5 Changes between 0.9.5a and 0.9.6 [xx XXX 2000]
7 *) New options added to the 'dgst' utility for signature
8 generation and verification.
11 *) Unrecognized PKCS#7 content types are now handled via a
12 catch all ASN1_TYPE structure. This allows unsupported
13 types to be stored as a "blob" and an application can
14 encode and decode it manually.
17 *) Fix various signed/unsigned issues to make a_strex.c
19 [Oscar Jacobsson <oscar.jacobsson@celocom.com>]
21 *) ASN1 fixes. i2d_ASN1_OBJECT was not returning the correct
22 length if passed a buffer. ASN1_INTEGER_to_BN failed
23 if passed a NULL BN and its argument was negative.
24 [Steve Henson, pointed out by Sven Heiberg <sven@tartu.cyber.ee>]
26 *) Modification to PKCS#7 encoding routines to output definite
27 length encoding. Since currently the whole structures are in
28 memory there's not real point in using indefinite length
29 constructed encoding. However if OpenSSL is compiled with
30 the flag PKCS7_INDEFINITE_ENCODING the old form is used.
33 *) Added BIO_vprintf() and BIO_vsnprintf().
36 *) Added more prefixes to parse for in the the strings written
37 through a logging bio, to cover all the levels that are available
38 through syslog. The prefixes are now:
40 PANIC, EMERG, EMR => LOG_EMERG
41 ALERT, ALR => LOG_ALERT
44 WARNING, WARN, WAR => LOG_WARNING
45 NOTICE, NOTE, NOT => LOG_NOTICE
47 DEBUG, DBG => LOG_DEBUG
49 and as before, if none of those prefixes are present at the
50 beginning of the string, LOG_ERR is chosen.
52 On Win32, the LOG_* levels are mapped according to this:
54 LOG_EMERG, LOG_ALERT, LOG_CRIT, LOG_ERR => EVENTLOG_ERROR_TYPE
55 LOG_WARNING => EVENTLOG_WARNING_TYPE
56 LOG_NOTICE, LOG_INFO, LOG_DEBUG => EVENTLOG_INFORMATION_TYPE
60 *) Made it possible to reconfigure with just the configuration
61 argument "reconf" or "reconfigure". The command line arguments
62 are stored in Makefile.ssl in the variable CONFIGURE_ARGS,
63 and are retrieved from there when reconfiguring.
67 [Assar Westerlund <assar@sics.se>, Richard Levitte]
69 *) Add the arguments -CAfile and -CApath to the pkcs12 utility.
72 *) The obj_dat.pl script was messing up the sorting of object
73 names. The reason was that it compared the quoted version
74 of strings as a result "OCSP" > "OCSP Signing" because
75 " > SPACE. Changed script to store unquoted versions of
76 names and add quotes on output. It was also omitting some
77 names from the lookup table if they were given a default
78 value (that is if SN is missing it is given the same
79 value as LN and vice versa), these are now added on the
80 grounds that if an object has a name we should be able to
81 look it up. Finally added warning output when duplicate
82 short or long names are found.
85 *) Changes needed for Tandem NSK.
86 [Scott Uroff scott@xypro.com]
88 *) Fix SSL 2.0 rollback checking: Due to an off-by-one error in
89 RSA_padding_check_SSLv23(), special padding was never detected
90 and thus the SSL 3.0/TLS 1.0 countermeasure against protocol
91 version rollback attacks was not effective.
93 In s23_clnt.c, don't use special rollback-attack detection padding
94 (RSA_SSLV23_PADDING) if SSL 2.0 is the only protocol enabled in the
95 client; similarly, in s23_srvr.c, don't do the rollback check if
96 SSL 2.0 is the only protocol enabled in the server.
99 *) Make it possible to get hexdumps of unprintable data with 'openssl
100 asn1parse'. By implication, the functions ASN1_parse_dump() and
101 BIO_dump_indent() are added.
104 *) New functions ASN1_STRING_print_ex() and X509_NAME_print_ex()
105 these print out strings and name structures based on various
106 flags including RFC2253 support and proper handling of
107 multibyte characters. Added options to the 'x509' utility
108 to allow the various flags to be set.
111 *) Various fixes to use ASN1_TIME instead of ASN1_UTCTIME.
112 Also change the functions X509_cmp_current_time() and
113 X509_gmtime_adj() work with an ASN1_TIME structure,
114 this will enable certificates using GeneralizedTime in validity
118 *) Make the NEG_PUBKEY_BUG code (which tolerates invalid
119 negative public key encodings) on by default,
120 NO_NEG_PUBKEY_BUG can be set to disable it.
123 *) New function c2i_ASN1_OBJECT() which acts on ASN1_OBJECT
124 content octets. An i2c_ASN1_OBJECT is unnecessary because
125 the encoding can be trivially obtained from the structure.
128 *) crypto/err.c locking bugfix: Use write locks (CRYPTO_w_[un]lock),
129 not read locks (CRYPTO_r_[un]lock).
132 *) A first attempt at creating official support for shared
133 libraries through configuration. I've kept it so the
134 default is static libraries only, and the OpenSSL programs
135 are always statically linked for now, but there are
136 preparations for dynamic linking in place.
137 This has been tested on Linux and True64.
140 *) Randomness polling function for Win9x, as described in:
141 Peter Gutmann, Software Generation of Practically Strong
145 *) Fix so PRNG is seeded in req if using an already existing
149 *) New options to smime application. -inform and -outform
150 allow alternative formats for the S/MIME message including
151 PEM and DER. The -content option allows the content to be
152 specified separately. This should allow things like Netscape
153 form signing output easier to verify.
156 *) Fix the ASN1 encoding of tags using the 'long form'.
159 *) New ASN1 functions, i2c_* and c2i_* for INTEGER and BIT
160 STRING types. These convert content octets to and from the
161 underlying type. The actual tag and length octets are
162 already assumed to have been read in and checked. These
163 are needed because all other string types have virtually
164 identical handling apart from the tag. By having versions
165 of the ASN1 functions that just operate on content octets
166 IMPLICIT tagging can be handled properly. It also allows
167 the ASN1_ENUMERATED code to be cut down because ASN1_ENUMERATED
168 and ASN1_INTEGER are identical apart from the tag.
171 *) Change the handling of OID objects as follows:
173 - New object identifiers are inserted in objects.txt, following
174 the syntax given in objects.README.
175 - objects.pl is used to process obj_mac.num and create a new
177 - obj_dat.pl is used to create a new obj_dat.h, using the data in
180 This is currently kind of a hack, and the perl code in objects.pl
181 isn't very elegant, but it works as I intended. The simplest way
182 to check that it worked correctly is to look in obj_dat.h and
183 check the array nid_objs and make sure the objects haven't moved
184 around (this is important!). Additions are OK, as well as
185 consistent name changes.
188 *) Add BSD-style MD5-based passwords to 'openssl passwd' (option '-1').
191 *) Addition of the command line parameter '-rand file' to 'openssl req'.
192 The given file adds to whatever has already been seeded into the
193 random pool through the RANDFILE configuration file option or
194 environment variable, or the default random state file.
197 *) mkstack.pl now sorts each macro group into lexical order.
198 Previously the output order depended on the order the files
199 appeared in the directory, resulting in needless rewriting
203 *) Patches to make OpenSSL compile under Win32 again. Mostly
204 work arounds for the VC++ problem that it treats func() as
205 func(void). Also stripped out the parts of mkdef.pl that
206 added extra typesafe functions: these no longer exist.
209 *) Reorganisation of the stack code. The macros are now all
210 collected in safestack.h . Each macro is defined in terms of
211 a "stack macro" of the form SKM_<name>(type, a, b). The
212 DEBUG_SAFESTACK is now handled in terms of function casts,
213 this has the advantage of retaining type safety without the
214 use of additional functions. If DEBUG_SAFESTACK is not defined
215 then the non typesafe macros are used instead. Also modified the
216 mkstack.pl script to handle the new form. Needs testing to see
217 if which (if any) compilers it chokes and maybe make DEBUG_SAFESTACK
218 the default if no major problems. Similar behaviour for ASN1_SET_OF
222 *) When some versions of IIS use the 'NET' form of private key the
223 key derivation algorithm is different. Normally MD5(password) is
224 used as a 128 bit RC4 key. In the modified case
225 MD5(MD5(password) + "SGCKEYSALT") is used insted. Added some
226 new functions i2d_RSA_NET(), d2i_RSA_NET() etc which are the same
227 as the old Netscape_RSA functions except they have an additional
228 'sgckey' parameter which uses the modified algorithm. Also added
229 an -sgckey command line option to the rsa utility. Thanks to
230 Adrian Peck <bertie@ncipher.com> for posting details of the modified
231 algorithm to openssl-dev.
234 *) The evp_local.h macros were using 'c.##kname' which resulted in
235 invalid expansion on some systems (SCO 5.0.5 for example).
236 Corrected to 'c.kname'.
237 [Phillip Porch <root@theporch.com>]
239 *) New X509_get1_email() and X509_REQ_get1_email() functions that return
240 a STACK of email addresses from a certificate or request, these look
241 in the subject name and the subject alternative name extensions and
242 omit any duplicate addresses.
245 *) Re-implement BN_mod_exp2_mont using independent (and larger) windows.
246 This makes DSA verification about 2 % faster.
249 *) Increase maximum window size in BN_mod_exp_... to 6 bits instead of 5
250 (meaning that now 2^5 values will be precomputed, which is only 4 KB
251 plus overhead for 1024 bit moduli).
252 This makes exponentiations about 0.5 % faster for 1024 bit
253 exponents (as measured by "openssl speed rsa2048").
256 *) Rename memory handling macros to avoid conflicts with other
258 Malloc => OPENSSL_malloc
259 Malloc_locked => OPENSSL_malloc_locked
260 Realloc => OPENSSL_realloc
264 *) New function BN_mod_exp_mont_word for small bases (roughly 15%
265 faster than BN_mod_exp_mont, i.e. 7% for a full DH exchange).
269 [John Jarvie <jjarvie@newsguy.com>]
271 *) The type-safe stack code has been rejigged. It is now only compiled
272 in when OpenSSL is configured with the DEBUG_SAFESTACK option and
273 by default all type-specific stack functions are "#define"d back to
274 standard stack functions. This results in more streamlined output
275 but retains the type-safety checking possibilities of the original
279 *) The STACK code has been cleaned up, and certain type declarations
280 that didn't make a lot of sense have been brought in line. This has
281 also involved a cleanup of sorts in safestack.h to more correctly
282 map type-safe stack functions onto their plain stack counterparts.
283 This work has also resulted in a variety of "const"ifications of
284 lots of the code, especially "_cmp" operations which should normally
285 be prototyped with "const" parameters anyway.
288 *) When generating bytes for the first time in md_rand.c, 'stir the pool'
289 by seeding with STATE_SIZE dummy bytes (with zero entropy count).
290 (The PRNG state consists of two parts, the large pool 'state' and 'md',
291 where all of 'md' is used each time the PRNG is used, but 'state'
292 is used only indexed by a cyclic counter. As entropy may not be
293 well distributed from the beginning, 'md' is important as a
294 chaining variable. However, the output function chains only half
295 of 'md', i.e. 80 bits. ssleay_rand_add, on the other hand, chains
296 all of 'md', and seeding with STATE_SIZE dummy bytes will result
297 in all of 'state' being rewritten, with the new values depending
298 on virtually all of 'md'. This overcomes the 80 bit limitation.)
301 *) In ssl/s2_clnt.c and ssl/s3_clnt.c, call ERR_clear_error() when
302 the handshake is continued after ssl_verify_cert_chain();
303 otherwise, if SSL_VERIFY_NONE is set, remaining error codes
304 can lead to 'unexplainable' connection aborts later.
305 [Bodo Moeller; problem tracked down by Lutz Jaenicke]
307 *) Major EVP API cipher revision.
308 Add hooks for extra EVP features. This allows various cipher
309 parameters to be set in the EVP interface. Support added for variable
310 key length ciphers via the EVP_CIPHER_CTX_set_key_length() function and
311 setting of RC2 and RC5 parameters.
313 Modify EVP_OpenInit() and EVP_SealInit() to cope with variable key length
316 Remove lots of duplicated code from the EVP library. For example *every*
317 cipher init() function handles the 'iv' in the same way according to the
318 cipher mode. They also all do nothing if the 'key' parameter is NULL and
319 for CFB and OFB modes they zero ctx->num.
321 New functionality allows removal of S/MIME code RC2 hack.
323 Most of the routines have the same form and so can be declared in terms
326 By shifting this to the top level EVP_CipherInit() it can be removed from
327 all individual ciphers. If the cipher wants to handle IVs or keys
328 differently it can set the EVP_CIPH_CUSTOM_IV or EVP_CIPH_ALWAYS_CALL_INIT
331 Change lots of functions like EVP_EncryptUpdate() to now return a
332 value: although software versions of the algorithms cannot fail
333 any installed hardware versions can.
336 *) Implement SSL_OP_TLS_ROLLBACK_BUG: In ssl3_get_client_key_exchange, if
337 this option is set, tolerate broken clients that send the negotiated
338 protocol version number instead of the requested protocol version
342 *) Call dh_tmp_cb (set by ..._TMP_DH_CB) with correct 'is_export' flag;
343 i.e. non-zero for export ciphersuites, zero otherwise.
344 Previous versions had this flag inverted, inconsistent with
345 rsa_tmp_cb (..._TMP_RSA_CB).
346 [Bodo Moeller; problem reported by Amit Chopra]
348 *) Add missing DSA library text string. Work around for some IIS
349 key files with invalid SEQUENCE encoding.
352 *) Add a document (doc/standards.txt) that list all kinds of standards
353 and so on that are implemented in OpenSSL.
356 *) Enhance c_rehash script. Old version would mishandle certificates
357 with the same subject name hash and wouldn't handle CRLs at all.
358 Added -fingerprint option to crl utility, to support new c_rehash
362 *) Eliminate non-ANSI declarations in crypto.h and stack.h.
365 *) Fix for SSL server purpose checking. Server checking was
366 rejecting certificates which had extended key usage present
367 but no ssl client purpose.
368 [Steve Henson, reported by Rene Grosser <grosser@hisolutions.com>]
370 *) Make PKCS#12 code work with no password. The PKCS#12 spec
371 is a little unclear about how a blank password is handled.
372 Since the password in encoded as a BMPString with terminating
373 double NULL a zero length password would end up as just the
374 double NULL. However no password at all is different and is
375 handled differently in the PKCS#12 key generation code. NS
376 treats a blank password as zero length. MSIE treats it as no
377 password on export: but it will try both on import. We now do
378 the same: PKCS12_parse() tries zero length and no password if
379 the password is set to "" or NULL (NULL is now a valid password:
380 it wasn't before) as does the pkcs12 application.
383 *) Bugfixes in apps/x509.c: Avoid a memory leak; and don't use
384 perror when PEM_read_bio_X509_REQ fails, the error message must
385 be obtained from the error queue.
388 *) Avoid 'thread_hash' memory leak in crypto/err/err.c by freeing
389 it in ERR_remove_state if appropriate, and change ERR_get_state
390 accordingly to avoid race conditions (this is necessary because
391 thread_hash is no longer constant once set).
394 *) Bugfix for linux-elf makefile.one.
397 *) RSA_get_default_method() will now cause a default
398 RSA_METHOD to be chosen if one doesn't exist already.
399 Previously this was only set during a call to RSA_new()
400 or RSA_new_method(NULL) meaning it was possible for
401 RSA_get_default_method() to return NULL.
404 *) Added native name translation to the existing DSO code
405 that will convert (if the flag to do so is set) filenames
406 that are sufficiently small and have no path information
407 into a canonical native form. Eg. "blah" converted to
408 "libblah.so" or "blah.dll" etc.
411 *) New function ERR_error_string_n(e, buf, len) which is like
412 ERR_error_string(e, buf), but writes at most 'len' bytes
413 including the 0 terminator. For ERR_error_string_n, 'buf'
415 [Damien Miller <djm@mindrot.org>, Bodo Moeller]
417 *) CONF library reworked to become more general. A new CONF
418 configuration file reader "class" is implemented as well as a
419 new functions (NCONF_*, for "New CONF") to handle it. The now
420 old CONF_* functions are still there, but are reimplemented to
421 work in terms of the new functions. Also, a set of functions
422 to handle the internal storage of the configuration data is
423 provided to make it easier to write new configuration file
424 reader "classes" (I can definitely see something reading a
425 configuration file in XML format, for example), called _CONF_*,
426 or "the configuration storage API"...
428 The new configuration file reading functions are:
430 NCONF_new, NCONF_free, NCONF_load, NCONF_load_fp, NCONF_load_bio,
431 NCONF_get_section, NCONF_get_string, NCONF_get_numbre
433 NCONF_default, NCONF_WIN32
435 NCONF_dump_fp, NCONF_dump_bio
437 NCONF_default and NCONF_WIN32 are method (or "class") choosers,
438 NCONF_new creates a new CONF object. This works in the same way
439 as other interfaces in OpenSSL, like the BIO interface.
440 NCONF_dump_* dump the internal storage of the configuration file,
441 which is useful for debugging. All other functions take the same
442 arguments as the old CONF_* functions wth the exception of the
443 first that must be a `CONF *' instead of a `LHASH *'.
445 To make it easer to use the new classes with the old CONF_* functions,
446 the function CONF_set_default_method is provided.
449 *) Add '-tls1' option to 'openssl ciphers', which was already
450 mentioned in the documentation but had not been implemented.
451 (This option is not yet really useful because even the additional
452 experimental TLS 1.0 ciphers are currently treated as SSL 3.0 ciphers.)
455 *) Initial DSO code added into libcrypto for letting OpenSSL (and
456 OpenSSL-based applications) load shared libraries and bind to
457 them in a portable way.
458 [Geoff Thorpe, with contributions from Richard Levitte]
460 Changes between 0.9.5 and 0.9.5a [1 Apr 2000]
462 *) Make sure _lrotl and _lrotr are only used with MSVC.
464 *) Use lock CRYPTO_LOCK_RAND correctly in ssleay_rand_status
465 (the default implementation of RAND_status).
467 *) Rename openssl x509 option '-crlext', which was added in 0.9.5,
468 to '-clrext' (= clear extensions), as intended and documented.
469 [Bodo Moeller; inconsistency pointed out by Michael Attili
472 *) Fix for HMAC. It wasn't zeroing the rest of the block if the key length
473 was larger than the MD block size.
474 [Steve Henson, pointed out by Yost William <YostW@tce.com>]
476 *) Modernise PKCS12_parse() so it uses STACK_OF(X509) for its ca argument
477 fix a leak when the ca argument was passed as NULL. Stop X509_PUBKEY_set()
478 using the passed key: if the passed key was a private key the result
479 of X509_print(), for example, would be to print out all the private key
483 *) des_quad_cksum() byte order bug fix.
484 [Ulf Möller, using the problem description in krb4-0.9.7, where
485 the solution is attributed to Derrick J Brashear <shadow@DEMENTIA.ORG>]
487 *) Fix so V_ASN1_APP_CHOOSE works again: however its use is strongly
489 [Steve Henson, pointed out by Brian Korver <briank@cs.stanford.edu>]
491 *) For easily testing in shell scripts whether some command
492 'openssl XXX' exists, the new pseudo-command 'openssl no-XXX'
493 returns with exit code 0 iff no command of the given name is available.
494 'no-XXX' is printed in this case, 'XXX' otherwise. In both cases,
495 the output goes to stdout and nothing is printed to stderr.
496 Additional arguments are always ignored.
498 Since for each cipher there is a command of the same name,
499 the 'no-cipher' compilation switches can be tested this way.
501 ('openssl no-XXX' is not able to detect pseudo-commands such
502 as 'quit', 'list-XXX-commands', or 'no-XXX' itself.)
505 *) Update test suite so that 'make test' succeeds in 'no-rsa' configuration.
508 *) For SSL_[CTX_]set_tmp_dh, don't create a DH key if SSL_OP_SINGLE_DH_USE
509 is set; it will be thrown away anyway because each handshake creates
511 ssl_cert_dup, which is used by SSL_new, now copies DH keys in addition
512 to parameters -- in previous versions (since OpenSSL 0.9.3) the
513 'default key' from SSL_CTX_set_tmp_dh would always be lost, meanining
514 you effectivly got SSL_OP_SINGLE_DH_USE when using this macro.
517 *) New s_client option -ign_eof: EOF at stdin is ignored, and
518 'Q' and 'R' lose their special meanings (quit/renegotiate).
519 This is part of what -quiet does; unlike -quiet, -ign_eof
520 does not suppress any output.
523 *) Add compatibility options to the purpose and trust code. The
524 purpose X509_PURPOSE_ANY is "any purpose" which automatically
525 accepts a certificate or CA, this was the previous behaviour,
526 with all the associated security issues.
528 X509_TRUST_COMPAT is the old trust behaviour: only and
529 automatically trust self signed roots in certificate store. A
530 new trust setting X509_TRUST_DEFAULT is used to specify that
531 a purpose has no associated trust setting and it should instead
532 use the value in the default purpose.
535 *) Fix the PKCS#8 DSA private key code so it decodes keys again
536 and fix a memory leak.
539 *) In util/mkerr.pl (which implements 'make errors'), preserve
540 reason strings from the previous version of the .c file, as
541 the default to have only downcase letters (and digits) in
542 automatically generated reasons codes is not always appropriate.
545 *) In ERR_load_ERR_strings(), build an ERR_LIB_SYS error reason table
546 using strerror. Previously, ERR_reason_error_string() returned
547 library names as reason strings for SYSerr; but SYSerr is a special
548 case where small numbers are errno values, not library numbers.
551 *) Add '-dsaparam' option to 'openssl dhparam' application. This
552 converts DSA parameters into DH parameters. (When creating parameters,
553 DSA_generate_parameters is used.)
556 *) Include 'length' (recommended exponent length) in C code generated
557 by 'openssl dhparam -C'.
560 *) The second argument to set_label in perlasm was already being used
561 so couldn't be used as a "file scope" flag. Moved to third argument
565 *) In PEM_ASN1_write_bio and some other functions, use RAND_pseudo_bytes
566 instead of RAND_bytes for encryption IVs and salts.
569 *) Include RAND_status() into RAND_METHOD instead of implementing
570 it only for md_rand.c Otherwise replacing the PRNG by calling
571 RAND_set_rand_method would be impossible.
574 *) Don't let DSA_generate_key() enter an infinite loop if the random
575 number generation fails.
578 *) New 'rand' application for creating pseudo-random output.
581 *) Added configuration support for Linux/IA64
582 [Rolf Haberrecker <rolf@suse.de>]
584 *) Assembler module support for Mingw32.
587 *) Shared library support for HPUX (in shlib/).
588 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE> and Anonymous]
590 *) Shared library support for Solaris gcc.
591 [Lutz Behnke <behnke@trustcenter.de>]
593 Changes between 0.9.4 and 0.9.5 [28 Feb 2000]
595 *) PKCS7_encrypt() was adding text MIME headers twice because they
596 were added manually and by SMIME_crlf_copy().
599 *) In bntest.c don't call BN_rand with zero bits argument.
600 [Steve Henson, pointed out by Andrew W. Gray <agray@iconsinc.com>]
602 *) BN_mul bugfix: In bn_mul_part_recursion() only the a>a[n] && b>b[n]
603 case was implemented. This caused BN_div_recp() to fail occasionally.
606 *) Add an optional second argument to the set_label() in the perl
607 assembly language builder. If this argument exists and is set
608 to 1 it signals that the assembler should use a symbol whose
609 scope is the entire file, not just the current function. This
610 is needed with MASM which uses the format label:: for this scope.
611 [Steve Henson, pointed out by Peter Runestig <peter@runestig.com>]
613 *) Change the ASN1 types so they are typedefs by default. Before
614 almost all types were #define'd to ASN1_STRING which was causing
615 STACK_OF() problems: you couldn't declare STACK_OF(ASN1_UTF8STRING)
619 *) Change names of new functions to the new get1/get0 naming
620 convention: After 'get1', the caller owns a reference count
621 and has to call ..._free; 'get0' returns a pointer to some
622 data structure without incrementing reference counters.
623 (Some of the existing 'get' functions increment a reference
624 counter, some don't.)
625 Similarly, 'set1' and 'add1' functions increase reference
626 counters or duplicate objects.
629 *) Allow for the possibility of temp RSA key generation failure:
630 the code used to assume it always worked and crashed on failure.
633 *) Fix potential buffer overrun problem in BIO_printf().
634 [Ulf Möller, using public domain code by Patrick Powell; problem
635 pointed out by David Sacerdote <das33@cornell.edu>]
637 *) Support EGD <http://www.lothar.com/tech/crypto/>. New functions
638 RAND_egd() and RAND_status(). In the command line application,
639 the EGD socket can be specified like a seed file using RANDFILE
643 *) Allow the string CERTIFICATE to be tolerated in PKCS#7 structures.
644 Some CAs (e.g. Verisign) distribute certificates in this form.
647 *) Remove the SSL_ALLOW_ADH compile option and set the default cipher
648 list to exclude them. This means that no special compilation option
649 is needed to use anonymous DH: it just needs to be included in the
653 *) Change the EVP_MD_CTX_type macro so its meaning consistent with
654 EVP_MD_type. The old functionality is available in a new macro called
655 EVP_MD_md(). Change code that uses it and update docs.
658 *) ..._ctrl functions now have corresponding ..._callback_ctrl functions
659 where the 'void *' argument is replaced by a function pointer argument.
660 Previously 'void *' was abused to point to functions, which works on
661 many platforms, but is not correct. As these functions are usually
662 called by macros defined in OpenSSL header files, most source code
663 should work without changes.
666 *) <openssl/opensslconf.h> (which is created by Configure) now contains
667 sections with information on -D... compiler switches used for
668 compiling the library so that applications can see them. To enable
669 one of these sections, a pre-processor symbol OPENSSL_..._DEFINES
670 must be defined. E.g.,
671 #define OPENSSL_ALGORITHM_DEFINES
672 #include <openssl/opensslconf.h>
673 defines all pertinent NO_<algo> symbols, such as NO_IDEA, NO_RSA, etc.
674 [Richard Levitte, Ulf and Bodo Möller]
676 *) Bugfix: Tolerate fragmentation and interleaving in the SSL 3/TLS
680 *) Change the 'other' type in certificate aux info to a STACK_OF
681 X509_ALGOR. Although not an AlgorithmIdentifier as such it has
682 the required ASN1 format: arbitrary types determined by an OID.
685 *) Add some PEM_write_X509_REQ_NEW() functions and a command line
686 argument to 'req'. This is not because the function is newer or
687 better than others it just uses the work 'NEW' in the certificate
688 request header lines. Some software needs this.
691 *) Reorganise password command line arguments: now passwords can be
692 obtained from various sources. Delete the PEM_cb function and make
693 it the default behaviour: i.e. if the callback is NULL and the
694 usrdata argument is not NULL interpret it as a null terminated pass
695 phrase. If usrdata and the callback are NULL then the pass phrase
696 is prompted for as usual.
699 *) Add support for the Compaq Atalla crypto accelerator. If it is installed,
700 the support is automatically enabled. The resulting binaries will
701 autodetect the card and use it if present.
702 [Ben Laurie and Compaq Inc.]
704 *) Work around for Netscape hang bug. This sends certificate request
705 and server done in one record. Since this is perfectly legal in the
706 SSL/TLS protocol it isn't a "bug" option and is on by default. See
707 the bugs/SSLv3 entry for more info.
710 *) HP-UX tune-up: new unified configs, HP C compiler bug workaround.
713 *) Add -rand argument to smime and pkcs12 applications and read/write
717 *) New 'passwd' tool for crypt(3) and apr1 password hashes.
720 *) Add command line password options to the remaining applications.
723 *) Bug fix for BN_div_recp() for numerators with an even number of
727 *) More tests in bntest.c, and changed test_bn output.
730 *) ./config recognizes MacOS X now.
733 *) Bug fix for BN_div() when the first words of num and divsor are
734 equal (it gave wrong results if (rem=(n1-q*d0)&BN_MASK2) < d0).
737 *) Add support for various broken PKCS#8 formats, and command line
738 options to produce them.
741 *) New functions BN_CTX_start(), BN_CTX_get() and BT_CTX_end() to
742 get temporary BIGNUMs from a BN_CTX.
745 *) Correct return values in BN_mod_exp_mont() and BN_mod_exp2_mont()
749 *) Change the SSLeay_add_all_*() functions to OpenSSL_add_all_*() and
750 include a #define from the old name to the new. The original intent
751 was that statically linked binaries could for example just call
752 SSLeay_add_all_ciphers() to just add ciphers to the table and not
753 link with digests. This never worked becayse SSLeay_add_all_digests()
754 and SSLeay_add_all_ciphers() were in the same source file so calling
755 one would link with the other. They are now in separate source files.
758 *) Add a new -notext option to 'ca' and a -pubkey option to 'spkac'.
761 *) Use a less unusual form of the Miller-Rabin primality test (it used
762 a binary algorithm for exponentiation integrated into the Miller-Rabin
763 loop, our standard modexp algorithms are faster).
766 *) Support for the EBCDIC character set completed.
767 [Martin Kraemer <Martin.Kraemer@Mch.SNI.De>]
769 *) Source code cleanups: use const where appropriate, eliminate casts,
770 use void * instead of char * in lhash.
773 *) Bugfix: ssl3_send_server_key_exchange was not restartable
774 (the state was not changed to SSL3_ST_SW_KEY_EXCH_B, and because of
775 this the server could overwrite ephemeral keys that the client
779 *) Turn DSA_is_prime into a macro that calls BN_is_prime,
780 using 50 iterations of the Rabin-Miller test.
782 DSA_generate_parameters now uses BN_is_prime_fasttest (with 50
783 iterations of the Rabin-Miller test as required by the appendix
784 to FIPS PUB 186[-1]) instead of DSA_is_prime.
785 As BN_is_prime_fasttest includes trial division, DSA parameter
786 generation becomes much faster.
788 This implies a change for the callback functions in DSA_is_prime
789 and DSA_generate_parameters: The callback function is called once
790 for each positive witness in the Rabin-Miller test, not just
791 occasionally in the inner loop; and the parameters to the
792 callback function now provide an iteration count for the outer
793 loop rather than for the current invocation of the inner loop.
794 DSA_generate_parameters additionally can call the callback
795 function with an 'iteration count' of -1, meaning that a
796 candidate has passed the trial division test (when q is generated
797 from an application-provided seed, trial division is skipped).
800 *) New function BN_is_prime_fasttest that optionally does trial
801 division before starting the Rabin-Miller test and has
802 an additional BN_CTX * argument (whereas BN_is_prime always
803 has to allocate at least one BN_CTX).
804 'callback(1, -1, cb_arg)' is called when a number has passed the
805 trial division stage.
808 *) Fix for bug in CRL encoding. The validity dates weren't being handled
812 *) New -pkcs12 option to CA.pl script to write out a PKCS#12 file.
815 *) New function BN_pseudo_rand().
818 *) Clean up BN_mod_mul_montgomery(): replace the broken (and unreadable)
819 bignum version of BN_from_montgomery() with the working code from
820 SSLeay 0.9.0 (the word based version is faster anyway), and clean up
824 *) Avoid a race condition in s2_clnt.c (function get_server_hello) that
825 made it impossible to use the same SSL_SESSION data structure in
826 SSL2 clients in multiple threads.
829 *) The return value of RAND_load_file() no longer counts bytes obtained
830 by stat(). RAND_load_file(..., -1) is new and uses the complete file
831 to seed the PRNG (previously an explicit byte count was required).
832 [Ulf Möller, Bodo Möller]
834 *) Clean up CRYPTO_EX_DATA functions, some of these didn't have prototypes
835 used (char *) instead of (void *) and had casts all over the place.
838 *) Make BN_generate_prime() return NULL on error if ret!=NULL.
841 *) Retain source code compatibility for BN_prime_checks macro:
842 BN_is_prime(..., BN_prime_checks, ...) now uses
843 BN_prime_checks_for_size to determine the appropriate number of
844 Rabin-Miller iterations.
847 *) Diffie-Hellman uses "safe" primes: DH_check() return code renamed to
848 DH_CHECK_P_NOT_SAFE_PRIME.
849 (Check if this is true? OpenPGP calls them "strong".)
852 *) Merge the functionality of "dh" and "gendh" programs into a new program
853 "dhparam". The old programs are retained for now but will handle DH keys
854 (instead of parameters) in future.
857 *) Make the ciphers, s_server and s_client programs check the return values
858 when a new cipher list is set.
861 *) Enhance the SSL/TLS cipher mechanism to correctly handle the TLS 56bit
862 ciphers. Before when the 56bit ciphers were enabled the sorting was
865 The syntax for the cipher sorting has been extended to support sorting by
866 cipher-strength (using the strength_bits hard coded in the tables).
867 The new command is "@STRENGTH" (see also doc/apps/ciphers.pod).
869 Fix a bug in the cipher-command parser: when supplying a cipher command
870 string with an "undefined" symbol (neither command nor alphanumeric
871 [A-Za-z0-9], ssl_set_cipher_list used to hang in an endless loop. Now
874 Due to the strength-sorting extension, the code of the
875 ssl_create_cipher_list() function was completely rearranged. I hope that
876 the readability was also increased :-)
877 [Lutz Jaenicke <Lutz.Jaenicke@aet.TU-Cottbus.DE>]
879 *) Minor change to 'x509' utility. The -CAcreateserial option now uses 1
880 for the first serial number and places 2 in the serial number file. This
881 avoids problems when the root CA is created with serial number zero and
882 the first user certificate has the same issuer name and serial number
886 *) Fixes to X509_ATTRIBUTE utilities, change the 'req' program so it uses
887 the new code. Add documentation for this stuff.
890 *) Changes to X509_ATTRIBUTE utilities. These have been renamed from
891 X509_*() to X509at_*() on the grounds that they don't handle X509
892 structures and behave in an analagous way to the X509v3 functions:
893 they shouldn't be called directly but wrapper functions should be used
896 So we also now have some wrapper functions that call the X509at functions
897 when passed certificate requests. (TO DO: similar things can be done with
898 PKCS#7 signed and unsigned attributes, PKCS#12 attributes and a few other
899 things. Some of these need some d2i or i2d and print functionality
900 because they handle more complex structures.)
903 *) Add missing #ifndefs that caused missing symbols when building libssl
904 as a shared library without RSA. Use #ifndef NO_SSL2 instead of
906 [Kris Kennaway <kris@hub.freebsd.org>, modified by Ulf Möller]
908 *) Precautions against using the PRNG uninitialized: RAND_bytes() now
909 has a return value which indicates the quality of the random data
910 (1 = ok, 0 = not seeded). Also an error is recorded on the thread's
911 error queue. New function RAND_pseudo_bytes() generates output that is
912 guaranteed to be unique but not unpredictable. RAND_add is like
913 RAND_seed, but takes an extra argument for an entropy estimate
914 (RAND_seed always assumes full entropy).
917 *) Do more iterations of Rabin-Miller probable prime test (specifically,
918 3 for 1024-bit primes, 6 for 512-bit primes, 12 for 256-bit primes
919 instead of only 2 for all lengths; see BN_prime_checks_for_size definition
920 in crypto/bn/bn_prime.c for the complete table). This guarantees a
921 false-positive rate of at most 2^-80 for random input.
924 *) Rewrite ssl3_read_n (ssl/s3_pkt.c) avoiding a couple of bugs.
927 *) New function X509_CTX_rget_chain() (renamed to X509_CTX_get1_chain
928 in the 0.9.5 release), this returns the chain
929 from an X509_CTX structure with a dup of the stack and all
930 the X509 reference counts upped: so the stack will exist
931 after X509_CTX_cleanup() has been called. Modify pkcs12.c
934 Also make SSL_SESSION_print() print out the verify return
938 *) Add manpage for the pkcs12 command. Also change the default
939 behaviour so MAC iteration counts are used unless the new
940 -nomaciter option is used. This improves file security and
941 only older versions of MSIE (4.0 for example) need it.
944 *) Honor the no-xxx Configure options when creating .DEF files.
947 *) Add PKCS#10 attributes to field table: challengePassword,
948 unstructuredName and unstructuredAddress. These are taken from
949 draft PKCS#9 v2.0 but are compatible with v1.2 provided no
950 international characters are used.
952 More changes to X509_ATTRIBUTE code: allow the setting of types
953 based on strings. Remove the 'loc' parameter when adding
954 attributes because these will be a SET OF encoding which is sorted
958 *) Initial changes to the 'req' utility to allow request generation
959 automation. This will allow an application to just generate a template
960 file containing all the field values and have req construct the
963 Initial support for X509_ATTRIBUTE handling. Stacks of these are
964 used all over the place including certificate requests and PKCS#7
965 structures. They are currently handled manually where necessary with
966 some primitive wrappers for PKCS#7. The new functions behave in a
967 manner analogous to the X509 extension functions: they allow
968 attributes to be looked up by NID and added.
970 Later something similar to the X509V3 code would be desirable to
971 automatically handle the encoding, decoding and printing of the
972 more complex types. The string types like challengePassword can
973 be handled by the string table functions.
975 Also modified the multi byte string table handling. Now there is
976 a 'global mask' which masks out certain types. The table itself
977 can use the flag STABLE_NO_MASK to ignore the mask setting: this
978 is useful when for example there is only one permissible type
979 (as in countryName) and using the mask might result in no valid
983 *) Clean up 'Finished' handling, and add functions SSL_get_finished and
984 SSL_get_peer_finished to allow applications to obtain the latest
985 Finished messages sent to the peer or expected from the peer,
986 respectively. (SSL_get_peer_finished is usually the Finished message
987 actually received from the peer, otherwise the protocol will be aborted.)
989 As the Finished message are message digests of the complete handshake
990 (with a total of 192 bits for TLS 1.0 and more for SSL 3.0), they can
991 be used for external authentication procedures when the authentication
992 provided by SSL/TLS is not desired or is not enough.
995 *) Enhanced support for Alpha Linux is added. Now ./config checks if
996 the host supports BWX extension and if Compaq C is present on the
997 $PATH. Just exploiting of the BWX extension results in 20-30%
998 performance kick for some algorithms, e.g. DES and RC4 to mention
999 a couple. Compaq C in turn generates ~20% faster code for MD5 and
1003 *) Add support for MS "fast SGC". This is arguably a violation of the
1004 SSL3/TLS protocol. Netscape SGC does two handshakes: the first with
1005 weak crypto and after checking the certificate is SGC a second one
1006 with strong crypto. MS SGC stops the first handshake after receiving
1007 the server certificate message and sends a second client hello. Since
1008 a server will typically do all the time consuming operations before
1009 expecting any further messages from the client (server key exchange
1010 is the most expensive) there is little difference between the two.
1012 To get OpenSSL to support MS SGC we have to permit a second client
1013 hello message after we have sent server done. In addition we have to
1014 reset the MAC if we do get this second client hello.
1017 *) Add a function 'd2i_AutoPrivateKey()' this will automatically decide
1018 if a DER encoded private key is RSA or DSA traditional format. Changed
1019 d2i_PrivateKey_bio() to use it. This is only needed for the "traditional"
1020 format DER encoded private key. Newer code should use PKCS#8 format which
1021 has the key type encoded in the ASN1 structure. Added DER private key
1022 support to pkcs8 application.
1025 *) SSL 3/TLS 1 servers now don't request certificates when an anonymous
1026 ciphersuites has been selected (as required by the SSL 3/TLS 1
1027 specifications). Exception: When SSL_VERIFY_FAIL_IF_NO_PEER_CERT
1028 is set, we interpret this as a request to violate the specification
1029 (the worst that can happen is a handshake failure, and 'correct'
1030 behaviour would result in a handshake failure anyway).
1033 *) In SSL_CTX_add_session, take into account that there might be multiple
1034 SSL_SESSION structures with the same session ID (e.g. when two threads
1035 concurrently obtain them from an external cache).
1036 The internal cache can handle only one SSL_SESSION with a given ID,
1037 so if there's a conflict, we now throw out the old one to achieve
1041 *) Add OIDs for idea and blowfish in CBC mode. This will allow both
1042 to be used in PKCS#5 v2.0 and S/MIME. Also add checking to
1043 some routines that use cipher OIDs: some ciphers do not have OIDs
1044 defined and so they cannot be used for S/MIME and PKCS#5 v2.0 for
1048 *) Simplify the trust setting structure and code. Now we just have
1049 two sequences of OIDs for trusted and rejected settings. These will
1050 typically have values the same as the extended key usage extension
1051 and any application specific purposes.
1053 The trust checking code now has a default behaviour: it will just
1054 check for an object with the same NID as the passed id. Functions can
1055 be provided to override either the default behaviour or the behaviour
1056 for a given id. SSL client, server and email already have functions
1057 in place for compatibility: they check the NID and also return "trusted"
1058 if the certificate is self signed.
1061 *) Add d2i,i2d bio/fp functions for PrivateKey: these convert the
1062 traditional format into an EVP_PKEY structure.
1065 *) Add a password callback function PEM_cb() which either prompts for
1066 a password if usr_data is NULL or otherwise assumes it is a null
1067 terminated password. Allow passwords to be passed on command line
1068 environment or config files in a few more utilities.
1071 *) Add a bunch of DER and PEM functions to handle PKCS#8 format private
1072 keys. Add some short names for PKCS#8 PBE algorithms and allow them
1073 to be specified on the command line for the pkcs8 and pkcs12 utilities.
1074 Update documentation.
1077 *) Support for ASN1 "NULL" type. This could be handled before by using
1078 ASN1_TYPE but there wasn't any function that would try to read a NULL
1079 and produce an error if it couldn't. For compatibility we also have
1080 ASN1_NULL_new() and ASN1_NULL_free() functions but these are faked and
1081 don't allocate anything because they don't need to.
1084 *) Initial support for MacOS is now provided. Examine INSTALL.MacOS
1086 [Andy Polyakov, Roy Woods <roy@centicsystems.ca>]
1088 *) Rebuild of the memory allocation routines used by OpenSSL code and
1089 possibly others as well. The purpose is to make an interface that
1090 provide hooks so anyone can build a separate set of allocation and
1091 deallocation routines to be used by OpenSSL, for example memory
1092 pool implementations, or something else, which was previously hard
1093 since Malloc(), Realloc() and Free() were defined as macros having
1094 the values malloc, realloc and free, respectively (except for Win32
1095 compilations). The same is provided for memory debugging code.
1096 OpenSSL already comes with functionality to find memory leaks, but
1097 this gives people a chance to debug other memory problems.
1099 With these changes, a new set of functions and macros have appeared:
1101 CRYPTO_set_mem_debug_functions() [F]
1102 CRYPTO_get_mem_debug_functions() [F]
1103 CRYPTO_dbg_set_options() [F]
1104 CRYPTO_dbg_get_options() [F]
1105 CRYPTO_malloc_debug_init() [M]
1107 The memory debug functions are NULL by default, unless the library
1108 is compiled with CRYPTO_MDEBUG or friends is defined. If someone
1109 wants to debug memory anyway, CRYPTO_malloc_debug_init() (which
1110 gives the standard debugging functions that come with OpenSSL) or
1111 CRYPTO_set_mem_debug_functions() (tells OpenSSL to use functions
1112 provided by the library user) must be used. When the standard
1113 debugging functions are used, CRYPTO_dbg_set_options can be used to
1114 request additional information:
1115 CRYPTO_dbg_set_options(V_CYRPTO_MDEBUG_xxx) corresponds to setting
1116 the CRYPTO_MDEBUG_xxx macro when compiling the library.
1118 Also, things like CRYPTO_set_mem_functions will always give the
1119 expected result (the new set of functions is used for allocation
1120 and deallocation) at all times, regardless of platform and compiler
1123 To finish it up, some functions that were never use in any other
1124 way than through macros have a new API and new semantic:
1127 CRYPTO_dbg_realloc()
1130 All macros of value have retained their old syntax.
1131 [Richard Levitte and Bodo Moeller]
1133 *) Some S/MIME fixes. The OID for SMIMECapabilities was wrong, the
1134 ordering of SMIMECapabilities wasn't in "strength order" and there
1135 was a missing NULL in the AlgorithmIdentifier for the SHA1 signature
1139 *) Some ASN1 types with illegal zero length encoding (INTEGER,
1140 ENUMERATED and OBJECT IDENTIFIER) choked the ASN1 routines.
1141 [Frans Heymans <fheymans@isaserver.be>, modified by Steve Henson]
1143 *) Merge in my S/MIME library for OpenSSL. This provides a simple
1144 S/MIME API on top of the PKCS#7 code, a MIME parser (with enough
1145 functionality to handle multipart/signed properly) and a utility
1146 called 'smime' to call all this stuff. This is based on code I
1147 originally wrote for Celo who have kindly allowed it to be
1148 included in OpenSSL.
1151 *) Add variants des_set_key_checked and des_set_key_unchecked of
1152 des_set_key (aka des_key_sched). Global variable des_check_key
1153 decides which of these is called by des_set_key; this way
1154 des_check_key behaves as it always did, but applications and
1155 the library itself, which was buggy for des_check_key == 1,
1156 have a cleaner way to pick the version they need.
1159 *) New function PKCS12_newpass() which changes the password of a
1163 *) Modify X509_TRUST and X509_PURPOSE so it also uses a static and
1164 dynamic mix. In both cases the ids can be used as an index into the
1165 table. Also modified the X509_TRUST_add() and X509_PURPOSE_add()
1166 functions so they accept a list of the field values and the
1167 application doesn't need to directly manipulate the X509_TRUST
1171 *) Modify the ASN1_STRING_TABLE stuff so it also uses bsearch and doesn't
1175 *) Modify the way the V3 extension code looks up extensions. This now
1176 works in a similar way to the object code: we have some "standard"
1177 extensions in a static table which is searched with OBJ_bsearch()
1178 and the application can add dynamic ones if needed. The file
1179 crypto/x509v3/ext_dat.h now has the info: this file needs to be
1180 updated whenever a new extension is added to the core code and kept
1181 in ext_nid order. There is a simple program 'tabtest.c' which checks
1182 this. New extensions are not added too often so this file can readily
1183 be maintained manually.
1185 There are two big advantages in doing things this way. The extensions
1186 can be looked up immediately and no longer need to be "added" using
1187 X509V3_add_standard_extensions(): this function now does nothing.
1188 [Side note: I get *lots* of email saying the extension code doesn't
1189 work because people forget to call this function]
1190 Also no dynamic allocation is done unless new extensions are added:
1191 so if we don't add custom extensions there is no need to call
1192 X509V3_EXT_cleanup().
1195 *) Modify enc utility's salting as follows: make salting the default. Add a
1196 magic header, so unsalted files fail gracefully instead of just decrypting
1197 to garbage. This is because not salting is a big security hole, so people
1198 should be discouraged from doing it.
1201 *) Fixes and enhancements to the 'x509' utility. It allowed a message
1202 digest to be passed on the command line but it only used this
1203 parameter when signing a certificate. Modified so all relevant
1204 operations are affected by the digest parameter including the
1205 -fingerprint and -x509toreq options. Also -x509toreq choked if a
1206 DSA key was used because it didn't fix the digest.
1209 *) Initial certificate chain verify code. Currently tests the untrusted
1210 certificates for consistency with the verify purpose (which is set
1211 when the X509_STORE_CTX structure is set up) and checks the pathlength.
1213 There is a NO_CHAIN_VERIFY compilation option to keep the old behaviour:
1214 this is because it will reject chains with invalid extensions whereas
1215 every previous version of OpenSSL and SSLeay made no checks at all.
1217 Trust code: checks the root CA for the relevant trust settings. Trust
1218 settings have an initial value consistent with the verify purpose: e.g.
1219 if the verify purpose is for SSL client use it expects the CA to be
1220 trusted for SSL client use. However the default value can be changed to
1221 permit custom trust settings: one example of this would be to only trust
1222 certificates from a specific "secure" set of CAs.
1224 Also added X509_STORE_CTX_new() and X509_STORE_CTX_free() functions
1225 which should be used for version portability: especially since the
1226 verify structure is likely to change more often now.
1228 SSL integration. Add purpose and trust to SSL_CTX and SSL and functions
1229 to set them. If not set then assume SSL clients will verify SSL servers
1232 Two new options to the verify program: -untrusted allows a set of
1233 untrusted certificates to be passed in and -purpose which sets the
1234 intended purpose of the certificate. If a purpose is set then the
1235 new chain verify code is used to check extension consistency.
1238 *) Support for the authority information access extension.
1241 *) Modify RSA and DSA PEM read routines to transparently handle
1242 PKCS#8 format private keys. New *_PUBKEY_* functions that handle
1243 public keys in a format compatible with certificate
1244 SubjectPublicKeyInfo structures. Unfortunately there were already
1245 functions called *_PublicKey_* which used various odd formats so
1246 these are retained for compatibility: however the DSA variants were
1247 never in a public release so they have been deleted. Changed dsa/rsa
1248 utilities to handle the new format: note no releases ever handled public
1249 keys so we should be OK.
1251 The primary motivation for this change is to avoid the same fiasco
1252 that dogs private keys: there are several incompatible private key
1253 formats some of which are standard and some OpenSSL specific and
1254 require various evil hacks to allow partial transparent handling and
1255 even then it doesn't work with DER formats. Given the option anything
1256 other than PKCS#8 should be dumped: but the other formats have to
1257 stay in the name of compatibility.
1259 With public keys and the benefit of hindsight one standard format
1260 is used which works with EVP_PKEY, RSA or DSA structures: though
1261 it clearly returns an error if you try to read the wrong kind of key.
1263 Added a -pubkey option to the 'x509' utility to output the public key.
1264 Also rename the EVP_PKEY_get_*() to EVP_PKEY_rget_*()
1265 (renamed to EVP_PKEY_get1_*() in the OpenSSL 0.9.5 release) and add
1266 EVP_PKEY_rset_*() functions (renamed to EVP_PKEY_set1_*())
1267 that do the same as the EVP_PKEY_assign_*() except they up the
1268 reference count of the added key (they don't "swallow" the
1272 *) Fixes to crypto/x509/by_file.c the code to read in certificates and
1273 CRLs would fail if the file contained no certificates or no CRLs:
1274 added a new function to read in both types and return the number
1275 read: this means that if none are read it will be an error. The
1276 DER versions of the certificate and CRL reader would always fail
1277 because it isn't possible to mix certificates and CRLs in DER format
1278 without choking one or the other routine. Changed this to just read
1279 a certificate: this is the best we can do. Also modified the code
1280 in apps/verify.c to take notice of return codes: it was previously
1281 attempting to read in certificates from NULL pointers and ignoring
1282 any errors: this is one reason why the cert and CRL reader seemed
1283 to work. It doesn't check return codes from the default certificate
1284 routines: these may well fail if the certificates aren't installed.
1287 *) Code to support otherName option in GeneralName.
1290 *) First update to verify code. Change the verify utility
1291 so it warns if it is passed a self signed certificate:
1292 for consistency with the normal behaviour. X509_verify
1293 has been modified to it will now verify a self signed
1294 certificate if *exactly* the same certificate appears
1295 in the store: it was previously impossible to trust a
1296 single self signed certificate. This means that:
1297 openssl verify ss.pem
1298 now gives a warning about a self signed certificate but
1299 openssl verify -CAfile ss.pem ss.pem
1303 *) For servers, store verify_result in SSL_SESSION data structure
1304 (and add it to external session representation).
1305 This is needed when client certificate verifications fails,
1306 but an application-provided verification callback (set by
1307 SSL_CTX_set_cert_verify_callback) allows accepting the session
1308 anyway (i.e. leaves x509_store_ctx->error != X509_V_OK
1309 but returns 1): When the session is reused, we have to set
1310 ssl->verify_result to the appropriate error code to avoid
1312 [Bodo Moeller, problem pointed out by Lutz Jaenicke]
1314 *) Fix a bug in the new PKCS#7 code: it didn't consider the
1315 case in PKCS7_dataInit() where the signed PKCS7 structure
1316 didn't contain any existing data because it was being created.
1317 [Po-Cheng Chen <pocheng@nst.com.tw>, slightly modified by Steve Henson]
1319 *) Add a salt to the key derivation routines in enc.c. This
1320 forms the first 8 bytes of the encrypted file. Also add a
1321 -S option to allow a salt to be input on the command line.
1324 *) New function X509_cmp(). Oddly enough there wasn't a function
1325 to compare two certificates. We do this by working out the SHA1
1326 hash and comparing that. X509_cmp() will be needed by the trust
1330 *) SSL_get1_session() is like SSL_get_session(), but increments
1331 the reference count in the SSL_SESSION returned.
1332 [Geoff Thorpe <geoff@eu.c2.net>]
1334 *) Fix for 'req': it was adding a null to request attributes.
1335 Also change the X509_LOOKUP and X509_INFO code to handle
1336 certificate auxiliary information.
1339 *) Add support for 40 and 64 bit RC2 and RC4 algorithms: document
1343 *) Add the possibility to add extra information to the memory leak
1344 detecting output, to form tracebacks, showing from where each
1345 allocation was originated: CRYPTO_push_info("constant string") adds
1346 the string plus current file name and line number to a per-thread
1347 stack, CRYPTO_pop_info() does the obvious, CRYPTO_remove_all_info()
1348 is like calling CYRPTO_pop_info() until the stack is empty.
1349 Also updated memory leak detection code to be multi-thread-safe.
1352 *) Add options -text and -noout to pkcs7 utility and delete the
1353 encryption options which never did anything. Update docs.
1356 *) Add options to some of the utilities to allow the pass phrase
1357 to be included on either the command line (not recommended on
1358 OSes like Unix) or read from the environment. Update the
1359 manpages and fix a few bugs.
1362 *) Add a few manpages for some of the openssl commands.
1365 *) Fix the -revoke option in ca. It was freeing up memory twice,
1366 leaking and not finding already revoked certificates.
1369 *) Extensive changes to support certificate auxiliary information.
1370 This involves the use of X509_CERT_AUX structure and X509_AUX
1371 functions. An X509_AUX function such as PEM_read_X509_AUX()
1372 can still read in a certificate file in the usual way but it
1373 will also read in any additional "auxiliary information". By
1374 doing things this way a fair degree of compatibility can be
1375 retained: existing certificates can have this information added
1376 using the new 'x509' options.
1378 Current auxiliary information includes an "alias" and some trust
1379 settings. The trust settings will ultimately be used in enhanced
1380 certificate chain verification routines: currently a certificate
1381 can only be trusted if it is self signed and then it is trusted
1385 *) Fix assembler for Alpha (tested only on DEC OSF not Linux or *BSD).
1386 The problem was that one of the replacement routines had not been working
1387 since SSLeay releases. For now the offending routine has been replaced
1388 with non-optimised assembler. Even so, this now gives around 95%
1389 performance improvement for 1024 bit RSA signs.
1392 *) Hack to fix PKCS#7 decryption when used with some unorthodox RC2
1393 handling. Most clients have the effective key size in bits equal to
1394 the key length in bits: so a 40 bit RC2 key uses a 40 bit (5 byte) key.
1395 A few however don't do this and instead use the size of the decrypted key
1396 to determine the RC2 key length and the AlgorithmIdentifier to determine
1397 the effective key length. In this case the effective key length can still
1398 be 40 bits but the key length can be 168 bits for example. This is fixed
1399 by manually forcing an RC2 key into the EVP_PKEY structure because the
1400 EVP code can't currently handle unusual RC2 key sizes: it always assumes
1401 the key length and effective key length are equal.
1404 *) Add a bunch of functions that should simplify the creation of
1405 X509_NAME structures. Now you should be able to do:
1406 X509_NAME_add_entry_by_txt(nm, "CN", MBSTRING_ASC, "Steve", -1, -1, 0);
1407 and have it automatically work out the correct field type and fill in
1408 the structures. The more adventurous can try:
1409 X509_NAME_add_entry_by_txt(nm, field, MBSTRING_UTF8, str, -1, -1, 0);
1410 and it will (hopefully) work out the correct multibyte encoding.
1413 *) Change the 'req' utility to use the new field handling and multibyte
1414 copy routines. Before the DN field creation was handled in an ad hoc
1415 way in req, ca, and x509 which was rather broken and didn't support
1416 BMPStrings or UTF8Strings. Since some software doesn't implement
1417 BMPStrings or UTF8Strings yet, they can be enabled using the config file
1418 using the dirstring_type option. See the new comment in the default
1419 openssl.cnf for more info.
1422 *) Make crypto/rand/md_rand.c more robust:
1423 - Assure unique random numbers after fork().
1424 - Make sure that concurrent threads access the global counter and
1425 md serializably so that we never lose entropy in them
1426 or use exactly the same state in multiple threads.
1427 Access to the large state is not always serializable because
1428 the additional locking could be a performance killer, and
1429 md should be large enough anyway.
1432 *) New file apps/app_rand.c with commonly needed functionality
1433 for handling the random seed file.
1435 Use the random seed file in some applications that previously did not:
1437 dsaparam -genkey (which also ignored its '-rand' option),
1440 x509 (when signing).
1441 Except on systems with /dev/urandom, it is crucial to have a random
1442 seed file at least for key creation, DSA signing, and for DH exchanges;
1443 for RSA signatures we could do without one.
1445 gendh and gendsa (unlike genrsa) used to read only the first byte
1446 of each file listed in the '-rand' option. The function as previously
1447 found in genrsa is now in app_rand.c and is used by all programs
1448 that support '-rand'.
1451 *) In RAND_write_file, use mode 0600 for creating files;
1452 don't just chmod when it may be too late.
1455 *) Report an error from X509_STORE_load_locations
1456 when X509_LOOKUP_load_file or X509_LOOKUP_add_dir failed.
1459 *) New function ASN1_mbstring_copy() this copies a string in either
1460 ASCII, Unicode, Universal (4 bytes per character) or UTF8 format
1461 into an ASN1_STRING type. A mask of permissible types is passed
1462 and it chooses the "minimal" type to use or an error if not type
1466 *) Add function equivalents to the various macros in asn1.h. The old
1467 macros are retained with an M_ prefix. Code inside the library can
1468 use the M_ macros. External code (including the openssl utility)
1469 should *NOT* in order to be "shared library friendly".
1472 *) Add various functions that can check a certificate's extensions
1473 to see if it usable for various purposes such as SSL client,
1474 server or S/MIME and CAs of these types. This is currently
1475 VERY EXPERIMENTAL but will ultimately be used for certificate chain
1476 verification. Also added a -purpose flag to x509 utility to
1477 print out all the purposes.
1480 *) Add a CRYPTO_EX_DATA to X509 certificate structure and associated
1484 *) New X509V3_{X509,CRL,REVOKED}_get_d2i() functions. These will search
1485 for, obtain and decode and extension and obtain its critical flag.
1486 This allows all the necessary extension code to be handled in a
1487 single function call.
1490 *) RC4 tune-up featuring 30-40% performance improvement on most RISC
1491 platforms. See crypto/rc4/rc4_enc.c for further details.
1494 *) New -noout option to asn1parse. This causes no output to be produced
1495 its main use is when combined with -strparse and -out to extract data
1496 from a file (which may not be in ASN.1 format).
1499 *) Fix for pkcs12 program. It was hashing an invalid certificate pointer
1500 when producing the local key id.
1501 [Richard Levitte <levitte@stacken.kth.se>]
1503 *) New option -dhparam in s_server. This allows a DH parameter file to be
1504 stated explicitly. If it is not stated then it tries the first server
1505 certificate file. The previous behaviour hard coded the filename
1509 *) Add -pubin and -pubout options to the rsa and dsa commands. These allow
1510 a public key to be input or output. For example:
1511 openssl rsa -in key.pem -pubout -out pubkey.pem
1512 Also added necessary DSA public key functions to handle this.
1515 *) Fix so PKCS7_dataVerify() doesn't crash if no certificates are contained
1516 in the message. This was handled by allowing
1517 X509_find_by_issuer_and_serial() to tolerate a NULL passed to it.
1518 [Steve Henson, reported by Sampo Kellomaki <sampo@mail.neuronio.pt>]
1520 *) Fix for bug in d2i_ASN1_bytes(): other ASN1 functions add an extra null
1521 to the end of the strings whereas this didn't. This would cause problems
1522 if strings read with d2i_ASN1_bytes() were later modified.
1523 [Steve Henson, reported by Arne Ansper <arne@ats.cyber.ee>]
1525 *) Fix for base64 decode bug. When a base64 bio reads only one line of
1526 data and it contains EOF it will end up returning an error. This is
1527 caused by input 46 bytes long. The cause is due to the way base64
1528 BIOs find the start of base64 encoded data. They do this by trying a
1529 trial decode on each line until they find one that works. When they
1530 do a flag is set and it starts again knowing it can pass all the
1531 data directly through the decoder. Unfortunately it doesn't reset
1532 the context it uses. This means that if EOF is reached an attempt
1533 is made to pass two EOFs through the context and this causes the
1534 resulting error. This can also cause other problems as well. As is
1535 usual with these problems it takes *ages* to find and the fix is
1536 trivial: move one line.
1537 [Steve Henson, reported by ian@uns.ns.ac.yu (Ivan Nejgebauer) ]
1539 *) Ugly workaround to get s_client and s_server working under Windows. The
1540 old code wouldn't work because it needed to select() on sockets and the
1541 tty (for keypresses and to see if data could be written). Win32 only
1542 supports select() on sockets so we select() with a 1s timeout on the
1543 sockets and then see if any characters are waiting to be read, if none
1544 are present then we retry, we also assume we can always write data to
1545 the tty. This isn't nice because the code then blocks until we've
1546 received a complete line of data and it is effectively polling the
1547 keyboard at 1s intervals: however it's quite a bit better than not
1548 working at all :-) A dedicated Windows application might handle this
1549 with an event loop for example.
1552 *) Enhance RSA_METHOD structure. Now there are two extra methods, rsa_sign
1553 and rsa_verify. When the RSA_FLAGS_SIGN_VER option is set these functions
1554 will be called when RSA_sign() and RSA_verify() are used. This is useful
1555 if rsa_pub_dec() and rsa_priv_enc() equivalents are not available.
1556 For this to work properly RSA_public_decrypt() and RSA_private_encrypt()
1557 should *not* be used: RSA_sign() and RSA_verify() must be used instead.
1558 This necessitated the support of an extra signature type NID_md5_sha1
1559 for SSL signatures and modifications to the SSL library to use it instead
1560 of calling RSA_public_decrypt() and RSA_private_encrypt().
1563 *) Add new -verify -CAfile and -CApath options to the crl program, these
1564 will lookup a CRL issuers certificate and verify the signature in a
1565 similar way to the verify program. Tidy up the crl program so it
1566 no longer accesses structures directly. Make the ASN1 CRL parsing a bit
1567 less strict. It will now permit CRL extensions even if it is not
1568 a V2 CRL: this will allow it to tolerate some broken CRLs.
1571 *) Initialize all non-automatic variables each time one of the openssl
1572 sub-programs is started (this is necessary as they may be started
1573 multiple times from the "OpenSSL>" prompt).
1574 [Lennart Bang, Bodo Moeller]
1576 *) Preliminary compilation option RSA_NULL which disables RSA crypto without
1577 removing all other RSA functionality (this is what NO_RSA does). This
1578 is so (for example) those in the US can disable those operations covered
1579 by the RSA patent while allowing storage and parsing of RSA keys and RSA
1583 *) Non-copying interface to BIO pairs.
1584 (still largely untested)
1587 *) New function ANS1_tag2str() to convert an ASN1 tag to a descriptive
1588 ASCII string. This was handled independently in various places before.
1591 *) New functions UTF8_getc() and UTF8_putc() that parse and generate
1592 UTF8 strings a character at a time.
1595 *) Use client_version from client hello to select the protocol
1596 (s23_srvr.c) and for RSA client key exchange verification
1597 (s3_srvr.c), as required by the SSL 3.0/TLS 1.0 specifications.
1600 *) Add various utility functions to handle SPKACs, these were previously
1601 handled by poking round in the structure internals. Added new function
1602 NETSCAPE_SPKI_print() to print out SPKAC and a new utility 'spkac' to
1603 print, verify and generate SPKACs. Based on an original idea from
1604 Massimiliano Pala <madwolf@comune.modena.it> but extensively modified.
1607 *) RIPEMD160 is operational on all platforms and is back in 'make test'.
1610 *) Allow the config file extension section to be overwritten on the
1611 command line. Based on an original idea from Massimiliano Pala
1612 <madwolf@comune.modena.it>. The new option is called -extensions
1613 and can be applied to ca, req and x509. Also -reqexts to override
1614 the request extensions in req and -crlexts to override the crl extensions
1618 *) Add new feature to the SPKAC handling in ca. Now you can include
1619 the same field multiple times by preceding it by "XXXX." for example:
1622 this is the same syntax as used in the req config file.
1625 *) Allow certificate extensions to be added to certificate requests. These
1626 are specified in a 'req_extensions' option of the req section of the
1627 config file. They can be printed out with the -text option to req but
1628 are otherwise ignored at present.
1631 *) Fix a horrible bug in enc_read() in crypto/evp/bio_enc.c: if the first
1632 data read consists of only the final block it would not decrypted because
1633 EVP_CipherUpdate() would correctly report zero bytes had been decrypted.
1634 A misplaced 'break' also meant the decrypted final block might not be
1635 copied until the next read.
1638 *) Initial support for DH_METHOD. Again based on RSA_METHOD. Also added
1639 a few extra parameters to the DH structure: these will be useful if
1640 for example we want the value of 'q' or implement X9.42 DH.
1643 *) Initial support for DSA_METHOD. This is based on the RSA_METHOD and
1644 provides hooks that allow the default DSA functions or functions on a
1645 "per key" basis to be replaced. This allows hardware acceleration and
1646 hardware key storage to be handled without major modification to the
1647 library. Also added low level modexp hooks and CRYPTO_EX structure and
1648 associated functions.
1651 *) Add a new flag to memory BIOs, BIO_FLAG_MEM_RDONLY. This marks the BIO
1652 as "read only": it can't be written to and the buffer it points to will
1653 not be freed. Reading from a read only BIO is much more efficient than
1654 a normal memory BIO. This was added because there are several times when
1655 an area of memory needs to be read from a BIO. The previous method was
1656 to create a memory BIO and write the data to it, this results in two
1657 copies of the data and an O(n^2) reading algorithm. There is a new
1658 function BIO_new_mem_buf() which creates a read only memory BIO from
1659 an area of memory. Also modified the PKCS#7 routines to use read only
1663 *) Bugfix: ssl23_get_client_hello did not work properly when called in
1664 state SSL23_ST_SR_CLNT_HELLO_B, i.e. when the first 7 bytes of
1665 a SSLv2-compatible client hello for SSLv3 or TLSv1 could be read,
1666 but a retry condition occured while trying to read the rest.
1669 *) The PKCS7_ENC_CONTENT_new() function was setting the content type as
1670 NID_pkcs7_encrypted by default: this was wrong since this should almost
1671 always be NID_pkcs7_data. Also modified the PKCS7_set_type() to handle
1672 the encrypted data type: this is a more sensible place to put it and it
1673 allows the PKCS#12 code to be tidied up that duplicated this
1677 *) Changed obj_dat.pl script so it takes its input and output files on
1678 the command line. This should avoid shell escape redirection problems
1682 *) Initial support for certificate extension requests, these are included
1683 in things like Xenroll certificate requests. Included functions to allow
1684 extensions to be obtained and added.
1687 *) -crlf option to s_client and s_server for sending newlines as
1688 CRLF (as required by many protocols).
1691 Changes between 0.9.3a and 0.9.4 [09 Aug 1999]
1693 *) Install libRSAglue.a when OpenSSL is built with RSAref.
1694 [Ralf S. Engelschall]
1696 *) A few more ``#ifndef NO_FP_API / #endif'' pairs for consistency.
1697 [Andrija Antonijevic <TheAntony2@bigfoot.com>]
1699 *) Fix -startdate and -enddate (which was missing) arguments to 'ca'
1703 *) New function DSA_dup_DH, which duplicates DSA parameters/keys as
1704 DH parameters/keys (q is lost during that conversion, but the resulting
1705 DH parameters contain its length).
1707 For 1024-bit p, DSA_generate_parameters followed by DSA_dup_DH is
1708 much faster than DH_generate_parameters (which creates parameters
1709 where p = 2*q + 1), and also the smaller q makes DH computations
1710 much more efficient (160-bit exponentiation instead of 1024-bit
1711 exponentiation); so this provides a convenient way to support DHE
1712 ciphersuites in SSL/TLS servers (see ssl/ssltest.c). It is of
1713 utter importance to use
1714 SSL_CTX_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
1716 SSL_set_options(s_ctx, SSL_OP_SINGLE_DH_USE);
1717 when such DH parameters are used, because otherwise small subgroup
1718 attacks may become possible!
1721 *) Avoid memory leak in i2d_DHparams.
1724 *) Allow the -k option to be used more than once in the enc program:
1725 this allows the same encrypted message to be read by multiple recipients.
1728 *) New function OBJ_obj2txt(buf, buf_len, a, no_name), this converts
1729 an ASN1_OBJECT to a text string. If the "no_name" parameter is set then
1730 it will always use the numerical form of the OID, even if it has a short
1734 *) Added an extra RSA flag: RSA_FLAG_EXT_PKEY. Previously the rsa_mod_exp
1735 method only got called if p,q,dmp1,dmq1,iqmp components were present,
1736 otherwise bn_mod_exp was called. In the case of hardware keys for example
1737 no private key components need be present and it might store extra data
1738 in the RSA structure, which cannot be accessed from bn_mod_exp.
1739 By setting RSA_FLAG_EXT_PKEY rsa_mod_exp will always be called for
1740 private key operations.
1743 *) Added support for SPARC Linux.
1746 *) pem_password_cb function type incompatibly changed from
1747 typedef int pem_password_cb(char *buf, int size, int rwflag);
1749 ....(char *buf, int size, int rwflag, void *userdata);
1750 so that applications can pass data to their callbacks:
1751 The PEM[_ASN1]_{read,write}... functions and macros now take an
1752 additional void * argument, which is just handed through whenever
1753 the password callback is called.
1754 [Damien Miller <dmiller@ilogic.com.au>; tiny changes by Bodo Moeller]
1756 New function SSL_CTX_set_default_passwd_cb_userdata.
1758 Compatibility note: As many C implementations push function arguments
1759 onto the stack in reverse order, the new library version is likely to
1760 interoperate with programs that have been compiled with the old
1761 pem_password_cb definition (PEM_whatever takes some data that
1762 happens to be on the stack as its last argument, and the callback
1763 just ignores this garbage); but there is no guarantee whatsoever that
1766 *) The -DPLATFORM="\"$(PLATFORM)\"" definition and the similar -DCFLAGS=...
1767 (both in crypto/Makefile.ssl for use by crypto/cversion.c) caused
1768 problems not only on Windows, but also on some Unix platforms.
1769 To avoid problematic command lines, these definitions are now in an
1770 auto-generated file crypto/buildinf.h (created by crypto/Makefile.ssl
1771 for standard "make" builds, by util/mk1mf.pl for "mk1mf" builds).
1774 *) MIPS III/IV assembler module is reimplemented.
1777 *) More DES library cleanups: remove references to srand/rand and
1778 delete an unused file.
1781 *) Add support for the the free Netwide assembler (NASM) under Win32,
1782 since not many people have MASM (ml) and it can be hard to obtain.
1783 This is currently experimental but it seems to work OK and pass all
1784 the tests. Check out INSTALL.W32 for info.
1787 *) Fix memory leaks in s3_clnt.c: All non-anonymous SSL3/TLS1 connections
1788 without temporary keys kept an extra copy of the server key,
1789 and connections with temporary keys did not free everything in case
1793 *) New function RSA_check_key and new openssl rsa option -check
1794 for verifying the consistency of RSA keys.
1795 [Ulf Moeller, Bodo Moeller]
1797 *) Various changes to make Win32 compile work:
1798 1. Casts to avoid "loss of data" warnings in p5_crpt2.c
1799 2. Change unsigned int to int in b_dump.c to avoid "signed/unsigned
1800 comparison" warnings.
1801 3. Add sk_<TYPE>_sort to DEF file generator and do make update.
1804 *) Add a debugging option to PKCS#5 v2 key generation function: when
1805 you #define DEBUG_PKCS5V2 passwords, salts, iteration counts and
1806 derived keys are printed to stderr.
1809 *) Copy the flags in ASN1_STRING_dup().
1810 [Roman E. Pavlov <pre@mo.msk.ru>]
1812 *) The x509 application mishandled signing requests containing DSA
1813 keys when the signing key was also DSA and the parameters didn't match.
1815 It was supposed to omit the parameters when they matched the signing key:
1816 the verifying software was then supposed to automatically use the CA's
1817 parameters if they were absent from the end user certificate.
1819 Omitting parameters is no longer recommended. The test was also
1820 the wrong way round! This was probably due to unusual behaviour in
1821 EVP_cmp_parameters() which returns 1 if the parameters match.
1822 This meant that parameters were omitted when they *didn't* match and
1823 the certificate was useless. Certificates signed with 'ca' didn't have
1825 [Steve Henson, reported by Doug Erickson <Doug.Erickson@Part.NET>]
1827 *) Memory leak checking (-DCRYPTO_MDEBUG) had some problems.
1828 The interface is as follows:
1829 Applications can use
1830 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ON) aka MemCheck_start(),
1831 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_OFF) aka MemCheck_stop();
1832 "off" is now the default.
1833 The library internally uses
1834 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_DISABLE) aka MemCheck_off(),
1835 CRYPTO_mem_ctrl(CRYPTO_MEM_CHECK_ENABLE) aka MemCheck_on()
1836 to disable memory-checking temporarily.
1838 Some inconsistent states that previously were possible (and were
1839 even the default) are now avoided.
1841 -DCRYPTO_MDEBUG_TIME is new and additionally stores the current time
1842 with each memory chunk allocated; this is occasionally more helpful
1843 than just having a counter.
1845 -DCRYPTO_MDEBUG_THREAD is also new and adds the thread ID.
1847 -DCRYPTO_MDEBUG_ALL enables all of the above, plus any future
1851 *) Introduce "mode" for SSL structures (with defaults in SSL_CTX),
1852 which largely parallels "options", but is for changing API behaviour,
1853 whereas "options" are about protocol behaviour.
1854 Initial "mode" flags are:
1856 SSL_MODE_ENABLE_PARTIAL_WRITE Allow SSL_write to report success when
1857 a single record has been written.
1858 SSL_MODE_ACCEPT_MOVING_WRITE_BUFFER Don't insist that SSL_write
1859 retries use the same buffer location.
1860 (But all of the contents must be
1864 *) Bugfix: SSL_set_mode ignored its parameter, only SSL_CTX_set_mode
1867 *) Fix problems with no-hmac etc.
1868 [Ulf Möller, pointed out by Brian Wellington <bwelling@tislabs.com>]
1870 *) New functions RSA_get_default_method(), RSA_set_method() and
1871 RSA_get_method(). These allows replacement of RSA_METHODs without having
1872 to mess around with the internals of an RSA structure.
1875 *) Fix memory leaks in DSA_do_sign and DSA_is_prime.
1876 Also really enable memory leak checks in openssl.c and in some
1878 [Chad C. Mulligan, Bodo Moeller]
1880 *) Fix a bug in d2i_ASN1_INTEGER() and i2d_ASN1_INTEGER() which can mess
1881 up the length of negative integers. This has now been simplified to just
1882 store the length when it is first determined and use it later, rather
1883 than trying to keep track of where data is copied and updating it to
1885 [Steve Henson, reported by Brien Wheeler
1886 <bwheeler@authentica-security.com>]
1888 *) Add a new function PKCS7_signatureVerify. This allows the verification
1889 of a PKCS#7 signature but with the signing certificate passed to the
1890 function itself. This contrasts with PKCS7_dataVerify which assumes the
1891 certificate is present in the PKCS#7 structure. This isn't always the
1892 case: certificates can be omitted from a PKCS#7 structure and be
1893 distributed by "out of band" means (such as a certificate database).
1896 *) Complete the PEM_* macros with DECLARE_PEM versions to replace the
1897 function prototypes in pem.h, also change util/mkdef.pl to add the
1898 necessary function names.
1901 *) mk1mf.pl (used by Windows builds) did not properly read the
1902 options set by Configure in the top level Makefile, and Configure
1903 was not even able to write more than one option correctly.
1904 Fixed, now "no-idea no-rc5 -DCRYPTO_MDEBUG" etc. works as intended.
1907 *) New functions CONF_load_bio() and CONF_load_fp() to allow a config
1908 file to be loaded from a BIO or FILE pointer. The BIO version will
1909 for example allow memory BIOs to contain config info.
1912 *) New function "CRYPTO_num_locks" that returns CRYPTO_NUM_LOCKS.
1913 Whoever hopes to achieve shared-library compatibility across versions
1914 must use this, not the compile-time macro.
1915 (Exercise 0.9.4: Which is the minimum library version required by
1917 Note: All this applies only to multi-threaded programs, others don't
1921 *) Add missing case to s3_clnt.c state machine -- one of the new SSL tests
1922 through a BIO pair triggered the default case, i.e.
1923 SSLerr(...,SSL_R_UNKNOWN_STATE).
1926 *) New "BIO pair" concept (crypto/bio/bss_bio.c) so that applications
1927 can use the SSL library even if none of the specific BIOs is
1931 *) Fix a bug in i2d_DSAPublicKey() which meant it returned the wrong value
1932 for the encoded length.
1933 [Jeon KyoungHo <khjeon@sds.samsung.co.kr>]
1935 *) Add initial documentation of the X509V3 functions.
1938 *) Add a new pair of functions PEM_write_PKCS8PrivateKey() and
1939 PEM_write_bio_PKCS8PrivateKey() that are equivalent to
1940 PEM_write_PrivateKey() and PEM_write_bio_PrivateKey() but use the more
1941 secure PKCS#8 private key format with a high iteration count.
1944 *) Fix determination of Perl interpreter: A perl or perl5
1945 _directory_ in $PATH was also accepted as the interpreter.
1946 [Ralf S. Engelschall]
1948 *) Fix demos/sign/sign.c: well there wasn't anything strictly speaking
1949 wrong with it but it was very old and did things like calling
1950 PEM_ASN1_read() directly and used MD5 for the hash not to mention some
1954 *) Fix demos/selfsign.c: it used obsolete and deleted functions, changed
1955 to use the new extension code.
1958 *) Implement the PEM_read/PEM_write functions in crypto/pem/pem_all.c
1959 with macros. This should make it easier to change their form, add extra
1960 arguments etc. Fix a few PEM prototypes which didn't have cipher as a
1964 *) Add to configuration table a new entry that can specify an alternative
1965 name for unistd.h (for pre-POSIX systems); we need this for NeXTstep,
1966 according to Mark Crispin <MRC@Panda.COM>.
1970 *) DES CBC did not update the IV. Weird.
1973 des_cbc_encrypt does not update the IV, but des_ncbc_encrypt does.
1974 Changing the behaviour of the former might break existing programs --
1975 where IV updating is needed, des_ncbc_encrypt can be used.
1978 *) When bntest is run from "make test" it drives bc to check its
1979 calculations, as well as internally checking them. If an internal check
1980 fails, it needs to cause bc to give a non-zero result or make test carries
1981 on without noticing the failure. Fixed.
1984 *) DES library cleanups.
1987 *) Add support for PKCS#5 v2.0 PBE algorithms. This will permit PKCS#8 to be
1988 used with any cipher unlike PKCS#5 v1.5 which can at most handle 64 bit
1989 ciphers. NOTE: although the key derivation function has been verified
1990 against some published test vectors it has not been extensively tested
1991 yet. Added a -v2 "cipher" option to pkcs8 application to allow the use
1995 *) Instead of "mkdir -p", which is not fully portable, use new
1996 Perl script "util/mkdir-p.pl".
1999 *) Rewrite the way password based encryption (PBE) is handled. It used to
2000 assume that the ASN1 AlgorithmIdentifier parameter was a PBEParameter
2001 structure. This was true for the PKCS#5 v1.5 and PKCS#12 PBE algorithms
2002 but doesn't apply to PKCS#5 v2.0 where it can be something else. Now
2003 the 'parameter' field of the AlgorithmIdentifier is passed to the
2004 underlying key generation function so it must do its own ASN1 parsing.
2005 This has also changed the EVP_PBE_CipherInit() function which now has a
2006 'parameter' argument instead of literal salt and iteration count values
2007 and the function EVP_PBE_ALGOR_CipherInit() has been deleted.
2010 *) Support for PKCS#5 v1.5 compatible password based encryption algorithms
2011 and PKCS#8 functionality. New 'pkcs8' application linked to openssl.
2012 Needed to change the PEM_STRING_EVP_PKEY value which was just "PRIVATE
2013 KEY" because this clashed with PKCS#8 unencrypted string. Since this
2014 value was just used as a "magic string" and not used directly its
2015 value doesn't matter.
2018 *) Introduce some semblance of const correctness to BN. Shame C doesn't
2022 *) "linux-sparc64" configuration (ultrapenguin).
2023 [Ray Miller <ray.miller@oucs.ox.ac.uk>]
2024 "linux-sparc" configuration.
2025 [Christian Forster <fo@hawo.stw.uni-erlangen.de>]
2027 *) config now generates no-xxx options for missing ciphers.
2030 *) Support the EBCDIC character set (work in progress).
2031 File ebcdic.c not yet included because it has a different license.
2032 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
2034 *) Support BS2000/OSD-POSIX.
2035 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>]
2037 *) Make callbacks for key generation use void * instead of char *.
2040 *) Make S/MIME samples compile (not yet tested).
2043 *) Additional typesafe stacks.
2046 *) New configuration variants "bsdi-elf-gcc" (BSD/OS 4.x).
2050 Changes between 0.9.3 and 0.9.3a [29 May 1999]
2052 *) New configuration variant "sco5-gcc".
2054 *) Updated some demos.
2055 [Sean O Riordain, Wade Scholine]
2057 *) Add missing BIO_free at exit of pkcs12 application.
2060 *) Fix memory leak in conf.c.
2063 *) Updates for Win32 to assembler version of MD5.
2066 *) Set #! path to perl in apps/der_chop to where we found it
2067 instead of using a fixed path.
2070 *) SHA library changes for irix64-mips4-cc.
2073 *) Improvements for VMS support.
2077 Changes between 0.9.2b and 0.9.3 [24 May 1999]
2079 *) Bignum library bug fix. IRIX 6 passes "make test" now!
2080 This also avoids the problems with SC4.2 and unpatched SC5.
2081 [Andy Polyakov <appro@fy.chalmers.se>]
2083 *) New functions sk_num, sk_value and sk_set to replace the previous macros.
2084 These are required because of the typesafe stack would otherwise break
2085 existing code. If old code used a structure member which used to be STACK
2086 and is now STACK_OF (for example cert in a PKCS7_SIGNED structure) with
2087 sk_num or sk_value it would produce an error because the num, data members
2088 are not present in STACK_OF. Now it just produces a warning. sk_set
2089 replaces the old method of assigning a value to sk_value
2090 (e.g. sk_value(x, i) = y) which the library used in a few cases. Any code
2091 that does this will no longer work (and should use sk_set instead) but
2092 this could be regarded as a "questionable" behaviour anyway.
2095 *) Fix most of the other PKCS#7 bugs. The "experimental" code can now
2096 correctly handle encrypted S/MIME data.
2099 *) Change type of various DES function arguments from des_cblock
2100 (which means, in function argument declarations, pointer to char)
2101 to des_cblock * (meaning pointer to array with 8 char elements),
2102 which allows the compiler to do more typechecking; it was like
2103 that back in SSLeay, but with lots of ugly casts.
2105 Introduce new type const_des_cblock.
2108 *) Reorganise the PKCS#7 library and get rid of some of the more obvious
2109 problems: find RecipientInfo structure that matches recipient certificate
2110 and initialise the ASN1 structures properly based on passed cipher.
2113 *) Belatedly make the BN tests actually check the results.
2116 *) Fix the encoding and decoding of negative ASN1 INTEGERS and conversion
2117 to and from BNs: it was completely broken. New compilation option
2118 NEG_PUBKEY_BUG to allow for some broken certificates that encode public
2119 key elements as negative integers.
2122 *) Reorganize and speed up MD5.
2123 [Andy Polyakov <appro@fy.chalmers.se>]
2126 [Richard Levitte <richard@levitte.org>]
2128 *) New option -out to asn1parse to allow the parsed structure to be
2129 output to a file. This is most useful when combined with the -strparse
2130 option to examine the output of things like OCTET STRINGS.
2133 *) Make SSL library a little more fool-proof by not requiring any longer
2134 that SSL_set_{accept,connect}_state be called before
2135 SSL_{accept,connect} may be used (SSL_set_..._state is omitted
2136 in many applications because usually everything *appeared* to work as
2137 intended anyway -- now it really works as intended).
2140 *) Move openssl.cnf out of lib/.
2143 *) Fix various things to let OpenSSL even pass ``egcc -pipe -O2 -Wall
2144 -Wshadow -Wpointer-arith -Wcast-align -Wmissing-prototypes
2145 -Wmissing-declarations -Wnested-externs -Winline'' with EGCS 1.1.2+
2146 [Ralf S. Engelschall]
2148 *) Various fixes to the EVP and PKCS#7 code. It may now be able to
2149 handle PKCS#7 enveloped data properly.
2150 [Sebastian Akerman <sak@parallelconsulting.com>, modified by Steve]
2152 *) Create a duplicate of the SSL_CTX's CERT in SSL_new instead of
2153 copying pointers. The cert_st handling is changed by this in
2154 various ways (and thus what used to be known as ctx->default_cert
2155 is now called ctx->cert, since we don't resort to s->ctx->[default_]cert
2156 any longer when s->cert does not give us what we need).
2157 ssl_cert_instantiate becomes obsolete by this change.
2158 As soon as we've got the new code right (possibly it already is?),
2159 we have solved a couple of bugs of the earlier code where s->cert
2160 was used as if it could not have been shared with other SSL structures.
2162 Note that using the SSL API in certain dirty ways now will result
2163 in different behaviour than observed with earlier library versions:
2164 Changing settings for an SSL_CTX *ctx after having done s = SSL_new(ctx)
2165 does not influence s as it used to.
2167 In order to clean up things more thoroughly, inside SSL_SESSION
2168 we don't use CERT any longer, but a new structure SESS_CERT
2169 that holds per-session data (if available); currently, this is
2170 the peer's certificate chain and, for clients, the server's certificate
2171 and temporary key. CERT holds only those values that can have
2172 meaningful defaults in an SSL_CTX.
2175 *) New function X509V3_EXT_i2d() to create an X509_EXTENSION structure
2176 from the internal representation. Various PKCS#7 fixes: remove some
2177 evil casts and set the enc_dig_alg field properly based on the signing
2181 *) Allow PKCS#12 password to be set from the command line or the
2182 environment. Let 'ca' get its config file name from the environment
2183 variables "OPENSSL_CONF" or "SSLEAY_CONF" (for consistency with 'req'
2187 *) Allow certificate policies extension to use an IA5STRING for the
2188 organization field. This is contrary to the PKIX definition but
2189 VeriSign uses it and IE5 only recognises this form. Document 'x509'
2193 *) Add PEDANTIC compiler flag to allow compilation with gcc -pedantic,
2194 without disallowing inline assembler and the like for non-pedantic builds.
2197 *) Support Borland C++ builder.
2198 [Janez Jere <jj@void.si>, modified by Ulf Möller]
2203 *) SHA-1 cleanups and performance enhancements.
2204 [Andy Polyakov <appro@fy.chalmers.se>]
2206 *) Sparc v8plus assembler for the bignum library.
2207 [Andy Polyakov <appro@fy.chalmers.se>]
2209 *) Accept any -xxx and +xxx compiler options in Configure.
2212 *) Update HPUX configuration.
2215 *) Add missing sk_<type>_unshift() function to safestack.h
2216 [Ralf S. Engelschall]
2218 *) New function SSL_CTX_use_certificate_chain_file that sets the
2219 "extra_cert"s in addition to the certificate. (This makes sense
2220 only for "PEM" format files, as chains as a whole are not
2224 *) Support verify_depth from the SSL API.
2225 x509_vfy.c had what can be considered an off-by-one-error:
2226 Its depth (which was not part of the external interface)
2227 was actually counting the number of certificates in a chain;
2228 now it really counts the depth.
2231 *) Bugfix in crypto/x509/x509_cmp.c: The SSLerr macro was used
2232 instead of X509err, which often resulted in confusing error
2233 messages since the error codes are not globally unique
2234 (e.g. an alleged error in ssl3_accept when a certificate
2235 didn't match the private key).
2237 *) New function SSL_CTX_set_session_id_context that allows to set a default
2238 value (so that you don't need SSL_set_session_id_context for each
2239 connection using the SSL_CTX).
2242 *) OAEP decoding bug fix.
2245 *) Support INSTALL_PREFIX for package builders, as proposed by
2249 *) New Configure options "threads" and "no-threads". For systems
2250 where the proper compiler options are known (currently Solaris
2251 and Linux), "threads" is the default.
2254 *) New script util/mklink.pl as a faster substitute for util/mklink.sh.
2257 *) Install various scripts to $(OPENSSLDIR)/misc, not to
2258 $(INSTALLTOP)/bin -- they shouldn't clutter directories
2259 such as /usr/local/bin.
2262 *) "make linux-shared" to build shared libraries.
2263 [Niels Poppe <niels@netbox.org>]
2265 *) New Configure option no-<cipher> (rsa, idea, rc5, ...).
2268 *) Add the PKCS#12 API documentation to openssl.txt. Preliminary support for
2269 extension adding in x509 utility.
2272 *) Remove NOPROTO sections and error code comments.
2275 *) Partial rewrite of the DEF file generator to now parse the ANSI
2279 *) New Configure options --prefix=DIR and --openssldir=DIR.
2282 *) Complete rewrite of the error code script(s). It is all now handled
2283 by one script at the top level which handles error code gathering,
2284 header rewriting and C source file generation. It should be much better
2285 than the old method: it now uses a modified version of Ulf's parser to
2286 read the ANSI prototypes in all header files (thus the old K&R definitions
2287 aren't needed for error creation any more) and do a better job of
2288 translating function codes into names. The old 'ASN1 error code imbedded
2289 in a comment' is no longer necessary and it doesn't use .err files which
2290 have now been deleted. Also the error code call doesn't have to appear all
2291 on one line (which resulted in some large lines...).
2294 *) Change #include filenames from <foo.h> to <openssl/foo.h>.
2297 *) Change behaviour of ssl2_read when facing length-0 packets: Don't return
2298 0 (which usually indicates a closed connection), but continue reading.
2301 *) Fix some race conditions.
2304 *) Add support for CRL distribution points extension. Add Certificate
2305 Policies and CRL distribution points documentation.
2308 *) Move the autogenerated header file parts to crypto/opensslconf.h.
2311 *) Fix new 56-bit DES export ciphersuites: they were using 7 bytes instead of
2312 8 of keying material. Merlin has also confirmed interop with this fix
2313 between OpenSSL and Baltimore C/SSL 2.0 and J/SSL 2.0.
2314 [Merlin Hughes <merlin@baltimore.ie>]
2316 *) Fix lots of warnings.
2317 [Richard Levitte <levitte@stacken.kth.se>]
2319 *) In add_cert_dir() in crypto/x509/by_dir.c, break out of the loop if
2320 the directory spec didn't end with a LIST_SEPARATOR_CHAR.
2321 [Richard Levitte <levitte@stacken.kth.se>]
2323 *) Fix problems with sizeof(long) == 8.
2324 [Andy Polyakov <appro@fy.chalmers.se>]
2326 *) Change functions to ANSI C.
2329 *) Fix typos in error codes.
2330 [Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>, Ulf Möller]
2332 *) Remove defunct assembler files from Configure.
2335 *) SPARC v8 assembler BIGNUM implementation.
2336 [Andy Polyakov <appro@fy.chalmers.se>]
2338 *) Support for Certificate Policies extension: both print and set.
2339 Various additions to support the r2i method this uses.
2342 *) A lot of constification, and fix a bug in X509_NAME_oneline() that could
2343 return a const string when you are expecting an allocated buffer.
2346 *) Add support for ASN1 types UTF8String and VISIBLESTRING, also the CHOICE
2347 types DirectoryString and DisplayText.
2350 *) Add code to allow r2i extensions to access the configuration database,
2351 add an LHASH database driver and add several ctx helper functions.
2354 *) Fix an evil bug in bn_expand2() which caused various BN functions to
2355 fail when they extended the size of a BIGNUM.
2358 *) Various utility functions to handle SXNet extension. Modify mkdef.pl to
2359 support typesafe stack.
2362 *) Fix typo in SSL_[gs]et_options().
2363 [Nils Frostberg <nils@medcom.se>]
2365 *) Delete various functions and files that belonged to the (now obsolete)
2366 old X509V3 handling code.
2369 *) New Configure option "rsaref".
2372 *) Don't auto-generate pem.h.
2375 *) Introduce type-safe ASN.1 SETs.
2378 *) Convert various additional casted stacks to type-safe STACK_OF() variants.
2379 [Ben Laurie, Ralf S. Engelschall, Steve Henson]
2381 *) Introduce type-safe STACKs. This will almost certainly break lots of code
2382 that links with OpenSSL (well at least cause lots of warnings), but fear
2383 not: the conversion is trivial, and it eliminates loads of evil casts. A
2384 few STACKed things have been converted already. Feel free to convert more.
2385 In the fullness of time, I'll do away with the STACK type altogether.
2388 *) Add `openssl ca -revoke <certfile>' facility which revokes a certificate
2389 specified in <certfile> by updating the entry in the index.txt file.
2390 This way one no longer has to edit the index.txt file manually for
2391 revoking a certificate. The -revoke option does the gory details now.
2392 [Massimiliano Pala <madwolf@openca.org>, Ralf S. Engelschall]
2394 *) Fix `openssl crl -noout -text' combination where `-noout' killed the
2395 `-text' option at all and this way the `-noout -text' combination was
2396 inconsistent in `openssl crl' with the friends in `openssl x509|rsa|dsa'.
2397 [Ralf S. Engelschall]
2399 *) Make sure a corresponding plain text error message exists for the
2400 X509_V_ERR_CERT_REVOKED/23 error number which can occur when a
2401 verify callback function determined that a certificate was revoked.
2402 [Ralf S. Engelschall]
2404 *) Bugfix: In test/testenc, don't test "openssl <cipher>" for
2405 ciphers that were excluded, e.g. by -DNO_IDEA. Also, test
2406 all available cipers including rc5, which was forgotten until now.
2407 In order to let the testing shell script know which algorithms
2408 are available, a new (up to now undocumented) command
2409 "openssl list-cipher-commands" is used.
2412 *) Bugfix: s_client occasionally would sleep in select() when
2413 it should have checked SSL_pending() first.
2416 *) New functions DSA_do_sign and DSA_do_verify to provide access to
2417 the raw DSA values prior to ASN.1 encoding.
2420 *) Tweaks to Configure
2421 [Niels Poppe <niels@netbox.org>]
2423 *) Add support for PKCS#5 v2.0 ASN1 PBES2 structures. No other support,
2427 *) New variables $(RANLIB) and $(PERL) in the Makefiles.
2430 *) New config option to avoid instructions that are illegal on the 80386.
2431 The default code is faster, but requires at least a 486.
2434 *) Got rid of old SSL2_CLIENT_VERSION (inconsistently used) and
2435 SSL2_SERVER_VERSION (not used at all) macros, which are now the
2436 same as SSL2_VERSION anyway.
2439 *) New "-showcerts" option for s_client.
2442 *) Still more PKCS#12 integration. Add pkcs12 application to openssl
2443 application. Various cleanups and fixes.
2446 *) More PKCS#12 integration. Add new pkcs12 directory with Makefile.ssl and
2447 modify error routines to work internally. Add error codes and PBE init
2448 to library startup routines.
2451 *) Further PKCS#12 integration. Added password based encryption, PKCS#8 and
2452 packing functions to asn1 and evp. Changed function names and error
2453 codes along the way.
2456 *) PKCS12 integration: and so it begins... First of several patches to
2457 slowly integrate PKCS#12 functionality into OpenSSL. Add PKCS#12
2458 objects to objects.h
2461 *) Add a new 'indent' option to some X509V3 extension code. Initial ASN1
2462 and display support for Thawte strong extranet extension.
2465 *) Add LinuxPPC support.
2466 [Jeff Dubrule <igor@pobox.org>]
2468 *) Get rid of redundant BN file bn_mulw.c, and rename bn_div64 to
2469 bn_div_words in alpha.s.
2470 [Hannes Reinecke <H.Reinecke@hw.ac.uk> and Ben Laurie]
2472 *) Make sure the RSA OAEP test is skipped under -DRSAref because
2473 OAEP isn't supported when OpenSSL is built with RSAref.
2474 [Ulf Moeller <ulf@fitug.de>]
2476 *) Move definitions of IS_SET/IS_SEQUENCE inside crypto/asn1/asn1.h
2477 so they no longer are missing under -DNOPROTO.
2478 [Soren S. Jorvang <soren@t.dk>]
2481 Changes between 0.9.1c and 0.9.2b [22 Mar 1999]
2483 *) Make SSL_get_peer_cert_chain() work in servers. Unfortunately, it still
2484 doesn't work when the session is reused. Coming soon!
2487 *) Fix a security hole, that allows sessions to be reused in the wrong
2488 context thus bypassing client cert protection! All software that uses
2489 client certs and session caches in multiple contexts NEEDS PATCHING to
2490 allow session reuse! A fuller solution is in the works.
2491 [Ben Laurie, problem pointed out by Holger Reif, Bodo Moeller (and ???)]
2493 *) Some more source tree cleanups (removed obsolete files
2494 crypto/bf/asm/bf586.pl, test/test.txt and crypto/sha/asm/f.s; changed
2495 permission on "config" script to be executable) and a fix for the INSTALL
2497 [Ulf Moeller <ulf@fitug.de>]
2499 *) Remove some legacy and erroneous uses of malloc, free instead of
2501 [Lennart Bang <lob@netstream.se>, with minor changes by Steve]
2503 *) Make rsa_oaep_test return non-zero on error.
2504 [Ulf Moeller <ulf@fitug.de>]
2506 *) Add support for native Solaris shared libraries. Configure
2507 solaris-sparc-sc4-pic, make, then run shlib/solaris-sc4.sh. It'd be nice
2508 if someone would make that last step automatic.
2509 [Matthias Loepfe <Matthias.Loepfe@AdNovum.CH>]
2511 *) ctx_size was not built with the right compiler during "make links". Fixed.
2514 *) Change the meaning of 'ALL' in the cipher list. It now means "everything
2515 except NULL ciphers". This means the default cipher list will no longer
2516 enable NULL ciphers. They need to be specifically enabled e.g. with
2517 the string "DEFAULT:eNULL".
2520 *) Fix to RSA private encryption routines: if p < q then it would
2521 occasionally produce an invalid result. This will only happen with
2522 externally generated keys because OpenSSL (and SSLeay) ensure p > q.
2525 *) Be less restrictive and allow also `perl util/perlpath.pl
2526 /path/to/bin/perl' in addition to `perl util/perlpath.pl /path/to/bin',
2527 because this way one can also use an interpreter named `perl5' (which is
2528 usually the name of Perl 5.xxx on platforms where an Perl 4.x is still
2529 installed as `perl').
2530 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2532 *) Let util/clean-depend.pl work also with older Perl 5.00x versions.
2533 [Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2535 *) Fix Makefile.org so CC,CFLAG etc are passed to 'make links' add
2536 advapi32.lib to Win32 build and change the pem test comparision
2537 to fc.exe (thanks to Ulrich Kroener <kroneru@yahoo.com> for the
2538 suggestion). Fix misplaced ASNI prototypes and declarations in evp.h
2539 and crypto/des/ede_cbcm_enc.c.
2542 *) DES quad checksum was broken on big-endian architectures. Fixed.
2545 *) Comment out two functions in bio.h that aren't implemented. Fix up the
2546 Win32 test batch file so it (might) work again. The Win32 test batch file
2547 is horrible: I feel ill....
2550 *) Move various #ifdefs around so NO_SYSLOG, NO_DIRENT etc are now selected
2551 in e_os.h. Audit of header files to check ANSI and non ANSI
2552 sections: 10 functions were absent from non ANSI section and not exported
2553 from Windows DLLs. Fixed up libeay.num for new functions.
2556 *) Make `openssl version' output lines consistent.
2557 [Ralf S. Engelschall]
2559 *) Fix Win32 symbol export lists for BIO functions: Added
2560 BIO_get_ex_new_index, BIO_get_ex_num, BIO_get_ex_data and BIO_set_ex_data
2561 to ms/libeay{16,32}.def.
2562 [Ralf S. Engelschall]
2564 *) Second round of fixing the OpenSSL perl/ stuff. It now at least compiled
2565 fine under Unix and passes some trivial tests I've now added. But the
2566 whole stuff is horribly incomplete, so a README.1ST with a disclaimer was
2567 added to make sure no one expects that this stuff really works in the
2568 OpenSSL 0.9.2 release. Additionally I've started to clean the XS sources
2569 up and fixed a few little bugs and inconsistencies in OpenSSL.{pm,xs} and
2571 [Ralf S. Engelschall]
2573 *) Fix the generation of two part addresses in perl.
2574 [Kenji Miyake <kenji@miyake.org>, integrated by Ben Laurie]
2576 *) Add config entry for Linux on MIPS.
2577 [John Tobey <jtobey@channel1.com>]
2579 *) Make links whenever Configure is run, unless we are on Windoze.
2582 *) Permit extensions to be added to CRLs using crl_section in openssl.cnf.
2583 Currently only issuerAltName and AuthorityKeyIdentifier make any sense
2587 *) Add a useful kludge to allow package maintainers to specify compiler and
2588 other platforms details on the command line without having to patch the
2589 Configure script everytime: One now can use ``perl Configure
2590 <id>:<details>'', i.e. platform ids are allowed to have details appended
2591 to them (seperated by colons). This is treated as there would be a static
2592 pre-configured entry in Configure's %table under key <id> with value
2593 <details> and ``perl Configure <id>'' is called. So, when you want to
2594 perform a quick test-compile under FreeBSD 3.1 with pgcc and without
2595 assembler stuff you can use ``perl Configure "FreeBSD-elf:pgcc:-O6:::"''
2596 now, which overrides the FreeBSD-elf entry on-the-fly.
2597 [Ralf S. Engelschall]
2599 *) Disable new TLS1 ciphersuites by default: they aren't official yet.
2602 *) Allow DSO flags like -fpic, -fPIC, -KPIC etc. to be specified
2603 on the `perl Configure ...' command line. This way one can compile
2604 OpenSSL libraries with Position Independent Code (PIC) which is needed
2605 for linking it into DSOs.
2606 [Ralf S. Engelschall]
2608 *) Remarkably, export ciphers were totally broken and no-one had noticed!
2612 *) Cleaned up the LICENSE document: The official contact for any license
2613 questions now is the OpenSSL core team under openssl-core@openssl.org.
2614 And add a paragraph about the dual-license situation to make sure people
2615 recognize that _BOTH_ the OpenSSL license _AND_ the SSLeay license apply
2616 to the OpenSSL toolkit.
2617 [Ralf S. Engelschall]
2619 *) General source tree makefile cleanups: Made `making xxx in yyy...'
2620 display consistent in the source tree and replaced `/bin/rm' by `rm'.
2621 Additonally cleaned up the `make links' target: Remove unnecessary
2622 semicolons, subsequent redundant removes, inline point.sh into mklink.sh
2623 to speed processing and no longer clutter the display with confusing
2624 stuff. Instead only the actually done links are displayed.
2625 [Ralf S. Engelschall]
2627 *) Permit null encryption ciphersuites, used for authentication only. It used
2628 to be necessary to set the preprocessor define SSL_ALLOW_ENULL to do this.
2629 It is now necessary to set SSL_FORBID_ENULL to prevent the use of null
2633 *) Add a bunch of fixes to the PKCS#7 stuff. It used to sometimes reorder
2634 signed attributes when verifying signatures (this would break them),
2635 the detached data encoding was wrong and public keys obtained using
2636 X509_get_pubkey() weren't freed.
2639 *) Add text documentation for the BUFFER functions. Also added a work around
2640 to a Win95 console bug. This was triggered by the password read stuff: the
2641 last character typed gets carried over to the next fread(). If you were
2642 generating a new cert request using 'req' for example then the last
2643 character of the passphrase would be CR which would then enter the first
2647 *) Added the new `Includes OpenSSL Cryptography Software' button as
2648 doc/openssl_button.{gif,html} which is similar in style to the old SSLeay
2649 button and can be used by applications based on OpenSSL to show the
2650 relationship to the OpenSSL project.
2651 [Ralf S. Engelschall]
2653 *) Remove confusing variables in function signatures in files
2654 ssl/ssl_lib.c and ssl/ssl.h.
2655 [Lennart Bong <lob@kulthea.stacken.kth.se>]
2657 *) Don't install bss_file.c under PREFIX/include/
2658 [Lennart Bong <lob@kulthea.stacken.kth.se>]
2660 *) Get the Win32 compile working again. Modify mkdef.pl so it can handle
2661 functions that return function pointers and has support for NT specific
2662 stuff. Fix mk1mf.pl and VC-32.pl to support NT differences also. Various
2663 #ifdef WIN32 and WINNTs sprinkled about the place and some changes from
2664 unsigned to signed types: this was killing the Win32 compile.
2667 *) Add new certificate file to stack functions,
2668 SSL_add_dir_cert_subjects_to_stack() and
2669 SSL_add_file_cert_subjects_to_stack(). These largely supplant
2670 SSL_load_client_CA_file(), and can be used to add multiple certs easily
2671 to a stack (usually this is then handed to SSL_CTX_set_client_CA_list()).
2672 This means that Apache-SSL and similar packages don't have to mess around
2673 to add as many CAs as they want to the preferred list.
2676 *) Experiment with doxygen documentation. Currently only partially applied to
2678 See http://www.stack.nl/~dimitri/doxygen/index.html, and run doxygen with
2679 openssl.doxy as the configuration file.
2682 *) Get rid of remaining C++-style comments which strict C compilers hate.
2683 [Ralf S. Engelschall, pointed out by Carlos Amengual]
2685 *) Changed BN_RECURSION in bn_mont.c to BN_RECURSION_MONT so it is not
2686 compiled in by default: it has problems with large keys.
2689 *) Add a bunch of SSL_xxx() functions for configuring the temporary RSA and
2690 DH private keys and/or callback functions which directly correspond to
2691 their SSL_CTX_xxx() counterparts but work on a per-connection basis. This
2692 is needed for applications which have to configure certificates on a
2693 per-connection basis (e.g. Apache+mod_ssl) instead of a per-context basis
2695 For the RSA certificate situation is makes no difference, but
2696 for the DSA certificate situation this fixes the "no shared cipher"
2697 problem where the OpenSSL cipher selection procedure failed because the
2698 temporary keys were not overtaken from the context and the API provided
2699 no way to reconfigure them.
2700 The new functions now let applications reconfigure the stuff and they
2701 are in detail: SSL_need_tmp_RSA, SSL_set_tmp_rsa, SSL_set_tmp_dh,
2702 SSL_set_tmp_rsa_callback and SSL_set_tmp_dh_callback. Additionally a new
2703 non-public-API function ssl_cert_instantiate() is used as a helper
2704 function and also to reduce code redundancy inside ssl_rsa.c.
2705 [Ralf S. Engelschall]
2707 *) Move s_server -dcert and -dkey options out of the undocumented feature
2708 area because they are useful for the DSA situation and should be
2709 recognized by the users.
2710 [Ralf S. Engelschall]
2712 *) Fix the cipher decision scheme for export ciphers: the export bits are
2713 *not* within SSL_MKEY_MASK or SSL_AUTH_MASK, they are within
2714 SSL_EXP_MASK. So, the original variable has to be used instead of the
2715 already masked variable.
2716 [Richard Levitte <levitte@stacken.kth.se>]
2718 *) Fix 'port' variable from `int' to `unsigned int' in crypto/bio/b_sock.c
2719 [Richard Levitte <levitte@stacken.kth.se>]
2721 *) Change type of another md_len variable in pk7_doit.c:PKCS7_dataFinal()
2722 from `int' to `unsigned int' because it's a length and initialized by
2723 EVP_DigestFinal() which expects an `unsigned int *'.
2724 [Richard Levitte <levitte@stacken.kth.se>]
2726 *) Don't hard-code path to Perl interpreter on shebang line of Configure
2727 script. Instead use the usual Shell->Perl transition trick.
2728 [Ralf S. Engelschall]
2730 *) Make `openssl x509 -noout -modulus' functional also for DSA certificates
2731 (in addition to RSA certificates) to match the behaviour of `openssl dsa
2732 -noout -modulus' as it's already the case for `openssl rsa -noout
2733 -modulus'. For RSA the -modulus is the real "modulus" while for DSA
2734 currently the public key is printed (a decision which was already done by
2735 `openssl dsa -modulus' in the past) which serves a similar purpose.
2736 Additionally the NO_RSA no longer completely removes the whole -modulus
2737 option; it now only avoids using the RSA stuff. Same applies to NO_DSA
2739 [Ralf S. Engelschall]
2741 *) Add Arne Ansper's reliable BIO - this is an encrypted, block-digested
2742 BIO. See the source (crypto/evp/bio_ok.c) for more info.
2743 [Arne Ansper <arne@ats.cyber.ee>]
2745 *) Dump the old yucky req code that tried (and failed) to allow raw OIDs
2746 to be added. Now both 'req' and 'ca' can use new objects defined in the
2750 *) Add cool BIO that does syslog (or event log on NT).
2751 [Arne Ansper <arne@ats.cyber.ee>, integrated by Ben Laurie]
2753 *) Add support for new TLS ciphersuites, TLS_RSA_EXPORT56_WITH_RC4_56_MD5,
2754 TLS_RSA_EXPORT56_WITH_RC2_CBC_56_MD5 and
2755 TLS_RSA_EXPORT56_WITH_DES_CBC_SHA, as specified in "56-bit Export Cipher
2756 Suites For TLS", draft-ietf-tls-56-bit-ciphersuites-00.txt.
2759 *) Add preliminary config info for new extension code.
2762 *) Make RSA_NO_PADDING really use no padding.
2763 [Ulf Moeller <ulf@fitug.de>]
2765 *) Generate errors when private/public key check is done.
2768 *) Overhaul for 'crl' utility. New function X509_CRL_print. Partial support
2769 for some CRL extensions and new objects added.
2772 *) Really fix the ASN1 IMPLICIT bug this time... Partial support for private
2773 key usage extension and fuller support for authority key id.
2776 *) Add OAEP encryption for the OpenSSL crypto library. OAEP is the improved
2777 padding method for RSA, which is recommended for new applications in PKCS
2778 #1 v2.0 (RFC 2437, October 1998).
2779 OAEP (Optimal Asymmetric Encryption Padding) has better theoretical
2780 foundations than the ad-hoc padding used in PKCS #1 v1.5. It is secure
2781 against Bleichbacher's attack on RSA.
2782 [Ulf Moeller <ulf@fitug.de>, reformatted, corrected and integrated by
2785 *) Updates to the new SSL compression code
2786 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2788 *) Fix so that the version number in the master secret, when passed
2789 via RSA, checks that if TLS was proposed, but we roll back to SSLv3
2790 (because the server will not accept higher), that the version number
2791 is 0x03,0x01, not 0x03,0x00
2792 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2794 *) Run extensive memory leak checks on SSL apps. Fixed *lots* of memory
2795 leaks in ssl/ relating to new X509_get_pubkey() behaviour. Also fixes
2796 in apps/ and an unrelated leak in crypto/dsa/dsa_vrf.c
2799 *) Support for RAW extensions where an arbitrary extension can be
2800 created by including its DER encoding. See apps/openssl.cnf for
2804 *) Make sure latest Perl versions don't interpret some generated C array
2805 code as Perl array code in the crypto/err/err_genc.pl script.
2806 [Lars Weber <3weber@informatik.uni-hamburg.de>]
2808 *) Modify ms/do_ms.bat to not generate assembly language makefiles since
2809 not many people have the assembler. Various Win32 compilation fixes and
2810 update to the INSTALL.W32 file with (hopefully) more accurate Win32
2814 *) Modify configure script 'Configure' to automatically create crypto/date.h
2815 file under Win32 and also build pem.h from pem.org. New script
2816 util/mkfiles.pl to create the MINFO file on environments that can't do a
2817 'make files': perl util/mkfiles.pl >MINFO should work.
2820 *) Major rework of DES function declarations, in the pursuit of correctness
2821 and purity. As a result, many evil casts evaporated, and some weirdness,
2822 too. You may find this causes warnings in your code. Zapping your evil
2823 casts will probably fix them. Mostly.
2826 *) Fix for a typo in asn1.h. Bug fix to object creation script
2827 obj_dat.pl. It considered a zero in an object definition to mean
2828 "end of object": none of the objects in objects.h have any zeros
2829 so it wasn't spotted.
2830 [Steve Henson, reported by Erwann ABALEA <eabalea@certplus.com>]
2832 *) Add support for Triple DES Cipher Block Chaining with Output Feedback
2833 Masking (CBCM). In the absence of test vectors, the best I have been able
2834 to do is check that the decrypt undoes the encrypt, so far. Send me test
2835 vectors if you have them.
2838 *) Correct calculation of key length for export ciphers (too much space was
2839 allocated for null ciphers). This has not been tested!
2842 *) Modifications to the mkdef.pl for Win32 DEF file creation. The usage
2843 message is now correct (it understands "crypto" and "ssl" on its
2844 command line). There is also now an "update" option. This will update
2845 the util/ssleay.num and util/libeay.num files with any new functions.
2847 perl util/mkdef.pl crypto ssl update
2848 it will update them.
2851 *) Overhauled the Perl interface (perl/*):
2852 - ported BN stuff to OpenSSL's different BN library
2853 - made the perl/ source tree CVS-aware
2854 - renamed the package from SSLeay to OpenSSL (the files still contain
2855 their history because I've copied them in the repository)
2856 - removed obsolete files (the test scripts will be replaced
2857 by better Test::Harness variants in the future)
2858 [Ralf S. Engelschall]
2860 *) First cut for a very conservative source tree cleanup:
2861 1. merge various obsolete readme texts into doc/ssleay.txt
2862 where we collect the old documents and readme texts.
2863 2. remove the first part of files where I'm already sure that we no
2864 longer need them because of three reasons: either they are just temporary
2865 files which were left by Eric or they are preserved original files where
2866 I've verified that the diff is also available in the CVS via "cvs diff
2867 -rSSLeay_0_8_1b" or they were renamed (as it was definitely the case for
2868 the crypto/md/ stuff).
2869 [Ralf S. Engelschall]
2871 *) More extension code. Incomplete support for subject and issuer alt
2872 name, issuer and authority key id. Change the i2v function parameters
2873 and add an extra 'crl' parameter in the X509V3_CTX structure: guess
2874 what that's for :-) Fix to ASN1 macro which messed up
2875 IMPLICIT tag and add f_enum.c which adds a2i, i2a for ENUMERATED.
2878 *) Preliminary support for ENUMERATED type. This is largely copied from the
2882 *) Add new function, EVP_MD_CTX_copy() to replace frequent use of memcpy.
2883 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2885 *) Make sure `make rehash' target really finds the `openssl' program.
2886 [Ralf S. Engelschall, Matthias Loepfe <Matthias.Loepfe@adnovum.ch>]
2888 *) Squeeze another 7% of speed out of MD5 assembler, at least on a P2. I'd
2889 like to hear about it if this slows down other processors.
2892 *) Add CygWin32 platform information to Configure script.
2893 [Alan Batie <batie@aahz.jf.intel.com>]
2895 *) Fixed ms/32all.bat script: `no_asm' -> `no-asm'
2896 [Rainer W. Gerling <gerling@mpg-gv.mpg.de>]
2898 *) New program nseq to manipulate netscape certificate sequences
2901 *) Modify crl2pkcs7 so it supports multiple -certfile arguments. Fix a
2905 *) Fixes to BN code. Previously the default was to define BN_RECURSION
2906 but the BN code had some problems that would cause failures when
2907 doing certificate verification and some other functions.
2908 [Eric A. Young, (from changes to C2Net SSLeay, integrated by Mark Cox)]
2910 *) Add ASN1 and PEM code to support netscape certificate sequences.
2913 *) Add ASN1 and PEM code to support netscape certificate sequences.
2916 *) Add several PKIX and private extended key usage OIDs.
2919 *) Modify the 'ca' program to handle the new extension code. Modify
2920 openssl.cnf for new extension format, add comments.
2923 *) More X509 V3 changes. Fix typo in v3_bitstr.c. Add support to 'req'
2924 and add a sample to openssl.cnf so req -x509 now adds appropriate
2928 *) Continued X509 V3 changes. Add to other makefiles, integrate with the
2929 error code, add initial support to X509_print() and x509 application.
2932 *) Takes a deep breath and start addding X509 V3 extension support code. Add
2933 files in crypto/x509v3. Move original stuff to crypto/x509v3/old. All this
2934 stuff is currently isolated and isn't even compiled yet.
2937 *) Continuing patches for GeneralizedTime. Fix up certificate and CRL
2938 ASN1 to use ASN1_TIME and modify print routines to use ASN1_TIME_print.
2939 Removed the versions check from X509 routines when loading extensions:
2940 this allows certain broken certificates that don't set the version
2941 properly to be processed.
2944 *) Deal with irritating shit to do with dependencies, in YAAHW (Yet Another
2945 Ad Hoc Way) - Makefile.ssls now all contain local dependencies, which
2946 can still be regenerated with "make depend".
2949 *) Spelling mistake in C version of CAST-128.
2950 [Ben Laurie, reported by Jeremy Hylton <jeremy@cnri.reston.va.us>]
2952 *) Changes to the error generation code. The perl script err-code.pl
2953 now reads in the old error codes and retains the old numbers, only
2954 adding new ones if necessary. It also only changes the .err files if new
2955 codes are added. The makefiles have been modified to only insert errors
2956 when needed (to avoid needlessly modifying header files). This is done
2957 by only inserting errors if the .err file is newer than the auto generated
2958 C file. To rebuild all the error codes from scratch (the old behaviour)
2959 either modify crypto/Makefile.ssl to pass the -regen flag to err_code.pl
2960 or delete all the .err files.
2963 *) CAST-128 was incorrectly implemented for short keys. The C version has
2964 been fixed, but is untested. The assembler versions are also fixed, but
2965 new assembler HAS NOT BEEN GENERATED FOR WIN32 - the Makefile needs fixing
2966 to regenerate it if needed.
2967 [Ben Laurie, reported (with fix for C version) by Jun-ichiro itojun
2968 Hagino <itojun@kame.net>]
2970 *) File was opened incorrectly in randfile.c.
2971 [Ulf Möller <ulf@fitug.de>]
2973 *) Beginning of support for GeneralizedTime. d2i, i2d, check and print
2974 functions. Also ASN1_TIME suite which is a CHOICE of UTCTime or
2975 GeneralizedTime. ASN1_TIME is the proper type used in certificates et
2976 al: it's just almost always a UTCTime. Note this patch adds new error
2977 codes so do a "make errors" if there are problems.
2980 *) Correct Linux 1 recognition in config.
2981 [Ulf Möller <ulf@fitug.de>]
2983 *) Remove pointless MD5 hash when using DSA keys in ca.
2984 [Anonymous <nobody@replay.com>]
2986 *) Generate an error if given an empty string as a cert directory. Also
2987 generate an error if handed NULL (previously returned 0 to indicate an
2988 error, but didn't set one).
2989 [Ben Laurie, reported by Anonymous <nobody@replay.com>]
2991 *) Add prototypes to SSL methods. Make SSL_write's buffer const, at last.
2994 *) Fix the dummy function BN_ref_mod_exp() in rsaref.c to have the correct
2995 parameters. This was causing a warning which killed off the Win32 compile.
2998 *) Remove C++ style comments from crypto/bn/bn_local.h.
2999 [Neil Costigan <neil.costigan@celocom.com>]
3001 *) The function OBJ_txt2nid was broken. It was supposed to return a nid
3002 based on a text string, looking up short and long names and finally
3003 "dot" format. The "dot" format stuff didn't work. Added new function
3004 OBJ_txt2obj to do the same but return an ASN1_OBJECT and rewrote
3005 OBJ_txt2nid to use it. OBJ_txt2obj can also return objects even if the
3006 OID is not part of the table.
3009 *) Add prototypes to X509 lookup/verify methods, fixing a bug in
3010 X509_LOOKUP_by_alias().
3013 *) Sort openssl functions by name.
3016 *) Get the gendsa program working (hopefully) and add it to app list. Remove
3017 encryption from sample DSA keys (in case anyone is interested the password
3021 *) Make _all_ *_free functions accept a NULL pointer.
3022 [Frans Heymans <fheymans@isaserver.be>]
3024 *) If a DH key is generated in s3_srvr.c, don't blow it by trying to use
3026 [Anonymous <nobody@replay.com>]
3028 *) s_server should send the CAfile as acceptable CAs, not its own cert.
3029 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
3031 *) Don't blow it for numeric -newkey arguments to apps/req.
3032 [Bodo Moeller <3moeller@informatik.uni-hamburg.de>]
3034 *) Temp key "for export" tests were wrong in s3_srvr.c.
3035 [Anonymous <nobody@replay.com>]
3037 *) Add prototype for temp key callback functions
3038 SSL_CTX_set_tmp_{rsa,dh}_callback().
3041 *) Make DH_free() tolerate being passed a NULL pointer (like RSA_free() and
3042 DSA_free()). Make X509_PUBKEY_set() check for errors in d2i_PublicKey().
3045 *) X509_name_add_entry() freed the wrong thing after an error.
3046 [Arne Ansper <arne@ats.cyber.ee>]
3048 *) rsa_eay.c would attempt to free a NULL context.
3049 [Arne Ansper <arne@ats.cyber.ee>]
3051 *) BIO_s_socket() had a broken should_retry() on Windoze.
3052 [Arne Ansper <arne@ats.cyber.ee>]
3054 *) BIO_f_buffer() didn't pass on BIO_CTRL_FLUSH.
3055 [Arne Ansper <arne@ats.cyber.ee>]
3057 *) Make sure the already existing X509_STORE->depth variable is initialized
3058 in X509_STORE_new(), but document the fact that this variable is still
3059 unused in the certificate verification process.
3060 [Ralf S. Engelschall]
3062 *) Fix the various library and apps files to free up pkeys obtained from
3063 X509_PUBKEY_get() et al. Also allow x509.c to handle netscape extensions.
3066 *) Fix reference counting in X509_PUBKEY_get(). This makes
3067 demos/maurice/example2.c work, amongst others, probably.
3068 [Steve Henson and Ben Laurie]
3070 *) First cut of a cleanup for apps/. First the `ssleay' program is now named
3071 `openssl' and second, the shortcut symlinks for the `openssl <command>'
3072 are no longer created. This way we have a single and consistent command
3073 line interface `openssl <command>', similar to `cvs <command>'.
3074 [Ralf S. Engelschall, Paul Sutton and Ben Laurie]
3076 *) ca.c: move test for DSA keys inside #ifndef NO_DSA. Make pubkey
3077 BIT STRING wrapper always have zero unused bits.
3080 *) Add CA.pl, perl version of CA.sh, add extended key usage OID.
3083 *) Make the top-level INSTALL documentation easier to understand.
3086 *) Makefiles updated to exit if an error occurs in a sub-directory
3087 make (including if user presses ^C) [Paul Sutton]
3089 *) Make Montgomery context stuff explicit in RSA data structure.
3092 *) Fix build order of pem and err to allow for generated pem.h.
3095 *) Fix renumbering bug in X509_NAME_delete_entry().
3098 *) Enhanced the err-ins.pl script so it makes the error library number
3099 global and can add a library name. This is needed for external ASN1 and
3100 other error libraries.
3103 *) Fixed sk_insert which never worked properly.
3106 *) Fix ASN1 macros so they can handle indefinite length construted
3107 EXPLICIT tags. Some non standard certificates use these: they can now
3111 *) Merged the various old/obsolete SSLeay documentation files (doc/xxx.doc)
3112 into a single doc/ssleay.txt bundle. This way the information is still
3113 preserved but no longer messes up this directory. Now it's new room for
3114 the new set of documenation files.
3115 [Ralf S. Engelschall]
3117 *) SETs were incorrectly DER encoded. This was a major pain, because they
3118 shared code with SEQUENCEs, which aren't coded the same. This means that
3119 almost everything to do with SETs or SEQUENCEs has either changed name or
3120 number of arguments.
3121 [Ben Laurie, based on a partial fix by GP Jayan <gp@nsj.co.jp>]
3123 *) Fix test data to work with the above.
3126 *) Fix the RSA header declarations that hid a bug I fixed in 0.9.0b but
3127 was already fixed by Eric for 0.9.1 it seems.
3128 [Ben Laurie - pointed out by Ulf Möller <ulf@fitug.de>]
3130 *) Autodetect FreeBSD3.
3133 *) Fix various bugs in Configure. This affects the following platforms:
3137 unixware-2.0-pentium
3141 *) Eliminate generated files from CVS. Reorder tests to regenerate files
3142 before they are needed.
3145 *) Generate Makefile.ssl from Makefile.org (to keep CVS happy).
3149 Changes between 0.9.1b and 0.9.1c [23-Dec-1998]
3151 *) Added OPENSSL_VERSION_NUMBER to crypto/crypto.h and
3152 changed SSLeay to OpenSSL in version strings.
3153 [Ralf S. Engelschall]
3155 *) Some fixups to the top-level documents.
3158 *) Fixed the nasty bug where rsaref.h was not found under compile-time
3159 because the symlink to include/ was missing.
3160 [Ralf S. Engelschall]
3162 *) Incorporated the popular no-RSA/DSA-only patches
3163 which allow to compile a RSA-free SSLeay.
3164 [Andrew Cooke / Interrader Ldt., Ralf S. Engelschall]
3166 *) Fixed nasty rehash problem under `make -f Makefile.ssl links'
3167 when "ssleay" is still not found.
3168 [Ralf S. Engelschall]
3170 *) Added more platforms to Configure: Cray T3E, HPUX 11,
3171 [Ralf S. Engelschall, Beckmann <beckman@acl.lanl.gov>]
3173 *) Updated the README file.
3174 [Ralf S. Engelschall]
3176 *) Added various .cvsignore files in the CVS repository subdirs
3177 to make a "cvs update" really silent.
3178 [Ralf S. Engelschall]
3180 *) Recompiled the error-definition header files and added
3181 missing symbols to the Win32 linker tables.
3182 [Ralf S. Engelschall]
3184 *) Cleaned up the top-level documents;
3185 o new files: CHANGES and LICENSE
3186 o merged VERSION, HISTORY* and README* files a CHANGES.SSLeay
3187 o merged COPYRIGHT into LICENSE
3188 o removed obsolete TODO file
3189 o renamed MICROSOFT to INSTALL.W32
3190 [Ralf S. Engelschall]
3192 *) Removed dummy files from the 0.9.1b source tree:
3193 crypto/asn1/x crypto/bio/cd crypto/bio/fg crypto/bio/grep crypto/bio/vi
3194 crypto/bn/asm/......add.c crypto/bn/asm/a.out crypto/dsa/f crypto/md5/f
3195 crypto/pem/gmon.out crypto/perlasm/f crypto/pkcs7/build crypto/rsa/f
3196 crypto/sha/asm/f crypto/threads/f ms/zzz ssl/f ssl/f.mak test/f
3197 util/f.mak util/pl/f util/pl/f.mak crypto/bf/bf_locl.old apps/f
3198 [Ralf S. Engelschall]
3200 *) Added various platform portability fixes.
3203 *) The Genesis of the OpenSSL rpject:
3204 We start with the latest (unreleased) SSLeay version 0.9.1b which Eric A.
3205 Young and Tim J. Hudson created while they were working for C2Net until
3207 [The OpenSSL Project]
3210 Changes between 0.9.0b and 0.9.1b [not released]
3212 *) Updated a few CA certificates under certs/
3215 *) Changed some BIGNUM api stuff.
3218 *) Various platform ports: OpenBSD, Ultrix, IRIX 64bit, NetBSD,
3219 DGUX x86, Linux Alpha, etc.
3222 *) New COMP library [crypto/comp/] for SSL Record Layer Compression:
3223 RLE (dummy implemented) and ZLIB (really implemented when ZLIB is
3227 *) Add -strparse option to asn1pars program which parses nested
3229 [Dr Stephen Henson <shenson@bigfoot.com>]
3231 *) Added "oid_file" to ssleay.cnf for "ca" and "req" programs.
3234 *) DSA fix for "ca" program.
3237 *) Added "-genkey" option to "dsaparam" program.
3240 *) Added RIPE MD160 (rmd160) message digest.
3243 *) Added -a (all) option to "ssleay version" command.
3246 *) Added PLATFORM define which is the id given to Configure.
3249 *) Added MemCheck_XXXX functions to crypto/mem.c for memory checking.
3252 *) Extended the ASN.1 parser routines.
3255 *) Extended BIO routines to support REUSEADDR, seek, tell, etc.
3258 *) Added a BN_CTX to the BN library.
3261 *) Fixed the weak key values in DES library
3264 *) Changed API in EVP library for cipher aliases.
3267 *) Added support for RC2/64bit cipher.
3270 *) Converted the lhash library to the crypto/mem.c functions.
3273 *) Added more recognized ASN.1 object ids.
3276 *) Added more RSA padding checks for SSL/TLS.
3279 *) Added BIO proxy/filter functionality.
3282 *) Added extra_certs to SSL_CTX which can be used
3283 send extra CA certificates to the client in the CA cert chain sending
3284 process. It can be configured with SSL_CTX_add_extra_chain_cert().
3287 *) Now Fortezza is denied in the authentication phase because
3288 this is key exchange mechanism is not supported by SSLeay at all.
3291 *) Additional PKCS1 checks.
3294 *) Support the string "TLSv1" for all TLS v1 ciphers.
3297 *) Added function SSL_get_ex_data_X509_STORE_CTX_idx() which gives the
3298 ex_data index of the SSL context in the X509_STORE_CTX ex_data.
3301 *) Fixed a few memory leaks.
3304 *) Fixed various code and comment typos.
3307 *) A minor bug in ssl/s3_clnt.c where there would always be 4 0
3308 bytes sent in the client random.
3309 [Edward Bishop <ebishop@spyglass.com>]