters if they were absent from the end user certificate.
16558 Omitting parameters is no longer recommended. The test was also
16559 the wrong way round! This was probably due to unusual behaviour in
16560 EVP_cmp_parameters() which returns 1 if the parameters match.
16561 This meant that parameters were omitted when they *didn't* match and
16562 the certificate was useless. Certificates signed with 'ca' didn't have
16565 *Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>*
16567 * Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
16568 The interface is as follows:
16569 Applications can use
16570 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
16571 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
16572 "off" is now the default.
16573 The library internally uses
16574 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
16575 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
16576 to disable memory-checking temporarily.
16578 Some inconsistent states that previously were possible (and were
16579 even the default) are now avoided.
16581 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
16582 with each memory chunk allocated; this is occasionally more helpful
16583 than just having a counter.
16585 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
16587 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
16592 * Introduce "mode" for SSL structures (with defaults in SSL_CTX),
16593 which largely parallels "options", but is for changing API behaviour,
16594 whereas "options" are about protocol behaviour.
16595 Initial "mode" flags are:
16597 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
16598 a single record has been written.
16599 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
16600 retries use the same buffer location.
16601 (But all of the contents must be
16606 * Bugfix: SSL_set_options ignored its parameter, only SSL_CTX_set_options
16609 * Fix problems with no-hmac etc.
16611 *Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>*
16613 * New functions RSA_get_default_method(), RSA_set_method() and
16614 RSA_get_method(). These allows replacement of RSA_METHODs without having
16615 to mess around with the internals of an RSA structure.
16619 * Fix memory leaks in DSA_do_sign and DSA_is_prime.
16620 Also really enable memory leak checks in openssl.c and in some
16623 *Chad C. Mulligan, Bodo Moeller*
16625 * Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
16626 up the length of negative integers. This has now been simplified to just
16627 store the length when it is first determined and use it later, rather
16628 than trying to keep track of where data is copied and updating it to
16630 *Steve Henson, reported by Brien Wheeler <bwheeler@authentica-security.com>*
16632 * Add a new function PKCS7_signatureVerify. This allows the verification
16633 of a PKCS#7 signature but with the signing certificate passed to the
16634 function itself. This contrasts with PKCS7_dataVerify which assumes the
16635 certificate is present in the PKCS#7 structure. This isn't always the
16636 case: certificates can be omitted from a PKCS#7 structure and be
16637 distributed by "out of band" means (such as a certificate database).
16641 * Complete the `PEM_*` macros with DECLARE_PEM versions to replace the
16642 function prototypes in pem.h, also change util/mkdef.pl to add the
16643 necessary function names.
16647 * mk1mf.pl (used by Windows builds) did not properly read the
16648 options set by Configure in the top level Makefile, and Configure
16649 was not even able to write more than one option correctly.
16650 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
16654 * New functions CONF_load_bio() and CONF_load_fp() to allow a config
16655 file to be loaded from a BIO or FILE pointer. The BIO version will
16656 for example allow memory BIOs to contain config info.
16660 * New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
16661 Whoever hopes to achieve shared-library compatibility across versions
16662 must use this, not the compile-time macro.
16663 (Exercise 0.9.4: Which is the minimum library version required by
16665 Note: All this applies only to multi-threaded programs, others don't
16670 * Add missing case to s3_clnt.c state machine -- one of the new SSL tests
16671 through a BIO pair triggered the default case, i.e.
16672 SSLerr(...,SSL_R_UNKNOWN_STATE).
16676 * New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
16677 can use the SSL library even if none of the specific BIOs is
16682 * Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
16683 for the encoded length.
16685 *Jeon KyoungHo <khjeon@sds.samsung.co.kr>*
16687 * Add initial documentation of the X509V3 functions.
16691 * Add a new pair of functions PEM_write_PKCS8PrivateKey() and
16692 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
16693 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
16694 secure PKCS#8 private key format with a high iteration count.
16698 * Fix determination of Perl interpreter: A perl or perl5
16699 *directory* in $PATH was also accepted as the interpreter.
16701 *Ralf S. Engelschall*
16703 * Fix demos/sign/sign.c: well there wasn't anything strictly speaking
16704 wrong with it but it was very old and did things like calling
16705 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
16706 unusual formatting.
16710 * Fix demos/selfsign.c: it used obsolete and deleted functions, changed
16711 to use the new extension code.
16715 * Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
16716 with macros. This should make it easier to change their form, add extra
16717 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
16722 * Add to configuration table a new entry that can specify an alternative
16723 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
16724 according to Mark Crispin <MRC@Panda.COM>.
16728 * DES CBC did not update the IV. Weird.
16732 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
16733 Changing the behaviour of the former might break existing programs --
16734 where IV updating is needed, des_ncbc_encrypt can be used.
16737 * When bntest is run from "make test" it drives bc to check its
16738 calculations, as well as internally checking them. If an internal check
16739 fails, it needs to cause bc to give a non-zero result or make test carries
16740 on without noticing the failure. Fixed.
16744 * DES library cleanups.
16748 * Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
16749 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
16750 ciphers. NOTE: although the key derivation function has been verified
16751 against some published test vectors it has not been extensively tested
16752 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
16757 * Instead of "mkdir -p", which is not fully portable, use new
16758 Perl script "util/mkdir-p.pl".
16762 * Rewrite the way password based encryption (PBE) is handled. It used to
16763 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
16764 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
16765 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
16766 the 'parameter' field of the AlgorithmIdentifier is passed to the
16767 underlying key generation function so it must do its own ASN1 parsing.
16768 This has also changed the EVP_PBE_CipherInit() function which now has a
16769 'parameter' argument instead of literal salt and iteration count values
16770 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
16774 * Support for PKCS#5 v1.5 compatible password based encryption algorithms
16775 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
16776 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
16777 KEY" because this clashed with PKCS#8 unencrypted string. Since this
16778 value was just used as a "magic string" and not used directly its
16779 value doesn't matter.
16783 * Introduce some semblance of const correctness to BN. Shame C doesn't
16788 * "linux-sparc64" configuration (ultrapenguin).
16790 *Ray Miller <ray.miller@oucs.ox.ac.uk>*
16791 "linux-sparc" configuration.
16793 *Christian Forster <fo@hawo.stw.uni-erlangen.de>*
16795 * config now generates no-xxx options for missing ciphers.
16799 * Support the EBCDIC character set (work in progress).
16800 File ebcdic.c not yet included because it has a different license.
16802 *Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>*
16804 * Support BS2000/OSD-POSIX.
16806 *Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>*
16808 * Make callbacks for key generation use `void *` instead of `char *`.
16812 * Make S/MIME samples compile (not yet tested).
16816 * Additional typesafe stacks.
16820 * New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
16824 ### Changes between 0.9.3 and 0.9.3a [29 May 1999]
16826 * New configuration variant "sco5-gcc".
16828 * Updated some demos.
16830 *Sean O Riordain, Wade Scholine*
16832 * Add missing BIO_free at exit of pkcs12 application.
16836 * Fix memory leak in conf.c.
16840 * Updates for Win32 to assembler version of MD5.
16844 * Set #! path to perl in apps/der_chop to where we found it
16845 instead of using a fixed path.
16849 * SHA library changes for irix64-mips4-cc.
16853 * Improvements for VMS support.
16857 ### Changes between 0.9.2b and 0.9.3 [24 May 1999]
16859 * Bignum library bug fix. IRIX 6 passes "make test" now!
16860 This also avoids the problems with SC4.2 and unpatched SC5.
16862 *Andy Polyakov <appro@fy.chalmers.se>*
16864 * New functions sk_num, sk_value and sk_set to replace the previous macros.
16865 These are required because of the typesafe stack would otherwise break
16866 existing code. If old code used a structure member which used to be STACK
16867 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
16868 sk_num or sk_value it would produce an error because the num, data members
16869 are not present in STACK_OF. Now it just produces a warning. sk_set
16870 replaces the old method of assigning a value to sk_value
16871 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
16872 that does this will no longer work (and should use sk_set instead) but
16873 this could be regarded as a "questionable" behaviour anyway.
16877 * Fix most of the other PKCS#7 bugs. The "experimental" code can now
16878 correctly handle encrypted S/MIME data.
16882 * Change type of various DES function arguments from des_cblock
16883 (which means, in function argument declarations, pointer to char)
16884 to des_cblock * (meaning pointer to array with 8 char elements),
16885 which allows the compiler to do more typechecking; it was like
16886 that back in SSLeay, but with lots of ugly casts.
16888 Introduce new type const_des_cblock.
16892 * Reorganise the PKCS#7 library and get rid of some of the more obvious
16893 problems: find RecipientInfo structure that matches recipient certificate
16894 and initialise the ASN1 structures properly based on passed cipher.
16898 * Belatedly make the BN tests actually check the results.
16902 * Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
16903 to and from BNs: it was completely broken. New compilation option
16904 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
16905 key elements as negative integers.
16909 * Reorganize and speed up MD5.
16911 *Andy Polyakov <appro@fy.chalmers.se>*
16915 *Richard Levitte <richard@levitte.org>*
16917 * New option -out to asn1parse to allow the parsed structure to be
16918 output to a file. This is most useful when combined with the -strparse
16919 option to examine the output of things like OCTET STRINGS.
16923 * Make SSL library a little more fool-proof by not requiring any longer
16924 that `SSL_set_{accept,connect}_state` be called before
16925 `SSL_{accept,connect}` may be used (`SSL_set_..._state` is omitted
16926 in many applications because usually everything *appeared* to work as
16927 intended anyway -- now it really works as intended).
16931 * Move openssl.cnf out of lib/.
16935 * Fix various things to let OpenSSL even pass "egcc -pipe -O2 -Wall
16936 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
16937 -Wmissing-declarations -Wnested-externs -Winline" with EGCS 1.1.2+
16939 *Ralf S. Engelschall*
16941 * Various fixes to the EVP and PKCS#7 code. It may now be able to
16942 handle PKCS#7 enveloped data properly.
16944 *Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve*
16946 * Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
16947 copying pointers. The cert_st handling is changed by this in
16948 various ways (and thus what used to be known as ctx->default_cert
16949 is now called ctx->cert, since we don't resort to `s->ctx->[default_]cert`
16950 any longer when s->cert does not give us what we need).
16951 ssl_cert_instantiate becomes obsolete by this change.
16952 As soon as we've got the new code right (possibly it already is?),
16953 we have solved a couple of bugs of the earlier code where s->cert
16954 was used as if it could not have been shared with other SSL structures.
16956 Note that using the SSL API in certain dirty ways now will result
16957 in different behaviour than observed with earlier library versions:
16958 Changing settings for an `SSL_CTX *ctx` after having done s = SSL_new(ctx)
16959 does not influence s as it used to.
16961 In order to clean up things more thoroughly, inside SSL_SESSION
16962 we don't use CERT any longer, but a new structure SESS_CERT
16963 that holds per-session data (if available); currently, this is
16964 the peer's certificate chain and, for clients, the server's certificate
16965 and temporary key. CERT holds only those values that can have
16966 meaningful defaults in an SSL_CTX.
16970 * New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
16971 from the internal representation. Various PKCS#7 fixes: remove some
16972 evil casts and set the enc_dig_alg field properly based on the signing
16977 * Allow PKCS#12 password to be set from the command line or the
16978 environment. Let 'ca' get its config file name from the environment
16979 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
16984 * Allow certificate policies extension to use an IA5STRING for the
16985 organization field. This is contrary to the PKIX definition but
16986 VeriSign uses it and IE5 only recognises this form. Document 'x509'
16991 * Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
16992 without disallowing inline assembler and the like for non-pedantic builds.
16996 * Support Borland C++ builder.
16998 *Janez Jere <jj@void.si>, modified by Ulf Möller*
17004 * SHA-1 cleanups and performance enhancements.
17006 *Andy Polyakov <appro@fy.chalmers.se>*
17008 * Sparc v8plus assembler for the bignum library.
17010 *Andy Polyakov <appro@fy.chalmers.se>*
17012 * Accept any -xxx and +xxx compiler options in Configure.
17016 * Update HPUX configuration.
17020 * Add missing `sk_<type>_unshift()` function to safestack.h
17022 *Ralf S. Engelschall*
17024 * New function SSL_CTX_use_certificate_chain_file that sets the
17025 "extra_cert"s in addition to the certificate. (This makes sense
17026 only for "PEM" format files, as chains as a whole are not
17031 * Support verify_depth from the SSL API.
17032 x509_vfy.c had what can be considered an off-by-one-error:
17033 Its depth (which was not part of the external interface)
17034 was actually counting the number of certificates in a chain;
17035 now it really counts the depth.
17039 * Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
17040 instead of X509err, which often resulted in confusing error
17041 messages since the error codes are not globally unique
17042 (e.g. an alleged error in ssl3_accept when a certificate
17043 didn't match the private key).
17045 * New function SSL_CTX_set_session_id_context that allows to set a default
17046 value (so that you don't need SSL_set_session_id_context for each
17047 connection using the SSL_CTX).
17051 * OAEP decoding bug fix.
17055 * Support INSTALL_PREFIX for package builders, as proposed by
17060 * New Configure options "threads" and "no-threads". For systems
17061 where the proper compiler options are known (currently Solaris
17062 and Linux), "threads" is the default.
17066 * New script util/mklink.pl as a faster substitute for util/mklink.sh.
17070 * Install various scripts to $(OPENSSLDIR)/misc, not to
17071 $(INSTALLTOP)/bin -- they shouldn't clutter directories
17072 such as /usr/local/bin.
17076 * "make linux-shared" to build shared libraries.
17078 *Niels Poppe <niels@netbox.org>*
17080 * New Configure option `no-<cipher>` (rsa, idea, rc5, ...).
17084 * Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
17085 extension adding in x509 utility.
17089 * Remove NOPROTO sections and error code comments.
17093 * Partial rewrite of the DEF file generator to now parse the ANSI
17098 * New Configure options --prefix=DIR and --openssldir=DIR.
17102 * Complete rewrite of the error code script(s). It is all now handled
17103 by one script at the top level which handles error code gathering,
17104 header rewriting and C source file generation. It should be much better
17105 than the old method: it now uses a modified version of Ulf's parser to
17106 read the ANSI prototypes in all header files (thus the old K&R definitions
17107 aren't needed for error creation any more) and do a better job of
17108 translating function codes into names. The old 'ASN1 error code embedded
17109 in a comment' is no longer necessary and it doesn't use .err files which
17110 have now been deleted. Also the error code call doesn't have to appear all
17111 on one line (which resulted in some large lines...).
17115 * Change #include filenames from `<foo.h>` to `<openssl/foo.h>`.
17119 * Change behaviour of ssl2_read when facing length-0 packets: Don't return
17120 0 (which usually indicates a closed connection), but continue reading.
17124 * Fix some race conditions.
17128 * Add support for CRL distribution points extension. Add Certificate
17129 Policies and CRL distribution points documentation.
17133 * Move the autogenerated header file parts to crypto/opensslconf.h.
17137 * Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
17138 8 of keying material. Merlin has also confirmed interop with this fix
17139 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
17141 *Merlin Hughes <merlin@baltimore.ie>*
17143 * Fix lots of warnings.
17145 *Richard Levitte <levitte@stacken.kth.se>*
17147 * In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
17148 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
17150 *Richard Levitte <levitte@stacken.kth.se>*
17152 * Fix problems with sizeof(long) == 8.
17154 *Andy Polyakov <appro@fy.chalmers.se>*
17156 * Change functions to ANSI C.
17160 * Fix typos in error codes.
17162 *Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller*
17164 * Remove defunct assembler files from Configure.
17168 * SPARC v8 assembler BIGNUM implementation.
17170 *Andy Polyakov <appro@fy.chalmers.se>*
17172 * Support for Certificate Policies extension: both print and set.
17173 Various additions to support the r2i method this uses.
17177 * A lot of constification, and fix a bug in X509_NAME_oneline() that could
17178 return a const string when you are expecting an allocated buffer.
17182 * Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
17183 types DirectoryString and DisplayText.
17187 * Add code to allow r2i extensions to access the configuration database,
17188 add an LHASH database driver and add several ctx helper functions.
17192 * Fix an evil bug in bn_expand2() which caused various BN functions to
17193 fail when they extended the size of a BIGNUM.
17197 * Various utility functions to handle SXNet extension. Modify mkdef.pl to
17198 support typesafe stack.
17202 * Fix typo in SSL_[gs]et_options().
17204 *Nils Frostberg <nils@medcom.se>*
17206 * Delete various functions and files that belonged to the (now obsolete)
17207 old X509V3 handling code.
17211 * New Configure option "rsaref".
17215 * Don't auto-generate pem.h.
17219 * Introduce type-safe ASN.1 SETs.
17223 * Convert various additional casted stacks to type-safe STACK_OF() variants.
17225 *Ben Laurie, Ralf S. Engelschall, Steve Henson*
17227 * Introduce type-safe STACKs. This will almost certainly break lots of code
17228 that links with OpenSSL (well at least cause lots of warnings), but fear
17229 not: the conversion is trivial, and it eliminates loads of evil casts. A
17230 few STACKed things have been converted already. Feel free to convert more.
17231 In the fullness of time, I'll do away with the STACK type altogether.
17235 * Add `openssl ca -revoke <certfile>` facility which revokes a certificate
17236 specified in `<certfile>` by updating the entry in the index.txt file.
17237 This way one no longer has to edit the index.txt file manually for
17238 revoking a certificate. The -revoke option does the gory details now.
17240 *Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall*
17242 * Fix `openssl crl -noout -text` combination where `-noout` killed the
17243 `-text` option at all and this way the `-noout -text` combination was
17244 inconsistent in `openssl crl` with the friends in `openssl x509|rsa|dsa`.
17246 *Ralf S. Engelschall*
17248 * Make sure a corresponding plain text error message exists for the
17249 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
17250 verify callback function determined that a certificate was revoked.
17252 *Ralf S. Engelschall*
17254 * Bugfix: In test/testenc, don't test `openssl <cipher>` for
17255 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
17256 all available ciphers including rc5, which was forgotten until now.
17257 In order to let the testing shell script know which algorithms
17258 are available, a new (up to now undocumented) command
17259 `openssl list-cipher-commands` is used.
17263 * Bugfix: s_client occasionally would sleep in select() when
17264 it should have checked SSL_pending() first.
17268 * New functions DSA_do_sign and DSA_do_verify to provide access to
17269 the raw DSA values prior to ASN.1 encoding.
17273 * Tweaks to Configure
17275 *Niels Poppe <niels@netbox.org>*
17277 * Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
17282 * New variables $(RANLIB) and $(PERL) in the Makefiles.
17286 * New config option to avoid instructions that are illegal on the 80386.
17287 The default code is faster, but requires at least a 486.
17291 * Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
17292 SSL2_SERVER_VERSION (not used at all) macros, which are now the
17293 same as SSL2_VERSION anyway.
17297 * New "-showcerts" option for s_client.
17301 * Still more PKCS#12 integration. Add pkcs12 application to openssl
17302 application. Various cleanups and fixes.
17306 * More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
17307 modify error routines to work internally. Add error codes and PBE init
17308 to library startup routines.
17312 * Further PKCS#12 integration. Added password based encryption, PKCS#8 and
17313 packing functions to asn1 and evp. Changed function names and error
17314 codes along the way.
17318 * PKCS12 integration: and so it begins... First of several patches to
17319 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
17320 objects to objects.h
17324 * Add a new 'indent' option to some X509V3 extension code. Initial ASN1
17325 and display support for Thawte strong extranet extension.
17329 * Add LinuxPPC support.
17331 *Jeff Dubrule <igor@pobox.org>*
17333 * Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
17334 bn_div_words in alpha.s.
17336 *Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie*
17338 * Make sure the RSA OAEP test is skipped under -DRSAref because
17339 OAEP isn't supported when OpenSSL is built with RSAref.
17341 *Ulf Moeller <ulf@fitug.de>*
17343 * Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
17344 so they no longer are missing under -DNOPROTO.
17346 *Soren S. Jorvang <soren@t.dk>*
17348 ### Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
17350 * Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
17351 doesn't work when the session is reused. Coming soon!
17355 * Fix a security hole, that allows sessions to be reused in the wrong
17356 context thus bypassing client cert protection! All software that uses
17357 client certs and session caches in multiple contexts NEEDS PATCHING to
17358 allow session reuse! A fuller solution is in the works.
17360 *Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)*
17362 * Some more source tree cleanups (removed obsolete files
17363 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
17364 permission on "config" script to be executable) and a fix for the INSTALL
17367 *Ulf Moeller <ulf@fitug.de>*
17369 * Remove some legacy and erroneous uses of malloc, free instead of
17372 *Lennart Bang <lob@netstream.se>, with minor changes by Steve*
17374 * Make rsa_oaep_test return non-zero on error.
17376 *Ulf Moeller <ulf@fitug.de>*
17378 * Add support for native Solaris shared libraries. Configure
17379 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
17380 if someone would make that last step automatic.
17382 *Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>*
17384 * ctx_size was not built with the right compiler during "make links". Fixed.
17388 * Change the meaning of 'ALL' in the cipher list. It now means "everything
17389 except NULL ciphers". This means the default cipher list will no longer
17390 enable NULL ciphers. They need to be specifically enabled e.g. with
17391 the string "DEFAULT:eNULL".
17395 * Fix to RSA private encryption routines: if p < q then it would
17396 occasionally produce an invalid result. This will only happen with
17397 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
17401 * Be less restrictive and allow also `perl util/perlpath.pl
17402 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin',
17403 because this way one can also use an interpreter named `perl5' (which is
17404 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
17405 installed as `perl').
17407 *Matthias Loepfe <Matthias.Loepfe@adnovum.ch>*
17409 * Let util/clean-depend.pl work also with older Perl 5.00x versions.
17411 *Matthias Loepfe <Matthias.Loepfe@adnovum.ch>*
17413 * Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
17414 advapi32.lib to Win32 build and change the pem test comparison
17415 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
17416 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
17417 and crypto/des/ede_cbcm_enc.c.
17421 * DES quad checksum was broken on big-endian architectures. Fixed.
17425 * Comment out two functions in bio.h that aren't implemented. Fix up the
17426 Win32 test batch file so it (might) work again. The Win32 test batch file
17427 is horrible: I feel ill....
17431 * Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
17432 in e_os.h. Audit of header files to check ANSI and non ANSI
17433 sections: 10 functions were absent from non ANSI section and not exported
17434 from Windows DLLs. Fixed up libeay.num for new functions.
17438 * Make `openssl version' output lines consistent.
17440 *Ralf S. Engelschall*
17442 * Fix Win32 symbol export lists for BIO functions: Added
17443 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
17444 to ms/libeay{16,32}.def.
17446 *Ralf S. Engelschall*
17448 * Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
17449 fine under Unix and passes some trivial tests I've now added. But the
17450 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
17451 added to make sure no one expects that this stuff really works in the
17452 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
17453 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
17456 *Ralf S. Engelschall*
17458 * Fix the generation of two part addresses in perl.
17460 *Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie*
17462 * Add config entry for Linux on MIPS.
17464 *John Tobey <jtobey@channel1.com>*
17466 * Make links whenever Configure is run, unless we are on Windoze.
17470 * Permit extensions to be added to CRLs using crl_section in openssl.cnf.
17471 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
17476 * Add a useful kludge to allow package maintainers to specify compiler and
17477 other platforms details on the command line without having to patch the
17478 Configure script every time: One now can use
17479 `perl Configure <id>:<details>`,
17480 i.e. platform ids are allowed to have details appended
17481 to them (separated by colons). This is treated as there would be a static
17482 pre-configured entry in Configure's %table under key `<id>` with value
17483 `<details>` and `perl Configure <id>` is called. So, when you want to
17484 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
17485 assembler stuff you can use `perl Configure "FreeBSD-elf:pgcc:-O6:::"`
17486 now, which overrides the FreeBSD-elf entry on-the-fly.
17488 *Ralf S. Engelschall*
17490 * Disable new TLS1 ciphersuites by default: they aren't official yet.
17494 * Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
17495 on the `perl Configure ...' command line. This way one can compile
17496 OpenSSL libraries with Position Independent Code (PIC) which is needed
17497 for linking it into DSOs.
17499 *Ralf S. Engelschall*
17501 * Remarkably, export ciphers were totally broken and no-one had noticed!
17506 * Cleaned up the LICENSE document: The official contact for any license
17507 questions now is the OpenSSL core team under openssl-core@openssl.org.
17508 And add a paragraph about the dual-license situation to make sure people
17509 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
17510 to the OpenSSL toolkit.
17512 *Ralf S. Engelschall*
17514 * General source tree makefile cleanups: Made `making xxx in yyy...'
17515 display consistent in the source tree and replaced `/bin/rm' by `rm'.
17516 Additionally cleaned up the `make links' target: Remove unnecessary
17517 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
17518 to speed processing and no longer clutter the display with confusing
17519 stuff. Instead only the actually done links are displayed.
17521 *Ralf S. Engelschall*
17523 * Permit null encryption ciphersuites, used for authentication only. It used
17524 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
17525 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
17530 * Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
17531 signed attributes when verifying signatures (this would break them),
17532 the detached data encoding was wrong and public keys obtained using
17533 X509_get_pubkey() weren't freed.
17537 * Add text documentation for the BUFFER functions. Also added a work around
17538 to a Win95 console bug. This was triggered by the password read stuff: the
17539 last character typed gets carried over to the next fread(). If you were
17540 generating a new cert request using 'req' for example then the last
17541 character of the passphrase would be CR which would then enter the first
17546 * Added the new 'Includes OpenSSL Cryptography Software' button as
17547 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
17548 button and can be used by applications based on OpenSSL to show the
17549 relationship to the OpenSSL project.
17551 *Ralf S. Engelschall*
17553 * Remove confusing variables in function signatures in files
17554 ssl/ssl_lib.c and ssl/ssl.h.
17556 *Lennart Bong <lob@kulthea.stacken.kth.se>*
17558 * Don't install bss_file.c under PREFIX/include/
17560 *Lennart Bong <lob@kulthea.stacken.kth.se>*
17562 * Get the Win32 compile working again. Modify mkdef.pl so it can handle
17563 functions that return function pointers and has support for NT specific
17564 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
17565 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
17566 unsigned to signed types: this was killing the Win32 compile.
17570 * Add new certificate file to stack functions,
17571 SSL_add_dir_cert_subjects_to_stack() and
17572 SSL_add_file_cert_subjects_to_stack(). These largely supplant
17573 SSL_load_client_CA_file(), and can be used to add multiple certs easily
17574 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
17575 This means that Apache-SSL and similar packages don't have to mess around
17576 to add as many CAs as they want to the preferred list.
17580 * Experiment with doxygen documentation. Currently only partially applied to
17582 See <http://www.stack.nl/~dimitri/doxygen/index.html>, and run doxygen with
17583 openssl.doxy as the configuration file.
17587 * Get rid of remaining C++-style comments which strict C compilers hate.
17589 *Ralf S. Engelschall, pointed out by Carlos Amengual*
17591 * Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
17592 compiled in by default: it has problems with large keys.
17596 * Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
17597 DH private keys and/or callback functions which directly correspond to
17598 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
17599 is needed for applications which have to configure certificates on a
17600 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
17602 For the RSA certificate situation is makes no difference, but
17603 for the DSA certificate situation this fixes the "no shared cipher"
17604 problem where the OpenSSL cipher selection procedure failed because the
17605 temporary keys were not overtaken from the context and the API provided
17606 no way to reconfigure them.
17607 The new functions now let applications reconfigure the stuff and they
17608 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
17609 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
17610 non-public-API function ssl_cert_instantiate() is used as a helper
17611 function and also to reduce code redundancy inside ssl_rsa.c.
17613 *Ralf S. Engelschall*
17615 * Move s_server -dcert and -dkey options out of the undocumented feature
17616 area because they are useful for the DSA situation and should be
17617 recognized by the users.
17619 *Ralf S. Engelschall*
17621 * Fix the cipher decision scheme for export ciphers: the export bits are
17622 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
17623 SSL_EXP_MASK. So, the original variable has to be used instead of the
17624 already masked variable.
17626 *Richard Levitte <levitte@stacken.kth.se>*
17628 * Fix `port` variable from `int` to `unsigned int` in crypto/bio/b_sock.c
17630 *Richard Levitte <levitte@stacken.kth.se>*
17632 * Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
17633 from `int` to `unsigned int` because it is a length and initialized by
17634 EVP_DigestFinal() which expects an `unsigned int *`.
17636 *Richard Levitte <levitte@stacken.kth.se>*
17638 * Don't hard-code path to Perl interpreter on shebang line of Configure
17639 script. Instead use the usual Shell->Perl transition trick.
17641 *Ralf S. Engelschall*
17643 * Make `openssl x509 -noout -modulus' functional also for DSA certificates
17644 (in addition to RSA certificates) to match the behaviour of `openssl dsa
17645 -noout -modulus' as it's already the case for `openssl rsa -noout
17646 -modulus'. For RSA the -modulus is the real "modulus" while for DSA
17647 currently the public key is printed (a decision which was already done by
17648 `openssl dsa -modulus' in the past) which serves a similar purpose.
17649 Additionally the NO_RSA no longer completely removes the whole -modulus
17650 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
17653 *Ralf S. Engelschall*
17655 * Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
17656 BIO. See the source (crypto/evp/bio_ok.c) for more info.
17658 *Arne Ansper <arne@ats.cyber.ee>*
17660 * Dump the old yucky req code that tried (and failed) to allow raw OIDs
17661 to be added. Now both 'req' and 'ca' can use new objects defined in the
17666 * Add cool BIO that does syslog (or event log on NT).
17668 *Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie*
17670 * Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
17671 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
17672 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
17673 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
17677 * Add preliminary config info for new extension code.
17681 * Make RSA_NO_PADDING really use no padding.
17683 *Ulf Moeller <ulf@fitug.de>*
17685 * Generate errors when private/public key check is done.
17689 * Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
17690 for some CRL extensions and new objects added.
17694 * Really fix the ASN1 IMPLICIT bug this time... Partial support for private
17695 key usage extension and fuller support for authority key id.
17699 * Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
17700 padding method for RSA, which is recommended for new applications in PKCS
17701 #1 v2.0 (RFC 2437, October 1998).
17702 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
17703 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
17704 against Bleichbacher's attack on RSA.
17705 *Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
17708 * Updates to the new SSL compression code
17710 *Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)*
17712 * Fix so that the version number in the master secret, when passed
17713 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
17714 (because the server will not accept higher), that the version number
17715 is 0x03,0x01, not 0x03,0x00
17717 *Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)*
17719 * Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
17720 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
17721 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
17725 * Support for RAW extensions where an arbitrary extension can be
17726 created by including its DER encoding. See apps/openssl.cnf for
17731 * Make sure latest Perl versions don't interpret some generated C array
17732 code as Perl array code in the crypto/err/err_genc.pl script.
17734 *Lars Weber <3weber@informatik.uni-hamburg.de>*
17736 * Modify ms/do_ms.bat to not generate assembly language makefiles since
17737 not many people have the assembler. Various Win32 compilation fixes and
17738 update to the INSTALL.W32 file with (hopefully) more accurate Win32
17739 build instructions.
17743 * Modify configure script 'Configure' to automatically create crypto/date.h
17744 file under Win32 and also build pem.h from pem.org. New script
17745 util/mkfiles.pl to create the MINFO file on environments that can't do a
17746 'make files': perl util/mkfiles.pl >MINFO should work.
17750 * Major rework of DES function declarations, in the pursuit of correctness
17751 and purity. As a result, many evil casts evaporated, and some weirdness,
17752 too. You may find this causes warnings in your code. Zapping your evil
17753 casts will probably fix them. Mostly.
17757 * Fix for a typo in asn1.h. Bug fix to object creation script
17758 obj_dat.pl. It considered a zero in an object definition to mean
17759 "end of object": none of the objects in objects.h have any zeros
17760 so it wasn't spotted.
17762 *Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>*
17764 * Add support for Triple DES Cipher Block Chaining with Output Feedback
17765 Masking (CBCM). In the absence of test vectors, the best I have been able
17766 to do is check that the decrypt undoes the encrypt, so far. Send me test
17767 vectors if you have them.
17771 * Correct calculation of key length for export ciphers (too much space was
17772 allocated for null ciphers). This has not been tested!
17776 * Modifications to the mkdef.pl for Win32 DEF file creation. The usage
17777 message is now correct (it understands "crypto" and "ssl" on its
17778 command line). There is also now an "update" option. This will update
17779 the util/ssleay.num and util/libeay.num files with any new functions.
17781 perl util/mkdef.pl crypto ssl update
17782 it will update them.
17786 * Overhauled the Perl interface:
17787 - ported BN stuff to OpenSSL's different BN library
17788 - made the perl/ source tree CVS-aware
17789 - renamed the package from SSLeay to OpenSSL (the files still contain
17790 their history because I've copied them in the repository)
17791 - removed obsolete files (the test scripts will be replaced
17792 by better Test::Harness variants in the future)
17794 *Ralf S. Engelschall*
17796 * First cut for a very conservative source tree cleanup:
17797 1. merge various obsolete readme texts into doc/ssleay.txt
17798 where we collect the old documents and readme texts.
17799 2. remove the first part of files where I'm already sure that we no
17800 longer need them because of three reasons: either they are just temporary
17801 files which were left by Eric or they are preserved original files where
17802 I've verified that the diff is also available in the CVS via "cvs diff
17803 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
17804 the crypto/md/ stuff).
17806 *Ralf S. Engelschall*
17808 * More extension code. Incomplete support for subject and issuer alt
17809 name, issuer and authority key id. Change the i2v function parameters
17810 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
17811 what that's for :-) Fix to ASN1 macro which messed up
17812 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
17816 * Preliminary support for ENUMERATED type. This is largely copied from the
17821 * Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
17823 *Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)*
17825 * Make sure `make rehash` target really finds the `openssl` program.
17827 *Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>*
17829 * Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
17830 like to hear about it if this slows down other processors.
17834 * Add CygWin32 platform information to Configure script.
17836 *Alan Batie <batie@aahz.jf.intel.com>*
17838 * Fixed ms/32all.bat script: `no_asm` -> `no-asm`
17840 *Rainer W. Gerling <gerling@mpg-gv.mpg.de>*
17842 * New program nseq to manipulate netscape certificate sequences
17846 * Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
17851 * Fixes to BN code. Previously the default was to define BN_RECURSION
17852 but the BN code had some problems that would cause failures when
17853 doing certificate verification and some other functions.
17855 *Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)*
17857 * Add ASN1 and PEM code to support netscape certificate sequences.
17861 * Add ASN1 and PEM code to support netscape certificate sequences.
17865 * Add several PKIX and private extended key usage OIDs.
17869 * Modify the 'ca' program to handle the new extension code. Modify
17870 openssl.cnf for new extension format, add comments.
17874 * More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
17875 and add a sample to openssl.cnf so req -x509 now adds appropriate
17880 * Continued X509 V3 changes. Add to other makefiles, integrate with the
17881 error code, add initial support to X509_print() and x509 application.
17885 * Takes a deep breath and start adding X509 V3 extension support code. Add
17886 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
17887 stuff is currently isolated and isn't even compiled yet.
17891 * Continuing patches for GeneralizedTime. Fix up certificate and CRL
17892 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
17893 Removed the versions check from X509 routines when loading extensions:
17894 this allows certain broken certificates that don't set the version
17895 properly to be processed.
17899 * Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
17900 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
17901 can still be regenerated with "make depend".
17905 * Spelling mistake in C version of CAST-128.
17907 *Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>*
17909 * Changes to the error generation code. The perl script err-code.pl
17910 now reads in the old error codes and retains the old numbers, only
17911 adding new ones if necessary. It also only changes the .err files if new
17912 codes are added. The makefiles have been modified to only insert errors
17913 when needed (to avoid needlessly modifying header files). This is done
17914 by only inserting errors if the .err file is newer than the auto generated
17915 C file. To rebuild all the error codes from scratch (the old behaviour)
17916 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
17917 or delete all the .err files.
17921 * CAST-128 was incorrectly implemented for short keys. The C version has
17922 been fixed, but is untested. The assembler versions are also fixed, but
17923 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
17924 to regenerate it if needed.
17925 *Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
17926 Hagino <itojun@kame.net>*
17928 * File was opened incorrectly in randfile.c.
17930 *Ulf Möller <ulf@fitug.de>*
17932 * Beginning of support for GeneralizedTime. d2i, i2d, check and print
17933 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
17934 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
17935 al: it's just almost always a UTCTime. Note this patch adds new error
17936 codes so do a "make errors" if there are problems.
17940 * Correct Linux 1 recognition in config.
17942 *Ulf Möller <ulf@fitug.de>*
17944 * Remove pointless MD5 hash when using DSA keys in ca.
17946 *Anonymous <nobody@replay.com>*
17948 * Generate an error if given an empty string as a cert directory. Also
17949 generate an error if handed NULL (previously returned 0 to indicate an
17950 error, but didn't set one).
17952 *Ben Laurie, reported by Anonymous <nobody@replay.com>*
17954 * Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
17958 * Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
17959 parameters. This was causing a warning which killed off the Win32 compile.
17963 * Remove C++ style comments from crypto/bn/bn_local.h.
17965 *Neil Costigan <neil.costigan@celocom.com>*
17967 * The function OBJ_txt2nid was broken. It was supposed to return a nid
17968 based on a text string, looking up short and long names and finally
17969 "dot" format. The "dot" format stuff didn't work. Added new function
17970 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
17971 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
17972 OID is not part of the table.
17976 * Add prototypes to X509 lookup/verify methods, fixing a bug in
17977 X509_LOOKUP_by_alias().
17981 * Sort openssl functions by name.
17985 * Get the gendsa program working (hopefully) and add it to app list. Remove
17986 encryption from sample DSA keys (in case anyone is interested the password
17991 * Make *all* `*_free` functions accept a NULL pointer.
17993 *Frans Heymans <fheymans@isaserver.be>*
17995 * If a DH key is generated in s3_srvr.c, don't blow it by trying to use
17998 *Anonymous <nobody@replay.com>*
18000 * s_server should send the CAfile as acceptable CAs, not its own cert.
18002 *Bodo Moeller <3moeller@informatik.uni-hamburg.de>*
18004 * Don't blow it for numeric -newkey arguments to apps/req.
18006 *Bodo Moeller <3moeller@informatik.uni-hamburg.de>*
18008 * Temp key "for export" tests were wrong in s3_srvr.c.
18010 *Anonymous <nobody@replay.com>*
18012 * Add prototype for temp key callback functions
18013 SSL_CTX_set_tmp_{rsa,dh}_callback().
18017 * Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
18018 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
18022 * X509_name_add_entry() freed the wrong thing after an error.
18024 *Arne Ansper <arne@ats.cyber.ee>*
18026 * rsa_eay.c would attempt to free a NULL context.
18028 *Arne Ansper <arne@ats.cyber.ee>*
18030 * BIO_s_socket() had a broken should_retry() on Windoze.
18032 *Arne Ansper <arne@ats.cyber.ee>*
18034 * BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
18036 *Arne Ansper <arne@ats.cyber.ee>*
18038 * Make sure the already existing X509_STORE->depth variable is initialized
18039 in X509_STORE_new(), but document the fact that this variable is still
18040 unused in the certificate verification process.
18042 *Ralf S. Engelschall*
18044 * Fix the various library and apps files to free up pkeys obtained from
18045 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
18049 * Fix reference counting in X509_PUBKEY_get(). This makes
18050 demos/maurice/example2.c work, amongst others, probably.
18052 *Steve Henson and Ben Laurie*
18054 * First cut of a cleanup for apps/. First the `ssleay` program is now named
18055 `openssl` and second, the shortcut symlinks for the `openssl <command>`
18056 are no longer created. This way we have a single and consistent command
18057 line interface `openssl <command>`, similar to `cvs <command>`.
18059 *Ralf S. Engelschall, Paul Sutton and Ben Laurie*
18061 * ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
18062 BIT STRING wrapper always have zero unused bits.
18066 * Add CA.pl, perl version of CA.sh, add extended key usage OID.
18070 * Make the top-level INSTALL documentation easier to understand.
18074 * Makefiles updated to exit if an error occurs in a sub-directory
18075 make (including if user presses ^C) [Paul Sutton]
18077 * Make Montgomery context stuff explicit in RSA data structure.
18081 * Fix build order of pem and err to allow for generated pem.h.
18085 * Fix renumbering bug in X509_NAME_delete_entry().
18089 * Enhanced the err-ins.pl script so it makes the error library number
18090 global and can add a library name. This is needed for external ASN1 and
18091 other error libraries.
18095 * Fixed sk_insert which never worked properly.
18099 * Fix ASN1 macros so they can handle indefinite length constructed
18100 EXPLICIT tags. Some non standard certificates use these: they can now
18105 * Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
18106 into a single doc/ssleay.txt bundle. This way the information is still
18107 preserved but no longer messes up this directory. Now it's new room for
18108 the new set of documentation files.
18110 *Ralf S. Engelschall*
18112 * SETs were incorrectly DER encoded. This was a major pain, because they
18113 shared code with SEQUENCEs, which aren't coded the same. This means that
18114 almost everything to do with SETs or SEQUENCEs has either changed name or
18115 number of arguments.
18117 *Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>*
18119 * Fix test data to work with the above.
18123 * Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
18124 was already fixed by Eric for 0.9.1 it seems.
18126 *Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>*
18128 * Autodetect FreeBSD3.
18132 * Fix various bugs in Configure. This affects the following platforms:
18136 unixware-2.0-pentium
18141 * Eliminate generated files from CVS. Reorder tests to regenerate files
18142 before they are needed.
18146 * Generate Makefile.ssl from Makefile.org (to keep CVS happy).
18150 ### Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
18152 * Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
18153 changed SSLeay to OpenSSL in version strings.
18155 *Ralf S. Engelschall*
18157 * Some fixups to the top-level documents.
18161 * Fixed the nasty bug where rsaref.h was not found under compile-time
18162 because the symlink to include/ was missing.
18164 *Ralf S. Engelschall*
18166 * Incorporated the popular no-RSA/DSA-only patches
18167 which allow to compile a RSA-free SSLeay.
18169 *Andrew Cooke / Interrader Ldt., Ralf S. Engelschall*
18171 * Fixed nasty rehash problem under `make -f Makefile.ssl links`
18172 when "ssleay" is still not found.
18174 *Ralf S. Engelschall*
18176 * Added more platforms to Configure: Cray T3E, HPUX 11,
18178 *Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>*
18180 * Updated the README file.
18182 *Ralf S. Engelschall*
18184 * Added various .cvsignore files in the CVS repository subdirs
18185 to make a "cvs update" really silent.
18187 *Ralf S. Engelschall*
18189 * Recompiled the error-definition header files and added
18190 missing symbols to the Win32 linker tables.
18192 *Ralf S. Engelschall*
18194 * Cleaned up the top-level documents;
18195 o new files: CHANGES and LICENSE
18196 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
18197 o merged COPYRIGHT into LICENSE
18198 o removed obsolete TODO file
18199 o renamed MICROSOFT to INSTALL.W32
18201 *Ralf S. Engelschall*
18203 * Removed dummy files from the 0.9.1b source tree:
18204 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
18205 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
18206 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
18207 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
18208 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
18210 *Ralf S. Engelschall*
18212 * Added various platform portability fixes.
18216 * The Genesis of the OpenSSL rpject:
18217 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
18218 Young and Tim J. Hudson created while they were working for C2Net until
18221 *The OpenSSL Project*
18223 ### Changes between 0.9.0b and 0.9.1b [not released]
18225 * Updated a few CA certificates under certs/
18229 * Changed some BIGNUM api stuff.
18233 * Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
18234 DGUX x86, Linux Alpha, etc.
18238 * New COMP library [crypto/comp/] for SSL Record Layer Compression:
18239 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
18244 * Add -strparse option to asn1pars program which parses nested
18247 *Dr Stephen Henson <shenson@bigfoot.com>*
18249 * Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
18253 * DSA fix for "ca" program.
18257 * Added "-genkey" option to "dsaparam" program.
18261 * Added RIPE MD160 (rmd160) message digest.
18265 * Added -a (all) option to "ssleay version" command.
18269 * Added PLATFORM define which is the id given to Configure.
18273 * Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
18277 * Extended the ASN.1 parser routines.
18281 * Extended BIO routines to support REUSEADDR, seek, tell, etc.
18285 * Added a BN_CTX to the BN library.
18289 * Fixed the weak key values in DES library
18293 * Changed API in EVP library for cipher aliases.
18297 * Added support for RC2/64bit cipher.
18301 * Converted the lhash library to the crypto/mem.c functions.
18305 * Added more recognized ASN.1 object ids.
18309 * Added more RSA padding checks for SSL/TLS.
18313 * Added BIO proxy/filter functionality.
18317 * Added extra_certs to SSL_CTX which can be used
18318 send extra CA certificates to the client in the CA cert chain sending
18319 process. It can be configured with SSL_CTX_add_extra_chain_cert().
18323 * Now Fortezza is denied in the authentication phase because
18324 this is key exchange mechanism is not supported by SSLeay at all.
18328 * Additional PKCS1 checks.
18332 * Support the string "TLSv1" for all TLS v1 ciphers.
18336 * Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
18337 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
18341 * Fixed a few memory leaks.
18345 * Fixed various code and comment typos.
18349 * A minor bug in ssl/s3_clnt.c where there would always be 4 0
18350 bytes sent in the client random.
18352 *Edward Bishop <ebishop@spyglass.com>*
18356 [CVE-2019-1563]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1563
18357 [CVE-2019-1559]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1559
18358 [CVE-2019-1552]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1552
18359 [CVE-2019-1551]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1551
18360 [CVE-2019-1549]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1549
18361 [CVE-2019-1547]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1547
18362 [CVE-2019-1543]: https://www.openssl.org/news/vulnerabilities.html#CVE-2019-1543
18363 [CVE-2018-5407]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-5407
18364 [CVE-2018-0739]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0739
18365 [CVE-2018-0737]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0737
18366 [CVE-2018-0735]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0735
18367 [CVE-2018-0734]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0734
18368 [CVE-2018-0733]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0733
18369 [CVE-2018-0732]: https://www.openssl.org/news/vulnerabilities.html#CVE-2018-0732
18370 [CVE-2017-3738]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3738
18371 [CVE-2017-3737]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3737
18372 [CVE-2017-3736]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3736
18373 [CVE-2017-3735]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3735
18374 [CVE-2017-3733]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3733
18375 [CVE-2017-3732]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3732
18376 [CVE-2017-3731]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3731
18377 [CVE-2017-3730]: https://www.openssl.org/news/vulnerabilities.html#CVE-2017-3730
18378 [CVE-2016-7055]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-7055
18379 [CVE-2016-7054]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-7054
18380 [CVE-2016-7053]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-7053
18381 [CVE-2016-7052]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-7052
18382 [CVE-2016-6309]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6309
18383 [CVE-2016-6308]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6308
18384 [CVE-2016-6307]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6307
18385 [CVE-2016-6306]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6306
18386 [CVE-2016-6305]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6305
18387 [CVE-2016-6304]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6304
18388 [CVE-2016-6303]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6303
18389 [CVE-2016-6302]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-6302
18390 [CVE-2016-2183]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2183
18391 [CVE-2016-2182]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2182
18392 [CVE-2016-2181]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2181
18393 [CVE-2016-2180]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2180
18394 [CVE-2016-2179]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2179
18395 [CVE-2016-2178]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2178
18396 [CVE-2016-2177]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2177
18397 [CVE-2016-2176]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2176
18398 [CVE-2016-2109]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2109
18399 [CVE-2016-2107]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2107
18400 [CVE-2016-2106]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2106
18401 [CVE-2016-2105]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-2105
18402 [CVE-2016-0800]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0800
18403 [CVE-2016-0799]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0799
18404 [CVE-2016-0798]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0798
18405 [CVE-2016-0797]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0797
18406 [CVE-2016-0705]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0705
18407 [CVE-2016-0702]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0702
18408 [CVE-2016-0701]: https://www.openssl.org/news/vulnerabilities.html#CVE-2016-0701
18409 [CVE-2015-3197]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-3197
18410 [CVE-2015-3196]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-3196
18411 [CVE-2015-3195]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-3195
18412 [CVE-2015-3194]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-3194
18413 [CVE-2015-3193]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-3193
18414 [CVE-2015-1793]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1793
18415 [CVE-2015-1792]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1792
18416 [CVE-2015-1791]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1791
18417 [CVE-2015-1790]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1790
18418 [CVE-2015-1789]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1789
18419 [CVE-2015-1788]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1788
18420 [CVE-2015-1787]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-1787
18421 [CVE-2015-0293]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0293
18422 [CVE-2015-0291]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0291
18423 [CVE-2015-0290]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0290
18424 [CVE-2015-0289]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0289
18425 [CVE-2015-0288]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0288
18426 [CVE-2015-0287]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0287
18427 [CVE-2015-0286]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0286
18428 [CVE-2015-0285]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0285
18429 [CVE-2015-0209]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0209
18430 [CVE-2015-0208]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0208
18431 [CVE-2015-0207]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0207
18432 [CVE-2015-0206]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0206
18433 [CVE-2015-0205]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0205
18434 [CVE-2015-0204]: https://www.openssl.org/news/vulnerabilities.html#CVE-2015-0204
18435 [CVE-2014-8275]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-8275
18436 [CVE-2014-5139]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-5139
18437 [CVE-2014-3572]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3572
18438 [CVE-2014-3571]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3571
18439 [CVE-2014-3570]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3570
18440 [CVE-2014-3569]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3569
18441 [CVE-2014-3568]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3568
18442 [CVE-2014-3567]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3567
18443 [CVE-2014-3566]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3566
18444 [CVE-2014-3513]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3513
18445 [CVE-2014-3512]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3512
18446 [CVE-2014-3511]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3511
18447 [CVE-2014-3510]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3510
18448 [CVE-2014-3509]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3509
18449 [CVE-2014-3508]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3508
18450 [CVE-2014-3507]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3507
18451 [CVE-2014-3506]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3506
18452 [CVE-2014-3505]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3505
18453 [CVE-2014-3470]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-3470
18454 [CVE-2014-0224]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-0224
18455 [CVE-2014-0221]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-0221
18456 [CVE-2014-0195]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-0195
18457 [CVE-2014-0160]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-0160
18458 [CVE-2014-0076]: https://www.openssl.org/news/vulnerabilities.html#CVE-2014-0076
18459 [CVE-2013-6450]: https://www.openssl.org/news/vulnerabilities.html#CVE-2013-6450
18460 [CVE-2013-4353]: https://www.openssl.org/news/vulnerabilities.html#CVE-2013-4353
18461 [CVE-2013-0169]: https://www.openssl.org/news/vulnerabilities.html#CVE-2013-0169
18462 [CVE-2013-0166]: https://www.openssl.org/news/vulnerabilities.html#CVE-2013-0166
18463 [CVE-2012-2686]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-2686
18464 [CVE-2012-2333]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-2333
18465 [CVE-2012-2110]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-2110
18466 [CVE-2012-0884]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-0884
18467 [CVE-2012-0050]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-0050
18468 [CVE-2012-0027]: https://www.openssl.org/news/vulnerabilities.html#CVE-2012-0027
18469 [CVE-2011-4619]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-4619
18470 [CVE-2011-4577]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-4577
18471 [CVE-2011-4576]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-4576
18472 [CVE-2011-4109]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-4109
18473 [CVE-2011-4108]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-4108
18474 [CVE-2011-3210]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-3210
18475 [CVE-2011-3207]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-3207
18476 [CVE-2011-0014]: https://www.openssl.org/news/vulnerabilities.html#CVE-2011-0014
18477 [CVE-2010-4252]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-4252
18478 [CVE-2010-4180]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-4180
18479 [CVE-2010-3864]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-3864
18480 [CVE-2010-1633]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-1633
18481 [CVE-2010-0740]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-0740
18482 [CVE-2010-0433]: https://www.openssl.org/news/vulnerabilities.html#CVE-2010-0433
18483 [CVE-2009-4355]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-4355
18484 [CVE-2009-3555]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-3555
18485 [CVE-2009-3245]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-3245
18486 [CVE-2009-1386]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-1386
18487 [CVE-2009-1379]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-1379
18488 [CVE-2009-1378]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-1378
18489 [CVE-2009-1377]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-1377
18490 [CVE-2009-0789]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-0789
18491 [CVE-2009-0591]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-0591
18492 [CVE-2009-0590]: https://www.openssl.org/news/vulnerabilities.html#CVE-2009-0590
18493 [CVE-2008-5077]: https://www.openssl.org/news/vulnerabilities.html#CVE-2008-5077
18494 [CVE-2008-1678]: https://www.openssl.org/news/vulnerabilities.html#CVE-2008-1678
18495 [CVE-2008-1672]: https://www.openssl.org/news/vulnerabilities.html#CVE-2008-1672
18496 [CVE-2008-0891]: https://www.openssl.org/news/vulnerabilities.html#CVE-2008-0891
18497 [CVE-2007-5135]: https://www.openssl.org/news/vulnerabilities.html#CVE-2007-5135
18498 [CVE-2007-4995]: https://www.openssl.org/news/vulnerabilities.html#CVE-2007-4995
18499 [CVE-2006-4343]: https://www.openssl.org/news/vulnerabilities.html#CVE-2006-4343
18500 [CVE-2006-4339]: https://www.openssl.org/news/vulnerabilities.html#CVE-2006-4339
18501 [CVE-2006-3738]: https://www.openssl.org/news/vulnerabilities.html#CVE-2006-3738
18502 [CVE-2006-2940]: https://www.openssl.org/news/vulnerabilities.html#CVE-2006-2940
18503 [CVE-2006-2937]: https://www.openssl.org/news/vulnerabilities.html#CVE-2006-2937
18504 [CVE-2005-2969]: https://www.openssl.org/news/vulnerabilities.html#CVE-2005-2969
18505 [CVE-2004-0112]: https://www.openssl.org/news/vulnerabilities.html#CVE-2004-0112
18506 [CVE-2004-0079]: https://www.openssl.org/news/vulnerabilities.html#CVE-2004-0079
18507 [CVE-2003-0851]: https://www.openssl.org/news/vulnerabilities.html#CVE-2003-0851
18508 [CVE-2003-0545]: https://www.openssl.org/news/vulnerabilities.html#CVE-2003-0545
18509 [CVE-2003-0544]: https://www.openssl.org/news/vulnerabilities.html#CVE-2003-0544
18510 [CVE-2003-0543]: https://www.openssl.org/news/vulnerabilities.html#CVE-2003-0543
18511 [CVE-2003-0078]: https://www.openssl.org/news/vulnerabilities.html#CVE-2003-0078
18512 [CVE-2002-0659]: https://www.openssl.org/news/vulnerabilities.html#CVE-2002-0659
18513 [CVE-2002-0657]: https://www.openssl.org/news/vulnerabilities.html#CVE-2002-0657
18514 [CVE-2002-0656]: https://www.openssl.org/news/vulnerabilities.html#CVE-2002-0656
18515 [CVE-2002-0655]: https://www.openssl.org/news/vulnerabilities.html#CVE-2002-0655