5 Changes between 1.0.2 and 1.1.0 [xx XXX xxxx]
7 *) Experimental support for a new, fast, unbiased prime candidate generator,
8 bn_probable_prime_dh_coprime(). Not currently used by any prime generator.
9 [Felix Laurie von Massenbach <felix@erbridge.co.uk>]
11 *) New output format NSS in the sess_id command line tool. This allows
12 exporting the session id and the master key in NSS keylog format.
13 [Martin Kaiser <martin@kaiser.cx>]
15 *) Harmonize version and its documentation. -f flag is used to display
17 [mancha <mancha1@zoho.com>]
19 *) Fix eckey_priv_encode so it immediately returns an error upon a failure
21 [mancha <mancha1@zoho.com>]
23 *) Fix some double frees. These are not thought to be exploitable.
24 [mancha <mancha1@zoho.com>]
26 *) A missing bounds check in the handling of the TLS heartbeat extension
27 can be used to reveal up to 64k of memory to a connected client or
30 Thanks for Neel Mehta of Google Security for discovering this bug and to
31 Adam Langley <agl@chromium.org> and Bodo Moeller <bmoeller@acm.org> for
32 preparing the fix (CVE-2014-0160)
33 [Adam Langley, Bodo Moeller]
35 *) Fix for the attack described in the paper "Recovering OpenSSL
36 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
37 by Yuval Yarom and Naomi Benger. Details can be obtained from:
38 http://eprint.iacr.org/2014/140
40 Thanks to Yuval Yarom and Naomi Benger for discovering this
41 flaw and to Yuval Yarom for supplying a fix (CVE-2014-0076)
42 [Yuval Yarom and Naomi Benger]
44 *) Use algorithm specific chains in SSL_CTX_use_certificate_chain_file():
45 this fixes a limitation in previous versions of OpenSSL.
48 *) Experimental encrypt-then-mac support.
50 Experimental support for encrypt then mac from
51 draft-gutmann-tls-encrypt-then-mac-02.txt
53 To enable it set the appropriate extension number (0x42 for the test
54 server) using e.g. -DTLSEXT_TYPE_encrypt_then_mac=0x42
56 For non-compliant peers (i.e. just about everything) this should have no
59 WARNING: EXPERIMENTAL, SUBJECT TO CHANGE.
63 *) Add EVP support for key wrapping algorithms, to avoid problems with
64 existing code the flag EVP_CIPHER_CTX_WRAP_ALLOW has to be set in
65 the EVP_CIPHER_CTX or an error is returned. Add AES and DES3 wrap
66 algorithms and include tests cases.
69 *) Extend CMS code to support RSA-PSS signatures and RSA-OAEP for
73 *) Extended RSA OAEP support via EVP_PKEY API. Options to specify digest,
74 MGF1 digest and OAEP label.
77 *) Make openssl verify return errors.
78 [Chris Palmer <palmer@google.com> and Ben Laurie]
80 *) New function ASN1_TIME_diff to calculate the difference between two
81 ASN1_TIME structures or one structure and the current time.
84 *) Update fips_test_suite to support multiple command line options. New
85 test to induce all self test errors in sequence and check expected
89 *) Add FIPS_{rsa,dsa,ecdsa}_{sign,verify} functions which digest and
90 sign or verify all in one operation.
93 *) Add fips_algvs: a multicall fips utility incorporating all the algorithm
94 test programs and fips_test_suite. Includes functionality to parse
95 the minimal script output of fipsalgest.pl directly.
98 *) Add authorisation parameter to FIPS_module_mode_set().
101 *) Add FIPS selftest for ECDH algorithm using P-224 and B-233 curves.
104 *) Use separate DRBG fields for internal and external flags. New function
105 FIPS_drbg_health_check() to perform on demand health checking. Add
106 generation tests to fips_test_suite with reduced health check interval to
107 demonstrate periodic health checking. Add "nodh" option to
108 fips_test_suite to skip very slow DH test.
111 *) New function FIPS_get_cipherbynid() to lookup FIPS supported ciphers
115 *) More extensive health check for DRBG checking many more failure modes.
116 New function FIPS_selftest_drbg_all() to handle every possible DRBG
117 combination: call this in fips_test_suite.
120 *) Add support for Dual EC DRBG from SP800-90. Update DRBG algorithm test
121 and POST to handle Dual EC cases.
124 *) Add support for canonical generation of DSA parameter 'g'. See
127 *) Add support for HMAC DRBG from SP800-90. Update DRBG algorithm test and
128 POST to handle HMAC cases.
131 *) Add functions FIPS_module_version() and FIPS_module_version_text()
132 to return numerical and string versions of the FIPS module number.
135 *) Rename FIPS_mode_set and FIPS_mode to FIPS_module_mode_set and
136 FIPS_module_mode. FIPS_mode and FIPS_mode_set will be implemented
137 outside the validated module in the FIPS capable OpenSSL.
140 *) Minor change to DRBG entropy callback semantics. In some cases
141 there is no multiple of the block length between min_len and
142 max_len. Allow the callback to return more than max_len bytes
143 of entropy but discard any extra: it is the callback's responsibility
144 to ensure that the extra data discarded does not impact the
145 requested amount of entropy.
148 *) Add PRNG security strength checks to RSA, DSA and ECDSA using
149 information in FIPS186-3, SP800-57 and SP800-131A.
152 *) CCM support via EVP. Interface is very similar to GCM case except we
153 must supply all data in one chunk (i.e. no update, final) and the
154 message length must be supplied if AAD is used. Add algorithm test
158 *) Initial version of POST overhaul. Add POST callback to allow the status
159 of POST to be monitored and/or failures induced. Modify fips_test_suite
160 to use callback. Always run all selftests even if one fails.
163 *) XTS support including algorithm test driver in the fips_gcmtest program.
164 Note: this does increase the maximum key length from 32 to 64 bytes but
165 there should be no binary compatibility issues as existing applications
166 will never use XTS mode.
169 *) Extensive reorganisation of FIPS PRNG behaviour. Remove all dependencies
170 to OpenSSL RAND code and replace with a tiny FIPS RAND API which also
171 performs algorithm blocking for unapproved PRNG types. Also do not
172 set PRNG type in FIPS_mode_set(): leave this to the application.
173 Add default OpenSSL DRBG handling: sets up FIPS PRNG and seeds with
174 the standard OpenSSL PRNG: set additional data to a date time vector.
177 *) Rename old X9.31 PRNG functions of the form FIPS_rand* to FIPS_x931*.
178 This shouldn't present any incompatibility problems because applications
179 shouldn't be using these directly and any that are will need to rethink
180 anyway as the X9.31 PRNG is now deprecated by FIPS 140-2
183 *) Extensive self tests and health checking required by SP800-90 DRBG.
184 Remove strength parameter from FIPS_drbg_instantiate and always
185 instantiate at maximum supported strength.
188 *) Add ECDH code to fips module and fips_ecdhvs for primitives only testing.
191 *) New algorithm test program fips_dhvs to handle DH primitives only testing.
194 *) New function DH_compute_key_padded() to compute a DH key and pad with
195 leading zeroes if needed: this complies with SP800-56A et al.
198 *) Initial implementation of SP800-90 DRBGs for Hash and CTR. Not used by
199 anything, incomplete, subject to change and largely untested at present.
202 *) Modify fipscanisteronly build option to only build the necessary object
203 files by filtering FIPS_EX_OBJ through a perl script in crypto/Makefile.
206 *) Add experimental option FIPSSYMS to give all symbols in
207 fipscanister.o and FIPS or fips prefix. This will avoid
208 conflicts with future versions of OpenSSL. Add perl script
209 util/fipsas.pl to preprocess assembly language source files
210 and rename any affected symbols.
213 *) Add selftest checks and algorithm block of non-fips algorithms in
214 FIPS mode. Remove DES2 from selftests.
217 *) Add ECDSA code to fips module. Add tiny fips_ecdsa_check to just
218 return internal method without any ENGINE dependencies. Add new
219 tiny fips sign and verify functions.
222 *) New build option no-ec2m to disable characteristic 2 code.
225 *) New build option "fipscanisteronly". This only builds fipscanister.o
226 and (currently) associated fips utilities. Uses the file Makefile.fips
227 instead of Makefile.org as the prototype.
230 *) Add some FIPS mode restrictions to GCM. Add internal IV generator.
231 Update fips_gcmtest to use IV generator.
234 *) Initial, experimental EVP support for AES-GCM. AAD can be input by
235 setting output buffer to NULL. The *Final function must be
236 called although it will not retrieve any additional data. The tag
237 can be set or retrieved with a ctrl. The IV length is by default 12
238 bytes (96 bits) but can be set to an alternative value. If the IV
239 length exceeds the maximum IV length (currently 16 bytes) it cannot be
243 *) New flag in ciphers: EVP_CIPH_FLAG_CUSTOM_CIPHER. This means the
244 underlying do_cipher function handles all cipher semantics itself
245 including padding and finalisation. This is useful if (for example)
246 an ENGINE cipher handles block padding itself. The behaviour of
247 do_cipher is subtly changed if this flag is set: the return value
248 is the number of characters written to the output buffer (zero is
249 no longer an error code) or a negative error code. Also if the
250 input buffer is NULL and length 0 finalisation should be performed.
253 *) If a candidate issuer certificate is already part of the constructed
254 path ignore it: new debug notification X509_V_ERR_PATH_LOOP for this case.
257 *) Improve forward-security support: add functions
259 void SSL_CTX_set_not_resumable_session_callback(SSL_CTX *ctx, int (*cb)(SSL *ssl, int is_forward_secure))
260 void SSL_set_not_resumable_session_callback(SSL *ssl, int (*cb)(SSL *ssl, int is_forward_secure))
262 for use by SSL/TLS servers; the callback function will be called whenever a
263 new session is created, and gets to decide whether the session may be
264 cached to make it resumable (return 0) or not (return 1). (As by the
265 SSL/TLS protocol specifications, the session_id sent by the server will be
266 empty to indicate that the session is not resumable; also, the server will
267 not generate RFC 4507 (RFC 5077) session tickets.)
269 A simple reasonable callback implementation is to return is_forward_secure.
270 This parameter will be set to 1 or 0 depending on the ciphersuite selected
271 by the SSL/TLS server library, indicating whether it can provide forward
273 [Emilia Käsper <emilia.kasper@esat.kuleuven.be> (Google)]
275 *) New -verify_name option in command line utilities to set verification
279 *) Initial CMAC implementation. WARNING: EXPERIMENTAL, API MAY CHANGE.
280 Add CMAC pkey methods.
283 *) Experimental renegotiation in s_server -www mode. If the client
284 browses /reneg connection is renegotiated. If /renegcert it is
285 renegotiated requesting a certificate.
288 *) Add an "external" session cache for debugging purposes to s_server. This
289 should help trace issues which normally are only apparent in deployed
290 multi-process servers.
293 *) Extensive audit of libcrypto with DEBUG_UNUSED. Fix many cases where
294 return value is ignored. NB. The functions RAND_add(), RAND_seed(),
295 BIO_set_cipher() and some obscure PEM functions were changed so they
296 can now return an error. The RAND changes required a change to the
297 RAND_METHOD structure.
300 *) New macro __owur for "OpenSSL Warn Unused Result". This makes use of
301 a gcc attribute to warn if the result of a function is ignored. This
302 is enable if DEBUG_UNUSED is set. Add to several functions in evp.h
303 whose return value is often ignored.
306 Changes between 1.0.1j and 1.0.2 [xx XXX xxxx]
308 *) Tighten client-side session ticket handling during renegotiation:
309 ensure that the client only accepts a session ticket if the server sends
310 the extension anew in the ServerHello. Previously, a TLS client would
311 reuse the old extension state and thus accept a session ticket if one was
312 announced in the initial ServerHello.
315 *) Accelerated NIST P-256 elliptic curve implementation for x86_64
316 (other platforms pending).
317 [Shay Gueron & Vlad Krasnov (Intel Corp), Andy Polyakov]
319 *) Add support for the SignedCertificateTimestampList certificate and
320 OCSP response extensions from RFC6962.
323 *) Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
324 for corner cases. (Certain input points at infinity could lead to
325 bogus results, with non-infinity inputs mapped to infinity too.)
328 *) Initial support for PowerISA 2.0.7, first implemented in POWER8.
329 This covers AES, SHA256/512 and GHASH. "Initial" means that most
330 common cases are optimized and there still is room for further
331 improvements. Vector Permutation AES for Altivec is also added.
334 *) Add support for little-endian ppc64 Linux target.
335 [Marcelo Cerri (IBM)]
337 *) Initial support for AMRv8 ISA crypto extensions. This covers AES,
338 SHA1, SHA256 and GHASH. "Initial" means that most common cases
339 are optimized and there still is room for further improvements.
340 Both 32- and 64-bit modes are supported.
341 [Andy Polyakov, Ard Biesheuvel (Linaro)]
343 *) Improved ARMv7 NEON support.
346 *) Support for SPARC Architecture 2011 crypto extensions, first
347 implemented in SPARC T4. This covers AES, DES, Camellia, SHA1,
348 SHA256/512, MD5, GHASH and modular exponentiation.
349 [Andy Polyakov, David Miller]
351 *) Accelerated modular exponentiation for Intel processors, a.k.a.
353 [Shay Gueron & Vlad Krasnov (Intel Corp)]
355 *) Support for new and upcoming Intel processors, including AVX2,
356 BMI and SHA ISA extensions. This includes additional "stitched"
357 implementations, AESNI-SHA256 and GCM, and multi-buffer support
360 This work was sponsored by Intel Corp.
363 *) Support for DTLS 1.2. This adds two sets of DTLS methods: DTLS_*_method()
364 supports both DTLS 1.2 and 1.0 and should use whatever version the peer
365 supports and DTLSv1_2_*_method() which supports DTLS 1.2 only.
368 *) Use algorithm specific chains in SSL_CTX_use_certificate_chain_file():
369 this fixes a limiation in previous versions of OpenSSL.
372 *) Extended RSA OAEP support via EVP_PKEY API. Options to specify digest,
373 MGF1 digest and OAEP label.
376 *) Add EVP support for key wrapping algorithms, to avoid problems with
377 existing code the flag EVP_CIPHER_CTX_WRAP_ALLOW has to be set in
378 the EVP_CIPHER_CTX or an error is returned. Add AES and DES3 wrap
379 algorithms and include tests cases.
382 *) Add functions to allocate and set the fields of an ECDSA_METHOD
384 [Douglas E. Engert, Steve Henson]
386 *) New functions OPENSSL_gmtime_diff and ASN1_TIME_diff to find the
387 difference in days and seconds between two tm or ASN1_TIME structures.
390 *) Add -rev test option to s_server to just reverse order of characters
391 received by client and send back to server. Also prints an abbreviated
392 summary of the connection parameters.
395 *) New option -brief for s_client and s_server to print out a brief summary
396 of connection parameters.
399 *) Add callbacks for arbitrary TLS extensions.
400 [Trevor Perrin <trevp@trevp.net> and Ben Laurie]
402 *) New option -crl_download in several openssl utilities to download CRLs
403 from CRLDP extension in certificates.
406 *) New options -CRL and -CRLform for s_client and s_server for CRLs.
409 *) New function X509_CRL_diff to generate a delta CRL from the difference
410 of two full CRLs. Add support to "crl" utility.
413 *) New functions to set lookup_crls function and to retrieve
414 X509_STORE from X509_STORE_CTX.
417 *) Print out deprecated issuer and subject unique ID fields in
421 *) Extend OCSP I/O functions so they can be used for simple general purpose
422 HTTP as well as OCSP. New wrapper function which can be used to download
423 CRLs using the OCSP API.
426 *) Delegate command line handling in s_client/s_server to SSL_CONF APIs.
429 *) SSL_CONF* functions. These provide a common framework for application
430 configuration using configuration files or command lines.
433 *) SSL/TLS tracing code. This parses out SSL/TLS records using the
434 message callback and prints the results. Needs compile time option
435 "enable-ssl-trace". New options to s_client and s_server to enable
439 *) New ctrl and macro to retrieve supported points extensions.
440 Print out extension in s_server and s_client.
443 *) New functions to retrieve certificate signature and signature
447 *) Add functions to retrieve and manipulate the raw cipherlist sent by a
451 *) New Suite B modes for TLS code. These use and enforce the requirements
452 of RFC6460: restrict ciphersuites, only permit Suite B algorithms and
453 only use Suite B curves. The Suite B modes can be set by using the
454 strings "SUITEB128", "SUITEB192" or "SUITEB128ONLY" for the cipherstring.
457 *) New chain verification flags for Suite B levels of security. Check
458 algorithms are acceptable when flags are set in X509_verify_cert.
461 *) Make tls1_check_chain return a set of flags indicating checks passed
462 by a certificate chain. Add additional tests to handle client
463 certificates: checks for matching certificate type and issuer name
467 *) If an attempt is made to use a signature algorithm not in the peer
468 preference list abort the handshake. If client has no suitable
469 signature algorithms in response to a certificate request do not
473 *) If server EC tmp key is not in client preference list abort handshake.
476 *) Add support for certificate stores in CERT structure. This makes it
477 possible to have different stores per SSL structure or one store in
478 the parent SSL_CTX. Include distint stores for certificate chain
479 verification and chain building. New ctrl SSL_CTRL_BUILD_CERT_CHAIN
480 to build and store a certificate chain in CERT structure: returing
481 an error if the chain cannot be built: this will allow applications
482 to test if a chain is correctly configured.
484 Note: if the CERT based stores are not set then the parent SSL_CTX
485 store is used to retain compatibility with existing behaviour.
489 *) New function ssl_set_client_disabled to set a ciphersuite disabled
490 mask based on the current session, check mask when sending client
491 hello and checking the requested ciphersuite.
494 *) New ctrls to retrieve and set certificate types in a certificate
495 request message. Print out received values in s_client. If certificate
496 types is not set with custom values set sensible values based on
497 supported signature algorithms.
500 *) Support for distinct client and server supported signature algorithms.
503 *) Add certificate callback. If set this is called whenever a certificate
504 is required by client or server. An application can decide which
505 certificate chain to present based on arbitrary criteria: for example
506 supported signature algorithms. Add very simple example to s_server.
507 This fixes many of the problems and restrictions of the existing client
508 certificate callback: for example you can now clear an existing
509 certificate and specify the whole chain.
512 *) Add new "valid_flags" field to CERT_PKEY structure which determines what
513 the certificate can be used for (if anything). Set valid_flags field
514 in new tls1_check_chain function. Simplify ssl_set_cert_masks which used
515 to have similar checks in it.
517 Add new "cert_flags" field to CERT structure and include a "strict mode".
518 This enforces some TLS certificate requirements (such as only permitting
519 certificate signature algorithms contained in the supported algorithms
520 extension) which some implementations ignore: this option should be used
521 with caution as it could cause interoperability issues.
524 *) Update and tidy signature algorithm extension processing. Work out
525 shared signature algorithms based on preferences and peer algorithms
526 and print them out in s_client and s_server. Abort handshake if no
527 shared signature algorithms.
530 *) Add new functions to allow customised supported signature algorithms
531 for SSL and SSL_CTX structures. Add options to s_client and s_server
535 *) New function SSL_certs_clear() to delete all references to certificates
536 from an SSL structure. Before this once a certificate had been added
537 it couldn't be removed.
540 *) Integrate hostname, email address and IP address checking with certificate
541 verification. New verify options supporting checking in opensl utility.
544 *) Fixes and wildcard matching support to hostname and email checking
545 functions. Add manual page.
546 [Florian Weimer (Red Hat Product Security Team)]
548 *) New functions to check a hostname email or IP address against a
549 certificate. Add options x509 utility to print results of checks against
553 *) Fix OCSP checking.
554 [Rob Stradling <rob.stradling@comodo.com> and Ben Laurie]
556 *) Initial experimental support for explicitly trusted non-root CAs.
557 OpenSSL still tries to build a complete chain to a root but if an
558 intermediate CA has a trust setting included that is used. The first
559 setting is used: whether to trust (e.g., -addtrust option to the x509
563 *) Add -trusted_first option which attempts to find certificates in the
564 trusted store even if an untrusted chain is also supplied.
567 *) MIPS assembly pack updates: support for MIPS32r2 and SmartMIPS ASE,
568 platform support for Linux and Android.
571 *) Support for linux-x32, ILP32 environment in x86_64 framework.
574 *) Experimental multi-implementation support for FIPS capable OpenSSL.
575 When in FIPS mode the approved implementations are used as normal,
576 when not in FIPS mode the internal unapproved versions are used instead.
577 This means that the FIPS capable OpenSSL isn't forced to use the
578 (often lower performance) FIPS implementations outside FIPS mode.
581 *) Transparently support X9.42 DH parameters when calling
582 PEM_read_bio_DHparameters. This means existing applications can handle
583 the new parameter format automatically.
586 *) Initial experimental support for X9.42 DH parameter format: mainly
587 to support use of 'q' parameter for RFC5114 parameters.
590 *) Add DH parameters from RFC5114 including test data to dhtest.
593 *) Support for automatic EC temporary key parameter selection. If enabled
594 the most preferred EC parameters are automatically used instead of
595 hardcoded fixed parameters. Now a server just has to call:
596 SSL_CTX_set_ecdh_auto(ctx, 1) and the server will automatically
597 support ECDH and use the most appropriate parameters.
600 *) Enhance and tidy EC curve and point format TLS extension code. Use
601 static structures instead of allocation if default values are used.
602 New ctrls to set curves we wish to support and to retrieve shared curves.
603 Print out shared curves in s_server. New options to s_server and s_client
604 to set list of supported curves.
607 *) New ctrls to retrieve supported signature algorithms and
608 supported curve values as an array of NIDs. Extend openssl utility
609 to print out received values.
612 *) Add new APIs EC_curve_nist2nid and EC_curve_nid2nist which convert
613 between NIDs and the more common NIST names such as "P-256". Enhance
614 ecparam utility and ECC method to recognise the NIST names for curves.
617 *) Enhance SSL/TLS certificate chain handling to support different
618 chains for each certificate instead of one chain in the parent SSL_CTX.
621 *) Support for fixed DH ciphersuite client authentication: where both
622 server and client use DH certificates with common parameters.
625 *) Support for fixed DH ciphersuites: those requiring DH server
629 *) New function i2d_re_X509_tbs for re-encoding the TBS portion of
631 Note: Related 1.0.2-beta specific macros X509_get_cert_info,
632 X509_CINF_set_modified, X509_CINF_get_issuer, X509_CINF_get_extensions and
633 X509_CINF_get_signature were reverted post internal team review.
635 Changes between 1.0.1j and 1.0.1k [xx XXX xxxx]
637 *) Tighten client-side session ticket handling during renegotiation:
638 ensure that the client only accepts a session ticket if the server sends
639 the extension anew in the ServerHello. Previously, a TLS client would
640 reuse the old extension state and thus accept a session ticket if one was
641 announced in the initial ServerHello.
644 Changes between 1.0.1i and 1.0.1j [15 Oct 2014]
648 A flaw in the DTLS SRTP extension parsing code allows an attacker, who
649 sends a carefully crafted handshake message, to cause OpenSSL to fail
650 to free up to 64k of memory causing a memory leak. This could be
651 exploited in a Denial Of Service attack. This issue affects OpenSSL
652 1.0.1 server implementations for both SSL/TLS and DTLS regardless of
653 whether SRTP is used or configured. Implementations of OpenSSL that
654 have been compiled with OPENSSL_NO_SRTP defined are not affected.
656 The fix was developed by the OpenSSL team.
660 *) Session Ticket Memory Leak.
662 When an OpenSSL SSL/TLS/DTLS server receives a session ticket the
663 integrity of that ticket is first verified. In the event of a session
664 ticket integrity check failing, OpenSSL will fail to free memory
665 causing a memory leak. By sending a large number of invalid session
666 tickets an attacker could exploit this issue in a Denial Of Service
671 *) Build option no-ssl3 is incomplete.
673 When OpenSSL is configured with "no-ssl3" as a build option, servers
674 could accept and complete a SSL 3.0 handshake, and clients could be
675 configured to send them.
677 [Akamai and the OpenSSL team]
679 *) Add support for TLS_FALLBACK_SCSV.
680 Client applications doing fallback retries should call
681 SSL_set_mode(s, SSL_MODE_SEND_FALLBACK_SCSV).
683 [Adam Langley, Bodo Moeller]
685 *) Add additional DigestInfo checks.
687 Reencode DigestInto in DER and check against the original when
688 verifying RSA signature: this will reject any improperly encoded
689 DigestInfo structures.
691 Note: this is a precautionary measure and no attacks are currently known.
695 Changes between 1.0.1h and 1.0.1i [6 Aug 2014]
697 *) Fix SRP buffer overrun vulnerability. Invalid parameters passed to the
698 SRP code can be overrun an internal buffer. Add sanity check that
699 g, A, B < N to SRP code.
701 Thanks to Sean Devlin and Watson Ladd of Cryptography Services, NCC
702 Group for discovering this issue.
706 *) A flaw in the OpenSSL SSL/TLS server code causes the server to negotiate
707 TLS 1.0 instead of higher protocol versions when the ClientHello message
708 is badly fragmented. This allows a man-in-the-middle attacker to force a
709 downgrade to TLS 1.0 even if both the server and the client support a
710 higher protocol version, by modifying the client's TLS records.
712 Thanks to David Benjamin and Adam Langley (Google) for discovering and
713 researching this issue.
717 *) OpenSSL DTLS clients enabling anonymous (EC)DH ciphersuites are subject
718 to a denial of service attack. A malicious server can crash the client
719 with a null pointer dereference (read) by specifying an anonymous (EC)DH
720 ciphersuite and sending carefully crafted handshake messages.
722 Thanks to Felix Gröbert (Google) for discovering and researching this
727 *) By sending carefully crafted DTLS packets an attacker could cause openssl
728 to leak memory. This can be exploited through a Denial of Service attack.
729 Thanks to Adam Langley for discovering and researching this issue.
733 *) An attacker can force openssl to consume large amounts of memory whilst
734 processing DTLS handshake messages. This can be exploited through a
735 Denial of Service attack.
736 Thanks to Adam Langley for discovering and researching this issue.
740 *) An attacker can force an error condition which causes openssl to crash
741 whilst processing DTLS packets due to memory being freed twice. This
742 can be exploited through a Denial of Service attack.
743 Thanks to Adam Langley and Wan-Teh Chang for discovering and researching
748 *) If a multithreaded client connects to a malicious server using a resumed
749 session and the server sends an ec point format extension it could write
750 up to 255 bytes to freed memory.
752 Thanks to Gabor Tyukasz (LogMeIn Inc) for discovering and researching this
757 *) A malicious server can crash an OpenSSL client with a null pointer
758 dereference (read) by specifying an SRP ciphersuite even though it was not
759 properly negotiated with the client. This can be exploited through a
760 Denial of Service attack.
762 Thanks to Joonas Kuorilehto and Riku Hietamäki (Codenomicon) for
763 discovering and researching this issue.
767 *) A flaw in OBJ_obj2txt may cause pretty printing functions such as
768 X509_name_oneline, X509_name_print_ex et al. to leak some information
769 from the stack. Applications may be affected if they echo pretty printing
770 output to the attacker.
772 Thanks to Ivan Fratric (Google) for discovering this issue.
774 [Emilia Käsper, and Steve Henson]
776 *) Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
777 for corner cases. (Certain input points at infinity could lead to
778 bogus results, with non-infinity inputs mapped to infinity too.)
781 Changes between 1.0.1g and 1.0.1h [5 Jun 2014]
783 *) Fix for SSL/TLS MITM flaw. An attacker using a carefully crafted
784 handshake can force the use of weak keying material in OpenSSL
785 SSL/TLS clients and servers.
787 Thanks to KIKUCHI Masashi (Lepidum Co. Ltd.) for discovering and
788 researching this issue. (CVE-2014-0224)
789 [KIKUCHI Masashi, Steve Henson]
791 *) Fix DTLS recursion flaw. By sending an invalid DTLS handshake to an
792 OpenSSL DTLS client the code can be made to recurse eventually crashing
795 Thanks to Imre Rad (Search-Lab Ltd.) for discovering this issue.
797 [Imre Rad, Steve Henson]
799 *) Fix DTLS invalid fragment vulnerability. A buffer overrun attack can
800 be triggered by sending invalid DTLS fragments to an OpenSSL DTLS
801 client or server. This is potentially exploitable to run arbitrary
802 code on a vulnerable client or server.
804 Thanks to Jüri Aedla for reporting this issue. (CVE-2014-0195)
805 [Jüri Aedla, Steve Henson]
807 *) Fix bug in TLS code where clients enable anonymous ECDH ciphersuites
808 are subject to a denial of service attack.
810 Thanks to Felix Gröbert and Ivan Fratric at Google for discovering
811 this issue. (CVE-2014-3470)
812 [Felix Gröbert, Ivan Fratric, Steve Henson]
814 *) Harmonize version and its documentation. -f flag is used to display
816 [mancha <mancha1@zoho.com>]
818 *) Fix eckey_priv_encode so it immediately returns an error upon a failure
820 [mancha <mancha1@zoho.com>]
822 *) Fix some double frees. These are not thought to be exploitable.
823 [mancha <mancha1@zoho.com>]
825 Changes between 1.0.1f and 1.0.1g [7 Apr 2014]
827 *) A missing bounds check in the handling of the TLS heartbeat extension
828 can be used to reveal up to 64k of memory to a connected client or
831 Thanks for Neel Mehta of Google Security for discovering this bug and to
832 Adam Langley <agl@chromium.org> and Bodo Moeller <bmoeller@acm.org> for
833 preparing the fix (CVE-2014-0160)
834 [Adam Langley, Bodo Moeller]
836 *) Fix for the attack described in the paper "Recovering OpenSSL
837 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
838 by Yuval Yarom and Naomi Benger. Details can be obtained from:
839 http://eprint.iacr.org/2014/140
841 Thanks to Yuval Yarom and Naomi Benger for discovering this
842 flaw and to Yuval Yarom for supplying a fix (CVE-2014-0076)
843 [Yuval Yarom and Naomi Benger]
845 *) TLS pad extension: draft-agl-tls-padding-03
847 Workaround for the "TLS hang bug" (see FAQ and PR#2771): if the
848 TLS client Hello record length value would otherwise be > 255 and
849 less that 512 pad with a dummy extension containing zeroes so it
850 is at least 512 bytes long.
852 [Adam Langley, Steve Henson]
854 Changes between 1.0.1e and 1.0.1f [6 Jan 2014]
856 *) Fix for TLS record tampering bug. A carefully crafted invalid
857 handshake could crash OpenSSL with a NULL pointer exception.
858 Thanks to Anton Johansson for reporting this issues.
861 *) Keep original DTLS digest and encryption contexts in retransmission
862 structures so we can use the previous session parameters if they need
863 to be resent. (CVE-2013-6450)
866 *) Add option SSL_OP_SAFARI_ECDHE_ECDSA_BUG (part of SSL_OP_ALL) which
867 avoids preferring ECDHE-ECDSA ciphers when the client appears to be
868 Safari on OS X. Safari on OS X 10.8..10.8.3 advertises support for
869 several ECDHE-ECDSA ciphers, but fails to negotiate them. The bug
870 is fixed in OS X 10.8.4, but Apple have ruled out both hot fixing
871 10.8..10.8.3 and forcing users to upgrade to 10.8.4 or newer.
872 [Rob Stradling, Adam Langley]
874 Changes between 1.0.1d and 1.0.1e [11 Feb 2013]
876 *) Correct fix for CVE-2013-0169. The original didn't work on AES-NI
877 supporting platforms or when small records were transferred.
878 [Andy Polyakov, Steve Henson]
880 Changes between 1.0.1c and 1.0.1d [5 Feb 2013]
882 *) Make the decoding of SSLv3, TLS and DTLS CBC records constant time.
884 This addresses the flaw in CBC record processing discovered by
885 Nadhem Alfardan and Kenny Paterson. Details of this attack can be found
886 at: http://www.isg.rhul.ac.uk/tls/
888 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
889 Security Group at Royal Holloway, University of London
890 (www.isg.rhul.ac.uk) for discovering this flaw and Adam Langley and
891 Emilia Käsper for the initial patch.
893 [Emilia Käsper, Adam Langley, Ben Laurie, Andy Polyakov, Steve Henson]
895 *) Fix flaw in AESNI handling of TLS 1.2 and 1.1 records for CBC mode
896 ciphersuites which can be exploited in a denial of service attack.
897 Thanks go to and to Adam Langley <agl@chromium.org> for discovering
898 and detecting this bug and to Wolfgang Ettlinger
899 <wolfgang.ettlinger@gmail.com> for independently discovering this issue.
903 *) Return an error when checking OCSP signatures when key is NULL.
904 This fixes a DoS attack. (CVE-2013-0166)
907 *) Make openssl verify return errors.
908 [Chris Palmer <palmer@google.com> and Ben Laurie]
910 *) Call OCSP Stapling callback after ciphersuite has been chosen, so
911 the right response is stapled. Also change SSL_get_certificate()
912 so it returns the certificate actually sent.
913 See http://rt.openssl.org/Ticket/Display.html?id=2836.
914 [Rob Stradling <rob.stradling@comodo.com>]
916 *) Fix possible deadlock when decoding public keys.
919 *) Don't use TLS 1.0 record version number in initial client hello
923 Changes between 1.0.1b and 1.0.1c [10 May 2012]
925 *) Sanity check record length before skipping explicit IV in TLS
926 1.2, 1.1 and DTLS to fix DoS attack.
928 Thanks to Codenomicon for discovering this issue using Fuzz-o-Matic
929 fuzzing as a service testing platform.
933 *) Initialise tkeylen properly when encrypting CMS messages.
934 Thanks to Solar Designer of Openwall for reporting this issue.
937 *) In FIPS mode don't try to use composite ciphers as they are not
941 Changes between 1.0.1a and 1.0.1b [26 Apr 2012]
943 *) OpenSSL 1.0.0 sets SSL_OP_ALL to 0x80000FFFL and OpenSSL 1.0.1 and
944 1.0.1a set SSL_OP_NO_TLSv1_1 to 0x00000400L which would unfortunately
945 mean any application compiled against OpenSSL 1.0.0 headers setting
946 SSL_OP_ALL would also set SSL_OP_NO_TLSv1_1, unintentionally disablng
947 TLS 1.1 also. Fix this by changing the value of SSL_OP_NO_TLSv1_1 to
948 0x10000000L Any application which was previously compiled against
949 OpenSSL 1.0.1 or 1.0.1a headers and which cares about SSL_OP_NO_TLSv1_1
950 will need to be recompiled as a result. Letting be results in
951 inability to disable specifically TLS 1.1 and in client context,
952 in unlike event, limit maximum offered version to TLS 1.0 [see below].
955 *) In order to ensure interoperabilty SSL_OP_NO_protocolX does not
956 disable just protocol X, but all protocols above X *if* there are
957 protocols *below* X still enabled. In more practical terms it means
958 that if application wants to disable TLS1.0 in favor of TLS1.1 and
959 above, it's not sufficient to pass SSL_OP_NO_TLSv1, one has to pass
960 SSL_OP_NO_TLSv1|SSL_OP_NO_SSLv3|SSL_OP_NO_SSLv2. This applies to
964 Changes between 1.0.1 and 1.0.1a [19 Apr 2012]
966 *) Check for potentially exploitable overflows in asn1_d2i_read_bio
967 BUF_mem_grow and BUF_mem_grow_clean. Refuse attempts to shrink buffer
968 in CRYPTO_realloc_clean.
970 Thanks to Tavis Ormandy, Google Security Team, for discovering this
971 issue and to Adam Langley <agl@chromium.org> for fixing it.
973 [Adam Langley (Google), Tavis Ormandy, Google Security Team]
975 *) Don't allow TLS 1.2 SHA-256 ciphersuites in TLS 1.0, 1.1 connections.
978 *) Workarounds for some broken servers that "hang" if a client hello
979 record length exceeds 255 bytes.
981 1. Do not use record version number > TLS 1.0 in initial client
982 hello: some (but not all) hanging servers will now work.
983 2. If we set OPENSSL_MAX_TLS1_2_CIPHER_LENGTH this will truncate
984 the number of ciphers sent in the client hello. This should be
985 set to an even number, such as 50, for example by passing:
986 -DOPENSSL_MAX_TLS1_2_CIPHER_LENGTH=50 to config or Configure.
987 Most broken servers should now work.
988 3. If all else fails setting OPENSSL_NO_TLS1_2_CLIENT will disable
989 TLS 1.2 client support entirely.
992 *) Fix SEGV in Vector Permutation AES module observed in OpenSSH.
995 Changes between 1.0.0h and 1.0.1 [14 Mar 2012]
997 *) Add compatibility with old MDC2 signatures which use an ASN1 OCTET
998 STRING form instead of a DigestInfo.
1001 *) The format used for MDC2 RSA signatures is inconsistent between EVP
1002 and the RSA_sign/RSA_verify functions. This was made more apparent when
1003 OpenSSL used RSA_sign/RSA_verify for some RSA signatures in particular
1004 those which went through EVP_PKEY_METHOD in 1.0.0 and later. Detect
1005 the correct format in RSA_verify so both forms transparently work.
1008 *) Some servers which support TLS 1.0 can choke if we initially indicate
1009 support for TLS 1.2 and later renegotiate using TLS 1.0 in the RSA
1010 encrypted premaster secret. As a workaround use the maximum permitted
1011 client version in client hello, this should keep such servers happy
1012 and still work with previous versions of OpenSSL.
1015 *) Add support for TLS/DTLS heartbeats.
1016 [Robin Seggelmann <seggelmann@fh-muenster.de>]
1018 *) Add support for SCTP.
1019 [Robin Seggelmann <seggelmann@fh-muenster.de>]
1021 *) Improved PRNG seeding for VOS.
1022 [Paul Green <Paul.Green@stratus.com>]
1024 *) Extensive assembler packs updates, most notably:
1026 - x86[_64]: AES-NI, PCLMULQDQ, RDRAND support;
1027 - x86[_64]: SSSE3 support (SHA1, vector-permutation AES);
1028 - x86_64: bit-sliced AES implementation;
1029 - ARM: NEON support, contemporary platforms optimizations;
1030 - s390x: z196 support;
1031 - *: GHASH and GF(2^m) multiplication implementations;
1035 *) Make TLS-SRP code conformant with RFC 5054 API cleanup
1036 (removal of unnecessary code)
1037 [Peter Sylvester <peter.sylvester@edelweb.fr>]
1039 *) Add TLS key material exporter from RFC 5705.
1042 *) Add DTLS-SRTP negotiation from RFC 5764.
1045 *) Add Next Protocol Negotiation,
1046 http://tools.ietf.org/html/draft-agl-tls-nextprotoneg-00. Can be
1047 disabled with a no-npn flag to config or Configure. Code donated
1049 [Adam Langley <agl@google.com> and Ben Laurie]
1051 *) Add optional 64-bit optimized implementations of elliptic curves NIST-P224,
1052 NIST-P256, NIST-P521, with constant-time single point multiplication on
1053 typical inputs. Compiler support for the nonstandard type __uint128_t is
1054 required to use this (present in gcc 4.4 and later, for 64-bit builds).
1055 Code made available under Apache License version 2.0.
1057 Specify "enable-ec_nistp_64_gcc_128" on the Configure (or config) command
1058 line to include this in your build of OpenSSL, and run "make depend" (or
1059 "make update"). This enables the following EC_METHODs:
1061 EC_GFp_nistp224_method()
1062 EC_GFp_nistp256_method()
1063 EC_GFp_nistp521_method()
1065 EC_GROUP_new_by_curve_name() will automatically use these (while
1066 EC_GROUP_new_curve_GFp() currently prefers the more flexible
1068 [Emilia Käsper, Adam Langley, Bodo Moeller (Google)]
1070 *) Use type ossl_ssize_t instad of ssize_t which isn't available on
1071 all platforms. Move ssize_t definition from e_os.h to the public
1072 header file e_os2.h as it now appears in public header file cms.h
1075 *) New -sigopt option to the ca, req and x509 utilities. Additional
1076 signature parameters can be passed using this option and in
1080 *) Add RSA PSS signing function. This will generate and set the
1081 appropriate AlgorithmIdentifiers for PSS based on those in the
1082 corresponding EVP_MD_CTX structure. No application support yet.
1085 *) Support for companion algorithm specific ASN1 signing routines.
1086 New function ASN1_item_sign_ctx() signs a pre-initialised
1087 EVP_MD_CTX structure and sets AlgorithmIdentifiers based on
1088 the appropriate parameters.
1091 *) Add new algorithm specific ASN1 verification initialisation function
1092 to EVP_PKEY_ASN1_METHOD: this is not in EVP_PKEY_METHOD since the ASN1
1093 handling will be the same no matter what EVP_PKEY_METHOD is used.
1094 Add a PSS handler to support verification of PSS signatures: checked
1095 against a number of sample certificates.
1098 *) Add signature printing for PSS. Add PSS OIDs.
1099 [Steve Henson, Martin Kaiser <lists@kaiser.cx>]
1101 *) Add algorithm specific signature printing. An individual ASN1 method
1102 can now print out signatures instead of the standard hex dump.
1104 More complex signatures (e.g. PSS) can print out more meaningful
1105 information. Include DSA version that prints out the signature
1109 *) Password based recipient info support for CMS library: implementing
1113 *) Split password based encryption into PBES2 and PBKDF2 functions. This
1114 neatly separates the code into cipher and PBE sections and is required
1115 for some algorithms that split PBES2 into separate pieces (such as
1116 password based CMS).
1119 *) Session-handling fixes:
1120 - Fix handling of connections that are resuming with a session ID,
1121 but also support Session Tickets.
1122 - Fix a bug that suppressed issuing of a new ticket if the client
1123 presented a ticket with an expired session.
1124 - Try to set the ticket lifetime hint to something reasonable.
1125 - Make tickets shorter by excluding irrelevant information.
1126 - On the client side, don't ignore renewed tickets.
1127 [Adam Langley, Bodo Moeller (Google)]
1129 *) Fix PSK session representation.
1132 *) Add RC4-MD5 and AESNI-SHA1 "stitched" implementations.
1134 This work was sponsored by Intel.
1137 *) Add GCM support to TLS library. Some custom code is needed to split
1138 the IV between the fixed (from PRF) and explicit (from TLS record)
1139 portions. This adds all GCM ciphersuites supported by RFC5288 and
1140 RFC5289. Generalise some AES* cipherstrings to include GCM and
1141 add a special AESGCM string for GCM only.
1144 *) Expand range of ctrls for AES GCM. Permit setting invocation
1145 field on decrypt and retrieval of invocation field only on encrypt.
1148 *) Add HMAC ECC ciphersuites from RFC5289. Include SHA384 PRF support.
1149 As required by RFC5289 these ciphersuites cannot be used if for
1150 versions of TLS earlier than 1.2.
1153 *) For FIPS capable OpenSSL interpret a NULL default public key method
1154 as unset and return the appropriate default but do *not* set the default.
1155 This means we can return the appropriate method in applications that
1156 switch between FIPS and non-FIPS modes.
1159 *) Redirect HMAC and CMAC operations to FIPS module in FIPS mode. If an
1160 ENGINE is used then we cannot handle that in the FIPS module so we
1161 keep original code iff non-FIPS operations are allowed.
1164 *) Add -attime option to openssl utilities.
1165 [Peter Eckersley <pde@eff.org>, Ben Laurie and Steve Henson]
1167 *) Redirect DSA and DH operations to FIPS module in FIPS mode.
1170 *) Redirect ECDSA and ECDH operations to FIPS module in FIPS mode. Also use
1171 FIPS EC methods unconditionally for now.
1174 *) New build option no-ec2m to disable characteristic 2 code.
1177 *) Backport libcrypto audit of return value checking from 1.1.0-dev; not
1178 all cases can be covered as some introduce binary incompatibilities.
1181 *) Redirect RSA operations to FIPS module including keygen,
1182 encrypt, decrypt, sign and verify. Block use of non FIPS RSA methods.
1185 *) Add similar low level API blocking to ciphers.
1188 *) Low level digest APIs are not approved in FIPS mode: any attempt
1189 to use these will cause a fatal error. Applications that *really* want
1190 to use them can use the private_* version instead.
1193 *) Redirect cipher operations to FIPS module for FIPS builds.
1196 *) Redirect digest operations to FIPS module for FIPS builds.
1199 *) Update build system to add "fips" flag which will link in fipscanister.o
1200 for static and shared library builds embedding a signature if needed.
1203 *) Output TLS supported curves in preference order instead of numerical
1204 order. This is currently hardcoded for the highest order curves first.
1205 This should be configurable so applications can judge speed vs strength.
1208 *) Add TLS v1.2 server support for client authentication.
1211 *) Add support for FIPS mode in ssl library: disable SSLv3, non-FIPS ciphers
1215 *) Functions FIPS_mode_set() and FIPS_mode() which call the underlying
1216 FIPS modules versions.
1219 *) Add TLS v1.2 client side support for client authentication. Keep cache
1220 of handshake records longer as we don't know the hash algorithm to use
1221 until after the certificate request message is received.
1224 *) Initial TLS v1.2 client support. Add a default signature algorithms
1225 extension including all the algorithms we support. Parse new signature
1226 format in client key exchange. Relax some ECC signing restrictions for
1227 TLS v1.2 as indicated in RFC5246.
1230 *) Add server support for TLS v1.2 signature algorithms extension. Switch
1231 to new signature format when needed using client digest preference.
1232 All server ciphersuites should now work correctly in TLS v1.2. No client
1233 support yet and no support for client certificates.
1236 *) Initial TLS v1.2 support. Add new SHA256 digest to ssl code, switch
1237 to SHA256 for PRF when using TLS v1.2 and later. Add new SHA256 based
1238 ciphersuites. At present only RSA key exchange ciphersuites work with
1239 TLS v1.2. Add new option for TLS v1.2 replacing the old and obsolete
1240 SSL_OP_PKCS1_CHECK flags with SSL_OP_NO_TLSv1_2. New TLSv1.2 methods
1241 and version checking.
1244 *) New option OPENSSL_NO_SSL_INTERN. If an application can be compiled
1245 with this defined it will not be affected by any changes to ssl internal
1246 structures. Add several utility functions to allow openssl application
1247 to work with OPENSSL_NO_SSL_INTERN defined.
1251 [Tom Wu <tjw@cs.stanford.edu> and Ben Laurie]
1253 *) Add functions to copy EVP_PKEY_METHOD and retrieve flags and id.
1256 *) Permit abbreviated handshakes when renegotiating using the function
1257 SSL_renegotiate_abbreviated().
1258 [Robin Seggelmann <seggelmann@fh-muenster.de>]
1260 *) Add call to ENGINE_register_all_complete() to
1261 ENGINE_load_builtin_engines(), so some implementations get used
1262 automatically instead of needing explicit application support.
1265 *) Add support for TLS key exporter as described in RFC5705.
1266 [Robin Seggelmann <seggelmann@fh-muenster.de>, Steve Henson]
1268 *) Initial TLSv1.1 support. Since TLSv1.1 is very similar to TLS v1.0 only
1269 a few changes are required:
1271 Add SSL_OP_NO_TLSv1_1 flag.
1272 Add TLSv1_1 methods.
1273 Update version checking logic to handle version 1.1.
1274 Add explicit IV handling (ported from DTLS code).
1275 Add command line options to s_client/s_server.
1278 Changes between 1.0.0j and 1.0.0k [5 Feb 2013]
1280 *) Make the decoding of SSLv3, TLS and DTLS CBC records constant time.
1282 This addresses the flaw in CBC record processing discovered by
1283 Nadhem Alfardan and Kenny Paterson. Details of this attack can be found
1284 at: http://www.isg.rhul.ac.uk/tls/
1286 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
1287 Security Group at Royal Holloway, University of London
1288 (www.isg.rhul.ac.uk) for discovering this flaw and Adam Langley and
1289 Emilia Käsper for the initial patch.
1291 [Emilia Käsper, Adam Langley, Ben Laurie, Andy Polyakov, Steve Henson]
1293 *) Return an error when checking OCSP signatures when key is NULL.
1294 This fixes a DoS attack. (CVE-2013-0166)
1297 *) Call OCSP Stapling callback after ciphersuite has been chosen, so
1298 the right response is stapled. Also change SSL_get_certificate()
1299 so it returns the certificate actually sent.
1300 See http://rt.openssl.org/Ticket/Display.html?id=2836.
1301 (This is a backport)
1302 [Rob Stradling <rob.stradling@comodo.com>]
1304 *) Fix possible deadlock when decoding public keys.
1307 Changes between 1.0.0i and 1.0.0j [10 May 2012]
1309 [NB: OpenSSL 1.0.0i and later 1.0.0 patch levels were released after
1312 *) Sanity check record length before skipping explicit IV in DTLS
1315 Thanks to Codenomicon for discovering this issue using Fuzz-o-Matic
1316 fuzzing as a service testing platform.
1320 *) Initialise tkeylen properly when encrypting CMS messages.
1321 Thanks to Solar Designer of Openwall for reporting this issue.
1324 Changes between 1.0.0h and 1.0.0i [19 Apr 2012]
1326 *) Check for potentially exploitable overflows in asn1_d2i_read_bio
1327 BUF_mem_grow and BUF_mem_grow_clean. Refuse attempts to shrink buffer
1328 in CRYPTO_realloc_clean.
1330 Thanks to Tavis Ormandy, Google Security Team, for discovering this
1331 issue and to Adam Langley <agl@chromium.org> for fixing it.
1333 [Adam Langley (Google), Tavis Ormandy, Google Security Team]
1335 Changes between 1.0.0g and 1.0.0h [12 Mar 2012]
1337 *) Fix MMA (Bleichenbacher's attack on PKCS #1 v1.5 RSA padding) weakness
1338 in CMS and PKCS7 code. When RSA decryption fails use a random key for
1339 content decryption and always return the same error. Note: this attack
1340 needs on average 2^20 messages so it only affects automated senders. The
1341 old behaviour can be reenabled in the CMS code by setting the
1342 CMS_DEBUG_DECRYPT flag: this is useful for debugging and testing where
1343 an MMA defence is not necessary.
1344 Thanks to Ivan Nestlerode <inestlerode@us.ibm.com> for discovering
1345 this issue. (CVE-2012-0884)
1348 *) Fix CVE-2011-4619: make sure we really are receiving a
1349 client hello before rejecting multiple SGC restarts. Thanks to
1350 Ivan Nestlerode <inestlerode@us.ibm.com> for discovering this bug.
1353 Changes between 1.0.0f and 1.0.0g [18 Jan 2012]
1355 *) Fix for DTLS DoS issue introduced by fix for CVE-2011-4109.
1356 Thanks to Antonio Martin, Enterprise Secure Access Research and
1357 Development, Cisco Systems, Inc. for discovering this bug and
1358 preparing a fix. (CVE-2012-0050)
1361 Changes between 1.0.0e and 1.0.0f [4 Jan 2012]
1363 *) Nadhem Alfardan and Kenny Paterson have discovered an extension
1364 of the Vaudenay padding oracle attack on CBC mode encryption
1365 which enables an efficient plaintext recovery attack against
1366 the OpenSSL implementation of DTLS. Their attack exploits timing
1367 differences arising during decryption processing. A research
1368 paper describing this attack can be found at:
1369 http://www.isg.rhul.ac.uk/~kp/dtls.pdf
1370 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
1371 Security Group at Royal Holloway, University of London
1372 (www.isg.rhul.ac.uk) for discovering this flaw and to Robin Seggelmann
1373 <seggelmann@fh-muenster.de> and Michael Tuexen <tuexen@fh-muenster.de>
1374 for preparing the fix. (CVE-2011-4108)
1375 [Robin Seggelmann, Michael Tuexen]
1377 *) Clear bytes used for block padding of SSL 3.0 records.
1379 [Adam Langley (Google)]
1381 *) Only allow one SGC handshake restart for SSL/TLS. Thanks to George
1382 Kadianakis <desnacked@gmail.com> for discovering this issue and
1383 Adam Langley for preparing the fix. (CVE-2011-4619)
1384 [Adam Langley (Google)]
1386 *) Check parameters are not NULL in GOST ENGINE. (CVE-2012-0027)
1387 [Andrey Kulikov <amdeich@gmail.com>]
1389 *) Prevent malformed RFC3779 data triggering an assertion failure.
1390 Thanks to Andrew Chi, BBN Technologies, for discovering the flaw
1391 and Rob Austein <sra@hactrn.net> for fixing it. (CVE-2011-4577)
1392 [Rob Austein <sra@hactrn.net>]
1394 *) Improved PRNG seeding for VOS.
1395 [Paul Green <Paul.Green@stratus.com>]
1397 *) Fix ssl_ciph.c set-up race.
1398 [Adam Langley (Google)]
1400 *) Fix spurious failures in ecdsatest.c.
1401 [Emilia Käsper (Google)]
1403 *) Fix the BIO_f_buffer() implementation (which was mixing different
1404 interpretations of the '..._len' fields).
1405 [Adam Langley (Google)]
1407 *) Fix handling of BN_BLINDING: now BN_BLINDING_invert_ex (rather than
1408 BN_BLINDING_invert_ex) calls BN_BLINDING_update, ensuring that concurrent
1409 threads won't reuse the same blinding coefficients.
1411 This also avoids the need to obtain the CRYPTO_LOCK_RSA_BLINDING
1412 lock to call BN_BLINDING_invert_ex, and avoids one use of
1413 BN_BLINDING_update for each BN_BLINDING structure (previously,
1414 the last update always remained unused).
1415 [Emilia Käsper (Google)]
1417 *) In ssl3_clear, preserve s3->init_extra along with s3->rbuf.
1418 [Bob Buckholz (Google)]
1420 Changes between 1.0.0d and 1.0.0e [6 Sep 2011]
1422 *) Fix bug where CRLs with nextUpdate in the past are sometimes accepted
1423 by initialising X509_STORE_CTX properly. (CVE-2011-3207)
1424 [Kaspar Brand <ossl@velox.ch>]
1426 *) Fix SSL memory handling for (EC)DH ciphersuites, in particular
1427 for multi-threaded use of ECDH. (CVE-2011-3210)
1428 [Adam Langley (Google)]
1430 *) Fix x509_name_ex_d2i memory leak on bad inputs.
1433 *) Remove hard coded ecdsaWithSHA1 signature tests in ssl code and check
1434 signature public key algorithm by using OID xref utilities instead.
1435 Before this you could only use some ECC ciphersuites with SHA1 only.
1438 *) Add protection against ECDSA timing attacks as mentioned in the paper
1439 by Billy Bob Brumley and Nicola Tuveri, see:
1441 http://eprint.iacr.org/2011/232.pdf
1443 [Billy Bob Brumley and Nicola Tuveri]
1445 Changes between 1.0.0c and 1.0.0d [8 Feb 2011]
1447 *) Fix parsing of OCSP stapling ClientHello extension. CVE-2011-0014
1448 [Neel Mehta, Adam Langley, Bodo Moeller (Google)]
1450 *) Fix bug in string printing code: if *any* escaping is enabled we must
1451 escape the escape character (backslash) or the resulting string is
1455 Changes between 1.0.0b and 1.0.0c [2 Dec 2010]
1457 *) Disable code workaround for ancient and obsolete Netscape browsers
1458 and servers: an attacker can use it in a ciphersuite downgrade attack.
1459 Thanks to Martin Rex for discovering this bug. CVE-2010-4180
1462 *) Fixed J-PAKE implementation error, originally discovered by
1463 Sebastien Martini, further info and confirmation from Stefan
1464 Arentz and Feng Hao. Note that this fix is a security fix. CVE-2010-4252
1467 Changes between 1.0.0a and 1.0.0b [16 Nov 2010]
1469 *) Fix extension code to avoid race conditions which can result in a buffer
1470 overrun vulnerability: resumed sessions must not be modified as they can
1471 be shared by multiple threads. CVE-2010-3864
1474 *) Fix WIN32 build system to correctly link an ENGINE directory into
1478 Changes between 1.0.0 and 1.0.0a [01 Jun 2010]
1480 *) Check return value of int_rsa_verify in pkey_rsa_verifyrecover
1482 [Steve Henson, Peter-Michael Hager <hager@dortmund.net>]
1484 Changes between 0.9.8n and 1.0.0 [29 Mar 2010]
1486 *) Add "missing" function EVP_CIPHER_CTX_copy(). This copies a cipher
1487 context. The operation can be customised via the ctrl mechanism in
1488 case ENGINEs want to include additional functionality.
1491 *) Tolerate yet another broken PKCS#8 key format: private key value negative.
1494 *) Add new -subject_hash_old and -issuer_hash_old options to x509 utility to
1495 output hashes compatible with older versions of OpenSSL.
1496 [Willy Weisz <weisz@vcpc.univie.ac.at>]
1498 *) Fix compression algorithm handling: if resuming a session use the
1499 compression algorithm of the resumed session instead of determining
1500 it from client hello again. Don't allow server to change algorithm.
1503 *) Add load_crls() function to apps tidying load_certs() too. Add option
1504 to verify utility to allow additional CRLs to be included.
1507 *) Update OCSP request code to permit adding custom headers to the request:
1508 some responders need this.
1511 *) The function EVP_PKEY_sign() returns <=0 on error: check return code
1513 [Julia Lawall <julia@diku.dk>]
1515 *) Update verify callback code in apps/s_cb.c and apps/verify.c, it
1516 needlessly dereferenced structures, used obsolete functions and
1517 didn't handle all updated verify codes correctly.
1520 *) Disable MD2 in the default configuration.
1523 *) In BIO_pop() and BIO_push() use the ctrl argument (which was NULL) to
1524 indicate the initial BIO being pushed or popped. This makes it possible
1525 to determine whether the BIO is the one explicitly called or as a result
1526 of the ctrl being passed down the chain. Fix BIO_pop() and SSL BIOs so
1527 it handles reference counts correctly and doesn't zero out the I/O bio
1528 when it is not being explicitly popped. WARNING: applications which
1529 included workarounds for the old buggy behaviour will need to be modified
1530 or they could free up already freed BIOs.
1533 *) Extend the uni2asc/asc2uni => OPENSSL_uni2asc/OPENSSL_asc2uni
1534 renaming to all platforms (within the 0.9.8 branch, this was
1535 done conditionally on Netware platforms to avoid a name clash).
1536 [Guenter <lists@gknw.net>]
1538 *) Add ECDHE and PSK support to DTLS.
1539 [Michael Tuexen <tuexen@fh-muenster.de>]
1541 *) Add CHECKED_STACK_OF macro to safestack.h, otherwise safestack can't
1545 *) Add "missing" function EVP_MD_flags() (without this the only way to
1546 retrieve a digest flags is by accessing the structure directly. Update
1547 EVP_MD_do_all*() and EVP_CIPHER_do_all*() to include the name a digest
1548 or cipher is registered as in the "from" argument. Print out all
1549 registered digests in the dgst usage message instead of manually
1550 attempting to work them out.
1553 *) If no SSLv2 ciphers are used don't use an SSLv2 compatible client hello:
1554 this allows the use of compression and extensions. Change default cipher
1555 string to remove SSLv2 ciphersuites. This effectively avoids ancient SSLv2
1556 by default unless an application cipher string requests it.
1559 *) Alter match criteria in PKCS12_parse(). It used to try to use local
1560 key ids to find matching certificates and keys but some PKCS#12 files
1561 don't follow the (somewhat unwritten) rules and this strategy fails.
1562 Now just gather all certificates together and the first private key
1563 then look for the first certificate that matches the key.
1566 *) Support use of registered digest and cipher names for dgst and cipher
1567 commands instead of having to add each one as a special case. So now
1574 openssl dgst -sha256 foo
1576 and this works for ENGINE based algorithms too.
1580 *) Update Gost ENGINE to support parameter files.
1581 [Victor B. Wagner <vitus@cryptocom.ru>]
1583 *) Support GeneralizedTime in ca utility.
1584 [Oliver Martin <oliver@volatilevoid.net>, Steve Henson]
1586 *) Enhance the hash format used for certificate directory links. The new
1587 form uses the canonical encoding (meaning equivalent names will work
1588 even if they aren't identical) and uses SHA1 instead of MD5. This form
1589 is incompatible with the older format and as a result c_rehash should
1590 be used to rebuild symbolic links.
1593 *) Make PKCS#8 the default write format for private keys, replacing the
1594 traditional format. This form is standardised, more secure and doesn't
1595 include an implicit MD5 dependency.
1598 *) Add a $gcc_devteam_warn option to Configure. The idea is that any code
1599 committed to OpenSSL should pass this lot as a minimum.
1602 *) Add session ticket override functionality for use by EAP-FAST.
1603 [Jouni Malinen <j@w1.fi>]
1605 *) Modify HMAC functions to return a value. Since these can be implemented
1606 in an ENGINE errors can occur.
1609 *) Type-checked OBJ_bsearch_ex.
1612 *) Type-checked OBJ_bsearch. Also some constification necessitated
1613 by type-checking. Still to come: TXT_DB, bsearch(?),
1614 OBJ_bsearch_ex, qsort, CRYPTO_EX_DATA, ASN1_VALUE, ASN1_STRING,
1618 *) New function OPENSSL_gmtime_adj() to add a specific number of days and
1619 seconds to a tm structure directly, instead of going through OS
1620 specific date routines. This avoids any issues with OS routines such
1621 as the year 2038 bug. New *_adj() functions for ASN1 time structures
1622 and X509_time_adj_ex() to cover the extended range. The existing
1623 X509_time_adj() is still usable and will no longer have any date issues.
1626 *) Delta CRL support. New use deltas option which will attempt to locate
1627 and search any appropriate delta CRLs available.
1629 This work was sponsored by Google.
1632 *) Support for CRLs partitioned by reason code. Reorganise CRL processing
1633 code and add additional score elements. Validate alternate CRL paths
1634 as part of the CRL checking and indicate a new error "CRL path validation
1635 error" in this case. Applications wanting additional details can use
1636 the verify callback and check the new "parent" field. If this is not
1637 NULL CRL path validation is taking place. Existing applications wont
1638 see this because it requires extended CRL support which is off by
1641 This work was sponsored by Google.
1644 *) Support for freshest CRL extension.
1646 This work was sponsored by Google.
1649 *) Initial indirect CRL support. Currently only supported in the CRLs
1650 passed directly and not via lookup. Process certificate issuer
1651 CRL entry extension and lookup CRL entries by bother issuer name
1652 and serial number. Check and process CRL issuer entry in IDP extension.
1654 This work was sponsored by Google.
1657 *) Add support for distinct certificate and CRL paths. The CRL issuer
1658 certificate is validated separately in this case. Only enabled if
1659 an extended CRL support flag is set: this flag will enable additional
1660 CRL functionality in future.
1662 This work was sponsored by Google.
1665 *) Add support for policy mappings extension.
1667 This work was sponsored by Google.
1670 *) Fixes to pathlength constraint, self issued certificate handling,
1671 policy processing to align with RFC3280 and PKITS tests.
1673 This work was sponsored by Google.
1676 *) Support for name constraints certificate extension. DN, email, DNS
1677 and URI types are currently supported.
1679 This work was sponsored by Google.
1682 *) To cater for systems that provide a pointer-based thread ID rather
1683 than numeric, deprecate the current numeric thread ID mechanism and
1684 replace it with a structure and associated callback type. This
1685 mechanism allows a numeric "hash" to be extracted from a thread ID in
1686 either case, and on platforms where pointers are larger than 'long',
1687 mixing is done to help ensure the numeric 'hash' is usable even if it
1688 can't be guaranteed unique. The default mechanism is to use "&errno"
1689 as a pointer-based thread ID to distinguish between threads.
1691 Applications that want to provide their own thread IDs should now use
1692 CRYPTO_THREADID_set_callback() to register a callback that will call
1693 either CRYPTO_THREADID_set_numeric() or CRYPTO_THREADID_set_pointer().
1695 Note that ERR_remove_state() is now deprecated, because it is tied
1696 to the assumption that thread IDs are numeric. ERR_remove_state(0)
1697 to free the current thread's error state should be replaced by
1698 ERR_remove_thread_state(NULL).
1700 (This new approach replaces the functions CRYPTO_set_idptr_callback(),
1701 CRYPTO_get_idptr_callback(), and CRYPTO_thread_idptr() that existed in
1702 OpenSSL 0.9.9-dev between June 2006 and August 2008. Also, if an
1703 application was previously providing a numeric thread callback that
1704 was inappropriate for distinguishing threads, then uniqueness might
1705 have been obtained with &errno that happened immediately in the
1706 intermediate development versions of OpenSSL; this is no longer the
1707 case, the numeric thread callback will now override the automatic use
1709 [Geoff Thorpe, with help from Bodo Moeller]
1711 *) Initial support for different CRL issuing certificates. This covers a
1712 simple case where the self issued certificates in the chain exist and
1713 the real CRL issuer is higher in the existing chain.
1715 This work was sponsored by Google.
1718 *) Removed effectively defunct crypto/store from the build.
1721 *) Revamp of STACK to provide stronger type-checking. Still to come:
1722 TXT_DB, bsearch(?), OBJ_bsearch, qsort, CRYPTO_EX_DATA, ASN1_VALUE,
1723 ASN1_STRING, CONF_VALUE.
1726 *) Add a new SSL_MODE_RELEASE_BUFFERS mode flag to release unused buffer
1727 RAM on SSL connections. This option can save about 34k per idle SSL.
1730 *) Revamp of LHASH to provide stronger type-checking. Still to come:
1731 STACK, TXT_DB, bsearch, qsort.
1734 *) Initial support for Cryptographic Message Syntax (aka CMS) based
1735 on RFC3850, RFC3851 and RFC3852. New cms directory and cms utility,
1736 support for data, signedData, compressedData, digestedData and
1737 encryptedData, envelopedData types included. Scripts to check against
1738 RFC4134 examples draft and interop and consistency checks of many
1739 content types and variants.
1742 *) Add options to enc utility to support use of zlib compression BIO.
1745 *) Extend mk1mf to support importing of options and assembly language
1746 files from Configure script, currently only included in VC-WIN32.
1747 The assembly language rules can now optionally generate the source
1748 files from the associated perl scripts.
1751 *) Implement remaining functionality needed to support GOST ciphersuites.
1752 Interop testing has been performed using CryptoPro implementations.
1753 [Victor B. Wagner <vitus@cryptocom.ru>]
1755 *) s390x assembler pack.
1758 *) ARMv4 assembler pack. ARMv4 refers to v4 and later ISA, not CPU
1762 *) Implement Opaque PRF Input TLS extension as specified in
1763 draft-rescorla-tls-opaque-prf-input-00.txt. Since this is not an
1764 official specification yet and no extension type assignment by
1765 IANA exists, this extension (for now) will have to be explicitly
1766 enabled when building OpenSSL by providing the extension number
1767 to use. For example, specify an option
1769 -DTLSEXT_TYPE_opaque_prf_input=0x9527
1771 to the "config" or "Configure" script to enable the extension,
1772 assuming extension number 0x9527 (which is a completely arbitrary
1773 and unofficial assignment based on the MD5 hash of the Internet
1774 Draft). Note that by doing so, you potentially lose
1775 interoperability with other TLS implementations since these might
1776 be using the same extension number for other purposes.
1778 SSL_set_tlsext_opaque_prf_input(ssl, src, len) is used to set the
1779 opaque PRF input value to use in the handshake. This will create
1780 an interal copy of the length-'len' string at 'src', and will
1781 return non-zero for success.
1783 To get more control and flexibility, provide a callback function
1786 SSL_CTX_set_tlsext_opaque_prf_input_callback(ctx, cb)
1787 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg(ctx, arg)
1791 int (*cb)(SSL *, void *peerinput, size_t len, void *arg);
1794 Callback function 'cb' will be called in handshakes, and is
1795 expected to use SSL_set_tlsext_opaque_prf_input() as appropriate.
1796 Argument 'arg' is for application purposes (the value as given to
1797 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg() will directly
1798 be provided to the callback function). The callback function
1799 has to return non-zero to report success: usually 1 to use opaque
1800 PRF input just if possible, or 2 to enforce use of the opaque PRF
1801 input. In the latter case, the library will abort the handshake
1802 if opaque PRF input is not successfully negotiated.
1804 Arguments 'peerinput' and 'len' given to the callback function
1805 will always be NULL and 0 in the case of a client. A server will
1806 see the client's opaque PRF input through these variables if
1807 available (NULL and 0 otherwise). Note that if the server
1808 provides an opaque PRF input, the length must be the same as the
1809 length of the client's opaque PRF input.
1811 Note that the callback function will only be called when creating
1812 a new session (session resumption can resume whatever was
1813 previously negotiated), and will not be called in SSL 2.0
1814 handshakes; thus, SSL_CTX_set_options(ctx, SSL_OP_NO_SSLv2) or
1815 SSL_set_options(ssl, SSL_OP_NO_SSLv2) is especially recommended
1816 for applications that need to enforce opaque PRF input.
1820 *) Update ssl code to support digests other than SHA1+MD5 for handshake
1823 [Victor B. Wagner <vitus@cryptocom.ru>]
1825 *) Add RFC4507 support to OpenSSL. This includes the corrections in
1826 RFC4507bis. The encrypted ticket format is an encrypted encoded
1827 SSL_SESSION structure, that way new session features are automatically
1830 If a client application caches session in an SSL_SESSION structure
1831 support is transparent because tickets are now stored in the encoded
1834 The SSL_CTX structure automatically generates keys for ticket
1835 protection in servers so again support should be possible
1836 with no application modification.
1838 If a client or server wishes to disable RFC4507 support then the option
1839 SSL_OP_NO_TICKET can be set.
1841 Add a TLS extension debugging callback to allow the contents of any client
1842 or server extensions to be examined.
1844 This work was sponsored by Google.
1847 *) Final changes to avoid use of pointer pointer casts in OpenSSL.
1848 OpenSSL should now compile cleanly on gcc 4.2
1849 [Peter Hartley <pdh@utter.chaos.org.uk>, Steve Henson]
1851 *) Update SSL library to use new EVP_PKEY MAC API. Include generic MAC
1852 support including streaming MAC support: this is required for GOST
1853 ciphersuite support.
1854 [Victor B. Wagner <vitus@cryptocom.ru>, Steve Henson]
1856 *) Add option -stream to use PKCS#7 streaming in smime utility. New
1857 function i2d_PKCS7_bio_stream() and PEM_write_PKCS7_bio_stream()
1858 to output in BER and PEM format.
1861 *) Experimental support for use of HMAC via EVP_PKEY interface. This
1862 allows HMAC to be handled via the EVP_DigestSign*() interface. The
1863 EVP_PKEY "key" in this case is the HMAC key, potentially allowing
1864 ENGINE support for HMAC keys which are unextractable. New -mac and
1865 -macopt options to dgst utility.
1868 *) New option -sigopt to dgst utility. Update dgst to use
1869 EVP_Digest{Sign,Verify}*. These two changes make it possible to use
1870 alternative signing paramaters such as X9.31 or PSS in the dgst
1874 *) Change ssl_cipher_apply_rule(), the internal function that does
1875 the work each time a ciphersuite string requests enabling
1876 ("foo+bar"), moving ("+foo+bar"), disabling ("-foo+bar", or
1877 removing ("!foo+bar") a class of ciphersuites: Now it maintains
1878 the order of disabled ciphersuites such that those ciphersuites
1879 that most recently went from enabled to disabled not only stay
1880 in order with respect to each other, but also have higher priority
1881 than other disabled ciphersuites the next time ciphersuites are
1884 This means that you can now say, e.g., "PSK:-PSK:HIGH" to enable
1885 the same ciphersuites as with "HIGH" alone, but in a specific
1886 order where the PSK ciphersuites come first (since they are the
1887 most recently disabled ciphersuites when "HIGH" is parsed).
1889 Also, change ssl_create_cipher_list() (using this new
1890 funcionality) such that between otherwise identical
1891 cihpersuites, ephemeral ECDH is preferred over ephemeral DH in
1895 *) Change ssl_create_cipher_list() so that it automatically
1896 arranges the ciphersuites in reasonable order before starting
1897 to process the rule string. Thus, the definition for "DEFAULT"
1898 (SSL_DEFAULT_CIPHER_LIST) now is just "ALL:!aNULL:!eNULL", but
1899 remains equivalent to "AES:ALL:!aNULL:!eNULL:+aECDH:+kRSA:+RC4:@STRENGTH".
1900 This makes it much easier to arrive at a reasonable default order
1901 in applications for which anonymous ciphers are OK (meaning
1902 that you can't actually use DEFAULT).
1903 [Bodo Moeller; suggested by Victor Duchovni]
1905 *) Split the SSL/TLS algorithm mask (as used for ciphersuite string
1906 processing) into multiple integers instead of setting
1907 "SSL_MKEY_MASK" bits, "SSL_AUTH_MASK" bits, "SSL_ENC_MASK",
1908 "SSL_MAC_MASK", and "SSL_SSL_MASK" bits all in a single integer.
1909 (These masks as well as the individual bit definitions are hidden
1910 away into the non-exported interface ssl/ssl_locl.h, so this
1911 change to the definition of the SSL_CIPHER structure shouldn't
1912 affect applications.) This give us more bits for each of these
1913 categories, so there is no longer a need to coagulate AES128 and
1914 AES256 into a single algorithm bit, and to coagulate Camellia128
1915 and Camellia256 into a single algorithm bit, which has led to all
1918 Thus, among other things, the kludge introduced in 0.9.7m and
1919 0.9.8e for masking out AES256 independently of AES128 or masking
1920 out Camellia256 independently of AES256 is not needed here in 0.9.9.
1922 With the change, we also introduce new ciphersuite aliases that
1923 so far were missing: "AES128", "AES256", "CAMELLIA128", and
1927 *) Add support for dsa-with-SHA224 and dsa-with-SHA256.
1928 Use the leftmost N bytes of the signature input if the input is
1929 larger than the prime q (with N being the size in bytes of q).
1932 *) Very *very* experimental PKCS#7 streaming encoder support. Nothing uses
1933 it yet and it is largely untested.
1936 *) Add support for the ecdsa-with-SHA224/256/384/512 signature types.
1939 *) Initial incomplete changes to avoid need for function casts in OpenSSL
1940 some compilers (gcc 4.2 and later) reject their use. Safestack is
1941 reimplemented. Update ASN1 to avoid use of legacy functions.
1944 *) Win32/64 targets are linked with Winsock2.
1947 *) Add an X509_CRL_METHOD structure to allow CRL processing to be redirected
1948 to external functions. This can be used to increase CRL handling
1949 efficiency especially when CRLs are very large by (for example) storing
1950 the CRL revoked certificates in a database.
1953 *) Overhaul of by_dir code. Add support for dynamic loading of CRLs so
1954 new CRLs added to a directory can be used. New command line option
1955 -verify_return_error to s_client and s_server. This causes real errors
1956 to be returned by the verify callback instead of carrying on no matter
1957 what. This reflects the way a "real world" verify callback would behave.
1960 *) GOST engine, supporting several GOST algorithms and public key formats.
1961 Kindly donated by Cryptocom.
1964 *) Partial support for Issuing Distribution Point CRL extension. CRLs
1965 partitioned by DP are handled but no indirect CRL or reason partitioning
1966 (yet). Complete overhaul of CRL handling: now the most suitable CRL is
1967 selected via a scoring technique which handles IDP and AKID in CRLs.
1970 *) New X509_STORE_CTX callbacks lookup_crls() and lookup_certs() which
1971 will ultimately be used for all verify operations: this will remove the
1972 X509_STORE dependency on certificate verification and allow alternative
1973 lookup methods. X509_STORE based implementations of these two callbacks.
1976 *) Allow multiple CRLs to exist in an X509_STORE with matching issuer names.
1977 Modify get_crl() to find a valid (unexpired) CRL if possible.
1980 *) New function X509_CRL_match() to check if two CRLs are identical. Normally
1981 this would be called X509_CRL_cmp() but that name is already used by
1982 a function that just compares CRL issuer names. Cache several CRL
1983 extensions in X509_CRL structure and cache CRLDP in X509.
1986 *) Store a "canonical" representation of X509_NAME structure (ASN1 Name)
1987 this maps equivalent X509_NAME structures into a consistent structure.
1988 Name comparison can then be performed rapidly using memcmp().
1991 *) Non-blocking OCSP request processing. Add -timeout option to ocsp
1995 *) Allow digests to supply their own micalg string for S/MIME type using
1996 the ctrl EVP_MD_CTRL_MICALG.
1999 *) During PKCS7 signing pass the PKCS7 SignerInfo structure to the
2000 EVP_PKEY_METHOD before and after signing via the EVP_PKEY_CTRL_PKCS7_SIGN
2001 ctrl. It can then customise the structure before and/or after signing
2005 *) New function OBJ_add_sigid() to allow application defined signature OIDs
2006 to be added to OpenSSLs internal tables. New function OBJ_sigid_free()
2007 to free up any added signature OIDs.
2010 *) New functions EVP_CIPHER_do_all(), EVP_CIPHER_do_all_sorted(),
2011 EVP_MD_do_all() and EVP_MD_do_all_sorted() to enumerate internal
2012 digest and cipher tables. New options added to openssl utility:
2013 list-message-digest-algorithms and list-cipher-algorithms.
2016 *) Change the array representation of binary polynomials: the list
2017 of degrees of non-zero coefficients is now terminated with -1.
2018 Previously it was terminated with 0, which was also part of the
2019 value; thus, the array representation was not applicable to
2020 polynomials where t^0 has coefficient zero. This change makes
2021 the array representation useful in a more general context.
2024 *) Various modifications and fixes to SSL/TLS cipher string
2025 handling. For ECC, the code now distinguishes between fixed ECDH
2026 with RSA certificates on the one hand and with ECDSA certificates
2027 on the other hand, since these are separate ciphersuites. The
2028 unused code for Fortezza ciphersuites has been removed.
2030 For consistency with EDH, ephemeral ECDH is now called "EECDH"
2031 (not "ECDHE"). For consistency with the code for DH
2032 certificates, use of ECDH certificates is now considered ECDH
2033 authentication, not RSA or ECDSA authentication (the latter is
2034 merely the CA's signing algorithm and not actively used in the
2037 The temporary ciphersuite alias "ECCdraft" is no longer
2038 available, and ECC ciphersuites are no longer excluded from "ALL"
2039 and "DEFAULT". The following aliases now exist for RFC 4492
2040 ciphersuites, most of these by analogy with the DH case:
2042 kECDHr - ECDH cert, signed with RSA
2043 kECDHe - ECDH cert, signed with ECDSA
2044 kECDH - ECDH cert (signed with either RSA or ECDSA)
2045 kEECDH - ephemeral ECDH
2046 ECDH - ECDH cert or ephemeral ECDH
2052 AECDH - anonymous ECDH
2053 EECDH - non-anonymous ephemeral ECDH (equivalent to "kEECDH:-AECDH")
2057 *) Add additional S/MIME capabilities for AES and GOST ciphers if supported.
2058 Use correct micalg parameters depending on digest(s) in signed message.
2061 *) Add engine support for EVP_PKEY_ASN1_METHOD. Add functions to process
2062 an ENGINE asn1 method. Support ENGINE lookups in the ASN1 code.
2065 *) Initial engine support for EVP_PKEY_METHOD. New functions to permit
2066 an engine to register a method. Add ENGINE lookups for methods and
2067 functional reference processing.
2070 *) New functions EVP_Digest{Sign,Verify)*. These are enchance versions of
2071 EVP_{Sign,Verify}* which allow an application to customise the signature
2075 *) New -resign option to smime utility. This adds one or more signers
2076 to an existing PKCS#7 signedData structure. Also -md option to use an
2077 alternative message digest algorithm for signing.
2080 *) Tidy up PKCS#7 routines and add new functions to make it easier to
2081 create PKCS7 structures containing multiple signers. Update smime
2082 application to support multiple signers.
2085 *) New -macalg option to pkcs12 utility to allow setting of an alternative
2089 *) Initial support for PKCS#5 v2.0 PRFs other than default SHA1 HMAC.
2090 Reorganize PBE internals to lookup from a static table using NIDs,
2091 add support for HMAC PBE OID translation. Add a EVP_CIPHER ctrl:
2092 EVP_CTRL_PBE_PRF_NID this allows a cipher to specify an alternative
2093 PRF which will be automatically used with PBES2.
2096 *) Replace the algorithm specific calls to generate keys in "req" with the
2100 *) Update PKCS#7 enveloped data routines to use new API. This is now
2101 supported by any public key method supporting the encrypt operation. A
2102 ctrl is added to allow the public key algorithm to examine or modify
2103 the PKCS#7 RecipientInfo structure if it needs to: for RSA this is
2107 *) Add a ctrl to asn1 method to allow a public key algorithm to express
2108 a default digest type to use. In most cases this will be SHA1 but some
2109 algorithms (such as GOST) need to specify an alternative digest. The
2110 return value indicates how strong the preference is 1 means optional and
2111 2 is mandatory (that is it is the only supported type). Modify
2112 ASN1_item_sign() to accept a NULL digest argument to indicate it should
2113 use the default md. Update openssl utilities to use the default digest
2114 type for signing if it is not explicitly indicated.
2117 *) Use OID cross reference table in ASN1_sign() and ASN1_verify(). New
2118 EVP_MD flag EVP_MD_FLAG_PKEY_METHOD_SIGNATURE. This uses the relevant
2119 signing method from the key type. This effectively removes the link
2120 between digests and public key types.
2123 *) Add an OID cross reference table and utility functions. Its purpose is to
2124 translate between signature OIDs such as SHA1WithrsaEncryption and SHA1,
2125 rsaEncryption. This will allow some of the algorithm specific hackery
2126 needed to use the correct OID to be removed.
2129 *) Remove algorithm specific dependencies when setting PKCS7_SIGNER_INFO
2130 structures for PKCS7_sign(). They are now set up by the relevant public
2134 *) Add provisional EC pkey method with support for ECDSA and ECDH.
2137 *) Add support for key derivation (agreement) in the API, DH method and
2141 *) Add DSA pkey method and DH pkey methods, extend DH ASN1 method to support
2142 public and private key formats. As a side effect these add additional
2143 command line functionality not previously available: DSA signatures can be
2144 generated and verified using pkeyutl and DH key support and generation in
2149 [Oliver Tappe <zooey@hirschkaefer.de>]
2151 *) New make target "install_html_docs" installs HTML renditions of the
2153 [Oliver Tappe <zooey@hirschkaefer.de>]
2155 *) New utility "genpkey" this is analogous to "genrsa" etc except it can
2156 generate keys for any algorithm. Extend and update EVP_PKEY_METHOD to
2157 support key and parameter generation and add initial key generation
2158 functionality for RSA.
2161 *) Add functions for main EVP_PKEY_method operations. The undocumented
2162 functions EVP_PKEY_{encrypt,decrypt} have been renamed to
2163 EVP_PKEY_{encrypt,decrypt}_old.
2166 *) Initial definitions for EVP_PKEY_METHOD. This will be a high level public
2167 key API, doesn't do much yet.
2170 *) New function EVP_PKEY_asn1_get0_info() to retrieve information about
2171 public key algorithms. New option to openssl utility:
2172 "list-public-key-algorithms" to print out info.
2175 *) Implement the Supported Elliptic Curves Extension for
2176 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
2179 *) Don't free up OIDs in OBJ_cleanup() if they are in use by EVP_MD or
2180 EVP_CIPHER structures to avoid later problems in EVP_cleanup().
2183 *) New utilities pkey and pkeyparam. These are similar to algorithm specific
2184 utilities such as rsa, dsa, dsaparam etc except they process any key
2188 *) Transfer public key printing routines to EVP_PKEY_ASN1_METHOD. New
2189 functions EVP_PKEY_print_public(), EVP_PKEY_print_private(),
2190 EVP_PKEY_print_param() to print public key data from an EVP_PKEY
2194 *) Initial support for pluggable public key ASN1.
2195 De-spaghettify the public key ASN1 handling. Move public and private
2196 key ASN1 handling to a new EVP_PKEY_ASN1_METHOD structure. Relocate
2197 algorithm specific handling to a single module within the relevant
2198 algorithm directory. Add functions to allow (near) opaque processing
2199 of public and private key structures.
2202 *) Implement the Supported Point Formats Extension for
2203 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
2206 *) Add initial support for RFC 4279 PSK TLS ciphersuites. Add members
2207 for the psk identity [hint] and the psk callback functions to the
2208 SSL_SESSION, SSL and SSL_CTX structure.
2211 PSK-RC4-SHA, PSK-3DES-EDE-CBC-SHA, PSK-AES128-CBC-SHA,
2215 SSL_CTX_use_psk_identity_hint
2216 SSL_get_psk_identity_hint
2217 SSL_get_psk_identity
2218 SSL_use_psk_identity_hint
2220 [Mika Kousa and Pasi Eronen of Nokia Corporation]
2222 *) Add RFC 3161 compliant time stamp request creation, response generation
2223 and response verification functionality.
2224 [Zoltán Glózik <zglozik@opentsa.org>, The OpenTSA Project]
2226 *) Add initial support for TLS extensions, specifically for the server_name
2227 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
2228 have new members for a host name. The SSL data structure has an
2229 additional member SSL_CTX *initial_ctx so that new sessions can be
2230 stored in that context to allow for session resumption, even after the
2231 SSL has been switched to a new SSL_CTX in reaction to a client's
2232 server_name extension.
2234 New functions (subject to change):
2236 SSL_get_servername()
2237 SSL_get_servername_type()
2240 New CTRL codes and macros (subject to change):
2242 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
2243 - SSL_CTX_set_tlsext_servername_callback()
2244 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
2245 - SSL_CTX_set_tlsext_servername_arg()
2246 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
2248 openssl s_client has a new '-servername ...' option.
2250 openssl s_server has new options '-servername_host ...', '-cert2 ...',
2251 '-key2 ...', '-servername_fatal' (subject to change). This allows
2252 testing the HostName extension for a specific single host name ('-cert'
2253 and '-key' remain fallbacks for handshakes without HostName
2254 negotiation). If the unrecognized_name alert has to be sent, this by
2255 default is a warning; it becomes fatal with the '-servername_fatal'
2258 [Peter Sylvester, Remy Allais, Christophe Renou]
2260 *) Whirlpool hash implementation is added.
2263 *) BIGNUM code on 64-bit SPARCv9 targets is switched from bn(64,64) to
2264 bn(64,32). Because of instruction set limitations it doesn't have
2265 any negative impact on performance. This was done mostly in order
2266 to make it possible to share assembler modules, such as bn_mul_mont
2267 implementations, between 32- and 64-bit builds without hassle.
2270 *) Move code previously exiled into file crypto/ec/ec2_smpt.c
2271 to ec2_smpl.c, and no longer require the OPENSSL_EC_BIN_PT_COMP
2275 *) New candidate for BIGNUM assembler implementation, bn_mul_mont,
2276 dedicated Montgomery multiplication procedure, is introduced.
2277 BN_MONT_CTX is modified to allow bn_mul_mont to reach for higher
2278 "64-bit" performance on certain 32-bit targets.
2281 *) New option SSL_OP_NO_COMP to disable use of compression selectively
2282 in SSL structures. New SSL ctrl to set maximum send fragment size.
2283 Save memory by seeting the I/O buffer sizes dynamically instead of
2284 using the maximum available value.
2287 *) New option -V for 'openssl ciphers'. This prints the ciphersuite code
2288 in addition to the text details.
2291 *) Very, very preliminary EXPERIMENTAL support for printing of general
2292 ASN1 structures. This currently produces rather ugly output and doesn't
2293 handle several customised structures at all.
2296 *) Integrated support for PVK file format and some related formats such
2297 as MS PUBLICKEYBLOB and PRIVATEKEYBLOB. Command line switches to support
2298 these in the 'rsa' and 'dsa' utilities.
2301 *) Support for PKCS#1 RSAPublicKey format on rsa utility command line.
2304 *) Remove the ancient ASN1_METHOD code. This was only ever used in one
2305 place for the (very old) "NETSCAPE" format certificates which are now
2306 handled using new ASN1 code equivalents.
2309 *) Let the TLSv1_method() etc. functions return a 'const' SSL_METHOD
2310 pointer and make the SSL_METHOD parameter in SSL_CTX_new,
2311 SSL_CTX_set_ssl_version and SSL_set_ssl_method 'const'.
2314 *) Modify CRL distribution points extension code to print out previously
2315 unsupported fields. Enhance extension setting code to allow setting of
2319 *) Add print and set support for Issuing Distribution Point CRL extension.
2322 *) Change 'Configure' script to enable Camellia by default.
2325 Changes between 0.9.8x and 0.9.8y [5 Feb 2013]
2327 *) Make the decoding of SSLv3, TLS and DTLS CBC records constant time.
2329 This addresses the flaw in CBC record processing discovered by
2330 Nadhem Alfardan and Kenny Paterson. Details of this attack can be found
2331 at: http://www.isg.rhul.ac.uk/tls/
2333 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
2334 Security Group at Royal Holloway, University of London
2335 (www.isg.rhul.ac.uk) for discovering this flaw and Adam Langley and
2336 Emilia Käsper for the initial patch.
2338 [Emilia Käsper, Adam Langley, Ben Laurie, Andy Polyakov, Steve Henson]
2340 *) Return an error when checking OCSP signatures when key is NULL.
2341 This fixes a DoS attack. (CVE-2013-0166)
2344 *) Call OCSP Stapling callback after ciphersuite has been chosen, so
2345 the right response is stapled. Also change SSL_get_certificate()
2346 so it returns the certificate actually sent.
2347 See http://rt.openssl.org/Ticket/Display.html?id=2836.
2348 (This is a backport)
2349 [Rob Stradling <rob.stradling@comodo.com>]
2351 *) Fix possible deadlock when decoding public keys.
2354 Changes between 0.9.8w and 0.9.8x [10 May 2012]
2356 *) Sanity check record length before skipping explicit IV in DTLS
2359 Thanks to Codenomicon for discovering this issue using Fuzz-o-Matic
2360 fuzzing as a service testing platform.
2364 *) Initialise tkeylen properly when encrypting CMS messages.
2365 Thanks to Solar Designer of Openwall for reporting this issue.
2368 Changes between 0.9.8v and 0.9.8w [23 Apr 2012]
2370 *) The fix for CVE-2012-2110 did not take into account that the
2371 'len' argument to BUF_MEM_grow and BUF_MEM_grow_clean is an
2372 int in OpenSSL 0.9.8, making it still vulnerable. Fix by
2373 rejecting negative len parameter. (CVE-2012-2131)
2374 [Tomas Hoger <thoger@redhat.com>]
2376 Changes between 0.9.8u and 0.9.8v [19 Apr 2012]
2378 *) Check for potentially exploitable overflows in asn1_d2i_read_bio
2379 BUF_mem_grow and BUF_mem_grow_clean. Refuse attempts to shrink buffer
2380 in CRYPTO_realloc_clean.
2382 Thanks to Tavis Ormandy, Google Security Team, for discovering this
2383 issue and to Adam Langley <agl@chromium.org> for fixing it.
2385 [Adam Langley (Google), Tavis Ormandy, Google Security Team]
2387 Changes between 0.9.8t and 0.9.8u [12 Mar 2012]
2389 *) Fix MMA (Bleichenbacher's attack on PKCS #1 v1.5 RSA padding) weakness
2390 in CMS and PKCS7 code. When RSA decryption fails use a random key for
2391 content decryption and always return the same error. Note: this attack
2392 needs on average 2^20 messages so it only affects automated senders. The
2393 old behaviour can be reenabled in the CMS code by setting the
2394 CMS_DEBUG_DECRYPT flag: this is useful for debugging and testing where
2395 an MMA defence is not necessary.
2396 Thanks to Ivan Nestlerode <inestlerode@us.ibm.com> for discovering
2397 this issue. (CVE-2012-0884)
2400 *) Fix CVE-2011-4619: make sure we really are receiving a
2401 client hello before rejecting multiple SGC restarts. Thanks to
2402 Ivan Nestlerode <inestlerode@us.ibm.com> for discovering this bug.
2405 Changes between 0.9.8s and 0.9.8t [18 Jan 2012]
2407 *) Fix for DTLS DoS issue introduced by fix for CVE-2011-4109.
2408 Thanks to Antonio Martin, Enterprise Secure Access Research and
2409 Development, Cisco Systems, Inc. for discovering this bug and
2410 preparing a fix. (CVE-2012-0050)
2413 Changes between 0.9.8r and 0.9.8s [4 Jan 2012]
2415 *) Nadhem Alfardan and Kenny Paterson have discovered an extension
2416 of the Vaudenay padding oracle attack on CBC mode encryption
2417 which enables an efficient plaintext recovery attack against
2418 the OpenSSL implementation of DTLS. Their attack exploits timing
2419 differences arising during decryption processing. A research
2420 paper describing this attack can be found at:
2421 http://www.isg.rhul.ac.uk/~kp/dtls.pdf
2422 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
2423 Security Group at Royal Holloway, University of London
2424 (www.isg.rhul.ac.uk) for discovering this flaw and to Robin Seggelmann
2425 <seggelmann@fh-muenster.de> and Michael Tuexen <tuexen@fh-muenster.de>
2426 for preparing the fix. (CVE-2011-4108)
2427 [Robin Seggelmann, Michael Tuexen]
2429 *) Stop policy check failure freeing same buffer twice. (CVE-2011-4109)
2430 [Ben Laurie, Kasper <ekasper@google.com>]
2432 *) Clear bytes used for block padding of SSL 3.0 records.
2434 [Adam Langley (Google)]
2436 *) Only allow one SGC handshake restart for SSL/TLS. Thanks to George
2437 Kadianakis <desnacked@gmail.com> for discovering this issue and
2438 Adam Langley for preparing the fix. (CVE-2011-4619)
2439 [Adam Langley (Google)]
2441 *) Prevent malformed RFC3779 data triggering an assertion failure.
2442 Thanks to Andrew Chi, BBN Technologies, for discovering the flaw
2443 and Rob Austein <sra@hactrn.net> for fixing it. (CVE-2011-4577)
2444 [Rob Austein <sra@hactrn.net>]
2446 *) Fix ssl_ciph.c set-up race.
2447 [Adam Langley (Google)]
2449 *) Fix spurious failures in ecdsatest.c.
2450 [Emilia Käsper (Google)]
2452 *) Fix the BIO_f_buffer() implementation (which was mixing different
2453 interpretations of the '..._len' fields).
2454 [Adam Langley (Google)]
2456 *) Fix handling of BN_BLINDING: now BN_BLINDING_invert_ex (rather than
2457 BN_BLINDING_invert_ex) calls BN_BLINDING_update, ensuring that concurrent
2458 threads won't reuse the same blinding coefficients.
2460 This also avoids the need to obtain the CRYPTO_LOCK_RSA_BLINDING
2461 lock to call BN_BLINDING_invert_ex, and avoids one use of
2462 BN_BLINDING_update for each BN_BLINDING structure (previously,
2463 the last update always remained unused).
2464 [Emilia Käsper (Google)]
2466 *) Fix SSL memory handling for (EC)DH ciphersuites, in particular
2467 for multi-threaded use of ECDH.
2468 [Adam Langley (Google)]
2470 *) Fix x509_name_ex_d2i memory leak on bad inputs.
2473 *) Add protection against ECDSA timing attacks as mentioned in the paper
2474 by Billy Bob Brumley and Nicola Tuveri, see:
2476 http://eprint.iacr.org/2011/232.pdf
2478 [Billy Bob Brumley and Nicola Tuveri]
2480 Changes between 0.9.8q and 0.9.8r [8 Feb 2011]
2482 *) Fix parsing of OCSP stapling ClientHello extension. CVE-2011-0014
2483 [Neel Mehta, Adam Langley, Bodo Moeller (Google)]
2485 *) Fix bug in string printing code: if *any* escaping is enabled we must
2486 escape the escape character (backslash) or the resulting string is
2490 Changes between 0.9.8p and 0.9.8q [2 Dec 2010]
2492 *) Disable code workaround for ancient and obsolete Netscape browsers
2493 and servers: an attacker can use it in a ciphersuite downgrade attack.
2494 Thanks to Martin Rex for discovering this bug. CVE-2010-4180
2497 *) Fixed J-PAKE implementation error, originally discovered by
2498 Sebastien Martini, further info and confirmation from Stefan
2499 Arentz and Feng Hao. Note that this fix is a security fix. CVE-2010-4252
2502 Changes between 0.9.8o and 0.9.8p [16 Nov 2010]
2504 *) Fix extension code to avoid race conditions which can result in a buffer
2505 overrun vulnerability: resumed sessions must not be modified as they can
2506 be shared by multiple threads. CVE-2010-3864
2509 *) Fix for double free bug in ssl/s3_clnt.c CVE-2010-2939
2512 *) Don't reencode certificate when calculating signature: cache and use
2513 the original encoding instead. This makes signature verification of
2514 some broken encodings work correctly.
2517 *) ec2_GF2m_simple_mul bugfix: compute correct result if the output EC_POINT
2518 is also one of the inputs.
2519 [Emilia Käsper <emilia.kasper@esat.kuleuven.be> (Google)]
2521 *) Don't repeatedly append PBE algorithms to table if they already exist.
2522 Sort table on each new add. This effectively makes the table read only
2523 after all algorithms are added and subsequent calls to PKCS12_pbe_add
2527 Changes between 0.9.8n and 0.9.8o [01 Jun 2010]
2529 [NB: OpenSSL 0.9.8o and later 0.9.8 patch levels were released after
2532 *) Correct a typo in the CMS ASN1 module which can result in invalid memory
2533 access or freeing data twice (CVE-2010-0742)
2534 [Steve Henson, Ronald Moesbergen <intercommit@gmail.com>]
2536 *) Add SHA2 algorithms to SSL_library_init(). SHA2 is becoming far more
2537 common in certificates and some applications which only call
2538 SSL_library_init and not OpenSSL_add_all_algorithms() will fail.
2542 Reduce copying into .apps and .test in makevms.com
2543 Don't try to use blank CA certificate in CA.com
2544 Allow use of C files from original directories in maketests.com
2545 [Steven M. Schweda" <sms@antinode.info>]
2547 Changes between 0.9.8m and 0.9.8n [24 Mar 2010]
2549 *) When rejecting SSL/TLS records due to an incorrect version number, never
2550 update s->server with a new major version number. As of
2551 - OpenSSL 0.9.8m if 'short' is a 16-bit type,
2552 - OpenSSL 0.9.8f if 'short' is longer than 16 bits,
2553 the previous behavior could result in a read attempt at NULL when
2554 receiving specific incorrect SSL/TLS records once record payload
2555 protection is active. (CVE-2010-0740)
2556 [Bodo Moeller, Adam Langley <agl@chromium.org>]
2558 *) Fix for CVE-2010-0433 where some kerberos enabled versions of OpenSSL
2559 could be crashed if the relevant tables were not present (e.g. chrooted).
2560 [Tomas Hoger <thoger@redhat.com>]
2562 Changes between 0.9.8l and 0.9.8m [25 Feb 2010]
2564 *) Always check bn_wexpend() return values for failure. (CVE-2009-3245)
2565 [Martin Olsson, Neel Mehta]
2567 *) Fix X509_STORE locking: Every 'objs' access requires a lock (to
2568 accommodate for stack sorting, always a write lock!).
2571 *) On some versions of WIN32 Heap32Next is very slow. This can cause
2572 excessive delays in the RAND_poll(): over a minute. As a workaround
2573 include a time check in the inner Heap32Next loop too.
2576 *) The code that handled flushing of data in SSL/TLS originally used the
2577 BIO_CTRL_INFO ctrl to see if any data was pending first. This caused
2578 the problem outlined in PR#1949. The fix suggested there however can
2579 trigger problems with buggy BIO_CTRL_WPENDING (e.g. some versions
2580 of Apache). So instead simplify the code to flush unconditionally.
2581 This should be fine since flushing with no data to flush is a no op.
2584 *) Handle TLS versions 2.0 and later properly and correctly use the
2585 highest version of TLS/SSL supported. Although TLS >= 2.0 is some way
2586 off ancient servers have a habit of sticking around for a while...
2589 *) Modify compression code so it frees up structures without using the
2590 ex_data callbacks. This works around a problem where some applications
2591 call CRYPTO_cleanup_all_ex_data() before application exit (e.g. when
2592 restarting) then use compression (e.g. SSL with compression) later.
2593 This results in significant per-connection memory leaks and
2594 has caused some security issues including CVE-2008-1678 and
2598 *) Constify crypto/cast (i.e., <openssl/cast.h>): a CAST_KEY doesn't
2599 change when encrypting or decrypting.
2602 *) Add option SSL_OP_LEGACY_SERVER_CONNECT which will allow clients to
2603 connect and renegotiate with servers which do not support RI.
2604 Until RI is more widely deployed this option is enabled by default.
2607 *) Add "missing" ssl ctrls to clear options and mode.
2610 *) If client attempts to renegotiate and doesn't support RI respond with
2611 a no_renegotiation alert as required by RFC5746. Some renegotiating
2612 TLS clients will continue a connection gracefully when they receive
2613 the alert. Unfortunately OpenSSL mishandled this alert and would hang
2614 waiting for a server hello which it will never receive. Now we treat a
2615 received no_renegotiation alert as a fatal error. This is because
2616 applications requesting a renegotiation might well expect it to succeed
2617 and would have no code in place to handle the server denying it so the
2618 only safe thing to do is to terminate the connection.
2621 *) Add ctrl macro SSL_get_secure_renegotiation_support() which returns 1 if
2622 peer supports secure renegotiation and 0 otherwise. Print out peer
2623 renegotiation support in s_client/s_server.
2626 *) Replace the highly broken and deprecated SPKAC certification method with
2627 the updated NID creation version. This should correctly handle UTF8.
2630 *) Implement RFC5746. Re-enable renegotiation but require the extension
2631 as needed. Unfortunately, SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION
2632 turns out to be a bad idea. It has been replaced by
2633 SSL_OP_ALLOW_UNSAFE_LEGACY_RENEGOTIATION which can be set with
2634 SSL_CTX_set_options(). This is really not recommended unless you
2635 know what you are doing.
2636 [Eric Rescorla <ekr@networkresonance.com>, Ben Laurie, Steve Henson]
2638 *) Fixes to stateless session resumption handling. Use initial_ctx when
2639 issuing and attempting to decrypt tickets in case it has changed during
2640 servername handling. Use a non-zero length session ID when attempting
2641 stateless session resumption: this makes it possible to determine if
2642 a resumption has occurred immediately after receiving server hello
2643 (several places in OpenSSL subtly assume this) instead of later in
2647 *) The functions ENGINE_ctrl(), OPENSSL_isservice(),
2648 CMS_get1_RecipientRequest() and RAND_bytes() can return <=0 on error
2649 fixes for a few places where the return code is not checked
2651 [Julia Lawall <julia@diku.dk>]
2653 *) Add --strict-warnings option to Configure script to include devteam
2654 warnings in other configurations.
2657 *) Add support for --libdir option and LIBDIR variable in makefiles. This
2658 makes it possible to install openssl libraries in locations which
2659 have names other than "lib", for example "/usr/lib64" which some
2661 [Steve Henson, based on patch from Jeremy Utley]
2663 *) Don't allow the use of leading 0x80 in OIDs. This is a violation of
2664 X690 8.9.12 and can produce some misleading textual output of OIDs.
2665 [Steve Henson, reported by Dan Kaminsky]
2667 *) Delete MD2 from algorithm tables. This follows the recommendation in
2668 several standards that it is not used in new applications due to
2669 several cryptographic weaknesses. For binary compatibility reasons
2670 the MD2 API is still compiled in by default.
2673 *) Add compression id to {d2i,i2d}_SSL_SESSION so it is correctly saved
2677 *) Rename uni2asc and asc2uni functions to OPENSSL_uni2asc and
2678 OPENSSL_asc2uni conditionally on Netware platforms to avoid a name
2680 [Guenter <lists@gknw.net>]
2682 *) Fix the server certificate chain building code to use X509_verify_cert(),
2683 it used to have an ad-hoc builder which was unable to cope with anything
2684 other than a simple chain.
2685 [David Woodhouse <dwmw2@infradead.org>, Steve Henson]
2687 *) Don't check self signed certificate signatures in X509_verify_cert()
2688 by default (a flag can override this): it just wastes time without
2689 adding any security. As a useful side effect self signed root CAs
2690 with non-FIPS digests are now usable in FIPS mode.
2693 *) In dtls1_process_out_of_seq_message() the check if the current message
2694 is already buffered was missing. For every new message was memory
2695 allocated, allowing an attacker to perform an denial of service attack
2696 with sending out of seq handshake messages until there is no memory
2697 left. Additionally every future messege was buffered, even if the
2698 sequence number made no sense and would be part of another handshake.
2699 So only messages with sequence numbers less than 10 in advance will be
2700 buffered. (CVE-2009-1378)
2701 [Robin Seggelmann, discovered by Daniel Mentz]
2703 *) Records are buffered if they arrive with a future epoch to be
2704 processed after finishing the corresponding handshake. There is
2705 currently no limitation to this buffer allowing an attacker to perform
2706 a DOS attack with sending records with future epochs until there is no
2707 memory left. This patch adds the pqueue_size() function to determine
2708 the size of a buffer and limits the record buffer to 100 entries.
2710 [Robin Seggelmann, discovered by Daniel Mentz]
2712 *) Keep a copy of frag->msg_header.frag_len so it can be used after the
2713 parent structure is freed. (CVE-2009-1379)
2716 *) Handle non-blocking I/O properly in SSL_shutdown() call.
2717 [Darryl Miles <darryl-mailinglists@netbauds.net>]
2720 [Ilya O. <vrghost@gmail.com>]
2722 Changes between 0.9.8k and 0.9.8l [5 Nov 2009]
2724 *) Disable renegotiation completely - this fixes a severe security
2725 problem (CVE-2009-3555) at the cost of breaking all
2726 renegotiation. Renegotiation can be re-enabled by setting
2727 SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION in s3->flags at
2728 run-time. This is really not recommended unless you know what
2732 Changes between 0.9.8j and 0.9.8k [25 Mar 2009]
2734 *) Don't set val to NULL when freeing up structures, it is freed up by
2735 underlying code. If sizeof(void *) > sizeof(long) this can result in
2736 zeroing past the valid field. (CVE-2009-0789)
2737 [Paolo Ganci <Paolo.Ganci@AdNovum.CH>]
2739 *) Fix bug where return value of CMS_SignerInfo_verify_content() was not
2740 checked correctly. This would allow some invalid signed attributes to
2741 appear to verify correctly. (CVE-2009-0591)
2742 [Ivan Nestlerode <inestlerode@us.ibm.com>]
2744 *) Reject UniversalString and BMPString types with invalid lengths. This
2745 prevents a crash in ASN1_STRING_print_ex() which assumes the strings have
2746 a legal length. (CVE-2009-0590)
2749 *) Set S/MIME signing as the default purpose rather than setting it
2750 unconditionally. This allows applications to override it at the store
2754 *) Permit restricted recursion of ASN1 strings. This is needed in practice
2755 to handle some structures.
2758 *) Improve efficiency of mem_gets: don't search whole buffer each time
2760 [Jeremy Shapiro <jnshapir@us.ibm.com>]
2762 *) New -hex option for openssl rand.
2765 *) Print out UTF8String and NumericString when parsing ASN1.
2768 *) Support NumericString type for name components.
2771 *) Allow CC in the environment to override the automatically chosen
2772 compiler. Note that nothing is done to ensure flags work with the
2776 Changes between 0.9.8i and 0.9.8j [07 Jan 2009]
2778 *) Properly check EVP_VerifyFinal() and similar return values
2780 [Ben Laurie, Bodo Moeller, Google Security Team]
2782 *) Enable TLS extensions by default.
2785 *) Allow the CHIL engine to be loaded, whether the application is
2786 multithreaded or not. (This does not release the developer from the
2787 obligation to set up the dynamic locking callbacks.)
2788 [Sander Temme <sander@temme.net>]
2790 *) Use correct exit code if there is an error in dgst command.
2791 [Steve Henson; problem pointed out by Roland Dirlewanger]
2793 *) Tweak Configure so that you need to say "experimental-jpake" to enable
2794 JPAKE, and need to use -DOPENSSL_EXPERIMENTAL_JPAKE in applications.
2797 *) Add experimental JPAKE support, including demo authentication in
2798 s_client and s_server.
2801 *) Set the comparison function in v3_addr_canonize().
2802 [Rob Austein <sra@hactrn.net>]
2804 *) Add support for XMPP STARTTLS in s_client.
2805 [Philip Paeps <philip@freebsd.org>]
2807 *) Change the server-side SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG behavior
2808 to ensure that even with this option, only ciphersuites in the
2809 server's preference list will be accepted. (Note that the option
2810 applies only when resuming a session, so the earlier behavior was
2811 just about the algorithm choice for symmetric cryptography.)
2814 Changes between 0.9.8h and 0.9.8i [15 Sep 2008]
2816 *) Fix NULL pointer dereference if a DTLS server received
2817 ChangeCipherSpec as first record (CVE-2009-1386).
2820 *) Fix a state transition in s3_srvr.c and d1_srvr.c
2821 (was using SSL3_ST_CW_CLNT_HELLO_B, should be ..._ST_SW_SRVR_...).
2824 *) The fix in 0.9.8c that supposedly got rid of unsafe
2825 double-checked locking was incomplete for RSA blinding,
2826 addressing just one layer of what turns out to have been
2827 doubly unsafe triple-checked locking.
2829 So now fix this for real by retiring the MONT_HELPER macro
2830 in crypto/rsa/rsa_eay.c.
2832 [Bodo Moeller; problem pointed out by Marius Schilder]
2834 *) Various precautionary measures:
2836 - Avoid size_t integer overflow in HASH_UPDATE (md32_common.h).
2838 - Avoid a buffer overflow in d2i_SSL_SESSION() (ssl_asn1.c).
2839 (NB: This would require knowledge of the secret session ticket key
2840 to exploit, in which case you'd be SOL either way.)
2842 - Change bn_nist.c so that it will properly handle input BIGNUMs
2843 outside the expected range.
2845 - Enforce the 'num' check in BN_div() (bn_div.c) for non-BN_DEBUG
2848 [Neel Mehta, Bodo Moeller]
2850 *) Allow engines to be "soft loaded" - i.e. optionally don't die if
2851 the load fails. Useful for distros.
2852 [Ben Laurie and the FreeBSD team]
2854 *) Add support for Local Machine Keyset attribute in PKCS#12 files.
2857 *) Fix BN_GF2m_mod_arr() top-bit cleanup code.
2860 *) Expand ENGINE to support engine supplied SSL client certificate functions.
2862 This work was sponsored by Logica.
2865 *) Add CryptoAPI ENGINE to support use of RSA and DSA keys held in Windows
2866 keystores. Support for SSL/TLS client authentication too.
2867 Not compiled unless enable-capieng specified to Configure.
2869 This work was sponsored by Logica.
2872 *) Fix bug in X509_ATTRIBUTE creation: dont set attribute using
2873 ASN1_TYPE_set1 if MBSTRING flag set. This bug would crash certain
2874 attribute creation routines such as certifcate requests and PKCS#12
2878 Changes between 0.9.8g and 0.9.8h [28 May 2008]
2880 *) Fix flaw if 'Server Key exchange message' is omitted from a TLS
2881 handshake which could lead to a cilent crash as found using the
2882 Codenomicon TLS test suite (CVE-2008-1672)
2883 [Steve Henson, Mark Cox]
2885 *) Fix double free in TLS server name extensions which could lead to
2886 a remote crash found by Codenomicon TLS test suite (CVE-2008-0891)
2889 *) Clear error queue in SSL_CTX_use_certificate_chain_file()
2891 Clear the error queue to ensure that error entries left from
2892 older function calls do not interfere with the correct operation.
2893 [Lutz Jaenicke, Erik de Castro Lopo]
2895 *) Remove root CA certificates of commercial CAs:
2897 The OpenSSL project does not recommend any specific CA and does not
2898 have any policy with respect to including or excluding any CA.
2899 Therefore it does not make any sense to ship an arbitrary selection
2900 of root CA certificates with the OpenSSL software.
2903 *) RSA OAEP patches to fix two separate invalid memory reads.
2904 The first one involves inputs when 'lzero' is greater than
2905 'SHA_DIGEST_LENGTH' (it would read about SHA_DIGEST_LENGTH bytes
2906 before the beginning of from). The second one involves inputs where
2907 the 'db' section contains nothing but zeroes (there is a one-byte
2908 invalid read after the end of 'db').
2909 [Ivan Nestlerode <inestlerode@us.ibm.com>]
2911 *) Partial backport from 0.9.9-dev:
2913 Introduce bn_mul_mont (dedicated Montgomery multiplication
2914 procedure) as a candidate for BIGNUM assembler implementation.
2915 While 0.9.9-dev uses assembler for various architectures, only
2916 x86_64 is available by default here in the 0.9.8 branch, and
2917 32-bit x86 is available through a compile-time setting.
2919 To try the 32-bit x86 assembler implementation, use Configure
2920 option "enable-montasm" (which exists only for this backport).
2922 As "enable-montasm" for 32-bit x86 disclaims code stability
2923 anyway, in this constellation we activate additional code
2924 backported from 0.9.9-dev for further performance improvements,
2925 namely BN_from_montgomery_word. (To enable this otherwise,
2926 e.g. x86_64, try "-DMONT_FROM_WORD___NON_DEFAULT_0_9_8_BUILD".)
2928 [Andy Polyakov (backport partially by Bodo Moeller)]
2930 *) Add TLS session ticket callback. This allows an application to set
2931 TLS ticket cipher and HMAC keys rather than relying on hardcoded fixed
2932 values. This is useful for key rollover for example where several key
2933 sets may exist with different names.
2936 *) Reverse ENGINE-internal logic for caching default ENGINE handles.
2937 This was broken until now in 0.9.8 releases, such that the only way
2938 a registered ENGINE could be used (assuming it initialises
2939 successfully on the host) was to explicitly set it as the default
2940 for the relevant algorithms. This is in contradiction with 0.9.7
2941 behaviour and the documentation. With this fix, when an ENGINE is
2942 registered into a given algorithm's table of implementations, the
2943 'uptodate' flag is reset so that auto-discovery will be used next
2944 time a new context for that algorithm attempts to select an
2946 [Ian Lister (tweaked by Geoff Thorpe)]
2948 *) Backport of CMS code to OpenSSL 0.9.8. This differs from the 0.9.9
2949 implemention in the following ways:
2951 Lack of EVP_PKEY_ASN1_METHOD means algorithm parameters have to be
2954 Lack of BER streaming support means one pass streaming processing is
2955 only supported if data is detached: setting the streaming flag is
2956 ignored for embedded content.
2958 CMS support is disabled by default and must be explicitly enabled
2959 with the enable-cms configuration option.
2962 *) Update the GMP engine glue to do direct copies between BIGNUM and
2963 mpz_t when openssl and GMP use the same limb size. Otherwise the
2964 existing "conversion via a text string export" trick is still used.
2965 [Paul Sheer <paulsheer@gmail.com>]
2967 *) Zlib compression BIO. This is a filter BIO which compressed and
2968 uncompresses any data passed through it.
2971 *) Add AES_wrap_key() and AES_unwrap_key() functions to implement
2972 RFC3394 compatible AES key wrapping.
2975 *) Add utility functions to handle ASN1 structures. ASN1_STRING_set0():
2976 sets string data without copying. X509_ALGOR_set0() and
2977 X509_ALGOR_get0(): set and retrieve X509_ALGOR (AlgorithmIdentifier)
2978 data. Attribute function X509at_get0_data_by_OBJ(): retrieves data
2979 from an X509_ATTRIBUTE structure optionally checking it occurs only
2980 once. ASN1_TYPE_set1(): set and ASN1_TYPE structure copying supplied
2984 *) Fix BN flag handling in RSA_eay_mod_exp() and BN_MONT_CTX_set()
2985 to get the expected BN_FLG_CONSTTIME behavior.
2986 [Bodo Moeller (Google)]
2990 - fixed wrong usage of ioctlsocket() when build for LIBC BSD sockets
2991 - fixed do_tests.pl to run the test suite with CLIB builds too (CLIB_OPT)
2992 - added some more tests to do_tests.pl
2993 - fixed RunningProcess usage so that it works with newer LIBC NDKs too
2994 - removed usage of BN_LLONG for CLIB builds to avoid runtime dependency
2995 - added new Configure targets netware-clib-bsdsock, netware-clib-gcc,
2996 netware-clib-bsdsock-gcc, netware-libc-bsdsock-gcc
2997 - various changes to netware.pl to enable gcc-cross builds on Win32
2999 - changed crypto/bio/b_sock.c to work with macro functions (CLIB BSD)
3000 - various changes to fix missing prototype warnings
3001 - fixed x86nasm.pl to create correct asm files for NASM COFF output
3002 - added AES, WHIRLPOOL and CPUID assembler code to build files
3003 - added missing AES assembler make rules to mk1mf.pl
3004 - fixed order of includes in apps/ocsp.c so that e_os.h settings apply
3005 [Guenter Knauf <eflash@gmx.net>]
3007 *) Implement certificate status request TLS extension defined in RFC3546.
3008 A client can set the appropriate parameters and receive the encoded
3009 OCSP response via a callback. A server can query the supplied parameters
3010 and set the encoded OCSP response in the callback. Add simplified examples
3011 to s_client and s_server.
3014 Changes between 0.9.8f and 0.9.8g [19 Oct 2007]
3016 *) Fix various bugs:
3017 + Binary incompatibility of ssl_ctx_st structure
3018 + DTLS interoperation with non-compliant servers
3019 + Don't call get_session_cb() without proposed session
3020 + Fix ia64 assembler code
3021 [Andy Polyakov, Steve Henson]
3023 Changes between 0.9.8e and 0.9.8f [11 Oct 2007]
3025 *) DTLS Handshake overhaul. There were longstanding issues with
3026 OpenSSL DTLS implementation, which were making it impossible for
3027 RFC 4347 compliant client to communicate with OpenSSL server.
3028 Unfortunately just fixing these incompatibilities would "cut off"
3029 pre-0.9.8f clients. To allow for hassle free upgrade post-0.9.8e
3030 server keeps tolerating non RFC compliant syntax. The opposite is
3031 not true, 0.9.8f client can not communicate with earlier server.
3032 This update even addresses CVE-2007-4995.
3035 *) Changes to avoid need for function casts in OpenSSL: some compilers
3036 (gcc 4.2 and later) reject their use.
3037 [Kurt Roeckx <kurt@roeckx.be>, Peter Hartley <pdh@utter.chaos.org.uk>,
3040 *) Add RFC4507 support to OpenSSL. This includes the corrections in
3041 RFC4507bis. The encrypted ticket format is an encrypted encoded
3042 SSL_SESSION structure, that way new session features are automatically
3045 If a client application caches session in an SSL_SESSION structure
3046 support is transparent because tickets are now stored in the encoded
3049 The SSL_CTX structure automatically generates keys for ticket
3050 protection in servers so again support should be possible
3051 with no application modification.
3053 If a client or server wishes to disable RFC4507 support then the option
3054 SSL_OP_NO_TICKET can be set.
3056 Add a TLS extension debugging callback to allow the contents of any client
3057 or server extensions to be examined.
3059 This work was sponsored by Google.
3062 *) Add initial support for TLS extensions, specifically for the server_name
3063 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
3064 have new members for a host name. The SSL data structure has an
3065 additional member SSL_CTX *initial_ctx so that new sessions can be
3066 stored in that context to allow for session resumption, even after the
3067 SSL has been switched to a new SSL_CTX in reaction to a client's
3068 server_name extension.
3070 New functions (subject to change):
3072 SSL_get_servername()
3073 SSL_get_servername_type()
3076 New CTRL codes and macros (subject to change):
3078 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
3079 - SSL_CTX_set_tlsext_servername_callback()
3080 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
3081 - SSL_CTX_set_tlsext_servername_arg()
3082 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
3084 openssl s_client has a new '-servername ...' option.
3086 openssl s_server has new options '-servername_host ...', '-cert2 ...',
3087 '-key2 ...', '-servername_fatal' (subject to change). This allows
3088 testing the HostName extension for a specific single host name ('-cert'
3089 and '-key' remain fallbacks for handshakes without HostName
3090 negotiation). If the unrecognized_name alert has to be sent, this by
3091 default is a warning; it becomes fatal with the '-servername_fatal'
3094 [Peter Sylvester, Remy Allais, Christophe Renou, Steve Henson]
3096 *) Add AES and SSE2 assembly language support to VC++ build.
3099 *) Mitigate attack on final subtraction in Montgomery reduction.
3102 *) Fix crypto/ec/ec_mult.c to work properly with scalars of value 0
3103 (which previously caused an internal error).
3106 *) Squeeze another 10% out of IGE mode when in != out.
3109 *) AES IGE mode speedup.
3110 [Dean Gaudet (Google)]
3112 *) Add the Korean symmetric 128-bit cipher SEED (see
3113 http://www.kisa.or.kr/kisa/seed/jsp/seed_eng.jsp) and
3114 add SEED ciphersuites from RFC 4162:
3116 TLS_RSA_WITH_SEED_CBC_SHA = "SEED-SHA"
3117 TLS_DHE_DSS_WITH_SEED_CBC_SHA = "DHE-DSS-SEED-SHA"
3118 TLS_DHE_RSA_WITH_SEED_CBC_SHA = "DHE-RSA-SEED-SHA"
3119 TLS_DH_anon_WITH_SEED_CBC_SHA = "ADH-SEED-SHA"
3121 To minimize changes between patchlevels in the OpenSSL 0.9.8
3122 series, SEED remains excluded from compilation unless OpenSSL
3123 is configured with 'enable-seed'.
3124 [KISA, Bodo Moeller]
3126 *) Mitigate branch prediction attacks, which can be practical if a
3127 single processor is shared, allowing a spy process to extract
3128 information. For detailed background information, see
3129 http://eprint.iacr.org/2007/039 (O. Aciicmez, S. Gueron,
3130 J.-P. Seifert, "New Branch Prediction Vulnerabilities in OpenSSL
3131 and Necessary Software Countermeasures"). The core of the change
3132 are new versions BN_div_no_branch() and
3133 BN_mod_inverse_no_branch() of BN_div() and BN_mod_inverse(),
3134 respectively, which are slower, but avoid the security-relevant
3135 conditional branches. These are automatically called by BN_div()
3136 and BN_mod_inverse() if the flag BN_FLG_CONSTTIME is set for one
3137 of the input BIGNUMs. Also, BN_is_bit_set() has been changed to
3138 remove a conditional branch.
3140 BN_FLG_CONSTTIME is the new name for the previous
3141 BN_FLG_EXP_CONSTTIME flag, since it now affects more than just
3142 modular exponentiation. (Since OpenSSL 0.9.7h, setting this flag
3143 in the exponent causes BN_mod_exp_mont() to use the alternative
3144 implementation in BN_mod_exp_mont_consttime().) The old name
3145 remains as a deprecated alias.
3147 Similary, RSA_FLAG_NO_EXP_CONSTTIME is replaced by a more general
3148 RSA_FLAG_NO_CONSTTIME flag since the RSA implementation now uses
3149 constant-time implementations for more than just exponentiation.
3150 Here too the old name is kept as a deprecated alias.
3152 BN_BLINDING_new() will now use BN_dup() for the modulus so that
3153 the BN_BLINDING structure gets an independent copy of the
3154 modulus. This means that the previous "BIGNUM *m" argument to
3155 BN_BLINDING_new() and to BN_BLINDING_create_param() now
3156 essentially becomes "const BIGNUM *m", although we can't actually
3157 change this in the header file before 0.9.9. It allows
3158 RSA_setup_blinding() to use BN_with_flags() on the modulus to
3159 enable BN_FLG_CONSTTIME.
3161 [Matthew D Wood (Intel Corp)]
3163 *) In the SSL/TLS server implementation, be strict about session ID
3164 context matching (which matters if an application uses a single
3165 external cache for different purposes). Previously,
3166 out-of-context reuse was forbidden only if SSL_VERIFY_PEER was
3167 set. This did ensure strict client verification, but meant that,
3168 with applications using a single external cache for quite
3169 different requirements, clients could circumvent ciphersuite
3170 restrictions for a given session ID context by starting a session
3171 in a different context.
3174 *) Include "!eNULL" in SSL_DEFAULT_CIPHER_LIST to make sure that
3175 a ciphersuite string such as "DEFAULT:RSA" cannot enable
3176 authentication-only ciphersuites.
3179 *) Update the SSL_get_shared_ciphers() fix CVE-2006-3738 which was
3180 not complete and could lead to a possible single byte overflow
3181 (CVE-2007-5135) [Ben Laurie]
3183 Changes between 0.9.8d and 0.9.8e [23 Feb 2007]
3185 *) Since AES128 and AES256 (and similarly Camellia128 and
3186 Camellia256) share a single mask bit in the logic of
3187 ssl/ssl_ciph.c, the code for masking out disabled ciphers needs a
3188 kludge to work properly if AES128 is available and AES256 isn't
3189 (or if Camellia128 is available and Camellia256 isn't).
3192 *) Fix the BIT STRING encoding generated by crypto/ec/ec_asn1.c
3193 (within i2d_ECPrivateKey, i2d_ECPKParameters, i2d_ECParameters):
3194 When a point or a seed is encoded in a BIT STRING, we need to
3195 prevent the removal of trailing zero bits to get the proper DER
3196 encoding. (By default, crypto/asn1/a_bitstr.c assumes the case
3197 of a NamedBitList, for which trailing 0 bits need to be removed.)
3200 *) Have SSL/TLS server implementation tolerate "mismatched" record
3201 protocol version while receiving ClientHello even if the
3202 ClientHello is fragmented. (The server can't insist on the
3203 particular protocol version it has chosen before the ServerHello
3204 message has informed the client about his choice.)
3207 *) Add RFC 3779 support.
3208 [Rob Austein for ARIN, Ben Laurie]
3210 *) Load error codes if they are not already present instead of using a
3211 static variable. This allows them to be cleanly unloaded and reloaded.
3212 Improve header file function name parsing.
3215 *) extend SMTP and IMAP protocol emulation in s_client to use EHLO
3216 or CAPABILITY handshake as required by RFCs.
3219 Changes between 0.9.8c and 0.9.8d [28 Sep 2006]
3221 *) Introduce limits to prevent malicious keys being able to
3222 cause a denial of service. (CVE-2006-2940)
3223 [Steve Henson, Bodo Moeller]
3225 *) Fix ASN.1 parsing of certain invalid structures that can result
3226 in a denial of service. (CVE-2006-2937) [Steve Henson]
3228 *) Fix buffer overflow in SSL_get_shared_ciphers() function.
3229 (CVE-2006-3738) [Tavis Ormandy and Will Drewry, Google Security Team]
3231 *) Fix SSL client code which could crash if connecting to a
3232 malicious SSLv2 server. (CVE-2006-4343)
3233 [Tavis Ormandy and Will Drewry, Google Security Team]
3235 *) Since 0.9.8b, ciphersuite strings naming explicit ciphersuites
3236 match only those. Before that, "AES256-SHA" would be interpreted
3237 as a pattern and match "AES128-SHA" too (since AES128-SHA got
3238 the same strength classification in 0.9.7h) as we currently only
3239 have a single AES bit in the ciphersuite description bitmap.
3240 That change, however, also applied to ciphersuite strings such as
3241 "RC4-MD5" that intentionally matched multiple ciphersuites --
3242 namely, SSL 2.0 ciphersuites in addition to the more common ones
3243 from SSL 3.0/TLS 1.0.
3245 So we change the selection algorithm again: Naming an explicit
3246 ciphersuite selects this one ciphersuite, and any other similar
3247 ciphersuite (same bitmap) from *other* protocol versions.
3248 Thus, "RC4-MD5" again will properly select both the SSL 2.0
3249 ciphersuite and the SSL 3.0/TLS 1.0 ciphersuite.
3251 Since SSL 2.0 does not have any ciphersuites for which the
3252 128/256 bit distinction would be relevant, this works for now.
3253 The proper fix will be to use different bits for AES128 and
3254 AES256, which would have avoided the problems from the beginning;
3255 however, bits are scarce, so we can only do this in a new release
3256 (not just a patchlevel) when we can change the SSL_CIPHER
3257 definition to split the single 'unsigned long mask' bitmap into
3258 multiple values to extend the available space.
3262 Changes between 0.9.8b and 0.9.8c [05 Sep 2006]
3264 *) Avoid PKCS #1 v1.5 signature attack discovered by Daniel Bleichenbacher
3265 (CVE-2006-4339) [Ben Laurie and Google Security Team]
3267 *) Add AES IGE and biIGE modes.
3270 *) Change the Unix randomness entropy gathering to use poll() when
3271 possible instead of select(), since the latter has some
3272 undesirable limitations.
3273 [Darryl Miles via Richard Levitte and Bodo Moeller]
3275 *) Disable "ECCdraft" ciphersuites more thoroughly. Now special
3276 treatment in ssl/ssl_ciph.s makes sure that these ciphersuites
3277 cannot be implicitly activated as part of, e.g., the "AES" alias.
3278 However, please upgrade to OpenSSL 0.9.9[-dev] for
3279 non-experimental use of the ECC ciphersuites to get TLS extension
3280 support, which is required for curve and point format negotiation
3281 to avoid potential handshake problems.
3284 *) Disable rogue ciphersuites:
3286 - SSLv2 0x08 0x00 0x80 ("RC4-64-MD5")
3287 - SSLv3/TLSv1 0x00 0x61 ("EXP1024-RC2-CBC-MD5")
3288 - SSLv3/TLSv1 0x00 0x60 ("EXP1024-RC4-MD5")
3290 The latter two were purportedly from
3291 draft-ietf-tls-56-bit-ciphersuites-0[01].txt, but do not really
3294 Also deactivate the remaining ciphersuites from
3295 draft-ietf-tls-56-bit-ciphersuites-01.txt. These are just as
3296 unofficial, and the ID has long expired.
3299 *) Fix RSA blinding Heisenbug (problems sometimes occured on
3300 dual-core machines) and other potential thread-safety issues.
3303 *) Add the symmetric cipher Camellia (128-bit, 192-bit, 256-bit key
3304 versions), which is now available for royalty-free use
3305 (see http://info.isl.ntt.co.jp/crypt/eng/info/chiteki.html).
3306 Also, add Camellia TLS ciphersuites from RFC 4132.
3308 To minimize changes between patchlevels in the OpenSSL 0.9.8
3309 series, Camellia remains excluded from compilation unless OpenSSL
3310 is configured with 'enable-camellia'.
3313 *) Disable the padding bug check when compression is in use. The padding
3314 bug check assumes the first packet is of even length, this is not
3315 necessarily true if compresssion is enabled and can result in false
3316 positives causing handshake failure. The actual bug test is ancient
3317 code so it is hoped that implementations will either have fixed it by
3318 now or any which still have the bug do not support compression.
3321 Changes between 0.9.8a and 0.9.8b [04 May 2006]
3323 *) When applying a cipher rule check to see if string match is an explicit
3324 cipher suite and only match that one cipher suite if it is.
3327 *) Link in manifests for VC++ if needed.
3328 [Austin Ziegler <halostatue@gmail.com>]
3330 *) Update support for ECC-based TLS ciphersuites according to
3331 draft-ietf-tls-ecc-12.txt with proposed changes (but without
3332 TLS extensions, which are supported starting with the 0.9.9
3333 branch, not in the OpenSSL 0.9.8 branch).
3336 *) New functions EVP_CIPHER_CTX_new() and EVP_CIPHER_CTX_free() to support
3337 opaque EVP_CIPHER_CTX handling.
3340 *) Fixes and enhancements to zlib compression code. We now only use
3341 "zlib1.dll" and use the default __cdecl calling convention on Win32
3342 to conform with the standards mentioned here:
3343 http://www.zlib.net/DLL_FAQ.txt
3344 Static zlib linking now works on Windows and the new --with-zlib-include
3345 --with-zlib-lib options to Configure can be used to supply the location
3346 of the headers and library. Gracefully handle case where zlib library
3350 *) Several fixes and enhancements to the OID generation code. The old code
3351 sometimes allowed invalid OIDs (1.X for X >= 40 for example), couldn't
3352 handle numbers larger than ULONG_MAX, truncated printing and had a
3353 non standard OBJ_obj2txt() behaviour.
3356 *) Add support for building of engines under engine/ as shared libraries
3357 under VC++ build system.
3360 *) Corrected the numerous bugs in the Win32 path splitter in DSO.
3361 Hopefully, we will not see any false combination of paths any more.
3364 Changes between 0.9.8 and 0.9.8a [11 Oct 2005]
3366 *) Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING
3367 (part of SSL_OP_ALL). This option used to disable the
3368 countermeasure against man-in-the-middle protocol-version
3369 rollback in the SSL 2.0 server implementation, which is a bad
3370 idea. (CVE-2005-2969)
3372 [Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center
3373 for Information Security, National Institute of Advanced Industrial
3374 Science and Technology [AIST], Japan)]
3376 *) Add two function to clear and return the verify parameter flags.
3379 *) Keep cipherlists sorted in the source instead of sorting them at
3380 runtime, thus removing the need for a lock.
3383 *) Avoid some small subgroup attacks in Diffie-Hellman.
3384 [Nick Mathewson and Ben Laurie]
3386 *) Add functions for well-known primes.
3389 *) Extended Windows CE support.
3390 [Satoshi Nakamura and Andy Polyakov]
3392 *) Initialize SSL_METHOD structures at compile time instead of during
3393 runtime, thus removing the need for a lock.
3396 *) Make PKCS7_decrypt() work even if no certificate is supplied by
3397 attempting to decrypt each encrypted key in turn. Add support to
3401 Changes between 0.9.7h and 0.9.8 [05 Jul 2005]
3403 [NB: OpenSSL 0.9.7i and later 0.9.7 patch levels were released after
3406 *) Add libcrypto.pc and libssl.pc for those who feel they need them.
3409 *) Change CA.sh and CA.pl so they don't bundle the CSR and the private
3410 key into the same file any more.
3413 *) Add initial support for Win64, both IA64 and AMD64/x64 flavors.
3416 *) Add -utf8 command line and config file option to 'ca'.
3417 [Stefan <stf@udoma.org]
3419 *) Removed the macro des_crypt(), as it seems to conflict with some
3420 libraries. Use DES_crypt().
3423 *) Correct naming of the 'chil' and '4758cca' ENGINEs. This
3424 involves renaming the source and generated shared-libs for
3425 both. The engines will accept the corrected or legacy ids
3426 ('ncipher' and '4758_cca' respectively) when binding. NB,
3427 this only applies when building 'shared'.
3428 [Corinna Vinschen <vinschen@redhat.com> and Geoff Thorpe]
3430 *) Add attribute functions to EVP_PKEY structure. Modify
3431 PKCS12_create() to recognize a CSP name attribute and
3432 use it. Make -CSP option work again in pkcs12 utility.
3435 *) Add new functionality to the bn blinding code:
3436 - automatic re-creation of the BN_BLINDING parameters after
3437 a fixed number of uses (currently 32)
3438 - add new function for parameter creation
3439 - introduce flags to control the update behaviour of the
3440 BN_BLINDING parameters
3441 - hide BN_BLINDING structure
3442 Add a second BN_BLINDING slot to the RSA structure to improve
3443 performance when a single RSA object is shared among several
3447 *) Add support for DTLS.
3448 [Nagendra Modadugu <nagendra@cs.stanford.edu> and Ben Laurie]
3450 *) Add support for DER encoded private keys (SSL_FILETYPE_ASN1)
3451 to SSL_CTX_use_PrivateKey_file() and SSL_use_PrivateKey_file()
3454 *) Remove buggy and incomplete DH cert support from
3455 ssl/ssl_rsa.c and ssl/s3_both.c
3458 *) Use SHA-1 instead of MD5 as the default digest algorithm for
3459 the apps/openssl applications.
3462 *) Compile clean with "-Wall -Wmissing-prototypes
3463 -Wstrict-prototypes -Wmissing-declarations -Werror". Currently
3464 DEBUG_SAFESTACK must also be set.
3467 *) Change ./Configure so that certain algorithms can be disabled by default.
3468 The new counterpiece to "no-xxx" is "enable-xxx".
3470 The patented RC5 and MDC2 algorithms will now be disabled unless
3471 "enable-rc5" and "enable-mdc2", respectively, are specified.
3473 (IDEA remains enabled despite being patented. This is because IDEA
3474 is frequently required for interoperability, and there is no license
3475 fee for non-commercial use. As before, "no-idea" can be used to
3476 avoid this algorithm.)
3480 *) Add processing of proxy certificates (see RFC 3820). This work was
3481 sponsored by KTH (The Royal Institute of Technology in Stockholm) and
3482 EGEE (Enabling Grids for E-science in Europe).
3485 *) RC4 performance overhaul on modern architectures/implementations, such
3486 as Intel P4, IA-64 and AMD64.
3489 *) New utility extract-section.pl. This can be used specify an alternative
3490 section number in a pod file instead of having to treat each file as
3491 a separate case in Makefile. This can be done by adding two lines to the
3494 =for comment openssl_section:XXX
3496 The blank line is mandatory.
3500 *) New arguments -certform, -keyform and -pass for s_client and s_server
3501 to allow alternative format key and certificate files and passphrase
3505 *) New structure X509_VERIFY_PARAM which combines current verify parameters,
3506 update associated structures and add various utility functions.
3508 Add new policy related verify parameters, include policy checking in
3509 standard verify code. Enhance 'smime' application with extra parameters
3510 to support policy checking and print out.
3513 *) Add a new engine to support VIA PadLock ACE extensions in the VIA C3
3514 Nehemiah processors. These extensions support AES encryption in hardware
3515 as well as RNG (though RNG support is currently disabled).
3516 [Michal Ludvig <michal@logix.cz>, with help from Andy Polyakov]
3518 *) Deprecate BN_[get|set]_params() functions (they were ignored internally).
3521 *) New FIPS 180-2 algorithms, SHA-224/-256/-384/-512 are implemented.
3522 [Andy Polyakov and a number of other people]
3524 *) Improved PowerPC platform support. Most notably BIGNUM assembler
3525 implementation contributed by IBM.
3526 [Suresh Chari, Peter Waltenberg, Andy Polyakov]
3528 *) The new 'RSA_generate_key_ex' function now takes a BIGNUM for the public
3529 exponent rather than 'unsigned long'. There is a corresponding change to
3530 the new 'rsa_keygen' element of the RSA_METHOD structure.
3531 [Jelte Jansen, Geoff Thorpe]
3533 *) Functionality for creating the initial serial number file is now
3534 moved from CA.pl to the 'ca' utility with a new option -create_serial.
3536 (Before OpenSSL 0.9.7e, CA.pl used to initialize the serial
3537 number file to 1, which is bound to cause problems. To avoid
3538 the problems while respecting compatibility between different 0.9.7
3539 patchlevels, 0.9.7e employed 'openssl x509 -next_serial' in
3540 CA.pl for serial number initialization. With the new release 0.9.8,
3541 we can fix the problem directly in the 'ca' utility.)
3544 *) Reduced header interdepencies by declaring more opaque objects in
3545 ossl_typ.h. As a consequence, including some headers (eg. engine.h) will
3546 give fewer recursive includes, which could break lazy source code - so
3547 this change is covered by the OPENSSL_NO_DEPRECATED symbol. As always,
3548 developers should define this symbol when building and using openssl to
3549 ensure they track the recommended behaviour, interfaces, [etc], but
3550 backwards-compatible behaviour prevails when this isn't defined.
3553 *) New function X509_POLICY_NODE_print() which prints out policy nodes.
3556 *) Add new EVP function EVP_CIPHER_CTX_rand_key and associated functionality.
3557 This will generate a random key of the appropriate length based on the
3558 cipher context. The EVP_CIPHER can provide its own random key generation
3559 routine to support keys of a specific form. This is used in the des and
3560 3des routines to generate a key of the correct parity. Update S/MIME
3561 code to use new functions and hence generate correct parity DES keys.
3562 Add EVP_CHECK_DES_KEY #define to return an error if the key is not
3563 valid (weak or incorrect parity).
3566 *) Add a local set of CRLs that can be used by X509_verify_cert() as well
3567 as looking them up. This is useful when the verified structure may contain
3568 CRLs, for example PKCS#7 signedData. Modify PKCS7_verify() to use any CRLs
3569 present unless the new PKCS7_NO_CRL flag is asserted.
3572 *) Extend ASN1 oid configuration module. It now additionally accepts the
3575 shortName = some long name, 1.2.3.4
3578 *) Reimplemented the BN_CTX implementation. There is now no more static
3579 limitation on the number of variables it can handle nor the depth of the
3580 "stack" handling for BN_CTX_start()/BN_CTX_end() pairs. The stack
3581 information can now expand as required, and rather than having a single
3582 static array of bignums, BN_CTX now uses a linked-list of such arrays
3583 allowing it to expand on demand whilst maintaining the usefulness of
3584 BN_CTX's "bundling".
3587 *) Add a missing BN_CTX parameter to the 'rsa_mod_exp' callback in RSA_METHOD
3588 to allow all RSA operations to function using a single BN_CTX.
3591 *) Preliminary support for certificate policy evaluation and checking. This
3592 is initially intended to pass the tests outlined in "Conformance Testing
3593 of Relying Party Client Certificate Path Processing Logic" v1.07.
3596 *) bn_dup_expand() has been deprecated, it was introduced in 0.9.7 and
3597 remained unused and not that useful. A variety of other little bignum
3598 tweaks and fixes have also been made continuing on from the audit (see
3602 *) Constify all or almost all d2i, c2i, s2i and r2i functions, along with
3603 associated ASN1, EVP and SSL functions and old ASN1 macros.
3606 *) BN_zero() only needs to set 'top' and 'neg' to zero for correct results,
3607 and this should never fail. So the return value from the use of
3608 BN_set_word() (which can fail due to needless expansion) is now deprecated;
3609 if OPENSSL_NO_DEPRECATED is defined, BN_zero() is a void macro.
3612 *) BN_CTX_get() should return zero-valued bignums, providing the same
3613 initialised value as BN_new().
3614 [Geoff Thorpe, suggested by Ulf Möller]
3616 *) Support for inhibitAnyPolicy certificate extension.
3619 *) An audit of the BIGNUM code is underway, for which debugging code is
3620 enabled when BN_DEBUG is defined. This makes stricter enforcements on what
3621 is considered valid when processing BIGNUMs, and causes execution to
3622 assert() when a problem is discovered. If BN_DEBUG_RAND is defined,
3623 further steps are taken to deliberately pollute unused data in BIGNUM
3624 structures to try and expose faulty code further on. For now, openssl will
3625 (in its default mode of operation) continue to tolerate the inconsistent
3626 forms that it has tolerated in the past, but authors and packagers should
3627 consider trying openssl and their own applications when compiled with
3628 these debugging symbols defined. It will help highlight potential bugs in
3629 their own code, and will improve the test coverage for OpenSSL itself. At
3630 some point, these tighter rules will become openssl's default to improve
3631 maintainability, though the assert()s and other overheads will remain only
3632 in debugging configurations. See bn.h for more details.
3633 [Geoff Thorpe, Nils Larsch, Ulf Möller]
3635 *) BN_CTX_init() has been deprecated, as BN_CTX is an opaque structure
3636 that can only be obtained through BN_CTX_new() (which implicitly
3637 initialises it). The presence of this function only made it possible
3638 to overwrite an existing structure (and cause memory leaks).
3641 *) Because of the callback-based approach for implementing LHASH as a
3642 template type, lh_insert() adds opaque objects to hash-tables and
3643 lh_doall() or lh_doall_arg() are typically used with a destructor callback
3644 to clean up those corresponding objects before destroying the hash table
3645 (and losing the object pointers). So some over-zealous constifications in
3646 LHASH have been relaxed so that lh_insert() does not take (nor store) the
3647 objects as "const" and the lh_doall[_arg] callback wrappers are not
3648 prototyped to have "const" restrictions on the object pointers they are
3649 given (and so aren't required to cast them away any more).
3652 *) The tmdiff.h API was so ugly and minimal that our own timing utility
3653 (speed) prefers to use its own implementation. The two implementations
3654 haven't been consolidated as yet (volunteers?) but the tmdiff API has had
3655 its object type properly exposed (MS_TM) instead of casting to/from "char
3656 *". This may still change yet if someone realises MS_TM and "ms_time_***"
3657 aren't necessarily the greatest nomenclatures - but this is what was used
3658 internally to the implementation so I've used that for now.
3661 *) Ensure that deprecated functions do not get compiled when
3662 OPENSSL_NO_DEPRECATED is defined. Some "openssl" subcommands and a few of
3663 the self-tests were still using deprecated key-generation functions so
3664 these have been updated also.
3667 *) Reorganise PKCS#7 code to separate the digest location functionality
3668 into PKCS7_find_digest(), digest addition into PKCS7_bio_add_digest().
3669 New function PKCS7_set_digest() to set the digest type for PKCS#7
3670 digestedData type. Add additional code to correctly generate the
3671 digestedData type and add support for this type in PKCS7 initialization
3675 *) New function PKCS7_set0_type_other() this initializes a PKCS7
3676 structure of type "other".
3679 *) Fix prime generation loop in crypto/bn/bn_prime.pl by making
3680 sure the loop does correctly stop and breaking ("division by zero")
3681 modulus operations are not performed. The (pre-generated) prime
3682 table crypto/bn/bn_prime.h was already correct, but it could not be
3683 re-generated on some platforms because of the "division by zero"
3684 situation in the script.
3685 [Ralf S. Engelschall]
3687 *) Update support for ECC-based TLS ciphersuites according to
3688 draft-ietf-tls-ecc-03.txt: the KDF1 key derivation function with
3689 SHA-1 now is only used for "small" curves (where the
3690 representation of a field element takes up to 24 bytes); for
3691 larger curves, the field element resulting from ECDH is directly
3692 used as premaster secret.
3693 [Douglas Stebila (Sun Microsystems Laboratories)]
3695 *) Add code for kP+lQ timings to crypto/ec/ectest.c, and add SEC2
3696 curve secp160r1 to the tests.
3697 [Douglas Stebila (Sun Microsystems Laboratories)]
3699 *) Add the possibility to load symbols globally with DSO.
3700 [Götz Babin-Ebell <babin-ebell@trustcenter.de> via Richard Levitte]
3702 *) Add the functions ERR_set_mark() and ERR_pop_to_mark() for better
3703 control of the error stack.
3706 *) Add support for STORE in ENGINE.
3709 *) Add the STORE type. The intention is to provide a common interface
3710 to certificate and key stores, be they simple file-based stores, or
3711 HSM-type store, or LDAP stores, or...
3712 NOTE: The code is currently UNTESTED and isn't really used anywhere.
3715 *) Add a generic structure called OPENSSL_ITEM. This can be used to
3716 pass a list of arguments to any function as well as provide a way
3717 for a function to pass data back to the caller.
3720 *) Add the functions BUF_strndup() and BUF_memdup(). BUF_strndup()
3721 works like BUF_strdup() but can be used to duplicate a portion of
3722 a string. The copy gets NUL-terminated. BUF_memdup() duplicates
3726 *) Add the function sk_find_ex() which works like sk_find(), but will
3727 return an index to an element even if an exact match couldn't be
3728 found. The index is guaranteed to point at the element where the
3729 searched-for key would be inserted to preserve sorting order.
3732 *) Add the function OBJ_bsearch_ex() which works like OBJ_bsearch() but
3733 takes an extra flags argument for optional functionality. Currently,
3734 the following flags are defined:
3736 OBJ_BSEARCH_VALUE_ON_NOMATCH
3737 This one gets OBJ_bsearch_ex() to return a pointer to the first
3738 element where the comparing function returns a negative or zero
projects / openssl.git / blob