CHANGES revision 89837
1 2 OpenSSL CHANGES 3 _______________ 4 5 Changes between 0.9.6b and 0.9.6c [21 dec 2001] 6 7 *) Fix BN_rand_range bug pointed out by Dominikus Scherkl 8 <Dominikus.Scherkl@biodata.com>. (The previous implementation 9 worked incorrectly for those cases where range = 10..._2 and 10 3*range is two bits longer than range.) 11 [Bodo Moeller] 12 13 *) Only add signing time to PKCS7 structures if it is not already 14 present. 15 [Steve Henson] 16 17 *) Fix crypto/objects/objects.h: "ld-ce" should be "id-ce", 18 OBJ_ld_ce should be OBJ_id_ce. 19 Also some ip-pda OIDs in crypto/objects/objects.txt were 20 incorrect (cf. RFC 3039). 21 [Matt Cooper, Frederic Giudicelli, Bodo Moeller] 22 23 *) Release CRYPTO_LOCK_DYNLOCK when CRYPTO_destroy_dynlockid() 24 returns early because it has nothing to do. 25 [Andy Schneider <andy.schneider@bjss.co.uk>] 26 27 *) [In 0.9.6c-engine release:] 28 Fix mutex callback return values in crypto/engine/hw_ncipher.c. 29 [Andy Schneider <andy.schneider@bjss.co.uk>] 30 31 *) [In 0.9.6c-engine release:] 32 Add support for Cryptographic Appliance's keyserver technology. 33 (Use engine 'keyclient') 34 [Cryptographic Appliances and Geoff Thorpe] 35 36 *) Add a configuration entry for OS/390 Unix. The C compiler 'c89' 37 is called via tools/c89.sh because arguments have to be 38 rearranged (all '-L' options must appear before the first object 39 modules). 40 [Richard Shapiro <rshapiro@abinitio.com>] 41 42 *) [In 0.9.6c-engine release:] 43 Add support for Broadcom crypto accelerator cards, backported 44 from 0.9.7. 45 [Broadcom, Nalin Dahyabhai <nalin@redhat.com>, Mark Cox] 46 47 *) [In 0.9.6c-engine release:] 48 Add support for SureWare crypto accelerator cards from 49 Baltimore Technologies. (Use engine 'sureware') 50 [Baltimore Technologies and Mark Cox] 51 52 *) [In 0.9.6c-engine release:] 53 Add support for crypto accelerator cards from Accelerated 54 Encryption Processing, www.aep.ie. (Use engine 'aep') 55 [AEP Inc. and Mark Cox] 56 57 *) Add a configuration entry for gcc on UnixWare. 58 [Gary Benson <gbenson@redhat.com>] 59 60 *) Change ssl/s2_clnt.c and ssl/s2_srvr.c so that received handshake 61 messages are stored in a single piece (fixed-length part and 62 variable-length part combined) and fix various bugs found on the way. 63 [Bodo Moeller] 64 65 *) Disable caching in BIO_gethostbyname(), directly use gethostbyname() 66 instead. BIO_gethostbyname() does not know what timeouts are 67 appropriate, so entries would stay in cache even when they have 68 become invalid. 69 [Bodo Moeller; problem pointed out by Rich Salz <rsalz@zolera.com> 70 71 *) Change ssl23_get_client_hello (ssl/s23_srvr.c) behaviour when 72 faced with a pathologically small ClientHello fragment that does 73 not contain client_version: Instead of aborting with an error, 74 simply choose the highest available protocol version (i.e., 75 TLS 1.0 unless it is disabled). In practice, ClientHello 76 messages are never sent like this, but this change gives us 77 strictly correct behaviour at least for TLS. 78 [Bodo Moeller] 79 80 *) Fix SSL handshake functions and SSL_clear() such that SSL_clear() 81 never resets s->method to s->ctx->method when called from within 82 one of the SSL handshake functions. 83 [Bodo Moeller; problem pointed out by Niko Baric] 84 85 *) In ssl3_get_client_hello (ssl/s3_srvr.c), generate a fatal alert 86 (sent using the client's version number) if client_version is 87 smaller than the protocol version in use. Also change 88 ssl23_get_client_hello (ssl/s23_srvr.c) to select TLS 1.0 if 89 the client demanded SSL 3.0 but only TLS 1.0 is enabled; then 90 the client will at least see that alert. 91 [Bodo Moeller] 92 93 *) Fix ssl3_get_message (ssl/s3_both.c) to handle message fragmentation 94 correctly. 95 [Bodo Moeller] 96 97 *) Avoid infinite loop in ssl3_get_message (ssl/s3_both.c) if a 98 client receives HelloRequest while in a handshake. 99 [Bodo Moeller; bug noticed by Andy Schneider <andy.schneider@bjss.co.uk>] 100 101 *) Bugfix in ssl3_accept (ssl/s3_srvr.c): Case SSL3_ST_SW_HELLO_REQ_C 102 should end in 'break', not 'goto end' which circuments various 103 cleanups done in state SSL_ST_OK. But session related stuff 104 must be disabled for SSL_ST_OK in the case that we just sent a 105 HelloRequest. 106 107 Also avoid some overhead by not calling ssl_init_wbio_buffer() 108 before just sending a HelloRequest. 109 [Bodo Moeller, Eric Rescorla <ekr@rtfm.com>] 110 111 *) Fix ssl/s3_enc.c, ssl/t1_enc.c and ssl/s3_pkt.c so that we don't 112 reveal whether illegal block cipher padding was found or a MAC 113 verification error occured. (Neither SSLerr() codes nor alerts 114 are directly visible to potential attackers, but the information 115 may leak via logfiles.) 116 117 Similar changes are not required for the SSL 2.0 implementation 118 because the number of padding bytes is sent in clear for SSL 2.0, 119 and the extra bytes are just ignored. However ssl/s2_pkt.c 120 failed to verify that the purported number of padding bytes is in 121 the legal range. 122 [Bodo Moeller] 123 124 *) Add OpenUNIX-8 support including shared libraries 125 (Boyd Lynn Gerber <gerberb@zenez.com>). 126 [Lutz Jaenicke] 127 128 *) Improve RSA_padding_check_PKCS1_OAEP() check again to avoid 129 'wristwatch attack' using huge encoding parameters (cf. 130 James H. Manger's CRYPTO 2001 paper). Note that the 131 RSA_PKCS1_OAEP_PADDING case of RSA_private_decrypt() does not use 132 encoding parameters and hence was not vulnerable. 133 [Bodo Moeller] 134 135 *) BN_sqr() bug fix. 136 [Ulf M�ller, reported by Jim Ellis <jim.ellis@cavium.com>] 137 138 *) Rabin-Miller test analyses assume uniformly distributed witnesses, 139 so use BN_pseudo_rand_range() instead of using BN_pseudo_rand() 140 followed by modular reduction. 141 [Bodo Moeller; pointed out by Adam Young <AYoung1@NCSUS.JNJ.COM>] 142 143 *) Add BN_pseudo_rand_range() with obvious functionality: BN_rand_range() 144 equivalent based on BN_pseudo_rand() instead of BN_rand(). 145 [Bodo Moeller] 146 147 *) s3_srvr.c: allow sending of large client certificate lists (> 16 kB). 148 This function was broken, as the check for a new client hello message 149 to handle SGC did not allow these large messages. 150 (Tracked down by "Douglas E. Engert" <deengert@anl.gov>.) 151 [Lutz Jaenicke] 152 153 *) Add alert descriptions for TLSv1 to SSL_alert_desc_string[_long](). 154 [Lutz Jaenicke] 155 156 *) Fix buggy behaviour of BIO_get_num_renegotiates() and BIO_ctrl() 157 for BIO_C_GET_WRITE_BUF_SIZE ("Stephen Hinton" <shinton@netopia.com>). 158 [Lutz Jaenicke] 159 160 *) Rework the configuration and shared library support for Tru64 Unix. 161 The configuration part makes use of modern compiler features and 162 still retains old compiler behavior for those that run older versions 163 of the OS. The shared library support part includes a variant that 164 uses the RPATH feature, and is available through the special 165 configuration target "alpha-cc-rpath", which will never be selected 166 automatically. 167 [Tim Mooney <mooney@dogbert.cc.ndsu.NoDak.edu> via Richard Levitte] 168 169 *) In ssl3_get_key_exchange (ssl/s3_clnt.c), call ssl3_get_message() 170 with the same message size as in ssl3_get_certificate_request(). 171 Otherwise, if no ServerKeyExchange message occurs, CertificateRequest 172 messages might inadvertently be reject as too long. 173 [Petr Lampa <lampa@fee.vutbr.cz>] 174 175 *) Enhanced support for IA-64 Unix platforms (well, Linux and HP-UX). 176 [Andy Polyakov] 177 178 *) Modified SSL library such that the verify_callback that has been set 179 specificly for an SSL object with SSL_set_verify() is actually being 180 used. Before the change, a verify_callback set with this function was 181 ignored and the verify_callback() set in the SSL_CTX at the time of 182 the call was used. New function X509_STORE_CTX_set_verify_cb() introduced 183 to allow the necessary settings. 184 [Lutz Jaenicke] 185 186 *) Initialize static variable in crypto/dsa/dsa_lib.c and crypto/dh/dh_lib.c 187 explicitly to NULL, as at least on Solaris 8 this seems not always to be 188 done automatically (in contradiction to the requirements of the C 189 standard). This made problems when used from OpenSSH. 190 [Lutz Jaenicke] 191 192 *) In OpenSSL 0.9.6a and 0.9.6b, crypto/dh/dh_key.c ignored 193 dh->length and always used 194 195 BN_rand_range(priv_key, dh->p). 196 197 BN_rand_range() is not necessary for Diffie-Hellman, and this 198 specific range makes Diffie-Hellman unnecessarily inefficient if 199 dh->length (recommended exponent length) is much smaller than the 200 length of dh->p. We could use BN_rand_range() if the order of 201 the subgroup was stored in the DH structure, but we only have 202 dh->length. 203 204 So switch back to 205 206 BN_rand(priv_key, l, ...) 207 208 where 'l' is dh->length if this is defined, or BN_num_bits(dh->p)-1 209 otherwise. 210 [Bodo Moeller] 211 212 *) In 213 214 RSA_eay_public_encrypt 215 RSA_eay_private_decrypt 216 RSA_eay_private_encrypt (signing) 217 RSA_eay_public_decrypt (signature verification) 218 219 (default implementations for RSA_public_encrypt, 220 RSA_private_decrypt, RSA_private_encrypt, RSA_public_decrypt), 221 always reject numbers >= n. 222 [Bodo Moeller] 223 224 *) In crypto/rand/md_rand.c, use a new short-time lock CRYPTO_LOCK_RAND2 225 to synchronize access to 'locking_thread'. This is necessary on 226 systems where access to 'locking_thread' (an 'unsigned long' 227 variable) is not atomic. 228 [Bodo Moeller] 229 230 *) In crypto/rand/md_rand.c, set 'locking_thread' to current thread's ID 231 *before* setting the 'crypto_lock_rand' flag. The previous code had 232 a race condition if 0 is a valid thread ID. 233 [Travis Vitek <vitek@roguewave.com>] 234 235 *) Add support for shared libraries under Irix. 236 [Albert Chin-A-Young <china@thewrittenword.com>] 237 238 *) Add configuration option to build on Linux on both big-endian and 239 little-endian MIPS. 240 [Ralf Baechle <ralf@uni-koblenz.de>] 241 242 *) Add the possibility to create shared libraries on HP-UX. 243 [Richard Levitte] 244 245 Changes between 0.9.6a and 0.9.6b [9 Jul 2001] 246 247 *) Change ssleay_rand_bytes (crypto/rand/md_rand.c) 248 to avoid a SSLeay/OpenSSL PRNG weakness pointed out by 249 Markku-Juhani O. Saarinen <markku-juhani.saarinen@nokia.com>: 250 PRNG state recovery was possible based on the output of 251 one PRNG request appropriately sized to gain knowledge on 252 'md' followed by enough consecutive 1-byte PRNG requests 253 to traverse all of 'state'. 254 255 1. When updating 'md_local' (the current thread's copy of 'md') 256 during PRNG output generation, hash all of the previous 257 'md_local' value, not just the half used for PRNG output. 258 259 2. Make the number of bytes from 'state' included into the hash 260 independent from the number of PRNG bytes requested. 261 262 The first measure alone would be sufficient to avoid 263 Markku-Juhani's attack. (Actually it had never occurred 264 to me that the half of 'md_local' used for chaining was the 265 half from which PRNG output bytes were taken -- I had always 266 assumed that the secret half would be used.) The second 267 measure makes sure that additional data from 'state' is never 268 mixed into 'md_local' in small portions; this heuristically 269 further strengthens the PRNG. 270 [Bodo Moeller] 271 272 *) Fix crypto/bn/asm/mips3.s. 273 [Andy Polyakov] 274 275 *) When only the key is given to "enc", the IV is undefined. Print out 276 an error message in this case. 277 [Lutz Jaenicke] 278 279 *) Handle special case when X509_NAME is empty in X509 printing routines. 280 [Steve Henson] 281 282 *) In dsa_do_verify (crypto/dsa/dsa_ossl.c), verify that r and s are 283 positive and less than q. 284 [Bodo Moeller] 285 286 *) Don't change *pointer in CRYPTO_add_lock() is add_lock_callback is 287 used: it isn't thread safe and the add_lock_callback should handle 288 that itself. 289 [Paul Rose <Paul.Rose@bridge.com>] 290 291 *) Verify that incoming data obeys the block size in 292 ssl3_enc (ssl/s3_enc.c) and tls1_enc (ssl/t1_enc.c). 293 [Bodo Moeller] 294 295 *) Fix OAEP check. 296 [Ulf M�ller, Bodo M�ller] 297 298 *) The countermeasure against Bleichbacher's attack on PKCS #1 v1.5 299 RSA encryption was accidentally removed in s3_srvr.c in OpenSSL 0.9.5 300 when fixing the server behaviour for backwards-compatible 'client 301 hello' messages. (Note that the attack is impractical against 302 SSL 3.0 and TLS 1.0 anyway because length and version checking 303 means that the probability of guessing a valid ciphertext is 304 around 2^-40; see section 5 in Bleichenbacher's CRYPTO '98 305 paper.) 306 307 Before 0.9.5, the countermeasure (hide the error by generating a 308 random 'decryption result') did not work properly because 309 ERR_clear_error() was missing, meaning that SSL_get_error() would 310 detect the supposedly ignored error. 311 312 Both problems are now fixed. 313 [Bodo Moeller] 314 315 *) In crypto/bio/bf_buff.c, increase DEFAULT_BUFFER_SIZE to 4096 316 (previously it was 1024). 317 [Bodo Moeller] 318 319 *) Fix for compatibility mode trust settings: ignore trust settings 320 unless some valid trust or reject settings are present. 321 [Steve Henson] 322 323 *) Fix for blowfish EVP: its a variable length cipher. 324 [Steve Henson] 325 326 *) Fix various bugs related to DSA S/MIME verification. Handle missing 327 parameters in DSA public key structures and return an error in the 328 DSA routines if parameters are absent. 329 [Steve Henson] 330 331 *) In versions up to 0.9.6, RAND_file_name() resorted to file ".rnd" 332 in the current directory if neither $RANDFILE nor $HOME was set. 333 RAND_file_name() in 0.9.6a returned NULL in this case. This has 334 caused some confusion to Windows users who haven't defined $HOME. 335 Thus RAND_file_name() is changed again: e_os.h can define a 336 DEFAULT_HOME, which will be used if $HOME is not set. 337 For Windows, we use "C:"; on other platforms, we still require 338 environment variables. 339 340 *) Move 'if (!initialized) RAND_poll()' into regions protected by 341 CRYPTO_LOCK_RAND. This is not strictly necessary, but avoids 342 having multiple threads call RAND_poll() concurrently. 343 [Bodo Moeller] 344 345 *) In crypto/rand/md_rand.c, replace 'add_do_not_lock' flag by a 346 combination of a flag and a thread ID variable. 347 Otherwise while one thread is in ssleay_rand_bytes (which sets the 348 flag), *other* threads can enter ssleay_add_bytes without obeying 349 the CRYPTO_LOCK_RAND lock (and may even illegally release the lock 350 that they do not hold after the first thread unsets add_do_not_lock). 351 [Bodo Moeller] 352 353 *) Change bctest again: '-x' expressions are not available in all 354 versions of 'test'. 355 [Bodo Moeller] 356 357 Changes between 0.9.6 and 0.9.6a [5 Apr 2001] 358 359 *) Fix a couple of memory leaks in PKCS7_dataDecode() 360 [Steve Henson, reported by Heyun Zheng <hzheng@atdsprint.com>] 361 362 *) Change Configure and Makefiles to provide EXE_EXT, which will contain 363 the default extension for executables, if any. Also, make the perl 364 scripts that use symlink() to test if it really exists and use "cp" 365 if it doesn't. All this made OpenSSL compilable and installable in 366 CygWin. 367 [Richard Levitte] 368 369 *) Fix for asn1_GetSequence() for indefinite length constructed data. 370 If SEQUENCE is length is indefinite just set c->slen to the total 371 amount of data available. 372 [Steve Henson, reported by shige@FreeBSD.org] 373 [This change does not apply to 0.9.7.] 374 375 *) Change bctest to avoid here-documents inside command substitution 376 (workaround for FreeBSD /bin/sh bug). 377 For compatibility with Ultrix, avoid shell functions (introduced 378 in the bctest version that searches along $PATH). 379 [Bodo Moeller] 380 381 *) Rename 'des_encrypt' to 'des_encrypt1'. This avoids the clashes 382 with des_encrypt() defined on some operating systems, like Solaris 383 and UnixWare. 384 [Richard Levitte] 385 386 *) Check the result of RSA-CRT (see D. Boneh, R. DeMillo, R. Lipton: 387 On the Importance of Eliminating Errors in Cryptographic 388 Computations, J. Cryptology 14 (2001) 2, 101-119, 389 http://theory.stanford.edu/~dabo/papers/faults.ps.gz). 390 [Ulf Moeller] 391 392 *) MIPS assembler BIGNUM division bug fix. 393 [Andy Polyakov] 394 395 *) Disabled incorrect Alpha assembler code. 396 [Richard Levitte] 397 398 *) Fix PKCS#7 decode routines so they correctly update the length 399 after reading an EOC for the EXPLICIT tag. 400 [Steve Henson] 401 [This change does not apply to 0.9.7.] 402 403 *) Fix bug in PKCS#12 key generation routines. This was triggered 404 if a 3DES key was generated with a 0 initial byte. Include 405 PKCS12_BROKEN_KEYGEN compilation option to retain the old 406 (but broken) behaviour. 407 [Steve Henson] 408 409 *) Enhance bctest to search for a working bc along $PATH and print 410 it when found. 411 [Tim Rice <tim@multitalents.net> via Richard Levitte] 412 413 *) Fix memory leaks in err.c: free err_data string if necessary; 414 don't write to the wrong index in ERR_set_error_data. 415 [Bodo Moeller] 416 417 *) Implement ssl23_peek (analogous to ssl23_read), which previously 418 did not exist. 419 [Bodo Moeller] 420 421 *) Replace rdtsc with _emit statements for VC++ version 5. 422 [Jeremy Cooper <jeremy@baymoo.org>] 423 424 *) Make it possible to reuse SSLv2 sessions. 425 [Richard Levitte] 426 427 *) In copy_email() check for >= 0 as a return value for 428 X509_NAME_get_index_by_NID() since 0 is a valid index. 429 [Steve Henson reported by Massimiliano Pala <madwolf@opensca.org>] 430 431 *) Avoid coredump with unsupported or invalid public keys by checking if 432 X509_get_pubkey() fails in PKCS7_verify(). Fix memory leak when 433 PKCS7_verify() fails with non detached data. 434 [Steve Henson] 435 436 *) Don't use getenv in library functions when run as setuid/setgid. 437 New function OPENSSL_issetugid(). 438 [Ulf Moeller] 439 440 *) Avoid false positives in memory leak detection code (crypto/mem_dbg.c) 441 due to incorrect handling of multi-threading: 442 443 1. Fix timing glitch in the MemCheck_off() portion of CRYPTO_mem_ctrl(). 444 445 2. Fix logical glitch in is_MemCheck_on() aka CRYPTO_is_mem_check_on(). 446 447 3. Count how many times MemCheck_off() has been called so that 448 nested use can be treated correctly. This also avoids 449 inband-signalling in the previous code (which relied on the 450 assumption that thread ID 0 is impossible). 451 [Bodo Moeller] 452 453 *) Add "-rand" option also to s_client and s_server. 454 [Lutz Jaenicke] 455 456 *) Fix CPU detection on Irix 6.x. 457 [Kurt Hockenbury <khockenb@stevens-tech.edu> and 458 "Bruce W. Forsberg" <bruce.forsberg@baesystems.com>] 459 460 *) Fix X509_NAME bug which produced incorrect encoding if X509_NAME 461 was empty. 462 [Steve Henson] 463 [This change does not apply to 0.9.7.] 464 465 *) Use the cached encoding of an X509_NAME structure rather than 466 copying it. This is apparently the reason for the libsafe "errors" 467 but the code is actually correct. 468 [Steve Henson] 469 470 *) Add new function BN_rand_range(), and fix DSA_sign_setup() to prevent 471 Bleichenbacher's DSA attack. 472 Extend BN_[pseudo_]rand: As before, top=1 forces the highest two bits 473 to be set and top=0 forces the highest bit to be set; top=-1 is new 474 and leaves the highest bit random. 475 [Ulf Moeller, Bodo Moeller] 476 477 *) In the NCONF_...-based implementations for CONF_... queries 478 (crypto/conf/conf_lib.c), if the input LHASH is NULL, avoid using 479 a temporary CONF structure with the data component set to NULL 480 (which gives segmentation faults in lh_retrieve). 481 Instead, use NULL for the CONF pointer in CONF_get_string and 482 CONF_get_number (which may use environment variables) and directly 483 return NULL from CONF_get_section. 484 [Bodo Moeller] 485 486 *) Fix potential buffer overrun for EBCDIC. 487 [Ulf Moeller] 488 489 *) Tolerate nonRepudiation as being valid for S/MIME signing and certSign 490 keyUsage if basicConstraints absent for a CA. 491 [Steve Henson] 492 493 *) Make SMIME_write_PKCS7() write mail header values with a format that 494 is more generally accepted (no spaces before the semicolon), since 495 some programs can't parse those values properly otherwise. Also make 496 sure BIO's that break lines after each write do not create invalid 497 headers. 498 [Richard Levitte] 499 500 *) Make the CRL encoding routines work with empty SEQUENCE OF. The 501 macros previously used would not encode an empty SEQUENCE OF 502 and break the signature. 503 [Steve Henson] 504 [This change does not apply to 0.9.7.] 505 506 *) Zero the premaster secret after deriving the master secret in 507 DH ciphersuites. 508 [Steve Henson] 509 510 *) Add some EVP_add_digest_alias registrations (as found in 511 OpenSSL_add_all_digests()) to SSL_library_init() 512 aka OpenSSL_add_ssl_algorithms(). This provides improved 513 compatibility with peers using X.509 certificates 514 with unconventional AlgorithmIdentifier OIDs. 515 [Bodo Moeller] 516 517 *) Fix for Irix with NO_ASM. 518 ["Bruce W. Forsberg" <bruce.forsberg@baesystems.com>] 519 520 *) ./config script fixes. 521 [Ulf Moeller, Richard Levitte] 522 523 *) Fix 'openssl passwd -1'. 524 [Bodo Moeller] 525 526 *) Change PKCS12_key_gen_asc() so it can cope with non null 527 terminated strings whose length is passed in the passlen 528 parameter, for example from PEM callbacks. This was done 529 by adding an extra length parameter to asc2uni(). 530 [Steve Henson, reported by <oddissey@samsung.co.kr>] 531 532 *) Fix C code generated by 'openssl dsaparam -C': If a BN_bin2bn 533 call failed, free the DSA structure. 534 [Bodo Moeller] 535 536 *) Fix to uni2asc() to cope with zero length Unicode strings. 537 These are present in some PKCS#12 files. 538 [Steve Henson] 539 540 *) Increase s2->wbuf allocation by one byte in ssl2_new (ssl/s2_lib.c). 541 Otherwise do_ssl_write (ssl/s2_pkt.c) will write beyond buffer limits 542 when writing a 32767 byte record. 543 [Bodo Moeller; problem reported by Eric Day <eday@concentric.net>] 544 545 *) In RSA_eay_public_{en,ed}crypt and RSA_eay_mod_exp (rsa_eay.c), 546 obtain lock CRYPTO_LOCK_RSA before setting rsa->_method_mod_{n,p,q}. 547 548 (RSA objects have a reference count access to which is protected 549 by CRYPTO_LOCK_RSA [see rsa_lib.c, s3_srvr.c, ssl_cert.c, ssl_rsa.c], 550 so they are meant to be shared between threads.) 551 [Bodo Moeller, Geoff Thorpe; original patch submitted by 552 "Reddie, Steven" <Steven.Reddie@ca.com>] 553 554 *) Fix a deadlock in CRYPTO_mem_leaks(). 555 [Bodo Moeller] 556 557 *) Use better test patterns in bntest. 558 [Ulf M�ller] 559 560 *) rand_win.c fix for Borland C. 561 [Ulf M�ller] 562 563 *) BN_rshift bugfix for n == 0. 564 [Bodo Moeller] 565 566 *) Add a 'bctest' script that checks for some known 'bc' bugs 567 so that 'make test' does not abort just because 'bc' is broken. 568 [Bodo Moeller] 569 570 *) Store verify_result within SSL_SESSION also for client side to 571 avoid potential security hole. (Re-used sessions on the client side 572 always resulted in verify_result==X509_V_OK, not using the original 573 result of the server certificate verification.) 574 [Lutz Jaenicke] 575 576 *) Fix ssl3_pending: If the record in s->s3->rrec is not of type 577 SSL3_RT_APPLICATION_DATA, return 0. 578 Similarly, change ssl2_pending to return 0 if SSL_in_init(s) is true. 579 [Bodo Moeller] 580 581 *) Fix SSL_peek: 582 Both ssl2_peek and ssl3_peek, which were totally broken in earlier 583 releases, have been re-implemented by renaming the previous 584 implementations of ssl2_read and ssl3_read to ssl2_read_internal 585 and ssl3_read_internal, respectively, and adding 'peek' parameters 586 to them. The new ssl[23]_{read,peek} functions are calls to 587 ssl[23]_read_internal with the 'peek' flag set appropriately. 588 A 'peek' parameter has also been added to ssl3_read_bytes, which 589 does the actual work for ssl3_read_internal. 590 [Bodo Moeller] 591 592 *) Initialise "ex_data" member of RSA/DSA/DH structures prior to calling 593 the method-specific "init()" handler. Also clean up ex_data after 594 calling the method-specific "finish()" handler. Previously, this was 595 happening the other way round. 596 [Geoff Thorpe] 597 598 *) Increase BN_CTX_NUM (the number of BIGNUMs in a BN_CTX) to 16. 599 The previous value, 12, was not always sufficient for BN_mod_exp(). 600 [Bodo Moeller] 601 602 *) Make sure that shared libraries get the internal name engine with 603 the full version number and not just 0. This should mark the 604 shared libraries as not backward compatible. Of course, this should 605 be changed again when we can guarantee backward binary compatibility. 606 [Richard Levitte] 607 608 *) Fix typo in get_cert_by_subject() in by_dir.c 609 [Jean-Marc Desperrier <jean-marc.desperrier@certplus.com>] 610 611 *) Rework the system to generate shared libraries: 612 613 - Make note of the expected extension for the shared libraries and 614 if there is a need for symbolic links from for example libcrypto.so.0 615 to libcrypto.so.0.9.7. There is extended info in Configure for 616 that. 617 618 - Make as few rebuilds of the shared libraries as possible. 619 620 - Still avoid linking the OpenSSL programs with the shared libraries. 621 622 - When installing, install the shared libraries separately from the 623 static ones. 624 [Richard Levitte] 625 626 *) Fix SSL_CTX_set_read_ahead macro to actually use its argument. 627 628 Copy SSL_CTX's read_ahead flag to SSL object directly in SSL_new 629 and not in SSL_clear because the latter is also used by the 630 accept/connect functions; previously, the settings made by 631 SSL_set_read_ahead would be lost during the handshake. 632 [Bodo Moeller; problems reported by Anders Gertz <gertz@epact.se>] 633 634 *) Correct util/mkdef.pl to be selective about disabled algorithms. 635 Previously, it would create entries for disableed algorithms no 636 matter what. 637 [Richard Levitte] 638 639 *) Added several new manual pages for SSL_* function. 640 [Lutz Jaenicke] 641 642 Changes between 0.9.5a and 0.9.6 [24 Sep 2000] 643 644 *) In ssl23_get_client_hello, generate an error message when faced 645 with an initial SSL 3.0/TLS record that is too small to contain the 646 first two bytes of the ClientHello message, i.e. client_version. 647 (Note that this is a pathologic case that probably has never happened 648 in real life.) The previous approach was to use the version number 649 from the record header as a substitute; but our protocol choice 650 should not depend on that one because it is not authenticated 651 by the Finished messages. 652 [Bodo Moeller] 653 654 *) More robust randomness gathering functions for Windows. 655 [Jeffrey Altman <jaltman@columbia.edu>] 656 657 *) For compatibility reasons if the flag X509_V_FLAG_ISSUER_CHECK is 658 not set then we don't setup the error code for issuer check errors 659 to avoid possibly overwriting other errors which the callback does 660 handle. If an application does set the flag then we assume it knows 661 what it is doing and can handle the new informational codes 662 appropriately. 663 [Steve Henson] 664 665 *) Fix for a nasty bug in ASN1_TYPE handling. ASN1_TYPE is used for 666 a general "ANY" type, as such it should be able to decode anything 667 including tagged types. However it didn't check the class so it would 668 wrongly interpret tagged types in the same way as their universal 669 counterpart and unknown types were just rejected. Changed so that the 670 tagged and unknown types are handled in the same way as a SEQUENCE: 671 that is the encoding is stored intact. There is also a new type 672 "V_ASN1_OTHER" which is used when the class is not universal, in this 673 case we have no idea what the actual type is so we just lump them all 674 together. 675 [Steve Henson] 676 677 *) On VMS, stdout may very well lead to a file that is written to 678 in a record-oriented fashion. That means that every write() will 679 write a separate record, which will be read separately by the 680 programs trying to read from it. This can be very confusing. 681 682 The solution is to put a BIO filter in the way that will buffer 683 text until a linefeed is reached, and then write everything a 684 line at a time, so every record written will be an actual line, 685 not chunks of lines and not (usually doesn't happen, but I've 686 seen it once) several lines in one record. BIO_f_linebuffer() is 687 the answer. 688 689 Currently, it's a VMS-only method, because that's where it has 690 been tested well enough. 691 [Richard Levitte] 692 693 *) Remove 'optimized' squaring variant in BN_mod_mul_montgomery, 694 it can return incorrect results. 695 (Note: The buggy variant was not enabled in OpenSSL 0.9.5a, 696 but it was in 0.9.6-beta[12].) 697 [Bodo Moeller] 698 699 *) Disable the check for content being present when verifying detached 700 signatures in pk7_smime.c. Some versions of Netscape (wrongly) 701 include zero length content when signing messages. 702 [Steve Henson] 703 704 *) New BIO_shutdown_wr macro, which invokes the BIO_C_SHUTDOWN_WR 705 BIO_ctrl (for BIO pairs). 706 [Bodo M�ller] 707 708 *) Add DSO method for VMS. 709 [Richard Levitte] 710 711 *) Bug fix: Montgomery multiplication could produce results with the 712 wrong sign. 713 [Ulf M�ller] 714 715 *) Add RPM specification openssl.spec and modify it to build three 716 packages. The default package contains applications, application 717 documentation and run-time libraries. The devel package contains 718 include files, static libraries and function documentation. The 719 doc package contains the contents of the doc directory. The original 720 openssl.spec was provided by Damien Miller <djm@mindrot.org>. 721 [Richard Levitte] 722 723 *) Add a large number of documentation files for many SSL routines. 724 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>] 725 726 *) Add a configuration entry for Sony News 4. 727 [NAKAJI Hiroyuki <nakaji@tutrp.tut.ac.jp>] 728 729 *) Don't set the two most significant bits to one when generating a 730 random number < q in the DSA library. 731 [Ulf M�ller] 732 733 *) New SSL API mode 'SSL_MODE_AUTO_RETRY'. This disables the default 734 behaviour that SSL_read may result in SSL_ERROR_WANT_READ (even if 735 the underlying transport is blocking) if a handshake took place. 736 (The default behaviour is needed by applications such as s_client 737 and s_server that use select() to determine when to use SSL_read; 738 but for applications that know in advance when to expect data, it 739 just makes things more complicated.) 740 [Bodo Moeller] 741 742 *) Add RAND_egd_bytes(), which gives control over the number of bytes read 743 from EGD. 744 [Ben Laurie] 745 746 *) Add a few more EBCDIC conditionals that make `req' and `x509' 747 work better on such systems. 748 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>] 749 750 *) Add two demo programs for PKCS12_parse() and PKCS12_create(). 751 Update PKCS12_parse() so it copies the friendlyName and the 752 keyid to the certificates aux info. 753 [Steve Henson] 754 755 *) Fix bug in PKCS7_verify() which caused an infinite loop 756 if there was more than one signature. 757 [Sven Uszpelkat <su@celocom.de>] 758 759 *) Major change in util/mkdef.pl to include extra information 760 about each symbol, as well as presentig variables as well 761 as functions. This change means that there's n more need 762 to rebuild the .num files when some algorithms are excluded. 763 [Richard Levitte] 764 765 *) Allow the verify time to be set by an application, 766 rather than always using the current time. 767 [Steve Henson] 768 769 *) Phase 2 verify code reorganisation. The certificate 770 verify code now looks up an issuer certificate by a 771 number of criteria: subject name, authority key id 772 and key usage. It also verifies self signed certificates 773 by the same criteria. The main comparison function is 774 X509_check_issued() which performs these checks. 775 776 Lot of changes were necessary in order to support this 777 without completely rewriting the lookup code. 778 779 Authority and subject key identifier are now cached. 780 781 The LHASH 'certs' is X509_STORE has now been replaced 782 by a STACK_OF(X509_OBJECT). This is mainly because an 783 LHASH can't store or retrieve multiple objects with 784 the same hash value. 785 786 As a result various functions (which were all internal 787 use only) have changed to handle the new X509_STORE 788 structure. This will break anything that messed round 789 with X509_STORE internally. 790 791 The functions X509_STORE_add_cert() now checks for an 792 exact match, rather than just subject name. 793 794 The X509_STORE API doesn't directly support the retrieval 795 of multiple certificates matching a given criteria, however 796 this can be worked round by performing a lookup first 797 (which will fill the cache with candidate certificates) 798 and then examining the cache for matches. This is probably 799 the best we can do without throwing out X509_LOOKUP 800 entirely (maybe later...). 801 802 The X509_VERIFY_CTX structure has been enhanced considerably. 803 804 All certificate lookup operations now go via a get_issuer() 805 callback. Although this currently uses an X509_STORE it 806 can be replaced by custom lookups. This is a simple way 807 to bypass the X509_STORE hackery necessary to make this 808 work and makes it possible to use more efficient techniques 809 in future. A very simple version which uses a simple 810 STACK for its trusted certificate store is also provided 811 using X509_STORE_CTX_trusted_stack(). 812 813 The verify_cb() and verify() callbacks now have equivalents 814 in the X509_STORE_CTX structure. 815 816 X509_STORE_CTX also has a 'flags' field which can be used 817 to customise the verify behaviour. 818 [Steve Henson] 819 820 *) Add new PKCS#7 signing option PKCS7_NOSMIMECAP which 821 excludes S/MIME capabilities. 822 [Steve Henson] 823 824 *) When a certificate request is read in keep a copy of the 825 original encoding of the signed data and use it when outputing 826 again. Signatures then use the original encoding rather than 827 a decoded, encoded version which may cause problems if the 828 request is improperly encoded. 829 [Steve Henson] 830 831 *) For consistency with other BIO_puts implementations, call 832 buffer_write(b, ...) directly in buffer_puts instead of calling 833 BIO_write(b, ...). 834 835 In BIO_puts, increment b->num_write as in BIO_write. 836 [Peter.Sylvester@EdelWeb.fr] 837 838 *) Fix BN_mul_word for the case where the word is 0. (We have to use 839 BN_zero, we may not return a BIGNUM with an array consisting of 840 words set to zero.) 841 [Bodo Moeller] 842 843 *) Avoid calling abort() from within the library when problems are 844 detected, except if preprocessor symbols have been defined 845 (such as REF_CHECK, BN_DEBUG etc.). 846 [Bodo Moeller] 847 848 *) New openssl application 'rsautl'. This utility can be 849 used for low level RSA operations. DER public key 850 BIO/fp routines also added. 851 [Steve Henson] 852 853 *) New Configure entry and patches for compiling on QNX 4. 854 [Andreas Schneider <andreas@ds3.etech.fh-hamburg.de>] 855 856 *) A demo state-machine implementation was sponsored by 857 Nuron (http://www.nuron.com/) and is now available in 858 demos/state_machine. 859 [Ben Laurie] 860 861 *) New options added to the 'dgst' utility for signature 862 generation and verification. 863 [Steve Henson] 864 865 *) Unrecognized PKCS#7 content types are now handled via a 866 catch all ASN1_TYPE structure. This allows unsupported 867 types to be stored as a "blob" and an application can 868 encode and decode it manually. 869 [Steve Henson] 870 871 *) Fix various signed/unsigned issues to make a_strex.c 872 compile under VC++. 873 [Oscar Jacobsson <oscar.jacobsson@celocom.com>] 874 875 *) ASN1 fixes. i2d_ASN1_OBJECT was not returning the correct 876 length if passed a buffer. ASN1_INTEGER_to_BN failed 877 if passed a NULL BN and its argument was negative. 878 [Steve Henson, pointed out by Sven Heiberg <sven@tartu.cyber.ee>] 879 880 *) Modification to PKCS#7 encoding routines to output definite 881 length encoding. Since currently the whole structures are in 882 memory there's not real point in using indefinite length 883 constructed encoding. However if OpenSSL is compiled with 884 the flag PKCS7_INDEFINITE_ENCODING the old form is used. 885 [Steve Henson] 886 887 *) Added BIO_vprintf() and BIO_vsnprintf(). 888 [Richard Levitte] 889 890 *) Added more prefixes to parse for in the the strings written 891 through a logging bio, to cover all the levels that are available 892 through syslog. The prefixes are now: 893 894 PANIC, EMERG, EMR => LOG_EMERG 895 ALERT, ALR => LOG_ALERT 896 CRIT, CRI => LOG_CRIT 897 ERROR, ERR => LOG_ERR 898 WARNING, WARN, WAR => LOG_WARNING 899 NOTICE, NOTE, NOT => LOG_NOTICE 900 INFO, INF => LOG_INFO 901 DEBUG, DBG => LOG_DEBUG 902 903 and as before, if none of those prefixes are present at the 904 beginning of the string, LOG_ERR is chosen. 905 906 On Win32, the LOG_* levels are mapped according to this: 907 908 LOG_EMERG, LOG_ALERT, LOG_CRIT, LOG_ERR => EVENTLOG_ERROR_TYPE 909 LOG_WARNING => EVENTLOG_WARNING_TYPE 910 LOG_NOTICE, LOG_INFO, LOG_DEBUG => EVENTLOG_INFORMATION_TYPE 911 912 [Richard Levitte] 913 914 *) Made it possible to reconfigure with just the configuration 915 argument "reconf" or "reconfigure". The command line arguments 916 are stored in Makefile.ssl in the variable CONFIGURE_ARGS, 917 and are retrieved from there when reconfiguring. 918 [Richard Levitte] 919 920 *) MD4 implemented. 921 [Assar Westerlund <assar@sics.se>, Richard Levitte] 922 923 *) Add the arguments -CAfile and -CApath to the pkcs12 utility. 924 [Richard Levitte] 925 926 *) The obj_dat.pl script was messing up the sorting of object 927 names. The reason was that it compared the quoted version 928 of strings as a result "OCSP" > "OCSP Signing" because 929 " > SPACE. Changed script to store unquoted versions of 930 names and add quotes on output. It was also omitting some 931 names from the lookup table if they were given a default 932 value (that is if SN is missing it is given the same 933 value as LN and vice versa), these are now added on the 934 grounds that if an object has a name we should be able to 935 look it up. Finally added warning output when duplicate 936 short or long names are found. 937 [Steve Henson] 938 939 *) Changes needed for Tandem NSK. 940 [Scott Uroff <scott@xypro.com>] 941 942 *) Fix SSL 2.0 rollback checking: Due to an off-by-one error in 943 RSA_padding_check_SSLv23(), special padding was never detected 944 and thus the SSL 3.0/TLS 1.0 countermeasure against protocol 945 version rollback attacks was not effective. 946 947 In s23_clnt.c, don't use special rollback-attack detection padding 948 (RSA_SSLV23_PADDING) if SSL 2.0 is the only protocol enabled in the 949 client; similarly, in s23_srvr.c, don't do the rollback check if 950 SSL 2.0 is the only protocol enabled in the server. 951 [Bodo Moeller] 952 953 *) Make it possible to get hexdumps of unprintable data with 'openssl 954 asn1parse'. By implication, the functions ASN1_parse_dump() and 955 BIO_dump_indent() are added. 956 [Richard Levitte] 957 958 *) New functions ASN1_STRING_print_ex() and X509_NAME_print_ex() 959 these print out strings and name structures based on various 960 flags including RFC2253 support and proper handling of 961 multibyte characters. Added options to the 'x509' utility 962 to allow the various flags to be set. 963 [Steve Henson] 964 965 *) Various fixes to use ASN1_TIME instead of ASN1_UTCTIME. 966 Also change the functions X509_cmp_current_time() and 967 X509_gmtime_adj() work with an ASN1_TIME structure, 968 this will enable certificates using GeneralizedTime in validity 969 dates to be checked. 970 [Steve Henson] 971 972 *) Make the NEG_PUBKEY_BUG code (which tolerates invalid 973 negative public key encodings) on by default, 974 NO_NEG_PUBKEY_BUG can be set to disable it. 975 [Steve Henson] 976 977 *) New function c2i_ASN1_OBJECT() which acts on ASN1_OBJECT 978 content octets. An i2c_ASN1_OBJECT is unnecessary because 979 the encoding can be trivially obtained from the structure. 980 [Steve Henson] 981 982 *) crypto/err.c locking bugfix: Use write locks (CRYPTO_w_[un]lock), 983 not read locks (CRYPTO_r_[un]lock). 984 [Bodo Moeller] 985 986 *) A first attempt at creating official support for shared 987 libraries through configuration. I've kept it so the 988 default is static libraries only, and the OpenSSL programs 989 are always statically linked for now, but there are 990 preparations for dynamic linking in place. 991 This has been tested on Linux and Tru64. 992 [Richard Levitte] 993 994 *) Randomness polling function for Win9x, as described in: 995 Peter Gutmann, Software Generation of Practically Strong 996 Random Numbers. 997 [Ulf M�ller] 998 999 *) Fix so PRNG is seeded in req if using an already existing 1000 DSA key. 1001 [Steve Henson] 1002 1003 *) New options to smime application. -inform and -outform 1004 allow alternative formats for the S/MIME message including 1005 PEM and DER. The -content option allows the content to be 1006 specified separately. This should allow things like Netscape 1007 form signing output easier to verify. 1008 [Steve Henson] 1009 1010 *) Fix the ASN1 encoding of tags using the 'long form'. 1011 [Steve Henson] 1012 1013 *) New ASN1 functions, i2c_* and c2i_* for INTEGER and BIT 1014 STRING types. These convert content octets to and from the 1015 underlying type. The actual tag and length octets are 1016 already assumed to have been read in and checked. These 1017 are needed because all other string types have virtually 1018 identical handling apart from the tag. By having versions 1019 of the ASN1 functions that just operate on content octets 1020 IMPLICIT tagging can be handled properly. It also allows 1021 the ASN1_ENUMERATED code to be cut down because ASN1_ENUMERATED 1022 and ASN1_INTEGER are identical apart from the tag. 1023 [Steve Henson] 1024 1025 *) Change the handling of OID objects as follows: 1026 1027 - New object identifiers are inserted in objects.txt, following 1028 the syntax given in objects.README. 1029 - objects.pl is used to process obj_mac.num and create a new 1030 obj_mac.h. 1031 - obj_dat.pl is used to create a new obj_dat.h, using the data in 1032 obj_mac.h. 1033 1034 This is currently kind of a hack, and the perl code in objects.pl 1035 isn't very elegant, but it works as I intended. The simplest way 1036 to check that it worked correctly is to look in obj_dat.h and 1037 check the array nid_objs and make sure the objects haven't moved 1038 around (this is important!). Additions are OK, as well as 1039 consistent name changes. 1040 [Richard Levitte] 1041 1042 *) Add BSD-style MD5-based passwords to 'openssl passwd' (option '-1'). 1043 [Bodo Moeller] 1044 1045 *) Addition of the command line parameter '-rand file' to 'openssl req'. 1046 The given file adds to whatever has already been seeded into the 1047 random pool through the RANDFILE configuration file option or 1048 environment variable, or the default random state file. 1049 [Richard Levitte] 1050 1051 *) mkstack.pl now sorts each macro group into lexical order. 1052 Previously the output order depended on the order the files 1053 appeared in the directory, resulting in needless rewriting 1054 of safestack.h . 1055 [Steve Henson] 1056 1057 *) Patches to make OpenSSL compile under Win32 again. Mostly 1058 work arounds for the VC++ problem that it treats func() as 1059 func(void). Also stripped out the parts of mkdef.pl that 1060 added extra typesafe functions: these no longer exist. 1061 [Steve Henson] 1062 1063 *) Reorganisation of the stack code. The macros are now all 1064 collected in safestack.h . Each macro is defined in terms of 1065 a "stack macro" of the form SKM_<name>(type, a, b). The 1066 DEBUG_SAFESTACK is now handled in terms of function casts, 1067 this has the advantage of retaining type safety without the 1068 use of additional functions. If DEBUG_SAFESTACK is not defined 1069 then the non typesafe macros are used instead. Also modified the 1070 mkstack.pl script to handle the new form. Needs testing to see 1071 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK 1072 the default if no major problems. Similar behaviour for ASN1_SET_OF 1073 and PKCS12_STACK_OF. 1074 [Steve Henson] 1075 1076 *) When some versions of IIS use the 'NET' form of private key the 1077 key derivation algorithm is different. Normally MD5(password) is 1078 used as a 128 bit RC4 key. In the modified case 1079 MD5(MD5(password) + "SGCKEYSALT") is used insted. Added some 1080 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same 1081 as the old Netscape_RSA functions except they have an additional 1082 'sgckey' parameter which uses the modified algorithm. Also added 1083 an -sgckey command line option to the rsa utility. Thanks to 1084 Adrian Peck <bertie@ncipher.com> for posting details of the modified 1085 algorithm to openssl-dev. 1086 [Steve Henson] 1087 1088 *) The evp_local.h macros were using 'c.##kname' which resulted in 1089 invalid expansion on some systems (SCO 5.0.5 for example). 1090 Corrected to 'c.kname'. 1091 [Phillip Porch <root@theporch.com>] 1092 1093 *) New X509_get1_email() and X509_REQ_get1_email() functions that return 1094 a STACK of email addresses from a certificate or request, these look 1095 in the subject name and the subject alternative name extensions and 1096 omit any duplicate addresses. 1097 [Steve Henson] 1098 1099 *) Re-implement BN_mod_exp2_mont using independent (and larger) windows. 1100 This makes DSA verification about 2 % faster. 1101 [Bodo Moeller] 1102 1103 *) Increase maximum window size in BN_mod_exp_... to 6 bits instead of 5 1104 (meaning that now 2^5 values will be precomputed, which is only 4 KB 1105 plus overhead for 1024 bit moduli). 1106 This makes exponentiations about 0.5 % faster for 1024 bit 1107 exponents (as measured by "openssl speed rsa2048"). 1108 [Bodo Moeller] 1109 1110 *) Rename memory handling macros to avoid conflicts with other 1111 software: 1112 Malloc => OPENSSL_malloc 1113 Malloc_locked => OPENSSL_malloc_locked 1114 Realloc => OPENSSL_realloc 1115 Free => OPENSSL_free 1116 [Richard Levitte] 1117 1118 *) New function BN_mod_exp_mont_word for small bases (roughly 15% 1119 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange). 1120 [Bodo Moeller] 1121 1122 *) CygWin32 support. 1123 [John Jarvie <jjarvie@newsguy.com>] 1124 1125 *) The type-safe stack code has been rejigged. It is now only compiled 1126 in when OpenSSL is configured with the DEBUG_SAFESTACK option and 1127 by default all type-specific stack functions are "#define"d back to 1128 standard stack functions. This results in more streamlined output 1129 but retains the type-safety checking possibilities of the original 1130 approach. 1131 [Geoff Thorpe] 1132 1133 *) The STACK code has been cleaned up, and certain type declarations 1134 that didn't make a lot of sense have been brought in line. This has 1135 also involved a cleanup of sorts in safestack.h to more correctly 1136 map type-safe stack functions onto their plain stack counterparts. 1137 This work has also resulted in a variety of "const"ifications of 1138 lots of the code, especially "_cmp" operations which should normally 1139 be prototyped with "const" parameters anyway. 1140 [Geoff Thorpe] 1141 1142 *) When generating bytes for the first time in md_rand.c, 'stir the pool' 1143 by seeding with STATE_SIZE dummy bytes (with zero entropy count). 1144 (The PRNG state consists of two parts, the large pool 'state' and 'md', 1145 where all of 'md' is used each time the PRNG is used, but 'state' 1146 is used only indexed by a cyclic counter. As entropy may not be 1147 well distributed from the beginning, 'md' is important as a 1148 chaining variable. However, the output function chains only half 1149 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains 1150 all of 'md', and seeding with STATE_SIZE dummy bytes will result 1151 in all of 'state' being rewritten, with the new values depending 1152 on virtually all of 'md'. This overcomes the 80 bit limitation.) 1153 [Bodo Moeller] 1154 1155 *) In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when 1156 the handshake is continued after ssl_verify_cert_chain(); 1157 otherwise, if SSL_VERIFY_NONE is set, remaining error codes 1158 can lead to 'unexplainable' connection aborts later. 1159 [Bodo Moeller; problem tracked down by Lutz Jaenicke] 1160 1161 *) Major EVP API cipher revision. 1162 Add hooks for extra EVP features. This allows various cipher 1163 parameters to be set in the EVP interface. Support added for variable 1164 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and 1165 setting of RC2 and RC5 parameters. 1166 1167 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length 1168 ciphers. 1169 1170 Remove lots of duplicated code from the EVP library. For example *every* 1171 cipher init() function handles the 'iv' in the same way according to the 1172 cipher mode. They also all do nothing if the 'key' parameter is NULL and 1173 for CFB and OFB modes they zero ctx->num. 1174 1175 New functionality allows removal of S/MIME code RC2 hack. 1176 1177 Most of the routines have the same form and so can be declared in terms 1178 of macros. 1179 1180 By shifting this to the top level EVP_CipherInit() it can be removed from 1181 all individual ciphers. If the cipher wants to handle IVs or keys 1182 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT 1183 flags. 1184 1185 Change lots of functions like EVP_EncryptUpdate() to now return a 1186 value: although software versions of the algorithms cannot fail 1187 any installed hardware versions can. 1188 [Steve Henson] 1189 1190 *) Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if 1191 this option is set, tolerate broken clients that send the negotiated 1192 protocol version number instead of the requested protocol version 1193 number. 1194 [Bodo Moeller] 1195 1196 *) Call dh_tmp_cb (set by ..._TMP_DH_CB) with correct 'is_export' flag; 1197 i.e. non-zero for export ciphersuites, zero otherwise. 1198 Previous versions had this flag inverted, inconsistent with 1199 rsa_tmp_cb (..._TMP_RSA_CB). 1200 [Bodo Moeller; problem reported by Amit Chopra] 1201 1202 *) Add missing DSA library text string. Work around for some IIS 1203 key files with invalid SEQUENCE encoding. 1204 [Steve Henson] 1205 1206 *) Add a document (doc/standards.txt) that list all kinds of standards 1207 and so on that are implemented in OpenSSL. 1208 [Richard Levitte] 1209 1210 *) Enhance c_rehash script. Old version would mishandle certificates 1211 with the same subject name hash and wouldn't handle CRLs at all. 1212 Added -fingerprint option to crl utility, to support new c_rehash 1213 features. 1214 [Steve Henson] 1215 1216 *) Eliminate non-ANSI declarations in crypto.h and stack.h. 1217 [Ulf M�ller] 1218 1219 *) Fix for SSL server purpose checking. Server checking was 1220 rejecting certificates which had extended key usage present 1221 but no ssl client purpose. 1222 [Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>] 1223 1224 *) Make PKCS#12 code work with no password. The PKCS#12 spec 1225 is a little unclear about how a blank password is handled. 1226 Since the password in encoded as a BMPString with terminating 1227 double NULL a zero length password would end up as just the 1228 double NULL. However no password at all is different and is 1229 handled differently in the PKCS#12 key generation code. NS 1230 treats a blank password as zero length. MSIE treats it as no 1231 password on export: but it will try both on import. We now do 1232 the same: PKCS12_parse() tries zero length and no password if 1233 the password is set to "" or NULL (NULL is now a valid password: 1234 it wasn't before) as does the pkcs12 application. 1235 [Steve Henson] 1236 1237 *) Bugfixes in apps/x509.c: Avoid a memory leak; and don't use 1238 perror when PEM_read_bio_X509_REQ fails, the error message must 1239 be obtained from the error queue. 1240 [Bodo Moeller] 1241 1242 *) Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing 1243 it in ERR_remove_state if appropriate, and change ERR_get_state 1244 accordingly to avoid race conditions (this is necessary because 1245 thread_hash is no longer constant once set). 1246 [Bodo Moeller] 1247 1248 *) Bugfix for linux-elf makefile.one. 1249 [Ulf M�ller] 1250 1251 *) RSA_get_default_method() will now cause a default 1252 RSA_METHOD to be chosen if one doesn't exist already. 1253 Previously this was only set during a call to RSA_new() 1254 or RSA_new_method(NULL) meaning it was possible for 1255 RSA_get_default_method() to return NULL. 1256 [Geoff Thorpe] 1257 1258 *) Added native name translation to the existing DSO code 1259 that will convert (if the flag to do so is set) filenames 1260 that are sufficiently small and have no path information 1261 into a canonical native form. Eg. "blah" converted to 1262 "libblah.so" or "blah.dll" etc. 1263 [Geoff Thorpe] 1264 1265 *) New function ERR_error_string_n(e, buf, len) which is like 1266 ERR_error_string(e, buf), but writes at most 'len' bytes 1267 including the 0 terminator. For ERR_error_string_n, 'buf' 1268 may not be NULL. 1269 [Damien Miller <djm@mindrot.org>, Bodo Moeller] 1270 1271 *) CONF library reworked to become more general. A new CONF 1272 configuration file reader "class" is implemented as well as a 1273 new functions (NCONF_*, for "New CONF") to handle it. The now 1274 old CONF_* functions are still there, but are reimplemented to 1275 work in terms of the new functions. Also, a set of functions 1276 to handle the internal storage of the configuration data is 1277 provided to make it easier to write new configuration file 1278 reader "classes" (I can definitely see something reading a 1279 configuration file in XML format, for example), called _CONF_*, 1280 or "the configuration storage API"... 1281 1282 The new configuration file reading functions are: 1283 1284 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio, 1285 NCONF_get_section, NCONF_get_string, NCONF_get_numbre 1286 1287 NCONF_default, NCONF_WIN32 1288 1289 NCONF_dump_fp, NCONF_dump_bio 1290 1291 NCONF_default and NCONF_WIN32 are method (or "class") choosers, 1292 NCONF_new creates a new CONF object. This works in the same way 1293 as other interfaces in OpenSSL, like the BIO interface. 1294 NCONF_dump_* dump the internal storage of the configuration file, 1295 which is useful for debugging. All other functions take the same 1296 arguments as the old CONF_* functions wth the exception of the 1297 first that must be a `CONF *' instead of a `LHASH *'. 1298 1299 To make it easer to use the new classes with the old CONF_* functions, 1300 the function CONF_set_default_method is provided. 1301 [Richard Levitte] 1302 1303 *) Add '-tls1' option to 'openssl ciphers', which was already 1304 mentioned in the documentation but had not been implemented. 1305 (This option is not yet really useful because even the additional 1306 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.) 1307 [Bodo Moeller] 1308 1309 *) Initial DSO code added into libcrypto for letting OpenSSL (and 1310 OpenSSL-based applications) load shared libraries and bind to 1311 them in a portable way. 1312 [Geoff Thorpe, with contributions from Richard Levitte] 1313 1314 Changes between 0.9.5 and 0.9.5a [1 Apr 2000] 1315 1316 *) Make sure _lrotl and _lrotr are only used with MSVC. 1317 1318 *) Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status 1319 (the default implementation of RAND_status). 1320 1321 *) Rename openssl x509 option '-crlext', which was added in 0.9.5, 1322 to '-clrext' (= clear extensions), as intended and documented. 1323 [Bodo Moeller; inconsistency pointed out by Michael Attili 1324 <attili@amaxo.com>] 1325 1326 *) Fix for HMAC. It wasn't zeroing the rest of the block if the key length 1327 was larger than the MD block size. 1328 [Steve Henson, pointed out by Yost William <YostW@tce.com>] 1329 1330 *) Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument 1331 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set() 1332 using the passed key: if the passed key was a private key the result 1333 of X509_print(), for example, would be to print out all the private key 1334 components. 1335 [Steve Henson] 1336 1337 *) des_quad_cksum() byte order bug fix. 1338 [Ulf M�ller, using the problem description in krb4-0.9.7, where 1339 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>] 1340 1341 *) Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly 1342 discouraged. 1343 [Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>] 1344 1345 *) For easily testing in shell scripts whether some command 1346 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX' 1347 returns with exit code 0 iff no command of the given name is available. 1348 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases, 1349 the output goes to stdout and nothing is printed to stderr. 1350 Additional arguments are always ignored. 1351 1352 Since for each cipher there is a command of the same name, 1353 the 'no-cipher' compilation switches can be tested this way. 1354 1355 ('openssl no-XXX' is not able to detect pseudo-commands such 1356 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.) 1357 [Bodo Moeller] 1358 1359 *) Update test suite so that 'make test' succeeds in 'no-rsa' configuration. 1360 [Bodo Moeller] 1361 1362 *) For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE 1363 is set; it will be thrown away anyway because each handshake creates 1364 its own key. 1365 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition 1366 to parameters -- in previous versions (since OpenSSL 0.9.3) the 1367 'default key' from SSL_CTX_set_tmp_dh would always be lost, meanining 1368 you effectivly got SSL_OP_SINGLE_DH_USE when using this macro. 1369 [Bodo Moeller] 1370 1371 *) New s_client option -ign_eof: EOF at stdin is ignored, and 1372 'Q' and 'R' lose their special meanings (quit/renegotiate). 1373 This is part of what -quiet does; unlike -quiet, -ign_eof 1374 does not suppress any output. 1375 [Richard Levitte] 1376 1377 *) Add compatibility options to the purpose and trust code. The 1378 purpose X509_PURPOSE_ANY is "any purpose" which automatically 1379 accepts a certificate or CA, this was the previous behaviour, 1380 with all the associated security issues. 1381 1382 X509_TRUST_COMPAT is the old trust behaviour: only and 1383 automatically trust self signed roots in certificate store. A 1384 new trust setting X509_TRUST_DEFAULT is used to specify that 1385 a purpose has no associated trust setting and it should instead 1386 use the value in the default purpose. 1387 [Steve Henson] 1388 1389 *) Fix the PKCS#8 DSA private key code so it decodes keys again 1390 and fix a memory leak. 1391 [Steve Henson] 1392 1393 *) In util/mkerr.pl (which implements 'make errors'), preserve 1394 reason strings from the previous version of the .c file, as 1395 the default to have only downcase letters (and digits) in 1396 automatically generated reasons codes is not always appropriate. 1397 [Bodo Moeller] 1398 1399 *) In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table 1400 using strerror. Previously, ERR_reason_error_string() returned 1401 library names as reason strings for SYSerr; but SYSerr is a special 1402 case where small numbers are errno values, not library numbers. 1403 [Bodo Moeller] 1404 1405 *) Add '-dsaparam' option to 'openssl dhparam' application. This 1406 converts DSA parameters into DH parameters. (When creating parameters, 1407 DSA_generate_parameters is used.) 1408 [Bodo Moeller] 1409 1410 *) Include 'length' (recommended exponent length) in C code generated 1411 by 'openssl dhparam -C'. 1412 [Bodo Moeller] 1413 1414 *) The second argument to set_label in perlasm was already being used 1415 so couldn't be used as a "file scope" flag. Moved to third argument 1416 which was free. 1417 [Steve Henson] 1418 1419 *) In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes 1420 instead of RAND_bytes for encryption IVs and salts. 1421 [Bodo Moeller] 1422 1423 *) Include RAND_status() into RAND_METHOD instead of implementing 1424 it only for md_rand.c Otherwise replacing the PRNG by calling 1425 RAND_set_rand_method would be impossible. 1426 [Bodo Moeller] 1427 1428 *) Don't let DSA_generate_key() enter an infinite loop if the random 1429 number generation fails. 1430 [Bodo Moeller] 1431 1432 *) New 'rand' application for creating pseudo-random output. 1433 [Bodo Moeller] 1434 1435 *) Added configuration support for Linux/IA64 1436 [Rolf Haberrecker <rolf@suse.de>] 1437 1438 *) Assembler module support for Mingw32. 1439 [Ulf M�ller] 1440 1441 *) Shared library support for HPUX (in shlib/). 1442 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous] 1443 1444 *) Shared library support for Solaris gcc. 1445 [Lutz Behnke <behnke@trustcenter.de>] 1446 1447 Changes between 0.9.4 and 0.9.5 [28 Feb 2000] 1448 1449 *) PKCS7_encrypt() was adding text MIME headers twice because they 1450 were added manually and by SMIME_crlf_copy(). 1451 [Steve Henson] 1452 1453 *) In bntest.c don't call BN_rand with zero bits argument. 1454 [Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>] 1455 1456 *) BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n] 1457 case was implemented. This caused BN_div_recp() to fail occasionally. 1458 [Ulf M�ller] 1459 1460 *) Add an optional second argument to the set_label() in the perl 1461 assembly language builder. If this argument exists and is set 1462 to 1 it signals that the assembler should use a symbol whose 1463 scope is the entire file, not just the current function. This 1464 is needed with MASM which uses the format label:: for this scope. 1465 [Steve Henson, pointed out by Peter Runestig <peter@runestig.com>] 1466 1467 *) Change the ASN1 types so they are typedefs by default. Before 1468 almost all types were #define'd to ASN1_STRING which was causing 1469 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING) 1470 for example. 1471 [Steve Henson] 1472 1473 *) Change names of new functions to the new get1/get0 naming 1474 convention: After 'get1', the caller owns a reference count 1475 and has to call ..._free; 'get0' returns a pointer to some 1476 data structure without incrementing reference counters. 1477 (Some of the existing 'get' functions increment a reference 1478 counter, some don't.) 1479 Similarly, 'set1' and 'add1' functions increase reference 1480 counters or duplicate objects. 1481 [Steve Henson] 1482 1483 *) Allow for the possibility of temp RSA key generation failure: 1484 the code used to assume it always worked and crashed on failure. 1485 [Steve Henson] 1486 1487 *) Fix potential buffer overrun problem in BIO_printf(). 1488 [Ulf M�ller, using public domain code by Patrick Powell; problem 1489 pointed out by David Sacerdote <das33@cornell.edu>] 1490 1491 *) Support EGD <http://www.lothar.com/tech/crypto/>. New functions 1492 RAND_egd() and RAND_status(). In the command line application, 1493 the EGD socket can be specified like a seed file using RANDFILE 1494 or -rand. 1495 [Ulf M�ller] 1496 1497 *) Allow the string CERTIFICATE to be tolerated in PKCS#7 structures. 1498 Some CAs (e.g. Verisign) distribute certificates in this form. 1499 [Steve Henson] 1500 1501 *) Remove the SSL_ALLOW_ADH compile option and set the default cipher 1502 list to exclude them. This means that no special compilation option 1503 is needed to use anonymous DH: it just needs to be included in the 1504 cipher list. 1505 [Steve Henson] 1506 1507 *) Change the EVP_MD_CTX_type macro so its meaning consistent with 1508 EVP_MD_type. The old functionality is available in a new macro called 1509 EVP_MD_md(). Change code that uses it and update docs. 1510 [Steve Henson] 1511 1512 *) ..._ctrl functions now have corresponding ..._callback_ctrl functions 1513 where the 'void *' argument is replaced by a function pointer argument. 1514 Previously 'void *' was abused to point to functions, which works on 1515 many platforms, but is not correct. As these functions are usually 1516 called by macros defined in OpenSSL header files, most source code 1517 should work without changes. 1518 [Richard Levitte] 1519 1520 *) <openssl/opensslconf.h> (which is created by Configure) now contains 1521 sections with information on -D... compiler switches used for 1522 compiling the library so that applications can see them. To enable 1523 one of these sections, a pre-processor symbol OPENSSL_..._DEFINES 1524 must be defined. E.g., 1525 #define OPENSSL_ALGORITHM_DEFINES 1526 #include <openssl/opensslconf.h> 1527 defines all pertinent NO_<algo> symbols, such as NO_IDEA, NO_RSA, etc. 1528 [Richard Levitte, Ulf and Bodo M�ller] 1529 1530 *) Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS 1531 record layer. 1532 [Bodo Moeller] 1533 1534 *) Change the 'other' type in certificate aux info to a STACK_OF 1535 X509_ALGOR. Although not an AlgorithmIdentifier as such it has 1536 the required ASN1 format: arbitrary types determined by an OID. 1537 [Steve Henson] 1538 1539 *) Add some PEM_write_X509_REQ_NEW() functions and a command line 1540 argument to 'req'. This is not because the function is newer or 1541 better than others it just uses the work 'NEW' in the certificate 1542 request header lines. Some software needs this. 1543 [Steve Henson] 1544 1545 *) Reorganise password command line arguments: now passwords can be 1546 obtained from various sources. Delete the PEM_cb function and make 1547 it the default behaviour: i.e. if the callback is NULL and the 1548 usrdata argument is not NULL interpret it as a null terminated pass 1549 phrase. If usrdata and the callback are NULL then the pass phrase 1550 is prompted for as usual. 1551 [Steve Henson] 1552 1553 *) Add support for the Compaq Atalla crypto accelerator. If it is installed, 1554 the support is automatically enabled. The resulting binaries will 1555 autodetect the card and use it if present. 1556 [Ben Laurie and Compaq Inc.] 1557 1558 *) Work around for Netscape hang bug. This sends certificate request 1559 and server done in one record. Since this is perfectly legal in the 1560 SSL/TLS protocol it isn't a "bug" option and is on by default. See 1561 the bugs/SSLv3 entry for more info. 1562 [Steve Henson] 1563 1564 *) HP-UX tune-up: new unified configs, HP C compiler bug workaround. 1565 [Andy Polyakov] 1566 1567 *) Add -rand argument to smime and pkcs12 applications and read/write 1568 of seed file. 1569 [Steve Henson] 1570 1571 *) New 'passwd' tool for crypt(3) and apr1 password hashes. 1572 [Bodo Moeller] 1573 1574 *) Add command line password options to the remaining applications. 1575 [Steve Henson] 1576 1577 *) Bug fix for BN_div_recp() for numerators with an even number of 1578 bits. 1579 [Ulf M�ller] 1580 1581 *) More tests in bntest.c, and changed test_bn output. 1582 [Ulf M�ller] 1583 1584 *) ./config recognizes MacOS X now. 1585 [Andy Polyakov] 1586 1587 *) Bug fix for BN_div() when the first words of num and divsor are 1588 equal (it gave wrong results if (rem=(n1-q*d0)&BN_MASK2) < d0). 1589 [Ulf M�ller] 1590 1591 *) Add support for various broken PKCS#8 formats, and command line 1592 options to produce them. 1593 [Steve Henson] 1594 1595 *) New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to 1596 get temporary BIGNUMs from a BN_CTX. 1597 [Ulf M�ller] 1598 1599 *) Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont() 1600 for p == 0. 1601 [Ulf M�ller] 1602 1603 *) Change the SSLeay_add_all_*() functions to OpenSSL_add_all_*() and 1604 include a #define from the old name to the new. The original intent 1605 was that statically linked binaries could for example just call 1606 SSLeay_add_all_ciphers() to just add ciphers to the table and not 1607 link with digests. This never worked becayse SSLeay_add_all_digests() 1608 and SSLeay_add_all_ciphers() were in the same source file so calling 1609 one would link with the other. They are now in separate source files. 1610 [Steve Henson] 1611 1612 *) Add a new -notext option to 'ca' and a -pubkey option to 'spkac'. 1613 [Steve Henson] 1614 1615 *) Use a less unusual form of the Miller-Rabin primality test (it used 1616 a binary algorithm for exponentiation integrated into the Miller-Rabin 1617 loop, our standard modexp algorithms are faster). 1618 [Bodo Moeller] 1619 1620 *) Support for the EBCDIC character set completed. 1621 [Martin Kraemer <Martin.Kraemer@Mch.SNI.De>] 1622 1623 *) Source code cleanups: use const where appropriate, eliminate casts, 1624 use void * instead of char * in lhash. 1625 [Ulf M�ller] 1626 1627 *) Bugfix: ssl3_send_server_key_exchange was not restartable 1628 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of 1629 this the server could overwrite ephemeral keys that the client 1630 has already seen). 1631 [Bodo Moeller] 1632 1633 *) Turn DSA_is_prime into a macro that calls BN_is_prime, 1634 using 50 iterations of the Rabin-Miller test. 1635 1636 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50 1637 iterations of the Rabin-Miller test as required by the appendix 1638 to FIPS PUB 186[-1]) instead of DSA_is_prime. 1639 As BN_is_prime_fasttest includes trial division, DSA parameter 1640 generation becomes much faster. 1641 1642 This implies a change for the callback functions in DSA_is_prime 1643 and DSA_generate_parameters: The callback function is called once 1644 for each positive witness in the Rabin-Miller test, not just 1645 occasionally in the inner loop; and the parameters to the 1646 callback function now provide an iteration count for the outer 1647 loop rather than for the current invocation of the inner loop. 1648 DSA_generate_parameters additionally can call the callback 1649 function with an 'iteration count' of -1, meaning that a 1650 candidate has passed the trial division test (when q is generated 1651 from an application-provided seed, trial division is skipped). 1652 [Bodo Moeller] 1653 1654 *) New function BN_is_prime_fasttest that optionally does trial 1655 division before starting the Rabin-Miller test and has 1656 an additional BN_CTX * argument (whereas BN_is_prime always 1657 has to allocate at least one BN_CTX). 1658 'callback(1, -1, cb_arg)' is called when a number has passed the 1659 trial division stage. 1660 [Bodo Moeller] 1661 1662 *) Fix for bug in CRL encoding. The validity dates weren't being handled 1663 as ASN1_TIME. 1664 [Steve Henson] 1665 1666 *) New -pkcs12 option to CA.pl script to write out a PKCS#12 file. 1667 [Steve Henson] 1668 1669 *) New function BN_pseudo_rand(). 1670 [Ulf M�ller] 1671 1672 *) Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable) 1673 bignum version of BN_from_montgomery() with the working code from 1674 SSLeay 0.9.0 (the word based version is faster anyway), and clean up 1675 the comments. 1676 [Ulf M�ller] 1677 1678 *) Avoid a race condition in s2_clnt.c (function get_server_hello) that 1679 made it impossible to use the same SSL_SESSION data structure in 1680 SSL2 clients in multiple threads. 1681 [Bodo Moeller] 1682 1683 *) The return value of RAND_load_file() no longer counts bytes obtained 1684 by stat(). RAND_load_file(..., -1) is new and uses the complete file 1685 to seed the PRNG (previously an explicit byte count was required). 1686 [Ulf M�ller, Bodo M�ller] 1687 1688 *) Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes 1689 used (char *) instead of (void *) and had casts all over the place. 1690 [Steve Henson] 1691 1692 *) Make BN_generate_prime() return NULL on error if ret!=NULL. 1693 [Ulf M�ller] 1694 1695 *) Retain source code compatibility for BN_prime_checks macro: 1696 BN_is_prime(..., BN_prime_checks, ...) now uses 1697 BN_prime_checks_for_size to determine the appropriate number of 1698 Rabin-Miller iterations. 1699 [Ulf M�ller] 1700 1701 *) Diffie-Hellman uses "safe" primes: DH_check() return code renamed to 1702 DH_CHECK_P_NOT_SAFE_PRIME. 1703 (Check if this is true? OpenPGP calls them "strong".) 1704 [Ulf M�ller] 1705 1706 *) Merge the functionality of "dh" and "gendh" programs into a new program 1707 "dhparam". The old programs are retained for now but will handle DH keys 1708 (instead of parameters) in future. 1709 [Steve Henson] 1710 1711 *) Make the ciphers, s_server and s_client programs check the return values 1712 when a new cipher list is set. 1713 [Steve Henson] 1714 1715 *) Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit 1716 ciphers. Before when the 56bit ciphers were enabled the sorting was 1717 wrong. 1718 1719 The syntax for the cipher sorting has been extended to support sorting by 1720 cipher-strength (using the strength_bits hard coded in the tables). 1721 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod). 1722 1723 Fix a bug in the cipher-command parser: when supplying a cipher command 1724 string with an "undefined" symbol (neither command nor alphanumeric 1725 [A-Za-z0-9], ssl_set_cipher_list used to hang in an endless loop. Now 1726 an error is flagged. 1727 1728 Due to the strength-sorting extension, the code of the 1729 ssl_create_cipher_list() function was completely rearranged. I hope that 1730 the readability was also increased :-) 1731 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>] 1732 1733 *) Minor change to 'x509' utility. The -CAcreateserial option now uses 1 1734 for the first serial number and places 2 in the serial number file. This 1735 avoids problems when the root CA is created with serial number zero and 1736 the first user certificate has the same issuer name and serial number 1737 as the root CA. 1738 [Steve Henson] 1739 1740 *) Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses 1741 the new code. Add documentation for this stuff. 1742 [Steve Henson] 1743 1744 *) Changes to X509_ATTRIBUTE utilities. These have been renamed from 1745 X509_*() to X509at_*() on the grounds that they don't handle X509 1746 structures and behave in an analagous way to the X509v3 functions: 1747 they shouldn't be called directly but wrapper functions should be used 1748 instead. 1749 1750 So we also now have some wrapper functions that call the X509at functions 1751 when passed certificate requests. (TO DO: similar things can be done with 1752 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other 1753 things. Some of these need some d2i or i2d and print functionality 1754 because they handle more complex structures.) 1755 [Steve Henson] 1756 1757 *) Add missing #ifndefs that caused missing symbols when building libssl 1758 as a shared library without RSA. Use #ifndef NO_SSL2 instead of 1759 NO_RSA in ssl/s2*.c. 1760 [Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf M�ller] 1761 1762 *) Precautions against using the PRNG uninitialized: RAND_bytes() now 1763 has a return value which indicates the quality of the random data 1764 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's 1765 error queue. New function RAND_pseudo_bytes() generates output that is 1766 guaranteed to be unique but not unpredictable. RAND_add is like 1767 RAND_seed, but takes an extra argument for an entropy estimate 1768 (RAND_seed always assumes full entropy). 1769 [Ulf M�ller] 1770 1771 *) Do more iterations of Rabin-Miller probable prime test (specifically, 1772 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes 1773 instead of only 2 for all lengths; see BN_prime_checks_for_size definition 1774 in crypto/bn/bn_prime.c for the complete table). This guarantees a 1775 false-positive rate of at most 2^-80 for random input. 1776 [Bodo Moeller] 1777 1778 *) Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs. 1779 [Bodo Moeller] 1780 1781 *) New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain 1782 in the 0.9.5 release), this returns the chain 1783 from an X509_CTX structure with a dup of the stack and all 1784 the X509 reference counts upped: so the stack will exist 1785 after X509_CTX_cleanup() has been called. Modify pkcs12.c 1786 to use this. 1787 1788 Also make SSL_SESSION_print() print out the verify return 1789 code. 1790 [Steve Henson] 1791 1792 *) Add manpage for the pkcs12 command. Also change the default 1793 behaviour so MAC iteration counts are used unless the new 1794 -nomaciter option is used. This improves file security and 1795 only older versions of MSIE (4.0 for example) need it. 1796 [Steve Henson] 1797 1798 *) Honor the no-xxx Configure options when creating .DEF files. 1799 [Ulf M�ller] 1800 1801 *) Add PKCS#10 attributes to field table: challengePassword, 1802 unstructuredName and unstructuredAddress. These are taken from 1803 draft PKCS#9 v2.0 but are compatible with v1.2 provided no 1804 international characters are used. 1805 1806 More changes to X509_ATTRIBUTE code: allow the setting of types 1807 based on strings. Remove the 'loc' parameter when adding 1808 attributes because these will be a SET OF encoding which is sorted 1809 in ASN1 order. 1810 [Steve Henson] 1811 1812 *) Initial changes to the 'req' utility to allow request generation 1813 automation. This will allow an application to just generate a template 1814 file containing all the field values and have req construct the 1815 request. 1816 1817 Initial support for X509_ATTRIBUTE handling. Stacks of these are 1818 used all over the place including certificate requests and PKCS#7 1819 structures. They are currently handled manually where necessary with 1820 some primitive wrappers for PKCS#7. The new functions behave in a 1821 manner analogous to the X509 extension functions: they allow 1822 attributes to be looked up by NID and added. 1823 1824 Later something similar to the X509V3 code would be desirable to 1825 automatically handle the encoding, decoding and printing of the 1826 more complex types. The string types like challengePassword can 1827 be handled by the string table functions. 1828 1829 Also modified the multi byte string table handling. Now there is 1830 a 'global mask' which masks out certain types. The table itself 1831 can use the flag STABLE_NO_MASK to ignore the mask setting: this 1832 is useful when for example there is only one permissible type 1833 (as in countryName) and using the mask might result in no valid 1834 types at all. 1835 [Steve Henson] 1836 1837 *) Clean up 'Finished' handling, and add functions SSL_get_finished and 1838 SSL_get_peer_finished to allow applications to obtain the latest 1839 Finished messages sent to the peer or expected from the peer, 1840 respectively. (SSL_get_peer_finished is usually the Finished message 1841 actually received from the peer, otherwise the protocol will be aborted.) 1842 1843 As the Finished message are message digests of the complete handshake 1844 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can 1845 be used for external authentication procedures when the authentication 1846 provided by SSL/TLS is not desired or is not enough. 1847 [Bodo Moeller] 1848 1849 *) Enhanced support for Alpha Linux is added. Now ./config checks if 1850 the host supports BWX extension and if Compaq C is present on the 1851 $PATH. Just exploiting of the BWX extension results in 20-30% 1852 performance kick for some algorithms, e.g. DES and RC4 to mention 1853 a couple. Compaq C in turn generates ~20% faster code for MD5 and 1854 SHA1. 1855 [Andy Polyakov] 1856 1857 *) Add support for MS "fast SGC". This is arguably a violation of the 1858 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with 1859 weak crypto and after checking the certificate is SGC a second one 1860 with strong crypto. MS SGC stops the first handshake after receiving 1861 the server certificate message and sends a second client hello. Since 1862 a server will typically do all the time consuming operations before 1863 expecting any further messages from the client (server key exchange 1864 is the most expensive) there is little difference between the two. 1865 1866 To get OpenSSL to support MS SGC we have to permit a second client 1867 hello message after we have sent server done. In addition we have to 1868 reset the MAC if we do get this second client hello. 1869 [Steve Henson] 1870 1871 *) Add a function 'd2i_AutoPrivateKey()' this will automatically decide 1872 if a DER encoded private key is RSA or DSA traditional format. Changed 1873 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional" 1874 format DER encoded private key. Newer code should use PKCS#8 format which 1875 has the key type encoded in the ASN1 structure. Added DER private key 1876 support to pkcs8 application. 1877 [Steve Henson] 1878 1879 *) SSL 3/TLS 1 servers now don't request certificates when an anonymous 1880 ciphersuites has been selected (as required by the SSL 3/TLS 1 1881 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT 1882 is set, we interpret this as a request to violate the specification 1883 (the worst that can happen is a handshake failure, and 'correct' 1884 behaviour would result in a handshake failure anyway). 1885 [Bodo Moeller] 1886 1887 *) In SSL_CTX_add_session, take into account that there might be multiple 1888 SSL_SESSION structures with the same session ID (e.g. when two threads 1889 concurrently obtain them from an external cache). 1890 The internal cache can handle only one SSL_SESSION with a given ID, 1891 so if there's a conflict, we now throw out the old one to achieve 1892 consistency. 1893 [Bodo Moeller] 1894 1895 *) Add OIDs for idea and blowfish in CBC mode. This will allow both 1896 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to 1897 some routines that use cipher OIDs: some ciphers do not have OIDs 1898 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for 1899 example. 1900 [Steve Henson] 1901 1902 *) Simplify the trust setting structure and code. Now we just have 1903 two sequences of OIDs for trusted and rejected settings. These will 1904 typically have values the same as the extended key usage extension 1905 and any application specific purposes. 1906 1907 The trust checking code now has a default behaviour: it will just 1908 check for an object with the same NID as the passed id. Functions can 1909 be provided to override either the default behaviour or the behaviour 1910 for a given id. SSL client, server and email already have functions 1911 in place for compatibility: they check the NID and also return "trusted" 1912 if the certificate is self signed. 1913 [Steve Henson] 1914 1915 *) Add d2i,i2d bio/fp functions for PrivateKey: these convert the 1916 traditional format into an EVP_PKEY structure. 1917 [Steve Henson] 1918 1919 *) Add a password callback function PEM_cb() which either prompts for 1920 a password if usr_data is NULL or otherwise assumes it is a null 1921 terminated password. Allow passwords to be passed on command line 1922 environment or config files in a few more utilities. 1923 [Steve Henson] 1924 1925 *) Add a bunch of DER and PEM functions to handle PKCS#8 format private 1926 keys. Add some short names for PKCS#8 PBE algorithms and allow them 1927 to be specified on the command line for the pkcs8 and pkcs12 utilities. 1928 Update documentation. 1929 [Steve Henson] 1930 1931 *) Support for ASN1 "NULL" type. This could be handled before by using 1932 ASN1_TYPE but there wasn't any function that would try to read a NULL 1933 and produce an error if it couldn't. For compatibility we also have 1934 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and 1935 don't allocate anything because they don't need to. 1936 [Steve Henson] 1937 1938 *) Initial support for MacOS is now provided. Examine INSTALL.MacOS 1939 for details. 1940 [Andy Polyakov, Roy Woods <roy@centicsystems.ca>] 1941 1942 *) Rebuild of the memory allocation routines used by OpenSSL code and 1943 possibly others as well. The purpose is to make an interface that 1944 provide hooks so anyone can build a separate set of allocation and 1945 deallocation routines to be used by OpenSSL, for example memory 1946 pool implementations, or something else, which was previously hard 1947 since Malloc(), Realloc() and Free() were defined as macros having 1948 the values malloc, realloc and free, respectively (except for Win32 1949 compilations). The same is provided for memory debugging code. 1950 OpenSSL already comes with functionality to find memory leaks, but 1951 this gives people a chance to debug other memory problems. 1952 1953 With these changes, a new set of functions and macros have appeared: 1954 1955 CRYPTO_set_mem_debug_functions() [F] 1956 CRYPTO_get_mem_debug_functions() [F] 1957 CRYPTO_dbg_set_options() [F] 1958 CRYPTO_dbg_get_options() [F] 1959 CRYPTO_malloc_debug_init() [M] 1960 1961 The memory debug functions are NULL by default, unless the library 1962 is compiled with CRYPTO_MDEBUG or friends is defined. If someone 1963 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which 1964 gives the standard debugging functions that come with OpenSSL) or 1965 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions 1966 provided by the library user) must be used. When the standard 1967 debugging functions are used, CRYPTO_dbg_set_options can be used to 1968 request additional information: 1969 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting 1970 the CRYPTO_MDEBUG_xxx macro when compiling the library. 1971 1972 Also, things like CRYPTO_set_mem_functions will always give the 1973 expected result (the new set of functions is used for allocation 1974 and deallocation) at all times, regardless of platform and compiler 1975 options. 1976 1977 To finish it up, some functions that were never use in any other 1978 way than through macros have a new API and new semantic: 1979 1980 CRYPTO_dbg_malloc() 1981 CRYPTO_dbg_realloc() 1982 CRYPTO_dbg_free() 1983 1984 All macros of value have retained their old syntax. 1985 [Richard Levitte and Bodo Moeller] 1986 1987 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the 1988 ordering of SMIMECapabilities wasn't in "strength order" and there 1989 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature 1990 algorithm. 1991 [Steve Henson] 1992 1993 *) Some ASN1 types with illegal zero length encoding (INTEGER, 1994 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines. 1995 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson] 1996 1997 *) Merge in my S/MIME library for OpenSSL. This provides a simple 1998 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough 1999 functionality to handle multipart/signed properly) and a utility 2000 called 'smime' to call all this stuff. This is based on code I 2001 originally wrote for Celo who have kindly allowed it to be 2002 included in OpenSSL. 2003 [Steve Henson] 2004 2005 *) Add variants des_set_key_checked and des_set_key_unchecked of 2006 des_set_key (aka des_key_sched). Global variable des_check_key 2007 decides which of these is called by des_set_key; this way 2008 des_check_key behaves as it always did, but applications and 2009 the library itself, which was buggy for des_check_key == 1, 2010 have a cleaner way to pick the version they need. 2011 [Bodo Moeller] 2012 2013 *) New function PKCS12_newpass() which changes the password of a 2014 PKCS12 structure. 2015 [Steve Henson] 2016 2017 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and 2018 dynamic mix. In both cases the ids can be used as an index into the 2019 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add() 2020 functions so they accept a list of the field values and the 2021 application doesn't need to directly manipulate the X509_TRUST 2022 structure. 2023 [Steve Henson] 2024 2025 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't 2026 need initialising. 2027 [Steve Henson] 2028 2029 *) Modify the way the V3 extension code looks up extensions. This now 2030 works in a similar way to the object code: we have some "standard" 2031 extensions in a static table which is searched with OBJ_bsearch() 2032 and the application can add dynamic ones if needed. The file 2033 crypto/x509v3/ext_dat.h now has the info: this file needs to be 2034 updated whenever a new extension is added to the core code and kept 2035 in ext_nid order. There is a simple program 'tabtest.c' which checks 2036 this. New extensions are not added too often so this file can readily 2037 be maintained manually. 2038 2039 There are two big advantages in doing things this way. The extensions 2040 can be looked up immediately and no longer need to be "added" using 2041 X509V3_add_standard_extensions(): this function now does nothing. 2042 [Side note: I get *lots* of email saying the extension code doesn't 2043 work because people forget to call this function] 2044 Also no dynamic allocation is done unless new extensions are added: 2045 so if we don't add custom extensions there is no need to call 2046 X509V3_EXT_cleanup(). 2047 [Steve Henson] 2048 2049 *) Modify enc utility's salting as follows: make salting the default. Add a 2050 magic header, so unsalted files fail gracefully instead of just decrypting 2051 to garbage. This is because not salting is a big security hole, so people 2052 should be discouraged from doing it. 2053 [Ben Laurie] 2054 2055 *) Fixes and enhancements to the 'x509' utility. It allowed a message 2056 digest to be passed on the command line but it only used this 2057 parameter when signing a certificate. Modified so all relevant 2058 operations are affected by the digest parameter including the 2059 -fingerprint and -x509toreq options. Also -x509toreq choked if a 2060 DSA key was used because it didn't fix the digest. 2061 [Steve Henson] 2062 2063 *) Initial certificate chain verify code. Currently tests the untrusted 2064 certificates for consistency with the verify purpose (which is set 2065 when the X509_STORE_CTX structure is set up) and checks the pathlength. 2066 2067 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour: 2068 this is because it will reject chains with invalid extensions whereas 2069 every previous version of OpenSSL and SSLeay made no checks at all. 2070 2071 Trust code: checks the root CA for the relevant trust settings. Trust 2072 settings have an initial value consistent with the verify purpose: e.g. 2073 if the verify purpose is for SSL client use it expects the CA to be 2074 trusted for SSL client use. However the default value can be changed to 2075 permit custom trust settings: one example of this would be to only trust 2076 certificates from a specific "secure" set of CAs. 2077 2078 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions 2079 which should be used for version portability: especially since the 2080 verify structure is likely to change more often now. 2081 2082 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions 2083 to set them. If not set then assume SSL clients will verify SSL servers 2084 and vice versa. 2085 2086 Two new options to the verify program: -untrusted allows a set of 2087 untrusted certificates to be passed in and -purpose which sets the 2088 intended purpose of the certificate. If a purpose is set then the 2089 new chain verify code is used to check extension consistency. 2090 [Steve Henson] 2091 2092 *) Support for the authority information access extension. 2093 [Steve Henson] 2094 2095 *) Modify RSA and DSA PEM read routines to transparently handle 2096 PKCS#8 format private keys. New *_PUBKEY_* functions that handle 2097 public keys in a format compatible with certificate 2098 SubjectPublicKeyInfo structures. Unfortunately there were already 2099 functions called *_PublicKey_* which used various odd formats so 2100 these are retained for compatibility: however the DSA variants were 2101 never in a public release so they have been deleted. Changed dsa/rsa 2102 utilities to handle the new format: note no releases ever handled public 2103 keys so we should be OK. 2104 2105 The primary motivation for this change is to avoid the same fiasco 2106 that dogs private keys: there are several incompatible private key 2107 formats some of which are standard and some OpenSSL specific and 2108 require various evil hacks to allow partial transparent handling and 2109 even then it doesn't work with DER formats. Given the option anything 2110 other than PKCS#8 should be dumped: but the other formats have to 2111 stay in the name of compatibility. 2112 2113 With public keys and the benefit of hindsight one standard format 2114 is used which works with EVP_PKEY, RSA or DSA structures: though 2115 it clearly returns an error if you try to read the wrong kind of key. 2116 2117 Added a -pubkey option to the 'x509' utility to output the public key. 2118 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*() 2119 (renamed to EVP_PKEY_get1_*() in the OpenSSL 0.9.5 release) and add 2120 EVP_PKEY_rset_*() functions (renamed to EVP_PKEY_set1_*()) 2121 that do the same as the EVP_PKEY_assign_*() except they up the 2122 reference count of the added key (they don't "swallow" the 2123 supplied key). 2124 [Steve Henson] 2125 2126 *) Fixes to crypto/x509/by_file.c the code to read in certificates and 2127 CRLs would fail if the file contained no certificates or no CRLs: 2128 added a new function to read in both types and return the number 2129 read: this means that if none are read it will be an error. The 2130 DER versions of the certificate and CRL reader would always fail 2131 because it isn't possible to mix certificates and CRLs in DER format 2132 without choking one or the other routine. Changed this to just read 2133 a certificate: this is the best we can do. Also modified the code 2134 in apps/verify.c to take notice of return codes: it was previously 2135 attempting to read in certificates from NULL pointers and ignoring 2136 any errors: this is one reason why the cert and CRL reader seemed 2137 to work. It doesn't check return codes from the default certificate 2138 routines: these may well fail if the certificates aren't installed. 2139 [Steve Henson] 2140 2141 *) Code to support otherName option in GeneralName. 2142 [Steve Henson] 2143 2144 *) First update to verify code. Change the verify utility 2145 so it warns if it is passed a self signed certificate: 2146 for consistency with the normal behaviour. X509_verify 2147 has been modified to it will now verify a self signed 2148 certificate if *exactly* the same certificate appears 2149 in the store: it was previously impossible to trust a 2150 single self signed certificate. This means that: 2151 openssl verify ss.pem 2152 now gives a warning about a self signed certificate but 2153 openssl verify -CAfile ss.pem ss.pem 2154 is OK. 2155 [Steve Henson] 2156 2157 *) For servers, store verify_result in SSL_SESSION data structure 2158 (and add it to external session representation). 2159 This is needed when client certificate verifications fails, 2160 but an application-provided verification callback (set by 2161 SSL_CTX_set_cert_verify_callback) allows accepting the session 2162 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK 2163 but returns 1): When the session is reused, we have to set 2164 ssl->verify_result to the appropriate error code to avoid 2165 security holes. 2166 [Bodo Moeller, problem pointed out by Lutz Jaenicke] 2167 2168 *) Fix a bug in the new PKCS#7 code: it didn't consider the 2169 case in PKCS7_dataInit() where the signed PKCS7 structure 2170 didn't contain any existing data because it was being created. 2171 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson] 2172 2173 *) Add a salt to the key derivation routines in enc.c. This 2174 forms the first 8 bytes of the encrypted file. Also add a 2175 -S option to allow a salt to be input on the command line. 2176 [Steve Henson] 2177 2178 *) New function X509_cmp(). Oddly enough there wasn't a function 2179 to compare two certificates. We do this by working out the SHA1 2180 hash and comparing that. X509_cmp() will be needed by the trust 2181 code. 2182 [Steve Henson] 2183 2184 *) SSL_get1_session() is like SSL_get_session(), but increments 2185 the reference count in the SSL_SESSION returned. 2186 [Geoff Thorpe <geoff@eu.c2.net>] 2187 2188 *) Fix for 'req': it was adding a null to request attributes. 2189 Also change the X509_LOOKUP and X509_INFO code to handle 2190 certificate auxiliary information. 2191 [Steve Henson] 2192 2193 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document 2194 the 'enc' command. 2195 [Steve Henson] 2196 2197 *) Add the possibility to add extra information to the memory leak 2198 detecting output, to form tracebacks, showing from where each 2199 allocation was originated: CRYPTO_push_info("constant string") adds 2200 the string plus current file name and line number to a per-thread 2201 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info() 2202 is like calling CYRPTO_pop_info() until the stack is empty. 2203 Also updated memory leak detection code to be multi-thread-safe. 2204 [Richard Levitte] 2205 2206 *) Add options -text and -noout to pkcs7 utility and delete the 2207 encryption options which never did anything. Update docs. 2208 [Steve Henson] 2209 2210 *) Add options to some of the utilities to allow the pass phrase 2211 to be included on either the command line (not recommended on 2212 OSes like Unix) or read from the environment. Update the 2213 manpages and fix a few bugs. 2214 [Steve Henson] 2215 2216 *) Add a few manpages for some of the openssl commands. 2217 [Steve Henson] 2218 2219 *) Fix the -revoke option in ca. It was freeing up memory twice, 2220 leaking and not finding already revoked certificates. 2221 [Steve Henson] 2222 2223 *) Extensive changes to support certificate auxiliary information. 2224 This involves the use of X509_CERT_AUX structure and X509_AUX 2225 functions. An X509_AUX function such as PEM_read_X509_AUX() 2226 can still read in a certificate file in the usual way but it 2227 will also read in any additional "auxiliary information". By 2228 doing things this way a fair degree of compatibility can be 2229 retained: existing certificates can have this information added 2230 using the new 'x509' options. 2231 2232 Current auxiliary information includes an "alias" and some trust 2233 settings. The trust settings will ultimately be used in enhanced 2234 certificate chain verification routines: currently a certificate 2235 can only be trusted if it is self signed and then it is trusted 2236 for all purposes. 2237 [Steve Henson] 2238 2239 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD). 2240 The problem was that one of the replacement routines had not been working 2241 since SSLeay releases. For now the offending routine has been replaced 2242 with non-optimised assembler. Even so, this now gives around 95% 2243 performance improvement for 1024 bit RSA signs. 2244 [Mark Cox] 2245 2246 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2 2247 handling. Most clients have the effective key size in bits equal to 2248 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key. 2249 A few however don't do this and instead use the size of the decrypted key 2250 to determine the RC2 key length and the AlgorithmIdentifier to determine 2251 the effective key length. In this case the effective key length can still 2252 be 40 bits but the key length can be 168 bits for example. This is fixed 2253 by manually forcing an RC2 key into the EVP_PKEY structure because the 2254 EVP code can't currently handle unusual RC2 key sizes: it always assumes 2255 the key length and effective key length are equal. 2256 [Steve Henson] 2257 2258 *) Add a bunch of functions that should simplify the creation of 2259 X509_NAME structures. Now you should be able to do: 2260 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0); 2261 and have it automatically work out the correct field type and fill in 2262 the structures. The more adventurous can try: 2263 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0); 2264 and it will (hopefully) work out the correct multibyte encoding. 2265 [Steve Henson] 2266 2267 *) Change the 'req' utility to use the new field handling and multibyte 2268 copy routines. Before the DN field creation was handled in an ad hoc 2269 way in req, ca, and x509 which was rather broken and didn't support 2270 BMPStrings or UTF8Strings. Since some software doesn't implement 2271 BMPStrings or UTF8Strings yet, they can be enabled using the config file 2272 using the dirstring_type option. See the new comment in the default 2273 openssl.cnf for more info. 2274 [Steve Henson] 2275 2276 *) Make crypto/rand/md_rand.c more robust: 2277 - Assure unique random numbers after fork(). 2278 - Make sure that concurrent threads access the global counter and 2279 md serializably so that we never lose entropy in them 2280 or use exactly the same state in multiple threads. 2281 Access to the large state is not always serializable because 2282 the additional locking could be a performance killer, and 2283 md should be large enough anyway. 2284 [Bodo Moeller] 2285 2286 *) New file apps/app_rand.c with commonly needed functionality 2287 for handling the random seed file. 2288 2289 Use the random seed file in some applications that previously did not: 2290 ca, 2291 dsaparam -genkey (which also ignored its '-rand' option), 2292 s_client, 2293 s_server, 2294 x509 (when signing). 2295 Except on systems with /dev/urandom, it is crucial to have a random 2296 seed file at least for key creation, DSA signing, and for DH exchanges; 2297 for RSA signatures we could do without one. 2298 2299 gendh and gendsa (unlike genrsa) used to read only the first byte 2300 of each file listed in the '-rand' option. The function as previously 2301 found in genrsa is now in app_rand.c and is used by all programs 2302 that support '-rand'. 2303 [Bodo Moeller] 2304 2305 *) In RAND_write_file, use mode 0600 for creating files; 2306 don't just chmod when it may be too late. 2307 [Bodo Moeller] 2308 2309 *) Report an error from X509_STORE_load_locations 2310 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed. 2311 [Bill Perry] 2312 2313 *) New function ASN1_mbstring_copy() this copies a string in either 2314 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format 2315 into an ASN1_STRING type. A mask of permissible types is passed 2316 and it chooses the "minimal" type to use or an error if not type 2317 is suitable. 2318 [Steve Henson] 2319 2320 *) Add function equivalents to the various macros in asn1.h. The old 2321 macros are retained with an M_ prefix. Code inside the library can 2322 use the M_ macros. External code (including the openssl utility) 2323 should *NOT* in order to be "shared library friendly". 2324 [Steve Henson] 2325 2326 *) Add various functions that can check a certificate's extensions 2327 to see if it usable for various purposes such as SSL client, 2328 server or S/MIME and CAs of these types. This is currently 2329 VERY EXPERIMENTAL but will ultimately be used for certificate chain 2330 verification. Also added a -purpose flag to x509 utility to 2331 print out all the purposes. 2332 [Steve Henson] 2333 2334 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated 2335 functions. 2336 [Steve Henson] 2337 2338 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search 2339 for, obtain and decode and extension and obtain its critical flag. 2340 This allows all the necessary extension code to be handled in a 2341 single function call. 2342 [Steve Henson] 2343 2344 *) RC4 tune-up featuring 30-40% performance improvement on most RISC 2345 platforms. See crypto/rc4/rc4_enc.c for further details. 2346 [Andy Polyakov] 2347 2348 *) New -noout option to asn1parse. This causes no output to be produced 2349 its main use is when combined with -strparse and -out to extract data 2350 from a file (which may not be in ASN.1 format). 2351 [Steve Henson] 2352 2353 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer 2354 when producing the local key id. 2355 [Richard Levitte <levitte@stacken.kth.se>] 2356 2357 *) New option -dhparam in s_server. This allows a DH parameter file to be 2358 stated explicitly. If it is not stated then it tries the first server 2359 certificate file. The previous behaviour hard coded the filename 2360 "server.pem". 2361 [Steve Henson] 2362 2363 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow 2364 a public key to be input or output. For example: 2365 openssl rsa -in key.pem -pubout -out pubkey.pem 2366 Also added necessary DSA public key functions to handle this. 2367 [Steve Henson] 2368 2369 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained 2370 in the message. This was handled by allowing 2371 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it. 2372 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>] 2373 2374 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null 2375 to the end of the strings whereas this didn't. This would cause problems 2376 if strings read with d2i_ASN1_bytes() were later modified. 2377 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>] 2378 2379 *) Fix for base64 decode bug. When a base64 bio reads only one line of 2380 data and it contains EOF it will end up returning an error. This is 2381 caused by input 46 bytes long. The cause is due to the way base64 2382 BIOs find the start of base64 encoded data. They do this by trying a 2383 trial decode on each line until they find one that works. When they 2384 do a flag is set and it starts again knowing it can pass all the 2385 data directly through the decoder. Unfortunately it doesn't reset 2386 the context it uses. This means that if EOF is reached an attempt 2387 is made to pass two EOFs through the context and this causes the 2388 resulting error. This can also cause other problems as well. As is 2389 usual with these problems it takes *ages* to find and the fix is 2390 trivial: move one line. 2391 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ] 2392 2393 *) Ugly workaround to get s_client and s_server working under Windows. The 2394 old code wouldn't work because it needed to select() on sockets and the 2395 tty (for keypresses and to see if data could be written). Win32 only 2396 supports select() on sockets so we select() with a 1s timeout on the 2397 sockets and then see if any characters are waiting to be read, if none 2398 are present then we retry, we also assume we can always write data to 2399 the tty. This isn't nice because the code then blocks until we've 2400 received a complete line of data and it is effectively polling the 2401 keyboard at 1s intervals: however it's quite a bit better than not 2402 working at all :-) A dedicated Windows application might handle this 2403 with an event loop for example. 2404 [Steve Henson] 2405 2406 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign 2407 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions 2408 will be called when RSA_sign() and RSA_verify() are used. This is useful 2409 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available. 2410 For this to work properly RSA_public_decrypt() and RSA_private_encrypt() 2411 should *not* be used: RSA_sign() and RSA_verify() must be used instead. 2412 This necessitated the support of an extra signature type NID_md5_sha1 2413 for SSL signatures and modifications to the SSL library to use it instead 2414 of calling RSA_public_decrypt() and RSA_private_encrypt(). 2415 [Steve Henson] 2416 2417 *) Add new -verify -CAfile and -CApath options to the crl program, these 2418 will lookup a CRL issuers certificate and verify the signature in a 2419 similar way to the verify program. Tidy up the crl program so it 2420 no longer accesses structures directly. Make the ASN1 CRL parsing a bit 2421 less strict. It will now permit CRL extensions even if it is not 2422 a V2 CRL: this will allow it to tolerate some broken CRLs. 2423 [Steve Henson] 2424 2425 *) Initialize all non-automatic variables each time one of the openssl 2426 sub-programs is started (this is necessary as they may be started 2427 multiple times from the "OpenSSL>" prompt). 2428 [Lennart Bang, Bodo Moeller] 2429 2430 *) Preliminary compilation option RSA_NULL which disables RSA crypto without 2431 removing all other RSA functionality (this is what NO_RSA does). This 2432 is so (for example) those in the US can disable those operations covered 2433 by the RSA patent while allowing storage and parsing of RSA keys and RSA 2434 key generation. 2435 [Steve Henson] 2436 2437 *) Non-copying interface to BIO pairs. 2438 (still largely untested) 2439 [Bodo Moeller] 2440 2441 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive 2442 ASCII string. This was handled independently in various places before. 2443 [Steve Henson] 2444 2445 *) New functions UTF8_getc() and UTF8_putc() that parse and generate 2446 UTF8 strings a character at a time. 2447 [Steve Henson] 2448 2449 *) Use client_version from client hello to select the protocol 2450 (s23_srvr.c) and for RSA client key exchange verification 2451 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications. 2452 [Bodo Moeller] 2453 2454 *) Add various utility functions to handle SPKACs, these were previously 2455 handled by poking round in the structure internals. Added new function 2456 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to 2457 print, verify and generate SPKACs. Based on an original idea from 2458 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified. 2459 [Steve Henson] 2460 2461 *) RIPEMD160 is operational on all platforms and is back in 'make test'. 2462 [Andy Polyakov] 2463 2464 *) Allow the config file extension section to be overwritten on the 2465 command line. Based on an original idea from Massimiliano Pala 2466 <madwolf@comune.modena.it>. The new option is called -extensions 2467 and can be applied to ca, req and x509. Also -reqexts to override 2468 the request extensions in req and -crlexts to override the crl extensions 2469 in ca. 2470 [Steve Henson] 2471 2472 *) Add new feature to the SPKAC handling in ca. Now you can include 2473 the same field multiple times by preceding it by "XXXX." for example: 2474 1.OU="Unit name 1" 2475 2.OU="Unit name 2" 2476 this is the same syntax as used in the req config file. 2477 [Steve Henson] 2478 2479 *) Allow certificate extensions to be added to certificate requests. These 2480 are specified in a 'req_extensions' option of the req section of the 2481 config file. They can be printed out with the -text option to req but 2482 are otherwise ignored at present. 2483 [Steve Henson] 2484 2485 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first 2486 data read consists of only the final block it would not decrypted because 2487 EVP_CipherUpdate() would correctly report zero bytes had been decrypted. 2488 A misplaced 'break' also meant the decrypted final block might not be 2489 copied until the next read. 2490 [Steve Henson] 2491 2492 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added 2493 a few extra parameters to the DH structure: these will be useful if 2494 for example we want the value of 'q' or implement X9.42 DH. 2495 [Steve Henson] 2496 2497 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and 2498 provides hooks that allow the default DSA functions or functions on a 2499 "per key" basis to be replaced. This allows hardware acceleration and 2500 hardware key storage to be handled without major modification to the 2501 library. Also added low level modexp hooks and CRYPTO_EX structure and 2502 associated functions. 2503 [Steve Henson] 2504 2505 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO 2506 as "read only": it can't be written to and the buffer it points to will 2507 not be freed. Reading from a read only BIO is much more efficient than 2508 a normal memory BIO. This was added because there are several times when 2509 an area of memory needs to be read from a BIO. The previous method was 2510 to create a memory BIO and write the data to it, this results in two 2511 copies of the data and an O(n^2) reading algorithm. There is a new 2512 function BIO_new_mem_buf() which creates a read only memory BIO from 2513 an area of memory. Also modified the PKCS#7 routines to use read only 2514 memory BIOs. 2515 [Steve Henson] 2516 2517 *) Bugfix: ssl23_get_client_hello did not work properly when called in 2518 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of 2519 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read, 2520 but a retry condition occured while trying to read the rest. 2521 [Bodo Moeller] 2522 2523 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as 2524 NID_pkcs7_encrypted by default: this was wrong since this should almost 2525 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle 2526 the encrypted data type: this is a more sensible place to put it and it 2527 allows the PKCS#12 code to be tidied up that duplicated this 2528 functionality. 2529 [Steve Henson] 2530 2531 *) Changed obj_dat.pl script so it takes its input and output files on 2532 the command line. This should avoid shell escape redirection problems 2533 under Win32. 2534 [Steve Henson] 2535 2536 *) Initial support for certificate extension requests, these are included 2537 in things like Xenroll certificate requests. Included functions to allow 2538 extensions to be obtained and added. 2539 [Steve Henson] 2540 2541 *) -crlf option to s_client and s_server for sending newlines as 2542 CRLF (as required by many protocols). 2543 [Bodo Moeller] 2544 2545 Changes between 0.9.3a and 0.9.4 [09 Aug 1999] 2546 2547 *) Install libRSAglue.a when OpenSSL is built with RSAref. 2548 [Ralf S. Engelschall] 2549 2550 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency. 2551 [Andrija Antonijevic <TheAntony2@bigfoot.com>] 2552 2553 *) Fix -startdate and -enddate (which was missing) arguments to 'ca' 2554 program. 2555 [Steve Henson] 2556 2557 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as 2558 DH parameters/keys (q is lost during that conversion, but the resulting 2559 DH parameters contain its length). 2560 2561 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is 2562 much faster than DH_generate_parameters (which creates parameters 2563 where p = 2*q + 1), and also the smaller q makes DH computations 2564 much more efficient (160-bit exponentiation instead of 1024-bit 2565 exponentiation); so this provides a convenient way to support DHE 2566 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of 2567 utter importance to use 2568 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE); 2569 or 2570 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE); 2571 when such DH parameters are used, because otherwise small subgroup 2572 attacks may become possible! 2573 [Bodo Moeller] 2574 2575 *) Avoid memory leak in i2d_DHparams. 2576 [Bodo Moeller] 2577 2578 *) Allow the -k option to be used more than once in the enc program: 2579 this allows the same encrypted message to be read by multiple recipients. 2580 [Steve Henson] 2581 2582 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts 2583 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then 2584 it will always use the numerical form of the OID, even if it has a short 2585 or long name. 2586 [Steve Henson] 2587 2588 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp 2589 method only got called if p,q,dmp1,dmq1,iqmp components were present, 2590 otherwise bn_mod_exp was called. In the case of hardware keys for example 2591 no private key components need be present and it might store extra data 2592 in the RSA structure, which cannot be accessed from bn_mod_exp. 2593 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for 2594 private key operations. 2595 [Steve Henson] 2596 2597 *) Added support for SPARC Linux. 2598 [Andy Polyakov] 2599 2600 *) pem_password_cb function type incompatibly changed from 2601 typedef int pem_password_cb(char *buf, int size, int rwflag); 2602 to 2603 ....(char *buf, int size, int rwflag, void *userdata); 2604 so that applications can pass data to their callbacks: 2605 The PEM[_ASN1]_{read,write}... functions and macros now take an 2606 additional void * argument, which is just handed through whenever 2607 the password callback is called. 2608 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller] 2609 2610 New function SSL_CTX_set_default_passwd_cb_userdata. 2611 2612 Compatibility note: As many C implementations push function arguments 2613 onto the stack in reverse order, the new library version is likely to 2614 interoperate with programs that have been compiled with the old 2615 pem_password_cb definition (PEM_whatever takes some data that 2616 happens to be on the stack as its last argument, and the callback 2617 just ignores this garbage); but there is no guarantee whatsoever that 2618 this will work. 2619 2620 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=... 2621 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused 2622 problems not only on Windows, but also on some Unix platforms. 2623 To avoid problematic command lines, these definitions are now in an 2624 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl 2625 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds). 2626 [Bodo Moeller] 2627 2628 *) MIPS III/IV assembler module is reimplemented. 2629 [Andy Polyakov] 2630 2631 *) More DES library cleanups: remove references to srand/rand and 2632 delete an unused file. 2633 [Ulf M�ller] 2634 2635 *) Add support for the the free Netwide assembler (NASM) under Win32, 2636 since not many people have MASM (ml) and it can be hard to obtain. 2637 This is currently experimental but it seems to work OK and pass all 2638 the tests. Check out INSTALL.W32 for info. 2639 [Steve Henson] 2640 2641 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections 2642 without temporary keys kept an extra copy of the server key, 2643 and connections with temporary keys did not free everything in case 2644 of an error. 2645 [Bodo Moeller] 2646 2647 *) New function RSA_check_key and new openssl rsa option -check 2648 for verifying the consistency of RSA keys. 2649 [Ulf Moeller, Bodo Moeller] 2650 2651 *) Various changes to make Win32 compile work: 2652 1. Casts to avoid "loss of data" warnings in p5_crpt2.c 2653 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned 2654 comparison" warnings. 2655 3. Add sk_<TYPE>_sort to DEF file generator and do make update. 2656 [Steve Henson] 2657 2658 *) Add a debugging option to PKCS#5 v2 key generation function: when 2659 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and 2660 derived keys are printed to stderr. 2661 [Steve Henson] 2662 2663 *) Copy the flags in ASN1_STRING_dup(). 2664 [Roman E. Pavlov <pre@mo.msk.ru>] 2665 2666 *) The x509 application mishandled signing requests containing DSA 2667 keys when the signing key was also DSA and the parameters didn't match. 2668 2669 It was supposed to omit the parameters when they matched the signing key: 2670 the verifying software was then supposed to automatically use the CA's 2671 parameters if they were absent from the end user certificate. 2672 2673 Omitting parameters is no longer recommended. The test was also 2674 the wrong way round! This was probably due to unusual behaviour in 2675 EVP_cmp_parameters() which returns 1 if the parameters match. 2676 This meant that parameters were omitted when they *didn't* match and 2677 the certificate was useless. Certificates signed with 'ca' didn't have 2678 this bug. 2679 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>] 2680 2681 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems. 2682 The interface is as follows: 2683 Applications can use 2684 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(), 2685 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop(); 2686 "off" is now the default. 2687 The library internally uses 2688 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(), 2689 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on() 2690 to disable memory-checking temporarily. 2691 2692 Some inconsistent states that previously were possible (and were 2693 even the default) are now avoided. 2694 2695 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time 2696 with each memory chunk allocated; this is occasionally more helpful 2697 than just having a counter. 2698 2699 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID. 2700 2701 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future 2702 extensions. 2703 [Bodo Moeller] 2704 2705 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX), 2706 which largely parallels "options", but is for changing API behaviour, 2707 whereas "options" are about protocol behaviour. 2708 Initial "mode" flags are: 2709 2710 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when 2711 a single record has been written. 2712 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write 2713 retries use the same buffer location. 2714 (But all of the contents must be 2715 copied!) 2716 [Bodo Moeller] 2717 2718 *) Bugfix: SSL_set_options ignored its parameter, only SSL_CTX_set_options 2719 worked. 2720 2721 *) Fix problems with no-hmac etc. 2722 [Ulf M�ller, pointed out by Brian Wellington <bwelling@tislabs.com>] 2723 2724 *) New functions RSA_get_default_method(), RSA_set_method() and 2725 RSA_get_method(). These allows replacement of RSA_METHODs without having 2726 to mess around with the internals of an RSA structure. 2727 [Steve Henson] 2728 2729 *) Fix memory leaks in DSA_do_sign and DSA_is_prime. 2730 Also really enable memory leak checks in openssl.c and in some 2731 test programs. 2732 [Chad C. Mulligan, Bodo Moeller] 2733 2734 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess 2735 up the length of negative integers. This has now been simplified to just 2736 store the length when it is first determined and use it later, rather 2737 than trying to keep track of where data is copied and updating it to 2738 point to the end. 2739 [Steve Henson, reported by Brien Wheeler 2740 <bwheeler@authentica-security.com>] 2741 2742 *) Add a new function PKCS7_signatureVerify. This allows the verification 2743 of a PKCS#7 signature but with the signing certificate passed to the 2744 function itself. This contrasts with PKCS7_dataVerify which assumes the 2745 certificate is present in the PKCS#7 structure. This isn't always the 2746 case: certificates can be omitted from a PKCS#7 structure and be 2747 distributed by "out of band" means (such as a certificate database). 2748 [Steve Henson] 2749 2750 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the 2751 function prototypes in pem.h, also change util/mkdef.pl to add the 2752 necessary function names. 2753 [Steve Henson] 2754 2755 *) mk1mf.pl (used by Windows builds) did not properly read the 2756 options set by Configure in the top level Makefile, and Configure 2757 was not even able to write more than one option correctly. 2758 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended. 2759 [Bodo Moeller] 2760 2761 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config 2762 file to be loaded from a BIO or FILE pointer. The BIO version will 2763 for example allow memory BIOs to contain config info. 2764 [Steve Henson] 2765 2766 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS. 2767 Whoever hopes to achieve shared-library compatibility across versions 2768 must use this, not the compile-time macro. 2769 (Exercise 0.9.4: Which is the minimum library version required by 2770 such programs?) 2771 Note: All this applies only to multi-threaded programs, others don't 2772 need locks. 2773 [Bodo Moeller] 2774 2775 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests 2776 through a BIO pair triggered the default case, i.e. 2777 SSLerr(...,SSL_R_UNKNOWN_STATE). 2778 [Bodo Moeller] 2779 2780 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications 2781 can use the SSL library even if none of the specific BIOs is 2782 appropriate. 2783 [Bodo Moeller] 2784 2785 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value 2786 for the encoded length. 2787 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>] 2788 2789 *) Add initial documentation of the X509V3 functions. 2790 [Steve Henson] 2791 2792 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and 2793 PEM_write_bio_PKCS8PrivateKey() that are equivalent to 2794 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more 2795 secure PKCS#8 private key format with a high iteration count. 2796 [Steve Henson] 2797 2798 *) Fix determination of Perl interpreter: A perl or perl5 2799 _directory_ in $PATH was also accepted as the interpreter. 2800 [Ralf S. Engelschall] 2801 2802 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking 2803 wrong with it but it was very old and did things like calling 2804 PEM_ASN1_read() directly and used MD5 for the hash not to mention some 2805 unusual formatting. 2806 [Steve Henson] 2807 2808 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed 2809 to use the new extension code. 2810 [Steve Henson] 2811 2812 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c 2813 with macros. This should make it easier to change their form, add extra 2814 arguments etc. Fix a few PEM prototypes which didn't have cipher as a 2815 constant. 2816 [Steve Henson] 2817 2818 *) Add to configuration table a new entry that can specify an alternative 2819 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep, 2820 according to Mark Crispin <MRC@Panda.COM>. 2821 [Bodo Moeller] 2822 2823#if 0 2824 *) DES CBC did not update the IV. Weird. 2825 [Ben Laurie] 2826#else 2827 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does. 2828 Changing the behaviour of the former might break existing programs -- 2829 where IV updating is needed, des_ncbc_encrypt can be used. 2830#endif 2831 2832 *) When bntest is run from "make test" it drives bc to check its 2833 calculations, as well as internally checking them. If an internal check 2834 fails, it needs to cause bc to give a non-zero result or make test carries 2835 on without noticing the failure. Fixed. 2836 [Ben Laurie] 2837 2838 *) DES library cleanups. 2839 [Ulf M�ller] 2840 2841 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be 2842 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit 2843 ciphers. NOTE: although the key derivation function has been verified 2844 against some published test vectors it has not been extensively tested 2845 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use 2846 of v2.0. 2847 [Steve Henson] 2848 2849 *) Instead of "mkdir -p", which is not fully portable, use new 2850 Perl script "util/mkdir-p.pl". 2851 [Bodo Moeller] 2852 2853 *) Rewrite the way password based encryption (PBE) is handled. It used to 2854 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter 2855 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms 2856 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now 2857 the 'parameter' field of the AlgorithmIdentifier is passed to the 2858 underlying key generation function so it must do its own ASN1 parsing. 2859 This has also changed the EVP_PBE_CipherInit() function which now has a 2860 'parameter' argument instead of literal salt and iteration count values 2861 and the function EVP_PBE_ALGOR_CipherInit() has been deleted. 2862 [Steve Henson] 2863 2864 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms 2865 and PKCS#8 functionality. New 'pkcs8' application linked to openssl. 2866 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE 2867 KEY" because this clashed with PKCS#8 unencrypted string. Since this 2868 value was just used as a "magic string" and not used directly its 2869 value doesn't matter. 2870 [Steve Henson] 2871 2872 *) Introduce some semblance of const correctness to BN. Shame C doesn't 2873 support mutable. 2874 [Ben Laurie] 2875 2876 *) "linux-sparc64" configuration (ultrapenguin). 2877 [Ray Miller <ray.miller@oucs.ox.ac.uk>] 2878 "linux-sparc" configuration. 2879 [Christian Forster <fo@hawo.stw.uni-erlangen.de>] 2880 2881 *) config now generates no-xxx options for missing ciphers. 2882 [Ulf M�ller] 2883 2884 *) Support the EBCDIC character set (work in progress). 2885 File ebcdic.c not yet included because it has a different license. 2886 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>] 2887 2888 *) Support BS2000/OSD-POSIX. 2889 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>] 2890 2891 *) Make callbacks for key generation use void * instead of char *. 2892 [Ben Laurie] 2893 2894 *) Make S/MIME samples compile (not yet tested). 2895 [Ben Laurie] 2896 2897 *) Additional typesafe stacks. 2898 [Ben Laurie] 2899 2900 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x). 2901 [Bodo Moeller] 2902 2903 2904 Changes between 0.9.3 and 0.9.3a [29 May 1999] 2905 2906 *) New configuration variant "sco5-gcc". 2907 2908 *) Updated some demos. 2909 [Sean O Riordain, Wade Scholine] 2910 2911 *) Add missing BIO_free at exit of pkcs12 application. 2912 [Wu Zhigang] 2913 2914 *) Fix memory leak in conf.c. 2915 [Steve Henson] 2916 2917 *) Updates for Win32 to assembler version of MD5. 2918 [Steve Henson] 2919 2920 *) Set #! path to perl in apps/der_chop to where we found it 2921 instead of using a fixed path. 2922 [Bodo Moeller] 2923 2924 *) SHA library changes for irix64-mips4-cc. 2925 [Andy Polyakov] 2926 2927 *) Improvements for VMS support. 2928 [Richard Levitte] 2929 2930 2931 Changes between 0.9.2b and 0.9.3 [24 May 1999] 2932 2933 *) Bignum library bug fix. IRIX 6 passes "make test" now! 2934 This also avoids the problems with SC4.2 and unpatched SC5. 2935 [Andy Polyakov <appro@fy.chalmers.se>] 2936 2937 *) New functions sk_num, sk_value and sk_set to replace the previous macros. 2938 These are required because of the typesafe stack would otherwise break 2939 existing code. If old code used a structure member which used to be STACK 2940 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with 2941 sk_num or sk_value it would produce an error because the num, data members 2942 are not present in STACK_OF. Now it just produces a warning. sk_set 2943 replaces the old method of assigning a value to sk_value 2944 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code 2945 that does this will no longer work (and should use sk_set instead) but 2946 this could be regarded as a "questionable" behaviour anyway. 2947 [Steve Henson] 2948 2949 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now 2950 correctly handle encrypted S/MIME data. 2951 [Steve Henson] 2952 2953 *) Change type of various DES function arguments from des_cblock 2954 (which means, in function argument declarations, pointer to char) 2955 to des_cblock * (meaning pointer to array with 8 char elements), 2956 which allows the compiler to do more typechecking; it was like 2957 that back in SSLeay, but with lots of ugly casts. 2958 2959 Introduce new type const_des_cblock. 2960 [Bodo Moeller] 2961 2962 *) Reorganise the PKCS#7 library and get rid of some of the more obvious 2963 problems: find RecipientInfo structure that matches recipient certificate 2964 and initialise the ASN1 structures properly based on passed cipher. 2965 [Steve Henson] 2966 2967 *) Belatedly make the BN tests actually check the results. 2968 [Ben Laurie] 2969 2970 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion 2971 to and from BNs: it was completely broken. New compilation option 2972 NEG_PUBKEY_BUG to allow for some broken certificates that encode public 2973 key elements as negative integers. 2974 [Steve Henson] 2975 2976 *) Reorganize and speed up MD5. 2977 [Andy Polyakov <appro@fy.chalmers.se>] 2978 2979 *) VMS support. 2980 [Richard Levitte <richard@levitte.org>] 2981 2982 *) New option -out to asn1parse to allow the parsed structure to be 2983 output to a file. This is most useful when combined with the -strparse 2984 option to examine the output of things like OCTET STRINGS. 2985 [Steve Henson] 2986 2987 *) Make SSL library a little more fool-proof by not requiring any longer 2988 that SSL_set_{accept,connect}_state be called before 2989 SSL_{accept,connect} may be used (SSL_set_..._state is omitted 2990 in many applications because usually everything *appeared* to work as 2991 intended anyway -- now it really works as intended). 2992 [Bodo Moeller] 2993 2994 *) Move openssl.cnf out of lib/. 2995 [Ulf M�ller] 2996 2997 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall 2998 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes 2999 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+ 3000 [Ralf S. Engelschall] 3001 3002 *) Various fixes to the EVP and PKCS#7 code. It may now be able to 3003 handle PKCS#7 enveloped data properly. 3004 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve] 3005 3006 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of 3007 copying pointers. The cert_st handling is changed by this in 3008 various ways (and thus what used to be known as ctx->default_cert 3009 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert 3010 any longer when s->cert does not give us what we need). 3011 ssl_cert_instantiate becomes obsolete by this change. 3012 As soon as we've got the new code right (possibly it already is?), 3013 we have solved a couple of bugs of the earlier code where s->cert 3014 was used as if it could not have been shared with other SSL structures. 3015 3016 Note that using the SSL API in certain dirty ways now will result 3017 in different behaviour than observed with earlier library versions: 3018 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx) 3019 does not influence s as it used to. 3020 3021 In order to clean up things more thoroughly, inside SSL_SESSION 3022 we don't use CERT any longer, but a new structure SESS_CERT 3023 that holds per-session data (if available); currently, this is 3024 the peer's certificate chain and, for clients, the server's certificate 3025 and temporary key. CERT holds only those values that can have 3026 meaningful defaults in an SSL_CTX. 3027 [Bodo Moeller] 3028 3029 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure 3030 from the internal representation. Various PKCS#7 fixes: remove some 3031 evil casts and set the enc_dig_alg field properly based on the signing 3032 key type. 3033 [Steve Henson] 3034 3035 *) Allow PKCS#12 password to be set from the command line or the 3036 environment. Let 'ca' get its config file name from the environment 3037 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req' 3038 and 'x509'). 3039 [Steve Henson] 3040 3041 *) Allow certificate policies extension to use an IA5STRING for the 3042 organization field. This is contrary to the PKIX definition but 3043 VeriSign uses it and IE5 only recognises this form. Document 'x509' 3044 extension option. 3045 [Steve Henson] 3046 3047 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic, 3048 without disallowing inline assembler and the like for non-pedantic builds. 3049 [Ben Laurie] 3050 3051 *) Support Borland C++ builder. 3052 [Janez Jere <jj@void.si>, modified by Ulf M�ller] 3053 3054 *) Support Mingw32. 3055 [Ulf M�ller] 3056 3057 *) SHA-1 cleanups and performance enhancements. 3058 [Andy Polyakov <appro@fy.chalmers.se>] 3059 3060 *) Sparc v8plus assembler for the bignum library. 3061 [Andy Polyakov <appro@fy.chalmers.se>] 3062 3063 *) Accept any -xxx and +xxx compiler options in Configure. 3064 [Ulf M�ller] 3065 3066 *) Update HPUX configuration. 3067 [Anonymous] 3068 3069 *) Add missing sk_<type>_unshift() function to safestack.h 3070 [Ralf S. Engelschall] 3071 3072 *) New function SSL_CTX_use_certificate_chain_file that sets the 3073 "extra_cert"s in addition to the certificate. (This makes sense 3074 only for "PEM" format files, as chains as a whole are not 3075 DER-encoded.) 3076 [Bodo Moeller] 3077 3078 *) Support verify_depth from the SSL API. 3079 x509_vfy.c had what can be considered an off-by-one-error: 3080 Its depth (which was not part of the external interface) 3081 was actually counting the number of certificates in a chain; 3082 now it really counts the depth. 3083 [Bodo Moeller] 3084 3085 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used 3086 instead of X509err, which often resulted in confusing error 3087 messages since the error codes are not globally unique 3088 (e.g. an alleged error in ssl3_accept when a certificate 3089 didn't match the private key). 3090 3091 *) New function SSL_CTX_set_session_id_context that allows to set a default 3092 value (so that you don't need SSL_set_session_id_context for each 3093 connection using the SSL_CTX). 3094 [Bodo Moeller] 3095 3096 *) OAEP decoding bug fix. 3097 [Ulf M�ller] 3098 3099 *) Support INSTALL_PREFIX for package builders, as proposed by 3100 David Harris. 3101 [Bodo Moeller] 3102 3103 *) New Configure options "threads" and "no-threads". For systems 3104 where the proper compiler options are known (currently Solaris 3105 and Linux), "threads" is the default. 3106 [Bodo Moeller] 3107 3108 *) New script util/mklink.pl as a faster substitute for util/mklink.sh. 3109 [Bodo Moeller] 3110 3111 *) Install various scripts to $(OPENSSLDIR)/misc, not to 3112 $(INSTALLTOP)/bin -- they shouldn't clutter directories 3113 such as /usr/local/bin. 3114 [Bodo Moeller] 3115 3116 *) "make linux-shared" to build shared libraries. 3117 [Niels Poppe <niels@netbox.org>] 3118 3119 *) New Configure option no-<cipher> (rsa, idea, rc5, ...). 3120 [Ulf M�ller] 3121 3122 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for 3123 extension adding in x509 utility. 3124 [Steve Henson] 3125 3126 *) Remove NOPROTO sections and error code comments. 3127 [Ulf M�ller] 3128 3129 *) Partial rewrite of the DEF file generator to now parse the ANSI 3130 prototypes. 3131 [Steve Henson] 3132 3133 *) New Configure options --prefix=DIR and --openssldir=DIR. 3134 [Ulf M�ller] 3135 3136 *) Complete rewrite of the error code script(s). It is all now handled 3137 by one script at the top level which handles error code gathering, 3138 header rewriting and C source file generation. It should be much better 3139 than the old method: it now uses a modified version of Ulf's parser to 3140 read the ANSI prototypes in all header files (thus the old K&R definitions 3141 aren't needed for error creation any more) and do a better job of 3142 translating function codes into names. The old 'ASN1 error code imbedded 3143 in a comment' is no longer necessary and it doesn't use .err files which 3144 have now been deleted. Also the error code call doesn't have to appear all 3145 on one line (which resulted in some large lines...). 3146 [Steve Henson] 3147 3148 *) Change #include filenames from <foo.h> to <openssl/foo.h>. 3149 [Bodo Moeller] 3150 3151 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return 3152 0 (which usually indicates a closed connection), but continue reading. 3153 [Bodo Moeller] 3154 3155 *) Fix some race conditions. 3156 [Bodo Moeller] 3157 3158 *) Add support for CRL distribution points extension. Add Certificate 3159 Policies and CRL distribution points documentation. 3160 [Steve Henson] 3161 3162 *) Move the autogenerated header file parts to crypto/opensslconf.h. 3163 [Ulf M�ller] 3164 3165 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of 3166 8 of keying material. Merlin has also confirmed interop with this fix 3167 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0. 3168 [Merlin Hughes <merlin@baltimore.ie>] 3169 3170 *) Fix lots of warnings. 3171 [Richard Levitte <levitte@stacken.kth.se>] 3172 3173 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if 3174 the directory spec didn't end with a LIST_SEPARATOR_CHAR. 3175 [Richard Levitte <levitte@stacken.kth.se>] 3176 3177 *) Fix problems with sizeof(long) == 8. 3178 [Andy Polyakov <appro@fy.chalmers.se>] 3179 3180 *) Change functions to ANSI C. 3181 [Ulf M�ller] 3182 3183 *) Fix typos in error codes. 3184 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf M�ller] 3185 3186 *) Remove defunct assembler files from Configure. 3187 [Ulf M�ller] 3188 3189 *) SPARC v8 assembler BIGNUM implementation. 3190 [Andy Polyakov <appro@fy.chalmers.se>] 3191 3192 *) Support for Certificate Policies extension: both print and set. 3193 Various additions to support the r2i method this uses. 3194 [Steve Henson] 3195 3196 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could 3197 return a const string when you are expecting an allocated buffer. 3198 [Ben Laurie] 3199 3200 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE 3201 types DirectoryString and DisplayText. 3202 [Steve Henson] 3203 3204 *) Add code to allow r2i extensions to access the configuration database, 3205 add an LHASH database driver and add several ctx helper functions. 3206 [Steve Henson] 3207 3208 *) Fix an evil bug in bn_expand2() which caused various BN functions to 3209 fail when they extended the size of a BIGNUM. 3210 [Steve Henson] 3211 3212 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to 3213 support typesafe stack. 3214 [Steve Henson] 3215 3216 *) Fix typo in SSL_[gs]et_options(). 3217 [Nils Frostberg <nils@medcom.se>] 3218 3219 *) Delete various functions and files that belonged to the (now obsolete) 3220 old X509V3 handling code. 3221 [Steve Henson] 3222 3223 *) New Configure option "rsaref". 3224 [Ulf M�ller] 3225 3226 *) Don't auto-generate pem.h. 3227 [Bodo Moeller] 3228 3229 *) Introduce type-safe ASN.1 SETs. 3230 [Ben Laurie] 3231 3232 *) Convert various additional casted stacks to type-safe STACK_OF() variants. 3233 [Ben Laurie, Ralf S. Engelschall, Steve Henson] 3234 3235 *) Introduce type-safe STACKs. This will almost certainly break lots of code 3236 that links with OpenSSL (well at least cause lots of warnings), but fear 3237 not: the conversion is trivial, and it eliminates loads of evil casts. A 3238 few STACKed things have been converted already. Feel free to convert more. 3239 In the fullness of time, I'll do away with the STACK type altogether. 3240 [Ben Laurie] 3241 3242 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate 3243 specified in <certfile> by updating the entry in the index.txt file. 3244 This way one no longer has to edit the index.txt file manually for 3245 revoking a certificate. The -revoke option does the gory details now. 3246 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall] 3247 3248 *) Fix `openssl crl -noout -text' combination where `-noout' killed the 3249 `-text' option at all and this way the `-noout -text' combination was 3250 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'. 3251 [Ralf S. Engelschall] 3252 3253 *) Make sure a corresponding plain text error message exists for the 3254 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a 3255 verify callback function determined that a certificate was revoked. 3256 [Ralf S. Engelschall] 3257 3258 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for 3259 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test 3260 all available cipers including rc5, which was forgotten until now. 3261 In order to let the testing shell script know which algorithms 3262 are available, a new (up to now undocumented) command 3263 "openssl list-cipher-commands" is used. 3264 [Bodo Moeller] 3265 3266 *) Bugfix: s_client occasionally would sleep in select() when 3267 it should have checked SSL_pending() first. 3268 [Bodo Moeller] 3269 3270 *) New functions DSA_do_sign and DSA_do_verify to provide access to 3271 the raw DSA values prior to ASN.1 encoding. 3272 [Ulf M�ller] 3273 3274 *) Tweaks to Configure 3275 [Niels Poppe <niels@netbox.org>] 3276 3277 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support, 3278 yet... 3279 [Steve Henson] 3280 3281 *) New variables $(RANLIB) and $(PERL) in the Makefiles. 3282 [Ulf M�ller] 3283 3284 *) New config option to avoid instructions that are illegal on the 80386. 3285 The default code is faster, but requires at least a 486. 3286 [Ulf M�ller] 3287 3288 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and 3289 SSL2_SERVER_VERSION (not used at all) macros, which are now the 3290 same as SSL2_VERSION anyway. 3291 [Bodo Moeller] 3292 3293 *) New "-showcerts" option for s_client. 3294 [Bodo Moeller] 3295 3296 *) Still more PKCS#12 integration. Add pkcs12 application to openssl 3297 application. Various cleanups and fixes. 3298 [Steve Henson] 3299 3300 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and 3301 modify error routines to work internally. Add error codes and PBE init 3302 to library startup routines. 3303 [Steve Henson] 3304 3305 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and 3306 packing functions to asn1 and evp. Changed function names and error 3307 codes along the way. 3308 [Steve Henson] 3309 3310 *) PKCS12 integration: and so it begins... First of several patches to 3311 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12 3312 objects to objects.h 3313 [Steve Henson] 3314 3315 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1 3316 and display support for Thawte strong extranet extension. 3317 [Steve Henson] 3318 3319 *) Add LinuxPPC support. 3320 [Jeff Dubrule <igor@pobox.org>] 3321 3322 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to 3323 bn_div_words in alpha.s. 3324 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie] 3325 3326 *) Make sure the RSA OAEP test is skipped under -DRSAref because 3327 OAEP isn't supported when OpenSSL is built with RSAref. 3328 [Ulf Moeller <ulf@fitug.de>] 3329 3330 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h 3331 so they no longer are missing under -DNOPROTO. 3332 [Soren S. Jorvang <soren@t.dk>] 3333 3334 3335 Changes between 0.9.1c and 0.9.2b [22 Mar 1999] 3336 3337 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still 3338 doesn't work when the session is reused. Coming soon! 3339 [Ben Laurie] 3340 3341 *) Fix a security hole, that allows sessions to be reused in the wrong 3342 context thus bypassing client cert protection! All software that uses 3343 client certs and session caches in multiple contexts NEEDS PATCHING to 3344 allow session reuse! A fuller solution is in the works. 3345 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)] 3346 3347 *) Some more source tree cleanups (removed obsolete files 3348 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed 3349 permission on "config" script to be executable) and a fix for the INSTALL 3350 document. 3351 [Ulf Moeller <ulf@fitug.de>] 3352 3353 *) Remove some legacy and erroneous uses of malloc, free instead of 3354 Malloc, Free. 3355 [Lennart Bang <lob@netstream.se>, with minor changes by Steve] 3356 3357 *) Make rsa_oaep_test return non-zero on error. 3358 [Ulf Moeller <ulf@fitug.de>] 3359 3360 *) Add support for native Solaris shared libraries. Configure 3361 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice 3362 if someone would make that last step automatic. 3363 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>] 3364 3365 *) ctx_size was not built with the right compiler during "make links". Fixed. 3366 [Ben Laurie] 3367 3368 *) Change the meaning of 'ALL' in the cipher list. It now means "everything 3369 except NULL ciphers". This means the default cipher list will no longer 3370 enable NULL ciphers. They need to be specifically enabled e.g. with 3371 the string "DEFAULT:eNULL". 3372 [Steve Henson] 3373 3374 *) Fix to RSA private encryption routines: if p < q then it would 3375 occasionally produce an invalid result. This will only happen with 3376 externally generated keys because OpenSSL (and SSLeay) ensure p > q. 3377 [Steve Henson] 3378 3379 *) Be less restrictive and allow also `perl util/perlpath.pl 3380 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin', 3381 because this way one can also use an interpreter named `perl5' (which is 3382 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still 3383 installed as `perl'). 3384 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>] 3385 3386 *) Let util/clean-depend.pl work also with older Perl 5.00x versions. 3387 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>] 3388 3389 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add 3390 advapi32.lib to Win32 build and change the pem test comparision 3391 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the 3392 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h 3393 and crypto/des/ede_cbcm_enc.c. 3394 [Steve Henson] 3395 3396 *) DES quad checksum was broken on big-endian architectures. Fixed. 3397 [Ben Laurie] 3398 3399 *) Comment out two functions in bio.h that aren't implemented. Fix up the 3400 Win32 test batch file so it (might) work again. The Win32 test batch file 3401 is horrible: I feel ill.... 3402 [Steve Henson] 3403 3404 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected 3405 in e_os.h. Audit of header files to check ANSI and non ANSI 3406 sections: 10 functions were absent from non ANSI section and not exported 3407 from Windows DLLs. Fixed up libeay.num for new functions. 3408 [Steve Henson] 3409 3410 *) Make `openssl version' output lines consistent. 3411 [Ralf S. Engelschall] 3412 3413 *) Fix Win32 symbol export lists for BIO functions: Added 3414 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data 3415 to ms/libeay{16,32}.def. 3416 [Ralf S. Engelschall] 3417 3418 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled 3419 fine under Unix and passes some trivial tests I've now added. But the 3420 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was 3421 added to make sure no one expects that this stuff really works in the 3422 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources 3423 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and 3424 openssl_bio.xs. 3425 [Ralf S. Engelschall] 3426 3427 *) Fix the generation of two part addresses in perl. 3428 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie] 3429 3430 *) Add config entry for Linux on MIPS. 3431 [John Tobey <jtobey@channel1.com>] 3432 3433 *) Make links whenever Configure is run, unless we are on Windoze. 3434 [Ben Laurie] 3435 3436 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf. 3437 Currently only issuerAltName and AuthorityKeyIdentifier make any sense 3438 in CRLs. 3439 [Steve Henson] 3440 3441 *) Add a useful kludge to allow package maintainers to specify compiler and 3442 other platforms details on the command line without having to patch the 3443 Configure script everytime: One now can use ``perl Configure 3444 <id>:<details>'', i.e. platform ids are allowed to have details appended 3445 to them (seperated by colons). This is treated as there would be a static 3446 pre-configured entry in Configure's %table under key <id> with value 3447 <details> and ``perl Configure <id>'' is called. So, when you want to 3448 perform a quick test-compile under FreeBSD 3.1 with pgcc and without 3449 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"'' 3450 now, which overrides the FreeBSD-elf entry on-the-fly. 3451 [Ralf S. Engelschall] 3452 3453 *) Disable new TLS1 ciphersuites by default: they aren't official yet. 3454 [Ben Laurie] 3455 3456 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified 3457 on the `perl Configure ...' command line. This way one can compile 3458 OpenSSL libraries with Position Independent Code (PIC) which is needed 3459 for linking it into DSOs. 3460 [Ralf S. Engelschall] 3461 3462 *) Remarkably, export ciphers were totally broken and no-one had noticed! 3463 Fixed. 3464 [Ben Laurie] 3465 3466 *) Cleaned up the LICENSE document: The official contact for any license 3467 questions now is the OpenSSL core team under openssl-core@openssl.org. 3468 And add a paragraph about the dual-license situation to make sure people 3469 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply 3470 to the OpenSSL toolkit. 3471 [Ralf S. Engelschall] 3472 3473 *) General source tree makefile cleanups: Made `making xxx in yyy...' 3474 display consistent in the source tree and replaced `/bin/rm' by `rm'. 3475 Additonally cleaned up the `make links' target: Remove unnecessary 3476 semicolons, subsequent redundant removes, inline point.sh into mklink.sh 3477 to speed processing and no longer clutter the display with confusing 3478 stuff. Instead only the actually done links are displayed. 3479 [Ralf S. Engelschall] 3480 3481 *) Permit null encryption ciphersuites, used for authentication only. It used 3482 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this. 3483 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null 3484 encryption. 3485 [Ben Laurie] 3486 3487 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder 3488 signed attributes when verifying signatures (this would break them), 3489 the detached data encoding was wrong and public keys obtained using 3490 X509_get_pubkey() weren't freed. 3491 [Steve Henson] 3492 3493 *) Add text documentation for the BUFFER functions. Also added a work around 3494 to a Win95 console bug. This was triggered by the password read stuff: the 3495 last character typed gets carried over to the next fread(). If you were 3496 generating a new cert request using 'req' for example then the last 3497 character of the passphrase would be CR which would then enter the first 3498 field as blank. 3499 [Steve Henson] 3500 3501 *) Added the new `Includes OpenSSL Cryptography Software' button as 3502 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay 3503 button and can be used by applications based on OpenSSL to show the 3504 relationship to the OpenSSL project. 3505 [Ralf S. Engelschall] 3506 3507 *) Remove confusing variables in function signatures in files 3508 ssl/ssl_lib.c and ssl/ssl.h. 3509 [Lennart Bong <lob@kulthea.stacken.kth.se>] 3510 3511 *) Don't install bss_file.c under PREFIX/include/ 3512 [Lennart Bong <lob@kulthea.stacken.kth.se>] 3513 3514 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle 3515 functions that return function pointers and has support for NT specific 3516 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various 3517 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from 3518 unsigned to signed types: this was killing the Win32 compile. 3519 [Steve Henson] 3520 3521 *) Add new certificate file to stack functions, 3522 SSL_add_dir_cert_subjects_to_stack() and 3523 SSL_add_file_cert_subjects_to_stack(). These largely supplant 3524 SSL_load_client_CA_file(), and can be used to add multiple certs easily 3525 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()). 3526 This means that Apache-SSL and similar packages don't have to mess around 3527 to add as many CAs as they want to the preferred list. 3528 [Ben Laurie] 3529 3530 *) Experiment with doxygen documentation. Currently only partially applied to 3531 ssl/ssl_lib.c. 3532 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with 3533 openssl.doxy as the configuration file. 3534 [Ben Laurie] 3535 3536 *) Get rid of remaining C++-style comments which strict C compilers hate. 3537 [Ralf S. Engelschall, pointed out by Carlos Amengual] 3538 3539 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not 3540 compiled in by default: it has problems with large keys. 3541 [Steve Henson] 3542 3543 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and 3544 DH private keys and/or callback functions which directly correspond to 3545 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This 3546 is needed for applications which have to configure certificates on a 3547 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis 3548 (e.g. s_server). 3549 For the RSA certificate situation is makes no difference, but 3550 for the DSA certificate situation this fixes the "no shared cipher" 3551 problem where the OpenSSL cipher selection procedure failed because the 3552 temporary keys were not overtaken from the context and the API provided 3553 no way to reconfigure them. 3554 The new functions now let applications reconfigure the stuff and they 3555 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh, 3556 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new 3557 non-public-API function ssl_cert_instantiate() is used as a helper 3558 function and also to reduce code redundancy inside ssl_rsa.c. 3559 [Ralf S. Engelschall] 3560 3561 *) Move s_server -dcert and -dkey options out of the undocumented feature 3562 area because they are useful for the DSA situation and should be 3563 recognized by the users. 3564 [Ralf S. Engelschall] 3565 3566 *) Fix the cipher decision scheme for export ciphers: the export bits are 3567 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within 3568 SSL_EXP_MASK. So, the original variable has to be used instead of the 3569 already masked variable. 3570 [Richard Levitte <levitte@stacken.kth.se>] 3571 3572 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c 3573 [Richard Levitte <levitte@stacken.kth.se>] 3574 3575 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal() 3576 from `int' to `unsigned int' because it's a length and initialized by 3577 EVP_DigestFinal() which expects an `unsigned int *'. 3578 [Richard Levitte <levitte@stacken.kth.se>] 3579 3580 *) Don't hard-code path to Perl interpreter on shebang line of Configure 3581 script. Instead use the usual Shell->Perl transition trick. 3582 [Ralf S. Engelschall] 3583 3584 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates 3585 (in addition to RSA certificates) to match the behaviour of `openssl dsa 3586 -noout -modulus' as it's already the case for `openssl rsa -noout 3587 -modulus'. For RSA the -modulus is the real "modulus" while for DSA 3588 currently the public key is printed (a decision which was already done by 3589 `openssl dsa -modulus' in the past) which serves a similar purpose. 3590 Additionally the NO_RSA no longer completely removes the whole -modulus 3591 option; it now only avoids using the RSA stuff. Same applies to NO_DSA 3592 now, too. 3593 [Ralf S. Engelschall] 3594 3595 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested 3596 BIO. See the source (crypto/evp/bio_ok.c) for more info. 3597 [Arne Ansper <arne@ats.cyber.ee>] 3598 3599 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs 3600 to be added. Now both 'req' and 'ca' can use new objects defined in the 3601 config file. 3602 [Steve Henson] 3603 3604 *) Add cool BIO that does syslog (or event log on NT). 3605 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie] 3606 3607 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5, 3608 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and 3609 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher 3610 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt. 3611 [Ben Laurie] 3612 3613 *) Add preliminary config info for new extension code. 3614 [Steve Henson] 3615 3616 *) Make RSA_NO_PADDING really use no padding. 3617 [Ulf Moeller <ulf@fitug.de>] 3618 3619 *) Generate errors when private/public key check is done. 3620 [Ben Laurie] 3621 3622 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support 3623 for some CRL extensions and new objects added. 3624 [Steve Henson] 3625 3626 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private 3627 key usage extension and fuller support for authority key id. 3628 [Steve Henson] 3629 3630 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved 3631 padding method for RSA, which is recommended for new applications in PKCS 3632 #1 v2.0 (RFC 2437, October 1998). 3633 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical 3634 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure 3635 against Bleichbacher's attack on RSA. 3636 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by 3637 Ben Laurie] 3638 3639 *) Updates to the new SSL compression code 3640 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)] 3641 3642 *) Fix so that the version number in the master secret, when passed 3643 via RSA, checks that if TLS was proposed, but we roll back to SSLv3 3644 (because the server will not accept higher), that the version number 3645 is 0x03,0x01, not 0x03,0x00 3646 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)] 3647 3648 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory 3649 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes 3650 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c 3651 [Steve Henson] 3652 3653 *) Support for RAW extensions where an arbitrary extension can be 3654 created by including its DER encoding. See apps/openssl.cnf for 3655 an example. 3656 [Steve Henson] 3657 3658 *) Make sure latest Perl versions don't interpret some generated C array 3659 code as Perl array code in the crypto/err/err_genc.pl script. 3660 [Lars Weber <3weber@informatik.uni-hamburg.de>] 3661 3662 *) Modify ms/do_ms.bat to not generate assembly language makefiles since 3663 not many people have the assembler. Various Win32 compilation fixes and 3664 update to the INSTALL.W32 file with (hopefully) more accurate Win32 3665 build instructions. 3666 [Steve Henson] 3667 3668 *) Modify configure script 'Configure' to automatically create crypto/date.h 3669 file under Win32 and also build pem.h from pem.org. New script 3670 util/mkfiles.pl to create the MINFO file on environments that can't do a 3671 'make files': perl util/mkfiles.pl >MINFO should work. 3672 [Steve Henson] 3673 3674 *) Major rework of DES function declarations, in the pursuit of correctness 3675 and purity. As a result, many evil casts evaporated, and some weirdness, 3676 too. You may find this causes warnings in your code. Zapping your evil 3677 casts will probably fix them. Mostly. 3678 [Ben Laurie] 3679 3680 *) Fix for a typo in asn1.h. Bug fix to object creation script 3681 obj_dat.pl. It considered a zero in an object definition to mean 3682 "end of object": none of the objects in objects.h have any zeros 3683 so it wasn't spotted. 3684 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>] 3685 3686 *) Add support for Triple DES Cipher Block Chaining with Output Feedback 3687 Masking (CBCM). In the absence of test vectors, the best I have been able 3688 to do is check that the decrypt undoes the encrypt, so far. Send me test 3689 vectors if you have them. 3690 [Ben Laurie] 3691 3692 *) Correct calculation of key length for export ciphers (too much space was 3693 allocated for null ciphers). This has not been tested! 3694 [Ben Laurie] 3695 3696 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage 3697 message is now correct (it understands "crypto" and "ssl" on its 3698 command line). There is also now an "update" option. This will update 3699 the util/ssleay.num and util/libeay.num files with any new functions. 3700 If you do a: 3701 perl util/mkdef.pl crypto ssl update 3702 it will update them. 3703 [Steve Henson] 3704 3705 *) Overhauled the Perl interface (perl/*): 3706 - ported BN stuff to OpenSSL's different BN library 3707 - made the perl/ source tree CVS-aware 3708 - renamed the package from SSLeay to OpenSSL (the files still contain 3709 their history because I've copied them in the repository) 3710 - removed obsolete files (the test scripts will be replaced 3711 by better Test::Harness variants in the future) 3712 [Ralf S. Engelschall] 3713 3714 *) First cut for a very conservative source tree cleanup: 3715 1. merge various obsolete readme texts into doc/ssleay.txt 3716 where we collect the old documents and readme texts. 3717 2. remove the first part of files where I'm already sure that we no 3718 longer need them because of three reasons: either they are just temporary 3719 files which were left by Eric or they are preserved original files where 3720 I've verified that the diff is also available in the CVS via "cvs diff 3721 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for 3722 the crypto/md/ stuff). 3723 [Ralf S. Engelschall] 3724 3725 *) More extension code. Incomplete support for subject and issuer alt 3726 name, issuer and authority key id. Change the i2v function parameters 3727 and add an extra 'crl' parameter in the X509V3_CTX structure: guess 3728 what that's for :-) Fix to ASN1 macro which messed up 3729 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED. 3730 [Steve Henson] 3731 3732 *) Preliminary support for ENUMERATED type. This is largely copied from the 3733 INTEGER code. 3734 [Steve Henson] 3735 3736 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy. 3737 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)] 3738 3739 *) Make sure `make rehash' target really finds the `openssl' program. 3740 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>] 3741 3742 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd 3743 like to hear about it if this slows down other processors. 3744 [Ben Laurie] 3745 3746 *) Add CygWin32 platform information to Configure script. 3747 [Alan Batie <batie@aahz.jf.intel.com>] 3748 3749 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm' 3750 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>] 3751 3752 *) New program nseq to manipulate netscape certificate sequences 3753 [Steve Henson] 3754 3755 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a 3756 few typos. 3757 [Steve Henson] 3758 3759 *) Fixes to BN code. Previously the default was to define BN_RECURSION 3760 but the BN code had some problems that would cause failures when 3761 doing certificate verification and some other functions. 3762 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)] 3763 3764 *) Add ASN1 and PEM code to support netscape certificate sequences. 3765 [Steve Henson] 3766 3767 *) Add ASN1 and PEM code to support netscape certificate sequences. 3768 [Steve Henson] 3769 3770 *) Add several PKIX and private extended key usage OIDs. 3771 [Steve Henson] 3772 3773 *) Modify the 'ca' program to handle the new extension code. Modify 3774 openssl.cnf for new extension format, add comments. 3775 [Steve Henson] 3776 3777 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req' 3778 and add a sample to openssl.cnf so req -x509 now adds appropriate 3779 CA extensions. 3780 [Steve Henson] 3781 3782 *) Continued X509 V3 changes. Add to other makefiles, integrate with the 3783 error code, add initial support to X509_print() and x509 application. 3784 [Steve Henson] 3785 3786 *) Takes a deep breath and start addding X509 V3 extension support code. Add 3787 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this 3788 stuff is currently isolated and isn't even compiled yet. 3789 [Steve Henson] 3790 3791 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL 3792 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print. 3793 Removed the versions check from X509 routines when loading extensions: 3794 this allows certain broken certificates that don't set the version 3795 properly to be processed. 3796 [Steve Henson] 3797 3798 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another 3799 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which 3800 can still be regenerated with "make depend". 3801 [Ben Laurie] 3802 3803 *) Spelling mistake in C version of CAST-128. 3804 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>] 3805 3806 *) Changes to the error generation code. The perl script err-code.pl 3807 now reads in the old error codes and retains the old numbers, only 3808 adding new ones if necessary. It also only changes the .err files if new 3809 codes are added. The makefiles have been modified to only insert errors 3810 when needed (to avoid needlessly modifying header files). This is done 3811 by only inserting errors if the .err file is newer than the auto generated 3812 C file. To rebuild all the error codes from scratch (the old behaviour) 3813 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl 3814 or delete all the .err files. 3815 [Steve Henson] 3816 3817 *) CAST-128 was incorrectly implemented for short keys. The C version has 3818 been fixed, but is untested. The assembler versions are also fixed, but 3819 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing 3820 to regenerate it if needed. 3821 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun 3822 Hagino <itojun@kame.net>] 3823 3824 *) File was opened incorrectly in randfile.c. 3825 [Ulf M�ller <ulf@fitug.de>] 3826 3827 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print 3828 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or 3829 GeneralizedTime. ASN1_TIME is the proper type used in certificates et 3830 al: it's just almost always a UTCTime. Note this patch adds new error 3831 codes so do a "make errors" if there are problems. 3832 [Steve Henson] 3833 3834 *) Correct Linux 1 recognition in config. 3835 [Ulf M�ller <ulf@fitug.de>] 3836 3837 *) Remove pointless MD5 hash when using DSA keys in ca. 3838 [Anonymous <nobody@replay.com>] 3839 3840 *) Generate an error if given an empty string as a cert directory. Also 3841 generate an error if handed NULL (previously returned 0 to indicate an 3842 error, but didn't set one). 3843 [Ben Laurie, reported by Anonymous <nobody@replay.com>] 3844 3845 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last. 3846 [Ben Laurie] 3847 3848 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct 3849 parameters. This was causing a warning which killed off the Win32 compile. 3850 [Steve Henson] 3851 3852 *) Remove C++ style comments from crypto/bn/bn_local.h. 3853 [Neil Costigan <neil.costigan@celocom.com>] 3854 3855 *) The function OBJ_txt2nid was broken. It was supposed to return a nid 3856 based on a text string, looking up short and long names and finally 3857 "dot" format. The "dot" format stuff didn't work. Added new function 3858 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote 3859 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the 3860 OID is not part of the table. 3861 [Steve Henson] 3862 3863 *) Add prototypes to X509 lookup/verify methods, fixing a bug in 3864 X509_LOOKUP_by_alias(). 3865 [Ben Laurie] 3866 3867 *) Sort openssl functions by name. 3868 [Ben Laurie] 3869 3870 *) Get the gendsa program working (hopefully) and add it to app list. Remove 3871 encryption from sample DSA keys (in case anyone is interested the password 3872 was "1234"). 3873 [Steve Henson] 3874 3875 *) Make _all_ *_free functions accept a NULL pointer. 3876 [Frans Heymans <fheymans@isaserver.be>] 3877 3878 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use 3879 NULL pointers. 3880 [Anonymous <nobody@replay.com>] 3881 3882 *) s_server should send the CAfile as acceptable CAs, not its own cert. 3883 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>] 3884 3885 *) Don't blow it for numeric -newkey arguments to apps/req. 3886 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>] 3887 3888 *) Temp key "for export" tests were wrong in s3_srvr.c. 3889 [Anonymous <nobody@replay.com>] 3890 3891 *) Add prototype for temp key callback functions 3892 SSL_CTX_set_tmp_{rsa,dh}_callback(). 3893 [Ben Laurie] 3894 3895 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and 3896 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey(). 3897 [Steve Henson] 3898 3899 *) X509_name_add_entry() freed the wrong thing after an error. 3900 [Arne Ansper <arne@ats.cyber.ee>] 3901 3902 *) rsa_eay.c would attempt to free a NULL context. 3903 [Arne Ansper <arne@ats.cyber.ee>] 3904 3905 *) BIO_s_socket() had a broken should_retry() on Windoze. 3906 [Arne Ansper <arne@ats.cyber.ee>] 3907 3908 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH. 3909 [Arne Ansper <arne@ats.cyber.ee>] 3910 3911 *) Make sure the already existing X509_STORE->depth variable is initialized 3912 in X509_STORE_new(), but document the fact that this variable is still 3913 unused in the certificate verification process. 3914 [Ralf S. Engelschall] 3915 3916 *) Fix the various library and apps files to free up pkeys obtained from 3917 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions. 3918 [Steve Henson] 3919 3920 *) Fix reference counting in X509_PUBKEY_get(). This makes 3921 demos/maurice/example2.c work, amongst others, probably. 3922 [Steve Henson and Ben Laurie] 3923 3924 *) First cut of a cleanup for apps/. First the `ssleay' program is now named 3925 `openssl' and second, the shortcut symlinks for the `openssl <command>' 3926 are no longer created. This way we have a single and consistent command 3927 line interface `openssl <command>', similar to `cvs <command>'. 3928 [Ralf S. Engelschall, Paul Sutton and Ben Laurie] 3929 3930 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey 3931 BIT STRING wrapper always have zero unused bits. 3932 [Steve Henson] 3933 3934 *) Add CA.pl, perl version of CA.sh, add extended key usage OID. 3935 [Steve Henson] 3936 3937 *) Make the top-level INSTALL documentation easier to understand. 3938 [Paul Sutton] 3939 3940 *) Makefiles updated to exit if an error occurs in a sub-directory 3941 make (including if user presses ^C) [Paul Sutton] 3942 3943 *) Make Montgomery context stuff explicit in RSA data structure. 3944 [Ben Laurie] 3945 3946 *) Fix build order of pem and err to allow for generated pem.h. 3947 [Ben Laurie] 3948 3949 *) Fix renumbering bug in X509_NAME_delete_entry(). 3950 [Ben Laurie] 3951 3952 *) Enhanced the err-ins.pl script so it makes the error library number 3953 global and can add a library name. This is needed for external ASN1 and 3954 other error libraries. 3955 [Steve Henson] 3956 3957 *) Fixed sk_insert which never worked properly. 3958 [Steve Henson] 3959 3960 *) Fix ASN1 macros so they can handle indefinite length construted 3961 EXPLICIT tags. Some non standard certificates use these: they can now 3962 be read in. 3963 [Steve Henson] 3964 3965 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc) 3966 into a single doc/ssleay.txt bundle. This way the information is still 3967 preserved but no longer messes up this directory. Now it's new room for 3968 the new set of documenation files. 3969 [Ralf S. Engelschall] 3970 3971 *) SETs were incorrectly DER encoded. This was a major pain, because they 3972 shared code with SEQUENCEs, which aren't coded the same. This means that 3973 almost everything to do with SETs or SEQUENCEs has either changed name or 3974 number of arguments. 3975 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>] 3976 3977 *) Fix test data to work with the above. 3978 [Ben Laurie] 3979 3980 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but 3981 was already fixed by Eric for 0.9.1 it seems. 3982 [Ben Laurie - pointed out by Ulf M�ller <ulf@fitug.de>] 3983 3984 *) Autodetect FreeBSD3. 3985 [Ben Laurie] 3986 3987 *) Fix various bugs in Configure. This affects the following platforms: 3988 nextstep 3989 ncr-scde 3990 unixware-2.0 3991 unixware-2.0-pentium 3992 sco5-cc. 3993 [Ben Laurie] 3994 3995 *) Eliminate generated files from CVS. Reorder tests to regenerate files 3996 before they are needed. 3997 [Ben Laurie] 3998 3999 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy). 4000 [Ben Laurie] 4001 4002 4003 Changes between 0.9.1b and 0.9.1c [23-Dec-1998] 4004 4005 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and 4006 changed SSLeay to OpenSSL in version strings. 4007 [Ralf S. Engelschall] 4008 4009 *) Some fixups to the top-level documents. 4010 [Paul Sutton] 4011 4012 *) Fixed the nasty bug where rsaref.h was not found under compile-time 4013 because the symlink to include/ was missing. 4014 [Ralf S. Engelschall] 4015 4016 *) Incorporated the popular no-RSA/DSA-only patches 4017 which allow to compile a RSA-free SSLeay. 4018 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall] 4019 4020 *) Fixed nasty rehash problem under `make -f Makefile.ssl links' 4021 when "ssleay" is still not found. 4022 [Ralf S. Engelschall] 4023 4024 *) Added more platforms to Configure: Cray T3E, HPUX 11, 4025 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>] 4026 4027 *) Updated the README file. 4028 [Ralf S. Engelschall] 4029 4030 *) Added various .cvsignore files in the CVS repository subdirs 4031 to make a "cvs update" really silent. 4032 [Ralf S. Engelschall] 4033 4034 *) Recompiled the error-definition header files and added 4035 missing symbols to the Win32 linker tables. 4036 [Ralf S. Engelschall] 4037 4038 *) Cleaned up the top-level documents; 4039 o new files: CHANGES and LICENSE 4040 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay 4041 o merged COPYRIGHT into LICENSE 4042 o removed obsolete TODO file 4043 o renamed MICROSOFT to INSTALL.W32 4044 [Ralf S. Engelschall] 4045 4046 *) Removed dummy files from the 0.9.1b source tree: 4047 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi 4048 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f 4049 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f 4050 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f 4051 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f 4052 [Ralf S. Engelschall] 4053 4054 *) Added various platform portability fixes. 4055 [Mark J. Cox] 4056 4057 *) The Genesis of the OpenSSL rpject: 4058 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A. 4059 Young and Tim J. Hudson created while they were working for C2Net until 4060 summer 1998. 4061 [The OpenSSL Project] 4062 4063 4064 Changes between 0.9.0b and 0.9.1b [not released] 4065 4066 *) Updated a few CA certificates under certs/ 4067 [Eric A. Young] 4068 4069 *) Changed some BIGNUM api stuff. 4070 [Eric A. Young] 4071 4072 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD, 4073 DGUX x86, Linux Alpha, etc. 4074 [Eric A. Young] 4075 4076 *) New COMP library [crypto/comp/] for SSL Record Layer Compression: 4077 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is 4078 available). 4079 [Eric A. Young] 4080 4081 *) Add -strparse option to asn1pars program which parses nested 4082 binary structures 4083 [Dr Stephen Henson <shenson@bigfoot.com>] 4084 4085 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs. 4086 [Eric A. Young] 4087 4088 *) DSA fix for "ca" program. 4089 [Eric A. Young] 4090 4091 *) Added "-genkey" option to "dsaparam" program. 4092 [Eric A. Young] 4093 4094 *) Added RIPE MD160 (rmd160) message digest. 4095 [Eric A. Young] 4096 4097 *) Added -a (all) option to "ssleay version" command. 4098 [Eric A. Young] 4099 4100 *) Added PLATFORM define which is the id given to Configure. 4101 [Eric A. Young] 4102 4103 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking. 4104 [Eric A. Young] 4105 4106 *) Extended the ASN.1 parser routines. 4107 [Eric A. Young] 4108 4109 *) Extended BIO routines to support REUSEADDR, seek, tell, etc. 4110 [Eric A. Young] 4111 4112 *) Added a BN_CTX to the BN library. 4113 [Eric A. Young] 4114 4115 *) Fixed the weak key values in DES library 4116 [Eric A. Young] 4117 4118 *) Changed API in EVP library for cipher aliases. 4119 [Eric A. Young] 4120 4121 *) Added support for RC2/64bit cipher. 4122 [Eric A. Young] 4123 4124 *) Converted the lhash library to the crypto/mem.c functions. 4125 [Eric A. Young] 4126 4127 *) Added more recognized ASN.1 object ids. 4128 [Eric A. Young] 4129 4130 *) Added more RSA padding checks for SSL/TLS. 4131 [Eric A. Young] 4132 4133 *) Added BIO proxy/filter functionality. 4134 [Eric A. Young] 4135 4136 *) Added extra_certs to SSL_CTX which can be used 4137 send extra CA certificates to the client in the CA cert chain sending 4138 process. It can be configured with SSL_CTX_add_extra_chain_cert(). 4139 [Eric A. Young] 4140 4141 *) Now Fortezza is denied in the authentication phase because 4142 this is key exchange mechanism is not supported by SSLeay at all. 4143 [Eric A. Young] 4144 4145 *) Additional PKCS1 checks. 4146 [Eric A. Young] 4147 4148 *) Support the string "TLSv1" for all TLS v1 ciphers. 4149 [Eric A. Young] 4150 4151 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the 4152 ex_data index of the SSL context in the X509_STORE_CTX ex_data. 4153 [Eric A. Young] 4154 4155 *) Fixed a few memory leaks. 4156 [Eric A. Young] 4157 4158 *) Fixed various code and comment typos. 4159 [Eric A. Young] 4160 4161 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0 4162 bytes sent in the client random. 4163 [Edward Bishop <ebishop@spyglass.com>] 4164 4165