5 Changes between 1.0.2 and 1.1.0 [xx XXX xxxx]
7 *) Added support for TLS extended master secret from
8 draft-ietf-tls-session-hash-03.txt. Thanks for Alfredo Pironti for an
9 initial patch which was a great help during development.
12 *) All libssl internal structures have been removed from the public header
13 files, and the OPENSSL_NO_SSL_INTERN option has been removed (since it is
14 now redundant). Users should not attempt to access internal structures
15 directly. Instead they should use the provided API functions.
18 *) config has been changed so that by default OPENSSL_NO_DEPRECATED is used.
19 Access to deprecated functions can be re-enabled by running config with
20 "enable-deprecated". In addition applications wishing to use deprecated
21 functions must define OPENSSL_USE_DEPRECATED. Note that this new behaviour
22 will, by default, disable some transitive includes that previously existed
23 in the header files (e.g. ec.h will no longer, by default, include bn.h)
26 *) Added support for OCB mode. OpenSSL has been granted a patent license
27 compatible with the OpenSSL license for use of OCB. Details are available
28 at https://www.openssl.org/docs/misc/OCB-patent-grant-OpenSSL.pdf. Support
29 for OCB can be removed by calling config with no-ocb.
32 *) SSLv2 support has been removed. It still supports receiving a SSLv2
33 compatible client hello.
36 *) Increased the minimal RSA keysize from 256 to 512 bits [Rich Salz],
37 done while fixing the error code for the key-too-small case.
38 [Annie Yousar <a.yousar@informatik.hu-berlin.de>]
40 *) Removed old DES API.
43 *) Remove various unsupported platforms:
49 Sinix/ReliantUNIX RM400
54 16-bit platforms such as WIN16
57 *) Clean up OPENSSL_NO_xxx #define's
58 Use setbuf() and remove OPENSSL_NO_SETVBUF_IONBF
59 Rename OPENSSL_SYSNAME_xxx to OPENSSL_SYS_xxx
60 OPENSSL_NO_EC{DH,DSA} merged into OPENSSL_NO_EC
61 OPENSSL_NO_RIPEMD160, OPENSSL_NO_RIPEMD merged into OPENSSL_NO_RMD160
62 OPENSSL_NO_FP_API merged into OPENSSL_NO_STDIO
63 Remove OPENSSL_NO_BIO OPENSSL_NO_BUFFER OPENSSL_NO_CHAIN_VERIFY
64 OPENSSL_NO_EVP OPENSSL_NO_FIPS_ERR OPENSSL_NO_HASH_COMP
65 OPENSSL_NO_LHASH OPENSSL_NO_OBJECT OPENSSL_NO_SPEED OPENSSL_NO_STACK
66 OPENSSL_NO_X509 OPENSSL_NO_X509_VERIFY
67 Remove MS_STATIC; it's a relic from platforms <32 bits.
70 *) Cleaned up dead code
71 Remove all but one '#ifdef undef' which is to be looked at.
74 *) Experimental support for a new, fast, unbiased prime candidate generator,
75 bn_probable_prime_dh_coprime(). Not currently used by any prime generator.
76 [Felix Laurie von Massenbach <felix@erbridge.co.uk>]
78 *) New output format NSS in the sess_id command line tool. This allows
79 exporting the session id and the master key in NSS keylog format.
80 [Martin Kaiser <martin@kaiser.cx>]
82 *) Harmonize version and its documentation. -f flag is used to display
84 [mancha <mancha1@zoho.com>]
86 *) Fix eckey_priv_encode so it immediately returns an error upon a failure
88 [mancha <mancha1@zoho.com>]
90 *) Fix some double frees. These are not thought to be exploitable.
91 [mancha <mancha1@zoho.com>]
93 *) A missing bounds check in the handling of the TLS heartbeat extension
94 can be used to reveal up to 64k of memory to a connected client or
97 Thanks for Neel Mehta of Google Security for discovering this bug and to
98 Adam Langley <agl@chromium.org> and Bodo Moeller <bmoeller@acm.org> for
99 preparing the fix (CVE-2014-0160)
100 [Adam Langley, Bodo Moeller]
102 *) Fix for the attack described in the paper "Recovering OpenSSL
103 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
104 by Yuval Yarom and Naomi Benger. Details can be obtained from:
105 http://eprint.iacr.org/2014/140
107 Thanks to Yuval Yarom and Naomi Benger for discovering this
108 flaw and to Yuval Yarom for supplying a fix (CVE-2014-0076)
109 [Yuval Yarom and Naomi Benger]
111 *) Use algorithm specific chains in SSL_CTX_use_certificate_chain_file():
112 this fixes a limitation in previous versions of OpenSSL.
115 *) Experimental encrypt-then-mac support.
117 Experimental support for encrypt then mac from
118 draft-gutmann-tls-encrypt-then-mac-02.txt
120 To enable it set the appropriate extension number (0x42 for the test
121 server) using e.g. -DTLSEXT_TYPE_encrypt_then_mac=0x42
123 For non-compliant peers (i.e. just about everything) this should have no
126 WARNING: EXPERIMENTAL, SUBJECT TO CHANGE.
130 *) Add EVP support for key wrapping algorithms, to avoid problems with
131 existing code the flag EVP_CIPHER_CTX_WRAP_ALLOW has to be set in
132 the EVP_CIPHER_CTX or an error is returned. Add AES and DES3 wrap
133 algorithms and include tests cases.
136 *) Extend CMS code to support RSA-PSS signatures and RSA-OAEP for
140 *) Extended RSA OAEP support via EVP_PKEY API. Options to specify digest,
141 MGF1 digest and OAEP label.
144 *) Make openssl verify return errors.
145 [Chris Palmer <palmer@google.com> and Ben Laurie]
147 *) New function ASN1_TIME_diff to calculate the difference between two
148 ASN1_TIME structures or one structure and the current time.
151 *) Update fips_test_suite to support multiple command line options. New
152 test to induce all self test errors in sequence and check expected
156 *) Add FIPS_{rsa,dsa,ecdsa}_{sign,verify} functions which digest and
157 sign or verify all in one operation.
160 *) Add fips_algvs: a multicall fips utility incorporating all the algorithm
161 test programs and fips_test_suite. Includes functionality to parse
162 the minimal script output of fipsalgest.pl directly.
165 *) Add authorisation parameter to FIPS_module_mode_set().
168 *) Add FIPS selftest for ECDH algorithm using P-224 and B-233 curves.
171 *) Use separate DRBG fields for internal and external flags. New function
172 FIPS_drbg_health_check() to perform on demand health checking. Add
173 generation tests to fips_test_suite with reduced health check interval to
174 demonstrate periodic health checking. Add "nodh" option to
175 fips_test_suite to skip very slow DH test.
178 *) New function FIPS_get_cipherbynid() to lookup FIPS supported ciphers
182 *) More extensive health check for DRBG checking many more failure modes.
183 New function FIPS_selftest_drbg_all() to handle every possible DRBG
184 combination: call this in fips_test_suite.
187 *) Add support for Dual EC DRBG from SP800-90. Update DRBG algorithm test
188 and POST to handle Dual EC cases.
191 *) Add support for canonical generation of DSA parameter 'g'. See
194 *) Add support for HMAC DRBG from SP800-90. Update DRBG algorithm test and
195 POST to handle HMAC cases.
198 *) Add functions FIPS_module_version() and FIPS_module_version_text()
199 to return numerical and string versions of the FIPS module number.
202 *) Rename FIPS_mode_set and FIPS_mode to FIPS_module_mode_set and
203 FIPS_module_mode. FIPS_mode and FIPS_mode_set will be implemented
204 outside the validated module in the FIPS capable OpenSSL.
207 *) Minor change to DRBG entropy callback semantics. In some cases
208 there is no multiple of the block length between min_len and
209 max_len. Allow the callback to return more than max_len bytes
210 of entropy but discard any extra: it is the callback's responsibility
211 to ensure that the extra data discarded does not impact the
212 requested amount of entropy.
215 *) Add PRNG security strength checks to RSA, DSA and ECDSA using
216 information in FIPS186-3, SP800-57 and SP800-131A.
219 *) CCM support via EVP. Interface is very similar to GCM case except we
220 must supply all data in one chunk (i.e. no update, final) and the
221 message length must be supplied if AAD is used. Add algorithm test
225 *) Initial version of POST overhaul. Add POST callback to allow the status
226 of POST to be monitored and/or failures induced. Modify fips_test_suite
227 to use callback. Always run all selftests even if one fails.
230 *) XTS support including algorithm test driver in the fips_gcmtest program.
231 Note: this does increase the maximum key length from 32 to 64 bytes but
232 there should be no binary compatibility issues as existing applications
233 will never use XTS mode.
236 *) Extensive reorganisation of FIPS PRNG behaviour. Remove all dependencies
237 to OpenSSL RAND code and replace with a tiny FIPS RAND API which also
238 performs algorithm blocking for unapproved PRNG types. Also do not
239 set PRNG type in FIPS_mode_set(): leave this to the application.
240 Add default OpenSSL DRBG handling: sets up FIPS PRNG and seeds with
241 the standard OpenSSL PRNG: set additional data to a date time vector.
244 *) Rename old X9.31 PRNG functions of the form FIPS_rand* to FIPS_x931*.
245 This shouldn't present any incompatibility problems because applications
246 shouldn't be using these directly and any that are will need to rethink
247 anyway as the X9.31 PRNG is now deprecated by FIPS 140-2
250 *) Extensive self tests and health checking required by SP800-90 DRBG.
251 Remove strength parameter from FIPS_drbg_instantiate and always
252 instantiate at maximum supported strength.
255 *) Add ECDH code to fips module and fips_ecdhvs for primitives only testing.
258 *) New algorithm test program fips_dhvs to handle DH primitives only testing.
261 *) New function DH_compute_key_padded() to compute a DH key and pad with
262 leading zeroes if needed: this complies with SP800-56A et al.
265 *) Initial implementation of SP800-90 DRBGs for Hash and CTR. Not used by
266 anything, incomplete, subject to change and largely untested at present.
269 *) Modify fipscanisteronly build option to only build the necessary object
270 files by filtering FIPS_EX_OBJ through a perl script in crypto/Makefile.
273 *) Add experimental option FIPSSYMS to give all symbols in
274 fipscanister.o and FIPS or fips prefix. This will avoid
275 conflicts with future versions of OpenSSL. Add perl script
276 util/fipsas.pl to preprocess assembly language source files
277 and rename any affected symbols.
280 *) Add selftest checks and algorithm block of non-fips algorithms in
281 FIPS mode. Remove DES2 from selftests.
284 *) Add ECDSA code to fips module. Add tiny fips_ecdsa_check to just
285 return internal method without any ENGINE dependencies. Add new
286 tiny fips sign and verify functions.
289 *) New build option no-ec2m to disable characteristic 2 code.
292 *) New build option "fipscanisteronly". This only builds fipscanister.o
293 and (currently) associated fips utilities. Uses the file Makefile.fips
294 instead of Makefile.org as the prototype.
297 *) Add some FIPS mode restrictions to GCM. Add internal IV generator.
298 Update fips_gcmtest to use IV generator.
301 *) Initial, experimental EVP support for AES-GCM. AAD can be input by
302 setting output buffer to NULL. The *Final function must be
303 called although it will not retrieve any additional data. The tag
304 can be set or retrieved with a ctrl. The IV length is by default 12
305 bytes (96 bits) but can be set to an alternative value. If the IV
306 length exceeds the maximum IV length (currently 16 bytes) it cannot be
310 *) New flag in ciphers: EVP_CIPH_FLAG_CUSTOM_CIPHER. This means the
311 underlying do_cipher function handles all cipher semantics itself
312 including padding and finalisation. This is useful if (for example)
313 an ENGINE cipher handles block padding itself. The behaviour of
314 do_cipher is subtly changed if this flag is set: the return value
315 is the number of characters written to the output buffer (zero is
316 no longer an error code) or a negative error code. Also if the
317 input buffer is NULL and length 0 finalisation should be performed.
320 *) If a candidate issuer certificate is already part of the constructed
321 path ignore it: new debug notification X509_V_ERR_PATH_LOOP for this case.
324 *) Improve forward-security support: add functions
326 void SSL_CTX_set_not_resumable_session_callback(SSL_CTX *ctx, int (*cb)(SSL *ssl, int is_forward_secure))
327 void SSL_set_not_resumable_session_callback(SSL *ssl, int (*cb)(SSL *ssl, int is_forward_secure))
329 for use by SSL/TLS servers; the callback function will be called whenever a
330 new session is created, and gets to decide whether the session may be
331 cached to make it resumable (return 0) or not (return 1). (As by the
332 SSL/TLS protocol specifications, the session_id sent by the server will be
333 empty to indicate that the session is not resumable; also, the server will
334 not generate RFC 4507 (RFC 5077) session tickets.)
336 A simple reasonable callback implementation is to return is_forward_secure.
337 This parameter will be set to 1 or 0 depending on the ciphersuite selected
338 by the SSL/TLS server library, indicating whether it can provide forward
340 [Emilia Käsper <emilia.kasper@esat.kuleuven.be> (Google)]
342 *) New -verify_name option in command line utilities to set verification
346 *) Initial CMAC implementation. WARNING: EXPERIMENTAL, API MAY CHANGE.
347 Add CMAC pkey methods.
350 *) Experimental renegotiation in s_server -www mode. If the client
351 browses /reneg connection is renegotiated. If /renegcert it is
352 renegotiated requesting a certificate.
355 *) Add an "external" session cache for debugging purposes to s_server. This
356 should help trace issues which normally are only apparent in deployed
357 multi-process servers.
360 *) Extensive audit of libcrypto with DEBUG_UNUSED. Fix many cases where
361 return value is ignored. NB. The functions RAND_add(), RAND_seed(),
362 BIO_set_cipher() and some obscure PEM functions were changed so they
363 can now return an error. The RAND changes required a change to the
364 RAND_METHOD structure.
367 *) New macro __owur for "OpenSSL Warn Unused Result". This makes use of
368 a gcc attribute to warn if the result of a function is ignored. This
369 is enable if DEBUG_UNUSED is set. Add to several functions in evp.h
370 whose return value is often ignored.
373 Changes between 1.0.2 and 1.0.2a [xx XXX xxxx]
375 *) ClientHello sigalgs DoS fix
377 If a client connects to an OpenSSL 1.0.2 server and renegotiates with an
378 invalid signature algorithms extension a NULL pointer dereference will
379 occur. This can be exploited in a DoS attack against the server.
381 This issue was was reported to OpenSSL by David Ramos of Stanford
384 [Stephen Henson and Matt Caswell]
386 *) Multiblock corrupted pointer fix
388 OpenSSL 1.0.2 introduced the "multiblock" performance improvement. This
389 feature only applies on 64 bit x86 architecture platforms that support AES
390 NI instructions. A defect in the implementation of "multiblock" can cause
391 OpenSSL's internal write buffer to become incorrectly set to NULL when
392 using non-blocking IO. Typically, when the user application is using a
393 socket BIO for writing, this will only result in a failed connection.
394 However if some other BIO is used then it is likely that a segmentation
395 fault will be triggered, thus enabling a potential DoS attack.
397 This issue was reported to OpenSSL by Daniel Danner and Rainer Mueller.
401 *) Segmentation fault in DTLSv1_listen fix
403 The DTLSv1_listen function is intended to be stateless and processes the
404 initial ClientHello from many peers. It is common for user code to loop
405 over the call to DTLSv1_listen until a valid ClientHello is received with
406 an associated cookie. A defect in the implementation of DTLSv1_listen means
407 that state is preserved in the SSL object from one invocation to the next
408 that can lead to a segmentation fault. Errors processing the initial
409 ClientHello can trigger this scenario. An example of such an error could be
410 that a DTLS1.0 only client is attempting to connect to a DTLS1.2 only
413 This issue was reported to OpenSSL by Per Allansson.
417 *) Segmentation fault in ASN1_TYPE_cmp fix
419 The function ASN1_TYPE_cmp will crash with an invalid read if an attempt is
420 made to compare ASN.1 boolean types. Since ASN1_TYPE_cmp is used to check
421 certificate signature algorithm consistency this can be used to crash any
422 certificate verification operation and exploited in a DoS attack. Any
423 application which performs certificate verification is vulnerable including
424 OpenSSL clients and servers which enable client authentication.
428 *) Segmentation fault for invalid PSS parameters fix
430 The signature verification routines will crash with a NULL pointer
431 dereference if presented with an ASN.1 signature using the RSA PSS
432 algorithm and invalid parameters. Since these routines are used to verify
433 certificate signature algorithms this can be used to crash any
434 certificate verification operation and exploited in a DoS attack. Any
435 application which performs certificate verification is vulnerable including
436 OpenSSL clients and servers which enable client authentication.
438 This issue was was reported to OpenSSL by Brian Carpenter.
442 *) ASN.1 structure reuse memory corruption fix
444 Reusing a structure in ASN.1 parsing may allow an attacker to cause
445 memory corruption via an invalid write. Such reuse is and has been
446 strongly discouraged and is believed to be rare.
448 Applications that parse structures containing CHOICE or ANY DEFINED BY
449 components may be affected. Certificate parsing (d2i_X509 and related
450 functions) are however not affected. OpenSSL clients and servers are
455 *) PKCS7 NULL pointer dereferences fix
457 The PKCS#7 parsing code does not handle missing outer ContentInfo
458 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs with
459 missing content and trigger a NULL pointer dereference on parsing.
461 Applications that verify PKCS#7 signatures, decrypt PKCS#7 data or
462 otherwise parse PKCS#7 structures from untrusted sources are
463 affected. OpenSSL clients and servers are not affected.
465 This issue was reported to OpenSSL by Michal Zalewski (Google).
469 *) DoS via reachable assert in SSLv2 servers fix
471 A malicious client can trigger an OPENSSL_assert (i.e., an abort) in
472 servers that both support SSLv2 and enable export cipher suites by sending
473 a specially crafted SSLv2 CLIENT-MASTER-KEY message.
475 This issue was discovered by Sean Burford (Google) and Emilia Käsper
476 (OpenSSL development team).
480 *) Empty CKE with client auth and DHE fix
482 If client auth is used then a server can seg fault in the event of a DHE
483 ciphersuite being selected and a zero length ClientKeyExchange message
484 being sent by the client. This could be exploited in a DoS attack.
488 *) Handshake with unseeded PRNG fix
490 Under certain conditions an OpenSSL 1.0.2 client can complete a handshake
491 with an unseeded PRNG. The conditions are:
492 - The client is on a platform where the PRNG has not been seeded
493 automatically, and the user has not seeded manually
494 - A protocol specific client method version has been used (i.e. not
495 SSL_client_methodv23)
496 - A ciphersuite is used that does not require additional random data from
497 the PRNG beyond the initial ClientHello client random (e.g. PSK-RC4-SHA).
499 If the handshake succeeds then the client random that has been used will
500 have been generated from a PRNG with insufficient entropy and therefore the
501 output may be predictable.
503 For example using the following command with an unseeded openssl will
504 succeed on an unpatched platform:
506 openssl s_client -psk 1a2b3c4d -tls1_2 -cipher PSK-RC4-SHA
510 *) Use After Free following d2i_ECPrivatekey error fix
512 A malformed EC private key file consumed via the d2i_ECPrivateKey function
513 could cause a use after free condition. This, in turn, could cause a double
514 free in several private key parsing functions (such as d2i_PrivateKey
515 or EVP_PKCS82PKEY) and could lead to a DoS attack or memory corruption
516 for applications that receive EC private keys from untrusted
517 sources. This scenario is considered rare.
519 This issue was discovered by the BoringSSL project and fixed in their
524 *) X509_to_X509_REQ NULL pointer deref fix
526 The function X509_to_X509_REQ will crash with a NULL pointer dereference if
527 the certificate key is invalid. This function is rarely used in practice.
529 This issue was discovered by Brian Carpenter.
533 *) Removed the export ciphers from the DEFAULT ciphers
536 Changes between 1.0.1l and 1.0.2 [22 Jan 2015]
538 *) Facilitate "universal" ARM builds targeting range of ARM ISAs, e.g.
539 ARMv5 through ARMv8, as opposite to "locking" it to single one.
540 So far those who have to target multiple plaforms would compromise
541 and argue that binary targeting say ARMv5 would still execute on
542 ARMv8. "Universal" build resolves this compromise by providing
543 near-optimal performance even on newer platforms.
546 *) Accelerated NIST P-256 elliptic curve implementation for x86_64
547 (other platforms pending).
548 [Shay Gueron & Vlad Krasnov (Intel Corp), Andy Polyakov]
550 *) Add support for the SignedCertificateTimestampList certificate and
551 OCSP response extensions from RFC6962.
554 *) Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
555 for corner cases. (Certain input points at infinity could lead to
556 bogus results, with non-infinity inputs mapped to infinity too.)
559 *) Initial support for PowerISA 2.0.7, first implemented in POWER8.
560 This covers AES, SHA256/512 and GHASH. "Initial" means that most
561 common cases are optimized and there still is room for further
562 improvements. Vector Permutation AES for Altivec is also added.
565 *) Add support for little-endian ppc64 Linux target.
566 [Marcelo Cerri (IBM)]
568 *) Initial support for AMRv8 ISA crypto extensions. This covers AES,
569 SHA1, SHA256 and GHASH. "Initial" means that most common cases
570 are optimized and there still is room for further improvements.
571 Both 32- and 64-bit modes are supported.
572 [Andy Polyakov, Ard Biesheuvel (Linaro)]
574 *) Improved ARMv7 NEON support.
577 *) Support for SPARC Architecture 2011 crypto extensions, first
578 implemented in SPARC T4. This covers AES, DES, Camellia, SHA1,
579 SHA256/512, MD5, GHASH and modular exponentiation.
580 [Andy Polyakov, David Miller]
582 *) Accelerated modular exponentiation for Intel processors, a.k.a.
584 [Shay Gueron & Vlad Krasnov (Intel Corp)]
586 *) Support for new and upcoming Intel processors, including AVX2,
587 BMI and SHA ISA extensions. This includes additional "stitched"
588 implementations, AESNI-SHA256 and GCM, and multi-buffer support
591 This work was sponsored by Intel Corp.
594 *) Support for DTLS 1.2. This adds two sets of DTLS methods: DTLS_*_method()
595 supports both DTLS 1.2 and 1.0 and should use whatever version the peer
596 supports and DTLSv1_2_*_method() which supports DTLS 1.2 only.
599 *) Use algorithm specific chains in SSL_CTX_use_certificate_chain_file():
600 this fixes a limiation in previous versions of OpenSSL.
603 *) Extended RSA OAEP support via EVP_PKEY API. Options to specify digest,
604 MGF1 digest and OAEP label.
607 *) Add EVP support for key wrapping algorithms, to avoid problems with
608 existing code the flag EVP_CIPHER_CTX_WRAP_ALLOW has to be set in
609 the EVP_CIPHER_CTX or an error is returned. Add AES and DES3 wrap
610 algorithms and include tests cases.
613 *) Add functions to allocate and set the fields of an ECDSA_METHOD
615 [Douglas E. Engert, Steve Henson]
617 *) New functions OPENSSL_gmtime_diff and ASN1_TIME_diff to find the
618 difference in days and seconds between two tm or ASN1_TIME structures.
621 *) Add -rev test option to s_server to just reverse order of characters
622 received by client and send back to server. Also prints an abbreviated
623 summary of the connection parameters.
626 *) New option -brief for s_client and s_server to print out a brief summary
627 of connection parameters.
630 *) Add callbacks for arbitrary TLS extensions.
631 [Trevor Perrin <trevp@trevp.net> and Ben Laurie]
633 *) New option -crl_download in several openssl utilities to download CRLs
634 from CRLDP extension in certificates.
637 *) New options -CRL and -CRLform for s_client and s_server for CRLs.
640 *) New function X509_CRL_diff to generate a delta CRL from the difference
641 of two full CRLs. Add support to "crl" utility.
644 *) New functions to set lookup_crls function and to retrieve
645 X509_STORE from X509_STORE_CTX.
648 *) Print out deprecated issuer and subject unique ID fields in
652 *) Extend OCSP I/O functions so they can be used for simple general purpose
653 HTTP as well as OCSP. New wrapper function which can be used to download
654 CRLs using the OCSP API.
657 *) Delegate command line handling in s_client/s_server to SSL_CONF APIs.
660 *) SSL_CONF* functions. These provide a common framework for application
661 configuration using configuration files or command lines.
664 *) SSL/TLS tracing code. This parses out SSL/TLS records using the
665 message callback and prints the results. Needs compile time option
666 "enable-ssl-trace". New options to s_client and s_server to enable
670 *) New ctrl and macro to retrieve supported points extensions.
671 Print out extension in s_server and s_client.
674 *) New functions to retrieve certificate signature and signature
678 *) Add functions to retrieve and manipulate the raw cipherlist sent by a
682 *) New Suite B modes for TLS code. These use and enforce the requirements
683 of RFC6460: restrict ciphersuites, only permit Suite B algorithms and
684 only use Suite B curves. The Suite B modes can be set by using the
685 strings "SUITEB128", "SUITEB192" or "SUITEB128ONLY" for the cipherstring.
688 *) New chain verification flags for Suite B levels of security. Check
689 algorithms are acceptable when flags are set in X509_verify_cert.
692 *) Make tls1_check_chain return a set of flags indicating checks passed
693 by a certificate chain. Add additional tests to handle client
694 certificates: checks for matching certificate type and issuer name
698 *) If an attempt is made to use a signature algorithm not in the peer
699 preference list abort the handshake. If client has no suitable
700 signature algorithms in response to a certificate request do not
704 *) If server EC tmp key is not in client preference list abort handshake.
707 *) Add support for certificate stores in CERT structure. This makes it
708 possible to have different stores per SSL structure or one store in
709 the parent SSL_CTX. Include distint stores for certificate chain
710 verification and chain building. New ctrl SSL_CTRL_BUILD_CERT_CHAIN
711 to build and store a certificate chain in CERT structure: returing
712 an error if the chain cannot be built: this will allow applications
713 to test if a chain is correctly configured.
715 Note: if the CERT based stores are not set then the parent SSL_CTX
716 store is used to retain compatibility with existing behaviour.
720 *) New function ssl_set_client_disabled to set a ciphersuite disabled
721 mask based on the current session, check mask when sending client
722 hello and checking the requested ciphersuite.
725 *) New ctrls to retrieve and set certificate types in a certificate
726 request message. Print out received values in s_client. If certificate
727 types is not set with custom values set sensible values based on
728 supported signature algorithms.
731 *) Support for distinct client and server supported signature algorithms.
734 *) Add certificate callback. If set this is called whenever a certificate
735 is required by client or server. An application can decide which
736 certificate chain to present based on arbitrary criteria: for example
737 supported signature algorithms. Add very simple example to s_server.
738 This fixes many of the problems and restrictions of the existing client
739 certificate callback: for example you can now clear an existing
740 certificate and specify the whole chain.
743 *) Add new "valid_flags" field to CERT_PKEY structure which determines what
744 the certificate can be used for (if anything). Set valid_flags field
745 in new tls1_check_chain function. Simplify ssl_set_cert_masks which used
746 to have similar checks in it.
748 Add new "cert_flags" field to CERT structure and include a "strict mode".
749 This enforces some TLS certificate requirements (such as only permitting
750 certificate signature algorithms contained in the supported algorithms
751 extension) which some implementations ignore: this option should be used
752 with caution as it could cause interoperability issues.
755 *) Update and tidy signature algorithm extension processing. Work out
756 shared signature algorithms based on preferences and peer algorithms
757 and print them out in s_client and s_server. Abort handshake if no
758 shared signature algorithms.
761 *) Add new functions to allow customised supported signature algorithms
762 for SSL and SSL_CTX structures. Add options to s_client and s_server
766 *) New function SSL_certs_clear() to delete all references to certificates
767 from an SSL structure. Before this once a certificate had been added
768 it couldn't be removed.
771 *) Integrate hostname, email address and IP address checking with certificate
772 verification. New verify options supporting checking in opensl utility.
775 *) Fixes and wildcard matching support to hostname and email checking
776 functions. Add manual page.
777 [Florian Weimer (Red Hat Product Security Team)]
779 *) New functions to check a hostname email or IP address against a
780 certificate. Add options x509 utility to print results of checks against
784 *) Fix OCSP checking.
785 [Rob Stradling <rob.stradling@comodo.com> and Ben Laurie]
787 *) Initial experimental support for explicitly trusted non-root CAs.
788 OpenSSL still tries to build a complete chain to a root but if an
789 intermediate CA has a trust setting included that is used. The first
790 setting is used: whether to trust (e.g., -addtrust option to the x509
794 *) Add -trusted_first option which attempts to find certificates in the
795 trusted store even if an untrusted chain is also supplied.
798 *) MIPS assembly pack updates: support for MIPS32r2 and SmartMIPS ASE,
799 platform support for Linux and Android.
802 *) Support for linux-x32, ILP32 environment in x86_64 framework.
805 *) Experimental multi-implementation support for FIPS capable OpenSSL.
806 When in FIPS mode the approved implementations are used as normal,
807 when not in FIPS mode the internal unapproved versions are used instead.
808 This means that the FIPS capable OpenSSL isn't forced to use the
809 (often lower performance) FIPS implementations outside FIPS mode.
812 *) Transparently support X9.42 DH parameters when calling
813 PEM_read_bio_DHparameters. This means existing applications can handle
814 the new parameter format automatically.
817 *) Initial experimental support for X9.42 DH parameter format: mainly
818 to support use of 'q' parameter for RFC5114 parameters.
821 *) Add DH parameters from RFC5114 including test data to dhtest.
824 *) Support for automatic EC temporary key parameter selection. If enabled
825 the most preferred EC parameters are automatically used instead of
826 hardcoded fixed parameters. Now a server just has to call:
827 SSL_CTX_set_ecdh_auto(ctx, 1) and the server will automatically
828 support ECDH and use the most appropriate parameters.
831 *) Enhance and tidy EC curve and point format TLS extension code. Use
832 static structures instead of allocation if default values are used.
833 New ctrls to set curves we wish to support and to retrieve shared curves.
834 Print out shared curves in s_server. New options to s_server and s_client
835 to set list of supported curves.
838 *) New ctrls to retrieve supported signature algorithms and
839 supported curve values as an array of NIDs. Extend openssl utility
840 to print out received values.
843 *) Add new APIs EC_curve_nist2nid and EC_curve_nid2nist which convert
844 between NIDs and the more common NIST names such as "P-256". Enhance
845 ecparam utility and ECC method to recognise the NIST names for curves.
848 *) Enhance SSL/TLS certificate chain handling to support different
849 chains for each certificate instead of one chain in the parent SSL_CTX.
852 *) Support for fixed DH ciphersuite client authentication: where both
853 server and client use DH certificates with common parameters.
856 *) Support for fixed DH ciphersuites: those requiring DH server
860 *) New function i2d_re_X509_tbs for re-encoding the TBS portion of
862 Note: Related 1.0.2-beta specific macros X509_get_cert_info,
863 X509_CINF_set_modified, X509_CINF_get_issuer, X509_CINF_get_extensions and
864 X509_CINF_get_signature were reverted post internal team review.
866 Changes between 1.0.1k and 1.0.1l [15 Jan 2015]
868 *) Build fixes for the Windows and OpenVMS platforms
869 [Matt Caswell and Richard Levitte]
871 Changes between 1.0.1j and 1.0.1k [8 Jan 2015]
873 *) Fix DTLS segmentation fault in dtls1_get_record. A carefully crafted DTLS
874 message can cause a segmentation fault in OpenSSL due to a NULL pointer
875 dereference. This could lead to a Denial Of Service attack. Thanks to
876 Markus Stenberg of Cisco Systems, Inc. for reporting this issue.
880 *) Fix DTLS memory leak in dtls1_buffer_record. A memory leak can occur in the
881 dtls1_buffer_record function under certain conditions. In particular this
882 could occur if an attacker sent repeated DTLS records with the same
883 sequence number but for the next epoch. The memory leak could be exploited
884 by an attacker in a Denial of Service attack through memory exhaustion.
885 Thanks to Chris Mueller for reporting this issue.
889 *) Fix issue where no-ssl3 configuration sets method to NULL. When openssl is
890 built with the no-ssl3 option and a SSL v3 ClientHello is received the ssl
891 method would be set to NULL which could later result in a NULL pointer
892 dereference. Thanks to Frank Schmirler for reporting this issue.
896 *) Abort handshake if server key exchange message is omitted for ephemeral
899 Thanks to Karthikeyan Bhargavan of the PROSECCO team at INRIA for
900 reporting this issue.
904 *) Remove non-export ephemeral RSA code on client and server. This code
905 violated the TLS standard by allowing the use of temporary RSA keys in
906 non-export ciphersuites and could be used by a server to effectively
907 downgrade the RSA key length used to a value smaller than the server
908 certificate. Thanks for Karthikeyan Bhargavan of the PROSECCO team at
909 INRIA or reporting this issue.
913 *) Fixed issue where DH client certificates are accepted without verification.
914 An OpenSSL server will accept a DH certificate for client authentication
915 without the certificate verify message. This effectively allows a client to
916 authenticate without the use of a private key. This only affects servers
917 which trust a client certificate authority which issues certificates
918 containing DH keys: these are extremely rare and hardly ever encountered.
919 Thanks for Karthikeyan Bhargavan of the PROSECCO team at INRIA or reporting
924 *) Ensure that the session ID context of an SSL is updated when its
925 SSL_CTX is updated via SSL_set_SSL_CTX.
927 The session ID context is typically set from the parent SSL_CTX,
928 and can vary with the CTX.
931 *) Fix various certificate fingerprint issues.
933 By using non-DER or invalid encodings outside the signed portion of a
934 certificate the fingerprint can be changed without breaking the signature.
935 Although no details of the signed portion of the certificate can be changed
936 this can cause problems with some applications: e.g. those using the
937 certificate fingerprint for blacklists.
939 1. Reject signatures with non zero unused bits.
941 If the BIT STRING containing the signature has non zero unused bits reject
942 the signature. All current signature algorithms require zero unused bits.
944 2. Check certificate algorithm consistency.
946 Check the AlgorithmIdentifier inside TBS matches the one in the
947 certificate signature. NB: this will result in signature failure
948 errors for some broken certificates.
950 Thanks to Konrad Kraszewski from Google for reporting this issue.
952 3. Check DSA/ECDSA signatures use DER.
954 Reencode DSA/ECDSA signatures and compare with the original received
955 signature. Return an error if there is a mismatch.
957 This will reject various cases including garbage after signature
958 (thanks to Antti Karjalainen and Tuomo Untinen from the Codenomicon CROSS
959 program for discovering this case) and use of BER or invalid ASN.1 INTEGERs
960 (negative or with leading zeroes).
962 Further analysis was conducted and fixes were developed by Stephen Henson
963 of the OpenSSL core team.
968 *) Correct Bignum squaring. Bignum squaring (BN_sqr) may produce incorrect
969 results on some platforms, including x86_64. This bug occurs at random
970 with a very low probability, and is not known to be exploitable in any
971 way, though its exact impact is difficult to determine. Thanks to Pieter
972 Wuille (Blockstream) who reported this issue and also suggested an initial
973 fix. Further analysis was conducted by the OpenSSL development team and
974 Adam Langley of Google. The final fix was developed by Andy Polyakov of
975 the OpenSSL core team.
979 *) Do not resume sessions on the server if the negotiated protocol
980 version does not match the session's version. Resuming with a different
981 version, while not strictly forbidden by the RFC, is of questionable
982 sanity and breaks all known clients.
983 [David Benjamin, Emilia Käsper]
985 *) Tighten handling of the ChangeCipherSpec (CCS) message: reject
986 early CCS messages during renegotiation. (Note that because
987 renegotiation is encrypted, this early CCS was not exploitable.)
990 *) Tighten client-side session ticket handling during renegotiation:
991 ensure that the client only accepts a session ticket if the server sends
992 the extension anew in the ServerHello. Previously, a TLS client would
993 reuse the old extension state and thus accept a session ticket if one was
994 announced in the initial ServerHello.
996 Similarly, ensure that the client requires a session ticket if one
997 was advertised in the ServerHello. Previously, a TLS client would
998 ignore a missing NewSessionTicket message.
1001 Changes between 1.0.1i and 1.0.1j [15 Oct 2014]
1003 *) SRTP Memory Leak.
1005 A flaw in the DTLS SRTP extension parsing code allows an attacker, who
1006 sends a carefully crafted handshake message, to cause OpenSSL to fail
1007 to free up to 64k of memory causing a memory leak. This could be
1008 exploited in a Denial Of Service attack. This issue affects OpenSSL
1009 1.0.1 server implementations for both SSL/TLS and DTLS regardless of
1010 whether SRTP is used or configured. Implementations of OpenSSL that
1011 have been compiled with OPENSSL_NO_SRTP defined are not affected.
1013 The fix was developed by the OpenSSL team.
1017 *) Session Ticket Memory Leak.
1019 When an OpenSSL SSL/TLS/DTLS server receives a session ticket the
1020 integrity of that ticket is first verified. In the event of a session
1021 ticket integrity check failing, OpenSSL will fail to free memory
1022 causing a memory leak. By sending a large number of invalid session
1023 tickets an attacker could exploit this issue in a Denial Of Service
1028 *) Build option no-ssl3 is incomplete.
1030 When OpenSSL is configured with "no-ssl3" as a build option, servers
1031 could accept and complete a SSL 3.0 handshake, and clients could be
1032 configured to send them.
1034 [Akamai and the OpenSSL team]
1036 *) Add support for TLS_FALLBACK_SCSV.
1037 Client applications doing fallback retries should call
1038 SSL_set_mode(s, SSL_MODE_SEND_FALLBACK_SCSV).
1040 [Adam Langley, Bodo Moeller]
1042 *) Add additional DigestInfo checks.
1044 Reencode DigestInto in DER and check against the original when
1045 verifying RSA signature: this will reject any improperly encoded
1046 DigestInfo structures.
1048 Note: this is a precautionary measure and no attacks are currently known.
1052 Changes between 1.0.1h and 1.0.1i [6 Aug 2014]
1054 *) Fix SRP buffer overrun vulnerability. Invalid parameters passed to the
1055 SRP code can be overrun an internal buffer. Add sanity check that
1056 g, A, B < N to SRP code.
1058 Thanks to Sean Devlin and Watson Ladd of Cryptography Services, NCC
1059 Group for discovering this issue.
1063 *) A flaw in the OpenSSL SSL/TLS server code causes the server to negotiate
1064 TLS 1.0 instead of higher protocol versions when the ClientHello message
1065 is badly fragmented. This allows a man-in-the-middle attacker to force a
1066 downgrade to TLS 1.0 even if both the server and the client support a
1067 higher protocol version, by modifying the client's TLS records.
1069 Thanks to David Benjamin and Adam Langley (Google) for discovering and
1070 researching this issue.
1074 *) OpenSSL DTLS clients enabling anonymous (EC)DH ciphersuites are subject
1075 to a denial of service attack. A malicious server can crash the client
1076 with a null pointer dereference (read) by specifying an anonymous (EC)DH
1077 ciphersuite and sending carefully crafted handshake messages.
1079 Thanks to Felix Gröbert (Google) for discovering and researching this
1084 *) By sending carefully crafted DTLS packets an attacker could cause openssl
1085 to leak memory. This can be exploited through a Denial of Service attack.
1086 Thanks to Adam Langley for discovering and researching this issue.
1090 *) An attacker can force openssl to consume large amounts of memory whilst
1091 processing DTLS handshake messages. This can be exploited through a
1092 Denial of Service attack.
1093 Thanks to Adam Langley for discovering and researching this issue.
1097 *) An attacker can force an error condition which causes openssl to crash
1098 whilst processing DTLS packets due to memory being freed twice. This
1099 can be exploited through a Denial of Service attack.
1100 Thanks to Adam Langley and Wan-Teh Chang for discovering and researching
1105 *) If a multithreaded client connects to a malicious server using a resumed
1106 session and the server sends an ec point format extension it could write
1107 up to 255 bytes to freed memory.
1109 Thanks to Gabor Tyukasz (LogMeIn Inc) for discovering and researching this
1114 *) A malicious server can crash an OpenSSL client with a null pointer
1115 dereference (read) by specifying an SRP ciphersuite even though it was not
1116 properly negotiated with the client. This can be exploited through a
1117 Denial of Service attack.
1119 Thanks to Joonas Kuorilehto and Riku Hietamäki (Codenomicon) for
1120 discovering and researching this issue.
1124 *) A flaw in OBJ_obj2txt may cause pretty printing functions such as
1125 X509_name_oneline, X509_name_print_ex et al. to leak some information
1126 from the stack. Applications may be affected if they echo pretty printing
1127 output to the attacker.
1129 Thanks to Ivan Fratric (Google) for discovering this issue.
1131 [Emilia Käsper, and Steve Henson]
1133 *) Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
1134 for corner cases. (Certain input points at infinity could lead to
1135 bogus results, with non-infinity inputs mapped to infinity too.)
1138 Changes between 1.0.1g and 1.0.1h [5 Jun 2014]
1140 *) Fix for SSL/TLS MITM flaw. An attacker using a carefully crafted
1141 handshake can force the use of weak keying material in OpenSSL
1142 SSL/TLS clients and servers.
1144 Thanks to KIKUCHI Masashi (Lepidum Co. Ltd.) for discovering and
1145 researching this issue. (CVE-2014-0224)
1146 [KIKUCHI Masashi, Steve Henson]
1148 *) Fix DTLS recursion flaw. By sending an invalid DTLS handshake to an
1149 OpenSSL DTLS client the code can be made to recurse eventually crashing
1152 Thanks to Imre Rad (Search-Lab Ltd.) for discovering this issue.
1154 [Imre Rad, Steve Henson]
1156 *) Fix DTLS invalid fragment vulnerability. A buffer overrun attack can
1157 be triggered by sending invalid DTLS fragments to an OpenSSL DTLS
1158 client or server. This is potentially exploitable to run arbitrary
1159 code on a vulnerable client or server.
1161 Thanks to Jüri Aedla for reporting this issue. (CVE-2014-0195)
1162 [Jüri Aedla, Steve Henson]
1164 *) Fix bug in TLS code where clients enable anonymous ECDH ciphersuites
1165 are subject to a denial of service attack.
1167 Thanks to Felix Gröbert and Ivan Fratric at Google for discovering
1168 this issue. (CVE-2014-3470)
1169 [Felix Gröbert, Ivan Fratric, Steve Henson]
1171 *) Harmonize version and its documentation. -f flag is used to display
1173 [mancha <mancha1@zoho.com>]
1175 *) Fix eckey_priv_encode so it immediately returns an error upon a failure
1176 in i2d_ECPrivateKey.
1177 [mancha <mancha1@zoho.com>]
1179 *) Fix some double frees. These are not thought to be exploitable.
1180 [mancha <mancha1@zoho.com>]
1182 Changes between 1.0.1f and 1.0.1g [7 Apr 2014]
1184 *) A missing bounds check in the handling of the TLS heartbeat extension
1185 can be used to reveal up to 64k of memory to a connected client or
1188 Thanks for Neel Mehta of Google Security for discovering this bug and to
1189 Adam Langley <agl@chromium.org> and Bodo Moeller <bmoeller@acm.org> for
1190 preparing the fix (CVE-2014-0160)
1191 [Adam Langley, Bodo Moeller]
1193 *) Fix for the attack described in the paper "Recovering OpenSSL
1194 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
1195 by Yuval Yarom and Naomi Benger. Details can be obtained from:
1196 http://eprint.iacr.org/2014/140
1198 Thanks to Yuval Yarom and Naomi Benger for discovering this
1199 flaw and to Yuval Yarom for supplying a fix (CVE-2014-0076)
1200 [Yuval Yarom and Naomi Benger]
1202 *) TLS pad extension: draft-agl-tls-padding-03
1204 Workaround for the "TLS hang bug" (see FAQ and PR#2771): if the
1205 TLS client Hello record length value would otherwise be > 255 and
1206 less that 512 pad with a dummy extension containing zeroes so it
1207 is at least 512 bytes long.
1209 [Adam Langley, Steve Henson]
1211 Changes between 1.0.1e and 1.0.1f [6 Jan 2014]
1213 *) Fix for TLS record tampering bug. A carefully crafted invalid
1214 handshake could crash OpenSSL with a NULL pointer exception.
1215 Thanks to Anton Johansson for reporting this issues.
1218 *) Keep original DTLS digest and encryption contexts in retransmission
1219 structures so we can use the previous session parameters if they need
1220 to be resent. (CVE-2013-6450)
1223 *) Add option SSL_OP_SAFARI_ECDHE_ECDSA_BUG (part of SSL_OP_ALL) which
1224 avoids preferring ECDHE-ECDSA ciphers when the client appears to be
1225 Safari on OS X. Safari on OS X 10.8..10.8.3 advertises support for
1226 several ECDHE-ECDSA ciphers, but fails to negotiate them. The bug
1227 is fixed in OS X 10.8.4, but Apple have ruled out both hot fixing
1228 10.8..10.8.3 and forcing users to upgrade to 10.8.4 or newer.
1229 [Rob Stradling, Adam Langley]
1231 Changes between 1.0.1d and 1.0.1e [11 Feb 2013]
1233 *) Correct fix for CVE-2013-0169. The original didn't work on AES-NI
1234 supporting platforms or when small records were transferred.
1235 [Andy Polyakov, Steve Henson]
1237 Changes between 1.0.1c and 1.0.1d [5 Feb 2013]
1239 *) Make the decoding of SSLv3, TLS and DTLS CBC records constant time.
1241 This addresses the flaw in CBC record processing discovered by
1242 Nadhem Alfardan and Kenny Paterson. Details of this attack can be found
1243 at: http://www.isg.rhul.ac.uk/tls/
1245 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
1246 Security Group at Royal Holloway, University of London
1247 (www.isg.rhul.ac.uk) for discovering this flaw and Adam Langley and
1248 Emilia Käsper for the initial patch.
1250 [Emilia Käsper, Adam Langley, Ben Laurie, Andy Polyakov, Steve Henson]
1252 *) Fix flaw in AESNI handling of TLS 1.2 and 1.1 records for CBC mode
1253 ciphersuites which can be exploited in a denial of service attack.
1254 Thanks go to and to Adam Langley <agl@chromium.org> for discovering
1255 and detecting this bug and to Wolfgang Ettlinger
1256 <wolfgang.ettlinger@gmail.com> for independently discovering this issue.
1260 *) Return an error when checking OCSP signatures when key is NULL.
1261 This fixes a DoS attack. (CVE-2013-0166)
1264 *) Make openssl verify return errors.
1265 [Chris Palmer <palmer@google.com> and Ben Laurie]
1267 *) Call OCSP Stapling callback after ciphersuite has been chosen, so
1268 the right response is stapled. Also change SSL_get_certificate()
1269 so it returns the certificate actually sent.
1270 See http://rt.openssl.org/Ticket/Display.html?id=2836.
1271 [Rob Stradling <rob.stradling@comodo.com>]
1273 *) Fix possible deadlock when decoding public keys.
1276 *) Don't use TLS 1.0 record version number in initial client hello
1280 Changes between 1.0.1b and 1.0.1c [10 May 2012]
1282 *) Sanity check record length before skipping explicit IV in TLS
1283 1.2, 1.1 and DTLS to fix DoS attack.
1285 Thanks to Codenomicon for discovering this issue using Fuzz-o-Matic
1286 fuzzing as a service testing platform.
1290 *) Initialise tkeylen properly when encrypting CMS messages.
1291 Thanks to Solar Designer of Openwall for reporting this issue.
1294 *) In FIPS mode don't try to use composite ciphers as they are not
1298 Changes between 1.0.1a and 1.0.1b [26 Apr 2012]
1300 *) OpenSSL 1.0.0 sets SSL_OP_ALL to 0x80000FFFL and OpenSSL 1.0.1 and
1301 1.0.1a set SSL_OP_NO_TLSv1_1 to 0x00000400L which would unfortunately
1302 mean any application compiled against OpenSSL 1.0.0 headers setting
1303 SSL_OP_ALL would also set SSL_OP_NO_TLSv1_1, unintentionally disablng
1304 TLS 1.1 also. Fix this by changing the value of SSL_OP_NO_TLSv1_1 to
1305 0x10000000L Any application which was previously compiled against
1306 OpenSSL 1.0.1 or 1.0.1a headers and which cares about SSL_OP_NO_TLSv1_1
1307 will need to be recompiled as a result. Letting be results in
1308 inability to disable specifically TLS 1.1 and in client context,
1309 in unlike event, limit maximum offered version to TLS 1.0 [see below].
1312 *) In order to ensure interoperabilty SSL_OP_NO_protocolX does not
1313 disable just protocol X, but all protocols above X *if* there are
1314 protocols *below* X still enabled. In more practical terms it means
1315 that if application wants to disable TLS1.0 in favor of TLS1.1 and
1316 above, it's not sufficient to pass SSL_OP_NO_TLSv1, one has to pass
1317 SSL_OP_NO_TLSv1|SSL_OP_NO_SSLv3|SSL_OP_NO_SSLv2. This applies to
1321 Changes between 1.0.1 and 1.0.1a [19 Apr 2012]
1323 *) Check for potentially exploitable overflows in asn1_d2i_read_bio
1324 BUF_mem_grow and BUF_mem_grow_clean. Refuse attempts to shrink buffer
1325 in CRYPTO_realloc_clean.
1327 Thanks to Tavis Ormandy, Google Security Team, for discovering this
1328 issue and to Adam Langley <agl@chromium.org> for fixing it.
1330 [Adam Langley (Google), Tavis Ormandy, Google Security Team]
1332 *) Don't allow TLS 1.2 SHA-256 ciphersuites in TLS 1.0, 1.1 connections.
1335 *) Workarounds for some broken servers that "hang" if a client hello
1336 record length exceeds 255 bytes.
1338 1. Do not use record version number > TLS 1.0 in initial client
1339 hello: some (but not all) hanging servers will now work.
1340 2. If we set OPENSSL_MAX_TLS1_2_CIPHER_LENGTH this will truncate
1341 the number of ciphers sent in the client hello. This should be
1342 set to an even number, such as 50, for example by passing:
1343 -DOPENSSL_MAX_TLS1_2_CIPHER_LENGTH=50 to config or Configure.
1344 Most broken servers should now work.
1345 3. If all else fails setting OPENSSL_NO_TLS1_2_CLIENT will disable
1346 TLS 1.2 client support entirely.
1349 *) Fix SEGV in Vector Permutation AES module observed in OpenSSH.
1352 Changes between 1.0.0h and 1.0.1 [14 Mar 2012]
1354 *) Add compatibility with old MDC2 signatures which use an ASN1 OCTET
1355 STRING form instead of a DigestInfo.
1358 *) The format used for MDC2 RSA signatures is inconsistent between EVP
1359 and the RSA_sign/RSA_verify functions. This was made more apparent when
1360 OpenSSL used RSA_sign/RSA_verify for some RSA signatures in particular
1361 those which went through EVP_PKEY_METHOD in 1.0.0 and later. Detect
1362 the correct format in RSA_verify so both forms transparently work.
1365 *) Some servers which support TLS 1.0 can choke if we initially indicate
1366 support for TLS 1.2 and later renegotiate using TLS 1.0 in the RSA
1367 encrypted premaster secret. As a workaround use the maximum permitted
1368 client version in client hello, this should keep such servers happy
1369 and still work with previous versions of OpenSSL.
1372 *) Add support for TLS/DTLS heartbeats.
1373 [Robin Seggelmann <seggelmann@fh-muenster.de>]
1375 *) Add support for SCTP.
1376 [Robin Seggelmann <seggelmann@fh-muenster.de>]
1378 *) Improved PRNG seeding for VOS.
1379 [Paul Green <Paul.Green@stratus.com>]
1381 *) Extensive assembler packs updates, most notably:
1383 - x86[_64]: AES-NI, PCLMULQDQ, RDRAND support;
1384 - x86[_64]: SSSE3 support (SHA1, vector-permutation AES);
1385 - x86_64: bit-sliced AES implementation;
1386 - ARM: NEON support, contemporary platforms optimizations;
1387 - s390x: z196 support;
1388 - *: GHASH and GF(2^m) multiplication implementations;
1392 *) Make TLS-SRP code conformant with RFC 5054 API cleanup
1393 (removal of unnecessary code)
1394 [Peter Sylvester <peter.sylvester@edelweb.fr>]
1396 *) Add TLS key material exporter from RFC 5705.
1399 *) Add DTLS-SRTP negotiation from RFC 5764.
1402 *) Add Next Protocol Negotiation,
1403 http://tools.ietf.org/html/draft-agl-tls-nextprotoneg-00. Can be
1404 disabled with a no-npn flag to config or Configure. Code donated
1406 [Adam Langley <agl@google.com> and Ben Laurie]
1408 *) Add optional 64-bit optimized implementations of elliptic curves NIST-P224,
1409 NIST-P256, NIST-P521, with constant-time single point multiplication on
1410 typical inputs. Compiler support for the nonstandard type __uint128_t is
1411 required to use this (present in gcc 4.4 and later, for 64-bit builds).
1412 Code made available under Apache License version 2.0.
1414 Specify "enable-ec_nistp_64_gcc_128" on the Configure (or config) command
1415 line to include this in your build of OpenSSL, and run "make depend" (or
1416 "make update"). This enables the following EC_METHODs:
1418 EC_GFp_nistp224_method()
1419 EC_GFp_nistp256_method()
1420 EC_GFp_nistp521_method()
1422 EC_GROUP_new_by_curve_name() will automatically use these (while
1423 EC_GROUP_new_curve_GFp() currently prefers the more flexible
1425 [Emilia Käsper, Adam Langley, Bodo Moeller (Google)]
1427 *) Use type ossl_ssize_t instad of ssize_t which isn't available on
1428 all platforms. Move ssize_t definition from e_os.h to the public
1429 header file e_os2.h as it now appears in public header file cms.h
1432 *) New -sigopt option to the ca, req and x509 utilities. Additional
1433 signature parameters can be passed using this option and in
1437 *) Add RSA PSS signing function. This will generate and set the
1438 appropriate AlgorithmIdentifiers for PSS based on those in the
1439 corresponding EVP_MD_CTX structure. No application support yet.
1442 *) Support for companion algorithm specific ASN1 signing routines.
1443 New function ASN1_item_sign_ctx() signs a pre-initialised
1444 EVP_MD_CTX structure and sets AlgorithmIdentifiers based on
1445 the appropriate parameters.
1448 *) Add new algorithm specific ASN1 verification initialisation function
1449 to EVP_PKEY_ASN1_METHOD: this is not in EVP_PKEY_METHOD since the ASN1
1450 handling will be the same no matter what EVP_PKEY_METHOD is used.
1451 Add a PSS handler to support verification of PSS signatures: checked
1452 against a number of sample certificates.
1455 *) Add signature printing for PSS. Add PSS OIDs.
1456 [Steve Henson, Martin Kaiser <lists@kaiser.cx>]
1458 *) Add algorithm specific signature printing. An individual ASN1 method
1459 can now print out signatures instead of the standard hex dump.
1461 More complex signatures (e.g. PSS) can print out more meaningful
1462 information. Include DSA version that prints out the signature
1466 *) Password based recipient info support for CMS library: implementing
1470 *) Split password based encryption into PBES2 and PBKDF2 functions. This
1471 neatly separates the code into cipher and PBE sections and is required
1472 for some algorithms that split PBES2 into separate pieces (such as
1473 password based CMS).
1476 *) Session-handling fixes:
1477 - Fix handling of connections that are resuming with a session ID,
1478 but also support Session Tickets.
1479 - Fix a bug that suppressed issuing of a new ticket if the client
1480 presented a ticket with an expired session.
1481 - Try to set the ticket lifetime hint to something reasonable.
1482 - Make tickets shorter by excluding irrelevant information.
1483 - On the client side, don't ignore renewed tickets.
1484 [Adam Langley, Bodo Moeller (Google)]
1486 *) Fix PSK session representation.
1489 *) Add RC4-MD5 and AESNI-SHA1 "stitched" implementations.
1491 This work was sponsored by Intel.
1494 *) Add GCM support to TLS library. Some custom code is needed to split
1495 the IV between the fixed (from PRF) and explicit (from TLS record)
1496 portions. This adds all GCM ciphersuites supported by RFC5288 and
1497 RFC5289. Generalise some AES* cipherstrings to include GCM and
1498 add a special AESGCM string for GCM only.
1501 *) Expand range of ctrls for AES GCM. Permit setting invocation
1502 field on decrypt and retrieval of invocation field only on encrypt.
1505 *) Add HMAC ECC ciphersuites from RFC5289. Include SHA384 PRF support.
1506 As required by RFC5289 these ciphersuites cannot be used if for
1507 versions of TLS earlier than 1.2.
1510 *) For FIPS capable OpenSSL interpret a NULL default public key method
1511 as unset and return the appropriate default but do *not* set the default.
1512 This means we can return the appropriate method in applications that
1513 switch between FIPS and non-FIPS modes.
1516 *) Redirect HMAC and CMAC operations to FIPS module in FIPS mode. If an
1517 ENGINE is used then we cannot handle that in the FIPS module so we
1518 keep original code iff non-FIPS operations are allowed.
1521 *) Add -attime option to openssl utilities.
1522 [Peter Eckersley <pde@eff.org>, Ben Laurie and Steve Henson]
1524 *) Redirect DSA and DH operations to FIPS module in FIPS mode.
1527 *) Redirect ECDSA and ECDH operations to FIPS module in FIPS mode. Also use
1528 FIPS EC methods unconditionally for now.
1531 *) New build option no-ec2m to disable characteristic 2 code.
1534 *) Backport libcrypto audit of return value checking from 1.1.0-dev; not
1535 all cases can be covered as some introduce binary incompatibilities.
1538 *) Redirect RSA operations to FIPS module including keygen,
1539 encrypt, decrypt, sign and verify. Block use of non FIPS RSA methods.
1542 *) Add similar low level API blocking to ciphers.
1545 *) Low level digest APIs are not approved in FIPS mode: any attempt
1546 to use these will cause a fatal error. Applications that *really* want
1547 to use them can use the private_* version instead.
1550 *) Redirect cipher operations to FIPS module for FIPS builds.
1553 *) Redirect digest operations to FIPS module for FIPS builds.
1556 *) Update build system to add "fips" flag which will link in fipscanister.o
1557 for static and shared library builds embedding a signature if needed.
1560 *) Output TLS supported curves in preference order instead of numerical
1561 order. This is currently hardcoded for the highest order curves first.
1562 This should be configurable so applications can judge speed vs strength.
1565 *) Add TLS v1.2 server support for client authentication.
1568 *) Add support for FIPS mode in ssl library: disable SSLv3, non-FIPS ciphers
1572 *) Functions FIPS_mode_set() and FIPS_mode() which call the underlying
1573 FIPS modules versions.
1576 *) Add TLS v1.2 client side support for client authentication. Keep cache
1577 of handshake records longer as we don't know the hash algorithm to use
1578 until after the certificate request message is received.
1581 *) Initial TLS v1.2 client support. Add a default signature algorithms
1582 extension including all the algorithms we support. Parse new signature
1583 format in client key exchange. Relax some ECC signing restrictions for
1584 TLS v1.2 as indicated in RFC5246.
1587 *) Add server support for TLS v1.2 signature algorithms extension. Switch
1588 to new signature format when needed using client digest preference.
1589 All server ciphersuites should now work correctly in TLS v1.2. No client
1590 support yet and no support for client certificates.
1593 *) Initial TLS v1.2 support. Add new SHA256 digest to ssl code, switch
1594 to SHA256 for PRF when using TLS v1.2 and later. Add new SHA256 based
1595 ciphersuites. At present only RSA key exchange ciphersuites work with
1596 TLS v1.2. Add new option for TLS v1.2 replacing the old and obsolete
1597 SSL_OP_PKCS1_CHECK flags with SSL_OP_NO_TLSv1_2. New TLSv1.2 methods
1598 and version checking.
1601 *) New option OPENSSL_NO_SSL_INTERN. If an application can be compiled
1602 with this defined it will not be affected by any changes to ssl internal
1603 structures. Add several utility functions to allow openssl application
1604 to work with OPENSSL_NO_SSL_INTERN defined.
1608 [Tom Wu <tjw@cs.stanford.edu> and Ben Laurie]
1610 *) Add functions to copy EVP_PKEY_METHOD and retrieve flags and id.
1613 *) Permit abbreviated handshakes when renegotiating using the function
1614 SSL_renegotiate_abbreviated().
1615 [Robin Seggelmann <seggelmann@fh-muenster.de>]
1617 *) Add call to ENGINE_register_all_complete() to
1618 ENGINE_load_builtin_engines(), so some implementations get used
1619 automatically instead of needing explicit application support.
1622 *) Add support for TLS key exporter as described in RFC5705.
1623 [Robin Seggelmann <seggelmann@fh-muenster.de>, Steve Henson]
1625 *) Initial TLSv1.1 support. Since TLSv1.1 is very similar to TLS v1.0 only
1626 a few changes are required:
1628 Add SSL_OP_NO_TLSv1_1 flag.
1629 Add TLSv1_1 methods.
1630 Update version checking logic to handle version 1.1.
1631 Add explicit IV handling (ported from DTLS code).
1632 Add command line options to s_client/s_server.
1635 Changes between 1.0.0g and 1.0.0h [12 Mar 2012]
1637 *) Fix MMA (Bleichenbacher's attack on PKCS #1 v1.5 RSA padding) weakness
1638 in CMS and PKCS7 code. When RSA decryption fails use a random key for
1639 content decryption and always return the same error. Note: this attack
1640 needs on average 2^20 messages so it only affects automated senders. The
1641 old behaviour can be reenabled in the CMS code by setting the
1642 CMS_DEBUG_DECRYPT flag: this is useful for debugging and testing where
1643 an MMA defence is not necessary.
1644 Thanks to Ivan Nestlerode <inestlerode@us.ibm.com> for discovering
1645 this issue. (CVE-2012-0884)
1648 *) Fix CVE-2011-4619: make sure we really are receiving a
1649 client hello before rejecting multiple SGC restarts. Thanks to
1650 Ivan Nestlerode <inestlerode@us.ibm.com> for discovering this bug.
1653 Changes between 1.0.0f and 1.0.0g [18 Jan 2012]
1655 *) Fix for DTLS DoS issue introduced by fix for CVE-2011-4109.
1656 Thanks to Antonio Martin, Enterprise Secure Access Research and
1657 Development, Cisco Systems, Inc. for discovering this bug and
1658 preparing a fix. (CVE-2012-0050)
1661 Changes between 1.0.0e and 1.0.0f [4 Jan 2012]
1663 *) Nadhem Alfardan and Kenny Paterson have discovered an extension
1664 of the Vaudenay padding oracle attack on CBC mode encryption
1665 which enables an efficient plaintext recovery attack against
1666 the OpenSSL implementation of DTLS. Their attack exploits timing
1667 differences arising during decryption processing. A research
1668 paper describing this attack can be found at:
1669 http://www.isg.rhul.ac.uk/~kp/dtls.pdf
1670 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
1671 Security Group at Royal Holloway, University of London
1672 (www.isg.rhul.ac.uk) for discovering this flaw and to Robin Seggelmann
1673 <seggelmann@fh-muenster.de> and Michael Tuexen <tuexen@fh-muenster.de>
1674 for preparing the fix. (CVE-2011-4108)
1675 [Robin Seggelmann, Michael Tuexen]
1677 *) Clear bytes used for block padding of SSL 3.0 records.
1679 [Adam Langley (Google)]
1681 *) Only allow one SGC handshake restart for SSL/TLS. Thanks to George
1682 Kadianakis <desnacked@gmail.com> for discovering this issue and
1683 Adam Langley for preparing the fix. (CVE-2011-4619)
1684 [Adam Langley (Google)]
1686 *) Check parameters are not NULL in GOST ENGINE. (CVE-2012-0027)
1687 [Andrey Kulikov <amdeich@gmail.com>]
1689 *) Prevent malformed RFC3779 data triggering an assertion failure.
1690 Thanks to Andrew Chi, BBN Technologies, for discovering the flaw
1691 and Rob Austein <sra@hactrn.net> for fixing it. (CVE-2011-4577)
1692 [Rob Austein <sra@hactrn.net>]
1694 *) Improved PRNG seeding for VOS.
1695 [Paul Green <Paul.Green@stratus.com>]
1697 *) Fix ssl_ciph.c set-up race.
1698 [Adam Langley (Google)]
1700 *) Fix spurious failures in ecdsatest.c.
1701 [Emilia Käsper (Google)]
1703 *) Fix the BIO_f_buffer() implementation (which was mixing different
1704 interpretations of the '..._len' fields).
1705 [Adam Langley (Google)]
1707 *) Fix handling of BN_BLINDING: now BN_BLINDING_invert_ex (rather than
1708 BN_BLINDING_invert_ex) calls BN_BLINDING_update, ensuring that concurrent
1709 threads won't reuse the same blinding coefficients.
1711 This also avoids the need to obtain the CRYPTO_LOCK_RSA_BLINDING
1712 lock to call BN_BLINDING_invert_ex, and avoids one use of
1713 BN_BLINDING_update for each BN_BLINDING structure (previously,
1714 the last update always remained unused).
1715 [Emilia Käsper (Google)]
1717 *) In ssl3_clear, preserve s3->init_extra along with s3->rbuf.
1718 [Bob Buckholz (Google)]
1720 Changes between 1.0.0d and 1.0.0e [6 Sep 2011]
1722 *) Fix bug where CRLs with nextUpdate in the past are sometimes accepted
1723 by initialising X509_STORE_CTX properly. (CVE-2011-3207)
1724 [Kaspar Brand <ossl@velox.ch>]
1726 *) Fix SSL memory handling for (EC)DH ciphersuites, in particular
1727 for multi-threaded use of ECDH. (CVE-2011-3210)
1728 [Adam Langley (Google)]
1730 *) Fix x509_name_ex_d2i memory leak on bad inputs.
1733 *) Remove hard coded ecdsaWithSHA1 signature tests in ssl code and check
1734 signature public key algorithm by using OID xref utilities instead.
1735 Before this you could only use some ECC ciphersuites with SHA1 only.
1738 *) Add protection against ECDSA timing attacks as mentioned in the paper
1739 by Billy Bob Brumley and Nicola Tuveri, see:
1741 http://eprint.iacr.org/2011/232.pdf
1743 [Billy Bob Brumley and Nicola Tuveri]
1745 Changes between 1.0.0c and 1.0.0d [8 Feb 2011]
1747 *) Fix parsing of OCSP stapling ClientHello extension. CVE-2011-0014
1748 [Neel Mehta, Adam Langley, Bodo Moeller (Google)]
1750 *) Fix bug in string printing code: if *any* escaping is enabled we must
1751 escape the escape character (backslash) or the resulting string is
1755 Changes between 1.0.0b and 1.0.0c [2 Dec 2010]
1757 *) Disable code workaround for ancient and obsolete Netscape browsers
1758 and servers: an attacker can use it in a ciphersuite downgrade attack.
1759 Thanks to Martin Rex for discovering this bug. CVE-2010-4180
1762 *) Fixed J-PAKE implementation error, originally discovered by
1763 Sebastien Martini, further info and confirmation from Stefan
1764 Arentz and Feng Hao. Note that this fix is a security fix. CVE-2010-4252
1767 Changes between 1.0.0a and 1.0.0b [16 Nov 2010]
1769 *) Fix extension code to avoid race conditions which can result in a buffer
1770 overrun vulnerability: resumed sessions must not be modified as they can
1771 be shared by multiple threads. CVE-2010-3864
1774 *) Fix WIN32 build system to correctly link an ENGINE directory into
1778 Changes between 1.0.0 and 1.0.0a [01 Jun 2010]
1780 *) Check return value of int_rsa_verify in pkey_rsa_verifyrecover
1782 [Steve Henson, Peter-Michael Hager <hager@dortmund.net>]
1784 Changes between 0.9.8n and 1.0.0 [29 Mar 2010]
1786 *) Add "missing" function EVP_CIPHER_CTX_copy(). This copies a cipher
1787 context. The operation can be customised via the ctrl mechanism in
1788 case ENGINEs want to include additional functionality.
1791 *) Tolerate yet another broken PKCS#8 key format: private key value negative.
1794 *) Add new -subject_hash_old and -issuer_hash_old options to x509 utility to
1795 output hashes compatible with older versions of OpenSSL.
1796 [Willy Weisz <weisz@vcpc.univie.ac.at>]
1798 *) Fix compression algorithm handling: if resuming a session use the
1799 compression algorithm of the resumed session instead of determining
1800 it from client hello again. Don't allow server to change algorithm.
1803 *) Add load_crls() function to apps tidying load_certs() too. Add option
1804 to verify utility to allow additional CRLs to be included.
1807 *) Update OCSP request code to permit adding custom headers to the request:
1808 some responders need this.
1811 *) The function EVP_PKEY_sign() returns <=0 on error: check return code
1813 [Julia Lawall <julia@diku.dk>]
1815 *) Update verify callback code in apps/s_cb.c and apps/verify.c, it
1816 needlessly dereferenced structures, used obsolete functions and
1817 didn't handle all updated verify codes correctly.
1820 *) Disable MD2 in the default configuration.
1823 *) In BIO_pop() and BIO_push() use the ctrl argument (which was NULL) to
1824 indicate the initial BIO being pushed or popped. This makes it possible
1825 to determine whether the BIO is the one explicitly called or as a result
1826 of the ctrl being passed down the chain. Fix BIO_pop() and SSL BIOs so
1827 it handles reference counts correctly and doesn't zero out the I/O bio
1828 when it is not being explicitly popped. WARNING: applications which
1829 included workarounds for the old buggy behaviour will need to be modified
1830 or they could free up already freed BIOs.
1833 *) Extend the uni2asc/asc2uni => OPENSSL_uni2asc/OPENSSL_asc2uni
1834 renaming to all platforms (within the 0.9.8 branch, this was
1835 done conditionally on Netware platforms to avoid a name clash).
1836 [Guenter <lists@gknw.net>]
1838 *) Add ECDHE and PSK support to DTLS.
1839 [Michael Tuexen <tuexen@fh-muenster.de>]
1841 *) Add CHECKED_STACK_OF macro to safestack.h, otherwise safestack can't
1845 *) Add "missing" function EVP_MD_flags() (without this the only way to
1846 retrieve a digest flags is by accessing the structure directly. Update
1847 EVP_MD_do_all*() and EVP_CIPHER_do_all*() to include the name a digest
1848 or cipher is registered as in the "from" argument. Print out all
1849 registered digests in the dgst usage message instead of manually
1850 attempting to work them out.
1853 *) If no SSLv2 ciphers are used don't use an SSLv2 compatible client hello:
1854 this allows the use of compression and extensions. Change default cipher
1855 string to remove SSLv2 ciphersuites. This effectively avoids ancient SSLv2
1856 by default unless an application cipher string requests it.
1859 *) Alter match criteria in PKCS12_parse(). It used to try to use local
1860 key ids to find matching certificates and keys but some PKCS#12 files
1861 don't follow the (somewhat unwritten) rules and this strategy fails.
1862 Now just gather all certificates together and the first private key
1863 then look for the first certificate that matches the key.
1866 *) Support use of registered digest and cipher names for dgst and cipher
1867 commands instead of having to add each one as a special case. So now
1874 openssl dgst -sha256 foo
1876 and this works for ENGINE based algorithms too.
1880 *) Update Gost ENGINE to support parameter files.
1881 [Victor B. Wagner <vitus@cryptocom.ru>]
1883 *) Support GeneralizedTime in ca utility.
1884 [Oliver Martin <oliver@volatilevoid.net>, Steve Henson]
1886 *) Enhance the hash format used for certificate directory links. The new
1887 form uses the canonical encoding (meaning equivalent names will work
1888 even if they aren't identical) and uses SHA1 instead of MD5. This form
1889 is incompatible with the older format and as a result c_rehash should
1890 be used to rebuild symbolic links.
1893 *) Make PKCS#8 the default write format for private keys, replacing the
1894 traditional format. This form is standardised, more secure and doesn't
1895 include an implicit MD5 dependency.
1898 *) Add a $gcc_devteam_warn option to Configure. The idea is that any code
1899 committed to OpenSSL should pass this lot as a minimum.
1902 *) Add session ticket override functionality for use by EAP-FAST.
1903 [Jouni Malinen <j@w1.fi>]
1905 *) Modify HMAC functions to return a value. Since these can be implemented
1906 in an ENGINE errors can occur.
1909 *) Type-checked OBJ_bsearch_ex.
1912 *) Type-checked OBJ_bsearch. Also some constification necessitated
1913 by type-checking. Still to come: TXT_DB, bsearch(?),
1914 OBJ_bsearch_ex, qsort, CRYPTO_EX_DATA, ASN1_VALUE, ASN1_STRING,
1918 *) New function OPENSSL_gmtime_adj() to add a specific number of days and
1919 seconds to a tm structure directly, instead of going through OS
1920 specific date routines. This avoids any issues with OS routines such
1921 as the year 2038 bug. New *_adj() functions for ASN1 time structures
1922 and X509_time_adj_ex() to cover the extended range. The existing
1923 X509_time_adj() is still usable and will no longer have any date issues.
1926 *) Delta CRL support. New use deltas option which will attempt to locate
1927 and search any appropriate delta CRLs available.
1929 This work was sponsored by Google.
1932 *) Support for CRLs partitioned by reason code. Reorganise CRL processing
1933 code and add additional score elements. Validate alternate CRL paths
1934 as part of the CRL checking and indicate a new error "CRL path validation
1935 error" in this case. Applications wanting additional details can use
1936 the verify callback and check the new "parent" field. If this is not
1937 NULL CRL path validation is taking place. Existing applications wont
1938 see this because it requires extended CRL support which is off by
1941 This work was sponsored by Google.
1944 *) Support for freshest CRL extension.
1946 This work was sponsored by Google.
1949 *) Initial indirect CRL support. Currently only supported in the CRLs
1950 passed directly and not via lookup. Process certificate issuer
1951 CRL entry extension and lookup CRL entries by bother issuer name
1952 and serial number. Check and process CRL issuer entry in IDP extension.
1954 This work was sponsored by Google.
1957 *) Add support for distinct certificate and CRL paths. The CRL issuer
1958 certificate is validated separately in this case. Only enabled if
1959 an extended CRL support flag is set: this flag will enable additional
1960 CRL functionality in future.
1962 This work was sponsored by Google.
1965 *) Add support for policy mappings extension.
1967 This work was sponsored by Google.
1970 *) Fixes to pathlength constraint, self issued certificate handling,
1971 policy processing to align with RFC3280 and PKITS tests.
1973 This work was sponsored by Google.
1976 *) Support for name constraints certificate extension. DN, email, DNS
1977 and URI types are currently supported.
1979 This work was sponsored by Google.
1982 *) To cater for systems that provide a pointer-based thread ID rather
1983 than numeric, deprecate the current numeric thread ID mechanism and
1984 replace it with a structure and associated callback type. This
1985 mechanism allows a numeric "hash" to be extracted from a thread ID in
1986 either case, and on platforms where pointers are larger than 'long',
1987 mixing is done to help ensure the numeric 'hash' is usable even if it
1988 can't be guaranteed unique. The default mechanism is to use "&errno"
1989 as a pointer-based thread ID to distinguish between threads.
1991 Applications that want to provide their own thread IDs should now use
1992 CRYPTO_THREADID_set_callback() to register a callback that will call
1993 either CRYPTO_THREADID_set_numeric() or CRYPTO_THREADID_set_pointer().
1995 Note that ERR_remove_state() is now deprecated, because it is tied
1996 to the assumption that thread IDs are numeric. ERR_remove_state(0)
1997 to free the current thread's error state should be replaced by
1998 ERR_remove_thread_state(NULL).
2000 (This new approach replaces the functions CRYPTO_set_idptr_callback(),
2001 CRYPTO_get_idptr_callback(), and CRYPTO_thread_idptr() that existed in
2002 OpenSSL 0.9.9-dev between June 2006 and August 2008. Also, if an
2003 application was previously providing a numeric thread callback that
2004 was inappropriate for distinguishing threads, then uniqueness might
2005 have been obtained with &errno that happened immediately in the
2006 intermediate development versions of OpenSSL; this is no longer the
2007 case, the numeric thread callback will now override the automatic use
2009 [Geoff Thorpe, with help from Bodo Moeller]
2011 *) Initial support for different CRL issuing certificates. This covers a
2012 simple case where the self issued certificates in the chain exist and
2013 the real CRL issuer is higher in the existing chain.
2015 This work was sponsored by Google.
2018 *) Removed effectively defunct crypto/store from the build.
2021 *) Revamp of STACK to provide stronger type-checking. Still to come:
2022 TXT_DB, bsearch(?), OBJ_bsearch, qsort, CRYPTO_EX_DATA, ASN1_VALUE,
2023 ASN1_STRING, CONF_VALUE.
2026 *) Add a new SSL_MODE_RELEASE_BUFFERS mode flag to release unused buffer
2027 RAM on SSL connections. This option can save about 34k per idle SSL.
2030 *) Revamp of LHASH to provide stronger type-checking. Still to come:
2031 STACK, TXT_DB, bsearch, qsort.
2034 *) Initial support for Cryptographic Message Syntax (aka CMS) based
2035 on RFC3850, RFC3851 and RFC3852. New cms directory and cms utility,
2036 support for data, signedData, compressedData, digestedData and
2037 encryptedData, envelopedData types included. Scripts to check against
2038 RFC4134 examples draft and interop and consistency checks of many
2039 content types and variants.
2042 *) Add options to enc utility to support use of zlib compression BIO.
2045 *) Extend mk1mf to support importing of options and assembly language
2046 files from Configure script, currently only included in VC-WIN32.
2047 The assembly language rules can now optionally generate the source
2048 files from the associated perl scripts.
2051 *) Implement remaining functionality needed to support GOST ciphersuites.
2052 Interop testing has been performed using CryptoPro implementations.
2053 [Victor B. Wagner <vitus@cryptocom.ru>]
2055 *) s390x assembler pack.
2058 *) ARMv4 assembler pack. ARMv4 refers to v4 and later ISA, not CPU
2062 *) Implement Opaque PRF Input TLS extension as specified in
2063 draft-rescorla-tls-opaque-prf-input-00.txt. Since this is not an
2064 official specification yet and no extension type assignment by
2065 IANA exists, this extension (for now) will have to be explicitly
2066 enabled when building OpenSSL by providing the extension number
2067 to use. For example, specify an option
2069 -DTLSEXT_TYPE_opaque_prf_input=0x9527
2071 to the "config" or "Configure" script to enable the extension,
2072 assuming extension number 0x9527 (which is a completely arbitrary
2073 and unofficial assignment based on the MD5 hash of the Internet
2074 Draft). Note that by doing so, you potentially lose
2075 interoperability with other TLS implementations since these might
2076 be using the same extension number for other purposes.
2078 SSL_set_tlsext_opaque_prf_input(ssl, src, len) is used to set the
2079 opaque PRF input value to use in the handshake. This will create
2080 an interal copy of the length-'len' string at 'src', and will
2081 return non-zero for success.
2083 To get more control and flexibility, provide a callback function
2086 SSL_CTX_set_tlsext_opaque_prf_input_callback(ctx, cb)
2087 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg(ctx, arg)
2091 int (*cb)(SSL *, void *peerinput, size_t len, void *arg);
2094 Callback function 'cb' will be called in handshakes, and is
2095 expected to use SSL_set_tlsext_opaque_prf_input() as appropriate.
2096 Argument 'arg' is for application purposes (the value as given to
2097 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg() will directly
2098 be provided to the callback function). The callback function
2099 has to return non-zero to report success: usually 1 to use opaque
2100 PRF input just if possible, or 2 to enforce use of the opaque PRF
2101 input. In the latter case, the library will abort the handshake
2102 if opaque PRF input is not successfully negotiated.
2104 Arguments 'peerinput' and 'len' given to the callback function
2105 will always be NULL and 0 in the case of a client. A server will
2106 see the client's opaque PRF input through these variables if
2107 available (NULL and 0 otherwise). Note that if the server
2108 provides an opaque PRF input, the length must be the same as the
2109 length of the client's opaque PRF input.
2111 Note that the callback function will only be called when creating
2112 a new session (session resumption can resume whatever was
2113 previously negotiated), and will not be called in SSL 2.0
2114 handshakes; thus, SSL_CTX_set_options(ctx, SSL_OP_NO_SSLv2) or
2115 SSL_set_options(ssl, SSL_OP_NO_SSLv2) is especially recommended
2116 for applications that need to enforce opaque PRF input.
2120 *) Update ssl code to support digests other than SHA1+MD5 for handshake
2123 [Victor B. Wagner <vitus@cryptocom.ru>]
2125 *) Add RFC4507 support to OpenSSL. This includes the corrections in
2126 RFC4507bis. The encrypted ticket format is an encrypted encoded
2127 SSL_SESSION structure, that way new session features are automatically
2130 If a client application caches session in an SSL_SESSION structure
2131 support is transparent because tickets are now stored in the encoded
2134 The SSL_CTX structure automatically generates keys for ticket
2135 protection in servers so again support should be possible
2136 with no application modification.
2138 If a client or server wishes to disable RFC4507 support then the option
2139 SSL_OP_NO_TICKET can be set.
2141 Add a TLS extension debugging callback to allow the contents of any client
2142 or server extensions to be examined.
2144 This work was sponsored by Google.
2147 *) Final changes to avoid use of pointer pointer casts in OpenSSL.
2148 OpenSSL should now compile cleanly on gcc 4.2
2149 [Peter Hartley <pdh@utter.chaos.org.uk>, Steve Henson]
2151 *) Update SSL library to use new EVP_PKEY MAC API. Include generic MAC
2152 support including streaming MAC support: this is required for GOST
2153 ciphersuite support.
2154 [Victor B. Wagner <vitus@cryptocom.ru>, Steve Henson]
2156 *) Add option -stream to use PKCS#7 streaming in smime utility. New
2157 function i2d_PKCS7_bio_stream() and PEM_write_PKCS7_bio_stream()
2158 to output in BER and PEM format.
2161 *) Experimental support for use of HMAC via EVP_PKEY interface. This
2162 allows HMAC to be handled via the EVP_DigestSign*() interface. The
2163 EVP_PKEY "key" in this case is the HMAC key, potentially allowing
2164 ENGINE support for HMAC keys which are unextractable. New -mac and
2165 -macopt options to dgst utility.
2168 *) New option -sigopt to dgst utility. Update dgst to use
2169 EVP_Digest{Sign,Verify}*. These two changes make it possible to use
2170 alternative signing paramaters such as X9.31 or PSS in the dgst
2174 *) Change ssl_cipher_apply_rule(), the internal function that does
2175 the work each time a ciphersuite string requests enabling
2176 ("foo+bar"), moving ("+foo+bar"), disabling ("-foo+bar", or
2177 removing ("!foo+bar") a class of ciphersuites: Now it maintains
2178 the order of disabled ciphersuites such that those ciphersuites
2179 that most recently went from enabled to disabled not only stay
2180 in order with respect to each other, but also have higher priority
2181 than other disabled ciphersuites the next time ciphersuites are
2184 This means that you can now say, e.g., "PSK:-PSK:HIGH" to enable
2185 the same ciphersuites as with "HIGH" alone, but in a specific
2186 order where the PSK ciphersuites come first (since they are the
2187 most recently disabled ciphersuites when "HIGH" is parsed).
2189 Also, change ssl_create_cipher_list() (using this new
2190 funcionality) such that between otherwise identical
2191 cihpersuites, ephemeral ECDH is preferred over ephemeral DH in
2195 *) Change ssl_create_cipher_list() so that it automatically
2196 arranges the ciphersuites in reasonable order before starting
2197 to process the rule string. Thus, the definition for "DEFAULT"
2198 (SSL_DEFAULT_CIPHER_LIST) now is just "ALL:!aNULL:!eNULL", but
2199 remains equivalent to "AES:ALL:!aNULL:!eNULL:+aECDH:+kRSA:+RC4:@STRENGTH".
2200 This makes it much easier to arrive at a reasonable default order
2201 in applications for which anonymous ciphers are OK (meaning
2202 that you can't actually use DEFAULT).
2203 [Bodo Moeller; suggested by Victor Duchovni]
2205 *) Split the SSL/TLS algorithm mask (as used for ciphersuite string
2206 processing) into multiple integers instead of setting
2207 "SSL_MKEY_MASK" bits, "SSL_AUTH_MASK" bits, "SSL_ENC_MASK",
2208 "SSL_MAC_MASK", and "SSL_SSL_MASK" bits all in a single integer.
2209 (These masks as well as the individual bit definitions are hidden
2210 away into the non-exported interface ssl/ssl_locl.h, so this
2211 change to the definition of the SSL_CIPHER structure shouldn't
2212 affect applications.) This give us more bits for each of these
2213 categories, so there is no longer a need to coagulate AES128 and
2214 AES256 into a single algorithm bit, and to coagulate Camellia128
2215 and Camellia256 into a single algorithm bit, which has led to all
2218 Thus, among other things, the kludge introduced in 0.9.7m and
2219 0.9.8e for masking out AES256 independently of AES128 or masking
2220 out Camellia256 independently of AES256 is not needed here in 0.9.9.
2222 With the change, we also introduce new ciphersuite aliases that
2223 so far were missing: "AES128", "AES256", "CAMELLIA128", and
2227 *) Add support for dsa-with-SHA224 and dsa-with-SHA256.
2228 Use the leftmost N bytes of the signature input if the input is
2229 larger than the prime q (with N being the size in bytes of q).
2232 *) Very *very* experimental PKCS#7 streaming encoder support. Nothing uses
2233 it yet and it is largely untested.
2236 *) Add support for the ecdsa-with-SHA224/256/384/512 signature types.
2239 *) Initial incomplete changes to avoid need for function casts in OpenSSL
2240 some compilers (gcc 4.2 and later) reject their use. Safestack is
2241 reimplemented. Update ASN1 to avoid use of legacy functions.
2244 *) Win32/64 targets are linked with Winsock2.
2247 *) Add an X509_CRL_METHOD structure to allow CRL processing to be redirected
2248 to external functions. This can be used to increase CRL handling
2249 efficiency especially when CRLs are very large by (for example) storing
2250 the CRL revoked certificates in a database.
2253 *) Overhaul of by_dir code. Add support for dynamic loading of CRLs so
2254 new CRLs added to a directory can be used. New command line option
2255 -verify_return_error to s_client and s_server. This causes real errors
2256 to be returned by the verify callback instead of carrying on no matter
2257 what. This reflects the way a "real world" verify callback would behave.
2260 *) GOST engine, supporting several GOST algorithms and public key formats.
2261 Kindly donated by Cryptocom.
2264 *) Partial support for Issuing Distribution Point CRL extension. CRLs
2265 partitioned by DP are handled but no indirect CRL or reason partitioning
2266 (yet). Complete overhaul of CRL handling: now the most suitable CRL is
2267 selected via a scoring technique which handles IDP and AKID in CRLs.
2270 *) New X509_STORE_CTX callbacks lookup_crls() and lookup_certs() which
2271 will ultimately be used for all verify operations: this will remove the
2272 X509_STORE dependency on certificate verification and allow alternative
2273 lookup methods. X509_STORE based implementations of these two callbacks.
2276 *) Allow multiple CRLs to exist in an X509_STORE with matching issuer names.
2277 Modify get_crl() to find a valid (unexpired) CRL if possible.
2280 *) New function X509_CRL_match() to check if two CRLs are identical. Normally
2281 this would be called X509_CRL_cmp() but that name is already used by
2282 a function that just compares CRL issuer names. Cache several CRL
2283 extensions in X509_CRL structure and cache CRLDP in X509.
2286 *) Store a "canonical" representation of X509_NAME structure (ASN1 Name)
2287 this maps equivalent X509_NAME structures into a consistent structure.
2288 Name comparison can then be performed rapidly using memcmp().
2291 *) Non-blocking OCSP request processing. Add -timeout option to ocsp
2295 *) Allow digests to supply their own micalg string for S/MIME type using
2296 the ctrl EVP_MD_CTRL_MICALG.
2299 *) During PKCS7 signing pass the PKCS7 SignerInfo structure to the
2300 EVP_PKEY_METHOD before and after signing via the EVP_PKEY_CTRL_PKCS7_SIGN
2301 ctrl. It can then customise the structure before and/or after signing
2305 *) New function OBJ_add_sigid() to allow application defined signature OIDs
2306 to be added to OpenSSLs internal tables. New function OBJ_sigid_free()
2307 to free up any added signature OIDs.
2310 *) New functions EVP_CIPHER_do_all(), EVP_CIPHER_do_all_sorted(),
2311 EVP_MD_do_all() and EVP_MD_do_all_sorted() to enumerate internal
2312 digest and cipher tables. New options added to openssl utility:
2313 list-message-digest-algorithms and list-cipher-algorithms.
2316 *) Change the array representation of binary polynomials: the list
2317 of degrees of non-zero coefficients is now terminated with -1.
2318 Previously it was terminated with 0, which was also part of the
2319 value; thus, the array representation was not applicable to
2320 polynomials where t^0 has coefficient zero. This change makes
2321 the array representation useful in a more general context.
2324 *) Various modifications and fixes to SSL/TLS cipher string
2325 handling. For ECC, the code now distinguishes between fixed ECDH
2326 with RSA certificates on the one hand and with ECDSA certificates
2327 on the other hand, since these are separate ciphersuites. The
2328 unused code for Fortezza ciphersuites has been removed.
2330 For consistency with EDH, ephemeral ECDH is now called "EECDH"
2331 (not "ECDHE"). For consistency with the code for DH
2332 certificates, use of ECDH certificates is now considered ECDH
2333 authentication, not RSA or ECDSA authentication (the latter is
2334 merely the CA's signing algorithm and not actively used in the
2337 The temporary ciphersuite alias "ECCdraft" is no longer
2338 available, and ECC ciphersuites are no longer excluded from "ALL"
2339 and "DEFAULT". The following aliases now exist for RFC 4492
2340 ciphersuites, most of these by analogy with the DH case:
2342 kECDHr - ECDH cert, signed with RSA
2343 kECDHe - ECDH cert, signed with ECDSA
2344 kECDH - ECDH cert (signed with either RSA or ECDSA)
2345 kEECDH - ephemeral ECDH
2346 ECDH - ECDH cert or ephemeral ECDH
2352 AECDH - anonymous ECDH
2353 EECDH - non-anonymous ephemeral ECDH (equivalent to "kEECDH:-AECDH")
2357 *) Add additional S/MIME capabilities for AES and GOST ciphers if supported.
2358 Use correct micalg parameters depending on digest(s) in signed message.
2361 *) Add engine support for EVP_PKEY_ASN1_METHOD. Add functions to process
2362 an ENGINE asn1 method. Support ENGINE lookups in the ASN1 code.
2365 *) Initial engine support for EVP_PKEY_METHOD. New functions to permit
2366 an engine to register a method. Add ENGINE lookups for methods and
2367 functional reference processing.
2370 *) New functions EVP_Digest{Sign,Verify)*. These are enchance versions of
2371 EVP_{Sign,Verify}* which allow an application to customise the signature
2375 *) New -resign option to smime utility. This adds one or more signers
2376 to an existing PKCS#7 signedData structure. Also -md option to use an
2377 alternative message digest algorithm for signing.
2380 *) Tidy up PKCS#7 routines and add new functions to make it easier to
2381 create PKCS7 structures containing multiple signers. Update smime
2382 application to support multiple signers.
2385 *) New -macalg option to pkcs12 utility to allow setting of an alternative
2389 *) Initial support for PKCS#5 v2.0 PRFs other than default SHA1 HMAC.
2390 Reorganize PBE internals to lookup from a static table using NIDs,
2391 add support for HMAC PBE OID translation. Add a EVP_CIPHER ctrl:
2392 EVP_CTRL_PBE_PRF_NID this allows a cipher to specify an alternative
2393 PRF which will be automatically used with PBES2.
2396 *) Replace the algorithm specific calls to generate keys in "req" with the
2400 *) Update PKCS#7 enveloped data routines to use new API. This is now
2401 supported by any public key method supporting the encrypt operation. A
2402 ctrl is added to allow the public key algorithm to examine or modify
2403 the PKCS#7 RecipientInfo structure if it needs to: for RSA this is
2407 *) Add a ctrl to asn1 method to allow a public key algorithm to express
2408 a default digest type to use. In most cases this will be SHA1 but some
2409 algorithms (such as GOST) need to specify an alternative digest. The
2410 return value indicates how strong the preference is 1 means optional and
2411 2 is mandatory (that is it is the only supported type). Modify
2412 ASN1_item_sign() to accept a NULL digest argument to indicate it should
2413 use the default md. Update openssl utilities to use the default digest
2414 type for signing if it is not explicitly indicated.
2417 *) Use OID cross reference table in ASN1_sign() and ASN1_verify(). New
2418 EVP_MD flag EVP_MD_FLAG_PKEY_METHOD_SIGNATURE. This uses the relevant
2419 signing method from the key type. This effectively removes the link
2420 between digests and public key types.
2423 *) Add an OID cross reference table and utility functions. Its purpose is to
2424 translate between signature OIDs such as SHA1WithrsaEncryption and SHA1,
2425 rsaEncryption. This will allow some of the algorithm specific hackery
2426 needed to use the correct OID to be removed.
2429 *) Remove algorithm specific dependencies when setting PKCS7_SIGNER_INFO
2430 structures for PKCS7_sign(). They are now set up by the relevant public
2434 *) Add provisional EC pkey method with support for ECDSA and ECDH.
2437 *) Add support for key derivation (agreement) in the API, DH method and
2441 *) Add DSA pkey method and DH pkey methods, extend DH ASN1 method to support
2442 public and private key formats. As a side effect these add additional
2443 command line functionality not previously available: DSA signatures can be
2444 generated and verified using pkeyutl and DH key support and generation in
2449 [Oliver Tappe <zooey@hirschkaefer.de>]
2451 *) New make target "install_html_docs" installs HTML renditions of the
2453 [Oliver Tappe <zooey@hirschkaefer.de>]
2455 *) New utility "genpkey" this is analogous to "genrsa" etc except it can
2456 generate keys for any algorithm. Extend and update EVP_PKEY_METHOD to
2457 support key and parameter generation and add initial key generation
2458 functionality for RSA.
2461 *) Add functions for main EVP_PKEY_method operations. The undocumented
2462 functions EVP_PKEY_{encrypt,decrypt} have been renamed to
2463 EVP_PKEY_{encrypt,decrypt}_old.
2466 *) Initial definitions for EVP_PKEY_METHOD. This will be a high level public
2467 key API, doesn't do much yet.
2470 *) New function EVP_PKEY_asn1_get0_info() to retrieve information about
2471 public key algorithms. New option to openssl utility:
2472 "list-public-key-algorithms" to print out info.
2475 *) Implement the Supported Elliptic Curves Extension for
2476 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
2479 *) Don't free up OIDs in OBJ_cleanup() if they are in use by EVP_MD or
2480 EVP_CIPHER structures to avoid later problems in EVP_cleanup().
2483 *) New utilities pkey and pkeyparam. These are similar to algorithm specific
2484 utilities such as rsa, dsa, dsaparam etc except they process any key
2488 *) Transfer public key printing routines to EVP_PKEY_ASN1_METHOD. New
2489 functions EVP_PKEY_print_public(), EVP_PKEY_print_private(),
2490 EVP_PKEY_print_param() to print public key data from an EVP_PKEY
2494 *) Initial support for pluggable public key ASN1.
2495 De-spaghettify the public key ASN1 handling. Move public and private
2496 key ASN1 handling to a new EVP_PKEY_ASN1_METHOD structure. Relocate
2497 algorithm specific handling to a single module within the relevant
2498 algorithm directory. Add functions to allow (near) opaque processing
2499 of public and private key structures.
2502 *) Implement the Supported Point Formats Extension for
2503 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
2506 *) Add initial support for RFC 4279 PSK TLS ciphersuites. Add members
2507 for the psk identity [hint] and the psk callback functions to the
2508 SSL_SESSION, SSL and SSL_CTX structure.
2511 PSK-RC4-SHA, PSK-3DES-EDE-CBC-SHA, PSK-AES128-CBC-SHA,
2515 SSL_CTX_use_psk_identity_hint
2516 SSL_get_psk_identity_hint
2517 SSL_get_psk_identity
2518 SSL_use_psk_identity_hint
2520 [Mika Kousa and Pasi Eronen of Nokia Corporation]
2522 *) Add RFC 3161 compliant time stamp request creation, response generation
2523 and response verification functionality.
2524 [Zoltán Glózik <zglozik@opentsa.org>, The OpenTSA Project]
2526 *) Add initial support for TLS extensions, specifically for the server_name
2527 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
2528 have new members for a host name. The SSL data structure has an
2529 additional member SSL_CTX *initial_ctx so that new sessions can be
2530 stored in that context to allow for session resumption, even after the
2531 SSL has been switched to a new SSL_CTX in reaction to a client's
2532 server_name extension.
2534 New functions (subject to change):
2536 SSL_get_servername()
2537 SSL_get_servername_type()
2540 New CTRL codes and macros (subject to change):
2542 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
2543 - SSL_CTX_set_tlsext_servername_callback()
2544 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
2545 - SSL_CTX_set_tlsext_servername_arg()
2546 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
2548 openssl s_client has a new '-servername ...' option.
2550 openssl s_server has new options '-servername_host ...', '-cert2 ...',
2551 '-key2 ...', '-servername_fatal' (subject to change). This allows
2552 testing the HostName extension for a specific single host name ('-cert'
2553 and '-key' remain fallbacks for handshakes without HostName
2554 negotiation). If the unrecognized_name alert has to be sent, this by
2555 default is a warning; it becomes fatal with the '-servername_fatal'
2558 [Peter Sylvester, Remy Allais, Christophe Renou]
2560 *) Whirlpool hash implementation is added.
2563 *) BIGNUM code on 64-bit SPARCv9 targets is switched from bn(64,64) to
2564 bn(64,32). Because of instruction set limitations it doesn't have
2565 any negative impact on performance. This was done mostly in order
2566 to make it possible to share assembler modules, such as bn_mul_mont
2567 implementations, between 32- and 64-bit builds without hassle.
2570 *) Move code previously exiled into file crypto/ec/ec2_smpt.c
2571 to ec2_smpl.c, and no longer require the OPENSSL_EC_BIN_PT_COMP
2575 *) New candidate for BIGNUM assembler implementation, bn_mul_mont,
2576 dedicated Montgomery multiplication procedure, is introduced.
2577 BN_MONT_CTX is modified to allow bn_mul_mont to reach for higher
2578 "64-bit" performance on certain 32-bit targets.
2581 *) New option SSL_OP_NO_COMP to disable use of compression selectively
2582 in SSL structures. New SSL ctrl to set maximum send fragment size.
2583 Save memory by seeting the I/O buffer sizes dynamically instead of
2584 using the maximum available value.
2587 *) New option -V for 'openssl ciphers'. This prints the ciphersuite code
2588 in addition to the text details.
2591 *) Very, very preliminary EXPERIMENTAL support for printing of general
2592 ASN1 structures. This currently produces rather ugly output and doesn't
2593 handle several customised structures at all.
2596 *) Integrated support for PVK file format and some related formats such
2597 as MS PUBLICKEYBLOB and PRIVATEKEYBLOB. Command line switches to support
2598 these in the 'rsa' and 'dsa' utilities.
2601 *) Support for PKCS#1 RSAPublicKey format on rsa utility command line.
2604 *) Remove the ancient ASN1_METHOD code. This was only ever used in one
2605 place for the (very old) "NETSCAPE" format certificates which are now
2606 handled using new ASN1 code equivalents.
2609 *) Let the TLSv1_method() etc. functions return a 'const' SSL_METHOD
2610 pointer and make the SSL_METHOD parameter in SSL_CTX_new,
2611 SSL_CTX_set_ssl_version and SSL_set_ssl_method 'const'.
2614 *) Modify CRL distribution points extension code to print out previously
2615 unsupported fields. Enhance extension setting code to allow setting of
2619 *) Add print and set support for Issuing Distribution Point CRL extension.
2622 *) Change 'Configure' script to enable Camellia by default.
2625 Changes between 0.9.8m and 0.9.8n [24 Mar 2010]
2627 *) When rejecting SSL/TLS records due to an incorrect version number, never
2628 update s->server with a new major version number. As of
2629 - OpenSSL 0.9.8m if 'short' is a 16-bit type,
2630 - OpenSSL 0.9.8f if 'short' is longer than 16 bits,
2631 the previous behavior could result in a read attempt at NULL when
2632 receiving specific incorrect SSL/TLS records once record payload
2633 protection is active. (CVE-2010-0740)
2634 [Bodo Moeller, Adam Langley <agl@chromium.org>]
2636 *) Fix for CVE-2010-0433 where some kerberos enabled versions of OpenSSL
2637 could be crashed if the relevant tables were not present (e.g. chrooted).
2638 [Tomas Hoger <thoger@redhat.com>]
2640 Changes between 0.9.8l and 0.9.8m [25 Feb 2010]
2642 *) Always check bn_wexpend() return values for failure. (CVE-2009-3245)
2643 [Martin Olsson, Neel Mehta]
2645 *) Fix X509_STORE locking: Every 'objs' access requires a lock (to
2646 accommodate for stack sorting, always a write lock!).
2649 *) On some versions of WIN32 Heap32Next is very slow. This can cause
2650 excessive delays in the RAND_poll(): over a minute. As a workaround
2651 include a time check in the inner Heap32Next loop too.
2654 *) The code that handled flushing of data in SSL/TLS originally used the
2655 BIO_CTRL_INFO ctrl to see if any data was pending first. This caused
2656 the problem outlined in PR#1949. The fix suggested there however can
2657 trigger problems with buggy BIO_CTRL_WPENDING (e.g. some versions
2658 of Apache). So instead simplify the code to flush unconditionally.
2659 This should be fine since flushing with no data to flush is a no op.
2662 *) Handle TLS versions 2.0 and later properly and correctly use the
2663 highest version of TLS/SSL supported. Although TLS >= 2.0 is some way
2664 off ancient servers have a habit of sticking around for a while...
2667 *) Modify compression code so it frees up structures without using the
2668 ex_data callbacks. This works around a problem where some applications
2669 call CRYPTO_cleanup_all_ex_data() before application exit (e.g. when
2670 restarting) then use compression (e.g. SSL with compression) later.
2671 This results in significant per-connection memory leaks and
2672 has caused some security issues including CVE-2008-1678 and
2676 *) Constify crypto/cast (i.e., <openssl/cast.h>): a CAST_KEY doesn't
2677 change when encrypting or decrypting.
2680 *) Add option SSL_OP_LEGACY_SERVER_CONNECT which will allow clients to
2681 connect and renegotiate with servers which do not support RI.
2682 Until RI is more widely deployed this option is enabled by default.
2685 *) Add "missing" ssl ctrls to clear options and mode.
2688 *) If client attempts to renegotiate and doesn't support RI respond with
2689 a no_renegotiation alert as required by RFC5746. Some renegotiating
2690 TLS clients will continue a connection gracefully when they receive
2691 the alert. Unfortunately OpenSSL mishandled this alert and would hang
2692 waiting for a server hello which it will never receive. Now we treat a
2693 received no_renegotiation alert as a fatal error. This is because
2694 applications requesting a renegotiation might well expect it to succeed
2695 and would have no code in place to handle the server denying it so the
2696 only safe thing to do is to terminate the connection.
2699 *) Add ctrl macro SSL_get_secure_renegotiation_support() which returns 1 if
2700 peer supports secure renegotiation and 0 otherwise. Print out peer
2701 renegotiation support in s_client/s_server.
2704 *) Replace the highly broken and deprecated SPKAC certification method with
2705 the updated NID creation version. This should correctly handle UTF8.
2708 *) Implement RFC5746. Re-enable renegotiation but require the extension
2709 as needed. Unfortunately, SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION
2710 turns out to be a bad idea. It has been replaced by
2711 SSL_OP_ALLOW_UNSAFE_LEGACY_RENEGOTIATION which can be set with
2712 SSL_CTX_set_options(). This is really not recommended unless you
2713 know what you are doing.
2714 [Eric Rescorla <ekr@networkresonance.com>, Ben Laurie, Steve Henson]
2716 *) Fixes to stateless session resumption handling. Use initial_ctx when
2717 issuing and attempting to decrypt tickets in case it has changed during
2718 servername handling. Use a non-zero length session ID when attempting
2719 stateless session resumption: this makes it possible to determine if
2720 a resumption has occurred immediately after receiving server hello
2721 (several places in OpenSSL subtly assume this) instead of later in
2725 *) The functions ENGINE_ctrl(), OPENSSL_isservice(),
2726 CMS_get1_RecipientRequest() and RAND_bytes() can return <=0 on error
2727 fixes for a few places where the return code is not checked
2729 [Julia Lawall <julia@diku.dk>]
2731 *) Add --strict-warnings option to Configure script to include devteam
2732 warnings in other configurations.
2735 *) Add support for --libdir option and LIBDIR variable in makefiles. This
2736 makes it possible to install openssl libraries in locations which
2737 have names other than "lib", for example "/usr/lib64" which some
2739 [Steve Henson, based on patch from Jeremy Utley]
2741 *) Don't allow the use of leading 0x80 in OIDs. This is a violation of
2742 X690 8.9.12 and can produce some misleading textual output of OIDs.
2743 [Steve Henson, reported by Dan Kaminsky]
2745 *) Delete MD2 from algorithm tables. This follows the recommendation in
2746 several standards that it is not used in new applications due to
2747 several cryptographic weaknesses. For binary compatibility reasons
2748 the MD2 API is still compiled in by default.
2751 *) Add compression id to {d2i,i2d}_SSL_SESSION so it is correctly saved
2755 *) Rename uni2asc and asc2uni functions to OPENSSL_uni2asc and
2756 OPENSSL_asc2uni conditionally on Netware platforms to avoid a name
2758 [Guenter <lists@gknw.net>]
2760 *) Fix the server certificate chain building code to use X509_verify_cert(),
2761 it used to have an ad-hoc builder which was unable to cope with anything
2762 other than a simple chain.
2763 [David Woodhouse <dwmw2@infradead.org>, Steve Henson]
2765 *) Don't check self signed certificate signatures in X509_verify_cert()
2766 by default (a flag can override this): it just wastes time without
2767 adding any security. As a useful side effect self signed root CAs
2768 with non-FIPS digests are now usable in FIPS mode.
2771 *) In dtls1_process_out_of_seq_message() the check if the current message
2772 is already buffered was missing. For every new message was memory
2773 allocated, allowing an attacker to perform an denial of service attack
2774 with sending out of seq handshake messages until there is no memory
2775 left. Additionally every future messege was buffered, even if the
2776 sequence number made no sense and would be part of another handshake.
2777 So only messages with sequence numbers less than 10 in advance will be
2778 buffered. (CVE-2009-1378)
2779 [Robin Seggelmann, discovered by Daniel Mentz]
2781 *) Records are buffered if they arrive with a future epoch to be
2782 processed after finishing the corresponding handshake. There is
2783 currently no limitation to this buffer allowing an attacker to perform
2784 a DOS attack with sending records with future epochs until there is no
2785 memory left. This patch adds the pqueue_size() function to determine
2786 the size of a buffer and limits the record buffer to 100 entries.
2788 [Robin Seggelmann, discovered by Daniel Mentz]
2790 *) Keep a copy of frag->msg_header.frag_len so it can be used after the
2791 parent structure is freed. (CVE-2009-1379)
2794 *) Handle non-blocking I/O properly in SSL_shutdown() call.
2795 [Darryl Miles <darryl-mailinglists@netbauds.net>]
2798 [Ilya O. <vrghost@gmail.com>]
2800 Changes between 0.9.8k and 0.9.8l [5 Nov 2009]
2802 *) Disable renegotiation completely - this fixes a severe security
2803 problem (CVE-2009-3555) at the cost of breaking all
2804 renegotiation. Renegotiation can be re-enabled by setting
2805 SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION in s3->flags at
2806 run-time. This is really not recommended unless you know what
2810 Changes between 0.9.8j and 0.9.8k [25 Mar 2009]
2812 *) Don't set val to NULL when freeing up structures, it is freed up by
2813 underlying code. If sizeof(void *) > sizeof(long) this can result in
2814 zeroing past the valid field. (CVE-2009-0789)
2815 [Paolo Ganci <Paolo.Ganci@AdNovum.CH>]
2817 *) Fix bug where return value of CMS_SignerInfo_verify_content() was not
2818 checked correctly. This would allow some invalid signed attributes to
2819 appear to verify correctly. (CVE-2009-0591)
2820 [Ivan Nestlerode <inestlerode@us.ibm.com>]
2822 *) Reject UniversalString and BMPString types with invalid lengths. This
2823 prevents a crash in ASN1_STRING_print_ex() which assumes the strings have
2824 a legal length. (CVE-2009-0590)
2827 *) Set S/MIME signing as the default purpose rather than setting it
2828 unconditionally. This allows applications to override it at the store
2832 *) Permit restricted recursion of ASN1 strings. This is needed in practice
2833 to handle some structures.
2836 *) Improve efficiency of mem_gets: don't search whole buffer each time
2838 [Jeremy Shapiro <jnshapir@us.ibm.com>]
2840 *) New -hex option for openssl rand.
2843 *) Print out UTF8String and NumericString when parsing ASN1.
2846 *) Support NumericString type for name components.
2849 *) Allow CC in the environment to override the automatically chosen
2850 compiler. Note that nothing is done to ensure flags work with the
2854 Changes between 0.9.8i and 0.9.8j [07 Jan 2009]
2856 *) Properly check EVP_VerifyFinal() and similar return values
2858 [Ben Laurie, Bodo Moeller, Google Security Team]
2860 *) Enable TLS extensions by default.
2863 *) Allow the CHIL engine to be loaded, whether the application is
2864 multithreaded or not. (This does not release the developer from the
2865 obligation to set up the dynamic locking callbacks.)
2866 [Sander Temme <sander@temme.net>]
2868 *) Use correct exit code if there is an error in dgst command.
2869 [Steve Henson; problem pointed out by Roland Dirlewanger]
2871 *) Tweak Configure so that you need to say "experimental-jpake" to enable
2872 JPAKE, and need to use -DOPENSSL_EXPERIMENTAL_JPAKE in applications.
2875 *) Add experimental JPAKE support, including demo authentication in
2876 s_client and s_server.
2879 *) Set the comparison function in v3_addr_canonize().
2880 [Rob Austein <sra@hactrn.net>]
2882 *) Add support for XMPP STARTTLS in s_client.
2883 [Philip Paeps <philip@freebsd.org>]
2885 *) Change the server-side SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG behavior
2886 to ensure that even with this option, only ciphersuites in the
2887 server's preference list will be accepted. (Note that the option
2888 applies only when resuming a session, so the earlier behavior was
2889 just about the algorithm choice for symmetric cryptography.)
2892 Changes between 0.9.8h and 0.9.8i [15 Sep 2008]
2894 *) Fix NULL pointer dereference if a DTLS server received
2895 ChangeCipherSpec as first record (CVE-2009-1386).
2898 *) Fix a state transition in s3_srvr.c and d1_srvr.c
2899 (was using SSL3_ST_CW_CLNT_HELLO_B, should be ..._ST_SW_SRVR_...).
2902 *) The fix in 0.9.8c that supposedly got rid of unsafe
2903 double-checked locking was incomplete for RSA blinding,
2904 addressing just one layer of what turns out to have been
2905 doubly unsafe triple-checked locking.
2907 So now fix this for real by retiring the MONT_HELPER macro
2908 in crypto/rsa/rsa_eay.c.
2910 [Bodo Moeller; problem pointed out by Marius Schilder]
2912 *) Various precautionary measures:
2914 - Avoid size_t integer overflow in HASH_UPDATE (md32_common.h).
2916 - Avoid a buffer overflow in d2i_SSL_SESSION() (ssl_asn1.c).
2917 (NB: This would require knowledge of the secret session ticket key
2918 to exploit, in which case you'd be SOL either way.)
2920 - Change bn_nist.c so that it will properly handle input BIGNUMs
2921 outside the expected range.
2923 - Enforce the 'num' check in BN_div() (bn_div.c) for non-BN_DEBUG
2926 [Neel Mehta, Bodo Moeller]
2928 *) Allow engines to be "soft loaded" - i.e. optionally don't die if
2929 the load fails. Useful for distros.
2930 [Ben Laurie and the FreeBSD team]
2932 *) Add support for Local Machine Keyset attribute in PKCS#12 files.
2935 *) Fix BN_GF2m_mod_arr() top-bit cleanup code.
2938 *) Expand ENGINE to support engine supplied SSL client certificate functions.
2940 This work was sponsored by Logica.
2943 *) Add CryptoAPI ENGINE to support use of RSA and DSA keys held in Windows
2944 keystores. Support for SSL/TLS client authentication too.
2945 Not compiled unless enable-capieng specified to Configure.
2947 This work was sponsored by Logica.
2950 *) Fix bug in X509_ATTRIBUTE creation: dont set attribute using
2951 ASN1_TYPE_set1 if MBSTRING flag set. This bug would crash certain
2952 attribute creation routines such as certifcate requests and PKCS#12
2956 Changes between 0.9.8g and 0.9.8h [28 May 2008]
2958 *) Fix flaw if 'Server Key exchange message' is omitted from a TLS
2959 handshake which could lead to a cilent crash as found using the
2960 Codenomicon TLS test suite (CVE-2008-1672)
2961 [Steve Henson, Mark Cox]
2963 *) Fix double free in TLS server name extensions which could lead to
2964 a remote crash found by Codenomicon TLS test suite (CVE-2008-0891)
2967 *) Clear error queue in SSL_CTX_use_certificate_chain_file()
2969 Clear the error queue to ensure that error entries left from
2970 older function calls do not interfere with the correct operation.
2971 [Lutz Jaenicke, Erik de Castro Lopo]
2973 *) Remove root CA certificates of commercial CAs:
2975 The OpenSSL project does not recommend any specific CA and does not
2976 have any policy with respect to including or excluding any CA.
2977 Therefore it does not make any sense to ship an arbitrary selection
2978 of root CA certificates with the OpenSSL software.
2981 *) RSA OAEP patches to fix two separate invalid memory reads.
2982 The first one involves inputs when 'lzero' is greater than
2983 'SHA_DIGEST_LENGTH' (it would read about SHA_DIGEST_LENGTH bytes
2984 before the beginning of from). The second one involves inputs where
2985 the 'db' section contains nothing but zeroes (there is a one-byte
2986 invalid read after the end of 'db').
2987 [Ivan Nestlerode <inestlerode@us.ibm.com>]
2989 *) Partial backport from 0.9.9-dev:
2991 Introduce bn_mul_mont (dedicated Montgomery multiplication
2992 procedure) as a candidate for BIGNUM assembler implementation.
2993 While 0.9.9-dev uses assembler for various architectures, only
2994 x86_64 is available by default here in the 0.9.8 branch, and
2995 32-bit x86 is available through a compile-time setting.
2997 To try the 32-bit x86 assembler implementation, use Configure
2998 option "enable-montasm" (which exists only for this backport).
3000 As "enable-montasm" for 32-bit x86 disclaims code stability
3001 anyway, in this constellation we activate additional code
3002 backported from 0.9.9-dev for further performance improvements,
3003 namely BN_from_montgomery_word. (To enable this otherwise,
3004 e.g. x86_64, try "-DMONT_FROM_WORD___NON_DEFAULT_0_9_8_BUILD".)
3006 [Andy Polyakov (backport partially by Bodo Moeller)]
3008 *) Add TLS session ticket callback. This allows an application to set
3009 TLS ticket cipher and HMAC keys rather than relying on hardcoded fixed
3010 values. This is useful for key rollover for example where several key
3011 sets may exist with different names.
3014 *) Reverse ENGINE-internal logic for caching default ENGINE handles.
3015 This was broken until now in 0.9.8 releases, such that the only way
3016 a registered ENGINE could be used (assuming it initialises
3017 successfully on the host) was to explicitly set it as the default
3018 for the relevant algorithms. This is in contradiction with 0.9.7
3019 behaviour and the documentation. With this fix, when an ENGINE is
3020 registered into a given algorithm's table of implementations, the
3021 'uptodate' flag is reset so that auto-discovery will be used next
3022 time a new context for that algorithm attempts to select an
3024 [Ian Lister (tweaked by Geoff Thorpe)]
3026 *) Backport of CMS code to OpenSSL 0.9.8. This differs from the 0.9.9
3027 implemention in the following ways:
3029 Lack of EVP_PKEY_ASN1_METHOD means algorithm parameters have to be
3032 Lack of BER streaming support means one pass streaming processing is
3033 only supported if data is detached: setting the streaming flag is
3034 ignored for embedded content.
3036 CMS support is disabled by default and must be explicitly enabled
3037 with the enable-cms configuration option.
3040 *) Update the GMP engine glue to do direct copies between BIGNUM and
3041 mpz_t when openssl and GMP use the same limb size. Otherwise the
3042 existing "conversion via a text string export" trick is still used.
3043 [Paul Sheer <paulsheer@gmail.com>]
3045 *) Zlib compression BIO. This is a filter BIO which compressed and
3046 uncompresses any data passed through it.
3049 *) Add AES_wrap_key() and AES_unwrap_key() functions to implement
3050 RFC3394 compatible AES key wrapping.
3053 *) Add utility functions to handle ASN1 structures. ASN1_STRING_set0():
3054 sets string data without copying. X509_ALGOR_set0() and
3055 X509_ALGOR_get0(): set and retrieve X509_ALGOR (AlgorithmIdentifier)
3056 data. Attribute function X509at_get0_data_by_OBJ(): retrieves data
3057 from an X509_ATTRIBUTE structure optionally checking it occurs only
3058 once. ASN1_TYPE_set1(): set and ASN1_TYPE structure copying supplied
3062 *) Fix BN flag handling in RSA_eay_mod_exp() and BN_MONT_CTX_set()
3063 to get the expected BN_FLG_CONSTTIME behavior.
3064 [Bodo Moeller (Google)]
3068 - fixed wrong usage of ioctlsocket() when build for LIBC BSD sockets
3069 - fixed do_tests.pl to run the test suite with CLIB builds too (CLIB_OPT)
3070 - added some more tests to do_tests.pl
3071 - fixed RunningProcess usage so that it works with newer LIBC NDKs too
3072 - removed usage of BN_LLONG for CLIB builds to avoid runtime dependency
3073 - added new Configure targets netware-clib-bsdsock, netware-clib-gcc,
3074 netware-clib-bsdsock-gcc, netware-libc-bsdsock-gcc
3075 - various changes to netware.pl to enable gcc-cross builds on Win32
3077 - changed crypto/bio/b_sock.c to work with macro functions (CLIB BSD)
3078 - various changes to fix missing prototype warnings
3079 - fixed x86nasm.pl to create correct asm files for NASM COFF output
3080 - added AES, WHIRLPOOL and CPUID assembler code to build files
3081 - added missing AES assembler make rules to mk1mf.pl
3082 - fixed order of includes in apps/ocsp.c so that e_os.h settings apply
3083 [Guenter Knauf <eflash@gmx.net>]
3085 *) Implement certificate status request TLS extension defined in RFC3546.
3086 A client can set the appropriate parameters and receive the encoded
3087 OCSP response via a callback. A server can query the supplied parameters
3088 and set the encoded OCSP response in the callback. Add simplified examples
3089 to s_client and s_server.
3092 Changes between 0.9.8f and 0.9.8g [19 Oct 2007]
3094 *) Fix various bugs:
3095 + Binary incompatibility of ssl_ctx_st structure
3096 + DTLS interoperation with non-compliant servers
3097 + Don't call get_session_cb() without proposed session
3098 + Fix ia64 assembler code
3099 [Andy Polyakov, Steve Henson]
3101 Changes between 0.9.8e and 0.9.8f [11 Oct 2007]
3103 *) DTLS Handshake overhaul. There were longstanding issues with
3104 OpenSSL DTLS implementation, which were making it impossible for
3105 RFC 4347 compliant client to communicate with OpenSSL server.
3106 Unfortunately just fixing these incompatibilities would "cut off"
3107 pre-0.9.8f clients. To allow for hassle free upgrade post-0.9.8e
3108 server keeps tolerating non RFC compliant syntax. The opposite is
3109 not true, 0.9.8f client can not communicate with earlier server.
3110 This update even addresses CVE-2007-4995.
3113 *) Changes to avoid need for function casts in OpenSSL: some compilers
3114 (gcc 4.2 and later) reject their use.
3115 [Kurt Roeckx <kurt@roeckx.be>, Peter Hartley <pdh@utter.chaos.org.uk>,
3118 *) Add RFC4507 support to OpenSSL. This includes the corrections in
3119 RFC4507bis. The encrypted ticket format is an encrypted encoded
3120 SSL_SESSION structure, that way new session features are automatically
3123 If a client application caches session in an SSL_SESSION structure
3124 support is transparent because tickets are now stored in the encoded
3127 The SSL_CTX structure automatically generates keys for ticket
3128 protection in servers so again support should be possible
3129 with no application modification.
3131 If a client or server wishes to disable RFC4507 support then the option
3132 SSL_OP_NO_TICKET can be set.
3134 Add a TLS extension debugging callback to allow the contents of any client
3135 or server extensions to be examined.
3137 This work was sponsored by Google.
3140 *) Add initial support for TLS extensions, specifically for the server_name
3141 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
3142 have new members for a host name. The SSL data structure has an
3143 additional member SSL_CTX *initial_ctx so that new sessions can be
3144 stored in that context to allow for session resumption, even after the
3145 SSL has been switched to a new SSL_CTX in reaction to a client's
3146 server_name extension.
3148 New functions (subject to change):
3150 SSL_get_servername()
3151 SSL_get_servername_type()
3154 New CTRL codes and macros (subject to change):
3156 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
3157 - SSL_CTX_set_tlsext_servername_callback()
3158 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
3159 - SSL_CTX_set_tlsext_servername_arg()
3160 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
3162 openssl s_client has a new '-servername ...' option.
3164 openssl s_server has new options '-servername_host ...', '-cert2 ...',
3165 '-key2 ...', '-servername_fatal' (subject to change). This allows
3166 testing the HostName extension for a specific single host name ('-cert'
3167 and '-key' remain fallbacks for handshakes without HostName
3168 negotiation). If the unrecognized_name alert has to be sent, this by
3169 default is a warning; it becomes fatal with the '-servername_fatal'
3172 [Peter Sylvester, Remy Allais, Christophe Renou, Steve Henson]
3174 *) Add AES and SSE2 assembly language support to VC++ build.
3177 *) Mitigate attack on final subtraction in Montgomery reduction.
3180 *) Fix crypto/ec/ec_mult.c to work properly with scalars of value 0
3181 (which previously caused an internal error).
3184 *) Squeeze another 10% out of IGE mode when in != out.
3187 *) AES IGE mode speedup.
3188 [Dean Gaudet (Google)]
3190 *) Add the Korean symmetric 128-bit cipher SEED (see
3191 http://www.kisa.or.kr/kisa/seed/jsp/seed_eng.jsp) and
3192 add SEED ciphersuites from RFC 4162:
3194 TLS_RSA_WITH_SEED_CBC_SHA = "SEED-SHA"
3195 TLS_DHE_DSS_WITH_SEED_CBC_SHA = "DHE-DSS-SEED-SHA"
3196 TLS_DHE_RSA_WITH_SEED_CBC_SHA = "DHE-RSA-SEED-SHA"
3197 TLS_DH_anon_WITH_SEED_CBC_SHA = "ADH-SEED-SHA"
3199 To minimize changes between patchlevels in the OpenSSL 0.9.8
3200 series, SEED remains excluded from compilation unless OpenSSL
3201 is configured with 'enable-seed'.
3202 [KISA, Bodo Moeller]
3204 *) Mitigate branch prediction attacks, which can be practical if a
3205 single processor is shared, allowing a spy process to extract
3206 information. For detailed background information, see
3207 http://eprint.iacr.org/2007/039 (O. Aciicmez, S. Gueron,
3208 J.-P. Seifert, "New Branch Prediction Vulnerabilities in OpenSSL
3209 and Necessary Software Countermeasures"). The core of the change
3210 are new versions BN_div_no_branch() and
3211 BN_mod_inverse_no_branch() of BN_div() and BN_mod_inverse(),
3212 respectively, which are slower, but avoid the security-relevant
3213 conditional branches. These are automatically called by BN_div()
3214 and BN_mod_inverse() if the flag BN_FLG_CONSTTIME is set for one
3215 of the input BIGNUMs. Also, BN_is_bit_set() has been changed to
3216 remove a conditional branch.
3218 BN_FLG_CONSTTIME is the new name for the previous
3219 BN_FLG_EXP_CONSTTIME flag, since it now affects more than just
3220 modular exponentiation. (Since OpenSSL 0.9.7h, setting this flag
3221 in the exponent causes BN_mod_exp_mont() to use the alternative
3222 implementation in BN_mod_exp_mont_consttime().) The old name
3223 remains as a deprecated alias.
3225 Similary, RSA_FLAG_NO_EXP_CONSTTIME is replaced by a more general
3226 RSA_FLAG_NO_CONSTTIME flag since the RSA implementation now uses
3227 constant-time implementations for more than just exponentiation.
3228 Here too the old name is kept as a deprecated alias.
3230 BN_BLINDING_new() will now use BN_dup() for the modulus so that
3231 the BN_BLINDING structure gets an independent copy of the
3232 modulus. This means that the previous "BIGNUM *m" argument to
3233 BN_BLINDING_new() and to BN_BLINDING_create_param() now
3234 essentially becomes "const BIGNUM *m", although we can't actually
3235 change this in the header file before 0.9.9. It allows
3236 RSA_setup_blinding() to use BN_with_flags() on the modulus to
3237 enable BN_FLG_CONSTTIME.
3239 [Matthew D Wood (Intel Corp)]
3241 *) In the SSL/TLS server implementation, be strict about session ID
3242 context matching (which matters if an application uses a single
3243 external cache for different purposes). Previously,
3244 out-of-context reuse was forbidden only if SSL_VERIFY_PEER was
3245 set. This did ensure strict client verification, but meant that,
3246 with applications using a single external cache for quite
3247 different requirements, clients could circumvent ciphersuite
3248 restrictions for a given session ID context by starting a session
3249 in a different context.
3252 *) Include "!eNULL" in SSL_DEFAULT_CIPHER_LIST to make sure that
3253 a ciphersuite string such as "DEFAULT:RSA" cannot enable
3254 authentication-only ciphersuites.
3257 *) Update the SSL_get_shared_ciphers() fix CVE-2006-3738 which was
3258 not complete and could lead to a possible single byte overflow
3259 (CVE-2007-5135) [Ben Laurie]
3261 Changes between 0.9.8d and 0.9.8e [23 Feb 2007]
3263 *) Since AES128 and AES256 (and similarly Camellia128 and
3264 Camellia256) share a single mask bit in the logic of
3265 ssl/ssl_ciph.c, the code for masking out disabled ciphers needs a
3266 kludge to work properly if AES128 is available and AES256 isn't
3267 (or if Camellia128 is available and Camellia256 isn't).
3270 *) Fix the BIT STRING encoding generated by crypto/ec/ec_asn1.c
3271 (within i2d_ECPrivateKey, i2d_ECPKParameters, i2d_ECParameters):
3272 When a point or a seed is encoded in a BIT STRING, we need to
3273 prevent the removal of trailing zero bits to get the proper DER
3274 encoding. (By default, crypto/asn1/a_bitstr.c assumes the case
3275 of a NamedBitList, for which trailing 0 bits need to be removed.)
3278 *) Have SSL/TLS server implementation tolerate "mismatched" record
3279 protocol version while receiving ClientHello even if the
3280 ClientHello is fragmented. (The server can't insist on the
3281 particular protocol version it has chosen before the ServerHello
3282 message has informed the client about his choice.)
3285 *) Add RFC 3779 support.
3286 [Rob Austein for ARIN, Ben Laurie]
3288 *) Load error codes if they are not already present instead of using a
3289 static variable. This allows them to be cleanly unloaded and reloaded.
3290 Improve header file function name parsing.
3293 *) extend SMTP and IMAP protocol emulation in s_client to use EHLO
3294 or CAPABILITY handshake as required by RFCs.
3297 Changes between 0.9.8c and 0.9.8d [28 Sep 2006]
3299 *) Introduce limits to prevent malicious keys being able to
3300 cause a denial of service. (CVE-2006-2940)
3301 [Steve Henson, Bodo Moeller]
3303 *) Fix ASN.1 parsing of certain invalid structures that can result
3304 in a denial of service. (CVE-2006-2937) [Steve Henson]
3306 *) Fix buffer overflow in SSL_get_shared_ciphers() function.
3307 (CVE-2006-3738) [Tavis Ormandy and Will Drewry, Google Security Team]
3309 *) Fix SSL client code which could crash if connecting to a
3310 malicious SSLv2 server. (CVE-2006-4343)
3311 [Tavis Ormandy and Will Drewry, Google Security Team]
3313 *) Since 0.9.8b, ciphersuite strings naming explicit ciphersuites
3314 match only those. Before that, "AES256-SHA" would be interpreted
3315 as a pattern and match "AES128-SHA" too (since AES128-SHA got
3316 the same strength classification in 0.9.7h) as we currently only
3317 have a single AES bit in the ciphersuite description bitmap.
3318 That change, however, also applied to ciphersuite strings such as
3319 "RC4-MD5" that intentionally matched multiple ciphersuites --
3320 namely, SSL 2.0 ciphersuites in addition to the more common ones
3321 from SSL 3.0/TLS 1.0.
3323 So we change the selection algorithm again: Naming an explicit
3324 ciphersuite selects this one ciphersuite, and any other similar
3325 ciphersuite (same bitmap) from *other* protocol versions.
3326 Thus, "RC4-MD5" again will properly select both the SSL 2.0
3327 ciphersuite and the SSL 3.0/TLS 1.0 ciphersuite.
3329 Since SSL 2.0 does not have any ciphersuites for which the
3330 128/256 bit distinction would be relevant, this works for now.
3331 The proper fix will be to use different bits for AES128 and
3332 AES256, which would have avoided the problems from the beginning;
3333 however, bits are scarce, so we can only do this in a new release
3334 (not just a patchlevel) when we can change the SSL_CIPHER
3335 definition to split the single 'unsigned long mask' bitmap into
3336 multiple values to extend the available space.
3340 Changes between 0.9.8b and 0.9.8c [05 Sep 2006]
3342 *) Avoid PKCS #1 v1.5 signature attack discovered by Daniel Bleichenbacher
3343 (CVE-2006-4339) [Ben Laurie and Google Security Team]
3345 *) Add AES IGE and biIGE modes.
3348 *) Change the Unix randomness entropy gathering to use poll() when
3349 possible instead of select(), since the latter has some
3350 undesirable limitations.
3351 [Darryl Miles via Richard Levitte and Bodo Moeller]
3353 *) Disable "ECCdraft" ciphersuites more thoroughly. Now special
3354 treatment in ssl/ssl_ciph.s makes sure that these ciphersuites
3355 cannot be implicitly activated as part of, e.g., the "AES" alias.
3356 However, please upgrade to OpenSSL 0.9.9[-dev] for
3357 non-experimental use of the ECC ciphersuites to get TLS extension
3358 support, which is required for curve and point format negotiation
3359 to avoid potential handshake problems.
3362 *) Disable rogue ciphersuites:
3364 - SSLv2 0x08 0x00 0x80 ("RC4-64-MD5")
3365 - SSLv3/TLSv1 0x00 0x61 ("EXP1024-RC2-CBC-MD5")
3366 - SSLv3/TLSv1 0x00 0x60 ("EXP1024-RC4-MD5")
3368 The latter two were purportedly from
3369 draft-ietf-tls-56-bit-ciphersuites-0[01].txt, but do not really
3372 Also deactivate the remaining ciphersuites from
3373 draft-ietf-tls-56-bit-ciphersuites-01.txt. These are just as
3374 unofficial, and the ID has long expired.
3377 *) Fix RSA blinding Heisenbug (problems sometimes occured on
3378 dual-core machines) and other potential thread-safety issues.
3381 *) Add the symmetric cipher Camellia (128-bit, 192-bit, 256-bit key
3382 versions), which is now available for royalty-free use
3383 (see http://info.isl.ntt.co.jp/crypt/eng/info/chiteki.html).
3384 Also, add Camellia TLS ciphersuites from RFC 4132.
3386 To minimize changes between patchlevels in the OpenSSL 0.9.8
3387 series, Camellia remains excluded from compilation unless OpenSSL
3388 is configured with 'enable-camellia'.
3391 *) Disable the padding bug check when compression is in use. The padding
3392 bug check assumes the first packet is of even length, this is not
3393 necessarily true if compresssion is enabled and can result in false
3394 positives causing handshake failure. The actual bug test is ancient
3395 code so it is hoped that implementations will either have fixed it by
3396 now or any which still have the bug do not support compression.
3399 Changes between 0.9.8a and 0.9.8b [04 May 2006]
3401 *) When applying a cipher rule check to see if string match is an explicit
3402 cipher suite and only match that one cipher suite if it is.
3405 *) Link in manifests for VC++ if needed.
3406 [Austin Ziegler <halostatue@gmail.com>]
3408 *) Update support for ECC-based TLS ciphersuites according to
3409 draft-ietf-tls-ecc-12.txt with proposed changes (but without
3410 TLS extensions, which are supported starting with the 0.9.9
3411 branch, not in the OpenSSL 0.9.8 branch).
3414 *) New functions EVP_CIPHER_CTX_new() and EVP_CIPHER_CTX_free() to support
3415 opaque EVP_CIPHER_CTX handling.
3418 *) Fixes and enhancements to zlib compression code. We now only use
3419 "zlib1.dll" and use the default __cdecl calling convention on Win32
3420 to conform with the standards mentioned here:
3421 http://www.zlib.net/DLL_FAQ.txt
3422 Static zlib linking now works on Windows and the new --with-zlib-include
3423 --with-zlib-lib options to Configure can be used to supply the location
3424 of the headers and library. Gracefully handle case where zlib library
3428 *) Several fixes and enhancements to the OID generation code. The old code
3429 sometimes allowed invalid OIDs (1.X for X >= 40 for example), couldn't
3430 handle numbers larger than ULONG_MAX, truncated printing and had a
3431 non standard OBJ_obj2txt() behaviour.
3434 *) Add support for building of engines under engine/ as shared libraries
3435 under VC++ build system.
3438 *) Corrected the numerous bugs in the Win32 path splitter in DSO.
3439 Hopefully, we will not see any false combination of paths any more.
3442 Changes between 0.9.8 and 0.9.8a [11 Oct 2005]
3444 *) Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING
3445 (part of SSL_OP_ALL). This option used to disable the
3446 countermeasure against man-in-the-middle protocol-version
3447 rollback in the SSL 2.0 server implementation, which is a bad
3448 idea. (CVE-2005-2969)
3450 [Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center
3451 for Information Security, National Institute of Advanced Industrial
3452 Science and Technology [AIST], Japan)]
3454 *) Add two function to clear and return the verify parameter flags.
3457 *) Keep cipherlists sorted in the source instead of sorting them at
3458 runtime, thus removing the need for a lock.
3461 *) Avoid some small subgroup attacks in Diffie-Hellman.
3462 [Nick Mathewson and Ben Laurie]
3464 *) Add functions for well-known primes.
3467 *) Extended Windows CE support.
3468 [Satoshi Nakamura and Andy Polyakov]
3470 *) Initialize SSL_METHOD structures at compile time instead of during
3471 runtime, thus removing the need for a lock.
3474 *) Make PKCS7_decrypt() work even if no certificate is supplied by
3475 attempting to decrypt each encrypted key in turn. Add support to
3479 Changes between 0.9.7h and 0.9.8 [05 Jul 2005]
3481 [NB: OpenSSL 0.9.7i and later 0.9.7 patch levels were released after
3484 *) Add libcrypto.pc and libssl.pc for those who feel they need them.
3487 *) Change CA.sh and CA.pl so they don't bundle the CSR and the private
3488 key into the same file any more.
3491 *) Add initial support for Win64, both IA64 and AMD64/x64 flavors.
3494 *) Add -utf8 command line and config file option to 'ca'.
3495 [Stefan <stf@udoma.org]
3497 *) Removed the macro des_crypt(), as it seems to conflict with some
3498 libraries. Use DES_crypt().
3501 *) Correct naming of the 'chil' and '4758cca' ENGINEs. This
3502 involves renaming the source and generated shared-libs for
3503 both. The engines will accept the corrected or legacy ids
3504 ('ncipher' and '4758_cca' respectively) when binding. NB,
3505 this only applies when building 'shared'.
3506 [Corinna Vinschen <vinschen@redhat.com> and Geoff Thorpe]
3508 *) Add attribute functions to EVP_PKEY structure. Modify
3509 PKCS12_create() to recognize a CSP name attribute and
3510 use it. Make -CSP option work again in pkcs12 utility.
3513 *) Add new functionality to the bn blinding code:
3514 - automatic re-creation of the BN_BLINDING parameters after
3515 a fixed number of uses (currently 32)
3516 - add new function for parameter creation
3517 - introduce flags to control the update behaviour of the
3518 BN_BLINDING parameters
3519 - hide BN_BLINDING structure
3520 Add a second BN_BLINDING slot to the RSA structure to improve
3521 performance when a single RSA object is shared among several
3525 *) Add support for DTLS.
3526 [Nagendra Modadugu <nagendra@cs.stanford.edu> and Ben Laurie]
3528 *) Add support for DER encoded private keys (SSL_FILETYPE_ASN1)
3529 to SSL_CTX_use_PrivateKey_file() and SSL_use_PrivateKey_file()
3532 *) Remove buggy and incomplete DH cert support from
3533 ssl/ssl_rsa.c and ssl/s3_both.c
3536 *) Use SHA-1 instead of MD5 as the default digest algorithm for
3537 the apps/openssl applications.
3540 *) Compile clean with "-Wall -Wmissing-prototypes
3541 -Wstrict-prototypes -Wmissing-declarations -Werror". Currently
3542 DEBUG_SAFESTACK must also be set.
3545 *) Change ./Configure so that certain algorithms can be disabled by default.
3546 The new counterpiece to "no-xxx" is "enable-xxx".
3548 The patented RC5 and MDC2 algorithms will now be disabled unless
3549 "enable-rc5" and "enable-mdc2", respectively, are specified.
3551 (IDEA remains enabled despite being patented. This is because IDEA
3552 is frequently required for interoperability, and there is no license
3553 fee for non-commercial use. As before, "no-idea" can be used to
3554 avoid this algorithm.)
3558 *) Add processing of proxy certificates (see RFC 3820). This work was
3559 sponsored by KTH (The Royal Institute of Technology in Stockholm) and
3560 EGEE (Enabling Grids for E-science in Europe).
3563 *) RC4 performance overhaul on modern architectures/implementations, such
3564 as Intel P4, IA-64 and AMD64.
3567 *) New utility extract-section.pl. This can be used specify an alternative
3568 section number in a pod file instead of having to treat each file as
3569 a separate case in Makefile. This can be done by adding two lines to the
3572 =for comment openssl_section:XXX
3574 The blank line is mandatory.
3578 *) New arguments -certform, -keyform and -pass for s_client and s_server
3579 to allow alternative format key and certificate files and passphrase
3583 *) New structure X509_VERIFY_PARAM which combines current verify parameters,
3584 update associated structures and add various utility functions.
3586 Add new policy related verify parameters, include policy checking in
3587 standard verify code. Enhance 'smime' application with extra parameters
3588 to support policy checking and print out.
3591 *) Add a new engine to support VIA PadLock ACE extensions in the VIA C3
3592 Nehemiah processors. These extensions support AES encryption in hardware
3593 as well as RNG (though RNG support is currently disabled).
3594 [Michal Ludvig <michal@logix.cz>, with help from Andy Polyakov]
3596 *) Deprecate BN_[get|set]_params() functions (they were ignored internally).
3599 *) New FIPS 180-2 algorithms, SHA-224/-256/-384/-512 are implemented.
3600 [Andy Polyakov and a number of other people]
3602 *) Improved PowerPC platform support. Most notably BIGNUM assembler
3603 implementation contributed by IBM.
3604 [Suresh Chari, Peter Waltenberg, Andy Polyakov]
3606 *) The new 'RSA_generate_key_ex' function now takes a BIGNUM for the public
3607 exponent rather than 'unsigned long'. There is a corresponding change to
3608 the new 'rsa_keygen' element of the RSA_METHOD structure.
3609 [Jelte Jansen, Geoff Thorpe]
3611 *) Functionality for creating the initial serial number file is now
3612 moved from CA.pl to the 'ca' utility with a new option -create_serial.
3614 (Before OpenSSL 0.9.7e, CA.pl used to initialize the serial
3615 number file to 1, which is bound to cause problems. To avoid
3616 the problems while respecting compatibility between different 0.9.7
3617 patchlevels, 0.9.7e employed 'openssl x509 -next_serial' in
3618 CA.pl for serial number initialization. With the new release 0.9.8,
3619 we can fix the problem directly in the 'ca' utility.)
3622 *) Reduced header interdepencies by declaring more opaque objects in
3623 ossl_typ.h. As a consequence, including some headers (eg. engine.h) will
3624 give fewer recursive includes, which could break lazy source code - so
3625 this change is covered by the OPENSSL_NO_DEPRECATED symbol. As always,
3626 developers should define this symbol when building and using openssl to
3627 ensure they track the recommended behaviour, interfaces, [etc], but
3628 backwards-compatible behaviour prevails when this isn't defined.
3631 *) New function X509_POLICY_NODE_print() which prints out policy nodes.
3634 *) Add new EVP function EVP_CIPHER_CTX_rand_key and associated functionality.
3635 This will generate a random key of the appropriate length based on the
3636 cipher context. The EVP_CIPHER can provide its own random key generation
3637 routine to support keys of a specific form. This is used in the des and
3638 3des routines to generate a key of the correct parity. Update S/MIME
3639 code to use new functions and hence generate correct parity DES keys.
3640 Add EVP_CHECK_DES_KEY #define to return an error if the key is not
3641 valid (weak or incorrect parity).
3644 *) Add a local set of CRLs that can be used by X509_verify_cert() as well
3645 as looking them up. This is useful when the verified structure may contain
3646 CRLs, for example PKCS#7 signedData. Modify PKCS7_verify() to use any CRLs
3647 present unless the new PKCS7_NO_CRL flag is asserted.
3650 *) Extend ASN1 oid configuration module. It now additionally accepts the
3653 shortName = some long name, 1.2.3.4
3656 *) Reimplemented the BN_CTX implementation. There is now no more static
3657 limitation on the number of variables it can handle nor the depth of the
3658 "stack" handling for BN_CTX_start()/BN_CTX_end() pairs. The stack
3659 information can now expand as required, and rather than having a single
3660 static array of bignums, BN_CTX now uses a linked-list of such arrays
3661 allowing it to expand on demand whilst maintaining the usefulness of
3662 BN_CTX's "bundling".
3665 *) Add a missing BN_CTX parameter to the 'rsa_mod_exp' callback in RSA_METHOD
3666 to allow all RSA operations to function using a single BN_CTX.
3669 *) Preliminary support for certificate policy evaluation and checking. This
3670 is initially intended to pass the tests outlined in "Conformance Testing
3671 of Relying Party Client Certificate Path Processing Logic" v1.07.
3674 *) bn_dup_expand() has been deprecated, it was introduced in 0.9.7 and
3675 remained unused and not that useful. A variety of other little bignum
3676 tweaks and fixes have also been made continuing on from the audit (see
3680 *) Constify all or almost all d2i, c2i, s2i and r2i functions, along with
3681 associated ASN1, EVP and SSL functions and old ASN1 macros.
3684 *) BN_zero() only needs to set 'top' and 'neg' to zero for correct results,
3685 and this should never fail. So the return value from the use of
3686 BN_set_word() (which can fail due to needless expansion) is now deprecated;
3687 if OPENSSL_NO_DEPRECATED is defined, BN_zero() is a void macro.
3690 *) BN_CTX_get() should return zero-valued bignums, providing the same
3691 initialised value as BN_new().
3692 [Geoff Thorpe, suggested by Ulf Möller]
3694 *) Support for inhibitAnyPolicy certificate extension.
3697 *) An audit of the BIGNUM code is underway, for which debugging code is
3698 enabled when BN_DEBUG is defined. This makes stricter enforcements on what
3699 is considered valid when processing BIGNUMs, and causes execution to
3700 assert() when a problem is discovered. If BN_DEBUG_RAND is defined,
3701 further steps are taken to deliberately pollute unused data in BIGNUM
3702 structures to try and expose faulty code further on. For now, openssl will
3703 (in its default mode of operation) continue to tolerate the inconsistent
3704 forms that it has tolerated in the past, but authors and packagers should
3705 consider trying openssl and their own applications when compiled with
3706 these debugging symbols defined. It will help highlight potential bugs in
3707 their own code, and will improve the test coverage for OpenSSL itself. At
3708 some point, these tighter rules will become openssl's default to improve
3709 maintainability, though the assert()s and other overheads will remain only
3710 in debugging configurations. See bn.h for more details.
3711 [Geoff Thorpe, Nils Larsch, Ulf Möller]
3713 *) BN_CTX_init() has been deprecated, as BN_CTX is an opaque structure
3714 that can only be obtained through BN_CTX_new() (which implicitly
3715 initialises it). The presence of this function only made it possible
3716 to overwrite an existing structure (and cause memory leaks).
3719 *) Because of the callback-based approach for implementing LHASH as a
3720 template type, lh_insert() adds opaque objects to hash-tables and
3721 lh_doall() or lh_doall_arg() are typically used with a destructor callback
3722 to clean up those corresponding objects before destroying the hash table
3723 (and losing the object pointers). So some over-zealous constifications in
3724 LHASH have been relaxed so that lh_insert() does not take (nor store) the
3725 objects as "const" and the lh_doall[_arg] callback wrappers are not
3726 prototyped to have "const" restrictions on the object pointers they are
3727 given (and so aren't required to cast them away any more).
3730 *) The tmdiff.h API was so ugly and minimal that our own timing utility
3731 (speed) prefers to use its own implementation. The two implementations
3732 haven't been consolidated as yet (volunteers?) but the tmdiff API has had
3733 its object type properly exposed (MS_TM) instead of casting to/from "char
3734 *". This may still change yet if someone realises MS_TM and "ms_time_***"
3735 aren't necessarily the greatest nomenclatures - but this is what was used
3736 internally to the implementation so I've used that for now.
3739 *) Ensure that deprecated functions do not get compiled when
3740 OPENSSL_NO_DEPRECATED is defined. Some "openssl" subcommands and a few of
3741 the self-tests were still using deprecated key-generation functions so
3742 these have been updated also.
3745 *) Reorganise PKCS#7 code to separate the digest location functionality
3746 into PKCS7_find_digest(), digest addition into PKCS7_bio_add_digest().
3747 New function PKCS7_set_digest() to set the digest type for PKCS#7
3748 digestedData type. Add additional code to correctly generate the
3749 digestedData type and add support for this type in PKCS7 initialization
3753 *) New function PKCS7_set0_type_other() this initializes a PKCS7
3754 structure of type "other".
3757 *) Fix prime generation loop in crypto/bn/bn_prime.pl by making
3758 sure the loop does correctly stop and breaking ("division by zero")
3759 modulus operations are not performed. The (pre-generated) prime
3760 table crypto/bn/bn_prime.h was already correct, but it could not be
3761 re-generated on some platforms because of the "division by zero"
3762 situation in the script.
3763 [Ralf S. Engelschall]
3765 *) Update support for ECC-based TLS ciphersuites according to
3766 draft-ietf-tls-ecc-03.txt: the KDF1 key derivation function with
3767 SHA-1 now is only used for "small" curves (where the
3768 representation of a field element takes up to 24 bytes); for
3769 larger curves, the field element resulting from ECDH is directly
3770 used as premaster secret.
3771 [Douglas Stebila (Sun Microsystems Laboratories)]
3773 *) Add code for kP+lQ timings to crypto/ec/ectest.c, and add SEC2
3774 curve secp160r1 to the tests.
3775 [Douglas Stebila (Sun Microsystems Laboratories)]
3777 *) Add the possibility to load symbols globally with DSO.
3778 [Götz Babin-Ebell <babin-ebell@trustcenter.de> via Richard Levitte]
3780 *) Add the functions ERR_set_mark() and ERR_pop_to_mark() for better
3781 control of the error stack.
3784 *) Add support for STORE in ENGINE.
3787 *) Add the STORE type. The intention is to provide a common interface
3788 to certificate and key stores, be they simple file-based stores, or
3789 HSM-type store, or LDAP stores, or...
3790 NOTE: The code is currently UNTESTED and isn't really used anywhere.
3793 *) Add a generic structure called OPENSSL_ITEM. This can be used to
3794 pass a list of arguments to any function as well as provide a way
3795 for a function to pass data back to the caller.
3798 *) Add the functions BUF_strndup() and BUF_memdup(). BUF_strndup()
3799 works like BUF_strdup() but can be used to duplicate a portion of
3800 a string. The copy gets NUL-terminated. BUF_memdup() duplicates
3804 *) Add the function sk_find_ex() which works like sk_find(), but will
3805 return an index to an element even if an exact match couldn't be
3806 found. The index is guaranteed to point at the element where the
3807 searched-for key would be inserted to preserve sorting order.
3810 *) Add the function OBJ_bsearch_ex() which works like OBJ_bsearch() but
3811 takes an extra flags argument for optional functionality. Currently,
3812 the following flags are defined:
3814 OBJ_BSEARCH_VALUE_ON_NOMATCH
3815 This one gets OBJ_bsearch_ex() to return a pointer to the first
3816 element where the comparing function returns a negative or zero
3819 OBJ_BSEARCH_FIRST_VALUE_ON_MATCH
3820 This one gets OBJ_bsearch_ex() to return a pointer to the first
3821 element where the comparing function returns zero. This is useful
3822 if there are more than one element where the comparing function
3826 *) Make it possible to create self-signed certificates with 'openssl ca'
3827 in such a way that the self-signed certificate becomes part of the
3828 CA database and uses the same mechanisms for serial number generation
3829 as all other certificate signing. The new flag '-selfsign' enables
3830 this functionality. Adapt CA.sh and CA.pl.in.
3833 *) Add functionality to check the public key of a certificate request
3834 against a given private. This is useful to check that a certificate
3835 request can be signed by that key (self-signing).
3838 *) Make it possible to have multiple active certificates with the same
3839 subject in the CA index file. This is done only if the keyword
3840 'unique_subject' is set to 'no' in the main CA section (default
3841 if 'CA_default') of the configuration file. The value is saved
3842 with the database itself in a separate index attribute file,
3843 named like the index file with '.attr' appended to the name.
3846 *) Generate muti valued AVAs using '+' notation in config files for
3850 *) Support for nameConstraints certificate extension.
3853 *) Support for policyConstraints certificate extension.
3856 *) Support for policyMappings certificate extension.
3859 *) Make sure the default DSA_METHOD implementation only uses its
3860 dsa_mod_exp() and/or bn_mod_exp() handlers if they are non-NULL,
3861 and change its own handlers to be NULL so as to remove unnecessary
3862 indirection. This lets alternative implementations fallback to the
3863 default implementation more easily.
3866 *) Support for directoryName in GeneralName related extensions
3870 *) Make it possible to link applications using Makefile.shared.
3871 Make that possible even when linking against static libraries!
3874 *) Support for single pass processing for S/MIME signing. This now
3875 means that S/MIME signing can be done from a pipe, in addition
3876 cleartext signing (multipart/signed type) is effectively streaming
3877 and the signed data does not need to be all held in memory.
3879 This is done with a new flag PKCS7_STREAM. When this flag is set
3880 PKCS7_sign() only initializes the PKCS7 structure and the actual signing
3881 is done after the data is output (and digests calculated) in
3882 SMIME_write_PKCS7().
3885 *) Add full support for -rpath/-R, both in shared libraries and
3886 applications, at least on the platforms where it's known how
3890 *) In crypto/ec/ec_mult.c, implement fast point multiplication with
3891 precomputation, based on wNAF splitting: EC_GROUP_precompute_mult()
3892 will now compute a table of multiples of the generator that
3893 makes subsequent invocations of EC_POINTs_mul() or EC_POINT_mul()
3894 faster (notably in the case of a single point multiplication,
3895 scalar * generator).
3896 [Nils Larsch, Bodo Moeller]
3898 *) IPv6 support for certificate extensions. The various extensions
3899 which use the IP:a.b.c.d can now take IPv6 addresses using the
3900 formats of RFC1884 2.2 . IPv6 addresses are now also displayed
3904 *) Added an ENGINE that implements RSA by performing private key
3905 exponentiations with the GMP library. The conversions to and from
3906 GMP's mpz_t format aren't optimised nor are any montgomery forms
3907 cached, and on x86 it appears OpenSSL's own performance has caught up.
3908 However there are likely to be other architectures where GMP could
3909 provide a boost. This ENGINE is not built in by default, but it can be
3910 specified at Configure time and should be accompanied by the necessary
3911 linker additions, eg;
3912 ./config -DOPENSSL_USE_GMP -lgmp
3915 *) "openssl engine" will not display ENGINE/DSO load failure errors when
3916 testing availability of engines with "-t" - the old behaviour is
3917 produced by increasing the feature's verbosity with "-tt".
3920 *) ECDSA routines: under certain error conditions uninitialized BN objects
3921 could be freed. Solution: make sure initialization is performed early
3922 enough. (Reported and fix supplied by Nils Larsch <nla@trustcenter.de>
3926 *) Key-generation can now be implemented in RSA_METHOD, DSA_METHOD
3927 and DH_METHOD (eg. by ENGINE implementations) to override the normal
3928 software implementations. For DSA and DH, parameter generation can
3929 also be overriden by providing the appropriate method callbacks.
3932 *) Change the "progress" mechanism used in key-generation and
3933 primality testing to functions that take a new BN_GENCB pointer in
3934 place of callback/argument pairs. The new API functions have "_ex"
3935 postfixes and the older functions are reimplemented as wrappers for
3936 the new ones. The OPENSSL_NO_DEPRECATED symbol can be used to hide
3937 declarations of the old functions to help (graceful) attempts to
3938 migrate to the new functions. Also, the new key-generation API
3939 functions operate on a caller-supplied key-structure and return
3940 success/failure rather than returning a key or NULL - this is to
3941 help make "keygen" another member function of RSA_METHOD etc.
3943 Example for using the new callback interface:
3945 int (*my_callback)(int a, int b, BN_GENCB *cb) = ...;
3949 BN_GENCB_set(&my_cb, my_callback, my_arg);
3951 return BN_is_prime_ex(some_bignum, BN_prime_checks, NULL, &cb);
3952 /* For the meaning of a, b in calls to my_callback(), see the
3953 * documentation of the function that calls the callback.
3954 * cb will point to my_cb; my_arg can be retrieved as cb->arg.
3955 * my_callback should return 1 if it wants BN_is_prime_ex()
3956 * to continue, or 0 to stop.
3961 *) Change the ZLIB compression method to be stateful, and make it
3962 available to TLS with the number defined in
3963 draft-ietf-tls-compression-04.txt.
3966 *) Add the ASN.1 structures and functions for CertificatePair, which
3967 is defined as follows (according to X.509_4thEditionDraftV6.pdf):
3969 CertificatePair ::= SEQUENCE {
3970 forward [0] Certificate OPTIONAL,
3971 reverse [1] Certificate OPTIONAL,
3972 -- at least one of the pair shall be present -- }
3974 Also implement the PEM functions to read and write certificate
3975 pairs, and defined the PEM tag as "CERTIFICATE PAIR".
3977 This needed to be defined, mostly for the sake of the LDAP
3978 attribute crossCertificatePair, but may prove useful elsewhere as
3982 *) Make it possible to inhibit symlinking of shared libraries in
3983 Makefile.shared, for Cygwin's sake.
3986 *) Extend the BIGNUM API by creating a function
3987 void BN_set_negative(BIGNUM *a, int neg);
3988 and a macro that behave like
3989 int BN_is_negative(const BIGNUM *a);
3991 to avoid the need to access 'a->neg' directly in applications.
3994 *) Implement fast modular reduction for pseudo-Mersenne primes
3995 used in NIST curves (crypto/bn/bn_nist.c, crypto/ec/ecp_nist.c).
3996 EC_GROUP_new_curve_GFp() will now automatically use this
3998 [Nils Larsch <nla@trustcenter.de>]
4000 *) Add new lock type (CRYPTO_LOCK_BN).
4003 *) Change the ENGINE framework to automatically load engines
4004 dynamically from specific directories unless they could be
4005 found to already be built in or loaded. Move all the
4006 current engines except for the cryptodev one to a new
4008 The engines in engines/ are built as shared libraries if
4009 the "shared" options was given to ./Configure or ./config.
4010 Otherwise, they are inserted in libcrypto.a.
4011 /usr/local/ssl/engines is the default directory for dynamic
4012 engines, but that can be overriden at configure time through
4013 the usual use of --prefix and/or --openssldir, and at run
4014 time with the environment variable OPENSSL_ENGINES.
4015 [Geoff Thorpe and Richard Levitte]
4017 *) Add Makefile.shared, a helper makefile to build shared
4018 libraries. Adapt Makefile.org.
4021 *) Add version info to Win32 DLLs.
4022 [Peter 'Luna' Runestig" <peter@runestig.com>]
4024 *) Add new 'medium level' PKCS#12 API. Certificates and keys
4025 can be added using this API to created arbitrary PKCS#12
4026 files while avoiding the low level API.
4028 New options to PKCS12_create(), key or cert can be NULL and
4029 will then be omitted from the output file. The encryption
4030 algorithm NIDs can be set to -1 for no encryption, the mac
4031 iteration count can be set to 0 to omit the mac.
4033 Enhance pkcs12 utility by making the -nokeys and -nocerts
4034 options work when creating a PKCS#12 file. New option -nomac
4035 to omit the mac, NONE can be set for an encryption algorithm.
4036 New code is modified to use the enhanced PKCS12_create()
4037 instead of the low level API.
4040 *) Extend ASN1 encoder to support indefinite length constructed
4041 encoding. This can output sequences tags and octet strings in
4042 this form. Modify pk7_asn1.c to support indefinite length
4043 encoding. This is experimental and needs additional code to
4044 be useful, such as an ASN1 bio and some enhanced streaming
4047 Extend template encode functionality so that tagging is passed
4048 down to the template encoder.
4051 *) Let 'openssl req' fail if an argument to '-newkey' is not
4052 recognized instead of using RSA as a default.
4055 *) Add support for ECC-based ciphersuites from draft-ietf-tls-ecc-01.txt.
4056 As these are not official, they are not included in "ALL";
4057 the "ECCdraft" ciphersuite group alias can be used to select them.
4058 [Vipul Gupta and Sumit Gupta (Sun Microsystems Laboratories)]
4060 *) Add ECDH engine support.
4061 [Nils Gura and Douglas Stebila (Sun Microsystems Laboratories)]
4063 *) Add ECDH in new directory crypto/ecdh/.
4064 [Douglas Stebila (Sun Microsystems Laboratories)]
4066 *) Let BN_rand_range() abort with an error after 100 iterations
4067 without success (which indicates a broken PRNG).
4070 *) Change BN_mod_sqrt() so that it verifies that the input value
4071 is really the square of the return value. (Previously,
4072 BN_mod_sqrt would show GIGO behaviour.)
4075 *) Add named elliptic curves over binary fields from X9.62, SECG,
4076 and WAP/WTLS; add OIDs that were still missing.
4078 [Sheueling Chang Shantz and Douglas Stebila
4079 (Sun Microsystems Laboratories)]
4081 *) Extend the EC library for elliptic curves over binary fields
4082 (new files ec2_smpl.c, ec2_smpt.c, ec2_mult.c in crypto/ec/).
4085 EC_GF2m_simple_method
4089 EC_GROUP_new_curve_GF2m
4090 EC_GROUP_set_curve_GF2m
4091 EC_GROUP_get_curve_GF2m
4092 EC_POINT_set_affine_coordinates_GF2m
4093 EC_POINT_get_affine_coordinates_GF2m
4094 EC_POINT_set_compressed_coordinates_GF2m
4096 Point compression for binary fields is disabled by default for
4097 patent reasons (compile with OPENSSL_EC_BIN_PT_COMP defined to
4100 As binary polynomials are represented as BIGNUMs, various members
4101 of the EC_GROUP and EC_POINT data structures can be shared
4102 between the implementations for prime fields and binary fields;
4103 the above ..._GF2m functions (except for EX_GROUP_new_curve_GF2m)
4104 are essentially identical to their ..._GFp counterparts.
4105 (For simplicity, the '..._GFp' prefix has been dropped from
4106 various internal method names.)
4108 An internal 'field_div' method (similar to 'field_mul' and
4109 'field_sqr') has been added; this is used only for binary fields.
4111 [Sheueling Chang Shantz and Douglas Stebila
4112 (Sun Microsystems Laboratories)]
4114 *) Optionally dispatch EC_POINT_mul(), EC_POINT_precompute_mult()
4115 through methods ('mul', 'precompute_mult').
4117 The generic implementations (now internally called 'ec_wNAF_mul'
4118 and 'ec_wNAF_precomputed_mult') remain the default if these
4119 methods are undefined.
4121 [Sheueling Chang Shantz and Douglas Stebila
4122 (Sun Microsystems Laboratories)]
4124 *) New function EC_GROUP_get_degree, which is defined through
4125 EC_METHOD. For curves over prime fields, this returns the bit
4126 length of the modulus.
4128 [Sheueling Chang Shantz and Douglas Stebila
4129 (Sun Microsystems Laboratories)]
4131 *) New functions EC_GROUP_dup, EC_POINT_dup.
4132 (These simply call ..._new and ..._copy).
4134 [Sheueling Chang Shantz and Douglas Stebila
4135 (Sun Microsystems Laboratories)]
4137 *) Add binary polynomial arithmetic software in crypto/bn/bn_gf2m.c.
4138 Polynomials are represented as BIGNUMs (where the sign bit is not
4139 used) in the following functions [macros]:
4142 BN_GF2m_sub [= BN_GF2m_add]
4143 BN_GF2m_mod [wrapper for BN_GF2m_mod_arr]
4144 BN_GF2m_mod_mul [wrapper for BN_GF2m_mod_mul_arr]
4145 BN_GF2m_mod_sqr [wrapper for BN_GF2m_mod_sqr_arr]
4147 BN_GF2m_mod_exp [wrapper for BN_GF2m_mod_exp_arr]
4148 BN_GF2m_mod_sqrt [wrapper for BN_GF2m_mod_sqrt_arr]
4149 BN_GF2m_mod_solve_quad [wrapper for BN_GF2m_mod_solve_quad_arr]
4150 BN_GF2m_cmp [= BN_ucmp]
4152 (Note that only the 'mod' functions are actually for fields GF(2^m).
4153 BN_GF2m_add() is misnomer, but this is for the sake of consistency.)
4155 For some functions, an the irreducible polynomial defining a
4156 field can be given as an 'unsigned int[]' with strictly
4157 decreasing elements giving the indices of those bits that are set;
4158 i.e., p[] represents the polynomial
4159 f(t) = t^p[0] + t^p[1] + ... + t^p[k]
4161 p[0] > p[1] > ... > p[k] = 0.
4162 This applies to the following functions:
4167 BN_GF2m_mod_inv_arr [wrapper for BN_GF2m_mod_inv]
4168 BN_GF2m_mod_div_arr [wrapper for BN_GF2m_mod_div]
4170 BN_GF2m_mod_sqrt_arr
4171 BN_GF2m_mod_solve_quad_arr
4175 Conversion can be performed by the following functions:
4180 bntest.c has additional tests for binary polynomial arithmetic.
4182 Two implementations for BN_GF2m_mod_div() are available.
4183 The default algorithm simply uses BN_GF2m_mod_inv() and
4184 BN_GF2m_mod_mul(). The alternative algorithm is compiled in only
4185 if OPENSSL_SUN_GF2M_DIV is defined (patent pending; read the
4186 copyright notice in crypto/bn/bn_gf2m.c before enabling it).
4188 [Sheueling Chang Shantz and Douglas Stebila
4189 (Sun Microsystems Laboratories)]
4191 *) Add new error code 'ERR_R_DISABLED' that can be used when some
4192 functionality is disabled at compile-time.
4193 [Douglas Stebila <douglas.stebila@sun.com>]
4195 *) Change default behaviour of 'openssl asn1parse' so that more
4196 information is visible when viewing, e.g., a certificate:
4198 Modify asn1_parse2 (crypto/asn1/asn1_par.c) so that in non-'dump'
4199 mode the content of non-printable OCTET STRINGs is output in a
4200 style similar to INTEGERs, but with '[HEX DUMP]' prepended to
4201 avoid the appearance of a printable string.
4202 [Nils Larsch <nla@trustcenter.de>]
4204 *) Add 'asn1_flag' and 'asn1_form' member to EC_GROUP with access
4206 EC_GROUP_set_asn1_flag()
4207 EC_GROUP_get_asn1_flag()
4208 EC_GROUP_set_point_conversion_form()
4209 EC_GROUP_get_point_conversion_form()
4210 These control ASN1 encoding details:
4211 - Curves (i.e., groups) are encoded explicitly unless asn1_flag
4212 has been set to OPENSSL_EC_NAMED_CURVE.
4213 - Points are encoded in uncompressed form by default; options for
4214 asn1_for are as for point2oct, namely
4215 POINT_CONVERSION_COMPRESSED
4216 POINT_CONVERSION_UNCOMPRESSED
4217 POINT_CONVERSION_HYBRID
4219 Also add 'seed' and 'seed_len' members to EC_GROUP with access
4222 EC_GROUP_get0_seed()
4223 EC_GROUP_get_seed_len()
4224 This is used only for ASN1 purposes (so far).
4225 [Nils Larsch <nla@trustcenter.de>]
4227 *) Add 'field_type' member to EC_METHOD, which holds the NID
4228 of the appropriate field type OID. The new function
4229 EC_METHOD_get_field_type() returns this value.
4230 [Nils Larsch <nla@trustcenter.de>]
4235 EC_POINT_point2hex()
4236 EC_POINT_hex2point()
4237 providing useful interfaces to EC_POINT_point2oct() and
4238 EC_POINT_oct2point().
4239 [Nils Larsch <nla@trustcenter.de>]
4241 *) Change internals of the EC library so that the functions
4242 EC_GROUP_set_generator()
4243 EC_GROUP_get_generator()
4244 EC_GROUP_get_order()
4245 EC_GROUP_get_cofactor()
4246 are implemented directly in crypto/ec/ec_lib.c and not dispatched
4247 to methods, which would lead to unnecessary code duplication when
4248 adding different types of curves.
4249 [Nils Larsch <nla@trustcenter.de> with input by Bodo Moeller]
4251 *) Implement compute_wNAF (crypto/ec/ec_mult.c) without BIGNUM
4252 arithmetic, and such that modified wNAFs are generated
4253 (which avoid length expansion in many cases).
4256 *) Add a function EC_GROUP_check_discriminant() (defined via
4257 EC_METHOD) that verifies that the curve discriminant is non-zero.
4259 Add a function EC_GROUP_check() that makes some sanity tests
4260 on a EC_GROUP, its generator and order. This includes
4261 EC_GROUP_check_discriminant().
4262 [Nils Larsch <nla@trustcenter.de>]
4264 *) Add ECDSA in new directory crypto/ecdsa/.
4266 Add applications 'openssl ecparam' and 'openssl ecdsa'
4267 (these are based on 'openssl dsaparam' and 'openssl dsa').
4269 ECDSA support is also included in various other files across the
4270 library. Most notably,
4271 - 'openssl req' now has a '-newkey ecdsa:file' option;
4272 - EVP_PKCS82PKEY (crypto/evp/evp_pkey.c) now can handle ECDSA;
4273 - X509_PUBKEY_get (crypto/asn1/x_pubkey.c) and
4274 d2i_PublicKey (crypto/asn1/d2i_pu.c) have been modified to make
4275 them suitable for ECDSA where domain parameters must be
4276 extracted before the specific public key;
4277 - ECDSA engine support has been added.
4278 [Nils Larsch <nla@trustcenter.de>]
4280 *) Include some named elliptic curves, and add OIDs from X9.62,
4281 SECG, and WAP/WTLS. Each curve can be obtained from the new
4283 EC_GROUP_new_by_curve_name(),
4284 and the list of available named curves can be obtained with
4285 EC_get_builtin_curves().
4286 Also add a 'curve_name' member to EC_GROUP objects, which can be
4288 EC_GROUP_set_curve_name()
4289 EC_GROUP_get_curve_name()
4290 [Nils Larsch <larsch@trustcenter.de, Bodo Moeller]
4292 *) Remove a few calls to bn_wexpand() in BN_sqr() (the one in there
4293 was actually never needed) and in BN_mul(). The removal in BN_mul()
4294 required a small change in bn_mul_part_recursive() and the addition
4295 of the functions bn_cmp_part_words(), bn_sub_part_words() and
4296 bn_add_part_words(), which do the same thing as bn_cmp_words(),
4297 bn_sub_words() and bn_add_words() except they take arrays with
4301 Changes between 0.9.7l and 0.9.7m [23 Feb 2007]
4303 *) Cleanse PEM buffers before freeing them since they may contain
4305 [Benjamin Bennett <ben@psc.edu>]
4307 *) Include "!eNULL" in SSL_DEFAULT_CIPHER_LIST to make sure that
4308 a ciphersuite string such as "DEFAULT:RSA" cannot enable
4309 authentication-only ciphersuites.
4312 *) Since AES128 and AES256 share a single mask bit in the logic of
4313 ssl/ssl_ciph.c, the code for masking out disabled ciphers needs a
4314 kludge to work properly if AES128 is available and AES256 isn't.
4317 *) Expand security boundary to match 1.1.1 module.
4320 *) Remove redundant features: hash file source, editing of test vectors
4321 modify fipsld to use external fips_premain.c signature.
4324 *) New perl script mkfipsscr.pl to create shell scripts or batch files to
4325 run algorithm test programs.
4328 *) Make algorithm test programs more tolerant of whitespace.
4331 *) Have SSL/TLS server implementation tolerate "mismatched" record
4332 protocol version while receiving ClientHello even if the
4333 ClientHello is fragmented. (The server can't insist on the
4334 particular protocol version it has chosen before the ServerHello
4335 message has informed the client about his choice.)
4338 *) Load error codes if they are not already present instead of using a
4339 static variable. This allows them to be cleanly unloaded and reloaded.
4342 Changes between 0.9.7k and 0.9.7l [28 Sep 2006]
4344 *) Introduce limits to prevent malicious keys being able to
4345 cause a denial of service. (CVE-2006-2940)
4346 [Steve Henson, Bodo Moeller]
4348 *) Fix ASN.1 parsing of certain invalid structures that can result
4349 in a denial of service. (CVE-2006-2937) [Steve Henson]
4351 *) Fix buffer overflow in SSL_get_shared_ciphers() function.
4352 (CVE-2006-3738) [Tavis Ormandy and Will Drewry, Google Security Team]
4354 *) Fix SSL client code which could crash if connecting to a
4355 malicious SSLv2 server. (CVE-2006-4343)
4356 [Tavis Ormandy and Will Drewry, Google Security Team]
4358 *) Change ciphersuite string processing so that an explicit
4359 ciphersuite selects this one ciphersuite (so that "AES256-SHA"
4360 will no longer include "AES128-SHA"), and any other similar
4361 ciphersuite (same bitmap) from *other* protocol versions (so that
4362 "RC4-MD5" will still include both the SSL 2.0 ciphersuite and the
4363 SSL 3.0/TLS 1.0 ciphersuite). This is a backport combining
4364 changes from 0.9.8b and 0.9.8d.
4367 Changes between 0.9.7j and 0.9.7k [05 Sep 2006]
4369 *) Avoid PKCS #1 v1.5 signature attack discovered by Daniel Bleichenbacher
4370 (CVE-2006-4339) [Ben Laurie and Google Security Team]
4372 *) Change the Unix randomness entropy gathering to use poll() when
4373 possible instead of select(), since the latter has some
4374 undesirable limitations.
4375 [Darryl Miles via Richard Levitte and Bodo Moeller]
4377 *) Disable rogue ciphersuites:
4379 - SSLv2 0x08 0x00 0x80 ("RC4-64-MD5")
4380 - SSLv3/TLSv1 0x00 0x61 ("EXP1024-RC2-CBC-MD5")
4381 - SSLv3/TLSv1 0x00 0x60 ("EXP1024-RC4-MD5")
4383 The latter two were purportedly from
4384 draft-ietf-tls-56-bit-ciphersuites-0[01].txt, but do not really
4387 Also deactive the remaining ciphersuites from
4388 draft-ietf-tls-56-bit-ciphersuites-01.txt. These are just as
4389 unofficial, and the ID has long expired.
4392 *) Fix RSA blinding Heisenbug (problems sometimes occured on
4393 dual-core machines) and other potential thread-safety issues.
4396 Changes between 0.9.7i and 0.9.7j [04 May 2006]
4398 *) Adapt fipsld and the build system to link against the validated FIPS
4399 module in FIPS mode.
4402 *) Fixes for VC++ 2005 build under Windows.
4405 *) Add new Windows build target VC-32-GMAKE for VC++. This uses GNU make
4406 from a Windows bash shell such as MSYS. It is autodetected from the
4407 "config" script when run from a VC++ environment. Modify standard VC++
4408 build to use fipscanister.o from the GNU make build.
4411 Changes between 0.9.7h and 0.9.7i [14 Oct 2005]
4413 *) Wrapped the definition of EVP_MAX_MD_SIZE in a #ifdef OPENSSL_FIPS.
4414 The value now differs depending on if you build for FIPS or not.
4415 BEWARE! A program linked with a shared FIPSed libcrypto can't be
4416 safely run with a non-FIPSed libcrypto, as it may crash because of
4417 the difference induced by this change.
4420 Changes between 0.9.7g and 0.9.7h [11 Oct 2005]
4422 *) Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING
4423 (part of SSL_OP_ALL). This option used to disable the
4424 countermeasure against man-in-the-middle protocol-version
4425 rollback in the SSL 2.0 server implementation, which is a bad
4426 idea. (CVE-2005-2969)
4428 [Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center
4429 for Information Security, National Institute of Advanced Industrial
4430 Science and Technology [AIST], Japan)]
4432 *) Minimal support for X9.31 signatures and PSS padding modes. This is
4433 mainly for FIPS compliance and not fully integrated at this stage.
4436 *) For DSA signing, unless DSA_FLAG_NO_EXP_CONSTTIME is set, perform
4437 the exponentiation using a fixed-length exponent. (Otherwise,
4438 the information leaked through timing could expose the secret key
4439 after many signatures; cf. Bleichenbacher's attack on DSA with
4443 *) Make a new fixed-window mod_exp implementation the default for
4444 RSA, DSA, and DH private-key operations so that the sequence of
4445 squares and multiplies and the memory access pattern are
4446 independent of the particular secret key. This will mitigate
4447 cache-timing and potential related attacks.
4449 BN_mod_exp_mont_consttime() is the new exponentiation implementation,
4450 and this is automatically used by BN_mod_exp_mont() if the new flag
4451 BN_FLG_EXP_CONSTTIME is set for the exponent. RSA, DSA, and DH
4452 will use this BN flag for private exponents unless the flag
4453 RSA_FLAG_NO_EXP_CONSTTIME, DSA_FLAG_NO_EXP_CONSTTIME, or
4454 DH_FLAG_NO_EXP_CONSTTIME, respectively, is set.
4456 [Matthew D Wood (Intel Corp), with some changes by Bodo Moeller]
4458 *) Change the client implementation for SSLv23_method() and
4459 SSLv23_client_method() so that is uses the SSL 3.0/TLS 1.0
4460 Client Hello message format if the SSL_OP_NO_SSLv2 option is set.
4461 (Previously, the SSL 2.0 backwards compatible Client Hello
4462 message format would be used even with SSL_OP_NO_SSLv2.)
4465 *) Add support for smime-type MIME parameter in S/MIME messages which some
4469 *) New function BN_MONT_CTX_set_locked() to set montgomery parameters in
4470 a threadsafe manner. Modify rsa code to use new function and add calls
4471 to dsa and dh code (which had race conditions before).
4474 *) Include the fixed error library code in the C error file definitions
4475 instead of fixing them up at runtime. This keeps the error code
4476 structures constant.
4479 Changes between 0.9.7f and 0.9.7g [11 Apr 2005]
4481 [NB: OpenSSL 0.9.7h and later 0.9.7 patch levels were released after
4484 *) Fixes for newer kerberos headers. NB: the casts are needed because
4485 the 'length' field is signed on one version and unsigned on another
4486 with no (?) obvious way to tell the difference, without these VC++
4487 complains. Also the "definition" of FAR (blank) is no longer included
4488 nor is the error ENOMEM. KRB5_PRIVATE has to be set to 1 to pick up
4489 some needed definitions.
4492 *) Undo Cygwin change.
4495 *) Added support for proxy certificates according to RFC 3820.
4496 Because they may be a security thread to unaware applications,
4497 they must be explicitely allowed in run-time. See
4498 docs/HOWTO/proxy_certificates.txt for further information.
4501 Changes between 0.9.7e and 0.9.7f [22 Mar 2005]
4503 *) Use (SSL_RANDOM_VALUE - 4) bytes of pseudo random data when generating
4504 server and client random values. Previously
4505 (SSL_RANDOM_VALUE - sizeof(time_t)) would be used which would result in
4506 less random data when sizeof(time_t) > 4 (some 64 bit platforms).
4508 This change has negligible security impact because:
4510 1. Server and client random values still have 24 bytes of pseudo random
4513 2. Server and client random values are sent in the clear in the initial
4516 3. The master secret is derived using the premaster secret (48 bytes in
4517 size for static RSA ciphersuites) as well as client server and random
4520 The OpenSSL team would like to thank the UK NISCC for bringing this issue
4523 [Stephen Henson, reported by UK NISCC]
4525 *) Use Windows randomness collection on Cygwin.
4528 *) Fix hang in EGD/PRNGD query when communication socket is closed
4529 prematurely by EGD/PRNGD.
4530 [Darren Tucker <dtucker@zip.com.au> via Lutz Jänicke, resolves #1014]
4532 *) Prompt for pass phrases when appropriate for PKCS12 input format.
4535 *) Back-port of selected performance improvements from development
4536 branch, as well as improved support for PowerPC platforms.
4539 *) Add lots of checks for memory allocation failure, error codes to indicate
4540 failure and freeing up memory if a failure occurs.
4541 [Nauticus Networks SSL Team <openssl@nauticusnet.com>, Steve Henson]
4543 *) Add new -passin argument to dgst.
4546 *) Perform some character comparisons of different types in X509_NAME_cmp:
4547 this is needed for some certificates that reencode DNs into UTF8Strings
4548 (in violation of RFC3280) and can't or wont issue name rollover
4552 *) Make an explicit check during certificate validation to see that
4553 the CA setting in each certificate on the chain is correct. As a
4554 side effect always do the following basic checks on extensions,
4555 not just when there's an associated purpose to the check:
4557 - if there is an unhandled critical extension (unless the user
4558 has chosen to ignore this fault)
4559 - if the path length has been exceeded (if one is set at all)
4560 - that certain extensions fit the associated purpose (if one has
4564 Changes between 0.9.7d and 0.9.7e [25 Oct 2004]
4566 *) Avoid a race condition when CRLs are checked in a multi threaded
4567 environment. This would happen due to the reordering of the revoked
4568 entries during signature checking and serial number lookup. Now the
4569 encoding is cached and the serial number sort performed under a lock.
4570 Add new STACK function sk_is_sorted().
4573 *) Add Delta CRL to the extension code.
4576 *) Various fixes to s3_pkt.c so alerts are sent properly.
4577 [David Holmes <d.holmes@f5.com>]
4579 *) Reduce the chances of duplicate issuer name and serial numbers (in
4580 violation of RFC3280) using the OpenSSL certificate creation utilities.
4581 This is done by creating a random 64 bit value for the initial serial
4582 number when a serial number file is created or when a self signed
4583 certificate is created using 'openssl req -x509'. The initial serial
4584 number file is created using 'openssl x509 -next_serial' in CA.pl
4585 rather than being initialized to 1.
4588 Changes between 0.9.7c and 0.9.7d [17 Mar 2004]
4590 *) Fix null-pointer assignment in do_change_cipher_spec() revealed
4591 by using the Codenomicon TLS Test Tool (CVE-2004-0079)
4592 [Joe Orton, Steve Henson]
4594 *) Fix flaw in SSL/TLS handshaking when using Kerberos ciphersuites
4596 [Joe Orton, Steve Henson]
4598 *) Make it possible to have multiple active certificates with the same
4599 subject in the CA index file. This is done only if the keyword
4600 'unique_subject' is set to 'no' in the main CA section (default
4601 if 'CA_default') of the configuration file. The value is saved
4602 with the database itself in a separate index attribute file,
4603 named like the index file with '.attr' appended to the name.
4606 *) X509 verify fixes. Disable broken certificate workarounds when
4607 X509_V_FLAGS_X509_STRICT is set. Check CRL issuer has cRLSign set if
4608 keyUsage extension present. Don't accept CRLs with unhandled critical
4609 extensions: since verify currently doesn't process CRL extensions this
4610 rejects a CRL with *any* critical extensions. Add new verify error codes
4614 *) When creating an OCSP nonce use an OCTET STRING inside the extnValue.
4615 A clarification of RFC2560 will require the use of OCTET STRINGs and
4616 some implementations cannot handle the current raw format. Since OpenSSL
4617 copies and compares OCSP nonces as opaque blobs without any attempt at
4618 parsing them this should not create any compatibility issues.
4621 *) New md flag EVP_MD_CTX_FLAG_REUSE this allows md_data to be reused when
4622 calling EVP_MD_CTX_copy_ex() to avoid calling OPENSSL_malloc(). Without
4623 this HMAC (and other) operations are several times slower than OpenSSL
4627 *) Print out GeneralizedTime and UTCTime in ASN1_STRING_print_ex().
4628 [Peter Sylvester <Peter.Sylvester@EdelWeb.fr>]
4630 *) Use the correct content when signing type "other".
4633 Changes between 0.9.7b and 0.9.7c [30 Sep 2003]
4635 *) Fix various bugs revealed by running the NISCC test suite:
4637 Stop out of bounds reads in the ASN1 code when presented with
4638 invalid tags (CVE-2003-0543 and CVE-2003-0544).
4640 Free up ASN1_TYPE correctly if ANY type is invalid (CVE-2003-0545).
4642 If verify callback ignores invalid public key errors don't try to check
4643 certificate signature with the NULL public key.
4647 *) New -ignore_err option in ocsp application to stop the server
4648 exiting on the first error in a request.
4651 *) In ssl3_accept() (ssl/s3_srvr.c) only accept a client certificate
4652 if the server requested one: as stated in TLS 1.0 and SSL 3.0
4656 *) In ssl3_get_client_hello() (ssl/s3_srvr.c), tolerate additional
4657 extra data after the compression methods not only for TLS 1.0
4658 but also for SSL 3.0 (as required by the specification).
4659 [Bodo Moeller; problem pointed out by Matthias Loepfe]
4661 *) Change X509_certificate_type() to mark the key as exported/exportable
4662 when it's 512 *bits* long, not 512 bytes.
4665 *) Change AES_cbc_encrypt() so it outputs exact multiple of
4666 blocks during encryption.
4669 *) Various fixes to base64 BIO and non blocking I/O. On write
4670 flushes were not handled properly if the BIO retried. On read
4671 data was not being buffered properly and had various logic bugs.
4672 This also affects blocking I/O when the data being decoded is a
4676 *) Various S/MIME bugfixes and compatibility changes:
4677 output correct application/pkcs7 MIME type if
4678 PKCS7_NOOLDMIMETYPE is set. Tolerate some broken signatures.
4679 Output CR+LF for EOL if PKCS7_CRLFEOL is set (this makes opening
4680 of files as .eml work). Correctly handle very long lines in MIME
4684 Changes between 0.9.7a and 0.9.7b [10 Apr 2003]
4686 *) Countermeasure against the Klima-Pokorny-Rosa extension of
4687 Bleichbacher's attack on PKCS #1 v1.5 padding: treat
4688 a protocol version number mismatch like a decryption error
4689 in ssl3_get_client_key_exchange (ssl/s3_srvr.c).
4692 *) Turn on RSA blinding by default in the default implementation
4693 to avoid a timing attack. Applications that don't want it can call
4694 RSA_blinding_off() or use the new flag RSA_FLAG_NO_BLINDING.
4695 They would be ill-advised to do so in most cases.
4696 [Ben Laurie, Steve Henson, Geoff Thorpe, Bodo Moeller]
4698 *) Change RSA blinding code so that it works when the PRNG is not
4699 seeded (in this case, the secret RSA exponent is abused as
4700 an unpredictable seed -- if it is not unpredictable, there
4701 is no point in blinding anyway). Make RSA blinding thread-safe
4702 by remembering the creator's thread ID in rsa->blinding and
4703 having all other threads use local one-time blinding factors
4704 (this requires more computation than sharing rsa->blinding, but
4705 avoids excessive locking; and if an RSA object is not shared
4706 between threads, blinding will still be very fast).
4709 *) Fixed a typo bug that would cause ENGINE_set_default() to set an
4710 ENGINE as defaults for all supported algorithms irrespective of
4711 the 'flags' parameter. 'flags' is now honoured, so applications
4712 should make sure they are passing it correctly.
4715 *) Target "mingw" now allows native Windows code to be generated in
4716 the Cygwin environment as well as with the MinGW compiler.
4719 Changes between 0.9.7 and 0.9.7a [19 Feb 2003]
4721 *) In ssl3_get_record (ssl/s3_pkt.c), minimize information leaked
4722 via timing by performing a MAC computation even if incorrect
4723 block cipher padding has been found. This is a countermeasure
4724 against active attacks where the attacker has to distinguish
4725 between bad padding and a MAC verification error. (CVE-2003-0078)
4727 [Bodo Moeller; problem pointed out by Brice Canvel (EPFL),
4728 Alain Hiltgen (UBS), Serge Vaudenay (EPFL), and
4729 Martin Vuagnoux (EPFL, Ilion)]
4731 *) Make the no-err option work as intended. The intention with no-err
4732 is not to have the whole error stack handling routines removed from
4733 libcrypto, it's only intended to remove all the function name and
4734 reason texts, thereby removing some of the footprint that may not
4735 be interesting if those errors aren't displayed anyway.
4737 NOTE: it's still possible for any application or module to have it's
4738 own set of error texts inserted. The routines are there, just not
4739 used by default when no-err is given.
4742 *) Add support for FreeBSD on IA64.
4743 [dirk.meyer@dinoex.sub.org via Richard Levitte, resolves #454]
4745 *) Adjust DES_cbc_cksum() so it returns the same value as the MIT
4746 Kerberos function mit_des_cbc_cksum(). Before this change,
4747 the value returned by DES_cbc_cksum() was like the one from
4748 mit_des_cbc_cksum(), except the bytes were swapped.
4749 [Kevin Greaney <Kevin.Greaney@hp.com> and Richard Levitte]
4751 *) Allow an application to disable the automatic SSL chain building.
4752 Before this a rather primitive chain build was always performed in
4753 ssl3_output_cert_chain(): an application had no way to send the
4754 correct chain if the automatic operation produced an incorrect result.
4756 Now the chain builder is disabled if either:
4758 1. Extra certificates are added via SSL_CTX_add_extra_chain_cert().
4760 2. The mode flag SSL_MODE_NO_AUTO_CHAIN is set.
4762 The reasoning behind this is that an application would not want the
4763 auto chain building to take place if extra chain certificates are
4764 present and it might also want a means of sending no additional
4765 certificates (for example the chain has two certificates and the
4769 *) Add the possibility to build without the ENGINE framework.
4770 [Steven Reddie <smr@essemer.com.au> via Richard Levitte]
4772 *) Under Win32 gmtime() can return NULL: check return value in
4773 OPENSSL_gmtime(). Add error code for case where gmtime() fails.
4776 *) DSA routines: under certain error conditions uninitialized BN objects
4777 could be freed. Solution: make sure initialization is performed early
4778 enough. (Reported and fix supplied by Ivan D Nestlerode <nestler@MIT.EDU>,
4779 Nils Larsch <nla@trustcenter.de> via PR#459)
4782 *) Another fix for SSLv2 session ID handling: the session ID was incorrectly
4783 checked on reconnect on the client side, therefore session resumption
4784 could still fail with a "ssl session id is different" error. This
4785 behaviour is masked when SSL_OP_ALL is used due to
4786 SSL_OP_MICROSOFT_SESS_ID_BUG being set.
4787 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as
4788 followup to PR #377.
4791 *) IA-32 assembler support enhancements: unified ELF targets, support
4792 for SCO/Caldera platforms, fix for Cygwin shared build.
4795 *) Add support for FreeBSD on sparc64. As a consequence, support for
4796 FreeBSD on non-x86 processors is separate from x86 processors on
4797 the config script, much like the NetBSD support.
4798 [Richard Levitte & Kris Kennaway <kris@obsecurity.org>]
4800 Changes between 0.9.6h and 0.9.7 [31 Dec 2002]
4802 [NB: OpenSSL 0.9.6i and later 0.9.6 patch levels were released after
4805 *) Fix session ID handling in SSLv2 client code: the SERVER FINISHED
4806 code (06) was taken as the first octet of the session ID and the last
4807 octet was ignored consequently. As a result SSLv2 client side session
4808 caching could not have worked due to the session ID mismatch between
4810 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as
4814 *) Change the declaration of needed Kerberos libraries to use EX_LIBS
4815 instead of the special (and badly supported) LIBKRB5. LIBKRB5 is
4819 *) The hw_ncipher.c engine requires dynamic locks. Unfortunately, it
4820 seems that in spite of existing for more than a year, many application
4821 author have done nothing to provide the necessary callbacks, which
4822 means that this particular engine will not work properly anywhere.
4823 This is a very unfortunate situation which forces us, in the name
4824 of usability, to give the hw_ncipher.c a static lock, which is part
4826 NOTE: This is for the 0.9.7 series ONLY. This hack will never
4827 appear in 0.9.8 or later. We EXPECT application authors to have
4828 dealt properly with this when 0.9.8 is released (unless we actually
4829 make such changes in the libcrypto locking code that changes will
4830 have to be made anyway).
4833 *) In asn1_d2i_read_bio() repeatedly call BIO_read() until all content
4834 octets have been read, EOF or an error occurs. Without this change
4835 some truncated ASN1 structures will not produce an error.
4838 *) Disable Heimdal support, since it hasn't been fully implemented.
4839 Still give the possibility to force the use of Heimdal, but with
4840 warnings and a request that patches get sent to openssl-dev.
4843 *) Add the VC-CE target, introduce the WINCE sysname, and add
4844 INSTALL.WCE and appropriate conditionals to make it build.
4845 [Steven Reddie <smr@essemer.com.au> via Richard Levitte]
4847 *) Change the DLL names for Cygwin to cygcrypto-x.y.z.dll and
4848 cygssl-x.y.z.dll, where x, y and z are the major, minor and
4849 edit numbers of the version.
4850 [Corinna Vinschen <vinschen@redhat.com> and Richard Levitte]
4852 *) Introduce safe string copy and catenation functions
4853 (BUF_strlcpy() and BUF_strlcat()).
4854 [Ben Laurie (CHATS) and Richard Levitte]
4856 *) Avoid using fixed-size buffers for one-line DNs.
4857 [Ben Laurie (CHATS)]
4859 *) Add BUF_MEM_grow_clean() to avoid information leakage when
4860 resizing buffers containing secrets, and use where appropriate.
4861 [Ben Laurie (CHATS)]
4863 *) Avoid using fixed size buffers for configuration file location.
4864 [Ben Laurie (CHATS)]
4866 *) Avoid filename truncation for various CA files.
4867 [Ben Laurie (CHATS)]
4869 *) Use sizeof in preference to magic numbers.
4870 [Ben Laurie (CHATS)]
4872 *) Avoid filename truncation in cert requests.
4873 [Ben Laurie (CHATS)]
4875 *) Add assertions to check for (supposedly impossible) buffer
4877 [Ben Laurie (CHATS)]
4879 *) Don't cache truncated DNS entries in the local cache (this could
4880 potentially lead to a spoofing attack).
4881 [Ben Laurie (CHATS)]
4883 *) Fix various buffers to be large enough for hex/decimal
4884 representations in a platform independent manner.
4885 [Ben Laurie (CHATS)]
4887 *) Add CRYPTO_realloc_clean() to avoid information leakage when
4888 resizing buffers containing secrets, and use where appropriate.
4889 [Ben Laurie (CHATS)]
4891 *) Add BIO_indent() to avoid much slightly worrying code to do
4893 [Ben Laurie (CHATS)]
4895 *) Convert sprintf()/BIO_puts() to BIO_printf().
4896 [Ben Laurie (CHATS)]
4898 *) buffer_gets() could terminate with the buffer only half
4900 [Ben Laurie (CHATS)]
4902 *) Add assertions to prevent user-supplied crypto functions from
4903 overflowing internal buffers by having large block sizes, etc.
4904 [Ben Laurie (CHATS)]
4906 *) New OPENSSL_assert() macro (similar to assert(), but enabled
4908 [Ben Laurie (CHATS)]
4910 *) Eliminate unused copy of key in RC4.
4911 [Ben Laurie (CHATS)]
4913 *) Eliminate unused and incorrectly sized buffers for IV in pem.h.
4914 [Ben Laurie (CHATS)]
4916 *) Fix off-by-one error in EGD path.
4917 [Ben Laurie (CHATS)]
4919 *) If RANDFILE path is too long, ignore instead of truncating.
4920 [Ben Laurie (CHATS)]
4922 *) Eliminate unused and incorrectly sized X.509 structure
4924 [Ben Laurie (CHATS)]
4926 *) Eliminate unused and dangerous function knumber().
4927 [Ben Laurie (CHATS)]
4929 *) Eliminate unused and dangerous structure, KSSL_ERR.
4930 [Ben Laurie (CHATS)]
4932 *) Protect against overlong session ID context length in an encoded
4933 session object. Since these are local, this does not appear to be
4935 [Ben Laurie (CHATS)]
4937 *) Change from security patch (see 0.9.6e below) that did not affect
4938 the 0.9.6 release series:
4940 Remote buffer overflow in SSL3 protocol - an attacker could
4941 supply an oversized master key in Kerberos-enabled versions.
4943 [Ben Laurie (CHATS)]
4945 *) Change the SSL kerb5 codes to match RFC 2712.
4948 *) Make -nameopt work fully for req and add -reqopt switch.
4949 [Michael Bell <michael.bell@rz.hu-berlin.de>, Steve Henson]
4951 *) The "block size" for block ciphers in CFB and OFB mode should be 1.
4952 [Steve Henson, reported by Yngve Nysaeter Pettersen <yngve@opera.com>]
4954 *) Make sure tests can be performed even if the corresponding algorithms
4955 have been removed entirely. This was also the last step to make
4956 OpenSSL compilable with DJGPP under all reasonable conditions.
4957 [Richard Levitte, Doug Kaufman <dkaufman@rahul.net>]
4959 *) Add cipher selection rules COMPLEMENTOFALL and COMPLEMENTOFDEFAULT
4960 to allow version independent disabling of normally unselected ciphers,
4961 which may be activated as a side-effect of selecting a single cipher.
4963 (E.g., cipher list string "RSA" enables ciphersuites that are left
4964 out of "ALL" because they do not provide symmetric encryption.
4965 "RSA:!COMPLEMEMENTOFALL" avoids these unsafe ciphersuites.)
4966 [Lutz Jaenicke, Bodo Moeller]
4968 *) Add appropriate support for separate platform-dependent build
4969 directories. The recommended way to make a platform-dependent
4970 build directory is the following (tested on Linux), maybe with
4973 # Place yourself outside of the OpenSSL source tree. In
4974 # this example, the environment variable OPENSSL_SOURCE
4975 # is assumed to contain the absolute OpenSSL source directory.
4976 mkdir -p objtree/"`uname -s`-`uname -r`-`uname -m`"
4977 cd objtree/"`uname -s`-`uname -r`-`uname -m`"
4978 (cd $OPENSSL_SOURCE; find . -type f) | while read F; do
4979 mkdir -p `dirname $F`
4980 ln -s $OPENSSL_SOURCE/$F $F
4983 To be absolutely sure not to disturb the source tree, a "make clean"
4984 is a good thing. If it isn't successfull, don't worry about it,
4985 it probably means the source directory is very clean.
4988 *) Make sure any ENGINE control commands make local copies of string
4989 pointers passed to them whenever necessary. Otherwise it is possible
4990 the caller may have overwritten (or deallocated) the original string
4991 data when a later ENGINE operation tries to use the stored values.
4992 [Götz Babin-Ebell <babinebell@trustcenter.de>]
4994 *) Improve diagnostics in file reading and command-line digests.
4995 [Ben Laurie aided and abetted by Solar Designer <solar@openwall.com>]
4997 *) Add AES modes CFB and OFB to the object database. Correct an
4998 error in AES-CFB decryption.
5001 *) Remove most calls to EVP_CIPHER_CTX_cleanup() in evp_enc.c, this
5002 allows existing EVP_CIPHER_CTX structures to be reused after
5003 calling EVP_*Final(). This behaviour is used by encryption
5004 BIOs and some applications. This has the side effect that
5005 applications must explicitly clean up cipher contexts with
5006 EVP_CIPHER_CTX_cleanup() or they will leak memory.
5009 *) Check the values of dna and dnb in bn_mul_recursive before calling
5010 bn_mul_comba (a non zero value means the a or b arrays do not contain
5011 n2 elements) and fallback to bn_mul_normal if either is not zero.
5014 *) Fix escaping of non-ASCII characters when using the -subj option
5015 of the "openssl req" command line tool. (Robert Joop <joop@fokus.gmd.de>)
5018 *) Make object definitions compliant to LDAP (RFC2256): SN is the short
5019 form for "surname", serialNumber has no short form.
5020 Use "mail" as the short name for "rfc822Mailbox" according to RFC2798;
5021 therefore remove "mail" short name for "internet 7".
5022 The OID for unique identifiers in X509 certificates is
5023 x500UniqueIdentifier, not uniqueIdentifier.
5024 Some more OID additions. (Michael Bell <michael.bell@rz.hu-berlin.de>)
5027 *) Add an "init" command to the ENGINE config module and auto initialize
5028 ENGINEs. Without any "init" command the ENGINE will be initialized
5029 after all ctrl commands have been executed on it. If init=1 the
5030 ENGINE is initailized at that point (ctrls before that point are run
5031 on the uninitialized ENGINE and after on the initialized one). If
5032 init=0 then the ENGINE will not be iniatialized at all.
5035 *) Fix the 'app_verify_callback' interface so that the user-defined
5036 argument is actually passed to the callback: In the
5037 SSL_CTX_set_cert_verify_callback() prototype, the callback
5038 declaration has been changed from
5041 int (*cb)(X509_STORE_CTX *,void *);
5042 in ssl_verify_cert_chain (ssl/ssl_cert.c), the call
5043 i=s->ctx->app_verify_callback(&ctx)
5044 has been changed into
5045 i=s->ctx->app_verify_callback(&ctx, s->ctx->app_verify_arg).
5047 To update applications using SSL_CTX_set_cert_verify_callback(),
5048 a dummy argument can be added to their callback functions.
5049 [D. K. Smetters <smetters@parc.xerox.com>]
5051 *) Added the '4758cca' ENGINE to support IBM 4758 cards.
5052 [Maurice Gittens <maurice@gittens.nl>, touchups by Geoff Thorpe]
5054 *) Add and OPENSSL_LOAD_CONF define which will cause
5055 OpenSSL_add_all_algorithms() to load the openssl.cnf config file.
5056 This allows older applications to transparently support certain
5057 OpenSSL features: such as crypto acceleration and dynamic ENGINE loading.
5058 Two new functions OPENSSL_add_all_algorithms_noconf() which will never
5059 load the config file and OPENSSL_add_all_algorithms_conf() which will
5060 always load it have also been added.
5063 *) Add the OFB, CFB and CTR (all with 128 bit feedback) to AES.
5064 Adjust NIDs and EVP layer.
5065 [Stephen Sprunk <stephen@sprunk.org> and Richard Levitte]
5067 *) Config modules support in openssl utility.
5069 Most commands now load modules from the config file,
5070 though in a few (such as version) this isn't done
5071 because it couldn't be used for anything.
5073 In the case of ca and req the config file used is
5074 the same as the utility itself: that is the -config
5075 command line option can be used to specify an
5079 *) Move default behaviour from OPENSSL_config(). If appname is NULL
5080 use "openssl_conf" if filename is NULL use default openssl config file.
5083 *) Add an argument to OPENSSL_config() to allow the use of an alternative
5084 config section name. Add a new flag to tolerate a missing config file
5085 and move code to CONF_modules_load_file().
5088 *) Support for crypto accelerator cards from Accelerated Encryption
5089 Processing, www.aep.ie. (Use engine 'aep')
5090 The support was copied from 0.9.6c [engine] and adapted/corrected
5091 to work with the new engine framework.
5092 [AEP Inc. and Richard Levitte]
5094 *) Support for SureWare crypto accelerator cards from Baltimore
5095 Technologies. (Use engine 'sureware')
5096 The support was copied from 0.9.6c [engine] and adapted
5097 to work with the new engine framework.
5100 *) Have the CHIL engine fork-safe (as defined by nCipher) and actually
5101 make the newer ENGINE framework commands for the CHIL engine work.
5102 [Toomas Kiisk <vix@cyber.ee> and Richard Levitte]
5104 *) Make it possible to produce shared libraries on ReliantUNIX.
5105 [Robert Dahlem <Robert.Dahlem@ffm2.siemens.de> via Richard Levitte]
5107 *) Add the configuration target debug-linux-ppro.
5108 Make 'openssl rsa' use the general key loading routines
5109 implemented in apps.c, and make those routines able to
5110 handle the key format FORMAT_NETSCAPE and the variant
5112 [Toomas Kiisk <vix@cyber.ee> via Richard Levitte]
5114 *) Fix a crashbug and a logic bug in hwcrhk_load_pubkey().
5115 [Toomas Kiisk <vix@cyber.ee> via Richard Levitte]
5117 *) Add -keyform to rsautl, and document -engine.
5118 [Richard Levitte, inspired by Toomas Kiisk <vix@cyber.ee>]
5120 *) Change BIO_new_file (crypto/bio/bss_file.c) to use new
5121 BIO_R_NO_SUCH_FILE error code rather than the generic
5122 ERR_R_SYS_LIB error code if fopen() fails with ENOENT.
5125 *) Add new functions
5127 ERR_peek_last_error_line
5128 ERR_peek_last_error_line_data.
5129 These are similar to
5132 ERR_peek_error_line_data,
5133 but report on the latest error recorded rather than the first one
5134 still in the error queue.
5135 [Ben Laurie, Bodo Moeller]
5137 *) default_algorithms option in ENGINE config module. This allows things
5139 default_algorithms = ALL
5140 default_algorithms = RSA, DSA, RAND, CIPHERS, DIGESTS
5143 *) Preliminary ENGINE config module.
5146 *) New experimental application configuration code.
5149 *) Change the AES code to follow the same name structure as all other
5150 symmetric ciphers, and behave the same way. Move everything to
5151 the directory crypto/aes, thereby obsoleting crypto/rijndael.
5152 [Stephen Sprunk <stephen@sprunk.org> and Richard Levitte]
5154 *) SECURITY: remove unsafe setjmp/signal interaction from ui_openssl.c.
5155 [Ben Laurie and Theo de Raadt]
5157 *) Add option to output public keys in req command.
5158 [Massimiliano Pala madwolf@openca.org]
5160 *) Use wNAFs in EC_POINTs_mul() for improved efficiency
5161 (up to about 10% better than before for P-192 and P-224).
5164 *) New functions/macros
5166 SSL_CTX_set_msg_callback(ctx, cb)
5167 SSL_CTX_set_msg_callback_arg(ctx, arg)
5168 SSL_set_msg_callback(ssl, cb)
5169 SSL_set_msg_callback_arg(ssl, arg)
5171 to request calling a callback function
5173 void cb(int write_p, int version, int content_type,
5174 const void *buf, size_t len, SSL *ssl, void *arg)
5176 whenever a protocol message has been completely received
5177 (write_p == 0) or sent (write_p == 1). Here 'version' is the
5178 protocol version according to which the SSL library interprets
5179 the current protocol message (SSL2_VERSION, SSL3_VERSION, or
5180 TLS1_VERSION). 'content_type' is 0 in the case of SSL 2.0, or
5181 the content type as defined in the SSL 3.0/TLS 1.0 protocol
5182 specification (change_cipher_spec(20), alert(21), handshake(22)).
5183 'buf' and 'len' point to the actual message, 'ssl' to the
5184 SSL object, and 'arg' is the application-defined value set by
5185 SSL[_CTX]_set_msg_callback_arg().
5187 'openssl s_client' and 'openssl s_server' have new '-msg' options
5188 to enable a callback that displays all protocol messages.
5191 *) Change the shared library support so shared libraries are built as
5192 soon as the corresponding static library is finished, and thereby get
5193 openssl and the test programs linked against the shared library.
5194 This still only happens when the keyword "shard" has been given to
5195 the configuration scripts.
5197 NOTE: shared library support is still an experimental thing, and
5198 backward binary compatibility is still not guaranteed.
5199 ["Maciej W. Rozycki" <macro@ds2.pg.gda.pl> and Richard Levitte]
5201 *) Add support for Subject Information Access extension.
5202 [Peter Sylvester <Peter.Sylvester@EdelWeb.fr>]
5204 *) Make BUF_MEM_grow() behaviour more consistent: Initialise to zero
5205 additional bytes when new memory had to be allocated, not just
5206 when reusing an existing buffer.
5209 *) New command line and configuration option 'utf8' for the req command.
5210 This allows field values to be specified as UTF8 strings.
5213 *) Add -multi and -mr options to "openssl speed" - giving multiple parallel
5214 runs for the former and machine-readable output for the latter.
5217 *) Add '-noemailDN' option to 'openssl ca'. This prevents inclusion
5218 of the e-mail address in the DN (i.e., it will go into a certificate
5219 extension only). The new configuration file option 'email_in_dn = no'
5220 has the same effect.
5221 [Massimiliano Pala madwolf@openca.org]
5223 *) Change all functions with names starting with des_ to be starting
5224 with DES_ instead. Add wrappers that are compatible with libdes,
5225 but are named _ossl_old_des_*. Finally, add macros that map the
5226 des_* symbols to the corresponding _ossl_old_des_* if libdes
5227 compatibility is desired. If OpenSSL 0.9.6c compatibility is
5228 desired, the des_* symbols will be mapped to DES_*, with one
5231 Since we provide two compatibility mappings, the user needs to
5232 define the macro OPENSSL_DES_LIBDES_COMPATIBILITY if libdes
5233 compatibility is desired. The default (i.e., when that macro
5234 isn't defined) is OpenSSL 0.9.6c compatibility.
5236 There are also macros that enable and disable the support of old
5237 des functions altogether. Those are OPENSSL_ENABLE_OLD_DES_SUPPORT
5238 and OPENSSL_DISABLE_OLD_DES_SUPPORT. If none or both of those
5239 are defined, the default will apply: to support the old des routines.
5241 In either case, one must include openssl/des.h to get the correct
5242 definitions. Do not try to just include openssl/des_old.h, that
5245 NOTE: This is a major break of an old API into a new one. Software
5246 authors are encouraged to switch to the DES_ style functions. Some
5247 time in the future, des_old.h and the libdes compatibility functions
5248 will be disable (i.e. OPENSSL_DISABLE_OLD_DES_SUPPORT will be the
5249 default), and then completely removed.
5252 *) Test for certificates which contain unsupported critical extensions.
5253 If such a certificate is found during a verify operation it is
5254 rejected by default: this behaviour can be overridden by either
5255 handling the new error X509_V_ERR_UNHANDLED_CRITICAL_EXTENSION or
5256 by setting the verify flag X509_V_FLAG_IGNORE_CRITICAL. A new function
5257 X509_supported_extension() has also been added which returns 1 if a
5258 particular extension is supported.
5261 *) Modify the behaviour of EVP cipher functions in similar way to digests
5262 to retain compatibility with existing code.
5265 *) Modify the behaviour of EVP_DigestInit() and EVP_DigestFinal() to retain
5266 compatibility with existing code. In particular the 'ctx' parameter does
5267 not have to be to be initialized before the call to EVP_DigestInit() and
5268 it is tidied up after a call to EVP_DigestFinal(). New function
5269 EVP_DigestFinal_ex() which does not tidy up the ctx. Similarly function
5270 EVP_MD_CTX_copy() changed to not require the destination to be
5271 initialized valid and new function EVP_MD_CTX_copy_ex() added which
5272 requires the destination to be valid.
5274 Modify all the OpenSSL digest calls to use EVP_DigestInit_ex(),
5275 EVP_DigestFinal_ex() and EVP_MD_CTX_copy_ex().
5278 *) Change ssl3_get_message (ssl/s3_both.c) and the functions using it
5279 so that complete 'Handshake' protocol structures are kept in memory
5280 instead of overwriting 'msg_type' and 'length' with 'body' data.
5283 *) Add an implementation of SSL_add_dir_cert_subjects_to_stack for Win32.
5284 [Massimo Santin via Richard Levitte]
5286 *) Major restructuring to the underlying ENGINE code. This includes
5287 reduction of linker bloat, separation of pure "ENGINE" manipulation
5288 (initialisation, etc) from functionality dealing with implementations
5289 of specific crypto iterfaces. This change also introduces integrated
5290 support for symmetric ciphers and digest implementations - so ENGINEs
5291 can now accelerate these by providing EVP_CIPHER and EVP_MD
5292 implementations of their own. This is detailed in crypto/engine/README
5293 as it couldn't be adequately described here. However, there are a few
5294 API changes worth noting - some RSA, DSA, DH, and RAND functions that
5295 were changed in the original introduction of ENGINE code have now
5296 reverted back - the hooking from this code to ENGINE is now a good
5297 deal more passive and at run-time, operations deal directly with
5298 RSA_METHODs, DSA_METHODs (etc) as they did before, rather than
5299 dereferencing through an ENGINE pointer any more. Also, the ENGINE
5300 functions dealing with BN_MOD_EXP[_CRT] handlers have been removed -
5301 they were not being used by the framework as there is no concept of a
5302 BIGNUM_METHOD and they could not be generalised to the new
5303 'ENGINE_TABLE' mechanism that underlies the new code. Similarly,
5304 ENGINE_cpy() has been removed as it cannot be consistently defined in
5308 *) Change ASN1_GENERALIZEDTIME_check() to allow fractional seconds.
5311 *) Change mkdef.pl to sort symbols that get the same entry number,
5312 and make sure the automatically generated functions ERR_load_*
5313 become part of libeay.num as well.
5316 *) New function SSL_renegotiate_pending(). This returns true once
5317 renegotiation has been requested (either SSL_renegotiate() call
5318 or HelloRequest/ClientHello received from the peer) and becomes
5319 false once a handshake has been completed.
5320 (For servers, SSL_renegotiate() followed by SSL_do_handshake()
5321 sends a HelloRequest, but does not ensure that a handshake takes
5322 place. SSL_renegotiate_pending() is useful for checking if the
5323 client has followed the request.)
5326 *) New SSL option SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION.
5327 By default, clients may request session resumption even during
5328 renegotiation (if session ID contexts permit); with this option,
5329 session resumption is possible only in the first handshake.
5331 SSL_OP_ALL is now 0x00000FFFL instead of 0x000FFFFFL. This makes
5332 more bits available for options that should not be part of
5333 SSL_OP_ALL (such as SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION).
5336 *) Add some demos for certificate and certificate request creation.
5339 *) Make maximum certificate chain size accepted from the peer application
5340 settable (SSL*_get/set_max_cert_list()), as proposed by
5341 "Douglas E. Engert" <deengert@anl.gov>.
5344 *) Add support for shared libraries for Unixware-7
5345 (Boyd Lynn Gerber <gerberb@zenez.com>).
5348 *) Add a "destroy" handler to ENGINEs that allows structural cleanup to
5349 be done prior to destruction. Use this to unload error strings from
5350 ENGINEs that load their own error strings. NB: This adds two new API
5351 functions to "get" and "set" this destroy handler in an ENGINE.
5354 *) Alter all existing ENGINE implementations (except "openssl" and
5355 "openbsd") to dynamically instantiate their own error strings. This
5356 makes them more flexible to be built both as statically-linked ENGINEs
5357 and self-contained shared-libraries loadable via the "dynamic" ENGINE.
5358 Also, add stub code to each that makes building them as self-contained
5359 shared-libraries easier (see README.ENGINE).
5362 *) Add a "dynamic" ENGINE that provides a mechanism for binding ENGINE
5363 implementations into applications that are completely implemented in
5364 self-contained shared-libraries. The "dynamic" ENGINE exposes control
5365 commands that can be used to configure what shared-library to load and
5366 to control aspects of the way it is handled. Also, made an update to
5367 the README.ENGINE file that brings its information up-to-date and
5368 provides some information and instructions on the "dynamic" ENGINE
5369 (ie. how to use it, how to build "dynamic"-loadable ENGINEs, etc).
5372 *) Make it possible to unload ranges of ERR strings with a new
5373 "ERR_unload_strings" function.
5376 *) Add a copy() function to EVP_MD.
5379 *) Make EVP_MD routines take a context pointer instead of just the
5380 md_data void pointer.
5383 *) Add flags to EVP_MD and EVP_MD_CTX. EVP_MD_FLAG_ONESHOT indicates
5384 that the digest can only process a single chunk of data
5385 (typically because it is provided by a piece of
5386 hardware). EVP_MD_CTX_FLAG_ONESHOT indicates that the application
5387 is only going to provide a single chunk of data, and hence the
5388 framework needn't accumulate the data for oneshot drivers.
5391 *) As with "ERR", make it possible to replace the underlying "ex_data"
5392 functions. This change also alters the storage and management of global
5393 ex_data state - it's now all inside ex_data.c and all "class" code (eg.
5394 RSA, BIO, SSL_CTX, etc) no longer stores its own STACKS and per-class
5395 index counters. The API functions that use this state have been changed
5396 to take a "class_index" rather than pointers to the class's local STACK
5397 and counter, and there is now an API function to dynamically create new
5398 classes. This centralisation allows us to (a) plug a lot of the
5399 thread-safety problems that existed, and (b) makes it possible to clean
5400 up all allocated state using "CRYPTO_cleanup_all_ex_data()". W.r.t. (b)
5401 such data would previously have always leaked in application code and
5402 workarounds were in place to make the memory debugging turn a blind eye
5403 to it. Application code that doesn't use this new function will still
5404 leak as before, but their memory debugging output will announce it now
5405 rather than letting it slide.
5407 Besides the addition of CRYPTO_cleanup_all_ex_data(), another API change
5408 induced by the "ex_data" overhaul is that X509_STORE_CTX_init() now
5409 has a return value to indicate success or failure.
5412 *) Make it possible to replace the underlying "ERR" functions such that the
5413 global state (2 LHASH tables and 2 locks) is only used by the "default"
5414 implementation. This change also adds two functions to "get" and "set"
5415 the implementation prior to it being automatically set the first time
5416 any other ERR function takes place. Ie. an application can call "get",
5417 pass the return value to a module it has just loaded, and that module
5418 can call its own "set" function using that value. This means the
5419 module's "ERR" operations will use (and modify) the error state in the
5420 application and not in its own statically linked copy of OpenSSL code.
5423 *) Give DH, DSA, and RSA types their own "**_up_ref()" function to increment
5424 reference counts. This performs normal REF_PRINT/REF_CHECK macros on
5425 the operation, and provides a more encapsulated way for external code
5426 (crypto/evp/ and ssl/) to do this. Also changed the evp and ssl code
5427 to use these functions rather than manually incrementing the counts.
5429 Also rename "DSO_up()" function to more descriptive "DSO_up_ref()".
5432 *) Add EVP test program.
5435 *) Add symmetric cipher support to ENGINE. Expect the API to change!
5438 *) New CRL functions: X509_CRL_set_version(), X509_CRL_set_issuer_name()
5439 X509_CRL_set_lastUpdate(), X509_CRL_set_nextUpdate(), X509_CRL_sort(),
5440 X509_REVOKED_set_serialNumber(), and X509_REVOKED_set_revocationDate().
5441 These allow a CRL to be built without having to access X509_CRL fields
5442 directly. Modify 'ca' application to use new functions.
5445 *) Move SSL_OP_TLS_ROLLBACK_BUG out of the SSL_OP_ALL list of recommended
5446 bug workarounds. Rollback attack detection is a security feature.
5447 The problem will only arise on OpenSSL servers when TLSv1 is not
5448 available (sslv3_server_method() or SSL_OP_NO_TLSv1).
5449 Software authors not wanting to support TLSv1 will have special reasons
5450 for their choice and can explicitly enable this option.
5451 [Bodo Moeller, Lutz Jaenicke]
5453 *) Rationalise EVP so it can be extended: don't include a union of
5454 cipher/digest structures, add init/cleanup functions for EVP_MD_CTX
5455 (similar to those existing for EVP_CIPHER_CTX).
5460 EVP_MD_CTX_init(&md); /* new function call */
5461 EVP_DigestInit(&md, EVP_sha1());
5462 EVP_DigestUpdate(&md, in, len);
5463 EVP_DigestFinal(&md, out, NULL);
5464 EVP_MD_CTX_cleanup(&md); /* new function call */
5468 *) Make DES key schedule conform to the usual scheme, as well as
5469 correcting its structure. This means that calls to DES functions
5470 now have to pass a pointer to a des_key_schedule instead of a
5471 plain des_key_schedule (which was actually always a pointer
5474 des_key_schedule ks;
5476 des_set_key_checked(..., &ks);
5477 des_ncbc_encrypt(..., &ks, ...);
5479 (Note that a later change renames 'des_...' into 'DES_...'.)
5482 *) Initial reduction of linker bloat: the use of some functions, such as
5483 PEM causes large amounts of unused functions to be linked in due to
5484 poor organisation. For example pem_all.c contains every PEM function
5485 which has a knock on effect of linking in large amounts of (unused)
5486 ASN1 code. Grouping together similar functions and splitting unrelated
5487 functions prevents this.
5490 *) Cleanup of EVP macros.
5493 *) Change historical references to {NID,SN,LN}_des_ede and ede3 to add the
5494 correct _ecb suffix.
5497 *) Add initial OCSP responder support to ocsp application. The
5498 revocation information is handled using the text based index
5499 use by the ca application. The responder can either handle
5500 requests generated internally, supplied in files (for example
5501 via a CGI script) or using an internal minimal server.
5504 *) Add configuration choices to get zlib compression for TLS.
5507 *) Changes to Kerberos SSL for RFC 2712 compliance:
5508 1. Implemented real KerberosWrapper, instead of just using
5509 KRB5 AP_REQ message. [Thanks to Simon Wilkinson <sxw@sxw.org.uk>]
5510 2. Implemented optional authenticator field of KerberosWrapper.
5512 Added openssl-style ASN.1 macros for Kerberos ticket, ap_req,
5513 and authenticator structs; see crypto/krb5/.
5515 Generalized Kerberos calls to support multiple Kerberos libraries.
5516 [Vern Staats <staatsvr@asc.hpc.mil>,
5517 Jeffrey Altman <jaltman@columbia.edu>
5518 via Richard Levitte]
5520 *) Cause 'openssl speed' to use fully hard-coded DSA keys as it
5521 already does with RSA. testdsa.h now has 'priv_key/pub_key'
5522 values for each of the key sizes rather than having just
5523 parameters (and 'speed' generating keys each time).
5526 *) Speed up EVP routines.
5529 type 8 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes
5530 des-cbc 4408.85k 5560.51k 5778.46k 5862.20k 5825.16k
5531 des-cbc 4389.55k 5571.17k 5792.23k 5846.91k 5832.11k
5532 des-cbc 4394.32k 5575.92k 5807.44k 5848.37k 5841.30k
5534 des-cbc 3482.66k 5069.49k 5496.39k 5614.16k 5639.28k
5535 des-cbc 3480.74k 5068.76k 5510.34k 5609.87k 5635.52k
5536 des-cbc 3483.72k 5067.62k 5504.60k 5708.01k 5724.80k
5539 des-cbc 4660.16k 5650.19k 5807.19k 5827.13k 5783.32k
5541 des-cbc 3624.96k 5258.21k 5530.91k 5624.30k 5628.26k
5544 *) Added the OS2-EMX target.
5545 ["Brian Havard" <brianh@kheldar.apana.org.au> and Richard Levitte]
5547 *) Rewrite apps to use NCONF routines instead of the old CONF. New functions
5548 to support NCONF routines in extension code. New function CONF_set_nconf()
5549 to allow functions which take an NCONF to also handle the old LHASH
5550 structure: this means that the old CONF compatible routines can be
5551 retained (in particular wrt extensions) without having to duplicate the
5552 code. New function X509V3_add_ext_nconf_sk to add extensions to a stack.
5555 *) Enhance the general user interface with mechanisms for inner control
5556 and with possibilities to have yes/no kind of prompts.
5559 *) Change all calls to low level digest routines in the library and
5560 applications to use EVP. Add missing calls to HMAC_cleanup() and
5561 don't assume HMAC_CTX can be copied using memcpy().
5562 [Verdon Walker <VWalker@novell.com>, Steve Henson]
5564 *) Add the possibility to control engines through control names but with
5565 arbitrary arguments instead of just a string.
5566 Change the key loaders to take a UI_METHOD instead of a callback
5567 function pointer. NOTE: this breaks binary compatibility with earlier
5568 versions of OpenSSL [engine].
5569 Adapt the nCipher code for these new conditions and add a card insertion
5573 *) Enhance the general user interface with mechanisms to better support
5574 dialog box interfaces, application-defined prompts, the possibility
5575 to use defaults (for example default passwords from somewhere else)
5576 and interrupts/cancellations.
5579 *) Tidy up PKCS#12 attribute handling. Add support for the CSP name
5580 attribute in PKCS#12 files, add new -CSP option to pkcs12 utility.
5583 *) Fix a memory leak in 'sk_dup()' in the case reallocation fails. (Also
5584 tidy up some unnecessarily weird code in 'sk_new()').
5585 [Geoff, reported by Diego Tartara <dtartara@novamens.com>]
5587 *) Change the key loading routines for ENGINEs to use the same kind
5588 callback (pem_password_cb) as all other routines that need this
5592 *) Increase ENTROPY_NEEDED to 32 bytes, as Rijndael can operate with
5593 256 bit (=32 byte) keys. Of course seeding with more entropy bytes
5594 than this minimum value is recommended.
5597 *) New random seeder for OpenVMS, using the system process statistics
5598 that are easily reachable.
5601 *) Windows apparently can't transparently handle global
5602 variables defined in DLLs. Initialisations such as:
5604 const ASN1_ITEM *it = &ASN1_INTEGER_it;
5606 wont compile. This is used by the any applications that need to
5607 declare their own ASN1 modules. This was fixed by adding the option
5608 EXPORT_VAR_AS_FN to all Win32 platforms, although this isn't strictly
5609 needed for static libraries under Win32.
5612 *) New functions X509_PURPOSE_set() and X509_TRUST_set() to handle
5613 setting of purpose and trust fields. New X509_STORE trust and
5614 purpose functions and tidy up setting in other SSL functions.
5617 *) Add copies of X509_STORE_CTX fields and callbacks to X509_STORE
5618 structure. These are inherited by X509_STORE_CTX when it is
5619 initialised. This allows various defaults to be set in the
5620 X509_STORE structure (such as flags for CRL checking and custom
5621 purpose or trust settings) for functions which only use X509_STORE_CTX
5622 internally such as S/MIME.
5624 Modify X509_STORE_CTX_purpose_inherit() so it only sets purposes and
5625 trust settings if they are not set in X509_STORE. This allows X509_STORE
5626 purposes and trust (in S/MIME for example) to override any set by default.
5628 Add command line options for CRL checking to smime, s_client and s_server
5632 *) Initial CRL based revocation checking. If the CRL checking flag(s)
5633 are set then the CRL is looked up in the X509_STORE structure and
5634 its validity and signature checked, then if the certificate is found
5635 in the CRL the verify fails with a revoked error.
5637 Various new CRL related callbacks added to X509_STORE_CTX structure.
5639 Command line options added to 'verify' application to support this.
5641 This needs some additional work, such as being able to handle multiple
5642 CRLs with different times, extension based lookup (rather than just
5643 by subject name) and ultimately more complete V2 CRL extension
5647 *) Add a general user interface API (crypto/ui/). This is designed
5648 to replace things like des_read_password and friends (backward
5649 compatibility functions using this new API are provided).
5650 The purpose is to remove prompting functions from the DES code
5651 section as well as provide for prompting through dialog boxes in
5652 a window system and the like.
5655 *) Add "ex_data" support to ENGINE so implementations can add state at a
5656 per-structure level rather than having to store it globally.
5659 *) Make it possible for ENGINE structures to be copied when retrieved by
5660 ENGINE_by_id() if the ENGINE specifies a new flag: ENGINE_FLAGS_BY_ID_COPY.
5661 This causes the "original" ENGINE structure to act like a template,
5662 analogous to the RSA vs. RSA_METHOD type of separation. Because of this
5663 operational state can be localised to each ENGINE structure, despite the
5664 fact they all share the same "methods". New ENGINE structures returned in
5665 this case have no functional references and the return value is the single
5666 structural reference. This matches the single structural reference returned
5667 by ENGINE_by_id() normally, when it is incremented on the pre-existing
5671 *) Fix ASN1 decoder when decoding type ANY and V_ASN1_OTHER: since this
5672 needs to match any other type at all we need to manually clear the
5676 *) Changes to the "openssl engine" utility to include;
5677 - verbosity levels ('-v', '-vv', and '-vvv') that provide information
5678 about an ENGINE's available control commands.
5679 - executing control commands from command line arguments using the
5680 '-pre' and '-post' switches. '-post' is only used if '-t' is
5681 specified and the ENGINE is successfully initialised. The syntax for
5682 the individual commands are colon-separated, for example;
5683 openssl engine chil -pre FORK_CHECK:0 -pre SO_PATH:/lib/test.so
5686 *) New dynamic control command support for ENGINEs. ENGINEs can now
5687 declare their own commands (numbers), names (strings), descriptions,
5688 and input types for run-time discovery by calling applications. A
5689 subset of these commands are implicitly classed as "executable"
5690 depending on their input type, and only these can be invoked through
5691 the new string-based API function ENGINE_ctrl_cmd_string(). (Eg. this
5692 can be based on user input, config files, etc). The distinction is
5693 that "executable" commands cannot return anything other than a boolean
5694 result and can only support numeric or string input, whereas some
5695 discoverable commands may only be for direct use through
5696 ENGINE_ctrl(), eg. supporting the exchange of binary data, function
5697 pointers, or other custom uses. The "executable" commands are to
5698 support parameterisations of ENGINE behaviour that can be
5699 unambiguously defined by ENGINEs and used consistently across any
5700 OpenSSL-based application. Commands have been added to all the
5701 existing hardware-supporting ENGINEs, noticeably "SO_PATH" to allow
5702 control over shared-library paths without source code alterations.
5705 *) Changed all ENGINE implementations to dynamically allocate their
5706 ENGINEs rather than declaring them statically. Apart from this being
5707 necessary with the removal of the ENGINE_FLAGS_MALLOCED distinction,
5708 this also allows the implementations to compile without using the
5709 internal engine_int.h header.
5712 *) Minor adjustment to "rand" code. RAND_get_rand_method() now returns a
5713 'const' value. Any code that should be able to modify a RAND_METHOD
5714 should already have non-const pointers to it (ie. they should only
5715 modify their own ones).
5718 *) Made a variety of little tweaks to the ENGINE code.
5719 - "atalla" and "ubsec" string definitions were moved from header files
5720 to C code. "nuron" string definitions were placed in variables
5721 rather than hard-coded - allowing parameterisation of these values
5722 later on via ctrl() commands.
5723 - Removed unused "#if 0"'d code.
5724 - Fixed engine list iteration code so it uses ENGINE_free() to release
5725 structural references.
5726 - Constified the RAND_METHOD element of ENGINE structures.
5727 - Constified various get/set functions as appropriate and added
5728 missing functions (including a catch-all ENGINE_cpy that duplicates
5729 all ENGINE values onto a new ENGINE except reference counts/state).
5730 - Removed NULL parameter checks in get/set functions. Setting a method
5731 or function to NULL is a way of cancelling out a previously set
5732 value. Passing a NULL ENGINE parameter is just plain stupid anyway
5733 and doesn't justify the extra error symbols and code.
5734 - Deprecate the ENGINE_FLAGS_MALLOCED define and move the area for
5735 flags from engine_int.h to engine.h.
5736 - Changed prototypes for ENGINE handler functions (init(), finish(),
5737 ctrl(), key-load functions, etc) to take an (ENGINE*) parameter.
5740 *) Implement binary inversion algorithm for BN_mod_inverse in addition
5741 to the algorithm using long division. The binary algorithm can be
5742 used only if the modulus is odd. On 32-bit systems, it is faster
5743 only for relatively small moduli (roughly 20-30% for 128-bit moduli,
5744 roughly 5-15% for 256-bit moduli), so we use it only for moduli
5745 up to 450 bits. In 64-bit environments, the binary algorithm
5746 appears to be advantageous for much longer moduli; here we use it
5747 for moduli up to 2048 bits.
5750 *) Rewrite CHOICE field setting in ASN1_item_ex_d2i(). The old code
5751 could not support the combine flag in choice fields.
5754 *) Add a 'copy_extensions' option to the 'ca' utility. This copies
5755 extensions from a certificate request to the certificate.
5758 *) Allow multiple 'certopt' and 'nameopt' options to be separated
5759 by commas. Add 'namopt' and 'certopt' options to the 'ca' config
5760 file: this allows the display of the certificate about to be
5761 signed to be customised, to allow certain fields to be included
5762 or excluded and extension details. The old system didn't display
5763 multicharacter strings properly, omitted fields not in the policy
5764 and couldn't display additional details such as extensions.
5767 *) Function EC_POINTs_mul for multiple scalar multiplication
5768 of an arbitrary number of elliptic curve points
5769 \sum scalars[i]*points[i],
5770 optionally including the generator defined for the EC_GROUP:
5771 scalar*generator + \sum scalars[i]*points[i].
5773 EC_POINT_mul is a simple wrapper function for the typical case
5774 that the point list has just one item (besides the optional
5778 *) First EC_METHODs for curves over GF(p):
5780 EC_GFp_simple_method() uses the basic BN_mod_mul and BN_mod_sqr
5781 operations and provides various method functions that can also
5782 operate with faster implementations of modular arithmetic.
5784 EC_GFp_mont_method() reuses most functions that are part of
5785 EC_GFp_simple_method, but uses Montgomery arithmetic.
5787 [Bodo Moeller; point addition and point doubling
5788 implementation directly derived from source code provided by
5789 Lenka Fibikova <fibikova@exp-math.uni-essen.de>]
5791 *) Framework for elliptic curves (crypto/ec/ec.h, crypto/ec/ec_lcl.h,
5792 crypto/ec/ec_lib.c):
5794 Curves are EC_GROUP objects (with an optional group generator)
5795 based on EC_METHODs that are built into the library.
5797 Points are EC_POINT objects based on EC_GROUP objects.
5799 Most of the framework would be able to handle curves over arbitrary
5800 finite fields, but as there are no obvious types for fields other
5801 than GF(p), some functions are limited to that for now.
5804 *) Add the -HTTP option to s_server. It is similar to -WWW, but requires
5805 that the file contains a complete HTTP response.
5808 *) Add the ec directory to mkdef.pl and mkfiles.pl. In mkdef.pl
5809 change the def and num file printf format specifier from "%-40sXXX"
5810 to "%-39s XXX". The latter will always guarantee a space after the
5811 field while the former will cause them to run together if the field
5812 is 40 of more characters long.
5815 *) Constify the cipher and digest 'method' functions and structures
5816 and modify related functions to take constant EVP_MD and EVP_CIPHER
5820 *) Hide BN_CTX structure details in bn_lcl.h instead of publishing them
5821 in <openssl/bn.h>. Also further increase BN_CTX_NUM to 32.
5824 *) Modify EVP_Digest*() routines so they now return values. Although the
5825 internal software routines can never fail additional hardware versions
5829 *) Clean up crypto/err/err.h and change some error codes to avoid conflicts:
5831 Previously ERR_R_FATAL was too small and coincided with ERR_LIB_PKCS7
5832 (= ERR_R_PKCS7_LIB); it is now 64 instead of 32.
5835 ERR_R_NESTED_ASN1_ERROR
5837 ERR_R_MISSING_ASN1_EOS
5838 were 4 .. 9, conflicting with
5839 ERR_LIB_RSA (= ERR_R_RSA_LIB)
5841 ERR_LIB_PEM (= ERR_R_PEM_LIB).
5842 They are now 58 .. 63 (i.e., just below ERR_R_FATAL).
5844 Add new error code 'ERR_R_INTERNAL_ERROR'.
5847 *) Don't overuse locks in crypto/err/err.c: For data retrieval, CRYPTO_r_lock
5851 *) New option '-subj arg' for 'openssl req' and 'openssl ca'. This
5852 sets the subject name for a new request or supersedes the
5853 subject name in a given request. Formats that can be parsed are
5854 'CN=Some Name, OU=myOU, C=IT'
5856 'CN=Some Name/OU=myOU/C=IT'.
5858 Add options '-batch' and '-verbose' to 'openssl req'.
5859 [Massimiliano Pala <madwolf@hackmasters.net>]
5861 *) Introduce the possibility to access global variables through
5862 functions on platform were that's the best way to handle exporting
5863 global variables in shared libraries. To enable this functionality,
5864 one must configure with "EXPORT_VAR_AS_FN" or defined the C macro
5865 "OPENSSL_EXPORT_VAR_AS_FUNCTION" in crypto/opensslconf.h (the latter
5866 is normally done by Configure or something similar).
5868 To implement a global variable, use the macro OPENSSL_IMPLEMENT_GLOBAL
5869 in the source file (foo.c) like this:
5871 OPENSSL_IMPLEMENT_GLOBAL(int,foo)=1;
5872 OPENSSL_IMPLEMENT_GLOBAL(double,bar);
5874 To declare a global variable, use the macros OPENSSL_DECLARE_GLOBAL
5875 and OPENSSL_GLOBAL_REF in the header file (foo.h) like this:
5877 OPENSSL_DECLARE_GLOBAL(int,foo);
5878 #define foo OPENSSL_GLOBAL_REF(foo)
5879 OPENSSL_DECLARE_GLOBAL(double,bar);
5880 #define bar OPENSSL_GLOBAL_REF(bar)
5882 The #defines are very important, and therefore so is including the
5883 header file everywhere where the defined globals are used.
5885 The macro OPENSSL_EXPORT_VAR_AS_FUNCTION also affects the definition
5886 of ASN.1 items, but that structure is a bit different.
5888 The largest change is in util/mkdef.pl which has been enhanced with
5889 better and easier to understand logic to choose which symbols should
5890 go into the Windows .def files as well as a number of fixes and code
5891 cleanup (among others, algorithm keywords are now sorted
5892 lexicographically to avoid constant rewrites).
5895 *) In BN_div() keep a copy of the sign of 'num' before writing the
5896 result to 'rm' because if rm==num the value will be overwritten
5897 and produce the wrong result if 'num' is negative: this caused
5898 problems with BN_mod() and BN_nnmod().
5901 *) Function OCSP_request_verify(). This checks the signature on an
5902 OCSP request and verifies the signer certificate. The signer
5903 certificate is just checked for a generic purpose and OCSP request
5907 *) Add OCSP_check_validity() function to check the validity of OCSP
5908 responses. OCSP responses are prepared in real time and may only
5909 be a few seconds old. Simply checking that the current time lies
5910 between thisUpdate and nextUpdate max reject otherwise valid responses
5911 caused by either OCSP responder or client clock inaccuracy. Instead
5912 we allow thisUpdate and nextUpdate to fall within a certain period of
5913 the current time. The age of the response can also optionally be
5914 checked. Two new options -validity_period and -status_age added to
5918 *) If signature or public key algorithm is unrecognized print out its
5919 OID rather that just UNKNOWN.
5922 *) Change OCSP_cert_to_id() to tolerate a NULL subject certificate and
5923 OCSP_cert_id_new() a NULL serialNumber. This allows a partial certificate
5924 ID to be generated from the issuer certificate alone which can then be
5925 passed to OCSP_id_issuer_cmp().
5928 *) New compilation option ASN1_ITEM_FUNCTIONS. This causes the new
5929 ASN1 modules to export functions returning ASN1_ITEM pointers
5930 instead of the ASN1_ITEM structures themselves. This adds several
5931 new macros which allow the underlying ASN1 function/structure to
5932 be accessed transparently. As a result code should not use ASN1_ITEM
5933 references directly (such as &X509_it) but instead use the relevant
5934 macros (such as ASN1_ITEM_rptr(X509)). This option is to allow
5935 use of the new ASN1 code on platforms where exporting structures
5936 is problematical (for example in shared libraries) but exporting
5937 functions returning pointers to structures is not.
5940 *) Add support for overriding the generation of SSL/TLS session IDs.
5941 These callbacks can be registered either in an SSL_CTX or per SSL.
5942 The purpose of this is to allow applications to control, if they wish,
5943 the arbitrary values chosen for use as session IDs, particularly as it
5944 can be useful for session caching in multiple-server environments. A
5945 command-line switch for testing this (and any client code that wishes
5946 to use such a feature) has been added to "s_server".
5947 [Geoff Thorpe, Lutz Jaenicke]
5949 *) Modify mkdef.pl to recognise and parse preprocessor conditionals
5950 of the form '#if defined(...) || defined(...) || ...' and
5951 '#if !defined(...) && !defined(...) && ...'. This also avoids
5952 the growing number of special cases it was previously handling.
5955 *) Make all configuration macros available for application by making
5956 sure they are available in opensslconf.h, by giving them names starting
5957 with "OPENSSL_" to avoid conflicts with other packages and by making
5958 sure e_os2.h will cover all platform-specific cases together with
5960 Additionally, it is now possible to define configuration/platform-
5961 specific names (called "system identities"). In the C code, these
5962 are prefixed with "OPENSSL_SYSNAME_". e_os2.h will create another
5963 macro with the name beginning with "OPENSSL_SYS_", which is determined
5964 from "OPENSSL_SYSNAME_*" or compiler-specific macros depending on
5968 *) New option -set_serial to 'req' and 'x509' this allows the serial
5969 number to use to be specified on the command line. Previously self
5970 signed certificates were hard coded with serial number 0 and the
5971 CA options of 'x509' had to use a serial number in a file which was
5975 *) New options to 'ca' utility to support V2 CRL entry extensions.
5976 Currently CRL reason, invalidity date and hold instruction are
5977 supported. Add new CRL extensions to V3 code and some new objects.
5980 *) New function EVP_CIPHER_CTX_set_padding() this is used to
5981 disable standard block padding (aka PKCS#5 padding) in the EVP
5982 API, which was previously mandatory. This means that the data is
5983 not padded in any way and so the total length much be a multiple
5984 of the block size, otherwise an error occurs.
5987 *) Initial (incomplete) OCSP SSL support.
5990 *) New function OCSP_parse_url(). This splits up a URL into its host,
5991 port and path components: primarily to parse OCSP URLs. New -url
5992 option to ocsp utility.
5995 *) New nonce behavior. The return value of OCSP_check_nonce() now
5996 reflects the various checks performed. Applications can decide
5997 whether to tolerate certain situations such as an absent nonce
5998 in a response when one was present in a request: the ocsp application
5999 just prints out a warning. New function OCSP_add1_basic_nonce()
6000 this is to allow responders to include a nonce in a response even if
6001 the request is nonce-less.
6004 *) Disable stdin buffering in load_cert (apps/apps.c) so that no certs are
6005 skipped when using openssl x509 multiple times on a single input file,
6006 e.g. "(openssl x509 -out cert1; openssl x509 -out cert2) <certs".
6009 *) Make ASN1_UTCTIME_set_string() and ASN1_GENERALIZEDTIME_set_string()
6010 set string type: to handle setting ASN1_TIME structures. Fix ca
6011 utility to correctly initialize revocation date of CRLs.
6014 *) New option SSL_OP_CIPHER_SERVER_PREFERENCE allows the server to override
6015 the clients preferred ciphersuites and rather use its own preferences.
6016 Should help to work around M$ SGC (Server Gated Cryptography) bug in
6017 Internet Explorer by ensuring unchanged hash method during stepup.
6018 (Also replaces the broken/deactivated SSL_OP_NON_EXPORT_FIRST option.)
6021 *) Make mkdef.pl recognise all DECLARE_ASN1 macros, change rijndael
6022 to aes and add a new 'exist' option to print out symbols that don't
6026 *) Additional options to ocsp utility to allow flags to be set and
6027 additional certificates supplied.
6030 *) Add the option -VAfile to 'openssl ocsp', so the user can give the
6031 OCSP client a number of certificate to only verify the response
6035 *) Update Rijndael code to version 3.0 and change EVP AES ciphers to
6036 handle the new API. Currently only ECB, CBC modes supported. Add new
6039 Add TLS AES ciphersuites as described in RFC3268, "Advanced
6040 Encryption Standard (AES) Ciphersuites for Transport Layer
6041 Security (TLS)". (In beta versions of OpenSSL 0.9.7, these were
6042 not enabled by default and were not part of the "ALL" ciphersuite
6043 alias because they were not yet official; they could be
6044 explicitly requested by specifying the "AESdraft" ciphersuite
6045 group alias. In the final release of OpenSSL 0.9.7, the group
6046 alias is called "AES" and is part of "ALL".)
6047 [Ben Laurie, Steve Henson, Bodo Moeller]
6049 *) New function OCSP_copy_nonce() to copy nonce value (if present) from
6050 request to response.
6053 *) Functions for OCSP responders. OCSP_request_onereq_count(),
6054 OCSP_request_onereq_get0(), OCSP_onereq_get0_id() and OCSP_id_get0_info()
6055 extract information from a certificate request. OCSP_response_create()
6056 creates a response and optionally adds a basic response structure.
6057 OCSP_basic_add1_status() adds a complete single response to a basic
6058 response and returns the OCSP_SINGLERESP structure just added (to allow
6059 extensions to be included for example). OCSP_basic_add1_cert() adds a
6060 certificate to a basic response and OCSP_basic_sign() signs a basic
6061 response with various flags. New helper functions ASN1_TIME_check()
6062 (checks validity of ASN1_TIME structure) and ASN1_TIME_to_generalizedtime()
6063 (converts ASN1_TIME to GeneralizedTime).
6066 *) Various new functions. EVP_Digest() combines EVP_Digest{Init,Update,Final}()
6067 in a single operation. X509_get0_pubkey_bitstr() extracts the public_key
6068 structure from a certificate. X509_pubkey_digest() digests the public_key
6069 contents: this is used in various key identifiers.
6072 *) Make sk_sort() tolerate a NULL argument.
6073 [Steve Henson reported by Massimiliano Pala <madwolf@comune.modena.it>]
6075 *) New OCSP verify flag OCSP_TRUSTOTHER. When set the "other" certificates
6076 passed by the function are trusted implicitly. If any of them signed the
6077 response then it is assumed to be valid and is not verified.
6080 *) In PKCS7_set_type() initialise content_type in PKCS7_ENC_CONTENT
6081 to data. This was previously part of the PKCS7 ASN1 code. This
6082 was causing problems with OpenSSL created PKCS#12 and PKCS#7 structures.
6083 [Steve Henson, reported by Kenneth R. Robinette
6084 <support@securenetterm.com>]
6086 *) Add CRYPTO_push_info() and CRYPTO_pop_info() calls to new ASN1
6087 routines: without these tracing memory leaks is very painful.
6088 Fix leaks in PKCS12 and PKCS7 routines.
6091 *) Make X509_time_adj() cope with the new behaviour of ASN1_TIME_new().
6092 Previously it initialised the 'type' argument to V_ASN1_UTCTIME which
6093 effectively meant GeneralizedTime would never be used. Now it
6094 is initialised to -1 but X509_time_adj() now has to check the value
6095 and use ASN1_TIME_set() if the value is not V_ASN1_UTCTIME or
6096 V_ASN1_GENERALIZEDTIME, without this it always uses GeneralizedTime.
6097 [Steve Henson, reported by Kenneth R. Robinette
6098 <support@securenetterm.com>]
6100 *) Fixes to BN_to_ASN1_INTEGER when bn is zero. This would previously
6101 result in a zero length in the ASN1_INTEGER structure which was
6102 not consistent with the structure when d2i_ASN1_INTEGER() was used
6103 and would cause ASN1_INTEGER_cmp() to fail. Enhance s2i_ASN1_INTEGER()
6104 to cope with hex and negative integers. Fix bug in i2a_ASN1_INTEGER()
6105 where it did not print out a minus for negative ASN1_INTEGER.
6108 *) Add summary printout to ocsp utility. The various functions which
6109 convert status values to strings have been renamed to:
6110 OCSP_response_status_str(), OCSP_cert_status_str() and
6111 OCSP_crl_reason_str() and are no longer static. New options
6112 to verify nonce values and to disable verification. OCSP response
6113 printout format cleaned up.
6116 *) Add additional OCSP certificate checks. These are those specified
6117 in RFC2560. This consists of two separate checks: the CA of the
6118 certificate being checked must either be the OCSP signer certificate
6119 or the issuer of the OCSP signer certificate. In the latter case the
6120 OCSP signer certificate must contain the OCSP signing extended key
6121 usage. This check is performed by attempting to match the OCSP
6122 signer or the OCSP signer CA to the issuerNameHash and issuerKeyHash
6123 in the OCSP_CERTID structures of the response.
6126 *) Initial OCSP certificate verification added to OCSP_basic_verify()
6127 and related routines. This uses the standard OpenSSL certificate
6128 verify routines to perform initial checks (just CA validity) and
6129 to obtain the certificate chain. Then additional checks will be
6130 performed on the chain. Currently the root CA is checked to see
6131 if it is explicitly trusted for OCSP signing. This is used to set
6132 a root CA as a global signing root: that is any certificate that
6133 chains to that CA is an acceptable OCSP signing certificate.
6136 *) New '-extfile ...' option to 'openssl ca' for reading X.509v3
6137 extensions from a separate configuration file.
6138 As when reading extensions from the main configuration file,
6139 the '-extensions ...' option may be used for specifying the
6141 [Massimiliano Pala <madwolf@comune.modena.it>]
6143 *) New OCSP utility. Allows OCSP requests to be generated or
6144 read. The request can be sent to a responder and the output
6145 parsed, outputed or printed in text form. Not complete yet:
6146 still needs to check the OCSP response validity.
6149 *) New subcommands for 'openssl ca':
6150 'openssl ca -status <serial>' prints the status of the cert with
6151 the given serial number (according to the index file).
6152 'openssl ca -updatedb' updates the expiry status of certificates
6154 [Massimiliano Pala <madwolf@comune.modena.it>]
6156 *) New '-newreq-nodes' command option to CA.pl. This is like
6157 '-newreq', but calls 'openssl req' with the '-nodes' option
6158 so that the resulting key is not encrypted.
6159 [Damien Miller <djm@mindrot.org>]
6161 *) New configuration for the GNU Hurd.
6162 [Jonathan Bartlett <johnnyb@wolfram.com> via Richard Levitte]
6164 *) Initial code to implement OCSP basic response verify. This
6165 is currently incomplete. Currently just finds the signer's
6166 certificate and verifies the signature on the response.
6169 *) New SSLeay_version code SSLEAY_DIR to determine the compiled-in
6170 value of OPENSSLDIR. This is available via the new '-d' option
6171 to 'openssl version', and is also included in 'openssl version -a'.
6174 *) Allowing defining memory allocation callbacks that will be given
6175 file name and line number information in additional arguments
6176 (a const char* and an int). The basic functionality remains, as
6177 well as the original possibility to just replace malloc(),
6178 realloc() and free() by functions that do not know about these
6179 additional arguments. To register and find out the current
6180 settings for extended allocation functions, the following
6181 functions are provided:
6183 CRYPTO_set_mem_ex_functions
6184 CRYPTO_set_locked_mem_ex_functions
6185 CRYPTO_get_mem_ex_functions
6186 CRYPTO_get_locked_mem_ex_functions
6188 These work the same way as CRYPTO_set_mem_functions and friends.
6189 CRYPTO_get_[locked_]mem_functions now writes 0 where such an
6190 extended allocation function is enabled.
6191 Similarly, CRYPTO_get_[locked_]mem_ex_functions writes 0 where
6192 a conventional allocation function is enabled.
6193 [Richard Levitte, Bodo Moeller]
6195 *) Finish off removing the remaining LHASH function pointer casts.
6196 There should no longer be any prototype-casting required when using
6197 the LHASH abstraction, and any casts that remain are "bugs". See
6198 the callback types and macros at the head of lhash.h for details
6199 (and "OBJ_cleanup" in crypto/objects/obj_dat.c as an example).
6202 *) Add automatic query of EGD sockets in RAND_poll() for the unix variant.
6203 If /dev/[u]random devices are not available or do not return enough
6204 entropy, EGD style sockets (served by EGD or PRNGD) will automatically
6206 The locations /var/run/egd-pool, /dev/egd-pool, /etc/egd-pool, and
6207 /etc/entropy will be queried once each in this sequence, quering stops
6208 when enough entropy was collected without querying more sockets.
6211 *) Change the Unix RAND_poll() variant to be able to poll several
6212 random devices, as specified by DEVRANDOM, until a sufficient amount
6213 of data has been collected. We spend at most 10 ms on each file
6214 (select timeout) and read in non-blocking mode. DEVRANDOM now
6215 defaults to the list "/dev/urandom", "/dev/random", "/dev/srandom"
6216 (previously it was just the string "/dev/urandom"), so on typical
6217 platforms the 10 ms delay will never occur.
6218 Also separate out the Unix variant to its own file, rand_unix.c.
6219 For VMS, there's a currently-empty rand_vms.c.
6222 *) Move OCSP client related routines to ocsp_cl.c. These
6223 provide utility functions which an application needing
6224 to issue a request to an OCSP responder and analyse the
6225 response will typically need: as opposed to those which an
6226 OCSP responder itself would need which will be added later.
6228 OCSP_request_sign() signs an OCSP request with an API similar
6229 to PKCS7_sign(). OCSP_response_status() returns status of OCSP
6230 response. OCSP_response_get1_basic() extracts basic response
6231 from response. OCSP_resp_find_status(): finds and extracts status
6232 information from an OCSP_CERTID structure (which will be created
6233 when the request structure is built). These are built from lower
6234 level functions which work on OCSP_SINGLERESP structures but
6235 wont normally be used unless the application wishes to examine
6236 extensions in the OCSP response for example.
6238 Replace nonce routines with a pair of functions.
6239 OCSP_request_add1_nonce() adds a nonce value and optionally
6240 generates a random value. OCSP_check_nonce() checks the
6241 validity of the nonce in an OCSP response.
6244 *) Change function OCSP_request_add() to OCSP_request_add0_id().
6245 This doesn't copy the supplied OCSP_CERTID and avoids the
6246 need to free up the newly created id. Change return type
6247 to OCSP_ONEREQ to return the internal OCSP_ONEREQ structure.
6248 This can then be used to add extensions to the request.
6249 Deleted OCSP_request_new(), since most of its functionality
6250 is now in OCSP_REQUEST_new() (and the case insensitive name
6251 clash) apart from the ability to set the request name which
6252 will be added elsewhere.
6255 *) Update OCSP API. Remove obsolete extensions argument from
6256 various functions. Extensions are now handled using the new
6257 OCSP extension code. New simple OCSP HTTP function which
6258 can be used to send requests and parse the response.
6261 *) Fix the PKCS#7 (S/MIME) code to work with new ASN1. Two new
6262 ASN1_ITEM structures help with sign and verify. PKCS7_ATTR_SIGN
6263 uses the special reorder version of SET OF to sort the attributes
6264 and reorder them to match the encoded order. This resolves a long
6265 standing problem: a verify on a PKCS7 structure just after signing
6266 it used to fail because the attribute order did not match the
6267 encoded order. PKCS7_ATTR_VERIFY does not reorder the attributes:
6268 it uses the received order. This is necessary to tolerate some broken
6269 software that does not order SET OF. This is handled by encoding
6270 as a SEQUENCE OF but using implicit tagging (with UNIVERSAL class)
6271 to produce the required SET OF.
6274 *) Have mk1mf.pl generate the macros OPENSSL_BUILD_SHLIBCRYPTO and
6275 OPENSSL_BUILD_SHLIBSSL and use them appropriately in the header
6276 files to get correct declarations of the ASN.1 item variables.
6279 *) Rewrite of PKCS#12 code to use new ASN1 functionality. Replace many
6280 PKCS#12 macros with real functions. Fix two unrelated ASN1 bugs:
6281 asn1_check_tlen() would sometimes attempt to use 'ctx' when it was
6282 NULL and ASN1_TYPE was not dereferenced properly in asn1_ex_c2i().
6283 New ASN1 macro: DECLARE_ASN1_ITEM() which just declares the relevant
6284 ASN1_ITEM and no wrapper functions.
6287 *) New functions or ASN1_item_d2i_fp() and ASN1_item_d2i_bio(). These
6288 replace the old function pointer based I/O routines. Change most of
6289 the *_d2i_bio() and *_d2i_fp() functions to use these.
6292 *) Enhance mkdef.pl to be more accepting about spacing in C preprocessor
6293 lines, recognice more "algorithms" that can be deselected, and make
6294 it complain about algorithm deselection that isn't recognised.
6297 *) New ASN1 functions to handle dup, sign, verify, digest, pack and
6298 unpack operations in terms of ASN1_ITEM. Modify existing wrappers
6299 to use new functions. Add NO_ASN1_OLD which can be set to remove
6300 some old style ASN1 functions: this can be used to determine if old
6301 code will still work when these eventually go away.
6304 *) New extension functions for OCSP structures, these follow the
6305 same conventions as certificates and CRLs.
6308 *) New function X509V3_add1_i2d(). This automatically encodes and
6309 adds an extension. Its behaviour can be customised with various
6310 flags to append, replace or delete. Various wrappers added for
6311 certifcates and CRLs.
6314 *) Fix to avoid calling the underlying ASN1 print routine when
6315 an extension cannot be parsed. Correct a typo in the
6316 OCSP_SERVICELOC extension. Tidy up print OCSP format.
6319 *) Make mkdef.pl parse some of the ASN1 macros and add appropriate
6320 entries for variables.
6323 *) Add functionality to apps/openssl.c for detecting locking
6324 problems: As the program is single-threaded, all we have
6325 to do is register a locking callback using an array for
6326 storing which locks are currently held by the program.
6329 *) Use a lock around the call to CRYPTO_get_ex_new_index() in
6330 SSL_get_ex_data_X509_STORE_idx(), which is used in
6331 ssl_verify_cert_chain() and thus can be called at any time
6332 during TLS/SSL handshakes so that thread-safety is essential.
6333 Unfortunately, the ex_data design is not at all suited
6334 for multi-threaded use, so it probably should be abolished.
6337 *) Added Broadcom "ubsec" ENGINE to OpenSSL.
6338 [Broadcom, tweaked and integrated by Geoff Thorpe]
6340 *) Move common extension printing code to new function
6341 X509V3_print_extensions(). Reorganise OCSP print routines and
6342 implement some needed OCSP ASN1 functions. Add OCSP extensions.
6345 *) New function X509_signature_print() to remove duplication in some
6349 *) Add a special meaning when SET OF and SEQUENCE OF flags are both
6350 set (this was treated exactly the same as SET OF previously). This
6351 is used to reorder the STACK representing the structure to match the
6352 encoding. This will be used to get round a problem where a PKCS7
6353 structure which was signed could not be verified because the STACK
6354 order did not reflect the encoded order.
6357 *) Reimplement the OCSP ASN1 module using the new code.
6360 *) Update the X509V3 code to permit the use of an ASN1_ITEM structure
6361 for its ASN1 operations. The old style function pointers still exist
6362 for now but they will eventually go away.
6365 *) Merge in replacement ASN1 code from the ASN1 branch. This almost
6366 completely replaces the old ASN1 functionality with a table driven
6367 encoder and decoder which interprets an ASN1_ITEM structure describing
6368 the ASN1 module. Compatibility with the existing ASN1 API (i2d,d2i) is
6369 largely maintained. Almost all of the old asn1_mac.h macro based ASN1
6370 has also been converted to the new form.
6373 *) Change BN_mod_exp_recp so that negative moduli are tolerated
6374 (the sign is ignored). Similarly, ignore the sign in BN_MONT_CTX_set
6375 so that BN_mod_exp_mont and BN_mod_exp_mont_word work
6376 for negative moduli.
6379 *) Fix BN_uadd and BN_usub: Always return non-negative results instead
6380 of not touching the result's sign bit.
6383 *) BN_div bugfix: If the result is 0, the sign (res->neg) must not be
6387 *) Changed the LHASH code to use prototypes for callbacks, and created
6388 macros to declare and implement thin (optionally static) functions
6389 that provide type-safety and avoid function pointer casting for the
6390 type-specific callbacks.
6393 *) Added Kerberos Cipher Suites to be used with TLS, as written in
6395 [Veers Staats <staatsvr@asc.hpc.mil>,
6396 Jeffrey Altman <jaltman@columbia.edu>, via Richard Levitte]
6398 *) Reformat the FAQ so the different questions and answers can be divided
6399 in sections depending on the subject.
6402 *) Have the zlib compression code load ZLIB.DLL dynamically under
6406 *) New function BN_mod_sqrt for computing square roots modulo a prime
6407 (using the probabilistic Tonelli-Shanks algorithm unless
6408 p == 3 (mod 4) or p == 5 (mod 8), which are cases that can
6409 be handled deterministically).
6410 [Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller]
6412 *) Make BN_mod_inverse faster by explicitly handling small quotients
6413 in the Euclid loop. (Speed gain about 20% for small moduli [256 or
6414 512 bits], about 30% for larger ones [1024 or 2048 bits].)
6417 *) New function BN_kronecker.
6420 *) Fix BN_gcd so that it works on negative inputs; the result is
6421 positive unless both parameters are zero.
6422 Previously something reasonably close to an infinite loop was
6423 possible because numbers could be growing instead of shrinking
6424 in the implementation of Euclid's algorithm.
6427 *) Fix BN_is_word() and BN_is_one() macros to take into account the
6428 sign of the number in question.
6430 Fix BN_is_word(a,w) to work correctly for w == 0.
6432 The old BN_is_word(a,w) macro is now called BN_abs_is_word(a,w)
6433 because its test if the absolute value of 'a' equals 'w'.
6434 Note that BN_abs_is_word does *not* handle w == 0 reliably;
6435 it exists mostly for use in the implementations of BN_is_zero(),
6436 BN_is_one(), and BN_is_word().
6439 *) New function BN_swap.
6442 *) Use BN_nnmod instead of BN_mod in crypto/bn/bn_exp.c so that
6443 the exponentiation functions are more likely to produce reasonable
6444 results on negative inputs.
6447 *) Change BN_mod_mul so that the result is always non-negative.
6448 Previously, it could be negative if one of the factors was negative;
6449 I don't think anyone really wanted that behaviour.
6452 *) Move BN_mod_... functions into new file crypto/bn/bn_mod.c
6453 (except for exponentiation, which stays in crypto/bn/bn_exp.c,
6454 and BN_mod_mul_reciprocal, which stays in crypto/bn/bn_recp.c)
6455 and add new functions:
6464 BN_mod_lshift1_quick
6468 These functions always generate non-negative results.
6470 BN_nnmod otherwise is like BN_mod (if BN_mod computes a remainder r
6471 such that |m| < r < 0, BN_nnmod will output rem + |m| instead).
6473 BN_mod_XXX_quick(r, a, [b,] m) generates the same result as
6474 BN_mod_XXX(r, a, [b,] m, ctx), but requires that a [and b]
6475 be reduced modulo m.
6476 [Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller]
6479 The following entry accidentally appeared in the CHANGES file
6480 distributed with OpenSSL 0.9.7. The modifications described in
6481 it do *not* apply to OpenSSL 0.9.7.
6483 *) Remove a few calls to bn_wexpand() in BN_sqr() (the one in there
6484 was actually never needed) and in BN_mul(). The removal in BN_mul()
6485 required a small change in bn_mul_part_recursive() and the addition
6486 of the functions bn_cmp_part_words(), bn_sub_part_words() and
6487 bn_add_part_words(), which do the same thing as bn_cmp_words(),
6488 bn_sub_words() and bn_add_words() except they take arrays with
6493 *) In 'openssl passwd', verify passwords read from the terminal
6494 unless the '-salt' option is used (which usually means that
6495 verification would just waste user's time since the resulting
6496 hash is going to be compared with some given password hash)
6497 or the new '-noverify' option is used.
6499 This is an incompatible change, but it does not affect
6500 non-interactive use of 'openssl passwd' (passwords on the command
6501 line, '-stdin' option, '-in ...' option) and thus should not
6505 *) Remove all references to RSAref, since there's no more need for it.
6508 *) Make DSO load along a path given through an environment variable
6509 (SHLIB_PATH) with shl_load().
6512 *) Constify the ENGINE code as a result of BIGNUM constification.
6513 Also constify the RSA code and most things related to it. In a
6514 few places, most notable in the depth of the ASN.1 code, ugly
6515 casts back to non-const were required (to be solved at a later
6519 *) Make it so the openssl application has all engines loaded by default.
6522 *) Constify the BIGNUM routines a little more.
6525 *) Add the following functions:
6527 ENGINE_load_cswift()
6529 ENGINE_load_atalla()
6531 ENGINE_load_builtin_engines()
6533 That way, an application can itself choose if external engines that
6534 are built-in in OpenSSL shall ever be used or not. The benefit is
6535 that applications won't have to be linked with libdl or other dso
6536 libraries unless it's really needed.
6538 Changed 'openssl engine' to load all engines on demand.
6539 Changed the engine header files to avoid the duplication of some
6540 declarations (they differed!).
6543 *) 'openssl engine' can now list capabilities.
6546 *) Better error reporting in 'openssl engine'.
6549 *) Never call load_dh_param(NULL) in s_server.
6552 *) Add engine application. It can currently list engines by name and
6553 identity, and test if they are actually available.
6556 *) Improve RPM specification file by forcing symbolic linking and making
6557 sure the installed documentation is also owned by root.root.
6558 [Damien Miller <djm@mindrot.org>]
6560 *) Give the OpenSSL applications more possibilities to make use of
6561 keys (public as well as private) handled by engines.
6564 *) Add OCSP code that comes from CertCo.
6567 *) Add VMS support for the Rijndael code.
6570 *) Added untested support for Nuron crypto accelerator.
6573 *) Add support for external cryptographic devices. This code was
6574 previously distributed separately as the "engine" branch.
6575 [Geoff Thorpe, Richard Levitte]
6577 *) Rework the filename-translation in the DSO code. It is now possible to
6578 have far greater control over how a "name" is turned into a filename
6579 depending on the operating environment and any oddities about the
6580 different shared library filenames on each system.
6583 *) Support threads on FreeBSD-elf in Configure.
6586 *) Fix for SHA1 assembly problem with MASM: it produces
6587 warnings about corrupt line number information when assembling
6588 with debugging information. This is caused by the overlapping
6590 [Bernd Matthes <mainbug@celocom.de>, Steve Henson]
6593 NCONF_get_number() has no error checking at all. As a replacement,
6594 NCONF_get_number_e() is defined (_e for "error checking") and is
6595 promoted strongly. The old NCONF_get_number is kept around for
6596 binary backward compatibility.
6597 Make it possible for methods to load from something other than a BIO,
6598 by providing a function pointer that is given a name instead of a BIO.
6599 For example, this could be used to load configuration data from an
6603 *) Fix for non blocking accept BIOs. Added new I/O special reason
6604 BIO_RR_ACCEPT to cover this case. Previously use of accept BIOs
6605 with non blocking I/O was not possible because no retry code was
6606 implemented. Also added new SSL code SSL_WANT_ACCEPT to cover
6610 *) Added the beginnings of Rijndael support.
6613 *) Fix for bug in DirectoryString mask setting. Add support for
6614 X509_NAME_print_ex() in 'req' and X509_print_ex() function
6615 to allow certificate printing to more controllable, additional
6616 'certopt' option to 'x509' to allow new printing options to be
6620 *) Clean old EAY MD5 hack from e_os.h.
6623 Changes between 0.9.6l and 0.9.6m [17 Mar 2004]
6625 *) Fix null-pointer assignment in do_change_cipher_spec() revealed
6626 by using the Codenomicon TLS Test Tool (CVE-2004-0079)
6627 [Joe Orton, Steve Henson]
6629 Changes between 0.9.6k and 0.9.6l [04 Nov 2003]
6631 *) Fix additional bug revealed by the NISCC test suite:
6633 Stop bug triggering large recursion when presented with
6634 certain ASN.1 tags (CVE-2003-0851)
6637 Changes between 0.9.6j and 0.9.6k [30 Sep 2003]
6639 *) Fix various bugs revealed by running the NISCC test suite:
6641 Stop out of bounds reads in the ASN1 code when presented with
6642 invalid tags (CVE-2003-0543 and CVE-2003-0544).
6644 If verify callback ignores invalid public key errors don't try to check
6645 certificate signature with the NULL public key.
6649 *) In ssl3_accept() (ssl/s3_srvr.c) only accept a client certificate
6650 if the server requested one: as stated in TLS 1.0 and SSL 3.0
6654 *) In ssl3_get_client_hello() (ssl/s3_srvr.c), tolerate additional
6655 extra data after the compression methods not only for TLS 1.0
6656 but also for SSL 3.0 (as required by the specification).
6657 [Bodo Moeller; problem pointed out by Matthias Loepfe]
6659 *) Change X509_certificate_type() to mark the key as exported/exportable
6660 when it's 512 *bits* long, not 512 bytes.
6663 Changes between 0.9.6i and 0.9.6j [10 Apr 2003]
6665 *) Countermeasure against the Klima-Pokorny-Rosa extension of
6666 Bleichbacher's attack on PKCS #1 v1.5 padding: treat
6667 a protocol version number mismatch like a decryption error
6668 in ssl3_get_client_key_exchange (ssl/s3_srvr.c).
6671 *) Turn on RSA blinding by default in the default implementation
6672 to avoid a timing attack. Applications that don't want it can call
6673 RSA_blinding_off() or use the new flag RSA_FLAG_NO_BLINDING.
6674 They would be ill-advised to do so in most cases.
6675 [Ben Laurie, Steve Henson, Geoff Thorpe, Bodo Moeller]
6677 *) Change RSA blinding code so that it works when the PRNG is not
6678 seeded (in this case, the secret RSA exponent is abused as
6679 an unpredictable seed -- if it is not unpredictable, there
6680 is no point in blinding anyway). Make RSA blinding thread-safe
6681 by remembering the creator's thread ID in rsa->blinding and
6682 having all other threads use local one-time blinding factors
6683 (this requires more computation than sharing rsa->blinding, but
6684 avoids excessive locking; and if an RSA object is not shared
6685 between threads, blinding will still be very fast).
6688 Changes between 0.9.6h and 0.9.6i [19 Feb 2003]
6690 *) In ssl3_get_record (ssl/s3_pkt.c), minimize information leaked
6691 via timing by performing a MAC computation even if incorrrect
6692 block cipher padding has been found. This is a countermeasure
6693 against active attacks where the attacker has to distinguish
6694 between bad padding and a MAC verification error. (CVE-2003-0078)
6696 [Bodo Moeller; problem pointed out by Brice Canvel (EPFL),
6697 Alain Hiltgen (UBS), Serge Vaudenay (EPFL), and
6698 Martin Vuagnoux (EPFL, Ilion)]
6700 Changes between 0.9.6g and 0.9.6h [5 Dec 2002]
6702 *) New function OPENSSL_cleanse(), which is used to cleanse a section of
6703 memory from it's contents. This is done with a counter that will
6704 place alternating values in each byte. This can be used to solve
6705 two issues: 1) the removal of calls to memset() by highly optimizing
6706 compilers, and 2) cleansing with other values than 0, since those can
6707 be read through on certain media, for example a swap space on disk.
6710 *) Bugfix: client side session caching did not work with external caching,
6711 because the session->cipher setting was not restored when reloading
6712 from the external cache. This problem was masked, when
6713 SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG (part of SSL_OP_ALL) was set.
6714 (Found by Steve Haslam <steve@araqnid.ddts.net>.)
6717 *) Fix client_certificate (ssl/s2_clnt.c): The permissible total
6718 length of the REQUEST-CERTIFICATE message is 18 .. 34, not 17 .. 33.
6719 [Zeev Lieber <zeev-l@yahoo.com>]
6721 *) Undo an undocumented change introduced in 0.9.6e which caused
6722 repeated calls to OpenSSL_add_all_ciphers() and
6723 OpenSSL_add_all_digests() to be ignored, even after calling
6727 *) Change the default configuration reader to deal with last line not
6728 being properly terminated.
6731 *) Change X509_NAME_cmp() so it applies the special rules on handling
6732 DN values that are of type PrintableString, as well as RDNs of type
6733 emailAddress where the value has the type ia5String.
6734 [stefank@valicert.com via Richard Levitte]
6736 *) Add a SSL_SESS_CACHE_NO_INTERNAL_STORE flag to take over half
6737 the job SSL_SESS_CACHE_NO_INTERNAL_LOOKUP was inconsistently
6738 doing, define a new flag (SSL_SESS_CACHE_NO_INTERNAL) to be
6739 the bitwise-OR of the two for use by the majority of applications
6740 wanting this behaviour, and update the docs. The documented
6741 behaviour and actual behaviour were inconsistent and had been
6742 changing anyway, so this is more a bug-fix than a behavioural
6744 [Geoff Thorpe, diagnosed by Nadav Har'El]
6746 *) Don't impose a 16-byte length minimum on session IDs in ssl/s3_clnt.c
6747 (the SSL 3.0 and TLS 1.0 specifications allow any length up to 32 bytes).
6750 *) Fix initialization code race conditions in
6751 SSLv23_method(), SSLv23_client_method(), SSLv23_server_method(),
6752 SSLv2_method(), SSLv2_client_method(), SSLv2_server_method(),
6753 SSLv3_method(), SSLv3_client_method(), SSLv3_server_method(),
6754 TLSv1_method(), TLSv1_client_method(), TLSv1_server_method(),
6755 ssl2_get_cipher_by_char(),
6756 ssl3_get_cipher_by_char().
6757 [Patrick McCormick <patrick@tellme.com>, Bodo Moeller]
6759 *) Reorder cleanup sequence in SSL_CTX_free(): only remove the ex_data after
6760 the cached sessions are flushed, as the remove_cb() might use ex_data
6761 contents. Bug found by Sam Varshavchik <mrsam@courier-mta.com>
6762 (see [openssl.org #212]).
6763 [Geoff Thorpe, Lutz Jaenicke]
6765 *) Fix typo in OBJ_txt2obj which incorrectly passed the content
6766 length, instead of the encoding length to d2i_ASN1_OBJECT.
6769 Changes between 0.9.6f and 0.9.6g [9 Aug 2002]
6771 *) [In 0.9.6g-engine release:]
6772 Fix crypto/engine/vendor_defns/cswift.h for WIN32 (use '_stdcall').
6773 [Lynn Gazis <lgazis@rainbow.com>]
6775 Changes between 0.9.6e and 0.9.6f [8 Aug 2002]
6777 *) Fix ASN1 checks. Check for overflow by comparing with LONG_MAX
6778 and get fix the header length calculation.
6779 [Florian Weimer <Weimer@CERT.Uni-Stuttgart.DE>,
6780 Alon Kantor <alonk@checkpoint.com> (and others),
6783 *) Use proper error handling instead of 'assertions' in buffer
6784 overflow checks added in 0.9.6e. This prevents DoS (the
6785 assertions could call abort()).
6786 [Arne Ansper <arne@ats.cyber.ee>, Bodo Moeller]
6788 Changes between 0.9.6d and 0.9.6e [30 Jul 2002]
6790 *) Add various sanity checks to asn1_get_length() to reject
6791 the ASN1 length bytes if they exceed sizeof(long), will appear
6792 negative or the content length exceeds the length of the
6794 [Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>]
6796 *) Fix cipher selection routines: ciphers without encryption had no flags
6797 for the cipher strength set and where therefore not handled correctly
6798 by the selection routines (PR #130).
6801 *) Fix EVP_dsa_sha macro.
6805 SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS
6806 for disabling the SSL 3.0/TLS 1.0 CBC vulnerability countermeasure
6807 that was added in OpenSSL 0.9.6d.
6809 As the countermeasure turned out to be incompatible with some
6810 broken SSL implementations, the new option is part of SSL_OP_ALL.
6811 SSL_OP_ALL is usually employed when compatibility with weird SSL
6812 implementations is desired (e.g. '-bugs' option to 's_client' and
6813 's_server'), so the new option is automatically set in many
6817 *) Changes in security patch:
6819 Changes marked "(CHATS)" were sponsored by the Defense Advanced
6820 Research Projects Agency (DARPA) and Air Force Research Laboratory,
6821 Air Force Materiel Command, USAF, under agreement number
6824 *) Add various sanity checks to asn1_get_length() to reject
6825 the ASN1 length bytes if they exceed sizeof(long), will appear
6826 negative or the content length exceeds the length of the
6827 supplied buffer. (CVE-2002-0659)
6828 [Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>]
6830 *) Assertions for various potential buffer overflows, not known to
6832 [Ben Laurie (CHATS)]
6834 *) Various temporary buffers to hold ASCII versions of integers were
6835 too small for 64 bit platforms. (CVE-2002-0655)
6836 [Matthew Byng-Maddick <mbm@aldigital.co.uk> and Ben Laurie (CHATS)>
6838 *) Remote buffer overflow in SSL3 protocol - an attacker could
6839 supply an oversized session ID to a client. (CVE-2002-0656)
6840 [Ben Laurie (CHATS)]
6842 *) Remote buffer overflow in SSL2 protocol - an attacker could
6843 supply an oversized client master key. (CVE-2002-0656)
6844 [Ben Laurie (CHATS)]
6846 Changes between 0.9.6c and 0.9.6d [9 May 2002]
6848 *) Fix crypto/asn1/a_sign.c so that 'parameters' is omitted (not
6849 encoded as NULL) with id-dsa-with-sha1.
6850 [Nils Larsch <nla@trustcenter.de>; problem pointed out by Bodo Moeller]
6852 *) Check various X509_...() return values in apps/req.c.
6853 [Nils Larsch <nla@trustcenter.de>]
6855 *) Fix BASE64 decode (EVP_DecodeUpdate) for data with CR/LF ended lines:
6856 an end-of-file condition would erronously be flagged, when the CRLF
6857 was just at the end of a processed block. The bug was discovered when
6858 processing data through a buffering memory BIO handing the data to a
6859 BASE64-decoding BIO. Bug fund and patch submitted by Pavel Tsekov
6860 <ptsekov@syntrex.com> and Nedelcho Stanev.
6863 *) Implement a countermeasure against a vulnerability recently found
6864 in CBC ciphersuites in SSL 3.0/TLS 1.0: Send an empty fragment
6865 before application data chunks to avoid the use of known IVs
6866 with data potentially chosen by the attacker.
6869 *) Fix length checks in ssl3_get_client_hello().
6872 *) TLS/SSL library bugfix: use s->s3->in_read_app_data differently
6873 to prevent ssl3_read_internal() from incorrectly assuming that
6874 ssl3_read_bytes() found application data while handshake
6875 processing was enabled when in fact s->s3->in_read_app_data was
6876 merely automatically cleared during the initial handshake.
6877 [Bodo Moeller; problem pointed out by Arne Ansper <arne@ats.cyber.ee>]
6879 *) Fix object definitions for Private and Enterprise: they were not
6880 recognized in their shortname (=lowercase) representation. Extend
6881 obj_dat.pl to issue an error when using undefined keywords instead
6882 of silently ignoring the problem (Svenning Sorensen
6883 <sss@sss.dnsalias.net>).
6886 *) Fix DH_generate_parameters() so that it works for 'non-standard'
6887 generators, i.e. generators other than 2 and 5. (Previously, the
6888 code did not properly initialise the 'add' and 'rem' values to
6889 BN_generate_prime().)
6891 In the new general case, we do not insist that 'generator' is
6892 actually a primitive root: This requirement is rather pointless;
6893 a generator of the order-q subgroup is just as good, if not
6897 *) Map new X509 verification errors to alerts. Discovered and submitted by
6898 Tom Wu <tom@arcot.com>.
6901 *) Fix ssl3_pending() (ssl/s3_lib.c) to prevent SSL_pending() from
6902 returning non-zero before the data has been completely received
6903 when using non-blocking I/O.
6904 [Bodo Moeller; problem pointed out by John Hughes]
6906 *) Some of the ciphers missed the strength entry (SSL_LOW etc).
6907 [Ben Laurie, Lutz Jaenicke]
6909 *) Fix bug in SSL_clear(): bad sessions were not removed (found by
6910 Yoram Zahavi <YoramZ@gilian.com>).
6913 *) Add information about CygWin 1.3 and on, and preserve proper
6914 configuration for the versions before that.
6915 [Corinna Vinschen <vinschen@redhat.com> and Richard Levitte]
6917 *) Make removal from session cache (SSL_CTX_remove_session()) more robust:
6918 check whether we deal with a copy of a session and do not delete from
6919 the cache in this case. Problem reported by "Izhar Shoshani Levi"
6920 <izhar@checkpoint.com>.
6923 *) Do not store session data into the internal session cache, if it
6924 is never intended to be looked up (SSL_SESS_CACHE_NO_INTERNAL_LOOKUP
6925 flag is set). Proposed by Aslam <aslam@funk.com>.
6928 *) Have ASN1_BIT_STRING_set_bit() really clear a bit when the requested
6932 *) [In 0.9.6d-engine release:]
6933 Fix a crashbug and a logic bug in hwcrhk_load_pubkey().
6934 [Toomas Kiisk <vix@cyber.ee> via Richard Levitte]
6936 *) Add the configuration target linux-s390x.
6937 [Neale Ferguson <Neale.Ferguson@SoftwareAG-USA.com> via Richard Levitte]
6939 *) The earlier bugfix for the SSL3_ST_SW_HELLO_REQ_C case of
6940 ssl3_accept (ssl/s3_srvr.c) incorrectly used a local flag
6941 variable as an indication that a ClientHello message has been
6942 received. As the flag value will be lost between multiple
6943 invocations of ssl3_accept when using non-blocking I/O, the
6944 function may not be aware that a handshake has actually taken
6945 place, thus preventing a new session from being added to the
6948 To avoid this problem, we now set s->new_session to 2 instead of
6949 using a local variable.
6950 [Lutz Jaenicke, Bodo Moeller]
6952 *) Bugfix: Return -1 from ssl3_get_server_done (ssl3/s3_clnt.c)
6953 if the SSL_R_LENGTH_MISMATCH error is detected.
6954 [Geoff Thorpe, Bodo Moeller]
6956 *) New 'shared_ldflag' column in Configure platform table.
6959 *) Fix EVP_CIPHER_mode macro.
6960 ["Dan S. Camper" <dan@bti.net>]
6962 *) Fix ssl3_read_bytes (ssl/s3_pkt.c): To ignore messages of unknown
6963 type, we must throw them away by setting rr->length to 0.
6964 [D P Chang <dpc@qualys.com>]
6966 Changes between 0.9.6b and 0.9.6c [21 dec 2001]
6968 *) Fix BN_rand_range bug pointed out by Dominikus Scherkl
6969 <Dominikus.Scherkl@biodata.com>. (The previous implementation
6970 worked incorrectly for those cases where range = 10..._2 and
6971 3*range is two bits longer than range.)
6974 *) Only add signing time to PKCS7 structures if it is not already
6978 *) Fix crypto/objects/objects.h: "ld-ce" should be "id-ce",
6979 OBJ_ld_ce should be OBJ_id_ce.
6980 Also some ip-pda OIDs in crypto/objects/objects.txt were
6981 incorrect (cf. RFC 3039).
6982 [Matt Cooper, Frederic Giudicelli, Bodo Moeller]
6984 *) Release CRYPTO_LOCK_DYNLOCK when CRYPTO_destroy_dynlockid()
6985 returns early because it has nothing to do.
6986 [Andy Schneider <andy.schneider@bjss.co.uk>]
6988 *) [In 0.9.6c-engine release:]
6989 Fix mutex callback return values in crypto/engine/hw_ncipher.c.
6990 [Andy Schneider <andy.schneider@bjss.co.uk>]
6992 *) [In 0.9.6c-engine release:]
6993 Add support for Cryptographic Appliance's keyserver technology.
6994 (Use engine 'keyclient')
6995 [Cryptographic Appliances and Geoff Thorpe]
6997 *) Add a configuration entry for OS/390 Unix. The C compiler 'c89'
6998 is called via tools/c89.sh because arguments have to be
6999 rearranged (all '-L' options must appear before the first object
7001 [Richard Shapiro <rshapiro@abinitio.com>]
7003 *) [In 0.9.6c-engine release:]
7004 Add support for Broadcom crypto accelerator cards, backported
7006 [Broadcom, Nalin Dahyabhai <nalin@redhat.com>, Mark Cox]
7008 *) [In 0.9.6c-engine release:]
7009 Add support for SureWare crypto accelerator cards from
7010 Baltimore Technologies. (Use engine 'sureware')
7011 [Baltimore Technologies and Mark Cox]
7013 *) [In 0.9.6c-engine release:]
7014 Add support for crypto accelerator cards from Accelerated
7015 Encryption Processing, www.aep.ie. (Use engine 'aep')
7016 [AEP Inc. and Mark Cox]
7018 *) Add a configuration entry for gcc on UnixWare.
7019 [Gary Benson <gbenson@redhat.com>]
7021 *) Change ssl/s2_clnt.c and ssl/s2_srvr.c so that received handshake
7022 messages are stored in a single piece (fixed-length part and
7023 variable-length part combined) and fix various bugs found on the way.
7026 *) Disable caching in BIO_gethostbyname(), directly use gethostbyname()
7027 instead. BIO_gethostbyname() does not know what timeouts are
7028 appropriate, so entries would stay in cache even when they have
7030 [Bodo Moeller; problem pointed out by Rich Salz <rsalz@zolera.com>
7032 *) Change ssl23_get_client_hello (ssl/s23_srvr.c) behaviour when
7033 faced with a pathologically small ClientHello fragment that does
7034 not contain client_version: Instead of aborting with an error,
7035 simply choose the highest available protocol version (i.e.,
7036 TLS 1.0 unless it is disabled). In practice, ClientHello
7037 messages are never sent like this, but this change gives us
7038 strictly correct behaviour at least for TLS.
7041 *) Fix SSL handshake functions and SSL_clear() such that SSL_clear()
7042 never resets s->method to s->ctx->method when called from within
7043 one of the SSL handshake functions.
7044 [Bodo Moeller; problem pointed out by Niko Baric]
7046 *) In ssl3_get_client_hello (ssl/s3_srvr.c), generate a fatal alert
7047 (sent using the client's version number) if client_version is
7048 smaller than the protocol version in use. Also change
7049 ssl23_get_client_hello (ssl/s23_srvr.c) to select TLS 1.0 if
7050 the client demanded SSL 3.0 but only TLS 1.0 is enabled; then
7051 the client will at least see that alert.
7054 *) Fix ssl3_get_message (ssl/s3_both.c) to handle message fragmentation
7058 *) Avoid infinite loop in ssl3_get_message (ssl/s3_both.c) if a
7059 client receives HelloRequest while in a handshake.
7060 [Bodo Moeller; bug noticed by Andy Schneider <andy.schneider@bjss.co.uk>]
7062 *) Bugfix in ssl3_accept (ssl/s3_srvr.c): Case SSL3_ST_SW_HELLO_REQ_C
7063 should end in 'break', not 'goto end' which circumvents various
7064 cleanups done in state SSL_ST_OK. But session related stuff
7065 must be disabled for SSL_ST_OK in the case that we just sent a
7068 Also avoid some overhead by not calling ssl_init_wbio_buffer()
7069 before just sending a HelloRequest.
7070 [Bodo Moeller, Eric Rescorla <ekr@rtfm.com>]
7072 *) Fix ssl/s3_enc.c, ssl/t1_enc.c and ssl/s3_pkt.c so that we don't
7073 reveal whether illegal block cipher padding was found or a MAC
7074 verification error occured. (Neither SSLerr() codes nor alerts
7075 are directly visible to potential attackers, but the information
7076 may leak via logfiles.)
7078 Similar changes are not required for the SSL 2.0 implementation
7079 because the number of padding bytes is sent in clear for SSL 2.0,
7080 and the extra bytes are just ignored. However ssl/s2_pkt.c
7081 failed to verify that the purported number of padding bytes is in
7085 *) Add OpenUNIX-8 support including shared libraries
7086 (Boyd Lynn Gerber <gerberb@zenez.com>).
7089 *) Improve RSA_padding_check_PKCS1_OAEP() check again to avoid
7090 'wristwatch attack' using huge encoding parameters (cf.
7091 James H. Manger's CRYPTO 2001 paper). Note that the
7092 RSA_PKCS1_OAEP_PADDING case of RSA_private_decrypt() does not use
7093 encoding parameters and hence was not vulnerable.
7096 *) BN_sqr() bug fix.
7097 [Ulf Möller, reported by Jim Ellis <jim.ellis@cavium.com>]
7099 *) Rabin-Miller test analyses assume uniformly distributed witnesses,
7100 so use BN_pseudo_rand_range() instead of using BN_pseudo_rand()
7101 followed by modular reduction.
7102 [Bodo Moeller; pointed out by Adam Young <AYoung1@NCSUS.JNJ.COM>]
7104 *) Add BN_pseudo_rand_range() with obvious functionality: BN_rand_range()
7105 equivalent based on BN_pseudo_rand() instead of BN_rand().
7108 *) s3_srvr.c: allow sending of large client certificate lists (> 16 kB).
7109 This function was broken, as the check for a new client hello message
7110 to handle SGC did not allow these large messages.
7111 (Tracked down by "Douglas E. Engert" <deengert@anl.gov>.)
7114 *) Add alert descriptions for TLSv1 to SSL_alert_desc_string[_long]().
7117 *) Fix buggy behaviour of BIO_get_num_renegotiates() and BIO_ctrl()
7118 for BIO_C_GET_WRITE_BUF_SIZE ("Stephen Hinton" <shinton@netopia.com>).
7121 *) Rework the configuration and shared library support for Tru64 Unix.
7122 The configuration part makes use of modern compiler features and
7123 still retains old compiler behavior for those that run older versions
7124 of the OS. The shared library support part includes a variant that
7125 uses the RPATH feature, and is available through the special
7126 configuration target "alpha-cc-rpath", which will never be selected
7128 [Tim Mooney <mooney@dogbert.cc.ndsu.NoDak.edu> via Richard Levitte]
7130 *) In ssl3_get_key_exchange (ssl/s3_clnt.c), call ssl3_get_message()
7131 with the same message size as in ssl3_get_certificate_request().
7132 Otherwise, if no ServerKeyExchange message occurs, CertificateRequest
7133 messages might inadvertently be reject as too long.
7134 [Petr Lampa <lampa@fee.vutbr.cz>]
7136 *) Enhanced support for IA-64 Unix platforms (well, Linux and HP-UX).
7139 *) Modified SSL library such that the verify_callback that has been set
7140 specificly for an SSL object with SSL_set_verify() is actually being
7141 used. Before the change, a verify_callback set with this function was
7142 ignored and the verify_callback() set in the SSL_CTX at the time of
7143 the call was used. New function X509_STORE_CTX_set_verify_cb() introduced
7144 to allow the necessary settings.
7147 *) Initialize static variable in crypto/dsa/dsa_lib.c and crypto/dh/dh_lib.c
7148 explicitly to NULL, as at least on Solaris 8 this seems not always to be
7149 done automatically (in contradiction to the requirements of the C
7150 standard). This made problems when used from OpenSSH.
7153 *) In OpenSSL 0.9.6a and 0.9.6b, crypto/dh/dh_key.c ignored
7154 dh->length and always used
7156 BN_rand_range(priv_key, dh->p).
7158 BN_rand_range() is not necessary for Diffie-Hellman, and this
7159 specific range makes Diffie-Hellman unnecessarily inefficient if
7160 dh->length (recommended exponent length) is much smaller than the
7161 length of dh->p. We could use BN_rand_range() if the order of
7162 the subgroup was stored in the DH structure, but we only have
7167 BN_rand(priv_key, l, ...)
7169 where 'l' is dh->length if this is defined, or BN_num_bits(dh->p)-1
7175 RSA_eay_public_encrypt
7176 RSA_eay_private_decrypt
7177 RSA_eay_private_encrypt (signing)
7178 RSA_eay_public_decrypt (signature verification)
7180 (default implementations for RSA_public_encrypt,
7181 RSA_private_decrypt, RSA_private_encrypt, RSA_public_decrypt),
7182 always reject numbers >= n.
7185 *) In crypto/rand/md_rand.c, use a new short-time lock CRYPTO_LOCK_RAND2
7186 to synchronize access to 'locking_thread'. This is necessary on
7187 systems where access to 'locking_thread' (an 'unsigned long'
7188 variable) is not atomic.
7191 *) In crypto/rand/md_rand.c, set 'locking_thread' to current thread's ID
7192 *before* setting the 'crypto_lock_rand' flag. The previous code had
7193 a race condition if 0 is a valid thread ID.
7194 [Travis Vitek <vitek@roguewave.com>]
7196 *) Add support for shared libraries under Irix.
7197 [Albert Chin-A-Young <china@thewrittenword.com>]
7199 *) Add configuration option to build on Linux on both big-endian and
7201 [Ralf Baechle <ralf@uni-koblenz.de>]
7203 *) Add the possibility to create shared libraries on HP-UX.
7206 Changes between 0.9.6a and 0.9.6b [9 Jul 2001]
7208 *) Change ssleay_rand_bytes (crypto/rand/md_rand.c)
7209 to avoid a SSLeay/OpenSSL PRNG weakness pointed out by
7210 Markku-Juhani O. Saarinen <markku-juhani.saarinen@nokia.com>:
7211 PRNG state recovery was possible based on the output of
7212 one PRNG request appropriately sized to gain knowledge on
7213 'md' followed by enough consecutive 1-byte PRNG requests
7214 to traverse all of 'state'.
7216 1. When updating 'md_local' (the current thread's copy of 'md')
7217 during PRNG output generation, hash all of the previous
7218 'md_local' value, not just the half used for PRNG output.
7220 2. Make the number of bytes from 'state' included into the hash
7221 independent from the number of PRNG bytes requested.
7223 The first measure alone would be sufficient to avoid
7224 Markku-Juhani's attack. (Actually it had never occurred
7225 to me that the half of 'md_local' used for chaining was the
7226 half from which PRNG output bytes were taken -- I had always
7227 assumed that the secret half would be used.) The second
7228 measure makes sure that additional data from 'state' is never
7229 mixed into 'md_local' in small portions; this heuristically
7230 further strengthens the PRNG.
7233 *) Fix crypto/bn/asm/mips3.s.
7236 *) When only the key is given to "enc", the IV is undefined. Print out
7237 an error message in this case.
7240 *) Handle special case when X509_NAME is empty in X509 printing routines.
7243 *) In dsa_do_verify (crypto/dsa/dsa_ossl.c), verify that r and s are
7244 positive and less than q.
7247 *) Don't change *pointer in CRYPTO_add_lock() is add_lock_callback is
7248 used: it isn't thread safe and the add_lock_callback should handle
7250 [Paul Rose <Paul.Rose@bridge.com>]
7252 *) Verify that incoming data obeys the block size in
7253 ssl3_enc (ssl/s3_enc.c) and tls1_enc (ssl/t1_enc.c).
7257 [Ulf Möller, Bodo Möller]
7259 *) The countermeasure against Bleichbacher's attack on PKCS #1 v1.5
7260 RSA encryption was accidentally removed in s3_srvr.c in OpenSSL 0.9.5
7261 when fixing the server behaviour for backwards-compatible 'client
7262 hello' messages. (Note that the attack is impractical against
7263 SSL 3.0 and TLS 1.0 anyway because length and version checking
7264 means that the probability of guessing a valid ciphertext is
7265 around 2^-40; see section 5 in Bleichenbacher's CRYPTO '98
7268 Before 0.9.5, the countermeasure (hide the error by generating a
7269 random 'decryption result') did not work properly because
7270 ERR_clear_error() was missing, meaning that SSL_get_error() would
7271 detect the supposedly ignored error.
7273 Both problems are now fixed.
7276 *) In crypto/bio/bf_buff.c, increase DEFAULT_BUFFER_SIZE to 4096
7277 (previously it was 1024).
7280 *) Fix for compatibility mode trust settings: ignore trust settings
7281 unless some valid trust or reject settings are present.
7284 *) Fix for blowfish EVP: its a variable length cipher.
7287 *) Fix various bugs related to DSA S/MIME verification. Handle missing
7288 parameters in DSA public key structures and return an error in the
7289 DSA routines if parameters are absent.
7292 *) In versions up to 0.9.6, RAND_file_name() resorted to file ".rnd"
7293 in the current directory if neither $RANDFILE nor $HOME was set.
7294 RAND_file_name() in 0.9.6a returned NULL in this case. This has
7295 caused some confusion to Windows users who haven't defined $HOME.
7296 Thus RAND_file_name() is changed again: e_os.h can define a
7297 DEFAULT_HOME, which will be used if $HOME is not set.
7298 For Windows, we use "C:"; on other platforms, we still require
7299 environment variables.
7301 *) Move 'if (!initialized) RAND_poll()' into regions protected by
7302 CRYPTO_LOCK_RAND. This is not strictly necessary, but avoids
7303 having multiple threads call RAND_poll() concurrently.
7306 *) In crypto/rand/md_rand.c, replace 'add_do_not_lock' flag by a
7307 combination of a flag and a thread ID variable.
7308 Otherwise while one thread is in ssleay_rand_bytes (which sets the
7309 flag), *other* threads can enter ssleay_add_bytes without obeying
7310 the CRYPTO_LOCK_RAND lock (and may even illegally release the lock
7311 that they do not hold after the first thread unsets add_do_not_lock).
7314 *) Change bctest again: '-x' expressions are not available in all
7318 Changes between 0.9.6 and 0.9.6a [5 Apr 2001]
7320 *) Fix a couple of memory leaks in PKCS7_dataDecode()
7321 [Steve Henson, reported by Heyun Zheng <hzheng@atdsprint.com>]
7323 *) Change Configure and Makefiles to provide EXE_EXT, which will contain
7324 the default extension for executables, if any. Also, make the perl
7325 scripts that use symlink() to test if it really exists and use "cp"
7326 if it doesn't. All this made OpenSSL compilable and installable in
7330 *) Fix for asn1_GetSequence() for indefinite length constructed data.
7331 If SEQUENCE is length is indefinite just set c->slen to the total
7332 amount of data available.
7333 [Steve Henson, reported by shige@FreeBSD.org]
7334 [This change does not apply to 0.9.7.]
7336 *) Change bctest to avoid here-documents inside command substitution
7337 (workaround for FreeBSD /bin/sh bug).
7338 For compatibility with Ultrix, avoid shell functions (introduced
7339 in the bctest version that searches along $PATH).
7342 *) Rename 'des_encrypt' to 'des_encrypt1'. This avoids the clashes
7343 with des_encrypt() defined on some operating systems, like Solaris
7347 *) Check the result of RSA-CRT (see D. Boneh, R. DeMillo, R. Lipton:
7348 On the Importance of Eliminating Errors in Cryptographic
7349 Computations, J. Cryptology 14 (2001) 2, 101-119,
7350 http://theory.stanford.edu/~dabo/papers/faults.ps.gz).
7353 *) MIPS assembler BIGNUM division bug fix.
7356 *) Disabled incorrect Alpha assembler code.
7359 *) Fix PKCS#7 decode routines so they correctly update the length
7360 after reading an EOC for the EXPLICIT tag.
7362 [This change does not apply to 0.9.7.]
7364 *) Fix bug in PKCS#12 key generation routines. This was triggered
7365 if a 3DES key was generated with a 0 initial byte. Include
7366 PKCS12_BROKEN_KEYGEN compilation option to retain the old
7367 (but broken) behaviour.
7370 *) Enhance bctest to search for a working bc along $PATH and print
7372 [Tim Rice <tim@multitalents.net> via Richard Levitte]
7374 *) Fix memory leaks in err.c: free err_data string if necessary;
7375 don't write to the wrong index in ERR_set_error_data.
7378 *) Implement ssl23_peek (analogous to ssl23_read), which previously
7382 *) Replace rdtsc with _emit statements for VC++ version 5.
7383 [Jeremy Cooper <jeremy@baymoo.org>]
7385 *) Make it possible to reuse SSLv2 sessions.
7388 *) In copy_email() check for >= 0 as a return value for
7389 X509_NAME_get_index_by_NID() since 0 is a valid index.
7390 [Steve Henson reported by Massimiliano Pala <madwolf@opensca.org>]
7392 *) Avoid coredump with unsupported or invalid public keys by checking if
7393 X509_get_pubkey() fails in PKCS7_verify(). Fix memory leak when
7394 PKCS7_verify() fails with non detached data.
7397 *) Don't use getenv in library functions when run as setuid/setgid.
7398 New function OPENSSL_issetugid().
7401 *) Avoid false positives in memory leak detection code (crypto/mem_dbg.c)
7402 due to incorrect handling of multi-threading:
7404 1. Fix timing glitch in the MemCheck_off() portion of CRYPTO_mem_ctrl().
7406 2. Fix logical glitch in is_MemCheck_on() aka CRYPTO_is_mem_check_on().
7408 3. Count how many times MemCheck_off() has been called so that
7409 nested use can be treated correctly. This also avoids
7410 inband-signalling in the previous code (which relied on the
7411 assumption that thread ID 0 is impossible).
7414 *) Add "-rand" option also to s_client and s_server.
7417 *) Fix CPU detection on Irix 6.x.
7418 [Kurt Hockenbury <khockenb@stevens-tech.edu> and
7419 "Bruce W. Forsberg" <bruce.forsberg@baesystems.com>]
7421 *) Fix X509_NAME bug which produced incorrect encoding if X509_NAME
7424 [This change does not apply to 0.9.7.]
7426 *) Use the cached encoding of an X509_NAME structure rather than
7427 copying it. This is apparently the reason for the libsafe "errors"
7428 but the code is actually correct.
7431 *) Add new function BN_rand_range(), and fix DSA_sign_setup() to prevent
7432 Bleichenbacher's DSA attack.
7433 Extend BN_[pseudo_]rand: As before, top=1 forces the highest two bits
7434 to be set and top=0 forces the highest bit to be set; top=-1 is new
7435 and leaves the highest bit random.
7436 [Ulf Moeller, Bodo Moeller]
7438 *) In the NCONF_...-based implementations for CONF_... queries
7439 (crypto/conf/conf_lib.c), if the input LHASH is NULL, avoid using
7440 a temporary CONF structure with the data component set to NULL
7441 (which gives segmentation faults in lh_retrieve).
7442 Instead, use NULL for the CONF pointer in CONF_get_string and
7443 CONF_get_number (which may use environment variables) and directly
7444 return NULL from CONF_get_section.
7447 *) Fix potential buffer overrun for EBCDIC.
7450 *) Tolerate nonRepudiation as being valid for S/MIME signing and certSign
7451 keyUsage if basicConstraints absent for a CA.
7454 *) Make SMIME_write_PKCS7() write mail header values with a format that
7455 is more generally accepted (no spaces before the semicolon), since
7456 some programs can't parse those values properly otherwise. Also make
7457 sure BIO's that break lines after each write do not create invalid
7461 *) Make the CRL encoding routines work with empty SEQUENCE OF. The
7462 macros previously used would not encode an empty SEQUENCE OF
7463 and break the signature.
7465 [This change does not apply to 0.9.7.]
7467 *) Zero the premaster secret after deriving the master secret in
7471 *) Add some EVP_add_digest_alias registrations (as found in
7472 OpenSSL_add_all_digests()) to SSL_library_init()
7473 aka OpenSSL_add_ssl_algorithms(). This provides improved
7474 compatibility with peers using X.509 certificates
7475 with unconventional AlgorithmIdentifier OIDs.
7478 *) Fix for Irix with NO_ASM.
7479 ["Bruce W. Forsberg" <bruce.forsberg@baesystems.com>]
7481 *) ./config script fixes.
7482 [Ulf Moeller, Richard Levitte]
7484 *) Fix 'openssl passwd -1'.
7487 *) Change PKCS12_key_gen_asc() so it can cope with non null
7488 terminated strings whose length is passed in the passlen
7489 parameter, for example from PEM callbacks. This was done
7490 by adding an extra length parameter to asc2uni().
7491 [Steve Henson, reported by <oddissey@samsung.co.kr>]
7493 *) Fix C code generated by 'openssl dsaparam -C': If a BN_bin2bn
7494 call failed, free the DSA structure.
7497 *) Fix to uni2asc() to cope with zero length Unicode strings.
7498 These are present in some PKCS#12 files.
7501 *) Increase s2->wbuf allocation by one byte in ssl2_new (ssl/s2_lib.c).
7502 Otherwise do_ssl_write (ssl/s2_pkt.c) will write beyond buffer limits
7503 when writing a 32767 byte record.
7504 [Bodo Moeller; problem reported by Eric Day <eday@concentric.net>]
7506 *) In RSA_eay_public_{en,ed}crypt and RSA_eay_mod_exp (rsa_eay.c),
7507 obtain lock CRYPTO_LOCK_RSA before setting rsa->_method_mod_{n,p,q}.
7509 (RSA objects have a reference count access to which is protected
7510 by CRYPTO_LOCK_RSA [see rsa_lib.c, s3_srvr.c, ssl_cert.c, ssl_rsa.c],
7511 so they are meant to be shared between threads.)
7512 [Bodo Moeller, Geoff Thorpe; original patch submitted by
7513 "Reddie, Steven" <Steven.Reddie@ca.com>]
7515 *) Fix a deadlock in CRYPTO_mem_leaks().
7518 *) Use better test patterns in bntest.
7521 *) rand_win.c fix for Borland C.
7524 *) BN_rshift bugfix for n == 0.
7527 *) Add a 'bctest' script that checks for some known 'bc' bugs
7528 so that 'make test' does not abort just because 'bc' is broken.
7531 *) Store verify_result within SSL_SESSION also for client side to
7532 avoid potential security hole. (Re-used sessions on the client side
7533 always resulted in verify_result==X509_V_OK, not using the original
7534 result of the server certificate verification.)
7537 *) Fix ssl3_pending: If the record in s->s3->rrec is not of type
7538 SSL3_RT_APPLICATION_DATA, return 0.
7539 Similarly, change ssl2_pending to return 0 if SSL_in_init(s) is true.
7543 Both ssl2_peek and ssl3_peek, which were totally broken in earlier
7544 releases, have been re-implemented by renaming the previous
7545 implementations of ssl2_read and ssl3_read to ssl2_read_internal
7546 and ssl3_read_internal, respectively, and adding 'peek' parameters
7547 to them. The new ssl[23]_{read,peek} functions are calls to
7548 ssl[23]_read_internal with the 'peek' flag set appropriately.
7549 A 'peek' parameter has also been added to ssl3_read_bytes, which
7550 does the actual work for ssl3_read_internal.
7553 *) Initialise "ex_data" member of RSA/DSA/DH structures prior to calling
7554 the method-specific "init()" handler. Also clean up ex_data after
7555 calling the method-specific "finish()" handler. Previously, this was
7556 happening the other way round.
7559 *) Increase BN_CTX_NUM (the number of BIGNUMs in a BN_CTX) to 16.
7560 The previous value, 12, was not always sufficient for BN_mod_exp().
7563 *) Make sure that shared libraries get the internal name engine with
7564 the full version number and not just 0. This should mark the
7565 shared libraries as not backward compatible. Of course, this should
7566 be changed again when we can guarantee backward binary compatibility.
7569 *) Fix typo in get_cert_by_subject() in by_dir.c
7570 [Jean-Marc Desperrier <jean-marc.desperrier@certplus.com>]
7572 *) Rework the system to generate shared libraries:
7574 - Make note of the expected extension for the shared libraries and
7575 if there is a need for symbolic links from for example libcrypto.so.0
7576 to libcrypto.so.0.9.7. There is extended info in Configure for
7579 - Make as few rebuilds of the shared libraries as possible.
7581 - Still avoid linking the OpenSSL programs with the shared libraries.
7583 - When installing, install the shared libraries separately from the
7587 *) Fix SSL_CTX_set_read_ahead macro to actually use its argument.
7589 Copy SSL_CTX's read_ahead flag to SSL object directly in SSL_new
7590 and not in SSL_clear because the latter is also used by the
7591 accept/connect functions; previously, the settings made by
7592 SSL_set_read_ahead would be lost during the handshake.
7593 [Bodo Moeller; problems reported by Anders Gertz <gertz@epact.se>]
7595 *) Correct util/mkdef.pl to be selective about disabled algorithms.
7596 Previously, it would create entries for disableed algorithms no
7600 *) Added several new manual pages for SSL_* function.
7603 Changes between 0.9.5a and 0.9.6 [24 Sep 2000]
7605 *) In ssl23_get_client_hello, generate an error message when faced
7606 with an initial SSL 3.0/TLS record that is too small to contain the
7607 first two bytes of the ClientHello message, i.e. client_version.
7608 (Note that this is a pathologic case that probably has never happened
7609 in real life.) The previous approach was to use the version number
7610 from the record header as a substitute; but our protocol choice
7611 should not depend on that one because it is not authenticated
7612 by the Finished messages.
7615 *) More robust randomness gathering functions for Windows.
7616 [Jeffrey Altman <jaltman@columbia.edu>]
7618 *) For compatibility reasons if the flag X509_V_FLAG_ISSUER_CHECK is
7619 not set then we don't setup the error code for issuer check errors
7620 to avoid possibly overwriting other errors which the callback does
7621 handle. If an application does set the flag then we assume it knows
7622 what it is doing and can handle the new informational codes
7626 *) Fix for a nasty bug in ASN1_TYPE handling. ASN1_TYPE is used for
7627 a general "ANY" type, as such it should be able to decode anything
7628 including tagged types. However it didn't check the class so it would
7629 wrongly interpret tagged types in the same way as their universal
7630 counterpart and unknown types were just rejected. Changed so that the
7631 tagged and unknown types are handled in the same way as a SEQUENCE:
7632 that is the encoding is stored intact. There is also a new type
7633 "V_ASN1_OTHER" which is used when the class is not universal, in this
7634 case we have no idea what the actual type is so we just lump them all
7638 *) On VMS, stdout may very well lead to a file that is written to
7639 in a record-oriented fashion. That means that every write() will
7640 write a separate record, which will be read separately by the
7641 programs trying to read from it. This can be very confusing.
7643 The solution is to put a BIO filter in the way that will buffer
7644 text until a linefeed is reached, and then write everything a
7645 line at a time, so every record written will be an actual line,
7646 not chunks of lines and not (usually doesn't happen, but I've
7647 seen it once) several lines in one record. BIO_f_linebuffer() is
7650 Currently, it's a VMS-only method, because that's where it has
7651 been tested well enough.
7654 *) Remove 'optimized' squaring variant in BN_mod_mul_montgomery,
7655 it can return incorrect results.
7656 (Note: The buggy variant was not enabled in OpenSSL 0.9.5a,
7657 but it was in 0.9.6-beta[12].)
7660 *) Disable the check for content being present when verifying detached
7661 signatures in pk7_smime.c. Some versions of Netscape (wrongly)
7662 include zero length content when signing messages.
7665 *) New BIO_shutdown_wr macro, which invokes the BIO_C_SHUTDOWN_WR
7666 BIO_ctrl (for BIO pairs).
7669 *) Add DSO method for VMS.
7672 *) Bug fix: Montgomery multiplication could produce results with the
7676 *) Add RPM specification openssl.spec and modify it to build three
7677 packages. The default package contains applications, application
7678 documentation and run-time libraries. The devel package contains
7679 include files, static libraries and function documentation. The
7680 doc package contains the contents of the doc directory. The original
7681 openssl.spec was provided by Damien Miller <djm@mindrot.org>.
7684 *) Add a large number of documentation files for many SSL routines.
7685 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
7687 *) Add a configuration entry for Sony News 4.
7688 [NAKAJI Hiroyuki <nakaji@tutrp.tut.ac.jp>]
7690 *) Don't set the two most significant bits to one when generating a
7691 random number < q in the DSA library.
7694 *) New SSL API mode 'SSL_MODE_AUTO_RETRY'. This disables the default
7695 behaviour that SSL_read may result in SSL_ERROR_WANT_READ (even if
7696 the underlying transport is blocking) if a handshake took place.
7697 (The default behaviour is needed by applications such as s_client
7698 and s_server that use select() to determine when to use SSL_read;
7699 but for applications that know in advance when to expect data, it
7700 just makes things more complicated.)
7703 *) Add RAND_egd_bytes(), which gives control over the number of bytes read
7707 *) Add a few more EBCDIC conditionals that make `req' and `x509'
7708 work better on such systems.
7709 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
7711 *) Add two demo programs for PKCS12_parse() and PKCS12_create().
7712 Update PKCS12_parse() so it copies the friendlyName and the
7713 keyid to the certificates aux info.
7716 *) Fix bug in PKCS7_verify() which caused an infinite loop
7717 if there was more than one signature.
7718 [Sven Uszpelkat <su@celocom.de>]
7720 *) Major change in util/mkdef.pl to include extra information
7721 about each symbol, as well as presenting variables as well
7722 as functions. This change means that there's n more need
7723 to rebuild the .num files when some algorithms are excluded.
7726 *) Allow the verify time to be set by an application,
7727 rather than always using the current time.
7730 *) Phase 2 verify code reorganisation. The certificate
7731 verify code now looks up an issuer certificate by a
7732 number of criteria: subject name, authority key id
7733 and key usage. It also verifies self signed certificates
7734 by the same criteria. The main comparison function is
7735 X509_check_issued() which performs these checks.
7737 Lot of changes were necessary in order to support this
7738 without completely rewriting the lookup code.
7740 Authority and subject key identifier are now cached.
7742 The LHASH 'certs' is X509_STORE has now been replaced
7743 by a STACK_OF(X509_OBJECT). This is mainly because an
7744 LHASH can't store or retrieve multiple objects with
7745 the same hash value.
7747 As a result various functions (which were all internal
7748 use only) have changed to handle the new X509_STORE
7749 structure. This will break anything that messed round
7750 with X509_STORE internally.
7752 The functions X509_STORE_add_cert() now checks for an
7753 exact match, rather than just subject name.
7755 The X509_STORE API doesn't directly support the retrieval
7756 of multiple certificates matching a given criteria, however
7757 this can be worked round by performing a lookup first
7758 (which will fill the cache with candidate certificates)
7759 and then examining the cache for matches. This is probably
7760 the best we can do without throwing out X509_LOOKUP
7761 entirely (maybe later...).
7763 The X509_VERIFY_CTX structure has been enhanced considerably.
7765 All certificate lookup operations now go via a get_issuer()
7766 callback. Although this currently uses an X509_STORE it
7767 can be replaced by custom lookups. This is a simple way
7768 to bypass the X509_STORE hackery necessary to make this
7769 work and makes it possible to use more efficient techniques
7770 in future. A very simple version which uses a simple
7771 STACK for its trusted certificate store is also provided
7772 using X509_STORE_CTX_trusted_stack().
7774 The verify_cb() and verify() callbacks now have equivalents
7775 in the X509_STORE_CTX structure.
7777 X509_STORE_CTX also has a 'flags' field which can be used
7778 to customise the verify behaviour.
7781 *) Add new PKCS#7 signing option PKCS7_NOSMIMECAP which
7782 excludes S/MIME capabilities.
7785 *) When a certificate request is read in keep a copy of the
7786 original encoding of the signed data and use it when outputing
7787 again. Signatures then use the original encoding rather than
7788 a decoded, encoded version which may cause problems if the
7789 request is improperly encoded.
7792 *) For consistency with other BIO_puts implementations, call
7793 buffer_write(b, ...) directly in buffer_puts instead of calling
7796 In BIO_puts, increment b->num_write as in BIO_write.
7797 [Peter.Sylvester@EdelWeb.fr]
7799 *) Fix BN_mul_word for the case where the word is 0. (We have to use
7800 BN_zero, we may not return a BIGNUM with an array consisting of
7804 *) Avoid calling abort() from within the library when problems are
7805 detected, except if preprocessor symbols have been defined
7806 (such as REF_CHECK, BN_DEBUG etc.).
7809 *) New openssl application 'rsautl'. This utility can be
7810 used for low level RSA operations. DER public key
7811 BIO/fp routines also added.
7814 *) New Configure entry and patches for compiling on QNX 4.
7815 [Andreas Schneider <andreas@ds3.etech.fh-hamburg.de>]
7817 *) A demo state-machine implementation was sponsored by
7818 Nuron (http://www.nuron.com/) and is now available in
7819 demos/state_machine.
7822 *) New options added to the 'dgst' utility for signature
7823 generation and verification.
7826 *) Unrecognized PKCS#7 content types are now handled via a
7827 catch all ASN1_TYPE structure. This allows unsupported
7828 types to be stored as a "blob" and an application can
7829 encode and decode it manually.
7832 *) Fix various signed/unsigned issues to make a_strex.c
7834 [Oscar Jacobsson <oscar.jacobsson@celocom.com>]
7836 *) ASN1 fixes. i2d_ASN1_OBJECT was not returning the correct
7837 length if passed a buffer. ASN1_INTEGER_to_BN failed
7838 if passed a NULL BN and its argument was negative.
7839 [Steve Henson, pointed out by Sven Heiberg <sven@tartu.cyber.ee>]
7841 *) Modification to PKCS#7 encoding routines to output definite
7842 length encoding. Since currently the whole structures are in
7843 memory there's not real point in using indefinite length
7844 constructed encoding. However if OpenSSL is compiled with
7845 the flag PKCS7_INDEFINITE_ENCODING the old form is used.
7848 *) Added BIO_vprintf() and BIO_vsnprintf().
7851 *) Added more prefixes to parse for in the the strings written
7852 through a logging bio, to cover all the levels that are available
7853 through syslog. The prefixes are now:
7855 PANIC, EMERG, EMR => LOG_EMERG
7856 ALERT, ALR => LOG_ALERT
7857 CRIT, CRI => LOG_CRIT
7858 ERROR, ERR => LOG_ERR
7859 WARNING, WARN, WAR => LOG_WARNING
7860 NOTICE, NOTE, NOT => LOG_NOTICE
7861 INFO, INF => LOG_INFO
7862 DEBUG, DBG => LOG_DEBUG
7864 and as before, if none of those prefixes are present at the
7865 beginning of the string, LOG_ERR is chosen.
7867 On Win32, the LOG_* levels are mapped according to this:
7869 LOG_EMERG, LOG_ALERT, LOG_CRIT, LOG_ERR => EVENTLOG_ERROR_TYPE
7870 LOG_WARNING => EVENTLOG_WARNING_TYPE
7871 LOG_NOTICE, LOG_INFO, LOG_DEBUG => EVENTLOG_INFORMATION_TYPE
7875 *) Made it possible to reconfigure with just the configuration
7876 argument "reconf" or "reconfigure". The command line arguments
7877 are stored in Makefile.ssl in the variable CONFIGURE_ARGS,
7878 and are retrieved from there when reconfiguring.
7882 [Assar Westerlund <assar@sics.se>, Richard Levitte]
7884 *) Add the arguments -CAfile and -CApath to the pkcs12 utility.
7887 *) The obj_dat.pl script was messing up the sorting of object
7888 names. The reason was that it compared the quoted version
7889 of strings as a result "OCSP" > "OCSP Signing" because
7890 " > SPACE. Changed script to store unquoted versions of
7891 names and add quotes on output. It was also omitting some
7892 names from the lookup table if they were given a default
7893 value (that is if SN is missing it is given the same
7894 value as LN and vice versa), these are now added on the
7895 grounds that if an object has a name we should be able to
7896 look it up. Finally added warning output when duplicate
7897 short or long names are found.
7900 *) Changes needed for Tandem NSK.
7901 [Scott Uroff <scott@xypro.com>]
7903 *) Fix SSL 2.0 rollback checking: Due to an off-by-one error in
7904 RSA_padding_check_SSLv23(), special padding was never detected
7905 and thus the SSL 3.0/TLS 1.0 countermeasure against protocol
7906 version rollback attacks was not effective.
7908 In s23_clnt.c, don't use special rollback-attack detection padding
7909 (RSA_SSLV23_PADDING) if SSL 2.0 is the only protocol enabled in the
7910 client; similarly, in s23_srvr.c, don't do the rollback check if
7911 SSL 2.0 is the only protocol enabled in the server.
7914 *) Make it possible to get hexdumps of unprintable data with 'openssl
7915 asn1parse'. By implication, the functions ASN1_parse_dump() and
7916 BIO_dump_indent() are added.
7919 *) New functions ASN1_STRING_print_ex() and X509_NAME_print_ex()
7920 these print out strings and name structures based on various
7921 flags including RFC2253 support and proper handling of
7922 multibyte characters. Added options to the 'x509' utility
7923 to allow the various flags to be set.
7926 *) Various fixes to use ASN1_TIME instead of ASN1_UTCTIME.
7927 Also change the functions X509_cmp_current_time() and
7928 X509_gmtime_adj() work with an ASN1_TIME structure,
7929 this will enable certificates using GeneralizedTime in validity
7930 dates to be checked.
7933 *) Make the NEG_PUBKEY_BUG code (which tolerates invalid
7934 negative public key encodings) on by default,
7935 NO_NEG_PUBKEY_BUG can be set to disable it.
7938 *) New function c2i_ASN1_OBJECT() which acts on ASN1_OBJECT
7939 content octets. An i2c_ASN1_OBJECT is unnecessary because
7940 the encoding can be trivially obtained from the structure.
7943 *) crypto/err.c locking bugfix: Use write locks (CRYPTO_w_[un]lock),
7944 not read locks (CRYPTO_r_[un]lock).
7947 *) A first attempt at creating official support for shared
7948 libraries through configuration. I've kept it so the
7949 default is static libraries only, and the OpenSSL programs
7950 are always statically linked for now, but there are
7951 preparations for dynamic linking in place.
7952 This has been tested on Linux and Tru64.
7955 *) Randomness polling function for Win9x, as described in:
7956 Peter Gutmann, Software Generation of Practically Strong
7960 *) Fix so PRNG is seeded in req if using an already existing
7964 *) New options to smime application. -inform and -outform
7965 allow alternative formats for the S/MIME message including
7966 PEM and DER. The -content option allows the content to be
7967 specified separately. This should allow things like Netscape
7968 form signing output easier to verify.
7971 *) Fix the ASN1 encoding of tags using the 'long form'.
7974 *) New ASN1 functions, i2c_* and c2i_* for INTEGER and BIT
7975 STRING types. These convert content octets to and from the
7976 underlying type. The actual tag and length octets are
7977 already assumed to have been read in and checked. These
7978 are needed because all other string types have virtually
7979 identical handling apart from the tag. By having versions
7980 of the ASN1 functions that just operate on content octets
7981 IMPLICIT tagging can be handled properly. It also allows
7982 the ASN1_ENUMERATED code to be cut down because ASN1_ENUMERATED
7983 and ASN1_INTEGER are identical apart from the tag.
7986 *) Change the handling of OID objects as follows:
7988 - New object identifiers are inserted in objects.txt, following
7989 the syntax given in objects.README.
7990 - objects.pl is used to process obj_mac.num and create a new
7992 - obj_dat.pl is used to create a new obj_dat.h, using the data in
7995 This is currently kind of a hack, and the perl code in objects.pl
7996 isn't very elegant, but it works as I intended. The simplest way
7997 to check that it worked correctly is to look in obj_dat.h and
7998 check the array nid_objs and make sure the objects haven't moved
7999 around (this is important!). Additions are OK, as well as
8000 consistent name changes.
8003 *) Add BSD-style MD5-based passwords to 'openssl passwd' (option '-1').
8006 *) Addition of the command line parameter '-rand file' to 'openssl req'.
8007 The given file adds to whatever has already been seeded into the
8008 random pool through the RANDFILE configuration file option or
8009 environment variable, or the default random state file.
8012 *) mkstack.pl now sorts each macro group into lexical order.
8013 Previously the output order depended on the order the files
8014 appeared in the directory, resulting in needless rewriting
8018 *) Patches to make OpenSSL compile under Win32 again. Mostly
8019 work arounds for the VC++ problem that it treats func() as
8020 func(void). Also stripped out the parts of mkdef.pl that
8021 added extra typesafe functions: these no longer exist.
8024 *) Reorganisation of the stack code. The macros are now all
8025 collected in safestack.h . Each macro is defined in terms of
8026 a "stack macro" of the form SKM_<name>(type, a, b). The
8027 DEBUG_SAFESTACK is now handled in terms of function casts,
8028 this has the advantage of retaining type safety without the
8029 use of additional functions. If DEBUG_SAFESTACK is not defined
8030 then the non typesafe macros are used instead. Also modified the
8031 mkstack.pl script to handle the new form. Needs testing to see
8032 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK
8033 the default if no major problems. Similar behaviour for ASN1_SET_OF
8034 and PKCS12_STACK_OF.
8037 *) When some versions of IIS use the 'NET' form of private key the
8038 key derivation algorithm is different. Normally MD5(password) is
8039 used as a 128 bit RC4 key. In the modified case
8040 MD5(MD5(password) + "SGCKEYSALT") is used instead. Added some
8041 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same
8042 as the old Netscape_RSA functions except they have an additional
8043 'sgckey' parameter which uses the modified algorithm. Also added
8044 an -sgckey command line option to the rsa utility. Thanks to
8045 Adrian Peck <bertie@ncipher.com> for posting details of the modified
8046 algorithm to openssl-dev.
8049 *) The evp_local.h macros were using 'c.##kname' which resulted in
8050 invalid expansion on some systems (SCO 5.0.5 for example).
8051 Corrected to 'c.kname'.
8052 [Phillip Porch <root@theporch.com>]
8054 *) New X509_get1_email() and X509_REQ_get1_email() functions that return
8055 a STACK of email addresses from a certificate or request, these look
8056 in the subject name and the subject alternative name extensions and
8057 omit any duplicate addresses.
8060 *) Re-implement BN_mod_exp2_mont using independent (and larger) windows.
8061 This makes DSA verification about 2 % faster.
8064 *) Increase maximum window size in BN_mod_exp_... to 6 bits instead of 5
8065 (meaning that now 2^5 values will be precomputed, which is only 4 KB
8066 plus overhead for 1024 bit moduli).
8067 This makes exponentiations about 0.5 % faster for 1024 bit
8068 exponents (as measured by "openssl speed rsa2048").
8071 *) Rename memory handling macros to avoid conflicts with other
8073 Malloc => OPENSSL_malloc
8074 Malloc_locked => OPENSSL_malloc_locked
8075 Realloc => OPENSSL_realloc
8076 Free => OPENSSL_free
8079 *) New function BN_mod_exp_mont_word for small bases (roughly 15%
8080 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange).
8083 *) CygWin32 support.
8084 [John Jarvie <jjarvie@newsguy.com>]
8086 *) The type-safe stack code has been rejigged. It is now only compiled
8087 in when OpenSSL is configured with the DEBUG_SAFESTACK option and
8088 by default all type-specific stack functions are "#define"d back to
8089 standard stack functions. This results in more streamlined output
8090 but retains the type-safety checking possibilities of the original
8094 *) The STACK code has been cleaned up, and certain type declarations
8095 that didn't make a lot of sense have been brought in line. This has
8096 also involved a cleanup of sorts in safestack.h to more correctly
8097 map type-safe stack functions onto their plain stack counterparts.
8098 This work has also resulted in a variety of "const"ifications of
8099 lots of the code, especially "_cmp" operations which should normally
8100 be prototyped with "const" parameters anyway.
8103 *) When generating bytes for the first time in md_rand.c, 'stir the pool'
8104 by seeding with STATE_SIZE dummy bytes (with zero entropy count).
8105 (The PRNG state consists of two parts, the large pool 'state' and 'md',
8106 where all of 'md' is used each time the PRNG is used, but 'state'
8107 is used only indexed by a cyclic counter. As entropy may not be
8108 well distributed from the beginning, 'md' is important as a
8109 chaining variable. However, the output function chains only half
8110 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains
8111 all of 'md', and seeding with STATE_SIZE dummy bytes will result
8112 in all of 'state' being rewritten, with the new values depending
8113 on virtually all of 'md'. This overcomes the 80 bit limitation.)
8116 *) In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when
8117 the handshake is continued after ssl_verify_cert_chain();
8118 otherwise, if SSL_VERIFY_NONE is set, remaining error codes
8119 can lead to 'unexplainable' connection aborts later.
8120 [Bodo Moeller; problem tracked down by Lutz Jaenicke]
8122 *) Major EVP API cipher revision.
8123 Add hooks for extra EVP features. This allows various cipher
8124 parameters to be set in the EVP interface. Support added for variable
8125 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and
8126 setting of RC2 and RC5 parameters.
8128 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length
8131 Remove lots of duplicated code from the EVP library. For example *every*
8132 cipher init() function handles the 'iv' in the same way according to the
8133 cipher mode. They also all do nothing if the 'key' parameter is NULL and
8134 for CFB and OFB modes they zero ctx->num.
8136 New functionality allows removal of S/MIME code RC2 hack.
8138 Most of the routines have the same form and so can be declared in terms
8141 By shifting this to the top level EVP_CipherInit() it can be removed from
8142 all individual ciphers. If the cipher wants to handle IVs or keys
8143 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT
8146 Change lots of functions like EVP_EncryptUpdate() to now return a
8147 value: although software versions of the algorithms cannot fail
8148 any installed hardware versions can.
8151 *) Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if
8152 this option is set, tolerate broken clients that send the negotiated
8153 protocol version number instead of the requested protocol version
8157 *) Call dh_tmp_cb (set by ..._TMP_DH_CB) with correct 'is_export' flag;
8158 i.e. non-zero for export ciphersuites, zero otherwise.
8159 Previous versions had this flag inverted, inconsistent with
8160 rsa_tmp_cb (..._TMP_RSA_CB).
8161 [Bodo Moeller; problem reported by Amit Chopra]
8163 *) Add missing DSA library text string. Work around for some IIS
8164 key files with invalid SEQUENCE encoding.
8167 *) Add a document (doc/standards.txt) that list all kinds of standards
8168 and so on that are implemented in OpenSSL.
8171 *) Enhance c_rehash script. Old version would mishandle certificates
8172 with the same subject name hash and wouldn't handle CRLs at all.
8173 Added -fingerprint option to crl utility, to support new c_rehash
8177 *) Eliminate non-ANSI declarations in crypto.h and stack.h.
8180 *) Fix for SSL server purpose checking. Server checking was
8181 rejecting certificates which had extended key usage present
8182 but no ssl client purpose.
8183 [Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>]
8185 *) Make PKCS#12 code work with no password. The PKCS#12 spec
8186 is a little unclear about how a blank password is handled.
8187 Since the password in encoded as a BMPString with terminating
8188 double NULL a zero length password would end up as just the
8189 double NULL. However no password at all is different and is
8190 handled differently in the PKCS#12 key generation code. NS
8191 treats a blank password as zero length. MSIE treats it as no
8192 password on export: but it will try both on import. We now do
8193 the same: PKCS12_parse() tries zero length and no password if
8194 the password is set to "" or NULL (NULL is now a valid password:
8195 it wasn't before) as does the pkcs12 application.
8198 *) Bugfixes in apps/x509.c: Avoid a memory leak; and don't use
8199 perror when PEM_read_bio_X509_REQ fails, the error message must
8200 be obtained from the error queue.
8203 *) Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing
8204 it in ERR_remove_state if appropriate, and change ERR_get_state
8205 accordingly to avoid race conditions (this is necessary because
8206 thread_hash is no longer constant once set).
8209 *) Bugfix for linux-elf makefile.one.
8212 *) RSA_get_default_method() will now cause a default
8213 RSA_METHOD to be chosen if one doesn't exist already.
8214 Previously this was only set during a call to RSA_new()
8215 or RSA_new_method(NULL) meaning it was possible for
8216 RSA_get_default_method() to return NULL.
8219 *) Added native name translation to the existing DSO code
8220 that will convert (if the flag to do so is set) filenames
8221 that are sufficiently small and have no path information
8222 into a canonical native form. Eg. "blah" converted to
8223 "libblah.so" or "blah.dll" etc.
8226 *) New function ERR_error_string_n(e, buf, len) which is like
8227 ERR_error_string(e, buf), but writes at most 'len' bytes
8228 including the 0 terminator. For ERR_error_string_n, 'buf'
8230 [Damien Miller <djm@mindrot.org>, Bodo Moeller]
8232 *) CONF library reworked to become more general. A new CONF
8233 configuration file reader "class" is implemented as well as a
8234 new functions (NCONF_*, for "New CONF") to handle it. The now
8235 old CONF_* functions are still there, but are reimplemented to
8236 work in terms of the new functions. Also, a set of functions
8237 to handle the internal storage of the configuration data is
8238 provided to make it easier to write new configuration file
8239 reader "classes" (I can definitely see something reading a
8240 configuration file in XML format, for example), called _CONF_*,
8241 or "the configuration storage API"...
8243 The new configuration file reading functions are:
8245 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio,
8246 NCONF_get_section, NCONF_get_string, NCONF_get_numbre
8248 NCONF_default, NCONF_WIN32
8250 NCONF_dump_fp, NCONF_dump_bio
8252 NCONF_default and NCONF_WIN32 are method (or "class") choosers,
8253 NCONF_new creates a new CONF object. This works in the same way
8254 as other interfaces in OpenSSL, like the BIO interface.
8255 NCONF_dump_* dump the internal storage of the configuration file,
8256 which is useful for debugging. All other functions take the same
8257 arguments as the old CONF_* functions wth the exception of the
8258 first that must be a `CONF *' instead of a `LHASH *'.
8260 To make it easer to use the new classes with the old CONF_* functions,
8261 the function CONF_set_default_method is provided.
8264 *) Add '-tls1' option to 'openssl ciphers', which was already
8265 mentioned in the documentation but had not been implemented.
8266 (This option is not yet really useful because even the additional
8267 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.)
8270 *) Initial DSO code added into libcrypto for letting OpenSSL (and
8271 OpenSSL-based applications) load shared libraries and bind to
8272 them in a portable way.
8273 [Geoff Thorpe, with contributions from Richard Levitte]
8275 Changes between 0.9.5 and 0.9.5a [1 Apr 2000]
8277 *) Make sure _lrotl and _lrotr are only used with MSVC.
8279 *) Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status
8280 (the default implementation of RAND_status).
8282 *) Rename openssl x509 option '-crlext', which was added in 0.9.5,
8283 to '-clrext' (= clear extensions), as intended and documented.
8284 [Bodo Moeller; inconsistency pointed out by Michael Attili
8287 *) Fix for HMAC. It wasn't zeroing the rest of the block if the key length
8288 was larger than the MD block size.
8289 [Steve Henson, pointed out by Yost William <YostW@tce.com>]
8291 *) Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument
8292 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set()
8293 using the passed key: if the passed key was a private key the result
8294 of X509_print(), for example, would be to print out all the private key
8298 *) des_quad_cksum() byte order bug fix.
8299 [Ulf Möller, using the problem description in krb4-0.9.7, where
8300 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>]
8302 *) Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly
8304 [Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>]
8306 *) For easily testing in shell scripts whether some command
8307 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX'
8308 returns with exit code 0 iff no command of the given name is available.
8309 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases,
8310 the output goes to stdout and nothing is printed to stderr.
8311 Additional arguments are always ignored.
8313 Since for each cipher there is a command of the same name,
8314 the 'no-cipher' compilation switches can be tested this way.
8316 ('openssl no-XXX' is not able to detect pseudo-commands such
8317 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.)
8320 *) Update test suite so that 'make test' succeeds in 'no-rsa' configuration.
8323 *) For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE
8324 is set; it will be thrown away anyway because each handshake creates
8326 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition
8327 to parameters -- in previous versions (since OpenSSL 0.9.3) the
8328 'default key' from SSL_CTX_set_tmp_dh would always be lost, meaning
8329 you effectivly got SSL_OP_SINGLE_DH_USE when using this macro.
8332 *) New s_client option -ign_eof: EOF at stdin is ignored, and
8333 'Q' and 'R' lose their special meanings (quit/renegotiate).
8334 This is part of what -quiet does; unlike -quiet, -ign_eof
8335 does not suppress any output.
8338 *) Add compatibility options to the purpose and trust code. The
8339 purpose X509_PURPOSE_ANY is "any purpose" which automatically
8340 accepts a certificate or CA, this was the previous behaviour,
8341 with all the associated security issues.
8343 X509_TRUST_COMPAT is the old trust behaviour: only and
8344 automatically trust self signed roots in certificate store. A
8345 new trust setting X509_TRUST_DEFAULT is used to specify that
8346 a purpose has no associated trust setting and it should instead
8347 use the value in the default purpose.
8350 *) Fix the PKCS#8 DSA private key code so it decodes keys again
8351 and fix a memory leak.
8354 *) In util/mkerr.pl (which implements 'make errors'), preserve
8355 reason strings from the previous version of the .c file, as
8356 the default to have only downcase letters (and digits) in
8357 automatically generated reasons codes is not always appropriate.
8360 *) In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table
8361 using strerror. Previously, ERR_reason_error_string() returned
8362 library names as reason strings for SYSerr; but SYSerr is a special
8363 case where small numbers are errno values, not library numbers.
8366 *) Add '-dsaparam' option to 'openssl dhparam' application. This
8367 converts DSA parameters into DH parameters. (When creating parameters,
8368 DSA_generate_parameters is used.)
8371 *) Include 'length' (recommended exponent length) in C code generated
8372 by 'openssl dhparam -C'.
8375 *) The second argument to set_label in perlasm was already being used
8376 so couldn't be used as a "file scope" flag. Moved to third argument
8380 *) In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes
8381 instead of RAND_bytes for encryption IVs and salts.
8384 *) Include RAND_status() into RAND_METHOD instead of implementing
8385 it only for md_rand.c Otherwise replacing the PRNG by calling
8386 RAND_set_rand_method would be impossible.
8389 *) Don't let DSA_generate_key() enter an infinite loop if the random
8390 number generation fails.
8393 *) New 'rand' application for creating pseudo-random output.
8396 *) Added configuration support for Linux/IA64
8397 [Rolf Haberrecker <rolf@suse.de>]
8399 *) Assembler module support for Mingw32.
8402 *) Shared library support for HPUX (in shlib/).
8403 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous]
8405 *) Shared library support for Solaris gcc.
8406 [Lutz Behnke <behnke@trustcenter.de>]
8408 Changes between 0.9.4 and 0.9.5 [28 Feb 2000]
8410 *) PKCS7_encrypt() was adding text MIME headers twice because they
8411 were added manually and by SMIME_crlf_copy().
8414 *) In bntest.c don't call BN_rand with zero bits argument.
8415 [Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>]
8417 *) BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n]
8418 case was implemented. This caused BN_div_recp() to fail occasionally.
8421 *) Add an optional second argument to the set_label() in the perl
8422 assembly language builder. If this argument exists and is set
8423 to 1 it signals that the assembler should use a symbol whose
8424 scope is the entire file, not just the current function. This
8425 is needed with MASM which uses the format label:: for this scope.
8426 [Steve Henson, pointed out by Peter Runestig <peter@runestig.com>]
8428 *) Change the ASN1 types so they are typedefs by default. Before
8429 almost all types were #define'd to ASN1_STRING which was causing
8430 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING)
8434 *) Change names of new functions to the new get1/get0 naming
8435 convention: After 'get1', the caller owns a reference count
8436 and has to call ..._free; 'get0' returns a pointer to some
8437 data structure without incrementing reference counters.
8438 (Some of the existing 'get' functions increment a reference
8439 counter, some don't.)
8440 Similarly, 'set1' and 'add1' functions increase reference
8441 counters or duplicate objects.
8444 *) Allow for the possibility of temp RSA key generation failure:
8445 the code used to assume it always worked and crashed on failure.
8448 *) Fix potential buffer overrun problem in BIO_printf().
8449 [Ulf Möller, using public domain code by Patrick Powell; problem
8450 pointed out by David Sacerdote <das33@cornell.edu>]
8452 *) Support EGD <http://www.lothar.com/tech/crypto/>. New functions
8453 RAND_egd() and RAND_status(). In the command line application,
8454 the EGD socket can be specified like a seed file using RANDFILE
8458 *) Allow the string CERTIFICATE to be tolerated in PKCS#7 structures.
8459 Some CAs (e.g. Verisign) distribute certificates in this form.
8462 *) Remove the SSL_ALLOW_ADH compile option and set the default cipher
8463 list to exclude them. This means that no special compilation option
8464 is needed to use anonymous DH: it just needs to be included in the
8468 *) Change the EVP_MD_CTX_type macro so its meaning consistent with
8469 EVP_MD_type. The old functionality is available in a new macro called
8470 EVP_MD_md(). Change code that uses it and update docs.
8473 *) ..._ctrl functions now have corresponding ..._callback_ctrl functions
8474 where the 'void *' argument is replaced by a function pointer argument.
8475 Previously 'void *' was abused to point to functions, which works on
8476 many platforms, but is not correct. As these functions are usually
8477 called by macros defined in OpenSSL header files, most source code
8478 should work without changes.
8481 *) <openssl/opensslconf.h> (which is created by Configure) now contains
8482 sections with information on -D... compiler switches used for
8483 compiling the library so that applications can see them. To enable
8484 one of these sections, a pre-processor symbol OPENSSL_..._DEFINES
8485 must be defined. E.g.,
8486 #define OPENSSL_ALGORITHM_DEFINES
8487 #include <openssl/opensslconf.h>
8488 defines all pertinent NO_<algo> symbols, such as NO_IDEA, NO_RSA, etc.
8489 [Richard Levitte, Ulf and Bodo Möller]
8491 *) Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS
8495 *) Change the 'other' type in certificate aux info to a STACK_OF
8496 X509_ALGOR. Although not an AlgorithmIdentifier as such it has
8497 the required ASN1 format: arbitrary types determined by an OID.
8500 *) Add some PEM_write_X509_REQ_NEW() functions and a command line
8501 argument to 'req'. This is not because the function is newer or
8502 better than others it just uses the work 'NEW' in the certificate
8503 request header lines. Some software needs this.
8506 *) Reorganise password command line arguments: now passwords can be
8507 obtained from various sources. Delete the PEM_cb function and make
8508 it the default behaviour: i.e. if the callback is NULL and the
8509 usrdata argument is not NULL interpret it as a null terminated pass
8510 phrase. If usrdata and the callback are NULL then the pass phrase
8511 is prompted for as usual.
8514 *) Add support for the Compaq Atalla crypto accelerator. If it is installed,
8515 the support is automatically enabled. The resulting binaries will
8516 autodetect the card and use it if present.
8517 [Ben Laurie and Compaq Inc.]
8519 *) Work around for Netscape hang bug. This sends certificate request
8520 and server done in one record. Since this is perfectly legal in the
8521 SSL/TLS protocol it isn't a "bug" option and is on by default. See
8522 the bugs/SSLv3 entry for more info.
8525 *) HP-UX tune-up: new unified configs, HP C compiler bug workaround.
8528 *) Add -rand argument to smime and pkcs12 applications and read/write
8532 *) New 'passwd' tool for crypt(3) and apr1 password hashes.
8535 *) Add command line password options to the remaining applications.
8538 *) Bug fix for BN_div_recp() for numerators with an even number of
8542 *) More tests in bntest.c, and changed test_bn output.
8545 *) ./config recognizes MacOS X now.
8548 *) Bug fix for BN_div() when the first words of num and divsor are
8549 equal (it gave wrong results if (rem=(n1-q*d0)&BN_MASK2) < d0).
8552 *) Add support for various broken PKCS#8 formats, and command line
8553 options to produce them.
8556 *) New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to
8557 get temporary BIGNUMs from a BN_CTX.
8560 *) Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont()
8564 *) Change the SSLeay_add_all_*() functions to OpenSSL_add_all_*() and
8565 include a #define from the old name to the new. The original intent
8566 was that statically linked binaries could for example just call
8567 SSLeay_add_all_ciphers() to just add ciphers to the table and not
8568 link with digests. This never worked because SSLeay_add_all_digests()
8569 and SSLeay_add_all_ciphers() were in the same source file so calling
8570 one would link with the other. They are now in separate source files.
8573 *) Add a new -notext option to 'ca' and a -pubkey option to 'spkac'.
8576 *) Use a less unusual form of the Miller-Rabin primality test (it used
8577 a binary algorithm for exponentiation integrated into the Miller-Rabin
8578 loop, our standard modexp algorithms are faster).
8581 *) Support for the EBCDIC character set completed.
8582 [Martin Kraemer <Martin.Kraemer@Mch.SNI.De>]
8584 *) Source code cleanups: use const where appropriate, eliminate casts,
8585 use void * instead of char * in lhash.
8588 *) Bugfix: ssl3_send_server_key_exchange was not restartable
8589 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of
8590 this the server could overwrite ephemeral keys that the client
8594 *) Turn DSA_is_prime into a macro that calls BN_is_prime,
8595 using 50 iterations of the Rabin-Miller test.
8597 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50
8598 iterations of the Rabin-Miller test as required by the appendix
8599 to FIPS PUB 186[-1]) instead of DSA_is_prime.
8600 As BN_is_prime_fasttest includes trial division, DSA parameter
8601 generation becomes much faster.
8603 This implies a change for the callback functions in DSA_is_prime
8604 and DSA_generate_parameters: The callback function is called once
8605 for each positive witness in the Rabin-Miller test, not just
8606 occasionally in the inner loop; and the parameters to the
8607 callback function now provide an iteration count for the outer
8608 loop rather than for the current invocation of the inner loop.
8609 DSA_generate_parameters additionally can call the callback
8610 function with an 'iteration count' of -1, meaning that a
8611 candidate has passed the trial division test (when q is generated
8612 from an application-provided seed, trial division is skipped).
8615 *) New function BN_is_prime_fasttest that optionally does trial
8616 division before starting the Rabin-Miller test and has
8617 an additional BN_CTX * argument (whereas BN_is_prime always
8618 has to allocate at least one BN_CTX).
8619 'callback(1, -1, cb_arg)' is called when a number has passed the
8620 trial division stage.
8623 *) Fix for bug in CRL encoding. The validity dates weren't being handled
8627 *) New -pkcs12 option to CA.pl script to write out a PKCS#12 file.
8630 *) New function BN_pseudo_rand().
8633 *) Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable)
8634 bignum version of BN_from_montgomery() with the working code from
8635 SSLeay 0.9.0 (the word based version is faster anyway), and clean up
8639 *) Avoid a race condition in s2_clnt.c (function get_server_hello) that
8640 made it impossible to use the same SSL_SESSION data structure in
8641 SSL2 clients in multiple threads.
8644 *) The return value of RAND_load_file() no longer counts bytes obtained
8645 by stat(). RAND_load_file(..., -1) is new and uses the complete file
8646 to seed the PRNG (previously an explicit byte count was required).
8647 [Ulf Möller, Bodo Möller]
8649 *) Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
8650 used (char *) instead of (void *) and had casts all over the place.
8653 *) Make BN_generate_prime() return NULL on error if ret!=NULL.
8656 *) Retain source code compatibility for BN_prime_checks macro:
8657 BN_is_prime(..., BN_prime_checks, ...) now uses
8658 BN_prime_checks_for_size to determine the appropriate number of
8659 Rabin-Miller iterations.
8662 *) Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
8663 DH_CHECK_P_NOT_SAFE_PRIME.
8664 (Check if this is true? OpenPGP calls them "strong".)
8667 *) Merge the functionality of "dh" and "gendh" programs into a new program
8668 "dhparam". The old programs are retained for now but will handle DH keys
8669 (instead of parameters) in future.
8672 *) Make the ciphers, s_server and s_client programs check the return values
8673 when a new cipher list is set.
8676 *) Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
8677 ciphers. Before when the 56bit ciphers were enabled the sorting was
8680 The syntax for the cipher sorting has been extended to support sorting by
8681 cipher-strength (using the strength_bits hard coded in the tables).
8682 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
8684 Fix a bug in the cipher-command parser: when supplying a cipher command
8685 string with an "undefined" symbol (neither command nor alphanumeric
8686 [A-Za-z0-9], ssl_set_cipher_list used to hang in an endless loop. Now
8687 an error is flagged.
8689 Due to the strength-sorting extension, the code of the
8690 ssl_create_cipher_list() function was completely rearranged. I hope that
8691 the readability was also increased :-)
8692 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
8694 *) Minor change to 'x509' utility. The -CAcreateserial option now uses 1
8695 for the first serial number and places 2 in the serial number file. This
8696 avoids problems when the root CA is created with serial number zero and
8697 the first user certificate has the same issuer name and serial number
8701 *) Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
8702 the new code. Add documentation for this stuff.
8705 *) Changes to X509_ATTRIBUTE utilities. These have been renamed from
8706 X509_*() to X509at_*() on the grounds that they don't handle X509
8707 structures and behave in an analogous way to the X509v3 functions:
8708 they shouldn't be called directly but wrapper functions should be used
8711 So we also now have some wrapper functions that call the X509at functions
8712 when passed certificate requests. (TO DO: similar things can be done with
8713 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
8714 things. Some of these need some d2i or i2d and print functionality
8715 because they handle more complex structures.)
8718 *) Add missing #ifndefs that caused missing symbols when building libssl
8719 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
8720 NO_RSA in ssl/s2*.c.
8721 [Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller]
8723 *) Precautions against using the PRNG uninitialized: RAND_bytes() now
8724 has a return value which indicates the quality of the random data
8725 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
8726 error queue. New function RAND_pseudo_bytes() generates output that is
8727 guaranteed to be unique but not unpredictable. RAND_add is like
8728 RAND_seed, but takes an extra argument for an entropy estimate
8729 (RAND_seed always assumes full entropy).
8732 *) Do more iterations of Rabin-Miller probable prime test (specifically,
8733 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
8734 instead of only 2 for all lengths; see BN_prime_checks_for_size definition
8735 in crypto/bn/bn_prime.c for the complete table). This guarantees a
8736 false-positive rate of at most 2^-80 for random input.
8739 *) Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
8742 *) New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain
8743 in the 0.9.5 release), this returns the chain
8744 from an X509_CTX structure with a dup of the stack and all
8745 the X509 reference counts upped: so the stack will exist
8746 after X509_CTX_cleanup() has been called. Modify pkcs12.c
8749 Also make SSL_SESSION_print() print out the verify return
8753 *) Add manpage for the pkcs12 command. Also change the default
8754 behaviour so MAC iteration counts are used unless the new
8755 -nomaciter option is used. This improves file security and
8756 only older versions of MSIE (4.0 for example) need it.
8759 *) Honor the no-xxx Configure options when creating .DEF files.
8762 *) Add PKCS#10 attributes to field table: challengePassword,
8763 unstructuredName and unstructuredAddress. These are taken from
8764 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
8765 international characters are used.
8767 More changes to X509_ATTRIBUTE code: allow the setting of types
8768 based on strings. Remove the 'loc' parameter when adding
8769 attributes because these will be a SET OF encoding which is sorted
8773 *) Initial changes to the 'req' utility to allow request generation
8774 automation. This will allow an application to just generate a template
8775 file containing all the field values and have req construct the
8778 Initial support for X509_ATTRIBUTE handling. Stacks of these are
8779 used all over the place including certificate requests and PKCS#7
8780 structures. They are currently handled manually where necessary with
8781 some primitive wrappers for PKCS#7. The new functions behave in a
8782 manner analogous to the X509 extension functions: they allow
8783 attributes to be looked up by NID and added.
8785 Later something similar to the X509V3 code would be desirable to
8786 automatically handle the encoding, decoding and printing of the
8787 more complex types. The string types like challengePassword can
8788 be handled by the string table functions.
8790 Also modified the multi byte string table handling. Now there is
8791 a 'global mask' which masks out certain types. The table itself
8792 can use the flag STABLE_NO_MASK to ignore the mask setting: this
8793 is useful when for example there is only one permissible type
8794 (as in countryName) and using the mask might result in no valid
8798 *) Clean up 'Finished' handling, and add functions SSL_get_finished and
8799 SSL_get_peer_finished to allow applications to obtain the latest
8800 Finished messages sent to the peer or expected from the peer,
8801 respectively. (SSL_get_peer_finished is usually the Finished message
8802 actually received from the peer, otherwise the protocol will be aborted.)
8804 As the Finished message are message digests of the complete handshake
8805 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
8806 be used for external authentication procedures when the authentication
8807 provided by SSL/TLS is not desired or is not enough.
8810 *) Enhanced support for Alpha Linux is added. Now ./config checks if
8811 the host supports BWX extension and if Compaq C is present on the
8812 $PATH. Just exploiting of the BWX extension results in 20-30%
8813 performance kick for some algorithms, e.g. DES and RC4 to mention
8814 a couple. Compaq C in turn generates ~20% faster code for MD5 and
8818 *) Add support for MS "fast SGC". This is arguably a violation of the
8819 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
8820 weak crypto and after checking the certificate is SGC a second one
8821 with strong crypto. MS SGC stops the first handshake after receiving
8822 the server certificate message and sends a second client hello. Since
8823 a server will typically do all the time consuming operations before
8824 expecting any further messages from the client (server key exchange
8825 is the most expensive) there is little difference between the two.
8827 To get OpenSSL to support MS SGC we have to permit a second client
8828 hello message after we have sent server done. In addition we have to
8829 reset the MAC if we do get this second client hello.
8832 *) Add a function 'd2i_AutoPrivateKey()' this will automatically decide
8833 if a DER encoded private key is RSA or DSA traditional format. Changed
8834 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
8835 format DER encoded private key. Newer code should use PKCS#8 format which
8836 has the key type encoded in the ASN1 structure. Added DER private key
8837 support to pkcs8 application.
8840 *) SSL 3/TLS 1 servers now don't request certificates when an anonymous
8841 ciphersuites has been selected (as required by the SSL 3/TLS 1
8842 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
8843 is set, we interpret this as a request to violate the specification
8844 (the worst that can happen is a handshake failure, and 'correct'
8845 behaviour would result in a handshake failure anyway).
8848 *) In SSL_CTX_add_session, take into account that there might be multiple
8849 SSL_SESSION structures with the same session ID (e.g. when two threads
8850 concurrently obtain them from an external cache).
8851 The internal cache can handle only one SSL_SESSION with a given ID,
8852 so if there's a conflict, we now throw out the old one to achieve
8856 *) Add OIDs for idea and blowfish in CBC mode. This will allow both
8857 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
8858 some routines that use cipher OIDs: some ciphers do not have OIDs
8859 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
8863 *) Simplify the trust setting structure and code. Now we just have
8864 two sequences of OIDs for trusted and rejected settings. These will
8865 typically have values the same as the extended key usage extension
8866 and any application specific purposes.
8868 The trust checking code now has a default behaviour: it will just
8869 check for an object with the same NID as the passed id. Functions can
8870 be provided to override either the default behaviour or the behaviour
8871 for a given id. SSL client, server and email already have functions
8872 in place for compatibility: they check the NID and also return "trusted"
8873 if the certificate is self signed.
8876 *) Add d2i,i2d bio/fp functions for PrivateKey: these convert the
8877 traditional format into an EVP_PKEY structure.
8880 *) Add a password callback function PEM_cb() which either prompts for
8881 a password if usr_data is NULL or otherwise assumes it is a null
8882 terminated password. Allow passwords to be passed on command line
8883 environment or config files in a few more utilities.
8886 *) Add a bunch of DER and PEM functions to handle PKCS#8 format private
8887 keys. Add some short names for PKCS#8 PBE algorithms and allow them
8888 to be specified on the command line for the pkcs8 and pkcs12 utilities.
8889 Update documentation.
8892 *) Support for ASN1 "NULL" type. This could be handled before by using
8893 ASN1_TYPE but there wasn't any function that would try to read a NULL
8894 and produce an error if it couldn't. For compatibility we also have
8895 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
8896 don't allocate anything because they don't need to.
8899 *) Initial support for MacOS is now provided. Examine INSTALL.MacOS
8901 [Andy Polyakov, Roy Woods <roy@centicsystems.ca>]
8903 *) Rebuild of the memory allocation routines used by OpenSSL code and
8904 possibly others as well. The purpose is to make an interface that
8905 provide hooks so anyone can build a separate set of allocation and
8906 deallocation routines to be used by OpenSSL, for example memory
8907 pool implementations, or something else, which was previously hard
8908 since Malloc(), Realloc() and Free() were defined as macros having
8909 the values malloc, realloc and free, respectively (except for Win32
8910 compilations). The same is provided for memory debugging code.
8911 OpenSSL already comes with functionality to find memory leaks, but
8912 this gives people a chance to debug other memory problems.
8914 With these changes, a new set of functions and macros have appeared:
8916 CRYPTO_set_mem_debug_functions() [F]
8917 CRYPTO_get_mem_debug_functions() [F]
8918 CRYPTO_dbg_set_options() [F]
8919 CRYPTO_dbg_get_options() [F]
8920 CRYPTO_malloc_debug_init() [M]
8922 The memory debug functions are NULL by default, unless the library
8923 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
8924 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
8925 gives the standard debugging functions that come with OpenSSL) or
8926 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
8927 provided by the library user) must be used. When the standard
8928 debugging functions are used, CRYPTO_dbg_set_options can be used to
8929 request additional information:
8930 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
8931 the CRYPTO_MDEBUG_xxx macro when compiling the library.
8933 Also, things like CRYPTO_set_mem_functions will always give the
8934 expected result (the new set of functions is used for allocation
8935 and deallocation) at all times, regardless of platform and compiler
8938 To finish it up, some functions that were never use in any other
8939 way than through macros have a new API and new semantic:
8942 CRYPTO_dbg_realloc()
8945 All macros of value have retained their old syntax.
8946 [Richard Levitte and Bodo Moeller]
8948 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
8949 ordering of SMIMECapabilities wasn't in "strength order" and there
8950 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
8954 *) Some ASN1 types with illegal zero length encoding (INTEGER,
8955 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
8956 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson]
8958 *) Merge in my S/MIME library for OpenSSL. This provides a simple
8959 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
8960 functionality to handle multipart/signed properly) and a utility
8961 called 'smime' to call all this stuff. This is based on code I
8962 originally wrote for Celo who have kindly allowed it to be
8963 included in OpenSSL.
8966 *) Add variants des_set_key_checked and des_set_key_unchecked of
8967 des_set_key (aka des_key_sched). Global variable des_check_key
8968 decides which of these is called by des_set_key; this way
8969 des_check_key behaves as it always did, but applications and
8970 the library itself, which was buggy for des_check_key == 1,
8971 have a cleaner way to pick the version they need.
8974 *) New function PKCS12_newpass() which changes the password of a
8978 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and
8979 dynamic mix. In both cases the ids can be used as an index into the
8980 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
8981 functions so they accept a list of the field values and the
8982 application doesn't need to directly manipulate the X509_TRUST
8986 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
8990 *) Modify the way the V3 extension code looks up extensions. This now
8991 works in a similar way to the object code: we have some "standard"
8992 extensions in a static table which is searched with OBJ_bsearch()
8993 and the application can add dynamic ones if needed. The file
8994 crypto/x509v3/ext_dat.h now has the info: this file needs to be
8995 updated whenever a new extension is added to the core code and kept
8996 in ext_nid order. There is a simple program 'tabtest.c' which checks
8997 this. New extensions are not added too often so this file can readily
8998 be maintained manually.
9000 There are two big advantages in doing things this way. The extensions
9001 can be looked up immediately and no longer need to be "added" using
9002 X509V3_add_standard_extensions(): this function now does nothing.
9003 [Side note: I get *lots* of email saying the extension code doesn't
9004 work because people forget to call this function]
9005 Also no dynamic allocation is done unless new extensions are added:
9006 so if we don't add custom extensions there is no need to call
9007 X509V3_EXT_cleanup().
9010 *) Modify enc utility's salting as follows: make salting the default. Add a
9011 magic header, so unsalted files fail gracefully instead of just decrypting
9012 to garbage. This is because not salting is a big security hole, so people
9013 should be discouraged from doing it.
9016 *) Fixes and enhancements to the 'x509' utility. It allowed a message
9017 digest to be passed on the command line but it only used this
9018 parameter when signing a certificate. Modified so all relevant
9019 operations are affected by the digest parameter including the
9020 -fingerprint and -x509toreq options. Also -x509toreq choked if a
9021 DSA key was used because it didn't fix the digest.
9024 *) Initial certificate chain verify code. Currently tests the untrusted
9025 certificates for consistency with the verify purpose (which is set
9026 when the X509_STORE_CTX structure is set up) and checks the pathlength.
9028 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
9029 this is because it will reject chains with invalid extensions whereas
9030 every previous version of OpenSSL and SSLeay made no checks at all.
9032 Trust code: checks the root CA for the relevant trust settings. Trust
9033 settings have an initial value consistent with the verify purpose: e.g.
9034 if the verify purpose is for SSL client use it expects the CA to be
9035 trusted for SSL client use. However the default value can be changed to
9036 permit custom trust settings: one example of this would be to only trust
9037 certificates from a specific "secure" set of CAs.
9039 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
9040 which should be used for version portability: especially since the
9041 verify structure is likely to change more often now.
9043 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
9044 to set them. If not set then assume SSL clients will verify SSL servers
9047 Two new options to the verify program: -untrusted allows a set of
9048 untrusted certificates to be passed in and -purpose which sets the
9049 intended purpose of the certificate. If a purpose is set then the
9050 new chain verify code is used to check extension consistency.
9053 *) Support for the authority information access extension.
9056 *) Modify RSA and DSA PEM read routines to transparently handle
9057 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
9058 public keys in a format compatible with certificate
9059 SubjectPublicKeyInfo structures. Unfortunately there were already
9060 functions called *_PublicKey_* which used various odd formats so
9061 these are retained for compatibility: however the DSA variants were
9062 never in a public release so they have been deleted. Changed dsa/rsa
9063 utilities to handle the new format: note no releases ever handled public
9064 keys so we should be OK.
9066 The primary motivation for this change is to avoid the same fiasco
9067 that dogs private keys: there are several incompatible private key
9068 formats some of which are standard and some OpenSSL specific and
9069 require various evil hacks to allow partial transparent handling and
9070 even then it doesn't work with DER formats. Given the option anything
9071 other than PKCS#8 should be dumped: but the other formats have to
9072 stay in the name of compatibility.
9074 With public keys and the benefit of hindsight one standard format
9075 is used which works with EVP_PKEY, RSA or DSA structures: though
9076 it clearly returns an error if you try to read the wrong kind of key.
9078 Added a -pubkey option to the 'x509' utility to output the public key.
9079 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*()
9080 (renamed to EVP_PKEY_get1_*() in the OpenSSL 0.9.5 release) and add
9081 EVP_PKEY_rset_*() functions (renamed to EVP_PKEY_set1_*())
9082 that do the same as the EVP_PKEY_assign_*() except they up the
9083 reference count of the added key (they don't "swallow" the
9087 *) Fixes to crypto/x509/by_file.c the code to read in certificates and
9088 CRLs would fail if the file contained no certificates or no CRLs:
9089 added a new function to read in both types and return the number
9090 read: this means that if none are read it will be an error. The
9091 DER versions of the certificate and CRL reader would always fail
9092 because it isn't possible to mix certificates and CRLs in DER format
9093 without choking one or the other routine. Changed this to just read
9094 a certificate: this is the best we can do. Also modified the code
9095 in apps/verify.c to take notice of return codes: it was previously
9096 attempting to read in certificates from NULL pointers and ignoring
9097 any errors: this is one reason why the cert and CRL reader seemed
9098 to work. It doesn't check return codes from the default certificate
9099 routines: these may well fail if the certificates aren't installed.
9102 *) Code to support otherName option in GeneralName.
9105 *) First update to verify code. Change the verify utility
9106 so it warns if it is passed a self signed certificate:
9107 for consistency with the normal behaviour. X509_verify
9108 has been modified to it will now verify a self signed
9109 certificate if *exactly* the same certificate appears
9110 in the store: it was previously impossible to trust a
9111 single self signed certificate. This means that:
9112 openssl verify ss.pem
9113 now gives a warning about a self signed certificate but
9114 openssl verify -CAfile ss.pem ss.pem
9118 *) For servers, store verify_result in SSL_SESSION data structure
9119 (and add it to external session representation).
9120 This is needed when client certificate verifications fails,
9121 but an application-provided verification callback (set by
9122 SSL_CTX_set_cert_verify_callback) allows accepting the session
9123 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
9124 but returns 1): When the session is reused, we have to set
9125 ssl->verify_result to the appropriate error code to avoid
9127 [Bodo Moeller, problem pointed out by Lutz Jaenicke]
9129 *) Fix a bug in the new PKCS#7 code: it didn't consider the
9130 case in PKCS7_dataInit() where the signed PKCS7 structure
9131 didn't contain any existing data because it was being created.
9132 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson]
9134 *) Add a salt to the key derivation routines in enc.c. This
9135 forms the first 8 bytes of the encrypted file. Also add a
9136 -S option to allow a salt to be input on the command line.
9139 *) New function X509_cmp(). Oddly enough there wasn't a function
9140 to compare two certificates. We do this by working out the SHA1
9141 hash and comparing that. X509_cmp() will be needed by the trust
9145 *) SSL_get1_session() is like SSL_get_session(), but increments
9146 the reference count in the SSL_SESSION returned.
9147 [Geoff Thorpe <geoff@eu.c2.net>]
9149 *) Fix for 'req': it was adding a null to request attributes.
9150 Also change the X509_LOOKUP and X509_INFO code to handle
9151 certificate auxiliary information.
9154 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document
9158 *) Add the possibility to add extra information to the memory leak
9159 detecting output, to form tracebacks, showing from where each
9160 allocation was originated: CRYPTO_push_info("constant string") adds
9161 the string plus current file name and line number to a per-thread
9162 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
9163 is like calling CYRPTO_pop_info() until the stack is empty.
9164 Also updated memory leak detection code to be multi-thread-safe.
9167 *) Add options -text and -noout to pkcs7 utility and delete the
9168 encryption options which never did anything. Update docs.
9171 *) Add options to some of the utilities to allow the pass phrase
9172 to be included on either the command line (not recommended on
9173 OSes like Unix) or read from the environment. Update the
9174 manpages and fix a few bugs.
9177 *) Add a few manpages for some of the openssl commands.
9180 *) Fix the -revoke option in ca. It was freeing up memory twice,
9181 leaking and not finding already revoked certificates.
9184 *) Extensive changes to support certificate auxiliary information.
9185 This involves the use of X509_CERT_AUX structure and X509_AUX
9186 functions. An X509_AUX function such as PEM_read_X509_AUX()
9187 can still read in a certificate file in the usual way but it
9188 will also read in any additional "auxiliary information". By
9189 doing things this way a fair degree of compatibility can be
9190 retained: existing certificates can have this information added
9191 using the new 'x509' options.
9193 Current auxiliary information includes an "alias" and some trust
9194 settings. The trust settings will ultimately be used in enhanced
9195 certificate chain verification routines: currently a certificate
9196 can only be trusted if it is self signed and then it is trusted
9200 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD).
9201 The problem was that one of the replacement routines had not been working
9202 since SSLeay releases. For now the offending routine has been replaced
9203 with non-optimised assembler. Even so, this now gives around 95%
9204 performance improvement for 1024 bit RSA signs.
9207 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2
9208 handling. Most clients have the effective key size in bits equal to
9209 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
9210 A few however don't do this and instead use the size of the decrypted key
9211 to determine the RC2 key length and the AlgorithmIdentifier to determine
9212 the effective key length. In this case the effective key length can still
9213 be 40 bits but the key length can be 168 bits for example. This is fixed
9214 by manually forcing an RC2 key into the EVP_PKEY structure because the
9215 EVP code can't currently handle unusual RC2 key sizes: it always assumes
9216 the key length and effective key length are equal.
9219 *) Add a bunch of functions that should simplify the creation of
9220 X509_NAME structures. Now you should be able to do:
9221 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
9222 and have it automatically work out the correct field type and fill in
9223 the structures. The more adventurous can try:
9224 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
9225 and it will (hopefully) work out the correct multibyte encoding.
9228 *) Change the 'req' utility to use the new field handling and multibyte
9229 copy routines. Before the DN field creation was handled in an ad hoc
9230 way in req, ca, and x509 which was rather broken and didn't support
9231 BMPStrings or UTF8Strings. Since some software doesn't implement
9232 BMPStrings or UTF8Strings yet, they can be enabled using the config file
9233 using the dirstring_type option. See the new comment in the default
9234 openssl.cnf for more info.
9237 *) Make crypto/rand/md_rand.c more robust:
9238 - Assure unique random numbers after fork().
9239 - Make sure that concurrent threads access the global counter and
9240 md serializably so that we never lose entropy in them
9241 or use exactly the same state in multiple threads.
9242 Access to the large state is not always serializable because
9243 the additional locking could be a performance killer, and
9244 md should be large enough anyway.
9247 *) New file apps/app_rand.c with commonly needed functionality
9248 for handling the random seed file.
9250 Use the random seed file in some applications that previously did not:
9252 dsaparam -genkey (which also ignored its '-rand' option),
9255 x509 (when signing).
9256 Except on systems with /dev/urandom, it is crucial to have a random
9257 seed file at least for key creation, DSA signing, and for DH exchanges;
9258 for RSA signatures we could do without one.
9260 gendh and gendsa (unlike genrsa) used to read only the first byte
9261 of each file listed in the '-rand' option. The function as previously
9262 found in genrsa is now in app_rand.c and is used by all programs
9263 that support '-rand'.
9266 *) In RAND_write_file, use mode 0600 for creating files;
9267 don't just chmod when it may be too late.
9270 *) Report an error from X509_STORE_load_locations
9271 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
9274 *) New function ASN1_mbstring_copy() this copies a string in either
9275 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
9276 into an ASN1_STRING type. A mask of permissible types is passed
9277 and it chooses the "minimal" type to use or an error if not type
9281 *) Add function equivalents to the various macros in asn1.h. The old
9282 macros are retained with an M_ prefix. Code inside the library can
9283 use the M_ macros. External code (including the openssl utility)
9284 should *NOT* in order to be "shared library friendly".
9287 *) Add various functions that can check a certificate's extensions
9288 to see if it usable for various purposes such as SSL client,
9289 server or S/MIME and CAs of these types. This is currently
9290 VERY EXPERIMENTAL but will ultimately be used for certificate chain
9291 verification. Also added a -purpose flag to x509 utility to
9292 print out all the purposes.
9295 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated
9299 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search
9300 for, obtain and decode and extension and obtain its critical flag.
9301 This allows all the necessary extension code to be handled in a
9302 single function call.
9305 *) RC4 tune-up featuring 30-40% performance improvement on most RISC
9306 platforms. See crypto/rc4/rc4_enc.c for further details.
9309 *) New -noout option to asn1parse. This causes no output to be produced
9310 its main use is when combined with -strparse and -out to extract data
9311 from a file (which may not be in ASN.1 format).
9314 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer
9315 when producing the local key id.
9316 [Richard Levitte <levitte@stacken.kth.se>]
9318 *) New option -dhparam in s_server. This allows a DH parameter file to be
9319 stated explicitly. If it is not stated then it tries the first server
9320 certificate file. The previous behaviour hard coded the filename
9324 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow
9325 a public key to be input or output. For example:
9326 openssl rsa -in key.pem -pubout -out pubkey.pem
9327 Also added necessary DSA public key functions to handle this.
9330 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
9331 in the message. This was handled by allowing
9332 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
9333 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>]
9335 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
9336 to the end of the strings whereas this didn't. This would cause problems
9337 if strings read with d2i_ASN1_bytes() were later modified.
9338 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>]
9340 *) Fix for base64 decode bug. When a base64 bio reads only one line of
9341 data and it contains EOF it will end up returning an error. This is
9342 caused by input 46 bytes long. The cause is due to the way base64
9343 BIOs find the start of base64 encoded data. They do this by trying a
9344 trial decode on each line until they find one that works. When they
9345 do a flag is set and it starts again knowing it can pass all the
9346 data directly through the decoder. Unfortunately it doesn't reset
9347 the context it uses. This means that if EOF is reached an attempt
9348 is made to pass two EOFs through the context and this causes the
9349 resulting error. This can also cause other problems as well. As is
9350 usual with these problems it takes *ages* to find and the fix is
9351 trivial: move one line.
9352 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ]
9354 *) Ugly workaround to get s_client and s_server working under Windows. The
9355 old code wouldn't work because it needed to select() on sockets and the
9356 tty (for keypresses and to see if data could be written). Win32 only
9357 supports select() on sockets so we select() with a 1s timeout on the
9358 sockets and then see if any characters are waiting to be read, if none
9359 are present then we retry, we also assume we can always write data to
9360 the tty. This isn't nice because the code then blocks until we've
9361 received a complete line of data and it is effectively polling the
9362 keyboard at 1s intervals: however it's quite a bit better than not
9363 working at all :-) A dedicated Windows application might handle this
9364 with an event loop for example.
9367 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
9368 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
9369 will be called when RSA_sign() and RSA_verify() are used. This is useful
9370 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
9371 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
9372 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
9373 This necessitated the support of an extra signature type NID_md5_sha1
9374 for SSL signatures and modifications to the SSL library to use it instead
9375 of calling RSA_public_decrypt() and RSA_private_encrypt().
9378 *) Add new -verify -CAfile and -CApath options to the crl program, these
9379 will lookup a CRL issuers certificate and verify the signature in a
9380 similar way to the verify program. Tidy up the crl program so it
9381 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
9382 less strict. It will now permit CRL extensions even if it is not
9383 a V2 CRL: this will allow it to tolerate some broken CRLs.
9386 *) Initialize all non-automatic variables each time one of the openssl
9387 sub-programs is started (this is necessary as they may be started
9388 multiple times from the "OpenSSL>" prompt).
9389 [Lennart Bang, Bodo Moeller]
9391 *) Preliminary compilation option RSA_NULL which disables RSA crypto without
9392 removing all other RSA functionality (this is what NO_RSA does). This
9393 is so (for example) those in the US can disable those operations covered
9394 by the RSA patent while allowing storage and parsing of RSA keys and RSA
9398 *) Non-copying interface to BIO pairs.
9399 (still largely untested)
9402 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive
9403 ASCII string. This was handled independently in various places before.
9406 *) New functions UTF8_getc() and UTF8_putc() that parse and generate
9407 UTF8 strings a character at a time.
9410 *) Use client_version from client hello to select the protocol
9411 (s23_srvr.c) and for RSA client key exchange verification
9412 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
9415 *) Add various utility functions to handle SPKACs, these were previously
9416 handled by poking round in the structure internals. Added new function
9417 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
9418 print, verify and generate SPKACs. Based on an original idea from
9419 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
9422 *) RIPEMD160 is operational on all platforms and is back in 'make test'.
9425 *) Allow the config file extension section to be overwritten on the
9426 command line. Based on an original idea from Massimiliano Pala
9427 <madwolf@comune.modena.it>. The new option is called -extensions
9428 and can be applied to ca, req and x509. Also -reqexts to override
9429 the request extensions in req and -crlexts to override the crl extensions
9433 *) Add new feature to the SPKAC handling in ca. Now you can include
9434 the same field multiple times by preceding it by "XXXX." for example:
9437 this is the same syntax as used in the req config file.
9440 *) Allow certificate extensions to be added to certificate requests. These
9441 are specified in a 'req_extensions' option of the req section of the
9442 config file. They can be printed out with the -text option to req but
9443 are otherwise ignored at present.
9446 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
9447 data read consists of only the final block it would not decrypted because
9448 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
9449 A misplaced 'break' also meant the decrypted final block might not be
9450 copied until the next read.
9453 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
9454 a few extra parameters to the DH structure: these will be useful if
9455 for example we want the value of 'q' or implement X9.42 DH.
9458 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and
9459 provides hooks that allow the default DSA functions or functions on a
9460 "per key" basis to be replaced. This allows hardware acceleration and
9461 hardware key storage to be handled without major modification to the
9462 library. Also added low level modexp hooks and CRYPTO_EX structure and
9463 associated functions.
9466 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
9467 as "read only": it can't be written to and the buffer it points to will
9468 not be freed. Reading from a read only BIO is much more efficient than
9469 a normal memory BIO. This was added because there are several times when
9470 an area of memory needs to be read from a BIO. The previous method was
9471 to create a memory BIO and write the data to it, this results in two
9472 copies of the data and an O(n^2) reading algorithm. There is a new
9473 function BIO_new_mem_buf() which creates a read only memory BIO from
9474 an area of memory. Also modified the PKCS#7 routines to use read only
9478 *) Bugfix: ssl23_get_client_hello did not work properly when called in
9479 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
9480 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
9481 but a retry condition occured while trying to read the rest.
9484 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as
9485 NID_pkcs7_encrypted by default: this was wrong since this should almost
9486 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
9487 the encrypted data type: this is a more sensible place to put it and it
9488 allows the PKCS#12 code to be tidied up that duplicated this
9492 *) Changed obj_dat.pl script so it takes its input and output files on
9493 the command line. This should avoid shell escape redirection problems
9497 *) Initial support for certificate extension requests, these are included
9498 in things like Xenroll certificate requests. Included functions to allow
9499 extensions to be obtained and added.
9502 *) -crlf option to s_client and s_server for sending newlines as
9503 CRLF (as required by many protocols).
9506 Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
9508 *) Install libRSAglue.a when OpenSSL is built with RSAref.
9509 [Ralf S. Engelschall]
9511 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency.
9512 [Andrija Antonijevic <TheAntony2@bigfoot.com>]
9514 *) Fix -startdate and -enddate (which was missing) arguments to 'ca'
9518 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as
9519 DH parameters/keys (q is lost during that conversion, but the resulting
9520 DH parameters contain its length).
9522 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
9523 much faster than DH_generate_parameters (which creates parameters
9524 where p = 2*q + 1), and also the smaller q makes DH computations
9525 much more efficient (160-bit exponentiation instead of 1024-bit
9526 exponentiation); so this provides a convenient way to support DHE
9527 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
9528 utter importance to use
9529 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
9531 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
9532 when such DH parameters are used, because otherwise small subgroup
9533 attacks may become possible!
9536 *) Avoid memory leak in i2d_DHparams.
9539 *) Allow the -k option to be used more than once in the enc program:
9540 this allows the same encrypted message to be read by multiple recipients.
9543 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
9544 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
9545 it will always use the numerical form of the OID, even if it has a short
9549 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
9550 method only got called if p,q,dmp1,dmq1,iqmp components were present,
9551 otherwise bn_mod_exp was called. In the case of hardware keys for example
9552 no private key components need be present and it might store extra data
9553 in the RSA structure, which cannot be accessed from bn_mod_exp.
9554 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
9555 private key operations.
9558 *) Added support for SPARC Linux.
9561 *) pem_password_cb function type incompatibly changed from
9562 typedef int pem_password_cb(char *buf, int size, int rwflag);
9564 ....(char *buf, int size, int rwflag, void *userdata);
9565 so that applications can pass data to their callbacks:
9566 The PEM[_ASN1]_{read,write}... functions and macros now take an
9567 additional void * argument, which is just handed through whenever
9568 the password callback is called.
9569 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller]
9571 New function SSL_CTX_set_default_passwd_cb_userdata.
9573 Compatibility note: As many C implementations push function arguments
9574 onto the stack in reverse order, the new library version is likely to
9575 interoperate with programs that have been compiled with the old
9576 pem_password_cb definition (PEM_whatever takes some data that
9577 happens to be on the stack as its last argument, and the callback
9578 just ignores this garbage); but there is no guarantee whatsoever that
9581 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
9582 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
9583 problems not only on Windows, but also on some Unix platforms.
9584 To avoid problematic command lines, these definitions are now in an
9585 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
9586 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
9589 *) MIPS III/IV assembler module is reimplemented.
9592 *) More DES library cleanups: remove references to srand/rand and
9593 delete an unused file.
9596 *) Add support for the the free Netwide assembler (NASM) under Win32,
9597 since not many people have MASM (ml) and it can be hard to obtain.
9598 This is currently experimental but it seems to work OK and pass all
9599 the tests. Check out INSTALL.W32 for info.
9602 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
9603 without temporary keys kept an extra copy of the server key,
9604 and connections with temporary keys did not free everything in case
9608 *) New function RSA_check_key and new openssl rsa option -check
9609 for verifying the consistency of RSA keys.
9610 [Ulf Moeller, Bodo Moeller]
9612 *) Various changes to make Win32 compile work:
9613 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
9614 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
9615 comparison" warnings.
9616 3. Add sk_<TYPE>_sort to DEF file generator and do make update.
9619 *) Add a debugging option to PKCS#5 v2 key generation function: when
9620 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
9621 derived keys are printed to stderr.
9624 *) Copy the flags in ASN1_STRING_dup().
9625 [Roman E. Pavlov <pre@mo.msk.ru>]
9627 *) The x509 application mishandled signing requests containing DSA
9628 keys when the signing key was also DSA and the parameters didn't match.
9630 It was supposed to omit the parameters when they matched the signing key:
9631 the verifying software was then supposed to automatically use the CA's
9632 parameters if they were absent from the end user certificate.
9634 Omitting parameters is no longer recommended. The test was also
9635 the wrong way round! This was probably due to unusual behaviour in
9636 EVP_cmp_parameters() which returns 1 if the parameters match.
9637 This meant that parameters were omitted when they *didn't* match and
9638 the certificate was useless. Certificates signed with 'ca' didn't have
9640 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>]
9642 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
9643 The interface is as follows:
9644 Applications can use
9645 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
9646 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
9647 "off" is now the default.
9648 The library internally uses
9649 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
9650 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
9651 to disable memory-checking temporarily.
9653 Some inconsistent states that previously were possible (and were
9654 even the default) are now avoided.
9656 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
9657 with each memory chunk allocated; this is occasionally more helpful
9658 than just having a counter.
9660 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
9662 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
9666 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX),
9667 which largely parallels "options", but is for changing API behaviour,
9668 whereas "options" are about protocol behaviour.
9669 Initial "mode" flags are:
9671 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
9672 a single record has been written.
9673 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
9674 retries use the same buffer location.
9675 (But all of the contents must be
9679 *) Bugfix: SSL_set_options ignored its parameter, only SSL_CTX_set_options
9682 *) Fix problems with no-hmac etc.
9683 [Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>]
9685 *) New functions RSA_get_default_method(), RSA_set_method() and
9686 RSA_get_method(). These allows replacement of RSA_METHODs without having
9687 to mess around with the internals of an RSA structure.
9690 *) Fix memory leaks in DSA_do_sign and DSA_is_prime.
9691 Also really enable memory leak checks in openssl.c and in some
9693 [Chad C. Mulligan, Bodo Moeller]
9695 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
9696 up the length of negative integers. This has now been simplified to just
9697 store the length when it is first determined and use it later, rather
9698 than trying to keep track of where data is copied and updating it to
9700 [Steve Henson, reported by Brien Wheeler
9701 <bwheeler@authentica-security.com>]
9703 *) Add a new function PKCS7_signatureVerify. This allows the verification
9704 of a PKCS#7 signature but with the signing certificate passed to the
9705 function itself. This contrasts with PKCS7_dataVerify which assumes the
9706 certificate is present in the PKCS#7 structure. This isn't always the
9707 case: certificates can be omitted from a PKCS#7 structure and be
9708 distributed by "out of band" means (such as a certificate database).
9711 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the
9712 function prototypes in pem.h, also change util/mkdef.pl to add the
9713 necessary function names.
9716 *) mk1mf.pl (used by Windows builds) did not properly read the
9717 options set by Configure in the top level Makefile, and Configure
9718 was not even able to write more than one option correctly.
9719 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
9722 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config
9723 file to be loaded from a BIO or FILE pointer. The BIO version will
9724 for example allow memory BIOs to contain config info.
9727 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
9728 Whoever hopes to achieve shared-library compatibility across versions
9729 must use this, not the compile-time macro.
9730 (Exercise 0.9.4: Which is the minimum library version required by
9732 Note: All this applies only to multi-threaded programs, others don't
9736 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests
9737 through a BIO pair triggered the default case, i.e.
9738 SSLerr(...,SSL_R_UNKNOWN_STATE).
9741 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
9742 can use the SSL library even if none of the specific BIOs is
9746 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
9747 for the encoded length.
9748 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>]
9750 *) Add initial documentation of the X509V3 functions.
9753 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and
9754 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
9755 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
9756 secure PKCS#8 private key format with a high iteration count.
9759 *) Fix determination of Perl interpreter: A perl or perl5
9760 _directory_ in $PATH was also accepted as the interpreter.
9761 [Ralf S. Engelschall]
9763 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking
9764 wrong with it but it was very old and did things like calling
9765 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
9769 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed
9770 to use the new extension code.
9773 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
9774 with macros. This should make it easier to change their form, add extra
9775 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
9779 *) Add to configuration table a new entry that can specify an alternative
9780 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
9781 according to Mark Crispin <MRC@Panda.COM>.
9785 *) DES CBC did not update the IV. Weird.
9788 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
9789 Changing the behaviour of the former might break existing programs --
9790 where IV updating is needed, des_ncbc_encrypt can be used.
9793 *) When bntest is run from "make test" it drives bc to check its
9794 calculations, as well as internally checking them. If an internal check
9795 fails, it needs to cause bc to give a non-zero result or make test carries
9796 on without noticing the failure. Fixed.
9799 *) DES library cleanups.
9802 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
9803 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
9804 ciphers. NOTE: although the key derivation function has been verified
9805 against some published test vectors it has not been extensively tested
9806 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
9810 *) Instead of "mkdir -p", which is not fully portable, use new
9811 Perl script "util/mkdir-p.pl".
9814 *) Rewrite the way password based encryption (PBE) is handled. It used to
9815 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
9816 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
9817 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
9818 the 'parameter' field of the AlgorithmIdentifier is passed to the
9819 underlying key generation function so it must do its own ASN1 parsing.
9820 This has also changed the EVP_PBE_CipherInit() function which now has a
9821 'parameter' argument instead of literal salt and iteration count values
9822 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
9825 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms
9826 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
9827 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
9828 KEY" because this clashed with PKCS#8 unencrypted string. Since this
9829 value was just used as a "magic string" and not used directly its
9830 value doesn't matter.
9833 *) Introduce some semblance of const correctness to BN. Shame C doesn't
9837 *) "linux-sparc64" configuration (ultrapenguin).
9838 [Ray Miller <ray.miller@oucs.ox.ac.uk>]
9839 "linux-sparc" configuration.
9840 [Christian Forster <fo@hawo.stw.uni-erlangen.de>]
9842 *) config now generates no-xxx options for missing ciphers.
9845 *) Support the EBCDIC character set (work in progress).
9846 File ebcdic.c not yet included because it has a different license.
9847 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
9849 *) Support BS2000/OSD-POSIX.
9850 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
9852 *) Make callbacks for key generation use void * instead of char *.
9855 *) Make S/MIME samples compile (not yet tested).
9858 *) Additional typesafe stacks.
9861 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
9865 Changes between 0.9.3 and 0.9.3a [29 May 1999]
9867 *) New configuration variant "sco5-gcc".
9869 *) Updated some demos.
9870 [Sean O Riordain, Wade Scholine]
9872 *) Add missing BIO_free at exit of pkcs12 application.
9875 *) Fix memory leak in conf.c.
9878 *) Updates for Win32 to assembler version of MD5.
9881 *) Set #! path to perl in apps/der_chop to where we found it
9882 instead of using a fixed path.
9885 *) SHA library changes for irix64-mips4-cc.
9888 *) Improvements for VMS support.
9892 Changes between 0.9.2b and 0.9.3 [24 May 1999]
9894 *) Bignum library bug fix. IRIX 6 passes "make test" now!
9895 This also avoids the problems with SC4.2 and unpatched SC5.
9896 [Andy Polyakov <appro@fy.chalmers.se>]
9898 *) New functions sk_num, sk_value and sk_set to replace the previous macros.
9899 These are required because of the typesafe stack would otherwise break
9900 existing code. If old code used a structure member which used to be STACK
9901 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
9902 sk_num or sk_value it would produce an error because the num, data members
9903 are not present in STACK_OF. Now it just produces a warning. sk_set
9904 replaces the old method of assigning a value to sk_value
9905 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
9906 that does this will no longer work (and should use sk_set instead) but
9907 this could be regarded as a "questionable" behaviour anyway.
9910 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now
9911 correctly handle encrypted S/MIME data.
9914 *) Change type of various DES function arguments from des_cblock
9915 (which means, in function argument declarations, pointer to char)
9916 to des_cblock * (meaning pointer to array with 8 char elements),
9917 which allows the compiler to do more typechecking; it was like
9918 that back in SSLeay, but with lots of ugly casts.
9920 Introduce new type const_des_cblock.
9923 *) Reorganise the PKCS#7 library and get rid of some of the more obvious
9924 problems: find RecipientInfo structure that matches recipient certificate
9925 and initialise the ASN1 structures properly based on passed cipher.
9928 *) Belatedly make the BN tests actually check the results.
9931 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
9932 to and from BNs: it was completely broken. New compilation option
9933 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
9934 key elements as negative integers.
9937 *) Reorganize and speed up MD5.
9938 [Andy Polyakov <appro@fy.chalmers.se>]
9941 [Richard Levitte <richard@levitte.org>]
9943 *) New option -out to asn1parse to allow the parsed structure to be
9944 output to a file. This is most useful when combined with the -strparse
9945 option to examine the output of things like OCTET STRINGS.
9948 *) Make SSL library a little more fool-proof by not requiring any longer
9949 that SSL_set_{accept,connect}_state be called before
9950 SSL_{accept,connect} may be used (SSL_set_..._state is omitted
9951 in many applications because usually everything *appeared* to work as
9952 intended anyway -- now it really works as intended).
9955 *) Move openssl.cnf out of lib/.
9958 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall
9959 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
9960 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+
9961 [Ralf S. Engelschall]
9963 *) Various fixes to the EVP and PKCS#7 code. It may now be able to
9964 handle PKCS#7 enveloped data properly.
9965 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve]
9967 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
9968 copying pointers. The cert_st handling is changed by this in
9969 various ways (and thus what used to be known as ctx->default_cert
9970 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert
9971 any longer when s->cert does not give us what we need).
9972 ssl_cert_instantiate becomes obsolete by this change.
9973 As soon as we've got the new code right (possibly it already is?),
9974 we have solved a couple of bugs of the earlier code where s->cert
9975 was used as if it could not have been shared with other SSL structures.
9977 Note that using the SSL API in certain dirty ways now will result
9978 in different behaviour than observed with earlier library versions:
9979 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx)
9980 does not influence s as it used to.
9982 In order to clean up things more thoroughly, inside SSL_SESSION
9983 we don't use CERT any longer, but a new structure SESS_CERT
9984 that holds per-session data (if available); currently, this is
9985 the peer's certificate chain and, for clients, the server's certificate
9986 and temporary key. CERT holds only those values that can have
9987 meaningful defaults in an SSL_CTX.
9990 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
9991 from the internal representation. Various PKCS#7 fixes: remove some
9992 evil casts and set the enc_dig_alg field properly based on the signing
9996 *) Allow PKCS#12 password to be set from the command line or the
9997 environment. Let 'ca' get its config file name from the environment
9998 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
10002 *) Allow certificate policies extension to use an IA5STRING for the
10003 organization field. This is contrary to the PKIX definition but
10004 VeriSign uses it and IE5 only recognises this form. Document 'x509'
10008 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
10009 without disallowing inline assembler and the like for non-pedantic builds.
10012 *) Support Borland C++ builder.
10013 [Janez Jere <jj@void.si>, modified by Ulf Möller]
10015 *) Support Mingw32.
10018 *) SHA-1 cleanups and performance enhancements.
10019 [Andy Polyakov <appro@fy.chalmers.se>]
10021 *) Sparc v8plus assembler for the bignum library.
10022 [Andy Polyakov <appro@fy.chalmers.se>]
10024 *) Accept any -xxx and +xxx compiler options in Configure.
10027 *) Update HPUX configuration.
10030 *) Add missing sk_<type>_unshift() function to safestack.h
10031 [Ralf S. Engelschall]
10033 *) New function SSL_CTX_use_certificate_chain_file that sets the
10034 "extra_cert"s in addition to the certificate. (This makes sense
10035 only for "PEM" format files, as chains as a whole are not
10039 *) Support verify_depth from the SSL API.
10040 x509_vfy.c had what can be considered an off-by-one-error:
10041 Its depth (which was not part of the external interface)
10042 was actually counting the number of certificates in a chain;
10043 now it really counts the depth.
10046 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
10047 instead of X509err, which often resulted in confusing error
10048 messages since the error codes are not globally unique
10049 (e.g. an alleged error in ssl3_accept when a certificate
10050 didn't match the private key).
10052 *) New function SSL_CTX_set_session_id_context that allows to set a default
10053 value (so that you don't need SSL_set_session_id_context for each
10054 connection using the SSL_CTX).
10057 *) OAEP decoding bug fix.
10060 *) Support INSTALL_PREFIX for package builders, as proposed by
10064 *) New Configure options "threads" and "no-threads". For systems
10065 where the proper compiler options are known (currently Solaris
10066 and Linux), "threads" is the default.
10069 *) New script util/mklink.pl as a faster substitute for util/mklink.sh.
10072 *) Install various scripts to $(OPENSSLDIR)/misc, not to
10073 $(INSTALLTOP)/bin -- they shouldn't clutter directories
10074 such as /usr/local/bin.
10077 *) "make linux-shared" to build shared libraries.
10078 [Niels Poppe <niels@netbox.org>]
10080 *) New Configure option no-<cipher> (rsa, idea, rc5, ...).
10083 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
10084 extension adding in x509 utility.
10087 *) Remove NOPROTO sections and error code comments.
10090 *) Partial rewrite of the DEF file generator to now parse the ANSI
10094 *) New Configure options --prefix=DIR and --openssldir=DIR.
10097 *) Complete rewrite of the error code script(s). It is all now handled
10098 by one script at the top level which handles error code gathering,
10099 header rewriting and C source file generation. It should be much better
10100 than the old method: it now uses a modified version of Ulf's parser to
10101 read the ANSI prototypes in all header files (thus the old K&R definitions
10102 aren't needed for error creation any more) and do a better job of
10103 translating function codes into names. The old 'ASN1 error code imbedded
10104 in a comment' is no longer necessary and it doesn't use .err files which
10105 have now been deleted. Also the error code call doesn't have to appear all
10106 on one line (which resulted in some large lines...).
10109 *) Change #include filenames from <foo.h> to <openssl/foo.h>.
10112 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return
10113 0 (which usually indicates a closed connection), but continue reading.
10116 *) Fix some race conditions.
10119 *) Add support for CRL distribution points extension. Add Certificate
10120 Policies and CRL distribution points documentation.
10123 *) Move the autogenerated header file parts to crypto/opensslconf.h.
10126 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
10127 8 of keying material. Merlin has also confirmed interop with this fix
10128 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
10129 [Merlin Hughes <merlin@baltimore.ie>]
10131 *) Fix lots of warnings.
10132 [Richard Levitte <levitte@stacken.kth.se>]
10134 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
10135 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
10136 [Richard Levitte <levitte@stacken.kth.se>]
10138 *) Fix problems with sizeof(long) == 8.
10139 [Andy Polyakov <appro@fy.chalmers.se>]
10141 *) Change functions to ANSI C.
10144 *) Fix typos in error codes.
10145 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller]
10147 *) Remove defunct assembler files from Configure.
10150 *) SPARC v8 assembler BIGNUM implementation.
10151 [Andy Polyakov <appro@fy.chalmers.se>]
10153 *) Support for Certificate Policies extension: both print and set.
10154 Various additions to support the r2i method this uses.
10157 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could
10158 return a const string when you are expecting an allocated buffer.
10161 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
10162 types DirectoryString and DisplayText.
10165 *) Add code to allow r2i extensions to access the configuration database,
10166 add an LHASH database driver and add several ctx helper functions.
10169 *) Fix an evil bug in bn_expand2() which caused various BN functions to
10170 fail when they extended the size of a BIGNUM.
10173 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to
10174 support typesafe stack.
10177 *) Fix typo in SSL_[gs]et_options().
10178 [Nils Frostberg <nils@medcom.se>]
10180 *) Delete various functions and files that belonged to the (now obsolete)
10181 old X509V3 handling code.
10184 *) New Configure option "rsaref".
10187 *) Don't auto-generate pem.h.
10190 *) Introduce type-safe ASN.1 SETs.
10193 *) Convert various additional casted stacks to type-safe STACK_OF() variants.
10194 [Ben Laurie, Ralf S. Engelschall, Steve Henson]
10196 *) Introduce type-safe STACKs. This will almost certainly break lots of code
10197 that links with OpenSSL (well at least cause lots of warnings), but fear
10198 not: the conversion is trivial, and it eliminates loads of evil casts. A
10199 few STACKed things have been converted already. Feel free to convert more.
10200 In the fullness of time, I'll do away with the STACK type altogether.
10203 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate
10204 specified in <certfile> by updating the entry in the index.txt file.
10205 This way one no longer has to edit the index.txt file manually for
10206 revoking a certificate. The -revoke option does the gory details now.
10207 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall]
10209 *) Fix `openssl crl -noout -text' combination where `-noout' killed the
10210 `-text' option at all and this way the `-noout -text' combination was
10211 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'.
10212 [Ralf S. Engelschall]
10214 *) Make sure a corresponding plain text error message exists for the
10215 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
10216 verify callback function determined that a certificate was revoked.
10217 [Ralf S. Engelschall]
10219 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for
10220 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
10221 all available cipers including rc5, which was forgotten until now.
10222 In order to let the testing shell script know which algorithms
10223 are available, a new (up to now undocumented) command
10224 "openssl list-cipher-commands" is used.
10227 *) Bugfix: s_client occasionally would sleep in select() when
10228 it should have checked SSL_pending() first.
10231 *) New functions DSA_do_sign and DSA_do_verify to provide access to
10232 the raw DSA values prior to ASN.1 encoding.
10235 *) Tweaks to Configure
10236 [Niels Poppe <niels@netbox.org>]
10238 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
10242 *) New variables $(RANLIB) and $(PERL) in the Makefiles.
10245 *) New config option to avoid instructions that are illegal on the 80386.
10246 The default code is faster, but requires at least a 486.
10249 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
10250 SSL2_SERVER_VERSION (not used at all) macros, which are now the
10251 same as SSL2_VERSION anyway.
10254 *) New "-showcerts" option for s_client.
10257 *) Still more PKCS#12 integration. Add pkcs12 application to openssl
10258 application. Various cleanups and fixes.
10261 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
10262 modify error routines to work internally. Add error codes and PBE init
10263 to library startup routines.
10266 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and
10267 packing functions to asn1 and evp. Changed function names and error
10268 codes along the way.
10271 *) PKCS12 integration: and so it begins... First of several patches to
10272 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
10273 objects to objects.h
10276 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1
10277 and display support for Thawte strong extranet extension.
10280 *) Add LinuxPPC support.
10281 [Jeff Dubrule <igor@pobox.org>]
10283 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
10284 bn_div_words in alpha.s.
10285 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie]
10287 *) Make sure the RSA OAEP test is skipped under -DRSAref because
10288 OAEP isn't supported when OpenSSL is built with RSAref.
10289 [Ulf Moeller <ulf@fitug.de>]
10291 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
10292 so they no longer are missing under -DNOPROTO.
10293 [Soren S. Jorvang <soren@t.dk>]
10296 Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
10298 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
10299 doesn't work when the session is reused. Coming soon!
10302 *) Fix a security hole, that allows sessions to be reused in the wrong
10303 context thus bypassing client cert protection! All software that uses
10304 client certs and session caches in multiple contexts NEEDS PATCHING to
10305 allow session reuse! A fuller solution is in the works.
10306 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)]
10308 *) Some more source tree cleanups (removed obsolete files
10309 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
10310 permission on "config" script to be executable) and a fix for the INSTALL
10312 [Ulf Moeller <ulf@fitug.de>]
10314 *) Remove some legacy and erroneous uses of malloc, free instead of
10316 [Lennart Bang <lob@netstream.se>, with minor changes by Steve]
10318 *) Make rsa_oaep_test return non-zero on error.
10319 [Ulf Moeller <ulf@fitug.de>]
10321 *) Add support for native Solaris shared libraries. Configure
10322 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
10323 if someone would make that last step automatic.
10324 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>]
10326 *) ctx_size was not built with the right compiler during "make links". Fixed.
10329 *) Change the meaning of 'ALL' in the cipher list. It now means "everything
10330 except NULL ciphers". This means the default cipher list will no longer
10331 enable NULL ciphers. They need to be specifically enabled e.g. with
10332 the string "DEFAULT:eNULL".
10335 *) Fix to RSA private encryption routines: if p < q then it would
10336 occasionally produce an invalid result. This will only happen with
10337 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
10340 *) Be less restrictive and allow also `perl util/perlpath.pl
10341 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin',
10342 because this way one can also use an interpreter named `perl5' (which is
10343 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
10344 installed as `perl').
10345 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
10347 *) Let util/clean-depend.pl work also with older Perl 5.00x versions.
10348 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
10350 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
10351 advapi32.lib to Win32 build and change the pem test comparison
10352 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
10353 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
10354 and crypto/des/ede_cbcm_enc.c.
10357 *) DES quad checksum was broken on big-endian architectures. Fixed.
10360 *) Comment out two functions in bio.h that aren't implemented. Fix up the
10361 Win32 test batch file so it (might) work again. The Win32 test batch file
10362 is horrible: I feel ill....
10365 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
10366 in e_os.h. Audit of header files to check ANSI and non ANSI
10367 sections: 10 functions were absent from non ANSI section and not exported
10368 from Windows DLLs. Fixed up libeay.num for new functions.
10371 *) Make `openssl version' output lines consistent.
10372 [Ralf S. Engelschall]
10374 *) Fix Win32 symbol export lists for BIO functions: Added
10375 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
10376 to ms/libeay{16,32}.def.
10377 [Ralf S. Engelschall]
10379 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
10380 fine under Unix and passes some trivial tests I've now added. But the
10381 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
10382 added to make sure no one expects that this stuff really works in the
10383 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
10384 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
10386 [Ralf S. Engelschall]
10388 *) Fix the generation of two part addresses in perl.
10389 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie]
10391 *) Add config entry for Linux on MIPS.
10392 [John Tobey <jtobey@channel1.com>]
10394 *) Make links whenever Configure is run, unless we are on Windoze.
10397 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf.
10398 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
10402 *) Add a useful kludge to allow package maintainers to specify compiler and
10403 other platforms details on the command line without having to patch the
10404 Configure script everytime: One now can use ``perl Configure
10405 <id>:<details>'', i.e. platform ids are allowed to have details appended
10406 to them (separated by colons). This is treated as there would be a static
10407 pre-configured entry in Configure's %table under key <id> with value
10408 <details> and ``perl Configure <id>'' is called. So, when you want to
10409 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
10410 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"''
10411 now, which overrides the FreeBSD-elf entry on-the-fly.
10412 [Ralf S. Engelschall]
10414 *) Disable new TLS1 ciphersuites by default: they aren't official yet.
10417 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
10418 on the `perl Configure ...' command line. This way one can compile
10419 OpenSSL libraries with Position Independent Code (PIC) which is needed
10420 for linking it into DSOs.
10421 [Ralf S. Engelschall]
10423 *) Remarkably, export ciphers were totally broken and no-one had noticed!
10427 *) Cleaned up the LICENSE document: The official contact for any license
10428 questions now is the OpenSSL core team under openssl-core@openssl.org.
10429 And add a paragraph about the dual-license situation to make sure people
10430 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
10431 to the OpenSSL toolkit.
10432 [Ralf S. Engelschall]
10434 *) General source tree makefile cleanups: Made `making xxx in yyy...'
10435 display consistent in the source tree and replaced `/bin/rm' by `rm'.
10436 Additionally cleaned up the `make links' target: Remove unnecessary
10437 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
10438 to speed processing and no longer clutter the display with confusing
10439 stuff. Instead only the actually done links are displayed.
10440 [Ralf S. Engelschall]
10442 *) Permit null encryption ciphersuites, used for authentication only. It used
10443 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
10444 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
10448 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
10449 signed attributes when verifying signatures (this would break them),
10450 the detached data encoding was wrong and public keys obtained using
10451 X509_get_pubkey() weren't freed.
10454 *) Add text documentation for the BUFFER functions. Also added a work around
10455 to a Win95 console bug. This was triggered by the password read stuff: the
10456 last character typed gets carried over to the next fread(). If you were
10457 generating a new cert request using 'req' for example then the last
10458 character of the passphrase would be CR which would then enter the first
10462 *) Added the new `Includes OpenSSL Cryptography Software' button as
10463 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
10464 button and can be used by applications based on OpenSSL to show the
10465 relationship to the OpenSSL project.
10466 [Ralf S. Engelschall]
10468 *) Remove confusing variables in function signatures in files
10469 ssl/ssl_lib.c and ssl/ssl.h.
10470 [Lennart Bong <lob@kulthea.stacken.kth.se>]
10472 *) Don't install bss_file.c under PREFIX/include/
10473 [Lennart Bong <lob@kulthea.stacken.kth.se>]
10475 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle
10476 functions that return function pointers and has support for NT specific
10477 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
10478 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
10479 unsigned to signed types: this was killing the Win32 compile.
10482 *) Add new certificate file to stack functions,
10483 SSL_add_dir_cert_subjects_to_stack() and
10484 SSL_add_file_cert_subjects_to_stack(). These largely supplant
10485 SSL_load_client_CA_file(), and can be used to add multiple certs easily
10486 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
10487 This means that Apache-SSL and similar packages don't have to mess around
10488 to add as many CAs as they want to the preferred list.
10491 *) Experiment with doxygen documentation. Currently only partially applied to
10493 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with
10494 openssl.doxy as the configuration file.
10497 *) Get rid of remaining C++-style comments which strict C compilers hate.
10498 [Ralf S. Engelschall, pointed out by Carlos Amengual]
10500 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
10501 compiled in by default: it has problems with large keys.
10504 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
10505 DH private keys and/or callback functions which directly correspond to
10506 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
10507 is needed for applications which have to configure certificates on a
10508 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
10510 For the RSA certificate situation is makes no difference, but
10511 for the DSA certificate situation this fixes the "no shared cipher"
10512 problem where the OpenSSL cipher selection procedure failed because the
10513 temporary keys were not overtaken from the context and the API provided
10514 no way to reconfigure them.
10515 The new functions now let applications reconfigure the stuff and they
10516 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
10517 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
10518 non-public-API function ssl_cert_instantiate() is used as a helper
10519 function and also to reduce code redundancy inside ssl_rsa.c.
10520 [Ralf S. Engelschall]
10522 *) Move s_server -dcert and -dkey options out of the undocumented feature
10523 area because they are useful for the DSA situation and should be
10524 recognized by the users.
10525 [Ralf S. Engelschall]
10527 *) Fix the cipher decision scheme for export ciphers: the export bits are
10528 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
10529 SSL_EXP_MASK. So, the original variable has to be used instead of the
10530 already masked variable.
10531 [Richard Levitte <levitte@stacken.kth.se>]
10533 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c
10534 [Richard Levitte <levitte@stacken.kth.se>]
10536 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
10537 from `int' to `unsigned int' because it's a length and initialized by
10538 EVP_DigestFinal() which expects an `unsigned int *'.
10539 [Richard Levitte <levitte@stacken.kth.se>]
10541 *) Don't hard-code path to Perl interpreter on shebang line of Configure
10542 script. Instead use the usual Shell->Perl transition trick.
10543 [Ralf S. Engelschall]
10545 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates
10546 (in addition to RSA certificates) to match the behaviour of `openssl dsa
10547 -noout -modulus' as it's already the case for `openssl rsa -noout
10548 -modulus'. For RSA the -modulus is the real "modulus" while for DSA
10549 currently the public key is printed (a decision which was already done by
10550 `openssl dsa -modulus' in the past) which serves a similar purpose.
10551 Additionally the NO_RSA no longer completely removes the whole -modulus
10552 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
10554 [Ralf S. Engelschall]
10556 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
10557 BIO. See the source (crypto/evp/bio_ok.c) for more info.
10558 [Arne Ansper <arne@ats.cyber.ee>]
10560 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs
10561 to be added. Now both 'req' and 'ca' can use new objects defined in the
10565 *) Add cool BIO that does syslog (or event log on NT).
10566 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie]
10568 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
10569 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
10570 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
10571 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
10574 *) Add preliminary config info for new extension code.
10577 *) Make RSA_NO_PADDING really use no padding.
10578 [Ulf Moeller <ulf@fitug.de>]
10580 *) Generate errors when private/public key check is done.
10583 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
10584 for some CRL extensions and new objects added.
10587 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private
10588 key usage extension and fuller support for authority key id.
10591 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
10592 padding method for RSA, which is recommended for new applications in PKCS
10593 #1 v2.0 (RFC 2437, October 1998).
10594 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
10595 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
10596 against Bleichbacher's attack on RSA.
10597 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
10600 *) Updates to the new SSL compression code
10601 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
10603 *) Fix so that the version number in the master secret, when passed
10604 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
10605 (because the server will not accept higher), that the version number
10606 is 0x03,0x01, not 0x03,0x00
10607 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
10609 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
10610 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
10611 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
10614 *) Support for RAW extensions where an arbitrary extension can be
10615 created by including its DER encoding. See apps/openssl.cnf for
10619 *) Make sure latest Perl versions don't interpret some generated C array
10620 code as Perl array code in the crypto/err/err_genc.pl script.
10621 [Lars Weber <3weber@informatik.uni-hamburg.de>]
10623 *) Modify ms/do_ms.bat to not generate assembly language makefiles since
10624 not many people have the assembler. Various Win32 compilation fixes and
10625 update to the INSTALL.W32 file with (hopefully) more accurate Win32
10626 build instructions.
10629 *) Modify configure script 'Configure' to automatically create crypto/date.h
10630 file under Win32 and also build pem.h from pem.org. New script
10631 util/mkfiles.pl to create the MINFO file on environments that can't do a
10632 'make files': perl util/mkfiles.pl >MINFO should work.
10635 *) Major rework of DES function declarations, in the pursuit of correctness
10636 and purity. As a result, many evil casts evaporated, and some weirdness,
10637 too. You may find this causes warnings in your code. Zapping your evil
10638 casts will probably fix them. Mostly.
10641 *) Fix for a typo in asn1.h. Bug fix to object creation script
10642 obj_dat.pl. It considered a zero in an object definition to mean
10643 "end of object": none of the objects in objects.h have any zeros
10644 so it wasn't spotted.
10645 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>]
10647 *) Add support for Triple DES Cipher Block Chaining with Output Feedback
10648 Masking (CBCM). In the absence of test vectors, the best I have been able
10649 to do is check that the decrypt undoes the encrypt, so far. Send me test
10650 vectors if you have them.
10653 *) Correct calculation of key length for export ciphers (too much space was
10654 allocated for null ciphers). This has not been tested!
10657 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage
10658 message is now correct (it understands "crypto" and "ssl" on its
10659 command line). There is also now an "update" option. This will update
10660 the util/ssleay.num and util/libeay.num files with any new functions.
10662 perl util/mkdef.pl crypto ssl update
10663 it will update them.
10666 *) Overhauled the Perl interface (perl/*):
10667 - ported BN stuff to OpenSSL's different BN library
10668 - made the perl/ source tree CVS-aware
10669 - renamed the package from SSLeay to OpenSSL (the files still contain
10670 their history because I've copied them in the repository)
10671 - removed obsolete files (the test scripts will be replaced
10672 by better Test::Harness variants in the future)
10673 [Ralf S. Engelschall]
10675 *) First cut for a very conservative source tree cleanup:
10676 1. merge various obsolete readme texts into doc/ssleay.txt
10677 where we collect the old documents and readme texts.
10678 2. remove the first part of files where I'm already sure that we no
10679 longer need them because of three reasons: either they are just temporary
10680 files which were left by Eric or they are preserved original files where
10681 I've verified that the diff is also available in the CVS via "cvs diff
10682 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
10683 the crypto/md/ stuff).
10684 [Ralf S. Engelschall]
10686 *) More extension code. Incomplete support for subject and issuer alt
10687 name, issuer and authority key id. Change the i2v function parameters
10688 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
10689 what that's for :-) Fix to ASN1 macro which messed up
10690 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
10693 *) Preliminary support for ENUMERATED type. This is largely copied from the
10697 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
10698 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
10700 *) Make sure `make rehash' target really finds the `openssl' program.
10701 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
10703 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
10704 like to hear about it if this slows down other processors.
10707 *) Add CygWin32 platform information to Configure script.
10708 [Alan Batie <batie@aahz.jf.intel.com>]
10710 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm'
10711 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>]
10713 *) New program nseq to manipulate netscape certificate sequences
10716 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
10720 *) Fixes to BN code. Previously the default was to define BN_RECURSION
10721 but the BN code had some problems that would cause failures when
10722 doing certificate verification and some other functions.
10723 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
10725 *) Add ASN1 and PEM code to support netscape certificate sequences.
10728 *) Add ASN1 and PEM code to support netscape certificate sequences.
10731 *) Add several PKIX and private extended key usage OIDs.
10734 *) Modify the 'ca' program to handle the new extension code. Modify
10735 openssl.cnf for new extension format, add comments.
10738 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
10739 and add a sample to openssl.cnf so req -x509 now adds appropriate
10743 *) Continued X509 V3 changes. Add to other makefiles, integrate with the
10744 error code, add initial support to X509_print() and x509 application.
10747 *) Takes a deep breath and start adding X509 V3 extension support code. Add
10748 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
10749 stuff is currently isolated and isn't even compiled yet.
10752 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL
10753 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
10754 Removed the versions check from X509 routines when loading extensions:
10755 this allows certain broken certificates that don't set the version
10756 properly to be processed.
10759 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
10760 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
10761 can still be regenerated with "make depend".
10764 *) Spelling mistake in C version of CAST-128.
10765 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>]
10767 *) Changes to the error generation code. The perl script err-code.pl
10768 now reads in the old error codes and retains the old numbers, only
10769 adding new ones if necessary. It also only changes the .err files if new
10770 codes are added. The makefiles have been modified to only insert errors
10771 when needed (to avoid needlessly modifying header files). This is done
10772 by only inserting errors if the .err file is newer than the auto generated
10773 C file. To rebuild all the error codes from scratch (the old behaviour)
10774 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
10775 or delete all the .err files.
10778 *) CAST-128 was incorrectly implemented for short keys. The C version has
10779 been fixed, but is untested. The assembler versions are also fixed, but
10780 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
10781 to regenerate it if needed.
10782 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
10783 Hagino <itojun@kame.net>]
10785 *) File was opened incorrectly in randfile.c.
10786 [Ulf Möller <ulf@fitug.de>]
10788 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print
10789 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
10790 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
10791 al: it's just almost always a UTCTime. Note this patch adds new error
10792 codes so do a "make errors" if there are problems.
10795 *) Correct Linux 1 recognition in config.
10796 [Ulf Möller <ulf@fitug.de>]
10798 *) Remove pointless MD5 hash when using DSA keys in ca.
10799 [Anonymous <nobody@replay.com>]
10801 *) Generate an error if given an empty string as a cert directory. Also
10802 generate an error if handed NULL (previously returned 0 to indicate an
10803 error, but didn't set one).
10804 [Ben Laurie, reported by Anonymous <nobody@replay.com>]
10806 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
10809 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
10810 parameters. This was causing a warning which killed off the Win32 compile.
10813 *) Remove C++ style comments from crypto/bn/bn_local.h.
10814 [Neil Costigan <neil.costigan@celocom.com>]
10816 *) The function OBJ_txt2nid was broken. It was supposed to return a nid
10817 based on a text string, looking up short and long names and finally
10818 "dot" format. The "dot" format stuff didn't work. Added new function
10819 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
10820 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
10821 OID is not part of the table.
10824 *) Add prototypes to X509 lookup/verify methods, fixing a bug in
10825 X509_LOOKUP_by_alias().
10828 *) Sort openssl functions by name.
10831 *) Get the gendsa program working (hopefully) and add it to app list. Remove
10832 encryption from sample DSA keys (in case anyone is interested the password
10836 *) Make _all_ *_free functions accept a NULL pointer.
10837 [Frans Heymans <fheymans@isaserver.be>]
10839 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use
10841 [Anonymous <nobody@replay.com>]
10843 *) s_server should send the CAfile as acceptable CAs, not its own cert.
10844 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
10846 *) Don't blow it for numeric -newkey arguments to apps/req.
10847 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
10849 *) Temp key "for export" tests were wrong in s3_srvr.c.
10850 [Anonymous <nobody@replay.com>]
10852 *) Add prototype for temp key callback functions
10853 SSL_CTX_set_tmp_{rsa,dh}_callback().
10856 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
10857 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
10860 *) X509_name_add_entry() freed the wrong thing after an error.
10861 [Arne Ansper <arne@ats.cyber.ee>]
10863 *) rsa_eay.c would attempt to free a NULL context.
10864 [Arne Ansper <arne@ats.cyber.ee>]
10866 *) BIO_s_socket() had a broken should_retry() on Windoze.
10867 [Arne Ansper <arne@ats.cyber.ee>]
10869 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
10870 [Arne Ansper <arne@ats.cyber.ee>]
10872 *) Make sure the already existing X509_STORE->depth variable is initialized
10873 in X509_STORE_new(), but document the fact that this variable is still
10874 unused in the certificate verification process.
10875 [Ralf S. Engelschall]
10877 *) Fix the various library and apps files to free up pkeys obtained from
10878 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
10881 *) Fix reference counting in X509_PUBKEY_get(). This makes
10882 demos/maurice/example2.c work, amongst others, probably.
10883 [Steve Henson and Ben Laurie]
10885 *) First cut of a cleanup for apps/. First the `ssleay' program is now named
10886 `openssl' and second, the shortcut symlinks for the `openssl <command>'
10887 are no longer created. This way we have a single and consistent command
10888 line interface `openssl <command>', similar to `cvs <command>'.
10889 [Ralf S. Engelschall, Paul Sutton and Ben Laurie]
10891 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
10892 BIT STRING wrapper always have zero unused bits.
10895 *) Add CA.pl, perl version of CA.sh, add extended key usage OID.
10898 *) Make the top-level INSTALL documentation easier to understand.
10901 *) Makefiles updated to exit if an error occurs in a sub-directory
10902 make (including if user presses ^C) [Paul Sutton]
10904 *) Make Montgomery context stuff explicit in RSA data structure.
10907 *) Fix build order of pem and err to allow for generated pem.h.
10910 *) Fix renumbering bug in X509_NAME_delete_entry().
10913 *) Enhanced the err-ins.pl script so it makes the error library number
10914 global and can add a library name. This is needed for external ASN1 and
10915 other error libraries.
10918 *) Fixed sk_insert which never worked properly.
10921 *) Fix ASN1 macros so they can handle indefinite length constructed
10922 EXPLICIT tags. Some non standard certificates use these: they can now
10926 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
10927 into a single doc/ssleay.txt bundle. This way the information is still
10928 preserved but no longer messes up this directory. Now it's new room for
10929 the new set of documentation files.
10930 [Ralf S. Engelschall]
10932 *) SETs were incorrectly DER encoded. This was a major pain, because they
10933 shared code with SEQUENCEs, which aren't coded the same. This means that
10934 almost everything to do with SETs or SEQUENCEs has either changed name or
10935 number of arguments.
10936 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>]
10938 *) Fix test data to work with the above.
10941 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
10942 was already fixed by Eric for 0.9.1 it seems.
10943 [Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>]
10945 *) Autodetect FreeBSD3.
10948 *) Fix various bugs in Configure. This affects the following platforms:
10952 unixware-2.0-pentium
10956 *) Eliminate generated files from CVS. Reorder tests to regenerate files
10957 before they are needed.
10960 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy).
10964 Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
10966 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
10967 changed SSLeay to OpenSSL in version strings.
10968 [Ralf S. Engelschall]
10970 *) Some fixups to the top-level documents.
10973 *) Fixed the nasty bug where rsaref.h was not found under compile-time
10974 because the symlink to include/ was missing.
10975 [Ralf S. Engelschall]
10977 *) Incorporated the popular no-RSA/DSA-only patches
10978 which allow to compile a RSA-free SSLeay.
10979 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall]
10981 *) Fixed nasty rehash problem under `make -f Makefile.ssl links'
10982 when "ssleay" is still not found.
10983 [Ralf S. Engelschall]
10985 *) Added more platforms to Configure: Cray T3E, HPUX 11,
10986 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>]
10988 *) Updated the README file.
10989 [Ralf S. Engelschall]
10991 *) Added various .cvsignore files in the CVS repository subdirs
10992 to make a "cvs update" really silent.
10993 [Ralf S. Engelschall]
10995 *) Recompiled the error-definition header files and added
10996 missing symbols to the Win32 linker tables.
10997 [Ralf S. Engelschall]
10999 *) Cleaned up the top-level documents;
11000 o new files: CHANGES and LICENSE
11001 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
11002 o merged COPYRIGHT into LICENSE
11003 o removed obsolete TODO file
11004 o renamed MICROSOFT to INSTALL.W32
11005 [Ralf S. Engelschall]
11007 *) Removed dummy files from the 0.9.1b source tree:
11008 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
11009 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
11010 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
11011 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
11012 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
11013 [Ralf S. Engelschall]
11015 *) Added various platform portability fixes.
11018 *) The Genesis of the OpenSSL rpject:
11019 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
11020 Young and Tim J. Hudson created while they were working for C2Net until
11022 [The OpenSSL Project]
11025 Changes between 0.9.0b and 0.9.1b [not released]
11027 *) Updated a few CA certificates under certs/
11030 *) Changed some BIGNUM api stuff.
11033 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
11034 DGUX x86, Linux Alpha, etc.
11037 *) New COMP library [crypto/comp/] for SSL Record Layer Compression:
11038 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
11042 *) Add -strparse option to asn1pars program which parses nested
11044 [Dr Stephen Henson <shenson@bigfoot.com>]
11046 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
11049 *) DSA fix for "ca" program.
11052 *) Added "-genkey" option to "dsaparam" program.
11055 *) Added RIPE MD160 (rmd160) message digest.
11058 *) Added -a (all) option to "ssleay version" command.
11061 *) Added PLATFORM define which is the id given to Configure.
11064 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
11067 *) Extended the ASN.1 parser routines.
11070 *) Extended BIO routines to support REUSEADDR, seek, tell, etc.
11073 *) Added a BN_CTX to the BN library.
11076 *) Fixed the weak key values in DES library
11079 *) Changed API in EVP library for cipher aliases.
11082 *) Added support for RC2/64bit cipher.
11085 *) Converted the lhash library to the crypto/mem.c functions.
11088 *) Added more recognized ASN.1 object ids.
11091 *) Added more RSA padding checks for SSL/TLS.
11094 *) Added BIO proxy/filter functionality.
11097 *) Added extra_certs to SSL_CTX which can be used
11098 send extra CA certificates to the client in the CA cert chain sending
11099 process. It can be configured with SSL_CTX_add_extra_chain_cert().
11102 *) Now Fortezza is denied in the authentication phase because
11103 this is key exchange mechanism is not supported by SSLeay at all.
11106 *) Additional PKCS1 checks.
11109 *) Support the string "TLSv1" for all TLS v1 ciphers.
11112 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
11113 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
11116 *) Fixed a few memory leaks.
11119 *) Fixed various code and comment typos.
11122 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0
11123 bytes sent in the client random.
11124 [Edward Bishop <ebishop@spyglass.com>]
projects / openssl.git / blob