]> git.ipfire.org Git - thirdparty/openssl.git/blob - CHANGES
Add support for distinct certificate chains per key type and per SSL
[thirdparty/openssl.git] / CHANGES
1
2 OpenSSL CHANGES
3 _______________
4
5 Changes between 1.0.1 and 1.1.0 [xx XXX xxxx]
6
7 *) Enhance SSL/TLS certificate chain handling to support different
8 chains for each certificate instead of one chain in the parent SSL_CTX.
9 [Steve Henson]
10
11 *) Support for fixed DH ciphersuite client authentication: where both
12 server and client use DH certificates with common parameters.
13 [Steve Henson]
14
15 *) Support for fixed DH ciphersuites: those requiring DH server
16 certificates.
17 [Steve Henson]
18
19 *) Transparently support X9.42 DH parameters when calling
20 PEM_read_bio_DHparameters. This means existing applications can handle
21 the new parameter format automatically.
22 [Steve Henson]
23
24 *) Initial experimental support for X9.42 DH parameter format: mainly
25 to support use of 'q' parameter for RFC5114 parameters.
26 [Steve Henson]
27
28 *) Add DH parameters from RFC5114 including test data to dhtest.
29 [Steve Henson]
30
31 *) Update fips_test_suite to support multiple command line options. New
32 test to induce all self test errors in sequence and check expected
33 failures.
34 [Steve Henson]
35
36 *) Add FIPS_{rsa,dsa,ecdsa}_{sign,verify} functions which digest and
37 sign or verify all in one operation.
38 [Steve Henson]
39
40 *) Add fips_algvs: a multicall fips utility incorporaing all the algorithm
41 test programs and fips_test_suite. Includes functionality to parse
42 the minimal script output of fipsalgest.pl directly.
43 [Steve Henson]
44
45 *) Add authorisation parameter to FIPS_module_mode_set().
46 [Steve Henson]
47
48 *) Add FIPS selftest for ECDH algorithm using P-224 and B-233 curves.
49 [Steve Henson]
50
51 *) Use separate DRBG fields for internal and external flags. New function
52 FIPS_drbg_health_check() to perform on demand health checking. Add
53 generation tests to fips_test_suite with reduced health check interval to
54 demonstrate periodic health checking. Add "nodh" option to
55 fips_test_suite to skip very slow DH test.
56 [Steve Henson]
57
58 *) New function FIPS_get_cipherbynid() to lookup FIPS supported ciphers
59 based on NID.
60 [Steve Henson]
61
62 *) More extensive health check for DRBG checking many more failure modes.
63 New function FIPS_selftest_drbg_all() to handle every possible DRBG
64 combination: call this in fips_test_suite.
65 [Steve Henson]
66
67 *) Add support for Dual EC DRBG from SP800-90. Update DRBG algorithm test
68 and POST to handle Dual EC cases.
69 [Steve Henson]
70
71 *) Add support for canonical generation of DSA parameter 'g'. See
72 FIPS 186-3 A.2.3.
73
74 *) Add support for HMAC DRBG from SP800-90. Update DRBG algorithm test and
75 POST to handle HMAC cases.
76 [Steve Henson]
77
78 *) Add functions FIPS_module_version() and FIPS_module_version_text()
79 to return numberical and string versions of the FIPS module number.
80 [Steve Henson]
81
82 *) Rename FIPS_mode_set and FIPS_mode to FIPS_module_mode_set and
83 FIPS_module_mode. FIPS_mode and FIPS_mode_set will be implmeneted
84 outside the validated module in the FIPS capable OpenSSL.
85 [Steve Henson]
86
87 *) Minor change to DRBG entropy callback semantics. In some cases
88 there is no mutiple of the block length between min_len and
89 max_len. Allow the callback to return more than max_len bytes
90 of entropy but discard any extra: it is the callback's responsibility
91 to ensure that the extra data discarded does not impact the
92 requested amount of entropy.
93 [Steve Henson]
94
95 *) Add PRNG security strength checks to RSA, DSA and ECDSA using
96 information in FIPS186-3, SP800-57 and SP800-131A.
97 [Steve Henson]
98
99 *) CCM support via EVP. Interface is very similar to GCM case except we
100 must supply all data in one chunk (i.e. no update, final) and the
101 message length must be supplied if AAD is used. Add algorithm test
102 support.
103 [Steve Henson]
104
105 *) Initial version of POST overhaul. Add POST callback to allow the status
106 of POST to be monitored and/or failures induced. Modify fips_test_suite
107 to use callback. Always run all selftests even if one fails.
108 [Steve Henson]
109
110 *) XTS support including algorithm test driver in the fips_gcmtest program.
111 Note: this does increase the maximum key length from 32 to 64 bytes but
112 there should be no binary compatibility issues as existing applications
113 will never use XTS mode.
114 [Steve Henson]
115
116 *) Extensive reorganisation of FIPS PRNG behaviour. Remove all dependencies
117 to OpenSSL RAND code and replace with a tiny FIPS RAND API which also
118 performs algorithm blocking for unapproved PRNG types. Also do not
119 set PRNG type in FIPS_mode_set(): leave this to the application.
120 Add default OpenSSL DRBG handling: sets up FIPS PRNG and seeds with
121 the standard OpenSSL PRNG: set additional data to a date time vector.
122 [Steve Henson]
123
124 *) Rename old X9.31 PRNG functions of the form FIPS_rand* to FIPS_x931*.
125 This shouldn't present any incompatibility problems because applications
126 shouldn't be using these directly and any that are will need to rethink
127 anyway as the X9.31 PRNG is now deprecated by FIPS 140-2
128 [Steve Henson]
129
130 *) Extensive self tests and health checking required by SP800-90 DRBG.
131 Remove strength parameter from FIPS_drbg_instantiate and always
132 instantiate at maximum supported strength.
133 [Steve Henson]
134
135 *) Add ECDH code to fips module and fips_ecdhvs for primitives only testing.
136 [Steve Henson]
137
138 *) New algorithm test program fips_dhvs to handle DH primitives only testing.
139 [Steve Henson]
140
141 *) New function DH_compute_key_padded() to compute a DH key and pad with
142 leading zeroes if needed: this complies with SP800-56A et al.
143 [Steve Henson]
144
145 *) Initial implementation of SP800-90 DRBGs for Hash and CTR. Not used by
146 anything, incomplete, subject to change and largely untested at present.
147 [Steve Henson]
148
149 *) Modify fipscanisteronly build option to only build the necessary object
150 files by filtering FIPS_EX_OBJ through a perl script in crypto/Makefile.
151 [Steve Henson]
152
153 *) Add experimental option FIPSSYMS to give all symbols in
154 fipscanister.o and FIPS or fips prefix. This will avoid
155 conflicts with future versions of OpenSSL. Add perl script
156 util/fipsas.pl to preprocess assembly language source files
157 and rename any affected symbols.
158 [Steve Henson]
159
160 *) Add selftest checks and algorithm block of non-fips algorithms in
161 FIPS mode. Remove DES2 from selftests.
162 [Steve Henson]
163
164 *) Add ECDSA code to fips module. Add tiny fips_ecdsa_check to just
165 return internal method without any ENGINE dependencies. Add new
166 tiny fips sign and verify functions.
167 [Steve Henson]
168
169 *) New build option no-ec2m to disable characteristic 2 code.
170 [Steve Henson]
171
172 *) New build option "fipscanisteronly". This only builds fipscanister.o
173 and (currently) associated fips utilities. Uses the file Makefile.fips
174 instead of Makefile.org as the prototype.
175 [Steve Henson]
176
177 *) Add some FIPS mode restrictions to GCM. Add internal IV generator.
178 Update fips_gcmtest to use IV generator.
179 [Steve Henson]
180
181 *) Initial, experimental EVP support for AES-GCM. AAD can be input by
182 setting output buffer to NULL. The *Final function must be
183 called although it will not retrieve any additional data. The tag
184 can be set or retrieved with a ctrl. The IV length is by default 12
185 bytes (96 bits) but can be set to an alternative value. If the IV
186 length exceeds the maximum IV length (currently 16 bytes) it cannot be
187 set before the key.
188 [Steve Henson]
189
190 *) New flag in ciphers: EVP_CIPH_FLAG_CUSTOM_CIPHER. This means the
191 underlying do_cipher function handles all cipher semantics itself
192 including padding and finalisation. This is useful if (for example)
193 an ENGINE cipher handles block padding itself. The behaviour of
194 do_cipher is subtly changed if this flag is set: the return value
195 is the number of characters written to the output buffer (zero is
196 no longer an error code) or a negative error code. Also if the
197 input buffer is NULL and length 0 finalisation should be performed.
198 [Steve Henson]
199
200 *) If a candidate issuer certificate is already part of the constructed
201 path ignore it: new debug notification X509_V_ERR_PATH_LOOP for this case.
202 [Steve Henson]
203
204 *) Improve forward-security support: add functions
205
206 void SSL_CTX_set_not_resumable_session_callback(SSL_CTX *ctx, int (*cb)(SSL *ssl, int is_forward_secure))
207 void SSL_set_not_resumable_session_callback(SSL *ssl, int (*cb)(SSL *ssl, int is_forward_secure))
208
209 for use by SSL/TLS servers; the callback function will be called whenever a
210 new session is created, and gets to decide whether the session may be
211 cached to make it resumable (return 0) or not (return 1). (As by the
212 SSL/TLS protocol specifications, the session_id sent by the server will be
213 empty to indicate that the session is not resumable; also, the server will
214 not generate RFC 4507 (RFC 5077) session tickets.)
215
216 A simple reasonable callback implementation is to return is_forward_secure.
217 This parameter will be set to 1 or 0 depending on the ciphersuite selected
218 by the SSL/TLS server library, indicating whether it can provide forward
219 security.
220 [Emilia Käsper <emilia.kasper@esat.kuleuven.be> (Google)]
221
222 *) New function OPENSSL_gmtime_diff to find the difference in days
223 and seconds between two tm structures. This will be used to provide
224 additional functionality for ASN1_TIME.
225 [Steve Henson]
226
227 *) Add -trusted_first option which attempts to find certificates in the
228 trusted store even if an untrusted chain is also supplied.
229 [Steve Henson]
230
231 *) Initial experimental support for explicitly trusted non-root CAs.
232 OpenSSL still tries to build a complete chain to a root but if an
233 intermediate CA has a trust setting included that is used. The first
234 setting is used: whether to trust or reject.
235 [Steve Henson]
236
237 *) New -verify_name option in command line utilities to set verification
238 parameters by name.
239 [Steve Henson]
240
241 *) Initial CMAC implementation. WARNING: EXPERIMENTAL, API MAY CHANGE.
242 Add CMAC pkey methods.
243 [Steve Henson]
244
245 *) Experimental regnegotiation in s_server -www mode. If the client
246 browses /reneg connection is renegotiated. If /renegcert it is
247 renegotiated requesting a certificate.
248 [Steve Henson]
249
250 *) Add an "external" session cache for debugging purposes to s_server. This
251 should help trace issues which normally are only apparent in deployed
252 multi-process servers.
253 [Steve Henson]
254
255 *) Extensive audit of libcrypto with DEBUG_UNUSED. Fix many cases where
256 return value is ignored. NB. The functions RAND_add(), RAND_seed(),
257 BIO_set_cipher() and some obscure PEM functions were changed so they
258 can now return an error. The RAND changes required a change to the
259 RAND_METHOD structure.
260 [Steve Henson]
261
262 *) New macro __owur for "OpenSSL Warn Unused Result". This makes use of
263 a gcc attribute to warn if the result of a function is ignored. This
264 is enable if DEBUG_UNUSED is set. Add to several functions in evp.h
265 whose return value is often ignored.
266 [Steve Henson]
267
268 Changes between 1.0.0f and 1.0.1 [xx XXX xxxx]
269
270 *) Add support for TLS/DTLS heartbeats.
271 [Robin Seggelmann <seggelmann@fh-muenster.de>]
272
273 *) Add support for SCTP.
274 [Robin Seggelmann <seggelmann@fh-muenster.de>]
275
276 *) Improved PRNG seeding for VOS.
277 [Paul Green <Paul.Green@stratus.com>]
278
279 *) Extensive assembler packs updates, most notably:
280
281 - x86[_64]: AES-NI, PCLMULQDQ, RDRAND support;
282 - x86[_64]: SSSE3 support (SHA1, vector-permutation AES);
283 - x86_64: bit-sliced AES implementation;
284 - ARM: NEON support, contemporary platforms optimizations;
285 - s390x: z196 support;
286 - *: GHASH and GF(2^m) multiplication implementations;
287
288 [Andy Polyakov]
289
290 *) Make TLS-SRP code conformant with RFC 5054 API cleanup
291 (removal of unnecessary code)
292 [Peter Sylvester <peter.sylvester@edelweb.fr>]
293
294 *) Add TLS key material exporter from RFC 5705.
295 [Eric Rescorla]
296
297 *) Add DTLS-SRTP negotiation from RFC 5764.
298 [Eric Rescorla]
299
300 *) Add Next Protocol Negotiation,
301 http://tools.ietf.org/html/draft-agl-tls-nextprotoneg-00. Can be
302 disabled with a no-npn flag to config or Configure. Code donated
303 by Google.
304 [Adam Langley <agl@google.com> and Ben Laurie]
305
306 *) Add optional 64-bit optimized implementations of elliptic curves NIST-P224,
307 NIST-P256, NIST-P521, with constant-time single point multiplication on
308 typical inputs. Compiler support for the nonstandard type __uint128_t is
309 required to use this (present in gcc 4.4 and later, for 64-bit builds).
310 Code made available under Apache License version 2.0.
311
312 Specify "enable-ec_nistp_64_gcc_128" on the Configure (or config) command
313 line to include this in your build of OpenSSL, and run "make depend" (or
314 "make update"). This enables the following EC_METHODs:
315
316 EC_GFp_nistp224_method()
317 EC_GFp_nistp256_method()
318 EC_GFp_nistp521_method()
319
320 EC_GROUP_new_by_curve_name() will automatically use these (while
321 EC_GROUP_new_curve_GFp() currently prefers the more flexible
322 implementations).
323 [Emilia Käsper, Adam Langley, Bodo Moeller (Google)]
324
325 *) Use type ossl_ssize_t instad of ssize_t which isn't available on
326 all platforms. Move ssize_t definition from e_os.h to the public
327 header file e_os2.h as it now appears in public header file cms.h
328 [Steve Henson]
329
330 *) New -sigopt option to the ca, req and x509 utilities. Additional
331 signature parameters can be passed using this option and in
332 particular PSS.
333 [Steve Henson]
334
335 *) Add RSA PSS signing function. This will generate and set the
336 appropriate AlgorithmIdentifiers for PSS based on those in the
337 corresponding EVP_MD_CTX structure. No application support yet.
338 [Steve Henson]
339
340 *) Support for companion algorithm specific ASN1 signing routines.
341 New function ASN1_item_sign_ctx() signs a pre-initialised
342 EVP_MD_CTX structure and sets AlgorithmIdentifiers based on
343 the appropriate parameters.
344 [Steve Henson]
345
346 *) Add new algorithm specific ASN1 verification initialisation function
347 to EVP_PKEY_ASN1_METHOD: this is not in EVP_PKEY_METHOD since the ASN1
348 handling will be the same no matter what EVP_PKEY_METHOD is used.
349 Add a PSS handler to support verification of PSS signatures: checked
350 against a number of sample certificates.
351 [Steve Henson]
352
353 *) Add signature printing for PSS. Add PSS OIDs.
354 [Steve Henson, Martin Kaiser <lists@kaiser.cx>]
355
356 *) Add algorithm specific signature printing. An individual ASN1 method
357 can now print out signatures instead of the standard hex dump.
358
359 More complex signatures (e.g. PSS) can print out more meaningful
360 information. Include DSA version that prints out the signature
361 parameters r, s.
362 [Steve Henson]
363
364 *) Password based recipient info support for CMS library: implementing
365 RFC3211.
366 [Steve Henson]
367
368 *) Split password based encryption into PBES2 and PBKDF2 functions. This
369 neatly separates the code into cipher and PBE sections and is required
370 for some algorithms that split PBES2 into separate pieces (such as
371 password based CMS).
372 [Steve Henson]
373
374 *) Session-handling fixes:
375 - Fix handling of connections that are resuming with a session ID,
376 but also support Session Tickets.
377 - Fix a bug that suppressed issuing of a new ticket if the client
378 presented a ticket with an expired session.
379 - Try to set the ticket lifetime hint to something reasonable.
380 - Make tickets shorter by excluding irrelevant information.
381 - On the client side, don't ignore renewed tickets.
382 [Adam Langley, Bodo Moeller (Google)]
383
384 *) Fix PSK session representation.
385 [Bodo Moeller]
386
387 *) Add RC4-MD5 and AESNI-SHA1 "stitched" implementations.
388
389 This work was sponsored by Intel.
390 [Andy Polyakov]
391
392 *) Add GCM support to TLS library. Some custom code is needed to split
393 the IV between the fixed (from PRF) and explicit (from TLS record)
394 portions. This adds all GCM ciphersuites supported by RFC5288 and
395 RFC5289. Generalise some AES* cipherstrings to inlclude GCM and
396 add a special AESGCM string for GCM only.
397 [Steve Henson]
398
399 *) Expand range of ctrls for AES GCM. Permit setting invocation
400 field on decrypt and retrieval of invocation field only on encrypt.
401 [Steve Henson]
402
403 *) Add HMAC ECC ciphersuites from RFC5289. Include SHA384 PRF support.
404 As required by RFC5289 these ciphersuites cannot be used if for
405 versions of TLS earlier than 1.2.
406 [Steve Henson]
407
408 *) For FIPS capable OpenSSL interpret a NULL default public key method
409 as unset and return the appopriate default but do *not* set the default.
410 This means we can return the appopriate method in applications that
411 swicth between FIPS and non-FIPS modes.
412 [Steve Henson]
413
414 *) Redirect HMAC and CMAC operations to FIPS module in FIPS mode. If an
415 ENGINE is used then we cannot handle that in the FIPS module so we
416 keep original code iff non-FIPS operations are allowed.
417 [Steve Henson]
418
419 *) Add -attime option to openssl utilities.
420 [Peter Eckersley <pde@eff.org>, Ben Laurie and Steve Henson]
421
422 *) Redirect DSA and DH operations to FIPS module in FIPS mode.
423 [Steve Henson]
424
425 *) Redirect ECDSA and ECDH operations to FIPS module in FIPS mode. Also use
426 FIPS EC methods unconditionally for now.
427 [Steve Henson]
428
429 *) New build option no-ec2m to disable characteristic 2 code.
430 [Steve Henson]
431
432 *) Backport libcrypto audit of return value checking from 1.1.0-dev; not
433 all cases can be covered as some introduce binary incompatibilities.
434 [Steve Henson]
435
436 *) Redirect RSA operations to FIPS module including keygen,
437 encrypt, decrypt, sign and verify. Block use of non FIPS RSA methods.
438 [Steve Henson]
439
440 *) Add similar low level API blocking to ciphers.
441 [Steve Henson]
442
443 *) Low level digest APIs are not approved in FIPS mode: any attempt
444 to use these will cause a fatal error. Applications that *really* want
445 to use them can use the private_* version instead.
446 [Steve Henson]
447
448 *) Redirect cipher operations to FIPS module for FIPS builds.
449 [Steve Henson]
450
451 *) Redirect digest operations to FIPS module for FIPS builds.
452 [Steve Henson]
453
454 *) Update build system to add "fips" flag which will link in fipscanister.o
455 for static and shared library builds embedding a signature if needed.
456 [Steve Henson]
457
458 *) Output TLS supported curves in preference order instead of numerical
459 order. This is currently hardcoded for the highest order curves first.
460 This should be configurable so applications can judge speed vs strength.
461 [Steve Henson]
462
463 *) Add TLS v1.2 server support for client authentication.
464 [Steve Henson]
465
466 *) Add support for FIPS mode in ssl library: disable SSLv3, non-FIPS ciphers
467 and enable MD5.
468 [Steve Henson]
469
470 *) Functions FIPS_mode_set() and FIPS_mode() which call the underlying
471 FIPS modules versions.
472 [Steve Henson]
473
474 *) Add TLS v1.2 client side support for client authentication. Keep cache
475 of handshake records longer as we don't know the hash algorithm to use
476 until after the certificate request message is received.
477 [Steve Henson]
478
479 *) Initial TLS v1.2 client support. Add a default signature algorithms
480 extension including all the algorithms we support. Parse new signature
481 format in client key exchange. Relax some ECC signing restrictions for
482 TLS v1.2 as indicated in RFC5246.
483 [Steve Henson]
484
485 *) Add server support for TLS v1.2 signature algorithms extension. Switch
486 to new signature format when needed using client digest preference.
487 All server ciphersuites should now work correctly in TLS v1.2. No client
488 support yet and no support for client certificates.
489 [Steve Henson]
490
491 *) Initial TLS v1.2 support. Add new SHA256 digest to ssl code, switch
492 to SHA256 for PRF when using TLS v1.2 and later. Add new SHA256 based
493 ciphersuites. At present only RSA key exchange ciphersuites work with
494 TLS v1.2. Add new option for TLS v1.2 replacing the old and obsolete
495 SSL_OP_PKCS1_CHECK flags with SSL_OP_NO_TLSv1_2. New TLSv1.2 methods
496 and version checking.
497 [Steve Henson]
498
499 *) New option OPENSSL_NO_SSL_INTERN. If an application can be compiled
500 with this defined it will not be affected by any changes to ssl internal
501 structures. Add several utility functions to allow openssl application
502 to work with OPENSSL_NO_SSL_INTERN defined.
503 [Steve Henson]
504
505 *) Add SRP support.
506 [Tom Wu <tjw@cs.stanford.edu> and Ben Laurie]
507
508 *) Add functions to copy EVP_PKEY_METHOD and retrieve flags and id.
509 [Steve Henson]
510
511 *) Permit abbreviated handshakes when renegotiating using the function
512 SSL_renegotiate_abbreviated().
513 [Robin Seggelmann <seggelmann@fh-muenster.de>]
514
515 *) Add call to ENGINE_register_all_complete() to
516 ENGINE_load_builtin_engines(), so some implementations get used
517 automatically instead of needing explicit application support.
518 [Steve Henson]
519
520 *) Add support for TLS key exporter as described in RFC5705.
521 [Robin Seggelmann <seggelmann@fh-muenster.de>, Steve Henson]
522
523 *) Initial TLSv1.1 support. Since TLSv1.1 is very similar to TLS v1.0 only
524 a few changes are required:
525
526 Add SSL_OP_NO_TLSv1_1 flag.
527 Add TLSv1_1 methods.
528 Update version checking logic to handle version 1.1.
529 Add explicit IV handling (ported from DTLS code).
530 Add command line options to s_client/s_server.
531 [Steve Henson]
532
533 Changes between 1.0.0f and 1.0.0g [18 Jan 2012]
534
535 *) Fix for DTLS DoS issue introduced by fix for CVE-2011-4109.
536 Thanks to Antonio Martin, Enterprise Secure Access Research and
537 Development, Cisco Systems, Inc. for discovering this bug and
538 preparing a fix. (CVE-2012-0050)
539 [Antonio Martin]
540
541 Changes between 1.0.0e and 1.0.0f [4 Jan 2012]
542
543 *) Nadhem Alfardan and Kenny Paterson have discovered an extension
544 of the Vaudenay padding oracle attack on CBC mode encryption
545 which enables an efficient plaintext recovery attack against
546 the OpenSSL implementation of DTLS. Their attack exploits timing
547 differences arising during decryption processing. A research
548 paper describing this attack can be found at:
549 http://www.isg.rhul.ac.uk/~kp/dtls.pdf
550 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
551 Security Group at Royal Holloway, University of London
552 (www.isg.rhul.ac.uk) for discovering this flaw and to Robin Seggelmann
553 <seggelmann@fh-muenster.de> and Michael Tuexen <tuexen@fh-muenster.de>
554 for preparing the fix. (CVE-2011-4108)
555 [Robin Seggelmann, Michael Tuexen]
556
557 *) Clear bytes used for block padding of SSL 3.0 records.
558 (CVE-2011-4576)
559 [Adam Langley (Google)]
560
561 *) Only allow one SGC handshake restart for SSL/TLS. Thanks to George
562 Kadianakis <desnacked@gmail.com> for discovering this issue and
563 Adam Langley for preparing the fix. (CVE-2011-4619)
564 [Adam Langley (Google)]
565
566 *) Check parameters are not NULL in GOST ENGINE. (CVE-2012-0027)
567 [Andrey Kulikov <amdeich@gmail.com>]
568
569 *) Prevent malformed RFC3779 data triggering an assertion failure.
570 Thanks to Andrew Chi, BBN Technologies, for discovering the flaw
571 and Rob Austein <sra@hactrn.net> for fixing it. (CVE-2011-4577)
572 [Rob Austein <sra@hactrn.net>]
573
574 *) Improved PRNG seeding for VOS.
575 [Paul Green <Paul.Green@stratus.com>]
576
577 *) Fix ssl_ciph.c set-up race.
578 [Adam Langley (Google)]
579
580 *) Fix spurious failures in ecdsatest.c.
581 [Emilia Käsper (Google)]
582
583 *) Fix the BIO_f_buffer() implementation (which was mixing different
584 interpretations of the '..._len' fields).
585 [Adam Langley (Google)]
586
587 *) Fix handling of BN_BLINDING: now BN_BLINDING_invert_ex (rather than
588 BN_BLINDING_invert_ex) calls BN_BLINDING_update, ensuring that concurrent
589 threads won't reuse the same blinding coefficients.
590
591 This also avoids the need to obtain the CRYPTO_LOCK_RSA_BLINDING
592 lock to call BN_BLINDING_invert_ex, and avoids one use of
593 BN_BLINDING_update for each BN_BLINDING structure (previously,
594 the last update always remained unused).
595 [Emilia Käsper (Google)]
596
597 *) In ssl3_clear, preserve s3->init_extra along with s3->rbuf.
598 [Bob Buckholz (Google)]
599
600 Changes between 1.0.0d and 1.0.0e [6 Sep 2011]
601
602 *) Fix bug where CRLs with nextUpdate in the past are sometimes accepted
603 by initialising X509_STORE_CTX properly. (CVE-2011-3207)
604 [Kaspar Brand <ossl@velox.ch>]
605
606 *) Fix SSL memory handling for (EC)DH ciphersuites, in particular
607 for multi-threaded use of ECDH. (CVE-2011-3210)
608 [Adam Langley (Google)]
609
610 *) Fix x509_name_ex_d2i memory leak on bad inputs.
611 [Bodo Moeller]
612
613 *) Remove hard coded ecdsaWithSHA1 signature tests in ssl code and check
614 signature public key algorithm by using OID xref utilities instead.
615 Before this you could only use some ECC ciphersuites with SHA1 only.
616 [Steve Henson]
617
618 *) Add protection against ECDSA timing attacks as mentioned in the paper
619 by Billy Bob Brumley and Nicola Tuveri, see:
620
621 http://eprint.iacr.org/2011/232.pdf
622
623 [Billy Bob Brumley and Nicola Tuveri]
624
625 Changes between 1.0.0c and 1.0.0d [8 Feb 2011]
626
627 *) Fix parsing of OCSP stapling ClientHello extension. CVE-2011-0014
628 [Neel Mehta, Adam Langley, Bodo Moeller (Google)]
629
630 *) Fix bug in string printing code: if *any* escaping is enabled we must
631 escape the escape character (backslash) or the resulting string is
632 ambiguous.
633 [Steve Henson]
634
635 Changes between 1.0.0b and 1.0.0c [2 Dec 2010]
636
637 *) Disable code workaround for ancient and obsolete Netscape browsers
638 and servers: an attacker can use it in a ciphersuite downgrade attack.
639 Thanks to Martin Rex for discovering this bug. CVE-2010-4180
640 [Steve Henson]
641
642 *) Fixed J-PAKE implementation error, originally discovered by
643 Sebastien Martini, further info and confirmation from Stefan
644 Arentz and Feng Hao. Note that this fix is a security fix. CVE-2010-4252
645 [Ben Laurie]
646
647 Changes between 1.0.0a and 1.0.0b [16 Nov 2010]
648
649 *) Fix extension code to avoid race conditions which can result in a buffer
650 overrun vulnerability: resumed sessions must not be modified as they can
651 be shared by multiple threads. CVE-2010-3864
652 [Steve Henson]
653
654 *) Fix WIN32 build system to correctly link an ENGINE directory into
655 a DLL.
656 [Steve Henson]
657
658 Changes between 1.0.0 and 1.0.0a [01 Jun 2010]
659
660 *) Check return value of int_rsa_verify in pkey_rsa_verifyrecover
661 (CVE-2010-1633)
662 [Steve Henson, Peter-Michael Hager <hager@dortmund.net>]
663
664 Changes between 0.9.8n and 1.0.0 [29 Mar 2010]
665
666 *) Add "missing" function EVP_CIPHER_CTX_copy(). This copies a cipher
667 context. The operation can be customised via the ctrl mechanism in
668 case ENGINEs want to include additional functionality.
669 [Steve Henson]
670
671 *) Tolerate yet another broken PKCS#8 key format: private key value negative.
672 [Steve Henson]
673
674 *) Add new -subject_hash_old and -issuer_hash_old options to x509 utility to
675 output hashes compatible with older versions of OpenSSL.
676 [Willy Weisz <weisz@vcpc.univie.ac.at>]
677
678 *) Fix compression algorithm handling: if resuming a session use the
679 compression algorithm of the resumed session instead of determining
680 it from client hello again. Don't allow server to change algorithm.
681 [Steve Henson]
682
683 *) Add load_crls() function to apps tidying load_certs() too. Add option
684 to verify utility to allow additional CRLs to be included.
685 [Steve Henson]
686
687 *) Update OCSP request code to permit adding custom headers to the request:
688 some responders need this.
689 [Steve Henson]
690
691 *) The function EVP_PKEY_sign() returns <=0 on error: check return code
692 correctly.
693 [Julia Lawall <julia@diku.dk>]
694
695 *) Update verify callback code in apps/s_cb.c and apps/verify.c, it
696 needlessly dereferenced structures, used obsolete functions and
697 didn't handle all updated verify codes correctly.
698 [Steve Henson]
699
700 *) Disable MD2 in the default configuration.
701 [Steve Henson]
702
703 *) In BIO_pop() and BIO_push() use the ctrl argument (which was NULL) to
704 indicate the initial BIO being pushed or popped. This makes it possible
705 to determine whether the BIO is the one explicitly called or as a result
706 of the ctrl being passed down the chain. Fix BIO_pop() and SSL BIOs so
707 it handles reference counts correctly and doesn't zero out the I/O bio
708 when it is not being explicitly popped. WARNING: applications which
709 included workarounds for the old buggy behaviour will need to be modified
710 or they could free up already freed BIOs.
711 [Steve Henson]
712
713 *) Extend the uni2asc/asc2uni => OPENSSL_uni2asc/OPENSSL_asc2uni
714 renaming to all platforms (within the 0.9.8 branch, this was
715 done conditionally on Netware platforms to avoid a name clash).
716 [Guenter <lists@gknw.net>]
717
718 *) Add ECDHE and PSK support to DTLS.
719 [Michael Tuexen <tuexen@fh-muenster.de>]
720
721 *) Add CHECKED_STACK_OF macro to safestack.h, otherwise safestack can't
722 be used on C++.
723 [Steve Henson]
724
725 *) Add "missing" function EVP_MD_flags() (without this the only way to
726 retrieve a digest flags is by accessing the structure directly. Update
727 EVP_MD_do_all*() and EVP_CIPHER_do_all*() to include the name a digest
728 or cipher is registered as in the "from" argument. Print out all
729 registered digests in the dgst usage message instead of manually
730 attempting to work them out.
731 [Steve Henson]
732
733 *) If no SSLv2 ciphers are used don't use an SSLv2 compatible client hello:
734 this allows the use of compression and extensions. Change default cipher
735 string to remove SSLv2 ciphersuites. This effectively avoids ancient SSLv2
736 by default unless an application cipher string requests it.
737 [Steve Henson]
738
739 *) Alter match criteria in PKCS12_parse(). It used to try to use local
740 key ids to find matching certificates and keys but some PKCS#12 files
741 don't follow the (somewhat unwritten) rules and this strategy fails.
742 Now just gather all certificates together and the first private key
743 then look for the first certificate that matches the key.
744 [Steve Henson]
745
746 *) Support use of registered digest and cipher names for dgst and cipher
747 commands instead of having to add each one as a special case. So now
748 you can do:
749
750 openssl sha256 foo
751
752 as well as:
753
754 openssl dgst -sha256 foo
755
756 and this works for ENGINE based algorithms too.
757
758 [Steve Henson]
759
760 *) Update Gost ENGINE to support parameter files.
761 [Victor B. Wagner <vitus@cryptocom.ru>]
762
763 *) Support GeneralizedTime in ca utility.
764 [Oliver Martin <oliver@volatilevoid.net>, Steve Henson]
765
766 *) Enhance the hash format used for certificate directory links. The new
767 form uses the canonical encoding (meaning equivalent names will work
768 even if they aren't identical) and uses SHA1 instead of MD5. This form
769 is incompatible with the older format and as a result c_rehash should
770 be used to rebuild symbolic links.
771 [Steve Henson]
772
773 *) Make PKCS#8 the default write format for private keys, replacing the
774 traditional format. This form is standardised, more secure and doesn't
775 include an implicit MD5 dependency.
776 [Steve Henson]
777
778 *) Add a $gcc_devteam_warn option to Configure. The idea is that any code
779 committed to OpenSSL should pass this lot as a minimum.
780 [Steve Henson]
781
782 *) Add session ticket override functionality for use by EAP-FAST.
783 [Jouni Malinen <j@w1.fi>]
784
785 *) Modify HMAC functions to return a value. Since these can be implemented
786 in an ENGINE errors can occur.
787 [Steve Henson]
788
789 *) Type-checked OBJ_bsearch_ex.
790 [Ben Laurie]
791
792 *) Type-checked OBJ_bsearch. Also some constification necessitated
793 by type-checking. Still to come: TXT_DB, bsearch(?),
794 OBJ_bsearch_ex, qsort, CRYPTO_EX_DATA, ASN1_VALUE, ASN1_STRING,
795 CONF_VALUE.
796 [Ben Laurie]
797
798 *) New function OPENSSL_gmtime_adj() to add a specific number of days and
799 seconds to a tm structure directly, instead of going through OS
800 specific date routines. This avoids any issues with OS routines such
801 as the year 2038 bug. New *_adj() functions for ASN1 time structures
802 and X509_time_adj_ex() to cover the extended range. The existing
803 X509_time_adj() is still usable and will no longer have any date issues.
804 [Steve Henson]
805
806 *) Delta CRL support. New use deltas option which will attempt to locate
807 and search any appropriate delta CRLs available.
808
809 This work was sponsored by Google.
810 [Steve Henson]
811
812 *) Support for CRLs partitioned by reason code. Reorganise CRL processing
813 code and add additional score elements. Validate alternate CRL paths
814 as part of the CRL checking and indicate a new error "CRL path validation
815 error" in this case. Applications wanting additional details can use
816 the verify callback and check the new "parent" field. If this is not
817 NULL CRL path validation is taking place. Existing applications wont
818 see this because it requires extended CRL support which is off by
819 default.
820
821 This work was sponsored by Google.
822 [Steve Henson]
823
824 *) Support for freshest CRL extension.
825
826 This work was sponsored by Google.
827 [Steve Henson]
828
829 *) Initial indirect CRL support. Currently only supported in the CRLs
830 passed directly and not via lookup. Process certificate issuer
831 CRL entry extension and lookup CRL entries by bother issuer name
832 and serial number. Check and process CRL issuer entry in IDP extension.
833
834 This work was sponsored by Google.
835 [Steve Henson]
836
837 *) Add support for distinct certificate and CRL paths. The CRL issuer
838 certificate is validated separately in this case. Only enabled if
839 an extended CRL support flag is set: this flag will enable additional
840 CRL functionality in future.
841
842 This work was sponsored by Google.
843 [Steve Henson]
844
845 *) Add support for policy mappings extension.
846
847 This work was sponsored by Google.
848 [Steve Henson]
849
850 *) Fixes to pathlength constraint, self issued certificate handling,
851 policy processing to align with RFC3280 and PKITS tests.
852
853 This work was sponsored by Google.
854 [Steve Henson]
855
856 *) Support for name constraints certificate extension. DN, email, DNS
857 and URI types are currently supported.
858
859 This work was sponsored by Google.
860 [Steve Henson]
861
862 *) To cater for systems that provide a pointer-based thread ID rather
863 than numeric, deprecate the current numeric thread ID mechanism and
864 replace it with a structure and associated callback type. This
865 mechanism allows a numeric "hash" to be extracted from a thread ID in
866 either case, and on platforms where pointers are larger than 'long',
867 mixing is done to help ensure the numeric 'hash' is usable even if it
868 can't be guaranteed unique. The default mechanism is to use "&errno"
869 as a pointer-based thread ID to distinguish between threads.
870
871 Applications that want to provide their own thread IDs should now use
872 CRYPTO_THREADID_set_callback() to register a callback that will call
873 either CRYPTO_THREADID_set_numeric() or CRYPTO_THREADID_set_pointer().
874
875 Note that ERR_remove_state() is now deprecated, because it is tied
876 to the assumption that thread IDs are numeric. ERR_remove_state(0)
877 to free the current thread's error state should be replaced by
878 ERR_remove_thread_state(NULL).
879
880 (This new approach replaces the functions CRYPTO_set_idptr_callback(),
881 CRYPTO_get_idptr_callback(), and CRYPTO_thread_idptr() that existed in
882 OpenSSL 0.9.9-dev between June 2006 and August 2008. Also, if an
883 application was previously providing a numeric thread callback that
884 was inappropriate for distinguishing threads, then uniqueness might
885 have been obtained with &errno that happened immediately in the
886 intermediate development versions of OpenSSL; this is no longer the
887 case, the numeric thread callback will now override the automatic use
888 of &errno.)
889 [Geoff Thorpe, with help from Bodo Moeller]
890
891 *) Initial support for different CRL issuing certificates. This covers a
892 simple case where the self issued certificates in the chain exist and
893 the real CRL issuer is higher in the existing chain.
894
895 This work was sponsored by Google.
896 [Steve Henson]
897
898 *) Removed effectively defunct crypto/store from the build.
899 [Ben Laurie]
900
901 *) Revamp of STACK to provide stronger type-checking. Still to come:
902 TXT_DB, bsearch(?), OBJ_bsearch, qsort, CRYPTO_EX_DATA, ASN1_VALUE,
903 ASN1_STRING, CONF_VALUE.
904 [Ben Laurie]
905
906 *) Add a new SSL_MODE_RELEASE_BUFFERS mode flag to release unused buffer
907 RAM on SSL connections. This option can save about 34k per idle SSL.
908 [Nick Mathewson]
909
910 *) Revamp of LHASH to provide stronger type-checking. Still to come:
911 STACK, TXT_DB, bsearch, qsort.
912 [Ben Laurie]
913
914 *) Initial support for Cryptographic Message Syntax (aka CMS) based
915 on RFC3850, RFC3851 and RFC3852. New cms directory and cms utility,
916 support for data, signedData, compressedData, digestedData and
917 encryptedData, envelopedData types included. Scripts to check against
918 RFC4134 examples draft and interop and consistency checks of many
919 content types and variants.
920 [Steve Henson]
921
922 *) Add options to enc utility to support use of zlib compression BIO.
923 [Steve Henson]
924
925 *) Extend mk1mf to support importing of options and assembly language
926 files from Configure script, currently only included in VC-WIN32.
927 The assembly language rules can now optionally generate the source
928 files from the associated perl scripts.
929 [Steve Henson]
930
931 *) Implement remaining functionality needed to support GOST ciphersuites.
932 Interop testing has been performed using CryptoPro implementations.
933 [Victor B. Wagner <vitus@cryptocom.ru>]
934
935 *) s390x assembler pack.
936 [Andy Polyakov]
937
938 *) ARMv4 assembler pack. ARMv4 refers to v4 and later ISA, not CPU
939 "family."
940 [Andy Polyakov]
941
942 *) Implement Opaque PRF Input TLS extension as specified in
943 draft-rescorla-tls-opaque-prf-input-00.txt. Since this is not an
944 official specification yet and no extension type assignment by
945 IANA exists, this extension (for now) will have to be explicitly
946 enabled when building OpenSSL by providing the extension number
947 to use. For example, specify an option
948
949 -DTLSEXT_TYPE_opaque_prf_input=0x9527
950
951 to the "config" or "Configure" script to enable the extension,
952 assuming extension number 0x9527 (which is a completely arbitrary
953 and unofficial assignment based on the MD5 hash of the Internet
954 Draft). Note that by doing so, you potentially lose
955 interoperability with other TLS implementations since these might
956 be using the same extension number for other purposes.
957
958 SSL_set_tlsext_opaque_prf_input(ssl, src, len) is used to set the
959 opaque PRF input value to use in the handshake. This will create
960 an interal copy of the length-'len' string at 'src', and will
961 return non-zero for success.
962
963 To get more control and flexibility, provide a callback function
964 by using
965
966 SSL_CTX_set_tlsext_opaque_prf_input_callback(ctx, cb)
967 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg(ctx, arg)
968
969 where
970
971 int (*cb)(SSL *, void *peerinput, size_t len, void *arg);
972 void *arg;
973
974 Callback function 'cb' will be called in handshakes, and is
975 expected to use SSL_set_tlsext_opaque_prf_input() as appropriate.
976 Argument 'arg' is for application purposes (the value as given to
977 SSL_CTX_set_tlsext_opaque_prf_input_callback_arg() will directly
978 be provided to the callback function). The callback function
979 has to return non-zero to report success: usually 1 to use opaque
980 PRF input just if possible, or 2 to enforce use of the opaque PRF
981 input. In the latter case, the library will abort the handshake
982 if opaque PRF input is not successfully negotiated.
983
984 Arguments 'peerinput' and 'len' given to the callback function
985 will always be NULL and 0 in the case of a client. A server will
986 see the client's opaque PRF input through these variables if
987 available (NULL and 0 otherwise). Note that if the server
988 provides an opaque PRF input, the length must be the same as the
989 length of the client's opaque PRF input.
990
991 Note that the callback function will only be called when creating
992 a new session (session resumption can resume whatever was
993 previously negotiated), and will not be called in SSL 2.0
994 handshakes; thus, SSL_CTX_set_options(ctx, SSL_OP_NO_SSLv2) or
995 SSL_set_options(ssl, SSL_OP_NO_SSLv2) is especially recommended
996 for applications that need to enforce opaque PRF input.
997
998 [Bodo Moeller]
999
1000 *) Update ssl code to support digests other than SHA1+MD5 for handshake
1001 MAC.
1002
1003 [Victor B. Wagner <vitus@cryptocom.ru>]
1004
1005 *) Add RFC4507 support to OpenSSL. This includes the corrections in
1006 RFC4507bis. The encrypted ticket format is an encrypted encoded
1007 SSL_SESSION structure, that way new session features are automatically
1008 supported.
1009
1010 If a client application caches session in an SSL_SESSION structure
1011 support is transparent because tickets are now stored in the encoded
1012 SSL_SESSION.
1013
1014 The SSL_CTX structure automatically generates keys for ticket
1015 protection in servers so again support should be possible
1016 with no application modification.
1017
1018 If a client or server wishes to disable RFC4507 support then the option
1019 SSL_OP_NO_TICKET can be set.
1020
1021 Add a TLS extension debugging callback to allow the contents of any client
1022 or server extensions to be examined.
1023
1024 This work was sponsored by Google.
1025 [Steve Henson]
1026
1027 *) Final changes to avoid use of pointer pointer casts in OpenSSL.
1028 OpenSSL should now compile cleanly on gcc 4.2
1029 [Peter Hartley <pdh@utter.chaos.org.uk>, Steve Henson]
1030
1031 *) Update SSL library to use new EVP_PKEY MAC API. Include generic MAC
1032 support including streaming MAC support: this is required for GOST
1033 ciphersuite support.
1034 [Victor B. Wagner <vitus@cryptocom.ru>, Steve Henson]
1035
1036 *) Add option -stream to use PKCS#7 streaming in smime utility. New
1037 function i2d_PKCS7_bio_stream() and PEM_write_PKCS7_bio_stream()
1038 to output in BER and PEM format.
1039 [Steve Henson]
1040
1041 *) Experimental support for use of HMAC via EVP_PKEY interface. This
1042 allows HMAC to be handled via the EVP_DigestSign*() interface. The
1043 EVP_PKEY "key" in this case is the HMAC key, potentially allowing
1044 ENGINE support for HMAC keys which are unextractable. New -mac and
1045 -macopt options to dgst utility.
1046 [Steve Henson]
1047
1048 *) New option -sigopt to dgst utility. Update dgst to use
1049 EVP_Digest{Sign,Verify}*. These two changes make it possible to use
1050 alternative signing paramaters such as X9.31 or PSS in the dgst
1051 utility.
1052 [Steve Henson]
1053
1054 *) Change ssl_cipher_apply_rule(), the internal function that does
1055 the work each time a ciphersuite string requests enabling
1056 ("foo+bar"), moving ("+foo+bar"), disabling ("-foo+bar", or
1057 removing ("!foo+bar") a class of ciphersuites: Now it maintains
1058 the order of disabled ciphersuites such that those ciphersuites
1059 that most recently went from enabled to disabled not only stay
1060 in order with respect to each other, but also have higher priority
1061 than other disabled ciphersuites the next time ciphersuites are
1062 enabled again.
1063
1064 This means that you can now say, e.g., "PSK:-PSK:HIGH" to enable
1065 the same ciphersuites as with "HIGH" alone, but in a specific
1066 order where the PSK ciphersuites come first (since they are the
1067 most recently disabled ciphersuites when "HIGH" is parsed).
1068
1069 Also, change ssl_create_cipher_list() (using this new
1070 funcionality) such that between otherwise identical
1071 cihpersuites, ephemeral ECDH is preferred over ephemeral DH in
1072 the default order.
1073 [Bodo Moeller]
1074
1075 *) Change ssl_create_cipher_list() so that it automatically
1076 arranges the ciphersuites in reasonable order before starting
1077 to process the rule string. Thus, the definition for "DEFAULT"
1078 (SSL_DEFAULT_CIPHER_LIST) now is just "ALL:!aNULL:!eNULL", but
1079 remains equivalent to "AES:ALL:!aNULL:!eNULL:+aECDH:+kRSA:+RC4:@STRENGTH".
1080 This makes it much easier to arrive at a reasonable default order
1081 in applications for which anonymous ciphers are OK (meaning
1082 that you can't actually use DEFAULT).
1083 [Bodo Moeller; suggested by Victor Duchovni]
1084
1085 *) Split the SSL/TLS algorithm mask (as used for ciphersuite string
1086 processing) into multiple integers instead of setting
1087 "SSL_MKEY_MASK" bits, "SSL_AUTH_MASK" bits, "SSL_ENC_MASK",
1088 "SSL_MAC_MASK", and "SSL_SSL_MASK" bits all in a single integer.
1089 (These masks as well as the individual bit definitions are hidden
1090 away into the non-exported interface ssl/ssl_locl.h, so this
1091 change to the definition of the SSL_CIPHER structure shouldn't
1092 affect applications.) This give us more bits for each of these
1093 categories, so there is no longer a need to coagulate AES128 and
1094 AES256 into a single algorithm bit, and to coagulate Camellia128
1095 and Camellia256 into a single algorithm bit, which has led to all
1096 kinds of kludges.
1097
1098 Thus, among other things, the kludge introduced in 0.9.7m and
1099 0.9.8e for masking out AES256 independently of AES128 or masking
1100 out Camellia256 independently of AES256 is not needed here in 0.9.9.
1101
1102 With the change, we also introduce new ciphersuite aliases that
1103 so far were missing: "AES128", "AES256", "CAMELLIA128", and
1104 "CAMELLIA256".
1105 [Bodo Moeller]
1106
1107 *) Add support for dsa-with-SHA224 and dsa-with-SHA256.
1108 Use the leftmost N bytes of the signature input if the input is
1109 larger than the prime q (with N being the size in bytes of q).
1110 [Nils Larsch]
1111
1112 *) Very *very* experimental PKCS#7 streaming encoder support. Nothing uses
1113 it yet and it is largely untested.
1114 [Steve Henson]
1115
1116 *) Add support for the ecdsa-with-SHA224/256/384/512 signature types.
1117 [Nils Larsch]
1118
1119 *) Initial incomplete changes to avoid need for function casts in OpenSSL
1120 some compilers (gcc 4.2 and later) reject their use. Safestack is
1121 reimplemented. Update ASN1 to avoid use of legacy functions.
1122 [Steve Henson]
1123
1124 *) Win32/64 targets are linked with Winsock2.
1125 [Andy Polyakov]
1126
1127 *) Add an X509_CRL_METHOD structure to allow CRL processing to be redirected
1128 to external functions. This can be used to increase CRL handling
1129 efficiency especially when CRLs are very large by (for example) storing
1130 the CRL revoked certificates in a database.
1131 [Steve Henson]
1132
1133 *) Overhaul of by_dir code. Add support for dynamic loading of CRLs so
1134 new CRLs added to a directory can be used. New command line option
1135 -verify_return_error to s_client and s_server. This causes real errors
1136 to be returned by the verify callback instead of carrying on no matter
1137 what. This reflects the way a "real world" verify callback would behave.
1138 [Steve Henson]
1139
1140 *) GOST engine, supporting several GOST algorithms and public key formats.
1141 Kindly donated by Cryptocom.
1142 [Cryptocom]
1143
1144 *) Partial support for Issuing Distribution Point CRL extension. CRLs
1145 partitioned by DP are handled but no indirect CRL or reason partitioning
1146 (yet). Complete overhaul of CRL handling: now the most suitable CRL is
1147 selected via a scoring technique which handles IDP and AKID in CRLs.
1148 [Steve Henson]
1149
1150 *) New X509_STORE_CTX callbacks lookup_crls() and lookup_certs() which
1151 will ultimately be used for all verify operations: this will remove the
1152 X509_STORE dependency on certificate verification and allow alternative
1153 lookup methods. X509_STORE based implementations of these two callbacks.
1154 [Steve Henson]
1155
1156 *) Allow multiple CRLs to exist in an X509_STORE with matching issuer names.
1157 Modify get_crl() to find a valid (unexpired) CRL if possible.
1158 [Steve Henson]
1159
1160 *) New function X509_CRL_match() to check if two CRLs are identical. Normally
1161 this would be called X509_CRL_cmp() but that name is already used by
1162 a function that just compares CRL issuer names. Cache several CRL
1163 extensions in X509_CRL structure and cache CRLDP in X509.
1164 [Steve Henson]
1165
1166 *) Store a "canonical" representation of X509_NAME structure (ASN1 Name)
1167 this maps equivalent X509_NAME structures into a consistent structure.
1168 Name comparison can then be performed rapidly using memcmp().
1169 [Steve Henson]
1170
1171 *) Non-blocking OCSP request processing. Add -timeout option to ocsp
1172 utility.
1173 [Steve Henson]
1174
1175 *) Allow digests to supply their own micalg string for S/MIME type using
1176 the ctrl EVP_MD_CTRL_MICALG.
1177 [Steve Henson]
1178
1179 *) During PKCS7 signing pass the PKCS7 SignerInfo structure to the
1180 EVP_PKEY_METHOD before and after signing via the EVP_PKEY_CTRL_PKCS7_SIGN
1181 ctrl. It can then customise the structure before and/or after signing
1182 if necessary.
1183 [Steve Henson]
1184
1185 *) New function OBJ_add_sigid() to allow application defined signature OIDs
1186 to be added to OpenSSLs internal tables. New function OBJ_sigid_free()
1187 to free up any added signature OIDs.
1188 [Steve Henson]
1189
1190 *) New functions EVP_CIPHER_do_all(), EVP_CIPHER_do_all_sorted(),
1191 EVP_MD_do_all() and EVP_MD_do_all_sorted() to enumerate internal
1192 digest and cipher tables. New options added to openssl utility:
1193 list-message-digest-algorithms and list-cipher-algorithms.
1194 [Steve Henson]
1195
1196 *) Change the array representation of binary polynomials: the list
1197 of degrees of non-zero coefficients is now terminated with -1.
1198 Previously it was terminated with 0, which was also part of the
1199 value; thus, the array representation was not applicable to
1200 polynomials where t^0 has coefficient zero. This change makes
1201 the array representation useful in a more general context.
1202 [Douglas Stebila]
1203
1204 *) Various modifications and fixes to SSL/TLS cipher string
1205 handling. For ECC, the code now distinguishes between fixed ECDH
1206 with RSA certificates on the one hand and with ECDSA certificates
1207 on the other hand, since these are separate ciphersuites. The
1208 unused code for Fortezza ciphersuites has been removed.
1209
1210 For consistency with EDH, ephemeral ECDH is now called "EECDH"
1211 (not "ECDHE"). For consistency with the code for DH
1212 certificates, use of ECDH certificates is now considered ECDH
1213 authentication, not RSA or ECDSA authentication (the latter is
1214 merely the CA's signing algorithm and not actively used in the
1215 protocol).
1216
1217 The temporary ciphersuite alias "ECCdraft" is no longer
1218 available, and ECC ciphersuites are no longer excluded from "ALL"
1219 and "DEFAULT". The following aliases now exist for RFC 4492
1220 ciphersuites, most of these by analogy with the DH case:
1221
1222 kECDHr - ECDH cert, signed with RSA
1223 kECDHe - ECDH cert, signed with ECDSA
1224 kECDH - ECDH cert (signed with either RSA or ECDSA)
1225 kEECDH - ephemeral ECDH
1226 ECDH - ECDH cert or ephemeral ECDH
1227
1228 aECDH - ECDH cert
1229 aECDSA - ECDSA cert
1230 ECDSA - ECDSA cert
1231
1232 AECDH - anonymous ECDH
1233 EECDH - non-anonymous ephemeral ECDH (equivalent to "kEECDH:-AECDH")
1234
1235 [Bodo Moeller]
1236
1237 *) Add additional S/MIME capabilities for AES and GOST ciphers if supported.
1238 Use correct micalg parameters depending on digest(s) in signed message.
1239 [Steve Henson]
1240
1241 *) Add engine support for EVP_PKEY_ASN1_METHOD. Add functions to process
1242 an ENGINE asn1 method. Support ENGINE lookups in the ASN1 code.
1243 [Steve Henson]
1244
1245 *) Initial engine support for EVP_PKEY_METHOD. New functions to permit
1246 an engine to register a method. Add ENGINE lookups for methods and
1247 functional reference processing.
1248 [Steve Henson]
1249
1250 *) New functions EVP_Digest{Sign,Verify)*. These are enchance versions of
1251 EVP_{Sign,Verify}* which allow an application to customise the signature
1252 process.
1253 [Steve Henson]
1254
1255 *) New -resign option to smime utility. This adds one or more signers
1256 to an existing PKCS#7 signedData structure. Also -md option to use an
1257 alternative message digest algorithm for signing.
1258 [Steve Henson]
1259
1260 *) Tidy up PKCS#7 routines and add new functions to make it easier to
1261 create PKCS7 structures containing multiple signers. Update smime
1262 application to support multiple signers.
1263 [Steve Henson]
1264
1265 *) New -macalg option to pkcs12 utility to allow setting of an alternative
1266 digest MAC.
1267 [Steve Henson]
1268
1269 *) Initial support for PKCS#5 v2.0 PRFs other than default SHA1 HMAC.
1270 Reorganize PBE internals to lookup from a static table using NIDs,
1271 add support for HMAC PBE OID translation. Add a EVP_CIPHER ctrl:
1272 EVP_CTRL_PBE_PRF_NID this allows a cipher to specify an alternative
1273 PRF which will be automatically used with PBES2.
1274 [Steve Henson]
1275
1276 *) Replace the algorithm specific calls to generate keys in "req" with the
1277 new API.
1278 [Steve Henson]
1279
1280 *) Update PKCS#7 enveloped data routines to use new API. This is now
1281 supported by any public key method supporting the encrypt operation. A
1282 ctrl is added to allow the public key algorithm to examine or modify
1283 the PKCS#7 RecipientInfo structure if it needs to: for RSA this is
1284 a no op.
1285 [Steve Henson]
1286
1287 *) Add a ctrl to asn1 method to allow a public key algorithm to express
1288 a default digest type to use. In most cases this will be SHA1 but some
1289 algorithms (such as GOST) need to specify an alternative digest. The
1290 return value indicates how strong the prefernce is 1 means optional and
1291 2 is mandatory (that is it is the only supported type). Modify
1292 ASN1_item_sign() to accept a NULL digest argument to indicate it should
1293 use the default md. Update openssl utilities to use the default digest
1294 type for signing if it is not explicitly indicated.
1295 [Steve Henson]
1296
1297 *) Use OID cross reference table in ASN1_sign() and ASN1_verify(). New
1298 EVP_MD flag EVP_MD_FLAG_PKEY_METHOD_SIGNATURE. This uses the relevant
1299 signing method from the key type. This effectively removes the link
1300 between digests and public key types.
1301 [Steve Henson]
1302
1303 *) Add an OID cross reference table and utility functions. Its purpose is to
1304 translate between signature OIDs such as SHA1WithrsaEncryption and SHA1,
1305 rsaEncryption. This will allow some of the algorithm specific hackery
1306 needed to use the correct OID to be removed.
1307 [Steve Henson]
1308
1309 *) Remove algorithm specific dependencies when setting PKCS7_SIGNER_INFO
1310 structures for PKCS7_sign(). They are now set up by the relevant public
1311 key ASN1 method.
1312 [Steve Henson]
1313
1314 *) Add provisional EC pkey method with support for ECDSA and ECDH.
1315 [Steve Henson]
1316
1317 *) Add support for key derivation (agreement) in the API, DH method and
1318 pkeyutl.
1319 [Steve Henson]
1320
1321 *) Add DSA pkey method and DH pkey methods, extend DH ASN1 method to support
1322 public and private key formats. As a side effect these add additional
1323 command line functionality not previously available: DSA signatures can be
1324 generated and verified using pkeyutl and DH key support and generation in
1325 pkey, genpkey.
1326 [Steve Henson]
1327
1328 *) BeOS support.
1329 [Oliver Tappe <zooey@hirschkaefer.de>]
1330
1331 *) New make target "install_html_docs" installs HTML renditions of the
1332 manual pages.
1333 [Oliver Tappe <zooey@hirschkaefer.de>]
1334
1335 *) New utility "genpkey" this is analagous to "genrsa" etc except it can
1336 generate keys for any algorithm. Extend and update EVP_PKEY_METHOD to
1337 support key and parameter generation and add initial key generation
1338 functionality for RSA.
1339 [Steve Henson]
1340
1341 *) Add functions for main EVP_PKEY_method operations. The undocumented
1342 functions EVP_PKEY_{encrypt,decrypt} have been renamed to
1343 EVP_PKEY_{encrypt,decrypt}_old.
1344 [Steve Henson]
1345
1346 *) Initial definitions for EVP_PKEY_METHOD. This will be a high level public
1347 key API, doesn't do much yet.
1348 [Steve Henson]
1349
1350 *) New function EVP_PKEY_asn1_get0_info() to retrieve information about
1351 public key algorithms. New option to openssl utility:
1352 "list-public-key-algorithms" to print out info.
1353 [Steve Henson]
1354
1355 *) Implement the Supported Elliptic Curves Extension for
1356 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
1357 [Douglas Stebila]
1358
1359 *) Don't free up OIDs in OBJ_cleanup() if they are in use by EVP_MD or
1360 EVP_CIPHER structures to avoid later problems in EVP_cleanup().
1361 [Steve Henson]
1362
1363 *) New utilities pkey and pkeyparam. These are similar to algorithm specific
1364 utilities such as rsa, dsa, dsaparam etc except they process any key
1365 type.
1366 [Steve Henson]
1367
1368 *) Transfer public key printing routines to EVP_PKEY_ASN1_METHOD. New
1369 functions EVP_PKEY_print_public(), EVP_PKEY_print_private(),
1370 EVP_PKEY_print_param() to print public key data from an EVP_PKEY
1371 structure.
1372 [Steve Henson]
1373
1374 *) Initial support for pluggable public key ASN1.
1375 De-spaghettify the public key ASN1 handling. Move public and private
1376 key ASN1 handling to a new EVP_PKEY_ASN1_METHOD structure. Relocate
1377 algorithm specific handling to a single module within the relevant
1378 algorithm directory. Add functions to allow (near) opaque processing
1379 of public and private key structures.
1380 [Steve Henson]
1381
1382 *) Implement the Supported Point Formats Extension for
1383 ECC ciphersuites from draft-ietf-tls-ecc-12.txt.
1384 [Douglas Stebila]
1385
1386 *) Add initial support for RFC 4279 PSK TLS ciphersuites. Add members
1387 for the psk identity [hint] and the psk callback functions to the
1388 SSL_SESSION, SSL and SSL_CTX structure.
1389
1390 New ciphersuites:
1391 PSK-RC4-SHA, PSK-3DES-EDE-CBC-SHA, PSK-AES128-CBC-SHA,
1392 PSK-AES256-CBC-SHA
1393
1394 New functions:
1395 SSL_CTX_use_psk_identity_hint
1396 SSL_get_psk_identity_hint
1397 SSL_get_psk_identity
1398 SSL_use_psk_identity_hint
1399
1400 [Mika Kousa and Pasi Eronen of Nokia Corporation]
1401
1402 *) Add RFC 3161 compliant time stamp request creation, response generation
1403 and response verification functionality.
1404 [Zoltán Glózik <zglozik@opentsa.org>, The OpenTSA Project]
1405
1406 *) Add initial support for TLS extensions, specifically for the server_name
1407 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
1408 have new members for a host name. The SSL data structure has an
1409 additional member SSL_CTX *initial_ctx so that new sessions can be
1410 stored in that context to allow for session resumption, even after the
1411 SSL has been switched to a new SSL_CTX in reaction to a client's
1412 server_name extension.
1413
1414 New functions (subject to change):
1415
1416 SSL_get_servername()
1417 SSL_get_servername_type()
1418 SSL_set_SSL_CTX()
1419
1420 New CTRL codes and macros (subject to change):
1421
1422 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
1423 - SSL_CTX_set_tlsext_servername_callback()
1424 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
1425 - SSL_CTX_set_tlsext_servername_arg()
1426 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
1427
1428 openssl s_client has a new '-servername ...' option.
1429
1430 openssl s_server has new options '-servername_host ...', '-cert2 ...',
1431 '-key2 ...', '-servername_fatal' (subject to change). This allows
1432 testing the HostName extension for a specific single host name ('-cert'
1433 and '-key' remain fallbacks for handshakes without HostName
1434 negotiation). If the unrecogninzed_name alert has to be sent, this by
1435 default is a warning; it becomes fatal with the '-servername_fatal'
1436 option.
1437
1438 [Peter Sylvester, Remy Allais, Christophe Renou]
1439
1440 *) Whirlpool hash implementation is added.
1441 [Andy Polyakov]
1442
1443 *) BIGNUM code on 64-bit SPARCv9 targets is switched from bn(64,64) to
1444 bn(64,32). Because of instruction set limitations it doesn't have
1445 any negative impact on performance. This was done mostly in order
1446 to make it possible to share assembler modules, such as bn_mul_mont
1447 implementations, between 32- and 64-bit builds without hassle.
1448 [Andy Polyakov]
1449
1450 *) Move code previously exiled into file crypto/ec/ec2_smpt.c
1451 to ec2_smpl.c, and no longer require the OPENSSL_EC_BIN_PT_COMP
1452 macro.
1453 [Bodo Moeller]
1454
1455 *) New candidate for BIGNUM assembler implementation, bn_mul_mont,
1456 dedicated Montgomery multiplication procedure, is introduced.
1457 BN_MONT_CTX is modified to allow bn_mul_mont to reach for higher
1458 "64-bit" performance on certain 32-bit targets.
1459 [Andy Polyakov]
1460
1461 *) New option SSL_OP_NO_COMP to disable use of compression selectively
1462 in SSL structures. New SSL ctrl to set maximum send fragment size.
1463 Save memory by seeting the I/O buffer sizes dynamically instead of
1464 using the maximum available value.
1465 [Steve Henson]
1466
1467 *) New option -V for 'openssl ciphers'. This prints the ciphersuite code
1468 in addition to the text details.
1469 [Bodo Moeller]
1470
1471 *) Very, very preliminary EXPERIMENTAL support for printing of general
1472 ASN1 structures. This currently produces rather ugly output and doesn't
1473 handle several customised structures at all.
1474 [Steve Henson]
1475
1476 *) Integrated support for PVK file format and some related formats such
1477 as MS PUBLICKEYBLOB and PRIVATEKEYBLOB. Command line switches to support
1478 these in the 'rsa' and 'dsa' utilities.
1479 [Steve Henson]
1480
1481 *) Support for PKCS#1 RSAPublicKey format on rsa utility command line.
1482 [Steve Henson]
1483
1484 *) Remove the ancient ASN1_METHOD code. This was only ever used in one
1485 place for the (very old) "NETSCAPE" format certificates which are now
1486 handled using new ASN1 code equivalents.
1487 [Steve Henson]
1488
1489 *) Let the TLSv1_method() etc. functions return a 'const' SSL_METHOD
1490 pointer and make the SSL_METHOD parameter in SSL_CTX_new,
1491 SSL_CTX_set_ssl_version and SSL_set_ssl_method 'const'.
1492 [Nils Larsch]
1493
1494 *) Modify CRL distribution points extension code to print out previously
1495 unsupported fields. Enhance extension setting code to allow setting of
1496 all fields.
1497 [Steve Henson]
1498
1499 *) Add print and set support for Issuing Distribution Point CRL extension.
1500 [Steve Henson]
1501
1502 *) Change 'Configure' script to enable Camellia by default.
1503 [NTT]
1504
1505 Changes between 0.9.8s and 0.9.8t [18 Jan 2012]
1506
1507 *) Fix for DTLS DoS issue introduced by fix for CVE-2011-4109.
1508 Thanks to Antonio Martin, Enterprise Secure Access Research and
1509 Development, Cisco Systems, Inc. for discovering this bug and
1510 preparing a fix. (CVE-2012-0050)
1511 [Antonio Martin]
1512
1513 Changes between 0.9.8r and 0.9.8s [4 Jan 2012]
1514
1515 *) Nadhem Alfardan and Kenny Paterson have discovered an extension
1516 of the Vaudenay padding oracle attack on CBC mode encryption
1517 which enables an efficient plaintext recovery attack against
1518 the OpenSSL implementation of DTLS. Their attack exploits timing
1519 differences arising during decryption processing. A research
1520 paper describing this attack can be found at:
1521 http://www.isg.rhul.ac.uk/~kp/dtls.pdf
1522 Thanks go to Nadhem Alfardan and Kenny Paterson of the Information
1523 Security Group at Royal Holloway, University of London
1524 (www.isg.rhul.ac.uk) for discovering this flaw and to Robin Seggelmann
1525 <seggelmann@fh-muenster.de> and Michael Tuexen <tuexen@fh-muenster.de>
1526 for preparing the fix. (CVE-2011-4108)
1527 [Robin Seggelmann, Michael Tuexen]
1528
1529 *) Stop policy check failure freeing same buffer twice. (CVE-2011-4109)
1530 [Ben Laurie, Kasper <ekasper@google.com>]
1531
1532 *) Clear bytes used for block padding of SSL 3.0 records.
1533 (CVE-2011-4576)
1534 [Adam Langley (Google)]
1535
1536 *) Only allow one SGC handshake restart for SSL/TLS. Thanks to George
1537 Kadianakis <desnacked@gmail.com> for discovering this issue and
1538 Adam Langley for preparing the fix. (CVE-2011-4619)
1539 [Adam Langley (Google)]
1540
1541 *) Prevent malformed RFC3779 data triggering an assertion failure.
1542 Thanks to Andrew Chi, BBN Technologies, for discovering the flaw
1543 and Rob Austein <sra@hactrn.net> for fixing it. (CVE-2011-4577)
1544 [Rob Austein <sra@hactrn.net>]
1545
1546 *) Fix ssl_ciph.c set-up race.
1547 [Adam Langley (Google)]
1548
1549 *) Fix spurious failures in ecdsatest.c.
1550 [Emilia Käsper (Google)]
1551
1552 *) Fix the BIO_f_buffer() implementation (which was mixing different
1553 interpretations of the '..._len' fields).
1554 [Adam Langley (Google)]
1555
1556 *) Fix handling of BN_BLINDING: now BN_BLINDING_invert_ex (rather than
1557 BN_BLINDING_invert_ex) calls BN_BLINDING_update, ensuring that concurrent
1558 threads won't reuse the same blinding coefficients.
1559
1560 This also avoids the need to obtain the CRYPTO_LOCK_RSA_BLINDING
1561 lock to call BN_BLINDING_invert_ex, and avoids one use of
1562 BN_BLINDING_update for each BN_BLINDING structure (previously,
1563 the last update always remained unused).
1564 [Emilia Käsper (Google)]
1565
1566 *) Fix SSL memory handling for (EC)DH ciphersuites, in particular
1567 for multi-threaded use of ECDH.
1568 [Adam Langley (Google)]
1569
1570 *) Fix x509_name_ex_d2i memory leak on bad inputs.
1571 [Bodo Moeller]
1572
1573 *) Add protection against ECDSA timing attacks as mentioned in the paper
1574 by Billy Bob Brumley and Nicola Tuveri, see:
1575
1576 http://eprint.iacr.org/2011/232.pdf
1577
1578 [Billy Bob Brumley and Nicola Tuveri]
1579
1580 Changes between 0.9.8q and 0.9.8r [8 Feb 2011]
1581
1582 *) Fix parsing of OCSP stapling ClientHello extension. CVE-2011-0014
1583 [Neel Mehta, Adam Langley, Bodo Moeller (Google)]
1584
1585 *) Fix bug in string printing code: if *any* escaping is enabled we must
1586 escape the escape character (backslash) or the resulting string is
1587 ambiguous.
1588 [Steve Henson]
1589
1590 Changes between 0.9.8p and 0.9.8q [2 Dec 2010]
1591
1592 *) Disable code workaround for ancient and obsolete Netscape browsers
1593 and servers: an attacker can use it in a ciphersuite downgrade attack.
1594 Thanks to Martin Rex for discovering this bug. CVE-2010-4180
1595 [Steve Henson]
1596
1597 *) Fixed J-PAKE implementation error, originally discovered by
1598 Sebastien Martini, further info and confirmation from Stefan
1599 Arentz and Feng Hao. Note that this fix is a security fix. CVE-2010-4252
1600 [Ben Laurie]
1601
1602 Changes between 0.9.8o and 0.9.8p [16 Nov 2010]
1603
1604 *) Fix extension code to avoid race conditions which can result in a buffer
1605 overrun vulnerability: resumed sessions must not be modified as they can
1606 be shared by multiple threads. CVE-2010-3864
1607 [Steve Henson]
1608
1609 *) Fix for double free bug in ssl/s3_clnt.c CVE-2010-2939
1610 [Steve Henson]
1611
1612 *) Don't reencode certificate when calculating signature: cache and use
1613 the original encoding instead. This makes signature verification of
1614 some broken encodings work correctly.
1615 [Steve Henson]
1616
1617 *) ec2_GF2m_simple_mul bugfix: compute correct result if the output EC_POINT
1618 is also one of the inputs.
1619 [Emilia Käsper <emilia.kasper@esat.kuleuven.be> (Google)]
1620
1621 *) Don't repeatedly append PBE algorithms to table if they already exist.
1622 Sort table on each new add. This effectively makes the table read only
1623 after all algorithms are added and subsequent calls to PKCS12_pbe_add
1624 etc are non-op.
1625 [Steve Henson]
1626
1627 Changes between 0.9.8n and 0.9.8o [01 Jun 2010]
1628
1629 [NB: OpenSSL 0.9.8o and later 0.9.8 patch levels were released after
1630 OpenSSL 1.0.0.]
1631
1632 *) Correct a typo in the CMS ASN1 module which can result in invalid memory
1633 access or freeing data twice (CVE-2010-0742)
1634 [Steve Henson, Ronald Moesbergen <intercommit@gmail.com>]
1635
1636 *) Add SHA2 algorithms to SSL_library_init(). SHA2 is becoming far more
1637 common in certificates and some applications which only call
1638 SSL_library_init and not OpenSSL_add_all_algorithms() will fail.
1639 [Steve Henson]
1640
1641 *) VMS fixes:
1642 Reduce copying into .apps and .test in makevms.com
1643 Don't try to use blank CA certificate in CA.com
1644 Allow use of C files from original directories in maketests.com
1645 [Steven M. Schweda" <sms@antinode.info>]
1646
1647 Changes between 0.9.8m and 0.9.8n [24 Mar 2010]
1648
1649 *) When rejecting SSL/TLS records due to an incorrect version number, never
1650 update s->server with a new major version number. As of
1651 - OpenSSL 0.9.8m if 'short' is a 16-bit type,
1652 - OpenSSL 0.9.8f if 'short' is longer than 16 bits,
1653 the previous behavior could result in a read attempt at NULL when
1654 receiving specific incorrect SSL/TLS records once record payload
1655 protection is active. (CVE-2010-0740)
1656 [Bodo Moeller, Adam Langley <agl@chromium.org>]
1657
1658 *) Fix for CVE-2010-0433 where some kerberos enabled versions of OpenSSL
1659 could be crashed if the relevant tables were not present (e.g. chrooted).
1660 [Tomas Hoger <thoger@redhat.com>]
1661
1662 Changes between 0.9.8l and 0.9.8m [25 Feb 2010]
1663
1664 *) Always check bn_wexpend() return values for failure. (CVE-2009-3245)
1665 [Martin Olsson, Neel Mehta]
1666
1667 *) Fix X509_STORE locking: Every 'objs' access requires a lock (to
1668 accommodate for stack sorting, always a write lock!).
1669 [Bodo Moeller]
1670
1671 *) On some versions of WIN32 Heap32Next is very slow. This can cause
1672 excessive delays in the RAND_poll(): over a minute. As a workaround
1673 include a time check in the inner Heap32Next loop too.
1674 [Steve Henson]
1675
1676 *) The code that handled flushing of data in SSL/TLS originally used the
1677 BIO_CTRL_INFO ctrl to see if any data was pending first. This caused
1678 the problem outlined in PR#1949. The fix suggested there however can
1679 trigger problems with buggy BIO_CTRL_WPENDING (e.g. some versions
1680 of Apache). So instead simplify the code to flush unconditionally.
1681 This should be fine since flushing with no data to flush is a no op.
1682 [Steve Henson]
1683
1684 *) Handle TLS versions 2.0 and later properly and correctly use the
1685 highest version of TLS/SSL supported. Although TLS >= 2.0 is some way
1686 off ancient servers have a habit of sticking around for a while...
1687 [Steve Henson]
1688
1689 *) Modify compression code so it frees up structures without using the
1690 ex_data callbacks. This works around a problem where some applications
1691 call CRYPTO_cleanup_all_ex_data() before application exit (e.g. when
1692 restarting) then use compression (e.g. SSL with compression) later.
1693 This results in significant per-connection memory leaks and
1694 has caused some security issues including CVE-2008-1678 and
1695 CVE-2009-4355.
1696 [Steve Henson]
1697
1698 *) Constify crypto/cast (i.e., <openssl/cast.h>): a CAST_KEY doesn't
1699 change when encrypting or decrypting.
1700 [Bodo Moeller]
1701
1702 *) Add option SSL_OP_LEGACY_SERVER_CONNECT which will allow clients to
1703 connect and renegotiate with servers which do not support RI.
1704 Until RI is more widely deployed this option is enabled by default.
1705 [Steve Henson]
1706
1707 *) Add "missing" ssl ctrls to clear options and mode.
1708 [Steve Henson]
1709
1710 *) If client attempts to renegotiate and doesn't support RI respond with
1711 a no_renegotiation alert as required by RFC5746. Some renegotiating
1712 TLS clients will continue a connection gracefully when they receive
1713 the alert. Unfortunately OpenSSL mishandled this alert and would hang
1714 waiting for a server hello which it will never receive. Now we treat a
1715 received no_renegotiation alert as a fatal error. This is because
1716 applications requesting a renegotiation might well expect it to succeed
1717 and would have no code in place to handle the server denying it so the
1718 only safe thing to do is to terminate the connection.
1719 [Steve Henson]
1720
1721 *) Add ctrl macro SSL_get_secure_renegotiation_support() which returns 1 if
1722 peer supports secure renegotiation and 0 otherwise. Print out peer
1723 renegotiation support in s_client/s_server.
1724 [Steve Henson]
1725
1726 *) Replace the highly broken and deprecated SPKAC certification method with
1727 the updated NID creation version. This should correctly handle UTF8.
1728 [Steve Henson]
1729
1730 *) Implement RFC5746. Re-enable renegotiation but require the extension
1731 as needed. Unfortunately, SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION
1732 turns out to be a bad idea. It has been replaced by
1733 SSL_OP_ALLOW_UNSAFE_LEGACY_RENEGOTIATION which can be set with
1734 SSL_CTX_set_options(). This is really not recommended unless you
1735 know what you are doing.
1736 [Eric Rescorla <ekr@networkresonance.com>, Ben Laurie, Steve Henson]
1737
1738 *) Fixes to stateless session resumption handling. Use initial_ctx when
1739 issuing and attempting to decrypt tickets in case it has changed during
1740 servername handling. Use a non-zero length session ID when attempting
1741 stateless session resumption: this makes it possible to determine if
1742 a resumption has occurred immediately after receiving server hello
1743 (several places in OpenSSL subtly assume this) instead of later in
1744 the handshake.
1745 [Steve Henson]
1746
1747 *) The functions ENGINE_ctrl(), OPENSSL_isservice(),
1748 CMS_get1_RecipientRequest() and RAND_bytes() can return <=0 on error
1749 fixes for a few places where the return code is not checked
1750 correctly.
1751 [Julia Lawall <julia@diku.dk>]
1752
1753 *) Add --strict-warnings option to Configure script to include devteam
1754 warnings in other configurations.
1755 [Steve Henson]
1756
1757 *) Add support for --libdir option and LIBDIR variable in makefiles. This
1758 makes it possible to install openssl libraries in locations which
1759 have names other than "lib", for example "/usr/lib64" which some
1760 systems need.
1761 [Steve Henson, based on patch from Jeremy Utley]
1762
1763 *) Don't allow the use of leading 0x80 in OIDs. This is a violation of
1764 X690 8.9.12 and can produce some misleading textual output of OIDs.
1765 [Steve Henson, reported by Dan Kaminsky]
1766
1767 *) Delete MD2 from algorithm tables. This follows the recommendation in
1768 several standards that it is not used in new applications due to
1769 several cryptographic weaknesses. For binary compatibility reasons
1770 the MD2 API is still compiled in by default.
1771 [Steve Henson]
1772
1773 *) Add compression id to {d2i,i2d}_SSL_SESSION so it is correctly saved
1774 and restored.
1775 [Steve Henson]
1776
1777 *) Rename uni2asc and asc2uni functions to OPENSSL_uni2asc and
1778 OPENSSL_asc2uni conditionally on Netware platforms to avoid a name
1779 clash.
1780 [Guenter <lists@gknw.net>]
1781
1782 *) Fix the server certificate chain building code to use X509_verify_cert(),
1783 it used to have an ad-hoc builder which was unable to cope with anything
1784 other than a simple chain.
1785 [David Woodhouse <dwmw2@infradead.org>, Steve Henson]
1786
1787 *) Don't check self signed certificate signatures in X509_verify_cert()
1788 by default (a flag can override this): it just wastes time without
1789 adding any security. As a useful side effect self signed root CAs
1790 with non-FIPS digests are now usable in FIPS mode.
1791 [Steve Henson]
1792
1793 *) In dtls1_process_out_of_seq_message() the check if the current message
1794 is already buffered was missing. For every new message was memory
1795 allocated, allowing an attacker to perform an denial of service attack
1796 with sending out of seq handshake messages until there is no memory
1797 left. Additionally every future messege was buffered, even if the
1798 sequence number made no sense and would be part of another handshake.
1799 So only messages with sequence numbers less than 10 in advance will be
1800 buffered. (CVE-2009-1378)
1801 [Robin Seggelmann, discovered by Daniel Mentz]
1802
1803 *) Records are buffered if they arrive with a future epoch to be
1804 processed after finishing the corresponding handshake. There is
1805 currently no limitation to this buffer allowing an attacker to perform
1806 a DOS attack with sending records with future epochs until there is no
1807 memory left. This patch adds the pqueue_size() function to detemine
1808 the size of a buffer and limits the record buffer to 100 entries.
1809 (CVE-2009-1377)
1810 [Robin Seggelmann, discovered by Daniel Mentz]
1811
1812 *) Keep a copy of frag->msg_header.frag_len so it can be used after the
1813 parent structure is freed. (CVE-2009-1379)
1814 [Daniel Mentz]
1815
1816 *) Handle non-blocking I/O properly in SSL_shutdown() call.
1817 [Darryl Miles <darryl-mailinglists@netbauds.net>]
1818
1819 *) Add 2.5.4.* OIDs
1820 [Ilya O. <vrghost@gmail.com>]
1821
1822 Changes between 0.9.8k and 0.9.8l [5 Nov 2009]
1823
1824 *) Disable renegotiation completely - this fixes a severe security
1825 problem (CVE-2009-3555) at the cost of breaking all
1826 renegotiation. Renegotiation can be re-enabled by setting
1827 SSL3_FLAGS_ALLOW_UNSAFE_LEGACY_RENEGOTIATION in s3->flags at
1828 run-time. This is really not recommended unless you know what
1829 you're doing.
1830 [Ben Laurie]
1831
1832 Changes between 0.9.8j and 0.9.8k [25 Mar 2009]
1833
1834 *) Don't set val to NULL when freeing up structures, it is freed up by
1835 underlying code. If sizeof(void *) > sizeof(long) this can result in
1836 zeroing past the valid field. (CVE-2009-0789)
1837 [Paolo Ganci <Paolo.Ganci@AdNovum.CH>]
1838
1839 *) Fix bug where return value of CMS_SignerInfo_verify_content() was not
1840 checked correctly. This would allow some invalid signed attributes to
1841 appear to verify correctly. (CVE-2009-0591)
1842 [Ivan Nestlerode <inestlerode@us.ibm.com>]
1843
1844 *) Reject UniversalString and BMPString types with invalid lengths. This
1845 prevents a crash in ASN1_STRING_print_ex() which assumes the strings have
1846 a legal length. (CVE-2009-0590)
1847 [Steve Henson]
1848
1849 *) Set S/MIME signing as the default purpose rather than setting it
1850 unconditionally. This allows applications to override it at the store
1851 level.
1852 [Steve Henson]
1853
1854 *) Permit restricted recursion of ASN1 strings. This is needed in practice
1855 to handle some structures.
1856 [Steve Henson]
1857
1858 *) Improve efficiency of mem_gets: don't search whole buffer each time
1859 for a '\n'
1860 [Jeremy Shapiro <jnshapir@us.ibm.com>]
1861
1862 *) New -hex option for openssl rand.
1863 [Matthieu Herrb]
1864
1865 *) Print out UTF8String and NumericString when parsing ASN1.
1866 [Steve Henson]
1867
1868 *) Support NumericString type for name components.
1869 [Steve Henson]
1870
1871 *) Allow CC in the environment to override the automatically chosen
1872 compiler. Note that nothing is done to ensure flags work with the
1873 chosen compiler.
1874 [Ben Laurie]
1875
1876 Changes between 0.9.8i and 0.9.8j [07 Jan 2009]
1877
1878 *) Properly check EVP_VerifyFinal() and similar return values
1879 (CVE-2008-5077).
1880 [Ben Laurie, Bodo Moeller, Google Security Team]
1881
1882 *) Enable TLS extensions by default.
1883 [Ben Laurie]
1884
1885 *) Allow the CHIL engine to be loaded, whether the application is
1886 multithreaded or not. (This does not release the developer from the
1887 obligation to set up the dynamic locking callbacks.)
1888 [Sander Temme <sander@temme.net>]
1889
1890 *) Use correct exit code if there is an error in dgst command.
1891 [Steve Henson; problem pointed out by Roland Dirlewanger]
1892
1893 *) Tweak Configure so that you need to say "experimental-jpake" to enable
1894 JPAKE, and need to use -DOPENSSL_EXPERIMENTAL_JPAKE in applications.
1895 [Bodo Moeller]
1896
1897 *) Add experimental JPAKE support, including demo authentication in
1898 s_client and s_server.
1899 [Ben Laurie]
1900
1901 *) Set the comparison function in v3_addr_canonize().
1902 [Rob Austein <sra@hactrn.net>]
1903
1904 *) Add support for XMPP STARTTLS in s_client.
1905 [Philip Paeps <philip@freebsd.org>]
1906
1907 *) Change the server-side SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG behavior
1908 to ensure that even with this option, only ciphersuites in the
1909 server's preference list will be accepted. (Note that the option
1910 applies only when resuming a session, so the earlier behavior was
1911 just about the algorithm choice for symmetric cryptography.)
1912 [Bodo Moeller]
1913
1914 Changes between 0.9.8h and 0.9.8i [15 Sep 2008]
1915
1916 *) Fix NULL pointer dereference if a DTLS server received
1917 ChangeCipherSpec as first record (CVE-2009-1386).
1918 [PR #1679]
1919
1920 *) Fix a state transitition in s3_srvr.c and d1_srvr.c
1921 (was using SSL3_ST_CW_CLNT_HELLO_B, should be ..._ST_SW_SRVR_...).
1922 [Nagendra Modadugu]
1923
1924 *) The fix in 0.9.8c that supposedly got rid of unsafe
1925 double-checked locking was incomplete for RSA blinding,
1926 addressing just one layer of what turns out to have been
1927 doubly unsafe triple-checked locking.
1928
1929 So now fix this for real by retiring the MONT_HELPER macro
1930 in crypto/rsa/rsa_eay.c.
1931
1932 [Bodo Moeller; problem pointed out by Marius Schilder]
1933
1934 *) Various precautionary measures:
1935
1936 - Avoid size_t integer overflow in HASH_UPDATE (md32_common.h).
1937
1938 - Avoid a buffer overflow in d2i_SSL_SESSION() (ssl_asn1.c).
1939 (NB: This would require knowledge of the secret session ticket key
1940 to exploit, in which case you'd be SOL either way.)
1941
1942 - Change bn_nist.c so that it will properly handle input BIGNUMs
1943 outside the expected range.
1944
1945 - Enforce the 'num' check in BN_div() (bn_div.c) for non-BN_DEBUG
1946 builds.
1947
1948 [Neel Mehta, Bodo Moeller]
1949
1950 *) Allow engines to be "soft loaded" - i.e. optionally don't die if
1951 the load fails. Useful for distros.
1952 [Ben Laurie and the FreeBSD team]
1953
1954 *) Add support for Local Machine Keyset attribute in PKCS#12 files.
1955 [Steve Henson]
1956
1957 *) Fix BN_GF2m_mod_arr() top-bit cleanup code.
1958 [Huang Ying]
1959
1960 *) Expand ENGINE to support engine supplied SSL client certificate functions.
1961
1962 This work was sponsored by Logica.
1963 [Steve Henson]
1964
1965 *) Add CryptoAPI ENGINE to support use of RSA and DSA keys held in Windows
1966 keystores. Support for SSL/TLS client authentication too.
1967 Not compiled unless enable-capieng specified to Configure.
1968
1969 This work was sponsored by Logica.
1970 [Steve Henson]
1971
1972 *) Fix bug in X509_ATTRIBUTE creation: dont set attribute using
1973 ASN1_TYPE_set1 if MBSTRING flag set. This bug would crash certain
1974 attribute creation routines such as certifcate requests and PKCS#12
1975 files.
1976 [Steve Henson]
1977
1978 Changes between 0.9.8g and 0.9.8h [28 May 2008]
1979
1980 *) Fix flaw if 'Server Key exchange message' is omitted from a TLS
1981 handshake which could lead to a cilent crash as found using the
1982 Codenomicon TLS test suite (CVE-2008-1672)
1983 [Steve Henson, Mark Cox]
1984
1985 *) Fix double free in TLS server name extensions which could lead to
1986 a remote crash found by Codenomicon TLS test suite (CVE-2008-0891)
1987 [Joe Orton]
1988
1989 *) Clear error queue in SSL_CTX_use_certificate_chain_file()
1990
1991 Clear the error queue to ensure that error entries left from
1992 older function calls do not interfere with the correct operation.
1993 [Lutz Jaenicke, Erik de Castro Lopo]
1994
1995 *) Remove root CA certificates of commercial CAs:
1996
1997 The OpenSSL project does not recommend any specific CA and does not
1998 have any policy with respect to including or excluding any CA.
1999 Therefore it does not make any sense to ship an arbitrary selection
2000 of root CA certificates with the OpenSSL software.
2001 [Lutz Jaenicke]
2002
2003 *) RSA OAEP patches to fix two separate invalid memory reads.
2004 The first one involves inputs when 'lzero' is greater than
2005 'SHA_DIGEST_LENGTH' (it would read about SHA_DIGEST_LENGTH bytes
2006 before the beginning of from). The second one involves inputs where
2007 the 'db' section contains nothing but zeroes (there is a one-byte
2008 invalid read after the end of 'db').
2009 [Ivan Nestlerode <inestlerode@us.ibm.com>]
2010
2011 *) Partial backport from 0.9.9-dev:
2012
2013 Introduce bn_mul_mont (dedicated Montgomery multiplication
2014 procedure) as a candidate for BIGNUM assembler implementation.
2015 While 0.9.9-dev uses assembler for various architectures, only
2016 x86_64 is available by default here in the 0.9.8 branch, and
2017 32-bit x86 is available through a compile-time setting.
2018
2019 To try the 32-bit x86 assembler implementation, use Configure
2020 option "enable-montasm" (which exists only for this backport).
2021
2022 As "enable-montasm" for 32-bit x86 disclaims code stability
2023 anyway, in this constellation we activate additional code
2024 backported from 0.9.9-dev for further performance improvements,
2025 namely BN_from_montgomery_word. (To enable this otherwise,
2026 e.g. x86_64, try "-DMONT_FROM_WORD___NON_DEFAULT_0_9_8_BUILD".)
2027
2028 [Andy Polyakov (backport partially by Bodo Moeller)]
2029
2030 *) Add TLS session ticket callback. This allows an application to set
2031 TLS ticket cipher and HMAC keys rather than relying on hardcoded fixed
2032 values. This is useful for key rollover for example where several key
2033 sets may exist with different names.
2034 [Steve Henson]
2035
2036 *) Reverse ENGINE-internal logic for caching default ENGINE handles.
2037 This was broken until now in 0.9.8 releases, such that the only way
2038 a registered ENGINE could be used (assuming it initialises
2039 successfully on the host) was to explicitly set it as the default
2040 for the relevant algorithms. This is in contradiction with 0.9.7
2041 behaviour and the documentation. With this fix, when an ENGINE is
2042 registered into a given algorithm's table of implementations, the
2043 'uptodate' flag is reset so that auto-discovery will be used next
2044 time a new context for that algorithm attempts to select an
2045 implementation.
2046 [Ian Lister (tweaked by Geoff Thorpe)]
2047
2048 *) Backport of CMS code to OpenSSL 0.9.8. This differs from the 0.9.9
2049 implemention in the following ways:
2050
2051 Lack of EVP_PKEY_ASN1_METHOD means algorithm parameters have to be
2052 hard coded.
2053
2054 Lack of BER streaming support means one pass streaming processing is
2055 only supported if data is detached: setting the streaming flag is
2056 ignored for embedded content.
2057
2058 CMS support is disabled by default and must be explicitly enabled
2059 with the enable-cms configuration option.
2060 [Steve Henson]
2061
2062 *) Update the GMP engine glue to do direct copies between BIGNUM and
2063 mpz_t when openssl and GMP use the same limb size. Otherwise the
2064 existing "conversion via a text string export" trick is still used.
2065 [Paul Sheer <paulsheer@gmail.com>]
2066
2067 *) Zlib compression BIO. This is a filter BIO which compressed and
2068 uncompresses any data passed through it.
2069 [Steve Henson]
2070
2071 *) Add AES_wrap_key() and AES_unwrap_key() functions to implement
2072 RFC3394 compatible AES key wrapping.
2073 [Steve Henson]
2074
2075 *) Add utility functions to handle ASN1 structures. ASN1_STRING_set0():
2076 sets string data without copying. X509_ALGOR_set0() and
2077 X509_ALGOR_get0(): set and retrieve X509_ALGOR (AlgorithmIdentifier)
2078 data. Attribute function X509at_get0_data_by_OBJ(): retrieves data
2079 from an X509_ATTRIBUTE structure optionally checking it occurs only
2080 once. ASN1_TYPE_set1(): set and ASN1_TYPE structure copying supplied
2081 data.
2082 [Steve Henson]
2083
2084 *) Fix BN flag handling in RSA_eay_mod_exp() and BN_MONT_CTX_set()
2085 to get the expected BN_FLG_CONSTTIME behavior.
2086 [Bodo Moeller (Google)]
2087
2088 *) Netware support:
2089
2090 - fixed wrong usage of ioctlsocket() when build for LIBC BSD sockets
2091 - fixed do_tests.pl to run the test suite with CLIB builds too (CLIB_OPT)
2092 - added some more tests to do_tests.pl
2093 - fixed RunningProcess usage so that it works with newer LIBC NDKs too
2094 - removed usage of BN_LLONG for CLIB builds to avoid runtime dependency
2095 - added new Configure targets netware-clib-bsdsock, netware-clib-gcc,
2096 netware-clib-bsdsock-gcc, netware-libc-bsdsock-gcc
2097 - various changes to netware.pl to enable gcc-cross builds on Win32
2098 platform
2099 - changed crypto/bio/b_sock.c to work with macro functions (CLIB BSD)
2100 - various changes to fix missing prototype warnings
2101 - fixed x86nasm.pl to create correct asm files for NASM COFF output
2102 - added AES, WHIRLPOOL and CPUID assembler code to build files
2103 - added missing AES assembler make rules to mk1mf.pl
2104 - fixed order of includes in apps/ocsp.c so that e_os.h settings apply
2105 [Guenter Knauf <eflash@gmx.net>]
2106
2107 *) Implement certificate status request TLS extension defined in RFC3546.
2108 A client can set the appropriate parameters and receive the encoded
2109 OCSP response via a callback. A server can query the supplied parameters
2110 and set the encoded OCSP response in the callback. Add simplified examples
2111 to s_client and s_server.
2112 [Steve Henson]
2113
2114 Changes between 0.9.8f and 0.9.8g [19 Oct 2007]
2115
2116 *) Fix various bugs:
2117 + Binary incompatibility of ssl_ctx_st structure
2118 + DTLS interoperation with non-compliant servers
2119 + Don't call get_session_cb() without proposed session
2120 + Fix ia64 assembler code
2121 [Andy Polyakov, Steve Henson]
2122
2123 Changes between 0.9.8e and 0.9.8f [11 Oct 2007]
2124
2125 *) DTLS Handshake overhaul. There were longstanding issues with
2126 OpenSSL DTLS implementation, which were making it impossible for
2127 RFC 4347 compliant client to communicate with OpenSSL server.
2128 Unfortunately just fixing these incompatibilities would "cut off"
2129 pre-0.9.8f clients. To allow for hassle free upgrade post-0.9.8e
2130 server keeps tolerating non RFC compliant syntax. The opposite is
2131 not true, 0.9.8f client can not communicate with earlier server.
2132 This update even addresses CVE-2007-4995.
2133 [Andy Polyakov]
2134
2135 *) Changes to avoid need for function casts in OpenSSL: some compilers
2136 (gcc 4.2 and later) reject their use.
2137 [Kurt Roeckx <kurt@roeckx.be>, Peter Hartley <pdh@utter.chaos.org.uk>,
2138 Steve Henson]
2139
2140 *) Add RFC4507 support to OpenSSL. This includes the corrections in
2141 RFC4507bis. The encrypted ticket format is an encrypted encoded
2142 SSL_SESSION structure, that way new session features are automatically
2143 supported.
2144
2145 If a client application caches session in an SSL_SESSION structure
2146 support is transparent because tickets are now stored in the encoded
2147 SSL_SESSION.
2148
2149 The SSL_CTX structure automatically generates keys for ticket
2150 protection in servers so again support should be possible
2151 with no application modification.
2152
2153 If a client or server wishes to disable RFC4507 support then the option
2154 SSL_OP_NO_TICKET can be set.
2155
2156 Add a TLS extension debugging callback to allow the contents of any client
2157 or server extensions to be examined.
2158
2159 This work was sponsored by Google.
2160 [Steve Henson]
2161
2162 *) Add initial support for TLS extensions, specifically for the server_name
2163 extension so far. The SSL_SESSION, SSL_CTX, and SSL data structures now
2164 have new members for a host name. The SSL data structure has an
2165 additional member SSL_CTX *initial_ctx so that new sessions can be
2166 stored in that context to allow for session resumption, even after the
2167 SSL has been switched to a new SSL_CTX in reaction to a client's
2168 server_name extension.
2169
2170 New functions (subject to change):
2171
2172 SSL_get_servername()
2173 SSL_get_servername_type()
2174 SSL_set_SSL_CTX()
2175
2176 New CTRL codes and macros (subject to change):
2177
2178 SSL_CTRL_SET_TLSEXT_SERVERNAME_CB
2179 - SSL_CTX_set_tlsext_servername_callback()
2180 SSL_CTRL_SET_TLSEXT_SERVERNAME_ARG
2181 - SSL_CTX_set_tlsext_servername_arg()
2182 SSL_CTRL_SET_TLSEXT_HOSTNAME - SSL_set_tlsext_host_name()
2183
2184 openssl s_client has a new '-servername ...' option.
2185
2186 openssl s_server has new options '-servername_host ...', '-cert2 ...',
2187 '-key2 ...', '-servername_fatal' (subject to change). This allows
2188 testing the HostName extension for a specific single host name ('-cert'
2189 and '-key' remain fallbacks for handshakes without HostName
2190 negotiation). If the unrecogninzed_name alert has to be sent, this by
2191 default is a warning; it becomes fatal with the '-servername_fatal'
2192 option.
2193
2194 [Peter Sylvester, Remy Allais, Christophe Renou, Steve Henson]
2195
2196 *) Add AES and SSE2 assembly language support to VC++ build.
2197 [Steve Henson]
2198
2199 *) Mitigate attack on final subtraction in Montgomery reduction.
2200 [Andy Polyakov]
2201
2202 *) Fix crypto/ec/ec_mult.c to work properly with scalars of value 0
2203 (which previously caused an internal error).
2204 [Bodo Moeller]
2205
2206 *) Squeeze another 10% out of IGE mode when in != out.
2207 [Ben Laurie]
2208
2209 *) AES IGE mode speedup.
2210 [Dean Gaudet (Google)]
2211
2212 *) Add the Korean symmetric 128-bit cipher SEED (see
2213 http://www.kisa.or.kr/kisa/seed/jsp/seed_eng.jsp) and
2214 add SEED ciphersuites from RFC 4162:
2215
2216 TLS_RSA_WITH_SEED_CBC_SHA = "SEED-SHA"
2217 TLS_DHE_DSS_WITH_SEED_CBC_SHA = "DHE-DSS-SEED-SHA"
2218 TLS_DHE_RSA_WITH_SEED_CBC_SHA = "DHE-RSA-SEED-SHA"
2219 TLS_DH_anon_WITH_SEED_CBC_SHA = "ADH-SEED-SHA"
2220
2221 To minimize changes between patchlevels in the OpenSSL 0.9.8
2222 series, SEED remains excluded from compilation unless OpenSSL
2223 is configured with 'enable-seed'.
2224 [KISA, Bodo Moeller]
2225
2226 *) Mitigate branch prediction attacks, which can be practical if a
2227 single processor is shared, allowing a spy process to extract
2228 information. For detailed background information, see
2229 http://eprint.iacr.org/2007/039 (O. Aciicmez, S. Gueron,
2230 J.-P. Seifert, "New Branch Prediction Vulnerabilities in OpenSSL
2231 and Necessary Software Countermeasures"). The core of the change
2232 are new versions BN_div_no_branch() and
2233 BN_mod_inverse_no_branch() of BN_div() and BN_mod_inverse(),
2234 respectively, which are slower, but avoid the security-relevant
2235 conditional branches. These are automatically called by BN_div()
2236 and BN_mod_inverse() if the flag BN_FLG_CONSTTIME is set for one
2237 of the input BIGNUMs. Also, BN_is_bit_set() has been changed to
2238 remove a conditional branch.
2239
2240 BN_FLG_CONSTTIME is the new name for the previous
2241 BN_FLG_EXP_CONSTTIME flag, since it now affects more than just
2242 modular exponentiation. (Since OpenSSL 0.9.7h, setting this flag
2243 in the exponent causes BN_mod_exp_mont() to use the alternative
2244 implementation in BN_mod_exp_mont_consttime().) The old name
2245 remains as a deprecated alias.
2246
2247 Similary, RSA_FLAG_NO_EXP_CONSTTIME is replaced by a more general
2248 RSA_FLAG_NO_CONSTTIME flag since the RSA implementation now uses
2249 constant-time implementations for more than just exponentiation.
2250 Here too the old name is kept as a deprecated alias.
2251
2252 BN_BLINDING_new() will now use BN_dup() for the modulus so that
2253 the BN_BLINDING structure gets an independent copy of the
2254 modulus. This means that the previous "BIGNUM *m" argument to
2255 BN_BLINDING_new() and to BN_BLINDING_create_param() now
2256 essentially becomes "const BIGNUM *m", although we can't actually
2257 change this in the header file before 0.9.9. It allows
2258 RSA_setup_blinding() to use BN_with_flags() on the modulus to
2259 enable BN_FLG_CONSTTIME.
2260
2261 [Matthew D Wood (Intel Corp)]
2262
2263 *) In the SSL/TLS server implementation, be strict about session ID
2264 context matching (which matters if an application uses a single
2265 external cache for different purposes). Previously,
2266 out-of-context reuse was forbidden only if SSL_VERIFY_PEER was
2267 set. This did ensure strict client verification, but meant that,
2268 with applications using a single external cache for quite
2269 different requirements, clients could circumvent ciphersuite
2270 restrictions for a given session ID context by starting a session
2271 in a different context.
2272 [Bodo Moeller]
2273
2274 *) Include "!eNULL" in SSL_DEFAULT_CIPHER_LIST to make sure that
2275 a ciphersuite string such as "DEFAULT:RSA" cannot enable
2276 authentication-only ciphersuites.
2277 [Bodo Moeller]
2278
2279 *) Update the SSL_get_shared_ciphers() fix CVE-2006-3738 which was
2280 not complete and could lead to a possible single byte overflow
2281 (CVE-2007-5135) [Ben Laurie]
2282
2283 Changes between 0.9.8d and 0.9.8e [23 Feb 2007]
2284
2285 *) Since AES128 and AES256 (and similarly Camellia128 and
2286 Camellia256) share a single mask bit in the logic of
2287 ssl/ssl_ciph.c, the code for masking out disabled ciphers needs a
2288 kludge to work properly if AES128 is available and AES256 isn't
2289 (or if Camellia128 is available and Camellia256 isn't).
2290 [Victor Duchovni]
2291
2292 *) Fix the BIT STRING encoding generated by crypto/ec/ec_asn1.c
2293 (within i2d_ECPrivateKey, i2d_ECPKParameters, i2d_ECParameters):
2294 When a point or a seed is encoded in a BIT STRING, we need to
2295 prevent the removal of trailing zero bits to get the proper DER
2296 encoding. (By default, crypto/asn1/a_bitstr.c assumes the case
2297 of a NamedBitList, for which trailing 0 bits need to be removed.)
2298 [Bodo Moeller]
2299
2300 *) Have SSL/TLS server implementation tolerate "mismatched" record
2301 protocol version while receiving ClientHello even if the
2302 ClientHello is fragmented. (The server can't insist on the
2303 particular protocol version it has chosen before the ServerHello
2304 message has informed the client about his choice.)
2305 [Bodo Moeller]
2306
2307 *) Add RFC 3779 support.
2308 [Rob Austein for ARIN, Ben Laurie]
2309
2310 *) Load error codes if they are not already present instead of using a
2311 static variable. This allows them to be cleanly unloaded and reloaded.
2312 Improve header file function name parsing.
2313 [Steve Henson]
2314
2315 *) extend SMTP and IMAP protocol emulation in s_client to use EHLO
2316 or CAPABILITY handshake as required by RFCs.
2317 [Goetz Babin-Ebell]
2318
2319 Changes between 0.9.8c and 0.9.8d [28 Sep 2006]
2320
2321 *) Introduce limits to prevent malicious keys being able to
2322 cause a denial of service. (CVE-2006-2940)
2323 [Steve Henson, Bodo Moeller]
2324
2325 *) Fix ASN.1 parsing of certain invalid structures that can result
2326 in a denial of service. (CVE-2006-2937) [Steve Henson]
2327
2328 *) Fix buffer overflow in SSL_get_shared_ciphers() function.
2329 (CVE-2006-3738) [Tavis Ormandy and Will Drewry, Google Security Team]
2330
2331 *) Fix SSL client code which could crash if connecting to a
2332 malicious SSLv2 server. (CVE-2006-4343)
2333 [Tavis Ormandy and Will Drewry, Google Security Team]
2334
2335 *) Since 0.9.8b, ciphersuite strings naming explicit ciphersuites
2336 match only those. Before that, "AES256-SHA" would be interpreted
2337 as a pattern and match "AES128-SHA" too (since AES128-SHA got
2338 the same strength classification in 0.9.7h) as we currently only
2339 have a single AES bit in the ciphersuite description bitmap.
2340 That change, however, also applied to ciphersuite strings such as
2341 "RC4-MD5" that intentionally matched multiple ciphersuites --
2342 namely, SSL 2.0 ciphersuites in addition to the more common ones
2343 from SSL 3.0/TLS 1.0.
2344
2345 So we change the selection algorithm again: Naming an explicit
2346 ciphersuite selects this one ciphersuite, and any other similar
2347 ciphersuite (same bitmap) from *other* protocol versions.
2348 Thus, "RC4-MD5" again will properly select both the SSL 2.0
2349 ciphersuite and the SSL 3.0/TLS 1.0 ciphersuite.
2350
2351 Since SSL 2.0 does not have any ciphersuites for which the
2352 128/256 bit distinction would be relevant, this works for now.
2353 The proper fix will be to use different bits for AES128 and
2354 AES256, which would have avoided the problems from the beginning;
2355 however, bits are scarce, so we can only do this in a new release
2356 (not just a patchlevel) when we can change the SSL_CIPHER
2357 definition to split the single 'unsigned long mask' bitmap into
2358 multiple values to extend the available space.
2359
2360 [Bodo Moeller]
2361
2362 Changes between 0.9.8b and 0.9.8c [05 Sep 2006]
2363
2364 *) Avoid PKCS #1 v1.5 signature attack discovered by Daniel Bleichenbacher
2365 (CVE-2006-4339) [Ben Laurie and Google Security Team]
2366
2367 *) Add AES IGE and biIGE modes.
2368 [Ben Laurie]
2369
2370 *) Change the Unix randomness entropy gathering to use poll() when
2371 possible instead of select(), since the latter has some
2372 undesirable limitations.
2373 [Darryl Miles via Richard Levitte and Bodo Moeller]
2374
2375 *) Disable "ECCdraft" ciphersuites more thoroughly. Now special
2376 treatment in ssl/ssl_ciph.s makes sure that these ciphersuites
2377 cannot be implicitly activated as part of, e.g., the "AES" alias.
2378 However, please upgrade to OpenSSL 0.9.9[-dev] for
2379 non-experimental use of the ECC ciphersuites to get TLS extension
2380 support, which is required for curve and point format negotiation
2381 to avoid potential handshake problems.
2382 [Bodo Moeller]
2383
2384 *) Disable rogue ciphersuites:
2385
2386 - SSLv2 0x08 0x00 0x80 ("RC4-64-MD5")
2387 - SSLv3/TLSv1 0x00 0x61 ("EXP1024-RC2-CBC-MD5")
2388 - SSLv3/TLSv1 0x00 0x60 ("EXP1024-RC4-MD5")
2389
2390 The latter two were purportedly from
2391 draft-ietf-tls-56-bit-ciphersuites-0[01].txt, but do not really
2392 appear there.
2393
2394 Also deactivate the remaining ciphersuites from
2395 draft-ietf-tls-56-bit-ciphersuites-01.txt. These are just as
2396 unofficial, and the ID has long expired.
2397 [Bodo Moeller]
2398
2399 *) Fix RSA blinding Heisenbug (problems sometimes occured on
2400 dual-core machines) and other potential thread-safety issues.
2401 [Bodo Moeller]
2402
2403 *) Add the symmetric cipher Camellia (128-bit, 192-bit, 256-bit key
2404 versions), which is now available for royalty-free use
2405 (see http://info.isl.ntt.co.jp/crypt/eng/info/chiteki.html).
2406 Also, add Camellia TLS ciphersuites from RFC 4132.
2407
2408 To minimize changes between patchlevels in the OpenSSL 0.9.8
2409 series, Camellia remains excluded from compilation unless OpenSSL
2410 is configured with 'enable-camellia'.
2411 [NTT]
2412
2413 *) Disable the padding bug check when compression is in use. The padding
2414 bug check assumes the first packet is of even length, this is not
2415 necessarily true if compresssion is enabled and can result in false
2416 positives causing handshake failure. The actual bug test is ancient
2417 code so it is hoped that implementations will either have fixed it by
2418 now or any which still have the bug do not support compression.
2419 [Steve Henson]
2420
2421 Changes between 0.9.8a and 0.9.8b [04 May 2006]
2422
2423 *) When applying a cipher rule check to see if string match is an explicit
2424 cipher suite and only match that one cipher suite if it is.
2425 [Steve Henson]
2426
2427 *) Link in manifests for VC++ if needed.
2428 [Austin Ziegler <halostatue@gmail.com>]
2429
2430 *) Update support for ECC-based TLS ciphersuites according to
2431 draft-ietf-tls-ecc-12.txt with proposed changes (but without
2432 TLS extensions, which are supported starting with the 0.9.9
2433 branch, not in the OpenSSL 0.9.8 branch).
2434 [Douglas Stebila]
2435
2436 *) New functions EVP_CIPHER_CTX_new() and EVP_CIPHER_CTX_free() to support
2437 opaque EVP_CIPHER_CTX handling.
2438 [Steve Henson]
2439
2440 *) Fixes and enhancements to zlib compression code. We now only use
2441 "zlib1.dll" and use the default __cdecl calling convention on Win32
2442 to conform with the standards mentioned here:
2443 http://www.zlib.net/DLL_FAQ.txt
2444 Static zlib linking now works on Windows and the new --with-zlib-include
2445 --with-zlib-lib options to Configure can be used to supply the location
2446 of the headers and library. Gracefully handle case where zlib library
2447 can't be loaded.
2448 [Steve Henson]
2449
2450 *) Several fixes and enhancements to the OID generation code. The old code
2451 sometimes allowed invalid OIDs (1.X for X >= 40 for example), couldn't
2452 handle numbers larger than ULONG_MAX, truncated printing and had a
2453 non standard OBJ_obj2txt() behaviour.
2454 [Steve Henson]
2455
2456 *) Add support for building of engines under engine/ as shared libraries
2457 under VC++ build system.
2458 [Steve Henson]
2459
2460 *) Corrected the numerous bugs in the Win32 path splitter in DSO.
2461 Hopefully, we will not see any false combination of paths any more.
2462 [Richard Levitte]
2463
2464 Changes between 0.9.8 and 0.9.8a [11 Oct 2005]
2465
2466 *) Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING
2467 (part of SSL_OP_ALL). This option used to disable the
2468 countermeasure against man-in-the-middle protocol-version
2469 rollback in the SSL 2.0 server implementation, which is a bad
2470 idea. (CVE-2005-2969)
2471
2472 [Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center
2473 for Information Security, National Institute of Advanced Industrial
2474 Science and Technology [AIST], Japan)]
2475
2476 *) Add two function to clear and return the verify parameter flags.
2477 [Steve Henson]
2478
2479 *) Keep cipherlists sorted in the source instead of sorting them at
2480 runtime, thus removing the need for a lock.
2481 [Nils Larsch]
2482
2483 *) Avoid some small subgroup attacks in Diffie-Hellman.
2484 [Nick Mathewson and Ben Laurie]
2485
2486 *) Add functions for well-known primes.
2487 [Nick Mathewson]
2488
2489 *) Extended Windows CE support.
2490 [Satoshi Nakamura and Andy Polyakov]
2491
2492 *) Initialize SSL_METHOD structures at compile time instead of during
2493 runtime, thus removing the need for a lock.
2494 [Steve Henson]
2495
2496 *) Make PKCS7_decrypt() work even if no certificate is supplied by
2497 attempting to decrypt each encrypted key in turn. Add support to
2498 smime utility.
2499 [Steve Henson]
2500
2501 Changes between 0.9.7h and 0.9.8 [05 Jul 2005]
2502
2503 [NB: OpenSSL 0.9.7i and later 0.9.7 patch levels were released after
2504 OpenSSL 0.9.8.]
2505
2506 *) Add libcrypto.pc and libssl.pc for those who feel they need them.
2507 [Richard Levitte]
2508
2509 *) Change CA.sh and CA.pl so they don't bundle the CSR and the private
2510 key into the same file any more.
2511 [Richard Levitte]
2512
2513 *) Add initial support for Win64, both IA64 and AMD64/x64 flavors.
2514 [Andy Polyakov]
2515
2516 *) Add -utf8 command line and config file option to 'ca'.
2517 [Stefan <stf@udoma.org]
2518
2519 *) Removed the macro des_crypt(), as it seems to conflict with some
2520 libraries. Use DES_crypt().
2521 [Richard Levitte]
2522
2523 *) Correct naming of the 'chil' and '4758cca' ENGINEs. This
2524 involves renaming the source and generated shared-libs for
2525 both. The engines will accept the corrected or legacy ids
2526 ('ncipher' and '4758_cca' respectively) when binding. NB,
2527 this only applies when building 'shared'.
2528 [Corinna Vinschen <vinschen@redhat.com> and Geoff Thorpe]
2529
2530 *) Add attribute functions to EVP_PKEY structure. Modify
2531 PKCS12_create() to recognize a CSP name attribute and
2532 use it. Make -CSP option work again in pkcs12 utility.
2533 [Steve Henson]
2534
2535 *) Add new functionality to the bn blinding code:
2536 - automatic re-creation of the BN_BLINDING parameters after
2537 a fixed number of uses (currently 32)
2538 - add new function for parameter creation
2539 - introduce flags to control the update behaviour of the
2540 BN_BLINDING parameters
2541 - hide BN_BLINDING structure
2542 Add a second BN_BLINDING slot to the RSA structure to improve
2543 performance when a single RSA object is shared among several
2544 threads.
2545 [Nils Larsch]
2546
2547 *) Add support for DTLS.
2548 [Nagendra Modadugu <nagendra@cs.stanford.edu> and Ben Laurie]
2549
2550 *) Add support for DER encoded private keys (SSL_FILETYPE_ASN1)
2551 to SSL_CTX_use_PrivateKey_file() and SSL_use_PrivateKey_file()
2552 [Walter Goulet]
2553
2554 *) Remove buggy and incompletet DH cert support from
2555 ssl/ssl_rsa.c and ssl/s3_both.c
2556 [Nils Larsch]
2557
2558 *) Use SHA-1 instead of MD5 as the default digest algorithm for
2559 the apps/openssl applications.
2560 [Nils Larsch]
2561
2562 *) Compile clean with "-Wall -Wmissing-prototypes
2563 -Wstrict-prototypes -Wmissing-declarations -Werror". Currently
2564 DEBUG_SAFESTACK must also be set.
2565 [Ben Laurie]
2566
2567 *) Change ./Configure so that certain algorithms can be disabled by default.
2568 The new counterpiece to "no-xxx" is "enable-xxx".
2569
2570 The patented RC5 and MDC2 algorithms will now be disabled unless
2571 "enable-rc5" and "enable-mdc2", respectively, are specified.
2572
2573 (IDEA remains enabled despite being patented. This is because IDEA
2574 is frequently required for interoperability, and there is no license
2575 fee for non-commercial use. As before, "no-idea" can be used to
2576 avoid this algorithm.)
2577
2578 [Bodo Moeller]
2579
2580 *) Add processing of proxy certificates (see RFC 3820). This work was
2581 sponsored by KTH (The Royal Institute of Technology in Stockholm) and
2582 EGEE (Enabling Grids for E-science in Europe).
2583 [Richard Levitte]
2584
2585 *) RC4 performance overhaul on modern architectures/implementations, such
2586 as Intel P4, IA-64 and AMD64.
2587 [Andy Polyakov]
2588
2589 *) New utility extract-section.pl. This can be used specify an alternative
2590 section number in a pod file instead of having to treat each file as
2591 a separate case in Makefile. This can be done by adding two lines to the
2592 pod file:
2593
2594 =for comment openssl_section:XXX
2595
2596 The blank line is mandatory.
2597
2598 [Steve Henson]
2599
2600 *) New arguments -certform, -keyform and -pass for s_client and s_server
2601 to allow alternative format key and certificate files and passphrase
2602 sources.
2603 [Steve Henson]
2604
2605 *) New structure X509_VERIFY_PARAM which combines current verify parameters,
2606 update associated structures and add various utility functions.
2607
2608 Add new policy related verify parameters, include policy checking in
2609 standard verify code. Enhance 'smime' application with extra parameters
2610 to support policy checking and print out.
2611 [Steve Henson]
2612
2613 *) Add a new engine to support VIA PadLock ACE extensions in the VIA C3
2614 Nehemiah processors. These extensions support AES encryption in hardware
2615 as well as RNG (though RNG support is currently disabled).
2616 [Michal Ludvig <michal@logix.cz>, with help from Andy Polyakov]
2617
2618 *) Deprecate BN_[get|set]_params() functions (they were ignored internally).
2619 [Geoff Thorpe]
2620
2621 *) New FIPS 180-2 algorithms, SHA-224/-256/-384/-512 are implemented.
2622 [Andy Polyakov and a number of other people]
2623
2624 *) Improved PowerPC platform support. Most notably BIGNUM assembler
2625 implementation contributed by IBM.
2626 [Suresh Chari, Peter Waltenberg, Andy Polyakov]
2627
2628 *) The new 'RSA_generate_key_ex' function now takes a BIGNUM for the public
2629 exponent rather than 'unsigned long'. There is a corresponding change to
2630 the new 'rsa_keygen' element of the RSA_METHOD structure.
2631 [Jelte Jansen, Geoff Thorpe]
2632
2633 *) Functionality for creating the initial serial number file is now
2634 moved from CA.pl to the 'ca' utility with a new option -create_serial.
2635
2636 (Before OpenSSL 0.9.7e, CA.pl used to initialize the serial
2637 number file to 1, which is bound to cause problems. To avoid
2638 the problems while respecting compatibility between different 0.9.7
2639 patchlevels, 0.9.7e employed 'openssl x509 -next_serial' in
2640 CA.pl for serial number initialization. With the new release 0.9.8,
2641 we can fix the problem directly in the 'ca' utility.)
2642 [Steve Henson]
2643
2644 *) Reduced header interdepencies by declaring more opaque objects in
2645 ossl_typ.h. As a consequence, including some headers (eg. engine.h) will
2646 give fewer recursive includes, which could break lazy source code - so
2647 this change is covered by the OPENSSL_NO_DEPRECATED symbol. As always,
2648 developers should define this symbol when building and using openssl to
2649 ensure they track the recommended behaviour, interfaces, [etc], but
2650 backwards-compatible behaviour prevails when this isn't defined.
2651 [Geoff Thorpe]
2652
2653 *) New function X509_POLICY_NODE_print() which prints out policy nodes.
2654 [Steve Henson]
2655
2656 *) Add new EVP function EVP_CIPHER_CTX_rand_key and associated functionality.
2657 This will generate a random key of the appropriate length based on the
2658 cipher context. The EVP_CIPHER can provide its own random key generation
2659 routine to support keys of a specific form. This is used in the des and
2660 3des routines to generate a key of the correct parity. Update S/MIME
2661 code to use new functions and hence generate correct parity DES keys.
2662 Add EVP_CHECK_DES_KEY #define to return an error if the key is not
2663 valid (weak or incorrect parity).
2664 [Steve Henson]
2665
2666 *) Add a local set of CRLs that can be used by X509_verify_cert() as well
2667 as looking them up. This is useful when the verified structure may contain
2668 CRLs, for example PKCS#7 signedData. Modify PKCS7_verify() to use any CRLs
2669 present unless the new PKCS7_NO_CRL flag is asserted.
2670 [Steve Henson]
2671
2672 *) Extend ASN1 oid configuration module. It now additionally accepts the
2673 syntax:
2674
2675 shortName = some long name, 1.2.3.4
2676 [Steve Henson]
2677
2678 *) Reimplemented the BN_CTX implementation. There is now no more static
2679 limitation on the number of variables it can handle nor the depth of the
2680 "stack" handling for BN_CTX_start()/BN_CTX_end() pairs. The stack
2681 information can now expand as required, and rather than having a single
2682 static array of bignums, BN_CTX now uses a linked-list of such arrays
2683 allowing it to expand on demand whilst maintaining the usefulness of
2684 BN_CTX's "bundling".
2685 [Geoff Thorpe]
2686
2687 *) Add a missing BN_CTX parameter to the 'rsa_mod_exp' callback in RSA_METHOD
2688 to allow all RSA operations to function using a single BN_CTX.
2689 [Geoff Thorpe]
2690
2691 *) Preliminary support for certificate policy evaluation and checking. This
2692 is initially intended to pass the tests outlined in "Conformance Testing
2693 of Relying Party Client Certificate Path Processing Logic" v1.07.
2694 [Steve Henson]
2695
2696 *) bn_dup_expand() has been deprecated, it was introduced in 0.9.7 and
2697 remained unused and not that useful. A variety of other little bignum
2698 tweaks and fixes have also been made continuing on from the audit (see
2699 below).
2700 [Geoff Thorpe]
2701
2702 *) Constify all or almost all d2i, c2i, s2i and r2i functions, along with
2703 associated ASN1, EVP and SSL functions and old ASN1 macros.
2704 [Richard Levitte]
2705
2706 *) BN_zero() only needs to set 'top' and 'neg' to zero for correct results,
2707 and this should never fail. So the return value from the use of
2708 BN_set_word() (which can fail due to needless expansion) is now deprecated;
2709 if OPENSSL_NO_DEPRECATED is defined, BN_zero() is a void macro.
2710 [Geoff Thorpe]
2711
2712 *) BN_CTX_get() should return zero-valued bignums, providing the same
2713 initialised value as BN_new().
2714 [Geoff Thorpe, suggested by Ulf Möller]
2715
2716 *) Support for inhibitAnyPolicy certificate extension.
2717 [Steve Henson]
2718
2719 *) An audit of the BIGNUM code is underway, for which debugging code is
2720 enabled when BN_DEBUG is defined. This makes stricter enforcements on what
2721 is considered valid when processing BIGNUMs, and causes execution to
2722 assert() when a problem is discovered. If BN_DEBUG_RAND is defined,
2723 further steps are taken to deliberately pollute unused data in BIGNUM
2724 structures to try and expose faulty code further on. For now, openssl will
2725 (in its default mode of operation) continue to tolerate the inconsistent
2726 forms that it has tolerated in the past, but authors and packagers should
2727 consider trying openssl and their own applications when compiled with
2728 these debugging symbols defined. It will help highlight potential bugs in
2729 their own code, and will improve the test coverage for OpenSSL itself. At
2730 some point, these tighter rules will become openssl's default to improve
2731 maintainability, though the assert()s and other overheads will remain only
2732 in debugging configurations. See bn.h for more details.
2733 [Geoff Thorpe, Nils Larsch, Ulf Möller]
2734
2735 *) BN_CTX_init() has been deprecated, as BN_CTX is an opaque structure
2736 that can only be obtained through BN_CTX_new() (which implicitly
2737 initialises it). The presence of this function only made it possible
2738 to overwrite an existing structure (and cause memory leaks).
2739 [Geoff Thorpe]
2740
2741 *) Because of the callback-based approach for implementing LHASH as a
2742 template type, lh_insert() adds opaque objects to hash-tables and
2743 lh_doall() or lh_doall_arg() are typically used with a destructor callback
2744 to clean up those corresponding objects before destroying the hash table
2745 (and losing the object pointers). So some over-zealous constifications in
2746 LHASH have been relaxed so that lh_insert() does not take (nor store) the
2747 objects as "const" and the lh_doall[_arg] callback wrappers are not
2748 prototyped to have "const" restrictions on the object pointers they are
2749 given (and so aren't required to cast them away any more).
2750 [Geoff Thorpe]
2751
2752 *) The tmdiff.h API was so ugly and minimal that our own timing utility
2753 (speed) prefers to use its own implementation. The two implementations
2754 haven't been consolidated as yet (volunteers?) but the tmdiff API has had
2755 its object type properly exposed (MS_TM) instead of casting to/from "char
2756 *". This may still change yet if someone realises MS_TM and "ms_time_***"
2757 aren't necessarily the greatest nomenclatures - but this is what was used
2758 internally to the implementation so I've used that for now.
2759 [Geoff Thorpe]
2760
2761 *) Ensure that deprecated functions do not get compiled when
2762 OPENSSL_NO_DEPRECATED is defined. Some "openssl" subcommands and a few of
2763 the self-tests were still using deprecated key-generation functions so
2764 these have been updated also.
2765 [Geoff Thorpe]
2766
2767 *) Reorganise PKCS#7 code to separate the digest location functionality
2768 into PKCS7_find_digest(), digest addtion into PKCS7_bio_add_digest().
2769 New function PKCS7_set_digest() to set the digest type for PKCS#7
2770 digestedData type. Add additional code to correctly generate the
2771 digestedData type and add support for this type in PKCS7 initialization
2772 functions.
2773 [Steve Henson]
2774
2775 *) New function PKCS7_set0_type_other() this initializes a PKCS7
2776 structure of type "other".
2777 [Steve Henson]
2778
2779 *) Fix prime generation loop in crypto/bn/bn_prime.pl by making
2780 sure the loop does correctly stop and breaking ("division by zero")
2781 modulus operations are not performed. The (pre-generated) prime
2782 table crypto/bn/bn_prime.h was already correct, but it could not be
2783 re-generated on some platforms because of the "division by zero"
2784 situation in the script.
2785 [Ralf S. Engelschall]
2786
2787 *) Update support for ECC-based TLS ciphersuites according to
2788 draft-ietf-tls-ecc-03.txt: the KDF1 key derivation function with
2789 SHA-1 now is only used for "small" curves (where the
2790 representation of a field element takes up to 24 bytes); for
2791 larger curves, the field element resulting from ECDH is directly
2792 used as premaster secret.
2793 [Douglas Stebila (Sun Microsystems Laboratories)]
2794
2795 *) Add code for kP+lQ timings to crypto/ec/ectest.c, and add SEC2
2796 curve secp160r1 to the tests.
2797 [Douglas Stebila (Sun Microsystems Laboratories)]
2798
2799 *) Add the possibility to load symbols globally with DSO.
2800 [Götz Babin-Ebell <babin-ebell@trustcenter.de> via Richard Levitte]
2801
2802 *) Add the functions ERR_set_mark() and ERR_pop_to_mark() for better
2803 control of the error stack.
2804 [Richard Levitte]
2805
2806 *) Add support for STORE in ENGINE.
2807 [Richard Levitte]
2808
2809 *) Add the STORE type. The intention is to provide a common interface
2810 to certificate and key stores, be they simple file-based stores, or
2811 HSM-type store, or LDAP stores, or...
2812 NOTE: The code is currently UNTESTED and isn't really used anywhere.
2813 [Richard Levitte]
2814
2815 *) Add a generic structure called OPENSSL_ITEM. This can be used to
2816 pass a list of arguments to any function as well as provide a way
2817 for a function to pass data back to the caller.
2818 [Richard Levitte]
2819
2820 *) Add the functions BUF_strndup() and BUF_memdup(). BUF_strndup()
2821 works like BUF_strdup() but can be used to duplicate a portion of
2822 a string. The copy gets NUL-terminated. BUF_memdup() duplicates
2823 a memory area.
2824 [Richard Levitte]
2825
2826 *) Add the function sk_find_ex() which works like sk_find(), but will
2827 return an index to an element even if an exact match couldn't be
2828 found. The index is guaranteed to point at the element where the
2829 searched-for key would be inserted to preserve sorting order.
2830 [Richard Levitte]
2831
2832 *) Add the function OBJ_bsearch_ex() which works like OBJ_bsearch() but
2833 takes an extra flags argument for optional functionality. Currently,
2834 the following flags are defined:
2835
2836 OBJ_BSEARCH_VALUE_ON_NOMATCH
2837 This one gets OBJ_bsearch_ex() to return a pointer to the first
2838 element where the comparing function returns a negative or zero
2839 number.
2840
2841 OBJ_BSEARCH_FIRST_VALUE_ON_MATCH
2842 This one gets OBJ_bsearch_ex() to return a pointer to the first
2843 element where the comparing function returns zero. This is useful
2844 if there are more than one element where the comparing function
2845 returns zero.
2846 [Richard Levitte]
2847
2848 *) Make it possible to create self-signed certificates with 'openssl ca'
2849 in such a way that the self-signed certificate becomes part of the
2850 CA database and uses the same mechanisms for serial number generation
2851 as all other certificate signing. The new flag '-selfsign' enables
2852 this functionality. Adapt CA.sh and CA.pl.in.
2853 [Richard Levitte]
2854
2855 *) Add functionality to check the public key of a certificate request
2856 against a given private. This is useful to check that a certificate
2857 request can be signed by that key (self-signing).
2858 [Richard Levitte]
2859
2860 *) Make it possible to have multiple active certificates with the same
2861 subject in the CA index file. This is done only if the keyword
2862 'unique_subject' is set to 'no' in the main CA section (default
2863 if 'CA_default') of the configuration file. The value is saved
2864 with the database itself in a separate index attribute file,
2865 named like the index file with '.attr' appended to the name.
2866 [Richard Levitte]
2867
2868 *) Generate muti valued AVAs using '+' notation in config files for
2869 req and dirName.
2870 [Steve Henson]
2871
2872 *) Support for nameConstraints certificate extension.
2873 [Steve Henson]
2874
2875 *) Support for policyConstraints certificate extension.
2876 [Steve Henson]
2877
2878 *) Support for policyMappings certificate extension.
2879 [Steve Henson]
2880
2881 *) Make sure the default DSA_METHOD implementation only uses its
2882 dsa_mod_exp() and/or bn_mod_exp() handlers if they are non-NULL,
2883 and change its own handlers to be NULL so as to remove unnecessary
2884 indirection. This lets alternative implementations fallback to the
2885 default implementation more easily.
2886 [Geoff Thorpe]
2887
2888 *) Support for directoryName in GeneralName related extensions
2889 in config files.
2890 [Steve Henson]
2891
2892 *) Make it possible to link applications using Makefile.shared.
2893 Make that possible even when linking against static libraries!
2894 [Richard Levitte]
2895
2896 *) Support for single pass processing for S/MIME signing. This now
2897 means that S/MIME signing can be done from a pipe, in addition
2898 cleartext signing (multipart/signed type) is effectively streaming
2899 and the signed data does not need to be all held in memory.
2900
2901 This is done with a new flag PKCS7_STREAM. When this flag is set
2902 PKCS7_sign() only initializes the PKCS7 structure and the actual signing
2903 is done after the data is output (and digests calculated) in
2904 SMIME_write_PKCS7().
2905 [Steve Henson]
2906
2907 *) Add full support for -rpath/-R, both in shared libraries and
2908 applications, at least on the platforms where it's known how
2909 to do it.
2910 [Richard Levitte]
2911
2912 *) In crypto/ec/ec_mult.c, implement fast point multiplication with
2913 precomputation, based on wNAF splitting: EC_GROUP_precompute_mult()
2914 will now compute a table of multiples of the generator that
2915 makes subsequent invocations of EC_POINTs_mul() or EC_POINT_mul()
2916 faster (notably in the case of a single point multiplication,
2917 scalar * generator).
2918 [Nils Larsch, Bodo Moeller]
2919
2920 *) IPv6 support for certificate extensions. The various extensions
2921 which use the IP:a.b.c.d can now take IPv6 addresses using the
2922 formats of RFC1884 2.2 . IPv6 addresses are now also displayed
2923 correctly.
2924 [Steve Henson]
2925
2926 *) Added an ENGINE that implements RSA by performing private key
2927 exponentiations with the GMP library. The conversions to and from
2928 GMP's mpz_t format aren't optimised nor are any montgomery forms
2929 cached, and on x86 it appears OpenSSL's own performance has caught up.
2930 However there are likely to be other architectures where GMP could
2931 provide a boost. This ENGINE is not built in by default, but it can be
2932 specified at Configure time and should be accompanied by the necessary
2933 linker additions, eg;
2934 ./config -DOPENSSL_USE_GMP -lgmp
2935 [Geoff Thorpe]
2936
2937 *) "openssl engine" will not display ENGINE/DSO load failure errors when
2938 testing availability of engines with "-t" - the old behaviour is
2939 produced by increasing the feature's verbosity with "-tt".
2940 [Geoff Thorpe]
2941
2942 *) ECDSA routines: under certain error conditions uninitialized BN objects
2943 could be freed. Solution: make sure initialization is performed early
2944 enough. (Reported and fix supplied by Nils Larsch <nla@trustcenter.de>
2945 via PR#459)
2946 [Lutz Jaenicke]
2947
2948 *) Key-generation can now be implemented in RSA_METHOD, DSA_METHOD
2949 and DH_METHOD (eg. by ENGINE implementations) to override the normal
2950 software implementations. For DSA and DH, parameter generation can
2951 also be overriden by providing the appropriate method callbacks.
2952 [Geoff Thorpe]
2953
2954 *) Change the "progress" mechanism used in key-generation and
2955 primality testing to functions that take a new BN_GENCB pointer in
2956 place of callback/argument pairs. The new API functions have "_ex"
2957 postfixes and the older functions are reimplemented as wrappers for
2958 the new ones. The OPENSSL_NO_DEPRECATED symbol can be used to hide
2959 declarations of the old functions to help (graceful) attempts to
2960 migrate to the new functions. Also, the new key-generation API
2961 functions operate on a caller-supplied key-structure and return
2962 success/failure rather than returning a key or NULL - this is to
2963 help make "keygen" another member function of RSA_METHOD etc.
2964
2965 Example for using the new callback interface:
2966
2967 int (*my_callback)(int a, int b, BN_GENCB *cb) = ...;
2968 void *my_arg = ...;
2969 BN_GENCB my_cb;
2970
2971 BN_GENCB_set(&my_cb, my_callback, my_arg);
2972
2973 return BN_is_prime_ex(some_bignum, BN_prime_checks, NULL, &cb);
2974 /* For the meaning of a, b in calls to my_callback(), see the
2975 * documentation of the function that calls the callback.
2976 * cb will point to my_cb; my_arg can be retrieved as cb->arg.
2977 * my_callback should return 1 if it wants BN_is_prime_ex()
2978 * to continue, or 0 to stop.
2979 */
2980
2981 [Geoff Thorpe]
2982
2983 *) Change the ZLIB compression method to be stateful, and make it
2984 available to TLS with the number defined in
2985 draft-ietf-tls-compression-04.txt.
2986 [Richard Levitte]
2987
2988 *) Add the ASN.1 structures and functions for CertificatePair, which
2989 is defined as follows (according to X.509_4thEditionDraftV6.pdf):
2990
2991 CertificatePair ::= SEQUENCE {
2992 forward [0] Certificate OPTIONAL,
2993 reverse [1] Certificate OPTIONAL,
2994 -- at least one of the pair shall be present -- }
2995
2996 Also implement the PEM functions to read and write certificate
2997 pairs, and defined the PEM tag as "CERTIFICATE PAIR".
2998
2999 This needed to be defined, mostly for the sake of the LDAP
3000 attribute crossCertificatePair, but may prove useful elsewhere as
3001 well.
3002 [Richard Levitte]
3003
3004 *) Make it possible to inhibit symlinking of shared libraries in
3005 Makefile.shared, for Cygwin's sake.
3006 [Richard Levitte]
3007
3008 *) Extend the BIGNUM API by creating a function
3009 void BN_set_negative(BIGNUM *a, int neg);
3010 and a macro that behave like
3011 int BN_is_negative(const BIGNUM *a);
3012
3013 to avoid the need to access 'a->neg' directly in applications.
3014 [Nils Larsch]
3015
3016 *) Implement fast modular reduction for pseudo-Mersenne primes
3017 used in NIST curves (crypto/bn/bn_nist.c, crypto/ec/ecp_nist.c).
3018 EC_GROUP_new_curve_GFp() will now automatically use this
3019 if applicable.
3020 [Nils Larsch <nla@trustcenter.de>]
3021
3022 *) Add new lock type (CRYPTO_LOCK_BN).
3023 [Bodo Moeller]
3024
3025 *) Change the ENGINE framework to automatically load engines
3026 dynamically from specific directories unless they could be
3027 found to already be built in or loaded. Move all the
3028 current engines except for the cryptodev one to a new
3029 directory engines/.
3030 The engines in engines/ are built as shared libraries if
3031 the "shared" options was given to ./Configure or ./config.
3032 Otherwise, they are inserted in libcrypto.a.
3033 /usr/local/ssl/engines is the default directory for dynamic
3034 engines, but that can be overriden at configure time through
3035 the usual use of --prefix and/or --openssldir, and at run
3036 time with the environment variable OPENSSL_ENGINES.
3037 [Geoff Thorpe and Richard Levitte]
3038
3039 *) Add Makefile.shared, a helper makefile to build shared
3040 libraries. Addapt Makefile.org.
3041 [Richard Levitte]
3042
3043 *) Add version info to Win32 DLLs.
3044 [Peter 'Luna' Runestig" <peter@runestig.com>]
3045
3046 *) Add new 'medium level' PKCS#12 API. Certificates and keys
3047 can be added using this API to created arbitrary PKCS#12
3048 files while avoiding the low level API.
3049
3050 New options to PKCS12_create(), key or cert can be NULL and
3051 will then be omitted from the output file. The encryption
3052 algorithm NIDs can be set to -1 for no encryption, the mac
3053 iteration count can be set to 0 to omit the mac.
3054
3055 Enhance pkcs12 utility by making the -nokeys and -nocerts
3056 options work when creating a PKCS#12 file. New option -nomac
3057 to omit the mac, NONE can be set for an encryption algorithm.
3058 New code is modified to use the enhanced PKCS12_create()
3059 instead of the low level API.
3060 [Steve Henson]
3061
3062 *) Extend ASN1 encoder to support indefinite length constructed
3063 encoding. This can output sequences tags and octet strings in
3064 this form. Modify pk7_asn1.c to support indefinite length
3065 encoding. This is experimental and needs additional code to
3066 be useful, such as an ASN1 bio and some enhanced streaming
3067 PKCS#7 code.
3068
3069 Extend template encode functionality so that tagging is passed
3070 down to the template encoder.
3071 [Steve Henson]
3072
3073 *) Let 'openssl req' fail if an argument to '-newkey' is not
3074 recognized instead of using RSA as a default.
3075 [Bodo Moeller]
3076
3077 *) Add support for ECC-based ciphersuites from draft-ietf-tls-ecc-01.txt.
3078 As these are not official, they are not included in "ALL";
3079 the "ECCdraft" ciphersuite group alias can be used to select them.
3080 [Vipul Gupta and Sumit Gupta (Sun Microsystems Laboratories)]
3081
3082 *) Add ECDH engine support.
3083 [Nils Gura and Douglas Stebila (Sun Microsystems Laboratories)]
3084
3085 *) Add ECDH in new directory crypto/ecdh/.
3086 [Douglas Stebila (Sun Microsystems Laboratories)]
3087
3088 *) Let BN_rand_range() abort with an error after 100 iterations
3089 without success (which indicates a broken PRNG).
3090 [Bodo Moeller]
3091
3092 *) Change BN_mod_sqrt() so that it verifies that the input value
3093 is really the square of the return value. (Previously,
3094 BN_mod_sqrt would show GIGO behaviour.)
3095 [Bodo Moeller]
3096
3097 *) Add named elliptic curves over binary fields from X9.62, SECG,
3098 and WAP/WTLS; add OIDs that were still missing.
3099
3100 [Sheueling Chang Shantz and Douglas Stebila
3101 (Sun Microsystems Laboratories)]
3102
3103 *) Extend the EC library for elliptic curves over binary fields
3104 (new files ec2_smpl.c, ec2_smpt.c, ec2_mult.c in crypto/ec/).
3105 New EC_METHOD:
3106
3107 EC_GF2m_simple_method
3108
3109 New API functions:
3110
3111 EC_GROUP_new_curve_GF2m
3112 EC_GROUP_set_curve_GF2m
3113 EC_GROUP_get_curve_GF2m
3114 EC_POINT_set_affine_coordinates_GF2m
3115 EC_POINT_get_affine_coordinates_GF2m
3116 EC_POINT_set_compressed_coordinates_GF2m
3117
3118 Point compression for binary fields is disabled by default for
3119 patent reasons (compile with OPENSSL_EC_BIN_PT_COMP defined to
3120 enable it).
3121
3122 As binary polynomials are represented as BIGNUMs, various members
3123 of the EC_GROUP and EC_POINT data structures can be shared
3124 between the implementations for prime fields and binary fields;
3125 the above ..._GF2m functions (except for EX_GROUP_new_curve_GF2m)
3126 are essentially identical to their ..._GFp counterparts.
3127 (For simplicity, the '..._GFp' prefix has been dropped from
3128 various internal method names.)
3129
3130 An internal 'field_div' method (similar to 'field_mul' and
3131 'field_sqr') has been added; this is used only for binary fields.
3132
3133 [Sheueling Chang Shantz and Douglas Stebila
3134 (Sun Microsystems Laboratories)]
3135
3136 *) Optionally dispatch EC_POINT_mul(), EC_POINT_precompute_mult()
3137 through methods ('mul', 'precompute_mult').
3138
3139 The generic implementations (now internally called 'ec_wNAF_mul'
3140 and 'ec_wNAF_precomputed_mult') remain the default if these
3141 methods are undefined.
3142
3143 [Sheueling Chang Shantz and Douglas Stebila
3144 (Sun Microsystems Laboratories)]
3145
3146 *) New function EC_GROUP_get_degree, which is defined through
3147 EC_METHOD. For curves over prime fields, this returns the bit
3148 length of the modulus.
3149
3150 [Sheueling Chang Shantz and Douglas Stebila
3151 (Sun Microsystems Laboratories)]
3152
3153 *) New functions EC_GROUP_dup, EC_POINT_dup.
3154 (These simply call ..._new and ..._copy).
3155
3156 [Sheueling Chang Shantz and Douglas Stebila
3157 (Sun Microsystems Laboratories)]
3158
3159 *) Add binary polynomial arithmetic software in crypto/bn/bn_gf2m.c.
3160 Polynomials are represented as BIGNUMs (where the sign bit is not
3161 used) in the following functions [macros]:
3162
3163 BN_GF2m_add
3164 BN_GF2m_sub [= BN_GF2m_add]
3165 BN_GF2m_mod [wrapper for BN_GF2m_mod_arr]
3166 BN_GF2m_mod_mul [wrapper for BN_GF2m_mod_mul_arr]
3167 BN_GF2m_mod_sqr [wrapper for BN_GF2m_mod_sqr_arr]
3168 BN_GF2m_mod_inv
3169 BN_GF2m_mod_exp [wrapper for BN_GF2m_mod_exp_arr]
3170 BN_GF2m_mod_sqrt [wrapper for BN_GF2m_mod_sqrt_arr]
3171 BN_GF2m_mod_solve_quad [wrapper for BN_GF2m_mod_solve_quad_arr]
3172 BN_GF2m_cmp [= BN_ucmp]
3173
3174 (Note that only the 'mod' functions are actually for fields GF(2^m).
3175 BN_GF2m_add() is misnomer, but this is for the sake of consistency.)
3176
3177 For some functions, an the irreducible polynomial defining a
3178 field can be given as an 'unsigned int[]' with strictly
3179 decreasing elements giving the indices of those bits that are set;
3180 i.e., p[] represents the polynomial
3181 f(t) = t^p[0] + t^p[1] + ... + t^p[k]
3182 where
3183 p[0] > p[1] > ... > p[k] = 0.
3184 This applies to the following functions:
3185
3186 BN_GF2m_mod_arr
3187 BN_GF2m_mod_mul_arr
3188 BN_GF2m_mod_sqr_arr
3189 BN_GF2m_mod_inv_arr [wrapper for BN_GF2m_mod_inv]
3190 BN_GF2m_mod_div_arr [wrapper for BN_GF2m_mod_div]
3191 BN_GF2m_mod_exp_arr
3192 BN_GF2m_mod_sqrt_arr
3193 BN_GF2m_mod_solve_quad_arr
3194 BN_GF2m_poly2arr
3195 BN_GF2m_arr2poly
3196
3197 Conversion can be performed by the following functions:
3198
3199 BN_GF2m_poly2arr
3200 BN_GF2m_arr2poly
3201
3202 bntest.c has additional tests for binary polynomial arithmetic.
3203
3204 Two implementations for BN_GF2m_mod_div() are available.
3205 The default algorithm simply uses BN_GF2m_mod_inv() and
3206 BN_GF2m_mod_mul(). The alternative algorithm is compiled in only
3207 if OPENSSL_SUN_GF2M_DIV is defined (patent pending; read the
3208 copyright notice in crypto/bn/bn_gf2m.c before enabling it).
3209
3210 [Sheueling Chang Shantz and Douglas Stebila
3211 (Sun Microsystems Laboratories)]
3212
3213 *) Add new error code 'ERR_R_DISABLED' that can be used when some
3214 functionality is disabled at compile-time.
3215 [Douglas Stebila <douglas.stebila@sun.com>]
3216
3217 *) Change default behaviour of 'openssl asn1parse' so that more
3218 information is visible when viewing, e.g., a certificate:
3219
3220 Modify asn1_parse2 (crypto/asn1/asn1_par.c) so that in non-'dump'
3221 mode the content of non-printable OCTET STRINGs is output in a
3222 style similar to INTEGERs, but with '[HEX DUMP]' prepended to
3223 avoid the appearance of a printable string.
3224 [Nils Larsch <nla@trustcenter.de>]
3225
3226 *) Add 'asn1_flag' and 'asn1_form' member to EC_GROUP with access
3227 functions
3228 EC_GROUP_set_asn1_flag()
3229 EC_GROUP_get_asn1_flag()
3230 EC_GROUP_set_point_conversion_form()
3231 EC_GROUP_get_point_conversion_form()
3232 These control ASN1 encoding details:
3233 - Curves (i.e., groups) are encoded explicitly unless asn1_flag
3234 has been set to OPENSSL_EC_NAMED_CURVE.
3235 - Points are encoded in uncompressed form by default; options for
3236 asn1_for are as for point2oct, namely
3237 POINT_CONVERSION_COMPRESSED
3238 POINT_CONVERSION_UNCOMPRESSED
3239 POINT_CONVERSION_HYBRID
3240
3241 Also add 'seed' and 'seed_len' members to EC_GROUP with access
3242 functions
3243 EC_GROUP_set_seed()
3244 EC_GROUP_get0_seed()
3245 EC_GROUP_get_seed_len()
3246 This is used only for ASN1 purposes (so far).
3247 [Nils Larsch <nla@trustcenter.de>]
3248
3249 *) Add 'field_type' member to EC_METHOD, which holds the NID
3250 of the appropriate field type OID. The new function
3251 EC_METHOD_get_field_type() returns this value.
3252 [Nils Larsch <nla@trustcenter.de>]
3253
3254 *) Add functions
3255 EC_POINT_point2bn()
3256 EC_POINT_bn2point()
3257 EC_POINT_point2hex()
3258 EC_POINT_hex2point()
3259 providing useful interfaces to EC_POINT_point2oct() and
3260 EC_POINT_oct2point().
3261 [Nils Larsch <nla@trustcenter.de>]
3262
3263 *) Change internals of the EC library so that the functions
3264 EC_GROUP_set_generator()
3265 EC_GROUP_get_generator()
3266 EC_GROUP_get_order()
3267 EC_GROUP_get_cofactor()
3268 are implemented directly in crypto/ec/ec_lib.c and not dispatched
3269 to methods, which would lead to unnecessary code duplication when
3270 adding different types of curves.
3271 [Nils Larsch <nla@trustcenter.de> with input by Bodo Moeller]
3272
3273 *) Implement compute_wNAF (crypto/ec/ec_mult.c) without BIGNUM
3274 arithmetic, and such that modified wNAFs are generated
3275 (which avoid length expansion in many cases).
3276 [Bodo Moeller]
3277
3278 *) Add a function EC_GROUP_check_discriminant() (defined via
3279 EC_METHOD) that verifies that the curve discriminant is non-zero.
3280
3281 Add a function EC_GROUP_check() that makes some sanity tests
3282 on a EC_GROUP, its generator and order. This includes
3283 EC_GROUP_check_discriminant().
3284 [Nils Larsch <nla@trustcenter.de>]
3285
3286 *) Add ECDSA in new directory crypto/ecdsa/.
3287
3288 Add applications 'openssl ecparam' and 'openssl ecdsa'
3289 (these are based on 'openssl dsaparam' and 'openssl dsa').
3290
3291 ECDSA support is also included in various other files across the
3292 library. Most notably,
3293 - 'openssl req' now has a '-newkey ecdsa:file' option;
3294 - EVP_PKCS82PKEY (crypto/evp/evp_pkey.c) now can handle ECDSA;
3295 - X509_PUBKEY_get (crypto/asn1/x_pubkey.c) and
3296 d2i_PublicKey (crypto/asn1/d2i_pu.c) have been modified to make
3297 them suitable for ECDSA where domain parameters must be
3298 extracted before the specific public key;
3299 - ECDSA engine support has been added.
3300 [Nils Larsch <nla@trustcenter.de>]
3301
3302 *) Include some named elliptic curves, and add OIDs from X9.62,
3303 SECG, and WAP/WTLS. Each curve can be obtained from the new
3304 function
3305 EC_GROUP_new_by_curve_name(),
3306 and the list of available named curves can be obtained with
3307 EC_get_builtin_curves().
3308 Also add a 'curve_name' member to EC_GROUP objects, which can be
3309 accessed via
3310 EC_GROUP_set_curve_name()
3311 EC_GROUP_get_curve_name()
3312 [Nils Larsch <larsch@trustcenter.de, Bodo Moeller]
3313
3314 *) Remove a few calls to bn_wexpand() in BN_sqr() (the one in there
3315 was actually never needed) and in BN_mul(). The removal in BN_mul()
3316 required a small change in bn_mul_part_recursive() and the addition
3317 of the functions bn_cmp_part_words(), bn_sub_part_words() and
3318 bn_add_part_words(), which do the same thing as bn_cmp_words(),
3319 bn_sub_words() and bn_add_words() except they take arrays with
3320 differing sizes.
3321 [Richard Levitte]
3322
3323 Changes between 0.9.7l and 0.9.7m [23 Feb 2007]
3324
3325 *) Cleanse PEM buffers before freeing them since they may contain
3326 sensitive data.
3327 [Benjamin Bennett <ben@psc.edu>]
3328
3329 *) Include "!eNULL" in SSL_DEFAULT_CIPHER_LIST to make sure that
3330 a ciphersuite string such as "DEFAULT:RSA" cannot enable
3331 authentication-only ciphersuites.
3332 [Bodo Moeller]
3333
3334 *) Since AES128 and AES256 share a single mask bit in the logic of
3335 ssl/ssl_ciph.c, the code for masking out disabled ciphers needs a
3336 kludge to work properly if AES128 is available and AES256 isn't.
3337 [Victor Duchovni]
3338
3339 *) Expand security boundary to match 1.1.1 module.
3340 [Steve Henson]
3341
3342 *) Remove redundant features: hash file source, editing of test vectors
3343 modify fipsld to use external fips_premain.c signature.
3344 [Steve Henson]
3345
3346 *) New perl script mkfipsscr.pl to create shell scripts or batch files to
3347 run algorithm test programs.
3348 [Steve Henson]
3349
3350 *) Make algorithm test programs more tolerant of whitespace.
3351 [Steve Henson]
3352
3353 *) Have SSL/TLS server implementation tolerate "mismatched" record
3354 protocol version while receiving ClientHello even if the
3355 ClientHello is fragmented. (The server can't insist on the
3356 particular protocol version it has chosen before the ServerHello
3357 message has informed the client about his choice.)
3358 [Bodo Moeller]
3359
3360 *) Load error codes if they are not already present instead of using a
3361 static variable. This allows them to be cleanly unloaded and reloaded.
3362 [Steve Henson]
3363
3364 Changes between 0.9.7k and 0.9.7l [28 Sep 2006]
3365
3366 *) Introduce limits to prevent malicious keys being able to
3367 cause a denial of service. (CVE-2006-2940)
3368 [Steve Henson, Bodo Moeller]
3369
3370 *) Fix ASN.1 parsing of certain invalid structures that can result
3371 in a denial of service. (CVE-2006-2937) [Steve Henson]
3372
3373 *) Fix buffer overflow in SSL_get_shared_ciphers() function.
3374 (CVE-2006-3738) [Tavis Ormandy and Will Drewry, Google Security Team]
3375
3376 *) Fix SSL client code which could crash if connecting to a
3377 malicious SSLv2 server. (CVE-2006-4343)
3378 [Tavis Ormandy and Will Drewry, Google Security Team]
3379
3380 *) Change ciphersuite string processing so that an explicit
3381 ciphersuite selects this one ciphersuite (so that "AES256-SHA"
3382 will no longer include "AES128-SHA"), and any other similar
3383 ciphersuite (same bitmap) from *other* protocol versions (so that
3384 "RC4-MD5" will still include both the SSL 2.0 ciphersuite and the
3385 SSL 3.0/TLS 1.0 ciphersuite). This is a backport combining
3386 changes from 0.9.8b and 0.9.8d.
3387 [Bodo Moeller]
3388
3389 Changes between 0.9.7j and 0.9.7k [05 Sep 2006]
3390
3391 *) Avoid PKCS #1 v1.5 signature attack discovered by Daniel Bleichenbacher
3392 (CVE-2006-4339) [Ben Laurie and Google Security Team]
3393
3394 *) Change the Unix randomness entropy gathering to use poll() when
3395 possible instead of select(), since the latter has some
3396 undesirable limitations.
3397 [Darryl Miles via Richard Levitte and Bodo Moeller]
3398
3399 *) Disable rogue ciphersuites:
3400
3401 - SSLv2 0x08 0x00 0x80 ("RC4-64-MD5")
3402 - SSLv3/TLSv1 0x00 0x61 ("EXP1024-RC2-CBC-MD5")
3403 - SSLv3/TLSv1 0x00 0x60 ("EXP1024-RC4-MD5")
3404
3405 The latter two were purportedly from
3406 draft-ietf-tls-56-bit-ciphersuites-0[01].txt, but do not really
3407 appear there.
3408
3409 Also deactive the remaining ciphersuites from
3410 draft-ietf-tls-56-bit-ciphersuites-01.txt. These are just as
3411 unofficial, and the ID has long expired.
3412 [Bodo Moeller]
3413
3414 *) Fix RSA blinding Heisenbug (problems sometimes occured on
3415 dual-core machines) and other potential thread-safety issues.
3416 [Bodo Moeller]
3417
3418 Changes between 0.9.7i and 0.9.7j [04 May 2006]
3419
3420 *) Adapt fipsld and the build system to link against the validated FIPS
3421 module in FIPS mode.
3422 [Steve Henson]
3423
3424 *) Fixes for VC++ 2005 build under Windows.
3425 [Steve Henson]
3426
3427 *) Add new Windows build target VC-32-GMAKE for VC++. This uses GNU make
3428 from a Windows bash shell such as MSYS. It is autodetected from the
3429 "config" script when run from a VC++ environment. Modify standard VC++
3430 build to use fipscanister.o from the GNU make build.
3431 [Steve Henson]
3432
3433 Changes between 0.9.7h and 0.9.7i [14 Oct 2005]
3434
3435 *) Wrapped the definition of EVP_MAX_MD_SIZE in a #ifdef OPENSSL_FIPS.
3436 The value now differs depending on if you build for FIPS or not.
3437 BEWARE! A program linked with a shared FIPSed libcrypto can't be
3438 safely run with a non-FIPSed libcrypto, as it may crash because of
3439 the difference induced by this change.
3440 [Andy Polyakov]
3441
3442 Changes between 0.9.7g and 0.9.7h [11 Oct 2005]
3443
3444 *) Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING
3445 (part of SSL_OP_ALL). This option used to disable the
3446 countermeasure against man-in-the-middle protocol-version
3447 rollback in the SSL 2.0 server implementation, which is a bad
3448 idea. (CVE-2005-2969)
3449
3450 [Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center
3451 for Information Security, National Institute of Advanced Industrial
3452 Science and Technology [AIST], Japan)]
3453
3454 *) Minimal support for X9.31 signatures and PSS padding modes. This is
3455 mainly for FIPS compliance and not fully integrated at this stage.
3456 [Steve Henson]
3457
3458 *) For DSA signing, unless DSA_FLAG_NO_EXP_CONSTTIME is set, perform
3459 the exponentiation using a fixed-length exponent. (Otherwise,
3460 the information leaked through timing could expose the secret key
3461 after many signatures; cf. Bleichenbacher's attack on DSA with
3462 biased k.)
3463 [Bodo Moeller]
3464
3465 *) Make a new fixed-window mod_exp implementation the default for
3466 RSA, DSA, and DH private-key operations so that the sequence of
3467 squares and multiplies and the memory access pattern are
3468 independent of the particular secret key. This will mitigate
3469 cache-timing and potential related attacks.
3470
3471 BN_mod_exp_mont_consttime() is the new exponentiation implementation,
3472 and this is automatically used by BN_mod_exp_mont() if the new flag
3473 BN_FLG_EXP_CONSTTIME is set for the exponent. RSA, DSA, and DH
3474 will use this BN flag for private exponents unless the flag
3475 RSA_FLAG_NO_EXP_CONSTTIME, DSA_FLAG_NO_EXP_CONSTTIME, or
3476 DH_FLAG_NO_EXP_CONSTTIME, respectively, is set.
3477
3478 [Matthew D Wood (Intel Corp), with some changes by Bodo Moeller]
3479
3480 *) Change the client implementation for SSLv23_method() and
3481 SSLv23_client_method() so that is uses the SSL 3.0/TLS 1.0
3482 Client Hello message format if the SSL_OP_NO_SSLv2 option is set.
3483 (Previously, the SSL 2.0 backwards compatible Client Hello
3484 message format would be used even with SSL_OP_NO_SSLv2.)
3485 [Bodo Moeller]
3486
3487 *) Add support for smime-type MIME parameter in S/MIME messages which some
3488 clients need.
3489 [Steve Henson]
3490
3491 *) New function BN_MONT_CTX_set_locked() to set montgomery parameters in
3492 a threadsafe manner. Modify rsa code to use new function and add calls
3493 to dsa and dh code (which had race conditions before).
3494 [Steve Henson]
3495
3496 *) Include the fixed error library code in the C error file definitions
3497 instead of fixing them up at runtime. This keeps the error code
3498 structures constant.
3499 [Steve Henson]
3500
3501 Changes between 0.9.7f and 0.9.7g [11 Apr 2005]
3502
3503 [NB: OpenSSL 0.9.7h and later 0.9.7 patch levels were released after
3504 OpenSSL 0.9.8.]
3505
3506 *) Fixes for newer kerberos headers. NB: the casts are needed because
3507 the 'length' field is signed on one version and unsigned on another
3508 with no (?) obvious way to tell the difference, without these VC++
3509 complains. Also the "definition" of FAR (blank) is no longer included
3510 nor is the error ENOMEM. KRB5_PRIVATE has to be set to 1 to pick up
3511 some needed definitions.
3512 [Steve Henson]
3513
3514 *) Undo Cygwin change.
3515 [Ulf Möller]
3516
3517 *) Added support for proxy certificates according to RFC 3820.
3518 Because they may be a security thread to unaware applications,
3519 they must be explicitely allowed in run-time. See
3520 docs/HOWTO/proxy_certificates.txt for further information.
3521 [Richard Levitte]
3522
3523 Changes between 0.9.7e and 0.9.7f [22 Mar 2005]
3524
3525 *) Use (SSL_RANDOM_VALUE - 4) bytes of pseudo random data when generating
3526 server and client random values. Previously
3527 (SSL_RANDOM_VALUE - sizeof(time_t)) would be used which would result in
3528 less random data when sizeof(time_t) > 4 (some 64 bit platforms).
3529
3530 This change has negligible security impact because:
3531
3532 1. Server and client random values still have 24 bytes of pseudo random
3533 data.
3534
3535 2. Server and client random values are sent in the clear in the initial
3536 handshake.
3537
3538 3. The master secret is derived using the premaster secret (48 bytes in
3539 size for static RSA ciphersuites) as well as client server and random
3540 values.
3541
3542 The OpenSSL team would like to thank the UK NISCC for bringing this issue
3543 to our attention.
3544
3545 [Stephen Henson, reported by UK NISCC]
3546
3547 *) Use Windows randomness collection on Cygwin.
3548 [Ulf Möller]
3549
3550 *) Fix hang in EGD/PRNGD query when communication socket is closed
3551 prematurely by EGD/PRNGD.
3552 [Darren Tucker <dtucker@zip.com.au> via Lutz Jänicke, resolves #1014]
3553
3554 *) Prompt for pass phrases when appropriate for PKCS12 input format.
3555 [Steve Henson]
3556
3557 *) Back-port of selected performance improvements from development
3558 branch, as well as improved support for PowerPC platforms.
3559 [Andy Polyakov]
3560
3561 *) Add lots of checks for memory allocation failure, error codes to indicate
3562 failure and freeing up memory if a failure occurs.
3563 [Nauticus Networks SSL Team <openssl@nauticusnet.com>, Steve Henson]
3564
3565 *) Add new -passin argument to dgst.
3566 [Steve Henson]
3567
3568 *) Perform some character comparisons of different types in X509_NAME_cmp:
3569 this is needed for some certificates that reencode DNs into UTF8Strings
3570 (in violation of RFC3280) and can't or wont issue name rollover
3571 certificates.
3572 [Steve Henson]
3573
3574 *) Make an explicit check during certificate validation to see that
3575 the CA setting in each certificate on the chain is correct. As a
3576 side effect always do the following basic checks on extensions,
3577 not just when there's an associated purpose to the check:
3578
3579 - if there is an unhandled critical extension (unless the user
3580 has chosen to ignore this fault)
3581 - if the path length has been exceeded (if one is set at all)
3582 - that certain extensions fit the associated purpose (if one has
3583 been given)
3584 [Richard Levitte]
3585
3586 Changes between 0.9.7d and 0.9.7e [25 Oct 2004]
3587
3588 *) Avoid a race condition when CRLs are checked in a multi threaded
3589 environment. This would happen due to the reordering of the revoked
3590 entries during signature checking and serial number lookup. Now the
3591 encoding is cached and the serial number sort performed under a lock.
3592 Add new STACK function sk_is_sorted().
3593 [Steve Henson]
3594
3595 *) Add Delta CRL to the extension code.
3596 [Steve Henson]
3597
3598 *) Various fixes to s3_pkt.c so alerts are sent properly.
3599 [David Holmes <d.holmes@f5.com>]
3600
3601 *) Reduce the chances of duplicate issuer name and serial numbers (in
3602 violation of RFC3280) using the OpenSSL certificate creation utilities.
3603 This is done by creating a random 64 bit value for the initial serial
3604 number when a serial number file is created or when a self signed
3605 certificate is created using 'openssl req -x509'. The initial serial
3606 number file is created using 'openssl x509 -next_serial' in CA.pl
3607 rather than being initialized to 1.
3608 [Steve Henson]
3609
3610 Changes between 0.9.7c and 0.9.7d [17 Mar 2004]
3611
3612 *) Fix null-pointer assignment in do_change_cipher_spec() revealed
3613 by using the Codenomicon TLS Test Tool (CVE-2004-0079)
3614 [Joe Orton, Steve Henson]
3615
3616 *) Fix flaw in SSL/TLS handshaking when using Kerberos ciphersuites
3617 (CVE-2004-0112)
3618 [Joe Orton, Steve Henson]
3619
3620 *) Make it possible to have multiple active certificates with the same
3621 subject in the CA index file. This is done only if the keyword
3622 'unique_subject' is set to 'no' in the main CA section (default
3623 if 'CA_default') of the configuration file. The value is saved
3624 with the database itself in a separate index attribute file,
3625 named like the index file with '.attr' appended to the name.
3626 [Richard Levitte]
3627
3628 *) X509 verify fixes. Disable broken certificate workarounds when
3629 X509_V_FLAGS_X509_STRICT is set. Check CRL issuer has cRLSign set if
3630 keyUsage extension present. Don't accept CRLs with unhandled critical
3631 extensions: since verify currently doesn't process CRL extensions this
3632 rejects a CRL with *any* critical extensions. Add new verify error codes
3633 for these cases.
3634 [Steve Henson]
3635
3636 *) When creating an OCSP nonce use an OCTET STRING inside the extnValue.
3637 A clarification of RFC2560 will require the use of OCTET STRINGs and
3638 some implementations cannot handle the current raw format. Since OpenSSL
3639 copies and compares OCSP nonces as opaque blobs without any attempt at
3640 parsing them this should not create any compatibility issues.
3641 [Steve Henson]
3642
3643 *) New md flag EVP_MD_CTX_FLAG_REUSE this allows md_data to be reused when
3644 calling EVP_MD_CTX_copy_ex() to avoid calling OPENSSL_malloc(). Without
3645 this HMAC (and other) operations are several times slower than OpenSSL
3646 < 0.9.7.
3647 [Steve Henson]
3648
3649 *) Print out GeneralizedTime and UTCTime in ASN1_STRING_print_ex().
3650 [Peter Sylvester <Peter.Sylvester@EdelWeb.fr>]
3651
3652 *) Use the correct content when signing type "other".
3653 [Steve Henson]
3654
3655 Changes between 0.9.7b and 0.9.7c [30 Sep 2003]
3656
3657 *) Fix various bugs revealed by running the NISCC test suite:
3658
3659 Stop out of bounds reads in the ASN1 code when presented with
3660 invalid tags (CVE-2003-0543 and CVE-2003-0544).
3661
3662 Free up ASN1_TYPE correctly if ANY type is invalid (CVE-2003-0545).
3663
3664 If verify callback ignores invalid public key errors don't try to check
3665 certificate signature with the NULL public key.
3666
3667 [Steve Henson]
3668
3669 *) New -ignore_err option in ocsp application to stop the server
3670 exiting on the first error in a request.
3671 [Steve Henson]
3672
3673 *) In ssl3_accept() (ssl/s3_srvr.c) only accept a client certificate
3674 if the server requested one: as stated in TLS 1.0 and SSL 3.0
3675 specifications.
3676 [Steve Henson]
3677
3678 *) In ssl3_get_client_hello() (ssl/s3_srvr.c), tolerate additional
3679 extra data after the compression methods not only for TLS 1.0
3680 but also for SSL 3.0 (as required by the specification).
3681 [Bodo Moeller; problem pointed out by Matthias Loepfe]
3682
3683 *) Change X509_certificate_type() to mark the key as exported/exportable
3684 when it's 512 *bits* long, not 512 bytes.
3685 [Richard Levitte]
3686
3687 *) Change AES_cbc_encrypt() so it outputs exact multiple of
3688 blocks during encryption.
3689 [Richard Levitte]
3690
3691 *) Various fixes to base64 BIO and non blocking I/O. On write
3692 flushes were not handled properly if the BIO retried. On read
3693 data was not being buffered properly and had various logic bugs.
3694 This also affects blocking I/O when the data being decoded is a
3695 certain size.
3696 [Steve Henson]
3697
3698 *) Various S/MIME bugfixes and compatibility changes:
3699 output correct application/pkcs7 MIME type if
3700 PKCS7_NOOLDMIMETYPE is set. Tolerate some broken signatures.
3701 Output CR+LF for EOL if PKCS7_CRLFEOL is set (this makes opening
3702 of files as .eml work). Correctly handle very long lines in MIME
3703 parser.
3704 [Steve Henson]
3705
3706 Changes between 0.9.7a and 0.9.7b [10 Apr 2003]
3707
3708 *) Countermeasure against the Klima-Pokorny-Rosa extension of
3709 Bleichbacher's attack on PKCS #1 v1.5 padding: treat
3710 a protocol version number mismatch like a decryption error
3711 in ssl3_get_client_key_exchange (ssl/s3_srvr.c).
3712 [Bodo Moeller]
3713
3714 *) Turn on RSA blinding by default in the default implementation
3715 to avoid a timing attack. Applications that don't want it can call
3716 RSA_blinding_off() or use the new flag RSA_FLAG_NO_BLINDING.
3717 They would be ill-advised to do so in most cases.
3718 [Ben Laurie, Steve Henson, Geoff Thorpe, Bodo Moeller]
3719
3720 *) Change RSA blinding code so that it works when the PRNG is not
3721 seeded (in this case, the secret RSA exponent is abused as
3722 an unpredictable seed -- if it is not unpredictable, there
3723 is no point in blinding anyway). Make RSA blinding thread-safe
3724 by remembering the creator's thread ID in rsa->blinding and
3725 having all other threads use local one-time blinding factors
3726 (this requires more computation than sharing rsa->blinding, but
3727 avoids excessive locking; and if an RSA object is not shared
3728 between threads, blinding will still be very fast).
3729 [Bodo Moeller]
3730
3731 *) Fixed a typo bug that would cause ENGINE_set_default() to set an
3732 ENGINE as defaults for all supported algorithms irrespective of
3733 the 'flags' parameter. 'flags' is now honoured, so applications
3734 should make sure they are passing it correctly.
3735 [Geoff Thorpe]
3736
3737 *) Target "mingw" now allows native Windows code to be generated in
3738 the Cygwin environment as well as with the MinGW compiler.
3739 [Ulf Moeller]
3740
3741 Changes between 0.9.7 and 0.9.7a [19 Feb 2003]
3742
3743 *) In ssl3_get_record (ssl/s3_pkt.c), minimize information leaked
3744 via timing by performing a MAC computation even if incorrrect
3745 block cipher padding has been found. This is a countermeasure
3746 against active attacks where the attacker has to distinguish
3747 between bad padding and a MAC verification error. (CVE-2003-0078)
3748
3749 [Bodo Moeller; problem pointed out by Brice Canvel (EPFL),
3750 Alain Hiltgen (UBS), Serge Vaudenay (EPFL), and
3751 Martin Vuagnoux (EPFL, Ilion)]
3752
3753 *) Make the no-err option work as intended. The intention with no-err
3754 is not to have the whole error stack handling routines removed from
3755 libcrypto, it's only intended to remove all the function name and
3756 reason texts, thereby removing some of the footprint that may not
3757 be interesting if those errors aren't displayed anyway.
3758
3759 NOTE: it's still possible for any application or module to have it's
3760 own set of error texts inserted. The routines are there, just not
3761 used by default when no-err is given.
3762 [Richard Levitte]
3763
3764 *) Add support for FreeBSD on IA64.
3765 [dirk.meyer@dinoex.sub.org via Richard Levitte, resolves #454]
3766
3767 *) Adjust DES_cbc_cksum() so it returns the same value as the MIT
3768 Kerberos function mit_des_cbc_cksum(). Before this change,
3769 the value returned by DES_cbc_cksum() was like the one from
3770 mit_des_cbc_cksum(), except the bytes were swapped.
3771 [Kevin Greaney <Kevin.Greaney@hp.com> and Richard Levitte]
3772
3773 *) Allow an application to disable the automatic SSL chain building.
3774 Before this a rather primitive chain build was always performed in
3775 ssl3_output_cert_chain(): an application had no way to send the
3776 correct chain if the automatic operation produced an incorrect result.
3777
3778 Now the chain builder is disabled if either:
3779
3780 1. Extra certificates are added via SSL_CTX_add_extra_chain_cert().
3781
3782 2. The mode flag SSL_MODE_NO_AUTO_CHAIN is set.
3783
3784 The reasoning behind this is that an application would not want the
3785 auto chain building to take place if extra chain certificates are
3786 present and it might also want a means of sending no additional
3787 certificates (for example the chain has two certificates and the
3788 root is omitted).
3789 [Steve Henson]
3790
3791 *) Add the possibility to build without the ENGINE framework.
3792 [Steven Reddie <smr@essemer.com.au> via Richard Levitte]
3793
3794 *) Under Win32 gmtime() can return NULL: check return value in
3795 OPENSSL_gmtime(). Add error code for case where gmtime() fails.
3796 [Steve Henson]
3797
3798 *) DSA routines: under certain error conditions uninitialized BN objects
3799 could be freed. Solution: make sure initialization is performed early
3800 enough. (Reported and fix supplied by Ivan D Nestlerode <nestler@MIT.EDU>,
3801 Nils Larsch <nla@trustcenter.de> via PR#459)
3802 [Lutz Jaenicke]
3803
3804 *) Another fix for SSLv2 session ID handling: the session ID was incorrectly
3805 checked on reconnect on the client side, therefore session resumption
3806 could still fail with a "ssl session id is different" error. This
3807 behaviour is masked when SSL_OP_ALL is used due to
3808 SSL_OP_MICROSOFT_SESS_ID_BUG being set.
3809 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as
3810 followup to PR #377.
3811 [Lutz Jaenicke]
3812
3813 *) IA-32 assembler support enhancements: unified ELF targets, support
3814 for SCO/Caldera platforms, fix for Cygwin shared build.
3815 [Andy Polyakov]
3816
3817 *) Add support for FreeBSD on sparc64. As a consequence, support for
3818 FreeBSD on non-x86 processors is separate from x86 processors on
3819 the config script, much like the NetBSD support.
3820 [Richard Levitte & Kris Kennaway <kris@obsecurity.org>]
3821
3822 Changes between 0.9.6h and 0.9.7 [31 Dec 2002]
3823
3824 [NB: OpenSSL 0.9.6i and later 0.9.6 patch levels were released after
3825 OpenSSL 0.9.7.]
3826
3827 *) Fix session ID handling in SSLv2 client code: the SERVER FINISHED
3828 code (06) was taken as the first octet of the session ID and the last
3829 octet was ignored consequently. As a result SSLv2 client side session
3830 caching could not have worked due to the session ID mismatch between
3831 client and server.
3832 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as
3833 PR #377.
3834 [Lutz Jaenicke]
3835
3836 *) Change the declaration of needed Kerberos libraries to use EX_LIBS
3837 instead of the special (and badly supported) LIBKRB5. LIBKRB5 is
3838 removed entirely.
3839 [Richard Levitte]
3840
3841 *) The hw_ncipher.c engine requires dynamic locks. Unfortunately, it
3842 seems that in spite of existing for more than a year, many application
3843 author have done nothing to provide the necessary callbacks, which
3844 means that this particular engine will not work properly anywhere.
3845 This is a very unfortunate situation which forces us, in the name
3846 of usability, to give the hw_ncipher.c a static lock, which is part
3847 of libcrypto.
3848 NOTE: This is for the 0.9.7 series ONLY. This hack will never
3849 appear in 0.9.8 or later. We EXPECT application authors to have
3850 dealt properly with this when 0.9.8 is released (unless we actually
3851 make such changes in the libcrypto locking code that changes will
3852 have to be made anyway).
3853 [Richard Levitte]
3854
3855 *) In asn1_d2i_read_bio() repeatedly call BIO_read() until all content
3856 octets have been read, EOF or an error occurs. Without this change
3857 some truncated ASN1 structures will not produce an error.
3858 [Steve Henson]
3859
3860 *) Disable Heimdal support, since it hasn't been fully implemented.
3861 Still give the possibility to force the use of Heimdal, but with
3862 warnings and a request that patches get sent to openssl-dev.
3863 [Richard Levitte]
3864
3865 *) Add the VC-CE target, introduce the WINCE sysname, and add
3866 INSTALL.WCE and appropriate conditionals to make it build.
3867 [Steven Reddie <smr@essemer.com.au> via Richard Levitte]
3868
3869 *) Change the DLL names for Cygwin to cygcrypto-x.y.z.dll and
3870 cygssl-x.y.z.dll, where x, y and z are the major, minor and
3871 edit numbers of the version.
3872 [Corinna Vinschen <vinschen@redhat.com> and Richard Levitte]
3873
3874 *) Introduce safe string copy and catenation functions
3875 (BUF_strlcpy() and BUF_strlcat()).
3876 [Ben Laurie (CHATS) and Richard Levitte]
3877
3878 *) Avoid using fixed-size buffers for one-line DNs.
3879 [Ben Laurie (CHATS)]
3880
3881 *) Add BUF_MEM_grow_clean() to avoid information leakage when
3882 resizing buffers containing secrets, and use where appropriate.
3883 [Ben Laurie (CHATS)]
3884
3885 *) Avoid using fixed size buffers for configuration file location.
3886 [Ben Laurie (CHATS)]
3887
3888 *) Avoid filename truncation for various CA files.
3889 [Ben Laurie (CHATS)]
3890
3891 *) Use sizeof in preference to magic numbers.
3892 [Ben Laurie (CHATS)]
3893
3894 *) Avoid filename truncation in cert requests.
3895 [Ben Laurie (CHATS)]
3896
3897 *) Add assertions to check for (supposedly impossible) buffer
3898 overflows.
3899 [Ben Laurie (CHATS)]
3900
3901 *) Don't cache truncated DNS entries in the local cache (this could
3902 potentially lead to a spoofing attack).
3903 [Ben Laurie (CHATS)]
3904
3905 *) Fix various buffers to be large enough for hex/decimal
3906 representations in a platform independent manner.
3907 [Ben Laurie (CHATS)]
3908
3909 *) Add CRYPTO_realloc_clean() to avoid information leakage when
3910 resizing buffers containing secrets, and use where appropriate.
3911 [Ben Laurie (CHATS)]
3912
3913 *) Add BIO_indent() to avoid much slightly worrying code to do
3914 indents.
3915 [Ben Laurie (CHATS)]
3916
3917 *) Convert sprintf()/BIO_puts() to BIO_printf().
3918 [Ben Laurie (CHATS)]
3919
3920 *) buffer_gets() could terminate with the buffer only half
3921 full. Fixed.
3922 [Ben Laurie (CHATS)]
3923
3924 *) Add assertions to prevent user-supplied crypto functions from
3925 overflowing internal buffers by having large block sizes, etc.
3926 [Ben Laurie (CHATS)]
3927
3928 *) New OPENSSL_assert() macro (similar to assert(), but enabled
3929 unconditionally).
3930 [Ben Laurie (CHATS)]
3931
3932 *) Eliminate unused copy of key in RC4.
3933 [Ben Laurie (CHATS)]
3934
3935 *) Eliminate unused and incorrectly sized buffers for IV in pem.h.
3936 [Ben Laurie (CHATS)]
3937
3938 *) Fix off-by-one error in EGD path.
3939 [Ben Laurie (CHATS)]
3940
3941 *) If RANDFILE path is too long, ignore instead of truncating.
3942 [Ben Laurie (CHATS)]
3943
3944 *) Eliminate unused and incorrectly sized X.509 structure
3945 CBCParameter.
3946 [Ben Laurie (CHATS)]
3947
3948 *) Eliminate unused and dangerous function knumber().
3949 [Ben Laurie (CHATS)]
3950
3951 *) Eliminate unused and dangerous structure, KSSL_ERR.
3952 [Ben Laurie (CHATS)]
3953
3954 *) Protect against overlong session ID context length in an encoded
3955 session object. Since these are local, this does not appear to be
3956 exploitable.
3957 [Ben Laurie (CHATS)]
3958
3959 *) Change from security patch (see 0.9.6e below) that did not affect
3960 the 0.9.6 release series:
3961
3962 Remote buffer overflow in SSL3 protocol - an attacker could
3963 supply an oversized master key in Kerberos-enabled versions.
3964 (CVE-2002-0657)
3965 [Ben Laurie (CHATS)]
3966
3967 *) Change the SSL kerb5 codes to match RFC 2712.
3968 [Richard Levitte]
3969
3970 *) Make -nameopt work fully for req and add -reqopt switch.
3971 [Michael Bell <michael.bell@rz.hu-berlin.de>, Steve Henson]
3972
3973 *) The "block size" for block ciphers in CFB and OFB mode should be 1.
3974 [Steve Henson, reported by Yngve Nysaeter Pettersen <yngve@opera.com>]
3975
3976 *) Make sure tests can be performed even if the corresponding algorithms
3977 have been removed entirely. This was also the last step to make
3978 OpenSSL compilable with DJGPP under all reasonable conditions.
3979 [Richard Levitte, Doug Kaufman <dkaufman@rahul.net>]
3980
3981 *) Add cipher selection rules COMPLEMENTOFALL and COMPLEMENTOFDEFAULT
3982 to allow version independent disabling of normally unselected ciphers,
3983 which may be activated as a side-effect of selecting a single cipher.
3984
3985 (E.g., cipher list string "RSA" enables ciphersuites that are left
3986 out of "ALL" because they do not provide symmetric encryption.
3987 "RSA:!COMPLEMEMENTOFALL" avoids these unsafe ciphersuites.)
3988 [Lutz Jaenicke, Bodo Moeller]
3989
3990 *) Add appropriate support for separate platform-dependent build
3991 directories. The recommended way to make a platform-dependent
3992 build directory is the following (tested on Linux), maybe with
3993 some local tweaks:
3994
3995 # Place yourself outside of the OpenSSL source tree. In
3996 # this example, the environment variable OPENSSL_SOURCE
3997 # is assumed to contain the absolute OpenSSL source directory.
3998 mkdir -p objtree/"`uname -s`-`uname -r`-`uname -m`"
3999 cd objtree/"`uname -s`-`uname -r`-`uname -m`"
4000 (cd $OPENSSL_SOURCE; find . -type f) | while read F; do
4001 mkdir -p `dirname $F`
4002 ln -s $OPENSSL_SOURCE/$F $F
4003 done
4004
4005 To be absolutely sure not to disturb the source tree, a "make clean"
4006 is a good thing. If it isn't successfull, don't worry about it,
4007 it probably means the source directory is very clean.
4008 [Richard Levitte]
4009
4010 *) Make sure any ENGINE control commands make local copies of string
4011 pointers passed to them whenever necessary. Otherwise it is possible
4012 the caller may have overwritten (or deallocated) the original string
4013 data when a later ENGINE operation tries to use the stored values.
4014 [Götz Babin-Ebell <babinebell@trustcenter.de>]
4015
4016 *) Improve diagnostics in file reading and command-line digests.
4017 [Ben Laurie aided and abetted by Solar Designer <solar@openwall.com>]
4018
4019 *) Add AES modes CFB and OFB to the object database. Correct an
4020 error in AES-CFB decryption.
4021 [Richard Levitte]
4022
4023 *) Remove most calls to EVP_CIPHER_CTX_cleanup() in evp_enc.c, this
4024 allows existing EVP_CIPHER_CTX structures to be reused after
4025 calling EVP_*Final(). This behaviour is used by encryption
4026 BIOs and some applications. This has the side effect that
4027 applications must explicitly clean up cipher contexts with
4028 EVP_CIPHER_CTX_cleanup() or they will leak memory.
4029 [Steve Henson]
4030
4031 *) Check the values of dna and dnb in bn_mul_recursive before calling
4032 bn_mul_comba (a non zero value means the a or b arrays do not contain
4033 n2 elements) and fallback to bn_mul_normal if either is not zero.
4034 [Steve Henson]
4035
4036 *) Fix escaping of non-ASCII characters when using the -subj option
4037 of the "openssl req" command line tool. (Robert Joop <joop@fokus.gmd.de>)
4038 [Lutz Jaenicke]
4039
4040 *) Make object definitions compliant to LDAP (RFC2256): SN is the short
4041 form for "surname", serialNumber has no short form.
4042 Use "mail" as the short name for "rfc822Mailbox" according to RFC2798;
4043 therefore remove "mail" short name for "internet 7".
4044 The OID for unique identifiers in X509 certificates is
4045 x500UniqueIdentifier, not uniqueIdentifier.
4046 Some more OID additions. (Michael Bell <michael.bell@rz.hu-berlin.de>)
4047 [Lutz Jaenicke]
4048
4049 *) Add an "init" command to the ENGINE config module and auto initialize
4050 ENGINEs. Without any "init" command the ENGINE will be initialized
4051 after all ctrl commands have been executed on it. If init=1 the
4052 ENGINE is initailized at that point (ctrls before that point are run
4053 on the uninitialized ENGINE and after on the initialized one). If
4054 init=0 then the ENGINE will not be iniatialized at all.
4055 [Steve Henson]
4056
4057 *) Fix the 'app_verify_callback' interface so that the user-defined
4058 argument is actually passed to the callback: In the
4059 SSL_CTX_set_cert_verify_callback() prototype, the callback
4060 declaration has been changed from
4061 int (*cb)()
4062 into
4063 int (*cb)(X509_STORE_CTX *,void *);
4064 in ssl_verify_cert_chain (ssl/ssl_cert.c), the call
4065 i=s->ctx->app_verify_callback(&ctx)
4066 has been changed into
4067 i=s->ctx->app_verify_callback(&ctx, s->ctx->app_verify_arg).
4068
4069 To update applications using SSL_CTX_set_cert_verify_callback(),
4070 a dummy argument can be added to their callback functions.
4071 [D. K. Smetters <smetters@parc.xerox.com>]
4072
4073 *) Added the '4758cca' ENGINE to support IBM 4758 cards.
4074 [Maurice Gittens <maurice@gittens.nl>, touchups by Geoff Thorpe]
4075
4076 *) Add and OPENSSL_LOAD_CONF define which will cause
4077 OpenSSL_add_all_algorithms() to load the openssl.cnf config file.
4078 This allows older applications to transparently support certain
4079 OpenSSL features: such as crypto acceleration and dynamic ENGINE loading.
4080 Two new functions OPENSSL_add_all_algorithms_noconf() which will never
4081 load the config file and OPENSSL_add_all_algorithms_conf() which will
4082 always load it have also been added.
4083 [Steve Henson]
4084
4085 *) Add the OFB, CFB and CTR (all with 128 bit feedback) to AES.
4086 Adjust NIDs and EVP layer.
4087 [Stephen Sprunk <stephen@sprunk.org> and Richard Levitte]
4088
4089 *) Config modules support in openssl utility.
4090
4091 Most commands now load modules from the config file,
4092 though in a few (such as version) this isn't done
4093 because it couldn't be used for anything.
4094
4095 In the case of ca and req the config file used is
4096 the same as the utility itself: that is the -config
4097 command line option can be used to specify an
4098 alternative file.
4099 [Steve Henson]
4100
4101 *) Move default behaviour from OPENSSL_config(). If appname is NULL
4102 use "openssl_conf" if filename is NULL use default openssl config file.
4103 [Steve Henson]
4104
4105 *) Add an argument to OPENSSL_config() to allow the use of an alternative
4106 config section name. Add a new flag to tolerate a missing config file
4107 and move code to CONF_modules_load_file().
4108 [Steve Henson]
4109
4110 *) Support for crypto accelerator cards from Accelerated Encryption
4111 Processing, www.aep.ie. (Use engine 'aep')
4112 The support was copied from 0.9.6c [engine] and adapted/corrected
4113 to work with the new engine framework.
4114 [AEP Inc. and Richard Levitte]
4115
4116 *) Support for SureWare crypto accelerator cards from Baltimore
4117 Technologies. (Use engine 'sureware')
4118 The support was copied from 0.9.6c [engine] and adapted
4119 to work with the new engine framework.
4120 [Richard Levitte]
4121
4122 *) Have the CHIL engine fork-safe (as defined by nCipher) and actually
4123 make the newer ENGINE framework commands for the CHIL engine work.
4124 [Toomas Kiisk <vix@cyber.ee> and Richard Levitte]
4125
4126 *) Make it possible to produce shared libraries on ReliantUNIX.
4127 [Robert Dahlem <Robert.Dahlem@ffm2.siemens.de> via Richard Levitte]
4128
4129 *) Add the configuration target debug-linux-ppro.
4130 Make 'openssl rsa' use the general key loading routines
4131 implemented in apps.c, and make those routines able to
4132 handle the key format FORMAT_NETSCAPE and the variant
4133 FORMAT_IISSGC.
4134 [Toomas Kiisk <vix@cyber.ee> via Richard Levitte]
4135
4136 *) Fix a crashbug and a logic bug in hwcrhk_load_pubkey().
4137 [Toomas Kiisk <vix@cyber.ee> via Richard Levitte]
4138
4139 *) Add -keyform to rsautl, and document -engine.
4140 [Richard Levitte, inspired by Toomas Kiisk <vix@cyber.ee>]
4141
4142 *) Change BIO_new_file (crypto/bio/bss_file.c) to use new
4143 BIO_R_NO_SUCH_FILE error code rather than the generic
4144 ERR_R_SYS_LIB error code if fopen() fails with ENOENT.
4145 [Ben Laurie]
4146
4147 *) Add new functions
4148 ERR_peek_last_error
4149 ERR_peek_last_error_line
4150 ERR_peek_last_error_line_data.
4151 These are similar to
4152 ERR_peek_error
4153 ERR_peek_error_line
4154 ERR_peek_error_line_data,
4155 but report on the latest error recorded rather than the first one
4156 still in the error queue.
4157 [Ben Laurie, Bodo Moeller]
4158
4159 *) default_algorithms option in ENGINE config module. This allows things
4160 like:
4161 default_algorithms = ALL
4162 default_algorithms = RSA, DSA, RAND, CIPHERS, DIGESTS
4163 [Steve Henson]
4164
4165 *) Prelminary ENGINE config module.
4166 [Steve Henson]
4167
4168 *) New experimental application configuration code.
4169 [Steve Henson]
4170
4171 *) Change the AES code to follow the same name structure as all other
4172 symmetric ciphers, and behave the same way. Move everything to
4173 the directory crypto/aes, thereby obsoleting crypto/rijndael.
4174 [Stephen Sprunk <stephen@sprunk.org> and Richard Levitte]
4175
4176 *) SECURITY: remove unsafe setjmp/signal interaction from ui_openssl.c.
4177 [Ben Laurie and Theo de Raadt]
4178
4179 *) Add option to output public keys in req command.
4180 [Massimiliano Pala madwolf@openca.org]
4181
4182 *) Use wNAFs in EC_POINTs_mul() for improved efficiency
4183 (up to about 10% better than before for P-192 and P-224).
4184 [Bodo Moeller]
4185
4186 *) New functions/macros
4187
4188 SSL_CTX_set_msg_callback(ctx, cb)
4189 SSL_CTX_set_msg_callback_arg(ctx, arg)
4190 SSL_set_msg_callback(ssl, cb)
4191 SSL_set_msg_callback_arg(ssl, arg)
4192
4193 to request calling a callback function
4194
4195 void cb(int write_p, int version, int content_type,
4196 const void *buf, size_t len, SSL *ssl, void *arg)
4197
4198 whenever a protocol message has been completely received
4199 (write_p == 0) or sent (write_p == 1). Here 'version' is the
4200 protocol version according to which the SSL library interprets
4201 the current protocol message (SSL2_VERSION, SSL3_VERSION, or
4202 TLS1_VERSION). 'content_type' is 0 in the case of SSL 2.0, or
4203 the content type as defined in the SSL 3.0/TLS 1.0 protocol
4204 specification (change_cipher_spec(20), alert(21), handshake(22)).
4205 'buf' and 'len' point to the actual message, 'ssl' to the
4206 SSL object, and 'arg' is the application-defined value set by
4207 SSL[_CTX]_set_msg_callback_arg().
4208
4209 'openssl s_client' and 'openssl s_server' have new '-msg' options
4210 to enable a callback that displays all protocol messages.
4211 [Bodo Moeller]
4212
4213 *) Change the shared library support so shared libraries are built as
4214 soon as the corresponding static library is finished, and thereby get
4215 openssl and the test programs linked against the shared library.
4216 This still only happens when the keyword "shard" has been given to
4217 the configuration scripts.
4218
4219 NOTE: shared library support is still an experimental thing, and
4220 backward binary compatibility is still not guaranteed.
4221 ["Maciej W. Rozycki" <macro@ds2.pg.gda.pl> and Richard Levitte]
4222
4223 *) Add support for Subject Information Access extension.
4224 [Peter Sylvester <Peter.Sylvester@EdelWeb.fr>]
4225
4226 *) Make BUF_MEM_grow() behaviour more consistent: Initialise to zero
4227 additional bytes when new memory had to be allocated, not just
4228 when reusing an existing buffer.
4229 [Bodo Moeller]
4230
4231 *) New command line and configuration option 'utf8' for the req command.
4232 This allows field values to be specified as UTF8 strings.
4233 [Steve Henson]
4234
4235 *) Add -multi and -mr options to "openssl speed" - giving multiple parallel
4236 runs for the former and machine-readable output for the latter.
4237 [Ben Laurie]
4238
4239 *) Add '-noemailDN' option to 'openssl ca'. This prevents inclusion
4240 of the e-mail address in the DN (i.e., it will go into a certificate
4241 extension only). The new configuration file option 'email_in_dn = no'
4242 has the same effect.
4243 [Massimiliano Pala madwolf@openca.org]
4244
4245 *) Change all functions with names starting with des_ to be starting
4246 with DES_ instead. Add wrappers that are compatible with libdes,
4247 but are named _ossl_old_des_*. Finally, add macros that map the
4248 des_* symbols to the corresponding _ossl_old_des_* if libdes
4249 compatibility is desired. If OpenSSL 0.9.6c compatibility is
4250 desired, the des_* symbols will be mapped to DES_*, with one
4251 exception.
4252
4253 Since we provide two compatibility mappings, the user needs to
4254 define the macro OPENSSL_DES_LIBDES_COMPATIBILITY if libdes
4255 compatibility is desired. The default (i.e., when that macro
4256 isn't defined) is OpenSSL 0.9.6c compatibility.
4257
4258 There are also macros that enable and disable the support of old
4259 des functions altogether. Those are OPENSSL_ENABLE_OLD_DES_SUPPORT
4260 and OPENSSL_DISABLE_OLD_DES_SUPPORT. If none or both of those
4261 are defined, the default will apply: to support the old des routines.
4262
4263 In either case, one must include openssl/des.h to get the correct
4264 definitions. Do not try to just include openssl/des_old.h, that
4265 won't work.
4266
4267 NOTE: This is a major break of an old API into a new one. Software
4268 authors are encouraged to switch to the DES_ style functions. Some
4269 time in the future, des_old.h and the libdes compatibility functions
4270 will be disable (i.e. OPENSSL_DISABLE_OLD_DES_SUPPORT will be the
4271 default), and then completely removed.
4272 [Richard Levitte]
4273
4274 *) Test for certificates which contain unsupported critical extensions.
4275 If such a certificate is found during a verify operation it is
4276 rejected by default: this behaviour can be overridden by either
4277 handling the new error X509_V_ERR_UNHANDLED_CRITICAL_EXTENSION or
4278 by setting the verify flag X509_V_FLAG_IGNORE_CRITICAL. A new function
4279 X509_supported_extension() has also been added which returns 1 if a
4280 particular extension is supported.
4281 [Steve Henson]
4282
4283 *) Modify the behaviour of EVP cipher functions in similar way to digests
4284 to retain compatibility with existing code.
4285 [Steve Henson]
4286
4287 *) Modify the behaviour of EVP_DigestInit() and EVP_DigestFinal() to retain
4288 compatibility with existing code. In particular the 'ctx' parameter does
4289 not have to be to be initialized before the call to EVP_DigestInit() and
4290 it is tidied up after a call to EVP_DigestFinal(). New function
4291 EVP_DigestFinal_ex() which does not tidy up the ctx. Similarly function
4292 EVP_MD_CTX_copy() changed to not require the destination to be
4293 initialized valid and new function EVP_MD_CTX_copy_ex() added which
4294 requires the destination to be valid.
4295
4296 Modify all the OpenSSL digest calls to use EVP_DigestInit_ex(),
4297 EVP_DigestFinal_ex() and EVP_MD_CTX_copy_ex().
4298 [Steve Henson]
4299
4300 *) Change ssl3_get_message (ssl/s3_both.c) and the functions using it
4301 so that complete 'Handshake' protocol structures are kept in memory
4302 instead of overwriting 'msg_type' and 'length' with 'body' data.
4303 [Bodo Moeller]
4304
4305 *) Add an implementation of SSL_add_dir_cert_subjects_to_stack for Win32.
4306 [Massimo Santin via Richard Levitte]
4307
4308 *) Major restructuring to the underlying ENGINE code. This includes
4309 reduction of linker bloat, separation of pure "ENGINE" manipulation
4310 (initialisation, etc) from functionality dealing with implementations
4311 of specific crypto iterfaces. This change also introduces integrated
4312 support for symmetric ciphers and digest implementations - so ENGINEs
4313 can now accelerate these by providing EVP_CIPHER and EVP_MD
4314 implementations of their own. This is detailed in crypto/engine/README
4315 as it couldn't be adequately described here. However, there are a few
4316 API changes worth noting - some RSA, DSA, DH, and RAND functions that
4317 were changed in the original introduction of ENGINE code have now
4318 reverted back - the hooking from this code to ENGINE is now a good
4319 deal more passive and at run-time, operations deal directly with
4320 RSA_METHODs, DSA_METHODs (etc) as they did before, rather than
4321 dereferencing through an ENGINE pointer any more. Also, the ENGINE
4322 functions dealing with BN_MOD_EXP[_CRT] handlers have been removed -
4323 they were not being used by the framework as there is no concept of a
4324 BIGNUM_METHOD and they could not be generalised to the new
4325 'ENGINE_TABLE' mechanism that underlies the new code. Similarly,
4326 ENGINE_cpy() has been removed as it cannot be consistently defined in
4327 the new code.
4328 [Geoff Thorpe]
4329
4330 *) Change ASN1_GENERALIZEDTIME_check() to allow fractional seconds.
4331 [Steve Henson]
4332
4333 *) Change mkdef.pl to sort symbols that get the same entry number,
4334 and make sure the automatically generated functions ERR_load_*
4335 become part of libeay.num as well.
4336 [Richard Levitte]
4337
4338 *) New function SSL_renegotiate_pending(). This returns true once
4339 renegotiation has been requested (either SSL_renegotiate() call
4340 or HelloRequest/ClientHello receveived from the peer) and becomes
4341 false once a handshake has been completed.
4342 (For servers, SSL_renegotiate() followed by SSL_do_handshake()
4343 sends a HelloRequest, but does not ensure that a handshake takes
4344 place. SSL_renegotiate_pending() is useful for checking if the
4345 client has followed the request.)
4346 [Bodo Moeller]
4347
4348 *) New SSL option SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION.
4349 By default, clients may request session resumption even during
4350 renegotiation (if session ID contexts permit); with this option,
4351 session resumption is possible only in the first handshake.
4352
4353 SSL_OP_ALL is now 0x00000FFFL instead of 0x000FFFFFL. This makes
4354 more bits available for options that should not be part of
4355 SSL_OP_ALL (such as SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION).
4356 [Bodo Moeller]
4357
4358 *) Add some demos for certificate and certificate request creation.
4359 [Steve Henson]
4360
4361 *) Make maximum certificate chain size accepted from the peer application
4362 settable (SSL*_get/set_max_cert_list()), as proposed by
4363 "Douglas E. Engert" <deengert@anl.gov>.
4364 [Lutz Jaenicke]
4365
4366 *) Add support for shared libraries for Unixware-7
4367 (Boyd Lynn Gerber <gerberb@zenez.com>).
4368 [Lutz Jaenicke]
4369
4370 *) Add a "destroy" handler to ENGINEs that allows structural cleanup to
4371 be done prior to destruction. Use this to unload error strings from
4372 ENGINEs that load their own error strings. NB: This adds two new API
4373 functions to "get" and "set" this destroy handler in an ENGINE.
4374 [Geoff Thorpe]
4375
4376 *) Alter all existing ENGINE implementations (except "openssl" and
4377 "openbsd") to dynamically instantiate their own error strings. This
4378 makes them more flexible to be built both as statically-linked ENGINEs
4379 and self-contained shared-libraries loadable via the "dynamic" ENGINE.
4380 Also, add stub code to each that makes building them as self-contained
4381 shared-libraries easier (see README.ENGINE).
4382 [Geoff Thorpe]
4383
4384 *) Add a "dynamic" ENGINE that provides a mechanism for binding ENGINE
4385 implementations into applications that are completely implemented in
4386 self-contained shared-libraries. The "dynamic" ENGINE exposes control
4387 commands that can be used to configure what shared-library to load and
4388 to control aspects of the way it is handled. Also, made an update to
4389 the README.ENGINE file that brings its information up-to-date and
4390 provides some information and instructions on the "dynamic" ENGINE
4391 (ie. how to use it, how to build "dynamic"-loadable ENGINEs, etc).
4392 [Geoff Thorpe]
4393
4394 *) Make it possible to unload ranges of ERR strings with a new
4395 "ERR_unload_strings" function.
4396 [Geoff Thorpe]
4397
4398 *) Add a copy() function to EVP_MD.
4399 [Ben Laurie]
4400
4401 *) Make EVP_MD routines take a context pointer instead of just the
4402 md_data void pointer.
4403 [Ben Laurie]
4404
4405 *) Add flags to EVP_MD and EVP_MD_CTX. EVP_MD_FLAG_ONESHOT indicates
4406 that the digest can only process a single chunk of data
4407 (typically because it is provided by a piece of
4408 hardware). EVP_MD_CTX_FLAG_ONESHOT indicates that the application
4409 is only going to provide a single chunk of data, and hence the
4410 framework needn't accumulate the data for oneshot drivers.
4411 [Ben Laurie]
4412
4413 *) As with "ERR", make it possible to replace the underlying "ex_data"
4414 functions. This change also alters the storage and management of global
4415 ex_data state - it's now all inside ex_data.c and all "class" code (eg.
4416 RSA, BIO, SSL_CTX, etc) no longer stores its own STACKS and per-class
4417 index counters. The API functions that use this state have been changed
4418 to take a "class_index" rather than pointers to the class's local STACK
4419 and counter, and there is now an API function to dynamically create new
4420 classes. This centralisation allows us to (a) plug a lot of the
4421 thread-safety problems that existed, and (b) makes it possible to clean
4422 up all allocated state using "CRYPTO_cleanup_all_ex_data()". W.r.t. (b)
4423 such data would previously have always leaked in application code and
4424 workarounds were in place to make the memory debugging turn a blind eye
4425 to it. Application code that doesn't use this new function will still
4426 leak as before, but their memory debugging output will announce it now
4427 rather than letting it slide.
4428
4429 Besides the addition of CRYPTO_cleanup_all_ex_data(), another API change
4430 induced by the "ex_data" overhaul is that X509_STORE_CTX_init() now
4431 has a return value to indicate success or failure.
4432 [Geoff Thorpe]
4433
4434 *) Make it possible to replace the underlying "ERR" functions such that the
4435 global state (2 LHASH tables and 2 locks) is only used by the "default"
4436 implementation. This change also adds two functions to "get" and "set"
4437 the implementation prior to it being automatically set the first time
4438 any other ERR function takes place. Ie. an application can call "get",
4439 pass the return value to a module it has just loaded, and that module
4440 can call its own "set" function using that value. This means the
4441 module's "ERR" operations will use (and modify) the error state in the
4442 application and not in its own statically linked copy of OpenSSL code.
4443 [Geoff Thorpe]
4444
4445 *) Give DH, DSA, and RSA types their own "**_up_ref()" function to increment
4446 reference counts. This performs normal REF_PRINT/REF_CHECK macros on
4447 the operation, and provides a more encapsulated way for external code
4448 (crypto/evp/ and ssl/) to do this. Also changed the evp and ssl code
4449 to use these functions rather than manually incrementing the counts.
4450
4451 Also rename "DSO_up()" function to more descriptive "DSO_up_ref()".
4452 [Geoff Thorpe]
4453
4454 *) Add EVP test program.
4455 [Ben Laurie]
4456
4457 *) Add symmetric cipher support to ENGINE. Expect the API to change!
4458 [Ben Laurie]
4459
4460 *) New CRL functions: X509_CRL_set_version(), X509_CRL_set_issuer_name()
4461 X509_CRL_set_lastUpdate(), X509_CRL_set_nextUpdate(), X509_CRL_sort(),
4462 X509_REVOKED_set_serialNumber(), and X509_REVOKED_set_revocationDate().
4463 These allow a CRL to be built without having to access X509_CRL fields
4464 directly. Modify 'ca' application to use new functions.
4465 [Steve Henson]
4466
4467 *) Move SSL_OP_TLS_ROLLBACK_BUG out of the SSL_OP_ALL list of recommended
4468 bug workarounds. Rollback attack detection is a security feature.
4469 The problem will only arise on OpenSSL servers when TLSv1 is not
4470 available (sslv3_server_method() or SSL_OP_NO_TLSv1).
4471 Software authors not wanting to support TLSv1 will have special reasons
4472 for their choice and can explicitly enable this option.
4473 [Bodo Moeller, Lutz Jaenicke]
4474
4475 *) Rationalise EVP so it can be extended: don't include a union of
4476 cipher/digest structures, add init/cleanup functions for EVP_MD_CTX
4477 (similar to those existing for EVP_CIPHER_CTX).
4478 Usage example:
4479
4480 EVP_MD_CTX md;
4481
4482 EVP_MD_CTX_init(&md); /* new function call */
4483 EVP_DigestInit(&md, EVP_sha1());
4484 EVP_DigestUpdate(&md, in, len);
4485 EVP_DigestFinal(&md, out, NULL);
4486 EVP_MD_CTX_cleanup(&md); /* new function call */
4487
4488 [Ben Laurie]
4489
4490 *) Make DES key schedule conform to the usual scheme, as well as
4491 correcting its structure. This means that calls to DES functions
4492 now have to pass a pointer to a des_key_schedule instead of a
4493 plain des_key_schedule (which was actually always a pointer
4494 anyway): E.g.,
4495
4496 des_key_schedule ks;
4497
4498 des_set_key_checked(..., &ks);
4499 des_ncbc_encrypt(..., &ks, ...);
4500
4501 (Note that a later change renames 'des_...' into 'DES_...'.)
4502 [Ben Laurie]
4503
4504 *) Initial reduction of linker bloat: the use of some functions, such as
4505 PEM causes large amounts of unused functions to be linked in due to
4506 poor organisation. For example pem_all.c contains every PEM function
4507 which has a knock on effect of linking in large amounts of (unused)
4508 ASN1 code. Grouping together similar functions and splitting unrelated
4509 functions prevents this.
4510 [Steve Henson]
4511
4512 *) Cleanup of EVP macros.
4513 [Ben Laurie]
4514
4515 *) Change historical references to {NID,SN,LN}_des_ede and ede3 to add the
4516 correct _ecb suffix.
4517 [Ben Laurie]
4518
4519 *) Add initial OCSP responder support to ocsp application. The
4520 revocation information is handled using the text based index
4521 use by the ca application. The responder can either handle
4522 requests generated internally, supplied in files (for example
4523 via a CGI script) or using an internal minimal server.
4524 [Steve Henson]
4525
4526 *) Add configuration choices to get zlib compression for TLS.
4527 [Richard Levitte]
4528
4529 *) Changes to Kerberos SSL for RFC 2712 compliance:
4530 1. Implemented real KerberosWrapper, instead of just using
4531 KRB5 AP_REQ message. [Thanks to Simon Wilkinson <sxw@sxw.org.uk>]
4532 2. Implemented optional authenticator field of KerberosWrapper.
4533
4534 Added openssl-style ASN.1 macros for Kerberos ticket, ap_req,
4535 and authenticator structs; see crypto/krb5/.
4536
4537 Generalized Kerberos calls to support multiple Kerberos libraries.
4538 [Vern Staats <staatsvr@asc.hpc.mil>,
4539 Jeffrey Altman <jaltman@columbia.edu>
4540 via Richard Levitte]
4541
4542 *) Cause 'openssl speed' to use fully hard-coded DSA keys as it
4543 already does with RSA. testdsa.h now has 'priv_key/pub_key'
4544 values for each of the key sizes rather than having just
4545 parameters (and 'speed' generating keys each time).
4546 [Geoff Thorpe]
4547
4548 *) Speed up EVP routines.
4549 Before:
4550 encrypt
4551 type 8 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes
4552 des-cbc 4408.85k 5560.51k 5778.46k 5862.20k 5825.16k
4553 des-cbc 4389.55k 5571.17k 5792.23k 5846.91k 5832.11k
4554 des-cbc 4394.32k 5575.92k 5807.44k 5848.37k 5841.30k
4555 decrypt
4556 des-cbc 3482.66k 5069.49k 5496.39k 5614.16k 5639.28k
4557 des-cbc 3480.74k 5068.76k 5510.34k 5609.87k 5635.52k
4558 des-cbc 3483.72k 5067.62k 5504.60k 5708.01k 5724.80k
4559 After:
4560 encrypt
4561 des-cbc 4660.16k 5650.19k 5807.19k 5827.13k 5783.32k
4562 decrypt
4563 des-cbc 3624.96k 5258.21k 5530.91k 5624.30k 5628.26k
4564 [Ben Laurie]
4565
4566 *) Added the OS2-EMX target.
4567 ["Brian Havard" <brianh@kheldar.apana.org.au> and Richard Levitte]
4568
4569 *) Rewrite apps to use NCONF routines instead of the old CONF. New functions
4570 to support NCONF routines in extension code. New function CONF_set_nconf()
4571 to allow functions which take an NCONF to also handle the old LHASH
4572 structure: this means that the old CONF compatible routines can be
4573 retained (in particular wrt extensions) without having to duplicate the
4574 code. New function X509V3_add_ext_nconf_sk to add extensions to a stack.
4575 [Steve Henson]
4576
4577 *) Enhance the general user interface with mechanisms for inner control
4578 and with possibilities to have yes/no kind of prompts.
4579 [Richard Levitte]
4580
4581 *) Change all calls to low level digest routines in the library and
4582 applications to use EVP. Add missing calls to HMAC_cleanup() and
4583 don't assume HMAC_CTX can be copied using memcpy().
4584 [Verdon Walker <VWalker@novell.com>, Steve Henson]
4585
4586 *) Add the possibility to control engines through control names but with
4587 arbitrary arguments instead of just a string.
4588 Change the key loaders to take a UI_METHOD instead of a callback
4589 function pointer. NOTE: this breaks binary compatibility with earlier
4590 versions of OpenSSL [engine].
4591 Adapt the nCipher code for these new conditions and add a card insertion
4592 callback.
4593 [Richard Levitte]
4594
4595 *) Enhance the general user interface with mechanisms to better support
4596 dialog box interfaces, application-defined prompts, the possibility
4597 to use defaults (for example default passwords from somewhere else)
4598 and interrupts/cancellations.
4599 [Richard Levitte]
4600
4601 *) Tidy up PKCS#12 attribute handling. Add support for the CSP name
4602 attribute in PKCS#12 files, add new -CSP option to pkcs12 utility.
4603 [Steve Henson]
4604
4605 *) Fix a memory leak in 'sk_dup()' in the case reallocation fails. (Also
4606 tidy up some unnecessarily weird code in 'sk_new()').
4607 [Geoff, reported by Diego Tartara <dtartara@novamens.com>]
4608
4609 *) Change the key loading routines for ENGINEs to use the same kind
4610 callback (pem_password_cb) as all other routines that need this
4611 kind of callback.
4612 [Richard Levitte]
4613
4614 *) Increase ENTROPY_NEEDED to 32 bytes, as Rijndael can operate with
4615 256 bit (=32 byte) keys. Of course seeding with more entropy bytes
4616 than this minimum value is recommended.
4617 [Lutz Jaenicke]
4618
4619 *) New random seeder for OpenVMS, using the system process statistics
4620 that are easily reachable.
4621 [Richard Levitte]
4622
4623 *) Windows apparently can't transparently handle global
4624 variables defined in DLLs. Initialisations such as:
4625
4626 const ASN1_ITEM *it = &ASN1_INTEGER_it;
4627
4628 wont compile. This is used by the any applications that need to
4629 declare their own ASN1 modules. This was fixed by adding the option
4630 EXPORT_VAR_AS_FN to all Win32 platforms, although this isn't strictly
4631 needed for static libraries under Win32.
4632 [Steve Henson]
4633
4634 *) New functions X509_PURPOSE_set() and X509_TRUST_set() to handle
4635 setting of purpose and trust fields. New X509_STORE trust and
4636 purpose functions and tidy up setting in other SSL functions.
4637 [Steve Henson]
4638
4639 *) Add copies of X509_STORE_CTX fields and callbacks to X509_STORE
4640 structure. These are inherited by X509_STORE_CTX when it is
4641 initialised. This allows various defaults to be set in the
4642 X509_STORE structure (such as flags for CRL checking and custom
4643 purpose or trust settings) for functions which only use X509_STORE_CTX
4644 internally such as S/MIME.
4645
4646 Modify X509_STORE_CTX_purpose_inherit() so it only sets purposes and
4647 trust settings if they are not set in X509_STORE. This allows X509_STORE
4648 purposes and trust (in S/MIME for example) to override any set by default.
4649
4650 Add command line options for CRL checking to smime, s_client and s_server
4651 applications.
4652 [Steve Henson]
4653
4654 *) Initial CRL based revocation checking. If the CRL checking flag(s)
4655 are set then the CRL is looked up in the X509_STORE structure and
4656 its validity and signature checked, then if the certificate is found
4657 in the CRL the verify fails with a revoked error.
4658
4659 Various new CRL related callbacks added to X509_STORE_CTX structure.
4660
4661 Command line options added to 'verify' application to support this.
4662
4663 This needs some additional work, such as being able to handle multiple
4664 CRLs with different times, extension based lookup (rather than just
4665 by subject name) and ultimately more complete V2 CRL extension
4666 handling.
4667 [Steve Henson]
4668
4669 *) Add a general user interface API (crypto/ui/). This is designed
4670 to replace things like des_read_password and friends (backward
4671 compatibility functions using this new API are provided).
4672 The purpose is to remove prompting functions from the DES code
4673 section as well as provide for prompting through dialog boxes in
4674 a window system and the like.
4675 [Richard Levitte]
4676
4677 *) Add "ex_data" support to ENGINE so implementations can add state at a
4678 per-structure level rather than having to store it globally.
4679 [Geoff]
4680
4681 *) Make it possible for ENGINE structures to be copied when retrieved by
4682 ENGINE_by_id() if the ENGINE specifies a new flag: ENGINE_FLAGS_BY_ID_COPY.
4683 This causes the "original" ENGINE structure to act like a template,
4684 analogous to the RSA vs. RSA_METHOD type of separation. Because of this
4685 operational state can be localised to each ENGINE structure, despite the
4686 fact they all share the same "methods". New ENGINE structures returned in
4687 this case have no functional references and the return value is the single
4688 structural reference. This matches the single structural reference returned
4689 by ENGINE_by_id() normally, when it is incremented on the pre-existing
4690 ENGINE structure.
4691 [Geoff]
4692
4693 *) Fix ASN1 decoder when decoding type ANY and V_ASN1_OTHER: since this
4694 needs to match any other type at all we need to manually clear the
4695 tag cache.
4696 [Steve Henson]
4697
4698 *) Changes to the "openssl engine" utility to include;
4699 - verbosity levels ('-v', '-vv', and '-vvv') that provide information
4700 about an ENGINE's available control commands.
4701 - executing control commands from command line arguments using the
4702 '-pre' and '-post' switches. '-post' is only used if '-t' is
4703 specified and the ENGINE is successfully initialised. The syntax for
4704 the individual commands are colon-separated, for example;
4705 openssl engine chil -pre FORK_CHECK:0 -pre SO_PATH:/lib/test.so
4706 [Geoff]
4707
4708 *) New dynamic control command support for ENGINEs. ENGINEs can now
4709 declare their own commands (numbers), names (strings), descriptions,
4710 and input types for run-time discovery by calling applications. A
4711 subset of these commands are implicitly classed as "executable"
4712 depending on their input type, and only these can be invoked through
4713 the new string-based API function ENGINE_ctrl_cmd_string(). (Eg. this
4714 can be based on user input, config files, etc). The distinction is
4715 that "executable" commands cannot return anything other than a boolean
4716 result and can only support numeric or string input, whereas some
4717 discoverable commands may only be for direct use through
4718 ENGINE_ctrl(), eg. supporting the exchange of binary data, function
4719 pointers, or other custom uses. The "executable" commands are to
4720 support parameterisations of ENGINE behaviour that can be
4721 unambiguously defined by ENGINEs and used consistently across any
4722 OpenSSL-based application. Commands have been added to all the
4723 existing hardware-supporting ENGINEs, noticeably "SO_PATH" to allow
4724 control over shared-library paths without source code alterations.
4725 [Geoff]
4726
4727 *) Changed all ENGINE implementations to dynamically allocate their
4728 ENGINEs rather than declaring them statically. Apart from this being
4729 necessary with the removal of the ENGINE_FLAGS_MALLOCED distinction,
4730 this also allows the implementations to compile without using the
4731 internal engine_int.h header.
4732 [Geoff]
4733
4734 *) Minor adjustment to "rand" code. RAND_get_rand_method() now returns a
4735 'const' value. Any code that should be able to modify a RAND_METHOD
4736 should already have non-const pointers to it (ie. they should only
4737 modify their own ones).
4738 [Geoff]
4739
4740 *) Made a variety of little tweaks to the ENGINE code.
4741 - "atalla" and "ubsec" string definitions were moved from header files
4742 to C code. "nuron" string definitions were placed in variables
4743 rather than hard-coded - allowing parameterisation of these values
4744 later on via ctrl() commands.
4745 - Removed unused "#if 0"'d code.
4746 - Fixed engine list iteration code so it uses ENGINE_free() to release
4747 structural references.
4748 - Constified the RAND_METHOD element of ENGINE structures.
4749 - Constified various get/set functions as appropriate and added
4750 missing functions (including a catch-all ENGINE_cpy that duplicates
4751 all ENGINE values onto a new ENGINE except reference counts/state).
4752 - Removed NULL parameter checks in get/set functions. Setting a method
4753 or function to NULL is a way of cancelling out a previously set
4754 value. Passing a NULL ENGINE parameter is just plain stupid anyway
4755 and doesn't justify the extra error symbols and code.
4756 - Deprecate the ENGINE_FLAGS_MALLOCED define and move the area for
4757 flags from engine_int.h to engine.h.
4758 - Changed prototypes for ENGINE handler functions (init(), finish(),
4759 ctrl(), key-load functions, etc) to take an (ENGINE*) parameter.
4760 [Geoff]
4761
4762 *) Implement binary inversion algorithm for BN_mod_inverse in addition
4763 to the algorithm using long division. The binary algorithm can be
4764 used only if the modulus is odd. On 32-bit systems, it is faster
4765 only for relatively small moduli (roughly 20-30% for 128-bit moduli,
4766 roughly 5-15% for 256-bit moduli), so we use it only for moduli
4767 up to 450 bits. In 64-bit environments, the binary algorithm
4768 appears to be advantageous for much longer moduli; here we use it
4769 for moduli up to 2048 bits.
4770 [Bodo Moeller]
4771
4772 *) Rewrite CHOICE field setting in ASN1_item_ex_d2i(). The old code
4773 could not support the combine flag in choice fields.
4774 [Steve Henson]
4775
4776 *) Add a 'copy_extensions' option to the 'ca' utility. This copies
4777 extensions from a certificate request to the certificate.
4778 [Steve Henson]
4779
4780 *) Allow multiple 'certopt' and 'nameopt' options to be separated
4781 by commas. Add 'namopt' and 'certopt' options to the 'ca' config
4782 file: this allows the display of the certificate about to be
4783 signed to be customised, to allow certain fields to be included
4784 or excluded and extension details. The old system didn't display
4785 multicharacter strings properly, omitted fields not in the policy
4786 and couldn't display additional details such as extensions.
4787 [Steve Henson]
4788
4789 *) Function EC_POINTs_mul for multiple scalar multiplication
4790 of an arbitrary number of elliptic curve points
4791 \sum scalars[i]*points[i],
4792 optionally including the generator defined for the EC_GROUP:
4793 scalar*generator + \sum scalars[i]*points[i].
4794
4795 EC_POINT_mul is a simple wrapper function for the typical case
4796 that the point list has just one item (besides the optional
4797 generator).
4798 [Bodo Moeller]
4799
4800 *) First EC_METHODs for curves over GF(p):
4801
4802 EC_GFp_simple_method() uses the basic BN_mod_mul and BN_mod_sqr
4803 operations and provides various method functions that can also
4804 operate with faster implementations of modular arithmetic.
4805
4806 EC_GFp_mont_method() reuses most functions that are part of
4807 EC_GFp_simple_method, but uses Montgomery arithmetic.
4808
4809 [Bodo Moeller; point addition and point doubling
4810 implementation directly derived from source code provided by
4811 Lenka Fibikova <fibikova@exp-math.uni-essen.de>]
4812
4813 *) Framework for elliptic curves (crypto/ec/ec.h, crypto/ec/ec_lcl.h,
4814 crypto/ec/ec_lib.c):
4815
4816 Curves are EC_GROUP objects (with an optional group generator)
4817 based on EC_METHODs that are built into the library.
4818
4819 Points are EC_POINT objects based on EC_GROUP objects.
4820
4821 Most of the framework would be able to handle curves over arbitrary
4822 finite fields, but as there are no obvious types for fields other
4823 than GF(p), some functions are limited to that for now.
4824 [Bodo Moeller]
4825
4826 *) Add the -HTTP option to s_server. It is similar to -WWW, but requires
4827 that the file contains a complete HTTP response.
4828 [Richard Levitte]
4829
4830 *) Add the ec directory to mkdef.pl and mkfiles.pl. In mkdef.pl
4831 change the def and num file printf format specifier from "%-40sXXX"
4832 to "%-39s XXX". The latter will always guarantee a space after the
4833 field while the former will cause them to run together if the field
4834 is 40 of more characters long.
4835 [Steve Henson]
4836
4837 *) Constify the cipher and digest 'method' functions and structures
4838 and modify related functions to take constant EVP_MD and EVP_CIPHER
4839 pointers.
4840 [Steve Henson]
4841
4842 *) Hide BN_CTX structure details in bn_lcl.h instead of publishing them
4843 in <openssl/bn.h>. Also further increase BN_CTX_NUM to 32.
4844 [Bodo Moeller]
4845
4846 *) Modify EVP_Digest*() routines so they now return values. Although the
4847 internal software routines can never fail additional hardware versions
4848 might.
4849 [Steve Henson]
4850
4851 *) Clean up crypto/err/err.h and change some error codes to avoid conflicts:
4852
4853 Previously ERR_R_FATAL was too small and coincided with ERR_LIB_PKCS7
4854 (= ERR_R_PKCS7_LIB); it is now 64 instead of 32.
4855
4856 ASN1 error codes
4857 ERR_R_NESTED_ASN1_ERROR
4858 ...
4859 ERR_R_MISSING_ASN1_EOS
4860 were 4 .. 9, conflicting with
4861 ERR_LIB_RSA (= ERR_R_RSA_LIB)
4862 ...
4863 ERR_LIB_PEM (= ERR_R_PEM_LIB).
4864 They are now 58 .. 63 (i.e., just below ERR_R_FATAL).
4865
4866 Add new error code 'ERR_R_INTERNAL_ERROR'.
4867 [Bodo Moeller]
4868
4869 *) Don't overuse locks in crypto/err/err.c: For data retrieval, CRYPTO_r_lock
4870 suffices.
4871 [Bodo Moeller]
4872
4873 *) New option '-subj arg' for 'openssl req' and 'openssl ca'. This
4874 sets the subject name for a new request or supersedes the
4875 subject name in a given request. Formats that can be parsed are
4876 'CN=Some Name, OU=myOU, C=IT'
4877 and
4878 'CN=Some Name/OU=myOU/C=IT'.
4879
4880 Add options '-batch' and '-verbose' to 'openssl req'.
4881 [Massimiliano Pala <madwolf@hackmasters.net>]
4882
4883 *) Introduce the possibility to access global variables through
4884 functions on platform were that's the best way to handle exporting
4885 global variables in shared libraries. To enable this functionality,
4886 one must configure with "EXPORT_VAR_AS_FN" or defined the C macro
4887 "OPENSSL_EXPORT_VAR_AS_FUNCTION" in crypto/opensslconf.h (the latter
4888 is normally done by Configure or something similar).
4889
4890 To implement a global variable, use the macro OPENSSL_IMPLEMENT_GLOBAL
4891 in the source file (foo.c) like this:
4892
4893 OPENSSL_IMPLEMENT_GLOBAL(int,foo)=1;
4894 OPENSSL_IMPLEMENT_GLOBAL(double,bar);
4895
4896 To declare a global variable, use the macros OPENSSL_DECLARE_GLOBAL
4897 and OPENSSL_GLOBAL_REF in the header file (foo.h) like this:
4898
4899 OPENSSL_DECLARE_GLOBAL(int,foo);
4900 #define foo OPENSSL_GLOBAL_REF(foo)
4901 OPENSSL_DECLARE_GLOBAL(double,bar);
4902 #define bar OPENSSL_GLOBAL_REF(bar)
4903
4904 The #defines are very important, and therefore so is including the
4905 header file everywhere where the defined globals are used.
4906
4907 The macro OPENSSL_EXPORT_VAR_AS_FUNCTION also affects the definition
4908 of ASN.1 items, but that structure is a bit different.
4909
4910 The largest change is in util/mkdef.pl which has been enhanced with
4911 better and easier to understand logic to choose which symbols should
4912 go into the Windows .def files as well as a number of fixes and code
4913 cleanup (among others, algorithm keywords are now sorted
4914 lexicographically to avoid constant rewrites).
4915 [Richard Levitte]
4916
4917 *) In BN_div() keep a copy of the sign of 'num' before writing the
4918 result to 'rm' because if rm==num the value will be overwritten
4919 and produce the wrong result if 'num' is negative: this caused
4920 problems with BN_mod() and BN_nnmod().
4921 [Steve Henson]
4922
4923 *) Function OCSP_request_verify(). This checks the signature on an
4924 OCSP request and verifies the signer certificate. The signer
4925 certificate is just checked for a generic purpose and OCSP request
4926 trust settings.
4927 [Steve Henson]
4928
4929 *) Add OCSP_check_validity() function to check the validity of OCSP
4930 responses. OCSP responses are prepared in real time and may only
4931 be a few seconds old. Simply checking that the current time lies
4932 between thisUpdate and nextUpdate max reject otherwise valid responses
4933 caused by either OCSP responder or client clock inaccuracy. Instead
4934 we allow thisUpdate and nextUpdate to fall within a certain period of
4935 the current time. The age of the response can also optionally be
4936 checked. Two new options -validity_period and -status_age added to
4937 ocsp utility.
4938 [Steve Henson]
4939
4940 *) If signature or public key algorithm is unrecognized print out its
4941 OID rather that just UNKNOWN.
4942 [Steve Henson]
4943
4944 *) Change OCSP_cert_to_id() to tolerate a NULL subject certificate and
4945 OCSP_cert_id_new() a NULL serialNumber. This allows a partial certificate
4946 ID to be generated from the issuer certificate alone which can then be
4947 passed to OCSP_id_issuer_cmp().
4948 [Steve Henson]
4949
4950 *) New compilation option ASN1_ITEM_FUNCTIONS. This causes the new
4951 ASN1 modules to export functions returning ASN1_ITEM pointers
4952 instead of the ASN1_ITEM structures themselves. This adds several
4953 new macros which allow the underlying ASN1 function/structure to
4954 be accessed transparently. As a result code should not use ASN1_ITEM
4955 references directly (such as &X509_it) but instead use the relevant
4956 macros (such as ASN1_ITEM_rptr(X509)). This option is to allow
4957 use of the new ASN1 code on platforms where exporting structures
4958 is problematical (for example in shared libraries) but exporting
4959 functions returning pointers to structures is not.
4960 [Steve Henson]
4961
4962 *) Add support for overriding the generation of SSL/TLS session IDs.
4963 These callbacks can be registered either in an SSL_CTX or per SSL.
4964 The purpose of this is to allow applications to control, if they wish,
4965 the arbitrary values chosen for use as session IDs, particularly as it
4966 can be useful for session caching in multiple-server environments. A
4967 command-line switch for testing this (and any client code that wishes
4968 to use such a feature) has been added to "s_server".
4969 [Geoff Thorpe, Lutz Jaenicke]
4970
4971 *) Modify mkdef.pl to recognise and parse preprocessor conditionals
4972 of the form '#if defined(...) || defined(...) || ...' and
4973 '#if !defined(...) && !defined(...) && ...'. This also avoids
4974 the growing number of special cases it was previously handling.
4975 [Richard Levitte]
4976
4977 *) Make all configuration macros available for application by making
4978 sure they are available in opensslconf.h, by giving them names starting
4979 with "OPENSSL_" to avoid conflicts with other packages and by making
4980 sure e_os2.h will cover all platform-specific cases together with
4981 opensslconf.h.
4982 Additionally, it is now possible to define configuration/platform-
4983 specific names (called "system identities"). In the C code, these
4984 are prefixed with "OPENSSL_SYSNAME_". e_os2.h will create another
4985 macro with the name beginning with "OPENSSL_SYS_", which is determined
4986 from "OPENSSL_SYSNAME_*" or compiler-specific macros depending on
4987 what is available.
4988 [Richard Levitte]
4989
4990 *) New option -set_serial to 'req' and 'x509' this allows the serial
4991 number to use to be specified on the command line. Previously self
4992 signed certificates were hard coded with serial number 0 and the
4993 CA options of 'x509' had to use a serial number in a file which was
4994 auto incremented.
4995 [Steve Henson]
4996
4997 *) New options to 'ca' utility to support V2 CRL entry extensions.
4998 Currently CRL reason, invalidity date and hold instruction are
4999 supported. Add new CRL extensions to V3 code and some new objects.
5000 [Steve Henson]
5001
5002 *) New function EVP_CIPHER_CTX_set_padding() this is used to
5003 disable standard block padding (aka PKCS#5 padding) in the EVP
5004 API, which was previously mandatory. This means that the data is
5005 not padded in any way and so the total length much be a multiple
5006 of the block size, otherwise an error occurs.
5007 [Steve Henson]
5008
5009 *) Initial (incomplete) OCSP SSL support.
5010 [Steve Henson]
5011
5012 *) New function OCSP_parse_url(). This splits up a URL into its host,
5013 port and path components: primarily to parse OCSP URLs. New -url
5014 option to ocsp utility.
5015 [Steve Henson]
5016
5017 *) New nonce behavior. The return value of OCSP_check_nonce() now
5018 reflects the various checks performed. Applications can decide
5019 whether to tolerate certain situations such as an absent nonce
5020 in a response when one was present in a request: the ocsp application
5021 just prints out a warning. New function OCSP_add1_basic_nonce()
5022 this is to allow responders to include a nonce in a response even if
5023 the request is nonce-less.
5024 [Steve Henson]
5025
5026 *) Disable stdin buffering in load_cert (apps/apps.c) so that no certs are
5027 skipped when using openssl x509 multiple times on a single input file,
5028 e.g. "(openssl x509 -out cert1; openssl x509 -out cert2) <certs".
5029 [Bodo Moeller]
5030
5031 *) Make ASN1_UTCTIME_set_string() and ASN1_GENERALIZEDTIME_set_string()
5032 set string type: to handle setting ASN1_TIME structures. Fix ca
5033 utility to correctly initialize revocation date of CRLs.
5034 [Steve Henson]
5035
5036 *) New option SSL_OP_CIPHER_SERVER_PREFERENCE allows the server to override
5037 the clients preferred ciphersuites and rather use its own preferences.
5038 Should help to work around M$ SGC (Server Gated Cryptography) bug in
5039 Internet Explorer by ensuring unchanged hash method during stepup.
5040 (Also replaces the broken/deactivated SSL_OP_NON_EXPORT_FIRST option.)
5041 [Lutz Jaenicke]
5042
5043 *) Make mkdef.pl recognise all DECLARE_ASN1 macros, change rijndael
5044 to aes and add a new 'exist' option to print out symbols that don't
5045 appear to exist.
5046 [Steve Henson]
5047
5048 *) Additional options to ocsp utility to allow flags to be set and
5049 additional certificates supplied.
5050 [Steve Henson]
5051
5052 *) Add the option -VAfile to 'openssl ocsp', so the user can give the
5053 OCSP client a number of certificate to only verify the response
5054 signature against.
5055 [Richard Levitte]
5056
5057 *) Update Rijndael code to version 3.0 and change EVP AES ciphers to
5058 handle the new API. Currently only ECB, CBC modes supported. Add new
5059 AES OIDs.
5060
5061 Add TLS AES ciphersuites as described in RFC3268, "Advanced
5062 Encryption Standard (AES) Ciphersuites for Transport Layer
5063 Security (TLS)". (In beta versions of OpenSSL 0.9.7, these were
5064 not enabled by default and were not part of the "ALL" ciphersuite
5065 alias because they were not yet official; they could be
5066 explicitly requested by specifying the "AESdraft" ciphersuite
5067 group alias. In the final release of OpenSSL 0.9.7, the group
5068 alias is called "AES" and is part of "ALL".)
5069 [Ben Laurie, Steve Henson, Bodo Moeller]
5070
5071 *) New function OCSP_copy_nonce() to copy nonce value (if present) from
5072 request to response.
5073 [Steve Henson]
5074
5075 *) Functions for OCSP responders. OCSP_request_onereq_count(),
5076 OCSP_request_onereq_get0(), OCSP_onereq_get0_id() and OCSP_id_get0_info()
5077 extract information from a certificate request. OCSP_response_create()
5078 creates a response and optionally adds a basic response structure.
5079 OCSP_basic_add1_status() adds a complete single response to a basic
5080 response and returns the OCSP_SINGLERESP structure just added (to allow
5081 extensions to be included for example). OCSP_basic_add1_cert() adds a
5082 certificate to a basic response and OCSP_basic_sign() signs a basic
5083 response with various flags. New helper functions ASN1_TIME_check()
5084 (checks validity of ASN1_TIME structure) and ASN1_TIME_to_generalizedtime()
5085 (converts ASN1_TIME to GeneralizedTime).
5086 [Steve Henson]
5087
5088 *) Various new functions. EVP_Digest() combines EVP_Digest{Init,Update,Final}()
5089 in a single operation. X509_get0_pubkey_bitstr() extracts the public_key
5090 structure from a certificate. X509_pubkey_digest() digests the public_key
5091 contents: this is used in various key identifiers.
5092 [Steve Henson]
5093
5094 *) Make sk_sort() tolerate a NULL argument.
5095 [Steve Henson reported by Massimiliano Pala <madwolf@comune.modena.it>]
5096
5097 *) New OCSP verify flag OCSP_TRUSTOTHER. When set the "other" certificates
5098 passed by the function are trusted implicitly. If any of them signed the
5099 response then it is assumed to be valid and is not verified.
5100 [Steve Henson]
5101
5102 *) In PKCS7_set_type() initialise content_type in PKCS7_ENC_CONTENT
5103 to data. This was previously part of the PKCS7 ASN1 code. This
5104 was causing problems with OpenSSL created PKCS#12 and PKCS#7 structures.
5105 [Steve Henson, reported by Kenneth R. Robinette
5106 <support@securenetterm.com>]
5107
5108 *) Add CRYPTO_push_info() and CRYPTO_pop_info() calls to new ASN1
5109 routines: without these tracing memory leaks is very painful.
5110 Fix leaks in PKCS12 and PKCS7 routines.
5111 [Steve Henson]
5112
5113 *) Make X509_time_adj() cope with the new behaviour of ASN1_TIME_new().
5114 Previously it initialised the 'type' argument to V_ASN1_UTCTIME which
5115 effectively meant GeneralizedTime would never be used. Now it
5116 is initialised to -1 but X509_time_adj() now has to check the value
5117 and use ASN1_TIME_set() if the value is not V_ASN1_UTCTIME or
5118 V_ASN1_GENERALIZEDTIME, without this it always uses GeneralizedTime.
5119 [Steve Henson, reported by Kenneth R. Robinette
5120 <support@securenetterm.com>]
5121
5122 *) Fixes to BN_to_ASN1_INTEGER when bn is zero. This would previously
5123 result in a zero length in the ASN1_INTEGER structure which was
5124 not consistent with the structure when d2i_ASN1_INTEGER() was used
5125 and would cause ASN1_INTEGER_cmp() to fail. Enhance s2i_ASN1_INTEGER()
5126 to cope with hex and negative integers. Fix bug in i2a_ASN1_INTEGER()
5127 where it did not print out a minus for negative ASN1_INTEGER.
5128 [Steve Henson]
5129
5130 *) Add summary printout to ocsp utility. The various functions which
5131 convert status values to strings have been renamed to:
5132 OCSP_response_status_str(), OCSP_cert_status_str() and
5133 OCSP_crl_reason_str() and are no longer static. New options
5134 to verify nonce values and to disable verification. OCSP response
5135 printout format cleaned up.
5136 [Steve Henson]
5137
5138 *) Add additional OCSP certificate checks. These are those specified
5139 in RFC2560. This consists of two separate checks: the CA of the
5140 certificate being checked must either be the OCSP signer certificate
5141 or the issuer of the OCSP signer certificate. In the latter case the
5142 OCSP signer certificate must contain the OCSP signing extended key
5143 usage. This check is performed by attempting to match the OCSP
5144 signer or the OCSP signer CA to the issuerNameHash and issuerKeyHash
5145 in the OCSP_CERTID structures of the response.
5146 [Steve Henson]
5147
5148 *) Initial OCSP certificate verification added to OCSP_basic_verify()
5149 and related routines. This uses the standard OpenSSL certificate
5150 verify routines to perform initial checks (just CA validity) and
5151 to obtain the certificate chain. Then additional checks will be
5152 performed on the chain. Currently the root CA is checked to see
5153 if it is explicitly trusted for OCSP signing. This is used to set
5154 a root CA as a global signing root: that is any certificate that
5155 chains to that CA is an acceptable OCSP signing certificate.
5156 [Steve Henson]
5157
5158 *) New '-extfile ...' option to 'openssl ca' for reading X.509v3
5159 extensions from a separate configuration file.
5160 As when reading extensions from the main configuration file,
5161 the '-extensions ...' option may be used for specifying the
5162 section to use.
5163 [Massimiliano Pala <madwolf@comune.modena.it>]
5164
5165 *) New OCSP utility. Allows OCSP requests to be generated or
5166 read. The request can be sent to a responder and the output
5167 parsed, outputed or printed in text form. Not complete yet:
5168 still needs to check the OCSP response validity.
5169 [Steve Henson]
5170
5171 *) New subcommands for 'openssl ca':
5172 'openssl ca -status <serial>' prints the status of the cert with
5173 the given serial number (according to the index file).
5174 'openssl ca -updatedb' updates the expiry status of certificates
5175 in the index file.
5176 [Massimiliano Pala <madwolf@comune.modena.it>]
5177
5178 *) New '-newreq-nodes' command option to CA.pl. This is like
5179 '-newreq', but calls 'openssl req' with the '-nodes' option
5180 so that the resulting key is not encrypted.
5181 [Damien Miller <djm@mindrot.org>]
5182
5183 *) New configuration for the GNU Hurd.
5184 [Jonathan Bartlett <johnnyb@wolfram.com> via Richard Levitte]
5185
5186 *) Initial code to implement OCSP basic response verify. This
5187 is currently incomplete. Currently just finds the signer's
5188 certificate and verifies the signature on the response.
5189 [Steve Henson]
5190
5191 *) New SSLeay_version code SSLEAY_DIR to determine the compiled-in
5192 value of OPENSSLDIR. This is available via the new '-d' option
5193 to 'openssl version', and is also included in 'openssl version -a'.
5194 [Bodo Moeller]
5195
5196 *) Allowing defining memory allocation callbacks that will be given
5197 file name and line number information in additional arguments
5198 (a const char* and an int). The basic functionality remains, as
5199 well as the original possibility to just replace malloc(),
5200 realloc() and free() by functions that do not know about these
5201 additional arguments. To register and find out the current
5202 settings for extended allocation functions, the following
5203 functions are provided:
5204
5205 CRYPTO_set_mem_ex_functions
5206 CRYPTO_set_locked_mem_ex_functions
5207 CRYPTO_get_mem_ex_functions
5208 CRYPTO_get_locked_mem_ex_functions
5209
5210 These work the same way as CRYPTO_set_mem_functions and friends.
5211 CRYPTO_get_[locked_]mem_functions now writes 0 where such an
5212 extended allocation function is enabled.
5213 Similarly, CRYPTO_get_[locked_]mem_ex_functions writes 0 where
5214 a conventional allocation function is enabled.
5215 [Richard Levitte, Bodo Moeller]
5216
5217 *) Finish off removing the remaining LHASH function pointer casts.
5218 There should no longer be any prototype-casting required when using
5219 the LHASH abstraction, and any casts that remain are "bugs". See
5220 the callback types and macros at the head of lhash.h for details
5221 (and "OBJ_cleanup" in crypto/objects/obj_dat.c as an example).
5222 [Geoff Thorpe]
5223
5224 *) Add automatic query of EGD sockets in RAND_poll() for the unix variant.
5225 If /dev/[u]random devices are not available or do not return enough
5226 entropy, EGD style sockets (served by EGD or PRNGD) will automatically
5227 be queried.
5228 The locations /var/run/egd-pool, /dev/egd-pool, /etc/egd-pool, and
5229 /etc/entropy will be queried once each in this sequence, quering stops
5230 when enough entropy was collected without querying more sockets.
5231 [Lutz Jaenicke]
5232
5233 *) Change the Unix RAND_poll() variant to be able to poll several
5234 random devices, as specified by DEVRANDOM, until a sufficient amount
5235 of data has been collected. We spend at most 10 ms on each file
5236 (select timeout) and read in non-blocking mode. DEVRANDOM now
5237 defaults to the list "/dev/urandom", "/dev/random", "/dev/srandom"
5238 (previously it was just the string "/dev/urandom"), so on typical
5239 platforms the 10 ms delay will never occur.
5240 Also separate out the Unix variant to its own file, rand_unix.c.
5241 For VMS, there's a currently-empty rand_vms.c.
5242 [Richard Levitte]
5243
5244 *) Move OCSP client related routines to ocsp_cl.c. These
5245 provide utility functions which an application needing
5246 to issue a request to an OCSP responder and analyse the
5247 response will typically need: as opposed to those which an
5248 OCSP responder itself would need which will be added later.
5249
5250 OCSP_request_sign() signs an OCSP request with an API similar
5251 to PKCS7_sign(). OCSP_response_status() returns status of OCSP
5252 response. OCSP_response_get1_basic() extracts basic response
5253 from response. OCSP_resp_find_status(): finds and extracts status
5254 information from an OCSP_CERTID structure (which will be created
5255 when the request structure is built). These are built from lower
5256 level functions which work on OCSP_SINGLERESP structures but
5257 wont normally be used unless the application wishes to examine
5258 extensions in the OCSP response for example.
5259
5260 Replace nonce routines with a pair of functions.
5261 OCSP_request_add1_nonce() adds a nonce value and optionally
5262 generates a random value. OCSP_check_nonce() checks the
5263 validity of the nonce in an OCSP response.
5264 [Steve Henson]
5265
5266 *) Change function OCSP_request_add() to OCSP_request_add0_id().
5267 This doesn't copy the supplied OCSP_CERTID and avoids the
5268 need to free up the newly created id. Change return type
5269 to OCSP_ONEREQ to return the internal OCSP_ONEREQ structure.
5270 This can then be used to add extensions to the request.
5271 Deleted OCSP_request_new(), since most of its functionality
5272 is now in OCSP_REQUEST_new() (and the case insensitive name
5273 clash) apart from the ability to set the request name which
5274 will be added elsewhere.
5275 [Steve Henson]
5276
5277 *) Update OCSP API. Remove obsolete extensions argument from
5278 various functions. Extensions are now handled using the new
5279 OCSP extension code. New simple OCSP HTTP function which
5280 can be used to send requests and parse the response.
5281 [Steve Henson]
5282
5283 *) Fix the PKCS#7 (S/MIME) code to work with new ASN1. Two new
5284 ASN1_ITEM structures help with sign and verify. PKCS7_ATTR_SIGN
5285 uses the special reorder version of SET OF to sort the attributes
5286 and reorder them to match the encoded order. This resolves a long
5287 standing problem: a verify on a PKCS7 structure just after signing
5288 it used to fail because the attribute order did not match the
5289 encoded order. PKCS7_ATTR_VERIFY does not reorder the attributes:
5290 it uses the received order. This is necessary to tolerate some broken
5291 software that does not order SET OF. This is handled by encoding
5292 as a SEQUENCE OF but using implicit tagging (with UNIVERSAL class)
5293 to produce the required SET OF.
5294 [Steve Henson]
5295
5296 *) Have mk1mf.pl generate the macros OPENSSL_BUILD_SHLIBCRYPTO and
5297 OPENSSL_BUILD_SHLIBSSL and use them appropriately in the header
5298 files to get correct declarations of the ASN.1 item variables.
5299 [Richard Levitte]
5300
5301 *) Rewrite of PKCS#12 code to use new ASN1 functionality. Replace many
5302 PKCS#12 macros with real functions. Fix two unrelated ASN1 bugs:
5303 asn1_check_tlen() would sometimes attempt to use 'ctx' when it was
5304 NULL and ASN1_TYPE was not dereferenced properly in asn1_ex_c2i().
5305 New ASN1 macro: DECLARE_ASN1_ITEM() which just declares the relevant
5306 ASN1_ITEM and no wrapper functions.
5307 [Steve Henson]
5308
5309 *) New functions or ASN1_item_d2i_fp() and ASN1_item_d2i_bio(). These
5310 replace the old function pointer based I/O routines. Change most of
5311 the *_d2i_bio() and *_d2i_fp() functions to use these.
5312 [Steve Henson]
5313
5314 *) Enhance mkdef.pl to be more accepting about spacing in C preprocessor
5315 lines, recognice more "algorithms" that can be deselected, and make
5316 it complain about algorithm deselection that isn't recognised.
5317 [Richard Levitte]
5318
5319 *) New ASN1 functions to handle dup, sign, verify, digest, pack and
5320 unpack operations in terms of ASN1_ITEM. Modify existing wrappers
5321 to use new functions. Add NO_ASN1_OLD which can be set to remove
5322 some old style ASN1 functions: this can be used to determine if old
5323 code will still work when these eventually go away.
5324 [Steve Henson]
5325
5326 *) New extension functions for OCSP structures, these follow the
5327 same conventions as certificates and CRLs.
5328 [Steve Henson]
5329
5330 *) New function X509V3_add1_i2d(). This automatically encodes and
5331 adds an extension. Its behaviour can be customised with various
5332 flags to append, replace or delete. Various wrappers added for
5333 certifcates and CRLs.
5334 [Steve Henson]
5335
5336 *) Fix to avoid calling the underlying ASN1 print routine when
5337 an extension cannot be parsed. Correct a typo in the
5338 OCSP_SERVICELOC extension. Tidy up print OCSP format.
5339 [Steve Henson]
5340
5341 *) Make mkdef.pl parse some of the ASN1 macros and add apropriate
5342 entries for variables.
5343 [Steve Henson]
5344
5345 *) Add functionality to apps/openssl.c for detecting locking
5346 problems: As the program is single-threaded, all we have
5347 to do is register a locking callback using an array for
5348 storing which locks are currently held by the program.
5349 [Bodo Moeller]
5350
5351 *) Use a lock around the call to CRYPTO_get_ex_new_index() in
5352 SSL_get_ex_data_X509_STORE_idx(), which is used in
5353 ssl_verify_cert_chain() and thus can be called at any time
5354 during TLS/SSL handshakes so that thread-safety is essential.
5355 Unfortunately, the ex_data design is not at all suited
5356 for multi-threaded use, so it probably should be abolished.
5357 [Bodo Moeller]
5358
5359 *) Added Broadcom "ubsec" ENGINE to OpenSSL.
5360 [Broadcom, tweaked and integrated by Geoff Thorpe]
5361
5362 *) Move common extension printing code to new function
5363 X509V3_print_extensions(). Reorganise OCSP print routines and
5364 implement some needed OCSP ASN1 functions. Add OCSP extensions.
5365 [Steve Henson]
5366
5367 *) New function X509_signature_print() to remove duplication in some
5368 print routines.
5369 [Steve Henson]
5370
5371 *) Add a special meaning when SET OF and SEQUENCE OF flags are both
5372 set (this was treated exactly the same as SET OF previously). This
5373 is used to reorder the STACK representing the structure to match the
5374 encoding. This will be used to get round a problem where a PKCS7
5375 structure which was signed could not be verified because the STACK
5376 order did not reflect the encoded order.
5377 [Steve Henson]
5378
5379 *) Reimplement the OCSP ASN1 module using the new code.
5380 [Steve Henson]
5381
5382 *) Update the X509V3 code to permit the use of an ASN1_ITEM structure
5383 for its ASN1 operations. The old style function pointers still exist
5384 for now but they will eventually go away.
5385 [Steve Henson]
5386
5387 *) Merge in replacement ASN1 code from the ASN1 branch. This almost
5388 completely replaces the old ASN1 functionality with a table driven
5389 encoder and decoder which interprets an ASN1_ITEM structure describing
5390 the ASN1 module. Compatibility with the existing ASN1 API (i2d,d2i) is
5391 largely maintained. Almost all of the old asn1_mac.h macro based ASN1
5392 has also been converted to the new form.
5393 [Steve Henson]
5394
5395 *) Change BN_mod_exp_recp so that negative moduli are tolerated
5396 (the sign is ignored). Similarly, ignore the sign in BN_MONT_CTX_set
5397 so that BN_mod_exp_mont and BN_mod_exp_mont_word work
5398 for negative moduli.
5399 [Bodo Moeller]
5400
5401 *) Fix BN_uadd and BN_usub: Always return non-negative results instead
5402 of not touching the result's sign bit.
5403 [Bodo Moeller]
5404
5405 *) BN_div bugfix: If the result is 0, the sign (res->neg) must not be
5406 set.
5407 [Bodo Moeller]
5408
5409 *) Changed the LHASH code to use prototypes for callbacks, and created
5410 macros to declare and implement thin (optionally static) functions
5411 that provide type-safety and avoid function pointer casting for the
5412 type-specific callbacks.
5413 [Geoff Thorpe]
5414
5415 *) Added Kerberos Cipher Suites to be used with TLS, as written in
5416 RFC 2712.
5417 [Veers Staats <staatsvr@asc.hpc.mil>,
5418 Jeffrey Altman <jaltman@columbia.edu>, via Richard Levitte]
5419
5420 *) Reformat the FAQ so the different questions and answers can be divided
5421 in sections depending on the subject.
5422 [Richard Levitte]
5423
5424 *) Have the zlib compression code load ZLIB.DLL dynamically under
5425 Windows.
5426 [Richard Levitte]
5427
5428 *) New function BN_mod_sqrt for computing square roots modulo a prime
5429 (using the probabilistic Tonelli-Shanks algorithm unless
5430 p == 3 (mod 4) or p == 5 (mod 8), which are cases that can
5431 be handled deterministically).
5432 [Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller]
5433
5434 *) Make BN_mod_inverse faster by explicitly handling small quotients
5435 in the Euclid loop. (Speed gain about 20% for small moduli [256 or
5436 512 bits], about 30% for larger ones [1024 or 2048 bits].)
5437 [Bodo Moeller]
5438
5439 *) New function BN_kronecker.
5440 [Bodo Moeller]
5441
5442 *) Fix BN_gcd so that it works on negative inputs; the result is
5443 positive unless both parameters are zero.
5444 Previously something reasonably close to an infinite loop was
5445 possible because numbers could be growing instead of shrinking
5446 in the implementation of Euclid's algorithm.
5447 [Bodo Moeller]
5448
5449 *) Fix BN_is_word() and BN_is_one() macros to take into account the
5450 sign of the number in question.
5451
5452 Fix BN_is_word(a,w) to work correctly for w == 0.
5453
5454 The old BN_is_word(a,w) macro is now called BN_abs_is_word(a,w)
5455 because its test if the absolute value of 'a' equals 'w'.
5456 Note that BN_abs_is_word does *not* handle w == 0 reliably;
5457 it exists mostly for use in the implementations of BN_is_zero(),
5458 BN_is_one(), and BN_is_word().
5459 [Bodo Moeller]
5460
5461 *) New function BN_swap.
5462 [Bodo Moeller]
5463
5464 *) Use BN_nnmod instead of BN_mod in crypto/bn/bn_exp.c so that
5465 the exponentiation functions are more likely to produce reasonable
5466 results on negative inputs.
5467 [Bodo Moeller]
5468
5469 *) Change BN_mod_mul so that the result is always non-negative.
5470 Previously, it could be negative if one of the factors was negative;
5471 I don't think anyone really wanted that behaviour.
5472 [Bodo Moeller]
5473
5474 *) Move BN_mod_... functions into new file crypto/bn/bn_mod.c
5475 (except for exponentiation, which stays in crypto/bn/bn_exp.c,
5476 and BN_mod_mul_reciprocal, which stays in crypto/bn/bn_recp.c)
5477 and add new functions:
5478
5479 BN_nnmod
5480 BN_mod_sqr
5481 BN_mod_add
5482 BN_mod_add_quick
5483 BN_mod_sub
5484 BN_mod_sub_quick
5485 BN_mod_lshift1
5486 BN_mod_lshift1_quick
5487 BN_mod_lshift
5488 BN_mod_lshift_quick
5489
5490 These functions always generate non-negative results.
5491
5492 BN_nnmod otherwise is like BN_mod (if BN_mod computes a remainder r
5493 such that |m| < r < 0, BN_nnmod will output rem + |m| instead).
5494
5495 BN_mod_XXX_quick(r, a, [b,] m) generates the same result as
5496 BN_mod_XXX(r, a, [b,] m, ctx), but requires that a [and b]
5497 be reduced modulo m.
5498 [Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller]
5499
5500 #if 0
5501 The following entry accidentily appeared in the CHANGES file
5502 distributed with OpenSSL 0.9.7. The modifications described in
5503 it do *not* apply to OpenSSL 0.9.7.
5504
5505 *) Remove a few calls to bn_wexpand() in BN_sqr() (the one in there
5506 was actually never needed) and in BN_mul(). The removal in BN_mul()
5507 required a small change in bn_mul_part_recursive() and the addition
5508 of the functions bn_cmp_part_words(), bn_sub_part_words() and
5509 bn_add_part_words(), which do the same thing as bn_cmp_words(),
5510 bn_sub_words() and bn_add_words() except they take arrays with
5511 differing sizes.
5512 [Richard Levitte]
5513 #endif
5514
5515 *) In 'openssl passwd', verify passwords read from the terminal
5516 unless the '-salt' option is used (which usually means that
5517 verification would just waste user's time since the resulting
5518 hash is going to be compared with some given password hash)
5519 or the new '-noverify' option is used.
5520
5521 This is an incompatible change, but it does not affect
5522 non-interactive use of 'openssl passwd' (passwords on the command
5523 line, '-stdin' option, '-in ...' option) and thus should not
5524 cause any problems.
5525 [Bodo Moeller]
5526
5527 *) Remove all references to RSAref, since there's no more need for it.
5528 [Richard Levitte]
5529
5530 *) Make DSO load along a path given through an environment variable
5531 (SHLIB_PATH) with shl_load().
5532 [Richard Levitte]
5533
5534 *) Constify the ENGINE code as a result of BIGNUM constification.
5535 Also constify the RSA code and most things related to it. In a
5536 few places, most notable in the depth of the ASN.1 code, ugly
5537 casts back to non-const were required (to be solved at a later
5538 time)
5539 [Richard Levitte]
5540
5541 *) Make it so the openssl application has all engines loaded by default.
5542 [Richard Levitte]
5543
5544 *) Constify the BIGNUM routines a little more.
5545 [Richard Levitte]
5546
5547 *) Add the following functions:
5548
5549 ENGINE_load_cswift()
5550 ENGINE_load_chil()
5551 ENGINE_load_atalla()
5552 ENGINE_load_nuron()
5553 ENGINE_load_builtin_engines()
5554
5555 That way, an application can itself choose if external engines that
5556 are built-in in OpenSSL shall ever be used or not. The benefit is
5557 that applications won't have to be linked with libdl or other dso
5558 libraries unless it's really needed.
5559
5560 Changed 'openssl engine' to load all engines on demand.
5561 Changed the engine header files to avoid the duplication of some
5562 declarations (they differed!).
5563 [Richard Levitte]
5564
5565 *) 'openssl engine' can now list capabilities.
5566 [Richard Levitte]
5567
5568 *) Better error reporting in 'openssl engine'.
5569 [Richard Levitte]
5570
5571 *) Never call load_dh_param(NULL) in s_server.
5572 [Bodo Moeller]
5573
5574 *) Add engine application. It can currently list engines by name and
5575 identity, and test if they are actually available.
5576 [Richard Levitte]
5577
5578 *) Improve RPM specification file by forcing symbolic linking and making
5579 sure the installed documentation is also owned by root.root.
5580 [Damien Miller <djm@mindrot.org>]
5581
5582 *) Give the OpenSSL applications more possibilities to make use of
5583 keys (public as well as private) handled by engines.
5584 [Richard Levitte]
5585
5586 *) Add OCSP code that comes from CertCo.
5587 [Richard Levitte]
5588
5589 *) Add VMS support for the Rijndael code.
5590 [Richard Levitte]
5591
5592 *) Added untested support for Nuron crypto accelerator.
5593 [Ben Laurie]
5594
5595 *) Add support for external cryptographic devices. This code was
5596 previously distributed separately as the "engine" branch.
5597 [Geoff Thorpe, Richard Levitte]
5598
5599 *) Rework the filename-translation in the DSO code. It is now possible to
5600 have far greater control over how a "name" is turned into a filename
5601 depending on the operating environment and any oddities about the
5602 different shared library filenames on each system.
5603 [Geoff Thorpe]
5604
5605 *) Support threads on FreeBSD-elf in Configure.
5606 [Richard Levitte]
5607
5608 *) Fix for SHA1 assembly problem with MASM: it produces
5609 warnings about corrupt line number information when assembling
5610 with debugging information. This is caused by the overlapping
5611 of two sections.
5612 [Bernd Matthes <mainbug@celocom.de>, Steve Henson]
5613
5614 *) NCONF changes.
5615 NCONF_get_number() has no error checking at all. As a replacement,
5616 NCONF_get_number_e() is defined (_e for "error checking") and is
5617 promoted strongly. The old NCONF_get_number is kept around for
5618 binary backward compatibility.
5619 Make it possible for methods to load from something other than a BIO,
5620 by providing a function pointer that is given a name instead of a BIO.
5621 For example, this could be used to load configuration data from an
5622 LDAP server.
5623 [Richard Levitte]
5624
5625 *) Fix for non blocking accept BIOs. Added new I/O special reason
5626 BIO_RR_ACCEPT to cover this case. Previously use of accept BIOs
5627 with non blocking I/O was not possible because no retry code was
5628 implemented. Also added new SSL code SSL_WANT_ACCEPT to cover
5629 this case.
5630 [Steve Henson]
5631
5632 *) Added the beginnings of Rijndael support.
5633 [Ben Laurie]
5634
5635 *) Fix for bug in DirectoryString mask setting. Add support for
5636 X509_NAME_print_ex() in 'req' and X509_print_ex() function
5637 to allow certificate printing to more controllable, additional
5638 'certopt' option to 'x509' to allow new printing options to be
5639 set.
5640 [Steve Henson]
5641
5642 *) Clean old EAY MD5 hack from e_os.h.
5643 [Richard Levitte]
5644
5645 Changes between 0.9.6l and 0.9.6m [17 Mar 2004]
5646
5647 *) Fix null-pointer assignment in do_change_cipher_spec() revealed
5648 by using the Codenomicon TLS Test Tool (CVE-2004-0079)
5649 [Joe Orton, Steve Henson]
5650
5651 Changes between 0.9.6k and 0.9.6l [04 Nov 2003]
5652
5653 *) Fix additional bug revealed by the NISCC test suite:
5654
5655 Stop bug triggering large recursion when presented with
5656 certain ASN.1 tags (CVE-2003-0851)
5657 [Steve Henson]
5658
5659 Changes between 0.9.6j and 0.9.6k [30 Sep 2003]
5660
5661 *) Fix various bugs revealed by running the NISCC test suite:
5662
5663 Stop out of bounds reads in the ASN1 code when presented with
5664 invalid tags (CVE-2003-0543 and CVE-2003-0544).
5665
5666 If verify callback ignores invalid public key errors don't try to check
5667 certificate signature with the NULL public key.
5668
5669 [Steve Henson]
5670
5671 *) In ssl3_accept() (ssl/s3_srvr.c) only accept a client certificate
5672 if the server requested one: as stated in TLS 1.0 and SSL 3.0
5673 specifications.
5674 [Steve Henson]
5675
5676 *) In ssl3_get_client_hello() (ssl/s3_srvr.c), tolerate additional
5677 extra data after the compression methods not only for TLS 1.0
5678 but also for SSL 3.0 (as required by the specification).
5679 [Bodo Moeller; problem pointed out by Matthias Loepfe]
5680
5681 *) Change X509_certificate_type() to mark the key as exported/exportable
5682 when it's 512 *bits* long, not 512 bytes.
5683 [Richard Levitte]
5684
5685 Changes between 0.9.6i and 0.9.6j [10 Apr 2003]
5686
5687 *) Countermeasure against the Klima-Pokorny-Rosa extension of
5688 Bleichbacher's attack on PKCS #1 v1.5 padding: treat
5689 a protocol version number mismatch like a decryption error
5690 in ssl3_get_client_key_exchange (ssl/s3_srvr.c).
5691 [Bodo Moeller]
5692
5693 *) Turn on RSA blinding by default in the default implementation
5694 to avoid a timing attack. Applications that don't want it can call
5695 RSA_blinding_off() or use the new flag RSA_FLAG_NO_BLINDING.
5696 They would be ill-advised to do so in most cases.
5697 [Ben Laurie, Steve Henson, Geoff Thorpe, Bodo Moeller]
5698
5699 *) Change RSA blinding code so that it works when the PRNG is not
5700 seeded (in this case, the secret RSA exponent is abused as
5701 an unpredictable seed -- if it is not unpredictable, there
5702 is no point in blinding anyway). Make RSA blinding thread-safe
5703 by remembering the creator's thread ID in rsa->blinding and
5704 having all other threads use local one-time blinding factors
5705 (this requires more computation than sharing rsa->blinding, but
5706 avoids excessive locking; and if an RSA object is not shared
5707 between threads, blinding will still be very fast).
5708 [Bodo Moeller]
5709
5710 Changes between 0.9.6h and 0.9.6i [19 Feb 2003]
5711
5712 *) In ssl3_get_record (ssl/s3_pkt.c), minimize information leaked
5713 via timing by performing a MAC computation even if incorrrect
5714 block cipher padding has been found. This is a countermeasure
5715 against active attacks where the attacker has to distinguish
5716 between bad padding and a MAC verification error. (CVE-2003-0078)
5717
5718 [Bodo Moeller; problem pointed out by Brice Canvel (EPFL),
5719 Alain Hiltgen (UBS), Serge Vaudenay (EPFL), and
5720 Martin Vuagnoux (EPFL, Ilion)]
5721
5722 Changes between 0.9.6g and 0.9.6h [5 Dec 2002]
5723
5724 *) New function OPENSSL_cleanse(), which is used to cleanse a section of
5725 memory from it's contents. This is done with a counter that will
5726 place alternating values in each byte. This can be used to solve
5727 two issues: 1) the removal of calls to memset() by highly optimizing
5728 compilers, and 2) cleansing with other values than 0, since those can
5729 be read through on certain media, for example a swap space on disk.
5730 [Geoff Thorpe]
5731
5732 *) Bugfix: client side session caching did not work with external caching,
5733 because the session->cipher setting was not restored when reloading
5734 from the external cache. This problem was masked, when
5735 SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG (part of SSL_OP_ALL) was set.
5736 (Found by Steve Haslam <steve@araqnid.ddts.net>.)
5737 [Lutz Jaenicke]
5738
5739 *) Fix client_certificate (ssl/s2_clnt.c): The permissible total
5740 length of the REQUEST-CERTIFICATE message is 18 .. 34, not 17 .. 33.
5741 [Zeev Lieber <zeev-l@yahoo.com>]
5742
5743 *) Undo an undocumented change introduced in 0.9.6e which caused
5744 repeated calls to OpenSSL_add_all_ciphers() and
5745 OpenSSL_add_all_digests() to be ignored, even after calling
5746 EVP_cleanup().
5747 [Richard Levitte]
5748
5749 *) Change the default configuration reader to deal with last line not
5750 being properly terminated.
5751 [Richard Levitte]
5752
5753 *) Change X509_NAME_cmp() so it applies the special rules on handling
5754 DN values that are of type PrintableString, as well as RDNs of type
5755 emailAddress where the value has the type ia5String.
5756 [stefank@valicert.com via Richard Levitte]
5757
5758 *) Add a SSL_SESS_CACHE_NO_INTERNAL_STORE flag to take over half
5759 the job SSL_SESS_CACHE_NO_INTERNAL_LOOKUP was inconsistently
5760 doing, define a new flag (SSL_SESS_CACHE_NO_INTERNAL) to be
5761 the bitwise-OR of the two for use by the majority of applications
5762 wanting this behaviour, and update the docs. The documented
5763 behaviour and actual behaviour were inconsistent and had been
5764 changing anyway, so this is more a bug-fix than a behavioural
5765 change.
5766 [Geoff Thorpe, diagnosed by Nadav Har'El]
5767
5768 *) Don't impose a 16-byte length minimum on session IDs in ssl/s3_clnt.c
5769 (the SSL 3.0 and TLS 1.0 specifications allow any length up to 32 bytes).
5770 [Bodo Moeller]
5771
5772 *) Fix initialization code race conditions in
5773 SSLv23_method(), SSLv23_client_method(), SSLv23_server_method(),
5774 SSLv2_method(), SSLv2_client_method(), SSLv2_server_method(),
5775 SSLv3_method(), SSLv3_client_method(), SSLv3_server_method(),
5776 TLSv1_method(), TLSv1_client_method(), TLSv1_server_method(),
5777 ssl2_get_cipher_by_char(),
5778 ssl3_get_cipher_by_char().
5779 [Patrick McCormick <patrick@tellme.com>, Bodo Moeller]
5780
5781 *) Reorder cleanup sequence in SSL_CTX_free(): only remove the ex_data after
5782 the cached sessions are flushed, as the remove_cb() might use ex_data
5783 contents. Bug found by Sam Varshavchik <mrsam@courier-mta.com>
5784 (see [openssl.org #212]).
5785 [Geoff Thorpe, Lutz Jaenicke]
5786
5787 *) Fix typo in OBJ_txt2obj which incorrectly passed the content
5788 length, instead of the encoding length to d2i_ASN1_OBJECT.
5789 [Steve Henson]
5790
5791 Changes between 0.9.6f and 0.9.6g [9 Aug 2002]
5792
5793 *) [In 0.9.6g-engine release:]
5794 Fix crypto/engine/vendor_defns/cswift.h for WIN32 (use '_stdcall').
5795 [Lynn Gazis <lgazis@rainbow.com>]
5796
5797 Changes between 0.9.6e and 0.9.6f [8 Aug 2002]
5798
5799 *) Fix ASN1 checks. Check for overflow by comparing with LONG_MAX
5800 and get fix the header length calculation.
5801 [Florian Weimer <Weimer@CERT.Uni-Stuttgart.DE>,
5802 Alon Kantor <alonk@checkpoint.com> (and others),
5803 Steve Henson]
5804
5805 *) Use proper error handling instead of 'assertions' in buffer
5806 overflow checks added in 0.9.6e. This prevents DoS (the
5807 assertions could call abort()).
5808 [Arne Ansper <arne@ats.cyber.ee>, Bodo Moeller]
5809
5810 Changes between 0.9.6d and 0.9.6e [30 Jul 2002]
5811
5812 *) Add various sanity checks to asn1_get_length() to reject
5813 the ASN1 length bytes if they exceed sizeof(long), will appear
5814 negative or the content length exceeds the length of the
5815 supplied buffer.
5816 [Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>]
5817
5818 *) Fix cipher selection routines: ciphers without encryption had no flags
5819 for the cipher strength set and where therefore not handled correctly
5820 by the selection routines (PR #130).
5821 [Lutz Jaenicke]
5822
5823 *) Fix EVP_dsa_sha macro.
5824 [Nils Larsch]
5825
5826 *) New option
5827 SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS
5828 for disabling the SSL 3.0/TLS 1.0 CBC vulnerability countermeasure
5829 that was added in OpenSSL 0.9.6d.
5830
5831 As the countermeasure turned out to be incompatible with some
5832 broken SSL implementations, the new option is part of SSL_OP_ALL.
5833 SSL_OP_ALL is usually employed when compatibility with weird SSL
5834 implementations is desired (e.g. '-bugs' option to 's_client' and
5835 's_server'), so the new option is automatically set in many
5836 applications.
5837 [Bodo Moeller]
5838
5839 *) Changes in security patch:
5840
5841 Changes marked "(CHATS)" were sponsored by the Defense Advanced
5842 Research Projects Agency (DARPA) and Air Force Research Laboratory,
5843 Air Force Materiel Command, USAF, under agreement number
5844 F30602-01-2-0537.
5845
5846 *) Add various sanity checks to asn1_get_length() to reject
5847 the ASN1 length bytes if they exceed sizeof(long), will appear
5848 negative or the content length exceeds the length of the
5849 supplied buffer. (CVE-2002-0659)
5850 [Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>]
5851
5852 *) Assertions for various potential buffer overflows, not known to
5853 happen in practice.
5854 [Ben Laurie (CHATS)]
5855
5856 *) Various temporary buffers to hold ASCII versions of integers were
5857 too small for 64 bit platforms. (CVE-2002-0655)
5858 [Matthew Byng-Maddick <mbm@aldigital.co.uk> and Ben Laurie (CHATS)>
5859
5860 *) Remote buffer overflow in SSL3 protocol - an attacker could
5861 supply an oversized session ID to a client. (CVE-2002-0656)
5862 [Ben Laurie (CHATS)]
5863
5864 *) Remote buffer overflow in SSL2 protocol - an attacker could
5865 supply an oversized client master key. (CVE-2002-0656)
5866 [Ben Laurie (CHATS)]
5867
5868 Changes between 0.9.6c and 0.9.6d [9 May 2002]
5869
5870 *) Fix crypto/asn1/a_sign.c so that 'parameters' is omitted (not
5871 encoded as NULL) with id-dsa-with-sha1.
5872 [Nils Larsch <nla@trustcenter.de>; problem pointed out by Bodo Moeller]
5873
5874 *) Check various X509_...() return values in apps/req.c.
5875 [Nils Larsch <nla@trustcenter.de>]
5876
5877 *) Fix BASE64 decode (EVP_DecodeUpdate) for data with CR/LF ended lines:
5878 an end-of-file condition would erronously be flagged, when the CRLF
5879 was just at the end of a processed block. The bug was discovered when
5880 processing data through a buffering memory BIO handing the data to a
5881 BASE64-decoding BIO. Bug fund and patch submitted by Pavel Tsekov
5882 <ptsekov@syntrex.com> and Nedelcho Stanev.
5883 [Lutz Jaenicke]
5884
5885 *) Implement a countermeasure against a vulnerability recently found
5886 in CBC ciphersuites in SSL 3.0/TLS 1.0: Send an empty fragment
5887 before application data chunks to avoid the use of known IVs
5888 with data potentially chosen by the attacker.
5889 [Bodo Moeller]
5890
5891 *) Fix length checks in ssl3_get_client_hello().
5892 [Bodo Moeller]
5893
5894 *) TLS/SSL library bugfix: use s->s3->in_read_app_data differently
5895 to prevent ssl3_read_internal() from incorrectly assuming that
5896 ssl3_read_bytes() found application data while handshake
5897 processing was enabled when in fact s->s3->in_read_app_data was
5898 merely automatically cleared during the initial handshake.
5899 [Bodo Moeller; problem pointed out by Arne Ansper <arne@ats.cyber.ee>]
5900
5901 *) Fix object definitions for Private and Enterprise: they were not
5902 recognized in their shortname (=lowercase) representation. Extend
5903 obj_dat.pl to issue an error when using undefined keywords instead
5904 of silently ignoring the problem (Svenning Sorensen
5905 <sss@sss.dnsalias.net>).
5906 [Lutz Jaenicke]
5907
5908 *) Fix DH_generate_parameters() so that it works for 'non-standard'
5909 generators, i.e. generators other than 2 and 5. (Previously, the
5910 code did not properly initialise the 'add' and 'rem' values to
5911 BN_generate_prime().)
5912
5913 In the new general case, we do not insist that 'generator' is
5914 actually a primitive root: This requirement is rather pointless;
5915 a generator of the order-q subgroup is just as good, if not
5916 better.
5917 [Bodo Moeller]
5918
5919 *) Map new X509 verification errors to alerts. Discovered and submitted by
5920 Tom Wu <tom@arcot.com>.
5921 [Lutz Jaenicke]
5922
5923 *) Fix ssl3_pending() (ssl/s3_lib.c) to prevent SSL_pending() from
5924 returning non-zero before the data has been completely received
5925 when using non-blocking I/O.
5926 [Bodo Moeller; problem pointed out by John Hughes]
5927
5928 *) Some of the ciphers missed the strength entry (SSL_LOW etc).
5929 [Ben Laurie, Lutz Jaenicke]
5930
5931 *) Fix bug in SSL_clear(): bad sessions were not removed (found by
5932 Yoram Zahavi <YoramZ@gilian.com>).
5933 [Lutz Jaenicke]
5934
5935 *) Add information about CygWin 1.3 and on, and preserve proper
5936 configuration for the versions before that.
5937 [Corinna Vinschen <vinschen@redhat.com> and Richard Levitte]
5938
5939 *) Make removal from session cache (SSL_CTX_remove_session()) more robust:
5940 check whether we deal with a copy of a session and do not delete from
5941 the cache in this case. Problem reported by "Izhar Shoshani Levi"
5942 <izhar@checkpoint.com>.
5943 [Lutz Jaenicke]
5944
5945 *) Do not store session data into the internal session cache, if it
5946 is never intended to be looked up (SSL_SESS_CACHE_NO_INTERNAL_LOOKUP
5947 flag is set). Proposed by Aslam <aslam@funk.com>.
5948 [Lutz Jaenicke]
5949
5950 *) Have ASN1_BIT_STRING_set_bit() really clear a bit when the requested
5951 value is 0.
5952 [Richard Levitte]
5953
5954 *) [In 0.9.6d-engine release:]
5955 Fix a crashbug and a logic bug in hwcrhk_load_pubkey().
5956 [Toomas Kiisk <vix@cyber.ee> via Richard Levitte]
5957
5958 *) Add the configuration target linux-s390x.
5959 [Neale Ferguson <Neale.Ferguson@SoftwareAG-USA.com> via Richard Levitte]
5960
5961 *) The earlier bugfix for the SSL3_ST_SW_HELLO_REQ_C case of
5962 ssl3_accept (ssl/s3_srvr.c) incorrectly used a local flag
5963 variable as an indication that a ClientHello message has been
5964 received. As the flag value will be lost between multiple
5965 invocations of ssl3_accept when using non-blocking I/O, the
5966 function may not be aware that a handshake has actually taken
5967 place, thus preventing a new session from being added to the
5968 session cache.
5969
5970 To avoid this problem, we now set s->new_session to 2 instead of
5971 using a local variable.
5972 [Lutz Jaenicke, Bodo Moeller]
5973
5974 *) Bugfix: Return -1 from ssl3_get_server_done (ssl3/s3_clnt.c)
5975 if the SSL_R_LENGTH_MISMATCH error is detected.
5976 [Geoff Thorpe, Bodo Moeller]
5977
5978 *) New 'shared_ldflag' column in Configure platform table.
5979 [Richard Levitte]
5980
5981 *) Fix EVP_CIPHER_mode macro.
5982 ["Dan S. Camper" <dan@bti.net>]
5983
5984 *) Fix ssl3_read_bytes (ssl/s3_pkt.c): To ignore messages of unknown
5985 type, we must throw them away by setting rr->length to 0.
5986 [D P Chang <dpc@qualys.com>]
5987
5988 Changes between 0.9.6b and 0.9.6c [21 dec 2001]
5989
5990 *) Fix BN_rand_range bug pointed out by Dominikus Scherkl
5991 <Dominikus.Scherkl@biodata.com>. (The previous implementation
5992 worked incorrectly for those cases where range = 10..._2 and
5993 3*range is two bits longer than range.)
5994 [Bodo Moeller]
5995
5996 *) Only add signing time to PKCS7 structures if it is not already
5997 present.
5998 [Steve Henson]
5999
6000 *) Fix crypto/objects/objects.h: "ld-ce" should be "id-ce",
6001 OBJ_ld_ce should be OBJ_id_ce.
6002 Also some ip-pda OIDs in crypto/objects/objects.txt were
6003 incorrect (cf. RFC 3039).
6004 [Matt Cooper, Frederic Giudicelli, Bodo Moeller]
6005
6006 *) Release CRYPTO_LOCK_DYNLOCK when CRYPTO_destroy_dynlockid()
6007 returns early because it has nothing to do.
6008 [Andy Schneider <andy.schneider@bjss.co.uk>]
6009
6010 *) [In 0.9.6c-engine release:]
6011 Fix mutex callback return values in crypto/engine/hw_ncipher.c.
6012 [Andy Schneider <andy.schneider@bjss.co.uk>]
6013
6014 *) [In 0.9.6c-engine release:]
6015 Add support for Cryptographic Appliance's keyserver technology.
6016 (Use engine 'keyclient')
6017 [Cryptographic Appliances and Geoff Thorpe]
6018
6019 *) Add a configuration entry for OS/390 Unix. The C compiler 'c89'
6020 is called via tools/c89.sh because arguments have to be
6021 rearranged (all '-L' options must appear before the first object
6022 modules).
6023 [Richard Shapiro <rshapiro@abinitio.com>]
6024
6025 *) [In 0.9.6c-engine release:]
6026 Add support for Broadcom crypto accelerator cards, backported
6027 from 0.9.7.
6028 [Broadcom, Nalin Dahyabhai <nalin@redhat.com>, Mark Cox]
6029
6030 *) [In 0.9.6c-engine release:]
6031 Add support for SureWare crypto accelerator cards from
6032 Baltimore Technologies. (Use engine 'sureware')
6033 [Baltimore Technologies and Mark Cox]
6034
6035 *) [In 0.9.6c-engine release:]
6036 Add support for crypto accelerator cards from Accelerated
6037 Encryption Processing, www.aep.ie. (Use engine 'aep')
6038 [AEP Inc. and Mark Cox]
6039
6040 *) Add a configuration entry for gcc on UnixWare.
6041 [Gary Benson <gbenson@redhat.com>]
6042
6043 *) Change ssl/s2_clnt.c and ssl/s2_srvr.c so that received handshake
6044 messages are stored in a single piece (fixed-length part and
6045 variable-length part combined) and fix various bugs found on the way.
6046 [Bodo Moeller]
6047
6048 *) Disable caching in BIO_gethostbyname(), directly use gethostbyname()
6049 instead. BIO_gethostbyname() does not know what timeouts are
6050 appropriate, so entries would stay in cache even when they have
6051 become invalid.
6052 [Bodo Moeller; problem pointed out by Rich Salz <rsalz@zolera.com>
6053
6054 *) Change ssl23_get_client_hello (ssl/s23_srvr.c) behaviour when
6055 faced with a pathologically small ClientHello fragment that does
6056 not contain client_version: Instead of aborting with an error,
6057 simply choose the highest available protocol version (i.e.,
6058 TLS 1.0 unless it is disabled). In practice, ClientHello
6059 messages are never sent like this, but this change gives us
6060 strictly correct behaviour at least for TLS.
6061 [Bodo Moeller]
6062
6063 *) Fix SSL handshake functions and SSL_clear() such that SSL_clear()
6064 never resets s->method to s->ctx->method when called from within
6065 one of the SSL handshake functions.
6066 [Bodo Moeller; problem pointed out by Niko Baric]
6067
6068 *) In ssl3_get_client_hello (ssl/s3_srvr.c), generate a fatal alert
6069 (sent using the client's version number) if client_version is
6070 smaller than the protocol version in use. Also change
6071 ssl23_get_client_hello (ssl/s23_srvr.c) to select TLS 1.0 if
6072 the client demanded SSL 3.0 but only TLS 1.0 is enabled; then
6073 the client will at least see that alert.
6074 [Bodo Moeller]
6075
6076 *) Fix ssl3_get_message (ssl/s3_both.c) to handle message fragmentation
6077 correctly.
6078 [Bodo Moeller]
6079
6080 *) Avoid infinite loop in ssl3_get_message (ssl/s3_both.c) if a
6081 client receives HelloRequest while in a handshake.
6082 [Bodo Moeller; bug noticed by Andy Schneider <andy.schneider@bjss.co.uk>]
6083
6084 *) Bugfix in ssl3_accept (ssl/s3_srvr.c): Case SSL3_ST_SW_HELLO_REQ_C
6085 should end in 'break', not 'goto end' which circuments various
6086 cleanups done in state SSL_ST_OK. But session related stuff
6087 must be disabled for SSL_ST_OK in the case that we just sent a
6088 HelloRequest.
6089
6090 Also avoid some overhead by not calling ssl_init_wbio_buffer()
6091 before just sending a HelloRequest.
6092 [Bodo Moeller, Eric Rescorla <ekr@rtfm.com>]
6093
6094 *) Fix ssl/s3_enc.c, ssl/t1_enc.c and ssl/s3_pkt.c so that we don't
6095 reveal whether illegal block cipher padding was found or a MAC
6096 verification error occured. (Neither SSLerr() codes nor alerts
6097 are directly visible to potential attackers, but the information
6098 may leak via logfiles.)
6099
6100 Similar changes are not required for the SSL 2.0 implementation
6101 because the number of padding bytes is sent in clear for SSL 2.0,
6102 and the extra bytes are just ignored. However ssl/s2_pkt.c
6103 failed to verify that the purported number of padding bytes is in
6104 the legal range.
6105 [Bodo Moeller]
6106
6107 *) Add OpenUNIX-8 support including shared libraries
6108 (Boyd Lynn Gerber <gerberb@zenez.com>).
6109 [Lutz Jaenicke]
6110
6111 *) Improve RSA_padding_check_PKCS1_OAEP() check again to avoid
6112 'wristwatch attack' using huge encoding parameters (cf.
6113 James H. Manger's CRYPTO 2001 paper). Note that the
6114 RSA_PKCS1_OAEP_PADDING case of RSA_private_decrypt() does not use
6115 encoding parameters and hence was not vulnerable.
6116 [Bodo Moeller]
6117
6118 *) BN_sqr() bug fix.
6119 [Ulf Möller, reported by Jim Ellis <jim.ellis@cavium.com>]
6120
6121 *) Rabin-Miller test analyses assume uniformly distributed witnesses,
6122 so use BN_pseudo_rand_range() instead of using BN_pseudo_rand()
6123 followed by modular reduction.
6124 [Bodo Moeller; pointed out by Adam Young <AYoung1@NCSUS.JNJ.COM>]
6125
6126 *) Add BN_pseudo_rand_range() with obvious functionality: BN_rand_range()
6127 equivalent based on BN_pseudo_rand() instead of BN_rand().
6128 [Bodo Moeller]
6129
6130 *) s3_srvr.c: allow sending of large client certificate lists (> 16 kB).
6131 This function was broken, as the check for a new client hello message
6132 to handle SGC did not allow these large messages.
6133 (Tracked down by "Douglas E. Engert" <deengert@anl.gov>.)
6134 [Lutz Jaenicke]
6135
6136 *) Add alert descriptions for TLSv1 to SSL_alert_desc_string[_long]().
6137 [Lutz Jaenicke]
6138
6139 *) Fix buggy behaviour of BIO_get_num_renegotiates() and BIO_ctrl()
6140 for BIO_C_GET_WRITE_BUF_SIZE ("Stephen Hinton" <shinton@netopia.com>).
6141 [Lutz Jaenicke]
6142
6143 *) Rework the configuration and shared library support for Tru64 Unix.
6144 The configuration part makes use of modern compiler features and
6145 still retains old compiler behavior for those that run older versions
6146 of the OS. The shared library support part includes a variant that
6147 uses the RPATH feature, and is available through the special
6148 configuration target "alpha-cc-rpath", which will never be selected
6149 automatically.
6150 [Tim Mooney <mooney@dogbert.cc.ndsu.NoDak.edu> via Richard Levitte]
6151
6152 *) In ssl3_get_key_exchange (ssl/s3_clnt.c), call ssl3_get_message()
6153 with the same message size as in ssl3_get_certificate_request().
6154 Otherwise, if no ServerKeyExchange message occurs, CertificateRequest
6155 messages might inadvertently be reject as too long.
6156 [Petr Lampa <lampa@fee.vutbr.cz>]
6157
6158 *) Enhanced support for IA-64 Unix platforms (well, Linux and HP-UX).
6159 [Andy Polyakov]
6160
6161 *) Modified SSL library such that the verify_callback that has been set
6162 specificly for an SSL object with SSL_set_verify() is actually being
6163 used. Before the change, a verify_callback set with this function was
6164 ignored and the verify_callback() set in the SSL_CTX at the time of
6165 the call was used. New function X509_STORE_CTX_set_verify_cb() introduced
6166 to allow the necessary settings.
6167 [Lutz Jaenicke]
6168
6169 *) Initialize static variable in crypto/dsa/dsa_lib.c and crypto/dh/dh_lib.c
6170 explicitly to NULL, as at least on Solaris 8 this seems not always to be
6171 done automatically (in contradiction to the requirements of the C
6172 standard). This made problems when used from OpenSSH.
6173 [Lutz Jaenicke]
6174
6175 *) In OpenSSL 0.9.6a and 0.9.6b, crypto/dh/dh_key.c ignored
6176 dh->length and always used
6177
6178 BN_rand_range(priv_key, dh->p).
6179
6180 BN_rand_range() is not necessary for Diffie-Hellman, and this
6181 specific range makes Diffie-Hellman unnecessarily inefficient if
6182 dh->length (recommended exponent length) is much smaller than the
6183 length of dh->p. We could use BN_rand_range() if the order of
6184 the subgroup was stored in the DH structure, but we only have
6185 dh->length.
6186
6187 So switch back to
6188
6189 BN_rand(priv_key, l, ...)
6190
6191 where 'l' is dh->length if this is defined, or BN_num_bits(dh->p)-1
6192 otherwise.
6193 [Bodo Moeller]
6194
6195 *) In
6196
6197 RSA_eay_public_encrypt
6198 RSA_eay_private_decrypt
6199 RSA_eay_private_encrypt (signing)
6200 RSA_eay_public_decrypt (signature verification)
6201
6202 (default implementations for RSA_public_encrypt,
6203 RSA_private_decrypt, RSA_private_encrypt, RSA_public_decrypt),
6204 always reject numbers >= n.
6205 [Bodo Moeller]
6206
6207 *) In crypto/rand/md_rand.c, use a new short-time lock CRYPTO_LOCK_RAND2
6208 to synchronize access to 'locking_thread'. This is necessary on
6209 systems where access to 'locking_thread' (an 'unsigned long'
6210 variable) is not atomic.
6211 [Bodo Moeller]
6212
6213 *) In crypto/rand/md_rand.c, set 'locking_thread' to current thread's ID
6214 *before* setting the 'crypto_lock_rand' flag. The previous code had
6215 a race condition if 0 is a valid thread ID.
6216 [Travis Vitek <vitek@roguewave.com>]
6217
6218 *) Add support for shared libraries under Irix.
6219 [Albert Chin-A-Young <china@thewrittenword.com>]
6220
6221 *) Add configuration option to build on Linux on both big-endian and
6222 little-endian MIPS.
6223 [Ralf Baechle <ralf@uni-koblenz.de>]
6224
6225 *) Add the possibility to create shared libraries on HP-UX.
6226 [Richard Levitte]
6227
6228 Changes between 0.9.6a and 0.9.6b [9 Jul 2001]
6229
6230 *) Change ssleay_rand_bytes (crypto/rand/md_rand.c)
6231 to avoid a SSLeay/OpenSSL PRNG weakness pointed out by
6232 Markku-Juhani O. Saarinen <markku-juhani.saarinen@nokia.com>:
6233 PRNG state recovery was possible based on the output of
6234 one PRNG request appropriately sized to gain knowledge on
6235 'md' followed by enough consecutive 1-byte PRNG requests
6236 to traverse all of 'state'.
6237
6238 1. When updating 'md_local' (the current thread's copy of 'md')
6239 during PRNG output generation, hash all of the previous
6240 'md_local' value, not just the half used for PRNG output.
6241
6242 2. Make the number of bytes from 'state' included into the hash
6243 independent from the number of PRNG bytes requested.
6244
6245 The first measure alone would be sufficient to avoid
6246 Markku-Juhani's attack. (Actually it had never occurred
6247 to me that the half of 'md_local' used for chaining was the
6248 half from which PRNG output bytes were taken -- I had always
6249 assumed that the secret half would be used.) The second
6250 measure makes sure that additional data from 'state' is never
6251 mixed into 'md_local' in small portions; this heuristically
6252 further strengthens the PRNG.
6253 [Bodo Moeller]
6254
6255 *) Fix crypto/bn/asm/mips3.s.
6256 [Andy Polyakov]
6257
6258 *) When only the key is given to "enc", the IV is undefined. Print out
6259 an error message in this case.
6260 [Lutz Jaenicke]
6261
6262 *) Handle special case when X509_NAME is empty in X509 printing routines.
6263 [Steve Henson]
6264
6265 *) In dsa_do_verify (crypto/dsa/dsa_ossl.c), verify that r and s are
6266 positive and less than q.
6267 [Bodo Moeller]
6268
6269 *) Don't change *pointer in CRYPTO_add_lock() is add_lock_callback is
6270 used: it isn't thread safe and the add_lock_callback should handle
6271 that itself.
6272 [Paul Rose <Paul.Rose@bridge.com>]
6273
6274 *) Verify that incoming data obeys the block size in
6275 ssl3_enc (ssl/s3_enc.c) and tls1_enc (ssl/t1_enc.c).
6276 [Bodo Moeller]
6277
6278 *) Fix OAEP check.
6279 [Ulf Möller, Bodo Möller]
6280
6281 *) The countermeasure against Bleichbacher's attack on PKCS #1 v1.5
6282 RSA encryption was accidentally removed in s3_srvr.c in OpenSSL 0.9.5
6283 when fixing the server behaviour for backwards-compatible 'client
6284 hello' messages. (Note that the attack is impractical against
6285 SSL 3.0 and TLS 1.0 anyway because length and version checking
6286 means that the probability of guessing a valid ciphertext is
6287 around 2^-40; see section 5 in Bleichenbacher's CRYPTO '98
6288 paper.)
6289
6290 Before 0.9.5, the countermeasure (hide the error by generating a
6291 random 'decryption result') did not work properly because
6292 ERR_clear_error() was missing, meaning that SSL_get_error() would
6293 detect the supposedly ignored error.
6294
6295 Both problems are now fixed.
6296 [Bodo Moeller]
6297
6298 *) In crypto/bio/bf_buff.c, increase DEFAULT_BUFFER_SIZE to 4096
6299 (previously it was 1024).
6300 [Bodo Moeller]
6301
6302 *) Fix for compatibility mode trust settings: ignore trust settings
6303 unless some valid trust or reject settings are present.
6304 [Steve Henson]
6305
6306 *) Fix for blowfish EVP: its a variable length cipher.
6307 [Steve Henson]
6308
6309 *) Fix various bugs related to DSA S/MIME verification. Handle missing
6310 parameters in DSA public key structures and return an error in the
6311 DSA routines if parameters are absent.
6312 [Steve Henson]
6313
6314 *) In versions up to 0.9.6, RAND_file_name() resorted to file ".rnd"
6315 in the current directory if neither $RANDFILE nor $HOME was set.
6316 RAND_file_name() in 0.9.6a returned NULL in this case. This has
6317 caused some confusion to Windows users who haven't defined $HOME.
6318 Thus RAND_file_name() is changed again: e_os.h can define a
6319 DEFAULT_HOME, which will be used if $HOME is not set.
6320 For Windows, we use "C:"; on other platforms, we still require
6321 environment variables.
6322
6323 *) Move 'if (!initialized) RAND_poll()' into regions protected by
6324 CRYPTO_LOCK_RAND. This is not strictly necessary, but avoids
6325 having multiple threads call RAND_poll() concurrently.
6326 [Bodo Moeller]
6327
6328 *) In crypto/rand/md_rand.c, replace 'add_do_not_lock' flag by a
6329 combination of a flag and a thread ID variable.
6330 Otherwise while one thread is in ssleay_rand_bytes (which sets the
6331 flag), *other* threads can enter ssleay_add_bytes without obeying
6332 the CRYPTO_LOCK_RAND lock (and may even illegally release the lock
6333 that they do not hold after the first thread unsets add_do_not_lock).
6334 [Bodo Moeller]
6335
6336 *) Change bctest again: '-x' expressions are not available in all
6337 versions of 'test'.
6338 [Bodo Moeller]
6339
6340 Changes between 0.9.6 and 0.9.6a [5 Apr 2001]
6341
6342 *) Fix a couple of memory leaks in PKCS7_dataDecode()
6343 [Steve Henson, reported by Heyun Zheng <hzheng@atdsprint.com>]
6344
6345 *) Change Configure and Makefiles to provide EXE_EXT, which will contain
6346 the default extension for executables, if any. Also, make the perl
6347 scripts that use symlink() to test if it really exists and use "cp"
6348 if it doesn't. All this made OpenSSL compilable and installable in
6349 CygWin.
6350 [Richard Levitte]
6351
6352 *) Fix for asn1_GetSequence() for indefinite length constructed data.
6353 If SEQUENCE is length is indefinite just set c->slen to the total
6354 amount of data available.
6355 [Steve Henson, reported by shige@FreeBSD.org]
6356 [This change does not apply to 0.9.7.]
6357
6358 *) Change bctest to avoid here-documents inside command substitution
6359 (workaround for FreeBSD /bin/sh bug).
6360 For compatibility with Ultrix, avoid shell functions (introduced
6361 in the bctest version that searches along $PATH).
6362 [Bodo Moeller]
6363
6364 *) Rename 'des_encrypt' to 'des_encrypt1'. This avoids the clashes
6365 with des_encrypt() defined on some operating systems, like Solaris
6366 and UnixWare.
6367 [Richard Levitte]
6368
6369 *) Check the result of RSA-CRT (see D. Boneh, R. DeMillo, R. Lipton:
6370 On the Importance of Eliminating Errors in Cryptographic
6371 Computations, J. Cryptology 14 (2001) 2, 101-119,
6372 http://theory.stanford.edu/~dabo/papers/faults.ps.gz).
6373 [Ulf Moeller]
6374
6375 *) MIPS assembler BIGNUM division bug fix.
6376 [Andy Polyakov]
6377
6378 *) Disabled incorrect Alpha assembler code.
6379 [Richard Levitte]
6380
6381 *) Fix PKCS#7 decode routines so they correctly update the length
6382 after reading an EOC for the EXPLICIT tag.
6383 [Steve Henson]
6384 [This change does not apply to 0.9.7.]
6385
6386 *) Fix bug in PKCS#12 key generation routines. This was triggered
6387 if a 3DES key was generated with a 0 initial byte. Include
6388 PKCS12_BROKEN_KEYGEN compilation option to retain the old
6389 (but broken) behaviour.
6390 [Steve Henson]
6391
6392 *) Enhance bctest to search for a working bc along $PATH and print
6393 it when found.
6394 [Tim Rice <tim@multitalents.net> via Richard Levitte]
6395
6396 *) Fix memory leaks in err.c: free err_data string if necessary;
6397 don't write to the wrong index in ERR_set_error_data.
6398 [Bodo Moeller]
6399
6400 *) Implement ssl23_peek (analogous to ssl23_read), which previously
6401 did not exist.
6402 [Bodo Moeller]
6403
6404 *) Replace rdtsc with _emit statements for VC++ version 5.
6405 [Jeremy Cooper <jeremy@baymoo.org>]
6406
6407 *) Make it possible to reuse SSLv2 sessions.
6408 [Richard Levitte]
6409
6410 *) In copy_email() check for >= 0 as a return value for
6411 X509_NAME_get_index_by_NID() since 0 is a valid index.
6412 [Steve Henson reported by Massimiliano Pala <madwolf@opensca.org>]
6413
6414 *) Avoid coredump with unsupported or invalid public keys by checking if
6415 X509_get_pubkey() fails in PKCS7_verify(). Fix memory leak when
6416 PKCS7_verify() fails with non detached data.
6417 [Steve Henson]
6418
6419 *) Don't use getenv in library functions when run as setuid/setgid.
6420 New function OPENSSL_issetugid().
6421 [Ulf Moeller]
6422
6423 *) Avoid false positives in memory leak detection code (crypto/mem_dbg.c)
6424 due to incorrect handling of multi-threading:
6425
6426 1. Fix timing glitch in the MemCheck_off() portion of CRYPTO_mem_ctrl().
6427
6428 2. Fix logical glitch in is_MemCheck_on() aka CRYPTO_is_mem_check_on().
6429
6430 3. Count how many times MemCheck_off() has been called so that
6431 nested use can be treated correctly. This also avoids
6432 inband-signalling in the previous code (which relied on the
6433 assumption that thread ID 0 is impossible).
6434 [Bodo Moeller]
6435
6436 *) Add "-rand" option also to s_client and s_server.
6437 [Lutz Jaenicke]
6438
6439 *) Fix CPU detection on Irix 6.x.
6440 [Kurt Hockenbury <khockenb@stevens-tech.edu> and
6441 "Bruce W. Forsberg" <bruce.forsberg@baesystems.com>]
6442
6443 *) Fix X509_NAME bug which produced incorrect encoding if X509_NAME
6444 was empty.
6445 [Steve Henson]
6446 [This change does not apply to 0.9.7.]
6447
6448 *) Use the cached encoding of an X509_NAME structure rather than
6449 copying it. This is apparently the reason for the libsafe "errors"
6450 but the code is actually correct.
6451 [Steve Henson]
6452
6453 *) Add new function BN_rand_range(), and fix DSA_sign_setup() to prevent
6454 Bleichenbacher's DSA attack.
6455 Extend BN_[pseudo_]rand: As before, top=1 forces the highest two bits
6456 to be set and top=0 forces the highest bit to be set; top=-1 is new
6457 and leaves the highest bit random.
6458 [Ulf Moeller, Bodo Moeller]
6459
6460 *) In the NCONF_...-based implementations for CONF_... queries
6461 (crypto/conf/conf_lib.c), if the input LHASH is NULL, avoid using
6462 a temporary CONF structure with the data component set to NULL
6463 (which gives segmentation faults in lh_retrieve).
6464 Instead, use NULL for the CONF pointer in CONF_get_string and
6465 CONF_get_number (which may use environment variables) and directly
6466 return NULL from CONF_get_section.
6467 [Bodo Moeller]
6468
6469 *) Fix potential buffer overrun for EBCDIC.
6470 [Ulf Moeller]
6471
6472 *) Tolerate nonRepudiation as being valid for S/MIME signing and certSign
6473 keyUsage if basicConstraints absent for a CA.
6474 [Steve Henson]
6475
6476 *) Make SMIME_write_PKCS7() write mail header values with a format that
6477 is more generally accepted (no spaces before the semicolon), since
6478 some programs can't parse those values properly otherwise. Also make
6479 sure BIO's that break lines after each write do not create invalid
6480 headers.
6481 [Richard Levitte]
6482
6483 *) Make the CRL encoding routines work with empty SEQUENCE OF. The
6484 macros previously used would not encode an empty SEQUENCE OF
6485 and break the signature.
6486 [Steve Henson]
6487 [This change does not apply to 0.9.7.]
6488
6489 *) Zero the premaster secret after deriving the master secret in
6490 DH ciphersuites.
6491 [Steve Henson]
6492
6493 *) Add some EVP_add_digest_alias registrations (as found in
6494 OpenSSL_add_all_digests()) to SSL_library_init()
6495 aka OpenSSL_add_ssl_algorithms(). This provides improved
6496 compatibility with peers using X.509 certificates
6497 with unconventional AlgorithmIdentifier OIDs.
6498 [Bodo Moeller]
6499
6500 *) Fix for Irix with NO_ASM.
6501 ["Bruce W. Forsberg" <bruce.forsberg@baesystems.com>]
6502
6503 *) ./config script fixes.
6504 [Ulf Moeller, Richard Levitte]
6505
6506 *) Fix 'openssl passwd -1'.
6507 [Bodo Moeller]
6508
6509 *) Change PKCS12_key_gen_asc() so it can cope with non null
6510 terminated strings whose length is passed in the passlen
6511 parameter, for example from PEM callbacks. This was done
6512 by adding an extra length parameter to asc2uni().
6513 [Steve Henson, reported by <oddissey@samsung.co.kr>]
6514
6515 *) Fix C code generated by 'openssl dsaparam -C': If a BN_bin2bn
6516 call failed, free the DSA structure.
6517 [Bodo Moeller]
6518
6519 *) Fix to uni2asc() to cope with zero length Unicode strings.
6520 These are present in some PKCS#12 files.
6521 [Steve Henson]
6522
6523 *) Increase s2->wbuf allocation by one byte in ssl2_new (ssl/s2_lib.c).
6524 Otherwise do_ssl_write (ssl/s2_pkt.c) will write beyond buffer limits
6525 when writing a 32767 byte record.
6526 [Bodo Moeller; problem reported by Eric Day <eday@concentric.net>]
6527
6528 *) In RSA_eay_public_{en,ed}crypt and RSA_eay_mod_exp (rsa_eay.c),
6529 obtain lock CRYPTO_LOCK_RSA before setting rsa->_method_mod_{n,p,q}.
6530
6531 (RSA objects have a reference count access to which is protected
6532 by CRYPTO_LOCK_RSA [see rsa_lib.c, s3_srvr.c, ssl_cert.c, ssl_rsa.c],
6533 so they are meant to be shared between threads.)
6534 [Bodo Moeller, Geoff Thorpe; original patch submitted by
6535 "Reddie, Steven" <Steven.Reddie@ca.com>]
6536
6537 *) Fix a deadlock in CRYPTO_mem_leaks().
6538 [Bodo Moeller]
6539
6540 *) Use better test patterns in bntest.
6541 [Ulf Möller]
6542
6543 *) rand_win.c fix for Borland C.
6544 [Ulf Möller]
6545
6546 *) BN_rshift bugfix for n == 0.
6547 [Bodo Moeller]
6548
6549 *) Add a 'bctest' script that checks for some known 'bc' bugs
6550 so that 'make test' does not abort just because 'bc' is broken.
6551 [Bodo Moeller]
6552
6553 *) Store verify_result within SSL_SESSION also for client side to
6554 avoid potential security hole. (Re-used sessions on the client side
6555 always resulted in verify_result==X509_V_OK, not using the original
6556 result of the server certificate verification.)
6557 [Lutz Jaenicke]
6558
6559 *) Fix ssl3_pending: If the record in s->s3->rrec is not of type
6560 SSL3_RT_APPLICATION_DATA, return 0.
6561 Similarly, change ssl2_pending to return 0 if SSL_in_init(s) is true.
6562 [Bodo Moeller]
6563
6564 *) Fix SSL_peek:
6565 Both ssl2_peek and ssl3_peek, which were totally broken in earlier
6566 releases, have been re-implemented by renaming the previous
6567 implementations of ssl2_read and ssl3_read to ssl2_read_internal
6568 and ssl3_read_internal, respectively, and adding 'peek' parameters
6569 to them. The new ssl[23]_{read,peek} functions are calls to
6570 ssl[23]_read_internal with the 'peek' flag set appropriately.
6571 A 'peek' parameter has also been added to ssl3_read_bytes, which
6572 does the actual work for ssl3_read_internal.
6573 [Bodo Moeller]
6574
6575 *) Initialise "ex_data" member of RSA/DSA/DH structures prior to calling
6576 the method-specific "init()" handler. Also clean up ex_data after
6577 calling the method-specific "finish()" handler. Previously, this was
6578 happening the other way round.
6579 [Geoff Thorpe]
6580
6581 *) Increase BN_CTX_NUM (the number of BIGNUMs in a BN_CTX) to 16.
6582 The previous value, 12, was not always sufficient for BN_mod_exp().
6583 [Bodo Moeller]
6584
6585 *) Make sure that shared libraries get the internal name engine with
6586 the full version number and not just 0. This should mark the
6587 shared libraries as not backward compatible. Of course, this should
6588 be changed again when we can guarantee backward binary compatibility.
6589 [Richard Levitte]
6590
6591 *) Fix typo in get_cert_by_subject() in by_dir.c
6592 [Jean-Marc Desperrier <jean-marc.desperrier@certplus.com>]
6593
6594 *) Rework the system to generate shared libraries:
6595
6596 - Make note of the expected extension for the shared libraries and
6597 if there is a need for symbolic links from for example libcrypto.so.0
6598 to libcrypto.so.0.9.7. There is extended info in Configure for
6599 that.
6600
6601 - Make as few rebuilds of the shared libraries as possible.
6602
6603 - Still avoid linking the OpenSSL programs with the shared libraries.
6604
6605 - When installing, install the shared libraries separately from the
6606 static ones.
6607 [Richard Levitte]
6608
6609 *) Fix SSL_CTX_set_read_ahead macro to actually use its argument.
6610
6611 Copy SSL_CTX's read_ahead flag to SSL object directly in SSL_new
6612 and not in SSL_clear because the latter is also used by the
6613 accept/connect functions; previously, the settings made by
6614 SSL_set_read_ahead would be lost during the handshake.
6615 [Bodo Moeller; problems reported by Anders Gertz <gertz@epact.se>]
6616
6617 *) Correct util/mkdef.pl to be selective about disabled algorithms.
6618 Previously, it would create entries for disableed algorithms no
6619 matter what.
6620 [Richard Levitte]
6621
6622 *) Added several new manual pages for SSL_* function.
6623 [Lutz Jaenicke]
6624
6625 Changes between 0.9.5a and 0.9.6 [24 Sep 2000]
6626
6627 *) In ssl23_get_client_hello, generate an error message when faced
6628 with an initial SSL 3.0/TLS record that is too small to contain the
6629 first two bytes of the ClientHello message, i.e. client_version.
6630 (Note that this is a pathologic case that probably has never happened
6631 in real life.) The previous approach was to use the version number
6632 from the record header as a substitute; but our protocol choice
6633 should not depend on that one because it is not authenticated
6634 by the Finished messages.
6635 [Bodo Moeller]
6636
6637 *) More robust randomness gathering functions for Windows.
6638 [Jeffrey Altman <jaltman@columbia.edu>]
6639
6640 *) For compatibility reasons if the flag X509_V_FLAG_ISSUER_CHECK is
6641 not set then we don't setup the error code for issuer check errors
6642 to avoid possibly overwriting other errors which the callback does
6643 handle. If an application does set the flag then we assume it knows
6644 what it is doing and can handle the new informational codes
6645 appropriately.
6646 [Steve Henson]
6647
6648 *) Fix for a nasty bug in ASN1_TYPE handling. ASN1_TYPE is used for
6649 a general "ANY" type, as such it should be able to decode anything
6650 including tagged types. However it didn't check the class so it would
6651 wrongly interpret tagged types in the same way as their universal
6652 counterpart and unknown types were just rejected. Changed so that the
6653 tagged and unknown types are handled in the same way as a SEQUENCE:
6654 that is the encoding is stored intact. There is also a new type
6655 "V_ASN1_OTHER" which is used when the class is not universal, in this
6656 case we have no idea what the actual type is so we just lump them all
6657 together.
6658 [Steve Henson]
6659
6660 *) On VMS, stdout may very well lead to a file that is written to
6661 in a record-oriented fashion. That means that every write() will
6662 write a separate record, which will be read separately by the
6663 programs trying to read from it. This can be very confusing.
6664
6665 The solution is to put a BIO filter in the way that will buffer
6666 text until a linefeed is reached, and then write everything a
6667 line at a time, so every record written will be an actual line,
6668 not chunks of lines and not (usually doesn't happen, but I've
6669 seen it once) several lines in one record. BIO_f_linebuffer() is
6670 the answer.
6671
6672 Currently, it's a VMS-only method, because that's where it has
6673 been tested well enough.
6674 [Richard Levitte]
6675
6676 *) Remove 'optimized' squaring variant in BN_mod_mul_montgomery,
6677 it can return incorrect results.
6678 (Note: The buggy variant was not enabled in OpenSSL 0.9.5a,
6679 but it was in 0.9.6-beta[12].)
6680 [Bodo Moeller]
6681
6682 *) Disable the check for content being present when verifying detached
6683 signatures in pk7_smime.c. Some versions of Netscape (wrongly)
6684 include zero length content when signing messages.
6685 [Steve Henson]
6686
6687 *) New BIO_shutdown_wr macro, which invokes the BIO_C_SHUTDOWN_WR
6688 BIO_ctrl (for BIO pairs).
6689 [Bodo Möller]
6690
6691 *) Add DSO method for VMS.
6692 [Richard Levitte]
6693
6694 *) Bug fix: Montgomery multiplication could produce results with the
6695 wrong sign.
6696 [Ulf Möller]
6697
6698 *) Add RPM specification openssl.spec and modify it to build three
6699 packages. The default package contains applications, application
6700 documentation and run-time libraries. The devel package contains
6701 include files, static libraries and function documentation. The
6702 doc package contains the contents of the doc directory. The original
6703 openssl.spec was provided by Damien Miller <djm@mindrot.org>.
6704 [Richard Levitte]
6705
6706 *) Add a large number of documentation files for many SSL routines.
6707 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
6708
6709 *) Add a configuration entry for Sony News 4.
6710 [NAKAJI Hiroyuki <nakaji@tutrp.tut.ac.jp>]
6711
6712 *) Don't set the two most significant bits to one when generating a
6713 random number < q in the DSA library.
6714 [Ulf Möller]
6715
6716 *) New SSL API mode 'SSL_MODE_AUTO_RETRY'. This disables the default
6717 behaviour that SSL_read may result in SSL_ERROR_WANT_READ (even if
6718 the underlying transport is blocking) if a handshake took place.
6719 (The default behaviour is needed by applications such as s_client
6720 and s_server that use select() to determine when to use SSL_read;
6721 but for applications that know in advance when to expect data, it
6722 just makes things more complicated.)
6723 [Bodo Moeller]
6724
6725 *) Add RAND_egd_bytes(), which gives control over the number of bytes read
6726 from EGD.
6727 [Ben Laurie]
6728
6729 *) Add a few more EBCDIC conditionals that make `req' and `x509'
6730 work better on such systems.
6731 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
6732
6733 *) Add two demo programs for PKCS12_parse() and PKCS12_create().
6734 Update PKCS12_parse() so it copies the friendlyName and the
6735 keyid to the certificates aux info.
6736 [Steve Henson]
6737
6738 *) Fix bug in PKCS7_verify() which caused an infinite loop
6739 if there was more than one signature.
6740 [Sven Uszpelkat <su@celocom.de>]
6741
6742 *) Major change in util/mkdef.pl to include extra information
6743 about each symbol, as well as presentig variables as well
6744 as functions. This change means that there's n more need
6745 to rebuild the .num files when some algorithms are excluded.
6746 [Richard Levitte]
6747
6748 *) Allow the verify time to be set by an application,
6749 rather than always using the current time.
6750 [Steve Henson]
6751
6752 *) Phase 2 verify code reorganisation. The certificate
6753 verify code now looks up an issuer certificate by a
6754 number of criteria: subject name, authority key id
6755 and key usage. It also verifies self signed certificates
6756 by the same criteria. The main comparison function is
6757 X509_check_issued() which performs these checks.
6758
6759 Lot of changes were necessary in order to support this
6760 without completely rewriting the lookup code.
6761
6762 Authority and subject key identifier are now cached.
6763
6764 The LHASH 'certs' is X509_STORE has now been replaced
6765 by a STACK_OF(X509_OBJECT). This is mainly because an
6766 LHASH can't store or retrieve multiple objects with
6767 the same hash value.
6768
6769 As a result various functions (which were all internal
6770 use only) have changed to handle the new X509_STORE
6771 structure. This will break anything that messed round
6772 with X509_STORE internally.
6773
6774 The functions X509_STORE_add_cert() now checks for an
6775 exact match, rather than just subject name.
6776
6777 The X509_STORE API doesn't directly support the retrieval
6778 of multiple certificates matching a given criteria, however
6779 this can be worked round by performing a lookup first
6780 (which will fill the cache with candidate certificates)
6781 and then examining the cache for matches. This is probably
6782 the best we can do without throwing out X509_LOOKUP
6783 entirely (maybe later...).
6784
6785 The X509_VERIFY_CTX structure has been enhanced considerably.
6786
6787 All certificate lookup operations now go via a get_issuer()
6788 callback. Although this currently uses an X509_STORE it
6789 can be replaced by custom lookups. This is a simple way
6790 to bypass the X509_STORE hackery necessary to make this
6791 work and makes it possible to use more efficient techniques
6792 in future. A very simple version which uses a simple
6793 STACK for its trusted certificate store is also provided
6794 using X509_STORE_CTX_trusted_stack().
6795
6796 The verify_cb() and verify() callbacks now have equivalents
6797 in the X509_STORE_CTX structure.
6798
6799 X509_STORE_CTX also has a 'flags' field which can be used
6800 to customise the verify behaviour.
6801 [Steve Henson]
6802
6803 *) Add new PKCS#7 signing option PKCS7_NOSMIMECAP which
6804 excludes S/MIME capabilities.
6805 [Steve Henson]
6806
6807 *) When a certificate request is read in keep a copy of the
6808 original encoding of the signed data and use it when outputing
6809 again. Signatures then use the original encoding rather than
6810 a decoded, encoded version which may cause problems if the
6811 request is improperly encoded.
6812 [Steve Henson]
6813
6814 *) For consistency with other BIO_puts implementations, call
6815 buffer_write(b, ...) directly in buffer_puts instead of calling
6816 BIO_write(b, ...).
6817
6818 In BIO_puts, increment b->num_write as in BIO_write.
6819 [Peter.Sylvester@EdelWeb.fr]
6820
6821 *) Fix BN_mul_word for the case where the word is 0. (We have to use
6822 BN_zero, we may not return a BIGNUM with an array consisting of
6823 words set to zero.)
6824 [Bodo Moeller]
6825
6826 *) Avoid calling abort() from within the library when problems are
6827 detected, except if preprocessor symbols have been defined
6828 (such as REF_CHECK, BN_DEBUG etc.).
6829 [Bodo Moeller]
6830
6831 *) New openssl application 'rsautl'. This utility can be
6832 used for low level RSA operations. DER public key
6833 BIO/fp routines also added.
6834 [Steve Henson]
6835
6836 *) New Configure entry and patches for compiling on QNX 4.
6837 [Andreas Schneider <andreas@ds3.etech.fh-hamburg.de>]
6838
6839 *) A demo state-machine implementation was sponsored by
6840 Nuron (http://www.nuron.com/) and is now available in
6841 demos/state_machine.
6842 [Ben Laurie]
6843
6844 *) New options added to the 'dgst' utility for signature
6845 generation and verification.
6846 [Steve Henson]
6847
6848 *) Unrecognized PKCS#7 content types are now handled via a
6849 catch all ASN1_TYPE structure. This allows unsupported
6850 types to be stored as a "blob" and an application can
6851 encode and decode it manually.
6852 [Steve Henson]
6853
6854 *) Fix various signed/unsigned issues to make a_strex.c
6855 compile under VC++.
6856 [Oscar Jacobsson <oscar.jacobsson@celocom.com>]
6857
6858 *) ASN1 fixes. i2d_ASN1_OBJECT was not returning the correct
6859 length if passed a buffer. ASN1_INTEGER_to_BN failed
6860 if passed a NULL BN and its argument was negative.
6861 [Steve Henson, pointed out by Sven Heiberg <sven@tartu.cyber.ee>]
6862
6863 *) Modification to PKCS#7 encoding routines to output definite
6864 length encoding. Since currently the whole structures are in
6865 memory there's not real point in using indefinite length
6866 constructed encoding. However if OpenSSL is compiled with
6867 the flag PKCS7_INDEFINITE_ENCODING the old form is used.
6868 [Steve Henson]
6869
6870 *) Added BIO_vprintf() and BIO_vsnprintf().
6871 [Richard Levitte]
6872
6873 *) Added more prefixes to parse for in the the strings written
6874 through a logging bio, to cover all the levels that are available
6875 through syslog. The prefixes are now:
6876
6877 PANIC, EMERG, EMR => LOG_EMERG
6878 ALERT, ALR => LOG_ALERT
6879 CRIT, CRI => LOG_CRIT
6880 ERROR, ERR => LOG_ERR
6881 WARNING, WARN, WAR => LOG_WARNING
6882 NOTICE, NOTE, NOT => LOG_NOTICE
6883 INFO, INF => LOG_INFO
6884 DEBUG, DBG => LOG_DEBUG
6885
6886 and as before, if none of those prefixes are present at the
6887 beginning of the string, LOG_ERR is chosen.
6888
6889 On Win32, the LOG_* levels are mapped according to this:
6890
6891 LOG_EMERG, LOG_ALERT, LOG_CRIT, LOG_ERR => EVENTLOG_ERROR_TYPE
6892 LOG_WARNING => EVENTLOG_WARNING_TYPE
6893 LOG_NOTICE, LOG_INFO, LOG_DEBUG => EVENTLOG_INFORMATION_TYPE
6894
6895 [Richard Levitte]
6896
6897 *) Made it possible to reconfigure with just the configuration
6898 argument "reconf" or "reconfigure". The command line arguments
6899 are stored in Makefile.ssl in the variable CONFIGURE_ARGS,
6900 and are retrieved from there when reconfiguring.
6901 [Richard Levitte]
6902
6903 *) MD4 implemented.
6904 [Assar Westerlund <assar@sics.se>, Richard Levitte]
6905
6906 *) Add the arguments -CAfile and -CApath to the pkcs12 utility.
6907 [Richard Levitte]
6908
6909 *) The obj_dat.pl script was messing up the sorting of object
6910 names. The reason was that it compared the quoted version
6911 of strings as a result "OCSP" > "OCSP Signing" because
6912 " > SPACE. Changed script to store unquoted versions of
6913 names and add quotes on output. It was also omitting some
6914 names from the lookup table if they were given a default
6915 value (that is if SN is missing it is given the same
6916 value as LN and vice versa), these are now added on the
6917 grounds that if an object has a name we should be able to
6918 look it up. Finally added warning output when duplicate
6919 short or long names are found.
6920 [Steve Henson]
6921
6922 *) Changes needed for Tandem NSK.
6923 [Scott Uroff <scott@xypro.com>]
6924
6925 *) Fix SSL 2.0 rollback checking: Due to an off-by-one error in
6926 RSA_padding_check_SSLv23(), special padding was never detected
6927 and thus the SSL 3.0/TLS 1.0 countermeasure against protocol
6928 version rollback attacks was not effective.
6929
6930 In s23_clnt.c, don't use special rollback-attack detection padding
6931 (RSA_SSLV23_PADDING) if SSL 2.0 is the only protocol enabled in the
6932 client; similarly, in s23_srvr.c, don't do the rollback check if
6933 SSL 2.0 is the only protocol enabled in the server.
6934 [Bodo Moeller]
6935
6936 *) Make it possible to get hexdumps of unprintable data with 'openssl
6937 asn1parse'. By implication, the functions ASN1_parse_dump() and
6938 BIO_dump_indent() are added.
6939 [Richard Levitte]
6940
6941 *) New functions ASN1_STRING_print_ex() and X509_NAME_print_ex()
6942 these print out strings and name structures based on various
6943 flags including RFC2253 support and proper handling of
6944 multibyte characters. Added options to the 'x509' utility
6945 to allow the various flags to be set.
6946 [Steve Henson]
6947
6948 *) Various fixes to use ASN1_TIME instead of ASN1_UTCTIME.
6949 Also change the functions X509_cmp_current_time() and
6950 X509_gmtime_adj() work with an ASN1_TIME structure,
6951 this will enable certificates using GeneralizedTime in validity
6952 dates to be checked.
6953 [Steve Henson]
6954
6955 *) Make the NEG_PUBKEY_BUG code (which tolerates invalid
6956 negative public key encodings) on by default,
6957 NO_NEG_PUBKEY_BUG can be set to disable it.
6958 [Steve Henson]
6959
6960 *) New function c2i_ASN1_OBJECT() which acts on ASN1_OBJECT
6961 content octets. An i2c_ASN1_OBJECT is unnecessary because
6962 the encoding can be trivially obtained from the structure.
6963 [Steve Henson]
6964
6965 *) crypto/err.c locking bugfix: Use write locks (CRYPTO_w_[un]lock),
6966 not read locks (CRYPTO_r_[un]lock).
6967 [Bodo Moeller]
6968
6969 *) A first attempt at creating official support for shared
6970 libraries through configuration. I've kept it so the
6971 default is static libraries only, and the OpenSSL programs
6972 are always statically linked for now, but there are
6973 preparations for dynamic linking in place.
6974 This has been tested on Linux and Tru64.
6975 [Richard Levitte]
6976
6977 *) Randomness polling function for Win9x, as described in:
6978 Peter Gutmann, Software Generation of Practically Strong
6979 Random Numbers.
6980 [Ulf Möller]
6981
6982 *) Fix so PRNG is seeded in req if using an already existing
6983 DSA key.
6984 [Steve Henson]
6985
6986 *) New options to smime application. -inform and -outform
6987 allow alternative formats for the S/MIME message including
6988 PEM and DER. The -content option allows the content to be
6989 specified separately. This should allow things like Netscape
6990 form signing output easier to verify.
6991 [Steve Henson]
6992
6993 *) Fix the ASN1 encoding of tags using the 'long form'.
6994 [Steve Henson]
6995
6996 *) New ASN1 functions, i2c_* and c2i_* for INTEGER and BIT
6997 STRING types. These convert content octets to and from the
6998 underlying type. The actual tag and length octets are
6999 already assumed to have been read in and checked. These
7000 are needed because all other string types have virtually
7001 identical handling apart from the tag. By having versions
7002 of the ASN1 functions that just operate on content octets
7003 IMPLICIT tagging can be handled properly. It also allows
7004 the ASN1_ENUMERATED code to be cut down because ASN1_ENUMERATED
7005 and ASN1_INTEGER are identical apart from the tag.
7006 [Steve Henson]
7007
7008 *) Change the handling of OID objects as follows:
7009
7010 - New object identifiers are inserted in objects.txt, following
7011 the syntax given in objects.README.
7012 - objects.pl is used to process obj_mac.num and create a new
7013 obj_mac.h.
7014 - obj_dat.pl is used to create a new obj_dat.h, using the data in
7015 obj_mac.h.
7016
7017 This is currently kind of a hack, and the perl code in objects.pl
7018 isn't very elegant, but it works as I intended. The simplest way
7019 to check that it worked correctly is to look in obj_dat.h and
7020 check the array nid_objs and make sure the objects haven't moved
7021 around (this is important!). Additions are OK, as well as
7022 consistent name changes.
7023 [Richard Levitte]
7024
7025 *) Add BSD-style MD5-based passwords to 'openssl passwd' (option '-1').
7026 [Bodo Moeller]
7027
7028 *) Addition of the command line parameter '-rand file' to 'openssl req'.
7029 The given file adds to whatever has already been seeded into the
7030 random pool through the RANDFILE configuration file option or
7031 environment variable, or the default random state file.
7032 [Richard Levitte]
7033
7034 *) mkstack.pl now sorts each macro group into lexical order.
7035 Previously the output order depended on the order the files
7036 appeared in the directory, resulting in needless rewriting
7037 of safestack.h .
7038 [Steve Henson]
7039
7040 *) Patches to make OpenSSL compile under Win32 again. Mostly
7041 work arounds for the VC++ problem that it treats func() as
7042 func(void). Also stripped out the parts of mkdef.pl that
7043 added extra typesafe functions: these no longer exist.
7044 [Steve Henson]
7045
7046 *) Reorganisation of the stack code. The macros are now all
7047 collected in safestack.h . Each macro is defined in terms of
7048 a "stack macro" of the form SKM_<name>(type, a, b). The
7049 DEBUG_SAFESTACK is now handled in terms of function casts,
7050 this has the advantage of retaining type safety without the
7051 use of additional functions. If DEBUG_SAFESTACK is not defined
7052 then the non typesafe macros are used instead. Also modified the
7053 mkstack.pl script to handle the new form. Needs testing to see
7054 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK
7055 the default if no major problems. Similar behaviour for ASN1_SET_OF
7056 and PKCS12_STACK_OF.
7057 [Steve Henson]
7058
7059 *) When some versions of IIS use the 'NET' form of private key the
7060 key derivation algorithm is different. Normally MD5(password) is
7061 used as a 128 bit RC4 key. In the modified case
7062 MD5(MD5(password) + "SGCKEYSALT") is used insted. Added some
7063 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same
7064 as the old Netscape_RSA functions except they have an additional
7065 'sgckey' parameter which uses the modified algorithm. Also added
7066 an -sgckey command line option to the rsa utility. Thanks to
7067 Adrian Peck <bertie@ncipher.com> for posting details of the modified
7068 algorithm to openssl-dev.
7069 [Steve Henson]
7070
7071 *) The evp_local.h macros were using 'c.##kname' which resulted in
7072 invalid expansion on some systems (SCO 5.0.5 for example).
7073 Corrected to 'c.kname'.
7074 [Phillip Porch <root@theporch.com>]
7075
7076 *) New X509_get1_email() and X509_REQ_get1_email() functions that return
7077 a STACK of email addresses from a certificate or request, these look
7078 in the subject name and the subject alternative name extensions and
7079 omit any duplicate addresses.
7080 [Steve Henson]
7081
7082 *) Re-implement BN_mod_exp2_mont using independent (and larger) windows.
7083 This makes DSA verification about 2 % faster.
7084 [Bodo Moeller]
7085
7086 *) Increase maximum window size in BN_mod_exp_... to 6 bits instead of 5
7087 (meaning that now 2^5 values will be precomputed, which is only 4 KB
7088 plus overhead for 1024 bit moduli).
7089 This makes exponentiations about 0.5 % faster for 1024 bit
7090 exponents (as measured by "openssl speed rsa2048").
7091 [Bodo Moeller]
7092
7093 *) Rename memory handling macros to avoid conflicts with other
7094 software:
7095 Malloc => OPENSSL_malloc
7096 Malloc_locked => OPENSSL_malloc_locked
7097 Realloc => OPENSSL_realloc
7098 Free => OPENSSL_free
7099 [Richard Levitte]
7100
7101 *) New function BN_mod_exp_mont_word for small bases (roughly 15%
7102 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange).
7103 [Bodo Moeller]
7104
7105 *) CygWin32 support.
7106 [John Jarvie <jjarvie@newsguy.com>]
7107
7108 *) The type-safe stack code has been rejigged. It is now only compiled
7109 in when OpenSSL is configured with the DEBUG_SAFESTACK option and
7110 by default all type-specific stack functions are "#define"d back to
7111 standard stack functions. This results in more streamlined output
7112 but retains the type-safety checking possibilities of the original
7113 approach.
7114 [Geoff Thorpe]
7115
7116 *) The STACK code has been cleaned up, and certain type declarations
7117 that didn't make a lot of sense have been brought in line. This has
7118 also involved a cleanup of sorts in safestack.h to more correctly
7119 map type-safe stack functions onto their plain stack counterparts.
7120 This work has also resulted in a variety of "const"ifications of
7121 lots of the code, especially "_cmp" operations which should normally
7122 be prototyped with "const" parameters anyway.
7123 [Geoff Thorpe]
7124
7125 *) When generating bytes for the first time in md_rand.c, 'stir the pool'
7126 by seeding with STATE_SIZE dummy bytes (with zero entropy count).
7127 (The PRNG state consists of two parts, the large pool 'state' and 'md',
7128 where all of 'md' is used each time the PRNG is used, but 'state'
7129 is used only indexed by a cyclic counter. As entropy may not be
7130 well distributed from the beginning, 'md' is important as a
7131 chaining variable. However, the output function chains only half
7132 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains
7133 all of 'md', and seeding with STATE_SIZE dummy bytes will result
7134 in all of 'state' being rewritten, with the new values depending
7135 on virtually all of 'md'. This overcomes the 80 bit limitation.)
7136 [Bodo Moeller]
7137
7138 *) In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when
7139 the handshake is continued after ssl_verify_cert_chain();
7140 otherwise, if SSL_VERIFY_NONE is set, remaining error codes
7141 can lead to 'unexplainable' connection aborts later.
7142 [Bodo Moeller; problem tracked down by Lutz Jaenicke]
7143
7144 *) Major EVP API cipher revision.
7145 Add hooks for extra EVP features. This allows various cipher
7146 parameters to be set in the EVP interface. Support added for variable
7147 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and
7148 setting of RC2 and RC5 parameters.
7149
7150 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length
7151 ciphers.
7152
7153 Remove lots of duplicated code from the EVP library. For example *every*
7154 cipher init() function handles the 'iv' in the same way according to the
7155 cipher mode. They also all do nothing if the 'key' parameter is NULL and
7156 for CFB and OFB modes they zero ctx->num.
7157
7158 New functionality allows removal of S/MIME code RC2 hack.
7159
7160 Most of the routines have the same form and so can be declared in terms
7161 of macros.
7162
7163 By shifting this to the top level EVP_CipherInit() it can be removed from
7164 all individual ciphers. If the cipher wants to handle IVs or keys
7165 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT
7166 flags.
7167
7168 Change lots of functions like EVP_EncryptUpdate() to now return a
7169 value: although software versions of the algorithms cannot fail
7170 any installed hardware versions can.
7171 [Steve Henson]
7172
7173 *) Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if
7174 this option is set, tolerate broken clients that send the negotiated
7175 protocol version number instead of the requested protocol version
7176 number.
7177 [Bodo Moeller]
7178
7179 *) Call dh_tmp_cb (set by ..._TMP_DH_CB) with correct 'is_export' flag;
7180 i.e. non-zero for export ciphersuites, zero otherwise.
7181 Previous versions had this flag inverted, inconsistent with
7182 rsa_tmp_cb (..._TMP_RSA_CB).
7183 [Bodo Moeller; problem reported by Amit Chopra]
7184
7185 *) Add missing DSA library text string. Work around for some IIS
7186 key files with invalid SEQUENCE encoding.
7187 [Steve Henson]
7188
7189 *) Add a document (doc/standards.txt) that list all kinds of standards
7190 and so on that are implemented in OpenSSL.
7191 [Richard Levitte]
7192
7193 *) Enhance c_rehash script. Old version would mishandle certificates
7194 with the same subject name hash and wouldn't handle CRLs at all.
7195 Added -fingerprint option to crl utility, to support new c_rehash
7196 features.
7197 [Steve Henson]
7198
7199 *) Eliminate non-ANSI declarations in crypto.h and stack.h.
7200 [Ulf Möller]
7201
7202 *) Fix for SSL server purpose checking. Server checking was
7203 rejecting certificates which had extended key usage present
7204 but no ssl client purpose.
7205 [Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>]
7206
7207 *) Make PKCS#12 code work with no password. The PKCS#12 spec
7208 is a little unclear about how a blank password is handled.
7209 Since the password in encoded as a BMPString with terminating
7210 double NULL a zero length password would end up as just the
7211 double NULL. However no password at all is different and is
7212 handled differently in the PKCS#12 key generation code. NS
7213 treats a blank password as zero length. MSIE treats it as no
7214 password on export: but it will try both on import. We now do
7215 the same: PKCS12_parse() tries zero length and no password if
7216 the password is set to "" or NULL (NULL is now a valid password:
7217 it wasn't before) as does the pkcs12 application.
7218 [Steve Henson]
7219
7220 *) Bugfixes in apps/x509.c: Avoid a memory leak; and don't use
7221 perror when PEM_read_bio_X509_REQ fails, the error message must
7222 be obtained from the error queue.
7223 [Bodo Moeller]
7224
7225 *) Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing
7226 it in ERR_remove_state if appropriate, and change ERR_get_state
7227 accordingly to avoid race conditions (this is necessary because
7228 thread_hash is no longer constant once set).
7229 [Bodo Moeller]
7230
7231 *) Bugfix for linux-elf makefile.one.
7232 [Ulf Möller]
7233
7234 *) RSA_get_default_method() will now cause a default
7235 RSA_METHOD to be chosen if one doesn't exist already.
7236 Previously this was only set during a call to RSA_new()
7237 or RSA_new_method(NULL) meaning it was possible for
7238 RSA_get_default_method() to return NULL.
7239 [Geoff Thorpe]
7240
7241 *) Added native name translation to the existing DSO code
7242 that will convert (if the flag to do so is set) filenames
7243 that are sufficiently small and have no path information
7244 into a canonical native form. Eg. "blah" converted to
7245 "libblah.so" or "blah.dll" etc.
7246 [Geoff Thorpe]
7247
7248 *) New function ERR_error_string_n(e, buf, len) which is like
7249 ERR_error_string(e, buf), but writes at most 'len' bytes
7250 including the 0 terminator. For ERR_error_string_n, 'buf'
7251 may not be NULL.
7252 [Damien Miller <djm@mindrot.org>, Bodo Moeller]
7253
7254 *) CONF library reworked to become more general. A new CONF
7255 configuration file reader "class" is implemented as well as a
7256 new functions (NCONF_*, for "New CONF") to handle it. The now
7257 old CONF_* functions are still there, but are reimplemented to
7258 work in terms of the new functions. Also, a set of functions
7259 to handle the internal storage of the configuration data is
7260 provided to make it easier to write new configuration file
7261 reader "classes" (I can definitely see something reading a
7262 configuration file in XML format, for example), called _CONF_*,
7263 or "the configuration storage API"...
7264
7265 The new configuration file reading functions are:
7266
7267 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio,
7268 NCONF_get_section, NCONF_get_string, NCONF_get_numbre
7269
7270 NCONF_default, NCONF_WIN32
7271
7272 NCONF_dump_fp, NCONF_dump_bio
7273
7274 NCONF_default and NCONF_WIN32 are method (or "class") choosers,
7275 NCONF_new creates a new CONF object. This works in the same way
7276 as other interfaces in OpenSSL, like the BIO interface.
7277 NCONF_dump_* dump the internal storage of the configuration file,
7278 which is useful for debugging. All other functions take the same
7279 arguments as the old CONF_* functions wth the exception of the
7280 first that must be a `CONF *' instead of a `LHASH *'.
7281
7282 To make it easer to use the new classes with the old CONF_* functions,
7283 the function CONF_set_default_method is provided.
7284 [Richard Levitte]
7285
7286 *) Add '-tls1' option to 'openssl ciphers', which was already
7287 mentioned in the documentation but had not been implemented.
7288 (This option is not yet really useful because even the additional
7289 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.)
7290 [Bodo Moeller]
7291
7292 *) Initial DSO code added into libcrypto for letting OpenSSL (and
7293 OpenSSL-based applications) load shared libraries and bind to
7294 them in a portable way.
7295 [Geoff Thorpe, with contributions from Richard Levitte]
7296
7297 Changes between 0.9.5 and 0.9.5a [1 Apr 2000]
7298
7299 *) Make sure _lrotl and _lrotr are only used with MSVC.
7300
7301 *) Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status
7302 (the default implementation of RAND_status).
7303
7304 *) Rename openssl x509 option '-crlext', which was added in 0.9.5,
7305 to '-clrext' (= clear extensions), as intended and documented.
7306 [Bodo Moeller; inconsistency pointed out by Michael Attili
7307 <attili@amaxo.com>]
7308
7309 *) Fix for HMAC. It wasn't zeroing the rest of the block if the key length
7310 was larger than the MD block size.
7311 [Steve Henson, pointed out by Yost William <YostW@tce.com>]
7312
7313 *) Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument
7314 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set()
7315 using the passed key: if the passed key was a private key the result
7316 of X509_print(), for example, would be to print out all the private key
7317 components.
7318 [Steve Henson]
7319
7320 *) des_quad_cksum() byte order bug fix.
7321 [Ulf Möller, using the problem description in krb4-0.9.7, where
7322 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>]
7323
7324 *) Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly
7325 discouraged.
7326 [Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>]
7327
7328 *) For easily testing in shell scripts whether some command
7329 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX'
7330 returns with exit code 0 iff no command of the given name is available.
7331 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases,
7332 the output goes to stdout and nothing is printed to stderr.
7333 Additional arguments are always ignored.
7334
7335 Since for each cipher there is a command of the same name,
7336 the 'no-cipher' compilation switches can be tested this way.
7337
7338 ('openssl no-XXX' is not able to detect pseudo-commands such
7339 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.)
7340 [Bodo Moeller]
7341
7342 *) Update test suite so that 'make test' succeeds in 'no-rsa' configuration.
7343 [Bodo Moeller]
7344
7345 *) For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE
7346 is set; it will be thrown away anyway because each handshake creates
7347 its own key.
7348 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition
7349 to parameters -- in previous versions (since OpenSSL 0.9.3) the
7350 'default key' from SSL_CTX_set_tmp_dh would always be lost, meanining
7351 you effectivly got SSL_OP_SINGLE_DH_USE when using this macro.
7352 [Bodo Moeller]
7353
7354 *) New s_client option -ign_eof: EOF at stdin is ignored, and
7355 'Q' and 'R' lose their special meanings (quit/renegotiate).
7356 This is part of what -quiet does; unlike -quiet, -ign_eof
7357 does not suppress any output.
7358 [Richard Levitte]
7359
7360 *) Add compatibility options to the purpose and trust code. The
7361 purpose X509_PURPOSE_ANY is "any purpose" which automatically
7362 accepts a certificate or CA, this was the previous behaviour,
7363 with all the associated security issues.
7364
7365 X509_TRUST_COMPAT is the old trust behaviour: only and
7366 automatically trust self signed roots in certificate store. A
7367 new trust setting X509_TRUST_DEFAULT is used to specify that
7368 a purpose has no associated trust setting and it should instead
7369 use the value in the default purpose.
7370 [Steve Henson]
7371
7372 *) Fix the PKCS#8 DSA private key code so it decodes keys again
7373 and fix a memory leak.
7374 [Steve Henson]
7375
7376 *) In util/mkerr.pl (which implements 'make errors'), preserve
7377 reason strings from the previous version of the .c file, as
7378 the default to have only downcase letters (and digits) in
7379 automatically generated reasons codes is not always appropriate.
7380 [Bodo Moeller]
7381
7382 *) In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table
7383 using strerror. Previously, ERR_reason_error_string() returned
7384 library names as reason strings for SYSerr; but SYSerr is a special
7385 case where small numbers are errno values, not library numbers.
7386 [Bodo Moeller]
7387
7388 *) Add '-dsaparam' option to 'openssl dhparam' application. This
7389 converts DSA parameters into DH parameters. (When creating parameters,
7390 DSA_generate_parameters is used.)
7391 [Bodo Moeller]
7392
7393 *) Include 'length' (recommended exponent length) in C code generated
7394 by 'openssl dhparam -C'.
7395 [Bodo Moeller]
7396
7397 *) The second argument to set_label in perlasm was already being used
7398 so couldn't be used as a "file scope" flag. Moved to third argument
7399 which was free.
7400 [Steve Henson]
7401
7402 *) In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes
7403 instead of RAND_bytes for encryption IVs and salts.
7404 [Bodo Moeller]
7405
7406 *) Include RAND_status() into RAND_METHOD instead of implementing
7407 it only for md_rand.c Otherwise replacing the PRNG by calling
7408 RAND_set_rand_method would be impossible.
7409 [Bodo Moeller]
7410
7411 *) Don't let DSA_generate_key() enter an infinite loop if the random
7412 number generation fails.
7413 [Bodo Moeller]
7414
7415 *) New 'rand' application for creating pseudo-random output.
7416 [Bodo Moeller]
7417
7418 *) Added configuration support for Linux/IA64
7419 [Rolf Haberrecker <rolf@suse.de>]
7420
7421 *) Assembler module support for Mingw32.
7422 [Ulf Möller]
7423
7424 *) Shared library support for HPUX (in shlib/).
7425 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous]
7426
7427 *) Shared library support for Solaris gcc.
7428 [Lutz Behnke <behnke@trustcenter.de>]
7429
7430 Changes between 0.9.4 and 0.9.5 [28 Feb 2000]
7431
7432 *) PKCS7_encrypt() was adding text MIME headers twice because they
7433 were added manually and by SMIME_crlf_copy().
7434 [Steve Henson]
7435
7436 *) In bntest.c don't call BN_rand with zero bits argument.
7437 [Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>]
7438
7439 *) BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n]
7440 case was implemented. This caused BN_div_recp() to fail occasionally.
7441 [Ulf Möller]
7442
7443 *) Add an optional second argument to the set_label() in the perl
7444 assembly language builder. If this argument exists and is set
7445 to 1 it signals that the assembler should use a symbol whose
7446 scope is the entire file, not just the current function. This
7447 is needed with MASM which uses the format label:: for this scope.
7448 [Steve Henson, pointed out by Peter Runestig <peter@runestig.com>]
7449
7450 *) Change the ASN1 types so they are typedefs by default. Before
7451 almost all types were #define'd to ASN1_STRING which was causing
7452 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING)
7453 for example.
7454 [Steve Henson]
7455
7456 *) Change names of new functions to the new get1/get0 naming
7457 convention: After 'get1', the caller owns a reference count
7458 and has to call ..._free; 'get0' returns a pointer to some
7459 data structure without incrementing reference counters.
7460 (Some of the existing 'get' functions increment a reference
7461 counter, some don't.)
7462 Similarly, 'set1' and 'add1' functions increase reference
7463 counters or duplicate objects.
7464 [Steve Henson]
7465
7466 *) Allow for the possibility of temp RSA key generation failure:
7467 the code used to assume it always worked and crashed on failure.
7468 [Steve Henson]
7469
7470 *) Fix potential buffer overrun problem in BIO_printf().
7471 [Ulf Möller, using public domain code by Patrick Powell; problem
7472 pointed out by David Sacerdote <das33@cornell.edu>]
7473
7474 *) Support EGD <http://www.lothar.com/tech/crypto/>. New functions
7475 RAND_egd() and RAND_status(). In the command line application,
7476 the EGD socket can be specified like a seed file using RANDFILE
7477 or -rand.
7478 [Ulf Möller]
7479
7480 *) Allow the string CERTIFICATE to be tolerated in PKCS#7 structures.
7481 Some CAs (e.g. Verisign) distribute certificates in this form.
7482 [Steve Henson]
7483
7484 *) Remove the SSL_ALLOW_ADH compile option and set the default cipher
7485 list to exclude them. This means that no special compilation option
7486 is needed to use anonymous DH: it just needs to be included in the
7487 cipher list.
7488 [Steve Henson]
7489
7490 *) Change the EVP_MD_CTX_type macro so its meaning consistent with
7491 EVP_MD_type. The old functionality is available in a new macro called
7492 EVP_MD_md(). Change code that uses it and update docs.
7493 [Steve Henson]
7494
7495 *) ..._ctrl functions now have corresponding ..._callback_ctrl functions
7496 where the 'void *' argument is replaced by a function pointer argument.
7497 Previously 'void *' was abused to point to functions, which works on
7498 many platforms, but is not correct. As these functions are usually
7499 called by macros defined in OpenSSL header files, most source code
7500 should work without changes.
7501 [Richard Levitte]
7502
7503 *) <openssl/opensslconf.h> (which is created by Configure) now contains
7504 sections with information on -D... compiler switches used for
7505 compiling the library so that applications can see them. To enable
7506 one of these sections, a pre-processor symbol OPENSSL_..._DEFINES
7507 must be defined. E.g.,
7508 #define OPENSSL_ALGORITHM_DEFINES
7509 #include <openssl/opensslconf.h>
7510 defines all pertinent NO_<algo> symbols, such as NO_IDEA, NO_RSA, etc.
7511 [Richard Levitte, Ulf and Bodo Möller]
7512
7513 *) Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS
7514 record layer.
7515 [Bodo Moeller]
7516
7517 *) Change the 'other' type in certificate aux info to a STACK_OF
7518 X509_ALGOR. Although not an AlgorithmIdentifier as such it has
7519 the required ASN1 format: arbitrary types determined by an OID.
7520 [Steve Henson]
7521
7522 *) Add some PEM_write_X509_REQ_NEW() functions and a command line
7523 argument to 'req'. This is not because the function is newer or
7524 better than others it just uses the work 'NEW' in the certificate
7525 request header lines. Some software needs this.
7526 [Steve Henson]
7527
7528 *) Reorganise password command line arguments: now passwords can be
7529 obtained from various sources. Delete the PEM_cb function and make
7530 it the default behaviour: i.e. if the callback is NULL and the
7531 usrdata argument is not NULL interpret it as a null terminated pass
7532 phrase. If usrdata and the callback are NULL then the pass phrase
7533 is prompted for as usual.
7534 [Steve Henson]
7535
7536 *) Add support for the Compaq Atalla crypto accelerator. If it is installed,
7537 the support is automatically enabled. The resulting binaries will
7538 autodetect the card and use it if present.
7539 [Ben Laurie and Compaq Inc.]
7540
7541 *) Work around for Netscape hang bug. This sends certificate request
7542 and server done in one record. Since this is perfectly legal in the
7543 SSL/TLS protocol it isn't a "bug" option and is on by default. See
7544 the bugs/SSLv3 entry for more info.
7545 [Steve Henson]
7546
7547 *) HP-UX tune-up: new unified configs, HP C compiler bug workaround.
7548 [Andy Polyakov]
7549
7550 *) Add -rand argument to smime and pkcs12 applications and read/write
7551 of seed file.
7552 [Steve Henson]
7553
7554 *) New 'passwd' tool for crypt(3) and apr1 password hashes.
7555 [Bodo Moeller]
7556
7557 *) Add command line password options to the remaining applications.
7558 [Steve Henson]
7559
7560 *) Bug fix for BN_div_recp() for numerators with an even number of
7561 bits.
7562 [Ulf Möller]
7563
7564 *) More tests in bntest.c, and changed test_bn output.
7565 [Ulf Möller]
7566
7567 *) ./config recognizes MacOS X now.
7568 [Andy Polyakov]
7569
7570 *) Bug fix for BN_div() when the first words of num and divsor are
7571 equal (it gave wrong results if (rem=(n1-q*d0)&BN_MASK2) < d0).
7572 [Ulf Möller]
7573
7574 *) Add support for various broken PKCS#8 formats, and command line
7575 options to produce them.
7576 [Steve Henson]
7577
7578 *) New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to
7579 get temporary BIGNUMs from a BN_CTX.
7580 [Ulf Möller]
7581
7582 *) Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont()
7583 for p == 0.
7584 [Ulf Möller]
7585
7586 *) Change the SSLeay_add_all_*() functions to OpenSSL_add_all_*() and
7587 include a #define from the old name to the new. The original intent
7588 was that statically linked binaries could for example just call
7589 SSLeay_add_all_ciphers() to just add ciphers to the table and not
7590 link with digests. This never worked becayse SSLeay_add_all_digests()
7591 and SSLeay_add_all_ciphers() were in the same source file so calling
7592 one would link with the other. They are now in separate source files.
7593 [Steve Henson]
7594
7595 *) Add a new -notext option to 'ca' and a -pubkey option to 'spkac'.
7596 [Steve Henson]
7597
7598 *) Use a less unusual form of the Miller-Rabin primality test (it used
7599 a binary algorithm for exponentiation integrated into the Miller-Rabin
7600 loop, our standard modexp algorithms are faster).
7601 [Bodo Moeller]
7602
7603 *) Support for the EBCDIC character set completed.
7604 [Martin Kraemer <Martin.Kraemer@Mch.SNI.De>]
7605
7606 *) Source code cleanups: use const where appropriate, eliminate casts,
7607 use void * instead of char * in lhash.
7608 [Ulf Möller]
7609
7610 *) Bugfix: ssl3_send_server_key_exchange was not restartable
7611 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of
7612 this the server could overwrite ephemeral keys that the client
7613 has already seen).
7614 [Bodo Moeller]
7615
7616 *) Turn DSA_is_prime into a macro that calls BN_is_prime,
7617 using 50 iterations of the Rabin-Miller test.
7618
7619 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50
7620 iterations of the Rabin-Miller test as required by the appendix
7621 to FIPS PUB 186[-1]) instead of DSA_is_prime.
7622 As BN_is_prime_fasttest includes trial division, DSA parameter
7623 generation becomes much faster.
7624
7625 This implies a change for the callback functions in DSA_is_prime
7626 and DSA_generate_parameters: The callback function is called once
7627 for each positive witness in the Rabin-Miller test, not just
7628 occasionally in the inner loop; and the parameters to the
7629 callback function now provide an iteration count for the outer
7630 loop rather than for the current invocation of the inner loop.
7631 DSA_generate_parameters additionally can call the callback
7632 function with an 'iteration count' of -1, meaning that a
7633 candidate has passed the trial division test (when q is generated
7634 from an application-provided seed, trial division is skipped).
7635 [Bodo Moeller]
7636
7637 *) New function BN_is_prime_fasttest that optionally does trial
7638 division before starting the Rabin-Miller test and has
7639 an additional BN_CTX * argument (whereas BN_is_prime always
7640 has to allocate at least one BN_CTX).
7641 'callback(1, -1, cb_arg)' is called when a number has passed the
7642 trial division stage.
7643 [Bodo Moeller]
7644
7645 *) Fix for bug in CRL encoding. The validity dates weren't being handled
7646 as ASN1_TIME.
7647 [Steve Henson]
7648
7649 *) New -pkcs12 option to CA.pl script to write out a PKCS#12 file.
7650 [Steve Henson]
7651
7652 *) New function BN_pseudo_rand().
7653 [Ulf Möller]
7654
7655 *) Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable)
7656 bignum version of BN_from_montgomery() with the working code from
7657 SSLeay 0.9.0 (the word based version is faster anyway), and clean up
7658 the comments.
7659 [Ulf Möller]
7660
7661 *) Avoid a race condition in s2_clnt.c (function get_server_hello) that
7662 made it impossible to use the same SSL_SESSION data structure in
7663 SSL2 clients in multiple threads.
7664 [Bodo Moeller]
7665
7666 *) The return value of RAND_load_file() no longer counts bytes obtained
7667 by stat(). RAND_load_file(..., -1) is new and uses the complete file
7668 to seed the PRNG (previously an explicit byte count was required).
7669 [Ulf Möller, Bodo Möller]
7670
7671 *) Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
7672 used (char *) instead of (void *) and had casts all over the place.
7673 [Steve Henson]
7674
7675 *) Make BN_generate_prime() return NULL on error if ret!=NULL.
7676 [Ulf Möller]
7677
7678 *) Retain source code compatibility for BN_prime_checks macro:
7679 BN_is_prime(..., BN_prime_checks, ...) now uses
7680 BN_prime_checks_for_size to determine the appropriate number of
7681 Rabin-Miller iterations.
7682 [Ulf Möller]
7683
7684 *) Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
7685 DH_CHECK_P_NOT_SAFE_PRIME.
7686 (Check if this is true? OpenPGP calls them "strong".)
7687 [Ulf Möller]
7688
7689 *) Merge the functionality of "dh" and "gendh" programs into a new program
7690 "dhparam". The old programs are retained for now but will handle DH keys
7691 (instead of parameters) in future.
7692 [Steve Henson]
7693
7694 *) Make the ciphers, s_server and s_client programs check the return values
7695 when a new cipher list is set.
7696 [Steve Henson]
7697
7698 *) Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
7699 ciphers. Before when the 56bit ciphers were enabled the sorting was
7700 wrong.
7701
7702 The syntax for the cipher sorting has been extended to support sorting by
7703 cipher-strength (using the strength_bits hard coded in the tables).
7704 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
7705
7706 Fix a bug in the cipher-command parser: when supplying a cipher command
7707 string with an "undefined" symbol (neither command nor alphanumeric
7708 [A-Za-z0-9], ssl_set_cipher_list used to hang in an endless loop. Now
7709 an error is flagged.
7710
7711 Due to the strength-sorting extension, the code of the
7712 ssl_create_cipher_list() function was completely rearranged. I hope that
7713 the readability was also increased :-)
7714 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
7715
7716 *) Minor change to 'x509' utility. The -CAcreateserial option now uses 1
7717 for the first serial number and places 2 in the serial number file. This
7718 avoids problems when the root CA is created with serial number zero and
7719 the first user certificate has the same issuer name and serial number
7720 as the root CA.
7721 [Steve Henson]
7722
7723 *) Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
7724 the new code. Add documentation for this stuff.
7725 [Steve Henson]
7726
7727 *) Changes to X509_ATTRIBUTE utilities. These have been renamed from
7728 X509_*() to X509at_*() on the grounds that they don't handle X509
7729 structures and behave in an analagous way to the X509v3 functions:
7730 they shouldn't be called directly but wrapper functions should be used
7731 instead.
7732
7733 So we also now have some wrapper functions that call the X509at functions
7734 when passed certificate requests. (TO DO: similar things can be done with
7735 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
7736 things. Some of these need some d2i or i2d and print functionality
7737 because they handle more complex structures.)
7738 [Steve Henson]
7739
7740 *) Add missing #ifndefs that caused missing symbols when building libssl
7741 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
7742 NO_RSA in ssl/s2*.c.
7743 [Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller]
7744
7745 *) Precautions against using the PRNG uninitialized: RAND_bytes() now
7746 has a return value which indicates the quality of the random data
7747 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
7748 error queue. New function RAND_pseudo_bytes() generates output that is
7749 guaranteed to be unique but not unpredictable. RAND_add is like
7750 RAND_seed, but takes an extra argument for an entropy estimate
7751 (RAND_seed always assumes full entropy).
7752 [Ulf Möller]
7753
7754 *) Do more iterations of Rabin-Miller probable prime test (specifically,
7755 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
7756 instead of only 2 for all lengths; see BN_prime_checks_for_size definition
7757 in crypto/bn/bn_prime.c for the complete table). This guarantees a
7758 false-positive rate of at most 2^-80 for random input.
7759 [Bodo Moeller]
7760
7761 *) Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
7762 [Bodo Moeller]
7763
7764 *) New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain
7765 in the 0.9.5 release), this returns the chain
7766 from an X509_CTX structure with a dup of the stack and all
7767 the X509 reference counts upped: so the stack will exist
7768 after X509_CTX_cleanup() has been called. Modify pkcs12.c
7769 to use this.
7770
7771 Also make SSL_SESSION_print() print out the verify return
7772 code.
7773 [Steve Henson]
7774
7775 *) Add manpage for the pkcs12 command. Also change the default
7776 behaviour so MAC iteration counts are used unless the new
7777 -nomaciter option is used. This improves file security and
7778 only older versions of MSIE (4.0 for example) need it.
7779 [Steve Henson]
7780
7781 *) Honor the no-xxx Configure options when creating .DEF files.
7782 [Ulf Möller]
7783
7784 *) Add PKCS#10 attributes to field table: challengePassword,
7785 unstructuredName and unstructuredAddress. These are taken from
7786 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
7787 international characters are used.
7788
7789 More changes to X509_ATTRIBUTE code: allow the setting of types
7790 based on strings. Remove the 'loc' parameter when adding
7791 attributes because these will be a SET OF encoding which is sorted
7792 in ASN1 order.
7793 [Steve Henson]
7794
7795 *) Initial changes to the 'req' utility to allow request generation
7796 automation. This will allow an application to just generate a template
7797 file containing all the field values and have req construct the
7798 request.
7799
7800 Initial support for X509_ATTRIBUTE handling. Stacks of these are
7801 used all over the place including certificate requests and PKCS#7
7802 structures. They are currently handled manually where necessary with
7803 some primitive wrappers for PKCS#7. The new functions behave in a
7804 manner analogous to the X509 extension functions: they allow
7805 attributes to be looked up by NID and added.
7806
7807 Later something similar to the X509V3 code would be desirable to
7808 automatically handle the encoding, decoding and printing of the
7809 more complex types. The string types like challengePassword can
7810 be handled by the string table functions.
7811
7812 Also modified the multi byte string table handling. Now there is
7813 a 'global mask' which masks out certain types. The table itself
7814 can use the flag STABLE_NO_MASK to ignore the mask setting: this
7815 is useful when for example there is only one permissible type
7816 (as in countryName) and using the mask might result in no valid
7817 types at all.
7818 [Steve Henson]
7819
7820 *) Clean up 'Finished' handling, and add functions SSL_get_finished and
7821 SSL_get_peer_finished to allow applications to obtain the latest
7822 Finished messages sent to the peer or expected from the peer,
7823 respectively. (SSL_get_peer_finished is usually the Finished message
7824 actually received from the peer, otherwise the protocol will be aborted.)
7825
7826 As the Finished message are message digests of the complete handshake
7827 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
7828 be used for external authentication procedures when the authentication
7829 provided by SSL/TLS is not desired or is not enough.
7830 [Bodo Moeller]
7831
7832 *) Enhanced support for Alpha Linux is added. Now ./config checks if
7833 the host supports BWX extension and if Compaq C is present on the
7834 $PATH. Just exploiting of the BWX extension results in 20-30%
7835 performance kick for some algorithms, e.g. DES and RC4 to mention
7836 a couple. Compaq C in turn generates ~20% faster code for MD5 and
7837 SHA1.
7838 [Andy Polyakov]
7839
7840 *) Add support for MS "fast SGC". This is arguably a violation of the
7841 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
7842 weak crypto and after checking the certificate is SGC a second one
7843 with strong crypto. MS SGC stops the first handshake after receiving
7844 the server certificate message and sends a second client hello. Since
7845 a server will typically do all the time consuming operations before
7846 expecting any further messages from the client (server key exchange
7847 is the most expensive) there is little difference between the two.
7848
7849 To get OpenSSL to support MS SGC we have to permit a second client
7850 hello message after we have sent server done. In addition we have to
7851 reset the MAC if we do get this second client hello.
7852 [Steve Henson]
7853
7854 *) Add a function 'd2i_AutoPrivateKey()' this will automatically decide
7855 if a DER encoded private key is RSA or DSA traditional format. Changed
7856 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
7857 format DER encoded private key. Newer code should use PKCS#8 format which
7858 has the key type encoded in the ASN1 structure. Added DER private key
7859 support to pkcs8 application.
7860 [Steve Henson]
7861
7862 *) SSL 3/TLS 1 servers now don't request certificates when an anonymous
7863 ciphersuites has been selected (as required by the SSL 3/TLS 1
7864 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
7865 is set, we interpret this as a request to violate the specification
7866 (the worst that can happen is a handshake failure, and 'correct'
7867 behaviour would result in a handshake failure anyway).
7868 [Bodo Moeller]
7869
7870 *) In SSL_CTX_add_session, take into account that there might be multiple
7871 SSL_SESSION structures with the same session ID (e.g. when two threads
7872 concurrently obtain them from an external cache).
7873 The internal cache can handle only one SSL_SESSION with a given ID,
7874 so if there's a conflict, we now throw out the old one to achieve
7875 consistency.
7876 [Bodo Moeller]
7877
7878 *) Add OIDs for idea and blowfish in CBC mode. This will allow both
7879 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
7880 some routines that use cipher OIDs: some ciphers do not have OIDs
7881 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
7882 example.
7883 [Steve Henson]
7884
7885 *) Simplify the trust setting structure and code. Now we just have
7886 two sequences of OIDs for trusted and rejected settings. These will
7887 typically have values the same as the extended key usage extension
7888 and any application specific purposes.
7889
7890 The trust checking code now has a default behaviour: it will just
7891 check for an object with the same NID as the passed id. Functions can
7892 be provided to override either the default behaviour or the behaviour
7893 for a given id. SSL client, server and email already have functions
7894 in place for compatibility: they check the NID and also return "trusted"
7895 if the certificate is self signed.
7896 [Steve Henson]
7897
7898 *) Add d2i,i2d bio/fp functions for PrivateKey: these convert the
7899 traditional format into an EVP_PKEY structure.
7900 [Steve Henson]
7901
7902 *) Add a password callback function PEM_cb() which either prompts for
7903 a password if usr_data is NULL or otherwise assumes it is a null
7904 terminated password. Allow passwords to be passed on command line
7905 environment or config files in a few more utilities.
7906 [Steve Henson]
7907
7908 *) Add a bunch of DER and PEM functions to handle PKCS#8 format private
7909 keys. Add some short names for PKCS#8 PBE algorithms and allow them
7910 to be specified on the command line for the pkcs8 and pkcs12 utilities.
7911 Update documentation.
7912 [Steve Henson]
7913
7914 *) Support for ASN1 "NULL" type. This could be handled before by using
7915 ASN1_TYPE but there wasn't any function that would try to read a NULL
7916 and produce an error if it couldn't. For compatibility we also have
7917 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
7918 don't allocate anything because they don't need to.
7919 [Steve Henson]
7920
7921 *) Initial support for MacOS is now provided. Examine INSTALL.MacOS
7922 for details.
7923 [Andy Polyakov, Roy Woods <roy@centicsystems.ca>]
7924
7925 *) Rebuild of the memory allocation routines used by OpenSSL code and
7926 possibly others as well. The purpose is to make an interface that
7927 provide hooks so anyone can build a separate set of allocation and
7928 deallocation routines to be used by OpenSSL, for example memory
7929 pool implementations, or something else, which was previously hard
7930 since Malloc(), Realloc() and Free() were defined as macros having
7931 the values malloc, realloc and free, respectively (except for Win32
7932 compilations). The same is provided for memory debugging code.
7933 OpenSSL already comes with functionality to find memory leaks, but
7934 this gives people a chance to debug other memory problems.
7935
7936 With these changes, a new set of functions and macros have appeared:
7937
7938 CRYPTO_set_mem_debug_functions() [F]
7939 CRYPTO_get_mem_debug_functions() [F]
7940 CRYPTO_dbg_set_options() [F]
7941 CRYPTO_dbg_get_options() [F]
7942 CRYPTO_malloc_debug_init() [M]
7943
7944 The memory debug functions are NULL by default, unless the library
7945 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
7946 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
7947 gives the standard debugging functions that come with OpenSSL) or
7948 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
7949 provided by the library user) must be used. When the standard
7950 debugging functions are used, CRYPTO_dbg_set_options can be used to
7951 request additional information:
7952 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
7953 the CRYPTO_MDEBUG_xxx macro when compiling the library.
7954
7955 Also, things like CRYPTO_set_mem_functions will always give the
7956 expected result (the new set of functions is used for allocation
7957 and deallocation) at all times, regardless of platform and compiler
7958 options.
7959
7960 To finish it up, some functions that were never use in any other
7961 way than through macros have a new API and new semantic:
7962
7963 CRYPTO_dbg_malloc()
7964 CRYPTO_dbg_realloc()
7965 CRYPTO_dbg_free()
7966
7967 All macros of value have retained their old syntax.
7968 [Richard Levitte and Bodo Moeller]
7969
7970 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
7971 ordering of SMIMECapabilities wasn't in "strength order" and there
7972 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
7973 algorithm.
7974 [Steve Henson]
7975
7976 *) Some ASN1 types with illegal zero length encoding (INTEGER,
7977 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
7978 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson]
7979
7980 *) Merge in my S/MIME library for OpenSSL. This provides a simple
7981 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
7982 functionality to handle multipart/signed properly) and a utility
7983 called 'smime' to call all this stuff. This is based on code I
7984 originally wrote for Celo who have kindly allowed it to be
7985 included in OpenSSL.
7986 [Steve Henson]
7987
7988 *) Add variants des_set_key_checked and des_set_key_unchecked of
7989 des_set_key (aka des_key_sched). Global variable des_check_key
7990 decides which of these is called by des_set_key; this way
7991 des_check_key behaves as it always did, but applications and
7992 the library itself, which was buggy for des_check_key == 1,
7993 have a cleaner way to pick the version they need.
7994 [Bodo Moeller]
7995
7996 *) New function PKCS12_newpass() which changes the password of a
7997 PKCS12 structure.
7998 [Steve Henson]
7999
8000 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and
8001 dynamic mix. In both cases the ids can be used as an index into the
8002 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
8003 functions so they accept a list of the field values and the
8004 application doesn't need to directly manipulate the X509_TRUST
8005 structure.
8006 [Steve Henson]
8007
8008 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
8009 need initialising.
8010 [Steve Henson]
8011
8012 *) Modify the way the V3 extension code looks up extensions. This now
8013 works in a similar way to the object code: we have some "standard"
8014 extensions in a static table which is searched with OBJ_bsearch()
8015 and the application can add dynamic ones if needed. The file
8016 crypto/x509v3/ext_dat.h now has the info: this file needs to be
8017 updated whenever a new extension is added to the core code and kept
8018 in ext_nid order. There is a simple program 'tabtest.c' which checks
8019 this. New extensions are not added too often so this file can readily
8020 be maintained manually.
8021
8022 There are two big advantages in doing things this way. The extensions
8023 can be looked up immediately and no longer need to be "added" using
8024 X509V3_add_standard_extensions(): this function now does nothing.
8025 [Side note: I get *lots* of email saying the extension code doesn't
8026 work because people forget to call this function]
8027 Also no dynamic allocation is done unless new extensions are added:
8028 so if we don't add custom extensions there is no need to call
8029 X509V3_EXT_cleanup().
8030 [Steve Henson]
8031
8032 *) Modify enc utility's salting as follows: make salting the default. Add a
8033 magic header, so unsalted files fail gracefully instead of just decrypting
8034 to garbage. This is because not salting is a big security hole, so people
8035 should be discouraged from doing it.
8036 [Ben Laurie]
8037
8038 *) Fixes and enhancements to the 'x509' utility. It allowed a message
8039 digest to be passed on the command line but it only used this
8040 parameter when signing a certificate. Modified so all relevant
8041 operations are affected by the digest parameter including the
8042 -fingerprint and -x509toreq options. Also -x509toreq choked if a
8043 DSA key was used because it didn't fix the digest.
8044 [Steve Henson]
8045
8046 *) Initial certificate chain verify code. Currently tests the untrusted
8047 certificates for consistency with the verify purpose (which is set
8048 when the X509_STORE_CTX structure is set up) and checks the pathlength.
8049
8050 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
8051 this is because it will reject chains with invalid extensions whereas
8052 every previous version of OpenSSL and SSLeay made no checks at all.
8053
8054 Trust code: checks the root CA for the relevant trust settings. Trust
8055 settings have an initial value consistent with the verify purpose: e.g.
8056 if the verify purpose is for SSL client use it expects the CA to be
8057 trusted for SSL client use. However the default value can be changed to
8058 permit custom trust settings: one example of this would be to only trust
8059 certificates from a specific "secure" set of CAs.
8060
8061 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
8062 which should be used for version portability: especially since the
8063 verify structure is likely to change more often now.
8064
8065 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
8066 to set them. If not set then assume SSL clients will verify SSL servers
8067 and vice versa.
8068
8069 Two new options to the verify program: -untrusted allows a set of
8070 untrusted certificates to be passed in and -purpose which sets the
8071 intended purpose of the certificate. If a purpose is set then the
8072 new chain verify code is used to check extension consistency.
8073 [Steve Henson]
8074
8075 *) Support for the authority information access extension.
8076 [Steve Henson]
8077
8078 *) Modify RSA and DSA PEM read routines to transparently handle
8079 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
8080 public keys in a format compatible with certificate
8081 SubjectPublicKeyInfo structures. Unfortunately there were already
8082 functions called *_PublicKey_* which used various odd formats so
8083 these are retained for compatibility: however the DSA variants were
8084 never in a public release so they have been deleted. Changed dsa/rsa
8085 utilities to handle the new format: note no releases ever handled public
8086 keys so we should be OK.
8087
8088 The primary motivation for this change is to avoid the same fiasco
8089 that dogs private keys: there are several incompatible private key
8090 formats some of which are standard and some OpenSSL specific and
8091 require various evil hacks to allow partial transparent handling and
8092 even then it doesn't work with DER formats. Given the option anything
8093 other than PKCS#8 should be dumped: but the other formats have to
8094 stay in the name of compatibility.
8095
8096 With public keys and the benefit of hindsight one standard format
8097 is used which works with EVP_PKEY, RSA or DSA structures: though
8098 it clearly returns an error if you try to read the wrong kind of key.
8099
8100 Added a -pubkey option to the 'x509' utility to output the public key.
8101 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*()
8102 (renamed to EVP_PKEY_get1_*() in the OpenSSL 0.9.5 release) and add
8103 EVP_PKEY_rset_*() functions (renamed to EVP_PKEY_set1_*())
8104 that do the same as the EVP_PKEY_assign_*() except they up the
8105 reference count of the added key (they don't "swallow" the
8106 supplied key).
8107 [Steve Henson]
8108
8109 *) Fixes to crypto/x509/by_file.c the code to read in certificates and
8110 CRLs would fail if the file contained no certificates or no CRLs:
8111 added a new function to read in both types and return the number
8112 read: this means that if none are read it will be an error. The
8113 DER versions of the certificate and CRL reader would always fail
8114 because it isn't possible to mix certificates and CRLs in DER format
8115 without choking one or the other routine. Changed this to just read
8116 a certificate: this is the best we can do. Also modified the code
8117 in apps/verify.c to take notice of return codes: it was previously
8118 attempting to read in certificates from NULL pointers and ignoring
8119 any errors: this is one reason why the cert and CRL reader seemed
8120 to work. It doesn't check return codes from the default certificate
8121 routines: these may well fail if the certificates aren't installed.
8122 [Steve Henson]
8123
8124 *) Code to support otherName option in GeneralName.
8125 [Steve Henson]
8126
8127 *) First update to verify code. Change the verify utility
8128 so it warns if it is passed a self signed certificate:
8129 for consistency with the normal behaviour. X509_verify
8130 has been modified to it will now verify a self signed
8131 certificate if *exactly* the same certificate appears
8132 in the store: it was previously impossible to trust a
8133 single self signed certificate. This means that:
8134 openssl verify ss.pem
8135 now gives a warning about a self signed certificate but
8136 openssl verify -CAfile ss.pem ss.pem
8137 is OK.
8138 [Steve Henson]
8139
8140 *) For servers, store verify_result in SSL_SESSION data structure
8141 (and add it to external session representation).
8142 This is needed when client certificate verifications fails,
8143 but an application-provided verification callback (set by
8144 SSL_CTX_set_cert_verify_callback) allows accepting the session
8145 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
8146 but returns 1): When the session is reused, we have to set
8147 ssl->verify_result to the appropriate error code to avoid
8148 security holes.
8149 [Bodo Moeller, problem pointed out by Lutz Jaenicke]
8150
8151 *) Fix a bug in the new PKCS#7 code: it didn't consider the
8152 case in PKCS7_dataInit() where the signed PKCS7 structure
8153 didn't contain any existing data because it was being created.
8154 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson]
8155
8156 *) Add a salt to the key derivation routines in enc.c. This
8157 forms the first 8 bytes of the encrypted file. Also add a
8158 -S option to allow a salt to be input on the command line.
8159 [Steve Henson]
8160
8161 *) New function X509_cmp(). Oddly enough there wasn't a function
8162 to compare two certificates. We do this by working out the SHA1
8163 hash and comparing that. X509_cmp() will be needed by the trust
8164 code.
8165 [Steve Henson]
8166
8167 *) SSL_get1_session() is like SSL_get_session(), but increments
8168 the reference count in the SSL_SESSION returned.
8169 [Geoff Thorpe <geoff@eu.c2.net>]
8170
8171 *) Fix for 'req': it was adding a null to request attributes.
8172 Also change the X509_LOOKUP and X509_INFO code to handle
8173 certificate auxiliary information.
8174 [Steve Henson]
8175
8176 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document
8177 the 'enc' command.
8178 [Steve Henson]
8179
8180 *) Add the possibility to add extra information to the memory leak
8181 detecting output, to form tracebacks, showing from where each
8182 allocation was originated: CRYPTO_push_info("constant string") adds
8183 the string plus current file name and line number to a per-thread
8184 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
8185 is like calling CYRPTO_pop_info() until the stack is empty.
8186 Also updated memory leak detection code to be multi-thread-safe.
8187 [Richard Levitte]
8188
8189 *) Add options -text and -noout to pkcs7 utility and delete the
8190 encryption options which never did anything. Update docs.
8191 [Steve Henson]
8192
8193 *) Add options to some of the utilities to allow the pass phrase
8194 to be included on either the command line (not recommended on
8195 OSes like Unix) or read from the environment. Update the
8196 manpages and fix a few bugs.
8197 [Steve Henson]
8198
8199 *) Add a few manpages for some of the openssl commands.
8200 [Steve Henson]
8201
8202 *) Fix the -revoke option in ca. It was freeing up memory twice,
8203 leaking and not finding already revoked certificates.
8204 [Steve Henson]
8205
8206 *) Extensive changes to support certificate auxiliary information.
8207 This involves the use of X509_CERT_AUX structure and X509_AUX
8208 functions. An X509_AUX function such as PEM_read_X509_AUX()
8209 can still read in a certificate file in the usual way but it
8210 will also read in any additional "auxiliary information". By
8211 doing things this way a fair degree of compatibility can be
8212 retained: existing certificates can have this information added
8213 using the new 'x509' options.
8214
8215 Current auxiliary information includes an "alias" and some trust
8216 settings. The trust settings will ultimately be used in enhanced
8217 certificate chain verification routines: currently a certificate
8218 can only be trusted if it is self signed and then it is trusted
8219 for all purposes.
8220 [Steve Henson]
8221
8222 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD).
8223 The problem was that one of the replacement routines had not been working
8224 since SSLeay releases. For now the offending routine has been replaced
8225 with non-optimised assembler. Even so, this now gives around 95%
8226 performance improvement for 1024 bit RSA signs.
8227 [Mark Cox]
8228
8229 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2
8230 handling. Most clients have the effective key size in bits equal to
8231 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
8232 A few however don't do this and instead use the size of the decrypted key
8233 to determine the RC2 key length and the AlgorithmIdentifier to determine
8234 the effective key length. In this case the effective key length can still
8235 be 40 bits but the key length can be 168 bits for example. This is fixed
8236 by manually forcing an RC2 key into the EVP_PKEY structure because the
8237 EVP code can't currently handle unusual RC2 key sizes: it always assumes
8238 the key length and effective key length are equal.
8239 [Steve Henson]
8240
8241 *) Add a bunch of functions that should simplify the creation of
8242 X509_NAME structures. Now you should be able to do:
8243 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
8244 and have it automatically work out the correct field type and fill in
8245 the structures. The more adventurous can try:
8246 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
8247 and it will (hopefully) work out the correct multibyte encoding.
8248 [Steve Henson]
8249
8250 *) Change the 'req' utility to use the new field handling and multibyte
8251 copy routines. Before the DN field creation was handled in an ad hoc
8252 way in req, ca, and x509 which was rather broken and didn't support
8253 BMPStrings or UTF8Strings. Since some software doesn't implement
8254 BMPStrings or UTF8Strings yet, they can be enabled using the config file
8255 using the dirstring_type option. See the new comment in the default
8256 openssl.cnf for more info.
8257 [Steve Henson]
8258
8259 *) Make crypto/rand/md_rand.c more robust:
8260 - Assure unique random numbers after fork().
8261 - Make sure that concurrent threads access the global counter and
8262 md serializably so that we never lose entropy in them
8263 or use exactly the same state in multiple threads.
8264 Access to the large state is not always serializable because
8265 the additional locking could be a performance killer, and
8266 md should be large enough anyway.
8267 [Bodo Moeller]
8268
8269 *) New file apps/app_rand.c with commonly needed functionality
8270 for handling the random seed file.
8271
8272 Use the random seed file in some applications that previously did not:
8273 ca,
8274 dsaparam -genkey (which also ignored its '-rand' option),
8275 s_client,
8276 s_server,
8277 x509 (when signing).
8278 Except on systems with /dev/urandom, it is crucial to have a random
8279 seed file at least for key creation, DSA signing, and for DH exchanges;
8280 for RSA signatures we could do without one.
8281
8282 gendh and gendsa (unlike genrsa) used to read only the first byte
8283 of each file listed in the '-rand' option. The function as previously
8284 found in genrsa is now in app_rand.c and is used by all programs
8285 that support '-rand'.
8286 [Bodo Moeller]
8287
8288 *) In RAND_write_file, use mode 0600 for creating files;
8289 don't just chmod when it may be too late.
8290 [Bodo Moeller]
8291
8292 *) Report an error from X509_STORE_load_locations
8293 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
8294 [Bill Perry]
8295
8296 *) New function ASN1_mbstring_copy() this copies a string in either
8297 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
8298 into an ASN1_STRING type. A mask of permissible types is passed
8299 and it chooses the "minimal" type to use or an error if not type
8300 is suitable.
8301 [Steve Henson]
8302
8303 *) Add function equivalents to the various macros in asn1.h. The old
8304 macros are retained with an M_ prefix. Code inside the library can
8305 use the M_ macros. External code (including the openssl utility)
8306 should *NOT* in order to be "shared library friendly".
8307 [Steve Henson]
8308
8309 *) Add various functions that can check a certificate's extensions
8310 to see if it usable for various purposes such as SSL client,
8311 server or S/MIME and CAs of these types. This is currently
8312 VERY EXPERIMENTAL but will ultimately be used for certificate chain
8313 verification. Also added a -purpose flag to x509 utility to
8314 print out all the purposes.
8315 [Steve Henson]
8316
8317 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated
8318 functions.
8319 [Steve Henson]
8320
8321 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search
8322 for, obtain and decode and extension and obtain its critical flag.
8323 This allows all the necessary extension code to be handled in a
8324 single function call.
8325 [Steve Henson]
8326
8327 *) RC4 tune-up featuring 30-40% performance improvement on most RISC
8328 platforms. See crypto/rc4/rc4_enc.c for further details.
8329 [Andy Polyakov]
8330
8331 *) New -noout option to asn1parse. This causes no output to be produced
8332 its main use is when combined with -strparse and -out to extract data
8333 from a file (which may not be in ASN.1 format).
8334 [Steve Henson]
8335
8336 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer
8337 when producing the local key id.
8338 [Richard Levitte <levitte@stacken.kth.se>]
8339
8340 *) New option -dhparam in s_server. This allows a DH parameter file to be
8341 stated explicitly. If it is not stated then it tries the first server
8342 certificate file. The previous behaviour hard coded the filename
8343 "server.pem".
8344 [Steve Henson]
8345
8346 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow
8347 a public key to be input or output. For example:
8348 openssl rsa -in key.pem -pubout -out pubkey.pem
8349 Also added necessary DSA public key functions to handle this.
8350 [Steve Henson]
8351
8352 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
8353 in the message. This was handled by allowing
8354 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
8355 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>]
8356
8357 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
8358 to the end of the strings whereas this didn't. This would cause problems
8359 if strings read with d2i_ASN1_bytes() were later modified.
8360 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>]
8361
8362 *) Fix for base64 decode bug. When a base64 bio reads only one line of
8363 data and it contains EOF it will end up returning an error. This is
8364 caused by input 46 bytes long. The cause is due to the way base64
8365 BIOs find the start of base64 encoded data. They do this by trying a
8366 trial decode on each line until they find one that works. When they
8367 do a flag is set and it starts again knowing it can pass all the
8368 data directly through the decoder. Unfortunately it doesn't reset
8369 the context it uses. This means that if EOF is reached an attempt
8370 is made to pass two EOFs through the context and this causes the
8371 resulting error. This can also cause other problems as well. As is
8372 usual with these problems it takes *ages* to find and the fix is
8373 trivial: move one line.
8374 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ]
8375
8376 *) Ugly workaround to get s_client and s_server working under Windows. The
8377 old code wouldn't work because it needed to select() on sockets and the
8378 tty (for keypresses and to see if data could be written). Win32 only
8379 supports select() on sockets so we select() with a 1s timeout on the
8380 sockets and then see if any characters are waiting to be read, if none
8381 are present then we retry, we also assume we can always write data to
8382 the tty. This isn't nice because the code then blocks until we've
8383 received a complete line of data and it is effectively polling the
8384 keyboard at 1s intervals: however it's quite a bit better than not
8385 working at all :-) A dedicated Windows application might handle this
8386 with an event loop for example.
8387 [Steve Henson]
8388
8389 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
8390 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
8391 will be called when RSA_sign() and RSA_verify() are used. This is useful
8392 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
8393 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
8394 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
8395 This necessitated the support of an extra signature type NID_md5_sha1
8396 for SSL signatures and modifications to the SSL library to use it instead
8397 of calling RSA_public_decrypt() and RSA_private_encrypt().
8398 [Steve Henson]
8399
8400 *) Add new -verify -CAfile and -CApath options to the crl program, these
8401 will lookup a CRL issuers certificate and verify the signature in a
8402 similar way to the verify program. Tidy up the crl program so it
8403 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
8404 less strict. It will now permit CRL extensions even if it is not
8405 a V2 CRL: this will allow it to tolerate some broken CRLs.
8406 [Steve Henson]
8407
8408 *) Initialize all non-automatic variables each time one of the openssl
8409 sub-programs is started (this is necessary as they may be started
8410 multiple times from the "OpenSSL>" prompt).
8411 [Lennart Bang, Bodo Moeller]
8412
8413 *) Preliminary compilation option RSA_NULL which disables RSA crypto without
8414 removing all other RSA functionality (this is what NO_RSA does). This
8415 is so (for example) those in the US can disable those operations covered
8416 by the RSA patent while allowing storage and parsing of RSA keys and RSA
8417 key generation.
8418 [Steve Henson]
8419
8420 *) Non-copying interface to BIO pairs.
8421 (still largely untested)
8422 [Bodo Moeller]
8423
8424 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive
8425 ASCII string. This was handled independently in various places before.
8426 [Steve Henson]
8427
8428 *) New functions UTF8_getc() and UTF8_putc() that parse and generate
8429 UTF8 strings a character at a time.
8430 [Steve Henson]
8431
8432 *) Use client_version from client hello to select the protocol
8433 (s23_srvr.c) and for RSA client key exchange verification
8434 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
8435 [Bodo Moeller]
8436
8437 *) Add various utility functions to handle SPKACs, these were previously
8438 handled by poking round in the structure internals. Added new function
8439 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
8440 print, verify and generate SPKACs. Based on an original idea from
8441 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
8442 [Steve Henson]
8443
8444 *) RIPEMD160 is operational on all platforms and is back in 'make test'.
8445 [Andy Polyakov]
8446
8447 *) Allow the config file extension section to be overwritten on the
8448 command line. Based on an original idea from Massimiliano Pala
8449 <madwolf@comune.modena.it>. The new option is called -extensions
8450 and can be applied to ca, req and x509. Also -reqexts to override
8451 the request extensions in req and -crlexts to override the crl extensions
8452 in ca.
8453 [Steve Henson]
8454
8455 *) Add new feature to the SPKAC handling in ca. Now you can include
8456 the same field multiple times by preceding it by "XXXX." for example:
8457 1.OU="Unit name 1"
8458 2.OU="Unit name 2"
8459 this is the same syntax as used in the req config file.
8460 [Steve Henson]
8461
8462 *) Allow certificate extensions to be added to certificate requests. These
8463 are specified in a 'req_extensions' option of the req section of the
8464 config file. They can be printed out with the -text option to req but
8465 are otherwise ignored at present.
8466 [Steve Henson]
8467
8468 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
8469 data read consists of only the final block it would not decrypted because
8470 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
8471 A misplaced 'break' also meant the decrypted final block might not be
8472 copied until the next read.
8473 [Steve Henson]
8474
8475 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
8476 a few extra parameters to the DH structure: these will be useful if
8477 for example we want the value of 'q' or implement X9.42 DH.
8478 [Steve Henson]
8479
8480 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and
8481 provides hooks that allow the default DSA functions or functions on a
8482 "per key" basis to be replaced. This allows hardware acceleration and
8483 hardware key storage to be handled without major modification to the
8484 library. Also added low level modexp hooks and CRYPTO_EX structure and
8485 associated functions.
8486 [Steve Henson]
8487
8488 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
8489 as "read only": it can't be written to and the buffer it points to will
8490 not be freed. Reading from a read only BIO is much more efficient than
8491 a normal memory BIO. This was added because there are several times when
8492 an area of memory needs to be read from a BIO. The previous method was
8493 to create a memory BIO and write the data to it, this results in two
8494 copies of the data and an O(n^2) reading algorithm. There is a new
8495 function BIO_new_mem_buf() which creates a read only memory BIO from
8496 an area of memory. Also modified the PKCS#7 routines to use read only
8497 memory BIOs.
8498 [Steve Henson]
8499
8500 *) Bugfix: ssl23_get_client_hello did not work properly when called in
8501 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
8502 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
8503 but a retry condition occured while trying to read the rest.
8504 [Bodo Moeller]
8505
8506 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as
8507 NID_pkcs7_encrypted by default: this was wrong since this should almost
8508 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
8509 the encrypted data type: this is a more sensible place to put it and it
8510 allows the PKCS#12 code to be tidied up that duplicated this
8511 functionality.
8512 [Steve Henson]
8513
8514 *) Changed obj_dat.pl script so it takes its input and output files on
8515 the command line. This should avoid shell escape redirection problems
8516 under Win32.
8517 [Steve Henson]
8518
8519 *) Initial support for certificate extension requests, these are included
8520 in things like Xenroll certificate requests. Included functions to allow
8521 extensions to be obtained and added.
8522 [Steve Henson]
8523
8524 *) -crlf option to s_client and s_server for sending newlines as
8525 CRLF (as required by many protocols).
8526 [Bodo Moeller]
8527
8528 Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
8529
8530 *) Install libRSAglue.a when OpenSSL is built with RSAref.
8531 [Ralf S. Engelschall]
8532
8533 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency.
8534 [Andrija Antonijevic <TheAntony2@bigfoot.com>]
8535
8536 *) Fix -startdate and -enddate (which was missing) arguments to 'ca'
8537 program.
8538 [Steve Henson]
8539
8540 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as
8541 DH parameters/keys (q is lost during that conversion, but the resulting
8542 DH parameters contain its length).
8543
8544 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
8545 much faster than DH_generate_parameters (which creates parameters
8546 where p = 2*q + 1), and also the smaller q makes DH computations
8547 much more efficient (160-bit exponentiation instead of 1024-bit
8548 exponentiation); so this provides a convenient way to support DHE
8549 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
8550 utter importance to use
8551 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
8552 or
8553 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
8554 when such DH parameters are used, because otherwise small subgroup
8555 attacks may become possible!
8556 [Bodo Moeller]
8557
8558 *) Avoid memory leak in i2d_DHparams.
8559 [Bodo Moeller]
8560
8561 *) Allow the -k option to be used more than once in the enc program:
8562 this allows the same encrypted message to be read by multiple recipients.
8563 [Steve Henson]
8564
8565 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
8566 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
8567 it will always use the numerical form of the OID, even if it has a short
8568 or long name.
8569 [Steve Henson]
8570
8571 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
8572 method only got called if p,q,dmp1,dmq1,iqmp components were present,
8573 otherwise bn_mod_exp was called. In the case of hardware keys for example
8574 no private key components need be present and it might store extra data
8575 in the RSA structure, which cannot be accessed from bn_mod_exp.
8576 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
8577 private key operations.
8578 [Steve Henson]
8579
8580 *) Added support for SPARC Linux.
8581 [Andy Polyakov]
8582
8583 *) pem_password_cb function type incompatibly changed from
8584 typedef int pem_password_cb(char *buf, int size, int rwflag);
8585 to
8586 ....(char *buf, int size, int rwflag, void *userdata);
8587 so that applications can pass data to their callbacks:
8588 The PEM[_ASN1]_{read,write}... functions and macros now take an
8589 additional void * argument, which is just handed through whenever
8590 the password callback is called.
8591 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller]
8592
8593 New function SSL_CTX_set_default_passwd_cb_userdata.
8594
8595 Compatibility note: As many C implementations push function arguments
8596 onto the stack in reverse order, the new library version is likely to
8597 interoperate with programs that have been compiled with the old
8598 pem_password_cb definition (PEM_whatever takes some data that
8599 happens to be on the stack as its last argument, and the callback
8600 just ignores this garbage); but there is no guarantee whatsoever that
8601 this will work.
8602
8603 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
8604 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
8605 problems not only on Windows, but also on some Unix platforms.
8606 To avoid problematic command lines, these definitions are now in an
8607 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
8608 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
8609 [Bodo Moeller]
8610
8611 *) MIPS III/IV assembler module is reimplemented.
8612 [Andy Polyakov]
8613
8614 *) More DES library cleanups: remove references to srand/rand and
8615 delete an unused file.
8616 [Ulf Möller]
8617
8618 *) Add support for the the free Netwide assembler (NASM) under Win32,
8619 since not many people have MASM (ml) and it can be hard to obtain.
8620 This is currently experimental but it seems to work OK and pass all
8621 the tests. Check out INSTALL.W32 for info.
8622 [Steve Henson]
8623
8624 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
8625 without temporary keys kept an extra copy of the server key,
8626 and connections with temporary keys did not free everything in case
8627 of an error.
8628 [Bodo Moeller]
8629
8630 *) New function RSA_check_key and new openssl rsa option -check
8631 for verifying the consistency of RSA keys.
8632 [Ulf Moeller, Bodo Moeller]
8633
8634 *) Various changes to make Win32 compile work:
8635 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
8636 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
8637 comparison" warnings.
8638 3. Add sk_<TYPE>_sort to DEF file generator and do make update.
8639 [Steve Henson]
8640
8641 *) Add a debugging option to PKCS#5 v2 key generation function: when
8642 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
8643 derived keys are printed to stderr.
8644 [Steve Henson]
8645
8646 *) Copy the flags in ASN1_STRING_dup().
8647 [Roman E. Pavlov <pre@mo.msk.ru>]
8648
8649 *) The x509 application mishandled signing requests containing DSA
8650 keys when the signing key was also DSA and the parameters didn't match.
8651
8652 It was supposed to omit the parameters when they matched the signing key:
8653 the verifying software was then supposed to automatically use the CA's
8654 parameters if they were absent from the end user certificate.
8655
8656 Omitting parameters is no longer recommended. The test was also
8657 the wrong way round! This was probably due to unusual behaviour in
8658 EVP_cmp_parameters() which returns 1 if the parameters match.
8659 This meant that parameters were omitted when they *didn't* match and
8660 the certificate was useless. Certificates signed with 'ca' didn't have
8661 this bug.
8662 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>]
8663
8664 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
8665 The interface is as follows:
8666 Applications can use
8667 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
8668 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
8669 "off" is now the default.
8670 The library internally uses
8671 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
8672 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
8673 to disable memory-checking temporarily.
8674
8675 Some inconsistent states that previously were possible (and were
8676 even the default) are now avoided.
8677
8678 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
8679 with each memory chunk allocated; this is occasionally more helpful
8680 than just having a counter.
8681
8682 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
8683
8684 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
8685 extensions.
8686 [Bodo Moeller]
8687
8688 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX),
8689 which largely parallels "options", but is for changing API behaviour,
8690 whereas "options" are about protocol behaviour.
8691 Initial "mode" flags are:
8692
8693 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
8694 a single record has been written.
8695 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
8696 retries use the same buffer location.
8697 (But all of the contents must be
8698 copied!)
8699 [Bodo Moeller]
8700
8701 *) Bugfix: SSL_set_options ignored its parameter, only SSL_CTX_set_options
8702 worked.
8703
8704 *) Fix problems with no-hmac etc.
8705 [Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>]
8706
8707 *) New functions RSA_get_default_method(), RSA_set_method() and
8708 RSA_get_method(). These allows replacement of RSA_METHODs without having
8709 to mess around with the internals of an RSA structure.
8710 [Steve Henson]
8711
8712 *) Fix memory leaks in DSA_do_sign and DSA_is_prime.
8713 Also really enable memory leak checks in openssl.c and in some
8714 test programs.
8715 [Chad C. Mulligan, Bodo Moeller]
8716
8717 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
8718 up the length of negative integers. This has now been simplified to just
8719 store the length when it is first determined and use it later, rather
8720 than trying to keep track of where data is copied and updating it to
8721 point to the end.
8722 [Steve Henson, reported by Brien Wheeler
8723 <bwheeler@authentica-security.com>]
8724
8725 *) Add a new function PKCS7_signatureVerify. This allows the verification
8726 of a PKCS#7 signature but with the signing certificate passed to the
8727 function itself. This contrasts with PKCS7_dataVerify which assumes the
8728 certificate is present in the PKCS#7 structure. This isn't always the
8729 case: certificates can be omitted from a PKCS#7 structure and be
8730 distributed by "out of band" means (such as a certificate database).
8731 [Steve Henson]
8732
8733 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the
8734 function prototypes in pem.h, also change util/mkdef.pl to add the
8735 necessary function names.
8736 [Steve Henson]
8737
8738 *) mk1mf.pl (used by Windows builds) did not properly read the
8739 options set by Configure in the top level Makefile, and Configure
8740 was not even able to write more than one option correctly.
8741 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
8742 [Bodo Moeller]
8743
8744 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config
8745 file to be loaded from a BIO or FILE pointer. The BIO version will
8746 for example allow memory BIOs to contain config info.
8747 [Steve Henson]
8748
8749 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
8750 Whoever hopes to achieve shared-library compatibility across versions
8751 must use this, not the compile-time macro.
8752 (Exercise 0.9.4: Which is the minimum library version required by
8753 such programs?)
8754 Note: All this applies only to multi-threaded programs, others don't
8755 need locks.
8756 [Bodo Moeller]
8757
8758 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests
8759 through a BIO pair triggered the default case, i.e.
8760 SSLerr(...,SSL_R_UNKNOWN_STATE).
8761 [Bodo Moeller]
8762
8763 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
8764 can use the SSL library even if none of the specific BIOs is
8765 appropriate.
8766 [Bodo Moeller]
8767
8768 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
8769 for the encoded length.
8770 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>]
8771
8772 *) Add initial documentation of the X509V3 functions.
8773 [Steve Henson]
8774
8775 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and
8776 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
8777 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
8778 secure PKCS#8 private key format with a high iteration count.
8779 [Steve Henson]
8780
8781 *) Fix determination of Perl interpreter: A perl or perl5
8782 _directory_ in $PATH was also accepted as the interpreter.
8783 [Ralf S. Engelschall]
8784
8785 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking
8786 wrong with it but it was very old and did things like calling
8787 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
8788 unusual formatting.
8789 [Steve Henson]
8790
8791 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed
8792 to use the new extension code.
8793 [Steve Henson]
8794
8795 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
8796 with macros. This should make it easier to change their form, add extra
8797 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
8798 constant.
8799 [Steve Henson]
8800
8801 *) Add to configuration table a new entry that can specify an alternative
8802 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
8803 according to Mark Crispin <MRC@Panda.COM>.
8804 [Bodo Moeller]
8805
8806 #if 0
8807 *) DES CBC did not update the IV. Weird.
8808 [Ben Laurie]
8809 #else
8810 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
8811 Changing the behaviour of the former might break existing programs --
8812 where IV updating is needed, des_ncbc_encrypt can be used.
8813 #endif
8814
8815 *) When bntest is run from "make test" it drives bc to check its
8816 calculations, as well as internally checking them. If an internal check
8817 fails, it needs to cause bc to give a non-zero result or make test carries
8818 on without noticing the failure. Fixed.
8819 [Ben Laurie]
8820
8821 *) DES library cleanups.
8822 [Ulf Möller]
8823
8824 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
8825 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
8826 ciphers. NOTE: although the key derivation function has been verified
8827 against some published test vectors it has not been extensively tested
8828 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
8829 of v2.0.
8830 [Steve Henson]
8831
8832 *) Instead of "mkdir -p", which is not fully portable, use new
8833 Perl script "util/mkdir-p.pl".
8834 [Bodo Moeller]
8835
8836 *) Rewrite the way password based encryption (PBE) is handled. It used to
8837 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
8838 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
8839 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
8840 the 'parameter' field of the AlgorithmIdentifier is passed to the
8841 underlying key generation function so it must do its own ASN1 parsing.
8842 This has also changed the EVP_PBE_CipherInit() function which now has a
8843 'parameter' argument instead of literal salt and iteration count values
8844 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
8845 [Steve Henson]
8846
8847 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms
8848 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
8849 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
8850 KEY" because this clashed with PKCS#8 unencrypted string. Since this
8851 value was just used as a "magic string" and not used directly its
8852 value doesn't matter.
8853 [Steve Henson]
8854
8855 *) Introduce some semblance of const correctness to BN. Shame C doesn't
8856 support mutable.
8857 [Ben Laurie]
8858
8859 *) "linux-sparc64" configuration (ultrapenguin).
8860 [Ray Miller <ray.miller@oucs.ox.ac.uk>]
8861 "linux-sparc" configuration.
8862 [Christian Forster <fo@hawo.stw.uni-erlangen.de>]
8863
8864 *) config now generates no-xxx options for missing ciphers.
8865 [Ulf Möller]
8866
8867 *) Support the EBCDIC character set (work in progress).
8868 File ebcdic.c not yet included because it has a different license.
8869 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
8870
8871 *) Support BS2000/OSD-POSIX.
8872 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
8873
8874 *) Make callbacks for key generation use void * instead of char *.
8875 [Ben Laurie]
8876
8877 *) Make S/MIME samples compile (not yet tested).
8878 [Ben Laurie]
8879
8880 *) Additional typesafe stacks.
8881 [Ben Laurie]
8882
8883 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
8884 [Bodo Moeller]
8885
8886
8887 Changes between 0.9.3 and 0.9.3a [29 May 1999]
8888
8889 *) New configuration variant "sco5-gcc".
8890
8891 *) Updated some demos.
8892 [Sean O Riordain, Wade Scholine]
8893
8894 *) Add missing BIO_free at exit of pkcs12 application.
8895 [Wu Zhigang]
8896
8897 *) Fix memory leak in conf.c.
8898 [Steve Henson]
8899
8900 *) Updates for Win32 to assembler version of MD5.
8901 [Steve Henson]
8902
8903 *) Set #! path to perl in apps/der_chop to where we found it
8904 instead of using a fixed path.
8905 [Bodo Moeller]
8906
8907 *) SHA library changes for irix64-mips4-cc.
8908 [Andy Polyakov]
8909
8910 *) Improvements for VMS support.
8911 [Richard Levitte]
8912
8913
8914 Changes between 0.9.2b and 0.9.3 [24 May 1999]
8915
8916 *) Bignum library bug fix. IRIX 6 passes "make test" now!
8917 This also avoids the problems with SC4.2 and unpatched SC5.
8918 [Andy Polyakov <appro@fy.chalmers.se>]
8919
8920 *) New functions sk_num, sk_value and sk_set to replace the previous macros.
8921 These are required because of the typesafe stack would otherwise break
8922 existing code. If old code used a structure member which used to be STACK
8923 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
8924 sk_num or sk_value it would produce an error because the num, data members
8925 are not present in STACK_OF. Now it just produces a warning. sk_set
8926 replaces the old method of assigning a value to sk_value
8927 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
8928 that does this will no longer work (and should use sk_set instead) but
8929 this could be regarded as a "questionable" behaviour anyway.
8930 [Steve Henson]
8931
8932 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now
8933 correctly handle encrypted S/MIME data.
8934 [Steve Henson]
8935
8936 *) Change type of various DES function arguments from des_cblock
8937 (which means, in function argument declarations, pointer to char)
8938 to des_cblock * (meaning pointer to array with 8 char elements),
8939 which allows the compiler to do more typechecking; it was like
8940 that back in SSLeay, but with lots of ugly casts.
8941
8942 Introduce new type const_des_cblock.
8943 [Bodo Moeller]
8944
8945 *) Reorganise the PKCS#7 library and get rid of some of the more obvious
8946 problems: find RecipientInfo structure that matches recipient certificate
8947 and initialise the ASN1 structures properly based on passed cipher.
8948 [Steve Henson]
8949
8950 *) Belatedly make the BN tests actually check the results.
8951 [Ben Laurie]
8952
8953 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
8954 to and from BNs: it was completely broken. New compilation option
8955 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
8956 key elements as negative integers.
8957 [Steve Henson]
8958
8959 *) Reorganize and speed up MD5.
8960 [Andy Polyakov <appro@fy.chalmers.se>]
8961
8962 *) VMS support.
8963 [Richard Levitte <richard@levitte.org>]
8964
8965 *) New option -out to asn1parse to allow the parsed structure to be
8966 output to a file. This is most useful when combined with the -strparse
8967 option to examine the output of things like OCTET STRINGS.
8968 [Steve Henson]
8969
8970 *) Make SSL library a little more fool-proof by not requiring any longer
8971 that SSL_set_{accept,connect}_state be called before
8972 SSL_{accept,connect} may be used (SSL_set_..._state is omitted
8973 in many applications because usually everything *appeared* to work as
8974 intended anyway -- now it really works as intended).
8975 [Bodo Moeller]
8976
8977 *) Move openssl.cnf out of lib/.
8978 [Ulf Möller]
8979
8980 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall
8981 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
8982 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+
8983 [Ralf S. Engelschall]
8984
8985 *) Various fixes to the EVP and PKCS#7 code. It may now be able to
8986 handle PKCS#7 enveloped data properly.
8987 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve]
8988
8989 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
8990 copying pointers. The cert_st handling is changed by this in
8991 various ways (and thus what used to be known as ctx->default_cert
8992 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert
8993 any longer when s->cert does not give us what we need).
8994 ssl_cert_instantiate becomes obsolete by this change.
8995 As soon as we've got the new code right (possibly it already is?),
8996 we have solved a couple of bugs of the earlier code where s->cert
8997 was used as if it could not have been shared with other SSL structures.
8998
8999 Note that using the SSL API in certain dirty ways now will result
9000 in different behaviour than observed with earlier library versions:
9001 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx)
9002 does not influence s as it used to.
9003
9004 In order to clean up things more thoroughly, inside SSL_SESSION
9005 we don't use CERT any longer, but a new structure SESS_CERT
9006 that holds per-session data (if available); currently, this is
9007 the peer's certificate chain and, for clients, the server's certificate
9008 and temporary key. CERT holds only those values that can have
9009 meaningful defaults in an SSL_CTX.
9010 [Bodo Moeller]
9011
9012 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
9013 from the internal representation. Various PKCS#7 fixes: remove some
9014 evil casts and set the enc_dig_alg field properly based on the signing
9015 key type.
9016 [Steve Henson]
9017
9018 *) Allow PKCS#12 password to be set from the command line or the
9019 environment. Let 'ca' get its config file name from the environment
9020 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
9021 and 'x509').
9022 [Steve Henson]
9023
9024 *) Allow certificate policies extension to use an IA5STRING for the
9025 organization field. This is contrary to the PKIX definition but
9026 VeriSign uses it and IE5 only recognises this form. Document 'x509'
9027 extension option.
9028 [Steve Henson]
9029
9030 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
9031 without disallowing inline assembler and the like for non-pedantic builds.
9032 [Ben Laurie]
9033
9034 *) Support Borland C++ builder.
9035 [Janez Jere <jj@void.si>, modified by Ulf Möller]
9036
9037 *) Support Mingw32.
9038 [Ulf Möller]
9039
9040 *) SHA-1 cleanups and performance enhancements.
9041 [Andy Polyakov <appro@fy.chalmers.se>]
9042
9043 *) Sparc v8plus assembler for the bignum library.
9044 [Andy Polyakov <appro@fy.chalmers.se>]
9045
9046 *) Accept any -xxx and +xxx compiler options in Configure.
9047 [Ulf Möller]
9048
9049 *) Update HPUX configuration.
9050 [Anonymous]
9051
9052 *) Add missing sk_<type>_unshift() function to safestack.h
9053 [Ralf S. Engelschall]
9054
9055 *) New function SSL_CTX_use_certificate_chain_file that sets the
9056 "extra_cert"s in addition to the certificate. (This makes sense
9057 only for "PEM" format files, as chains as a whole are not
9058 DER-encoded.)
9059 [Bodo Moeller]
9060
9061 *) Support verify_depth from the SSL API.
9062 x509_vfy.c had what can be considered an off-by-one-error:
9063 Its depth (which was not part of the external interface)
9064 was actually counting the number of certificates in a chain;
9065 now it really counts the depth.
9066 [Bodo Moeller]
9067
9068 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
9069 instead of X509err, which often resulted in confusing error
9070 messages since the error codes are not globally unique
9071 (e.g. an alleged error in ssl3_accept when a certificate
9072 didn't match the private key).
9073
9074 *) New function SSL_CTX_set_session_id_context that allows to set a default
9075 value (so that you don't need SSL_set_session_id_context for each
9076 connection using the SSL_CTX).
9077 [Bodo Moeller]
9078
9079 *) OAEP decoding bug fix.
9080 [Ulf Möller]
9081
9082 *) Support INSTALL_PREFIX for package builders, as proposed by
9083 David Harris.
9084 [Bodo Moeller]
9085
9086 *) New Configure options "threads" and "no-threads". For systems
9087 where the proper compiler options are known (currently Solaris
9088 and Linux), "threads" is the default.
9089 [Bodo Moeller]
9090
9091 *) New script util/mklink.pl as a faster substitute for util/mklink.sh.
9092 [Bodo Moeller]
9093
9094 *) Install various scripts to $(OPENSSLDIR)/misc, not to
9095 $(INSTALLTOP)/bin -- they shouldn't clutter directories
9096 such as /usr/local/bin.
9097 [Bodo Moeller]
9098
9099 *) "make linux-shared" to build shared libraries.
9100 [Niels Poppe <niels@netbox.org>]
9101
9102 *) New Configure option no-<cipher> (rsa, idea, rc5, ...).
9103 [Ulf Möller]
9104
9105 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
9106 extension adding in x509 utility.
9107 [Steve Henson]
9108
9109 *) Remove NOPROTO sections and error code comments.
9110 [Ulf Möller]
9111
9112 *) Partial rewrite of the DEF file generator to now parse the ANSI
9113 prototypes.
9114 [Steve Henson]
9115
9116 *) New Configure options --prefix=DIR and --openssldir=DIR.
9117 [Ulf Möller]
9118
9119 *) Complete rewrite of the error code script(s). It is all now handled
9120 by one script at the top level which handles error code gathering,
9121 header rewriting and C source file generation. It should be much better
9122 than the old method: it now uses a modified version of Ulf's parser to
9123 read the ANSI prototypes in all header files (thus the old K&R definitions
9124 aren't needed for error creation any more) and do a better job of
9125 translating function codes into names. The old 'ASN1 error code imbedded
9126 in a comment' is no longer necessary and it doesn't use .err files which
9127 have now been deleted. Also the error code call doesn't have to appear all
9128 on one line (which resulted in some large lines...).
9129 [Steve Henson]
9130
9131 *) Change #include filenames from <foo.h> to <openssl/foo.h>.
9132 [Bodo Moeller]
9133
9134 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return
9135 0 (which usually indicates a closed connection), but continue reading.
9136 [Bodo Moeller]
9137
9138 *) Fix some race conditions.
9139 [Bodo Moeller]
9140
9141 *) Add support for CRL distribution points extension. Add Certificate
9142 Policies and CRL distribution points documentation.
9143 [Steve Henson]
9144
9145 *) Move the autogenerated header file parts to crypto/opensslconf.h.
9146 [Ulf Möller]
9147
9148 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
9149 8 of keying material. Merlin has also confirmed interop with this fix
9150 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
9151 [Merlin Hughes <merlin@baltimore.ie>]
9152
9153 *) Fix lots of warnings.
9154 [Richard Levitte <levitte@stacken.kth.se>]
9155
9156 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
9157 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
9158 [Richard Levitte <levitte@stacken.kth.se>]
9159
9160 *) Fix problems with sizeof(long) == 8.
9161 [Andy Polyakov <appro@fy.chalmers.se>]
9162
9163 *) Change functions to ANSI C.
9164 [Ulf Möller]
9165
9166 *) Fix typos in error codes.
9167 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller]
9168
9169 *) Remove defunct assembler files from Configure.
9170 [Ulf Möller]
9171
9172 *) SPARC v8 assembler BIGNUM implementation.
9173 [Andy Polyakov <appro@fy.chalmers.se>]
9174
9175 *) Support for Certificate Policies extension: both print and set.
9176 Various additions to support the r2i method this uses.
9177 [Steve Henson]
9178
9179 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could
9180 return a const string when you are expecting an allocated buffer.
9181 [Ben Laurie]
9182
9183 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
9184 types DirectoryString and DisplayText.
9185 [Steve Henson]
9186
9187 *) Add code to allow r2i extensions to access the configuration database,
9188 add an LHASH database driver and add several ctx helper functions.
9189 [Steve Henson]
9190
9191 *) Fix an evil bug in bn_expand2() which caused various BN functions to
9192 fail when they extended the size of a BIGNUM.
9193 [Steve Henson]
9194
9195 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to
9196 support typesafe stack.
9197 [Steve Henson]
9198
9199 *) Fix typo in SSL_[gs]et_options().
9200 [Nils Frostberg <nils@medcom.se>]
9201
9202 *) Delete various functions and files that belonged to the (now obsolete)
9203 old X509V3 handling code.
9204 [Steve Henson]
9205
9206 *) New Configure option "rsaref".
9207 [Ulf Möller]
9208
9209 *) Don't auto-generate pem.h.
9210 [Bodo Moeller]
9211
9212 *) Introduce type-safe ASN.1 SETs.
9213 [Ben Laurie]
9214
9215 *) Convert various additional casted stacks to type-safe STACK_OF() variants.
9216 [Ben Laurie, Ralf S. Engelschall, Steve Henson]
9217
9218 *) Introduce type-safe STACKs. This will almost certainly break lots of code
9219 that links with OpenSSL (well at least cause lots of warnings), but fear
9220 not: the conversion is trivial, and it eliminates loads of evil casts. A
9221 few STACKed things have been converted already. Feel free to convert more.
9222 In the fullness of time, I'll do away with the STACK type altogether.
9223 [Ben Laurie]
9224
9225 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate
9226 specified in <certfile> by updating the entry in the index.txt file.
9227 This way one no longer has to edit the index.txt file manually for
9228 revoking a certificate. The -revoke option does the gory details now.
9229 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall]
9230
9231 *) Fix `openssl crl -noout -text' combination where `-noout' killed the
9232 `-text' option at all and this way the `-noout -text' combination was
9233 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'.
9234 [Ralf S. Engelschall]
9235
9236 *) Make sure a corresponding plain text error message exists for the
9237 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
9238 verify callback function determined that a certificate was revoked.
9239 [Ralf S. Engelschall]
9240
9241 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for
9242 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
9243 all available cipers including rc5, which was forgotten until now.
9244 In order to let the testing shell script know which algorithms
9245 are available, a new (up to now undocumented) command
9246 "openssl list-cipher-commands" is used.
9247 [Bodo Moeller]
9248
9249 *) Bugfix: s_client occasionally would sleep in select() when
9250 it should have checked SSL_pending() first.
9251 [Bodo Moeller]
9252
9253 *) New functions DSA_do_sign and DSA_do_verify to provide access to
9254 the raw DSA values prior to ASN.1 encoding.
9255 [Ulf Möller]
9256
9257 *) Tweaks to Configure
9258 [Niels Poppe <niels@netbox.org>]
9259
9260 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
9261 yet...
9262 [Steve Henson]
9263
9264 *) New variables $(RANLIB) and $(PERL) in the Makefiles.
9265 [Ulf Möller]
9266
9267 *) New config option to avoid instructions that are illegal on the 80386.
9268 The default code is faster, but requires at least a 486.
9269 [Ulf Möller]
9270
9271 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
9272 SSL2_SERVER_VERSION (not used at all) macros, which are now the
9273 same as SSL2_VERSION anyway.
9274 [Bodo Moeller]
9275
9276 *) New "-showcerts" option for s_client.
9277 [Bodo Moeller]
9278
9279 *) Still more PKCS#12 integration. Add pkcs12 application to openssl
9280 application. Various cleanups and fixes.
9281 [Steve Henson]
9282
9283 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
9284 modify error routines to work internally. Add error codes and PBE init
9285 to library startup routines.
9286 [Steve Henson]
9287
9288 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and
9289 packing functions to asn1 and evp. Changed function names and error
9290 codes along the way.
9291 [Steve Henson]
9292
9293 *) PKCS12 integration: and so it begins... First of several patches to
9294 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
9295 objects to objects.h
9296 [Steve Henson]
9297
9298 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1
9299 and display support for Thawte strong extranet extension.
9300 [Steve Henson]
9301
9302 *) Add LinuxPPC support.
9303 [Jeff Dubrule <igor@pobox.org>]
9304
9305 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
9306 bn_div_words in alpha.s.
9307 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie]
9308
9309 *) Make sure the RSA OAEP test is skipped under -DRSAref because
9310 OAEP isn't supported when OpenSSL is built with RSAref.
9311 [Ulf Moeller <ulf@fitug.de>]
9312
9313 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
9314 so they no longer are missing under -DNOPROTO.
9315 [Soren S. Jorvang <soren@t.dk>]
9316
9317
9318 Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
9319
9320 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
9321 doesn't work when the session is reused. Coming soon!
9322 [Ben Laurie]
9323
9324 *) Fix a security hole, that allows sessions to be reused in the wrong
9325 context thus bypassing client cert protection! All software that uses
9326 client certs and session caches in multiple contexts NEEDS PATCHING to
9327 allow session reuse! A fuller solution is in the works.
9328 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)]
9329
9330 *) Some more source tree cleanups (removed obsolete files
9331 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
9332 permission on "config" script to be executable) and a fix for the INSTALL
9333 document.
9334 [Ulf Moeller <ulf@fitug.de>]
9335
9336 *) Remove some legacy and erroneous uses of malloc, free instead of
9337 Malloc, Free.
9338 [Lennart Bang <lob@netstream.se>, with minor changes by Steve]
9339
9340 *) Make rsa_oaep_test return non-zero on error.
9341 [Ulf Moeller <ulf@fitug.de>]
9342
9343 *) Add support for native Solaris shared libraries. Configure
9344 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
9345 if someone would make that last step automatic.
9346 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>]
9347
9348 *) ctx_size was not built with the right compiler during "make links". Fixed.
9349 [Ben Laurie]
9350
9351 *) Change the meaning of 'ALL' in the cipher list. It now means "everything
9352 except NULL ciphers". This means the default cipher list will no longer
9353 enable NULL ciphers. They need to be specifically enabled e.g. with
9354 the string "DEFAULT:eNULL".
9355 [Steve Henson]
9356
9357 *) Fix to RSA private encryption routines: if p < q then it would
9358 occasionally produce an invalid result. This will only happen with
9359 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
9360 [Steve Henson]
9361
9362 *) Be less restrictive and allow also `perl util/perlpath.pl
9363 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin',
9364 because this way one can also use an interpreter named `perl5' (which is
9365 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
9366 installed as `perl').
9367 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
9368
9369 *) Let util/clean-depend.pl work also with older Perl 5.00x versions.
9370 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
9371
9372 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
9373 advapi32.lib to Win32 build and change the pem test comparision
9374 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
9375 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
9376 and crypto/des/ede_cbcm_enc.c.
9377 [Steve Henson]
9378
9379 *) DES quad checksum was broken on big-endian architectures. Fixed.
9380 [Ben Laurie]
9381
9382 *) Comment out two functions in bio.h that aren't implemented. Fix up the
9383 Win32 test batch file so it (might) work again. The Win32 test batch file
9384 is horrible: I feel ill....
9385 [Steve Henson]
9386
9387 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
9388 in e_os.h. Audit of header files to check ANSI and non ANSI
9389 sections: 10 functions were absent from non ANSI section and not exported
9390 from Windows DLLs. Fixed up libeay.num for new functions.
9391 [Steve Henson]
9392
9393 *) Make `openssl version' output lines consistent.
9394 [Ralf S. Engelschall]
9395
9396 *) Fix Win32 symbol export lists for BIO functions: Added
9397 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
9398 to ms/libeay{16,32}.def.
9399 [Ralf S. Engelschall]
9400
9401 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
9402 fine under Unix and passes some trivial tests I've now added. But the
9403 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
9404 added to make sure no one expects that this stuff really works in the
9405 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
9406 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
9407 openssl_bio.xs.
9408 [Ralf S. Engelschall]
9409
9410 *) Fix the generation of two part addresses in perl.
9411 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie]
9412
9413 *) Add config entry for Linux on MIPS.
9414 [John Tobey <jtobey@channel1.com>]
9415
9416 *) Make links whenever Configure is run, unless we are on Windoze.
9417 [Ben Laurie]
9418
9419 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf.
9420 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
9421 in CRLs.
9422 [Steve Henson]
9423
9424 *) Add a useful kludge to allow package maintainers to specify compiler and
9425 other platforms details on the command line without having to patch the
9426 Configure script everytime: One now can use ``perl Configure
9427 <id>:<details>'', i.e. platform ids are allowed to have details appended
9428 to them (seperated by colons). This is treated as there would be a static
9429 pre-configured entry in Configure's %table under key <id> with value
9430 <details> and ``perl Configure <id>'' is called. So, when you want to
9431 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
9432 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"''
9433 now, which overrides the FreeBSD-elf entry on-the-fly.
9434 [Ralf S. Engelschall]
9435
9436 *) Disable new TLS1 ciphersuites by default: they aren't official yet.
9437 [Ben Laurie]
9438
9439 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
9440 on the `perl Configure ...' command line. This way one can compile
9441 OpenSSL libraries with Position Independent Code (PIC) which is needed
9442 for linking it into DSOs.
9443 [Ralf S. Engelschall]
9444
9445 *) Remarkably, export ciphers were totally broken and no-one had noticed!
9446 Fixed.
9447 [Ben Laurie]
9448
9449 *) Cleaned up the LICENSE document: The official contact for any license
9450 questions now is the OpenSSL core team under openssl-core@openssl.org.
9451 And add a paragraph about the dual-license situation to make sure people
9452 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
9453 to the OpenSSL toolkit.
9454 [Ralf S. Engelschall]
9455
9456 *) General source tree makefile cleanups: Made `making xxx in yyy...'
9457 display consistent in the source tree and replaced `/bin/rm' by `rm'.
9458 Additonally cleaned up the `make links' target: Remove unnecessary
9459 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
9460 to speed processing and no longer clutter the display with confusing
9461 stuff. Instead only the actually done links are displayed.
9462 [Ralf S. Engelschall]
9463
9464 *) Permit null encryption ciphersuites, used for authentication only. It used
9465 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
9466 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
9467 encryption.
9468 [Ben Laurie]
9469
9470 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
9471 signed attributes when verifying signatures (this would break them),
9472 the detached data encoding was wrong and public keys obtained using
9473 X509_get_pubkey() weren't freed.
9474 [Steve Henson]
9475
9476 *) Add text documentation for the BUFFER functions. Also added a work around
9477 to a Win95 console bug. This was triggered by the password read stuff: the
9478 last character typed gets carried over to the next fread(). If you were
9479 generating a new cert request using 'req' for example then the last
9480 character of the passphrase would be CR which would then enter the first
9481 field as blank.
9482 [Steve Henson]
9483
9484 *) Added the new `Includes OpenSSL Cryptography Software' button as
9485 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
9486 button and can be used by applications based on OpenSSL to show the
9487 relationship to the OpenSSL project.
9488 [Ralf S. Engelschall]
9489
9490 *) Remove confusing variables in function signatures in files
9491 ssl/ssl_lib.c and ssl/ssl.h.
9492 [Lennart Bong <lob@kulthea.stacken.kth.se>]
9493
9494 *) Don't install bss_file.c under PREFIX/include/
9495 [Lennart Bong <lob@kulthea.stacken.kth.se>]
9496
9497 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle
9498 functions that return function pointers and has support for NT specific
9499 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
9500 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
9501 unsigned to signed types: this was killing the Win32 compile.
9502 [Steve Henson]
9503
9504 *) Add new certificate file to stack functions,
9505 SSL_add_dir_cert_subjects_to_stack() and
9506 SSL_add_file_cert_subjects_to_stack(). These largely supplant
9507 SSL_load_client_CA_file(), and can be used to add multiple certs easily
9508 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
9509 This means that Apache-SSL and similar packages don't have to mess around
9510 to add as many CAs as they want to the preferred list.
9511 [Ben Laurie]
9512
9513 *) Experiment with doxygen documentation. Currently only partially applied to
9514 ssl/ssl_lib.c.
9515 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with
9516 openssl.doxy as the configuration file.
9517 [Ben Laurie]
9518
9519 *) Get rid of remaining C++-style comments which strict C compilers hate.
9520 [Ralf S. Engelschall, pointed out by Carlos Amengual]
9521
9522 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
9523 compiled in by default: it has problems with large keys.
9524 [Steve Henson]
9525
9526 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
9527 DH private keys and/or callback functions which directly correspond to
9528 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
9529 is needed for applications which have to configure certificates on a
9530 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
9531 (e.g. s_server).
9532 For the RSA certificate situation is makes no difference, but
9533 for the DSA certificate situation this fixes the "no shared cipher"
9534 problem where the OpenSSL cipher selection procedure failed because the
9535 temporary keys were not overtaken from the context and the API provided
9536 no way to reconfigure them.
9537 The new functions now let applications reconfigure the stuff and they
9538 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
9539 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
9540 non-public-API function ssl_cert_instantiate() is used as a helper
9541 function and also to reduce code redundancy inside ssl_rsa.c.
9542 [Ralf S. Engelschall]
9543
9544 *) Move s_server -dcert and -dkey options out of the undocumented feature
9545 area because they are useful for the DSA situation and should be
9546 recognized by the users.
9547 [Ralf S. Engelschall]
9548
9549 *) Fix the cipher decision scheme for export ciphers: the export bits are
9550 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
9551 SSL_EXP_MASK. So, the original variable has to be used instead of the
9552 already masked variable.
9553 [Richard Levitte <levitte@stacken.kth.se>]
9554
9555 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c
9556 [Richard Levitte <levitte@stacken.kth.se>]
9557
9558 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
9559 from `int' to `unsigned int' because it's a length and initialized by
9560 EVP_DigestFinal() which expects an `unsigned int *'.
9561 [Richard Levitte <levitte@stacken.kth.se>]
9562
9563 *) Don't hard-code path to Perl interpreter on shebang line of Configure
9564 script. Instead use the usual Shell->Perl transition trick.
9565 [Ralf S. Engelschall]
9566
9567 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates
9568 (in addition to RSA certificates) to match the behaviour of `openssl dsa
9569 -noout -modulus' as it's already the case for `openssl rsa -noout
9570 -modulus'. For RSA the -modulus is the real "modulus" while for DSA
9571 currently the public key is printed (a decision which was already done by
9572 `openssl dsa -modulus' in the past) which serves a similar purpose.
9573 Additionally the NO_RSA no longer completely removes the whole -modulus
9574 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
9575 now, too.
9576 [Ralf S. Engelschall]
9577
9578 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
9579 BIO. See the source (crypto/evp/bio_ok.c) for more info.
9580 [Arne Ansper <arne@ats.cyber.ee>]
9581
9582 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs
9583 to be added. Now both 'req' and 'ca' can use new objects defined in the
9584 config file.
9585 [Steve Henson]
9586
9587 *) Add cool BIO that does syslog (or event log on NT).
9588 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie]
9589
9590 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
9591 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
9592 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
9593 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
9594 [Ben Laurie]
9595
9596 *) Add preliminary config info for new extension code.
9597 [Steve Henson]
9598
9599 *) Make RSA_NO_PADDING really use no padding.
9600 [Ulf Moeller <ulf@fitug.de>]
9601
9602 *) Generate errors when private/public key check is done.
9603 [Ben Laurie]
9604
9605 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
9606 for some CRL extensions and new objects added.
9607 [Steve Henson]
9608
9609 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private
9610 key usage extension and fuller support for authority key id.
9611 [Steve Henson]
9612
9613 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
9614 padding method for RSA, which is recommended for new applications in PKCS
9615 #1 v2.0 (RFC 2437, October 1998).
9616 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
9617 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
9618 against Bleichbacher's attack on RSA.
9619 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
9620 Ben Laurie]
9621
9622 *) Updates to the new SSL compression code
9623 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
9624
9625 *) Fix so that the version number in the master secret, when passed
9626 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
9627 (because the server will not accept higher), that the version number
9628 is 0x03,0x01, not 0x03,0x00
9629 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
9630
9631 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
9632 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
9633 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
9634 [Steve Henson]
9635
9636 *) Support for RAW extensions where an arbitrary extension can be
9637 created by including its DER encoding. See apps/openssl.cnf for
9638 an example.
9639 [Steve Henson]
9640
9641 *) Make sure latest Perl versions don't interpret some generated C array
9642 code as Perl array code in the crypto/err/err_genc.pl script.
9643 [Lars Weber <3weber@informatik.uni-hamburg.de>]
9644
9645 *) Modify ms/do_ms.bat to not generate assembly language makefiles since
9646 not many people have the assembler. Various Win32 compilation fixes and
9647 update to the INSTALL.W32 file with (hopefully) more accurate Win32
9648 build instructions.
9649 [Steve Henson]
9650
9651 *) Modify configure script 'Configure' to automatically create crypto/date.h
9652 file under Win32 and also build pem.h from pem.org. New script
9653 util/mkfiles.pl to create the MINFO file on environments that can't do a
9654 'make files': perl util/mkfiles.pl >MINFO should work.
9655 [Steve Henson]
9656
9657 *) Major rework of DES function declarations, in the pursuit of correctness
9658 and purity. As a result, many evil casts evaporated, and some weirdness,
9659 too. You may find this causes warnings in your code. Zapping your evil
9660 casts will probably fix them. Mostly.
9661 [Ben Laurie]
9662
9663 *) Fix for a typo in asn1.h. Bug fix to object creation script
9664 obj_dat.pl. It considered a zero in an object definition to mean
9665 "end of object": none of the objects in objects.h have any zeros
9666 so it wasn't spotted.
9667 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>]
9668
9669 *) Add support for Triple DES Cipher Block Chaining with Output Feedback
9670 Masking (CBCM). In the absence of test vectors, the best I have been able
9671 to do is check that the decrypt undoes the encrypt, so far. Send me test
9672 vectors if you have them.
9673 [Ben Laurie]
9674
9675 *) Correct calculation of key length for export ciphers (too much space was
9676 allocated for null ciphers). This has not been tested!
9677 [Ben Laurie]
9678
9679 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage
9680 message is now correct (it understands "crypto" and "ssl" on its
9681 command line). There is also now an "update" option. This will update
9682 the util/ssleay.num and util/libeay.num files with any new functions.
9683 If you do a:
9684 perl util/mkdef.pl crypto ssl update
9685 it will update them.
9686 [Steve Henson]
9687
9688 *) Overhauled the Perl interface (perl/*):
9689 - ported BN stuff to OpenSSL's different BN library
9690 - made the perl/ source tree CVS-aware
9691 - renamed the package from SSLeay to OpenSSL (the files still contain
9692 their history because I've copied them in the repository)
9693 - removed obsolete files (the test scripts will be replaced
9694 by better Test::Harness variants in the future)
9695 [Ralf S. Engelschall]
9696
9697 *) First cut for a very conservative source tree cleanup:
9698 1. merge various obsolete readme texts into doc/ssleay.txt
9699 where we collect the old documents and readme texts.
9700 2. remove the first part of files where I'm already sure that we no
9701 longer need them because of three reasons: either they are just temporary
9702 files which were left by Eric or they are preserved original files where
9703 I've verified that the diff is also available in the CVS via "cvs diff
9704 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
9705 the crypto/md/ stuff).
9706 [Ralf S. Engelschall]
9707
9708 *) More extension code. Incomplete support for subject and issuer alt
9709 name, issuer and authority key id. Change the i2v function parameters
9710 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
9711 what that's for :-) Fix to ASN1 macro which messed up
9712 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
9713 [Steve Henson]
9714
9715 *) Preliminary support for ENUMERATED type. This is largely copied from the
9716 INTEGER code.
9717 [Steve Henson]
9718
9719 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
9720 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
9721
9722 *) Make sure `make rehash' target really finds the `openssl' program.
9723 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
9724
9725 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
9726 like to hear about it if this slows down other processors.
9727 [Ben Laurie]
9728
9729 *) Add CygWin32 platform information to Configure script.
9730 [Alan Batie <batie@aahz.jf.intel.com>]
9731
9732 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm'
9733 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>]
9734
9735 *) New program nseq to manipulate netscape certificate sequences
9736 [Steve Henson]
9737
9738 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
9739 few typos.
9740 [Steve Henson]
9741
9742 *) Fixes to BN code. Previously the default was to define BN_RECURSION
9743 but the BN code had some problems that would cause failures when
9744 doing certificate verification and some other functions.
9745 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
9746
9747 *) Add ASN1 and PEM code to support netscape certificate sequences.
9748 [Steve Henson]
9749
9750 *) Add ASN1 and PEM code to support netscape certificate sequences.
9751 [Steve Henson]
9752
9753 *) Add several PKIX and private extended key usage OIDs.
9754 [Steve Henson]
9755
9756 *) Modify the 'ca' program to handle the new extension code. Modify
9757 openssl.cnf for new extension format, add comments.
9758 [Steve Henson]
9759
9760 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
9761 and add a sample to openssl.cnf so req -x509 now adds appropriate
9762 CA extensions.
9763 [Steve Henson]
9764
9765 *) Continued X509 V3 changes. Add to other makefiles, integrate with the
9766 error code, add initial support to X509_print() and x509 application.
9767 [Steve Henson]
9768
9769 *) Takes a deep breath and start addding X509 V3 extension support code. Add
9770 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
9771 stuff is currently isolated and isn't even compiled yet.
9772 [Steve Henson]
9773
9774 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL
9775 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
9776 Removed the versions check from X509 routines when loading extensions:
9777 this allows certain broken certificates that don't set the version
9778 properly to be processed.
9779 [Steve Henson]
9780
9781 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
9782 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
9783 can still be regenerated with "make depend".
9784 [Ben Laurie]
9785
9786 *) Spelling mistake in C version of CAST-128.
9787 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>]
9788
9789 *) Changes to the error generation code. The perl script err-code.pl
9790 now reads in the old error codes and retains the old numbers, only
9791 adding new ones if necessary. It also only changes the .err files if new
9792 codes are added. The makefiles have been modified to only insert errors
9793 when needed (to avoid needlessly modifying header files). This is done
9794 by only inserting errors if the .err file is newer than the auto generated
9795 C file. To rebuild all the error codes from scratch (the old behaviour)
9796 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
9797 or delete all the .err files.
9798 [Steve Henson]
9799
9800 *) CAST-128 was incorrectly implemented for short keys. The C version has
9801 been fixed, but is untested. The assembler versions are also fixed, but
9802 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
9803 to regenerate it if needed.
9804 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
9805 Hagino <itojun@kame.net>]
9806
9807 *) File was opened incorrectly in randfile.c.
9808 [Ulf Möller <ulf@fitug.de>]
9809
9810 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print
9811 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
9812 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
9813 al: it's just almost always a UTCTime. Note this patch adds new error
9814 codes so do a "make errors" if there are problems.
9815 [Steve Henson]
9816
9817 *) Correct Linux 1 recognition in config.
9818 [Ulf Möller <ulf@fitug.de>]
9819
9820 *) Remove pointless MD5 hash when using DSA keys in ca.
9821 [Anonymous <nobody@replay.com>]
9822
9823 *) Generate an error if given an empty string as a cert directory. Also
9824 generate an error if handed NULL (previously returned 0 to indicate an
9825 error, but didn't set one).
9826 [Ben Laurie, reported by Anonymous <nobody@replay.com>]
9827
9828 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
9829 [Ben Laurie]
9830
9831 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
9832 parameters. This was causing a warning which killed off the Win32 compile.
9833 [Steve Henson]
9834
9835 *) Remove C++ style comments from crypto/bn/bn_local.h.
9836 [Neil Costigan <neil.costigan@celocom.com>]
9837
9838 *) The function OBJ_txt2nid was broken. It was supposed to return a nid
9839 based on a text string, looking up short and long names and finally
9840 "dot" format. The "dot" format stuff didn't work. Added new function
9841 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
9842 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
9843 OID is not part of the table.
9844 [Steve Henson]
9845
9846 *) Add prototypes to X509 lookup/verify methods, fixing a bug in
9847 X509_LOOKUP_by_alias().
9848 [Ben Laurie]
9849
9850 *) Sort openssl functions by name.
9851 [Ben Laurie]
9852
9853 *) Get the gendsa program working (hopefully) and add it to app list. Remove
9854 encryption from sample DSA keys (in case anyone is interested the password
9855 was "1234").
9856 [Steve Henson]
9857
9858 *) Make _all_ *_free functions accept a NULL pointer.
9859 [Frans Heymans <fheymans@isaserver.be>]
9860
9861 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use
9862 NULL pointers.
9863 [Anonymous <nobody@replay.com>]
9864
9865 *) s_server should send the CAfile as acceptable CAs, not its own cert.
9866 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
9867
9868 *) Don't blow it for numeric -newkey arguments to apps/req.
9869 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
9870
9871 *) Temp key "for export" tests were wrong in s3_srvr.c.
9872 [Anonymous <nobody@replay.com>]
9873
9874 *) Add prototype for temp key callback functions
9875 SSL_CTX_set_tmp_{rsa,dh}_callback().
9876 [Ben Laurie]
9877
9878 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
9879 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
9880 [Steve Henson]
9881
9882 *) X509_name_add_entry() freed the wrong thing after an error.
9883 [Arne Ansper <arne@ats.cyber.ee>]
9884
9885 *) rsa_eay.c would attempt to free a NULL context.
9886 [Arne Ansper <arne@ats.cyber.ee>]
9887
9888 *) BIO_s_socket() had a broken should_retry() on Windoze.
9889 [Arne Ansper <arne@ats.cyber.ee>]
9890
9891 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
9892 [Arne Ansper <arne@ats.cyber.ee>]
9893
9894 *) Make sure the already existing X509_STORE->depth variable is initialized
9895 in X509_STORE_new(), but document the fact that this variable is still
9896 unused in the certificate verification process.
9897 [Ralf S. Engelschall]
9898
9899 *) Fix the various library and apps files to free up pkeys obtained from
9900 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
9901 [Steve Henson]
9902
9903 *) Fix reference counting in X509_PUBKEY_get(). This makes
9904 demos/maurice/example2.c work, amongst others, probably.
9905 [Steve Henson and Ben Laurie]
9906
9907 *) First cut of a cleanup for apps/. First the `ssleay' program is now named
9908 `openssl' and second, the shortcut symlinks for the `openssl <command>'
9909 are no longer created. This way we have a single and consistent command
9910 line interface `openssl <command>', similar to `cvs <command>'.
9911 [Ralf S. Engelschall, Paul Sutton and Ben Laurie]
9912
9913 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
9914 BIT STRING wrapper always have zero unused bits.
9915 [Steve Henson]
9916
9917 *) Add CA.pl, perl version of CA.sh, add extended key usage OID.
9918 [Steve Henson]
9919
9920 *) Make the top-level INSTALL documentation easier to understand.
9921 [Paul Sutton]
9922
9923 *) Makefiles updated to exit if an error occurs in a sub-directory
9924 make (including if user presses ^C) [Paul Sutton]
9925
9926 *) Make Montgomery context stuff explicit in RSA data structure.
9927 [Ben Laurie]
9928
9929 *) Fix build order of pem and err to allow for generated pem.h.
9930 [Ben Laurie]
9931
9932 *) Fix renumbering bug in X509_NAME_delete_entry().
9933 [Ben Laurie]
9934
9935 *) Enhanced the err-ins.pl script so it makes the error library number
9936 global and can add a library name. This is needed for external ASN1 and
9937 other error libraries.
9938 [Steve Henson]
9939
9940 *) Fixed sk_insert which never worked properly.
9941 [Steve Henson]
9942
9943 *) Fix ASN1 macros so they can handle indefinite length construted
9944 EXPLICIT tags. Some non standard certificates use these: they can now
9945 be read in.
9946 [Steve Henson]
9947
9948 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
9949 into a single doc/ssleay.txt bundle. This way the information is still
9950 preserved but no longer messes up this directory. Now it's new room for
9951 the new set of documenation files.
9952 [Ralf S. Engelschall]
9953
9954 *) SETs were incorrectly DER encoded. This was a major pain, because they
9955 shared code with SEQUENCEs, which aren't coded the same. This means that
9956 almost everything to do with SETs or SEQUENCEs has either changed name or
9957 number of arguments.
9958 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>]
9959
9960 *) Fix test data to work with the above.
9961 [Ben Laurie]
9962
9963 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
9964 was already fixed by Eric for 0.9.1 it seems.
9965 [Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>]
9966
9967 *) Autodetect FreeBSD3.
9968 [Ben Laurie]
9969
9970 *) Fix various bugs in Configure. This affects the following platforms:
9971 nextstep
9972 ncr-scde
9973 unixware-2.0
9974 unixware-2.0-pentium
9975 sco5-cc.
9976 [Ben Laurie]
9977
9978 *) Eliminate generated files from CVS. Reorder tests to regenerate files
9979 before they are needed.
9980 [Ben Laurie]
9981
9982 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy).
9983 [Ben Laurie]
9984
9985
9986 Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
9987
9988 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
9989 changed SSLeay to OpenSSL in version strings.
9990 [Ralf S. Engelschall]
9991
9992 *) Some fixups to the top-level documents.
9993 [Paul Sutton]
9994
9995 *) Fixed the nasty bug where rsaref.h was not found under compile-time
9996 because the symlink to include/ was missing.
9997 [Ralf S. Engelschall]
9998
9999 *) Incorporated the popular no-RSA/DSA-only patches
10000 which allow to compile a RSA-free SSLeay.
10001 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall]
10002
10003 *) Fixed nasty rehash problem under `make -f Makefile.ssl links'
10004 when "ssleay" is still not found.
10005 [Ralf S. Engelschall]
10006
10007 *) Added more platforms to Configure: Cray T3E, HPUX 11,
10008 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>]
10009
10010 *) Updated the README file.
10011 [Ralf S. Engelschall]
10012
10013 *) Added various .cvsignore files in the CVS repository subdirs
10014 to make a "cvs update" really silent.
10015 [Ralf S. Engelschall]
10016
10017 *) Recompiled the error-definition header files and added
10018 missing symbols to the Win32 linker tables.
10019 [Ralf S. Engelschall]
10020
10021 *) Cleaned up the top-level documents;
10022 o new files: CHANGES and LICENSE
10023 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
10024 o merged COPYRIGHT into LICENSE
10025 o removed obsolete TODO file
10026 o renamed MICROSOFT to INSTALL.W32
10027 [Ralf S. Engelschall]
10028
10029 *) Removed dummy files from the 0.9.1b source tree:
10030 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
10031 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
10032 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
10033 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
10034 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
10035 [Ralf S. Engelschall]
10036
10037 *) Added various platform portability fixes.
10038 [Mark J. Cox]
10039
10040 *) The Genesis of the OpenSSL rpject:
10041 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
10042 Young and Tim J. Hudson created while they were working for C2Net until
10043 summer 1998.
10044 [The OpenSSL Project]
10045
10046
10047 Changes between 0.9.0b and 0.9.1b [not released]
10048
10049 *) Updated a few CA certificates under certs/
10050 [Eric A. Young]
10051
10052 *) Changed some BIGNUM api stuff.
10053 [Eric A. Young]
10054
10055 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
10056 DGUX x86, Linux Alpha, etc.
10057 [Eric A. Young]
10058
10059 *) New COMP library [crypto/comp/] for SSL Record Layer Compression:
10060 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
10061 available).
10062 [Eric A. Young]
10063
10064 *) Add -strparse option to asn1pars program which parses nested
10065 binary structures
10066 [Dr Stephen Henson <shenson@bigfoot.com>]
10067
10068 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
10069 [Eric A. Young]
10070
10071 *) DSA fix for "ca" program.
10072 [Eric A. Young]
10073
10074 *) Added "-genkey" option to "dsaparam" program.
10075 [Eric A. Young]
10076
10077 *) Added RIPE MD160 (rmd160) message digest.
10078 [Eric A. Young]
10079
10080 *) Added -a (all) option to "ssleay version" command.
10081 [Eric A. Young]
10082
10083 *) Added PLATFORM define which is the id given to Configure.
10084 [Eric A. Young]
10085
10086 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
10087 [Eric A. Young]
10088
10089 *) Extended the ASN.1 parser routines.
10090 [Eric A. Young]
10091
10092 *) Extended BIO routines to support REUSEADDR, seek, tell, etc.
10093 [Eric A. Young]
10094
10095 *) Added a BN_CTX to the BN library.
10096 [Eric A. Young]
10097
10098 *) Fixed the weak key values in DES library
10099 [Eric A. Young]
10100
10101 *) Changed API in EVP library for cipher aliases.
10102 [Eric A. Young]
10103
10104 *) Added support for RC2/64bit cipher.
10105 [Eric A. Young]
10106
10107 *) Converted the lhash library to the crypto/mem.c functions.
10108 [Eric A. Young]
10109
10110 *) Added more recognized ASN.1 object ids.
10111 [Eric A. Young]
10112
10113 *) Added more RSA padding checks for SSL/TLS.
10114 [Eric A. Young]
10115
10116 *) Added BIO proxy/filter functionality.
10117 [Eric A. Young]
10118
10119 *) Added extra_certs to SSL_CTX which can be used
10120 send extra CA certificates to the client in the CA cert chain sending
10121 process. It can be configured with SSL_CTX_add_extra_chain_cert().
10122 [Eric A. Young]
10123
10124 *) Now Fortezza is denied in the authentication phase because
10125 this is key exchange mechanism is not supported by SSLeay at all.
10126 [Eric A. Young]
10127
10128 *) Additional PKCS1 checks.
10129 [Eric A. Young]
10130
10131 *) Support the string "TLSv1" for all TLS v1 ciphers.
10132 [Eric A. Young]
10133
10134 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
10135 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
10136 [Eric A. Young]
10137
10138 *) Fixed a few memory leaks.
10139 [Eric A. Young]
10140
10141 *) Fixed various code and comment typos.
10142 [Eric A. Young]
10143
10144 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0
10145 bytes sent in the client random.
10146 [Edward Bishop <ebishop@spyglass.com>]
10147