5 Changes between 1.0.2 and 1.1.0 [xx XXX xxxx]
7 *) config has been changed so that by default OPENSSL_NO_DEPRECATED is used.
8 Access to deprecated functions can be re-enabled by running config with
9 "enable-deprecated". In addition applications wishing to use deprecated
10 functions must define OPENSSL_USE_DEPRECATED. Note that this new behaviour
11 will, by default, disable some transitive includes that previously existed
12 in the header files (e.g. ec.h will no longer, by default, include bn.h)
15 *) Added support for OCB mode. OpenSSL has been granted a patent license
16 compatible with the OpenSSL license for use of OCB. Details are available
17 at https://www.openssl.org/docs/misc/OCB-patent-grant-OpenSSL.pdf. Support
18 for OCB can be removed by calling config with no-ocb.
21 *) SSLv2 support has been removed. It still supports receiving a SSLv2
22 compatible client hello.
25 *) Increased the minimal RSA keysize from 256 to 512 bits [Rich Salz],
26 done while fixing the error code for the key-too-small case.
27 [Annie Yousar <a.yousar@informatik.hu-berlin.de>]
29 *) Remove various unsupported platforms:
37 *) Experimental support for a new, fast, unbiased prime candidate generator,
38 bn_probable_prime_dh_coprime(). Not currently used by any prime generator.
39 [Felix Laurie von Massenbach <felix@erbridge.co.uk>]
41 *) New output format NSS in the sess_id command line tool. This allows
42 exporting the session id and the master key in NSS keylog format.
43 [Martin Kaiser <martin@kaiser.cx>]
45 *) Harmonize version and its documentation. -f flag is used to display
47 [mancha <mancha1@zoho.com>]
49 *) Fix eckey_priv_encode so it immediately returns an error upon a failure
51 [mancha <mancha1@zoho.com>]
53 *) Fix some double frees. These are not thought to be exploitable.
54 [mancha <mancha1@zoho.com>]
56 *) A missing bounds check in the handling of the TLS heartbeat extension
57 can be used to reveal up to 64k of memory to a connected client or
60 Thanks for Neel Mehta of Google Security for discovering this bug and to
61 Adam Langley <agl@chromium.org> and Bodo Moeller <bmoeller@acm.org> for
62 preparing the fix (CVE-2014-0160)
63 [Adam Langley, Bodo Moeller]
65 *) Fix for the attack described in the paper "Recovering OpenSSL
66 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
67 by Yuval Yarom and Naomi Benger. Details can be obtained from:
68 http://eprint.iacr.org/2014/140
70 Thanks to Yuval Yarom and Naomi Benger for discovering this
71 flaw and to Yuval Yarom for supplying a fix (CVE-2014-0076)
72 [Yuval Yarom and Naomi Benger]
74 *) Use algorithm specific chains in SSL_CTX_use_certificate_chain_file():
75 this fixes a limitation in previous versions of OpenSSL.
78 *) Experimental encrypt-then-mac support.
80 Experimental support for encrypt then mac from
81 draft-gutmann-tls-encrypt-then-mac-02.txt
83 To enable it set the appropriate extension number (0x42 for the test
84 server) using e.g. -DTLSEXT_TYPE_encrypt_then_mac=0x42
86 For non-compliant peers (i.e. just about everything) this should have no
89 WARNING: EXPERIMENTAL, SUBJECT TO CHANGE.
93 *) Add EVP support for key wrapping algorithms, to avoid problems with
94 existing code the flag EVP_CIPHER_CTX_WRAP_ALLOW has to be set in
95 the EVP_CIPHER_CTX or an error is returned. Add AES and DES3 wrap
96 algorithms and include tests cases.
99 *) Extend CMS code to support RSA-PSS signatures and RSA-OAEP for
103 *) Extended RSA OAEP support via EVP_PKEY API. Options to specify digest,
104 MGF1 digest and OAEP label.
107 *) Make openssl verify return errors.
108 [Chris Palmer <palmer@google.com> and Ben Laurie]
110 *) New function ASN1_TIME_diff to calculate the difference between two
111 ASN1_TIME structures or one structure and the current time.
114 *) Update fips_test_suite to support multiple command line options. New
115 test to induce all self test errors in sequence and check expected
119 *) Add FIPS_{rsa,dsa,ecdsa}_{sign,verify} functions which digest and
120 sign or verify all in one operation.
123 *) Add fips_algvs: a multicall fips utility incorporating all the algorithm
124 test programs and fips_test_suite. Includes functionality to parse
125 the minimal script output of fipsalgest.pl directly.
128 *) Add authorisation parameter to FIPS_module_mode_set().
131 *) Add FIPS selftest for ECDH algorithm using P-224 and B-233 curves.
134 *) Use separate DRBG fields for internal and external flags. New function
135 FIPS_drbg_health_check() to perform on demand health checking. Add
136 generation tests to fips_test_suite with reduced health check interval to
137 demonstrate periodic health checking. Add "nodh" option to
138 fips_test_suite to skip very slow DH test.
141 *) New function FIPS_get_cipherbynid() to lookup FIPS supported ciphers
145 *) More extensive health check for DRBG checking many more failure modes.
146 New function FIPS_selftest_drbg_all() to handle every possible DRBG
147 combination: call this in fips_test_suite.
150 *) Add support for Dual EC DRBG from SP800-90. Update DRBG algorithm test
151 and POST to handle Dual EC cases.
154 *) Add support for canonical generation of DSA parameter 'g'. See
157 *) Add support for HMAC DRBG from SP800-90. Update DRBG algorithm test and
158 POST to handle HMAC cases.
161 *) Add functions FIPS_module_version() and FIPS_module_version_text()
162 to return numerical and string versions of the FIPS module number.
165 *) Rename FIPS_mode_set and FIPS_mode to FIPS_module_mode_set and
166 FIPS_module_mode. FIPS_mode and FIPS_mode_set will be implemented
167 outside the validated module in the FIPS capable OpenSSL.
170 *) Minor change to DRBG entropy callback semantics. In some cases
171 there is no multiple of the block length between min_len and
172 max_len. Allow the callback to return more than max_len bytes
173 of entropy but discard any extra: it is the callback's responsibility
174 to ensure that the extra data discarded does not impact the
175 requested amount of entropy.
178 *) Add PRNG security strength checks to RSA, DSA and ECDSA using
179 information in FIPS186-3, SP800-57 and SP800-131A.
182 *) CCM support via EVP. Interface is very similar to GCM case except we
183 must supply all data in one chunk (i.e. no update, final) and the
184 message length must be supplied if AAD is used. Add algorithm test
188 *) Initial version of POST overhaul. Add POST callback to allow the status
189 of POST to be monitored and/or failures induced. Modify fips_test_suite
190 to use callback. Always run all selftests even if one fails.
193 *) XTS support including algorithm test driver in the fips_gcmtest program.
194 Note: this does increase the maximum key length from 32 to 64 bytes but
195 there should be no binary compatibility issues as existing applications
196 will never use XTS mode.
199 *) Extensive reorganisation of FIPS PRNG behaviour. Remove all dependencies
200 to OpenSSL RAND code and replace with a tiny FIPS RAND API which also
201 performs algorithm blocking for unapproved PRNG types. Also do not
202 set PRNG type in FIPS_mode_set(): leave this to the application.
203 Add default OpenSSL DRBG handling: sets up FIPS PRNG and seeds with
204 the standard OpenSSL PRNG: set additional data to a date time vector.
207 *) Rename old X9.31 PRNG functions of the form FIPS_rand* to FIPS_x931*.
208 This shouldn't present any incompatibility problems because applications
209 shouldn't be using these directly and any that are will need to rethink
210 anyway as the X9.31 PRNG is now deprecated by FIPS 140-2
213 *) Extensive self tests and health checking required by SP800-90 DRBG.
214 Remove strength parameter from FIPS_drbg_instantiate and always
215 instantiate at maximum supported strength.
218 *) Add ECDH code to fips module and fips_ecdhvs for primitives only testing.
221 *) New algorithm test program fips_dhvs to handle DH primitives only testing.
224 *) New function DH_compute_key_padded() to compute a DH key and pad with
225 leading zeroes if needed: this complies with SP800-56A et al.
228 *) Initial implementation of SP800-90 DRBGs for Hash and CTR. Not used by
229 anything, incomplete, subject to change and largely untested at present.
232 *) Modify fipscanisteronly build option to only build the necessary object
233 files by filtering FIPS_EX_OBJ through a perl script in crypto/Makefile.
236 *) Add experimental option FIPSSYMS to give all symbols in
237 fipscanister.o and FIPS or fips prefix. This will avoid
238 conflicts with future versions of OpenSSL. Add perl script
239 util/fipsas.pl to preprocess assembly language source files
240 and rename any affected symbols.
243 *) Add selftest checks and algorithm block of non-fips algorithms in
244 FIPS mode. Remove DES2 from selftests.
247 *) Add ECDSA code to fips module. Add tiny fips_ecdsa_check to just
248 return internal method without any ENGINE dependencies. Add new
249 tiny fips sign and verify functions.
252 *) New build option no-ec2m to disable characteristic 2 code.
255 *) New build option "fipscanisteronly". This only builds fipscanister.o
256 and (currently) associated fips utilities. Uses the file Makefile.fips
257 instead of Makefile.org as the prototype.
260 *) Add some FIPS mode restrictions to GCM. Add internal IV generator.
261 Update fips_gcmtest to use IV generator.
264 *) Initial, experimental EVP support for AES-GCM. AAD can be input by
265 setting output buffer to NULL. The *Final function must be
266 called although it will not retrieve any additional data. The tag
267 can be set or retrieved with a ctrl. The IV length is by default 12
268 bytes (96 bits) but can be set to an alternative value. If the IV
269 length exceeds the maximum IV length (currently 16 bytes) it cannot be
273 *) New flag in ciphers: EVP_CIPH_FLAG_CUSTOM_CIPHER. This means the
274 underlying do_cipher function handles all cipher semantics itself
275 including padding and finalisation. This is useful if (for example)
276 an ENGINE cipher handles block padding itself. The behaviour of
277 do_cipher is subtly changed if this flag is set: the return value
278 is the number of characters written to the output buffer (zero is
279 no longer an error code) or a negative error code. Also if the
280 input buffer is NULL and length 0 finalisation should be performed.
283 *) If a candidate issuer certificate is already part of the constructed
284 path ignore it: new debug notification X509_V_ERR_PATH_LOOP for this case.
287 *) Improve forward-security support: add functions
289 void SSL_CTX_set_not_resumable_session_callback(SSL_CTX *ctx, int (*cb)(SSL *ssl, int is_forward_secure))
290 void SSL_set_not_resumable_session_callback(SSL *ssl, int (*cb)(SSL *ssl, int is_forward_secure))
292 for use by SSL/TLS servers; the callback function will be called whenever a
293 new session is created, and gets to decide whether the session may be
294 cached to make it resumable (return 0) or not (return 1). (As by the
295 SSL/TLS protocol specifications, the session_id sent by the server will be
296 empty to indicate that the session is not resumable; also, the server will
297 not generate RFC 4507 (RFC 5077) session tickets.)
299 A simple reasonable callback implementation is to return is_forward_secure.
300 This parameter will be set to 1 or 0 depending on the ciphersuite selected
301 by the SSL/TLS server library, indicating whether it can provide forward
303 [Emilia Käsper <emilia.kasper@esat.kuleuven.be> (Google)]
305 *) New -verify_name option in command line utilities to set verification
309 *) Initial CMAC implementation. WARNING: EXPERIMENTAL, API MAY CHANGE.
310 Add CMAC pkey methods.
313 *) Experimental renegotiation in s_server -www mode. If the client
314 browses /reneg connection is renegotiated. If /renegcert it is
315 renegotiated requesting a certificate.
318 *) Add an "external" session cache for debugging purposes to s_server. This
319 should help trace issues which normally are only apparent in deployed
320 multi-process servers.
323 *) Extensive audit of libcrypto with DEBUG_UNUSED. Fix many cases where
324 return value is ignored. NB. The functions RAND_add(), RAND_seed(),
325 BIO_set_cipher() and some obscure PEM functions were changed so they
326 can now return an error. The RAND changes required a change to the
327 RAND_METHOD structure.
330 *) New macro __owur for "OpenSSL Warn Unused Result". This makes use of
331 a gcc attribute to warn if the result of a function is ignored. This
332 is enable if DEBUG_UNUSED is set. Add to several functions in evp.h
333 whose return value is often ignored.
336 Changes between 1.0.1k and 1.0.2 [xx XXX xxxx]
338 *) Accelerated NIST P-256 elliptic curve implementation for x86_64
339 (other platforms pending).
340 [Shay Gueron & Vlad Krasnov (Intel Corp), Andy Polyakov]
342 *) Add support for the SignedCertificateTimestampList certificate and
343 OCSP response extensions from RFC6962.
346 *) Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
347 for corner cases. (Certain input points at infinity could lead to
348 bogus results, with non-infinity inputs mapped to infinity too.)
351 *) Initial support for PowerISA 2.0.7, first implemented in POWER8.
352 This covers AES, SHA256/512 and GHASH. "Initial" means that most
353 common cases are optimized and there still is room for further
354 improvements. Vector Permutation AES for Altivec is also added.
357 *) Add support for little-endian ppc64 Linux target.
358 [Marcelo Cerri (IBM)]
360 *) Initial support for AMRv8 ISA crypto extensions. This covers AES,
361 SHA1, SHA256 and GHASH. "Initial" means that most common cases
362 are optimized and there still is room for further improvements.
363 Both 32- and 64-bit modes are supported.
364 [Andy Polyakov, Ard Biesheuvel (Linaro)]
366 *) Improved ARMv7 NEON support.
369 *) Support for SPARC Architecture 2011 crypto extensions, first
370 implemented in SPARC T4. This covers AES, DES, Camellia, SHA1,
371 SHA256/512, MD5, GHASH and modular exponentiation.
372 [Andy Polyakov, David Miller]
374 *) Accelerated modular exponentiation for Intel processors, a.k.a.
376 [Shay Gueron & Vlad Krasnov (Intel Corp)]
378 *) Support for new and upcoming Intel processors, including AVX2,
379 BMI and SHA ISA extensions. This includes additional "stitched"
380 implementations, AESNI-SHA256 and GCM, and multi-buffer support
383 This work was sponsored by Intel Corp.
386 *) Support for DTLS 1.2. This adds two sets of DTLS methods: DTLS_*_method()
387 supports both DTLS 1.2 and 1.0 and should use whatever version the peer
388 supports and DTLSv1_2_*_method() which supports DTLS 1.2 only.
391 *) Use algorithm specific chains in SSL_CTX_use_certificate_chain_file():
392 this fixes a limiation in previous versions of OpenSSL.
395 *) Extended RSA OAEP support via EVP_PKEY API. Options to specify digest,
396 MGF1 digest and OAEP label.
399 *) Add EVP support for key wrapping algorithms, to avoid problems with
400 existing code the flag EVP_CIPHER_CTX_WRAP_ALLOW has to be set in
401 the EVP_CIPHER_CTX or an error is returned. Add AES and DES3 wrap
402 algorithms and include tests cases.
405 *) Add functions to allocate and set the fields of an ECDSA_METHOD
407 [Douglas E. Engert, Steve Henson]
409 *) New functions OPENSSL_gmtime_diff and ASN1_TIME_diff to find the
410 difference in days and seconds between two tm or ASN1_TIME structures.
413 *) Add -rev test option to s_server to just reverse order of characters
414 received by client and send back to server. Also prints an abbreviated
415 summary of the connection parameters.
418 *) New option -brief for s_client and s_server to print out a brief summary
419 of connection parameters.
422 *) Add callbacks for arbitrary TLS extensions.
423 [Trevor Perrin <trevp@trevp.net> and Ben Laurie]
425 *) New option -crl_download in several openssl utilities to download CRLs
426 from CRLDP extension in certificates.
429 *) New options -CRL and -CRLform for s_client and s_server for CRLs.
432 *) New function X509_CRL_diff to generate a delta CRL from the difference
433 of two full CRLs. Add support to "crl" utility.
436 *) New functions to set lookup_crls function and to retrieve
437 X509_STORE from X509_STORE_CTX.
440 *) Print out deprecated issuer and subject unique ID fields in
444 *) Extend OCSP I/O functions so they can be used for simple general purpose
445 HTTP as well as OCSP. New wrapper function which can be used to download
446 CRLs using the OCSP API.
449 *) Delegate command line handling in s_client/s_server to SSL_CONF APIs.
452 *) SSL_CONF* functions. These provide a common framework for application
453 configuration using configuration files or command lines.
456 *) SSL/TLS tracing code. This parses out SSL/TLS records using the
457 message callback and prints the results. Needs compile time option
458 "enable-ssl-trace". New options to s_client and s_server to enable
462 *) New ctrl and macro to retrieve supported points extensions.
463 Print out extension in s_server and s_client.
466 *) New functions to retrieve certificate signature and signature
470 *) Add functions to retrieve and manipulate the raw cipherlist sent by a
474 *) New Suite B modes for TLS code. These use and enforce the requirements
475 of RFC6460: restrict ciphersuites, only permit Suite B algorithms and
476 only use Suite B curves. The Suite B modes can be set by using the
477 strings "SUITEB128", "SUITEB192" or "SUITEB128ONLY" for the cipherstring.
480 *) New chain verification flags for Suite B levels of security. Check
481 algorithms are acceptable when flags are set in X509_verify_cert.
484 *) Make tls1_check_chain return a set of flags indicating checks passed
485 by a certificate chain. Add additional tests to handle client
486 certificates: checks for matching certificate type and issuer name
490 *) If an attempt is made to use a signature algorithm not in the peer
491 preference list abort the handshake. If client has no suitable
492 signature algorithms in response to a certificate request do not
496 *) If server EC tmp key is not in client preference list abort handshake.
499 *) Add support for certificate stores in CERT structure. This makes it
500 possible to have different stores per SSL structure or one store in
501 the parent SSL_CTX. Include distint stores for certificate chain
502 verification and chain building. New ctrl SSL_CTRL_BUILD_CERT_CHAIN
503 to build and store a certificate chain in CERT structure: returing
504 an error if the chain cannot be built: this will allow applications
505 to test if a chain is correctly configured.
507 Note: if the CERT based stores are not set then the parent SSL_CTX
508 store is used to retain compatibility with existing behaviour.
512 *) New function ssl_set_client_disabled to set a ciphersuite disabled
513 mask based on the current session, check mask when sending client
514 hello and checking the requested ciphersuite.
517 *) New ctrls to retrieve and set certificate types in a certificate
518 request message. Print out received values in s_client. If certificate
519 types is not set with custom values set sensible values based on
520 supported signature algorithms.
523 *) Support for distinct client and server supported signature algorithms.
526 *) Add certificate callback. If set this is called whenever a certificate
527 is required by client or server. An application can decide which
528 certificate chain to present based on arbitrary criteria: for example
529 supported signature algorithms. Add very simple example to s_server.
530 This fixes many of the problems and restrictions of the existing client
531 certificate callback: for example you can now clear an existing
532 certificate and specify the whole chain.
535 *) Add new "valid_flags" field to CERT_PKEY structure which determines what
536 the certificate can be used for (if anything). Set valid_flags field
537 in new tls1_check_chain function. Simplify ssl_set_cert_masks which used
538 to have similar checks in it.
540 Add new "cert_flags" field to CERT structure and include a "strict mode".
541 This enforces some TLS certificate requirements (such as only permitting
542 certificate signature algorithms contained in the supported algorithms
543 extension) which some implementations ignore: this option should be used
544 with caution as it could cause interoperability issues.
547 *) Update and tidy signature algorithm extension processing. Work out
548 shared signature algorithms based on preferences and peer algorithms
549 and print them out in s_client and s_server. Abort handshake if no
550 shared signature algorithms.
553 *) Add new functions to allow customised supported signature algorithms
554 for SSL and SSL_CTX structures. Add options to s_client and s_server
558 *) New function SSL_certs_clear() to delete all references to certificates
559 from an SSL structure. Before this once a certificate had been added
560 it couldn't be removed.
563 *) Integrate hostname, email address and IP address checking with certificate
564 verification. New verify options supporting checking in opensl utility.
567 *) Fixes and wildcard matching support to hostname and email checking
568 functions. Add manual page.
569 [Florian Weimer (Red Hat Product Security Team)]
571 *) New functions to check a hostname email or IP address against a
572 certificate. Add options x509 utility to print results of checks against
576 *) Fix OCSP checking.
577 [Rob Stradling <rob.stradling@comodo.com> and Ben Laurie]
579 *) Initial experimental support for explicitly trusted non-root CAs.
580 OpenSSL still tries to build a complete chain to a root but if an
581 intermediate CA has a trust setting included that is used. The first
582 setting is used: whether to trust (e.g., -addtrust option to the x509
586 *) Add -trusted_first option which attempts to find certificates in the
587 trusted store even if an untrusted chain is also supplied.
590 *) MIPS assembly pack updates: support for MIPS32r2 and SmartMIPS ASE,
591 platform support for Linux and Android.
594 *) Support for linux-x32, ILP32 environment in x86_64 framework.
597 *) Experimental multi-implementation support for FIPS capable OpenSSL.
598 When in FIPS mode the approved implementations are used as normal,
599 when not in FIPS mode the internal unapproved versions are used instead.
600 This means that the FIPS capable OpenSSL isn't forced to use the
601 (often lower performance) FIPS implementations outside FIPS mode.
604 *) Transparently support X9.42 DH parameters when calling
605 PEM_read_bio_DHparameters. This means existing applications can handle
606 the new parameter format automatically.
609 *) Initial experimental support for X9.42 DH parameter format: mainly
610 to support use of 'q' parameter for RFC5114 parameters.
613 *) Add DH parameters from RFC5114 including test data to dhtest.
616 *) Support for automatic EC temporary key parameter selection. If enabled
617 the most preferred EC parameters are automatically used instead of
618 hardcoded fixed parameters. Now a server just has to call:
619 SSL_CTX_set_ecdh_auto(ctx, 1) and the server will automatically
620 support ECDH and use the most appropriate parameters.
623 *) Enhance and tidy EC curve and point format TLS extension code. Use
624 static structures instead of allocation if default values are used.
625 New ctrls to set curves we wish to support and to retrieve shared curves.
626 Print out shared curves in s_server. New options to s_server and s_client
627 to set list of supported curves.
630 *) New ctrls to retrieve supported signature algorithms and
631 supported curve values as an array of NIDs. Extend openssl utility
632 to print out received values.
635 *) Add new APIs EC_curve_nist2nid and EC_curve_nid2nist which convert
636 between NIDs and the more common NIST names such as "P-256". Enhance
637 ecparam utility and ECC method to recognise the NIST names for curves.
640 *) Enhance SSL/TLS certificate chain handling to support different
641 chains for each certificate instead of one chain in the parent SSL_CTX.
644 *) Support for fixed DH ciphersuite client authentication: where both
645 server and client use DH certificates with common parameters.
648 *) Support for fixed DH ciphersuites: those requiring DH server
652 *) New function i2d_re_X509_tbs for re-encoding the TBS portion of
654 Note: Related 1.0.2-beta specific macros X509_get_cert_info,
655 X509_CINF_set_modified, X509_CINF_get_issuer, X509_CINF_get_extensions and
656 X509_CINF_get_signature were reverted post internal team review.
658 Changes between 1.0.1j and 1.0.1k [xx XXX xxxx]
660 *) Do not resume sessions on the server if the negotiated protocol
661 version does not match the session's version. Resuming with a different
662 version, while not strictly forbidden by the RFC, is of questionable
663 sanity and breaks all known clients.
664 [David Benjamin, Emilia Käsper]
666 *) Tighten handling of the ChangeCipherSpec (CCS) message: reject
667 early CCS messages during renegotiation. (Note that because
668 renegotiation is encrypted, this early CCS was not exploitable.)
671 *) Tighten client-side session ticket handling during renegotiation:
672 ensure that the client only accepts a session ticket if the server sends
673 the extension anew in the ServerHello. Previously, a TLS client would
674 reuse the old extension state and thus accept a session ticket if one was
675 announced in the initial ServerHello.
677 Similarly, ensure that the client requires a session ticket if one
678 was advertised in the ServerHello. Previously, a TLS client would
679 ignore a missing NewSessionTicket message.
682 Changes between 1.0.1i and 1.0.1j [15 Oct 2014]
686 A flaw in the DTLS SRTP extension parsing code allows an attacker, who
687 sends a carefully crafted handshake message, to cause OpenSSL to fail
688 to free up to 64k of memory causing a memory leak. This could be
689 exploited in a Denial Of Service attack. This issue affects OpenSSL
690 1.0.1 server implementations for both SSL/TLS and DTLS regardless of
691 whether SRTP is used or configured. Implementations of OpenSSL that
692 have been compiled with OPENSSL_NO_SRTP defined are not affected.
694 The fix was developed by the OpenSSL team.
698 *) Session Ticket Memory Leak.
700 When an OpenSSL SSL/TLS/DTLS server receives a session ticket the
701 integrity of that ticket is first verified. In the event of a session
702 ticket integrity check failing, OpenSSL will fail to free memory
703 causing a memory leak. By sending a large number of invalid session
704 tickets an attacker could exploit this issue in a Denial Of Service
709 *) Build option no-ssl3 is incomplete.
711 When OpenSSL is configured with "no-ssl3" as a build option, servers
712 could accept and complete a SSL 3.0 handshake, and clients could be
713 configured to send them.
715 [Akamai and the OpenSSL team]
717 *) Add support for TLS_FALLBACK_SCSV.
718 Client applications doing fallback retries should call
719 SSL_set_mode(s, SSL_MODE_SEND_FALLBACK_SCSV).
721 [Adam Langley, Bodo Moeller]
723 *) Add additional DigestInfo checks.
725 Reencode DigestInto in DER and check against the original when
726 verifying RSA signature: this will reject any improperly encoded
727 DigestInfo structures.
729 Note: this is a precautionary measure and no attacks are currently known.
733 Changes between 1.0.1h and 1.0.1i [6 Aug 2014]
735 *) Fix SRP buffer overrun vulnerability. Invalid parameters passed to the
736 SRP code can be overrun an internal buffer. Add sanity check that
737 g, A, B < N to SRP code.
739 Thanks to Sean Devlin and Watson Ladd of Cryptography Services, NCC
740 Group for discovering this issue.
744 *) A flaw in the OpenSSL SSL/TLS server code causes the server to negotiate
745 TLS 1.0 instead of higher protocol versions when the ClientHello message
746 is badly fragmented. This allows a man-in-the-middle attacker to force a
747 downgrade to TLS 1.0 even if both the server and the client support a
748 higher protocol version, by modifying the client's TLS records.
750 Thanks to David Benjamin and Adam Langley (Google) for discovering and
751 researching this issue.
755 *) OpenSSL DTLS clients enabling anonymous (EC)DH ciphersuites are subject
756 to a denial of service attack. A malicious server can crash the client
757 with a null pointer dereference (read) by specifying an anonymous (EC)DH
758 ciphersuite and sending carefully crafted handshake messages.
760 Thanks to Felix Gröbert (Google) for discovering and researching this
765 *) By sending carefully crafted DTLS packets an attacker could cause openssl
766 to leak memory. This can be exploited through a Denial of Service attack.
767 Thanks to Adam Langley for discovering and researching this issue.
771 *) An attacker can force openssl to consume large amounts of memory whilst
772 processing DTLS handshake messages. This can be exploited through a
773 Denial of Service attack.
774 Thanks to Adam Langley for discovering and researching this issue.
778 *) An attacker can force an error condition which causes openssl to crash
779 whilst processing DTLS packets due to memory being freed twice. This
780 can be exploited through a Denial of Service attack.
781 Thanks to Adam Langley and Wan-Teh Chang for discovering and researching
786 *) If a multithreaded client connects to a malicious server using a resumed
787 session and the server sends an ec point format extension it could write
788 up to 255 bytes to freed memory.
790 Thanks to Gabor Tyukasz (LogMeIn Inc) for discovering and researching this
795 *) A malicious server can crash an OpenSSL client with a null pointer
796 dereference (read) by specifying an SRP ciphersuite even though it was not
797 properly negotiated with the client. This can be exploited through a
798 Denial of Service attack.
800 Thanks to Joonas Kuorilehto and Riku Hietamäki (Codenomicon) for
801 discovering and researching this issue.
805 *) A flaw in OBJ_obj2txt may cause pretty printing functions such as
806 X509_name_oneline, X509_name_print_ex et al. to leak some information
807 from the stack. Applications may be affected if they echo pretty printing
808 output to the attacker.
810 Thanks to Ivan Fratric (Google) for discovering this issue.
812 [Emilia Käsper, and Steve Henson]
814 *) Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
815 for corner cases. (Certain input points at infinity could lead to
816 bogus results, with non-infinity inputs mapped to infinity too.)
819 Changes between 1.0.1g and 1.0.1h [5 Jun 2014]
821 *) Fix for SSL/TLS MITM flaw. An attacker using a carefully crafted
822 handshake can force the use of weak keying material in OpenSSL
823 SSL/TLS clients and servers.
825 Thanks to KIKUCHI Masashi (Lepidum Co. Ltd.) for discovering and
826 researching this issue. (CVE-2014-0224)
827 [KIKUCHI Masashi, Steve Henson]
829 *) Fix DTLS recursion flaw. By sending an invalid DTLS handshake to an
830 OpenSSL DTLS client the code can be made to recurse eventually crashing
833 Thanks to Imre Rad (Search-Lab Ltd.) for discovering this issue.
835 [Imre Rad, Steve Henson]
837 *) Fix DTLS invalid fragment vulnerability. A buffer overrun attack can
838 be triggered by sending invalid DTLS fragments to an OpenSSL DTLS
839 client or server. This is potentially exploitable to run arbitrary
840 code on a vulnerable client or server.
842 Thanks to Jüri Aedla for reporting this issue. (CVE-2014-0195)
843 [Jüri Aedla, Steve Henson]
845 *) Fix bug in TLS code where clients enable anonymous ECDH ciphersuites
846 are subject to a denial of service attack.
848 Thanks to Felix Gröbert and Ivan Fratric at Google for discovering
849 this issue. (CVE-2014-3470)
850 [Felix Gröbert, Ivan Fratric, Steve Henson]
852 *) Harmonize version and its documentation. -f flag is used to display
854 [mancha <mancha1@zoho.com>]
856 *) Fix eckey_priv_encode so it immediately returns an error upon a failure
858 [mancha <mancha1@zoho.com>]
860 *) Fix some double frees. These are not thought to be exploitable.
861 [mancha <mancha1@zoho.com>]
863 Changes between 1.0.1f and 1.0.1g [7 Apr 2014]
865 *) A missing bounds check in the handling of the TLS heartbeat extension
866 can be used to reveal up to 64k of memory to a connected client or
869 Thanks for Neel Mehta of Google Security for discovering this bug and to
870 Adam Langley <agl@chromium.org> and Bodo Moeller <bmoeller@acm.org> for
871 preparing the fix (CVE-2014-0160)
872 [Adam Langley, Bodo Moeller]
874 *) Fix for the attack described in the paper "Recovering OpenSSL
875 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
876 by Yuval Yarom and Naomi Benger. Details can be obtained from:
877 http://eprint.iacr.org/2014/140
879 Thanks to Yuval Yarom and Naomi Benger for discovering this
880 flaw and to Yuval Yarom for supplying a fix (CVE-2014-0076)
881 [Yuval Yarom and Naomi Benger]
883 *) TLS pad extension: draft-agl-tls-padding-03
885 Workaround for the "TLS hang bug" (see FAQ and PR#2771): if the
886 TLS client Hello record length value would otherwise be > 255 and
887 less that 512 pad with a dummy extension containing zeroes so it
888 is at least 512 bytes long.
890 [Adam Langley, Steve Henson]
892 Changes between 1.0.1e and 1.0.1f [6 Jan 2014]
894 *) Fix for TLS record tampering bug. A carefully crafted invalid
895 handshake could crash OpenSSL with a NULL pointer exception.
896 Thanks to Anton Johansson for reporting this issues.
899 *) Keep original DTLS digest and encryption contexts in retransmission
900 structures so we can use the previous session parameters if they need
901 to be resent. (CVE-2013-6450)
904 *) Add option SSL_OP_SAFARI_ECDHE_ECDSA_BUG (part of SSL_OP_ALL) which
905 avoids preferring ECDHE-ECDSA ciphers when the client appears to be
906 Safari on OS X. Safari on OS X 10.8..10.8.3 advertises support for
907 several ECDHE-ECDSA ciphers, but fails to negotiate them. The bug
908 is fixed in OS X 10.8.4, but Apple have ruled out both hot fixing
909 10.8..10.8.3 and forcing users to upgrade to 10.8.4 or newer.
910 [Rob Stradling, Adam Langley]
912 Changes between 1.0.1d and 1.0.1e [11 Feb 2013]
914 *) Correct fix for CVE-2013-0169. The original didn't work on AES-NI
915 supporting platforms or when small records were transferred.
916 [Andy Polyakov, Steve Henson]
918 Changes between 1.0.1c and 1.0.1d [5 Feb 2013]
920 *) Make the decoding of SSLv3, TLS and DTLS CBC records constant time.
922 This addresses the flaw in CBC record processing discovered by
923 Nadhem Alfardan and Kenny Paterson. Details of this attack can be found
924 at: http://www.isg.rhul.ac.uk/tls/
926 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
927 Security Group at Royal Holloway, University of London
928 (www.isg.rhul.ac.uk) for discovering this flaw and Adam Langley and
929 Emilia Käsper for the initial patch.
931 [Emilia Käsper, Adam Langley, Ben Laurie, Andy Polyakov, Steve Henson]
933 *) Fix flaw in AESNI handling of TLS 1.2 and 1.1 records for CBC mode
934 ciphersuites which can be exploited in a denial of service attack.
935 Thanks go to and to Adam Langley <agl@chromium.org> for discovering
936 and detecting this bug and to Wolfgang Ettlinger
937 <wolfgang.ettlinger@gmail.com> for independently discovering this issue.
941 *) Return an error when checking OCSP signatures when key is NULL.
942 This fixes a DoS attack. (CVE-2013-0166)
945 *) Make openssl verify return errors.
946 [Chris Palmer <palmer@google.com> and Ben Laurie]
948 *) Call OCSP Stapling callback after ciphersuite has been chosen, so
949 the right response is stapled. Also change SSL_get_certificate()
950 so it returns the certificate actually sent.
951 See http://rt.openssl.org/Ticket/Display.html?id=2836.
952 [Rob Stradling <rob.stradling@comodo.com>]
954 *) Fix possible deadlock when decoding public keys.
957 *) Don't use TLS 1.0 record version number in initial client hello
961 Changes between 1.0.1b and 1.0.1c [10 May 2012]
963 *) Sanity check record length before skipping explicit IV in TLS
964 1.2, 1.1 and DTLS to fix DoS attack.
966 Thanks to Codenomicon for discovering this issue using Fuzz-o-Matic
967 fuzzing as a service testing platform.
971 *) Initialise tkeylen properly when encrypting CMS messages.
972 Thanks to Solar Designer of Openwall for reporting this issue.
975 *) In FIPS mode don't try to use composite ciphers as they are not
979 Changes between 1.0.1a and 1.0.1b [26 Apr 2012]
981 *) OpenSSL 1.0.0 sets SSL_OP_ALL to 0x80000FFFL and OpenSSL 1.0.1 and
982 1.0.1a set SSL_OP_NO_TLSv1_1 to 0x00000400L which would unfortunately
983 mean any application compiled against OpenSSL 1.0.0 headers setting
984 SSL_OP_ALL would also set SSL_OP_NO_TLSv1_1, unintentionally disablng
985 TLS 1.1 also. Fix this by changing the value of SSL_OP_NO_TLSv1_1 to
986 0x10000000L Any application which was previously compiled against
987 OpenSSL 1.0.1 or 1.0.1a headers and which cares about SSL_OP_NO_TLSv1_1
988 will need to be recompiled as a result. Letting be results in
989 inability to disable specifically TLS 1.1 and in client context,
990 in unlike event, limit maximum offered version to TLS 1.0 [see below].
993 *) In order to ensure interoperabilty SSL_OP_NO_protocolX does not
994 disable just protocol X, but all protocols above X *if* there are
995 protocols *below* X still enabled. In more practical terms it means
996 that if application wants to disable TLS1.0 in favor of TLS1.1 and
997 above, it's not sufficient to pass SSL_OP_NO_TLSv1, one has to pass
998 SSL_OP_NO_TLSv1|SSL_OP_NO_SSLv3|SSL_OP_NO_SSLv2. This applies to
1002 Changes between 1.0.1 and 1.0.1a [19 Apr 2012]
1004 *) Check for potentially exploitable overflows in asn1_d2i_read_bio
1005 BUF_mem_grow and BUF_mem_grow_clean. Refuse attempts to shrink buffer
1006 in CRYPTO_realloc_clean.
1008 Thanks to Tavis Ormandy, Google Security Team, for discovering this
1009 issue and to Adam Langley <agl@chromium.org> for fixing it.
1011 [Adam Langley (Google), Tavis Ormandy, Google Security Team]
1013 *) Don't allow TLS 1.2 SHA-256 ciphersuites in TLS 1.0, 1.1 connections.
1016 *) Workarounds for some broken servers that "hang" if a client hello
1017 record length exceeds 255 bytes.
1019 1. Do not use record version number > TLS 1.0 in initial client
1020 hello: some (but not all) hanging servers will now work.
1021 2. If we set OPENSSL_MAX_TLS1_2_CIPHER_LENGTH this will truncate
1022 the number of ciphers sent in the client hello. This should be
1023 set to an even number, such as 50, for example by passing:
1024 -DOPENSSL_MAX_TLS1_2_CIPHER_LENGTH=50 to config or Configure.
1025 Most broken servers should now work.
1026 3. If all else fails setting OPENSSL_NO_TLS1_2_CLIENT will disable
1027 TLS 1.2 client support entirely.
1030 *) Fix SEGV in Vector Permutation AES module observed in OpenSSH.
1033 Changes between 1.0.0h and 1.0.1 [14 Mar 2012]
1035 *) Add compatibility with old MDC2 signatures which use an ASN1 OCTET
1036 STRING form instead of a DigestInfo.
1039 *) The format used for MDC2 RSA signatures is inconsistent between EVP
1040 and the RSA_sign/RSA_verify functions. This was made more apparent when
1041 OpenSSL used RSA_sign/RSA_verify for some RSA signatures in particular
1042 those which went through EVP_PKEY_METHOD in 1.0.0 and later. Detect
1043 the correct format in RSA_verify so both forms transparently work.
1046 *) Some servers which support TLS 1.0 can choke if we initially indicate
1047 support for TLS 1.2 and later renegotiate using TLS 1.0 in the RSA
1048 encrypted premaster secret. As a workaround use the maximum permitted
1049 client version in client hello, this should keep such servers happy
1050 and still work with previous versions of OpenSSL.
1053 *) Add support for TLS/DTLS heartbeats.
1054 [Robin Seggelmann <seggelmann@fh-muenster.de>]
1056 *) Add support for SCTP.
1057 [Robin Seggelmann <seggelmann@fh-muenster.de>]
1059 *) Improved PRNG seeding for VOS.
1060 [Paul Green <Paul.Green@stratus.com>]
1062 *) Extensive assembler packs updates, most notably:
1064 - x86[_64]: AES-NI, PCLMULQDQ, RDRAND support;
1065 - x86[_64]: SSSE3 support (SHA1, vector-permutation AES);
1066 - x86_64: bit-sliced AES implementation;
1067 - ARM: NEON support, contemporary platforms optimizations;
1068 - s390x: z196 support;
1069 - *: GHASH and GF(2^m) multiplication implementations;
1073 *) Make TLS-SRP code conformant with RFC 5054 API cleanup
1074 (removal of unnecessary code)
1075 [Peter Sylvester <peter.sylvester@edelweb.fr>]
1077 *) Add TLS key material exporter from RFC 5705.
1080 *) Add DTLS-SRTP negotiation from RFC 5764.
1083 *) Add Next Protocol Negotiation,
1084 http://tools.ietf.org/html/draft-agl-tls-nextprotoneg-00. Can be
1085 disabled with a no-npn flag to config or Configure. Code donated
1087 [Adam Langley <agl@google.com> and Ben Laurie]
1089 *) Add optional 64-bit optimized implementations of elliptic curves NIST-P224,
1090 NIST-P256, NIST-P521, with constant-time single point multiplication on
1091 typical inputs. Compiler support for the nonstandard type __uint128_t is
1092 required to use this (present in gcc 4.4 and later, for 64-bit builds).
1093 Code made available under Apache License version 2.0.
1095 Specify "enable-ec_nistp_64_gcc_128" on the Configure (or config) command
1096 line to include this in your build of OpenSSL, and run "make depend" (or
1097 "make update"). This enables the following EC_METHODs:
1099 EC_GFp_nistp224_method()
1100 EC_GFp_nistp256_method()
1101 EC_GFp_nistp521_method()
1103 EC_GROUP_new_by_curve_name() will automatically use these (while
1104 EC_GROUP_new_curve_GFp() currently prefers the more flexible
1106 [Emilia Käsper, Adam Langley, Bodo Moeller (Google)]
1108 *) Use type ossl_ssize_t instad of ssize_t which isn't available on
1109 all platforms. Move ssize_t definition from e_os.h to the public
1110 header file e_os2.h as it now appears in public header file cms.h
1113 *) New -sigopt option to the ca, req and x509 utilities. Additional
1114 signature parameters can be passed using this option and in
1118 *) Add RSA PSS signing function. This will generate and set the
1119 appropriate AlgorithmIdentifiers for PSS based on those in the
1120 corresponding EVP_MD_CTX structure. No application support yet.
1123 *) Support for companion algorithm specific ASN1 signing routines.
1124 New function ASN1_item_sign_ctx() signs a pre-initialised
1125 EVP_MD_CTX structure and sets AlgorithmIdentifiers based on
1126 the appropriate parameters.
1129 *) Add new algorithm specific ASN1 verification initialisation function
1130 to EVP_PKEY_ASN1_METHOD: this is not in EVP_PKEY_METHOD since the ASN1
1131 handling will be the same no matter what EVP_PKEY_METHOD is used.
1132 Add a PSS handler to support verification of PSS signatures: checked
1133 against a number of sample certificates.
1136 *) Add signature printing for PSS. Add PSS OIDs.
1137 [Steve Henson, Martin Kaiser <lists@kaiser.cx>]
1139 *) Add algorithm specific signature printing. An individual ASN1 method
1140 can now print out signatures instead of the standard hex dump.
1142 More complex signatures (e.g. PSS) can print out more meaningful
1143 information. Include DSA version that prints out the signature
1147 *) Password based recipient info support for CMS library: implementing
1151 *) Split password based encryption into PBES2 and PBKDF2 functions. This
1152 neatly separates the code into cipher and PBE sections and is required
1153 for some algorithms that split PBES2 into separate pieces (such as
1154 password based CMS).
1157 *) Session-handling fixes:
1158 - Fix handling of connections that are resuming with a session ID,
1159 but also support Session Tickets.
1160 - Fix a bug that suppressed issuing of a new ticket if the client
1161 presented a ticket with an expired session.
1162 - Try to set the ticket lifetime hint to something reasonable.
1163 - Make tickets shorter by excluding irrelevant information.
1164 - On the client side, don't ignore renewed tickets.
1165 [Adam Langley, Bodo Moeller (Google)]
1167 *) Fix PSK session representation.
1170 *) Add RC4-MD5 and AESNI-SHA1 "stitched" implementations.
1172 This work was sponsored by Intel.
1175 *) Add GCM support to TLS library. Some custom code is needed to split
1176 the IV between the fixed (from PRF) and explicit (from TLS record)
1177 portions. This adds all GCM ciphersuites supported by RFC5288 and
1178 RFC5289. Generalise some AES* cipherstrings to include GCM and
1179 add a special AESGCM string for GCM only.
1182 *) Expand range of ctrls for AES GCM. Permit setting invocation
1183 field on decrypt and retrieval of invocation field only on encrypt.
1186 *) Add HMAC ECC ciphersuites from RFC5289. Include SHA384 PRF support.
1187 As required by RFC5289 these ciphersuites cannot be used if for
1188 versions of TLS earlier than 1.2.
1191 *) For FIPS capable OpenSSL interpret a NULL default public key method
1192 as unset and return the appropriate default but do *not* set the default.
1193 This means we can return the appropriate method in applications that
1194 switch between FIPS and non-FIPS modes.
1197 *) Redirect HMAC and CMAC operations to FIPS module in FIPS mode. If an
1198 ENGINE is used then we cannot handle that in the FIPS module so we
1199 keep original code iff non-FIPS operations are allowed.
1202 *) Add -attime option to openssl utilities.
1203 [Peter Eckersley <pde@eff.org>, Ben Laurie and Steve Henson]
1205 *) Redirect DSA and DH operations to FIPS module in FIPS mode.
1208 *) Redirect ECDSA and ECDH operations to FIPS module in FIPS mode. Also use
1209 FIPS EC methods unconditionally for now.
1212 *) New build option no-ec2m to disable characteristic 2 code.
1215 *) Backport libcrypto audit of return value checking from 1.1.0-dev; not
1216 all cases can be covered as some introduce binary incompatibilities.
1219 *) Redirect RSA operations to FIPS module including keygen,
1220 encrypt, decrypt, sign and verify. Block use of non FIPS RSA methods.
1223 *) Add similar low level API blocking to ciphers.
1226 *) Low level digest APIs are not approved in FIPS mode: any attempt
1227 to use these will cause a fatal error. Applications that *really* want
1228 to use them can use the private_* version instead.
1231 *) Redirect cipher operations to FIPS module for FIPS builds.
1234 *) Redirect digest operations to FIPS module for FIPS builds.
1237 *) Update build system to add "fips" flag which will link in fipscanister.o
1238 for static and shared library builds embedding a signature if needed.
1241 *) Output TLS supported curves in preference order instead of numerical
1242 order. This is currently hardcoded for the highest order curves first.
1243 This should be configurable so applications can judge speed vs strength.
1246 *) Add TLS v1.2 server support for client authentication.
1249 *) Add support for FIPS mode in ssl library: disable SSLv3, non-FIPS ciphers
1253 *) Functions FIPS_mode_set() and FIPS_mode() which call the underlying
1254 FIPS modules versions.
1257 *) Add TLS v1.2 client side support for client authentication. Keep cache
1258 of handshake records longer as we don't know the hash algorithm to use
1259 until after the certificate request message is received.
1262 *) Initial TLS v1.2 client support. Add a default signature algorithms
1263 extension including all the algorithms we support. Parse new signature
1264 format in client key exchange. Relax some ECC signing restrictions for
1265 TLS v1.2 as indicated in RFC5246.
1268 *) Add server support for TLS v1.2 signature algorithms extension. Switch
1269 to new signature format when needed using client digest preference.
1270 All server ciphersuites should now work correctly in TLS v1.2. No client
1271 support yet and no support for client certificates.
1274 *) Initial TLS v1.2 support. Add new SHA256 digest to ssl code, switch
1275 to SHA256 for PRF when using TLS v1.2 and later. Add new SHA256 based
1276 ciphersuites. At present only RSA key exchange ciphersuites work with
1277 TLS v1.2. Add new option for TLS v1.2 replacing the old and obsolete
1278 SSL_OP_PKCS1_CHECK flags with SSL_OP_NO_TLSv1_2. New TLSv1.2 methods
1279 and version checking.
1282 *) New option OPENSSL_NO_SSL_INTERN. If an application can be compiled
1283 with this defined it will not be affected by any changes to ssl internal
1284 structures. Add several utility functions to allow openssl application
1285 to work with OPENSSL_NO_SSL_INTERN defined.
1289 [Tom Wu <tjw@cs.stanford.edu> and Ben Laurie]
1291 *) Add functions to copy EVP_PKEY_METHOD and retrieve flags and id.
1294 *) Permit abbreviated handshakes when renegotiating using the function
1295 SSL_renegotiate_abbreviated().
1296 [Robin Seggelmann <seggelmann@fh-muenster.de>]
1298 *) Add call to ENGINE_register_all_complete() to
1299 ENGINE_load_builtin_engines(), so some implementations get used
1300 automatically instead of needing explicit application support.
1303 *) Add support for TLS key exporter as described in RFC5705.
1304 [Robin Seggelmann <seggelmann@fh-muenster.de>, Steve Henson]
1306 *) Initial TLSv1.1 support. Since TLSv1.1 is very similar to TLS v1.0 only
1307 a few changes are required:
1309 Add SSL_OP_NO_TLSv1_1 flag.
1310 Add TLSv1_1 methods.
1311 Update version checking logic to handle version 1.1.
1312 Add explicit IV handling (ported from DTLS code).
1313 Add command line options to s_client/s_server.
1316 Changes between 1.0.0j and 1.0.0k [5 Feb 2013]
1318 *) Make the decoding of SSLv3, TLS and DTLS CBC records constant time.
1320 This addresses the flaw in CBC record processing discovered by
1321 Nadhem Alfardan and Kenny Paterson. Details of this attack can be found
1322 at: http://www.isg.rhul.ac.uk/tls/
1324 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
1325 Security Group at Royal Holloway, University of London
1326 (www.isg.rhul.ac.uk) for discovering this flaw and Adam Langley and
1327 Emilia Käsper for the initial patch.
1329 [Emilia Käsper, Adam Langley, Ben Laurie, Andy Polyakov, Steve Henson]
1331 *) Return an error when checking OCSP signatures when key is NULL.
1332 This fixes a DoS attack. (CVE-2013-0166)
1335 *) Call OCSP Stapling callback after ciphersuite has been chosen, so
1336 the right response is stapled. Also change SSL_get_certificate()
1337 so it returns the certificate actually sent.
1338 See http://rt.openssl.org/Ticket/Display.html?id=2836.
1339 (This is a backport)
1340 [Rob Stradling <rob.stradling@comodo.com>]
1342 *) Fix possible deadlock when decoding public keys.
1345 Changes between 1.0.0i and 1.0.0j [10 May 2012]
1347 [NB: OpenSSL 1.0.0i and later 1.0.0 patch levels were released after
1350 *) Sanity check record length before skipping explicit IV in DTLS
1353 Thanks to Codenomicon for discovering this issue using Fuzz-o-Matic
1354 fuzzing as a service testing platform.
1358 *) Initialise tkeylen properly when encrypting CMS messages.
1359 Thanks to Solar Designer of Openwall for reporting this issue.
1362 Changes between 1.0.0h and 1.0.0i [19 Apr 2012]
1364 *) Check for potentially exploitable overflows in asn1_d2i_read_bio
1365 BUF_mem_grow and BUF_mem_grow_clean. Refuse attempts to shrink buffer
1366 in CRYPTO_realloc_clean.
1368 Thanks to Tavis Ormandy, Google Security Team, for discovering this
1369 issue and to Adam Langley <agl@chromium.org> for fixing it.
1371 [Adam Langley (Google), Tavis Ormandy, Google Security Team]
1373 Changes between 1.0.0g and 1.0.0h [12 Mar 2012]
1375 *) Fix MMA (Bleichenbacher's attack on PKCS #1 v1.5 RSA padding) weakness
1376 in CMS and PKCS7 code. When RSA decryption fails use a random key for
1377 content decryption and always return the same error. Note: this attack
1378 needs on average 2^20 messages so it only affects automated senders. The
1379 old behaviour can be reenabled in the CMS code by setting the
1380 CMS_DEBUG_DECRYPT flag: this is useful for debugging and testing where
1381 an MMA defence is not necessary.
1382 Thanks to Ivan Nestlerode <inestlerode@us.ibm.com> for discovering
1383 this issue. (CVE-2012-0884)
1386 *) Fix CVE-2011-4619: make sure we really are receiving a
1387 client hello before rejecting multiple SGC restarts. Thanks to
1388 Ivan Nestlerode <inestlerode@us.ibm.com> for discovering this bug.
1391 Changes between 1.0.0f and 1.0.0g [18 Jan 2012]
1393 *) Fix for DTLS DoS issue introduced by fix for CVE-2011-4109.
1394 Thanks to Antonio Martin, Enterprise Secure Access Research and
1395 Development, Cisco Systems, Inc. for discovering this bug and
1396 preparing a fix. (CVE-2012-0050)
1399 Changes between 1.0.0e and 1.0.0f [4 Jan 2012]
1401 *) Nadhem Alfardan and Kenny Paterson have discovered an extension
1402 of the Vaudenay padding oracle attack on CBC mode encryption
1403 which enables an efficient plaintext recovery attack against
1404 the OpenSSL implementation of DTLS. Their attack exploits timing
1405 differences arising during decryption processing. A research
1406 paper describing this attack can be found at:
1407 http://www.isg.rhul.ac.uk/~kp/dtls.pdf
1408 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
1409 Security Group at Royal Holloway, University of London
1410 (www.isg.rhul.ac.uk) for discovering this flaw and to Robin Seggelmann
1411 <seggelmann@fh-muenster.de> and Michael Tuexen <tuexen@fh-muenster.de>
1412 for preparing the fix. (CVE-2011-4108)
1413 [Robin Seggelmann, Michael Tuexen]
1415 *) Clear bytes used for block padding of SSL 3.0 records.
1417 [Adam Langley (Google)]
1419 *) Only allow one SGC handshake restart for SSL/TLS. Thanks to George
1420 Kadianakis <desnacked@gmail.com> for discovering this issue and
1421 Adam Langley for preparing the fix. (CVE-2011-4619)
1422 [Adam Langley (Google)]
1424 *) Check parameters are not NULL in GOST ENGINE. (CVE-2012-0027)
1425 [Andrey Kulikov <amdeich@gmail.com>]
1427 *) Prevent malformed RFC3779 data triggering an assertion failure.
1428 Thanks to Andrew Chi, BBN Technologies, for discovering the flaw
1429 and Rob Austein <sra@hactrn.net> for fixing it. (CVE-2011-4577)
1430 [Rob Austein <sra@hactrn.net>]
1432 *) Improved PRNG seeding for VOS.
1433 [Paul Green <Paul.Green@stratus.com>]
1435 *) Fix ssl_ciph.c set-up race.
1436 [Adam Langley (Google)]
1438 *) Fix spurious failures in ecdsatest.c.
1439 [Emilia Käsper (Google)]
1441 *) Fix the BIO_f_buffer() implementation (which was mixing different
1442 interpretations of the '..._len' fields).
1443 [Adam Langley (Google)]
1445 *) Fix handling of BN_BLINDING: now BN_BLINDING_invert_ex (rather than
1446 BN_BLINDING_invert_ex) calls BN_BLINDING_update, ensuring that concurrent
1447 threads won't reuse the same blinding coefficients.
1449 This also avoids the need to obtain the CRYPTO_LOCK_RSA_BLINDING
1450 lock to call BN_BLINDING_invert_ex, and avoids one use of
1451 BN_BLINDING_update for each BN_BLINDING structure (previously,
1452 the last update always remained unused).
1453 [Emilia Käsper (Google)]
1455 *) In ssl3_clear, preserve s3->init_extra along with s3->rbuf.
1456 [Bob Buckholz (Google)]
1458 Changes between 1.0.0d and 1.0.0e [6 Sep 2011]
1460 *) Fix bug where CRLs with nextUpdate in the past are sometimes accepted
1461 by initialising X509_STORE_CTX properly. (CVE-2011-3207)
1462 [Kaspar Brand <ossl@velox.ch>]
1464 *) Fix SSL memory handling for (EC)DH ciphersuites, in particular
1465 for multi-threaded use of ECDH. (CVE-2011-3210)
1466 [Adam Langley (Google)]
1468 *) Fix x509_name_ex_d2i memory leak on bad inputs.
1471 *) Remove hard coded ecdsaWithSHA1 signature tests in ssl code and check
1472 signature public key algorithm by using OID xref utilities instead.
1473 Before this you could only use some ECC ciphersuites with SHA1 only.
1476 *) Add protection against ECDSA timing attacks as mentioned in the paper
1477 by Billy Bob Brumley and Nicola Tuveri, see:
1479 http://eprint.iacr.org/2011/232.pdf
1481 [Billy Bob Brumley and Nicola Tuveri]
1483 Changes between 1.0.0c and 1.0.0d [8 Feb 2011]
1485 *) Fix parsing of OCSP stapling ClientHello extension. CVE-2011-0014
1486 [Neel Mehta, Adam Langley, Bodo Moeller (Google)]
1488 *) Fix bug in string printing code: if *any* escaping is enabled we must
1489 escape the escape character (backslash) or the resulting string is
1493 Changes between 1.0.0b and 1.0.0c [2 Dec 2010]
1495 *) Disable code workaround for ancient and obsolete Netscape browsers
1496 and servers: an attacker can use it in a ciphersuite downgrade attack.
1497 Thanks to Martin Rex for discovering this bug. CVE-2010-4180
1500 *) Fixed J-PAKE implementation error, originally discovered by
1501 Sebastien Martini, further info and confirmation from Stefan
1502 Arentz and Feng Hao. Note that this fix is a security fix. CVE-2010-4252
1505 Changes between 1.0.0a and 1.0.0b [16 Nov 2010]
1507 *) Fix extension code to avoid race conditions which can result in a buffer
1508 overrun vulnerability: resumed sessions must not be modified as they can
1509 be shared by multiple threads. CVE-2010-3864
1512 *) Fix WIN32 build system to correctly link an ENGINE directory into
1516 Changes between 1.0.0 and 1.0.0a [01 Jun 2010]
1518 *) Check return value of int_rsa_verify in pkey_rsa_verifyrecover
1520 [Steve Henson, Peter-Michael Hager <hager@dortmund.net>]
1522 Changes between 0.9.8n and 1.0.0 [29 Mar 2010]
1524 *) Add "missing" function EVP_CIPHER_CTX_copy(). This copies a cipher
1525 context. The operation can be customised via the ctrl mechanism in
1526 case ENGINEs want to include additional functionality.
1529 *) Tolerate yet another broken PKCS#8 key format: private key value negative.
1532 *) Add new -subject_hash_old and -issuer_hash_old options to x509 utility to
1533 output hashes compatible with older versions of OpenSSL.
1534 [Willy Weisz <weisz@vcpc.univie.ac.at>]
1536 *) Fix compression algorithm handling: if resuming a session use the
1537 compression algorithm of the resumed session instead of determining
1538 it from client hello again. Don't allow server to change algorithm.
1541 *) Add load_crls() function to apps tidying load_certs() too. Add option
1542 to verify utility to allow additional CRLs to be included.
1545 *) Update OCSP request code to permit adding custom headers to the request:
1546 some responders need this.
1549 *) The function EVP_PKEY_sign() returns <=0 on error: check return code
1551 [Julia Lawall <julia@diku.dk>]
1553 *) Update verify callback code in apps/s_cb.c and apps/verify.c, it
1554 needlessly dereferenced structures, used obsolete functions and
1555 didn't handle all updated verify codes correctly.
1558 *) Disable MD2 in the default configuration.
1561 *) In BIO_pop() and BIO_push() use the ctrl argument (which was NULL) to
1562 indicate the initial BIO being pushed or popped. This makes it possible
1563 to determine whether the BIO is the one explicitly called or as a result
1564 of the ctrl being passed down the chain. Fix BIO_pop() and SSL BIOs so
1565 it handles reference counts correctly and doesn't zero out the I/O bio
1566 when it is not being explicitly popped. WARNING: applications which
1567 included workarounds for the old buggy behaviour will need to be modified
1568 or they could free up already freed BIOs.
1571 *) Extend the uni2asc/asc2uni => OPENSSL_uni2asc/OPENSSL_asc2uni
1572 renaming to all platforms (within the 0.9.8 branch, this was
1573 done conditionally on Netware platforms to avoid a name clash).
1574 [Guenter <lists@gknw.net>]
1576 *) Add ECDHE and PSK support to DTLS.
1577 [Michael Tuexen <tuexen@fh-muenster.de>]
1579 *) Add CHECKED_STACK_OF macro to safestack.h, otherwise safestack can't
1583 *) Add "missing" function EVP_MD_flags() (without this the only way to
1584 retrieve a digest flags is by accessing the structure directly. Update
1585 EVP_MD_do_all*() and EVP_CIPHER_do_all*() to include the name a digest
1586 or cipher is registered as in the "from" argument. Print out all
1587 registered digests in the dgst usage message instead of manually
1588 attempting to work them out.
1591 *) If no SSLv2 ciphers are used don't use an SSLv2 compatible client hello:
1592 this allows the use of compression and extensions. Change default cipher
1593 string to remove SSLv2 ciphersuites. This effectively avoids ancient SSLv2
1594 by default unless an application cipher string requests it.
1597 *) Alter match criteria in PKCS12_parse(). It used to try to use local
1598 key ids to find matching certificates and keys but some PKCS#12 files
1599 don't follow the (somewhat unwritten) rules and this strategy fails.
1600 Now just gather all certificates together and the first private key
1601 then look for the first certificate that matches the key.
1604 *) Support use of registered digest and cipher names for dgst and cipher
1605 commands instead of having to add each one as a special case. So now
1612 openssl dgst -sha256 foo
1614 and this works for ENGINE based algorithms too.
1618 *) Update Gost ENGINE to support parameter files.
1619 [Victor B. Wagner <vitus@cryptocom.ru>]
1621 *) Support GeneralizedTime in ca utility.
1622 [Oliver Martin <oliver@volatilevoid.net>, Steve Henson]
1624 *) Enhance the hash format used for certificate directory links. The new
1625 form uses the canonical encoding (meaning equivalent names will work
1626 even if they aren't identical) and uses SHA1 instead of MD5. This form
1627 is incompatible with the older format and as a result c_rehash should
1628 be used to rebuild symbolic links.
1631 *) Make PKCS#8 the default write format for private keys, replacing the
1632 traditional format. This form is standardised, more secure and doesn't
1633 include an implicit MD5 dependency.
1636 *) Add a $gcc_devteam_warn option to Configure. The idea is that any code
1637 committed to OpenSSL should pass this lot as a minimum.
1640 *) Add session ticket override functionality for use by EAP-FAST.
1641 [Jouni Malinen <j@w1.fi>]
1643 *) Modify HMAC functions to return a value. Since these can be implemented
1644 in an ENGINE errors can occur.
1647 *) Type-checked OBJ_bsearch_ex.
1650 *) Type-checked OBJ_bsearch. Also some constification necessitated
1651 by type-checking. Still to come: TXT_DB, bsearch(?),
1652 OBJ_bsearch_ex, qsort, CRYPTO_EX_DATA, ASN1_VALUE, ASN1_STRING,
1656 *) New function OPENSSL_gmtime_adj() to add a specific number of days and
1657 seconds to a tm structure directly, instead of going through OS
1658 specific date routines. This avoids any issues with OS routines such
1659 as the year 2038 bug. New *_adj() functions for ASN1 time structures
1660 and X509_time_adj_ex() to cover the extended range. The existing
1661 X509_time_adj() is still usable and will no longer have any date issues.
1664 *) Delta CRL support. New use deltas option which will attempt to locate
1665 and search any appropriate delta CRLs available.
1667 This work was sponsored by Google.
1670 *) Support for CRLs partitioned by reason code. Reorganise CRL processing
1671 code and add additional score elements. Validate alternate CRL paths
1672 as part of the CRL checking and indicate a new error "CRL path validation
1673 error" in this case. Applications wanting additional details can use
1674 the verify callback and check the new "parent" field. If this is not
1675 NULL CRL path validation is taking place. Existing applications wont
1676 see this because it requires extended CRL support which is off by
1679 This work was sponsored by Google.
1682 *) Support for freshest CRL extension.
1684 This work was sponsored by Google.
1687 *) Initial indirect CRL support. Currently only supported in the CRLs
1688 passed directly and not via lookup. Process certificate issuer
1689 CRL entry extension and lookup CRL entries by bother issuer name
1690 and serial number. Check and process CRL issuer entry in IDP extension.
1692 This work was sponsored by Google.
1695 *) Add support for distinct certificate and CRL paths. The CRL issuer
1696 certificate is validated separately in this case. Only enabled if
1697 an extended CRL support flag is set: this flag will enable additional
1698 CRL functionality in future.
1700 This work was sponsored by Google.
1703 *) Add support for policy mappings extension.
1705 This work was sponsored by Google.
1708 *) Fixes to pathlength constraint, self issued certificate handling,
1709 policy processing to align with RFC3280 and PKITS tests.
1711 This work was sponsored by Google.
1714 *) Support for name constraints certificate extension. DN, email, DNS
1715 and URI types are currently supported.
1717 This work was sponsored by Google.
1720 *) To cater for systems that provide a pointer-based thread ID rather
1721 than numeric, deprecate the current numeric thread ID mechanism and
1722 replace it with a structure and associated callback type. This
1723 mechanism allows a numeric "hash" to be extracted from a thread ID in
1724 either case, and on platforms where pointers are larger than 'long',
1725 mixing is done to help ensure the numeric 'hash' is usable even if it
1726 can't be guaranteed unique. The default mechanism is to use "&errno"
1727 as a pointer-based thread ID to distinguish between threads.
1729 Applications that want to provide their own thread IDs should now use
1730 CRYPTO_THREADID_set_callback() to register a callback that will call
1731 either CRYPTO_THREADID_set_numeric() or CRYPTO_THREADID_set_pointer().
1733 Note that ERR_remove_state() is now deprecated, because it is tied
1734 to the assumption that thread IDs are numeric. ERR_remove_state(0)
1735 to free the current thread's error state should be replaced by
1736 ERR_remove_thread_state(NULL).
1738 (This new approach replaces the functions CRYPTO_set_idptr_callback(),
1739 CRYPTO_get_idptr_callback(), and CRYPTO_thread_idptr() that existed in
1740 OpenSSL 0.9.9-dev between June 2006 and August 2008. Also, if an
1741 application was previously providing a numeric thread callback that
1742 was inappropriate for distinguishing threads, then uniqueness might
1743 have been obtained with &errno that happened immediately in the
1744 intermediate development versions of OpenSSL; this is no longer the
1745 case, the numeric thread callback will now override the automatic use
1747 [Geoff Thorpe, with help from Bodo Moeller]
1749 *) Initial support for different CRL issuing certificates. This covers a
1750 simple case where the self issued certificates in the chain exist and
1751 the real CRL issuer is higher in the existing chain.
1753 This work was sponsored by Google.
1756 *) Removed effectively defunct crypto/store from the build.
1759 *) Revamp of STACK to provide stronger type-checking. Still to come:
1760 TXT_DB, bsearch(?), OBJ_bsearch, qsort, CRYPTO_EX_DATA, ASN1_VALUE,
1761 ASN1_STRING, CONF_VALUE.
1764 *) Add a new SSL_MODE_RELEASE_BUFFERS mode flag to release unused buffer
1765 RAM on SSL connections. This option can save about 34k per idle SSL.
1768 *) Revamp of LHASH to provide stronger type-checking. Still to come:
1769 STACK, TXT_DB, bsearch, qsort.
1772 *) Initial support for Cryptographic Message Syntax (aka CMS) based
1773 on RFC3850, RFC3851 and RFC3852. New cms directory and cms utility,
1774 support for data, signedData, compressedData, digestedData and
1775 encryptedData, envelopedData types included. Scripts to check against
1776 RFC4134 examples draft and interop and consistency checks of many
1777 content types and variants.
1780 *) Add options to enc utility to support use of zlib compression BIO.
1783 *) Extend mk1mf to support importing of options and assembly language
1784 files from Configure script, currently only included in VC-WIN32.
1785 The assembly language rules can now optionally generate the source
1786 files from the associated perl scripts.
1789 *) Implement remaining functionality needed to support GOST ciphersuites.
1790 Interop testing has been performed using CryptoPro implementations.
1791 [Victor B. Wagner <vitus@cryptocom.ru>]
1793 *) s390x assembler pack.
1796 *) ARMv4 assembler pack. ARMv4 refers to v4 and later ISA, not CPU
1800 *) Implement Opaque PRF Input TLS extension as specified in
1801 draft-rescorla-tls-opaque-prf-input-00.txt. Since this is not an
1802 official specification yet and no extension type assignment by
1803 IANA exists, this extension (for now) will have to be explicitly
1804 enabled when building OpenSSL by providing the extension number
1805 to use. For example, specify an option
1807 -DTLSEXT_TYPE_opaque_prf_input=0x9527
1809 to the "config" or "Configure" script to enable the extension,
1810 assuming extension number 0x9527 (which is a completely arbitrary
1811 and unofficial assignment based on the MD5 hash of the Internet
1812 Draft). Note that by doing so, you potentially lose
1813 interoperability with other TLS implementations since these might
1814 be using the same extension number for other purposes.
1816 SSL_set_tlsext_opaque_prf_input(ssl, src, len) is used to set the
1817 opaque PRF input value to use in the handshake. This will create
1818 an interal copy of the length-'len' string at 'src', and will
1819 return non-zero for success.
1821 To get more control and flexibility, provide a callback function
1824 SSL_CTX_set_tlsext_opaque_prf_input_callback(ctx, cb)
1825 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg(ctx, arg)
1829 int (*cb)(SSL *, void *peerinput, size_t len, void *arg);
1832 Callback function 'cb' will be called in handshakes, and is
1833 expected to use SSL_set_tlsext_opaque_prf_input() as appropriate.
1834 Argument 'arg' is for application purposes (the value as given to
1835 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg() will directly
1836 be provided to the callback function). The callback function
1837 has to return non-zero to report success: usually 1 to use opaque
1838 PRF input just if possible, or 2 to enforce use of the opaque PRF
1839 input. In the latter case, the library will abort the handshake
1840 if opaque PRF input is not successfully negotiated.
1842 Arguments 'peerinput' and 'len' given to the callback function
1843 will always be NULL and 0 in the case of a client. A server will
1844 see the client's opaque PRF input through these variables if
1845 available (NULL and 0 otherwise). Note that if the server
1846 provides an opaque PRF input, the length must be the same as the
1847 length of the client's opaque PRF input.
1849 Note that the callback function will only be called when creating
1850 a new session (session resumption can resume whatever was
1851 previously negotiated), and will not be called in SSL 2.0
1852 handshakes; thus, SSL_CTX_set_options(ctx, SSL_OP_NO_SSLv2) or
1853 SSL_set_options(ssl, SSL_OP_NO_SSLv2) is especially recommended
1854 for applications that need to enforce opaque PRF input.
1858 *) Update ssl code to support digests other than SHA1+MD5 for handshake
1861 [Victor B. Wagner <vitus@cryptocom.ru>]
1863 *) Add RFC4507 support to OpenSSL. This includes the corrections in
1864 RFC4507bis. The encrypted ticket format is an encrypted encoded
1865 SSL_SESSION structure, that way new session features are automatically
1868 If a client application caches session in an SSL_SESSION structure
1869 support is transparent because tickets are now stored in the encoded
1872 The SSL_CTX structure automatically generates keys for ticket
1873 protection in servers so again support should be possible
1874 with no application modification.
1876 If a client or server wishes to disable RFC4507 support then the option
1877 SSL_OP_NO_TICKET can be set.
1879 Add a TLS extension debugging callback to allow the contents of any client
1880 or server extensions to be examined.
1882 This work was sponsored by Google.
1885 *) Final changes to avoid use of pointer pointer casts in OpenSSL.
1886 OpenSSL should now compile cleanly on gcc 4.2
1887 [Peter Hartley <pdh@utter.chaos.org.uk>, Steve Henson]
1889 *) Update SSL library to use new EVP_PKEY MAC API. Include generic MAC
1890 support including streaming MAC support: this is required for GOST
1891 ciphersuite support.
1892 [Victor B. Wagner <vitus@cryptocom.ru>, Steve Henson]
1894 *) Add option -stream to use PKCS#7 streaming in smime utility. New
1895 function i2d_PKCS7_bio_stream() and PEM_write_PKCS7_bio_stream()
1896 to output in BER and PEM format.
1899 *) Experimental support for use of HMAC via EVP_PKEY interface. This
1900 allows HMAC to be handled via the EVP_DigestSign*() interface. The
1901 EVP_PKEY "key" in this case is the HMAC key, potentially allowing
1902 ENGINE support for HMAC keys which are unextractable. New -mac and
1903 -macopt options to dgst utility.
1906 *) New option -sigopt to dgst utility. Update dgst to use
1907 EVP_Digest{Sign,Verify}*. These two changes make it possible to use
1908 alternative signing paramaters such as X9.31 or PSS in the dgst
1912 *) Change ssl_cipher_apply_rule(), the internal function that does
1913 the work each time a ciphersuite string requests enabling
1914 ("foo+bar"), moving ("+foo+bar"), disabling ("-foo+bar", or
1915 removing ("!foo+bar") a class of ciphersuites: Now it maintains
1916 the order of disabled ciphersuites such that those ciphersuites
1917 that most recently went from enabled to disabled not only stay
1918 in order with respect to each other, but also have higher priority
1919 than other disabled ciphersuites the next time ciphersuites are
1922 This means that you can now say, e.g., "PSK:-PSK:HIGH" to enable
1923 the same ciphersuites as with "HIGH" alone, but in a specific
1924 order where the PSK ciphersuites come first (since they are the
1925 most recently disabled ciphersuites when "HIGH" is parsed).
1927 Also, change ssl_create_cipher_list() (using this new
1928 funcionality) such that between otherwise identical
1929 cihpersuites, ephemeral ECDH is preferred over ephemeral DH in
1933 *) Change ssl_create_cipher_list() so that it automatically
1934 arranges the ciphersuites in reasonable order before starting
1935 to process the rule string. Thus, the definition for "DEFAULT"
1936 (SSL_DEFAULT_CIPHER_LIST) now is just "ALL:!aNULL:!eNULL", but
1937 remains equivalent to "AES:ALL:!aNULL:!eNULL:+aECDH:+kRSA:+RC4:@STRENGTH".
1938 This makes it much easier to arrive at a reasonable default order
1939 in applications for which anonymous ciphers are OK (meaning
1940 that you can't actually use DEFAULT).
1941 [Bodo Moeller; suggested by Victor Duchovni]
1943 *) Split the SSL/TLS algorithm mask (as used for ciphersuite string
1944 processing) into multiple integers instead of setting
1945 "SSL_MKEY_MASK" bits, "SSL_AUTH_MASK" bits, "SSL_ENC_MASK",
1946 "SSL_MAC_MASK", and "SSL_SSL_MASK" bits all in a single integer.
1947 (These masks as well as the individual bit definitions are hidden
1948 away into the non-exported interface ssl/ssl_locl.h, so this
1949 change to the definition of the SSL_CIPHER structure shouldn't
1950 affect applications.) This give us more bits for each of these
1951 categories, so there is no longer a need to coagulate AES128 and
1952 AES256 into a single algorithm bit, and to coagulate Camellia128
1953 and Camellia256 into a single algorithm bit, which has led to all
1956 Thus, among other things, the kludge introduced in 0.9.7m and
1957 0.9.8e for masking out AES256 independently of AES128 or masking
1958 out Camellia256 independently of AES256 is not needed here in 0.9.9.
1960 With the change, we also introduce new ciphersuite aliases that
1961 so far were missing: "AES128", "AES256", "CAMELLIA128", and
1965 *) Add support for dsa-with-SHA224 and dsa-with-SHA256.
1966 Use the leftmost N bytes of the signature input if the input is
1967 larger than the prime q (with N being the size in bytes of q).
1970 *) Very *very* experimental PKCS#7 streaming encoder support. Nothing uses
1971 it yet and it is largely untested.
1974 *) Add support for the ecdsa-with-SHA224/256/384/512 signature types.
1977 *) Initial incomplete changes to avoid need for function casts in OpenSSL
1978 some compilers (gcc 4.2 and later) reject their use. Safestack is
1979 reimplemented. Update ASN1 to avoid use of legacy functions.
1982 *) Win32/64 targets are linked with Winsock2.
1985 *) Add an X509_CRL_METHOD structure to allow CRL processing to be redirected
1986 to external functions. This can be used to increase CRL handling
1987 efficiency especially when CRLs are very large by (for example) storing
1988 the CRL revoked certificates in a database.
1991 *) Overhaul of by_dir code. Add support for dynamic loading of CRLs so
1992 new CRLs added to a directory can be used. New command line option
1993 -verify_return_error to s_client and s_server. This causes real errors
1994 to be returned by the verify callback instead of carrying on no matter
1995 what. This reflects the way a "real world" verify callback would behave.
1998 *) GOST engine, supporting several GOST algorithms and public key formats.
1999 Kindly donated by Cryptocom.
2002 *) Partial support for Issuing Distribution Point CRL extension. CRLs
2003 partitioned by DP are handled but no indirect CRL or reason partitioning
2004 (yet). Complete overhaul of CRL handling: now the most suitable CRL is
2005 selected via a scoring technique which handles IDP and AKID in CRLs.
2008 *) New X509_STORE_CTX callbacks lookup_crls() and lookup_certs() which
2009 will ultimately be used for all verify operations: this will remove the
2010 X509_STORE dependency on certificate verification and allow alternative
2011 lookup methods. X509_STORE based implementations of these two callbacks.
2014 *) Allow multiple CRLs to exist in an X509_STORE with matching issuer names.
2015 Modify get_crl() to find a valid (unexpired) CRL if possible.
2018 *) New function X509_CRL_match() to check if two CRLs are identical. Normally
2019 this would be called X509_CRL_cmp() but that name is already used by
2020 a function that just compares CRL issuer names. Cache several CRL
2021 extensions in X509_CRL structure and cache CRLDP in X509.
2024 *) Store a "canonical" representation of X509_NAME structure (ASN1 Name)
2025 this maps equivalent X509_NAME structures into a consistent structure.
2026 Name comparison can then be performed rapidly using memcmp().
2029 *) Non-blocking OCSP request processing. Add -timeout option to ocsp
2033 *) Allow digests to supply their own micalg string for S/MIME type using
2034 the ctrl EVP_MD_CTRL_MICALG.
2037 *) During PKCS7 signing pass the PKCS7 SignerInfo structure to the
2038 EVP_PKEY_METHOD before and after signing via the EVP_PKEY_CTRL_PKCS7_SIGN
2039 ctrl. It can then customise the structure before and/or after signing
2043 *) New function OBJ_add_sigid() to allow application defined signature OIDs
2044 to be added to OpenSSLs internal tables. New function OBJ_sigid_free()
2045 to free up any added signature OIDs.
2048 *) New functions EVP_CIPHER_do_all(), EVP_CIPHER_do_all_sorted(),
2049 EVP_MD_do_all() and EVP_MD_do_all_sorted() to enumerate internal
2050 digest and cipher tables. New options added to openssl utility:
2051 list-message-digest-algorithms and list-cipher-algorithms.
2054 *) Change the array representation of binary polynomials: the list
2055 of degrees of non-zero coefficients is now terminated with -1.
2056 Previously it was terminated with 0, which was also part of the
2057 value; thus, the array representation was not applicable to
2058 polynomials where t^0 has coefficient zero. This change makes
2059 the array representation useful in a more general context.
2062 *) Various modifications and fixes to SSL/TLS cipher string
2063 handling. For ECC, the code now distinguishes between fixed ECDH
2064 with RSA certificates on the one hand and with ECDSA certificates
2065 on the other hand, since these are separate ciphersuites. The
2066 unused code for Fortezza ciphersuites has been removed.
2068 For consistency with EDH, ephemeral ECDH is now called "EECDH"
2069 (not "ECDHE"). For consistency with the code for DH
2070 certificates, use of ECDH certificates is now considered ECDH
2071 authentication, not RSA or ECDSA authentication (the latter is
2072 merely the CA's signing algorithm and not actively used in the
2075 The temporary ciphersuite alias "ECCdraft" is no longer
2076 available, and ECC ciphersuites are no longer excluded from "ALL"
2077 and "DEFAULT". The following aliases now exist for RFC 4492
2078 ciphersuites, most of these by analogy with the DH case:
2080 kECDHr - ECDH cert, signed with RSA
2081 kECDHe - ECDH cert, signed with ECDSA
2082 kECDH - ECDH cert (signed with either RSA or ECDSA)
2083 kEECDH - ephemeral ECDH
2084 ECDH - ECDH cert or ephemeral ECDH
2090 AECDH - anonymous ECDH
2091 EECDH - non-anonymous ephemeral ECDH (equivalent to "kEECDH:-AECDH")
2095 *) Add additional S/MIME capabilities for AES and GOST ciphers if supported.
2096 Use correct micalg parameters depending on digest(s) in signed message.
2099 *) Add engine support for EVP_PKEY_ASN1_METHOD. Add functions to process
2100 an ENGINE asn1 method. Support ENGINE lookups in the ASN1 code.
2103 *) Initial engine support for EVP_PKEY_METHOD. New functions to permit
2104 an engine to register a method. Add ENGINE lookups for methods and
2105 functional reference processing.
2108 *) New functions EVP_Digest{Sign,Verify)*. These are enchance versions of
2109 EVP_{Sign,Verify}* which allow an application to customise the signature
2113 *) New -resign option to smime utility. This adds one or more signers
2114 to an existing PKCS#7 signedData structure. Also -md option to use an
2115 alternative message digest algorithm for signing.
2118 *) Tidy up PKCS#7 routines and add new functions to make it easier to
2119 create PKCS7 structures containing multiple signers. Update smime
2120 application to support multiple signers.
2123 *) New -macalg option to pkcs12 utility to allow setting of an alternative
2127 *) Initial support for PKCS#5 v2.0 PRFs other than default SHA1 HMAC.
2128 Reorganize PBE internals to lookup from a static table using NIDs,
2129 add support for HMAC PBE OID translation. Add a EVP_CIPHER ctrl:
2130 EVP_CTRL_PBE_PRF_NID this allows a cipher to specify an alternative
2131 PRF which will be automatically used with PBES2.
2134 *) Replace the algorithm specific calls to generate keys in "req" with the
2138 *) Update PKCS#7 enveloped data routines to use new API. This is now
2139 supported by any public key method supporting the encrypt operation. A
2140 ctrl is added to allow the public key algorithm to examine or modify
2141 the PKCS#7 RecipientInfo structure if it needs to: for RSA this is
2145 *) Add a ctrl to asn1 method to allow a public key algorithm to express
2146 a default digest type to use. In most cases this will be SHA1 but some
2147 algorithms (such as GOST) need to specify an alternative digest. The
2148 return value indicates how strong the preference is 1 means optional and
2149 2 is mandatory (that is it is the only supported type). Modify
2150 ASN1_item_sign() to accept a NULL digest argument to indicate it should
2151 use the default md. Update openssl utilities to use the default digest
2152 type for signing if it is not explicitly indicated.
2155 *) Use OID cross reference table in ASN1_sign() and ASN1_verify(). New
2156 EVP_MD flag EVP_MD_FLAG_PKEY_METHOD_SIGNATURE. This uses the relevant
2157 signing method from the key type. This effectively removes the link
2158 between digests and public key types.
2161 *) Add an OID cross reference table and utility functions. Its purpose is to
2162 translate between signature OIDs such as SHA1WithrsaEncryption and SHA1,
2163 rsaEncryption. This will allow some of the algorithm specific hackery
2164 needed to use the correct OID to be removed.
2167 *) Remove algorithm specific dependencies when setting PKCS7_SIGNER_INFO
2168 structures for PKCS7_sign(). They are now set up by the relevant public
2172 *) Add provisional EC pkey method with support for ECDSA and ECDH.
2175 *) Add support for key derivation (agreement) in the API, DH method and
2179 *) Add DSA pkey method and DH pkey methods, extend DH ASN1 method to support
2180 public and private key formats. As a side effect these add additional
2181 command line functionality not previously available: DSA signatures can be
2182 generated and verified using pkeyutl and DH key support and generation in
2187 [Oliver Tappe <zooey@hirschkaefer.de>]
2189 *) New make target "install_html_docs" installs HTML renditions of the
2191 [Oliver Tappe <zooey@hirschkaefer.de>]
2193 *) New utility "genpkey" this is analogous to "genrsa" etc except it can
2194 generate keys for any algorithm. Extend and update EVP_PKEY_METHOD to
2195 support key and parameter generation and add initial key generation
2196 functionality for RSA.
2199 *) Add functions for main EVP_PKEY_method operations. The undocumented
2200 functions EVP_PKEY_{encrypt,decrypt} have been renamed to
2201 EVP_PKEY_{encrypt,decrypt}_old.
2204 *) Initial definitions for EVP_PKEY_METHOD. This will be a high level public
2205 key API, doesn't do much yet.
2208 *) New function EVP_PKEY_asn1_get0_info() to retrieve information about
2209 public key algorithms. New option to openssl utility:
2210 "list-public-key-algorithms" to print out info.
2213 *) Implement the Supported Elliptic Curves Extension for
2214 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
2217 *) Don't free up OIDs in OBJ_cleanup() if they are in use by EVP_MD or
2218 EVP_CIPHER structures to avoid later problems in EVP_cleanup().
2221 *) New utilities pkey and pkeyparam. These are similar to algorithm specific
2222 utilities such as rsa, dsa, dsaparam etc except they process any key
2226 *) Transfer public key printing routines to EVP_PKEY_ASN1_METHOD. New
2227 functions EVP_PKEY_print_public(), EVP_PKEY_print_private(),
2228 EVP_PKEY_print_param() to print public key data from an EVP_PKEY
2232 *) Initial support for pluggable public key ASN1.
2233 De-spaghettify the public key ASN1 handling. Move public and private
2234 key ASN1 handling to a new EVP_PKEY_ASN1_METHOD structure. Relocate
2235 algorithm specific handling to a single module within the relevant
2236 algorithm directory. Add functions to allow (near) opaque processing
2237 of public and private key structures.
2240 *) Implement the Supported Point Formats Extension for
2241 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
2244 *) Add initial support for RFC 4279 PSK TLS ciphersuites. Add members
2245 for the psk identity [hint] and the psk callback functions to the
2246 SSL_SESSION, SSL and SSL_CTX structure.
2249 PSK-RC4-SHA, PSK-3DES-EDE-CBC-SHA, PSK-AES128-CBC-SHA,
2253 SSL_CTX_use_psk_identity_hint
2254 SSL_get_psk_identity_hint
2255 SSL_get_psk_identity
2256 SSL_use_psk_identity_hint
2258 [Mika Kousa and Pasi Eronen of Nokia Corporation]
2260 *) Add RFC 3161 compliant time stamp request creation, response generation
2261 and response verification functionality.
2262 [Zoltán Glózik <zglozik@opentsa.org>, The OpenTSA Project]
2264 *) Add initial support for TLS extensions, specifically for the server_name
2265 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
2266 have new members for a host name. The SSL data structure has an
2267 additional member SSL_CTX *initial_ctx so that new sessions can be
2268 stored in that context to allow for session resumption, even after the
2269 SSL has been switched to a new SSL_CTX in reaction to a client's
2270 server_name extension.
2272 New functions (subject to change):
2274 SSL_get_servername()
2275 SSL_get_servername_type()
2278 New CTRL codes and macros (subject to change):
2280 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
2281 - SSL_CTX_set_tlsext_servername_callback()
2282 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
2283 - SSL_CTX_set_tlsext_servername_arg()
2284 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
2286 openssl s_client has a new '-servername ...' option.
2288 openssl s_server has new options '-servername_host ...', '-cert2 ...',
2289 '-key2 ...', '-servername_fatal' (subject to change). This allows
2290 testing the HostName extension for a specific single host name ('-cert'
2291 and '-key' remain fallbacks for handshakes without HostName
2292 negotiation). If the unrecognized_name alert has to be sent, this by
2293 default is a warning; it becomes fatal with the '-servername_fatal'
2296 [Peter Sylvester, Remy Allais, Christophe Renou]
2298 *) Whirlpool hash implementation is added.
2301 *) BIGNUM code on 64-bit SPARCv9 targets is switched from bn(64,64) to
2302 bn(64,32). Because of instruction set limitations it doesn't have
2303 any negative impact on performance. This was done mostly in order
2304 to make it possible to share assembler modules, such as bn_mul_mont
2305 implementations, between 32- and 64-bit builds without hassle.
2308 *) Move code previously exiled into file crypto/ec/ec2_smpt.c
2309 to ec2_smpl.c, and no longer require the OPENSSL_EC_BIN_PT_COMP
2313 *) New candidate for BIGNUM assembler implementation, bn_mul_mont,
2314 dedicated Montgomery multiplication procedure, is introduced.
2315 BN_MONT_CTX is modified to allow bn_mul_mont to reach for higher
2316 "64-bit" performance on certain 32-bit targets.
2319 *) New option SSL_OP_NO_COMP to disable use of compression selectively
2320 in SSL structures. New SSL ctrl to set maximum send fragment size.
2321 Save memory by seeting the I/O buffer sizes dynamically instead of
2322 using the maximum available value.
2325 *) New option -V for 'openssl ciphers'. This prints the ciphersuite code
2326 in addition to the text details.
2329 *) Very, very preliminary EXPERIMENTAL support for printing of general
2330 ASN1 structures. This currently produces rather ugly output and doesn't
2331 handle several customised structures at all.
2334 *) Integrated support for PVK file format and some related formats such
2335 as MS PUBLICKEYBLOB and PRIVATEKEYBLOB. Command line switches to support
2336 these in the 'rsa' and 'dsa' utilities.
2339 *) Support for PKCS#1 RSAPublicKey format on rsa utility command line.
2342 *) Remove the ancient ASN1_METHOD code. This was only ever used in one
2343 place for the (very old) "NETSCAPE" format certificates which are now
2344 handled using new ASN1 code equivalents.
2347 *) Let the TLSv1_method() etc. functions return a 'const' SSL_METHOD
2348 pointer and make the SSL_METHOD parameter in SSL_CTX_new,
2349 SSL_CTX_set_ssl_version and SSL_set_ssl_method 'const'.
2352 *) Modify CRL distribution points extension code to print out previously
2353 unsupported fields. Enhance extension setting code to allow setting of
2357 *) Add print and set support for Issuing Distribution Point CRL extension.
2360 *) Change 'Configure' script to enable Camellia by default.
2363 Changes between 0.9.8x and 0.9.8y [5 Feb 2013]
2365 *) Make the decoding of SSLv3, TLS and DTLS CBC records constant time.
2367 This addresses the flaw in CBC record processing discovered by
2368 Nadhem Alfardan and Kenny Paterson. Details of this attack can be found
2369 at: http://www.isg.rhul.ac.uk/tls/
2371 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
2372 Security Group at Royal Holloway, University of London
2373 (www.isg.rhul.ac.uk) for discovering this flaw and Adam Langley and
2374 Emilia Käsper for the initial patch.
2376 [Emilia Käsper, Adam Langley, Ben Laurie, Andy Polyakov, Steve Henson]
2378 *) Return an error when checking OCSP signatures when key is NULL.
2379 This fixes a DoS attack. (CVE-2013-0166)
2382 *) Call OCSP Stapling callback after ciphersuite has been chosen, so
2383 the right response is stapled. Also change SSL_get_certificate()
2384 so it returns the certificate actually sent.
2385 See http://rt.openssl.org/Ticket/Display.html?id=2836.
2386 (This is a backport)
2387 [Rob Stradling <rob.stradling@comodo.com>]
2389 *) Fix possible deadlock when decoding public keys.
2392 Changes between 0.9.8w and 0.9.8x [10 May 2012]
2394 *) Sanity check record length before skipping explicit IV in DTLS
2397 Thanks to Codenomicon for discovering this issue using Fuzz-o-Matic
2398 fuzzing as a service testing platform.
2402 *) Initialise tkeylen properly when encrypting CMS messages.
2403 Thanks to Solar Designer of Openwall for reporting this issue.
2406 Changes between 0.9.8v and 0.9.8w [23 Apr 2012]
2408 *) The fix for CVE-2012-2110 did not take into account that the
2409 'len' argument to BUF_MEM_grow and BUF_MEM_grow_clean is an
2410 int in OpenSSL 0.9.8, making it still vulnerable. Fix by
2411 rejecting negative len parameter. (CVE-2012-2131)
2412 [Tomas Hoger <thoger@redhat.com>]
2414 Changes between 0.9.8u and 0.9.8v [19 Apr 2012]
2416 *) Check for potentially exploitable overflows in asn1_d2i_read_bio
2417 BUF_mem_grow and BUF_mem_grow_clean. Refuse attempts to shrink buffer
2418 in CRYPTO_realloc_clean.
2420 Thanks to Tavis Ormandy, Google Security Team, for discovering this
2421 issue and to Adam Langley <agl@chromium.org> for fixing it.
2423 [Adam Langley (Google), Tavis Ormandy, Google Security Team]
2425 Changes between 0.9.8t and 0.9.8u [12 Mar 2012]
2427 *) Fix MMA (Bleichenbacher's attack on PKCS #1 v1.5 RSA padding) weakness
2428 in CMS and PKCS7 code. When RSA decryption fails use a random key for
2429 content decryption and always return the same error. Note: this attack
2430 needs on average 2^20 messages so it only affects automated senders. The
2431 old behaviour can be reenabled in the CMS code by setting the
2432 CMS_DEBUG_DECRYPT flag: this is useful for debugging and testing where
2433 an MMA defence is not necessary.
2434 Thanks to Ivan Nestlerode <inestlerode@us.ibm.com> for discovering
2435 this issue. (CVE-2012-0884)
2438 *) Fix CVE-2011-4619: make sure we really are receiving a
2439 client hello before rejecting multiple SGC restarts. Thanks to
2440 Ivan Nestlerode <inestlerode@us.ibm.com> for discovering this bug.
2443 Changes between 0.9.8s and 0.9.8t [18 Jan 2012]
2445 *) Fix for DTLS DoS issue introduced by fix for CVE-2011-4109.
2446 Thanks to Antonio Martin, Enterprise Secure Access Research and
2447 Development, Cisco Systems, Inc. for discovering this bug and
2448 preparing a fix. (CVE-2012-0050)
2451 Changes between 0.9.8r and 0.9.8s [4 Jan 2012]
2453 *) Nadhem Alfardan and Kenny Paterson have discovered an extension
2454 of the Vaudenay padding oracle attack on CBC mode encryption
2455 which enables an efficient plaintext recovery attack against
2456 the OpenSSL implementation of DTLS. Their attack exploits timing
2457 differences arising during decryption processing. A research
2458 paper describing this attack can be found at:
2459 http://www.isg.rhul.ac.uk/~kp/dtls.pdf
2460 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
2461 Security Group at Royal Holloway, University of London
2462 (www.isg.rhul.ac.uk) for discovering this flaw and to Robin Seggelmann
2463 <seggelmann@fh-muenster.de> and Michael Tuexen <tuexen@fh-muenster.de>
2464 for preparing the fix. (CVE-2011-4108)
2465 [Robin Seggelmann, Michael Tuexen]
2467 *) Stop policy check failure freeing same buffer twice. (CVE-2011-4109)
2468 [Ben Laurie, Kasper <ekasper@google.com>]
2470 *) Clear bytes used for block padding of SSL 3.0 records.
2472 [Adam Langley (Google)]
2474 *) Only allow one SGC handshake restart for SSL/TLS. Thanks to George
2475 Kadianakis <desnacked@gmail.com> for discovering this issue and
2476 Adam Langley for preparing the fix. (CVE-2011-4619)
2477 [Adam Langley (Google)]
2479 *) Prevent malformed RFC3779 data triggering an assertion failure.
2480 Thanks to Andrew Chi, BBN Technologies, for discovering the flaw
2481 and Rob Austein <sra@hactrn.net> for fixing it. (CVE-2011-4577)
2482 [Rob Austein <sra@hactrn.net>]
2484 *) Fix ssl_ciph.c set-up race.
2485 [Adam Langley (Google)]
2487 *) Fix spurious failures in ecdsatest.c.
2488 [Emilia Käsper (Google)]
2490 *) Fix the BIO_f_buffer() implementation (which was mixing different
2491 interpretations of the '..._len' fields).
2492 [Adam Langley (Google)]
2494 *) Fix handling of BN_BLINDING: now BN_BLINDING_invert_ex (rather than
2495 BN_BLINDING_invert_ex) calls BN_BLINDING_update, ensuring that concurrent
2496 threads won't reuse the same blinding coefficients.
2498 This also avoids the need to obtain the CRYPTO_LOCK_RSA_BLINDING
2499 lock to call BN_BLINDING_invert_ex, and avoids one use of
2500 BN_BLINDING_update for each BN_BLINDING structure (previously,
2501 the last update always remained unused).
2502 [Emilia Käsper (Google)]
2504 *) Fix SSL memory handling for (EC)DH ciphersuites, in particular
2505 for multi-threaded use of ECDH.
2506 [Adam Langley (Google)]
2508 *) Fix x509_name_ex_d2i memory leak on bad inputs.
2511 *) Add protection against ECDSA timing attacks as mentioned in the paper
2512 by Billy Bob Brumley and Nicola Tuveri, see:
2514 http://eprint.iacr.org/2011/232.pdf
2516 [Billy Bob Brumley and Nicola Tuveri]
2518 Changes between 0.9.8q and 0.9.8r [8 Feb 2011]
2520 *) Fix parsing of OCSP stapling ClientHello extension. CVE-2011-0014
2521 [Neel Mehta, Adam Langley, Bodo Moeller (Google)]
2523 *) Fix bug in string printing code: if *any* escaping is enabled we must
2524 escape the escape character (backslash) or the resulting string is
2528 Changes between 0.9.8p and 0.9.8q [2 Dec 2010]
2530 *) Disable code workaround for ancient and obsolete Netscape browsers
2531 and servers: an attacker can use it in a ciphersuite downgrade attack.
2532 Thanks to Martin Rex for discovering this bug. CVE-2010-4180
2535 *) Fixed J-PAKE implementation error, originally discovered by
2536 Sebastien Martini, further info and confirmation from Stefan
2537 Arentz and Feng Hao. Note that this fix is a security fix. CVE-2010-4252
2540 Changes between 0.9.8o and 0.9.8p [16 Nov 2010]
2542 *) Fix extension code to avoid race conditions which can result in a buffer
2543 overrun vulnerability: resumed sessions must not be modified as they can
2544 be shared by multiple threads. CVE-2010-3864
2547 *) Fix for double free bug in ssl/s3_clnt.c CVE-2010-2939
2550 *) Don't reencode certificate when calculating signature: cache and use
2551 the original encoding instead. This makes signature verification of
2552 some broken encodings work correctly.
2555 *) ec2_GF2m_simple_mul bugfix: compute correct result if the output EC_POINT
2556 is also one of the inputs.
2557 [Emilia Käsper <emilia.kasper@esat.kuleuven.be> (Google)]
2559 *) Don't repeatedly append PBE algorithms to table if they already exist.
2560 Sort table on each new add. This effectively makes the table read only
2561 after all algorithms are added and subsequent calls to PKCS12_pbe_add
2565 Changes between 0.9.8n and 0.9.8o [01 Jun 2010]
2567 [NB: OpenSSL 0.9.8o and later 0.9.8 patch levels were released after
2570 *) Correct a typo in the CMS ASN1 module which can result in invalid memory
2571 access or freeing data twice (CVE-2010-0742)
2572 [Steve Henson, Ronald Moesbergen <intercommit@gmail.com>]
2574 *) Add SHA2 algorithms to SSL_library_init(). SHA2 is becoming far more
2575 common in certificates and some applications which only call
2576 SSL_library_init and not OpenSSL_add_all_algorithms() will fail.
2580 Reduce copying into .apps and .test in makevms.com
2581 Don't try to use blank CA certificate in CA.com
2582 Allow use of C files from original directories in maketests.com
2583 [Steven M. Schweda" <sms@antinode.info>]
2585 Changes between 0.9.8m and 0.9.8n [24 Mar 2010]
2587 *) When rejecting SSL/TLS records due to an incorrect version number, never
2588 update s->server with a new major version number. As of
2589 - OpenSSL 0.9.8m if 'short' is a 16-bit type,
2590 - OpenSSL 0.9.8f if 'short' is longer than 16 bits,
2591 the previous behavior could result in a read attempt at NULL when
2592 receiving specific incorrect SSL/TLS records once record payload
2593 protection is active. (CVE-2010-0740)
2594 [Bodo Moeller, Adam Langley <agl@chromium.org>]
2596 *) Fix for CVE-2010-0433 where some kerberos enabled versions of OpenSSL
2597 could be crashed if the relevant tables were not present (e.g. chrooted).
2598 [Tomas Hoger <thoger@redhat.com>]
2600 Changes between 0.9.8l and 0.9.8m [25 Feb 2010]
2602 *) Always check bn_wexpend() return values for failure. (CVE-2009-3245)
2603 [Martin Olsson, Neel Mehta]
2605 *) Fix X509_STORE locking: Every 'objs' access requires a lock (to
2606 accommodate for stack sorting, always a write lock!).
2609 *) On some versions of WIN32 Heap32Next is very slow. This can cause
2610 excessive delays in the RAND_poll(): over a minute. As a workaround
2611 include a time check in the inner Heap32Next loop too.
2614 *) The code that handled flushing of data in SSL/TLS originally used the
2615 BIO_CTRL_INFO ctrl to see if any data was pending first. This caused
2616 the problem outlined in PR#1949. The fix suggested there however can
2617 trigger problems with buggy BIO_CTRL_WPENDING (e.g. some versions
2618 of Apache). So instead simplify the code to flush unconditionally.
2619 This should be fine since flushing with no data to flush is a no op.
2622 *) Handle TLS versions 2.0 and later properly and correctly use the
2623 highest version of TLS/SSL supported. Although TLS >= 2.0 is some way
2624 off ancient servers have a habit of sticking around for a while...
2627 *) Modify compression code so it frees up structures without using the
2628 ex_data callbacks. This works around a problem where some applications
2629 call CRYPTO_cleanup_all_ex_data() before application exit (e.g. when
2630 restarting) then use compression (e.g. SSL with compression) later.
2631 This results in significant per-connection memory leaks and
2632 has caused some security issues including CVE-2008-1678 and
2636 *) Constify crypto/cast (i.e., <openssl/cast.h>): a CAST_KEY doesn't
2637 change when encrypting or decrypting.
2640 *) Add option SSL_OP_LEGACY_SERVER_CONNECT which will allow clients to
2641 connect and renegotiate with servers which do not support RI.
2642 Until RI is more widely deployed this option is enabled by default.
2645 *) Add "missing" ssl ctrls to clear options and mode.
2648 *) If client attempts to renegotiate and doesn't support RI respond with
2649 a no_renegotiation alert as required by RFC5746. Some renegotiating
2650 TLS clients will continue a connection gracefully when they receive
2651 the alert. Unfortunately OpenSSL mishandled this alert and would hang
2652 waiting for a server hello which it will never receive. Now we treat a
2653 received no_renegotiation alert as a fatal error. This is because
2654 applications requesting a renegotiation might well expect it to succeed
2655 and would have no code in place to handle the server denying it so the
2656 only safe thing to do is to terminate the connection.
2659 *) Add ctrl macro SSL_get_secure_renegotiation_support() which returns 1 if
2660 peer supports secure renegotiation and 0 otherwise. Print out peer
2661 renegotiation support in s_client/s_server.
2664 *) Replace the highly broken and deprecated SPKAC certification method with
2665 the updated NID creation version. This should correctly handle UTF8.
2668 *) Implement RFC5746. Re-enable renegotiation but require the extension
2669 as needed. Unfortunately, SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION
2670 turns out to be a bad idea. It has been replaced by
2671 SSL_OP_ALLOW_UNSAFE_LEGACY_RENEGOTIATION which can be set with
2672 SSL_CTX_set_options(). This is really not recommended unless you
2673 know what you are doing.
2674 [Eric Rescorla <ekr@networkresonance.com>, Ben Laurie, Steve Henson]
2676 *) Fixes to stateless session resumption handling. Use initial_ctx when
2677 issuing and attempting to decrypt tickets in case it has changed during
2678 servername handling. Use a non-zero length session ID when attempting
2679 stateless session resumption: this makes it possible to determine if
2680 a resumption has occurred immediately after receiving server hello
2681 (several places in OpenSSL subtly assume this) instead of later in
2685 *) The functions ENGINE_ctrl(), OPENSSL_isservice(),
2686 CMS_get1_RecipientRequest() and RAND_bytes() can return <=0 on error
2687 fixes for a few places where the return code is not checked
2689 [Julia Lawall <julia@diku.dk>]
2691 *) Add --strict-warnings option to Configure script to include devteam
2692 warnings in other configurations.
2695 *) Add support for --libdir option and LIBDIR variable in makefiles. This
2696 makes it possible to install openssl libraries in locations which
2697 have names other than "lib", for example "/usr/lib64" which some
2699 [Steve Henson, based on patch from Jeremy Utley]
2701 *) Don't allow the use of leading 0x80 in OIDs. This is a violation of
2702 X690 8.9.12 and can produce some misleading textual output of OIDs.
2703 [Steve Henson, reported by Dan Kaminsky]
2705 *) Delete MD2 from algorithm tables. This follows the recommendation in
2706 several standards that it is not used in new applications due to
2707 several cryptographic weaknesses. For binary compatibility reasons
2708 the MD2 API is still compiled in by default.
2711 *) Add compression id to {d2i,i2d}_SSL_SESSION so it is correctly saved
2715 *) Rename uni2asc and asc2uni functions to OPENSSL_uni2asc and
2716 OPENSSL_asc2uni conditionally on Netware platforms to avoid a name
2718 [Guenter <lists@gknw.net>]
2720 *) Fix the server certificate chain building code to use X509_verify_cert(),
2721 it used to have an ad-hoc builder which was unable to cope with anything
2722 other than a simple chain.
2723 [David Woodhouse <dwmw2@infradead.org>, Steve Henson]
2725 *) Don't check self signed certificate signatures in X509_verify_cert()
2726 by default (a flag can override this): it just wastes time without
2727 adding any security. As a useful side effect self signed root CAs
2728 with non-FIPS digests are now usable in FIPS mode.
2731 *) In dtls1_process_out_of_seq_message() the check if the current message
2732 is already buffered was missing. For every new message was memory
2733 allocated, allowing an attacker to perform an denial of service attack
2734 with sending out of seq handshake messages until there is no memory
2735 left. Additionally every future messege was buffered, even if the
2736 sequence number made no sense and would be part of another handshake.
2737 So only messages with sequence numbers less than 10 in advance will be
2738 buffered. (CVE-2009-1378)
2739 [Robin Seggelmann, discovered by Daniel Mentz]
2741 *) Records are buffered if they arrive with a future epoch to be
2742 processed after finishing the corresponding handshake. There is
2743 currently no limitation to this buffer allowing an attacker to perform
2744 a DOS attack with sending records with future epochs until there is no
2745 memory left. This patch adds the pqueue_size() function to determine
2746 the size of a buffer and limits the record buffer to 100 entries.
2748 [Robin Seggelmann, discovered by Daniel Mentz]
2750 *) Keep a copy of frag->msg_header.frag_len so it can be used after the
2751 parent structure is freed. (CVE-2009-1379)
2754 *) Handle non-blocking I/O properly in SSL_shutdown() call.
2755 [Darryl Miles <darryl-mailinglists@netbauds.net>]
2758 [Ilya O. <vrghost@gmail.com>]
2760 Changes between 0.9.8k and 0.9.8l [5 Nov 2009]
2762 *) Disable renegotiation completely - this fixes a severe security
2763 problem (CVE-2009-3555) at the cost of breaking all
2764 renegotiation. Renegotiation can be re-enabled by setting
2765 SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION in s3->flags at
2766 run-time. This is really not recommended unless you know what
2770 Changes between 0.9.8j and 0.9.8k [25 Mar 2009]
2772 *) Don't set val to NULL when freeing up structures, it is freed up by
2773 underlying code. If sizeof(void *) > sizeof(long) this can result in
2774 zeroing past the valid field. (CVE-2009-0789)
2775 [Paolo Ganci <Paolo.Ganci@AdNovum.CH>]
2777 *) Fix bug where return value of CMS_SignerInfo_verify_content() was not
2778 checked correctly. This would allow some invalid signed attributes to
2779 appear to verify correctly. (CVE-2009-0591)
2780 [Ivan Nestlerode <inestlerode@us.ibm.com>]
2782 *) Reject UniversalString and BMPString types with invalid lengths. This
2783 prevents a crash in ASN1_STRING_print_ex() which assumes the strings have
2784 a legal length. (CVE-2009-0590)
2787 *) Set S/MIME signing as the default purpose rather than setting it
2788 unconditionally. This allows applications to override it at the store
2792 *) Permit restricted recursion of ASN1 strings. This is needed in practice
2793 to handle some structures.
2796 *) Improve efficiency of mem_gets: don't search whole buffer each time
2798 [Jeremy Shapiro <jnshapir@us.ibm.com>]
2800 *) New -hex option for openssl rand.
2803 *) Print out UTF8String and NumericString when parsing ASN1.
2806 *) Support NumericString type for name components.
2809 *) Allow CC in the environment to override the automatically chosen
2810 compiler. Note that nothing is done to ensure flags work with the
2814 Changes between 0.9.8i and 0.9.8j [07 Jan 2009]
2816 *) Properly check EVP_VerifyFinal() and similar return values
2818 [Ben Laurie, Bodo Moeller, Google Security Team]
2820 *) Enable TLS extensions by default.
2823 *) Allow the CHIL engine to be loaded, whether the application is
2824 multithreaded or not. (This does not release the developer from the
2825 obligation to set up the dynamic locking callbacks.)
2826 [Sander Temme <sander@temme.net>]
2828 *) Use correct exit code if there is an error in dgst command.
2829 [Steve Henson; problem pointed out by Roland Dirlewanger]
2831 *) Tweak Configure so that you need to say "experimental-jpake" to enable
2832 JPAKE, and need to use -DOPENSSL_EXPERIMENTAL_JPAKE in applications.
2835 *) Add experimental JPAKE support, including demo authentication in
2836 s_client and s_server.
2839 *) Set the comparison function in v3_addr_canonize().
2840 [Rob Austein <sra@hactrn.net>]
2842 *) Add support for XMPP STARTTLS in s_client.
2843 [Philip Paeps <philip@freebsd.org>]
2845 *) Change the server-side SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG behavior
2846 to ensure that even with this option, only ciphersuites in the
2847 server's preference list will be accepted. (Note that the option
2848 applies only when resuming a session, so the earlier behavior was
2849 just about the algorithm choice for symmetric cryptography.)
2852 Changes between 0.9.8h and 0.9.8i [15 Sep 2008]
2854 *) Fix NULL pointer dereference if a DTLS server received
2855 ChangeCipherSpec as first record (CVE-2009-1386).
2858 *) Fix a state transition in s3_srvr.c and d1_srvr.c
2859 (was using SSL3_ST_CW_CLNT_HELLO_B, should be ..._ST_SW_SRVR_...).
2862 *) The fix in 0.9.8c that supposedly got rid of unsafe
2863 double-checked locking was incomplete for RSA blinding,
2864 addressing just one layer of what turns out to have been
2865 doubly unsafe triple-checked locking.
2867 So now fix this for real by retiring the MONT_HELPER macro
2868 in crypto/rsa/rsa_eay.c.
2870 [Bodo Moeller; problem pointed out by Marius Schilder]
2872 *) Various precautionary measures:
2874 - Avoid size_t integer overflow in HASH_UPDATE (md32_common.h).
2876 - Avoid a buffer overflow in d2i_SSL_SESSION() (ssl_asn1.c).
2877 (NB: This would require knowledge of the secret session ticket key
2878 to exploit, in which case you'd be SOL either way.)
2880 - Change bn_nist.c so that it will properly handle input BIGNUMs
2881 outside the expected range.
2883 - Enforce the 'num' check in BN_div() (bn_div.c) for non-BN_DEBUG
2886 [Neel Mehta, Bodo Moeller]
2888 *) Allow engines to be "soft loaded" - i.e. optionally don't die if
2889 the load fails. Useful for distros.
2890 [Ben Laurie and the FreeBSD team]
2892 *) Add support for Local Machine Keyset attribute in PKCS#12 files.
2895 *) Fix BN_GF2m_mod_arr() top-bit cleanup code.
2898 *) Expand ENGINE to support engine supplied SSL client certificate functions.
2900 This work was sponsored by Logica.
2903 *) Add CryptoAPI ENGINE to support use of RSA and DSA keys held in Windows
2904 keystores. Support for SSL/TLS client authentication too.
2905 Not compiled unless enable-capieng specified to Configure.
2907 This work was sponsored by Logica.
2910 *) Fix bug in X509_ATTRIBUTE creation: dont set attribute using
2911 ASN1_TYPE_set1 if MBSTRING flag set. This bug would crash certain
2912 attribute creation routines such as certifcate requests and PKCS#12
2916 Changes between 0.9.8g and 0.9.8h [28 May 2008]
2918 *) Fix flaw if 'Server Key exchange message' is omitted from a TLS
2919 handshake which could lead to a cilent crash as found using the
2920 Codenomicon TLS test suite (CVE-2008-1672)
2921 [Steve Henson, Mark Cox]
2923 *) Fix double free in TLS server name extensions which could lead to
2924 a remote crash found by Codenomicon TLS test suite (CVE-2008-0891)
2927 *) Clear error queue in SSL_CTX_use_certificate_chain_file()
2929 Clear the error queue to ensure that error entries left from
2930 older function calls do not interfere with the correct operation.
2931 [Lutz Jaenicke, Erik de Castro Lopo]
2933 *) Remove root CA certificates of commercial CAs:
2935 The OpenSSL project does not recommend any specific CA and does not
2936 have any policy with respect to including or excluding any CA.
2937 Therefore it does not make any sense to ship an arbitrary selection
2938 of root CA certificates with the OpenSSL software.
2941 *) RSA OAEP patches to fix two separate invalid memory reads.
2942 The first one involves inputs when 'lzero' is greater than
2943 'SHA_DIGEST_LENGTH' (it would read about SHA_DIGEST_LENGTH bytes
2944 before the beginning of from). The second one involves inputs where
2945 the 'db' section contains nothing but zeroes (there is a one-byte
2946 invalid read after the end of 'db').
2947 [Ivan Nestlerode <inestlerode@us.ibm.com>]
2949 *) Partial backport from 0.9.9-dev:
2951 Introduce bn_mul_mont (dedicated Montgomery multiplication
2952 procedure) as a candidate for BIGNUM assembler implementation.
2953 While 0.9.9-dev uses assembler for various architectures, only
2954 x86_64 is available by default here in the 0.9.8 branch, and
2955 32-bit x86 is available through a compile-time setting.
2957 To try the 32-bit x86 assembler implementation, use Configure
2958 option "enable-montasm" (which exists only for this backport).
2960 As "enable-montasm" for 32-bit x86 disclaims code stability
2961 anyway, in this constellation we activate additional code
2962 backported from 0.9.9-dev for further performance improvements,
2963 namely BN_from_montgomery_word. (To enable this otherwise,
2964 e.g. x86_64, try "-DMONT_FROM_WORD___NON_DEFAULT_0_9_8_BUILD".)
2966 [Andy Polyakov (backport partially by Bodo Moeller)]
2968 *) Add TLS session ticket callback. This allows an application to set
2969 TLS ticket cipher and HMAC keys rather than relying on hardcoded fixed
2970 values. This is useful for key rollover for example where several key
2971 sets may exist with different names.
2974 *) Reverse ENGINE-internal logic for caching default ENGINE handles.
2975 This was broken until now in 0.9.8 releases, such that the only way
2976 a registered ENGINE could be used (assuming it initialises
2977 successfully on the host) was to explicitly set it as the default
2978 for the relevant algorithms. This is in contradiction with 0.9.7
2979 behaviour and the documentation. With this fix, when an ENGINE is
2980 registered into a given algorithm's table of implementations, the
2981 'uptodate' flag is reset so that auto-discovery will be used next
2982 time a new context for that algorithm attempts to select an
2984 [Ian Lister (tweaked by Geoff Thorpe)]
2986 *) Backport of CMS code to OpenSSL 0.9.8. This differs from the 0.9.9
2987 implemention in the following ways:
2989 Lack of EVP_PKEY_ASN1_METHOD means algorithm parameters have to be
2992 Lack of BER streaming support means one pass streaming processing is
2993 only supported if data is detached: setting the streaming flag is
2994 ignored for embedded content.
2996 CMS support is disabled by default and must be explicitly enabled
2997 with the enable-cms configuration option.
3000 *) Update the GMP engine glue to do direct copies between BIGNUM and
3001 mpz_t when openssl and GMP use the same limb size. Otherwise the
3002 existing "conversion via a text string export" trick is still used.
3003 [Paul Sheer <paulsheer@gmail.com>]
3005 *) Zlib compression BIO. This is a filter BIO which compressed and
3006 uncompresses any data passed through it.
3009 *) Add AES_wrap_key() and AES_unwrap_key() functions to implement
3010 RFC3394 compatible AES key wrapping.
3013 *) Add utility functions to handle ASN1 structures. ASN1_STRING_set0():
3014 sets string data without copying. X509_ALGOR_set0() and
3015 X509_ALGOR_get0(): set and retrieve X509_ALGOR (AlgorithmIdentifier)
3016 data. Attribute function X509at_get0_data_by_OBJ(): retrieves data
3017 from an X509_ATTRIBUTE structure optionally checking it occurs only
3018 once. ASN1_TYPE_set1(): set and ASN1_TYPE structure copying supplied
3022 *) Fix BN flag handling in RSA_eay_mod_exp() and BN_MONT_CTX_set()
3023 to get the expected BN_FLG_CONSTTIME behavior.
3024 [Bodo Moeller (Google)]
3028 - fixed wrong usage of ioctlsocket() when build for LIBC BSD sockets
3029 - fixed do_tests.pl to run the test suite with CLIB builds too (CLIB_OPT)
3030 - added some more tests to do_tests.pl
3031 - fixed RunningProcess usage so that it works with newer LIBC NDKs too
3032 - removed usage of BN_LLONG for CLIB builds to avoid runtime dependency
3033 - added new Configure targets netware-clib-bsdsock, netware-clib-gcc,
3034 netware-clib-bsdsock-gcc, netware-libc-bsdsock-gcc
3035 - various changes to netware.pl to enable gcc-cross builds on Win32
3037 - changed crypto/bio/b_sock.c to work with macro functions (CLIB BSD)
3038 - various changes to fix missing prototype warnings
3039 - fixed x86nasm.pl to create correct asm files for NASM COFF output
3040 - added AES, WHIRLPOOL and CPUID assembler code to build files
3041 - added missing AES assembler make rules to mk1mf.pl
3042 - fixed order of includes in apps/ocsp.c so that e_os.h settings apply
3043 [Guenter Knauf <eflash@gmx.net>]
3045 *) Implement certificate status request TLS extension defined in RFC3546.
3046 A client can set the appropriate parameters and receive the encoded
3047 OCSP response via a callback. A server can query the supplied parameters
3048 and set the encoded OCSP response in the callback. Add simplified examples
3049 to s_client and s_server.
3052 Changes between 0.9.8f and 0.9.8g [19 Oct 2007]
3054 *) Fix various bugs:
3055 + Binary incompatibility of ssl_ctx_st structure
3056 + DTLS interoperation with non-compliant servers
3057 + Don't call get_session_cb() without proposed session
3058 + Fix ia64 assembler code
3059 [Andy Polyakov, Steve Henson]
3061 Changes between 0.9.8e and 0.9.8f [11 Oct 2007]
3063 *) DTLS Handshake overhaul. There were longstanding issues with
3064 OpenSSL DTLS implementation, which were making it impossible for
3065 RFC 4347 compliant client to communicate with OpenSSL server.
3066 Unfortunately just fixing these incompatibilities would "cut off"
3067 pre-0.9.8f clients. To allow for hassle free upgrade post-0.9.8e
3068 server keeps tolerating non RFC compliant syntax. The opposite is
3069 not true, 0.9.8f client can not communicate with earlier server.
3070 This update even addresses CVE-2007-4995.
3073 *) Changes to avoid need for function casts in OpenSSL: some compilers
3074 (gcc 4.2 and later) reject their use.
3075 [Kurt Roeckx <kurt@roeckx.be>, Peter Hartley <pdh@utter.chaos.org.uk>,
3078 *) Add RFC4507 support to OpenSSL. This includes the corrections in
3079 RFC4507bis. The encrypted ticket format is an encrypted encoded
3080 SSL_SESSION structure, that way new session features are automatically
3083 If a client application caches session in an SSL_SESSION structure
3084 support is transparent because tickets are now stored in the encoded
3087 The SSL_CTX structure automatically generates keys for ticket
3088 protection in servers so again support should be possible
3089 with no application modification.
3091 If a client or server wishes to disable RFC4507 support then the option
3092 SSL_OP_NO_TICKET can be set.
3094 Add a TLS extension debugging callback to allow the contents of any client
3095 or server extensions to be examined.
3097 This work was sponsored by Google.
3100 *) Add initial support for TLS extensions, specifically for the server_name
3101 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
3102 have new members for a host name. The SSL data structure has an
3103 additional member SSL_CTX *initial_ctx so that new sessions can be
3104 stored in that context to allow for session resumption, even after the
3105 SSL has been switched to a new SSL_CTX in reaction to a client's
3106 server_name extension.
3108 New functions (subject to change):
3110 SSL_get_servername()
3111 SSL_get_servername_type()
3114 New CTRL codes and macros (subject to change):
3116 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
3117 - SSL_CTX_set_tlsext_servername_callback()
3118 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
3119 - SSL_CTX_set_tlsext_servername_arg()
3120 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
3122 openssl s_client has a new '-servername ...' option.
3124 openssl s_server has new options '-servername_host ...', '-cert2 ...',
3125 '-key2 ...', '-servername_fatal' (subject to change). This allows
3126 testing the HostName extension for a specific single host name ('-cert'
3127 and '-key' remain fallbacks for handshakes without HostName
3128 negotiation). If the unrecognized_name alert has to be sent, this by
3129 default is a warning; it becomes fatal with the '-servername_fatal'
3132 [Peter Sylvester, Remy Allais, Christophe Renou, Steve Henson]
3134 *) Add AES and SSE2 assembly language support to VC++ build.
3137 *) Mitigate attack on final subtraction in Montgomery reduction.
3140 *) Fix crypto/ec/ec_mult.c to work properly with scalars of value 0
3141 (which previously caused an internal error).
3144 *) Squeeze another 10% out of IGE mode when in != out.
3147 *) AES IGE mode speedup.
3148 [Dean Gaudet (Google)]
3150 *) Add the Korean symmetric 128-bit cipher SEED (see
3151 http://www.kisa.or.kr/kisa/seed/jsp/seed_eng.jsp) and
3152 add SEED ciphersuites from RFC 4162:
3154 TLS_RSA_WITH_SEED_CBC_SHA = "SEED-SHA"
3155 TLS_DHE_DSS_WITH_SEED_CBC_SHA = "DHE-DSS-SEED-SHA"
3156 TLS_DHE_RSA_WITH_SEED_CBC_SHA = "DHE-RSA-SEED-SHA"
3157 TLS_DH_anon_WITH_SEED_CBC_SHA = "ADH-SEED-SHA"
3159 To minimize changes between patchlevels in the OpenSSL 0.9.8
3160 series, SEED remains excluded from compilation unless OpenSSL
3161 is configured with 'enable-seed'.
3162 [KISA, Bodo Moeller]
3164 *) Mitigate branch prediction attacks, which can be practical if a
3165 single processor is shared, allowing a spy process to extract
3166 information. For detailed background information, see
3167 http://eprint.iacr.org/2007/039 (O. Aciicmez, S. Gueron,
3168 J.-P. Seifert, "New Branch Prediction Vulnerabilities in OpenSSL
3169 and Necessary Software Countermeasures"). The core of the change
3170 are new versions BN_div_no_branch() and
3171 BN_mod_inverse_no_branch() of BN_div() and BN_mod_inverse(),
3172 respectively, which are slower, but avoid the security-relevant
3173 conditional branches. These are automatically called by BN_div()
3174 and BN_mod_inverse() if the flag BN_FLG_CONSTTIME is set for one
3175 of the input BIGNUMs. Also, BN_is_bit_set() has been changed to
3176 remove a conditional branch.
3178 BN_FLG_CONSTTIME is the new name for the previous
3179 BN_FLG_EXP_CONSTTIME flag, since it now affects more than just
3180 modular exponentiation. (Since OpenSSL 0.9.7h, setting this flag
3181 in the exponent causes BN_mod_exp_mont() to use the alternative
3182 implementation in BN_mod_exp_mont_consttime().) The old name
3183 remains as a deprecated alias.
3185 Similary, RSA_FLAG_NO_EXP_CONSTTIME is replaced by a more general
3186 RSA_FLAG_NO_CONSTTIME flag since the RSA implementation now uses
3187 constant-time implementations for more than just exponentiation.
3188 Here too the old name is kept as a deprecated alias.
3190 BN_BLINDING_new() will now use BN_dup() for the modulus so that
3191 the BN_BLINDING structure gets an independent copy of the
3192 modulus. This means that the previous "BIGNUM *m" argument to
3193 BN_BLINDING_new() and to BN_BLINDING_create_param() now
3194 essentially becomes "const BIGNUM *m", although we can't actually
3195 change this in the header file before 0.9.9. It allows
3196 RSA_setup_blinding() to use BN_with_flags() on the modulus to
3197 enable BN_FLG_CONSTTIME.
3199 [Matthew D Wood (Intel Corp)]
3201 *) In the SSL/TLS server implementation, be strict about session ID
3202 context matching (which matters if an application uses a single
3203 external cache for different purposes). Previously,
3204 out-of-context reuse was forbidden only if SSL_VERIFY_PEER was
3205 set. This did ensure strict client verification, but meant that,
3206 with applications using a single external cache for quite
3207 different requirements, clients could circumvent ciphersuite
3208 restrictions for a given session ID context by starting a session
3209 in a different context.
3212 *) Include "!eNULL" in SSL_DEFAULT_CIPHER_LIST to make sure that
3213 a ciphersuite string such as "DEFAULT:RSA" cannot enable
3214 authentication-only ciphersuites.
3217 *) Update the SSL_get_shared_ciphers() fix CVE-2006-3738 which was
3218 not complete and could lead to a possible single byte overflow
3219 (CVE-2007-5135) [Ben Laurie]
3221 Changes between 0.9.8d and 0.9.8e [23 Feb 2007]
3223 *) Since AES128 and AES256 (and similarly Camellia128 and
3224 Camellia256) share a single mask bit in the logic of
3225 ssl/ssl_ciph.c, the code for masking out disabled ciphers needs a
3226 kludge to work properly if AES128 is available and AES256 isn't
3227 (or if Camellia128 is available and Camellia256 isn't).
3230 *) Fix the BIT STRING encoding generated by crypto/ec/ec_asn1.c
3231 (within i2d_ECPrivateKey, i2d_ECPKParameters, i2d_ECParameters):
3232 When a point or a seed is encoded in a BIT STRING, we need to
3233 prevent the removal of trailing zero bits to get the proper DER
3234 encoding. (By default, crypto/asn1/a_bitstr.c assumes the case
3235 of a NamedBitList, for which trailing 0 bits need to be removed.)
3238 *) Have SSL/TLS server implementation tolerate "mismatched" record
3239 protocol version while receiving ClientHello even if the
3240 ClientHello is fragmented. (The server can't insist on the
3241 particular protocol version it has chosen before the ServerHello
3242 message has informed the client about his choice.)
3245 *) Add RFC 3779 support.
3246 [Rob Austein for ARIN, Ben Laurie]
3248 *) Load error codes if they are not already present instead of using a
3249 static variable. This allows them to be cleanly unloaded and reloaded.
3250 Improve header file function name parsing.
3253 *) extend SMTP and IMAP protocol emulation in s_client to use EHLO
3254 or CAPABILITY handshake as required by RFCs.
3257 Changes between 0.9.8c and 0.9.8d [28 Sep 2006]
3259 *) Introduce limits to prevent malicious keys being able to
3260 cause a denial of service. (CVE-2006-2940)
3261 [Steve Henson, Bodo Moeller]
3263 *) Fix ASN.1 parsing of certain invalid structures that can result
3264 in a denial of service. (CVE-2006-2937) [Steve Henson]
3266 *) Fix buffer overflow in SSL_get_shared_ciphers() function.
3267 (CVE-2006-3738) [Tavis Ormandy and Will Drewry, Google Security Team]
3269 *) Fix SSL client code which could crash if connecting to a
3270 malicious SSLv2 server. (CVE-2006-4343)
3271 [Tavis Ormandy and Will Drewry, Google Security Team]
3273 *) Since 0.9.8b, ciphersuite strings naming explicit ciphersuites
3274 match only those. Before that, "AES256-SHA" would be interpreted
3275 as a pattern and match "AES128-SHA" too (since AES128-SHA got
3276 the same strength classification in 0.9.7h) as we currently only
3277 have a single AES bit in the ciphersuite description bitmap.
3278 That change, however, also applied to ciphersuite strings such as
3279 "RC4-MD5" that intentionally matched multiple ciphersuites --
3280 namely, SSL 2.0 ciphersuites in addition to the more common ones
3281 from SSL 3.0/TLS 1.0.
3283 So we change the selection algorithm again: Naming an explicit
3284 ciphersuite selects this one ciphersuite, and any other similar
3285 ciphersuite (same bitmap) from *other* protocol versions.
3286 Thus, "RC4-MD5" again will properly select both the SSL 2.0
3287 ciphersuite and the SSL 3.0/TLS 1.0 ciphersuite.
3289 Since SSL 2.0 does not have any ciphersuites for which the
3290 128/256 bit distinction would be relevant, this works for now.
3291 The proper fix will be to use different bits for AES128 and
3292 AES256, which would have avoided the problems from the beginning;
3293 however, bits are scarce, so we can only do this in a new release
3294 (not just a patchlevel) when we can change the SSL_CIPHER
3295 definition to split the single 'unsigned long mask' bitmap into
3296 multiple values to extend the available space.
3300 Changes between 0.9.8b and 0.9.8c [05 Sep 2006]
3302 *) Avoid PKCS #1 v1.5 signature attack discovered by Daniel Bleichenbacher
3303 (CVE-2006-4339) [Ben Laurie and Google Security Team]
3305 *) Add AES IGE and biIGE modes.
3308 *) Change the Unix randomness entropy gathering to use poll() when
3309 possible instead of select(), since the latter has some
3310 undesirable limitations.
3311 [Darryl Miles via Richard Levitte and Bodo Moeller]
3313 *) Disable "ECCdraft" ciphersuites more thoroughly. Now special
3314 treatment in ssl/ssl_ciph.s makes sure that these ciphersuites
3315 cannot be implicitly activated as part of, e.g., the "AES" alias.
3316 However, please upgrade to OpenSSL 0.9.9[-dev] for
3317 non-experimental use of the ECC ciphersuites to get TLS extension
3318 support, which is required for curve and point format negotiation
3319 to avoid potential handshake problems.
3322 *) Disable rogue ciphersuites:
3324 - SSLv2 0x08 0x00 0x80 ("RC4-64-MD5")
3325 - SSLv3/TLSv1 0x00 0x61 ("EXP1024-RC2-CBC-MD5")
3326 - SSLv3/TLSv1 0x00 0x60 ("EXP1024-RC4-MD5")
3328 The latter two were purportedly from
3329 draft-ietf-tls-56-bit-ciphersuites-0[01].txt, but do not really
3332 Also deactivate the remaining ciphersuites from
3333 draft-ietf-tls-56-bit-ciphersuites-01.txt. These are just as
3334 unofficial, and the ID has long expired.
3337 *) Fix RSA blinding Heisenbug (problems sometimes occured on
3338 dual-core machines) and other potential thread-safety issues.
3341 *) Add the symmetric cipher Camellia (128-bit, 192-bit, 256-bit key
3342 versions), which is now available for royalty-free use
3343 (see http://info.isl.ntt.co.jp/crypt/eng/info/chiteki.html).
3344 Also, add Camellia TLS ciphersuites from RFC 4132.
3346 To minimize changes between patchlevels in the OpenSSL 0.9.8
3347 series, Camellia remains excluded from compilation unless OpenSSL
3348 is configured with 'enable-camellia'.
3351 *) Disable the padding bug check when compression is in use. The padding
3352 bug check assumes the first packet is of even length, this is not
3353 necessarily true if compresssion is enabled and can result in false
3354 positives causing handshake failure. The actual bug test is ancient
3355 code so it is hoped that implementations will either have fixed it by
3356 now or any which still have the bug do not support compression.
3359 Changes between 0.9.8a and 0.9.8b [04 May 2006]
3361 *) When applying a cipher rule check to see if string match is an explicit
3362 cipher suite and only match that one cipher suite if it is.
3365 *) Link in manifests for VC++ if needed.
3366 [Austin Ziegler <halostatue@gmail.com>]
3368 *) Update support for ECC-based TLS ciphersuites according to
3369 draft-ietf-tls-ecc-12.txt with proposed changes (but without
3370 TLS extensions, which are supported starting with the 0.9.9
3371 branch, not in the OpenSSL 0.9.8 branch).
3374 *) New functions EVP_CIPHER_CTX_new() and EVP_CIPHER_CTX_free() to support
3375 opaque EVP_CIPHER_CTX handling.
3378 *) Fixes and enhancements to zlib compression code. We now only use
3379 "zlib1.dll" and use the default __cdecl calling convention on Win32
3380 to conform with the standards mentioned here:
3381 http://www.zlib.net/DLL_FAQ.txt
3382 Static zlib linking now works on Windows and the new --with-zlib-include
3383 --with-zlib-lib options to Configure can be used to supply the location
3384 of the headers and library. Gracefully handle case where zlib library
3388 *) Several fixes and enhancements to the OID generation code. The old code
3389 sometimes allowed invalid OIDs (1.X for X >= 40 for example), couldn't
3390 handle numbers larger than ULONG_MAX, truncated printing and had a
3391 non standard OBJ_obj2txt() behaviour.
3394 *) Add support for building of engines under engine/ as shared libraries
3395 under VC++ build system.
3398 *) Corrected the numerous bugs in the Win32 path splitter in DSO.
3399 Hopefully, we will not see any false combination of paths any more.
3402 Changes between 0.9.8 and 0.9.8a [11 Oct 2005]
3404 *) Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING
3405 (part of SSL_OP_ALL). This option used to disable the
3406 countermeasure against man-in-the-middle protocol-version
3407 rollback in the SSL 2.0 server implementation, which is a bad
3408 idea. (CVE-2005-2969)
3410 [Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center
3411 for Information Security, National Institute of Advanced Industrial
3412 Science and Technology [AIST], Japan)]
3414 *) Add two function to clear and return the verify parameter flags.
3417 *) Keep cipherlists sorted in the source instead of sorting them at
3418 runtime, thus removing the need for a lock.
3421 *) Avoid some small subgroup attacks in Diffie-Hellman.
3422 [Nick Mathewson and Ben Laurie]
3424 *) Add functions for well-known primes.
3427 *) Extended Windows CE support.
3428 [Satoshi Nakamura and Andy Polyakov]
3430 *) Initialize SSL_METHOD structures at compile time instead of during
3431 runtime, thus removing the need for a lock.
3434 *) Make PKCS7_decrypt() work even if no certificate is supplied by
3435 attempting to decrypt each encrypted key in turn. Add support to
3439 Changes between 0.9.7h and 0.9.8 [05 Jul 2005]
3441 [NB: OpenSSL 0.9.7i and later 0.9.7 patch levels were released after
3444 *) Add libcrypto.pc and libssl.pc for those who feel they need them.
3447 *) Change CA.sh and CA.pl so they don't bundle the CSR and the private
3448 key into the same file any more.
3451 *) Add initial support for Win64, both IA64 and AMD64/x64 flavors.
3454 *) Add -utf8 command line and config file option to 'ca'.
3455 [Stefan <stf@udoma.org]
3457 *) Removed the macro des_crypt(), as it seems to conflict with some
3458 libraries. Use DES_crypt().
3461 *) Correct naming of the 'chil' and '4758cca' ENGINEs. This
3462 involves renaming the source and generated shared-libs for
3463 both. The engines will accept the corrected or legacy ids
3464 ('ncipher' and '4758_cca' respectively) when binding. NB,
3465 this only applies when building 'shared'.
3466 [Corinna Vinschen <vinschen@redhat.com> and Geoff Thorpe]
3468 *) Add attribute functions to EVP_PKEY structure. Modify
3469 PKCS12_create() to recognize a CSP name attribute and
3470 use it. Make -CSP option work again in pkcs12 utility.
3473 *) Add new functionality to the bn blinding code:
3474 - automatic re-creation of the BN_BLINDING parameters after
3475 a fixed number of uses (currently 32)
3476 - add new function for parameter creation
3477 - introduce flags to control the update behaviour of the
3478 BN_BLINDING parameters
3479 - hide BN_BLINDING structure
3480 Add a second BN_BLINDING slot to the RSA structure to improve
3481 performance when a single RSA object is shared among several
3485 *) Add support for DTLS.
3486 [Nagendra Modadugu <nagendra@cs.stanford.edu> and Ben Laurie]
3488 *) Add support for DER encoded private keys (SSL_FILETYPE_ASN1)
3489 to SSL_CTX_use_PrivateKey_file() and SSL_use_PrivateKey_file()
3492 *) Remove buggy and incomplete DH cert support from
3493 ssl/ssl_rsa.c and ssl/s3_both.c
3496 *) Use SHA-1 instead of MD5 as the default digest algorithm for
3497 the apps/openssl applications.
3500 *) Compile clean with "-Wall -Wmissing-prototypes
3501 -Wstrict-prototypes -Wmissing-declarations -Werror". Currently
3502 DEBUG_SAFESTACK must also be set.
3505 *) Change ./Configure so that certain algorithms can be disabled by default.
3506 The new counterpiece to "no-xxx" is "enable-xxx".
3508 The patented RC5 and MDC2 algorithms will now be disabled unless
3509 "enable-rc5" and "enable-mdc2", respectively, are specified.
3511 (IDEA remains enabled despite being patented. This is because IDEA
3512 is frequently required for interoperability, and there is no license
3513 fee for non-commercial use. As before, "no-idea" can be used to
3514 avoid this algorithm.)
3518 *) Add processing of proxy certificates (see RFC 3820). This work was
3519 sponsored by KTH (The Royal Institute of Technology in Stockholm) and
3520 EGEE (Enabling Grids for E-science in Europe).
3523 *) RC4 performance overhaul on modern architectures/implementations, such
3524 as Intel P4, IA-64 and AMD64.
3527 *) New utility extract-section.pl. This can be used specify an alternative
3528 section number in a pod file instead of having to treat each file as
3529 a separate case in Makefile. This can be done by adding two lines to the
3532 =for comment openssl_section:XXX
3534 The blank line is mandatory.
3538 *) New arguments -certform, -keyform and -pass for s_client and s_server
3539 to allow alternative format key and certificate files and passphrase
3543 *) New structure X509_VERIFY_PARAM which combines current verify parameters,
3544 update associated structures and add various utility functions.
3546 Add new policy related verify parameters, include policy checking in
3547 standard verify code. Enhance 'smime' application with extra parameters
3548 to support policy checking and print out.
3551 *) Add a new engine to support VIA PadLock ACE extensions in the VIA C3
3552 Nehemiah processors. These extensions support AES encryption in hardware
3553 as well as RNG (though RNG support is currently disabled).
3554 [Michal Ludvig <michal@logix.cz>, with help from Andy Polyakov]
3556 *) Deprecate BN_[get|set]_params() functions (they were ignored internally).
3559 *) New FIPS 180-2 algorithms, SHA-224/-256/-384/-512 are implemented.
3560 [Andy Polyakov and a number of other people]
3562 *) Improved PowerPC platform support. Most notably BIGNUM assembler
3563 implementation contributed by IBM.
3564 [Suresh Chari, Peter Waltenberg, Andy Polyakov]
3566 *) The new 'RSA_generate_key_ex' function now takes a BIGNUM for the public
3567 exponent rather than 'unsigned long'. There is a corresponding change to
3568 the new 'rsa_keygen' element of the RSA_METHOD structure.
3569 [Jelte Jansen, Geoff Thorpe]
3571 *) Functionality for creating the initial serial number file is now
3572 moved from CA.pl to the 'ca' utility with a new option -create_serial.
3574 (Before OpenSSL 0.9.7e, CA.pl used to initialize the serial
3575 number file to 1, which is bound to cause problems. To avoid
3576 the problems while respecting compatibility between different 0.9.7
3577 patchlevels, 0.9.7e employed 'openssl x509 -next_serial' in
3578 CA.pl for serial number initialization. With the new release 0.9.8,
3579 we can fix the problem directly in the 'ca' utility.)
3582 *) Reduced header interdepencies by declaring more opaque objects in
3583 ossl_typ.h. As a consequence, including some headers (eg. engine.h) will
3584 give fewer recursive includes, which could break lazy source code - so
3585 this change is covered by the OPENSSL_NO_DEPRECATED symbol. As always,
3586 developers should define this symbol when building and using openssl to
3587 ensure they track the recommended behaviour, interfaces, [etc], but
3588 backwards-compatible behaviour prevails when this isn't defined.
3591 *) New function X509_POLICY_NODE_print() which prints out policy nodes.
3594 *) Add new EVP function EVP_CIPHER_CTX_rand_key and associated functionality.
3595 This will generate a random key of the appropriate length based on the
3596 cipher context. The EVP_CIPHER can provide its own random key generation
3597 routine to support keys of a specific form. This is used in the des and
3598 3des routines to generate a key of the correct parity. Update S/MIME
3599 code to use new functions and hence generate correct parity DES keys.
3600 Add EVP_CHECK_DES_KEY #define to return an error if the key is not
3601 valid (weak or incorrect parity).
3604 *) Add a local set of CRLs that can be used by X509_verify_cert() as well
3605 as looking them up. This is useful when the verified structure may contain
3606 CRLs, for example PKCS#7 signedData. Modify PKCS7_verify() to use any CRLs
3607 present unless the new PKCS7_NO_CRL flag is asserted.
3610 *) Extend ASN1 oid configuration module. It now additionally accepts the
3613 shortName = some long name, 1.2.3.4
3616 *) Reimplemented the BN_CTX implementation. There is now no more static
3617 limitation on the number of variables it can handle nor the depth of the
3618 "stack" handling for BN_CTX_start()/BN_CTX_end() pairs. The stack
3619 information can now expand as required, and rather than having a single
3620 static array of bignums, BN_CTX now uses a linked-list of such arrays
3621 allowing it to expand on demand whilst maintaining the usefulness of
3622 BN_CTX's "bundling".
3625 *) Add a missing BN_CTX parameter to the 'rsa_mod_exp' callback in RSA_METHOD
3626 to allow all RSA operations to function using a single BN_CTX.
3629 *) Preliminary support for certificate policy evaluation and checking. This
3630 is initially intended to pass the tests outlined in "Conformance Testing
3631 of Relying Party Client Certificate Path Processing Logic" v1.07.
3634 *) bn_dup_expand() has been deprecated, it was introduced in 0.9.7 and
3635 remained unused and not that useful. A variety of other little bignum
3636 tweaks and fixes have also been made continuing on from the audit (see
3640 *) Constify all or almost all d2i, c2i, s2i and r2i functions, along with
3641 associated ASN1, EVP and SSL functions and old ASN1 macros.
3644 *) BN_zero() only needs to set 'top' and 'neg' to zero for correct results,
3645 and this should never fail. So the return value from the use of
3646 BN_set_word() (which can fail due to needless expansion) is now deprecated;
3647 if OPENSSL_NO_DEPRECATED is defined, BN_zero() is a void macro.
3650 *) BN_CTX_get() should return zero-valued bignums, providing the same
3651 initialised value as BN_new().
3652 [Geoff Thorpe, suggested by Ulf Möller]
3654 *) Support for inhibitAnyPolicy certificate extension.
3657 *) An audit of the BIGNUM code is underway, for which debugging code is
3658 enabled when BN_DEBUG is defined. This makes stricter enforcements on what
3659 is considered valid when processing BIGNUMs, and causes execution to
3660 assert() when a problem is discovered. If BN_DEBUG_RAND is defined,
3661 further steps are taken to deliberately pollute unused data in BIGNUM
3662 structures to try and expose faulty code further on. For now, openssl will
3663 (in its default mode of operation) continue to tolerate the inconsistent
3664 forms that it has tolerated in the past, but authors and packagers should
3665 consider trying openssl and their own applications when compiled with
3666 these debugging symbols defined. It will help highlight potential bugs in
3667 their own code, and will improve the test coverage for OpenSSL itself. At
3668 some point, these tighter rules will become openssl's default to improve
3669 maintainability, though the assert()s and other overheads will remain only
3670 in debugging configurations. See bn.h for more details.
3671 [Geoff Thorpe, Nils Larsch, Ulf Möller]
3673 *) BN_CTX_init() has been deprecated, as BN_CTX is an opaque structure
3674 that can only be obtained through BN_CTX_new() (which implicitly
3675 initialises it). The presence of this function only made it possible
3676 to overwrite an existing structure (and cause memory leaks).
3679 *) Because of the callback-based approach for implementing LHASH as a
3680 template type, lh_insert() adds opaque objects to hash-tables and
3681 lh_doall() or lh_doall_arg() are typically used with a destructor callback
3682 to clean up those corresponding objects before destroying the hash table
3683 (and losing the object pointers). So some over-zealous constifications in
3684 LHASH have been relaxed so that lh_insert() does not take (nor store) the
3685 objects as "const" and the lh_doall[_arg] callback wrappers are not
3686 prototyped to have "const" restrictions on the object pointers they are
3687 given (and so aren't required to cast them away any more).
3690 *) The tmdiff.h API was so ugly and minimal that our own timing utility
3691 (speed) prefers to use its own implementation. The two implementations
3692 haven't been consolidated as yet (volunteers?) but the tmdiff API has had
3693 its object type properly exposed (MS_TM) instead of casting to/from "char
3694 *". This may still change yet if someone realises MS_TM and "ms_time_***"
3695 aren't necessarily the greatest nomenclatures - but this is what was used
3696 internally to the implementation so I've used that for now.
3699 *) Ensure that deprecated functions do not get compiled when
3700 OPENSSL_NO_DEPRECATED is defined. Some "openssl" subcommands and a few of
3701 the self-tests were still using deprecated key-generation functions so
3702 these have been updated also.
3705 *) Reorganise PKCS#7 code to separate the digest location functionality
3706 into PKCS7_find_digest(), digest addition into PKCS7_bio_add_digest().
3707 New function PKCS7_set_digest() to set the digest type for PKCS#7
3708 digestedData type. Add additional code to correctly generate the
3709 digestedData type and add support for this type in PKCS7 initialization
3713 *) New function PKCS7_set0_type_other() this initializes a PKCS7
3714 structure of type "other".
3717 *) Fix prime generation loop in crypto/bn/bn_prime.pl by making
3718 sure the loop does correctly stop and breaking ("division by zero")
3719 modulus operations are not performed. The (pre-generated) prime
3720 table crypto/bn/bn_prime.h was already correct, but it could not be
3721 re-generated on some platforms because of the "division by zero"
3722 situation in the script.
3723 [Ralf S. Engelschall]
3725 *) Update support for ECC-based TLS ciphersuites according to
3726 draft-ietf-tls-ecc-03.txt: the KDF1 key derivation function with
3727 SHA-1 now is only used for "small" curves (where the
3728 representation of a field element takes up to 24 bytes); for
3729 larger curves, the field element resulting from ECDH is directly
3730 used as premaster secret.
3731 [Douglas Stebila (Sun Microsystems Laboratories)]
3733 *) Add code for kP+lQ timings to crypto/ec/ectest.c, and add SEC2
3734 curve secp160r1 to the tests.
3735 [Douglas Stebila (Sun Microsystems Laboratories)]
3737 *) Add the possibility to load symbols globally with DSO.
3738 [Götz Babin-Ebell <babin-ebell@trustcenter.de> via Richard Levitte]
3740 *) Add the functions ERR_set_mark() and ERR_pop_to_mark() for better
3741 control of the error stack.
projects / openssl.git / blob