5 Changes between 1.0.2d and 1.0.2e [xx XXX xxxx]
7 *) In DSA_generate_parameters_ex, if the provided seed is too short,
9 [Rich Salz and Ismo Puustinen <ismo.puustinen@intel.com>]
11 *) Rewrite PSK to support ECDHE_PSK, DHE_PSK and RSA_PSK. Add ciphersuites
12 from RFC4279, RFC4785, RFC5487, RFC5489.
14 Thanks to Christian J. Dietrich and Giuseppe D'Angelo for the
15 original RSA_PSK patch.
18 *) Dropped support for the SSL3_FLAGS_DELAY_CLIENT_FINISHED flag. This SSLeay
19 era flag was never set throughout the codebase (only read). Also removed
20 SSL3_FLAGS_POP_BUFFER which was only used if
21 SSL3_FLAGS_DELAY_CLIENT_FINISHED was also set.
24 *) Changed the default name options in the "ca", "crl", "req" and "x509"
25 to be "oneline" instead of "compat".
28 *) Remove SSL_OP_TLS_BLOCK_PADDING_BUG. This is SSLeay legacy, we're
29 not aware of clients that still exhibit this bug, and the workaround
30 hasn't been working properly for a while.
33 *) The return type of BIO_number_read() and BIO_number_written() as well as
34 the corresponding num_read and num_write members in the BIO structure has
35 changed from unsigned long to uint64_t. On platforms where an unsigned
36 long is 32 bits (e.g. Windows) these counters could overflow if >4Gb is
40 *) Given the pervasive nature of TLS extensions it is inadvisable to run
41 OpenSSL without support for them. It also means that maintaining
42 the OPENSSL_NO_TLSEXT option within the code is very invasive (and probably
43 not well tested). Therefore the OPENSSL_NO_TLSEXT option has been removed.
46 *) Removed support for the two export grade static DH ciphersuites
47 EXP-DH-RSA-DES-CBC-SHA and EXP-DH-DSS-DES-CBC-SHA. These two ciphersuites
48 were newly added (along with a number of other static DH ciphersuites) to
49 1.0.2. However the two export ones have *never* worked since they were
50 introduced. It seems strange in any case to be adding new export
51 ciphersuites, and given "logjam" it also does not seem correct to fix them.
54 *) Version negotiation has been rewritten. In particular SSLv23_method(),
55 SSLv23_client_method() and SSLv23_server_method() have been deprecated,
56 and turned into macros which simply call the new preferred function names
57 TLS_method(), TLS_client_method() and TLS_server_method(). All new code
58 should use the new names instead. Also as part of this change the ssl23.h
59 header file has been removed.
62 *) Support for Kerberos ciphersuites in TLS (RFC2712) has been removed. This
63 code and the associated standard is no longer considered fit-for-purpose.
66 *) RT2547 was closed. When generating a private key, try to make the
67 output file readable only by the owner. This behavior change might
68 be noticeable when interacting with other software.
70 *) Added HTTP GET support to the ocsp command.
73 *) RAND_pseudo_bytes has been deprecated. Users should use RAND_bytes instead.
76 *) Added support for TLS extended master secret from
77 draft-ietf-tls-session-hash-03.txt. Thanks for Alfredo Pironti for an
78 initial patch which was a great help during development.
81 *) All libssl internal structures have been removed from the public header
82 files, and the OPENSSL_NO_SSL_INTERN option has been removed (since it is
83 now redundant). Users should not attempt to access internal structures
84 directly. Instead they should use the provided API functions.
87 *) config has been changed so that by default OPENSSL_NO_DEPRECATED is used.
88 Access to deprecated functions can be re-enabled by running config with
89 "enable-deprecated". In addition applications wishing to use deprecated
90 functions must define OPENSSL_USE_DEPRECATED. Note that this new behaviour
91 will, by default, disable some transitive includes that previously existed
92 in the header files (e.g. ec.h will no longer, by default, include bn.h)
95 *) Added support for OCB mode. OpenSSL has been granted a patent license
96 compatible with the OpenSSL license for use of OCB. Details are available
97 at https://www.openssl.org/docs/misc/OCB-patent-grant-OpenSSL.pdf. Support
98 for OCB can be removed by calling config with no-ocb.
101 *) SSLv2 support has been removed. It still supports receiving a SSLv2
102 compatible client hello.
105 *) Increased the minimal RSA keysize from 256 to 512 bits [Rich Salz],
106 done while fixing the error code for the key-too-small case.
107 [Annie Yousar <a.yousar@informatik.hu-berlin.de>]
109 *) CA.sh has been removmed; use CA.pl instead.
112 *) Removed old DES API.
115 *) Remove various unsupported platforms:
121 Sinix/ReliantUNIX RM400
126 16-bit platforms such as WIN16
129 *) Clean up OPENSSL_NO_xxx #define's
130 Use setbuf() and remove OPENSSL_NO_SETVBUF_IONBF
131 Rename OPENSSL_SYSNAME_xxx to OPENSSL_SYS_xxx
132 OPENSSL_NO_EC{DH,DSA} merged into OPENSSL_NO_EC
133 OPENSSL_NO_RIPEMD160, OPENSSL_NO_RIPEMD merged into OPENSSL_NO_RMD160
134 OPENSSL_NO_FP_API merged into OPENSSL_NO_STDIO
135 Remove OPENSSL_NO_BIO OPENSSL_NO_BUFFER OPENSSL_NO_CHAIN_VERIFY
136 OPENSSL_NO_EVP OPENSSL_NO_FIPS_ERR OPENSSL_NO_HASH_COMP
137 OPENSSL_NO_LHASH OPENSSL_NO_OBJECT OPENSSL_NO_SPEED OPENSSL_NO_STACK
138 OPENSSL_NO_X509 OPENSSL_NO_X509_VERIFY
139 Remove MS_STATIC; it's a relic from platforms <32 bits.
142 *) Cleaned up dead code
143 Remove all but one '#ifdef undef' which is to be looked at.
146 *) Clean up calling of xxx_free routines.
147 Just like free(), fix most of the xxx_free routines to accept
148 NULL. Remove the non-null checks from callers. Save much code.
151 *) Add secure heap for storage of private keys (when possible).
152 Add BIO_s_secmem(), CBIGNUM, etc.
153 Contributed by Akamai Technologies under our Corporate CLA.
156 *) Experimental support for a new, fast, unbiased prime candidate generator,
157 bn_probable_prime_dh_coprime(). Not currently used by any prime generator.
158 [Felix Laurie von Massenbach <felix@erbridge.co.uk>]
160 *) New output format NSS in the sess_id command line tool. This allows
161 exporting the session id and the master key in NSS keylog format.
162 [Martin Kaiser <martin@kaiser.cx>]
164 *) Harmonize version and its documentation. -f flag is used to display
166 [mancha <mancha1@zoho.com>]
168 *) Fix eckey_priv_encode so it immediately returns an error upon a failure
169 in i2d_ECPrivateKey. Thanks to Ted Unangst for feedback on this issue.
170 [mancha <mancha1@zoho.com>]
172 *) Fix some double frees. These are not thought to be exploitable.
173 [mancha <mancha1@zoho.com>]
175 *) A missing bounds check in the handling of the TLS heartbeat extension
176 can be used to reveal up to 64k of memory to a connected client or
179 Thanks for Neel Mehta of Google Security for discovering this bug and to
180 Adam Langley <agl@chromium.org> and Bodo Moeller <bmoeller@acm.org> for
181 preparing the fix (CVE-2014-0160)
182 [Adam Langley, Bodo Moeller]
184 *) Fix for the attack described in the paper "Recovering OpenSSL
185 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
186 by Yuval Yarom and Naomi Benger. Details can be obtained from:
187 http://eprint.iacr.org/2014/140
189 Thanks to Yuval Yarom and Naomi Benger for discovering this
190 flaw and to Yuval Yarom for supplying a fix (CVE-2014-0076)
191 [Yuval Yarom and Naomi Benger]
193 *) Use algorithm specific chains in SSL_CTX_use_certificate_chain_file():
194 this fixes a limitation in previous versions of OpenSSL.
197 *) Experimental encrypt-then-mac support.
199 Experimental support for encrypt then mac from
200 draft-gutmann-tls-encrypt-then-mac-02.txt
202 To enable it set the appropriate extension number (0x42 for the test
203 server) using e.g. -DTLSEXT_TYPE_encrypt_then_mac=0x42
205 For non-compliant peers (i.e. just about everything) this should have no
208 WARNING: EXPERIMENTAL, SUBJECT TO CHANGE.
212 *) Add EVP support for key wrapping algorithms, to avoid problems with
213 existing code the flag EVP_CIPHER_CTX_WRAP_ALLOW has to be set in
214 the EVP_CIPHER_CTX or an error is returned. Add AES and DES3 wrap
215 algorithms and include tests cases.
218 *) Extend CMS code to support RSA-PSS signatures and RSA-OAEP for
222 *) Extended RSA OAEP support via EVP_PKEY API. Options to specify digest,
223 MGF1 digest and OAEP label.
226 *) Make openssl verify return errors.
227 [Chris Palmer <palmer@google.com> and Ben Laurie]
229 *) New function ASN1_TIME_diff to calculate the difference between two
230 ASN1_TIME structures or one structure and the current time.
233 *) Update fips_test_suite to support multiple command line options. New
234 test to induce all self test errors in sequence and check expected
238 *) Add FIPS_{rsa,dsa,ecdsa}_{sign,verify} functions which digest and
239 sign or verify all in one operation.
242 *) Add fips_algvs: a multicall fips utility incorporating all the algorithm
243 test programs and fips_test_suite. Includes functionality to parse
244 the minimal script output of fipsalgest.pl directly.
247 *) Add authorisation parameter to FIPS_module_mode_set().
250 *) Add FIPS selftest for ECDH algorithm using P-224 and B-233 curves.
253 *) Use separate DRBG fields for internal and external flags. New function
254 FIPS_drbg_health_check() to perform on demand health checking. Add
255 generation tests to fips_test_suite with reduced health check interval to
256 demonstrate periodic health checking. Add "nodh" option to
257 fips_test_suite to skip very slow DH test.
260 *) New function FIPS_get_cipherbynid() to lookup FIPS supported ciphers
264 *) More extensive health check for DRBG checking many more failure modes.
265 New function FIPS_selftest_drbg_all() to handle every possible DRBG
266 combination: call this in fips_test_suite.
269 *) Add support for Dual EC DRBG from SP800-90. Update DRBG algorithm test
270 and POST to handle Dual EC cases.
273 *) Add support for canonical generation of DSA parameter 'g'. See
276 *) Add support for HMAC DRBG from SP800-90. Update DRBG algorithm test and
277 POST to handle HMAC cases.
280 *) Add functions FIPS_module_version() and FIPS_module_version_text()
281 to return numerical and string versions of the FIPS module number.
284 *) Rename FIPS_mode_set and FIPS_mode to FIPS_module_mode_set and
285 FIPS_module_mode. FIPS_mode and FIPS_mode_set will be implemented
286 outside the validated module in the FIPS capable OpenSSL.
289 *) Minor change to DRBG entropy callback semantics. In some cases
290 there is no multiple of the block length between min_len and
291 max_len. Allow the callback to return more than max_len bytes
292 of entropy but discard any extra: it is the callback's responsibility
293 to ensure that the extra data discarded does not impact the
294 requested amount of entropy.
297 *) Add PRNG security strength checks to RSA, DSA and ECDSA using
298 information in FIPS186-3, SP800-57 and SP800-131A.
301 *) CCM support via EVP. Interface is very similar to GCM case except we
302 must supply all data in one chunk (i.e. no update, final) and the
303 message length must be supplied if AAD is used. Add algorithm test
307 *) Initial version of POST overhaul. Add POST callback to allow the status
308 of POST to be monitored and/or failures induced. Modify fips_test_suite
309 to use callback. Always run all selftests even if one fails.
312 *) XTS support including algorithm test driver in the fips_gcmtest program.
313 Note: this does increase the maximum key length from 32 to 64 bytes but
314 there should be no binary compatibility issues as existing applications
315 will never use XTS mode.
318 *) Extensive reorganisation of FIPS PRNG behaviour. Remove all dependencies
319 to OpenSSL RAND code and replace with a tiny FIPS RAND API which also
320 performs algorithm blocking for unapproved PRNG types. Also do not
321 set PRNG type in FIPS_mode_set(): leave this to the application.
322 Add default OpenSSL DRBG handling: sets up FIPS PRNG and seeds with
323 the standard OpenSSL PRNG: set additional data to a date time vector.
326 *) Rename old X9.31 PRNG functions of the form FIPS_rand* to FIPS_x931*.
327 This shouldn't present any incompatibility problems because applications
328 shouldn't be using these directly and any that are will need to rethink
329 anyway as the X9.31 PRNG is now deprecated by FIPS 140-2
332 *) Extensive self tests and health checking required by SP800-90 DRBG.
333 Remove strength parameter from FIPS_drbg_instantiate and always
334 instantiate at maximum supported strength.
337 *) Add ECDH code to fips module and fips_ecdhvs for primitives only testing.
340 *) New algorithm test program fips_dhvs to handle DH primitives only testing.
343 *) New function DH_compute_key_padded() to compute a DH key and pad with
344 leading zeroes if needed: this complies with SP800-56A et al.
347 *) Initial implementation of SP800-90 DRBGs for Hash and CTR. Not used by
348 anything, incomplete, subject to change and largely untested at present.
351 *) Modify fipscanisteronly build option to only build the necessary object
352 files by filtering FIPS_EX_OBJ through a perl script in crypto/Makefile.
355 *) Add experimental option FIPSSYMS to give all symbols in
356 fipscanister.o and FIPS or fips prefix. This will avoid
357 conflicts with future versions of OpenSSL. Add perl script
358 util/fipsas.pl to preprocess assembly language source files
359 and rename any affected symbols.
362 *) Add selftest checks and algorithm block of non-fips algorithms in
363 FIPS mode. Remove DES2 from selftests.
366 *) Add ECDSA code to fips module. Add tiny fips_ecdsa_check to just
367 return internal method without any ENGINE dependencies. Add new
368 tiny fips sign and verify functions.
371 *) New build option no-ec2m to disable characteristic 2 code.
374 *) New build option "fipscanisteronly". This only builds fipscanister.o
375 and (currently) associated fips utilities. Uses the file Makefile.fips
376 instead of Makefile.org as the prototype.
379 *) Add some FIPS mode restrictions to GCM. Add internal IV generator.
380 Update fips_gcmtest to use IV generator.
383 *) Initial, experimental EVP support for AES-GCM. AAD can be input by
384 setting output buffer to NULL. The *Final function must be
385 called although it will not retrieve any additional data. The tag
386 can be set or retrieved with a ctrl. The IV length is by default 12
387 bytes (96 bits) but can be set to an alternative value. If the IV
388 length exceeds the maximum IV length (currently 16 bytes) it cannot be
392 *) New flag in ciphers: EVP_CIPH_FLAG_CUSTOM_CIPHER. This means the
393 underlying do_cipher function handles all cipher semantics itself
394 including padding and finalisation. This is useful if (for example)
395 an ENGINE cipher handles block padding itself. The behaviour of
396 do_cipher is subtly changed if this flag is set: the return value
397 is the number of characters written to the output buffer (zero is
398 no longer an error code) or a negative error code. Also if the
399 input buffer is NULL and length 0 finalisation should be performed.
402 *) If a candidate issuer certificate is already part of the constructed
403 path ignore it: new debug notification X509_V_ERR_PATH_LOOP for this case.
406 *) Improve forward-security support: add functions
408 void SSL_CTX_set_not_resumable_session_callback(SSL_CTX *ctx, int (*cb)(SSL *ssl, int is_forward_secure))
409 void SSL_set_not_resumable_session_callback(SSL *ssl, int (*cb)(SSL *ssl, int is_forward_secure))
411 for use by SSL/TLS servers; the callback function will be called whenever a
412 new session is created, and gets to decide whether the session may be
413 cached to make it resumable (return 0) or not (return 1). (As by the
414 SSL/TLS protocol specifications, the session_id sent by the server will be
415 empty to indicate that the session is not resumable; also, the server will
416 not generate RFC 4507 (RFC 5077) session tickets.)
418 A simple reasonable callback implementation is to return is_forward_secure.
419 This parameter will be set to 1 or 0 depending on the ciphersuite selected
420 by the SSL/TLS server library, indicating whether it can provide forward
422 [Emilia Käsper <emilia.kasper@esat.kuleuven.be> (Google)]
424 *) New -verify_name option in command line utilities to set verification
428 *) Initial CMAC implementation. WARNING: EXPERIMENTAL, API MAY CHANGE.
429 Add CMAC pkey methods.
432 *) Experimental renegotiation in s_server -www mode. If the client
433 browses /reneg connection is renegotiated. If /renegcert it is
434 renegotiated requesting a certificate.
437 *) Add an "external" session cache for debugging purposes to s_server. This
438 should help trace issues which normally are only apparent in deployed
439 multi-process servers.
442 *) Extensive audit of libcrypto with DEBUG_UNUSED. Fix many cases where
443 return value is ignored. NB. The functions RAND_add(), RAND_seed(),
444 BIO_set_cipher() and some obscure PEM functions were changed so they
445 can now return an error. The RAND changes required a change to the
446 RAND_METHOD structure.
449 *) New macro __owur for "OpenSSL Warn Unused Result". This makes use of
450 a gcc attribute to warn if the result of a function is ignored. This
451 is enable if DEBUG_UNUSED is set. Add to several functions in evp.h
452 whose return value is often ignored.
454 >>>>>>> f00a10b... GH367: Fix dsa keygen for too-short seed
456 Changes between 1.0.2c and 1.0.2d [9 Jul 2015]
458 *) Alternate chains certificate forgery
460 During certificate verfification, OpenSSL will attempt to find an
461 alternative certificate chain if the first attempt to build such a chain
462 fails. An error in the implementation of this logic can mean that an
463 attacker could cause certain checks on untrusted certificates to be
464 bypassed, such as the CA flag, enabling them to use a valid leaf
465 certificate to act as a CA and "issue" an invalid certificate.
467 This issue was reported to OpenSSL by Adam Langley/David Benjamin
471 Changes between 1.0.2b and 1.0.2c [12 Jun 2015]
473 *) Fix HMAC ABI incompatibility. The previous version introduced an ABI
474 incompatibility in the handling of HMAC. The previous ABI has now been
477 Changes between 1.0.2a and 1.0.2b [11 Jun 2015]
479 *) Malformed ECParameters causes infinite loop
481 When processing an ECParameters structure OpenSSL enters an infinite loop
482 if the curve specified is over a specially malformed binary polynomial
485 This can be used to perform denial of service against any
486 system which processes public keys, certificate requests or
487 certificates. This includes TLS clients and TLS servers with
488 client authentication enabled.
490 This issue was reported to OpenSSL by Joseph Barr-Pixton.
494 *) Exploitable out-of-bounds read in X509_cmp_time
496 X509_cmp_time does not properly check the length of the ASN1_TIME
497 string and can read a few bytes out of bounds. In addition,
498 X509_cmp_time accepts an arbitrary number of fractional seconds in the
501 An attacker can use this to craft malformed certificates and CRLs of
502 various sizes and potentially cause a segmentation fault, resulting in
503 a DoS on applications that verify certificates or CRLs. TLS clients
504 that verify CRLs are affected. TLS clients and servers with client
505 authentication enabled may be affected if they use custom verification
508 This issue was reported to OpenSSL by Robert Swiecki (Google), and
509 independently by Hanno Böck.
513 *) PKCS7 crash with missing EnvelopedContent
515 The PKCS#7 parsing code does not handle missing inner EncryptedContent
516 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs
517 with missing content and trigger a NULL pointer dereference on parsing.
519 Applications that decrypt PKCS#7 data or otherwise parse PKCS#7
520 structures from untrusted sources are affected. OpenSSL clients and
521 servers are not affected.
523 This issue was reported to OpenSSL by Michal Zalewski (Google).
527 *) CMS verify infinite loop with unknown hash function
529 When verifying a signedData message the CMS code can enter an infinite loop
530 if presented with an unknown hash function OID. This can be used to perform
531 denial of service against any system which verifies signedData messages using
533 This issue was reported to OpenSSL by Johannes Bauer.
537 *) Race condition handling NewSessionTicket
539 If a NewSessionTicket is received by a multi-threaded client when attempting to
540 reuse a previous ticket then a race condition can occur potentially leading to
541 a double free of the ticket data.
545 *) Removed support for the two export grade static DH ciphersuites
546 EXP-DH-RSA-DES-CBC-SHA and EXP-DH-DSS-DES-CBC-SHA. These two ciphersuites
547 were newly added (along with a number of other static DH ciphersuites) to
548 1.0.2. However the two export ones have *never* worked since they were
549 introduced. It seems strange in any case to be adding new export
550 ciphersuites, and given "logjam" it also does not seem correct to fix them.
553 *) Only support 256-bit or stronger elliptic curves with the
554 'ecdh_auto' setting (server) or by default (client). Of supported
555 curves, prefer P-256 (both).
558 *) Reject DH handshakes with parameters shorter than 768 bits.
559 [Kurt Roeckx and Emilia Kasper]
561 Changes between 1.0.2 and 1.0.2a [19 Mar 2015]
563 *) ClientHello sigalgs DoS fix
565 If a client connects to an OpenSSL 1.0.2 server and renegotiates with an
566 invalid signature algorithms extension a NULL pointer dereference will
567 occur. This can be exploited in a DoS attack against the server.
569 This issue was was reported to OpenSSL by David Ramos of Stanford
572 [Stephen Henson and Matt Caswell]
574 *) Multiblock corrupted pointer fix
576 OpenSSL 1.0.2 introduced the "multiblock" performance improvement. This
577 feature only applies on 64 bit x86 architecture platforms that support AES
578 NI instructions. A defect in the implementation of "multiblock" can cause
579 OpenSSL's internal write buffer to become incorrectly set to NULL when
580 using non-blocking IO. Typically, when the user application is using a
581 socket BIO for writing, this will only result in a failed connection.
582 However if some other BIO is used then it is likely that a segmentation
583 fault will be triggered, thus enabling a potential DoS attack.
585 This issue was reported to OpenSSL by Daniel Danner and Rainer Mueller.
589 *) Segmentation fault in DTLSv1_listen fix
591 The DTLSv1_listen function is intended to be stateless and processes the
592 initial ClientHello from many peers. It is common for user code to loop
593 over the call to DTLSv1_listen until a valid ClientHello is received with
594 an associated cookie. A defect in the implementation of DTLSv1_listen means
595 that state is preserved in the SSL object from one invocation to the next
596 that can lead to a segmentation fault. Errors processing the initial
597 ClientHello can trigger this scenario. An example of such an error could be
598 that a DTLS1.0 only client is attempting to connect to a DTLS1.2 only
601 This issue was reported to OpenSSL by Per Allansson.
605 *) Segmentation fault in ASN1_TYPE_cmp fix
607 The function ASN1_TYPE_cmp will crash with an invalid read if an attempt is
608 made to compare ASN.1 boolean types. Since ASN1_TYPE_cmp is used to check
609 certificate signature algorithm consistency this can be used to crash any
610 certificate verification operation and exploited in a DoS attack. Any
611 application which performs certificate verification is vulnerable including
612 OpenSSL clients and servers which enable client authentication.
616 *) Segmentation fault for invalid PSS parameters fix
618 The signature verification routines will crash with a NULL pointer
619 dereference if presented with an ASN.1 signature using the RSA PSS
620 algorithm and invalid parameters. Since these routines are used to verify
621 certificate signature algorithms this can be used to crash any
622 certificate verification operation and exploited in a DoS attack. Any
623 application which performs certificate verification is vulnerable including
624 OpenSSL clients and servers which enable client authentication.
626 This issue was was reported to OpenSSL by Brian Carpenter.
630 *) ASN.1 structure reuse memory corruption fix
632 Reusing a structure in ASN.1 parsing may allow an attacker to cause
633 memory corruption via an invalid write. Such reuse is and has been
634 strongly discouraged and is believed to be rare.
636 Applications that parse structures containing CHOICE or ANY DEFINED BY
637 components may be affected. Certificate parsing (d2i_X509 and related
638 functions) are however not affected. OpenSSL clients and servers are
643 *) PKCS7 NULL pointer dereferences fix
645 The PKCS#7 parsing code does not handle missing outer ContentInfo
646 correctly. An attacker can craft malformed ASN.1-encoded PKCS#7 blobs with
647 missing content and trigger a NULL pointer dereference on parsing.
649 Applications that verify PKCS#7 signatures, decrypt PKCS#7 data or
650 otherwise parse PKCS#7 structures from untrusted sources are
651 affected. OpenSSL clients and servers are not affected.
653 This issue was reported to OpenSSL by Michal Zalewski (Google).
657 *) DoS via reachable assert in SSLv2 servers fix
659 A malicious client can trigger an OPENSSL_assert (i.e., an abort) in
660 servers that both support SSLv2 and enable export cipher suites by sending
661 a specially crafted SSLv2 CLIENT-MASTER-KEY message.
663 This issue was discovered by Sean Burford (Google) and Emilia Käsper
664 (OpenSSL development team).
668 *) Empty CKE with client auth and DHE fix
670 If client auth is used then a server can seg fault in the event of a DHE
671 ciphersuite being selected and a zero length ClientKeyExchange message
672 being sent by the client. This could be exploited in a DoS attack.
676 *) Handshake with unseeded PRNG fix
678 Under certain conditions an OpenSSL 1.0.2 client can complete a handshake
679 with an unseeded PRNG. The conditions are:
680 - The client is on a platform where the PRNG has not been seeded
681 automatically, and the user has not seeded manually
682 - A protocol specific client method version has been used (i.e. not
683 SSL_client_methodv23)
684 - A ciphersuite is used that does not require additional random data from
685 the PRNG beyond the initial ClientHello client random (e.g. PSK-RC4-SHA).
687 If the handshake succeeds then the client random that has been used will
688 have been generated from a PRNG with insufficient entropy and therefore the
689 output may be predictable.
691 For example using the following command with an unseeded openssl will
692 succeed on an unpatched platform:
694 openssl s_client -psk 1a2b3c4d -tls1_2 -cipher PSK-RC4-SHA
698 *) Use After Free following d2i_ECPrivatekey error fix
700 A malformed EC private key file consumed via the d2i_ECPrivateKey function
701 could cause a use after free condition. This, in turn, could cause a double
702 free in several private key parsing functions (such as d2i_PrivateKey
703 or EVP_PKCS82PKEY) and could lead to a DoS attack or memory corruption
704 for applications that receive EC private keys from untrusted
705 sources. This scenario is considered rare.
707 This issue was discovered by the BoringSSL project and fixed in their
712 *) X509_to_X509_REQ NULL pointer deref fix
714 The function X509_to_X509_REQ will crash with a NULL pointer dereference if
715 the certificate key is invalid. This function is rarely used in practice.
717 This issue was discovered by Brian Carpenter.
721 *) Removed the export ciphers from the DEFAULT ciphers
724 Changes between 1.0.1l and 1.0.2 [22 Jan 2015]
726 *) Facilitate "universal" ARM builds targeting range of ARM ISAs, e.g.
727 ARMv5 through ARMv8, as opposite to "locking" it to single one.
728 So far those who have to target multiple plaforms would compromise
729 and argue that binary targeting say ARMv5 would still execute on
730 ARMv8. "Universal" build resolves this compromise by providing
731 near-optimal performance even on newer platforms.
734 *) Accelerated NIST P-256 elliptic curve implementation for x86_64
735 (other platforms pending).
736 [Shay Gueron & Vlad Krasnov (Intel Corp), Andy Polyakov]
738 *) Add support for the SignedCertificateTimestampList certificate and
739 OCSP response extensions from RFC6962.
742 *) Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
743 for corner cases. (Certain input points at infinity could lead to
744 bogus results, with non-infinity inputs mapped to infinity too.)
747 *) Initial support for PowerISA 2.0.7, first implemented in POWER8.
748 This covers AES, SHA256/512 and GHASH. "Initial" means that most
749 common cases are optimized and there still is room for further
750 improvements. Vector Permutation AES for Altivec is also added.
753 *) Add support for little-endian ppc64 Linux target.
754 [Marcelo Cerri (IBM)]
756 *) Initial support for AMRv8 ISA crypto extensions. This covers AES,
757 SHA1, SHA256 and GHASH. "Initial" means that most common cases
758 are optimized and there still is room for further improvements.
759 Both 32- and 64-bit modes are supported.
760 [Andy Polyakov, Ard Biesheuvel (Linaro)]
762 *) Improved ARMv7 NEON support.
765 *) Support for SPARC Architecture 2011 crypto extensions, first
766 implemented in SPARC T4. This covers AES, DES, Camellia, SHA1,
767 SHA256/512, MD5, GHASH and modular exponentiation.
768 [Andy Polyakov, David Miller]
770 *) Accelerated modular exponentiation for Intel processors, a.k.a.
772 [Shay Gueron & Vlad Krasnov (Intel Corp)]
774 *) Support for new and upcoming Intel processors, including AVX2,
775 BMI and SHA ISA extensions. This includes additional "stitched"
776 implementations, AESNI-SHA256 and GCM, and multi-buffer support
779 This work was sponsored by Intel Corp.
782 *) Support for DTLS 1.2. This adds two sets of DTLS methods: DTLS_*_method()
783 supports both DTLS 1.2 and 1.0 and should use whatever version the peer
784 supports and DTLSv1_2_*_method() which supports DTLS 1.2 only.
787 *) Use algorithm specific chains in SSL_CTX_use_certificate_chain_file():
788 this fixes a limiation in previous versions of OpenSSL.
791 *) Extended RSA OAEP support via EVP_PKEY API. Options to specify digest,
792 MGF1 digest and OAEP label.
795 *) Add EVP support for key wrapping algorithms, to avoid problems with
796 existing code the flag EVP_CIPHER_CTX_WRAP_ALLOW has to be set in
797 the EVP_CIPHER_CTX or an error is returned. Add AES and DES3 wrap
798 algorithms and include tests cases.
801 *) Add functions to allocate and set the fields of an ECDSA_METHOD
803 [Douglas E. Engert, Steve Henson]
805 *) New functions OPENSSL_gmtime_diff and ASN1_TIME_diff to find the
806 difference in days and seconds between two tm or ASN1_TIME structures.
809 *) Add -rev test option to s_server to just reverse order of characters
810 received by client and send back to server. Also prints an abbreviated
811 summary of the connection parameters.
814 *) New option -brief for s_client and s_server to print out a brief summary
815 of connection parameters.
818 *) Add callbacks for arbitrary TLS extensions.
819 [Trevor Perrin <trevp@trevp.net> and Ben Laurie]
821 *) New option -crl_download in several openssl utilities to download CRLs
822 from CRLDP extension in certificates.
825 *) New options -CRL and -CRLform for s_client and s_server for CRLs.
828 *) New function X509_CRL_diff to generate a delta CRL from the difference
829 of two full CRLs. Add support to "crl" utility.
832 *) New functions to set lookup_crls function and to retrieve
833 X509_STORE from X509_STORE_CTX.
836 *) Print out deprecated issuer and subject unique ID fields in
840 *) Extend OCSP I/O functions so they can be used for simple general purpose
841 HTTP as well as OCSP. New wrapper function which can be used to download
842 CRLs using the OCSP API.
845 *) Delegate command line handling in s_client/s_server to SSL_CONF APIs.
848 *) SSL_CONF* functions. These provide a common framework for application
849 configuration using configuration files or command lines.
852 *) SSL/TLS tracing code. This parses out SSL/TLS records using the
853 message callback and prints the results. Needs compile time option
854 "enable-ssl-trace". New options to s_client and s_server to enable
858 *) New ctrl and macro to retrieve supported points extensions.
859 Print out extension in s_server and s_client.
862 *) New functions to retrieve certificate signature and signature
866 *) Add functions to retrieve and manipulate the raw cipherlist sent by a
870 *) New Suite B modes for TLS code. These use and enforce the requirements
871 of RFC6460: restrict ciphersuites, only permit Suite B algorithms and
872 only use Suite B curves. The Suite B modes can be set by using the
873 strings "SUITEB128", "SUITEB192" or "SUITEB128ONLY" for the cipherstring.
876 *) New chain verification flags for Suite B levels of security. Check
877 algorithms are acceptable when flags are set in X509_verify_cert.
880 *) Make tls1_check_chain return a set of flags indicating checks passed
881 by a certificate chain. Add additional tests to handle client
882 certificates: checks for matching certificate type and issuer name
886 *) If an attempt is made to use a signature algorithm not in the peer
887 preference list abort the handshake. If client has no suitable
888 signature algorithms in response to a certificate request do not
892 *) If server EC tmp key is not in client preference list abort handshake.
895 *) Add support for certificate stores in CERT structure. This makes it
896 possible to have different stores per SSL structure or one store in
897 the parent SSL_CTX. Include distint stores for certificate chain
898 verification and chain building. New ctrl SSL_CTRL_BUILD_CERT_CHAIN
899 to build and store a certificate chain in CERT structure: returing
900 an error if the chain cannot be built: this will allow applications
901 to test if a chain is correctly configured.
903 Note: if the CERT based stores are not set then the parent SSL_CTX
904 store is used to retain compatibility with existing behaviour.
908 *) New function ssl_set_client_disabled to set a ciphersuite disabled
909 mask based on the current session, check mask when sending client
910 hello and checking the requested ciphersuite.
913 *) New ctrls to retrieve and set certificate types in a certificate
914 request message. Print out received values in s_client. If certificate
915 types is not set with custom values set sensible values based on
916 supported signature algorithms.
919 *) Support for distinct client and server supported signature algorithms.
922 *) Add certificate callback. If set this is called whenever a certificate
923 is required by client or server. An application can decide which
924 certificate chain to present based on arbitrary criteria: for example
925 supported signature algorithms. Add very simple example to s_server.
926 This fixes many of the problems and restrictions of the existing client
927 certificate callback: for example you can now clear an existing
928 certificate and specify the whole chain.
931 *) Add new "valid_flags" field to CERT_PKEY structure which determines what
932 the certificate can be used for (if anything). Set valid_flags field
933 in new tls1_check_chain function. Simplify ssl_set_cert_masks which used
934 to have similar checks in it.
936 Add new "cert_flags" field to CERT structure and include a "strict mode".
937 This enforces some TLS certificate requirements (such as only permitting
938 certificate signature algorithms contained in the supported algorithms
939 extension) which some implementations ignore: this option should be used
940 with caution as it could cause interoperability issues.
943 *) Update and tidy signature algorithm extension processing. Work out
944 shared signature algorithms based on preferences and peer algorithms
945 and print them out in s_client and s_server. Abort handshake if no
946 shared signature algorithms.
949 *) Add new functions to allow customised supported signature algorithms
950 for SSL and SSL_CTX structures. Add options to s_client and s_server
954 *) New function SSL_certs_clear() to delete all references to certificates
955 from an SSL structure. Before this once a certificate had been added
956 it couldn't be removed.
959 *) Integrate hostname, email address and IP address checking with certificate
960 verification. New verify options supporting checking in opensl utility.
963 *) Fixes and wildcard matching support to hostname and email checking
964 functions. Add manual page.
965 [Florian Weimer (Red Hat Product Security Team)]
967 *) New functions to check a hostname email or IP address against a
968 certificate. Add options x509 utility to print results of checks against
972 *) Fix OCSP checking.
973 [Rob Stradling <rob.stradling@comodo.com> and Ben Laurie]
975 *) Initial experimental support for explicitly trusted non-root CAs.
976 OpenSSL still tries to build a complete chain to a root but if an
977 intermediate CA has a trust setting included that is used. The first
978 setting is used: whether to trust (e.g., -addtrust option to the x509
982 *) Add -trusted_first option which attempts to find certificates in the
983 trusted store even if an untrusted chain is also supplied.
986 *) MIPS assembly pack updates: support for MIPS32r2 and SmartMIPS ASE,
987 platform support for Linux and Android.
990 *) Support for linux-x32, ILP32 environment in x86_64 framework.
993 *) Experimental multi-implementation support for FIPS capable OpenSSL.
994 When in FIPS mode the approved implementations are used as normal,
995 when not in FIPS mode the internal unapproved versions are used instead.
996 This means that the FIPS capable OpenSSL isn't forced to use the
997 (often lower perfomance) FIPS implementations outside FIPS mode.
1000 *) Transparently support X9.42 DH parameters when calling
1001 PEM_read_bio_DHparameters. This means existing applications can handle
1002 the new parameter format automatically.
1005 *) Initial experimental support for X9.42 DH parameter format: mainly
1006 to support use of 'q' parameter for RFC5114 parameters.
1009 *) Add DH parameters from RFC5114 including test data to dhtest.
1012 *) Support for automatic EC temporary key parameter selection. If enabled
1013 the most preferred EC parameters are automatically used instead of
1014 hardcoded fixed parameters. Now a server just has to call:
1015 SSL_CTX_set_ecdh_auto(ctx, 1) and the server will automatically
1016 support ECDH and use the most appropriate parameters.
1019 *) Enhance and tidy EC curve and point format TLS extension code. Use
1020 static structures instead of allocation if default values are used.
1021 New ctrls to set curves we wish to support and to retrieve shared curves.
1022 Print out shared curves in s_server. New options to s_server and s_client
1023 to set list of supported curves.
1026 *) New ctrls to retrieve supported signature algorithms and
1027 supported curve values as an array of NIDs. Extend openssl utility
1028 to print out received values.
1031 *) Add new APIs EC_curve_nist2nid and EC_curve_nid2nist which convert
1032 between NIDs and the more common NIST names such as "P-256". Enhance
1033 ecparam utility and ECC method to recognise the NIST names for curves.
1036 *) Enhance SSL/TLS certificate chain handling to support different
1037 chains for each certificate instead of one chain in the parent SSL_CTX.
1040 *) Support for fixed DH ciphersuite client authentication: where both
1041 server and client use DH certificates with common parameters.
1044 *) Support for fixed DH ciphersuites: those requiring DH server
1048 *) New function i2d_re_X509_tbs for re-encoding the TBS portion of
1050 Note: Related 1.0.2-beta specific macros X509_get_cert_info,
1051 X509_CINF_set_modified, X509_CINF_get_issuer, X509_CINF_get_extensions and
1052 X509_CINF_get_signature were reverted post internal team review.
1054 Changes between 1.0.1k and 1.0.1l [15 Jan 2015]
1056 *) Build fixes for the Windows and OpenVMS platforms
1057 [Matt Caswell and Richard Levitte]
1059 Changes between 1.0.1j and 1.0.1k [8 Jan 2015]
1061 *) Fix DTLS segmentation fault in dtls1_get_record. A carefully crafted DTLS
1062 message can cause a segmentation fault in OpenSSL due to a NULL pointer
1063 dereference. This could lead to a Denial Of Service attack. Thanks to
1064 Markus Stenberg of Cisco Systems, Inc. for reporting this issue.
1068 *) Fix DTLS memory leak in dtls1_buffer_record. A memory leak can occur in the
1069 dtls1_buffer_record function under certain conditions. In particular this
1070 could occur if an attacker sent repeated DTLS records with the same
1071 sequence number but for the next epoch. The memory leak could be exploited
1072 by an attacker in a Denial of Service attack through memory exhaustion.
1073 Thanks to Chris Mueller for reporting this issue.
1077 *) Fix issue where no-ssl3 configuration sets method to NULL. When openssl is
1078 built with the no-ssl3 option and a SSL v3 ClientHello is received the ssl
1079 method would be set to NULL which could later result in a NULL pointer
1080 dereference. Thanks to Frank Schmirler for reporting this issue.
1084 *) Abort handshake if server key exchange message is omitted for ephemeral
1087 Thanks to Karthikeyan Bhargavan of the PROSECCO team at INRIA for
1088 reporting this issue.
1092 *) Remove non-export ephemeral RSA code on client and server. This code
1093 violated the TLS standard by allowing the use of temporary RSA keys in
1094 non-export ciphersuites and could be used by a server to effectively
1095 downgrade the RSA key length used to a value smaller than the server
1096 certificate. Thanks for Karthikeyan Bhargavan of the PROSECCO team at
1097 INRIA or reporting this issue.
1101 *) Fixed issue where DH client certificates are accepted without verification.
1102 An OpenSSL server will accept a DH certificate for client authentication
1103 without the certificate verify message. This effectively allows a client to
1104 authenticate without the use of a private key. This only affects servers
1105 which trust a client certificate authority which issues certificates
1106 containing DH keys: these are extremely rare and hardly ever encountered.
1107 Thanks for Karthikeyan Bhargavan of the PROSECCO team at INRIA or reporting
1112 *) Ensure that the session ID context of an SSL is updated when its
1113 SSL_CTX is updated via SSL_set_SSL_CTX.
1115 The session ID context is typically set from the parent SSL_CTX,
1116 and can vary with the CTX.
1119 *) Fix various certificate fingerprint issues.
1121 By using non-DER or invalid encodings outside the signed portion of a
1122 certificate the fingerprint can be changed without breaking the signature.
1123 Although no details of the signed portion of the certificate can be changed
1124 this can cause problems with some applications: e.g. those using the
1125 certificate fingerprint for blacklists.
1127 1. Reject signatures with non zero unused bits.
1129 If the BIT STRING containing the signature has non zero unused bits reject
1130 the signature. All current signature algorithms require zero unused bits.
1132 2. Check certificate algorithm consistency.
1134 Check the AlgorithmIdentifier inside TBS matches the one in the
1135 certificate signature. NB: this will result in signature failure
1136 errors for some broken certificates.
1138 Thanks to Konrad Kraszewski from Google for reporting this issue.
1140 3. Check DSA/ECDSA signatures use DER.
1142 Reencode DSA/ECDSA signatures and compare with the original received
1143 signature. Return an error if there is a mismatch.
1145 This will reject various cases including garbage after signature
1146 (thanks to Antti Karjalainen and Tuomo Untinen from the Codenomicon CROSS
1147 program for discovering this case) and use of BER or invalid ASN.1 INTEGERs
1148 (negative or with leading zeroes).
1150 Further analysis was conducted and fixes were developed by Stephen Henson
1151 of the OpenSSL core team.
1156 *) Correct Bignum squaring. Bignum squaring (BN_sqr) may produce incorrect
1157 results on some platforms, including x86_64. This bug occurs at random
1158 with a very low probability, and is not known to be exploitable in any
1159 way, though its exact impact is difficult to determine. Thanks to Pieter
1160 Wuille (Blockstream) who reported this issue and also suggested an initial
1161 fix. Further analysis was conducted by the OpenSSL development team and
1162 Adam Langley of Google. The final fix was developed by Andy Polyakov of
1163 the OpenSSL core team.
1167 *) Do not resume sessions on the server if the negotiated protocol
1168 version does not match the session's version. Resuming with a different
1169 version, while not strictly forbidden by the RFC, is of questionable
1170 sanity and breaks all known clients.
1171 [David Benjamin, Emilia Käsper]
1173 *) Tighten handling of the ChangeCipherSpec (CCS) message: reject
1174 early CCS messages during renegotiation. (Note that because
1175 renegotiation is encrypted, this early CCS was not exploitable.)
1178 *) Tighten client-side session ticket handling during renegotiation:
1179 ensure that the client only accepts a session ticket if the server sends
1180 the extension anew in the ServerHello. Previously, a TLS client would
1181 reuse the old extension state and thus accept a session ticket if one was
1182 announced in the initial ServerHello.
1184 Similarly, ensure that the client requires a session ticket if one
1185 was advertised in the ServerHello. Previously, a TLS client would
1186 ignore a missing NewSessionTicket message.
1189 Changes between 1.0.1i and 1.0.1j [15 Oct 2014]
1191 *) SRTP Memory Leak.
1193 A flaw in the DTLS SRTP extension parsing code allows an attacker, who
1194 sends a carefully crafted handshake message, to cause OpenSSL to fail
1195 to free up to 64k of memory causing a memory leak. This could be
1196 exploited in a Denial Of Service attack. This issue affects OpenSSL
1197 1.0.1 server implementations for both SSL/TLS and DTLS regardless of
1198 whether SRTP is used or configured. Implementations of OpenSSL that
1199 have been compiled with OPENSSL_NO_SRTP defined are not affected.
1201 The fix was developed by the OpenSSL team.
1205 *) Session Ticket Memory Leak.
1207 When an OpenSSL SSL/TLS/DTLS server receives a session ticket the
1208 integrity of that ticket is first verified. In the event of a session
1209 ticket integrity check failing, OpenSSL will fail to free memory
1210 causing a memory leak. By sending a large number of invalid session
1211 tickets an attacker could exploit this issue in a Denial Of Service
1216 *) Build option no-ssl3 is incomplete.
1218 When OpenSSL is configured with "no-ssl3" as a build option, servers
1219 could accept and complete a SSL 3.0 handshake, and clients could be
1220 configured to send them.
1222 [Akamai and the OpenSSL team]
1224 *) Add support for TLS_FALLBACK_SCSV.
1225 Client applications doing fallback retries should call
1226 SSL_set_mode(s, SSL_MODE_SEND_FALLBACK_SCSV).
1228 [Adam Langley, Bodo Moeller]
1230 *) Add additional DigestInfo checks.
1232 Reencode DigestInto in DER and check against the original when
1233 verifying RSA signature: this will reject any improperly encoded
1234 DigestInfo structures.
1236 Note: this is a precautionary measure and no attacks are currently known.
1240 Changes between 1.0.1h and 1.0.1i [6 Aug 2014]
1242 *) Fix SRP buffer overrun vulnerability. Invalid parameters passed to the
1243 SRP code can be overrun an internal buffer. Add sanity check that
1244 g, A, B < N to SRP code.
1246 Thanks to Sean Devlin and Watson Ladd of Cryptography Services, NCC
1247 Group for discovering this issue.
1251 *) A flaw in the OpenSSL SSL/TLS server code causes the server to negotiate
1252 TLS 1.0 instead of higher protocol versions when the ClientHello message
1253 is badly fragmented. This allows a man-in-the-middle attacker to force a
1254 downgrade to TLS 1.0 even if both the server and the client support a
1255 higher protocol version, by modifying the client's TLS records.
1257 Thanks to David Benjamin and Adam Langley (Google) for discovering and
1258 researching this issue.
1262 *) OpenSSL DTLS clients enabling anonymous (EC)DH ciphersuites are subject
1263 to a denial of service attack. A malicious server can crash the client
1264 with a null pointer dereference (read) by specifying an anonymous (EC)DH
1265 ciphersuite and sending carefully crafted handshake messages.
1267 Thanks to Felix Gröbert (Google) for discovering and researching this
1272 *) By sending carefully crafted DTLS packets an attacker could cause openssl
1273 to leak memory. This can be exploited through a Denial of Service attack.
1274 Thanks to Adam Langley for discovering and researching this issue.
1278 *) An attacker can force openssl to consume large amounts of memory whilst
1279 processing DTLS handshake messages. This can be exploited through a
1280 Denial of Service attack.
1281 Thanks to Adam Langley for discovering and researching this issue.
1285 *) An attacker can force an error condition which causes openssl to crash
1286 whilst processing DTLS packets due to memory being freed twice. This
1287 can be exploited through a Denial of Service attack.
1288 Thanks to Adam Langley and Wan-Teh Chang for discovering and researching
1293 *) If a multithreaded client connects to a malicious server using a resumed
1294 session and the server sends an ec point format extension it could write
1295 up to 255 bytes to freed memory.
1297 Thanks to Gabor Tyukasz (LogMeIn Inc) for discovering and researching this
1302 *) A malicious server can crash an OpenSSL client with a null pointer
1303 dereference (read) by specifying an SRP ciphersuite even though it was not
1304 properly negotiated with the client. This can be exploited through a
1305 Denial of Service attack.
1307 Thanks to Joonas Kuorilehto and Riku Hietamäki (Codenomicon) for
1308 discovering and researching this issue.
1312 *) A flaw in OBJ_obj2txt may cause pretty printing functions such as
1313 X509_name_oneline, X509_name_print_ex et al. to leak some information
1314 from the stack. Applications may be affected if they echo pretty printing
1315 output to the attacker.
1317 Thanks to Ivan Fratric (Google) for discovering this issue.
1319 [Emilia Käsper, and Steve Henson]
1321 *) Fix ec_GFp_simple_points_make_affine (thus, EC_POINTs_mul etc.)
1322 for corner cases. (Certain input points at infinity could lead to
1323 bogus results, with non-infinity inputs mapped to infinity too.)
1326 Changes between 1.0.1g and 1.0.1h [5 Jun 2014]
1328 *) Fix for SSL/TLS MITM flaw. An attacker using a carefully crafted
1329 handshake can force the use of weak keying material in OpenSSL
1330 SSL/TLS clients and servers.
1332 Thanks to KIKUCHI Masashi (Lepidum Co. Ltd.) for discovering and
1333 researching this issue. (CVE-2014-0224)
1334 [KIKUCHI Masashi, Steve Henson]
1336 *) Fix DTLS recursion flaw. By sending an invalid DTLS handshake to an
1337 OpenSSL DTLS client the code can be made to recurse eventually crashing
1340 Thanks to Imre Rad (Search-Lab Ltd.) for discovering this issue.
1342 [Imre Rad, Steve Henson]
1344 *) Fix DTLS invalid fragment vulnerability. A buffer overrun attack can
1345 be triggered by sending invalid DTLS fragments to an OpenSSL DTLS
1346 client or server. This is potentially exploitable to run arbitrary
1347 code on a vulnerable client or server.
1349 Thanks to Jüri Aedla for reporting this issue. (CVE-2014-0195)
1350 [Jüri Aedla, Steve Henson]
1352 *) Fix bug in TLS code where clients enable anonymous ECDH ciphersuites
1353 are subject to a denial of service attack.
1355 Thanks to Felix Gröbert and Ivan Fratric at Google for discovering
1356 this issue. (CVE-2014-3470)
1357 [Felix Gröbert, Ivan Fratric, Steve Henson]
1359 *) Harmonize version and its documentation. -f flag is used to display
1361 [mancha <mancha1@zoho.com>]
1363 *) Fix eckey_priv_encode so it immediately returns an error upon a failure
1364 in i2d_ECPrivateKey. Thanks to Ted Unangst for feedback on this issue.
1365 [mancha <mancha1@zoho.com>]
1367 *) Fix some double frees. These are not thought to be exploitable.
1368 [mancha <mancha1@zoho.com>]
1370 Changes between 1.0.1f and 1.0.1g [7 Apr 2014]
1372 *) A missing bounds check in the handling of the TLS heartbeat extension
1373 can be used to reveal up to 64k of memory to a connected client or
1376 Thanks for Neel Mehta of Google Security for discovering this bug and to
1377 Adam Langley <agl@chromium.org> and Bodo Moeller <bmoeller@acm.org> for
1378 preparing the fix (CVE-2014-0160)
1379 [Adam Langley, Bodo Moeller]
1381 *) Fix for the attack described in the paper "Recovering OpenSSL
1382 ECDSA Nonces Using the FLUSH+RELOAD Cache Side-channel Attack"
1383 by Yuval Yarom and Naomi Benger. Details can be obtained from:
1384 http://eprint.iacr.org/2014/140
1386 Thanks to Yuval Yarom and Naomi Benger for discovering this
1387 flaw and to Yuval Yarom for supplying a fix (CVE-2014-0076)
1388 [Yuval Yarom and Naomi Benger]
1390 *) TLS pad extension: draft-agl-tls-padding-03
1392 Workaround for the "TLS hang bug" (see FAQ and PR#2771): if the
1393 TLS client Hello record length value would otherwise be > 255 and
1394 less that 512 pad with a dummy extension containing zeroes so it
1395 is at least 512 bytes long.
1397 [Adam Langley, Steve Henson]
1399 Changes between 1.0.1e and 1.0.1f [6 Jan 2014]
1401 *) Fix for TLS record tampering bug. A carefully crafted invalid
1402 handshake could crash OpenSSL with a NULL pointer exception.
1403 Thanks to Anton Johansson for reporting this issues.
1406 *) Keep original DTLS digest and encryption contexts in retransmission
1407 structures so we can use the previous session parameters if they need
1408 to be resent. (CVE-2013-6450)
1411 *) Add option SSL_OP_SAFARI_ECDHE_ECDSA_BUG (part of SSL_OP_ALL) which
1412 avoids preferring ECDHE-ECDSA ciphers when the client appears to be
1413 Safari on OS X. Safari on OS X 10.8..10.8.3 advertises support for
1414 several ECDHE-ECDSA ciphers, but fails to negotiate them. The bug
1415 is fixed in OS X 10.8.4, but Apple have ruled out both hot fixing
1416 10.8..10.8.3 and forcing users to upgrade to 10.8.4 or newer.
1417 [Rob Stradling, Adam Langley]
1419 Changes between 1.0.1d and 1.0.1e [11 Feb 2013]
1421 *) Correct fix for CVE-2013-0169. The original didn't work on AES-NI
1422 supporting platforms or when small records were transferred.
1423 [Andy Polyakov, Steve Henson]
1425 Changes between 1.0.1c and 1.0.1d [5 Feb 2013]
1427 *) Make the decoding of SSLv3, TLS and DTLS CBC records constant time.
1429 This addresses the flaw in CBC record processing discovered by
1430 Nadhem Alfardan and Kenny Paterson. Details of this attack can be found
1431 at: http://www.isg.rhul.ac.uk/tls/
1433 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
1434 Security Group at Royal Holloway, University of London
1435 (www.isg.rhul.ac.uk) for discovering this flaw and Adam Langley and
1436 Emilia Käsper for the initial patch.
1438 [Emilia Käsper, Adam Langley, Ben Laurie, Andy Polyakov, Steve Henson]
1440 *) Fix flaw in AESNI handling of TLS 1.2 and 1.1 records for CBC mode
1441 ciphersuites which can be exploited in a denial of service attack.
1442 Thanks go to and to Adam Langley <agl@chromium.org> for discovering
1443 and detecting this bug and to Wolfgang Ettlinger
1444 <wolfgang.ettlinger@gmail.com> for independently discovering this issue.
1448 *) Return an error when checking OCSP signatures when key is NULL.
1449 This fixes a DoS attack. (CVE-2013-0166)
1452 *) Make openssl verify return errors.
1453 [Chris Palmer <palmer@google.com> and Ben Laurie]
1455 *) Call OCSP Stapling callback after ciphersuite has been chosen, so
1456 the right response is stapled. Also change SSL_get_certificate()
1457 so it returns the certificate actually sent.
1458 See http://rt.openssl.org/Ticket/Display.html?id=2836.
1459 [Rob Stradling <rob.stradling@comodo.com>]
1461 *) Fix possible deadlock when decoding public keys.
1464 *) Don't use TLS 1.0 record version number in initial client hello
1468 Changes between 1.0.1b and 1.0.1c [10 May 2012]
1470 *) Sanity check record length before skipping explicit IV in TLS
1471 1.2, 1.1 and DTLS to fix DoS attack.
1473 Thanks to Codenomicon for discovering this issue using Fuzz-o-Matic
1474 fuzzing as a service testing platform.
1478 *) Initialise tkeylen properly when encrypting CMS messages.
1479 Thanks to Solar Designer of Openwall for reporting this issue.
1482 *) In FIPS mode don't try to use composite ciphers as they are not
1486 Changes between 1.0.1a and 1.0.1b [26 Apr 2012]
1488 *) OpenSSL 1.0.0 sets SSL_OP_ALL to 0x80000FFFL and OpenSSL 1.0.1 and
1489 1.0.1a set SSL_OP_NO_TLSv1_1 to 0x00000400L which would unfortunately
1490 mean any application compiled against OpenSSL 1.0.0 headers setting
1491 SSL_OP_ALL would also set SSL_OP_NO_TLSv1_1, unintentionally disablng
1492 TLS 1.1 also. Fix this by changing the value of SSL_OP_NO_TLSv1_1 to
1493 0x10000000L Any application which was previously compiled against
1494 OpenSSL 1.0.1 or 1.0.1a headers and which cares about SSL_OP_NO_TLSv1_1
1495 will need to be recompiled as a result. Letting be results in
1496 inability to disable specifically TLS 1.1 and in client context,
1497 in unlike event, limit maximum offered version to TLS 1.0 [see below].
1500 *) In order to ensure interoperabilty SSL_OP_NO_protocolX does not
1501 disable just protocol X, but all protocols above X *if* there are
1502 protocols *below* X still enabled. In more practical terms it means
1503 that if application wants to disable TLS1.0 in favor of TLS1.1 and
1504 above, it's not sufficient to pass SSL_OP_NO_TLSv1, one has to pass
1505 SSL_OP_NO_TLSv1|SSL_OP_NO_SSLv3|SSL_OP_NO_SSLv2. This applies to
1509 Changes between 1.0.1 and 1.0.1a [19 Apr 2012]
1511 *) Check for potentially exploitable overflows in asn1_d2i_read_bio
1512 BUF_mem_grow and BUF_mem_grow_clean. Refuse attempts to shrink buffer
1513 in CRYPTO_realloc_clean.
1515 Thanks to Tavis Ormandy, Google Security Team, for discovering this
1516 issue and to Adam Langley <agl@chromium.org> for fixing it.
1518 [Adam Langley (Google), Tavis Ormandy, Google Security Team]
1520 *) Don't allow TLS 1.2 SHA-256 ciphersuites in TLS 1.0, 1.1 connections.
1523 *) Workarounds for some broken servers that "hang" if a client hello
1524 record length exceeds 255 bytes.
1526 1. Do not use record version number > TLS 1.0 in initial client
1527 hello: some (but not all) hanging servers will now work.
1528 2. If we set OPENSSL_MAX_TLS1_2_CIPHER_LENGTH this will truncate
1529 the number of ciphers sent in the client hello. This should be
1530 set to an even number, such as 50, for example by passing:
1531 -DOPENSSL_MAX_TLS1_2_CIPHER_LENGTH=50 to config or Configure.
1532 Most broken servers should now work.
1533 3. If all else fails setting OPENSSL_NO_TLS1_2_CLIENT will disable
1534 TLS 1.2 client support entirely.
1537 *) Fix SEGV in Vector Permutation AES module observed in OpenSSH.
1540 Changes between 1.0.0h and 1.0.1 [14 Mar 2012]
1542 *) Add compatibility with old MDC2 signatures which use an ASN1 OCTET
1543 STRING form instead of a DigestInfo.
1546 *) The format used for MDC2 RSA signatures is inconsistent between EVP
1547 and the RSA_sign/RSA_verify functions. This was made more apparent when
1548 OpenSSL used RSA_sign/RSA_verify for some RSA signatures in particular
1549 those which went through EVP_PKEY_METHOD in 1.0.0 and later. Detect
1550 the correct format in RSA_verify so both forms transparently work.
1553 *) Some servers which support TLS 1.0 can choke if we initially indicate
1554 support for TLS 1.2 and later renegotiate using TLS 1.0 in the RSA
1555 encrypted premaster secret. As a workaround use the maximum pemitted
1556 client version in client hello, this should keep such servers happy
1557 and still work with previous versions of OpenSSL.
1560 *) Add support for TLS/DTLS heartbeats.
1561 [Robin Seggelmann <seggelmann@fh-muenster.de>]
1563 *) Add support for SCTP.
1564 [Robin Seggelmann <seggelmann@fh-muenster.de>]
1566 *) Improved PRNG seeding for VOS.
1567 [Paul Green <Paul.Green@stratus.com>]
1569 *) Extensive assembler packs updates, most notably:
1571 - x86[_64]: AES-NI, PCLMULQDQ, RDRAND support;
1572 - x86[_64]: SSSE3 support (SHA1, vector-permutation AES);
1573 - x86_64: bit-sliced AES implementation;
1574 - ARM: NEON support, contemporary platforms optimizations;
1575 - s390x: z196 support;
1576 - *: GHASH and GF(2^m) multiplication implementations;
1580 *) Make TLS-SRP code conformant with RFC 5054 API cleanup
1581 (removal of unnecessary code)
1582 [Peter Sylvester <peter.sylvester@edelweb.fr>]
1584 *) Add TLS key material exporter from RFC 5705.
1587 *) Add DTLS-SRTP negotiation from RFC 5764.
1590 *) Add Next Protocol Negotiation,
1591 http://tools.ietf.org/html/draft-agl-tls-nextprotoneg-00. Can be
1592 disabled with a no-npn flag to config or Configure. Code donated
1594 [Adam Langley <agl@google.com> and Ben Laurie]
1596 *) Add optional 64-bit optimized implementations of elliptic curves NIST-P224,
1597 NIST-P256, NIST-P521, with constant-time single point multiplication on
1598 typical inputs. Compiler support for the nonstandard type __uint128_t is
1599 required to use this (present in gcc 4.4 and later, for 64-bit builds).
1600 Code made available under Apache License version 2.0.
1602 Specify "enable-ec_nistp_64_gcc_128" on the Configure (or config) command
1603 line to include this in your build of OpenSSL, and run "make depend" (or
1604 "make update"). This enables the following EC_METHODs:
1606 EC_GFp_nistp224_method()
1607 EC_GFp_nistp256_method()
1608 EC_GFp_nistp521_method()
1610 EC_GROUP_new_by_curve_name() will automatically use these (while
1611 EC_GROUP_new_curve_GFp() currently prefers the more flexible
1613 [Emilia Käsper, Adam Langley, Bodo Moeller (Google)]
1615 *) Use type ossl_ssize_t instad of ssize_t which isn't available on
1616 all platforms. Move ssize_t definition from e_os.h to the public
1617 header file e_os2.h as it now appears in public header file cms.h
1620 *) New -sigopt option to the ca, req and x509 utilities. Additional
1621 signature parameters can be passed using this option and in
1625 *) Add RSA PSS signing function. This will generate and set the
1626 appropriate AlgorithmIdentifiers for PSS based on those in the
1627 corresponding EVP_MD_CTX structure. No application support yet.
1630 *) Support for companion algorithm specific ASN1 signing routines.
1631 New function ASN1_item_sign_ctx() signs a pre-initialised
1632 EVP_MD_CTX structure and sets AlgorithmIdentifiers based on
1633 the appropriate parameters.
1636 *) Add new algorithm specific ASN1 verification initialisation function
1637 to EVP_PKEY_ASN1_METHOD: this is not in EVP_PKEY_METHOD since the ASN1
1638 handling will be the same no matter what EVP_PKEY_METHOD is used.
1639 Add a PSS handler to support verification of PSS signatures: checked
1640 against a number of sample certificates.
1643 *) Add signature printing for PSS. Add PSS OIDs.
1644 [Steve Henson, Martin Kaiser <lists@kaiser.cx>]
1646 *) Add algorithm specific signature printing. An individual ASN1 method
1647 can now print out signatures instead of the standard hex dump.
1649 More complex signatures (e.g. PSS) can print out more meaningful
1650 information. Include DSA version that prints out the signature
1654 *) Password based recipient info support for CMS library: implementing
1658 *) Split password based encryption into PBES2 and PBKDF2 functions. This
1659 neatly separates the code into cipher and PBE sections and is required
1660 for some algorithms that split PBES2 into separate pieces (such as
1661 password based CMS).
1664 *) Session-handling fixes:
1665 - Fix handling of connections that are resuming with a session ID,
1666 but also support Session Tickets.
1667 - Fix a bug that suppressed issuing of a new ticket if the client
1668 presented a ticket with an expired session.
1669 - Try to set the ticket lifetime hint to something reasonable.
1670 - Make tickets shorter by excluding irrelevant information.
1671 - On the client side, don't ignore renewed tickets.
1672 [Adam Langley, Bodo Moeller (Google)]
1674 *) Fix PSK session representation.
1677 *) Add RC4-MD5 and AESNI-SHA1 "stitched" implementations.
1679 This work was sponsored by Intel.
1682 *) Add GCM support to TLS library. Some custom code is needed to split
1683 the IV between the fixed (from PRF) and explicit (from TLS record)
1684 portions. This adds all GCM ciphersuites supported by RFC5288 and
1685 RFC5289. Generalise some AES* cipherstrings to inlclude GCM and
1686 add a special AESGCM string for GCM only.
1689 *) Expand range of ctrls for AES GCM. Permit setting invocation
1690 field on decrypt and retrieval of invocation field only on encrypt.
1693 *) Add HMAC ECC ciphersuites from RFC5289. Include SHA384 PRF support.
1694 As required by RFC5289 these ciphersuites cannot be used if for
1695 versions of TLS earlier than 1.2.
1698 *) For FIPS capable OpenSSL interpret a NULL default public key method
1699 as unset and return the appopriate default but do *not* set the default.
1700 This means we can return the appopriate method in applications that
1701 swicth between FIPS and non-FIPS modes.
1704 *) Redirect HMAC and CMAC operations to FIPS module in FIPS mode. If an
1705 ENGINE is used then we cannot handle that in the FIPS module so we
1706 keep original code iff non-FIPS operations are allowed.
1709 *) Add -attime option to openssl utilities.
1710 [Peter Eckersley <pde@eff.org>, Ben Laurie and Steve Henson]
1712 *) Redirect DSA and DH operations to FIPS module in FIPS mode.
1715 *) Redirect ECDSA and ECDH operations to FIPS module in FIPS mode. Also use
1716 FIPS EC methods unconditionally for now.
1719 *) New build option no-ec2m to disable characteristic 2 code.
1722 *) Backport libcrypto audit of return value checking from 1.1.0-dev; not
1723 all cases can be covered as some introduce binary incompatibilities.
1726 *) Redirect RSA operations to FIPS module including keygen,
1727 encrypt, decrypt, sign and verify. Block use of non FIPS RSA methods.
1730 *) Add similar low level API blocking to ciphers.
1733 *) Low level digest APIs are not approved in FIPS mode: any attempt
1734 to use these will cause a fatal error. Applications that *really* want
1735 to use them can use the private_* version instead.
1738 *) Redirect cipher operations to FIPS module for FIPS builds.
1741 *) Redirect digest operations to FIPS module for FIPS builds.
1744 *) Update build system to add "fips" flag which will link in fipscanister.o
1745 for static and shared library builds embedding a signature if needed.
1748 *) Output TLS supported curves in preference order instead of numerical
1749 order. This is currently hardcoded for the highest order curves first.
1750 This should be configurable so applications can judge speed vs strength.
1753 *) Add TLS v1.2 server support for client authentication.
1756 *) Add support for FIPS mode in ssl library: disable SSLv3, non-FIPS ciphers
1760 *) Functions FIPS_mode_set() and FIPS_mode() which call the underlying
1761 FIPS modules versions.
1764 *) Add TLS v1.2 client side support for client authentication. Keep cache
1765 of handshake records longer as we don't know the hash algorithm to use
1766 until after the certificate request message is received.
1769 *) Initial TLS v1.2 client support. Add a default signature algorithms
1770 extension including all the algorithms we support. Parse new signature
1771 format in client key exchange. Relax some ECC signing restrictions for
1772 TLS v1.2 as indicated in RFC5246.
1775 *) Add server support for TLS v1.2 signature algorithms extension. Switch
1776 to new signature format when needed using client digest preference.
1777 All server ciphersuites should now work correctly in TLS v1.2. No client
1778 support yet and no support for client certificates.
1781 *) Initial TLS v1.2 support. Add new SHA256 digest to ssl code, switch
1782 to SHA256 for PRF when using TLS v1.2 and later. Add new SHA256 based
1783 ciphersuites. At present only RSA key exchange ciphersuites work with
1784 TLS v1.2. Add new option for TLS v1.2 replacing the old and obsolete
1785 SSL_OP_PKCS1_CHECK flags with SSL_OP_NO_TLSv1_2. New TLSv1.2 methods
1786 and version checking.
1789 *) New option OPENSSL_NO_SSL_INTERN. If an application can be compiled
1790 with this defined it will not be affected by any changes to ssl internal
1791 structures. Add several utility functions to allow openssl application
1792 to work with OPENSSL_NO_SSL_INTERN defined.
1796 [Tom Wu <tjw@cs.stanford.edu> and Ben Laurie]
1798 *) Add functions to copy EVP_PKEY_METHOD and retrieve flags and id.
1801 *) Permit abbreviated handshakes when renegotiating using the function
1802 SSL_renegotiate_abbreviated().
1803 [Robin Seggelmann <seggelmann@fh-muenster.de>]
1805 *) Add call to ENGINE_register_all_complete() to
1806 ENGINE_load_builtin_engines(), so some implementations get used
1807 automatically instead of needing explicit application support.
1810 *) Add support for TLS key exporter as described in RFC5705.
1811 [Robin Seggelmann <seggelmann@fh-muenster.de>, Steve Henson]
1813 *) Initial TLSv1.1 support. Since TLSv1.1 is very similar to TLS v1.0 only
1814 a few changes are required:
1816 Add SSL_OP_NO_TLSv1_1 flag.
1817 Add TLSv1_1 methods.
1818 Update version checking logic to handle version 1.1.
1819 Add explicit IV handling (ported from DTLS code).
1820 Add command line options to s_client/s_server.
1823 Changes between 1.0.0g and 1.0.0h [12 Mar 2012]
1825 *) Fix MMA (Bleichenbacher's attack on PKCS #1 v1.5 RSA padding) weakness
1826 in CMS and PKCS7 code. When RSA decryption fails use a random key for
1827 content decryption and always return the same error. Note: this attack
1828 needs on average 2^20 messages so it only affects automated senders. The
1829 old behaviour can be reenabled in the CMS code by setting the
1830 CMS_DEBUG_DECRYPT flag: this is useful for debugging and testing where
1831 an MMA defence is not necessary.
1832 Thanks to Ivan Nestlerode <inestlerode@us.ibm.com> for discovering
1833 this issue. (CVE-2012-0884)
1836 *) Fix CVE-2011-4619: make sure we really are receiving a
1837 client hello before rejecting multiple SGC restarts. Thanks to
1838 Ivan Nestlerode <inestlerode@us.ibm.com> for discovering this bug.
1841 Changes between 1.0.0f and 1.0.0g [18 Jan 2012]
1843 *) Fix for DTLS DoS issue introduced by fix for CVE-2011-4109.
1844 Thanks to Antonio Martin, Enterprise Secure Access Research and
1845 Development, Cisco Systems, Inc. for discovering this bug and
1846 preparing a fix. (CVE-2012-0050)
1849 Changes between 1.0.0e and 1.0.0f [4 Jan 2012]
1851 *) Nadhem Alfardan and Kenny Paterson have discovered an extension
1852 of the Vaudenay padding oracle attack on CBC mode encryption
1853 which enables an efficient plaintext recovery attack against
1854 the OpenSSL implementation of DTLS. Their attack exploits timing
1855 differences arising during decryption processing. A research
1856 paper describing this attack can be found at:
1857 http://www.isg.rhul.ac.uk/~kp/dtls.pdf
1858 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
1859 Security Group at Royal Holloway, University of London
1860 (www.isg.rhul.ac.uk) for discovering this flaw and to Robin Seggelmann
1861 <seggelmann@fh-muenster.de> and Michael Tuexen <tuexen@fh-muenster.de>
1862 for preparing the fix. (CVE-2011-4108)
1863 [Robin Seggelmann, Michael Tuexen]
1865 *) Clear bytes used for block padding of SSL 3.0 records.
1867 [Adam Langley (Google)]
1869 *) Only allow one SGC handshake restart for SSL/TLS. Thanks to George
1870 Kadianakis <desnacked@gmail.com> for discovering this issue and
1871 Adam Langley for preparing the fix. (CVE-2011-4619)
1872 [Adam Langley (Google)]
1874 *) Check parameters are not NULL in GOST ENGINE. (CVE-2012-0027)
1875 [Andrey Kulikov <amdeich@gmail.com>]
1877 *) Prevent malformed RFC3779 data triggering an assertion failure.
1878 Thanks to Andrew Chi, BBN Technologies, for discovering the flaw
1879 and Rob Austein <sra@hactrn.net> for fixing it. (CVE-2011-4577)
1880 [Rob Austein <sra@hactrn.net>]
1882 *) Improved PRNG seeding for VOS.
1883 [Paul Green <Paul.Green@stratus.com>]
1885 *) Fix ssl_ciph.c set-up race.
1886 [Adam Langley (Google)]
1888 *) Fix spurious failures in ecdsatest.c.
1889 [Emilia Käsper (Google)]
1891 *) Fix the BIO_f_buffer() implementation (which was mixing different
1892 interpretations of the '..._len' fields).
1893 [Adam Langley (Google)]
1895 *) Fix handling of BN_BLINDING: now BN_BLINDING_invert_ex (rather than
1896 BN_BLINDING_invert_ex) calls BN_BLINDING_update, ensuring that concurrent
1897 threads won't reuse the same blinding coefficients.
1899 This also avoids the need to obtain the CRYPTO_LOCK_RSA_BLINDING
1900 lock to call BN_BLINDING_invert_ex, and avoids one use of
1901 BN_BLINDING_update for each BN_BLINDING structure (previously,
1902 the last update always remained unused).
1903 [Emilia Käsper (Google)]
1905 *) In ssl3_clear, preserve s3->init_extra along with s3->rbuf.
1906 [Bob Buckholz (Google)]
1908 Changes between 1.0.0d and 1.0.0e [6 Sep 2011]
1910 *) Fix bug where CRLs with nextUpdate in the past are sometimes accepted
1911 by initialising X509_STORE_CTX properly. (CVE-2011-3207)
1912 [Kaspar Brand <ossl@velox.ch>]
1914 *) Fix SSL memory handling for (EC)DH ciphersuites, in particular
1915 for multi-threaded use of ECDH. (CVE-2011-3210)
1916 [Adam Langley (Google)]
1918 *) Fix x509_name_ex_d2i memory leak on bad inputs.
1921 *) Remove hard coded ecdsaWithSHA1 signature tests in ssl code and check
1922 signature public key algorithm by using OID xref utilities instead.
1923 Before this you could only use some ECC ciphersuites with SHA1 only.
1926 *) Add protection against ECDSA timing attacks as mentioned in the paper
1927 by Billy Bob Brumley and Nicola Tuveri, see:
1929 http://eprint.iacr.org/2011/232.pdf
1931 [Billy Bob Brumley and Nicola Tuveri]
1933 Changes between 1.0.0c and 1.0.0d [8 Feb 2011]
1935 *) Fix parsing of OCSP stapling ClientHello extension. CVE-2011-0014
1936 [Neel Mehta, Adam Langley, Bodo Moeller (Google)]
1938 *) Fix bug in string printing code: if *any* escaping is enabled we must
1939 escape the escape character (backslash) or the resulting string is
1943 Changes between 1.0.0b and 1.0.0c [2 Dec 2010]
1945 *) Disable code workaround for ancient and obsolete Netscape browsers
1946 and servers: an attacker can use it in a ciphersuite downgrade attack.
1947 Thanks to Martin Rex for discovering this bug. CVE-2010-4180
1950 *) Fixed J-PAKE implementation error, originally discovered by
1951 Sebastien Martini, further info and confirmation from Stefan
1952 Arentz and Feng Hao. Note that this fix is a security fix. CVE-2010-4252
1955 Changes between 1.0.0a and 1.0.0b [16 Nov 2010]
1957 *) Fix extension code to avoid race conditions which can result in a buffer
1958 overrun vulnerability: resumed sessions must not be modified as they can
1959 be shared by multiple threads. CVE-2010-3864
1962 *) Fix WIN32 build system to correctly link an ENGINE directory into
1966 Changes between 1.0.0 and 1.0.0a [01 Jun 2010]
1968 *) Check return value of int_rsa_verify in pkey_rsa_verifyrecover
1970 [Steve Henson, Peter-Michael Hager <hager@dortmund.net>]
1972 Changes between 0.9.8n and 1.0.0 [29 Mar 2010]
1974 *) Add "missing" function EVP_CIPHER_CTX_copy(). This copies a cipher
1975 context. The operation can be customised via the ctrl mechanism in
1976 case ENGINEs want to include additional functionality.
1979 *) Tolerate yet another broken PKCS#8 key format: private key value negative.
1982 *) Add new -subject_hash_old and -issuer_hash_old options to x509 utility to
1983 output hashes compatible with older versions of OpenSSL.
1984 [Willy Weisz <weisz@vcpc.univie.ac.at>]
1986 *) Fix compression algorithm handling: if resuming a session use the
1987 compression algorithm of the resumed session instead of determining
1988 it from client hello again. Don't allow server to change algorithm.
1991 *) Add load_crls() function to apps tidying load_certs() too. Add option
1992 to verify utility to allow additional CRLs to be included.
1995 *) Update OCSP request code to permit adding custom headers to the request:
1996 some responders need this.
1999 *) The function EVP_PKEY_sign() returns <=0 on error: check return code
2001 [Julia Lawall <julia@diku.dk>]
2003 *) Update verify callback code in apps/s_cb.c and apps/verify.c, it
2004 needlessly dereferenced structures, used obsolete functions and
2005 didn't handle all updated verify codes correctly.
2008 *) Disable MD2 in the default configuration.
2011 *) In BIO_pop() and BIO_push() use the ctrl argument (which was NULL) to
2012 indicate the initial BIO being pushed or popped. This makes it possible
2013 to determine whether the BIO is the one explicitly called or as a result
2014 of the ctrl being passed down the chain. Fix BIO_pop() and SSL BIOs so
2015 it handles reference counts correctly and doesn't zero out the I/O bio
2016 when it is not being explicitly popped. WARNING: applications which
2017 included workarounds for the old buggy behaviour will need to be modified
2018 or they could free up already freed BIOs.
2021 *) Extend the uni2asc/asc2uni => OPENSSL_uni2asc/OPENSSL_asc2uni
2022 renaming to all platforms (within the 0.9.8 branch, this was
2023 done conditionally on Netware platforms to avoid a name clash).
2024 [Guenter <lists@gknw.net>]
2026 *) Add ECDHE and PSK support to DTLS.
2027 [Michael Tuexen <tuexen@fh-muenster.de>]
2029 *) Add CHECKED_STACK_OF macro to safestack.h, otherwise safestack can't
2033 *) Add "missing" function EVP_MD_flags() (without this the only way to
2034 retrieve a digest flags is by accessing the structure directly. Update
2035 EVP_MD_do_all*() and EVP_CIPHER_do_all*() to include the name a digest
2036 or cipher is registered as in the "from" argument. Print out all
2037 registered digests in the dgst usage message instead of manually
2038 attempting to work them out.
2041 *) If no SSLv2 ciphers are used don't use an SSLv2 compatible client hello:
2042 this allows the use of compression and extensions. Change default cipher
2043 string to remove SSLv2 ciphersuites. This effectively avoids ancient SSLv2
2044 by default unless an application cipher string requests it.
2047 *) Alter match criteria in PKCS12_parse(). It used to try to use local
2048 key ids to find matching certificates and keys but some PKCS#12 files
2049 don't follow the (somewhat unwritten) rules and this strategy fails.
2050 Now just gather all certificates together and the first private key
2051 then look for the first certificate that matches the key.
2054 *) Support use of registered digest and cipher names for dgst and cipher
2055 commands instead of having to add each one as a special case. So now
2062 openssl dgst -sha256 foo
2064 and this works for ENGINE based algorithms too.
2068 *) Update Gost ENGINE to support parameter files.
2069 [Victor B. Wagner <vitus@cryptocom.ru>]
2071 *) Support GeneralizedTime in ca utility.
2072 [Oliver Martin <oliver@volatilevoid.net>, Steve Henson]
2074 *) Enhance the hash format used for certificate directory links. The new
2075 form uses the canonical encoding (meaning equivalent names will work
2076 even if they aren't identical) and uses SHA1 instead of MD5. This form
2077 is incompatible with the older format and as a result c_rehash should
2078 be used to rebuild symbolic links.
2081 *) Make PKCS#8 the default write format for private keys, replacing the
2082 traditional format. This form is standardised, more secure and doesn't
2083 include an implicit MD5 dependency.
2086 *) Add a $gcc_devteam_warn option to Configure. The idea is that any code
2087 committed to OpenSSL should pass this lot as a minimum.
2090 *) Add session ticket override functionality for use by EAP-FAST.
2091 [Jouni Malinen <j@w1.fi>]
2093 *) Modify HMAC functions to return a value. Since these can be implemented
2094 in an ENGINE errors can occur.
2097 *) Type-checked OBJ_bsearch_ex.
2100 *) Type-checked OBJ_bsearch. Also some constification necessitated
2101 by type-checking. Still to come: TXT_DB, bsearch(?),
2102 OBJ_bsearch_ex, qsort, CRYPTO_EX_DATA, ASN1_VALUE, ASN1_STRING,
2106 *) New function OPENSSL_gmtime_adj() to add a specific number of days and
2107 seconds to a tm structure directly, instead of going through OS
2108 specific date routines. This avoids any issues with OS routines such
2109 as the year 2038 bug. New *_adj() functions for ASN1 time structures
2110 and X509_time_adj_ex() to cover the extended range. The existing
2111 X509_time_adj() is still usable and will no longer have any date issues.
2114 *) Delta CRL support. New use deltas option which will attempt to locate
2115 and search any appropriate delta CRLs available.
2117 This work was sponsored by Google.
2120 *) Support for CRLs partitioned by reason code. Reorganise CRL processing
2121 code and add additional score elements. Validate alternate CRL paths
2122 as part of the CRL checking and indicate a new error "CRL path validation
2123 error" in this case. Applications wanting additional details can use
2124 the verify callback and check the new "parent" field. If this is not
2125 NULL CRL path validation is taking place. Existing applications wont
2126 see this because it requires extended CRL support which is off by
2129 This work was sponsored by Google.
2132 *) Support for freshest CRL extension.
2134 This work was sponsored by Google.
2137 *) Initial indirect CRL support. Currently only supported in the CRLs
2138 passed directly and not via lookup. Process certificate issuer
2139 CRL entry extension and lookup CRL entries by bother issuer name
2140 and serial number. Check and process CRL issuer entry in IDP extension.
2142 This work was sponsored by Google.
2145 *) Add support for distinct certificate and CRL paths. The CRL issuer
2146 certificate is validated separately in this case. Only enabled if
2147 an extended CRL support flag is set: this flag will enable additional
2148 CRL functionality in future.
2150 This work was sponsored by Google.
2153 *) Add support for policy mappings extension.
2155 This work was sponsored by Google.
2158 *) Fixes to pathlength constraint, self issued certificate handling,
2159 policy processing to align with RFC3280 and PKITS tests.
2161 This work was sponsored by Google.
2164 *) Support for name constraints certificate extension. DN, email, DNS
2165 and URI types are currently supported.
2167 This work was sponsored by Google.
2170 *) To cater for systems that provide a pointer-based thread ID rather
2171 than numeric, deprecate the current numeric thread ID mechanism and
2172 replace it with a structure and associated callback type. This
2173 mechanism allows a numeric "hash" to be extracted from a thread ID in
2174 either case, and on platforms where pointers are larger than 'long',
2175 mixing is done to help ensure the numeric 'hash' is usable even if it
2176 can't be guaranteed unique. The default mechanism is to use "&errno"
2177 as a pointer-based thread ID to distinguish between threads.
2179 Applications that want to provide their own thread IDs should now use
2180 CRYPTO_THREADID_set_callback() to register a callback that will call
2181 either CRYPTO_THREADID_set_numeric() or CRYPTO_THREADID_set_pointer().
2183 Note that ERR_remove_state() is now deprecated, because it is tied
2184 to the assumption that thread IDs are numeric. ERR_remove_state(0)
2185 to free the current thread's error state should be replaced by
2186 ERR_remove_thread_state(NULL).
2188 (This new approach replaces the functions CRYPTO_set_idptr_callback(),
2189 CRYPTO_get_idptr_callback(), and CRYPTO_thread_idptr() that existed in
2190 OpenSSL 0.9.9-dev between June 2006 and August 2008. Also, if an
2191 application was previously providing a numeric thread callback that
2192 was inappropriate for distinguishing threads, then uniqueness might
2193 have been obtained with &errno that happened immediately in the
2194 intermediate development versions of OpenSSL; this is no longer the
2195 case, the numeric thread callback will now override the automatic use
2197 [Geoff Thorpe, with help from Bodo Moeller]
2199 *) Initial support for different CRL issuing certificates. This covers a
2200 simple case where the self issued certificates in the chain exist and
2201 the real CRL issuer is higher in the existing chain.
2203 This work was sponsored by Google.
2206 *) Removed effectively defunct crypto/store from the build.
2209 *) Revamp of STACK to provide stronger type-checking. Still to come:
2210 TXT_DB, bsearch(?), OBJ_bsearch, qsort, CRYPTO_EX_DATA, ASN1_VALUE,
2211 ASN1_STRING, CONF_VALUE.
2214 *) Add a new SSL_MODE_RELEASE_BUFFERS mode flag to release unused buffer
2215 RAM on SSL connections. This option can save about 34k per idle SSL.
2218 *) Revamp of LHASH to provide stronger type-checking. Still to come:
2219 STACK, TXT_DB, bsearch, qsort.
2222 *) Initial support for Cryptographic Message Syntax (aka CMS) based
2223 on RFC3850, RFC3851 and RFC3852. New cms directory and cms utility,
2224 support for data, signedData, compressedData, digestedData and
2225 encryptedData, envelopedData types included. Scripts to check against
2226 RFC4134 examples draft and interop and consistency checks of many
2227 content types and variants.
2230 *) Add options to enc utility to support use of zlib compression BIO.
2233 *) Extend mk1mf to support importing of options and assembly language
2234 files from Configure script, currently only included in VC-WIN32.
2235 The assembly language rules can now optionally generate the source
2236 files from the associated perl scripts.
2239 *) Implement remaining functionality needed to support GOST ciphersuites.
2240 Interop testing has been performed using CryptoPro implementations.
2241 [Victor B. Wagner <vitus@cryptocom.ru>]
2243 *) s390x assembler pack.
2246 *) ARMv4 assembler pack. ARMv4 refers to v4 and later ISA, not CPU
2250 *) Implement Opaque PRF Input TLS extension as specified in
2251 draft-rescorla-tls-opaque-prf-input-00.txt. Since this is not an
2252 official specification yet and no extension type assignment by
2253 IANA exists, this extension (for now) will have to be explicitly
2254 enabled when building OpenSSL by providing the extension number
2255 to use. For example, specify an option
2257 -DTLSEXT_TYPE_opaque_prf_input=0x9527
2259 to the "config" or "Configure" script to enable the extension,
2260 assuming extension number 0x9527 (which is a completely arbitrary
2261 and unofficial assignment based on the MD5 hash of the Internet
2262 Draft). Note that by doing so, you potentially lose
2263 interoperability with other TLS implementations since these might
2264 be using the same extension number for other purposes.
2266 SSL_set_tlsext_opaque_prf_input(ssl, src, len) is used to set the
2267 opaque PRF input value to use in the handshake. This will create
2268 an interal copy of the length-'len' string at 'src', and will
2269 return non-zero for success.
2271 To get more control and flexibility, provide a callback function
2274 SSL_CTX_set_tlsext_opaque_prf_input_callback(ctx, cb)
2275 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg(ctx, arg)
2279 int (*cb)(SSL *, void *peerinput, size_t len, void *arg);
2282 Callback function 'cb' will be called in handshakes, and is
2283 expected to use SSL_set_tlsext_opaque_prf_input() as appropriate.
2284 Argument 'arg' is for application purposes (the value as given to
2285 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg() will directly
2286 be provided to the callback function). The callback function
2287 has to return non-zero to report success: usually 1 to use opaque
2288 PRF input just if possible, or 2 to enforce use of the opaque PRF
2289 input. In the latter case, the library will abort the handshake
2290 if opaque PRF input is not successfully negotiated.
2292 Arguments 'peerinput' and 'len' given to the callback function
2293 will always be NULL and 0 in the case of a client. A server will
2294 see the client's opaque PRF input through these variables if
2295 available (NULL and 0 otherwise). Note that if the server
2296 provides an opaque PRF input, the length must be the same as the
2297 length of the client's opaque PRF input.
2299 Note that the callback function will only be called when creating
2300 a new session (session resumption can resume whatever was
2301 previously negotiated), and will not be called in SSL 2.0
2302 handshakes; thus, SSL_CTX_set_options(ctx, SSL_OP_NO_SSLv2) or
2303 SSL_set_options(ssl, SSL_OP_NO_SSLv2) is especially recommended
2304 for applications that need to enforce opaque PRF input.
2308 *) Update ssl code to support digests other than SHA1+MD5 for handshake
2311 [Victor B. Wagner <vitus@cryptocom.ru>]
2313 *) Add RFC4507 support to OpenSSL. This includes the corrections in
2314 RFC4507bis. The encrypted ticket format is an encrypted encoded
2315 SSL_SESSION structure, that way new session features are automatically
2318 If a client application caches session in an SSL_SESSION structure
2319 support is transparent because tickets are now stored in the encoded
2322 The SSL_CTX structure automatically generates keys for ticket
2323 protection in servers so again support should be possible
2324 with no application modification.
2326 If a client or server wishes to disable RFC4507 support then the option
2327 SSL_OP_NO_TICKET can be set.
2329 Add a TLS extension debugging callback to allow the contents of any client
2330 or server extensions to be examined.
2332 This work was sponsored by Google.
2335 *) Final changes to avoid use of pointer pointer casts in OpenSSL.
2336 OpenSSL should now compile cleanly on gcc 4.2
2337 [Peter Hartley <pdh@utter.chaos.org.uk>, Steve Henson]
2339 *) Update SSL library to use new EVP_PKEY MAC API. Include generic MAC
2340 support including streaming MAC support: this is required for GOST
2341 ciphersuite support.
2342 [Victor B. Wagner <vitus@cryptocom.ru>, Steve Henson]
2344 *) Add option -stream to use PKCS#7 streaming in smime utility. New
2345 function i2d_PKCS7_bio_stream() and PEM_write_PKCS7_bio_stream()
2346 to output in BER and PEM format.
2349 *) Experimental support for use of HMAC via EVP_PKEY interface. This
2350 allows HMAC to be handled via the EVP_DigestSign*() interface. The
2351 EVP_PKEY "key" in this case is the HMAC key, potentially allowing
2352 ENGINE support for HMAC keys which are unextractable. New -mac and
2353 -macopt options to dgst utility.
2356 *) New option -sigopt to dgst utility. Update dgst to use
2357 EVP_Digest{Sign,Verify}*. These two changes make it possible to use
2358 alternative signing paramaters such as X9.31 or PSS in the dgst
2362 *) Change ssl_cipher_apply_rule(), the internal function that does
2363 the work each time a ciphersuite string requests enabling
2364 ("foo+bar"), moving ("+foo+bar"), disabling ("-foo+bar", or
2365 removing ("!foo+bar") a class of ciphersuites: Now it maintains
2366 the order of disabled ciphersuites such that those ciphersuites
2367 that most recently went from enabled to disabled not only stay
2368 in order with respect to each other, but also have higher priority
2369 than other disabled ciphersuites the next time ciphersuites are
2372 This means that you can now say, e.g., "PSK:-PSK:HIGH" to enable
2373 the same ciphersuites as with "HIGH" alone, but in a specific
2374 order where the PSK ciphersuites come first (since they are the
2375 most recently disabled ciphersuites when "HIGH" is parsed).
2377 Also, change ssl_create_cipher_list() (using this new
2378 funcionality) such that between otherwise identical
2379 cihpersuites, ephemeral ECDH is preferred over ephemeral DH in
2383 *) Change ssl_create_cipher_list() so that it automatically
2384 arranges the ciphersuites in reasonable order before starting
2385 to process the rule string. Thus, the definition for "DEFAULT"
2386 (SSL_DEFAULT_CIPHER_LIST) now is just "ALL:!aNULL:!eNULL", but
2387 remains equivalent to "AES:ALL:!aNULL:!eNULL:+aECDH:+kRSA:+RC4:@STRENGTH".
2388 This makes it much easier to arrive at a reasonable default order
2389 in applications for which anonymous ciphers are OK (meaning
2390 that you can't actually use DEFAULT).
2391 [Bodo Moeller; suggested by Victor Duchovni]
2393 *) Split the SSL/TLS algorithm mask (as used for ciphersuite string
2394 processing) into multiple integers instead of setting
2395 "SSL_MKEY_MASK" bits, "SSL_AUTH_MASK" bits, "SSL_ENC_MASK",
2396 "SSL_MAC_MASK", and "SSL_SSL_MASK" bits all in a single integer.
2397 (These masks as well as the individual bit definitions are hidden
2398 away into the non-exported interface ssl/ssl_locl.h, so this
2399 change to the definition of the SSL_CIPHER structure shouldn't
2400 affect applications.) This give us more bits for each of these
2401 categories, so there is no longer a need to coagulate AES128 and
2402 AES256 into a single algorithm bit, and to coagulate Camellia128
2403 and Camellia256 into a single algorithm bit, which has led to all
2406 Thus, among other things, the kludge introduced in 0.9.7m and
2407 0.9.8e for masking out AES256 independently of AES128 or masking
2408 out Camellia256 independently of AES256 is not needed here in 0.9.9.
2410 With the change, we also introduce new ciphersuite aliases that
2411 so far were missing: "AES128", "AES256", "CAMELLIA128", and
2415 *) Add support for dsa-with-SHA224 and dsa-with-SHA256.
2416 Use the leftmost N bytes of the signature input if the input is
2417 larger than the prime q (with N being the size in bytes of q).
2420 *) Very *very* experimental PKCS#7 streaming encoder support. Nothing uses
2421 it yet and it is largely untested.
2424 *) Add support for the ecdsa-with-SHA224/256/384/512 signature types.
2427 *) Initial incomplete changes to avoid need for function casts in OpenSSL
2428 some compilers (gcc 4.2 and later) reject their use. Safestack is
2429 reimplemented. Update ASN1 to avoid use of legacy functions.
2432 *) Win32/64 targets are linked with Winsock2.
2435 *) Add an X509_CRL_METHOD structure to allow CRL processing to be redirected
2436 to external functions. This can be used to increase CRL handling
2437 efficiency especially when CRLs are very large by (for example) storing
2438 the CRL revoked certificates in a database.
2441 *) Overhaul of by_dir code. Add support for dynamic loading of CRLs so
2442 new CRLs added to a directory can be used. New command line option
2443 -verify_return_error to s_client and s_server. This causes real errors
2444 to be returned by the verify callback instead of carrying on no matter
2445 what. This reflects the way a "real world" verify callback would behave.
2448 *) GOST engine, supporting several GOST algorithms and public key formats.
2449 Kindly donated by Cryptocom.
2452 *) Partial support for Issuing Distribution Point CRL extension. CRLs
2453 partitioned by DP are handled but no indirect CRL or reason partitioning
2454 (yet). Complete overhaul of CRL handling: now the most suitable CRL is
2455 selected via a scoring technique which handles IDP and AKID in CRLs.
2458 *) New X509_STORE_CTX callbacks lookup_crls() and lookup_certs() which
2459 will ultimately be used for all verify operations: this will remove the
2460 X509_STORE dependency on certificate verification and allow alternative
2461 lookup methods. X509_STORE based implementations of these two callbacks.
2464 *) Allow multiple CRLs to exist in an X509_STORE with matching issuer names.
2465 Modify get_crl() to find a valid (unexpired) CRL if possible.
2468 *) New function X509_CRL_match() to check if two CRLs are identical. Normally
2469 this would be called X509_CRL_cmp() but that name is already used by
2470 a function that just compares CRL issuer names. Cache several CRL
2471 extensions in X509_CRL structure and cache CRLDP in X509.
2474 *) Store a "canonical" representation of X509_NAME structure (ASN1 Name)
2475 this maps equivalent X509_NAME structures into a consistent structure.
2476 Name comparison can then be performed rapidly using memcmp().
2479 *) Non-blocking OCSP request processing. Add -timeout option to ocsp
2483 *) Allow digests to supply their own micalg string for S/MIME type using
2484 the ctrl EVP_MD_CTRL_MICALG.
2487 *) During PKCS7 signing pass the PKCS7 SignerInfo structure to the
2488 EVP_PKEY_METHOD before and after signing via the EVP_PKEY_CTRL_PKCS7_SIGN
2489 ctrl. It can then customise the structure before and/or after signing
2493 *) New function OBJ_add_sigid() to allow application defined signature OIDs
2494 to be added to OpenSSLs internal tables. New function OBJ_sigid_free()
2495 to free up any added signature OIDs.
2498 *) New functions EVP_CIPHER_do_all(), EVP_CIPHER_do_all_sorted(),
2499 EVP_MD_do_all() and EVP_MD_do_all_sorted() to enumerate internal
2500 digest and cipher tables. New options added to openssl utility:
2501 list-message-digest-algorithms and list-cipher-algorithms.
2504 *) Change the array representation of binary polynomials: the list
2505 of degrees of non-zero coefficients is now terminated with -1.
2506 Previously it was terminated with 0, which was also part of the
2507 value; thus, the array representation was not applicable to
2508 polynomials where t^0 has coefficient zero. This change makes
2509 the array representation useful in a more general context.
2512 *) Various modifications and fixes to SSL/TLS cipher string
2513 handling. For ECC, the code now distinguishes between fixed ECDH
2514 with RSA certificates on the one hand and with ECDSA certificates
2515 on the other hand, since these are separate ciphersuites. The
2516 unused code for Fortezza ciphersuites has been removed.
2518 For consistency with EDH, ephemeral ECDH is now called "EECDH"
2519 (not "ECDHE"). For consistency with the code for DH
2520 certificates, use of ECDH certificates is now considered ECDH
2521 authentication, not RSA or ECDSA authentication (the latter is
2522 merely the CA's signing algorithm and not actively used in the
2525 The temporary ciphersuite alias "ECCdraft" is no longer
2526 available, and ECC ciphersuites are no longer excluded from "ALL"
2527 and "DEFAULT". The following aliases now exist for RFC 4492
2528 ciphersuites, most of these by analogy with the DH case:
2530 kECDHr - ECDH cert, signed with RSA
2531 kECDHe - ECDH cert, signed with ECDSA
2532 kECDH - ECDH cert (signed with either RSA or ECDSA)
2533 kEECDH - ephemeral ECDH
2534 ECDH - ECDH cert or ephemeral ECDH
2540 AECDH - anonymous ECDH
2541 EECDH - non-anonymous ephemeral ECDH (equivalent to "kEECDH:-AECDH")
2545 *) Add additional S/MIME capabilities for AES and GOST ciphers if supported.
2546 Use correct micalg parameters depending on digest(s) in signed message.
2549 *) Add engine support for EVP_PKEY_ASN1_METHOD. Add functions to process
2550 an ENGINE asn1 method. Support ENGINE lookups in the ASN1 code.
2553 *) Initial engine support for EVP_PKEY_METHOD. New functions to permit
2554 an engine to register a method. Add ENGINE lookups for methods and
2555 functional reference processing.
2558 *) New functions EVP_Digest{Sign,Verify)*. These are enchance versions of
2559 EVP_{Sign,Verify}* which allow an application to customise the signature
2563 *) New -resign option to smime utility. This adds one or more signers
2564 to an existing PKCS#7 signedData structure. Also -md option to use an
2565 alternative message digest algorithm for signing.
2568 *) Tidy up PKCS#7 routines and add new functions to make it easier to
2569 create PKCS7 structures containing multiple signers. Update smime
2570 application to support multiple signers.
2573 *) New -macalg option to pkcs12 utility to allow setting of an alternative
2577 *) Initial support for PKCS#5 v2.0 PRFs other than default SHA1 HMAC.
2578 Reorganize PBE internals to lookup from a static table using NIDs,
2579 add support for HMAC PBE OID translation. Add a EVP_CIPHER ctrl:
2580 EVP_CTRL_PBE_PRF_NID this allows a cipher to specify an alternative
2581 PRF which will be automatically used with PBES2.
2584 *) Replace the algorithm specific calls to generate keys in "req" with the
2588 *) Update PKCS#7 enveloped data routines to use new API. This is now
2589 supported by any public key method supporting the encrypt operation. A
2590 ctrl is added to allow the public key algorithm to examine or modify
2591 the PKCS#7 RecipientInfo structure if it needs to: for RSA this is
2595 *) Add a ctrl to asn1 method to allow a public key algorithm to express
2596 a default digest type to use. In most cases this will be SHA1 but some
2597 algorithms (such as GOST) need to specify an alternative digest. The
2598 return value indicates how strong the prefernce is 1 means optional and
2599 2 is mandatory (that is it is the only supported type). Modify
2600 ASN1_item_sign() to accept a NULL digest argument to indicate it should
2601 use the default md. Update openssl utilities to use the default digest
2602 type for signing if it is not explicitly indicated.
2605 *) Use OID cross reference table in ASN1_sign() and ASN1_verify(). New
2606 EVP_MD flag EVP_MD_FLAG_PKEY_METHOD_SIGNATURE. This uses the relevant
2607 signing method from the key type. This effectively removes the link
2608 between digests and public key types.
2611 *) Add an OID cross reference table and utility functions. Its purpose is to
2612 translate between signature OIDs such as SHA1WithrsaEncryption and SHA1,
2613 rsaEncryption. This will allow some of the algorithm specific hackery
2614 needed to use the correct OID to be removed.
2617 *) Remove algorithm specific dependencies when setting PKCS7_SIGNER_INFO
2618 structures for PKCS7_sign(). They are now set up by the relevant public
2622 *) Add provisional EC pkey method with support for ECDSA and ECDH.
2625 *) Add support for key derivation (agreement) in the API, DH method and
2629 *) Add DSA pkey method and DH pkey methods, extend DH ASN1 method to support
2630 public and private key formats. As a side effect these add additional
2631 command line functionality not previously available: DSA signatures can be
2632 generated and verified using pkeyutl and DH key support and generation in
2637 [Oliver Tappe <zooey@hirschkaefer.de>]
2639 *) New make target "install_html_docs" installs HTML renditions of the
2641 [Oliver Tappe <zooey@hirschkaefer.de>]
2643 *) New utility "genpkey" this is analagous to "genrsa" etc except it can
2644 generate keys for any algorithm. Extend and update EVP_PKEY_METHOD to
2645 support key and parameter generation and add initial key generation
2646 functionality for RSA.
2649 *) Add functions for main EVP_PKEY_method operations. The undocumented
2650 functions EVP_PKEY_{encrypt,decrypt} have been renamed to
2651 EVP_PKEY_{encrypt,decrypt}_old.
2654 *) Initial definitions for EVP_PKEY_METHOD. This will be a high level public
2655 key API, doesn't do much yet.
2658 *) New function EVP_PKEY_asn1_get0_info() to retrieve information about
2659 public key algorithms. New option to openssl utility:
2660 "list-public-key-algorithms" to print out info.
2663 *) Implement the Supported Elliptic Curves Extension for
2664 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
2667 *) Don't free up OIDs in OBJ_cleanup() if they are in use by EVP_MD or
2668 EVP_CIPHER structures to avoid later problems in EVP_cleanup().
2671 *) New utilities pkey and pkeyparam. These are similar to algorithm specific
2672 utilities such as rsa, dsa, dsaparam etc except they process any key
2676 *) Transfer public key printing routines to EVP_PKEY_ASN1_METHOD. New
2677 functions EVP_PKEY_print_public(), EVP_PKEY_print_private(),
2678 EVP_PKEY_print_param() to print public key data from an EVP_PKEY
2682 *) Initial support for pluggable public key ASN1.
2683 De-spaghettify the public key ASN1 handling. Move public and private
2684 key ASN1 handling to a new EVP_PKEY_ASN1_METHOD structure. Relocate
2685 algorithm specific handling to a single module within the relevant
2686 algorithm directory. Add functions to allow (near) opaque processing
2687 of public and private key structures.
2690 *) Implement the Supported Point Formats Extension for
2691 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
2694 *) Add initial support for RFC 4279 PSK TLS ciphersuites. Add members
2695 for the psk identity [hint] and the psk callback functions to the
2696 SSL_SESSION, SSL and SSL_CTX structure.
2699 PSK-RC4-SHA, PSK-3DES-EDE-CBC-SHA, PSK-AES128-CBC-SHA,
2703 SSL_CTX_use_psk_identity_hint
2704 SSL_get_psk_identity_hint
2705 SSL_get_psk_identity
2706 SSL_use_psk_identity_hint
2708 [Mika Kousa and Pasi Eronen of Nokia Corporation]
2710 *) Add RFC 3161 compliant time stamp request creation, response generation
2711 and response verification functionality.
2712 [Zoltán Glózik <zglozik@opentsa.org>, The OpenTSA Project]
2714 *) Add initial support for TLS extensions, specifically for the server_name
2715 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
2716 have new members for a host name. The SSL data structure has an
2717 additional member SSL_CTX *initial_ctx so that new sessions can be
2718 stored in that context to allow for session resumption, even after the
2719 SSL has been switched to a new SSL_CTX in reaction to a client's
2720 server_name extension.
2722 New functions (subject to change):
2724 SSL_get_servername()
2725 SSL_get_servername_type()
2728 New CTRL codes and macros (subject to change):
2730 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
2731 - SSL_CTX_set_tlsext_servername_callback()
2732 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
2733 - SSL_CTX_set_tlsext_servername_arg()
2734 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
2736 openssl s_client has a new '-servername ...' option.
2738 openssl s_server has new options '-servername_host ...', '-cert2 ...',
2739 '-key2 ...', '-servername_fatal' (subject to change). This allows
2740 testing the HostName extension for a specific single host name ('-cert'
2741 and '-key' remain fallbacks for handshakes without HostName
2742 negotiation). If the unrecogninzed_name alert has to be sent, this by
2743 default is a warning; it becomes fatal with the '-servername_fatal'
2746 [Peter Sylvester, Remy Allais, Christophe Renou]
2748 *) Whirlpool hash implementation is added.
2751 *) BIGNUM code on 64-bit SPARCv9 targets is switched from bn(64,64) to
2752 bn(64,32). Because of instruction set limitations it doesn't have
2753 any negative impact on performance. This was done mostly in order
2754 to make it possible to share assembler modules, such as bn_mul_mont
2755 implementations, between 32- and 64-bit builds without hassle.
2758 *) Move code previously exiled into file crypto/ec/ec2_smpt.c
2759 to ec2_smpl.c, and no longer require the OPENSSL_EC_BIN_PT_COMP
2763 *) New candidate for BIGNUM assembler implementation, bn_mul_mont,
2764 dedicated Montgomery multiplication procedure, is introduced.
2765 BN_MONT_CTX is modified to allow bn_mul_mont to reach for higher
2766 "64-bit" performance on certain 32-bit targets.
2769 *) New option SSL_OP_NO_COMP to disable use of compression selectively
2770 in SSL structures. New SSL ctrl to set maximum send fragment size.
2771 Save memory by seeting the I/O buffer sizes dynamically instead of
2772 using the maximum available value.
2775 *) New option -V for 'openssl ciphers'. This prints the ciphersuite code
2776 in addition to the text details.
2779 *) Very, very preliminary EXPERIMENTAL support for printing of general
2780 ASN1 structures. This currently produces rather ugly output and doesn't
2781 handle several customised structures at all.
2784 *) Integrated support for PVK file format and some related formats such
2785 as MS PUBLICKEYBLOB and PRIVATEKEYBLOB. Command line switches to support
2786 these in the 'rsa' and 'dsa' utilities.
2789 *) Support for PKCS#1 RSAPublicKey format on rsa utility command line.
2792 *) Remove the ancient ASN1_METHOD code. This was only ever used in one
2793 place for the (very old) "NETSCAPE" format certificates which are now
2794 handled using new ASN1 code equivalents.
2797 *) Let the TLSv1_method() etc. functions return a 'const' SSL_METHOD
2798 pointer and make the SSL_METHOD parameter in SSL_CTX_new,
2799 SSL_CTX_set_ssl_version and SSL_set_ssl_method 'const'.
2802 *) Modify CRL distribution points extension code to print out previously
2803 unsupported fields. Enhance extension setting code to allow setting of
2807 *) Add print and set support for Issuing Distribution Point CRL extension.
2810 *) Change 'Configure' script to enable Camellia by default.
2813 Changes between 0.9.8m and 0.9.8n [24 Mar 2010]
2815 *) When rejecting SSL/TLS records due to an incorrect version number, never
2816 update s->server with a new major version number. As of
2817 - OpenSSL 0.9.8m if 'short' is a 16-bit type,
2818 - OpenSSL 0.9.8f if 'short' is longer than 16 bits,
2819 the previous behavior could result in a read attempt at NULL when
2820 receiving specific incorrect SSL/TLS records once record payload
2821 protection is active. (CVE-2010-0740)
2822 [Bodo Moeller, Adam Langley <agl@chromium.org>]
2824 *) Fix for CVE-2010-0433 where some kerberos enabled versions of OpenSSL
2825 could be crashed if the relevant tables were not present (e.g. chrooted).
2826 [Tomas Hoger <thoger@redhat.com>]
2828 Changes between 0.9.8l and 0.9.8m [25 Feb 2010]
2830 *) Always check bn_wexpend() return values for failure. (CVE-2009-3245)
2831 [Martin Olsson, Neel Mehta]
2833 *) Fix X509_STORE locking: Every 'objs' access requires a lock (to
2834 accommodate for stack sorting, always a write lock!).
2837 *) On some versions of WIN32 Heap32Next is very slow. This can cause
2838 excessive delays in the RAND_poll(): over a minute. As a workaround
2839 include a time check in the inner Heap32Next loop too.
2842 *) The code that handled flushing of data in SSL/TLS originally used the
2843 BIO_CTRL_INFO ctrl to see if any data was pending first. This caused
2844 the problem outlined in PR#1949. The fix suggested there however can
2845 trigger problems with buggy BIO_CTRL_WPENDING (e.g. some versions
2846 of Apache). So instead simplify the code to flush unconditionally.
2847 This should be fine since flushing with no data to flush is a no op.
2850 *) Handle TLS versions 2.0 and later properly and correctly use the
2851 highest version of TLS/SSL supported. Although TLS >= 2.0 is some way
2852 off ancient servers have a habit of sticking around for a while...
2855 *) Modify compression code so it frees up structures without using the
2856 ex_data callbacks. This works around a problem where some applications
2857 call CRYPTO_cleanup_all_ex_data() before application exit (e.g. when
2858 restarting) then use compression (e.g. SSL with compression) later.
2859 This results in significant per-connection memory leaks and
2860 has caused some security issues including CVE-2008-1678 and
2864 *) Constify crypto/cast (i.e., <openssl/cast.h>): a CAST_KEY doesn't
2865 change when encrypting or decrypting.
2868 *) Add option SSL_OP_LEGACY_SERVER_CONNECT which will allow clients to
2869 connect and renegotiate with servers which do not support RI.
2870 Until RI is more widely deployed this option is enabled by default.
2873 *) Add "missing" ssl ctrls to clear options and mode.
2876 *) If client attempts to renegotiate and doesn't support RI respond with
2877 a no_renegotiation alert as required by RFC5746. Some renegotiating
2878 TLS clients will continue a connection gracefully when they receive
2879 the alert. Unfortunately OpenSSL mishandled this alert and would hang
2880 waiting for a server hello which it will never receive. Now we treat a
2881 received no_renegotiation alert as a fatal error. This is because
2882 applications requesting a renegotiation might well expect it to succeed
2883 and would have no code in place to handle the server denying it so the
2884 only safe thing to do is to terminate the connection.
2887 *) Add ctrl macro SSL_get_secure_renegotiation_support() which returns 1 if
2888 peer supports secure renegotiation and 0 otherwise. Print out peer
2889 renegotiation support in s_client/s_server.
2892 *) Replace the highly broken and deprecated SPKAC certification method with
2893 the updated NID creation version. This should correctly handle UTF8.
2896 *) Implement RFC5746. Re-enable renegotiation but require the extension
2897 as needed. Unfortunately, SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION
2898 turns out to be a bad idea. It has been replaced by
2899 SSL_OP_ALLOW_UNSAFE_LEGACY_RENEGOTIATION which can be set with
2900 SSL_CTX_set_options(). This is really not recommended unless you
2901 know what you are doing.
2902 [Eric Rescorla <ekr@networkresonance.com>, Ben Laurie, Steve Henson]
2904 *) Fixes to stateless session resumption handling. Use initial_ctx when
2905 issuing and attempting to decrypt tickets in case it has changed during
2906 servername handling. Use a non-zero length session ID when attempting
2907 stateless session resumption: this makes it possible to determine if
2908 a resumption has occurred immediately after receiving server hello
2909 (several places in OpenSSL subtly assume this) instead of later in
2913 *) The functions ENGINE_ctrl(), OPENSSL_isservice(),
2914 CMS_get1_RecipientRequest() and RAND_bytes() can return <=0 on error
2915 fixes for a few places where the return code is not checked
2917 [Julia Lawall <julia@diku.dk>]
2919 *) Add --strict-warnings option to Configure script to include devteam
2920 warnings in other configurations.
2923 *) Add support for --libdir option and LIBDIR variable in makefiles. This
2924 makes it possible to install openssl libraries in locations which
2925 have names other than "lib", for example "/usr/lib64" which some
2927 [Steve Henson, based on patch from Jeremy Utley]
2929 *) Don't allow the use of leading 0x80 in OIDs. This is a violation of
2930 X690 8.9.12 and can produce some misleading textual output of OIDs.
2931 [Steve Henson, reported by Dan Kaminsky]
2933 *) Delete MD2 from algorithm tables. This follows the recommendation in
2934 several standards that it is not used in new applications due to
2935 several cryptographic weaknesses. For binary compatibility reasons
2936 the MD2 API is still compiled in by default.
2939 *) Add compression id to {d2i,i2d}_SSL_SESSION so it is correctly saved
2943 *) Rename uni2asc and asc2uni functions to OPENSSL_uni2asc and
2944 OPENSSL_asc2uni conditionally on Netware platforms to avoid a name
2946 [Guenter <lists@gknw.net>]
2948 *) Fix the server certificate chain building code to use X509_verify_cert(),
2949 it used to have an ad-hoc builder which was unable to cope with anything
2950 other than a simple chain.
2951 [David Woodhouse <dwmw2@infradead.org>, Steve Henson]
2953 *) Don't check self signed certificate signatures in X509_verify_cert()
2954 by default (a flag can override this): it just wastes time without
2955 adding any security. As a useful side effect self signed root CAs
2956 with non-FIPS digests are now usable in FIPS mode.
2959 *) In dtls1_process_out_of_seq_message() the check if the current message
2960 is already buffered was missing. For every new message was memory
2961 allocated, allowing an attacker to perform an denial of service attack
2962 with sending out of seq handshake messages until there is no memory
2963 left. Additionally every future messege was buffered, even if the
2964 sequence number made no sense and would be part of another handshake.
2965 So only messages with sequence numbers less than 10 in advance will be
2966 buffered. (CVE-2009-1378)
2967 [Robin Seggelmann, discovered by Daniel Mentz]
2969 *) Records are buffered if they arrive with a future epoch to be
2970 processed after finishing the corresponding handshake. There is
2971 currently no limitation to this buffer allowing an attacker to perform
2972 a DOS attack with sending records with future epochs until there is no
2973 memory left. This patch adds the pqueue_size() function to detemine
2974 the size of a buffer and limits the record buffer to 100 entries.
2976 [Robin Seggelmann, discovered by Daniel Mentz]
2978 *) Keep a copy of frag->msg_header.frag_len so it can be used after the
2979 parent structure is freed. (CVE-2009-1379)
2982 *) Handle non-blocking I/O properly in SSL_shutdown() call.
2983 [Darryl Miles <darryl-mailinglists@netbauds.net>]
2986 [Ilya O. <vrghost@gmail.com>]
2988 Changes between 0.9.8k and 0.9.8l [5 Nov 2009]
2990 *) Disable renegotiation completely - this fixes a severe security
2991 problem (CVE-2009-3555) at the cost of breaking all
2992 renegotiation. Renegotiation can be re-enabled by setting
2993 SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION in s3->flags at
2994 run-time. This is really not recommended unless you know what
2998 Changes between 0.9.8j and 0.9.8k [25 Mar 2009]
3000 *) Don't set val to NULL when freeing up structures, it is freed up by
3001 underlying code. If sizeof(void *) > sizeof(long) this can result in
3002 zeroing past the valid field. (CVE-2009-0789)
3003 [Paolo Ganci <Paolo.Ganci@AdNovum.CH>]
3005 *) Fix bug where return value of CMS_SignerInfo_verify_content() was not
3006 checked correctly. This would allow some invalid signed attributes to
3007 appear to verify correctly. (CVE-2009-0591)
3008 [Ivan Nestlerode <inestlerode@us.ibm.com>]
3010 *) Reject UniversalString and BMPString types with invalid lengths. This
3011 prevents a crash in ASN1_STRING_print_ex() which assumes the strings have
3012 a legal length. (CVE-2009-0590)
3015 *) Set S/MIME signing as the default purpose rather than setting it
3016 unconditionally. This allows applications to override it at the store
3020 *) Permit restricted recursion of ASN1 strings. This is needed in practice
3021 to handle some structures.
3024 *) Improve efficiency of mem_gets: don't search whole buffer each time
3026 [Jeremy Shapiro <jnshapir@us.ibm.com>]
3028 *) New -hex option for openssl rand.
3031 *) Print out UTF8String and NumericString when parsing ASN1.
3034 *) Support NumericString type for name components.
3037 *) Allow CC in the environment to override the automatically chosen
3038 compiler. Note that nothing is done to ensure flags work with the
3042 Changes between 0.9.8i and 0.9.8j [07 Jan 2009]
3044 *) Properly check EVP_VerifyFinal() and similar return values
3046 [Ben Laurie, Bodo Moeller, Google Security Team]
3048 *) Enable TLS extensions by default.
3051 *) Allow the CHIL engine to be loaded, whether the application is
3052 multithreaded or not. (This does not release the developer from the
3053 obligation to set up the dynamic locking callbacks.)
3054 [Sander Temme <sander@temme.net>]
3056 *) Use correct exit code if there is an error in dgst command.
3057 [Steve Henson; problem pointed out by Roland Dirlewanger]
3059 *) Tweak Configure so that you need to say "experimental-jpake" to enable
3060 JPAKE, and need to use -DOPENSSL_EXPERIMENTAL_JPAKE in applications.
3063 *) Add experimental JPAKE support, including demo authentication in
3064 s_client and s_server.
3067 *) Set the comparison function in v3_addr_canonize().
3068 [Rob Austein <sra@hactrn.net>]
3070 *) Add support for XMPP STARTTLS in s_client.
3071 [Philip Paeps <philip@freebsd.org>]
3073 *) Change the server-side SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG behavior
3074 to ensure that even with this option, only ciphersuites in the
3075 server's preference list will be accepted. (Note that the option
3076 applies only when resuming a session, so the earlier behavior was
3077 just about the algorithm choice for symmetric cryptography.)
3080 Changes between 0.9.8h and 0.9.8i [15 Sep 2008]
3082 *) Fix NULL pointer dereference if a DTLS server received
3083 ChangeCipherSpec as first record (CVE-2009-1386).
3086 *) Fix a state transitition in s3_srvr.c and d1_srvr.c
3087 (was using SSL3_ST_CW_CLNT_HELLO_B, should be ..._ST_SW_SRVR_...).
3090 *) The fix in 0.9.8c that supposedly got rid of unsafe
3091 double-checked locking was incomplete for RSA blinding,
3092 addressing just one layer of what turns out to have been
3093 doubly unsafe triple-checked locking.
3095 So now fix this for real by retiring the MONT_HELPER macro
3096 in crypto/rsa/rsa_eay.c.
3098 [Bodo Moeller; problem pointed out by Marius Schilder]
3100 *) Various precautionary measures:
3102 - Avoid size_t integer overflow in HASH_UPDATE (md32_common.h).
3104 - Avoid a buffer overflow in d2i_SSL_SESSION() (ssl_asn1.c).
3105 (NB: This would require knowledge of the secret session ticket key
3106 to exploit, in which case you'd be SOL either way.)
3108 - Change bn_nist.c so that it will properly handle input BIGNUMs
3109 outside the expected range.
3111 - Enforce the 'num' check in BN_div() (bn_div.c) for non-BN_DEBUG
3114 [Neel Mehta, Bodo Moeller]
3116 *) Allow engines to be "soft loaded" - i.e. optionally don't die if
3117 the load fails. Useful for distros.
3118 [Ben Laurie and the FreeBSD team]
3120 *) Add support for Local Machine Keyset attribute in PKCS#12 files.
3123 *) Fix BN_GF2m_mod_arr() top-bit cleanup code.
3126 *) Expand ENGINE to support engine supplied SSL client certificate functions.
3128 This work was sponsored by Logica.
3131 *) Add CryptoAPI ENGINE to support use of RSA and DSA keys held in Windows
3132 keystores. Support for SSL/TLS client authentication too.
3133 Not compiled unless enable-capieng specified to Configure.
3135 This work was sponsored by Logica.
3138 *) Fix bug in X509_ATTRIBUTE creation: dont set attribute using
3139 ASN1_TYPE_set1 if MBSTRING flag set. This bug would crash certain
3140 attribute creation routines such as certifcate requests and PKCS#12
3144 Changes between 0.9.8g and 0.9.8h [28 May 2008]
3146 *) Fix flaw if 'Server Key exchange message' is omitted from a TLS
3147 handshake which could lead to a cilent crash as found using the
3148 Codenomicon TLS test suite (CVE-2008-1672)
3149 [Steve Henson, Mark Cox]
3151 *) Fix double free in TLS server name extensions which could lead to
3152 a remote crash found by Codenomicon TLS test suite (CVE-2008-0891)
3155 *) Clear error queue in SSL_CTX_use_certificate_chain_file()
3157 Clear the error queue to ensure that error entries left from
3158 older function calls do not interfere with the correct operation.
3159 [Lutz Jaenicke, Erik de Castro Lopo]
3161 *) Remove root CA certificates of commercial CAs:
3163 The OpenSSL project does not recommend any specific CA and does not
3164 have any policy with respect to including or excluding any CA.
3165 Therefore it does not make any sense to ship an arbitrary selection
3166 of root CA certificates with the OpenSSL software.
3169 *) RSA OAEP patches to fix two separate invalid memory reads.
3170 The first one involves inputs when 'lzero' is greater than
3171 'SHA_DIGEST_LENGTH' (it would read about SHA_DIGEST_LENGTH bytes
3172 before the beginning of from). The second one involves inputs where
3173 the 'db' section contains nothing but zeroes (there is a one-byte
3174 invalid read after the end of 'db').
3175 [Ivan Nestlerode <inestlerode@us.ibm.com>]
3177 *) Partial backport from 0.9.9-dev:
3179 Introduce bn_mul_mont (dedicated Montgomery multiplication
3180 procedure) as a candidate for BIGNUM assembler implementation.
3181 While 0.9.9-dev uses assembler for various architectures, only
3182 x86_64 is available by default here in the 0.9.8 branch, and
3183 32-bit x86 is available through a compile-time setting.
3185 To try the 32-bit x86 assembler implementation, use Configure
3186 option "enable-montasm" (which exists only for this backport).
3188 As "enable-montasm" for 32-bit x86 disclaims code stability
3189 anyway, in this constellation we activate additional code
3190 backported from 0.9.9-dev for further performance improvements,
3191 namely BN_from_montgomery_word. (To enable this otherwise,
3192 e.g. x86_64, try "-DMONT_FROM_WORD___NON_DEFAULT_0_9_8_BUILD".)
3194 [Andy Polyakov (backport partially by Bodo Moeller)]
3196 *) Add TLS session ticket callback. This allows an application to set
3197 TLS ticket cipher and HMAC keys rather than relying on hardcoded fixed
3198 values. This is useful for key rollover for example where several key
3199 sets may exist with different names.
3202 *) Reverse ENGINE-internal logic for caching default ENGINE handles.
3203 This was broken until now in 0.9.8 releases, such that the only way
3204 a registered ENGINE could be used (assuming it initialises
3205 successfully on the host) was to explicitly set it as the default
3206 for the relevant algorithms. This is in contradiction with 0.9.7
3207 behaviour and the documentation. With this fix, when an ENGINE is
3208 registered into a given algorithm's table of implementations, the
3209 'uptodate' flag is reset so that auto-discovery will be used next
3210 time a new context for that algorithm attempts to select an
3212 [Ian Lister (tweaked by Geoff Thorpe)]
3214 *) Backport of CMS code to OpenSSL 0.9.8. This differs from the 0.9.9
3215 implemention in the following ways:
3217 Lack of EVP_PKEY_ASN1_METHOD means algorithm parameters have to be
3220 Lack of BER streaming support means one pass streaming processing is
3221 only supported if data is detached: setting the streaming flag is
3222 ignored for embedded content.
3224 CMS support is disabled by default and must be explicitly enabled
3225 with the enable-cms configuration option.
3228 *) Update the GMP engine glue to do direct copies between BIGNUM and
3229 mpz_t when openssl and GMP use the same limb size. Otherwise the
3230 existing "conversion via a text string export" trick is still used.
3231 [Paul Sheer <paulsheer@gmail.com>]
3233 *) Zlib compression BIO. This is a filter BIO which compressed and
3234 uncompresses any data passed through it.
3237 *) Add AES_wrap_key() and AES_unwrap_key() functions to implement
3238 RFC3394 compatible AES key wrapping.
3241 *) Add utility functions to handle ASN1 structures. ASN1_STRING_set0():
3242 sets string data without copying. X509_ALGOR_set0() and
3243 X509_ALGOR_get0(): set and retrieve X509_ALGOR (AlgorithmIdentifier)
3244 data. Attribute function X509at_get0_data_by_OBJ(): retrieves data
3245 from an X509_ATTRIBUTE structure optionally checking it occurs only
3246 once. ASN1_TYPE_set1(): set and ASN1_TYPE structure copying supplied
3250 *) Fix BN flag handling in RSA_eay_mod_exp() and BN_MONT_CTX_set()
3251 to get the expected BN_FLG_CONSTTIME behavior.
3252 [Bodo Moeller (Google)]
3256 - fixed wrong usage of ioctlsocket() when build for LIBC BSD sockets
3257 - fixed do_tests.pl to run the test suite with CLIB builds too (CLIB_OPT)
3258 - added some more tests to do_tests.pl
3259 - fixed RunningProcess usage so that it works with newer LIBC NDKs too
3260 - removed usage of BN_LLONG for CLIB builds to avoid runtime dependency
3261 - added new Configure targets netware-clib-bsdsock, netware-clib-gcc,
3262 netware-clib-bsdsock-gcc, netware-libc-bsdsock-gcc
3263 - various changes to netware.pl to enable gcc-cross builds on Win32
3265 - changed crypto/bio/b_sock.c to work with macro functions (CLIB BSD)
3266 - various changes to fix missing prototype warnings
3267 - fixed x86nasm.pl to create correct asm files for NASM COFF output
3268 - added AES, WHIRLPOOL and CPUID assembler code to build files
3269 - added missing AES assembler make rules to mk1mf.pl
3270 - fixed order of includes in apps/ocsp.c so that e_os.h settings apply
3271 [Guenter Knauf <eflash@gmx.net>]
3273 *) Implement certificate status request TLS extension defined in RFC3546.
3274 A client can set the appropriate parameters and receive the encoded
3275 OCSP response via a callback. A server can query the supplied parameters
3276 and set the encoded OCSP response in the callback. Add simplified examples
3277 to s_client and s_server.
3280 Changes between 0.9.8f and 0.9.8g [19 Oct 2007]
3282 *) Fix various bugs:
3283 + Binary incompatibility of ssl_ctx_st structure
3284 + DTLS interoperation with non-compliant servers
3285 + Don't call get_session_cb() without proposed session
3286 + Fix ia64 assembler code
3287 [Andy Polyakov, Steve Henson]
3289 Changes between 0.9.8e and 0.9.8f [11 Oct 2007]
3291 *) DTLS Handshake overhaul. There were longstanding issues with
3292 OpenSSL DTLS implementation, which were making it impossible for
3293 RFC 4347 compliant client to communicate with OpenSSL server.
3294 Unfortunately just fixing these incompatibilities would "cut off"
3295 pre-0.9.8f clients. To allow for hassle free upgrade post-0.9.8e
3296 server keeps tolerating non RFC compliant syntax. The opposite is
3297 not true, 0.9.8f client can not communicate with earlier server.
3298 This update even addresses CVE-2007-4995.
3301 *) Changes to avoid need for function casts in OpenSSL: some compilers
3302 (gcc 4.2 and later) reject their use.
3303 [Kurt Roeckx <kurt@roeckx.be>, Peter Hartley <pdh@utter.chaos.org.uk>,
3306 *) Add RFC4507 support to OpenSSL. This includes the corrections in
3307 RFC4507bis. The encrypted ticket format is an encrypted encoded
3308 SSL_SESSION structure, that way new session features are automatically
3311 If a client application caches session in an SSL_SESSION structure
3312 support is transparent because tickets are now stored in the encoded
3315 The SSL_CTX structure automatically generates keys for ticket
3316 protection in servers so again support should be possible
3317 with no application modification.
3319 If a client or server wishes to disable RFC4507 support then the option
3320 SSL_OP_NO_TICKET can be set.
3322 Add a TLS extension debugging callback to allow the contents of any client
3323 or server extensions to be examined.
3325 This work was sponsored by Google.
3328 *) Add initial support for TLS extensions, specifically for the server_name
3329 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
3330 have new members for a host name. The SSL data structure has an
3331 additional member SSL_CTX *initial_ctx so that new sessions can be
3332 stored in that context to allow for session resumption, even after the
3333 SSL has been switched to a new SSL_CTX in reaction to a client's
3334 server_name extension.
3336 New functions (subject to change):
3338 SSL_get_servername()
3339 SSL_get_servername_type()
3342 New CTRL codes and macros (subject to change):
3344 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
3345 - SSL_CTX_set_tlsext_servername_callback()
3346 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
3347 - SSL_CTX_set_tlsext_servername_arg()
3348 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
3350 openssl s_client has a new '-servername ...' option.
3352 openssl s_server has new options '-servername_host ...', '-cert2 ...',
3353 '-key2 ...', '-servername_fatal' (subject to change). This allows
3354 testing the HostName extension for a specific single host name ('-cert'
3355 and '-key' remain fallbacks for handshakes without HostName
3356 negotiation). If the unrecogninzed_name alert has to be sent, this by
3357 default is a warning; it becomes fatal with the '-servername_fatal'
3360 [Peter Sylvester, Remy Allais, Christophe Renou, Steve Henson]
3362 *) Add AES and SSE2 assembly language support to VC++ build.
3365 *) Mitigate attack on final subtraction in Montgomery reduction.
3368 *) Fix crypto/ec/ec_mult.c to work properly with scalars of value 0
3369 (which previously caused an internal error).
3372 *) Squeeze another 10% out of IGE mode when in != out.
3375 *) AES IGE mode speedup.
3376 [Dean Gaudet (Google)]
3378 *) Add the Korean symmetric 128-bit cipher SEED (see
3379 http://www.kisa.or.kr/kisa/seed/jsp/seed_eng.jsp) and
3380 add SEED ciphersuites from RFC 4162:
3382 TLS_RSA_WITH_SEED_CBC_SHA = "SEED-SHA"
3383 TLS_DHE_DSS_WITH_SEED_CBC_SHA = "DHE-DSS-SEED-SHA"
3384 TLS_DHE_RSA_WITH_SEED_CBC_SHA = "DHE-RSA-SEED-SHA"
3385 TLS_DH_anon_WITH_SEED_CBC_SHA = "ADH-SEED-SHA"
3387 To minimize changes between patchlevels in the OpenSSL 0.9.8
3388 series, SEED remains excluded from compilation unless OpenSSL
3389 is configured with 'enable-seed'.
3390 [KISA, Bodo Moeller]
3392 *) Mitigate branch prediction attacks, which can be practical if a
3393 single processor is shared, allowing a spy process to extract
3394 information. For detailed background information, see
3395 http://eprint.iacr.org/2007/039 (O. Aciicmez, S. Gueron,
3396 J.-P. Seifert, "New Branch Prediction Vulnerabilities in OpenSSL
3397 and Necessary Software Countermeasures"). The core of the change
3398 are new versions BN_div_no_branch() and
3399 BN_mod_inverse_no_branch() of BN_div() and BN_mod_inverse(),
3400 respectively, which are slower, but avoid the security-relevant
3401 conditional branches. These are automatically called by BN_div()
3402 and BN_mod_inverse() if the flag BN_FLG_CONSTTIME is set for one
3403 of the input BIGNUMs. Also, BN_is_bit_set() has been changed to
3404 remove a conditional branch.
3406 BN_FLG_CONSTTIME is the new name for the previous
3407 BN_FLG_EXP_CONSTTIME flag, since it now affects more than just
3408 modular exponentiation. (Since OpenSSL 0.9.7h, setting this flag
3409 in the exponent causes BN_mod_exp_mont() to use the alternative
3410 implementation in BN_mod_exp_mont_consttime().) The old name
3411 remains as a deprecated alias.
3413 Similary, RSA_FLAG_NO_EXP_CONSTTIME is replaced by a more general
3414 RSA_FLAG_NO_CONSTTIME flag since the RSA implementation now uses
3415 constant-time implementations for more than just exponentiation.
3416 Here too the old name is kept as a deprecated alias.
3418 BN_BLINDING_new() will now use BN_dup() for the modulus so that
3419 the BN_BLINDING structure gets an independent copy of the
3420 modulus. This means that the previous "BIGNUM *m" argument to
3421 BN_BLINDING_new() and to BN_BLINDING_create_param() now
3422 essentially becomes "const BIGNUM *m", although we can't actually
3423 change this in the header file before 0.9.9. It allows
3424 RSA_setup_blinding() to use BN_with_flags() on the modulus to
3425 enable BN_FLG_CONSTTIME.
3427 [Matthew D Wood (Intel Corp)]
3429 *) In the SSL/TLS server implementation, be strict about session ID
3430 context matching (which matters if an application uses a single
3431 external cache for different purposes). Previously,
3432 out-of-context reuse was forbidden only if SSL_VERIFY_PEER was
3433 set. This did ensure strict client verification, but meant that,
3434 with applications using a single external cache for quite
3435 different requirements, clients could circumvent ciphersuite
3436 restrictions for a given session ID context by starting a session
3437 in a different context.
3440 *) Include "!eNULL" in SSL_DEFAULT_CIPHER_LIST to make sure that
3441 a ciphersuite string such as "DEFAULT:RSA" cannot enable
3442 authentication-only ciphersuites.
3445 *) Update the SSL_get_shared_ciphers() fix CVE-2006-3738 which was
3446 not complete and could lead to a possible single byte overflow
3447 (CVE-2007-5135) [Ben Laurie]
3449 Changes between 0.9.8d and 0.9.8e [23 Feb 2007]
3451 *) Since AES128 and AES256 (and similarly Camellia128 and
3452 Camellia256) share a single mask bit in the logic of
3453 ssl/ssl_ciph.c, the code for masking out disabled ciphers needs a
3454 kludge to work properly if AES128 is available and AES256 isn't
3455 (or if Camellia128 is available and Camellia256 isn't).
3458 *) Fix the BIT STRING encoding generated by crypto/ec/ec_asn1.c
3459 (within i2d_ECPrivateKey, i2d_ECPKParameters, i2d_ECParameters):
3460 When a point or a seed is encoded in a BIT STRING, we need to
3461 prevent the removal of trailing zero bits to get the proper DER
3462 encoding. (By default, crypto/asn1/a_bitstr.c assumes the case
3463 of a NamedBitList, for which trailing 0 bits need to be removed.)
3466 *) Have SSL/TLS server implementation tolerate "mismatched" record
3467 protocol version while receiving ClientHello even if the
3468 ClientHello is fragmented. (The server can't insist on the
3469 particular protocol version it has chosen before the ServerHello
3470 message has informed the client about his choice.)
3473 *) Add RFC 3779 support.
3474 [Rob Austein for ARIN, Ben Laurie]
3476 *) Load error codes if they are not already present instead of using a
3477 static variable. This allows them to be cleanly unloaded and reloaded.
3478 Improve header file function name parsing.
3481 *) extend SMTP and IMAP protocol emulation in s_client to use EHLO
3482 or CAPABILITY handshake as required by RFCs.
3485 Changes between 0.9.8c and 0.9.8d [28 Sep 2006]
3487 *) Introduce limits to prevent malicious keys being able to
3488 cause a denial of service. (CVE-2006-2940)
3489 [Steve Henson, Bodo Moeller]
3491 *) Fix ASN.1 parsing of certain invalid structures that can result
3492 in a denial of service. (CVE-2006-2937) [Steve Henson]
3494 *) Fix buffer overflow in SSL_get_shared_ciphers() function.
3495 (CVE-2006-3738) [Tavis Ormandy and Will Drewry, Google Security Team]
3497 *) Fix SSL client code which could crash if connecting to a
3498 malicious SSLv2 server. (CVE-2006-4343)
3499 [Tavis Ormandy and Will Drewry, Google Security Team]
3501 *) Since 0.9.8b, ciphersuite strings naming explicit ciphersuites
3502 match only those. Before that, "AES256-SHA" would be interpreted
3503 as a pattern and match "AES128-SHA" too (since AES128-SHA got
3504 the same strength classification in 0.9.7h) as we currently only
3505 have a single AES bit in the ciphersuite description bitmap.
3506 That change, however, also applied to ciphersuite strings such as
3507 "RC4-MD5" that intentionally matched multiple ciphersuites --
3508 namely, SSL 2.0 ciphersuites in addition to the more common ones
3509 from SSL 3.0/TLS 1.0.
3511 So we change the selection algorithm again: Naming an explicit
3512 ciphersuite selects this one ciphersuite, and any other similar
3513 ciphersuite (same bitmap) from *other* protocol versions.
3514 Thus, "RC4-MD5" again will properly select both the SSL 2.0
3515 ciphersuite and the SSL 3.0/TLS 1.0 ciphersuite.
3517 Since SSL 2.0 does not have any ciphersuites for which the
3518 128/256 bit distinction would be relevant, this works for now.
3519 The proper fix will be to use different bits for AES128 and
3520 AES256, which would have avoided the problems from the beginning;
3521 however, bits are scarce, so we can only do this in a new release
3522 (not just a patchlevel) when we can change the SSL_CIPHER
3523 definition to split the single 'unsigned long mask' bitmap into
3524 multiple values to extend the available space.
3528 Changes between 0.9.8b and 0.9.8c [05 Sep 2006]
3530 *) Avoid PKCS #1 v1.5 signature attack discovered by Daniel Bleichenbacher
3531 (CVE-2006-4339) [Ben Laurie and Google Security Team]
3533 *) Add AES IGE and biIGE modes.
3536 *) Change the Unix randomness entropy gathering to use poll() when
3537 possible instead of select(), since the latter has some
3538 undesirable limitations.
3539 [Darryl Miles via Richard Levitte and Bodo Moeller]
3541 *) Disable "ECCdraft" ciphersuites more thoroughly. Now special
3542 treatment in ssl/ssl_ciph.s makes sure that these ciphersuites
3543 cannot be implicitly activated as part of, e.g., the "AES" alias.
3544 However, please upgrade to OpenSSL 0.9.9[-dev] for
3545 non-experimental use of the ECC ciphersuites to get TLS extension
3546 support, which is required for curve and point format negotiation
3547 to avoid potential handshake problems.
3550 *) Disable rogue ciphersuites:
3552 - SSLv2 0x08 0x00 0x80 ("RC4-64-MD5")
3553 - SSLv3/TLSv1 0x00 0x61 ("EXP1024-RC2-CBC-MD5")
3554 - SSLv3/TLSv1 0x00 0x60 ("EXP1024-RC4-MD5")
3556 The latter two were purportedly from
3557 draft-ietf-tls-56-bit-ciphersuites-0[01].txt, but do not really
3560 Also deactivate the remaining ciphersuites from
3561 draft-ietf-tls-56-bit-ciphersuites-01.txt. These are just as
3562 unofficial, and the ID has long expired.
3565 *) Fix RSA blinding Heisenbug (problems sometimes occured on
3566 dual-core machines) and other potential thread-safety issues.
3569 *) Add the symmetric cipher Camellia (128-bit, 192-bit, 256-bit key
3570 versions), which is now available for royalty-free use
3571 (see http://info.isl.ntt.co.jp/crypt/eng/info/chiteki.html).
3572 Also, add Camellia TLS ciphersuites from RFC 4132.
3574 To minimize changes between patchlevels in the OpenSSL 0.9.8
3575 series, Camellia remains excluded from compilation unless OpenSSL
3576 is configured with 'enable-camellia'.
3579 *) Disable the padding bug check when compression is in use. The padding
3580 bug check assumes the first packet is of even length, this is not
3581 necessarily true if compresssion is enabled and can result in false
3582 positives causing handshake failure. The actual bug test is ancient
3583 code so it is hoped that implementations will either have fixed it by
3584 now or any which still have the bug do not support compression.
3587 Changes between 0.9.8a and 0.9.8b [04 May 2006]
3589 *) When applying a cipher rule check to see if string match is an explicit
3590 cipher suite and only match that one cipher suite if it is.
3593 *) Link in manifests for VC++ if needed.
3594 [Austin Ziegler <halostatue@gmail.com>]
3596 *) Update support for ECC-based TLS ciphersuites according to
3597 draft-ietf-tls-ecc-12.txt with proposed changes (but without
3598 TLS extensions, which are supported starting with the 0.9.9
3599 branch, not in the OpenSSL 0.9.8 branch).
3602 *) New functions EVP_CIPHER_CTX_new() and EVP_CIPHER_CTX_free() to support
3603 opaque EVP_CIPHER_CTX handling.
3606 *) Fixes and enhancements to zlib compression code. We now only use
3607 "zlib1.dll" and use the default __cdecl calling convention on Win32
3608 to conform with the standards mentioned here:
3609 http://www.zlib.net/DLL_FAQ.txt
3610 Static zlib linking now works on Windows and the new --with-zlib-include
3611 --with-zlib-lib options to Configure can be used to supply the location
3612 of the headers and library. Gracefully handle case where zlib library
3616 *) Several fixes and enhancements to the OID generation code. The old code
3617 sometimes allowed invalid OIDs (1.X for X >= 40 for example), couldn't
3618 handle numbers larger than ULONG_MAX, truncated printing and had a
3619 non standard OBJ_obj2txt() behaviour.
3622 *) Add support for building of engines under engine/ as shared libraries
3623 under VC++ build system.
3626 *) Corrected the numerous bugs in the Win32 path splitter in DSO.
3627 Hopefully, we will not see any false combination of paths any more.
3630 Changes between 0.9.8 and 0.9.8a [11 Oct 2005]
3632 *) Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING
3633 (part of SSL_OP_ALL). This option used to disable the
3634 countermeasure against man-in-the-middle protocol-version
3635 rollback in the SSL 2.0 server implementation, which is a bad
3636 idea. (CVE-2005-2969)
3638 [Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center
3639 for Information Security, National Institute of Advanced Industrial
3640 Science and Technology [AIST], Japan)]
3642 *) Add two function to clear and return the verify parameter flags.
3645 *) Keep cipherlists sorted in the source instead of sorting them at
3646 runtime, thus removing the need for a lock.
3649 *) Avoid some small subgroup attacks in Diffie-Hellman.
3650 [Nick Mathewson and Ben Laurie]
3652 *) Add functions for well-known primes.
3655 *) Extended Windows CE support.
3656 [Satoshi Nakamura and Andy Polyakov]
3658 *) Initialize SSL_METHOD structures at compile time instead of during
3659 runtime, thus removing the need for a lock.
3662 *) Make PKCS7_decrypt() work even if no certificate is supplied by
3663 attempting to decrypt each encrypted key in turn. Add support to
3667 Changes between 0.9.7h and 0.9.8 [05 Jul 2005]
3669 [NB: OpenSSL 0.9.7i and later 0.9.7 patch levels were released after
3672 *) Add libcrypto.pc and libssl.pc for those who feel they need them.
3675 *) Change CA.sh and CA.pl so they don't bundle the CSR and the private
3676 key into the same file any more.
3679 *) Add initial support for Win64, both IA64 and AMD64/x64 flavors.
3682 *) Add -utf8 command line and config file option to 'ca'.
3683 [Stefan <stf@udoma.org]
3685 *) Removed the macro des_crypt(), as it seems to conflict with some
3686 libraries. Use DES_crypt().
3689 *) Correct naming of the 'chil' and '4758cca' ENGINEs. This
3690 involves renaming the source and generated shared-libs for
3691 both. The engines will accept the corrected or legacy ids
3692 ('ncipher' and '4758_cca' respectively) when binding. NB,
3693 this only applies when building 'shared'.
3694 [Corinna Vinschen <vinschen@redhat.com> and Geoff Thorpe]
3696 *) Add attribute functions to EVP_PKEY structure. Modify
3697 PKCS12_create() to recognize a CSP name attribute and
3698 use it. Make -CSP option work again in pkcs12 utility.
3701 *) Add new functionality to the bn blinding code:
3702 - automatic re-creation of the BN_BLINDING parameters after
3703 a fixed number of uses (currently 32)
3704 - add new function for parameter creation
3705 - introduce flags to control the update behaviour of the
3706 BN_BLINDING parameters
3707 - hide BN_BLINDING structure
3708 Add a second BN_BLINDING slot to the RSA structure to improve
3709 performance when a single RSA object is shared among several
3713 *) Add support for DTLS.
3714 [Nagendra Modadugu <nagendra@cs.stanford.edu> and Ben Laurie]
3716 *) Add support for DER encoded private keys (SSL_FILETYPE_ASN1)
3717 to SSL_CTX_use_PrivateKey_file() and SSL_use_PrivateKey_file()
3720 *) Remove buggy and incompletet DH cert support from
3721 ssl/ssl_rsa.c and ssl/s3_both.c
3724 *) Use SHA-1 instead of MD5 as the default digest algorithm for
3725 the apps/openssl applications.
3728 *) Compile clean with "-Wall -Wmissing-prototypes
3729 -Wstrict-prototypes -Wmissing-declarations -Werror". Currently
3730 DEBUG_SAFESTACK must also be set.
3733 *) Change ./Configure so that certain algorithms can be disabled by default.
3734 The new counterpiece to "no-xxx" is "enable-xxx".
3736 The patented RC5 and MDC2 algorithms will now be disabled unless
3737 "enable-rc5" and "enable-mdc2", respectively, are specified.
3739 (IDEA remains enabled despite being patented. This is because IDEA
3740 is frequently required for interoperability, and there is no license
3741 fee for non-commercial use. As before, "no-idea" can be used to
3742 avoid this algorithm.)
3746 *) Add processing of proxy certificates (see RFC 3820). This work was
3747 sponsored by KTH (The Royal Institute of Technology in Stockholm) and
3748 EGEE (Enabling Grids for E-science in Europe).
3751 *) RC4 performance overhaul on modern architectures/implementations, such
3752 as Intel P4, IA-64 and AMD64.
3755 *) New utility extract-section.pl. This can be used specify an alternative
3756 section number in a pod file instead of having to treat each file as
3757 a separate case in Makefile. This can be done by adding two lines to the
3760 =for comment openssl_section:XXX
3762 The blank line is mandatory.
3766 *) New arguments -certform, -keyform and -pass for s_client and s_server
3767 to allow alternative format key and certificate files and passphrase
3771 *) New structure X509_VERIFY_PARAM which combines current verify parameters,
3772 update associated structures and add various utility functions.
3774 Add new policy related verify parameters, include policy checking in
3775 standard verify code. Enhance 'smime' application with extra parameters
3776 to support policy checking and print out.
3779 *) Add a new engine to support VIA PadLock ACE extensions in the VIA C3
3780 Nehemiah processors. These extensions support AES encryption in hardware
3781 as well as RNG (though RNG support is currently disabled).
3782 [Michal Ludvig <michal@logix.cz>, with help from Andy Polyakov]
3784 *) Deprecate BN_[get|set]_params() functions (they were ignored internally).
3787 *) New FIPS 180-2 algorithms, SHA-224/-256/-384/-512 are implemented.
3788 [Andy Polyakov and a number of other people]
3790 *) Improved PowerPC platform support. Most notably BIGNUM assembler
3791 implementation contributed by IBM.
3792 [Suresh Chari, Peter Waltenberg, Andy Polyakov]
3794 *) The new 'RSA_generate_key_ex' function now takes a BIGNUM for the public
3795 exponent rather than 'unsigned long'. There is a corresponding change to
3796 the new 'rsa_keygen' element of the RSA_METHOD structure.
3797 [Jelte Jansen, Geoff Thorpe]
3799 *) Functionality for creating the initial serial number file is now
3800 moved from CA.pl to the 'ca' utility with a new option -create_serial.
3802 (Before OpenSSL 0.9.7e, CA.pl used to initialize the serial
3803 number file to 1, which is bound to cause problems. To avoid
3804 the problems while respecting compatibility between different 0.9.7
3805 patchlevels, 0.9.7e employed 'openssl x509 -next_serial' in
3806 CA.pl for serial number initialization. With the new release 0.9.8,
3807 we can fix the problem directly in the 'ca' utility.)
3810 *) Reduced header interdepencies by declaring more opaque objects in
3811 ossl_typ.h. As a consequence, including some headers (eg. engine.h) will
3812 give fewer recursive includes, which could break lazy source code - so
3813 this change is covered by the OPENSSL_NO_DEPRECATED symbol. As always,
3814 developers should define this symbol when building and using openssl to
3815 ensure they track the recommended behaviour, interfaces, [etc], but
3816 backwards-compatible behaviour prevails when this isn't defined.
3819 *) New function X509_POLICY_NODE_print() which prints out policy nodes.
3822 *) Add new EVP function EVP_CIPHER_CTX_rand_key and associated functionality.
3823 This will generate a random key of the appropriate length based on the
3824 cipher context. The EVP_CIPHER can provide its own random key generation
3825 routine to support keys of a specific form. This is used in the des and
3826 3des routines to generate a key of the correct parity. Update S/MIME
3827 code to use new functions and hence generate correct parity DES keys.
3828 Add EVP_CHECK_DES_KEY #define to return an error if the key is not
3829 valid (weak or incorrect parity).
3832 *) Add a local set of CRLs that can be used by X509_verify_cert() as well
3833 as looking them up. This is useful when the verified structure may contain
3834 CRLs, for example PKCS#7 signedData. Modify PKCS7_verify() to use any CRLs
3835 present unless the new PKCS7_NO_CRL flag is asserted.
3838 *) Extend ASN1 oid configuration module. It now additionally accepts the
3841 shortName = some long name, 1.2.3.4
3844 *) Reimplemented the BN_CTX implementation. There is now no more static
3845 limitation on the number of variables it can handle nor the depth of the
3846 "stack" handling for BN_CTX_start()/BN_CTX_end() pairs. The stack
3847 information can now expand as required, and rather than having a single
3848 static array of bignums, BN_CTX now uses a linked-list of such arrays
3849 allowing it to expand on demand whilst maintaining the usefulness of
3850 BN_CTX's "bundling".
3853 *) Add a missing BN_CTX parameter to the 'rsa_mod_exp' callback in RSA_METHOD
3854 to allow all RSA operations to function using a single BN_CTX.
3857 *) Preliminary support for certificate policy evaluation and checking. This
3858 is initially intended to pass the tests outlined in "Conformance Testing
3859 of Relying Party Client Certificate Path Processing Logic" v1.07.
3862 *) bn_dup_expand() has been deprecated, it was introduced in 0.9.7 and
3863 remained unused and not that useful. A variety of other little bignum
3864 tweaks and fixes have also been made continuing on from the audit (see
3868 *) Constify all or almost all d2i, c2i, s2i and r2i functions, along with
3869 associated ASN1, EVP and SSL functions and old ASN1 macros.
3872 *) BN_zero() only needs to set 'top' and 'neg' to zero for correct results,
3873 and this should never fail. So the return value from the use of
3874 BN_set_word() (which can fail due to needless expansion) is now deprecated;
3875 if OPENSSL_NO_DEPRECATED is defined, BN_zero() is a void macro.
3878 *) BN_CTX_get() should return zero-valued bignums, providing the same
3879 initialised value as BN_new().
3880 [Geoff Thorpe, suggested by Ulf Möller]
3882 *) Support for inhibitAnyPolicy certificate extension.
3885 *) An audit of the BIGNUM code is underway, for which debugging code is
3886 enabled when BN_DEBUG is defined. This makes stricter enforcements on what
3887 is considered valid when processing BIGNUMs, and causes execution to
3888 assert() when a problem is discovered. If BN_DEBUG_RAND is defined,
3889 further steps are taken to deliberately pollute unused data in BIGNUM
3890 structures to try and expose faulty code further on. For now, openssl will
3891 (in its default mode of operation) continue to tolerate the inconsistent
3892 forms that it has tolerated in the past, but authors and packagers should
3893 consider trying openssl and their own applications when compiled with
3894 these debugging symbols defined. It will help highlight potential bugs in
3895 their own code, and will improve the test coverage for OpenSSL itself. At
3896 some point, these tighter rules will become openssl's default to improve
3897 maintainability, though the assert()s and other overheads will remain only
3898 in debugging configurations. See bn.h for more details.
3899 [Geoff Thorpe, Nils Larsch, Ulf Möller]
3901 *) BN_CTX_init() has been deprecated, as BN_CTX is an opaque structure
3902 that can only be obtained through BN_CTX_new() (which implicitly
3903 initialises it). The presence of this function only made it possible
3904 to overwrite an existing structure (and cause memory leaks).
3907 *) Because of the callback-based approach for implementing LHASH as a
3908 template type, lh_insert() adds opaque objects to hash-tables and
3909 lh_doall() or lh_doall_arg() are typically used with a destructor callback
3910 to clean up those corresponding objects before destroying the hash table
3911 (and losing the object pointers). So some over-zealous constifications in
3912 LHASH have been relaxed so that lh_insert() does not take (nor store) the
3913 objects as "const" and the lh_doall[_arg] callback wrappers are not
3914 prototyped to have "const" restrictions on the object pointers they are
3915 given (and so aren't required to cast them away any more).
3918 *) The tmdiff.h API was so ugly and minimal that our own timing utility
3919 (speed) prefers to use its own implementation. The two implementations
3920 haven't been consolidated as yet (volunteers?) but the tmdiff API has had
3921 its object type properly exposed (MS_TM) instead of casting to/from "char
3922 *". This may still change yet if someone realises MS_TM and "ms_time_***"
3923 aren't necessarily the greatest nomenclatures - but this is what was used
3924 internally to the implementation so I've used that for now.
3927 *) Ensure that deprecated functions do not get compiled when
3928 OPENSSL_NO_DEPRECATED is defined. Some "openssl" subcommands and a few of
3929 the self-tests were still using deprecated key-generation functions so
3930 these have been updated also.
3933 *) Reorganise PKCS#7 code to separate the digest location functionality
3934 into PKCS7_find_digest(), digest addtion into PKCS7_bio_add_digest().
3935 New function PKCS7_set_digest() to set the digest type for PKCS#7
3936 digestedData type. Add additional code to correctly generate the
3937 digestedData type and add support for this type in PKCS7 initialization
3941 *) New function PKCS7_set0_type_other() this initializes a PKCS7
3942 structure of type "other".
3945 *) Fix prime generation loop in crypto/bn/bn_prime.pl by making
3946 sure the loop does correctly stop and breaking ("division by zero")
3947 modulus operations are not performed. The (pre-generated) prime
3948 table crypto/bn/bn_prime.h was already correct, but it could not be
3949 re-generated on some platforms because of the "division by zero"
3950 situation in the script.
3951 [Ralf S. Engelschall]
3953 *) Update support for ECC-based TLS ciphersuites according to
3954 draft-ietf-tls-ecc-03.txt: the KDF1 key derivation function with
3955 SHA-1 now is only used for "small" curves (where the
3956 representation of a field element takes up to 24 bytes); for
3957 larger curves, the field element resulting from ECDH is directly
3958 used as premaster secret.
3959 [Douglas Stebila (Sun Microsystems Laboratories)]
3961 *) Add code for kP+lQ timings to crypto/ec/ectest.c, and add SEC2
3962 curve secp160r1 to the tests.
3963 [Douglas Stebila (Sun Microsystems Laboratories)]
3965 *) Add the possibility to load symbols globally with DSO.
3966 [Götz Babin-Ebell <babin-ebell@trustcenter.de> via Richard Levitte]
3968 *) Add the functions ERR_set_mark() and ERR_pop_to_mark() for better
3969 control of the error stack.
3972 *) Add support for STORE in ENGINE.
3975 *) Add the STORE type. The intention is to provide a common interface
3976 to certificate and key stores, be they simple file-based stores, or
3977 HSM-type store, or LDAP stores, or...
3978 NOTE: The code is currently UNTESTED and isn't really used anywhere.
3981 *) Add a generic structure called OPENSSL_ITEM. This can be used to
3982 pass a list of arguments to any function as well as provide a way
3983 for a function to pass data back to the caller.
3986 *) Add the functions BUF_strndup() and BUF_memdup(). BUF_strndup()
3987 works like BUF_strdup() but can be used to duplicate a portion of
3988 a string. The copy gets NUL-terminated. BUF_memdup() duplicates
3992 *) Add the function sk_find_ex() which works like sk_find(), but will
3993 return an index to an element even if an exact match couldn't be
3994 found. The index is guaranteed to point at the element where the
3995 searched-for key would be inserted to preserve sorting order.
3998 *) Add the function OBJ_bsearch_ex() which works like OBJ_bsearch() but
3999 takes an extra flags argument for optional functionality. Currently,
4000 the following flags are defined:
4002 OBJ_BSEARCH_VALUE_ON_NOMATCH
4003 This one gets OBJ_bsearch_ex() to return a pointer to the first
4004 element where the comparing function returns a negative or zero
4007 OBJ_BSEARCH_FIRST_VALUE_ON_MATCH
4008 This one gets OBJ_bsearch_ex() to return a pointer to the first
4009 element where the comparing function returns zero. This is useful
4010 if there are more than one element where the comparing function
4014 *) Make it possible to create self-signed certificates with 'openssl ca'
4015 in such a way that the self-signed certificate becomes part of the
4016 CA database and uses the same mechanisms for serial number generation
4017 as all other certificate signing. The new flag '-selfsign' enables
4018 this functionality. Adapt CA.sh and CA.pl.in.
4021 *) Add functionality to check the public key of a certificate request
4022 against a given private. This is useful to check that a certificate
4023 request can be signed by that key (self-signing).
4026 *) Make it possible to have multiple active certificates with the same
4027 subject in the CA index file. This is done only if the keyword
4028 'unique_subject' is set to 'no' in the main CA section (default
4029 if 'CA_default') of the configuration file. The value is saved
4030 with the database itself in a separate index attribute file,
4031 named like the index file with '.attr' appended to the name.
4034 *) Generate muti valued AVAs using '+' notation in config files for
4038 *) Support for nameConstraints certificate extension.
4041 *) Support for policyConstraints certificate extension.
4044 *) Support for policyMappings certificate extension.
4047 *) Make sure the default DSA_METHOD implementation only uses its
4048 dsa_mod_exp() and/or bn_mod_exp() handlers if they are non-NULL,
4049 and change its own handlers to be NULL so as to remove unnecessary
4050 indirection. This lets alternative implementations fallback to the
4051 default implementation more easily.
4054 *) Support for directoryName in GeneralName related extensions
4058 *) Make it possible to link applications using Makefile.shared.
4059 Make that possible even when linking against static libraries!
4062 *) Support for single pass processing for S/MIME signing. This now
4063 means that S/MIME signing can be done from a pipe, in addition
4064 cleartext signing (multipart/signed type) is effectively streaming
4065 and the signed data does not need to be all held in memory.
4067 This is done with a new flag PKCS7_STREAM. When this flag is set
4068 PKCS7_sign() only initializes the PKCS7 structure and the actual signing
4069 is done after the data is output (and digests calculated) in
4070 SMIME_write_PKCS7().
4073 *) Add full support for -rpath/-R, both in shared libraries and
4074 applications, at least on the platforms where it's known how
4078 *) In crypto/ec/ec_mult.c, implement fast point multiplication with
4079 precomputation, based on wNAF splitting: EC_GROUP_precompute_mult()
4080 will now compute a table of multiples of the generator that
4081 makes subsequent invocations of EC_POINTs_mul() or EC_POINT_mul()
4082 faster (notably in the case of a single point multiplication,
4083 scalar * generator).
4084 [Nils Larsch, Bodo Moeller]
4086 *) IPv6 support for certificate extensions. The various extensions
4087 which use the IP:a.b.c.d can now take IPv6 addresses using the
4088 formats of RFC1884 2.2 . IPv6 addresses are now also displayed
4092 *) Added an ENGINE that implements RSA by performing private key
4093 exponentiations with the GMP library. The conversions to and from
4094 GMP's mpz_t format aren't optimised nor are any montgomery forms
4095 cached, and on x86 it appears OpenSSL's own performance has caught up.
4096 However there are likely to be other architectures where GMP could
4097 provide a boost. This ENGINE is not built in by default, but it can be
4098 specified at Configure time and should be accompanied by the necessary
4099 linker additions, eg;
4100 ./config -DOPENSSL_USE_GMP -lgmp
4103 *) "openssl engine" will not display ENGINE/DSO load failure errors when
4104 testing availability of engines with "-t" - the old behaviour is
4105 produced by increasing the feature's verbosity with "-tt".
4108 *) ECDSA routines: under certain error conditions uninitialized BN objects
4109 could be freed. Solution: make sure initialization is performed early
4110 enough. (Reported and fix supplied by Nils Larsch <nla@trustcenter.de>
4114 *) Key-generation can now be implemented in RSA_METHOD, DSA_METHOD
4115 and DH_METHOD (eg. by ENGINE implementations) to override the normal
4116 software implementations. For DSA and DH, parameter generation can
4117 also be overriden by providing the appropriate method callbacks.
4120 *) Change the "progress" mechanism used in key-generation and
4121 primality testing to functions that take a new BN_GENCB pointer in
4122 place of callback/argument pairs. The new API functions have "_ex"
4123 postfixes and the older functions are reimplemented as wrappers for
4124 the new ones. The OPENSSL_NO_DEPRECATED symbol can be used to hide
4125 declarations of the old functions to help (graceful) attempts to
4126 migrate to the new functions. Also, the new key-generation API
4127 functions operate on a caller-supplied key-structure and return
4128 success/failure rather than returning a key or NULL - this is to
4129 help make "keygen" another member function of RSA_METHOD etc.
4131 Example for using the new callback interface:
4133 int (*my_callback)(int a, int b, BN_GENCB *cb) = ...;
4137 BN_GENCB_set(&my_cb, my_callback, my_arg);
4139 return BN_is_prime_ex(some_bignum, BN_prime_checks, NULL, &cb);
4140 /* For the meaning of a, b in calls to my_callback(), see the
4141 * documentation of the function that calls the callback.
4142 * cb will point to my_cb; my_arg can be retrieved as cb->arg.
4143 * my_callback should return 1 if it wants BN_is_prime_ex()
4144 * to continue, or 0 to stop.
4149 *) Change the ZLIB compression method to be stateful, and make it
4150 available to TLS with the number defined in
4151 draft-ietf-tls-compression-04.txt.
4154 *) Add the ASN.1 structures and functions for CertificatePair, which
4155 is defined as follows (according to X.509_4thEditionDraftV6.pdf):
4157 CertificatePair ::= SEQUENCE {
4158 forward [0] Certificate OPTIONAL,
4159 reverse [1] Certificate OPTIONAL,
4160 -- at least one of the pair shall be present -- }
4162 Also implement the PEM functions to read and write certificate
4163 pairs, and defined the PEM tag as "CERTIFICATE PAIR".
4165 This needed to be defined, mostly for the sake of the LDAP
4166 attribute crossCertificatePair, but may prove useful elsewhere as
4170 *) Make it possible to inhibit symlinking of shared libraries in
4171 Makefile.shared, for Cygwin's sake.
4174 *) Extend the BIGNUM API by creating a function
4175 void BN_set_negative(BIGNUM *a, int neg);
4176 and a macro that behave like
4177 int BN_is_negative(const BIGNUM *a);
4179 to avoid the need to access 'a->neg' directly in applications.
4182 *) Implement fast modular reduction for pseudo-Mersenne primes
4183 used in NIST curves (crypto/bn/bn_nist.c, crypto/ec/ecp_nist.c).
4184 EC_GROUP_new_curve_GFp() will now automatically use this
4186 [Nils Larsch <nla@trustcenter.de>]
4188 *) Add new lock type (CRYPTO_LOCK_BN).
4191 *) Change the ENGINE framework to automatically load engines
4192 dynamically from specific directories unless they could be
4193 found to already be built in or loaded. Move all the
4194 current engines except for the cryptodev one to a new
4196 The engines in engines/ are built as shared libraries if
4197 the "shared" options was given to ./Configure or ./config.
4198 Otherwise, they are inserted in libcrypto.a.
4199 /usr/local/ssl/engines is the default directory for dynamic
4200 engines, but that can be overriden at configure time through
4201 the usual use of --prefix and/or --openssldir, and at run
4202 time with the environment variable OPENSSL_ENGINES.
4203 [Geoff Thorpe and Richard Levitte]
4205 *) Add Makefile.shared, a helper makefile to build shared
4206 libraries. Addapt Makefile.org.
4209 *) Add version info to Win32 DLLs.
4210 [Peter 'Luna' Runestig" <peter@runestig.com>]
4212 *) Add new 'medium level' PKCS#12 API. Certificates and keys
4213 can be added using this API to created arbitrary PKCS#12
4214 files while avoiding the low level API.
4216 New options to PKCS12_create(), key or cert can be NULL and
4217 will then be omitted from the output file. The encryption
4218 algorithm NIDs can be set to -1 for no encryption, the mac
4219 iteration count can be set to 0 to omit the mac.
4221 Enhance pkcs12 utility by making the -nokeys and -nocerts
4222 options work when creating a PKCS#12 file. New option -nomac
4223 to omit the mac, NONE can be set for an encryption algorithm.
4224 New code is modified to use the enhanced PKCS12_create()
4225 instead of the low level API.
4228 *) Extend ASN1 encoder to support indefinite length constructed
4229 encoding. This can output sequences tags and octet strings in
4230 this form. Modify pk7_asn1.c to support indefinite length
4231 encoding. This is experimental and needs additional code to
4232 be useful, such as an ASN1 bio and some enhanced streaming
4235 Extend template encode functionality so that tagging is passed
4236 down to the template encoder.
4239 *) Let 'openssl req' fail if an argument to '-newkey' is not
4240 recognized instead of using RSA as a default.
4243 *) Add support for ECC-based ciphersuites from draft-ietf-tls-ecc-01.txt.
4244 As these are not official, they are not included in "ALL";
4245 the "ECCdraft" ciphersuite group alias can be used to select them.
4246 [Vipul Gupta and Sumit Gupta (Sun Microsystems Laboratories)]
4248 *) Add ECDH engine support.
4249 [Nils Gura and Douglas Stebila (Sun Microsystems Laboratories)]
4251 *) Add ECDH in new directory crypto/ecdh/.
4252 [Douglas Stebila (Sun Microsystems Laboratories)]
4254 *) Let BN_rand_range() abort with an error after 100 iterations
4255 without success (which indicates a broken PRNG).
4258 *) Change BN_mod_sqrt() so that it verifies that the input value
4259 is really the square of the return value. (Previously,
4260 BN_mod_sqrt would show GIGO behaviour.)
4263 *) Add named elliptic curves over binary fields from X9.62, SECG,
4264 and WAP/WTLS; add OIDs that were still missing.
4266 [Sheueling Chang Shantz and Douglas Stebila
4267 (Sun Microsystems Laboratories)]
4269 *) Extend the EC library for elliptic curves over binary fields
4270 (new files ec2_smpl.c, ec2_smpt.c, ec2_mult.c in crypto/ec/).
4273 EC_GF2m_simple_method
4277 EC_GROUP_new_curve_GF2m
4278 EC_GROUP_set_curve_GF2m
4279 EC_GROUP_get_curve_GF2m
4280 EC_POINT_set_affine_coordinates_GF2m
4281 EC_POINT_get_affine_coordinates_GF2m
4282 EC_POINT_set_compressed_coordinates_GF2m
4284 Point compression for binary fields is disabled by default for
4285 patent reasons (compile with OPENSSL_EC_BIN_PT_COMP defined to
4288 As binary polynomials are represented as BIGNUMs, various members
4289 of the EC_GROUP and EC_POINT data structures can be shared
4290 between the implementations for prime fields and binary fields;
4291 the above ..._GF2m functions (except for EX_GROUP_new_curve_GF2m)
4292 are essentially identical to their ..._GFp counterparts.
4293 (For simplicity, the '..._GFp' prefix has been dropped from
4294 various internal method names.)
4296 An internal 'field_div' method (similar to 'field_mul' and
4297 'field_sqr') has been added; this is used only for binary fields.
4299 [Sheueling Chang Shantz and Douglas Stebila
4300 (Sun Microsystems Laboratories)]
4302 *) Optionally dispatch EC_POINT_mul(), EC_POINT_precompute_mult()
4303 through methods ('mul', 'precompute_mult').
4305 The generic implementations (now internally called 'ec_wNAF_mul'
4306 and 'ec_wNAF_precomputed_mult') remain the default if these
4307 methods are undefined.
4309 [Sheueling Chang Shantz and Douglas Stebila
4310 (Sun Microsystems Laboratories)]
4312 *) New function EC_GROUP_get_degree, which is defined through
4313 EC_METHOD. For curves over prime fields, this returns the bit
4314 length of the modulus.
4316 [Sheueling Chang Shantz and Douglas Stebila
4317 (Sun Microsystems Laboratories)]
4319 *) New functions EC_GROUP_dup, EC_POINT_dup.
4320 (These simply call ..._new and ..._copy).
4322 [Sheueling Chang Shantz and Douglas Stebila
4323 (Sun Microsystems Laboratories)]
4325 *) Add binary polynomial arithmetic software in crypto/bn/bn_gf2m.c.
4326 Polynomials are represented as BIGNUMs (where the sign bit is not
4327 used) in the following functions [macros]:
4330 BN_GF2m_sub [= BN_GF2m_add]
4331 BN_GF2m_mod [wrapper for BN_GF2m_mod_arr]
4332 BN_GF2m_mod_mul [wrapper for BN_GF2m_mod_mul_arr]
4333 BN_GF2m_mod_sqr [wrapper for BN_GF2m_mod_sqr_arr]
4335 BN_GF2m_mod_exp [wrapper for BN_GF2m_mod_exp_arr]
4336 BN_GF2m_mod_sqrt [wrapper for BN_GF2m_mod_sqrt_arr]
4337 BN_GF2m_mod_solve_quad [wrapper for BN_GF2m_mod_solve_quad_arr]
4338 BN_GF2m_cmp [= BN_ucmp]
4340 (Note that only the 'mod' functions are actually for fields GF(2^m).
4341 BN_GF2m_add() is misnomer, but this is for the sake of consistency.)
4343 For some functions, an the irreducible polynomial defining a
4344 field can be given as an 'unsigned int[]' with strictly
4345 decreasing elements giving the indices of those bits that are set;
4346 i.e., p[] represents the polynomial
4347 f(t) = t^p[0] + t^p[1] + ... + t^p[k]
4349 p[0] > p[1] > ... > p[k] = 0.
4350 This applies to the following functions:
4355 BN_GF2m_mod_inv_arr [wrapper for BN_GF2m_mod_inv]
4356 BN_GF2m_mod_div_arr [wrapper for BN_GF2m_mod_div]
4358 BN_GF2m_mod_sqrt_arr
4359 BN_GF2m_mod_solve_quad_arr
4363 Conversion can be performed by the following functions:
4368 bntest.c has additional tests for binary polynomial arithmetic.
4370 Two implementations for BN_GF2m_mod_div() are available.
4371 The default algorithm simply uses BN_GF2m_mod_inv() and
4372 BN_GF2m_mod_mul(). The alternative algorithm is compiled in only
4373 if OPENSSL_SUN_GF2M_DIV is defined (patent pending; read the
4374 copyright notice in crypto/bn/bn_gf2m.c before enabling it).
4376 [Sheueling Chang Shantz and Douglas Stebila
4377 (Sun Microsystems Laboratories)]
4379 *) Add new error code 'ERR_R_DISABLED' that can be used when some
4380 functionality is disabled at compile-time.
4381 [Douglas Stebila <douglas.stebila@sun.com>]
4383 *) Change default behaviour of 'openssl asn1parse' so that more
4384 information is visible when viewing, e.g., a certificate:
4386 Modify asn1_parse2 (crypto/asn1/asn1_par.c) so that in non-'dump'
4387 mode the content of non-printable OCTET STRINGs is output in a
4388 style similar to INTEGERs, but with '[HEX DUMP]' prepended to
4389 avoid the appearance of a printable string.
4390 [Nils Larsch <nla@trustcenter.de>]
4392 *) Add 'asn1_flag' and 'asn1_form' member to EC_GROUP with access
4394 EC_GROUP_set_asn1_flag()
4395 EC_GROUP_get_asn1_flag()
4396 EC_GROUP_set_point_conversion_form()
4397 EC_GROUP_get_point_conversion_form()
4398 These control ASN1 encoding details:
4399 - Curves (i.e., groups) are encoded explicitly unless asn1_flag
4400 has been set to OPENSSL_EC_NAMED_CURVE.
4401 - Points are encoded in uncompressed form by default; options for
4402 asn1_for are as for point2oct, namely
4403 POINT_CONVERSION_COMPRESSED
4404 POINT_CONVERSION_UNCOMPRESSED
4405 POINT_CONVERSION_HYBRID
4407 Also add 'seed' and 'seed_len' members to EC_GROUP with access
4410 EC_GROUP_get0_seed()
4411 EC_GROUP_get_seed_len()
4412 This is used only for ASN1 purposes (so far).
4413 [Nils Larsch <nla@trustcenter.de>]
4415 *) Add 'field_type' member to EC_METHOD, which holds the NID
4416 of the appropriate field type OID. The new function
4417 EC_METHOD_get_field_type() returns this value.
4418 [Nils Larsch <nla@trustcenter.de>]
4423 EC_POINT_point2hex()
4424 EC_POINT_hex2point()
4425 providing useful interfaces to EC_POINT_point2oct() and
4426 EC_POINT_oct2point().
4427 [Nils Larsch <nla@trustcenter.de>]
4429 *) Change internals of the EC library so that the functions
4430 EC_GROUP_set_generator()
4431 EC_GROUP_get_generator()
4432 EC_GROUP_get_order()
4433 EC_GROUP_get_cofactor()
4434 are implemented directly in crypto/ec/ec_lib.c and not dispatched
4435 to methods, which would lead to unnecessary code duplication when
4436 adding different types of curves.
4437 [Nils Larsch <nla@trustcenter.de> with input by Bodo Moeller]
4439 *) Implement compute_wNAF (crypto/ec/ec_mult.c) without BIGNUM
4440 arithmetic, and such that modified wNAFs are generated
4441 (which avoid length expansion in many cases).
4444 *) Add a function EC_GROUP_check_discriminant() (defined via
4445 EC_METHOD) that verifies that the curve discriminant is non-zero.
4447 Add a function EC_GROUP_check() that makes some sanity tests
4448 on a EC_GROUP, its generator and order. This includes
4449 EC_GROUP_check_discriminant().
4450 [Nils Larsch <nla@trustcenter.de>]
4452 *) Add ECDSA in new directory crypto/ecdsa/.
4454 Add applications 'openssl ecparam' and 'openssl ecdsa'
4455 (these are based on 'openssl dsaparam' and 'openssl dsa').
4457 ECDSA support is also included in various other files across the
4458 library. Most notably,
4459 - 'openssl req' now has a '-newkey ecdsa:file' option;
4460 - EVP_PKCS82PKEY (crypto/evp/evp_pkey.c) now can handle ECDSA;
4461 - X509_PUBKEY_get (crypto/asn1/x_pubkey.c) and
4462 d2i_PublicKey (crypto/asn1/d2i_pu.c) have been modified to make
4463 them suitable for ECDSA where domain parameters must be
4464 extracted before the specific public key;
4465 - ECDSA engine support has been added.
4466 [Nils Larsch <nla@trustcenter.de>]
4468 *) Include some named elliptic curves, and add OIDs from X9.62,
4469 SECG, and WAP/WTLS. Each curve can be obtained from the new
4471 EC_GROUP_new_by_curve_name(),
4472 and the list of available named curves can be obtained with
4473 EC_get_builtin_curves().
4474 Also add a 'curve_name' member to EC_GROUP objects, which can be
4476 EC_GROUP_set_curve_name()
4477 EC_GROUP_get_curve_name()
4478 [Nils Larsch <larsch@trustcenter.de, Bodo Moeller]
4480 *) Remove a few calls to bn_wexpand() in BN_sqr() (the one in there
4481 was actually never needed) and in BN_mul(). The removal in BN_mul()
4482 required a small change in bn_mul_part_recursive() and the addition
4483 of the functions bn_cmp_part_words(), bn_sub_part_words() and
4484 bn_add_part_words(), which do the same thing as bn_cmp_words(),
4485 bn_sub_words() and bn_add_words() except they take arrays with
4489 Changes between 0.9.7l and 0.9.7m [23 Feb 2007]
4491 *) Cleanse PEM buffers before freeing them since they may contain
4493 [Benjamin Bennett <ben@psc.edu>]
4495 *) Include "!eNULL" in SSL_DEFAULT_CIPHER_LIST to make sure that
4496 a ciphersuite string such as "DEFAULT:RSA" cannot enable
4497 authentication-only ciphersuites.
4500 *) Since AES128 and AES256 share a single mask bit in the logic of
4501 ssl/ssl_ciph.c, the code for masking out disabled ciphers needs a
4502 kludge to work properly if AES128 is available and AES256 isn't.
4505 *) Expand security boundary to match 1.1.1 module.
4508 *) Remove redundant features: hash file source, editing of test vectors
4509 modify fipsld to use external fips_premain.c signature.
4512 *) New perl script mkfipsscr.pl to create shell scripts or batch files to
4513 run algorithm test programs.
4516 *) Make algorithm test programs more tolerant of whitespace.
4519 *) Have SSL/TLS server implementation tolerate "mismatched" record
4520 protocol version while receiving ClientHello even if the
4521 ClientHello is fragmented. (The server can't insist on the
4522 particular protocol version it has chosen before the ServerHello
4523 message has informed the client about his choice.)
4526 *) Load error codes if they are not already present instead of using a
4527 static variable. This allows them to be cleanly unloaded and reloaded.
4530 Changes between 0.9.7k and 0.9.7l [28 Sep 2006]
4532 *) Introduce limits to prevent malicious keys being able to
4533 cause a denial of service. (CVE-2006-2940)
4534 [Steve Henson, Bodo Moeller]
4536 *) Fix ASN.1 parsing of certain invalid structures that can result
4537 in a denial of service. (CVE-2006-2937) [Steve Henson]
4539 *) Fix buffer overflow in SSL_get_shared_ciphers() function.
4540 (CVE-2006-3738) [Tavis Ormandy and Will Drewry, Google Security Team]
4542 *) Fix SSL client code which could crash if connecting to a
4543 malicious SSLv2 server. (CVE-2006-4343)
4544 [Tavis Ormandy and Will Drewry, Google Security Team]
4546 *) Change ciphersuite string processing so that an explicit
4547 ciphersuite selects this one ciphersuite (so that "AES256-SHA"
4548 will no longer include "AES128-SHA"), and any other similar
4549 ciphersuite (same bitmap) from *other* protocol versions (so that
4550 "RC4-MD5" will still include both the SSL 2.0 ciphersuite and the
4551 SSL 3.0/TLS 1.0 ciphersuite). This is a backport combining
4552 changes from 0.9.8b and 0.9.8d.
4555 Changes between 0.9.7j and 0.9.7k [05 Sep 2006]
4557 *) Avoid PKCS #1 v1.5 signature attack discovered by Daniel Bleichenbacher
4558 (CVE-2006-4339) [Ben Laurie and Google Security Team]
4560 *) Change the Unix randomness entropy gathering to use poll() when
4561 possible instead of select(), since the latter has some
4562 undesirable limitations.
4563 [Darryl Miles via Richard Levitte and Bodo Moeller]
4565 *) Disable rogue ciphersuites:
4567 - SSLv2 0x08 0x00 0x80 ("RC4-64-MD5")
4568 - SSLv3/TLSv1 0x00 0x61 ("EXP1024-RC2-CBC-MD5")
4569 - SSLv3/TLSv1 0x00 0x60 ("EXP1024-RC4-MD5")
4571 The latter two were purportedly from
4572 draft-ietf-tls-56-bit-ciphersuites-0[01].txt, but do not really
4575 Also deactive the remaining ciphersuites from
4576 draft-ietf-tls-56-bit-ciphersuites-01.txt. These are just as
4577 unofficial, and the ID has long expired.
4580 *) Fix RSA blinding Heisenbug (problems sometimes occured on
4581 dual-core machines) and other potential thread-safety issues.
4584 Changes between 0.9.7i and 0.9.7j [04 May 2006]
4586 *) Adapt fipsld and the build system to link against the validated FIPS
4587 module in FIPS mode.
4590 *) Fixes for VC++ 2005 build under Windows.
4593 *) Add new Windows build target VC-32-GMAKE for VC++. This uses GNU make
4594 from a Windows bash shell such as MSYS. It is autodetected from the
4595 "config" script when run from a VC++ environment. Modify standard VC++
4596 build to use fipscanister.o from the GNU make build.
4599 Changes between 0.9.7h and 0.9.7i [14 Oct 2005]
4601 *) Wrapped the definition of EVP_MAX_MD_SIZE in a #ifdef OPENSSL_FIPS.
4602 The value now differs depending on if you build for FIPS or not.
4603 BEWARE! A program linked with a shared FIPSed libcrypto can't be
4604 safely run with a non-FIPSed libcrypto, as it may crash because of
4605 the difference induced by this change.
4608 Changes between 0.9.7g and 0.9.7h [11 Oct 2005]
4610 *) Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING
4611 (part of SSL_OP_ALL). This option used to disable the
4612 countermeasure against man-in-the-middle protocol-version
4613 rollback in the SSL 2.0 server implementation, which is a bad
4614 idea. (CVE-2005-2969)
4616 [Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center
4617 for Information Security, National Institute of Advanced Industrial
4618 Science and Technology [AIST], Japan)]
4620 *) Minimal support for X9.31 signatures and PSS padding modes. This is
4621 mainly for FIPS compliance and not fully integrated at this stage.
4624 *) For DSA signing, unless DSA_FLAG_NO_EXP_CONSTTIME is set, perform
4625 the exponentiation using a fixed-length exponent. (Otherwise,
4626 the information leaked through timing could expose the secret key
4627 after many signatures; cf. Bleichenbacher's attack on DSA with
4631 *) Make a new fixed-window mod_exp implementation the default for
4632 RSA, DSA, and DH private-key operations so that the sequence of
4633 squares and multiplies and the memory access pattern are
4634 independent of the particular secret key. This will mitigate
4635 cache-timing and potential related attacks.
4637 BN_mod_exp_mont_consttime() is the new exponentiation implementation,
4638 and this is automatically used by BN_mod_exp_mont() if the new flag
4639 BN_FLG_EXP_CONSTTIME is set for the exponent. RSA, DSA, and DH
4640 will use this BN flag for private exponents unless the flag
4641 RSA_FLAG_NO_EXP_CONSTTIME, DSA_FLAG_NO_EXP_CONSTTIME, or
4642 DH_FLAG_NO_EXP_CONSTTIME, respectively, is set.
4644 [Matthew D Wood (Intel Corp), with some changes by Bodo Moeller]
4646 *) Change the client implementation for SSLv23_method() and
4647 SSLv23_client_method() so that is uses the SSL 3.0/TLS 1.0
4648 Client Hello message format if the SSL_OP_NO_SSLv2 option is set.
4649 (Previously, the SSL 2.0 backwards compatible Client Hello
4650 message format would be used even with SSL_OP_NO_SSLv2.)
4653 *) Add support for smime-type MIME parameter in S/MIME messages which some
4657 *) New function BN_MONT_CTX_set_locked() to set montgomery parameters in
4658 a threadsafe manner. Modify rsa code to use new function and add calls
4659 to dsa and dh code (which had race conditions before).
4662 *) Include the fixed error library code in the C error file definitions
4663 instead of fixing them up at runtime. This keeps the error code
4664 structures constant.
4667 Changes between 0.9.7f and 0.9.7g [11 Apr 2005]
4669 [NB: OpenSSL 0.9.7h and later 0.9.7 patch levels were released after
4672 *) Fixes for newer kerberos headers. NB: the casts are needed because
4673 the 'length' field is signed on one version and unsigned on another
4674 with no (?) obvious way to tell the difference, without these VC++
4675 complains. Also the "definition" of FAR (blank) is no longer included
4676 nor is the error ENOMEM. KRB5_PRIVATE has to be set to 1 to pick up
4677 some needed definitions.
4680 *) Undo Cygwin change.
4683 *) Added support for proxy certificates according to RFC 3820.
4684 Because they may be a security thread to unaware applications,
4685 they must be explicitely allowed in run-time. See
4686 docs/HOWTO/proxy_certificates.txt for further information.
4689 Changes between 0.9.7e and 0.9.7f [22 Mar 2005]
4691 *) Use (SSL_RANDOM_VALUE - 4) bytes of pseudo random data when generating
4692 server and client random values. Previously
4693 (SSL_RANDOM_VALUE - sizeof(time_t)) would be used which would result in
4694 less random data when sizeof(time_t) > 4 (some 64 bit platforms).
4696 This change has negligible security impact because:
4698 1. Server and client random values still have 24 bytes of pseudo random
4701 2. Server and client random values are sent in the clear in the initial
4704 3. The master secret is derived using the premaster secret (48 bytes in
4705 size for static RSA ciphersuites) as well as client server and random
4708 The OpenSSL team would like to thank the UK NISCC for bringing this issue
4711 [Stephen Henson, reported by UK NISCC]
4713 *) Use Windows randomness collection on Cygwin.
4716 *) Fix hang in EGD/PRNGD query when communication socket is closed
4717 prematurely by EGD/PRNGD.
4718 [Darren Tucker <dtucker@zip.com.au> via Lutz Jänicke, resolves #1014]
4720 *) Prompt for pass phrases when appropriate for PKCS12 input format.
4723 *) Back-port of selected performance improvements from development
4724 branch, as well as improved support for PowerPC platforms.
4727 *) Add lots of checks for memory allocation failure, error codes to indicate
4728 failure and freeing up memory if a failure occurs.
4729 [Nauticus Networks SSL Team <openssl@nauticusnet.com>, Steve Henson]
4731 *) Add new -passin argument to dgst.
4734 *) Perform some character comparisons of different types in X509_NAME_cmp:
4735 this is needed for some certificates that reencode DNs into UTF8Strings
4736 (in violation of RFC3280) and can't or wont issue name rollover
4740 *) Make an explicit check during certificate validation to see that
4741 the CA setting in each certificate on the chain is correct. As a
4742 side effect always do the following basic checks on extensions,
4743 not just when there's an associated purpose to the check:
4745 - if there is an unhandled critical extension (unless the user
4746 has chosen to ignore this fault)
4747 - if the path length has been exceeded (if one is set at all)
4748 - that certain extensions fit the associated purpose (if one has
4752 Changes between 0.9.7d and 0.9.7e [25 Oct 2004]
4754 *) Avoid a race condition when CRLs are checked in a multi threaded
4755 environment. This would happen due to the reordering of the revoked
4756 entries during signature checking and serial number lookup. Now the
4757 encoding is cached and the serial number sort performed under a lock.
4758 Add new STACK function sk_is_sorted().
4761 *) Add Delta CRL to the extension code.
4764 *) Various fixes to s3_pkt.c so alerts are sent properly.
4765 [David Holmes <d.holmes@f5.com>]
4767 *) Reduce the chances of duplicate issuer name and serial numbers (in
4768 violation of RFC3280) using the OpenSSL certificate creation utilities.
4769 This is done by creating a random 64 bit value for the initial serial
4770 number when a serial number file is created or when a self signed
4771 certificate is created using 'openssl req -x509'. The initial serial
4772 number file is created using 'openssl x509 -next_serial' in CA.pl
4773 rather than being initialized to 1.
4776 Changes between 0.9.7c and 0.9.7d [17 Mar 2004]
4778 *) Fix null-pointer assignment in do_change_cipher_spec() revealed
4779 by using the Codenomicon TLS Test Tool (CVE-2004-0079)
4780 [Joe Orton, Steve Henson]
4782 *) Fix flaw in SSL/TLS handshaking when using Kerberos ciphersuites
4784 [Joe Orton, Steve Henson]
4786 *) Make it possible to have multiple active certificates with the same
4787 subject in the CA index file. This is done only if the keyword
4788 'unique_subject' is set to 'no' in the main CA section (default
4789 if 'CA_default') of the configuration file. The value is saved
4790 with the database itself in a separate index attribute file,
4791 named like the index file with '.attr' appended to the name.
4794 *) X509 verify fixes. Disable broken certificate workarounds when
4795 X509_V_FLAGS_X509_STRICT is set. Check CRL issuer has cRLSign set if
4796 keyUsage extension present. Don't accept CRLs with unhandled critical
4797 extensions: since verify currently doesn't process CRL extensions this
4798 rejects a CRL with *any* critical extensions. Add new verify error codes
4802 *) When creating an OCSP nonce use an OCTET STRING inside the extnValue.
4803 A clarification of RFC2560 will require the use of OCTET STRINGs and
4804 some implementations cannot handle the current raw format. Since OpenSSL
4805 copies and compares OCSP nonces as opaque blobs without any attempt at
4806 parsing them this should not create any compatibility issues.
4809 *) New md flag EVP_MD_CTX_FLAG_REUSE this allows md_data to be reused when
4810 calling EVP_MD_CTX_copy_ex() to avoid calling OPENSSL_malloc(). Without
4811 this HMAC (and other) operations are several times slower than OpenSSL
4815 *) Print out GeneralizedTime and UTCTime in ASN1_STRING_print_ex().
4816 [Peter Sylvester <Peter.Sylvester@EdelWeb.fr>]
4818 *) Use the correct content when signing type "other".
4821 Changes between 0.9.7b and 0.9.7c [30 Sep 2003]
4823 *) Fix various bugs revealed by running the NISCC test suite:
4825 Stop out of bounds reads in the ASN1 code when presented with
4826 invalid tags (CVE-2003-0543 and CVE-2003-0544).
4828 Free up ASN1_TYPE correctly if ANY type is invalid (CVE-2003-0545).
4830 If verify callback ignores invalid public key errors don't try to check
4831 certificate signature with the NULL public key.
4835 *) New -ignore_err option in ocsp application to stop the server
4836 exiting on the first error in a request.
4839 *) In ssl3_accept() (ssl/s3_srvr.c) only accept a client certificate
4840 if the server requested one: as stated in TLS 1.0 and SSL 3.0
4844 *) In ssl3_get_client_hello() (ssl/s3_srvr.c), tolerate additional
4845 extra data after the compression methods not only for TLS 1.0
4846 but also for SSL 3.0 (as required by the specification).
4847 [Bodo Moeller; problem pointed out by Matthias Loepfe]
4849 *) Change X509_certificate_type() to mark the key as exported/exportable
4850 when it's 512 *bits* long, not 512 bytes.
4853 *) Change AES_cbc_encrypt() so it outputs exact multiple of
4854 blocks during encryption.
4857 *) Various fixes to base64 BIO and non blocking I/O. On write
4858 flushes were not handled properly if the BIO retried. On read
4859 data was not being buffered properly and had various logic bugs.
4860 This also affects blocking I/O when the data being decoded is a
4864 *) Various S/MIME bugfixes and compatibility changes:
4865 output correct application/pkcs7 MIME type if
4866 PKCS7_NOOLDMIMETYPE is set. Tolerate some broken signatures.
4867 Output CR+LF for EOL if PKCS7_CRLFEOL is set (this makes opening
4868 of files as .eml work). Correctly handle very long lines in MIME
4872 Changes between 0.9.7a and 0.9.7b [10 Apr 2003]
4874 *) Countermeasure against the Klima-Pokorny-Rosa extension of
4875 Bleichbacher's attack on PKCS #1 v1.5 padding: treat
4876 a protocol version number mismatch like a decryption error
4877 in ssl3_get_client_key_exchange (ssl/s3_srvr.c).
4880 *) Turn on RSA blinding by default in the default implementation
4881 to avoid a timing attack. Applications that don't want it can call
4882 RSA_blinding_off() or use the new flag RSA_FLAG_NO_BLINDING.
4883 They would be ill-advised to do so in most cases.
4884 [Ben Laurie, Steve Henson, Geoff Thorpe, Bodo Moeller]
4886 *) Change RSA blinding code so that it works when the PRNG is not
4887 seeded (in this case, the secret RSA exponent is abused as
4888 an unpredictable seed -- if it is not unpredictable, there
4889 is no point in blinding anyway). Make RSA blinding thread-safe
4890 by remembering the creator's thread ID in rsa->blinding and
4891 having all other threads use local one-time blinding factors
4892 (this requires more computation than sharing rsa->blinding, but
4893 avoids excessive locking; and if an RSA object is not shared
4894 between threads, blinding will still be very fast).
4897 *) Fixed a typo bug that would cause ENGINE_set_default() to set an
4898 ENGINE as defaults for all supported algorithms irrespective of
4899 the 'flags' parameter. 'flags' is now honoured, so applications
4900 should make sure they are passing it correctly.
4903 *) Target "mingw" now allows native Windows code to be generated in
4904 the Cygwin environment as well as with the MinGW compiler.
4907 Changes between 0.9.7 and 0.9.7a [19 Feb 2003]
4909 *) In ssl3_get_record (ssl/s3_pkt.c), minimize information leaked
4910 via timing by performing a MAC computation even if incorrrect
4911 block cipher padding has been found. This is a countermeasure
4912 against active attacks where the attacker has to distinguish
4913 between bad padding and a MAC verification error. (CVE-2003-0078)
4915 [Bodo Moeller; problem pointed out by Brice Canvel (EPFL),
4916 Alain Hiltgen (UBS), Serge Vaudenay (EPFL), and
4917 Martin Vuagnoux (EPFL, Ilion)]
4919 *) Make the no-err option work as intended. The intention with no-err
4920 is not to have the whole error stack handling routines removed from
4921 libcrypto, it's only intended to remove all the function name and
4922 reason texts, thereby removing some of the footprint that may not
4923 be interesting if those errors aren't displayed anyway.
4925 NOTE: it's still possible for any application or module to have it's
4926 own set of error texts inserted. The routines are there, just not
4927 used by default when no-err is given.
4930 *) Add support for FreeBSD on IA64.
4931 [dirk.meyer@dinoex.sub.org via Richard Levitte, resolves #454]
4933 *) Adjust DES_cbc_cksum() so it returns the same value as the MIT
4934 Kerberos function mit_des_cbc_cksum(). Before this change,
4935 the value returned by DES_cbc_cksum() was like the one from
4936 mit_des_cbc_cksum(), except the bytes were swapped.
4937 [Kevin Greaney <Kevin.Greaney@hp.com> and Richard Levitte]
4939 *) Allow an application to disable the automatic SSL chain building.
4940 Before this a rather primitive chain build was always performed in
4941 ssl3_output_cert_chain(): an application had no way to send the
4942 correct chain if the automatic operation produced an incorrect result.
4944 Now the chain builder is disabled if either:
4946 1. Extra certificates are added via SSL_CTX_add_extra_chain_cert().
4948 2. The mode flag SSL_MODE_NO_AUTO_CHAIN is set.
4950 The reasoning behind this is that an application would not want the
4951 auto chain building to take place if extra chain certificates are
4952 present and it might also want a means of sending no additional
4953 certificates (for example the chain has two certificates and the
4957 *) Add the possibility to build without the ENGINE framework.
4958 [Steven Reddie <smr@essemer.com.au> via Richard Levitte]
4960 *) Under Win32 gmtime() can return NULL: check return value in
4961 OPENSSL_gmtime(). Add error code for case where gmtime() fails.
4964 *) DSA routines: under certain error conditions uninitialized BN objects
4965 could be freed. Solution: make sure initialization is performed early
4966 enough. (Reported and fix supplied by Ivan D Nestlerode <nestler@MIT.EDU>,
4967 Nils Larsch <nla@trustcenter.de> via PR#459)
4970 *) Another fix for SSLv2 session ID handling: the session ID was incorrectly
4971 checked on reconnect on the client side, therefore session resumption
4972 could still fail with a "ssl session id is different" error. This
4973 behaviour is masked when SSL_OP_ALL is used due to
4974 SSL_OP_MICROSOFT_SESS_ID_BUG being set.
4975 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as
4976 followup to PR #377.
4979 *) IA-32 assembler support enhancements: unified ELF targets, support
4980 for SCO/Caldera platforms, fix for Cygwin shared build.
4983 *) Add support for FreeBSD on sparc64. As a consequence, support for
4984 FreeBSD on non-x86 processors is separate from x86 processors on
4985 the config script, much like the NetBSD support.
4986 [Richard Levitte & Kris Kennaway <kris@obsecurity.org>]
4988 Changes between 0.9.6h and 0.9.7 [31 Dec 2002]
4990 [NB: OpenSSL 0.9.6i and later 0.9.6 patch levels were released after
4993 *) Fix session ID handling in SSLv2 client code: the SERVER FINISHED
4994 code (06) was taken as the first octet of the session ID and the last
4995 octet was ignored consequently. As a result SSLv2 client side session
4996 caching could not have worked due to the session ID mismatch between
4998 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as
5002 *) Change the declaration of needed Kerberos libraries to use EX_LIBS
5003 instead of the special (and badly supported) LIBKRB5. LIBKRB5 is
5007 *) The hw_ncipher.c engine requires dynamic locks. Unfortunately, it
5008 seems that in spite of existing for more than a year, many application
5009 author have done nothing to provide the necessary callbacks, which
5010 means that this particular engine will not work properly anywhere.
5011 This is a very unfortunate situation which forces us, in the name
5012 of usability, to give the hw_ncipher.c a static lock, which is part
5014 NOTE: This is for the 0.9.7 series ONLY. This hack will never
5015 appear in 0.9.8 or later. We EXPECT application authors to have
5016 dealt properly with this when 0.9.8 is released (unless we actually
5017 make such changes in the libcrypto locking code that changes will
5018 have to be made anyway).
5021 *) In asn1_d2i_read_bio() repeatedly call BIO_read() until all content
5022 octets have been read, EOF or an error occurs. Without this change
5023 some truncated ASN1 structures will not produce an error.
5026 *) Disable Heimdal support, since it hasn't been fully implemented.
5027 Still give the possibility to force the use of Heimdal, but with
5028 warnings and a request that patches get sent to openssl-dev.
5031 *) Add the VC-CE target, introduce the WINCE sysname, and add
5032 INSTALL.WCE and appropriate conditionals to make it build.
5033 [Steven Reddie <smr@essemer.com.au> via Richard Levitte]
5035 *) Change the DLL names for Cygwin to cygcrypto-x.y.z.dll and
5036 cygssl-x.y.z.dll, where x, y and z are the major, minor and
5037 edit numbers of the version.
5038 [Corinna Vinschen <vinschen@redhat.com> and Richard Levitte]
5040 *) Introduce safe string copy and catenation functions
5041 (BUF_strlcpy() and BUF_strlcat()).
5042 [Ben Laurie (CHATS) and Richard Levitte]
5044 *) Avoid using fixed-size buffers for one-line DNs.
5045 [Ben Laurie (CHATS)]
5047 *) Add BUF_MEM_grow_clean() to avoid information leakage when
5048 resizing buffers containing secrets, and use where appropriate.
5049 [Ben Laurie (CHATS)]
5051 *) Avoid using fixed size buffers for configuration file location.
5052 [Ben Laurie (CHATS)]
5054 *) Avoid filename truncation for various CA files.
5055 [Ben Laurie (CHATS)]
5057 *) Use sizeof in preference to magic numbers.
5058 [Ben Laurie (CHATS)]
5060 *) Avoid filename truncation in cert requests.
5061 [Ben Laurie (CHATS)]
5063 *) Add assertions to check for (supposedly impossible) buffer
5065 [Ben Laurie (CHATS)]
5067 *) Don't cache truncated DNS entries in the local cache (this could
5068 potentially lead to a spoofing attack).
5069 [Ben Laurie (CHATS)]
5071 *) Fix various buffers to be large enough for hex/decimal
5072 representations in a platform independent manner.
5073 [Ben Laurie (CHATS)]
5075 *) Add CRYPTO_realloc_clean() to avoid information leakage when
5076 resizing buffers containing secrets, and use where appropriate.
5077 [Ben Laurie (CHATS)]
5079 *) Add BIO_indent() to avoid much slightly worrying code to do
5081 [Ben Laurie (CHATS)]
5083 *) Convert sprintf()/BIO_puts() to BIO_printf().
5084 [Ben Laurie (CHATS)]
5086 *) buffer_gets() could terminate with the buffer only half
5088 [Ben Laurie (CHATS)]
5090 *) Add assertions to prevent user-supplied crypto functions from
5091 overflowing internal buffers by having large block sizes, etc.
5092 [Ben Laurie (CHATS)]
5094 *) New OPENSSL_assert() macro (similar to assert(), but enabled
5096 [Ben Laurie (CHATS)]
5098 *) Eliminate unused copy of key in RC4.
5099 [Ben Laurie (CHATS)]
5101 *) Eliminate unused and incorrectly sized buffers for IV in pem.h.
5102 [Ben Laurie (CHATS)]
5104 *) Fix off-by-one error in EGD path.
5105 [Ben Laurie (CHATS)]
5107 *) If RANDFILE path is too long, ignore instead of truncating.
5108 [Ben Laurie (CHATS)]
5110 *) Eliminate unused and incorrectly sized X.509 structure
5112 [Ben Laurie (CHATS)]
5114 *) Eliminate unused and dangerous function knumber().
5115 [Ben Laurie (CHATS)]
5117 *) Eliminate unused and dangerous structure, KSSL_ERR.
5118 [Ben Laurie (CHATS)]
5120 *) Protect against overlong session ID context length in an encoded
5121 session object. Since these are local, this does not appear to be
5123 [Ben Laurie (CHATS)]
5125 *) Change from security patch (see 0.9.6e below) that did not affect
5126 the 0.9.6 release series:
5128 Remote buffer overflow in SSL3 protocol - an attacker could
5129 supply an oversized master key in Kerberos-enabled versions.
5131 [Ben Laurie (CHATS)]
5133 *) Change the SSL kerb5 codes to match RFC 2712.
5136 *) Make -nameopt work fully for req and add -reqopt switch.
5137 [Michael Bell <michael.bell@rz.hu-berlin.de>, Steve Henson]
5139 *) The "block size" for block ciphers in CFB and OFB mode should be 1.
5140 [Steve Henson, reported by Yngve Nysaeter Pettersen <yngve@opera.com>]
5142 *) Make sure tests can be performed even if the corresponding algorithms
5143 have been removed entirely. This was also the last step to make
5144 OpenSSL compilable with DJGPP under all reasonable conditions.
5145 [Richard Levitte, Doug Kaufman <dkaufman@rahul.net>]
5147 *) Add cipher selection rules COMPLEMENTOFALL and COMPLEMENTOFDEFAULT
5148 to allow version independent disabling of normally unselected ciphers,
5149 which may be activated as a side-effect of selecting a single cipher.
5151 (E.g., cipher list string "RSA" enables ciphersuites that are left
5152 out of "ALL" because they do not provide symmetric encryption.
5153 "RSA:!COMPLEMEMENTOFALL" avoids these unsafe ciphersuites.)
5154 [Lutz Jaenicke, Bodo Moeller]
5156 *) Add appropriate support for separate platform-dependent build
5157 directories. The recommended way to make a platform-dependent
5158 build directory is the following (tested on Linux), maybe with
5161 # Place yourself outside of the OpenSSL source tree. In
5162 # this example, the environment variable OPENSSL_SOURCE
5163 # is assumed to contain the absolute OpenSSL source directory.
5164 mkdir -p objtree/"`uname -s`-`uname -r`-`uname -m`"
5165 cd objtree/"`uname -s`-`uname -r`-`uname -m`"
5166 (cd $OPENSSL_SOURCE; find . -type f) | while read F; do
5167 mkdir -p `dirname $F`
5168 ln -s $OPENSSL_SOURCE/$F $F
5171 To be absolutely sure not to disturb the source tree, a "make clean"
5172 is a good thing. If it isn't successfull, don't worry about it,
5173 it probably means the source directory is very clean.
5176 *) Make sure any ENGINE control commands make local copies of string
5177 pointers passed to them whenever necessary. Otherwise it is possible
5178 the caller may have overwritten (or deallocated) the original string
5179 data when a later ENGINE operation tries to use the stored values.
5180 [Götz Babin-Ebell <babinebell@trustcenter.de>]
5182 *) Improve diagnostics in file reading and command-line digests.
5183 [Ben Laurie aided and abetted by Solar Designer <solar@openwall.com>]
5185 *) Add AES modes CFB and OFB to the object database. Correct an
5186 error in AES-CFB decryption.
5189 *) Remove most calls to EVP_CIPHER_CTX_cleanup() in evp_enc.c, this
5190 allows existing EVP_CIPHER_CTX structures to be reused after
5191 calling EVP_*Final(). This behaviour is used by encryption
5192 BIOs and some applications. This has the side effect that
5193 applications must explicitly clean up cipher contexts with
5194 EVP_CIPHER_CTX_cleanup() or they will leak memory.
5197 *) Check the values of dna and dnb in bn_mul_recursive before calling
5198 bn_mul_comba (a non zero value means the a or b arrays do not contain
5199 n2 elements) and fallback to bn_mul_normal if either is not zero.
5202 *) Fix escaping of non-ASCII characters when using the -subj option
5203 of the "openssl req" command line tool. (Robert Joop <joop@fokus.gmd.de>)
5206 *) Make object definitions compliant to LDAP (RFC2256): SN is the short
5207 form for "surname", serialNumber has no short form.
5208 Use "mail" as the short name for "rfc822Mailbox" according to RFC2798;
5209 therefore remove "mail" short name for "internet 7".
5210 The OID for unique identifiers in X509 certificates is
5211 x500UniqueIdentifier, not uniqueIdentifier.
5212 Some more OID additions. (Michael Bell <michael.bell@rz.hu-berlin.de>)
5215 *) Add an "init" command to the ENGINE config module and auto initialize
5216 ENGINEs. Without any "init" command the ENGINE will be initialized
5217 after all ctrl commands have been executed on it. If init=1 the
5218 ENGINE is initailized at that point (ctrls before that point are run
5219 on the uninitialized ENGINE and after on the initialized one). If
5220 init=0 then the ENGINE will not be iniatialized at all.
5223 *) Fix the 'app_verify_callback' interface so that the user-defined
5224 argument is actually passed to the callback: In the
5225 SSL_CTX_set_cert_verify_callback() prototype, the callback
5226 declaration has been changed from
5229 int (*cb)(X509_STORE_CTX *,void *);
5230 in ssl_verify_cert_chain (ssl/ssl_cert.c), the call
5231 i=s->ctx->app_verify_callback(&ctx)
5232 has been changed into
5233 i=s->ctx->app_verify_callback(&ctx, s->ctx->app_verify_arg).
5235 To update applications using SSL_CTX_set_cert_verify_callback(),
5236 a dummy argument can be added to their callback functions.
5237 [D. K. Smetters <smetters@parc.xerox.com>]
5239 *) Added the '4758cca' ENGINE to support IBM 4758 cards.
5240 [Maurice Gittens <maurice@gittens.nl>, touchups by Geoff Thorpe]
5242 *) Add and OPENSSL_LOAD_CONF define which will cause
5243 OpenSSL_add_all_algorithms() to load the openssl.cnf config file.
5244 This allows older applications to transparently support certain
5245 OpenSSL features: such as crypto acceleration and dynamic ENGINE loading.
5246 Two new functions OPENSSL_add_all_algorithms_noconf() which will never
5247 load the config file and OPENSSL_add_all_algorithms_conf() which will
5248 always load it have also been added.
5251 *) Add the OFB, CFB and CTR (all with 128 bit feedback) to AES.
5252 Adjust NIDs and EVP layer.
5253 [Stephen Sprunk <stephen@sprunk.org> and Richard Levitte]
5255 *) Config modules support in openssl utility.
5257 Most commands now load modules from the config file,
5258 though in a few (such as version) this isn't done
5259 because it couldn't be used for anything.
5261 In the case of ca and req the config file used is
5262 the same as the utility itself: that is the -config
5263 command line option can be used to specify an
5267 *) Move default behaviour from OPENSSL_config(). If appname is NULL
5268 use "openssl_conf" if filename is NULL use default openssl config file.
5271 *) Add an argument to OPENSSL_config() to allow the use of an alternative
5272 config section name. Add a new flag to tolerate a missing config file
5273 and move code to CONF_modules_load_file().
5276 *) Support for crypto accelerator cards from Accelerated Encryption
5277 Processing, www.aep.ie. (Use engine 'aep')
5278 The support was copied from 0.9.6c [engine] and adapted/corrected
5279 to work with the new engine framework.
5280 [AEP Inc. and Richard Levitte]
5282 *) Support for SureWare crypto accelerator cards from Baltimore
5283 Technologies. (Use engine 'sureware')
5284 The support was copied from 0.9.6c [engine] and adapted
5285 to work with the new engine framework.
5288 *) Have the CHIL engine fork-safe (as defined by nCipher) and actually
5289 make the newer ENGINE framework commands for the CHIL engine work.
5290 [Toomas Kiisk <vix@cyber.ee> and Richard Levitte]
5292 *) Make it possible to produce shared libraries on ReliantUNIX.
5293 [Robert Dahlem <Robert.Dahlem@ffm2.siemens.de> via Richard Levitte]
5295 *) Add the configuration target debug-linux-ppro.
5296 Make 'openssl rsa' use the general key loading routines
5297 implemented in apps.c, and make those routines able to
5298 handle the key format FORMAT_NETSCAPE and the variant
5300 [Toomas Kiisk <vix@cyber.ee> via Richard Levitte]
5302 *) Fix a crashbug and a logic bug in hwcrhk_load_pubkey().
5303 [Toomas Kiisk <vix@cyber.ee> via Richard Levitte]
5305 *) Add -keyform to rsautl, and document -engine.
5306 [Richard Levitte, inspired by Toomas Kiisk <vix@cyber.ee>]
5308 *) Change BIO_new_file (crypto/bio/bss_file.c) to use new
5309 BIO_R_NO_SUCH_FILE error code rather than the generic
5310 ERR_R_SYS_LIB error code if fopen() fails with ENOENT.
5313 *) Add new functions
5315 ERR_peek_last_error_line
5316 ERR_peek_last_error_line_data.
5317 These are similar to
5320 ERR_peek_error_line_data,
5321 but report on the latest error recorded rather than the first one
5322 still in the error queue.
5323 [Ben Laurie, Bodo Moeller]
5325 *) default_algorithms option in ENGINE config module. This allows things
5327 default_algorithms = ALL
5328 default_algorithms = RSA, DSA, RAND, CIPHERS, DIGESTS
5331 *) Prelminary ENGINE config module.
5334 *) New experimental application configuration code.
5337 *) Change the AES code to follow the same name structure as all other
5338 symmetric ciphers, and behave the same way. Move everything to
5339 the directory crypto/aes, thereby obsoleting crypto/rijndael.
5340 [Stephen Sprunk <stephen@sprunk.org> and Richard Levitte]
5342 *) SECURITY: remove unsafe setjmp/signal interaction from ui_openssl.c.
5343 [Ben Laurie and Theo de Raadt]
5345 *) Add option to output public keys in req command.
5346 [Massimiliano Pala madwolf@openca.org]
5348 *) Use wNAFs in EC_POINTs_mul() for improved efficiency
5349 (up to about 10% better than before for P-192 and P-224).
5352 *) New functions/macros
5354 SSL_CTX_set_msg_callback(ctx, cb)
5355 SSL_CTX_set_msg_callback_arg(ctx, arg)
5356 SSL_set_msg_callback(ssl, cb)
5357 SSL_set_msg_callback_arg(ssl, arg)
5359 to request calling a callback function
5361 void cb(int write_p, int version, int content_type,
5362 const void *buf, size_t len, SSL *ssl, void *arg)
5364 whenever a protocol message has been completely received
5365 (write_p == 0) or sent (write_p == 1). Here 'version' is the
5366 protocol version according to which the SSL library interprets
5367 the current protocol message (SSL2_VERSION, SSL3_VERSION, or
5368 TLS1_VERSION). 'content_type' is 0 in the case of SSL 2.0, or
5369 the content type as defined in the SSL 3.0/TLS 1.0 protocol
5370 specification (change_cipher_spec(20), alert(21), handshake(22)).
5371 'buf' and 'len' point to the actual message, 'ssl' to the
5372 SSL object, and 'arg' is the application-defined value set by
5373 SSL[_CTX]_set_msg_callback_arg().
5375 'openssl s_client' and 'openssl s_server' have new '-msg' options
5376 to enable a callback that displays all protocol messages.
5379 *) Change the shared library support so shared libraries are built as
5380 soon as the corresponding static library is finished, and thereby get
5381 openssl and the test programs linked against the shared library.
5382 This still only happens when the keyword "shard" has been given to
5383 the configuration scripts.
5385 NOTE: shared library support is still an experimental thing, and
5386 backward binary compatibility is still not guaranteed.
5387 ["Maciej W. Rozycki" <macro@ds2.pg.gda.pl> and Richard Levitte]
5389 *) Add support for Subject Information Access extension.
5390 [Peter Sylvester <Peter.Sylvester@EdelWeb.fr>]
5392 *) Make BUF_MEM_grow() behaviour more consistent: Initialise to zero
5393 additional bytes when new memory had to be allocated, not just
5394 when reusing an existing buffer.
5397 *) New command line and configuration option 'utf8' for the req command.
5398 This allows field values to be specified as UTF8 strings.
5401 *) Add -multi and -mr options to "openssl speed" - giving multiple parallel
5402 runs for the former and machine-readable output for the latter.
5405 *) Add '-noemailDN' option to 'openssl ca'. This prevents inclusion
5406 of the e-mail address in the DN (i.e., it will go into a certificate
5407 extension only). The new configuration file option 'email_in_dn = no'
5408 has the same effect.
5409 [Massimiliano Pala madwolf@openca.org]
5411 *) Change all functions with names starting with des_ to be starting
5412 with DES_ instead. Add wrappers that are compatible with libdes,
5413 but are named _ossl_old_des_*. Finally, add macros that map the
5414 des_* symbols to the corresponding _ossl_old_des_* if libdes
5415 compatibility is desired. If OpenSSL 0.9.6c compatibility is
5416 desired, the des_* symbols will be mapped to DES_*, with one
5419 Since we provide two compatibility mappings, the user needs to
5420 define the macro OPENSSL_DES_LIBDES_COMPATIBILITY if libdes
5421 compatibility is desired. The default (i.e., when that macro
5422 isn't defined) is OpenSSL 0.9.6c compatibility.
5424 There are also macros that enable and disable the support of old
5425 des functions altogether. Those are OPENSSL_ENABLE_OLD_DES_SUPPORT
5426 and OPENSSL_DISABLE_OLD_DES_SUPPORT. If none or both of those
5427 are defined, the default will apply: to support the old des routines.
5429 In either case, one must include openssl/des.h to get the correct
5430 definitions. Do not try to just include openssl/des_old.h, that
5433 NOTE: This is a major break of an old API into a new one. Software
5434 authors are encouraged to switch to the DES_ style functions. Some
5435 time in the future, des_old.h and the libdes compatibility functions
5436 will be disable (i.e. OPENSSL_DISABLE_OLD_DES_SUPPORT will be the
5437 default), and then completely removed.
5440 *) Test for certificates which contain unsupported critical extensions.
5441 If such a certificate is found during a verify operation it is
5442 rejected by default: this behaviour can be overridden by either
5443 handling the new error X509_V_ERR_UNHANDLED_CRITICAL_EXTENSION or
5444 by setting the verify flag X509_V_FLAG_IGNORE_CRITICAL. A new function
5445 X509_supported_extension() has also been added which returns 1 if a
5446 particular extension is supported.
5449 *) Modify the behaviour of EVP cipher functions in similar way to digests
5450 to retain compatibility with existing code.
5453 *) Modify the behaviour of EVP_DigestInit() and EVP_DigestFinal() to retain
5454 compatibility with existing code. In particular the 'ctx' parameter does
5455 not have to be to be initialized before the call to EVP_DigestInit() and
5456 it is tidied up after a call to EVP_DigestFinal(). New function
5457 EVP_DigestFinal_ex() which does not tidy up the ctx. Similarly function
5458 EVP_MD_CTX_copy() changed to not require the destination to be
5459 initialized valid and new function EVP_MD_CTX_copy_ex() added which
5460 requires the destination to be valid.
5462 Modify all the OpenSSL digest calls to use EVP_DigestInit_ex(),
5463 EVP_DigestFinal_ex() and EVP_MD_CTX_copy_ex().
5466 *) Change ssl3_get_message (ssl/s3_both.c) and the functions using it
5467 so that complete 'Handshake' protocol structures are kept in memory
5468 instead of overwriting 'msg_type' and 'length' with 'body' data.
5471 *) Add an implementation of SSL_add_dir_cert_subjects_to_stack for Win32.
5472 [Massimo Santin via Richard Levitte]
5474 *) Major restructuring to the underlying ENGINE code. This includes
5475 reduction of linker bloat, separation of pure "ENGINE" manipulation
5476 (initialisation, etc) from functionality dealing with implementations
5477 of specific crypto iterfaces. This change also introduces integrated
5478 support for symmetric ciphers and digest implementations - so ENGINEs
5479 can now accelerate these by providing EVP_CIPHER and EVP_MD
5480 implementations of their own. This is detailed in crypto/engine/README
5481 as it couldn't be adequately described here. However, there are a few
5482 API changes worth noting - some RSA, DSA, DH, and RAND functions that
5483 were changed in the original introduction of ENGINE code have now
5484 reverted back - the hooking from this code to ENGINE is now a good
5485 deal more passive and at run-time, operations deal directly with
5486 RSA_METHODs, DSA_METHODs (etc) as they did before, rather than
5487 dereferencing through an ENGINE pointer any more. Also, the ENGINE
5488 functions dealing with BN_MOD_EXP[_CRT] handlers have been removed -
5489 they were not being used by the framework as there is no concept of a
5490 BIGNUM_METHOD and they could not be generalised to the new
5491 'ENGINE_TABLE' mechanism that underlies the new code. Similarly,
5492 ENGINE_cpy() has been removed as it cannot be consistently defined in
5496 *) Change ASN1_GENERALIZEDTIME_check() to allow fractional seconds.
5499 *) Change mkdef.pl to sort symbols that get the same entry number,
5500 and make sure the automatically generated functions ERR_load_*
5501 become part of libeay.num as well.
5504 *) New function SSL_renegotiate_pending(). This returns true once
5505 renegotiation has been requested (either SSL_renegotiate() call
5506 or HelloRequest/ClientHello receveived from the peer) and becomes
5507 false once a handshake has been completed.
5508 (For servers, SSL_renegotiate() followed by SSL_do_handshake()
5509 sends a HelloRequest, but does not ensure that a handshake takes
5510 place. SSL_renegotiate_pending() is useful for checking if the
5511 client has followed the request.)
5514 *) New SSL option SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION.
5515 By default, clients may request session resumption even during
5516 renegotiation (if session ID contexts permit); with this option,
5517 session resumption is possible only in the first handshake.
5519 SSL_OP_ALL is now 0x00000FFFL instead of 0x000FFFFFL. This makes
5520 more bits available for options that should not be part of
5521 SSL_OP_ALL (such as SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION).
5524 *) Add some demos for certificate and certificate request creation.
5527 *) Make maximum certificate chain size accepted from the peer application
5528 settable (SSL*_get/set_max_cert_list()), as proposed by
5529 "Douglas E. Engert" <deengert@anl.gov>.
5532 *) Add support for shared libraries for Unixware-7
5533 (Boyd Lynn Gerber <gerberb@zenez.com>).
5536 *) Add a "destroy" handler to ENGINEs that allows structural cleanup to
5537 be done prior to destruction. Use this to unload error strings from
5538 ENGINEs that load their own error strings. NB: This adds two new API
5539 functions to "get" and "set" this destroy handler in an ENGINE.
5542 *) Alter all existing ENGINE implementations (except "openssl" and
5543 "openbsd") to dynamically instantiate their own error strings. This
5544 makes them more flexible to be built both as statically-linked ENGINEs
5545 and self-contained shared-libraries loadable via the "dynamic" ENGINE.
5546 Also, add stub code to each that makes building them as self-contained
5547 shared-libraries easier (see README.ENGINE).
5550 *) Add a "dynamic" ENGINE that provides a mechanism for binding ENGINE
5551 implementations into applications that are completely implemented in
5552 self-contained shared-libraries. The "dynamic" ENGINE exposes control
5553 commands that can be used to configure what shared-library to load and
5554 to control aspects of the way it is handled. Also, made an update to
5555 the README.ENGINE file that brings its information up-to-date and
5556 provides some information and instructions on the "dynamic" ENGINE
5557 (ie. how to use it, how to build "dynamic"-loadable ENGINEs, etc).
5560 *) Make it possible to unload ranges of ERR strings with a new
5561 "ERR_unload_strings" function.
5564 *) Add a copy() function to EVP_MD.
5567 *) Make EVP_MD routines take a context pointer instead of just the
5568 md_data void pointer.
5571 *) Add flags to EVP_MD and EVP_MD_CTX. EVP_MD_FLAG_ONESHOT indicates
5572 that the digest can only process a single chunk of data
5573 (typically because it is provided by a piece of
5574 hardware). EVP_MD_CTX_FLAG_ONESHOT indicates that the application
5575 is only going to provide a single chunk of data, and hence the
5576 framework needn't accumulate the data for oneshot drivers.
5579 *) As with "ERR", make it possible to replace the underlying "ex_data"
5580 functions. This change also alters the storage and management of global
5581 ex_data state - it's now all inside ex_data.c and all "class" code (eg.
5582 RSA, BIO, SSL_CTX, etc) no longer stores its own STACKS and per-class
5583 index counters. The API functions that use this state have been changed
5584 to take a "class_index" rather than pointers to the class's local STACK
5585 and counter, and there is now an API function to dynamically create new
5586 classes. This centralisation allows us to (a) plug a lot of the
5587 thread-safety problems that existed, and (b) makes it possible to clean
5588 up all allocated state using "CRYPTO_cleanup_all_ex_data()". W.r.t. (b)
5589 such data would previously have always leaked in application code and
5590 workarounds were in place to make the memory debugging turn a blind eye
5591 to it. Application code that doesn't use this new function will still
5592 leak as before, but their memory debugging output will announce it now
5593 rather than letting it slide.
5595 Besides the addition of CRYPTO_cleanup_all_ex_data(), another API change
5596 induced by the "ex_data" overhaul is that X509_STORE_CTX_init() now
5597 has a return value to indicate success or failure.
5600 *) Make it possible to replace the underlying "ERR" functions such that the
5601 global state (2 LHASH tables and 2 locks) is only used by the "default"
5602 implementation. This change also adds two functions to "get" and "set"
5603 the implementation prior to it being automatically set the first time
5604 any other ERR function takes place. Ie. an application can call "get",
5605 pass the return value to a module it has just loaded, and that module
5606 can call its own "set" function using that value. This means the
5607 module's "ERR" operations will use (and modify) the error state in the
5608 application and not in its own statically linked copy of OpenSSL code.
5611 *) Give DH, DSA, and RSA types their own "**_up_ref()" function to increment
5612 reference counts. This performs normal REF_PRINT/REF_CHECK macros on
5613 the operation, and provides a more encapsulated way for external code
5614 (crypto/evp/ and ssl/) to do this. Also changed the evp and ssl code
5615 to use these functions rather than manually incrementing the counts.
5617 Also rename "DSO_up()" function to more descriptive "DSO_up_ref()".
5620 *) Add EVP test program.
5623 *) Add symmetric cipher support to ENGINE. Expect the API to change!
5626 *) New CRL functions: X509_CRL_set_version(), X509_CRL_set_issuer_name()
5627 X509_CRL_set_lastUpdate(), X509_CRL_set_nextUpdate(), X509_CRL_sort(),
5628 X509_REVOKED_set_serialNumber(), and X509_REVOKED_set_revocationDate().
5629 These allow a CRL to be built without having to access X509_CRL fields
5630 directly. Modify 'ca' application to use new functions.
5633 *) Move SSL_OP_TLS_ROLLBACK_BUG out of the SSL_OP_ALL list of recommended
5634 bug workarounds. Rollback attack detection is a security feature.
5635 The problem will only arise on OpenSSL servers when TLSv1 is not
5636 available (sslv3_server_method() or SSL_OP_NO_TLSv1).
5637 Software authors not wanting to support TLSv1 will have special reasons
5638 for their choice and can explicitly enable this option.
5639 [Bodo Moeller, Lutz Jaenicke]
5641 *) Rationalise EVP so it can be extended: don't include a union of
5642 cipher/digest structures, add init/cleanup functions for EVP_MD_CTX
5643 (similar to those existing for EVP_CIPHER_CTX).
5648 EVP_MD_CTX_init(&md); /* new function call */
5649 EVP_DigestInit(&md, EVP_sha1());
5650 EVP_DigestUpdate(&md, in, len);
5651 EVP_DigestFinal(&md, out, NULL);
5652 EVP_MD_CTX_cleanup(&md); /* new function call */
5656 *) Make DES key schedule conform to the usual scheme, as well as
5657 correcting its structure. This means that calls to DES functions
5658 now have to pass a pointer to a des_key_schedule instead of a
5659 plain des_key_schedule (which was actually always a pointer
5662 des_key_schedule ks;
5664 des_set_key_checked(..., &ks);
5665 des_ncbc_encrypt(..., &ks, ...);
5667 (Note that a later change renames 'des_...' into 'DES_...'.)
5670 *) Initial reduction of linker bloat: the use of some functions, such as
5671 PEM causes large amounts of unused functions to be linked in due to
5672 poor organisation. For example pem_all.c contains every PEM function
5673 which has a knock on effect of linking in large amounts of (unused)
5674 ASN1 code. Grouping together similar functions and splitting unrelated
5675 functions prevents this.
5678 *) Cleanup of EVP macros.
5681 *) Change historical references to {NID,SN,LN}_des_ede and ede3 to add the
5682 correct _ecb suffix.
5685 *) Add initial OCSP responder support to ocsp application. The
5686 revocation information is handled using the text based index
5687 use by the ca application. The responder can either handle
5688 requests generated internally, supplied in files (for example
5689 via a CGI script) or using an internal minimal server.
5692 *) Add configuration choices to get zlib compression for TLS.
5695 *) Changes to Kerberos SSL for RFC 2712 compliance:
5696 1. Implemented real KerberosWrapper, instead of just using
5697 KRB5 AP_REQ message. [Thanks to Simon Wilkinson <sxw@sxw.org.uk>]
5698 2. Implemented optional authenticator field of KerberosWrapper.
5700 Added openssl-style ASN.1 macros for Kerberos ticket, ap_req,
5701 and authenticator structs; see crypto/krb5/.
5703 Generalized Kerberos calls to support multiple Kerberos libraries.
5704 [Vern Staats <staatsvr@asc.hpc.mil>,
5705 Jeffrey Altman <jaltman@columbia.edu>
5706 via Richard Levitte]
5708 *) Cause 'openssl speed' to use fully hard-coded DSA keys as it
5709 already does with RSA. testdsa.h now has 'priv_key/pub_key'
5710 values for each of the key sizes rather than having just
5711 parameters (and 'speed' generating keys each time).
5714 *) Speed up EVP routines.
5717 type 8 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes
5718 des-cbc 4408.85k 5560.51k 5778.46k 5862.20k 5825.16k
5719 des-cbc 4389.55k 5571.17k 5792.23k 5846.91k 5832.11k
5720 des-cbc 4394.32k 5575.92k 5807.44k 5848.37k 5841.30k
5722 des-cbc 3482.66k 5069.49k 5496.39k 5614.16k 5639.28k
5723 des-cbc 3480.74k 5068.76k 5510.34k 5609.87k 5635.52k
5724 des-cbc 3483.72k 5067.62k 5504.60k 5708.01k 5724.80k
5727 des-cbc 4660.16k 5650.19k 5807.19k 5827.13k 5783.32k
5729 des-cbc 3624.96k 5258.21k 5530.91k 5624.30k 5628.26k
5732 *) Added the OS2-EMX target.
5733 ["Brian Havard" <brianh@kheldar.apana.org.au> and Richard Levitte]
5735 *) Rewrite apps to use NCONF routines instead of the old CONF. New functions
5736 to support NCONF routines in extension code. New function CONF_set_nconf()
5737 to allow functions which take an NCONF to also handle the old LHASH
5738 structure: this means that the old CONF compatible routines can be
5739 retained (in particular wrt extensions) without having to duplicate the
5740 code. New function X509V3_add_ext_nconf_sk to add extensions to a stack.
5743 *) Enhance the general user interface with mechanisms for inner control
5744 and with possibilities to have yes/no kind of prompts.
5747 *) Change all calls to low level digest routines in the library and
5748 applications to use EVP. Add missing calls to HMAC_cleanup() and
5749 don't assume HMAC_CTX can be copied using memcpy().
5750 [Verdon Walker <VWalker@novell.com>, Steve Henson]
5752 *) Add the possibility to control engines through control names but with
5753 arbitrary arguments instead of just a string.
5754 Change the key loaders to take a UI_METHOD instead of a callback
5755 function pointer. NOTE: this breaks binary compatibility with earlier
5756 versions of OpenSSL [engine].
5757 Adapt the nCipher code for these new conditions and add a card insertion
5761 *) Enhance the general user interface with mechanisms to better support
5762 dialog box interfaces, application-defined prompts, the possibility
5763 to use defaults (for example default passwords from somewhere else)
5764 and interrupts/cancellations.
5767 *) Tidy up PKCS#12 attribute handling. Add support for the CSP name
5768 attribute in PKCS#12 files, add new -CSP option to pkcs12 utility.
5771 *) Fix a memory leak in 'sk_dup()' in the case reallocation fails. (Also
5772 tidy up some unnecessarily weird code in 'sk_new()').
5773 [Geoff, reported by Diego Tartara <dtartara@novamens.com>]
5775 *) Change the key loading routines for ENGINEs to use the same kind
5776 callback (pem_password_cb) as all other routines that need this
5780 *) Increase ENTROPY_NEEDED to 32 bytes, as Rijndael can operate with
5781 256 bit (=32 byte) keys. Of course seeding with more entropy bytes
5782 than this minimum value is recommended.
5785 *) New random seeder for OpenVMS, using the system process statistics
5786 that are easily reachable.
5789 *) Windows apparently can't transparently handle global
5790 variables defined in DLLs. Initialisations such as:
5792 const ASN1_ITEM *it = &ASN1_INTEGER_it;
5794 wont compile. This is used by the any applications that need to
5795 declare their own ASN1 modules. This was fixed by adding the option
5796 EXPORT_VAR_AS_FN to all Win32 platforms, although this isn't strictly
5797 needed for static libraries under Win32.
5800 *) New functions X509_PURPOSE_set() and X509_TRUST_set() to handle
5801 setting of purpose and trust fields. New X509_STORE trust and
5802 purpose functions and tidy up setting in other SSL functions.
5805 *) Add copies of X509_STORE_CTX fields and callbacks to X509_STORE
5806 structure. These are inherited by X509_STORE_CTX when it is
5807 initialised. This allows various defaults to be set in the
5808 X509_STORE structure (such as flags for CRL checking and custom
5809 purpose or trust settings) for functions which only use X509_STORE_CTX
5810 internally such as S/MIME.
5812 Modify X509_STORE_CTX_purpose_inherit() so it only sets purposes and
5813 trust settings if they are not set in X509_STORE. This allows X509_STORE
5814 purposes and trust (in S/MIME for example) to override any set by default.
5816 Add command line options for CRL checking to smime, s_client and s_server
5820 *) Initial CRL based revocation checking. If the CRL checking flag(s)
5821 are set then the CRL is looked up in the X509_STORE structure and
5822 its validity and signature checked, then if the certificate is found
5823 in the CRL the verify fails with a revoked error.
5825 Various new CRL related callbacks added to X509_STORE_CTX structure.
5827 Command line options added to 'verify' application to support this.
5829 This needs some additional work, such as being able to handle multiple
5830 CRLs with different times, extension based lookup (rather than just
5831 by subject name) and ultimately more complete V2 CRL extension
5835 *) Add a general user interface API (crypto/ui/). This is designed
5836 to replace things like des_read_password and friends (backward
5837 compatibility functions using this new API are provided).
5838 The purpose is to remove prompting functions from the DES code
5839 section as well as provide for prompting through dialog boxes in
5840 a window system and the like.
5843 *) Add "ex_data" support to ENGINE so implementations can add state at a
5844 per-structure level rather than having to store it globally.
5847 *) Make it possible for ENGINE structures to be copied when retrieved by
5848 ENGINE_by_id() if the ENGINE specifies a new flag: ENGINE_FLAGS_BY_ID_COPY.
5849 This causes the "original" ENGINE structure to act like a template,
5850 analogous to the RSA vs. RSA_METHOD type of separation. Because of this
5851 operational state can be localised to each ENGINE structure, despite the
5852 fact they all share the same "methods". New ENGINE structures returned in
5853 this case have no functional references and the return value is the single
5854 structural reference. This matches the single structural reference returned
5855 by ENGINE_by_id() normally, when it is incremented on the pre-existing
5859 *) Fix ASN1 decoder when decoding type ANY and V_ASN1_OTHER: since this
5860 needs to match any other type at all we need to manually clear the
5864 *) Changes to the "openssl engine" utility to include;
5865 - verbosity levels ('-v', '-vv', and '-vvv') that provide information
5866 about an ENGINE's available control commands.
5867 - executing control commands from command line arguments using the
5868 '-pre' and '-post' switches. '-post' is only used if '-t' is
5869 specified and the ENGINE is successfully initialised. The syntax for
5870 the individual commands are colon-separated, for example;
5871 openssl engine chil -pre FORK_CHECK:0 -pre SO_PATH:/lib/test.so
5874 *) New dynamic control command support for ENGINEs. ENGINEs can now
5875 declare their own commands (numbers), names (strings), descriptions,
5876 and input types for run-time discovery by calling applications. A
5877 subset of these commands are implicitly classed as "executable"
5878 depending on their input type, and only these can be invoked through
5879 the new string-based API function ENGINE_ctrl_cmd_string(). (Eg. this
5880 can be based on user input, config files, etc). The distinction is
5881 that "executable" commands cannot return anything other than a boolean
5882 result and can only support numeric or string input, whereas some
5883 discoverable commands may only be for direct use through
5884 ENGINE_ctrl(), eg. supporting the exchange of binary data, function
5885 pointers, or other custom uses. The "executable" commands are to
5886 support parameterisations of ENGINE behaviour that can be
5887 unambiguously defined by ENGINEs and used consistently across any
5888 OpenSSL-based application. Commands have been added to all the
5889 existing hardware-supporting ENGINEs, noticeably "SO_PATH" to allow
5890 control over shared-library paths without source code alterations.
5893 *) Changed all ENGINE implementations to dynamically allocate their
5894 ENGINEs rather than declaring them statically. Apart from this being
5895 necessary with the removal of the ENGINE_FLAGS_MALLOCED distinction,
5896 this also allows the implementations to compile without using the
5897 internal engine_int.h header.
5900 *) Minor adjustment to "rand" code. RAND_get_rand_method() now returns a
5901 'const' value. Any code that should be able to modify a RAND_METHOD
5902 should already have non-const pointers to it (ie. they should only
5903 modify their own ones).
5906 *) Made a variety of little tweaks to the ENGINE code.
5907 - "atalla" and "ubsec" string definitions were moved from header files
5908 to C code. "nuron" string definitions were placed in variables
5909 rather than hard-coded - allowing parameterisation of these values
5910 later on via ctrl() commands.
5911 - Removed unused "#if 0"'d code.
5912 - Fixed engine list iteration code so it uses ENGINE_free() to release
5913 structural references.
5914 - Constified the RAND_METHOD element of ENGINE structures.
5915 - Constified various get/set functions as appropriate and added
5916 missing functions (including a catch-all ENGINE_cpy that duplicates
5917 all ENGINE values onto a new ENGINE except reference counts/state).
5918 - Removed NULL parameter checks in get/set functions. Setting a method
5919 or function to NULL is a way of cancelling out a previously set
5920 value. Passing a NULL ENGINE parameter is just plain stupid anyway
5921 and doesn't justify the extra error symbols and code.
5922 - Deprecate the ENGINE_FLAGS_MALLOCED define and move the area for
5923 flags from engine_int.h to engine.h.
5924 - Changed prototypes for ENGINE handler functions (init(), finish(),
5925 ctrl(), key-load functions, etc) to take an (ENGINE*) parameter.
5928 *) Implement binary inversion algorithm for BN_mod_inverse in addition
5929 to the algorithm using long division. The binary algorithm can be
5930 used only if the modulus is odd. On 32-bit systems, it is faster
5931 only for relatively small moduli (roughly 20-30% for 128-bit moduli,
5932 roughly 5-15% for 256-bit moduli), so we use it only for moduli
5933 up to 450 bits. In 64-bit environments, the binary algorithm
5934 appears to be advantageous for much longer moduli; here we use it
5935 for moduli up to 2048 bits.
5938 *) Rewrite CHOICE field setting in ASN1_item_ex_d2i(). The old code
5939 could not support the combine flag in choice fields.
5942 *) Add a 'copy_extensions' option to the 'ca' utility. This copies
5943 extensions from a certificate request to the certificate.
5946 *) Allow multiple 'certopt' and 'nameopt' options to be separated
5947 by commas. Add 'namopt' and 'certopt' options to the 'ca' config
5948 file: this allows the display of the certificate about to be
5949 signed to be customised, to allow certain fields to be included
5950 or excluded and extension details. The old system didn't display
5951 multicharacter strings properly, omitted fields not in the policy
5952 and couldn't display additional details such as extensions.
5955 *) Function EC_POINTs_mul for multiple scalar multiplication
5956 of an arbitrary number of elliptic curve points
5957 \sum scalars[i]*points[i],
5958 optionally including the generator defined for the EC_GROUP:
5959 scalar*generator + \sum scalars[i]*points[i].
5961 EC_POINT_mul is a simple wrapper function for the typical case
5962 that the point list has just one item (besides the optional
5966 *) First EC_METHODs for curves over GF(p):
5968 EC_GFp_simple_method() uses the basic BN_mod_mul and BN_mod_sqr
5969 operations and provides various method functions that can also
5970 operate with faster implementations of modular arithmetic.
5972 EC_GFp_mont_method() reuses most functions that are part of
5973 EC_GFp_simple_method, but uses Montgomery arithmetic.
5975 [Bodo Moeller; point addition and point doubling
5976 implementation directly derived from source code provided by
5977 Lenka Fibikova <fibikova@exp-math.uni-essen.de>]
5979 *) Framework for elliptic curves (crypto/ec/ec.h, crypto/ec/ec_lcl.h,
5980 crypto/ec/ec_lib.c):
5982 Curves are EC_GROUP objects (with an optional group generator)
5983 based on EC_METHODs that are built into the library.
5985 Points are EC_POINT objects based on EC_GROUP objects.
5987 Most of the framework would be able to handle curves over arbitrary
5988 finite fields, but as there are no obvious types for fields other
5989 than GF(p), some functions are limited to that for now.
5992 *) Add the -HTTP option to s_server. It is similar to -WWW, but requires
5993 that the file contains a complete HTTP response.
5996 *) Add the ec directory to mkdef.pl and mkfiles.pl. In mkdef.pl
5997 change the def and num file printf format specifier from "%-40sXXX"
5998 to "%-39s XXX". The latter will always guarantee a space after the
5999 field while the former will cause them to run together if the field
6000 is 40 of more characters long.
6003 *) Constify the cipher and digest 'method' functions and structures
6004 and modify related functions to take constant EVP_MD and EVP_CIPHER
6008 *) Hide BN_CTX structure details in bn_lcl.h instead of publishing them
6009 in <openssl/bn.h>. Also further increase BN_CTX_NUM to 32.
6012 *) Modify EVP_Digest*() routines so they now return values. Although the
6013 internal software routines can never fail additional hardware versions
6017 *) Clean up crypto/err/err.h and change some error codes to avoid conflicts:
6019 Previously ERR_R_FATAL was too small and coincided with ERR_LIB_PKCS7
6020 (= ERR_R_PKCS7_LIB); it is now 64 instead of 32.
6023 ERR_R_NESTED_ASN1_ERROR
6025 ERR_R_MISSING_ASN1_EOS
6026 were 4 .. 9, conflicting with
6027 ERR_LIB_RSA (= ERR_R_RSA_LIB)
6029 ERR_LIB_PEM (= ERR_R_PEM_LIB).
6030 They are now 58 .. 63 (i.e., just below ERR_R_FATAL).
6032 Add new error code 'ERR_R_INTERNAL_ERROR'.
6035 *) Don't overuse locks in crypto/err/err.c: For data retrieval, CRYPTO_r_lock
6039 *) New option '-subj arg' for 'openssl req' and 'openssl ca'. This
6040 sets the subject name for a new request or supersedes the
6041 subject name in a given request. Formats that can be parsed are
6042 'CN=Some Name, OU=myOU, C=IT'
6044 'CN=Some Name/OU=myOU/C=IT'.
6046 Add options '-batch' and '-verbose' to 'openssl req'.
6047 [Massimiliano Pala <madwolf@hackmasters.net>]
6049 *) Introduce the possibility to access global variables through
6050 functions on platform were that's the best way to handle exporting
6051 global variables in shared libraries. To enable this functionality,
6052 one must configure with "EXPORT_VAR_AS_FN" or defined the C macro
6053 "OPENSSL_EXPORT_VAR_AS_FUNCTION" in crypto/opensslconf.h (the latter
6054 is normally done by Configure or something similar).
6056 To implement a global variable, use the macro OPENSSL_IMPLEMENT_GLOBAL
6057 in the source file (foo.c) like this:
6059 OPENSSL_IMPLEMENT_GLOBAL(int,foo)=1;
6060 OPENSSL_IMPLEMENT_GLOBAL(double,bar);
6062 To declare a global variable, use the macros OPENSSL_DECLARE_GLOBAL
6063 and OPENSSL_GLOBAL_REF in the header file (foo.h) like this:
6065 OPENSSL_DECLARE_GLOBAL(int,foo);
6066 #define foo OPENSSL_GLOBAL_REF(foo)
6067 OPENSSL_DECLARE_GLOBAL(double,bar);
6068 #define bar OPENSSL_GLOBAL_REF(bar)
6070 The #defines are very important, and therefore so is including the
6071 header file everywhere where the defined globals are used.
6073 The macro OPENSSL_EXPORT_VAR_AS_FUNCTION also affects the definition
6074 of ASN.1 items, but that structure is a bit different.
6076 The largest change is in util/mkdef.pl which has been enhanced with
6077 better and easier to understand logic to choose which symbols should
6078 go into the Windows .def files as well as a number of fixes and code
6079 cleanup (among others, algorithm keywords are now sorted
6080 lexicographically to avoid constant rewrites).
6083 *) In BN_div() keep a copy of the sign of 'num' before writing the
6084 result to 'rm' because if rm==num the value will be overwritten
6085 and produce the wrong result if 'num' is negative: this caused
6086 problems with BN_mod() and BN_nnmod().
6089 *) Function OCSP_request_verify(). This checks the signature on an
6090 OCSP request and verifies the signer certificate. The signer
6091 certificate is just checked for a generic purpose and OCSP request
6095 *) Add OCSP_check_validity() function to check the validity of OCSP
6096 responses. OCSP responses are prepared in real time and may only
6097 be a few seconds old. Simply checking that the current time lies
6098 between thisUpdate and nextUpdate max reject otherwise valid responses
6099 caused by either OCSP responder or client clock inaccuracy. Instead
6100 we allow thisUpdate and nextUpdate to fall within a certain period of
6101 the current time. The age of the response can also optionally be
6102 checked. Two new options -validity_period and -status_age added to
6106 *) If signature or public key algorithm is unrecognized print out its
6107 OID rather that just UNKNOWN.
6110 *) Change OCSP_cert_to_id() to tolerate a NULL subject certificate and
6111 OCSP_cert_id_new() a NULL serialNumber. This allows a partial certificate
6112 ID to be generated from the issuer certificate alone which can then be
6113 passed to OCSP_id_issuer_cmp().
6116 *) New compilation option ASN1_ITEM_FUNCTIONS. This causes the new
6117 ASN1 modules to export functions returning ASN1_ITEM pointers
6118 instead of the ASN1_ITEM structures themselves. This adds several
6119 new macros which allow the underlying ASN1 function/structure to
6120 be accessed transparently. As a result code should not use ASN1_ITEM
6121 references directly (such as &X509_it) but instead use the relevant
6122 macros (such as ASN1_ITEM_rptr(X509)). This option is to allow
6123 use of the new ASN1 code on platforms where exporting structures
6124 is problematical (for example in shared libraries) but exporting
6125 functions returning pointers to structures is not.
6128 *) Add support for overriding the generation of SSL/TLS session IDs.
6129 These callbacks can be registered either in an SSL_CTX or per SSL.
6130 The purpose of this is to allow applications to control, if they wish,
6131 the arbitrary values chosen for use as session IDs, particularly as it
6132 can be useful for session caching in multiple-server environments. A
6133 command-line switch for testing this (and any client code that wishes
6134 to use such a feature) has been added to "s_server".
6135 [Geoff Thorpe, Lutz Jaenicke]
6137 *) Modify mkdef.pl to recognise and parse preprocessor conditionals
6138 of the form '#if defined(...) || defined(...) || ...' and
6139 '#if !defined(...) && !defined(...) && ...'. This also avoids
6140 the growing number of special cases it was previously handling.
6143 *) Make all configuration macros available for application by making
6144 sure they are available in opensslconf.h, by giving them names starting
6145 with "OPENSSL_" to avoid conflicts with other packages and by making
6146 sure e_os2.h will cover all platform-specific cases together with
6148 Additionally, it is now possible to define configuration/platform-
6149 specific names (called "system identities"). In the C code, these
6150 are prefixed with "OPENSSL_SYSNAME_". e_os2.h will create another
6151 macro with the name beginning with "OPENSSL_SYS_", which is determined
6152 from "OPENSSL_SYSNAME_*" or compiler-specific macros depending on
6156 *) New option -set_serial to 'req' and 'x509' this allows the serial
6157 number to use to be specified on the command line. Previously self
6158 signed certificates were hard coded with serial number 0 and the
6159 CA options of 'x509' had to use a serial number in a file which was
6163 *) New options to 'ca' utility to support V2 CRL entry extensions.
6164 Currently CRL reason, invalidity date and hold instruction are
6165 supported. Add new CRL extensions to V3 code and some new objects.
6168 *) New function EVP_CIPHER_CTX_set_padding() this is used to
6169 disable standard block padding (aka PKCS#5 padding) in the EVP
6170 API, which was previously mandatory. This means that the data is
6171 not padded in any way and so the total length much be a multiple
6172 of the block size, otherwise an error occurs.
6175 *) Initial (incomplete) OCSP SSL support.
6178 *) New function OCSP_parse_url(). This splits up a URL into its host,
6179 port and path components: primarily to parse OCSP URLs. New -url
6180 option to ocsp utility.
6183 *) New nonce behavior. The return value of OCSP_check_nonce() now
6184 reflects the various checks performed. Applications can decide
6185 whether to tolerate certain situations such as an absent nonce
6186 in a response when one was present in a request: the ocsp application
6187 just prints out a warning. New function OCSP_add1_basic_nonce()
6188 this is to allow responders to include a nonce in a response even if
6189 the request is nonce-less.
6192 *) Disable stdin buffering in load_cert (apps/apps.c) so that no certs are
6193 skipped when using openssl x509 multiple times on a single input file,
6194 e.g. "(openssl x509 -out cert1; openssl x509 -out cert2) <certs".
6197 *) Make ASN1_UTCTIME_set_string() and ASN1_GENERALIZEDTIME_set_string()
6198 set string type: to handle setting ASN1_TIME structures. Fix ca
6199 utility to correctly initialize revocation date of CRLs.
6202 *) New option SSL_OP_CIPHER_SERVER_PREFERENCE allows the server to override
6203 the clients preferred ciphersuites and rather use its own preferences.
6204 Should help to work around M$ SGC (Server Gated Cryptography) bug in
6205 Internet Explorer by ensuring unchanged hash method during stepup.
6206 (Also replaces the broken/deactivated SSL_OP_NON_EXPORT_FIRST option.)
6209 *) Make mkdef.pl recognise all DECLARE_ASN1 macros, change rijndael
6210 to aes and add a new 'exist' option to print out symbols that don't
6214 *) Additional options to ocsp utility to allow flags to be set and
6215 additional certificates supplied.
6218 *) Add the option -VAfile to 'openssl ocsp', so the user can give the
6219 OCSP client a number of certificate to only verify the response
6223 *) Update Rijndael code to version 3.0 and change EVP AES ciphers to
6224 handle the new API. Currently only ECB, CBC modes supported. Add new
6227 Add TLS AES ciphersuites as described in RFC3268, "Advanced
6228 Encryption Standard (AES) Ciphersuites for Transport Layer
6229 Security (TLS)". (In beta versions of OpenSSL 0.9.7, these were
6230 not enabled by default and were not part of the "ALL" ciphersuite
6231 alias because they were not yet official; they could be
6232 explicitly requested by specifying the "AESdraft" ciphersuite
6233 group alias. In the final release of OpenSSL 0.9.7, the group
6234 alias is called "AES" and is part of "ALL".)
6235 [Ben Laurie, Steve Henson, Bodo Moeller]
6237 *) New function OCSP_copy_nonce() to copy nonce value (if present) from
6238 request to response.
6241 *) Functions for OCSP responders. OCSP_request_onereq_count(),
6242 OCSP_request_onereq_get0(), OCSP_onereq_get0_id() and OCSP_id_get0_info()
6243 extract information from a certificate request. OCSP_response_create()
6244 creates a response and optionally adds a basic response structure.
6245 OCSP_basic_add1_status() adds a complete single response to a basic
6246 response and returns the OCSP_SINGLERESP structure just added (to allow
6247 extensions to be included for example). OCSP_basic_add1_cert() adds a
6248 certificate to a basic response and OCSP_basic_sign() signs a basic
6249 response with various flags. New helper functions ASN1_TIME_check()
6250 (checks validity of ASN1_TIME structure) and ASN1_TIME_to_generalizedtime()
6251 (converts ASN1_TIME to GeneralizedTime).
6254 *) Various new functions. EVP_Digest() combines EVP_Digest{Init,Update,Final}()
6255 in a single operation. X509_get0_pubkey_bitstr() extracts the public_key
6256 structure from a certificate. X509_pubkey_digest() digests the public_key
6257 contents: this is used in various key identifiers.
6260 *) Make sk_sort() tolerate a NULL argument.
6261 [Steve Henson reported by Massimiliano Pala <madwolf@comune.modena.it>]
6263 *) New OCSP verify flag OCSP_TRUSTOTHER. When set the "other" certificates
6264 passed by the function are trusted implicitly. If any of them signed the
6265 response then it is assumed to be valid and is not verified.
6268 *) In PKCS7_set_type() initialise content_type in PKCS7_ENC_CONTENT
6269 to data. This was previously part of the PKCS7 ASN1 code. This
6270 was causing problems with OpenSSL created PKCS#12 and PKCS#7 structures.
6271 [Steve Henson, reported by Kenneth R. Robinette
6272 <support@securenetterm.com>]
6274 *) Add CRYPTO_push_info() and CRYPTO_pop_info() calls to new ASN1
6275 routines: without these tracing memory leaks is very painful.
6276 Fix leaks in PKCS12 and PKCS7 routines.
6279 *) Make X509_time_adj() cope with the new behaviour of ASN1_TIME_new().
6280 Previously it initialised the 'type' argument to V_ASN1_UTCTIME which
6281 effectively meant GeneralizedTime would never be used. Now it
6282 is initialised to -1 but X509_time_adj() now has to check the value
6283 and use ASN1_TIME_set() if the value is not V_ASN1_UTCTIME or
6284 V_ASN1_GENERALIZEDTIME, without this it always uses GeneralizedTime.
6285 [Steve Henson, reported by Kenneth R. Robinette
6286 <support@securenetterm.com>]
6288 *) Fixes to BN_to_ASN1_INTEGER when bn is zero. This would previously
6289 result in a zero length in the ASN1_INTEGER structure which was
6290 not consistent with the structure when d2i_ASN1_INTEGER() was used
6291 and would cause ASN1_INTEGER_cmp() to fail. Enhance s2i_ASN1_INTEGER()
6292 to cope with hex and negative integers. Fix bug in i2a_ASN1_INTEGER()
6293 where it did not print out a minus for negative ASN1_INTEGER.
6296 *) Add summary printout to ocsp utility. The various functions which
6297 convert status values to strings have been renamed to:
6298 OCSP_response_status_str(), OCSP_cert_status_str() and
6299 OCSP_crl_reason_str() and are no longer static. New options
6300 to verify nonce values and to disable verification. OCSP response
6301 printout format cleaned up.
6304 *) Add additional OCSP certificate checks. These are those specified
6305 in RFC2560. This consists of two separate checks: the CA of the
6306 certificate being checked must either be the OCSP signer certificate
6307 or the issuer of the OCSP signer certificate. In the latter case the
6308 OCSP signer certificate must contain the OCSP signing extended key
6309 usage. This check is performed by attempting to match the OCSP
6310 signer or the OCSP signer CA to the issuerNameHash and issuerKeyHash
6311 in the OCSP_CERTID structures of the response.
6314 *) Initial OCSP certificate verification added to OCSP_basic_verify()
6315 and related routines. This uses the standard OpenSSL certificate
6316 verify routines to perform initial checks (just CA validity) and
6317 to obtain the certificate chain. Then additional checks will be
6318 performed on the chain. Currently the root CA is checked to see
6319 if it is explicitly trusted for OCSP signing. This is used to set
6320 a root CA as a global signing root: that is any certificate that
6321 chains to that CA is an acceptable OCSP signing certificate.
6324 *) New '-extfile ...' option to 'openssl ca' for reading X.509v3
6325 extensions from a separate configuration file.
6326 As when reading extensions from the main configuration file,
6327 the '-extensions ...' option may be used for specifying the
6329 [Massimiliano Pala <madwolf@comune.modena.it>]
6331 *) New OCSP utility. Allows OCSP requests to be generated or
6332 read. The request can be sent to a responder and the output
6333 parsed, outputed or printed in text form. Not complete yet:
6334 still needs to check the OCSP response validity.
6337 *) New subcommands for 'openssl ca':
6338 'openssl ca -status <serial>' prints the status of the cert with
6339 the given serial number (according to the index file).
6340 'openssl ca -updatedb' updates the expiry status of certificates
6342 [Massimiliano Pala <madwolf@comune.modena.it>]
6344 *) New '-newreq-nodes' command option to CA.pl. This is like
6345 '-newreq', but calls 'openssl req' with the '-nodes' option
6346 so that the resulting key is not encrypted.
6347 [Damien Miller <djm@mindrot.org>]
6349 *) New configuration for the GNU Hurd.
6350 [Jonathan Bartlett <johnnyb@wolfram.com> via Richard Levitte]
6352 *) Initial code to implement OCSP basic response verify. This
6353 is currently incomplete. Currently just finds the signer's
6354 certificate and verifies the signature on the response.
6357 *) New SSLeay_version code SSLEAY_DIR to determine the compiled-in
6358 value of OPENSSLDIR. This is available via the new '-d' option
6359 to 'openssl version', and is also included in 'openssl version -a'.
6362 *) Allowing defining memory allocation callbacks that will be given
6363 file name and line number information in additional arguments
6364 (a const char* and an int). The basic functionality remains, as
6365 well as the original possibility to just replace malloc(),
6366 realloc() and free() by functions that do not know about these
6367 additional arguments. To register and find out the current
6368 settings for extended allocation functions, the following
6369 functions are provided:
6371 CRYPTO_set_mem_ex_functions
6372 CRYPTO_set_locked_mem_ex_functions
6373 CRYPTO_get_mem_ex_functions
6374 CRYPTO_get_locked_mem_ex_functions
6376 These work the same way as CRYPTO_set_mem_functions and friends.
6377 CRYPTO_get_[locked_]mem_functions now writes 0 where such an
6378 extended allocation function is enabled.
6379 Similarly, CRYPTO_get_[locked_]mem_ex_functions writes 0 where
6380 a conventional allocation function is enabled.
6381 [Richard Levitte, Bodo Moeller]
6383 *) Finish off removing the remaining LHASH function pointer casts.
6384 There should no longer be any prototype-casting required when using
6385 the LHASH abstraction, and any casts that remain are "bugs". See
6386 the callback types and macros at the head of lhash.h for details
6387 (and "OBJ_cleanup" in crypto/objects/obj_dat.c as an example).
6390 *) Add automatic query of EGD sockets in RAND_poll() for the unix variant.
6391 If /dev/[u]random devices are not available or do not return enough
6392 entropy, EGD style sockets (served by EGD or PRNGD) will automatically
6394 The locations /var/run/egd-pool, /dev/egd-pool, /etc/egd-pool, and
6395 /etc/entropy will be queried once each in this sequence, quering stops
6396 when enough entropy was collected without querying more sockets.
6399 *) Change the Unix RAND_poll() variant to be able to poll several
6400 random devices, as specified by DEVRANDOM, until a sufficient amount
6401 of data has been collected. We spend at most 10 ms on each file
6402 (select timeout) and read in non-blocking mode. DEVRANDOM now
6403 defaults to the list "/dev/urandom", "/dev/random", "/dev/srandom"
6404 (previously it was just the string "/dev/urandom"), so on typical
6405 platforms the 10 ms delay will never occur.
6406 Also separate out the Unix variant to its own file, rand_unix.c.
6407 For VMS, there's a currently-empty rand_vms.c.
6410 *) Move OCSP client related routines to ocsp_cl.c. These
6411 provide utility functions which an application needing
6412 to issue a request to an OCSP responder and analyse the
6413 response will typically need: as opposed to those which an
6414 OCSP responder itself would need which will be added later.
6416 OCSP_request_sign() signs an OCSP request with an API similar
6417 to PKCS7_sign(). OCSP_response_status() returns status of OCSP
6418 response. OCSP_response_get1_basic() extracts basic response
6419 from response. OCSP_resp_find_status(): finds and extracts status
6420 information from an OCSP_CERTID structure (which will be created
6421 when the request structure is built). These are built from lower
6422 level functions which work on OCSP_SINGLERESP structures but
6423 wont normally be used unless the application wishes to examine
6424 extensions in the OCSP response for example.
6426 Replace nonce routines with a pair of functions.
6427 OCSP_request_add1_nonce() adds a nonce value and optionally
6428 generates a random value. OCSP_check_nonce() checks the
6429 validity of the nonce in an OCSP response.
6432 *) Change function OCSP_request_add() to OCSP_request_add0_id().
6433 This doesn't copy the supplied OCSP_CERTID and avoids the
6434 need to free up the newly created id. Change return type
6435 to OCSP_ONEREQ to return the internal OCSP_ONEREQ structure.
6436 This can then be used to add extensions to the request.
6437 Deleted OCSP_request_new(), since most of its functionality
6438 is now in OCSP_REQUEST_new() (and the case insensitive name
6439 clash) apart from the ability to set the request name which
6440 will be added elsewhere.
6443 *) Update OCSP API. Remove obsolete extensions argument from
6444 various functions. Extensions are now handled using the new
6445 OCSP extension code. New simple OCSP HTTP function which
6446 can be used to send requests and parse the response.
6449 *) Fix the PKCS#7 (S/MIME) code to work with new ASN1. Two new
6450 ASN1_ITEM structures help with sign and verify. PKCS7_ATTR_SIGN
6451 uses the special reorder version of SET OF to sort the attributes
6452 and reorder them to match the encoded order. This resolves a long
6453 standing problem: a verify on a PKCS7 structure just after signing
6454 it used to fail because the attribute order did not match the
6455 encoded order. PKCS7_ATTR_VERIFY does not reorder the attributes:
6456 it uses the received order. This is necessary to tolerate some broken
6457 software that does not order SET OF. This is handled by encoding
6458 as a SEQUENCE OF but using implicit tagging (with UNIVERSAL class)
6459 to produce the required SET OF.
6462 *) Have mk1mf.pl generate the macros OPENSSL_BUILD_SHLIBCRYPTO and
6463 OPENSSL_BUILD_SHLIBSSL and use them appropriately in the header
6464 files to get correct declarations of the ASN.1 item variables.
6467 *) Rewrite of PKCS#12 code to use new ASN1 functionality. Replace many
6468 PKCS#12 macros with real functions. Fix two unrelated ASN1 bugs:
6469 asn1_check_tlen() would sometimes attempt to use 'ctx' when it was
6470 NULL and ASN1_TYPE was not dereferenced properly in asn1_ex_c2i().
6471 New ASN1 macro: DECLARE_ASN1_ITEM() which just declares the relevant
6472 ASN1_ITEM and no wrapper functions.
6475 *) New functions or ASN1_item_d2i_fp() and ASN1_item_d2i_bio(). These
6476 replace the old function pointer based I/O routines. Change most of
6477 the *_d2i_bio() and *_d2i_fp() functions to use these.
6480 *) Enhance mkdef.pl to be more accepting about spacing in C preprocessor
6481 lines, recognice more "algorithms" that can be deselected, and make
6482 it complain about algorithm deselection that isn't recognised.
6485 *) New ASN1 functions to handle dup, sign, verify, digest, pack and
6486 unpack operations in terms of ASN1_ITEM. Modify existing wrappers
6487 to use new functions. Add NO_ASN1_OLD which can be set to remove
6488 some old style ASN1 functions: this can be used to determine if old
6489 code will still work when these eventually go away.
6492 *) New extension functions for OCSP structures, these follow the
6493 same conventions as certificates and CRLs.
6496 *) New function X509V3_add1_i2d(). This automatically encodes and
6497 adds an extension. Its behaviour can be customised with various
6498 flags to append, replace or delete. Various wrappers added for
6499 certifcates and CRLs.
6502 *) Fix to avoid calling the underlying ASN1 print routine when
6503 an extension cannot be parsed. Correct a typo in the
6504 OCSP_SERVICELOC extension. Tidy up print OCSP format.
6507 *) Make mkdef.pl parse some of the ASN1 macros and add apropriate
6508 entries for variables.
6511 *) Add functionality to apps/openssl.c for detecting locking
6512 problems: As the program is single-threaded, all we have
6513 to do is register a locking callback using an array for
6514 storing which locks are currently held by the program.
6517 *) Use a lock around the call to CRYPTO_get_ex_new_index() in
6518 SSL_get_ex_data_X509_STORE_idx(), which is used in
6519 ssl_verify_cert_chain() and thus can be called at any time
6520 during TLS/SSL handshakes so that thread-safety is essential.
6521 Unfortunately, the ex_data design is not at all suited
6522 for multi-threaded use, so it probably should be abolished.
6525 *) Added Broadcom "ubsec" ENGINE to OpenSSL.
6526 [Broadcom, tweaked and integrated by Geoff Thorpe]
6528 *) Move common extension printing code to new function
6529 X509V3_print_extensions(). Reorganise OCSP print routines and
6530 implement some needed OCSP ASN1 functions. Add OCSP extensions.
6533 *) New function X509_signature_print() to remove duplication in some
6537 *) Add a special meaning when SET OF and SEQUENCE OF flags are both
6538 set (this was treated exactly the same as SET OF previously). This
6539 is used to reorder the STACK representing the structure to match the
6540 encoding. This will be used to get round a problem where a PKCS7
6541 structure which was signed could not be verified because the STACK
6542 order did not reflect the encoded order.
6545 *) Reimplement the OCSP ASN1 module using the new code.
6548 *) Update the X509V3 code to permit the use of an ASN1_ITEM structure
6549 for its ASN1 operations. The old style function pointers still exist
6550 for now but they will eventually go away.
6553 *) Merge in replacement ASN1 code from the ASN1 branch. This almost
6554 completely replaces the old ASN1 functionality with a table driven
6555 encoder and decoder which interprets an ASN1_ITEM structure describing
6556 the ASN1 module. Compatibility with the existing ASN1 API (i2d,d2i) is
6557 largely maintained. Almost all of the old asn1_mac.h macro based ASN1
6558 has also been converted to the new form.
6561 *) Change BN_mod_exp_recp so that negative moduli are tolerated
6562 (the sign is ignored). Similarly, ignore the sign in BN_MONT_CTX_set
6563 so that BN_mod_exp_mont and BN_mod_exp_mont_word work
6564 for negative moduli.
6567 *) Fix BN_uadd and BN_usub: Always return non-negative results instead
6568 of not touching the result's sign bit.
6571 *) BN_div bugfix: If the result is 0, the sign (res->neg) must not be
6575 *) Changed the LHASH code to use prototypes for callbacks, and created
6576 macros to declare and implement thin (optionally static) functions
6577 that provide type-safety and avoid function pointer casting for the
6578 type-specific callbacks.
6581 *) Added Kerberos Cipher Suites to be used with TLS, as written in
6583 [Veers Staats <staatsvr@asc.hpc.mil>,
6584 Jeffrey Altman <jaltman@columbia.edu>, via Richard Levitte]
6586 *) Reformat the FAQ so the different questions and answers can be divided
6587 in sections depending on the subject.
6590 *) Have the zlib compression code load ZLIB.DLL dynamically under
6594 *) New function BN_mod_sqrt for computing square roots modulo a prime
6595 (using the probabilistic Tonelli-Shanks algorithm unless
6596 p == 3 (mod 4) or p == 5 (mod 8), which are cases that can
6597 be handled deterministically).
6598 [Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller]
6600 *) Make BN_mod_inverse faster by explicitly handling small quotients
6601 in the Euclid loop. (Speed gain about 20% for small moduli [256 or
6602 512 bits], about 30% for larger ones [1024 or 2048 bits].)
6605 *) New function BN_kronecker.
6608 *) Fix BN_gcd so that it works on negative inputs; the result is
6609 positive unless both parameters are zero.
6610 Previously something reasonably close to an infinite loop was
6611 possible because numbers could be growing instead of shrinking
6612 in the implementation of Euclid's algorithm.
6615 *) Fix BN_is_word() and BN_is_one() macros to take into account the
6616 sign of the number in question.
6618 Fix BN_is_word(a,w) to work correctly for w == 0.
6620 The old BN_is_word(a,w) macro is now called BN_abs_is_word(a,w)
6621 because its test if the absolute value of 'a' equals 'w'.
6622 Note that BN_abs_is_word does *not* handle w == 0 reliably;
6623 it exists mostly for use in the implementations of BN_is_zero(),
6624 BN_is_one(), and BN_is_word().
6627 *) New function BN_swap.
6630 *) Use BN_nnmod instead of BN_mod in crypto/bn/bn_exp.c so that
6631 the exponentiation functions are more likely to produce reasonable
6632 results on negative inputs.
6635 *) Change BN_mod_mul so that the result is always non-negative.
6636 Previously, it could be negative if one of the factors was negative;
6637 I don't think anyone really wanted that behaviour.
6640 *) Move BN_mod_... functions into new file crypto/bn/bn_mod.c
6641 (except for exponentiation, which stays in crypto/bn/bn_exp.c,
6642 and BN_mod_mul_reciprocal, which stays in crypto/bn/bn_recp.c)
6643 and add new functions:
6652 BN_mod_lshift1_quick
6656 These functions always generate non-negative results.
6658 BN_nnmod otherwise is like BN_mod (if BN_mod computes a remainder r
6659 such that |m| < r < 0, BN_nnmod will output rem + |m| instead).
6661 BN_mod_XXX_quick(r, a, [b,] m) generates the same result as
6662 BN_mod_XXX(r, a, [b,] m, ctx), but requires that a [and b]
6663 be reduced modulo m.
6664 [Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller]
6667 The following entry accidentily appeared in the CHANGES file
6668 distributed with OpenSSL 0.9.7. The modifications described in
6669 it do *not* apply to OpenSSL 0.9.7.
6671 *) Remove a few calls to bn_wexpand() in BN_sqr() (the one in there
6672 was actually never needed) and in BN_mul(). The removal in BN_mul()
6673 required a small change in bn_mul_part_recursive() and the addition
6674 of the functions bn_cmp_part_words(), bn_sub_part_words() and
6675 bn_add_part_words(), which do the same thing as bn_cmp_words(),
6676 bn_sub_words() and bn_add_words() except they take arrays with
6681 *) In 'openssl passwd', verify passwords read from the terminal
6682 unless the '-salt' option is used (which usually means that
6683 verification would just waste user's time since the resulting
6684 hash is going to be compared with some given password hash)
6685 or the new '-noverify' option is used.
6687 This is an incompatible change, but it does not affect
6688 non-interactive use of 'openssl passwd' (passwords on the command
6689 line, '-stdin' option, '-in ...' option) and thus should not
6693 *) Remove all references to RSAref, since there's no more need for it.
6696 *) Make DSO load along a path given through an environment variable
6697 (SHLIB_PATH) with shl_load().
6700 *) Constify the ENGINE code as a result of BIGNUM constification.
6701 Also constify the RSA code and most things related to it. In a
6702 few places, most notable in the depth of the ASN.1 code, ugly
6703 casts back to non-const were required (to be solved at a later
6707 *) Make it so the openssl application has all engines loaded by default.
6710 *) Constify the BIGNUM routines a little more.
6713 *) Add the following functions:
6715 ENGINE_load_cswift()
6717 ENGINE_load_atalla()
6719 ENGINE_load_builtin_engines()
6721 That way, an application can itself choose if external engines that
6722 are built-in in OpenSSL shall ever be used or not. The benefit is
6723 that applications won't have to be linked with libdl or other dso
6724 libraries unless it's really needed.
6726 Changed 'openssl engine' to load all engines on demand.
6727 Changed the engine header files to avoid the duplication of some
6728 declarations (they differed!).
6731 *) 'openssl engine' can now list capabilities.
6734 *) Better error reporting in 'openssl engine'.
6737 *) Never call load_dh_param(NULL) in s_server.
6740 *) Add engine application. It can currently list engines by name and
6741 identity, and test if they are actually available.
6744 *) Improve RPM specification file by forcing symbolic linking and making
6745 sure the installed documentation is also owned by root.root.
6746 [Damien Miller <djm@mindrot.org>]
6748 *) Give the OpenSSL applications more possibilities to make use of
6749 keys (public as well as private) handled by engines.
6752 *) Add OCSP code that comes from CertCo.
6755 *) Add VMS support for the Rijndael code.
6758 *) Added untested support for Nuron crypto accelerator.
6761 *) Add support for external cryptographic devices. This code was
6762 previously distributed separately as the "engine" branch.
6763 [Geoff Thorpe, Richard Levitte]
6765 *) Rework the filename-translation in the DSO code. It is now possible to
6766 have far greater control over how a "name" is turned into a filename
6767 depending on the operating environment and any oddities about the
6768 different shared library filenames on each system.
6771 *) Support threads on FreeBSD-elf in Configure.
6774 *) Fix for SHA1 assembly problem with MASM: it produces
6775 warnings about corrupt line number information when assembling
6776 with debugging information. This is caused by the overlapping
6778 [Bernd Matthes <mainbug@celocom.de>, Steve Henson]
6781 NCONF_get_number() has no error checking at all. As a replacement,
6782 NCONF_get_number_e() is defined (_e for "error checking") and is
6783 promoted strongly. The old NCONF_get_number is kept around for
6784 binary backward compatibility.
6785 Make it possible for methods to load from something other than a BIO,
6786 by providing a function pointer that is given a name instead of a BIO.
6787 For example, this could be used to load configuration data from an
6791 *) Fix for non blocking accept BIOs. Added new I/O special reason
6792 BIO_RR_ACCEPT to cover this case. Previously use of accept BIOs
6793 with non blocking I/O was not possible because no retry code was
6794 implemented. Also added new SSL code SSL_WANT_ACCEPT to cover
6798 *) Added the beginnings of Rijndael support.
6801 *) Fix for bug in DirectoryString mask setting. Add support for
6802 X509_NAME_print_ex() in 'req' and X509_print_ex() function
6803 to allow certificate printing to more controllable, additional
6804 'certopt' option to 'x509' to allow new printing options to be
6808 *) Clean old EAY MD5 hack from e_os.h.
6811 Changes between 0.9.6l and 0.9.6m [17 Mar 2004]
6813 *) Fix null-pointer assignment in do_change_cipher_spec() revealed
6814 by using the Codenomicon TLS Test Tool (CVE-2004-0079)
6815 [Joe Orton, Steve Henson]
6817 Changes between 0.9.6k and 0.9.6l [04 Nov 2003]
6819 *) Fix additional bug revealed by the NISCC test suite:
6821 Stop bug triggering large recursion when presented with
6822 certain ASN.1 tags (CVE-2003-0851)
6825 Changes between 0.9.6j and 0.9.6k [30 Sep 2003]
6827 *) Fix various bugs revealed by running the NISCC test suite:
6829 Stop out of bounds reads in the ASN1 code when presented with
6830 invalid tags (CVE-2003-0543 and CVE-2003-0544).
6832 If verify callback ignores invalid public key errors don't try to check
6833 certificate signature with the NULL public key.
6837 *) In ssl3_accept() (ssl/s3_srvr.c) only accept a client certificate
6838 if the server requested one: as stated in TLS 1.0 and SSL 3.0
6842 *) In ssl3_get_client_hello() (ssl/s3_srvr.c), tolerate additional
6843 extra data after the compression methods not only for TLS 1.0
6844 but also for SSL 3.0 (as required by the specification).
6845 [Bodo Moeller; problem pointed out by Matthias Loepfe]
6847 *) Change X509_certificate_type() to mark the key as exported/exportable
6848 when it's 512 *bits* long, not 512 bytes.
6851 Changes between 0.9.6i and 0.9.6j [10 Apr 2003]
6853 *) Countermeasure against the Klima-Pokorny-Rosa extension of
6854 Bleichbacher's attack on PKCS #1 v1.5 padding: treat
6855 a protocol version number mismatch like a decryption error
6856 in ssl3_get_client_key_exchange (ssl/s3_srvr.c).
6859 *) Turn on RSA blinding by default in the default implementation
6860 to avoid a timing attack. Applications that don't want it can call
6861 RSA_blinding_off() or use the new flag RSA_FLAG_NO_BLINDING.
6862 They would be ill-advised to do so in most cases.
6863 [Ben Laurie, Steve Henson, Geoff Thorpe, Bodo Moeller]
6865 *) Change RSA blinding code so that it works when the PRNG is not
6866 seeded (in this case, the secret RSA exponent is abused as
6867 an unpredictable seed -- if it is not unpredictable, there
6868 is no point in blinding anyway). Make RSA blinding thread-safe
6869 by remembering the creator's thread ID in rsa->blinding and
6870 having all other threads use local one-time blinding factors
6871 (this requires more computation than sharing rsa->blinding, but
6872 avoids excessive locking; and if an RSA object is not shared
6873 between threads, blinding will still be very fast).
6876 Changes between 0.9.6h and 0.9.6i [19 Feb 2003]
6878 *) In ssl3_get_record (ssl/s3_pkt.c), minimize information leaked
6879 via timing by performing a MAC computation even if incorrrect
6880 block cipher padding has been found. This is a countermeasure
6881 against active attacks where the attacker has to distinguish
6882 between bad padding and a MAC verification error. (CVE-2003-0078)
6884 [Bodo Moeller; problem pointed out by Brice Canvel (EPFL),
6885 Alain Hiltgen (UBS), Serge Vaudenay (EPFL), and
6886 Martin Vuagnoux (EPFL, Ilion)]
6888 Changes between 0.9.6g and 0.9.6h [5 Dec 2002]
6890 *) New function OPENSSL_cleanse(), which is used to cleanse a section of
6891 memory from it's contents. This is done with a counter that will
6892 place alternating values in each byte. This can be used to solve
6893 two issues: 1) the removal of calls to memset() by highly optimizing
6894 compilers, and 2) cleansing with other values than 0, since those can
6895 be read through on certain media, for example a swap space on disk.
6898 *) Bugfix: client side session caching did not work with external caching,
6899 because the session->cipher setting was not restored when reloading
6900 from the external cache. This problem was masked, when
6901 SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG (part of SSL_OP_ALL) was set.
6902 (Found by Steve Haslam <steve@araqnid.ddts.net>.)
6905 *) Fix client_certificate (ssl/s2_clnt.c): The permissible total
6906 length of the REQUEST-CERTIFICATE message is 18 .. 34, not 17 .. 33.
6907 [Zeev Lieber <zeev-l@yahoo.com>]
6909 *) Undo an undocumented change introduced in 0.9.6e which caused
6910 repeated calls to OpenSSL_add_all_ciphers() and
6911 OpenSSL_add_all_digests() to be ignored, even after calling
6915 *) Change the default configuration reader to deal with last line not
6916 being properly terminated.
6919 *) Change X509_NAME_cmp() so it applies the special rules on handling
6920 DN values that are of type PrintableString, as well as RDNs of type
6921 emailAddress where the value has the type ia5String.
6922 [stefank@valicert.com via Richard Levitte]
6924 *) Add a SSL_SESS_CACHE_NO_INTERNAL_STORE flag to take over half
6925 the job SSL_SESS_CACHE_NO_INTERNAL_LOOKUP was inconsistently
6926 doing, define a new flag (SSL_SESS_CACHE_NO_INTERNAL) to be
6927 the bitwise-OR of the two for use by the majority of applications
6928 wanting this behaviour, and update the docs. The documented
6929 behaviour and actual behaviour were inconsistent and had been
6930 changing anyway, so this is more a bug-fix than a behavioural
6932 [Geoff Thorpe, diagnosed by Nadav Har'El]
6934 *) Don't impose a 16-byte length minimum on session IDs in ssl/s3_clnt.c
6935 (the SSL 3.0 and TLS 1.0 specifications allow any length up to 32 bytes).
6938 *) Fix initialization code race conditions in
6939 SSLv23_method(), SSLv23_client_method(), SSLv23_server_method(),
6940 SSLv2_method(), SSLv2_client_method(), SSLv2_server_method(),
6941 SSLv3_method(), SSLv3_client_method(), SSLv3_server_method(),
6942 TLSv1_method(), TLSv1_client_method(), TLSv1_server_method(),
6943 ssl2_get_cipher_by_char(),
6944 ssl3_get_cipher_by_char().
6945 [Patrick McCormick <patrick@tellme.com>, Bodo Moeller]
6947 *) Reorder cleanup sequence in SSL_CTX_free(): only remove the ex_data after
6948 the cached sessions are flushed, as the remove_cb() might use ex_data
6949 contents. Bug found by Sam Varshavchik <mrsam@courier-mta.com>
6950 (see [openssl.org #212]).
6951 [Geoff Thorpe, Lutz Jaenicke]
6953 *) Fix typo in OBJ_txt2obj which incorrectly passed the content
6954 length, instead of the encoding length to d2i_ASN1_OBJECT.
6957 Changes between 0.9.6f and 0.9.6g [9 Aug 2002]
6959 *) [In 0.9.6g-engine release:]
6960 Fix crypto/engine/vendor_defns/cswift.h for WIN32 (use '_stdcall').
6961 [Lynn Gazis <lgazis@rainbow.com>]
6963 Changes between 0.9.6e and 0.9.6f [8 Aug 2002]
6965 *) Fix ASN1 checks. Check for overflow by comparing with LONG_MAX
6966 and get fix the header length calculation.
6967 [Florian Weimer <Weimer@CERT.Uni-Stuttgart.DE>,
6968 Alon Kantor <alonk@checkpoint.com> (and others),
6971 *) Use proper error handling instead of 'assertions' in buffer
6972 overflow checks added in 0.9.6e. This prevents DoS (the
6973 assertions could call abort()).
6974 [Arne Ansper <arne@ats.cyber.ee>, Bodo Moeller]
6976 Changes between 0.9.6d and 0.9.6e [30 Jul 2002]
6978 *) Add various sanity checks to asn1_get_length() to reject
6979 the ASN1 length bytes if they exceed sizeof(long), will appear
6980 negative or the content length exceeds the length of the
6982 [Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>]
6984 *) Fix cipher selection routines: ciphers without encryption had no flags
6985 for the cipher strength set and where therefore not handled correctly
6986 by the selection routines (PR #130).
6989 *) Fix EVP_dsa_sha macro.
6993 SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS
6994 for disabling the SSL 3.0/TLS 1.0 CBC vulnerability countermeasure
6995 that was added in OpenSSL 0.9.6d.
6997 As the countermeasure turned out to be incompatible with some
6998 broken SSL implementations, the new option is part of SSL_OP_ALL.
6999 SSL_OP_ALL is usually employed when compatibility with weird SSL
7000 implementations is desired (e.g. '-bugs' option to 's_client' and
7001 's_server'), so the new option is automatically set in many
7005 *) Changes in security patch:
7007 Changes marked "(CHATS)" were sponsored by the Defense Advanced
7008 Research Projects Agency (DARPA) and Air Force Research Laboratory,
7009 Air Force Materiel Command, USAF, under agreement number
7012 *) Add various sanity checks to asn1_get_length() to reject
7013 the ASN1 length bytes if they exceed sizeof(long), will appear
7014 negative or the content length exceeds the length of the
7015 supplied buffer. (CVE-2002-0659)
7016 [Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>]
7018 *) Assertions for various potential buffer overflows, not known to
7020 [Ben Laurie (CHATS)]
7022 *) Various temporary buffers to hold ASCII versions of integers were
7023 too small for 64 bit platforms. (CVE-2002-0655)
7024 [Matthew Byng-Maddick <mbm@aldigital.co.uk> and Ben Laurie (CHATS)>
7026 *) Remote buffer overflow in SSL3 protocol - an attacker could
7027 supply an oversized session ID to a client. (CVE-2002-0656)
7028 [Ben Laurie (CHATS)]
7030 *) Remote buffer overflow in SSL2 protocol - an attacker could
7031 supply an oversized client master key. (CVE-2002-0656)
7032 [Ben Laurie (CHATS)]
7034 Changes between 0.9.6c and 0.9.6d [9 May 2002]
7036 *) Fix crypto/asn1/a_sign.c so that 'parameters' is omitted (not
7037 encoded as NULL) with id-dsa-with-sha1.
7038 [Nils Larsch <nla@trustcenter.de>; problem pointed out by Bodo Moeller]
7040 *) Check various X509_...() return values in apps/req.c.
7041 [Nils Larsch <nla@trustcenter.de>]
7043 *) Fix BASE64 decode (EVP_DecodeUpdate) for data with CR/LF ended lines:
7044 an end-of-file condition would erronously be flagged, when the CRLF
7045 was just at the end of a processed block. The bug was discovered when
7046 processing data through a buffering memory BIO handing the data to a
7047 BASE64-decoding BIO. Bug fund and patch submitted by Pavel Tsekov
7048 <ptsekov@syntrex.com> and Nedelcho Stanev.
7051 *) Implement a countermeasure against a vulnerability recently found
7052 in CBC ciphersuites in SSL 3.0/TLS 1.0: Send an empty fragment
7053 before application data chunks to avoid the use of known IVs
7054 with data potentially chosen by the attacker.
7057 *) Fix length checks in ssl3_get_client_hello().
7060 *) TLS/SSL library bugfix: use s->s3->in_read_app_data differently
7061 to prevent ssl3_read_internal() from incorrectly assuming that
7062 ssl3_read_bytes() found application data while handshake
7063 processing was enabled when in fact s->s3->in_read_app_data was
7064 merely automatically cleared during the initial handshake.
7065 [Bodo Moeller; problem pointed out by Arne Ansper <arne@ats.cyber.ee>]
7067 *) Fix object definitions for Private and Enterprise: they were not
7068 recognized in their shortname (=lowercase) representation. Extend
7069 obj_dat.pl to issue an error when using undefined keywords instead
7070 of silently ignoring the problem (Svenning Sorensen
7071 <sss@sss.dnsalias.net>).
7074 *) Fix DH_generate_parameters() so that it works for 'non-standard'
7075 generators, i.e. generators other than 2 and 5. (Previously, the
7076 code did not properly initialise the 'add' and 'rem' values to
7077 BN_generate_prime().)
7079 In the new general case, we do not insist that 'generator' is
7080 actually a primitive root: This requirement is rather pointless;
7081 a generator of the order-q subgroup is just as good, if not
7085 *) Map new X509 verification errors to alerts. Discovered and submitted by
7086 Tom Wu <tom@arcot.com>.
7089 *) Fix ssl3_pending() (ssl/s3_lib.c) to prevent SSL_pending() from
7090 returning non-zero before the data has been completely received
7091 when using non-blocking I/O.
7092 [Bodo Moeller; problem pointed out by John Hughes]
7094 *) Some of the ciphers missed the strength entry (SSL_LOW etc).
7095 [Ben Laurie, Lutz Jaenicke]
7097 *) Fix bug in SSL_clear(): bad sessions were not removed (found by
7098 Yoram Zahavi <YoramZ@gilian.com>).
7101 *) Add information about CygWin 1.3 and on, and preserve proper
7102 configuration for the versions before that.
7103 [Corinna Vinschen <vinschen@redhat.com> and Richard Levitte]
7105 *) Make removal from session cache (SSL_CTX_remove_session()) more robust:
7106 check whether we deal with a copy of a session and do not delete from
7107 the cache in this case. Problem reported by "Izhar Shoshani Levi"
7108 <izhar@checkpoint.com>.
7111 *) Do not store session data into the internal session cache, if it
7112 is never intended to be looked up (SSL_SESS_CACHE_NO_INTERNAL_LOOKUP
7113 flag is set). Proposed by Aslam <aslam@funk.com>.
7116 *) Have ASN1_BIT_STRING_set_bit() really clear a bit when the requested
7120 *) [In 0.9.6d-engine release:]
7121 Fix a crashbug and a logic bug in hwcrhk_load_pubkey().
7122 [Toomas Kiisk <vix@cyber.ee> via Richard Levitte]
7124 *) Add the configuration target linux-s390x.
7125 [Neale Ferguson <Neale.Ferguson@SoftwareAG-USA.com> via Richard Levitte]
7127 *) The earlier bugfix for the SSL3_ST_SW_HELLO_REQ_C case of
7128 ssl3_accept (ssl/s3_srvr.c) incorrectly used a local flag
7129 variable as an indication that a ClientHello message has been
7130 received. As the flag value will be lost between multiple
7131 invocations of ssl3_accept when using non-blocking I/O, the
7132 function may not be aware that a handshake has actually taken
7133 place, thus preventing a new session from being added to the
7136 To avoid this problem, we now set s->new_session to 2 instead of
7137 using a local variable.
7138 [Lutz Jaenicke, Bodo Moeller]
7140 *) Bugfix: Return -1 from ssl3_get_server_done (ssl3/s3_clnt.c)
7141 if the SSL_R_LENGTH_MISMATCH error is detected.
7142 [Geoff Thorpe, Bodo Moeller]
7144 *) New 'shared_ldflag' column in Configure platform table.
7147 *) Fix EVP_CIPHER_mode macro.
7148 ["Dan S. Camper" <dan@bti.net>]
7150 *) Fix ssl3_read_bytes (ssl/s3_pkt.c): To ignore messages of unknown
7151 type, we must throw them away by setting rr->length to 0.
7152 [D P Chang <dpc@qualys.com>]
7154 Changes between 0.9.6b and 0.9.6c [21 dec 2001]
7156 *) Fix BN_rand_range bug pointed out by Dominikus Scherkl
7157 <Dominikus.Scherkl@biodata.com>. (The previous implementation
7158 worked incorrectly for those cases where range = 10..._2 and
7159 3*range is two bits longer than range.)
7162 *) Only add signing time to PKCS7 structures if it is not already
7166 *) Fix crypto/objects/objects.h: "ld-ce" should be "id-ce",
7167 OBJ_ld_ce should be OBJ_id_ce.
7168 Also some ip-pda OIDs in crypto/objects/objects.txt were
7169 incorrect (cf. RFC 3039).
7170 [Matt Cooper, Frederic Giudicelli, Bodo Moeller]
7172 *) Release CRYPTO_LOCK_DYNLOCK when CRYPTO_destroy_dynlockid()
7173 returns early because it has nothing to do.
7174 [Andy Schneider <andy.schneider@bjss.co.uk>]
7176 *) [In 0.9.6c-engine release:]
7177 Fix mutex callback return values in crypto/engine/hw_ncipher.c.
7178 [Andy Schneider <andy.schneider@bjss.co.uk>]
7180 *) [In 0.9.6c-engine release:]
7181 Add support for Cryptographic Appliance's keyserver technology.
7182 (Use engine 'keyclient')
7183 [Cryptographic Appliances and Geoff Thorpe]
7185 *) Add a configuration entry for OS/390 Unix. The C compiler 'c89'
7186 is called via tools/c89.sh because arguments have to be
7187 rearranged (all '-L' options must appear before the first object
7189 [Richard Shapiro <rshapiro@abinitio.com>]
7191 *) [In 0.9.6c-engine release:]
7192 Add support for Broadcom crypto accelerator cards, backported
7194 [Broadcom, Nalin Dahyabhai <nalin@redhat.com>, Mark Cox]
7196 *) [In 0.9.6c-engine release:]
7197 Add support for SureWare crypto accelerator cards from
7198 Baltimore Technologies. (Use engine 'sureware')
7199 [Baltimore Technologies and Mark Cox]
7201 *) [In 0.9.6c-engine release:]
7202 Add support for crypto accelerator cards from Accelerated
7203 Encryption Processing, www.aep.ie. (Use engine 'aep')
7204 [AEP Inc. and Mark Cox]
7206 *) Add a configuration entry for gcc on UnixWare.
7207 [Gary Benson <gbenson@redhat.com>]
7209 *) Change ssl/s2_clnt.c and ssl/s2_srvr.c so that received handshake
7210 messages are stored in a single piece (fixed-length part and
7211 variable-length part combined) and fix various bugs found on the way.
7214 *) Disable caching in BIO_gethostbyname(), directly use gethostbyname()
7215 instead. BIO_gethostbyname() does not know what timeouts are
7216 appropriate, so entries would stay in cache even when they have
7218 [Bodo Moeller; problem pointed out by Rich Salz <rsalz@zolera.com>
7220 *) Change ssl23_get_client_hello (ssl/s23_srvr.c) behaviour when
7221 faced with a pathologically small ClientHello fragment that does
7222 not contain client_version: Instead of aborting with an error,
7223 simply choose the highest available protocol version (i.e.,
7224 TLS 1.0 unless it is disabled). In practice, ClientHello
7225 messages are never sent like this, but this change gives us
7226 strictly correct behaviour at least for TLS.
7229 *) Fix SSL handshake functions and SSL_clear() such that SSL_clear()
7230 never resets s->method to s->ctx->method when called from within
7231 one of the SSL handshake functions.
7232 [Bodo Moeller; problem pointed out by Niko Baric]
7234 *) In ssl3_get_client_hello (ssl/s3_srvr.c), generate a fatal alert
7235 (sent using the client's version number) if client_version is
7236 smaller than the protocol version in use. Also change
7237 ssl23_get_client_hello (ssl/s23_srvr.c) to select TLS 1.0 if
7238 the client demanded SSL 3.0 but only TLS 1.0 is enabled; then
7239 the client will at least see that alert.
7242 *) Fix ssl3_get_message (ssl/s3_both.c) to handle message fragmentation
7246 *) Avoid infinite loop in ssl3_get_message (ssl/s3_both.c) if a
7247 client receives HelloRequest while in a handshake.
7248 [Bodo Moeller; bug noticed by Andy Schneider <andy.schneider@bjss.co.uk>]
7250 *) Bugfix in ssl3_accept (ssl/s3_srvr.c): Case SSL3_ST_SW_HELLO_REQ_C
7251 should end in 'break', not 'goto end' which circuments various
7252 cleanups done in state SSL_ST_OK. But session related stuff
7253 must be disabled for SSL_ST_OK in the case that we just sent a
7256 Also avoid some overhead by not calling ssl_init_wbio_buffer()
7257 before just sending a HelloRequest.
7258 [Bodo Moeller, Eric Rescorla <ekr@rtfm.com>]
7260 *) Fix ssl/s3_enc.c, ssl/t1_enc.c and ssl/s3_pkt.c so that we don't
7261 reveal whether illegal block cipher padding was found or a MAC
7262 verification error occured. (Neither SSLerr() codes nor alerts
7263 are directly visible to potential attackers, but the information
7264 may leak via logfiles.)
7266 Similar changes are not required for the SSL 2.0 implementation
7267 because the number of padding bytes is sent in clear for SSL 2.0,
7268 and the extra bytes are just ignored. However ssl/s2_pkt.c
7269 failed to verify that the purported number of padding bytes is in
7273 *) Add OpenUNIX-8 support including shared libraries
7274 (Boyd Lynn Gerber <gerberb@zenez.com>).
7277 *) Improve RSA_padding_check_PKCS1_OAEP() check again to avoid
7278 'wristwatch attack' using huge encoding parameters (cf.
7279 James H. Manger's CRYPTO 2001 paper). Note that the
7280 RSA_PKCS1_OAEP_PADDING case of RSA_private_decrypt() does not use
7281 encoding parameters and hence was not vulnerable.
7284 *) BN_sqr() bug fix.
7285 [Ulf Möller, reported by Jim Ellis <jim.ellis@cavium.com>]
7287 *) Rabin-Miller test analyses assume uniformly distributed witnesses,
7288 so use BN_pseudo_rand_range() instead of using BN_pseudo_rand()
7289 followed by modular reduction.
7290 [Bodo Moeller; pointed out by Adam Young <AYoung1@NCSUS.JNJ.COM>]
7292 *) Add BN_pseudo_rand_range() with obvious functionality: BN_rand_range()
7293 equivalent based on BN_pseudo_rand() instead of BN_rand().
7296 *) s3_srvr.c: allow sending of large client certificate lists (> 16 kB).
7297 This function was broken, as the check for a new client hello message
7298 to handle SGC did not allow these large messages.
7299 (Tracked down by "Douglas E. Engert" <deengert@anl.gov>.)
7302 *) Add alert descriptions for TLSv1 to SSL_alert_desc_string[_long]().
7305 *) Fix buggy behaviour of BIO_get_num_renegotiates() and BIO_ctrl()
7306 for BIO_C_GET_WRITE_BUF_SIZE ("Stephen Hinton" <shinton@netopia.com>).
7309 *) Rework the configuration and shared library support for Tru64 Unix.
7310 The configuration part makes use of modern compiler features and
7311 still retains old compiler behavior for those that run older versions
7312 of the OS. The shared library support part includes a variant that
7313 uses the RPATH feature, and is available through the special
7314 configuration target "alpha-cc-rpath", which will never be selected
7316 [Tim Mooney <mooney@dogbert.cc.ndsu.NoDak.edu> via Richard Levitte]
7318 *) In ssl3_get_key_exchange (ssl/s3_clnt.c), call ssl3_get_message()
7319 with the same message size as in ssl3_get_certificate_request().
7320 Otherwise, if no ServerKeyExchange message occurs, CertificateRequest
7321 messages might inadvertently be reject as too long.
7322 [Petr Lampa <lampa@fee.vutbr.cz>]
7324 *) Enhanced support for IA-64 Unix platforms (well, Linux and HP-UX).
7327 *) Modified SSL library such that the verify_callback that has been set
7328 specificly for an SSL object with SSL_set_verify() is actually being
7329 used. Before the change, a verify_callback set with this function was
7330 ignored and the verify_callback() set in the SSL_CTX at the time of
7331 the call was used. New function X509_STORE_CTX_set_verify_cb() introduced
7332 to allow the necessary settings.
7335 *) Initialize static variable in crypto/dsa/dsa_lib.c and crypto/dh/dh_lib.c
7336 explicitly to NULL, as at least on Solaris 8 this seems not always to be
7337 done automatically (in contradiction to the requirements of the C
7338 standard). This made problems when used from OpenSSH.
7341 *) In OpenSSL 0.9.6a and 0.9.6b, crypto/dh/dh_key.c ignored
7342 dh->length and always used
7344 BN_rand_range(priv_key, dh->p).
7346 BN_rand_range() is not necessary for Diffie-Hellman, and this
7347 specific range makes Diffie-Hellman unnecessarily inefficient if
7348 dh->length (recommended exponent length) is much smaller than the
7349 length of dh->p. We could use BN_rand_range() if the order of
7350 the subgroup was stored in the DH structure, but we only have
7355 BN_rand(priv_key, l, ...)
7357 where 'l' is dh->length if this is defined, or BN_num_bits(dh->p)-1
7363 RSA_eay_public_encrypt
7364 RSA_eay_private_decrypt
7365 RSA_eay_private_encrypt (signing)
7366 RSA_eay_public_decrypt (signature verification)
7368 (default implementations for RSA_public_encrypt,
7369 RSA_private_decrypt, RSA_private_encrypt, RSA_public_decrypt),
7370 always reject numbers >= n.
7373 *) In crypto/rand/md_rand.c, use a new short-time lock CRYPTO_LOCK_RAND2
7374 to synchronize access to 'locking_thread'. This is necessary on
7375 systems where access to 'locking_thread' (an 'unsigned long'
7376 variable) is not atomic.
7379 *) In crypto/rand/md_rand.c, set 'locking_thread' to current thread's ID
7380 *before* setting the 'crypto_lock_rand' flag. The previous code had
7381 a race condition if 0 is a valid thread ID.
7382 [Travis Vitek <vitek@roguewave.com>]
7384 *) Add support for shared libraries under Irix.
7385 [Albert Chin-A-Young <china@thewrittenword.com>]
7387 *) Add configuration option to build on Linux on both big-endian and
7389 [Ralf Baechle <ralf@uni-koblenz.de>]
7391 *) Add the possibility to create shared libraries on HP-UX.
7394 Changes between 0.9.6a and 0.9.6b [9 Jul 2001]
7396 *) Change ssleay_rand_bytes (crypto/rand/md_rand.c)
7397 to avoid a SSLeay/OpenSSL PRNG weakness pointed out by
7398 Markku-Juhani O. Saarinen <markku-juhani.saarinen@nokia.com>:
7399 PRNG state recovery was possible based on the output of
7400 one PRNG request appropriately sized to gain knowledge on
7401 'md' followed by enough consecutive 1-byte PRNG requests
7402 to traverse all of 'state'.
7404 1. When updating 'md_local' (the current thread's copy of 'md')
7405 during PRNG output generation, hash all of the previous
7406 'md_local' value, not just the half used for PRNG output.
7408 2. Make the number of bytes from 'state' included into the hash
7409 independent from the number of PRNG bytes requested.
7411 The first measure alone would be sufficient to avoid
7412 Markku-Juhani's attack. (Actually it had never occurred
7413 to me that the half of 'md_local' used for chaining was the
7414 half from which PRNG output bytes were taken -- I had always
7415 assumed that the secret half would be used.) The second
7416 measure makes sure that additional data from 'state' is never
7417 mixed into 'md_local' in small portions; this heuristically
7418 further strengthens the PRNG.
7421 *) Fix crypto/bn/asm/mips3.s.
7424 *) When only the key is given to "enc", the IV is undefined. Print out
7425 an error message in this case.
7428 *) Handle special case when X509_NAME is empty in X509 printing routines.
7431 *) In dsa_do_verify (crypto/dsa/dsa_ossl.c), verify that r and s are
7432 positive and less than q.
7435 *) Don't change *pointer in CRYPTO_add_lock() is add_lock_callback is
7436 used: it isn't thread safe and the add_lock_callback should handle
7438 [Paul Rose <Paul.Rose@bridge.com>]
7440 *) Verify that incoming data obeys the block size in
7441 ssl3_enc (ssl/s3_enc.c) and tls1_enc (ssl/t1_enc.c).
7445 [Ulf Möller, Bodo Möller]
7447 *) The countermeasure against Bleichbacher's attack on PKCS #1 v1.5
7448 RSA encryption was accidentally removed in s3_srvr.c in OpenSSL 0.9.5
7449 when fixing the server behaviour for backwards-compatible 'client
7450 hello' messages. (Note that the attack is impractical against
7451 SSL 3.0 and TLS 1.0 anyway because length and version checking
7452 means that the probability of guessing a valid ciphertext is
7453 around 2^-40; see section 5 in Bleichenbacher's CRYPTO '98
7456 Before 0.9.5, the countermeasure (hide the error by generating a
7457 random 'decryption result') did not work properly because
7458 ERR_clear_error() was missing, meaning that SSL_get_error() would
7459 detect the supposedly ignored error.
7461 Both problems are now fixed.
7464 *) In crypto/bio/bf_buff.c, increase DEFAULT_BUFFER_SIZE to 4096
7465 (previously it was 1024).
7468 *) Fix for compatibility mode trust settings: ignore trust settings
7469 unless some valid trust or reject settings are present.
7472 *) Fix for blowfish EVP: its a variable length cipher.
7475 *) Fix various bugs related to DSA S/MIME verification. Handle missing
7476 parameters in DSA public key structures and return an error in the
7477 DSA routines if parameters are absent.
7480 *) In versions up to 0.9.6, RAND_file_name() resorted to file ".rnd"
7481 in the current directory if neither $RANDFILE nor $HOME was set.
7482 RAND_file_name() in 0.9.6a returned NULL in this case. This has
7483 caused some confusion to Windows users who haven't defined $HOME.
7484 Thus RAND_file_name() is changed again: e_os.h can define a
7485 DEFAULT_HOME, which will be used if $HOME is not set.
7486 For Windows, we use "C:"; on other platforms, we still require
7487 environment variables.
7489 *) Move 'if (!initialized) RAND_poll()' into regions protected by
7490 CRYPTO_LOCK_RAND. This is not strictly necessary, but avoids
7491 having multiple threads call RAND_poll() concurrently.
7494 *) In crypto/rand/md_rand.c, replace 'add_do_not_lock' flag by a
7495 combination of a flag and a thread ID variable.
7496 Otherwise while one thread is in ssleay_rand_bytes (which sets the
7497 flag), *other* threads can enter ssleay_add_bytes without obeying
7498 the CRYPTO_LOCK_RAND lock (and may even illegally release the lock
7499 that they do not hold after the first thread unsets add_do_not_lock).
7502 *) Change bctest again: '-x' expressions are not available in all
7506 Changes between 0.9.6 and 0.9.6a [5 Apr 2001]
7508 *) Fix a couple of memory leaks in PKCS7_dataDecode()
7509 [Steve Henson, reported by Heyun Zheng <hzheng@atdsprint.com>]
7511 *) Change Configure and Makefiles to provide EXE_EXT, which will contain
7512 the default extension for executables, if any. Also, make the perl
7513 scripts that use symlink() to test if it really exists and use "cp"
7514 if it doesn't. All this made OpenSSL compilable and installable in
7518 *) Fix for asn1_GetSequence() for indefinite length constructed data.
7519 If SEQUENCE is length is indefinite just set c->slen to the total
7520 amount of data available.
7521 [Steve Henson, reported by shige@FreeBSD.org]
7522 [This change does not apply to 0.9.7.]
7524 *) Change bctest to avoid here-documents inside command substitution
7525 (workaround for FreeBSD /bin/sh bug).
7526 For compatibility with Ultrix, avoid shell functions (introduced
7527 in the bctest version that searches along $PATH).
7530 *) Rename 'des_encrypt' to 'des_encrypt1'. This avoids the clashes
7531 with des_encrypt() defined on some operating systems, like Solaris
7535 *) Check the result of RSA-CRT (see D. Boneh, R. DeMillo, R. Lipton:
7536 On the Importance of Eliminating Errors in Cryptographic
7537 Computations, J. Cryptology 14 (2001) 2, 101-119,
7538 http://theory.stanford.edu/~dabo/papers/faults.ps.gz).
7541 *) MIPS assembler BIGNUM division bug fix.
7544 *) Disabled incorrect Alpha assembler code.
7547 *) Fix PKCS#7 decode routines so they correctly update the length
7548 after reading an EOC for the EXPLICIT tag.
7550 [This change does not apply to 0.9.7.]
7552 *) Fix bug in PKCS#12 key generation routines. This was triggered
7553 if a 3DES key was generated with a 0 initial byte. Include
7554 PKCS12_BROKEN_KEYGEN compilation option to retain the old
7555 (but broken) behaviour.
7558 *) Enhance bctest to search for a working bc along $PATH and print
7560 [Tim Rice <tim@multitalents.net> via Richard Levitte]
7562 *) Fix memory leaks in err.c: free err_data string if necessary;
7563 don't write to the wrong index in ERR_set_error_data.
7566 *) Implement ssl23_peek (analogous to ssl23_read), which previously
7570 *) Replace rdtsc with _emit statements for VC++ version 5.
7571 [Jeremy Cooper <jeremy@baymoo.org>]
7573 *) Make it possible to reuse SSLv2 sessions.
7576 *) In copy_email() check for >= 0 as a return value for
7577 X509_NAME_get_index_by_NID() since 0 is a valid index.
7578 [Steve Henson reported by Massimiliano Pala <madwolf@opensca.org>]
7580 *) Avoid coredump with unsupported or invalid public keys by checking if
7581 X509_get_pubkey() fails in PKCS7_verify(). Fix memory leak when
7582 PKCS7_verify() fails with non detached data.
7585 *) Don't use getenv in library functions when run as setuid/setgid.
7586 New function OPENSSL_issetugid().
7589 *) Avoid false positives in memory leak detection code (crypto/mem_dbg.c)
7590 due to incorrect handling of multi-threading:
7592 1. Fix timing glitch in the MemCheck_off() portion of CRYPTO_mem_ctrl().
7594 2. Fix logical glitch in is_MemCheck_on() aka CRYPTO_is_mem_check_on().
7596 3. Count how many times MemCheck_off() has been called so that
7597 nested use can be treated correctly. This also avoids
7598 inband-signalling in the previous code (which relied on the
7599 assumption that thread ID 0 is impossible).
7602 *) Add "-rand" option also to s_client and s_server.
7605 *) Fix CPU detection on Irix 6.x.
7606 [Kurt Hockenbury <khockenb@stevens-tech.edu> and
7607 "Bruce W. Forsberg" <bruce.forsberg@baesystems.com>]
7609 *) Fix X509_NAME bug which produced incorrect encoding if X509_NAME
7612 [This change does not apply to 0.9.7.]
7614 *) Use the cached encoding of an X509_NAME structure rather than
7615 copying it. This is apparently the reason for the libsafe "errors"
7616 but the code is actually correct.
7619 *) Add new function BN_rand_range(), and fix DSA_sign_setup() to prevent
7620 Bleichenbacher's DSA attack.
7621 Extend BN_[pseudo_]rand: As before, top=1 forces the highest two bits
7622 to be set and top=0 forces the highest bit to be set; top=-1 is new
7623 and leaves the highest bit random.
7624 [Ulf Moeller, Bodo Moeller]
7626 *) In the NCONF_...-based implementations for CONF_... queries
7627 (crypto/conf/conf_lib.c), if the input LHASH is NULL, avoid using
7628 a temporary CONF structure with the data component set to NULL
7629 (which gives segmentation faults in lh_retrieve).
7630 Instead, use NULL for the CONF pointer in CONF_get_string and
7631 CONF_get_number (which may use environment variables) and directly
7632 return NULL from CONF_get_section.
7635 *) Fix potential buffer overrun for EBCDIC.
7638 *) Tolerate nonRepudiation as being valid for S/MIME signing and certSign
7639 keyUsage if basicConstraints absent for a CA.
7642 *) Make SMIME_write_PKCS7() write mail header values with a format that
7643 is more generally accepted (no spaces before the semicolon), since
7644 some programs can't parse those values properly otherwise. Also make
7645 sure BIO's that break lines after each write do not create invalid
7649 *) Make the CRL encoding routines work with empty SEQUENCE OF. The
7650 macros previously used would not encode an empty SEQUENCE OF
7651 and break the signature.
7653 [This change does not apply to 0.9.7.]
7655 *) Zero the premaster secret after deriving the master secret in
7659 *) Add some EVP_add_digest_alias registrations (as found in
7660 OpenSSL_add_all_digests()) to SSL_library_init()
7661 aka OpenSSL_add_ssl_algorithms(). This provides improved
7662 compatibility with peers using X.509 certificates
7663 with unconventional AlgorithmIdentifier OIDs.
7666 *) Fix for Irix with NO_ASM.
7667 ["Bruce W. Forsberg" <bruce.forsberg@baesystems.com>]
7669 *) ./config script fixes.
7670 [Ulf Moeller, Richard Levitte]
7672 *) Fix 'openssl passwd -1'.
7675 *) Change PKCS12_key_gen_asc() so it can cope with non null
7676 terminated strings whose length is passed in the passlen
7677 parameter, for example from PEM callbacks. This was done
7678 by adding an extra length parameter to asc2uni().
7679 [Steve Henson, reported by <oddissey@samsung.co.kr>]
7681 *) Fix C code generated by 'openssl dsaparam -C': If a BN_bin2bn
7682 call failed, free the DSA structure.
7685 *) Fix to uni2asc() to cope with zero length Unicode strings.
7686 These are present in some PKCS#12 files.
7689 *) Increase s2->wbuf allocation by one byte in ssl2_new (ssl/s2_lib.c).
7690 Otherwise do_ssl_write (ssl/s2_pkt.c) will write beyond buffer limits
7691 when writing a 32767 byte record.
7692 [Bodo Moeller; problem reported by Eric Day <eday@concentric.net>]
7694 *) In RSA_eay_public_{en,ed}crypt and RSA_eay_mod_exp (rsa_eay.c),
7695 obtain lock CRYPTO_LOCK_RSA before setting rsa->_method_mod_{n,p,q}.
7697 (RSA objects have a reference count access to which is protected
7698 by CRYPTO_LOCK_RSA [see rsa_lib.c, s3_srvr.c, ssl_cert.c, ssl_rsa.c],
7699 so they are meant to be shared between threads.)
7700 [Bodo Moeller, Geoff Thorpe; original patch submitted by
7701 "Reddie, Steven" <Steven.Reddie@ca.com>]
7703 *) Fix a deadlock in CRYPTO_mem_leaks().
7706 *) Use better test patterns in bntest.
7709 *) rand_win.c fix for Borland C.
7712 *) BN_rshift bugfix for n == 0.
7715 *) Add a 'bctest' script that checks for some known 'bc' bugs
7716 so that 'make test' does not abort just because 'bc' is broken.
7719 *) Store verify_result within SSL_SESSION also for client side to
7720 avoid potential security hole. (Re-used sessions on the client side
7721 always resulted in verify_result==X509_V_OK, not using the original
7722 result of the server certificate verification.)
7725 *) Fix ssl3_pending: If the record in s->s3->rrec is not of type
7726 SSL3_RT_APPLICATION_DATA, return 0.
7727 Similarly, change ssl2_pending to return 0 if SSL_in_init(s) is true.
7731 Both ssl2_peek and ssl3_peek, which were totally broken in earlier
7732 releases, have been re-implemented by renaming the previous
7733 implementations of ssl2_read and ssl3_read to ssl2_read_internal
7734 and ssl3_read_internal, respectively, and adding 'peek' parameters
7735 to them. The new ssl[23]_{read,peek} functions are calls to
7736 ssl[23]_read_internal with the 'peek' flag set appropriately.
7737 A 'peek' parameter has also been added to ssl3_read_bytes, which
7738 does the actual work for ssl3_read_internal.
7741 *) Initialise "ex_data" member of RSA/DSA/DH structures prior to calling
7742 the method-specific "init()" handler. Also clean up ex_data after
7743 calling the method-specific "finish()" handler. Previously, this was
7744 happening the other way round.
7747 *) Increase BN_CTX_NUM (the number of BIGNUMs in a BN_CTX) to 16.
7748 The previous value, 12, was not always sufficient for BN_mod_exp().
7751 *) Make sure that shared libraries get the internal name engine with
7752 the full version number and not just 0. This should mark the
7753 shared libraries as not backward compatible. Of course, this should
7754 be changed again when we can guarantee backward binary compatibility.
7757 *) Fix typo in get_cert_by_subject() in by_dir.c
7758 [Jean-Marc Desperrier <jean-marc.desperrier@certplus.com>]
7760 *) Rework the system to generate shared libraries:
7762 - Make note of the expected extension for the shared libraries and
7763 if there is a need for symbolic links from for example libcrypto.so.0
7764 to libcrypto.so.0.9.7. There is extended info in Configure for
7767 - Make as few rebuilds of the shared libraries as possible.
7769 - Still avoid linking the OpenSSL programs with the shared libraries.
7771 - When installing, install the shared libraries separately from the
7775 *) Fix SSL_CTX_set_read_ahead macro to actually use its argument.
7777 Copy SSL_CTX's read_ahead flag to SSL object directly in SSL_new
7778 and not in SSL_clear because the latter is also used by the
7779 accept/connect functions; previously, the settings made by
7780 SSL_set_read_ahead would be lost during the handshake.
7781 [Bodo Moeller; problems reported by Anders Gertz <gertz@epact.se>]
7783 *) Correct util/mkdef.pl to be selective about disabled algorithms.
7784 Previously, it would create entries for disableed algorithms no
7788 *) Added several new manual pages for SSL_* function.
7791 Changes between 0.9.5a and 0.9.6 [24 Sep 2000]
7793 *) In ssl23_get_client_hello, generate an error message when faced
7794 with an initial SSL 3.0/TLS record that is too small to contain the
7795 first two bytes of the ClientHello message, i.e. client_version.
7796 (Note that this is a pathologic case that probably has never happened
7797 in real life.) The previous approach was to use the version number
7798 from the record header as a substitute; but our protocol choice
7799 should not depend on that one because it is not authenticated
7800 by the Finished messages.
7803 *) More robust randomness gathering functions for Windows.
7804 [Jeffrey Altman <jaltman@columbia.edu>]
7806 *) For compatibility reasons if the flag X509_V_FLAG_ISSUER_CHECK is
7807 not set then we don't setup the error code for issuer check errors
7808 to avoid possibly overwriting other errors which the callback does
7809 handle. If an application does set the flag then we assume it knows
7810 what it is doing and can handle the new informational codes
7814 *) Fix for a nasty bug in ASN1_TYPE handling. ASN1_TYPE is used for
7815 a general "ANY" type, as such it should be able to decode anything
7816 including tagged types. However it didn't check the class so it would
7817 wrongly interpret tagged types in the same way as their universal
7818 counterpart and unknown types were just rejected. Changed so that the
7819 tagged and unknown types are handled in the same way as a SEQUENCE:
7820 that is the encoding is stored intact. There is also a new type
7821 "V_ASN1_OTHER" which is used when the class is not universal, in this
7822 case we have no idea what the actual type is so we just lump them all
7826 *) On VMS, stdout may very well lead to a file that is written to
7827 in a record-oriented fashion. That means that every write() will
7828 write a separate record, which will be read separately by the
7829 programs trying to read from it. This can be very confusing.
7831 The solution is to put a BIO filter in the way that will buffer
7832 text until a linefeed is reached, and then write everything a
7833 line at a time, so every record written will be an actual line,
7834 not chunks of lines and not (usually doesn't happen, but I've
7835 seen it once) several lines in one record. BIO_f_linebuffer() is
7838 Currently, it's a VMS-only method, because that's where it has
7839 been tested well enough.
7842 *) Remove 'optimized' squaring variant in BN_mod_mul_montgomery,
7843 it can return incorrect results.
7844 (Note: The buggy variant was not enabled in OpenSSL 0.9.5a,
7845 but it was in 0.9.6-beta[12].)
7848 *) Disable the check for content being present when verifying detached
7849 signatures in pk7_smime.c. Some versions of Netscape (wrongly)
7850 include zero length content when signing messages.
7853 *) New BIO_shutdown_wr macro, which invokes the BIO_C_SHUTDOWN_WR
7854 BIO_ctrl (for BIO pairs).
7857 *) Add DSO method for VMS.
7860 *) Bug fix: Montgomery multiplication could produce results with the
7864 *) Add RPM specification openssl.spec and modify it to build three
7865 packages. The default package contains applications, application
7866 documentation and run-time libraries. The devel package contains
7867 include files, static libraries and function documentation. The
7868 doc package contains the contents of the doc directory. The original
7869 openssl.spec was provided by Damien Miller <djm@mindrot.org>.
7872 *) Add a large number of documentation files for many SSL routines.
7873 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
7875 *) Add a configuration entry for Sony News 4.
7876 [NAKAJI Hiroyuki <nakaji@tutrp.tut.ac.jp>]
7878 *) Don't set the two most significant bits to one when generating a
7879 random number < q in the DSA library.
7882 *) New SSL API mode 'SSL_MODE_AUTO_RETRY'. This disables the default
7883 behaviour that SSL_read may result in SSL_ERROR_WANT_READ (even if
7884 the underlying transport is blocking) if a handshake took place.
7885 (The default behaviour is needed by applications such as s_client
7886 and s_server that use select() to determine when to use SSL_read;
7887 but for applications that know in advance when to expect data, it
7888 just makes things more complicated.)
7891 *) Add RAND_egd_bytes(), which gives control over the number of bytes read
7895 *) Add a few more EBCDIC conditionals that make `req' and `x509'
7896 work better on such systems.
7897 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
7899 *) Add two demo programs for PKCS12_parse() and PKCS12_create().
7900 Update PKCS12_parse() so it copies the friendlyName and the
7901 keyid to the certificates aux info.
7904 *) Fix bug in PKCS7_verify() which caused an infinite loop
7905 if there was more than one signature.
7906 [Sven Uszpelkat <su@celocom.de>]
7908 *) Major change in util/mkdef.pl to include extra information
7909 about each symbol, as well as presentig variables as well
7910 as functions. This change means that there's n more need
7911 to rebuild the .num files when some algorithms are excluded.
7914 *) Allow the verify time to be set by an application,
7915 rather than always using the current time.
7918 *) Phase 2 verify code reorganisation. The certificate
7919 verify code now looks up an issuer certificate by a
7920 number of criteria: subject name, authority key id
7921 and key usage. It also verifies self signed certificates
7922 by the same criteria. The main comparison function is
7923 X509_check_issued() which performs these checks.
7925 Lot of changes were necessary in order to support this
7926 without completely rewriting the lookup code.
7928 Authority and subject key identifier are now cached.
7930 The LHASH 'certs' is X509_STORE has now been replaced
7931 by a STACK_OF(X509_OBJECT). This is mainly because an
7932 LHASH can't store or retrieve multiple objects with
7933 the same hash value.
7935 As a result various functions (which were all internal
7936 use only) have changed to handle the new X509_STORE
7937 structure. This will break anything that messed round
7938 with X509_STORE internally.
7940 The functions X509_STORE_add_cert() now checks for an
7941 exact match, rather than just subject name.
7943 The X509_STORE API doesn't directly support the retrieval
7944 of multiple certificates matching a given criteria, however
7945 this can be worked round by performing a lookup first
7946 (which will fill the cache with candidate certificates)
7947 and then examining the cache for matches. This is probably
7948 the best we can do without throwing out X509_LOOKUP
7949 entirely (maybe later...).
7951 The X509_VERIFY_CTX structure has been enhanced considerably.
7953 All certificate lookup operations now go via a get_issuer()
7954 callback. Although this currently uses an X509_STORE it
7955 can be replaced by custom lookups. This is a simple way
7956 to bypass the X509_STORE hackery necessary to make this
7957 work and makes it possible to use more efficient techniques
7958 in future. A very simple version which uses a simple
7959 STACK for its trusted certificate store is also provided
7960 using X509_STORE_CTX_trusted_stack().
7962 The verify_cb() and verify() callbacks now have equivalents
7963 in the X509_STORE_CTX structure.
7965 X509_STORE_CTX also has a 'flags' field which can be used
7966 to customise the verify behaviour.
7969 *) Add new PKCS#7 signing option PKCS7_NOSMIMECAP which
7970 excludes S/MIME capabilities.
7973 *) When a certificate request is read in keep a copy of the
7974 original encoding of the signed data and use it when outputing
7975 again. Signatures then use the original encoding rather than
7976 a decoded, encoded version which may cause problems if the
7977 request is improperly encoded.
7980 *) For consistency with other BIO_puts implementations, call
7981 buffer_write(b, ...) directly in buffer_puts instead of calling
7984 In BIO_puts, increment b->num_write as in BIO_write.
7985 [Peter.Sylvester@EdelWeb.fr]
7987 *) Fix BN_mul_word for the case where the word is 0. (We have to use
7988 BN_zero, we may not return a BIGNUM with an array consisting of
7992 *) Avoid calling abort() from within the library when problems are
7993 detected, except if preprocessor symbols have been defined
7994 (such as REF_CHECK, BN_DEBUG etc.).
7997 *) New openssl application 'rsautl'. This utility can be
7998 used for low level RSA operations. DER public key
7999 BIO/fp routines also added.
8002 *) New Configure entry and patches for compiling on QNX 4.
8003 [Andreas Schneider <andreas@ds3.etech.fh-hamburg.de>]
8005 *) A demo state-machine implementation was sponsored by
8006 Nuron (http://www.nuron.com/) and is now available in
8007 demos/state_machine.
8010 *) New options added to the 'dgst' utility for signature
8011 generation and verification.
8014 *) Unrecognized PKCS#7 content types are now handled via a
8015 catch all ASN1_TYPE structure. This allows unsupported
8016 types to be stored as a "blob" and an application can
8017 encode and decode it manually.
8020 *) Fix various signed/unsigned issues to make a_strex.c
8022 [Oscar Jacobsson <oscar.jacobsson@celocom.com>]
8024 *) ASN1 fixes. i2d_ASN1_OBJECT was not returning the correct
8025 length if passed a buffer. ASN1_INTEGER_to_BN failed
8026 if passed a NULL BN and its argument was negative.
8027 [Steve Henson, pointed out by Sven Heiberg <sven@tartu.cyber.ee>]
8029 *) Modification to PKCS#7 encoding routines to output definite
8030 length encoding. Since currently the whole structures are in
8031 memory there's not real point in using indefinite length
8032 constructed encoding. However if OpenSSL is compiled with
8033 the flag PKCS7_INDEFINITE_ENCODING the old form is used.
8036 *) Added BIO_vprintf() and BIO_vsnprintf().
8039 *) Added more prefixes to parse for in the the strings written
8040 through a logging bio, to cover all the levels that are available
8041 through syslog. The prefixes are now:
8043 PANIC, EMERG, EMR => LOG_EMERG
8044 ALERT, ALR => LOG_ALERT
8045 CRIT, CRI => LOG_CRIT
8046 ERROR, ERR => LOG_ERR
8047 WARNING, WARN, WAR => LOG_WARNING
8048 NOTICE, NOTE, NOT => LOG_NOTICE
8049 INFO, INF => LOG_INFO
8050 DEBUG, DBG => LOG_DEBUG
8052 and as before, if none of those prefixes are present at the
8053 beginning of the string, LOG_ERR is chosen.
8055 On Win32, the LOG_* levels are mapped according to this:
8057 LOG_EMERG, LOG_ALERT, LOG_CRIT, LOG_ERR => EVENTLOG_ERROR_TYPE
8058 LOG_WARNING => EVENTLOG_WARNING_TYPE
8059 LOG_NOTICE, LOG_INFO, LOG_DEBUG => EVENTLOG_INFORMATION_TYPE
8063 *) Made it possible to reconfigure with just the configuration
8064 argument "reconf" or "reconfigure". The command line arguments
8065 are stored in Makefile.ssl in the variable CONFIGURE_ARGS,
8066 and are retrieved from there when reconfiguring.
8070 [Assar Westerlund <assar@sics.se>, Richard Levitte]
8072 *) Add the arguments -CAfile and -CApath to the pkcs12 utility.
8075 *) The obj_dat.pl script was messing up the sorting of object
8076 names. The reason was that it compared the quoted version
8077 of strings as a result "OCSP" > "OCSP Signing" because
8078 " > SPACE. Changed script to store unquoted versions of
8079 names and add quotes on output. It was also omitting some
8080 names from the lookup table if they were given a default
8081 value (that is if SN is missing it is given the same
8082 value as LN and vice versa), these are now added on the
8083 grounds that if an object has a name we should be able to
8084 look it up. Finally added warning output when duplicate
8085 short or long names are found.
8088 *) Changes needed for Tandem NSK.
8089 [Scott Uroff <scott@xypro.com>]
8091 *) Fix SSL 2.0 rollback checking: Due to an off-by-one error in
8092 RSA_padding_check_SSLv23(), special padding was never detected
8093 and thus the SSL 3.0/TLS 1.0 countermeasure against protocol
8094 version rollback attacks was not effective.
8096 In s23_clnt.c, don't use special rollback-attack detection padding
8097 (RSA_SSLV23_PADDING) if SSL 2.0 is the only protocol enabled in the
8098 client; similarly, in s23_srvr.c, don't do the rollback check if
8099 SSL 2.0 is the only protocol enabled in the server.
8102 *) Make it possible to get hexdumps of unprintable data with 'openssl
8103 asn1parse'. By implication, the functions ASN1_parse_dump() and
8104 BIO_dump_indent() are added.
8107 *) New functions ASN1_STRING_print_ex() and X509_NAME_print_ex()
8108 these print out strings and name structures based on various
8109 flags including RFC2253 support and proper handling of
8110 multibyte characters. Added options to the 'x509' utility
8111 to allow the various flags to be set.
8114 *) Various fixes to use ASN1_TIME instead of ASN1_UTCTIME.
8115 Also change the functions X509_cmp_current_time() and
8116 X509_gmtime_adj() work with an ASN1_TIME structure,
8117 this will enable certificates using GeneralizedTime in validity
8118 dates to be checked.
8121 *) Make the NEG_PUBKEY_BUG code (which tolerates invalid
8122 negative public key encodings) on by default,
8123 NO_NEG_PUBKEY_BUG can be set to disable it.
8126 *) New function c2i_ASN1_OBJECT() which acts on ASN1_OBJECT
8127 content octets. An i2c_ASN1_OBJECT is unnecessary because
8128 the encoding can be trivially obtained from the structure.
8131 *) crypto/err.c locking bugfix: Use write locks (CRYPTO_w_[un]lock),
8132 not read locks (CRYPTO_r_[un]lock).
8135 *) A first attempt at creating official support for shared
8136 libraries through configuration. I've kept it so the
8137 default is static libraries only, and the OpenSSL programs
8138 are always statically linked for now, but there are
8139 preparations for dynamic linking in place.
8140 This has been tested on Linux and Tru64.
8143 *) Randomness polling function for Win9x, as described in:
8144 Peter Gutmann, Software Generation of Practically Strong
8148 *) Fix so PRNG is seeded in req if using an already existing
8152 *) New options to smime application. -inform and -outform
8153 allow alternative formats for the S/MIME message including
8154 PEM and DER. The -content option allows the content to be
8155 specified separately. This should allow things like Netscape
8156 form signing output easier to verify.
8159 *) Fix the ASN1 encoding of tags using the 'long form'.
8162 *) New ASN1 functions, i2c_* and c2i_* for INTEGER and BIT
8163 STRING types. These convert content octets to and from the
8164 underlying type. The actual tag and length octets are
8165 already assumed to have been read in and checked. These
8166 are needed because all other string types have virtually
8167 identical handling apart from the tag. By having versions
8168 of the ASN1 functions that just operate on content octets
8169 IMPLICIT tagging can be handled properly. It also allows
8170 the ASN1_ENUMERATED code to be cut down because ASN1_ENUMERATED
8171 and ASN1_INTEGER are identical apart from the tag.
8174 *) Change the handling of OID objects as follows:
8176 - New object identifiers are inserted in objects.txt, following
8177 the syntax given in objects.README.
8178 - objects.pl is used to process obj_mac.num and create a new
8180 - obj_dat.pl is used to create a new obj_dat.h, using the data in
8183 This is currently kind of a hack, and the perl code in objects.pl
8184 isn't very elegant, but it works as I intended. The simplest way
8185 to check that it worked correctly is to look in obj_dat.h and
8186 check the array nid_objs and make sure the objects haven't moved
8187 around (this is important!). Additions are OK, as well as
8188 consistent name changes.
8191 *) Add BSD-style MD5-based passwords to 'openssl passwd' (option '-1').
8194 *) Addition of the command line parameter '-rand file' to 'openssl req'.
8195 The given file adds to whatever has already been seeded into the
8196 random pool through the RANDFILE configuration file option or
8197 environment variable, or the default random state file.
8200 *) mkstack.pl now sorts each macro group into lexical order.
8201 Previously the output order depended on the order the files
8202 appeared in the directory, resulting in needless rewriting
8206 *) Patches to make OpenSSL compile under Win32 again. Mostly
8207 work arounds for the VC++ problem that it treats func() as
8208 func(void). Also stripped out the parts of mkdef.pl that
8209 added extra typesafe functions: these no longer exist.
8212 *) Reorganisation of the stack code. The macros are now all
8213 collected in safestack.h . Each macro is defined in terms of
8214 a "stack macro" of the form SKM_<name>(type, a, b). The
8215 DEBUG_SAFESTACK is now handled in terms of function casts,
8216 this has the advantage of retaining type safety without the
8217 use of additional functions. If DEBUG_SAFESTACK is not defined
8218 then the non typesafe macros are used instead. Also modified the
8219 mkstack.pl script to handle the new form. Needs testing to see
8220 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK
8221 the default if no major problems. Similar behaviour for ASN1_SET_OF
8222 and PKCS12_STACK_OF.
8225 *) When some versions of IIS use the 'NET' form of private key the
8226 key derivation algorithm is different. Normally MD5(password) is
8227 used as a 128 bit RC4 key. In the modified case
8228 MD5(MD5(password) + "SGCKEYSALT") is used insted. Added some
8229 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same
8230 as the old Netscape_RSA functions except they have an additional
8231 'sgckey' parameter which uses the modified algorithm. Also added
8232 an -sgckey command line option to the rsa utility. Thanks to
8233 Adrian Peck <bertie@ncipher.com> for posting details of the modified
8234 algorithm to openssl-dev.
8237 *) The evp_local.h macros were using 'c.##kname' which resulted in
8238 invalid expansion on some systems (SCO 5.0.5 for example).
8239 Corrected to 'c.kname'.
8240 [Phillip Porch <root@theporch.com>]
8242 *) New X509_get1_email() and X509_REQ_get1_email() functions that return
8243 a STACK of email addresses from a certificate or request, these look
8244 in the subject name and the subject alternative name extensions and
8245 omit any duplicate addresses.
8248 *) Re-implement BN_mod_exp2_mont using independent (and larger) windows.
8249 This makes DSA verification about 2 % faster.
8252 *) Increase maximum window size in BN_mod_exp_... to 6 bits instead of 5
8253 (meaning that now 2^5 values will be precomputed, which is only 4 KB
8254 plus overhead for 1024 bit moduli).
8255 This makes exponentiations about 0.5 % faster for 1024 bit
8256 exponents (as measured by "openssl speed rsa2048").
8259 *) Rename memory handling macros to avoid conflicts with other
8261 Malloc => OPENSSL_malloc
8262 Malloc_locked => OPENSSL_malloc_locked
8263 Realloc => OPENSSL_realloc
8264 Free => OPENSSL_free
8267 *) New function BN_mod_exp_mont_word for small bases (roughly 15%
8268 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange).
8271 *) CygWin32 support.
8272 [John Jarvie <jjarvie@newsguy.com>]
8274 *) The type-safe stack code has been rejigged. It is now only compiled
8275 in when OpenSSL is configured with the DEBUG_SAFESTACK option and
8276 by default all type-specific stack functions are "#define"d back to
8277 standard stack functions. This results in more streamlined output
8278 but retains the type-safety checking possibilities of the original
8282 *) The STACK code has been cleaned up, and certain type declarations
8283 that didn't make a lot of sense have been brought in line. This has
8284 also involved a cleanup of sorts in safestack.h to more correctly
8285 map type-safe stack functions onto their plain stack counterparts.
8286 This work has also resulted in a variety of "const"ifications of
8287 lots of the code, especially "_cmp" operations which should normally
8288 be prototyped with "const" parameters anyway.
8291 *) When generating bytes for the first time in md_rand.c, 'stir the pool'
8292 by seeding with STATE_SIZE dummy bytes (with zero entropy count).
8293 (The PRNG state consists of two parts, the large pool 'state' and 'md',
8294 where all of 'md' is used each time the PRNG is used, but 'state'
8295 is used only indexed by a cyclic counter. As entropy may not be
8296 well distributed from the beginning, 'md' is important as a
8297 chaining variable. However, the output function chains only half
8298 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains
8299 all of 'md', and seeding with STATE_SIZE dummy bytes will result
8300 in all of 'state' being rewritten, with the new values depending
8301 on virtually all of 'md'. This overcomes the 80 bit limitation.)
8304 *) In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when
8305 the handshake is continued after ssl_verify_cert_chain();
8306 otherwise, if SSL_VERIFY_NONE is set, remaining error codes
8307 can lead to 'unexplainable' connection aborts later.
8308 [Bodo Moeller; problem tracked down by Lutz Jaenicke]
8310 *) Major EVP API cipher revision.
8311 Add hooks for extra EVP features. This allows various cipher
8312 parameters to be set in the EVP interface. Support added for variable
8313 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and
8314 setting of RC2 and RC5 parameters.
8316 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length
8319 Remove lots of duplicated code from the EVP library. For example *every*
8320 cipher init() function handles the 'iv' in the same way according to the
8321 cipher mode. They also all do nothing if the 'key' parameter is NULL and
8322 for CFB and OFB modes they zero ctx->num.
8324 New functionality allows removal of S/MIME code RC2 hack.
8326 Most of the routines have the same form and so can be declared in terms
8329 By shifting this to the top level EVP_CipherInit() it can be removed from
8330 all individual ciphers. If the cipher wants to handle IVs or keys
8331 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT
8334 Change lots of functions like EVP_EncryptUpdate() to now return a
8335 value: although software versions of the algorithms cannot fail
8336 any installed hardware versions can.
8339 *) Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if
8340 this option is set, tolerate broken clients that send the negotiated
8341 protocol version number instead of the requested protocol version
8345 *) Call dh_tmp_cb (set by ..._TMP_DH_CB) with correct 'is_export' flag;
8346 i.e. non-zero for export ciphersuites, zero otherwise.
8347 Previous versions had this flag inverted, inconsistent with
8348 rsa_tmp_cb (..._TMP_RSA_CB).
8349 [Bodo Moeller; problem reported by Amit Chopra]
8351 *) Add missing DSA library text string. Work around for some IIS
8352 key files with invalid SEQUENCE encoding.
8355 *) Add a document (doc/standards.txt) that list all kinds of standards
8356 and so on that are implemented in OpenSSL.
8359 *) Enhance c_rehash script. Old version would mishandle certificates
8360 with the same subject name hash and wouldn't handle CRLs at all.
8361 Added -fingerprint option to crl utility, to support new c_rehash
8365 *) Eliminate non-ANSI declarations in crypto.h and stack.h.
8368 *) Fix for SSL server purpose checking. Server checking was
8369 rejecting certificates which had extended key usage present
8370 but no ssl client purpose.
8371 [Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>]
8373 *) Make PKCS#12 code work with no password. The PKCS#12 spec
8374 is a little unclear about how a blank password is handled.
8375 Since the password in encoded as a BMPString with terminating
8376 double NULL a zero length password would end up as just the
8377 double NULL. However no password at all is different and is
8378 handled differently in the PKCS#12 key generation code. NS
8379 treats a blank password as zero length. MSIE treats it as no
8380 password on export: but it will try both on import. We now do
8381 the same: PKCS12_parse() tries zero length and no password if
8382 the password is set to "" or NULL (NULL is now a valid password:
8383 it wasn't before) as does the pkcs12 application.
8386 *) Bugfixes in apps/x509.c: Avoid a memory leak; and don't use
8387 perror when PEM_read_bio_X509_REQ fails, the error message must
8388 be obtained from the error queue.
8391 *) Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing
8392 it in ERR_remove_state if appropriate, and change ERR_get_state
8393 accordingly to avoid race conditions (this is necessary because
8394 thread_hash is no longer constant once set).
8397 *) Bugfix for linux-elf makefile.one.
8400 *) RSA_get_default_method() will now cause a default
8401 RSA_METHOD to be chosen if one doesn't exist already.
8402 Previously this was only set during a call to RSA_new()
8403 or RSA_new_method(NULL) meaning it was possible for
8404 RSA_get_default_method() to return NULL.
8407 *) Added native name translation to the existing DSO code
8408 that will convert (if the flag to do so is set) filenames
8409 that are sufficiently small and have no path information
8410 into a canonical native form. Eg. "blah" converted to
8411 "libblah.so" or "blah.dll" etc.
8414 *) New function ERR_error_string_n(e, buf, len) which is like
8415 ERR_error_string(e, buf), but writes at most 'len' bytes
8416 including the 0 terminator. For ERR_error_string_n, 'buf'
8418 [Damien Miller <djm@mindrot.org>, Bodo Moeller]
8420 *) CONF library reworked to become more general. A new CONF
8421 configuration file reader "class" is implemented as well as a
8422 new functions (NCONF_*, for "New CONF") to handle it. The now
8423 old CONF_* functions are still there, but are reimplemented to
8424 work in terms of the new functions. Also, a set of functions
8425 to handle the internal storage of the configuration data is
8426 provided to make it easier to write new configuration file
8427 reader "classes" (I can definitely see something reading a
8428 configuration file in XML format, for example), called _CONF_*,
8429 or "the configuration storage API"...
8431 The new configuration file reading functions are:
8433 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio,
8434 NCONF_get_section, NCONF_get_string, NCONF_get_numbre
8436 NCONF_default, NCONF_WIN32
8438 NCONF_dump_fp, NCONF_dump_bio
8440 NCONF_default and NCONF_WIN32 are method (or "class") choosers,
8441 NCONF_new creates a new CONF object. This works in the same way
8442 as other interfaces in OpenSSL, like the BIO interface.
8443 NCONF_dump_* dump the internal storage of the configuration file,
8444 which is useful for debugging. All other functions take the same
8445 arguments as the old CONF_* functions wth the exception of the
8446 first that must be a `CONF *' instead of a `LHASH *'.
8448 To make it easer to use the new classes with the old CONF_* functions,
8449 the function CONF_set_default_method is provided.
8452 *) Add '-tls1' option to 'openssl ciphers', which was already
8453 mentioned in the documentation but had not been implemented.
8454 (This option is not yet really useful because even the additional
8455 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.)
8458 *) Initial DSO code added into libcrypto for letting OpenSSL (and
8459 OpenSSL-based applications) load shared libraries and bind to
8460 them in a portable way.
8461 [Geoff Thorpe, with contributions from Richard Levitte]
8463 Changes between 0.9.5 and 0.9.5a [1 Apr 2000]
8465 *) Make sure _lrotl and _lrotr are only used with MSVC.
8467 *) Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status
8468 (the default implementation of RAND_status).
8470 *) Rename openssl x509 option '-crlext', which was added in 0.9.5,
8471 to '-clrext' (= clear extensions), as intended and documented.
8472 [Bodo Moeller; inconsistency pointed out by Michael Attili
8475 *) Fix for HMAC. It wasn't zeroing the rest of the block if the key length
8476 was larger than the MD block size.
8477 [Steve Henson, pointed out by Yost William <YostW@tce.com>]
8479 *) Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument
8480 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set()
8481 using the passed key: if the passed key was a private key the result
8482 of X509_print(), for example, would be to print out all the private key
8486 *) des_quad_cksum() byte order bug fix.
8487 [Ulf Möller, using the problem description in krb4-0.9.7, where
8488 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>]
8490 *) Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly
8492 [Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>]
8494 *) For easily testing in shell scripts whether some command
8495 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX'
8496 returns with exit code 0 iff no command of the given name is available.
8497 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases,
8498 the output goes to stdout and nothing is printed to stderr.
8499 Additional arguments are always ignored.
8501 Since for each cipher there is a command of the same name,
8502 the 'no-cipher' compilation switches can be tested this way.
8504 ('openssl no-XXX' is not able to detect pseudo-commands such
8505 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.)
8508 *) Update test suite so that 'make test' succeeds in 'no-rsa' configuration.
8511 *) For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE
8512 is set; it will be thrown away anyway because each handshake creates
8514 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition
8515 to parameters -- in previous versions (since OpenSSL 0.9.3) the
8516 'default key' from SSL_CTX_set_tmp_dh would always be lost, meanining
8517 you effectivly got SSL_OP_SINGLE_DH_USE when using this macro.
8520 *) New s_client option -ign_eof: EOF at stdin is ignored, and
8521 'Q' and 'R' lose their special meanings (quit/renegotiate).
8522 This is part of what -quiet does; unlike -quiet, -ign_eof
8523 does not suppress any output.
8526 *) Add compatibility options to the purpose and trust code. The
8527 purpose X509_PURPOSE_ANY is "any purpose" which automatically
8528 accepts a certificate or CA, this was the previous behaviour,
8529 with all the associated security issues.
8531 X509_TRUST_COMPAT is the old trust behaviour: only and
8532 automatically trust self signed roots in certificate store. A
8533 new trust setting X509_TRUST_DEFAULT is used to specify that
8534 a purpose has no associated trust setting and it should instead
8535 use the value in the default purpose.
8538 *) Fix the PKCS#8 DSA private key code so it decodes keys again
8539 and fix a memory leak.
8542 *) In util/mkerr.pl (which implements 'make errors'), preserve
8543 reason strings from the previous version of the .c file, as
8544 the default to have only downcase letters (and digits) in
8545 automatically generated reasons codes is not always appropriate.
8548 *) In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table
8549 using strerror. Previously, ERR_reason_error_string() returned
8550 library names as reason strings for SYSerr; but SYSerr is a special
8551 case where small numbers are errno values, not library numbers.
8554 *) Add '-dsaparam' option to 'openssl dhparam' application. This
8555 converts DSA parameters into DH parameters. (When creating parameters,
8556 DSA_generate_parameters is used.)
8559 *) Include 'length' (recommended exponent length) in C code generated
8560 by 'openssl dhparam -C'.
8563 *) The second argument to set_label in perlasm was already being used
8564 so couldn't be used as a "file scope" flag. Moved to third argument
8568 *) In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes
8569 instead of RAND_bytes for encryption IVs and salts.
8572 *) Include RAND_status() into RAND_METHOD instead of implementing
8573 it only for md_rand.c Otherwise replacing the PRNG by calling
8574 RAND_set_rand_method would be impossible.
8577 *) Don't let DSA_generate_key() enter an infinite loop if the random
8578 number generation fails.
8581 *) New 'rand' application for creating pseudo-random output.
8584 *) Added configuration support for Linux/IA64
8585 [Rolf Haberrecker <rolf@suse.de>]
8587 *) Assembler module support for Mingw32.
8590 *) Shared library support for HPUX (in shlib/).
8591 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous]
8593 *) Shared library support for Solaris gcc.
8594 [Lutz Behnke <behnke@trustcenter.de>]
8596 Changes between 0.9.4 and 0.9.5 [28 Feb 2000]
8598 *) PKCS7_encrypt() was adding text MIME headers twice because they
8599 were added manually and by SMIME_crlf_copy().
8602 *) In bntest.c don't call BN_rand with zero bits argument.
8603 [Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>]
8605 *) BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n]
8606 case was implemented. This caused BN_div_recp() to fail occasionally.
8609 *) Add an optional second argument to the set_label() in the perl
8610 assembly language builder. If this argument exists and is set
8611 to 1 it signals that the assembler should use a symbol whose
8612 scope is the entire file, not just the current function. This
8613 is needed with MASM which uses the format label:: for this scope.
8614 [Steve Henson, pointed out by Peter Runestig <peter@runestig.com>]
8616 *) Change the ASN1 types so they are typedefs by default. Before
8617 almost all types were #define'd to ASN1_STRING which was causing
8618 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING)
8622 *) Change names of new functions to the new get1/get0 naming
8623 convention: After 'get1', the caller owns a reference count
8624 and has to call ..._free; 'get0' returns a pointer to some
8625 data structure without incrementing reference counters.
8626 (Some of the existing 'get' functions increment a reference
8627 counter, some don't.)
8628 Similarly, 'set1' and 'add1' functions increase reference
8629 counters or duplicate objects.
8632 *) Allow for the possibility of temp RSA key generation failure:
8633 the code used to assume it always worked and crashed on failure.
8636 *) Fix potential buffer overrun problem in BIO_printf().
8637 [Ulf Möller, using public domain code by Patrick Powell; problem
8638 pointed out by David Sacerdote <das33@cornell.edu>]
8640 *) Support EGD <http://www.lothar.com/tech/crypto/>. New functions
8641 RAND_egd() and RAND_status(). In the command line application,
8642 the EGD socket can be specified like a seed file using RANDFILE
8646 *) Allow the string CERTIFICATE to be tolerated in PKCS#7 structures.
8647 Some CAs (e.g. Verisign) distribute certificates in this form.
8650 *) Remove the SSL_ALLOW_ADH compile option and set the default cipher
8651 list to exclude them. This means that no special compilation option
8652 is needed to use anonymous DH: it just needs to be included in the
8656 *) Change the EVP_MD_CTX_type macro so its meaning consistent with
8657 EVP_MD_type. The old functionality is available in a new macro called
8658 EVP_MD_md(). Change code that uses it and update docs.
8661 *) ..._ctrl functions now have corresponding ..._callback_ctrl functions
8662 where the 'void *' argument is replaced by a function pointer argument.
8663 Previously 'void *' was abused to point to functions, which works on
8664 many platforms, but is not correct. As these functions are usually
8665 called by macros defined in OpenSSL header files, most source code
8666 should work without changes.
8669 *) <openssl/opensslconf.h> (which is created by Configure) now contains
8670 sections with information on -D... compiler switches used for
8671 compiling the library so that applications can see them. To enable
8672 one of these sections, a pre-processor symbol OPENSSL_..._DEFINES
8673 must be defined. E.g.,
8674 #define OPENSSL_ALGORITHM_DEFINES
8675 #include <openssl/opensslconf.h>
8676 defines all pertinent NO_<algo> symbols, such as NO_IDEA, NO_RSA, etc.
8677 [Richard Levitte, Ulf and Bodo Möller]
8679 *) Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS
8683 *) Change the 'other' type in certificate aux info to a STACK_OF
8684 X509_ALGOR. Although not an AlgorithmIdentifier as such it has
8685 the required ASN1 format: arbitrary types determined by an OID.
8688 *) Add some PEM_write_X509_REQ_NEW() functions and a command line
8689 argument to 'req'. This is not because the function is newer or
8690 better than others it just uses the work 'NEW' in the certificate
8691 request header lines. Some software needs this.
8694 *) Reorganise password command line arguments: now passwords can be
8695 obtained from various sources. Delete the PEM_cb function and make
8696 it the default behaviour: i.e. if the callback is NULL and the
8697 usrdata argument is not NULL interpret it as a null terminated pass
8698 phrase. If usrdata and the callback are NULL then the pass phrase
8699 is prompted for as usual.
8702 *) Add support for the Compaq Atalla crypto accelerator. If it is installed,
8703 the support is automatically enabled. The resulting binaries will
8704 autodetect the card and use it if present.
8705 [Ben Laurie and Compaq Inc.]
8707 *) Work around for Netscape hang bug. This sends certificate request
8708 and server done in one record. Since this is perfectly legal in the
8709 SSL/TLS protocol it isn't a "bug" option and is on by default. See
8710 the bugs/SSLv3 entry for more info.
8713 *) HP-UX tune-up: new unified configs, HP C compiler bug workaround.
8716 *) Add -rand argument to smime and pkcs12 applications and read/write
8720 *) New 'passwd' tool for crypt(3) and apr1 password hashes.
8723 *) Add command line password options to the remaining applications.
8726 *) Bug fix for BN_div_recp() for numerators with an even number of
8730 *) More tests in bntest.c, and changed test_bn output.
8733 *) ./config recognizes MacOS X now.
8736 *) Bug fix for BN_div() when the first words of num and divsor are
8737 equal (it gave wrong results if (rem=(n1-q*d0)&BN_MASK2) < d0).
8740 *) Add support for various broken PKCS#8 formats, and command line
8741 options to produce them.
8744 *) New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to
8745 get temporary BIGNUMs from a BN_CTX.
8748 *) Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont()
8752 *) Change the SSLeay_add_all_*() functions to OpenSSL_add_all_*() and
8753 include a #define from the old name to the new. The original intent
8754 was that statically linked binaries could for example just call
8755 SSLeay_add_all_ciphers() to just add ciphers to the table and not
8756 link with digests. This never worked becayse SSLeay_add_all_digests()
8757 and SSLeay_add_all_ciphers() were in the same source file so calling
8758 one would link with the other. They are now in separate source files.
8761 *) Add a new -notext option to 'ca' and a -pubkey option to 'spkac'.
8764 *) Use a less unusual form of the Miller-Rabin primality test (it used
8765 a binary algorithm for exponentiation integrated into the Miller-Rabin
8766 loop, our standard modexp algorithms are faster).
8769 *) Support for the EBCDIC character set completed.
8770 [Martin Kraemer <Martin.Kraemer@Mch.SNI.De>]
8772 *) Source code cleanups: use const where appropriate, eliminate casts,
8773 use void * instead of char * in lhash.
8776 *) Bugfix: ssl3_send_server_key_exchange was not restartable
8777 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of
8778 this the server could overwrite ephemeral keys that the client
8782 *) Turn DSA_is_prime into a macro that calls BN_is_prime,
8783 using 50 iterations of the Rabin-Miller test.
8785 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50
8786 iterations of the Rabin-Miller test as required by the appendix
8787 to FIPS PUB 186[-1]) instead of DSA_is_prime.
8788 As BN_is_prime_fasttest includes trial division, DSA parameter
8789 generation becomes much faster.
8791 This implies a change for the callback functions in DSA_is_prime
8792 and DSA_generate_parameters: The callback function is called once
8793 for each positive witness in the Rabin-Miller test, not just
8794 occasionally in the inner loop; and the parameters to the
8795 callback function now provide an iteration count for the outer
8796 loop rather than for the current invocation of the inner loop.
8797 DSA_generate_parameters additionally can call the callback
8798 function with an 'iteration count' of -1, meaning that a
8799 candidate has passed the trial division test (when q is generated
8800 from an application-provided seed, trial division is skipped).
8803 *) New function BN_is_prime_fasttest that optionally does trial
8804 division before starting the Rabin-Miller test and has
8805 an additional BN_CTX * argument (whereas BN_is_prime always
8806 has to allocate at least one BN_CTX).
8807 'callback(1, -1, cb_arg)' is called when a number has passed the
8808 trial division stage.
8811 *) Fix for bug in CRL encoding. The validity dates weren't being handled
8815 *) New -pkcs12 option to CA.pl script to write out a PKCS#12 file.
8818 *) New function BN_pseudo_rand().
8821 *) Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable)
8822 bignum version of BN_from_montgomery() with the working code from
8823 SSLeay 0.9.0 (the word based version is faster anyway), and clean up
8827 *) Avoid a race condition in s2_clnt.c (function get_server_hello) that
8828 made it impossible to use the same SSL_SESSION data structure in
8829 SSL2 clients in multiple threads.
8832 *) The return value of RAND_load_file() no longer counts bytes obtained
8833 by stat(). RAND_load_file(..., -1) is new and uses the complete file
8834 to seed the PRNG (previously an explicit byte count was required).
8835 [Ulf Möller, Bodo Möller]
8837 *) Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
8838 used (char *) instead of (void *) and had casts all over the place.
8841 *) Make BN_generate_prime() return NULL on error if ret!=NULL.
8844 *) Retain source code compatibility for BN_prime_checks macro:
8845 BN_is_prime(..., BN_prime_checks, ...) now uses
8846 BN_prime_checks_for_size to determine the appropriate number of
8847 Rabin-Miller iterations.
8850 *) Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
8851 DH_CHECK_P_NOT_SAFE_PRIME.
8852 (Check if this is true? OpenPGP calls them "strong".)
8855 *) Merge the functionality of "dh" and "gendh" programs into a new program
8856 "dhparam". The old programs are retained for now but will handle DH keys
8857 (instead of parameters) in future.
8860 *) Make the ciphers, s_server and s_client programs check the return values
8861 when a new cipher list is set.
8864 *) Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
8865 ciphers. Before when the 56bit ciphers were enabled the sorting was
8868 The syntax for the cipher sorting has been extended to support sorting by
8869 cipher-strength (using the strength_bits hard coded in the tables).
8870 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
8872 Fix a bug in the cipher-command parser: when supplying a cipher command
8873 string with an "undefined" symbol (neither command nor alphanumeric
8874 [A-Za-z0-9], ssl_set_cipher_list used to hang in an endless loop. Now
8875 an error is flagged.
8877 Due to the strength-sorting extension, the code of the
8878 ssl_create_cipher_list() function was completely rearranged. I hope that
8879 the readability was also increased :-)
8880 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
8882 *) Minor change to 'x509' utility. The -CAcreateserial option now uses 1
8883 for the first serial number and places 2 in the serial number file. This
8884 avoids problems when the root CA is created with serial number zero and
8885 the first user certificate has the same issuer name and serial number
8889 *) Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
8890 the new code. Add documentation for this stuff.
8893 *) Changes to X509_ATTRIBUTE utilities. These have been renamed from
8894 X509_*() to X509at_*() on the grounds that they don't handle X509
8895 structures and behave in an analagous way to the X509v3 functions:
8896 they shouldn't be called directly but wrapper functions should be used
8899 So we also now have some wrapper functions that call the X509at functions
8900 when passed certificate requests. (TO DO: similar things can be done with
8901 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
8902 things. Some of these need some d2i or i2d and print functionality
8903 because they handle more complex structures.)
8906 *) Add missing #ifndefs that caused missing symbols when building libssl
8907 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
8908 NO_RSA in ssl/s2*.c.
8909 [Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller]
8911 *) Precautions against using the PRNG uninitialized: RAND_bytes() now
8912 has a return value which indicates the quality of the random data
8913 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
8914 error queue. New function RAND_pseudo_bytes() generates output that is
8915 guaranteed to be unique but not unpredictable. RAND_add is like
8916 RAND_seed, but takes an extra argument for an entropy estimate
8917 (RAND_seed always assumes full entropy).
8920 *) Do more iterations of Rabin-Miller probable prime test (specifically,
8921 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
8922 instead of only 2 for all lengths; see BN_prime_checks_for_size definition
8923 in crypto/bn/bn_prime.c for the complete table). This guarantees a
8924 false-positive rate of at most 2^-80 for random input.
8927 *) Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
8930 *) New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain
8931 in the 0.9.5 release), this returns the chain
8932 from an X509_CTX structure with a dup of the stack and all
8933 the X509 reference counts upped: so the stack will exist
8934 after X509_CTX_cleanup() has been called. Modify pkcs12.c
8937 Also make SSL_SESSION_print() print out the verify return
8941 *) Add manpage for the pkcs12 command. Also change the default
8942 behaviour so MAC iteration counts are used unless the new
8943 -nomaciter option is used. This improves file security and
8944 only older versions of MSIE (4.0 for example) need it.
8947 *) Honor the no-xxx Configure options when creating .DEF files.
8950 *) Add PKCS#10 attributes to field table: challengePassword,
8951 unstructuredName and unstructuredAddress. These are taken from
8952 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
8953 international characters are used.
8955 More changes to X509_ATTRIBUTE code: allow the setting of types
8956 based on strings. Remove the 'loc' parameter when adding
8957 attributes because these will be a SET OF encoding which is sorted
8961 *) Initial changes to the 'req' utility to allow request generation
8962 automation. This will allow an application to just generate a template
8963 file containing all the field values and have req construct the
8966 Initial support for X509_ATTRIBUTE handling. Stacks of these are
8967 used all over the place including certificate requests and PKCS#7
8968 structures. They are currently handled manually where necessary with
8969 some primitive wrappers for PKCS#7. The new functions behave in a
8970 manner analogous to the X509 extension functions: they allow
8971 attributes to be looked up by NID and added.
8973 Later something similar to the X509V3 code would be desirable to
8974 automatically handle the encoding, decoding and printing of the
8975 more complex types. The string types like challengePassword can
8976 be handled by the string table functions.
8978 Also modified the multi byte string table handling. Now there is
8979 a 'global mask' which masks out certain types. The table itself
8980 can use the flag STABLE_NO_MASK to ignore the mask setting: this
8981 is useful when for example there is only one permissible type
8982 (as in countryName) and using the mask might result in no valid
8986 *) Clean up 'Finished' handling, and add functions SSL_get_finished and
8987 SSL_get_peer_finished to allow applications to obtain the latest
8988 Finished messages sent to the peer or expected from the peer,
8989 respectively. (SSL_get_peer_finished is usually the Finished message
8990 actually received from the peer, otherwise the protocol will be aborted.)
8992 As the Finished message are message digests of the complete handshake
8993 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
8994 be used for external authentication procedures when the authentication
8995 provided by SSL/TLS is not desired or is not enough.
8998 *) Enhanced support for Alpha Linux is added. Now ./config checks if
8999 the host supports BWX extension and if Compaq C is present on the
9000 $PATH. Just exploiting of the BWX extension results in 20-30%
9001 performance kick for some algorithms, e.g. DES and RC4 to mention
9002 a couple. Compaq C in turn generates ~20% faster code for MD5 and
9006 *) Add support for MS "fast SGC". This is arguably a violation of the
9007 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
9008 weak crypto and after checking the certificate is SGC a second one
9009 with strong crypto. MS SGC stops the first handshake after receiving
9010 the server certificate message and sends a second client hello. Since
9011 a server will typically do all the time consuming operations before
9012 expecting any further messages from the client (server key exchange
9013 is the most expensive) there is little difference between the two.
9015 To get OpenSSL to support MS SGC we have to permit a second client
9016 hello message after we have sent server done. In addition we have to
9017 reset the MAC if we do get this second client hello.
9020 *) Add a function 'd2i_AutoPrivateKey()' this will automatically decide
9021 if a DER encoded private key is RSA or DSA traditional format. Changed
9022 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
9023 format DER encoded private key. Newer code should use PKCS#8 format which
9024 has the key type encoded in the ASN1 structure. Added DER private key
9025 support to pkcs8 application.
9028 *) SSL 3/TLS 1 servers now don't request certificates when an anonymous
9029 ciphersuites has been selected (as required by the SSL 3/TLS 1
9030 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
9031 is set, we interpret this as a request to violate the specification
9032 (the worst that can happen is a handshake failure, and 'correct'
9033 behaviour would result in a handshake failure anyway).
9036 *) In SSL_CTX_add_session, take into account that there might be multiple
9037 SSL_SESSION structures with the same session ID (e.g. when two threads
9038 concurrently obtain them from an external cache).
9039 The internal cache can handle only one SSL_SESSION with a given ID,
9040 so if there's a conflict, we now throw out the old one to achieve
9044 *) Add OIDs for idea and blowfish in CBC mode. This will allow both
9045 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
9046 some routines that use cipher OIDs: some ciphers do not have OIDs
9047 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
9051 *) Simplify the trust setting structure and code. Now we just have
9052 two sequences of OIDs for trusted and rejected settings. These will
9053 typically have values the same as the extended key usage extension
9054 and any application specific purposes.
9056 The trust checking code now has a default behaviour: it will just
9057 check for an object with the same NID as the passed id. Functions can
9058 be provided to override either the default behaviour or the behaviour
9059 for a given id. SSL client, server and email already have functions
9060 in place for compatibility: they check the NID and also return "trusted"
9061 if the certificate is self signed.
9064 *) Add d2i,i2d bio/fp functions for PrivateKey: these convert the
9065 traditional format into an EVP_PKEY structure.
9068 *) Add a password callback function PEM_cb() which either prompts for
9069 a password if usr_data is NULL or otherwise assumes it is a null
9070 terminated password. Allow passwords to be passed on command line
9071 environment or config files in a few more utilities.
9074 *) Add a bunch of DER and PEM functions to handle PKCS#8 format private
9075 keys. Add some short names for PKCS#8 PBE algorithms and allow them
9076 to be specified on the command line for the pkcs8 and pkcs12 utilities.
9077 Update documentation.
9080 *) Support for ASN1 "NULL" type. This could be handled before by using
9081 ASN1_TYPE but there wasn't any function that would try to read a NULL
9082 and produce an error if it couldn't. For compatibility we also have
9083 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
9084 don't allocate anything because they don't need to.
9087 *) Initial support for MacOS is now provided. Examine INSTALL.MacOS
9089 [Andy Polyakov, Roy Woods <roy@centicsystems.ca>]
9091 *) Rebuild of the memory allocation routines used by OpenSSL code and
9092 possibly others as well. The purpose is to make an interface that
9093 provide hooks so anyone can build a separate set of allocation and
9094 deallocation routines to be used by OpenSSL, for example memory
9095 pool implementations, or something else, which was previously hard
9096 since Malloc(), Realloc() and Free() were defined as macros having
9097 the values malloc, realloc and free, respectively (except for Win32
9098 compilations). The same is provided for memory debugging code.
9099 OpenSSL already comes with functionality to find memory leaks, but
9100 this gives people a chance to debug other memory problems.
9102 With these changes, a new set of functions and macros have appeared:
9104 CRYPTO_set_mem_debug_functions() [F]
9105 CRYPTO_get_mem_debug_functions() [F]
9106 CRYPTO_dbg_set_options() [F]
9107 CRYPTO_dbg_get_options() [F]
9108 CRYPTO_malloc_debug_init() [M]
9110 The memory debug functions are NULL by default, unless the library
9111 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
9112 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
9113 gives the standard debugging functions that come with OpenSSL) or
9114 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
9115 provided by the library user) must be used. When the standard
9116 debugging functions are used, CRYPTO_dbg_set_options can be used to
9117 request additional information:
9118 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
9119 the CRYPTO_MDEBUG_xxx macro when compiling the library.
9121 Also, things like CRYPTO_set_mem_functions will always give the
9122 expected result (the new set of functions is used for allocation
9123 and deallocation) at all times, regardless of platform and compiler
9126 To finish it up, some functions that were never use in any other
9127 way than through macros have a new API and new semantic:
9130 CRYPTO_dbg_realloc()
9133 All macros of value have retained their old syntax.
9134 [Richard Levitte and Bodo Moeller]
9136 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
9137 ordering of SMIMECapabilities wasn't in "strength order" and there
9138 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
9142 *) Some ASN1 types with illegal zero length encoding (INTEGER,
9143 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
9144 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson]
9146 *) Merge in my S/MIME library for OpenSSL. This provides a simple
9147 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
9148 functionality to handle multipart/signed properly) and a utility
9149 called 'smime' to call all this stuff. This is based on code I
9150 originally wrote for Celo who have kindly allowed it to be
9151 included in OpenSSL.
9154 *) Add variants des_set_key_checked and des_set_key_unchecked of
9155 des_set_key (aka des_key_sched). Global variable des_check_key
9156 decides which of these is called by des_set_key; this way
9157 des_check_key behaves as it always did, but applications and
9158 the library itself, which was buggy for des_check_key == 1,
9159 have a cleaner way to pick the version they need.
9162 *) New function PKCS12_newpass() which changes the password of a
9166 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and
9167 dynamic mix. In both cases the ids can be used as an index into the
9168 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
9169 functions so they accept a list of the field values and the
9170 application doesn't need to directly manipulate the X509_TRUST
9174 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
9178 *) Modify the way the V3 extension code looks up extensions. This now
9179 works in a similar way to the object code: we have some "standard"
9180 extensions in a static table which is searched with OBJ_bsearch()
9181 and the application can add dynamic ones if needed. The file
9182 crypto/x509v3/ext_dat.h now has the info: this file needs to be
9183 updated whenever a new extension is added to the core code and kept
9184 in ext_nid order. There is a simple program 'tabtest.c' which checks
9185 this. New extensions are not added too often so this file can readily
9186 be maintained manually.
9188 There are two big advantages in doing things this way. The extensions
9189 can be looked up immediately and no longer need to be "added" using
9190 X509V3_add_standard_extensions(): this function now does nothing.
9191 [Side note: I get *lots* of email saying the extension code doesn't
9192 work because people forget to call this function]
9193 Also no dynamic allocation is done unless new extensions are added:
9194 so if we don't add custom extensions there is no need to call
9195 X509V3_EXT_cleanup().
9198 *) Modify enc utility's salting as follows: make salting the default. Add a
9199 magic header, so unsalted files fail gracefully instead of just decrypting
9200 to garbage. This is because not salting is a big security hole, so people
9201 should be discouraged from doing it.
9204 *) Fixes and enhancements to the 'x509' utility. It allowed a message
9205 digest to be passed on the command line but it only used this
9206 parameter when signing a certificate. Modified so all relevant
9207 operations are affected by the digest parameter including the
9208 -fingerprint and -x509toreq options. Also -x509toreq choked if a
9209 DSA key was used because it didn't fix the digest.
9212 *) Initial certificate chain verify code. Currently tests the untrusted
9213 certificates for consistency with the verify purpose (which is set
9214 when the X509_STORE_CTX structure is set up) and checks the pathlength.
9216 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
9217 this is because it will reject chains with invalid extensions whereas
9218 every previous version of OpenSSL and SSLeay made no checks at all.
9220 Trust code: checks the root CA for the relevant trust settings. Trust
9221 settings have an initial value consistent with the verify purpose: e.g.
9222 if the verify purpose is for SSL client use it expects the CA to be
9223 trusted for SSL client use. However the default value can be changed to
9224 permit custom trust settings: one example of this would be to only trust
9225 certificates from a specific "secure" set of CAs.
9227 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
9228 which should be used for version portability: especially since the
9229 verify structure is likely to change more often now.
9231 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
9232 to set them. If not set then assume SSL clients will verify SSL servers
9235 Two new options to the verify program: -untrusted allows a set of
9236 untrusted certificates to be passed in and -purpose which sets the
9237 intended purpose of the certificate. If a purpose is set then the
9238 new chain verify code is used to check extension consistency.
9241 *) Support for the authority information access extension.
9244 *) Modify RSA and DSA PEM read routines to transparently handle
9245 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
9246 public keys in a format compatible with certificate
9247 SubjectPublicKeyInfo structures. Unfortunately there were already
9248 functions called *_PublicKey_* which used various odd formats so
9249 these are retained for compatibility: however the DSA variants were
9250 never in a public release so they have been deleted. Changed dsa/rsa
9251 utilities to handle the new format: note no releases ever handled public
9252 keys so we should be OK.
9254 The primary motivation for this change is to avoid the same fiasco
9255 that dogs private keys: there are several incompatible private key
9256 formats some of which are standard and some OpenSSL specific and
9257 require various evil hacks to allow partial transparent handling and
9258 even then it doesn't work with DER formats. Given the option anything
9259 other than PKCS#8 should be dumped: but the other formats have to
9260 stay in the name of compatibility.
9262 With public keys and the benefit of hindsight one standard format
9263 is used which works with EVP_PKEY, RSA or DSA structures: though
9264 it clearly returns an error if you try to read the wrong kind of key.
9266 Added a -pubkey option to the 'x509' utility to output the public key.
9267 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*()
9268 (renamed to EVP_PKEY_get1_*() in the OpenSSL 0.9.5 release) and add
9269 EVP_PKEY_rset_*() functions (renamed to EVP_PKEY_set1_*())
9270 that do the same as the EVP_PKEY_assign_*() except they up the
9271 reference count of the added key (they don't "swallow" the
9275 *) Fixes to crypto/x509/by_file.c the code to read in certificates and
9276 CRLs would fail if the file contained no certificates or no CRLs:
9277 added a new function to read in both types and return the number
9278 read: this means that if none are read it will be an error. The
9279 DER versions of the certificate and CRL reader would always fail
9280 because it isn't possible to mix certificates and CRLs in DER format
9281 without choking one or the other routine. Changed this to just read
9282 a certificate: this is the best we can do. Also modified the code
9283 in apps/verify.c to take notice of return codes: it was previously
9284 attempting to read in certificates from NULL pointers and ignoring
9285 any errors: this is one reason why the cert and CRL reader seemed
9286 to work. It doesn't check return codes from the default certificate
9287 routines: these may well fail if the certificates aren't installed.
9290 *) Code to support otherName option in GeneralName.
9293 *) First update to verify code. Change the verify utility
9294 so it warns if it is passed a self signed certificate:
9295 for consistency with the normal behaviour. X509_verify
9296 has been modified to it will now verify a self signed
9297 certificate if *exactly* the same certificate appears
9298 in the store: it was previously impossible to trust a
9299 single self signed certificate. This means that:
9300 openssl verify ss.pem
9301 now gives a warning about a self signed certificate but
9302 openssl verify -CAfile ss.pem ss.pem
9306 *) For servers, store verify_result in SSL_SESSION data structure
9307 (and add it to external session representation).
9308 This is needed when client certificate verifications fails,
9309 but an application-provided verification callback (set by
9310 SSL_CTX_set_cert_verify_callback) allows accepting the session
9311 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
9312 but returns 1): When the session is reused, we have to set
9313 ssl->verify_result to the appropriate error code to avoid
9315 [Bodo Moeller, problem pointed out by Lutz Jaenicke]
9317 *) Fix a bug in the new PKCS#7 code: it didn't consider the
9318 case in PKCS7_dataInit() where the signed PKCS7 structure
9319 didn't contain any existing data because it was being created.
9320 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson]
9322 *) Add a salt to the key derivation routines in enc.c. This
9323 forms the first 8 bytes of the encrypted file. Also add a
9324 -S option to allow a salt to be input on the command line.
9327 *) New function X509_cmp(). Oddly enough there wasn't a function
9328 to compare two certificates. We do this by working out the SHA1
9329 hash and comparing that. X509_cmp() will be needed by the trust
9333 *) SSL_get1_session() is like SSL_get_session(), but increments
9334 the reference count in the SSL_SESSION returned.
9335 [Geoff Thorpe <geoff@eu.c2.net>]
9337 *) Fix for 'req': it was adding a null to request attributes.
9338 Also change the X509_LOOKUP and X509_INFO code to handle
9339 certificate auxiliary information.
9342 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document
9346 *) Add the possibility to add extra information to the memory leak
9347 detecting output, to form tracebacks, showing from where each
9348 allocation was originated: CRYPTO_push_info("constant string") adds
9349 the string plus current file name and line number to a per-thread
9350 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
9351 is like calling CYRPTO_pop_info() until the stack is empty.
9352 Also updated memory leak detection code to be multi-thread-safe.
9355 *) Add options -text and -noout to pkcs7 utility and delete the
9356 encryption options which never did anything. Update docs.
9359 *) Add options to some of the utilities to allow the pass phrase
9360 to be included on either the command line (not recommended on
9361 OSes like Unix) or read from the environment. Update the
9362 manpages and fix a few bugs.
9365 *) Add a few manpages for some of the openssl commands.
9368 *) Fix the -revoke option in ca. It was freeing up memory twice,
9369 leaking and not finding already revoked certificates.
9372 *) Extensive changes to support certificate auxiliary information.
9373 This involves the use of X509_CERT_AUX structure and X509_AUX
9374 functions. An X509_AUX function such as PEM_read_X509_AUX()
9375 can still read in a certificate file in the usual way but it
9376 will also read in any additional "auxiliary information". By
9377 doing things this way a fair degree of compatibility can be
9378 retained: existing certificates can have this information added
9379 using the new 'x509' options.
9381 Current auxiliary information includes an "alias" and some trust
9382 settings. The trust settings will ultimately be used in enhanced
9383 certificate chain verification routines: currently a certificate
9384 can only be trusted if it is self signed and then it is trusted
9388 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD).
9389 The problem was that one of the replacement routines had not been working
9390 since SSLeay releases. For now the offending routine has been replaced
9391 with non-optimised assembler. Even so, this now gives around 95%
9392 performance improvement for 1024 bit RSA signs.
9395 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2
9396 handling. Most clients have the effective key size in bits equal to
9397 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
9398 A few however don't do this and instead use the size of the decrypted key
9399 to determine the RC2 key length and the AlgorithmIdentifier to determine
9400 the effective key length. In this case the effective key length can still
9401 be 40 bits but the key length can be 168 bits for example. This is fixed
9402 by manually forcing an RC2 key into the EVP_PKEY structure because the
9403 EVP code can't currently handle unusual RC2 key sizes: it always assumes
9404 the key length and effective key length are equal.
9407 *) Add a bunch of functions that should simplify the creation of
9408 X509_NAME structures. Now you should be able to do:
9409 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
9410 and have it automatically work out the correct field type and fill in
9411 the structures. The more adventurous can try:
9412 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
9413 and it will (hopefully) work out the correct multibyte encoding.
9416 *) Change the 'req' utility to use the new field handling and multibyte
9417 copy routines. Before the DN field creation was handled in an ad hoc
9418 way in req, ca, and x509 which was rather broken and didn't support
9419 BMPStrings or UTF8Strings. Since some software doesn't implement
9420 BMPStrings or UTF8Strings yet, they can be enabled using the config file
9421 using the dirstring_type option. See the new comment in the default
9422 openssl.cnf for more info.
9425 *) Make crypto/rand/md_rand.c more robust:
9426 - Assure unique random numbers after fork().
9427 - Make sure that concurrent threads access the global counter and
9428 md serializably so that we never lose entropy in them
9429 or use exactly the same state in multiple threads.
9430 Access to the large state is not always serializable because
9431 the additional locking could be a performance killer, and
9432 md should be large enough anyway.
9435 *) New file apps/app_rand.c with commonly needed functionality
9436 for handling the random seed file.
9438 Use the random seed file in some applications that previously did not:
9440 dsaparam -genkey (which also ignored its '-rand' option),
9443 x509 (when signing).
9444 Except on systems with /dev/urandom, it is crucial to have a random
9445 seed file at least for key creation, DSA signing, and for DH exchanges;
9446 for RSA signatures we could do without one.
9448 gendh and gendsa (unlike genrsa) used to read only the first byte
9449 of each file listed in the '-rand' option. The function as previously
9450 found in genrsa is now in app_rand.c and is used by all programs
9451 that support '-rand'.
9454 *) In RAND_write_file, use mode 0600 for creating files;
9455 don't just chmod when it may be too late.
9458 *) Report an error from X509_STORE_load_locations
9459 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
9462 *) New function ASN1_mbstring_copy() this copies a string in either
9463 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
9464 into an ASN1_STRING type. A mask of permissible types is passed
9465 and it chooses the "minimal" type to use or an error if not type
9469 *) Add function equivalents to the various macros in asn1.h. The old
9470 macros are retained with an M_ prefix. Code inside the library can
9471 use the M_ macros. External code (including the openssl utility)
9472 should *NOT* in order to be "shared library friendly".
9475 *) Add various functions that can check a certificate's extensions
9476 to see if it usable for various purposes such as SSL client,
9477 server or S/MIME and CAs of these types. This is currently
9478 VERY EXPERIMENTAL but will ultimately be used for certificate chain
9479 verification. Also added a -purpose flag to x509 utility to
9480 print out all the purposes.
9483 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated
9487 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search
9488 for, obtain and decode and extension and obtain its critical flag.
9489 This allows all the necessary extension code to be handled in a
9490 single function call.
9493 *) RC4 tune-up featuring 30-40% performance improvement on most RISC
9494 platforms. See crypto/rc4/rc4_enc.c for further details.
9497 *) New -noout option to asn1parse. This causes no output to be produced
9498 its main use is when combined with -strparse and -out to extract data
9499 from a file (which may not be in ASN.1 format).
9502 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer
9503 when producing the local key id.
9504 [Richard Levitte <levitte@stacken.kth.se>]
9506 *) New option -dhparam in s_server. This allows a DH parameter file to be
9507 stated explicitly. If it is not stated then it tries the first server
9508 certificate file. The previous behaviour hard coded the filename
9512 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow
9513 a public key to be input or output. For example:
9514 openssl rsa -in key.pem -pubout -out pubkey.pem
9515 Also added necessary DSA public key functions to handle this.
9518 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
9519 in the message. This was handled by allowing
9520 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
9521 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>]
9523 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
9524 to the end of the strings whereas this didn't. This would cause problems
9525 if strings read with d2i_ASN1_bytes() were later modified.
9526 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>]
9528 *) Fix for base64 decode bug. When a base64 bio reads only one line of
9529 data and it contains EOF it will end up returning an error. This is
9530 caused by input 46 bytes long. The cause is due to the way base64
9531 BIOs find the start of base64 encoded data. They do this by trying a
9532 trial decode on each line until they find one that works. When they
9533 do a flag is set and it starts again knowing it can pass all the
9534 data directly through the decoder. Unfortunately it doesn't reset
9535 the context it uses. This means that if EOF is reached an attempt
9536 is made to pass two EOFs through the context and this causes the
9537 resulting error. This can also cause other problems as well. As is
9538 usual with these problems it takes *ages* to find and the fix is
9539 trivial: move one line.
9540 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ]
9542 *) Ugly workaround to get s_client and s_server working under Windows. The
9543 old code wouldn't work because it needed to select() on sockets and the
9544 tty (for keypresses and to see if data could be written). Win32 only
9545 supports select() on sockets so we select() with a 1s timeout on the
9546 sockets and then see if any characters are waiting to be read, if none
9547 are present then we retry, we also assume we can always write data to
9548 the tty. This isn't nice because the code then blocks until we've
9549 received a complete line of data and it is effectively polling the
9550 keyboard at 1s intervals: however it's quite a bit better than not
9551 working at all :-) A dedicated Windows application might handle this
9552 with an event loop for example.
9555 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
9556 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
9557 will be called when RSA_sign() and RSA_verify() are used. This is useful
9558 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
9559 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
9560 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
9561 This necessitated the support of an extra signature type NID_md5_sha1
9562 for SSL signatures and modifications to the SSL library to use it instead
9563 of calling RSA_public_decrypt() and RSA_private_encrypt().
9566 *) Add new -verify -CAfile and -CApath options to the crl program, these
9567 will lookup a CRL issuers certificate and verify the signature in a
9568 similar way to the verify program. Tidy up the crl program so it
9569 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
9570 less strict. It will now permit CRL extensions even if it is not
9571 a V2 CRL: this will allow it to tolerate some broken CRLs.
9574 *) Initialize all non-automatic variables each time one of the openssl
9575 sub-programs is started (this is necessary as they may be started
9576 multiple times from the "OpenSSL>" prompt).
9577 [Lennart Bang, Bodo Moeller]
9579 *) Preliminary compilation option RSA_NULL which disables RSA crypto without
9580 removing all other RSA functionality (this is what NO_RSA does). This
9581 is so (for example) those in the US can disable those operations covered
9582 by the RSA patent while allowing storage and parsing of RSA keys and RSA
9586 *) Non-copying interface to BIO pairs.
9587 (still largely untested)
9590 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive
9591 ASCII string. This was handled independently in various places before.
9594 *) New functions UTF8_getc() and UTF8_putc() that parse and generate
9595 UTF8 strings a character at a time.
9598 *) Use client_version from client hello to select the protocol
9599 (s23_srvr.c) and for RSA client key exchange verification
9600 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
9603 *) Add various utility functions to handle SPKACs, these were previously
9604 handled by poking round in the structure internals. Added new function
9605 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
9606 print, verify and generate SPKACs. Based on an original idea from
9607 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
9610 *) RIPEMD160 is operational on all platforms and is back in 'make test'.
9613 *) Allow the config file extension section to be overwritten on the
9614 command line. Based on an original idea from Massimiliano Pala
9615 <madwolf@comune.modena.it>. The new option is called -extensions
9616 and can be applied to ca, req and x509. Also -reqexts to override
9617 the request extensions in req and -crlexts to override the crl extensions
9621 *) Add new feature to the SPKAC handling in ca. Now you can include
9622 the same field multiple times by preceding it by "XXXX." for example:
9625 this is the same syntax as used in the req config file.
9628 *) Allow certificate extensions to be added to certificate requests. These
9629 are specified in a 'req_extensions' option of the req section of the
9630 config file. They can be printed out with the -text option to req but
9631 are otherwise ignored at present.
9634 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
9635 data read consists of only the final block it would not decrypted because
9636 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
9637 A misplaced 'break' also meant the decrypted final block might not be
9638 copied until the next read.
9641 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
9642 a few extra parameters to the DH structure: these will be useful if
9643 for example we want the value of 'q' or implement X9.42 DH.
9646 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and
9647 provides hooks that allow the default DSA functions or functions on a
9648 "per key" basis to be replaced. This allows hardware acceleration and
9649 hardware key storage to be handled without major modification to the
9650 library. Also added low level modexp hooks and CRYPTO_EX structure and
9651 associated functions.
9654 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
9655 as "read only": it can't be written to and the buffer it points to will
9656 not be freed. Reading from a read only BIO is much more efficient than
9657 a normal memory BIO. This was added because there are several times when
9658 an area of memory needs to be read from a BIO. The previous method was
9659 to create a memory BIO and write the data to it, this results in two
9660 copies of the data and an O(n^2) reading algorithm. There is a new
9661 function BIO_new_mem_buf() which creates a read only memory BIO from
9662 an area of memory. Also modified the PKCS#7 routines to use read only
9666 *) Bugfix: ssl23_get_client_hello did not work properly when called in
9667 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
9668 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
9669 but a retry condition occured while trying to read the rest.
9672 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as
9673 NID_pkcs7_encrypted by default: this was wrong since this should almost
9674 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
9675 the encrypted data type: this is a more sensible place to put it and it
9676 allows the PKCS#12 code to be tidied up that duplicated this
9680 *) Changed obj_dat.pl script so it takes its input and output files on
9681 the command line. This should avoid shell escape redirection problems
9685 *) Initial support for certificate extension requests, these are included
9686 in things like Xenroll certificate requests. Included functions to allow
9687 extensions to be obtained and added.
9690 *) -crlf option to s_client and s_server for sending newlines as
9691 CRLF (as required by many protocols).
9694 Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
9696 *) Install libRSAglue.a when OpenSSL is built with RSAref.
9697 [Ralf S. Engelschall]
9699 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency.
9700 [Andrija Antonijevic <TheAntony2@bigfoot.com>]
9702 *) Fix -startdate and -enddate (which was missing) arguments to 'ca'
9706 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as
9707 DH parameters/keys (q is lost during that conversion, but the resulting
9708 DH parameters contain its length).
9710 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
9711 much faster than DH_generate_parameters (which creates parameters
9712 where p = 2*q + 1), and also the smaller q makes DH computations
9713 much more efficient (160-bit exponentiation instead of 1024-bit
9714 exponentiation); so this provides a convenient way to support DHE
9715 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
9716 utter importance to use
9717 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
9719 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
9720 when such DH parameters are used, because otherwise small subgroup
9721 attacks may become possible!
9724 *) Avoid memory leak in i2d_DHparams.
9727 *) Allow the -k option to be used more than once in the enc program:
9728 this allows the same encrypted message to be read by multiple recipients.
9731 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
9732 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
9733 it will always use the numerical form of the OID, even if it has a short
9737 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
9738 method only got called if p,q,dmp1,dmq1,iqmp components were present,
9739 otherwise bn_mod_exp was called. In the case of hardware keys for example
9740 no private key components need be present and it might store extra data
9741 in the RSA structure, which cannot be accessed from bn_mod_exp.
9742 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
9743 private key operations.
9746 *) Added support for SPARC Linux.
9749 *) pem_password_cb function type incompatibly changed from
9750 typedef int pem_password_cb(char *buf, int size, int rwflag);
9752 ....(char *buf, int size, int rwflag, void *userdata);
9753 so that applications can pass data to their callbacks:
9754 The PEM[_ASN1]_{read,write}... functions and macros now take an
9755 additional void * argument, which is just handed through whenever
9756 the password callback is called.
9757 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller]
9759 New function SSL_CTX_set_default_passwd_cb_userdata.
9761 Compatibility note: As many C implementations push function arguments
9762 onto the stack in reverse order, the new library version is likely to
9763 interoperate with programs that have been compiled with the old
9764 pem_password_cb definition (PEM_whatever takes some data that
9765 happens to be on the stack as its last argument, and the callback
9766 just ignores this garbage); but there is no guarantee whatsoever that
9769 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
9770 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
9771 problems not only on Windows, but also on some Unix platforms.
9772 To avoid problematic command lines, these definitions are now in an
9773 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
9774 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
9777 *) MIPS III/IV assembler module is reimplemented.
9780 *) More DES library cleanups: remove references to srand/rand and
9781 delete an unused file.
9784 *) Add support for the the free Netwide assembler (NASM) under Win32,
9785 since not many people have MASM (ml) and it can be hard to obtain.
9786 This is currently experimental but it seems to work OK and pass all
9787 the tests. Check out INSTALL.W32 for info.
9790 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
9791 without temporary keys kept an extra copy of the server key,
9792 and connections with temporary keys did not free everything in case
9796 *) New function RSA_check_key and new openssl rsa option -check
9797 for verifying the consistency of RSA keys.
9798 [Ulf Moeller, Bodo Moeller]
9800 *) Various changes to make Win32 compile work:
9801 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
9802 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
9803 comparison" warnings.
9804 3. Add sk_<TYPE>_sort to DEF file generator and do make update.
9807 *) Add a debugging option to PKCS#5 v2 key generation function: when
9808 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
9809 derived keys are printed to stderr.
9812 *) Copy the flags in ASN1_STRING_dup().
9813 [Roman E. Pavlov <pre@mo.msk.ru>]
9815 *) The x509 application mishandled signing requests containing DSA
9816 keys when the signing key was also DSA and the parameters didn't match.
9818 It was supposed to omit the parameters when they matched the signing key:
9819 the verifying software was then supposed to automatically use the CA's
9820 parameters if they were absent from the end user certificate.
9822 Omitting parameters is no longer recommended. The test was also
9823 the wrong way round! This was probably due to unusual behaviour in
9824 EVP_cmp_parameters() which returns 1 if the parameters match.
9825 This meant that parameters were omitted when they *didn't* match and
9826 the certificate was useless. Certificates signed with 'ca' didn't have
9828 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>]
9830 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
9831 The interface is as follows:
9832 Applications can use
9833 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
9834 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
9835 "off" is now the default.
9836 The library internally uses
9837 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
9838 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
9839 to disable memory-checking temporarily.
9841 Some inconsistent states that previously were possible (and were
9842 even the default) are now avoided.
9844 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
9845 with each memory chunk allocated; this is occasionally more helpful
9846 than just having a counter.
9848 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
9850 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
9854 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX),
9855 which largely parallels "options", but is for changing API behaviour,
9856 whereas "options" are about protocol behaviour.
9857 Initial "mode" flags are:
9859 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
9860 a single record has been written.
9861 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
9862 retries use the same buffer location.
9863 (But all of the contents must be
9867 *) Bugfix: SSL_set_options ignored its parameter, only SSL_CTX_set_options
9870 *) Fix problems with no-hmac etc.
9871 [Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>]
9873 *) New functions RSA_get_default_method(), RSA_set_method() and
9874 RSA_get_method(). These allows replacement of RSA_METHODs without having
9875 to mess around with the internals of an RSA structure.
9878 *) Fix memory leaks in DSA_do_sign and DSA_is_prime.
9879 Also really enable memory leak checks in openssl.c and in some
9881 [Chad C. Mulligan, Bodo Moeller]
9883 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
9884 up the length of negative integers. This has now been simplified to just
9885 store the length when it is first determined and use it later, rather
9886 than trying to keep track of where data is copied and updating it to
9888 [Steve Henson, reported by Brien Wheeler
9889 <bwheeler@authentica-security.com>]
9891 *) Add a new function PKCS7_signatureVerify. This allows the verification
9892 of a PKCS#7 signature but with the signing certificate passed to the
9893 function itself. This contrasts with PKCS7_dataVerify which assumes the
9894 certificate is present in the PKCS#7 structure. This isn't always the
9895 case: certificates can be omitted from a PKCS#7 structure and be
9896 distributed by "out of band" means (such as a certificate database).
9899 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the
9900 function prototypes in pem.h, also change util/mkdef.pl to add the
9901 necessary function names.
9904 *) mk1mf.pl (used by Windows builds) did not properly read the
9905 options set by Configure in the top level Makefile, and Configure
9906 was not even able to write more than one option correctly.
9907 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
9910 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config
9911 file to be loaded from a BIO or FILE pointer. The BIO version will
9912 for example allow memory BIOs to contain config info.
9915 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
9916 Whoever hopes to achieve shared-library compatibility across versions
9917 must use this, not the compile-time macro.
9918 (Exercise 0.9.4: Which is the minimum library version required by
9920 Note: All this applies only to multi-threaded programs, others don't
9924 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests
9925 through a BIO pair triggered the default case, i.e.
9926 SSLerr(...,SSL_R_UNKNOWN_STATE).
9929 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
9930 can use the SSL library even if none of the specific BIOs is
9934 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
9935 for the encoded length.
9936 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>]
9938 *) Add initial documentation of the X509V3 functions.
9941 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and
9942 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
9943 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
9944 secure PKCS#8 private key format with a high iteration count.
9947 *) Fix determination of Perl interpreter: A perl or perl5
9948 _directory_ in $PATH was also accepted as the interpreter.
9949 [Ralf S. Engelschall]
9951 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking
9952 wrong with it but it was very old and did things like calling
9953 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
9957 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed
9958 to use the new extension code.
9961 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
9962 with macros. This should make it easier to change their form, add extra
9963 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
9967 *) Add to configuration table a new entry that can specify an alternative
9968 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
9969 according to Mark Crispin <MRC@Panda.COM>.
9973 *) DES CBC did not update the IV. Weird.
9976 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
9977 Changing the behaviour of the former might break existing programs --
9978 where IV updating is needed, des_ncbc_encrypt can be used.
9981 *) When bntest is run from "make test" it drives bc to check its
9982 calculations, as well as internally checking them. If an internal check
9983 fails, it needs to cause bc to give a non-zero result or make test carries
9984 on without noticing the failure. Fixed.
9987 *) DES library cleanups.
9990 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
9991 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
9992 ciphers. NOTE: although the key derivation function has been verified
9993 against some published test vectors it has not been extensively tested
9994 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
9998 *) Instead of "mkdir -p", which is not fully portable, use new
9999 Perl script "util/mkdir-p.pl".
10002 *) Rewrite the way password based encryption (PBE) is handled. It used to
10003 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
10004 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
10005 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
10006 the 'parameter' field of the AlgorithmIdentifier is passed to the
10007 underlying key generation function so it must do its own ASN1 parsing.
10008 This has also changed the EVP_PBE_CipherInit() function which now has a
10009 'parameter' argument instead of literal salt and iteration count values
10010 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
10013 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms
10014 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
10015 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
10016 KEY" because this clashed with PKCS#8 unencrypted string. Since this
10017 value was just used as a "magic string" and not used directly its
10018 value doesn't matter.
10021 *) Introduce some semblance of const correctness to BN. Shame C doesn't
10025 *) "linux-sparc64" configuration (ultrapenguin).
10026 [Ray Miller <ray.miller@oucs.ox.ac.uk>]
10027 "linux-sparc" configuration.
10028 [Christian Forster <fo@hawo.stw.uni-erlangen.de>]
10030 *) config now generates no-xxx options for missing ciphers.
10033 *) Support the EBCDIC character set (work in progress).
10034 File ebcdic.c not yet included because it has a different license.
10035 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
10037 *) Support BS2000/OSD-POSIX.
10038 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
10040 *) Make callbacks for key generation use void * instead of char *.
10043 *) Make S/MIME samples compile (not yet tested).
10046 *) Additional typesafe stacks.
10049 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
10053 Changes between 0.9.3 and 0.9.3a [29 May 1999]
10055 *) New configuration variant "sco5-gcc".
10057 *) Updated some demos.
10058 [Sean O Riordain, Wade Scholine]
10060 *) Add missing BIO_free at exit of pkcs12 application.
10063 *) Fix memory leak in conf.c.
10066 *) Updates for Win32 to assembler version of MD5.
10069 *) Set #! path to perl in apps/der_chop to where we found it
10070 instead of using a fixed path.
10073 *) SHA library changes for irix64-mips4-cc.
10076 *) Improvements for VMS support.
10080 Changes between 0.9.2b and 0.9.3 [24 May 1999]
10082 *) Bignum library bug fix. IRIX 6 passes "make test" now!
10083 This also avoids the problems with SC4.2 and unpatched SC5.
10084 [Andy Polyakov <appro@fy.chalmers.se>]
10086 *) New functions sk_num, sk_value and sk_set to replace the previous macros.
10087 These are required because of the typesafe stack would otherwise break
10088 existing code. If old code used a structure member which used to be STACK
10089 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
10090 sk_num or sk_value it would produce an error because the num, data members
10091 are not present in STACK_OF. Now it just produces a warning. sk_set
10092 replaces the old method of assigning a value to sk_value
10093 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
10094 that does this will no longer work (and should use sk_set instead) but
10095 this could be regarded as a "questionable" behaviour anyway.
10098 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now
10099 correctly handle encrypted S/MIME data.
10102 *) Change type of various DES function arguments from des_cblock
10103 (which means, in function argument declarations, pointer to char)
10104 to des_cblock * (meaning pointer to array with 8 char elements),
10105 which allows the compiler to do more typechecking; it was like
10106 that back in SSLeay, but with lots of ugly casts.
10108 Introduce new type const_des_cblock.
10111 *) Reorganise the PKCS#7 library and get rid of some of the more obvious
10112 problems: find RecipientInfo structure that matches recipient certificate
10113 and initialise the ASN1 structures properly based on passed cipher.
10116 *) Belatedly make the BN tests actually check the results.
10119 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
10120 to and from BNs: it was completely broken. New compilation option
10121 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
10122 key elements as negative integers.
10125 *) Reorganize and speed up MD5.
10126 [Andy Polyakov <appro@fy.chalmers.se>]
10129 [Richard Levitte <richard@levitte.org>]
10131 *) New option -out to asn1parse to allow the parsed structure to be
10132 output to a file. This is most useful when combined with the -strparse
10133 option to examine the output of things like OCTET STRINGS.
10136 *) Make SSL library a little more fool-proof by not requiring any longer
10137 that SSL_set_{accept,connect}_state be called before
10138 SSL_{accept,connect} may be used (SSL_set_..._state is omitted
10139 in many applications because usually everything *appeared* to work as
10140 intended anyway -- now it really works as intended).
10143 *) Move openssl.cnf out of lib/.
10146 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall
10147 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
10148 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+
10149 [Ralf S. Engelschall]
10151 *) Various fixes to the EVP and PKCS#7 code. It may now be able to
10152 handle PKCS#7 enveloped data properly.
10153 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve]
10155 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
10156 copying pointers. The cert_st handling is changed by this in
10157 various ways (and thus what used to be known as ctx->default_cert
10158 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert
10159 any longer when s->cert does not give us what we need).
10160 ssl_cert_instantiate becomes obsolete by this change.
10161 As soon as we've got the new code right (possibly it already is?),
10162 we have solved a couple of bugs of the earlier code where s->cert
10163 was used as if it could not have been shared with other SSL structures.
10165 Note that using the SSL API in certain dirty ways now will result
10166 in different behaviour than observed with earlier library versions:
10167 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx)
10168 does not influence s as it used to.
10170 In order to clean up things more thoroughly, inside SSL_SESSION
10171 we don't use CERT any longer, but a new structure SESS_CERT
10172 that holds per-session data (if available); currently, this is
10173 the peer's certificate chain and, for clients, the server's certificate
10174 and temporary key. CERT holds only those values that can have
10175 meaningful defaults in an SSL_CTX.
10178 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
10179 from the internal representation. Various PKCS#7 fixes: remove some
10180 evil casts and set the enc_dig_alg field properly based on the signing
10184 *) Allow PKCS#12 password to be set from the command line or the
10185 environment. Let 'ca' get its config file name from the environment
10186 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
10190 *) Allow certificate policies extension to use an IA5STRING for the
10191 organization field. This is contrary to the PKIX definition but
10192 VeriSign uses it and IE5 only recognises this form. Document 'x509'
10196 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
10197 without disallowing inline assembler and the like for non-pedantic builds.
10200 *) Support Borland C++ builder.
10201 [Janez Jere <jj@void.si>, modified by Ulf Möller]
10203 *) Support Mingw32.
10206 *) SHA-1 cleanups and performance enhancements.
10207 [Andy Polyakov <appro@fy.chalmers.se>]
10209 *) Sparc v8plus assembler for the bignum library.
10210 [Andy Polyakov <appro@fy.chalmers.se>]
10212 *) Accept any -xxx and +xxx compiler options in Configure.
10215 *) Update HPUX configuration.
10218 *) Add missing sk_<type>_unshift() function to safestack.h
10219 [Ralf S. Engelschall]
10221 *) New function SSL_CTX_use_certificate_chain_file that sets the
10222 "extra_cert"s in addition to the certificate. (This makes sense
10223 only for "PEM" format files, as chains as a whole are not
10227 *) Support verify_depth from the SSL API.
10228 x509_vfy.c had what can be considered an off-by-one-error:
10229 Its depth (which was not part of the external interface)
10230 was actually counting the number of certificates in a chain;
10231 now it really counts the depth.
10234 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
10235 instead of X509err, which often resulted in confusing error
10236 messages since the error codes are not globally unique
10237 (e.g. an alleged error in ssl3_accept when a certificate
10238 didn't match the private key).
10240 *) New function SSL_CTX_set_session_id_context that allows to set a default
10241 value (so that you don't need SSL_set_session_id_context for each
10242 connection using the SSL_CTX).
10245 *) OAEP decoding bug fix.
10248 *) Support INSTALL_PREFIX for package builders, as proposed by
10252 *) New Configure options "threads" and "no-threads". For systems
10253 where the proper compiler options are known (currently Solaris
10254 and Linux), "threads" is the default.
10257 *) New script util/mklink.pl as a faster substitute for util/mklink.sh.
10260 *) Install various scripts to $(OPENSSLDIR)/misc, not to
10261 $(INSTALLTOP)/bin -- they shouldn't clutter directories
10262 such as /usr/local/bin.
10265 *) "make linux-shared" to build shared libraries.
10266 [Niels Poppe <niels@netbox.org>]
10268 *) New Configure option no-<cipher> (rsa, idea, rc5, ...).
10271 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
10272 extension adding in x509 utility.
10275 *) Remove NOPROTO sections and error code comments.
10278 *) Partial rewrite of the DEF file generator to now parse the ANSI
10282 *) New Configure options --prefix=DIR and --openssldir=DIR.
10285 *) Complete rewrite of the error code script(s). It is all now handled
10286 by one script at the top level which handles error code gathering,
10287 header rewriting and C source file generation. It should be much better
10288 than the old method: it now uses a modified version of Ulf's parser to
10289 read the ANSI prototypes in all header files (thus the old K&R definitions
10290 aren't needed for error creation any more) and do a better job of
10291 translating function codes into names. The old 'ASN1 error code imbedded
10292 in a comment' is no longer necessary and it doesn't use .err files which
10293 have now been deleted. Also the error code call doesn't have to appear all
10294 on one line (which resulted in some large lines...).
10297 *) Change #include filenames from <foo.h> to <openssl/foo.h>.
10300 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return
10301 0 (which usually indicates a closed connection), but continue reading.
10304 *) Fix some race conditions.
10307 *) Add support for CRL distribution points extension. Add Certificate
10308 Policies and CRL distribution points documentation.
10311 *) Move the autogenerated header file parts to crypto/opensslconf.h.
10314 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
10315 8 of keying material. Merlin has also confirmed interop with this fix
10316 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
10317 [Merlin Hughes <merlin@baltimore.ie>]
10319 *) Fix lots of warnings.
10320 [Richard Levitte <levitte@stacken.kth.se>]
10322 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
10323 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
10324 [Richard Levitte <levitte@stacken.kth.se>]
10326 *) Fix problems with sizeof(long) == 8.
10327 [Andy Polyakov <appro@fy.chalmers.se>]
10329 *) Change functions to ANSI C.
10332 *) Fix typos in error codes.
10333 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller]
10335 *) Remove defunct assembler files from Configure.
10338 *) SPARC v8 assembler BIGNUM implementation.
10339 [Andy Polyakov <appro@fy.chalmers.se>]
10341 *) Support for Certificate Policies extension: both print and set.
10342 Various additions to support the r2i method this uses.
10345 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could
10346 return a const string when you are expecting an allocated buffer.
10349 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
10350 types DirectoryString and DisplayText.
10353 *) Add code to allow r2i extensions to access the configuration database,
10354 add an LHASH database driver and add several ctx helper functions.
10357 *) Fix an evil bug in bn_expand2() which caused various BN functions to
10358 fail when they extended the size of a BIGNUM.
10361 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to
10362 support typesafe stack.
10365 *) Fix typo in SSL_[gs]et_options().
10366 [Nils Frostberg <nils@medcom.se>]
10368 *) Delete various functions and files that belonged to the (now obsolete)
10369 old X509V3 handling code.
10372 *) New Configure option "rsaref".
10375 *) Don't auto-generate pem.h.
10378 *) Introduce type-safe ASN.1 SETs.
10381 *) Convert various additional casted stacks to type-safe STACK_OF() variants.
10382 [Ben Laurie, Ralf S. Engelschall, Steve Henson]
10384 *) Introduce type-safe STACKs. This will almost certainly break lots of code
10385 that links with OpenSSL (well at least cause lots of warnings), but fear
10386 not: the conversion is trivial, and it eliminates loads of evil casts. A
10387 few STACKed things have been converted already. Feel free to convert more.
10388 In the fullness of time, I'll do away with the STACK type altogether.
10391 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate
10392 specified in <certfile> by updating the entry in the index.txt file.
10393 This way one no longer has to edit the index.txt file manually for
10394 revoking a certificate. The -revoke option does the gory details now.
10395 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall]
10397 *) Fix `openssl crl -noout -text' combination where `-noout' killed the
10398 `-text' option at all and this way the `-noout -text' combination was
10399 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'.
10400 [Ralf S. Engelschall]
10402 *) Make sure a corresponding plain text error message exists for the
10403 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
10404 verify callback function determined that a certificate was revoked.
10405 [Ralf S. Engelschall]
10407 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for
10408 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
10409 all available cipers including rc5, which was forgotten until now.
10410 In order to let the testing shell script know which algorithms
10411 are available, a new (up to now undocumented) command
10412 "openssl list-cipher-commands" is used.
10415 *) Bugfix: s_client occasionally would sleep in select() when
10416 it should have checked SSL_pending() first.
10419 *) New functions DSA_do_sign and DSA_do_verify to provide access to
10420 the raw DSA values prior to ASN.1 encoding.
10423 *) Tweaks to Configure
10424 [Niels Poppe <niels@netbox.org>]
10426 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
10430 *) New variables $(RANLIB) and $(PERL) in the Makefiles.
10433 *) New config option to avoid instructions that are illegal on the 80386.
10434 The default code is faster, but requires at least a 486.
10437 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
10438 SSL2_SERVER_VERSION (not used at all) macros, which are now the
10439 same as SSL2_VERSION anyway.
10442 *) New "-showcerts" option for s_client.
10445 *) Still more PKCS#12 integration. Add pkcs12 application to openssl
10446 application. Various cleanups and fixes.
10449 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
10450 modify error routines to work internally. Add error codes and PBE init
10451 to library startup routines.
10454 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and
10455 packing functions to asn1 and evp. Changed function names and error
10456 codes along the way.
10459 *) PKCS12 integration: and so it begins... First of several patches to
10460 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
10461 objects to objects.h
10464 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1
10465 and display support for Thawte strong extranet extension.
10468 *) Add LinuxPPC support.
10469 [Jeff Dubrule <igor@pobox.org>]
10471 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
10472 bn_div_words in alpha.s.
10473 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie]
10475 *) Make sure the RSA OAEP test is skipped under -DRSAref because
10476 OAEP isn't supported when OpenSSL is built with RSAref.
10477 [Ulf Moeller <ulf@fitug.de>]
10479 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
10480 so they no longer are missing under -DNOPROTO.
10481 [Soren S. Jorvang <soren@t.dk>]
10484 Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
10486 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
10487 doesn't work when the session is reused. Coming soon!
10490 *) Fix a security hole, that allows sessions to be reused in the wrong
10491 context thus bypassing client cert protection! All software that uses
10492 client certs and session caches in multiple contexts NEEDS PATCHING to
10493 allow session reuse! A fuller solution is in the works.
10494 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)]
10496 *) Some more source tree cleanups (removed obsolete files
10497 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
10498 permission on "config" script to be executable) and a fix for the INSTALL
10500 [Ulf Moeller <ulf@fitug.de>]
10502 *) Remove some legacy and erroneous uses of malloc, free instead of
10504 [Lennart Bang <lob@netstream.se>, with minor changes by Steve]
10506 *) Make rsa_oaep_test return non-zero on error.
10507 [Ulf Moeller <ulf@fitug.de>]
10509 *) Add support for native Solaris shared libraries. Configure
10510 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
10511 if someone would make that last step automatic.
10512 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>]
10514 *) ctx_size was not built with the right compiler during "make links". Fixed.
10517 *) Change the meaning of 'ALL' in the cipher list. It now means "everything
10518 except NULL ciphers". This means the default cipher list will no longer
10519 enable NULL ciphers. They need to be specifically enabled e.g. with
10520 the string "DEFAULT:eNULL".
10523 *) Fix to RSA private encryption routines: if p < q then it would
10524 occasionally produce an invalid result. This will only happen with
10525 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
10528 *) Be less restrictive and allow also `perl util/perlpath.pl
10529 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin',
10530 because this way one can also use an interpreter named `perl5' (which is
10531 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
10532 installed as `perl').
10533 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
10535 *) Let util/clean-depend.pl work also with older Perl 5.00x versions.
10536 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
10538 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
10539 advapi32.lib to Win32 build and change the pem test comparision
10540 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
10541 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
10542 and crypto/des/ede_cbcm_enc.c.
10545 *) DES quad checksum was broken on big-endian architectures. Fixed.
10548 *) Comment out two functions in bio.h that aren't implemented. Fix up the
10549 Win32 test batch file so it (might) work again. The Win32 test batch file
10550 is horrible: I feel ill....
10553 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
10554 in e_os.h. Audit of header files to check ANSI and non ANSI
10555 sections: 10 functions were absent from non ANSI section and not exported
10556 from Windows DLLs. Fixed up libeay.num for new functions.
10559 *) Make `openssl version' output lines consistent.
10560 [Ralf S. Engelschall]
10562 *) Fix Win32 symbol export lists for BIO functions: Added
10563 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
10564 to ms/libeay{16,32}.def.
10565 [Ralf S. Engelschall]
10567 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
10568 fine under Unix and passes some trivial tests I've now added. But the
10569 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
10570 added to make sure no one expects that this stuff really works in the
10571 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
10572 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
10574 [Ralf S. Engelschall]
10576 *) Fix the generation of two part addresses in perl.
10577 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie]
10579 *) Add config entry for Linux on MIPS.
10580 [John Tobey <jtobey@channel1.com>]
10582 *) Make links whenever Configure is run, unless we are on Windoze.
10585 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf.
10586 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
10590 *) Add a useful kludge to allow package maintainers to specify compiler and
10591 other platforms details on the command line without having to patch the
10592 Configure script everytime: One now can use ``perl Configure
10593 <id>:<details>'', i.e. platform ids are allowed to have details appended
10594 to them (seperated by colons). This is treated as there would be a static
10595 pre-configured entry in Configure's %table under key <id> with value
10596 <details> and ``perl Configure <id>'' is called. So, when you want to
10597 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
10598 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"''
10599 now, which overrides the FreeBSD-elf entry on-the-fly.
10600 [Ralf S. Engelschall]
10602 *) Disable new TLS1 ciphersuites by default: they aren't official yet.
10605 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
10606 on the `perl Configure ...' command line. This way one can compile
10607 OpenSSL libraries with Position Independent Code (PIC) which is needed
10608 for linking it into DSOs.
10609 [Ralf S. Engelschall]
10611 *) Remarkably, export ciphers were totally broken and no-one had noticed!
10615 *) Cleaned up the LICENSE document: The official contact for any license
10616 questions now is the OpenSSL core team under openssl-core@openssl.org.
10617 And add a paragraph about the dual-license situation to make sure people
10618 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
10619 to the OpenSSL toolkit.
10620 [Ralf S. Engelschall]
10622 *) General source tree makefile cleanups: Made `making xxx in yyy...'
10623 display consistent in the source tree and replaced `/bin/rm' by `rm'.
10624 Additonally cleaned up the `make links' target: Remove unnecessary
10625 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
10626 to speed processing and no longer clutter the display with confusing
10627 stuff. Instead only the actually done links are displayed.
10628 [Ralf S. Engelschall]
10630 *) Permit null encryption ciphersuites, used for authentication only. It used
10631 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
10632 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
10636 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
10637 signed attributes when verifying signatures (this would break them),
10638 the detached data encoding was wrong and public keys obtained using
10639 X509_get_pubkey() weren't freed.
10642 *) Add text documentation for the BUFFER functions. Also added a work around
10643 to a Win95 console bug. This was triggered by the password read stuff: the
10644 last character typed gets carried over to the next fread(). If you were
10645 generating a new cert request using 'req' for example then the last
10646 character of the passphrase would be CR which would then enter the first
10650 *) Added the new `Includes OpenSSL Cryptography Software' button as
10651 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
10652 button and can be used by applications based on OpenSSL to show the
10653 relationship to the OpenSSL project.
10654 [Ralf S. Engelschall]
10656 *) Remove confusing variables in function signatures in files
10657 ssl/ssl_lib.c and ssl/ssl.h.
10658 [Lennart Bong <lob@kulthea.stacken.kth.se>]
10660 *) Don't install bss_file.c under PREFIX/include/
10661 [Lennart Bong <lob@kulthea.stacken.kth.se>]
10663 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle
10664 functions that return function pointers and has support for NT specific
10665 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
10666 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
10667 unsigned to signed types: this was killing the Win32 compile.
10670 *) Add new certificate file to stack functions,
10671 SSL_add_dir_cert_subjects_to_stack() and
10672 SSL_add_file_cert_subjects_to_stack(). These largely supplant
10673 SSL_load_client_CA_file(), and can be used to add multiple certs easily
10674 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
10675 This means that Apache-SSL and similar packages don't have to mess around
10676 to add as many CAs as they want to the preferred list.
10679 *) Experiment with doxygen documentation. Currently only partially applied to
10681 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with
10682 openssl.doxy as the configuration file.
10685 *) Get rid of remaining C++-style comments which strict C compilers hate.
10686 [Ralf S. Engelschall, pointed out by Carlos Amengual]
10688 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
10689 compiled in by default: it has problems with large keys.
10692 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
10693 DH private keys and/or callback functions which directly correspond to
10694 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
10695 is needed for applications which have to configure certificates on a
10696 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
10698 For the RSA certificate situation is makes no difference, but
10699 for the DSA certificate situation this fixes the "no shared cipher"
10700 problem where the OpenSSL cipher selection procedure failed because the
10701 temporary keys were not overtaken from the context and the API provided
10702 no way to reconfigure them.
10703 The new functions now let applications reconfigure the stuff and they
10704 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
10705 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
10706 non-public-API function ssl_cert_instantiate() is used as a helper
10707 function and also to reduce code redundancy inside ssl_rsa.c.
10708 [Ralf S. Engelschall]
10710 *) Move s_server -dcert and -dkey options out of the undocumented feature
10711 area because they are useful for the DSA situation and should be
10712 recognized by the users.
10713 [Ralf S. Engelschall]
10715 *) Fix the cipher decision scheme for export ciphers: the export bits are
10716 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
10717 SSL_EXP_MASK. So, the original variable has to be used instead of the
10718 already masked variable.
10719 [Richard Levitte <levitte@stacken.kth.se>]
10721 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c
10722 [Richard Levitte <levitte@stacken.kth.se>]
10724 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
10725 from `int' to `unsigned int' because it's a length and initialized by
10726 EVP_DigestFinal() which expects an `unsigned int *'.
10727 [Richard Levitte <levitte@stacken.kth.se>]
10729 *) Don't hard-code path to Perl interpreter on shebang line of Configure
10730 script. Instead use the usual Shell->Perl transition trick.
10731 [Ralf S. Engelschall]
10733 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates
10734 (in addition to RSA certificates) to match the behaviour of `openssl dsa
10735 -noout -modulus' as it's already the case for `openssl rsa -noout
10736 -modulus'. For RSA the -modulus is the real "modulus" while for DSA
10737 currently the public key is printed (a decision which was already done by
10738 `openssl dsa -modulus' in the past) which serves a similar purpose.
10739 Additionally the NO_RSA no longer completely removes the whole -modulus
10740 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
10742 [Ralf S. Engelschall]
10744 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
10745 BIO. See the source (crypto/evp/bio_ok.c) for more info.
10746 [Arne Ansper <arne@ats.cyber.ee>]
10748 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs
10749 to be added. Now both 'req' and 'ca' can use new objects defined in the
10753 *) Add cool BIO that does syslog (or event log on NT).
10754 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie]
10756 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
10757 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
10758 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
10759 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
10762 *) Add preliminary config info for new extension code.
10765 *) Make RSA_NO_PADDING really use no padding.
10766 [Ulf Moeller <ulf@fitug.de>]
10768 *) Generate errors when private/public key check is done.
10771 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
10772 for some CRL extensions and new objects added.
10775 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private
10776 key usage extension and fuller support for authority key id.
10779 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
10780 padding method for RSA, which is recommended for new applications in PKCS
10781 #1 v2.0 (RFC 2437, October 1998).
10782 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
10783 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
10784 against Bleichbacher's attack on RSA.
10785 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
10788 *) Updates to the new SSL compression code
10789 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
10791 *) Fix so that the version number in the master secret, when passed
10792 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
10793 (because the server will not accept higher), that the version number
10794 is 0x03,0x01, not 0x03,0x00
10795 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
10797 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
10798 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
10799 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
10802 *) Support for RAW extensions where an arbitrary extension can be
10803 created by including its DER encoding. See apps/openssl.cnf for
10807 *) Make sure latest Perl versions don't interpret some generated C array
10808 code as Perl array code in the crypto/err/err_genc.pl script.
10809 [Lars Weber <3weber@informatik.uni-hamburg.de>]
10811 *) Modify ms/do_ms.bat to not generate assembly language makefiles since
10812 not many people have the assembler. Various Win32 compilation fixes and
10813 update to the INSTALL.W32 file with (hopefully) more accurate Win32
10814 build instructions.
10817 *) Modify configure script 'Configure' to automatically create crypto/date.h
10818 file under Win32 and also build pem.h from pem.org. New script
10819 util/mkfiles.pl to create the MINFO file on environments that can't do a
10820 'make files': perl util/mkfiles.pl >MINFO should work.
10823 *) Major rework of DES function declarations, in the pursuit of correctness
10824 and purity. As a result, many evil casts evaporated, and some weirdness,
10825 too. You may find this causes warnings in your code. Zapping your evil
10826 casts will probably fix them. Mostly.
10829 *) Fix for a typo in asn1.h. Bug fix to object creation script
10830 obj_dat.pl. It considered a zero in an object definition to mean
10831 "end of object": none of the objects in objects.h have any zeros
10832 so it wasn't spotted.
10833 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>]
10835 *) Add support for Triple DES Cipher Block Chaining with Output Feedback
10836 Masking (CBCM). In the absence of test vectors, the best I have been able
10837 to do is check that the decrypt undoes the encrypt, so far. Send me test
10838 vectors if you have them.
10841 *) Correct calculation of key length for export ciphers (too much space was
10842 allocated for null ciphers). This has not been tested!
10845 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage
10846 message is now correct (it understands "crypto" and "ssl" on its
10847 command line). There is also now an "update" option. This will update
10848 the util/ssleay.num and util/libeay.num files with any new functions.
10850 perl util/mkdef.pl crypto ssl update
10851 it will update them.
10854 *) Overhauled the Perl interface (perl/*):
10855 - ported BN stuff to OpenSSL's different BN library
10856 - made the perl/ source tree CVS-aware
10857 - renamed the package from SSLeay to OpenSSL (the files still contain
10858 their history because I've copied them in the repository)
10859 - removed obsolete files (the test scripts will be replaced
10860 by better Test::Harness variants in the future)
10861 [Ralf S. Engelschall]
10863 *) First cut for a very conservative source tree cleanup:
10864 1. merge various obsolete readme texts into doc/ssleay.txt
10865 where we collect the old documents and readme texts.
10866 2. remove the first part of files where I'm already sure that we no
10867 longer need them because of three reasons: either they are just temporary
10868 files which were left by Eric or they are preserved original files where
10869 I've verified that the diff is also available in the CVS via "cvs diff
10870 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
10871 the crypto/md/ stuff).
10872 [Ralf S. Engelschall]
10874 *) More extension code. Incomplete support for subject and issuer alt
10875 name, issuer and authority key id. Change the i2v function parameters
10876 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
10877 what that's for :-) Fix to ASN1 macro which messed up
10878 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
10881 *) Preliminary support for ENUMERATED type. This is largely copied from the
10885 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
10886 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
10888 *) Make sure `make rehash' target really finds the `openssl' program.
10889 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
10891 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
10892 like to hear about it if this slows down other processors.
10895 *) Add CygWin32 platform information to Configure script.
10896 [Alan Batie <batie@aahz.jf.intel.com>]
10898 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm'
10899 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>]
10901 *) New program nseq to manipulate netscape certificate sequences
10904 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
10908 *) Fixes to BN code. Previously the default was to define BN_RECURSION
10909 but the BN code had some problems that would cause failures when
10910 doing certificate verification and some other functions.
10911 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
10913 *) Add ASN1 and PEM code to support netscape certificate sequences.
10916 *) Add ASN1 and PEM code to support netscape certificate sequences.
10919 *) Add several PKIX and private extended key usage OIDs.
10922 *) Modify the 'ca' program to handle the new extension code. Modify
10923 openssl.cnf for new extension format, add comments.
10926 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
10927 and add a sample to openssl.cnf so req -x509 now adds appropriate
10931 *) Continued X509 V3 changes. Add to other makefiles, integrate with the
10932 error code, add initial support to X509_print() and x509 application.
10935 *) Takes a deep breath and start addding X509 V3 extension support code. Add
10936 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
10937 stuff is currently isolated and isn't even compiled yet.
10940 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL
10941 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
10942 Removed the versions check from X509 routines when loading extensions:
10943 this allows certain broken certificates that don't set the version
10944 properly to be processed.
10947 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
10948 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
10949 can still be regenerated with "make depend".
10952 *) Spelling mistake in C version of CAST-128.
10953 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>]
10955 *) Changes to the error generation code. The perl script err-code.pl
10956 now reads in the old error codes and retains the old numbers, only
10957 adding new ones if necessary. It also only changes the .err files if new
10958 codes are added. The makefiles have been modified to only insert errors
10959 when needed (to avoid needlessly modifying header files). This is done
10960 by only inserting errors if the .err file is newer than the auto generated
10961 C file. To rebuild all the error codes from scratch (the old behaviour)
10962 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
10963 or delete all the .err files.
10966 *) CAST-128 was incorrectly implemented for short keys. The C version has
10967 been fixed, but is untested. The assembler versions are also fixed, but
10968 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
10969 to regenerate it if needed.
10970 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
10971 Hagino <itojun@kame.net>]
10973 *) File was opened incorrectly in randfile.c.
10974 [Ulf Möller <ulf@fitug.de>]
10976 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print
10977 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
10978 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
10979 al: it's just almost always a UTCTime. Note this patch adds new error
10980 codes so do a "make errors" if there are problems.
10983 *) Correct Linux 1 recognition in config.
10984 [Ulf Möller <ulf@fitug.de>]
10986 *) Remove pointless MD5 hash when using DSA keys in ca.
10987 [Anonymous <nobody@replay.com>]
10989 *) Generate an error if given an empty string as a cert directory. Also
10990 generate an error if handed NULL (previously returned 0 to indicate an
10991 error, but didn't set one).
10992 [Ben Laurie, reported by Anonymous <nobody@replay.com>]
10994 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
10997 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
10998 parameters. This was causing a warning which killed off the Win32 compile.
11001 *) Remove C++ style comments from crypto/bn/bn_local.h.
11002 [Neil Costigan <neil.costigan@celocom.com>]
11004 *) The function OBJ_txt2nid was broken. It was supposed to return a nid
11005 based on a text string, looking up short and long names and finally
11006 "dot" format. The "dot" format stuff didn't work. Added new function
11007 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
11008 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
11009 OID is not part of the table.
11012 *) Add prototypes to X509 lookup/verify methods, fixing a bug in
11013 X509_LOOKUP_by_alias().
11016 *) Sort openssl functions by name.
11019 *) Get the gendsa program working (hopefully) and add it to app list. Remove
11020 encryption from sample DSA keys (in case anyone is interested the password
11024 *) Make _all_ *_free functions accept a NULL pointer.
11025 [Frans Heymans <fheymans@isaserver.be>]
11027 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use
11029 [Anonymous <nobody@replay.com>]
11031 *) s_server should send the CAfile as acceptable CAs, not its own cert.
11032 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
11034 *) Don't blow it for numeric -newkey arguments to apps/req.
11035 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
11037 *) Temp key "for export" tests were wrong in s3_srvr.c.
11038 [Anonymous <nobody@replay.com>]
11040 *) Add prototype for temp key callback functions
11041 SSL_CTX_set_tmp_{rsa,dh}_callback().
11044 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
11045 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
11048 *) X509_name_add_entry() freed the wrong thing after an error.
11049 [Arne Ansper <arne@ats.cyber.ee>]
11051 *) rsa_eay.c would attempt to free a NULL context.
11052 [Arne Ansper <arne@ats.cyber.ee>]
11054 *) BIO_s_socket() had a broken should_retry() on Windoze.
11055 [Arne Ansper <arne@ats.cyber.ee>]
11057 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
11058 [Arne Ansper <arne@ats.cyber.ee>]
11060 *) Make sure the already existing X509_STORE->depth variable is initialized
11061 in X509_STORE_new(), but document the fact that this variable is still
11062 unused in the certificate verification process.
11063 [Ralf S. Engelschall]
11065 *) Fix the various library and apps files to free up pkeys obtained from
11066 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
11069 *) Fix reference counting in X509_PUBKEY_get(). This makes
11070 demos/maurice/example2.c work, amongst others, probably.
11071 [Steve Henson and Ben Laurie]
11073 *) First cut of a cleanup for apps/. First the `ssleay' program is now named
11074 `openssl' and second, the shortcut symlinks for the `openssl <command>'
11075 are no longer created. This way we have a single and consistent command
11076 line interface `openssl <command>', similar to `cvs <command>'.
11077 [Ralf S. Engelschall, Paul Sutton and Ben Laurie]
11079 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
11080 BIT STRING wrapper always have zero unused bits.
11083 *) Add CA.pl, perl version of CA.sh, add extended key usage OID.
11086 *) Make the top-level INSTALL documentation easier to understand.
11089 *) Makefiles updated to exit if an error occurs in a sub-directory
11090 make (including if user presses ^C) [Paul Sutton]
11092 *) Make Montgomery context stuff explicit in RSA data structure.
11095 *) Fix build order of pem and err to allow for generated pem.h.
11098 *) Fix renumbering bug in X509_NAME_delete_entry().
11101 *) Enhanced the err-ins.pl script so it makes the error library number
11102 global and can add a library name. This is needed for external ASN1 and
11103 other error libraries.
11106 *) Fixed sk_insert which never worked properly.
11109 *) Fix ASN1 macros so they can handle indefinite length construted
11110 EXPLICIT tags. Some non standard certificates use these: they can now
11114 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
11115 into a single doc/ssleay.txt bundle. This way the information is still
11116 preserved but no longer messes up this directory. Now it's new room for
11117 the new set of documenation files.
11118 [Ralf S. Engelschall]
11120 *) SETs were incorrectly DER encoded. This was a major pain, because they
11121 shared code with SEQUENCEs, which aren't coded the same. This means that
11122 almost everything to do with SETs or SEQUENCEs has either changed name or
11123 number of arguments.
11124 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>]
11126 *) Fix test data to work with the above.
11129 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
11130 was already fixed by Eric for 0.9.1 it seems.
11131 [Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>]
11133 *) Autodetect FreeBSD3.
11136 *) Fix various bugs in Configure. This affects the following platforms:
11140 unixware-2.0-pentium
11144 *) Eliminate generated files from CVS. Reorder tests to regenerate files
11145 before they are needed.
11148 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy).
11152 Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
11154 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
11155 changed SSLeay to OpenSSL in version strings.
11156 [Ralf S. Engelschall]
11158 *) Some fixups to the top-level documents.
11161 *) Fixed the nasty bug where rsaref.h was not found under compile-time
11162 because the symlink to include/ was missing.
11163 [Ralf S. Engelschall]
11165 *) Incorporated the popular no-RSA/DSA-only patches
11166 which allow to compile a RSA-free SSLeay.
11167 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall]
11169 *) Fixed nasty rehash problem under `make -f Makefile.ssl links'
11170 when "ssleay" is still not found.
11171 [Ralf S. Engelschall]
11173 *) Added more platforms to Configure: Cray T3E, HPUX 11,
11174 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>]
11176 *) Updated the README file.
11177 [Ralf S. Engelschall]
11179 *) Added various .cvsignore files in the CVS repository subdirs
11180 to make a "cvs update" really silent.
11181 [Ralf S. Engelschall]
11183 *) Recompiled the error-definition header files and added
11184 missing symbols to the Win32 linker tables.
11185 [Ralf S. Engelschall]
11187 *) Cleaned up the top-level documents;
11188 o new files: CHANGES and LICENSE
11189 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
11190 o merged COPYRIGHT into LICENSE
11191 o removed obsolete TODO file
11192 o renamed MICROSOFT to INSTALL.W32
11193 [Ralf S. Engelschall]
11195 *) Removed dummy files from the 0.9.1b source tree:
11196 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
11197 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
11198 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
11199 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
11200 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
11201 [Ralf S. Engelschall]
11203 *) Added various platform portability fixes.
11206 *) The Genesis of the OpenSSL rpject:
11207 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
11208 Young and Tim J. Hudson created while they were working for C2Net until
11210 [The OpenSSL Project]
11213 Changes between 0.9.0b and 0.9.1b [not released]
11215 *) Updated a few CA certificates under certs/
11218 *) Changed some BIGNUM api stuff.
11221 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
11222 DGUX x86, Linux Alpha, etc.
11225 *) New COMP library [crypto/comp/] for SSL Record Layer Compression:
11226 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
11230 *) Add -strparse option to asn1pars program which parses nested
11232 [Dr Stephen Henson <shenson@bigfoot.com>]
11234 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
11237 *) DSA fix for "ca" program.
11240 *) Added "-genkey" option to "dsaparam" program.
11243 *) Added RIPE MD160 (rmd160) message digest.
11246 *) Added -a (all) option to "ssleay version" command.
11249 *) Added PLATFORM define which is the id given to Configure.
11252 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
11255 *) Extended the ASN.1 parser routines.
11258 *) Extended BIO routines to support REUSEADDR, seek, tell, etc.
11261 *) Added a BN_CTX to the BN library.
11264 *) Fixed the weak key values in DES library
11267 *) Changed API in EVP library for cipher aliases.
11270 *) Added support for RC2/64bit cipher.
11273 *) Converted the lhash library to the crypto/mem.c functions.
11276 *) Added more recognized ASN.1 object ids.
11279 *) Added more RSA padding checks for SSL/TLS.
11282 *) Added BIO proxy/filter functionality.
11285 *) Added extra_certs to SSL_CTX which can be used
11286 send extra CA certificates to the client in the CA cert chain sending
11287 process. It can be configured with SSL_CTX_add_extra_chain_cert().
11290 *) Now Fortezza is denied in the authentication phase because
11291 this is key exchange mechanism is not supported by SSLeay at all.
11294 *) Additional PKCS1 checks.
11297 *) Support the string "TLSv1" for all TLS v1 ciphers.
11300 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
11301 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
11304 *) Fixed a few memory leaks.
11307 *) Fixed various code and comment typos.
11310 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0
11311 bytes sent in the client random.
11312 [Edward Bishop <ebishop@spyglass.com>]
projects / openssl.git / blob