5 Changes between 0.9.5a and 0.9.6 [xx XXX 2000]
7 *) mkstack.pl now sorts each macro group into lexical order.
8 Previously the output order depended on the order the files
9 appeared in the directory, resulting in needless rewriting
13 *) Patches to make OpenSSL compile under Win32 again. Mostly
14 work arounds for the VC++ problem that it treats func() as
15 func(void). Also stripped out the parts of mkdef.pl that
16 added extra typesafe functions: these no longer exist.
19 *) Reorganisation of the stack code. The macros are now all
20 collected in safestack.h . Each macro is defined in terms of
21 a "stack macro" of the form SKM_<name>(type, a, b). The
22 DEBUG_SAFESTACK is now handled in terms of function casts,
23 this has the advantage of retaining type safety without the
24 use of additional functions. If DEBUG_SAFESTACK is not defined
25 then the non typesafe macros are used instead. Also modified the
26 mkstack.pl script to handle the new form. Needs testing to see
27 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK
28 the default if no major problems. Similar behaviour for ASN1_SET_OF
32 *) When some versions of IIS use the 'NET' form of private key the
33 key derivation algorithm is different. Normally MD5(password) is
34 used as a 128 bit RC4 key. In the modified case
35 MD5(MD5(password) + "SGCKEYSALT") is used insted. Added some
36 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same
37 as the old Netscape_RSA functions except they have an additional
38 'sgckey' parameter which uses the modified algorithm. Also added
39 an -sgckey command line option to the rsa utility. Thanks to
40 Adrian Peck <bertie@ncipher.com> for posting details of the modified
41 algorithm to openssl-dev.
44 *) The evp_local.h macros were using 'c.##kname' which resulted in
45 invalid expansion on some systems (SCO 5.0.5 for example).
46 Corrected to 'c.kname'.
47 [Phillip Porch <root@theporch.com>]
49 *) New X509_get1_email() and X509_REQ_get1_email() functions that return
50 a STACK of email addresses from a certificate or request, these look
51 in the subject name and the subject alternative name extensions and
52 omit any duplicate addresses.
55 *) Re-implement BN_mod_exp2_mont using independent (and larger) windows.
56 This makes DSA verification about 2 % faster.
59 *) Increase maximum window size in BN_mod_exp_... to 6 bits instead of 5
60 (meaning that now 2^5 values will be precomputed, which is only 4 KB
61 plus overhead for 1024 bit moduli).
62 This makes exponentiations about 0.5 % faster for 1024 bit
63 exponents (as measured by "openssl speed rsa2048").
66 *) Rename memory handling macros to avoid conflicts with other
68 Malloc => OPENSSL_malloc
69 Malloc_locked => OPENSSL_malloc_locked
70 Realloc => OPENSSL_realloc
74 *) New function BN_mod_exp_mont_word for small bases (roughly 15%
75 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange).
79 [John Jarvie <jjarvie@newsguy.com>]
81 *) The type-safe stack code has been rejigged. It is now only compiled
82 in when OpenSSL is configured with the DEBUG_SAFESTACK option and
83 by default all type-specific stack functions are "#define"d back to
84 standard stack functions. This results in more streamlined output
85 but retains the type-safety checking possibilities of the original
89 *) The STACK code has been cleaned up, and certain type declarations
90 that didn't make a lot of sense have been brought in line. This has
91 also involved a cleanup of sorts in safestack.h to more correctly
92 map type-safe stack functions onto their plain stack counterparts.
93 This work has also resulted in a variety of "const"ifications of
94 lots of the code, especially "_cmp" operations which should normally
95 be prototyped with "const" parameters anyway.
98 *) When generating bytes for the first time in md_rand.c, 'stir the pool'
99 by seeding with STATE_SIZE dummy bytes (with zero entropy count).
100 (The PRNG state consists of two parts, the large pool 'state' and 'md',
101 where all of 'md' is used each time the PRNG is used, but 'state'
102 is used only indexed by a cyclic counter. As entropy may not be
103 well distributed from the beginning, 'md' is important as a
104 chaining variable. However, the output function chains only half
105 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains
106 all of 'md', and seeding with STATE_SIZE dummy bytes will result
107 in all of 'state' being rewritten, with the new values depending
108 on virtually all of 'md'. This overcomes the 80 bit limitation.)
111 *) In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when
112 the handshake is continued after ssl_verify_cert_chain();
113 otherwise, if SSL_VERIFY_NONE is set, remaining error codes
114 can lead to 'unexplainable' connection aborts later.
115 [Bodo Moeller; problem tracked down by Lutz Jaenicke]
117 *) Major EVP API cipher revision.
118 Add hooks for extra EVP features. This allows various cipher
119 parameters to be set in the EVP interface. Support added for variable
120 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and
121 setting of RC2 and RC5 parameters.
123 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length
126 Remove lots of duplicated code from the EVP library. For example *every*
127 cipher init() function handles the 'iv' in the same way according to the
128 cipher mode. They also all do nothing if the 'key' parameter is NULL and
129 for CFB and OFB modes they zero ctx->num.
131 New functionality allows removal of S/MIME code RC2 hack.
133 Most of the routines have the same form and so can be declared in terms
136 By shifting this to the top level EVP_CipherInit() it can be removed from
137 all individual ciphers. If the cipher wants to handle IVs or keys
138 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT
141 Change lots of functions like EVP_EncryptUpdate() to now return a
142 value: although software versions of the algorithms cannot fail
143 any installed hardware versions can.
146 *) Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if
147 this option is set, tolerate broken clients that send the negotiated
148 protocol version number instead of the requested protocol version
152 *) Call dh_tmp_cb (set by ..._TMP_DH_CB) with correct 'is_export' flag;
153 i.e. non-zero for export ciphersuites, zero otherwise.
154 Previous versions had this flag inverted, inconsistent with
155 rsa_tmp_cb (..._TMP_RSA_CB).
156 [Bodo Moeller; problem reported by Amit Chopra]
158 *) Add missing DSA library text string. Work around for some IIS
159 key files with invalid SEQUENCE encoding.
162 *) Add a document (doc/standards.txt) that list all kinds of standards
163 and so on that are implemented in OpenSSL.
166 *) Enhance c_rehash script. Old version would mishandle certificates
167 with the same subject name hash and wouldn't handle CRLs at all.
168 Added -fingerprint option to crl utility, to support new c_rehash
172 *) Eliminate non-ANSI declarations in crypto.h and stack.h.
175 *) Fix for SSL server purpose checking. Server checking was
176 rejecting certificates which had extended key usage present
177 but no ssl client purpose.
178 [Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>]
180 *) Make PKCS#12 code work with no password. The PKCS#12 spec
181 is a little unclear about how a blank password is handled.
182 Since the password in encoded as a BMPString with terminating
183 double NULL a zero length password would end up as just the
184 double NULL. However no password at all is different and is
185 handled differently in the PKCS#12 key generation code. NS
186 treats a blank password as zero length. MSIE treats it as no
187 password on export: but it will try both on import. We now do
188 the same: PKCS12_parse() tries zero length and no password if
189 the password is set to "" or NULL (NULL is now a valid password:
190 it wasn't before) as does the pkcs12 application.
193 *) Bugfixes in apps/x509.c: Avoid a memory leak; and don't use
194 perror when PEM_read_bio_X509_REQ fails, the error message must
195 be obtained from the error queue.
198 *) Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing
199 it in ERR_remove_state if appropriate, and change ERR_get_state
200 accordingly to avoid race conditions (this is necessary because
201 thread_hash is no longer constant once set).
204 *) Bugfix for linux-elf makefile.one.
207 *) RSA_get_default_method() will now cause a default
208 RSA_METHOD to be chosen if one doesn't exist already.
209 Previously this was only set during a call to RSA_new()
210 or RSA_new_method(NULL) meaning it was possible for
211 RSA_get_default_method() to return NULL.
214 *) Added native name translation to the existing DSO code
215 that will convert (if the flag to do so is set) filenames
216 that are sufficiently small and have no path information
217 into a canonical native form. Eg. "blah" converted to
218 "libblah.so" or "blah.dll" etc.
221 *) New function ERR_error_string_n(e, buf, len) which is like
222 ERR_error_string(e, buf), but writes at most 'len' bytes
223 including the 0 terminator. For ERR_error_string_n, 'buf'
225 [Damien Miller <djm@mindrot.org>, Bodo Moeller]
227 *) CONF library reworked to become more general. A new CONF
228 configuration file reader "class" is implemented as well as a
229 new functions (NCONF_*, for "New CONF") to handle it. The now
230 old CONF_* functions are still there, but are reimplemented to
231 work in terms of the new functions. Also, a set of functions
232 to handle the internal storage of the configuration data is
233 provided to make it easier to write new configuration file
234 reader "classes" (I can definitely see something reading a
235 configuration file in XML format, for example), called _CONF_*,
236 or "the configuration storage API"...
238 The new configuration file reading functions are:
240 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio,
241 NCONF_get_section, NCONF_get_string, NCONF_get_numbre
243 NCONF_default, NCONF_WIN32
245 NCONF_dump_fp, NCONF_dump_bio
247 NCONF_default and NCONF_WIN32 are method (or "class") choosers,
248 NCONF_new creates a new CONF object. This works in the same way
249 as other interfaces in OpenSSL, like the BIO interface.
250 NCONF_dump_* dump the internal storage of the configuration file,
251 which is useful for debugging. All other functions take the same
252 arguments as the old CONF_* functions wth the exception of the
253 first that must be a `CONF *' instead of a `LHASH *'.
255 To make it easer to use the new classes with the old CONF_* functions,
256 the function CONF_set_default_method is provided.
259 *) Add '-tls1' option to 'openssl ciphers', which was already
260 mentioned in the documentation but had not been implemented.
261 (This option is not yet really useful because even the additional
262 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.)
265 *) Initial DSO code added into libcrypto for letting OpenSSL (and
266 OpenSSL-based applications) load shared libraries and bind to
267 them in a portable way.
268 [Geoff Thorpe, with contributions from Richard Levitte]
270 Changes between 0.9.5 and 0.9.5a [1 Apr 2000]
272 *) Make sure _lrotl and _lrotr are only used with MSVC.
274 *) Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status
275 (the default implementation of RAND_status).
277 *) Rename openssl x509 option '-crlext', which was added in 0.9.5,
278 to '-clrext' (= clear extensions), as intended and documented.
279 [Bodo Moeller; inconsistency pointed out by Michael Attili
282 *) Fix for HMAC. It wasn't zeroing the rest of the block if the key length
283 was larger than the MD block size.
284 [Steve Henson, pointed out by Yost William <YostW@tce.com>]
286 *) Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument
287 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set()
288 using the passed key: if the passed key was a private key the result
289 of X509_print(), for example, would be to print out all the private key
293 *) des_quad_cksum() byte order bug fix.
294 [Ulf Möller, using the problem description in krb4-0.9.7, where
295 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>]
297 *) Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly
299 [Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>]
301 *) For easily testing in shell scripts whether some command
302 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX'
303 returns with exit code 0 iff no command of the given name is available.
304 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases,
305 the output goes to stdout and nothing is printed to stderr.
306 Additional arguments are always ignored.
308 Since for each cipher there is a command of the same name,
309 the 'no-cipher' compilation switches can be tested this way.
311 ('openssl no-XXX' is not able to detect pseudo-commands such
312 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.)
315 *) Update test suite so that 'make test' succeeds in 'no-rsa' configuration.
318 *) For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE
319 is set; it will be thrown away anyway because each handshake creates
321 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition
322 to parameters -- in previous versions (since OpenSSL 0.9.3) the
323 'default key' from SSL_CTX_set_tmp_dh would always be lost, meanining
324 you effectivly got SSL_OP_SINGLE_DH_USE when using this macro.
327 *) New s_client option -ign_eof: EOF at stdin is ignored, and
328 'Q' and 'R' lose their special meanings (quit/renegotiate).
329 This is part of what -quiet does; unlike -quiet, -ign_eof
330 does not suppress any output.
333 *) Add compatibility options to the purpose and trust code. The
334 purpose X509_PURPOSE_ANY is "any purpose" which automatically
335 accepts a certificate or CA, this was the previous behaviour,
336 with all the associated security issues.
338 X509_TRUST_COMPAT is the old trust behaviour: only and
339 automatically trust self signed roots in certificate store. A
340 new trust setting X509_TRUST_DEFAULT is used to specify that
341 a purpose has no associated trust setting and it should instead
342 use the value in the default purpose.
345 *) Fix the PKCS#8 DSA private key code so it decodes keys again
346 and fix a memory leak.
349 *) In util/mkerr.pl (which implements 'make errors'), preserve
350 reason strings from the previous version of the .c file, as
351 the default to have only downcase letters (and digits) in
352 automatically generated reasons codes is not always appropriate.
355 *) In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table
356 using strerror. Previously, ERR_reason_error_string() returned
357 library names as reason strings for SYSerr; but SYSerr is a special
358 case where small numbers are errno values, not library numbers.
361 *) Add '-dsaparam' option to 'openssl dhparam' application. This
362 converts DSA parameters into DH parameters. (When creating parameters,
363 DSA_generate_parameters is used.)
366 *) Include 'length' (recommended exponent length) in C code generated
367 by 'openssl dhparam -C'.
370 *) The second argument to set_label in perlasm was already being used
371 so couldn't be used as a "file scope" flag. Moved to third argument
375 *) In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes
376 instead of RAND_bytes for encryption IVs and salts.
379 *) Include RAND_status() into RAND_METHOD instead of implementing
380 it only for md_rand.c Otherwise replacing the PRNG by calling
381 RAND_set_rand_method would be impossible.
384 *) Don't let DSA_generate_key() enter an infinite loop if the random
385 number generation fails.
388 *) New 'rand' application for creating pseudo-random output.
391 *) Added configuration support for Linux/IA64
392 [Rolf Haberrecker <rolf@suse.de>]
394 *) Assembler module support for Mingw32.
397 *) Shared library support for HPUX (in shlib/).
398 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous]
400 *) Shared library support for Solaris gcc.
401 [Lutz Behnke <behnke@trustcenter.de>]
403 Changes between 0.9.4 and 0.9.5 [28 Feb 2000]
405 *) PKCS7_encrypt() was adding text MIME headers twice because they
406 were added manually and by SMIME_crlf_copy().
409 *) In bntest.c don't call BN_rand with zero bits argument.
410 [Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>]
412 *) BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n]
413 case was implemented. This caused BN_div_recp() to fail occasionally.
416 *) Add an optional second argument to the set_label() in the perl
417 assembly language builder. If this argument exists and is set
418 to 1 it signals that the assembler should use a symbol whose
419 scope is the entire file, not just the current function. This
420 is needed with MASM which uses the format label:: for this scope.
421 [Steve Henson, pointed out by Peter Runestig <peter@runestig.com>]
423 *) Change the ASN1 types so they are typedefs by default. Before
424 almost all types were #define'd to ASN1_STRING which was causing
425 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING)
429 *) Change names of new functions to the new get1/get0 naming
430 convention: After 'get1', the caller owns a reference count
431 and has to call ..._free; 'get0' returns a pointer to some
432 data structure without incrementing reference counters.
433 (Some of the existing 'get' functions increment a reference
434 counter, some don't.)
435 Similarly, 'set1' and 'add1' functions increase reference
436 counters or duplicate objects.
439 *) Allow for the possibility of temp RSA key generation failure:
440 the code used to assume it always worked and crashed on failure.
443 *) Fix potential buffer overrun problem in BIO_printf().
444 [Ulf Möller, using public domain code by Patrick Powell; problem
445 pointed out by David Sacerdote <das33@cornell.edu>]
447 *) Support EGD <http://www.lothar.com/tech/crypto/>. New functions
448 RAND_egd() and RAND_status(). In the command line application,
449 the EGD socket can be specified like a seed file using RANDFILE
453 *) Allow the string CERTIFICATE to be tolerated in PKCS#7 structures.
454 Some CAs (e.g. Verisign) distribute certificates in this form.
457 *) Remove the SSL_ALLOW_ADH compile option and set the default cipher
458 list to exclude them. This means that no special compilation option
459 is needed to use anonymous DH: it just needs to be included in the
463 *) Change the EVP_MD_CTX_type macro so its meaning consistent with
464 EVP_MD_type. The old functionality is available in a new macro called
465 EVP_MD_md(). Change code that uses it and update docs.
468 *) ..._ctrl functions now have corresponding ..._callback_ctrl functions
469 where the 'void *' argument is replaced by a function pointer argument.
470 Previously 'void *' was abused to point to functions, which works on
471 many platforms, but is not correct. As these functions are usually
472 called by macros defined in OpenSSL header files, most source code
473 should work without changes.
476 *) <openssl/opensslconf.h> (which is created by Configure) now contains
477 sections with information on -D... compiler switches used for
478 compiling the library so that applications can see them. To enable
479 one of these sections, a pre-processor symbol OPENSSL_..._DEFINES
480 must be defined. E.g.,
481 #define OPENSSL_ALGORITHM_DEFINES
482 #include <openssl/opensslconf.h>
483 defines all pertinent NO_<algo> symbols, such as NO_IDEA, NO_RSA, etc.
484 [Richard Levitte, Ulf and Bodo Möller]
486 *) Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS
490 *) Change the 'other' type in certificate aux info to a STACK_OF
491 X509_ALGOR. Although not an AlgorithmIdentifier as such it has
492 the required ASN1 format: arbitrary types determined by an OID.
495 *) Add some PEM_write_X509_REQ_NEW() functions and a command line
496 argument to 'req'. This is not because the function is newer or
497 better than others it just uses the work 'NEW' in the certificate
498 request header lines. Some software needs this.
501 *) Reorganise password command line arguments: now passwords can be
502 obtained from various sources. Delete the PEM_cb function and make
503 it the default behaviour: i.e. if the callback is NULL and the
504 usrdata argument is not NULL interpret it as a null terminated pass
505 phrase. If usrdata and the callback are NULL then the pass phrase
506 is prompted for as usual.
509 *) Add support for the Compaq Atalla crypto accelerator. If it is installed,
510 the support is automatically enabled. The resulting binaries will
511 autodetect the card and use it if present.
512 [Ben Laurie and Compaq Inc.]
514 *) Work around for Netscape hang bug. This sends certificate request
515 and server done in one record. Since this is perfectly legal in the
516 SSL/TLS protocol it isn't a "bug" option and is on by default. See
517 the bugs/SSLv3 entry for more info.
520 *) HP-UX tune-up: new unified configs, HP C compiler bug workaround.
523 *) Add -rand argument to smime and pkcs12 applications and read/write
527 *) New 'passwd' tool for crypt(3) and apr1 password hashes.
530 *) Add command line password options to the remaining applications.
533 *) Bug fix for BN_div_recp() for numerators with an even number of
537 *) More tests in bntest.c, and changed test_bn output.
540 *) ./config recognizes MacOS X now.
543 *) Bug fix for BN_div() when the first words of num and divsor are
544 equal (it gave wrong results if (rem=(n1-q*d0)&BN_MASK2) < d0).
547 *) Add support for various broken PKCS#8 formats, and command line
548 options to produce them.
551 *) New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to
552 get temporary BIGNUMs from a BN_CTX.
555 *) Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont()
559 *) Change the SSLeay_add_all_*() functions to OpenSSL_add_all_*() and
560 include a #define from the old name to the new. The original intent
561 was that statically linked binaries could for example just call
562 SSLeay_add_all_ciphers() to just add ciphers to the table and not
563 link with digests. This never worked becayse SSLeay_add_all_digests()
564 and SSLeay_add_all_ciphers() were in the same source file so calling
565 one would link with the other. They are now in separate source files.
568 *) Add a new -notext option to 'ca' and a -pubkey option to 'spkac'.
571 *) Use a less unusual form of the Miller-Rabin primality test (it used
572 a binary algorithm for exponentiation integrated into the Miller-Rabin
573 loop, our standard modexp algorithms are faster).
576 *) Support for the EBCDIC character set completed.
577 [Martin Kraemer <Martin.Kraemer@Mch.SNI.De>]
579 *) Source code cleanups: use const where appropriate, eliminate casts,
580 use void * instead of char * in lhash.
583 *) Bugfix: ssl3_send_server_key_exchange was not restartable
584 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of
585 this the server could overwrite ephemeral keys that the client
589 *) Turn DSA_is_prime into a macro that calls BN_is_prime,
590 using 50 iterations of the Rabin-Miller test.
592 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50
593 iterations of the Rabin-Miller test as required by the appendix
594 to FIPS PUB 186[-1]) instead of DSA_is_prime.
595 As BN_is_prime_fasttest includes trial division, DSA parameter
596 generation becomes much faster.
598 This implies a change for the callback functions in DSA_is_prime
599 and DSA_generate_parameters: The callback function is called once
600 for each positive witness in the Rabin-Miller test, not just
601 occasionally in the inner loop; and the parameters to the
602 callback function now provide an iteration count for the outer
603 loop rather than for the current invocation of the inner loop.
604 DSA_generate_parameters additionally can call the callback
605 function with an 'iteration count' of -1, meaning that a
606 candidate has passed the trial division test (when q is generated
607 from an application-provided seed, trial division is skipped).
610 *) New function BN_is_prime_fasttest that optionally does trial
611 division before starting the Rabin-Miller test and has
612 an additional BN_CTX * argument (whereas BN_is_prime always
613 has to allocate at least one BN_CTX).
614 'callback(1, -1, cb_arg)' is called when a number has passed the
615 trial division stage.
618 *) Fix for bug in CRL encoding. The validity dates weren't being handled
622 *) New -pkcs12 option to CA.pl script to write out a PKCS#12 file.
625 *) New function BN_pseudo_rand().
628 *) Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable)
629 bignum version of BN_from_montgomery() with the working code from
630 SSLeay 0.9.0 (the word based version is faster anyway), and clean up
634 *) Avoid a race condition in s2_clnt.c (function get_server_hello) that
635 made it impossible to use the same SSL_SESSION data structure in
636 SSL2 clients in multiple threads.
639 *) The return value of RAND_load_file() no longer counts bytes obtained
640 by stat(). RAND_load_file(..., -1) is new and uses the complete file
641 to seed the PRNG (previously an explicit byte count was required).
642 [Ulf Möller, Bodo Möller]
644 *) Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
645 used (char *) instead of (void *) and had casts all over the place.
648 *) Make BN_generate_prime() return NULL on error if ret!=NULL.
651 *) Retain source code compatibility for BN_prime_checks macro:
652 BN_is_prime(..., BN_prime_checks, ...) now uses
653 BN_prime_checks_for_size to determine the appropriate number of
654 Rabin-Miller iterations.
657 *) Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
658 DH_CHECK_P_NOT_SAFE_PRIME.
659 (Check if this is true? OpenPGP calls them "strong".)
662 *) Merge the functionality of "dh" and "gendh" programs into a new program
663 "dhparam". The old programs are retained for now but will handle DH keys
664 (instead of parameters) in future.
667 *) Make the ciphers, s_server and s_client programs check the return values
668 when a new cipher list is set.
671 *) Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
672 ciphers. Before when the 56bit ciphers were enabled the sorting was
675 The syntax for the cipher sorting has been extended to support sorting by
676 cipher-strength (using the strength_bits hard coded in the tables).
677 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
679 Fix a bug in the cipher-command parser: when supplying a cipher command
680 string with an "undefined" symbol (neither command nor alphanumeric
681 [A-Za-z0-9], ssl_set_cipher_list used to hang in an endless loop. Now
684 Due to the strength-sorting extension, the code of the
685 ssl_create_cipher_list() function was completely rearranged. I hope that
686 the readability was also increased :-)
687 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
689 *) Minor change to 'x509' utility. The -CAcreateserial option now uses 1
690 for the first serial number and places 2 in the serial number file. This
691 avoids problems when the root CA is created with serial number zero and
692 the first user certificate has the same issuer name and serial number
696 *) Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
697 the new code. Add documentation for this stuff.
700 *) Changes to X509_ATTRIBUTE utilities. These have been renamed from
701 X509_*() to X509at_*() on the grounds that they don't handle X509
702 structures and behave in an analagous way to the X509v3 functions:
703 they shouldn't be called directly but wrapper functions should be used
706 So we also now have some wrapper functions that call the X509at functions
707 when passed certificate requests. (TO DO: similar things can be done with
708 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
709 things. Some of these need some d2i or i2d and print functionality
710 because they handle more complex structures.)
713 *) Add missing #ifndefs that caused missing symbols when building libssl
714 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
716 [Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller]
718 *) Precautions against using the PRNG uninitialized: RAND_bytes() now
719 has a return value which indicates the quality of the random data
720 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
721 error queue. New function RAND_pseudo_bytes() generates output that is
722 guaranteed to be unique but not unpredictable. RAND_add is like
723 RAND_seed, but takes an extra argument for an entropy estimate
724 (RAND_seed always assumes full entropy).
727 *) Do more iterations of Rabin-Miller probable prime test (specifically,
728 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
729 instead of only 2 for all lengths; see BN_prime_checks_for_size definition
730 in crypto/bn/bn_prime.c for the complete table). This guarantees a
731 false-positive rate of at most 2^-80 for random input.
734 *) Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
737 *) New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain
738 in the 0.9.5 release), this returns the chain
739 from an X509_CTX structure with a dup of the stack and all
740 the X509 reference counts upped: so the stack will exist
741 after X509_CTX_cleanup() has been called. Modify pkcs12.c
744 Also make SSL_SESSION_print() print out the verify return
748 *) Add manpage for the pkcs12 command. Also change the default
749 behaviour so MAC iteration counts are used unless the new
750 -nomaciter option is used. This improves file security and
751 only older versions of MSIE (4.0 for example) need it.
754 *) Honor the no-xxx Configure options when creating .DEF files.
757 *) Add PKCS#10 attributes to field table: challengePassword,
758 unstructuredName and unstructuredAddress. These are taken from
759 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
760 international characters are used.
762 More changes to X509_ATTRIBUTE code: allow the setting of types
763 based on strings. Remove the 'loc' parameter when adding
764 attributes because these will be a SET OF encoding which is sorted
768 *) Initial changes to the 'req' utility to allow request generation
769 automation. This will allow an application to just generate a template
770 file containing all the field values and have req construct the
773 Initial support for X509_ATTRIBUTE handling. Stacks of these are
774 used all over the place including certificate requests and PKCS#7
775 structures. They are currently handled manually where necessary with
776 some primitive wrappers for PKCS#7. The new functions behave in a
777 manner analogous to the X509 extension functions: they allow
778 attributes to be looked up by NID and added.
780 Later something similar to the X509V3 code would be desirable to
781 automatically handle the encoding, decoding and printing of the
782 more complex types. The string types like challengePassword can
783 be handled by the string table functions.
785 Also modified the multi byte string table handling. Now there is
786 a 'global mask' which masks out certain types. The table itself
787 can use the flag STABLE_NO_MASK to ignore the mask setting: this
788 is useful when for example there is only one permissible type
789 (as in countryName) and using the mask might result in no valid
793 *) Clean up 'Finished' handling, and add functions SSL_get_finished and
794 SSL_get_peer_finished to allow applications to obtain the latest
795 Finished messages sent to the peer or expected from the peer,
796 respectively. (SSL_get_peer_finished is usually the Finished message
797 actually received from the peer, otherwise the protocol will be aborted.)
799 As the Finished message are message digests of the complete handshake
800 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
801 be used for external authentication procedures when the authentication
802 provided by SSL/TLS is not desired or is not enough.
805 *) Enhanced support for Alpha Linux is added. Now ./config checks if
806 the host supports BWX extension and if Compaq C is present on the
807 $PATH. Just exploiting of the BWX extension results in 20-30%
808 performance kick for some algorithms, e.g. DES and RC4 to mention
809 a couple. Compaq C in turn generates ~20% faster code for MD5 and
813 *) Add support for MS "fast SGC". This is arguably a violation of the
814 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
815 weak crypto and after checking the certificate is SGC a second one
816 with strong crypto. MS SGC stops the first handshake after receiving
817 the server certificate message and sends a second client hello. Since
818 a server will typically do all the time consuming operations before
819 expecting any further messages from the client (server key exchange
820 is the most expensive) there is little difference between the two.
822 To get OpenSSL to support MS SGC we have to permit a second client
823 hello message after we have sent server done. In addition we have to
824 reset the MAC if we do get this second client hello.
827 *) Add a function 'd2i_AutoPrivateKey()' this will automatically decide
828 if a DER encoded private key is RSA or DSA traditional format. Changed
829 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
830 format DER encoded private key. Newer code should use PKCS#8 format which
831 has the key type encoded in the ASN1 structure. Added DER private key
832 support to pkcs8 application.
835 *) SSL 3/TLS 1 servers now don't request certificates when an anonymous
836 ciphersuites has been selected (as required by the SSL 3/TLS 1
837 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
838 is set, we interpret this as a request to violate the specification
839 (the worst that can happen is a handshake failure, and 'correct'
840 behaviour would result in a handshake failure anyway).
843 *) In SSL_CTX_add_session, take into account that there might be multiple
844 SSL_SESSION structures with the same session ID (e.g. when two threads
845 concurrently obtain them from an external cache).
846 The internal cache can handle only one SSL_SESSION with a given ID,
847 so if there's a conflict, we now throw out the old one to achieve
851 *) Add OIDs for idea and blowfish in CBC mode. This will allow both
852 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
853 some routines that use cipher OIDs: some ciphers do not have OIDs
854 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
858 *) Simplify the trust setting structure and code. Now we just have
859 two sequences of OIDs for trusted and rejected settings. These will
860 typically have values the same as the extended key usage extension
861 and any application specific purposes.
863 The trust checking code now has a default behaviour: it will just
864 check for an object with the same NID as the passed id. Functions can
865 be provided to override either the default behaviour or the behaviour
866 for a given id. SSL client, server and email already have functions
867 in place for compatibility: they check the NID and also return "trusted"
868 if the certificate is self signed.
871 *) Add d2i,i2d bio/fp functions for PrivateKey: these convert the
872 traditional format into an EVP_PKEY structure.
875 *) Add a password callback function PEM_cb() which either prompts for
876 a password if usr_data is NULL or otherwise assumes it is a null
877 terminated password. Allow passwords to be passed on command line
878 environment or config files in a few more utilities.
881 *) Add a bunch of DER and PEM functions to handle PKCS#8 format private
882 keys. Add some short names for PKCS#8 PBE algorithms and allow them
883 to be specified on the command line for the pkcs8 and pkcs12 utilities.
884 Update documentation.
887 *) Support for ASN1 "NULL" type. This could be handled before by using
888 ASN1_TYPE but there wasn't any function that would try to read a NULL
889 and produce an error if it couldn't. For compatibility we also have
890 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
891 don't allocate anything because they don't need to.
894 *) Initial support for MacOS is now provided. Examine INSTALL.MacOS
896 [Andy Polyakov, Roy Woods <roy@centicsystems.ca>]
898 *) Rebuild of the memory allocation routines used by OpenSSL code and
899 possibly others as well. The purpose is to make an interface that
900 provide hooks so anyone can build a separate set of allocation and
901 deallocation routines to be used by OpenSSL, for example memory
902 pool implementations, or something else, which was previously hard
903 since Malloc(), Realloc() and Free() were defined as macros having
904 the values malloc, realloc and free, respectively (except for Win32
905 compilations). The same is provided for memory debugging code.
906 OpenSSL already comes with functionality to find memory leaks, but
907 this gives people a chance to debug other memory problems.
909 With these changes, a new set of functions and macros have appeared:
911 CRYPTO_set_mem_debug_functions() [F]
912 CRYPTO_get_mem_debug_functions() [F]
913 CRYPTO_dbg_set_options() [F]
914 CRYPTO_dbg_get_options() [F]
915 CRYPTO_malloc_debug_init() [M]
917 The memory debug functions are NULL by default, unless the library
918 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
919 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
920 gives the standard debugging functions that come with OpenSSL) or
921 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
922 provided by the library user) must be used. When the standard
923 debugging functions are used, CRYPTO_dbg_set_options can be used to
924 request additional information:
925 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
926 the CRYPTO_MDEBUG_xxx macro when compiling the library.
928 Also, things like CRYPTO_set_mem_functions will always give the
929 expected result (the new set of functions is used for allocation
930 and deallocation) at all times, regardless of platform and compiler
933 To finish it up, some functions that were never use in any other
934 way than through macros have a new API and new semantic:
940 All macros of value have retained their old syntax.
941 [Richard Levitte and Bodo Moeller]
943 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
944 ordering of SMIMECapabilities wasn't in "strength order" and there
945 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
949 *) Some ASN1 types with illegal zero length encoding (INTEGER,
950 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
951 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson]
953 *) Merge in my S/MIME library for OpenSSL. This provides a simple
954 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
955 functionality to handle multipart/signed properly) and a utility
956 called 'smime' to call all this stuff. This is based on code I
957 originally wrote for Celo who have kindly allowed it to be
961 *) Add variants des_set_key_checked and des_set_key_unchecked of
962 des_set_key (aka des_key_sched). Global variable des_check_key
963 decides which of these is called by des_set_key; this way
964 des_check_key behaves as it always did, but applications and
965 the library itself, which was buggy for des_check_key == 1,
966 have a cleaner way to pick the version they need.
969 *) New function PKCS12_newpass() which changes the password of a
973 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and
974 dynamic mix. In both cases the ids can be used as an index into the
975 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
976 functions so they accept a list of the field values and the
977 application doesn't need to directly manipulate the X509_TRUST
981 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
985 *) Modify the way the V3 extension code looks up extensions. This now
986 works in a similar way to the object code: we have some "standard"
987 extensions in a static table which is searched with OBJ_bsearch()
988 and the application can add dynamic ones if needed. The file
989 crypto/x509v3/ext_dat.h now has the info: this file needs to be
990 updated whenever a new extension is added to the core code and kept
991 in ext_nid order. There is a simple program 'tabtest.c' which checks
992 this. New extensions are not added too often so this file can readily
993 be maintained manually.
995 There are two big advantages in doing things this way. The extensions
996 can be looked up immediately and no longer need to be "added" using
997 X509V3_add_standard_extensions(): this function now does nothing.
998 [Side note: I get *lots* of email saying the extension code doesn't
999 work because people forget to call this function]
1000 Also no dynamic allocation is done unless new extensions are added:
1001 so if we don't add custom extensions there is no need to call
1002 X509V3_EXT_cleanup().
1005 *) Modify enc utility's salting as follows: make salting the default. Add a
1006 magic header, so unsalted files fail gracefully instead of just decrypting
1007 to garbage. This is because not salting is a big security hole, so people
1008 should be discouraged from doing it.
1011 *) Fixes and enhancements to the 'x509' utility. It allowed a message
1012 digest to be passed on the command line but it only used this
1013 parameter when signing a certificate. Modified so all relevant
1014 operations are affected by the digest parameter including the
1015 -fingerprint and -x509toreq options. Also -x509toreq choked if a
1016 DSA key was used because it didn't fix the digest.
1019 *) Initial certificate chain verify code. Currently tests the untrusted
1020 certificates for consistency with the verify purpose (which is set
1021 when the X509_STORE_CTX structure is set up) and checks the pathlength.
1023 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
1024 this is because it will reject chains with invalid extensions whereas
1025 every previous version of OpenSSL and SSLeay made no checks at all.
1027 Trust code: checks the root CA for the relevant trust settings. Trust
1028 settings have an initial value consistent with the verify purpose: e.g.
1029 if the verify purpose is for SSL client use it expects the CA to be
1030 trusted for SSL client use. However the default value can be changed to
1031 permit custom trust settings: one example of this would be to only trust
1032 certificates from a specific "secure" set of CAs.
1034 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
1035 which should be used for version portability: especially since the
1036 verify structure is likely to change more often now.
1038 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
1039 to set them. If not set then assume SSL clients will verify SSL servers
1042 Two new options to the verify program: -untrusted allows a set of
1043 untrusted certificates to be passed in and -purpose which sets the
1044 intended purpose of the certificate. If a purpose is set then the
1045 new chain verify code is used to check extension consistency.
1048 *) Support for the authority information access extension.
1051 *) Modify RSA and DSA PEM read routines to transparently handle
1052 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
1053 public keys in a format compatible with certificate
1054 SubjectPublicKeyInfo structures. Unfortunately there were already
1055 functions called *_PublicKey_* which used various odd formats so
1056 these are retained for compatibility: however the DSA variants were
1057 never in a public release so they have been deleted. Changed dsa/rsa
1058 utilities to handle the new format: note no releases ever handled public
1059 keys so we should be OK.
1061 The primary motivation for this change is to avoid the same fiasco
1062 that dogs private keys: there are several incompatible private key
1063 formats some of which are standard and some OpenSSL specific and
1064 require various evil hacks to allow partial transparent handling and
1065 even then it doesn't work with DER formats. Given the option anything
1066 other than PKCS#8 should be dumped: but the other formats have to
1067 stay in the name of compatibility.
1069 With public keys and the benefit of hindsight one standard format
1070 is used which works with EVP_PKEY, RSA or DSA structures: though
1071 it clearly returns an error if you try to read the wrong kind of key.
1073 Added a -pubkey option to the 'x509' utility to output the public key.
1074 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*()
1075 (renamed to EVP_PKEY_get1_*() in the OpenSSL 0.9.5 release) and add
1076 EVP_PKEY_rset_*() functions (renamed to EVP_PKEY_set1_*())
1077 that do the same as the EVP_PKEY_assign_*() except they up the
1078 reference count of the added key (they don't "swallow" the
1082 *) Fixes to crypto/x509/by_file.c the code to read in certificates and
1083 CRLs would fail if the file contained no certificates or no CRLs:
1084 added a new function to read in both types and return the number
1085 read: this means that if none are read it will be an error. The
1086 DER versions of the certificate and CRL reader would always fail
1087 because it isn't possible to mix certificates and CRLs in DER format
1088 without choking one or the other routine. Changed this to just read
1089 a certificate: this is the best we can do. Also modified the code
1090 in apps/verify.c to take notice of return codes: it was previously
1091 attempting to read in certificates from NULL pointers and ignoring
1092 any errors: this is one reason why the cert and CRL reader seemed
1093 to work. It doesn't check return codes from the default certificate
1094 routines: these may well fail if the certificates aren't installed.
1097 *) Code to support otherName option in GeneralName.
1100 *) First update to verify code. Change the verify utility
1101 so it warns if it is passed a self signed certificate:
1102 for consistency with the normal behaviour. X509_verify
1103 has been modified to it will now verify a self signed
1104 certificate if *exactly* the same certificate appears
1105 in the store: it was previously impossible to trust a
1106 single self signed certificate. This means that:
1107 openssl verify ss.pem
1108 now gives a warning about a self signed certificate but
1109 openssl verify -CAfile ss.pem ss.pem
1113 *) For servers, store verify_result in SSL_SESSION data structure
1114 (and add it to external session representation).
1115 This is needed when client certificate verifications fails,
1116 but an application-provided verification callback (set by
1117 SSL_CTX_set_cert_verify_callback) allows accepting the session
1118 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
1119 but returns 1): When the session is reused, we have to set
1120 ssl->verify_result to the appropriate error code to avoid
1122 [Bodo Moeller, problem pointed out by Lutz Jaenicke]
1124 *) Fix a bug in the new PKCS#7 code: it didn't consider the
1125 case in PKCS7_dataInit() where the signed PKCS7 structure
1126 didn't contain any existing data because it was being created.
1127 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson]
1129 *) Add a salt to the key derivation routines in enc.c. This
1130 forms the first 8 bytes of the encrypted file. Also add a
1131 -S option to allow a salt to be input on the command line.
1134 *) New function X509_cmp(). Oddly enough there wasn't a function
1135 to compare two certificates. We do this by working out the SHA1
1136 hash and comparing that. X509_cmp() will be needed by the trust
1140 *) SSL_get1_session() is like SSL_get_session(), but increments
1141 the reference count in the SSL_SESSION returned.
1142 [Geoff Thorpe <geoff@eu.c2.net>]
1144 *) Fix for 'req': it was adding a null to request attributes.
1145 Also change the X509_LOOKUP and X509_INFO code to handle
1146 certificate auxiliary information.
1149 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document
1153 *) Add the possibility to add extra information to the memory leak
1154 detecting output, to form tracebacks, showing from where each
1155 allocation was originated: CRYPTO_push_info("constant string") adds
1156 the string plus current file name and line number to a per-thread
1157 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
1158 is like calling CYRPTO_pop_info() until the stack is empty.
1159 Also updated memory leak detection code to be multi-thread-safe.
1162 *) Add options -text and -noout to pkcs7 utility and delete the
1163 encryption options which never did anything. Update docs.
1166 *) Add options to some of the utilities to allow the pass phrase
1167 to be included on either the command line (not recommended on
1168 OSes like Unix) or read from the environment. Update the
1169 manpages and fix a few bugs.
1172 *) Add a few manpages for some of the openssl commands.
1175 *) Fix the -revoke option in ca. It was freeing up memory twice,
1176 leaking and not finding already revoked certificates.
1179 *) Extensive changes to support certificate auxiliary information.
1180 This involves the use of X509_CERT_AUX structure and X509_AUX
1181 functions. An X509_AUX function such as PEM_read_X509_AUX()
1182 can still read in a certificate file in the usual way but it
1183 will also read in any additional "auxiliary information". By
1184 doing things this way a fair degree of compatibility can be
1185 retained: existing certificates can have this information added
1186 using the new 'x509' options.
1188 Current auxiliary information includes an "alias" and some trust
1189 settings. The trust settings will ultimately be used in enhanced
1190 certificate chain verification routines: currently a certificate
1191 can only be trusted if it is self signed and then it is trusted
1195 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD).
1196 The problem was that one of the replacement routines had not been working
1197 since SSLeay releases. For now the offending routine has been replaced
1198 with non-optimised assembler. Even so, this now gives around 95%
1199 performance improvement for 1024 bit RSA signs.
1202 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2
1203 handling. Most clients have the effective key size in bits equal to
1204 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
1205 A few however don't do this and instead use the size of the decrypted key
1206 to determine the RC2 key length and the AlgorithmIdentifier to determine
1207 the effective key length. In this case the effective key length can still
1208 be 40 bits but the key length can be 168 bits for example. This is fixed
1209 by manually forcing an RC2 key into the EVP_PKEY structure because the
1210 EVP code can't currently handle unusual RC2 key sizes: it always assumes
1211 the key length and effective key length are equal.
1214 *) Add a bunch of functions that should simplify the creation of
1215 X509_NAME structures. Now you should be able to do:
1216 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
1217 and have it automatically work out the correct field type and fill in
1218 the structures. The more adventurous can try:
1219 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
1220 and it will (hopefully) work out the correct multibyte encoding.
1223 *) Change the 'req' utility to use the new field handling and multibyte
1224 copy routines. Before the DN field creation was handled in an ad hoc
1225 way in req, ca, and x509 which was rather broken and didn't support
1226 BMPStrings or UTF8Strings. Since some software doesn't implement
1227 BMPStrings or UTF8Strings yet, they can be enabled using the config file
1228 using the dirstring_type option. See the new comment in the default
1229 openssl.cnf for more info.
1232 *) Make crypto/rand/md_rand.c more robust:
1233 - Assure unique random numbers after fork().
1234 - Make sure that concurrent threads access the global counter and
1235 md serializably so that we never lose entropy in them
1236 or use exactly the same state in multiple threads.
1237 Access to the large state is not always serializable because
1238 the additional locking could be a performance killer, and
1239 md should be large enough anyway.
1242 *) New file apps/app_rand.c with commonly needed functionality
1243 for handling the random seed file.
1245 Use the random seed file in some applications that previously did not:
1247 dsaparam -genkey (which also ignored its '-rand' option),
1250 x509 (when signing).
1251 Except on systems with /dev/urandom, it is crucial to have a random
1252 seed file at least for key creation, DSA signing, and for DH exchanges;
1253 for RSA signatures we could do without one.
1255 gendh and gendsa (unlike genrsa) used to read only the first byte
1256 of each file listed in the '-rand' option. The function as previously
1257 found in genrsa is now in app_rand.c and is used by all programs
1258 that support '-rand'.
1261 *) In RAND_write_file, use mode 0600 for creating files;
1262 don't just chmod when it may be too late.
1265 *) Report an error from X509_STORE_load_locations
1266 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
1269 *) New function ASN1_mbstring_copy() this copies a string in either
1270 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
1271 into an ASN1_STRING type. A mask of permissible types is passed
1272 and it chooses the "minimal" type to use or an error if not type
1276 *) Add function equivalents to the various macros in asn1.h. The old
1277 macros are retained with an M_ prefix. Code inside the library can
1278 use the M_ macros. External code (including the openssl utility)
1279 should *NOT* in order to be "shared library friendly".
1282 *) Add various functions that can check a certificate's extensions
1283 to see if it usable for various purposes such as SSL client,
1284 server or S/MIME and CAs of these types. This is currently
1285 VERY EXPERIMENTAL but will ultimately be used for certificate chain
1286 verification. Also added a -purpose flag to x509 utility to
1287 print out all the purposes.
1290 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated
1294 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search
1295 for, obtain and decode and extension and obtain its critical flag.
1296 This allows all the necessary extension code to be handled in a
1297 single function call.
1300 *) RC4 tune-up featuring 30-40% performance improvement on most RISC
1301 platforms. See crypto/rc4/rc4_enc.c for further details.
1304 *) New -noout option to asn1parse. This causes no output to be produced
1305 its main use is when combined with -strparse and -out to extract data
1306 from a file (which may not be in ASN.1 format).
1309 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer
1310 when producing the local key id.
1311 [Richard Levitte <levitte@stacken.kth.se>]
1313 *) New option -dhparam in s_server. This allows a DH parameter file to be
1314 stated explicitly. If it is not stated then it tries the first server
1315 certificate file. The previous behaviour hard coded the filename
1319 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow
1320 a public key to be input or output. For example:
1321 openssl rsa -in key.pem -pubout -out pubkey.pem
1322 Also added necessary DSA public key functions to handle this.
1325 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
1326 in the message. This was handled by allowing
1327 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
1328 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>]
1330 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
1331 to the end of the strings whereas this didn't. This would cause problems
1332 if strings read with d2i_ASN1_bytes() were later modified.
1333 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>]
1335 *) Fix for base64 decode bug. When a base64 bio reads only one line of
1336 data and it contains EOF it will end up returning an error. This is
1337 caused by input 46 bytes long. The cause is due to the way base64
1338 BIOs find the start of base64 encoded data. They do this by trying a
1339 trial decode on each line until they find one that works. When they
1340 do a flag is set and it starts again knowing it can pass all the
1341 data directly through the decoder. Unfortunately it doesn't reset
1342 the context it uses. This means that if EOF is reached an attempt
1343 is made to pass two EOFs through the context and this causes the
1344 resulting error. This can also cause other problems as well. As is
1345 usual with these problems it takes *ages* to find and the fix is
1346 trivial: move one line.
1347 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ]
1349 *) Ugly workaround to get s_client and s_server working under Windows. The
1350 old code wouldn't work because it needed to select() on sockets and the
1351 tty (for keypresses and to see if data could be written). Win32 only
1352 supports select() on sockets so we select() with a 1s timeout on the
1353 sockets and then see if any characters are waiting to be read, if none
1354 are present then we retry, we also assume we can always write data to
1355 the tty. This isn't nice because the code then blocks until we've
1356 received a complete line of data and it is effectively polling the
1357 keyboard at 1s intervals: however it's quite a bit better than not
1358 working at all :-) A dedicated Windows application might handle this
1359 with an event loop for example.
1362 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
1363 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
1364 will be called when RSA_sign() and RSA_verify() are used. This is useful
1365 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
1366 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
1367 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
1368 This necessitated the support of an extra signature type NID_md5_sha1
1369 for SSL signatures and modifications to the SSL library to use it instead
1370 of calling RSA_public_decrypt() and RSA_private_encrypt().
1373 *) Add new -verify -CAfile and -CApath options to the crl program, these
1374 will lookup a CRL issuers certificate and verify the signature in a
1375 similar way to the verify program. Tidy up the crl program so it
1376 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
1377 less strict. It will now permit CRL extensions even if it is not
1378 a V2 CRL: this will allow it to tolerate some broken CRLs.
1381 *) Initialize all non-automatic variables each time one of the openssl
1382 sub-programs is started (this is necessary as they may be started
1383 multiple times from the "OpenSSL>" prompt).
1384 [Lennart Bang, Bodo Moeller]
1386 *) Preliminary compilation option RSA_NULL which disables RSA crypto without
1387 removing all other RSA functionality (this is what NO_RSA does). This
1388 is so (for example) those in the US can disable those operations covered
1389 by the RSA patent while allowing storage and parsing of RSA keys and RSA
1393 *) Non-copying interface to BIO pairs.
1394 (still largely untested)
1397 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive
1398 ASCII string. This was handled independently in various places before.
1401 *) New functions UTF8_getc() and UTF8_putc() that parse and generate
1402 UTF8 strings a character at a time.
1405 *) Use client_version from client hello to select the protocol
1406 (s23_srvr.c) and for RSA client key exchange verification
1407 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
1410 *) Add various utility functions to handle SPKACs, these were previously
1411 handled by poking round in the structure internals. Added new function
1412 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
1413 print, verify and generate SPKACs. Based on an original idea from
1414 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
1417 *) RIPEMD160 is operational on all platforms and is back in 'make test'.
1420 *) Allow the config file extension section to be overwritten on the
1421 command line. Based on an original idea from Massimiliano Pala
1422 <madwolf@comune.modena.it>. The new option is called -extensions
1423 and can be applied to ca, req and x509. Also -reqexts to override
1424 the request extensions in req and -crlexts to override the crl extensions
1428 *) Add new feature to the SPKAC handling in ca. Now you can include
1429 the same field multiple times by preceding it by "XXXX." for example:
1432 this is the same syntax as used in the req config file.
1435 *) Allow certificate extensions to be added to certificate requests. These
1436 are specified in a 'req_extensions' option of the req section of the
1437 config file. They can be printed out with the -text option to req but
1438 are otherwise ignored at present.
1441 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
1442 data read consists of only the final block it would not decrypted because
1443 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
1444 A misplaced 'break' also meant the decrypted final block might not be
1445 copied until the next read.
1448 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
1449 a few extra parameters to the DH structure: these will be useful if
1450 for example we want the value of 'q' or implement X9.42 DH.
1453 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and
1454 provides hooks that allow the default DSA functions or functions on a
1455 "per key" basis to be replaced. This allows hardware acceleration and
1456 hardware key storage to be handled without major modification to the
1457 library. Also added low level modexp hooks and CRYPTO_EX structure and
1458 associated functions.
1461 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
1462 as "read only": it can't be written to and the buffer it points to will
1463 not be freed. Reading from a read only BIO is much more efficient than
1464 a normal memory BIO. This was added because there are several times when
1465 an area of memory needs to be read from a BIO. The previous method was
1466 to create a memory BIO and write the data to it, this results in two
1467 copies of the data and an O(n^2) reading algorithm. There is a new
1468 function BIO_new_mem_buf() which creates a read only memory BIO from
1469 an area of memory. Also modified the PKCS#7 routines to use read only
1473 *) Bugfix: ssl23_get_client_hello did not work properly when called in
1474 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
1475 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
1476 but a retry condition occured while trying to read the rest.
1479 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as
1480 NID_pkcs7_encrypted by default: this was wrong since this should almost
1481 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
1482 the encrypted data type: this is a more sensible place to put it and it
1483 allows the PKCS#12 code to be tidied up that duplicated this
1487 *) Changed obj_dat.pl script so it takes its input and output files on
1488 the command line. This should avoid shell escape redirection problems
1492 *) Initial support for certificate extension requests, these are included
1493 in things like Xenroll certificate requests. Included functions to allow
1494 extensions to be obtained and added.
1497 *) -crlf option to s_client and s_server for sending newlines as
1498 CRLF (as required by many protocols).
1501 Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
1503 *) Install libRSAglue.a when OpenSSL is built with RSAref.
1504 [Ralf S. Engelschall]
1506 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency.
1507 [Andrija Antonijevic <TheAntony2@bigfoot.com>]
1509 *) Fix -startdate and -enddate (which was missing) arguments to 'ca'
1513 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as
1514 DH parameters/keys (q is lost during that conversion, but the resulting
1515 DH parameters contain its length).
1517 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
1518 much faster than DH_generate_parameters (which creates parameters
1519 where p = 2*q + 1), and also the smaller q makes DH computations
1520 much more efficient (160-bit exponentiation instead of 1024-bit
1521 exponentiation); so this provides a convenient way to support DHE
1522 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
1523 utter importance to use
1524 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
1526 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
1527 when such DH parameters are used, because otherwise small subgroup
1528 attacks may become possible!
1531 *) Avoid memory leak in i2d_DHparams.
1534 *) Allow the -k option to be used more than once in the enc program:
1535 this allows the same encrypted message to be read by multiple recipients.
1538 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
1539 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
1540 it will always use the numerical form of the OID, even if it has a short
1544 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
1545 method only got called if p,q,dmp1,dmq1,iqmp components were present,
1546 otherwise bn_mod_exp was called. In the case of hardware keys for example
1547 no private key components need be present and it might store extra data
1548 in the RSA structure, which cannot be accessed from bn_mod_exp.
1549 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
1550 private key operations.
1553 *) Added support for SPARC Linux.
1556 *) pem_password_cb function type incompatibly changed from
1557 typedef int pem_password_cb(char *buf, int size, int rwflag);
1559 ....(char *buf, int size, int rwflag, void *userdata);
1560 so that applications can pass data to their callbacks:
1561 The PEM[_ASN1]_{read,write}... functions and macros now take an
1562 additional void * argument, which is just handed through whenever
1563 the password callback is called.
1564 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller]
1566 New function SSL_CTX_set_default_passwd_cb_userdata.
1568 Compatibility note: As many C implementations push function arguments
1569 onto the stack in reverse order, the new library version is likely to
1570 interoperate with programs that have been compiled with the old
1571 pem_password_cb definition (PEM_whatever takes some data that
1572 happens to be on the stack as its last argument, and the callback
1573 just ignores this garbage); but there is no guarantee whatsoever that
1576 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
1577 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
1578 problems not only on Windows, but also on some Unix platforms.
1579 To avoid problematic command lines, these definitions are now in an
1580 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
1581 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
1584 *) MIPS III/IV assembler module is reimplemented.
1587 *) More DES library cleanups: remove references to srand/rand and
1588 delete an unused file.
1591 *) Add support for the the free Netwide assembler (NASM) under Win32,
1592 since not many people have MASM (ml) and it can be hard to obtain.
1593 This is currently experimental but it seems to work OK and pass all
1594 the tests. Check out INSTALL.W32 for info.
1597 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
1598 without temporary keys kept an extra copy of the server key,
1599 and connections with temporary keys did not free everything in case
1603 *) New function RSA_check_key and new openssl rsa option -check
1604 for verifying the consistency of RSA keys.
1605 [Ulf Moeller, Bodo Moeller]
1607 *) Various changes to make Win32 compile work:
1608 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
1609 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
1610 comparison" warnings.
1611 3. Add sk_<TYPE>_sort to DEF file generator and do make update.
1614 *) Add a debugging option to PKCS#5 v2 key generation function: when
1615 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
1616 derived keys are printed to stderr.
1619 *) Copy the flags in ASN1_STRING_dup().
1620 [Roman E. Pavlov <pre@mo.msk.ru>]
1622 *) The x509 application mishandled signing requests containing DSA
1623 keys when the signing key was also DSA and the parameters didn't match.
1625 It was supposed to omit the parameters when they matched the signing key:
1626 the verifying software was then supposed to automatically use the CA's
1627 parameters if they were absent from the end user certificate.
1629 Omitting parameters is no longer recommended. The test was also
1630 the wrong way round! This was probably due to unusual behaviour in
1631 EVP_cmp_parameters() which returns 1 if the parameters match.
1632 This meant that parameters were omitted when they *didn't* match and
1633 the certificate was useless. Certificates signed with 'ca' didn't have
1635 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>]
1637 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
1638 The interface is as follows:
1639 Applications can use
1640 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
1641 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
1642 "off" is now the default.
1643 The library internally uses
1644 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
1645 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
1646 to disable memory-checking temporarily.
1648 Some inconsistent states that previously were possible (and were
1649 even the default) are now avoided.
1651 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
1652 with each memory chunk allocated; this is occasionally more helpful
1653 than just having a counter.
1655 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
1657 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
1661 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX),
1662 which largely parallels "options", but is for changing API behaviour,
1663 whereas "options" are about protocol behaviour.
1664 Initial "mode" flags are:
1666 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
1667 a single record has been written.
1668 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
1669 retries use the same buffer location.
1670 (But all of the contents must be
1674 *) Bugfix: SSL_set_mode ignored its parameter, only SSL_CTX_set_mode
1677 *) Fix problems with no-hmac etc.
1678 [Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>]
1680 *) New functions RSA_get_default_method(), RSA_set_method() and
1681 RSA_get_method(). These allows replacement of RSA_METHODs without having
1682 to mess around with the internals of an RSA structure.
1685 *) Fix memory leaks in DSA_do_sign and DSA_is_prime.
1686 Also really enable memory leak checks in openssl.c and in some
1688 [Chad C. Mulligan, Bodo Moeller]
1690 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
1691 up the length of negative integers. This has now been simplified to just
1692 store the length when it is first determined and use it later, rather
1693 than trying to keep track of where data is copied and updating it to
1695 [Steve Henson, reported by Brien Wheeler
1696 <bwheeler@authentica-security.com>]
1698 *) Add a new function PKCS7_signatureVerify. This allows the verification
1699 of a PKCS#7 signature but with the signing certificate passed to the
1700 function itself. This contrasts with PKCS7_dataVerify which assumes the
1701 certificate is present in the PKCS#7 structure. This isn't always the
1702 case: certificates can be omitted from a PKCS#7 structure and be
1703 distributed by "out of band" means (such as a certificate database).
1706 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the
1707 function prototypes in pem.h, also change util/mkdef.pl to add the
1708 necessary function names.
1711 *) mk1mf.pl (used by Windows builds) did not properly read the
1712 options set by Configure in the top level Makefile, and Configure
1713 was not even able to write more than one option correctly.
1714 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
1717 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config
1718 file to be loaded from a BIO or FILE pointer. The BIO version will
1719 for example allow memory BIOs to contain config info.
1722 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
1723 Whoever hopes to achieve shared-library compatibility across versions
1724 must use this, not the compile-time macro.
1725 (Exercise 0.9.4: Which is the minimum library version required by
1727 Note: All this applies only to multi-threaded programs, others don't
1731 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests
1732 through a BIO pair triggered the default case, i.e.
1733 SSLerr(...,SSL_R_UNKNOWN_STATE).
1736 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
1737 can use the SSL library even if none of the specific BIOs is
1741 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
1742 for the encoded length.
1743 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>]
1745 *) Add initial documentation of the X509V3 functions.
1748 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and
1749 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
1750 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
1751 secure PKCS#8 private key format with a high iteration count.
1754 *) Fix determination of Perl interpreter: A perl or perl5
1755 _directory_ in $PATH was also accepted as the interpreter.
1756 [Ralf S. Engelschall]
1758 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking
1759 wrong with it but it was very old and did things like calling
1760 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
1764 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed
1765 to use the new extension code.
1768 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
1769 with macros. This should make it easier to change their form, add extra
1770 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
1774 *) Add to configuration table a new entry that can specify an alternative
1775 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
1776 according to Mark Crispin <MRC@Panda.COM>.
1780 *) DES CBC did not update the IV. Weird.
1783 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
1784 Changing the behaviour of the former might break existing programs --
1785 where IV updating is needed, des_ncbc_encrypt can be used.
1788 *) When bntest is run from "make test" it drives bc to check its
1789 calculations, as well as internally checking them. If an internal check
1790 fails, it needs to cause bc to give a non-zero result or make test carries
1791 on without noticing the failure. Fixed.
1794 *) DES library cleanups.
1797 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
1798 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
1799 ciphers. NOTE: although the key derivation function has been verified
1800 against some published test vectors it has not been extensively tested
1801 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
1805 *) Instead of "mkdir -p", which is not fully portable, use new
1806 Perl script "util/mkdir-p.pl".
1809 *) Rewrite the way password based encryption (PBE) is handled. It used to
1810 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
1811 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
1812 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
1813 the 'parameter' field of the AlgorithmIdentifier is passed to the
1814 underlying key generation function so it must do its own ASN1 parsing.
1815 This has also changed the EVP_PBE_CipherInit() function which now has a
1816 'parameter' argument instead of literal salt and iteration count values
1817 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
1820 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms
1821 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
1822 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
1823 KEY" because this clashed with PKCS#8 unencrypted string. Since this
1824 value was just used as a "magic string" and not used directly its
1825 value doesn't matter.
1828 *) Introduce some semblance of const correctness to BN. Shame C doesn't
1832 *) "linux-sparc64" configuration (ultrapenguin).
1833 [Ray Miller <ray.miller@oucs.ox.ac.uk>]
1834 "linux-sparc" configuration.
1835 [Christian Forster <fo@hawo.stw.uni-erlangen.de>]
1837 *) config now generates no-xxx options for missing ciphers.
1840 *) Support the EBCDIC character set (work in progress).
1841 File ebcdic.c not yet included because it has a different license.
1842 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
1844 *) Support BS2000/OSD-POSIX.
1845 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
1847 *) Make callbacks for key generation use void * instead of char *.
1850 *) Make S/MIME samples compile (not yet tested).
1853 *) Additional typesafe stacks.
1856 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
1860 Changes between 0.9.3 and 0.9.3a [29 May 1999]
1862 *) New configuration variant "sco5-gcc".
1864 *) Updated some demos.
1865 [Sean O Riordain, Wade Scholine]
1867 *) Add missing BIO_free at exit of pkcs12 application.
1870 *) Fix memory leak in conf.c.
1873 *) Updates for Win32 to assembler version of MD5.
1876 *) Set #! path to perl in apps/der_chop to where we found it
1877 instead of using a fixed path.
1880 *) SHA library changes for irix64-mips4-cc.
1883 *) Improvements for VMS support.
1887 Changes between 0.9.2b and 0.9.3 [24 May 1999]
1889 *) Bignum library bug fix. IRIX 6 passes "make test" now!
1890 This also avoids the problems with SC4.2 and unpatched SC5.
1891 [Andy Polyakov <appro@fy.chalmers.se>]
1893 *) New functions sk_num, sk_value and sk_set to replace the previous macros.
1894 These are required because of the typesafe stack would otherwise break
1895 existing code. If old code used a structure member which used to be STACK
1896 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
1897 sk_num or sk_value it would produce an error because the num, data members
1898 are not present in STACK_OF. Now it just produces a warning. sk_set
1899 replaces the old method of assigning a value to sk_value
1900 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
1901 that does this will no longer work (and should use sk_set instead) but
1902 this could be regarded as a "questionable" behaviour anyway.
1905 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now
1906 correctly handle encrypted S/MIME data.
1909 *) Change type of various DES function arguments from des_cblock
1910 (which means, in function argument declarations, pointer to char)
1911 to des_cblock * (meaning pointer to array with 8 char elements),
1912 which allows the compiler to do more typechecking; it was like
1913 that back in SSLeay, but with lots of ugly casts.
1915 Introduce new type const_des_cblock.
1918 *) Reorganise the PKCS#7 library and get rid of some of the more obvious
1919 problems: find RecipientInfo structure that matches recipient certificate
1920 and initialise the ASN1 structures properly based on passed cipher.
1923 *) Belatedly make the BN tests actually check the results.
1926 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
1927 to and from BNs: it was completely broken. New compilation option
1928 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
1929 key elements as negative integers.
1932 *) Reorganize and speed up MD5.
1933 [Andy Polyakov <appro@fy.chalmers.se>]
1936 [Richard Levitte <richard@levitte.org>]
1938 *) New option -out to asn1parse to allow the parsed structure to be
1939 output to a file. This is most useful when combined with the -strparse
1940 option to examine the output of things like OCTET STRINGS.
1943 *) Make SSL library a little more fool-proof by not requiring any longer
1944 that SSL_set_{accept,connect}_state be called before
1945 SSL_{accept,connect} may be used (SSL_set_..._state is omitted
1946 in many applications because usually everything *appeared* to work as
1947 intended anyway -- now it really works as intended).
1950 *) Move openssl.cnf out of lib/.
1953 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall
1954 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
1955 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+
1956 [Ralf S. Engelschall]
1958 *) Various fixes to the EVP and PKCS#7 code. It may now be able to
1959 handle PKCS#7 enveloped data properly.
1960 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve]
1962 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
1963 copying pointers. The cert_st handling is changed by this in
1964 various ways (and thus what used to be known as ctx->default_cert
1965 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert
1966 any longer when s->cert does not give us what we need).
1967 ssl_cert_instantiate becomes obsolete by this change.
1968 As soon as we've got the new code right (possibly it already is?),
1969 we have solved a couple of bugs of the earlier code where s->cert
1970 was used as if it could not have been shared with other SSL structures.
1972 Note that using the SSL API in certain dirty ways now will result
1973 in different behaviour than observed with earlier library versions:
1974 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx)
1975 does not influence s as it used to.
1977 In order to clean up things more thoroughly, inside SSL_SESSION
1978 we don't use CERT any longer, but a new structure SESS_CERT
1979 that holds per-session data (if available); currently, this is
1980 the peer's certificate chain and, for clients, the server's certificate
1981 and temporary key. CERT holds only those values that can have
1982 meaningful defaults in an SSL_CTX.
1985 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
1986 from the internal representation. Various PKCS#7 fixes: remove some
1987 evil casts and set the enc_dig_alg field properly based on the signing
1991 *) Allow PKCS#12 password to be set from the command line or the
1992 environment. Let 'ca' get its config file name from the environment
1993 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
1997 *) Allow certificate policies extension to use an IA5STRING for the
1998 organization field. This is contrary to the PKIX definition but
1999 VeriSign uses it and IE5 only recognises this form. Document 'x509'
2003 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
2004 without disallowing inline assembler and the like for non-pedantic builds.
2007 *) Support Borland C++ builder.
2008 [Janez Jere <jj@void.si>, modified by Ulf Möller]
2013 *) SHA-1 cleanups and performance enhancements.
2014 [Andy Polyakov <appro@fy.chalmers.se>]
2016 *) Sparc v8plus assembler for the bignum library.
2017 [Andy Polyakov <appro@fy.chalmers.se>]
2019 *) Accept any -xxx and +xxx compiler options in Configure.
2022 *) Update HPUX configuration.
2025 *) Add missing sk_<type>_unshift() function to safestack.h
2026 [Ralf S. Engelschall]
2028 *) New function SSL_CTX_use_certificate_chain_file that sets the
2029 "extra_cert"s in addition to the certificate. (This makes sense
2030 only for "PEM" format files, as chains as a whole are not
2034 *) Support verify_depth from the SSL API.
2035 x509_vfy.c had what can be considered an off-by-one-error:
2036 Its depth (which was not part of the external interface)
2037 was actually counting the number of certificates in a chain;
2038 now it really counts the depth.
2041 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
2042 instead of X509err, which often resulted in confusing error
2043 messages since the error codes are not globally unique
2044 (e.g. an alleged error in ssl3_accept when a certificate
2045 didn't match the private key).
2047 *) New function SSL_CTX_set_session_id_context that allows to set a default
2048 value (so that you don't need SSL_set_session_id_context for each
2049 connection using the SSL_CTX).
2052 *) OAEP decoding bug fix.
2055 *) Support INSTALL_PREFIX for package builders, as proposed by
2059 *) New Configure options "threads" and "no-threads". For systems
2060 where the proper compiler options are known (currently Solaris
2061 and Linux), "threads" is the default.
2064 *) New script util/mklink.pl as a faster substitute for util/mklink.sh.
2067 *) Install various scripts to $(OPENSSLDIR)/misc, not to
2068 $(INSTALLTOP)/bin -- they shouldn't clutter directories
2069 such as /usr/local/bin.
2072 *) "make linux-shared" to build shared libraries.
2073 [Niels Poppe <niels@netbox.org>]
2075 *) New Configure option no-<cipher> (rsa, idea, rc5, ...).
2078 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
2079 extension adding in x509 utility.
2082 *) Remove NOPROTO sections and error code comments.
2085 *) Partial rewrite of the DEF file generator to now parse the ANSI
2089 *) New Configure options --prefix=DIR and --openssldir=DIR.
2092 *) Complete rewrite of the error code script(s). It is all now handled
2093 by one script at the top level which handles error code gathering,
2094 header rewriting and C source file generation. It should be much better
2095 than the old method: it now uses a modified version of Ulf's parser to
2096 read the ANSI prototypes in all header files (thus the old K&R definitions
2097 aren't needed for error creation any more) and do a better job of
2098 translating function codes into names. The old 'ASN1 error code imbedded
2099 in a comment' is no longer necessary and it doesn't use .err files which
2100 have now been deleted. Also the error code call doesn't have to appear all
2101 on one line (which resulted in some large lines...).
2104 *) Change #include filenames from <foo.h> to <openssl/foo.h>.
2107 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return
2108 0 (which usually indicates a closed connection), but continue reading.
2111 *) Fix some race conditions.
2114 *) Add support for CRL distribution points extension. Add Certificate
2115 Policies and CRL distribution points documentation.
2118 *) Move the autogenerated header file parts to crypto/opensslconf.h.
2121 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
2122 8 of keying material. Merlin has also confirmed interop with this fix
2123 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
2124 [Merlin Hughes <merlin@baltimore.ie>]
2126 *) Fix lots of warnings.
2127 [Richard Levitte <levitte@stacken.kth.se>]
2129 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
2130 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
2131 [Richard Levitte <levitte@stacken.kth.se>]
2133 *) Fix problems with sizeof(long) == 8.
2134 [Andy Polyakov <appro@fy.chalmers.se>]
2136 *) Change functions to ANSI C.
2139 *) Fix typos in error codes.
2140 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller]
2142 *) Remove defunct assembler files from Configure.
2145 *) SPARC v8 assembler BIGNUM implementation.
2146 [Andy Polyakov <appro@fy.chalmers.se>]
2148 *) Support for Certificate Policies extension: both print and set.
2149 Various additions to support the r2i method this uses.
2152 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could
2153 return a const string when you are expecting an allocated buffer.
2156 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
2157 types DirectoryString and DisplayText.
2160 *) Add code to allow r2i extensions to access the configuration database,
2161 add an LHASH database driver and add several ctx helper functions.
2164 *) Fix an evil bug in bn_expand2() which caused various BN functions to
2165 fail when they extended the size of a BIGNUM.
2168 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to
2169 support typesafe stack.
2172 *) Fix typo in SSL_[gs]et_options().
2173 [Nils Frostberg <nils@medcom.se>]
2175 *) Delete various functions and files that belonged to the (now obsolete)
2176 old X509V3 handling code.
2179 *) New Configure option "rsaref".
2182 *) Don't auto-generate pem.h.
2185 *) Introduce type-safe ASN.1 SETs.
2188 *) Convert various additional casted stacks to type-safe STACK_OF() variants.
2189 [Ben Laurie, Ralf S. Engelschall, Steve Henson]
2191 *) Introduce type-safe STACKs. This will almost certainly break lots of code
2192 that links with OpenSSL (well at least cause lots of warnings), but fear
2193 not: the conversion is trivial, and it eliminates loads of evil casts. A
2194 few STACKed things have been converted already. Feel free to convert more.
2195 In the fullness of time, I'll do away with the STACK type altogether.
2198 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate
2199 specified in <certfile> by updating the entry in the index.txt file.
2200 This way one no longer has to edit the index.txt file manually for
2201 revoking a certificate. The -revoke option does the gory details now.
2202 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall]
2204 *) Fix `openssl crl -noout -text' combination where `-noout' killed the
2205 `-text' option at all and this way the `-noout -text' combination was
2206 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'.
2207 [Ralf S. Engelschall]
2209 *) Make sure a corresponding plain text error message exists for the
2210 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
2211 verify callback function determined that a certificate was revoked.
2212 [Ralf S. Engelschall]
2214 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for
2215 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
2216 all available cipers including rc5, which was forgotten until now.
2217 In order to let the testing shell script know which algorithms
2218 are available, a new (up to now undocumented) command
2219 "openssl list-cipher-commands" is used.
2222 *) Bugfix: s_client occasionally would sleep in select() when
2223 it should have checked SSL_pending() first.
2226 *) New functions DSA_do_sign and DSA_do_verify to provide access to
2227 the raw DSA values prior to ASN.1 encoding.
2230 *) Tweaks to Configure
2231 [Niels Poppe <niels@netbox.org>]
2233 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
2237 *) New variables $(RANLIB) and $(PERL) in the Makefiles.
2240 *) New config option to avoid instructions that are illegal on the 80386.
2241 The default code is faster, but requires at least a 486.
2244 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
2245 SSL2_SERVER_VERSION (not used at all) macros, which are now the
2246 same as SSL2_VERSION anyway.
2249 *) New "-showcerts" option for s_client.
2252 *) Still more PKCS#12 integration. Add pkcs12 application to openssl
2253 application. Various cleanups and fixes.
2256 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
2257 modify error routines to work internally. Add error codes and PBE init
2258 to library startup routines.
2261 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and
2262 packing functions to asn1 and evp. Changed function names and error
2263 codes along the way.
2266 *) PKCS12 integration: and so it begins... First of several patches to
2267 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
2268 objects to objects.h
2271 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1
2272 and display support for Thawte strong extranet extension.
2275 *) Add LinuxPPC support.
2276 [Jeff Dubrule <igor@pobox.org>]
2278 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
2279 bn_div_words in alpha.s.
2280 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie]
2282 *) Make sure the RSA OAEP test is skipped under -DRSAref because
2283 OAEP isn't supported when OpenSSL is built with RSAref.
2284 [Ulf Moeller <ulf@fitug.de>]
2286 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
2287 so they no longer are missing under -DNOPROTO.
2288 [Soren S. Jorvang <soren@t.dk>]
2291 Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
2293 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
2294 doesn't work when the session is reused. Coming soon!
2297 *) Fix a security hole, that allows sessions to be reused in the wrong
2298 context thus bypassing client cert protection! All software that uses
2299 client certs and session caches in multiple contexts NEEDS PATCHING to
2300 allow session reuse! A fuller solution is in the works.
2301 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)]
2303 *) Some more source tree cleanups (removed obsolete files
2304 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
2305 permission on "config" script to be executable) and a fix for the INSTALL
2307 [Ulf Moeller <ulf@fitug.de>]
2309 *) Remove some legacy and erroneous uses of malloc, free instead of
2311 [Lennart Bang <lob@netstream.se>, with minor changes by Steve]
2313 *) Make rsa_oaep_test return non-zero on error.
2314 [Ulf Moeller <ulf@fitug.de>]
2316 *) Add support for native Solaris shared libraries. Configure
2317 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
2318 if someone would make that last step automatic.
2319 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>]
2321 *) ctx_size was not built with the right compiler during "make links". Fixed.
2324 *) Change the meaning of 'ALL' in the cipher list. It now means "everything
2325 except NULL ciphers". This means the default cipher list will no longer
2326 enable NULL ciphers. They need to be specifically enabled e.g. with
2327 the string "DEFAULT:eNULL".
2330 *) Fix to RSA private encryption routines: if p < q then it would
2331 occasionally produce an invalid result. This will only happen with
2332 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
2335 *) Be less restrictive and allow also `perl util/perlpath.pl
2336 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin',
2337 because this way one can also use an interpreter named `perl5' (which is
2338 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
2339 installed as `perl').
2340 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2342 *) Let util/clean-depend.pl work also with older Perl 5.00x versions.
2343 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2345 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
2346 advapi32.lib to Win32 build and change the pem test comparision
2347 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
2348 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
2349 and crypto/des/ede_cbcm_enc.c.
2352 *) DES quad checksum was broken on big-endian architectures. Fixed.
2355 *) Comment out two functions in bio.h that aren't implemented. Fix up the
2356 Win32 test batch file so it (might) work again. The Win32 test batch file
2357 is horrible: I feel ill....
2360 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
2361 in e_os.h. Audit of header files to check ANSI and non ANSI
2362 sections: 10 functions were absent from non ANSI section and not exported
2363 from Windows DLLs. Fixed up libeay.num for new functions.
2366 *) Make `openssl version' output lines consistent.
2367 [Ralf S. Engelschall]
2369 *) Fix Win32 symbol export lists for BIO functions: Added
2370 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
2371 to ms/libeay{16,32}.def.
2372 [Ralf S. Engelschall]
2374 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
2375 fine under Unix and passes some trivial tests I've now added. But the
2376 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
2377 added to make sure no one expects that this stuff really works in the
2378 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
2379 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
2381 [Ralf S. Engelschall]
2383 *) Fix the generation of two part addresses in perl.
2384 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie]
2386 *) Add config entry for Linux on MIPS.
2387 [John Tobey <jtobey@channel1.com>]
2389 *) Make links whenever Configure is run, unless we are on Windoze.
2392 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf.
2393 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
2397 *) Add a useful kludge to allow package maintainers to specify compiler and
2398 other platforms details on the command line without having to patch the
2399 Configure script everytime: One now can use ``perl Configure
2400 <id>:<details>'', i.e. platform ids are allowed to have details appended
2401 to them (seperated by colons). This is treated as there would be a static
2402 pre-configured entry in Configure's %table under key <id> with value
2403 <details> and ``perl Configure <id>'' is called. So, when you want to
2404 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
2405 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"''
2406 now, which overrides the FreeBSD-elf entry on-the-fly.
2407 [Ralf S. Engelschall]
2409 *) Disable new TLS1 ciphersuites by default: they aren't official yet.
2412 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
2413 on the `perl Configure ...' command line. This way one can compile
2414 OpenSSL libraries with Position Independent Code (PIC) which is needed
2415 for linking it into DSOs.
2416 [Ralf S. Engelschall]
2418 *) Remarkably, export ciphers were totally broken and no-one had noticed!
2422 *) Cleaned up the LICENSE document: The official contact for any license
2423 questions now is the OpenSSL core team under openssl-core@openssl.org.
2424 And add a paragraph about the dual-license situation to make sure people
2425 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
2426 to the OpenSSL toolkit.
2427 [Ralf S. Engelschall]
2429 *) General source tree makefile cleanups: Made `making xxx in yyy...'
2430 display consistent in the source tree and replaced `/bin/rm' by `rm'.
2431 Additonally cleaned up the `make links' target: Remove unnecessary
2432 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
2433 to speed processing and no longer clutter the display with confusing
2434 stuff. Instead only the actually done links are displayed.
2435 [Ralf S. Engelschall]
2437 *) Permit null encryption ciphersuites, used for authentication only. It used
2438 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
2439 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
2443 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
2444 signed attributes when verifying signatures (this would break them),
2445 the detached data encoding was wrong and public keys obtained using
2446 X509_get_pubkey() weren't freed.
2449 *) Add text documentation for the BUFFER functions. Also added a work around
2450 to a Win95 console bug. This was triggered by the password read stuff: the
2451 last character typed gets carried over to the next fread(). If you were
2452 generating a new cert request using 'req' for example then the last
2453 character of the passphrase would be CR which would then enter the first
2457 *) Added the new `Includes OpenSSL Cryptography Software' button as
2458 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
2459 button and can be used by applications based on OpenSSL to show the
2460 relationship to the OpenSSL project.
2461 [Ralf S. Engelschall]
2463 *) Remove confusing variables in function signatures in files
2464 ssl/ssl_lib.c and ssl/ssl.h.
2465 [Lennart Bong <lob@kulthea.stacken.kth.se>]
2467 *) Don't install bss_file.c under PREFIX/include/
2468 [Lennart Bong <lob@kulthea.stacken.kth.se>]
2470 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle
2471 functions that return function pointers and has support for NT specific
2472 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
2473 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
2474 unsigned to signed types: this was killing the Win32 compile.
2477 *) Add new certificate file to stack functions,
2478 SSL_add_dir_cert_subjects_to_stack() and
2479 SSL_add_file_cert_subjects_to_stack(). These largely supplant
2480 SSL_load_client_CA_file(), and can be used to add multiple certs easily
2481 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
2482 This means that Apache-SSL and similar packages don't have to mess around
2483 to add as many CAs as they want to the preferred list.
2486 *) Experiment with doxygen documentation. Currently only partially applied to
2488 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with
2489 openssl.doxy as the configuration file.
2492 *) Get rid of remaining C++-style comments which strict C compilers hate.
2493 [Ralf S. Engelschall, pointed out by Carlos Amengual]
2495 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
2496 compiled in by default: it has problems with large keys.
2499 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
2500 DH private keys and/or callback functions which directly correspond to
2501 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
2502 is needed for applications which have to configure certificates on a
2503 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
2505 For the RSA certificate situation is makes no difference, but
2506 for the DSA certificate situation this fixes the "no shared cipher"
2507 problem where the OpenSSL cipher selection procedure failed because the
2508 temporary keys were not overtaken from the context and the API provided
2509 no way to reconfigure them.
2510 The new functions now let applications reconfigure the stuff and they
2511 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
2512 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
2513 non-public-API function ssl_cert_instantiate() is used as a helper
2514 function and also to reduce code redundancy inside ssl_rsa.c.
2515 [Ralf S. Engelschall]
2517 *) Move s_server -dcert and -dkey options out of the undocumented feature
2518 area because they are useful for the DSA situation and should be
2519 recognized by the users.
2520 [Ralf S. Engelschall]
2522 *) Fix the cipher decision scheme for export ciphers: the export bits are
2523 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
2524 SSL_EXP_MASK. So, the original variable has to be used instead of the
2525 already masked variable.
2526 [Richard Levitte <levitte@stacken.kth.se>]
2528 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c
2529 [Richard Levitte <levitte@stacken.kth.se>]
2531 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
2532 from `int' to `unsigned int' because it's a length and initialized by
2533 EVP_DigestFinal() which expects an `unsigned int *'.
2534 [Richard Levitte <levitte@stacken.kth.se>]
2536 *) Don't hard-code path to Perl interpreter on shebang line of Configure
2537 script. Instead use the usual Shell->Perl transition trick.
2538 [Ralf S. Engelschall]
2540 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates
2541 (in addition to RSA certificates) to match the behaviour of `openssl dsa
2542 -noout -modulus' as it's already the case for `openssl rsa -noout
2543 -modulus'. For RSA the -modulus is the real "modulus" while for DSA
2544 currently the public key is printed (a decision which was already done by
2545 `openssl dsa -modulus' in the past) which serves a similar purpose.
2546 Additionally the NO_RSA no longer completely removes the whole -modulus
2547 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
2549 [Ralf S. Engelschall]
2551 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
2552 BIO. See the source (crypto/evp/bio_ok.c) for more info.
2553 [Arne Ansper <arne@ats.cyber.ee>]
2555 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs
2556 to be added. Now both 'req' and 'ca' can use new objects defined in the
2560 *) Add cool BIO that does syslog (or event log on NT).
2561 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie]
2563 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
2564 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
2565 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
2566 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
2569 *) Add preliminary config info for new extension code.
2572 *) Make RSA_NO_PADDING really use no padding.
2573 [Ulf Moeller <ulf@fitug.de>]
2575 *) Generate errors when private/public key check is done.
2578 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
2579 for some CRL extensions and new objects added.
2582 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private
2583 key usage extension and fuller support for authority key id.
2586 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
2587 padding method for RSA, which is recommended for new applications in PKCS
2588 #1 v2.0 (RFC 2437, October 1998).
2589 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
2590 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
2591 against Bleichbacher's attack on RSA.
2592 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
2595 *) Updates to the new SSL compression code
2596 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2598 *) Fix so that the version number in the master secret, when passed
2599 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
2600 (because the server will not accept higher), that the version number
2601 is 0x03,0x01, not 0x03,0x00
2602 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2604 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
2605 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
2606 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
2609 *) Support for RAW extensions where an arbitrary extension can be
2610 created by including its DER encoding. See apps/openssl.cnf for
2614 *) Make sure latest Perl versions don't interpret some generated C array
2615 code as Perl array code in the crypto/err/err_genc.pl script.
2616 [Lars Weber <3weber@informatik.uni-hamburg.de>]
2618 *) Modify ms/do_ms.bat to not generate assembly language makefiles since
2619 not many people have the assembler. Various Win32 compilation fixes and
2620 update to the INSTALL.W32 file with (hopefully) more accurate Win32
2624 *) Modify configure script 'Configure' to automatically create crypto/date.h
2625 file under Win32 and also build pem.h from pem.org. New script
2626 util/mkfiles.pl to create the MINFO file on environments that can't do a
2627 'make files': perl util/mkfiles.pl >MINFO should work.
2630 *) Major rework of DES function declarations, in the pursuit of correctness
2631 and purity. As a result, many evil casts evaporated, and some weirdness,
2632 too. You may find this causes warnings in your code. Zapping your evil
2633 casts will probably fix them. Mostly.
2636 *) Fix for a typo in asn1.h. Bug fix to object creation script
2637 obj_dat.pl. It considered a zero in an object definition to mean
2638 "end of object": none of the objects in objects.h have any zeros
2639 so it wasn't spotted.
2640 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>]
2642 *) Add support for Triple DES Cipher Block Chaining with Output Feedback
2643 Masking (CBCM). In the absence of test vectors, the best I have been able
2644 to do is check that the decrypt undoes the encrypt, so far. Send me test
2645 vectors if you have them.
2648 *) Correct calculation of key length for export ciphers (too much space was
2649 allocated for null ciphers). This has not been tested!
2652 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage
2653 message is now correct (it understands "crypto" and "ssl" on its
2654 command line). There is also now an "update" option. This will update
2655 the util/ssleay.num and util/libeay.num files with any new functions.
2657 perl util/mkdef.pl crypto ssl update
2658 it will update them.
2661 *) Overhauled the Perl interface (perl/*):
2662 - ported BN stuff to OpenSSL's different BN library
2663 - made the perl/ source tree CVS-aware
2664 - renamed the package from SSLeay to OpenSSL (the files still contain
2665 their history because I've copied them in the repository)
2666 - removed obsolete files (the test scripts will be replaced
2667 by better Test::Harness variants in the future)
2668 [Ralf S. Engelschall]
2670 *) First cut for a very conservative source tree cleanup:
2671 1. merge various obsolete readme texts into doc/ssleay.txt
2672 where we collect the old documents and readme texts.
2673 2. remove the first part of files where I'm already sure that we no
2674 longer need them because of three reasons: either they are just temporary
2675 files which were left by Eric or they are preserved original files where
2676 I've verified that the diff is also available in the CVS via "cvs diff
2677 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
2678 the crypto/md/ stuff).
2679 [Ralf S. Engelschall]
2681 *) More extension code. Incomplete support for subject and issuer alt
2682 name, issuer and authority key id. Change the i2v function parameters
2683 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
2684 what that's for :-) Fix to ASN1 macro which messed up
2685 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
2688 *) Preliminary support for ENUMERATED type. This is largely copied from the
2692 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
2693 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2695 *) Make sure `make rehash' target really finds the `openssl' program.
2696 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2698 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
2699 like to hear about it if this slows down other processors.
2702 *) Add CygWin32 platform information to Configure script.
2703 [Alan Batie <batie@aahz.jf.intel.com>]
2705 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm'
2706 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>]
2708 *) New program nseq to manipulate netscape certificate sequences
2711 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
2715 *) Fixes to BN code. Previously the default was to define BN_RECURSION
2716 but the BN code had some problems that would cause failures when
2717 doing certificate verification and some other functions.
2718 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2720 *) Add ASN1 and PEM code to support netscape certificate sequences.
2723 *) Add ASN1 and PEM code to support netscape certificate sequences.
2726 *) Add several PKIX and private extended key usage OIDs.
2729 *) Modify the 'ca' program to handle the new extension code. Modify
2730 openssl.cnf for new extension format, add comments.
2733 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
2734 and add a sample to openssl.cnf so req -x509 now adds appropriate
2738 *) Continued X509 V3 changes. Add to other makefiles, integrate with the
2739 error code, add initial support to X509_print() and x509 application.
2742 *) Takes a deep breath and start addding X509 V3 extension support code. Add
2743 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
2744 stuff is currently isolated and isn't even compiled yet.
2747 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL
2748 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
2749 Removed the versions check from X509 routines when loading extensions:
2750 this allows certain broken certificates that don't set the version
2751 properly to be processed.
2754 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
2755 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
2756 can still be regenerated with "make depend".
2759 *) Spelling mistake in C version of CAST-128.
2760 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>]
2762 *) Changes to the error generation code. The perl script err-code.pl
2763 now reads in the old error codes and retains the old numbers, only
2764 adding new ones if necessary. It also only changes the .err files if new
2765 codes are added. The makefiles have been modified to only insert errors
2766 when needed (to avoid needlessly modifying header files). This is done
2767 by only inserting errors if the .err file is newer than the auto generated
2768 C file. To rebuild all the error codes from scratch (the old behaviour)
2769 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
2770 or delete all the .err files.
2773 *) CAST-128 was incorrectly implemented for short keys. The C version has
2774 been fixed, but is untested. The assembler versions are also fixed, but
2775 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
2776 to regenerate it if needed.
2777 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
2778 Hagino <itojun@kame.net>]
2780 *) File was opened incorrectly in randfile.c.
2781 [Ulf Möller <ulf@fitug.de>]
2783 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print
2784 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
2785 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
2786 al: it's just almost always a UTCTime. Note this patch adds new error
2787 codes so do a "make errors" if there are problems.
2790 *) Correct Linux 1 recognition in config.
2791 [Ulf Möller <ulf@fitug.de>]
2793 *) Remove pointless MD5 hash when using DSA keys in ca.
2794 [Anonymous <nobody@replay.com>]
2796 *) Generate an error if given an empty string as a cert directory. Also
2797 generate an error if handed NULL (previously returned 0 to indicate an
2798 error, but didn't set one).
2799 [Ben Laurie, reported by Anonymous <nobody@replay.com>]
2801 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
2804 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
2805 parameters. This was causing a warning which killed off the Win32 compile.
2808 *) Remove C++ style comments from crypto/bn/bn_local.h.
2809 [Neil Costigan <neil.costigan@celocom.com>]
2811 *) The function OBJ_txt2nid was broken. It was supposed to return a nid
2812 based on a text string, looking up short and long names and finally
2813 "dot" format. The "dot" format stuff didn't work. Added new function
2814 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
2815 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
2816 OID is not part of the table.
2819 *) Add prototypes to X509 lookup/verify methods, fixing a bug in
2820 X509_LOOKUP_by_alias().
2823 *) Sort openssl functions by name.
2826 *) Get the gendsa program working (hopefully) and add it to app list. Remove
2827 encryption from sample DSA keys (in case anyone is interested the password
2831 *) Make _all_ *_free functions accept a NULL pointer.
2832 [Frans Heymans <fheymans@isaserver.be>]
2834 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use
2836 [Anonymous <nobody@replay.com>]
2838 *) s_server should send the CAfile as acceptable CAs, not its own cert.
2839 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
2841 *) Don't blow it for numeric -newkey arguments to apps/req.
2842 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
2844 *) Temp key "for export" tests were wrong in s3_srvr.c.
2845 [Anonymous <nobody@replay.com>]
2847 *) Add prototype for temp key callback functions
2848 SSL_CTX_set_tmp_{rsa,dh}_callback().
2851 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
2852 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
2855 *) X509_name_add_entry() freed the wrong thing after an error.
2856 [Arne Ansper <arne@ats.cyber.ee>]
2858 *) rsa_eay.c would attempt to free a NULL context.
2859 [Arne Ansper <arne@ats.cyber.ee>]
2861 *) BIO_s_socket() had a broken should_retry() on Windoze.
2862 [Arne Ansper <arne@ats.cyber.ee>]
2864 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
2865 [Arne Ansper <arne@ats.cyber.ee>]
2867 *) Make sure the already existing X509_STORE->depth variable is initialized
2868 in X509_STORE_new(), but document the fact that this variable is still
2869 unused in the certificate verification process.
2870 [Ralf S. Engelschall]
2872 *) Fix the various library and apps files to free up pkeys obtained from
2873 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
2876 *) Fix reference counting in X509_PUBKEY_get(). This makes
2877 demos/maurice/example2.c work, amongst others, probably.
2878 [Steve Henson and Ben Laurie]
2880 *) First cut of a cleanup for apps/. First the `ssleay' program is now named
2881 `openssl' and second, the shortcut symlinks for the `openssl <command>'
2882 are no longer created. This way we have a single and consistent command
2883 line interface `openssl <command>', similar to `cvs <command>'.
2884 [Ralf S. Engelschall, Paul Sutton and Ben Laurie]
2886 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
2887 BIT STRING wrapper always have zero unused bits.
2890 *) Add CA.pl, perl version of CA.sh, add extended key usage OID.
2893 *) Make the top-level INSTALL documentation easier to understand.
2896 *) Makefiles updated to exit if an error occurs in a sub-directory
2897 make (including if user presses ^C) [Paul Sutton]
2899 *) Make Montgomery context stuff explicit in RSA data structure.
2902 *) Fix build order of pem and err to allow for generated pem.h.
2905 *) Fix renumbering bug in X509_NAME_delete_entry().
2908 *) Enhanced the err-ins.pl script so it makes the error library number
2909 global and can add a library name. This is needed for external ASN1 and
2910 other error libraries.
2913 *) Fixed sk_insert which never worked properly.
2916 *) Fix ASN1 macros so they can handle indefinite length construted
2917 EXPLICIT tags. Some non standard certificates use these: they can now
2921 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
2922 into a single doc/ssleay.txt bundle. This way the information is still
2923 preserved but no longer messes up this directory. Now it's new room for
2924 the new set of documenation files.
2925 [Ralf S. Engelschall]
2927 *) SETs were incorrectly DER encoded. This was a major pain, because they
2928 shared code with SEQUENCEs, which aren't coded the same. This means that
2929 almost everything to do with SETs or SEQUENCEs has either changed name or
2930 number of arguments.
2931 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>]
2933 *) Fix test data to work with the above.
2936 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
2937 was already fixed by Eric for 0.9.1 it seems.
2938 [Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>]
2940 *) Autodetect FreeBSD3.
2943 *) Fix various bugs in Configure. This affects the following platforms:
2947 unixware-2.0-pentium
2951 *) Eliminate generated files from CVS. Reorder tests to regenerate files
2952 before they are needed.
2955 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy).
2959 Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
2961 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
2962 changed SSLeay to OpenSSL in version strings.
2963 [Ralf S. Engelschall]
2965 *) Some fixups to the top-level documents.
2968 *) Fixed the nasty bug where rsaref.h was not found under compile-time
2969 because the symlink to include/ was missing.
2970 [Ralf S. Engelschall]
2972 *) Incorporated the popular no-RSA/DSA-only patches
2973 which allow to compile a RSA-free SSLeay.
2974 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall]
2976 *) Fixed nasty rehash problem under `make -f Makefile.ssl links'
2977 when "ssleay" is still not found.
2978 [Ralf S. Engelschall]
2980 *) Added more platforms to Configure: Cray T3E, HPUX 11,
2981 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>]
2983 *) Updated the README file.
2984 [Ralf S. Engelschall]
2986 *) Added various .cvsignore files in the CVS repository subdirs
2987 to make a "cvs update" really silent.
2988 [Ralf S. Engelschall]
2990 *) Recompiled the error-definition header files and added
2991 missing symbols to the Win32 linker tables.
2992 [Ralf S. Engelschall]
2994 *) Cleaned up the top-level documents;
2995 o new files: CHANGES and LICENSE
2996 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
2997 o merged COPYRIGHT into LICENSE
2998 o removed obsolete TODO file
2999 o renamed MICROSOFT to INSTALL.W32
3000 [Ralf S. Engelschall]
3002 *) Removed dummy files from the 0.9.1b source tree:
3003 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
3004 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
3005 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
3006 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
3007 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
3008 [Ralf S. Engelschall]
3010 *) Added various platform portability fixes.
3013 *) The Genesis of the OpenSSL rpject:
3014 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
3015 Young and Tim J. Hudson created while they were working for C2Net until
3017 [The OpenSSL Project]
3020 Changes between 0.9.0b and 0.9.1b [not released]
3022 *) Updated a few CA certificates under certs/
3025 *) Changed some BIGNUM api stuff.
3028 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
3029 DGUX x86, Linux Alpha, etc.
3032 *) New COMP library [crypto/comp/] for SSL Record Layer Compression:
3033 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
3037 *) Add -strparse option to asn1pars program which parses nested
3039 [Dr Stephen Henson <shenson@bigfoot.com>]
3041 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
3044 *) DSA fix for "ca" program.
3047 *) Added "-genkey" option to "dsaparam" program.
3050 *) Added RIPE MD160 (rmd160) message digest.
3053 *) Added -a (all) option to "ssleay version" command.
3056 *) Added PLATFORM define which is the id given to Configure.
3059 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
3062 *) Extended the ASN.1 parser routines.
3065 *) Extended BIO routines to support REUSEADDR, seek, tell, etc.
3068 *) Added a BN_CTX to the BN library.
3071 *) Fixed the weak key values in DES library
3074 *) Changed API in EVP library for cipher aliases.
3077 *) Added support for RC2/64bit cipher.
3080 *) Converted the lhash library to the crypto/mem.c functions.
3083 *) Added more recognized ASN.1 object ids.
3086 *) Added more RSA padding checks for SSL/TLS.
3089 *) Added BIO proxy/filter functionality.
3092 *) Added extra_certs to SSL_CTX which can be used
3093 send extra CA certificates to the client in the CA cert chain sending
3094 process. It can be configured with SSL_CTX_add_extra_chain_cert().
3097 *) Now Fortezza is denied in the authentication phase because
3098 this is key exchange mechanism is not supported by SSLeay at all.
3101 *) Additional PKCS1 checks.
3104 *) Support the string "TLSv1" for all TLS v1 ciphers.
3107 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
3108 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
3111 *) Fixed a few memory leaks.
3114 *) Fixed various code and comment typos.
3117 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0
3118 bytes sent in the client random.
3119 [Edward Bishop <ebishop@spyglass.com>]