5 Changes between 0.9.5a and 0.9.6 [xx XXX 2000]
7 *) New openssl application 'rsautl'. This utility can be
8 used for low level RSA operations. DER public key
9 BIO/fp routines also added.
12 *) New Configure entry and patches for compiling on QNX 4.
13 [Andreas Schneider <andreas@ds3.etech.fh-hamburg.de>]
15 *) A demo state-machine implementation was sponsored by
16 Nuron (http://www.nuron.com/) and is now available in
20 *) New options added to the 'dgst' utility for signature
21 generation and verification.
24 *) Unrecognized PKCS#7 content types are now handled via a
25 catch all ASN1_TYPE structure. This allows unsupported
26 types to be stored as a "blob" and an application can
27 encode and decode it manually.
30 *) Fix various signed/unsigned issues to make a_strex.c
32 [Oscar Jacobsson <oscar.jacobsson@celocom.com>]
34 *) ASN1 fixes. i2d_ASN1_OBJECT was not returning the correct
35 length if passed a buffer. ASN1_INTEGER_to_BN failed
36 if passed a NULL BN and its argument was negative.
37 [Steve Henson, pointed out by Sven Heiberg <sven@tartu.cyber.ee>]
39 *) Modification to PKCS#7 encoding routines to output definite
40 length encoding. Since currently the whole structures are in
41 memory there's not real point in using indefinite length
42 constructed encoding. However if OpenSSL is compiled with
43 the flag PKCS7_INDEFINITE_ENCODING the old form is used.
46 *) Added BIO_vprintf() and BIO_vsnprintf().
49 *) Added more prefixes to parse for in the the strings written
50 through a logging bio, to cover all the levels that are available
51 through syslog. The prefixes are now:
53 PANIC, EMERG, EMR => LOG_EMERG
54 ALERT, ALR => LOG_ALERT
57 WARNING, WARN, WAR => LOG_WARNING
58 NOTICE, NOTE, NOT => LOG_NOTICE
60 DEBUG, DBG => LOG_DEBUG
62 and as before, if none of those prefixes are present at the
63 beginning of the string, LOG_ERR is chosen.
65 On Win32, the LOG_* levels are mapped according to this:
67 LOG_EMERG, LOG_ALERT, LOG_CRIT, LOG_ERR => EVENTLOG_ERROR_TYPE
68 LOG_WARNING => EVENTLOG_WARNING_TYPE
69 LOG_NOTICE, LOG_INFO, LOG_DEBUG => EVENTLOG_INFORMATION_TYPE
73 *) Made it possible to reconfigure with just the configuration
74 argument "reconf" or "reconfigure". The command line arguments
75 are stored in Makefile.ssl in the variable CONFIGURE_ARGS,
76 and are retrieved from there when reconfiguring.
80 [Assar Westerlund <assar@sics.se>, Richard Levitte]
82 *) Add the arguments -CAfile and -CApath to the pkcs12 utility.
85 *) The obj_dat.pl script was messing up the sorting of object
86 names. The reason was that it compared the quoted version
87 of strings as a result "OCSP" > "OCSP Signing" because
88 " > SPACE. Changed script to store unquoted versions of
89 names and add quotes on output. It was also omitting some
90 names from the lookup table if they were given a default
91 value (that is if SN is missing it is given the same
92 value as LN and vice versa), these are now added on the
93 grounds that if an object has a name we should be able to
94 look it up. Finally added warning output when duplicate
95 short or long names are found.
98 *) Changes needed for Tandem NSK.
99 [Scott Uroff scott@xypro.com]
101 *) Fix SSL 2.0 rollback checking: Due to an off-by-one error in
102 RSA_padding_check_SSLv23(), special padding was never detected
103 and thus the SSL 3.0/TLS 1.0 countermeasure against protocol
104 version rollback attacks was not effective.
106 In s23_clnt.c, don't use special rollback-attack detection padding
107 (RSA_SSLV23_PADDING) if SSL 2.0 is the only protocol enabled in the
108 client; similarly, in s23_srvr.c, don't do the rollback check if
109 SSL 2.0 is the only protocol enabled in the server.
112 *) Make it possible to get hexdumps of unprintable data with 'openssl
113 asn1parse'. By implication, the functions ASN1_parse_dump() and
114 BIO_dump_indent() are added.
117 *) New functions ASN1_STRING_print_ex() and X509_NAME_print_ex()
118 these print out strings and name structures based on various
119 flags including RFC2253 support and proper handling of
120 multibyte characters. Added options to the 'x509' utility
121 to allow the various flags to be set.
124 *) Various fixes to use ASN1_TIME instead of ASN1_UTCTIME.
125 Also change the functions X509_cmp_current_time() and
126 X509_gmtime_adj() work with an ASN1_TIME structure,
127 this will enable certificates using GeneralizedTime in validity
131 *) Make the NEG_PUBKEY_BUG code (which tolerates invalid
132 negative public key encodings) on by default,
133 NO_NEG_PUBKEY_BUG can be set to disable it.
136 *) New function c2i_ASN1_OBJECT() which acts on ASN1_OBJECT
137 content octets. An i2c_ASN1_OBJECT is unnecessary because
138 the encoding can be trivially obtained from the structure.
141 *) crypto/err.c locking bugfix: Use write locks (CRYPTO_w_[un]lock),
142 not read locks (CRYPTO_r_[un]lock).
145 *) A first attempt at creating official support for shared
146 libraries through configuration. I've kept it so the
147 default is static libraries only, and the OpenSSL programs
148 are always statically linked for now, but there are
149 preparations for dynamic linking in place.
150 This has been tested on Linux and True64.
153 *) Randomness polling function for Win9x, as described in:
154 Peter Gutmann, Software Generation of Practically Strong
158 *) Fix so PRNG is seeded in req if using an already existing
162 *) New options to smime application. -inform and -outform
163 allow alternative formats for the S/MIME message including
164 PEM and DER. The -content option allows the content to be
165 specified separately. This should allow things like Netscape
166 form signing output easier to verify.
169 *) Fix the ASN1 encoding of tags using the 'long form'.
172 *) New ASN1 functions, i2c_* and c2i_* for INTEGER and BIT
173 STRING types. These convert content octets to and from the
174 underlying type. The actual tag and length octets are
175 already assumed to have been read in and checked. These
176 are needed because all other string types have virtually
177 identical handling apart from the tag. By having versions
178 of the ASN1 functions that just operate on content octets
179 IMPLICIT tagging can be handled properly. It also allows
180 the ASN1_ENUMERATED code to be cut down because ASN1_ENUMERATED
181 and ASN1_INTEGER are identical apart from the tag.
184 *) Change the handling of OID objects as follows:
186 - New object identifiers are inserted in objects.txt, following
187 the syntax given in objects.README.
188 - objects.pl is used to process obj_mac.num and create a new
190 - obj_dat.pl is used to create a new obj_dat.h, using the data in
193 This is currently kind of a hack, and the perl code in objects.pl
194 isn't very elegant, but it works as I intended. The simplest way
195 to check that it worked correctly is to look in obj_dat.h and
196 check the array nid_objs and make sure the objects haven't moved
197 around (this is important!). Additions are OK, as well as
198 consistent name changes.
201 *) Add BSD-style MD5-based passwords to 'openssl passwd' (option '-1').
204 *) Addition of the command line parameter '-rand file' to 'openssl req'.
205 The given file adds to whatever has already been seeded into the
206 random pool through the RANDFILE configuration file option or
207 environment variable, or the default random state file.
210 *) mkstack.pl now sorts each macro group into lexical order.
211 Previously the output order depended on the order the files
212 appeared in the directory, resulting in needless rewriting
216 *) Patches to make OpenSSL compile under Win32 again. Mostly
217 work arounds for the VC++ problem that it treats func() as
218 func(void). Also stripped out the parts of mkdef.pl that
219 added extra typesafe functions: these no longer exist.
222 *) Reorganisation of the stack code. The macros are now all
223 collected in safestack.h . Each macro is defined in terms of
224 a "stack macro" of the form SKM_<name>(type, a, b). The
225 DEBUG_SAFESTACK is now handled in terms of function casts,
226 this has the advantage of retaining type safety without the
227 use of additional functions. If DEBUG_SAFESTACK is not defined
228 then the non typesafe macros are used instead. Also modified the
229 mkstack.pl script to handle the new form. Needs testing to see
230 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK
231 the default if no major problems. Similar behaviour for ASN1_SET_OF
235 *) When some versions of IIS use the 'NET' form of private key the
236 key derivation algorithm is different. Normally MD5(password) is
237 used as a 128 bit RC4 key. In the modified case
238 MD5(MD5(password) + "SGCKEYSALT") is used insted. Added some
239 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same
240 as the old Netscape_RSA functions except they have an additional
241 'sgckey' parameter which uses the modified algorithm. Also added
242 an -sgckey command line option to the rsa utility. Thanks to
243 Adrian Peck <bertie@ncipher.com> for posting details of the modified
244 algorithm to openssl-dev.
247 *) The evp_local.h macros were using 'c.##kname' which resulted in
248 invalid expansion on some systems (SCO 5.0.5 for example).
249 Corrected to 'c.kname'.
250 [Phillip Porch <root@theporch.com>]
252 *) New X509_get1_email() and X509_REQ_get1_email() functions that return
253 a STACK of email addresses from a certificate or request, these look
254 in the subject name and the subject alternative name extensions and
255 omit any duplicate addresses.
258 *) Re-implement BN_mod_exp2_mont using independent (and larger) windows.
259 This makes DSA verification about 2 % faster.
262 *) Increase maximum window size in BN_mod_exp_... to 6 bits instead of 5
263 (meaning that now 2^5 values will be precomputed, which is only 4 KB
264 plus overhead for 1024 bit moduli).
265 This makes exponentiations about 0.5 % faster for 1024 bit
266 exponents (as measured by "openssl speed rsa2048").
269 *) Rename memory handling macros to avoid conflicts with other
271 Malloc => OPENSSL_malloc
272 Malloc_locked => OPENSSL_malloc_locked
273 Realloc => OPENSSL_realloc
277 *) New function BN_mod_exp_mont_word for small bases (roughly 15%
278 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange).
282 [John Jarvie <jjarvie@newsguy.com>]
284 *) The type-safe stack code has been rejigged. It is now only compiled
285 in when OpenSSL is configured with the DEBUG_SAFESTACK option and
286 by default all type-specific stack functions are "#define"d back to
287 standard stack functions. This results in more streamlined output
288 but retains the type-safety checking possibilities of the original
292 *) The STACK code has been cleaned up, and certain type declarations
293 that didn't make a lot of sense have been brought in line. This has
294 also involved a cleanup of sorts in safestack.h to more correctly
295 map type-safe stack functions onto their plain stack counterparts.
296 This work has also resulted in a variety of "const"ifications of
297 lots of the code, especially "_cmp" operations which should normally
298 be prototyped with "const" parameters anyway.
301 *) When generating bytes for the first time in md_rand.c, 'stir the pool'
302 by seeding with STATE_SIZE dummy bytes (with zero entropy count).
303 (The PRNG state consists of two parts, the large pool 'state' and 'md',
304 where all of 'md' is used each time the PRNG is used, but 'state'
305 is used only indexed by a cyclic counter. As entropy may not be
306 well distributed from the beginning, 'md' is important as a
307 chaining variable. However, the output function chains only half
308 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains
309 all of 'md', and seeding with STATE_SIZE dummy bytes will result
310 in all of 'state' being rewritten, with the new values depending
311 on virtually all of 'md'. This overcomes the 80 bit limitation.)
314 *) In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when
315 the handshake is continued after ssl_verify_cert_chain();
316 otherwise, if SSL_VERIFY_NONE is set, remaining error codes
317 can lead to 'unexplainable' connection aborts later.
318 [Bodo Moeller; problem tracked down by Lutz Jaenicke]
320 *) Major EVP API cipher revision.
321 Add hooks for extra EVP features. This allows various cipher
322 parameters to be set in the EVP interface. Support added for variable
323 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and
324 setting of RC2 and RC5 parameters.
326 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length
329 Remove lots of duplicated code from the EVP library. For example *every*
330 cipher init() function handles the 'iv' in the same way according to the
331 cipher mode. They also all do nothing if the 'key' parameter is NULL and
332 for CFB and OFB modes they zero ctx->num.
334 New functionality allows removal of S/MIME code RC2 hack.
336 Most of the routines have the same form and so can be declared in terms
339 By shifting this to the top level EVP_CipherInit() it can be removed from
340 all individual ciphers. If the cipher wants to handle IVs or keys
341 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT
344 Change lots of functions like EVP_EncryptUpdate() to now return a
345 value: although software versions of the algorithms cannot fail
346 any installed hardware versions can.
349 *) Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if
350 this option is set, tolerate broken clients that send the negotiated
351 protocol version number instead of the requested protocol version
355 *) Call dh_tmp_cb (set by ..._TMP_DH_CB) with correct 'is_export' flag;
356 i.e. non-zero for export ciphersuites, zero otherwise.
357 Previous versions had this flag inverted, inconsistent with
358 rsa_tmp_cb (..._TMP_RSA_CB).
359 [Bodo Moeller; problem reported by Amit Chopra]
361 *) Add missing DSA library text string. Work around for some IIS
362 key files with invalid SEQUENCE encoding.
365 *) Add a document (doc/standards.txt) that list all kinds of standards
366 and so on that are implemented in OpenSSL.
369 *) Enhance c_rehash script. Old version would mishandle certificates
370 with the same subject name hash and wouldn't handle CRLs at all.
371 Added -fingerprint option to crl utility, to support new c_rehash
375 *) Eliminate non-ANSI declarations in crypto.h and stack.h.
378 *) Fix for SSL server purpose checking. Server checking was
379 rejecting certificates which had extended key usage present
380 but no ssl client purpose.
381 [Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>]
383 *) Make PKCS#12 code work with no password. The PKCS#12 spec
384 is a little unclear about how a blank password is handled.
385 Since the password in encoded as a BMPString with terminating
386 double NULL a zero length password would end up as just the
387 double NULL. However no password at all is different and is
388 handled differently in the PKCS#12 key generation code. NS
389 treats a blank password as zero length. MSIE treats it as no
390 password on export: but it will try both on import. We now do
391 the same: PKCS12_parse() tries zero length and no password if
392 the password is set to "" or NULL (NULL is now a valid password:
393 it wasn't before) as does the pkcs12 application.
396 *) Bugfixes in apps/x509.c: Avoid a memory leak; and don't use
397 perror when PEM_read_bio_X509_REQ fails, the error message must
398 be obtained from the error queue.
401 *) Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing
402 it in ERR_remove_state if appropriate, and change ERR_get_state
403 accordingly to avoid race conditions (this is necessary because
404 thread_hash is no longer constant once set).
407 *) Bugfix for linux-elf makefile.one.
410 *) RSA_get_default_method() will now cause a default
411 RSA_METHOD to be chosen if one doesn't exist already.
412 Previously this was only set during a call to RSA_new()
413 or RSA_new_method(NULL) meaning it was possible for
414 RSA_get_default_method() to return NULL.
417 *) Added native name translation to the existing DSO code
418 that will convert (if the flag to do so is set) filenames
419 that are sufficiently small and have no path information
420 into a canonical native form. Eg. "blah" converted to
421 "libblah.so" or "blah.dll" etc.
424 *) New function ERR_error_string_n(e, buf, len) which is like
425 ERR_error_string(e, buf), but writes at most 'len' bytes
426 including the 0 terminator. For ERR_error_string_n, 'buf'
428 [Damien Miller <djm@mindrot.org>, Bodo Moeller]
430 *) CONF library reworked to become more general. A new CONF
431 configuration file reader "class" is implemented as well as a
432 new functions (NCONF_*, for "New CONF") to handle it. The now
433 old CONF_* functions are still there, but are reimplemented to
434 work in terms of the new functions. Also, a set of functions
435 to handle the internal storage of the configuration data is
436 provided to make it easier to write new configuration file
437 reader "classes" (I can definitely see something reading a
438 configuration file in XML format, for example), called _CONF_*,
439 or "the configuration storage API"...
441 The new configuration file reading functions are:
443 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio,
444 NCONF_get_section, NCONF_get_string, NCONF_get_numbre
446 NCONF_default, NCONF_WIN32
448 NCONF_dump_fp, NCONF_dump_bio
450 NCONF_default and NCONF_WIN32 are method (or "class") choosers,
451 NCONF_new creates a new CONF object. This works in the same way
452 as other interfaces in OpenSSL, like the BIO interface.
453 NCONF_dump_* dump the internal storage of the configuration file,
454 which is useful for debugging. All other functions take the same
455 arguments as the old CONF_* functions wth the exception of the
456 first that must be a `CONF *' instead of a `LHASH *'.
458 To make it easer to use the new classes with the old CONF_* functions,
459 the function CONF_set_default_method is provided.
462 *) Add '-tls1' option to 'openssl ciphers', which was already
463 mentioned in the documentation but had not been implemented.
464 (This option is not yet really useful because even the additional
465 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.)
468 *) Initial DSO code added into libcrypto for letting OpenSSL (and
469 OpenSSL-based applications) load shared libraries and bind to
470 them in a portable way.
471 [Geoff Thorpe, with contributions from Richard Levitte]
473 Changes between 0.9.5 and 0.9.5a [1 Apr 2000]
475 *) Make sure _lrotl and _lrotr are only used with MSVC.
477 *) Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status
478 (the default implementation of RAND_status).
480 *) Rename openssl x509 option '-crlext', which was added in 0.9.5,
481 to '-clrext' (= clear extensions), as intended and documented.
482 [Bodo Moeller; inconsistency pointed out by Michael Attili
485 *) Fix for HMAC. It wasn't zeroing the rest of the block if the key length
486 was larger than the MD block size.
487 [Steve Henson, pointed out by Yost William <YostW@tce.com>]
489 *) Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument
490 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set()
491 using the passed key: if the passed key was a private key the result
492 of X509_print(), for example, would be to print out all the private key
496 *) des_quad_cksum() byte order bug fix.
497 [Ulf Möller, using the problem description in krb4-0.9.7, where
498 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>]
500 *) Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly
502 [Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>]
504 *) For easily testing in shell scripts whether some command
505 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX'
506 returns with exit code 0 iff no command of the given name is available.
507 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases,
508 the output goes to stdout and nothing is printed to stderr.
509 Additional arguments are always ignored.
511 Since for each cipher there is a command of the same name,
512 the 'no-cipher' compilation switches can be tested this way.
514 ('openssl no-XXX' is not able to detect pseudo-commands such
515 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.)
518 *) Update test suite so that 'make test' succeeds in 'no-rsa' configuration.
521 *) For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE
522 is set; it will be thrown away anyway because each handshake creates
524 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition
525 to parameters -- in previous versions (since OpenSSL 0.9.3) the
526 'default key' from SSL_CTX_set_tmp_dh would always be lost, meanining
527 you effectivly got SSL_OP_SINGLE_DH_USE when using this macro.
530 *) New s_client option -ign_eof: EOF at stdin is ignored, and
531 'Q' and 'R' lose their special meanings (quit/renegotiate).
532 This is part of what -quiet does; unlike -quiet, -ign_eof
533 does not suppress any output.
536 *) Add compatibility options to the purpose and trust code. The
537 purpose X509_PURPOSE_ANY is "any purpose" which automatically
538 accepts a certificate or CA, this was the previous behaviour,
539 with all the associated security issues.
541 X509_TRUST_COMPAT is the old trust behaviour: only and
542 automatically trust self signed roots in certificate store. A
543 new trust setting X509_TRUST_DEFAULT is used to specify that
544 a purpose has no associated trust setting and it should instead
545 use the value in the default purpose.
548 *) Fix the PKCS#8 DSA private key code so it decodes keys again
549 and fix a memory leak.
552 *) In util/mkerr.pl (which implements 'make errors'), preserve
553 reason strings from the previous version of the .c file, as
554 the default to have only downcase letters (and digits) in
555 automatically generated reasons codes is not always appropriate.
558 *) In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table
559 using strerror. Previously, ERR_reason_error_string() returned
560 library names as reason strings for SYSerr; but SYSerr is a special
561 case where small numbers are errno values, not library numbers.
564 *) Add '-dsaparam' option to 'openssl dhparam' application. This
565 converts DSA parameters into DH parameters. (When creating parameters,
566 DSA_generate_parameters is used.)
569 *) Include 'length' (recommended exponent length) in C code generated
570 by 'openssl dhparam -C'.
573 *) The second argument to set_label in perlasm was already being used
574 so couldn't be used as a "file scope" flag. Moved to third argument
578 *) In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes
579 instead of RAND_bytes for encryption IVs and salts.
582 *) Include RAND_status() into RAND_METHOD instead of implementing
583 it only for md_rand.c Otherwise replacing the PRNG by calling
584 RAND_set_rand_method would be impossible.
587 *) Don't let DSA_generate_key() enter an infinite loop if the random
588 number generation fails.
591 *) New 'rand' application for creating pseudo-random output.
594 *) Added configuration support for Linux/IA64
595 [Rolf Haberrecker <rolf@suse.de>]
597 *) Assembler module support for Mingw32.
600 *) Shared library support for HPUX (in shlib/).
601 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous]
603 *) Shared library support for Solaris gcc.
604 [Lutz Behnke <behnke@trustcenter.de>]
606 Changes between 0.9.4 and 0.9.5 [28 Feb 2000]
608 *) PKCS7_encrypt() was adding text MIME headers twice because they
609 were added manually and by SMIME_crlf_copy().
612 *) In bntest.c don't call BN_rand with zero bits argument.
613 [Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>]
615 *) BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n]
616 case was implemented. This caused BN_div_recp() to fail occasionally.
619 *) Add an optional second argument to the set_label() in the perl
620 assembly language builder. If this argument exists and is set
621 to 1 it signals that the assembler should use a symbol whose
622 scope is the entire file, not just the current function. This
623 is needed with MASM which uses the format label:: for this scope.
624 [Steve Henson, pointed out by Peter Runestig <peter@runestig.com>]
626 *) Change the ASN1 types so they are typedefs by default. Before
627 almost all types were #define'd to ASN1_STRING which was causing
628 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING)
632 *) Change names of new functions to the new get1/get0 naming
633 convention: After 'get1', the caller owns a reference count
634 and has to call ..._free; 'get0' returns a pointer to some
635 data structure without incrementing reference counters.
636 (Some of the existing 'get' functions increment a reference
637 counter, some don't.)
638 Similarly, 'set1' and 'add1' functions increase reference
639 counters or duplicate objects.
642 *) Allow for the possibility of temp RSA key generation failure:
643 the code used to assume it always worked and crashed on failure.
646 *) Fix potential buffer overrun problem in BIO_printf().
647 [Ulf Möller, using public domain code by Patrick Powell; problem
648 pointed out by David Sacerdote <das33@cornell.edu>]
650 *) Support EGD <http://www.lothar.com/tech/crypto/>. New functions
651 RAND_egd() and RAND_status(). In the command line application,
652 the EGD socket can be specified like a seed file using RANDFILE
656 *) Allow the string CERTIFICATE to be tolerated in PKCS#7 structures.
657 Some CAs (e.g. Verisign) distribute certificates in this form.
660 *) Remove the SSL_ALLOW_ADH compile option and set the default cipher
661 list to exclude them. This means that no special compilation option
662 is needed to use anonymous DH: it just needs to be included in the
666 *) Change the EVP_MD_CTX_type macro so its meaning consistent with
667 EVP_MD_type. The old functionality is available in a new macro called
668 EVP_MD_md(). Change code that uses it and update docs.
671 *) ..._ctrl functions now have corresponding ..._callback_ctrl functions
672 where the 'void *' argument is replaced by a function pointer argument.
673 Previously 'void *' was abused to point to functions, which works on
674 many platforms, but is not correct. As these functions are usually
675 called by macros defined in OpenSSL header files, most source code
676 should work without changes.
679 *) <openssl/opensslconf.h> (which is created by Configure) now contains
680 sections with information on -D... compiler switches used for
681 compiling the library so that applications can see them. To enable
682 one of these sections, a pre-processor symbol OPENSSL_..._DEFINES
683 must be defined. E.g.,
684 #define OPENSSL_ALGORITHM_DEFINES
685 #include <openssl/opensslconf.h>
686 defines all pertinent NO_<algo> symbols, such as NO_IDEA, NO_RSA, etc.
687 [Richard Levitte, Ulf and Bodo Möller]
689 *) Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS
693 *) Change the 'other' type in certificate aux info to a STACK_OF
694 X509_ALGOR. Although not an AlgorithmIdentifier as such it has
695 the required ASN1 format: arbitrary types determined by an OID.
698 *) Add some PEM_write_X509_REQ_NEW() functions and a command line
699 argument to 'req'. This is not because the function is newer or
700 better than others it just uses the work 'NEW' in the certificate
701 request header lines. Some software needs this.
704 *) Reorganise password command line arguments: now passwords can be
705 obtained from various sources. Delete the PEM_cb function and make
706 it the default behaviour: i.e. if the callback is NULL and the
707 usrdata argument is not NULL interpret it as a null terminated pass
708 phrase. If usrdata and the callback are NULL then the pass phrase
709 is prompted for as usual.
712 *) Add support for the Compaq Atalla crypto accelerator. If it is installed,
713 the support is automatically enabled. The resulting binaries will
714 autodetect the card and use it if present.
715 [Ben Laurie and Compaq Inc.]
717 *) Work around for Netscape hang bug. This sends certificate request
718 and server done in one record. Since this is perfectly legal in the
719 SSL/TLS protocol it isn't a "bug" option and is on by default. See
720 the bugs/SSLv3 entry for more info.
723 *) HP-UX tune-up: new unified configs, HP C compiler bug workaround.
726 *) Add -rand argument to smime and pkcs12 applications and read/write
730 *) New 'passwd' tool for crypt(3) and apr1 password hashes.
733 *) Add command line password options to the remaining applications.
736 *) Bug fix for BN_div_recp() for numerators with an even number of
740 *) More tests in bntest.c, and changed test_bn output.
743 *) ./config recognizes MacOS X now.
746 *) Bug fix for BN_div() when the first words of num and divsor are
747 equal (it gave wrong results if (rem=(n1-q*d0)&BN_MASK2) < d0).
750 *) Add support for various broken PKCS#8 formats, and command line
751 options to produce them.
754 *) New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to
755 get temporary BIGNUMs from a BN_CTX.
758 *) Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont()
762 *) Change the SSLeay_add_all_*() functions to OpenSSL_add_all_*() and
763 include a #define from the old name to the new. The original intent
764 was that statically linked binaries could for example just call
765 SSLeay_add_all_ciphers() to just add ciphers to the table and not
766 link with digests. This never worked becayse SSLeay_add_all_digests()
767 and SSLeay_add_all_ciphers() were in the same source file so calling
768 one would link with the other. They are now in separate source files.
771 *) Add a new -notext option to 'ca' and a -pubkey option to 'spkac'.
774 *) Use a less unusual form of the Miller-Rabin primality test (it used
775 a binary algorithm for exponentiation integrated into the Miller-Rabin
776 loop, our standard modexp algorithms are faster).
779 *) Support for the EBCDIC character set completed.
780 [Martin Kraemer <Martin.Kraemer@Mch.SNI.De>]
782 *) Source code cleanups: use const where appropriate, eliminate casts,
783 use void * instead of char * in lhash.
786 *) Bugfix: ssl3_send_server_key_exchange was not restartable
787 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of
788 this the server could overwrite ephemeral keys that the client
792 *) Turn DSA_is_prime into a macro that calls BN_is_prime,
793 using 50 iterations of the Rabin-Miller test.
795 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50
796 iterations of the Rabin-Miller test as required by the appendix
797 to FIPS PUB 186[-1]) instead of DSA_is_prime.
798 As BN_is_prime_fasttest includes trial division, DSA parameter
799 generation becomes much faster.
801 This implies a change for the callback functions in DSA_is_prime
802 and DSA_generate_parameters: The callback function is called once
803 for each positive witness in the Rabin-Miller test, not just
804 occasionally in the inner loop; and the parameters to the
805 callback function now provide an iteration count for the outer
806 loop rather than for the current invocation of the inner loop.
807 DSA_generate_parameters additionally can call the callback
808 function with an 'iteration count' of -1, meaning that a
809 candidate has passed the trial division test (when q is generated
810 from an application-provided seed, trial division is skipped).
813 *) New function BN_is_prime_fasttest that optionally does trial
814 division before starting the Rabin-Miller test and has
815 an additional BN_CTX * argument (whereas BN_is_prime always
816 has to allocate at least one BN_CTX).
817 'callback(1, -1, cb_arg)' is called when a number has passed the
818 trial division stage.
821 *) Fix for bug in CRL encoding. The validity dates weren't being handled
825 *) New -pkcs12 option to CA.pl script to write out a PKCS#12 file.
828 *) New function BN_pseudo_rand().
831 *) Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable)
832 bignum version of BN_from_montgomery() with the working code from
833 SSLeay 0.9.0 (the word based version is faster anyway), and clean up
837 *) Avoid a race condition in s2_clnt.c (function get_server_hello) that
838 made it impossible to use the same SSL_SESSION data structure in
839 SSL2 clients in multiple threads.
842 *) The return value of RAND_load_file() no longer counts bytes obtained
843 by stat(). RAND_load_file(..., -1) is new and uses the complete file
844 to seed the PRNG (previously an explicit byte count was required).
845 [Ulf Möller, Bodo Möller]
847 *) Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
848 used (char *) instead of (void *) and had casts all over the place.
851 *) Make BN_generate_prime() return NULL on error if ret!=NULL.
854 *) Retain source code compatibility for BN_prime_checks macro:
855 BN_is_prime(..., BN_prime_checks, ...) now uses
856 BN_prime_checks_for_size to determine the appropriate number of
857 Rabin-Miller iterations.
860 *) Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
861 DH_CHECK_P_NOT_SAFE_PRIME.
862 (Check if this is true? OpenPGP calls them "strong".)
865 *) Merge the functionality of "dh" and "gendh" programs into a new program
866 "dhparam". The old programs are retained for now but will handle DH keys
867 (instead of parameters) in future.
870 *) Make the ciphers, s_server and s_client programs check the return values
871 when a new cipher list is set.
874 *) Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
875 ciphers. Before when the 56bit ciphers were enabled the sorting was
878 The syntax for the cipher sorting has been extended to support sorting by
879 cipher-strength (using the strength_bits hard coded in the tables).
880 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
882 Fix a bug in the cipher-command parser: when supplying a cipher command
883 string with an "undefined" symbol (neither command nor alphanumeric
884 [A-Za-z0-9], ssl_set_cipher_list used to hang in an endless loop. Now
887 Due to the strength-sorting extension, the code of the
888 ssl_create_cipher_list() function was completely rearranged. I hope that
889 the readability was also increased :-)
890 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
892 *) Minor change to 'x509' utility. The -CAcreateserial option now uses 1
893 for the first serial number and places 2 in the serial number file. This
894 avoids problems when the root CA is created with serial number zero and
895 the first user certificate has the same issuer name and serial number
899 *) Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
900 the new code. Add documentation for this stuff.
903 *) Changes to X509_ATTRIBUTE utilities. These have been renamed from
904 X509_*() to X509at_*() on the grounds that they don't handle X509
905 structures and behave in an analagous way to the X509v3 functions:
906 they shouldn't be called directly but wrapper functions should be used
909 So we also now have some wrapper functions that call the X509at functions
910 when passed certificate requests. (TO DO: similar things can be done with
911 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
912 things. Some of these need some d2i or i2d and print functionality
913 because they handle more complex structures.)
916 *) Add missing #ifndefs that caused missing symbols when building libssl
917 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
919 [Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller]
921 *) Precautions against using the PRNG uninitialized: RAND_bytes() now
922 has a return value which indicates the quality of the random data
923 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
924 error queue. New function RAND_pseudo_bytes() generates output that is
925 guaranteed to be unique but not unpredictable. RAND_add is like
926 RAND_seed, but takes an extra argument for an entropy estimate
927 (RAND_seed always assumes full entropy).
930 *) Do more iterations of Rabin-Miller probable prime test (specifically,
931 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
932 instead of only 2 for all lengths; see BN_prime_checks_for_size definition
933 in crypto/bn/bn_prime.c for the complete table). This guarantees a
934 false-positive rate of at most 2^-80 for random input.
937 *) Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
940 *) New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain
941 in the 0.9.5 release), this returns the chain
942 from an X509_CTX structure with a dup of the stack and all
943 the X509 reference counts upped: so the stack will exist
944 after X509_CTX_cleanup() has been called. Modify pkcs12.c
947 Also make SSL_SESSION_print() print out the verify return
951 *) Add manpage for the pkcs12 command. Also change the default
952 behaviour so MAC iteration counts are used unless the new
953 -nomaciter option is used. This improves file security and
954 only older versions of MSIE (4.0 for example) need it.
957 *) Honor the no-xxx Configure options when creating .DEF files.
960 *) Add PKCS#10 attributes to field table: challengePassword,
961 unstructuredName and unstructuredAddress. These are taken from
962 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
963 international characters are used.
965 More changes to X509_ATTRIBUTE code: allow the setting of types
966 based on strings. Remove the 'loc' parameter when adding
967 attributes because these will be a SET OF encoding which is sorted
971 *) Initial changes to the 'req' utility to allow request generation
972 automation. This will allow an application to just generate a template
973 file containing all the field values and have req construct the
976 Initial support for X509_ATTRIBUTE handling. Stacks of these are
977 used all over the place including certificate requests and PKCS#7
978 structures. They are currently handled manually where necessary with
979 some primitive wrappers for PKCS#7. The new functions behave in a
980 manner analogous to the X509 extension functions: they allow
981 attributes to be looked up by NID and added.
983 Later something similar to the X509V3 code would be desirable to
984 automatically handle the encoding, decoding and printing of the
985 more complex types. The string types like challengePassword can
986 be handled by the string table functions.
988 Also modified the multi byte string table handling. Now there is
989 a 'global mask' which masks out certain types. The table itself
990 can use the flag STABLE_NO_MASK to ignore the mask setting: this
991 is useful when for example there is only one permissible type
992 (as in countryName) and using the mask might result in no valid
996 *) Clean up 'Finished' handling, and add functions SSL_get_finished and
997 SSL_get_peer_finished to allow applications to obtain the latest
998 Finished messages sent to the peer or expected from the peer,
999 respectively. (SSL_get_peer_finished is usually the Finished message
1000 actually received from the peer, otherwise the protocol will be aborted.)
1002 As the Finished message are message digests of the complete handshake
1003 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
1004 be used for external authentication procedures when the authentication
1005 provided by SSL/TLS is not desired or is not enough.
1008 *) Enhanced support for Alpha Linux is added. Now ./config checks if
1009 the host supports BWX extension and if Compaq C is present on the
1010 $PATH. Just exploiting of the BWX extension results in 20-30%
1011 performance kick for some algorithms, e.g. DES and RC4 to mention
1012 a couple. Compaq C in turn generates ~20% faster code for MD5 and
1016 *) Add support for MS "fast SGC". This is arguably a violation of the
1017 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
1018 weak crypto and after checking the certificate is SGC a second one
1019 with strong crypto. MS SGC stops the first handshake after receiving
1020 the server certificate message and sends a second client hello. Since
1021 a server will typically do all the time consuming operations before
1022 expecting any further messages from the client (server key exchange
1023 is the most expensive) there is little difference between the two.
1025 To get OpenSSL to support MS SGC we have to permit a second client
1026 hello message after we have sent server done. In addition we have to
1027 reset the MAC if we do get this second client hello.
1030 *) Add a function 'd2i_AutoPrivateKey()' this will automatically decide
1031 if a DER encoded private key is RSA or DSA traditional format. Changed
1032 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
1033 format DER encoded private key. Newer code should use PKCS#8 format which
1034 has the key type encoded in the ASN1 structure. Added DER private key
1035 support to pkcs8 application.
1038 *) SSL 3/TLS 1 servers now don't request certificates when an anonymous
1039 ciphersuites has been selected (as required by the SSL 3/TLS 1
1040 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
1041 is set, we interpret this as a request to violate the specification
1042 (the worst that can happen is a handshake failure, and 'correct'
1043 behaviour would result in a handshake failure anyway).
1046 *) In SSL_CTX_add_session, take into account that there might be multiple
1047 SSL_SESSION structures with the same session ID (e.g. when two threads
1048 concurrently obtain them from an external cache).
1049 The internal cache can handle only one SSL_SESSION with a given ID,
1050 so if there's a conflict, we now throw out the old one to achieve
1054 *) Add OIDs for idea and blowfish in CBC mode. This will allow both
1055 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
1056 some routines that use cipher OIDs: some ciphers do not have OIDs
1057 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
1061 *) Simplify the trust setting structure and code. Now we just have
1062 two sequences of OIDs for trusted and rejected settings. These will
1063 typically have values the same as the extended key usage extension
1064 and any application specific purposes.
1066 The trust checking code now has a default behaviour: it will just
1067 check for an object with the same NID as the passed id. Functions can
1068 be provided to override either the default behaviour or the behaviour
1069 for a given id. SSL client, server and email already have functions
1070 in place for compatibility: they check the NID and also return "trusted"
1071 if the certificate is self signed.
1074 *) Add d2i,i2d bio/fp functions for PrivateKey: these convert the
1075 traditional format into an EVP_PKEY structure.
1078 *) Add a password callback function PEM_cb() which either prompts for
1079 a password if usr_data is NULL or otherwise assumes it is a null
1080 terminated password. Allow passwords to be passed on command line
1081 environment or config files in a few more utilities.
1084 *) Add a bunch of DER and PEM functions to handle PKCS#8 format private
1085 keys. Add some short names for PKCS#8 PBE algorithms and allow them
1086 to be specified on the command line for the pkcs8 and pkcs12 utilities.
1087 Update documentation.
1090 *) Support for ASN1 "NULL" type. This could be handled before by using
1091 ASN1_TYPE but there wasn't any function that would try to read a NULL
1092 and produce an error if it couldn't. For compatibility we also have
1093 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
1094 don't allocate anything because they don't need to.
1097 *) Initial support for MacOS is now provided. Examine INSTALL.MacOS
1099 [Andy Polyakov, Roy Woods <roy@centicsystems.ca>]
1101 *) Rebuild of the memory allocation routines used by OpenSSL code and
1102 possibly others as well. The purpose is to make an interface that
1103 provide hooks so anyone can build a separate set of allocation and
1104 deallocation routines to be used by OpenSSL, for example memory
1105 pool implementations, or something else, which was previously hard
1106 since Malloc(), Realloc() and Free() were defined as macros having
1107 the values malloc, realloc and free, respectively (except for Win32
1108 compilations). The same is provided for memory debugging code.
1109 OpenSSL already comes with functionality to find memory leaks, but
1110 this gives people a chance to debug other memory problems.
1112 With these changes, a new set of functions and macros have appeared:
1114 CRYPTO_set_mem_debug_functions() [F]
1115 CRYPTO_get_mem_debug_functions() [F]
1116 CRYPTO_dbg_set_options() [F]
1117 CRYPTO_dbg_get_options() [F]
1118 CRYPTO_malloc_debug_init() [M]
1120 The memory debug functions are NULL by default, unless the library
1121 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
1122 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
1123 gives the standard debugging functions that come with OpenSSL) or
1124 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
1125 provided by the library user) must be used. When the standard
1126 debugging functions are used, CRYPTO_dbg_set_options can be used to
1127 request additional information:
1128 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
1129 the CRYPTO_MDEBUG_xxx macro when compiling the library.
1131 Also, things like CRYPTO_set_mem_functions will always give the
1132 expected result (the new set of functions is used for allocation
1133 and deallocation) at all times, regardless of platform and compiler
1136 To finish it up, some functions that were never use in any other
1137 way than through macros have a new API and new semantic:
1140 CRYPTO_dbg_realloc()
1143 All macros of value have retained their old syntax.
1144 [Richard Levitte and Bodo Moeller]
1146 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
1147 ordering of SMIMECapabilities wasn't in "strength order" and there
1148 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
1152 *) Some ASN1 types with illegal zero length encoding (INTEGER,
1153 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
1154 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson]
1156 *) Merge in my S/MIME library for OpenSSL. This provides a simple
1157 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
1158 functionality to handle multipart/signed properly) and a utility
1159 called 'smime' to call all this stuff. This is based on code I
1160 originally wrote for Celo who have kindly allowed it to be
1161 included in OpenSSL.
1164 *) Add variants des_set_key_checked and des_set_key_unchecked of
1165 des_set_key (aka des_key_sched). Global variable des_check_key
1166 decides which of these is called by des_set_key; this way
1167 des_check_key behaves as it always did, but applications and
1168 the library itself, which was buggy for des_check_key == 1,
1169 have a cleaner way to pick the version they need.
1172 *) New function PKCS12_newpass() which changes the password of a
1176 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and
1177 dynamic mix. In both cases the ids can be used as an index into the
1178 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
1179 functions so they accept a list of the field values and the
1180 application doesn't need to directly manipulate the X509_TRUST
1184 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
1188 *) Modify the way the V3 extension code looks up extensions. This now
1189 works in a similar way to the object code: we have some "standard"
1190 extensions in a static table which is searched with OBJ_bsearch()
1191 and the application can add dynamic ones if needed. The file
1192 crypto/x509v3/ext_dat.h now has the info: this file needs to be
1193 updated whenever a new extension is added to the core code and kept
1194 in ext_nid order. There is a simple program 'tabtest.c' which checks
1195 this. New extensions are not added too often so this file can readily
1196 be maintained manually.
1198 There are two big advantages in doing things this way. The extensions
1199 can be looked up immediately and no longer need to be "added" using
1200 X509V3_add_standard_extensions(): this function now does nothing.
1201 [Side note: I get *lots* of email saying the extension code doesn't
1202 work because people forget to call this function]
1203 Also no dynamic allocation is done unless new extensions are added:
1204 so if we don't add custom extensions there is no need to call
1205 X509V3_EXT_cleanup().
1208 *) Modify enc utility's salting as follows: make salting the default. Add a
1209 magic header, so unsalted files fail gracefully instead of just decrypting
1210 to garbage. This is because not salting is a big security hole, so people
1211 should be discouraged from doing it.
1214 *) Fixes and enhancements to the 'x509' utility. It allowed a message
1215 digest to be passed on the command line but it only used this
1216 parameter when signing a certificate. Modified so all relevant
1217 operations are affected by the digest parameter including the
1218 -fingerprint and -x509toreq options. Also -x509toreq choked if a
1219 DSA key was used because it didn't fix the digest.
1222 *) Initial certificate chain verify code. Currently tests the untrusted
1223 certificates for consistency with the verify purpose (which is set
1224 when the X509_STORE_CTX structure is set up) and checks the pathlength.
1226 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
1227 this is because it will reject chains with invalid extensions whereas
1228 every previous version of OpenSSL and SSLeay made no checks at all.
1230 Trust code: checks the root CA for the relevant trust settings. Trust
1231 settings have an initial value consistent with the verify purpose: e.g.
1232 if the verify purpose is for SSL client use it expects the CA to be
1233 trusted for SSL client use. However the default value can be changed to
1234 permit custom trust settings: one example of this would be to only trust
1235 certificates from a specific "secure" set of CAs.
1237 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
1238 which should be used for version portability: especially since the
1239 verify structure is likely to change more often now.
1241 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
1242 to set them. If not set then assume SSL clients will verify SSL servers
1245 Two new options to the verify program: -untrusted allows a set of
1246 untrusted certificates to be passed in and -purpose which sets the
1247 intended purpose of the certificate. If a purpose is set then the
1248 new chain verify code is used to check extension consistency.
1251 *) Support for the authority information access extension.
1254 *) Modify RSA and DSA PEM read routines to transparently handle
1255 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
1256 public keys in a format compatible with certificate
1257 SubjectPublicKeyInfo structures. Unfortunately there were already
1258 functions called *_PublicKey_* which used various odd formats so
1259 these are retained for compatibility: however the DSA variants were
1260 never in a public release so they have been deleted. Changed dsa/rsa
1261 utilities to handle the new format: note no releases ever handled public
1262 keys so we should be OK.
1264 The primary motivation for this change is to avoid the same fiasco
1265 that dogs private keys: there are several incompatible private key
1266 formats some of which are standard and some OpenSSL specific and
1267 require various evil hacks to allow partial transparent handling and
1268 even then it doesn't work with DER formats. Given the option anything
1269 other than PKCS#8 should be dumped: but the other formats have to
1270 stay in the name of compatibility.
1272 With public keys and the benefit of hindsight one standard format
1273 is used which works with EVP_PKEY, RSA or DSA structures: though
1274 it clearly returns an error if you try to read the wrong kind of key.
1276 Added a -pubkey option to the 'x509' utility to output the public key.
1277 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*()
1278 (renamed to EVP_PKEY_get1_*() in the OpenSSL 0.9.5 release) and add
1279 EVP_PKEY_rset_*() functions (renamed to EVP_PKEY_set1_*())
1280 that do the same as the EVP_PKEY_assign_*() except they up the
1281 reference count of the added key (they don't "swallow" the
1285 *) Fixes to crypto/x509/by_file.c the code to read in certificates and
1286 CRLs would fail if the file contained no certificates or no CRLs:
1287 added a new function to read in both types and return the number
1288 read: this means that if none are read it will be an error. The
1289 DER versions of the certificate and CRL reader would always fail
1290 because it isn't possible to mix certificates and CRLs in DER format
1291 without choking one or the other routine. Changed this to just read
1292 a certificate: this is the best we can do. Also modified the code
1293 in apps/verify.c to take notice of return codes: it was previously
1294 attempting to read in certificates from NULL pointers and ignoring
1295 any errors: this is one reason why the cert and CRL reader seemed
1296 to work. It doesn't check return codes from the default certificate
1297 routines: these may well fail if the certificates aren't installed.
1300 *) Code to support otherName option in GeneralName.
1303 *) First update to verify code. Change the verify utility
1304 so it warns if it is passed a self signed certificate:
1305 for consistency with the normal behaviour. X509_verify
1306 has been modified to it will now verify a self signed
1307 certificate if *exactly* the same certificate appears
1308 in the store: it was previously impossible to trust a
1309 single self signed certificate. This means that:
1310 openssl verify ss.pem
1311 now gives a warning about a self signed certificate but
1312 openssl verify -CAfile ss.pem ss.pem
1316 *) For servers, store verify_result in SSL_SESSION data structure
1317 (and add it to external session representation).
1318 This is needed when client certificate verifications fails,
1319 but an application-provided verification callback (set by
1320 SSL_CTX_set_cert_verify_callback) allows accepting the session
1321 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
1322 but returns 1): When the session is reused, we have to set
1323 ssl->verify_result to the appropriate error code to avoid
1325 [Bodo Moeller, problem pointed out by Lutz Jaenicke]
1327 *) Fix a bug in the new PKCS#7 code: it didn't consider the
1328 case in PKCS7_dataInit() where the signed PKCS7 structure
1329 didn't contain any existing data because it was being created.
1330 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson]
1332 *) Add a salt to the key derivation routines in enc.c. This
1333 forms the first 8 bytes of the encrypted file. Also add a
1334 -S option to allow a salt to be input on the command line.
1337 *) New function X509_cmp(). Oddly enough there wasn't a function
1338 to compare two certificates. We do this by working out the SHA1
1339 hash and comparing that. X509_cmp() will be needed by the trust
1343 *) SSL_get1_session() is like SSL_get_session(), but increments
1344 the reference count in the SSL_SESSION returned.
1345 [Geoff Thorpe <geoff@eu.c2.net>]
1347 *) Fix for 'req': it was adding a null to request attributes.
1348 Also change the X509_LOOKUP and X509_INFO code to handle
1349 certificate auxiliary information.
1352 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document
1356 *) Add the possibility to add extra information to the memory leak
1357 detecting output, to form tracebacks, showing from where each
1358 allocation was originated: CRYPTO_push_info("constant string") adds
1359 the string plus current file name and line number to a per-thread
1360 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
1361 is like calling CYRPTO_pop_info() until the stack is empty.
1362 Also updated memory leak detection code to be multi-thread-safe.
1365 *) Add options -text and -noout to pkcs7 utility and delete the
1366 encryption options which never did anything. Update docs.
1369 *) Add options to some of the utilities to allow the pass phrase
1370 to be included on either the command line (not recommended on
1371 OSes like Unix) or read from the environment. Update the
1372 manpages and fix a few bugs.
1375 *) Add a few manpages for some of the openssl commands.
1378 *) Fix the -revoke option in ca. It was freeing up memory twice,
1379 leaking and not finding already revoked certificates.
1382 *) Extensive changes to support certificate auxiliary information.
1383 This involves the use of X509_CERT_AUX structure and X509_AUX
1384 functions. An X509_AUX function such as PEM_read_X509_AUX()
1385 can still read in a certificate file in the usual way but it
1386 will also read in any additional "auxiliary information". By
1387 doing things this way a fair degree of compatibility can be
1388 retained: existing certificates can have this information added
1389 using the new 'x509' options.
1391 Current auxiliary information includes an "alias" and some trust
1392 settings. The trust settings will ultimately be used in enhanced
1393 certificate chain verification routines: currently a certificate
1394 can only be trusted if it is self signed and then it is trusted
1398 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD).
1399 The problem was that one of the replacement routines had not been working
1400 since SSLeay releases. For now the offending routine has been replaced
1401 with non-optimised assembler. Even so, this now gives around 95%
1402 performance improvement for 1024 bit RSA signs.
1405 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2
1406 handling. Most clients have the effective key size in bits equal to
1407 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
1408 A few however don't do this and instead use the size of the decrypted key
1409 to determine the RC2 key length and the AlgorithmIdentifier to determine
1410 the effective key length. In this case the effective key length can still
1411 be 40 bits but the key length can be 168 bits for example. This is fixed
1412 by manually forcing an RC2 key into the EVP_PKEY structure because the
1413 EVP code can't currently handle unusual RC2 key sizes: it always assumes
1414 the key length and effective key length are equal.
1417 *) Add a bunch of functions that should simplify the creation of
1418 X509_NAME structures. Now you should be able to do:
1419 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
1420 and have it automatically work out the correct field type and fill in
1421 the structures. The more adventurous can try:
1422 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
1423 and it will (hopefully) work out the correct multibyte encoding.
1426 *) Change the 'req' utility to use the new field handling and multibyte
1427 copy routines. Before the DN field creation was handled in an ad hoc
1428 way in req, ca, and x509 which was rather broken and didn't support
1429 BMPStrings or UTF8Strings. Since some software doesn't implement
1430 BMPStrings or UTF8Strings yet, they can be enabled using the config file
1431 using the dirstring_type option. See the new comment in the default
1432 openssl.cnf for more info.
1435 *) Make crypto/rand/md_rand.c more robust:
1436 - Assure unique random numbers after fork().
1437 - Make sure that concurrent threads access the global counter and
1438 md serializably so that we never lose entropy in them
1439 or use exactly the same state in multiple threads.
1440 Access to the large state is not always serializable because
1441 the additional locking could be a performance killer, and
1442 md should be large enough anyway.
1445 *) New file apps/app_rand.c with commonly needed functionality
1446 for handling the random seed file.
1448 Use the random seed file in some applications that previously did not:
1450 dsaparam -genkey (which also ignored its '-rand' option),
1453 x509 (when signing).
1454 Except on systems with /dev/urandom, it is crucial to have a random
1455 seed file at least for key creation, DSA signing, and for DH exchanges;
1456 for RSA signatures we could do without one.
1458 gendh and gendsa (unlike genrsa) used to read only the first byte
1459 of each file listed in the '-rand' option. The function as previously
1460 found in genrsa is now in app_rand.c and is used by all programs
1461 that support '-rand'.
1464 *) In RAND_write_file, use mode 0600 for creating files;
1465 don't just chmod when it may be too late.
1468 *) Report an error from X509_STORE_load_locations
1469 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
1472 *) New function ASN1_mbstring_copy() this copies a string in either
1473 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
1474 into an ASN1_STRING type. A mask of permissible types is passed
1475 and it chooses the "minimal" type to use or an error if not type
1479 *) Add function equivalents to the various macros in asn1.h. The old
1480 macros are retained with an M_ prefix. Code inside the library can
1481 use the M_ macros. External code (including the openssl utility)
1482 should *NOT* in order to be "shared library friendly".
1485 *) Add various functions that can check a certificate's extensions
1486 to see if it usable for various purposes such as SSL client,
1487 server or S/MIME and CAs of these types. This is currently
1488 VERY EXPERIMENTAL but will ultimately be used for certificate chain
1489 verification. Also added a -purpose flag to x509 utility to
1490 print out all the purposes.
1493 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated
1497 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search
1498 for, obtain and decode and extension and obtain its critical flag.
1499 This allows all the necessary extension code to be handled in a
1500 single function call.
1503 *) RC4 tune-up featuring 30-40% performance improvement on most RISC
1504 platforms. See crypto/rc4/rc4_enc.c for further details.
1507 *) New -noout option to asn1parse. This causes no output to be produced
1508 its main use is when combined with -strparse and -out to extract data
1509 from a file (which may not be in ASN.1 format).
1512 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer
1513 when producing the local key id.
1514 [Richard Levitte <levitte@stacken.kth.se>]
1516 *) New option -dhparam in s_server. This allows a DH parameter file to be
1517 stated explicitly. If it is not stated then it tries the first server
1518 certificate file. The previous behaviour hard coded the filename
1522 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow
1523 a public key to be input or output. For example:
1524 openssl rsa -in key.pem -pubout -out pubkey.pem
1525 Also added necessary DSA public key functions to handle this.
1528 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
1529 in the message. This was handled by allowing
1530 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
1531 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>]
1533 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
1534 to the end of the strings whereas this didn't. This would cause problems
1535 if strings read with d2i_ASN1_bytes() were later modified.
1536 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>]
1538 *) Fix for base64 decode bug. When a base64 bio reads only one line of
1539 data and it contains EOF it will end up returning an error. This is
1540 caused by input 46 bytes long. The cause is due to the way base64
1541 BIOs find the start of base64 encoded data. They do this by trying a
1542 trial decode on each line until they find one that works. When they
1543 do a flag is set and it starts again knowing it can pass all the
1544 data directly through the decoder. Unfortunately it doesn't reset
1545 the context it uses. This means that if EOF is reached an attempt
1546 is made to pass two EOFs through the context and this causes the
1547 resulting error. This can also cause other problems as well. As is
1548 usual with these problems it takes *ages* to find and the fix is
1549 trivial: move one line.
1550 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ]
1552 *) Ugly workaround to get s_client and s_server working under Windows. The
1553 old code wouldn't work because it needed to select() on sockets and the
1554 tty (for keypresses and to see if data could be written). Win32 only
1555 supports select() on sockets so we select() with a 1s timeout on the
1556 sockets and then see if any characters are waiting to be read, if none
1557 are present then we retry, we also assume we can always write data to
1558 the tty. This isn't nice because the code then blocks until we've
1559 received a complete line of data and it is effectively polling the
1560 keyboard at 1s intervals: however it's quite a bit better than not
1561 working at all :-) A dedicated Windows application might handle this
1562 with an event loop for example.
1565 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
1566 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
1567 will be called when RSA_sign() and RSA_verify() are used. This is useful
1568 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
1569 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
1570 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
1571 This necessitated the support of an extra signature type NID_md5_sha1
1572 for SSL signatures and modifications to the SSL library to use it instead
1573 of calling RSA_public_decrypt() and RSA_private_encrypt().
1576 *) Add new -verify -CAfile and -CApath options to the crl program, these
1577 will lookup a CRL issuers certificate and verify the signature in a
1578 similar way to the verify program. Tidy up the crl program so it
1579 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
1580 less strict. It will now permit CRL extensions even if it is not
1581 a V2 CRL: this will allow it to tolerate some broken CRLs.
1584 *) Initialize all non-automatic variables each time one of the openssl
1585 sub-programs is started (this is necessary as they may be started
1586 multiple times from the "OpenSSL>" prompt).
1587 [Lennart Bang, Bodo Moeller]
1589 *) Preliminary compilation option RSA_NULL which disables RSA crypto without
1590 removing all other RSA functionality (this is what NO_RSA does). This
1591 is so (for example) those in the US can disable those operations covered
1592 by the RSA patent while allowing storage and parsing of RSA keys and RSA
1596 *) Non-copying interface to BIO pairs.
1597 (still largely untested)
1600 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive
1601 ASCII string. This was handled independently in various places before.
1604 *) New functions UTF8_getc() and UTF8_putc() that parse and generate
1605 UTF8 strings a character at a time.
1608 *) Use client_version from client hello to select the protocol
1609 (s23_srvr.c) and for RSA client key exchange verification
1610 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
1613 *) Add various utility functions to handle SPKACs, these were previously
1614 handled by poking round in the structure internals. Added new function
1615 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
1616 print, verify and generate SPKACs. Based on an original idea from
1617 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
1620 *) RIPEMD160 is operational on all platforms and is back in 'make test'.
1623 *) Allow the config file extension section to be overwritten on the
1624 command line. Based on an original idea from Massimiliano Pala
1625 <madwolf@comune.modena.it>. The new option is called -extensions
1626 and can be applied to ca, req and x509. Also -reqexts to override
1627 the request extensions in req and -crlexts to override the crl extensions
1631 *) Add new feature to the SPKAC handling in ca. Now you can include
1632 the same field multiple times by preceding it by "XXXX." for example:
1635 this is the same syntax as used in the req config file.
1638 *) Allow certificate extensions to be added to certificate requests. These
1639 are specified in a 'req_extensions' option of the req section of the
1640 config file. They can be printed out with the -text option to req but
1641 are otherwise ignored at present.
1644 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
1645 data read consists of only the final block it would not decrypted because
1646 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
1647 A misplaced 'break' also meant the decrypted final block might not be
1648 copied until the next read.
1651 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
1652 a few extra parameters to the DH structure: these will be useful if
1653 for example we want the value of 'q' or implement X9.42 DH.
1656 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and
1657 provides hooks that allow the default DSA functions or functions on a
1658 "per key" basis to be replaced. This allows hardware acceleration and
1659 hardware key storage to be handled without major modification to the
1660 library. Also added low level modexp hooks and CRYPTO_EX structure and
1661 associated functions.
1664 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
1665 as "read only": it can't be written to and the buffer it points to will
1666 not be freed. Reading from a read only BIO is much more efficient than
1667 a normal memory BIO. This was added because there are several times when
1668 an area of memory needs to be read from a BIO. The previous method was
1669 to create a memory BIO and write the data to it, this results in two
1670 copies of the data and an O(n^2) reading algorithm. There is a new
1671 function BIO_new_mem_buf() which creates a read only memory BIO from
1672 an area of memory. Also modified the PKCS#7 routines to use read only
1676 *) Bugfix: ssl23_get_client_hello did not work properly when called in
1677 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
1678 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
1679 but a retry condition occured while trying to read the rest.
1682 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as
1683 NID_pkcs7_encrypted by default: this was wrong since this should almost
1684 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
1685 the encrypted data type: this is a more sensible place to put it and it
1686 allows the PKCS#12 code to be tidied up that duplicated this
1690 *) Changed obj_dat.pl script so it takes its input and output files on
1691 the command line. This should avoid shell escape redirection problems
1695 *) Initial support for certificate extension requests, these are included
1696 in things like Xenroll certificate requests. Included functions to allow
1697 extensions to be obtained and added.
1700 *) -crlf option to s_client and s_server for sending newlines as
1701 CRLF (as required by many protocols).
1704 Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
1706 *) Install libRSAglue.a when OpenSSL is built with RSAref.
1707 [Ralf S. Engelschall]
1709 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency.
1710 [Andrija Antonijevic <TheAntony2@bigfoot.com>]
1712 *) Fix -startdate and -enddate (which was missing) arguments to 'ca'
1716 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as
1717 DH parameters/keys (q is lost during that conversion, but the resulting
1718 DH parameters contain its length).
1720 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
1721 much faster than DH_generate_parameters (which creates parameters
1722 where p = 2*q + 1), and also the smaller q makes DH computations
1723 much more efficient (160-bit exponentiation instead of 1024-bit
1724 exponentiation); so this provides a convenient way to support DHE
1725 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
1726 utter importance to use
1727 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
1729 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
1730 when such DH parameters are used, because otherwise small subgroup
1731 attacks may become possible!
1734 *) Avoid memory leak in i2d_DHparams.
1737 *) Allow the -k option to be used more than once in the enc program:
1738 this allows the same encrypted message to be read by multiple recipients.
1741 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
1742 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
1743 it will always use the numerical form of the OID, even if it has a short
1747 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
1748 method only got called if p,q,dmp1,dmq1,iqmp components were present,
1749 otherwise bn_mod_exp was called. In the case of hardware keys for example
1750 no private key components need be present and it might store extra data
1751 in the RSA structure, which cannot be accessed from bn_mod_exp.
1752 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
1753 private key operations.
1756 *) Added support for SPARC Linux.
1759 *) pem_password_cb function type incompatibly changed from
1760 typedef int pem_password_cb(char *buf, int size, int rwflag);
1762 ....(char *buf, int size, int rwflag, void *userdata);
1763 so that applications can pass data to their callbacks:
1764 The PEM[_ASN1]_{read,write}... functions and macros now take an
1765 additional void * argument, which is just handed through whenever
1766 the password callback is called.
1767 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller]
1769 New function SSL_CTX_set_default_passwd_cb_userdata.
1771 Compatibility note: As many C implementations push function arguments
1772 onto the stack in reverse order, the new library version is likely to
1773 interoperate with programs that have been compiled with the old
1774 pem_password_cb definition (PEM_whatever takes some data that
1775 happens to be on the stack as its last argument, and the callback
1776 just ignores this garbage); but there is no guarantee whatsoever that
1779 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
1780 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
1781 problems not only on Windows, but also on some Unix platforms.
1782 To avoid problematic command lines, these definitions are now in an
1783 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
1784 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
1787 *) MIPS III/IV assembler module is reimplemented.
1790 *) More DES library cleanups: remove references to srand/rand and
1791 delete an unused file.
1794 *) Add support for the the free Netwide assembler (NASM) under Win32,
1795 since not many people have MASM (ml) and it can be hard to obtain.
1796 This is currently experimental but it seems to work OK and pass all
1797 the tests. Check out INSTALL.W32 for info.
1800 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
1801 without temporary keys kept an extra copy of the server key,
1802 and connections with temporary keys did not free everything in case
1806 *) New function RSA_check_key and new openssl rsa option -check
1807 for verifying the consistency of RSA keys.
1808 [Ulf Moeller, Bodo Moeller]
1810 *) Various changes to make Win32 compile work:
1811 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
1812 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
1813 comparison" warnings.
1814 3. Add sk_<TYPE>_sort to DEF file generator and do make update.
1817 *) Add a debugging option to PKCS#5 v2 key generation function: when
1818 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
1819 derived keys are printed to stderr.
1822 *) Copy the flags in ASN1_STRING_dup().
1823 [Roman E. Pavlov <pre@mo.msk.ru>]
1825 *) The x509 application mishandled signing requests containing DSA
1826 keys when the signing key was also DSA and the parameters didn't match.
1828 It was supposed to omit the parameters when they matched the signing key:
1829 the verifying software was then supposed to automatically use the CA's
1830 parameters if they were absent from the end user certificate.
1832 Omitting parameters is no longer recommended. The test was also
1833 the wrong way round! This was probably due to unusual behaviour in
1834 EVP_cmp_parameters() which returns 1 if the parameters match.
1835 This meant that parameters were omitted when they *didn't* match and
1836 the certificate was useless. Certificates signed with 'ca' didn't have
1838 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>]
1840 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
1841 The interface is as follows:
1842 Applications can use
1843 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
1844 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
1845 "off" is now the default.
1846 The library internally uses
1847 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
1848 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
1849 to disable memory-checking temporarily.
1851 Some inconsistent states that previously were possible (and were
1852 even the default) are now avoided.
1854 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
1855 with each memory chunk allocated; this is occasionally more helpful
1856 than just having a counter.
1858 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
1860 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
1864 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX),
1865 which largely parallels "options", but is for changing API behaviour,
1866 whereas "options" are about protocol behaviour.
1867 Initial "mode" flags are:
1869 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
1870 a single record has been written.
1871 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
1872 retries use the same buffer location.
1873 (But all of the contents must be
1877 *) Bugfix: SSL_set_mode ignored its parameter, only SSL_CTX_set_mode
1880 *) Fix problems with no-hmac etc.
1881 [Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>]
1883 *) New functions RSA_get_default_method(), RSA_set_method() and
1884 RSA_get_method(). These allows replacement of RSA_METHODs without having
1885 to mess around with the internals of an RSA structure.
1888 *) Fix memory leaks in DSA_do_sign and DSA_is_prime.
1889 Also really enable memory leak checks in openssl.c and in some
1891 [Chad C. Mulligan, Bodo Moeller]
1893 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
1894 up the length of negative integers. This has now been simplified to just
1895 store the length when it is first determined and use it later, rather
1896 than trying to keep track of where data is copied and updating it to
1898 [Steve Henson, reported by Brien Wheeler
1899 <bwheeler@authentica-security.com>]
1901 *) Add a new function PKCS7_signatureVerify. This allows the verification
1902 of a PKCS#7 signature but with the signing certificate passed to the
1903 function itself. This contrasts with PKCS7_dataVerify which assumes the
1904 certificate is present in the PKCS#7 structure. This isn't always the
1905 case: certificates can be omitted from a PKCS#7 structure and be
1906 distributed by "out of band" means (such as a certificate database).
1909 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the
1910 function prototypes in pem.h, also change util/mkdef.pl to add the
1911 necessary function names.
1914 *) mk1mf.pl (used by Windows builds) did not properly read the
1915 options set by Configure in the top level Makefile, and Configure
1916 was not even able to write more than one option correctly.
1917 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
1920 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config
1921 file to be loaded from a BIO or FILE pointer. The BIO version will
1922 for example allow memory BIOs to contain config info.
1925 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
1926 Whoever hopes to achieve shared-library compatibility across versions
1927 must use this, not the compile-time macro.
1928 (Exercise 0.9.4: Which is the minimum library version required by
1930 Note: All this applies only to multi-threaded programs, others don't
1934 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests
1935 through a BIO pair triggered the default case, i.e.
1936 SSLerr(...,SSL_R_UNKNOWN_STATE).
1939 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
1940 can use the SSL library even if none of the specific BIOs is
1944 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
1945 for the encoded length.
1946 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>]
1948 *) Add initial documentation of the X509V3 functions.
1951 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and
1952 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
1953 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
1954 secure PKCS#8 private key format with a high iteration count.
1957 *) Fix determination of Perl interpreter: A perl or perl5
1958 _directory_ in $PATH was also accepted as the interpreter.
1959 [Ralf S. Engelschall]
1961 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking
1962 wrong with it but it was very old and did things like calling
1963 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
1967 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed
1968 to use the new extension code.
1971 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
1972 with macros. This should make it easier to change their form, add extra
1973 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
1977 *) Add to configuration table a new entry that can specify an alternative
1978 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
1979 according to Mark Crispin <MRC@Panda.COM>.
1983 *) DES CBC did not update the IV. Weird.
1986 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
1987 Changing the behaviour of the former might break existing programs --
1988 where IV updating is needed, des_ncbc_encrypt can be used.
1991 *) When bntest is run from "make test" it drives bc to check its
1992 calculations, as well as internally checking them. If an internal check
1993 fails, it needs to cause bc to give a non-zero result or make test carries
1994 on without noticing the failure. Fixed.
1997 *) DES library cleanups.
2000 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
2001 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
2002 ciphers. NOTE: although the key derivation function has been verified
2003 against some published test vectors it has not been extensively tested
2004 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
2008 *) Instead of "mkdir -p", which is not fully portable, use new
2009 Perl script "util/mkdir-p.pl".
2012 *) Rewrite the way password based encryption (PBE) is handled. It used to
2013 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
2014 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
2015 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
2016 the 'parameter' field of the AlgorithmIdentifier is passed to the
2017 underlying key generation function so it must do its own ASN1 parsing.
2018 This has also changed the EVP_PBE_CipherInit() function which now has a
2019 'parameter' argument instead of literal salt and iteration count values
2020 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
2023 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms
2024 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
2025 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
2026 KEY" because this clashed with PKCS#8 unencrypted string. Since this
2027 value was just used as a "magic string" and not used directly its
2028 value doesn't matter.
2031 *) Introduce some semblance of const correctness to BN. Shame C doesn't
2035 *) "linux-sparc64" configuration (ultrapenguin).
2036 [Ray Miller <ray.miller@oucs.ox.ac.uk>]
2037 "linux-sparc" configuration.
2038 [Christian Forster <fo@hawo.stw.uni-erlangen.de>]
2040 *) config now generates no-xxx options for missing ciphers.
2043 *) Support the EBCDIC character set (work in progress).
2044 File ebcdic.c not yet included because it has a different license.
2045 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
2047 *) Support BS2000/OSD-POSIX.
2048 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
2050 *) Make callbacks for key generation use void * instead of char *.
2053 *) Make S/MIME samples compile (not yet tested).
2056 *) Additional typesafe stacks.
2059 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
2063 Changes between 0.9.3 and 0.9.3a [29 May 1999]
2065 *) New configuration variant "sco5-gcc".
2067 *) Updated some demos.
2068 [Sean O Riordain, Wade Scholine]
2070 *) Add missing BIO_free at exit of pkcs12 application.
2073 *) Fix memory leak in conf.c.
2076 *) Updates for Win32 to assembler version of MD5.
2079 *) Set #! path to perl in apps/der_chop to where we found it
2080 instead of using a fixed path.
2083 *) SHA library changes for irix64-mips4-cc.
2086 *) Improvements for VMS support.
2090 Changes between 0.9.2b and 0.9.3 [24 May 1999]
2092 *) Bignum library bug fix. IRIX 6 passes "make test" now!
2093 This also avoids the problems with SC4.2 and unpatched SC5.
2094 [Andy Polyakov <appro@fy.chalmers.se>]
2096 *) New functions sk_num, sk_value and sk_set to replace the previous macros.
2097 These are required because of the typesafe stack would otherwise break
2098 existing code. If old code used a structure member which used to be STACK
2099 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
2100 sk_num or sk_value it would produce an error because the num, data members
2101 are not present in STACK_OF. Now it just produces a warning. sk_set
2102 replaces the old method of assigning a value to sk_value
2103 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
2104 that does this will no longer work (and should use sk_set instead) but
2105 this could be regarded as a "questionable" behaviour anyway.
2108 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now
2109 correctly handle encrypted S/MIME data.
2112 *) Change type of various DES function arguments from des_cblock
2113 (which means, in function argument declarations, pointer to char)
2114 to des_cblock * (meaning pointer to array with 8 char elements),
2115 which allows the compiler to do more typechecking; it was like
2116 that back in SSLeay, but with lots of ugly casts.
2118 Introduce new type const_des_cblock.
2121 *) Reorganise the PKCS#7 library and get rid of some of the more obvious
2122 problems: find RecipientInfo structure that matches recipient certificate
2123 and initialise the ASN1 structures properly based on passed cipher.
2126 *) Belatedly make the BN tests actually check the results.
2129 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
2130 to and from BNs: it was completely broken. New compilation option
2131 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
2132 key elements as negative integers.
2135 *) Reorganize and speed up MD5.
2136 [Andy Polyakov <appro@fy.chalmers.se>]
2139 [Richard Levitte <richard@levitte.org>]
2141 *) New option -out to asn1parse to allow the parsed structure to be
2142 output to a file. This is most useful when combined with the -strparse
2143 option to examine the output of things like OCTET STRINGS.
2146 *) Make SSL library a little more fool-proof by not requiring any longer
2147 that SSL_set_{accept,connect}_state be called before
2148 SSL_{accept,connect} may be used (SSL_set_..._state is omitted
2149 in many applications because usually everything *appeared* to work as
2150 intended anyway -- now it really works as intended).
2153 *) Move openssl.cnf out of lib/.
2156 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall
2157 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
2158 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+
2159 [Ralf S. Engelschall]
2161 *) Various fixes to the EVP and PKCS#7 code. It may now be able to
2162 handle PKCS#7 enveloped data properly.
2163 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve]
2165 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
2166 copying pointers. The cert_st handling is changed by this in
2167 various ways (and thus what used to be known as ctx->default_cert
2168 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert
2169 any longer when s->cert does not give us what we need).
2170 ssl_cert_instantiate becomes obsolete by this change.
2171 As soon as we've got the new code right (possibly it already is?),
2172 we have solved a couple of bugs of the earlier code where s->cert
2173 was used as if it could not have been shared with other SSL structures.
2175 Note that using the SSL API in certain dirty ways now will result
2176 in different behaviour than observed with earlier library versions:
2177 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx)
2178 does not influence s as it used to.
2180 In order to clean up things more thoroughly, inside SSL_SESSION
2181 we don't use CERT any longer, but a new structure SESS_CERT
2182 that holds per-session data (if available); currently, this is
2183 the peer's certificate chain and, for clients, the server's certificate
2184 and temporary key. CERT holds only those values that can have
2185 meaningful defaults in an SSL_CTX.
2188 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
2189 from the internal representation. Various PKCS#7 fixes: remove some
2190 evil casts and set the enc_dig_alg field properly based on the signing
2194 *) Allow PKCS#12 password to be set from the command line or the
2195 environment. Let 'ca' get its config file name from the environment
2196 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
2200 *) Allow certificate policies extension to use an IA5STRING for the
2201 organization field. This is contrary to the PKIX definition but
2202 VeriSign uses it and IE5 only recognises this form. Document 'x509'
2206 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
2207 without disallowing inline assembler and the like for non-pedantic builds.
2210 *) Support Borland C++ builder.
2211 [Janez Jere <jj@void.si>, modified by Ulf Möller]
2216 *) SHA-1 cleanups and performance enhancements.
2217 [Andy Polyakov <appro@fy.chalmers.se>]
2219 *) Sparc v8plus assembler for the bignum library.
2220 [Andy Polyakov <appro@fy.chalmers.se>]
2222 *) Accept any -xxx and +xxx compiler options in Configure.
2225 *) Update HPUX configuration.
2228 *) Add missing sk_<type>_unshift() function to safestack.h
2229 [Ralf S. Engelschall]
2231 *) New function SSL_CTX_use_certificate_chain_file that sets the
2232 "extra_cert"s in addition to the certificate. (This makes sense
2233 only for "PEM" format files, as chains as a whole are not
2237 *) Support verify_depth from the SSL API.
2238 x509_vfy.c had what can be considered an off-by-one-error:
2239 Its depth (which was not part of the external interface)
2240 was actually counting the number of certificates in a chain;
2241 now it really counts the depth.
2244 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
2245 instead of X509err, which often resulted in confusing error
2246 messages since the error codes are not globally unique
2247 (e.g. an alleged error in ssl3_accept when a certificate
2248 didn't match the private key).
2250 *) New function SSL_CTX_set_session_id_context that allows to set a default
2251 value (so that you don't need SSL_set_session_id_context for each
2252 connection using the SSL_CTX).
2255 *) OAEP decoding bug fix.
2258 *) Support INSTALL_PREFIX for package builders, as proposed by
2262 *) New Configure options "threads" and "no-threads". For systems
2263 where the proper compiler options are known (currently Solaris
2264 and Linux), "threads" is the default.
2267 *) New script util/mklink.pl as a faster substitute for util/mklink.sh.
2270 *) Install various scripts to $(OPENSSLDIR)/misc, not to
2271 $(INSTALLTOP)/bin -- they shouldn't clutter directories
2272 such as /usr/local/bin.
2275 *) "make linux-shared" to build shared libraries.
2276 [Niels Poppe <niels@netbox.org>]
2278 *) New Configure option no-<cipher> (rsa, idea, rc5, ...).
2281 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
2282 extension adding in x509 utility.
2285 *) Remove NOPROTO sections and error code comments.
2288 *) Partial rewrite of the DEF file generator to now parse the ANSI
2292 *) New Configure options --prefix=DIR and --openssldir=DIR.
2295 *) Complete rewrite of the error code script(s). It is all now handled
2296 by one script at the top level which handles error code gathering,
2297 header rewriting and C source file generation. It should be much better
2298 than the old method: it now uses a modified version of Ulf's parser to
2299 read the ANSI prototypes in all header files (thus the old K&R definitions
2300 aren't needed for error creation any more) and do a better job of
2301 translating function codes into names. The old 'ASN1 error code imbedded
2302 in a comment' is no longer necessary and it doesn't use .err files which
2303 have now been deleted. Also the error code call doesn't have to appear all
2304 on one line (which resulted in some large lines...).
2307 *) Change #include filenames from <foo.h> to <openssl/foo.h>.
2310 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return
2311 0 (which usually indicates a closed connection), but continue reading.
2314 *) Fix some race conditions.
2317 *) Add support for CRL distribution points extension. Add Certificate
2318 Policies and CRL distribution points documentation.
2321 *) Move the autogenerated header file parts to crypto/opensslconf.h.
2324 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
2325 8 of keying material. Merlin has also confirmed interop with this fix
2326 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
2327 [Merlin Hughes <merlin@baltimore.ie>]
2329 *) Fix lots of warnings.
2330 [Richard Levitte <levitte@stacken.kth.se>]
2332 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
2333 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
2334 [Richard Levitte <levitte@stacken.kth.se>]
2336 *) Fix problems with sizeof(long) == 8.
2337 [Andy Polyakov <appro@fy.chalmers.se>]
2339 *) Change functions to ANSI C.
2342 *) Fix typos in error codes.
2343 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller]
2345 *) Remove defunct assembler files from Configure.
2348 *) SPARC v8 assembler BIGNUM implementation.
2349 [Andy Polyakov <appro@fy.chalmers.se>]
2351 *) Support for Certificate Policies extension: both print and set.
2352 Various additions to support the r2i method this uses.
2355 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could
2356 return a const string when you are expecting an allocated buffer.
2359 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
2360 types DirectoryString and DisplayText.
2363 *) Add code to allow r2i extensions to access the configuration database,
2364 add an LHASH database driver and add several ctx helper functions.
2367 *) Fix an evil bug in bn_expand2() which caused various BN functions to
2368 fail when they extended the size of a BIGNUM.
2371 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to
2372 support typesafe stack.
2375 *) Fix typo in SSL_[gs]et_options().
2376 [Nils Frostberg <nils@medcom.se>]
2378 *) Delete various functions and files that belonged to the (now obsolete)
2379 old X509V3 handling code.
2382 *) New Configure option "rsaref".
2385 *) Don't auto-generate pem.h.
2388 *) Introduce type-safe ASN.1 SETs.
2391 *) Convert various additional casted stacks to type-safe STACK_OF() variants.
2392 [Ben Laurie, Ralf S. Engelschall, Steve Henson]
2394 *) Introduce type-safe STACKs. This will almost certainly break lots of code
2395 that links with OpenSSL (well at least cause lots of warnings), but fear
2396 not: the conversion is trivial, and it eliminates loads of evil casts. A
2397 few STACKed things have been converted already. Feel free to convert more.
2398 In the fullness of time, I'll do away with the STACK type altogether.
2401 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate
2402 specified in <certfile> by updating the entry in the index.txt file.
2403 This way one no longer has to edit the index.txt file manually for
2404 revoking a certificate. The -revoke option does the gory details now.
2405 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall]
2407 *) Fix `openssl crl -noout -text' combination where `-noout' killed the
2408 `-text' option at all and this way the `-noout -text' combination was
2409 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'.
2410 [Ralf S. Engelschall]
2412 *) Make sure a corresponding plain text error message exists for the
2413 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
2414 verify callback function determined that a certificate was revoked.
2415 [Ralf S. Engelschall]
2417 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for
2418 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
2419 all available cipers including rc5, which was forgotten until now.
2420 In order to let the testing shell script know which algorithms
2421 are available, a new (up to now undocumented) command
2422 "openssl list-cipher-commands" is used.
2425 *) Bugfix: s_client occasionally would sleep in select() when
2426 it should have checked SSL_pending() first.
2429 *) New functions DSA_do_sign and DSA_do_verify to provide access to
2430 the raw DSA values prior to ASN.1 encoding.
2433 *) Tweaks to Configure
2434 [Niels Poppe <niels@netbox.org>]
2436 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
2440 *) New variables $(RANLIB) and $(PERL) in the Makefiles.
2443 *) New config option to avoid instructions that are illegal on the 80386.
2444 The default code is faster, but requires at least a 486.
2447 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
2448 SSL2_SERVER_VERSION (not used at all) macros, which are now the
2449 same as SSL2_VERSION anyway.
2452 *) New "-showcerts" option for s_client.
2455 *) Still more PKCS#12 integration. Add pkcs12 application to openssl
2456 application. Various cleanups and fixes.
2459 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
2460 modify error routines to work internally. Add error codes and PBE init
2461 to library startup routines.
2464 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and
2465 packing functions to asn1 and evp. Changed function names and error
2466 codes along the way.
2469 *) PKCS12 integration: and so it begins... First of several patches to
2470 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
2471 objects to objects.h
2474 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1
2475 and display support for Thawte strong extranet extension.
2478 *) Add LinuxPPC support.
2479 [Jeff Dubrule <igor@pobox.org>]
2481 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
2482 bn_div_words in alpha.s.
2483 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie]
2485 *) Make sure the RSA OAEP test is skipped under -DRSAref because
2486 OAEP isn't supported when OpenSSL is built with RSAref.
2487 [Ulf Moeller <ulf@fitug.de>]
2489 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
2490 so they no longer are missing under -DNOPROTO.
2491 [Soren S. Jorvang <soren@t.dk>]
2494 Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
2496 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
2497 doesn't work when the session is reused. Coming soon!
2500 *) Fix a security hole, that allows sessions to be reused in the wrong
2501 context thus bypassing client cert protection! All software that uses
2502 client certs and session caches in multiple contexts NEEDS PATCHING to
2503 allow session reuse! A fuller solution is in the works.
2504 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)]
2506 *) Some more source tree cleanups (removed obsolete files
2507 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
2508 permission on "config" script to be executable) and a fix for the INSTALL
2510 [Ulf Moeller <ulf@fitug.de>]
2512 *) Remove some legacy and erroneous uses of malloc, free instead of
2514 [Lennart Bang <lob@netstream.se>, with minor changes by Steve]
2516 *) Make rsa_oaep_test return non-zero on error.
2517 [Ulf Moeller <ulf@fitug.de>]
2519 *) Add support for native Solaris shared libraries. Configure
2520 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
2521 if someone would make that last step automatic.
2522 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>]
2524 *) ctx_size was not built with the right compiler during "make links". Fixed.
2527 *) Change the meaning of 'ALL' in the cipher list. It now means "everything
2528 except NULL ciphers". This means the default cipher list will no longer
2529 enable NULL ciphers. They need to be specifically enabled e.g. with
2530 the string "DEFAULT:eNULL".
2533 *) Fix to RSA private encryption routines: if p < q then it would
2534 occasionally produce an invalid result. This will only happen with
2535 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
2538 *) Be less restrictive and allow also `perl util/perlpath.pl
2539 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin',
2540 because this way one can also use an interpreter named `perl5' (which is
2541 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
2542 installed as `perl').
2543 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2545 *) Let util/clean-depend.pl work also with older Perl 5.00x versions.
2546 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2548 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
2549 advapi32.lib to Win32 build and change the pem test comparision
2550 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
2551 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
2552 and crypto/des/ede_cbcm_enc.c.
2555 *) DES quad checksum was broken on big-endian architectures. Fixed.
2558 *) Comment out two functions in bio.h that aren't implemented. Fix up the
2559 Win32 test batch file so it (might) work again. The Win32 test batch file
2560 is horrible: I feel ill....
2563 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
2564 in e_os.h. Audit of header files to check ANSI and non ANSI
2565 sections: 10 functions were absent from non ANSI section and not exported
2566 from Windows DLLs. Fixed up libeay.num for new functions.
2569 *) Make `openssl version' output lines consistent.
2570 [Ralf S. Engelschall]
2572 *) Fix Win32 symbol export lists for BIO functions: Added
2573 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
2574 to ms/libeay{16,32}.def.
2575 [Ralf S. Engelschall]
2577 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
2578 fine under Unix and passes some trivial tests I've now added. But the
2579 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
2580 added to make sure no one expects that this stuff really works in the
2581 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
2582 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
2584 [Ralf S. Engelschall]
2586 *) Fix the generation of two part addresses in perl.
2587 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie]
2589 *) Add config entry for Linux on MIPS.
2590 [John Tobey <jtobey@channel1.com>]
2592 *) Make links whenever Configure is run, unless we are on Windoze.
2595 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf.
2596 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
2600 *) Add a useful kludge to allow package maintainers to specify compiler and
2601 other platforms details on the command line without having to patch the
2602 Configure script everytime: One now can use ``perl Configure
2603 <id>:<details>'', i.e. platform ids are allowed to have details appended
2604 to them (seperated by colons). This is treated as there would be a static
2605 pre-configured entry in Configure's %table under key <id> with value
2606 <details> and ``perl Configure <id>'' is called. So, when you want to
2607 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
2608 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"''
2609 now, which overrides the FreeBSD-elf entry on-the-fly.
2610 [Ralf S. Engelschall]
2612 *) Disable new TLS1 ciphersuites by default: they aren't official yet.
2615 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
2616 on the `perl Configure ...' command line. This way one can compile
2617 OpenSSL libraries with Position Independent Code (PIC) which is needed
2618 for linking it into DSOs.
2619 [Ralf S. Engelschall]
2621 *) Remarkably, export ciphers were totally broken and no-one had noticed!
2625 *) Cleaned up the LICENSE document: The official contact for any license
2626 questions now is the OpenSSL core team under openssl-core@openssl.org.
2627 And add a paragraph about the dual-license situation to make sure people
2628 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
2629 to the OpenSSL toolkit.
2630 [Ralf S. Engelschall]
2632 *) General source tree makefile cleanups: Made `making xxx in yyy...'
2633 display consistent in the source tree and replaced `/bin/rm' by `rm'.
2634 Additonally cleaned up the `make links' target: Remove unnecessary
2635 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
2636 to speed processing and no longer clutter the display with confusing
2637 stuff. Instead only the actually done links are displayed.
2638 [Ralf S. Engelschall]
2640 *) Permit null encryption ciphersuites, used for authentication only. It used
2641 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
2642 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
2646 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
2647 signed attributes when verifying signatures (this would break them),
2648 the detached data encoding was wrong and public keys obtained using
2649 X509_get_pubkey() weren't freed.
2652 *) Add text documentation for the BUFFER functions. Also added a work around
2653 to a Win95 console bug. This was triggered by the password read stuff: the
2654 last character typed gets carried over to the next fread(). If you were
2655 generating a new cert request using 'req' for example then the last
2656 character of the passphrase would be CR which would then enter the first
2660 *) Added the new `Includes OpenSSL Cryptography Software' button as
2661 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
2662 button and can be used by applications based on OpenSSL to show the
2663 relationship to the OpenSSL project.
2664 [Ralf S. Engelschall]
2666 *) Remove confusing variables in function signatures in files
2667 ssl/ssl_lib.c and ssl/ssl.h.
2668 [Lennart Bong <lob@kulthea.stacken.kth.se>]
2670 *) Don't install bss_file.c under PREFIX/include/
2671 [Lennart Bong <lob@kulthea.stacken.kth.se>]
2673 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle
2674 functions that return function pointers and has support for NT specific
2675 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
2676 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
2677 unsigned to signed types: this was killing the Win32 compile.
2680 *) Add new certificate file to stack functions,
2681 SSL_add_dir_cert_subjects_to_stack() and
2682 SSL_add_file_cert_subjects_to_stack(). These largely supplant
2683 SSL_load_client_CA_file(), and can be used to add multiple certs easily
2684 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
2685 This means that Apache-SSL and similar packages don't have to mess around
2686 to add as many CAs as they want to the preferred list.
2689 *) Experiment with doxygen documentation. Currently only partially applied to
2691 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with
2692 openssl.doxy as the configuration file.
2695 *) Get rid of remaining C++-style comments which strict C compilers hate.
2696 [Ralf S. Engelschall, pointed out by Carlos Amengual]
2698 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
2699 compiled in by default: it has problems with large keys.
2702 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
2703 DH private keys and/or callback functions which directly correspond to
2704 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
2705 is needed for applications which have to configure certificates on a
2706 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
2708 For the RSA certificate situation is makes no difference, but
2709 for the DSA certificate situation this fixes the "no shared cipher"
2710 problem where the OpenSSL cipher selection procedure failed because the
2711 temporary keys were not overtaken from the context and the API provided
2712 no way to reconfigure them.
2713 The new functions now let applications reconfigure the stuff and they
2714 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
2715 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
2716 non-public-API function ssl_cert_instantiate() is used as a helper
2717 function and also to reduce code redundancy inside ssl_rsa.c.
2718 [Ralf S. Engelschall]
2720 *) Move s_server -dcert and -dkey options out of the undocumented feature
2721 area because they are useful for the DSA situation and should be
2722 recognized by the users.
2723 [Ralf S. Engelschall]
2725 *) Fix the cipher decision scheme for export ciphers: the export bits are
2726 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
2727 SSL_EXP_MASK. So, the original variable has to be used instead of the
2728 already masked variable.
2729 [Richard Levitte <levitte@stacken.kth.se>]
2731 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c
2732 [Richard Levitte <levitte@stacken.kth.se>]
2734 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
2735 from `int' to `unsigned int' because it's a length and initialized by
2736 EVP_DigestFinal() which expects an `unsigned int *'.
2737 [Richard Levitte <levitte@stacken.kth.se>]
2739 *) Don't hard-code path to Perl interpreter on shebang line of Configure
2740 script. Instead use the usual Shell->Perl transition trick.
2741 [Ralf S. Engelschall]
2743 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates
2744 (in addition to RSA certificates) to match the behaviour of `openssl dsa
2745 -noout -modulus' as it's already the case for `openssl rsa -noout
2746 -modulus'. For RSA the -modulus is the real "modulus" while for DSA
2747 currently the public key is printed (a decision which was already done by
2748 `openssl dsa -modulus' in the past) which serves a similar purpose.
2749 Additionally the NO_RSA no longer completely removes the whole -modulus
2750 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
2752 [Ralf S. Engelschall]
2754 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
2755 BIO. See the source (crypto/evp/bio_ok.c) for more info.
2756 [Arne Ansper <arne@ats.cyber.ee>]
2758 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs
2759 to be added. Now both 'req' and 'ca' can use new objects defined in the
2763 *) Add cool BIO that does syslog (or event log on NT).
2764 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie]
2766 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
2767 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
2768 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
2769 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
2772 *) Add preliminary config info for new extension code.
2775 *) Make RSA_NO_PADDING really use no padding.
2776 [Ulf Moeller <ulf@fitug.de>]
2778 *) Generate errors when private/public key check is done.
2781 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
2782 for some CRL extensions and new objects added.
2785 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private
2786 key usage extension and fuller support for authority key id.
2789 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
2790 padding method for RSA, which is recommended for new applications in PKCS
2791 #1 v2.0 (RFC 2437, October 1998).
2792 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
2793 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
2794 against Bleichbacher's attack on RSA.
2795 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
2798 *) Updates to the new SSL compression code
2799 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2801 *) Fix so that the version number in the master secret, when passed
2802 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
2803 (because the server will not accept higher), that the version number
2804 is 0x03,0x01, not 0x03,0x00
2805 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2807 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
2808 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
2809 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
2812 *) Support for RAW extensions where an arbitrary extension can be
2813 created by including its DER encoding. See apps/openssl.cnf for
2817 *) Make sure latest Perl versions don't interpret some generated C array
2818 code as Perl array code in the crypto/err/err_genc.pl script.
2819 [Lars Weber <3weber@informatik.uni-hamburg.de>]
2821 *) Modify ms/do_ms.bat to not generate assembly language makefiles since
2822 not many people have the assembler. Various Win32 compilation fixes and
2823 update to the INSTALL.W32 file with (hopefully) more accurate Win32
2827 *) Modify configure script 'Configure' to automatically create crypto/date.h
2828 file under Win32 and also build pem.h from pem.org. New script
2829 util/mkfiles.pl to create the MINFO file on environments that can't do a
2830 'make files': perl util/mkfiles.pl >MINFO should work.
2833 *) Major rework of DES function declarations, in the pursuit of correctness
2834 and purity. As a result, many evil casts evaporated, and some weirdness,
2835 too. You may find this causes warnings in your code. Zapping your evil
2836 casts will probably fix them. Mostly.
2839 *) Fix for a typo in asn1.h. Bug fix to object creation script
2840 obj_dat.pl. It considered a zero in an object definition to mean
2841 "end of object": none of the objects in objects.h have any zeros
2842 so it wasn't spotted.
2843 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>]
2845 *) Add support for Triple DES Cipher Block Chaining with Output Feedback
2846 Masking (CBCM). In the absence of test vectors, the best I have been able
2847 to do is check that the decrypt undoes the encrypt, so far. Send me test
2848 vectors if you have them.
2851 *) Correct calculation of key length for export ciphers (too much space was
2852 allocated for null ciphers). This has not been tested!
2855 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage
2856 message is now correct (it understands "crypto" and "ssl" on its
2857 command line). There is also now an "update" option. This will update
2858 the util/ssleay.num and util/libeay.num files with any new functions.
2860 perl util/mkdef.pl crypto ssl update
2861 it will update them.
2864 *) Overhauled the Perl interface (perl/*):
2865 - ported BN stuff to OpenSSL's different BN library
2866 - made the perl/ source tree CVS-aware
2867 - renamed the package from SSLeay to OpenSSL (the files still contain
2868 their history because I've copied them in the repository)
2869 - removed obsolete files (the test scripts will be replaced
2870 by better Test::Harness variants in the future)
2871 [Ralf S. Engelschall]
2873 *) First cut for a very conservative source tree cleanup:
2874 1. merge various obsolete readme texts into doc/ssleay.txt
2875 where we collect the old documents and readme texts.
2876 2. remove the first part of files where I'm already sure that we no
2877 longer need them because of three reasons: either they are just temporary
2878 files which were left by Eric or they are preserved original files where
2879 I've verified that the diff is also available in the CVS via "cvs diff
2880 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
2881 the crypto/md/ stuff).
2882 [Ralf S. Engelschall]
2884 *) More extension code. Incomplete support for subject and issuer alt
2885 name, issuer and authority key id. Change the i2v function parameters
2886 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
2887 what that's for :-) Fix to ASN1 macro which messed up
2888 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
2891 *) Preliminary support for ENUMERATED type. This is largely copied from the
2895 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
2896 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2898 *) Make sure `make rehash' target really finds the `openssl' program.
2899 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2901 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
2902 like to hear about it if this slows down other processors.
2905 *) Add CygWin32 platform information to Configure script.
2906 [Alan Batie <batie@aahz.jf.intel.com>]
2908 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm'
2909 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>]
2911 *) New program nseq to manipulate netscape certificate sequences
2914 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
2918 *) Fixes to BN code. Previously the default was to define BN_RECURSION
2919 but the BN code had some problems that would cause failures when
2920 doing certificate verification and some other functions.
2921 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2923 *) Add ASN1 and PEM code to support netscape certificate sequences.
2926 *) Add ASN1 and PEM code to support netscape certificate sequences.
2929 *) Add several PKIX and private extended key usage OIDs.
2932 *) Modify the 'ca' program to handle the new extension code. Modify
2933 openssl.cnf for new extension format, add comments.
2936 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
2937 and add a sample to openssl.cnf so req -x509 now adds appropriate
2941 *) Continued X509 V3 changes. Add to other makefiles, integrate with the
2942 error code, add initial support to X509_print() and x509 application.
2945 *) Takes a deep breath and start addding X509 V3 extension support code. Add
2946 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
2947 stuff is currently isolated and isn't even compiled yet.
2950 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL
2951 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
2952 Removed the versions check from X509 routines when loading extensions:
2953 this allows certain broken certificates that don't set the version
2954 properly to be processed.
2957 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
2958 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
2959 can still be regenerated with "make depend".
2962 *) Spelling mistake in C version of CAST-128.
2963 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>]
2965 *) Changes to the error generation code. The perl script err-code.pl
2966 now reads in the old error codes and retains the old numbers, only
2967 adding new ones if necessary. It also only changes the .err files if new
2968 codes are added. The makefiles have been modified to only insert errors
2969 when needed (to avoid needlessly modifying header files). This is done
2970 by only inserting errors if the .err file is newer than the auto generated
2971 C file. To rebuild all the error codes from scratch (the old behaviour)
2972 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
2973 or delete all the .err files.
2976 *) CAST-128 was incorrectly implemented for short keys. The C version has
2977 been fixed, but is untested. The assembler versions are also fixed, but
2978 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
2979 to regenerate it if needed.
2980 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
2981 Hagino <itojun@kame.net>]
2983 *) File was opened incorrectly in randfile.c.
2984 [Ulf Möller <ulf@fitug.de>]
2986 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print
2987 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
2988 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
2989 al: it's just almost always a UTCTime. Note this patch adds new error
2990 codes so do a "make errors" if there are problems.
2993 *) Correct Linux 1 recognition in config.
2994 [Ulf Möller <ulf@fitug.de>]
2996 *) Remove pointless MD5 hash when using DSA keys in ca.
2997 [Anonymous <nobody@replay.com>]
2999 *) Generate an error if given an empty string as a cert directory. Also
3000 generate an error if handed NULL (previously returned 0 to indicate an
3001 error, but didn't set one).
3002 [Ben Laurie, reported by Anonymous <nobody@replay.com>]
3004 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
3007 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
3008 parameters. This was causing a warning which killed off the Win32 compile.
3011 *) Remove C++ style comments from crypto/bn/bn_local.h.
3012 [Neil Costigan <neil.costigan@celocom.com>]
3014 *) The function OBJ_txt2nid was broken. It was supposed to return a nid
3015 based on a text string, looking up short and long names and finally
3016 "dot" format. The "dot" format stuff didn't work. Added new function
3017 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
3018 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
3019 OID is not part of the table.
3022 *) Add prototypes to X509 lookup/verify methods, fixing a bug in
3023 X509_LOOKUP_by_alias().
3026 *) Sort openssl functions by name.
3029 *) Get the gendsa program working (hopefully) and add it to app list. Remove
3030 encryption from sample DSA keys (in case anyone is interested the password
3034 *) Make _all_ *_free functions accept a NULL pointer.
3035 [Frans Heymans <fheymans@isaserver.be>]
3037 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use
3039 [Anonymous <nobody@replay.com>]
3041 *) s_server should send the CAfile as acceptable CAs, not its own cert.
3042 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
3044 *) Don't blow it for numeric -newkey arguments to apps/req.
3045 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
3047 *) Temp key "for export" tests were wrong in s3_srvr.c.
3048 [Anonymous <nobody@replay.com>]
3050 *) Add prototype for temp key callback functions
3051 SSL_CTX_set_tmp_{rsa,dh}_callback().
3054 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
3055 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
3058 *) X509_name_add_entry() freed the wrong thing after an error.
3059 [Arne Ansper <arne@ats.cyber.ee>]
3061 *) rsa_eay.c would attempt to free a NULL context.
3062 [Arne Ansper <arne@ats.cyber.ee>]
3064 *) BIO_s_socket() had a broken should_retry() on Windoze.
3065 [Arne Ansper <arne@ats.cyber.ee>]
3067 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
3068 [Arne Ansper <arne@ats.cyber.ee>]
3070 *) Make sure the already existing X509_STORE->depth variable is initialized
3071 in X509_STORE_new(), but document the fact that this variable is still
3072 unused in the certificate verification process.
3073 [Ralf S. Engelschall]
3075 *) Fix the various library and apps files to free up pkeys obtained from
3076 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
3079 *) Fix reference counting in X509_PUBKEY_get(). This makes
3080 demos/maurice/example2.c work, amongst others, probably.
3081 [Steve Henson and Ben Laurie]
3083 *) First cut of a cleanup for apps/. First the `ssleay' program is now named
3084 `openssl' and second, the shortcut symlinks for the `openssl <command>'
3085 are no longer created. This way we have a single and consistent command
3086 line interface `openssl <command>', similar to `cvs <command>'.
3087 [Ralf S. Engelschall, Paul Sutton and Ben Laurie]
3089 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
3090 BIT STRING wrapper always have zero unused bits.
3093 *) Add CA.pl, perl version of CA.sh, add extended key usage OID.
3096 *) Make the top-level INSTALL documentation easier to understand.
3099 *) Makefiles updated to exit if an error occurs in a sub-directory
3100 make (including if user presses ^C) [Paul Sutton]
3102 *) Make Montgomery context stuff explicit in RSA data structure.
3105 *) Fix build order of pem and err to allow for generated pem.h.
3108 *) Fix renumbering bug in X509_NAME_delete_entry().
3111 *) Enhanced the err-ins.pl script so it makes the error library number
3112 global and can add a library name. This is needed for external ASN1 and
3113 other error libraries.
3116 *) Fixed sk_insert which never worked properly.
3119 *) Fix ASN1 macros so they can handle indefinite length construted
3120 EXPLICIT tags. Some non standard certificates use these: they can now
3124 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
3125 into a single doc/ssleay.txt bundle. This way the information is still
3126 preserved but no longer messes up this directory. Now it's new room for
3127 the new set of documenation files.
3128 [Ralf S. Engelschall]
3130 *) SETs were incorrectly DER encoded. This was a major pain, because they
3131 shared code with SEQUENCEs, which aren't coded the same. This means that
3132 almost everything to do with SETs or SEQUENCEs has either changed name or
3133 number of arguments.
3134 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>]
3136 *) Fix test data to work with the above.
3139 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
3140 was already fixed by Eric for 0.9.1 it seems.
3141 [Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>]
3143 *) Autodetect FreeBSD3.
3146 *) Fix various bugs in Configure. This affects the following platforms:
3150 unixware-2.0-pentium
3154 *) Eliminate generated files from CVS. Reorder tests to regenerate files
3155 before they are needed.
3158 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy).
3162 Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
3164 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
3165 changed SSLeay to OpenSSL in version strings.
3166 [Ralf S. Engelschall]
3168 *) Some fixups to the top-level documents.
3171 *) Fixed the nasty bug where rsaref.h was not found under compile-time
3172 because the symlink to include/ was missing.
3173 [Ralf S. Engelschall]
3175 *) Incorporated the popular no-RSA/DSA-only patches
3176 which allow to compile a RSA-free SSLeay.
3177 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall]
3179 *) Fixed nasty rehash problem under `make -f Makefile.ssl links'
3180 when "ssleay" is still not found.
3181 [Ralf S. Engelschall]
3183 *) Added more platforms to Configure: Cray T3E, HPUX 11,
3184 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>]
3186 *) Updated the README file.
3187 [Ralf S. Engelschall]
3189 *) Added various .cvsignore files in the CVS repository subdirs
3190 to make a "cvs update" really silent.
3191 [Ralf S. Engelschall]
3193 *) Recompiled the error-definition header files and added
3194 missing symbols to the Win32 linker tables.
3195 [Ralf S. Engelschall]
3197 *) Cleaned up the top-level documents;
3198 o new files: CHANGES and LICENSE
3199 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
3200 o merged COPYRIGHT into LICENSE
3201 o removed obsolete TODO file
3202 o renamed MICROSOFT to INSTALL.W32
3203 [Ralf S. Engelschall]
3205 *) Removed dummy files from the 0.9.1b source tree:
3206 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
3207 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
3208 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
3209 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
3210 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
3211 [Ralf S. Engelschall]
3213 *) Added various platform portability fixes.
3216 *) The Genesis of the OpenSSL rpject:
3217 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
3218 Young and Tim J. Hudson created while they were working for C2Net until
3220 [The OpenSSL Project]
3223 Changes between 0.9.0b and 0.9.1b [not released]
3225 *) Updated a few CA certificates under certs/
3228 *) Changed some BIGNUM api stuff.
3231 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
3232 DGUX x86, Linux Alpha, etc.
3235 *) New COMP library [crypto/comp/] for SSL Record Layer Compression:
3236 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
3240 *) Add -strparse option to asn1pars program which parses nested
3242 [Dr Stephen Henson <shenson@bigfoot.com>]
3244 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
3247 *) DSA fix for "ca" program.
3250 *) Added "-genkey" option to "dsaparam" program.
3253 *) Added RIPE MD160 (rmd160) message digest.
3256 *) Added -a (all) option to "ssleay version" command.
3259 *) Added PLATFORM define which is the id given to Configure.
3262 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
3265 *) Extended the ASN.1 parser routines.
3268 *) Extended BIO routines to support REUSEADDR, seek, tell, etc.
3271 *) Added a BN_CTX to the BN library.
3274 *) Fixed the weak key values in DES library
3277 *) Changed API in EVP library for cipher aliases.
3280 *) Added support for RC2/64bit cipher.
3283 *) Converted the lhash library to the crypto/mem.c functions.
3286 *) Added more recognized ASN.1 object ids.
3289 *) Added more RSA padding checks for SSL/TLS.
3292 *) Added BIO proxy/filter functionality.
3295 *) Added extra_certs to SSL_CTX which can be used
3296 send extra CA certificates to the client in the CA cert chain sending
3297 process. It can be configured with SSL_CTX_add_extra_chain_cert().
3300 *) Now Fortezza is denied in the authentication phase because
3301 this is key exchange mechanism is not supported by SSLeay at all.
3304 *) Additional PKCS1 checks.
3307 *) Support the string "TLSv1" for all TLS v1 ciphers.
3310 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
3311 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
3314 *) Fixed a few memory leaks.
3317 *) Fixed various code and comment typos.
3320 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0
3321 bytes sent in the client random.
3322 [Edward Bishop <ebishop@spyglass.com>]