5 Changes between 0.9.5a and 0.9.6 [xx XXX 2000]
7 *) A first attempt at creating official support for shared
8 libraries through configuration. I've kept it so the
9 default is static libraries only, and the OpenSSL programs
10 are always statically linked for now, but there are
11 preparations for dynamic linking in place.
12 This has been tested on Linux and True64.
15 *) Randomness polling function for Win9x, as described in:
16 Peter Gutmann, Software Generation of Practically Strong
20 *) Fix so PRNG is seeded in req if using an already existing
24 *) New options to smime application. -inform and -outform
25 allow alternative formats for the S/MIME message including
26 PEM and DER. The -content option allows the content to be
27 specified separately. This should allow things like Netscape
28 form signing output easier to verify.
31 *) Fix the ASN1 encoding of tags using the 'long form'.
34 *) New ASN1 functions, i2c_* and c2i_* for INTEGER and BIT
35 STRING types. These convert content octets to and from the
36 underlying type. The actual tag and length octets are
37 already assumed to have been read in and checked. These
38 are needed because all other string types have virtually
39 identical handling apart from the tag. By having versions
40 of the ASN1 functions that just operate on content octets
41 IMPLICIT tagging can be handled properly. It also allows
42 the ASN1_ENUMERATED code to be cut down because ASN1_ENUMERATED
43 and ASN1_INTEGER are identical apart from the tag.
46 *) Change the handling of OID objects as follows:
48 - New object identifiers are inserted in objects.txt, following
49 the syntax given in objects.README.
50 - objects.pl is used to process obj_mac.num and create a new
52 - obj_dat.pl is used to create a new obj_dat.h, using the data in
55 This is currently kind of a hack, and the perl code in objects.pl
56 isn't very elegant, but it works as I intended. The simplest way
57 to check that it worked correctly is to look in obj_dat.h and
58 check the array nid_objs and make sure the objects haven't moved
59 around (this is important!). Additions are OK, as well as
60 consistent name changes.
63 *) Add BSD-style MD5-based passwords to 'openssl passwd' (option '-1').
66 *) Addition of the command line parameter '-rand file' to 'openssl req'.
67 The given file adds to whatever has already been seeded into the
68 random pool through the RANDFILE configuration file option or
69 environment variable, or the default random state file.
72 *) mkstack.pl now sorts each macro group into lexical order.
73 Previously the output order depended on the order the files
74 appeared in the directory, resulting in needless rewriting
78 *) Patches to make OpenSSL compile under Win32 again. Mostly
79 work arounds for the VC++ problem that it treats func() as
80 func(void). Also stripped out the parts of mkdef.pl that
81 added extra typesafe functions: these no longer exist.
84 *) Reorganisation of the stack code. The macros are now all
85 collected in safestack.h . Each macro is defined in terms of
86 a "stack macro" of the form SKM_<name>(type, a, b). The
87 DEBUG_SAFESTACK is now handled in terms of function casts,
88 this has the advantage of retaining type safety without the
89 use of additional functions. If DEBUG_SAFESTACK is not defined
90 then the non typesafe macros are used instead. Also modified the
91 mkstack.pl script to handle the new form. Needs testing to see
92 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK
93 the default if no major problems. Similar behaviour for ASN1_SET_OF
97 *) When some versions of IIS use the 'NET' form of private key the
98 key derivation algorithm is different. Normally MD5(password) is
99 used as a 128 bit RC4 key. In the modified case
100 MD5(MD5(password) + "SGCKEYSALT") is used insted. Added some
101 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same
102 as the old Netscape_RSA functions except they have an additional
103 'sgckey' parameter which uses the modified algorithm. Also added
104 an -sgckey command line option to the rsa utility. Thanks to
105 Adrian Peck <bertie@ncipher.com> for posting details of the modified
106 algorithm to openssl-dev.
109 *) The evp_local.h macros were using 'c.##kname' which resulted in
110 invalid expansion on some systems (SCO 5.0.5 for example).
111 Corrected to 'c.kname'.
112 [Phillip Porch <root@theporch.com>]
114 *) New X509_get1_email() and X509_REQ_get1_email() functions that return
115 a STACK of email addresses from a certificate or request, these look
116 in the subject name and the subject alternative name extensions and
117 omit any duplicate addresses.
120 *) Re-implement BN_mod_exp2_mont using independent (and larger) windows.
121 This makes DSA verification about 2 % faster.
124 *) Increase maximum window size in BN_mod_exp_... to 6 bits instead of 5
125 (meaning that now 2^5 values will be precomputed, which is only 4 KB
126 plus overhead for 1024 bit moduli).
127 This makes exponentiations about 0.5 % faster for 1024 bit
128 exponents (as measured by "openssl speed rsa2048").
131 *) Rename memory handling macros to avoid conflicts with other
133 Malloc => OPENSSL_malloc
134 Malloc_locked => OPENSSL_malloc_locked
135 Realloc => OPENSSL_realloc
139 *) New function BN_mod_exp_mont_word for small bases (roughly 15%
140 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange).
144 [John Jarvie <jjarvie@newsguy.com>]
146 *) The type-safe stack code has been rejigged. It is now only compiled
147 in when OpenSSL is configured with the DEBUG_SAFESTACK option and
148 by default all type-specific stack functions are "#define"d back to
149 standard stack functions. This results in more streamlined output
150 but retains the type-safety checking possibilities of the original
154 *) The STACK code has been cleaned up, and certain type declarations
155 that didn't make a lot of sense have been brought in line. This has
156 also involved a cleanup of sorts in safestack.h to more correctly
157 map type-safe stack functions onto their plain stack counterparts.
158 This work has also resulted in a variety of "const"ifications of
159 lots of the code, especially "_cmp" operations which should normally
160 be prototyped with "const" parameters anyway.
163 *) When generating bytes for the first time in md_rand.c, 'stir the pool'
164 by seeding with STATE_SIZE dummy bytes (with zero entropy count).
165 (The PRNG state consists of two parts, the large pool 'state' and 'md',
166 where all of 'md' is used each time the PRNG is used, but 'state'
167 is used only indexed by a cyclic counter. As entropy may not be
168 well distributed from the beginning, 'md' is important as a
169 chaining variable. However, the output function chains only half
170 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains
171 all of 'md', and seeding with STATE_SIZE dummy bytes will result
172 in all of 'state' being rewritten, with the new values depending
173 on virtually all of 'md'. This overcomes the 80 bit limitation.)
176 *) In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when
177 the handshake is continued after ssl_verify_cert_chain();
178 otherwise, if SSL_VERIFY_NONE is set, remaining error codes
179 can lead to 'unexplainable' connection aborts later.
180 [Bodo Moeller; problem tracked down by Lutz Jaenicke]
182 *) Major EVP API cipher revision.
183 Add hooks for extra EVP features. This allows various cipher
184 parameters to be set in the EVP interface. Support added for variable
185 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and
186 setting of RC2 and RC5 parameters.
188 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length
191 Remove lots of duplicated code from the EVP library. For example *every*
192 cipher init() function handles the 'iv' in the same way according to the
193 cipher mode. They also all do nothing if the 'key' parameter is NULL and
194 for CFB and OFB modes they zero ctx->num.
196 New functionality allows removal of S/MIME code RC2 hack.
198 Most of the routines have the same form and so can be declared in terms
201 By shifting this to the top level EVP_CipherInit() it can be removed from
202 all individual ciphers. If the cipher wants to handle IVs or keys
203 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT
206 Change lots of functions like EVP_EncryptUpdate() to now return a
207 value: although software versions of the algorithms cannot fail
208 any installed hardware versions can.
211 *) Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if
212 this option is set, tolerate broken clients that send the negotiated
213 protocol version number instead of the requested protocol version
217 *) Call dh_tmp_cb (set by ..._TMP_DH_CB) with correct 'is_export' flag;
218 i.e. non-zero for export ciphersuites, zero otherwise.
219 Previous versions had this flag inverted, inconsistent with
220 rsa_tmp_cb (..._TMP_RSA_CB).
221 [Bodo Moeller; problem reported by Amit Chopra]
223 *) Add missing DSA library text string. Work around for some IIS
224 key files with invalid SEQUENCE encoding.
227 *) Add a document (doc/standards.txt) that list all kinds of standards
228 and so on that are implemented in OpenSSL.
231 *) Enhance c_rehash script. Old version would mishandle certificates
232 with the same subject name hash and wouldn't handle CRLs at all.
233 Added -fingerprint option to crl utility, to support new c_rehash
237 *) Eliminate non-ANSI declarations in crypto.h and stack.h.
240 *) Fix for SSL server purpose checking. Server checking was
241 rejecting certificates which had extended key usage present
242 but no ssl client purpose.
243 [Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>]
245 *) Make PKCS#12 code work with no password. The PKCS#12 spec
246 is a little unclear about how a blank password is handled.
247 Since the password in encoded as a BMPString with terminating
248 double NULL a zero length password would end up as just the
249 double NULL. However no password at all is different and is
250 handled differently in the PKCS#12 key generation code. NS
251 treats a blank password as zero length. MSIE treats it as no
252 password on export: but it will try both on import. We now do
253 the same: PKCS12_parse() tries zero length and no password if
254 the password is set to "" or NULL (NULL is now a valid password:
255 it wasn't before) as does the pkcs12 application.
258 *) Bugfixes in apps/x509.c: Avoid a memory leak; and don't use
259 perror when PEM_read_bio_X509_REQ fails, the error message must
260 be obtained from the error queue.
263 *) Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing
264 it in ERR_remove_state if appropriate, and change ERR_get_state
265 accordingly to avoid race conditions (this is necessary because
266 thread_hash is no longer constant once set).
269 *) Bugfix for linux-elf makefile.one.
272 *) RSA_get_default_method() will now cause a default
273 RSA_METHOD to be chosen if one doesn't exist already.
274 Previously this was only set during a call to RSA_new()
275 or RSA_new_method(NULL) meaning it was possible for
276 RSA_get_default_method() to return NULL.
279 *) Added native name translation to the existing DSO code
280 that will convert (if the flag to do so is set) filenames
281 that are sufficiently small and have no path information
282 into a canonical native form. Eg. "blah" converted to
283 "libblah.so" or "blah.dll" etc.
286 *) New function ERR_error_string_n(e, buf, len) which is like
287 ERR_error_string(e, buf), but writes at most 'len' bytes
288 including the 0 terminator. For ERR_error_string_n, 'buf'
290 [Damien Miller <djm@mindrot.org>, Bodo Moeller]
292 *) CONF library reworked to become more general. A new CONF
293 configuration file reader "class" is implemented as well as a
294 new functions (NCONF_*, for "New CONF") to handle it. The now
295 old CONF_* functions are still there, but are reimplemented to
296 work in terms of the new functions. Also, a set of functions
297 to handle the internal storage of the configuration data is
298 provided to make it easier to write new configuration file
299 reader "classes" (I can definitely see something reading a
300 configuration file in XML format, for example), called _CONF_*,
301 or "the configuration storage API"...
303 The new configuration file reading functions are:
305 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio,
306 NCONF_get_section, NCONF_get_string, NCONF_get_numbre
308 NCONF_default, NCONF_WIN32
310 NCONF_dump_fp, NCONF_dump_bio
312 NCONF_default and NCONF_WIN32 are method (or "class") choosers,
313 NCONF_new creates a new CONF object. This works in the same way
314 as other interfaces in OpenSSL, like the BIO interface.
315 NCONF_dump_* dump the internal storage of the configuration file,
316 which is useful for debugging. All other functions take the same
317 arguments as the old CONF_* functions wth the exception of the
318 first that must be a `CONF *' instead of a `LHASH *'.
320 To make it easer to use the new classes with the old CONF_* functions,
321 the function CONF_set_default_method is provided.
324 *) Add '-tls1' option to 'openssl ciphers', which was already
325 mentioned in the documentation but had not been implemented.
326 (This option is not yet really useful because even the additional
327 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.)
330 *) Initial DSO code added into libcrypto for letting OpenSSL (and
331 OpenSSL-based applications) load shared libraries and bind to
332 them in a portable way.
333 [Geoff Thorpe, with contributions from Richard Levitte]
335 Changes between 0.9.5 and 0.9.5a [1 Apr 2000]
337 *) Make sure _lrotl and _lrotr are only used with MSVC.
339 *) Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status
340 (the default implementation of RAND_status).
342 *) Rename openssl x509 option '-crlext', which was added in 0.9.5,
343 to '-clrext' (= clear extensions), as intended and documented.
344 [Bodo Moeller; inconsistency pointed out by Michael Attili
347 *) Fix for HMAC. It wasn't zeroing the rest of the block if the key length
348 was larger than the MD block size.
349 [Steve Henson, pointed out by Yost William <YostW@tce.com>]
351 *) Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument
352 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set()
353 using the passed key: if the passed key was a private key the result
354 of X509_print(), for example, would be to print out all the private key
358 *) des_quad_cksum() byte order bug fix.
359 [Ulf Möller, using the problem description in krb4-0.9.7, where
360 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>]
362 *) Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly
364 [Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>]
366 *) For easily testing in shell scripts whether some command
367 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX'
368 returns with exit code 0 iff no command of the given name is available.
369 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases,
370 the output goes to stdout and nothing is printed to stderr.
371 Additional arguments are always ignored.
373 Since for each cipher there is a command of the same name,
374 the 'no-cipher' compilation switches can be tested this way.
376 ('openssl no-XXX' is not able to detect pseudo-commands such
377 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.)
380 *) Update test suite so that 'make test' succeeds in 'no-rsa' configuration.
383 *) For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE
384 is set; it will be thrown away anyway because each handshake creates
386 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition
387 to parameters -- in previous versions (since OpenSSL 0.9.3) the
388 'default key' from SSL_CTX_set_tmp_dh would always be lost, meanining
389 you effectivly got SSL_OP_SINGLE_DH_USE when using this macro.
392 *) New s_client option -ign_eof: EOF at stdin is ignored, and
393 'Q' and 'R' lose their special meanings (quit/renegotiate).
394 This is part of what -quiet does; unlike -quiet, -ign_eof
395 does not suppress any output.
398 *) Add compatibility options to the purpose and trust code. The
399 purpose X509_PURPOSE_ANY is "any purpose" which automatically
400 accepts a certificate or CA, this was the previous behaviour,
401 with all the associated security issues.
403 X509_TRUST_COMPAT is the old trust behaviour: only and
404 automatically trust self signed roots in certificate store. A
405 new trust setting X509_TRUST_DEFAULT is used to specify that
406 a purpose has no associated trust setting and it should instead
407 use the value in the default purpose.
410 *) Fix the PKCS#8 DSA private key code so it decodes keys again
411 and fix a memory leak.
414 *) In util/mkerr.pl (which implements 'make errors'), preserve
415 reason strings from the previous version of the .c file, as
416 the default to have only downcase letters (and digits) in
417 automatically generated reasons codes is not always appropriate.
420 *) In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table
421 using strerror. Previously, ERR_reason_error_string() returned
422 library names as reason strings for SYSerr; but SYSerr is a special
423 case where small numbers are errno values, not library numbers.
426 *) Add '-dsaparam' option to 'openssl dhparam' application. This
427 converts DSA parameters into DH parameters. (When creating parameters,
428 DSA_generate_parameters is used.)
431 *) Include 'length' (recommended exponent length) in C code generated
432 by 'openssl dhparam -C'.
435 *) The second argument to set_label in perlasm was already being used
436 so couldn't be used as a "file scope" flag. Moved to third argument
440 *) In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes
441 instead of RAND_bytes for encryption IVs and salts.
444 *) Include RAND_status() into RAND_METHOD instead of implementing
445 it only for md_rand.c Otherwise replacing the PRNG by calling
446 RAND_set_rand_method would be impossible.
449 *) Don't let DSA_generate_key() enter an infinite loop if the random
450 number generation fails.
453 *) New 'rand' application for creating pseudo-random output.
456 *) Added configuration support for Linux/IA64
457 [Rolf Haberrecker <rolf@suse.de>]
459 *) Assembler module support for Mingw32.
462 *) Shared library support for HPUX (in shlib/).
463 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous]
465 *) Shared library support for Solaris gcc.
466 [Lutz Behnke <behnke@trustcenter.de>]
468 Changes between 0.9.4 and 0.9.5 [28 Feb 2000]
470 *) PKCS7_encrypt() was adding text MIME headers twice because they
471 were added manually and by SMIME_crlf_copy().
474 *) In bntest.c don't call BN_rand with zero bits argument.
475 [Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>]
477 *) BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n]
478 case was implemented. This caused BN_div_recp() to fail occasionally.
481 *) Add an optional second argument to the set_label() in the perl
482 assembly language builder. If this argument exists and is set
483 to 1 it signals that the assembler should use a symbol whose
484 scope is the entire file, not just the current function. This
485 is needed with MASM which uses the format label:: for this scope.
486 [Steve Henson, pointed out by Peter Runestig <peter@runestig.com>]
488 *) Change the ASN1 types so they are typedefs by default. Before
489 almost all types were #define'd to ASN1_STRING which was causing
490 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING)
494 *) Change names of new functions to the new get1/get0 naming
495 convention: After 'get1', the caller owns a reference count
496 and has to call ..._free; 'get0' returns a pointer to some
497 data structure without incrementing reference counters.
498 (Some of the existing 'get' functions increment a reference
499 counter, some don't.)
500 Similarly, 'set1' and 'add1' functions increase reference
501 counters or duplicate objects.
504 *) Allow for the possibility of temp RSA key generation failure:
505 the code used to assume it always worked and crashed on failure.
508 *) Fix potential buffer overrun problem in BIO_printf().
509 [Ulf Möller, using public domain code by Patrick Powell; problem
510 pointed out by David Sacerdote <das33@cornell.edu>]
512 *) Support EGD <http://www.lothar.com/tech/crypto/>. New functions
513 RAND_egd() and RAND_status(). In the command line application,
514 the EGD socket can be specified like a seed file using RANDFILE
518 *) Allow the string CERTIFICATE to be tolerated in PKCS#7 structures.
519 Some CAs (e.g. Verisign) distribute certificates in this form.
522 *) Remove the SSL_ALLOW_ADH compile option and set the default cipher
523 list to exclude them. This means that no special compilation option
524 is needed to use anonymous DH: it just needs to be included in the
528 *) Change the EVP_MD_CTX_type macro so its meaning consistent with
529 EVP_MD_type. The old functionality is available in a new macro called
530 EVP_MD_md(). Change code that uses it and update docs.
533 *) ..._ctrl functions now have corresponding ..._callback_ctrl functions
534 where the 'void *' argument is replaced by a function pointer argument.
535 Previously 'void *' was abused to point to functions, which works on
536 many platforms, but is not correct. As these functions are usually
537 called by macros defined in OpenSSL header files, most source code
538 should work without changes.
541 *) <openssl/opensslconf.h> (which is created by Configure) now contains
542 sections with information on -D... compiler switches used for
543 compiling the library so that applications can see them. To enable
544 one of these sections, a pre-processor symbol OPENSSL_..._DEFINES
545 must be defined. E.g.,
546 #define OPENSSL_ALGORITHM_DEFINES
547 #include <openssl/opensslconf.h>
548 defines all pertinent NO_<algo> symbols, such as NO_IDEA, NO_RSA, etc.
549 [Richard Levitte, Ulf and Bodo Möller]
551 *) Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS
555 *) Change the 'other' type in certificate aux info to a STACK_OF
556 X509_ALGOR. Although not an AlgorithmIdentifier as such it has
557 the required ASN1 format: arbitrary types determined by an OID.
560 *) Add some PEM_write_X509_REQ_NEW() functions and a command line
561 argument to 'req'. This is not because the function is newer or
562 better than others it just uses the work 'NEW' in the certificate
563 request header lines. Some software needs this.
566 *) Reorganise password command line arguments: now passwords can be
567 obtained from various sources. Delete the PEM_cb function and make
568 it the default behaviour: i.e. if the callback is NULL and the
569 usrdata argument is not NULL interpret it as a null terminated pass
570 phrase. If usrdata and the callback are NULL then the pass phrase
571 is prompted for as usual.
574 *) Add support for the Compaq Atalla crypto accelerator. If it is installed,
575 the support is automatically enabled. The resulting binaries will
576 autodetect the card and use it if present.
577 [Ben Laurie and Compaq Inc.]
579 *) Work around for Netscape hang bug. This sends certificate request
580 and server done in one record. Since this is perfectly legal in the
581 SSL/TLS protocol it isn't a "bug" option and is on by default. See
582 the bugs/SSLv3 entry for more info.
585 *) HP-UX tune-up: new unified configs, HP C compiler bug workaround.
588 *) Add -rand argument to smime and pkcs12 applications and read/write
592 *) New 'passwd' tool for crypt(3) and apr1 password hashes.
595 *) Add command line password options to the remaining applications.
598 *) Bug fix for BN_div_recp() for numerators with an even number of
602 *) More tests in bntest.c, and changed test_bn output.
605 *) ./config recognizes MacOS X now.
608 *) Bug fix for BN_div() when the first words of num and divsor are
609 equal (it gave wrong results if (rem=(n1-q*d0)&BN_MASK2) < d0).
612 *) Add support for various broken PKCS#8 formats, and command line
613 options to produce them.
616 *) New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to
617 get temporary BIGNUMs from a BN_CTX.
620 *) Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont()
624 *) Change the SSLeay_add_all_*() functions to OpenSSL_add_all_*() and
625 include a #define from the old name to the new. The original intent
626 was that statically linked binaries could for example just call
627 SSLeay_add_all_ciphers() to just add ciphers to the table and not
628 link with digests. This never worked becayse SSLeay_add_all_digests()
629 and SSLeay_add_all_ciphers() were in the same source file so calling
630 one would link with the other. They are now in separate source files.
633 *) Add a new -notext option to 'ca' and a -pubkey option to 'spkac'.
636 *) Use a less unusual form of the Miller-Rabin primality test (it used
637 a binary algorithm for exponentiation integrated into the Miller-Rabin
638 loop, our standard modexp algorithms are faster).
641 *) Support for the EBCDIC character set completed.
642 [Martin Kraemer <Martin.Kraemer@Mch.SNI.De>]
644 *) Source code cleanups: use const where appropriate, eliminate casts,
645 use void * instead of char * in lhash.
648 *) Bugfix: ssl3_send_server_key_exchange was not restartable
649 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of
650 this the server could overwrite ephemeral keys that the client
654 *) Turn DSA_is_prime into a macro that calls BN_is_prime,
655 using 50 iterations of the Rabin-Miller test.
657 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50
658 iterations of the Rabin-Miller test as required by the appendix
659 to FIPS PUB 186[-1]) instead of DSA_is_prime.
660 As BN_is_prime_fasttest includes trial division, DSA parameter
661 generation becomes much faster.
663 This implies a change for the callback functions in DSA_is_prime
664 and DSA_generate_parameters: The callback function is called once
665 for each positive witness in the Rabin-Miller test, not just
666 occasionally in the inner loop; and the parameters to the
667 callback function now provide an iteration count for the outer
668 loop rather than for the current invocation of the inner loop.
669 DSA_generate_parameters additionally can call the callback
670 function with an 'iteration count' of -1, meaning that a
671 candidate has passed the trial division test (when q is generated
672 from an application-provided seed, trial division is skipped).
675 *) New function BN_is_prime_fasttest that optionally does trial
676 division before starting the Rabin-Miller test and has
677 an additional BN_CTX * argument (whereas BN_is_prime always
678 has to allocate at least one BN_CTX).
679 'callback(1, -1, cb_arg)' is called when a number has passed the
680 trial division stage.
683 *) Fix for bug in CRL encoding. The validity dates weren't being handled
687 *) New -pkcs12 option to CA.pl script to write out a PKCS#12 file.
690 *) New function BN_pseudo_rand().
693 *) Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable)
694 bignum version of BN_from_montgomery() with the working code from
695 SSLeay 0.9.0 (the word based version is faster anyway), and clean up
699 *) Avoid a race condition in s2_clnt.c (function get_server_hello) that
700 made it impossible to use the same SSL_SESSION data structure in
701 SSL2 clients in multiple threads.
704 *) The return value of RAND_load_file() no longer counts bytes obtained
705 by stat(). RAND_load_file(..., -1) is new and uses the complete file
706 to seed the PRNG (previously an explicit byte count was required).
707 [Ulf Möller, Bodo Möller]
709 *) Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
710 used (char *) instead of (void *) and had casts all over the place.
713 *) Make BN_generate_prime() return NULL on error if ret!=NULL.
716 *) Retain source code compatibility for BN_prime_checks macro:
717 BN_is_prime(..., BN_prime_checks, ...) now uses
718 BN_prime_checks_for_size to determine the appropriate number of
719 Rabin-Miller iterations.
722 *) Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
723 DH_CHECK_P_NOT_SAFE_PRIME.
724 (Check if this is true? OpenPGP calls them "strong".)
727 *) Merge the functionality of "dh" and "gendh" programs into a new program
728 "dhparam". The old programs are retained for now but will handle DH keys
729 (instead of parameters) in future.
732 *) Make the ciphers, s_server and s_client programs check the return values
733 when a new cipher list is set.
736 *) Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
737 ciphers. Before when the 56bit ciphers were enabled the sorting was
740 The syntax for the cipher sorting has been extended to support sorting by
741 cipher-strength (using the strength_bits hard coded in the tables).
742 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
744 Fix a bug in the cipher-command parser: when supplying a cipher command
745 string with an "undefined" symbol (neither command nor alphanumeric
746 [A-Za-z0-9], ssl_set_cipher_list used to hang in an endless loop. Now
749 Due to the strength-sorting extension, the code of the
750 ssl_create_cipher_list() function was completely rearranged. I hope that
751 the readability was also increased :-)
752 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
754 *) Minor change to 'x509' utility. The -CAcreateserial option now uses 1
755 for the first serial number and places 2 in the serial number file. This
756 avoids problems when the root CA is created with serial number zero and
757 the first user certificate has the same issuer name and serial number
761 *) Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
762 the new code. Add documentation for this stuff.
765 *) Changes to X509_ATTRIBUTE utilities. These have been renamed from
766 X509_*() to X509at_*() on the grounds that they don't handle X509
767 structures and behave in an analagous way to the X509v3 functions:
768 they shouldn't be called directly but wrapper functions should be used
771 So we also now have some wrapper functions that call the X509at functions
772 when passed certificate requests. (TO DO: similar things can be done with
773 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
774 things. Some of these need some d2i or i2d and print functionality
775 because they handle more complex structures.)
778 *) Add missing #ifndefs that caused missing symbols when building libssl
779 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
781 [Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller]
783 *) Precautions against using the PRNG uninitialized: RAND_bytes() now
784 has a return value which indicates the quality of the random data
785 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
786 error queue. New function RAND_pseudo_bytes() generates output that is
787 guaranteed to be unique but not unpredictable. RAND_add is like
788 RAND_seed, but takes an extra argument for an entropy estimate
789 (RAND_seed always assumes full entropy).
792 *) Do more iterations of Rabin-Miller probable prime test (specifically,
793 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
794 instead of only 2 for all lengths; see BN_prime_checks_for_size definition
795 in crypto/bn/bn_prime.c for the complete table). This guarantees a
796 false-positive rate of at most 2^-80 for random input.
799 *) Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
802 *) New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain
803 in the 0.9.5 release), this returns the chain
804 from an X509_CTX structure with a dup of the stack and all
805 the X509 reference counts upped: so the stack will exist
806 after X509_CTX_cleanup() has been called. Modify pkcs12.c
809 Also make SSL_SESSION_print() print out the verify return
813 *) Add manpage for the pkcs12 command. Also change the default
814 behaviour so MAC iteration counts are used unless the new
815 -nomaciter option is used. This improves file security and
816 only older versions of MSIE (4.0 for example) need it.
819 *) Honor the no-xxx Configure options when creating .DEF files.
822 *) Add PKCS#10 attributes to field table: challengePassword,
823 unstructuredName and unstructuredAddress. These are taken from
824 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
825 international characters are used.
827 More changes to X509_ATTRIBUTE code: allow the setting of types
828 based on strings. Remove the 'loc' parameter when adding
829 attributes because these will be a SET OF encoding which is sorted
833 *) Initial changes to the 'req' utility to allow request generation
834 automation. This will allow an application to just generate a template
835 file containing all the field values and have req construct the
838 Initial support for X509_ATTRIBUTE handling. Stacks of these are
839 used all over the place including certificate requests and PKCS#7
840 structures. They are currently handled manually where necessary with
841 some primitive wrappers for PKCS#7. The new functions behave in a
842 manner analogous to the X509 extension functions: they allow
843 attributes to be looked up by NID and added.
845 Later something similar to the X509V3 code would be desirable to
846 automatically handle the encoding, decoding and printing of the
847 more complex types. The string types like challengePassword can
848 be handled by the string table functions.
850 Also modified the multi byte string table handling. Now there is
851 a 'global mask' which masks out certain types. The table itself
852 can use the flag STABLE_NO_MASK to ignore the mask setting: this
853 is useful when for example there is only one permissible type
854 (as in countryName) and using the mask might result in no valid
858 *) Clean up 'Finished' handling, and add functions SSL_get_finished and
859 SSL_get_peer_finished to allow applications to obtain the latest
860 Finished messages sent to the peer or expected from the peer,
861 respectively. (SSL_get_peer_finished is usually the Finished message
862 actually received from the peer, otherwise the protocol will be aborted.)
864 As the Finished message are message digests of the complete handshake
865 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
866 be used for external authentication procedures when the authentication
867 provided by SSL/TLS is not desired or is not enough.
870 *) Enhanced support for Alpha Linux is added. Now ./config checks if
871 the host supports BWX extension and if Compaq C is present on the
872 $PATH. Just exploiting of the BWX extension results in 20-30%
873 performance kick for some algorithms, e.g. DES and RC4 to mention
874 a couple. Compaq C in turn generates ~20% faster code for MD5 and
878 *) Add support for MS "fast SGC". This is arguably a violation of the
879 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
880 weak crypto and after checking the certificate is SGC a second one
881 with strong crypto. MS SGC stops the first handshake after receiving
882 the server certificate message and sends a second client hello. Since
883 a server will typically do all the time consuming operations before
884 expecting any further messages from the client (server key exchange
885 is the most expensive) there is little difference between the two.
887 To get OpenSSL to support MS SGC we have to permit a second client
888 hello message after we have sent server done. In addition we have to
889 reset the MAC if we do get this second client hello.
892 *) Add a function 'd2i_AutoPrivateKey()' this will automatically decide
893 if a DER encoded private key is RSA or DSA traditional format. Changed
894 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
895 format DER encoded private key. Newer code should use PKCS#8 format which
896 has the key type encoded in the ASN1 structure. Added DER private key
897 support to pkcs8 application.
900 *) SSL 3/TLS 1 servers now don't request certificates when an anonymous
901 ciphersuites has been selected (as required by the SSL 3/TLS 1
902 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
903 is set, we interpret this as a request to violate the specification
904 (the worst that can happen is a handshake failure, and 'correct'
905 behaviour would result in a handshake failure anyway).
908 *) In SSL_CTX_add_session, take into account that there might be multiple
909 SSL_SESSION structures with the same session ID (e.g. when two threads
910 concurrently obtain them from an external cache).
911 The internal cache can handle only one SSL_SESSION with a given ID,
912 so if there's a conflict, we now throw out the old one to achieve
916 *) Add OIDs for idea and blowfish in CBC mode. This will allow both
917 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
918 some routines that use cipher OIDs: some ciphers do not have OIDs
919 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
923 *) Simplify the trust setting structure and code. Now we just have
924 two sequences of OIDs for trusted and rejected settings. These will
925 typically have values the same as the extended key usage extension
926 and any application specific purposes.
928 The trust checking code now has a default behaviour: it will just
929 check for an object with the same NID as the passed id. Functions can
930 be provided to override either the default behaviour or the behaviour
931 for a given id. SSL client, server and email already have functions
932 in place for compatibility: they check the NID and also return "trusted"
933 if the certificate is self signed.
936 *) Add d2i,i2d bio/fp functions for PrivateKey: these convert the
937 traditional format into an EVP_PKEY structure.
940 *) Add a password callback function PEM_cb() which either prompts for
941 a password if usr_data is NULL or otherwise assumes it is a null
942 terminated password. Allow passwords to be passed on command line
943 environment or config files in a few more utilities.
946 *) Add a bunch of DER and PEM functions to handle PKCS#8 format private
947 keys. Add some short names for PKCS#8 PBE algorithms and allow them
948 to be specified on the command line for the pkcs8 and pkcs12 utilities.
949 Update documentation.
952 *) Support for ASN1 "NULL" type. This could be handled before by using
953 ASN1_TYPE but there wasn't any function that would try to read a NULL
954 and produce an error if it couldn't. For compatibility we also have
955 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
956 don't allocate anything because they don't need to.
959 *) Initial support for MacOS is now provided. Examine INSTALL.MacOS
961 [Andy Polyakov, Roy Woods <roy@centicsystems.ca>]
963 *) Rebuild of the memory allocation routines used by OpenSSL code and
964 possibly others as well. The purpose is to make an interface that
965 provide hooks so anyone can build a separate set of allocation and
966 deallocation routines to be used by OpenSSL, for example memory
967 pool implementations, or something else, which was previously hard
968 since Malloc(), Realloc() and Free() were defined as macros having
969 the values malloc, realloc and free, respectively (except for Win32
970 compilations). The same is provided for memory debugging code.
971 OpenSSL already comes with functionality to find memory leaks, but
972 this gives people a chance to debug other memory problems.
974 With these changes, a new set of functions and macros have appeared:
976 CRYPTO_set_mem_debug_functions() [F]
977 CRYPTO_get_mem_debug_functions() [F]
978 CRYPTO_dbg_set_options() [F]
979 CRYPTO_dbg_get_options() [F]
980 CRYPTO_malloc_debug_init() [M]
982 The memory debug functions are NULL by default, unless the library
983 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
984 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
985 gives the standard debugging functions that come with OpenSSL) or
986 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
987 provided by the library user) must be used. When the standard
988 debugging functions are used, CRYPTO_dbg_set_options can be used to
989 request additional information:
990 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
991 the CRYPTO_MDEBUG_xxx macro when compiling the library.
993 Also, things like CRYPTO_set_mem_functions will always give the
994 expected result (the new set of functions is used for allocation
995 and deallocation) at all times, regardless of platform and compiler
998 To finish it up, some functions that were never use in any other
999 way than through macros have a new API and new semantic:
1002 CRYPTO_dbg_realloc()
1005 All macros of value have retained their old syntax.
1006 [Richard Levitte and Bodo Moeller]
1008 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
1009 ordering of SMIMECapabilities wasn't in "strength order" and there
1010 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
1014 *) Some ASN1 types with illegal zero length encoding (INTEGER,
1015 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
1016 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson]
1018 *) Merge in my S/MIME library for OpenSSL. This provides a simple
1019 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
1020 functionality to handle multipart/signed properly) and a utility
1021 called 'smime' to call all this stuff. This is based on code I
1022 originally wrote for Celo who have kindly allowed it to be
1023 included in OpenSSL.
1026 *) Add variants des_set_key_checked and des_set_key_unchecked of
1027 des_set_key (aka des_key_sched). Global variable des_check_key
1028 decides which of these is called by des_set_key; this way
1029 des_check_key behaves as it always did, but applications and
1030 the library itself, which was buggy for des_check_key == 1,
1031 have a cleaner way to pick the version they need.
1034 *) New function PKCS12_newpass() which changes the password of a
1038 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and
1039 dynamic mix. In both cases the ids can be used as an index into the
1040 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
1041 functions so they accept a list of the field values and the
1042 application doesn't need to directly manipulate the X509_TRUST
1046 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
1050 *) Modify the way the V3 extension code looks up extensions. This now
1051 works in a similar way to the object code: we have some "standard"
1052 extensions in a static table which is searched with OBJ_bsearch()
1053 and the application can add dynamic ones if needed. The file
1054 crypto/x509v3/ext_dat.h now has the info: this file needs to be
1055 updated whenever a new extension is added to the core code and kept
1056 in ext_nid order. There is a simple program 'tabtest.c' which checks
1057 this. New extensions are not added too often so this file can readily
1058 be maintained manually.
1060 There are two big advantages in doing things this way. The extensions
1061 can be looked up immediately and no longer need to be "added" using
1062 X509V3_add_standard_extensions(): this function now does nothing.
1063 [Side note: I get *lots* of email saying the extension code doesn't
1064 work because people forget to call this function]
1065 Also no dynamic allocation is done unless new extensions are added:
1066 so if we don't add custom extensions there is no need to call
1067 X509V3_EXT_cleanup().
1070 *) Modify enc utility's salting as follows: make salting the default. Add a
1071 magic header, so unsalted files fail gracefully instead of just decrypting
1072 to garbage. This is because not salting is a big security hole, so people
1073 should be discouraged from doing it.
1076 *) Fixes and enhancements to the 'x509' utility. It allowed a message
1077 digest to be passed on the command line but it only used this
1078 parameter when signing a certificate. Modified so all relevant
1079 operations are affected by the digest parameter including the
1080 -fingerprint and -x509toreq options. Also -x509toreq choked if a
1081 DSA key was used because it didn't fix the digest.
1084 *) Initial certificate chain verify code. Currently tests the untrusted
1085 certificates for consistency with the verify purpose (which is set
1086 when the X509_STORE_CTX structure is set up) and checks the pathlength.
1088 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
1089 this is because it will reject chains with invalid extensions whereas
1090 every previous version of OpenSSL and SSLeay made no checks at all.
1092 Trust code: checks the root CA for the relevant trust settings. Trust
1093 settings have an initial value consistent with the verify purpose: e.g.
1094 if the verify purpose is for SSL client use it expects the CA to be
1095 trusted for SSL client use. However the default value can be changed to
1096 permit custom trust settings: one example of this would be to only trust
1097 certificates from a specific "secure" set of CAs.
1099 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
1100 which should be used for version portability: especially since the
1101 verify structure is likely to change more often now.
1103 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
1104 to set them. If not set then assume SSL clients will verify SSL servers
1107 Two new options to the verify program: -untrusted allows a set of
1108 untrusted certificates to be passed in and -purpose which sets the
1109 intended purpose of the certificate. If a purpose is set then the
1110 new chain verify code is used to check extension consistency.
1113 *) Support for the authority information access extension.
1116 *) Modify RSA and DSA PEM read routines to transparently handle
1117 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
1118 public keys in a format compatible with certificate
1119 SubjectPublicKeyInfo structures. Unfortunately there were already
1120 functions called *_PublicKey_* which used various odd formats so
1121 these are retained for compatibility: however the DSA variants were
1122 never in a public release so they have been deleted. Changed dsa/rsa
1123 utilities to handle the new format: note no releases ever handled public
1124 keys so we should be OK.
1126 The primary motivation for this change is to avoid the same fiasco
1127 that dogs private keys: there are several incompatible private key
1128 formats some of which are standard and some OpenSSL specific and
1129 require various evil hacks to allow partial transparent handling and
1130 even then it doesn't work with DER formats. Given the option anything
1131 other than PKCS#8 should be dumped: but the other formats have to
1132 stay in the name of compatibility.
1134 With public keys and the benefit of hindsight one standard format
1135 is used which works with EVP_PKEY, RSA or DSA structures: though
1136 it clearly returns an error if you try to read the wrong kind of key.
1138 Added a -pubkey option to the 'x509' utility to output the public key.
1139 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*()
1140 (renamed to EVP_PKEY_get1_*() in the OpenSSL 0.9.5 release) and add
1141 EVP_PKEY_rset_*() functions (renamed to EVP_PKEY_set1_*())
1142 that do the same as the EVP_PKEY_assign_*() except they up the
1143 reference count of the added key (they don't "swallow" the
1147 *) Fixes to crypto/x509/by_file.c the code to read in certificates and
1148 CRLs would fail if the file contained no certificates or no CRLs:
1149 added a new function to read in both types and return the number
1150 read: this means that if none are read it will be an error. The
1151 DER versions of the certificate and CRL reader would always fail
1152 because it isn't possible to mix certificates and CRLs in DER format
1153 without choking one or the other routine. Changed this to just read
1154 a certificate: this is the best we can do. Also modified the code
1155 in apps/verify.c to take notice of return codes: it was previously
1156 attempting to read in certificates from NULL pointers and ignoring
1157 any errors: this is one reason why the cert and CRL reader seemed
1158 to work. It doesn't check return codes from the default certificate
1159 routines: these may well fail if the certificates aren't installed.
1162 *) Code to support otherName option in GeneralName.
1165 *) First update to verify code. Change the verify utility
1166 so it warns if it is passed a self signed certificate:
1167 for consistency with the normal behaviour. X509_verify
1168 has been modified to it will now verify a self signed
1169 certificate if *exactly* the same certificate appears
1170 in the store: it was previously impossible to trust a
1171 single self signed certificate. This means that:
1172 openssl verify ss.pem
1173 now gives a warning about a self signed certificate but
1174 openssl verify -CAfile ss.pem ss.pem
1178 *) For servers, store verify_result in SSL_SESSION data structure
1179 (and add it to external session representation).
1180 This is needed when client certificate verifications fails,
1181 but an application-provided verification callback (set by
1182 SSL_CTX_set_cert_verify_callback) allows accepting the session
1183 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
1184 but returns 1): When the session is reused, we have to set
1185 ssl->verify_result to the appropriate error code to avoid
1187 [Bodo Moeller, problem pointed out by Lutz Jaenicke]
1189 *) Fix a bug in the new PKCS#7 code: it didn't consider the
1190 case in PKCS7_dataInit() where the signed PKCS7 structure
1191 didn't contain any existing data because it was being created.
1192 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson]
1194 *) Add a salt to the key derivation routines in enc.c. This
1195 forms the first 8 bytes of the encrypted file. Also add a
1196 -S option to allow a salt to be input on the command line.
1199 *) New function X509_cmp(). Oddly enough there wasn't a function
1200 to compare two certificates. We do this by working out the SHA1
1201 hash and comparing that. X509_cmp() will be needed by the trust
1205 *) SSL_get1_session() is like SSL_get_session(), but increments
1206 the reference count in the SSL_SESSION returned.
1207 [Geoff Thorpe <geoff@eu.c2.net>]
1209 *) Fix for 'req': it was adding a null to request attributes.
1210 Also change the X509_LOOKUP and X509_INFO code to handle
1211 certificate auxiliary information.
1214 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document
1218 *) Add the possibility to add extra information to the memory leak
1219 detecting output, to form tracebacks, showing from where each
1220 allocation was originated: CRYPTO_push_info("constant string") adds
1221 the string plus current file name and line number to a per-thread
1222 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
1223 is like calling CYRPTO_pop_info() until the stack is empty.
1224 Also updated memory leak detection code to be multi-thread-safe.
1227 *) Add options -text and -noout to pkcs7 utility and delete the
1228 encryption options which never did anything. Update docs.
1231 *) Add options to some of the utilities to allow the pass phrase
1232 to be included on either the command line (not recommended on
1233 OSes like Unix) or read from the environment. Update the
1234 manpages and fix a few bugs.
1237 *) Add a few manpages for some of the openssl commands.
1240 *) Fix the -revoke option in ca. It was freeing up memory twice,
1241 leaking and not finding already revoked certificates.
1244 *) Extensive changes to support certificate auxiliary information.
1245 This involves the use of X509_CERT_AUX structure and X509_AUX
1246 functions. An X509_AUX function such as PEM_read_X509_AUX()
1247 can still read in a certificate file in the usual way but it
1248 will also read in any additional "auxiliary information". By
1249 doing things this way a fair degree of compatibility can be
1250 retained: existing certificates can have this information added
1251 using the new 'x509' options.
1253 Current auxiliary information includes an "alias" and some trust
1254 settings. The trust settings will ultimately be used in enhanced
1255 certificate chain verification routines: currently a certificate
1256 can only be trusted if it is self signed and then it is trusted
1260 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD).
1261 The problem was that one of the replacement routines had not been working
1262 since SSLeay releases. For now the offending routine has been replaced
1263 with non-optimised assembler. Even so, this now gives around 95%
1264 performance improvement for 1024 bit RSA signs.
1267 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2
1268 handling. Most clients have the effective key size in bits equal to
1269 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
1270 A few however don't do this and instead use the size of the decrypted key
1271 to determine the RC2 key length and the AlgorithmIdentifier to determine
1272 the effective key length. In this case the effective key length can still
1273 be 40 bits but the key length can be 168 bits for example. This is fixed
1274 by manually forcing an RC2 key into the EVP_PKEY structure because the
1275 EVP code can't currently handle unusual RC2 key sizes: it always assumes
1276 the key length and effective key length are equal.
1279 *) Add a bunch of functions that should simplify the creation of
1280 X509_NAME structures. Now you should be able to do:
1281 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
1282 and have it automatically work out the correct field type and fill in
1283 the structures. The more adventurous can try:
1284 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
1285 and it will (hopefully) work out the correct multibyte encoding.
1288 *) Change the 'req' utility to use the new field handling and multibyte
1289 copy routines. Before the DN field creation was handled in an ad hoc
1290 way in req, ca, and x509 which was rather broken and didn't support
1291 BMPStrings or UTF8Strings. Since some software doesn't implement
1292 BMPStrings or UTF8Strings yet, they can be enabled using the config file
1293 using the dirstring_type option. See the new comment in the default
1294 openssl.cnf for more info.
1297 *) Make crypto/rand/md_rand.c more robust:
1298 - Assure unique random numbers after fork().
1299 - Make sure that concurrent threads access the global counter and
1300 md serializably so that we never lose entropy in them
1301 or use exactly the same state in multiple threads.
1302 Access to the large state is not always serializable because
1303 the additional locking could be a performance killer, and
1304 md should be large enough anyway.
1307 *) New file apps/app_rand.c with commonly needed functionality
1308 for handling the random seed file.
1310 Use the random seed file in some applications that previously did not:
1312 dsaparam -genkey (which also ignored its '-rand' option),
1315 x509 (when signing).
1316 Except on systems with /dev/urandom, it is crucial to have a random
1317 seed file at least for key creation, DSA signing, and for DH exchanges;
1318 for RSA signatures we could do without one.
1320 gendh and gendsa (unlike genrsa) used to read only the first byte
1321 of each file listed in the '-rand' option. The function as previously
1322 found in genrsa is now in app_rand.c and is used by all programs
1323 that support '-rand'.
1326 *) In RAND_write_file, use mode 0600 for creating files;
1327 don't just chmod when it may be too late.
1330 *) Report an error from X509_STORE_load_locations
1331 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
1334 *) New function ASN1_mbstring_copy() this copies a string in either
1335 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
1336 into an ASN1_STRING type. A mask of permissible types is passed
1337 and it chooses the "minimal" type to use or an error if not type
1341 *) Add function equivalents to the various macros in asn1.h. The old
1342 macros are retained with an M_ prefix. Code inside the library can
1343 use the M_ macros. External code (including the openssl utility)
1344 should *NOT* in order to be "shared library friendly".
1347 *) Add various functions that can check a certificate's extensions
1348 to see if it usable for various purposes such as SSL client,
1349 server or S/MIME and CAs of these types. This is currently
1350 VERY EXPERIMENTAL but will ultimately be used for certificate chain
1351 verification. Also added a -purpose flag to x509 utility to
1352 print out all the purposes.
1355 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated
1359 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search
1360 for, obtain and decode and extension and obtain its critical flag.
1361 This allows all the necessary extension code to be handled in a
1362 single function call.
1365 *) RC4 tune-up featuring 30-40% performance improvement on most RISC
1366 platforms. See crypto/rc4/rc4_enc.c for further details.
1369 *) New -noout option to asn1parse. This causes no output to be produced
1370 its main use is when combined with -strparse and -out to extract data
1371 from a file (which may not be in ASN.1 format).
1374 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer
1375 when producing the local key id.
1376 [Richard Levitte <levitte@stacken.kth.se>]
1378 *) New option -dhparam in s_server. This allows a DH parameter file to be
1379 stated explicitly. If it is not stated then it tries the first server
1380 certificate file. The previous behaviour hard coded the filename
1384 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow
1385 a public key to be input or output. For example:
1386 openssl rsa -in key.pem -pubout -out pubkey.pem
1387 Also added necessary DSA public key functions to handle this.
1390 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
1391 in the message. This was handled by allowing
1392 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
1393 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>]
1395 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
1396 to the end of the strings whereas this didn't. This would cause problems
1397 if strings read with d2i_ASN1_bytes() were later modified.
1398 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>]
1400 *) Fix for base64 decode bug. When a base64 bio reads only one line of
1401 data and it contains EOF it will end up returning an error. This is
1402 caused by input 46 bytes long. The cause is due to the way base64
1403 BIOs find the start of base64 encoded data. They do this by trying a
1404 trial decode on each line until they find one that works. When they
1405 do a flag is set and it starts again knowing it can pass all the
1406 data directly through the decoder. Unfortunately it doesn't reset
1407 the context it uses. This means that if EOF is reached an attempt
1408 is made to pass two EOFs through the context and this causes the
1409 resulting error. This can also cause other problems as well. As is
1410 usual with these problems it takes *ages* to find and the fix is
1411 trivial: move one line.
1412 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ]
1414 *) Ugly workaround to get s_client and s_server working under Windows. The
1415 old code wouldn't work because it needed to select() on sockets and the
1416 tty (for keypresses and to see if data could be written). Win32 only
1417 supports select() on sockets so we select() with a 1s timeout on the
1418 sockets and then see if any characters are waiting to be read, if none
1419 are present then we retry, we also assume we can always write data to
1420 the tty. This isn't nice because the code then blocks until we've
1421 received a complete line of data and it is effectively polling the
1422 keyboard at 1s intervals: however it's quite a bit better than not
1423 working at all :-) A dedicated Windows application might handle this
1424 with an event loop for example.
1427 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
1428 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
1429 will be called when RSA_sign() and RSA_verify() are used. This is useful
1430 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
1431 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
1432 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
1433 This necessitated the support of an extra signature type NID_md5_sha1
1434 for SSL signatures and modifications to the SSL library to use it instead
1435 of calling RSA_public_decrypt() and RSA_private_encrypt().
1438 *) Add new -verify -CAfile and -CApath options to the crl program, these
1439 will lookup a CRL issuers certificate and verify the signature in a
1440 similar way to the verify program. Tidy up the crl program so it
1441 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
1442 less strict. It will now permit CRL extensions even if it is not
1443 a V2 CRL: this will allow it to tolerate some broken CRLs.
1446 *) Initialize all non-automatic variables each time one of the openssl
1447 sub-programs is started (this is necessary as they may be started
1448 multiple times from the "OpenSSL>" prompt).
1449 [Lennart Bang, Bodo Moeller]
1451 *) Preliminary compilation option RSA_NULL which disables RSA crypto without
1452 removing all other RSA functionality (this is what NO_RSA does). This
1453 is so (for example) those in the US can disable those operations covered
1454 by the RSA patent while allowing storage and parsing of RSA keys and RSA
1458 *) Non-copying interface to BIO pairs.
1459 (still largely untested)
1462 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive
1463 ASCII string. This was handled independently in various places before.
1466 *) New functions UTF8_getc() and UTF8_putc() that parse and generate
1467 UTF8 strings a character at a time.
1470 *) Use client_version from client hello to select the protocol
1471 (s23_srvr.c) and for RSA client key exchange verification
1472 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
1475 *) Add various utility functions to handle SPKACs, these were previously
1476 handled by poking round in the structure internals. Added new function
1477 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
1478 print, verify and generate SPKACs. Based on an original idea from
1479 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
1482 *) RIPEMD160 is operational on all platforms and is back in 'make test'.
1485 *) Allow the config file extension section to be overwritten on the
1486 command line. Based on an original idea from Massimiliano Pala
1487 <madwolf@comune.modena.it>. The new option is called -extensions
1488 and can be applied to ca, req and x509. Also -reqexts to override
1489 the request extensions in req and -crlexts to override the crl extensions
1493 *) Add new feature to the SPKAC handling in ca. Now you can include
1494 the same field multiple times by preceding it by "XXXX." for example:
1497 this is the same syntax as used in the req config file.
1500 *) Allow certificate extensions to be added to certificate requests. These
1501 are specified in a 'req_extensions' option of the req section of the
1502 config file. They can be printed out with the -text option to req but
1503 are otherwise ignored at present.
1506 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
1507 data read consists of only the final block it would not decrypted because
1508 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
1509 A misplaced 'break' also meant the decrypted final block might not be
1510 copied until the next read.
1513 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
1514 a few extra parameters to the DH structure: these will be useful if
1515 for example we want the value of 'q' or implement X9.42 DH.
1518 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and
1519 provides hooks that allow the default DSA functions or functions on a
1520 "per key" basis to be replaced. This allows hardware acceleration and
1521 hardware key storage to be handled without major modification to the
1522 library. Also added low level modexp hooks and CRYPTO_EX structure and
1523 associated functions.
1526 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
1527 as "read only": it can't be written to and the buffer it points to will
1528 not be freed. Reading from a read only BIO is much more efficient than
1529 a normal memory BIO. This was added because there are several times when
1530 an area of memory needs to be read from a BIO. The previous method was
1531 to create a memory BIO and write the data to it, this results in two
1532 copies of the data and an O(n^2) reading algorithm. There is a new
1533 function BIO_new_mem_buf() which creates a read only memory BIO from
1534 an area of memory. Also modified the PKCS#7 routines to use read only
1538 *) Bugfix: ssl23_get_client_hello did not work properly when called in
1539 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
1540 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
1541 but a retry condition occured while trying to read the rest.
1544 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as
1545 NID_pkcs7_encrypted by default: this was wrong since this should almost
1546 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
1547 the encrypted data type: this is a more sensible place to put it and it
1548 allows the PKCS#12 code to be tidied up that duplicated this
1552 *) Changed obj_dat.pl script so it takes its input and output files on
1553 the command line. This should avoid shell escape redirection problems
1557 *) Initial support for certificate extension requests, these are included
1558 in things like Xenroll certificate requests. Included functions to allow
1559 extensions to be obtained and added.
1562 *) -crlf option to s_client and s_server for sending newlines as
1563 CRLF (as required by many protocols).
1566 Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
1568 *) Install libRSAglue.a when OpenSSL is built with RSAref.
1569 [Ralf S. Engelschall]
1571 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency.
1572 [Andrija Antonijevic <TheAntony2@bigfoot.com>]
1574 *) Fix -startdate and -enddate (which was missing) arguments to 'ca'
1578 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as
1579 DH parameters/keys (q is lost during that conversion, but the resulting
1580 DH parameters contain its length).
1582 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
1583 much faster than DH_generate_parameters (which creates parameters
1584 where p = 2*q + 1), and also the smaller q makes DH computations
1585 much more efficient (160-bit exponentiation instead of 1024-bit
1586 exponentiation); so this provides a convenient way to support DHE
1587 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
1588 utter importance to use
1589 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
1591 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
1592 when such DH parameters are used, because otherwise small subgroup
1593 attacks may become possible!
1596 *) Avoid memory leak in i2d_DHparams.
1599 *) Allow the -k option to be used more than once in the enc program:
1600 this allows the same encrypted message to be read by multiple recipients.
1603 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
1604 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
1605 it will always use the numerical form of the OID, even if it has a short
1609 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
1610 method only got called if p,q,dmp1,dmq1,iqmp components were present,
1611 otherwise bn_mod_exp was called. In the case of hardware keys for example
1612 no private key components need be present and it might store extra data
1613 in the RSA structure, which cannot be accessed from bn_mod_exp.
1614 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
1615 private key operations.
1618 *) Added support for SPARC Linux.
1621 *) pem_password_cb function type incompatibly changed from
1622 typedef int pem_password_cb(char *buf, int size, int rwflag);
1624 ....(char *buf, int size, int rwflag, void *userdata);
1625 so that applications can pass data to their callbacks:
1626 The PEM[_ASN1]_{read,write}... functions and macros now take an
1627 additional void * argument, which is just handed through whenever
1628 the password callback is called.
1629 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller]
1631 New function SSL_CTX_set_default_passwd_cb_userdata.
1633 Compatibility note: As many C implementations push function arguments
1634 onto the stack in reverse order, the new library version is likely to
1635 interoperate with programs that have been compiled with the old
1636 pem_password_cb definition (PEM_whatever takes some data that
1637 happens to be on the stack as its last argument, and the callback
1638 just ignores this garbage); but there is no guarantee whatsoever that
1641 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
1642 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
1643 problems not only on Windows, but also on some Unix platforms.
1644 To avoid problematic command lines, these definitions are now in an
1645 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
1646 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
1649 *) MIPS III/IV assembler module is reimplemented.
1652 *) More DES library cleanups: remove references to srand/rand and
1653 delete an unused file.
1656 *) Add support for the the free Netwide assembler (NASM) under Win32,
1657 since not many people have MASM (ml) and it can be hard to obtain.
1658 This is currently experimental but it seems to work OK and pass all
1659 the tests. Check out INSTALL.W32 for info.
1662 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
1663 without temporary keys kept an extra copy of the server key,
1664 and connections with temporary keys did not free everything in case
1668 *) New function RSA_check_key and new openssl rsa option -check
1669 for verifying the consistency of RSA keys.
1670 [Ulf Moeller, Bodo Moeller]
1672 *) Various changes to make Win32 compile work:
1673 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
1674 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
1675 comparison" warnings.
1676 3. Add sk_<TYPE>_sort to DEF file generator and do make update.
1679 *) Add a debugging option to PKCS#5 v2 key generation function: when
1680 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
1681 derived keys are printed to stderr.
1684 *) Copy the flags in ASN1_STRING_dup().
1685 [Roman E. Pavlov <pre@mo.msk.ru>]
1687 *) The x509 application mishandled signing requests containing DSA
1688 keys when the signing key was also DSA and the parameters didn't match.
1690 It was supposed to omit the parameters when they matched the signing key:
1691 the verifying software was then supposed to automatically use the CA's
1692 parameters if they were absent from the end user certificate.
1694 Omitting parameters is no longer recommended. The test was also
1695 the wrong way round! This was probably due to unusual behaviour in
1696 EVP_cmp_parameters() which returns 1 if the parameters match.
1697 This meant that parameters were omitted when they *didn't* match and
1698 the certificate was useless. Certificates signed with 'ca' didn't have
1700 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>]
1702 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
1703 The interface is as follows:
1704 Applications can use
1705 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
1706 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
1707 "off" is now the default.
1708 The library internally uses
1709 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
1710 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
1711 to disable memory-checking temporarily.
1713 Some inconsistent states that previously were possible (and were
1714 even the default) are now avoided.
1716 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
1717 with each memory chunk allocated; this is occasionally more helpful
1718 than just having a counter.
1720 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
1722 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
1726 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX),
1727 which largely parallels "options", but is for changing API behaviour,
1728 whereas "options" are about protocol behaviour.
1729 Initial "mode" flags are:
1731 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
1732 a single record has been written.
1733 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
1734 retries use the same buffer location.
1735 (But all of the contents must be
1739 *) Bugfix: SSL_set_mode ignored its parameter, only SSL_CTX_set_mode
1742 *) Fix problems with no-hmac etc.
1743 [Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>]
1745 *) New functions RSA_get_default_method(), RSA_set_method() and
1746 RSA_get_method(). These allows replacement of RSA_METHODs without having
1747 to mess around with the internals of an RSA structure.
1750 *) Fix memory leaks in DSA_do_sign and DSA_is_prime.
1751 Also really enable memory leak checks in openssl.c and in some
1753 [Chad C. Mulligan, Bodo Moeller]
1755 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
1756 up the length of negative integers. This has now been simplified to just
1757 store the length when it is first determined and use it later, rather
1758 than trying to keep track of where data is copied and updating it to
1760 [Steve Henson, reported by Brien Wheeler
1761 <bwheeler@authentica-security.com>]
1763 *) Add a new function PKCS7_signatureVerify. This allows the verification
1764 of a PKCS#7 signature but with the signing certificate passed to the
1765 function itself. This contrasts with PKCS7_dataVerify which assumes the
1766 certificate is present in the PKCS#7 structure. This isn't always the
1767 case: certificates can be omitted from a PKCS#7 structure and be
1768 distributed by "out of band" means (such as a certificate database).
1771 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the
1772 function prototypes in pem.h, also change util/mkdef.pl to add the
1773 necessary function names.
1776 *) mk1mf.pl (used by Windows builds) did not properly read the
1777 options set by Configure in the top level Makefile, and Configure
1778 was not even able to write more than one option correctly.
1779 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
1782 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config
1783 file to be loaded from a BIO or FILE pointer. The BIO version will
1784 for example allow memory BIOs to contain config info.
1787 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
1788 Whoever hopes to achieve shared-library compatibility across versions
1789 must use this, not the compile-time macro.
1790 (Exercise 0.9.4: Which is the minimum library version required by
1792 Note: All this applies only to multi-threaded programs, others don't
1796 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests
1797 through a BIO pair triggered the default case, i.e.
1798 SSLerr(...,SSL_R_UNKNOWN_STATE).
1801 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
1802 can use the SSL library even if none of the specific BIOs is
1806 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
1807 for the encoded length.
1808 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>]
1810 *) Add initial documentation of the X509V3 functions.
1813 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and
1814 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
1815 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
1816 secure PKCS#8 private key format with a high iteration count.
1819 *) Fix determination of Perl interpreter: A perl or perl5
1820 _directory_ in $PATH was also accepted as the interpreter.
1821 [Ralf S. Engelschall]
1823 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking
1824 wrong with it but it was very old and did things like calling
1825 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
1829 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed
1830 to use the new extension code.
1833 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
1834 with macros. This should make it easier to change their form, add extra
1835 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
1839 *) Add to configuration table a new entry that can specify an alternative
1840 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
1841 according to Mark Crispin <MRC@Panda.COM>.
1845 *) DES CBC did not update the IV. Weird.
1848 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
1849 Changing the behaviour of the former might break existing programs --
1850 where IV updating is needed, des_ncbc_encrypt can be used.
1853 *) When bntest is run from "make test" it drives bc to check its
1854 calculations, as well as internally checking them. If an internal check
1855 fails, it needs to cause bc to give a non-zero result or make test carries
1856 on without noticing the failure. Fixed.
1859 *) DES library cleanups.
1862 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
1863 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
1864 ciphers. NOTE: although the key derivation function has been verified
1865 against some published test vectors it has not been extensively tested
1866 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
1870 *) Instead of "mkdir -p", which is not fully portable, use new
1871 Perl script "util/mkdir-p.pl".
1874 *) Rewrite the way password based encryption (PBE) is handled. It used to
1875 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
1876 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
1877 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
1878 the 'parameter' field of the AlgorithmIdentifier is passed to the
1879 underlying key generation function so it must do its own ASN1 parsing.
1880 This has also changed the EVP_PBE_CipherInit() function which now has a
1881 'parameter' argument instead of literal salt and iteration count values
1882 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
1885 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms
1886 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
1887 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
1888 KEY" because this clashed with PKCS#8 unencrypted string. Since this
1889 value was just used as a "magic string" and not used directly its
1890 value doesn't matter.
1893 *) Introduce some semblance of const correctness to BN. Shame C doesn't
1897 *) "linux-sparc64" configuration (ultrapenguin).
1898 [Ray Miller <ray.miller@oucs.ox.ac.uk>]
1899 "linux-sparc" configuration.
1900 [Christian Forster <fo@hawo.stw.uni-erlangen.de>]
1902 *) config now generates no-xxx options for missing ciphers.
1905 *) Support the EBCDIC character set (work in progress).
1906 File ebcdic.c not yet included because it has a different license.
1907 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
1909 *) Support BS2000/OSD-POSIX.
1910 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
1912 *) Make callbacks for key generation use void * instead of char *.
1915 *) Make S/MIME samples compile (not yet tested).
1918 *) Additional typesafe stacks.
1921 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
1925 Changes between 0.9.3 and 0.9.3a [29 May 1999]
1927 *) New configuration variant "sco5-gcc".
1929 *) Updated some demos.
1930 [Sean O Riordain, Wade Scholine]
1932 *) Add missing BIO_free at exit of pkcs12 application.
1935 *) Fix memory leak in conf.c.
1938 *) Updates for Win32 to assembler version of MD5.
1941 *) Set #! path to perl in apps/der_chop to where we found it
1942 instead of using a fixed path.
1945 *) SHA library changes for irix64-mips4-cc.
1948 *) Improvements for VMS support.
1952 Changes between 0.9.2b and 0.9.3 [24 May 1999]
1954 *) Bignum library bug fix. IRIX 6 passes "make test" now!
1955 This also avoids the problems with SC4.2 and unpatched SC5.
1956 [Andy Polyakov <appro@fy.chalmers.se>]
1958 *) New functions sk_num, sk_value and sk_set to replace the previous macros.
1959 These are required because of the typesafe stack would otherwise break
1960 existing code. If old code used a structure member which used to be STACK
1961 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
1962 sk_num or sk_value it would produce an error because the num, data members
1963 are not present in STACK_OF. Now it just produces a warning. sk_set
1964 replaces the old method of assigning a value to sk_value
1965 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
1966 that does this will no longer work (and should use sk_set instead) but
1967 this could be regarded as a "questionable" behaviour anyway.
1970 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now
1971 correctly handle encrypted S/MIME data.
1974 *) Change type of various DES function arguments from des_cblock
1975 (which means, in function argument declarations, pointer to char)
1976 to des_cblock * (meaning pointer to array with 8 char elements),
1977 which allows the compiler to do more typechecking; it was like
1978 that back in SSLeay, but with lots of ugly casts.
1980 Introduce new type const_des_cblock.
1983 *) Reorganise the PKCS#7 library and get rid of some of the more obvious
1984 problems: find RecipientInfo structure that matches recipient certificate
1985 and initialise the ASN1 structures properly based on passed cipher.
1988 *) Belatedly make the BN tests actually check the results.
1991 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
1992 to and from BNs: it was completely broken. New compilation option
1993 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
1994 key elements as negative integers.
1997 *) Reorganize and speed up MD5.
1998 [Andy Polyakov <appro@fy.chalmers.se>]
2001 [Richard Levitte <richard@levitte.org>]
2003 *) New option -out to asn1parse to allow the parsed structure to be
2004 output to a file. This is most useful when combined with the -strparse
2005 option to examine the output of things like OCTET STRINGS.
2008 *) Make SSL library a little more fool-proof by not requiring any longer
2009 that SSL_set_{accept,connect}_state be called before
2010 SSL_{accept,connect} may be used (SSL_set_..._state is omitted
2011 in many applications because usually everything *appeared* to work as
2012 intended anyway -- now it really works as intended).
2015 *) Move openssl.cnf out of lib/.
2018 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall
2019 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
2020 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+
2021 [Ralf S. Engelschall]
2023 *) Various fixes to the EVP and PKCS#7 code. It may now be able to
2024 handle PKCS#7 enveloped data properly.
2025 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve]
2027 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
2028 copying pointers. The cert_st handling is changed by this in
2029 various ways (and thus what used to be known as ctx->default_cert
2030 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert
2031 any longer when s->cert does not give us what we need).
2032 ssl_cert_instantiate becomes obsolete by this change.
2033 As soon as we've got the new code right (possibly it already is?),
2034 we have solved a couple of bugs of the earlier code where s->cert
2035 was used as if it could not have been shared with other SSL structures.
2037 Note that using the SSL API in certain dirty ways now will result
2038 in different behaviour than observed with earlier library versions:
2039 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx)
2040 does not influence s as it used to.
2042 In order to clean up things more thoroughly, inside SSL_SESSION
2043 we don't use CERT any longer, but a new structure SESS_CERT
2044 that holds per-session data (if available); currently, this is
2045 the peer's certificate chain and, for clients, the server's certificate
2046 and temporary key. CERT holds only those values that can have
2047 meaningful defaults in an SSL_CTX.
2050 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
2051 from the internal representation. Various PKCS#7 fixes: remove some
2052 evil casts and set the enc_dig_alg field properly based on the signing
2056 *) Allow PKCS#12 password to be set from the command line or the
2057 environment. Let 'ca' get its config file name from the environment
2058 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
2062 *) Allow certificate policies extension to use an IA5STRING for the
2063 organization field. This is contrary to the PKIX definition but
2064 VeriSign uses it and IE5 only recognises this form. Document 'x509'
2068 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
2069 without disallowing inline assembler and the like for non-pedantic builds.
2072 *) Support Borland C++ builder.
2073 [Janez Jere <jj@void.si>, modified by Ulf Möller]
2078 *) SHA-1 cleanups and performance enhancements.
2079 [Andy Polyakov <appro@fy.chalmers.se>]
2081 *) Sparc v8plus assembler for the bignum library.
2082 [Andy Polyakov <appro@fy.chalmers.se>]
2084 *) Accept any -xxx and +xxx compiler options in Configure.
2087 *) Update HPUX configuration.
2090 *) Add missing sk_<type>_unshift() function to safestack.h
2091 [Ralf S. Engelschall]
2093 *) New function SSL_CTX_use_certificate_chain_file that sets the
2094 "extra_cert"s in addition to the certificate. (This makes sense
2095 only for "PEM" format files, as chains as a whole are not
2099 *) Support verify_depth from the SSL API.
2100 x509_vfy.c had what can be considered an off-by-one-error:
2101 Its depth (which was not part of the external interface)
2102 was actually counting the number of certificates in a chain;
2103 now it really counts the depth.
2106 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
2107 instead of X509err, which often resulted in confusing error
2108 messages since the error codes are not globally unique
2109 (e.g. an alleged error in ssl3_accept when a certificate
2110 didn't match the private key).
2112 *) New function SSL_CTX_set_session_id_context that allows to set a default
2113 value (so that you don't need SSL_set_session_id_context for each
2114 connection using the SSL_CTX).
2117 *) OAEP decoding bug fix.
2120 *) Support INSTALL_PREFIX for package builders, as proposed by
2124 *) New Configure options "threads" and "no-threads". For systems
2125 where the proper compiler options are known (currently Solaris
2126 and Linux), "threads" is the default.
2129 *) New script util/mklink.pl as a faster substitute for util/mklink.sh.
2132 *) Install various scripts to $(OPENSSLDIR)/misc, not to
2133 $(INSTALLTOP)/bin -- they shouldn't clutter directories
2134 such as /usr/local/bin.
2137 *) "make linux-shared" to build shared libraries.
2138 [Niels Poppe <niels@netbox.org>]
2140 *) New Configure option no-<cipher> (rsa, idea, rc5, ...).
2143 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
2144 extension adding in x509 utility.
2147 *) Remove NOPROTO sections and error code comments.
2150 *) Partial rewrite of the DEF file generator to now parse the ANSI
2154 *) New Configure options --prefix=DIR and --openssldir=DIR.
2157 *) Complete rewrite of the error code script(s). It is all now handled
2158 by one script at the top level which handles error code gathering,
2159 header rewriting and C source file generation. It should be much better
2160 than the old method: it now uses a modified version of Ulf's parser to
2161 read the ANSI prototypes in all header files (thus the old K&R definitions
2162 aren't needed for error creation any more) and do a better job of
2163 translating function codes into names. The old 'ASN1 error code imbedded
2164 in a comment' is no longer necessary and it doesn't use .err files which
2165 have now been deleted. Also the error code call doesn't have to appear all
2166 on one line (which resulted in some large lines...).
2169 *) Change #include filenames from <foo.h> to <openssl/foo.h>.
2172 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return
2173 0 (which usually indicates a closed connection), but continue reading.
2176 *) Fix some race conditions.
2179 *) Add support for CRL distribution points extension. Add Certificate
2180 Policies and CRL distribution points documentation.
2183 *) Move the autogenerated header file parts to crypto/opensslconf.h.
2186 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
2187 8 of keying material. Merlin has also confirmed interop with this fix
2188 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
2189 [Merlin Hughes <merlin@baltimore.ie>]
2191 *) Fix lots of warnings.
2192 [Richard Levitte <levitte@stacken.kth.se>]
2194 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
2195 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
2196 [Richard Levitte <levitte@stacken.kth.se>]
2198 *) Fix problems with sizeof(long) == 8.
2199 [Andy Polyakov <appro@fy.chalmers.se>]
2201 *) Change functions to ANSI C.
2204 *) Fix typos in error codes.
2205 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller]
2207 *) Remove defunct assembler files from Configure.
2210 *) SPARC v8 assembler BIGNUM implementation.
2211 [Andy Polyakov <appro@fy.chalmers.se>]
2213 *) Support for Certificate Policies extension: both print and set.
2214 Various additions to support the r2i method this uses.
2217 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could
2218 return a const string when you are expecting an allocated buffer.
2221 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
2222 types DirectoryString and DisplayText.
2225 *) Add code to allow r2i extensions to access the configuration database,
2226 add an LHASH database driver and add several ctx helper functions.
2229 *) Fix an evil bug in bn_expand2() which caused various BN functions to
2230 fail when they extended the size of a BIGNUM.
2233 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to
2234 support typesafe stack.
2237 *) Fix typo in SSL_[gs]et_options().
2238 [Nils Frostberg <nils@medcom.se>]
2240 *) Delete various functions and files that belonged to the (now obsolete)
2241 old X509V3 handling code.
2244 *) New Configure option "rsaref".
2247 *) Don't auto-generate pem.h.
2250 *) Introduce type-safe ASN.1 SETs.
2253 *) Convert various additional casted stacks to type-safe STACK_OF() variants.
2254 [Ben Laurie, Ralf S. Engelschall, Steve Henson]
2256 *) Introduce type-safe STACKs. This will almost certainly break lots of code
2257 that links with OpenSSL (well at least cause lots of warnings), but fear
2258 not: the conversion is trivial, and it eliminates loads of evil casts. A
2259 few STACKed things have been converted already. Feel free to convert more.
2260 In the fullness of time, I'll do away with the STACK type altogether.
2263 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate
2264 specified in <certfile> by updating the entry in the index.txt file.
2265 This way one no longer has to edit the index.txt file manually for
2266 revoking a certificate. The -revoke option does the gory details now.
2267 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall]
2269 *) Fix `openssl crl -noout -text' combination where `-noout' killed the
2270 `-text' option at all and this way the `-noout -text' combination was
2271 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'.
2272 [Ralf S. Engelschall]
2274 *) Make sure a corresponding plain text error message exists for the
2275 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
2276 verify callback function determined that a certificate was revoked.
2277 [Ralf S. Engelschall]
2279 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for
2280 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
2281 all available cipers including rc5, which was forgotten until now.
2282 In order to let the testing shell script know which algorithms
2283 are available, a new (up to now undocumented) command
2284 "openssl list-cipher-commands" is used.
2287 *) Bugfix: s_client occasionally would sleep in select() when
2288 it should have checked SSL_pending() first.
2291 *) New functions DSA_do_sign and DSA_do_verify to provide access to
2292 the raw DSA values prior to ASN.1 encoding.
2295 *) Tweaks to Configure
2296 [Niels Poppe <niels@netbox.org>]
2298 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
2302 *) New variables $(RANLIB) and $(PERL) in the Makefiles.
2305 *) New config option to avoid instructions that are illegal on the 80386.
2306 The default code is faster, but requires at least a 486.
2309 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
2310 SSL2_SERVER_VERSION (not used at all) macros, which are now the
2311 same as SSL2_VERSION anyway.
2314 *) New "-showcerts" option for s_client.
2317 *) Still more PKCS#12 integration. Add pkcs12 application to openssl
2318 application. Various cleanups and fixes.
2321 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
2322 modify error routines to work internally. Add error codes and PBE init
2323 to library startup routines.
2326 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and
2327 packing functions to asn1 and evp. Changed function names and error
2328 codes along the way.
2331 *) PKCS12 integration: and so it begins... First of several patches to
2332 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
2333 objects to objects.h
2336 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1
2337 and display support for Thawte strong extranet extension.
2340 *) Add LinuxPPC support.
2341 [Jeff Dubrule <igor@pobox.org>]
2343 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
2344 bn_div_words in alpha.s.
2345 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie]
2347 *) Make sure the RSA OAEP test is skipped under -DRSAref because
2348 OAEP isn't supported when OpenSSL is built with RSAref.
2349 [Ulf Moeller <ulf@fitug.de>]
2351 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
2352 so they no longer are missing under -DNOPROTO.
2353 [Soren S. Jorvang <soren@t.dk>]
2356 Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
2358 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
2359 doesn't work when the session is reused. Coming soon!
2362 *) Fix a security hole, that allows sessions to be reused in the wrong
2363 context thus bypassing client cert protection! All software that uses
2364 client certs and session caches in multiple contexts NEEDS PATCHING to
2365 allow session reuse! A fuller solution is in the works.
2366 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)]
2368 *) Some more source tree cleanups (removed obsolete files
2369 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
2370 permission on "config" script to be executable) and a fix for the INSTALL
2372 [Ulf Moeller <ulf@fitug.de>]
2374 *) Remove some legacy and erroneous uses of malloc, free instead of
2376 [Lennart Bang <lob@netstream.se>, with minor changes by Steve]
2378 *) Make rsa_oaep_test return non-zero on error.
2379 [Ulf Moeller <ulf@fitug.de>]
2381 *) Add support for native Solaris shared libraries. Configure
2382 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
2383 if someone would make that last step automatic.
2384 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>]
2386 *) ctx_size was not built with the right compiler during "make links". Fixed.
2389 *) Change the meaning of 'ALL' in the cipher list. It now means "everything
2390 except NULL ciphers". This means the default cipher list will no longer
2391 enable NULL ciphers. They need to be specifically enabled e.g. with
2392 the string "DEFAULT:eNULL".
2395 *) Fix to RSA private encryption routines: if p < q then it would
2396 occasionally produce an invalid result. This will only happen with
2397 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
2400 *) Be less restrictive and allow also `perl util/perlpath.pl
2401 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin',
2402 because this way one can also use an interpreter named `perl5' (which is
2403 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
2404 installed as `perl').
2405 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2407 *) Let util/clean-depend.pl work also with older Perl 5.00x versions.
2408 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2410 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
2411 advapi32.lib to Win32 build and change the pem test comparision
2412 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
2413 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
2414 and crypto/des/ede_cbcm_enc.c.
2417 *) DES quad checksum was broken on big-endian architectures. Fixed.
2420 *) Comment out two functions in bio.h that aren't implemented. Fix up the
2421 Win32 test batch file so it (might) work again. The Win32 test batch file
2422 is horrible: I feel ill....
2425 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
2426 in e_os.h. Audit of header files to check ANSI and non ANSI
2427 sections: 10 functions were absent from non ANSI section and not exported
2428 from Windows DLLs. Fixed up libeay.num for new functions.
2431 *) Make `openssl version' output lines consistent.
2432 [Ralf S. Engelschall]
2434 *) Fix Win32 symbol export lists for BIO functions: Added
2435 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
2436 to ms/libeay{16,32}.def.
2437 [Ralf S. Engelschall]
2439 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
2440 fine under Unix and passes some trivial tests I've now added. But the
2441 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
2442 added to make sure no one expects that this stuff really works in the
2443 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
2444 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
2446 [Ralf S. Engelschall]
2448 *) Fix the generation of two part addresses in perl.
2449 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie]
2451 *) Add config entry for Linux on MIPS.
2452 [John Tobey <jtobey@channel1.com>]
2454 *) Make links whenever Configure is run, unless we are on Windoze.
2457 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf.
2458 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
2462 *) Add a useful kludge to allow package maintainers to specify compiler and
2463 other platforms details on the command line without having to patch the
2464 Configure script everytime: One now can use ``perl Configure
2465 <id>:<details>'', i.e. platform ids are allowed to have details appended
2466 to them (seperated by colons). This is treated as there would be a static
2467 pre-configured entry in Configure's %table under key <id> with value
2468 <details> and ``perl Configure <id>'' is called. So, when you want to
2469 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
2470 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"''
2471 now, which overrides the FreeBSD-elf entry on-the-fly.
2472 [Ralf S. Engelschall]
2474 *) Disable new TLS1 ciphersuites by default: they aren't official yet.
2477 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
2478 on the `perl Configure ...' command line. This way one can compile
2479 OpenSSL libraries with Position Independent Code (PIC) which is needed
2480 for linking it into DSOs.
2481 [Ralf S. Engelschall]
2483 *) Remarkably, export ciphers were totally broken and no-one had noticed!
2487 *) Cleaned up the LICENSE document: The official contact for any license
2488 questions now is the OpenSSL core team under openssl-core@openssl.org.
2489 And add a paragraph about the dual-license situation to make sure people
2490 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
2491 to the OpenSSL toolkit.
2492 [Ralf S. Engelschall]
2494 *) General source tree makefile cleanups: Made `making xxx in yyy...'
2495 display consistent in the source tree and replaced `/bin/rm' by `rm'.
2496 Additonally cleaned up the `make links' target: Remove unnecessary
2497 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
2498 to speed processing and no longer clutter the display with confusing
2499 stuff. Instead only the actually done links are displayed.
2500 [Ralf S. Engelschall]
2502 *) Permit null encryption ciphersuites, used for authentication only. It used
2503 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
2504 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
2508 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
2509 signed attributes when verifying signatures (this would break them),
2510 the detached data encoding was wrong and public keys obtained using
2511 X509_get_pubkey() weren't freed.
2514 *) Add text documentation for the BUFFER functions. Also added a work around
2515 to a Win95 console bug. This was triggered by the password read stuff: the
2516 last character typed gets carried over to the next fread(). If you were
2517 generating a new cert request using 'req' for example then the last
2518 character of the passphrase would be CR which would then enter the first
2522 *) Added the new `Includes OpenSSL Cryptography Software' button as
2523 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
2524 button and can be used by applications based on OpenSSL to show the
2525 relationship to the OpenSSL project.
2526 [Ralf S. Engelschall]
2528 *) Remove confusing variables in function signatures in files
2529 ssl/ssl_lib.c and ssl/ssl.h.
2530 [Lennart Bong <lob@kulthea.stacken.kth.se>]
2532 *) Don't install bss_file.c under PREFIX/include/
2533 [Lennart Bong <lob@kulthea.stacken.kth.se>]
2535 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle
2536 functions that return function pointers and has support for NT specific
2537 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
2538 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
2539 unsigned to signed types: this was killing the Win32 compile.
2542 *) Add new certificate file to stack functions,
2543 SSL_add_dir_cert_subjects_to_stack() and
2544 SSL_add_file_cert_subjects_to_stack(). These largely supplant
2545 SSL_load_client_CA_file(), and can be used to add multiple certs easily
2546 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
2547 This means that Apache-SSL and similar packages don't have to mess around
2548 to add as many CAs as they want to the preferred list.
2551 *) Experiment with doxygen documentation. Currently only partially applied to
2553 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with
2554 openssl.doxy as the configuration file.
2557 *) Get rid of remaining C++-style comments which strict C compilers hate.
2558 [Ralf S. Engelschall, pointed out by Carlos Amengual]
2560 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
2561 compiled in by default: it has problems with large keys.
2564 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
2565 DH private keys and/or callback functions which directly correspond to
2566 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
2567 is needed for applications which have to configure certificates on a
2568 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
2570 For the RSA certificate situation is makes no difference, but
2571 for the DSA certificate situation this fixes the "no shared cipher"
2572 problem where the OpenSSL cipher selection procedure failed because the
2573 temporary keys were not overtaken from the context and the API provided
2574 no way to reconfigure them.
2575 The new functions now let applications reconfigure the stuff and they
2576 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
2577 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
2578 non-public-API function ssl_cert_instantiate() is used as a helper
2579 function and also to reduce code redundancy inside ssl_rsa.c.
2580 [Ralf S. Engelschall]
2582 *) Move s_server -dcert and -dkey options out of the undocumented feature
2583 area because they are useful for the DSA situation and should be
2584 recognized by the users.
2585 [Ralf S. Engelschall]
2587 *) Fix the cipher decision scheme for export ciphers: the export bits are
2588 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
2589 SSL_EXP_MASK. So, the original variable has to be used instead of the
2590 already masked variable.
2591 [Richard Levitte <levitte@stacken.kth.se>]
2593 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c
2594 [Richard Levitte <levitte@stacken.kth.se>]
2596 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
2597 from `int' to `unsigned int' because it's a length and initialized by
2598 EVP_DigestFinal() which expects an `unsigned int *'.
2599 [Richard Levitte <levitte@stacken.kth.se>]
2601 *) Don't hard-code path to Perl interpreter on shebang line of Configure
2602 script. Instead use the usual Shell->Perl transition trick.
2603 [Ralf S. Engelschall]
2605 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates
2606 (in addition to RSA certificates) to match the behaviour of `openssl dsa
2607 -noout -modulus' as it's already the case for `openssl rsa -noout
2608 -modulus'. For RSA the -modulus is the real "modulus" while for DSA
2609 currently the public key is printed (a decision which was already done by
2610 `openssl dsa -modulus' in the past) which serves a similar purpose.
2611 Additionally the NO_RSA no longer completely removes the whole -modulus
2612 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
2614 [Ralf S. Engelschall]
2616 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
2617 BIO. See the source (crypto/evp/bio_ok.c) for more info.
2618 [Arne Ansper <arne@ats.cyber.ee>]
2620 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs
2621 to be added. Now both 'req' and 'ca' can use new objects defined in the
2625 *) Add cool BIO that does syslog (or event log on NT).
2626 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie]
2628 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
2629 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
2630 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
2631 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
2634 *) Add preliminary config info for new extension code.
2637 *) Make RSA_NO_PADDING really use no padding.
2638 [Ulf Moeller <ulf@fitug.de>]
2640 *) Generate errors when private/public key check is done.
2643 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
2644 for some CRL extensions and new objects added.
2647 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private
2648 key usage extension and fuller support for authority key id.
2651 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
2652 padding method for RSA, which is recommended for new applications in PKCS
2653 #1 v2.0 (RFC 2437, October 1998).
2654 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
2655 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
2656 against Bleichbacher's attack on RSA.
2657 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
2660 *) Updates to the new SSL compression code
2661 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2663 *) Fix so that the version number in the master secret, when passed
2664 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
2665 (because the server will not accept higher), that the version number
2666 is 0x03,0x01, not 0x03,0x00
2667 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2669 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
2670 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
2671 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
2674 *) Support for RAW extensions where an arbitrary extension can be
2675 created by including its DER encoding. See apps/openssl.cnf for
2679 *) Make sure latest Perl versions don't interpret some generated C array
2680 code as Perl array code in the crypto/err/err_genc.pl script.
2681 [Lars Weber <3weber@informatik.uni-hamburg.de>]
2683 *) Modify ms/do_ms.bat to not generate assembly language makefiles since
2684 not many people have the assembler. Various Win32 compilation fixes and
2685 update to the INSTALL.W32 file with (hopefully) more accurate Win32
2689 *) Modify configure script 'Configure' to automatically create crypto/date.h
2690 file under Win32 and also build pem.h from pem.org. New script
2691 util/mkfiles.pl to create the MINFO file on environments that can't do a
2692 'make files': perl util/mkfiles.pl >MINFO should work.
2695 *) Major rework of DES function declarations, in the pursuit of correctness
2696 and purity. As a result, many evil casts evaporated, and some weirdness,
2697 too. You may find this causes warnings in your code. Zapping your evil
2698 casts will probably fix them. Mostly.
2701 *) Fix for a typo in asn1.h. Bug fix to object creation script
2702 obj_dat.pl. It considered a zero in an object definition to mean
2703 "end of object": none of the objects in objects.h have any zeros
2704 so it wasn't spotted.
2705 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>]
2707 *) Add support for Triple DES Cipher Block Chaining with Output Feedback
2708 Masking (CBCM). In the absence of test vectors, the best I have been able
2709 to do is check that the decrypt undoes the encrypt, so far. Send me test
2710 vectors if you have them.
2713 *) Correct calculation of key length for export ciphers (too much space was
2714 allocated for null ciphers). This has not been tested!
2717 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage
2718 message is now correct (it understands "crypto" and "ssl" on its
2719 command line). There is also now an "update" option. This will update
2720 the util/ssleay.num and util/libeay.num files with any new functions.
2722 perl util/mkdef.pl crypto ssl update
2723 it will update them.
2726 *) Overhauled the Perl interface (perl/*):
2727 - ported BN stuff to OpenSSL's different BN library
2728 - made the perl/ source tree CVS-aware
2729 - renamed the package from SSLeay to OpenSSL (the files still contain
2730 their history because I've copied them in the repository)
2731 - removed obsolete files (the test scripts will be replaced
2732 by better Test::Harness variants in the future)
2733 [Ralf S. Engelschall]
2735 *) First cut for a very conservative source tree cleanup:
2736 1. merge various obsolete readme texts into doc/ssleay.txt
2737 where we collect the old documents and readme texts.
2738 2. remove the first part of files where I'm already sure that we no
2739 longer need them because of three reasons: either they are just temporary
2740 files which were left by Eric or they are preserved original files where
2741 I've verified that the diff is also available in the CVS via "cvs diff
2742 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
2743 the crypto/md/ stuff).
2744 [Ralf S. Engelschall]
2746 *) More extension code. Incomplete support for subject and issuer alt
2747 name, issuer and authority key id. Change the i2v function parameters
2748 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
2749 what that's for :-) Fix to ASN1 macro which messed up
2750 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
2753 *) Preliminary support for ENUMERATED type. This is largely copied from the
2757 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
2758 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2760 *) Make sure `make rehash' target really finds the `openssl' program.
2761 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2763 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
2764 like to hear about it if this slows down other processors.
2767 *) Add CygWin32 platform information to Configure script.
2768 [Alan Batie <batie@aahz.jf.intel.com>]
2770 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm'
2771 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>]
2773 *) New program nseq to manipulate netscape certificate sequences
2776 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
2780 *) Fixes to BN code. Previously the default was to define BN_RECURSION
2781 but the BN code had some problems that would cause failures when
2782 doing certificate verification and some other functions.
2783 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2785 *) Add ASN1 and PEM code to support netscape certificate sequences.
2788 *) Add ASN1 and PEM code to support netscape certificate sequences.
2791 *) Add several PKIX and private extended key usage OIDs.
2794 *) Modify the 'ca' program to handle the new extension code. Modify
2795 openssl.cnf for new extension format, add comments.
2798 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
2799 and add a sample to openssl.cnf so req -x509 now adds appropriate
2803 *) Continued X509 V3 changes. Add to other makefiles, integrate with the
2804 error code, add initial support to X509_print() and x509 application.
2807 *) Takes a deep breath and start addding X509 V3 extension support code. Add
2808 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
2809 stuff is currently isolated and isn't even compiled yet.
2812 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL
2813 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
2814 Removed the versions check from X509 routines when loading extensions:
2815 this allows certain broken certificates that don't set the version
2816 properly to be processed.
2819 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
2820 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
2821 can still be regenerated with "make depend".
2824 *) Spelling mistake in C version of CAST-128.
2825 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>]
2827 *) Changes to the error generation code. The perl script err-code.pl
2828 now reads in the old error codes and retains the old numbers, only
2829 adding new ones if necessary. It also only changes the .err files if new
2830 codes are added. The makefiles have been modified to only insert errors
2831 when needed (to avoid needlessly modifying header files). This is done
2832 by only inserting errors if the .err file is newer than the auto generated
2833 C file. To rebuild all the error codes from scratch (the old behaviour)
2834 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
2835 or delete all the .err files.
2838 *) CAST-128 was incorrectly implemented for short keys. The C version has
2839 been fixed, but is untested. The assembler versions are also fixed, but
2840 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
2841 to regenerate it if needed.
2842 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
2843 Hagino <itojun@kame.net>]
2845 *) File was opened incorrectly in randfile.c.
2846 [Ulf Möller <ulf@fitug.de>]
2848 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print
2849 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
2850 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
2851 al: it's just almost always a UTCTime. Note this patch adds new error
2852 codes so do a "make errors" if there are problems.
2855 *) Correct Linux 1 recognition in config.
2856 [Ulf Möller <ulf@fitug.de>]
2858 *) Remove pointless MD5 hash when using DSA keys in ca.
2859 [Anonymous <nobody@replay.com>]
2861 *) Generate an error if given an empty string as a cert directory. Also
2862 generate an error if handed NULL (previously returned 0 to indicate an
2863 error, but didn't set one).
2864 [Ben Laurie, reported by Anonymous <nobody@replay.com>]
2866 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
2869 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
2870 parameters. This was causing a warning which killed off the Win32 compile.
2873 *) Remove C++ style comments from crypto/bn/bn_local.h.
2874 [Neil Costigan <neil.costigan@celocom.com>]
2876 *) The function OBJ_txt2nid was broken. It was supposed to return a nid
2877 based on a text string, looking up short and long names and finally
2878 "dot" format. The "dot" format stuff didn't work. Added new function
2879 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
2880 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
2881 OID is not part of the table.
2884 *) Add prototypes to X509 lookup/verify methods, fixing a bug in
2885 X509_LOOKUP_by_alias().
2888 *) Sort openssl functions by name.
2891 *) Get the gendsa program working (hopefully) and add it to app list. Remove
2892 encryption from sample DSA keys (in case anyone is interested the password
2896 *) Make _all_ *_free functions accept a NULL pointer.
2897 [Frans Heymans <fheymans@isaserver.be>]
2899 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use
2901 [Anonymous <nobody@replay.com>]
2903 *) s_server should send the CAfile as acceptable CAs, not its own cert.
2904 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
2906 *) Don't blow it for numeric -newkey arguments to apps/req.
2907 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
2909 *) Temp key "for export" tests were wrong in s3_srvr.c.
2910 [Anonymous <nobody@replay.com>]
2912 *) Add prototype for temp key callback functions
2913 SSL_CTX_set_tmp_{rsa,dh}_callback().
2916 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
2917 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
2920 *) X509_name_add_entry() freed the wrong thing after an error.
2921 [Arne Ansper <arne@ats.cyber.ee>]
2923 *) rsa_eay.c would attempt to free a NULL context.
2924 [Arne Ansper <arne@ats.cyber.ee>]
2926 *) BIO_s_socket() had a broken should_retry() on Windoze.
2927 [Arne Ansper <arne@ats.cyber.ee>]
2929 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
2930 [Arne Ansper <arne@ats.cyber.ee>]
2932 *) Make sure the already existing X509_STORE->depth variable is initialized
2933 in X509_STORE_new(), but document the fact that this variable is still
2934 unused in the certificate verification process.
2935 [Ralf S. Engelschall]
2937 *) Fix the various library and apps files to free up pkeys obtained from
2938 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
2941 *) Fix reference counting in X509_PUBKEY_get(). This makes
2942 demos/maurice/example2.c work, amongst others, probably.
2943 [Steve Henson and Ben Laurie]
2945 *) First cut of a cleanup for apps/. First the `ssleay' program is now named
2946 `openssl' and second, the shortcut symlinks for the `openssl <command>'
2947 are no longer created. This way we have a single and consistent command
2948 line interface `openssl <command>', similar to `cvs <command>'.
2949 [Ralf S. Engelschall, Paul Sutton and Ben Laurie]
2951 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
2952 BIT STRING wrapper always have zero unused bits.
2955 *) Add CA.pl, perl version of CA.sh, add extended key usage OID.
2958 *) Make the top-level INSTALL documentation easier to understand.
2961 *) Makefiles updated to exit if an error occurs in a sub-directory
2962 make (including if user presses ^C) [Paul Sutton]
2964 *) Make Montgomery context stuff explicit in RSA data structure.
2967 *) Fix build order of pem and err to allow for generated pem.h.
2970 *) Fix renumbering bug in X509_NAME_delete_entry().
2973 *) Enhanced the err-ins.pl script so it makes the error library number
2974 global and can add a library name. This is needed for external ASN1 and
2975 other error libraries.
2978 *) Fixed sk_insert which never worked properly.
2981 *) Fix ASN1 macros so they can handle indefinite length construted
2982 EXPLICIT tags. Some non standard certificates use these: they can now
2986 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
2987 into a single doc/ssleay.txt bundle. This way the information is still
2988 preserved but no longer messes up this directory. Now it's new room for
2989 the new set of documenation files.
2990 [Ralf S. Engelschall]
2992 *) SETs were incorrectly DER encoded. This was a major pain, because they
2993 shared code with SEQUENCEs, which aren't coded the same. This means that
2994 almost everything to do with SETs or SEQUENCEs has either changed name or
2995 number of arguments.
2996 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>]
2998 *) Fix test data to work with the above.
3001 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
3002 was already fixed by Eric for 0.9.1 it seems.
3003 [Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>]
3005 *) Autodetect FreeBSD3.
3008 *) Fix various bugs in Configure. This affects the following platforms:
3012 unixware-2.0-pentium
3016 *) Eliminate generated files from CVS. Reorder tests to regenerate files
3017 before they are needed.
3020 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy).
3024 Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
3026 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
3027 changed SSLeay to OpenSSL in version strings.
3028 [Ralf S. Engelschall]
3030 *) Some fixups to the top-level documents.
3033 *) Fixed the nasty bug where rsaref.h was not found under compile-time
3034 because the symlink to include/ was missing.
3035 [Ralf S. Engelschall]
3037 *) Incorporated the popular no-RSA/DSA-only patches
3038 which allow to compile a RSA-free SSLeay.
3039 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall]
3041 *) Fixed nasty rehash problem under `make -f Makefile.ssl links'
3042 when "ssleay" is still not found.
3043 [Ralf S. Engelschall]
3045 *) Added more platforms to Configure: Cray T3E, HPUX 11,
3046 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>]
3048 *) Updated the README file.
3049 [Ralf S. Engelschall]
3051 *) Added various .cvsignore files in the CVS repository subdirs
3052 to make a "cvs update" really silent.
3053 [Ralf S. Engelschall]
3055 *) Recompiled the error-definition header files and added
3056 missing symbols to the Win32 linker tables.
3057 [Ralf S. Engelschall]
3059 *) Cleaned up the top-level documents;
3060 o new files: CHANGES and LICENSE
3061 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
3062 o merged COPYRIGHT into LICENSE
3063 o removed obsolete TODO file
3064 o renamed MICROSOFT to INSTALL.W32
3065 [Ralf S. Engelschall]
3067 *) Removed dummy files from the 0.9.1b source tree:
3068 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
3069 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
3070 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
3071 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
3072 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
3073 [Ralf S. Engelschall]
3075 *) Added various platform portability fixes.
3078 *) The Genesis of the OpenSSL rpject:
3079 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
3080 Young and Tim J. Hudson created while they were working for C2Net until
3082 [The OpenSSL Project]
3085 Changes between 0.9.0b and 0.9.1b [not released]
3087 *) Updated a few CA certificates under certs/
3090 *) Changed some BIGNUM api stuff.
3093 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
3094 DGUX x86, Linux Alpha, etc.
3097 *) New COMP library [crypto/comp/] for SSL Record Layer Compression:
3098 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
3102 *) Add -strparse option to asn1pars program which parses nested
3104 [Dr Stephen Henson <shenson@bigfoot.com>]
3106 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
3109 *) DSA fix for "ca" program.
3112 *) Added "-genkey" option to "dsaparam" program.
3115 *) Added RIPE MD160 (rmd160) message digest.
3118 *) Added -a (all) option to "ssleay version" command.
3121 *) Added PLATFORM define which is the id given to Configure.
3124 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
3127 *) Extended the ASN.1 parser routines.
3130 *) Extended BIO routines to support REUSEADDR, seek, tell, etc.
3133 *) Added a BN_CTX to the BN library.
3136 *) Fixed the weak key values in DES library
3139 *) Changed API in EVP library for cipher aliases.
3142 *) Added support for RC2/64bit cipher.
3145 *) Converted the lhash library to the crypto/mem.c functions.
3148 *) Added more recognized ASN.1 object ids.
3151 *) Added more RSA padding checks for SSL/TLS.
3154 *) Added BIO proxy/filter functionality.
3157 *) Added extra_certs to SSL_CTX which can be used
3158 send extra CA certificates to the client in the CA cert chain sending
3159 process. It can be configured with SSL_CTX_add_extra_chain_cert().
3162 *) Now Fortezza is denied in the authentication phase because
3163 this is key exchange mechanism is not supported by SSLeay at all.
3166 *) Additional PKCS1 checks.
3169 *) Support the string "TLSv1" for all TLS v1 ciphers.
3172 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
3173 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
3176 *) Fixed a few memory leaks.
3179 *) Fixed various code and comment typos.
3182 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0
3183 bytes sent in the client random.
3184 [Edward Bishop <ebishop@spyglass.com>]