5 Changes between 0.9.5a and 0.9.6 [xx XXX 2000]
7 *) Modification to PKCS#7 encoding routines to output definite
8 length encoding. Since currently the whole structures are in
9 memory there's not real point in using indefinite length
10 constructed encoding. However if OpenSSL is compiled with
11 the flag PKCS7_INDEFINITE_ENCODING the old form is used.
14 *) Added BIO_vprintf() and BIO_vsnprintf().
17 *) Added more prefixes to parse for in the the strings written
18 through a logging bio, to cover all the levels that are available
19 through syslog. The prefixes are now:
21 PANIC, EMERG, EMR => LOG_EMERG
22 ALERT, ALR => LOG_ALERT
25 WARNING, WARN, WAR => LOG_WARNING
26 NOTICE, NOTE, NOT => LOG_NOTICE
28 DEBUG, DBG => LOG_DEBUG
30 and as before, if none of those prefixes are present at the
31 beginning of the string, LOG_ERR is chosen.
33 On Win32, the LOG_* levels are mapped according to this:
35 LOG_EMERG, LOG_ALERT, LOG_CRIT, LOG_ERR => EVENTLOG_ERROR_TYPE
36 LOG_WARNING => EVENTLOG_WARNING_TYPE
37 LOG_NOTICE, LOG_INFO, LOG_DEBUG => EVENTLOG_INFORMATION_TYPE
41 *) Made it possible to reconfigure with just the configuration
42 argument "reconf" or "reconfigure". The command line arguments
43 are stored in Makefile.ssl in the variable CONFIGURE_ARGS,
44 and are retrieved from there when reconfiguring.
48 [Assar Westerlund <assar@sics.se>, Richard Levitte]
50 *) Add the arguments -CAfile and -CApath to the pkcs12 utility.
53 *) The obj_dat.pl script was messing up the sorting of object
54 names. The reason was that it compared the quoted version
55 of strings as a result "OCSP" > "OCSP Signing" because
56 " > SPACE. Changed script to store unquoted versions of
57 names and add quotes on output. It was also omitting some
58 names from the lookup table if they were given a default
59 value (that is if SN is missing it is given the same
60 value as LN and vice versa), these are now added on the
61 grounds that if an object has a name we should be able to
62 look it up. Finally added warning output when duplicate
63 short or long names are found.
66 *) Changes needed for Tandem NSK.
67 [Scott Uroff scott@xypro.com]
69 *) Fix SSL 2.0 rollback checking: Due to an off-by-one error in
70 RSA_padding_check_SSLv23(), special padding was never detected
71 and thus the SSL 3.0/TLS 1.0 countermeasure against protocol
72 version rollback attacks was not effective.
74 In s23_clnt.c, don't use special rollback-attack detection padding
75 (RSA_SSLV23_PADDING) if SSL 2.0 is the only protocol enabled in the
76 client; similarly, in s23_srvr.c, don't do the rollback check if
77 SSL 2.0 is the only protocol enabled in the server.
80 *) Make it possible to get hexdumps of unprintable data with 'openssl
81 asn1parse'. By implication, the functions ASN1_parse_dump() and
82 BIO_dump_indent() are added.
85 *) New functions ASN1_STRING_print_ex() and X509_NAME_print_ex()
86 these print out strings and name structures based on various
87 flags including RFC2253 support and proper handling of
88 multibyte characters. Added options to the 'x509' utility
89 to allow the various flags to be set.
92 *) Various fixes to use ASN1_TIME instead of ASN1_UTCTIME.
93 Also change the functions X509_cmp_current_time() and
94 X509_gmtime_adj() work with an ASN1_TIME structure,
95 this will enable certificates using GeneralizedTime in validity
99 *) Make the NEG_PUBKEY_BUG code (which tolerates invalid
100 negative public key encodings) on by default,
101 NO_NEG_PUBKEY_BUG can be set to disable it.
104 *) New function c2i_ASN1_OBJECT() which acts on ASN1_OBJECT
105 content octets. An i2c_ASN1_OBJECT is unnecessary because
106 the encoding can be trivially obtained from the structure.
109 *) crypto/err.c locking bugfix: Use write locks (CRYPTO_w_[un]lock),
110 not read locks (CRYPTO_r_[un]lock).
113 *) A first attempt at creating official support for shared
114 libraries through configuration. I've kept it so the
115 default is static libraries only, and the OpenSSL programs
116 are always statically linked for now, but there are
117 preparations for dynamic linking in place.
118 This has been tested on Linux and True64.
121 *) Randomness polling function for Win9x, as described in:
122 Peter Gutmann, Software Generation of Practically Strong
126 *) Fix so PRNG is seeded in req if using an already existing
130 *) New options to smime application. -inform and -outform
131 allow alternative formats for the S/MIME message including
132 PEM and DER. The -content option allows the content to be
133 specified separately. This should allow things like Netscape
134 form signing output easier to verify.
137 *) Fix the ASN1 encoding of tags using the 'long form'.
140 *) New ASN1 functions, i2c_* and c2i_* for INTEGER and BIT
141 STRING types. These convert content octets to and from the
142 underlying type. The actual tag and length octets are
143 already assumed to have been read in and checked. These
144 are needed because all other string types have virtually
145 identical handling apart from the tag. By having versions
146 of the ASN1 functions that just operate on content octets
147 IMPLICIT tagging can be handled properly. It also allows
148 the ASN1_ENUMERATED code to be cut down because ASN1_ENUMERATED
149 and ASN1_INTEGER are identical apart from the tag.
152 *) Change the handling of OID objects as follows:
154 - New object identifiers are inserted in objects.txt, following
155 the syntax given in objects.README.
156 - objects.pl is used to process obj_mac.num and create a new
158 - obj_dat.pl is used to create a new obj_dat.h, using the data in
161 This is currently kind of a hack, and the perl code in objects.pl
162 isn't very elegant, but it works as I intended. The simplest way
163 to check that it worked correctly is to look in obj_dat.h and
164 check the array nid_objs and make sure the objects haven't moved
165 around (this is important!). Additions are OK, as well as
166 consistent name changes.
169 *) Add BSD-style MD5-based passwords to 'openssl passwd' (option '-1').
172 *) Addition of the command line parameter '-rand file' to 'openssl req'.
173 The given file adds to whatever has already been seeded into the
174 random pool through the RANDFILE configuration file option or
175 environment variable, or the default random state file.
178 *) mkstack.pl now sorts each macro group into lexical order.
179 Previously the output order depended on the order the files
180 appeared in the directory, resulting in needless rewriting
184 *) Patches to make OpenSSL compile under Win32 again. Mostly
185 work arounds for the VC++ problem that it treats func() as
186 func(void). Also stripped out the parts of mkdef.pl that
187 added extra typesafe functions: these no longer exist.
190 *) Reorganisation of the stack code. The macros are now all
191 collected in safestack.h . Each macro is defined in terms of
192 a "stack macro" of the form SKM_<name>(type, a, b). The
193 DEBUG_SAFESTACK is now handled in terms of function casts,
194 this has the advantage of retaining type safety without the
195 use of additional functions. If DEBUG_SAFESTACK is not defined
196 then the non typesafe macros are used instead. Also modified the
197 mkstack.pl script to handle the new form. Needs testing to see
198 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK
199 the default if no major problems. Similar behaviour for ASN1_SET_OF
203 *) When some versions of IIS use the 'NET' form of private key the
204 key derivation algorithm is different. Normally MD5(password) is
205 used as a 128 bit RC4 key. In the modified case
206 MD5(MD5(password) + "SGCKEYSALT") is used insted. Added some
207 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same
208 as the old Netscape_RSA functions except they have an additional
209 'sgckey' parameter which uses the modified algorithm. Also added
210 an -sgckey command line option to the rsa utility. Thanks to
211 Adrian Peck <bertie@ncipher.com> for posting details of the modified
212 algorithm to openssl-dev.
215 *) The evp_local.h macros were using 'c.##kname' which resulted in
216 invalid expansion on some systems (SCO 5.0.5 for example).
217 Corrected to 'c.kname'.
218 [Phillip Porch <root@theporch.com>]
220 *) New X509_get1_email() and X509_REQ_get1_email() functions that return
221 a STACK of email addresses from a certificate or request, these look
222 in the subject name and the subject alternative name extensions and
223 omit any duplicate addresses.
226 *) Re-implement BN_mod_exp2_mont using independent (and larger) windows.
227 This makes DSA verification about 2 % faster.
230 *) Increase maximum window size in BN_mod_exp_... to 6 bits instead of 5
231 (meaning that now 2^5 values will be precomputed, which is only 4 KB
232 plus overhead for 1024 bit moduli).
233 This makes exponentiations about 0.5 % faster for 1024 bit
234 exponents (as measured by "openssl speed rsa2048").
237 *) Rename memory handling macros to avoid conflicts with other
239 Malloc => OPENSSL_malloc
240 Malloc_locked => OPENSSL_malloc_locked
241 Realloc => OPENSSL_realloc
245 *) New function BN_mod_exp_mont_word for small bases (roughly 15%
246 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange).
250 [John Jarvie <jjarvie@newsguy.com>]
252 *) The type-safe stack code has been rejigged. It is now only compiled
253 in when OpenSSL is configured with the DEBUG_SAFESTACK option and
254 by default all type-specific stack functions are "#define"d back to
255 standard stack functions. This results in more streamlined output
256 but retains the type-safety checking possibilities of the original
260 *) The STACK code has been cleaned up, and certain type declarations
261 that didn't make a lot of sense have been brought in line. This has
262 also involved a cleanup of sorts in safestack.h to more correctly
263 map type-safe stack functions onto their plain stack counterparts.
264 This work has also resulted in a variety of "const"ifications of
265 lots of the code, especially "_cmp" operations which should normally
266 be prototyped with "const" parameters anyway.
269 *) When generating bytes for the first time in md_rand.c, 'stir the pool'
270 by seeding with STATE_SIZE dummy bytes (with zero entropy count).
271 (The PRNG state consists of two parts, the large pool 'state' and 'md',
272 where all of 'md' is used each time the PRNG is used, but 'state'
273 is used only indexed by a cyclic counter. As entropy may not be
274 well distributed from the beginning, 'md' is important as a
275 chaining variable. However, the output function chains only half
276 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains
277 all of 'md', and seeding with STATE_SIZE dummy bytes will result
278 in all of 'state' being rewritten, with the new values depending
279 on virtually all of 'md'. This overcomes the 80 bit limitation.)
282 *) In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when
283 the handshake is continued after ssl_verify_cert_chain();
284 otherwise, if SSL_VERIFY_NONE is set, remaining error codes
285 can lead to 'unexplainable' connection aborts later.
286 [Bodo Moeller; problem tracked down by Lutz Jaenicke]
288 *) Major EVP API cipher revision.
289 Add hooks for extra EVP features. This allows various cipher
290 parameters to be set in the EVP interface. Support added for variable
291 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and
292 setting of RC2 and RC5 parameters.
294 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length
297 Remove lots of duplicated code from the EVP library. For example *every*
298 cipher init() function handles the 'iv' in the same way according to the
299 cipher mode. They also all do nothing if the 'key' parameter is NULL and
300 for CFB and OFB modes they zero ctx->num.
302 New functionality allows removal of S/MIME code RC2 hack.
304 Most of the routines have the same form and so can be declared in terms
307 By shifting this to the top level EVP_CipherInit() it can be removed from
308 all individual ciphers. If the cipher wants to handle IVs or keys
309 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT
312 Change lots of functions like EVP_EncryptUpdate() to now return a
313 value: although software versions of the algorithms cannot fail
314 any installed hardware versions can.
317 *) Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if
318 this option is set, tolerate broken clients that send the negotiated
319 protocol version number instead of the requested protocol version
323 *) Call dh_tmp_cb (set by ..._TMP_DH_CB) with correct 'is_export' flag;
324 i.e. non-zero for export ciphersuites, zero otherwise.
325 Previous versions had this flag inverted, inconsistent with
326 rsa_tmp_cb (..._TMP_RSA_CB).
327 [Bodo Moeller; problem reported by Amit Chopra]
329 *) Add missing DSA library text string. Work around for some IIS
330 key files with invalid SEQUENCE encoding.
333 *) Add a document (doc/standards.txt) that list all kinds of standards
334 and so on that are implemented in OpenSSL.
337 *) Enhance c_rehash script. Old version would mishandle certificates
338 with the same subject name hash and wouldn't handle CRLs at all.
339 Added -fingerprint option to crl utility, to support new c_rehash
343 *) Eliminate non-ANSI declarations in crypto.h and stack.h.
346 *) Fix for SSL server purpose checking. Server checking was
347 rejecting certificates which had extended key usage present
348 but no ssl client purpose.
349 [Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>]
351 *) Make PKCS#12 code work with no password. The PKCS#12 spec
352 is a little unclear about how a blank password is handled.
353 Since the password in encoded as a BMPString with terminating
354 double NULL a zero length password would end up as just the
355 double NULL. However no password at all is different and is
356 handled differently in the PKCS#12 key generation code. NS
357 treats a blank password as zero length. MSIE treats it as no
358 password on export: but it will try both on import. We now do
359 the same: PKCS12_parse() tries zero length and no password if
360 the password is set to "" or NULL (NULL is now a valid password:
361 it wasn't before) as does the pkcs12 application.
364 *) Bugfixes in apps/x509.c: Avoid a memory leak; and don't use
365 perror when PEM_read_bio_X509_REQ fails, the error message must
366 be obtained from the error queue.
369 *) Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing
370 it in ERR_remove_state if appropriate, and change ERR_get_state
371 accordingly to avoid race conditions (this is necessary because
372 thread_hash is no longer constant once set).
375 *) Bugfix for linux-elf makefile.one.
378 *) RSA_get_default_method() will now cause a default
379 RSA_METHOD to be chosen if one doesn't exist already.
380 Previously this was only set during a call to RSA_new()
381 or RSA_new_method(NULL) meaning it was possible for
382 RSA_get_default_method() to return NULL.
385 *) Added native name translation to the existing DSO code
386 that will convert (if the flag to do so is set) filenames
387 that are sufficiently small and have no path information
388 into a canonical native form. Eg. "blah" converted to
389 "libblah.so" or "blah.dll" etc.
392 *) New function ERR_error_string_n(e, buf, len) which is like
393 ERR_error_string(e, buf), but writes at most 'len' bytes
394 including the 0 terminator. For ERR_error_string_n, 'buf'
396 [Damien Miller <djm@mindrot.org>, Bodo Moeller]
398 *) CONF library reworked to become more general. A new CONF
399 configuration file reader "class" is implemented as well as a
400 new functions (NCONF_*, for "New CONF") to handle it. The now
401 old CONF_* functions are still there, but are reimplemented to
402 work in terms of the new functions. Also, a set of functions
403 to handle the internal storage of the configuration data is
404 provided to make it easier to write new configuration file
405 reader "classes" (I can definitely see something reading a
406 configuration file in XML format, for example), called _CONF_*,
407 or "the configuration storage API"...
409 The new configuration file reading functions are:
411 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio,
412 NCONF_get_section, NCONF_get_string, NCONF_get_numbre
414 NCONF_default, NCONF_WIN32
416 NCONF_dump_fp, NCONF_dump_bio
418 NCONF_default and NCONF_WIN32 are method (or "class") choosers,
419 NCONF_new creates a new CONF object. This works in the same way
420 as other interfaces in OpenSSL, like the BIO interface.
421 NCONF_dump_* dump the internal storage of the configuration file,
422 which is useful for debugging. All other functions take the same
423 arguments as the old CONF_* functions wth the exception of the
424 first that must be a `CONF *' instead of a `LHASH *'.
426 To make it easer to use the new classes with the old CONF_* functions,
427 the function CONF_set_default_method is provided.
430 *) Add '-tls1' option to 'openssl ciphers', which was already
431 mentioned in the documentation but had not been implemented.
432 (This option is not yet really useful because even the additional
433 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.)
436 *) Initial DSO code added into libcrypto for letting OpenSSL (and
437 OpenSSL-based applications) load shared libraries and bind to
438 them in a portable way.
439 [Geoff Thorpe, with contributions from Richard Levitte]
441 Changes between 0.9.5 and 0.9.5a [1 Apr 2000]
443 *) Make sure _lrotl and _lrotr are only used with MSVC.
445 *) Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status
446 (the default implementation of RAND_status).
448 *) Rename openssl x509 option '-crlext', which was added in 0.9.5,
449 to '-clrext' (= clear extensions), as intended and documented.
450 [Bodo Moeller; inconsistency pointed out by Michael Attili
453 *) Fix for HMAC. It wasn't zeroing the rest of the block if the key length
454 was larger than the MD block size.
455 [Steve Henson, pointed out by Yost William <YostW@tce.com>]
457 *) Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument
458 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set()
459 using the passed key: if the passed key was a private key the result
460 of X509_print(), for example, would be to print out all the private key
464 *) des_quad_cksum() byte order bug fix.
465 [Ulf Möller, using the problem description in krb4-0.9.7, where
466 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>]
468 *) Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly
470 [Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>]
472 *) For easily testing in shell scripts whether some command
473 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX'
474 returns with exit code 0 iff no command of the given name is available.
475 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases,
476 the output goes to stdout and nothing is printed to stderr.
477 Additional arguments are always ignored.
479 Since for each cipher there is a command of the same name,
480 the 'no-cipher' compilation switches can be tested this way.
482 ('openssl no-XXX' is not able to detect pseudo-commands such
483 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.)
486 *) Update test suite so that 'make test' succeeds in 'no-rsa' configuration.
489 *) For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE
490 is set; it will be thrown away anyway because each handshake creates
492 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition
493 to parameters -- in previous versions (since OpenSSL 0.9.3) the
494 'default key' from SSL_CTX_set_tmp_dh would always be lost, meanining
495 you effectivly got SSL_OP_SINGLE_DH_USE when using this macro.
498 *) New s_client option -ign_eof: EOF at stdin is ignored, and
499 'Q' and 'R' lose their special meanings (quit/renegotiate).
500 This is part of what -quiet does; unlike -quiet, -ign_eof
501 does not suppress any output.
504 *) Add compatibility options to the purpose and trust code. The
505 purpose X509_PURPOSE_ANY is "any purpose" which automatically
506 accepts a certificate or CA, this was the previous behaviour,
507 with all the associated security issues.
509 X509_TRUST_COMPAT is the old trust behaviour: only and
510 automatically trust self signed roots in certificate store. A
511 new trust setting X509_TRUST_DEFAULT is used to specify that
512 a purpose has no associated trust setting and it should instead
513 use the value in the default purpose.
516 *) Fix the PKCS#8 DSA private key code so it decodes keys again
517 and fix a memory leak.
520 *) In util/mkerr.pl (which implements 'make errors'), preserve
521 reason strings from the previous version of the .c file, as
522 the default to have only downcase letters (and digits) in
523 automatically generated reasons codes is not always appropriate.
526 *) In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table
527 using strerror. Previously, ERR_reason_error_string() returned
528 library names as reason strings for SYSerr; but SYSerr is a special
529 case where small numbers are errno values, not library numbers.
532 *) Add '-dsaparam' option to 'openssl dhparam' application. This
533 converts DSA parameters into DH parameters. (When creating parameters,
534 DSA_generate_parameters is used.)
537 *) Include 'length' (recommended exponent length) in C code generated
538 by 'openssl dhparam -C'.
541 *) The second argument to set_label in perlasm was already being used
542 so couldn't be used as a "file scope" flag. Moved to third argument
546 *) In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes
547 instead of RAND_bytes for encryption IVs and salts.
550 *) Include RAND_status() into RAND_METHOD instead of implementing
551 it only for md_rand.c Otherwise replacing the PRNG by calling
552 RAND_set_rand_method would be impossible.
555 *) Don't let DSA_generate_key() enter an infinite loop if the random
556 number generation fails.
559 *) New 'rand' application for creating pseudo-random output.
562 *) Added configuration support for Linux/IA64
563 [Rolf Haberrecker <rolf@suse.de>]
565 *) Assembler module support for Mingw32.
568 *) Shared library support for HPUX (in shlib/).
569 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous]
571 *) Shared library support for Solaris gcc.
572 [Lutz Behnke <behnke@trustcenter.de>]
574 Changes between 0.9.4 and 0.9.5 [28 Feb 2000]
576 *) PKCS7_encrypt() was adding text MIME headers twice because they
577 were added manually and by SMIME_crlf_copy().
580 *) In bntest.c don't call BN_rand with zero bits argument.
581 [Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>]
583 *) BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n]
584 case was implemented. This caused BN_div_recp() to fail occasionally.
587 *) Add an optional second argument to the set_label() in the perl
588 assembly language builder. If this argument exists and is set
589 to 1 it signals that the assembler should use a symbol whose
590 scope is the entire file, not just the current function. This
591 is needed with MASM which uses the format label:: for this scope.
592 [Steve Henson, pointed out by Peter Runestig <peter@runestig.com>]
594 *) Change the ASN1 types so they are typedefs by default. Before
595 almost all types were #define'd to ASN1_STRING which was causing
596 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING)
600 *) Change names of new functions to the new get1/get0 naming
601 convention: After 'get1', the caller owns a reference count
602 and has to call ..._free; 'get0' returns a pointer to some
603 data structure without incrementing reference counters.
604 (Some of the existing 'get' functions increment a reference
605 counter, some don't.)
606 Similarly, 'set1' and 'add1' functions increase reference
607 counters or duplicate objects.
610 *) Allow for the possibility of temp RSA key generation failure:
611 the code used to assume it always worked and crashed on failure.
614 *) Fix potential buffer overrun problem in BIO_printf().
615 [Ulf Möller, using public domain code by Patrick Powell; problem
616 pointed out by David Sacerdote <das33@cornell.edu>]
618 *) Support EGD <http://www.lothar.com/tech/crypto/>. New functions
619 RAND_egd() and RAND_status(). In the command line application,
620 the EGD socket can be specified like a seed file using RANDFILE
624 *) Allow the string CERTIFICATE to be tolerated in PKCS#7 structures.
625 Some CAs (e.g. Verisign) distribute certificates in this form.
628 *) Remove the SSL_ALLOW_ADH compile option and set the default cipher
629 list to exclude them. This means that no special compilation option
630 is needed to use anonymous DH: it just needs to be included in the
634 *) Change the EVP_MD_CTX_type macro so its meaning consistent with
635 EVP_MD_type. The old functionality is available in a new macro called
636 EVP_MD_md(). Change code that uses it and update docs.
639 *) ..._ctrl functions now have corresponding ..._callback_ctrl functions
640 where the 'void *' argument is replaced by a function pointer argument.
641 Previously 'void *' was abused to point to functions, which works on
642 many platforms, but is not correct. As these functions are usually
643 called by macros defined in OpenSSL header files, most source code
644 should work without changes.
647 *) <openssl/opensslconf.h> (which is created by Configure) now contains
648 sections with information on -D... compiler switches used for
649 compiling the library so that applications can see them. To enable
650 one of these sections, a pre-processor symbol OPENSSL_..._DEFINES
651 must be defined. E.g.,
652 #define OPENSSL_ALGORITHM_DEFINES
653 #include <openssl/opensslconf.h>
654 defines all pertinent NO_<algo> symbols, such as NO_IDEA, NO_RSA, etc.
655 [Richard Levitte, Ulf and Bodo Möller]
657 *) Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS
661 *) Change the 'other' type in certificate aux info to a STACK_OF
662 X509_ALGOR. Although not an AlgorithmIdentifier as such it has
663 the required ASN1 format: arbitrary types determined by an OID.
666 *) Add some PEM_write_X509_REQ_NEW() functions and a command line
667 argument to 'req'. This is not because the function is newer or
668 better than others it just uses the work 'NEW' in the certificate
669 request header lines. Some software needs this.
672 *) Reorganise password command line arguments: now passwords can be
673 obtained from various sources. Delete the PEM_cb function and make
674 it the default behaviour: i.e. if the callback is NULL and the
675 usrdata argument is not NULL interpret it as a null terminated pass
676 phrase. If usrdata and the callback are NULL then the pass phrase
677 is prompted for as usual.
680 *) Add support for the Compaq Atalla crypto accelerator. If it is installed,
681 the support is automatically enabled. The resulting binaries will
682 autodetect the card and use it if present.
683 [Ben Laurie and Compaq Inc.]
685 *) Work around for Netscape hang bug. This sends certificate request
686 and server done in one record. Since this is perfectly legal in the
687 SSL/TLS protocol it isn't a "bug" option and is on by default. See
688 the bugs/SSLv3 entry for more info.
691 *) HP-UX tune-up: new unified configs, HP C compiler bug workaround.
694 *) Add -rand argument to smime and pkcs12 applications and read/write
698 *) New 'passwd' tool for crypt(3) and apr1 password hashes.
701 *) Add command line password options to the remaining applications.
704 *) Bug fix for BN_div_recp() for numerators with an even number of
708 *) More tests in bntest.c, and changed test_bn output.
711 *) ./config recognizes MacOS X now.
714 *) Bug fix for BN_div() when the first words of num and divsor are
715 equal (it gave wrong results if (rem=(n1-q*d0)&BN_MASK2) < d0).
718 *) Add support for various broken PKCS#8 formats, and command line
719 options to produce them.
722 *) New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to
723 get temporary BIGNUMs from a BN_CTX.
726 *) Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont()
730 *) Change the SSLeay_add_all_*() functions to OpenSSL_add_all_*() and
731 include a #define from the old name to the new. The original intent
732 was that statically linked binaries could for example just call
733 SSLeay_add_all_ciphers() to just add ciphers to the table and not
734 link with digests. This never worked becayse SSLeay_add_all_digests()
735 and SSLeay_add_all_ciphers() were in the same source file so calling
736 one would link with the other. They are now in separate source files.
739 *) Add a new -notext option to 'ca' and a -pubkey option to 'spkac'.
742 *) Use a less unusual form of the Miller-Rabin primality test (it used
743 a binary algorithm for exponentiation integrated into the Miller-Rabin
744 loop, our standard modexp algorithms are faster).
747 *) Support for the EBCDIC character set completed.
748 [Martin Kraemer <Martin.Kraemer@Mch.SNI.De>]
750 *) Source code cleanups: use const where appropriate, eliminate casts,
751 use void * instead of char * in lhash.
754 *) Bugfix: ssl3_send_server_key_exchange was not restartable
755 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of
756 this the server could overwrite ephemeral keys that the client
760 *) Turn DSA_is_prime into a macro that calls BN_is_prime,
761 using 50 iterations of the Rabin-Miller test.
763 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50
764 iterations of the Rabin-Miller test as required by the appendix
765 to FIPS PUB 186[-1]) instead of DSA_is_prime.
766 As BN_is_prime_fasttest includes trial division, DSA parameter
767 generation becomes much faster.
769 This implies a change for the callback functions in DSA_is_prime
770 and DSA_generate_parameters: The callback function is called once
771 for each positive witness in the Rabin-Miller test, not just
772 occasionally in the inner loop; and the parameters to the
773 callback function now provide an iteration count for the outer
774 loop rather than for the current invocation of the inner loop.
775 DSA_generate_parameters additionally can call the callback
776 function with an 'iteration count' of -1, meaning that a
777 candidate has passed the trial division test (when q is generated
778 from an application-provided seed, trial division is skipped).
781 *) New function BN_is_prime_fasttest that optionally does trial
782 division before starting the Rabin-Miller test and has
783 an additional BN_CTX * argument (whereas BN_is_prime always
784 has to allocate at least one BN_CTX).
785 'callback(1, -1, cb_arg)' is called when a number has passed the
786 trial division stage.
789 *) Fix for bug in CRL encoding. The validity dates weren't being handled
793 *) New -pkcs12 option to CA.pl script to write out a PKCS#12 file.
796 *) New function BN_pseudo_rand().
799 *) Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable)
800 bignum version of BN_from_montgomery() with the working code from
801 SSLeay 0.9.0 (the word based version is faster anyway), and clean up
805 *) Avoid a race condition in s2_clnt.c (function get_server_hello) that
806 made it impossible to use the same SSL_SESSION data structure in
807 SSL2 clients in multiple threads.
810 *) The return value of RAND_load_file() no longer counts bytes obtained
811 by stat(). RAND_load_file(..., -1) is new and uses the complete file
812 to seed the PRNG (previously an explicit byte count was required).
813 [Ulf Möller, Bodo Möller]
815 *) Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
816 used (char *) instead of (void *) and had casts all over the place.
819 *) Make BN_generate_prime() return NULL on error if ret!=NULL.
822 *) Retain source code compatibility for BN_prime_checks macro:
823 BN_is_prime(..., BN_prime_checks, ...) now uses
824 BN_prime_checks_for_size to determine the appropriate number of
825 Rabin-Miller iterations.
828 *) Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
829 DH_CHECK_P_NOT_SAFE_PRIME.
830 (Check if this is true? OpenPGP calls them "strong".)
833 *) Merge the functionality of "dh" and "gendh" programs into a new program
834 "dhparam". The old programs are retained for now but will handle DH keys
835 (instead of parameters) in future.
838 *) Make the ciphers, s_server and s_client programs check the return values
839 when a new cipher list is set.
842 *) Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
843 ciphers. Before when the 56bit ciphers were enabled the sorting was
846 The syntax for the cipher sorting has been extended to support sorting by
847 cipher-strength (using the strength_bits hard coded in the tables).
848 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
850 Fix a bug in the cipher-command parser: when supplying a cipher command
851 string with an "undefined" symbol (neither command nor alphanumeric
852 [A-Za-z0-9], ssl_set_cipher_list used to hang in an endless loop. Now
855 Due to the strength-sorting extension, the code of the
856 ssl_create_cipher_list() function was completely rearranged. I hope that
857 the readability was also increased :-)
858 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
860 *) Minor change to 'x509' utility. The -CAcreateserial option now uses 1
861 for the first serial number and places 2 in the serial number file. This
862 avoids problems when the root CA is created with serial number zero and
863 the first user certificate has the same issuer name and serial number
867 *) Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
868 the new code. Add documentation for this stuff.
871 *) Changes to X509_ATTRIBUTE utilities. These have been renamed from
872 X509_*() to X509at_*() on the grounds that they don't handle X509
873 structures and behave in an analagous way to the X509v3 functions:
874 they shouldn't be called directly but wrapper functions should be used
877 So we also now have some wrapper functions that call the X509at functions
878 when passed certificate requests. (TO DO: similar things can be done with
879 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
880 things. Some of these need some d2i or i2d and print functionality
881 because they handle more complex structures.)
884 *) Add missing #ifndefs that caused missing symbols when building libssl
885 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
887 [Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller]
889 *) Precautions against using the PRNG uninitialized: RAND_bytes() now
890 has a return value which indicates the quality of the random data
891 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
892 error queue. New function RAND_pseudo_bytes() generates output that is
893 guaranteed to be unique but not unpredictable. RAND_add is like
894 RAND_seed, but takes an extra argument for an entropy estimate
895 (RAND_seed always assumes full entropy).
898 *) Do more iterations of Rabin-Miller probable prime test (specifically,
899 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
900 instead of only 2 for all lengths; see BN_prime_checks_for_size definition
901 in crypto/bn/bn_prime.c for the complete table). This guarantees a
902 false-positive rate of at most 2^-80 for random input.
905 *) Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
908 *) New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain
909 in the 0.9.5 release), this returns the chain
910 from an X509_CTX structure with a dup of the stack and all
911 the X509 reference counts upped: so the stack will exist
912 after X509_CTX_cleanup() has been called. Modify pkcs12.c
915 Also make SSL_SESSION_print() print out the verify return
919 *) Add manpage for the pkcs12 command. Also change the default
920 behaviour so MAC iteration counts are used unless the new
921 -nomaciter option is used. This improves file security and
922 only older versions of MSIE (4.0 for example) need it.
925 *) Honor the no-xxx Configure options when creating .DEF files.
928 *) Add PKCS#10 attributes to field table: challengePassword,
929 unstructuredName and unstructuredAddress. These are taken from
930 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
931 international characters are used.
933 More changes to X509_ATTRIBUTE code: allow the setting of types
934 based on strings. Remove the 'loc' parameter when adding
935 attributes because these will be a SET OF encoding which is sorted
939 *) Initial changes to the 'req' utility to allow request generation
940 automation. This will allow an application to just generate a template
941 file containing all the field values and have req construct the
944 Initial support for X509_ATTRIBUTE handling. Stacks of these are
945 used all over the place including certificate requests and PKCS#7
946 structures. They are currently handled manually where necessary with
947 some primitive wrappers for PKCS#7. The new functions behave in a
948 manner analogous to the X509 extension functions: they allow
949 attributes to be looked up by NID and added.
951 Later something similar to the X509V3 code would be desirable to
952 automatically handle the encoding, decoding and printing of the
953 more complex types. The string types like challengePassword can
954 be handled by the string table functions.
956 Also modified the multi byte string table handling. Now there is
957 a 'global mask' which masks out certain types. The table itself
958 can use the flag STABLE_NO_MASK to ignore the mask setting: this
959 is useful when for example there is only one permissible type
960 (as in countryName) and using the mask might result in no valid
964 *) Clean up 'Finished' handling, and add functions SSL_get_finished and
965 SSL_get_peer_finished to allow applications to obtain the latest
966 Finished messages sent to the peer or expected from the peer,
967 respectively. (SSL_get_peer_finished is usually the Finished message
968 actually received from the peer, otherwise the protocol will be aborted.)
970 As the Finished message are message digests of the complete handshake
971 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
972 be used for external authentication procedures when the authentication
973 provided by SSL/TLS is not desired or is not enough.
976 *) Enhanced support for Alpha Linux is added. Now ./config checks if
977 the host supports BWX extension and if Compaq C is present on the
978 $PATH. Just exploiting of the BWX extension results in 20-30%
979 performance kick for some algorithms, e.g. DES and RC4 to mention
980 a couple. Compaq C in turn generates ~20% faster code for MD5 and
984 *) Add support for MS "fast SGC". This is arguably a violation of the
985 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
986 weak crypto and after checking the certificate is SGC a second one
987 with strong crypto. MS SGC stops the first handshake after receiving
988 the server certificate message and sends a second client hello. Since
989 a server will typically do all the time consuming operations before
990 expecting any further messages from the client (server key exchange
991 is the most expensive) there is little difference between the two.
993 To get OpenSSL to support MS SGC we have to permit a second client
994 hello message after we have sent server done. In addition we have to
995 reset the MAC if we do get this second client hello.
998 *) Add a function 'd2i_AutoPrivateKey()' this will automatically decide
999 if a DER encoded private key is RSA or DSA traditional format. Changed
1000 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
1001 format DER encoded private key. Newer code should use PKCS#8 format which
1002 has the key type encoded in the ASN1 structure. Added DER private key
1003 support to pkcs8 application.
1006 *) SSL 3/TLS 1 servers now don't request certificates when an anonymous
1007 ciphersuites has been selected (as required by the SSL 3/TLS 1
1008 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
1009 is set, we interpret this as a request to violate the specification
1010 (the worst that can happen is a handshake failure, and 'correct'
1011 behaviour would result in a handshake failure anyway).
1014 *) In SSL_CTX_add_session, take into account that there might be multiple
1015 SSL_SESSION structures with the same session ID (e.g. when two threads
1016 concurrently obtain them from an external cache).
1017 The internal cache can handle only one SSL_SESSION with a given ID,
1018 so if there's a conflict, we now throw out the old one to achieve
1022 *) Add OIDs for idea and blowfish in CBC mode. This will allow both
1023 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
1024 some routines that use cipher OIDs: some ciphers do not have OIDs
1025 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
1029 *) Simplify the trust setting structure and code. Now we just have
1030 two sequences of OIDs for trusted and rejected settings. These will
1031 typically have values the same as the extended key usage extension
1032 and any application specific purposes.
1034 The trust checking code now has a default behaviour: it will just
1035 check for an object with the same NID as the passed id. Functions can
1036 be provided to override either the default behaviour or the behaviour
1037 for a given id. SSL client, server and email already have functions
1038 in place for compatibility: they check the NID and also return "trusted"
1039 if the certificate is self signed.
1042 *) Add d2i,i2d bio/fp functions for PrivateKey: these convert the
1043 traditional format into an EVP_PKEY structure.
1046 *) Add a password callback function PEM_cb() which either prompts for
1047 a password if usr_data is NULL or otherwise assumes it is a null
1048 terminated password. Allow passwords to be passed on command line
1049 environment or config files in a few more utilities.
1052 *) Add a bunch of DER and PEM functions to handle PKCS#8 format private
1053 keys. Add some short names for PKCS#8 PBE algorithms and allow them
1054 to be specified on the command line for the pkcs8 and pkcs12 utilities.
1055 Update documentation.
1058 *) Support for ASN1 "NULL" type. This could be handled before by using
1059 ASN1_TYPE but there wasn't any function that would try to read a NULL
1060 and produce an error if it couldn't. For compatibility we also have
1061 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
1062 don't allocate anything because they don't need to.
1065 *) Initial support for MacOS is now provided. Examine INSTALL.MacOS
1067 [Andy Polyakov, Roy Woods <roy@centicsystems.ca>]
1069 *) Rebuild of the memory allocation routines used by OpenSSL code and
1070 possibly others as well. The purpose is to make an interface that
1071 provide hooks so anyone can build a separate set of allocation and
1072 deallocation routines to be used by OpenSSL, for example memory
1073 pool implementations, or something else, which was previously hard
1074 since Malloc(), Realloc() and Free() were defined as macros having
1075 the values malloc, realloc and free, respectively (except for Win32
1076 compilations). The same is provided for memory debugging code.
1077 OpenSSL already comes with functionality to find memory leaks, but
1078 this gives people a chance to debug other memory problems.
1080 With these changes, a new set of functions and macros have appeared:
1082 CRYPTO_set_mem_debug_functions() [F]
1083 CRYPTO_get_mem_debug_functions() [F]
1084 CRYPTO_dbg_set_options() [F]
1085 CRYPTO_dbg_get_options() [F]
1086 CRYPTO_malloc_debug_init() [M]
1088 The memory debug functions are NULL by default, unless the library
1089 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
1090 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
1091 gives the standard debugging functions that come with OpenSSL) or
1092 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
1093 provided by the library user) must be used. When the standard
1094 debugging functions are used, CRYPTO_dbg_set_options can be used to
1095 request additional information:
1096 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
1097 the CRYPTO_MDEBUG_xxx macro when compiling the library.
1099 Also, things like CRYPTO_set_mem_functions will always give the
1100 expected result (the new set of functions is used for allocation
1101 and deallocation) at all times, regardless of platform and compiler
1104 To finish it up, some functions that were never use in any other
1105 way than through macros have a new API and new semantic:
1108 CRYPTO_dbg_realloc()
1111 All macros of value have retained their old syntax.
1112 [Richard Levitte and Bodo Moeller]
1114 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
1115 ordering of SMIMECapabilities wasn't in "strength order" and there
1116 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
1120 *) Some ASN1 types with illegal zero length encoding (INTEGER,
1121 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
1122 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson]
1124 *) Merge in my S/MIME library for OpenSSL. This provides a simple
1125 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
1126 functionality to handle multipart/signed properly) and a utility
1127 called 'smime' to call all this stuff. This is based on code I
1128 originally wrote for Celo who have kindly allowed it to be
1129 included in OpenSSL.
1132 *) Add variants des_set_key_checked and des_set_key_unchecked of
1133 des_set_key (aka des_key_sched). Global variable des_check_key
1134 decides which of these is called by des_set_key; this way
1135 des_check_key behaves as it always did, but applications and
1136 the library itself, which was buggy for des_check_key == 1,
1137 have a cleaner way to pick the version they need.
1140 *) New function PKCS12_newpass() which changes the password of a
1144 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and
1145 dynamic mix. In both cases the ids can be used as an index into the
1146 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
1147 functions so they accept a list of the field values and the
1148 application doesn't need to directly manipulate the X509_TRUST
1152 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
1156 *) Modify the way the V3 extension code looks up extensions. This now
1157 works in a similar way to the object code: we have some "standard"
1158 extensions in a static table which is searched with OBJ_bsearch()
1159 and the application can add dynamic ones if needed. The file
1160 crypto/x509v3/ext_dat.h now has the info: this file needs to be
1161 updated whenever a new extension is added to the core code and kept
1162 in ext_nid order. There is a simple program 'tabtest.c' which checks
1163 this. New extensions are not added too often so this file can readily
1164 be maintained manually.
1166 There are two big advantages in doing things this way. The extensions
1167 can be looked up immediately and no longer need to be "added" using
1168 X509V3_add_standard_extensions(): this function now does nothing.
1169 [Side note: I get *lots* of email saying the extension code doesn't
1170 work because people forget to call this function]
1171 Also no dynamic allocation is done unless new extensions are added:
1172 so if we don't add custom extensions there is no need to call
1173 X509V3_EXT_cleanup().
1176 *) Modify enc utility's salting as follows: make salting the default. Add a
1177 magic header, so unsalted files fail gracefully instead of just decrypting
1178 to garbage. This is because not salting is a big security hole, so people
1179 should be discouraged from doing it.
1182 *) Fixes and enhancements to the 'x509' utility. It allowed a message
1183 digest to be passed on the command line but it only used this
1184 parameter when signing a certificate. Modified so all relevant
1185 operations are affected by the digest parameter including the
1186 -fingerprint and -x509toreq options. Also -x509toreq choked if a
1187 DSA key was used because it didn't fix the digest.
1190 *) Initial certificate chain verify code. Currently tests the untrusted
1191 certificates for consistency with the verify purpose (which is set
1192 when the X509_STORE_CTX structure is set up) and checks the pathlength.
1194 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
1195 this is because it will reject chains with invalid extensions whereas
1196 every previous version of OpenSSL and SSLeay made no checks at all.
1198 Trust code: checks the root CA for the relevant trust settings. Trust
1199 settings have an initial value consistent with the verify purpose: e.g.
1200 if the verify purpose is for SSL client use it expects the CA to be
1201 trusted for SSL client use. However the default value can be changed to
1202 permit custom trust settings: one example of this would be to only trust
1203 certificates from a specific "secure" set of CAs.
1205 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
1206 which should be used for version portability: especially since the
1207 verify structure is likely to change more often now.
1209 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
1210 to set them. If not set then assume SSL clients will verify SSL servers
1213 Two new options to the verify program: -untrusted allows a set of
1214 untrusted certificates to be passed in and -purpose which sets the
1215 intended purpose of the certificate. If a purpose is set then the
1216 new chain verify code is used to check extension consistency.
1219 *) Support for the authority information access extension.
1222 *) Modify RSA and DSA PEM read routines to transparently handle
1223 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
1224 public keys in a format compatible with certificate
1225 SubjectPublicKeyInfo structures. Unfortunately there were already
1226 functions called *_PublicKey_* which used various odd formats so
1227 these are retained for compatibility: however the DSA variants were
1228 never in a public release so they have been deleted. Changed dsa/rsa
1229 utilities to handle the new format: note no releases ever handled public
1230 keys so we should be OK.
1232 The primary motivation for this change is to avoid the same fiasco
1233 that dogs private keys: there are several incompatible private key
1234 formats some of which are standard and some OpenSSL specific and
1235 require various evil hacks to allow partial transparent handling and
1236 even then it doesn't work with DER formats. Given the option anything
1237 other than PKCS#8 should be dumped: but the other formats have to
1238 stay in the name of compatibility.
1240 With public keys and the benefit of hindsight one standard format
1241 is used which works with EVP_PKEY, RSA or DSA structures: though
1242 it clearly returns an error if you try to read the wrong kind of key.
1244 Added a -pubkey option to the 'x509' utility to output the public key.
1245 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*()
1246 (renamed to EVP_PKEY_get1_*() in the OpenSSL 0.9.5 release) and add
1247 EVP_PKEY_rset_*() functions (renamed to EVP_PKEY_set1_*())
1248 that do the same as the EVP_PKEY_assign_*() except they up the
1249 reference count of the added key (they don't "swallow" the
1253 *) Fixes to crypto/x509/by_file.c the code to read in certificates and
1254 CRLs would fail if the file contained no certificates or no CRLs:
1255 added a new function to read in both types and return the number
1256 read: this means that if none are read it will be an error. The
1257 DER versions of the certificate and CRL reader would always fail
1258 because it isn't possible to mix certificates and CRLs in DER format
1259 without choking one or the other routine. Changed this to just read
1260 a certificate: this is the best we can do. Also modified the code
1261 in apps/verify.c to take notice of return codes: it was previously
1262 attempting to read in certificates from NULL pointers and ignoring
1263 any errors: this is one reason why the cert and CRL reader seemed
1264 to work. It doesn't check return codes from the default certificate
1265 routines: these may well fail if the certificates aren't installed.
1268 *) Code to support otherName option in GeneralName.
1271 *) First update to verify code. Change the verify utility
1272 so it warns if it is passed a self signed certificate:
1273 for consistency with the normal behaviour. X509_verify
1274 has been modified to it will now verify a self signed
1275 certificate if *exactly* the same certificate appears
1276 in the store: it was previously impossible to trust a
1277 single self signed certificate. This means that:
1278 openssl verify ss.pem
1279 now gives a warning about a self signed certificate but
1280 openssl verify -CAfile ss.pem ss.pem
1284 *) For servers, store verify_result in SSL_SESSION data structure
1285 (and add it to external session representation).
1286 This is needed when client certificate verifications fails,
1287 but an application-provided verification callback (set by
1288 SSL_CTX_set_cert_verify_callback) allows accepting the session
1289 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
1290 but returns 1): When the session is reused, we have to set
1291 ssl->verify_result to the appropriate error code to avoid
1293 [Bodo Moeller, problem pointed out by Lutz Jaenicke]
1295 *) Fix a bug in the new PKCS#7 code: it didn't consider the
1296 case in PKCS7_dataInit() where the signed PKCS7 structure
1297 didn't contain any existing data because it was being created.
1298 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson]
1300 *) Add a salt to the key derivation routines in enc.c. This
1301 forms the first 8 bytes of the encrypted file. Also add a
1302 -S option to allow a salt to be input on the command line.
1305 *) New function X509_cmp(). Oddly enough there wasn't a function
1306 to compare two certificates. We do this by working out the SHA1
1307 hash and comparing that. X509_cmp() will be needed by the trust
1311 *) SSL_get1_session() is like SSL_get_session(), but increments
1312 the reference count in the SSL_SESSION returned.
1313 [Geoff Thorpe <geoff@eu.c2.net>]
1315 *) Fix for 'req': it was adding a null to request attributes.
1316 Also change the X509_LOOKUP and X509_INFO code to handle
1317 certificate auxiliary information.
1320 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document
1324 *) Add the possibility to add extra information to the memory leak
1325 detecting output, to form tracebacks, showing from where each
1326 allocation was originated: CRYPTO_push_info("constant string") adds
1327 the string plus current file name and line number to a per-thread
1328 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
1329 is like calling CYRPTO_pop_info() until the stack is empty.
1330 Also updated memory leak detection code to be multi-thread-safe.
1333 *) Add options -text and -noout to pkcs7 utility and delete the
1334 encryption options which never did anything. Update docs.
1337 *) Add options to some of the utilities to allow the pass phrase
1338 to be included on either the command line (not recommended on
1339 OSes like Unix) or read from the environment. Update the
1340 manpages and fix a few bugs.
1343 *) Add a few manpages for some of the openssl commands.
1346 *) Fix the -revoke option in ca. It was freeing up memory twice,
1347 leaking and not finding already revoked certificates.
1350 *) Extensive changes to support certificate auxiliary information.
1351 This involves the use of X509_CERT_AUX structure and X509_AUX
1352 functions. An X509_AUX function such as PEM_read_X509_AUX()
1353 can still read in a certificate file in the usual way but it
1354 will also read in any additional "auxiliary information". By
1355 doing things this way a fair degree of compatibility can be
1356 retained: existing certificates can have this information added
1357 using the new 'x509' options.
1359 Current auxiliary information includes an "alias" and some trust
1360 settings. The trust settings will ultimately be used in enhanced
1361 certificate chain verification routines: currently a certificate
1362 can only be trusted if it is self signed and then it is trusted
1366 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD).
1367 The problem was that one of the replacement routines had not been working
1368 since SSLeay releases. For now the offending routine has been replaced
1369 with non-optimised assembler. Even so, this now gives around 95%
1370 performance improvement for 1024 bit RSA signs.
1373 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2
1374 handling. Most clients have the effective key size in bits equal to
1375 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
1376 A few however don't do this and instead use the size of the decrypted key
1377 to determine the RC2 key length and the AlgorithmIdentifier to determine
1378 the effective key length. In this case the effective key length can still
1379 be 40 bits but the key length can be 168 bits for example. This is fixed
1380 by manually forcing an RC2 key into the EVP_PKEY structure because the
1381 EVP code can't currently handle unusual RC2 key sizes: it always assumes
1382 the key length and effective key length are equal.
1385 *) Add a bunch of functions that should simplify the creation of
1386 X509_NAME structures. Now you should be able to do:
1387 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
1388 and have it automatically work out the correct field type and fill in
1389 the structures. The more adventurous can try:
1390 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
1391 and it will (hopefully) work out the correct multibyte encoding.
1394 *) Change the 'req' utility to use the new field handling and multibyte
1395 copy routines. Before the DN field creation was handled in an ad hoc
1396 way in req, ca, and x509 which was rather broken and didn't support
1397 BMPStrings or UTF8Strings. Since some software doesn't implement
1398 BMPStrings or UTF8Strings yet, they can be enabled using the config file
1399 using the dirstring_type option. See the new comment in the default
1400 openssl.cnf for more info.
1403 *) Make crypto/rand/md_rand.c more robust:
1404 - Assure unique random numbers after fork().
1405 - Make sure that concurrent threads access the global counter and
1406 md serializably so that we never lose entropy in them
1407 or use exactly the same state in multiple threads.
1408 Access to the large state is not always serializable because
1409 the additional locking could be a performance killer, and
1410 md should be large enough anyway.
1413 *) New file apps/app_rand.c with commonly needed functionality
1414 for handling the random seed file.
1416 Use the random seed file in some applications that previously did not:
1418 dsaparam -genkey (which also ignored its '-rand' option),
1421 x509 (when signing).
1422 Except on systems with /dev/urandom, it is crucial to have a random
1423 seed file at least for key creation, DSA signing, and for DH exchanges;
1424 for RSA signatures we could do without one.
1426 gendh and gendsa (unlike genrsa) used to read only the first byte
1427 of each file listed in the '-rand' option. The function as previously
1428 found in genrsa is now in app_rand.c and is used by all programs
1429 that support '-rand'.
1432 *) In RAND_write_file, use mode 0600 for creating files;
1433 don't just chmod when it may be too late.
1436 *) Report an error from X509_STORE_load_locations
1437 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
1440 *) New function ASN1_mbstring_copy() this copies a string in either
1441 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
1442 into an ASN1_STRING type. A mask of permissible types is passed
1443 and it chooses the "minimal" type to use or an error if not type
1447 *) Add function equivalents to the various macros in asn1.h. The old
1448 macros are retained with an M_ prefix. Code inside the library can
1449 use the M_ macros. External code (including the openssl utility)
1450 should *NOT* in order to be "shared library friendly".
1453 *) Add various functions that can check a certificate's extensions
1454 to see if it usable for various purposes such as SSL client,
1455 server or S/MIME and CAs of these types. This is currently
1456 VERY EXPERIMENTAL but will ultimately be used for certificate chain
1457 verification. Also added a -purpose flag to x509 utility to
1458 print out all the purposes.
1461 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated
1465 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search
1466 for, obtain and decode and extension and obtain its critical flag.
1467 This allows all the necessary extension code to be handled in a
1468 single function call.
1471 *) RC4 tune-up featuring 30-40% performance improvement on most RISC
1472 platforms. See crypto/rc4/rc4_enc.c for further details.
1475 *) New -noout option to asn1parse. This causes no output to be produced
1476 its main use is when combined with -strparse and -out to extract data
1477 from a file (which may not be in ASN.1 format).
1480 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer
1481 when producing the local key id.
1482 [Richard Levitte <levitte@stacken.kth.se>]
1484 *) New option -dhparam in s_server. This allows a DH parameter file to be
1485 stated explicitly. If it is not stated then it tries the first server
1486 certificate file. The previous behaviour hard coded the filename
1490 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow
1491 a public key to be input or output. For example:
1492 openssl rsa -in key.pem -pubout -out pubkey.pem
1493 Also added necessary DSA public key functions to handle this.
1496 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
1497 in the message. This was handled by allowing
1498 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
1499 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>]
1501 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
1502 to the end of the strings whereas this didn't. This would cause problems
1503 if strings read with d2i_ASN1_bytes() were later modified.
1504 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>]
1506 *) Fix for base64 decode bug. When a base64 bio reads only one line of
1507 data and it contains EOF it will end up returning an error. This is
1508 caused by input 46 bytes long. The cause is due to the way base64
1509 BIOs find the start of base64 encoded data. They do this by trying a
1510 trial decode on each line until they find one that works. When they
1511 do a flag is set and it starts again knowing it can pass all the
1512 data directly through the decoder. Unfortunately it doesn't reset
1513 the context it uses. This means that if EOF is reached an attempt
1514 is made to pass two EOFs through the context and this causes the
1515 resulting error. This can also cause other problems as well. As is
1516 usual with these problems it takes *ages* to find and the fix is
1517 trivial: move one line.
1518 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ]
1520 *) Ugly workaround to get s_client and s_server working under Windows. The
1521 old code wouldn't work because it needed to select() on sockets and the
1522 tty (for keypresses and to see if data could be written). Win32 only
1523 supports select() on sockets so we select() with a 1s timeout on the
1524 sockets and then see if any characters are waiting to be read, if none
1525 are present then we retry, we also assume we can always write data to
1526 the tty. This isn't nice because the code then blocks until we've
1527 received a complete line of data and it is effectively polling the
1528 keyboard at 1s intervals: however it's quite a bit better than not
1529 working at all :-) A dedicated Windows application might handle this
1530 with an event loop for example.
1533 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
1534 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
1535 will be called when RSA_sign() and RSA_verify() are used. This is useful
1536 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
1537 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
1538 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
1539 This necessitated the support of an extra signature type NID_md5_sha1
1540 for SSL signatures and modifications to the SSL library to use it instead
1541 of calling RSA_public_decrypt() and RSA_private_encrypt().
1544 *) Add new -verify -CAfile and -CApath options to the crl program, these
1545 will lookup a CRL issuers certificate and verify the signature in a
1546 similar way to the verify program. Tidy up the crl program so it
1547 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
1548 less strict. It will now permit CRL extensions even if it is not
1549 a V2 CRL: this will allow it to tolerate some broken CRLs.
1552 *) Initialize all non-automatic variables each time one of the openssl
1553 sub-programs is started (this is necessary as they may be started
1554 multiple times from the "OpenSSL>" prompt).
1555 [Lennart Bang, Bodo Moeller]
1557 *) Preliminary compilation option RSA_NULL which disables RSA crypto without
1558 removing all other RSA functionality (this is what NO_RSA does). This
1559 is so (for example) those in the US can disable those operations covered
1560 by the RSA patent while allowing storage and parsing of RSA keys and RSA
1564 *) Non-copying interface to BIO pairs.
1565 (still largely untested)
1568 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive
1569 ASCII string. This was handled independently in various places before.
1572 *) New functions UTF8_getc() and UTF8_putc() that parse and generate
1573 UTF8 strings a character at a time.
1576 *) Use client_version from client hello to select the protocol
1577 (s23_srvr.c) and for RSA client key exchange verification
1578 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
1581 *) Add various utility functions to handle SPKACs, these were previously
1582 handled by poking round in the structure internals. Added new function
1583 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
1584 print, verify and generate SPKACs. Based on an original idea from
1585 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
1588 *) RIPEMD160 is operational on all platforms and is back in 'make test'.
1591 *) Allow the config file extension section to be overwritten on the
1592 command line. Based on an original idea from Massimiliano Pala
1593 <madwolf@comune.modena.it>. The new option is called -extensions
1594 and can be applied to ca, req and x509. Also -reqexts to override
1595 the request extensions in req and -crlexts to override the crl extensions
1599 *) Add new feature to the SPKAC handling in ca. Now you can include
1600 the same field multiple times by preceding it by "XXXX." for example:
1603 this is the same syntax as used in the req config file.
1606 *) Allow certificate extensions to be added to certificate requests. These
1607 are specified in a 'req_extensions' option of the req section of the
1608 config file. They can be printed out with the -text option to req but
1609 are otherwise ignored at present.
1612 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
1613 data read consists of only the final block it would not decrypted because
1614 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
1615 A misplaced 'break' also meant the decrypted final block might not be
1616 copied until the next read.
1619 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
1620 a few extra parameters to the DH structure: these will be useful if
1621 for example we want the value of 'q' or implement X9.42 DH.
1624 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and
1625 provides hooks that allow the default DSA functions or functions on a
1626 "per key" basis to be replaced. This allows hardware acceleration and
1627 hardware key storage to be handled without major modification to the
1628 library. Also added low level modexp hooks and CRYPTO_EX structure and
1629 associated functions.
1632 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
1633 as "read only": it can't be written to and the buffer it points to will
1634 not be freed. Reading from a read only BIO is much more efficient than
1635 a normal memory BIO. This was added because there are several times when
1636 an area of memory needs to be read from a BIO. The previous method was
1637 to create a memory BIO and write the data to it, this results in two
1638 copies of the data and an O(n^2) reading algorithm. There is a new
1639 function BIO_new_mem_buf() which creates a read only memory BIO from
1640 an area of memory. Also modified the PKCS#7 routines to use read only
1644 *) Bugfix: ssl23_get_client_hello did not work properly when called in
1645 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
1646 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
1647 but a retry condition occured while trying to read the rest.
1650 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as
1651 NID_pkcs7_encrypted by default: this was wrong since this should almost
1652 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
1653 the encrypted data type: this is a more sensible place to put it and it
1654 allows the PKCS#12 code to be tidied up that duplicated this
1658 *) Changed obj_dat.pl script so it takes its input and output files on
1659 the command line. This should avoid shell escape redirection problems
1663 *) Initial support for certificate extension requests, these are included
1664 in things like Xenroll certificate requests. Included functions to allow
1665 extensions to be obtained and added.
1668 *) -crlf option to s_client and s_server for sending newlines as
1669 CRLF (as required by many protocols).
1672 Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
1674 *) Install libRSAglue.a when OpenSSL is built with RSAref.
1675 [Ralf S. Engelschall]
1677 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency.
1678 [Andrija Antonijevic <TheAntony2@bigfoot.com>]
1680 *) Fix -startdate and -enddate (which was missing) arguments to 'ca'
1684 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as
1685 DH parameters/keys (q is lost during that conversion, but the resulting
1686 DH parameters contain its length).
1688 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
1689 much faster than DH_generate_parameters (which creates parameters
1690 where p = 2*q + 1), and also the smaller q makes DH computations
1691 much more efficient (160-bit exponentiation instead of 1024-bit
1692 exponentiation); so this provides a convenient way to support DHE
1693 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
1694 utter importance to use
1695 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
1697 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
1698 when such DH parameters are used, because otherwise small subgroup
1699 attacks may become possible!
1702 *) Avoid memory leak in i2d_DHparams.
1705 *) Allow the -k option to be used more than once in the enc program:
1706 this allows the same encrypted message to be read by multiple recipients.
1709 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
1710 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
1711 it will always use the numerical form of the OID, even if it has a short
1715 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
1716 method only got called if p,q,dmp1,dmq1,iqmp components were present,
1717 otherwise bn_mod_exp was called. In the case of hardware keys for example
1718 no private key components need be present and it might store extra data
1719 in the RSA structure, which cannot be accessed from bn_mod_exp.
1720 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
1721 private key operations.
1724 *) Added support for SPARC Linux.
1727 *) pem_password_cb function type incompatibly changed from
1728 typedef int pem_password_cb(char *buf, int size, int rwflag);
1730 ....(char *buf, int size, int rwflag, void *userdata);
1731 so that applications can pass data to their callbacks:
1732 The PEM[_ASN1]_{read,write}... functions and macros now take an
1733 additional void * argument, which is just handed through whenever
1734 the password callback is called.
1735 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller]
1737 New function SSL_CTX_set_default_passwd_cb_userdata.
1739 Compatibility note: As many C implementations push function arguments
1740 onto the stack in reverse order, the new library version is likely to
1741 interoperate with programs that have been compiled with the old
1742 pem_password_cb definition (PEM_whatever takes some data that
1743 happens to be on the stack as its last argument, and the callback
1744 just ignores this garbage); but there is no guarantee whatsoever that
1747 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
1748 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
1749 problems not only on Windows, but also on some Unix platforms.
1750 To avoid problematic command lines, these definitions are now in an
1751 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
1752 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
1755 *) MIPS III/IV assembler module is reimplemented.
1758 *) More DES library cleanups: remove references to srand/rand and
1759 delete an unused file.
1762 *) Add support for the the free Netwide assembler (NASM) under Win32,
1763 since not many people have MASM (ml) and it can be hard to obtain.
1764 This is currently experimental but it seems to work OK and pass all
1765 the tests. Check out INSTALL.W32 for info.
1768 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
1769 without temporary keys kept an extra copy of the server key,
1770 and connections with temporary keys did not free everything in case
1774 *) New function RSA_check_key and new openssl rsa option -check
1775 for verifying the consistency of RSA keys.
1776 [Ulf Moeller, Bodo Moeller]
1778 *) Various changes to make Win32 compile work:
1779 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
1780 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
1781 comparison" warnings.
1782 3. Add sk_<TYPE>_sort to DEF file generator and do make update.
1785 *) Add a debugging option to PKCS#5 v2 key generation function: when
1786 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
1787 derived keys are printed to stderr.
1790 *) Copy the flags in ASN1_STRING_dup().
1791 [Roman E. Pavlov <pre@mo.msk.ru>]
1793 *) The x509 application mishandled signing requests containing DSA
1794 keys when the signing key was also DSA and the parameters didn't match.
1796 It was supposed to omit the parameters when they matched the signing key:
1797 the verifying software was then supposed to automatically use the CA's
1798 parameters if they were absent from the end user certificate.
1800 Omitting parameters is no longer recommended. The test was also
1801 the wrong way round! This was probably due to unusual behaviour in
1802 EVP_cmp_parameters() which returns 1 if the parameters match.
1803 This meant that parameters were omitted when they *didn't* match and
1804 the certificate was useless. Certificates signed with 'ca' didn't have
1806 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>]
1808 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
1809 The interface is as follows:
1810 Applications can use
1811 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
1812 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
1813 "off" is now the default.
1814 The library internally uses
1815 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
1816 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
1817 to disable memory-checking temporarily.
1819 Some inconsistent states that previously were possible (and were
1820 even the default) are now avoided.
1822 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
1823 with each memory chunk allocated; this is occasionally more helpful
1824 than just having a counter.
1826 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
1828 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
1832 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX),
1833 which largely parallels "options", but is for changing API behaviour,
1834 whereas "options" are about protocol behaviour.
1835 Initial "mode" flags are:
1837 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
1838 a single record has been written.
1839 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
1840 retries use the same buffer location.
1841 (But all of the contents must be
1845 *) Bugfix: SSL_set_mode ignored its parameter, only SSL_CTX_set_mode
1848 *) Fix problems with no-hmac etc.
1849 [Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>]
1851 *) New functions RSA_get_default_method(), RSA_set_method() and
1852 RSA_get_method(). These allows replacement of RSA_METHODs without having
1853 to mess around with the internals of an RSA structure.
1856 *) Fix memory leaks in DSA_do_sign and DSA_is_prime.
1857 Also really enable memory leak checks in openssl.c and in some
1859 [Chad C. Mulligan, Bodo Moeller]
1861 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
1862 up the length of negative integers. This has now been simplified to just
1863 store the length when it is first determined and use it later, rather
1864 than trying to keep track of where data is copied and updating it to
1866 [Steve Henson, reported by Brien Wheeler
1867 <bwheeler@authentica-security.com>]
1869 *) Add a new function PKCS7_signatureVerify. This allows the verification
1870 of a PKCS#7 signature but with the signing certificate passed to the
1871 function itself. This contrasts with PKCS7_dataVerify which assumes the
1872 certificate is present in the PKCS#7 structure. This isn't always the
1873 case: certificates can be omitted from a PKCS#7 structure and be
1874 distributed by "out of band" means (such as a certificate database).
1877 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the
1878 function prototypes in pem.h, also change util/mkdef.pl to add the
1879 necessary function names.
1882 *) mk1mf.pl (used by Windows builds) did not properly read the
1883 options set by Configure in the top level Makefile, and Configure
1884 was not even able to write more than one option correctly.
1885 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
1888 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config
1889 file to be loaded from a BIO or FILE pointer. The BIO version will
1890 for example allow memory BIOs to contain config info.
1893 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
1894 Whoever hopes to achieve shared-library compatibility across versions
1895 must use this, not the compile-time macro.
1896 (Exercise 0.9.4: Which is the minimum library version required by
1898 Note: All this applies only to multi-threaded programs, others don't
1902 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests
1903 through a BIO pair triggered the default case, i.e.
1904 SSLerr(...,SSL_R_UNKNOWN_STATE).
1907 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
1908 can use the SSL library even if none of the specific BIOs is
1912 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
1913 for the encoded length.
1914 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>]
1916 *) Add initial documentation of the X509V3 functions.
1919 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and
1920 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
1921 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
1922 secure PKCS#8 private key format with a high iteration count.
1925 *) Fix determination of Perl interpreter: A perl or perl5
1926 _directory_ in $PATH was also accepted as the interpreter.
1927 [Ralf S. Engelschall]
1929 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking
1930 wrong with it but it was very old and did things like calling
1931 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
1935 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed
1936 to use the new extension code.
1939 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
1940 with macros. This should make it easier to change their form, add extra
1941 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
1945 *) Add to configuration table a new entry that can specify an alternative
1946 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
1947 according to Mark Crispin <MRC@Panda.COM>.
1951 *) DES CBC did not update the IV. Weird.
1954 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
1955 Changing the behaviour of the former might break existing programs --
1956 where IV updating is needed, des_ncbc_encrypt can be used.
1959 *) When bntest is run from "make test" it drives bc to check its
1960 calculations, as well as internally checking them. If an internal check
1961 fails, it needs to cause bc to give a non-zero result or make test carries
1962 on without noticing the failure. Fixed.
1965 *) DES library cleanups.
1968 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
1969 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
1970 ciphers. NOTE: although the key derivation function has been verified
1971 against some published test vectors it has not been extensively tested
1972 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
1976 *) Instead of "mkdir -p", which is not fully portable, use new
1977 Perl script "util/mkdir-p.pl".
1980 *) Rewrite the way password based encryption (PBE) is handled. It used to
1981 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
1982 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
1983 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
1984 the 'parameter' field of the AlgorithmIdentifier is passed to the
1985 underlying key generation function so it must do its own ASN1 parsing.
1986 This has also changed the EVP_PBE_CipherInit() function which now has a
1987 'parameter' argument instead of literal salt and iteration count values
1988 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
1991 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms
1992 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
1993 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
1994 KEY" because this clashed with PKCS#8 unencrypted string. Since this
1995 value was just used as a "magic string" and not used directly its
1996 value doesn't matter.
1999 *) Introduce some semblance of const correctness to BN. Shame C doesn't
2003 *) "linux-sparc64" configuration (ultrapenguin).
2004 [Ray Miller <ray.miller@oucs.ox.ac.uk>]
2005 "linux-sparc" configuration.
2006 [Christian Forster <fo@hawo.stw.uni-erlangen.de>]
2008 *) config now generates no-xxx options for missing ciphers.
2011 *) Support the EBCDIC character set (work in progress).
2012 File ebcdic.c not yet included because it has a different license.
2013 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
2015 *) Support BS2000/OSD-POSIX.
2016 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
2018 *) Make callbacks for key generation use void * instead of char *.
2021 *) Make S/MIME samples compile (not yet tested).
2024 *) Additional typesafe stacks.
2027 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
2031 Changes between 0.9.3 and 0.9.3a [29 May 1999]
2033 *) New configuration variant "sco5-gcc".
2035 *) Updated some demos.
2036 [Sean O Riordain, Wade Scholine]
2038 *) Add missing BIO_free at exit of pkcs12 application.
2041 *) Fix memory leak in conf.c.
2044 *) Updates for Win32 to assembler version of MD5.
2047 *) Set #! path to perl in apps/der_chop to where we found it
2048 instead of using a fixed path.
2051 *) SHA library changes for irix64-mips4-cc.
2054 *) Improvements for VMS support.
2058 Changes between 0.9.2b and 0.9.3 [24 May 1999]
2060 *) Bignum library bug fix. IRIX 6 passes "make test" now!
2061 This also avoids the problems with SC4.2 and unpatched SC5.
2062 [Andy Polyakov <appro@fy.chalmers.se>]
2064 *) New functions sk_num, sk_value and sk_set to replace the previous macros.
2065 These are required because of the typesafe stack would otherwise break
2066 existing code. If old code used a structure member which used to be STACK
2067 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
2068 sk_num or sk_value it would produce an error because the num, data members
2069 are not present in STACK_OF. Now it just produces a warning. sk_set
2070 replaces the old method of assigning a value to sk_value
2071 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
2072 that does this will no longer work (and should use sk_set instead) but
2073 this could be regarded as a "questionable" behaviour anyway.
2076 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now
2077 correctly handle encrypted S/MIME data.
2080 *) Change type of various DES function arguments from des_cblock
2081 (which means, in function argument declarations, pointer to char)
2082 to des_cblock * (meaning pointer to array with 8 char elements),
2083 which allows the compiler to do more typechecking; it was like
2084 that back in SSLeay, but with lots of ugly casts.
2086 Introduce new type const_des_cblock.
2089 *) Reorganise the PKCS#7 library and get rid of some of the more obvious
2090 problems: find RecipientInfo structure that matches recipient certificate
2091 and initialise the ASN1 structures properly based on passed cipher.
2094 *) Belatedly make the BN tests actually check the results.
2097 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
2098 to and from BNs: it was completely broken. New compilation option
2099 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
2100 key elements as negative integers.
2103 *) Reorganize and speed up MD5.
2104 [Andy Polyakov <appro@fy.chalmers.se>]
2107 [Richard Levitte <richard@levitte.org>]
2109 *) New option -out to asn1parse to allow the parsed structure to be
2110 output to a file. This is most useful when combined with the -strparse
2111 option to examine the output of things like OCTET STRINGS.
2114 *) Make SSL library a little more fool-proof by not requiring any longer
2115 that SSL_set_{accept,connect}_state be called before
2116 SSL_{accept,connect} may be used (SSL_set_..._state is omitted
2117 in many applications because usually everything *appeared* to work as
2118 intended anyway -- now it really works as intended).
2121 *) Move openssl.cnf out of lib/.
2124 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall
2125 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
2126 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+
2127 [Ralf S. Engelschall]
2129 *) Various fixes to the EVP and PKCS#7 code. It may now be able to
2130 handle PKCS#7 enveloped data properly.
2131 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve]
2133 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
2134 copying pointers. The cert_st handling is changed by this in
2135 various ways (and thus what used to be known as ctx->default_cert
2136 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert
2137 any longer when s->cert does not give us what we need).
2138 ssl_cert_instantiate becomes obsolete by this change.
2139 As soon as we've got the new code right (possibly it already is?),
2140 we have solved a couple of bugs of the earlier code where s->cert
2141 was used as if it could not have been shared with other SSL structures.
2143 Note that using the SSL API in certain dirty ways now will result
2144 in different behaviour than observed with earlier library versions:
2145 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx)
2146 does not influence s as it used to.
2148 In order to clean up things more thoroughly, inside SSL_SESSION
2149 we don't use CERT any longer, but a new structure SESS_CERT
2150 that holds per-session data (if available); currently, this is
2151 the peer's certificate chain and, for clients, the server's certificate
2152 and temporary key. CERT holds only those values that can have
2153 meaningful defaults in an SSL_CTX.
2156 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
2157 from the internal representation. Various PKCS#7 fixes: remove some
2158 evil casts and set the enc_dig_alg field properly based on the signing
2162 *) Allow PKCS#12 password to be set from the command line or the
2163 environment. Let 'ca' get its config file name from the environment
2164 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
2168 *) Allow certificate policies extension to use an IA5STRING for the
2169 organization field. This is contrary to the PKIX definition but
2170 VeriSign uses it and IE5 only recognises this form. Document 'x509'
2174 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
2175 without disallowing inline assembler and the like for non-pedantic builds.
2178 *) Support Borland C++ builder.
2179 [Janez Jere <jj@void.si>, modified by Ulf Möller]
2184 *) SHA-1 cleanups and performance enhancements.
2185 [Andy Polyakov <appro@fy.chalmers.se>]
2187 *) Sparc v8plus assembler for the bignum library.
2188 [Andy Polyakov <appro@fy.chalmers.se>]
2190 *) Accept any -xxx and +xxx compiler options in Configure.
2193 *) Update HPUX configuration.
2196 *) Add missing sk_<type>_unshift() function to safestack.h
2197 [Ralf S. Engelschall]
2199 *) New function SSL_CTX_use_certificate_chain_file that sets the
2200 "extra_cert"s in addition to the certificate. (This makes sense
2201 only for "PEM" format files, as chains as a whole are not
2205 *) Support verify_depth from the SSL API.
2206 x509_vfy.c had what can be considered an off-by-one-error:
2207 Its depth (which was not part of the external interface)
2208 was actually counting the number of certificates in a chain;
2209 now it really counts the depth.
2212 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
2213 instead of X509err, which often resulted in confusing error
2214 messages since the error codes are not globally unique
2215 (e.g. an alleged error in ssl3_accept when a certificate
2216 didn't match the private key).
2218 *) New function SSL_CTX_set_session_id_context that allows to set a default
2219 value (so that you don't need SSL_set_session_id_context for each
2220 connection using the SSL_CTX).
2223 *) OAEP decoding bug fix.
2226 *) Support INSTALL_PREFIX for package builders, as proposed by
2230 *) New Configure options "threads" and "no-threads". For systems
2231 where the proper compiler options are known (currently Solaris
2232 and Linux), "threads" is the default.
2235 *) New script util/mklink.pl as a faster substitute for util/mklink.sh.
2238 *) Install various scripts to $(OPENSSLDIR)/misc, not to
2239 $(INSTALLTOP)/bin -- they shouldn't clutter directories
2240 such as /usr/local/bin.
2243 *) "make linux-shared" to build shared libraries.
2244 [Niels Poppe <niels@netbox.org>]
2246 *) New Configure option no-<cipher> (rsa, idea, rc5, ...).
2249 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
2250 extension adding in x509 utility.
2253 *) Remove NOPROTO sections and error code comments.
2256 *) Partial rewrite of the DEF file generator to now parse the ANSI
2260 *) New Configure options --prefix=DIR and --openssldir=DIR.
2263 *) Complete rewrite of the error code script(s). It is all now handled
2264 by one script at the top level which handles error code gathering,
2265 header rewriting and C source file generation. It should be much better
2266 than the old method: it now uses a modified version of Ulf's parser to
2267 read the ANSI prototypes in all header files (thus the old K&R definitions
2268 aren't needed for error creation any more) and do a better job of
2269 translating function codes into names. The old 'ASN1 error code imbedded
2270 in a comment' is no longer necessary and it doesn't use .err files which
2271 have now been deleted. Also the error code call doesn't have to appear all
2272 on one line (which resulted in some large lines...).
2275 *) Change #include filenames from <foo.h> to <openssl/foo.h>.
2278 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return
2279 0 (which usually indicates a closed connection), but continue reading.
2282 *) Fix some race conditions.
2285 *) Add support for CRL distribution points extension. Add Certificate
2286 Policies and CRL distribution points documentation.
2289 *) Move the autogenerated header file parts to crypto/opensslconf.h.
2292 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
2293 8 of keying material. Merlin has also confirmed interop with this fix
2294 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
2295 [Merlin Hughes <merlin@baltimore.ie>]
2297 *) Fix lots of warnings.
2298 [Richard Levitte <levitte@stacken.kth.se>]
2300 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
2301 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
2302 [Richard Levitte <levitte@stacken.kth.se>]
2304 *) Fix problems with sizeof(long) == 8.
2305 [Andy Polyakov <appro@fy.chalmers.se>]
2307 *) Change functions to ANSI C.
2310 *) Fix typos in error codes.
2311 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller]
2313 *) Remove defunct assembler files from Configure.
2316 *) SPARC v8 assembler BIGNUM implementation.
2317 [Andy Polyakov <appro@fy.chalmers.se>]
2319 *) Support for Certificate Policies extension: both print and set.
2320 Various additions to support the r2i method this uses.
2323 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could
2324 return a const string when you are expecting an allocated buffer.
2327 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
2328 types DirectoryString and DisplayText.
2331 *) Add code to allow r2i extensions to access the configuration database,
2332 add an LHASH database driver and add several ctx helper functions.
2335 *) Fix an evil bug in bn_expand2() which caused various BN functions to
2336 fail when they extended the size of a BIGNUM.
2339 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to
2340 support typesafe stack.
2343 *) Fix typo in SSL_[gs]et_options().
2344 [Nils Frostberg <nils@medcom.se>]
2346 *) Delete various functions and files that belonged to the (now obsolete)
2347 old X509V3 handling code.
2350 *) New Configure option "rsaref".
2353 *) Don't auto-generate pem.h.
2356 *) Introduce type-safe ASN.1 SETs.
2359 *) Convert various additional casted stacks to type-safe STACK_OF() variants.
2360 [Ben Laurie, Ralf S. Engelschall, Steve Henson]
2362 *) Introduce type-safe STACKs. This will almost certainly break lots of code
2363 that links with OpenSSL (well at least cause lots of warnings), but fear
2364 not: the conversion is trivial, and it eliminates loads of evil casts. A
2365 few STACKed things have been converted already. Feel free to convert more.
2366 In the fullness of time, I'll do away with the STACK type altogether.
2369 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate
2370 specified in <certfile> by updating the entry in the index.txt file.
2371 This way one no longer has to edit the index.txt file manually for
2372 revoking a certificate. The -revoke option does the gory details now.
2373 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall]
2375 *) Fix `openssl crl -noout -text' combination where `-noout' killed the
2376 `-text' option at all and this way the `-noout -text' combination was
2377 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'.
2378 [Ralf S. Engelschall]
2380 *) Make sure a corresponding plain text error message exists for the
2381 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
2382 verify callback function determined that a certificate was revoked.
2383 [Ralf S. Engelschall]
2385 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for
2386 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
2387 all available cipers including rc5, which was forgotten until now.
2388 In order to let the testing shell script know which algorithms
2389 are available, a new (up to now undocumented) command
2390 "openssl list-cipher-commands" is used.
2393 *) Bugfix: s_client occasionally would sleep in select() when
2394 it should have checked SSL_pending() first.
2397 *) New functions DSA_do_sign and DSA_do_verify to provide access to
2398 the raw DSA values prior to ASN.1 encoding.
2401 *) Tweaks to Configure
2402 [Niels Poppe <niels@netbox.org>]
2404 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
2408 *) New variables $(RANLIB) and $(PERL) in the Makefiles.
2411 *) New config option to avoid instructions that are illegal on the 80386.
2412 The default code is faster, but requires at least a 486.
2415 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
2416 SSL2_SERVER_VERSION (not used at all) macros, which are now the
2417 same as SSL2_VERSION anyway.
2420 *) New "-showcerts" option for s_client.
2423 *) Still more PKCS#12 integration. Add pkcs12 application to openssl
2424 application. Various cleanups and fixes.
2427 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
2428 modify error routines to work internally. Add error codes and PBE init
2429 to library startup routines.
2432 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and
2433 packing functions to asn1 and evp. Changed function names and error
2434 codes along the way.
2437 *) PKCS12 integration: and so it begins... First of several patches to
2438 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
2439 objects to objects.h
2442 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1
2443 and display support for Thawte strong extranet extension.
2446 *) Add LinuxPPC support.
2447 [Jeff Dubrule <igor@pobox.org>]
2449 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
2450 bn_div_words in alpha.s.
2451 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie]
2453 *) Make sure the RSA OAEP test is skipped under -DRSAref because
2454 OAEP isn't supported when OpenSSL is built with RSAref.
2455 [Ulf Moeller <ulf@fitug.de>]
2457 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
2458 so they no longer are missing under -DNOPROTO.
2459 [Soren S. Jorvang <soren@t.dk>]
2462 Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
2464 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
2465 doesn't work when the session is reused. Coming soon!
2468 *) Fix a security hole, that allows sessions to be reused in the wrong
2469 context thus bypassing client cert protection! All software that uses
2470 client certs and session caches in multiple contexts NEEDS PATCHING to
2471 allow session reuse! A fuller solution is in the works.
2472 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)]
2474 *) Some more source tree cleanups (removed obsolete files
2475 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
2476 permission on "config" script to be executable) and a fix for the INSTALL
2478 [Ulf Moeller <ulf@fitug.de>]
2480 *) Remove some legacy and erroneous uses of malloc, free instead of
2482 [Lennart Bang <lob@netstream.se>, with minor changes by Steve]
2484 *) Make rsa_oaep_test return non-zero on error.
2485 [Ulf Moeller <ulf@fitug.de>]
2487 *) Add support for native Solaris shared libraries. Configure
2488 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
2489 if someone would make that last step automatic.
2490 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>]
2492 *) ctx_size was not built with the right compiler during "make links". Fixed.
2495 *) Change the meaning of 'ALL' in the cipher list. It now means "everything
2496 except NULL ciphers". This means the default cipher list will no longer
2497 enable NULL ciphers. They need to be specifically enabled e.g. with
2498 the string "DEFAULT:eNULL".
2501 *) Fix to RSA private encryption routines: if p < q then it would
2502 occasionally produce an invalid result. This will only happen with
2503 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
2506 *) Be less restrictive and allow also `perl util/perlpath.pl
2507 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin',
2508 because this way one can also use an interpreter named `perl5' (which is
2509 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
2510 installed as `perl').
2511 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2513 *) Let util/clean-depend.pl work also with older Perl 5.00x versions.
2514 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2516 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
2517 advapi32.lib to Win32 build and change the pem test comparision
2518 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
2519 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
2520 and crypto/des/ede_cbcm_enc.c.
2523 *) DES quad checksum was broken on big-endian architectures. Fixed.
2526 *) Comment out two functions in bio.h that aren't implemented. Fix up the
2527 Win32 test batch file so it (might) work again. The Win32 test batch file
2528 is horrible: I feel ill....
2531 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
2532 in e_os.h. Audit of header files to check ANSI and non ANSI
2533 sections: 10 functions were absent from non ANSI section and not exported
2534 from Windows DLLs. Fixed up libeay.num for new functions.
2537 *) Make `openssl version' output lines consistent.
2538 [Ralf S. Engelschall]
2540 *) Fix Win32 symbol export lists for BIO functions: Added
2541 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
2542 to ms/libeay{16,32}.def.
2543 [Ralf S. Engelschall]
2545 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
2546 fine under Unix and passes some trivial tests I've now added. But the
2547 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
2548 added to make sure no one expects that this stuff really works in the
2549 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
2550 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
2552 [Ralf S. Engelschall]
2554 *) Fix the generation of two part addresses in perl.
2555 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie]
2557 *) Add config entry for Linux on MIPS.
2558 [John Tobey <jtobey@channel1.com>]
2560 *) Make links whenever Configure is run, unless we are on Windoze.
2563 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf.
2564 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
2568 *) Add a useful kludge to allow package maintainers to specify compiler and
2569 other platforms details on the command line without having to patch the
2570 Configure script everytime: One now can use ``perl Configure
2571 <id>:<details>'', i.e. platform ids are allowed to have details appended
2572 to them (seperated by colons). This is treated as there would be a static
2573 pre-configured entry in Configure's %table under key <id> with value
2574 <details> and ``perl Configure <id>'' is called. So, when you want to
2575 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
2576 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"''
2577 now, which overrides the FreeBSD-elf entry on-the-fly.
2578 [Ralf S. Engelschall]
2580 *) Disable new TLS1 ciphersuites by default: they aren't official yet.
2583 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
2584 on the `perl Configure ...' command line. This way one can compile
2585 OpenSSL libraries with Position Independent Code (PIC) which is needed
2586 for linking it into DSOs.
2587 [Ralf S. Engelschall]
2589 *) Remarkably, export ciphers were totally broken and no-one had noticed!
2593 *) Cleaned up the LICENSE document: The official contact for any license
2594 questions now is the OpenSSL core team under openssl-core@openssl.org.
2595 And add a paragraph about the dual-license situation to make sure people
2596 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
2597 to the OpenSSL toolkit.
2598 [Ralf S. Engelschall]
2600 *) General source tree makefile cleanups: Made `making xxx in yyy...'
2601 display consistent in the source tree and replaced `/bin/rm' by `rm'.
2602 Additonally cleaned up the `make links' target: Remove unnecessary
2603 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
2604 to speed processing and no longer clutter the display with confusing
2605 stuff. Instead only the actually done links are displayed.
2606 [Ralf S. Engelschall]
2608 *) Permit null encryption ciphersuites, used for authentication only. It used
2609 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
2610 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
2614 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
2615 signed attributes when verifying signatures (this would break them),
2616 the detached data encoding was wrong and public keys obtained using
2617 X509_get_pubkey() weren't freed.
2620 *) Add text documentation for the BUFFER functions. Also added a work around
2621 to a Win95 console bug. This was triggered by the password read stuff: the
2622 last character typed gets carried over to the next fread(). If you were
2623 generating a new cert request using 'req' for example then the last
2624 character of the passphrase would be CR which would then enter the first
2628 *) Added the new `Includes OpenSSL Cryptography Software' button as
2629 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
2630 button and can be used by applications based on OpenSSL to show the
2631 relationship to the OpenSSL project.
2632 [Ralf S. Engelschall]
2634 *) Remove confusing variables in function signatures in files
2635 ssl/ssl_lib.c and ssl/ssl.h.
2636 [Lennart Bong <lob@kulthea.stacken.kth.se>]
2638 *) Don't install bss_file.c under PREFIX/include/
2639 [Lennart Bong <lob@kulthea.stacken.kth.se>]
2641 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle
2642 functions that return function pointers and has support for NT specific
2643 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
2644 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
2645 unsigned to signed types: this was killing the Win32 compile.
2648 *) Add new certificate file to stack functions,
2649 SSL_add_dir_cert_subjects_to_stack() and
2650 SSL_add_file_cert_subjects_to_stack(). These largely supplant
2651 SSL_load_client_CA_file(), and can be used to add multiple certs easily
2652 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
2653 This means that Apache-SSL and similar packages don't have to mess around
2654 to add as many CAs as they want to the preferred list.
2657 *) Experiment with doxygen documentation. Currently only partially applied to
2659 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with
2660 openssl.doxy as the configuration file.
2663 *) Get rid of remaining C++-style comments which strict C compilers hate.
2664 [Ralf S. Engelschall, pointed out by Carlos Amengual]
2666 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
2667 compiled in by default: it has problems with large keys.
2670 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
2671 DH private keys and/or callback functions which directly correspond to
2672 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
2673 is needed for applications which have to configure certificates on a
2674 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
2676 For the RSA certificate situation is makes no difference, but
2677 for the DSA certificate situation this fixes the "no shared cipher"
2678 problem where the OpenSSL cipher selection procedure failed because the
2679 temporary keys were not overtaken from the context and the API provided
2680 no way to reconfigure them.
2681 The new functions now let applications reconfigure the stuff and they
2682 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
2683 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
2684 non-public-API function ssl_cert_instantiate() is used as a helper
2685 function and also to reduce code redundancy inside ssl_rsa.c.
2686 [Ralf S. Engelschall]
2688 *) Move s_server -dcert and -dkey options out of the undocumented feature
2689 area because they are useful for the DSA situation and should be
2690 recognized by the users.
2691 [Ralf S. Engelschall]
2693 *) Fix the cipher decision scheme for export ciphers: the export bits are
2694 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
2695 SSL_EXP_MASK. So, the original variable has to be used instead of the
2696 already masked variable.
2697 [Richard Levitte <levitte@stacken.kth.se>]
2699 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c
2700 [Richard Levitte <levitte@stacken.kth.se>]
2702 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
2703 from `int' to `unsigned int' because it's a length and initialized by
2704 EVP_DigestFinal() which expects an `unsigned int *'.
2705 [Richard Levitte <levitte@stacken.kth.se>]
2707 *) Don't hard-code path to Perl interpreter on shebang line of Configure
2708 script. Instead use the usual Shell->Perl transition trick.
2709 [Ralf S. Engelschall]
2711 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates
2712 (in addition to RSA certificates) to match the behaviour of `openssl dsa
2713 -noout -modulus' as it's already the case for `openssl rsa -noout
2714 -modulus'. For RSA the -modulus is the real "modulus" while for DSA
2715 currently the public key is printed (a decision which was already done by
2716 `openssl dsa -modulus' in the past) which serves a similar purpose.
2717 Additionally the NO_RSA no longer completely removes the whole -modulus
2718 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
2720 [Ralf S. Engelschall]
2722 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
2723 BIO. See the source (crypto/evp/bio_ok.c) for more info.
2724 [Arne Ansper <arne@ats.cyber.ee>]
2726 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs
2727 to be added. Now both 'req' and 'ca' can use new objects defined in the
2731 *) Add cool BIO that does syslog (or event log on NT).
2732 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie]
2734 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
2735 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
2736 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
2737 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
2740 *) Add preliminary config info for new extension code.
2743 *) Make RSA_NO_PADDING really use no padding.
2744 [Ulf Moeller <ulf@fitug.de>]
2746 *) Generate errors when private/public key check is done.
2749 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
2750 for some CRL extensions and new objects added.
2753 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private
2754 key usage extension and fuller support for authority key id.
2757 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
2758 padding method for RSA, which is recommended for new applications in PKCS
2759 #1 v2.0 (RFC 2437, October 1998).
2760 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
2761 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
2762 against Bleichbacher's attack on RSA.
2763 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
2766 *) Updates to the new SSL compression code
2767 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2769 *) Fix so that the version number in the master secret, when passed
2770 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
2771 (because the server will not accept higher), that the version number
2772 is 0x03,0x01, not 0x03,0x00
2773 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2775 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
2776 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
2777 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
2780 *) Support for RAW extensions where an arbitrary extension can be
2781 created by including its DER encoding. See apps/openssl.cnf for
2785 *) Make sure latest Perl versions don't interpret some generated C array
2786 code as Perl array code in the crypto/err/err_genc.pl script.
2787 [Lars Weber <3weber@informatik.uni-hamburg.de>]
2789 *) Modify ms/do_ms.bat to not generate assembly language makefiles since
2790 not many people have the assembler. Various Win32 compilation fixes and
2791 update to the INSTALL.W32 file with (hopefully) more accurate Win32
2795 *) Modify configure script 'Configure' to automatically create crypto/date.h
2796 file under Win32 and also build pem.h from pem.org. New script
2797 util/mkfiles.pl to create the MINFO file on environments that can't do a
2798 'make files': perl util/mkfiles.pl >MINFO should work.
2801 *) Major rework of DES function declarations, in the pursuit of correctness
2802 and purity. As a result, many evil casts evaporated, and some weirdness,
2803 too. You may find this causes warnings in your code. Zapping your evil
2804 casts will probably fix them. Mostly.
2807 *) Fix for a typo in asn1.h. Bug fix to object creation script
2808 obj_dat.pl. It considered a zero in an object definition to mean
2809 "end of object": none of the objects in objects.h have any zeros
2810 so it wasn't spotted.
2811 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>]
2813 *) Add support for Triple DES Cipher Block Chaining with Output Feedback
2814 Masking (CBCM). In the absence of test vectors, the best I have been able
2815 to do is check that the decrypt undoes the encrypt, so far. Send me test
2816 vectors if you have them.
2819 *) Correct calculation of key length for export ciphers (too much space was
2820 allocated for null ciphers). This has not been tested!
2823 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage
2824 message is now correct (it understands "crypto" and "ssl" on its
2825 command line). There is also now an "update" option. This will update
2826 the util/ssleay.num and util/libeay.num files with any new functions.
2828 perl util/mkdef.pl crypto ssl update
2829 it will update them.
2832 *) Overhauled the Perl interface (perl/*):
2833 - ported BN stuff to OpenSSL's different BN library
2834 - made the perl/ source tree CVS-aware
2835 - renamed the package from SSLeay to OpenSSL (the files still contain
2836 their history because I've copied them in the repository)
2837 - removed obsolete files (the test scripts will be replaced
2838 by better Test::Harness variants in the future)
2839 [Ralf S. Engelschall]
2841 *) First cut for a very conservative source tree cleanup:
2842 1. merge various obsolete readme texts into doc/ssleay.txt
2843 where we collect the old documents and readme texts.
2844 2. remove the first part of files where I'm already sure that we no
2845 longer need them because of three reasons: either they are just temporary
2846 files which were left by Eric or they are preserved original files where
2847 I've verified that the diff is also available in the CVS via "cvs diff
2848 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
2849 the crypto/md/ stuff).
2850 [Ralf S. Engelschall]
2852 *) More extension code. Incomplete support for subject and issuer alt
2853 name, issuer and authority key id. Change the i2v function parameters
2854 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
2855 what that's for :-) Fix to ASN1 macro which messed up
2856 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
2859 *) Preliminary support for ENUMERATED type. This is largely copied from the
2863 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
2864 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2866 *) Make sure `make rehash' target really finds the `openssl' program.
2867 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2869 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
2870 like to hear about it if this slows down other processors.
2873 *) Add CygWin32 platform information to Configure script.
2874 [Alan Batie <batie@aahz.jf.intel.com>]
2876 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm'
2877 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>]
2879 *) New program nseq to manipulate netscape certificate sequences
2882 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
2886 *) Fixes to BN code. Previously the default was to define BN_RECURSION
2887 but the BN code had some problems that would cause failures when
2888 doing certificate verification and some other functions.
2889 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2891 *) Add ASN1 and PEM code to support netscape certificate sequences.
2894 *) Add ASN1 and PEM code to support netscape certificate sequences.
2897 *) Add several PKIX and private extended key usage OIDs.
2900 *) Modify the 'ca' program to handle the new extension code. Modify
2901 openssl.cnf for new extension format, add comments.
2904 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
2905 and add a sample to openssl.cnf so req -x509 now adds appropriate
2909 *) Continued X509 V3 changes. Add to other makefiles, integrate with the
2910 error code, add initial support to X509_print() and x509 application.
2913 *) Takes a deep breath and start addding X509 V3 extension support code. Add
2914 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
2915 stuff is currently isolated and isn't even compiled yet.
2918 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL
2919 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
2920 Removed the versions check from X509 routines when loading extensions:
2921 this allows certain broken certificates that don't set the version
2922 properly to be processed.
2925 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
2926 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
2927 can still be regenerated with "make depend".
2930 *) Spelling mistake in C version of CAST-128.
2931 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>]
2933 *) Changes to the error generation code. The perl script err-code.pl
2934 now reads in the old error codes and retains the old numbers, only
2935 adding new ones if necessary. It also only changes the .err files if new
2936 codes are added. The makefiles have been modified to only insert errors
2937 when needed (to avoid needlessly modifying header files). This is done
2938 by only inserting errors if the .err file is newer than the auto generated
2939 C file. To rebuild all the error codes from scratch (the old behaviour)
2940 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
2941 or delete all the .err files.
2944 *) CAST-128 was incorrectly implemented for short keys. The C version has
2945 been fixed, but is untested. The assembler versions are also fixed, but
2946 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
2947 to regenerate it if needed.
2948 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
2949 Hagino <itojun@kame.net>]
2951 *) File was opened incorrectly in randfile.c.
2952 [Ulf Möller <ulf@fitug.de>]
2954 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print
2955 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
2956 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
2957 al: it's just almost always a UTCTime. Note this patch adds new error
2958 codes so do a "make errors" if there are problems.
2961 *) Correct Linux 1 recognition in config.
2962 [Ulf Möller <ulf@fitug.de>]
2964 *) Remove pointless MD5 hash when using DSA keys in ca.
2965 [Anonymous <nobody@replay.com>]
2967 *) Generate an error if given an empty string as a cert directory. Also
2968 generate an error if handed NULL (previously returned 0 to indicate an
2969 error, but didn't set one).
2970 [Ben Laurie, reported by Anonymous <nobody@replay.com>]
2972 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
2975 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
2976 parameters. This was causing a warning which killed off the Win32 compile.
2979 *) Remove C++ style comments from crypto/bn/bn_local.h.
2980 [Neil Costigan <neil.costigan@celocom.com>]
2982 *) The function OBJ_txt2nid was broken. It was supposed to return a nid
2983 based on a text string, looking up short and long names and finally
2984 "dot" format. The "dot" format stuff didn't work. Added new function
2985 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
2986 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
2987 OID is not part of the table.
2990 *) Add prototypes to X509 lookup/verify methods, fixing a bug in
2991 X509_LOOKUP_by_alias().
2994 *) Sort openssl functions by name.
2997 *) Get the gendsa program working (hopefully) and add it to app list. Remove
2998 encryption from sample DSA keys (in case anyone is interested the password
3002 *) Make _all_ *_free functions accept a NULL pointer.
3003 [Frans Heymans <fheymans@isaserver.be>]
3005 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use
3007 [Anonymous <nobody@replay.com>]
3009 *) s_server should send the CAfile as acceptable CAs, not its own cert.
3010 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
3012 *) Don't blow it for numeric -newkey arguments to apps/req.
3013 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
3015 *) Temp key "for export" tests were wrong in s3_srvr.c.
3016 [Anonymous <nobody@replay.com>]
3018 *) Add prototype for temp key callback functions
3019 SSL_CTX_set_tmp_{rsa,dh}_callback().
3022 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
3023 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
3026 *) X509_name_add_entry() freed the wrong thing after an error.
3027 [Arne Ansper <arne@ats.cyber.ee>]
3029 *) rsa_eay.c would attempt to free a NULL context.
3030 [Arne Ansper <arne@ats.cyber.ee>]
3032 *) BIO_s_socket() had a broken should_retry() on Windoze.
3033 [Arne Ansper <arne@ats.cyber.ee>]
3035 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
3036 [Arne Ansper <arne@ats.cyber.ee>]
3038 *) Make sure the already existing X509_STORE->depth variable is initialized
3039 in X509_STORE_new(), but document the fact that this variable is still
3040 unused in the certificate verification process.
3041 [Ralf S. Engelschall]
3043 *) Fix the various library and apps files to free up pkeys obtained from
3044 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
3047 *) Fix reference counting in X509_PUBKEY_get(). This makes
3048 demos/maurice/example2.c work, amongst others, probably.
3049 [Steve Henson and Ben Laurie]
3051 *) First cut of a cleanup for apps/. First the `ssleay' program is now named
3052 `openssl' and second, the shortcut symlinks for the `openssl <command>'
3053 are no longer created. This way we have a single and consistent command
3054 line interface `openssl <command>', similar to `cvs <command>'.
3055 [Ralf S. Engelschall, Paul Sutton and Ben Laurie]
3057 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
3058 BIT STRING wrapper always have zero unused bits.
3061 *) Add CA.pl, perl version of CA.sh, add extended key usage OID.
3064 *) Make the top-level INSTALL documentation easier to understand.
3067 *) Makefiles updated to exit if an error occurs in a sub-directory
3068 make (including if user presses ^C) [Paul Sutton]
3070 *) Make Montgomery context stuff explicit in RSA data structure.
3073 *) Fix build order of pem and err to allow for generated pem.h.
3076 *) Fix renumbering bug in X509_NAME_delete_entry().
3079 *) Enhanced the err-ins.pl script so it makes the error library number
3080 global and can add a library name. This is needed for external ASN1 and
3081 other error libraries.
3084 *) Fixed sk_insert which never worked properly.
3087 *) Fix ASN1 macros so they can handle indefinite length construted
3088 EXPLICIT tags. Some non standard certificates use these: they can now
3092 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
3093 into a single doc/ssleay.txt bundle. This way the information is still
3094 preserved but no longer messes up this directory. Now it's new room for
3095 the new set of documenation files.
3096 [Ralf S. Engelschall]
3098 *) SETs were incorrectly DER encoded. This was a major pain, because they
3099 shared code with SEQUENCEs, which aren't coded the same. This means that
3100 almost everything to do with SETs or SEQUENCEs has either changed name or
3101 number of arguments.
3102 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>]
3104 *) Fix test data to work with the above.
3107 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
3108 was already fixed by Eric for 0.9.1 it seems.
3109 [Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>]
3111 *) Autodetect FreeBSD3.
3114 *) Fix various bugs in Configure. This affects the following platforms:
3118 unixware-2.0-pentium
3122 *) Eliminate generated files from CVS. Reorder tests to regenerate files
3123 before they are needed.
3126 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy).
3130 Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
3132 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
3133 changed SSLeay to OpenSSL in version strings.
3134 [Ralf S. Engelschall]
3136 *) Some fixups to the top-level documents.
3139 *) Fixed the nasty bug where rsaref.h was not found under compile-time
3140 because the symlink to include/ was missing.
3141 [Ralf S. Engelschall]
3143 *) Incorporated the popular no-RSA/DSA-only patches
3144 which allow to compile a RSA-free SSLeay.
3145 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall]
3147 *) Fixed nasty rehash problem under `make -f Makefile.ssl links'
3148 when "ssleay" is still not found.
3149 [Ralf S. Engelschall]
3151 *) Added more platforms to Configure: Cray T3E, HPUX 11,
3152 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>]
3154 *) Updated the README file.
3155 [Ralf S. Engelschall]
3157 *) Added various .cvsignore files in the CVS repository subdirs
3158 to make a "cvs update" really silent.
3159 [Ralf S. Engelschall]
3161 *) Recompiled the error-definition header files and added
3162 missing symbols to the Win32 linker tables.
3163 [Ralf S. Engelschall]
3165 *) Cleaned up the top-level documents;
3166 o new files: CHANGES and LICENSE
3167 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
3168 o merged COPYRIGHT into LICENSE
3169 o removed obsolete TODO file
3170 o renamed MICROSOFT to INSTALL.W32
3171 [Ralf S. Engelschall]
3173 *) Removed dummy files from the 0.9.1b source tree:
3174 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
3175 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
3176 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
3177 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
3178 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
3179 [Ralf S. Engelschall]
3181 *) Added various platform portability fixes.
3184 *) The Genesis of the OpenSSL rpject:
3185 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
3186 Young and Tim J. Hudson created while they were working for C2Net until
3188 [The OpenSSL Project]
3191 Changes between 0.9.0b and 0.9.1b [not released]
3193 *) Updated a few CA certificates under certs/
3196 *) Changed some BIGNUM api stuff.
3199 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
3200 DGUX x86, Linux Alpha, etc.
3203 *) New COMP library [crypto/comp/] for SSL Record Layer Compression:
3204 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
3208 *) Add -strparse option to asn1pars program which parses nested
3210 [Dr Stephen Henson <shenson@bigfoot.com>]
3212 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
3215 *) DSA fix for "ca" program.
3218 *) Added "-genkey" option to "dsaparam" program.
3221 *) Added RIPE MD160 (rmd160) message digest.
3224 *) Added -a (all) option to "ssleay version" command.
3227 *) Added PLATFORM define which is the id given to Configure.
3230 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
3233 *) Extended the ASN.1 parser routines.
3236 *) Extended BIO routines to support REUSEADDR, seek, tell, etc.
3239 *) Added a BN_CTX to the BN library.
3242 *) Fixed the weak key values in DES library
3245 *) Changed API in EVP library for cipher aliases.
3248 *) Added support for RC2/64bit cipher.
3251 *) Converted the lhash library to the crypto/mem.c functions.
3254 *) Added more recognized ASN.1 object ids.
3257 *) Added more RSA padding checks for SSL/TLS.
3260 *) Added BIO proxy/filter functionality.
3263 *) Added extra_certs to SSL_CTX which can be used
3264 send extra CA certificates to the client in the CA cert chain sending
3265 process. It can be configured with SSL_CTX_add_extra_chain_cert().
3268 *) Now Fortezza is denied in the authentication phase because
3269 this is key exchange mechanism is not supported by SSLeay at all.
3272 *) Additional PKCS1 checks.
3275 *) Support the string "TLSv1" for all TLS v1 ciphers.
3278 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
3279 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
3282 *) Fixed a few memory leaks.
3285 *) Fixed various code and comment typos.
3288 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0
3289 bytes sent in the client random.
3290 [Edward Bishop <ebishop@spyglass.com>]