CHANGES revision 160814
1 2 OpenSSL CHANGES 3 _______________ 4 5 Changes between 0.9.8a and 0.9.8b [04 May 2006] 6 7 *) When applying a cipher rule check to see if string match is an explicit 8 cipher suite and only match that one cipher suite if it is. 9 [Steve Henson] 10 11 *) Link in manifests for VC++ if needed. 12 [Austin Ziegler <halostatue@gmail.com>] 13 14 *) Update support for ECC-based TLS ciphersuites according to 15 draft-ietf-tls-ecc-12.txt with proposed changes (but without 16 TLS extensions, which are supported starting with the 0.9.9 17 branch, not in the OpenSSL 0.9.8 branch). 18 [Douglas Stebila] 19 20 *) New functions EVP_CIPHER_CTX_new() and EVP_CIPHER_CTX_free() to support 21 opaque EVP_CIPHER_CTX handling. 22 [Steve Henson] 23 24 *) Fixes and enhancements to zlib compression code. We now only use 25 "zlib1.dll" and use the default __cdecl calling convention on Win32 26 to conform with the standards mentioned here: 27 http://www.zlib.net/DLL_FAQ.txt 28 Static zlib linking now works on Windows and the new --with-zlib-include 29 --with-zlib-lib options to Configure can be used to supply the location 30 of the headers and library. Gracefully handle case where zlib library 31 can't be loaded. 32 [Steve Henson] 33 34 *) Several fixes and enhancements to the OID generation code. The old code 35 sometimes allowed invalid OIDs (1.X for X >= 40 for example), couldn't 36 handle numbers larger than ULONG_MAX, truncated printing and had a 37 non standard OBJ_obj2txt() behaviour. 38 [Steve Henson] 39 40 *) Add support for building of engines under engine/ as shared libraries 41 under VC++ build system. 42 [Steve Henson] 43 44 *) Corrected the numerous bugs in the Win32 path splitter in DSO. 45 Hopefully, we will not see any false combination of paths any more. 46 [Richard Levitte] 47 48 Changes between 0.9.8 and 0.9.8a [11 Oct 2005] 49 50 *) Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING 51 (part of SSL_OP_ALL). This option used to disable the 52 countermeasure against man-in-the-middle protocol-version 53 rollback in the SSL 2.0 server implementation, which is a bad 54 idea. (CVE-2005-2969) 55 56 [Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center 57 for Information Security, National Institute of Advanced Industrial 58 Science and Technology [AIST], Japan)] 59 60 *) Add two function to clear and return the verify parameter flags. 61 [Steve Henson] 62 63 *) Keep cipherlists sorted in the source instead of sorting them at 64 runtime, thus removing the need for a lock. 65 [Nils Larsch] 66 67 *) Avoid some small subgroup attacks in Diffie-Hellman. 68 [Nick Mathewson and Ben Laurie] 69 70 *) Add functions for well-known primes. 71 [Nick Mathewson] 72 73 *) Extended Windows CE support. 74 [Satoshi Nakamura and Andy Polyakov] 75 76 *) Initialize SSL_METHOD structures at compile time instead of during 77 runtime, thus removing the need for a lock. 78 [Steve Henson] 79 80 *) Make PKCS7_decrypt() work even if no certificate is supplied by 81 attempting to decrypt each encrypted key in turn. Add support to 82 smime utility. 83 [Steve Henson] 84 85 Changes between 0.9.7h and 0.9.8 [05 Jul 2005] 86 87 *) Add libcrypto.pc and libssl.pc for those who feel they need them. 88 [Richard Levitte] 89 90 *) Change CA.sh and CA.pl so they don't bundle the CSR and the private 91 key into the same file any more. 92 [Richard Levitte] 93 94 *) Add initial support for Win64, both IA64 and AMD64/x64 flavors. 95 [Andy Polyakov] 96 97 *) Add -utf8 command line and config file option to 'ca'. 98 [Stefan <stf@udoma.org] 99 100 *) Removed the macro des_crypt(), as it seems to conflict with some 101 libraries. Use DES_crypt(). 102 [Richard Levitte] 103 104 *) Correct naming of the 'chil' and '4758cca' ENGINEs. This 105 involves renaming the source and generated shared-libs for 106 both. The engines will accept the corrected or legacy ids 107 ('ncipher' and '4758_cca' respectively) when binding. NB, 108 this only applies when building 'shared'. 109 [Corinna Vinschen <vinschen@redhat.com> and Geoff Thorpe] 110 111 *) Add attribute functions to EVP_PKEY structure. Modify 112 PKCS12_create() to recognize a CSP name attribute and 113 use it. Make -CSP option work again in pkcs12 utility. 114 [Steve Henson] 115 116 *) Add new functionality to the bn blinding code: 117 - automatic re-creation of the BN_BLINDING parameters after 118 a fixed number of uses (currently 32) 119 - add new function for parameter creation 120 - introduce flags to control the update behaviour of the 121 BN_BLINDING parameters 122 - hide BN_BLINDING structure 123 Add a second BN_BLINDING slot to the RSA structure to improve 124 performance when a single RSA object is shared among several 125 threads. 126 [Nils Larsch] 127 128 *) Add support for DTLS. 129 [Nagendra Modadugu <nagendra@cs.stanford.edu> and Ben Laurie] 130 131 *) Add support for DER encoded private keys (SSL_FILETYPE_ASN1) 132 to SSL_CTX_use_PrivateKey_file() and SSL_use_PrivateKey_file() 133 [Walter Goulet] 134 135 *) Remove buggy and incompletet DH cert support from 136 ssl/ssl_rsa.c and ssl/s3_both.c 137 [Nils Larsch] 138 139 *) Use SHA-1 instead of MD5 as the default digest algorithm for 140 the apps/openssl applications. 141 [Nils Larsch] 142 143 *) Compile clean with "-Wall -Wmissing-prototypes 144 -Wstrict-prototypes -Wmissing-declarations -Werror". Currently 145 DEBUG_SAFESTACK must also be set. 146 [Ben Laurie] 147 148 *) Change ./Configure so that certain algorithms can be disabled by default. 149 The new counterpiece to "no-xxx" is "enable-xxx". 150 151 The patented RC5 and MDC2 algorithms will now be disabled unless 152 "enable-rc5" and "enable-mdc2", respectively, are specified. 153 154 (IDEA remains enabled despite being patented. This is because IDEA 155 is frequently required for interoperability, and there is no license 156 fee for non-commercial use. As before, "no-idea" can be used to 157 avoid this algorithm.) 158 159 [Bodo Moeller] 160 161 *) Add processing of proxy certificates (see RFC 3820). This work was 162 sponsored by KTH (The Royal Institute of Technology in Stockholm) and 163 EGEE (Enabling Grids for E-science in Europe). 164 [Richard Levitte] 165 166 *) RC4 performance overhaul on modern architectures/implementations, such 167 as Intel P4, IA-64 and AMD64. 168 [Andy Polyakov] 169 170 *) New utility extract-section.pl. This can be used specify an alternative 171 section number in a pod file instead of having to treat each file as 172 a separate case in Makefile. This can be done by adding two lines to the 173 pod file: 174 175 =for comment openssl_section:XXX 176 177 The blank line is mandatory. 178 179 [Steve Henson] 180 181 *) New arguments -certform, -keyform and -pass for s_client and s_server 182 to allow alternative format key and certificate files and passphrase 183 sources. 184 [Steve Henson] 185 186 *) New structure X509_VERIFY_PARAM which combines current verify parameters, 187 update associated structures and add various utility functions. 188 189 Add new policy related verify parameters, include policy checking in 190 standard verify code. Enhance 'smime' application with extra parameters 191 to support policy checking and print out. 192 [Steve Henson] 193 194 *) Add a new engine to support VIA PadLock ACE extensions in the VIA C3 195 Nehemiah processors. These extensions support AES encryption in hardware 196 as well as RNG (though RNG support is currently disabled). 197 [Michal Ludvig <michal@logix.cz>, with help from Andy Polyakov] 198 199 *) Deprecate BN_[get|set]_params() functions (they were ignored internally). 200 [Geoff Thorpe] 201 202 *) New FIPS 180-2 algorithms, SHA-224/-256/-384/-512 are implemented. 203 [Andy Polyakov and a number of other people] 204 205 *) Improved PowerPC platform support. Most notably BIGNUM assembler 206 implementation contributed by IBM. 207 [Suresh Chari, Peter Waltenberg, Andy Polyakov] 208 209 *) The new 'RSA_generate_key_ex' function now takes a BIGNUM for the public 210 exponent rather than 'unsigned long'. There is a corresponding change to 211 the new 'rsa_keygen' element of the RSA_METHOD structure. 212 [Jelte Jansen, Geoff Thorpe] 213 214 *) Functionality for creating the initial serial number file is now 215 moved from CA.pl to the 'ca' utility with a new option -create_serial. 216 217 (Before OpenSSL 0.9.7e, CA.pl used to initialize the serial 218 number file to 1, which is bound to cause problems. To avoid 219 the problems while respecting compatibility between different 0.9.7 220 patchlevels, 0.9.7e employed 'openssl x509 -next_serial' in 221 CA.pl for serial number initialization. With the new release 0.9.8, 222 we can fix the problem directly in the 'ca' utility.) 223 [Steve Henson] 224 225 *) Reduced header interdepencies by declaring more opaque objects in 226 ossl_typ.h. As a consequence, including some headers (eg. engine.h) will 227 give fewer recursive includes, which could break lazy source code - so 228 this change is covered by the OPENSSL_NO_DEPRECATED symbol. As always, 229 developers should define this symbol when building and using openssl to 230 ensure they track the recommended behaviour, interfaces, [etc], but 231 backwards-compatible behaviour prevails when this isn't defined. 232 [Geoff Thorpe] 233 234 *) New function X509_POLICY_NODE_print() which prints out policy nodes. 235 [Steve Henson] 236 237 *) Add new EVP function EVP_CIPHER_CTX_rand_key and associated functionality. 238 This will generate a random key of the appropriate length based on the 239 cipher context. The EVP_CIPHER can provide its own random key generation 240 routine to support keys of a specific form. This is used in the des and 241 3des routines to generate a key of the correct parity. Update S/MIME 242 code to use new functions and hence generate correct parity DES keys. 243 Add EVP_CHECK_DES_KEY #define to return an error if the key is not 244 valid (weak or incorrect parity). 245 [Steve Henson] 246 247 *) Add a local set of CRLs that can be used by X509_verify_cert() as well 248 as looking them up. This is useful when the verified structure may contain 249 CRLs, for example PKCS#7 signedData. Modify PKCS7_verify() to use any CRLs 250 present unless the new PKCS7_NO_CRL flag is asserted. 251 [Steve Henson] 252 253 *) Extend ASN1 oid configuration module. It now additionally accepts the 254 syntax: 255 256 shortName = some long name, 1.2.3.4 257 [Steve Henson] 258 259 *) Reimplemented the BN_CTX implementation. There is now no more static 260 limitation on the number of variables it can handle nor the depth of the 261 "stack" handling for BN_CTX_start()/BN_CTX_end() pairs. The stack 262 information can now expand as required, and rather than having a single 263 static array of bignums, BN_CTX now uses a linked-list of such arrays 264 allowing it to expand on demand whilst maintaining the usefulness of 265 BN_CTX's "bundling". 266 [Geoff Thorpe] 267 268 *) Add a missing BN_CTX parameter to the 'rsa_mod_exp' callback in RSA_METHOD 269 to allow all RSA operations to function using a single BN_CTX. 270 [Geoff Thorpe] 271 272 *) Preliminary support for certificate policy evaluation and checking. This 273 is initially intended to pass the tests outlined in "Conformance Testing 274 of Relying Party Client Certificate Path Processing Logic" v1.07. 275 [Steve Henson] 276 277 *) bn_dup_expand() has been deprecated, it was introduced in 0.9.7 and 278 remained unused and not that useful. A variety of other little bignum 279 tweaks and fixes have also been made continuing on from the audit (see 280 below). 281 [Geoff Thorpe] 282 283 *) Constify all or almost all d2i, c2i, s2i and r2i functions, along with 284 associated ASN1, EVP and SSL functions and old ASN1 macros. 285 [Richard Levitte] 286 287 *) BN_zero() only needs to set 'top' and 'neg' to zero for correct results, 288 and this should never fail. So the return value from the use of 289 BN_set_word() (which can fail due to needless expansion) is now deprecated; 290 if OPENSSL_NO_DEPRECATED is defined, BN_zero() is a void macro. 291 [Geoff Thorpe] 292 293 *) BN_CTX_get() should return zero-valued bignums, providing the same 294 initialised value as BN_new(). 295 [Geoff Thorpe, suggested by Ulf M�ller] 296 297 *) Support for inhibitAnyPolicy certificate extension. 298 [Steve Henson] 299 300 *) An audit of the BIGNUM code is underway, for which debugging code is 301 enabled when BN_DEBUG is defined. This makes stricter enforcements on what 302 is considered valid when processing BIGNUMs, and causes execution to 303 assert() when a problem is discovered. If BN_DEBUG_RAND is defined, 304 further steps are taken to deliberately pollute unused data in BIGNUM 305 structures to try and expose faulty code further on. For now, openssl will 306 (in its default mode of operation) continue to tolerate the inconsistent 307 forms that it has tolerated in the past, but authors and packagers should 308 consider trying openssl and their own applications when compiled with 309 these debugging symbols defined. It will help highlight potential bugs in 310 their own code, and will improve the test coverage for OpenSSL itself. At 311 some point, these tighter rules will become openssl's default to improve 312 maintainability, though the assert()s and other overheads will remain only 313 in debugging configurations. See bn.h for more details. 314 [Geoff Thorpe, Nils Larsch, Ulf M�ller] 315 316 *) BN_CTX_init() has been deprecated, as BN_CTX is an opaque structure 317 that can only be obtained through BN_CTX_new() (which implicitly 318 initialises it). The presence of this function only made it possible 319 to overwrite an existing structure (and cause memory leaks). 320 [Geoff Thorpe] 321 322 *) Because of the callback-based approach for implementing LHASH as a 323 template type, lh_insert() adds opaque objects to hash-tables and 324 lh_doall() or lh_doall_arg() are typically used with a destructor callback 325 to clean up those corresponding objects before destroying the hash table 326 (and losing the object pointers). So some over-zealous constifications in 327 LHASH have been relaxed so that lh_insert() does not take (nor store) the 328 objects as "const" and the lh_doall[_arg] callback wrappers are not 329 prototyped to have "const" restrictions on the object pointers they are 330 given (and so aren't required to cast them away any more). 331 [Geoff Thorpe] 332 333 *) The tmdiff.h API was so ugly and minimal that our own timing utility 334 (speed) prefers to use its own implementation. The two implementations 335 haven't been consolidated as yet (volunteers?) but the tmdiff API has had 336 its object type properly exposed (MS_TM) instead of casting to/from "char 337 *". This may still change yet if someone realises MS_TM and "ms_time_***" 338 aren't necessarily the greatest nomenclatures - but this is what was used 339 internally to the implementation so I've used that for now. 340 [Geoff Thorpe] 341 342 *) Ensure that deprecated functions do not get compiled when 343 OPENSSL_NO_DEPRECATED is defined. Some "openssl" subcommands and a few of 344 the self-tests were still using deprecated key-generation functions so 345 these have been updated also. 346 [Geoff Thorpe] 347 348 *) Reorganise PKCS#7 code to separate the digest location functionality 349 into PKCS7_find_digest(), digest addtion into PKCS7_bio_add_digest(). 350 New function PKCS7_set_digest() to set the digest type for PKCS#7 351 digestedData type. Add additional code to correctly generate the 352 digestedData type and add support for this type in PKCS7 initialization 353 functions. 354 [Steve Henson] 355 356 *) New function PKCS7_set0_type_other() this initializes a PKCS7 357 structure of type "other". 358 [Steve Henson] 359 360 *) Fix prime generation loop in crypto/bn/bn_prime.pl by making 361 sure the loop does correctly stop and breaking ("division by zero") 362 modulus operations are not performed. The (pre-generated) prime 363 table crypto/bn/bn_prime.h was already correct, but it could not be 364 re-generated on some platforms because of the "division by zero" 365 situation in the script. 366 [Ralf S. Engelschall] 367 368 *) Update support for ECC-based TLS ciphersuites according to 369 draft-ietf-tls-ecc-03.txt: the KDF1 key derivation function with 370 SHA-1 now is only used for "small" curves (where the 371 representation of a field element takes up to 24 bytes); for 372 larger curves, the field element resulting from ECDH is directly 373 used as premaster secret. 374 [Douglas Stebila (Sun Microsystems Laboratories)] 375 376 *) Add code for kP+lQ timings to crypto/ec/ectest.c, and add SEC2 377 curve secp160r1 to the tests. 378 [Douglas Stebila (Sun Microsystems Laboratories)] 379 380 *) Add the possibility to load symbols globally with DSO. 381 [G�tz Babin-Ebell <babin-ebell@trustcenter.de> via Richard Levitte] 382 383 *) Add the functions ERR_set_mark() and ERR_pop_to_mark() for better 384 control of the error stack. 385 [Richard Levitte] 386 387 *) Add support for STORE in ENGINE. 388 [Richard Levitte] 389 390 *) Add the STORE type. The intention is to provide a common interface 391 to certificate and key stores, be they simple file-based stores, or 392 HSM-type store, or LDAP stores, or... 393 NOTE: The code is currently UNTESTED and isn't really used anywhere. 394 [Richard Levitte] 395 396 *) Add a generic structure called OPENSSL_ITEM. This can be used to 397 pass a list of arguments to any function as well as provide a way 398 for a function to pass data back to the caller. 399 [Richard Levitte] 400 401 *) Add the functions BUF_strndup() and BUF_memdup(). BUF_strndup() 402 works like BUF_strdup() but can be used to duplicate a portion of 403 a string. The copy gets NUL-terminated. BUF_memdup() duplicates 404 a memory area. 405 [Richard Levitte] 406 407 *) Add the function sk_find_ex() which works like sk_find(), but will 408 return an index to an element even if an exact match couldn't be 409 found. The index is guaranteed to point at the element where the 410 searched-for key would be inserted to preserve sorting order. 411 [Richard Levitte] 412 413 *) Add the function OBJ_bsearch_ex() which works like OBJ_bsearch() but 414 takes an extra flags argument for optional functionality. Currently, 415 the following flags are defined: 416 417 OBJ_BSEARCH_VALUE_ON_NOMATCH 418 This one gets OBJ_bsearch_ex() to return a pointer to the first 419 element where the comparing function returns a negative or zero 420 number. 421 422 OBJ_BSEARCH_FIRST_VALUE_ON_MATCH 423 This one gets OBJ_bsearch_ex() to return a pointer to the first 424 element where the comparing function returns zero. This is useful 425 if there are more than one element where the comparing function 426 returns zero. 427 [Richard Levitte] 428 429 *) Make it possible to create self-signed certificates with 'openssl ca' 430 in such a way that the self-signed certificate becomes part of the 431 CA database and uses the same mechanisms for serial number generation 432 as all other certificate signing. The new flag '-selfsign' enables 433 this functionality. Adapt CA.sh and CA.pl.in. 434 [Richard Levitte] 435 436 *) Add functionality to check the public key of a certificate request 437 against a given private. This is useful to check that a certificate 438 request can be signed by that key (self-signing). 439 [Richard Levitte] 440 441 *) Make it possible to have multiple active certificates with the same 442 subject in the CA index file. This is done only if the keyword 443 'unique_subject' is set to 'no' in the main CA section (default 444 if 'CA_default') of the configuration file. The value is saved 445 with the database itself in a separate index attribute file, 446 named like the index file with '.attr' appended to the name. 447 [Richard Levitte] 448 449 *) Generate muti valued AVAs using '+' notation in config files for 450 req and dirName. 451 [Steve Henson] 452 453 *) Support for nameConstraints certificate extension. 454 [Steve Henson] 455 456 *) Support for policyConstraints certificate extension. 457 [Steve Henson] 458 459 *) Support for policyMappings certificate extension. 460 [Steve Henson] 461 462 *) Make sure the default DSA_METHOD implementation only uses its 463 dsa_mod_exp() and/or bn_mod_exp() handlers if they are non-NULL, 464 and change its own handlers to be NULL so as to remove unnecessary 465 indirection. This lets alternative implementations fallback to the 466 default implementation more easily. 467 [Geoff Thorpe] 468 469 *) Support for directoryName in GeneralName related extensions 470 in config files. 471 [Steve Henson] 472 473 *) Make it possible to link applications using Makefile.shared. 474 Make that possible even when linking against static libraries! 475 [Richard Levitte] 476 477 *) Support for single pass processing for S/MIME signing. This now 478 means that S/MIME signing can be done from a pipe, in addition 479 cleartext signing (multipart/signed type) is effectively streaming 480 and the signed data does not need to be all held in memory. 481 482 This is done with a new flag PKCS7_STREAM. When this flag is set 483 PKCS7_sign() only initializes the PKCS7 structure and the actual signing 484 is done after the data is output (and digests calculated) in 485 SMIME_write_PKCS7(). 486 [Steve Henson] 487 488 *) Add full support for -rpath/-R, both in shared libraries and 489 applications, at least on the platforms where it's known how 490 to do it. 491 [Richard Levitte] 492 493 *) In crypto/ec/ec_mult.c, implement fast point multiplication with 494 precomputation, based on wNAF splitting: EC_GROUP_precompute_mult() 495 will now compute a table of multiples of the generator that 496 makes subsequent invocations of EC_POINTs_mul() or EC_POINT_mul() 497 faster (notably in the case of a single point multiplication, 498 scalar * generator). 499 [Nils Larsch, Bodo Moeller] 500 501 *) IPv6 support for certificate extensions. The various extensions 502 which use the IP:a.b.c.d can now take IPv6 addresses using the 503 formats of RFC1884 2.2 . IPv6 addresses are now also displayed 504 correctly. 505 [Steve Henson] 506 507 *) Added an ENGINE that implements RSA by performing private key 508 exponentiations with the GMP library. The conversions to and from 509 GMP's mpz_t format aren't optimised nor are any montgomery forms 510 cached, and on x86 it appears OpenSSL's own performance has caught up. 511 However there are likely to be other architectures where GMP could 512 provide a boost. This ENGINE is not built in by default, but it can be 513 specified at Configure time and should be accompanied by the necessary 514 linker additions, eg; 515 ./config -DOPENSSL_USE_GMP -lgmp 516 [Geoff Thorpe] 517 518 *) "openssl engine" will not display ENGINE/DSO load failure errors when 519 testing availability of engines with "-t" - the old behaviour is 520 produced by increasing the feature's verbosity with "-tt". 521 [Geoff Thorpe] 522 523 *) ECDSA routines: under certain error conditions uninitialized BN objects 524 could be freed. Solution: make sure initialization is performed early 525 enough. (Reported and fix supplied by Nils Larsch <nla@trustcenter.de> 526 via PR#459) 527 [Lutz Jaenicke] 528 529 *) Key-generation can now be implemented in RSA_METHOD, DSA_METHOD 530 and DH_METHOD (eg. by ENGINE implementations) to override the normal 531 software implementations. For DSA and DH, parameter generation can 532 also be overriden by providing the appropriate method callbacks. 533 [Geoff Thorpe] 534 535 *) Change the "progress" mechanism used in key-generation and 536 primality testing to functions that take a new BN_GENCB pointer in 537 place of callback/argument pairs. The new API functions have "_ex" 538 postfixes and the older functions are reimplemented as wrappers for 539 the new ones. The OPENSSL_NO_DEPRECATED symbol can be used to hide 540 declarations of the old functions to help (graceful) attempts to 541 migrate to the new functions. Also, the new key-generation API 542 functions operate on a caller-supplied key-structure and return 543 success/failure rather than returning a key or NULL - this is to 544 help make "keygen" another member function of RSA_METHOD etc. 545 546 Example for using the new callback interface: 547 548 int (*my_callback)(int a, int b, BN_GENCB *cb) = ...; 549 void *my_arg = ...; 550 BN_GENCB my_cb; 551 552 BN_GENCB_set(&my_cb, my_callback, my_arg); 553 554 return BN_is_prime_ex(some_bignum, BN_prime_checks, NULL, &cb); 555 /* For the meaning of a, b in calls to my_callback(), see the 556 * documentation of the function that calls the callback. 557 * cb will point to my_cb; my_arg can be retrieved as cb->arg. 558 * my_callback should return 1 if it wants BN_is_prime_ex() 559 * to continue, or 0 to stop. 560 */ 561 562 [Geoff Thorpe] 563 564 *) Change the ZLIB compression method to be stateful, and make it 565 available to TLS with the number defined in 566 draft-ietf-tls-compression-04.txt. 567 [Richard Levitte] 568 569 *) Add the ASN.1 structures and functions for CertificatePair, which 570 is defined as follows (according to X.509_4thEditionDraftV6.pdf): 571 572 CertificatePair ::= SEQUENCE { 573 forward [0] Certificate OPTIONAL, 574 reverse [1] Certificate OPTIONAL, 575 -- at least one of the pair shall be present -- } 576 577 Also implement the PEM functions to read and write certificate 578 pairs, and defined the PEM tag as "CERTIFICATE PAIR". 579 580 This needed to be defined, mostly for the sake of the LDAP 581 attribute crossCertificatePair, but may prove useful elsewhere as 582 well. 583 [Richard Levitte] 584 585 *) Make it possible to inhibit symlinking of shared libraries in 586 Makefile.shared, for Cygwin's sake. 587 [Richard Levitte] 588 589 *) Extend the BIGNUM API by creating a function 590 void BN_set_negative(BIGNUM *a, int neg); 591 and a macro that behave like 592 int BN_is_negative(const BIGNUM *a); 593 594 to avoid the need to access 'a->neg' directly in applications. 595 [Nils Larsch] 596 597 *) Implement fast modular reduction for pseudo-Mersenne primes 598 used in NIST curves (crypto/bn/bn_nist.c, crypto/ec/ecp_nist.c). 599 EC_GROUP_new_curve_GFp() will now automatically use this 600 if applicable. 601 [Nils Larsch <nla@trustcenter.de>] 602 603 *) Add new lock type (CRYPTO_LOCK_BN). 604 [Bodo Moeller] 605 606 *) Change the ENGINE framework to automatically load engines 607 dynamically from specific directories unless they could be 608 found to already be built in or loaded. Move all the 609 current engines except for the cryptodev one to a new 610 directory engines/. 611 The engines in engines/ are built as shared libraries if 612 the "shared" options was given to ./Configure or ./config. 613 Otherwise, they are inserted in libcrypto.a. 614 /usr/local/ssl/engines is the default directory for dynamic 615 engines, but that can be overriden at configure time through 616 the usual use of --prefix and/or --openssldir, and at run 617 time with the environment variable OPENSSL_ENGINES. 618 [Geoff Thorpe and Richard Levitte] 619 620 *) Add Makefile.shared, a helper makefile to build shared 621 libraries. Addapt Makefile.org. 622 [Richard Levitte] 623 624 *) Add version info to Win32 DLLs. 625 [Peter 'Luna' Runestig" <peter@runestig.com>] 626 627 *) Add new 'medium level' PKCS#12 API. Certificates and keys 628 can be added using this API to created arbitrary PKCS#12 629 files while avoiding the low level API. 630 631 New options to PKCS12_create(), key or cert can be NULL and 632 will then be omitted from the output file. The encryption 633 algorithm NIDs can be set to -1 for no encryption, the mac 634 iteration count can be set to 0 to omit the mac. 635 636 Enhance pkcs12 utility by making the -nokeys and -nocerts 637 options work when creating a PKCS#12 file. New option -nomac 638 to omit the mac, NONE can be set for an encryption algorithm. 639 New code is modified to use the enhanced PKCS12_create() 640 instead of the low level API. 641 [Steve Henson] 642 643 *) Extend ASN1 encoder to support indefinite length constructed 644 encoding. This can output sequences tags and octet strings in 645 this form. Modify pk7_asn1.c to support indefinite length 646 encoding. This is experimental and needs additional code to 647 be useful, such as an ASN1 bio and some enhanced streaming 648 PKCS#7 code. 649 650 Extend template encode functionality so that tagging is passed 651 down to the template encoder. 652 [Steve Henson] 653 654 *) Let 'openssl req' fail if an argument to '-newkey' is not 655 recognized instead of using RSA as a default. 656 [Bodo Moeller] 657 658 *) Add support for ECC-based ciphersuites from draft-ietf-tls-ecc-01.txt. 659 As these are not official, they are not included in "ALL"; 660 the "ECCdraft" ciphersuite group alias can be used to select them. 661 [Vipul Gupta and Sumit Gupta (Sun Microsystems Laboratories)] 662 663 *) Add ECDH engine support. 664 [Nils Gura and Douglas Stebila (Sun Microsystems Laboratories)] 665 666 *) Add ECDH in new directory crypto/ecdh/. 667 [Douglas Stebila (Sun Microsystems Laboratories)] 668 669 *) Let BN_rand_range() abort with an error after 100 iterations 670 without success (which indicates a broken PRNG). 671 [Bodo Moeller] 672 673 *) Change BN_mod_sqrt() so that it verifies that the input value 674 is really the square of the return value. (Previously, 675 BN_mod_sqrt would show GIGO behaviour.) 676 [Bodo Moeller] 677 678 *) Add named elliptic curves over binary fields from X9.62, SECG, 679 and WAP/WTLS; add OIDs that were still missing. 680 681 [Sheueling Chang Shantz and Douglas Stebila 682 (Sun Microsystems Laboratories)] 683 684 *) Extend the EC library for elliptic curves over binary fields 685 (new files ec2_smpl.c, ec2_smpt.c, ec2_mult.c in crypto/ec/). 686 New EC_METHOD: 687 688 EC_GF2m_simple_method 689 690 New API functions: 691 692 EC_GROUP_new_curve_GF2m 693 EC_GROUP_set_curve_GF2m 694 EC_GROUP_get_curve_GF2m 695 EC_POINT_set_affine_coordinates_GF2m 696 EC_POINT_get_affine_coordinates_GF2m 697 EC_POINT_set_compressed_coordinates_GF2m 698 699 Point compression for binary fields is disabled by default for 700 patent reasons (compile with OPENSSL_EC_BIN_PT_COMP defined to 701 enable it). 702 703 As binary polynomials are represented as BIGNUMs, various members 704 of the EC_GROUP and EC_POINT data structures can be shared 705 between the implementations for prime fields and binary fields; 706 the above ..._GF2m functions (except for EX_GROUP_new_curve_GF2m) 707 are essentially identical to their ..._GFp counterparts. 708 (For simplicity, the '..._GFp' prefix has been dropped from 709 various internal method names.) 710 711 An internal 'field_div' method (similar to 'field_mul' and 712 'field_sqr') has been added; this is used only for binary fields. 713 714 [Sheueling Chang Shantz and Douglas Stebila 715 (Sun Microsystems Laboratories)] 716 717 *) Optionally dispatch EC_POINT_mul(), EC_POINT_precompute_mult() 718 through methods ('mul', 'precompute_mult'). 719 720 The generic implementations (now internally called 'ec_wNAF_mul' 721 and 'ec_wNAF_precomputed_mult') remain the default if these 722 methods are undefined. 723 724 [Sheueling Chang Shantz and Douglas Stebila 725 (Sun Microsystems Laboratories)] 726 727 *) New function EC_GROUP_get_degree, which is defined through 728 EC_METHOD. For curves over prime fields, this returns the bit 729 length of the modulus. 730 731 [Sheueling Chang Shantz and Douglas Stebila 732 (Sun Microsystems Laboratories)] 733 734 *) New functions EC_GROUP_dup, EC_POINT_dup. 735 (These simply call ..._new and ..._copy). 736 737 [Sheueling Chang Shantz and Douglas Stebila 738 (Sun Microsystems Laboratories)] 739 740 *) Add binary polynomial arithmetic software in crypto/bn/bn_gf2m.c. 741 Polynomials are represented as BIGNUMs (where the sign bit is not 742 used) in the following functions [macros]: 743 744 BN_GF2m_add 745 BN_GF2m_sub [= BN_GF2m_add] 746 BN_GF2m_mod [wrapper for BN_GF2m_mod_arr] 747 BN_GF2m_mod_mul [wrapper for BN_GF2m_mod_mul_arr] 748 BN_GF2m_mod_sqr [wrapper for BN_GF2m_mod_sqr_arr] 749 BN_GF2m_mod_inv 750 BN_GF2m_mod_exp [wrapper for BN_GF2m_mod_exp_arr] 751 BN_GF2m_mod_sqrt [wrapper for BN_GF2m_mod_sqrt_arr] 752 BN_GF2m_mod_solve_quad [wrapper for BN_GF2m_mod_solve_quad_arr] 753 BN_GF2m_cmp [= BN_ucmp] 754 755 (Note that only the 'mod' functions are actually for fields GF(2^m). 756 BN_GF2m_add() is misnomer, but this is for the sake of consistency.) 757 758 For some functions, an the irreducible polynomial defining a 759 field can be given as an 'unsigned int[]' with strictly 760 decreasing elements giving the indices of those bits that are set; 761 i.e., p[] represents the polynomial 762 f(t) = t^p[0] + t^p[1] + ... + t^p[k] 763 where 764 p[0] > p[1] > ... > p[k] = 0. 765 This applies to the following functions: 766 767 BN_GF2m_mod_arr 768 BN_GF2m_mod_mul_arr 769 BN_GF2m_mod_sqr_arr 770 BN_GF2m_mod_inv_arr [wrapper for BN_GF2m_mod_inv] 771 BN_GF2m_mod_div_arr [wrapper for BN_GF2m_mod_div] 772 BN_GF2m_mod_exp_arr 773 BN_GF2m_mod_sqrt_arr 774 BN_GF2m_mod_solve_quad_arr 775 BN_GF2m_poly2arr 776 BN_GF2m_arr2poly 777 778 Conversion can be performed by the following functions: 779 780 BN_GF2m_poly2arr 781 BN_GF2m_arr2poly 782 783 bntest.c has additional tests for binary polynomial arithmetic. 784 785 Two implementations for BN_GF2m_mod_div() are available. 786 The default algorithm simply uses BN_GF2m_mod_inv() and 787 BN_GF2m_mod_mul(). The alternative algorithm is compiled in only 788 if OPENSSL_SUN_GF2M_DIV is defined (patent pending; read the 789 copyright notice in crypto/bn/bn_gf2m.c before enabling it). 790 791 [Sheueling Chang Shantz and Douglas Stebila 792 (Sun Microsystems Laboratories)] 793 794 *) Add new error code 'ERR_R_DISABLED' that can be used when some 795 functionality is disabled at compile-time. 796 [Douglas Stebila <douglas.stebila@sun.com>] 797 798 *) Change default behaviour of 'openssl asn1parse' so that more 799 information is visible when viewing, e.g., a certificate: 800 801 Modify asn1_parse2 (crypto/asn1/asn1_par.c) so that in non-'dump' 802 mode the content of non-printable OCTET STRINGs is output in a 803 style similar to INTEGERs, but with '[HEX DUMP]' prepended to 804 avoid the appearance of a printable string. 805 [Nils Larsch <nla@trustcenter.de>] 806 807 *) Add 'asn1_flag' and 'asn1_form' member to EC_GROUP with access 808 functions 809 EC_GROUP_set_asn1_flag() 810 EC_GROUP_get_asn1_flag() 811 EC_GROUP_set_point_conversion_form() 812 EC_GROUP_get_point_conversion_form() 813 These control ASN1 encoding details: 814 - Curves (i.e., groups) are encoded explicitly unless asn1_flag 815 has been set to OPENSSL_EC_NAMED_CURVE. 816 - Points are encoded in uncompressed form by default; options for 817 asn1_for are as for point2oct, namely 818 POINT_CONVERSION_COMPRESSED 819 POINT_CONVERSION_UNCOMPRESSED 820 POINT_CONVERSION_HYBRID 821 822 Also add 'seed' and 'seed_len' members to EC_GROUP with access 823 functions 824 EC_GROUP_set_seed() 825 EC_GROUP_get0_seed() 826 EC_GROUP_get_seed_len() 827 This is used only for ASN1 purposes (so far). 828 [Nils Larsch <nla@trustcenter.de>] 829 830 *) Add 'field_type' member to EC_METHOD, which holds the NID 831 of the appropriate field type OID. The new function 832 EC_METHOD_get_field_type() returns this value. 833 [Nils Larsch <nla@trustcenter.de>] 834 835 *) Add functions 836 EC_POINT_point2bn() 837 EC_POINT_bn2point() 838 EC_POINT_point2hex() 839 EC_POINT_hex2point() 840 providing useful interfaces to EC_POINT_point2oct() and 841 EC_POINT_oct2point(). 842 [Nils Larsch <nla@trustcenter.de>] 843 844 *) Change internals of the EC library so that the functions 845 EC_GROUP_set_generator() 846 EC_GROUP_get_generator() 847 EC_GROUP_get_order() 848 EC_GROUP_get_cofactor() 849 are implemented directly in crypto/ec/ec_lib.c and not dispatched 850 to methods, which would lead to unnecessary code duplication when 851 adding different types of curves. 852 [Nils Larsch <nla@trustcenter.de> with input by Bodo Moeller] 853 854 *) Implement compute_wNAF (crypto/ec/ec_mult.c) without BIGNUM 855 arithmetic, and such that modified wNAFs are generated 856 (which avoid length expansion in many cases). 857 [Bodo Moeller] 858 859 *) Add a function EC_GROUP_check_discriminant() (defined via 860 EC_METHOD) that verifies that the curve discriminant is non-zero. 861 862 Add a function EC_GROUP_check() that makes some sanity tests 863 on a EC_GROUP, its generator and order. This includes 864 EC_GROUP_check_discriminant(). 865 [Nils Larsch <nla@trustcenter.de>] 866 867 *) Add ECDSA in new directory crypto/ecdsa/. 868 869 Add applications 'openssl ecparam' and 'openssl ecdsa' 870 (these are based on 'openssl dsaparam' and 'openssl dsa'). 871 872 ECDSA support is also included in various other files across the 873 library. Most notably, 874 - 'openssl req' now has a '-newkey ecdsa:file' option; 875 - EVP_PKCS82PKEY (crypto/evp/evp_pkey.c) now can handle ECDSA; 876 - X509_PUBKEY_get (crypto/asn1/x_pubkey.c) and 877 d2i_PublicKey (crypto/asn1/d2i_pu.c) have been modified to make 878 them suitable for ECDSA where domain parameters must be 879 extracted before the specific public key; 880 - ECDSA engine support has been added. 881 [Nils Larsch <nla@trustcenter.de>] 882 883 *) Include some named elliptic curves, and add OIDs from X9.62, 884 SECG, and WAP/WTLS. Each curve can be obtained from the new 885 function 886 EC_GROUP_new_by_curve_name(), 887 and the list of available named curves can be obtained with 888 EC_get_builtin_curves(). 889 Also add a 'curve_name' member to EC_GROUP objects, which can be 890 accessed via 891 EC_GROUP_set_curve_name() 892 EC_GROUP_get_curve_name() 893 [Nils Larsch <larsch@trustcenter.de, Bodo Moeller] 894 895 *) Remove a few calls to bn_wexpand() in BN_sqr() (the one in there 896 was actually never needed) and in BN_mul(). The removal in BN_mul() 897 required a small change in bn_mul_part_recursive() and the addition 898 of the functions bn_cmp_part_words(), bn_sub_part_words() and 899 bn_add_part_words(), which do the same thing as bn_cmp_words(), 900 bn_sub_words() and bn_add_words() except they take arrays with 901 differing sizes. 902 [Richard Levitte] 903 904 Changes between 0.9.7h and 0.9.7i [14 Oct 2005] 905 906 *) Wrapped the definition of EVP_MAX_MD_SIZE in a #ifdef OPENSSL_FIPS. 907 The value now differs depending on if you build for FIPS or not. 908 BEWARE! A program linked with a shared FIPSed libcrypto can't be 909 safely run with a non-FIPSed libcrypto, as it may crash because of 910 the difference induced by this change. 911 [Andy Polyakov] 912 913 Changes between 0.9.7g and 0.9.7h [11 Oct 2005] 914 915 *) Remove the functionality of SSL_OP_MSIE_SSLV2_RSA_PADDING 916 (part of SSL_OP_ALL). This option used to disable the 917 countermeasure against man-in-the-middle protocol-version 918 rollback in the SSL 2.0 server implementation, which is a bad 919 idea. (CVE-2005-2969) 920 921 [Bodo Moeller; problem pointed out by Yutaka Oiwa (Research Center 922 for Information Security, National Institute of Advanced Industrial 923 Science and Technology [AIST], Japan)] 924 925 *) Minimal support for X9.31 signatures and PSS padding modes. This is 926 mainly for FIPS compliance and not fully integrated at this stage. 927 [Steve Henson] 928 929 *) For DSA signing, unless DSA_FLAG_NO_EXP_CONSTTIME is set, perform 930 the exponentiation using a fixed-length exponent. (Otherwise, 931 the information leaked through timing could expose the secret key 932 after many signatures; cf. Bleichenbacher's attack on DSA with 933 biased k.) 934 [Bodo Moeller] 935 936 *) Make a new fixed-window mod_exp implementation the default for 937 RSA, DSA, and DH private-key operations so that the sequence of 938 squares and multiplies and the memory access pattern are 939 independent of the particular secret key. This will mitigate 940 cache-timing and potential related attacks. 941 942 BN_mod_exp_mont_consttime() is the new exponentiation implementation, 943 and this is automatically used by BN_mod_exp_mont() if the new flag 944 BN_FLG_EXP_CONSTTIME is set for the exponent. RSA, DSA, and DH 945 will use this BN flag for private exponents unless the flag 946 RSA_FLAG_NO_EXP_CONSTTIME, DSA_FLAG_NO_EXP_CONSTTIME, or 947 DH_FLAG_NO_EXP_CONSTTIME, respectively, is set. 948 949 [Matthew D Wood (Intel Corp), with some changes by Bodo Moeller] 950 951 *) Change the client implementation for SSLv23_method() and 952 SSLv23_client_method() so that is uses the SSL 3.0/TLS 1.0 953 Client Hello message format if the SSL_OP_NO_SSLv2 option is set. 954 (Previously, the SSL 2.0 backwards compatible Client Hello 955 message format would be used even with SSL_OP_NO_SSLv2.) 956 [Bodo Moeller] 957 958 *) Add support for smime-type MIME parameter in S/MIME messages which some 959 clients need. 960 [Steve Henson] 961 962 *) New function BN_MONT_CTX_set_locked() to set montgomery parameters in 963 a threadsafe manner. Modify rsa code to use new function and add calls 964 to dsa and dh code (which had race conditions before). 965 [Steve Henson] 966 967 *) Include the fixed error library code in the C error file definitions 968 instead of fixing them up at runtime. This keeps the error code 969 structures constant. 970 [Steve Henson] 971 972 Changes between 0.9.7f and 0.9.7g [11 Apr 2005] 973 974 [NB: OpenSSL 0.9.7h and later 0.9.7 patch levels were released after 975 OpenSSL 0.9.8.] 976 977 *) Fixes for newer kerberos headers. NB: the casts are needed because 978 the 'length' field is signed on one version and unsigned on another 979 with no (?) obvious way to tell the difference, without these VC++ 980 complains. Also the "definition" of FAR (blank) is no longer included 981 nor is the error ENOMEM. KRB5_PRIVATE has to be set to 1 to pick up 982 some needed definitions. 983 [Steve Henson] 984 985 *) Undo Cygwin change. 986 [Ulf M�ller] 987 988 *) Added support for proxy certificates according to RFC 3820. 989 Because they may be a security thread to unaware applications, 990 they must be explicitely allowed in run-time. See 991 docs/HOWTO/proxy_certificates.txt for further information. 992 [Richard Levitte] 993 994 Changes between 0.9.7e and 0.9.7f [22 Mar 2005] 995 996 *) Use (SSL_RANDOM_VALUE - 4) bytes of pseudo random data when generating 997 server and client random values. Previously 998 (SSL_RANDOM_VALUE - sizeof(time_t)) would be used which would result in 999 less random data when sizeof(time_t) > 4 (some 64 bit platforms). 1000 1001 This change has negligible security impact because: 1002 1003 1. Server and client random values still have 24 bytes of pseudo random 1004 data. 1005 1006 2. Server and client random values are sent in the clear in the initial 1007 handshake. 1008 1009 3. The master secret is derived using the premaster secret (48 bytes in 1010 size for static RSA ciphersuites) as well as client server and random 1011 values. 1012 1013 The OpenSSL team would like to thank the UK NISCC for bringing this issue 1014 to our attention. 1015 1016 [Stephen Henson, reported by UK NISCC] 1017 1018 *) Use Windows randomness collection on Cygwin. 1019 [Ulf M�ller] 1020 1021 *) Fix hang in EGD/PRNGD query when communication socket is closed 1022 prematurely by EGD/PRNGD. 1023 [Darren Tucker <dtucker@zip.com.au> via Lutz J�nicke, resolves #1014] 1024 1025 *) Prompt for pass phrases when appropriate for PKCS12 input format. 1026 [Steve Henson] 1027 1028 *) Back-port of selected performance improvements from development 1029 branch, as well as improved support for PowerPC platforms. 1030 [Andy Polyakov] 1031 1032 *) Add lots of checks for memory allocation failure, error codes to indicate 1033 failure and freeing up memory if a failure occurs. 1034 [Nauticus Networks SSL Team <openssl@nauticusnet.com>, Steve Henson] 1035 1036 *) Add new -passin argument to dgst. 1037 [Steve Henson] 1038 1039 *) Perform some character comparisons of different types in X509_NAME_cmp: 1040 this is needed for some certificates that reencode DNs into UTF8Strings 1041 (in violation of RFC3280) and can't or wont issue name rollover 1042 certificates. 1043 [Steve Henson] 1044 1045 *) Make an explicit check during certificate validation to see that 1046 the CA setting in each certificate on the chain is correct. As a 1047 side effect always do the following basic checks on extensions, 1048 not just when there's an associated purpose to the check: 1049 1050 - if there is an unhandled critical extension (unless the user 1051 has chosen to ignore this fault) 1052 - if the path length has been exceeded (if one is set at all) 1053 - that certain extensions fit the associated purpose (if one has 1054 been given) 1055 [Richard Levitte] 1056 1057 Changes between 0.9.7d and 0.9.7e [25 Oct 2004] 1058 1059 *) Avoid a race condition when CRLs are checked in a multi threaded 1060 environment. This would happen due to the reordering of the revoked 1061 entries during signature checking and serial number lookup. Now the 1062 encoding is cached and the serial number sort performed under a lock. 1063 Add new STACK function sk_is_sorted(). 1064 [Steve Henson] 1065 1066 *) Add Delta CRL to the extension code. 1067 [Steve Henson] 1068 1069 *) Various fixes to s3_pkt.c so alerts are sent properly. 1070 [David Holmes <d.holmes@f5.com>] 1071 1072 *) Reduce the chances of duplicate issuer name and serial numbers (in 1073 violation of RFC3280) using the OpenSSL certificate creation utilities. 1074 This is done by creating a random 64 bit value for the initial serial 1075 number when a serial number file is created or when a self signed 1076 certificate is created using 'openssl req -x509'. The initial serial 1077 number file is created using 'openssl x509 -next_serial' in CA.pl 1078 rather than being initialized to 1. 1079 [Steve Henson] 1080 1081 Changes between 0.9.7c and 0.9.7d [17 Mar 2004] 1082 1083 *) Fix null-pointer assignment in do_change_cipher_spec() revealed 1084 by using the Codenomicon TLS Test Tool (CVE-2004-0079) 1085 [Joe Orton, Steve Henson] 1086 1087 *) Fix flaw in SSL/TLS handshaking when using Kerberos ciphersuites 1088 (CVE-2004-0112) 1089 [Joe Orton, Steve Henson] 1090 1091 *) Make it possible to have multiple active certificates with the same 1092 subject in the CA index file. This is done only if the keyword 1093 'unique_subject' is set to 'no' in the main CA section (default 1094 if 'CA_default') of the configuration file. The value is saved 1095 with the database itself in a separate index attribute file, 1096 named like the index file with '.attr' appended to the name. 1097 [Richard Levitte] 1098 1099 *) X509 verify fixes. Disable broken certificate workarounds when 1100 X509_V_FLAGS_X509_STRICT is set. Check CRL issuer has cRLSign set if 1101 keyUsage extension present. Don't accept CRLs with unhandled critical 1102 extensions: since verify currently doesn't process CRL extensions this 1103 rejects a CRL with *any* critical extensions. Add new verify error codes 1104 for these cases. 1105 [Steve Henson] 1106 1107 *) When creating an OCSP nonce use an OCTET STRING inside the extnValue. 1108 A clarification of RFC2560 will require the use of OCTET STRINGs and 1109 some implementations cannot handle the current raw format. Since OpenSSL 1110 copies and compares OCSP nonces as opaque blobs without any attempt at 1111 parsing them this should not create any compatibility issues. 1112 [Steve Henson] 1113 1114 *) New md flag EVP_MD_CTX_FLAG_REUSE this allows md_data to be reused when 1115 calling EVP_MD_CTX_copy_ex() to avoid calling OPENSSL_malloc(). Without 1116 this HMAC (and other) operations are several times slower than OpenSSL 1117 < 0.9.7. 1118 [Steve Henson] 1119 1120 *) Print out GeneralizedTime and UTCTime in ASN1_STRING_print_ex(). 1121 [Peter Sylvester <Peter.Sylvester@EdelWeb.fr>] 1122 1123 *) Use the correct content when signing type "other". 1124 [Steve Henson] 1125 1126 Changes between 0.9.7b and 0.9.7c [30 Sep 2003] 1127 1128 *) Fix various bugs revealed by running the NISCC test suite: 1129 1130 Stop out of bounds reads in the ASN1 code when presented with 1131 invalid tags (CVE-2003-0543 and CVE-2003-0544). 1132 1133 Free up ASN1_TYPE correctly if ANY type is invalid (CVE-2003-0545). 1134 1135 If verify callback ignores invalid public key errors don't try to check 1136 certificate signature with the NULL public key. 1137 1138 [Steve Henson] 1139 1140 *) New -ignore_err option in ocsp application to stop the server 1141 exiting on the first error in a request. 1142 [Steve Henson] 1143 1144 *) In ssl3_accept() (ssl/s3_srvr.c) only accept a client certificate 1145 if the server requested one: as stated in TLS 1.0 and SSL 3.0 1146 specifications. 1147 [Steve Henson] 1148 1149 *) In ssl3_get_client_hello() (ssl/s3_srvr.c), tolerate additional 1150 extra data after the compression methods not only for TLS 1.0 1151 but also for SSL 3.0 (as required by the specification). 1152 [Bodo Moeller; problem pointed out by Matthias Loepfe] 1153 1154 *) Change X509_certificate_type() to mark the key as exported/exportable 1155 when it's 512 *bits* long, not 512 bytes. 1156 [Richard Levitte] 1157 1158 *) Change AES_cbc_encrypt() so it outputs exact multiple of 1159 blocks during encryption. 1160 [Richard Levitte] 1161 1162 *) Various fixes to base64 BIO and non blocking I/O. On write 1163 flushes were not handled properly if the BIO retried. On read 1164 data was not being buffered properly and had various logic bugs. 1165 This also affects blocking I/O when the data being decoded is a 1166 certain size. 1167 [Steve Henson] 1168 1169 *) Various S/MIME bugfixes and compatibility changes: 1170 output correct application/pkcs7 MIME type if 1171 PKCS7_NOOLDMIMETYPE is set. Tolerate some broken signatures. 1172 Output CR+LF for EOL if PKCS7_CRLFEOL is set (this makes opening 1173 of files as .eml work). Correctly handle very long lines in MIME 1174 parser. 1175 [Steve Henson] 1176 1177 Changes between 0.9.7a and 0.9.7b [10 Apr 2003] 1178 1179 *) Countermeasure against the Klima-Pokorny-Rosa extension of 1180 Bleichbacher's attack on PKCS #1 v1.5 padding: treat 1181 a protocol version number mismatch like a decryption error 1182 in ssl3_get_client_key_exchange (ssl/s3_srvr.c). 1183 [Bodo Moeller] 1184 1185 *) Turn on RSA blinding by default in the default implementation 1186 to avoid a timing attack. Applications that don't want it can call 1187 RSA_blinding_off() or use the new flag RSA_FLAG_NO_BLINDING. 1188 They would be ill-advised to do so in most cases. 1189 [Ben Laurie, Steve Henson, Geoff Thorpe, Bodo Moeller] 1190 1191 *) Change RSA blinding code so that it works when the PRNG is not 1192 seeded (in this case, the secret RSA exponent is abused as 1193 an unpredictable seed -- if it is not unpredictable, there 1194 is no point in blinding anyway). Make RSA blinding thread-safe 1195 by remembering the creator's thread ID in rsa->blinding and 1196 having all other threads use local one-time blinding factors 1197 (this requires more computation than sharing rsa->blinding, but 1198 avoids excessive locking; and if an RSA object is not shared 1199 between threads, blinding will still be very fast). 1200 [Bodo Moeller] 1201 1202 *) Fixed a typo bug that would cause ENGINE_set_default() to set an 1203 ENGINE as defaults for all supported algorithms irrespective of 1204 the 'flags' parameter. 'flags' is now honoured, so applications 1205 should make sure they are passing it correctly. 1206 [Geoff Thorpe] 1207 1208 *) Target "mingw" now allows native Windows code to be generated in 1209 the Cygwin environment as well as with the MinGW compiler. 1210 [Ulf Moeller] 1211 1212 Changes between 0.9.7 and 0.9.7a [19 Feb 2003] 1213 1214 *) In ssl3_get_record (ssl/s3_pkt.c), minimize information leaked 1215 via timing by performing a MAC computation even if incorrrect 1216 block cipher padding has been found. This is a countermeasure 1217 against active attacks where the attacker has to distinguish 1218 between bad padding and a MAC verification error. (CVE-2003-0078) 1219 1220 [Bodo Moeller; problem pointed out by Brice Canvel (EPFL), 1221 Alain Hiltgen (UBS), Serge Vaudenay (EPFL), and 1222 Martin Vuagnoux (EPFL, Ilion)] 1223 1224 *) Make the no-err option work as intended. The intention with no-err 1225 is not to have the whole error stack handling routines removed from 1226 libcrypto, it's only intended to remove all the function name and 1227 reason texts, thereby removing some of the footprint that may not 1228 be interesting if those errors aren't displayed anyway. 1229 1230 NOTE: it's still possible for any application or module to have it's 1231 own set of error texts inserted. The routines are there, just not 1232 used by default when no-err is given. 1233 [Richard Levitte] 1234 1235 *) Add support for FreeBSD on IA64. 1236 [dirk.meyer@dinoex.sub.org via Richard Levitte, resolves #454] 1237 1238 *) Adjust DES_cbc_cksum() so it returns the same value as the MIT 1239 Kerberos function mit_des_cbc_cksum(). Before this change, 1240 the value returned by DES_cbc_cksum() was like the one from 1241 mit_des_cbc_cksum(), except the bytes were swapped. 1242 [Kevin Greaney <Kevin.Greaney@hp.com> and Richard Levitte] 1243 1244 *) Allow an application to disable the automatic SSL chain building. 1245 Before this a rather primitive chain build was always performed in 1246 ssl3_output_cert_chain(): an application had no way to send the 1247 correct chain if the automatic operation produced an incorrect result. 1248 1249 Now the chain builder is disabled if either: 1250 1251 1. Extra certificates are added via SSL_CTX_add_extra_chain_cert(). 1252 1253 2. The mode flag SSL_MODE_NO_AUTO_CHAIN is set. 1254 1255 The reasoning behind this is that an application would not want the 1256 auto chain building to take place if extra chain certificates are 1257 present and it might also want a means of sending no additional 1258 certificates (for example the chain has two certificates and the 1259 root is omitted). 1260 [Steve Henson] 1261 1262 *) Add the possibility to build without the ENGINE framework. 1263 [Steven Reddie <smr@essemer.com.au> via Richard Levitte] 1264 1265 *) Under Win32 gmtime() can return NULL: check return value in 1266 OPENSSL_gmtime(). Add error code for case where gmtime() fails. 1267 [Steve Henson] 1268 1269 *) DSA routines: under certain error conditions uninitialized BN objects 1270 could be freed. Solution: make sure initialization is performed early 1271 enough. (Reported and fix supplied by Ivan D Nestlerode <nestler@MIT.EDU>, 1272 Nils Larsch <nla@trustcenter.de> via PR#459) 1273 [Lutz Jaenicke] 1274 1275 *) Another fix for SSLv2 session ID handling: the session ID was incorrectly 1276 checked on reconnect on the client side, therefore session resumption 1277 could still fail with a "ssl session id is different" error. This 1278 behaviour is masked when SSL_OP_ALL is used due to 1279 SSL_OP_MICROSOFT_SESS_ID_BUG being set. 1280 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as 1281 followup to PR #377. 1282 [Lutz Jaenicke] 1283 1284 *) IA-32 assembler support enhancements: unified ELF targets, support 1285 for SCO/Caldera platforms, fix for Cygwin shared build. 1286 [Andy Polyakov] 1287 1288 *) Add support for FreeBSD on sparc64. As a consequence, support for 1289 FreeBSD on non-x86 processors is separate from x86 processors on 1290 the config script, much like the NetBSD support. 1291 [Richard Levitte & Kris Kennaway <kris@obsecurity.org>] 1292 1293 Changes between 0.9.6h and 0.9.7 [31 Dec 2002] 1294 1295 [NB: OpenSSL 0.9.6i and later 0.9.6 patch levels were released after 1296 OpenSSL 0.9.7.] 1297 1298 *) Fix session ID handling in SSLv2 client code: the SERVER FINISHED 1299 code (06) was taken as the first octet of the session ID and the last 1300 octet was ignored consequently. As a result SSLv2 client side session 1301 caching could not have worked due to the session ID mismatch between 1302 client and server. 1303 Behaviour observed by Crispin Flowerday <crispin@flowerday.cx> as 1304 PR #377. 1305 [Lutz Jaenicke] 1306 1307 *) Change the declaration of needed Kerberos libraries to use EX_LIBS 1308 instead of the special (and badly supported) LIBKRB5. LIBKRB5 is 1309 removed entirely. 1310 [Richard Levitte] 1311 1312 *) The hw_ncipher.c engine requires dynamic locks. Unfortunately, it 1313 seems that in spite of existing for more than a year, many application 1314 author have done nothing to provide the necessary callbacks, which 1315 means that this particular engine will not work properly anywhere. 1316 This is a very unfortunate situation which forces us, in the name 1317 of usability, to give the hw_ncipher.c a static lock, which is part 1318 of libcrypto. 1319 NOTE: This is for the 0.9.7 series ONLY. This hack will never 1320 appear in 0.9.8 or later. We EXPECT application authors to have 1321 dealt properly with this when 0.9.8 is released (unless we actually 1322 make such changes in the libcrypto locking code that changes will 1323 have to be made anyway). 1324 [Richard Levitte] 1325 1326 *) In asn1_d2i_read_bio() repeatedly call BIO_read() until all content 1327 octets have been read, EOF or an error occurs. Without this change 1328 some truncated ASN1 structures will not produce an error. 1329 [Steve Henson] 1330 1331 *) Disable Heimdal support, since it hasn't been fully implemented. 1332 Still give the possibility to force the use of Heimdal, but with 1333 warnings and a request that patches get sent to openssl-dev. 1334 [Richard Levitte] 1335 1336 *) Add the VC-CE target, introduce the WINCE sysname, and add 1337 INSTALL.WCE and appropriate conditionals to make it build. 1338 [Steven Reddie <smr@essemer.com.au> via Richard Levitte] 1339 1340 *) Change the DLL names for Cygwin to cygcrypto-x.y.z.dll and 1341 cygssl-x.y.z.dll, where x, y and z are the major, minor and 1342 edit numbers of the version. 1343 [Corinna Vinschen <vinschen@redhat.com> and Richard Levitte] 1344 1345 *) Introduce safe string copy and catenation functions 1346 (BUF_strlcpy() and BUF_strlcat()). 1347 [Ben Laurie (CHATS) and Richard Levitte] 1348 1349 *) Avoid using fixed-size buffers for one-line DNs. 1350 [Ben Laurie (CHATS)] 1351 1352 *) Add BUF_MEM_grow_clean() to avoid information leakage when 1353 resizing buffers containing secrets, and use where appropriate. 1354 [Ben Laurie (CHATS)] 1355 1356 *) Avoid using fixed size buffers for configuration file location. 1357 [Ben Laurie (CHATS)] 1358 1359 *) Avoid filename truncation for various CA files. 1360 [Ben Laurie (CHATS)] 1361 1362 *) Use sizeof in preference to magic numbers. 1363 [Ben Laurie (CHATS)] 1364 1365 *) Avoid filename truncation in cert requests. 1366 [Ben Laurie (CHATS)] 1367 1368 *) Add assertions to check for (supposedly impossible) buffer 1369 overflows. 1370 [Ben Laurie (CHATS)] 1371 1372 *) Don't cache truncated DNS entries in the local cache (this could 1373 potentially lead to a spoofing attack). 1374 [Ben Laurie (CHATS)] 1375 1376 *) Fix various buffers to be large enough for hex/decimal 1377 representations in a platform independent manner. 1378 [Ben Laurie (CHATS)] 1379 1380 *) Add CRYPTO_realloc_clean() to avoid information leakage when 1381 resizing buffers containing secrets, and use where appropriate. 1382 [Ben Laurie (CHATS)] 1383 1384 *) Add BIO_indent() to avoid much slightly worrying code to do 1385 indents. 1386 [Ben Laurie (CHATS)] 1387 1388 *) Convert sprintf()/BIO_puts() to BIO_printf(). 1389 [Ben Laurie (CHATS)] 1390 1391 *) buffer_gets() could terminate with the buffer only half 1392 full. Fixed. 1393 [Ben Laurie (CHATS)] 1394 1395 *) Add assertions to prevent user-supplied crypto functions from 1396 overflowing internal buffers by having large block sizes, etc. 1397 [Ben Laurie (CHATS)] 1398 1399 *) New OPENSSL_assert() macro (similar to assert(), but enabled 1400 unconditionally). 1401 [Ben Laurie (CHATS)] 1402 1403 *) Eliminate unused copy of key in RC4. 1404 [Ben Laurie (CHATS)] 1405 1406 *) Eliminate unused and incorrectly sized buffers for IV in pem.h. 1407 [Ben Laurie (CHATS)] 1408 1409 *) Fix off-by-one error in EGD path. 1410 [Ben Laurie (CHATS)] 1411 1412 *) If RANDFILE path is too long, ignore instead of truncating. 1413 [Ben Laurie (CHATS)] 1414 1415 *) Eliminate unused and incorrectly sized X.509 structure 1416 CBCParameter. 1417 [Ben Laurie (CHATS)] 1418 1419 *) Eliminate unused and dangerous function knumber(). 1420 [Ben Laurie (CHATS)] 1421 1422 *) Eliminate unused and dangerous structure, KSSL_ERR. 1423 [Ben Laurie (CHATS)] 1424 1425 *) Protect against overlong session ID context length in an encoded 1426 session object. Since these are local, this does not appear to be 1427 exploitable. 1428 [Ben Laurie (CHATS)] 1429 1430 *) Change from security patch (see 0.9.6e below) that did not affect 1431 the 0.9.6 release series: 1432 1433 Remote buffer overflow in SSL3 protocol - an attacker could 1434 supply an oversized master key in Kerberos-enabled versions. 1435 (CVE-2002-0657) 1436 [Ben Laurie (CHATS)] 1437 1438 *) Change the SSL kerb5 codes to match RFC 2712. 1439 [Richard Levitte] 1440 1441 *) Make -nameopt work fully for req and add -reqopt switch. 1442 [Michael Bell <michael.bell@rz.hu-berlin.de>, Steve Henson] 1443 1444 *) The "block size" for block ciphers in CFB and OFB mode should be 1. 1445 [Steve Henson, reported by Yngve Nysaeter Pettersen <yngve@opera.com>] 1446 1447 *) Make sure tests can be performed even if the corresponding algorithms 1448 have been removed entirely. This was also the last step to make 1449 OpenSSL compilable with DJGPP under all reasonable conditions. 1450 [Richard Levitte, Doug Kaufman <dkaufman@rahul.net>] 1451 1452 *) Add cipher selection rules COMPLEMENTOFALL and COMPLEMENTOFDEFAULT 1453 to allow version independent disabling of normally unselected ciphers, 1454 which may be activated as a side-effect of selecting a single cipher. 1455 1456 (E.g., cipher list string "RSA" enables ciphersuites that are left 1457 out of "ALL" because they do not provide symmetric encryption. 1458 "RSA:!COMPLEMEMENTOFALL" avoids these unsafe ciphersuites.) 1459 [Lutz Jaenicke, Bodo Moeller] 1460 1461 *) Add appropriate support for separate platform-dependent build 1462 directories. The recommended way to make a platform-dependent 1463 build directory is the following (tested on Linux), maybe with 1464 some local tweaks: 1465 1466 # Place yourself outside of the OpenSSL source tree. In 1467 # this example, the environment variable OPENSSL_SOURCE 1468 # is assumed to contain the absolute OpenSSL source directory. 1469 mkdir -p objtree/"`uname -s`-`uname -r`-`uname -m`" 1470 cd objtree/"`uname -s`-`uname -r`-`uname -m`" 1471 (cd $OPENSSL_SOURCE; find . -type f) | while read F; do 1472 mkdir -p `dirname $F` 1473 ln -s $OPENSSL_SOURCE/$F $F 1474 done 1475 1476 To be absolutely sure not to disturb the source tree, a "make clean" 1477 is a good thing. If it isn't successfull, don't worry about it, 1478 it probably means the source directory is very clean. 1479 [Richard Levitte] 1480 1481 *) Make sure any ENGINE control commands make local copies of string 1482 pointers passed to them whenever necessary. Otherwise it is possible 1483 the caller may have overwritten (or deallocated) the original string 1484 data when a later ENGINE operation tries to use the stored values. 1485 [G�tz Babin-Ebell <babinebell@trustcenter.de>] 1486 1487 *) Improve diagnostics in file reading and command-line digests. 1488 [Ben Laurie aided and abetted by Solar Designer <solar@openwall.com>] 1489 1490 *) Add AES modes CFB and OFB to the object database. Correct an 1491 error in AES-CFB decryption. 1492 [Richard Levitte] 1493 1494 *) Remove most calls to EVP_CIPHER_CTX_cleanup() in evp_enc.c, this 1495 allows existing EVP_CIPHER_CTX structures to be reused after 1496 calling EVP_*Final(). This behaviour is used by encryption 1497 BIOs and some applications. This has the side effect that 1498 applications must explicitly clean up cipher contexts with 1499 EVP_CIPHER_CTX_cleanup() or they will leak memory. 1500 [Steve Henson] 1501 1502 *) Check the values of dna and dnb in bn_mul_recursive before calling 1503 bn_mul_comba (a non zero value means the a or b arrays do not contain 1504 n2 elements) and fallback to bn_mul_normal if either is not zero. 1505 [Steve Henson] 1506 1507 *) Fix escaping of non-ASCII characters when using the -subj option 1508 of the "openssl req" command line tool. (Robert Joop <joop@fokus.gmd.de>) 1509 [Lutz Jaenicke] 1510 1511 *) Make object definitions compliant to LDAP (RFC2256): SN is the short 1512 form for "surname", serialNumber has no short form. 1513 Use "mail" as the short name for "rfc822Mailbox" according to RFC2798; 1514 therefore remove "mail" short name for "internet 7". 1515 The OID for unique identifiers in X509 certificates is 1516 x500UniqueIdentifier, not uniqueIdentifier. 1517 Some more OID additions. (Michael Bell <michael.bell@rz.hu-berlin.de>) 1518 [Lutz Jaenicke] 1519 1520 *) Add an "init" command to the ENGINE config module and auto initialize 1521 ENGINEs. Without any "init" command the ENGINE will be initialized 1522 after all ctrl commands have been executed on it. If init=1 the 1523 ENGINE is initailized at that point (ctrls before that point are run 1524 on the uninitialized ENGINE and after on the initialized one). If 1525 init=0 then the ENGINE will not be iniatialized at all. 1526 [Steve Henson] 1527 1528 *) Fix the 'app_verify_callback' interface so that the user-defined 1529 argument is actually passed to the callback: In the 1530 SSL_CTX_set_cert_verify_callback() prototype, the callback 1531 declaration has been changed from 1532 int (*cb)() 1533 into 1534 int (*cb)(X509_STORE_CTX *,void *); 1535 in ssl_verify_cert_chain (ssl/ssl_cert.c), the call 1536 i=s->ctx->app_verify_callback(&ctx) 1537 has been changed into 1538 i=s->ctx->app_verify_callback(&ctx, s->ctx->app_verify_arg). 1539 1540 To update applications using SSL_CTX_set_cert_verify_callback(), 1541 a dummy argument can be added to their callback functions. 1542 [D. K. Smetters <smetters@parc.xerox.com>] 1543 1544 *) Added the '4758cca' ENGINE to support IBM 4758 cards. 1545 [Maurice Gittens <maurice@gittens.nl>, touchups by Geoff Thorpe] 1546 1547 *) Add and OPENSSL_LOAD_CONF define which will cause 1548 OpenSSL_add_all_algorithms() to load the openssl.cnf config file. 1549 This allows older applications to transparently support certain 1550 OpenSSL features: such as crypto acceleration and dynamic ENGINE loading. 1551 Two new functions OPENSSL_add_all_algorithms_noconf() which will never 1552 load the config file and OPENSSL_add_all_algorithms_conf() which will 1553 always load it have also been added. 1554 [Steve Henson] 1555 1556 *) Add the OFB, CFB and CTR (all with 128 bit feedback) to AES. 1557 Adjust NIDs and EVP layer. 1558 [Stephen Sprunk <stephen@sprunk.org> and Richard Levitte] 1559 1560 *) Config modules support in openssl utility. 1561 1562 Most commands now load modules from the config file, 1563 though in a few (such as version) this isn't done 1564 because it couldn't be used for anything. 1565 1566 In the case of ca and req the config file used is 1567 the same as the utility itself: that is the -config 1568 command line option can be used to specify an 1569 alternative file. 1570 [Steve Henson] 1571 1572 *) Move default behaviour from OPENSSL_config(). If appname is NULL 1573 use "openssl_conf" if filename is NULL use default openssl config file. 1574 [Steve Henson] 1575 1576 *) Add an argument to OPENSSL_config() to allow the use of an alternative 1577 config section name. Add a new flag to tolerate a missing config file 1578 and move code to CONF_modules_load_file(). 1579 [Steve Henson] 1580 1581 *) Support for crypto accelerator cards from Accelerated Encryption 1582 Processing, www.aep.ie. (Use engine 'aep') 1583 The support was copied from 0.9.6c [engine] and adapted/corrected 1584 to work with the new engine framework. 1585 [AEP Inc. and Richard Levitte] 1586 1587 *) Support for SureWare crypto accelerator cards from Baltimore 1588 Technologies. (Use engine 'sureware') 1589 The support was copied from 0.9.6c [engine] and adapted 1590 to work with the new engine framework. 1591 [Richard Levitte] 1592 1593 *) Have the CHIL engine fork-safe (as defined by nCipher) and actually 1594 make the newer ENGINE framework commands for the CHIL engine work. 1595 [Toomas Kiisk <vix@cyber.ee> and Richard Levitte] 1596 1597 *) Make it possible to produce shared libraries on ReliantUNIX. 1598 [Robert Dahlem <Robert.Dahlem@ffm2.siemens.de> via Richard Levitte] 1599 1600 *) Add the configuration target debug-linux-ppro. 1601 Make 'openssl rsa' use the general key loading routines 1602 implemented in apps.c, and make those routines able to 1603 handle the key format FORMAT_NETSCAPE and the variant 1604 FORMAT_IISSGC. 1605 [Toomas Kiisk <vix@cyber.ee> via Richard Levitte] 1606 1607 *) Fix a crashbug and a logic bug in hwcrhk_load_pubkey(). 1608 [Toomas Kiisk <vix@cyber.ee> via Richard Levitte] 1609 1610 *) Add -keyform to rsautl, and document -engine. 1611 [Richard Levitte, inspired by Toomas Kiisk <vix@cyber.ee>] 1612 1613 *) Change BIO_new_file (crypto/bio/bss_file.c) to use new 1614 BIO_R_NO_SUCH_FILE error code rather than the generic 1615 ERR_R_SYS_LIB error code if fopen() fails with ENOENT. 1616 [Ben Laurie] 1617 1618 *) Add new functions 1619 ERR_peek_last_error 1620 ERR_peek_last_error_line 1621 ERR_peek_last_error_line_data. 1622 These are similar to 1623 ERR_peek_error 1624 ERR_peek_error_line 1625 ERR_peek_error_line_data, 1626 but report on the latest error recorded rather than the first one 1627 still in the error queue. 1628 [Ben Laurie, Bodo Moeller] 1629 1630 *) default_algorithms option in ENGINE config module. This allows things 1631 like: 1632 default_algorithms = ALL 1633 default_algorithms = RSA, DSA, RAND, CIPHERS, DIGESTS 1634 [Steve Henson] 1635 1636 *) Prelminary ENGINE config module. 1637 [Steve Henson] 1638 1639 *) New experimental application configuration code. 1640 [Steve Henson] 1641 1642 *) Change the AES code to follow the same name structure as all other 1643 symmetric ciphers, and behave the same way. Move everything to 1644 the directory crypto/aes, thereby obsoleting crypto/rijndael. 1645 [Stephen Sprunk <stephen@sprunk.org> and Richard Levitte] 1646 1647 *) SECURITY: remove unsafe setjmp/signal interaction from ui_openssl.c. 1648 [Ben Laurie and Theo de Raadt] 1649 1650 *) Add option to output public keys in req command. 1651 [Massimiliano Pala madwolf@openca.org] 1652 1653 *) Use wNAFs in EC_POINTs_mul() for improved efficiency 1654 (up to about 10% better than before for P-192 and P-224). 1655 [Bodo Moeller] 1656 1657 *) New functions/macros 1658 1659 SSL_CTX_set_msg_callback(ctx, cb) 1660 SSL_CTX_set_msg_callback_arg(ctx, arg) 1661 SSL_set_msg_callback(ssl, cb) 1662 SSL_set_msg_callback_arg(ssl, arg) 1663 1664 to request calling a callback function 1665 1666 void cb(int write_p, int version, int content_type, 1667 const void *buf, size_t len, SSL *ssl, void *arg) 1668 1669 whenever a protocol message has been completely received 1670 (write_p == 0) or sent (write_p == 1). Here 'version' is the 1671 protocol version according to which the SSL library interprets 1672 the current protocol message (SSL2_VERSION, SSL3_VERSION, or 1673 TLS1_VERSION). 'content_type' is 0 in the case of SSL 2.0, or 1674 the content type as defined in the SSL 3.0/TLS 1.0 protocol 1675 specification (change_cipher_spec(20), alert(21), handshake(22)). 1676 'buf' and 'len' point to the actual message, 'ssl' to the 1677 SSL object, and 'arg' is the application-defined value set by 1678 SSL[_CTX]_set_msg_callback_arg(). 1679 1680 'openssl s_client' and 'openssl s_server' have new '-msg' options 1681 to enable a callback that displays all protocol messages. 1682 [Bodo Moeller] 1683 1684 *) Change the shared library support so shared libraries are built as 1685 soon as the corresponding static library is finished, and thereby get 1686 openssl and the test programs linked against the shared library. 1687 This still only happens when the keyword "shard" has been given to 1688 the configuration scripts. 1689 1690 NOTE: shared library support is still an experimental thing, and 1691 backward binary compatibility is still not guaranteed. 1692 ["Maciej W. Rozycki" <macro@ds2.pg.gda.pl> and Richard Levitte] 1693 1694 *) Add support for Subject Information Access extension. 1695 [Peter Sylvester <Peter.Sylvester@EdelWeb.fr>] 1696 1697 *) Make BUF_MEM_grow() behaviour more consistent: Initialise to zero 1698 additional bytes when new memory had to be allocated, not just 1699 when reusing an existing buffer. 1700 [Bodo Moeller] 1701 1702 *) New command line and configuration option 'utf8' for the req command. 1703 This allows field values to be specified as UTF8 strings. 1704 [Steve Henson] 1705 1706 *) Add -multi and -mr options to "openssl speed" - giving multiple parallel 1707 runs for the former and machine-readable output for the latter. 1708 [Ben Laurie] 1709 1710 *) Add '-noemailDN' option to 'openssl ca'. This prevents inclusion 1711 of the e-mail address in the DN (i.e., it will go into a certificate 1712 extension only). The new configuration file option 'email_in_dn = no' 1713 has the same effect. 1714 [Massimiliano Pala madwolf@openca.org] 1715 1716 *) Change all functions with names starting with des_ to be starting 1717 with DES_ instead. Add wrappers that are compatible with libdes, 1718 but are named _ossl_old_des_*. Finally, add macros that map the 1719 des_* symbols to the corresponding _ossl_old_des_* if libdes 1720 compatibility is desired. If OpenSSL 0.9.6c compatibility is 1721 desired, the des_* symbols will be mapped to DES_*, with one 1722 exception. 1723 1724 Since we provide two compatibility mappings, the user needs to 1725 define the macro OPENSSL_DES_LIBDES_COMPATIBILITY if libdes 1726 compatibility is desired. The default (i.e., when that macro 1727 isn't defined) is OpenSSL 0.9.6c compatibility. 1728 1729 There are also macros that enable and disable the support of old 1730 des functions altogether. Those are OPENSSL_ENABLE_OLD_DES_SUPPORT 1731 and OPENSSL_DISABLE_OLD_DES_SUPPORT. If none or both of those 1732 are defined, the default will apply: to support the old des routines. 1733 1734 In either case, one must include openssl/des.h to get the correct 1735 definitions. Do not try to just include openssl/des_old.h, that 1736 won't work. 1737 1738 NOTE: This is a major break of an old API into a new one. Software 1739 authors are encouraged to switch to the DES_ style functions. Some 1740 time in the future, des_old.h and the libdes compatibility functions 1741 will be disable (i.e. OPENSSL_DISABLE_OLD_DES_SUPPORT will be the 1742 default), and then completely removed. 1743 [Richard Levitte] 1744 1745 *) Test for certificates which contain unsupported critical extensions. 1746 If such a certificate is found during a verify operation it is 1747 rejected by default: this behaviour can be overridden by either 1748 handling the new error X509_V_ERR_UNHANDLED_CRITICAL_EXTENSION or 1749 by setting the verify flag X509_V_FLAG_IGNORE_CRITICAL. A new function 1750 X509_supported_extension() has also been added which returns 1 if a 1751 particular extension is supported. 1752 [Steve Henson] 1753 1754 *) Modify the behaviour of EVP cipher functions in similar way to digests 1755 to retain compatibility with existing code. 1756 [Steve Henson] 1757 1758 *) Modify the behaviour of EVP_DigestInit() and EVP_DigestFinal() to retain 1759 compatibility with existing code. In particular the 'ctx' parameter does 1760 not have to be to be initialized before the call to EVP_DigestInit() and 1761 it is tidied up after a call to EVP_DigestFinal(). New function 1762 EVP_DigestFinal_ex() which does not tidy up the ctx. Similarly function 1763 EVP_MD_CTX_copy() changed to not require the destination to be 1764 initialized valid and new function EVP_MD_CTX_copy_ex() added which 1765 requires the destination to be valid. 1766 1767 Modify all the OpenSSL digest calls to use EVP_DigestInit_ex(), 1768 EVP_DigestFinal_ex() and EVP_MD_CTX_copy_ex(). 1769 [Steve Henson] 1770 1771 *) Change ssl3_get_message (ssl/s3_both.c) and the functions using it 1772 so that complete 'Handshake' protocol structures are kept in memory 1773 instead of overwriting 'msg_type' and 'length' with 'body' data. 1774 [Bodo Moeller] 1775 1776 *) Add an implementation of SSL_add_dir_cert_subjects_to_stack for Win32. 1777 [Massimo Santin via Richard Levitte] 1778 1779 *) Major restructuring to the underlying ENGINE code. This includes 1780 reduction of linker bloat, separation of pure "ENGINE" manipulation 1781 (initialisation, etc) from functionality dealing with implementations 1782 of specific crypto iterfaces. This change also introduces integrated 1783 support for symmetric ciphers and digest implementations - so ENGINEs 1784 can now accelerate these by providing EVP_CIPHER and EVP_MD 1785 implementations of their own. This is detailed in crypto/engine/README 1786 as it couldn't be adequately described here. However, there are a few 1787 API changes worth noting - some RSA, DSA, DH, and RAND functions that 1788 were changed in the original introduction of ENGINE code have now 1789 reverted back - the hooking from this code to ENGINE is now a good 1790 deal more passive and at run-time, operations deal directly with 1791 RSA_METHODs, DSA_METHODs (etc) as they did before, rather than 1792 dereferencing through an ENGINE pointer any more. Also, the ENGINE 1793 functions dealing with BN_MOD_EXP[_CRT] handlers have been removed - 1794 they were not being used by the framework as there is no concept of a 1795 BIGNUM_METHOD and they could not be generalised to the new 1796 'ENGINE_TABLE' mechanism that underlies the new code. Similarly, 1797 ENGINE_cpy() has been removed as it cannot be consistently defined in 1798 the new code. 1799 [Geoff Thorpe] 1800 1801 *) Change ASN1_GENERALIZEDTIME_check() to allow fractional seconds. 1802 [Steve Henson] 1803 1804 *) Change mkdef.pl to sort symbols that get the same entry number, 1805 and make sure the automatically generated functions ERR_load_* 1806 become part of libeay.num as well. 1807 [Richard Levitte] 1808 1809 *) New function SSL_renegotiate_pending(). This returns true once 1810 renegotiation has been requested (either SSL_renegotiate() call 1811 or HelloRequest/ClientHello receveived from the peer) and becomes 1812 false once a handshake has been completed. 1813 (For servers, SSL_renegotiate() followed by SSL_do_handshake() 1814 sends a HelloRequest, but does not ensure that a handshake takes 1815 place. SSL_renegotiate_pending() is useful for checking if the 1816 client has followed the request.) 1817 [Bodo Moeller] 1818 1819 *) New SSL option SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION. 1820 By default, clients may request session resumption even during 1821 renegotiation (if session ID contexts permit); with this option, 1822 session resumption is possible only in the first handshake. 1823 1824 SSL_OP_ALL is now 0x00000FFFL instead of 0x000FFFFFL. This makes 1825 more bits available for options that should not be part of 1826 SSL_OP_ALL (such as SSL_OP_NO_SESSION_RESUMPTION_ON_RENEGOTIATION). 1827 [Bodo Moeller] 1828 1829 *) Add some demos for certificate and certificate request creation. 1830 [Steve Henson] 1831 1832 *) Make maximum certificate chain size accepted from the peer application 1833 settable (SSL*_get/set_max_cert_list()), as proposed by 1834 "Douglas E. Engert" <deengert@anl.gov>. 1835 [Lutz Jaenicke] 1836 1837 *) Add support for shared libraries for Unixware-7 1838 (Boyd Lynn Gerber <gerberb@zenez.com>). 1839 [Lutz Jaenicke] 1840 1841 *) Add a "destroy" handler to ENGINEs that allows structural cleanup to 1842 be done prior to destruction. Use this to unload error strings from 1843 ENGINEs that load their own error strings. NB: This adds two new API 1844 functions to "get" and "set" this destroy handler in an ENGINE. 1845 [Geoff Thorpe] 1846 1847 *) Alter all existing ENGINE implementations (except "openssl" and 1848 "openbsd") to dynamically instantiate their own error strings. This 1849 makes them more flexible to be built both as statically-linked ENGINEs 1850 and self-contained shared-libraries loadable via the "dynamic" ENGINE. 1851 Also, add stub code to each that makes building them as self-contained 1852 shared-libraries easier (see README.ENGINE). 1853 [Geoff Thorpe] 1854 1855 *) Add a "dynamic" ENGINE that provides a mechanism for binding ENGINE 1856 implementations into applications that are completely implemented in 1857 self-contained shared-libraries. The "dynamic" ENGINE exposes control 1858 commands that can be used to configure what shared-library to load and 1859 to control aspects of the way it is handled. Also, made an update to 1860 the README.ENGINE file that brings its information up-to-date and 1861 provides some information and instructions on the "dynamic" ENGINE 1862 (ie. how to use it, how to build "dynamic"-loadable ENGINEs, etc). 1863 [Geoff Thorpe] 1864 1865 *) Make it possible to unload ranges of ERR strings with a new 1866 "ERR_unload_strings" function. 1867 [Geoff Thorpe] 1868 1869 *) Add a copy() function to EVP_MD. 1870 [Ben Laurie] 1871 1872 *) Make EVP_MD routines take a context pointer instead of just the 1873 md_data void pointer. 1874 [Ben Laurie] 1875 1876 *) Add flags to EVP_MD and EVP_MD_CTX. EVP_MD_FLAG_ONESHOT indicates 1877 that the digest can only process a single chunk of data 1878 (typically because it is provided by a piece of 1879 hardware). EVP_MD_CTX_FLAG_ONESHOT indicates that the application 1880 is only going to provide a single chunk of data, and hence the 1881 framework needn't accumulate the data for oneshot drivers. 1882 [Ben Laurie] 1883 1884 *) As with "ERR", make it possible to replace the underlying "ex_data" 1885 functions. This change also alters the storage and management of global 1886 ex_data state - it's now all inside ex_data.c and all "class" code (eg. 1887 RSA, BIO, SSL_CTX, etc) no longer stores its own STACKS and per-class 1888 index counters. The API functions that use this state have been changed 1889 to take a "class_index" rather than pointers to the class's local STACK 1890 and counter, and there is now an API function to dynamically create new 1891 classes. This centralisation allows us to (a) plug a lot of the 1892 thread-safety problems that existed, and (b) makes it possible to clean 1893 up all allocated state using "CRYPTO_cleanup_all_ex_data()". W.r.t. (b) 1894 such data would previously have always leaked in application code and 1895 workarounds were in place to make the memory debugging turn a blind eye 1896 to it. Application code that doesn't use this new function will still 1897 leak as before, but their memory debugging output will announce it now 1898 rather than letting it slide. 1899 1900 Besides the addition of CRYPTO_cleanup_all_ex_data(), another API change 1901 induced by the "ex_data" overhaul is that X509_STORE_CTX_init() now 1902 has a return value to indicate success or failure. 1903 [Geoff Thorpe] 1904 1905 *) Make it possible to replace the underlying "ERR" functions such that the 1906 global state (2 LHASH tables and 2 locks) is only used by the "default" 1907 implementation. This change also adds two functions to "get" and "set" 1908 the implementation prior to it being automatically set the first time 1909 any other ERR function takes place. Ie. an application can call "get", 1910 pass the return value to a module it has just loaded, and that module 1911 can call its own "set" function using that value. This means the 1912 module's "ERR" operations will use (and modify) the error state in the 1913 application and not in its own statically linked copy of OpenSSL code. 1914 [Geoff Thorpe] 1915 1916 *) Give DH, DSA, and RSA types their own "**_up_ref()" function to increment 1917 reference counts. This performs normal REF_PRINT/REF_CHECK macros on 1918 the operation, and provides a more encapsulated way for external code 1919 (crypto/evp/ and ssl/) to do this. Also changed the evp and ssl code 1920 to use these functions rather than manually incrementing the counts. 1921 1922 Also rename "DSO_up()" function to more descriptive "DSO_up_ref()". 1923 [Geoff Thorpe] 1924 1925 *) Add EVP test program. 1926 [Ben Laurie] 1927 1928 *) Add symmetric cipher support to ENGINE. Expect the API to change! 1929 [Ben Laurie] 1930 1931 *) New CRL functions: X509_CRL_set_version(), X509_CRL_set_issuer_name() 1932 X509_CRL_set_lastUpdate(), X509_CRL_set_nextUpdate(), X509_CRL_sort(), 1933 X509_REVOKED_set_serialNumber(), and X509_REVOKED_set_revocationDate(). 1934 These allow a CRL to be built without having to access X509_CRL fields 1935 directly. Modify 'ca' application to use new functions. 1936 [Steve Henson] 1937 1938 *) Move SSL_OP_TLS_ROLLBACK_BUG out of the SSL_OP_ALL list of recommended 1939 bug workarounds. Rollback attack detection is a security feature. 1940 The problem will only arise on OpenSSL servers when TLSv1 is not 1941 available (sslv3_server_method() or SSL_OP_NO_TLSv1). 1942 Software authors not wanting to support TLSv1 will have special reasons 1943 for their choice and can explicitly enable this option. 1944 [Bodo Moeller, Lutz Jaenicke] 1945 1946 *) Rationalise EVP so it can be extended: don't include a union of 1947 cipher/digest structures, add init/cleanup functions for EVP_MD_CTX 1948 (similar to those existing for EVP_CIPHER_CTX). 1949 Usage example: 1950 1951 EVP_MD_CTX md; 1952 1953 EVP_MD_CTX_init(&md); /* new function call */ 1954 EVP_DigestInit(&md, EVP_sha1()); 1955 EVP_DigestUpdate(&md, in, len); 1956 EVP_DigestFinal(&md, out, NULL); 1957 EVP_MD_CTX_cleanup(&md); /* new function call */ 1958 1959 [Ben Laurie] 1960 1961 *) Make DES key schedule conform to the usual scheme, as well as 1962 correcting its structure. This means that calls to DES functions 1963 now have to pass a pointer to a des_key_schedule instead of a 1964 plain des_key_schedule (which was actually always a pointer 1965 anyway): E.g., 1966 1967 des_key_schedule ks; 1968 1969 des_set_key_checked(..., &ks); 1970 des_ncbc_encrypt(..., &ks, ...); 1971 1972 (Note that a later change renames 'des_...' into 'DES_...'.) 1973 [Ben Laurie] 1974 1975 *) Initial reduction of linker bloat: the use of some functions, such as 1976 PEM causes large amounts of unused functions to be linked in due to 1977 poor organisation. For example pem_all.c contains every PEM function 1978 which has a knock on effect of linking in large amounts of (unused) 1979 ASN1 code. Grouping together similar functions and splitting unrelated 1980 functions prevents this. 1981 [Steve Henson] 1982 1983 *) Cleanup of EVP macros. 1984 [Ben Laurie] 1985 1986 *) Change historical references to {NID,SN,LN}_des_ede and ede3 to add the 1987 correct _ecb suffix. 1988 [Ben Laurie] 1989 1990 *) Add initial OCSP responder support to ocsp application. The 1991 revocation information is handled using the text based index 1992 use by the ca application. The responder can either handle 1993 requests generated internally, supplied in files (for example 1994 via a CGI script) or using an internal minimal server. 1995 [Steve Henson] 1996 1997 *) Add configuration choices to get zlib compression for TLS. 1998 [Richard Levitte] 1999 2000 *) Changes to Kerberos SSL for RFC 2712 compliance: 2001 1. Implemented real KerberosWrapper, instead of just using 2002 KRB5 AP_REQ message. [Thanks to Simon Wilkinson <sxw@sxw.org.uk>] 2003 2. Implemented optional authenticator field of KerberosWrapper. 2004 2005 Added openssl-style ASN.1 macros for Kerberos ticket, ap_req, 2006 and authenticator structs; see crypto/krb5/. 2007 2008 Generalized Kerberos calls to support multiple Kerberos libraries. 2009 [Vern Staats <staatsvr@asc.hpc.mil>, 2010 Jeffrey Altman <jaltman@columbia.edu> 2011 via Richard Levitte] 2012 2013 *) Cause 'openssl speed' to use fully hard-coded DSA keys as it 2014 already does with RSA. testdsa.h now has 'priv_key/pub_key' 2015 values for each of the key sizes rather than having just 2016 parameters (and 'speed' generating keys each time). 2017 [Geoff Thorpe] 2018 2019 *) Speed up EVP routines. 2020 Before: 2021encrypt 2022type 8 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes 2023des-cbc 4408.85k 5560.51k 5778.46k 5862.20k 5825.16k 2024des-cbc 4389.55k 5571.17k 5792.23k 5846.91k 5832.11k 2025des-cbc 4394.32k 5575.92k 5807.44k 5848.37k 5841.30k 2026decrypt 2027des-cbc 3482.66k 5069.49k 5496.39k 5614.16k 5639.28k 2028des-cbc 3480.74k 5068.76k 5510.34k 5609.87k 5635.52k 2029des-cbc 3483.72k 5067.62k 5504.60k 5708.01k 5724.80k 2030 After: 2031encrypt 2032des-cbc 4660.16k 5650.19k 5807.19k 5827.13k 5783.32k 2033decrypt 2034des-cbc 3624.96k 5258.21k 5530.91k 5624.30k 5628.26k 2035 [Ben Laurie] 2036 2037 *) Added the OS2-EMX target. 2038 ["Brian Havard" <brianh@kheldar.apana.org.au> and Richard Levitte] 2039 2040 *) Rewrite apps to use NCONF routines instead of the old CONF. New functions 2041 to support NCONF routines in extension code. New function CONF_set_nconf() 2042 to allow functions which take an NCONF to also handle the old LHASH 2043 structure: this means that the old CONF compatible routines can be 2044 retained (in particular wrt extensions) without having to duplicate the 2045 code. New function X509V3_add_ext_nconf_sk to add extensions to a stack. 2046 [Steve Henson] 2047 2048 *) Enhance the general user interface with mechanisms for inner control 2049 and with possibilities to have yes/no kind of prompts. 2050 [Richard Levitte] 2051 2052 *) Change all calls to low level digest routines in the library and 2053 applications to use EVP. Add missing calls to HMAC_cleanup() and 2054 don't assume HMAC_CTX can be copied using memcpy(). 2055 [Verdon Walker <VWalker@novell.com>, Steve Henson] 2056 2057 *) Add the possibility to control engines through control names but with 2058 arbitrary arguments instead of just a string. 2059 Change the key loaders to take a UI_METHOD instead of a callback 2060 function pointer. NOTE: this breaks binary compatibility with earlier 2061 versions of OpenSSL [engine]. 2062 Adapt the nCipher code for these new conditions and add a card insertion 2063 callback. 2064 [Richard Levitte] 2065 2066 *) Enhance the general user interface with mechanisms to better support 2067 dialog box interfaces, application-defined prompts, the possibility 2068 to use defaults (for example default passwords from somewhere else) 2069 and interrupts/cancellations. 2070 [Richard Levitte] 2071 2072 *) Tidy up PKCS#12 attribute handling. Add support for the CSP name 2073 attribute in PKCS#12 files, add new -CSP option to pkcs12 utility. 2074 [Steve Henson] 2075 2076 *) Fix a memory leak in 'sk_dup()' in the case reallocation fails. (Also 2077 tidy up some unnecessarily weird code in 'sk_new()'). 2078 [Geoff, reported by Diego Tartara <dtartara@novamens.com>] 2079 2080 *) Change the key loading routines for ENGINEs to use the same kind 2081 callback (pem_password_cb) as all other routines that need this 2082 kind of callback. 2083 [Richard Levitte] 2084 2085 *) Increase ENTROPY_NEEDED to 32 bytes, as Rijndael can operate with 2086 256 bit (=32 byte) keys. Of course seeding with more entropy bytes 2087 than this minimum value is recommended. 2088 [Lutz Jaenicke] 2089 2090 *) New random seeder for OpenVMS, using the system process statistics 2091 that are easily reachable. 2092 [Richard Levitte] 2093 2094 *) Windows apparently can't transparently handle global 2095 variables defined in DLLs. Initialisations such as: 2096 2097 const ASN1_ITEM *it = &ASN1_INTEGER_it; 2098 2099 wont compile. This is used by the any applications that need to 2100 declare their own ASN1 modules. This was fixed by adding the option 2101 EXPORT_VAR_AS_FN to all Win32 platforms, although this isn't strictly 2102 needed for static libraries under Win32. 2103 [Steve Henson] 2104 2105 *) New functions X509_PURPOSE_set() and X509_TRUST_set() to handle 2106 setting of purpose and trust fields. New X509_STORE trust and 2107 purpose functions and tidy up setting in other SSL functions. 2108 [Steve Henson] 2109 2110 *) Add copies of X509_STORE_CTX fields and callbacks to X509_STORE 2111 structure. These are inherited by X509_STORE_CTX when it is 2112 initialised. This allows various defaults to be set in the 2113 X509_STORE structure (such as flags for CRL checking and custom 2114 purpose or trust settings) for functions which only use X509_STORE_CTX 2115 internally such as S/MIME. 2116 2117 Modify X509_STORE_CTX_purpose_inherit() so it only sets purposes and 2118 trust settings if they are not set in X509_STORE. This allows X509_STORE 2119 purposes and trust (in S/MIME for example) to override any set by default. 2120 2121 Add command line options for CRL checking to smime, s_client and s_server 2122 applications. 2123 [Steve Henson] 2124 2125 *) Initial CRL based revocation checking. If the CRL checking flag(s) 2126 are set then the CRL is looked up in the X509_STORE structure and 2127 its validity and signature checked, then if the certificate is found 2128 in the CRL the verify fails with a revoked error. 2129 2130 Various new CRL related callbacks added to X509_STORE_CTX structure. 2131 2132 Command line options added to 'verify' application to support this. 2133 2134 This needs some additional work, such as being able to handle multiple 2135 CRLs with different times, extension based lookup (rather than just 2136 by subject name) and ultimately more complete V2 CRL extension 2137 handling. 2138 [Steve Henson] 2139 2140 *) Add a general user interface API (crypto/ui/). This is designed 2141 to replace things like des_read_password and friends (backward 2142 compatibility functions using this new API are provided). 2143 The purpose is to remove prompting functions from the DES code 2144 section as well as provide for prompting through dialog boxes in 2145 a window system and the like. 2146 [Richard Levitte] 2147 2148 *) Add "ex_data" support to ENGINE so implementations can add state at a 2149 per-structure level rather than having to store it globally. 2150 [Geoff] 2151 2152 *) Make it possible for ENGINE structures to be copied when retrieved by 2153 ENGINE_by_id() if the ENGINE specifies a new flag: ENGINE_FLAGS_BY_ID_COPY. 2154 This causes the "original" ENGINE structure to act like a template, 2155 analogous to the RSA vs. RSA_METHOD type of separation. Because of this 2156 operational state can be localised to each ENGINE structure, despite the 2157 fact they all share the same "methods". New ENGINE structures returned in 2158 this case have no functional references and the return value is the single 2159 structural reference. This matches the single structural reference returned 2160 by ENGINE_by_id() normally, when it is incremented on the pre-existing 2161 ENGINE structure. 2162 [Geoff] 2163 2164 *) Fix ASN1 decoder when decoding type ANY and V_ASN1_OTHER: since this 2165 needs to match any other type at all we need to manually clear the 2166 tag cache. 2167 [Steve Henson] 2168 2169 *) Changes to the "openssl engine" utility to include; 2170 - verbosity levels ('-v', '-vv', and '-vvv') that provide information 2171 about an ENGINE's available control commands. 2172 - executing control commands from command line arguments using the 2173 '-pre' and '-post' switches. '-post' is only used if '-t' is 2174 specified and the ENGINE is successfully initialised. The syntax for 2175 the individual commands are colon-separated, for example; 2176 openssl engine chil -pre FORK_CHECK:0 -pre SO_PATH:/lib/test.so 2177 [Geoff] 2178 2179 *) New dynamic control command support for ENGINEs. ENGINEs can now 2180 declare their own commands (numbers), names (strings), descriptions, 2181 and input types for run-time discovery by calling applications. A 2182 subset of these commands are implicitly classed as "executable" 2183 depending on their input type, and only these can be invoked through 2184 the new string-based API function ENGINE_ctrl_cmd_string(). (Eg. this 2185 can be based on user input, config files, etc). The distinction is 2186 that "executable" commands cannot return anything other than a boolean 2187 result and can only support numeric or string input, whereas some 2188 discoverable commands may only be for direct use through 2189 ENGINE_ctrl(), eg. supporting the exchange of binary data, function 2190 pointers, or other custom uses. The "executable" commands are to 2191 support parameterisations of ENGINE behaviour that can be 2192 unambiguously defined by ENGINEs and used consistently across any 2193 OpenSSL-based application. Commands have been added to all the 2194 existing hardware-supporting ENGINEs, noticeably "SO_PATH" to allow 2195 control over shared-library paths without source code alterations. 2196 [Geoff] 2197 2198 *) Changed all ENGINE implementations to dynamically allocate their 2199 ENGINEs rather than declaring them statically. Apart from this being 2200 necessary with the removal of the ENGINE_FLAGS_MALLOCED distinction, 2201 this also allows the implementations to compile without using the 2202 internal engine_int.h header. 2203 [Geoff] 2204 2205 *) Minor adjustment to "rand" code. RAND_get_rand_method() now returns a 2206 'const' value. Any code that should be able to modify a RAND_METHOD 2207 should already have non-const pointers to it (ie. they should only 2208 modify their own ones). 2209 [Geoff] 2210 2211 *) Made a variety of little tweaks to the ENGINE code. 2212 - "atalla" and "ubsec" string definitions were moved from header files 2213 to C code. "nuron" string definitions were placed in variables 2214 rather than hard-coded - allowing parameterisation of these values 2215 later on via ctrl() commands. 2216 - Removed unused "#if 0"'d code. 2217 - Fixed engine list iteration code so it uses ENGINE_free() to release 2218 structural references. 2219 - Constified the RAND_METHOD element of ENGINE structures. 2220 - Constified various get/set functions as appropriate and added 2221 missing functions (including a catch-all ENGINE_cpy that duplicates 2222 all ENGINE values onto a new ENGINE except reference counts/state). 2223 - Removed NULL parameter checks in get/set functions. Setting a method 2224 or function to NULL is a way of cancelling out a previously set 2225 value. Passing a NULL ENGINE parameter is just plain stupid anyway 2226 and doesn't justify the extra error symbols and code. 2227 - Deprecate the ENGINE_FLAGS_MALLOCED define and move the area for 2228 flags from engine_int.h to engine.h. 2229 - Changed prototypes for ENGINE handler functions (init(), finish(), 2230 ctrl(), key-load functions, etc) to take an (ENGINE*) parameter. 2231 [Geoff] 2232 2233 *) Implement binary inversion algorithm for BN_mod_inverse in addition 2234 to the algorithm using long division. The binary algorithm can be 2235 used only if the modulus is odd. On 32-bit systems, it is faster 2236 only for relatively small moduli (roughly 20-30% for 128-bit moduli, 2237 roughly 5-15% for 256-bit moduli), so we use it only for moduli 2238 up to 450 bits. In 64-bit environments, the binary algorithm 2239 appears to be advantageous for much longer moduli; here we use it 2240 for moduli up to 2048 bits. 2241 [Bodo Moeller] 2242 2243 *) Rewrite CHOICE field setting in ASN1_item_ex_d2i(). The old code 2244 could not support the combine flag in choice fields. 2245 [Steve Henson] 2246 2247 *) Add a 'copy_extensions' option to the 'ca' utility. This copies 2248 extensions from a certificate request to the certificate. 2249 [Steve Henson] 2250 2251 *) Allow multiple 'certopt' and 'nameopt' options to be separated 2252 by commas. Add 'namopt' and 'certopt' options to the 'ca' config 2253 file: this allows the display of the certificate about to be 2254 signed to be customised, to allow certain fields to be included 2255 or excluded and extension details. The old system didn't display 2256 multicharacter strings properly, omitted fields not in the policy 2257 and couldn't display additional details such as extensions. 2258 [Steve Henson] 2259 2260 *) Function EC_POINTs_mul for multiple scalar multiplication 2261 of an arbitrary number of elliptic curve points 2262 \sum scalars[i]*points[i], 2263 optionally including the generator defined for the EC_GROUP: 2264 scalar*generator + \sum scalars[i]*points[i]. 2265 2266 EC_POINT_mul is a simple wrapper function for the typical case 2267 that the point list has just one item (besides the optional 2268 generator). 2269 [Bodo Moeller] 2270 2271 *) First EC_METHODs for curves over GF(p): 2272 2273 EC_GFp_simple_method() uses the basic BN_mod_mul and BN_mod_sqr 2274 operations and provides various method functions that can also 2275 operate with faster implementations of modular arithmetic. 2276 2277 EC_GFp_mont_method() reuses most functions that are part of 2278 EC_GFp_simple_method, but uses Montgomery arithmetic. 2279 2280 [Bodo Moeller; point addition and point doubling 2281 implementation directly derived from source code provided by 2282 Lenka Fibikova <fibikova@exp-math.uni-essen.de>] 2283 2284 *) Framework for elliptic curves (crypto/ec/ec.h, crypto/ec/ec_lcl.h, 2285 crypto/ec/ec_lib.c): 2286 2287 Curves are EC_GROUP objects (with an optional group generator) 2288 based on EC_METHODs that are built into the library. 2289 2290 Points are EC_POINT objects based on EC_GROUP objects. 2291 2292 Most of the framework would be able to handle curves over arbitrary 2293 finite fields, but as there are no obvious types for fields other 2294 than GF(p), some functions are limited to that for now. 2295 [Bodo Moeller] 2296 2297 *) Add the -HTTP option to s_server. It is similar to -WWW, but requires 2298 that the file contains a complete HTTP response. 2299 [Richard Levitte] 2300 2301 *) Add the ec directory to mkdef.pl and mkfiles.pl. In mkdef.pl 2302 change the def and num file printf format specifier from "%-40sXXX" 2303 to "%-39s XXX". The latter will always guarantee a space after the 2304 field while the former will cause them to run together if the field 2305 is 40 of more characters long. 2306 [Steve Henson] 2307 2308 *) Constify the cipher and digest 'method' functions and structures 2309 and modify related functions to take constant EVP_MD and EVP_CIPHER 2310 pointers. 2311 [Steve Henson] 2312 2313 *) Hide BN_CTX structure details in bn_lcl.h instead of publishing them 2314 in <openssl/bn.h>. Also further increase BN_CTX_NUM to 32. 2315 [Bodo Moeller] 2316 2317 *) Modify EVP_Digest*() routines so they now return values. Although the 2318 internal software routines can never fail additional hardware versions 2319 might. 2320 [Steve Henson] 2321 2322 *) Clean up crypto/err/err.h and change some error codes to avoid conflicts: 2323 2324 Previously ERR_R_FATAL was too small and coincided with ERR_LIB_PKCS7 2325 (= ERR_R_PKCS7_LIB); it is now 64 instead of 32. 2326 2327 ASN1 error codes 2328 ERR_R_NESTED_ASN1_ERROR 2329 ... 2330 ERR_R_MISSING_ASN1_EOS 2331 were 4 .. 9, conflicting with 2332 ERR_LIB_RSA (= ERR_R_RSA_LIB) 2333 ... 2334 ERR_LIB_PEM (= ERR_R_PEM_LIB). 2335 They are now 58 .. 63 (i.e., just below ERR_R_FATAL). 2336 2337 Add new error code 'ERR_R_INTERNAL_ERROR'. 2338 [Bodo Moeller] 2339 2340 *) Don't overuse locks in crypto/err/err.c: For data retrieval, CRYPTO_r_lock 2341 suffices. 2342 [Bodo Moeller] 2343 2344 *) New option '-subj arg' for 'openssl req' and 'openssl ca'. This 2345 sets the subject name for a new request or supersedes the 2346 subject name in a given request. Formats that can be parsed are 2347 'CN=Some Name, OU=myOU, C=IT' 2348 and 2349 'CN=Some Name/OU=myOU/C=IT'. 2350 2351 Add options '-batch' and '-verbose' to 'openssl req'. 2352 [Massimiliano Pala <madwolf@hackmasters.net>] 2353 2354 *) Introduce the possibility to access global variables through 2355 functions on platform were that's the best way to handle exporting 2356 global variables in shared libraries. To enable this functionality, 2357 one must configure with "EXPORT_VAR_AS_FN" or defined the C macro 2358 "OPENSSL_EXPORT_VAR_AS_FUNCTION" in crypto/opensslconf.h (the latter 2359 is normally done by Configure or something similar). 2360 2361 To implement a global variable, use the macro OPENSSL_IMPLEMENT_GLOBAL 2362 in the source file (foo.c) like this: 2363 2364 OPENSSL_IMPLEMENT_GLOBAL(int,foo)=1; 2365 OPENSSL_IMPLEMENT_GLOBAL(double,bar); 2366 2367 To declare a global variable, use the macros OPENSSL_DECLARE_GLOBAL 2368 and OPENSSL_GLOBAL_REF in the header file (foo.h) like this: 2369 2370 OPENSSL_DECLARE_GLOBAL(int,foo); 2371 #define foo OPENSSL_GLOBAL_REF(foo) 2372 OPENSSL_DECLARE_GLOBAL(double,bar); 2373 #define bar OPENSSL_GLOBAL_REF(bar) 2374 2375 The #defines are very important, and therefore so is including the 2376 header file everywhere where the defined globals are used. 2377 2378 The macro OPENSSL_EXPORT_VAR_AS_FUNCTION also affects the definition 2379 of ASN.1 items, but that structure is a bit different. 2380 2381 The largest change is in util/mkdef.pl which has been enhanced with 2382 better and easier to understand logic to choose which symbols should 2383 go into the Windows .def files as well as a number of fixes and code 2384 cleanup (among others, algorithm keywords are now sorted 2385 lexicographically to avoid constant rewrites). 2386 [Richard Levitte] 2387 2388 *) In BN_div() keep a copy of the sign of 'num' before writing the 2389 result to 'rm' because if rm==num the value will be overwritten 2390 and produce the wrong result if 'num' is negative: this caused 2391 problems with BN_mod() and BN_nnmod(). 2392 [Steve Henson] 2393 2394 *) Function OCSP_request_verify(). This checks the signature on an 2395 OCSP request and verifies the signer certificate. The signer 2396 certificate is just checked for a generic purpose and OCSP request 2397 trust settings. 2398 [Steve Henson] 2399 2400 *) Add OCSP_check_validity() function to check the validity of OCSP 2401 responses. OCSP responses are prepared in real time and may only 2402 be a few seconds old. Simply checking that the current time lies 2403 between thisUpdate and nextUpdate max reject otherwise valid responses 2404 caused by either OCSP responder or client clock inaccuracy. Instead 2405 we allow thisUpdate and nextUpdate to fall within a certain period of 2406 the current time. The age of the response can also optionally be 2407 checked. Two new options -validity_period and -status_age added to 2408 ocsp utility. 2409 [Steve Henson] 2410 2411 *) If signature or public key algorithm is unrecognized print out its 2412 OID rather that just UNKNOWN. 2413 [Steve Henson] 2414 2415 *) Change OCSP_cert_to_id() to tolerate a NULL subject certificate and 2416 OCSP_cert_id_new() a NULL serialNumber. This allows a partial certificate 2417 ID to be generated from the issuer certificate alone which can then be 2418 passed to OCSP_id_issuer_cmp(). 2419 [Steve Henson] 2420 2421 *) New compilation option ASN1_ITEM_FUNCTIONS. This causes the new 2422 ASN1 modules to export functions returning ASN1_ITEM pointers 2423 instead of the ASN1_ITEM structures themselves. This adds several 2424 new macros which allow the underlying ASN1 function/structure to 2425 be accessed transparently. As a result code should not use ASN1_ITEM 2426 references directly (such as &X509_it) but instead use the relevant 2427 macros (such as ASN1_ITEM_rptr(X509)). This option is to allow 2428 use of the new ASN1 code on platforms where exporting structures 2429 is problematical (for example in shared libraries) but exporting 2430 functions returning pointers to structures is not. 2431 [Steve Henson] 2432 2433 *) Add support for overriding the generation of SSL/TLS session IDs. 2434 These callbacks can be registered either in an SSL_CTX or per SSL. 2435 The purpose of this is to allow applications to control, if they wish, 2436 the arbitrary values chosen for use as session IDs, particularly as it 2437 can be useful for session caching in multiple-server environments. A 2438 command-line switch for testing this (and any client code that wishes 2439 to use such a feature) has been added to "s_server". 2440 [Geoff Thorpe, Lutz Jaenicke] 2441 2442 *) Modify mkdef.pl to recognise and parse preprocessor conditionals 2443 of the form '#if defined(...) || defined(...) || ...' and 2444 '#if !defined(...) && !defined(...) && ...'. This also avoids 2445 the growing number of special cases it was previously handling. 2446 [Richard Levitte] 2447 2448 *) Make all configuration macros available for application by making 2449 sure they are available in opensslconf.h, by giving them names starting 2450 with "OPENSSL_" to avoid conflicts with other packages and by making 2451 sure e_os2.h will cover all platform-specific cases together with 2452 opensslconf.h. 2453 Additionally, it is now possible to define configuration/platform- 2454 specific names (called "system identities"). In the C code, these 2455 are prefixed with "OPENSSL_SYSNAME_". e_os2.h will create another 2456 macro with the name beginning with "OPENSSL_SYS_", which is determined 2457 from "OPENSSL_SYSNAME_*" or compiler-specific macros depending on 2458 what is available. 2459 [Richard Levitte] 2460 2461 *) New option -set_serial to 'req' and 'x509' this allows the serial 2462 number to use to be specified on the command line. Previously self 2463 signed certificates were hard coded with serial number 0 and the 2464 CA options of 'x509' had to use a serial number in a file which was 2465 auto incremented. 2466 [Steve Henson] 2467 2468 *) New options to 'ca' utility to support V2 CRL entry extensions. 2469 Currently CRL reason, invalidity date and hold instruction are 2470 supported. Add new CRL extensions to V3 code and some new objects. 2471 [Steve Henson] 2472 2473 *) New function EVP_CIPHER_CTX_set_padding() this is used to 2474 disable standard block padding (aka PKCS#5 padding) in the EVP 2475 API, which was previously mandatory. This means that the data is 2476 not padded in any way and so the total length much be a multiple 2477 of the block size, otherwise an error occurs. 2478 [Steve Henson] 2479 2480 *) Initial (incomplete) OCSP SSL support. 2481 [Steve Henson] 2482 2483 *) New function OCSP_parse_url(). This splits up a URL into its host, 2484 port and path components: primarily to parse OCSP URLs. New -url 2485 option to ocsp utility. 2486 [Steve Henson] 2487 2488 *) New nonce behavior. The return value of OCSP_check_nonce() now 2489 reflects the various checks performed. Applications can decide 2490 whether to tolerate certain situations such as an absent nonce 2491 in a response when one was present in a request: the ocsp application 2492 just prints out a warning. New function OCSP_add1_basic_nonce() 2493 this is to allow responders to include a nonce in a response even if 2494 the request is nonce-less. 2495 [Steve Henson] 2496 2497 *) Disable stdin buffering in load_cert (apps/apps.c) so that no certs are 2498 skipped when using openssl x509 multiple times on a single input file, 2499 e.g. "(openssl x509 -out cert1; openssl x509 -out cert2) <certs". 2500 [Bodo Moeller] 2501 2502 *) Make ASN1_UTCTIME_set_string() and ASN1_GENERALIZEDTIME_set_string() 2503 set string type: to handle setting ASN1_TIME structures. Fix ca 2504 utility to correctly initialize revocation date of CRLs. 2505 [Steve Henson] 2506 2507 *) New option SSL_OP_CIPHER_SERVER_PREFERENCE allows the server to override 2508 the clients preferred ciphersuites and rather use its own preferences. 2509 Should help to work around M$ SGC (Server Gated Cryptography) bug in 2510 Internet Explorer by ensuring unchanged hash method during stepup. 2511 (Also replaces the broken/deactivated SSL_OP_NON_EXPORT_FIRST option.) 2512 [Lutz Jaenicke] 2513 2514 *) Make mkdef.pl recognise all DECLARE_ASN1 macros, change rijndael 2515 to aes and add a new 'exist' option to print out symbols that don't 2516 appear to exist. 2517 [Steve Henson] 2518 2519 *) Additional options to ocsp utility to allow flags to be set and 2520 additional certificates supplied. 2521 [Steve Henson] 2522 2523 *) Add the option -VAfile to 'openssl ocsp', so the user can give the 2524 OCSP client a number of certificate to only verify the response 2525 signature against. 2526 [Richard Levitte] 2527 2528 *) Update Rijndael code to version 3.0 and change EVP AES ciphers to 2529 handle the new API. Currently only ECB, CBC modes supported. Add new 2530 AES OIDs. 2531 2532 Add TLS AES ciphersuites as described in RFC3268, "Advanced 2533 Encryption Standard (AES) Ciphersuites for Transport Layer 2534 Security (TLS)". (In beta versions of OpenSSL 0.9.7, these were 2535 not enabled by default and were not part of the "ALL" ciphersuite 2536 alias because they were not yet official; they could be 2537 explicitly requested by specifying the "AESdraft" ciphersuite 2538 group alias. In the final release of OpenSSL 0.9.7, the group 2539 alias is called "AES" and is part of "ALL".) 2540 [Ben Laurie, Steve Henson, Bodo Moeller] 2541 2542 *) New function OCSP_copy_nonce() to copy nonce value (if present) from 2543 request to response. 2544 [Steve Henson] 2545 2546 *) Functions for OCSP responders. OCSP_request_onereq_count(), 2547 OCSP_request_onereq_get0(), OCSP_onereq_get0_id() and OCSP_id_get0_info() 2548 extract information from a certificate request. OCSP_response_create() 2549 creates a response and optionally adds a basic response structure. 2550 OCSP_basic_add1_status() adds a complete single response to a basic 2551 response and returns the OCSP_SINGLERESP structure just added (to allow 2552 extensions to be included for example). OCSP_basic_add1_cert() adds a 2553 certificate to a basic response and OCSP_basic_sign() signs a basic 2554 response with various flags. New helper functions ASN1_TIME_check() 2555 (checks validity of ASN1_TIME structure) and ASN1_TIME_to_generalizedtime() 2556 (converts ASN1_TIME to GeneralizedTime). 2557 [Steve Henson] 2558 2559 *) Various new functions. EVP_Digest() combines EVP_Digest{Init,Update,Final}() 2560 in a single operation. X509_get0_pubkey_bitstr() extracts the public_key 2561 structure from a certificate. X509_pubkey_digest() digests the public_key 2562 contents: this is used in various key identifiers. 2563 [Steve Henson] 2564 2565 *) Make sk_sort() tolerate a NULL argument. 2566 [Steve Henson reported by Massimiliano Pala <madwolf@comune.modena.it>] 2567 2568 *) New OCSP verify flag OCSP_TRUSTOTHER. When set the "other" certificates 2569 passed by the function are trusted implicitly. If any of them signed the 2570 response then it is assumed to be valid and is not verified. 2571 [Steve Henson] 2572 2573 *) In PKCS7_set_type() initialise content_type in PKCS7_ENC_CONTENT 2574 to data. This was previously part of the PKCS7 ASN1 code. This 2575 was causing problems with OpenSSL created PKCS#12 and PKCS#7 structures. 2576 [Steve Henson, reported by Kenneth R. Robinette 2577 <support@securenetterm.com>] 2578 2579 *) Add CRYPTO_push_info() and CRYPTO_pop_info() calls to new ASN1 2580 routines: without these tracing memory leaks is very painful. 2581 Fix leaks in PKCS12 and PKCS7 routines. 2582 [Steve Henson] 2583 2584 *) Make X509_time_adj() cope with the new behaviour of ASN1_TIME_new(). 2585 Previously it initialised the 'type' argument to V_ASN1_UTCTIME which 2586 effectively meant GeneralizedTime would never be used. Now it 2587 is initialised to -1 but X509_time_adj() now has to check the value 2588 and use ASN1_TIME_set() if the value is not V_ASN1_UTCTIME or 2589 V_ASN1_GENERALIZEDTIME, without this it always uses GeneralizedTime. 2590 [Steve Henson, reported by Kenneth R. Robinette 2591 <support@securenetterm.com>] 2592 2593 *) Fixes to BN_to_ASN1_INTEGER when bn is zero. This would previously 2594 result in a zero length in the ASN1_INTEGER structure which was 2595 not consistent with the structure when d2i_ASN1_INTEGER() was used 2596 and would cause ASN1_INTEGER_cmp() to fail. Enhance s2i_ASN1_INTEGER() 2597 to cope with hex and negative integers. Fix bug in i2a_ASN1_INTEGER() 2598 where it did not print out a minus for negative ASN1_INTEGER. 2599 [Steve Henson] 2600 2601 *) Add summary printout to ocsp utility. The various functions which 2602 convert status values to strings have been renamed to: 2603 OCSP_response_status_str(), OCSP_cert_status_str() and 2604 OCSP_crl_reason_str() and are no longer static. New options 2605 to verify nonce values and to disable verification. OCSP response 2606 printout format cleaned up. 2607 [Steve Henson] 2608 2609 *) Add additional OCSP certificate checks. These are those specified 2610 in RFC2560. This consists of two separate checks: the CA of the 2611 certificate being checked must either be the OCSP signer certificate 2612 or the issuer of the OCSP signer certificate. In the latter case the 2613 OCSP signer certificate must contain the OCSP signing extended key 2614 usage. This check is performed by attempting to match the OCSP 2615 signer or the OCSP signer CA to the issuerNameHash and issuerKeyHash 2616 in the OCSP_CERTID structures of the response. 2617 [Steve Henson] 2618 2619 *) Initial OCSP certificate verification added to OCSP_basic_verify() 2620 and related routines. This uses the standard OpenSSL certificate 2621 verify routines to perform initial checks (just CA validity) and 2622 to obtain the certificate chain. Then additional checks will be 2623 performed on the chain. Currently the root CA is checked to see 2624 if it is explicitly trusted for OCSP signing. This is used to set 2625 a root CA as a global signing root: that is any certificate that 2626 chains to that CA is an acceptable OCSP signing certificate. 2627 [Steve Henson] 2628 2629 *) New '-extfile ...' option to 'openssl ca' for reading X.509v3 2630 extensions from a separate configuration file. 2631 As when reading extensions from the main configuration file, 2632 the '-extensions ...' option may be used for specifying the 2633 section to use. 2634 [Massimiliano Pala <madwolf@comune.modena.it>] 2635 2636 *) New OCSP utility. Allows OCSP requests to be generated or 2637 read. The request can be sent to a responder and the output 2638 parsed, outputed or printed in text form. Not complete yet: 2639 still needs to check the OCSP response validity. 2640 [Steve Henson] 2641 2642 *) New subcommands for 'openssl ca': 2643 'openssl ca -status <serial>' prints the status of the cert with 2644 the given serial number (according to the index file). 2645 'openssl ca -updatedb' updates the expiry status of certificates 2646 in the index file. 2647 [Massimiliano Pala <madwolf@comune.modena.it>] 2648 2649 *) New '-newreq-nodes' command option to CA.pl. This is like 2650 '-newreq', but calls 'openssl req' with the '-nodes' option 2651 so that the resulting key is not encrypted. 2652 [Damien Miller <djm@mindrot.org>] 2653 2654 *) New configuration for the GNU Hurd. 2655 [Jonathan Bartlett <johnnyb@wolfram.com> via Richard Levitte] 2656 2657 *) Initial code to implement OCSP basic response verify. This 2658 is currently incomplete. Currently just finds the signer's 2659 certificate and verifies the signature on the response. 2660 [Steve Henson] 2661 2662 *) New SSLeay_version code SSLEAY_DIR to determine the compiled-in 2663 value of OPENSSLDIR. This is available via the new '-d' option 2664 to 'openssl version', and is also included in 'openssl version -a'. 2665 [Bodo Moeller] 2666 2667 *) Allowing defining memory allocation callbacks that will be given 2668 file name and line number information in additional arguments 2669 (a const char* and an int). The basic functionality remains, as 2670 well as the original possibility to just replace malloc(), 2671 realloc() and free() by functions that do not know about these 2672 additional arguments. To register and find out the current 2673 settings for extended allocation functions, the following 2674 functions are provided: 2675 2676 CRYPTO_set_mem_ex_functions 2677 CRYPTO_set_locked_mem_ex_functions 2678 CRYPTO_get_mem_ex_functions 2679 CRYPTO_get_locked_mem_ex_functions 2680 2681 These work the same way as CRYPTO_set_mem_functions and friends. 2682 CRYPTO_get_[locked_]mem_functions now writes 0 where such an 2683 extended allocation function is enabled. 2684 Similarly, CRYPTO_get_[locked_]mem_ex_functions writes 0 where 2685 a conventional allocation function is enabled. 2686 [Richard Levitte, Bodo Moeller] 2687 2688 *) Finish off removing the remaining LHASH function pointer casts. 2689 There should no longer be any prototype-casting required when using 2690 the LHASH abstraction, and any casts that remain are "bugs". See 2691 the callback types and macros at the head of lhash.h for details 2692 (and "OBJ_cleanup" in crypto/objects/obj_dat.c as an example). 2693 [Geoff Thorpe] 2694 2695 *) Add automatic query of EGD sockets in RAND_poll() for the unix variant. 2696 If /dev/[u]random devices are not available or do not return enough 2697 entropy, EGD style sockets (served by EGD or PRNGD) will automatically 2698 be queried. 2699 The locations /var/run/egd-pool, /dev/egd-pool, /etc/egd-pool, and 2700 /etc/entropy will be queried once each in this sequence, quering stops 2701 when enough entropy was collected without querying more sockets. 2702 [Lutz Jaenicke] 2703 2704 *) Change the Unix RAND_poll() variant to be able to poll several 2705 random devices, as specified by DEVRANDOM, until a sufficient amount 2706 of data has been collected. We spend at most 10 ms on each file 2707 (select timeout) and read in non-blocking mode. DEVRANDOM now 2708 defaults to the list "/dev/urandom", "/dev/random", "/dev/srandom" 2709 (previously it was just the string "/dev/urandom"), so on typical 2710 platforms the 10 ms delay will never occur. 2711 Also separate out the Unix variant to its own file, rand_unix.c. 2712 For VMS, there's a currently-empty rand_vms.c. 2713 [Richard Levitte] 2714 2715 *) Move OCSP client related routines to ocsp_cl.c. These 2716 provide utility functions which an application needing 2717 to issue a request to an OCSP responder and analyse the 2718 response will typically need: as opposed to those which an 2719 OCSP responder itself would need which will be added later. 2720 2721 OCSP_request_sign() signs an OCSP request with an API similar 2722 to PKCS7_sign(). OCSP_response_status() returns status of OCSP 2723 response. OCSP_response_get1_basic() extracts basic response 2724 from response. OCSP_resp_find_status(): finds and extracts status 2725 information from an OCSP_CERTID structure (which will be created 2726 when the request structure is built). These are built from lower 2727 level functions which work on OCSP_SINGLERESP structures but 2728 wont normally be used unless the application wishes to examine 2729 extensions in the OCSP response for example. 2730 2731 Replace nonce routines with a pair of functions. 2732 OCSP_request_add1_nonce() adds a nonce value and optionally 2733 generates a random value. OCSP_check_nonce() checks the 2734 validity of the nonce in an OCSP response. 2735 [Steve Henson] 2736 2737 *) Change function OCSP_request_add() to OCSP_request_add0_id(). 2738 This doesn't copy the supplied OCSP_CERTID and avoids the 2739 need to free up the newly created id. Change return type 2740 to OCSP_ONEREQ to return the internal OCSP_ONEREQ structure. 2741 This can then be used to add extensions to the request. 2742 Deleted OCSP_request_new(), since most of its functionality 2743 is now in OCSP_REQUEST_new() (and the case insensitive name 2744 clash) apart from the ability to set the request name which 2745 will be added elsewhere. 2746 [Steve Henson] 2747 2748 *) Update OCSP API. Remove obsolete extensions argument from 2749 various functions. Extensions are now handled using the new 2750 OCSP extension code. New simple OCSP HTTP function which 2751 can be used to send requests and parse the response. 2752 [Steve Henson] 2753 2754 *) Fix the PKCS#7 (S/MIME) code to work with new ASN1. Two new 2755 ASN1_ITEM structures help with sign and verify. PKCS7_ATTR_SIGN 2756 uses the special reorder version of SET OF to sort the attributes 2757 and reorder them to match the encoded order. This resolves a long 2758 standing problem: a verify on a PKCS7 structure just after signing 2759 it used to fail because the attribute order did not match the 2760 encoded order. PKCS7_ATTR_VERIFY does not reorder the attributes: 2761 it uses the received order. This is necessary to tolerate some broken 2762 software that does not order SET OF. This is handled by encoding 2763 as a SEQUENCE OF but using implicit tagging (with UNIVERSAL class) 2764 to produce the required SET OF. 2765 [Steve Henson] 2766 2767 *) Have mk1mf.pl generate the macros OPENSSL_BUILD_SHLIBCRYPTO and 2768 OPENSSL_BUILD_SHLIBSSL and use them appropriately in the header 2769 files to get correct declarations of the ASN.1 item variables. 2770 [Richard Levitte] 2771 2772 *) Rewrite of PKCS#12 code to use new ASN1 functionality. Replace many 2773 PKCS#12 macros with real functions. Fix two unrelated ASN1 bugs: 2774 asn1_check_tlen() would sometimes attempt to use 'ctx' when it was 2775 NULL and ASN1_TYPE was not dereferenced properly in asn1_ex_c2i(). 2776 New ASN1 macro: DECLARE_ASN1_ITEM() which just declares the relevant 2777 ASN1_ITEM and no wrapper functions. 2778 [Steve Henson] 2779 2780 *) New functions or ASN1_item_d2i_fp() and ASN1_item_d2i_bio(). These 2781 replace the old function pointer based I/O routines. Change most of 2782 the *_d2i_bio() and *_d2i_fp() functions to use these. 2783 [Steve Henson] 2784 2785 *) Enhance mkdef.pl to be more accepting about spacing in C preprocessor 2786 lines, recognice more "algorithms" that can be deselected, and make 2787 it complain about algorithm deselection that isn't recognised. 2788 [Richard Levitte] 2789 2790 *) New ASN1 functions to handle dup, sign, verify, digest, pack and 2791 unpack operations in terms of ASN1_ITEM. Modify existing wrappers 2792 to use new functions. Add NO_ASN1_OLD which can be set to remove 2793 some old style ASN1 functions: this can be used to determine if old 2794 code will still work when these eventually go away. 2795 [Steve Henson] 2796 2797 *) New extension functions for OCSP structures, these follow the 2798 same conventions as certificates and CRLs. 2799 [Steve Henson] 2800 2801 *) New function X509V3_add1_i2d(). This automatically encodes and 2802 adds an extension. Its behaviour can be customised with various 2803 flags to append, replace or delete. Various wrappers added for 2804 certifcates and CRLs. 2805 [Steve Henson] 2806 2807 *) Fix to avoid calling the underlying ASN1 print routine when 2808 an extension cannot be parsed. Correct a typo in the 2809 OCSP_SERVICELOC extension. Tidy up print OCSP format. 2810 [Steve Henson] 2811 2812 *) Make mkdef.pl parse some of the ASN1 macros and add apropriate 2813 entries for variables. 2814 [Steve Henson] 2815 2816 *) Add functionality to apps/openssl.c for detecting locking 2817 problems: As the program is single-threaded, all we have 2818 to do is register a locking callback using an array for 2819 storing which locks are currently held by the program. 2820 [Bodo Moeller] 2821 2822 *) Use a lock around the call to CRYPTO_get_ex_new_index() in 2823 SSL_get_ex_data_X509_STORE_idx(), which is used in 2824 ssl_verify_cert_chain() and thus can be called at any time 2825 during TLS/SSL handshakes so that thread-safety is essential. 2826 Unfortunately, the ex_data design is not at all suited 2827 for multi-threaded use, so it probably should be abolished. 2828 [Bodo Moeller] 2829 2830 *) Added Broadcom "ubsec" ENGINE to OpenSSL. 2831 [Broadcom, tweaked and integrated by Geoff Thorpe] 2832 2833 *) Move common extension printing code to new function 2834 X509V3_print_extensions(). Reorganise OCSP print routines and 2835 implement some needed OCSP ASN1 functions. Add OCSP extensions. 2836 [Steve Henson] 2837 2838 *) New function X509_signature_print() to remove duplication in some 2839 print routines. 2840 [Steve Henson] 2841 2842 *) Add a special meaning when SET OF and SEQUENCE OF flags are both 2843 set (this was treated exactly the same as SET OF previously). This 2844 is used to reorder the STACK representing the structure to match the 2845 encoding. This will be used to get round a problem where a PKCS7 2846 structure which was signed could not be verified because the STACK 2847 order did not reflect the encoded order. 2848 [Steve Henson] 2849 2850 *) Reimplement the OCSP ASN1 module using the new code. 2851 [Steve Henson] 2852 2853 *) Update the X509V3 code to permit the use of an ASN1_ITEM structure 2854 for its ASN1 operations. The old style function pointers still exist 2855 for now but they will eventually go away. 2856 [Steve Henson] 2857 2858 *) Merge in replacement ASN1 code from the ASN1 branch. This almost 2859 completely replaces the old ASN1 functionality with a table driven 2860 encoder and decoder which interprets an ASN1_ITEM structure describing 2861 the ASN1 module. Compatibility with the existing ASN1 API (i2d,d2i) is 2862 largely maintained. Almost all of the old asn1_mac.h macro based ASN1 2863 has also been converted to the new form. 2864 [Steve Henson] 2865 2866 *) Change BN_mod_exp_recp so that negative moduli are tolerated 2867 (the sign is ignored). Similarly, ignore the sign in BN_MONT_CTX_set 2868 so that BN_mod_exp_mont and BN_mod_exp_mont_word work 2869 for negative moduli. 2870 [Bodo Moeller] 2871 2872 *) Fix BN_uadd and BN_usub: Always return non-negative results instead 2873 of not touching the result's sign bit. 2874 [Bodo Moeller] 2875 2876 *) BN_div bugfix: If the result is 0, the sign (res->neg) must not be 2877 set. 2878 [Bodo Moeller] 2879 2880 *) Changed the LHASH code to use prototypes for callbacks, and created 2881 macros to declare and implement thin (optionally static) functions 2882 that provide type-safety and avoid function pointer casting for the 2883 type-specific callbacks. 2884 [Geoff Thorpe] 2885 2886 *) Added Kerberos Cipher Suites to be used with TLS, as written in 2887 RFC 2712. 2888 [Veers Staats <staatsvr@asc.hpc.mil>, 2889 Jeffrey Altman <jaltman@columbia.edu>, via Richard Levitte] 2890 2891 *) Reformat the FAQ so the different questions and answers can be divided 2892 in sections depending on the subject. 2893 [Richard Levitte] 2894 2895 *) Have the zlib compression code load ZLIB.DLL dynamically under 2896 Windows. 2897 [Richard Levitte] 2898 2899 *) New function BN_mod_sqrt for computing square roots modulo a prime 2900 (using the probabilistic Tonelli-Shanks algorithm unless 2901 p == 3 (mod 4) or p == 5 (mod 8), which are cases that can 2902 be handled deterministically). 2903 [Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller] 2904 2905 *) Make BN_mod_inverse faster by explicitly handling small quotients 2906 in the Euclid loop. (Speed gain about 20% for small moduli [256 or 2907 512 bits], about 30% for larger ones [1024 or 2048 bits].) 2908 [Bodo Moeller] 2909 2910 *) New function BN_kronecker. 2911 [Bodo Moeller] 2912 2913 *) Fix BN_gcd so that it works on negative inputs; the result is 2914 positive unless both parameters are zero. 2915 Previously something reasonably close to an infinite loop was 2916 possible because numbers could be growing instead of shrinking 2917 in the implementation of Euclid's algorithm. 2918 [Bodo Moeller] 2919 2920 *) Fix BN_is_word() and BN_is_one() macros to take into account the 2921 sign of the number in question. 2922 2923 Fix BN_is_word(a,w) to work correctly for w == 0. 2924 2925 The old BN_is_word(a,w) macro is now called BN_abs_is_word(a,w) 2926 because its test if the absolute value of 'a' equals 'w'. 2927 Note that BN_abs_is_word does *not* handle w == 0 reliably; 2928 it exists mostly for use in the implementations of BN_is_zero(), 2929 BN_is_one(), and BN_is_word(). 2930 [Bodo Moeller] 2931 2932 *) New function BN_swap. 2933 [Bodo Moeller] 2934 2935 *) Use BN_nnmod instead of BN_mod in crypto/bn/bn_exp.c so that 2936 the exponentiation functions are more likely to produce reasonable 2937 results on negative inputs. 2938 [Bodo Moeller] 2939 2940 *) Change BN_mod_mul so that the result is always non-negative. 2941 Previously, it could be negative if one of the factors was negative; 2942 I don't think anyone really wanted that behaviour. 2943 [Bodo Moeller] 2944 2945 *) Move BN_mod_... functions into new file crypto/bn/bn_mod.c 2946 (except for exponentiation, which stays in crypto/bn/bn_exp.c, 2947 and BN_mod_mul_reciprocal, which stays in crypto/bn/bn_recp.c) 2948 and add new functions: 2949 2950 BN_nnmod 2951 BN_mod_sqr 2952 BN_mod_add 2953 BN_mod_add_quick 2954 BN_mod_sub 2955 BN_mod_sub_quick 2956 BN_mod_lshift1 2957 BN_mod_lshift1_quick 2958 BN_mod_lshift 2959 BN_mod_lshift_quick 2960 2961 These functions always generate non-negative results. 2962 2963 BN_nnmod otherwise is like BN_mod (if BN_mod computes a remainder r 2964 such that |m| < r < 0, BN_nnmod will output rem + |m| instead). 2965 2966 BN_mod_XXX_quick(r, a, [b,] m) generates the same result as 2967 BN_mod_XXX(r, a, [b,] m, ctx), but requires that a [and b] 2968 be reduced modulo m. 2969 [Lenka Fibikova <fibikova@exp-math.uni-essen.de>, Bodo Moeller] 2970 2971#if 0 2972 The following entry accidentily appeared in the CHANGES file 2973 distributed with OpenSSL 0.9.7. The modifications described in 2974 it do *not* apply to OpenSSL 0.9.7. 2975 2976 *) Remove a few calls to bn_wexpand() in BN_sqr() (the one in there 2977 was actually never needed) and in BN_mul(). The removal in BN_mul() 2978 required a small change in bn_mul_part_recursive() and the addition 2979 of the functions bn_cmp_part_words(), bn_sub_part_words() and 2980 bn_add_part_words(), which do the same thing as bn_cmp_words(), 2981 bn_sub_words() and bn_add_words() except they take arrays with 2982 differing sizes. 2983 [Richard Levitte] 2984#endif 2985 2986 *) In 'openssl passwd', verify passwords read from the terminal 2987 unless the '-salt' option is used (which usually means that 2988 verification would just waste user's time since the resulting 2989 hash is going to be compared with some given password hash) 2990 or the new '-noverify' option is used. 2991 2992 This is an incompatible change, but it does not affect 2993 non-interactive use of 'openssl passwd' (passwords on the command 2994 line, '-stdin' option, '-in ...' option) and thus should not 2995 cause any problems. 2996 [Bodo Moeller] 2997 2998 *) Remove all references to RSAref, since there's no more need for it. 2999 [Richard Levitte] 3000 3001 *) Make DSO load along a path given through an environment variable 3002 (SHLIB_PATH) with shl_load(). 3003 [Richard Levitte] 3004 3005 *) Constify the ENGINE code as a result of BIGNUM constification. 3006 Also constify the RSA code and most things related to it. In a 3007 few places, most notable in the depth of the ASN.1 code, ugly 3008 casts back to non-const were required (to be solved at a later 3009 time) 3010 [Richard Levitte] 3011 3012 *) Make it so the openssl application has all engines loaded by default. 3013 [Richard Levitte] 3014 3015 *) Constify the BIGNUM routines a little more. 3016 [Richard Levitte] 3017 3018 *) Add the following functions: 3019 3020 ENGINE_load_cswift() 3021 ENGINE_load_chil() 3022 ENGINE_load_atalla() 3023 ENGINE_load_nuron() 3024 ENGINE_load_builtin_engines() 3025 3026 That way, an application can itself choose if external engines that 3027 are built-in in OpenSSL shall ever be used or not. The benefit is 3028 that applications won't have to be linked with libdl or other dso 3029 libraries unless it's really needed. 3030 3031 Changed 'openssl engine' to load all engines on demand. 3032 Changed the engine header files to avoid the duplication of some 3033 declarations (they differed!). 3034 [Richard Levitte] 3035 3036 *) 'openssl engine' can now list capabilities. 3037 [Richard Levitte] 3038 3039 *) Better error reporting in 'openssl engine'. 3040 [Richard Levitte] 3041 3042 *) Never call load_dh_param(NULL) in s_server. 3043 [Bodo Moeller] 3044 3045 *) Add engine application. It can currently list engines by name and 3046 identity, and test if they are actually available. 3047 [Richard Levitte] 3048 3049 *) Improve RPM specification file by forcing symbolic linking and making 3050 sure the installed documentation is also owned by root.root. 3051 [Damien Miller <djm@mindrot.org>] 3052 3053 *) Give the OpenSSL applications more possibilities to make use of 3054 keys (public as well as private) handled by engines. 3055 [Richard Levitte] 3056 3057 *) Add OCSP code that comes from CertCo. 3058 [Richard Levitte] 3059 3060 *) Add VMS support for the Rijndael code. 3061 [Richard Levitte] 3062 3063 *) Added untested support for Nuron crypto accelerator. 3064 [Ben Laurie] 3065 3066 *) Add support for external cryptographic devices. This code was 3067 previously distributed separately as the "engine" branch. 3068 [Geoff Thorpe, Richard Levitte] 3069 3070 *) Rework the filename-translation in the DSO code. It is now possible to 3071 have far greater control over how a "name" is turned into a filename 3072 depending on the operating environment and any oddities about the 3073 different shared library filenames on each system. 3074 [Geoff Thorpe] 3075 3076 *) Support threads on FreeBSD-elf in Configure. 3077 [Richard Levitte] 3078 3079 *) Fix for SHA1 assembly problem with MASM: it produces 3080 warnings about corrupt line number information when assembling 3081 with debugging information. This is caused by the overlapping 3082 of two sections. 3083 [Bernd Matthes <mainbug@celocom.de>, Steve Henson] 3084 3085 *) NCONF changes. 3086 NCONF_get_number() has no error checking at all. As a replacement, 3087 NCONF_get_number_e() is defined (_e for "error checking") and is 3088 promoted strongly. The old NCONF_get_number is kept around for 3089 binary backward compatibility. 3090 Make it possible for methods to load from something other than a BIO, 3091 by providing a function pointer that is given a name instead of a BIO. 3092 For example, this could be used to load configuration data from an 3093 LDAP server. 3094 [Richard Levitte] 3095 3096 *) Fix for non blocking accept BIOs. Added new I/O special reason 3097 BIO_RR_ACCEPT to cover this case. Previously use of accept BIOs 3098 with non blocking I/O was not possible because no retry code was 3099 implemented. Also added new SSL code SSL_WANT_ACCEPT to cover 3100 this case. 3101 [Steve Henson] 3102 3103 *) Added the beginnings of Rijndael support. 3104 [Ben Laurie] 3105 3106 *) Fix for bug in DirectoryString mask setting. Add support for 3107 X509_NAME_print_ex() in 'req' and X509_print_ex() function 3108 to allow certificate printing to more controllable, additional 3109 'certopt' option to 'x509' to allow new printing options to be 3110 set. 3111 [Steve Henson] 3112 3113 *) Clean old EAY MD5 hack from e_os.h. 3114 [Richard Levitte] 3115 3116 Changes between 0.9.6l and 0.9.6m [17 Mar 2004] 3117 3118 *) Fix null-pointer assignment in do_change_cipher_spec() revealed 3119 by using the Codenomicon TLS Test Tool (CVE-2004-0079) 3120 [Joe Orton, Steve Henson] 3121 3122 Changes between 0.9.6k and 0.9.6l [04 Nov 2003] 3123 3124 *) Fix additional bug revealed by the NISCC test suite: 3125 3126 Stop bug triggering large recursion when presented with 3127 certain ASN.1 tags (CVE-2003-0851) 3128 [Steve Henson] 3129 3130 Changes between 0.9.6j and 0.9.6k [30 Sep 2003] 3131 3132 *) Fix various bugs revealed by running the NISCC test suite: 3133 3134 Stop out of bounds reads in the ASN1 code when presented with 3135 invalid tags (CVE-2003-0543 and CVE-2003-0544). 3136 3137 If verify callback ignores invalid public key errors don't try to check 3138 certificate signature with the NULL public key. 3139 3140 [Steve Henson] 3141 3142 *) In ssl3_accept() (ssl/s3_srvr.c) only accept a client certificate 3143 if the server requested one: as stated in TLS 1.0 and SSL 3.0 3144 specifications. 3145 [Steve Henson] 3146 3147 *) In ssl3_get_client_hello() (ssl/s3_srvr.c), tolerate additional 3148 extra data after the compression methods not only for TLS 1.0 3149 but also for SSL 3.0 (as required by the specification). 3150 [Bodo Moeller; problem pointed out by Matthias Loepfe] 3151 3152 *) Change X509_certificate_type() to mark the key as exported/exportable 3153 when it's 512 *bits* long, not 512 bytes. 3154 [Richard Levitte] 3155 3156 Changes between 0.9.6i and 0.9.6j [10 Apr 2003] 3157 3158 *) Countermeasure against the Klima-Pokorny-Rosa extension of 3159 Bleichbacher's attack on PKCS #1 v1.5 padding: treat 3160 a protocol version number mismatch like a decryption error 3161 in ssl3_get_client_key_exchange (ssl/s3_srvr.c). 3162 [Bodo Moeller] 3163 3164 *) Turn on RSA blinding by default in the default implementation 3165 to avoid a timing attack. Applications that don't want it can call 3166 RSA_blinding_off() or use the new flag RSA_FLAG_NO_BLINDING. 3167 They would be ill-advised to do so in most cases. 3168 [Ben Laurie, Steve Henson, Geoff Thorpe, Bodo Moeller] 3169 3170 *) Change RSA blinding code so that it works when the PRNG is not 3171 seeded (in this case, the secret RSA exponent is abused as 3172 an unpredictable seed -- if it is not unpredictable, there 3173 is no point in blinding anyway). Make RSA blinding thread-safe 3174 by remembering the creator's thread ID in rsa->blinding and 3175 having all other threads use local one-time blinding factors 3176 (this requires more computation than sharing rsa->blinding, but 3177 avoids excessive locking; and if an RSA object is not shared 3178 between threads, blinding will still be very fast). 3179 [Bodo Moeller] 3180 3181 Changes between 0.9.6h and 0.9.6i [19 Feb 2003] 3182 3183 *) In ssl3_get_record (ssl/s3_pkt.c), minimize information leaked 3184 via timing by performing a MAC computation even if incorrrect 3185 block cipher padding has been found. This is a countermeasure 3186 against active attacks where the attacker has to distinguish 3187 between bad padding and a MAC verification error. (CVE-2003-0078) 3188 3189 [Bodo Moeller; problem pointed out by Brice Canvel (EPFL), 3190 Alain Hiltgen (UBS), Serge Vaudenay (EPFL), and 3191 Martin Vuagnoux (EPFL, Ilion)] 3192 3193 Changes between 0.9.6g and 0.9.6h [5 Dec 2002] 3194 3195 *) New function OPENSSL_cleanse(), which is used to cleanse a section of 3196 memory from it's contents. This is done with a counter that will 3197 place alternating values in each byte. This can be used to solve 3198 two issues: 1) the removal of calls to memset() by highly optimizing 3199 compilers, and 2) cleansing with other values than 0, since those can 3200 be read through on certain media, for example a swap space on disk. 3201 [Geoff Thorpe] 3202 3203 *) Bugfix: client side session caching did not work with external caching, 3204 because the session->cipher setting was not restored when reloading 3205 from the external cache. This problem was masked, when 3206 SSL_OP_NETSCAPE_REUSE_CIPHER_CHANGE_BUG (part of SSL_OP_ALL) was set. 3207 (Found by Steve Haslam <steve@araqnid.ddts.net>.) 3208 [Lutz Jaenicke] 3209 3210 *) Fix client_certificate (ssl/s2_clnt.c): The permissible total 3211 length of the REQUEST-CERTIFICATE message is 18 .. 34, not 17 .. 33. 3212 [Zeev Lieber <zeev-l@yahoo.com>] 3213 3214 *) Undo an undocumented change introduced in 0.9.6e which caused 3215 repeated calls to OpenSSL_add_all_ciphers() and 3216 OpenSSL_add_all_digests() to be ignored, even after calling 3217 EVP_cleanup(). 3218 [Richard Levitte] 3219 3220 *) Change the default configuration reader to deal with last line not 3221 being properly terminated. 3222 [Richard Levitte] 3223 3224 *) Change X509_NAME_cmp() so it applies the special rules on handling 3225 DN values that are of type PrintableString, as well as RDNs of type 3226 emailAddress where the value has the type ia5String. 3227 [stefank@valicert.com via Richard Levitte] 3228 3229 *) Add a SSL_SESS_CACHE_NO_INTERNAL_STORE flag to take over half 3230 the job SSL_SESS_CACHE_NO_INTERNAL_LOOKUP was inconsistently 3231 doing, define a new flag (SSL_SESS_CACHE_NO_INTERNAL) to be 3232 the bitwise-OR of the two for use by the majority of applications 3233 wanting this behaviour, and update the docs. The documented 3234 behaviour and actual behaviour were inconsistent and had been 3235 changing anyway, so this is more a bug-fix than a behavioural 3236 change. 3237 [Geoff Thorpe, diagnosed by Nadav Har'El] 3238 3239 *) Don't impose a 16-byte length minimum on session IDs in ssl/s3_clnt.c 3240 (the SSL 3.0 and TLS 1.0 specifications allow any length up to 32 bytes). 3241 [Bodo Moeller] 3242 3243 *) Fix initialization code race conditions in 3244 SSLv23_method(), SSLv23_client_method(), SSLv23_server_method(), 3245 SSLv2_method(), SSLv2_client_method(), SSLv2_server_method(), 3246 SSLv3_method(), SSLv3_client_method(), SSLv3_server_method(), 3247 TLSv1_method(), TLSv1_client_method(), TLSv1_server_method(), 3248 ssl2_get_cipher_by_char(), 3249 ssl3_get_cipher_by_char(). 3250 [Patrick McCormick <patrick@tellme.com>, Bodo Moeller] 3251 3252 *) Reorder cleanup sequence in SSL_CTX_free(): only remove the ex_data after 3253 the cached sessions are flushed, as the remove_cb() might use ex_data 3254 contents. Bug found by Sam Varshavchik <mrsam@courier-mta.com> 3255 (see [openssl.org #212]). 3256 [Geoff Thorpe, Lutz Jaenicke] 3257 3258 *) Fix typo in OBJ_txt2obj which incorrectly passed the content 3259 length, instead of the encoding length to d2i_ASN1_OBJECT. 3260 [Steve Henson] 3261 3262 Changes between 0.9.6f and 0.9.6g [9 Aug 2002] 3263 3264 *) [In 0.9.6g-engine release:] 3265 Fix crypto/engine/vendor_defns/cswift.h for WIN32 (use '_stdcall'). 3266 [Lynn Gazis <lgazis@rainbow.com>] 3267 3268 Changes between 0.9.6e and 0.9.6f [8 Aug 2002] 3269 3270 *) Fix ASN1 checks. Check for overflow by comparing with LONG_MAX 3271 and get fix the header length calculation. 3272 [Florian Weimer <Weimer@CERT.Uni-Stuttgart.DE>, 3273 Alon Kantor <alonk@checkpoint.com> (and others), 3274 Steve Henson] 3275 3276 *) Use proper error handling instead of 'assertions' in buffer 3277 overflow checks added in 0.9.6e. This prevents DoS (the 3278 assertions could call abort()). 3279 [Arne Ansper <arne@ats.cyber.ee>, Bodo Moeller] 3280 3281 Changes between 0.9.6d and 0.9.6e [30 Jul 2002] 3282 3283 *) Add various sanity checks to asn1_get_length() to reject 3284 the ASN1 length bytes if they exceed sizeof(long), will appear 3285 negative or the content length exceeds the length of the 3286 supplied buffer. 3287 [Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>] 3288 3289 *) Fix cipher selection routines: ciphers without encryption had no flags 3290 for the cipher strength set and where therefore not handled correctly 3291 by the selection routines (PR #130). 3292 [Lutz Jaenicke] 3293 3294 *) Fix EVP_dsa_sha macro. 3295 [Nils Larsch] 3296 3297 *) New option 3298 SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS 3299 for disabling the SSL 3.0/TLS 1.0 CBC vulnerability countermeasure 3300 that was added in OpenSSL 0.9.6d. 3301 3302 As the countermeasure turned out to be incompatible with some 3303 broken SSL implementations, the new option is part of SSL_OP_ALL. 3304 SSL_OP_ALL is usually employed when compatibility with weird SSL 3305 implementations is desired (e.g. '-bugs' option to 's_client' and 3306 's_server'), so the new option is automatically set in many 3307 applications. 3308 [Bodo Moeller] 3309 3310 *) Changes in security patch: 3311 3312 Changes marked "(CHATS)" were sponsored by the Defense Advanced 3313 Research Projects Agency (DARPA) and Air Force Research Laboratory, 3314 Air Force Materiel Command, USAF, under agreement number 3315 F30602-01-2-0537. 3316 3317 *) Add various sanity checks to asn1_get_length() to reject 3318 the ASN1 length bytes if they exceed sizeof(long), will appear 3319 negative or the content length exceeds the length of the 3320 supplied buffer. (CVE-2002-0659) 3321 [Steve Henson, Adi Stav <stav@mercury.co.il>, James Yonan <jim@ntlp.com>] 3322 3323 *) Assertions for various potential buffer overflows, not known to 3324 happen in practice. 3325 [Ben Laurie (CHATS)] 3326 3327 *) Various temporary buffers to hold ASCII versions of integers were 3328 too small for 64 bit platforms. (CVE-2002-0655) 3329 [Matthew Byng-Maddick <mbm@aldigital.co.uk> and Ben Laurie (CHATS)> 3330 3331 *) Remote buffer overflow in SSL3 protocol - an attacker could 3332 supply an oversized session ID to a client. (CVE-2002-0656) 3333 [Ben Laurie (CHATS)] 3334 3335 *) Remote buffer overflow in SSL2 protocol - an attacker could 3336 supply an oversized client master key. (CVE-2002-0656) 3337 [Ben Laurie (CHATS)] 3338 3339 Changes between 0.9.6c and 0.9.6d [9 May 2002] 3340 3341 *) Fix crypto/asn1/a_sign.c so that 'parameters' is omitted (not 3342 encoded as NULL) with id-dsa-with-sha1. 3343 [Nils Larsch <nla@trustcenter.de>; problem pointed out by Bodo Moeller] 3344 3345 *) Check various X509_...() return values in apps/req.c. 3346 [Nils Larsch <nla@trustcenter.de>] 3347 3348 *) Fix BASE64 decode (EVP_DecodeUpdate) for data with CR/LF ended lines: 3349 an end-of-file condition would erronously be flagged, when the CRLF 3350 was just at the end of a processed block. The bug was discovered when 3351 processing data through a buffering memory BIO handing the data to a 3352 BASE64-decoding BIO. Bug fund and patch submitted by Pavel Tsekov 3353 <ptsekov@syntrex.com> and Nedelcho Stanev. 3354 [Lutz Jaenicke] 3355 3356 *) Implement a countermeasure against a vulnerability recently found 3357 in CBC ciphersuites in SSL 3.0/TLS 1.0: Send an empty fragment 3358 before application data chunks to avoid the use of known IVs 3359 with data potentially chosen by the attacker. 3360 [Bodo Moeller] 3361 3362 *) Fix length checks in ssl3_get_client_hello(). 3363 [Bodo Moeller] 3364 3365 *) TLS/SSL library bugfix: use s->s3->in_read_app_data differently 3366 to prevent ssl3_read_internal() from incorrectly assuming that 3367 ssl3_read_bytes() found application data while handshake 3368 processing was enabled when in fact s->s3->in_read_app_data was 3369 merely automatically cleared during the initial handshake. 3370 [Bodo Moeller; problem pointed out by Arne Ansper <arne@ats.cyber.ee>] 3371 3372 *) Fix object definitions for Private and Enterprise: they were not 3373 recognized in their shortname (=lowercase) representation. Extend 3374 obj_dat.pl to issue an error when using undefined keywords instead 3375 of silently ignoring the problem (Svenning Sorensen 3376 <sss@sss.dnsalias.net>). 3377 [Lutz Jaenicke] 3378 3379 *) Fix DH_generate_parameters() so that it works for 'non-standard' 3380 generators, i.e. generators other than 2 and 5. (Previously, the 3381 code did not properly initialise the 'add' and 'rem' values to 3382 BN_generate_prime().) 3383 3384 In the new general case, we do not insist that 'generator' is 3385 actually a primitive root: This requirement is rather pointless; 3386 a generator of the order-q subgroup is just as good, if not 3387 better. 3388 [Bodo Moeller] 3389 3390 *) Map new X509 verification errors to alerts. Discovered and submitted by 3391 Tom Wu <tom@arcot.com>. 3392 [Lutz Jaenicke] 3393 3394 *) Fix ssl3_pending() (ssl/s3_lib.c) to prevent SSL_pending() from 3395 returning non-zero before the data has been completely received 3396 when using non-blocking I/O. 3397 [Bodo Moeller; problem pointed out by John Hughes] 3398 3399 *) Some of the ciphers missed the strength entry (SSL_LOW etc). 3400 [Ben Laurie, Lutz Jaenicke] 3401 3402 *) Fix bug in SSL_clear(): bad sessions were not removed (found by 3403 Yoram Zahavi <YoramZ@gilian.com>). 3404 [Lutz Jaenicke] 3405 3406 *) Add information about CygWin 1.3 and on, and preserve proper 3407 configuration for the versions before that. 3408 [Corinna Vinschen <vinschen@redhat.com> and Richard Levitte] 3409 3410 *) Make removal from session cache (SSL_CTX_remove_session()) more robust: 3411 check whether we deal with a copy of a session and do not delete from 3412 the cache in this case. Problem reported by "Izhar Shoshani Levi" 3413 <izhar@checkpoint.com>. 3414 [Lutz Jaenicke] 3415 3416 *) Do not store session data into the internal session cache, if it 3417 is never intended to be looked up (SSL_SESS_CACHE_NO_INTERNAL_LOOKUP 3418 flag is set). Proposed by Aslam <aslam@funk.com>. 3419 [Lutz Jaenicke] 3420 3421 *) Have ASN1_BIT_STRING_set_bit() really clear a bit when the requested 3422 value is 0. 3423 [Richard Levitte] 3424 3425 *) [In 0.9.6d-engine release:] 3426 Fix a crashbug and a logic bug in hwcrhk_load_pubkey(). 3427 [Toomas Kiisk <vix@cyber.ee> via Richard Levitte] 3428 3429 *) Add the configuration target linux-s390x. 3430 [Neale Ferguson <Neale.Ferguson@SoftwareAG-USA.com> via Richard Levitte] 3431 3432 *) The earlier bugfix for the SSL3_ST_SW_HELLO_REQ_C case of 3433 ssl3_accept (ssl/s3_srvr.c) incorrectly used a local flag 3434 variable as an indication that a ClientHello message has been 3435 received. As the flag value will be lost between multiple 3436 invocations of ssl3_accept when using non-blocking I/O, the 3437 function may not be aware that a handshake has actually taken 3438 place, thus preventing a new session from being added to the 3439 session cache. 3440 3441 To avoid this problem, we now set s->new_session to 2 instead of 3442 using a local variable. 3443 [Lutz Jaenicke, Bodo Moeller] 3444 3445 *) Bugfix: Return -1 from ssl3_get_server_done (ssl3/s3_clnt.c) 3446 if the SSL_R_LENGTH_MISMATCH error is detected. 3447 [Geoff Thorpe, Bodo Moeller] 3448 3449 *) New 'shared_ldflag' column in Configure platform table. 3450 [Richard Levitte] 3451 3452 *) Fix EVP_CIPHER_mode macro. 3453 ["Dan S. Camper" <dan@bti.net>] 3454 3455 *) Fix ssl3_read_bytes (ssl/s3_pkt.c): To ignore messages of unknown 3456 type, we must throw them away by setting rr->length to 0. 3457 [D P Chang <dpc@qualys.com>] 3458 3459 Changes between 0.9.6b and 0.9.6c [21 dec 2001] 3460 3461 *) Fix BN_rand_range bug pointed out by Dominikus Scherkl 3462 <Dominikus.Scherkl@biodata.com>. (The previous implementation 3463 worked incorrectly for those cases where range = 10..._2 and 3464 3*range is two bits longer than range.) 3465 [Bodo Moeller] 3466 3467 *) Only add signing time to PKCS7 structures if it is not already 3468 present. 3469 [Steve Henson] 3470 3471 *) Fix crypto/objects/objects.h: "ld-ce" should be "id-ce", 3472 OBJ_ld_ce should be OBJ_id_ce. 3473 Also some ip-pda OIDs in crypto/objects/objects.txt were 3474 incorrect (cf. RFC 3039). 3475 [Matt Cooper, Frederic Giudicelli, Bodo Moeller] 3476 3477 *) Release CRYPTO_LOCK_DYNLOCK when CRYPTO_destroy_dynlockid() 3478 returns early because it has nothing to do. 3479 [Andy Schneider <andy.schneider@bjss.co.uk>] 3480 3481 *) [In 0.9.6c-engine release:] 3482 Fix mutex callback return values in crypto/engine/hw_ncipher.c. 3483 [Andy Schneider <andy.schneider@bjss.co.uk>] 3484 3485 *) [In 0.9.6c-engine release:] 3486 Add support for Cryptographic Appliance's keyserver technology. 3487 (Use engine 'keyclient') 3488 [Cryptographic Appliances and Geoff Thorpe] 3489 3490 *) Add a configuration entry for OS/390 Unix. The C compiler 'c89' 3491 is called via tools/c89.sh because arguments have to be 3492 rearranged (all '-L' options must appear before the first object 3493 modules). 3494 [Richard Shapiro <rshapiro@abinitio.com>] 3495 3496 *) [In 0.9.6c-engine release:] 3497 Add support for Broadcom crypto accelerator cards, backported 3498 from 0.9.7. 3499 [Broadcom, Nalin Dahyabhai <nalin@redhat.com>, Mark Cox] 3500 3501 *) [In 0.9.6c-engine release:] 3502 Add support for SureWare crypto accelerator cards from 3503 Baltimore Technologies. (Use engine 'sureware') 3504 [Baltimore Technologies and Mark Cox] 3505 3506 *) [In 0.9.6c-engine release:] 3507 Add support for crypto accelerator cards from Accelerated 3508 Encryption Processing, www.aep.ie. (Use engine 'aep') 3509 [AEP Inc. and Mark Cox] 3510 3511 *) Add a configuration entry for gcc on UnixWare. 3512 [Gary Benson <gbenson@redhat.com>] 3513 3514 *) Change ssl/s2_clnt.c and ssl/s2_srvr.c so that received handshake 3515 messages are stored in a single piece (fixed-length part and 3516 variable-length part combined) and fix various bugs found on the way. 3517 [Bodo Moeller] 3518 3519 *) Disable caching in BIO_gethostbyname(), directly use gethostbyname() 3520 instead. BIO_gethostbyname() does not know what timeouts are 3521 appropriate, so entries would stay in cache even when they have 3522 become invalid. 3523 [Bodo Moeller; problem pointed out by Rich Salz <rsalz@zolera.com> 3524 3525 *) Change ssl23_get_client_hello (ssl/s23_srvr.c) behaviour when 3526 faced with a pathologically small ClientHello fragment that does 3527 not contain client_version: Instead of aborting with an error, 3528 simply choose the highest available protocol version (i.e., 3529 TLS 1.0 unless it is disabled). In practice, ClientHello 3530 messages are never sent like this, but this change gives us 3531 strictly correct behaviour at least for TLS. 3532 [Bodo Moeller] 3533 3534 *) Fix SSL handshake functions and SSL_clear() such that SSL_clear() 3535 never resets s->method to s->ctx->method when called from within 3536 one of the SSL handshake functions. 3537 [Bodo Moeller; problem pointed out by Niko Baric] 3538 3539 *) In ssl3_get_client_hello (ssl/s3_srvr.c), generate a fatal alert 3540 (sent using the client's version number) if client_version is 3541 smaller than the protocol version in use. Also change 3542 ssl23_get_client_hello (ssl/s23_srvr.c) to select TLS 1.0 if 3543 the client demanded SSL 3.0 but only TLS 1.0 is enabled; then 3544 the client will at least see that alert. 3545 [Bodo Moeller] 3546 3547 *) Fix ssl3_get_message (ssl/s3_both.c) to handle message fragmentation 3548 correctly. 3549 [Bodo Moeller] 3550 3551 *) Avoid infinite loop in ssl3_get_message (ssl/s3_both.c) if a 3552 client receives HelloRequest while in a handshake. 3553 [Bodo Moeller; bug noticed by Andy Schneider <andy.schneider@bjss.co.uk>] 3554 3555 *) Bugfix in ssl3_accept (ssl/s3_srvr.c): Case SSL3_ST_SW_HELLO_REQ_C 3556 should end in 'break', not 'goto end' which circuments various 3557 cleanups done in state SSL_ST_OK. But session related stuff 3558 must be disabled for SSL_ST_OK in the case that we just sent a 3559 HelloRequest. 3560 3561 Also avoid some overhead by not calling ssl_init_wbio_buffer() 3562 before just sending a HelloRequest. 3563 [Bodo Moeller, Eric Rescorla <ekr@rtfm.com>] 3564 3565 *) Fix ssl/s3_enc.c, ssl/t1_enc.c and ssl/s3_pkt.c so that we don't 3566 reveal whether illegal block cipher padding was found or a MAC 3567 verification error occured. (Neither SSLerr() codes nor alerts 3568 are directly visible to potential attackers, but the information 3569 may leak via logfiles.) 3570 3571 Similar changes are not required for the SSL 2.0 implementation 3572 because the number of padding bytes is sent in clear for SSL 2.0, 3573 and the extra bytes are just ignored. However ssl/s2_pkt.c 3574 failed to verify that the purported number of padding bytes is in 3575 the legal range. 3576 [Bodo Moeller] 3577 3578 *) Add OpenUNIX-8 support including shared libraries 3579 (Boyd Lynn Gerber <gerberb@zenez.com>). 3580 [Lutz Jaenicke] 3581 3582 *) Improve RSA_padding_check_PKCS1_OAEP() check again to avoid 3583 'wristwatch attack' using huge encoding parameters (cf. 3584 James H. Manger's CRYPTO 2001 paper). Note that the 3585 RSA_PKCS1_OAEP_PADDING case of RSA_private_decrypt() does not use 3586 encoding parameters and hence was not vulnerable. 3587 [Bodo Moeller] 3588 3589 *) BN_sqr() bug fix. 3590 [Ulf M�ller, reported by Jim Ellis <jim.ellis@cavium.com>] 3591 3592 *) Rabin-Miller test analyses assume uniformly distributed witnesses, 3593 so use BN_pseudo_rand_range() instead of using BN_pseudo_rand() 3594 followed by modular reduction. 3595 [Bodo Moeller; pointed out by Adam Young <AYoung1@NCSUS.JNJ.COM>] 3596 3597 *) Add BN_pseudo_rand_range() with obvious functionality: BN_rand_range() 3598 equivalent based on BN_pseudo_rand() instead of BN_rand(). 3599 [Bodo Moeller] 3600 3601 *) s3_srvr.c: allow sending of large client certificate lists (> 16 kB). 3602 This function was broken, as the check for a new client hello message 3603 to handle SGC did not allow these large messages. 3604 (Tracked down by "Douglas E. Engert" <deengert@anl.gov>.) 3605 [Lutz Jaenicke] 3606 3607 *) Add alert descriptions for TLSv1 to SSL_alert_desc_string[_long](). 3608 [Lutz Jaenicke] 3609 3610 *) Fix buggy behaviour of BIO_get_num_renegotiates() and BIO_ctrl() 3611 for BIO_C_GET_WRITE_BUF_SIZE ("Stephen Hinton" <shinton@netopia.com>). 3612 [Lutz Jaenicke] 3613 3614 *) Rework the configuration and shared library support for Tru64 Unix. 3615 The configuration part makes use of modern compiler features and 3616 still retains old compiler behavior for those that run older versions 3617 of the OS. The shared library support part includes a variant that 3618 uses the RPATH feature, and is available through the special 3619 configuration target "alpha-cc-rpath", which will never be selected 3620 automatically. 3621 [Tim Mooney <mooney@dogbert.cc.ndsu.NoDak.edu> via Richard Levitte] 3622 3623 *) In ssl3_get_key_exchange (ssl/s3_clnt.c), call ssl3_get_message() 3624 with the same message size as in ssl3_get_certificate_request(). 3625 Otherwise, if no ServerKeyExchange message occurs, CertificateRequest 3626 messages might inadvertently be reject as too long. 3627 [Petr Lampa <lampa@fee.vutbr.cz>] 3628 3629 *) Enhanced support for IA-64 Unix platforms (well, Linux and HP-UX). 3630 [Andy Polyakov] 3631 3632 *) Modified SSL library such that the verify_callback that has been set 3633 specificly for an SSL object with SSL_set_verify() is actually being 3634 used. Before the change, a verify_callback set with this function was 3635 ignored and the verify_callback() set in the SSL_CTX at the time of 3636 the call was used. New function X509_STORE_CTX_set_verify_cb() introduced 3637 to allow the necessary settings. 3638 [Lutz Jaenicke] 3639 3640 *) Initialize static variable in crypto/dsa/dsa_lib.c and crypto/dh/dh_lib.c 3641 explicitly to NULL, as at least on Solaris 8 this seems not always to be 3642 done automatically (in contradiction to the requirements of the C 3643 standard). This made problems when used from OpenSSH. 3644 [Lutz Jaenicke] 3645 3646 *) In OpenSSL 0.9.6a and 0.9.6b, crypto/dh/dh_key.c ignored 3647 dh->length and always used 3648 3649 BN_rand_range(priv_key, dh->p). 3650 3651 BN_rand_range() is not necessary for Diffie-Hellman, and this 3652 specific range makes Diffie-Hellman unnecessarily inefficient if 3653 dh->length (recommended exponent length) is much smaller than the 3654 length of dh->p. We could use BN_rand_range() if the order of 3655 the subgroup was stored in the DH structure, but we only have 3656 dh->length. 3657 3658 So switch back to 3659 3660 BN_rand(priv_key, l, ...) 3661 3662 where 'l' is dh->length if this is defined, or BN_num_bits(dh->p)-1 3663 otherwise. 3664 [Bodo Moeller] 3665 3666 *) In 3667 3668 RSA_eay_public_encrypt 3669 RSA_eay_private_decrypt 3670 RSA_eay_private_encrypt (signing) 3671 RSA_eay_public_decrypt (signature verification) 3672 3673 (default implementations for RSA_public_encrypt, 3674 RSA_private_decrypt, RSA_private_encrypt, RSA_public_decrypt), 3675 always reject numbers >= n. 3676 [Bodo Moeller] 3677 3678 *) In crypto/rand/md_rand.c, use a new short-time lock CRYPTO_LOCK_RAND2 3679 to synchronize access to 'locking_thread'. This is necessary on 3680 systems where access to 'locking_thread' (an 'unsigned long' 3681 variable) is not atomic. 3682 [Bodo Moeller] 3683 3684 *) In crypto/rand/md_rand.c, set 'locking_thread' to current thread's ID 3685 *before* setting the 'crypto_lock_rand' flag. The previous code had 3686 a race condition if 0 is a valid thread ID. 3687 [Travis Vitek <vitek@roguewave.com>] 3688 3689 *) Add support for shared libraries under Irix. 3690 [Albert Chin-A-Young <china@thewrittenword.com>] 3691 3692 *) Add configuration option to build on Linux on both big-endian and 3693 little-endian MIPS. 3694 [Ralf Baechle <ralf@uni-koblenz.de>] 3695 3696 *) Add the possibility to create shared libraries on HP-UX. 3697 [Richard Levitte] 3698 3699 Changes between 0.9.6a and 0.9.6b [9 Jul 2001] 3700 3701 *) Change ssleay_rand_bytes (crypto/rand/md_rand.c) 3702 to avoid a SSLeay/OpenSSL PRNG weakness pointed out by 3703 Markku-Juhani O. Saarinen <markku-juhani.saarinen@nokia.com>: 3704 PRNG state recovery was possible based on the output of 3705 one PRNG request appropriately sized to gain knowledge on 3706 'md' followed by enough consecutive 1-byte PRNG requests 3707 to traverse all of 'state'. 3708 3709 1. When updating 'md_local' (the current thread's copy of 'md') 3710 during PRNG output generation, hash all of the previous 3711 'md_local' value, not just the half used for PRNG output. 3712 3713 2. Make the number of bytes from 'state' included into the hash 3714 independent from the number of PRNG bytes requested. 3715 3716 The first measure alone would be sufficient to avoid 3717 Markku-Juhani's attack. (Actually it had never occurred 3718 to me that the half of 'md_local' used for chaining was the 3719 half from which PRNG output bytes were taken -- I had always 3720 assumed that the secret half would be used.) The second 3721 measure makes sure that additional data from 'state' is never 3722 mixed into 'md_local' in small portions; this heuristically 3723 further strengthens the PRNG. 3724 [Bodo Moeller] 3725 3726 *) Fix crypto/bn/asm/mips3.s. 3727 [Andy Polyakov] 3728 3729 *) When only the key is given to "enc", the IV is undefined. Print out 3730 an error message in this case. 3731 [Lutz Jaenicke] 3732 3733 *) Handle special case when X509_NAME is empty in X509 printing routines. 3734 [Steve Henson] 3735 3736 *) In dsa_do_verify (crypto/dsa/dsa_ossl.c), verify that r and s are 3737 positive and less than q. 3738 [Bodo Moeller] 3739 3740 *) Don't change *pointer in CRYPTO_add_lock() is add_lock_callback is 3741 used: it isn't thread safe and the add_lock_callback should handle 3742 that itself. 3743 [Paul Rose <Paul.Rose@bridge.com>] 3744 3745 *) Verify that incoming data obeys the block size in 3746 ssl3_enc (ssl/s3_enc.c) and tls1_enc (ssl/t1_enc.c). 3747 [Bodo Moeller] 3748 3749 *) Fix OAEP check. 3750 [Ulf M�ller, Bodo M�ller] 3751 3752 *) The countermeasure against Bleichbacher's attack on PKCS #1 v1.5 3753 RSA encryption was accidentally removed in s3_srvr.c in OpenSSL 0.9.5 3754 when fixing the server behaviour for backwards-compatible 'client 3755 hello' messages. (Note that the attack is impractical against 3756 SSL 3.0 and TLS 1.0 anyway because length and version checking 3757 means that the probability of guessing a valid ciphertext is 3758 around 2^-40; see section 5 in Bleichenbacher's CRYPTO '98 3759 paper.) 3760 3761 Before 0.9.5, the countermeasure (hide the error by generating a 3762 random 'decryption result') did not work properly because 3763 ERR_clear_error() was missing, meaning that SSL_get_error() would 3764 detect the supposedly ignored error. 3765 3766 Both problems are now fixed. 3767 [Bodo Moeller] 3768 3769 *) In crypto/bio/bf_buff.c, increase DEFAULT_BUFFER_SIZE to 4096 3770 (previously it was 1024). 3771 [Bodo Moeller] 3772 3773 *) Fix for compatibility mode trust settings: ignore trust settings 3774 unless some valid trust or reject settings are present. 3775 [Steve Henson] 3776 3777 *) Fix for blowfish EVP: its a variable length cipher. 3778 [Steve Henson] 3779 3780 *) Fix various bugs related to DSA S/MIME verification. Handle missing 3781 parameters in DSA public key structures and return an error in the 3782 DSA routines if parameters are absent. 3783 [Steve Henson] 3784 3785 *) In versions up to 0.9.6, RAND_file_name() resorted to file ".rnd" 3786 in the current directory if neither $RANDFILE nor $HOME was set. 3787 RAND_file_name() in 0.9.6a returned NULL in this case. This has 3788 caused some confusion to Windows users who haven't defined $HOME. 3789 Thus RAND_file_name() is changed again: e_os.h can define a 3790 DEFAULT_HOME, which will be used if $HOME is not set. 3791 For Windows, we use "C:"; on other platforms, we still require 3792 environment variables. 3793 3794 *) Move 'if (!initialized) RAND_poll()' into regions protected by 3795 CRYPTO_LOCK_RAND. This is not strictly necessary, but avoids 3796 having multiple threads call RAND_poll() concurrently. 3797 [Bodo Moeller] 3798 3799 *) In crypto/rand/md_rand.c, replace 'add_do_not_lock' flag by a 3800 combination of a flag and a thread ID variable. 3801 Otherwise while one thread is in ssleay_rand_bytes (which sets the 3802 flag), *other* threads can enter ssleay_add_bytes without obeying 3803 the CRYPTO_LOCK_RAND lock (and may even illegally release the lock 3804 that they do not hold after the first thread unsets add_do_not_lock). 3805 [Bodo Moeller] 3806 3807 *) Change bctest again: '-x' expressions are not available in all 3808 versions of 'test'. 3809 [Bodo Moeller] 3810 3811 Changes between 0.9.6 and 0.9.6a [5 Apr 2001] 3812 3813 *) Fix a couple of memory leaks in PKCS7_dataDecode() 3814 [Steve Henson, reported by Heyun Zheng <hzheng@atdsprint.com>] 3815 3816 *) Change Configure and Makefiles to provide EXE_EXT, which will contain 3817 the default extension for executables, if any. Also, make the perl 3818 scripts that use symlink() to test if it really exists and use "cp" 3819 if it doesn't. All this made OpenSSL compilable and installable in 3820 CygWin. 3821 [Richard Levitte] 3822 3823 *) Fix for asn1_GetSequence() for indefinite length constructed data. 3824 If SEQUENCE is length is indefinite just set c->slen to the total 3825 amount of data available. 3826 [Steve Henson, reported by shige@FreeBSD.org] 3827 [This change does not apply to 0.9.7.] 3828 3829 *) Change bctest to avoid here-documents inside command substitution 3830 (workaround for FreeBSD /bin/sh bug). 3831 For compatibility with Ultrix, avoid shell functions (introduced 3832 in the bctest version that searches along $PATH). 3833 [Bodo Moeller] 3834 3835 *) Rename 'des_encrypt' to 'des_encrypt1'. This avoids the clashes 3836 with des_encrypt() defined on some operating systems, like Solaris 3837 and UnixWare. 3838 [Richard Levitte] 3839 3840 *) Check the result of RSA-CRT (see D. Boneh, R. DeMillo, R. Lipton: 3841 On the Importance of Eliminating Errors in Cryptographic 3842 Computations, J. Cryptology 14 (2001) 2, 101-119, 3843 http://theory.stanford.edu/~dabo/papers/faults.ps.gz). 3844 [Ulf Moeller] 3845 3846 *) MIPS assembler BIGNUM division bug fix. 3847 [Andy Polyakov] 3848 3849 *) Disabled incorrect Alpha assembler code. 3850 [Richard Levitte] 3851 3852 *) Fix PKCS#7 decode routines so they correctly update the length 3853 after reading an EOC for the EXPLICIT tag. 3854 [Steve Henson] 3855 [This change does not apply to 0.9.7.] 3856 3857 *) Fix bug in PKCS#12 key generation routines. This was triggered 3858 if a 3DES key was generated with a 0 initial byte. Include 3859 PKCS12_BROKEN_KEYGEN compilation option to retain the old 3860 (but broken) behaviour. 3861 [Steve Henson] 3862 3863 *) Enhance bctest to search for a working bc along $PATH and print 3864 it when found. 3865 [Tim Rice <tim@multitalents.net> via Richard Levitte] 3866 3867 *) Fix memory leaks in err.c: free err_data string if necessary; 3868 don't write to the wrong index in ERR_set_error_data. 3869 [Bodo Moeller] 3870 3871 *) Implement ssl23_peek (analogous to ssl23_read), which previously 3872 did not exist. 3873 [Bodo Moeller] 3874 3875 *) Replace rdtsc with _emit statements for VC++ version 5. 3876 [Jeremy Cooper <jeremy@baymoo.org>] 3877 3878 *) Make it possible to reuse SSLv2 sessions. 3879 [Richard Levitte] 3880 3881 *) In copy_email() check for >= 0 as a return value for 3882 X509_NAME_get_index_by_NID() since 0 is a valid index. 3883 [Steve Henson reported by Massimiliano Pala <madwolf@opensca.org>] 3884 3885 *) Avoid coredump with unsupported or invalid public keys by checking if 3886 X509_get_pubkey() fails in PKCS7_verify(). Fix memory leak when 3887 PKCS7_verify() fails with non detached data. 3888 [Steve Henson] 3889 3890 *) Don't use getenv in library functions when run as setuid/setgid. 3891 New function OPENSSL_issetugid(). 3892 [Ulf Moeller] 3893 3894 *) Avoid false positives in memory leak detection code (crypto/mem_dbg.c) 3895 due to incorrect handling of multi-threading: 3896 3897 1. Fix timing glitch in the MemCheck_off() portion of CRYPTO_mem_ctrl(). 3898 3899 2. Fix logical glitch in is_MemCheck_on() aka CRYPTO_is_mem_check_on(). 3900 3901 3. Count how many times MemCheck_off() has been called so that 3902 nested use can be treated correctly. This also avoids 3903 inband-signalling in the previous code (which relied on the 3904 assumption that thread ID 0 is impossible). 3905 [Bodo Moeller] 3906 3907 *) Add "-rand" option also to s_client and s_server. 3908 [Lutz Jaenicke] 3909 3910 *) Fix CPU detection on Irix 6.x. 3911 [Kurt Hockenbury <khockenb@stevens-tech.edu> and 3912 "Bruce W. Forsberg" <bruce.forsberg@baesystems.com>] 3913 3914 *) Fix X509_NAME bug which produced incorrect encoding if X509_NAME 3915 was empty. 3916 [Steve Henson] 3917 [This change does not apply to 0.9.7.] 3918 3919 *) Use the cached encoding of an X509_NAME structure rather than 3920 copying it. This is apparently the reason for the libsafe "errors" 3921 but the code is actually correct. 3922 [Steve Henson] 3923 3924 *) Add new function BN_rand_range(), and fix DSA_sign_setup() to prevent 3925 Bleichenbacher's DSA attack. 3926 Extend BN_[pseudo_]rand: As before, top=1 forces the highest two bits 3927 to be set and top=0 forces the highest bit to be set; top=-1 is new 3928 and leaves the highest bit random. 3929 [Ulf Moeller, Bodo Moeller] 3930 3931 *) In the NCONF_...-based implementations for CONF_... queries 3932 (crypto/conf/conf_lib.c), if the input LHASH is NULL, avoid using 3933 a temporary CONF structure with the data component set to NULL 3934 (which gives segmentation faults in lh_retrieve). 3935 Instead, use NULL for the CONF pointer in CONF_get_string and 3936 CONF_get_number (which may use environment variables) and directly 3937 return NULL from CONF_get_section. 3938 [Bodo Moeller] 3939 3940 *) Fix potential buffer overrun for EBCDIC. 3941 [Ulf Moeller] 3942 3943 *) Tolerate nonRepudiation as being valid for S/MIME signing and certSign 3944 keyUsage if basicConstraints absent for a CA. 3945 [Steve Henson] 3946 3947 *) Make SMIME_write_PKCS7() write mail header values with a format that 3948 is more generally accepted (no spaces before the semicolon), since 3949 some programs can't parse those values properly otherwise. Also make 3950 sure BIO's that break lines after each write do not create invalid 3951 headers. 3952 [Richard Levitte] 3953 3954 *) Make the CRL encoding routines work with empty SEQUENCE OF. The 3955 macros previously used would not encode an empty SEQUENCE OF 3956 and break the signature. 3957 [Steve Henson] 3958 [This change does not apply to 0.9.7.] 3959 3960 *) Zero the premaster secret after deriving the master secret in 3961 DH ciphersuites. 3962 [Steve Henson] 3963 3964 *) Add some EVP_add_digest_alias registrations (as found in 3965 OpenSSL_add_all_digests()) to SSL_library_init() 3966 aka OpenSSL_add_ssl_algorithms(). This provides improved 3967 compatibility with peers using X.509 certificates 3968 with unconventional AlgorithmIdentifier OIDs. 3969 [Bodo Moeller] 3970 3971 *) Fix for Irix with NO_ASM. 3972 ["Bruce W. Forsberg" <bruce.forsberg@baesystems.com>] 3973 3974 *) ./config script fixes. 3975 [Ulf Moeller, Richard Levitte] 3976 3977 *) Fix 'openssl passwd -1'. 3978 [Bodo Moeller] 3979 3980 *) Change PKCS12_key_gen_asc() so it can cope with non null 3981 terminated strings whose length is passed in the passlen 3982 parameter, for example from PEM callbacks. This was done 3983 by adding an extra length parameter to asc2uni(). 3984 [Steve Henson, reported by <oddissey@samsung.co.kr>] 3985 3986 *) Fix C code generated by 'openssl dsaparam -C': If a BN_bin2bn 3987 call failed, free the DSA structure. 3988 [Bodo Moeller] 3989 3990 *) Fix to uni2asc() to cope with zero length Unicode strings. 3991 These are present in some PKCS#12 files. 3992 [Steve Henson] 3993 3994 *) Increase s2->wbuf allocation by one byte in ssl2_new (ssl/s2_lib.c). 3995 Otherwise do_ssl_write (ssl/s2_pkt.c) will write beyond buffer limits 3996 when writing a 32767 byte record. 3997 [Bodo Moeller; problem reported by Eric Day <eday@concentric.net>] 3998 3999 *) In RSA_eay_public_{en,ed}crypt and RSA_eay_mod_exp (rsa_eay.c), 4000 obtain lock CRYPTO_LOCK_RSA before setting rsa->_method_mod_{n,p,q}. 4001 4002 (RSA objects have a reference count access to which is protected 4003 by CRYPTO_LOCK_RSA [see rsa_lib.c, s3_srvr.c, ssl_cert.c, ssl_rsa.c], 4004 so they are meant to be shared between threads.) 4005 [Bodo Moeller, Geoff Thorpe; original patch submitted by 4006 "Reddie, Steven" <Steven.Reddie@ca.com>] 4007 4008 *) Fix a deadlock in CRYPTO_mem_leaks(). 4009 [Bodo Moeller] 4010 4011 *) Use better test patterns in bntest. 4012 [Ulf M�ller] 4013 4014 *) rand_win.c fix for Borland C. 4015 [Ulf M�ller] 4016 4017 *) BN_rshift bugfix for n == 0. 4018 [Bodo Moeller] 4019 4020 *) Add a 'bctest' script that checks for some known 'bc' bugs 4021 so that 'make test' does not abort just because 'bc' is broken. 4022 [Bodo Moeller] 4023 4024 *) Store verify_result within SSL_SESSION also for client side to 4025 avoid potential security hole. (Re-used sessions on the client side 4026 always resulted in verify_result==X509_V_OK, not using the original 4027 result of the server certificate verification.) 4028 [Lutz Jaenicke] 4029 4030 *) Fix ssl3_pending: If the record in s->s3->rrec is not of type 4031 SSL3_RT_APPLICATION_DATA, return 0. 4032 Similarly, change ssl2_pending to return 0 if SSL_in_init(s) is true. 4033 [Bodo Moeller] 4034 4035 *) Fix SSL_peek: 4036 Both ssl2_peek and ssl3_peek, which were totally broken in earlier 4037 releases, have been re-implemented by renaming the previous 4038 implementations of ssl2_read and ssl3_read to ssl2_read_internal 4039 and ssl3_read_internal, respectively, and adding 'peek' parameters 4040 to them. The new ssl[23]_{read,peek} functions are calls to 4041 ssl[23]_read_internal with the 'peek' flag set appropriately. 4042 A 'peek' parameter has also been added to ssl3_read_bytes, which 4043 does the actual work for ssl3_read_internal. 4044 [Bodo Moeller] 4045 4046 *) Initialise "ex_data" member of RSA/DSA/DH structures prior to calling 4047 the method-specific "init()" handler. Also clean up ex_data after 4048 calling the method-specific "finish()" handler. Previously, this was 4049 happening the other way round. 4050 [Geoff Thorpe] 4051 4052 *) Increase BN_CTX_NUM (the number of BIGNUMs in a BN_CTX) to 16. 4053 The previous value, 12, was not always sufficient for BN_mod_exp(). 4054 [Bodo Moeller] 4055 4056 *) Make sure that shared libraries get the internal name engine with 4057 the full version number and not just 0. This should mark the 4058 shared libraries as not backward compatible. Of course, this should 4059 be changed again when we can guarantee backward binary compatibility. 4060 [Richard Levitte] 4061 4062 *) Fix typo in get_cert_by_subject() in by_dir.c 4063 [Jean-Marc Desperrier <jean-marc.desperrier@certplus.com>] 4064 4065 *) Rework the system to generate shared libraries: 4066 4067 - Make note of the expected extension for the shared libraries and 4068 if there is a need for symbolic links from for example libcrypto.so.0 4069 to libcrypto.so.0.9.7. There is extended info in Configure for 4070 that. 4071 4072 - Make as few rebuilds of the shared libraries as possible. 4073 4074 - Still avoid linking the OpenSSL programs with the shared libraries. 4075 4076 - When installing, install the shared libraries separately from the 4077 static ones. 4078 [Richard Levitte] 4079 4080 *) Fix SSL_CTX_set_read_ahead macro to actually use its argument. 4081 4082 Copy SSL_CTX's read_ahead flag to SSL object directly in SSL_new 4083 and not in SSL_clear because the latter is also used by the 4084 accept/connect functions; previously, the settings made by 4085 SSL_set_read_ahead would be lost during the handshake. 4086 [Bodo Moeller; problems reported by Anders Gertz <gertz@epact.se>] 4087 4088 *) Correct util/mkdef.pl to be selective about disabled algorithms. 4089 Previously, it would create entries for disableed algorithms no 4090 matter what. 4091 [Richard Levitte] 4092 4093 *) Added several new manual pages for SSL_* function. 4094 [Lutz Jaenicke] 4095 4096 Changes between 0.9.5a and 0.9.6 [24 Sep 2000] 4097 4098 *) In ssl23_get_client_hello, generate an error message when faced 4099 with an initial SSL 3.0/TLS record that is too small to contain the 4100 first two bytes of the ClientHello message, i.e. client_version. 4101 (Note that this is a pathologic case that probably has never happened 4102 in real life.) The previous approach was to use the version number 4103 from the record header as a substitute; but our protocol choice 4104 should not depend on that one because it is not authenticated 4105 by the Finished messages. 4106 [Bodo Moeller] 4107 4108 *) More robust randomness gathering functions for Windows. 4109 [Jeffrey Altman <jaltman@columbia.edu>] 4110 4111 *) For compatibility reasons if the flag X509_V_FLAG_ISSUER_CHECK is 4112 not set then we don't setup the error code for issuer check errors 4113 to avoid possibly overwriting other errors which the callback does 4114 handle. If an application does set the flag then we assume it knows 4115 what it is doing and can handle the new informational codes 4116 appropriately. 4117 [Steve Henson] 4118 4119 *) Fix for a nasty bug in ASN1_TYPE handling. ASN1_TYPE is used for 4120 a general "ANY" type, as such it should be able to decode anything 4121 including tagged types. However it didn't check the class so it would 4122 wrongly interpret tagged types in the same way as their universal 4123 counterpart and unknown types were just rejected. Changed so that the 4124 tagged and unknown types are handled in the same way as a SEQUENCE: 4125 that is the encoding is stored intact. There is also a new type 4126 "V_ASN1_OTHER" which is used when the class is not universal, in this 4127 case we have no idea what the actual type is so we just lump them all 4128 together. 4129 [Steve Henson] 4130 4131 *) On VMS, stdout may very well lead to a file that is written to 4132 in a record-oriented fashion. That means that every write() will 4133 write a separate record, which will be read separately by the 4134 programs trying to read from it. This can be very confusing. 4135 4136 The solution is to put a BIO filter in the way that will buffer 4137 text until a linefeed is reached, and then write everything a 4138 line at a time, so every record written will be an actual line, 4139 not chunks of lines and not (usually doesn't happen, but I've 4140 seen it once) several lines in one record. BIO_f_linebuffer() is 4141 the answer. 4142 4143 Currently, it's a VMS-only method, because that's where it has 4144 been tested well enough. 4145 [Richard Levitte] 4146 4147 *) Remove 'optimized' squaring variant in BN_mod_mul_montgomery, 4148 it can return incorrect results. 4149 (Note: The buggy variant was not enabled in OpenSSL 0.9.5a, 4150 but it was in 0.9.6-beta[12].) 4151 [Bodo Moeller] 4152 4153 *) Disable the check for content being present when verifying detached 4154 signatures in pk7_smime.c. Some versions of Netscape (wrongly) 4155 include zero length content when signing messages. 4156 [Steve Henson] 4157 4158 *) New BIO_shutdown_wr macro, which invokes the BIO_C_SHUTDOWN_WR 4159 BIO_ctrl (for BIO pairs). 4160 [Bodo M�ller] 4161 4162 *) Add DSO method for VMS. 4163 [Richard Levitte] 4164 4165 *) Bug fix: Montgomery multiplication could produce results with the 4166 wrong sign. 4167 [Ulf M�ller] 4168 4169 *) Add RPM specification openssl.spec and modify it to build three 4170 packages. The default package contains applications, application 4171 documentation and run-time libraries. The devel package contains 4172 include files, static libraries and function documentation. The 4173 doc package contains the contents of the doc directory. The original 4174 openssl.spec was provided by Damien Miller <djm@mindrot.org>. 4175 [Richard Levitte] 4176 4177 *) Add a large number of documentation files for many SSL routines. 4178 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>] 4179 4180 *) Add a configuration entry for Sony News 4. 4181 [NAKAJI Hiroyuki <nakaji@tutrp.tut.ac.jp>] 4182 4183 *) Don't set the two most significant bits to one when generating a 4184 random number < q in the DSA library. 4185 [Ulf M�ller] 4186 4187 *) New SSL API mode 'SSL_MODE_AUTO_RETRY'. This disables the default 4188 behaviour that SSL_read may result in SSL_ERROR_WANT_READ (even if 4189 the underlying transport is blocking) if a handshake took place. 4190 (The default behaviour is needed by applications such as s_client 4191 and s_server that use select() to determine when to use SSL_read; 4192 but for applications that know in advance when to expect data, it 4193 just makes things more complicated.) 4194 [Bodo Moeller] 4195 4196 *) Add RAND_egd_bytes(), which gives control over the number of bytes read 4197 from EGD. 4198 [Ben Laurie] 4199 4200 *) Add a few more EBCDIC conditionals that make `req' and `x509' 4201 work better on such systems. 4202 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>] 4203 4204 *) Add two demo programs for PKCS12_parse() and PKCS12_create(). 4205 Update PKCS12_parse() so it copies the friendlyName and the 4206 keyid to the certificates aux info. 4207 [Steve Henson] 4208 4209 *) Fix bug in PKCS7_verify() which caused an infinite loop 4210 if there was more than one signature. 4211 [Sven Uszpelkat <su@celocom.de>] 4212 4213 *) Major change in util/mkdef.pl to include extra information 4214 about each symbol, as well as presentig variables as well 4215 as functions. This change means that there's n more need 4216 to rebuild the .num files when some algorithms are excluded. 4217 [Richard Levitte] 4218 4219 *) Allow the verify time to be set by an application, 4220 rather than always using the current time. 4221 [Steve Henson] 4222 4223 *) Phase 2 verify code reorganisation. The certificate 4224 verify code now looks up an issuer certificate by a 4225 number of criteria: subject name, authority key id 4226 and key usage. It also verifies self signed certificates 4227 by the same criteria. The main comparison function is 4228 X509_check_issued() which performs these checks. 4229 4230 Lot of changes were necessary in order to support this 4231 without completely rewriting the lookup code. 4232 4233 Authority and subject key identifier are now cached. 4234 4235 The LHASH 'certs' is X509_STORE has now been replaced 4236 by a STACK_OF(X509_OBJECT). This is mainly because an 4237 LHASH can't store or retrieve multiple objects with 4238 the same hash value. 4239 4240 As a result various functions (which were all internal 4241 use only) have changed to handle the new X509_STORE 4242 structure. This will break anything that messed round 4243 with X509_STORE internally. 4244 4245 The functions X509_STORE_add_cert() now checks for an 4246 exact match, rather than just subject name. 4247 4248 The X509_STORE API doesn't directly support the retrieval 4249 of multiple certificates matching a given criteria, however 4250 this can be worked round by performing a lookup first 4251 (which will fill the cache with candidate certificates) 4252 and then examining the cache for matches. This is probably 4253 the best we can do without throwing out X509_LOOKUP 4254 entirely (maybe later...). 4255 4256 The X509_VERIFY_CTX structure has been enhanced considerably. 4257 4258 All certificate lookup operations now go via a get_issuer() 4259 callback. Although this currently uses an X509_STORE it 4260 can be replaced by custom lookups. This is a simple way 4261 to bypass the X509_STORE hackery necessary to make this 4262 work and makes it possible to use more efficient techniques 4263 in future. A very simple version which uses a simple 4264 STACK for its trusted certificate store is also provided 4265 using X509_STORE_CTX_trusted_stack(). 4266 4267 The verify_cb() and verify() callbacks now have equivalents 4268 in the X509_STORE_CTX structure. 4269 4270 X509_STORE_CTX also has a 'flags' field which can be used 4271 to customise the verify behaviour. 4272 [Steve Henson] 4273 4274 *) Add new PKCS#7 signing option PKCS7_NOSMIMECAP which 4275 excludes S/MIME capabilities. 4276 [Steve Henson] 4277 4278 *) When a certificate request is read in keep a copy of the 4279 original encoding of the signed data and use it when outputing 4280 again. Signatures then use the original encoding rather than 4281 a decoded, encoded version which may cause problems if the 4282 request is improperly encoded. 4283 [Steve Henson] 4284 4285 *) For consistency with other BIO_puts implementations, call 4286 buffer_write(b, ...) directly in buffer_puts instead of calling 4287 BIO_write(b, ...). 4288 4289 In BIO_puts, increment b->num_write as in BIO_write. 4290 [Peter.Sylvester@EdelWeb.fr] 4291 4292 *) Fix BN_mul_word for the case where the word is 0. (We have to use 4293 BN_zero, we may not return a BIGNUM with an array consisting of 4294 words set to zero.) 4295 [Bodo Moeller] 4296 4297 *) Avoid calling abort() from within the library when problems are 4298 detected, except if preprocessor symbols have been defined 4299 (such as REF_CHECK, BN_DEBUG etc.). 4300 [Bodo Moeller] 4301 4302 *) New openssl application 'rsautl'. This utility can be 4303 used for low level RSA operations. DER public key 4304 BIO/fp routines also added. 4305 [Steve Henson] 4306 4307 *) New Configure entry and patches for compiling on QNX 4. 4308 [Andreas Schneider <andreas@ds3.etech.fh-hamburg.de>] 4309 4310 *) A demo state-machine implementation was sponsored by 4311 Nuron (http://www.nuron.com/) and is now available in 4312 demos/state_machine. 4313 [Ben Laurie] 4314 4315 *) New options added to the 'dgst' utility for signature 4316 generation and verification. 4317 [Steve Henson] 4318 4319 *) Unrecognized PKCS#7 content types are now handled via a 4320 catch all ASN1_TYPE structure. This allows unsupported 4321 types to be stored as a "blob" and an application can 4322 encode and decode it manually. 4323 [Steve Henson] 4324 4325 *) Fix various signed/unsigned issues to make a_strex.c 4326 compile under VC++. 4327 [Oscar Jacobsson <oscar.jacobsson@celocom.com>] 4328 4329 *) ASN1 fixes. i2d_ASN1_OBJECT was not returning the correct 4330 length if passed a buffer. ASN1_INTEGER_to_BN failed 4331 if passed a NULL BN and its argument was negative. 4332 [Steve Henson, pointed out by Sven Heiberg <sven@tartu.cyber.ee>] 4333 4334 *) Modification to PKCS#7 encoding routines to output definite 4335 length encoding. Since currently the whole structures are in 4336 memory there's not real point in using indefinite length 4337 constructed encoding. However if OpenSSL is compiled with 4338 the flag PKCS7_INDEFINITE_ENCODING the old form is used. 4339 [Steve Henson] 4340 4341 *) Added BIO_vprintf() and BIO_vsnprintf(). 4342 [Richard Levitte] 4343 4344 *) Added more prefixes to parse for in the the strings written 4345 through a logging bio, to cover all the levels that are available 4346 through syslog. The prefixes are now: 4347 4348 PANIC, EMERG, EMR => LOG_EMERG 4349 ALERT, ALR => LOG_ALERT 4350 CRIT, CRI => LOG_CRIT 4351 ERROR, ERR => LOG_ERR 4352 WARNING, WARN, WAR => LOG_WARNING 4353 NOTICE, NOTE, NOT => LOG_NOTICE 4354 INFO, INF => LOG_INFO 4355 DEBUG, DBG => LOG_DEBUG 4356 4357 and as before, if none of those prefixes are present at the 4358 beginning of the string, LOG_ERR is chosen. 4359 4360 On Win32, the LOG_* levels are mapped according to this: 4361 4362 LOG_EMERG, LOG_ALERT, LOG_CRIT, LOG_ERR => EVENTLOG_ERROR_TYPE 4363 LOG_WARNING => EVENTLOG_WARNING_TYPE 4364 LOG_NOTICE, LOG_INFO, LOG_DEBUG => EVENTLOG_INFORMATION_TYPE 4365 4366 [Richard Levitte] 4367 4368 *) Made it possible to reconfigure with just the configuration 4369 argument "reconf" or "reconfigure". The command line arguments 4370 are stored in Makefile.ssl in the variable CONFIGURE_ARGS, 4371 and are retrieved from there when reconfiguring. 4372 [Richard Levitte] 4373 4374 *) MD4 implemented. 4375 [Assar Westerlund <assar@sics.se>, Richard Levitte] 4376 4377 *) Add the arguments -CAfile and -CApath to the pkcs12 utility. 4378 [Richard Levitte] 4379 4380 *) The obj_dat.pl script was messing up the sorting of object 4381 names. The reason was that it compared the quoted version 4382 of strings as a result "OCSP" > "OCSP Signing" because 4383 " > SPACE. Changed script to store unquoted versions of 4384 names and add quotes on output. It was also omitting some 4385 names from the lookup table if they were given a default 4386 value (that is if SN is missing it is given the same 4387 value as LN and vice versa), these are now added on the 4388 grounds that if an object has a name we should be able to 4389 look it up. Finally added warning output when duplicate 4390 short or long names are found. 4391 [Steve Henson] 4392 4393 *) Changes needed for Tandem NSK. 4394 [Scott Uroff <scott@xypro.com>] 4395 4396 *) Fix SSL 2.0 rollback checking: Due to an off-by-one error in 4397 RSA_padding_check_SSLv23(), special padding was never detected 4398 and thus the SSL 3.0/TLS 1.0 countermeasure against protocol 4399 version rollback attacks was not effective. 4400 4401 In s23_clnt.c, don't use special rollback-attack detection padding 4402 (RSA_SSLV23_PADDING) if SSL 2.0 is the only protocol enabled in the 4403 client; similarly, in s23_srvr.c, don't do the rollback check if 4404 SSL 2.0 is the only protocol enabled in the server. 4405 [Bodo Moeller] 4406 4407 *) Make it possible to get hexdumps of unprintable data with 'openssl 4408 asn1parse'. By implication, the functions ASN1_parse_dump() and 4409 BIO_dump_indent() are added. 4410 [Richard Levitte] 4411 4412 *) New functions ASN1_STRING_print_ex() and X509_NAME_print_ex() 4413 these print out strings and name structures based on various 4414 flags including RFC2253 support and proper handling of 4415 multibyte characters. Added options to the 'x509' utility 4416 to allow the various flags to be set. 4417 [Steve Henson] 4418 4419 *) Various fixes to use ASN1_TIME instead of ASN1_UTCTIME. 4420 Also change the functions X509_cmp_current_time() and 4421 X509_gmtime_adj() work with an ASN1_TIME structure, 4422 this will enable certificates using GeneralizedTime in validity 4423 dates to be checked. 4424 [Steve Henson] 4425 4426 *) Make the NEG_PUBKEY_BUG code (which tolerates invalid 4427 negative public key encodings) on by default, 4428 NO_NEG_PUBKEY_BUG can be set to disable it. 4429 [Steve Henson] 4430 4431 *) New function c2i_ASN1_OBJECT() which acts on ASN1_OBJECT 4432 content octets. An i2c_ASN1_OBJECT is unnecessary because 4433 the encoding can be trivially obtained from the structure. 4434 [Steve Henson] 4435 4436 *) crypto/err.c locking bugfix: Use write locks (CRYPTO_w_[un]lock), 4437 not read locks (CRYPTO_r_[un]lock). 4438 [Bodo Moeller] 4439 4440 *) A first attempt at creating official support for shared 4441 libraries through configuration. I've kept it so the 4442 default is static libraries only, and the OpenSSL programs 4443 are always statically linked for now, but there are 4444 preparations for dynamic linking in place. 4445 This has been tested on Linux and Tru64. 4446 [Richard Levitte] 4447 4448 *) Randomness polling function for Win9x, as described in: 4449 Peter Gutmann, Software Generation of Practically Strong 4450 Random Numbers. 4451 [Ulf M�ller] 4452 4453 *) Fix so PRNG is seeded in req if using an already existing 4454 DSA key. 4455 [Steve Henson] 4456 4457 *) New options to smime application. -inform and -outform 4458 allow alternative formats for the S/MIME message including 4459 PEM and DER. The -content option allows the content to be 4460 specified separately. This should allow things like Netscape 4461 form signing output easier to verify. 4462 [Steve Henson] 4463 4464 *) Fix the ASN1 encoding of tags using the 'long form'. 4465 [Steve Henson] 4466 4467 *) New ASN1 functions, i2c_* and c2i_* for INTEGER and BIT 4468 STRING types. These convert content octets to and from the 4469 underlying type. The actual tag and length octets are 4470 already assumed to have been read in and checked. These 4471 are needed because all other string types have virtually 4472 identical handling apart from the tag. By having versions 4473 of the ASN1 functions that just operate on content octets 4474 IMPLICIT tagging can be handled properly. It also allows 4475 the ASN1_ENUMERATED code to be cut down because ASN1_ENUMERATED 4476 and ASN1_INTEGER are identical apart from the tag. 4477 [Steve Henson] 4478 4479 *) Change the handling of OID objects as follows: 4480 4481 - New object identifiers are inserted in objects.txt, following 4482 the syntax given in objects.README. 4483 - objects.pl is used to process obj_mac.num and create a new 4484 obj_mac.h. 4485 - obj_dat.pl is used to create a new obj_dat.h, using the data in 4486 obj_mac.h. 4487 4488 This is currently kind of a hack, and the perl code in objects.pl 4489 isn't very elegant, but it works as I intended. The simplest way 4490 to check that it worked correctly is to look in obj_dat.h and 4491 check the array nid_objs and make sure the objects haven't moved 4492 around (this is important!). Additions are OK, as well as 4493 consistent name changes. 4494 [Richard Levitte] 4495 4496 *) Add BSD-style MD5-based passwords to 'openssl passwd' (option '-1'). 4497 [Bodo Moeller] 4498 4499 *) Addition of the command line parameter '-rand file' to 'openssl req'. 4500 The given file adds to whatever has already been seeded into the 4501 random pool through the RANDFILE configuration file option or 4502 environment variable, or the default random state file. 4503 [Richard Levitte] 4504 4505 *) mkstack.pl now sorts each macro group into lexical order. 4506 Previously the output order depended on the order the files 4507 appeared in the directory, resulting in needless rewriting 4508 of safestack.h . 4509 [Steve Henson] 4510 4511 *) Patches to make OpenSSL compile under Win32 again. Mostly 4512 work arounds for the VC++ problem that it treats func() as 4513 func(void). Also stripped out the parts of mkdef.pl that 4514 added extra typesafe functions: these no longer exist. 4515 [Steve Henson] 4516 4517 *) Reorganisation of the stack code. The macros are now all 4518 collected in safestack.h . Each macro is defined in terms of 4519 a "stack macro" of the form SKM_<name>(type, a, b). The 4520 DEBUG_SAFESTACK is now handled in terms of function casts, 4521 this has the advantage of retaining type safety without the 4522 use of additional functions. If DEBUG_SAFESTACK is not defined 4523 then the non typesafe macros are used instead. Also modified the 4524 mkstack.pl script to handle the new form. Needs testing to see 4525 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK 4526 the default if no major problems. Similar behaviour for ASN1_SET_OF 4527 and PKCS12_STACK_OF. 4528 [Steve Henson] 4529 4530 *) When some versions of IIS use the 'NET' form of private key the 4531 key derivation algorithm is different. Normally MD5(password) is 4532 used as a 128 bit RC4 key. In the modified case 4533 MD5(MD5(password) + "SGCKEYSALT") is used insted. Added some 4534 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same 4535 as the old Netscape_RSA functions except they have an additional 4536 'sgckey' parameter which uses the modified algorithm. Also added 4537 an -sgckey command line option to the rsa utility. Thanks to 4538 Adrian Peck <bertie@ncipher.com> for posting details of the modified 4539 algorithm to openssl-dev. 4540 [Steve Henson] 4541 4542 *) The evp_local.h macros were using 'c.##kname' which resulted in 4543 invalid expansion on some systems (SCO 5.0.5 for example). 4544 Corrected to 'c.kname'. 4545 [Phillip Porch <root@theporch.com>] 4546 4547 *) New X509_get1_email() and X509_REQ_get1_email() functions that return 4548 a STACK of email addresses from a certificate or request, these look 4549 in the subject name and the subject alternative name extensions and 4550 omit any duplicate addresses. 4551 [Steve Henson] 4552 4553 *) Re-implement BN_mod_exp2_mont using independent (and larger) windows. 4554 This makes DSA verification about 2 % faster. 4555 [Bodo Moeller] 4556 4557 *) Increase maximum window size in BN_mod_exp_... to 6 bits instead of 5 4558 (meaning that now 2^5 values will be precomputed, which is only 4 KB 4559 plus overhead for 1024 bit moduli). 4560 This makes exponentiations about 0.5 % faster for 1024 bit 4561 exponents (as measured by "openssl speed rsa2048"). 4562 [Bodo Moeller] 4563 4564 *) Rename memory handling macros to avoid conflicts with other 4565 software: 4566 Malloc => OPENSSL_malloc 4567 Malloc_locked => OPENSSL_malloc_locked 4568 Realloc => OPENSSL_realloc 4569 Free => OPENSSL_free 4570 [Richard Levitte] 4571 4572 *) New function BN_mod_exp_mont_word for small bases (roughly 15% 4573 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange). 4574 [Bodo Moeller] 4575 4576 *) CygWin32 support. 4577 [John Jarvie <jjarvie@newsguy.com>] 4578 4579 *) The type-safe stack code has been rejigged. It is now only compiled 4580 in when OpenSSL is configured with the DEBUG_SAFESTACK option and 4581 by default all type-specific stack functions are "#define"d back to 4582 standard stack functions. This results in more streamlined output 4583 but retains the type-safety checking possibilities of the original 4584 approach. 4585 [Geoff Thorpe] 4586 4587 *) The STACK code has been cleaned up, and certain type declarations 4588 that didn't make a lot of sense have been brought in line. This has 4589 also involved a cleanup of sorts in safestack.h to more correctly 4590 map type-safe stack functions onto their plain stack counterparts. 4591 This work has also resulted in a variety of "const"ifications of 4592 lots of the code, especially "_cmp" operations which should normally 4593 be prototyped with "const" parameters anyway. 4594 [Geoff Thorpe] 4595 4596 *) When generating bytes for the first time in md_rand.c, 'stir the pool' 4597 by seeding with STATE_SIZE dummy bytes (with zero entropy count). 4598 (The PRNG state consists of two parts, the large pool 'state' and 'md', 4599 where all of 'md' is used each time the PRNG is used, but 'state' 4600 is used only indexed by a cyclic counter. As entropy may not be 4601 well distributed from the beginning, 'md' is important as a 4602 chaining variable. However, the output function chains only half 4603 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains 4604 all of 'md', and seeding with STATE_SIZE dummy bytes will result 4605 in all of 'state' being rewritten, with the new values depending 4606 on virtually all of 'md'. This overcomes the 80 bit limitation.) 4607 [Bodo Moeller] 4608 4609 *) In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when 4610 the handshake is continued after ssl_verify_cert_chain(); 4611 otherwise, if SSL_VERIFY_NONE is set, remaining error codes 4612 can lead to 'unexplainable' connection aborts later. 4613 [Bodo Moeller; problem tracked down by Lutz Jaenicke] 4614 4615 *) Major EVP API cipher revision. 4616 Add hooks for extra EVP features. This allows various cipher 4617 parameters to be set in the EVP interface. Support added for variable 4618 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and 4619 setting of RC2 and RC5 parameters. 4620 4621 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length 4622 ciphers. 4623 4624 Remove lots of duplicated code from the EVP library. For example *every* 4625 cipher init() function handles the 'iv' in the same way according to the 4626 cipher mode. They also all do nothing if the 'key' parameter is NULL and 4627 for CFB and OFB modes they zero ctx->num. 4628 4629 New functionality allows removal of S/MIME code RC2 hack. 4630 4631 Most of the routines have the same form and so can be declared in terms 4632 of macros. 4633 4634 By shifting this to the top level EVP_CipherInit() it can be removed from 4635 all individual ciphers. If the cipher wants to handle IVs or keys 4636 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT 4637 flags. 4638 4639 Change lots of functions like EVP_EncryptUpdate() to now return a 4640 value: although software versions of the algorithms cannot fail 4641 any installed hardware versions can. 4642 [Steve Henson] 4643 4644 *) Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if 4645 this option is set, tolerate broken clients that send the negotiated 4646 protocol version number instead of the requested protocol version 4647 number. 4648 [Bodo Moeller] 4649 4650 *) Call dh_tmp_cb (set by ..._TMP_DH_CB) with correct 'is_export' flag; 4651 i.e. non-zero for export ciphersuites, zero otherwise. 4652 Previous versions had this flag inverted, inconsistent with 4653 rsa_tmp_cb (..._TMP_RSA_CB). 4654 [Bodo Moeller; problem reported by Amit Chopra] 4655 4656 *) Add missing DSA library text string. Work around for some IIS 4657 key files with invalid SEQUENCE encoding. 4658 [Steve Henson] 4659 4660 *) Add a document (doc/standards.txt) that list all kinds of standards 4661 and so on that are implemented in OpenSSL. 4662 [Richard Levitte] 4663 4664 *) Enhance c_rehash script. Old version would mishandle certificates 4665 with the same subject name hash and wouldn't handle CRLs at all. 4666 Added -fingerprint option to crl utility, to support new c_rehash 4667 features. 4668 [Steve Henson] 4669 4670 *) Eliminate non-ANSI declarations in crypto.h and stack.h. 4671 [Ulf M�ller] 4672 4673 *) Fix for SSL server purpose checking. Server checking was 4674 rejecting certificates which had extended key usage present 4675 but no ssl client purpose. 4676 [Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>] 4677 4678 *) Make PKCS#12 code work with no password. The PKCS#12 spec 4679 is a little unclear about how a blank password is handled. 4680 Since the password in encoded as a BMPString with terminating 4681 double NULL a zero length password would end up as just the 4682 double NULL. However no password at all is different and is 4683 handled differently in the PKCS#12 key generation code. NS 4684 treats a blank password as zero length. MSIE treats it as no 4685 password on export: but it will try both on import. We now do 4686 the same: PKCS12_parse() tries zero length and no password if 4687 the password is set to "" or NULL (NULL is now a valid password: 4688 it wasn't before) as does the pkcs12 application. 4689 [Steve Henson] 4690 4691 *) Bugfixes in apps/x509.c: Avoid a memory leak; and don't use 4692 perror when PEM_read_bio_X509_REQ fails, the error message must 4693 be obtained from the error queue. 4694 [Bodo Moeller] 4695 4696 *) Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing 4697 it in ERR_remove_state if appropriate, and change ERR_get_state 4698 accordingly to avoid race conditions (this is necessary because 4699 thread_hash is no longer constant once set). 4700 [Bodo Moeller] 4701 4702 *) Bugfix for linux-elf makefile.one. 4703 [Ulf M�ller] 4704 4705 *) RSA_get_default_method() will now cause a default 4706 RSA_METHOD to be chosen if one doesn't exist already. 4707 Previously this was only set during a call to RSA_new() 4708 or RSA_new_method(NULL) meaning it was possible for 4709 RSA_get_default_method() to return NULL. 4710 [Geoff Thorpe] 4711 4712 *) Added native name translation to the existing DSO code 4713 that will convert (if the flag to do so is set) filenames 4714 that are sufficiently small and have no path information 4715 into a canonical native form. Eg. "blah" converted to 4716 "libblah.so" or "blah.dll" etc. 4717 [Geoff Thorpe] 4718 4719 *) New function ERR_error_string_n(e, buf, len) which is like 4720 ERR_error_string(e, buf), but writes at most 'len' bytes 4721 including the 0 terminator. For ERR_error_string_n, 'buf' 4722 may not be NULL. 4723 [Damien Miller <djm@mindrot.org>, Bodo Moeller] 4724 4725 *) CONF library reworked to become more general. A new CONF 4726 configuration file reader "class" is implemented as well as a 4727 new functions (NCONF_*, for "New CONF") to handle it. The now 4728 old CONF_* functions are still there, but are reimplemented to 4729 work in terms of the new functions. Also, a set of functions 4730 to handle the internal storage of the configuration data is 4731 provided to make it easier to write new configuration file 4732 reader "classes" (I can definitely see something reading a 4733 configuration file in XML format, for example), called _CONF_*, 4734 or "the configuration storage API"... 4735 4736 The new configuration file reading functions are: 4737 4738 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio, 4739 NCONF_get_section, NCONF_get_string, NCONF_get_numbre 4740 4741 NCONF_default, NCONF_WIN32 4742 4743 NCONF_dump_fp, NCONF_dump_bio 4744 4745 NCONF_default and NCONF_WIN32 are method (or "class") choosers, 4746 NCONF_new creates a new CONF object. This works in the same way 4747 as other interfaces in OpenSSL, like the BIO interface. 4748 NCONF_dump_* dump the internal storage of the configuration file, 4749 which is useful for debugging. All other functions take the same 4750 arguments as the old CONF_* functions wth the exception of the 4751 first that must be a `CONF *' instead of a `LHASH *'. 4752 4753 To make it easer to use the new classes with the old CONF_* functions, 4754 the function CONF_set_default_method is provided. 4755 [Richard Levitte] 4756 4757 *) Add '-tls1' option to 'openssl ciphers', which was already 4758 mentioned in the documentation but had not been implemented. 4759 (This option is not yet really useful because even the additional 4760 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.) 4761 [Bodo Moeller] 4762 4763 *) Initial DSO code added into libcrypto for letting OpenSSL (and 4764 OpenSSL-based applications) load shared libraries and bind to 4765 them in a portable way. 4766 [Geoff Thorpe, with contributions from Richard Levitte] 4767 4768 Changes between 0.9.5 and 0.9.5a [1 Apr 2000] 4769 4770 *) Make sure _lrotl and _lrotr are only used with MSVC. 4771 4772 *) Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status 4773 (the default implementation of RAND_status). 4774 4775 *) Rename openssl x509 option '-crlext', which was added in 0.9.5, 4776 to '-clrext' (= clear extensions), as intended and documented. 4777 [Bodo Moeller; inconsistency pointed out by Michael Attili 4778 <attili@amaxo.com>] 4779 4780 *) Fix for HMAC. It wasn't zeroing the rest of the block if the key length 4781 was larger than the MD block size. 4782 [Steve Henson, pointed out by Yost William <YostW@tce.com>] 4783 4784 *) Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument 4785 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set() 4786 using the passed key: if the passed key was a private key the result 4787 of X509_print(), for example, would be to print out all the private key 4788 components. 4789 [Steve Henson] 4790 4791 *) des_quad_cksum() byte order bug fix. 4792 [Ulf M�ller, using the problem description in krb4-0.9.7, where 4793 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>] 4794 4795 *) Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly 4796 discouraged. 4797 [Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>] 4798 4799 *) For easily testing in shell scripts whether some command 4800 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX' 4801 returns with exit code 0 iff no command of the given name is available. 4802 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases, 4803 the output goes to stdout and nothing is printed to stderr. 4804 Additional arguments are always ignored. 4805 4806 Since for each cipher there is a command of the same name, 4807 the 'no-cipher' compilation switches can be tested this way. 4808 4809 ('openssl no-XXX' is not able to detect pseudo-commands such 4810 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.) 4811 [Bodo Moeller] 4812 4813 *) Update test suite so that 'make test' succeeds in 'no-rsa' configuration. 4814 [Bodo Moeller] 4815 4816 *) For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE 4817 is set; it will be thrown away anyway because each handshake creates 4818 its own key. 4819 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition 4820 to parameters -- in previous versions (since OpenSSL 0.9.3) the 4821 'default key' from SSL_CTX_set_tmp_dh would always be lost, meanining 4822 you effectivly got SSL_OP_SINGLE_DH_USE when using this macro. 4823 [Bodo Moeller] 4824 4825 *) New s_client option -ign_eof: EOF at stdin is ignored, and 4826 'Q' and 'R' lose their special meanings (quit/renegotiate). 4827 This is part of what -quiet does; unlike -quiet, -ign_eof 4828 does not suppress any output. 4829 [Richard Levitte] 4830 4831 *) Add compatibility options to the purpose and trust code. The 4832 purpose X509_PURPOSE_ANY is "any purpose" which automatically 4833 accepts a certificate or CA, this was the previous behaviour, 4834 with all the associated security issues. 4835 4836 X509_TRUST_COMPAT is the old trust behaviour: only and 4837 automatically trust self signed roots in certificate store. A 4838 new trust setting X509_TRUST_DEFAULT is used to specify that 4839 a purpose has no associated trust setting and it should instead 4840 use the value in the default purpose. 4841 [Steve Henson] 4842 4843 *) Fix the PKCS#8 DSA private key code so it decodes keys again 4844 and fix a memory leak. 4845 [Steve Henson] 4846 4847 *) In util/mkerr.pl (which implements 'make errors'), preserve 4848 reason strings from the previous version of the .c file, as 4849 the default to have only downcase letters (and digits) in 4850 automatically generated reasons codes is not always appropriate. 4851 [Bodo Moeller] 4852 4853 *) In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table 4854 using strerror. Previously, ERR_reason_error_string() returned 4855 library names as reason strings for SYSerr; but SYSerr is a special 4856 case where small numbers are errno values, not library numbers. 4857 [Bodo Moeller] 4858 4859 *) Add '-dsaparam' option to 'openssl dhparam' application. This 4860 converts DSA parameters into DH parameters. (When creating parameters, 4861 DSA_generate_parameters is used.) 4862 [Bodo Moeller] 4863 4864 *) Include 'length' (recommended exponent length) in C code generated 4865 by 'openssl dhparam -C'. 4866 [Bodo Moeller] 4867 4868 *) The second argument to set_label in perlasm was already being used 4869 so couldn't be used as a "file scope" flag. Moved to third argument 4870 which was free. 4871 [Steve Henson] 4872 4873 *) In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes 4874 instead of RAND_bytes for encryption IVs and salts. 4875 [Bodo Moeller] 4876 4877 *) Include RAND_status() into RAND_METHOD instead of implementing 4878 it only for md_rand.c Otherwise replacing the PRNG by calling 4879 RAND_set_rand_method would be impossible. 4880 [Bodo Moeller] 4881 4882 *) Don't let DSA_generate_key() enter an infinite loop if the random 4883 number generation fails. 4884 [Bodo Moeller] 4885 4886 *) New 'rand' application for creating pseudo-random output. 4887 [Bodo Moeller] 4888 4889 *) Added configuration support for Linux/IA64 4890 [Rolf Haberrecker <rolf@suse.de>] 4891 4892 *) Assembler module support for Mingw32. 4893 [Ulf M�ller] 4894 4895 *) Shared library support for HPUX (in shlib/). 4896 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous] 4897 4898 *) Shared library support for Solaris gcc. 4899 [Lutz Behnke <behnke@trustcenter.de>] 4900 4901 Changes between 0.9.4 and 0.9.5 [28 Feb 2000] 4902 4903 *) PKCS7_encrypt() was adding text MIME headers twice because they 4904 were added manually and by SMIME_crlf_copy(). 4905 [Steve Henson] 4906 4907 *) In bntest.c don't call BN_rand with zero bits argument. 4908 [Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>] 4909 4910 *) BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n] 4911 case was implemented. This caused BN_div_recp() to fail occasionally. 4912 [Ulf M�ller] 4913 4914 *) Add an optional second argument to the set_label() in the perl 4915 assembly language builder. If this argument exists and is set 4916 to 1 it signals that the assembler should use a symbol whose 4917 scope is the entire file, not just the current function. This 4918 is needed with MASM which uses the format label:: for this scope. 4919 [Steve Henson, pointed out by Peter Runestig <peter@runestig.com>] 4920 4921 *) Change the ASN1 types so they are typedefs by default. Before 4922 almost all types were #define'd to ASN1_STRING which was causing 4923 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING) 4924 for example. 4925 [Steve Henson] 4926 4927 *) Change names of new functions to the new get1/get0 naming 4928 convention: After 'get1', the caller owns a reference count 4929 and has to call ..._free; 'get0' returns a pointer to some 4930 data structure without incrementing reference counters. 4931 (Some of the existing 'get' functions increment a reference 4932 counter, some don't.) 4933 Similarly, 'set1' and 'add1' functions increase reference 4934 counters or duplicate objects. 4935 [Steve Henson] 4936 4937 *) Allow for the possibility of temp RSA key generation failure: 4938 the code used to assume it always worked and crashed on failure. 4939 [Steve Henson] 4940 4941 *) Fix potential buffer overrun problem in BIO_printf(). 4942 [Ulf M�ller, using public domain code by Patrick Powell; problem 4943 pointed out by David Sacerdote <das33@cornell.edu>] 4944 4945 *) Support EGD <http://www.lothar.com/tech/crypto/>. New functions 4946 RAND_egd() and RAND_status(). In the command line application, 4947 the EGD socket can be specified like a seed file using RANDFILE 4948 or -rand. 4949 [Ulf M�ller] 4950 4951 *) Allow the string CERTIFICATE to be tolerated in PKCS#7 structures. 4952 Some CAs (e.g. Verisign) distribute certificates in this form. 4953 [Steve Henson] 4954 4955 *) Remove the SSL_ALLOW_ADH compile option and set the default cipher 4956 list to exclude them. This means that no special compilation option 4957 is needed to use anonymous DH: it just needs to be included in the 4958 cipher list. 4959 [Steve Henson] 4960 4961 *) Change the EVP_MD_CTX_type macro so its meaning consistent with 4962 EVP_MD_type. The old functionality is available in a new macro called 4963 EVP_MD_md(). Change code that uses it and update docs. 4964 [Steve Henson] 4965 4966 *) ..._ctrl functions now have corresponding ..._callback_ctrl functions 4967 where the 'void *' argument is replaced by a function pointer argument. 4968 Previously 'void *' was abused to point to functions, which works on 4969 many platforms, but is not correct. As these functions are usually 4970 called by macros defined in OpenSSL header files, most source code 4971 should work without changes. 4972 [Richard Levitte] 4973 4974 *) <openssl/opensslconf.h> (which is created by Configure) now contains 4975 sections with information on -D... compiler switches used for 4976 compiling the library so that applications can see them. To enable 4977 one of these sections, a pre-processor symbol OPENSSL_..._DEFINES 4978 must be defined. E.g., 4979 #define OPENSSL_ALGORITHM_DEFINES 4980 #include <openssl/opensslconf.h> 4981 defines all pertinent NO_<algo> symbols, such as NO_IDEA, NO_RSA, etc. 4982 [Richard Levitte, Ulf and Bodo M�ller] 4983 4984 *) Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS 4985 record layer. 4986 [Bodo Moeller] 4987 4988 *) Change the 'other' type in certificate aux info to a STACK_OF 4989 X509_ALGOR. Although not an AlgorithmIdentifier as such it has 4990 the required ASN1 format: arbitrary types determined by an OID. 4991 [Steve Henson] 4992 4993 *) Add some PEM_write_X509_REQ_NEW() functions and a command line 4994 argument to 'req'. This is not because the function is newer or 4995 better than others it just uses the work 'NEW' in the certificate 4996 request header lines. Some software needs this. 4997 [Steve Henson] 4998 4999 *) Reorganise password command line arguments: now passwords can be 5000 obtained from various sources. Delete the PEM_cb function and make 5001 it the default behaviour: i.e. if the callback is NULL and the 5002 usrdata argument is not NULL interpret it as a null terminated pass 5003 phrase. If usrdata and the callback are NULL then the pass phrase 5004 is prompted for as usual. 5005 [Steve Henson] 5006 5007 *) Add support for the Compaq Atalla crypto accelerator. If it is installed, 5008 the support is automatically enabled. The resulting binaries will 5009 autodetect the card and use it if present. 5010 [Ben Laurie and Compaq Inc.] 5011 5012 *) Work around for Netscape hang bug. This sends certificate request 5013 and server done in one record. Since this is perfectly legal in the 5014 SSL/TLS protocol it isn't a "bug" option and is on by default. See 5015 the bugs/SSLv3 entry for more info. 5016 [Steve Henson] 5017 5018 *) HP-UX tune-up: new unified configs, HP C compiler bug workaround. 5019 [Andy Polyakov] 5020 5021 *) Add -rand argument to smime and pkcs12 applications and read/write 5022 of seed file. 5023 [Steve Henson] 5024 5025 *) New 'passwd' tool for crypt(3) and apr1 password hashes. 5026 [Bodo Moeller] 5027 5028 *) Add command line password options to the remaining applications. 5029 [Steve Henson] 5030 5031 *) Bug fix for BN_div_recp() for numerators with an even number of 5032 bits. 5033 [Ulf M�ller] 5034 5035 *) More tests in bntest.c, and changed test_bn output. 5036 [Ulf M�ller] 5037 5038 *) ./config recognizes MacOS X now. 5039 [Andy Polyakov] 5040 5041 *) Bug fix for BN_div() when the first words of num and divsor are 5042 equal (it gave wrong results if (rem=(n1-q*d0)&BN_MASK2) < d0). 5043 [Ulf M�ller] 5044 5045 *) Add support for various broken PKCS#8 formats, and command line 5046 options to produce them. 5047 [Steve Henson] 5048 5049 *) New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to 5050 get temporary BIGNUMs from a BN_CTX. 5051 [Ulf M�ller] 5052 5053 *) Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont() 5054 for p == 0. 5055 [Ulf M�ller] 5056 5057 *) Change the SSLeay_add_all_*() functions to OpenSSL_add_all_*() and 5058 include a #define from the old name to the new. The original intent 5059 was that statically linked binaries could for example just call 5060 SSLeay_add_all_ciphers() to just add ciphers to the table and not 5061 link with digests. This never worked becayse SSLeay_add_all_digests() 5062 and SSLeay_add_all_ciphers() were in the same source file so calling 5063 one would link with the other. They are now in separate source files. 5064 [Steve Henson] 5065 5066 *) Add a new -notext option to 'ca' and a -pubkey option to 'spkac'. 5067 [Steve Henson] 5068 5069 *) Use a less unusual form of the Miller-Rabin primality test (it used 5070 a binary algorithm for exponentiation integrated into the Miller-Rabin 5071 loop, our standard modexp algorithms are faster). 5072 [Bodo Moeller] 5073 5074 *) Support for the EBCDIC character set completed. 5075 [Martin Kraemer <Martin.Kraemer@Mch.SNI.De>] 5076 5077 *) Source code cleanups: use const where appropriate, eliminate casts, 5078 use void * instead of char * in lhash. 5079 [Ulf M�ller] 5080 5081 *) Bugfix: ssl3_send_server_key_exchange was not restartable 5082 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of 5083 this the server could overwrite ephemeral keys that the client 5084 has already seen). 5085 [Bodo Moeller] 5086 5087 *) Turn DSA_is_prime into a macro that calls BN_is_prime, 5088 using 50 iterations of the Rabin-Miller test. 5089 5090 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50 5091 iterations of the Rabin-Miller test as required by the appendix 5092 to FIPS PUB 186[-1]) instead of DSA_is_prime. 5093 As BN_is_prime_fasttest includes trial division, DSA parameter 5094 generation becomes much faster. 5095 5096 This implies a change for the callback functions in DSA_is_prime 5097 and DSA_generate_parameters: The callback function is called once 5098 for each positive witness in the Rabin-Miller test, not just 5099 occasionally in the inner loop; and the parameters to the 5100 callback function now provide an iteration count for the outer 5101 loop rather than for the current invocation of the inner loop. 5102 DSA_generate_parameters additionally can call the callback 5103 function with an 'iteration count' of -1, meaning that a 5104 candidate has passed the trial division test (when q is generated 5105 from an application-provided seed, trial division is skipped). 5106 [Bodo Moeller] 5107 5108 *) New function BN_is_prime_fasttest that optionally does trial 5109 division before starting the Rabin-Miller test and has 5110 an additional BN_CTX * argument (whereas BN_is_prime always 5111 has to allocate at least one BN_CTX). 5112 'callback(1, -1, cb_arg)' is called when a number has passed the 5113 trial division stage. 5114 [Bodo Moeller] 5115 5116 *) Fix for bug in CRL encoding. The validity dates weren't being handled 5117 as ASN1_TIME. 5118 [Steve Henson] 5119 5120 *) New -pkcs12 option to CA.pl script to write out a PKCS#12 file. 5121 [Steve Henson] 5122 5123 *) New function BN_pseudo_rand(). 5124 [Ulf M�ller] 5125 5126 *) Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable) 5127 bignum version of BN_from_montgomery() with the working code from 5128 SSLeay 0.9.0 (the word based version is faster anyway), and clean up 5129 the comments. 5130 [Ulf M�ller] 5131 5132 *) Avoid a race condition in s2_clnt.c (function get_server_hello) that 5133 made it impossible to use the same SSL_SESSION data structure in 5134 SSL2 clients in multiple threads. 5135 [Bodo Moeller] 5136 5137 *) The return value of RAND_load_file() no longer counts bytes obtained 5138 by stat(). RAND_load_file(..., -1) is new and uses the complete file 5139 to seed the PRNG (previously an explicit byte count was required). 5140 [Ulf M�ller, Bodo M�ller] 5141 5142 *) Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes 5143 used (char *) instead of (void *) and had casts all over the place. 5144 [Steve Henson] 5145 5146 *) Make BN_generate_prime() return NULL on error if ret!=NULL. 5147 [Ulf M�ller] 5148 5149 *) Retain source code compatibility for BN_prime_checks macro: 5150 BN_is_prime(..., BN_prime_checks, ...) now uses 5151 BN_prime_checks_for_size to determine the appropriate number of 5152 Rabin-Miller iterations. 5153 [Ulf M�ller] 5154 5155 *) Diffie-Hellman uses "safe" primes: DH_check() return code renamed to 5156 DH_CHECK_P_NOT_SAFE_PRIME. 5157 (Check if this is true? OpenPGP calls them "strong".) 5158 [Ulf M�ller] 5159 5160 *) Merge the functionality of "dh" and "gendh" programs into a new program 5161 "dhparam". The old programs are retained for now but will handle DH keys 5162 (instead of parameters) in future. 5163 [Steve Henson] 5164 5165 *) Make the ciphers, s_server and s_client programs check the return values 5166 when a new cipher list is set. 5167 [Steve Henson] 5168 5169 *) Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit 5170 ciphers. Before when the 56bit ciphers were enabled the sorting was 5171 wrong. 5172 5173 The syntax for the cipher sorting has been extended to support sorting by 5174 cipher-strength (using the strength_bits hard coded in the tables). 5175 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod). 5176 5177 Fix a bug in the cipher-command parser: when supplying a cipher command 5178 string with an "undefined" symbol (neither command nor alphanumeric 5179 [A-Za-z0-9], ssl_set_cipher_list used to hang in an endless loop. Now 5180 an error is flagged. 5181 5182 Due to the strength-sorting extension, the code of the 5183 ssl_create_cipher_list() function was completely rearranged. I hope that 5184 the readability was also increased :-) 5185 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>] 5186 5187 *) Minor change to 'x509' utility. The -CAcreateserial option now uses 1 5188 for the first serial number and places 2 in the serial number file. This 5189 avoids problems when the root CA is created with serial number zero and 5190 the first user certificate has the same issuer name and serial number 5191 as the root CA. 5192 [Steve Henson] 5193 5194 *) Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses 5195 the new code. Add documentation for this stuff. 5196 [Steve Henson] 5197 5198 *) Changes to X509_ATTRIBUTE utilities. These have been renamed from 5199 X509_*() to X509at_*() on the grounds that they don't handle X509 5200 structures and behave in an analagous way to the X509v3 functions: 5201 they shouldn't be called directly but wrapper functions should be used 5202 instead. 5203 5204 So we also now have some wrapper functions that call the X509at functions 5205 when passed certificate requests. (TO DO: similar things can be done with 5206 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other 5207 things. Some of these need some d2i or i2d and print functionality 5208 because they handle more complex structures.) 5209 [Steve Henson] 5210 5211 *) Add missing #ifndefs that caused missing symbols when building libssl 5212 as a shared library without RSA. Use #ifndef NO_SSL2 instead of 5213 NO_RSA in ssl/s2*.c. 5214 [Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf M�ller] 5215 5216 *) Precautions against using the PRNG uninitialized: RAND_bytes() now 5217 has a return value which indicates the quality of the random data 5218 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's 5219 error queue. New function RAND_pseudo_bytes() generates output that is 5220 guaranteed to be unique but not unpredictable. RAND_add is like 5221 RAND_seed, but takes an extra argument for an entropy estimate 5222 (RAND_seed always assumes full entropy). 5223 [Ulf M�ller] 5224 5225 *) Do more iterations of Rabin-Miller probable prime test (specifically, 5226 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes 5227 instead of only 2 for all lengths; see BN_prime_checks_for_size definition 5228 in crypto/bn/bn_prime.c for the complete table). This guarantees a 5229 false-positive rate of at most 2^-80 for random input. 5230 [Bodo Moeller] 5231 5232 *) Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs. 5233 [Bodo Moeller] 5234 5235 *) New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain 5236 in the 0.9.5 release), this returns the chain 5237 from an X509_CTX structure with a dup of the stack and all 5238 the X509 reference counts upped: so the stack will exist 5239 after X509_CTX_cleanup() has been called. Modify pkcs12.c 5240 to use this. 5241 5242 Also make SSL_SESSION_print() print out the verify return 5243 code. 5244 [Steve Henson] 5245 5246 *) Add manpage for the pkcs12 command. Also change the default 5247 behaviour so MAC iteration counts are used unless the new 5248 -nomaciter option is used. This improves file security and 5249 only older versions of MSIE (4.0 for example) need it. 5250 [Steve Henson] 5251 5252 *) Honor the no-xxx Configure options when creating .DEF files. 5253 [Ulf M�ller] 5254 5255 *) Add PKCS#10 attributes to field table: challengePassword, 5256 unstructuredName and unstructuredAddress. These are taken from 5257 draft PKCS#9 v2.0 but are compatible with v1.2 provided no 5258 international characters are used. 5259 5260 More changes to X509_ATTRIBUTE code: allow the setting of types 5261 based on strings. Remove the 'loc' parameter when adding 5262 attributes because these will be a SET OF encoding which is sorted 5263 in ASN1 order. 5264 [Steve Henson] 5265 5266 *) Initial changes to the 'req' utility to allow request generation 5267 automation. This will allow an application to just generate a template 5268 file containing all the field values and have req construct the 5269 request. 5270 5271 Initial support for X509_ATTRIBUTE handling. Stacks of these are 5272 used all over the place including certificate requests and PKCS#7 5273 structures. They are currently handled manually where necessary with 5274 some primitive wrappers for PKCS#7. The new functions behave in a 5275 manner analogous to the X509 extension functions: they allow 5276 attributes to be looked up by NID and added. 5277 5278 Later something similar to the X509V3 code would be desirable to 5279 automatically handle the encoding, decoding and printing of the 5280 more complex types. The string types like challengePassword can 5281 be handled by the string table functions. 5282 5283 Also modified the multi byte string table handling. Now there is 5284 a 'global mask' which masks out certain types. The table itself 5285 can use the flag STABLE_NO_MASK to ignore the mask setting: this 5286 is useful when for example there is only one permissible type 5287 (as in countryName) and using the mask might result in no valid 5288 types at all. 5289 [Steve Henson] 5290 5291 *) Clean up 'Finished' handling, and add functions SSL_get_finished and 5292 SSL_get_peer_finished to allow applications to obtain the latest 5293 Finished messages sent to the peer or expected from the peer, 5294 respectively. (SSL_get_peer_finished is usually the Finished message 5295 actually received from the peer, otherwise the protocol will be aborted.) 5296 5297 As the Finished message are message digests of the complete handshake 5298 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can 5299 be used for external authentication procedures when the authentication 5300 provided by SSL/TLS is not desired or is not enough. 5301 [Bodo Moeller] 5302 5303 *) Enhanced support for Alpha Linux is added. Now ./config checks if 5304 the host supports BWX extension and if Compaq C is present on the 5305 $PATH. Just exploiting of the BWX extension results in 20-30% 5306 performance kick for some algorithms, e.g. DES and RC4 to mention 5307 a couple. Compaq C in turn generates ~20% faster code for MD5 and 5308 SHA1. 5309 [Andy Polyakov] 5310 5311 *) Add support for MS "fast SGC". This is arguably a violation of the 5312 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with 5313 weak crypto and after checking the certificate is SGC a second one 5314 with strong crypto. MS SGC stops the first handshake after receiving 5315 the server certificate message and sends a second client hello. Since 5316 a server will typically do all the time consuming operations before 5317 expecting any further messages from the client (server key exchange 5318 is the most expensive) there is little difference between the two. 5319 5320 To get OpenSSL to support MS SGC we have to permit a second client 5321 hello message after we have sent server done. In addition we have to 5322 reset the MAC if we do get this second client hello. 5323 [Steve Henson] 5324 5325 *) Add a function 'd2i_AutoPrivateKey()' this will automatically decide 5326 if a DER encoded private key is RSA or DSA traditional format. Changed 5327 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional" 5328 format DER encoded private key. Newer code should use PKCS#8 format which 5329 has the key type encoded in the ASN1 structure. Added DER private key 5330 support to pkcs8 application. 5331 [Steve Henson] 5332 5333 *) SSL 3/TLS 1 servers now don't request certificates when an anonymous 5334 ciphersuites has been selected (as required by the SSL 3/TLS 1 5335 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT 5336 is set, we interpret this as a request to violate the specification 5337 (the worst that can happen is a handshake failure, and 'correct' 5338 behaviour would result in a handshake failure anyway). 5339 [Bodo Moeller] 5340 5341 *) In SSL_CTX_add_session, take into account that there might be multiple 5342 SSL_SESSION structures with the same session ID (e.g. when two threads 5343 concurrently obtain them from an external cache). 5344 The internal cache can handle only one SSL_SESSION with a given ID, 5345 so if there's a conflict, we now throw out the old one to achieve 5346 consistency. 5347 [Bodo Moeller] 5348 5349 *) Add OIDs for idea and blowfish in CBC mode. This will allow both 5350 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to 5351 some routines that use cipher OIDs: some ciphers do not have OIDs 5352 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for 5353 example. 5354 [Steve Henson] 5355 5356 *) Simplify the trust setting structure and code. Now we just have 5357 two sequences of OIDs for trusted and rejected settings. These will 5358 typically have values the same as the extended key usage extension 5359 and any application specific purposes. 5360 5361 The trust checking code now has a default behaviour: it will just 5362 check for an object with the same NID as the passed id. Functions can 5363 be provided to override either the default behaviour or the behaviour 5364 for a given id. SSL client, server and email already have functions 5365 in place for compatibility: they check the NID and also return "trusted" 5366 if the certificate is self signed. 5367 [Steve Henson] 5368 5369 *) Add d2i,i2d bio/fp functions for PrivateKey: these convert the 5370 traditional format into an EVP_PKEY structure. 5371 [Steve Henson] 5372 5373 *) Add a password callback function PEM_cb() which either prompts for 5374 a password if usr_data is NULL or otherwise assumes it is a null 5375 terminated password. Allow passwords to be passed on command line 5376 environment or config files in a few more utilities. 5377 [Steve Henson] 5378 5379 *) Add a bunch of DER and PEM functions to handle PKCS#8 format private 5380 keys. Add some short names for PKCS#8 PBE algorithms and allow them 5381 to be specified on the command line for the pkcs8 and pkcs12 utilities. 5382 Update documentation. 5383 [Steve Henson] 5384 5385 *) Support for ASN1 "NULL" type. This could be handled before by using 5386 ASN1_TYPE but there wasn't any function that would try to read a NULL 5387 and produce an error if it couldn't. For compatibility we also have 5388 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and 5389 don't allocate anything because they don't need to. 5390 [Steve Henson] 5391 5392 *) Initial support for MacOS is now provided. Examine INSTALL.MacOS 5393 for details. 5394 [Andy Polyakov, Roy Woods <roy@centicsystems.ca>] 5395 5396 *) Rebuild of the memory allocation routines used by OpenSSL code and 5397 possibly others as well. The purpose is to make an interface that 5398 provide hooks so anyone can build a separate set of allocation and 5399 deallocation routines to be used by OpenSSL, for example memory 5400 pool implementations, or something else, which was previously hard 5401 since Malloc(), Realloc() and Free() were defined as macros having 5402 the values malloc, realloc and free, respectively (except for Win32 5403 compilations). The same is provided for memory debugging code. 5404 OpenSSL already comes with functionality to find memory leaks, but 5405 this gives people a chance to debug other memory problems. 5406 5407 With these changes, a new set of functions and macros have appeared: 5408 5409 CRYPTO_set_mem_debug_functions() [F] 5410 CRYPTO_get_mem_debug_functions() [F] 5411 CRYPTO_dbg_set_options() [F] 5412 CRYPTO_dbg_get_options() [F] 5413 CRYPTO_malloc_debug_init() [M] 5414 5415 The memory debug functions are NULL by default, unless the library 5416 is compiled with CRYPTO_MDEBUG or friends is defined. If someone 5417 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which 5418 gives the standard debugging functions that come with OpenSSL) or 5419 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions 5420 provided by the library user) must be used. When the standard 5421 debugging functions are used, CRYPTO_dbg_set_options can be used to 5422 request additional information: 5423 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting 5424 the CRYPTO_MDEBUG_xxx macro when compiling the library. 5425 5426 Also, things like CRYPTO_set_mem_functions will always give the 5427 expected result (the new set of functions is used for allocation 5428 and deallocation) at all times, regardless of platform and compiler 5429 options. 5430 5431 To finish it up, some functions that were never use in any other 5432 way than through macros have a new API and new semantic: 5433 5434 CRYPTO_dbg_malloc() 5435 CRYPTO_dbg_realloc() 5436 CRYPTO_dbg_free() 5437 5438 All macros of value have retained their old syntax. 5439 [Richard Levitte and Bodo Moeller] 5440 5441 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the 5442 ordering of SMIMECapabilities wasn't in "strength order" and there 5443 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature 5444 algorithm. 5445 [Steve Henson] 5446 5447 *) Some ASN1 types with illegal zero length encoding (INTEGER, 5448 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines. 5449 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson] 5450 5451 *) Merge in my S/MIME library for OpenSSL. This provides a simple 5452 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough 5453 functionality to handle multipart/signed properly) and a utility 5454 called 'smime' to call all this stuff. This is based on code I 5455 originally wrote for Celo who have kindly allowed it to be 5456 included in OpenSSL. 5457 [Steve Henson] 5458 5459 *) Add variants des_set_key_checked and des_set_key_unchecked of 5460 des_set_key (aka des_key_sched). Global variable des_check_key 5461 decides which of these is called by des_set_key; this way 5462 des_check_key behaves as it always did, but applications and 5463 the library itself, which was buggy for des_check_key == 1, 5464 have a cleaner way to pick the version they need. 5465 [Bodo Moeller] 5466 5467 *) New function PKCS12_newpass() which changes the password of a 5468 PKCS12 structure. 5469 [Steve Henson] 5470 5471 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and 5472 dynamic mix. In both cases the ids can be used as an index into the 5473 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add() 5474 functions so they accept a list of the field values and the 5475 application doesn't need to directly manipulate the X509_TRUST 5476 structure. 5477 [Steve Henson] 5478 5479 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't 5480 need initialising. 5481 [Steve Henson] 5482 5483 *) Modify the way the V3 extension code looks up extensions. This now 5484 works in a similar way to the object code: we have some "standard" 5485 extensions in a static table which is searched with OBJ_bsearch() 5486 and the application can add dynamic ones if needed. The file 5487 crypto/x509v3/ext_dat.h now has the info: this file needs to be 5488 updated whenever a new extension is added to the core code and kept 5489 in ext_nid order. There is a simple program 'tabtest.c' which checks 5490 this. New extensions are not added too often so this file can readily 5491 be maintained manually. 5492 5493 There are two big advantages in doing things this way. The extensions 5494 can be looked up immediately and no longer need to be "added" using 5495 X509V3_add_standard_extensions(): this function now does nothing. 5496 [Side note: I get *lots* of email saying the extension code doesn't 5497 work because people forget to call this function] 5498 Also no dynamic allocation is done unless new extensions are added: 5499 so if we don't add custom extensions there is no need to call 5500 X509V3_EXT_cleanup(). 5501 [Steve Henson] 5502 5503 *) Modify enc utility's salting as follows: make salting the default. Add a 5504 magic header, so unsalted files fail gracefully instead of just decrypting 5505 to garbage. This is because not salting is a big security hole, so people 5506 should be discouraged from doing it. 5507 [Ben Laurie] 5508 5509 *) Fixes and enhancements to the 'x509' utility. It allowed a message 5510 digest to be passed on the command line but it only used this 5511 parameter when signing a certificate. Modified so all relevant 5512 operations are affected by the digest parameter including the 5513 -fingerprint and -x509toreq options. Also -x509toreq choked if a 5514 DSA key was used because it didn't fix the digest. 5515 [Steve Henson] 5516 5517 *) Initial certificate chain verify code. Currently tests the untrusted 5518 certificates for consistency with the verify purpose (which is set 5519 when the X509_STORE_CTX structure is set up) and checks the pathlength. 5520 5521 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour: 5522 this is because it will reject chains with invalid extensions whereas 5523 every previous version of OpenSSL and SSLeay made no checks at all. 5524 5525 Trust code: checks the root CA for the relevant trust settings. Trust 5526 settings have an initial value consistent with the verify purpose: e.g. 5527 if the verify purpose is for SSL client use it expects the CA to be 5528 trusted for SSL client use. However the default value can be changed to 5529 permit custom trust settings: one example of this would be to only trust 5530 certificates from a specific "secure" set of CAs. 5531 5532 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions 5533 which should be used for version portability: especially since the 5534 verify structure is likely to change more often now. 5535 5536 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions 5537 to set them. If not set then assume SSL clients will verify SSL servers 5538 and vice versa. 5539 5540 Two new options to the verify program: -untrusted allows a set of 5541 untrusted certificates to be passed in and -purpose which sets the 5542 intended purpose of the certificate. If a purpose is set then the 5543 new chain verify code is used to check extension consistency. 5544 [Steve Henson] 5545 5546 *) Support for the authority information access extension. 5547 [Steve Henson] 5548 5549 *) Modify RSA and DSA PEM read routines to transparently handle 5550 PKCS#8 format private keys. New *_PUBKEY_* functions that handle 5551 public keys in a format compatible with certificate 5552 SubjectPublicKeyInfo structures. Unfortunately there were already 5553 functions called *_PublicKey_* which used various odd formats so 5554 these are retained for compatibility: however the DSA variants were 5555 never in a public release so they have been deleted. Changed dsa/rsa 5556 utilities to handle the new format: note no releases ever handled public 5557 keys so we should be OK. 5558 5559 The primary motivation for this change is to avoid the same fiasco 5560 that dogs private keys: there are several incompatible private key 5561 formats some of which are standard and some OpenSSL specific and 5562 require various evil hacks to allow partial transparent handling and 5563 even then it doesn't work with DER formats. Given the option anything 5564 other than PKCS#8 should be dumped: but the other formats have to 5565 stay in the name of compatibility. 5566 5567 With public keys and the benefit of hindsight one standard format 5568 is used which works with EVP_PKEY, RSA or DSA structures: though 5569 it clearly returns an error if you try to read the wrong kind of key. 5570 5571 Added a -pubkey option to the 'x509' utility to output the public key. 5572 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*() 5573 (renamed to EVP_PKEY_get1_*() in the OpenSSL 0.9.5 release) and add 5574 EVP_PKEY_rset_*() functions (renamed to EVP_PKEY_set1_*()) 5575 that do the same as the EVP_PKEY_assign_*() except they up the 5576 reference count of the added key (they don't "swallow" the 5577 supplied key). 5578 [Steve Henson] 5579 5580 *) Fixes to crypto/x509/by_file.c the code to read in certificates and 5581 CRLs would fail if the file contained no certificates or no CRLs: 5582 added a new function to read in both types and return the number 5583 read: this means that if none are read it will be an error. The 5584 DER versions of the certificate and CRL reader would always fail 5585 because it isn't possible to mix certificates and CRLs in DER format 5586 without choking one or the other routine. Changed this to just read 5587 a certificate: this is the best we can do. Also modified the code 5588 in apps/verify.c to take notice of return codes: it was previously 5589 attempting to read in certificates from NULL pointers and ignoring 5590 any errors: this is one reason why the cert and CRL reader seemed 5591 to work. It doesn't check return codes from the default certificate 5592 routines: these may well fail if the certificates aren't installed. 5593 [Steve Henson] 5594 5595 *) Code to support otherName option in GeneralName. 5596 [Steve Henson] 5597 5598 *) First update to verify code. Change the verify utility 5599 so it warns if it is passed a self signed certificate: 5600 for consistency with the normal behaviour. X509_verify 5601 has been modified to it will now verify a self signed 5602 certificate if *exactly* the same certificate appears 5603 in the store: it was previously impossible to trust a 5604 single self signed certificate. This means that: 5605 openssl verify ss.pem 5606 now gives a warning about a self signed certificate but 5607 openssl verify -CAfile ss.pem ss.pem 5608 is OK. 5609 [Steve Henson] 5610 5611 *) For servers, store verify_result in SSL_SESSION data structure 5612 (and add it to external session representation). 5613 This is needed when client certificate verifications fails, 5614 but an application-provided verification callback (set by 5615 SSL_CTX_set_cert_verify_callback) allows accepting the session 5616 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK 5617 but returns 1): When the session is reused, we have to set 5618 ssl->verify_result to the appropriate error code to avoid 5619 security holes. 5620 [Bodo Moeller, problem pointed out by Lutz Jaenicke] 5621 5622 *) Fix a bug in the new PKCS#7 code: it didn't consider the 5623 case in PKCS7_dataInit() where the signed PKCS7 structure 5624 didn't contain any existing data because it was being created. 5625 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson] 5626 5627 *) Add a salt to the key derivation routines in enc.c. This 5628 forms the first 8 bytes of the encrypted file. Also add a 5629 -S option to allow a salt to be input on the command line. 5630 [Steve Henson] 5631 5632 *) New function X509_cmp(). Oddly enough there wasn't a function 5633 to compare two certificates. We do this by working out the SHA1 5634 hash and comparing that. X509_cmp() will be needed by the trust 5635 code. 5636 [Steve Henson] 5637 5638 *) SSL_get1_session() is like SSL_get_session(), but increments 5639 the reference count in the SSL_SESSION returned. 5640 [Geoff Thorpe <geoff@eu.c2.net>] 5641 5642 *) Fix for 'req': it was adding a null to request attributes. 5643 Also change the X509_LOOKUP and X509_INFO code to handle 5644 certificate auxiliary information. 5645 [Steve Henson] 5646 5647 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document 5648 the 'enc' command. 5649 [Steve Henson] 5650 5651 *) Add the possibility to add extra information to the memory leak 5652 detecting output, to form tracebacks, showing from where each 5653 allocation was originated: CRYPTO_push_info("constant string") adds 5654 the string plus current file name and line number to a per-thread 5655 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info() 5656 is like calling CYRPTO_pop_info() until the stack is empty. 5657 Also updated memory leak detection code to be multi-thread-safe. 5658 [Richard Levitte] 5659 5660 *) Add options -text and -noout to pkcs7 utility and delete the 5661 encryption options which never did anything. Update docs. 5662 [Steve Henson] 5663 5664 *) Add options to some of the utilities to allow the pass phrase 5665 to be included on either the command line (not recommended on 5666 OSes like Unix) or read from the environment. Update the 5667 manpages and fix a few bugs. 5668 [Steve Henson] 5669 5670 *) Add a few manpages for some of the openssl commands. 5671 [Steve Henson] 5672 5673 *) Fix the -revoke option in ca. It was freeing up memory twice, 5674 leaking and not finding already revoked certificates. 5675 [Steve Henson] 5676 5677 *) Extensive changes to support certificate auxiliary information. 5678 This involves the use of X509_CERT_AUX structure and X509_AUX 5679 functions. An X509_AUX function such as PEM_read_X509_AUX() 5680 can still read in a certificate file in the usual way but it 5681 will also read in any additional "auxiliary information". By 5682 doing things this way a fair degree of compatibility can be 5683 retained: existing certificates can have this information added 5684 using the new 'x509' options. 5685 5686 Current auxiliary information includes an "alias" and some trust 5687 settings. The trust settings will ultimately be used in enhanced 5688 certificate chain verification routines: currently a certificate 5689 can only be trusted if it is self signed and then it is trusted 5690 for all purposes. 5691 [Steve Henson] 5692 5693 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD). 5694 The problem was that one of the replacement routines had not been working 5695 since SSLeay releases. For now the offending routine has been replaced 5696 with non-optimised assembler. Even so, this now gives around 95% 5697 performance improvement for 1024 bit RSA signs. 5698 [Mark Cox] 5699 5700 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2 5701 handling. Most clients have the effective key size in bits equal to 5702 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key. 5703 A few however don't do this and instead use the size of the decrypted key 5704 to determine the RC2 key length and the AlgorithmIdentifier to determine 5705 the effective key length. In this case the effective key length can still 5706 be 40 bits but the key length can be 168 bits for example. This is fixed 5707 by manually forcing an RC2 key into the EVP_PKEY structure because the 5708 EVP code can't currently handle unusual RC2 key sizes: it always assumes 5709 the key length and effective key length are equal. 5710 [Steve Henson] 5711 5712 *) Add a bunch of functions that should simplify the creation of 5713 X509_NAME structures. Now you should be able to do: 5714 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0); 5715 and have it automatically work out the correct field type and fill in 5716 the structures. The more adventurous can try: 5717 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0); 5718 and it will (hopefully) work out the correct multibyte encoding. 5719 [Steve Henson] 5720 5721 *) Change the 'req' utility to use the new field handling and multibyte 5722 copy routines. Before the DN field creation was handled in an ad hoc 5723 way in req, ca, and x509 which was rather broken and didn't support 5724 BMPStrings or UTF8Strings. Since some software doesn't implement 5725 BMPStrings or UTF8Strings yet, they can be enabled using the config file 5726 using the dirstring_type option. See the new comment in the default 5727 openssl.cnf for more info. 5728 [Steve Henson] 5729 5730 *) Make crypto/rand/md_rand.c more robust: 5731 - Assure unique random numbers after fork(). 5732 - Make sure that concurrent threads access the global counter and 5733 md serializably so that we never lose entropy in them 5734 or use exactly the same state in multiple threads. 5735 Access to the large state is not always serializable because 5736 the additional locking could be a performance killer, and 5737 md should be large enough anyway. 5738 [Bodo Moeller] 5739 5740 *) New file apps/app_rand.c with commonly needed functionality 5741 for handling the random seed file. 5742 5743 Use the random seed file in some applications that previously did not: 5744 ca, 5745 dsaparam -genkey (which also ignored its '-rand' option), 5746 s_client, 5747 s_server, 5748 x509 (when signing). 5749 Except on systems with /dev/urandom, it is crucial to have a random 5750 seed file at least for key creation, DSA signing, and for DH exchanges; 5751 for RSA signatures we could do without one. 5752 5753 gendh and gendsa (unlike genrsa) used to read only the first byte 5754 of each file listed in the '-rand' option. The function as previously 5755 found in genrsa is now in app_rand.c and is used by all programs 5756 that support '-rand'. 5757 [Bodo Moeller] 5758 5759 *) In RAND_write_file, use mode 0600 for creating files; 5760 don't just chmod when it may be too late. 5761 [Bodo Moeller] 5762 5763 *) Report an error from X509_STORE_load_locations 5764 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed. 5765 [Bill Perry] 5766 5767 *) New function ASN1_mbstring_copy() this copies a string in either 5768 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format 5769 into an ASN1_STRING type. A mask of permissible types is passed 5770 and it chooses the "minimal" type to use or an error if not type 5771 is suitable. 5772 [Steve Henson] 5773 5774 *) Add function equivalents to the various macros in asn1.h. The old 5775 macros are retained with an M_ prefix. Code inside the library can 5776 use the M_ macros. External code (including the openssl utility) 5777 should *NOT* in order to be "shared library friendly". 5778 [Steve Henson] 5779 5780 *) Add various functions that can check a certificate's extensions 5781 to see if it usable for various purposes such as SSL client, 5782 server or S/MIME and CAs of these types. This is currently 5783 VERY EXPERIMENTAL but will ultimately be used for certificate chain 5784 verification. Also added a -purpose flag to x509 utility to 5785 print out all the purposes. 5786 [Steve Henson] 5787 5788 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated 5789 functions. 5790 [Steve Henson] 5791 5792 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search 5793 for, obtain and decode and extension and obtain its critical flag. 5794 This allows all the necessary extension code to be handled in a 5795 single function call. 5796 [Steve Henson] 5797 5798 *) RC4 tune-up featuring 30-40% performance improvement on most RISC 5799 platforms. See crypto/rc4/rc4_enc.c for further details. 5800 [Andy Polyakov] 5801 5802 *) New -noout option to asn1parse. This causes no output to be produced 5803 its main use is when combined with -strparse and -out to extract data 5804 from a file (which may not be in ASN.1 format). 5805 [Steve Henson] 5806 5807 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer 5808 when producing the local key id. 5809 [Richard Levitte <levitte@stacken.kth.se>] 5810 5811 *) New option -dhparam in s_server. This allows a DH parameter file to be 5812 stated explicitly. If it is not stated then it tries the first server 5813 certificate file. The previous behaviour hard coded the filename 5814 "server.pem". 5815 [Steve Henson] 5816 5817 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow 5818 a public key to be input or output. For example: 5819 openssl rsa -in key.pem -pubout -out pubkey.pem 5820 Also added necessary DSA public key functions to handle this. 5821 [Steve Henson] 5822 5823 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained 5824 in the message. This was handled by allowing 5825 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it. 5826 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>] 5827 5828 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null 5829 to the end of the strings whereas this didn't. This would cause problems 5830 if strings read with d2i_ASN1_bytes() were later modified. 5831 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>] 5832 5833 *) Fix for base64 decode bug. When a base64 bio reads only one line of 5834 data and it contains EOF it will end up returning an error. This is 5835 caused by input 46 bytes long. The cause is due to the way base64 5836 BIOs find the start of base64 encoded data. They do this by trying a 5837 trial decode on each line until they find one that works. When they 5838 do a flag is set and it starts again knowing it can pass all the 5839 data directly through the decoder. Unfortunately it doesn't reset 5840 the context it uses. This means that if EOF is reached an attempt 5841 is made to pass two EOFs through the context and this causes the 5842 resulting error. This can also cause other problems as well. As is 5843 usual with these problems it takes *ages* to find and the fix is 5844 trivial: move one line. 5845 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ] 5846 5847 *) Ugly workaround to get s_client and s_server working under Windows. The 5848 old code wouldn't work because it needed to select() on sockets and the 5849 tty (for keypresses and to see if data could be written). Win32 only 5850 supports select() on sockets so we select() with a 1s timeout on the 5851 sockets and then see if any characters are waiting to be read, if none 5852 are present then we retry, we also assume we can always write data to 5853 the tty. This isn't nice because the code then blocks until we've 5854 received a complete line of data and it is effectively polling the 5855 keyboard at 1s intervals: however it's quite a bit better than not 5856 working at all :-) A dedicated Windows application might handle this 5857 with an event loop for example. 5858 [Steve Henson] 5859 5860 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign 5861 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions 5862 will be called when RSA_sign() and RSA_verify() are used. This is useful 5863 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available. 5864 For this to work properly RSA_public_decrypt() and RSA_private_encrypt() 5865 should *not* be used: RSA_sign() and RSA_verify() must be used instead. 5866 This necessitated the support of an extra signature type NID_md5_sha1 5867 for SSL signatures and modifications to the SSL library to use it instead 5868 of calling RSA_public_decrypt() and RSA_private_encrypt(). 5869 [Steve Henson] 5870 5871 *) Add new -verify -CAfile and -CApath options to the crl program, these 5872 will lookup a CRL issuers certificate and verify the signature in a 5873 similar way to the verify program. Tidy up the crl program so it 5874 no longer accesses structures directly. Make the ASN1 CRL parsing a bit 5875 less strict. It will now permit CRL extensions even if it is not 5876 a V2 CRL: this will allow it to tolerate some broken CRLs. 5877 [Steve Henson] 5878 5879 *) Initialize all non-automatic variables each time one of the openssl 5880 sub-programs is started (this is necessary as they may be started 5881 multiple times from the "OpenSSL>" prompt). 5882 [Lennart Bang, Bodo Moeller] 5883 5884 *) Preliminary compilation option RSA_NULL which disables RSA crypto without 5885 removing all other RSA functionality (this is what NO_RSA does). This 5886 is so (for example) those in the US can disable those operations covered 5887 by the RSA patent while allowing storage and parsing of RSA keys and RSA 5888 key generation. 5889 [Steve Henson] 5890 5891 *) Non-copying interface to BIO pairs. 5892 (still largely untested) 5893 [Bodo Moeller] 5894 5895 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive 5896 ASCII string. This was handled independently in various places before. 5897 [Steve Henson] 5898 5899 *) New functions UTF8_getc() and UTF8_putc() that parse and generate 5900 UTF8 strings a character at a time. 5901 [Steve Henson] 5902 5903 *) Use client_version from client hello to select the protocol 5904 (s23_srvr.c) and for RSA client key exchange verification 5905 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications. 5906 [Bodo Moeller] 5907 5908 *) Add various utility functions to handle SPKACs, these were previously 5909 handled by poking round in the structure internals. Added new function 5910 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to 5911 print, verify and generate SPKACs. Based on an original idea from 5912 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified. 5913 [Steve Henson] 5914 5915 *) RIPEMD160 is operational on all platforms and is back in 'make test'. 5916 [Andy Polyakov] 5917 5918 *) Allow the config file extension section to be overwritten on the 5919 command line. Based on an original idea from Massimiliano Pala 5920 <madwolf@comune.modena.it>. The new option is called -extensions 5921 and can be applied to ca, req and x509. Also -reqexts to override 5922 the request extensions in req and -crlexts to override the crl extensions 5923 in ca. 5924 [Steve Henson] 5925 5926 *) Add new feature to the SPKAC handling in ca. Now you can include 5927 the same field multiple times by preceding it by "XXXX." for example: 5928 1.OU="Unit name 1" 5929 2.OU="Unit name 2" 5930 this is the same syntax as used in the req config file. 5931 [Steve Henson] 5932 5933 *) Allow certificate extensions to be added to certificate requests. These 5934 are specified in a 'req_extensions' option of the req section of the 5935 config file. They can be printed out with the -text option to req but 5936 are otherwise ignored at present. 5937 [Steve Henson] 5938 5939 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first 5940 data read consists of only the final block it would not decrypted because 5941 EVP_CipherUpdate() would correctly report zero bytes had been decrypted. 5942 A misplaced 'break' also meant the decrypted final block might not be 5943 copied until the next read. 5944 [Steve Henson] 5945 5946 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added 5947 a few extra parameters to the DH structure: these will be useful if 5948 for example we want the value of 'q' or implement X9.42 DH. 5949 [Steve Henson] 5950 5951 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and 5952 provides hooks that allow the default DSA functions or functions on a 5953 "per key" basis to be replaced. This allows hardware acceleration and 5954 hardware key storage to be handled without major modification to the 5955 library. Also added low level modexp hooks and CRYPTO_EX structure and 5956 associated functions. 5957 [Steve Henson] 5958 5959 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO 5960 as "read only": it can't be written to and the buffer it points to will 5961 not be freed. Reading from a read only BIO is much more efficient than 5962 a normal memory BIO. This was added because there are several times when 5963 an area of memory needs to be read from a BIO. The previous method was 5964 to create a memory BIO and write the data to it, this results in two 5965 copies of the data and an O(n^2) reading algorithm. There is a new 5966 function BIO_new_mem_buf() which creates a read only memory BIO from 5967 an area of memory. Also modified the PKCS#7 routines to use read only 5968 memory BIOs. 5969 [Steve Henson] 5970 5971 *) Bugfix: ssl23_get_client_hello did not work properly when called in 5972 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of 5973 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read, 5974 but a retry condition occured while trying to read the rest. 5975 [Bodo Moeller] 5976 5977 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as 5978 NID_pkcs7_encrypted by default: this was wrong since this should almost 5979 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle 5980 the encrypted data type: this is a more sensible place to put it and it 5981 allows the PKCS#12 code to be tidied up that duplicated this 5982 functionality. 5983 [Steve Henson] 5984 5985 *) Changed obj_dat.pl script so it takes its input and output files on 5986 the command line. This should avoid shell escape redirection problems 5987 under Win32. 5988 [Steve Henson] 5989 5990 *) Initial support for certificate extension requests, these are included 5991 in things like Xenroll certificate requests. Included functions to allow 5992 extensions to be obtained and added. 5993 [Steve Henson] 5994 5995 *) -crlf option to s_client and s_server for sending newlines as 5996 CRLF (as required by many protocols). 5997 [Bodo Moeller] 5998 5999 Changes between 0.9.3a and 0.9.4 [09 Aug 1999] 6000 6001 *) Install libRSAglue.a when OpenSSL is built with RSAref. 6002 [Ralf S. Engelschall] 6003 6004 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency. 6005 [Andrija Antonijevic <TheAntony2@bigfoot.com>] 6006 6007 *) Fix -startdate and -enddate (which was missing) arguments to 'ca' 6008 program. 6009 [Steve Henson] 6010 6011 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as 6012 DH parameters/keys (q is lost during that conversion, but the resulting 6013 DH parameters contain its length). 6014 6015 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is 6016 much faster than DH_generate_parameters (which creates parameters 6017 where p = 2*q + 1), and also the smaller q makes DH computations 6018 much more efficient (160-bit exponentiation instead of 1024-bit 6019 exponentiation); so this provides a convenient way to support DHE 6020 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of 6021 utter importance to use 6022 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE); 6023 or 6024 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE); 6025 when such DH parameters are used, because otherwise small subgroup 6026 attacks may become possible! 6027 [Bodo Moeller] 6028 6029 *) Avoid memory leak in i2d_DHparams. 6030 [Bodo Moeller] 6031 6032 *) Allow the -k option to be used more than once in the enc program: 6033 this allows the same encrypted message to be read by multiple recipients. 6034 [Steve Henson] 6035 6036 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts 6037 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then 6038 it will always use the numerical form of the OID, even if it has a short 6039 or long name. 6040 [Steve Henson] 6041 6042 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp 6043 method only got called if p,q,dmp1,dmq1,iqmp components were present, 6044 otherwise bn_mod_exp was called. In the case of hardware keys for example 6045 no private key components need be present and it might store extra data 6046 in the RSA structure, which cannot be accessed from bn_mod_exp. 6047 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for 6048 private key operations. 6049 [Steve Henson] 6050 6051 *) Added support for SPARC Linux. 6052 [Andy Polyakov] 6053 6054 *) pem_password_cb function type incompatibly changed from 6055 typedef int pem_password_cb(char *buf, int size, int rwflag); 6056 to 6057 ....(char *buf, int size, int rwflag, void *userdata); 6058 so that applications can pass data to their callbacks: 6059 The PEM[_ASN1]_{read,write}... functions and macros now take an 6060 additional void * argument, which is just handed through whenever 6061 the password callback is called. 6062 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller] 6063 6064 New function SSL_CTX_set_default_passwd_cb_userdata. 6065 6066 Compatibility note: As many C implementations push function arguments 6067 onto the stack in reverse order, the new library version is likely to 6068 interoperate with programs that have been compiled with the old 6069 pem_password_cb definition (PEM_whatever takes some data that 6070 happens to be on the stack as its last argument, and the callback 6071 just ignores this garbage); but there is no guarantee whatsoever that 6072 this will work. 6073 6074 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=... 6075 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused 6076 problems not only on Windows, but also on some Unix platforms. 6077 To avoid problematic command lines, these definitions are now in an 6078 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl 6079 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds). 6080 [Bodo Moeller] 6081 6082 *) MIPS III/IV assembler module is reimplemented. 6083 [Andy Polyakov] 6084 6085 *) More DES library cleanups: remove references to srand/rand and 6086 delete an unused file. 6087 [Ulf M�ller] 6088 6089 *) Add support for the the free Netwide assembler (NASM) under Win32, 6090 since not many people have MASM (ml) and it can be hard to obtain. 6091 This is currently experimental but it seems to work OK and pass all 6092 the tests. Check out INSTALL.W32 for info. 6093 [Steve Henson] 6094 6095 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections 6096 without temporary keys kept an extra copy of the server key, 6097 and connections with temporary keys did not free everything in case 6098 of an error. 6099 [Bodo Moeller] 6100 6101 *) New function RSA_check_key and new openssl rsa option -check 6102 for verifying the consistency of RSA keys. 6103 [Ulf Moeller, Bodo Moeller] 6104 6105 *) Various changes to make Win32 compile work: 6106 1. Casts to avoid "loss of data" warnings in p5_crpt2.c 6107 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned 6108 comparison" warnings. 6109 3. Add sk_<TYPE>_sort to DEF file generator and do make update. 6110 [Steve Henson] 6111 6112 *) Add a debugging option to PKCS#5 v2 key generation function: when 6113 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and 6114 derived keys are printed to stderr. 6115 [Steve Henson] 6116 6117 *) Copy the flags in ASN1_STRING_dup(). 6118 [Roman E. Pavlov <pre@mo.msk.ru>] 6119 6120 *) The x509 application mishandled signing requests containing DSA 6121 keys when the signing key was also DSA and the parameters didn't match. 6122 6123 It was supposed to omit the parameters when they matched the signing key: 6124 the verifying software was then supposed to automatically use the CA's 6125 parameters if they were absent from the end user certificate. 6126 6127 Omitting parameters is no longer recommended. The test was also 6128 the wrong way round! This was probably due to unusual behaviour in 6129 EVP_cmp_parameters() which returns 1 if the parameters match. 6130 This meant that parameters were omitted when they *didn't* match and 6131 the certificate was useless. Certificates signed with 'ca' didn't have 6132 this bug. 6133 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>] 6134 6135 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems. 6136 The interface is as follows: 6137 Applications can use 6138 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(), 6139 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop(); 6140 "off" is now the default. 6141 The library internally uses 6142 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(), 6143 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on() 6144 to disable memory-checking temporarily. 6145 6146 Some inconsistent states that previously were possible (and were 6147 even the default) are now avoided. 6148 6149 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time 6150 with each memory chunk allocated; this is occasionally more helpful 6151 than just having a counter. 6152 6153 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID. 6154 6155 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future 6156 extensions. 6157 [Bodo Moeller] 6158 6159 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX), 6160 which largely parallels "options", but is for changing API behaviour, 6161 whereas "options" are about protocol behaviour. 6162 Initial "mode" flags are: 6163 6164 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when 6165 a single record has been written. 6166 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write 6167 retries use the same buffer location. 6168 (But all of the contents must be 6169 copied!) 6170 [Bodo Moeller] 6171 6172 *) Bugfix: SSL_set_options ignored its parameter, only SSL_CTX_set_options 6173 worked. 6174 6175 *) Fix problems with no-hmac etc. 6176 [Ulf M�ller, pointed out by Brian Wellington <bwelling@tislabs.com>] 6177 6178 *) New functions RSA_get_default_method(), RSA_set_method() and 6179 RSA_get_method(). These allows replacement of RSA_METHODs without having 6180 to mess around with the internals of an RSA structure. 6181 [Steve Henson] 6182 6183 *) Fix memory leaks in DSA_do_sign and DSA_is_prime. 6184 Also really enable memory leak checks in openssl.c and in some 6185 test programs. 6186 [Chad C. Mulligan, Bodo Moeller] 6187 6188 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess 6189 up the length of negative integers. This has now been simplified to just 6190 store the length when it is first determined and use it later, rather 6191 than trying to keep track of where data is copied and updating it to 6192 point to the end. 6193 [Steve Henson, reported by Brien Wheeler 6194 <bwheeler@authentica-security.com>] 6195 6196 *) Add a new function PKCS7_signatureVerify. This allows the verification 6197 of a PKCS#7 signature but with the signing certificate passed to the 6198 function itself. This contrasts with PKCS7_dataVerify which assumes the 6199 certificate is present in the PKCS#7 structure. This isn't always the 6200 case: certificates can be omitted from a PKCS#7 structure and be 6201 distributed by "out of band" means (such as a certificate database). 6202 [Steve Henson] 6203 6204 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the 6205 function prototypes in pem.h, also change util/mkdef.pl to add the 6206 necessary function names. 6207 [Steve Henson] 6208 6209 *) mk1mf.pl (used by Windows builds) did not properly read the 6210 options set by Configure in the top level Makefile, and Configure 6211 was not even able to write more than one option correctly. 6212 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended. 6213 [Bodo Moeller] 6214 6215 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config 6216 file to be loaded from a BIO or FILE pointer. The BIO version will 6217 for example allow memory BIOs to contain config info. 6218 [Steve Henson] 6219 6220 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS. 6221 Whoever hopes to achieve shared-library compatibility across versions 6222 must use this, not the compile-time macro. 6223 (Exercise 0.9.4: Which is the minimum library version required by 6224 such programs?) 6225 Note: All this applies only to multi-threaded programs, others don't 6226 need locks. 6227 [Bodo Moeller] 6228 6229 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests 6230 through a BIO pair triggered the default case, i.e. 6231 SSLerr(...,SSL_R_UNKNOWN_STATE). 6232 [Bodo Moeller] 6233 6234 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications 6235 can use the SSL library even if none of the specific BIOs is 6236 appropriate. 6237 [Bodo Moeller] 6238 6239 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value 6240 for the encoded length. 6241 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>] 6242 6243 *) Add initial documentation of the X509V3 functions. 6244 [Steve Henson] 6245 6246 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and 6247 PEM_write_bio_PKCS8PrivateKey() that are equivalent to 6248 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more 6249 secure PKCS#8 private key format with a high iteration count. 6250 [Steve Henson] 6251 6252 *) Fix determination of Perl interpreter: A perl or perl5 6253 _directory_ in $PATH was also accepted as the interpreter. 6254 [Ralf S. Engelschall] 6255 6256 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking 6257 wrong with it but it was very old and did things like calling 6258 PEM_ASN1_read() directly and used MD5 for the hash not to mention some 6259 unusual formatting. 6260 [Steve Henson] 6261 6262 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed 6263 to use the new extension code. 6264 [Steve Henson] 6265 6266 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c 6267 with macros. This should make it easier to change their form, add extra 6268 arguments etc. Fix a few PEM prototypes which didn't have cipher as a 6269 constant. 6270 [Steve Henson] 6271 6272 *) Add to configuration table a new entry that can specify an alternative 6273 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep, 6274 according to Mark Crispin <MRC@Panda.COM>. 6275 [Bodo Moeller] 6276 6277#if 0 6278 *) DES CBC did not update the IV. Weird. 6279 [Ben Laurie] 6280#else 6281 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does. 6282 Changing the behaviour of the former might break existing programs -- 6283 where IV updating is needed, des_ncbc_encrypt can be used. 6284#endif 6285 6286 *) When bntest is run from "make test" it drives bc to check its 6287 calculations, as well as internally checking them. If an internal check 6288 fails, it needs to cause bc to give a non-zero result or make test carries 6289 on without noticing the failure. Fixed. 6290 [Ben Laurie] 6291 6292 *) DES library cleanups. 6293 [Ulf M�ller] 6294 6295 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be 6296 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit 6297 ciphers. NOTE: although the key derivation function has been verified 6298 against some published test vectors it has not been extensively tested 6299 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use 6300 of v2.0. 6301 [Steve Henson] 6302 6303 *) Instead of "mkdir -p", which is not fully portable, use new 6304 Perl script "util/mkdir-p.pl". 6305 [Bodo Moeller] 6306 6307 *) Rewrite the way password based encryption (PBE) is handled. It used to 6308 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter 6309 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms 6310 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now 6311 the 'parameter' field of the AlgorithmIdentifier is passed to the 6312 underlying key generation function so it must do its own ASN1 parsing. 6313 This has also changed the EVP_PBE_CipherInit() function which now has a 6314 'parameter' argument instead of literal salt and iteration count values 6315 and the function EVP_PBE_ALGOR_CipherInit() has been deleted. 6316 [Steve Henson] 6317 6318 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms 6319 and PKCS#8 functionality. New 'pkcs8' application linked to openssl. 6320 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE 6321 KEY" because this clashed with PKCS#8 unencrypted string. Since this 6322 value was just used as a "magic string" and not used directly its 6323 value doesn't matter. 6324 [Steve Henson] 6325 6326 *) Introduce some semblance of const correctness to BN. Shame C doesn't 6327 support mutable. 6328 [Ben Laurie] 6329 6330 *) "linux-sparc64" configuration (ultrapenguin). 6331 [Ray Miller <ray.miller@oucs.ox.ac.uk>] 6332 "linux-sparc" configuration. 6333 [Christian Forster <fo@hawo.stw.uni-erlangen.de>] 6334 6335 *) config now generates no-xxx options for missing ciphers. 6336 [Ulf M�ller] 6337 6338 *) Support the EBCDIC character set (work in progress). 6339 File ebcdic.c not yet included because it has a different license. 6340 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>] 6341 6342 *) Support BS2000/OSD-POSIX. 6343 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>] 6344 6345 *) Make callbacks for key generation use void * instead of char *. 6346 [Ben Laurie] 6347 6348 *) Make S/MIME samples compile (not yet tested). 6349 [Ben Laurie] 6350 6351 *) Additional typesafe stacks. 6352 [Ben Laurie] 6353 6354 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x). 6355 [Bodo Moeller] 6356 6357 6358 Changes between 0.9.3 and 0.9.3a [29 May 1999] 6359 6360 *) New configuration variant "sco5-gcc". 6361 6362 *) Updated some demos. 6363 [Sean O Riordain, Wade Scholine] 6364 6365 *) Add missing BIO_free at exit of pkcs12 application. 6366 [Wu Zhigang] 6367 6368 *) Fix memory leak in conf.c. 6369 [Steve Henson] 6370 6371 *) Updates for Win32 to assembler version of MD5. 6372 [Steve Henson] 6373 6374 *) Set #! path to perl in apps/der_chop to where we found it 6375 instead of using a fixed path. 6376 [Bodo Moeller] 6377 6378 *) SHA library changes for irix64-mips4-cc. 6379 [Andy Polyakov] 6380 6381 *) Improvements for VMS support. 6382 [Richard Levitte] 6383 6384 6385 Changes between 0.9.2b and 0.9.3 [24 May 1999] 6386 6387 *) Bignum library bug fix. IRIX 6 passes "make test" now! 6388 This also avoids the problems with SC4.2 and unpatched SC5. 6389 [Andy Polyakov <appro@fy.chalmers.se>] 6390 6391 *) New functions sk_num, sk_value and sk_set to replace the previous macros. 6392 These are required because of the typesafe stack would otherwise break 6393 existing code. If old code used a structure member which used to be STACK 6394 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with 6395 sk_num or sk_value it would produce an error because the num, data members 6396 are not present in STACK_OF. Now it just produces a warning. sk_set 6397 replaces the old method of assigning a value to sk_value 6398 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code 6399 that does this will no longer work (and should use sk_set instead) but 6400 this could be regarded as a "questionable" behaviour anyway. 6401 [Steve Henson] 6402 6403 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now 6404 correctly handle encrypted S/MIME data. 6405 [Steve Henson] 6406 6407 *) Change type of various DES function arguments from des_cblock 6408 (which means, in function argument declarations, pointer to char) 6409 to des_cblock * (meaning pointer to array with 8 char elements), 6410 which allows the compiler to do more typechecking; it was like 6411 that back in SSLeay, but with lots of ugly casts. 6412 6413 Introduce new type const_des_cblock. 6414 [Bodo Moeller] 6415 6416 *) Reorganise the PKCS#7 library and get rid of some of the more obvious 6417 problems: find RecipientInfo structure that matches recipient certificate 6418 and initialise the ASN1 structures properly based on passed cipher. 6419 [Steve Henson] 6420 6421 *) Belatedly make the BN tests actually check the results. 6422 [Ben Laurie] 6423 6424 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion 6425 to and from BNs: it was completely broken. New compilation option 6426 NEG_PUBKEY_BUG to allow for some broken certificates that encode public 6427 key elements as negative integers. 6428 [Steve Henson] 6429 6430 *) Reorganize and speed up MD5. 6431 [Andy Polyakov <appro@fy.chalmers.se>] 6432 6433 *) VMS support. 6434 [Richard Levitte <richard@levitte.org>] 6435 6436 *) New option -out to asn1parse to allow the parsed structure to be 6437 output to a file. This is most useful when combined with the -strparse 6438 option to examine the output of things like OCTET STRINGS. 6439 [Steve Henson] 6440 6441 *) Make SSL library a little more fool-proof by not requiring any longer 6442 that SSL_set_{accept,connect}_state be called before 6443 SSL_{accept,connect} may be used (SSL_set_..._state is omitted 6444 in many applications because usually everything *appeared* to work as 6445 intended anyway -- now it really works as intended). 6446 [Bodo Moeller] 6447 6448 *) Move openssl.cnf out of lib/. 6449 [Ulf M�ller] 6450 6451 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall 6452 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes 6453 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+ 6454 [Ralf S. Engelschall] 6455 6456 *) Various fixes to the EVP and PKCS#7 code. It may now be able to 6457 handle PKCS#7 enveloped data properly. 6458 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve] 6459 6460 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of 6461 copying pointers. The cert_st handling is changed by this in 6462 various ways (and thus what used to be known as ctx->default_cert 6463 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert 6464 any longer when s->cert does not give us what we need). 6465 ssl_cert_instantiate becomes obsolete by this change. 6466 As soon as we've got the new code right (possibly it already is?), 6467 we have solved a couple of bugs of the earlier code where s->cert 6468 was used as if it could not have been shared with other SSL structures. 6469 6470 Note that using the SSL API in certain dirty ways now will result 6471 in different behaviour than observed with earlier library versions: 6472 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx) 6473 does not influence s as it used to. 6474 6475 In order to clean up things more thoroughly, inside SSL_SESSION 6476 we don't use CERT any longer, but a new structure SESS_CERT 6477 that holds per-session data (if available); currently, this is 6478 the peer's certificate chain and, for clients, the server's certificate 6479 and temporary key. CERT holds only those values that can have 6480 meaningful defaults in an SSL_CTX. 6481 [Bodo Moeller] 6482 6483 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure 6484 from the internal representation. Various PKCS#7 fixes: remove some 6485 evil casts and set the enc_dig_alg field properly based on the signing 6486 key type. 6487 [Steve Henson] 6488 6489 *) Allow PKCS#12 password to be set from the command line or the 6490 environment. Let 'ca' get its config file name from the environment 6491 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req' 6492 and 'x509'). 6493 [Steve Henson] 6494 6495 *) Allow certificate policies extension to use an IA5STRING for the 6496 organization field. This is contrary to the PKIX definition but 6497 VeriSign uses it and IE5 only recognises this form. Document 'x509' 6498 extension option. 6499 [Steve Henson] 6500 6501 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic, 6502 without disallowing inline assembler and the like for non-pedantic builds. 6503 [Ben Laurie] 6504 6505 *) Support Borland C++ builder. 6506 [Janez Jere <jj@void.si>, modified by Ulf M�ller] 6507 6508 *) Support Mingw32. 6509 [Ulf M�ller] 6510 6511 *) SHA-1 cleanups and performance enhancements. 6512 [Andy Polyakov <appro@fy.chalmers.se>] 6513 6514 *) Sparc v8plus assembler for the bignum library. 6515 [Andy Polyakov <appro@fy.chalmers.se>] 6516 6517 *) Accept any -xxx and +xxx compiler options in Configure. 6518 [Ulf M�ller] 6519 6520 *) Update HPUX configuration. 6521 [Anonymous] 6522 6523 *) Add missing sk_<type>_unshift() function to safestack.h 6524 [Ralf S. Engelschall] 6525 6526 *) New function SSL_CTX_use_certificate_chain_file that sets the 6527 "extra_cert"s in addition to the certificate. (This makes sense 6528 only for "PEM" format files, as chains as a whole are not 6529 DER-encoded.) 6530 [Bodo Moeller] 6531 6532 *) Support verify_depth from the SSL API. 6533 x509_vfy.c had what can be considered an off-by-one-error: 6534 Its depth (which was not part of the external interface) 6535 was actually counting the number of certificates in a chain; 6536 now it really counts the depth. 6537 [Bodo Moeller] 6538 6539 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used 6540 instead of X509err, which often resulted in confusing error 6541 messages since the error codes are not globally unique 6542 (e.g. an alleged error in ssl3_accept when a certificate 6543 didn't match the private key). 6544 6545 *) New function SSL_CTX_set_session_id_context that allows to set a default 6546 value (so that you don't need SSL_set_session_id_context for each 6547 connection using the SSL_CTX). 6548 [Bodo Moeller] 6549 6550 *) OAEP decoding bug fix. 6551 [Ulf M�ller] 6552 6553 *) Support INSTALL_PREFIX for package builders, as proposed by 6554 David Harris. 6555 [Bodo Moeller] 6556 6557 *) New Configure options "threads" and "no-threads". For systems 6558 where the proper compiler options are known (currently Solaris 6559 and Linux), "threads" is the default. 6560 [Bodo Moeller] 6561 6562 *) New script util/mklink.pl as a faster substitute for util/mklink.sh. 6563 [Bodo Moeller] 6564 6565 *) Install various scripts to $(OPENSSLDIR)/misc, not to 6566 $(INSTALLTOP)/bin -- they shouldn't clutter directories 6567 such as /usr/local/bin. 6568 [Bodo Moeller] 6569 6570 *) "make linux-shared" to build shared libraries. 6571 [Niels Poppe <niels@netbox.org>] 6572 6573 *) New Configure option no-<cipher> (rsa, idea, rc5, ...). 6574 [Ulf M�ller] 6575 6576 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for 6577 extension adding in x509 utility. 6578 [Steve Henson] 6579 6580 *) Remove NOPROTO sections and error code comments. 6581 [Ulf M�ller] 6582 6583 *) Partial rewrite of the DEF file generator to now parse the ANSI 6584 prototypes. 6585 [Steve Henson] 6586 6587 *) New Configure options --prefix=DIR and --openssldir=DIR. 6588 [Ulf M�ller] 6589 6590 *) Complete rewrite of the error code script(s). It is all now handled 6591 by one script at the top level which handles error code gathering, 6592 header rewriting and C source file generation. It should be much better 6593 than the old method: it now uses a modified version of Ulf's parser to 6594 read the ANSI prototypes in all header files (thus the old K&R definitions 6595 aren't needed for error creation any more) and do a better job of 6596 translating function codes into names. The old 'ASN1 error code imbedded 6597 in a comment' is no longer necessary and it doesn't use .err files which 6598 have now been deleted. Also the error code call doesn't have to appear all 6599 on one line (which resulted in some large lines...). 6600 [Steve Henson] 6601 6602 *) Change #include filenames from <foo.h> to <openssl/foo.h>. 6603 [Bodo Moeller] 6604 6605 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return 6606 0 (which usually indicates a closed connection), but continue reading. 6607 [Bodo Moeller] 6608 6609 *) Fix some race conditions. 6610 [Bodo Moeller] 6611 6612 *) Add support for CRL distribution points extension. Add Certificate 6613 Policies and CRL distribution points documentation. 6614 [Steve Henson] 6615 6616 *) Move the autogenerated header file parts to crypto/opensslconf.h. 6617 [Ulf M�ller] 6618 6619 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of 6620 8 of keying material. Merlin has also confirmed interop with this fix 6621 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0. 6622 [Merlin Hughes <merlin@baltimore.ie>] 6623 6624 *) Fix lots of warnings. 6625 [Richard Levitte <levitte@stacken.kth.se>] 6626 6627 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if 6628 the directory spec didn't end with a LIST_SEPARATOR_CHAR. 6629 [Richard Levitte <levitte@stacken.kth.se>] 6630 6631 *) Fix problems with sizeof(long) == 8. 6632 [Andy Polyakov <appro@fy.chalmers.se>] 6633 6634 *) Change functions to ANSI C. 6635 [Ulf M�ller] 6636 6637 *) Fix typos in error codes. 6638 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf M�ller] 6639 6640 *) Remove defunct assembler files from Configure. 6641 [Ulf M�ller] 6642 6643 *) SPARC v8 assembler BIGNUM implementation. 6644 [Andy Polyakov <appro@fy.chalmers.se>] 6645 6646 *) Support for Certificate Policies extension: both print and set. 6647 Various additions to support the r2i method this uses. 6648 [Steve Henson] 6649 6650 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could 6651 return a const string when you are expecting an allocated buffer. 6652 [Ben Laurie] 6653 6654 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE 6655 types DirectoryString and DisplayText. 6656 [Steve Henson] 6657 6658 *) Add code to allow r2i extensions to access the configuration database, 6659 add an LHASH database driver and add several ctx helper functions. 6660 [Steve Henson] 6661 6662 *) Fix an evil bug in bn_expand2() which caused various BN functions to 6663 fail when they extended the size of a BIGNUM. 6664 [Steve Henson] 6665 6666 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to 6667 support typesafe stack. 6668 [Steve Henson] 6669 6670 *) Fix typo in SSL_[gs]et_options(). 6671 [Nils Frostberg <nils@medcom.se>] 6672 6673 *) Delete various functions and files that belonged to the (now obsolete) 6674 old X509V3 handling code. 6675 [Steve Henson] 6676 6677 *) New Configure option "rsaref". 6678 [Ulf M�ller] 6679 6680 *) Don't auto-generate pem.h. 6681 [Bodo Moeller] 6682 6683 *) Introduce type-safe ASN.1 SETs. 6684 [Ben Laurie] 6685 6686 *) Convert various additional casted stacks to type-safe STACK_OF() variants. 6687 [Ben Laurie, Ralf S. Engelschall, Steve Henson] 6688 6689 *) Introduce type-safe STACKs. This will almost certainly break lots of code 6690 that links with OpenSSL (well at least cause lots of warnings), but fear 6691 not: the conversion is trivial, and it eliminates loads of evil casts. A 6692 few STACKed things have been converted already. Feel free to convert more. 6693 In the fullness of time, I'll do away with the STACK type altogether. 6694 [Ben Laurie] 6695 6696 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate 6697 specified in <certfile> by updating the entry in the index.txt file. 6698 This way one no longer has to edit the index.txt file manually for 6699 revoking a certificate. The -revoke option does the gory details now. 6700 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall] 6701 6702 *) Fix `openssl crl -noout -text' combination where `-noout' killed the 6703 `-text' option at all and this way the `-noout -text' combination was 6704 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'. 6705 [Ralf S. Engelschall] 6706 6707 *) Make sure a corresponding plain text error message exists for the 6708 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a 6709 verify callback function determined that a certificate was revoked. 6710 [Ralf S. Engelschall] 6711 6712 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for 6713 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test 6714 all available cipers including rc5, which was forgotten until now. 6715 In order to let the testing shell script know which algorithms 6716 are available, a new (up to now undocumented) command 6717 "openssl list-cipher-commands" is used. 6718 [Bodo Moeller] 6719 6720 *) Bugfix: s_client occasionally would sleep in select() when 6721 it should have checked SSL_pending() first. 6722 [Bodo Moeller] 6723 6724 *) New functions DSA_do_sign and DSA_do_verify to provide access to 6725 the raw DSA values prior to ASN.1 encoding. 6726 [Ulf M�ller] 6727 6728 *) Tweaks to Configure 6729 [Niels Poppe <niels@netbox.org>] 6730 6731 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support, 6732 yet... 6733 [Steve Henson] 6734 6735 *) New variables $(RANLIB) and $(PERL) in the Makefiles. 6736 [Ulf M�ller] 6737 6738 *) New config option to avoid instructions that are illegal on the 80386. 6739 The default code is faster, but requires at least a 486. 6740 [Ulf M�ller] 6741 6742 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and 6743 SSL2_SERVER_VERSION (not used at all) macros, which are now the 6744 same as SSL2_VERSION anyway. 6745 [Bodo Moeller] 6746 6747 *) New "-showcerts" option for s_client. 6748 [Bodo Moeller] 6749 6750 *) Still more PKCS#12 integration. Add pkcs12 application to openssl 6751 application. Various cleanups and fixes. 6752 [Steve Henson] 6753 6754 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and 6755 modify error routines to work internally. Add error codes and PBE init 6756 to library startup routines. 6757 [Steve Henson] 6758 6759 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and 6760 packing functions to asn1 and evp. Changed function names and error 6761 codes along the way. 6762 [Steve Henson] 6763 6764 *) PKCS12 integration: and so it begins... First of several patches to 6765 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12 6766 objects to objects.h 6767 [Steve Henson] 6768 6769 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1 6770 and display support for Thawte strong extranet extension. 6771 [Steve Henson] 6772 6773 *) Add LinuxPPC support. 6774 [Jeff Dubrule <igor@pobox.org>] 6775 6776 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to 6777 bn_div_words in alpha.s. 6778 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie] 6779 6780 *) Make sure the RSA OAEP test is skipped under -DRSAref because 6781 OAEP isn't supported when OpenSSL is built with RSAref. 6782 [Ulf Moeller <ulf@fitug.de>] 6783 6784 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h 6785 so they no longer are missing under -DNOPROTO. 6786 [Soren S. Jorvang <soren@t.dk>] 6787 6788 6789 Changes between 0.9.1c and 0.9.2b [22 Mar 1999] 6790 6791 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still 6792 doesn't work when the session is reused. Coming soon! 6793 [Ben Laurie] 6794 6795 *) Fix a security hole, that allows sessions to be reused in the wrong 6796 context thus bypassing client cert protection! All software that uses 6797 client certs and session caches in multiple contexts NEEDS PATCHING to 6798 allow session reuse! A fuller solution is in the works. 6799 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)] 6800 6801 *) Some more source tree cleanups (removed obsolete files 6802 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed 6803 permission on "config" script to be executable) and a fix for the INSTALL 6804 document. 6805 [Ulf Moeller <ulf@fitug.de>] 6806 6807 *) Remove some legacy and erroneous uses of malloc, free instead of 6808 Malloc, Free. 6809 [Lennart Bang <lob@netstream.se>, with minor changes by Steve] 6810 6811 *) Make rsa_oaep_test return non-zero on error. 6812 [Ulf Moeller <ulf@fitug.de>] 6813 6814 *) Add support for native Solaris shared libraries. Configure 6815 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice 6816 if someone would make that last step automatic. 6817 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>] 6818 6819 *) ctx_size was not built with the right compiler during "make links". Fixed. 6820 [Ben Laurie] 6821 6822 *) Change the meaning of 'ALL' in the cipher list. It now means "everything 6823 except NULL ciphers". This means the default cipher list will no longer 6824 enable NULL ciphers. They need to be specifically enabled e.g. with 6825 the string "DEFAULT:eNULL". 6826 [Steve Henson] 6827 6828 *) Fix to RSA private encryption routines: if p < q then it would 6829 occasionally produce an invalid result. This will only happen with 6830 externally generated keys because OpenSSL (and SSLeay) ensure p > q. 6831 [Steve Henson] 6832 6833 *) Be less restrictive and allow also `perl util/perlpath.pl 6834 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin', 6835 because this way one can also use an interpreter named `perl5' (which is 6836 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still 6837 installed as `perl'). 6838 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>] 6839 6840 *) Let util/clean-depend.pl work also with older Perl 5.00x versions. 6841 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>] 6842 6843 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add 6844 advapi32.lib to Win32 build and change the pem test comparision 6845 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the 6846 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h 6847 and crypto/des/ede_cbcm_enc.c. 6848 [Steve Henson] 6849 6850 *) DES quad checksum was broken on big-endian architectures. Fixed. 6851 [Ben Laurie] 6852 6853 *) Comment out two functions in bio.h that aren't implemented. Fix up the 6854 Win32 test batch file so it (might) work again. The Win32 test batch file 6855 is horrible: I feel ill.... 6856 [Steve Henson] 6857 6858 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected 6859 in e_os.h. Audit of header files to check ANSI and non ANSI 6860 sections: 10 functions were absent from non ANSI section and not exported 6861 from Windows DLLs. Fixed up libeay.num for new functions. 6862 [Steve Henson] 6863 6864 *) Make `openssl version' output lines consistent. 6865 [Ralf S. Engelschall] 6866 6867 *) Fix Win32 symbol export lists for BIO functions: Added 6868 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data 6869 to ms/libeay{16,32}.def. 6870 [Ralf S. Engelschall] 6871 6872 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled 6873 fine under Unix and passes some trivial tests I've now added. But the 6874 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was 6875 added to make sure no one expects that this stuff really works in the 6876 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources 6877 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and 6878 openssl_bio.xs. 6879 [Ralf S. Engelschall] 6880 6881 *) Fix the generation of two part addresses in perl. 6882 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie] 6883 6884 *) Add config entry for Linux on MIPS. 6885 [John Tobey <jtobey@channel1.com>] 6886 6887 *) Make links whenever Configure is run, unless we are on Windoze. 6888 [Ben Laurie] 6889 6890 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf. 6891 Currently only issuerAltName and AuthorityKeyIdentifier make any sense 6892 in CRLs. 6893 [Steve Henson] 6894 6895 *) Add a useful kludge to allow package maintainers to specify compiler and 6896 other platforms details on the command line without having to patch the 6897 Configure script everytime: One now can use ``perl Configure 6898 <id>:<details>'', i.e. platform ids are allowed to have details appended 6899 to them (seperated by colons). This is treated as there would be a static 6900 pre-configured entry in Configure's %table under key <id> with value 6901 <details> and ``perl Configure <id>'' is called. So, when you want to 6902 perform a quick test-compile under FreeBSD 3.1 with pgcc and without 6903 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"'' 6904 now, which overrides the FreeBSD-elf entry on-the-fly. 6905 [Ralf S. Engelschall] 6906 6907 *) Disable new TLS1 ciphersuites by default: they aren't official yet. 6908 [Ben Laurie] 6909 6910 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified 6911 on the `perl Configure ...' command line. This way one can compile 6912 OpenSSL libraries with Position Independent Code (PIC) which is needed 6913 for linking it into DSOs. 6914 [Ralf S. Engelschall] 6915 6916 *) Remarkably, export ciphers were totally broken and no-one had noticed! 6917 Fixed. 6918 [Ben Laurie] 6919 6920 *) Cleaned up the LICENSE document: The official contact for any license 6921 questions now is the OpenSSL core team under openssl-core@openssl.org. 6922 And add a paragraph about the dual-license situation to make sure people 6923 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply 6924 to the OpenSSL toolkit. 6925 [Ralf S. Engelschall] 6926 6927 *) General source tree makefile cleanups: Made `making xxx in yyy...' 6928 display consistent in the source tree and replaced `/bin/rm' by `rm'. 6929 Additonally cleaned up the `make links' target: Remove unnecessary 6930 semicolons, subsequent redundant removes, inline point.sh into mklink.sh 6931 to speed processing and no longer clutter the display with confusing 6932 stuff. Instead only the actually done links are displayed. 6933 [Ralf S. Engelschall] 6934 6935 *) Permit null encryption ciphersuites, used for authentication only. It used 6936 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this. 6937 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null 6938 encryption. 6939 [Ben Laurie] 6940 6941 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder 6942 signed attributes when verifying signatures (this would break them), 6943 the detached data encoding was wrong and public keys obtained using 6944 X509_get_pubkey() weren't freed. 6945 [Steve Henson] 6946 6947 *) Add text documentation for the BUFFER functions. Also added a work around 6948 to a Win95 console bug. This was triggered by the password read stuff: the 6949 last character typed gets carried over to the next fread(). If you were 6950 generating a new cert request using 'req' for example then the last 6951 character of the passphrase would be CR which would then enter the first 6952 field as blank. 6953 [Steve Henson] 6954 6955 *) Added the new `Includes OpenSSL Cryptography Software' button as 6956 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay 6957 button and can be used by applications based on OpenSSL to show the 6958 relationship to the OpenSSL project. 6959 [Ralf S. Engelschall] 6960 6961 *) Remove confusing variables in function signatures in files 6962 ssl/ssl_lib.c and ssl/ssl.h. 6963 [Lennart Bong <lob@kulthea.stacken.kth.se>] 6964 6965 *) Don't install bss_file.c under PREFIX/include/ 6966 [Lennart Bong <lob@kulthea.stacken.kth.se>] 6967 6968 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle 6969 functions that return function pointers and has support for NT specific 6970 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various 6971 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from 6972 unsigned to signed types: this was killing the Win32 compile. 6973 [Steve Henson] 6974 6975 *) Add new certificate file to stack functions, 6976 SSL_add_dir_cert_subjects_to_stack() and 6977 SSL_add_file_cert_subjects_to_stack(). These largely supplant 6978 SSL_load_client_CA_file(), and can be used to add multiple certs easily 6979 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()). 6980 This means that Apache-SSL and similar packages don't have to mess around 6981 to add as many CAs as they want to the preferred list. 6982 [Ben Laurie] 6983 6984 *) Experiment with doxygen documentation. Currently only partially applied to 6985 ssl/ssl_lib.c. 6986 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with 6987 openssl.doxy as the configuration file. 6988 [Ben Laurie] 6989 6990 *) Get rid of remaining C++-style comments which strict C compilers hate. 6991 [Ralf S. Engelschall, pointed out by Carlos Amengual] 6992 6993 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not 6994 compiled in by default: it has problems with large keys. 6995 [Steve Henson] 6996 6997 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and 6998 DH private keys and/or callback functions which directly correspond to 6999 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This 7000 is needed for applications which have to configure certificates on a 7001 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis 7002 (e.g. s_server). 7003 For the RSA certificate situation is makes no difference, but 7004 for the DSA certificate situation this fixes the "no shared cipher" 7005 problem where the OpenSSL cipher selection procedure failed because the 7006 temporary keys were not overtaken from the context and the API provided 7007 no way to reconfigure them. 7008 The new functions now let applications reconfigure the stuff and they 7009 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh, 7010 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new 7011 non-public-API function ssl_cert_instantiate() is used as a helper 7012 function and also to reduce code redundancy inside ssl_rsa.c. 7013 [Ralf S. Engelschall] 7014 7015 *) Move s_server -dcert and -dkey options out of the undocumented feature 7016 area because they are useful for the DSA situation and should be 7017 recognized by the users. 7018 [Ralf S. Engelschall] 7019 7020 *) Fix the cipher decision scheme for export ciphers: the export bits are 7021 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within 7022 SSL_EXP_MASK. So, the original variable has to be used instead of the 7023 already masked variable. 7024 [Richard Levitte <levitte@stacken.kth.se>] 7025 7026 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c 7027 [Richard Levitte <levitte@stacken.kth.se>] 7028 7029 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal() 7030 from `int' to `unsigned int' because it's a length and initialized by 7031 EVP_DigestFinal() which expects an `unsigned int *'. 7032 [Richard Levitte <levitte@stacken.kth.se>] 7033 7034 *) Don't hard-code path to Perl interpreter on shebang line of Configure 7035 script. Instead use the usual Shell->Perl transition trick. 7036 [Ralf S. Engelschall] 7037 7038 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates 7039 (in addition to RSA certificates) to match the behaviour of `openssl dsa 7040 -noout -modulus' as it's already the case for `openssl rsa -noout 7041 -modulus'. For RSA the -modulus is the real "modulus" while for DSA 7042 currently the public key is printed (a decision which was already done by 7043 `openssl dsa -modulus' in the past) which serves a similar purpose. 7044 Additionally the NO_RSA no longer completely removes the whole -modulus 7045 option; it now only avoids using the RSA stuff. Same applies to NO_DSA 7046 now, too. 7047 [Ralf S. Engelschall] 7048 7049 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested 7050 BIO. See the source (crypto/evp/bio_ok.c) for more info. 7051 [Arne Ansper <arne@ats.cyber.ee>] 7052 7053 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs 7054 to be added. Now both 'req' and 'ca' can use new objects defined in the 7055 config file. 7056 [Steve Henson] 7057 7058 *) Add cool BIO that does syslog (or event log on NT). 7059 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie] 7060 7061 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5, 7062 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and 7063 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher 7064 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt. 7065 [Ben Laurie] 7066 7067 *) Add preliminary config info for new extension code. 7068 [Steve Henson] 7069 7070 *) Make RSA_NO_PADDING really use no padding. 7071 [Ulf Moeller <ulf@fitug.de>] 7072 7073 *) Generate errors when private/public key check is done. 7074 [Ben Laurie] 7075 7076 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support 7077 for some CRL extensions and new objects added. 7078 [Steve Henson] 7079 7080 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private 7081 key usage extension and fuller support for authority key id. 7082 [Steve Henson] 7083 7084 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved 7085 padding method for RSA, which is recommended for new applications in PKCS 7086 #1 v2.0 (RFC 2437, October 1998). 7087 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical 7088 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure 7089 against Bleichbacher's attack on RSA. 7090 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by 7091 Ben Laurie] 7092 7093 *) Updates to the new SSL compression code 7094 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)] 7095 7096 *) Fix so that the version number in the master secret, when passed 7097 via RSA, checks that if TLS was proposed, but we roll back to SSLv3 7098 (because the server will not accept higher), that the version number 7099 is 0x03,0x01, not 0x03,0x00 7100 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)] 7101 7102 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory 7103 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes 7104 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c 7105 [Steve Henson] 7106 7107 *) Support for RAW extensions where an arbitrary extension can be 7108 created by including its DER encoding. See apps/openssl.cnf for 7109 an example. 7110 [Steve Henson] 7111 7112 *) Make sure latest Perl versions don't interpret some generated C array 7113 code as Perl array code in the crypto/err/err_genc.pl script. 7114 [Lars Weber <3weber@informatik.uni-hamburg.de>] 7115 7116 *) Modify ms/do_ms.bat to not generate assembly language makefiles since 7117 not many people have the assembler. Various Win32 compilation fixes and 7118 update to the INSTALL.W32 file with (hopefully) more accurate Win32 7119 build instructions. 7120 [Steve Henson] 7121 7122 *) Modify configure script 'Configure' to automatically create crypto/date.h 7123 file under Win32 and also build pem.h from pem.org. New script 7124 util/mkfiles.pl to create the MINFO file on environments that can't do a 7125 'make files': perl util/mkfiles.pl >MINFO should work. 7126 [Steve Henson] 7127 7128 *) Major rework of DES function declarations, in the pursuit of correctness 7129 and purity. As a result, many evil casts evaporated, and some weirdness, 7130 too. You may find this causes warnings in your code. Zapping your evil 7131 casts will probably fix them. Mostly. 7132 [Ben Laurie] 7133 7134 *) Fix for a typo in asn1.h. Bug fix to object creation script 7135 obj_dat.pl. It considered a zero in an object definition to mean 7136 "end of object": none of the objects in objects.h have any zeros 7137 so it wasn't spotted. 7138 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>] 7139 7140 *) Add support for Triple DES Cipher Block Chaining with Output Feedback 7141 Masking (CBCM). In the absence of test vectors, the best I have been able 7142 to do is check that the decrypt undoes the encrypt, so far. Send me test 7143 vectors if you have them. 7144 [Ben Laurie] 7145 7146 *) Correct calculation of key length for export ciphers (too much space was 7147 allocated for null ciphers). This has not been tested! 7148 [Ben Laurie] 7149 7150 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage 7151 message is now correct (it understands "crypto" and "ssl" on its 7152 command line). There is also now an "update" option. This will update 7153 the util/ssleay.num and util/libeay.num files with any new functions. 7154 If you do a: 7155 perl util/mkdef.pl crypto ssl update 7156 it will update them. 7157 [Steve Henson] 7158 7159 *) Overhauled the Perl interface (perl/*): 7160 - ported BN stuff to OpenSSL's different BN library 7161 - made the perl/ source tree CVS-aware 7162 - renamed the package from SSLeay to OpenSSL (the files still contain 7163 their history because I've copied them in the repository) 7164 - removed obsolete files (the test scripts will be replaced 7165 by better Test::Harness variants in the future) 7166 [Ralf S. Engelschall] 7167 7168 *) First cut for a very conservative source tree cleanup: 7169 1. merge various obsolete readme texts into doc/ssleay.txt 7170 where we collect the old documents and readme texts. 7171 2. remove the first part of files where I'm already sure that we no 7172 longer need them because of three reasons: either they are just temporary 7173 files which were left by Eric or they are preserved original files where 7174 I've verified that the diff is also available in the CVS via "cvs diff 7175 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for 7176 the crypto/md/ stuff). 7177 [Ralf S. Engelschall] 7178 7179 *) More extension code. Incomplete support for subject and issuer alt 7180 name, issuer and authority key id. Change the i2v function parameters 7181 and add an extra 'crl' parameter in the X509V3_CTX structure: guess 7182 what that's for :-) Fix to ASN1 macro which messed up 7183 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED. 7184 [Steve Henson] 7185 7186 *) Preliminary support for ENUMERATED type. This is largely copied from the 7187 INTEGER code. 7188 [Steve Henson] 7189 7190 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy. 7191 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)] 7192 7193 *) Make sure `make rehash' target really finds the `openssl' program. 7194 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>] 7195 7196 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd 7197 like to hear about it if this slows down other processors. 7198 [Ben Laurie] 7199 7200 *) Add CygWin32 platform information to Configure script. 7201 [Alan Batie <batie@aahz.jf.intel.com>] 7202 7203 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm' 7204 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>] 7205 7206 *) New program nseq to manipulate netscape certificate sequences 7207 [Steve Henson] 7208 7209 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a 7210 few typos. 7211 [Steve Henson] 7212 7213 *) Fixes to BN code. Previously the default was to define BN_RECURSION 7214 but the BN code had some problems that would cause failures when 7215 doing certificate verification and some other functions. 7216 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)] 7217 7218 *) Add ASN1 and PEM code to support netscape certificate sequences. 7219 [Steve Henson] 7220 7221 *) Add ASN1 and PEM code to support netscape certificate sequences. 7222 [Steve Henson] 7223 7224 *) Add several PKIX and private extended key usage OIDs. 7225 [Steve Henson] 7226 7227 *) Modify the 'ca' program to handle the new extension code. Modify 7228 openssl.cnf for new extension format, add comments. 7229 [Steve Henson] 7230 7231 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req' 7232 and add a sample to openssl.cnf so req -x509 now adds appropriate 7233 CA extensions. 7234 [Steve Henson] 7235 7236 *) Continued X509 V3 changes. Add to other makefiles, integrate with the 7237 error code, add initial support to X509_print() and x509 application. 7238 [Steve Henson] 7239 7240 *) Takes a deep breath and start addding X509 V3 extension support code. Add 7241 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this 7242 stuff is currently isolated and isn't even compiled yet. 7243 [Steve Henson] 7244 7245 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL 7246 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print. 7247 Removed the versions check from X509 routines when loading extensions: 7248 this allows certain broken certificates that don't set the version 7249 properly to be processed. 7250 [Steve Henson] 7251 7252 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another 7253 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which 7254 can still be regenerated with "make depend". 7255 [Ben Laurie] 7256 7257 *) Spelling mistake in C version of CAST-128. 7258 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>] 7259 7260 *) Changes to the error generation code. The perl script err-code.pl 7261 now reads in the old error codes and retains the old numbers, only 7262 adding new ones if necessary. It also only changes the .err files if new 7263 codes are added. The makefiles have been modified to only insert errors 7264 when needed (to avoid needlessly modifying header files). This is done 7265 by only inserting errors if the .err file is newer than the auto generated 7266 C file. To rebuild all the error codes from scratch (the old behaviour) 7267 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl 7268 or delete all the .err files. 7269 [Steve Henson] 7270 7271 *) CAST-128 was incorrectly implemented for short keys. The C version has 7272 been fixed, but is untested. The assembler versions are also fixed, but 7273 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing 7274 to regenerate it if needed. 7275 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun 7276 Hagino <itojun@kame.net>] 7277 7278 *) File was opened incorrectly in randfile.c. 7279 [Ulf M�ller <ulf@fitug.de>] 7280 7281 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print 7282 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or 7283 GeneralizedTime. ASN1_TIME is the proper type used in certificates et 7284 al: it's just almost always a UTCTime. Note this patch adds new error 7285 codes so do a "make errors" if there are problems. 7286 [Steve Henson] 7287 7288 *) Correct Linux 1 recognition in config. 7289 [Ulf M�ller <ulf@fitug.de>] 7290 7291 *) Remove pointless MD5 hash when using DSA keys in ca. 7292 [Anonymous <nobody@replay.com>] 7293 7294 *) Generate an error if given an empty string as a cert directory. Also 7295 generate an error if handed NULL (previously returned 0 to indicate an 7296 error, but didn't set one). 7297 [Ben Laurie, reported by Anonymous <nobody@replay.com>] 7298 7299 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last. 7300 [Ben Laurie] 7301 7302 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct 7303 parameters. This was causing a warning which killed off the Win32 compile. 7304 [Steve Henson] 7305 7306 *) Remove C++ style comments from crypto/bn/bn_local.h. 7307 [Neil Costigan <neil.costigan@celocom.com>] 7308 7309 *) The function OBJ_txt2nid was broken. It was supposed to return a nid 7310 based on a text string, looking up short and long names and finally 7311 "dot" format. The "dot" format stuff didn't work. Added new function 7312 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote 7313 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the 7314 OID is not part of the table. 7315 [Steve Henson] 7316 7317 *) Add prototypes to X509 lookup/verify methods, fixing a bug in 7318 X509_LOOKUP_by_alias(). 7319 [Ben Laurie] 7320 7321 *) Sort openssl functions by name. 7322 [Ben Laurie] 7323 7324 *) Get the gendsa program working (hopefully) and add it to app list. Remove 7325 encryption from sample DSA keys (in case anyone is interested the password 7326 was "1234"). 7327 [Steve Henson] 7328 7329 *) Make _all_ *_free functions accept a NULL pointer. 7330 [Frans Heymans <fheymans@isaserver.be>] 7331 7332 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use 7333 NULL pointers. 7334 [Anonymous <nobody@replay.com>] 7335 7336 *) s_server should send the CAfile as acceptable CAs, not its own cert. 7337 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>] 7338 7339 *) Don't blow it for numeric -newkey arguments to apps/req. 7340 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>] 7341 7342 *) Temp key "for export" tests were wrong in s3_srvr.c. 7343 [Anonymous <nobody@replay.com>] 7344 7345 *) Add prototype for temp key callback functions 7346 SSL_CTX_set_tmp_{rsa,dh}_callback(). 7347 [Ben Laurie] 7348 7349 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and 7350 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey(). 7351 [Steve Henson] 7352 7353 *) X509_name_add_entry() freed the wrong thing after an error. 7354 [Arne Ansper <arne@ats.cyber.ee>] 7355 7356 *) rsa_eay.c would attempt to free a NULL context. 7357 [Arne Ansper <arne@ats.cyber.ee>] 7358 7359 *) BIO_s_socket() had a broken should_retry() on Windoze. 7360 [Arne Ansper <arne@ats.cyber.ee>] 7361 7362 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH. 7363 [Arne Ansper <arne@ats.cyber.ee>] 7364 7365 *) Make sure the already existing X509_STORE->depth variable is initialized 7366 in X509_STORE_new(), but document the fact that this variable is still 7367 unused in the certificate verification process. 7368 [Ralf S. Engelschall] 7369 7370 *) Fix the various library and apps files to free up pkeys obtained from 7371 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions. 7372 [Steve Henson] 7373 7374 *) Fix reference counting in X509_PUBKEY_get(). This makes 7375 demos/maurice/example2.c work, amongst others, probably. 7376 [Steve Henson and Ben Laurie] 7377 7378 *) First cut of a cleanup for apps/. First the `ssleay' program is now named 7379 `openssl' and second, the shortcut symlinks for the `openssl <command>' 7380 are no longer created. This way we have a single and consistent command 7381 line interface `openssl <command>', similar to `cvs <command>'. 7382 [Ralf S. Engelschall, Paul Sutton and Ben Laurie] 7383 7384 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey 7385 BIT STRING wrapper always have zero unused bits. 7386 [Steve Henson] 7387 7388 *) Add CA.pl, perl version of CA.sh, add extended key usage OID. 7389 [Steve Henson] 7390 7391 *) Make the top-level INSTALL documentation easier to understand. 7392 [Paul Sutton] 7393 7394 *) Makefiles updated to exit if an error occurs in a sub-directory 7395 make (including if user presses ^C) [Paul Sutton] 7396 7397 *) Make Montgomery context stuff explicit in RSA data structure. 7398 [Ben Laurie] 7399 7400 *) Fix build order of pem and err to allow for generated pem.h. 7401 [Ben Laurie] 7402 7403 *) Fix renumbering bug in X509_NAME_delete_entry(). 7404 [Ben Laurie] 7405 7406 *) Enhanced the err-ins.pl script so it makes the error library number 7407 global and can add a library name. This is needed for external ASN1 and 7408 other error libraries. 7409 [Steve Henson] 7410 7411 *) Fixed sk_insert which never worked properly. 7412 [Steve Henson] 7413 7414 *) Fix ASN1 macros so they can handle indefinite length construted 7415 EXPLICIT tags. Some non standard certificates use these: they can now 7416 be read in. 7417 [Steve Henson] 7418 7419 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc) 7420 into a single doc/ssleay.txt bundle. This way the information is still 7421 preserved but no longer messes up this directory. Now it's new room for 7422 the new set of documenation files. 7423 [Ralf S. Engelschall] 7424 7425 *) SETs were incorrectly DER encoded. This was a major pain, because they 7426 shared code with SEQUENCEs, which aren't coded the same. This means that 7427 almost everything to do with SETs or SEQUENCEs has either changed name or 7428 number of arguments. 7429 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>] 7430 7431 *) Fix test data to work with the above. 7432 [Ben Laurie] 7433 7434 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but 7435 was already fixed by Eric for 0.9.1 it seems. 7436 [Ben Laurie - pointed out by Ulf M�ller <ulf@fitug.de>] 7437 7438 *) Autodetect FreeBSD3. 7439 [Ben Laurie] 7440 7441 *) Fix various bugs in Configure. This affects the following platforms: 7442 nextstep 7443 ncr-scde 7444 unixware-2.0 7445 unixware-2.0-pentium 7446 sco5-cc. 7447 [Ben Laurie] 7448 7449 *) Eliminate generated files from CVS. Reorder tests to regenerate files 7450 before they are needed. 7451 [Ben Laurie] 7452 7453 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy). 7454 [Ben Laurie] 7455 7456 7457 Changes between 0.9.1b and 0.9.1c [23-Dec-1998] 7458 7459 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and 7460 changed SSLeay to OpenSSL in version strings. 7461 [Ralf S. Engelschall] 7462 7463 *) Some fixups to the top-level documents. 7464 [Paul Sutton] 7465 7466 *) Fixed the nasty bug where rsaref.h was not found under compile-time 7467 because the symlink to include/ was missing. 7468 [Ralf S. Engelschall] 7469 7470 *) Incorporated the popular no-RSA/DSA-only patches 7471 which allow to compile a RSA-free SSLeay. 7472 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall] 7473 7474 *) Fixed nasty rehash problem under `make -f Makefile.ssl links' 7475 when "ssleay" is still not found. 7476 [Ralf S. Engelschall] 7477 7478 *) Added more platforms to Configure: Cray T3E, HPUX 11, 7479 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>] 7480 7481 *) Updated the README file. 7482 [Ralf S. Engelschall] 7483 7484 *) Added various .cvsignore files in the CVS repository subdirs 7485 to make a "cvs update" really silent. 7486 [Ralf S. Engelschall] 7487 7488 *) Recompiled the error-definition header files and added 7489 missing symbols to the Win32 linker tables. 7490 [Ralf S. Engelschall] 7491 7492 *) Cleaned up the top-level documents; 7493 o new files: CHANGES and LICENSE 7494 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay 7495 o merged COPYRIGHT into LICENSE 7496 o removed obsolete TODO file 7497 o renamed MICROSOFT to INSTALL.W32 7498 [Ralf S. Engelschall] 7499 7500 *) Removed dummy files from the 0.9.1b source tree: 7501 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi 7502 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f 7503 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f 7504 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f 7505 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f 7506 [Ralf S. Engelschall] 7507 7508 *) Added various platform portability fixes. 7509 [Mark J. Cox] 7510 7511 *) The Genesis of the OpenSSL rpject: 7512 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A. 7513 Young and Tim J. Hudson created while they were working for C2Net until 7514 summer 1998. 7515 [The OpenSSL Project] 7516 7517 7518 Changes between 0.9.0b and 0.9.1b [not released] 7519 7520 *) Updated a few CA certificates under certs/ 7521 [Eric A. Young] 7522 7523 *) Changed some BIGNUM api stuff. 7524 [Eric A. Young] 7525 7526 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD, 7527 DGUX x86, Linux Alpha, etc. 7528 [Eric A. Young] 7529 7530 *) New COMP library [crypto/comp/] for SSL Record Layer Compression: 7531 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is 7532 available). 7533 [Eric A. Young] 7534 7535 *) Add -strparse option to asn1pars program which parses nested 7536 binary structures 7537 [Dr Stephen Henson <shenson@bigfoot.com>] 7538 7539 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs. 7540 [Eric A. Young] 7541 7542 *) DSA fix for "ca" program. 7543 [Eric A. Young] 7544 7545 *) Added "-genkey" option to "dsaparam" program. 7546 [Eric A. Young] 7547 7548 *) Added RIPE MD160 (rmd160) message digest. 7549 [Eric A. Young] 7550 7551 *) Added -a (all) option to "ssleay version" command. 7552 [Eric A. Young] 7553 7554 *) Added PLATFORM define which is the id given to Configure. 7555 [Eric A. Young] 7556 7557 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking. 7558 [Eric A. Young] 7559 7560 *) Extended the ASN.1 parser routines. 7561 [Eric A. Young] 7562 7563 *) Extended BIO routines to support REUSEADDR, seek, tell, etc. 7564 [Eric A. Young] 7565 7566 *) Added a BN_CTX to the BN library. 7567 [Eric A. Young] 7568 7569 *) Fixed the weak key values in DES library 7570 [Eric A. Young] 7571 7572 *) Changed API in EVP library for cipher aliases. 7573 [Eric A. Young] 7574 7575 *) Added support for RC2/64bit cipher. 7576 [Eric A. Young] 7577 7578 *) Converted the lhash library to the crypto/mem.c functions. 7579 [Eric A. Young] 7580 7581 *) Added more recognized ASN.1 object ids. 7582 [Eric A. Young] 7583 7584 *) Added more RSA padding checks for SSL/TLS. 7585 [Eric A. Young] 7586 7587 *) Added BIO proxy/filter functionality. 7588 [Eric A. Young] 7589 7590 *) Added extra_certs to SSL_CTX which can be used 7591 send extra CA certificates to the client in the CA cert chain sending 7592 process. It can be configured with SSL_CTX_add_extra_chain_cert(). 7593 [Eric A. Young] 7594 7595 *) Now Fortezza is denied in the authentication phase because 7596 this is key exchange mechanism is not supported by SSLeay at all. 7597 [Eric A. Young] 7598 7599 *) Additional PKCS1 checks. 7600 [Eric A. Young] 7601 7602 *) Support the string "TLSv1" for all TLS v1 ciphers. 7603 [Eric A. Young] 7604 7605 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the 7606 ex_data index of the SSL context in the X509_STORE_CTX ex_data. 7607 [Eric A. Young] 7608 7609 *) Fixed a few memory leaks. 7610 [Eric A. Young] 7611 7612 *) Fixed various code and comment typos. 7613 [Eric A. Young] 7614 7615 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0 7616 bytes sent in the client random. 7617 [Edward Bishop <ebishop@spyglass.com>] 7618 7619